3 @c Copyright (C) 1991-2021 Free Software Foundation, Inc.
6 @include configdoc.texi
7 @c (configdoc.texi is generated by the Makefile)
13 @macro gcctabopt{body}
19 @c Configure for the generation of man pages
47 @dircategory Software development
49 * Ld: (ld). The GNU linker.
54 This file documents the @sc{gnu} linker LD
55 @ifset VERSION_PACKAGE
56 @value{VERSION_PACKAGE}
58 version @value{VERSION}.
60 Copyright @copyright{} 1991-2021 Free Software Foundation, Inc.
62 Permission is granted to copy, distribute and/or modify this document
63 under the terms of the GNU Free Documentation License, Version 1.3
64 or any later version published by the Free Software Foundation;
65 with no Invariant Sections, with no Front-Cover Texts, and with no
66 Back-Cover Texts. A copy of the license is included in the
67 section entitled ``GNU Free Documentation License''.
71 @setchapternewpage odd
72 @settitle The GNU linker
77 @ifset VERSION_PACKAGE
78 @subtitle @value{VERSION_PACKAGE}
80 @subtitle Version @value{VERSION}
81 @author Steve Chamberlain
82 @author Ian Lance Taylor
87 \hfill Red Hat Inc\par
88 \hfill nickc\@credhat.com, doc\@redhat.com\par
89 \hfill {\it The GNU linker}\par
90 \hfill Edited by Jeffrey Osier (jeffrey\@cygnus.com)\par
92 \global\parindent=0pt % Steve likes it this way.
95 @vskip 0pt plus 1filll
96 @c man begin COPYRIGHT
97 Copyright @copyright{} 1991-2021 Free Software Foundation, Inc.
99 Permission is granted to copy, distribute and/or modify this document
100 under the terms of the GNU Free Documentation License, Version 1.3
101 or any later version published by the Free Software Foundation;
102 with no Invariant Sections, with no Front-Cover Texts, and with no
103 Back-Cover Texts. A copy of the license is included in the
104 section entitled ``GNU Free Documentation License''.
110 @c FIXME: Talk about importance of *order* of args, cmds to linker!
115 This file documents the @sc{gnu} linker ld
116 @ifset VERSION_PACKAGE
117 @value{VERSION_PACKAGE}
119 version @value{VERSION}.
121 This document is distributed under the terms of the GNU Free
122 Documentation License version 1.3. A copy of the license is included
123 in the section entitled ``GNU Free Documentation License''.
126 * Overview:: Overview
127 * Invocation:: Invocation
128 * Scripts:: Linker Scripts
129 * Plugins:: Linker Plugins
131 * Machine Dependent:: Machine Dependent Features
135 * H8/300:: ld and the H8/300
138 * Renesas:: ld and other Renesas micros
141 * ARM:: ld and the ARM family
144 * M68HC11/68HC12:: ld and the Motorola 68HC11 and 68HC12 families
147 * HPPA ELF32:: ld and HPPA 32-bit ELF
150 * M68K:: ld and Motorola 68K family
153 * MIPS:: ld and MIPS family
156 * PowerPC ELF32:: ld and PowerPC 32-bit ELF Support
159 * PowerPC64 ELF64:: ld and PowerPC64 64-bit ELF Support
162 * S/390 ELF:: ld and S/390 ELF Support
165 * SPU ELF:: ld and SPU ELF Support
168 * TI COFF:: ld and the TI COFF
171 * Win32:: ld and WIN32 (cygwin/mingw)
174 * Xtensa:: ld and Xtensa Processors
177 @ifclear SingleFormat
180 @c Following blank line required for remaining bug in makeinfo conds/menus
182 * Reporting Bugs:: Reporting Bugs
183 * MRI:: MRI Compatible Script Files
184 * GNU Free Documentation License:: GNU Free Documentation License
185 * LD Index:: LD Index
192 @cindex @sc{gnu} linker
193 @cindex what is this?
196 @c man begin SYNOPSIS
197 ld [@b{options}] @var{objfile} @dots{}
201 ar(1), nm(1), objcopy(1), objdump(1), readelf(1) and
202 the Info entries for @file{binutils} and
207 @c man begin DESCRIPTION
209 @command{ld} combines a number of object and archive files, relocates
210 their data and ties up symbol references. Usually the last step in
211 compiling a program is to run @command{ld}.
213 @command{ld} accepts Linker Command Language files written in
214 a superset of AT&T's Link Editor Command Language syntax,
215 to provide explicit and total control over the linking process.
219 This man page does not describe the command language; see the
220 @command{ld} entry in @code{info} for full details on the command
221 language and on other aspects of the GNU linker.
224 @ifclear SingleFormat
225 This version of @command{ld} uses the general purpose BFD libraries
226 to operate on object files. This allows @command{ld} to read, combine, and
227 write object files in many different formats---for example, COFF or
228 @code{a.out}. Different formats may be linked together to produce any
229 available kind of object file. @xref{BFD}, for more information.
232 Aside from its flexibility, the @sc{gnu} linker is more helpful than other
233 linkers in providing diagnostic information. Many linkers abandon
234 execution immediately upon encountering an error; whenever possible,
235 @command{ld} continues executing, allowing you to identify other errors
236 (or, in some cases, to get an output file in spite of the error).
243 @c man begin DESCRIPTION
245 The @sc{gnu} linker @command{ld} is meant to cover a broad range of situations,
246 and to be as compatible as possible with other linkers. As a result,
247 you have many choices to control its behavior.
253 * Options:: Command-line Options
254 * Environment:: Environment Variables
258 @section Command-line Options
266 The linker supports a plethora of command-line options, but in actual
267 practice few of them are used in any particular context.
268 @cindex standard Unix system
269 For instance, a frequent use of @command{ld} is to link standard Unix
270 object files on a standard, supported Unix system. On such a system, to
271 link a file @code{hello.o}:
274 ld -o @var{output} /lib/crt0.o hello.o -lc
277 This tells @command{ld} to produce a file called @var{output} as the
278 result of linking the file @code{/lib/crt0.o} with @code{hello.o} and
279 the library @code{libc.a}, which will come from the standard search
280 directories. (See the discussion of the @samp{-l} option below.)
282 Some of the command-line options to @command{ld} may be specified at any
283 point in the command line. However, options which refer to files, such
284 as @samp{-l} or @samp{-T}, cause the file to be read at the point at
285 which the option appears in the command line, relative to the object
286 files and other file options. Repeating non-file options with a
287 different argument will either have no further effect, or override prior
288 occurrences (those further to the left on the command line) of that
289 option. Options which may be meaningfully specified more than once are
290 noted in the descriptions below.
293 Non-option arguments are object files or archives which are to be linked
294 together. They may follow, precede, or be mixed in with command-line
295 options, except that an object file argument may not be placed between
296 an option and its argument.
298 Usually the linker is invoked with at least one object file, but you can
299 specify other forms of binary input files using @samp{-l}, @samp{-R},
300 and the script command language. If @emph{no} binary input files at all
301 are specified, the linker does not produce any output, and issues the
302 message @samp{No input files}.
304 If the linker cannot recognize the format of an object file, it will
305 assume that it is a linker script. A script specified in this way
306 augments the main linker script used for the link (either the default
307 linker script or the one specified by using @samp{-T}). This feature
308 permits the linker to link against a file which appears to be an object
309 or an archive, but actually merely defines some symbol values, or uses
310 @code{INPUT} or @code{GROUP} to load other objects. Specifying a
311 script in this way merely augments the main linker script, with the
312 extra commands placed after the main script; use the @samp{-T} option
313 to replace the default linker script entirely, but note the effect of
314 the @code{INSERT} command. @xref{Scripts}.
316 For options whose names are a single letter,
317 option arguments must either follow the option letter without intervening
318 whitespace, or be given as separate arguments immediately following the
319 option that requires them.
321 For options whose names are multiple letters, either one dash or two can
322 precede the option name; for example, @samp{-trace-symbol} and
323 @samp{--trace-symbol} are equivalent. Note---there is one exception to
324 this rule. Multiple letter options that start with a lower case 'o' can
325 only be preceded by two dashes. This is to reduce confusion with the
326 @samp{-o} option. So for example @samp{-omagic} sets the output file
327 name to @samp{magic} whereas @samp{--omagic} sets the NMAGIC flag on the
330 Arguments to multiple-letter options must either be separated from the
331 option name by an equals sign, or be given as separate arguments
332 immediately following the option that requires them. For example,
333 @samp{--trace-symbol foo} and @samp{--trace-symbol=foo} are equivalent.
334 Unique abbreviations of the names of multiple-letter options are
337 Note---if the linker is being invoked indirectly, via a compiler driver
338 (e.g. @samp{gcc}) then all the linker command-line options should be
339 prefixed by @samp{-Wl,} (or whatever is appropriate for the particular
340 compiler driver) like this:
343 gcc -Wl,--start-group foo.o bar.o -Wl,--end-group
346 This is important, because otherwise the compiler driver program may
347 silently drop the linker options, resulting in a bad link. Confusion
348 may also arise when passing options that require values through a
349 driver, as the use of a space between option and argument acts as
350 a separator, and causes the driver to pass only the option to the linker
351 and the argument to the compiler. In this case, it is simplest to use
352 the joined forms of both single- and multiple-letter options, such as:
355 gcc foo.o bar.o -Wl,-eENTRY -Wl,-Map=a.map
358 Here is a table of the generic command-line switches accepted by the GNU
362 @include at-file.texi
364 @kindex -a @var{keyword}
365 @item -a @var{keyword}
366 This option is supported for HP/UX compatibility. The @var{keyword}
367 argument must be one of the strings @samp{archive}, @samp{shared}, or
368 @samp{default}. @samp{-aarchive} is functionally equivalent to
369 @samp{-Bstatic}, and the other two keywords are functionally equivalent
370 to @samp{-Bdynamic}. This option may be used any number of times.
372 @kindex --audit @var{AUDITLIB}
373 @item --audit @var{AUDITLIB}
374 Adds @var{AUDITLIB} to the @code{DT_AUDIT} entry of the dynamic section.
375 @var{AUDITLIB} is not checked for existence, nor will it use the DT_SONAME
376 specified in the library. If specified multiple times @code{DT_AUDIT}
377 will contain a colon separated list of audit interfaces to use. If the linker
378 finds an object with an audit entry while searching for shared libraries,
379 it will add a corresponding @code{DT_DEPAUDIT} entry in the output file.
380 This option is only meaningful on ELF platforms supporting the rtld-audit
383 @ifclear SingleFormat
384 @cindex binary input format
385 @kindex -b @var{format}
386 @kindex --format=@var{format}
389 @item -b @var{input-format}
390 @itemx --format=@var{input-format}
391 @command{ld} may be configured to support more than one kind of object
392 file. If your @command{ld} is configured this way, you can use the
393 @samp{-b} option to specify the binary format for input object files
394 that follow this option on the command line. Even when @command{ld} is
395 configured to support alternative object formats, you don't usually need
396 to specify this, as @command{ld} should be configured to expect as a
397 default input format the most usual format on each machine.
398 @var{input-format} is a text string, the name of a particular format
399 supported by the BFD libraries. (You can list the available binary
400 formats with @samp{objdump -i}.)
403 You may want to use this option if you are linking files with an unusual
404 binary format. You can also use @samp{-b} to switch formats explicitly (when
405 linking object files of different formats), by including
406 @samp{-b @var{input-format}} before each group of object files in a
409 The default format is taken from the environment variable
414 You can also define the input format from a script, using the command
417 see @ref{Format Commands}.
421 @kindex -c @var{MRI-cmdfile}
422 @kindex --mri-script=@var{MRI-cmdfile}
423 @cindex compatibility, MRI
424 @item -c @var{MRI-commandfile}
425 @itemx --mri-script=@var{MRI-commandfile}
426 For compatibility with linkers produced by MRI, @command{ld} accepts script
427 files written in an alternate, restricted command language, described in
429 @ref{MRI,,MRI Compatible Script Files}.
432 the MRI Compatible Script Files section of GNU ld documentation.
434 Introduce MRI script files with
435 the option @samp{-c}; use the @samp{-T} option to run linker
436 scripts written in the general-purpose @command{ld} scripting language.
437 If @var{MRI-cmdfile} does not exist, @command{ld} looks for it in the directories
438 specified by any @samp{-L} options.
440 @cindex common allocation
447 These three options are equivalent; multiple forms are supported for
448 compatibility with other linkers. They assign space to common symbols
449 even if a relocatable output file is specified (with @samp{-r}). The
450 script command @code{FORCE_COMMON_ALLOCATION} has the same effect.
451 @xref{Miscellaneous Commands}.
453 @kindex --depaudit @var{AUDITLIB}
454 @kindex -P @var{AUDITLIB}
455 @item --depaudit @var{AUDITLIB}
456 @itemx -P @var{AUDITLIB}
457 Adds @var{AUDITLIB} to the @code{DT_DEPAUDIT} entry of the dynamic section.
458 @var{AUDITLIB} is not checked for existence, nor will it use the DT_SONAME
459 specified in the library. If specified multiple times @code{DT_DEPAUDIT}
460 will contain a colon separated list of audit interfaces to use. This
461 option is only meaningful on ELF platforms supporting the rtld-audit interface.
462 The -P option is provided for Solaris compatibility.
464 @kindex --enable-non-contiguous-regions
465 @item --enable-non-contiguous-regions
466 This option avoids generating an error if an input section does not
467 fit a matching output section. The linker tries to allocate the input
468 section to subseque nt matching output sections, and generates an
469 error only if no output section is large enough. This is useful when
470 several non-contiguous memory regions are available and the input
471 section does not require a particular one. The order in which input
472 sections are evaluated does not change, for instance:
476 MEM1 (rwx) : ORIGIN : 0x1000, LENGTH = 0x14
477 MEM2 (rwx) : ORIGIN : 0x1000, LENGTH = 0x40
478 MEM3 (rwx) : ORIGIN : 0x2000, LENGTH = 0x40
481 mem1 : @{ *(.data.*); @} > MEM1
482 mem2 : @{ *(.data.*); @} > MEM2
483 mem3 : @{ *(.data.*); @} > MEM2
491 results in .data.1 affected to mem1, and .data.2 and .data.3
492 affected to mem2, even though .data.3 would fit in mem3.
495 This option is incompatible with INSERT statements because it changes
496 the way input sections are mapped to output sections.
498 @kindex --enable-non-contiguous-regions-warnings
499 @item --enable-non-contiguous-regions-warnings
500 This option enables warnings when
501 @code{--enable-non-contiguous-regions} allows possibly unexpected
502 matches in sections mapping, potentially leading to silently
503 discarding a section instead of failing because it does not fit any
506 @cindex entry point, from command line
507 @kindex -e @var{entry}
508 @kindex --entry=@var{entry}
510 @itemx --entry=@var{entry}
511 Use @var{entry} as the explicit symbol for beginning execution of your
512 program, rather than the default entry point. If there is no symbol
513 named @var{entry}, the linker will try to parse @var{entry} as a number,
514 and use that as the entry address (the number will be interpreted in
515 base 10; you may use a leading @samp{0x} for base 16, or a leading
516 @samp{0} for base 8). @xref{Entry Point}, for a discussion of defaults
517 and other ways of specifying the entry point.
519 @kindex --exclude-libs
520 @item --exclude-libs @var{lib},@var{lib},...
521 Specifies a list of archive libraries from which symbols should not be automatically
522 exported. The library names may be delimited by commas or colons. Specifying
523 @code{--exclude-libs ALL} excludes symbols in all archive libraries from
524 automatic export. This option is available only for the i386 PE targeted
525 port of the linker and for ELF targeted ports. For i386 PE, symbols
526 explicitly listed in a .def file are still exported, regardless of this
527 option. For ELF targeted ports, symbols affected by this option will
528 be treated as hidden.
530 @kindex --exclude-modules-for-implib
531 @item --exclude-modules-for-implib @var{module},@var{module},...
532 Specifies a list of object files or archive members, from which symbols
533 should not be automatically exported, but which should be copied wholesale
534 into the import library being generated during the link. The module names
535 may be delimited by commas or colons, and must match exactly the filenames
536 used by @command{ld} to open the files; for archive members, this is simply
537 the member name, but for object files the name listed must include and
538 match precisely any path used to specify the input file on the linker's
539 command-line. This option is available only for the i386 PE targeted port
540 of the linker. Symbols explicitly listed in a .def file are still exported,
541 regardless of this option.
543 @cindex dynamic symbol table
545 @kindex --export-dynamic
546 @kindex --no-export-dynamic
548 @itemx --export-dynamic
549 @itemx --no-export-dynamic
550 When creating a dynamically linked executable, using the @option{-E}
551 option or the @option{--export-dynamic} option causes the linker to add
552 all symbols to the dynamic symbol table. The dynamic symbol table is the
553 set of symbols which are visible from dynamic objects at run time.
555 If you do not use either of these options (or use the
556 @option{--no-export-dynamic} option to restore the default behavior), the
557 dynamic symbol table will normally contain only those symbols which are
558 referenced by some dynamic object mentioned in the link.
560 If you use @code{dlopen} to load a dynamic object which needs to refer
561 back to the symbols defined by the program, rather than some other
562 dynamic object, then you will probably need to use this option when
563 linking the program itself.
565 You can also use the dynamic list to control what symbols should
566 be added to the dynamic symbol table if the output format supports it.
567 See the description of @samp{--dynamic-list}.
569 Note that this option is specific to ELF targeted ports. PE targets
570 support a similar function to export all symbols from a DLL or EXE; see
571 the description of @samp{--export-all-symbols} below.
573 @kindex --export-dynamic-symbol=@var{glob}
574 @cindex export dynamic symbol
575 @item --export-dynamic-symbol=@var{glob}
576 When creating a dynamically linked executable, symbols matching
577 @var{glob} will be added to the dynamic symbol table. When creating a
578 shared library, references to symbols matching @var{glob} will not be
579 bound to the definitions within the shared library. This option is a
580 no-op when creating a shared library and @samp{-Bsymbolic} or
581 @samp{--dynamic-list} are not specified. This option is only meaningful
582 on ELF platforms which support shared libraries.
584 @kindex --export-dynamic-symbol-list=@var{file}
585 @cindex export dynamic symbol list
586 @item --export-dynamic-symbol-list=@var{file}
587 Specify a @samp{--export-dynamic-symbol} for each pattern in the file.
588 The format of the file is the same as the version node without
589 scope and node name. See @ref{VERSION} for more information.
591 @ifclear SingleFormat
592 @cindex big-endian objects
596 Link big-endian objects. This affects the default output format.
598 @cindex little-endian objects
601 Link little-endian objects. This affects the default output format.
604 @kindex -f @var{name}
605 @kindex --auxiliary=@var{name}
607 @itemx --auxiliary=@var{name}
608 When creating an ELF shared object, set the internal DT_AUXILIARY field
609 to the specified name. This tells the dynamic linker that the symbol
610 table of the shared object should be used as an auxiliary filter on the
611 symbol table of the shared object @var{name}.
613 If you later link a program against this filter object, then, when you
614 run the program, the dynamic linker will see the DT_AUXILIARY field. If
615 the dynamic linker resolves any symbols from the filter object, it will
616 first check whether there is a definition in the shared object
617 @var{name}. If there is one, it will be used instead of the definition
618 in the filter object. The shared object @var{name} need not exist.
619 Thus the shared object @var{name} may be used to provide an alternative
620 implementation of certain functions, perhaps for debugging or for
621 machine-specific performance.
623 This option may be specified more than once. The DT_AUXILIARY entries
624 will be created in the order in which they appear on the command line.
626 @kindex -F @var{name}
627 @kindex --filter=@var{name}
629 @itemx --filter=@var{name}
630 When creating an ELF shared object, set the internal DT_FILTER field to
631 the specified name. This tells the dynamic linker that the symbol table
632 of the shared object which is being created should be used as a filter
633 on the symbol table of the shared object @var{name}.
635 If you later link a program against this filter object, then, when you
636 run the program, the dynamic linker will see the DT_FILTER field. The
637 dynamic linker will resolve symbols according to the symbol table of the
638 filter object as usual, but it will actually link to the definitions
639 found in the shared object @var{name}. Thus the filter object can be
640 used to select a subset of the symbols provided by the object
643 Some older linkers used the @option{-F} option throughout a compilation
644 toolchain for specifying object-file format for both input and output
646 @ifclear SingleFormat
647 The @sc{gnu} linker uses other mechanisms for this purpose: the
648 @option{-b}, @option{--format}, @option{--oformat} options, the
649 @code{TARGET} command in linker scripts, and the @code{GNUTARGET}
650 environment variable.
652 The @sc{gnu} linker will ignore the @option{-F} option when not
653 creating an ELF shared object.
655 @cindex finalization function
656 @kindex -fini=@var{name}
657 @item -fini=@var{name}
658 When creating an ELF executable or shared object, call NAME when the
659 executable or shared object is unloaded, by setting DT_FINI to the
660 address of the function. By default, the linker uses @code{_fini} as
661 the function to call.
665 Ignored. Provided for compatibility with other tools.
667 @kindex -G @var{value}
668 @kindex --gpsize=@var{value}
671 @itemx --gpsize=@var{value}
672 Set the maximum size of objects to be optimized using the GP register to
673 @var{size}. This is only meaningful for object file formats such as
674 MIPS ELF that support putting large and small objects into different
675 sections. This is ignored for other object file formats.
677 @cindex runtime library name
678 @kindex -h @var{name}
679 @kindex -soname=@var{name}
681 @itemx -soname=@var{name}
682 When creating an ELF shared object, set the internal DT_SONAME field to
683 the specified name. When an executable is linked with a shared object
684 which has a DT_SONAME field, then when the executable is run the dynamic
685 linker will attempt to load the shared object specified by the DT_SONAME
686 field rather than using the file name given to the linker.
689 @cindex incremental link
691 Perform an incremental link (same as option @samp{-r}).
693 @cindex initialization function
694 @kindex -init=@var{name}
695 @item -init=@var{name}
696 When creating an ELF executable or shared object, call NAME when the
697 executable or shared object is loaded, by setting DT_INIT to the address
698 of the function. By default, the linker uses @code{_init} as the
701 @cindex archive files, from cmd line
702 @kindex -l @var{namespec}
703 @kindex --library=@var{namespec}
704 @item -l @var{namespec}
705 @itemx --library=@var{namespec}
706 Add the archive or object file specified by @var{namespec} to the
707 list of files to link. This option may be used any number of times.
708 If @var{namespec} is of the form @file{:@var{filename}}, @command{ld}
709 will search the library path for a file called @var{filename}, otherwise it
710 will search the library path for a file called @file{lib@var{namespec}.a}.
712 On systems which support shared libraries, @command{ld} may also search for
713 files other than @file{lib@var{namespec}.a}. Specifically, on ELF
714 and SunOS systems, @command{ld} will search a directory for a library
715 called @file{lib@var{namespec}.so} before searching for one called
716 @file{lib@var{namespec}.a}. (By convention, a @code{.so} extension
717 indicates a shared library.) Note that this behavior does not apply
718 to @file{:@var{filename}}, which always specifies a file called
721 The linker will search an archive only once, at the location where it is
722 specified on the command line. If the archive defines a symbol which
723 was undefined in some object which appeared before the archive on the
724 command line, the linker will include the appropriate file(s) from the
725 archive. However, an undefined symbol in an object appearing later on
726 the command line will not cause the linker to search the archive again.
728 See the @option{-(} option for a way to force the linker to search
729 archives multiple times.
731 You may list the same archive multiple times on the command line.
734 This type of archive searching is standard for Unix linkers. However,
735 if you are using @command{ld} on AIX, note that it is different from the
736 behaviour of the AIX linker.
739 @cindex search directory, from cmd line
741 @kindex --library-path=@var{dir}
742 @item -L @var{searchdir}
743 @itemx --library-path=@var{searchdir}
744 Add path @var{searchdir} to the list of paths that @command{ld} will search
745 for archive libraries and @command{ld} control scripts. You may use this
746 option any number of times. The directories are searched in the order
747 in which they are specified on the command line. Directories specified
748 on the command line are searched before the default directories. All
749 @option{-L} options apply to all @option{-l} options, regardless of the
750 order in which the options appear. @option{-L} options do not affect
751 how @command{ld} searches for a linker script unless @option{-T}
754 If @var{searchdir} begins with @code{=} or @code{$SYSROOT}, then this
755 prefix will be replaced by the @dfn{sysroot prefix}, controlled by the
756 @samp{--sysroot} option, or specified when the linker is configured.
759 The default set of paths searched (without being specified with
760 @samp{-L}) depends on which emulation mode @command{ld} is using, and in
761 some cases also on how it was configured. @xref{Environment}.
764 The paths can also be specified in a link script with the
765 @code{SEARCH_DIR} command. Directories specified this way are searched
766 at the point in which the linker script appears in the command line.
769 @kindex -m @var{emulation}
770 @item -m @var{emulation}
771 Emulate the @var{emulation} linker. You can list the available
772 emulations with the @samp{--verbose} or @samp{-V} options.
774 If the @samp{-m} option is not used, the emulation is taken from the
775 @code{LDEMULATION} environment variable, if that is defined.
777 Otherwise, the default emulation depends upon how the linker was
785 Print a link map to the standard output. A link map provides
786 information about the link, including the following:
790 Where object files are mapped into memory.
792 How common symbols are allocated.
794 All archive members included in the link, with a mention of the symbol
795 which caused the archive member to be brought in.
797 The values assigned to symbols.
799 Note - symbols whose values are computed by an expression which
800 involves a reference to a previous value of the same symbol may not
801 have correct result displayed in the link map. This is because the
802 linker discards intermediate results and only retains the final value
803 of an expression. Under such circumstances the linker will display
804 the final value enclosed by square brackets. Thus for example a
805 linker script containing:
813 will produce the following output in the link map if the @option{-M}
818 [0x0000000c] foo = (foo * 0x4)
819 [0x0000000c] foo = (foo + 0x8)
822 See @ref{Expressions} for more information about expressions in linker
826 How GNU properties are merged.
828 When the linker merges input .note.gnu.property sections into one output
829 .note.gnu.property section, some properties are removed or updated.
830 These actions are reported in the link map. For example:
833 Removed property 0xc0000002 to merge foo.o (0x1) and bar.o (not found)
836 This indicates that property 0xc0000002 is removed from output when
837 merging properties in @file{foo.o}, whose property 0xc0000002 value
838 is 0x1, and @file{bar.o}, which doesn't have property 0xc0000002.
841 Updated property 0xc0010001 (0x1) to merge foo.o (0x1) and bar.o (0x1)
844 This indicates that property 0xc0010001 value is updated to 0x1 in output
845 when merging properties in @file{foo.o}, whose 0xc0010001 property value
846 is 0x1, and @file{bar.o}, whose 0xc0010001 property value is 0x1.
849 @cindex link map discarded
850 @kindex --print-map-discarded
851 @kindex --no-print-map-discarded
852 @item --print-map-discarded
853 @itemx --no-print-map-discarded
854 Print (or do not print) the list of discarded and garbage collected sections
855 in the link map. Enabled by default.
858 @cindex read-only text
863 Turn off page alignment of sections, and disable linking against shared
864 libraries. If the output format supports Unix style magic numbers,
865 mark the output as @code{NMAGIC}.
869 @cindex read/write from cmd line
873 Set the text and data sections to be readable and writable. Also, do
874 not page-align the data segment, and disable linking against shared
875 libraries. If the output format supports Unix style magic numbers,
876 mark the output as @code{OMAGIC}. Note: Although a writable text section
877 is allowed for PE-COFF targets, it does not conform to the format
878 specification published by Microsoft.
883 This option negates most of the effects of the @option{-N} option. It
884 sets the text section to be read-only, and forces the data segment to
885 be page-aligned. Note - this option does not enable linking against
886 shared libraries. Use @option{-Bdynamic} for this.
888 @kindex -o @var{output}
889 @kindex --output=@var{output}
890 @cindex naming the output file
891 @item -o @var{output}
892 @itemx --output=@var{output}
893 Use @var{output} as the name for the program produced by @command{ld}; if this
894 option is not specified, the name @file{a.out} is used by default. The
895 script command @code{OUTPUT} can also specify the output file name.
897 @kindex --dependency-file=@var{depfile}
898 @cindex dependency file
899 @item --dependency-file=@var{depfile}
900 Write a @dfn{dependency file} to @var{depfile}. This file contains a rule
901 suitable for @code{make} describing the output file and all the input files
902 that were read to produce it. The output is similar to the compiler's
903 output with @samp{-M -MP} (@pxref{Preprocessor Options,, Options
904 Controlling the Preprocessor, gcc.info, Using the GNU Compiler
905 Collection}). Note that there is no option like the compiler's @samp{-MM},
906 to exclude ``system files'' (which is not a well-specified concept in the
907 linker, unlike ``system headers'' in the compiler). So the output from
908 @samp{--dependency-file} is always specific to the exact state of the
909 installation where it was produced, and should not be copied into
910 distributed makefiles without careful editing.
912 @kindex -O @var{level}
913 @cindex generating optimized output
915 If @var{level} is a numeric values greater than zero @command{ld} optimizes
916 the output. This might take significantly longer and therefore probably
917 should only be enabled for the final binary. At the moment this
918 option only affects ELF shared library generation. Future releases of
919 the linker may make more use of this option. Also currently there is
920 no difference in the linker's behaviour for different non-zero values
921 of this option. Again this may change with future releases.
923 @kindex -plugin @var{name}
924 @item -plugin @var{name}
925 Involve a plugin in the linking process. The @var{name} parameter is
926 the absolute filename of the plugin. Usually this parameter is
927 automatically added by the complier, when using link time
928 optimization, but users can also add their own plugins if they so
931 Note that the location of the compiler originated plugins is different
932 from the place where the @command{ar}, @command{nm} and
933 @command{ranlib} programs search for their plugins. In order for
934 those commands to make use of a compiler based plugin it must first be
935 copied into the @file{$@{libdir@}/bfd-plugins} directory. All gcc
936 based linker plugins are backward compatible, so it is sufficient to
937 just copy in the newest one.
940 @cindex push state governing input file handling
942 The @option{--push-state} allows one to preserve the current state of the
943 flags which govern the input file handling so that they can all be
944 restored with one corresponding @option{--pop-state} option.
946 The option which are covered are: @option{-Bdynamic}, @option{-Bstatic},
947 @option{-dn}, @option{-dy}, @option{-call_shared}, @option{-non_shared},
948 @option{-static}, @option{-N}, @option{-n}, @option{--whole-archive},
949 @option{--no-whole-archive}, @option{-r}, @option{-Ur},
950 @option{--copy-dt-needed-entries}, @option{--no-copy-dt-needed-entries},
951 @option{--as-needed}, @option{--no-as-needed}, and @option{-a}.
953 One target for this option are specifications for @file{pkg-config}. When
954 used with the @option{--libs} option all possibly needed libraries are
955 listed and then possibly linked with all the time. It is better to return
956 something as follows:
959 -Wl,--push-state,--as-needed -libone -libtwo -Wl,--pop-state
963 @cindex pop state governing input file handling
965 Undoes the effect of --push-state, restores the previous values of the
966 flags governing input file handling.
969 @kindex --emit-relocs
970 @cindex retain relocations in final executable
973 Leave relocation sections and contents in fully linked executables.
974 Post link analysis and optimization tools may need this information in
975 order to perform correct modifications of executables. This results
976 in larger executables.
978 This option is currently only supported on ELF platforms.
980 @kindex --force-dynamic
981 @cindex forcing the creation of dynamic sections
982 @item --force-dynamic
983 Force the output file to have dynamic sections. This option is specific
987 @cindex relocatable output
989 @kindex --relocatable
992 Generate relocatable output---i.e., generate an output file that can in
993 turn serve as input to @command{ld}. This is often called @dfn{partial
994 linking}. As a side effect, in environments that support standard Unix
995 magic numbers, this option also sets the output file's magic number to
997 @c ; see @option{-N}.
998 If this option is not specified, an absolute file is produced. When
999 linking C++ programs, this option @emph{will not} resolve references to
1000 constructors; to do that, use @samp{-Ur}.
1002 When an input file does not have the same format as the output file,
1003 partial linking is only supported if that input file does not contain any
1004 relocations. Different output formats can have further restrictions; for
1005 example some @code{a.out}-based formats do not support partial linking
1006 with input files in other formats at all.
1008 This option does the same thing as @samp{-i}.
1010 @kindex -R @var{file}
1011 @kindex --just-symbols=@var{file}
1012 @cindex symbol-only input
1013 @item -R @var{filename}
1014 @itemx --just-symbols=@var{filename}
1015 Read symbol names and their addresses from @var{filename}, but do not
1016 relocate it or include it in the output. This allows your output file
1017 to refer symbolically to absolute locations of memory defined in other
1018 programs. You may use this option more than once.
1020 For compatibility with other ELF linkers, if the @option{-R} option is
1021 followed by a directory name, rather than a file name, it is treated as
1022 the @option{-rpath} option.
1026 @cindex strip all symbols
1029 Omit all symbol information from the output file.
1032 @kindex --strip-debug
1033 @cindex strip debugger symbols
1035 @itemx --strip-debug
1036 Omit debugger symbol information (but not all symbols) from the output file.
1038 @kindex --strip-discarded
1039 @kindex --no-strip-discarded
1040 @item --strip-discarded
1041 @itemx --no-strip-discarded
1042 Omit (or do not omit) global symbols defined in discarded sections.
1047 @cindex input files, displaying
1050 Print the names of the input files as @command{ld} processes them. If
1051 @samp{-t} is given twice then members within archives are also printed.
1052 @samp{-t} output is useful to generate a list of all the object files
1053 and scripts involved in linking, for example, when packaging files for
1054 a linker bug report.
1056 @kindex -T @var{script}
1057 @kindex --script=@var{script}
1058 @cindex script files
1059 @item -T @var{scriptfile}
1060 @itemx --script=@var{scriptfile}
1061 Use @var{scriptfile} as the linker script. This script replaces
1062 @command{ld}'s default linker script (rather than adding to it), so
1063 @var{commandfile} must specify everything necessary to describe the
1064 output file. @xref{Scripts}. If @var{scriptfile} does not exist in
1065 the current directory, @code{ld} looks for it in the directories
1066 specified by any preceding @samp{-L} options. Multiple @samp{-T}
1069 @kindex -dT @var{script}
1070 @kindex --default-script=@var{script}
1071 @cindex script files
1072 @item -dT @var{scriptfile}
1073 @itemx --default-script=@var{scriptfile}
1074 Use @var{scriptfile} as the default linker script. @xref{Scripts}.
1076 This option is similar to the @option{--script} option except that
1077 processing of the script is delayed until after the rest of the
1078 command line has been processed. This allows options placed after the
1079 @option{--default-script} option on the command line to affect the
1080 behaviour of the linker script, which can be important when the linker
1081 command line cannot be directly controlled by the user. (eg because
1082 the command line is being constructed by another tool, such as
1085 @kindex -u @var{symbol}
1086 @kindex --undefined=@var{symbol}
1087 @cindex undefined symbol
1088 @item -u @var{symbol}
1089 @itemx --undefined=@var{symbol}
1090 Force @var{symbol} to be entered in the output file as an undefined
1091 symbol. Doing this may, for example, trigger linking of additional
1092 modules from standard libraries. @samp{-u} may be repeated with
1093 different option arguments to enter additional undefined symbols. This
1094 option is equivalent to the @code{EXTERN} linker script command.
1096 If this option is being used to force additional modules to be pulled
1097 into the link, and if it is an error for the symbol to remain
1098 undefined, then the option @option{--require-defined} should be used
1101 @kindex --require-defined=@var{symbol}
1102 @cindex symbols, require defined
1103 @cindex defined symbol
1104 @item --require-defined=@var{symbol}
1105 Require that @var{symbol} is defined in the output file. This option
1106 is the same as option @option{--undefined} except that if @var{symbol}
1107 is not defined in the output file then the linker will issue an error
1108 and exit. The same effect can be achieved in a linker script by using
1109 @code{EXTERN}, @code{ASSERT} and @code{DEFINED} together. This option
1110 can be used multiple times to require additional symbols.
1113 @cindex constructors
1115 For anything other than C++ programs, this option is equivalent to
1116 @samp{-r}: it generates relocatable output---i.e., an output file that can in
1117 turn serve as input to @command{ld}. When linking C++ programs, @samp{-Ur}
1118 @emph{does} resolve references to constructors, unlike @samp{-r}.
1119 It does not work to use @samp{-Ur} on files that were themselves linked
1120 with @samp{-Ur}; once the constructor table has been built, it cannot
1121 be added to. Use @samp{-Ur} only for the last partial link, and
1122 @samp{-r} for the others.
1124 @kindex --orphan-handling=@var{MODE}
1125 @cindex orphan sections
1126 @cindex sections, orphan
1127 @item --orphan-handling=@var{MODE}
1128 Control how orphan sections are handled. An orphan section is one not
1129 specifically mentioned in a linker script. @xref{Orphan Sections}.
1131 @var{MODE} can have any of the following values:
1135 Orphan sections are placed into a suitable output section following
1136 the strategy described in @ref{Orphan Sections}. The option
1137 @samp{--unique} also affects how sections are placed.
1140 All orphan sections are discarded, by placing them in the
1141 @samp{/DISCARD/} section (@pxref{Output Section Discarding}).
1144 The linker will place the orphan section as for @code{place} and also
1148 The linker will exit with an error if any orphan section is found.
1151 The default if @samp{--orphan-handling} is not given is @code{place}.
1153 @kindex --unique[=@var{SECTION}]
1154 @item --unique[=@var{SECTION}]
1155 Creates a separate output section for every input section matching
1156 @var{SECTION}, or if the optional wildcard @var{SECTION} argument is
1157 missing, for every orphan input section. An orphan section is one not
1158 specifically mentioned in a linker script. You may use this option
1159 multiple times on the command line; It prevents the normal merging of
1160 input sections with the same name, overriding output section assignments
1170 Display the version number for @command{ld}. The @option{-V} option also
1171 lists the supported emulations.
1174 @kindex --discard-all
1175 @cindex deleting local symbols
1177 @itemx --discard-all
1178 Delete all local symbols.
1181 @kindex --discard-locals
1182 @cindex local symbols, deleting
1184 @itemx --discard-locals
1185 Delete all temporary local symbols. (These symbols start with
1186 system-specific local label prefixes, typically @samp{.L} for ELF systems
1187 or @samp{L} for traditional a.out systems.)
1189 @kindex -y @var{symbol}
1190 @kindex --trace-symbol=@var{symbol}
1191 @cindex symbol tracing
1192 @item -y @var{symbol}
1193 @itemx --trace-symbol=@var{symbol}
1194 Print the name of each linked file in which @var{symbol} appears. This
1195 option may be given any number of times. On many systems it is necessary
1196 to prepend an underscore.
1198 This option is useful when you have an undefined symbol in your link but
1199 don't know where the reference is coming from.
1201 @kindex -Y @var{path}
1203 Add @var{path} to the default library search path. This option exists
1204 for Solaris compatibility.
1206 @kindex -z @var{keyword}
1207 @item -z @var{keyword}
1208 The recognized keywords are:
1212 Always generate BND prefix in PLT entries. Supported for Linux/x86_64.
1214 @item call-nop=prefix-addr
1215 @itemx call-nop=suffix-nop
1216 @itemx call-nop=prefix-@var{byte}
1217 @itemx call-nop=suffix-@var{byte}
1218 Specify the 1-byte @code{NOP} padding when transforming indirect call
1219 to a locally defined function, foo, via its GOT slot.
1220 @option{call-nop=prefix-addr} generates @code{0x67 call foo}.
1221 @option{call-nop=suffix-nop} generates @code{call foo 0x90}.
1222 @option{call-nop=prefix-@var{byte}} generates @code{@var{byte} call foo}.
1223 @option{call-nop=suffix-@var{byte}} generates @code{call foo @var{byte}}.
1224 Supported for i386 and x86_64.
1226 @item cet-report=none
1227 @itemx cet-report=warning
1228 @itemx cet-report=error
1229 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_IBT and
1230 GNU_PROPERTY_X86_FEATURE_1_SHSTK properties in input .note.gnu.property
1231 section. @option{cet-report=none}, which is the default, will make the
1232 linker not report missing properties in input files.
1233 @option{cet-report=warning} will make the linker issue a warning for
1234 missing properties in input files. @option{cet-report=error} will make
1235 the linker issue an error for missing properties in input files.
1236 Note that @option{ibt} will turn off the missing
1237 GNU_PROPERTY_X86_FEATURE_1_IBT property report and @option{shstk} will
1238 turn off the missing GNU_PROPERTY_X86_FEATURE_1_SHSTK property report.
1239 Supported for Linux/i386 and Linux/x86_64.
1243 Combine multiple dynamic relocation sections and sort to improve
1244 dynamic symbol lookup caching. Do not do this if @samp{nocombreloc}.
1248 Generate common symbols with STT_COMMON type during a relocatable
1249 link. Use STT_OBJECT type if @samp{nocommon}.
1251 @item common-page-size=@var{value}
1252 Set the page size most commonly used to @var{value}. Memory image
1253 layout will be optimized to minimize memory pages if the system is
1254 using pages of this size.
1257 Report unresolved symbol references from regular object files. This
1258 is done even if the linker is creating a non-symbolic shared library.
1259 This option is the inverse of @samp{-z undefs}.
1261 @item dynamic-undefined-weak
1262 @itemx nodynamic-undefined-weak
1263 Make undefined weak symbols dynamic when building a dynamic object,
1264 if they are referenced from a regular object file and not forced local
1265 by symbol visibility or versioning. Do not make them dynamic if
1266 @samp{nodynamic-undefined-weak}. If neither option is given, a target
1267 may default to either option being in force, or make some other
1268 selection of undefined weak symbols dynamic. Not all targets support
1272 Marks the object as requiring executable stack.
1275 This option is only meaningful when building a shared object. It makes
1276 the symbols defined by this shared object available for symbol resolution
1277 of subsequently loaded libraries.
1280 This option is only meaningful when building a dynamic executable.
1281 This option marks the executable as requiring global auditing by
1282 setting the @code{DF_1_GLOBAUDIT} bit in the @code{DT_FLAGS_1} dynamic
1283 tag. Global auditing requires that any auditing library defined via
1284 the @option{--depaudit} or @option{-P} command-line options be run for
1285 all dynamic objects loaded by the application.
1288 Generate Intel Indirect Branch Tracking (IBT) enabled PLT entries.
1289 Supported for Linux/i386 and Linux/x86_64.
1292 Generate GNU_PROPERTY_X86_FEATURE_1_IBT in .note.gnu.property section
1293 to indicate compatibility with IBT. This also implies @option{ibtplt}.
1294 Supported for Linux/i386 and Linux/x86_64.
1296 @item indirect-extern-access
1297 @itemx noindirect-extern-access
1298 Generate GNU_PROPERTY_1_NEEDED_INDIRECT_EXTERN_ACCESS in
1299 .note.gnu.property section to indicate that object file requires
1300 canonical function pointers and cannot be used with copy relocation.
1301 This option also implies @option{noextern-protected-data} and
1302 @option{nocopyreloc}. Supported for i386 and x86-64.
1304 @option{noindirect-extern-access} removes
1305 GNU_PROPERTY_1_NEEDED_INDIRECT_EXTERN_ACCESS from .note.gnu.property
1309 This option is only meaningful when building a shared object.
1310 It marks the object so that its runtime initialization will occur
1311 before the runtime initialization of any other objects brought into
1312 the process at the same time. Similarly the runtime finalization of
1313 the object will occur after the runtime finalization of any other
1317 Specify that the dynamic loader should modify its symbol search order
1318 so that symbols in this shared library interpose all other shared
1319 libraries not so marked.
1323 When generating a shared library or other dynamically loadable ELF
1324 object mark it as one that should (by default) only ever be loaded once,
1325 and only in the main namespace (when using @code{dlmopen}). This is
1326 primarily used to mark fundamental libraries such as libc, libpthread et
1327 al which do not usually function correctly unless they are the sole instances
1328 of themselves. This behaviour can be overridden by the @code{dlmopen} caller
1329 and does not apply to certain loading mechanisms (such as audit libraries).
1332 Generate GNU_PROPERTY_X86_FEATURE_1_LAM_U48 in .note.gnu.property section
1333 to indicate compatibility with Intel LAM_U48. Supported for Linux/x86_64.
1336 Generate GNU_PROPERTY_X86_FEATURE_1_LAM_U57 in .note.gnu.property section
1337 to indicate compatibility with Intel LAM_U57. Supported for Linux/x86_64.
1339 @item lam-u48-report=none
1340 @itemx lam-u48-report=warning
1341 @itemx lam-u48-report=error
1342 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U48
1343 property in input .note.gnu.property section.
1344 @option{lam-u48-report=none}, which is the default, will make the
1345 linker not report missing properties in input files.
1346 @option{lam-u48-report=warning} will make the linker issue a warning for
1347 missing properties in input files. @option{lam-u48-report=error} will
1348 make the linker issue an error for missing properties in input files.
1349 Supported for Linux/x86_64.
1351 @item lam-u57-report=none
1352 @itemx lam-u57-report=warning
1353 @itemx lam-u57-report=error
1354 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U57
1355 property in input .note.gnu.property section.
1356 @option{lam-u57-report=none}, which is the default, will make the
1357 linker not report missing properties in input files.
1358 @option{lam-u57-report=warning} will make the linker issue a warning for
1359 missing properties in input files. @option{lam-u57-report=error} will
1360 make the linker issue an error for missing properties in input files.
1361 Supported for Linux/x86_64.
1363 @item lam-report=none
1364 @itemx lam-report=warning
1365 @itemx lam-report=error
1366 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U48 and
1367 GNU_PROPERTY_X86_FEATURE_1_LAM_U57 properties in input .note.gnu.property
1368 section. @option{lam-report=none}, which is the default, will make the
1369 linker not report missing properties in input files.
1370 @option{lam-report=warning} will make the linker issue a warning for
1371 missing properties in input files. @option{lam-report=error} will make
1372 the linker issue an error for missing properties in input files.
1373 Supported for Linux/x86_64.
1376 When generating an executable or shared library, mark it to tell the
1377 dynamic linker to defer function call resolution to the point when
1378 the function is called (lazy binding), rather than at load time.
1379 Lazy binding is the default.
1382 Specify that the object's filters be processed immediately at runtime.
1384 @item max-page-size=@var{value}
1385 Set the maximum memory page size supported to @var{value}.
1388 Allow multiple definitions.
1391 Disable linker generated .dynbss variables used in place of variables
1392 defined in shared libraries. May result in dynamic text relocations.
1395 Specify that the dynamic loader search for dependencies of this object
1396 should ignore any default library search paths.
1399 Specify that the object shouldn't be unloaded at runtime.
1402 Specify that the object is not available to @code{dlopen}.
1405 Specify that the object can not be dumped by @code{dldump}.
1408 Marks the object as not requiring executable stack.
1410 @item noextern-protected-data
1411 Don't treat protected data symbols as external when building a shared
1412 library. This option overrides the linker backend default. It can be
1413 used to work around incorrect relocations against protected data symbols
1414 generated by compiler. Updates on protected data symbols by another
1415 module aren't visible to the resulting shared library. Supported for
1418 @item noreloc-overflow
1419 Disable relocation overflow check. This can be used to disable
1420 relocation overflow check if there will be no dynamic relocation
1421 overflow at run-time. Supported for x86_64.
1424 When generating an executable or shared library, mark it to tell the
1425 dynamic linker to resolve all symbols when the program is started, or
1426 when the shared library is loaded by dlopen, instead of deferring
1427 function call resolution to the point when the function is first
1431 Specify that the object requires @samp{$ORIGIN} handling in paths.
1435 Create an ELF @code{PT_GNU_RELRO} segment header in the object. This
1436 specifies a memory segment that should be made read-only after
1437 relocation, if supported. Specifying @samp{common-page-size} smaller
1438 than the system page size will render this protection ineffective.
1439 Don't create an ELF @code{PT_GNU_RELRO} segment if @samp{norelro}.
1441 @item report-relative-reloc
1442 Report dynamic relative relocations generated by linker. Supported for
1443 Linux/i386 and Linux/x86_64.
1446 @itemx noseparate-code
1447 Create separate code @code{PT_LOAD} segment header in the object. This
1448 specifies a memory segment that should contain only instructions and must
1449 be in wholly disjoint pages from any other data. Don't create separate
1450 code @code{PT_LOAD} segment if @samp{noseparate-code} is used.
1453 Generate GNU_PROPERTY_X86_FEATURE_1_SHSTK in .note.gnu.property section
1454 to indicate compatibility with Intel Shadow Stack. Supported for
1455 Linux/i386 and Linux/x86_64.
1457 @item stack-size=@var{value}
1458 Specify a stack size for an ELF @code{PT_GNU_STACK} segment.
1459 Specifying zero will override any default non-zero sized
1460 @code{PT_GNU_STACK} segment creation.
1463 @itemx nostart-stop-gc
1464 @cindex start-stop-gc
1465 When @samp{--gc-sections} is in effect, a reference from a retained
1466 section to @code{__start_SECNAME} or @code{__stop_SECNAME} causes all
1467 input sections named @code{SECNAME} to also be retained, if
1468 @code{SECNAME} is representable as a C identifier and either
1469 @code{__start_SECNAME} or @code{__stop_SECNAME} is synthesized by the
1470 linker. @samp{-z start-stop-gc} disables this effect, allowing
1471 sections to be garbage collected as if the special synthesized symbols
1472 were not defined. @samp{-z start-stop-gc} has no effect on a
1473 definition of @code{__start_SECNAME} or @code{__stop_SECNAME} in an
1474 object file or linker script. Such a definition will prevent the
1475 linker providing a synthesized @code{__start_SECNAME} or
1476 @code{__stop_SECNAME} respectively, and therefore the special
1477 treatment by garbage collection for those references.
1479 @item start-stop-visibility=@var{value}
1481 @cindex ELF symbol visibility
1482 Specify the ELF symbol visibility for synthesized
1483 @code{__start_SECNAME} and @code{__stop_SECNAME} symbols (@pxref{Input
1484 Section Example}). @var{value} must be exactly @samp{default},
1485 @samp{internal}, @samp{hidden}, or @samp{protected}. If no @samp{-z
1486 start-stop-visibility} option is given, @samp{protected} is used for
1487 compatibility with historical practice. However, it's highly
1488 recommended to use @samp{-z start-stop-visibility=hidden} in new
1489 programs and shared libraries so that these symbols are not exported
1490 between shared objects, which is not usually what's intended.
1495 Report an error if DT_TEXTREL is set, i.e., if the position-independent
1496 or shared object has dynamic relocations in read-only sections. Don't
1497 report an error if @samp{notext} or @samp{textoff}.
1500 Do not report unresolved symbol references from regular object files,
1501 either when creating an executable, or when creating a shared library.
1502 This option is the inverse of @samp{-z defs}.
1505 @itemx nounique-symbol
1506 Avoid duplicated local symbol names in the symbol string table. Append
1507 ".@code{number}" to duplicated local symbol names if @samp{unique-symbol}
1508 is used. @option{nounique-symbol} is the default.
1510 @item x86-64-baseline
1514 Specify the x86-64 ISA level needed in .note.gnu.property section.
1515 @option{x86-64-baseline} generates @code{GNU_PROPERTY_X86_ISA_1_BASELINE}.
1516 @option{x86-64-v2} generates @code{GNU_PROPERTY_X86_ISA_1_V2}.
1517 @option{x86-64-v3} generates @code{GNU_PROPERTY_X86_ISA_1_V3}.
1518 @option{x86-64-v4} generates @code{GNU_PROPERTY_X86_ISA_1_V4}.
1519 Supported for Linux/i386 and Linux/x86_64.
1523 Other keywords are ignored for Solaris compatibility.
1526 @cindex groups of archives
1527 @item -( @var{archives} -)
1528 @itemx --start-group @var{archives} --end-group
1529 The @var{archives} should be a list of archive files. They may be
1530 either explicit file names, or @samp{-l} options.
1532 The specified archives are searched repeatedly until no new undefined
1533 references are created. Normally, an archive is searched only once in
1534 the order that it is specified on the command line. If a symbol in that
1535 archive is needed to resolve an undefined symbol referred to by an
1536 object in an archive that appears later on the command line, the linker
1537 would not be able to resolve that reference. By grouping the archives,
1538 they will all be searched repeatedly until all possible references are
1541 Using this option has a significant performance cost. It is best to use
1542 it only when there are unavoidable circular references between two or
1545 @kindex --accept-unknown-input-arch
1546 @kindex --no-accept-unknown-input-arch
1547 @item --accept-unknown-input-arch
1548 @itemx --no-accept-unknown-input-arch
1549 Tells the linker to accept input files whose architecture cannot be
1550 recognised. The assumption is that the user knows what they are doing
1551 and deliberately wants to link in these unknown input files. This was
1552 the default behaviour of the linker, before release 2.14. The default
1553 behaviour from release 2.14 onwards is to reject such input files, and
1554 so the @samp{--accept-unknown-input-arch} option has been added to
1555 restore the old behaviour.
1558 @kindex --no-as-needed
1560 @itemx --no-as-needed
1561 This option affects ELF DT_NEEDED tags for dynamic libraries mentioned
1562 on the command line after the @option{--as-needed} option. Normally
1563 the linker will add a DT_NEEDED tag for each dynamic library mentioned
1564 on the command line, regardless of whether the library is actually
1565 needed or not. @option{--as-needed} causes a DT_NEEDED tag to only be
1566 emitted for a library that @emph{at that point in the link} satisfies a
1567 non-weak undefined symbol reference from a regular object file or, if
1568 the library is not found in the DT_NEEDED lists of other needed libraries, a
1569 non-weak undefined symbol reference from another needed dynamic library.
1570 Object files or libraries appearing on the command line @emph{after}
1571 the library in question do not affect whether the library is seen as
1572 needed. This is similar to the rules for extraction of object files
1573 from archives. @option{--no-as-needed} restores the default behaviour.
1575 Note: On Linux based systems the @option{--as-needed} option also has
1576 an affect on the behaviour of the @option{--rpath} and
1577 @option{--rpath-link} options. See the description of
1578 @option{--rpath-link} for more details.
1580 @kindex --add-needed
1581 @kindex --no-add-needed
1583 @itemx --no-add-needed
1584 These two options have been deprecated because of the similarity of
1585 their names to the @option{--as-needed} and @option{--no-as-needed}
1586 options. They have been replaced by @option{--copy-dt-needed-entries}
1587 and @option{--no-copy-dt-needed-entries}.
1589 @kindex -assert @var{keyword}
1590 @item -assert @var{keyword}
1591 This option is ignored for SunOS compatibility.
1595 @kindex -call_shared
1599 Link against dynamic libraries. This is only meaningful on platforms
1600 for which shared libraries are supported. This option is normally the
1601 default on such platforms. The different variants of this option are
1602 for compatibility with various systems. You may use this option
1603 multiple times on the command line: it affects library searching for
1604 @option{-l} options which follow it.
1608 Set the @code{DF_1_GROUP} flag in the @code{DT_FLAGS_1} entry in the dynamic
1609 section. This causes the runtime linker to handle lookups in this
1610 object and its dependencies to be performed only inside the group.
1611 @option{--unresolved-symbols=report-all} is implied. This option is
1612 only meaningful on ELF platforms which support shared libraries.
1622 Do not link against shared libraries. This is only meaningful on
1623 platforms for which shared libraries are supported. The different
1624 variants of this option are for compatibility with various systems. You
1625 may use this option multiple times on the command line: it affects
1626 library searching for @option{-l} options which follow it. This
1627 option also implies @option{--unresolved-symbols=report-all}. This
1628 option can be used with @option{-shared}. Doing so means that a
1629 shared library is being created but that all of the library's external
1630 references must be resolved by pulling in entries from static
1635 When creating a shared library, bind references to global symbols to the
1636 definition within the shared library, if any. Normally, it is possible
1637 for a program linked against a shared library to override the definition
1638 within the shared library. This option is only meaningful on ELF
1639 platforms which support shared libraries.
1641 @kindex -Bsymbolic-functions
1642 @item -Bsymbolic-functions
1643 When creating a shared library, bind references to global function
1644 symbols to the definition within the shared library, if any.
1645 This option is only meaningful on ELF platforms which support shared
1648 @kindex -Bno-symbolic
1650 This option can cancel previously specified @samp{-Bsymbolic} and
1651 @samp{-Bsymbolic-functions}.
1653 @kindex --dynamic-list=@var{dynamic-list-file}
1654 @item --dynamic-list=@var{dynamic-list-file}
1655 Specify the name of a dynamic list file to the linker. This is
1656 typically used when creating shared libraries to specify a list of
1657 global symbols whose references shouldn't be bound to the definition
1658 within the shared library, or creating dynamically linked executables
1659 to specify a list of symbols which should be added to the symbol table
1660 in the executable. This option is only meaningful on ELF platforms
1661 which support shared libraries.
1663 The format of the dynamic list is the same as the version node without
1664 scope and node name. See @ref{VERSION} for more information.
1666 @kindex --dynamic-list-data
1667 @item --dynamic-list-data
1668 Include all global data symbols to the dynamic list.
1670 @kindex --dynamic-list-cpp-new
1671 @item --dynamic-list-cpp-new
1672 Provide the builtin dynamic list for C++ operator new and delete. It
1673 is mainly useful for building shared libstdc++.
1675 @kindex --dynamic-list-cpp-typeinfo
1676 @item --dynamic-list-cpp-typeinfo
1677 Provide the builtin dynamic list for C++ runtime type identification.
1679 @kindex --check-sections
1680 @kindex --no-check-sections
1681 @item --check-sections
1682 @itemx --no-check-sections
1683 Asks the linker @emph{not} to check section addresses after they have
1684 been assigned to see if there are any overlaps. Normally the linker will
1685 perform this check, and if it finds any overlaps it will produce
1686 suitable error messages. The linker does know about, and does make
1687 allowances for sections in overlays. The default behaviour can be
1688 restored by using the command-line switch @option{--check-sections}.
1689 Section overlap is not usually checked for relocatable links. You can
1690 force checking in that case by using the @option{--check-sections}
1693 @kindex --copy-dt-needed-entries
1694 @kindex --no-copy-dt-needed-entries
1695 @item --copy-dt-needed-entries
1696 @itemx --no-copy-dt-needed-entries
1697 This option affects the treatment of dynamic libraries referred to
1698 by DT_NEEDED tags @emph{inside} ELF dynamic libraries mentioned on the
1699 command line. Normally the linker won't add a DT_NEEDED tag to the
1700 output binary for each library mentioned in a DT_NEEDED tag in an
1701 input dynamic library. With @option{--copy-dt-needed-entries}
1702 specified on the command line however any dynamic libraries that
1703 follow it will have their DT_NEEDED entries added. The default
1704 behaviour can be restored with @option{--no-copy-dt-needed-entries}.
1706 This option also has an effect on the resolution of symbols in dynamic
1707 libraries. With @option{--copy-dt-needed-entries} dynamic libraries
1708 mentioned on the command line will be recursively searched, following
1709 their DT_NEEDED tags to other libraries, in order to resolve symbols
1710 required by the output binary. With the default setting however
1711 the searching of dynamic libraries that follow it will stop with the
1712 dynamic library itself. No DT_NEEDED links will be traversed to resolve
1715 @cindex cross reference table
1718 Output a cross reference table. If a linker map file is being
1719 generated, the cross reference table is printed to the map file.
1720 Otherwise, it is printed on the standard output.
1722 The format of the table is intentionally simple, so that it may be
1723 easily processed by a script if necessary. The symbols are printed out,
1724 sorted by name. For each symbol, a list of file names is given. If the
1725 symbol is defined, the first file listed is the location of the
1726 definition. If the symbol is defined as a common value then any files
1727 where this happens appear next. Finally any files that reference the
1730 @cindex ctf variables
1731 @kindex --ctf-variables
1732 @kindex --no-ctf-variables
1733 @item --ctf-variables
1734 @item --no-ctf-variables
1735 The CTF debuginfo format supports a section which encodes the names and
1736 types of variables found in the program which do not appear in any symbol
1737 table. These variables clearly cannot be looked up by address by
1738 conventional debuggers, so the space used for their types and names is
1739 usually wasted: the types are usually small but the names are often not.
1740 @option{--ctf-variables} causes the generation of such a section.
1741 The default behaviour can be restored with @option{--no-ctf-variables}.
1743 @cindex ctf type sharing
1744 @kindex --ctf-share-types
1745 @item --ctf-share-types=@var{method}
1746 Adjust the method used to share types between translation units in CTF.
1749 @item share-unconflicted
1750 Put all types that do not have ambiguous definitions into the shared dictionary,
1751 where debuggers can easily access them, even if they only occur in one
1752 translation unit. This is the default.
1754 @item share-duplicated
1755 Put only types that occur in multiple translation units into the shared
1756 dictionary: types with only one definition go into per-translation-unit
1757 dictionaries. Types with ambiguous definitions in multiple translation units
1758 always go into per-translation-unit dictionaries. This tends to make the CTF
1759 larger, but may reduce the amount of CTF in the shared dictionary. For very
1760 large projects this may speed up opening the CTF and save memory in the CTF
1761 consumer at runtime.
1764 @cindex common allocation
1765 @kindex --no-define-common
1766 @item --no-define-common
1767 This option inhibits the assignment of addresses to common symbols.
1768 The script command @code{INHIBIT_COMMON_ALLOCATION} has the same effect.
1769 @xref{Miscellaneous Commands}.
1771 The @samp{--no-define-common} option allows decoupling
1772 the decision to assign addresses to Common symbols from the choice
1773 of the output file type; otherwise a non-Relocatable output type
1774 forces assigning addresses to Common symbols.
1775 Using @samp{--no-define-common} allows Common symbols that are referenced
1776 from a shared library to be assigned addresses only in the main program.
1777 This eliminates the unused duplicate space in the shared library,
1778 and also prevents any possible confusion over resolving to the wrong
1779 duplicate when there are many dynamic modules with specialized search
1780 paths for runtime symbol resolution.
1782 @cindex group allocation in linker script
1783 @cindex section groups
1785 @kindex --force-group-allocation
1786 @item --force-group-allocation
1787 This option causes the linker to place section group members like
1788 normal input sections, and to delete the section groups. This is the
1789 default behaviour for a final link but this option can be used to
1790 change the behaviour of a relocatable link (@samp{-r}). The script
1791 command @code{FORCE_GROUP_ALLOCATION} has the same
1792 effect. @xref{Miscellaneous Commands}.
1794 @cindex symbols, from command line
1795 @kindex --defsym=@var{symbol}=@var{exp}
1796 @item --defsym=@var{symbol}=@var{expression}
1797 Create a global symbol in the output file, containing the absolute
1798 address given by @var{expression}. You may use this option as many
1799 times as necessary to define multiple symbols in the command line. A
1800 limited form of arithmetic is supported for the @var{expression} in this
1801 context: you may give a hexadecimal constant or the name of an existing
1802 symbol, or use @code{+} and @code{-} to add or subtract hexadecimal
1803 constants or symbols. If you need more elaborate expressions, consider
1804 using the linker command language from a script (@pxref{Assignments}).
1805 @emph{Note:} there should be no white space between @var{symbol}, the
1806 equals sign (``@key{=}''), and @var{expression}.
1808 The linker processes @samp{--defsym} arguments and @samp{-T} arguments
1809 in order, placing @samp{--defsym} before @samp{-T} will define the
1810 symbol before the linker script from @samp{-T} is processed, while
1811 placing @samp{--defsym} after @samp{-T} will define the symbol after
1812 the linker script has been processed. This difference has
1813 consequences for expressions within the linker script that use the
1814 @samp{--defsym} symbols, which order is correct will depend on what
1815 you are trying to achieve.
1817 @cindex demangling, from command line
1818 @kindex --demangle[=@var{style}]
1819 @kindex --no-demangle
1820 @item --demangle[=@var{style}]
1821 @itemx --no-demangle
1822 These options control whether to demangle symbol names in error messages
1823 and other output. When the linker is told to demangle, it tries to
1824 present symbol names in a readable fashion: it strips leading
1825 underscores if they are used by the object file format, and converts C++
1826 mangled symbol names into user readable names. Different compilers have
1827 different mangling styles. The optional demangling style argument can be used
1828 to choose an appropriate demangling style for your compiler. The linker will
1829 demangle by default unless the environment variable @samp{COLLECT_NO_DEMANGLE}
1830 is set. These options may be used to override the default.
1832 @cindex dynamic linker, from command line
1833 @kindex -I@var{file}
1834 @kindex --dynamic-linker=@var{file}
1836 @itemx --dynamic-linker=@var{file}
1837 Set the name of the dynamic linker. This is only meaningful when
1838 generating dynamically linked ELF executables. The default dynamic
1839 linker is normally correct; don't use this unless you know what you are
1842 @kindex --no-dynamic-linker
1843 @item --no-dynamic-linker
1844 When producing an executable file, omit the request for a dynamic
1845 linker to be used at load-time. This is only meaningful for ELF
1846 executables that contain dynamic relocations, and usually requires
1847 entry point code that is capable of processing these relocations.
1849 @kindex --embedded-relocs
1850 @item --embedded-relocs
1851 This option is similar to the @option{--emit-relocs} option except
1852 that the relocs are stored in a target-specific section. This option
1853 is only supported by the @samp{BFIN}, @samp{CR16} and @emph{M68K}
1856 @kindex --disable-multiple-abs-defs
1857 @item --disable-multiple-abs-defs
1858 Do not allow multiple definitions with symbols included
1859 in filename invoked by -R or --just-symbols
1861 @kindex --fatal-warnings
1862 @kindex --no-fatal-warnings
1863 @item --fatal-warnings
1864 @itemx --no-fatal-warnings
1865 Treat all warnings as errors. The default behaviour can be restored
1866 with the option @option{--no-fatal-warnings}.
1868 @kindex --force-exe-suffix
1869 @item --force-exe-suffix
1870 Make sure that an output file has a .exe suffix.
1872 If a successfully built fully linked output file does not have a
1873 @code{.exe} or @code{.dll} suffix, this option forces the linker to copy
1874 the output file to one of the same name with a @code{.exe} suffix. This
1875 option is useful when using unmodified Unix makefiles on a Microsoft
1876 Windows host, since some versions of Windows won't run an image unless
1877 it ends in a @code{.exe} suffix.
1879 @kindex --gc-sections
1880 @kindex --no-gc-sections
1881 @cindex garbage collection
1883 @itemx --no-gc-sections
1884 Enable garbage collection of unused input sections. It is ignored on
1885 targets that do not support this option. The default behaviour (of not
1886 performing this garbage collection) can be restored by specifying
1887 @samp{--no-gc-sections} on the command line. Note that garbage
1888 collection for COFF and PE format targets is supported, but the
1889 implementation is currently considered to be experimental.
1891 @samp{--gc-sections} decides which input sections are used by
1892 examining symbols and relocations. The section containing the entry
1893 symbol and all sections containing symbols undefined on the
1894 command-line will be kept, as will sections containing symbols
1895 referenced by dynamic objects. Note that when building shared
1896 libraries, the linker must assume that any visible symbol is
1897 referenced. Once this initial set of sections has been determined,
1898 the linker recursively marks as used any section referenced by their
1899 relocations. See @samp{--entry}, @samp{--undefined}, and
1900 @samp{--gc-keep-exported}.
1902 This option can be set when doing a partial link (enabled with option
1903 @samp{-r}). In this case the root of symbols kept must be explicitly
1904 specified either by one of the options @samp{--entry},
1905 @samp{--undefined}, or @samp{--gc-keep-exported} or by a @code{ENTRY}
1906 command in the linker script.
1908 As a GNU extension, ELF input sections marked with the
1909 @code{SHF_GNU_RETAIN} flag will not be garbage collected.
1911 @kindex --print-gc-sections
1912 @kindex --no-print-gc-sections
1913 @cindex garbage collection
1914 @item --print-gc-sections
1915 @itemx --no-print-gc-sections
1916 List all sections removed by garbage collection. The listing is
1917 printed on stderr. This option is only effective if garbage
1918 collection has been enabled via the @samp{--gc-sections}) option. The
1919 default behaviour (of not listing the sections that are removed) can
1920 be restored by specifying @samp{--no-print-gc-sections} on the command
1923 @kindex --gc-keep-exported
1924 @cindex garbage collection
1925 @item --gc-keep-exported
1926 When @samp{--gc-sections} is enabled, this option prevents garbage
1927 collection of unused input sections that contain global symbols having
1928 default or protected visibility. This option is intended to be used for
1929 executables where unreferenced sections would otherwise be garbage
1930 collected regardless of the external visibility of contained symbols.
1931 Note that this option has no effect when linking shared objects since
1932 it is already the default behaviour. This option is only supported for
1935 @kindex --print-output-format
1936 @cindex output format
1937 @item --print-output-format
1938 Print the name of the default output format (perhaps influenced by
1939 other command-line options). This is the string that would appear
1940 in an @code{OUTPUT_FORMAT} linker script command (@pxref{File Commands}).
1942 @kindex --print-memory-usage
1943 @cindex memory usage
1944 @item --print-memory-usage
1945 Print used size, total size and used size of memory regions created with
1946 the @ref{MEMORY} command. This is useful on embedded targets to have a
1947 quick view of amount of free memory. The format of the output has one
1948 headline and one line per region. It is both human readable and easily
1949 parsable by tools. Here is an example of an output:
1952 Memory region Used Size Region Size %age Used
1953 ROM: 256 KB 1 MB 25.00%
1954 RAM: 32 B 2 GB 0.00%
1961 Print a summary of the command-line options on the standard output and exit.
1963 @kindex --target-help
1965 Print a summary of all target-specific options on the standard output and exit.
1967 @kindex -Map=@var{mapfile}
1968 @item -Map=@var{mapfile}
1969 Print a link map to the file @var{mapfile}. See the description of the
1970 @option{-M} option, above. If @var{mapfile} is just the character
1971 @code{-} then the map will be written to stdout.
1973 Specifying a directory as @var{mapfile} causes the linker map to be
1974 written as a file inside the directory. Normally name of the file
1975 inside the directory is computed as the basename of the @var{output}
1976 file with @code{.map} appended. If however the special character
1977 @code{%} is used then this will be replaced by the full path of the
1978 output file. Additionally if there are any characters after the
1979 @var{%} symbol then @code{.map} will no longer be appended.
1982 -o foo.exe -Map=bar [Creates ./bar]
1983 -o ../dir/foo.exe -Map=bar [Creates ./bar]
1984 -o foo.exe -Map=../dir [Creates ../dir/foo.exe.map]
1985 -o ../dir2/foo.exe -Map=../dir [Creates ../dir/foo.exe.map]
1986 -o foo.exe -Map=% [Creates ./foo.exe.map]
1987 -o ../dir/foo.exe -Map=% [Creates ../dir/foo.exe.map]
1988 -o foo.exe -Map=%.bar [Creates ./foo.exe.bar]
1989 -o ../dir/foo.exe -Map=%.bar [Creates ../dir/foo.exe.bar]
1990 -o ../dir2/foo.exe -Map=../dir/% [Creates ../dir/../dir2/foo.exe.map]
1991 -o ../dir2/foo.exe -Map=../dir/%.bar [Creates ../dir/../dir2/foo.exe.bar]
1994 It is an error to specify more than one @code{%} character.
1996 If the map file already exists then it will be overwritten by this
1999 @cindex memory usage
2000 @kindex --no-keep-memory
2001 @item --no-keep-memory
2002 @command{ld} normally optimizes for speed over memory usage by caching the
2003 symbol tables of input files in memory. This option tells @command{ld} to
2004 instead optimize for memory usage, by rereading the symbol tables as
2005 necessary. This may be required if @command{ld} runs out of memory space
2006 while linking a large executable.
2008 @kindex --no-undefined
2011 @item --no-undefined
2013 Report unresolved symbol references from regular object files. This
2014 is done even if the linker is creating a non-symbolic shared library.
2015 The switch @option{--[no-]allow-shlib-undefined} controls the
2016 behaviour for reporting unresolved references found in shared
2017 libraries being linked in.
2019 The effects of this option can be reverted by using @code{-z undefs}.
2021 @kindex --allow-multiple-definition
2023 @item --allow-multiple-definition
2025 Normally when a symbol is defined multiple times, the linker will
2026 report a fatal error. These options allow multiple definitions and the
2027 first definition will be used.
2029 @kindex --allow-shlib-undefined
2030 @kindex --no-allow-shlib-undefined
2031 @item --allow-shlib-undefined
2032 @itemx --no-allow-shlib-undefined
2033 Allows or disallows undefined symbols in shared libraries.
2034 This switch is similar to @option{--no-undefined} except that it
2035 determines the behaviour when the undefined symbols are in a
2036 shared library rather than a regular object file. It does not affect
2037 how undefined symbols in regular object files are handled.
2039 The default behaviour is to report errors for any undefined symbols
2040 referenced in shared libraries if the linker is being used to create
2041 an executable, but to allow them if the linker is being used to create
2044 The reasons for allowing undefined symbol references in shared
2045 libraries specified at link time are that:
2049 A shared library specified at link time may not be the same as the one
2050 that is available at load time, so the symbol might actually be
2051 resolvable at load time.
2053 There are some operating systems, eg BeOS and HPPA, where undefined
2054 symbols in shared libraries are normal.
2056 The BeOS kernel for example patches shared libraries at load time to
2057 select whichever function is most appropriate for the current
2058 architecture. This is used, for example, to dynamically select an
2059 appropriate memset function.
2062 @kindex --error-handling-script=@var{scriptname}
2063 @item --error-handling-script=@var{scriptname}
2064 If this option is provided then the linker will invoke
2065 @var{scriptname} whenever an error is encountered. Currently however
2066 only two kinds of error are supported: missing symbols and missing
2067 libraries. Two arguments will be passed to script: the keyword
2068 ``undefined-symbol'' or `missing-lib'' and the @var{name} of the
2069 undefined symbol or missing library. The intention is that the script
2070 will provide suggestions to the user as to where the symbol or library
2071 might be found. After the script has finished then the normal linker
2072 error message will be displayed.
2074 The availability of this option is controlled by a configure time
2075 switch, so it may not be present in specific implementations.
2077 @kindex --no-undefined-version
2078 @item --no-undefined-version
2079 Normally when a symbol has an undefined version, the linker will ignore
2080 it. This option disallows symbols with undefined version and a fatal error
2081 will be issued instead.
2083 @kindex --default-symver
2084 @item --default-symver
2085 Create and use a default symbol version (the soname) for unversioned
2088 @kindex --default-imported-symver
2089 @item --default-imported-symver
2090 Create and use a default symbol version (the soname) for unversioned
2093 @kindex --no-warn-mismatch
2094 @item --no-warn-mismatch
2095 Normally @command{ld} will give an error if you try to link together input
2096 files that are mismatched for some reason, perhaps because they have
2097 been compiled for different processors or for different endiannesses.
2098 This option tells @command{ld} that it should silently permit such possible
2099 errors. This option should only be used with care, in cases when you
2100 have taken some special action that ensures that the linker errors are
2103 @kindex --no-warn-search-mismatch
2104 @item --no-warn-search-mismatch
2105 Normally @command{ld} will give a warning if it finds an incompatible
2106 library during a library search. This option silences the warning.
2108 @kindex --no-whole-archive
2109 @item --no-whole-archive
2110 Turn off the effect of the @option{--whole-archive} option for subsequent
2113 @cindex output file after errors
2114 @kindex --noinhibit-exec
2115 @item --noinhibit-exec
2116 Retain the executable output file whenever it is still usable.
2117 Normally, the linker will not produce an output file if it encounters
2118 errors during the link process; it exits without writing an output file
2119 when it issues any error whatsoever.
2123 Only search library directories explicitly specified on the
2124 command line. Library directories specified in linker scripts
2125 (including linker scripts specified on the command line) are ignored.
2127 @ifclear SingleFormat
2128 @kindex --oformat=@var{output-format}
2129 @item --oformat=@var{output-format}
2130 @command{ld} may be configured to support more than one kind of object
2131 file. If your @command{ld} is configured this way, you can use the
2132 @samp{--oformat} option to specify the binary format for the output
2133 object file. Even when @command{ld} is configured to support alternative
2134 object formats, you don't usually need to specify this, as @command{ld}
2135 should be configured to produce as a default output format the most
2136 usual format on each machine. @var{output-format} is a text string, the
2137 name of a particular format supported by the BFD libraries. (You can
2138 list the available binary formats with @samp{objdump -i}.) The script
2139 command @code{OUTPUT_FORMAT} can also specify the output format, but
2140 this option overrides it. @xref{BFD}.
2143 @kindex --out-implib
2144 @item --out-implib @var{file}
2145 Create an import library in @var{file} corresponding to the executable
2146 the linker is generating (eg. a DLL or ELF program). This import
2147 library (which should be called @code{*.dll.a} or @code{*.a} for DLLs)
2148 may be used to link clients against the generated executable; this
2149 behaviour makes it possible to skip a separate import library creation
2150 step (eg. @code{dlltool} for DLLs). This option is only available for
2151 the i386 PE and ELF targetted ports of the linker.
2154 @kindex --pic-executable
2156 @itemx --pic-executable
2157 @cindex position independent executables
2158 Create a position independent executable. This is currently only supported on
2159 ELF platforms. Position independent executables are similar to shared
2160 libraries in that they are relocated by the dynamic linker to the virtual
2161 address the OS chooses for them (which can vary between invocations). Like
2162 normal dynamically linked executables they can be executed and symbols
2163 defined in the executable cannot be overridden by shared libraries.
2167 @cindex position dependent executables
2168 Create a position dependent executable. This is the default.
2172 This option is ignored for Linux compatibility.
2176 This option is ignored for SVR4 compatibility.
2179 @cindex synthesizing linker
2180 @cindex relaxing addressing modes
2184 An option with machine dependent effects.
2186 This option is only supported on a few targets.
2189 @xref{H8/300,,@command{ld} and the H8/300}.
2192 @xref{Xtensa,, @command{ld} and Xtensa Processors}.
2195 @xref{M68HC11/68HC12,,@command{ld} and the 68HC11 and 68HC12}.
2198 @xref{Nios II,,@command{ld} and the Altera Nios II}.
2201 @xref{PowerPC ELF32,,@command{ld} and PowerPC 32-bit ELF Support}.
2204 On some platforms the @option{--relax} option performs target specific,
2205 global optimizations that become possible when the linker resolves
2206 addressing in the program, such as relaxing address modes,
2207 synthesizing new instructions, selecting shorter version of current
2208 instructions, and combining constant values.
2210 On some platforms these link time global optimizations may make symbolic
2211 debugging of the resulting executable impossible.
2213 This is known to be the case for the Matsushita MN10200 and MN10300
2214 family of processors.
2217 On platforms where the feature is supported, the option
2218 @option{--no-relax} will disable it.
2220 On platforms where the feature is not supported, both @option{--relax}
2221 and @option{--no-relax} are accepted, but ignored.
2223 @cindex retaining specified symbols
2224 @cindex stripping all but some symbols
2225 @cindex symbols, retaining selectively
2226 @kindex --retain-symbols-file=@var{filename}
2227 @item --retain-symbols-file=@var{filename}
2228 Retain @emph{only} the symbols listed in the file @var{filename},
2229 discarding all others. @var{filename} is simply a flat file, with one
2230 symbol name per line. This option is especially useful in environments
2234 where a large global symbol table is accumulated gradually, to conserve
2237 @samp{--retain-symbols-file} does @emph{not} discard undefined symbols,
2238 or symbols needed for relocations.
2240 You may only specify @samp{--retain-symbols-file} once in the command
2241 line. It overrides @samp{-s} and @samp{-S}.
2244 @item -rpath=@var{dir}
2245 @cindex runtime library search path
2246 @kindex -rpath=@var{dir}
2247 Add a directory to the runtime library search path. This is used when
2248 linking an ELF executable with shared objects. All @option{-rpath}
2249 arguments are concatenated and passed to the runtime linker, which uses
2250 them to locate shared objects at runtime.
2252 The @option{-rpath} option is also used when locating shared objects which
2253 are needed by shared objects explicitly included in the link; see the
2254 description of the @option{-rpath-link} option. Searching @option{-rpath}
2255 in this way is only supported by native linkers and cross linkers which
2256 have been configured with the @option{--with-sysroot} option.
2258 If @option{-rpath} is not used when linking an ELF executable, the
2259 contents of the environment variable @code{LD_RUN_PATH} will be used if it
2262 The @option{-rpath} option may also be used on SunOS. By default, on
2263 SunOS, the linker will form a runtime search path out of all the
2264 @option{-L} options it is given. If a @option{-rpath} option is used, the
2265 runtime search path will be formed exclusively using the @option{-rpath}
2266 options, ignoring the @option{-L} options. This can be useful when using
2267 gcc, which adds many @option{-L} options which may be on NFS mounted
2270 For compatibility with other ELF linkers, if the @option{-R} option is
2271 followed by a directory name, rather than a file name, it is treated as
2272 the @option{-rpath} option.
2276 @cindex link-time runtime library search path
2277 @kindex -rpath-link=@var{dir}
2278 @item -rpath-link=@var{dir}
2279 When using ELF or SunOS, one shared library may require another. This
2280 happens when an @code{ld -shared} link includes a shared library as one
2283 When the linker encounters such a dependency when doing a non-shared,
2284 non-relocatable link, it will automatically try to locate the required
2285 shared library and include it in the link, if it is not included
2286 explicitly. In such a case, the @option{-rpath-link} option
2287 specifies the first set of directories to search. The
2288 @option{-rpath-link} option may specify a sequence of directory names
2289 either by specifying a list of names separated by colons, or by
2290 appearing multiple times.
2292 The tokens @var{$ORIGIN} and @var{$LIB} can appear in these search
2293 directories. They will be replaced by the full path to the directory
2294 containing the program or shared object in the case of @var{$ORIGIN}
2295 and either @samp{lib} - for 32-bit binaries - or @samp{lib64} - for
2296 64-bit binaries - in the case of @var{$LIB}.
2298 The alternative form of these tokens - @var{$@{ORIGIN@}} and
2299 @var{$@{LIB@}} can also be used. The token @var{$PLATFORM} is not
2302 This option should be used with caution as it overrides the search path
2303 that may have been hard compiled into a shared library. In such a case it
2304 is possible to use unintentionally a different search path than the
2305 runtime linker would do.
2307 The linker uses the following search paths to locate required shared
2312 Any directories specified by @option{-rpath-link} options.
2314 Any directories specified by @option{-rpath} options. The difference
2315 between @option{-rpath} and @option{-rpath-link} is that directories
2316 specified by @option{-rpath} options are included in the executable and
2317 used at runtime, whereas the @option{-rpath-link} option is only effective
2318 at link time. Searching @option{-rpath} in this way is only supported
2319 by native linkers and cross linkers which have been configured with
2320 the @option{--with-sysroot} option.
2322 On an ELF system, for native linkers, if the @option{-rpath} and
2323 @option{-rpath-link} options were not used, search the contents of the
2324 environment variable @code{LD_RUN_PATH}.
2326 On SunOS, if the @option{-rpath} option was not used, search any
2327 directories specified using @option{-L} options.
2329 For a native linker, search the contents of the environment
2330 variable @code{LD_LIBRARY_PATH}.
2332 For a native ELF linker, the directories in @code{DT_RUNPATH} or
2333 @code{DT_RPATH} of a shared library are searched for shared
2334 libraries needed by it. The @code{DT_RPATH} entries are ignored if
2335 @code{DT_RUNPATH} entries exist.
2337 For a linker for a Linux system, if the file @file{/etc/ld.so.conf}
2338 exists, the list of directories found in that file. Note: the path
2339 to this file is prefixed with the @code{sysroot} value, if that is
2340 defined, and then any @code{prefix} string if the linker was
2341 configured with the @command{--prefix=<path>} option.
2343 For a native linker on a FreeBSD system, any directories specified by
2344 the @code{_PATH_ELF_HINTS} macro defined in the @file{elf-hints.h}
2347 Any directories specified by a @code{SEARCH_DIR} command in a
2348 linker script given on the command line, including scripts specified
2349 by @option{-T} (but not @option{-dT}).
2351 The default directories, normally @file{/lib} and @file{/usr/lib}.
2353 Any directories specified by a plugin LDPT_SET_EXTRA_LIBRARY_PATH.
2355 Any directories specified by a @code{SEARCH_DIR} command in a default
2359 Note however on Linux based systems there is an additional caveat: If
2360 the @option{--as-needed} option is active @emph{and} a shared library
2361 is located which would normally satisfy the search @emph{and} this
2362 library does not have DT_NEEDED tag for @file{libc.so}
2363 @emph{and} there is a shared library later on in the set of search
2364 directories which also satisfies the search @emph{and}
2365 this second shared library does have a DT_NEEDED tag for
2366 @file{libc.so} @emph{then} the second library will be selected instead
2369 If the required shared library is not found, the linker will issue a
2370 warning and continue with the link.
2378 @cindex shared libraries
2379 Create a shared library. This is currently only supported on ELF, XCOFF
2380 and SunOS platforms. On SunOS, the linker will automatically create a
2381 shared library if the @option{-e} option is not used and there are
2382 undefined symbols in the link.
2384 @kindex --sort-common
2386 @itemx --sort-common=ascending
2387 @itemx --sort-common=descending
2388 This option tells @command{ld} to sort the common symbols by alignment in
2389 ascending or descending order when it places them in the appropriate output
2390 sections. The symbol alignments considered are sixteen-byte or larger,
2391 eight-byte, four-byte, two-byte, and one-byte. This is to prevent gaps
2392 between symbols due to alignment constraints. If no sorting order is
2393 specified, then descending order is assumed.
2395 @kindex --sort-section=name
2396 @item --sort-section=name
2397 This option will apply @code{SORT_BY_NAME} to all wildcard section
2398 patterns in the linker script.
2400 @kindex --sort-section=alignment
2401 @item --sort-section=alignment
2402 This option will apply @code{SORT_BY_ALIGNMENT} to all wildcard section
2403 patterns in the linker script.
2405 @kindex --spare-dynamic-tags
2406 @item --spare-dynamic-tags=@var{count}
2407 This option specifies the number of empty slots to leave in the
2408 .dynamic section of ELF shared objects. Empty slots may be needed by
2409 post processing tools, such as the prelinker. The default is 5.
2411 @kindex --split-by-file
2412 @item --split-by-file[=@var{size}]
2413 Similar to @option{--split-by-reloc} but creates a new output section for
2414 each input file when @var{size} is reached. @var{size} defaults to a
2415 size of 1 if not given.
2417 @kindex --split-by-reloc
2418 @item --split-by-reloc[=@var{count}]
2419 Tries to creates extra sections in the output file so that no single
2420 output section in the file contains more than @var{count} relocations.
2421 This is useful when generating huge relocatable files for downloading into
2422 certain real time kernels with the COFF object file format; since COFF
2423 cannot represent more than 65535 relocations in a single section. Note
2424 that this will fail to work with object file formats which do not
2425 support arbitrary sections. The linker will not split up individual
2426 input sections for redistribution, so if a single input section contains
2427 more than @var{count} relocations one output section will contain that
2428 many relocations. @var{count} defaults to a value of 32768.
2432 Compute and display statistics about the operation of the linker, such
2433 as execution time and memory usage.
2435 @kindex --sysroot=@var{directory}
2436 @item --sysroot=@var{directory}
2437 Use @var{directory} as the location of the sysroot, overriding the
2438 configure-time default. This option is only supported by linkers
2439 that were configured using @option{--with-sysroot}.
2443 This is used by COFF/PE based targets to create a task-linked object
2444 file where all of the global symbols have been converted to statics.
2446 @kindex --traditional-format
2447 @cindex traditional format
2448 @item --traditional-format
2449 For some targets, the output of @command{ld} is different in some ways from
2450 the output of some existing linker. This switch requests @command{ld} to
2451 use the traditional format instead.
2454 For example, on SunOS, @command{ld} combines duplicate entries in the
2455 symbol string table. This can reduce the size of an output file with
2456 full debugging information by over 30 percent. Unfortunately, the SunOS
2457 @code{dbx} program can not read the resulting program (@code{gdb} has no
2458 trouble). The @samp{--traditional-format} switch tells @command{ld} to not
2459 combine duplicate entries.
2461 @kindex --section-start=@var{sectionname}=@var{org}
2462 @item --section-start=@var{sectionname}=@var{org}
2463 Locate a section in the output file at the absolute
2464 address given by @var{org}. You may use this option as many
2465 times as necessary to locate multiple sections in the command
2467 @var{org} must be a single hexadecimal integer;
2468 for compatibility with other linkers, you may omit the leading
2469 @samp{0x} usually associated with hexadecimal values. @emph{Note:} there
2470 should be no white space between @var{sectionname}, the equals
2471 sign (``@key{=}''), and @var{org}.
2473 @kindex -Tbss=@var{org}
2474 @kindex -Tdata=@var{org}
2475 @kindex -Ttext=@var{org}
2476 @cindex segment origins, cmd line
2477 @item -Tbss=@var{org}
2478 @itemx -Tdata=@var{org}
2479 @itemx -Ttext=@var{org}
2480 Same as @option{--section-start}, with @code{.bss}, @code{.data} or
2481 @code{.text} as the @var{sectionname}.
2483 @kindex -Ttext-segment=@var{org}
2484 @item -Ttext-segment=@var{org}
2485 @cindex text segment origin, cmd line
2486 When creating an ELF executable, it will set the address of the first
2487 byte of the text segment.
2489 @kindex -Trodata-segment=@var{org}
2490 @item -Trodata-segment=@var{org}
2491 @cindex rodata segment origin, cmd line
2492 When creating an ELF executable or shared object for a target where
2493 the read-only data is in its own segment separate from the executable
2494 text, it will set the address of the first byte of the read-only data segment.
2496 @kindex -Tldata-segment=@var{org}
2497 @item -Tldata-segment=@var{org}
2498 @cindex ldata segment origin, cmd line
2499 When creating an ELF executable or shared object for x86-64 medium memory
2500 model, it will set the address of the first byte of the ldata segment.
2502 @kindex --unresolved-symbols
2503 @item --unresolved-symbols=@var{method}
2504 Determine how to handle unresolved symbols. There are four possible
2505 values for @samp{method}:
2509 Do not report any unresolved symbols.
2512 Report all unresolved symbols. This is the default.
2514 @item ignore-in-object-files
2515 Report unresolved symbols that are contained in shared libraries, but
2516 ignore them if they come from regular object files.
2518 @item ignore-in-shared-libs
2519 Report unresolved symbols that come from regular object files, but
2520 ignore them if they come from shared libraries. This can be useful
2521 when creating a dynamic binary and it is known that all the shared
2522 libraries that it should be referencing are included on the linker's
2526 The behaviour for shared libraries on their own can also be controlled
2527 by the @option{--[no-]allow-shlib-undefined} option.
2529 Normally the linker will generate an error message for each reported
2530 unresolved symbol but the option @option{--warn-unresolved-symbols}
2531 can change this to a warning.
2533 @kindex --verbose[=@var{NUMBER}]
2534 @cindex verbose[=@var{NUMBER}]
2536 @itemx --verbose[=@var{NUMBER}]
2537 Display the version number for @command{ld} and list the linker emulations
2538 supported. Display which input files can and cannot be opened. Display
2539 the linker script being used by the linker. If the optional @var{NUMBER}
2540 argument > 1, plugin symbol status will also be displayed.
2542 @kindex --version-script=@var{version-scriptfile}
2543 @cindex version script, symbol versions
2544 @item --version-script=@var{version-scriptfile}
2545 Specify the name of a version script to the linker. This is typically
2546 used when creating shared libraries to specify additional information
2547 about the version hierarchy for the library being created. This option
2548 is only fully supported on ELF platforms which support shared libraries;
2549 see @ref{VERSION}. It is partially supported on PE platforms, which can
2550 use version scripts to filter symbol visibility in auto-export mode: any
2551 symbols marked @samp{local} in the version script will not be exported.
2554 @kindex --warn-common
2555 @cindex warnings, on combining symbols
2556 @cindex combining symbols, warnings on
2558 Warn when a common symbol is combined with another common symbol or with
2559 a symbol definition. Unix linkers allow this somewhat sloppy practice,
2560 but linkers on some other operating systems do not. This option allows
2561 you to find potential problems from combining global symbols.
2562 Unfortunately, some C libraries use this practice, so you may get some
2563 warnings about symbols in the libraries as well as in your programs.
2565 There are three kinds of global symbols, illustrated here by C examples:
2569 A definition, which goes in the initialized data section of the output
2573 An undefined reference, which does not allocate space.
2574 There must be either a definition or a common symbol for the
2578 A common symbol. If there are only (one or more) common symbols for a
2579 variable, it goes in the uninitialized data area of the output file.
2580 The linker merges multiple common symbols for the same variable into a
2581 single symbol. If they are of different sizes, it picks the largest
2582 size. The linker turns a common symbol into a declaration, if there is
2583 a definition of the same variable.
2586 The @samp{--warn-common} option can produce five kinds of warnings.
2587 Each warning consists of a pair of lines: the first describes the symbol
2588 just encountered, and the second describes the previous symbol
2589 encountered with the same name. One or both of the two symbols will be
2594 Turning a common symbol into a reference, because there is already a
2595 definition for the symbol.
2597 @var{file}(@var{section}): warning: common of `@var{symbol}'
2598 overridden by definition
2599 @var{file}(@var{section}): warning: defined here
2603 Turning a common symbol into a reference, because a later definition for
2604 the symbol is encountered. This is the same as the previous case,
2605 except that the symbols are encountered in a different order.
2607 @var{file}(@var{section}): warning: definition of `@var{symbol}'
2609 @var{file}(@var{section}): warning: common is here
2613 Merging a common symbol with a previous same-sized common symbol.
2615 @var{file}(@var{section}): warning: multiple common
2617 @var{file}(@var{section}): warning: previous common is here
2621 Merging a common symbol with a previous larger common symbol.
2623 @var{file}(@var{section}): warning: common of `@var{symbol}'
2624 overridden by larger common
2625 @var{file}(@var{section}): warning: larger common is here
2629 Merging a common symbol with a previous smaller common symbol. This is
2630 the same as the previous case, except that the symbols are
2631 encountered in a different order.
2633 @var{file}(@var{section}): warning: common of `@var{symbol}'
2634 overriding smaller common
2635 @var{file}(@var{section}): warning: smaller common is here
2639 @kindex --warn-constructors
2640 @item --warn-constructors
2641 Warn if any global constructors are used. This is only useful for a few
2642 object file formats. For formats like COFF or ELF, the linker can not
2643 detect the use of global constructors.
2645 @kindex --warn-multiple-gp
2646 @item --warn-multiple-gp
2647 Warn if multiple global pointer values are required in the output file.
2648 This is only meaningful for certain processors, such as the Alpha.
2649 Specifically, some processors put large-valued constants in a special
2650 section. A special register (the global pointer) points into the middle
2651 of this section, so that constants can be loaded efficiently via a
2652 base-register relative addressing mode. Since the offset in
2653 base-register relative mode is fixed and relatively small (e.g., 16
2654 bits), this limits the maximum size of the constant pool. Thus, in
2655 large programs, it is often necessary to use multiple global pointer
2656 values in order to be able to address all possible constants. This
2657 option causes a warning to be issued whenever this case occurs.
2660 @cindex warnings, on undefined symbols
2661 @cindex undefined symbols, warnings on
2663 Only warn once for each undefined symbol, rather than once per module
2666 @kindex --warn-section-align
2667 @cindex warnings, on section alignment
2668 @cindex section alignment, warnings on
2669 @item --warn-section-align
2670 Warn if the address of an output section is changed because of
2671 alignment. Typically, the alignment will be set by an input section.
2672 The address will only be changed if it not explicitly specified; that
2673 is, if the @code{SECTIONS} command does not specify a start address for
2674 the section (@pxref{SECTIONS}).
2676 @kindex --warn-textrel
2677 @item --warn-textrel
2678 Warn if the linker adds DT_TEXTREL to a position-independent executable
2681 @kindex --warn-alternate-em
2682 @item --warn-alternate-em
2683 Warn if an object has alternate ELF machine code.
2685 @kindex --warn-unresolved-symbols
2686 @item --warn-unresolved-symbols
2687 If the linker is going to report an unresolved symbol (see the option
2688 @option{--unresolved-symbols}) it will normally generate an error.
2689 This option makes it generate a warning instead.
2691 @kindex --error-unresolved-symbols
2692 @item --error-unresolved-symbols
2693 This restores the linker's default behaviour of generating errors when
2694 it is reporting unresolved symbols.
2696 @kindex --whole-archive
2697 @cindex including an entire archive
2698 @item --whole-archive
2699 For each archive mentioned on the command line after the
2700 @option{--whole-archive} option, include every object file in the archive
2701 in the link, rather than searching the archive for the required object
2702 files. This is normally used to turn an archive file into a shared
2703 library, forcing every object to be included in the resulting shared
2704 library. This option may be used more than once.
2706 Two notes when using this option from gcc: First, gcc doesn't know
2707 about this option, so you have to use @option{-Wl,-whole-archive}.
2708 Second, don't forget to use @option{-Wl,-no-whole-archive} after your
2709 list of archives, because gcc will add its own list of archives to
2710 your link and you may not want this flag to affect those as well.
2712 @kindex --wrap=@var{symbol}
2713 @item --wrap=@var{symbol}
2714 Use a wrapper function for @var{symbol}. Any undefined reference to
2715 @var{symbol} will be resolved to @code{__wrap_@var{symbol}}. Any
2716 undefined reference to @code{__real_@var{symbol}} will be resolved to
2719 This can be used to provide a wrapper for a system function. The
2720 wrapper function should be called @code{__wrap_@var{symbol}}. If it
2721 wishes to call the system function, it should call
2722 @code{__real_@var{symbol}}.
2724 Here is a trivial example:
2728 __wrap_malloc (size_t c)
2730 printf ("malloc called with %zu\n", c);
2731 return __real_malloc (c);
2735 If you link other code with this file using @option{--wrap malloc}, then
2736 all calls to @code{malloc} will call the function @code{__wrap_malloc}
2737 instead. The call to @code{__real_malloc} in @code{__wrap_malloc} will
2738 call the real @code{malloc} function.
2740 You may wish to provide a @code{__real_malloc} function as well, so that
2741 links without the @option{--wrap} option will succeed. If you do this,
2742 you should not put the definition of @code{__real_malloc} in the same
2743 file as @code{__wrap_malloc}; if you do, the assembler may resolve the
2744 call before the linker has a chance to wrap it to @code{malloc}.
2746 Only undefined references are replaced by the linker. So, translation unit
2747 internal references to @var{symbol} are not resolved to
2748 @code{__wrap_@var{symbol}}. In the next example, the call to @code{f} in
2749 @code{g} is not resolved to @code{__wrap_f}.
2765 @kindex --eh-frame-hdr
2766 @kindex --no-eh-frame-hdr
2767 @item --eh-frame-hdr
2768 @itemx --no-eh-frame-hdr
2769 Request (@option{--eh-frame-hdr}) or suppress
2770 (@option{--no-eh-frame-hdr}) the creation of @code{.eh_frame_hdr}
2771 section and ELF @code{PT_GNU_EH_FRAME} segment header.
2773 @kindex --ld-generated-unwind-info
2774 @item --no-ld-generated-unwind-info
2775 Request creation of @code{.eh_frame} unwind info for linker
2776 generated code sections like PLT. This option is on by default
2777 if linker generated unwind info is supported.
2779 @kindex --enable-new-dtags
2780 @kindex --disable-new-dtags
2781 @item --enable-new-dtags
2782 @itemx --disable-new-dtags
2783 This linker can create the new dynamic tags in ELF. But the older ELF
2784 systems may not understand them. If you specify
2785 @option{--enable-new-dtags}, the new dynamic tags will be created as needed
2786 and older dynamic tags will be omitted.
2787 If you specify @option{--disable-new-dtags}, no new dynamic tags will be
2788 created. By default, the new dynamic tags are not created. Note that
2789 those options are only available for ELF systems.
2791 @kindex --hash-size=@var{number}
2792 @item --hash-size=@var{number}
2793 Set the default size of the linker's hash tables to a prime number
2794 close to @var{number}. Increasing this value can reduce the length of
2795 time it takes the linker to perform its tasks, at the expense of
2796 increasing the linker's memory requirements. Similarly reducing this
2797 value can reduce the memory requirements at the expense of speed.
2799 @kindex --hash-style=@var{style}
2800 @item --hash-style=@var{style}
2801 Set the type of linker's hash table(s). @var{style} can be either
2802 @code{sysv} for classic ELF @code{.hash} section, @code{gnu} for
2803 new style GNU @code{.gnu.hash} section or @code{both} for both
2804 the classic ELF @code{.hash} and new style GNU @code{.gnu.hash}
2805 hash tables. The default depends upon how the linker was configured,
2806 but for most Linux based systems it will be @code{both}.
2808 @kindex --compress-debug-sections=none
2809 @kindex --compress-debug-sections=zlib
2810 @kindex --compress-debug-sections=zlib-gnu
2811 @kindex --compress-debug-sections=zlib-gabi
2812 @item --compress-debug-sections=none
2813 @itemx --compress-debug-sections=zlib
2814 @itemx --compress-debug-sections=zlib-gnu
2815 @itemx --compress-debug-sections=zlib-gabi
2816 On ELF platforms, these options control how DWARF debug sections are
2817 compressed using zlib.
2819 @option{--compress-debug-sections=none} doesn't compress DWARF debug
2820 sections. @option{--compress-debug-sections=zlib-gnu} compresses
2821 DWARF debug sections and renames them to begin with @samp{.zdebug}
2822 instead of @samp{.debug}. @option{--compress-debug-sections=zlib-gabi}
2823 also compresses DWARF debug sections, but rather than renaming them it
2824 sets the SHF_COMPRESSED flag in the sections' headers.
2826 The @option{--compress-debug-sections=zlib} option is an alias for
2827 @option{--compress-debug-sections=zlib-gabi}.
2829 Note that this option overrides any compression in input debug
2830 sections, so if a binary is linked with @option{--compress-debug-sections=none}
2831 for example, then any compressed debug sections in input files will be
2832 uncompressed before they are copied into the output binary.
2834 The default compression behaviour varies depending upon the target
2835 involved and the configure options used to build the toolchain. The
2836 default can be determined by examining the output from the linker's
2837 @option{--help} option.
2839 @kindex --reduce-memory-overheads
2840 @item --reduce-memory-overheads
2841 This option reduces memory requirements at ld runtime, at the expense of
2842 linking speed. This was introduced to select the old O(n^2) algorithm
2843 for link map file generation, rather than the new O(n) algorithm which uses
2844 about 40% more memory for symbol storage.
2846 Another effect of the switch is to set the default hash table size to
2847 1021, which again saves memory at the cost of lengthening the linker's
2848 run time. This is not done however if the @option{--hash-size} switch
2851 The @option{--reduce-memory-overheads} switch may be also be used to
2852 enable other tradeoffs in future versions of the linker.
2854 @kindex --max-cache-size=@var{size}
2855 @item --max-cache-size=@var{size}
2856 @command{ld} normally caches the relocation information and symbol tables
2857 of input files in memory with the unlimited size. This option sets the
2858 maximum cache size to @var{size}.
2861 @kindex --build-id=@var{style}
2863 @itemx --build-id=@var{style}
2864 Request the creation of a @code{.note.gnu.build-id} ELF note section
2865 or a @code{.buildid} COFF section. The contents of the note are
2866 unique bits identifying this linked file. @var{style} can be
2867 @code{uuid} to use 128 random bits, @code{sha1} to use a 160-bit
2868 @sc{SHA1} hash on the normative parts of the output contents,
2869 @code{md5} to use a 128-bit @sc{MD5} hash on the normative parts of
2870 the output contents, or @code{0x@var{hexstring}} to use a chosen bit
2871 string specified as an even number of hexadecimal digits (@code{-} and
2872 @code{:} characters between digit pairs are ignored). If @var{style}
2873 is omitted, @code{sha1} is used.
2875 The @code{md5} and @code{sha1} styles produces an identifier
2876 that is always the same in an identical output file, but will be
2877 unique among all nonidentical output files. It is not intended
2878 to be compared as a checksum for the file's contents. A linked
2879 file may be changed later by other tools, but the build ID bit
2880 string identifying the original linked file does not change.
2882 Passing @code{none} for @var{style} disables the setting from any
2883 @code{--build-id} options earlier on the command line.
2888 @subsection Options Specific to i386 PE Targets
2890 @c man begin OPTIONS
2892 The i386 PE linker supports the @option{-shared} option, which causes
2893 the output to be a dynamically linked library (DLL) instead of a
2894 normal executable. You should name the output @code{*.dll} when you
2895 use this option. In addition, the linker fully supports the standard
2896 @code{*.def} files, which may be specified on the linker command line
2897 like an object file (in fact, it should precede archives it exports
2898 symbols from, to ensure that they get linked in, just like a normal
2901 In addition to the options common to all targets, the i386 PE linker
2902 support additional command-line options that are specific to the i386
2903 PE target. Options that take values may be separated from their
2904 values by either a space or an equals sign.
2908 @kindex --add-stdcall-alias
2909 @item --add-stdcall-alias
2910 If given, symbols with a stdcall suffix (@@@var{nn}) will be exported
2911 as-is and also with the suffix stripped.
2912 [This option is specific to the i386 PE targeted port of the linker]
2915 @item --base-file @var{file}
2916 Use @var{file} as the name of a file in which to save the base
2917 addresses of all the relocations needed for generating DLLs with
2919 [This is an i386 PE specific option]
2923 Create a DLL instead of a regular executable. You may also use
2924 @option{-shared} or specify a @code{LIBRARY} in a given @code{.def}
2926 [This option is specific to the i386 PE targeted port of the linker]
2928 @kindex --enable-long-section-names
2929 @kindex --disable-long-section-names
2930 @item --enable-long-section-names
2931 @itemx --disable-long-section-names
2932 The PE variants of the COFF object format add an extension that permits
2933 the use of section names longer than eight characters, the normal limit
2934 for COFF. By default, these names are only allowed in object files, as
2935 fully-linked executable images do not carry the COFF string table required
2936 to support the longer names. As a GNU extension, it is possible to
2937 allow their use in executable images as well, or to (probably pointlessly!)
2938 disallow it in object files, by using these two options. Executable images
2939 generated with these long section names are slightly non-standard, carrying
2940 as they do a string table, and may generate confusing output when examined
2941 with non-GNU PE-aware tools, such as file viewers and dumpers. However,
2942 GDB relies on the use of PE long section names to find Dwarf-2 debug
2943 information sections in an executable image at runtime, and so if neither
2944 option is specified on the command-line, @command{ld} will enable long
2945 section names, overriding the default and technically correct behaviour,
2946 when it finds the presence of debug information while linking an executable
2947 image and not stripping symbols.
2948 [This option is valid for all PE targeted ports of the linker]
2950 @kindex --enable-stdcall-fixup
2951 @kindex --disable-stdcall-fixup
2952 @item --enable-stdcall-fixup
2953 @itemx --disable-stdcall-fixup
2954 If the link finds a symbol that it cannot resolve, it will attempt to
2955 do ``fuzzy linking'' by looking for another defined symbol that differs
2956 only in the format of the symbol name (cdecl vs stdcall) and will
2957 resolve that symbol by linking to the match. For example, the
2958 undefined symbol @code{_foo} might be linked to the function
2959 @code{_foo@@12}, or the undefined symbol @code{_bar@@16} might be linked
2960 to the function @code{_bar}. When the linker does this, it prints a
2961 warning, since it normally should have failed to link, but sometimes
2962 import libraries generated from third-party dlls may need this feature
2963 to be usable. If you specify @option{--enable-stdcall-fixup}, this
2964 feature is fully enabled and warnings are not printed. If you specify
2965 @option{--disable-stdcall-fixup}, this feature is disabled and such
2966 mismatches are considered to be errors.
2967 [This option is specific to the i386 PE targeted port of the linker]
2969 @kindex --leading-underscore
2970 @kindex --no-leading-underscore
2971 @item --leading-underscore
2972 @itemx --no-leading-underscore
2973 For most targets default symbol-prefix is an underscore and is defined
2974 in target's description. By this option it is possible to
2975 disable/enable the default underscore symbol-prefix.
2977 @cindex DLLs, creating
2978 @kindex --export-all-symbols
2979 @item --export-all-symbols
2980 If given, all global symbols in the objects used to build a DLL will
2981 be exported by the DLL. Note that this is the default if there
2982 otherwise wouldn't be any exported symbols. When symbols are
2983 explicitly exported via DEF files or implicitly exported via function
2984 attributes, the default is to not export anything else unless this
2985 option is given. Note that the symbols @code{DllMain@@12},
2986 @code{DllEntryPoint@@0}, @code{DllMainCRTStartup@@12}, and
2987 @code{impure_ptr} will not be automatically
2988 exported. Also, symbols imported from other DLLs will not be
2989 re-exported, nor will symbols specifying the DLL's internal layout
2990 such as those beginning with @code{_head_} or ending with
2991 @code{_iname}. In addition, no symbols from @code{libgcc},
2992 @code{libstd++}, @code{libmingw32}, or @code{crtX.o} will be exported.
2993 Symbols whose names begin with @code{__rtti_} or @code{__builtin_} will
2994 not be exported, to help with C++ DLLs. Finally, there is an
2995 extensive list of cygwin-private symbols that are not exported
2996 (obviously, this applies on when building DLLs for cygwin targets).
2997 These cygwin-excludes are: @code{_cygwin_dll_entry@@12},
2998 @code{_cygwin_crt0_common@@8}, @code{_cygwin_noncygwin_dll_entry@@12},
2999 @code{_fmode}, @code{_impure_ptr}, @code{cygwin_attach_dll},
3000 @code{cygwin_premain0}, @code{cygwin_premain1}, @code{cygwin_premain2},
3001 @code{cygwin_premain3}, and @code{environ}.
3002 [This option is specific to the i386 PE targeted port of the linker]
3004 @kindex --exclude-symbols
3005 @item --exclude-symbols @var{symbol},@var{symbol},...
3006 Specifies a list of symbols which should not be automatically
3007 exported. The symbol names may be delimited by commas or colons.
3008 [This option is specific to the i386 PE targeted port of the linker]
3010 @kindex --exclude-all-symbols
3011 @item --exclude-all-symbols
3012 Specifies no symbols should be automatically exported.
3013 [This option is specific to the i386 PE targeted port of the linker]
3015 @kindex --file-alignment
3016 @item --file-alignment
3017 Specify the file alignment. Sections in the file will always begin at
3018 file offsets which are multiples of this number. This defaults to
3020 [This option is specific to the i386 PE targeted port of the linker]
3024 @item --heap @var{reserve}
3025 @itemx --heap @var{reserve},@var{commit}
3026 Specify the number of bytes of memory to reserve (and optionally commit)
3027 to be used as heap for this program. The default is 1MB reserved, 4K
3029 [This option is specific to the i386 PE targeted port of the linker]
3032 @kindex --image-base
3033 @item --image-base @var{value}
3034 Use @var{value} as the base address of your program or dll. This is
3035 the lowest memory location that will be used when your program or dll
3036 is loaded. To reduce the need to relocate and improve performance of
3037 your dlls, each should have a unique base address and not overlap any
3038 other dlls. The default is 0x400000 for executables, and 0x10000000
3040 [This option is specific to the i386 PE targeted port of the linker]
3044 If given, the stdcall suffixes (@@@var{nn}) will be stripped from
3045 symbols before they are exported.
3046 [This option is specific to the i386 PE targeted port of the linker]
3048 @kindex --large-address-aware
3049 @item --large-address-aware
3050 If given, the appropriate bit in the ``Characteristics'' field of the COFF
3051 header is set to indicate that this executable supports virtual addresses
3052 greater than 2 gigabytes. This should be used in conjunction with the /3GB
3053 or /USERVA=@var{value} megabytes switch in the ``[operating systems]''
3054 section of the BOOT.INI. Otherwise, this bit has no effect.
3055 [This option is specific to PE targeted ports of the linker]
3057 @kindex --disable-large-address-aware
3058 @item --disable-large-address-aware
3059 Reverts the effect of a previous @samp{--large-address-aware} option.
3060 This is useful if @samp{--large-address-aware} is always set by the compiler
3061 driver (e.g. Cygwin gcc) and the executable does not support virtual
3062 addresses greater than 2 gigabytes.
3063 [This option is specific to PE targeted ports of the linker]
3065 @kindex --major-image-version
3066 @item --major-image-version @var{value}
3067 Sets the major number of the ``image version''. Defaults to 1.
3068 [This option is specific to the i386 PE targeted port of the linker]
3070 @kindex --major-os-version
3071 @item --major-os-version @var{value}
3072 Sets the major number of the ``os version''. Defaults to 4.
3073 [This option is specific to the i386 PE targeted port of the linker]
3075 @kindex --major-subsystem-version
3076 @item --major-subsystem-version @var{value}
3077 Sets the major number of the ``subsystem version''. Defaults to 4.
3078 [This option is specific to the i386 PE targeted port of the linker]
3080 @kindex --minor-image-version
3081 @item --minor-image-version @var{value}
3082 Sets the minor number of the ``image version''. Defaults to 0.
3083 [This option is specific to the i386 PE targeted port of the linker]
3085 @kindex --minor-os-version
3086 @item --minor-os-version @var{value}
3087 Sets the minor number of the ``os version''. Defaults to 0.
3088 [This option is specific to the i386 PE targeted port of the linker]
3090 @kindex --minor-subsystem-version
3091 @item --minor-subsystem-version @var{value}
3092 Sets the minor number of the ``subsystem version''. Defaults to 0.
3093 [This option is specific to the i386 PE targeted port of the linker]
3095 @cindex DEF files, creating
3096 @cindex DLLs, creating
3097 @kindex --output-def
3098 @item --output-def @var{file}
3099 The linker will create the file @var{file} which will contain a DEF
3100 file corresponding to the DLL the linker is generating. This DEF file
3101 (which should be called @code{*.def}) may be used to create an import
3102 library with @code{dlltool} or may be used as a reference to
3103 automatically or implicitly exported symbols.
3104 [This option is specific to the i386 PE targeted port of the linker]
3106 @cindex DLLs, creating
3107 @kindex --enable-auto-image-base
3108 @item --enable-auto-image-base
3109 @itemx --enable-auto-image-base=@var{value}
3110 Automatically choose the image base for DLLs, optionally starting with base
3111 @var{value}, unless one is specified using the @code{--image-base} argument.
3112 By using a hash generated from the dllname to create unique image bases
3113 for each DLL, in-memory collisions and relocations which can delay program
3114 execution are avoided.
3115 [This option is specific to the i386 PE targeted port of the linker]
3117 @kindex --disable-auto-image-base
3118 @item --disable-auto-image-base
3119 Do not automatically generate a unique image base. If there is no
3120 user-specified image base (@code{--image-base}) then use the platform
3122 [This option is specific to the i386 PE targeted port of the linker]
3124 @cindex DLLs, linking to
3125 @kindex --dll-search-prefix
3126 @item --dll-search-prefix @var{string}
3127 When linking dynamically to a dll without an import library,
3128 search for @code{<string><basename>.dll} in preference to
3129 @code{lib<basename>.dll}. This behaviour allows easy distinction
3130 between DLLs built for the various "subplatforms": native, cygwin,
3131 uwin, pw, etc. For instance, cygwin DLLs typically use
3132 @code{--dll-search-prefix=cyg}.
3133 [This option is specific to the i386 PE targeted port of the linker]
3135 @kindex --enable-auto-import
3136 @item --enable-auto-import
3137 Do sophisticated linking of @code{_symbol} to @code{__imp__symbol} for
3138 DATA imports from DLLs, thus making it possible to bypass the dllimport
3139 mechanism on the user side and to reference unmangled symbol names.
3140 [This option is specific to the i386 PE targeted port of the linker]
3142 The following remarks pertain to the original implementation of the
3143 feature and are obsolete nowadays for Cygwin and MinGW targets.
3145 Note: Use of the 'auto-import' extension will cause the text section
3146 of the image file to be made writable. This does not conform to the
3147 PE-COFF format specification published by Microsoft.
3149 Note - use of the 'auto-import' extension will also cause read only
3150 data which would normally be placed into the .rdata section to be
3151 placed into the .data section instead. This is in order to work
3152 around a problem with consts that is described here:
3153 http://www.cygwin.com/ml/cygwin/2004-09/msg01101.html
3155 Using 'auto-import' generally will 'just work' -- but sometimes you may
3158 "variable '<var>' can't be auto-imported. Please read the
3159 documentation for ld's @code{--enable-auto-import} for details."
3161 This message occurs when some (sub)expression accesses an address
3162 ultimately given by the sum of two constants (Win32 import tables only
3163 allow one). Instances where this may occur include accesses to member
3164 fields of struct variables imported from a DLL, as well as using a
3165 constant index into an array variable imported from a DLL. Any
3166 multiword variable (arrays, structs, long long, etc) may trigger
3167 this error condition. However, regardless of the exact data type
3168 of the offending exported variable, ld will always detect it, issue
3169 the warning, and exit.
3171 There are several ways to address this difficulty, regardless of the
3172 data type of the exported variable:
3174 One way is to use --enable-runtime-pseudo-reloc switch. This leaves the task
3175 of adjusting references in your client code for runtime environment, so
3176 this method works only when runtime environment supports this feature.
3178 A second solution is to force one of the 'constants' to be a variable --
3179 that is, unknown and un-optimizable at compile time. For arrays,
3180 there are two possibilities: a) make the indexee (the array's address)
3181 a variable, or b) make the 'constant' index a variable. Thus:
3184 extern type extern_array[];
3186 @{ volatile type *t=extern_array; t[1] @}
3192 extern type extern_array[];
3194 @{ volatile int t=1; extern_array[t] @}
3197 For structs (and most other multiword data types) the only option
3198 is to make the struct itself (or the long long, or the ...) variable:
3201 extern struct s extern_struct;
3202 extern_struct.field -->
3203 @{ volatile struct s *t=&extern_struct; t->field @}
3209 extern long long extern_ll;
3211 @{ volatile long long * local_ll=&extern_ll; *local_ll @}
3214 A third method of dealing with this difficulty is to abandon
3215 'auto-import' for the offending symbol and mark it with
3216 @code{__declspec(dllimport)}. However, in practice that
3217 requires using compile-time #defines to indicate whether you are
3218 building a DLL, building client code that will link to the DLL, or
3219 merely building/linking to a static library. In making the choice
3220 between the various methods of resolving the 'direct address with
3221 constant offset' problem, you should consider typical real-world usage:
3229 void main(int argc, char **argv)@{
3230 printf("%d\n",arr[1]);
3240 void main(int argc, char **argv)@{
3241 /* This workaround is for win32 and cygwin; do not "optimize" */
3242 volatile int *parr = arr;
3243 printf("%d\n",parr[1]);
3250 /* Note: auto-export is assumed (no __declspec(dllexport)) */
3251 #if (defined(_WIN32) || defined(__CYGWIN__)) && \
3252 !(defined(FOO_BUILD_DLL) || defined(FOO_STATIC))
3253 #define FOO_IMPORT __declspec(dllimport)
3257 extern FOO_IMPORT int arr[];
3260 void main(int argc, char **argv)@{
3261 printf("%d\n",arr[1]);
3265 A fourth way to avoid this problem is to re-code your
3266 library to use a functional interface rather than a data interface
3267 for the offending variables (e.g. set_foo() and get_foo() accessor
3270 @kindex --disable-auto-import
3271 @item --disable-auto-import
3272 Do not attempt to do sophisticated linking of @code{_symbol} to
3273 @code{__imp__symbol} for DATA imports from DLLs.
3274 [This option is specific to the i386 PE targeted port of the linker]
3276 @kindex --enable-runtime-pseudo-reloc
3277 @item --enable-runtime-pseudo-reloc
3278 If your code contains expressions described in --enable-auto-import section,
3279 that is, DATA imports from DLL with non-zero offset, this switch will create
3280 a vector of 'runtime pseudo relocations' which can be used by runtime
3281 environment to adjust references to such data in your client code.
3282 [This option is specific to the i386 PE targeted port of the linker]
3284 @kindex --disable-runtime-pseudo-reloc
3285 @item --disable-runtime-pseudo-reloc
3286 Do not create pseudo relocations for non-zero offset DATA imports from DLLs.
3287 [This option is specific to the i386 PE targeted port of the linker]
3289 @kindex --enable-extra-pe-debug
3290 @item --enable-extra-pe-debug
3291 Show additional debug info related to auto-import symbol thunking.
3292 [This option is specific to the i386 PE targeted port of the linker]
3294 @kindex --section-alignment
3295 @item --section-alignment
3296 Sets the section alignment. Sections in memory will always begin at
3297 addresses which are a multiple of this number. Defaults to 0x1000.
3298 [This option is specific to the i386 PE targeted port of the linker]
3302 @item --stack @var{reserve}
3303 @itemx --stack @var{reserve},@var{commit}
3304 Specify the number of bytes of memory to reserve (and optionally commit)
3305 to be used as stack for this program. The default is 2MB reserved, 4K
3307 [This option is specific to the i386 PE targeted port of the linker]
3310 @item --subsystem @var{which}
3311 @itemx --subsystem @var{which}:@var{major}
3312 @itemx --subsystem @var{which}:@var{major}.@var{minor}
3313 Specifies the subsystem under which your program will execute. The
3314 legal values for @var{which} are @code{native}, @code{windows},
3315 @code{console}, @code{posix}, and @code{xbox}. You may optionally set
3316 the subsystem version also. Numeric values are also accepted for
3318 [This option is specific to the i386 PE targeted port of the linker]
3320 The following options set flags in the @code{DllCharacteristics} field
3321 of the PE file header:
3322 [These options are specific to PE targeted ports of the linker]
3324 @kindex --high-entropy-va
3325 @item --high-entropy-va
3326 @itemx --disable-high-entropy-va
3327 Image is compatible with 64-bit address space layout randomization
3328 (ASLR). This option is enabled by default for 64-bit PE images.
3330 This option also implies @option{--dynamicbase} and
3331 @option{--enable-reloc-section}.
3333 @kindex --dynamicbase
3335 @itemx --disable-dynamicbase
3336 The image base address may be relocated using address space layout
3337 randomization (ASLR). This feature was introduced with MS Windows
3338 Vista for i386 PE targets. This option is enabled by default but
3339 can be disabled via the @option{--disable-dynamicbase} option.
3340 This option also implies @option{--enable-reloc-section}.
3342 @kindex --forceinteg
3344 @itemx --disable-forceinteg
3345 Code integrity checks are enforced. This option is disabled by
3350 @item --disable-nxcompat
3351 The image is compatible with the Data Execution Prevention.
3352 This feature was introduced with MS Windows XP SP2 for i386 PE
3353 targets. The option is enabled by default.
3355 @kindex --no-isolation
3356 @item --no-isolation
3357 @itemx --disable-no-isolation
3358 Although the image understands isolation, do not isolate the image.
3359 This option is disabled by default.
3363 @itemx --disable-no-seh
3364 The image does not use SEH. No SE handler may be called from
3365 this image. This option is disabled by default.
3369 @itemx --disable-no-bind
3370 Do not bind this image. This option is disabled by default.
3374 @itemx --disable-wdmdriver
3375 The driver uses the MS Windows Driver Model. This option is disabled
3380 @itemx --disable-tsaware
3381 The image is Terminal Server aware. This option is disabled by
3384 @kindex --insert-timestamp
3385 @item --insert-timestamp
3386 @itemx --no-insert-timestamp
3387 Insert a real timestamp into the image. This is the default behaviour
3388 as it matches legacy code and it means that the image will work with
3389 other, proprietary tools. The problem with this default is that it
3390 will result in slightly different images being produced each time the
3391 same sources are linked. The option @option{--no-insert-timestamp}
3392 can be used to insert a zero value for the timestamp, this ensuring
3393 that binaries produced from identical sources will compare
3396 @kindex --enable-reloc-section
3397 @item --enable-reloc-section
3398 @itemx --disable-reloc-section
3399 Create the base relocation table, which is necessary if the image
3400 is loaded at a different image base than specified in the PE header.
3401 This option is enabled by default.
3407 @subsection Options specific to C6X uClinux targets
3409 @c man begin OPTIONS
3411 The C6X uClinux target uses a binary format called DSBT to support shared
3412 libraries. Each shared library in the system needs to have a unique index;
3413 all executables use an index of 0.
3418 @item --dsbt-size @var{size}
3419 This option sets the number of entries in the DSBT of the current executable
3420 or shared library to @var{size}. The default is to create a table with 64
3423 @kindex --dsbt-index
3424 @item --dsbt-index @var{index}
3425 This option sets the DSBT index of the current executable or shared library
3426 to @var{index}. The default is 0, which is appropriate for generating
3427 executables. If a shared library is generated with a DSBT index of 0, the
3428 @code{R_C6000_DSBT_INDEX} relocs are copied into the output file.
3430 @kindex --no-merge-exidx-entries
3431 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent
3432 exidx entries in frame unwind info.
3440 @subsection Options specific to C-SKY targets
3442 @c man begin OPTIONS
3446 @kindex --branch-stub on C-SKY
3448 This option enables linker branch relaxation by inserting branch stub
3449 sections when needed to extend the range of branches. This option is
3450 usually not required since C-SKY supports branch and call instructions that
3451 can access the full memory range and branch relaxation is normally handled by
3452 the compiler or assembler.
3454 @kindex --stub-group-size on C-SKY
3455 @item --stub-group-size=@var{N}
3456 This option allows finer control of linker branch stub creation.
3457 It sets the maximum size of a group of input sections that can
3458 be handled by one stub section. A negative value of @var{N} locates
3459 stub sections after their branches, while a positive value allows stub
3460 sections to appear either before or after the branches. Values of
3461 @samp{1} or @samp{-1} indicate that the
3462 linker should choose suitable defaults.
3470 @subsection Options specific to Motorola 68HC11 and 68HC12 targets
3472 @c man begin OPTIONS
3474 The 68HC11 and 68HC12 linkers support specific options to control the
3475 memory bank switching mapping and trampoline code generation.
3479 @kindex --no-trampoline
3480 @item --no-trampoline
3481 This option disables the generation of trampoline. By default a trampoline
3482 is generated for each far function which is called using a @code{jsr}
3483 instruction (this happens when a pointer to a far function is taken).
3485 @kindex --bank-window
3486 @item --bank-window @var{name}
3487 This option indicates to the linker the name of the memory region in
3488 the @samp{MEMORY} specification that describes the memory bank window.
3489 The definition of such region is then used by the linker to compute
3490 paging and addresses within the memory window.
3498 @subsection Options specific to Motorola 68K target
3500 @c man begin OPTIONS
3502 The following options are supported to control handling of GOT generation
3503 when linking for 68K targets.
3508 @item --got=@var{type}
3509 This option tells the linker which GOT generation scheme to use.
3510 @var{type} should be one of @samp{single}, @samp{negative},
3511 @samp{multigot} or @samp{target}. For more information refer to the
3512 Info entry for @file{ld}.
3520 @subsection Options specific to MIPS targets
3522 @c man begin OPTIONS
3524 The following options are supported to control microMIPS instruction
3525 generation and branch relocation checks for ISA mode transitions when
3526 linking for MIPS targets.
3534 These options control the choice of microMIPS instructions used in code
3535 generated by the linker, such as that in the PLT or lazy binding stubs,
3536 or in relaxation. If @samp{--insn32} is used, then the linker only uses
3537 32-bit instruction encodings. By default or if @samp{--no-insn32} is
3538 used, all instruction encodings are used, including 16-bit ones where
3541 @kindex --ignore-branch-isa
3542 @item --ignore-branch-isa
3543 @kindex --no-ignore-branch-isa
3544 @itemx --no-ignore-branch-isa
3545 These options control branch relocation checks for invalid ISA mode
3546 transitions. If @samp{--ignore-branch-isa} is used, then the linker
3547 accepts any branch relocations and any ISA mode transition required
3548 is lost in relocation calculation, except for some cases of @code{BAL}
3549 instructions which meet relaxation conditions and are converted to
3550 equivalent @code{JALX} instructions as the associated relocation is
3551 calculated. By default or if @samp{--no-ignore-branch-isa} is used
3552 a check is made causing the loss of an ISA mode transition to produce
3555 @kindex --compact-branches
3556 @item --compact-branches
3557 @kindex --no-compact-branches
3558 @itemx --no-compact-branches
3559 These options control the generation of compact instructions by the linker
3560 in the PLT entries for MIPS R6.
3569 @subsection Options specific to PDP11 targets
3571 @c man begin OPTIONS
3573 For the pdp11-aout target, three variants of the output format can be
3574 produced as selected by the following options. The default variant
3575 for pdp11-aout is the @samp{--omagic} option, whereas for other
3576 targets @samp{--nmagic} is the default. The @samp{--imagic} option is
3577 defined only for the pdp11-aout target, while the others are described
3578 here as they apply to the pdp11-aout target.
3587 Mark the output as @code{OMAGIC} (0407) in the @file{a.out} header to
3588 indicate that the text segment is not to be write-protected and
3589 shared. Since the text and data sections are both readable and
3590 writable, the data section is allocated immediately contiguous after
3591 the text segment. This is the oldest format for PDP11 executable
3592 programs and is the default for @command{ld} on PDP11 Unix systems
3593 from the beginning through 2.11BSD.
3600 Mark the output as @code{NMAGIC} (0410) in the @file{a.out} header to
3601 indicate that when the output file is executed, the text portion will
3602 be read-only and shareable among all processes executing the same
3603 file. This involves moving the data areas up to the first possible 8K
3604 byte page boundary following the end of the text. This option creates
3605 a @emph{pure executable} format.
3612 Mark the output as @code{IMAGIC} (0411) in the @file{a.out} header to
3613 indicate that when the output file is executed, the program text and
3614 data areas will be loaded into separate address spaces using the split
3615 instruction and data space feature of the memory management unit in
3616 larger models of the PDP11. This doubles the address space available
3617 to the program. The text segment is again pure, write-protected, and
3618 shareable. The only difference in the output format between this
3619 option and the others, besides the magic number, is that both the text
3620 and data sections start at location 0. The @samp{-z} option selected
3621 this format in 2.11BSD. This option creates a @emph{separate
3627 Equivalent to @samp{--nmagic} for pdp11-aout.
3636 @section Environment Variables
3638 @c man begin ENVIRONMENT
3640 You can change the behaviour of @command{ld} with the environment variables
3641 @ifclear SingleFormat
3644 @code{LDEMULATION} and @code{COLLECT_NO_DEMANGLE}.
3646 @ifclear SingleFormat
3648 @cindex default input format
3649 @code{GNUTARGET} determines the input-file object format if you don't
3650 use @samp{-b} (or its synonym @samp{--format}). Its value should be one
3651 of the BFD names for an input format (@pxref{BFD}). If there is no
3652 @code{GNUTARGET} in the environment, @command{ld} uses the natural format
3653 of the target. If @code{GNUTARGET} is set to @code{default} then BFD
3654 attempts to discover the input format by examining binary input files;
3655 this method often succeeds, but there are potential ambiguities, since
3656 there is no method of ensuring that the magic number used to specify
3657 object-file formats is unique. However, the configuration procedure for
3658 BFD on each system places the conventional format for that system first
3659 in the search-list, so ambiguities are resolved in favor of convention.
3663 @cindex default emulation
3664 @cindex emulation, default
3665 @code{LDEMULATION} determines the default emulation if you don't use the
3666 @samp{-m} option. The emulation can affect various aspects of linker
3667 behaviour, particularly the default linker script. You can list the
3668 available emulations with the @samp{--verbose} or @samp{-V} options. If
3669 the @samp{-m} option is not used, and the @code{LDEMULATION} environment
3670 variable is not defined, the default emulation depends upon how the
3671 linker was configured.
3673 @kindex COLLECT_NO_DEMANGLE
3674 @cindex demangling, default
3675 Normally, the linker will default to demangling symbols. However, if
3676 @code{COLLECT_NO_DEMANGLE} is set in the environment, then it will
3677 default to not demangling symbols. This environment variable is used in
3678 a similar fashion by the @code{gcc} linker wrapper program. The default
3679 may be overridden by the @samp{--demangle} and @samp{--no-demangle}
3686 @chapter Linker Scripts
3689 @cindex linker scripts
3690 @cindex command files
3691 Every link is controlled by a @dfn{linker script}. This script is
3692 written in the linker command language.
3694 The main purpose of the linker script is to describe how the sections in
3695 the input files should be mapped into the output file, and to control
3696 the memory layout of the output file. Most linker scripts do nothing
3697 more than this. However, when necessary, the linker script can also
3698 direct the linker to perform many other operations, using the commands
3701 The linker always uses a linker script. If you do not supply one
3702 yourself, the linker will use a default script that is compiled into the
3703 linker executable. You can use the @samp{--verbose} command-line option
3704 to display the default linker script. Certain command-line options,
3705 such as @samp{-r} or @samp{-N}, will affect the default linker script.
3707 You may supply your own linker script by using the @samp{-T} command
3708 line option. When you do this, your linker script will replace the
3709 default linker script.
3711 You may also use linker scripts implicitly by naming them as input files
3712 to the linker, as though they were files to be linked. @xref{Implicit
3716 * Basic Script Concepts:: Basic Linker Script Concepts
3717 * Script Format:: Linker Script Format
3718 * Simple Example:: Simple Linker Script Example
3719 * Simple Commands:: Simple Linker Script Commands
3720 * Assignments:: Assigning Values to Symbols
3721 * SECTIONS:: SECTIONS Command
3722 * MEMORY:: MEMORY Command
3723 * PHDRS:: PHDRS Command
3724 * VERSION:: VERSION Command
3725 * Expressions:: Expressions in Linker Scripts
3726 * Implicit Linker Scripts:: Implicit Linker Scripts
3729 @node Basic Script Concepts
3730 @section Basic Linker Script Concepts
3731 @cindex linker script concepts
3732 We need to define some basic concepts and vocabulary in order to
3733 describe the linker script language.
3735 The linker combines input files into a single output file. The output
3736 file and each input file are in a special data format known as an
3737 @dfn{object file format}. Each file is called an @dfn{object file}.
3738 The output file is often called an @dfn{executable}, but for our
3739 purposes we will also call it an object file. Each object file has,
3740 among other things, a list of @dfn{sections}. We sometimes refer to a
3741 section in an input file as an @dfn{input section}; similarly, a section
3742 in the output file is an @dfn{output section}.
3744 Each section in an object file has a name and a size. Most sections
3745 also have an associated block of data, known as the @dfn{section
3746 contents}. A section may be marked as @dfn{loadable}, which means that
3747 the contents should be loaded into memory when the output file is run.
3748 A section with no contents may be @dfn{allocatable}, which means that an
3749 area in memory should be set aside, but nothing in particular should be
3750 loaded there (in some cases this memory must be zeroed out). A section
3751 which is neither loadable nor allocatable typically contains some sort
3752 of debugging information.
3754 Every loadable or allocatable output section has two addresses. The
3755 first is the @dfn{VMA}, or virtual memory address. This is the address
3756 the section will have when the output file is run. The second is the
3757 @dfn{LMA}, or load memory address. This is the address at which the
3758 section will be loaded. In most cases the two addresses will be the
3759 same. An example of when they might be different is when a data section
3760 is loaded into ROM, and then copied into RAM when the program starts up
3761 (this technique is often used to initialize global variables in a ROM
3762 based system). In this case the ROM address would be the LMA, and the
3763 RAM address would be the VMA.
3765 You can see the sections in an object file by using the @code{objdump}
3766 program with the @samp{-h} option.
3768 Every object file also has a list of @dfn{symbols}, known as the
3769 @dfn{symbol table}. A symbol may be defined or undefined. Each symbol
3770 has a name, and each defined symbol has an address, among other
3771 information. If you compile a C or C++ program into an object file, you
3772 will get a defined symbol for every defined function and global or
3773 static variable. Every undefined function or global variable which is
3774 referenced in the input file will become an undefined symbol.
3776 You can see the symbols in an object file by using the @code{nm}
3777 program, or by using the @code{objdump} program with the @samp{-t}
3781 @section Linker Script Format
3782 @cindex linker script format
3783 Linker scripts are text files.
3785 You write a linker script as a series of commands. Each command is
3786 either a keyword, possibly followed by arguments, or an assignment to a
3787 symbol. You may separate commands using semicolons. Whitespace is
3790 Strings such as file or format names can normally be entered directly.
3791 If the file name contains a character such as a comma which would
3792 otherwise serve to separate file names, you may put the file name in
3793 double quotes. There is no way to use a double quote character in a
3796 You may include comments in linker scripts just as in C, delimited by
3797 @samp{/*} and @samp{*/}. As in C, comments are syntactically equivalent
3800 @node Simple Example
3801 @section Simple Linker Script Example
3802 @cindex linker script example
3803 @cindex example of linker script
3804 Many linker scripts are fairly simple.
3806 The simplest possible linker script has just one command:
3807 @samp{SECTIONS}. You use the @samp{SECTIONS} command to describe the
3808 memory layout of the output file.
3810 The @samp{SECTIONS} command is a powerful command. Here we will
3811 describe a simple use of it. Let's assume your program consists only of
3812 code, initialized data, and uninitialized data. These will be in the
3813 @samp{.text}, @samp{.data}, and @samp{.bss} sections, respectively.
3814 Let's assume further that these are the only sections which appear in
3817 For this example, let's say that the code should be loaded at address
3818 0x10000, and that the data should start at address 0x8000000. Here is a
3819 linker script which will do that:
3824 .text : @{ *(.text) @}
3826 .data : @{ *(.data) @}
3827 .bss : @{ *(.bss) @}
3831 You write the @samp{SECTIONS} command as the keyword @samp{SECTIONS},
3832 followed by a series of symbol assignments and output section
3833 descriptions enclosed in curly braces.
3835 The first line inside the @samp{SECTIONS} command of the above example
3836 sets the value of the special symbol @samp{.}, which is the location
3837 counter. If you do not specify the address of an output section in some
3838 other way (other ways are described later), the address is set from the
3839 current value of the location counter. The location counter is then
3840 incremented by the size of the output section. At the start of the
3841 @samp{SECTIONS} command, the location counter has the value @samp{0}.
3843 The second line defines an output section, @samp{.text}. The colon is
3844 required syntax which may be ignored for now. Within the curly braces
3845 after the output section name, you list the names of the input sections
3846 which should be placed into this output section. The @samp{*} is a
3847 wildcard which matches any file name. The expression @samp{*(.text)}
3848 means all @samp{.text} input sections in all input files.
3850 Since the location counter is @samp{0x10000} when the output section
3851 @samp{.text} is defined, the linker will set the address of the
3852 @samp{.text} section in the output file to be @samp{0x10000}.
3854 The remaining lines define the @samp{.data} and @samp{.bss} sections in
3855 the output file. The linker will place the @samp{.data} output section
3856 at address @samp{0x8000000}. After the linker places the @samp{.data}
3857 output section, the value of the location counter will be
3858 @samp{0x8000000} plus the size of the @samp{.data} output section. The
3859 effect is that the linker will place the @samp{.bss} output section
3860 immediately after the @samp{.data} output section in memory.
3862 The linker will ensure that each output section has the required
3863 alignment, by increasing the location counter if necessary. In this
3864 example, the specified addresses for the @samp{.text} and @samp{.data}
3865 sections will probably satisfy any alignment constraints, but the linker
3866 may have to create a small gap between the @samp{.data} and @samp{.bss}
3869 That's it! That's a simple and complete linker script.
3871 @node Simple Commands
3872 @section Simple Linker Script Commands
3873 @cindex linker script simple commands
3874 In this section we describe the simple linker script commands.
3877 * Entry Point:: Setting the entry point
3878 * File Commands:: Commands dealing with files
3879 @ifclear SingleFormat
3880 * Format Commands:: Commands dealing with object file formats
3883 * REGION_ALIAS:: Assign alias names to memory regions
3884 * Miscellaneous Commands:: Other linker script commands
3888 @subsection Setting the Entry Point
3889 @kindex ENTRY(@var{symbol})
3890 @cindex start of execution
3891 @cindex first instruction
3893 The first instruction to execute in a program is called the @dfn{entry
3894 point}. You can use the @code{ENTRY} linker script command to set the
3895 entry point. The argument is a symbol name:
3900 There are several ways to set the entry point. The linker will set the
3901 entry point by trying each of the following methods in order, and
3902 stopping when one of them succeeds:
3905 the @samp{-e} @var{entry} command-line option;
3907 the @code{ENTRY(@var{symbol})} command in a linker script;
3909 the value of a target-specific symbol, if it is defined; For many
3910 targets this is @code{start}, but PE- and BeOS-based systems for example
3911 check a list of possible entry symbols, matching the first one found.
3913 the address of the first byte of the code section, if present and an
3914 executable is being created - the code section is usually
3915 @samp{.text}, but can be something else;
3917 The address @code{0}.
3921 @subsection Commands Dealing with Files
3922 @cindex linker script file commands
3923 Several linker script commands deal with files.
3926 @item INCLUDE @var{filename}
3927 @kindex INCLUDE @var{filename}
3928 @cindex including a linker script
3929 Include the linker script @var{filename} at this point. The file will
3930 be searched for in the current directory, and in any directory specified
3931 with the @option{-L} option. You can nest calls to @code{INCLUDE} up to
3934 You can place @code{INCLUDE} directives at the top level, in @code{MEMORY} or
3935 @code{SECTIONS} commands, or in output section descriptions.
3937 @item INPUT(@var{file}, @var{file}, @dots{})
3938 @itemx INPUT(@var{file} @var{file} @dots{})
3939 @kindex INPUT(@var{files})
3940 @cindex input files in linker scripts
3941 @cindex input object files in linker scripts
3942 @cindex linker script input object files
3943 The @code{INPUT} command directs the linker to include the named files
3944 in the link, as though they were named on the command line.
3946 For example, if you always want to include @file{subr.o} any time you do
3947 a link, but you can't be bothered to put it on every link command line,
3948 then you can put @samp{INPUT (subr.o)} in your linker script.
3950 In fact, if you like, you can list all of your input files in the linker
3951 script, and then invoke the linker with nothing but a @samp{-T} option.
3953 In case a @dfn{sysroot prefix} is configured, and the filename starts
3954 with the @samp{/} character, and the script being processed was
3955 located inside the @dfn{sysroot prefix}, the filename will be looked
3956 for in the @dfn{sysroot prefix}. The @dfn{sysroot prefix} can also be forced by specifying
3957 @code{=} as the first character in the filename path, or prefixing the
3958 filename path with @code{$SYSROOT}. See also the description of
3959 @samp{-L} in @ref{Options,,Command-line Options}.
3961 If a @dfn{sysroot prefix} is not used then the linker will try to open
3962 the file in the directory containing the linker script. If it is not
3963 found the linker will then search the current directory. If it is still
3964 not found the linker will search through the archive library search
3967 If you use @samp{INPUT (-l@var{file})}, @command{ld} will transform the
3968 name to @code{lib@var{file}.a}, as with the command-line argument
3971 When you use the @code{INPUT} command in an implicit linker script, the
3972 files will be included in the link at the point at which the linker
3973 script file is included. This can affect archive searching.
3975 @item GROUP(@var{file}, @var{file}, @dots{})
3976 @itemx GROUP(@var{file} @var{file} @dots{})
3977 @kindex GROUP(@var{files})
3978 @cindex grouping input files
3979 The @code{GROUP} command is like @code{INPUT}, except that the named
3980 files should all be archives, and they are searched repeatedly until no
3981 new undefined references are created. See the description of @samp{-(}
3982 in @ref{Options,,Command-line Options}.
3984 @item AS_NEEDED(@var{file}, @var{file}, @dots{})
3985 @itemx AS_NEEDED(@var{file} @var{file} @dots{})
3986 @kindex AS_NEEDED(@var{files})
3987 This construct can appear only inside of the @code{INPUT} or @code{GROUP}
3988 commands, among other filenames. The files listed will be handled
3989 as if they appear directly in the @code{INPUT} or @code{GROUP} commands,
3990 with the exception of ELF shared libraries, that will be added only
3991 when they are actually needed. This construct essentially enables
3992 @option{--as-needed} option for all the files listed inside of it
3993 and restores previous @option{--as-needed} resp. @option{--no-as-needed}
3996 @item OUTPUT(@var{filename})
3997 @kindex OUTPUT(@var{filename})
3998 @cindex output file name in linker script
3999 The @code{OUTPUT} command names the output file. Using
4000 @code{OUTPUT(@var{filename})} in the linker script is exactly like using
4001 @samp{-o @var{filename}} on the command line (@pxref{Options,,Command
4002 Line Options}). If both are used, the command-line option takes
4005 You can use the @code{OUTPUT} command to define a default name for the
4006 output file other than the usual default of @file{a.out}.
4008 @item SEARCH_DIR(@var{path})
4009 @kindex SEARCH_DIR(@var{path})
4010 @cindex library search path in linker script
4011 @cindex archive search path in linker script
4012 @cindex search path in linker script
4013 The @code{SEARCH_DIR} command adds @var{path} to the list of paths where
4014 @command{ld} looks for archive libraries. Using
4015 @code{SEARCH_DIR(@var{path})} is exactly like using @samp{-L @var{path}}
4016 on the command line (@pxref{Options,,Command-line Options}). If both
4017 are used, then the linker will search both paths. Paths specified using
4018 the command-line option are searched first.
4020 @item STARTUP(@var{filename})
4021 @kindex STARTUP(@var{filename})
4022 @cindex first input file
4023 The @code{STARTUP} command is just like the @code{INPUT} command, except
4024 that @var{filename} will become the first input file to be linked, as
4025 though it were specified first on the command line. This may be useful
4026 when using a system in which the entry point is always the start of the
4030 @ifclear SingleFormat
4031 @node Format Commands
4032 @subsection Commands Dealing with Object File Formats
4033 A couple of linker script commands deal with object file formats.
4036 @item OUTPUT_FORMAT(@var{bfdname})
4037 @itemx OUTPUT_FORMAT(@var{default}, @var{big}, @var{little})
4038 @kindex OUTPUT_FORMAT(@var{bfdname})
4039 @cindex output file format in linker script
4040 The @code{OUTPUT_FORMAT} command names the BFD format to use for the
4041 output file (@pxref{BFD}). Using @code{OUTPUT_FORMAT(@var{bfdname})} is
4042 exactly like using @samp{--oformat @var{bfdname}} on the command line
4043 (@pxref{Options,,Command-line Options}). If both are used, the command
4044 line option takes precedence.
4046 You can use @code{OUTPUT_FORMAT} with three arguments to use different
4047 formats based on the @samp{-EB} and @samp{-EL} command-line options.
4048 This permits the linker script to set the output format based on the
4051 If neither @samp{-EB} nor @samp{-EL} are used, then the output format
4052 will be the first argument, @var{default}. If @samp{-EB} is used, the
4053 output format will be the second argument, @var{big}. If @samp{-EL} is
4054 used, the output format will be the third argument, @var{little}.
4056 For example, the default linker script for the MIPS ELF target uses this
4059 OUTPUT_FORMAT(elf32-bigmips, elf32-bigmips, elf32-littlemips)
4061 This says that the default format for the output file is
4062 @samp{elf32-bigmips}, but if the user uses the @samp{-EL} command-line
4063 option, the output file will be created in the @samp{elf32-littlemips}
4066 @item TARGET(@var{bfdname})
4067 @kindex TARGET(@var{bfdname})
4068 @cindex input file format in linker script
4069 The @code{TARGET} command names the BFD format to use when reading input
4070 files. It affects subsequent @code{INPUT} and @code{GROUP} commands.
4071 This command is like using @samp{-b @var{bfdname}} on the command line
4072 (@pxref{Options,,Command-line Options}). If the @code{TARGET} command
4073 is used but @code{OUTPUT_FORMAT} is not, then the last @code{TARGET}
4074 command is also used to set the format for the output file. @xref{BFD}.
4079 @subsection Assign alias names to memory regions
4080 @kindex REGION_ALIAS(@var{alias}, @var{region})
4081 @cindex region alias
4082 @cindex region names
4084 Alias names can be added to existing memory regions created with the
4085 @ref{MEMORY} command. Each name corresponds to at most one memory region.
4088 REGION_ALIAS(@var{alias}, @var{region})
4091 The @code{REGION_ALIAS} function creates an alias name @var{alias} for the
4092 memory region @var{region}. This allows a flexible mapping of output sections
4093 to memory regions. An example follows.
4095 Suppose we have an application for embedded systems which come with various
4096 memory storage devices. All have a general purpose, volatile memory @code{RAM}
4097 that allows code execution or data storage. Some may have a read-only,
4098 non-volatile memory @code{ROM} that allows code execution and read-only data
4099 access. The last variant is a read-only, non-volatile memory @code{ROM2} with
4100 read-only data access and no code execution capability. We have four output
4105 @code{.text} program code;
4107 @code{.rodata} read-only data;
4109 @code{.data} read-write initialized data;
4111 @code{.bss} read-write zero initialized data.
4114 The goal is to provide a linker command file that contains a system independent
4115 part defining the output sections and a system dependent part mapping the
4116 output sections to the memory regions available on the system. Our embedded
4117 systems come with three different memory setups @code{A}, @code{B} and
4119 @multitable @columnfractions .25 .25 .25 .25
4120 @item Section @tab Variant A @tab Variant B @tab Variant C
4121 @item .text @tab RAM @tab ROM @tab ROM
4122 @item .rodata @tab RAM @tab ROM @tab ROM2
4123 @item .data @tab RAM @tab RAM/ROM @tab RAM/ROM2
4124 @item .bss @tab RAM @tab RAM @tab RAM
4126 The notation @code{RAM/ROM} or @code{RAM/ROM2} means that this section is
4127 loaded into region @code{ROM} or @code{ROM2} respectively. Please note that
4128 the load address of the @code{.data} section starts in all three variants at
4129 the end of the @code{.rodata} section.
4131 The base linker script that deals with the output sections follows. It
4132 includes the system dependent @code{linkcmds.memory} file that describes the
4135 INCLUDE linkcmds.memory
4148 .data : AT (rodata_end)
4153 data_size = SIZEOF(.data);
4154 data_load_start = LOADADDR(.data);
4162 Now we need three different @code{linkcmds.memory} files to define memory
4163 regions and alias names. The content of @code{linkcmds.memory} for the three
4164 variants @code{A}, @code{B} and @code{C}:
4167 Here everything goes into the @code{RAM}.
4171 RAM : ORIGIN = 0, LENGTH = 4M
4174 REGION_ALIAS("REGION_TEXT", RAM);
4175 REGION_ALIAS("REGION_RODATA", RAM);
4176 REGION_ALIAS("REGION_DATA", RAM);
4177 REGION_ALIAS("REGION_BSS", RAM);
4180 Program code and read-only data go into the @code{ROM}. Read-write data goes
4181 into the @code{RAM}. An image of the initialized data is loaded into the
4182 @code{ROM} and will be copied during system start into the @code{RAM}.
4186 ROM : ORIGIN = 0, LENGTH = 3M
4187 RAM : ORIGIN = 0x10000000, LENGTH = 1M
4190 REGION_ALIAS("REGION_TEXT", ROM);
4191 REGION_ALIAS("REGION_RODATA", ROM);
4192 REGION_ALIAS("REGION_DATA", RAM);
4193 REGION_ALIAS("REGION_BSS", RAM);
4196 Program code goes into the @code{ROM}. Read-only data goes into the
4197 @code{ROM2}. Read-write data goes into the @code{RAM}. An image of the
4198 initialized data is loaded into the @code{ROM2} and will be copied during
4199 system start into the @code{RAM}.
4203 ROM : ORIGIN = 0, LENGTH = 2M
4204 ROM2 : ORIGIN = 0x10000000, LENGTH = 1M
4205 RAM : ORIGIN = 0x20000000, LENGTH = 1M
4208 REGION_ALIAS("REGION_TEXT", ROM);
4209 REGION_ALIAS("REGION_RODATA", ROM2);
4210 REGION_ALIAS("REGION_DATA", RAM);
4211 REGION_ALIAS("REGION_BSS", RAM);
4215 It is possible to write a common system initialization routine to copy the
4216 @code{.data} section from @code{ROM} or @code{ROM2} into the @code{RAM} if
4221 extern char data_start [];
4222 extern char data_size [];
4223 extern char data_load_start [];
4225 void copy_data(void)
4227 if (data_start != data_load_start)
4229 memcpy(data_start, data_load_start, (size_t) data_size);
4234 @node Miscellaneous Commands
4235 @subsection Other Linker Script Commands
4236 There are a few other linker scripts commands.
4239 @item ASSERT(@var{exp}, @var{message})
4241 @cindex assertion in linker script
4242 Ensure that @var{exp} is non-zero. If it is zero, then exit the linker
4243 with an error code, and print @var{message}.
4245 Note that assertions are checked before the final stages of linking
4246 take place. This means that expressions involving symbols PROVIDEd
4247 inside section definitions will fail if the user has not set values
4248 for those symbols. The only exception to this rule is PROVIDEd
4249 symbols that just reference dot. Thus an assertion like this:
4254 PROVIDE (__stack = .);
4255 PROVIDE (__stack_size = 0x100);
4256 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
4260 will fail if @code{__stack_size} is not defined elsewhere. Symbols
4261 PROVIDEd outside of section definitions are evaluated earlier, so they
4262 can be used inside ASSERTions. Thus:
4265 PROVIDE (__stack_size = 0x100);
4268 PROVIDE (__stack = .);
4269 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
4275 @item EXTERN(@var{symbol} @var{symbol} @dots{})
4277 @cindex undefined symbol in linker script
4278 Force @var{symbol} to be entered in the output file as an undefined
4279 symbol. Doing this may, for example, trigger linking of additional
4280 modules from standard libraries. You may list several @var{symbol}s for
4281 each @code{EXTERN}, and you may use @code{EXTERN} multiple times. This
4282 command has the same effect as the @samp{-u} command-line option.
4284 @item FORCE_COMMON_ALLOCATION
4285 @kindex FORCE_COMMON_ALLOCATION
4286 @cindex common allocation in linker script
4287 This command has the same effect as the @samp{-d} command-line option:
4288 to make @command{ld} assign space to common symbols even if a relocatable
4289 output file is specified (@samp{-r}).
4291 @item INHIBIT_COMMON_ALLOCATION
4292 @kindex INHIBIT_COMMON_ALLOCATION
4293 @cindex common allocation in linker script
4294 This command has the same effect as the @samp{--no-define-common}
4295 command-line option: to make @code{ld} omit the assignment of addresses
4296 to common symbols even for a non-relocatable output file.
4298 @item FORCE_GROUP_ALLOCATION
4299 @kindex FORCE_GROUP_ALLOCATION
4300 @cindex group allocation in linker script
4301 @cindex section groups
4303 This command has the same effect as the
4304 @samp{--force-group-allocation} command-line option: to make
4305 @command{ld} place section group members like normal input sections,
4306 and to delete the section groups even if a relocatable output file is
4307 specified (@samp{-r}).
4309 @item INSERT [ AFTER | BEFORE ] @var{output_section}
4311 @cindex insert user script into default script
4312 This command is typically used in a script specified by @samp{-T} to
4313 augment the default @code{SECTIONS} with, for example, overlays. It
4314 inserts all prior linker script statements after (or before)
4315 @var{output_section}, and also causes @samp{-T} to not override the
4316 default linker script. The exact insertion point is as for orphan
4317 sections. @xref{Location Counter}. The insertion happens after the
4318 linker has mapped input sections to output sections. Prior to the
4319 insertion, since @samp{-T} scripts are parsed before the default
4320 linker script, statements in the @samp{-T} script occur before the
4321 default linker script statements in the internal linker representation
4322 of the script. In particular, input section assignments will be made
4323 to @samp{-T} output sections before those in the default script. Here
4324 is an example of how a @samp{-T} script using @code{INSERT} might look:
4331 .ov1 @{ ov1*(.text) @}
4332 .ov2 @{ ov2*(.text) @}
4338 @item NOCROSSREFS(@var{section} @var{section} @dots{})
4339 @kindex NOCROSSREFS(@var{sections})
4340 @cindex cross references
4341 This command may be used to tell @command{ld} to issue an error about any
4342 references among certain output sections.
4344 In certain types of programs, particularly on embedded systems when
4345 using overlays, when one section is loaded into memory, another section
4346 will not be. Any direct references between the two sections would be
4347 errors. For example, it would be an error if code in one section called
4348 a function defined in the other section.
4350 The @code{NOCROSSREFS} command takes a list of output section names. If
4351 @command{ld} detects any cross references between the sections, it reports
4352 an error and returns a non-zero exit status. Note that the
4353 @code{NOCROSSREFS} command uses output section names, not input section
4356 @item NOCROSSREFS_TO(@var{tosection} @var{fromsection} @dots{})
4357 @kindex NOCROSSREFS_TO(@var{tosection} @var{fromsections})
4358 @cindex cross references
4359 This command may be used to tell @command{ld} to issue an error about any
4360 references to one section from a list of other sections.
4362 The @code{NOCROSSREFS} command is useful when ensuring that two or more
4363 output sections are entirely independent but there are situations where
4364 a one-way dependency is needed. For example, in a multi-core application
4365 there may be shared code that can be called from each core but for safety
4366 must never call back.
4368 The @code{NOCROSSREFS_TO} command takes a list of output section names.
4369 The first section can not be referenced from any of the other sections.
4370 If @command{ld} detects any references to the first section from any of
4371 the other sections, it reports an error and returns a non-zero exit
4372 status. Note that the @code{NOCROSSREFS_TO} command uses output section
4373 names, not input section names.
4375 @ifclear SingleFormat
4376 @item OUTPUT_ARCH(@var{bfdarch})
4377 @kindex OUTPUT_ARCH(@var{bfdarch})
4378 @cindex machine architecture
4379 @cindex architecture
4380 Specify a particular output machine architecture. The argument is one
4381 of the names used by the BFD library (@pxref{BFD}). You can see the
4382 architecture of an object file by using the @code{objdump} program with
4383 the @samp{-f} option.
4386 @item LD_FEATURE(@var{string})
4387 @kindex LD_FEATURE(@var{string})
4388 This command may be used to modify @command{ld} behavior. If
4389 @var{string} is @code{"SANE_EXPR"} then absolute symbols and numbers
4390 in a script are simply treated as numbers everywhere.
4391 @xref{Expression Section}.
4395 @section Assigning Values to Symbols
4396 @cindex assignment in scripts
4397 @cindex symbol definition, scripts
4398 @cindex variables, defining
4399 You may assign a value to a symbol in a linker script. This will define
4400 the symbol and place it into the symbol table with a global scope.
4403 * Simple Assignments:: Simple Assignments
4406 * PROVIDE_HIDDEN:: PROVIDE_HIDDEN
4407 * Source Code Reference:: How to use a linker script defined symbol in source code
4410 @node Simple Assignments
4411 @subsection Simple Assignments
4413 You may assign to a symbol using any of the C assignment operators:
4416 @item @var{symbol} = @var{expression} ;
4417 @itemx @var{symbol} += @var{expression} ;
4418 @itemx @var{symbol} -= @var{expression} ;
4419 @itemx @var{symbol} *= @var{expression} ;
4420 @itemx @var{symbol} /= @var{expression} ;
4421 @itemx @var{symbol} <<= @var{expression} ;
4422 @itemx @var{symbol} >>= @var{expression} ;
4423 @itemx @var{symbol} &= @var{expression} ;
4424 @itemx @var{symbol} |= @var{expression} ;
4427 The first case will define @var{symbol} to the value of
4428 @var{expression}. In the other cases, @var{symbol} must already be
4429 defined, and the value will be adjusted accordingly.
4431 The special symbol name @samp{.} indicates the location counter. You
4432 may only use this within a @code{SECTIONS} command. @xref{Location Counter}.
4434 The semicolon after @var{expression} is required.
4436 Expressions are defined below; see @ref{Expressions}.
4438 You may write symbol assignments as commands in their own right, or as
4439 statements within a @code{SECTIONS} command, or as part of an output
4440 section description in a @code{SECTIONS} command.
4442 The section of the symbol will be set from the section of the
4443 expression; for more information, see @ref{Expression Section}.
4445 Here is an example showing the three different places that symbol
4446 assignments may be used:
4457 _bdata = (. + 3) & ~ 3;
4458 .data : @{ *(.data) @}
4462 In this example, the symbol @samp{floating_point} will be defined as
4463 zero. The symbol @samp{_etext} will be defined as the address following
4464 the last @samp{.text} input section. The symbol @samp{_bdata} will be
4465 defined as the address following the @samp{.text} output section aligned
4466 upward to a 4 byte boundary.
4471 For ELF targeted ports, define a symbol that will be hidden and won't be
4472 exported. The syntax is @code{HIDDEN(@var{symbol} = @var{expression})}.
4474 Here is the example from @ref{Simple Assignments}, rewritten to use
4478 HIDDEN(floating_point = 0);
4486 HIDDEN(_bdata = (. + 3) & ~ 3);
4487 .data : @{ *(.data) @}
4491 In this case none of the three symbols will be visible outside this module.
4496 In some cases, it is desirable for a linker script to define a symbol
4497 only if it is referenced and is not defined by any object included in
4498 the link. For example, traditional linkers defined the symbol
4499 @samp{etext}. However, ANSI C requires that the user be able to use
4500 @samp{etext} as a function name without encountering an error. The
4501 @code{PROVIDE} keyword may be used to define a symbol, such as
4502 @samp{etext}, only if it is referenced but not defined. The syntax is
4503 @code{PROVIDE(@var{symbol} = @var{expression})}.
4505 Here is an example of using @code{PROVIDE} to define @samp{etext}:
4518 In this example, if the program defines @samp{_etext} (with a leading
4519 underscore), the linker will give a multiple definition diagnostic. If,
4520 on the other hand, the program defines @samp{etext} (with no leading
4521 underscore), the linker will silently use the definition in the program.
4522 If the program references @samp{etext} but does not define it, the
4523 linker will use the definition in the linker script.
4525 Note - the @code{PROVIDE} directive considers a common symbol to be
4526 defined, even though such a symbol could be combined with the symbol
4527 that the @code{PROVIDE} would create. This is particularly important
4528 when considering constructor and destructor list symbols such as
4529 @samp{__CTOR_LIST__} as these are often defined as common symbols.
4531 @node PROVIDE_HIDDEN
4532 @subsection PROVIDE_HIDDEN
4533 @cindex PROVIDE_HIDDEN
4534 Similar to @code{PROVIDE}. For ELF targeted ports, the symbol will be
4535 hidden and won't be exported.
4537 @node Source Code Reference
4538 @subsection Source Code Reference
4540 Accessing a linker script defined variable from source code is not
4541 intuitive. In particular a linker script symbol is not equivalent to
4542 a variable declaration in a high level language, it is instead a
4543 symbol that does not have a value.
4545 Before going further, it is important to note that compilers often
4546 transform names in the source code into different names when they are
4547 stored in the symbol table. For example, Fortran compilers commonly
4548 prepend or append an underscore, and C++ performs extensive @samp{name
4549 mangling}. Therefore there might be a discrepancy between the name
4550 of a variable as it is used in source code and the name of the same
4551 variable as it is defined in a linker script. For example in C a
4552 linker script variable might be referred to as:
4558 But in the linker script it might be defined as:
4564 In the remaining examples however it is assumed that no name
4565 transformation has taken place.
4567 When a symbol is declared in a high level language such as C, two
4568 things happen. The first is that the compiler reserves enough space
4569 in the program's memory to hold the @emph{value} of the symbol. The
4570 second is that the compiler creates an entry in the program's symbol
4571 table which holds the symbol's @emph{address}. ie the symbol table
4572 contains the address of the block of memory holding the symbol's
4573 value. So for example the following C declaration, at file scope:
4579 creates an entry called @samp{foo} in the symbol table. This entry
4580 holds the address of an @samp{int} sized block of memory where the
4581 number 1000 is initially stored.
4583 When a program references a symbol the compiler generates code that
4584 first accesses the symbol table to find the address of the symbol's
4585 memory block and then code to read the value from that memory block.
4592 looks up the symbol @samp{foo} in the symbol table, gets the address
4593 associated with this symbol and then writes the value 1 into that
4600 looks up the symbol @samp{foo} in the symbol table, gets its address
4601 and then copies this address into the block of memory associated with
4602 the variable @samp{a}.
4604 Linker scripts symbol declarations, by contrast, create an entry in
4605 the symbol table but do not assign any memory to them. Thus they are
4606 an address without a value. So for example the linker script definition:
4612 creates an entry in the symbol table called @samp{foo} which holds
4613 the address of memory location 1000, but nothing special is stored at
4614 address 1000. This means that you cannot access the @emph{value} of a
4615 linker script defined symbol - it has no value - all you can do is
4616 access the @emph{address} of a linker script defined symbol.
4618 Hence when you are using a linker script defined symbol in source code
4619 you should always take the address of the symbol, and never attempt to
4620 use its value. For example suppose you want to copy the contents of a
4621 section of memory called .ROM into a section called .FLASH and the
4622 linker script contains these declarations:
4626 start_of_ROM = .ROM;
4627 end_of_ROM = .ROM + sizeof (.ROM);
4628 start_of_FLASH = .FLASH;
4632 Then the C source code to perform the copy would be:
4636 extern char start_of_ROM, end_of_ROM, start_of_FLASH;
4638 memcpy (& start_of_FLASH, & start_of_ROM, & end_of_ROM - & start_of_ROM);
4642 Note the use of the @samp{&} operators. These are correct.
4643 Alternatively the symbols can be treated as the names of vectors or
4644 arrays and then the code will again work as expected:
4648 extern char start_of_ROM[], end_of_ROM[], start_of_FLASH[];
4650 memcpy (start_of_FLASH, start_of_ROM, end_of_ROM - start_of_ROM);
4654 Note how using this method does not require the use of @samp{&}
4658 @section SECTIONS Command
4660 The @code{SECTIONS} command tells the linker how to map input sections
4661 into output sections, and how to place the output sections in memory.
4663 The format of the @code{SECTIONS} command is:
4667 @var{sections-command}
4668 @var{sections-command}
4673 Each @var{sections-command} may of be one of the following:
4677 an @code{ENTRY} command (@pxref{Entry Point,,Entry command})
4679 a symbol assignment (@pxref{Assignments})
4681 an output section description
4683 an overlay description
4686 The @code{ENTRY} command and symbol assignments are permitted inside the
4687 @code{SECTIONS} command for convenience in using the location counter in
4688 those commands. This can also make the linker script easier to
4689 understand because you can use those commands at meaningful points in
4690 the layout of the output file.
4692 Output section descriptions and overlay descriptions are described
4695 If you do not use a @code{SECTIONS} command in your linker script, the
4696 linker will place each input section into an identically named output
4697 section in the order that the sections are first encountered in the
4698 input files. If all input sections are present in the first file, for
4699 example, the order of sections in the output file will match the order
4700 in the first input file. The first section will be at address zero.
4703 * Output Section Description:: Output section description
4704 * Output Section Name:: Output section name
4705 * Output Section Address:: Output section address
4706 * Input Section:: Input section description
4707 * Output Section Data:: Output section data
4708 * Output Section Keywords:: Output section keywords
4709 * Output Section Discarding:: Output section discarding
4710 * Output Section Attributes:: Output section attributes
4711 * Overlay Description:: Overlay description
4714 @node Output Section Description
4715 @subsection Output Section Description
4716 The full description of an output section looks like this:
4719 @var{section} [@var{address}] [(@var{type})] :
4721 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
4722 [SUBALIGN(@var{subsection_align})]
4725 @var{output-section-command}
4726 @var{output-section-command}
4728 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}] [,]
4732 Most output sections do not use most of the optional section attributes.
4734 The whitespace around @var{section} is required, so that the section
4735 name is unambiguous. The colon and the curly braces are also required.
4736 The comma at the end may be required if a @var{fillexp} is used and
4737 the next @var{sections-command} looks like a continuation of the expression.
4738 The line breaks and other white space are optional.
4740 Each @var{output-section-command} may be one of the following:
4744 a symbol assignment (@pxref{Assignments})
4746 an input section description (@pxref{Input Section})
4748 data values to include directly (@pxref{Output Section Data})
4750 a special output section keyword (@pxref{Output Section Keywords})
4753 @node Output Section Name
4754 @subsection Output Section Name
4755 @cindex name, section
4756 @cindex section name
4757 The name of the output section is @var{section}. @var{section} must
4758 meet the constraints of your output format. In formats which only
4759 support a limited number of sections, such as @code{a.out}, the name
4760 must be one of the names supported by the format (@code{a.out}, for
4761 example, allows only @samp{.text}, @samp{.data} or @samp{.bss}). If the
4762 output format supports any number of sections, but with numbers and not
4763 names (as is the case for Oasys), the name should be supplied as a
4764 quoted numeric string. A section name may consist of any sequence of
4765 characters, but a name which contains any unusual characters such as
4766 commas must be quoted.
4768 The output section name @samp{/DISCARD/} is special; @ref{Output Section
4771 @node Output Section Address
4772 @subsection Output Section Address
4773 @cindex address, section
4774 @cindex section address
4775 The @var{address} is an expression for the VMA (the virtual memory
4776 address) of the output section. This address is optional, but if it
4777 is provided then the output address will be set exactly as specified.
4779 If the output address is not specified then one will be chosen for the
4780 section, based on the heuristic below. This address will be adjusted
4781 to fit the alignment requirement of the output section. The
4782 alignment requirement is the strictest alignment of any input section
4783 contained within the output section.
4785 The output section address heuristic is as follows:
4789 If an output memory @var{region} is set for the section then it
4790 is added to this region and its address will be the next free address
4794 If the MEMORY command has been used to create a list of memory
4795 regions then the first region which has attributes compatible with the
4796 section is selected to contain it. The section's output address will
4797 be the next free address in that region; @ref{MEMORY}.
4800 If no memory regions were specified, or none match the section then
4801 the output address will be based on the current value of the location
4809 .text . : @{ *(.text) @}
4816 .text : @{ *(.text) @}
4820 are subtly different. The first will set the address of the
4821 @samp{.text} output section to the current value of the location
4822 counter. The second will set it to the current value of the location
4823 counter aligned to the strictest alignment of any of the @samp{.text}
4826 The @var{address} may be an arbitrary expression; @ref{Expressions}.
4827 For example, if you want to align the section on a 0x10 byte boundary,
4828 so that the lowest four bits of the section address are zero, you could
4829 do something like this:
4831 .text ALIGN(0x10) : @{ *(.text) @}
4834 This works because @code{ALIGN} returns the current location counter
4835 aligned upward to the specified value.
4837 Specifying @var{address} for a section will change the value of the
4838 location counter, provided that the section is non-empty. (Empty
4839 sections are ignored).
4842 @subsection Input Section Description
4843 @cindex input sections
4844 @cindex mapping input sections to output sections
4845 The most common output section command is an input section description.
4847 The input section description is the most basic linker script operation.
4848 You use output sections to tell the linker how to lay out your program
4849 in memory. You use input section descriptions to tell the linker how to
4850 map the input files into your memory layout.
4853 * Input Section Basics:: Input section basics
4854 * Input Section Wildcards:: Input section wildcard patterns
4855 * Input Section Common:: Input section for common symbols
4856 * Input Section Keep:: Input section and garbage collection
4857 * Input Section Example:: Input section example
4860 @node Input Section Basics
4861 @subsubsection Input Section Basics
4862 @cindex input section basics
4863 An input section description consists of a file name optionally followed
4864 by a list of section names in parentheses.
4866 The file name and the section name may be wildcard patterns, which we
4867 describe further below (@pxref{Input Section Wildcards}).
4869 The most common input section description is to include all input
4870 sections with a particular name in the output section. For example, to
4871 include all input @samp{.text} sections, you would write:
4876 Here the @samp{*} is a wildcard which matches any file name. To exclude a list
4877 @cindex EXCLUDE_FILE
4878 of files from matching the file name wildcard, EXCLUDE_FILE may be used to
4879 match all files except the ones specified in the EXCLUDE_FILE list. For
4882 EXCLUDE_FILE (*crtend.o *otherfile.o) *(.ctors)
4885 will cause all .ctors sections from all files except @file{crtend.o}
4886 and @file{otherfile.o} to be included. The EXCLUDE_FILE can also be
4887 placed inside the section list, for example:
4889 *(EXCLUDE_FILE (*crtend.o *otherfile.o) .ctors)
4892 The result of this is identically to the previous example. Supporting
4893 two syntaxes for EXCLUDE_FILE is useful if the section list contains
4894 more than one section, as described below.
4896 There are two ways to include more than one section:
4902 The difference between these is the order in which the @samp{.text} and
4903 @samp{.rdata} input sections will appear in the output section. In the
4904 first example, they will be intermingled, appearing in the same order as
4905 they are found in the linker input. In the second example, all
4906 @samp{.text} input sections will appear first, followed by all
4907 @samp{.rdata} input sections.
4909 When using EXCLUDE_FILE with more than one section, if the exclusion
4910 is within the section list then the exclusion only applies to the
4911 immediately following section, for example:
4913 *(EXCLUDE_FILE (*somefile.o) .text .rdata)
4916 will cause all @samp{.text} sections from all files except
4917 @file{somefile.o} to be included, while all @samp{.rdata} sections
4918 from all files, including @file{somefile.o}, will be included. To
4919 exclude the @samp{.rdata} sections from @file{somefile.o} the example
4920 could be modified to:
4922 *(EXCLUDE_FILE (*somefile.o) .text EXCLUDE_FILE (*somefile.o) .rdata)
4925 Alternatively, placing the EXCLUDE_FILE outside of the section list,
4926 before the input file selection, will cause the exclusion to apply for
4927 all sections. Thus the previous example can be rewritten as:
4929 EXCLUDE_FILE (*somefile.o) *(.text .rdata)
4932 You can specify a file name to include sections from a particular file.
4933 You would do this if one or more of your files contain special data that
4934 needs to be at a particular location in memory. For example:
4939 To refine the sections that are included based on the section flags
4940 of an input section, INPUT_SECTION_FLAGS may be used.
4942 Here is a simple example for using Section header flags for ELF sections:
4947 .text : @{ INPUT_SECTION_FLAGS (SHF_MERGE & SHF_STRINGS) *(.text) @}
4948 .text2 : @{ INPUT_SECTION_FLAGS (!SHF_WRITE) *(.text) @}
4953 In this example, the output section @samp{.text} will be comprised of any
4954 input section matching the name *(.text) whose section header flags
4955 @code{SHF_MERGE} and @code{SHF_STRINGS} are set. The output section
4956 @samp{.text2} will be comprised of any input section matching the name *(.text)
4957 whose section header flag @code{SHF_WRITE} is clear.
4959 You can also specify files within archives by writing a pattern
4960 matching the archive, a colon, then the pattern matching the file,
4961 with no whitespace around the colon.
4965 matches file within archive
4967 matches the whole archive
4969 matches file but not one in an archive
4972 Either one or both of @samp{archive} and @samp{file} can contain shell
4973 wildcards. On DOS based file systems, the linker will assume that a
4974 single letter followed by a colon is a drive specifier, so
4975 @samp{c:myfile.o} is a simple file specification, not @samp{myfile.o}
4976 within an archive called @samp{c}. @samp{archive:file} filespecs may
4977 also be used within an @code{EXCLUDE_FILE} list, but may not appear in
4978 other linker script contexts. For instance, you cannot extract a file
4979 from an archive by using @samp{archive:file} in an @code{INPUT}
4982 If you use a file name without a list of sections, then all sections in
4983 the input file will be included in the output section. This is not
4984 commonly done, but it may by useful on occasion. For example:
4989 When you use a file name which is not an @samp{archive:file} specifier
4990 and does not contain any wild card
4991 characters, the linker will first see if you also specified the file
4992 name on the linker command line or in an @code{INPUT} command. If you
4993 did not, the linker will attempt to open the file as an input file, as
4994 though it appeared on the command line. Note that this differs from an
4995 @code{INPUT} command, because the linker will not search for the file in
4996 the archive search path.
4998 @node Input Section Wildcards
4999 @subsubsection Input Section Wildcard Patterns
5000 @cindex input section wildcards
5001 @cindex wildcard file name patterns
5002 @cindex file name wildcard patterns
5003 @cindex section name wildcard patterns
5004 In an input section description, either the file name or the section
5005 name or both may be wildcard patterns.
5007 The file name of @samp{*} seen in many examples is a simple wildcard
5008 pattern for the file name.
5010 The wildcard patterns are like those used by the Unix shell.
5014 matches any number of characters
5016 matches any single character
5018 matches a single instance of any of the @var{chars}; the @samp{-}
5019 character may be used to specify a range of characters, as in
5020 @samp{[a-z]} to match any lower case letter
5022 quotes the following character
5025 File name wildcard patterns only match files which are explicitly
5026 specified on the command line or in an @code{INPUT} command. The linker
5027 does not search directories to expand wildcards.
5029 If a file name matches more than one wildcard pattern, or if a file name
5030 appears explicitly and is also matched by a wildcard pattern, the linker
5031 will use the first match in the linker script. For example, this
5032 sequence of input section descriptions is probably in error, because the
5033 @file{data.o} rule will not be used:
5035 .data : @{ *(.data) @}
5036 .data1 : @{ data.o(.data) @}
5039 @cindex SORT_BY_NAME
5040 Normally, the linker will place files and sections matched by wildcards
5041 in the order in which they are seen during the link. You can change
5042 this by using the @code{SORT_BY_NAME} keyword, which appears before a wildcard
5043 pattern in parentheses (e.g., @code{SORT_BY_NAME(.text*)}). When the
5044 @code{SORT_BY_NAME} keyword is used, the linker will sort the files or sections
5045 into ascending order by name before placing them in the output file.
5047 @cindex SORT_BY_ALIGNMENT
5048 @code{SORT_BY_ALIGNMENT} is similar to @code{SORT_BY_NAME}.
5049 @code{SORT_BY_ALIGNMENT} will sort sections into descending order of
5050 alignment before placing them in the output file. Placing larger
5051 alignments before smaller alignments can reduce the amount of padding
5054 @cindex SORT_BY_INIT_PRIORITY
5055 @code{SORT_BY_INIT_PRIORITY} is also similar to @code{SORT_BY_NAME}.
5056 @code{SORT_BY_INIT_PRIORITY} will sort sections into ascending
5057 numerical order of the GCC init_priority attribute encoded in the
5058 section name before placing them in the output file. In
5059 @code{.init_array.NNNNN} and @code{.fini_array.NNNNN}, @code{NNNNN} is
5060 the init_priority. In @code{.ctors.NNNNN} and @code{.dtors.NNNNN},
5061 @code{NNNNN} is 65535 minus the init_priority.
5064 @code{SORT} is an alias for @code{SORT_BY_NAME}.
5066 When there are nested section sorting commands in linker script, there
5067 can be at most 1 level of nesting for section sorting commands.
5071 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
5072 It will sort the input sections by name first, then by alignment if two
5073 sections have the same name.
5075 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
5076 It will sort the input sections by alignment first, then by name if two
5077 sections have the same alignment.
5079 @code{SORT_BY_NAME} (@code{SORT_BY_NAME} (wildcard section pattern)) is
5080 treated the same as @code{SORT_BY_NAME} (wildcard section pattern).
5082 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern))
5083 is treated the same as @code{SORT_BY_ALIGNMENT} (wildcard section pattern).
5085 All other nested section sorting commands are invalid.
5088 When both command-line section sorting option and linker script
5089 section sorting command are used, section sorting command always
5090 takes precedence over the command-line option.
5092 If the section sorting command in linker script isn't nested, the
5093 command-line option will make the section sorting command to be
5094 treated as nested sorting command.
5098 @code{SORT_BY_NAME} (wildcard section pattern ) with
5099 @option{--sort-sections alignment} is equivalent to
5100 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
5102 @code{SORT_BY_ALIGNMENT} (wildcard section pattern) with
5103 @option{--sort-section name} is equivalent to
5104 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
5107 If the section sorting command in linker script is nested, the
5108 command-line option will be ignored.
5111 @code{SORT_NONE} disables section sorting by ignoring the command-line
5112 section sorting option.
5114 If you ever get confused about where input sections are going, use the
5115 @samp{-M} linker option to generate a map file. The map file shows
5116 precisely how input sections are mapped to output sections.
5118 This example shows how wildcard patterns might be used to partition
5119 files. This linker script directs the linker to place all @samp{.text}
5120 sections in @samp{.text} and all @samp{.bss} sections in @samp{.bss}.
5121 The linker will place the @samp{.data} section from all files beginning
5122 with an upper case character in @samp{.DATA}; for all other files, the
5123 linker will place the @samp{.data} section in @samp{.data}.
5127 .text : @{ *(.text) @}
5128 .DATA : @{ [A-Z]*(.data) @}
5129 .data : @{ *(.data) @}
5130 .bss : @{ *(.bss) @}
5135 @node Input Section Common
5136 @subsubsection Input Section for Common Symbols
5137 @cindex common symbol placement
5138 @cindex uninitialized data placement
5139 A special notation is needed for common symbols, because in many object
5140 file formats common symbols do not have a particular input section. The
5141 linker treats common symbols as though they are in an input section
5142 named @samp{COMMON}.
5144 You may use file names with the @samp{COMMON} section just as with any
5145 other input sections. You can use this to place common symbols from a
5146 particular input file in one section while common symbols from other
5147 input files are placed in another section.
5149 In most cases, common symbols in input files will be placed in the
5150 @samp{.bss} section in the output file. For example:
5152 .bss @{ *(.bss) *(COMMON) @}
5155 @cindex scommon section
5156 @cindex small common symbols
5157 Some object file formats have more than one type of common symbol. For
5158 example, the MIPS ELF object file format distinguishes standard common
5159 symbols and small common symbols. In this case, the linker will use a
5160 different special section name for other types of common symbols. In
5161 the case of MIPS ELF, the linker uses @samp{COMMON} for standard common
5162 symbols and @samp{.scommon} for small common symbols. This permits you
5163 to map the different types of common symbols into memory at different
5167 You will sometimes see @samp{[COMMON]} in old linker scripts. This
5168 notation is now considered obsolete. It is equivalent to
5171 @node Input Section Keep
5172 @subsubsection Input Section and Garbage Collection
5174 @cindex garbage collection
5175 When link-time garbage collection is in use (@samp{--gc-sections}),
5176 it is often useful to mark sections that should not be eliminated.
5177 This is accomplished by surrounding an input section's wildcard entry
5178 with @code{KEEP()}, as in @code{KEEP(*(.init))} or
5179 @code{KEEP(SORT_BY_NAME(*)(.ctors))}.
5181 @node Input Section Example
5182 @subsubsection Input Section Example
5183 The following example is a complete linker script. It tells the linker
5184 to read all of the sections from file @file{all.o} and place them at the
5185 start of output section @samp{outputa} which starts at location
5186 @samp{0x10000}. All of section @samp{.input1} from file @file{foo.o}
5187 follows immediately, in the same output section. All of section
5188 @samp{.input2} from @file{foo.o} goes into output section
5189 @samp{outputb}, followed by section @samp{.input1} from @file{foo1.o}.
5190 All of the remaining @samp{.input1} and @samp{.input2} sections from any
5191 files are written to output section @samp{outputc}.
5219 If an output section's name is the same as the input section's name
5220 and is representable as a C identifier, then the linker will
5221 automatically @pxref{PROVIDE} two symbols: __start_SECNAME and
5222 __stop_SECNAME, where SECNAME is the name of the section. These
5223 indicate the start address and end address of the output section
5224 respectively. Note: most section names are not representable as
5225 C identifiers because they contain a @samp{.} character.
5227 @node Output Section Data
5228 @subsection Output Section Data
5230 @cindex section data
5231 @cindex output section data
5232 @kindex BYTE(@var{expression})
5233 @kindex SHORT(@var{expression})
5234 @kindex LONG(@var{expression})
5235 @kindex QUAD(@var{expression})
5236 @kindex SQUAD(@var{expression})
5237 You can include explicit bytes of data in an output section by using
5238 @code{BYTE}, @code{SHORT}, @code{LONG}, @code{QUAD}, or @code{SQUAD} as
5239 an output section command. Each keyword is followed by an expression in
5240 parentheses providing the value to store (@pxref{Expressions}). The
5241 value of the expression is stored at the current value of the location
5244 The @code{BYTE}, @code{SHORT}, @code{LONG}, and @code{QUAD} commands
5245 store one, two, four, and eight bytes (respectively). After storing the
5246 bytes, the location counter is incremented by the number of bytes
5249 For example, this will store the byte 1 followed by the four byte value
5250 of the symbol @samp{addr}:
5256 When using a 64 bit host or target, @code{QUAD} and @code{SQUAD} are the
5257 same; they both store an 8 byte, or 64 bit, value. When both host and
5258 target are 32 bits, an expression is computed as 32 bits. In this case
5259 @code{QUAD} stores a 32 bit value zero extended to 64 bits, and
5260 @code{SQUAD} stores a 32 bit value sign extended to 64 bits.
5262 If the object file format of the output file has an explicit endianness,
5263 which is the normal case, the value will be stored in that endianness.
5264 When the object file format does not have an explicit endianness, as is
5265 true of, for example, S-records, the value will be stored in the
5266 endianness of the first input object file.
5268 Note---these commands only work inside a section description and not
5269 between them, so the following will produce an error from the linker:
5271 SECTIONS @{@ .text : @{@ *(.text) @}@ LONG(1) .data : @{@ *(.data) @}@ @}@
5273 whereas this will work:
5275 SECTIONS @{@ .text : @{@ *(.text) ; LONG(1) @}@ .data : @{@ *(.data) @}@ @}@
5278 @kindex FILL(@var{expression})
5279 @cindex holes, filling
5280 @cindex unspecified memory
5281 You may use the @code{FILL} command to set the fill pattern for the
5282 current section. It is followed by an expression in parentheses. Any
5283 otherwise unspecified regions of memory within the section (for example,
5284 gaps left due to the required alignment of input sections) are filled
5285 with the value of the expression, repeated as
5286 necessary. A @code{FILL} statement covers memory locations after the
5287 point at which it occurs in the section definition; by including more
5288 than one @code{FILL} statement, you can have different fill patterns in
5289 different parts of an output section.
5291 This example shows how to fill unspecified regions of memory with the
5297 The @code{FILL} command is similar to the @samp{=@var{fillexp}} output
5298 section attribute, but it only affects the
5299 part of the section following the @code{FILL} command, rather than the
5300 entire section. If both are used, the @code{FILL} command takes
5301 precedence. @xref{Output Section Fill}, for details on the fill
5304 @node Output Section Keywords
5305 @subsection Output Section Keywords
5306 There are a couple of keywords which can appear as output section
5310 @kindex CREATE_OBJECT_SYMBOLS
5311 @cindex input filename symbols
5312 @cindex filename symbols
5313 @item CREATE_OBJECT_SYMBOLS
5314 The command tells the linker to create a symbol for each input file.
5315 The name of each symbol will be the name of the corresponding input
5316 file. The section of each symbol will be the output section in which
5317 the @code{CREATE_OBJECT_SYMBOLS} command appears.
5319 This is conventional for the a.out object file format. It is not
5320 normally used for any other object file format.
5322 @kindex CONSTRUCTORS
5323 @cindex C++ constructors, arranging in link
5324 @cindex constructors, arranging in link
5326 When linking using the a.out object file format, the linker uses an
5327 unusual set construct to support C++ global constructors and
5328 destructors. When linking object file formats which do not support
5329 arbitrary sections, such as ECOFF and XCOFF, the linker will
5330 automatically recognize C++ global constructors and destructors by name.
5331 For these object file formats, the @code{CONSTRUCTORS} command tells the
5332 linker to place constructor information in the output section where the
5333 @code{CONSTRUCTORS} command appears. The @code{CONSTRUCTORS} command is
5334 ignored for other object file formats.
5336 The symbol @w{@code{__CTOR_LIST__}} marks the start of the global
5337 constructors, and the symbol @w{@code{__CTOR_END__}} marks the end.
5338 Similarly, @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_END__}} mark
5339 the start and end of the global destructors. The
5340 first word in the list is the number of entries, followed by the address
5341 of each constructor or destructor, followed by a zero word. The
5342 compiler must arrange to actually run the code. For these object file
5343 formats @sc{gnu} C++ normally calls constructors from a subroutine
5344 @code{__main}; a call to @code{__main} is automatically inserted into
5345 the startup code for @code{main}. @sc{gnu} C++ normally runs
5346 destructors either by using @code{atexit}, or directly from the function
5349 For object file formats such as @code{COFF} or @code{ELF} which support
5350 arbitrary section names, @sc{gnu} C++ will normally arrange to put the
5351 addresses of global constructors and destructors into the @code{.ctors}
5352 and @code{.dtors} sections. Placing the following sequence into your
5353 linker script will build the sort of table which the @sc{gnu} C++
5354 runtime code expects to see.
5358 LONG((__CTOR_END__ - __CTOR_LIST__) / 4 - 2)
5363 LONG((__DTOR_END__ - __DTOR_LIST__) / 4 - 2)
5369 If you are using the @sc{gnu} C++ support for initialization priority,
5370 which provides some control over the order in which global constructors
5371 are run, you must sort the constructors at link time to ensure that they
5372 are executed in the correct order. When using the @code{CONSTRUCTORS}
5373 command, use @samp{SORT_BY_NAME(CONSTRUCTORS)} instead. When using the
5374 @code{.ctors} and @code{.dtors} sections, use @samp{*(SORT_BY_NAME(.ctors))} and
5375 @samp{*(SORT_BY_NAME(.dtors))} instead of just @samp{*(.ctors)} and
5378 Normally the compiler and linker will handle these issues automatically,
5379 and you will not need to concern yourself with them. However, you may
5380 need to consider this if you are using C++ and writing your own linker
5385 @node Output Section Discarding
5386 @subsection Output Section Discarding
5387 @cindex discarding sections
5388 @cindex sections, discarding
5389 @cindex removing sections
5390 The linker will not normally create output sections with no contents.
5391 This is for convenience when referring to input sections that may or
5392 may not be present in any of the input files. For example:
5394 .foo : @{ *(.foo) @}
5397 will only create a @samp{.foo} section in the output file if there is a
5398 @samp{.foo} section in at least one input file, and if the input
5399 sections are not all empty. Other link script directives that allocate
5400 space in an output section will also create the output section. So
5401 too will assignments to dot even if the assignment does not create
5402 space, except for @samp{. = 0}, @samp{. = . + 0}, @samp{. = sym},
5403 @samp{. = . + sym} and @samp{. = ALIGN (. != 0, expr, 1)} when
5404 @samp{sym} is an absolute symbol of value 0 defined in the script.
5405 This allows you to force output of an empty section with @samp{. = .}.
5407 The linker will ignore address assignments (@pxref{Output Section Address})
5408 on discarded output sections, except when the linker script defines
5409 symbols in the output section. In that case the linker will obey
5410 the address assignments, possibly advancing dot even though the
5411 section is discarded.
5414 The special output section name @samp{/DISCARD/} may be used to discard
5415 input sections. Any input sections which are assigned to an output
5416 section named @samp{/DISCARD/} are not included in the output file.
5418 This can be used to discard input sections marked with the ELF flag
5419 @code{SHF_GNU_RETAIN}, which would otherwise have been saved from linker
5422 Note, sections that match the @samp{/DISCARD/} output section will be
5423 discarded even if they are in an ELF section group which has other
5424 members which are not being discarded. This is deliberate.
5425 Discarding takes precedence over grouping.
5427 @node Output Section Attributes
5428 @subsection Output Section Attributes
5429 @cindex output section attributes
5430 We showed above that the full description of an output section looked
5435 @var{section} [@var{address}] [(@var{type})] :
5437 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
5438 [SUBALIGN(@var{subsection_align})]
5441 @var{output-section-command}
5442 @var{output-section-command}
5444 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}]
5448 We've already described @var{section}, @var{address}, and
5449 @var{output-section-command}. In this section we will describe the
5450 remaining section attributes.
5453 * Output Section Type:: Output section type
5454 * Output Section LMA:: Output section LMA
5455 * Forced Output Alignment:: Forced Output Alignment
5456 * Forced Input Alignment:: Forced Input Alignment
5457 * Output Section Constraint:: Output section constraint
5458 * Output Section Region:: Output section region
5459 * Output Section Phdr:: Output section phdr
5460 * Output Section Fill:: Output section fill
5463 @node Output Section Type
5464 @subsubsection Output Section Type
5465 Each output section may have a type. The type is a keyword in
5466 parentheses. The following types are defined:
5470 The section should be marked as not loadable, so that it will not be
5471 loaded into memory when the program is run.
5473 The section should be marked as read-only.
5478 These type names are supported for backward compatibility, and are
5479 rarely used. They all have the same effect: the section should be
5480 marked as not allocatable, so that no memory is allocated for the
5481 section when the program is run.
5485 @cindex prevent unnecessary loading
5486 @cindex loading, preventing
5487 The linker normally sets the attributes of an output section based on
5488 the input sections which map into it. You can override this by using
5489 the section type. For example, in the script sample below, the
5490 @samp{ROM} section is addressed at memory location @samp{0} and does not
5491 need to be loaded when the program is run.
5495 ROM 0 (NOLOAD) : @{ @dots{} @}
5501 @node Output Section LMA
5502 @subsubsection Output Section LMA
5503 @kindex AT>@var{lma_region}
5504 @kindex AT(@var{lma})
5505 @cindex load address
5506 @cindex section load address
5507 Every section has a virtual address (VMA) and a load address (LMA); see
5508 @ref{Basic Script Concepts}. The virtual address is specified by the
5509 @pxref{Output Section Address} described earlier. The load address is
5510 specified by the @code{AT} or @code{AT>} keywords. Specifying a load
5511 address is optional.
5513 The @code{AT} keyword takes an expression as an argument. This
5514 specifies the exact load address of the section. The @code{AT>} keyword
5515 takes the name of a memory region as an argument. @xref{MEMORY}. The
5516 load address of the section is set to the next free address in the
5517 region, aligned to the section's alignment requirements.
5519 If neither @code{AT} nor @code{AT>} is specified for an allocatable
5520 section, the linker will use the following heuristic to determine the
5525 If the section has a specific VMA address, then this is used as
5526 the LMA address as well.
5529 If the section is not allocatable then its LMA is set to its VMA.
5532 Otherwise if a memory region can be found that is compatible
5533 with the current section, and this region contains at least one
5534 section, then the LMA is set so the difference between the
5535 VMA and LMA is the same as the difference between the VMA and LMA of
5536 the last section in the located region.
5539 If no memory regions have been declared then a default region
5540 that covers the entire address space is used in the previous step.
5543 If no suitable region could be found, or there was no previous
5544 section then the LMA is set equal to the VMA.
5547 @cindex ROM initialized data
5548 @cindex initialized data in ROM
5549 This feature is designed to make it easy to build a ROM image. For
5550 example, the following linker script creates three output sections: one
5551 called @samp{.text}, which starts at @code{0x1000}, one called
5552 @samp{.mdata}, which is loaded at the end of the @samp{.text} section
5553 even though its VMA is @code{0x2000}, and one called @samp{.bss} to hold
5554 uninitialized data at address @code{0x3000}. The symbol @code{_data} is
5555 defined with the value @code{0x2000}, which shows that the location
5556 counter holds the VMA value, not the LMA value.
5562 .text 0x1000 : @{ *(.text) _etext = . ; @}
5564 AT ( ADDR (.text) + SIZEOF (.text) )
5565 @{ _data = . ; *(.data); _edata = . ; @}
5567 @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@}
5572 The run-time initialization code for use with a program generated with
5573 this linker script would include something like the following, to copy
5574 the initialized data from the ROM image to its runtime address. Notice
5575 how this code takes advantage of the symbols defined by the linker
5580 extern char _etext, _data, _edata, _bstart, _bend;
5581 char *src = &_etext;
5584 /* ROM has data at end of text; copy it. */
5585 while (dst < &_edata)
5589 for (dst = &_bstart; dst< &_bend; dst++)
5594 @node Forced Output Alignment
5595 @subsubsection Forced Output Alignment
5596 @kindex ALIGN(@var{section_align})
5597 @cindex forcing output section alignment
5598 @cindex output section alignment
5599 You can increase an output section's alignment by using ALIGN. As an
5600 alternative you can enforce that the difference between the VMA and LMA remains
5601 intact throughout this output section with the ALIGN_WITH_INPUT attribute.
5603 @node Forced Input Alignment
5604 @subsubsection Forced Input Alignment
5605 @kindex SUBALIGN(@var{subsection_align})
5606 @cindex forcing input section alignment
5607 @cindex input section alignment
5608 You can force input section alignment within an output section by using
5609 SUBALIGN. The value specified overrides any alignment given by input
5610 sections, whether larger or smaller.
5612 @node Output Section Constraint
5613 @subsubsection Output Section Constraint
5616 @cindex constraints on output sections
5617 You can specify that an output section should only be created if all
5618 of its input sections are read-only or all of its input sections are
5619 read-write by using the keyword @code{ONLY_IF_RO} and
5620 @code{ONLY_IF_RW} respectively.
5622 @node Output Section Region
5623 @subsubsection Output Section Region
5624 @kindex >@var{region}
5625 @cindex section, assigning to memory region
5626 @cindex memory regions and sections
5627 You can assign a section to a previously defined region of memory by
5628 using @samp{>@var{region}}. @xref{MEMORY}.
5630 Here is a simple example:
5633 MEMORY @{ rom : ORIGIN = 0x1000, LENGTH = 0x1000 @}
5634 SECTIONS @{ ROM : @{ *(.text) @} >rom @}
5638 @node Output Section Phdr
5639 @subsubsection Output Section Phdr
5641 @cindex section, assigning to program header
5642 @cindex program headers and sections
5643 You can assign a section to a previously defined program segment by
5644 using @samp{:@var{phdr}}. @xref{PHDRS}. If a section is assigned to
5645 one or more segments, then all subsequent allocated sections will be
5646 assigned to those segments as well, unless they use an explicitly
5647 @code{:@var{phdr}} modifier. You can use @code{:NONE} to tell the
5648 linker to not put the section in any segment at all.
5650 Here is a simple example:
5653 PHDRS @{ text PT_LOAD ; @}
5654 SECTIONS @{ .text : @{ *(.text) @} :text @}
5658 @node Output Section Fill
5659 @subsubsection Output Section Fill
5660 @kindex =@var{fillexp}
5661 @cindex section fill pattern
5662 @cindex fill pattern, entire section
5663 You can set the fill pattern for an entire section by using
5664 @samp{=@var{fillexp}}. @var{fillexp} is an expression
5665 (@pxref{Expressions}). Any otherwise unspecified regions of memory
5666 within the output section (for example, gaps left due to the required
5667 alignment of input sections) will be filled with the value, repeated as
5668 necessary. If the fill expression is a simple hex number, ie. a string
5669 of hex digit starting with @samp{0x} and without a trailing @samp{k} or @samp{M}, then
5670 an arbitrarily long sequence of hex digits can be used to specify the
5671 fill pattern; Leading zeros become part of the pattern too. For all
5672 other cases, including extra parentheses or a unary @code{+}, the fill
5673 pattern is the four least significant bytes of the value of the
5674 expression. In all cases, the number is big-endian.
5676 You can also change the fill value with a @code{FILL} command in the
5677 output section commands; (@pxref{Output Section Data}).
5679 Here is a simple example:
5682 SECTIONS @{ .text : @{ *(.text) @} =0x90909090 @}
5686 @node Overlay Description
5687 @subsection Overlay Description
5690 An overlay description provides an easy way to describe sections which
5691 are to be loaded as part of a single memory image but are to be run at
5692 the same memory address. At run time, some sort of overlay manager will
5693 copy the overlaid sections in and out of the runtime memory address as
5694 required, perhaps by simply manipulating addressing bits. This approach
5695 can be useful, for example, when a certain region of memory is faster
5698 Overlays are described using the @code{OVERLAY} command. The
5699 @code{OVERLAY} command is used within a @code{SECTIONS} command, like an
5700 output section description. The full syntax of the @code{OVERLAY}
5701 command is as follows:
5704 OVERLAY [@var{start}] : [NOCROSSREFS] [AT ( @var{ldaddr} )]
5708 @var{output-section-command}
5709 @var{output-section-command}
5711 @} [:@var{phdr}@dots{}] [=@var{fill}]
5714 @var{output-section-command}
5715 @var{output-section-command}
5717 @} [:@var{phdr}@dots{}] [=@var{fill}]
5719 @} [>@var{region}] [:@var{phdr}@dots{}] [=@var{fill}] [,]
5723 Everything is optional except @code{OVERLAY} (a keyword), and each
5724 section must have a name (@var{secname1} and @var{secname2} above). The
5725 section definitions within the @code{OVERLAY} construct are identical to
5726 those within the general @code{SECTIONS} construct (@pxref{SECTIONS}),
5727 except that no addresses and no memory regions may be defined for
5728 sections within an @code{OVERLAY}.
5730 The comma at the end may be required if a @var{fill} is used and
5731 the next @var{sections-command} looks like a continuation of the expression.
5733 The sections are all defined with the same starting address. The load
5734 addresses of the sections are arranged such that they are consecutive in
5735 memory starting at the load address used for the @code{OVERLAY} as a
5736 whole (as with normal section definitions, the load address is optional,
5737 and defaults to the start address; the start address is also optional,
5738 and defaults to the current value of the location counter).
5740 If the @code{NOCROSSREFS} keyword is used, and there are any
5741 references among the sections, the linker will report an error. Since
5742 the sections all run at the same address, it normally does not make
5743 sense for one section to refer directly to another.
5744 @xref{Miscellaneous Commands, NOCROSSREFS}.
5746 For each section within the @code{OVERLAY}, the linker automatically
5747 provides two symbols. The symbol @code{__load_start_@var{secname}} is
5748 defined as the starting load address of the section. The symbol
5749 @code{__load_stop_@var{secname}} is defined as the final load address of
5750 the section. Any characters within @var{secname} which are not legal
5751 within C identifiers are removed. C (or assembler) code may use these
5752 symbols to move the overlaid sections around as necessary.
5754 At the end of the overlay, the value of the location counter is set to
5755 the start address of the overlay plus the size of the largest section.
5757 Here is an example. Remember that this would appear inside a
5758 @code{SECTIONS} construct.
5761 OVERLAY 0x1000 : AT (0x4000)
5763 .text0 @{ o1/*.o(.text) @}
5764 .text1 @{ o2/*.o(.text) @}
5769 This will define both @samp{.text0} and @samp{.text1} to start at
5770 address 0x1000. @samp{.text0} will be loaded at address 0x4000, and
5771 @samp{.text1} will be loaded immediately after @samp{.text0}. The
5772 following symbols will be defined if referenced: @code{__load_start_text0},
5773 @code{__load_stop_text0}, @code{__load_start_text1},
5774 @code{__load_stop_text1}.
5776 C code to copy overlay @code{.text1} into the overlay area might look
5781 extern char __load_start_text1, __load_stop_text1;
5782 memcpy ((char *) 0x1000, &__load_start_text1,
5783 &__load_stop_text1 - &__load_start_text1);
5787 Note that the @code{OVERLAY} command is just syntactic sugar, since
5788 everything it does can be done using the more basic commands. The above
5789 example could have been written identically as follows.
5793 .text0 0x1000 : AT (0x4000) @{ o1/*.o(.text) @}
5794 PROVIDE (__load_start_text0 = LOADADDR (.text0));
5795 PROVIDE (__load_stop_text0 = LOADADDR (.text0) + SIZEOF (.text0));
5796 .text1 0x1000 : AT (0x4000 + SIZEOF (.text0)) @{ o2/*.o(.text) @}
5797 PROVIDE (__load_start_text1 = LOADADDR (.text1));
5798 PROVIDE (__load_stop_text1 = LOADADDR (.text1) + SIZEOF (.text1));
5799 . = 0x1000 + MAX (SIZEOF (.text0), SIZEOF (.text1));
5804 @section MEMORY Command
5806 @cindex memory regions
5807 @cindex regions of memory
5808 @cindex allocating memory
5809 @cindex discontinuous memory
5810 The linker's default configuration permits allocation of all available
5811 memory. You can override this by using the @code{MEMORY} command.
5813 The @code{MEMORY} command describes the location and size of blocks of
5814 memory in the target. You can use it to describe which memory regions
5815 may be used by the linker, and which memory regions it must avoid. You
5816 can then assign sections to particular memory regions. The linker will
5817 set section addresses based on the memory regions, and will warn about
5818 regions that become too full. The linker will not shuffle sections
5819 around to fit into the available regions.
5821 A linker script may contain many uses of the @code{MEMORY} command,
5822 however, all memory blocks defined are treated as if they were
5823 specified inside a single @code{MEMORY} command. The syntax for
5829 @var{name} [(@var{attr})] : ORIGIN = @var{origin}, LENGTH = @var{len}
5835 The @var{name} is a name used in the linker script to refer to the
5836 region. The region name has no meaning outside of the linker script.
5837 Region names are stored in a separate name space, and will not conflict
5838 with symbol names, file names, or section names. Each memory region
5839 must have a distinct name within the @code{MEMORY} command. However you can
5840 add later alias names to existing memory regions with the @ref{REGION_ALIAS}
5843 @cindex memory region attributes
5844 The @var{attr} string is an optional list of attributes that specify
5845 whether to use a particular memory region for an input section which is
5846 not explicitly mapped in the linker script. As described in
5847 @ref{SECTIONS}, if you do not specify an output section for some input
5848 section, the linker will create an output section with the same name as
5849 the input section. If you define region attributes, the linker will use
5850 them to select the memory region for the output section that it creates.
5852 The @var{attr} string must consist only of the following characters:
5867 Invert the sense of any of the attributes that follow
5870 If an unmapped section matches any of the listed attributes other than
5871 @samp{!}, it will be placed in the memory region. The @samp{!}
5872 attribute reverses the test for the characters that follow, so that an
5873 unmapped section will be placed in the memory region only if it does
5874 not match any of the attributes listed afterwards. Thus an attribute
5875 string of @samp{RW!X} will match any unmapped section that has either
5876 or both of the @samp{R} and @samp{W} attributes, but only as long as
5877 the section does not also have the @samp{X} attribute.
5882 The @var{origin} is an numerical expression for the start address of
5883 the memory region. The expression must evaluate to a constant and it
5884 cannot involve any symbols. The keyword @code{ORIGIN} may be
5885 abbreviated to @code{org} or @code{o} (but not, for example,
5891 The @var{len} is an expression for the size in bytes of the memory
5892 region. As with the @var{origin} expression, the expression must
5893 be numerical only and must evaluate to a constant. The keyword
5894 @code{LENGTH} may be abbreviated to @code{len} or @code{l}.
5896 In the following example, we specify that there are two memory regions
5897 available for allocation: one starting at @samp{0} for 256 kilobytes,
5898 and the other starting at @samp{0x40000000} for four megabytes. The
5899 linker will place into the @samp{rom} memory region every section which
5900 is not explicitly mapped into a memory region, and is either read-only
5901 or executable. The linker will place other sections which are not
5902 explicitly mapped into a memory region into the @samp{ram} memory
5909 rom (rx) : ORIGIN = 0, LENGTH = 256K
5910 ram (!rx) : org = 0x40000000, l = 4M
5915 Once you define a memory region, you can direct the linker to place
5916 specific output sections into that memory region by using the
5917 @samp{>@var{region}} output section attribute. For example, if you have
5918 a memory region named @samp{mem}, you would use @samp{>mem} in the
5919 output section definition. @xref{Output Section Region}. If no address
5920 was specified for the output section, the linker will set the address to
5921 the next available address within the memory region. If the combined
5922 output sections directed to a memory region are too large for the
5923 region, the linker will issue an error message.
5925 It is possible to access the origin and length of a memory in an
5926 expression via the @code{ORIGIN(@var{memory})} and
5927 @code{LENGTH(@var{memory})} functions:
5931 _fstack = ORIGIN(ram) + LENGTH(ram) - 4;
5936 @section PHDRS Command
5938 @cindex program headers
5939 @cindex ELF program headers
5940 @cindex program segments
5941 @cindex segments, ELF
5942 The ELF object file format uses @dfn{program headers}, also knows as
5943 @dfn{segments}. The program headers describe how the program should be
5944 loaded into memory. You can print them out by using the @code{objdump}
5945 program with the @samp{-p} option.
5947 When you run an ELF program on a native ELF system, the system loader
5948 reads the program headers in order to figure out how to load the
5949 program. This will only work if the program headers are set correctly.
5950 This manual does not describe the details of how the system loader
5951 interprets program headers; for more information, see the ELF ABI.
5953 The linker will create reasonable program headers by default. However,
5954 in some cases, you may need to specify the program headers more
5955 precisely. You may use the @code{PHDRS} command for this purpose. When
5956 the linker sees the @code{PHDRS} command in the linker script, it will
5957 not create any program headers other than the ones specified.
5959 The linker only pays attention to the @code{PHDRS} command when
5960 generating an ELF output file. In other cases, the linker will simply
5961 ignore @code{PHDRS}.
5963 This is the syntax of the @code{PHDRS} command. The words @code{PHDRS},
5964 @code{FILEHDR}, @code{AT}, and @code{FLAGS} are keywords.
5970 @var{name} @var{type} [ FILEHDR ] [ PHDRS ] [ AT ( @var{address} ) ]
5971 [ FLAGS ( @var{flags} ) ] ;
5976 The @var{name} is used only for reference in the @code{SECTIONS} command
5977 of the linker script. It is not put into the output file. Program
5978 header names are stored in a separate name space, and will not conflict
5979 with symbol names, file names, or section names. Each program header
5980 must have a distinct name. The headers are processed in order and it
5981 is usual for them to map to sections in ascending load address order.
5983 Certain program header types describe segments of memory which the
5984 system loader will load from the file. In the linker script, you
5985 specify the contents of these segments by placing allocatable output
5986 sections in the segments. You use the @samp{:@var{phdr}} output section
5987 attribute to place a section in a particular segment. @xref{Output
5990 It is normal to put certain sections in more than one segment. This
5991 merely implies that one segment of memory contains another. You may
5992 repeat @samp{:@var{phdr}}, using it once for each segment which should
5993 contain the section.
5995 If you place a section in one or more segments using @samp{:@var{phdr}},
5996 then the linker will place all subsequent allocatable sections which do
5997 not specify @samp{:@var{phdr}} in the same segments. This is for
5998 convenience, since generally a whole set of contiguous sections will be
5999 placed in a single segment. You can use @code{:NONE} to override the
6000 default segment and tell the linker to not put the section in any
6005 You may use the @code{FILEHDR} and @code{PHDRS} keywords after
6006 the program header type to further describe the contents of the segment.
6007 The @code{FILEHDR} keyword means that the segment should include the ELF
6008 file header. The @code{PHDRS} keyword means that the segment should
6009 include the ELF program headers themselves. If applied to a loadable
6010 segment (@code{PT_LOAD}), all prior loadable segments must have one of
6013 The @var{type} may be one of the following. The numbers indicate the
6014 value of the keyword.
6017 @item @code{PT_NULL} (0)
6018 Indicates an unused program header.
6020 @item @code{PT_LOAD} (1)
6021 Indicates that this program header describes a segment to be loaded from
6024 @item @code{PT_DYNAMIC} (2)
6025 Indicates a segment where dynamic linking information can be found.
6027 @item @code{PT_INTERP} (3)
6028 Indicates a segment where the name of the program interpreter may be
6031 @item @code{PT_NOTE} (4)
6032 Indicates a segment holding note information.
6034 @item @code{PT_SHLIB} (5)
6035 A reserved program header type, defined but not specified by the ELF
6038 @item @code{PT_PHDR} (6)
6039 Indicates a segment where the program headers may be found.
6041 @item @code{PT_TLS} (7)
6042 Indicates a segment containing thread local storage.
6044 @item @var{expression}
6045 An expression giving the numeric type of the program header. This may
6046 be used for types not defined above.
6049 You can specify that a segment should be loaded at a particular address
6050 in memory by using an @code{AT} expression. This is identical to the
6051 @code{AT} command used as an output section attribute (@pxref{Output
6052 Section LMA}). The @code{AT} command for a program header overrides the
6053 output section attribute.
6055 The linker will normally set the segment flags based on the sections
6056 which comprise the segment. You may use the @code{FLAGS} keyword to
6057 explicitly specify the segment flags. The value of @var{flags} must be
6058 an integer. It is used to set the @code{p_flags} field of the program
6061 Here is an example of @code{PHDRS}. This shows a typical set of program
6062 headers used on a native ELF system.
6068 headers PT_PHDR PHDRS ;
6070 text PT_LOAD FILEHDR PHDRS ;
6072 dynamic PT_DYNAMIC ;
6078 .interp : @{ *(.interp) @} :text :interp
6079 .text : @{ *(.text) @} :text
6080 .rodata : @{ *(.rodata) @} /* defaults to :text */
6082 . = . + 0x1000; /* move to a new page in memory */
6083 .data : @{ *(.data) @} :data
6084 .dynamic : @{ *(.dynamic) @} :data :dynamic
6091 @section VERSION Command
6092 @kindex VERSION @{script text@}
6093 @cindex symbol versions
6094 @cindex version script
6095 @cindex versions of symbols
6096 The linker supports symbol versions when using ELF. Symbol versions are
6097 only useful when using shared libraries. The dynamic linker can use
6098 symbol versions to select a specific version of a function when it runs
6099 a program that may have been linked against an earlier version of the
6102 You can include a version script directly in the main linker script, or
6103 you can supply the version script as an implicit linker script. You can
6104 also use the @samp{--version-script} linker option.
6106 The syntax of the @code{VERSION} command is simply
6108 VERSION @{ version-script-commands @}
6111 The format of the version script commands is identical to that used by
6112 Sun's linker in Solaris 2.5. The version script defines a tree of
6113 version nodes. You specify the node names and interdependencies in the
6114 version script. You can specify which symbols are bound to which
6115 version nodes, and you can reduce a specified set of symbols to local
6116 scope so that they are not globally visible outside of the shared
6119 The easiest way to demonstrate the version script language is with a few
6145 This example version script defines three version nodes. The first
6146 version node defined is @samp{VERS_1.1}; it has no other dependencies.
6147 The script binds the symbol @samp{foo1} to @samp{VERS_1.1}. It reduces
6148 a number of symbols to local scope so that they are not visible outside
6149 of the shared library; this is done using wildcard patterns, so that any
6150 symbol whose name begins with @samp{old}, @samp{original}, or @samp{new}
6151 is matched. The wildcard patterns available are the same as those used
6152 in the shell when matching filenames (also known as ``globbing'').
6153 However, if you specify the symbol name inside double quotes, then the
6154 name is treated as literal, rather than as a glob pattern.
6156 Next, the version script defines node @samp{VERS_1.2}. This node
6157 depends upon @samp{VERS_1.1}. The script binds the symbol @samp{foo2}
6158 to the version node @samp{VERS_1.2}.
6160 Finally, the version script defines node @samp{VERS_2.0}. This node
6161 depends upon @samp{VERS_1.2}. The scripts binds the symbols @samp{bar1}
6162 and @samp{bar2} are bound to the version node @samp{VERS_2.0}.
6164 When the linker finds a symbol defined in a library which is not
6165 specifically bound to a version node, it will effectively bind it to an
6166 unspecified base version of the library. You can bind all otherwise
6167 unspecified symbols to a given version node by using @samp{global: *;}
6168 somewhere in the version script. Note that it's slightly crazy to use
6169 wildcards in a global spec except on the last version node. Global
6170 wildcards elsewhere run the risk of accidentally adding symbols to the
6171 set exported for an old version. That's wrong since older versions
6172 ought to have a fixed set of symbols.
6174 The names of the version nodes have no specific meaning other than what
6175 they might suggest to the person reading them. The @samp{2.0} version
6176 could just as well have appeared in between @samp{1.1} and @samp{1.2}.
6177 However, this would be a confusing way to write a version script.
6179 Node name can be omitted, provided it is the only version node
6180 in the version script. Such version script doesn't assign any versions to
6181 symbols, only selects which symbols will be globally visible out and which
6185 @{ global: foo; bar; local: *; @};
6188 When you link an application against a shared library that has versioned
6189 symbols, the application itself knows which version of each symbol it
6190 requires, and it also knows which version nodes it needs from each
6191 shared library it is linked against. Thus at runtime, the dynamic
6192 loader can make a quick check to make sure that the libraries you have
6193 linked against do in fact supply all of the version nodes that the
6194 application will need to resolve all of the dynamic symbols. In this
6195 way it is possible for the dynamic linker to know with certainty that
6196 all external symbols that it needs will be resolvable without having to
6197 search for each symbol reference.
6199 The symbol versioning is in effect a much more sophisticated way of
6200 doing minor version checking that SunOS does. The fundamental problem
6201 that is being addressed here is that typically references to external
6202 functions are bound on an as-needed basis, and are not all bound when
6203 the application starts up. If a shared library is out of date, a
6204 required interface may be missing; when the application tries to use
6205 that interface, it may suddenly and unexpectedly fail. With symbol
6206 versioning, the user will get a warning when they start their program if
6207 the libraries being used with the application are too old.
6209 There are several GNU extensions to Sun's versioning approach. The
6210 first of these is the ability to bind a symbol to a version node in the
6211 source file where the symbol is defined instead of in the versioning
6212 script. This was done mainly to reduce the burden on the library
6213 maintainer. You can do this by putting something like:
6215 __asm__(".symver original_foo,foo@@VERS_1.1");
6218 in the C source file. This renames the function @samp{original_foo} to
6219 be an alias for @samp{foo} bound to the version node @samp{VERS_1.1}.
6220 The @samp{local:} directive can be used to prevent the symbol
6221 @samp{original_foo} from being exported. A @samp{.symver} directive
6222 takes precedence over a version script.
6224 The second GNU extension is to allow multiple versions of the same
6225 function to appear in a given shared library. In this way you can make
6226 an incompatible change to an interface without increasing the major
6227 version number of the shared library, while still allowing applications
6228 linked against the old interface to continue to function.
6230 To do this, you must use multiple @samp{.symver} directives in the
6231 source file. Here is an example:
6234 __asm__(".symver original_foo,foo@@");
6235 __asm__(".symver old_foo,foo@@VERS_1.1");
6236 __asm__(".symver old_foo1,foo@@VERS_1.2");
6237 __asm__(".symver new_foo,foo@@@@VERS_2.0");
6240 In this example, @samp{foo@@} represents the symbol @samp{foo} bound to the
6241 unspecified base version of the symbol. The source file that contains this
6242 example would define 4 C functions: @samp{original_foo}, @samp{old_foo},
6243 @samp{old_foo1}, and @samp{new_foo}.
6245 When you have multiple definitions of a given symbol, there needs to be
6246 some way to specify a default version to which external references to
6247 this symbol will be bound. You can do this with the
6248 @samp{foo@@@@VERS_2.0} type of @samp{.symver} directive. You can only
6249 declare one version of a symbol as the default in this manner; otherwise
6250 you would effectively have multiple definitions of the same symbol.
6252 If you wish to bind a reference to a specific version of the symbol
6253 within the shared library, you can use the aliases of convenience
6254 (i.e., @samp{old_foo}), or you can use the @samp{.symver} directive to
6255 specifically bind to an external version of the function in question.
6257 You can also specify the language in the version script:
6260 VERSION extern "lang" @{ version-script-commands @}
6263 The supported @samp{lang}s are @samp{C}, @samp{C++}, and @samp{Java}.
6264 The linker will iterate over the list of symbols at the link time and
6265 demangle them according to @samp{lang} before matching them to the
6266 patterns specified in @samp{version-script-commands}. The default
6267 @samp{lang} is @samp{C}.
6269 Demangled names may contains spaces and other special characters. As
6270 described above, you can use a glob pattern to match demangled names,
6271 or you can use a double-quoted string to match the string exactly. In
6272 the latter case, be aware that minor differences (such as differing
6273 whitespace) between the version script and the demangler output will
6274 cause a mismatch. As the exact string generated by the demangler
6275 might change in the future, even if the mangled name does not, you
6276 should check that all of your version directives are behaving as you
6277 expect when you upgrade.
6280 @section Expressions in Linker Scripts
6283 The syntax for expressions in the linker script language is identical to
6284 that of C expressions, except that whitespace is required in some
6285 places to resolve syntactic ambiguities. All expressions are
6286 evaluated as integers. All expressions are evaluated in the same
6287 size, which is 32 bits if both the host and target are 32 bits, and is
6290 You can use and set symbol values in expressions.
6292 The linker defines several special purpose builtin functions for use in
6296 * Constants:: Constants
6297 * Symbolic Constants:: Symbolic constants
6298 * Symbols:: Symbol Names
6299 * Orphan Sections:: Orphan Sections
6300 * Location Counter:: The Location Counter
6301 * Operators:: Operators
6302 * Evaluation:: Evaluation
6303 * Expression Section:: The Section of an Expression
6304 * Builtin Functions:: Builtin Functions
6308 @subsection Constants
6309 @cindex integer notation
6310 @cindex constants in linker scripts
6311 All constants are integers.
6313 As in C, the linker considers an integer beginning with @samp{0} to be
6314 octal, and an integer beginning with @samp{0x} or @samp{0X} to be
6315 hexadecimal. Alternatively the linker accepts suffixes of @samp{h} or
6316 @samp{H} for hexadecimal, @samp{o} or @samp{O} for octal, @samp{b} or
6317 @samp{B} for binary and @samp{d} or @samp{D} for decimal. Any integer
6318 value without a prefix or a suffix is considered to be decimal.
6320 @cindex scaled integers
6321 @cindex K and M integer suffixes
6322 @cindex M and K integer suffixes
6323 @cindex suffixes for integers
6324 @cindex integer suffixes
6325 In addition, you can use the suffixes @code{K} and @code{M} to scale a
6329 @c END TEXI2ROFF-KILL
6330 @code{1024} or @code{1024*1024}
6334 ${\rm 1024}$ or ${\rm 1024}^2$
6336 @c END TEXI2ROFF-KILL
6337 respectively. For example, the following
6338 all refer to the same quantity:
6347 Note - the @code{K} and @code{M} suffixes cannot be used in
6348 conjunction with the base suffixes mentioned above.
6350 @node Symbolic Constants
6351 @subsection Symbolic Constants
6352 @cindex symbolic constants
6354 It is possible to refer to target-specific constants via the use of
6355 the @code{CONSTANT(@var{name})} operator, where @var{name} is one of:
6360 The target's maximum page size.
6362 @item COMMONPAGESIZE
6363 @kindex COMMONPAGESIZE
6364 The target's default page size.
6370 .text ALIGN (CONSTANT (MAXPAGESIZE)) : @{ *(.text) @}
6373 will create a text section aligned to the largest page boundary
6374 supported by the target.
6377 @subsection Symbol Names
6378 @cindex symbol names
6380 @cindex quoted symbol names
6382 Unless quoted, symbol names start with a letter, underscore, or period
6383 and may include letters, digits, underscores, periods, and hyphens.
6384 Unquoted symbol names must not conflict with any keywords. You can
6385 specify a symbol which contains odd characters or has the same name as a
6386 keyword by surrounding the symbol name in double quotes:
6389 "with a space" = "also with a space" + 10;
6392 Since symbols can contain many non-alphabetic characters, it is safest
6393 to delimit symbols with spaces. For example, @samp{A-B} is one symbol,
6394 whereas @samp{A - B} is an expression involving subtraction.
6396 @node Orphan Sections
6397 @subsection Orphan Sections
6399 Orphan sections are sections present in the input files which
6400 are not explicitly placed into the output file by the linker
6401 script. The linker will still copy these sections into the
6402 output file by either finding, or creating a suitable output section
6403 in which to place the orphaned input section.
6405 If the name of an orphaned input section exactly matches the name of
6406 an existing output section, then the orphaned input section will be
6407 placed at the end of that output section.
6409 If there is no output section with a matching name then new output
6410 sections will be created. Each new output section will have the same
6411 name as the orphan section placed within it. If there are multiple
6412 orphan sections with the same name, these will all be combined into
6413 one new output section.
6415 If new output sections are created to hold orphaned input sections,
6416 then the linker must decide where to place these new output sections
6417 in relation to existing output sections. On most modern targets, the
6418 linker attempts to place orphan sections after sections of the same
6419 attribute, such as code vs data, loadable vs non-loadable, etc. If no
6420 sections with matching attributes are found, or your target lacks this
6421 support, the orphan section is placed at the end of the file.
6423 The command-line options @samp{--orphan-handling} and @samp{--unique}
6424 (@pxref{Options,,Command-line Options}) can be used to control which
6425 output sections an orphan is placed in.
6427 @node Location Counter
6428 @subsection The Location Counter
6431 @cindex location counter
6432 @cindex current output location
6433 The special linker variable @dfn{dot} @samp{.} always contains the
6434 current output location counter. Since the @code{.} always refers to a
6435 location in an output section, it may only appear in an expression
6436 within a @code{SECTIONS} command. The @code{.} symbol may appear
6437 anywhere that an ordinary symbol is allowed in an expression.
6440 Assigning a value to @code{.} will cause the location counter to be
6441 moved. This may be used to create holes in the output section. The
6442 location counter may not be moved backwards inside an output section,
6443 and may not be moved backwards outside of an output section if so
6444 doing creates areas with overlapping LMAs.
6460 In the previous example, the @samp{.text} section from @file{file1} is
6461 located at the beginning of the output section @samp{output}. It is
6462 followed by a 1000 byte gap. Then the @samp{.text} section from
6463 @file{file2} appears, also with a 1000 byte gap following before the
6464 @samp{.text} section from @file{file3}. The notation @samp{= 0x12345678}
6465 specifies what data to write in the gaps (@pxref{Output Section Fill}).
6467 @cindex dot inside sections
6468 Note: @code{.} actually refers to the byte offset from the start of the
6469 current containing object. Normally this is the @code{SECTIONS}
6470 statement, whose start address is 0, hence @code{.} can be used as an
6471 absolute address. If @code{.} is used inside a section description
6472 however, it refers to the byte offset from the start of that section,
6473 not an absolute address. Thus in a script like this:
6491 The @samp{.text} section will be assigned a starting address of 0x100
6492 and a size of exactly 0x200 bytes, even if there is not enough data in
6493 the @samp{.text} input sections to fill this area. (If there is too
6494 much data, an error will be produced because this would be an attempt to
6495 move @code{.} backwards). The @samp{.data} section will start at 0x500
6496 and it will have an extra 0x600 bytes worth of space after the end of
6497 the values from the @samp{.data} input sections and before the end of
6498 the @samp{.data} output section itself.
6500 @cindex dot outside sections
6501 Setting symbols to the value of the location counter outside of an
6502 output section statement can result in unexpected values if the linker
6503 needs to place orphan sections. For example, given the following:
6509 .text: @{ *(.text) @}
6513 .data: @{ *(.data) @}
6518 If the linker needs to place some input section, e.g. @code{.rodata},
6519 not mentioned in the script, it might choose to place that section
6520 between @code{.text} and @code{.data}. You might think the linker
6521 should place @code{.rodata} on the blank line in the above script, but
6522 blank lines are of no particular significance to the linker. As well,
6523 the linker doesn't associate the above symbol names with their
6524 sections. Instead, it assumes that all assignments or other
6525 statements belong to the previous output section, except for the
6526 special case of an assignment to @code{.}. I.e., the linker will
6527 place the orphan @code{.rodata} section as if the script was written
6534 .text: @{ *(.text) @}
6538 .rodata: @{ *(.rodata) @}
6539 .data: @{ *(.data) @}
6544 This may or may not be the script author's intention for the value of
6545 @code{start_of_data}. One way to influence the orphan section
6546 placement is to assign the location counter to itself, as the linker
6547 assumes that an assignment to @code{.} is setting the start address of
6548 a following output section and thus should be grouped with that
6549 section. So you could write:
6555 .text: @{ *(.text) @}
6560 .data: @{ *(.data) @}
6565 Now, the orphan @code{.rodata} section will be placed between
6566 @code{end_of_text} and @code{start_of_data}.
6570 @subsection Operators
6571 @cindex operators for arithmetic
6572 @cindex arithmetic operators
6573 @cindex precedence in expressions
6574 The linker recognizes the standard C set of arithmetic operators, with
6575 the standard bindings and precedence levels:
6578 @c END TEXI2ROFF-KILL
6580 precedence associativity Operators Notes
6586 5 left == != > < <= >=
6592 11 right &= += -= *= /= (2)
6596 (1) Prefix operators
6597 (2) @xref{Assignments}.
6601 \vskip \baselineskip
6602 %"lispnarrowing" is the extra indent used generally for smallexample
6603 \hskip\lispnarrowing\vbox{\offinterlineskip
6606 {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr
6607 height2pt&\omit&&\omit&&\omit&\cr
6608 &Precedence&& Associativity &&{\rm Operators}&\cr
6609 height2pt&\omit&&\omit&&\omit&\cr
6611 height2pt&\omit&&\omit&&\omit&\cr
6613 % '176 is tilde, '~' in tt font
6614 &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr
6615 &2&&left&&* / \%&\cr
6618 &5&&left&&== != > < <= >=&\cr
6621 &8&&left&&{\&\&}&\cr
6624 &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr
6626 height2pt&\omit&&\omit&&\omit&\cr}
6631 @obeylines@parskip=0pt@parindent=0pt
6632 @dag@quad Prefix operators.
6633 @ddag@quad @xref{Assignments}.
6636 @c END TEXI2ROFF-KILL
6639 @subsection Evaluation
6640 @cindex lazy evaluation
6641 @cindex expression evaluation order
6642 The linker evaluates expressions lazily. It only computes the value of
6643 an expression when absolutely necessary.
6645 The linker needs some information, such as the value of the start
6646 address of the first section, and the origins and lengths of memory
6647 regions, in order to do any linking at all. These values are computed
6648 as soon as possible when the linker reads in the linker script.
6650 However, other values (such as symbol values) are not known or needed
6651 until after storage allocation. Such values are evaluated later, when
6652 other information (such as the sizes of output sections) is available
6653 for use in the symbol assignment expression.
6655 The sizes of sections cannot be known until after allocation, so
6656 assignments dependent upon these are not performed until after
6659 Some expressions, such as those depending upon the location counter
6660 @samp{.}, must be evaluated during section allocation.
6662 If the result of an expression is required, but the value is not
6663 available, then an error results. For example, a script like the
6669 .text 9+this_isnt_constant :
6675 will cause the error message @samp{non constant expression for initial
6678 @node Expression Section
6679 @subsection The Section of an Expression
6680 @cindex expression sections
6681 @cindex absolute expressions
6682 @cindex relative expressions
6683 @cindex absolute and relocatable symbols
6684 @cindex relocatable and absolute symbols
6685 @cindex symbols, relocatable and absolute
6686 Addresses and symbols may be section relative, or absolute. A section
6687 relative symbol is relocatable. If you request relocatable output
6688 using the @samp{-r} option, a further link operation may change the
6689 value of a section relative symbol. On the other hand, an absolute
6690 symbol will retain the same value throughout any further link
6693 Some terms in linker expressions are addresses. This is true of
6694 section relative symbols and for builtin functions that return an
6695 address, such as @code{ADDR}, @code{LOADADDR}, @code{ORIGIN} and
6696 @code{SEGMENT_START}. Other terms are simply numbers, or are builtin
6697 functions that return a non-address value, such as @code{LENGTH}.
6698 One complication is that unless you set @code{LD_FEATURE ("SANE_EXPR")}
6699 (@pxref{Miscellaneous Commands}), numbers and absolute symbols are treated
6700 differently depending on their location, for compatibility with older
6701 versions of @code{ld}. Expressions appearing outside an output
6702 section definition treat all numbers as absolute addresses.
6703 Expressions appearing inside an output section definition treat
6704 absolute symbols as numbers. If @code{LD_FEATURE ("SANE_EXPR")} is
6705 given, then absolute symbols and numbers are simply treated as numbers
6708 In the following simple example,
6715 __executable_start = 0x100;
6719 __data_start = 0x10;
6727 both @code{.} and @code{__executable_start} are set to the absolute
6728 address 0x100 in the first two assignments, then both @code{.} and
6729 @code{__data_start} are set to 0x10 relative to the @code{.data}
6730 section in the second two assignments.
6732 For expressions involving numbers, relative addresses and absolute
6733 addresses, ld follows these rules to evaluate terms:
6737 Unary operations on an absolute address or number, and binary
6738 operations on two absolute addresses or two numbers, or between one
6739 absolute address and a number, apply the operator to the value(s).
6741 Unary operations on a relative address, and binary operations on two
6742 relative addresses in the same section or between one relative address
6743 and a number, apply the operator to the offset part of the address(es).
6745 Other binary operations, that is, between two relative addresses not
6746 in the same section, or between a relative address and an absolute
6747 address, first convert any non-absolute term to an absolute address
6748 before applying the operator.
6751 The result section of each sub-expression is as follows:
6755 An operation involving only numbers results in a number.
6757 The result of comparisons, @samp{&&} and @samp{||} is also a number.
6759 The result of other binary arithmetic and logical operations on two
6760 relative addresses in the same section or two absolute addresses
6761 (after above conversions) is also a number when
6762 @code{LD_FEATURE ("SANE_EXPR")} or inside an output section definition
6763 but an absolute address otherwise.
6765 The result of other operations on relative addresses or one
6766 relative address and a number, is a relative address in the same
6767 section as the relative operand(s).
6769 The result of other operations on absolute addresses (after above
6770 conversions) is an absolute address.
6773 You can use the builtin function @code{ABSOLUTE} to force an expression
6774 to be absolute when it would otherwise be relative. For example, to
6775 create an absolute symbol set to the address of the end of the output
6776 section @samp{.data}:
6780 .data : @{ *(.data) _edata = ABSOLUTE(.); @}
6784 If @samp{ABSOLUTE} were not used, @samp{_edata} would be relative to the
6785 @samp{.data} section.
6787 Using @code{LOADADDR} also forces an expression absolute, since this
6788 particular builtin function returns an absolute address.
6790 @node Builtin Functions
6791 @subsection Builtin Functions
6792 @cindex functions in expressions
6793 The linker script language includes a number of builtin functions for
6794 use in linker script expressions.
6797 @item ABSOLUTE(@var{exp})
6798 @kindex ABSOLUTE(@var{exp})
6799 @cindex expression, absolute
6800 Return the absolute (non-relocatable, as opposed to non-negative) value
6801 of the expression @var{exp}. Primarily useful to assign an absolute
6802 value to a symbol within a section definition, where symbol values are
6803 normally section relative. @xref{Expression Section}.
6805 @item ADDR(@var{section})
6806 @kindex ADDR(@var{section})
6807 @cindex section address in expression
6808 Return the address (VMA) of the named @var{section}. Your
6809 script must previously have defined the location of that section. In
6810 the following example, @code{start_of_output_1}, @code{symbol_1} and
6811 @code{symbol_2} are assigned equivalent values, except that
6812 @code{symbol_1} will be relative to the @code{.output1} section while
6813 the other two will be absolute:
6819 start_of_output_1 = ABSOLUTE(.);
6824 symbol_1 = ADDR(.output1);
6825 symbol_2 = start_of_output_1;
6831 @item ALIGN(@var{align})
6832 @itemx ALIGN(@var{exp},@var{align})
6833 @kindex ALIGN(@var{align})
6834 @kindex ALIGN(@var{exp},@var{align})
6835 @cindex round up location counter
6836 @cindex align location counter
6837 @cindex round up expression
6838 @cindex align expression
6839 Return the location counter (@code{.}) or arbitrary expression aligned
6840 to the next @var{align} boundary. The single operand @code{ALIGN}
6841 doesn't change the value of the location counter---it just does
6842 arithmetic on it. The two operand @code{ALIGN} allows an arbitrary
6843 expression to be aligned upwards (@code{ALIGN(@var{align})} is
6844 equivalent to @code{ALIGN(ABSOLUTE(.), @var{align})}).
6846 Here is an example which aligns the output @code{.data} section to the
6847 next @code{0x2000} byte boundary after the preceding section and sets a
6848 variable within the section to the next @code{0x8000} boundary after the
6853 .data ALIGN(0x2000): @{
6855 variable = ALIGN(0x8000);
6861 The first use of @code{ALIGN} in this example specifies the location of
6862 a section because it is used as the optional @var{address} attribute of
6863 a section definition (@pxref{Output Section Address}). The second use
6864 of @code{ALIGN} is used to defines the value of a symbol.
6866 The builtin function @code{NEXT} is closely related to @code{ALIGN}.
6868 @item ALIGNOF(@var{section})
6869 @kindex ALIGNOF(@var{section})
6870 @cindex section alignment
6871 Return the alignment in bytes of the named @var{section}, if that section has
6872 been allocated. If the section has not been allocated when this is
6873 evaluated, the linker will report an error. In the following example,
6874 the alignment of the @code{.output} section is stored as the first
6875 value in that section.
6880 LONG (ALIGNOF (.output))
6887 @item BLOCK(@var{exp})
6888 @kindex BLOCK(@var{exp})
6889 This is a synonym for @code{ALIGN}, for compatibility with older linker
6890 scripts. It is most often seen when setting the address of an output
6893 @item DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6894 @kindex DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6895 This is equivalent to either
6897 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - 1)))
6901 (ALIGN(@var{maxpagesize})
6902 + ((. + @var{commonpagesize} - 1) & (@var{maxpagesize} - @var{commonpagesize})))
6905 depending on whether the latter uses fewer @var{commonpagesize} sized pages
6906 for the data segment (area between the result of this expression and
6907 @code{DATA_SEGMENT_END}) than the former or not.
6908 If the latter form is used, it means @var{commonpagesize} bytes of runtime
6909 memory will be saved at the expense of up to @var{commonpagesize} wasted
6910 bytes in the on-disk file.
6912 This expression can only be used directly in @code{SECTIONS} commands, not in
6913 any output section descriptions and only once in the linker script.
6914 @var{commonpagesize} should be less or equal to @var{maxpagesize} and should
6915 be the system page size the object wants to be optimized for while still
6916 running on system page sizes up to @var{maxpagesize}. Note however
6917 that @samp{-z relro} protection will not be effective if the system
6918 page size is larger than @var{commonpagesize}.
6923 . = DATA_SEGMENT_ALIGN(0x10000, 0x2000);
6926 @item DATA_SEGMENT_END(@var{exp})
6927 @kindex DATA_SEGMENT_END(@var{exp})
6928 This defines the end of data segment for @code{DATA_SEGMENT_ALIGN}
6929 evaluation purposes.
6932 . = DATA_SEGMENT_END(.);
6935 @item DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6936 @kindex DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6937 This defines the end of the @code{PT_GNU_RELRO} segment when
6938 @samp{-z relro} option is used.
6939 When @samp{-z relro} option is not present, @code{DATA_SEGMENT_RELRO_END}
6940 does nothing, otherwise @code{DATA_SEGMENT_ALIGN} is padded so that
6941 @var{exp} + @var{offset} is aligned to the @var{commonpagesize}
6942 argument given to @code{DATA_SEGMENT_ALIGN}. If present in the linker
6943 script, it must be placed between @code{DATA_SEGMENT_ALIGN} and
6944 @code{DATA_SEGMENT_END}. Evaluates to the second argument plus any
6945 padding needed at the end of the @code{PT_GNU_RELRO} segment due to
6949 . = DATA_SEGMENT_RELRO_END(24, .);
6952 @item DEFINED(@var{symbol})
6953 @kindex DEFINED(@var{symbol})
6954 @cindex symbol defaults
6955 Return 1 if @var{symbol} is in the linker global symbol table and is
6956 defined before the statement using DEFINED in the script, otherwise
6957 return 0. You can use this function to provide
6958 default values for symbols. For example, the following script fragment
6959 shows how to set a global symbol @samp{begin} to the first location in
6960 the @samp{.text} section---but if a symbol called @samp{begin} already
6961 existed, its value is preserved:
6967 begin = DEFINED(begin) ? begin : . ;
6975 @item LENGTH(@var{memory})
6976 @kindex LENGTH(@var{memory})
6977 Return the length of the memory region named @var{memory}.
6979 @item LOADADDR(@var{section})
6980 @kindex LOADADDR(@var{section})
6981 @cindex section load address in expression
6982 Return the absolute LMA of the named @var{section}. (@pxref{Output
6985 @item LOG2CEIL(@var{exp})
6986 @kindex LOG2CEIL(@var{exp})
6987 Return the binary logarithm of @var{exp} rounded towards infinity.
6988 @code{LOG2CEIL(0)} returns 0.
6991 @item MAX(@var{exp1}, @var{exp2})
6992 Returns the maximum of @var{exp1} and @var{exp2}.
6995 @item MIN(@var{exp1}, @var{exp2})
6996 Returns the minimum of @var{exp1} and @var{exp2}.
6998 @item NEXT(@var{exp})
6999 @kindex NEXT(@var{exp})
7000 @cindex unallocated address, next
7001 Return the next unallocated address that is a multiple of @var{exp}.
7002 This function is closely related to @code{ALIGN(@var{exp})}; unless you
7003 use the @code{MEMORY} command to define discontinuous memory for the
7004 output file, the two functions are equivalent.
7006 @item ORIGIN(@var{memory})
7007 @kindex ORIGIN(@var{memory})
7008 Return the origin of the memory region named @var{memory}.
7010 @item SEGMENT_START(@var{segment}, @var{default})
7011 @kindex SEGMENT_START(@var{segment}, @var{default})
7012 Return the base address of the named @var{segment}. If an explicit
7013 value has already been given for this segment (with a command-line
7014 @samp{-T} option) then that value will be returned otherwise the value
7015 will be @var{default}. At present, the @samp{-T} command-line option
7016 can only be used to set the base address for the ``text'', ``data'', and
7017 ``bss'' sections, but you can use @code{SEGMENT_START} with any segment
7020 @item SIZEOF(@var{section})
7021 @kindex SIZEOF(@var{section})
7022 @cindex section size
7023 Return the size in bytes of the named @var{section}, if that section has
7024 been allocated. If the section has not been allocated when this is
7025 evaluated, the linker will report an error. In the following example,
7026 @code{symbol_1} and @code{symbol_2} are assigned identical values:
7035 symbol_1 = .end - .start ;
7036 symbol_2 = SIZEOF(.output);
7041 @item SIZEOF_HEADERS
7042 @kindex SIZEOF_HEADERS
7044 Return the size in bytes of the output file's headers. This is
7045 information which appears at the start of the output file. You can use
7046 this number when setting the start address of the first section, if you
7047 choose, to facilitate paging.
7049 @cindex not enough room for program headers
7050 @cindex program headers, not enough room
7051 When producing an ELF output file, if the linker script uses the
7052 @code{SIZEOF_HEADERS} builtin function, the linker must compute the
7053 number of program headers before it has determined all the section
7054 addresses and sizes. If the linker later discovers that it needs
7055 additional program headers, it will report an error @samp{not enough
7056 room for program headers}. To avoid this error, you must avoid using
7057 the @code{SIZEOF_HEADERS} function, or you must rework your linker
7058 script to avoid forcing the linker to use additional program headers, or
7059 you must define the program headers yourself using the @code{PHDRS}
7060 command (@pxref{PHDRS}).
7063 @node Implicit Linker Scripts
7064 @section Implicit Linker Scripts
7065 @cindex implicit linker scripts
7066 If you specify a linker input file which the linker can not recognize as
7067 an object file or an archive file, it will try to read the file as a
7068 linker script. If the file can not be parsed as a linker script, the
7069 linker will report an error.
7071 An implicit linker script will not replace the default linker script.
7073 Typically an implicit linker script would contain only symbol
7074 assignments, or the @code{INPUT}, @code{GROUP}, or @code{VERSION}
7077 Any input files read because of an implicit linker script will be read
7078 at the position in the command line where the implicit linker script was
7079 read. This can affect archive searching.
7082 @chapter Linker Plugins
7085 @cindex linker plugins
7086 The linker can use dynamically loaded plugins to modify its behavior.
7087 For example, the link-time optimization feature that some compilers
7088 support is implemented with a linker plugin.
7090 Currently there is only one plugin shipped by default, but more may
7091 be added here later.
7094 * libdep Plugin:: Static Library Dependencies Plugin
7098 @section Static Library Dependencies Plugin
7099 @cindex static library dependencies
7100 Originally, static libraries were contained in an archive file consisting
7101 just of a collection of relocatable object files. Later they evolved to
7102 optionally include a symbol table, to assist in finding the needed objects
7103 within a library. There their evolution ended, and dynamic libraries
7106 One useful feature of dynamic libraries was that, more than just collecting
7107 multiple objects into a single file, they also included a list of their
7108 dependencies, such that one could specify just the name of a single dynamic
7109 library at link time, and all of its dependencies would be implicitly
7110 referenced as well. But static libraries lacked this feature, so if a
7111 link invocation was switched from using dynamic libraries to static
7112 libraries, the link command would usually fail unless it was rewritten to
7113 explicitly list the dependencies of the static library.
7115 The GNU @command{ar} utility now supports a @option{--record-libdeps} option
7116 to embed dependency lists into static libraries as well, and the @file{libdep}
7117 plugin may be used to read this dependency information at link time. The
7118 dependency information is stored as a single string, carrying @option{-l}
7119 and @option{-L} arguments as they would normally appear in a linker
7120 command line. As such, the information can be written with any text
7121 utility and stored into any archive, even if GNU @command{ar} is not
7122 being used to create the archive. The information is stored in an
7123 archive member named @samp{__.LIBDEP}.
7125 For example, given a library @file{libssl.a} that depends on another
7126 library @file{libcrypto.a} which may be found in @file{/usr/local/lib},
7127 the @samp{__.LIBDEP} member of @file{libssl.a} would contain
7130 -L/usr/local/lib -lcrypto
7134 @node Machine Dependent
7135 @chapter Machine Dependent Features
7137 @cindex machine dependencies
7138 @command{ld} has additional features on some platforms; the following
7139 sections describe them. Machines where @command{ld} has no additional
7140 functionality are not listed.
7144 * H8/300:: @command{ld} and the H8/300
7147 * M68HC11/68HC12:: @code{ld} and the Motorola 68HC11 and 68HC12 families
7150 * ARM:: @command{ld} and the ARM family
7153 * HPPA ELF32:: @command{ld} and HPPA 32-bit ELF
7156 * M68K:: @command{ld} and the Motorola 68K family
7159 * MIPS:: @command{ld} and the MIPS family
7162 * MMIX:: @command{ld} and MMIX
7165 * MSP430:: @command{ld} and MSP430
7168 * NDS32:: @command{ld} and NDS32
7171 * Nios II:: @command{ld} and the Altera Nios II
7174 * PowerPC ELF32:: @command{ld} and PowerPC 32-bit ELF Support
7177 * PowerPC64 ELF64:: @command{ld} and PowerPC64 64-bit ELF Support
7180 * S/390 ELF:: @command{ld} and S/390 ELF Support
7183 * SPU ELF:: @command{ld} and SPU ELF Support
7186 * TI COFF:: @command{ld} and TI COFF
7189 * WIN32:: @command{ld} and WIN32 (cygwin/mingw)
7192 * Xtensa:: @command{ld} and Xtensa Processors
7203 @section @command{ld} and the H8/300
7205 @cindex H8/300 support
7206 For the H8/300, @command{ld} can perform these global optimizations when
7207 you specify the @samp{--relax} command-line option.
7210 @cindex relaxing on H8/300
7211 @item relaxing address modes
7212 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
7213 targets are within eight bits, and turns them into eight-bit
7214 program-counter relative @code{bsr} and @code{bra} instructions,
7217 @cindex synthesizing on H8/300
7218 @item synthesizing instructions
7219 @c FIXME: specifically mov.b, or any mov instructions really? -> mov.b only, at least on H8, H8H, H8S
7220 @command{ld} finds all @code{mov.b} instructions which use the
7221 sixteen-bit absolute address form, but refer to the top
7222 page of memory, and changes them to use the eight-bit address form.
7223 (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into
7224 @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the
7225 top page of memory).
7227 @command{ld} finds all @code{mov} instructions which use the register
7228 indirect with 32-bit displacement addressing mode, but use a small
7229 displacement inside 16-bit displacement range, and changes them to use
7230 the 16-bit displacement form. (That is: the linker turns @samp{mov.b
7231 @code{@@}@var{d}:32,ERx} into @samp{mov.b @code{@@}@var{d}:16,ERx}
7232 whenever the displacement @var{d} is in the 16 bit signed integer
7233 range. Only implemented in ELF-format ld).
7235 @item bit manipulation instructions
7236 @command{ld} finds all bit manipulation instructions like @code{band, bclr,
7237 biand, bild, bior, bist, bixor, bld, bnot, bor, bset, bst, btst, bxor}
7238 which use 32 bit and 16 bit absolute address form, but refer to the top
7239 page of memory, and changes them to use the 8 bit address form.
7240 (That is: the linker turns @samp{bset #xx:3,@code{@@}@var{aa}:32} into
7241 @samp{bset #xx:3,@code{@@}@var{aa}:8} whenever the address @var{aa} is in
7242 the top page of memory).
7244 @item system control instructions
7245 @command{ld} finds all @code{ldc.w, stc.w} instructions which use the
7246 32 bit absolute address form, but refer to the top page of memory, and
7247 changes them to use 16 bit address form.
7248 (That is: the linker turns @samp{ldc.w @code{@@}@var{aa}:32,ccr} into
7249 @samp{ldc.w @code{@@}@var{aa}:16,ccr} whenever the address @var{aa} is in
7250 the top page of memory).
7260 @c This stuff is pointless to say unless you're especially concerned
7261 @c with Renesas chips; don't enable it for generic case, please.
7263 @chapter @command{ld} and Other Renesas Chips
7265 @command{ld} also supports the Renesas (formerly Hitachi) H8/300H,
7266 H8/500, and SH chips. No special features, commands, or command-line
7267 options are required for these chips.
7281 @node M68HC11/68HC12
7282 @section @command{ld} and the Motorola 68HC11 and 68HC12 families
7284 @cindex M68HC11 and 68HC12 support
7286 @subsection Linker Relaxation
7288 For the Motorola 68HC11, @command{ld} can perform these global
7289 optimizations when you specify the @samp{--relax} command-line option.
7292 @cindex relaxing on M68HC11
7293 @item relaxing address modes
7294 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
7295 targets are within eight bits, and turns them into eight-bit
7296 program-counter relative @code{bsr} and @code{bra} instructions,
7299 @command{ld} also looks at all 16-bit extended addressing modes and
7300 transforms them in a direct addressing mode when the address is in
7301 page 0 (between 0 and 0x0ff).
7303 @item relaxing gcc instruction group
7304 When @command{gcc} is called with @option{-mrelax}, it can emit group
7305 of instructions that the linker can optimize to use a 68HC11 direct
7306 addressing mode. These instructions consists of @code{bclr} or
7307 @code{bset} instructions.
7311 @subsection Trampoline Generation
7313 @cindex trampoline generation on M68HC11
7314 @cindex trampoline generation on M68HC12
7315 For 68HC11 and 68HC12, @command{ld} can generate trampoline code to
7316 call a far function using a normal @code{jsr} instruction. The linker
7317 will also change the relocation to some far function to use the
7318 trampoline address instead of the function address. This is typically the
7319 case when a pointer to a function is taken. The pointer will in fact
7320 point to the function trampoline.
7328 @section @command{ld} and the ARM family
7330 @cindex ARM interworking support
7331 @kindex --support-old-code
7332 For the ARM, @command{ld} will generate code stubs to allow functions calls
7333 between ARM and Thumb code. These stubs only work with code that has
7334 been compiled and assembled with the @samp{-mthumb-interwork} command
7335 line option. If it is necessary to link with old ARM object files or
7336 libraries, which have not been compiled with the -mthumb-interwork
7337 option then the @samp{--support-old-code} command-line switch should be
7338 given to the linker. This will make it generate larger stub functions
7339 which will work with non-interworking aware ARM code. Note, however,
7340 the linker does not support generating stubs for function calls to
7341 non-interworking aware Thumb code.
7343 @cindex thumb entry point
7344 @cindex entry point, thumb
7345 @kindex --thumb-entry=@var{entry}
7346 The @samp{--thumb-entry} switch is a duplicate of the generic
7347 @samp{--entry} switch, in that it sets the program's starting address.
7348 But it also sets the bottom bit of the address, so that it can be
7349 branched to using a BX instruction, and the program will start
7350 executing in Thumb mode straight away.
7352 @cindex PE import table prefixing
7353 @kindex --use-nul-prefixed-import-tables
7354 The @samp{--use-nul-prefixed-import-tables} switch is specifying, that
7355 the import tables idata4 and idata5 have to be generated with a zero
7356 element prefix for import libraries. This is the old style to generate
7357 import tables. By default this option is turned off.
7361 The @samp{--be8} switch instructs @command{ld} to generate BE8 format
7362 executables. This option is only valid when linking big-endian
7363 objects - ie ones which have been assembled with the @option{-EB}
7364 option. The resulting image will contain big-endian data and
7368 @kindex --target1-rel
7369 @kindex --target1-abs
7370 The @samp{R_ARM_TARGET1} relocation is typically used for entries in the
7371 @samp{.init_array} section. It is interpreted as either @samp{R_ARM_REL32}
7372 or @samp{R_ARM_ABS32}, depending on the target. The @samp{--target1-rel}
7373 and @samp{--target1-abs} switches override the default.
7376 @kindex --target2=@var{type}
7377 The @samp{--target2=type} switch overrides the default definition of the
7378 @samp{R_ARM_TARGET2} relocation. Valid values for @samp{type}, their
7379 meanings, and target defaults are as follows:
7382 @samp{R_ARM_REL32} (arm*-*-elf, arm*-*-eabi)
7386 @samp{R_ARM_GOT_PREL} (arm*-*-linux, arm*-*-*bsd)
7391 The @samp{R_ARM_V4BX} relocation (defined by the ARM AAELF
7392 specification) enables objects compiled for the ARMv4 architecture to be
7393 interworking-safe when linked with other objects compiled for ARMv4t, but
7394 also allows pure ARMv4 binaries to be built from the same ARMv4 objects.
7396 In the latter case, the switch @option{--fix-v4bx} must be passed to the
7397 linker, which causes v4t @code{BX rM} instructions to be rewritten as
7398 @code{MOV PC,rM}, since v4 processors do not have a @code{BX} instruction.
7400 In the former case, the switch should not be used, and @samp{R_ARM_V4BX}
7401 relocations are ignored.
7403 @cindex FIX_V4BX_INTERWORKING
7404 @kindex --fix-v4bx-interworking
7405 Replace @code{BX rM} instructions identified by @samp{R_ARM_V4BX}
7406 relocations with a branch to the following veneer:
7414 This allows generation of libraries/applications that work on ARMv4 cores
7415 and are still interworking safe. Note that the above veneer clobbers the
7416 condition flags, so may cause incorrect program behavior in rare cases.
7420 The @samp{--use-blx} switch enables the linker to use ARM/Thumb
7421 BLX instructions (available on ARMv5t and above) in various
7422 situations. Currently it is used to perform calls via the PLT from Thumb
7423 code using BLX rather than using BX and a mode-switching stub before
7424 each PLT entry. This should lead to such calls executing slightly faster.
7426 @cindex VFP11_DENORM_FIX
7427 @kindex --vfp11-denorm-fix
7428 The @samp{--vfp11-denorm-fix} switch enables a link-time workaround for a
7429 bug in certain VFP11 coprocessor hardware, which sometimes allows
7430 instructions with denorm operands (which must be handled by support code)
7431 to have those operands overwritten by subsequent instructions before
7432 the support code can read the intended values.
7434 The bug may be avoided in scalar mode if you allow at least one
7435 intervening instruction between a VFP11 instruction which uses a register
7436 and another instruction which writes to the same register, or at least two
7437 intervening instructions if vector mode is in use. The bug only affects
7438 full-compliance floating-point mode: you do not need this workaround if
7439 you are using "runfast" mode. Please contact ARM for further details.
7441 If you know you are using buggy VFP11 hardware, you can
7442 enable this workaround by specifying the linker option
7443 @samp{--vfp-denorm-fix=scalar} if you are using the VFP11 scalar
7444 mode only, or @samp{--vfp-denorm-fix=vector} if you are using
7445 vector mode (the latter also works for scalar code). The default is
7446 @samp{--vfp-denorm-fix=none}.
7448 If the workaround is enabled, instructions are scanned for
7449 potentially-troublesome sequences, and a veneer is created for each
7450 such sequence which may trigger the erratum. The veneer consists of the
7451 first instruction of the sequence and a branch back to the subsequent
7452 instruction. The original instruction is then replaced with a branch to
7453 the veneer. The extra cycles required to call and return from the veneer
7454 are sufficient to avoid the erratum in both the scalar and vector cases.
7456 @cindex ARM1176 erratum workaround
7457 @kindex --fix-arm1176
7458 @kindex --no-fix-arm1176
7459 The @samp{--fix-arm1176} switch enables a link-time workaround for an erratum
7460 in certain ARM1176 processors. The workaround is enabled by default if you
7461 are targeting ARM v6 (excluding ARM v6T2) or earlier. It can be disabled
7462 unconditionally by specifying @samp{--no-fix-arm1176}.
7464 Further information is available in the ``ARM1176JZ-S and ARM1176JZF-S
7465 Programmer Advice Notice'' available on the ARM documentation website at:
7466 http://infocenter.arm.com/.
7468 @cindex STM32L4xx erratum workaround
7469 @kindex --fix-stm32l4xx-629360
7471 The @samp{--fix-stm32l4xx-629360} switch enables a link-time
7472 workaround for a bug in the bus matrix / memory controller for some of
7473 the STM32 Cortex-M4 based products (STM32L4xx). When accessing
7474 off-chip memory via the affected bus for bus reads of 9 words or more,
7475 the bus can generate corrupt data and/or abort. These are only
7476 core-initiated accesses (not DMA), and might affect any access:
7477 integer loads such as LDM, POP and floating-point loads such as VLDM,
7478 VPOP. Stores are not affected.
7480 The bug can be avoided by splitting memory accesses into the
7481 necessary chunks to keep bus reads below 8 words.
7483 The workaround is not enabled by default, this is equivalent to use
7484 @samp{--fix-stm32l4xx-629360=none}. If you know you are using buggy
7485 STM32L4xx hardware, you can enable the workaround by specifying the
7486 linker option @samp{--fix-stm32l4xx-629360}, or the equivalent
7487 @samp{--fix-stm32l4xx-629360=default}.
7489 If the workaround is enabled, instructions are scanned for
7490 potentially-troublesome sequences, and a veneer is created for each
7491 such sequence which may trigger the erratum. The veneer consists in a
7492 replacement sequence emulating the behaviour of the original one and a
7493 branch back to the subsequent instruction. The original instruction is
7494 then replaced with a branch to the veneer.
7496 The workaround does not always preserve the memory access order for
7497 the LDMDB instruction, when the instruction loads the PC.
7499 The workaround is not able to handle problematic instructions when
7500 they are in the middle of an IT block, since a branch is not allowed
7501 there. In that case, the linker reports a warning and no replacement
7504 The workaround is not able to replace problematic instructions with a
7505 PC-relative branch instruction if the @samp{.text} section is too
7506 large. In that case, when the branch that replaces the original code
7507 cannot be encoded, the linker reports a warning and no replacement
7510 @cindex NO_ENUM_SIZE_WARNING
7511 @kindex --no-enum-size-warning
7512 The @option{--no-enum-size-warning} switch prevents the linker from
7513 warning when linking object files that specify incompatible EABI
7514 enumeration size attributes. For example, with this switch enabled,
7515 linking of an object file using 32-bit enumeration values with another
7516 using enumeration values fitted into the smallest possible space will
7519 @cindex NO_WCHAR_SIZE_WARNING
7520 @kindex --no-wchar-size-warning
7521 The @option{--no-wchar-size-warning} switch prevents the linker from
7522 warning when linking object files that specify incompatible EABI
7523 @code{wchar_t} size attributes. For example, with this switch enabled,
7524 linking of an object file using 32-bit @code{wchar_t} values with another
7525 using 16-bit @code{wchar_t} values will not be diagnosed.
7528 @kindex --pic-veneer
7529 The @samp{--pic-veneer} switch makes the linker use PIC sequences for
7530 ARM/Thumb interworking veneers, even if the rest of the binary
7531 is not PIC. This avoids problems on uClinux targets where
7532 @samp{--emit-relocs} is used to generate relocatable binaries.
7534 @cindex STUB_GROUP_SIZE
7535 @kindex --stub-group-size=@var{N}
7536 The linker will automatically generate and insert small sequences of
7537 code into a linked ARM ELF executable whenever an attempt is made to
7538 perform a function call to a symbol that is too far away. The
7539 placement of these sequences of instructions - called stubs - is
7540 controlled by the command-line option @option{--stub-group-size=N}.
7541 The placement is important because a poor choice can create a need for
7542 duplicate stubs, increasing the code size. The linker will try to
7543 group stubs together in order to reduce interruptions to the flow of
7544 code, but it needs guidance as to how big these groups should be and
7545 where they should be placed.
7547 The value of @samp{N}, the parameter to the
7548 @option{--stub-group-size=} option controls where the stub groups are
7549 placed. If it is negative then all stubs are placed after the first
7550 branch that needs them. If it is positive then the stubs can be
7551 placed either before or after the branches that need them. If the
7552 value of @samp{N} is 1 (either +1 or -1) then the linker will choose
7553 exactly where to place groups of stubs, using its built in heuristics.
7554 A value of @samp{N} greater than 1 (or smaller than -1) tells the
7555 linker that a single group of stubs can service at most @samp{N} bytes
7556 from the input sections.
7558 The default, if @option{--stub-group-size=} is not specified, is
7561 Farcalls stubs insertion is fully supported for the ARM-EABI target
7562 only, because it relies on object files properties not present
7565 @cindex Cortex-A8 erratum workaround
7566 @kindex --fix-cortex-a8
7567 @kindex --no-fix-cortex-a8
7568 The @samp{--fix-cortex-a8} switch enables a link-time workaround for an erratum in certain Cortex-A8 processors. The workaround is enabled by default if you are targeting the ARM v7-A architecture profile. It can be enabled otherwise by specifying @samp{--fix-cortex-a8}, or disabled unconditionally by specifying @samp{--no-fix-cortex-a8}.
7570 The erratum only affects Thumb-2 code. Please contact ARM for further details.
7572 @cindex Cortex-A53 erratum 835769 workaround
7573 @kindex --fix-cortex-a53-835769
7574 @kindex --no-fix-cortex-a53-835769
7575 The @samp{--fix-cortex-a53-835769} switch enables a link-time workaround for erratum 835769 present on certain early revisions of Cortex-A53 processors. The workaround is disabled by default. It can be enabled by specifying @samp{--fix-cortex-a53-835769}, or disabled unconditionally by specifying @samp{--no-fix-cortex-a53-835769}.
7577 Please contact ARM for further details.
7579 @kindex --merge-exidx-entries
7580 @kindex --no-merge-exidx-entries
7581 @cindex Merging exidx entries
7582 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent exidx entries in debuginfo.
7585 @cindex 32-bit PLT entries
7586 The @samp{--long-plt} option enables the use of 16 byte PLT entries
7587 which support up to 4Gb of code. The default is to use 12 byte PLT
7588 entries which only support 512Mb of code.
7590 @kindex --no-apply-dynamic-relocs
7591 @cindex AArch64 rela addend
7592 The @samp{--no-apply-dynamic-relocs} option makes AArch64 linker do not apply
7593 link-time values for dynamic relocations.
7595 @cindex Placement of SG veneers
7596 All SG veneers are placed in the special output section @code{.gnu.sgstubs}.
7597 Its start address must be set, either with the command-line option
7598 @samp{--section-start} or in a linker script, to indicate where to place these
7601 @kindex --cmse-implib
7602 @cindex Secure gateway import library
7603 The @samp{--cmse-implib} option requests that the import libraries
7604 specified by the @samp{--out-implib} and @samp{--in-implib} options are
7605 secure gateway import libraries, suitable for linking a non-secure
7606 executable against secure code as per ARMv8-M Security Extensions.
7608 @kindex --in-implib=@var{file}
7609 @cindex Input import library
7610 The @samp{--in-implib=file} specifies an input import library whose symbols
7611 must keep the same address in the executable being produced. A warning is
7612 given if no @samp{--out-implib} is given but new symbols have been introduced
7613 in the executable that should be listed in its import library. Otherwise, if
7614 @samp{--out-implib} is specified, the symbols are added to the output import
7615 library. A warning is also given if some symbols present in the input import
7616 library have disappeared from the executable. This option is only effective
7617 for Secure Gateway import libraries, ie. when @samp{--cmse-implib} is
7631 @section @command{ld} and HPPA 32-bit ELF Support
7632 @cindex HPPA multiple sub-space stubs
7633 @kindex --multi-subspace
7634 When generating a shared library, @command{ld} will by default generate
7635 import stubs suitable for use with a single sub-space application.
7636 The @samp{--multi-subspace} switch causes @command{ld} to generate export
7637 stubs, and different (larger) import stubs suitable for use with
7638 multiple sub-spaces.
7640 @cindex HPPA stub grouping
7641 @kindex --stub-group-size=@var{N}
7642 Long branch stubs and import/export stubs are placed by @command{ld} in
7643 stub sections located between groups of input sections.
7644 @samp{--stub-group-size} specifies the maximum size of a group of input
7645 sections handled by one stub section. Since branch offsets are signed,
7646 a stub section may serve two groups of input sections, one group before
7647 the stub section, and one group after it. However, when using
7648 conditional branches that require stubs, it may be better (for branch
7649 prediction) that stub sections only serve one group of input sections.
7650 A negative value for @samp{N} chooses this scheme, ensuring that
7651 branches to stubs always use a negative offset. Two special values of
7652 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
7653 @command{ld} to automatically size input section groups for the branch types
7654 detected, with the same behaviour regarding stub placement as other
7655 positive or negative values of @samp{N} respectively.
7657 Note that @samp{--stub-group-size} does not split input sections. A
7658 single input section larger than the group size specified will of course
7659 create a larger group (of one section). If input sections are too
7660 large, it may not be possible for a branch to reach its stub.
7673 @section @command{ld} and the Motorola 68K family
7675 @cindex Motorola 68K GOT generation
7676 @kindex --got=@var{type}
7677 The @samp{--got=@var{type}} option lets you choose the GOT generation scheme.
7678 The choices are @samp{single}, @samp{negative}, @samp{multigot} and
7679 @samp{target}. When @samp{target} is selected the linker chooses
7680 the default GOT generation scheme for the current target.
7681 @samp{single} tells the linker to generate a single GOT with
7682 entries only at non-negative offsets.
7683 @samp{negative} instructs the linker to generate a single GOT with
7684 entries at both negative and positive offsets. Not all environments
7686 @samp{multigot} allows the linker to generate several GOTs in the
7687 output file. All GOT references from a single input object
7688 file access the same GOT, but references from different input object
7689 files might access different GOTs. Not all environments support such GOTs.
7702 @section @command{ld} and the MIPS family
7704 @cindex MIPS microMIPS instruction choice selection
7707 The @samp{--insn32} and @samp{--no-insn32} options control the choice of
7708 microMIPS instructions used in code generated by the linker, such as that
7709 in the PLT or lazy binding stubs, or in relaxation. If @samp{--insn32} is
7710 used, then the linker only uses 32-bit instruction encodings. By default
7711 or if @samp{--no-insn32} is used, all instruction encodings are used,
7712 including 16-bit ones where possible.
7714 @cindex MIPS branch relocation check control
7715 @kindex --ignore-branch-isa
7716 @kindex --no-ignore-branch-isa
7717 The @samp{--ignore-branch-isa} and @samp{--no-ignore-branch-isa} options
7718 control branch relocation checks for invalid ISA mode transitions. If
7719 @samp{--ignore-branch-isa} is used, then the linker accepts any branch
7720 relocations and any ISA mode transition required is lost in relocation
7721 calculation, except for some cases of @code{BAL} instructions which meet
7722 relaxation conditions and are converted to equivalent @code{JALX}
7723 instructions as the associated relocation is calculated. By default
7724 or if @samp{--no-ignore-branch-isa} is used a check is made causing
7725 the loss of an ISA mode transition to produce an error.
7738 @section @code{ld} and MMIX
7739 For MMIX, there is a choice of generating @code{ELF} object files or
7740 @code{mmo} object files when linking. The simulator @code{mmix}
7741 understands the @code{mmo} format. The binutils @code{objcopy} utility
7742 can translate between the two formats.
7744 There is one special section, the @samp{.MMIX.reg_contents} section.
7745 Contents in this section is assumed to correspond to that of global
7746 registers, and symbols referring to it are translated to special symbols,
7747 equal to registers. In a final link, the start address of the
7748 @samp{.MMIX.reg_contents} section corresponds to the first allocated
7749 global register multiplied by 8. Register @code{$255} is not included in
7750 this section; it is always set to the program entry, which is at the
7751 symbol @code{Main} for @code{mmo} files.
7753 Global symbols with the prefix @code{__.MMIX.start.}, for example
7754 @code{__.MMIX.start..text} and @code{__.MMIX.start..data} are special.
7755 The default linker script uses these to set the default start address
7758 Initial and trailing multiples of zero-valued 32-bit words in a section,
7759 are left out from an mmo file.
7772 @section @code{ld} and MSP430
7773 For the MSP430 it is possible to select the MPU architecture. The flag @samp{-m [mpu type]}
7774 will select an appropriate linker script for selected MPU type. (To get a list of known MPUs
7775 just pass @samp{-m help} option to the linker).
7777 @cindex MSP430 extra sections
7778 The linker will recognize some extra sections which are MSP430 specific:
7781 @item @samp{.vectors}
7782 Defines a portion of ROM where interrupt vectors located.
7784 @item @samp{.bootloader}
7785 Defines the bootloader portion of the ROM (if applicable). Any code
7786 in this section will be uploaded to the MPU.
7788 @item @samp{.infomem}
7789 Defines an information memory section (if applicable). Any code in
7790 this section will be uploaded to the MPU.
7792 @item @samp{.infomemnobits}
7793 This is the same as the @samp{.infomem} section except that any code
7794 in this section will not be uploaded to the MPU.
7796 @item @samp{.noinit}
7797 Denotes a portion of RAM located above @samp{.bss} section.
7799 The last two sections are used by gcc.
7803 @cindex MSP430 Options
7804 @kindex --code-region
7805 @item --code-region=[either,lower,upper,none]
7806 This will transform .text* sections to [either,lower,upper].text* sections. The
7807 argument passed to GCC for -mcode-region is propagated to the linker
7810 @kindex --data-region
7811 @item --data-region=[either,lower,upper,none]
7812 This will transform .data*, .bss* and .rodata* sections to
7813 [either,lower,upper].[data,bss,rodata]* sections. The argument passed to GCC
7814 for -mdata-region is propagated to the linker using this option.
7816 @kindex --disable-sec-transformation
7817 @item --disable-sec-transformation
7818 Prevent the transformation of sections as specified by the @code{--code-region}
7819 and @code{--data-region} options.
7820 This is useful if you are compiling and linking using a single call to the GCC
7821 wrapper, and want to compile the source files using -m[code,data]-region but
7822 not transform the sections for prebuilt libraries and objects.
7836 @section @code{ld} and NDS32
7837 @kindex relaxing on NDS32
7838 For NDS32, there are some options to select relaxation behavior. The linker
7839 relaxes objects according to these options.
7842 @item @samp{--m[no-]fp-as-gp}
7843 Disable/enable fp-as-gp relaxation.
7845 @item @samp{--mexport-symbols=FILE}
7846 Exporting symbols and their address into FILE as linker script.
7848 @item @samp{--m[no-]ex9}
7849 Disable/enable link-time EX9 relaxation.
7851 @item @samp{--mexport-ex9=FILE}
7852 Export the EX9 table after linking.
7854 @item @samp{--mimport-ex9=FILE}
7855 Import the Ex9 table for EX9 relaxation.
7857 @item @samp{--mupdate-ex9}
7858 Update the existing EX9 table.
7860 @item @samp{--mex9-limit=NUM}
7861 Maximum number of entries in the ex9 table.
7863 @item @samp{--mex9-loop-aware}
7864 Avoid generating the EX9 instruction inside the loop.
7866 @item @samp{--m[no-]ifc}
7867 Disable/enable the link-time IFC optimization.
7869 @item @samp{--mifc-loop-aware}
7870 Avoid generating the IFC instruction inside the loop.
7884 @section @command{ld} and the Altera Nios II
7885 @cindex Nios II call relaxation
7886 @kindex --relax on Nios II
7888 Call and immediate jump instructions on Nios II processors are limited to
7889 transferring control to addresses in the same 256MB memory segment,
7890 which may result in @command{ld} giving
7891 @samp{relocation truncated to fit} errors with very large programs.
7892 The command-line option @option{--relax} enables the generation of
7893 trampolines that can access the entire 32-bit address space for calls
7894 outside the normal @code{call} and @code{jmpi} address range. These
7895 trampolines are inserted at section boundaries, so may not themselves
7896 be reachable if an input section and its associated call trampolines are
7899 The @option{--relax} option is enabled by default unless @option{-r}
7900 is also specified. You can disable trampoline generation by using the
7901 @option{--no-relax} linker option. You can also disable this optimization
7902 locally by using the @samp{set .noat} directive in assembly-language
7903 source files, as the linker-inserted trampolines use the @code{at}
7904 register as a temporary.
7906 Note that the linker @option{--relax} option is independent of assembler
7907 relaxation options, and that using the GNU assembler's @option{-relax-all}
7908 option interferes with the linker's more selective call instruction relaxation.
7921 @section @command{ld} and PowerPC 32-bit ELF Support
7922 @cindex PowerPC long branches
7923 @kindex --relax on PowerPC
7924 Branches on PowerPC processors are limited to a signed 26-bit
7925 displacement, which may result in @command{ld} giving
7926 @samp{relocation truncated to fit} errors with very large programs.
7927 @samp{--relax} enables the generation of trampolines that can access
7928 the entire 32-bit address space. These trampolines are inserted at
7929 section boundaries, so may not themselves be reachable if an input
7930 section exceeds 33M in size. You may combine @samp{-r} and
7931 @samp{--relax} to add trampolines in a partial link. In that case
7932 both branches to undefined symbols and inter-section branches are also
7933 considered potentially out of range, and trampolines inserted.
7935 @cindex PowerPC ELF32 options
7940 Current PowerPC GCC accepts a @samp{-msecure-plt} option that
7941 generates code capable of using a newer PLT and GOT layout that has
7942 the security advantage of no executable section ever needing to be
7943 writable and no writable section ever being executable. PowerPC
7944 @command{ld} will generate this layout, including stubs to access the
7945 PLT, if all input files (including startup and static libraries) were
7946 compiled with @samp{-msecure-plt}. @samp{--bss-plt} forces the old
7947 BSS PLT (and GOT layout) which can give slightly better performance.
7949 @kindex --secure-plt
7951 @command{ld} will use the new PLT and GOT layout if it is linking new
7952 @samp{-fpic} or @samp{-fPIC} code, but does not do so automatically
7953 when linking non-PIC code. This option requests the new PLT and GOT
7954 layout. A warning will be given if some object file requires the old
7960 The new secure PLT and GOT are placed differently relative to other
7961 sections compared to older BSS PLT and GOT placement. The location of
7962 @code{.plt} must change because the new secure PLT is an initialized
7963 section while the old PLT is uninitialized. The reason for the
7964 @code{.got} change is more subtle: The new placement allows
7965 @code{.got} to be read-only in applications linked with
7966 @samp{-z relro -z now}. However, this placement means that
7967 @code{.sdata} cannot always be used in shared libraries, because the
7968 PowerPC ABI accesses @code{.sdata} in shared libraries from the GOT
7969 pointer. @samp{--sdata-got} forces the old GOT placement. PowerPC
7970 GCC doesn't use @code{.sdata} in shared libraries, so this option is
7971 really only useful for other compilers that may do so.
7973 @cindex PowerPC stub symbols
7974 @kindex --emit-stub-syms
7975 @item --emit-stub-syms
7976 This option causes @command{ld} to label linker stubs with a local
7977 symbol that encodes the stub type and destination.
7979 @cindex PowerPC TLS optimization
7980 @kindex --no-tls-optimize
7981 @item --no-tls-optimize
7982 PowerPC @command{ld} normally performs some optimization of code
7983 sequences used to access Thread-Local Storage. Use this option to
7984 disable the optimization.
7997 @node PowerPC64 ELF64
7998 @section @command{ld} and PowerPC64 64-bit ELF Support
8000 @cindex PowerPC64 ELF64 options
8002 @cindex PowerPC64 stub grouping
8003 @kindex --stub-group-size
8004 @item --stub-group-size
8005 Long branch stubs, PLT call stubs and TOC adjusting stubs are placed
8006 by @command{ld} in stub sections located between groups of input sections.
8007 @samp{--stub-group-size} specifies the maximum size of a group of input
8008 sections handled by one stub section. Since branch offsets are signed,
8009 a stub section may serve two groups of input sections, one group before
8010 the stub section, and one group after it. However, when using
8011 conditional branches that require stubs, it may be better (for branch
8012 prediction) that stub sections only serve one group of input sections.
8013 A negative value for @samp{N} chooses this scheme, ensuring that
8014 branches to stubs always use a negative offset. Two special values of
8015 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
8016 @command{ld} to automatically size input section groups for the branch types
8017 detected, with the same behaviour regarding stub placement as other
8018 positive or negative values of @samp{N} respectively.
8020 Note that @samp{--stub-group-size} does not split input sections. A
8021 single input section larger than the group size specified will of course
8022 create a larger group (of one section). If input sections are too
8023 large, it may not be possible for a branch to reach its stub.
8025 @cindex PowerPC64 stub symbols
8026 @kindex --emit-stub-syms
8027 @item --emit-stub-syms
8028 This option causes @command{ld} to label linker stubs with a local
8029 symbol that encodes the stub type and destination.
8031 @cindex PowerPC64 dot symbols
8033 @kindex --no-dotsyms
8036 These two options control how @command{ld} interprets version patterns
8037 in a version script. Older PowerPC64 compilers emitted both a
8038 function descriptor symbol with the same name as the function, and a
8039 code entry symbol with the name prefixed by a dot (@samp{.}). To
8040 properly version a function @samp{foo}, the version script thus needs
8041 to control both @samp{foo} and @samp{.foo}. The option
8042 @samp{--dotsyms}, on by default, automatically adds the required
8043 dot-prefixed patterns. Use @samp{--no-dotsyms} to disable this
8046 @cindex PowerPC64 register save/restore functions
8047 @kindex --save-restore-funcs
8048 @kindex --no-save-restore-funcs
8049 @item --save-restore-funcs
8050 @itemx --no-save-restore-funcs
8051 These two options control whether PowerPC64 @command{ld} automatically
8052 provides out-of-line register save and restore functions used by
8053 @samp{-Os} code. The default is to provide any such referenced
8054 function for a normal final link, and to not do so for a relocatable
8057 @cindex PowerPC64 TLS optimization
8058 @kindex --no-tls-optimize
8059 @item --no-tls-optimize
8060 PowerPC64 @command{ld} normally performs some optimization of code
8061 sequences used to access Thread-Local Storage. Use this option to
8062 disable the optimization.
8064 @cindex PowerPC64 __tls_get_addr optimization
8065 @kindex --tls-get-addr-optimize
8066 @kindex --no-tls-get-addr-optimize
8067 @kindex --tls-get-addr-regsave
8068 @kindex --no-tls-get-addr-regsave
8069 @item --tls-get-addr-optimize
8070 @itemx --no-tls-get-addr-optimize
8071 These options control how PowerPC64 @command{ld} uses a special
8072 stub to call __tls_get_addr. PowerPC64 glibc 2.22 and later support
8073 an optimization that allows the second and subsequent calls to
8074 @code{__tls_get_addr} for a given symbol to be resolved by the special
8075 stub without calling in to glibc. By default the linker enables
8076 generation of the stub when glibc advertises the availability of
8078 Using @option{--tls-get-addr-optimize} with an older glibc won't do
8079 much besides slow down your applications, but may be useful if linking
8080 an application against an older glibc with the expectation that it
8081 will normally be used on systems having a newer glibc.
8082 @option{--tls-get-addr-regsave} forces generation of a stub that saves
8083 and restores volatile registers around the call into glibc. Normally,
8084 this is done when the linker detects a call to __tls_get_addr_desc.
8085 Such calls then go via the register saving stub to __tls_get_addr_opt.
8086 @option{--no-tls-get-addr-regsave} disables generation of the
8089 @cindex PowerPC64 OPD optimization
8090 @kindex --no-opd-optimize
8091 @item --no-opd-optimize
8092 PowerPC64 @command{ld} normally removes @code{.opd} section entries
8093 corresponding to deleted link-once functions, or functions removed by
8094 the action of @samp{--gc-sections} or linker script @code{/DISCARD/}.
8095 Use this option to disable @code{.opd} optimization.
8097 @cindex PowerPC64 OPD spacing
8098 @kindex --non-overlapping-opd
8099 @item --non-overlapping-opd
8100 Some PowerPC64 compilers have an option to generate compressed
8101 @code{.opd} entries spaced 16 bytes apart, overlapping the third word,
8102 the static chain pointer (unused in C) with the first word of the next
8103 entry. This option expands such entries to the full 24 bytes.
8105 @cindex PowerPC64 TOC optimization
8106 @kindex --no-toc-optimize
8107 @item --no-toc-optimize
8108 PowerPC64 @command{ld} normally removes unused @code{.toc} section
8109 entries. Such entries are detected by examining relocations that
8110 reference the TOC in code sections. A reloc in a deleted code section
8111 marks a TOC word as unneeded, while a reloc in a kept code section
8112 marks a TOC word as needed. Since the TOC may reference itself, TOC
8113 relocs are also examined. TOC words marked as both needed and
8114 unneeded will of course be kept. TOC words without any referencing
8115 reloc are assumed to be part of a multi-word entry, and are kept or
8116 discarded as per the nearest marked preceding word. This works
8117 reliably for compiler generated code, but may be incorrect if assembly
8118 code is used to insert TOC entries. Use this option to disable the
8121 @cindex PowerPC64 inline PLT call optimization
8122 @kindex --no-inline-optimize
8123 @item --no-inline-optimize
8124 PowerPC64 @command{ld} normally replaces inline PLT call sequences
8125 marked with @code{R_PPC64_PLTSEQ}, @code{R_PPC64_PLTCALL},
8126 @code{R_PPC64_PLT16_HA} and @code{R_PPC64_PLT16_LO_DS} relocations by
8127 a number of @code{nop}s and a direct call when the function is defined
8128 locally and can't be overridden by some other definition. This option
8129 disables that optimization.
8131 @cindex PowerPC64 multi-TOC
8132 @kindex --no-multi-toc
8133 @item --no-multi-toc
8134 If given any toc option besides @code{-mcmodel=medium} or
8135 @code{-mcmodel=large}, PowerPC64 GCC generates code for a TOC model
8137 entries are accessed with a 16-bit offset from r2. This limits the
8138 total TOC size to 64K. PowerPC64 @command{ld} extends this limit by
8139 grouping code sections such that each group uses less than 64K for its
8140 TOC entries, then inserts r2 adjusting stubs between inter-group
8141 calls. @command{ld} does not split apart input sections, so cannot
8142 help if a single input file has a @code{.toc} section that exceeds
8143 64K, most likely from linking multiple files with @command{ld -r}.
8144 Use this option to turn off this feature.
8146 @cindex PowerPC64 TOC sorting
8147 @kindex --no-toc-sort
8149 By default, @command{ld} sorts TOC sections so that those whose file
8150 happens to have a section called @code{.init} or @code{.fini} are
8151 placed first, followed by TOC sections referenced by code generated
8152 with PowerPC64 gcc's @code{-mcmodel=small}, and lastly TOC sections
8153 referenced only by code generated with PowerPC64 gcc's
8154 @code{-mcmodel=medium} or @code{-mcmodel=large} options. Doing this
8155 results in better TOC grouping for multi-TOC. Use this option to turn
8158 @cindex PowerPC64 PLT stub alignment
8160 @kindex --no-plt-align
8162 @itemx --no-plt-align
8163 Use these options to control whether individual PLT call stubs are
8164 aligned to a 32-byte boundary, or to the specified power of two
8165 boundary when using @code{--plt-align=}. A negative value may be
8166 specified to pad PLT call stubs so that they do not cross the
8167 specified power of two boundary (or the minimum number of boundaries
8168 if a PLT stub is so large that it must cross a boundary). By default
8169 PLT call stubs are aligned to 32-byte boundaries.
8171 @cindex PowerPC64 PLT call stub static chain
8172 @kindex --plt-static-chain
8173 @kindex --no-plt-static-chain
8174 @item --plt-static-chain
8175 @itemx --no-plt-static-chain
8176 Use these options to control whether PLT call stubs load the static
8177 chain pointer (r11). @code{ld} defaults to not loading the static
8178 chain since there is never any need to do so on a PLT call.
8180 @cindex PowerPC64 PLT call stub thread safety
8181 @kindex --plt-thread-safe
8182 @kindex --no-plt-thread-safe
8183 @item --plt-thread-safe
8184 @itemx --no-plt-thread-safe
8185 With power7's weakly ordered memory model, it is possible when using
8186 lazy binding for ld.so to update a plt entry in one thread and have
8187 another thread see the individual plt entry words update in the wrong
8188 order, despite ld.so carefully writing in the correct order and using
8189 memory write barriers. To avoid this we need some sort of read
8190 barrier in the call stub, or use LD_BIND_NOW=1. By default, @code{ld}
8191 looks for calls to commonly used functions that create threads, and if
8192 seen, adds the necessary barriers. Use these options to change the
8195 @cindex PowerPC64 ELFv2 PLT localentry optimization
8196 @kindex --plt-localentry
8197 @kindex --no-plt-localentry
8198 @item --plt-localentry
8199 @itemx --no-localentry
8200 ELFv2 functions with localentry:0 are those with a single entry point,
8201 ie. global entry == local entry, and that have no requirement on r2
8202 (the TOC/GOT pointer) or r12, and guarantee r2 is unchanged on return.
8203 Such an external function can be called via the PLT without saving r2
8204 or restoring it on return, avoiding a common load-hit-store for small
8205 functions. The optimization is attractive, with up to 40% reduction
8206 in execution time for a small function, but can result in symbol
8207 interposition failures. Also, minor changes in a shared library,
8208 including system libraries, can cause a function that was localentry:0
8209 to become localentry:8. This will result in a dynamic loader
8210 complaint and failure to run. The option is experimental, use with
8211 care. @option{--no-plt-localentry} is the default.
8213 @cindex PowerPC64 Power10 stubs
8214 @kindex --power10-stubs
8215 @kindex --no-power10-stubs
8216 @item --power10-stubs
8217 @itemx --no-power10-stubs
8218 When PowerPC64 @command{ld} links input object files containing
8219 relocations used on power10 prefixed instructions it normally creates
8220 linkage stubs (PLT call and long branch) using power10 instructions
8221 for @code{@@notoc} PLT calls where @code{r2} is not known. The
8222 power10 notoc stubs are smaller and faster, so are preferred for
8223 power10. @option{--power10-stubs} and @option{--no-power10-stubs}
8224 allow you to override the linker's selection of stub instructions.
8225 @option{--power10-stubs=auto} allows the user to select the default
8240 @section @command{ld} and S/390 ELF Support
8242 @cindex S/390 ELF options
8246 @kindex --s390-pgste
8248 This option marks the result file with a @code{PT_S390_PGSTE}
8249 segment. The Linux kernel is supposed to allocate 4k page tables for
8250 binaries marked that way.
8264 @section @command{ld} and SPU ELF Support
8266 @cindex SPU ELF options
8272 This option marks an executable as a PIC plugin module.
8274 @cindex SPU overlays
8275 @kindex --no-overlays
8277 Normally, @command{ld} recognizes calls to functions within overlay
8278 regions, and redirects such calls to an overlay manager via a stub.
8279 @command{ld} also provides a built-in overlay manager. This option
8280 turns off all this special overlay handling.
8282 @cindex SPU overlay stub symbols
8283 @kindex --emit-stub-syms
8284 @item --emit-stub-syms
8285 This option causes @command{ld} to label overlay stubs with a local
8286 symbol that encodes the stub type and destination.
8288 @cindex SPU extra overlay stubs
8289 @kindex --extra-overlay-stubs
8290 @item --extra-overlay-stubs
8291 This option causes @command{ld} to add overlay call stubs on all
8292 function calls out of overlay regions. Normally stubs are not added
8293 on calls to non-overlay regions.
8295 @cindex SPU local store size
8296 @kindex --local-store=lo:hi
8297 @item --local-store=lo:hi
8298 @command{ld} usually checks that a final executable for SPU fits in
8299 the address range 0 to 256k. This option may be used to change the
8300 range. Disable the check entirely with @option{--local-store=0:0}.
8303 @kindex --stack-analysis
8304 @item --stack-analysis
8305 SPU local store space is limited. Over-allocation of stack space
8306 unnecessarily limits space available for code and data, while
8307 under-allocation results in runtime failures. If given this option,
8308 @command{ld} will provide an estimate of maximum stack usage.
8309 @command{ld} does this by examining symbols in code sections to
8310 determine the extents of functions, and looking at function prologues
8311 for stack adjusting instructions. A call-graph is created by looking
8312 for relocations on branch instructions. The graph is then searched
8313 for the maximum stack usage path. Note that this analysis does not
8314 find calls made via function pointers, and does not handle recursion
8315 and other cycles in the call graph. Stack usage may be
8316 under-estimated if your code makes such calls. Also, stack usage for
8317 dynamic allocation, e.g. alloca, will not be detected. If a link map
8318 is requested, detailed information about each function's stack usage
8319 and calls will be given.
8322 @kindex --emit-stack-syms
8323 @item --emit-stack-syms
8324 This option, if given along with @option{--stack-analysis} will result
8325 in @command{ld} emitting stack sizing symbols for each function.
8326 These take the form @code{__stack_<function_name>} for global
8327 functions, and @code{__stack_<number>_<function_name>} for static
8328 functions. @code{<number>} is the section id in hex. The value of
8329 such symbols is the stack requirement for the corresponding function.
8330 The symbol size will be zero, type @code{STT_NOTYPE}, binding
8331 @code{STB_LOCAL}, and section @code{SHN_ABS}.
8345 @section @command{ld}'s Support for Various TI COFF Versions
8346 @cindex TI COFF versions
8347 @kindex --format=@var{version}
8348 The @samp{--format} switch allows selection of one of the various
8349 TI COFF versions. The latest of this writing is 2; versions 0 and 1 are
8350 also supported. The TI COFF versions also vary in header byte-order
8351 format; @command{ld} will read any version or byte order, but the output
8352 header format depends on the default specified by the specific target.
8365 @section @command{ld} and WIN32 (cygwin/mingw)
8367 This section describes some of the win32 specific @command{ld} issues.
8368 See @ref{Options,,Command-line Options} for detailed description of the
8369 command-line options mentioned here.
8372 @cindex import libraries
8373 @item import libraries
8374 The standard Windows linker creates and uses so-called import
8375 libraries, which contains information for linking to dll's. They are
8376 regular static archives and are handled as any other static
8377 archive. The cygwin and mingw ports of @command{ld} have specific
8378 support for creating such libraries provided with the
8379 @samp{--out-implib} command-line option.
8381 @item exporting DLL symbols
8382 @cindex exporting DLL symbols
8383 The cygwin/mingw @command{ld} has several ways to export symbols for dll's.
8386 @item using auto-export functionality
8387 @cindex using auto-export functionality
8388 By default @command{ld} exports symbols with the auto-export functionality,
8389 which is controlled by the following command-line options:
8392 @item --export-all-symbols [This is the default]
8393 @item --exclude-symbols
8394 @item --exclude-libs
8395 @item --exclude-modules-for-implib
8396 @item --version-script
8399 When auto-export is in operation, @command{ld} will export all the non-local
8400 (global and common) symbols it finds in a DLL, with the exception of a few
8401 symbols known to belong to the system's runtime and libraries. As it will
8402 often not be desirable to export all of a DLL's symbols, which may include
8403 private functions that are not part of any public interface, the command-line
8404 options listed above may be used to filter symbols out from the list for
8405 exporting. The @samp{--output-def} option can be used in order to see the
8406 final list of exported symbols with all exclusions taken into effect.
8408 If @samp{--export-all-symbols} is not given explicitly on the
8409 command line, then the default auto-export behavior will be @emph{disabled}
8410 if either of the following are true:
8413 @item A DEF file is used.
8414 @item Any symbol in any object file was marked with the __declspec(dllexport) attribute.
8417 @item using a DEF file
8418 @cindex using a DEF file
8419 Another way of exporting symbols is using a DEF file. A DEF file is
8420 an ASCII file containing definitions of symbols which should be
8421 exported when a dll is created. Usually it is named @samp{<dll
8422 name>.def} and is added as any other object file to the linker's
8423 command line. The file's name must end in @samp{.def} or @samp{.DEF}.
8426 gcc -o <output> <objectfiles> <dll name>.def
8429 Using a DEF file turns off the normal auto-export behavior, unless the
8430 @samp{--export-all-symbols} option is also used.
8432 Here is an example of a DEF file for a shared library called @samp{xyz.dll}:
8435 LIBRARY "xyz.dll" BASE=0x20000000
8441 another_foo = abc.dll.afoo
8447 This example defines a DLL with a non-default base address and seven
8448 symbols in the export table. The third exported symbol @code{_bar} is an
8449 alias for the second. The fourth symbol, @code{another_foo} is resolved
8450 by "forwarding" to another module and treating it as an alias for
8451 @code{afoo} exported from the DLL @samp{abc.dll}. The final symbol
8452 @code{var1} is declared to be a data object. The @samp{doo} symbol in
8453 export library is an alias of @samp{foo}, which gets the string name
8454 in export table @samp{foo2}. The @samp{eoo} symbol is an data export
8455 symbol, which gets in export table the name @samp{var1}.
8457 The optional @code{LIBRARY <name>} command indicates the @emph{internal}
8458 name of the output DLL. If @samp{<name>} does not include a suffix,
8459 the default library suffix, @samp{.DLL} is appended.
8461 When the .DEF file is used to build an application, rather than a
8462 library, the @code{NAME <name>} command should be used instead of
8463 @code{LIBRARY}. If @samp{<name>} does not include a suffix, the default
8464 executable suffix, @samp{.EXE} is appended.
8466 With either @code{LIBRARY <name>} or @code{NAME <name>} the optional
8467 specification @code{BASE = <number>} may be used to specify a
8468 non-default base address for the image.
8470 If neither @code{LIBRARY <name>} nor @code{NAME <name>} is specified,
8471 or they specify an empty string, the internal name is the same as the
8472 filename specified on the command line.
8474 The complete specification of an export symbol is:
8478 ( ( ( <name1> [ = <name2> ] )
8479 | ( <name1> = <module-name> . <external-name>))
8480 [ @@ <integer> ] [NONAME] [DATA] [CONSTANT] [PRIVATE] [== <name3>] ) *
8483 Declares @samp{<name1>} as an exported symbol from the DLL, or declares
8484 @samp{<name1>} as an exported alias for @samp{<name2>}; or declares
8485 @samp{<name1>} as a "forward" alias for the symbol
8486 @samp{<external-name>} in the DLL @samp{<module-name>}.
8487 Optionally, the symbol may be exported by the specified ordinal
8488 @samp{<integer>} alias. The optional @samp{<name3>} is the to be used
8489 string in import/export table for the symbol.
8491 The optional keywords that follow the declaration indicate:
8493 @code{NONAME}: Do not put the symbol name in the DLL's export table. It
8494 will still be exported by its ordinal alias (either the value specified
8495 by the .def specification or, otherwise, the value assigned by the
8496 linker). The symbol name, however, does remain visible in the import
8497 library (if any), unless @code{PRIVATE} is also specified.
8499 @code{DATA}: The symbol is a variable or object, rather than a function.
8500 The import lib will export only an indirect reference to @code{foo} as
8501 the symbol @code{_imp__foo} (ie, @code{foo} must be resolved as
8504 @code{CONSTANT}: Like @code{DATA}, but put the undecorated @code{foo} as
8505 well as @code{_imp__foo} into the import library. Both refer to the
8506 read-only import address table's pointer to the variable, not to the
8507 variable itself. This can be dangerous. If the user code fails to add
8508 the @code{dllimport} attribute and also fails to explicitly add the
8509 extra indirection that the use of the attribute enforces, the
8510 application will behave unexpectedly.
8512 @code{PRIVATE}: Put the symbol in the DLL's export table, but do not put
8513 it into the static import library used to resolve imports at link time. The
8514 symbol can still be imported using the @code{LoadLibrary/GetProcAddress}
8515 API at runtime or by using the GNU ld extension of linking directly to
8516 the DLL without an import library.
8518 See ld/deffilep.y in the binutils sources for the full specification of
8519 other DEF file statements
8521 @cindex creating a DEF file
8522 While linking a shared dll, @command{ld} is able to create a DEF file
8523 with the @samp{--output-def <file>} command-line option.
8525 @item Using decorations
8526 @cindex Using decorations
8527 Another way of marking symbols for export is to modify the source code
8528 itself, so that when building the DLL each symbol to be exported is
8532 __declspec(dllexport) int a_variable
8533 __declspec(dllexport) void a_function(int with_args)
8536 All such symbols will be exported from the DLL. If, however,
8537 any of the object files in the DLL contain symbols decorated in
8538 this way, then the normal auto-export behavior is disabled, unless
8539 the @samp{--export-all-symbols} option is also used.
8541 Note that object files that wish to access these symbols must @emph{not}
8542 decorate them with dllexport. Instead, they should use dllimport,
8546 __declspec(dllimport) int a_variable
8547 __declspec(dllimport) void a_function(int with_args)
8550 This complicates the structure of library header files, because
8551 when included by the library itself the header must declare the
8552 variables and functions as dllexport, but when included by client
8553 code the header must declare them as dllimport. There are a number
8554 of idioms that are typically used to do this; often client code can
8555 omit the __declspec() declaration completely. See
8556 @samp{--enable-auto-import} and @samp{automatic data imports} for more
8560 @cindex automatic data imports
8561 @item automatic data imports
8562 The standard Windows dll format supports data imports from dlls only
8563 by adding special decorations (dllimport/dllexport), which let the
8564 compiler produce specific assembler instructions to deal with this
8565 issue. This increases the effort necessary to port existing Un*x
8566 code to these platforms, especially for large
8567 c++ libraries and applications. The auto-import feature, which was
8568 initially provided by Paul Sokolovsky, allows one to omit the
8569 decorations to achieve a behavior that conforms to that on POSIX/Un*x
8570 platforms. This feature is enabled with the @samp{--enable-auto-import}
8571 command-line option, although it is enabled by default on cygwin/mingw.
8572 The @samp{--enable-auto-import} option itself now serves mainly to
8573 suppress any warnings that are ordinarily emitted when linked objects
8574 trigger the feature's use.
8576 auto-import of variables does not always work flawlessly without
8577 additional assistance. Sometimes, you will see this message
8579 "variable '<var>' can't be auto-imported. Please read the
8580 documentation for ld's @code{--enable-auto-import} for details."
8582 The @samp{--enable-auto-import} documentation explains why this error
8583 occurs, and several methods that can be used to overcome this difficulty.
8584 One of these methods is the @emph{runtime pseudo-relocs} feature, described
8587 @cindex runtime pseudo-relocation
8588 For complex variables imported from DLLs (such as structs or classes),
8589 object files typically contain a base address for the variable and an
8590 offset (@emph{addend}) within the variable--to specify a particular
8591 field or public member, for instance. Unfortunately, the runtime loader used
8592 in win32 environments is incapable of fixing these references at runtime
8593 without the additional information supplied by dllimport/dllexport decorations.
8594 The standard auto-import feature described above is unable to resolve these
8597 The @samp{--enable-runtime-pseudo-relocs} switch allows these references to
8598 be resolved without error, while leaving the task of adjusting the references
8599 themselves (with their non-zero addends) to specialized code provided by the
8600 runtime environment. Recent versions of the cygwin and mingw environments and
8601 compilers provide this runtime support; older versions do not. However, the
8602 support is only necessary on the developer's platform; the compiled result will
8603 run without error on an older system.
8605 @samp{--enable-runtime-pseudo-relocs} is not the default; it must be explicitly
8608 @cindex direct linking to a dll
8609 @item direct linking to a dll
8610 The cygwin/mingw ports of @command{ld} support the direct linking,
8611 including data symbols, to a dll without the usage of any import
8612 libraries. This is much faster and uses much less memory than does the
8613 traditional import library method, especially when linking large
8614 libraries or applications. When @command{ld} creates an import lib, each
8615 function or variable exported from the dll is stored in its own bfd, even
8616 though a single bfd could contain many exports. The overhead involved in
8617 storing, loading, and processing so many bfd's is quite large, and explains the
8618 tremendous time, memory, and storage needed to link against particularly
8619 large or complex libraries when using import libs.
8621 Linking directly to a dll uses no extra command-line switches other than
8622 @samp{-L} and @samp{-l}, because @command{ld} already searches for a number
8623 of names to match each library. All that is needed from the developer's
8624 perspective is an understanding of this search, in order to force ld to
8625 select the dll instead of an import library.
8628 For instance, when ld is called with the argument @samp{-lxxx} it will attempt
8629 to find, in the first directory of its search path,
8642 before moving on to the next directory in the search path.
8644 (*) Actually, this is not @samp{cygxxx.dll} but in fact is @samp{<prefix>xxx.dll},
8645 where @samp{<prefix>} is set by the @command{ld} option
8646 @samp{--dll-search-prefix=<prefix>}. In the case of cygwin, the standard gcc spec
8647 file includes @samp{--dll-search-prefix=cyg}, so in effect we actually search for
8650 Other win32-based unix environments, such as mingw or pw32, may use other
8651 @samp{<prefix>}es, although at present only cygwin makes use of this feature. It
8652 was originally intended to help avoid name conflicts among dll's built for the
8653 various win32/un*x environments, so that (for example) two versions of a zlib dll
8654 could coexist on the same machine.
8656 The generic cygwin/mingw path layout uses a @samp{bin} directory for
8657 applications and dll's and a @samp{lib} directory for the import
8658 libraries (using cygwin nomenclature):
8664 libxxx.dll.a (in case of dll's)
8665 libxxx.a (in case of static archive)
8668 Linking directly to a dll without using the import library can be
8671 1. Use the dll directly by adding the @samp{bin} path to the link line
8673 gcc -Wl,-verbose -o a.exe -L../bin/ -lxxx
8676 However, as the dll's often have version numbers appended to their names
8677 (@samp{cygncurses-5.dll}) this will often fail, unless one specifies
8678 @samp{-L../bin -lncurses-5} to include the version. Import libs are generally
8679 not versioned, and do not have this difficulty.
8681 2. Create a symbolic link from the dll to a file in the @samp{lib}
8682 directory according to the above mentioned search pattern. This
8683 should be used to avoid unwanted changes in the tools needed for
8687 ln -s bin/cygxxx.dll lib/[cyg|lib|]xxx.dll[.a]
8690 Then you can link without any make environment changes.
8693 gcc -Wl,-verbose -o a.exe -L../lib/ -lxxx
8696 This technique also avoids the version number problems, because the following is
8703 libxxx.dll.a -> ../bin/cygxxx-5.dll
8706 Linking directly to a dll without using an import lib will work
8707 even when auto-import features are exercised, and even when
8708 @samp{--enable-runtime-pseudo-relocs} is used.
8710 Given the improvements in speed and memory usage, one might justifiably
8711 wonder why import libraries are used at all. There are three reasons:
8713 1. Until recently, the link-directly-to-dll functionality did @emph{not}
8714 work with auto-imported data.
8716 2. Sometimes it is necessary to include pure static objects within the
8717 import library (which otherwise contains only bfd's for indirection
8718 symbols that point to the exports of a dll). Again, the import lib
8719 for the cygwin kernel makes use of this ability, and it is not
8720 possible to do this without an import lib.
8722 3. Symbol aliases can only be resolved using an import lib. This is
8723 critical when linking against OS-supplied dll's (eg, the win32 API)
8724 in which symbols are usually exported as undecorated aliases of their
8725 stdcall-decorated assembly names.
8727 So, import libs are not going away. But the ability to replace
8728 true import libs with a simple symbolic link to (or a copy of)
8729 a dll, in many cases, is a useful addition to the suite of tools
8730 binutils makes available to the win32 developer. Given the
8731 massive improvements in memory requirements during linking, storage
8732 requirements, and linking speed, we expect that many developers
8733 will soon begin to use this feature whenever possible.
8735 @item symbol aliasing
8737 @item adding additional names
8738 Sometimes, it is useful to export symbols with additional names.
8739 A symbol @samp{foo} will be exported as @samp{foo}, but it can also be
8740 exported as @samp{_foo} by using special directives in the DEF file
8741 when creating the dll. This will affect also the optional created
8742 import library. Consider the following DEF file:
8745 LIBRARY "xyz.dll" BASE=0x61000000
8752 The line @samp{_foo = foo} maps the symbol @samp{foo} to @samp{_foo}.
8754 Another method for creating a symbol alias is to create it in the
8755 source code using the "weak" attribute:
8758 void foo () @{ /* Do something. */; @}
8759 void _foo () __attribute__ ((weak, alias ("foo")));
8762 See the gcc manual for more information about attributes and weak
8765 @item renaming symbols
8766 Sometimes it is useful to rename exports. For instance, the cygwin
8767 kernel does this regularly. A symbol @samp{_foo} can be exported as
8768 @samp{foo} but not as @samp{_foo} by using special directives in the
8769 DEF file. (This will also affect the import library, if it is
8770 created). In the following example:
8773 LIBRARY "xyz.dll" BASE=0x61000000
8779 The line @samp{_foo = foo} maps the exported symbol @samp{foo} to
8783 Note: using a DEF file disables the default auto-export behavior,
8784 unless the @samp{--export-all-symbols} command-line option is used.
8785 If, however, you are trying to rename symbols, then you should list
8786 @emph{all} desired exports in the DEF file, including the symbols
8787 that are not being renamed, and do @emph{not} use the
8788 @samp{--export-all-symbols} option. If you list only the
8789 renamed symbols in the DEF file, and use @samp{--export-all-symbols}
8790 to handle the other symbols, then the both the new names @emph{and}
8791 the original names for the renamed symbols will be exported.
8792 In effect, you'd be aliasing those symbols, not renaming them,
8793 which is probably not what you wanted.
8795 @cindex weak externals
8796 @item weak externals
8797 The Windows object format, PE, specifies a form of weak symbols called
8798 weak externals. When a weak symbol is linked and the symbol is not
8799 defined, the weak symbol becomes an alias for some other symbol. There
8800 are three variants of weak externals:
8802 @item Definition is searched for in objects and libraries, historically
8803 called lazy externals.
8804 @item Definition is searched for only in other objects, not in libraries.
8805 This form is not presently implemented.
8806 @item No search; the symbol is an alias. This form is not presently
8809 As a GNU extension, weak symbols that do not specify an alternate symbol
8810 are supported. If the symbol is undefined when linking, the symbol
8811 uses a default value.
8813 @cindex aligned common symbols
8814 @item aligned common symbols
8815 As a GNU extension to the PE file format, it is possible to specify the
8816 desired alignment for a common symbol. This information is conveyed from
8817 the assembler or compiler to the linker by means of GNU-specific commands
8818 carried in the object file's @samp{.drectve} section, which are recognized
8819 by @command{ld} and respected when laying out the common symbols. Native
8820 tools will be able to process object files employing this GNU extension,
8821 but will fail to respect the alignment instructions, and may issue noisy
8822 warnings about unknown linker directives.
8837 @section @code{ld} and Xtensa Processors
8839 @cindex Xtensa processors
8840 The default @command{ld} behavior for Xtensa processors is to interpret
8841 @code{SECTIONS} commands so that lists of explicitly named sections in a
8842 specification with a wildcard file will be interleaved when necessary to
8843 keep literal pools within the range of PC-relative load offsets. For
8844 example, with the command:
8856 @command{ld} may interleave some of the @code{.literal}
8857 and @code{.text} sections from different object files to ensure that the
8858 literal pools are within the range of PC-relative load offsets. A valid
8859 interleaving might place the @code{.literal} sections from an initial
8860 group of files followed by the @code{.text} sections of that group of
8861 files. Then, the @code{.literal} sections from the rest of the files
8862 and the @code{.text} sections from the rest of the files would follow.
8864 @cindex @option{--relax} on Xtensa
8865 @cindex relaxing on Xtensa
8866 Relaxation is enabled by default for the Xtensa version of @command{ld} and
8867 provides two important link-time optimizations. The first optimization
8868 is to combine identical literal values to reduce code size. A redundant
8869 literal will be removed and all the @code{L32R} instructions that use it
8870 will be changed to reference an identical literal, as long as the
8871 location of the replacement literal is within the offset range of all
8872 the @code{L32R} instructions. The second optimization is to remove
8873 unnecessary overhead from assembler-generated ``longcall'' sequences of
8874 @code{L32R}/@code{CALLX@var{n}} when the target functions are within
8875 range of direct @code{CALL@var{n}} instructions.
8877 For each of these cases where an indirect call sequence can be optimized
8878 to a direct call, the linker will change the @code{CALLX@var{n}}
8879 instruction to a @code{CALL@var{n}} instruction, remove the @code{L32R}
8880 instruction, and remove the literal referenced by the @code{L32R}
8881 instruction if it is not used for anything else. Removing the
8882 @code{L32R} instruction always reduces code size but can potentially
8883 hurt performance by changing the alignment of subsequent branch targets.
8884 By default, the linker will always preserve alignments, either by
8885 switching some instructions between 24-bit encodings and the equivalent
8886 density instructions or by inserting a no-op in place of the @code{L32R}
8887 instruction that was removed. If code size is more important than
8888 performance, the @option{--size-opt} option can be used to prevent the
8889 linker from widening density instructions or inserting no-ops, except in
8890 a few cases where no-ops are required for correctness.
8892 The following Xtensa-specific command-line options can be used to
8895 @cindex Xtensa options
8898 When optimizing indirect calls to direct calls, optimize for code size
8899 more than performance. With this option, the linker will not insert
8900 no-ops or widen density instructions to preserve branch target
8901 alignment. There may still be some cases where no-ops are required to
8902 preserve the correctness of the code.
8904 @item --abi-windowed
8906 Choose ABI for the output object and for the generated PLT code.
8907 PLT code inserted by the linker must match ABI of the output object
8908 because windowed and call0 ABI use incompatible function call
8910 Default ABI is chosen by the ABI tag in the @code{.xtensa.info} section
8911 of the first input object.
8912 A warning is issued if ABI tags of input objects do not match each other
8913 or the chosen output object ABI.
8921 @ifclear SingleFormat
8926 @cindex object file management
8927 @cindex object formats available
8929 The linker accesses object and archive files using the BFD libraries.
8930 These libraries allow the linker to use the same routines to operate on
8931 object files whatever the object file format. A different object file
8932 format can be supported simply by creating a new BFD back end and adding
8933 it to the library. To conserve runtime memory, however, the linker and
8934 associated tools are usually configured to support only a subset of the
8935 object file formats available. You can use @code{objdump -i}
8936 (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to
8937 list all the formats available for your configuration.
8939 @cindex BFD requirements
8940 @cindex requirements for BFD
8941 As with most implementations, BFD is a compromise between
8942 several conflicting requirements. The major factor influencing
8943 BFD design was efficiency: any time used converting between
8944 formats is time which would not have been spent had BFD not
8945 been involved. This is partly offset by abstraction payback; since
8946 BFD simplifies applications and back ends, more time and care
8947 may be spent optimizing algorithms for a greater speed.
8949 One minor artifact of the BFD solution which you should bear in
8950 mind is the potential for information loss. There are two places where
8951 useful information can be lost using the BFD mechanism: during
8952 conversion and during output. @xref{BFD information loss}.
8955 * BFD outline:: How it works: an outline of BFD
8959 @section How It Works: An Outline of BFD
8960 @cindex opening object files
8961 @include bfdsumm.texi
8964 @node Reporting Bugs
8965 @chapter Reporting Bugs
8966 @cindex bugs in @command{ld}
8967 @cindex reporting bugs in @command{ld}
8969 Your bug reports play an essential role in making @command{ld} reliable.
8971 Reporting a bug may help you by bringing a solution to your problem, or
8972 it may not. But in any case the principal function of a bug report is
8973 to help the entire community by making the next version of @command{ld}
8974 work better. Bug reports are your contribution to the maintenance of
8977 In order for a bug report to serve its purpose, you must include the
8978 information that enables us to fix the bug.
8981 * Bug Criteria:: Have you found a bug?
8982 * Bug Reporting:: How to report bugs
8986 @section Have You Found a Bug?
8987 @cindex bug criteria
8989 If you are not sure whether you have found a bug, here are some guidelines:
8992 @cindex fatal signal
8993 @cindex linker crash
8994 @cindex crash of linker
8996 If the linker gets a fatal signal, for any input whatever, that is a
8997 @command{ld} bug. Reliable linkers never crash.
8999 @cindex error on valid input
9001 If @command{ld} produces an error message for valid input, that is a bug.
9003 @cindex invalid input
9005 If @command{ld} does not produce an error message for invalid input, that
9006 may be a bug. In the general case, the linker can not verify that
9007 object files are correct.
9010 If you are an experienced user of linkers, your suggestions for
9011 improvement of @command{ld} are welcome in any case.
9015 @section How to Report Bugs
9017 @cindex @command{ld} bugs, reporting
9019 A number of companies and individuals offer support for @sc{gnu}
9020 products. If you obtained @command{ld} from a support organization, we
9021 recommend you contact that organization first.
9023 You can find contact information for many support companies and
9024 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
9028 Otherwise, send bug reports for @command{ld} to
9032 The fundamental principle of reporting bugs usefully is this:
9033 @strong{report all the facts}. If you are not sure whether to state a
9034 fact or leave it out, state it!
9036 Often people omit facts because they think they know what causes the
9037 problem and assume that some details do not matter. Thus, you might
9038 assume that the name of a symbol you use in an example does not
9039 matter. Well, probably it does not, but one cannot be sure. Perhaps
9040 the bug is a stray memory reference which happens to fetch from the
9041 location where that name is stored in memory; perhaps, if the name
9042 were different, the contents of that location would fool the linker
9043 into doing the right thing despite the bug. Play it safe and give a
9044 specific, complete example. That is the easiest thing for you to do,
9045 and the most helpful.
9047 Keep in mind that the purpose of a bug report is to enable us to fix
9048 the bug if it is new to us. Therefore, always write your bug reports
9049 on the assumption that the bug has not been reported previously.
9051 Sometimes people give a few sketchy facts and ask, ``Does this ring a
9052 bell?'' This cannot help us fix a bug, so it is basically useless. We
9053 respond by asking for enough details to enable us to investigate.
9054 You might as well expedite matters by sending them to begin with.
9056 To enable us to fix the bug, you should include all these things:
9060 The version of @command{ld}. @command{ld} announces it if you start it with
9061 the @samp{--version} argument.
9063 Without this, we will not know whether there is any point in looking for
9064 the bug in the current version of @command{ld}.
9067 Any patches you may have applied to the @command{ld} source, including any
9068 patches made to the @code{BFD} library.
9071 The type of machine you are using, and the operating system name and
9075 What compiler (and its version) was used to compile @command{ld}---e.g.
9079 The command arguments you gave the linker to link your example and
9080 observe the bug. To guarantee you will not omit something important,
9081 list them all. A copy of the Makefile (or the output from make) is
9084 If we were to try to guess the arguments, we would probably guess wrong
9085 and then we might not encounter the bug.
9088 A complete input file, or set of input files, that will reproduce the
9089 bug. It is generally most helpful to send the actual object files
9090 provided that they are reasonably small. Say no more than 10K. For
9091 bigger files you can either make them available by FTP or HTTP or else
9092 state that you are willing to send the object file(s) to whomever
9093 requests them. (Note - your email will be going to a mailing list, so
9094 we do not want to clog it up with large attachments). But small
9095 attachments are best.
9097 If the source files were assembled using @code{gas} or compiled using
9098 @code{gcc}, then it may be OK to send the source files rather than the
9099 object files. In this case, be sure to say exactly what version of
9100 @code{gas} or @code{gcc} was used to produce the object files. Also say
9101 how @code{gas} or @code{gcc} were configured.
9104 A description of what behavior you observe that you believe is
9105 incorrect. For example, ``It gets a fatal signal.''
9107 Of course, if the bug is that @command{ld} gets a fatal signal, then we
9108 will certainly notice it. But if the bug is incorrect output, we might
9109 not notice unless it is glaringly wrong. You might as well not give us
9110 a chance to make a mistake.
9112 Even if the problem you experience is a fatal signal, you should still
9113 say so explicitly. Suppose something strange is going on, such as, your
9114 copy of @command{ld} is out of sync, or you have encountered a bug in the
9115 C library on your system. (This has happened!) Your copy might crash
9116 and ours would not. If you told us to expect a crash, then when ours
9117 fails to crash, we would know that the bug was not happening for us. If
9118 you had not told us to expect a crash, then we would not be able to draw
9119 any conclusion from our observations.
9122 If you wish to suggest changes to the @command{ld} source, send us context
9123 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or
9124 @samp{-p} option. Always send diffs from the old file to the new file.
9125 If you even discuss something in the @command{ld} source, refer to it by
9126 context, not by line number.
9128 The line numbers in our development sources will not match those in your
9129 sources. Your line numbers would convey no useful information to us.
9132 Here are some things that are not necessary:
9136 A description of the envelope of the bug.
9138 Often people who encounter a bug spend a lot of time investigating
9139 which changes to the input file will make the bug go away and which
9140 changes will not affect it.
9142 This is often time consuming and not very useful, because the way we
9143 will find the bug is by running a single example under the debugger
9144 with breakpoints, not by pure deduction from a series of examples.
9145 We recommend that you save your time for something else.
9147 Of course, if you can find a simpler example to report @emph{instead}
9148 of the original one, that is a convenience for us. Errors in the
9149 output will be easier to spot, running under the debugger will take
9150 less time, and so on.
9152 However, simplification is not vital; if you do not want to do this,
9153 report the bug anyway and send us the entire test case you used.
9156 A patch for the bug.
9158 A patch for the bug does help us if it is a good one. But do not omit
9159 the necessary information, such as the test case, on the assumption that
9160 a patch is all we need. We might see problems with your patch and decide
9161 to fix the problem another way, or we might not understand it at all.
9163 Sometimes with a program as complicated as @command{ld} it is very hard to
9164 construct an example that will make the program follow a certain path
9165 through the code. If you do not send us the example, we will not be
9166 able to construct one, so we will not be able to verify that the bug is
9169 And if we cannot understand what bug you are trying to fix, or why your
9170 patch should be an improvement, we will not install it. A test case will
9171 help us to understand.
9174 A guess about what the bug is or what it depends on.
9176 Such guesses are usually wrong. Even we cannot guess right about such
9177 things without first using the debugger to find the facts.
9181 @appendix MRI Compatible Script Files
9182 @cindex MRI compatibility
9183 To aid users making the transition to @sc{gnu} @command{ld} from the MRI
9184 linker, @command{ld} can use MRI compatible linker scripts as an
9185 alternative to the more general-purpose linker scripting language
9186 described in @ref{Scripts}. MRI compatible linker scripts have a much
9187 simpler command set than the scripting language otherwise used with
9188 @command{ld}. @sc{gnu} @command{ld} supports the most commonly used MRI
9189 linker commands; these commands are described here.
9191 In general, MRI scripts aren't of much use with the @code{a.out} object
9192 file format, since it only has three sections and MRI scripts lack some
9193 features to make use of them.
9195 You can specify a file containing an MRI-compatible script using the
9196 @samp{-c} command-line option.
9198 Each command in an MRI-compatible script occupies its own line; each
9199 command line starts with the keyword that identifies the command (though
9200 blank lines are also allowed for punctuation). If a line of an
9201 MRI-compatible script begins with an unrecognized keyword, @command{ld}
9202 issues a warning message, but continues processing the script.
9204 Lines beginning with @samp{*} are comments.
9206 You can write these commands using all upper-case letters, or all
9207 lower case; for example, @samp{chip} is the same as @samp{CHIP}.
9208 The following list shows only the upper-case form of each command.
9211 @cindex @code{ABSOLUTE} (MRI)
9212 @item ABSOLUTE @var{secname}
9213 @itemx ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname}
9214 Normally, @command{ld} includes in the output file all sections from all
9215 the input files. However, in an MRI-compatible script, you can use the
9216 @code{ABSOLUTE} command to restrict the sections that will be present in
9217 your output program. If the @code{ABSOLUTE} command is used at all in a
9218 script, then only the sections named explicitly in @code{ABSOLUTE}
9219 commands will appear in the linker output. You can still use other
9220 input sections (whatever you select on the command line, or using
9221 @code{LOAD}) to resolve addresses in the output file.
9223 @cindex @code{ALIAS} (MRI)
9224 @item ALIAS @var{out-secname}, @var{in-secname}
9225 Use this command to place the data from input section @var{in-secname}
9226 in a section called @var{out-secname} in the linker output file.
9228 @var{in-secname} may be an integer.
9230 @cindex @code{ALIGN} (MRI)
9231 @item ALIGN @var{secname} = @var{expression}
9232 Align the section called @var{secname} to @var{expression}. The
9233 @var{expression} should be a power of two.
9235 @cindex @code{BASE} (MRI)
9236 @item BASE @var{expression}
9237 Use the value of @var{expression} as the lowest address (other than
9238 absolute addresses) in the output file.
9240 @cindex @code{CHIP} (MRI)
9241 @item CHIP @var{expression}
9242 @itemx CHIP @var{expression}, @var{expression}
9243 This command does nothing; it is accepted only for compatibility.
9245 @cindex @code{END} (MRI)
9247 This command does nothing whatever; it's only accepted for compatibility.
9249 @cindex @code{FORMAT} (MRI)
9250 @item FORMAT @var{output-format}
9251 Similar to the @code{OUTPUT_FORMAT} command in the more general linker
9252 language, but restricted to S-records, if @var{output-format} is @samp{S}
9254 @cindex @code{LIST} (MRI)
9255 @item LIST @var{anything}@dots{}
9256 Print (to the standard output file) a link map, as produced by the
9257 @command{ld} command-line option @samp{-M}.
9259 The keyword @code{LIST} may be followed by anything on the
9260 same line, with no change in its effect.
9262 @cindex @code{LOAD} (MRI)
9263 @item LOAD @var{filename}
9264 @itemx LOAD @var{filename}, @var{filename}, @dots{} @var{filename}
9265 Include one or more object file @var{filename} in the link; this has the
9266 same effect as specifying @var{filename} directly on the @command{ld}
9269 @cindex @code{NAME} (MRI)
9270 @item NAME @var{output-name}
9271 @var{output-name} is the name for the program produced by @command{ld}; the
9272 MRI-compatible command @code{NAME} is equivalent to the command-line
9273 option @samp{-o} or the general script language command @code{OUTPUT}.
9275 @cindex @code{ORDER} (MRI)
9276 @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname}
9277 @itemx ORDER @var{secname} @var{secname} @var{secname}
9278 Normally, @command{ld} orders the sections in its output file in the
9279 order in which they first appear in the input files. In an MRI-compatible
9280 script, you can override this ordering with the @code{ORDER} command. The
9281 sections you list with @code{ORDER} will appear first in your output
9282 file, in the order specified.
9284 @cindex @code{PUBLIC} (MRI)
9285 @item PUBLIC @var{name}=@var{expression}
9286 @itemx PUBLIC @var{name},@var{expression}
9287 @itemx PUBLIC @var{name} @var{expression}
9288 Supply a value (@var{expression}) for external symbol
9289 @var{name} used in the linker input files.
9291 @cindex @code{SECT} (MRI)
9292 @item SECT @var{secname}, @var{expression}
9293 @itemx SECT @var{secname}=@var{expression}
9294 @itemx SECT @var{secname} @var{expression}
9295 You can use any of these three forms of the @code{SECT} command to
9296 specify the start address (@var{expression}) for section @var{secname}.
9297 If you have more than one @code{SECT} statement for the same
9298 @var{secname}, only the @emph{first} sets the start address.
9301 @node GNU Free Documentation License
9302 @appendix GNU Free Documentation License
9306 @unnumbered LD Index
9311 % I think something like @@colophon should be in texinfo. In the
9313 \long\def\colophon{\hbox to0pt{}\vfill
9314 \centerline{The body of this manual is set in}
9315 \centerline{\fontname\tenrm,}
9316 \centerline{with headings in {\bf\fontname\tenbf}}
9317 \centerline{and examples in {\tt\fontname\tentt}.}
9318 \centerline{{\it\fontname\tenit\/} and}
9319 \centerline{{\sl\fontname\tensl\/}}
9320 \centerline{are used for emphasis.}\vfill}
9322 % Blame: doc@@cygnus.com, 28mar91.