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 the 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.
1297 This option is only meaningful when building a shared object.
1298 It marks the object so that its runtime initialization will occur
1299 before the runtime initialization of any other objects brought into
1300 the process at the same time. Similarly the runtime finalization of
1301 the object will occur after the runtime finalization of any other
1305 Specify that the dynamic loader should modify its symbol search order
1306 so that symbols in this shared library interpose all other shared
1307 libraries not so marked.
1311 When generating a shared library or other dynamically loadable ELF
1312 object mark it as one that should (by default) only ever be loaded once,
1313 and only in the main namespace (when using @code{dlmopen}). This is
1314 primarily used to mark fundamental libraries such as libc, libpthread et
1315 al which do not usually function correctly unless they are the sole instances
1316 of themselves. This behaviour can be overridden by the @code{dlmopen} caller
1317 and does not apply to certain loading mechanisms (such as audit libraries).
1320 Generate GNU_PROPERTY_X86_FEATURE_1_LAM_U48 in .note.gnu.property section
1321 to indicate compatibility with Intel LAM_U48. Supported for Linux/x86_64.
1324 Generate GNU_PROPERTY_X86_FEATURE_1_LAM_U57 in .note.gnu.property section
1325 to indicate compatibility with Intel LAM_U57. Supported for Linux/x86_64.
1327 @item lam-u48-report=none
1328 @itemx lam-u48-report=warning
1329 @itemx lam-u48-report=error
1330 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U48
1331 property in input .note.gnu.property section.
1332 @option{lam-u48-report=none}, which is the default, will make the
1333 linker not report missing properties in input files.
1334 @option{lam-u48-report=warning} will make the linker issue a warning for
1335 missing properties in input files. @option{lam-u48-report=error} will
1336 make the linker issue an error for missing properties in input files.
1337 Supported for Linux/x86_64.
1339 @item lam-u57-report=none
1340 @itemx lam-u57-report=warning
1341 @itemx lam-u57-report=error
1342 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U57
1343 property in input .note.gnu.property section.
1344 @option{lam-u57-report=none}, which is the default, will make the
1345 linker not report missing properties in input files.
1346 @option{lam-u57-report=warning} will make the linker issue a warning for
1347 missing properties in input files. @option{lam-u57-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-report=none
1352 @itemx lam-report=warning
1353 @itemx lam-report=error
1354 Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U48 and
1355 GNU_PROPERTY_X86_FEATURE_1_LAM_U57 properties in input .note.gnu.property
1356 section. @option{lam-report=none}, which is the default, will make the
1357 linker not report missing properties in input files.
1358 @option{lam-report=warning} will make the linker issue a warning for
1359 missing properties in input files. @option{lam-report=error} will make
1360 the linker issue an error for missing properties in input files.
1361 Supported for Linux/x86_64.
1364 When generating an executable or shared library, mark it to tell the
1365 dynamic linker to defer function call resolution to the point when
1366 the function is called (lazy binding), rather than at load time.
1367 Lazy binding is the default.
1370 Specify that the object's filters be processed immediately at runtime.
1372 @item max-page-size=@var{value}
1373 Set the maximum memory page size supported to @var{value}.
1376 Allow multiple definitions.
1379 Disable linker generated .dynbss variables used in place of variables
1380 defined in shared libraries. May result in dynamic text relocations.
1383 Specify that the dynamic loader search for dependencies of this object
1384 should ignore any default library search paths.
1387 Specify that the object shouldn't be unloaded at runtime.
1390 Specify that the object is not available to @code{dlopen}.
1393 Specify that the object can not be dumped by @code{dldump}.
1396 Marks the object as not requiring executable stack.
1398 @item noextern-protected-data
1399 Don't treat protected data symbols as external when building a shared
1400 library. This option overrides the linker backend default. It can be
1401 used to work around incorrect relocations against protected data symbols
1402 generated by compiler. Updates on protected data symbols by another
1403 module aren't visible to the resulting shared library. Supported for
1406 @item noreloc-overflow
1407 Disable relocation overflow check. This can be used to disable
1408 relocation overflow check if there will be no dynamic relocation
1409 overflow at run-time. Supported for x86_64.
1412 When generating an executable or shared library, mark it to tell the
1413 dynamic linker to resolve all symbols when the program is started, or
1414 when the shared library is loaded by dlopen, instead of deferring
1415 function call resolution to the point when the function is first
1419 Specify that the object requires @samp{$ORIGIN} handling in paths.
1423 Create an ELF @code{PT_GNU_RELRO} segment header in the object. This
1424 specifies a memory segment that should be made read-only after
1425 relocation, if supported. Specifying @samp{common-page-size} smaller
1426 than the system page size will render this protection ineffective.
1427 Don't create an ELF @code{PT_GNU_RELRO} segment if @samp{norelro}.
1429 @item report-relative-reloc
1430 Report dynamic relative relocations generated by linker. Supported for
1431 Linux/i386 and Linux/x86_64.
1434 @itemx noseparate-code
1435 Create separate code @code{PT_LOAD} segment header in the object. This
1436 specifies a memory segment that should contain only instructions and must
1437 be in wholly disjoint pages from any other data. Don't create separate
1438 code @code{PT_LOAD} segment if @samp{noseparate-code} is used.
1441 @itemx nounique-symbol
1442 Avoid duplicated local symbol names in the symbol string table. Append
1443 ".@code{number}" to duplicated local symbol names if @samp{unique-symbol}
1444 is used. @option{nounique-symbol} is the default.
1447 Generate GNU_PROPERTY_X86_FEATURE_1_SHSTK in .note.gnu.property section
1448 to indicate compatibility with Intel Shadow Stack. Supported for
1449 Linux/i386 and Linux/x86_64.
1451 @item stack-size=@var{value}
1452 Specify a stack size for an ELF @code{PT_GNU_STACK} segment.
1453 Specifying zero will override any default non-zero sized
1454 @code{PT_GNU_STACK} segment creation.
1457 @itemx nostart-stop-gc
1458 @cindex start-stop-gc
1459 When @samp{--gc-sections} is in effect, a reference from a retained
1460 section to @code{__start_SECNAME} or @code{__stop_SECNAME} causes all
1461 input sections named @code{SECNAME} to also be retained, if
1462 @code{SECNAME} is representable as a C identifier and either
1463 @code{__start_SECNAME} or @code{__stop_SECNAME} is synthesized by the
1464 linker. @samp{-z start-stop-gc} disables this effect, allowing
1465 sections to be garbage collected as if the special synthesized symbols
1466 were not defined. @samp{-z start-stop-gc} has no effect on a
1467 definition of @code{__start_SECNAME} or @code{__stop_SECNAME} in an
1468 object file or linker script. Such a definition will prevent the
1469 linker providing a synthesized @code{__start_SECNAME} or
1470 @code{__stop_SECNAME} respectively, and therefore the special
1471 treatment by garbage collection for those references.
1473 @item start-stop-visibility=@var{value}
1475 @cindex ELF symbol visibility
1476 Specify the ELF symbol visibility for synthesized
1477 @code{__start_SECNAME} and @code{__stop_SECNAME} symbols (@pxref{Input
1478 Section Example}). @var{value} must be exactly @samp{default},
1479 @samp{internal}, @samp{hidden}, or @samp{protected}. If no @samp{-z
1480 start-stop-visibility} option is given, @samp{protected} is used for
1481 compatibility with historical practice. However, it's highly
1482 recommended to use @samp{-z start-stop-visibility=hidden} in new
1483 programs and shared libraries so that these symbols are not exported
1484 between shared objects, which is not usually what's intended.
1489 Report an error if DT_TEXTREL is set, i.e., if the position-independent
1490 or shared object has dynamic relocations in read-only sections. Don't
1491 report an error if @samp{notext} or @samp{textoff}.
1494 Do not report unresolved symbol references from regular object files,
1495 either when creating an executable, or when creating a shared library.
1496 This option is the inverse of @samp{-z defs}.
1498 @item x86-64-baseline
1502 Specify the x86-64 ISA level needed in .note.gnu.property section.
1503 @option{x86-64-baseline} generates @code{GNU_PROPERTY_X86_ISA_1_BASELINE}.
1504 @option{x86-64-v2} generates @code{GNU_PROPERTY_X86_ISA_1_V2}.
1505 @option{x86-64-v3} generates @code{GNU_PROPERTY_X86_ISA_1_V3}.
1506 @option{x86-64-v4} generates @code{GNU_PROPERTY_X86_ISA_1_V4}.
1507 Supported for Linux/i386 and Linux/x86_64.
1511 Other keywords are ignored for Solaris compatibility.
1514 @cindex groups of archives
1515 @item -( @var{archives} -)
1516 @itemx --start-group @var{archives} --end-group
1517 The @var{archives} should be a list of archive files. They may be
1518 either explicit file names, or @samp{-l} options.
1520 The specified archives are searched repeatedly until no new undefined
1521 references are created. Normally, an archive is searched only once in
1522 the order that it is specified on the command line. If a symbol in that
1523 archive is needed to resolve an undefined symbol referred to by an
1524 object in an archive that appears later on the command line, the linker
1525 would not be able to resolve that reference. By grouping the archives,
1526 they will all be searched repeatedly until all possible references are
1529 Using this option has a significant performance cost. It is best to use
1530 it only when there are unavoidable circular references between two or
1533 @kindex --accept-unknown-input-arch
1534 @kindex --no-accept-unknown-input-arch
1535 @item --accept-unknown-input-arch
1536 @itemx --no-accept-unknown-input-arch
1537 Tells the linker to accept input files whose architecture cannot be
1538 recognised. The assumption is that the user knows what they are doing
1539 and deliberately wants to link in these unknown input files. This was
1540 the default behaviour of the linker, before release 2.14. The default
1541 behaviour from release 2.14 onwards is to reject such input files, and
1542 so the @samp{--accept-unknown-input-arch} option has been added to
1543 restore the old behaviour.
1546 @kindex --no-as-needed
1548 @itemx --no-as-needed
1549 This option affects ELF DT_NEEDED tags for dynamic libraries mentioned
1550 on the command line after the @option{--as-needed} option. Normally
1551 the linker will add a DT_NEEDED tag for each dynamic library mentioned
1552 on the command line, regardless of whether the library is actually
1553 needed or not. @option{--as-needed} causes a DT_NEEDED tag to only be
1554 emitted for a library that @emph{at that point in the link} satisfies a
1555 non-weak undefined symbol reference from a regular object file or, if
1556 the library is not found in the DT_NEEDED lists of other needed libraries, a
1557 non-weak undefined symbol reference from another needed dynamic library.
1558 Object files or libraries appearing on the command line @emph{after}
1559 the library in question do not affect whether the library is seen as
1560 needed. This is similar to the rules for extraction of object files
1561 from archives. @option{--no-as-needed} restores the default behaviour.
1563 @kindex --add-needed
1564 @kindex --no-add-needed
1566 @itemx --no-add-needed
1567 These two options have been deprecated because of the similarity of
1568 their names to the @option{--as-needed} and @option{--no-as-needed}
1569 options. They have been replaced by @option{--copy-dt-needed-entries}
1570 and @option{--no-copy-dt-needed-entries}.
1572 @kindex -assert @var{keyword}
1573 @item -assert @var{keyword}
1574 This option is ignored for SunOS compatibility.
1578 @kindex -call_shared
1582 Link against dynamic libraries. This is only meaningful on platforms
1583 for which shared libraries are supported. This option is normally the
1584 default on such platforms. The different variants of this option are
1585 for compatibility with various systems. You may use this option
1586 multiple times on the command line: it affects library searching for
1587 @option{-l} options which follow it.
1591 Set the @code{DF_1_GROUP} flag in the @code{DT_FLAGS_1} entry in the dynamic
1592 section. This causes the runtime linker to handle lookups in this
1593 object and its dependencies to be performed only inside the group.
1594 @option{--unresolved-symbols=report-all} is implied. This option is
1595 only meaningful on ELF platforms which support shared libraries.
1605 Do not link against shared libraries. This is only meaningful on
1606 platforms for which shared libraries are supported. The different
1607 variants of this option are for compatibility with various systems. You
1608 may use this option multiple times on the command line: it affects
1609 library searching for @option{-l} options which follow it. This
1610 option also implies @option{--unresolved-symbols=report-all}. This
1611 option can be used with @option{-shared}. Doing so means that a
1612 shared library is being created but that all of the library's external
1613 references must be resolved by pulling in entries from static
1618 When creating a shared library, bind references to global symbols to the
1619 definition within the shared library, if any. Normally, it is possible
1620 for a program linked against a shared library to override the definition
1621 within the shared library. This option is only meaningful on ELF
1622 platforms which support shared libraries.
1624 @kindex -Bsymbolic-functions
1625 @item -Bsymbolic-functions
1626 When creating a shared library, bind references to global function
1627 symbols to the definition within the shared library, if any.
1628 This option is only meaningful on ELF platforms which support shared
1631 @kindex --dynamic-list=@var{dynamic-list-file}
1632 @item --dynamic-list=@var{dynamic-list-file}
1633 Specify the name of a dynamic list file to the linker. This is
1634 typically used when creating shared libraries to specify a list of
1635 global symbols whose references shouldn't be bound to the definition
1636 within the shared library, or creating dynamically linked executables
1637 to specify a list of symbols which should be added to the symbol table
1638 in the executable. This option is only meaningful on ELF platforms
1639 which support shared libraries.
1641 The format of the dynamic list is the same as the version node without
1642 scope and node name. See @ref{VERSION} for more information.
1644 @kindex --dynamic-list-data
1645 @item --dynamic-list-data
1646 Include all global data symbols to the dynamic list.
1648 @kindex --dynamic-list-cpp-new
1649 @item --dynamic-list-cpp-new
1650 Provide the builtin dynamic list for C++ operator new and delete. It
1651 is mainly useful for building shared libstdc++.
1653 @kindex --dynamic-list-cpp-typeinfo
1654 @item --dynamic-list-cpp-typeinfo
1655 Provide the builtin dynamic list for C++ runtime type identification.
1657 @kindex --check-sections
1658 @kindex --no-check-sections
1659 @item --check-sections
1660 @itemx --no-check-sections
1661 Asks the linker @emph{not} to check section addresses after they have
1662 been assigned to see if there are any overlaps. Normally the linker will
1663 perform this check, and if it finds any overlaps it will produce
1664 suitable error messages. The linker does know about, and does make
1665 allowances for sections in overlays. The default behaviour can be
1666 restored by using the command-line switch @option{--check-sections}.
1667 Section overlap is not usually checked for relocatable links. You can
1668 force checking in that case by using the @option{--check-sections}
1671 @kindex --copy-dt-needed-entries
1672 @kindex --no-copy-dt-needed-entries
1673 @item --copy-dt-needed-entries
1674 @itemx --no-copy-dt-needed-entries
1675 This option affects the treatment of dynamic libraries referred to
1676 by DT_NEEDED tags @emph{inside} ELF dynamic libraries mentioned on the
1677 command line. Normally the linker won't add a DT_NEEDED tag to the
1678 output binary for each library mentioned in a DT_NEEDED tag in an
1679 input dynamic library. With @option{--copy-dt-needed-entries}
1680 specified on the command line however any dynamic libraries that
1681 follow it will have their DT_NEEDED entries added. The default
1682 behaviour can be restored with @option{--no-copy-dt-needed-entries}.
1684 This option also has an effect on the resolution of symbols in dynamic
1685 libraries. With @option{--copy-dt-needed-entries} dynamic libraries
1686 mentioned on the command line will be recursively searched, following
1687 their DT_NEEDED tags to other libraries, in order to resolve symbols
1688 required by the output binary. With the default setting however
1689 the searching of dynamic libraries that follow it will stop with the
1690 dynamic library itself. No DT_NEEDED links will be traversed to resolve
1693 @cindex cross reference table
1696 Output a cross reference table. If a linker map file is being
1697 generated, the cross reference table is printed to the map file.
1698 Otherwise, it is printed on the standard output.
1700 The format of the table is intentionally simple, so that it may be
1701 easily processed by a script if necessary. The symbols are printed out,
1702 sorted by name. For each symbol, a list of file names is given. If the
1703 symbol is defined, the first file listed is the location of the
1704 definition. If the symbol is defined as a common value then any files
1705 where this happens appear next. Finally any files that reference the
1708 @cindex ctf variables
1709 @kindex --ctf-variables
1710 @kindex --no-ctf-variables
1711 @item --ctf-variables
1712 @item --no-ctf-variables
1713 The CTF debuginfo format supports a section which encodes the names and
1714 types of variables found in the program which do not appear in any symbol
1715 table. These variables clearly cannot be looked up by address by
1716 conventional debuggers, so the space used for their types and names is
1717 usually wasted: the types are usually small but the names are often not.
1718 @option{--ctf-variables} causes the generation of such a section.
1719 The default behaviour can be restored with @option{--no-ctf-variables}.
1721 @cindex ctf type sharing
1722 @kindex --ctf-share-types
1723 @item --ctf-share-types=@var{method}
1724 Adjust the method used to share types between translation units in CTF.
1727 @item share-unconflicted
1728 Put all types that do not have ambiguous definitions into the shared dictionary,
1729 where debuggers can easily access them, even if they only occur in one
1730 translation unit. This is the default.
1732 @item share-duplicated
1733 Put only types that occur in multiple translation units into the shared
1734 dictionary: types with only one definition go into per-translation-unit
1735 dictionaries. Types with ambiguous definitions in multiple translation units
1736 always go into per-translation-unit dictionaries. This tends to make the CTF
1737 larger, but may reduce the amount of CTF in the shared dictionary. For very
1738 large projects this may speed up opening the CTF and save memory in the CTF
1739 consumer at runtime.
1742 @cindex common allocation
1743 @kindex --no-define-common
1744 @item --no-define-common
1745 This option inhibits the assignment of addresses to common symbols.
1746 The script command @code{INHIBIT_COMMON_ALLOCATION} has the same effect.
1747 @xref{Miscellaneous Commands}.
1749 The @samp{--no-define-common} option allows decoupling
1750 the decision to assign addresses to Common symbols from the choice
1751 of the output file type; otherwise a non-Relocatable output type
1752 forces assigning addresses to Common symbols.
1753 Using @samp{--no-define-common} allows Common symbols that are referenced
1754 from a shared library to be assigned addresses only in the main program.
1755 This eliminates the unused duplicate space in the shared library,
1756 and also prevents any possible confusion over resolving to the wrong
1757 duplicate when there are many dynamic modules with specialized search
1758 paths for runtime symbol resolution.
1760 @cindex group allocation in linker script
1761 @cindex section groups
1763 @kindex --force-group-allocation
1764 @item --force-group-allocation
1765 This option causes the linker to place section group members like
1766 normal input sections, and to delete the section groups. This is the
1767 default behaviour for a final link but this option can be used to
1768 change the behaviour of a relocatable link (@samp{-r}). The script
1769 command @code{FORCE_GROUP_ALLOCATION} has the same
1770 effect. @xref{Miscellaneous Commands}.
1772 @cindex symbols, from command line
1773 @kindex --defsym=@var{symbol}=@var{exp}
1774 @item --defsym=@var{symbol}=@var{expression}
1775 Create a global symbol in the output file, containing the absolute
1776 address given by @var{expression}. You may use this option as many
1777 times as necessary to define multiple symbols in the command line. A
1778 limited form of arithmetic is supported for the @var{expression} in this
1779 context: you may give a hexadecimal constant or the name of an existing
1780 symbol, or use @code{+} and @code{-} to add or subtract hexadecimal
1781 constants or symbols. If you need more elaborate expressions, consider
1782 using the linker command language from a script (@pxref{Assignments}).
1783 @emph{Note:} there should be no white space between @var{symbol}, the
1784 equals sign (``@key{=}''), and @var{expression}.
1786 The linker processes @samp{--defsym} arguments and @samp{-T} arguments
1787 in order, placing @samp{--defsym} before @samp{-T} will define the
1788 symbol before the linker script from @samp{-T} is processed, while
1789 placing @samp{--defsym} after @samp{-T} will define the symbol after
1790 the linker script has been processed. This difference has
1791 consequences for expressions within the linker script that use the
1792 @samp{--defsym} symbols, which order is correct will depend on what
1793 you are trying to achieve.
1795 @cindex demangling, from command line
1796 @kindex --demangle[=@var{style}]
1797 @kindex --no-demangle
1798 @item --demangle[=@var{style}]
1799 @itemx --no-demangle
1800 These options control whether to demangle symbol names in error messages
1801 and other output. When the linker is told to demangle, it tries to
1802 present symbol names in a readable fashion: it strips leading
1803 underscores if they are used by the object file format, and converts C++
1804 mangled symbol names into user readable names. Different compilers have
1805 different mangling styles. The optional demangling style argument can be used
1806 to choose an appropriate demangling style for your compiler. The linker will
1807 demangle by default unless the environment variable @samp{COLLECT_NO_DEMANGLE}
1808 is set. These options may be used to override the default.
1810 @cindex dynamic linker, from command line
1811 @kindex -I@var{file}
1812 @kindex --dynamic-linker=@var{file}
1814 @itemx --dynamic-linker=@var{file}
1815 Set the name of the dynamic linker. This is only meaningful when
1816 generating dynamically linked ELF executables. The default dynamic
1817 linker is normally correct; don't use this unless you know what you are
1820 @kindex --no-dynamic-linker
1821 @item --no-dynamic-linker
1822 When producing an executable file, omit the request for a dynamic
1823 linker to be used at load-time. This is only meaningful for ELF
1824 executables that contain dynamic relocations, and usually requires
1825 entry point code that is capable of processing these relocations.
1827 @kindex --embedded-relocs
1828 @item --embedded-relocs
1829 This option is similar to the @option{--emit-relocs} option except
1830 that the relocs are stored in a target-specific section. This option
1831 is only supported by the @samp{BFIN}, @samp{CR16} and @emph{M68K}
1834 @kindex --disable-multiple-abs-defs
1835 @item --disable-multiple-abs-defs
1836 Do not allow multiple definitions with symbols included
1837 in filename invoked by -R or --just-symbols
1839 @kindex --fatal-warnings
1840 @kindex --no-fatal-warnings
1841 @item --fatal-warnings
1842 @itemx --no-fatal-warnings
1843 Treat all warnings as errors. The default behaviour can be restored
1844 with the option @option{--no-fatal-warnings}.
1846 @kindex --force-exe-suffix
1847 @item --force-exe-suffix
1848 Make sure that an output file has a .exe suffix.
1850 If a successfully built fully linked output file does not have a
1851 @code{.exe} or @code{.dll} suffix, this option forces the linker to copy
1852 the output file to one of the same name with a @code{.exe} suffix. This
1853 option is useful when using unmodified Unix makefiles on a Microsoft
1854 Windows host, since some versions of Windows won't run an image unless
1855 it ends in a @code{.exe} suffix.
1857 @kindex --gc-sections
1858 @kindex --no-gc-sections
1859 @cindex garbage collection
1861 @itemx --no-gc-sections
1862 Enable garbage collection of unused input sections. It is ignored on
1863 targets that do not support this option. The default behaviour (of not
1864 performing this garbage collection) can be restored by specifying
1865 @samp{--no-gc-sections} on the command line. Note that garbage
1866 collection for COFF and PE format targets is supported, but the
1867 implementation is currently considered to be experimental.
1869 @samp{--gc-sections} decides which input sections are used by
1870 examining symbols and relocations. The section containing the entry
1871 symbol and all sections containing symbols undefined on the
1872 command-line will be kept, as will sections containing symbols
1873 referenced by dynamic objects. Note that when building shared
1874 libraries, the linker must assume that any visible symbol is
1875 referenced. Once this initial set of sections has been determined,
1876 the linker recursively marks as used any section referenced by their
1877 relocations. See @samp{--entry}, @samp{--undefined}, and
1878 @samp{--gc-keep-exported}.
1880 This option can be set when doing a partial link (enabled with option
1881 @samp{-r}). In this case the root of symbols kept must be explicitly
1882 specified either by one of the options @samp{--entry},
1883 @samp{--undefined}, or @samp{--gc-keep-exported} or by a @code{ENTRY}
1884 command in the linker script.
1886 As a GNU extension, ELF input sections marked with the
1887 @code{SHF_GNU_RETAIN} flag will not be garbage collected.
1889 @kindex --print-gc-sections
1890 @kindex --no-print-gc-sections
1891 @cindex garbage collection
1892 @item --print-gc-sections
1893 @itemx --no-print-gc-sections
1894 List all sections removed by garbage collection. The listing is
1895 printed on stderr. This option is only effective if garbage
1896 collection has been enabled via the @samp{--gc-sections}) option. The
1897 default behaviour (of not listing the sections that are removed) can
1898 be restored by specifying @samp{--no-print-gc-sections} on the command
1901 @kindex --gc-keep-exported
1902 @cindex garbage collection
1903 @item --gc-keep-exported
1904 When @samp{--gc-sections} is enabled, this option prevents garbage
1905 collection of unused input sections that contain global symbols having
1906 default or protected visibility. This option is intended to be used for
1907 executables where unreferenced sections would otherwise be garbage
1908 collected regardless of the external visibility of contained symbols.
1909 Note that this option has no effect when linking shared objects since
1910 it is already the default behaviour. This option is only supported for
1913 @kindex --print-output-format
1914 @cindex output format
1915 @item --print-output-format
1916 Print the name of the default output format (perhaps influenced by
1917 other command-line options). This is the string that would appear
1918 in an @code{OUTPUT_FORMAT} linker script command (@pxref{File Commands}).
1920 @kindex --print-memory-usage
1921 @cindex memory usage
1922 @item --print-memory-usage
1923 Print used size, total size and used size of memory regions created with
1924 the @ref{MEMORY} command. This is useful on embedded targets to have a
1925 quick view of amount of free memory. The format of the output has one
1926 headline and one line per region. It is both human readable and easily
1927 parsable by tools. Here is an example of an output:
1930 Memory region Used Size Region Size %age Used
1931 ROM: 256 KB 1 MB 25.00%
1932 RAM: 32 B 2 GB 0.00%
1939 Print a summary of the command-line options on the standard output and exit.
1941 @kindex --target-help
1943 Print a summary of all target-specific options on the standard output and exit.
1945 @kindex -Map=@var{mapfile}
1946 @item -Map=@var{mapfile}
1947 Print a link map to the file @var{mapfile}. See the description of the
1948 @option{-M} option, above. If @var{mapfile} is just the character
1949 @code{-} then the map will be written to stdout.
1951 Specifying a directory as @var{mapfile} causes the linker map to be
1952 written as a file inside the directory. Normally name of the file
1953 inside the directory is computed as the basename of the @var{output}
1954 file with @code{.map} appended. If however the special character
1955 @code{%} is used then this will be replaced by the full path of the
1956 output file. Additionally if there are any characters after the
1957 @var{%} symbol then @code{.map} will no longer be appended.
1960 -o foo.exe -Map=bar [Creates ./bar]
1961 -o ../dir/foo.exe -Map=bar [Creates ./bar]
1962 -o foo.exe -Map=../dir [Creates ../dir/foo.exe.map]
1963 -o ../dir2/foo.exe -Map=../dir [Creates ../dir/foo.exe.map]
1964 -o foo.exe -Map=% [Creates ./foo.exe.map]
1965 -o ../dir/foo.exe -Map=% [Creates ../dir/foo.exe.map]
1966 -o foo.exe -Map=%.bar [Creates ./foo.exe.bar]
1967 -o ../dir/foo.exe -Map=%.bar [Creates ../dir/foo.exe.bar]
1968 -o ../dir2/foo.exe -Map=../dir/% [Creates ../dir/../dir2/foo.exe.map]
1969 -o ../dir2/foo.exe -Map=../dir/%.bar [Creates ../dir/../dir2/foo.exe.bar]
1972 It is an error to specify more than one @code{%} character.
1974 If the map file already exists then it will be overwritten by this
1977 @cindex memory usage
1978 @kindex --no-keep-memory
1979 @item --no-keep-memory
1980 @command{ld} normally optimizes for speed over memory usage by caching the
1981 symbol tables of input files in memory. This option tells @command{ld} to
1982 instead optimize for memory usage, by rereading the symbol tables as
1983 necessary. This may be required if @command{ld} runs out of memory space
1984 while linking a large executable.
1986 @kindex --no-undefined
1989 @item --no-undefined
1991 Report unresolved symbol references from regular object files. This
1992 is done even if the linker is creating a non-symbolic shared library.
1993 The switch @option{--[no-]allow-shlib-undefined} controls the
1994 behaviour for reporting unresolved references found in shared
1995 libraries being linked in.
1997 The effects of this option can be reverted by using @code{-z undefs}.
1999 @kindex --allow-multiple-definition
2001 @item --allow-multiple-definition
2003 Normally when a symbol is defined multiple times, the linker will
2004 report a fatal error. These options allow multiple definitions and the
2005 first definition will be used.
2007 @kindex --allow-shlib-undefined
2008 @kindex --no-allow-shlib-undefined
2009 @item --allow-shlib-undefined
2010 @itemx --no-allow-shlib-undefined
2011 Allows or disallows undefined symbols in shared libraries.
2012 This switch is similar to @option{--no-undefined} except that it
2013 determines the behaviour when the undefined symbols are in a
2014 shared library rather than a regular object file. It does not affect
2015 how undefined symbols in regular object files are handled.
2017 The default behaviour is to report errors for any undefined symbols
2018 referenced in shared libraries if the linker is being used to create
2019 an executable, but to allow them if the linker is being used to create
2022 The reasons for allowing undefined symbol references in shared
2023 libraries specified at link time are that:
2027 A shared library specified at link time may not be the same as the one
2028 that is available at load time, so the symbol might actually be
2029 resolvable at load time.
2031 There are some operating systems, eg BeOS and HPPA, where undefined
2032 symbols in shared libraries are normal.
2034 The BeOS kernel for example patches shared libraries at load time to
2035 select whichever function is most appropriate for the current
2036 architecture. This is used, for example, to dynamically select an
2037 appropriate memset function.
2040 @kindex --error-handling-script=@var{scriptname}
2041 @item --error-handling-script=@var{scriptname}
2042 If this option is provided then the linker will invoke
2043 @var{scriptname} whenever an error is encountered. Currently however
2044 only two kinds of error are supported: missing symbols and missing
2045 libraries. Two arguments will be passed to script: the keyword
2046 ``undefined-symbol'' or `missing-lib'' and the @var{name} of the
2047 undefined symbol or missing library. The intention is that the script
2048 will provide suggestions to the user as to where the symbol or library
2049 might be found. After the script has finished then the normal linker
2050 error message will be displayed.
2052 The availability of this option is controlled by a configure time
2053 switch, so it may not be present in specific implementations.
2055 @kindex --no-undefined-version
2056 @item --no-undefined-version
2057 Normally when a symbol has an undefined version, the linker will ignore
2058 it. This option disallows symbols with undefined version and a fatal error
2059 will be issued instead.
2061 @kindex --default-symver
2062 @item --default-symver
2063 Create and use a default symbol version (the soname) for unversioned
2066 @kindex --default-imported-symver
2067 @item --default-imported-symver
2068 Create and use a default symbol version (the soname) for unversioned
2071 @kindex --no-warn-mismatch
2072 @item --no-warn-mismatch
2073 Normally @command{ld} will give an error if you try to link together input
2074 files that are mismatched for some reason, perhaps because they have
2075 been compiled for different processors or for different endiannesses.
2076 This option tells @command{ld} that it should silently permit such possible
2077 errors. This option should only be used with care, in cases when you
2078 have taken some special action that ensures that the linker errors are
2081 @kindex --no-warn-search-mismatch
2082 @item --no-warn-search-mismatch
2083 Normally @command{ld} will give a warning if it finds an incompatible
2084 library during a library search. This option silences the warning.
2086 @kindex --no-whole-archive
2087 @item --no-whole-archive
2088 Turn off the effect of the @option{--whole-archive} option for subsequent
2091 @cindex output file after errors
2092 @kindex --noinhibit-exec
2093 @item --noinhibit-exec
2094 Retain the executable output file whenever it is still usable.
2095 Normally, the linker will not produce an output file if it encounters
2096 errors during the link process; it exits without writing an output file
2097 when it issues any error whatsoever.
2101 Only search library directories explicitly specified on the
2102 command line. Library directories specified in linker scripts
2103 (including linker scripts specified on the command line) are ignored.
2105 @ifclear SingleFormat
2106 @kindex --oformat=@var{output-format}
2107 @item --oformat=@var{output-format}
2108 @command{ld} may be configured to support more than one kind of object
2109 file. If your @command{ld} is configured this way, you can use the
2110 @samp{--oformat} option to specify the binary format for the output
2111 object file. Even when @command{ld} is configured to support alternative
2112 object formats, you don't usually need to specify this, as @command{ld}
2113 should be configured to produce as a default output format the most
2114 usual format on each machine. @var{output-format} is a text string, the
2115 name of a particular format supported by the BFD libraries. (You can
2116 list the available binary formats with @samp{objdump -i}.) The script
2117 command @code{OUTPUT_FORMAT} can also specify the output format, but
2118 this option overrides it. @xref{BFD}.
2121 @kindex --out-implib
2122 @item --out-implib @var{file}
2123 Create an import library in @var{file} corresponding to the executable
2124 the linker is generating (eg. a DLL or ELF program). This import
2125 library (which should be called @code{*.dll.a} or @code{*.a} for DLLs)
2126 may be used to link clients against the generated executable; this
2127 behaviour makes it possible to skip a separate import library creation
2128 step (eg. @code{dlltool} for DLLs). This option is only available for
2129 the i386 PE and ELF targetted ports of the linker.
2132 @kindex --pic-executable
2134 @itemx --pic-executable
2135 @cindex position independent executables
2136 Create a position independent executable. This is currently only supported on
2137 ELF platforms. Position independent executables are similar to shared
2138 libraries in that they are relocated by the dynamic linker to the virtual
2139 address the OS chooses for them (which can vary between invocations). Like
2140 normal dynamically linked executables they can be executed and symbols
2141 defined in the executable cannot be overridden by shared libraries.
2145 This option is ignored for Linux compatibility.
2149 This option is ignored for SVR4 compatibility.
2152 @cindex synthesizing linker
2153 @cindex relaxing addressing modes
2157 An option with machine dependent effects.
2159 This option is only supported on a few targets.
2162 @xref{H8/300,,@command{ld} and the H8/300}.
2165 @xref{Xtensa,, @command{ld} and Xtensa Processors}.
2168 @xref{M68HC11/68HC12,,@command{ld} and the 68HC11 and 68HC12}.
2171 @xref{Nios II,,@command{ld} and the Altera Nios II}.
2174 @xref{PowerPC ELF32,,@command{ld} and PowerPC 32-bit ELF Support}.
2177 On some platforms the @option{--relax} option performs target specific,
2178 global optimizations that become possible when the linker resolves
2179 addressing in the program, such as relaxing address modes,
2180 synthesizing new instructions, selecting shorter version of current
2181 instructions, and combining constant values.
2183 On some platforms these link time global optimizations may make symbolic
2184 debugging of the resulting executable impossible.
2186 This is known to be the case for the Matsushita MN10200 and MN10300
2187 family of processors.
2190 On platforms where the feature is supported, the option
2191 @option{--no-relax} will disable it.
2193 On platforms where the feature is not supported, both @option{--relax}
2194 and @option{--no-relax} are accepted, but ignored.
2196 @cindex retaining specified symbols
2197 @cindex stripping all but some symbols
2198 @cindex symbols, retaining selectively
2199 @kindex --retain-symbols-file=@var{filename}
2200 @item --retain-symbols-file=@var{filename}
2201 Retain @emph{only} the symbols listed in the file @var{filename},
2202 discarding all others. @var{filename} is simply a flat file, with one
2203 symbol name per line. This option is especially useful in environments
2207 where a large global symbol table is accumulated gradually, to conserve
2210 @samp{--retain-symbols-file} does @emph{not} discard undefined symbols,
2211 or symbols needed for relocations.
2213 You may only specify @samp{--retain-symbols-file} once in the command
2214 line. It overrides @samp{-s} and @samp{-S}.
2217 @item -rpath=@var{dir}
2218 @cindex runtime library search path
2219 @kindex -rpath=@var{dir}
2220 Add a directory to the runtime library search path. This is used when
2221 linking an ELF executable with shared objects. All @option{-rpath}
2222 arguments are concatenated and passed to the runtime linker, which uses
2223 them to locate shared objects at runtime.
2225 The @option{-rpath} option is also used when locating shared objects which
2226 are needed by shared objects explicitly included in the link; see the
2227 description of the @option{-rpath-link} option. Searching @option{-rpath}
2228 in this way is only supported by native linkers and cross linkers which
2229 have been configured with the @option{--with-sysroot} option.
2231 If @option{-rpath} is not used when linking an ELF executable, the
2232 contents of the environment variable @code{LD_RUN_PATH} will be used if it
2235 The @option{-rpath} option may also be used on SunOS. By default, on
2236 SunOS, the linker will form a runtime search path out of all the
2237 @option{-L} options it is given. If a @option{-rpath} option is used, the
2238 runtime search path will be formed exclusively using the @option{-rpath}
2239 options, ignoring the @option{-L} options. This can be useful when using
2240 gcc, which adds many @option{-L} options which may be on NFS mounted
2243 For compatibility with other ELF linkers, if the @option{-R} option is
2244 followed by a directory name, rather than a file name, it is treated as
2245 the @option{-rpath} option.
2249 @cindex link-time runtime library search path
2250 @kindex -rpath-link=@var{dir}
2251 @item -rpath-link=@var{dir}
2252 When using ELF or SunOS, one shared library may require another. This
2253 happens when an @code{ld -shared} link includes a shared library as one
2256 When the linker encounters such a dependency when doing a non-shared,
2257 non-relocatable link, it will automatically try to locate the required
2258 shared library and include it in the link, if it is not included
2259 explicitly. In such a case, the @option{-rpath-link} option
2260 specifies the first set of directories to search. The
2261 @option{-rpath-link} option may specify a sequence of directory names
2262 either by specifying a list of names separated by colons, or by
2263 appearing multiple times.
2265 The tokens @var{$ORIGIN} and @var{$LIB} can appear in these search
2266 directories. They will be replaced by the full path to the directory
2267 containing the program or shared object in the case of @var{$ORIGIN}
2268 and either @samp{lib} - for 32-bit binaries - or @samp{lib64} - for
2269 64-bit binaries - in the case of @var{$LIB}.
2271 The alternative form of these tokens - @var{$@{ORIGIN@}} and
2272 @var{$@{LIB@}} can also be used. The token @var{$PLATFORM} is not
2275 This option should be used with caution as it overrides the search path
2276 that may have been hard compiled into a shared library. In such a case it
2277 is possible to use unintentionally a different search path than the
2278 runtime linker would do.
2280 The linker uses the following search paths to locate required shared
2285 Any directories specified by @option{-rpath-link} options.
2287 Any directories specified by @option{-rpath} options. The difference
2288 between @option{-rpath} and @option{-rpath-link} is that directories
2289 specified by @option{-rpath} options are included in the executable and
2290 used at runtime, whereas the @option{-rpath-link} option is only effective
2291 at link time. Searching @option{-rpath} in this way is only supported
2292 by native linkers and cross linkers which have been configured with
2293 the @option{--with-sysroot} option.
2295 On an ELF system, for native linkers, if the @option{-rpath} and
2296 @option{-rpath-link} options were not used, search the contents of the
2297 environment variable @code{LD_RUN_PATH}.
2299 On SunOS, if the @option{-rpath} option was not used, search any
2300 directories specified using @option{-L} options.
2302 For a native linker, search the contents of the environment
2303 variable @code{LD_LIBRARY_PATH}.
2305 For a native ELF linker, the directories in @code{DT_RUNPATH} or
2306 @code{DT_RPATH} of a shared library are searched for shared
2307 libraries needed by it. The @code{DT_RPATH} entries are ignored if
2308 @code{DT_RUNPATH} entries exist.
2310 The default directories, normally @file{/lib} and @file{/usr/lib}.
2312 For a linker for a Linux system, if the file @file{/etc/ld.so.conf}
2313 exists, the list of directories found in that file. Note: the path
2314 to this file is prefixed with the @code{sysroot} value, if that is
2315 defined, and then any @code{prefix} string if the linker was
2316 configured with the @command{--prefix=<path>} option.
2318 For a native linker on a FreeBSD system, any directories specified by
2319 the @code{_PATH_ELF_HINTS} macro defined in the @file{elf-hints.h}
2322 Any directories specifed by a @code{SEARCH_DIR} command in the
2323 linker script being used.
2326 If the required shared library is not found, the linker will issue a
2327 warning and continue with the link.
2334 @cindex shared libraries
2335 Create a shared library. This is currently only supported on ELF, XCOFF
2336 and SunOS platforms. On SunOS, the linker will automatically create a
2337 shared library if the @option{-e} option is not used and there are
2338 undefined symbols in the link.
2340 @kindex --sort-common
2342 @itemx --sort-common=ascending
2343 @itemx --sort-common=descending
2344 This option tells @command{ld} to sort the common symbols by alignment in
2345 ascending or descending order when it places them in the appropriate output
2346 sections. The symbol alignments considered are sixteen-byte or larger,
2347 eight-byte, four-byte, two-byte, and one-byte. This is to prevent gaps
2348 between symbols due to alignment constraints. If no sorting order is
2349 specified, then descending order is assumed.
2351 @kindex --sort-section=name
2352 @item --sort-section=name
2353 This option will apply @code{SORT_BY_NAME} to all wildcard section
2354 patterns in the linker script.
2356 @kindex --sort-section=alignment
2357 @item --sort-section=alignment
2358 This option will apply @code{SORT_BY_ALIGNMENT} to all wildcard section
2359 patterns in the linker script.
2361 @kindex --spare-dynamic-tags
2362 @item --spare-dynamic-tags=@var{count}
2363 This option specifies the number of empty slots to leave in the
2364 .dynamic section of ELF shared objects. Empty slots may be needed by
2365 post processing tools, such as the prelinker. The default is 5.
2367 @kindex --split-by-file
2368 @item --split-by-file[=@var{size}]
2369 Similar to @option{--split-by-reloc} but creates a new output section for
2370 each input file when @var{size} is reached. @var{size} defaults to a
2371 size of 1 if not given.
2373 @kindex --split-by-reloc
2374 @item --split-by-reloc[=@var{count}]
2375 Tries to creates extra sections in the output file so that no single
2376 output section in the file contains more than @var{count} relocations.
2377 This is useful when generating huge relocatable files for downloading into
2378 certain real time kernels with the COFF object file format; since COFF
2379 cannot represent more than 65535 relocations in a single section. Note
2380 that this will fail to work with object file formats which do not
2381 support arbitrary sections. The linker will not split up individual
2382 input sections for redistribution, so if a single input section contains
2383 more than @var{count} relocations one output section will contain that
2384 many relocations. @var{count} defaults to a value of 32768.
2388 Compute and display statistics about the operation of the linker, such
2389 as execution time and memory usage.
2391 @kindex --sysroot=@var{directory}
2392 @item --sysroot=@var{directory}
2393 Use @var{directory} as the location of the sysroot, overriding the
2394 configure-time default. This option is only supported by linkers
2395 that were configured using @option{--with-sysroot}.
2399 This is used by COFF/PE based targets to create a task-linked object
2400 file where all of the global symbols have been converted to statics.
2402 @kindex --traditional-format
2403 @cindex traditional format
2404 @item --traditional-format
2405 For some targets, the output of @command{ld} is different in some ways from
2406 the output of some existing linker. This switch requests @command{ld} to
2407 use the traditional format instead.
2410 For example, on SunOS, @command{ld} combines duplicate entries in the
2411 symbol string table. This can reduce the size of an output file with
2412 full debugging information by over 30 percent. Unfortunately, the SunOS
2413 @code{dbx} program can not read the resulting program (@code{gdb} has no
2414 trouble). The @samp{--traditional-format} switch tells @command{ld} to not
2415 combine duplicate entries.
2417 @kindex --section-start=@var{sectionname}=@var{org}
2418 @item --section-start=@var{sectionname}=@var{org}
2419 Locate a section in the output file at the absolute
2420 address given by @var{org}. You may use this option as many
2421 times as necessary to locate multiple sections in the command
2423 @var{org} must be a single hexadecimal integer;
2424 for compatibility with other linkers, you may omit the leading
2425 @samp{0x} usually associated with hexadecimal values. @emph{Note:} there
2426 should be no white space between @var{sectionname}, the equals
2427 sign (``@key{=}''), and @var{org}.
2429 @kindex -Tbss=@var{org}
2430 @kindex -Tdata=@var{org}
2431 @kindex -Ttext=@var{org}
2432 @cindex segment origins, cmd line
2433 @item -Tbss=@var{org}
2434 @itemx -Tdata=@var{org}
2435 @itemx -Ttext=@var{org}
2436 Same as @option{--section-start}, with @code{.bss}, @code{.data} or
2437 @code{.text} as the @var{sectionname}.
2439 @kindex -Ttext-segment=@var{org}
2440 @item -Ttext-segment=@var{org}
2441 @cindex text segment origin, cmd line
2442 When creating an ELF executable, it will set the address of the first
2443 byte of the text segment.
2445 @kindex -Trodata-segment=@var{org}
2446 @item -Trodata-segment=@var{org}
2447 @cindex rodata segment origin, cmd line
2448 When creating an ELF executable or shared object for a target where
2449 the read-only data is in its own segment separate from the executable
2450 text, it will set the address of the first byte of the read-only data segment.
2452 @kindex -Tldata-segment=@var{org}
2453 @item -Tldata-segment=@var{org}
2454 @cindex ldata segment origin, cmd line
2455 When creating an ELF executable or shared object for x86-64 medium memory
2456 model, it will set the address of the first byte of the ldata segment.
2458 @kindex --unresolved-symbols
2459 @item --unresolved-symbols=@var{method}
2460 Determine how to handle unresolved symbols. There are four possible
2461 values for @samp{method}:
2465 Do not report any unresolved symbols.
2468 Report all unresolved symbols. This is the default.
2470 @item ignore-in-object-files
2471 Report unresolved symbols that are contained in shared libraries, but
2472 ignore them if they come from regular object files.
2474 @item ignore-in-shared-libs
2475 Report unresolved symbols that come from regular object files, but
2476 ignore them if they come from shared libraries. This can be useful
2477 when creating a dynamic binary and it is known that all the shared
2478 libraries that it should be referencing are included on the linker's
2482 The behaviour for shared libraries on their own can also be controlled
2483 by the @option{--[no-]allow-shlib-undefined} option.
2485 Normally the linker will generate an error message for each reported
2486 unresolved symbol but the option @option{--warn-unresolved-symbols}
2487 can change this to a warning.
2489 @kindex --verbose[=@var{NUMBER}]
2490 @cindex verbose[=@var{NUMBER}]
2492 @itemx --verbose[=@var{NUMBER}]
2493 Display the version number for @command{ld} and list the linker emulations
2494 supported. Display which input files can and cannot be opened. Display
2495 the linker script being used by the linker. If the optional @var{NUMBER}
2496 argument > 1, plugin symbol status will also be displayed.
2498 @kindex --version-script=@var{version-scriptfile}
2499 @cindex version script, symbol versions
2500 @item --version-script=@var{version-scriptfile}
2501 Specify the name of a version script to the linker. This is typically
2502 used when creating shared libraries to specify additional information
2503 about the version hierarchy for the library being created. This option
2504 is only fully supported on ELF platforms which support shared libraries;
2505 see @ref{VERSION}. It is partially supported on PE platforms, which can
2506 use version scripts to filter symbol visibility in auto-export mode: any
2507 symbols marked @samp{local} in the version script will not be exported.
2510 @kindex --warn-common
2511 @cindex warnings, on combining symbols
2512 @cindex combining symbols, warnings on
2514 Warn when a common symbol is combined with another common symbol or with
2515 a symbol definition. Unix linkers allow this somewhat sloppy practice,
2516 but linkers on some other operating systems do not. This option allows
2517 you to find potential problems from combining global symbols.
2518 Unfortunately, some C libraries use this practice, so you may get some
2519 warnings about symbols in the libraries as well as in your programs.
2521 There are three kinds of global symbols, illustrated here by C examples:
2525 A definition, which goes in the initialized data section of the output
2529 An undefined reference, which does not allocate space.
2530 There must be either a definition or a common symbol for the
2534 A common symbol. If there are only (one or more) common symbols for a
2535 variable, it goes in the uninitialized data area of the output file.
2536 The linker merges multiple common symbols for the same variable into a
2537 single symbol. If they are of different sizes, it picks the largest
2538 size. The linker turns a common symbol into a declaration, if there is
2539 a definition of the same variable.
2542 The @samp{--warn-common} option can produce five kinds of warnings.
2543 Each warning consists of a pair of lines: the first describes the symbol
2544 just encountered, and the second describes the previous symbol
2545 encountered with the same name. One or both of the two symbols will be
2550 Turning a common symbol into a reference, because there is already a
2551 definition for the symbol.
2553 @var{file}(@var{section}): warning: common of `@var{symbol}'
2554 overridden by definition
2555 @var{file}(@var{section}): warning: defined here
2559 Turning a common symbol into a reference, because a later definition for
2560 the symbol is encountered. This is the same as the previous case,
2561 except that the symbols are encountered in a different order.
2563 @var{file}(@var{section}): warning: definition of `@var{symbol}'
2565 @var{file}(@var{section}): warning: common is here
2569 Merging a common symbol with a previous same-sized common symbol.
2571 @var{file}(@var{section}): warning: multiple common
2573 @var{file}(@var{section}): warning: previous common is here
2577 Merging a common symbol with a previous larger common symbol.
2579 @var{file}(@var{section}): warning: common of `@var{symbol}'
2580 overridden by larger common
2581 @var{file}(@var{section}): warning: larger common is here
2585 Merging a common symbol with a previous smaller common symbol. This is
2586 the same as the previous case, except that the symbols are
2587 encountered in a different order.
2589 @var{file}(@var{section}): warning: common of `@var{symbol}'
2590 overriding smaller common
2591 @var{file}(@var{section}): warning: smaller common is here
2595 @kindex --warn-constructors
2596 @item --warn-constructors
2597 Warn if any global constructors are used. This is only useful for a few
2598 object file formats. For formats like COFF or ELF, the linker can not
2599 detect the use of global constructors.
2601 @kindex --warn-multiple-gp
2602 @item --warn-multiple-gp
2603 Warn if multiple global pointer values are required in the output file.
2604 This is only meaningful for certain processors, such as the Alpha.
2605 Specifically, some processors put large-valued constants in a special
2606 section. A special register (the global pointer) points into the middle
2607 of this section, so that constants can be loaded efficiently via a
2608 base-register relative addressing mode. Since the offset in
2609 base-register relative mode is fixed and relatively small (e.g., 16
2610 bits), this limits the maximum size of the constant pool. Thus, in
2611 large programs, it is often necessary to use multiple global pointer
2612 values in order to be able to address all possible constants. This
2613 option causes a warning to be issued whenever this case occurs.
2616 @cindex warnings, on undefined symbols
2617 @cindex undefined symbols, warnings on
2619 Only warn once for each undefined symbol, rather than once per module
2622 @kindex --warn-section-align
2623 @cindex warnings, on section alignment
2624 @cindex section alignment, warnings on
2625 @item --warn-section-align
2626 Warn if the address of an output section is changed because of
2627 alignment. Typically, the alignment will be set by an input section.
2628 The address will only be changed if it not explicitly specified; that
2629 is, if the @code{SECTIONS} command does not specify a start address for
2630 the section (@pxref{SECTIONS}).
2632 @kindex --warn-textrel
2633 @item --warn-textrel
2634 Warn if the linker adds DT_TEXTREL to a position-independent executable
2637 @kindex --warn-alternate-em
2638 @item --warn-alternate-em
2639 Warn if an object has alternate ELF machine code.
2641 @kindex --warn-unresolved-symbols
2642 @item --warn-unresolved-symbols
2643 If the linker is going to report an unresolved symbol (see the option
2644 @option{--unresolved-symbols}) it will normally generate an error.
2645 This option makes it generate a warning instead.
2647 @kindex --error-unresolved-symbols
2648 @item --error-unresolved-symbols
2649 This restores the linker's default behaviour of generating errors when
2650 it is reporting unresolved symbols.
2652 @kindex --whole-archive
2653 @cindex including an entire archive
2654 @item --whole-archive
2655 For each archive mentioned on the command line after the
2656 @option{--whole-archive} option, include every object file in the archive
2657 in the link, rather than searching the archive for the required object
2658 files. This is normally used to turn an archive file into a shared
2659 library, forcing every object to be included in the resulting shared
2660 library. This option may be used more than once.
2662 Two notes when using this option from gcc: First, gcc doesn't know
2663 about this option, so you have to use @option{-Wl,-whole-archive}.
2664 Second, don't forget to use @option{-Wl,-no-whole-archive} after your
2665 list of archives, because gcc will add its own list of archives to
2666 your link and you may not want this flag to affect those as well.
2668 @kindex --wrap=@var{symbol}
2669 @item --wrap=@var{symbol}
2670 Use a wrapper function for @var{symbol}. Any undefined reference to
2671 @var{symbol} will be resolved to @code{__wrap_@var{symbol}}. Any
2672 undefined reference to @code{__real_@var{symbol}} will be resolved to
2675 This can be used to provide a wrapper for a system function. The
2676 wrapper function should be called @code{__wrap_@var{symbol}}. If it
2677 wishes to call the system function, it should call
2678 @code{__real_@var{symbol}}.
2680 Here is a trivial example:
2684 __wrap_malloc (size_t c)
2686 printf ("malloc called with %zu\n", c);
2687 return __real_malloc (c);
2691 If you link other code with this file using @option{--wrap malloc}, then
2692 all calls to @code{malloc} will call the function @code{__wrap_malloc}
2693 instead. The call to @code{__real_malloc} in @code{__wrap_malloc} will
2694 call the real @code{malloc} function.
2696 You may wish to provide a @code{__real_malloc} function as well, so that
2697 links without the @option{--wrap} option will succeed. If you do this,
2698 you should not put the definition of @code{__real_malloc} in the same
2699 file as @code{__wrap_malloc}; if you do, the assembler may resolve the
2700 call before the linker has a chance to wrap it to @code{malloc}.
2702 Only undefined references are replaced by the linker. So, translation unit
2703 internal references to @var{symbol} are not resolved to
2704 @code{__wrap_@var{symbol}}. In the next example, the call to @code{f} in
2705 @code{g} is not resolved to @code{__wrap_f}.
2721 @kindex --eh-frame-hdr
2722 @kindex --no-eh-frame-hdr
2723 @item --eh-frame-hdr
2724 @itemx --no-eh-frame-hdr
2725 Request (@option{--eh-frame-hdr}) or suppress
2726 (@option{--no-eh-frame-hdr}) the creation of @code{.eh_frame_hdr}
2727 section and ELF @code{PT_GNU_EH_FRAME} segment header.
2729 @kindex --ld-generated-unwind-info
2730 @item --no-ld-generated-unwind-info
2731 Request creation of @code{.eh_frame} unwind info for linker
2732 generated code sections like PLT. This option is on by default
2733 if linker generated unwind info is supported.
2735 @kindex --enable-new-dtags
2736 @kindex --disable-new-dtags
2737 @item --enable-new-dtags
2738 @itemx --disable-new-dtags
2739 This linker can create the new dynamic tags in ELF. But the older ELF
2740 systems may not understand them. If you specify
2741 @option{--enable-new-dtags}, the new dynamic tags will be created as needed
2742 and older dynamic tags will be omitted.
2743 If you specify @option{--disable-new-dtags}, no new dynamic tags will be
2744 created. By default, the new dynamic tags are not created. Note that
2745 those options are only available for ELF systems.
2747 @kindex --hash-size=@var{number}
2748 @item --hash-size=@var{number}
2749 Set the default size of the linker's hash tables to a prime number
2750 close to @var{number}. Increasing this value can reduce the length of
2751 time it takes the linker to perform its tasks, at the expense of
2752 increasing the linker's memory requirements. Similarly reducing this
2753 value can reduce the memory requirements at the expense of speed.
2755 @kindex --hash-style=@var{style}
2756 @item --hash-style=@var{style}
2757 Set the type of linker's hash table(s). @var{style} can be either
2758 @code{sysv} for classic ELF @code{.hash} section, @code{gnu} for
2759 new style GNU @code{.gnu.hash} section or @code{both} for both
2760 the classic ELF @code{.hash} and new style GNU @code{.gnu.hash}
2761 hash tables. The default depends upon how the linker was configured,
2762 but for most Linux based systems it will be @code{both}.
2764 @kindex --compress-debug-sections=none
2765 @kindex --compress-debug-sections=zlib
2766 @kindex --compress-debug-sections=zlib-gnu
2767 @kindex --compress-debug-sections=zlib-gabi
2768 @item --compress-debug-sections=none
2769 @itemx --compress-debug-sections=zlib
2770 @itemx --compress-debug-sections=zlib-gnu
2771 @itemx --compress-debug-sections=zlib-gabi
2772 On ELF platforms, these options control how DWARF debug sections are
2773 compressed using zlib.
2775 @option{--compress-debug-sections=none} doesn't compress DWARF debug
2776 sections. @option{--compress-debug-sections=zlib-gnu} compresses
2777 DWARF debug sections and renames them to begin with @samp{.zdebug}
2778 instead of @samp{.debug}. @option{--compress-debug-sections=zlib-gabi}
2779 also compresses DWARF debug sections, but rather than renaming them it
2780 sets the SHF_COMPRESSED flag in the sections' headers.
2782 The @option{--compress-debug-sections=zlib} option is an alias for
2783 @option{--compress-debug-sections=zlib-gabi}.
2785 Note that this option overrides any compression in input debug
2786 sections, so if a binary is linked with @option{--compress-debug-sections=none}
2787 for example, then any compressed debug sections in input files will be
2788 uncompressed before they are copied into the output binary.
2790 The default compression behaviour varies depending upon the target
2791 involved and the configure options used to build the toolchain. The
2792 default can be determined by examining the output from the linker's
2793 @option{--help} option.
2795 @kindex --reduce-memory-overheads
2796 @item --reduce-memory-overheads
2797 This option reduces memory requirements at ld runtime, at the expense of
2798 linking speed. This was introduced to select the old O(n^2) algorithm
2799 for link map file generation, rather than the new O(n) algorithm which uses
2800 about 40% more memory for symbol storage.
2802 Another effect of the switch is to set the default hash table size to
2803 1021, which again saves memory at the cost of lengthening the linker's
2804 run time. This is not done however if the @option{--hash-size} switch
2807 The @option{--reduce-memory-overheads} switch may be also be used to
2808 enable other tradeoffs in future versions of the linker.
2811 @kindex --build-id=@var{style}
2813 @itemx --build-id=@var{style}
2814 Request the creation of a @code{.note.gnu.build-id} ELF note section
2815 or a @code{.buildid} COFF section. The contents of the note are
2816 unique bits identifying this linked file. @var{style} can be
2817 @code{uuid} to use 128 random bits, @code{sha1} to use a 160-bit
2818 @sc{SHA1} hash on the normative parts of the output contents,
2819 @code{md5} to use a 128-bit @sc{MD5} hash on the normative parts of
2820 the output contents, or @code{0x@var{hexstring}} to use a chosen bit
2821 string specified as an even number of hexadecimal digits (@code{-} and
2822 @code{:} characters between digit pairs are ignored). If @var{style}
2823 is omitted, @code{sha1} is used.
2825 The @code{md5} and @code{sha1} styles produces an identifier
2826 that is always the same in an identical output file, but will be
2827 unique among all nonidentical output files. It is not intended
2828 to be compared as a checksum for the file's contents. A linked
2829 file may be changed later by other tools, but the build ID bit
2830 string identifying the original linked file does not change.
2832 Passing @code{none} for @var{style} disables the setting from any
2833 @code{--build-id} options earlier on the command line.
2838 @subsection Options Specific to i386 PE Targets
2840 @c man begin OPTIONS
2842 The i386 PE linker supports the @option{-shared} option, which causes
2843 the output to be a dynamically linked library (DLL) instead of a
2844 normal executable. You should name the output @code{*.dll} when you
2845 use this option. In addition, the linker fully supports the standard
2846 @code{*.def} files, which may be specified on the linker command line
2847 like an object file (in fact, it should precede archives it exports
2848 symbols from, to ensure that they get linked in, just like a normal
2851 In addition to the options common to all targets, the i386 PE linker
2852 support additional command-line options that are specific to the i386
2853 PE target. Options that take values may be separated from their
2854 values by either a space or an equals sign.
2858 @kindex --add-stdcall-alias
2859 @item --add-stdcall-alias
2860 If given, symbols with a stdcall suffix (@@@var{nn}) will be exported
2861 as-is and also with the suffix stripped.
2862 [This option is specific to the i386 PE targeted port of the linker]
2865 @item --base-file @var{file}
2866 Use @var{file} as the name of a file in which to save the base
2867 addresses of all the relocations needed for generating DLLs with
2869 [This is an i386 PE specific option]
2873 Create a DLL instead of a regular executable. You may also use
2874 @option{-shared} or specify a @code{LIBRARY} in a given @code{.def}
2876 [This option is specific to the i386 PE targeted port of the linker]
2878 @kindex --enable-long-section-names
2879 @kindex --disable-long-section-names
2880 @item --enable-long-section-names
2881 @itemx --disable-long-section-names
2882 The PE variants of the COFF object format add an extension that permits
2883 the use of section names longer than eight characters, the normal limit
2884 for COFF. By default, these names are only allowed in object files, as
2885 fully-linked executable images do not carry the COFF string table required
2886 to support the longer names. As a GNU extension, it is possible to
2887 allow their use in executable images as well, or to (probably pointlessly!)
2888 disallow it in object files, by using these two options. Executable images
2889 generated with these long section names are slightly non-standard, carrying
2890 as they do a string table, and may generate confusing output when examined
2891 with non-GNU PE-aware tools, such as file viewers and dumpers. However,
2892 GDB relies on the use of PE long section names to find Dwarf-2 debug
2893 information sections in an executable image at runtime, and so if neither
2894 option is specified on the command-line, @command{ld} will enable long
2895 section names, overriding the default and technically correct behaviour,
2896 when it finds the presence of debug information while linking an executable
2897 image and not stripping symbols.
2898 [This option is valid for all PE targeted ports of the linker]
2900 @kindex --enable-stdcall-fixup
2901 @kindex --disable-stdcall-fixup
2902 @item --enable-stdcall-fixup
2903 @itemx --disable-stdcall-fixup
2904 If the link finds a symbol that it cannot resolve, it will attempt to
2905 do ``fuzzy linking'' by looking for another defined symbol that differs
2906 only in the format of the symbol name (cdecl vs stdcall) and will
2907 resolve that symbol by linking to the match. For example, the
2908 undefined symbol @code{_foo} might be linked to the function
2909 @code{_foo@@12}, or the undefined symbol @code{_bar@@16} might be linked
2910 to the function @code{_bar}. When the linker does this, it prints a
2911 warning, since it normally should have failed to link, but sometimes
2912 import libraries generated from third-party dlls may need this feature
2913 to be usable. If you specify @option{--enable-stdcall-fixup}, this
2914 feature is fully enabled and warnings are not printed. If you specify
2915 @option{--disable-stdcall-fixup}, this feature is disabled and such
2916 mismatches are considered to be errors.
2917 [This option is specific to the i386 PE targeted port of the linker]
2919 @kindex --leading-underscore
2920 @kindex --no-leading-underscore
2921 @item --leading-underscore
2922 @itemx --no-leading-underscore
2923 For most targets default symbol-prefix is an underscore and is defined
2924 in target's description. By this option it is possible to
2925 disable/enable the default underscore symbol-prefix.
2927 @cindex DLLs, creating
2928 @kindex --export-all-symbols
2929 @item --export-all-symbols
2930 If given, all global symbols in the objects used to build a DLL will
2931 be exported by the DLL. Note that this is the default if there
2932 otherwise wouldn't be any exported symbols. When symbols are
2933 explicitly exported via DEF files or implicitly exported via function
2934 attributes, the default is to not export anything else unless this
2935 option is given. Note that the symbols @code{DllMain@@12},
2936 @code{DllEntryPoint@@0}, @code{DllMainCRTStartup@@12}, and
2937 @code{impure_ptr} will not be automatically
2938 exported. Also, symbols imported from other DLLs will not be
2939 re-exported, nor will symbols specifying the DLL's internal layout
2940 such as those beginning with @code{_head_} or ending with
2941 @code{_iname}. In addition, no symbols from @code{libgcc},
2942 @code{libstd++}, @code{libmingw32}, or @code{crtX.o} will be exported.
2943 Symbols whose names begin with @code{__rtti_} or @code{__builtin_} will
2944 not be exported, to help with C++ DLLs. Finally, there is an
2945 extensive list of cygwin-private symbols that are not exported
2946 (obviously, this applies on when building DLLs for cygwin targets).
2947 These cygwin-excludes are: @code{_cygwin_dll_entry@@12},
2948 @code{_cygwin_crt0_common@@8}, @code{_cygwin_noncygwin_dll_entry@@12},
2949 @code{_fmode}, @code{_impure_ptr}, @code{cygwin_attach_dll},
2950 @code{cygwin_premain0}, @code{cygwin_premain1}, @code{cygwin_premain2},
2951 @code{cygwin_premain3}, and @code{environ}.
2952 [This option is specific to the i386 PE targeted port of the linker]
2954 @kindex --exclude-symbols
2955 @item --exclude-symbols @var{symbol},@var{symbol},...
2956 Specifies a list of symbols which should not be automatically
2957 exported. The symbol names may be delimited by commas or colons.
2958 [This option is specific to the i386 PE targeted port of the linker]
2960 @kindex --exclude-all-symbols
2961 @item --exclude-all-symbols
2962 Specifies no symbols should be automatically exported.
2963 [This option is specific to the i386 PE targeted port of the linker]
2965 @kindex --file-alignment
2966 @item --file-alignment
2967 Specify the file alignment. Sections in the file will always begin at
2968 file offsets which are multiples of this number. This defaults to
2970 [This option is specific to the i386 PE targeted port of the linker]
2974 @item --heap @var{reserve}
2975 @itemx --heap @var{reserve},@var{commit}
2976 Specify the number of bytes of memory to reserve (and optionally commit)
2977 to be used as heap for this program. The default is 1MB reserved, 4K
2979 [This option is specific to the i386 PE targeted port of the linker]
2982 @kindex --image-base
2983 @item --image-base @var{value}
2984 Use @var{value} as the base address of your program or dll. This is
2985 the lowest memory location that will be used when your program or dll
2986 is loaded. To reduce the need to relocate and improve performance of
2987 your dlls, each should have a unique base address and not overlap any
2988 other dlls. The default is 0x400000 for executables, and 0x10000000
2990 [This option is specific to the i386 PE targeted port of the linker]
2994 If given, the stdcall suffixes (@@@var{nn}) will be stripped from
2995 symbols before they are exported.
2996 [This option is specific to the i386 PE targeted port of the linker]
2998 @kindex --large-address-aware
2999 @item --large-address-aware
3000 If given, the appropriate bit in the ``Characteristics'' field of the COFF
3001 header is set to indicate that this executable supports virtual addresses
3002 greater than 2 gigabytes. This should be used in conjunction with the /3GB
3003 or /USERVA=@var{value} megabytes switch in the ``[operating systems]''
3004 section of the BOOT.INI. Otherwise, this bit has no effect.
3005 [This option is specific to PE targeted ports of the linker]
3007 @kindex --disable-large-address-aware
3008 @item --disable-large-address-aware
3009 Reverts the effect of a previous @samp{--large-address-aware} option.
3010 This is useful if @samp{--large-address-aware} is always set by the compiler
3011 driver (e.g. Cygwin gcc) and the executable does not support virtual
3012 addresses greater than 2 gigabytes.
3013 [This option is specific to PE targeted ports of the linker]
3015 @kindex --major-image-version
3016 @item --major-image-version @var{value}
3017 Sets the major number of the ``image version''. Defaults to 1.
3018 [This option is specific to the i386 PE targeted port of the linker]
3020 @kindex --major-os-version
3021 @item --major-os-version @var{value}
3022 Sets the major number of the ``os version''. Defaults to 4.
3023 [This option is specific to the i386 PE targeted port of the linker]
3025 @kindex --major-subsystem-version
3026 @item --major-subsystem-version @var{value}
3027 Sets the major number of the ``subsystem version''. Defaults to 4.
3028 [This option is specific to the i386 PE targeted port of the linker]
3030 @kindex --minor-image-version
3031 @item --minor-image-version @var{value}
3032 Sets the minor number of the ``image version''. Defaults to 0.
3033 [This option is specific to the i386 PE targeted port of the linker]
3035 @kindex --minor-os-version
3036 @item --minor-os-version @var{value}
3037 Sets the minor number of the ``os version''. Defaults to 0.
3038 [This option is specific to the i386 PE targeted port of the linker]
3040 @kindex --minor-subsystem-version
3041 @item --minor-subsystem-version @var{value}
3042 Sets the minor number of the ``subsystem version''. Defaults to 0.
3043 [This option is specific to the i386 PE targeted port of the linker]
3045 @cindex DEF files, creating
3046 @cindex DLLs, creating
3047 @kindex --output-def
3048 @item --output-def @var{file}
3049 The linker will create the file @var{file} which will contain a DEF
3050 file corresponding to the DLL the linker is generating. This DEF file
3051 (which should be called @code{*.def}) may be used to create an import
3052 library with @code{dlltool} or may be used as a reference to
3053 automatically or implicitly exported symbols.
3054 [This option is specific to the i386 PE targeted port of the linker]
3056 @cindex DLLs, creating
3057 @kindex --enable-auto-image-base
3058 @item --enable-auto-image-base
3059 @itemx --enable-auto-image-base=@var{value}
3060 Automatically choose the image base for DLLs, optionally starting with base
3061 @var{value}, unless one is specified using the @code{--image-base} argument.
3062 By using a hash generated from the dllname to create unique image bases
3063 for each DLL, in-memory collisions and relocations which can delay program
3064 execution are avoided.
3065 [This option is specific to the i386 PE targeted port of the linker]
3067 @kindex --disable-auto-image-base
3068 @item --disable-auto-image-base
3069 Do not automatically generate a unique image base. If there is no
3070 user-specified image base (@code{--image-base}) then use the platform
3072 [This option is specific to the i386 PE targeted port of the linker]
3074 @cindex DLLs, linking to
3075 @kindex --dll-search-prefix
3076 @item --dll-search-prefix @var{string}
3077 When linking dynamically to a dll without an import library,
3078 search for @code{<string><basename>.dll} in preference to
3079 @code{lib<basename>.dll}. This behaviour allows easy distinction
3080 between DLLs built for the various "subplatforms": native, cygwin,
3081 uwin, pw, etc. For instance, cygwin DLLs typically use
3082 @code{--dll-search-prefix=cyg}.
3083 [This option is specific to the i386 PE targeted port of the linker]
3085 @kindex --enable-auto-import
3086 @item --enable-auto-import
3087 Do sophisticated linking of @code{_symbol} to @code{__imp__symbol} for
3088 DATA imports from DLLs, thus making it possible to bypass the dllimport
3089 mechanism on the user side and to reference unmangled symbol names.
3090 [This option is specific to the i386 PE targeted port of the linker]
3092 The following remarks pertain to the original implementation of the
3093 feature and are obsolete nowadays for Cygwin and MinGW targets.
3095 Note: Use of the 'auto-import' extension will cause the text section
3096 of the image file to be made writable. This does not conform to the
3097 PE-COFF format specification published by Microsoft.
3099 Note - use of the 'auto-import' extension will also cause read only
3100 data which would normally be placed into the .rdata section to be
3101 placed into the .data section instead. This is in order to work
3102 around a problem with consts that is described here:
3103 http://www.cygwin.com/ml/cygwin/2004-09/msg01101.html
3105 Using 'auto-import' generally will 'just work' -- but sometimes you may
3108 "variable '<var>' can't be auto-imported. Please read the
3109 documentation for ld's @code{--enable-auto-import} for details."
3111 This message occurs when some (sub)expression accesses an address
3112 ultimately given by the sum of two constants (Win32 import tables only
3113 allow one). Instances where this may occur include accesses to member
3114 fields of struct variables imported from a DLL, as well as using a
3115 constant index into an array variable imported from a DLL. Any
3116 multiword variable (arrays, structs, long long, etc) may trigger
3117 this error condition. However, regardless of the exact data type
3118 of the offending exported variable, ld will always detect it, issue
3119 the warning, and exit.
3121 There are several ways to address this difficulty, regardless of the
3122 data type of the exported variable:
3124 One way is to use --enable-runtime-pseudo-reloc switch. This leaves the task
3125 of adjusting references in your client code for runtime environment, so
3126 this method works only when runtime environment supports this feature.
3128 A second solution is to force one of the 'constants' to be a variable --
3129 that is, unknown and un-optimizable at compile time. For arrays,
3130 there are two possibilities: a) make the indexee (the array's address)
3131 a variable, or b) make the 'constant' index a variable. Thus:
3134 extern type extern_array[];
3136 @{ volatile type *t=extern_array; t[1] @}
3142 extern type extern_array[];
3144 @{ volatile int t=1; extern_array[t] @}
3147 For structs (and most other multiword data types) the only option
3148 is to make the struct itself (or the long long, or the ...) variable:
3151 extern struct s extern_struct;
3152 extern_struct.field -->
3153 @{ volatile struct s *t=&extern_struct; t->field @}
3159 extern long long extern_ll;
3161 @{ volatile long long * local_ll=&extern_ll; *local_ll @}
3164 A third method of dealing with this difficulty is to abandon
3165 'auto-import' for the offending symbol and mark it with
3166 @code{__declspec(dllimport)}. However, in practice that
3167 requires using compile-time #defines to indicate whether you are
3168 building a DLL, building client code that will link to the DLL, or
3169 merely building/linking to a static library. In making the choice
3170 between the various methods of resolving the 'direct address with
3171 constant offset' problem, you should consider typical real-world usage:
3179 void main(int argc, char **argv)@{
3180 printf("%d\n",arr[1]);
3190 void main(int argc, char **argv)@{
3191 /* This workaround is for win32 and cygwin; do not "optimize" */
3192 volatile int *parr = arr;
3193 printf("%d\n",parr[1]);
3200 /* Note: auto-export is assumed (no __declspec(dllexport)) */
3201 #if (defined(_WIN32) || defined(__CYGWIN__)) && \
3202 !(defined(FOO_BUILD_DLL) || defined(FOO_STATIC))
3203 #define FOO_IMPORT __declspec(dllimport)
3207 extern FOO_IMPORT int arr[];
3210 void main(int argc, char **argv)@{
3211 printf("%d\n",arr[1]);
3215 A fourth way to avoid this problem is to re-code your
3216 library to use a functional interface rather than a data interface
3217 for the offending variables (e.g. set_foo() and get_foo() accessor
3220 @kindex --disable-auto-import
3221 @item --disable-auto-import
3222 Do not attempt to do sophisticated linking of @code{_symbol} to
3223 @code{__imp__symbol} for DATA imports from DLLs.
3224 [This option is specific to the i386 PE targeted port of the linker]
3226 @kindex --enable-runtime-pseudo-reloc
3227 @item --enable-runtime-pseudo-reloc
3228 If your code contains expressions described in --enable-auto-import section,
3229 that is, DATA imports from DLL with non-zero offset, this switch will create
3230 a vector of 'runtime pseudo relocations' which can be used by runtime
3231 environment to adjust references to such data in your client code.
3232 [This option is specific to the i386 PE targeted port of the linker]
3234 @kindex --disable-runtime-pseudo-reloc
3235 @item --disable-runtime-pseudo-reloc
3236 Do not create pseudo relocations for non-zero offset DATA imports from DLLs.
3237 [This option is specific to the i386 PE targeted port of the linker]
3239 @kindex --enable-extra-pe-debug
3240 @item --enable-extra-pe-debug
3241 Show additional debug info related to auto-import symbol thunking.
3242 [This option is specific to the i386 PE targeted port of the linker]
3244 @kindex --section-alignment
3245 @item --section-alignment
3246 Sets the section alignment. Sections in memory will always begin at
3247 addresses which are a multiple of this number. Defaults to 0x1000.
3248 [This option is specific to the i386 PE targeted port of the linker]
3252 @item --stack @var{reserve}
3253 @itemx --stack @var{reserve},@var{commit}
3254 Specify the number of bytes of memory to reserve (and optionally commit)
3255 to be used as stack for this program. The default is 2MB reserved, 4K
3257 [This option is specific to the i386 PE targeted port of the linker]
3260 @item --subsystem @var{which}
3261 @itemx --subsystem @var{which}:@var{major}
3262 @itemx --subsystem @var{which}:@var{major}.@var{minor}
3263 Specifies the subsystem under which your program will execute. The
3264 legal values for @var{which} are @code{native}, @code{windows},
3265 @code{console}, @code{posix}, and @code{xbox}. You may optionally set
3266 the subsystem version also. Numeric values are also accepted for
3268 [This option is specific to the i386 PE targeted port of the linker]
3270 The following options set flags in the @code{DllCharacteristics} field
3271 of the PE file header:
3272 [These options are specific to PE targeted ports of the linker]
3274 @kindex --high-entropy-va
3275 @item --high-entropy-va
3276 @itemx --disable-high-entropy-va
3277 Image is compatible with 64-bit address space layout randomization
3278 (ASLR). This option is enabled by default for 64-bit PE images.
3280 This option also implies @option{--dynamicbase} and
3281 @option{--enable-reloc-section}.
3283 @kindex --dynamicbase
3285 @itemx --disable-dynamicbase
3286 The image base address may be relocated using address space layout
3287 randomization (ASLR). This feature was introduced with MS Windows
3288 Vista for i386 PE targets. This option is enabled by default but
3289 can be disabled via the @option{--disable-dynamicbase} option.
3290 This option also implies @option{--enable-reloc-section}.
3292 @kindex --forceinteg
3294 @itemx --disable-forceinteg
3295 Code integrity checks are enforced. This option is disabled by
3300 @item --disable-nxcompat
3301 The image is compatible with the Data Execution Prevention.
3302 This feature was introduced with MS Windows XP SP2 for i386 PE
3303 targets. The option is enabled by default.
3305 @kindex --no-isolation
3306 @item --no-isolation
3307 @itemx --disable-no-isolation
3308 Although the image understands isolation, do not isolate the image.
3309 This option is disabled by default.
3313 @itemx --disable-no-seh
3314 The image does not use SEH. No SE handler may be called from
3315 this image. This option is disabled by default.
3319 @itemx --disable-no-bind
3320 Do not bind this image. This option is disabled by default.
3324 @itemx --disable-wdmdriver
3325 The driver uses the MS Windows Driver Model. This option is disabled
3330 @itemx --disable-tsaware
3331 The image is Terminal Server aware. This option is disabled by
3334 @kindex --insert-timestamp
3335 @item --insert-timestamp
3336 @itemx --no-insert-timestamp
3337 Insert a real timestamp into the image. This is the default behaviour
3338 as it matches legacy code and it means that the image will work with
3339 other, proprietary tools. The problem with this default is that it
3340 will result in slightly different images being produced each time the
3341 same sources are linked. The option @option{--no-insert-timestamp}
3342 can be used to insert a zero value for the timestamp, this ensuring
3343 that binaries produced from identical sources will compare
3346 @kindex --enable-reloc-section
3347 @item --enable-reloc-section
3348 @itemx --disable-reloc-section
3349 Create the base relocation table, which is necessary if the image
3350 is loaded at a different image base than specified in the PE header.
3351 This option is enabled by default.
3357 @subsection Options specific to C6X uClinux targets
3359 @c man begin OPTIONS
3361 The C6X uClinux target uses a binary format called DSBT to support shared
3362 libraries. Each shared library in the system needs to have a unique index;
3363 all executables use an index of 0.
3368 @item --dsbt-size @var{size}
3369 This option sets the number of entries in the DSBT of the current executable
3370 or shared library to @var{size}. The default is to create a table with 64
3373 @kindex --dsbt-index
3374 @item --dsbt-index @var{index}
3375 This option sets the DSBT index of the current executable or shared library
3376 to @var{index}. The default is 0, which is appropriate for generating
3377 executables. If a shared library is generated with a DSBT index of 0, the
3378 @code{R_C6000_DSBT_INDEX} relocs are copied into the output file.
3380 @kindex --no-merge-exidx-entries
3381 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent
3382 exidx entries in frame unwind info.
3390 @subsection Options specific to C-SKY targets
3392 @c man begin OPTIONS
3396 @kindex --branch-stub on C-SKY
3398 This option enables linker branch relaxation by inserting branch stub
3399 sections when needed to extend the range of branches. This option is
3400 usually not required since C-SKY supports branch and call instructions that
3401 can access the full memory range and branch relaxation is normally handled by
3402 the compiler or assembler.
3404 @kindex --stub-group-size on C-SKY
3405 @item --stub-group-size=@var{N}
3406 This option allows finer control of linker branch stub creation.
3407 It sets the maximum size of a group of input sections that can
3408 be handled by one stub section. A negative value of @var{N} locates
3409 stub sections after their branches, while a positive value allows stub
3410 sections to appear either before or after the branches. Values of
3411 @samp{1} or @samp{-1} indicate that the
3412 linker should choose suitable defaults.
3420 @subsection Options specific to Motorola 68HC11 and 68HC12 targets
3422 @c man begin OPTIONS
3424 The 68HC11 and 68HC12 linkers support specific options to control the
3425 memory bank switching mapping and trampoline code generation.
3429 @kindex --no-trampoline
3430 @item --no-trampoline
3431 This option disables the generation of trampoline. By default a trampoline
3432 is generated for each far function which is called using a @code{jsr}
3433 instruction (this happens when a pointer to a far function is taken).
3435 @kindex --bank-window
3436 @item --bank-window @var{name}
3437 This option indicates to the linker the name of the memory region in
3438 the @samp{MEMORY} specification that describes the memory bank window.
3439 The definition of such region is then used by the linker to compute
3440 paging and addresses within the memory window.
3448 @subsection Options specific to Motorola 68K target
3450 @c man begin OPTIONS
3452 The following options are supported to control handling of GOT generation
3453 when linking for 68K targets.
3458 @item --got=@var{type}
3459 This option tells the linker which GOT generation scheme to use.
3460 @var{type} should be one of @samp{single}, @samp{negative},
3461 @samp{multigot} or @samp{target}. For more information refer to the
3462 Info entry for @file{ld}.
3470 @subsection Options specific to MIPS targets
3472 @c man begin OPTIONS
3474 The following options are supported to control microMIPS instruction
3475 generation and branch relocation checks for ISA mode transitions when
3476 linking for MIPS targets.
3484 These options control the choice of microMIPS instructions used in code
3485 generated by the linker, such as that in the PLT or lazy binding stubs,
3486 or in relaxation. If @samp{--insn32} is used, then the linker only uses
3487 32-bit instruction encodings. By default or if @samp{--no-insn32} is
3488 used, all instruction encodings are used, including 16-bit ones where
3491 @kindex --ignore-branch-isa
3492 @item --ignore-branch-isa
3493 @kindex --no-ignore-branch-isa
3494 @itemx --no-ignore-branch-isa
3495 These options control branch relocation checks for invalid ISA mode
3496 transitions. If @samp{--ignore-branch-isa} is used, then the linker
3497 accepts any branch relocations and any ISA mode transition required
3498 is lost in relocation calculation, except for some cases of @code{BAL}
3499 instructions which meet relaxation conditions and are converted to
3500 equivalent @code{JALX} instructions as the associated relocation is
3501 calculated. By default or if @samp{--no-ignore-branch-isa} is used
3502 a check is made causing the loss of an ISA mode transition to produce
3505 @kindex --compact-branches
3506 @item --compact-branches
3507 @kindex --no-compact-branches
3508 @itemx --no-compact-branches
3509 These options control the generation of compact instructions by the linker
3510 in the PLT entries for MIPS R6.
3519 @subsection Options specific to PDP11 targets
3521 @c man begin OPTIONS
3523 For the pdp11-aout target, three variants of the output format can be
3524 produced as selected by the following options. The default variant
3525 for pdp11-aout is the @samp{--omagic} option, whereas for other
3526 targets @samp{--nmagic} is the default. The @samp{--imagic} option is
3527 defined only for the pdp11-aout target, while the others are described
3528 here as they apply to the pdp11-aout target.
3537 Mark the output as @code{OMAGIC} (0407) in the @file{a.out} header to
3538 indicate that the text segment is not to be write-protected and
3539 shared. Since the text and data sections are both readable and
3540 writable, the data section is allocated immediately contiguous after
3541 the text segment. This is the oldest format for PDP11 executable
3542 programs and is the default for @command{ld} on PDP11 Unix systems
3543 from the beginning through 2.11BSD.
3550 Mark the output as @code{NMAGIC} (0410) in the @file{a.out} header to
3551 indicate that when the output file is executed, the text portion will
3552 be read-only and shareable among all processes executing the same
3553 file. This involves moving the data areas up to the first possible 8K
3554 byte page boundary following the end of the text. This option creates
3555 a @emph{pure executable} format.
3562 Mark the output as @code{IMAGIC} (0411) in the @file{a.out} header to
3563 indicate that when the output file is executed, the program text and
3564 data areas will be loaded into separate address spaces using the split
3565 instruction and data space feature of the memory management unit in
3566 larger models of the PDP11. This doubles the address space available
3567 to the program. The text segment is again pure, write-protected, and
3568 shareable. The only difference in the output format between this
3569 option and the others, besides the magic number, is that both the text
3570 and data sections start at location 0. The @samp{-z} option selected
3571 this format in 2.11BSD. This option creates a @emph{separate
3577 Equivalent to @samp{--nmagic} for pdp11-aout.
3586 @section Environment Variables
3588 @c man begin ENVIRONMENT
3590 You can change the behaviour of @command{ld} with the environment variables
3591 @ifclear SingleFormat
3594 @code{LDEMULATION} and @code{COLLECT_NO_DEMANGLE}.
3596 @ifclear SingleFormat
3598 @cindex default input format
3599 @code{GNUTARGET} determines the input-file object format if you don't
3600 use @samp{-b} (or its synonym @samp{--format}). Its value should be one
3601 of the BFD names for an input format (@pxref{BFD}). If there is no
3602 @code{GNUTARGET} in the environment, @command{ld} uses the natural format
3603 of the target. If @code{GNUTARGET} is set to @code{default} then BFD
3604 attempts to discover the input format by examining binary input files;
3605 this method often succeeds, but there are potential ambiguities, since
3606 there is no method of ensuring that the magic number used to specify
3607 object-file formats is unique. However, the configuration procedure for
3608 BFD on each system places the conventional format for that system first
3609 in the search-list, so ambiguities are resolved in favor of convention.
3613 @cindex default emulation
3614 @cindex emulation, default
3615 @code{LDEMULATION} determines the default emulation if you don't use the
3616 @samp{-m} option. The emulation can affect various aspects of linker
3617 behaviour, particularly the default linker script. You can list the
3618 available emulations with the @samp{--verbose} or @samp{-V} options. If
3619 the @samp{-m} option is not used, and the @code{LDEMULATION} environment
3620 variable is not defined, the default emulation depends upon how the
3621 linker was configured.
3623 @kindex COLLECT_NO_DEMANGLE
3624 @cindex demangling, default
3625 Normally, the linker will default to demangling symbols. However, if
3626 @code{COLLECT_NO_DEMANGLE} is set in the environment, then it will
3627 default to not demangling symbols. This environment variable is used in
3628 a similar fashion by the @code{gcc} linker wrapper program. The default
3629 may be overridden by the @samp{--demangle} and @samp{--no-demangle}
3636 @chapter Linker Scripts
3639 @cindex linker scripts
3640 @cindex command files
3641 Every link is controlled by a @dfn{linker script}. This script is
3642 written in the linker command language.
3644 The main purpose of the linker script is to describe how the sections in
3645 the input files should be mapped into the output file, and to control
3646 the memory layout of the output file. Most linker scripts do nothing
3647 more than this. However, when necessary, the linker script can also
3648 direct the linker to perform many other operations, using the commands
3651 The linker always uses a linker script. If you do not supply one
3652 yourself, the linker will use a default script that is compiled into the
3653 linker executable. You can use the @samp{--verbose} command-line option
3654 to display the default linker script. Certain command-line options,
3655 such as @samp{-r} or @samp{-N}, will affect the default linker script.
3657 You may supply your own linker script by using the @samp{-T} command
3658 line option. When you do this, your linker script will replace the
3659 default linker script.
3661 You may also use linker scripts implicitly by naming them as input files
3662 to the linker, as though they were files to be linked. @xref{Implicit
3666 * Basic Script Concepts:: Basic Linker Script Concepts
3667 * Script Format:: Linker Script Format
3668 * Simple Example:: Simple Linker Script Example
3669 * Simple Commands:: Simple Linker Script Commands
3670 * Assignments:: Assigning Values to Symbols
3671 * SECTIONS:: SECTIONS Command
3672 * MEMORY:: MEMORY Command
3673 * PHDRS:: PHDRS Command
3674 * VERSION:: VERSION Command
3675 * Expressions:: Expressions in Linker Scripts
3676 * Implicit Linker Scripts:: Implicit Linker Scripts
3679 @node Basic Script Concepts
3680 @section Basic Linker Script Concepts
3681 @cindex linker script concepts
3682 We need to define some basic concepts and vocabulary in order to
3683 describe the linker script language.
3685 The linker combines input files into a single output file. The output
3686 file and each input file are in a special data format known as an
3687 @dfn{object file format}. Each file is called an @dfn{object file}.
3688 The output file is often called an @dfn{executable}, but for our
3689 purposes we will also call it an object file. Each object file has,
3690 among other things, a list of @dfn{sections}. We sometimes refer to a
3691 section in an input file as an @dfn{input section}; similarly, a section
3692 in the output file is an @dfn{output section}.
3694 Each section in an object file has a name and a size. Most sections
3695 also have an associated block of data, known as the @dfn{section
3696 contents}. A section may be marked as @dfn{loadable}, which means that
3697 the contents should be loaded into memory when the output file is run.
3698 A section with no contents may be @dfn{allocatable}, which means that an
3699 area in memory should be set aside, but nothing in particular should be
3700 loaded there (in some cases this memory must be zeroed out). A section
3701 which is neither loadable nor allocatable typically contains some sort
3702 of debugging information.
3704 Every loadable or allocatable output section has two addresses. The
3705 first is the @dfn{VMA}, or virtual memory address. This is the address
3706 the section will have when the output file is run. The second is the
3707 @dfn{LMA}, or load memory address. This is the address at which the
3708 section will be loaded. In most cases the two addresses will be the
3709 same. An example of when they might be different is when a data section
3710 is loaded into ROM, and then copied into RAM when the program starts up
3711 (this technique is often used to initialize global variables in a ROM
3712 based system). In this case the ROM address would be the LMA, and the
3713 RAM address would be the VMA.
3715 You can see the sections in an object file by using the @code{objdump}
3716 program with the @samp{-h} option.
3718 Every object file also has a list of @dfn{symbols}, known as the
3719 @dfn{symbol table}. A symbol may be defined or undefined. Each symbol
3720 has a name, and each defined symbol has an address, among other
3721 information. If you compile a C or C++ program into an object file, you
3722 will get a defined symbol for every defined function and global or
3723 static variable. Every undefined function or global variable which is
3724 referenced in the input file will become an undefined symbol.
3726 You can see the symbols in an object file by using the @code{nm}
3727 program, or by using the @code{objdump} program with the @samp{-t}
3731 @section Linker Script Format
3732 @cindex linker script format
3733 Linker scripts are text files.
3735 You write a linker script as a series of commands. Each command is
3736 either a keyword, possibly followed by arguments, or an assignment to a
3737 symbol. You may separate commands using semicolons. Whitespace is
3740 Strings such as file or format names can normally be entered directly.
3741 If the file name contains a character such as a comma which would
3742 otherwise serve to separate file names, you may put the file name in
3743 double quotes. There is no way to use a double quote character in a
3746 You may include comments in linker scripts just as in C, delimited by
3747 @samp{/*} and @samp{*/}. As in C, comments are syntactically equivalent
3750 @node Simple Example
3751 @section Simple Linker Script Example
3752 @cindex linker script example
3753 @cindex example of linker script
3754 Many linker scripts are fairly simple.
3756 The simplest possible linker script has just one command:
3757 @samp{SECTIONS}. You use the @samp{SECTIONS} command to describe the
3758 memory layout of the output file.
3760 The @samp{SECTIONS} command is a powerful command. Here we will
3761 describe a simple use of it. Let's assume your program consists only of
3762 code, initialized data, and uninitialized data. These will be in the
3763 @samp{.text}, @samp{.data}, and @samp{.bss} sections, respectively.
3764 Let's assume further that these are the only sections which appear in
3767 For this example, let's say that the code should be loaded at address
3768 0x10000, and that the data should start at address 0x8000000. Here is a
3769 linker script which will do that:
3774 .text : @{ *(.text) @}
3776 .data : @{ *(.data) @}
3777 .bss : @{ *(.bss) @}
3781 You write the @samp{SECTIONS} command as the keyword @samp{SECTIONS},
3782 followed by a series of symbol assignments and output section
3783 descriptions enclosed in curly braces.
3785 The first line inside the @samp{SECTIONS} command of the above example
3786 sets the value of the special symbol @samp{.}, which is the location
3787 counter. If you do not specify the address of an output section in some
3788 other way (other ways are described later), the address is set from the
3789 current value of the location counter. The location counter is then
3790 incremented by the size of the output section. At the start of the
3791 @samp{SECTIONS} command, the location counter has the value @samp{0}.
3793 The second line defines an output section, @samp{.text}. The colon is
3794 required syntax which may be ignored for now. Within the curly braces
3795 after the output section name, you list the names of the input sections
3796 which should be placed into this output section. The @samp{*} is a
3797 wildcard which matches any file name. The expression @samp{*(.text)}
3798 means all @samp{.text} input sections in all input files.
3800 Since the location counter is @samp{0x10000} when the output section
3801 @samp{.text} is defined, the linker will set the address of the
3802 @samp{.text} section in the output file to be @samp{0x10000}.
3804 The remaining lines define the @samp{.data} and @samp{.bss} sections in
3805 the output file. The linker will place the @samp{.data} output section
3806 at address @samp{0x8000000}. After the linker places the @samp{.data}
3807 output section, the value of the location counter will be
3808 @samp{0x8000000} plus the size of the @samp{.data} output section. The
3809 effect is that the linker will place the @samp{.bss} output section
3810 immediately after the @samp{.data} output section in memory.
3812 The linker will ensure that each output section has the required
3813 alignment, by increasing the location counter if necessary. In this
3814 example, the specified addresses for the @samp{.text} and @samp{.data}
3815 sections will probably satisfy any alignment constraints, but the linker
3816 may have to create a small gap between the @samp{.data} and @samp{.bss}
3819 That's it! That's a simple and complete linker script.
3821 @node Simple Commands
3822 @section Simple Linker Script Commands
3823 @cindex linker script simple commands
3824 In this section we describe the simple linker script commands.
3827 * Entry Point:: Setting the entry point
3828 * File Commands:: Commands dealing with files
3829 @ifclear SingleFormat
3830 * Format Commands:: Commands dealing with object file formats
3833 * REGION_ALIAS:: Assign alias names to memory regions
3834 * Miscellaneous Commands:: Other linker script commands
3838 @subsection Setting the Entry Point
3839 @kindex ENTRY(@var{symbol})
3840 @cindex start of execution
3841 @cindex first instruction
3843 The first instruction to execute in a program is called the @dfn{entry
3844 point}. You can use the @code{ENTRY} linker script command to set the
3845 entry point. The argument is a symbol name:
3850 There are several ways to set the entry point. The linker will set the
3851 entry point by trying each of the following methods in order, and
3852 stopping when one of them succeeds:
3855 the @samp{-e} @var{entry} command-line option;
3857 the @code{ENTRY(@var{symbol})} command in a linker script;
3859 the value of a target-specific symbol, if it is defined; For many
3860 targets this is @code{start}, but PE- and BeOS-based systems for example
3861 check a list of possible entry symbols, matching the first one found.
3863 the address of the first byte of the @samp{.text} section, if present;
3865 The address @code{0}.
3869 @subsection Commands Dealing with Files
3870 @cindex linker script file commands
3871 Several linker script commands deal with files.
3874 @item INCLUDE @var{filename}
3875 @kindex INCLUDE @var{filename}
3876 @cindex including a linker script
3877 Include the linker script @var{filename} at this point. The file will
3878 be searched for in the current directory, and in any directory specified
3879 with the @option{-L} option. You can nest calls to @code{INCLUDE} up to
3882 You can place @code{INCLUDE} directives at the top level, in @code{MEMORY} or
3883 @code{SECTIONS} commands, or in output section descriptions.
3885 @item INPUT(@var{file}, @var{file}, @dots{})
3886 @itemx INPUT(@var{file} @var{file} @dots{})
3887 @kindex INPUT(@var{files})
3888 @cindex input files in linker scripts
3889 @cindex input object files in linker scripts
3890 @cindex linker script input object files
3891 The @code{INPUT} command directs the linker to include the named files
3892 in the link, as though they were named on the command line.
3894 For example, if you always want to include @file{subr.o} any time you do
3895 a link, but you can't be bothered to put it on every link command line,
3896 then you can put @samp{INPUT (subr.o)} in your linker script.
3898 In fact, if you like, you can list all of your input files in the linker
3899 script, and then invoke the linker with nothing but a @samp{-T} option.
3901 In case a @dfn{sysroot prefix} is configured, and the filename starts
3902 with the @samp{/} character, and the script being processed was
3903 located inside the @dfn{sysroot prefix}, the filename will be looked
3904 for in the @dfn{sysroot prefix}. The @dfn{sysroot prefix} can also be forced by specifying
3905 @code{=} as the first character in the filename path, or prefixing the
3906 filename path with @code{$SYSROOT}. See also the description of
3907 @samp{-L} in @ref{Options,,Command-line Options}.
3909 If a @dfn{sysroot prefix} is not used then the linker will try to open
3910 the file in the directory containing the linker script. If it is not
3911 found the linker will then search the current directory. If it is still
3912 not found the linker will search through the archive library search
3915 If you use @samp{INPUT (-l@var{file})}, @command{ld} will transform the
3916 name to @code{lib@var{file}.a}, as with the command-line argument
3919 When you use the @code{INPUT} command in an implicit linker script, the
3920 files will be included in the link at the point at which the linker
3921 script file is included. This can affect archive searching.
3923 @item GROUP(@var{file}, @var{file}, @dots{})
3924 @itemx GROUP(@var{file} @var{file} @dots{})
3925 @kindex GROUP(@var{files})
3926 @cindex grouping input files
3927 The @code{GROUP} command is like @code{INPUT}, except that the named
3928 files should all be archives, and they are searched repeatedly until no
3929 new undefined references are created. See the description of @samp{-(}
3930 in @ref{Options,,Command-line Options}.
3932 @item AS_NEEDED(@var{file}, @var{file}, @dots{})
3933 @itemx AS_NEEDED(@var{file} @var{file} @dots{})
3934 @kindex AS_NEEDED(@var{files})
3935 This construct can appear only inside of the @code{INPUT} or @code{GROUP}
3936 commands, among other filenames. The files listed will be handled
3937 as if they appear directly in the @code{INPUT} or @code{GROUP} commands,
3938 with the exception of ELF shared libraries, that will be added only
3939 when they are actually needed. This construct essentially enables
3940 @option{--as-needed} option for all the files listed inside of it
3941 and restores previous @option{--as-needed} resp. @option{--no-as-needed}
3944 @item OUTPUT(@var{filename})
3945 @kindex OUTPUT(@var{filename})
3946 @cindex output file name in linker script
3947 The @code{OUTPUT} command names the output file. Using
3948 @code{OUTPUT(@var{filename})} in the linker script is exactly like using
3949 @samp{-o @var{filename}} on the command line (@pxref{Options,,Command
3950 Line Options}). If both are used, the command-line option takes
3953 You can use the @code{OUTPUT} command to define a default name for the
3954 output file other than the usual default of @file{a.out}.
3956 @item SEARCH_DIR(@var{path})
3957 @kindex SEARCH_DIR(@var{path})
3958 @cindex library search path in linker script
3959 @cindex archive search path in linker script
3960 @cindex search path in linker script
3961 The @code{SEARCH_DIR} command adds @var{path} to the list of paths where
3962 @command{ld} looks for archive libraries. Using
3963 @code{SEARCH_DIR(@var{path})} is exactly like using @samp{-L @var{path}}
3964 on the command line (@pxref{Options,,Command-line Options}). If both
3965 are used, then the linker will search both paths. Paths specified using
3966 the command-line option are searched first.
3968 @item STARTUP(@var{filename})
3969 @kindex STARTUP(@var{filename})
3970 @cindex first input file
3971 The @code{STARTUP} command is just like the @code{INPUT} command, except
3972 that @var{filename} will become the first input file to be linked, as
3973 though it were specified first on the command line. This may be useful
3974 when using a system in which the entry point is always the start of the
3978 @ifclear SingleFormat
3979 @node Format Commands
3980 @subsection Commands Dealing with Object File Formats
3981 A couple of linker script commands deal with object file formats.
3984 @item OUTPUT_FORMAT(@var{bfdname})
3985 @itemx OUTPUT_FORMAT(@var{default}, @var{big}, @var{little})
3986 @kindex OUTPUT_FORMAT(@var{bfdname})
3987 @cindex output file format in linker script
3988 The @code{OUTPUT_FORMAT} command names the BFD format to use for the
3989 output file (@pxref{BFD}). Using @code{OUTPUT_FORMAT(@var{bfdname})} is
3990 exactly like using @samp{--oformat @var{bfdname}} on the command line
3991 (@pxref{Options,,Command-line Options}). If both are used, the command
3992 line option takes precedence.
3994 You can use @code{OUTPUT_FORMAT} with three arguments to use different
3995 formats based on the @samp{-EB} and @samp{-EL} command-line options.
3996 This permits the linker script to set the output format based on the
3999 If neither @samp{-EB} nor @samp{-EL} are used, then the output format
4000 will be the first argument, @var{default}. If @samp{-EB} is used, the
4001 output format will be the second argument, @var{big}. If @samp{-EL} is
4002 used, the output format will be the third argument, @var{little}.
4004 For example, the default linker script for the MIPS ELF target uses this
4007 OUTPUT_FORMAT(elf32-bigmips, elf32-bigmips, elf32-littlemips)
4009 This says that the default format for the output file is
4010 @samp{elf32-bigmips}, but if the user uses the @samp{-EL} command-line
4011 option, the output file will be created in the @samp{elf32-littlemips}
4014 @item TARGET(@var{bfdname})
4015 @kindex TARGET(@var{bfdname})
4016 @cindex input file format in linker script
4017 The @code{TARGET} command names the BFD format to use when reading input
4018 files. It affects subsequent @code{INPUT} and @code{GROUP} commands.
4019 This command is like using @samp{-b @var{bfdname}} on the command line
4020 (@pxref{Options,,Command-line Options}). If the @code{TARGET} command
4021 is used but @code{OUTPUT_FORMAT} is not, then the last @code{TARGET}
4022 command is also used to set the format for the output file. @xref{BFD}.
4027 @subsection Assign alias names to memory regions
4028 @kindex REGION_ALIAS(@var{alias}, @var{region})
4029 @cindex region alias
4030 @cindex region names
4032 Alias names can be added to existing memory regions created with the
4033 @ref{MEMORY} command. Each name corresponds to at most one memory region.
4036 REGION_ALIAS(@var{alias}, @var{region})
4039 The @code{REGION_ALIAS} function creates an alias name @var{alias} for the
4040 memory region @var{region}. This allows a flexible mapping of output sections
4041 to memory regions. An example follows.
4043 Suppose we have an application for embedded systems which come with various
4044 memory storage devices. All have a general purpose, volatile memory @code{RAM}
4045 that allows code execution or data storage. Some may have a read-only,
4046 non-volatile memory @code{ROM} that allows code execution and read-only data
4047 access. The last variant is a read-only, non-volatile memory @code{ROM2} with
4048 read-only data access and no code execution capability. We have four output
4053 @code{.text} program code;
4055 @code{.rodata} read-only data;
4057 @code{.data} read-write initialized data;
4059 @code{.bss} read-write zero initialized data.
4062 The goal is to provide a linker command file that contains a system independent
4063 part defining the output sections and a system dependent part mapping the
4064 output sections to the memory regions available on the system. Our embedded
4065 systems come with three different memory setups @code{A}, @code{B} and
4067 @multitable @columnfractions .25 .25 .25 .25
4068 @item Section @tab Variant A @tab Variant B @tab Variant C
4069 @item .text @tab RAM @tab ROM @tab ROM
4070 @item .rodata @tab RAM @tab ROM @tab ROM2
4071 @item .data @tab RAM @tab RAM/ROM @tab RAM/ROM2
4072 @item .bss @tab RAM @tab RAM @tab RAM
4074 The notation @code{RAM/ROM} or @code{RAM/ROM2} means that this section is
4075 loaded into region @code{ROM} or @code{ROM2} respectively. Please note that
4076 the load address of the @code{.data} section starts in all three variants at
4077 the end of the @code{.rodata} section.
4079 The base linker script that deals with the output sections follows. It
4080 includes the system dependent @code{linkcmds.memory} file that describes the
4083 INCLUDE linkcmds.memory
4096 .data : AT (rodata_end)
4101 data_size = SIZEOF(.data);
4102 data_load_start = LOADADDR(.data);
4110 Now we need three different @code{linkcmds.memory} files to define memory
4111 regions and alias names. The content of @code{linkcmds.memory} for the three
4112 variants @code{A}, @code{B} and @code{C}:
4115 Here everything goes into the @code{RAM}.
4119 RAM : ORIGIN = 0, LENGTH = 4M
4122 REGION_ALIAS("REGION_TEXT", RAM);
4123 REGION_ALIAS("REGION_RODATA", RAM);
4124 REGION_ALIAS("REGION_DATA", RAM);
4125 REGION_ALIAS("REGION_BSS", RAM);
4128 Program code and read-only data go into the @code{ROM}. Read-write data goes
4129 into the @code{RAM}. An image of the initialized data is loaded into the
4130 @code{ROM} and will be copied during system start into the @code{RAM}.
4134 ROM : ORIGIN = 0, LENGTH = 3M
4135 RAM : ORIGIN = 0x10000000, LENGTH = 1M
4138 REGION_ALIAS("REGION_TEXT", ROM);
4139 REGION_ALIAS("REGION_RODATA", ROM);
4140 REGION_ALIAS("REGION_DATA", RAM);
4141 REGION_ALIAS("REGION_BSS", RAM);
4144 Program code goes into the @code{ROM}. Read-only data goes into the
4145 @code{ROM2}. Read-write data goes into the @code{RAM}. An image of the
4146 initialized data is loaded into the @code{ROM2} and will be copied during
4147 system start into the @code{RAM}.
4151 ROM : ORIGIN = 0, LENGTH = 2M
4152 ROM2 : ORIGIN = 0x10000000, LENGTH = 1M
4153 RAM : ORIGIN = 0x20000000, LENGTH = 1M
4156 REGION_ALIAS("REGION_TEXT", ROM);
4157 REGION_ALIAS("REGION_RODATA", ROM2);
4158 REGION_ALIAS("REGION_DATA", RAM);
4159 REGION_ALIAS("REGION_BSS", RAM);
4163 It is possible to write a common system initialization routine to copy the
4164 @code{.data} section from @code{ROM} or @code{ROM2} into the @code{RAM} if
4169 extern char data_start [];
4170 extern char data_size [];
4171 extern char data_load_start [];
4173 void copy_data(void)
4175 if (data_start != data_load_start)
4177 memcpy(data_start, data_load_start, (size_t) data_size);
4182 @node Miscellaneous Commands
4183 @subsection Other Linker Script Commands
4184 There are a few other linker scripts commands.
4187 @item ASSERT(@var{exp}, @var{message})
4189 @cindex assertion in linker script
4190 Ensure that @var{exp} is non-zero. If it is zero, then exit the linker
4191 with an error code, and print @var{message}.
4193 Note that assertions are checked before the final stages of linking
4194 take place. This means that expressions involving symbols PROVIDEd
4195 inside section definitions will fail if the user has not set values
4196 for those symbols. The only exception to this rule is PROVIDEd
4197 symbols that just reference dot. Thus an assertion like this:
4202 PROVIDE (__stack = .);
4203 PROVIDE (__stack_size = 0x100);
4204 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
4208 will fail if @code{__stack_size} is not defined elsewhere. Symbols
4209 PROVIDEd outside of section definitions are evaluated earlier, so they
4210 can be used inside ASSERTions. Thus:
4213 PROVIDE (__stack_size = 0x100);
4216 PROVIDE (__stack = .);
4217 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
4223 @item EXTERN(@var{symbol} @var{symbol} @dots{})
4225 @cindex undefined symbol in linker script
4226 Force @var{symbol} to be entered in the output file as an undefined
4227 symbol. Doing this may, for example, trigger linking of additional
4228 modules from standard libraries. You may list several @var{symbol}s for
4229 each @code{EXTERN}, and you may use @code{EXTERN} multiple times. This
4230 command has the same effect as the @samp{-u} command-line option.
4232 @item FORCE_COMMON_ALLOCATION
4233 @kindex FORCE_COMMON_ALLOCATION
4234 @cindex common allocation in linker script
4235 This command has the same effect as the @samp{-d} command-line option:
4236 to make @command{ld} assign space to common symbols even if a relocatable
4237 output file is specified (@samp{-r}).
4239 @item INHIBIT_COMMON_ALLOCATION
4240 @kindex INHIBIT_COMMON_ALLOCATION
4241 @cindex common allocation in linker script
4242 This command has the same effect as the @samp{--no-define-common}
4243 command-line option: to make @code{ld} omit the assignment of addresses
4244 to common symbols even for a non-relocatable output file.
4246 @item FORCE_GROUP_ALLOCATION
4247 @kindex FORCE_GROUP_ALLOCATION
4248 @cindex group allocation in linker script
4249 @cindex section groups
4251 This command has the same effect as the
4252 @samp{--force-group-allocation} command-line option: to make
4253 @command{ld} place section group members like normal input sections,
4254 and to delete the section groups even if a relocatable output file is
4255 specified (@samp{-r}).
4257 @item INSERT [ AFTER | BEFORE ] @var{output_section}
4259 @cindex insert user script into default script
4260 This command is typically used in a script specified by @samp{-T} to
4261 augment the default @code{SECTIONS} with, for example, overlays. It
4262 inserts all prior linker script statements after (or before)
4263 @var{output_section}, and also causes @samp{-T} to not override the
4264 default linker script. The exact insertion point is as for orphan
4265 sections. @xref{Location Counter}. The insertion happens after the
4266 linker has mapped input sections to output sections. Prior to the
4267 insertion, since @samp{-T} scripts are parsed before the default
4268 linker script, statements in the @samp{-T} script occur before the
4269 default linker script statements in the internal linker representation
4270 of the script. In particular, input section assignments will be made
4271 to @samp{-T} output sections before those in the default script. Here
4272 is an example of how a @samp{-T} script using @code{INSERT} might look:
4279 .ov1 @{ ov1*(.text) @}
4280 .ov2 @{ ov2*(.text) @}
4286 @item NOCROSSREFS(@var{section} @var{section} @dots{})
4287 @kindex NOCROSSREFS(@var{sections})
4288 @cindex cross references
4289 This command may be used to tell @command{ld} to issue an error about any
4290 references among certain output sections.
4292 In certain types of programs, particularly on embedded systems when
4293 using overlays, when one section is loaded into memory, another section
4294 will not be. Any direct references between the two sections would be
4295 errors. For example, it would be an error if code in one section called
4296 a function defined in the other section.
4298 The @code{NOCROSSREFS} command takes a list of output section names. If
4299 @command{ld} detects any cross references between the sections, it reports
4300 an error and returns a non-zero exit status. Note that the
4301 @code{NOCROSSREFS} command uses output section names, not input section
4304 @item NOCROSSREFS_TO(@var{tosection} @var{fromsection} @dots{})
4305 @kindex NOCROSSREFS_TO(@var{tosection} @var{fromsections})
4306 @cindex cross references
4307 This command may be used to tell @command{ld} to issue an error about any
4308 references to one section from a list of other sections.
4310 The @code{NOCROSSREFS} command is useful when ensuring that two or more
4311 output sections are entirely independent but there are situations where
4312 a one-way dependency is needed. For example, in a multi-core application
4313 there may be shared code that can be called from each core but for safety
4314 must never call back.
4316 The @code{NOCROSSREFS_TO} command takes a list of output section names.
4317 The first section can not be referenced from any of the other sections.
4318 If @command{ld} detects any references to the first section from any of
4319 the other sections, it reports an error and returns a non-zero exit
4320 status. Note that the @code{NOCROSSREFS_TO} command uses output section
4321 names, not input section names.
4323 @ifclear SingleFormat
4324 @item OUTPUT_ARCH(@var{bfdarch})
4325 @kindex OUTPUT_ARCH(@var{bfdarch})
4326 @cindex machine architecture
4327 @cindex architecture
4328 Specify a particular output machine architecture. The argument is one
4329 of the names used by the BFD library (@pxref{BFD}). You can see the
4330 architecture of an object file by using the @code{objdump} program with
4331 the @samp{-f} option.
4334 @item LD_FEATURE(@var{string})
4335 @kindex LD_FEATURE(@var{string})
4336 This command may be used to modify @command{ld} behavior. If
4337 @var{string} is @code{"SANE_EXPR"} then absolute symbols and numbers
4338 in a script are simply treated as numbers everywhere.
4339 @xref{Expression Section}.
4343 @section Assigning Values to Symbols
4344 @cindex assignment in scripts
4345 @cindex symbol definition, scripts
4346 @cindex variables, defining
4347 You may assign a value to a symbol in a linker script. This will define
4348 the symbol and place it into the symbol table with a global scope.
4351 * Simple Assignments:: Simple Assignments
4354 * PROVIDE_HIDDEN:: PROVIDE_HIDDEN
4355 * Source Code Reference:: How to use a linker script defined symbol in source code
4358 @node Simple Assignments
4359 @subsection Simple Assignments
4361 You may assign to a symbol using any of the C assignment operators:
4364 @item @var{symbol} = @var{expression} ;
4365 @itemx @var{symbol} += @var{expression} ;
4366 @itemx @var{symbol} -= @var{expression} ;
4367 @itemx @var{symbol} *= @var{expression} ;
4368 @itemx @var{symbol} /= @var{expression} ;
4369 @itemx @var{symbol} <<= @var{expression} ;
4370 @itemx @var{symbol} >>= @var{expression} ;
4371 @itemx @var{symbol} &= @var{expression} ;
4372 @itemx @var{symbol} |= @var{expression} ;
4375 The first case will define @var{symbol} to the value of
4376 @var{expression}. In the other cases, @var{symbol} must already be
4377 defined, and the value will be adjusted accordingly.
4379 The special symbol name @samp{.} indicates the location counter. You
4380 may only use this within a @code{SECTIONS} command. @xref{Location Counter}.
4382 The semicolon after @var{expression} is required.
4384 Expressions are defined below; see @ref{Expressions}.
4386 You may write symbol assignments as commands in their own right, or as
4387 statements within a @code{SECTIONS} command, or as part of an output
4388 section description in a @code{SECTIONS} command.
4390 The section of the symbol will be set from the section of the
4391 expression; for more information, see @ref{Expression Section}.
4393 Here is an example showing the three different places that symbol
4394 assignments may be used:
4405 _bdata = (. + 3) & ~ 3;
4406 .data : @{ *(.data) @}
4410 In this example, the symbol @samp{floating_point} will be defined as
4411 zero. The symbol @samp{_etext} will be defined as the address following
4412 the last @samp{.text} input section. The symbol @samp{_bdata} will be
4413 defined as the address following the @samp{.text} output section aligned
4414 upward to a 4 byte boundary.
4419 For ELF targeted ports, define a symbol that will be hidden and won't be
4420 exported. The syntax is @code{HIDDEN(@var{symbol} = @var{expression})}.
4422 Here is the example from @ref{Simple Assignments}, rewritten to use
4426 HIDDEN(floating_point = 0);
4434 HIDDEN(_bdata = (. + 3) & ~ 3);
4435 .data : @{ *(.data) @}
4439 In this case none of the three symbols will be visible outside this module.
4444 In some cases, it is desirable for a linker script to define a symbol
4445 only if it is referenced and is not defined by any object included in
4446 the link. For example, traditional linkers defined the symbol
4447 @samp{etext}. However, ANSI C requires that the user be able to use
4448 @samp{etext} as a function name without encountering an error. The
4449 @code{PROVIDE} keyword may be used to define a symbol, such as
4450 @samp{etext}, only if it is referenced but not defined. The syntax is
4451 @code{PROVIDE(@var{symbol} = @var{expression})}.
4453 Here is an example of using @code{PROVIDE} to define @samp{etext}:
4466 In this example, if the program defines @samp{_etext} (with a leading
4467 underscore), the linker will give a multiple definition diagnostic. If,
4468 on the other hand, the program defines @samp{etext} (with no leading
4469 underscore), the linker will silently use the definition in the program.
4470 If the program references @samp{etext} but does not define it, the
4471 linker will use the definition in the linker script.
4473 Note - the @code{PROVIDE} directive considers a common symbol to be
4474 defined, even though such a symbol could be combined with the symbol
4475 that the @code{PROVIDE} would create. This is particularly important
4476 when considering constructor and destructor list symbols such as
4477 @samp{__CTOR_LIST__} as these are often defined as common symbols.
4479 @node PROVIDE_HIDDEN
4480 @subsection PROVIDE_HIDDEN
4481 @cindex PROVIDE_HIDDEN
4482 Similar to @code{PROVIDE}. For ELF targeted ports, the symbol will be
4483 hidden and won't be exported.
4485 @node Source Code Reference
4486 @subsection Source Code Reference
4488 Accessing a linker script defined variable from source code is not
4489 intuitive. In particular a linker script symbol is not equivalent to
4490 a variable declaration in a high level language, it is instead a
4491 symbol that does not have a value.
4493 Before going further, it is important to note that compilers often
4494 transform names in the source code into different names when they are
4495 stored in the symbol table. For example, Fortran compilers commonly
4496 prepend or append an underscore, and C++ performs extensive @samp{name
4497 mangling}. Therefore there might be a discrepancy between the name
4498 of a variable as it is used in source code and the name of the same
4499 variable as it is defined in a linker script. For example in C a
4500 linker script variable might be referred to as:
4506 But in the linker script it might be defined as:
4512 In the remaining examples however it is assumed that no name
4513 transformation has taken place.
4515 When a symbol is declared in a high level language such as C, two
4516 things happen. The first is that the compiler reserves enough space
4517 in the program's memory to hold the @emph{value} of the symbol. The
4518 second is that the compiler creates an entry in the program's symbol
4519 table which holds the symbol's @emph{address}. ie the symbol table
4520 contains the address of the block of memory holding the symbol's
4521 value. So for example the following C declaration, at file scope:
4527 creates an entry called @samp{foo} in the symbol table. This entry
4528 holds the address of an @samp{int} sized block of memory where the
4529 number 1000 is initially stored.
4531 When a program references a symbol the compiler generates code that
4532 first accesses the symbol table to find the address of the symbol's
4533 memory block and then code to read the value from that memory block.
4540 looks up the symbol @samp{foo} in the symbol table, gets the address
4541 associated with this symbol and then writes the value 1 into that
4548 looks up the symbol @samp{foo} in the symbol table, gets its address
4549 and then copies this address into the block of memory associated with
4550 the variable @samp{a}.
4552 Linker scripts symbol declarations, by contrast, create an entry in
4553 the symbol table but do not assign any memory to them. Thus they are
4554 an address without a value. So for example the linker script definition:
4560 creates an entry in the symbol table called @samp{foo} which holds
4561 the address of memory location 1000, but nothing special is stored at
4562 address 1000. This means that you cannot access the @emph{value} of a
4563 linker script defined symbol - it has no value - all you can do is
4564 access the @emph{address} of a linker script defined symbol.
4566 Hence when you are using a linker script defined symbol in source code
4567 you should always take the address of the symbol, and never attempt to
4568 use its value. For example suppose you want to copy the contents of a
4569 section of memory called .ROM into a section called .FLASH and the
4570 linker script contains these declarations:
4574 start_of_ROM = .ROM;
4575 end_of_ROM = .ROM + sizeof (.ROM);
4576 start_of_FLASH = .FLASH;
4580 Then the C source code to perform the copy would be:
4584 extern char start_of_ROM, end_of_ROM, start_of_FLASH;
4586 memcpy (& start_of_FLASH, & start_of_ROM, & end_of_ROM - & start_of_ROM);
4590 Note the use of the @samp{&} operators. These are correct.
4591 Alternatively the symbols can be treated as the names of vectors or
4592 arrays and then the code will again work as expected:
4596 extern char start_of_ROM[], end_of_ROM[], start_of_FLASH[];
4598 memcpy (start_of_FLASH, start_of_ROM, end_of_ROM - start_of_ROM);
4602 Note how using this method does not require the use of @samp{&}
4606 @section SECTIONS Command
4608 The @code{SECTIONS} command tells the linker how to map input sections
4609 into output sections, and how to place the output sections in memory.
4611 The format of the @code{SECTIONS} command is:
4615 @var{sections-command}
4616 @var{sections-command}
4621 Each @var{sections-command} may of be one of the following:
4625 an @code{ENTRY} command (@pxref{Entry Point,,Entry command})
4627 a symbol assignment (@pxref{Assignments})
4629 an output section description
4631 an overlay description
4634 The @code{ENTRY} command and symbol assignments are permitted inside the
4635 @code{SECTIONS} command for convenience in using the location counter in
4636 those commands. This can also make the linker script easier to
4637 understand because you can use those commands at meaningful points in
4638 the layout of the output file.
4640 Output section descriptions and overlay descriptions are described
4643 If you do not use a @code{SECTIONS} command in your linker script, the
4644 linker will place each input section into an identically named output
4645 section in the order that the sections are first encountered in the
4646 input files. If all input sections are present in the first file, for
4647 example, the order of sections in the output file will match the order
4648 in the first input file. The first section will be at address zero.
4651 * Output Section Description:: Output section description
4652 * Output Section Name:: Output section name
4653 * Output Section Address:: Output section address
4654 * Input Section:: Input section description
4655 * Output Section Data:: Output section data
4656 * Output Section Keywords:: Output section keywords
4657 * Output Section Discarding:: Output section discarding
4658 * Output Section Attributes:: Output section attributes
4659 * Overlay Description:: Overlay description
4662 @node Output Section Description
4663 @subsection Output Section Description
4664 The full description of an output section looks like this:
4667 @var{section} [@var{address}] [(@var{type})] :
4669 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
4670 [SUBALIGN(@var{subsection_align})]
4673 @var{output-section-command}
4674 @var{output-section-command}
4676 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}] [,]
4680 Most output sections do not use most of the optional section attributes.
4682 The whitespace around @var{section} is required, so that the section
4683 name is unambiguous. The colon and the curly braces are also required.
4684 The comma at the end may be required if a @var{fillexp} is used and
4685 the next @var{sections-command} looks like a continuation of the expression.
4686 The line breaks and other white space are optional.
4688 Each @var{output-section-command} may be one of the following:
4692 a symbol assignment (@pxref{Assignments})
4694 an input section description (@pxref{Input Section})
4696 data values to include directly (@pxref{Output Section Data})
4698 a special output section keyword (@pxref{Output Section Keywords})
4701 @node Output Section Name
4702 @subsection Output Section Name
4703 @cindex name, section
4704 @cindex section name
4705 The name of the output section is @var{section}. @var{section} must
4706 meet the constraints of your output format. In formats which only
4707 support a limited number of sections, such as @code{a.out}, the name
4708 must be one of the names supported by the format (@code{a.out}, for
4709 example, allows only @samp{.text}, @samp{.data} or @samp{.bss}). If the
4710 output format supports any number of sections, but with numbers and not
4711 names (as is the case for Oasys), the name should be supplied as a
4712 quoted numeric string. A section name may consist of any sequence of
4713 characters, but a name which contains any unusual characters such as
4714 commas must be quoted.
4716 The output section name @samp{/DISCARD/} is special; @ref{Output Section
4719 @node Output Section Address
4720 @subsection Output Section Address
4721 @cindex address, section
4722 @cindex section address
4723 The @var{address} is an expression for the VMA (the virtual memory
4724 address) of the output section. This address is optional, but if it
4725 is provided then the output address will be set exactly as specified.
4727 If the output address is not specified then one will be chosen for the
4728 section, based on the heuristic below. This address will be adjusted
4729 to fit the alignment requirement of the output section. The
4730 alignment requirement is the strictest alignment of any input section
4731 contained within the output section.
4733 The output section address heuristic is as follows:
4737 If an output memory @var{region} is set for the section then it
4738 is added to this region and its address will be the next free address
4742 If the MEMORY command has been used to create a list of memory
4743 regions then the first region which has attributes compatible with the
4744 section is selected to contain it. The section's output address will
4745 be the next free address in that region; @ref{MEMORY}.
4748 If no memory regions were specified, or none match the section then
4749 the output address will be based on the current value of the location
4757 .text . : @{ *(.text) @}
4764 .text : @{ *(.text) @}
4768 are subtly different. The first will set the address of the
4769 @samp{.text} output section to the current value of the location
4770 counter. The second will set it to the current value of the location
4771 counter aligned to the strictest alignment of any of the @samp{.text}
4774 The @var{address} may be an arbitrary expression; @ref{Expressions}.
4775 For example, if you want to align the section on a 0x10 byte boundary,
4776 so that the lowest four bits of the section address are zero, you could
4777 do something like this:
4779 .text ALIGN(0x10) : @{ *(.text) @}
4782 This works because @code{ALIGN} returns the current location counter
4783 aligned upward to the specified value.
4785 Specifying @var{address} for a section will change the value of the
4786 location counter, provided that the section is non-empty. (Empty
4787 sections are ignored).
4790 @subsection Input Section Description
4791 @cindex input sections
4792 @cindex mapping input sections to output sections
4793 The most common output section command is an input section description.
4795 The input section description is the most basic linker script operation.
4796 You use output sections to tell the linker how to lay out your program
4797 in memory. You use input section descriptions to tell the linker how to
4798 map the input files into your memory layout.
4801 * Input Section Basics:: Input section basics
4802 * Input Section Wildcards:: Input section wildcard patterns
4803 * Input Section Common:: Input section for common symbols
4804 * Input Section Keep:: Input section and garbage collection
4805 * Input Section Example:: Input section example
4808 @node Input Section Basics
4809 @subsubsection Input Section Basics
4810 @cindex input section basics
4811 An input section description consists of a file name optionally followed
4812 by a list of section names in parentheses.
4814 The file name and the section name may be wildcard patterns, which we
4815 describe further below (@pxref{Input Section Wildcards}).
4817 The most common input section description is to include all input
4818 sections with a particular name in the output section. For example, to
4819 include all input @samp{.text} sections, you would write:
4824 Here the @samp{*} is a wildcard which matches any file name. To exclude a list
4825 @cindex EXCLUDE_FILE
4826 of files from matching the file name wildcard, EXCLUDE_FILE may be used to
4827 match all files except the ones specified in the EXCLUDE_FILE list. For
4830 EXCLUDE_FILE (*crtend.o *otherfile.o) *(.ctors)
4833 will cause all .ctors sections from all files except @file{crtend.o}
4834 and @file{otherfile.o} to be included. The EXCLUDE_FILE can also be
4835 placed inside the section list, for example:
4837 *(EXCLUDE_FILE (*crtend.o *otherfile.o) .ctors)
4840 The result of this is identically to the previous example. Supporting
4841 two syntaxes for EXCLUDE_FILE is useful if the section list contains
4842 more than one section, as described below.
4844 There are two ways to include more than one section:
4850 The difference between these is the order in which the @samp{.text} and
4851 @samp{.rdata} input sections will appear in the output section. In the
4852 first example, they will be intermingled, appearing in the same order as
4853 they are found in the linker input. In the second example, all
4854 @samp{.text} input sections will appear first, followed by all
4855 @samp{.rdata} input sections.
4857 When using EXCLUDE_FILE with more than one section, if the exclusion
4858 is within the section list then the exclusion only applies to the
4859 immediately following section, for example:
4861 *(EXCLUDE_FILE (*somefile.o) .text .rdata)
4864 will cause all @samp{.text} sections from all files except
4865 @file{somefile.o} to be included, while all @samp{.rdata} sections
4866 from all files, including @file{somefile.o}, will be included. To
4867 exclude the @samp{.rdata} sections from @file{somefile.o} the example
4868 could be modified to:
4870 *(EXCLUDE_FILE (*somefile.o) .text EXCLUDE_FILE (*somefile.o) .rdata)
4873 Alternatively, placing the EXCLUDE_FILE outside of the section list,
4874 before the input file selection, will cause the exclusion to apply for
4875 all sections. Thus the previous example can be rewritten as:
4877 EXCLUDE_FILE (*somefile.o) *(.text .rdata)
4880 You can specify a file name to include sections from a particular file.
4881 You would do this if one or more of your files contain special data that
4882 needs to be at a particular location in memory. For example:
4887 To refine the sections that are included based on the section flags
4888 of an input section, INPUT_SECTION_FLAGS may be used.
4890 Here is a simple example for using Section header flags for ELF sections:
4895 .text : @{ INPUT_SECTION_FLAGS (SHF_MERGE & SHF_STRINGS) *(.text) @}
4896 .text2 : @{ INPUT_SECTION_FLAGS (!SHF_WRITE) *(.text) @}
4901 In this example, the output section @samp{.text} will be comprised of any
4902 input section matching the name *(.text) whose section header flags
4903 @code{SHF_MERGE} and @code{SHF_STRINGS} are set. The output section
4904 @samp{.text2} will be comprised of any input section matching the name *(.text)
4905 whose section header flag @code{SHF_WRITE} is clear.
4907 You can also specify files within archives by writing a pattern
4908 matching the archive, a colon, then the pattern matching the file,
4909 with no whitespace around the colon.
4913 matches file within archive
4915 matches the whole archive
4917 matches file but not one in an archive
4920 Either one or both of @samp{archive} and @samp{file} can contain shell
4921 wildcards. On DOS based file systems, the linker will assume that a
4922 single letter followed by a colon is a drive specifier, so
4923 @samp{c:myfile.o} is a simple file specification, not @samp{myfile.o}
4924 within an archive called @samp{c}. @samp{archive:file} filespecs may
4925 also be used within an @code{EXCLUDE_FILE} list, but may not appear in
4926 other linker script contexts. For instance, you cannot extract a file
4927 from an archive by using @samp{archive:file} in an @code{INPUT}
4930 If you use a file name without a list of sections, then all sections in
4931 the input file will be included in the output section. This is not
4932 commonly done, but it may by useful on occasion. For example:
4937 When you use a file name which is not an @samp{archive:file} specifier
4938 and does not contain any wild card
4939 characters, the linker will first see if you also specified the file
4940 name on the linker command line or in an @code{INPUT} command. If you
4941 did not, the linker will attempt to open the file as an input file, as
4942 though it appeared on the command line. Note that this differs from an
4943 @code{INPUT} command, because the linker will not search for the file in
4944 the archive search path.
4946 @node Input Section Wildcards
4947 @subsubsection Input Section Wildcard Patterns
4948 @cindex input section wildcards
4949 @cindex wildcard file name patterns
4950 @cindex file name wildcard patterns
4951 @cindex section name wildcard patterns
4952 In an input section description, either the file name or the section
4953 name or both may be wildcard patterns.
4955 The file name of @samp{*} seen in many examples is a simple wildcard
4956 pattern for the file name.
4958 The wildcard patterns are like those used by the Unix shell.
4962 matches any number of characters
4964 matches any single character
4966 matches a single instance of any of the @var{chars}; the @samp{-}
4967 character may be used to specify a range of characters, as in
4968 @samp{[a-z]} to match any lower case letter
4970 quotes the following character
4973 When a file name is matched with a wildcard, the wildcard characters
4974 will not match a @samp{/} character (used to separate directory names on
4975 Unix). A pattern consisting of a single @samp{*} character is an
4976 exception; it will always match any file name, whether it contains a
4977 @samp{/} or not. In a section name, the wildcard characters will match
4978 a @samp{/} character.
4980 File name wildcard patterns only match files which are explicitly
4981 specified on the command line or in an @code{INPUT} command. The linker
4982 does not search directories to expand wildcards.
4984 If a file name matches more than one wildcard pattern, or if a file name
4985 appears explicitly and is also matched by a wildcard pattern, the linker
4986 will use the first match in the linker script. For example, this
4987 sequence of input section descriptions is probably in error, because the
4988 @file{data.o} rule will not be used:
4990 .data : @{ *(.data) @}
4991 .data1 : @{ data.o(.data) @}
4994 @cindex SORT_BY_NAME
4995 Normally, the linker will place files and sections matched by wildcards
4996 in the order in which they are seen during the link. You can change
4997 this by using the @code{SORT_BY_NAME} keyword, which appears before a wildcard
4998 pattern in parentheses (e.g., @code{SORT_BY_NAME(.text*)}). When the
4999 @code{SORT_BY_NAME} keyword is used, the linker will sort the files or sections
5000 into ascending order by name before placing them in the output file.
5002 @cindex SORT_BY_ALIGNMENT
5003 @code{SORT_BY_ALIGNMENT} is similar to @code{SORT_BY_NAME}.
5004 @code{SORT_BY_ALIGNMENT} will sort sections into descending order of
5005 alignment before placing them in the output file. Placing larger
5006 alignments before smaller alignments can reduce the amount of padding
5009 @cindex SORT_BY_INIT_PRIORITY
5010 @code{SORT_BY_INIT_PRIORITY} is also similar to @code{SORT_BY_NAME}.
5011 @code{SORT_BY_INIT_PRIORITY} will sort sections into ascending
5012 numerical order of the GCC init_priority attribute encoded in the
5013 section name before placing them in the output file. In
5014 @code{.init_array.NNNNN} and @code{.fini_array.NNNNN}, @code{NNNNN} is
5015 the init_priority. In @code{.ctors.NNNNN} and @code{.dtors.NNNNN},
5016 @code{NNNNN} is 65535 minus the init_priority.
5019 @code{SORT} is an alias for @code{SORT_BY_NAME}.
5021 When there are nested section sorting commands in linker script, there
5022 can be at most 1 level of nesting for section sorting commands.
5026 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
5027 It will sort the input sections by name first, then by alignment if two
5028 sections have the same name.
5030 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
5031 It will sort the input sections by alignment first, then by name if two
5032 sections have the same alignment.
5034 @code{SORT_BY_NAME} (@code{SORT_BY_NAME} (wildcard section pattern)) is
5035 treated the same as @code{SORT_BY_NAME} (wildcard section pattern).
5037 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern))
5038 is treated the same as @code{SORT_BY_ALIGNMENT} (wildcard section pattern).
5040 All other nested section sorting commands are invalid.
5043 When both command-line section sorting option and linker script
5044 section sorting command are used, section sorting command always
5045 takes precedence over the command-line option.
5047 If the section sorting command in linker script isn't nested, the
5048 command-line option will make the section sorting command to be
5049 treated as nested sorting command.
5053 @code{SORT_BY_NAME} (wildcard section pattern ) with
5054 @option{--sort-sections alignment} is equivalent to
5055 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
5057 @code{SORT_BY_ALIGNMENT} (wildcard section pattern) with
5058 @option{--sort-section name} is equivalent to
5059 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
5062 If the section sorting command in linker script is nested, the
5063 command-line option will be ignored.
5066 @code{SORT_NONE} disables section sorting by ignoring the command-line
5067 section sorting option.
5069 If you ever get confused about where input sections are going, use the
5070 @samp{-M} linker option to generate a map file. The map file shows
5071 precisely how input sections are mapped to output sections.
5073 This example shows how wildcard patterns might be used to partition
5074 files. This linker script directs the linker to place all @samp{.text}
5075 sections in @samp{.text} and all @samp{.bss} sections in @samp{.bss}.
5076 The linker will place the @samp{.data} section from all files beginning
5077 with an upper case character in @samp{.DATA}; for all other files, the
5078 linker will place the @samp{.data} section in @samp{.data}.
5082 .text : @{ *(.text) @}
5083 .DATA : @{ [A-Z]*(.data) @}
5084 .data : @{ *(.data) @}
5085 .bss : @{ *(.bss) @}
5090 @node Input Section Common
5091 @subsubsection Input Section for Common Symbols
5092 @cindex common symbol placement
5093 @cindex uninitialized data placement
5094 A special notation is needed for common symbols, because in many object
5095 file formats common symbols do not have a particular input section. The
5096 linker treats common symbols as though they are in an input section
5097 named @samp{COMMON}.
5099 You may use file names with the @samp{COMMON} section just as with any
5100 other input sections. You can use this to place common symbols from a
5101 particular input file in one section while common symbols from other
5102 input files are placed in another section.
5104 In most cases, common symbols in input files will be placed in the
5105 @samp{.bss} section in the output file. For example:
5107 .bss @{ *(.bss) *(COMMON) @}
5110 @cindex scommon section
5111 @cindex small common symbols
5112 Some object file formats have more than one type of common symbol. For
5113 example, the MIPS ELF object file format distinguishes standard common
5114 symbols and small common symbols. In this case, the linker will use a
5115 different special section name for other types of common symbols. In
5116 the case of MIPS ELF, the linker uses @samp{COMMON} for standard common
5117 symbols and @samp{.scommon} for small common symbols. This permits you
5118 to map the different types of common symbols into memory at different
5122 You will sometimes see @samp{[COMMON]} in old linker scripts. This
5123 notation is now considered obsolete. It is equivalent to
5126 @node Input Section Keep
5127 @subsubsection Input Section and Garbage Collection
5129 @cindex garbage collection
5130 When link-time garbage collection is in use (@samp{--gc-sections}),
5131 it is often useful to mark sections that should not be eliminated.
5132 This is accomplished by surrounding an input section's wildcard entry
5133 with @code{KEEP()}, as in @code{KEEP(*(.init))} or
5134 @code{KEEP(SORT_BY_NAME(*)(.ctors))}.
5136 @node Input Section Example
5137 @subsubsection Input Section Example
5138 The following example is a complete linker script. It tells the linker
5139 to read all of the sections from file @file{all.o} and place them at the
5140 start of output section @samp{outputa} which starts at location
5141 @samp{0x10000}. All of section @samp{.input1} from file @file{foo.o}
5142 follows immediately, in the same output section. All of section
5143 @samp{.input2} from @file{foo.o} goes into output section
5144 @samp{outputb}, followed by section @samp{.input1} from @file{foo1.o}.
5145 All of the remaining @samp{.input1} and @samp{.input2} sections from any
5146 files are written to output section @samp{outputc}.
5174 If an output section's name is the same as the input section's name
5175 and is representable as a C identifier, then the linker will
5176 automatically @pxref{PROVIDE} two symbols: __start_SECNAME and
5177 __stop_SECNAME, where SECNAME is the name of the section. These
5178 indicate the start address and end address of the output section
5179 respectively. Note: most section names are not representable as
5180 C identifiers because they contain a @samp{.} character.
5182 @node Output Section Data
5183 @subsection Output Section Data
5185 @cindex section data
5186 @cindex output section data
5187 @kindex BYTE(@var{expression})
5188 @kindex SHORT(@var{expression})
5189 @kindex LONG(@var{expression})
5190 @kindex QUAD(@var{expression})
5191 @kindex SQUAD(@var{expression})
5192 You can include explicit bytes of data in an output section by using
5193 @code{BYTE}, @code{SHORT}, @code{LONG}, @code{QUAD}, or @code{SQUAD} as
5194 an output section command. Each keyword is followed by an expression in
5195 parentheses providing the value to store (@pxref{Expressions}). The
5196 value of the expression is stored at the current value of the location
5199 The @code{BYTE}, @code{SHORT}, @code{LONG}, and @code{QUAD} commands
5200 store one, two, four, and eight bytes (respectively). After storing the
5201 bytes, the location counter is incremented by the number of bytes
5204 For example, this will store the byte 1 followed by the four byte value
5205 of the symbol @samp{addr}:
5211 When using a 64 bit host or target, @code{QUAD} and @code{SQUAD} are the
5212 same; they both store an 8 byte, or 64 bit, value. When both host and
5213 target are 32 bits, an expression is computed as 32 bits. In this case
5214 @code{QUAD} stores a 32 bit value zero extended to 64 bits, and
5215 @code{SQUAD} stores a 32 bit value sign extended to 64 bits.
5217 If the object file format of the output file has an explicit endianness,
5218 which is the normal case, the value will be stored in that endianness.
5219 When the object file format does not have an explicit endianness, as is
5220 true of, for example, S-records, the value will be stored in the
5221 endianness of the first input object file.
5223 Note---these commands only work inside a section description and not
5224 between them, so the following will produce an error from the linker:
5226 SECTIONS @{@ .text : @{@ *(.text) @}@ LONG(1) .data : @{@ *(.data) @}@ @}@
5228 whereas this will work:
5230 SECTIONS @{@ .text : @{@ *(.text) ; LONG(1) @}@ .data : @{@ *(.data) @}@ @}@
5233 @kindex FILL(@var{expression})
5234 @cindex holes, filling
5235 @cindex unspecified memory
5236 You may use the @code{FILL} command to set the fill pattern for the
5237 current section. It is followed by an expression in parentheses. Any
5238 otherwise unspecified regions of memory within the section (for example,
5239 gaps left due to the required alignment of input sections) are filled
5240 with the value of the expression, repeated as
5241 necessary. A @code{FILL} statement covers memory locations after the
5242 point at which it occurs in the section definition; by including more
5243 than one @code{FILL} statement, you can have different fill patterns in
5244 different parts of an output section.
5246 This example shows how to fill unspecified regions of memory with the
5252 The @code{FILL} command is similar to the @samp{=@var{fillexp}} output
5253 section attribute, but it only affects the
5254 part of the section following the @code{FILL} command, rather than the
5255 entire section. If both are used, the @code{FILL} command takes
5256 precedence. @xref{Output Section Fill}, for details on the fill
5259 @node Output Section Keywords
5260 @subsection Output Section Keywords
5261 There are a couple of keywords which can appear as output section
5265 @kindex CREATE_OBJECT_SYMBOLS
5266 @cindex input filename symbols
5267 @cindex filename symbols
5268 @item CREATE_OBJECT_SYMBOLS
5269 The command tells the linker to create a symbol for each input file.
5270 The name of each symbol will be the name of the corresponding input
5271 file. The section of each symbol will be the output section in which
5272 the @code{CREATE_OBJECT_SYMBOLS} command appears.
5274 This is conventional for the a.out object file format. It is not
5275 normally used for any other object file format.
5277 @kindex CONSTRUCTORS
5278 @cindex C++ constructors, arranging in link
5279 @cindex constructors, arranging in link
5281 When linking using the a.out object file format, the linker uses an
5282 unusual set construct to support C++ global constructors and
5283 destructors. When linking object file formats which do not support
5284 arbitrary sections, such as ECOFF and XCOFF, the linker will
5285 automatically recognize C++ global constructors and destructors by name.
5286 For these object file formats, the @code{CONSTRUCTORS} command tells the
5287 linker to place constructor information in the output section where the
5288 @code{CONSTRUCTORS} command appears. The @code{CONSTRUCTORS} command is
5289 ignored for other object file formats.
5291 The symbol @w{@code{__CTOR_LIST__}} marks the start of the global
5292 constructors, and the symbol @w{@code{__CTOR_END__}} marks the end.
5293 Similarly, @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_END__}} mark
5294 the start and end of the global destructors. The
5295 first word in the list is the number of entries, followed by the address
5296 of each constructor or destructor, followed by a zero word. The
5297 compiler must arrange to actually run the code. For these object file
5298 formats @sc{gnu} C++ normally calls constructors from a subroutine
5299 @code{__main}; a call to @code{__main} is automatically inserted into
5300 the startup code for @code{main}. @sc{gnu} C++ normally runs
5301 destructors either by using @code{atexit}, or directly from the function
5304 For object file formats such as @code{COFF} or @code{ELF} which support
5305 arbitrary section names, @sc{gnu} C++ will normally arrange to put the
5306 addresses of global constructors and destructors into the @code{.ctors}
5307 and @code{.dtors} sections. Placing the following sequence into your
5308 linker script will build the sort of table which the @sc{gnu} C++
5309 runtime code expects to see.
5313 LONG((__CTOR_END__ - __CTOR_LIST__) / 4 - 2)
5318 LONG((__DTOR_END__ - __DTOR_LIST__) / 4 - 2)
5324 If you are using the @sc{gnu} C++ support for initialization priority,
5325 which provides some control over the order in which global constructors
5326 are run, you must sort the constructors at link time to ensure that they
5327 are executed in the correct order. When using the @code{CONSTRUCTORS}
5328 command, use @samp{SORT_BY_NAME(CONSTRUCTORS)} instead. When using the
5329 @code{.ctors} and @code{.dtors} sections, use @samp{*(SORT_BY_NAME(.ctors))} and
5330 @samp{*(SORT_BY_NAME(.dtors))} instead of just @samp{*(.ctors)} and
5333 Normally the compiler and linker will handle these issues automatically,
5334 and you will not need to concern yourself with them. However, you may
5335 need to consider this if you are using C++ and writing your own linker
5340 @node Output Section Discarding
5341 @subsection Output Section Discarding
5342 @cindex discarding sections
5343 @cindex sections, discarding
5344 @cindex removing sections
5345 The linker will not normally create output sections with no contents.
5346 This is for convenience when referring to input sections that may or
5347 may not be present in any of the input files. For example:
5349 .foo : @{ *(.foo) @}
5352 will only create a @samp{.foo} section in the output file if there is a
5353 @samp{.foo} section in at least one input file, and if the input
5354 sections are not all empty. Other link script directives that allocate
5355 space in an output section will also create the output section. So
5356 too will assignments to dot even if the assignment does not create
5357 space, except for @samp{. = 0}, @samp{. = . + 0}, @samp{. = sym},
5358 @samp{. = . + sym} and @samp{. = ALIGN (. != 0, expr, 1)} when
5359 @samp{sym} is an absolute symbol of value 0 defined in the script.
5360 This allows you to force output of an empty section with @samp{. = .}.
5362 The linker will ignore address assignments (@pxref{Output Section Address})
5363 on discarded output sections, except when the linker script defines
5364 symbols in the output section. In that case the linker will obey
5365 the address assignments, possibly advancing dot even though the
5366 section is discarded.
5369 The special output section name @samp{/DISCARD/} may be used to discard
5370 input sections. Any input sections which are assigned to an output
5371 section named @samp{/DISCARD/} are not included in the output file.
5373 This can be used to discard input sections marked with the ELF flag
5374 @code{SHF_GNU_RETAIN}, which would otherwise have been saved from linker
5377 Note, sections that match the @samp{/DISCARD/} output section will be
5378 discarded even if they are in an ELF section group which has other
5379 members which are not being discarded. This is deliberate.
5380 Discarding takes precedence over grouping.
5382 @node Output Section Attributes
5383 @subsection Output Section Attributes
5384 @cindex output section attributes
5385 We showed above that the full description of an output section looked
5390 @var{section} [@var{address}] [(@var{type})] :
5392 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
5393 [SUBALIGN(@var{subsection_align})]
5396 @var{output-section-command}
5397 @var{output-section-command}
5399 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}]
5403 We've already described @var{section}, @var{address}, and
5404 @var{output-section-command}. In this section we will describe the
5405 remaining section attributes.
5408 * Output Section Type:: Output section type
5409 * Output Section LMA:: Output section LMA
5410 * Forced Output Alignment:: Forced Output Alignment
5411 * Forced Input Alignment:: Forced Input Alignment
5412 * Output Section Constraint:: Output section constraint
5413 * Output Section Region:: Output section region
5414 * Output Section Phdr:: Output section phdr
5415 * Output Section Fill:: Output section fill
5418 @node Output Section Type
5419 @subsubsection Output Section Type
5420 Each output section may have a type. The type is a keyword in
5421 parentheses. The following types are defined:
5425 The section should be marked as not loadable, so that it will not be
5426 loaded into memory when the program is run.
5431 These type names are supported for backward compatibility, and are
5432 rarely used. They all have the same effect: the section should be
5433 marked as not allocatable, so that no memory is allocated for the
5434 section when the program is run.
5438 @cindex prevent unnecessary loading
5439 @cindex loading, preventing
5440 The linker normally sets the attributes of an output section based on
5441 the input sections which map into it. You can override this by using
5442 the section type. For example, in the script sample below, the
5443 @samp{ROM} section is addressed at memory location @samp{0} and does not
5444 need to be loaded when the program is run.
5448 ROM 0 (NOLOAD) : @{ @dots{} @}
5454 @node Output Section LMA
5455 @subsubsection Output Section LMA
5456 @kindex AT>@var{lma_region}
5457 @kindex AT(@var{lma})
5458 @cindex load address
5459 @cindex section load address
5460 Every section has a virtual address (VMA) and a load address (LMA); see
5461 @ref{Basic Script Concepts}. The virtual address is specified by the
5462 @pxref{Output Section Address} described earlier. The load address is
5463 specified by the @code{AT} or @code{AT>} keywords. Specifying a load
5464 address is optional.
5466 The @code{AT} keyword takes an expression as an argument. This
5467 specifies the exact load address of the section. The @code{AT>} keyword
5468 takes the name of a memory region as an argument. @xref{MEMORY}. The
5469 load address of the section is set to the next free address in the
5470 region, aligned to the section's alignment requirements.
5472 If neither @code{AT} nor @code{AT>} is specified for an allocatable
5473 section, the linker will use the following heuristic to determine the
5478 If the section has a specific VMA address, then this is used as
5479 the LMA address as well.
5482 If the section is not allocatable then its LMA is set to its VMA.
5485 Otherwise if a memory region can be found that is compatible
5486 with the current section, and this region contains at least one
5487 section, then the LMA is set so the difference between the
5488 VMA and LMA is the same as the difference between the VMA and LMA of
5489 the last section in the located region.
5492 If no memory regions have been declared then a default region
5493 that covers the entire address space is used in the previous step.
5496 If no suitable region could be found, or there was no previous
5497 section then the LMA is set equal to the VMA.
5500 @cindex ROM initialized data
5501 @cindex initialized data in ROM
5502 This feature is designed to make it easy to build a ROM image. For
5503 example, the following linker script creates three output sections: one
5504 called @samp{.text}, which starts at @code{0x1000}, one called
5505 @samp{.mdata}, which is loaded at the end of the @samp{.text} section
5506 even though its VMA is @code{0x2000}, and one called @samp{.bss} to hold
5507 uninitialized data at address @code{0x3000}. The symbol @code{_data} is
5508 defined with the value @code{0x2000}, which shows that the location
5509 counter holds the VMA value, not the LMA value.
5515 .text 0x1000 : @{ *(.text) _etext = . ; @}
5517 AT ( ADDR (.text) + SIZEOF (.text) )
5518 @{ _data = . ; *(.data); _edata = . ; @}
5520 @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@}
5525 The run-time initialization code for use with a program generated with
5526 this linker script would include something like the following, to copy
5527 the initialized data from the ROM image to its runtime address. Notice
5528 how this code takes advantage of the symbols defined by the linker
5533 extern char _etext, _data, _edata, _bstart, _bend;
5534 char *src = &_etext;
5537 /* ROM has data at end of text; copy it. */
5538 while (dst < &_edata)
5542 for (dst = &_bstart; dst< &_bend; dst++)
5547 @node Forced Output Alignment
5548 @subsubsection Forced Output Alignment
5549 @kindex ALIGN(@var{section_align})
5550 @cindex forcing output section alignment
5551 @cindex output section alignment
5552 You can increase an output section's alignment by using ALIGN. As an
5553 alternative you can enforce that the difference between the VMA and LMA remains
5554 intact throughout this output section with the ALIGN_WITH_INPUT attribute.
5556 @node Forced Input Alignment
5557 @subsubsection Forced Input Alignment
5558 @kindex SUBALIGN(@var{subsection_align})
5559 @cindex forcing input section alignment
5560 @cindex input section alignment
5561 You can force input section alignment within an output section by using
5562 SUBALIGN. The value specified overrides any alignment given by input
5563 sections, whether larger or smaller.
5565 @node Output Section Constraint
5566 @subsubsection Output Section Constraint
5569 @cindex constraints on output sections
5570 You can specify that an output section should only be created if all
5571 of its input sections are read-only or all of its input sections are
5572 read-write by using the keyword @code{ONLY_IF_RO} and
5573 @code{ONLY_IF_RW} respectively.
5575 @node Output Section Region
5576 @subsubsection Output Section Region
5577 @kindex >@var{region}
5578 @cindex section, assigning to memory region
5579 @cindex memory regions and sections
5580 You can assign a section to a previously defined region of memory by
5581 using @samp{>@var{region}}. @xref{MEMORY}.
5583 Here is a simple example:
5586 MEMORY @{ rom : ORIGIN = 0x1000, LENGTH = 0x1000 @}
5587 SECTIONS @{ ROM : @{ *(.text) @} >rom @}
5591 @node Output Section Phdr
5592 @subsubsection Output Section Phdr
5594 @cindex section, assigning to program header
5595 @cindex program headers and sections
5596 You can assign a section to a previously defined program segment by
5597 using @samp{:@var{phdr}}. @xref{PHDRS}. If a section is assigned to
5598 one or more segments, then all subsequent allocated sections will be
5599 assigned to those segments as well, unless they use an explicitly
5600 @code{:@var{phdr}} modifier. You can use @code{:NONE} to tell the
5601 linker to not put the section in any segment at all.
5603 Here is a simple example:
5606 PHDRS @{ text PT_LOAD ; @}
5607 SECTIONS @{ .text : @{ *(.text) @} :text @}
5611 @node Output Section Fill
5612 @subsubsection Output Section Fill
5613 @kindex =@var{fillexp}
5614 @cindex section fill pattern
5615 @cindex fill pattern, entire section
5616 You can set the fill pattern for an entire section by using
5617 @samp{=@var{fillexp}}. @var{fillexp} is an expression
5618 (@pxref{Expressions}). Any otherwise unspecified regions of memory
5619 within the output section (for example, gaps left due to the required
5620 alignment of input sections) will be filled with the value, repeated as
5621 necessary. If the fill expression is a simple hex number, ie. a string
5622 of hex digit starting with @samp{0x} and without a trailing @samp{k} or @samp{M}, then
5623 an arbitrarily long sequence of hex digits can be used to specify the
5624 fill pattern; Leading zeros become part of the pattern too. For all
5625 other cases, including extra parentheses or a unary @code{+}, the fill
5626 pattern is the four least significant bytes of the value of the
5627 expression. In all cases, the number is big-endian.
5629 You can also change the fill value with a @code{FILL} command in the
5630 output section commands; (@pxref{Output Section Data}).
5632 Here is a simple example:
5635 SECTIONS @{ .text : @{ *(.text) @} =0x90909090 @}
5639 @node Overlay Description
5640 @subsection Overlay Description
5643 An overlay description provides an easy way to describe sections which
5644 are to be loaded as part of a single memory image but are to be run at
5645 the same memory address. At run time, some sort of overlay manager will
5646 copy the overlaid sections in and out of the runtime memory address as
5647 required, perhaps by simply manipulating addressing bits. This approach
5648 can be useful, for example, when a certain region of memory is faster
5651 Overlays are described using the @code{OVERLAY} command. The
5652 @code{OVERLAY} command is used within a @code{SECTIONS} command, like an
5653 output section description. The full syntax of the @code{OVERLAY}
5654 command is as follows:
5657 OVERLAY [@var{start}] : [NOCROSSREFS] [AT ( @var{ldaddr} )]
5661 @var{output-section-command}
5662 @var{output-section-command}
5664 @} [:@var{phdr}@dots{}] [=@var{fill}]
5667 @var{output-section-command}
5668 @var{output-section-command}
5670 @} [:@var{phdr}@dots{}] [=@var{fill}]
5672 @} [>@var{region}] [:@var{phdr}@dots{}] [=@var{fill}] [,]
5676 Everything is optional except @code{OVERLAY} (a keyword), and each
5677 section must have a name (@var{secname1} and @var{secname2} above). The
5678 section definitions within the @code{OVERLAY} construct are identical to
5679 those within the general @code{SECTIONS} construct (@pxref{SECTIONS}),
5680 except that no addresses and no memory regions may be defined for
5681 sections within an @code{OVERLAY}.
5683 The comma at the end may be required if a @var{fill} is used and
5684 the next @var{sections-command} looks like a continuation of the expression.
5686 The sections are all defined with the same starting address. The load
5687 addresses of the sections are arranged such that they are consecutive in
5688 memory starting at the load address used for the @code{OVERLAY} as a
5689 whole (as with normal section definitions, the load address is optional,
5690 and defaults to the start address; the start address is also optional,
5691 and defaults to the current value of the location counter).
5693 If the @code{NOCROSSREFS} keyword is used, and there are any
5694 references among the sections, the linker will report an error. Since
5695 the sections all run at the same address, it normally does not make
5696 sense for one section to refer directly to another.
5697 @xref{Miscellaneous Commands, NOCROSSREFS}.
5699 For each section within the @code{OVERLAY}, the linker automatically
5700 provides two symbols. The symbol @code{__load_start_@var{secname}} is
5701 defined as the starting load address of the section. The symbol
5702 @code{__load_stop_@var{secname}} is defined as the final load address of
5703 the section. Any characters within @var{secname} which are not legal
5704 within C identifiers are removed. C (or assembler) code may use these
5705 symbols to move the overlaid sections around as necessary.
5707 At the end of the overlay, the value of the location counter is set to
5708 the start address of the overlay plus the size of the largest section.
5710 Here is an example. Remember that this would appear inside a
5711 @code{SECTIONS} construct.
5714 OVERLAY 0x1000 : AT (0x4000)
5716 .text0 @{ o1/*.o(.text) @}
5717 .text1 @{ o2/*.o(.text) @}
5722 This will define both @samp{.text0} and @samp{.text1} to start at
5723 address 0x1000. @samp{.text0} will be loaded at address 0x4000, and
5724 @samp{.text1} will be loaded immediately after @samp{.text0}. The
5725 following symbols will be defined if referenced: @code{__load_start_text0},
5726 @code{__load_stop_text0}, @code{__load_start_text1},
5727 @code{__load_stop_text1}.
5729 C code to copy overlay @code{.text1} into the overlay area might look
5734 extern char __load_start_text1, __load_stop_text1;
5735 memcpy ((char *) 0x1000, &__load_start_text1,
5736 &__load_stop_text1 - &__load_start_text1);
5740 Note that the @code{OVERLAY} command is just syntactic sugar, since
5741 everything it does can be done using the more basic commands. The above
5742 example could have been written identically as follows.
5746 .text0 0x1000 : AT (0x4000) @{ o1/*.o(.text) @}
5747 PROVIDE (__load_start_text0 = LOADADDR (.text0));
5748 PROVIDE (__load_stop_text0 = LOADADDR (.text0) + SIZEOF (.text0));
5749 .text1 0x1000 : AT (0x4000 + SIZEOF (.text0)) @{ o2/*.o(.text) @}
5750 PROVIDE (__load_start_text1 = LOADADDR (.text1));
5751 PROVIDE (__load_stop_text1 = LOADADDR (.text1) + SIZEOF (.text1));
5752 . = 0x1000 + MAX (SIZEOF (.text0), SIZEOF (.text1));
5757 @section MEMORY Command
5759 @cindex memory regions
5760 @cindex regions of memory
5761 @cindex allocating memory
5762 @cindex discontinuous memory
5763 The linker's default configuration permits allocation of all available
5764 memory. You can override this by using the @code{MEMORY} command.
5766 The @code{MEMORY} command describes the location and size of blocks of
5767 memory in the target. You can use it to describe which memory regions
5768 may be used by the linker, and which memory regions it must avoid. You
5769 can then assign sections to particular memory regions. The linker will
5770 set section addresses based on the memory regions, and will warn about
5771 regions that become too full. The linker will not shuffle sections
5772 around to fit into the available regions.
5774 A linker script may contain many uses of the @code{MEMORY} command,
5775 however, all memory blocks defined are treated as if they were
5776 specified inside a single @code{MEMORY} command. The syntax for
5782 @var{name} [(@var{attr})] : ORIGIN = @var{origin}, LENGTH = @var{len}
5788 The @var{name} is a name used in the linker script to refer to the
5789 region. The region name has no meaning outside of the linker script.
5790 Region names are stored in a separate name space, and will not conflict
5791 with symbol names, file names, or section names. Each memory region
5792 must have a distinct name within the @code{MEMORY} command. However you can
5793 add later alias names to existing memory regions with the @ref{REGION_ALIAS}
5796 @cindex memory region attributes
5797 The @var{attr} string is an optional list of attributes that specify
5798 whether to use a particular memory region for an input section which is
5799 not explicitly mapped in the linker script. As described in
5800 @ref{SECTIONS}, if you do not specify an output section for some input
5801 section, the linker will create an output section with the same name as
5802 the input section. If you define region attributes, the linker will use
5803 them to select the memory region for the output section that it creates.
5805 The @var{attr} string must consist only of the following characters:
5820 Invert the sense of any of the attributes that follow
5823 If an unmapped section matches any of the listed attributes other than
5824 @samp{!}, it will be placed in the memory region. The @samp{!}
5825 attribute reverses the test for the characters that follow, so that an
5826 unmapped section will be placed in the memory region only if it does
5827 not match any of the attributes listed afterwards. Thus an attribute
5828 string of @samp{RW!X} will match any unmapped section that has either
5829 or both of the @samp{R} and @samp{W} attributes, but only as long as
5830 the section does not also have the @samp{X} attribute.
5835 The @var{origin} is an numerical expression for the start address of
5836 the memory region. The expression must evaluate to a constant and it
5837 cannot involve any symbols. The keyword @code{ORIGIN} may be
5838 abbreviated to @code{org} or @code{o} (but not, for example,
5844 The @var{len} is an expression for the size in bytes of the memory
5845 region. As with the @var{origin} expression, the expression must
5846 be numerical only and must evaluate to a constant. The keyword
5847 @code{LENGTH} may be abbreviated to @code{len} or @code{l}.
5849 In the following example, we specify that there are two memory regions
5850 available for allocation: one starting at @samp{0} for 256 kilobytes,
5851 and the other starting at @samp{0x40000000} for four megabytes. The
5852 linker will place into the @samp{rom} memory region every section which
5853 is not explicitly mapped into a memory region, and is either read-only
5854 or executable. The linker will place other sections which are not
5855 explicitly mapped into a memory region into the @samp{ram} memory
5862 rom (rx) : ORIGIN = 0, LENGTH = 256K
5863 ram (!rx) : org = 0x40000000, l = 4M
5868 Once you define a memory region, you can direct the linker to place
5869 specific output sections into that memory region by using the
5870 @samp{>@var{region}} output section attribute. For example, if you have
5871 a memory region named @samp{mem}, you would use @samp{>mem} in the
5872 output section definition. @xref{Output Section Region}. If no address
5873 was specified for the output section, the linker will set the address to
5874 the next available address within the memory region. If the combined
5875 output sections directed to a memory region are too large for the
5876 region, the linker will issue an error message.
5878 It is possible to access the origin and length of a memory in an
5879 expression via the @code{ORIGIN(@var{memory})} and
5880 @code{LENGTH(@var{memory})} functions:
5884 _fstack = ORIGIN(ram) + LENGTH(ram) - 4;
5889 @section PHDRS Command
5891 @cindex program headers
5892 @cindex ELF program headers
5893 @cindex program segments
5894 @cindex segments, ELF
5895 The ELF object file format uses @dfn{program headers}, also knows as
5896 @dfn{segments}. The program headers describe how the program should be
5897 loaded into memory. You can print them out by using the @code{objdump}
5898 program with the @samp{-p} option.
5900 When you run an ELF program on a native ELF system, the system loader
5901 reads the program headers in order to figure out how to load the
5902 program. This will only work if the program headers are set correctly.
5903 This manual does not describe the details of how the system loader
5904 interprets program headers; for more information, see the ELF ABI.
5906 The linker will create reasonable program headers by default. However,
5907 in some cases, you may need to specify the program headers more
5908 precisely. You may use the @code{PHDRS} command for this purpose. When
5909 the linker sees the @code{PHDRS} command in the linker script, it will
5910 not create any program headers other than the ones specified.
5912 The linker only pays attention to the @code{PHDRS} command when
5913 generating an ELF output file. In other cases, the linker will simply
5914 ignore @code{PHDRS}.
5916 This is the syntax of the @code{PHDRS} command. The words @code{PHDRS},
5917 @code{FILEHDR}, @code{AT}, and @code{FLAGS} are keywords.
5923 @var{name} @var{type} [ FILEHDR ] [ PHDRS ] [ AT ( @var{address} ) ]
5924 [ FLAGS ( @var{flags} ) ] ;
5929 The @var{name} is used only for reference in the @code{SECTIONS} command
5930 of the linker script. It is not put into the output file. Program
5931 header names are stored in a separate name space, and will not conflict
5932 with symbol names, file names, or section names. Each program header
5933 must have a distinct name. The headers are processed in order and it
5934 is usual for them to map to sections in ascending load address order.
5936 Certain program header types describe segments of memory which the
5937 system loader will load from the file. In the linker script, you
5938 specify the contents of these segments by placing allocatable output
5939 sections in the segments. You use the @samp{:@var{phdr}} output section
5940 attribute to place a section in a particular segment. @xref{Output
5943 It is normal to put certain sections in more than one segment. This
5944 merely implies that one segment of memory contains another. You may
5945 repeat @samp{:@var{phdr}}, using it once for each segment which should
5946 contain the section.
5948 If you place a section in one or more segments using @samp{:@var{phdr}},
5949 then the linker will place all subsequent allocatable sections which do
5950 not specify @samp{:@var{phdr}} in the same segments. This is for
5951 convenience, since generally a whole set of contiguous sections will be
5952 placed in a single segment. You can use @code{:NONE} to override the
5953 default segment and tell the linker to not put the section in any
5958 You may use the @code{FILEHDR} and @code{PHDRS} keywords after
5959 the program header type to further describe the contents of the segment.
5960 The @code{FILEHDR} keyword means that the segment should include the ELF
5961 file header. The @code{PHDRS} keyword means that the segment should
5962 include the ELF program headers themselves. If applied to a loadable
5963 segment (@code{PT_LOAD}), all prior loadable segments must have one of
5966 The @var{type} may be one of the following. The numbers indicate the
5967 value of the keyword.
5970 @item @code{PT_NULL} (0)
5971 Indicates an unused program header.
5973 @item @code{PT_LOAD} (1)
5974 Indicates that this program header describes a segment to be loaded from
5977 @item @code{PT_DYNAMIC} (2)
5978 Indicates a segment where dynamic linking information can be found.
5980 @item @code{PT_INTERP} (3)
5981 Indicates a segment where the name of the program interpreter may be
5984 @item @code{PT_NOTE} (4)
5985 Indicates a segment holding note information.
5987 @item @code{PT_SHLIB} (5)
5988 A reserved program header type, defined but not specified by the ELF
5991 @item @code{PT_PHDR} (6)
5992 Indicates a segment where the program headers may be found.
5994 @item @code{PT_TLS} (7)
5995 Indicates a segment containing thread local storage.
5997 @item @var{expression}
5998 An expression giving the numeric type of the program header. This may
5999 be used for types not defined above.
6002 You can specify that a segment should be loaded at a particular address
6003 in memory by using an @code{AT} expression. This is identical to the
6004 @code{AT} command used as an output section attribute (@pxref{Output
6005 Section LMA}). The @code{AT} command for a program header overrides the
6006 output section attribute.
6008 The linker will normally set the segment flags based on the sections
6009 which comprise the segment. You may use the @code{FLAGS} keyword to
6010 explicitly specify the segment flags. The value of @var{flags} must be
6011 an integer. It is used to set the @code{p_flags} field of the program
6014 Here is an example of @code{PHDRS}. This shows a typical set of program
6015 headers used on a native ELF system.
6021 headers PT_PHDR PHDRS ;
6023 text PT_LOAD FILEHDR PHDRS ;
6025 dynamic PT_DYNAMIC ;
6031 .interp : @{ *(.interp) @} :text :interp
6032 .text : @{ *(.text) @} :text
6033 .rodata : @{ *(.rodata) @} /* defaults to :text */
6035 . = . + 0x1000; /* move to a new page in memory */
6036 .data : @{ *(.data) @} :data
6037 .dynamic : @{ *(.dynamic) @} :data :dynamic
6044 @section VERSION Command
6045 @kindex VERSION @{script text@}
6046 @cindex symbol versions
6047 @cindex version script
6048 @cindex versions of symbols
6049 The linker supports symbol versions when using ELF. Symbol versions are
6050 only useful when using shared libraries. The dynamic linker can use
6051 symbol versions to select a specific version of a function when it runs
6052 a program that may have been linked against an earlier version of the
6055 You can include a version script directly in the main linker script, or
6056 you can supply the version script as an implicit linker script. You can
6057 also use the @samp{--version-script} linker option.
6059 The syntax of the @code{VERSION} command is simply
6061 VERSION @{ version-script-commands @}
6064 The format of the version script commands is identical to that used by
6065 Sun's linker in Solaris 2.5. The version script defines a tree of
6066 version nodes. You specify the node names and interdependencies in the
6067 version script. You can specify which symbols are bound to which
6068 version nodes, and you can reduce a specified set of symbols to local
6069 scope so that they are not globally visible outside of the shared
6072 The easiest way to demonstrate the version script language is with a few
6098 This example version script defines three version nodes. The first
6099 version node defined is @samp{VERS_1.1}; it has no other dependencies.
6100 The script binds the symbol @samp{foo1} to @samp{VERS_1.1}. It reduces
6101 a number of symbols to local scope so that they are not visible outside
6102 of the shared library; this is done using wildcard patterns, so that any
6103 symbol whose name begins with @samp{old}, @samp{original}, or @samp{new}
6104 is matched. The wildcard patterns available are the same as those used
6105 in the shell when matching filenames (also known as ``globbing'').
6106 However, if you specify the symbol name inside double quotes, then the
6107 name is treated as literal, rather than as a glob pattern.
6109 Next, the version script defines node @samp{VERS_1.2}. This node
6110 depends upon @samp{VERS_1.1}. The script binds the symbol @samp{foo2}
6111 to the version node @samp{VERS_1.2}.
6113 Finally, the version script defines node @samp{VERS_2.0}. This node
6114 depends upon @samp{VERS_1.2}. The scripts binds the symbols @samp{bar1}
6115 and @samp{bar2} are bound to the version node @samp{VERS_2.0}.
6117 When the linker finds a symbol defined in a library which is not
6118 specifically bound to a version node, it will effectively bind it to an
6119 unspecified base version of the library. You can bind all otherwise
6120 unspecified symbols to a given version node by using @samp{global: *;}
6121 somewhere in the version script. Note that it's slightly crazy to use
6122 wildcards in a global spec except on the last version node. Global
6123 wildcards elsewhere run the risk of accidentally adding symbols to the
6124 set exported for an old version. That's wrong since older versions
6125 ought to have a fixed set of symbols.
6127 The names of the version nodes have no specific meaning other than what
6128 they might suggest to the person reading them. The @samp{2.0} version
6129 could just as well have appeared in between @samp{1.1} and @samp{1.2}.
6130 However, this would be a confusing way to write a version script.
6132 Node name can be omitted, provided it is the only version node
6133 in the version script. Such version script doesn't assign any versions to
6134 symbols, only selects which symbols will be globally visible out and which
6138 @{ global: foo; bar; local: *; @};
6141 When you link an application against a shared library that has versioned
6142 symbols, the application itself knows which version of each symbol it
6143 requires, and it also knows which version nodes it needs from each
6144 shared library it is linked against. Thus at runtime, the dynamic
6145 loader can make a quick check to make sure that the libraries you have
6146 linked against do in fact supply all of the version nodes that the
6147 application will need to resolve all of the dynamic symbols. In this
6148 way it is possible for the dynamic linker to know with certainty that
6149 all external symbols that it needs will be resolvable without having to
6150 search for each symbol reference.
6152 The symbol versioning is in effect a much more sophisticated way of
6153 doing minor version checking that SunOS does. The fundamental problem
6154 that is being addressed here is that typically references to external
6155 functions are bound on an as-needed basis, and are not all bound when
6156 the application starts up. If a shared library is out of date, a
6157 required interface may be missing; when the application tries to use
6158 that interface, it may suddenly and unexpectedly fail. With symbol
6159 versioning, the user will get a warning when they start their program if
6160 the libraries being used with the application are too old.
6162 There are several GNU extensions to Sun's versioning approach. The
6163 first of these is the ability to bind a symbol to a version node in the
6164 source file where the symbol is defined instead of in the versioning
6165 script. This was done mainly to reduce the burden on the library
6166 maintainer. You can do this by putting something like:
6168 __asm__(".symver original_foo,foo@@VERS_1.1");
6171 in the C source file. This renames the function @samp{original_foo} to
6172 be an alias for @samp{foo} bound to the version node @samp{VERS_1.1}.
6173 The @samp{local:} directive can be used to prevent the symbol
6174 @samp{original_foo} from being exported. A @samp{.symver} directive
6175 takes precedence over a version script.
6177 The second GNU extension is to allow multiple versions of the same
6178 function to appear in a given shared library. In this way you can make
6179 an incompatible change to an interface without increasing the major
6180 version number of the shared library, while still allowing applications
6181 linked against the old interface to continue to function.
6183 To do this, you must use multiple @samp{.symver} directives in the
6184 source file. Here is an example:
6187 __asm__(".symver original_foo,foo@@");
6188 __asm__(".symver old_foo,foo@@VERS_1.1");
6189 __asm__(".symver old_foo1,foo@@VERS_1.2");
6190 __asm__(".symver new_foo,foo@@@@VERS_2.0");
6193 In this example, @samp{foo@@} represents the symbol @samp{foo} bound to the
6194 unspecified base version of the symbol. The source file that contains this
6195 example would define 4 C functions: @samp{original_foo}, @samp{old_foo},
6196 @samp{old_foo1}, and @samp{new_foo}.
6198 When you have multiple definitions of a given symbol, there needs to be
6199 some way to specify a default version to which external references to
6200 this symbol will be bound. You can do this with the
6201 @samp{foo@@@@VERS_2.0} type of @samp{.symver} directive. You can only
6202 declare one version of a symbol as the default in this manner; otherwise
6203 you would effectively have multiple definitions of the same symbol.
6205 If you wish to bind a reference to a specific version of the symbol
6206 within the shared library, you can use the aliases of convenience
6207 (i.e., @samp{old_foo}), or you can use the @samp{.symver} directive to
6208 specifically bind to an external version of the function in question.
6210 You can also specify the language in the version script:
6213 VERSION extern "lang" @{ version-script-commands @}
6216 The supported @samp{lang}s are @samp{C}, @samp{C++}, and @samp{Java}.
6217 The linker will iterate over the list of symbols at the link time and
6218 demangle them according to @samp{lang} before matching them to the
6219 patterns specified in @samp{version-script-commands}. The default
6220 @samp{lang} is @samp{C}.
6222 Demangled names may contains spaces and other special characters. As
6223 described above, you can use a glob pattern to match demangled names,
6224 or you can use a double-quoted string to match the string exactly. In
6225 the latter case, be aware that minor differences (such as differing
6226 whitespace) between the version script and the demangler output will
6227 cause a mismatch. As the exact string generated by the demangler
6228 might change in the future, even if the mangled name does not, you
6229 should check that all of your version directives are behaving as you
6230 expect when you upgrade.
6233 @section Expressions in Linker Scripts
6236 The syntax for expressions in the linker script language is identical to
6237 that of C expressions. All expressions are evaluated as integers. All
6238 expressions are evaluated in the same size, which is 32 bits if both the
6239 host and target are 32 bits, and is otherwise 64 bits.
6241 You can use and set symbol values in expressions.
6243 The linker defines several special purpose builtin functions for use in
6247 * Constants:: Constants
6248 * Symbolic Constants:: Symbolic constants
6249 * Symbols:: Symbol Names
6250 * Orphan Sections:: Orphan Sections
6251 * Location Counter:: The Location Counter
6252 * Operators:: Operators
6253 * Evaluation:: Evaluation
6254 * Expression Section:: The Section of an Expression
6255 * Builtin Functions:: Builtin Functions
6259 @subsection Constants
6260 @cindex integer notation
6261 @cindex constants in linker scripts
6262 All constants are integers.
6264 As in C, the linker considers an integer beginning with @samp{0} to be
6265 octal, and an integer beginning with @samp{0x} or @samp{0X} to be
6266 hexadecimal. Alternatively the linker accepts suffixes of @samp{h} or
6267 @samp{H} for hexadecimal, @samp{o} or @samp{O} for octal, @samp{b} or
6268 @samp{B} for binary and @samp{d} or @samp{D} for decimal. Any integer
6269 value without a prefix or a suffix is considered to be decimal.
6271 @cindex scaled integers
6272 @cindex K and M integer suffixes
6273 @cindex M and K integer suffixes
6274 @cindex suffixes for integers
6275 @cindex integer suffixes
6276 In addition, you can use the suffixes @code{K} and @code{M} to scale a
6280 @c END TEXI2ROFF-KILL
6281 @code{1024} or @code{1024*1024}
6285 ${\rm 1024}$ or ${\rm 1024}^2$
6287 @c END TEXI2ROFF-KILL
6288 respectively. For example, the following
6289 all refer to the same quantity:
6298 Note - the @code{K} and @code{M} suffixes cannot be used in
6299 conjunction with the base suffixes mentioned above.
6301 @node Symbolic Constants
6302 @subsection Symbolic Constants
6303 @cindex symbolic constants
6305 It is possible to refer to target-specific constants via the use of
6306 the @code{CONSTANT(@var{name})} operator, where @var{name} is one of:
6311 The target's maximum page size.
6313 @item COMMONPAGESIZE
6314 @kindex COMMONPAGESIZE
6315 The target's default page size.
6321 .text ALIGN (CONSTANT (MAXPAGESIZE)) : @{ *(.text) @}
6324 will create a text section aligned to the largest page boundary
6325 supported by the target.
6328 @subsection Symbol Names
6329 @cindex symbol names
6331 @cindex quoted symbol names
6333 Unless quoted, symbol names start with a letter, underscore, or period
6334 and may include letters, digits, underscores, periods, and hyphens.
6335 Unquoted symbol names must not conflict with any keywords. You can
6336 specify a symbol which contains odd characters or has the same name as a
6337 keyword by surrounding the symbol name in double quotes:
6340 "with a space" = "also with a space" + 10;
6343 Since symbols can contain many non-alphabetic characters, it is safest
6344 to delimit symbols with spaces. For example, @samp{A-B} is one symbol,
6345 whereas @samp{A - B} is an expression involving subtraction.
6347 @node Orphan Sections
6348 @subsection Orphan Sections
6350 Orphan sections are sections present in the input files which
6351 are not explicitly placed into the output file by the linker
6352 script. The linker will still copy these sections into the
6353 output file by either finding, or creating a suitable output section
6354 in which to place the orphaned input section.
6356 If the name of an orphaned input section exactly matches the name of
6357 an existing output section, then the orphaned input section will be
6358 placed at the end of that output section.
6360 If there is no output section with a matching name then new output
6361 sections will be created. Each new output section will have the same
6362 name as the orphan section placed within it. If there are multiple
6363 orphan sections with the same name, these will all be combined into
6364 one new output section.
6366 If new output sections are created to hold orphaned input sections,
6367 then the linker must decide where to place these new output sections
6368 in relation to existing output sections. On most modern targets, the
6369 linker attempts to place orphan sections after sections of the same
6370 attribute, such as code vs data, loadable vs non-loadable, etc. If no
6371 sections with matching attributes are found, or your target lacks this
6372 support, the orphan section is placed at the end of the file.
6374 The command-line options @samp{--orphan-handling} and @samp{--unique}
6375 (@pxref{Options,,Command-line Options}) can be used to control which
6376 output sections an orphan is placed in.
6378 @node Location Counter
6379 @subsection The Location Counter
6382 @cindex location counter
6383 @cindex current output location
6384 The special linker variable @dfn{dot} @samp{.} always contains the
6385 current output location counter. Since the @code{.} always refers to a
6386 location in an output section, it may only appear in an expression
6387 within a @code{SECTIONS} command. The @code{.} symbol may appear
6388 anywhere that an ordinary symbol is allowed in an expression.
6391 Assigning a value to @code{.} will cause the location counter to be
6392 moved. This may be used to create holes in the output section. The
6393 location counter may not be moved backwards inside an output section,
6394 and may not be moved backwards outside of an output section if so
6395 doing creates areas with overlapping LMAs.
6411 In the previous example, the @samp{.text} section from @file{file1} is
6412 located at the beginning of the output section @samp{output}. It is
6413 followed by a 1000 byte gap. Then the @samp{.text} section from
6414 @file{file2} appears, also with a 1000 byte gap following before the
6415 @samp{.text} section from @file{file3}. The notation @samp{= 0x12345678}
6416 specifies what data to write in the gaps (@pxref{Output Section Fill}).
6418 @cindex dot inside sections
6419 Note: @code{.} actually refers to the byte offset from the start of the
6420 current containing object. Normally this is the @code{SECTIONS}
6421 statement, whose start address is 0, hence @code{.} can be used as an
6422 absolute address. If @code{.} is used inside a section description
6423 however, it refers to the byte offset from the start of that section,
6424 not an absolute address. Thus in a script like this:
6442 The @samp{.text} section will be assigned a starting address of 0x100
6443 and a size of exactly 0x200 bytes, even if there is not enough data in
6444 the @samp{.text} input sections to fill this area. (If there is too
6445 much data, an error will be produced because this would be an attempt to
6446 move @code{.} backwards). The @samp{.data} section will start at 0x500
6447 and it will have an extra 0x600 bytes worth of space after the end of
6448 the values from the @samp{.data} input sections and before the end of
6449 the @samp{.data} output section itself.
6451 @cindex dot outside sections
6452 Setting symbols to the value of the location counter outside of an
6453 output section statement can result in unexpected values if the linker
6454 needs to place orphan sections. For example, given the following:
6460 .text: @{ *(.text) @}
6464 .data: @{ *(.data) @}
6469 If the linker needs to place some input section, e.g. @code{.rodata},
6470 not mentioned in the script, it might choose to place that section
6471 between @code{.text} and @code{.data}. You might think the linker
6472 should place @code{.rodata} on the blank line in the above script, but
6473 blank lines are of no particular significance to the linker. As well,
6474 the linker doesn't associate the above symbol names with their
6475 sections. Instead, it assumes that all assignments or other
6476 statements belong to the previous output section, except for the
6477 special case of an assignment to @code{.}. I.e., the linker will
6478 place the orphan @code{.rodata} section as if the script was written
6485 .text: @{ *(.text) @}
6489 .rodata: @{ *(.rodata) @}
6490 .data: @{ *(.data) @}
6495 This may or may not be the script author's intention for the value of
6496 @code{start_of_data}. One way to influence the orphan section
6497 placement is to assign the location counter to itself, as the linker
6498 assumes that an assignment to @code{.} is setting the start address of
6499 a following output section and thus should be grouped with that
6500 section. So you could write:
6506 .text: @{ *(.text) @}
6511 .data: @{ *(.data) @}
6516 Now, the orphan @code{.rodata} section will be placed between
6517 @code{end_of_text} and @code{start_of_data}.
6521 @subsection Operators
6522 @cindex operators for arithmetic
6523 @cindex arithmetic operators
6524 @cindex precedence in expressions
6525 The linker recognizes the standard C set of arithmetic operators, with
6526 the standard bindings and precedence levels:
6529 @c END TEXI2ROFF-KILL
6531 precedence associativity Operators Notes
6537 5 left == != > < <= >=
6543 11 right &= += -= *= /= (2)
6547 (1) Prefix operators
6548 (2) @xref{Assignments}.
6552 \vskip \baselineskip
6553 %"lispnarrowing" is the extra indent used generally for smallexample
6554 \hskip\lispnarrowing\vbox{\offinterlineskip
6557 {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr
6558 height2pt&\omit&&\omit&&\omit&\cr
6559 &Precedence&& Associativity &&{\rm Operators}&\cr
6560 height2pt&\omit&&\omit&&\omit&\cr
6562 height2pt&\omit&&\omit&&\omit&\cr
6564 % '176 is tilde, '~' in tt font
6565 &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr
6566 &2&&left&&* / \%&\cr
6569 &5&&left&&== != > < <= >=&\cr
6572 &8&&left&&{\&\&}&\cr
6575 &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr
6577 height2pt&\omit&&\omit&&\omit&\cr}
6582 @obeylines@parskip=0pt@parindent=0pt
6583 @dag@quad Prefix operators.
6584 @ddag@quad @xref{Assignments}.
6587 @c END TEXI2ROFF-KILL
6590 @subsection Evaluation
6591 @cindex lazy evaluation
6592 @cindex expression evaluation order
6593 The linker evaluates expressions lazily. It only computes the value of
6594 an expression when absolutely necessary.
6596 The linker needs some information, such as the value of the start
6597 address of the first section, and the origins and lengths of memory
6598 regions, in order to do any linking at all. These values are computed
6599 as soon as possible when the linker reads in the linker script.
6601 However, other values (such as symbol values) are not known or needed
6602 until after storage allocation. Such values are evaluated later, when
6603 other information (such as the sizes of output sections) is available
6604 for use in the symbol assignment expression.
6606 The sizes of sections cannot be known until after allocation, so
6607 assignments dependent upon these are not performed until after
6610 Some expressions, such as those depending upon the location counter
6611 @samp{.}, must be evaluated during section allocation.
6613 If the result of an expression is required, but the value is not
6614 available, then an error results. For example, a script like the
6620 .text 9+this_isnt_constant :
6626 will cause the error message @samp{non constant expression for initial
6629 @node Expression Section
6630 @subsection The Section of an Expression
6631 @cindex expression sections
6632 @cindex absolute expressions
6633 @cindex relative expressions
6634 @cindex absolute and relocatable symbols
6635 @cindex relocatable and absolute symbols
6636 @cindex symbols, relocatable and absolute
6637 Addresses and symbols may be section relative, or absolute. A section
6638 relative symbol is relocatable. If you request relocatable output
6639 using the @samp{-r} option, a further link operation may change the
6640 value of a section relative symbol. On the other hand, an absolute
6641 symbol will retain the same value throughout any further link
6644 Some terms in linker expressions are addresses. This is true of
6645 section relative symbols and for builtin functions that return an
6646 address, such as @code{ADDR}, @code{LOADADDR}, @code{ORIGIN} and
6647 @code{SEGMENT_START}. Other terms are simply numbers, or are builtin
6648 functions that return a non-address value, such as @code{LENGTH}.
6649 One complication is that unless you set @code{LD_FEATURE ("SANE_EXPR")}
6650 (@pxref{Miscellaneous Commands}), numbers and absolute symbols are treated
6651 differently depending on their location, for compatibility with older
6652 versions of @code{ld}. Expressions appearing outside an output
6653 section definition treat all numbers as absolute addresses.
6654 Expressions appearing inside an output section definition treat
6655 absolute symbols as numbers. If @code{LD_FEATURE ("SANE_EXPR")} is
6656 given, then absolute symbols and numbers are simply treated as numbers
6659 In the following simple example,
6666 __executable_start = 0x100;
6670 __data_start = 0x10;
6678 both @code{.} and @code{__executable_start} are set to the absolute
6679 address 0x100 in the first two assignments, then both @code{.} and
6680 @code{__data_start} are set to 0x10 relative to the @code{.data}
6681 section in the second two assignments.
6683 For expressions involving numbers, relative addresses and absolute
6684 addresses, ld follows these rules to evaluate terms:
6688 Unary operations on an absolute address or number, and binary
6689 operations on two absolute addresses or two numbers, or between one
6690 absolute address and a number, apply the operator to the value(s).
6692 Unary operations on a relative address, and binary operations on two
6693 relative addresses in the same section or between one relative address
6694 and a number, apply the operator to the offset part of the address(es).
6696 Other binary operations, that is, between two relative addresses not
6697 in the same section, or between a relative address and an absolute
6698 address, first convert any non-absolute term to an absolute address
6699 before applying the operator.
6702 The result section of each sub-expression is as follows:
6706 An operation involving only numbers results in a number.
6708 The result of comparisons, @samp{&&} and @samp{||} is also a number.
6710 The result of other binary arithmetic and logical operations on two
6711 relative addresses in the same section or two absolute addresses
6712 (after above conversions) is also a number when
6713 @code{LD_FEATURE ("SANE_EXPR")} or inside an output section definition
6714 but an absolute address otherwise.
6716 The result of other operations on relative addresses or one
6717 relative address and a number, is a relative address in the same
6718 section as the relative operand(s).
6720 The result of other operations on absolute addresses (after above
6721 conversions) is an absolute address.
6724 You can use the builtin function @code{ABSOLUTE} to force an expression
6725 to be absolute when it would otherwise be relative. For example, to
6726 create an absolute symbol set to the address of the end of the output
6727 section @samp{.data}:
6731 .data : @{ *(.data) _edata = ABSOLUTE(.); @}
6735 If @samp{ABSOLUTE} were not used, @samp{_edata} would be relative to the
6736 @samp{.data} section.
6738 Using @code{LOADADDR} also forces an expression absolute, since this
6739 particular builtin function returns an absolute address.
6741 @node Builtin Functions
6742 @subsection Builtin Functions
6743 @cindex functions in expressions
6744 The linker script language includes a number of builtin functions for
6745 use in linker script expressions.
6748 @item ABSOLUTE(@var{exp})
6749 @kindex ABSOLUTE(@var{exp})
6750 @cindex expression, absolute
6751 Return the absolute (non-relocatable, as opposed to non-negative) value
6752 of the expression @var{exp}. Primarily useful to assign an absolute
6753 value to a symbol within a section definition, where symbol values are
6754 normally section relative. @xref{Expression Section}.
6756 @item ADDR(@var{section})
6757 @kindex ADDR(@var{section})
6758 @cindex section address in expression
6759 Return the address (VMA) of the named @var{section}. Your
6760 script must previously have defined the location of that section. In
6761 the following example, @code{start_of_output_1}, @code{symbol_1} and
6762 @code{symbol_2} are assigned equivalent values, except that
6763 @code{symbol_1} will be relative to the @code{.output1} section while
6764 the other two will be absolute:
6770 start_of_output_1 = ABSOLUTE(.);
6775 symbol_1 = ADDR(.output1);
6776 symbol_2 = start_of_output_1;
6782 @item ALIGN(@var{align})
6783 @itemx ALIGN(@var{exp},@var{align})
6784 @kindex ALIGN(@var{align})
6785 @kindex ALIGN(@var{exp},@var{align})
6786 @cindex round up location counter
6787 @cindex align location counter
6788 @cindex round up expression
6789 @cindex align expression
6790 Return the location counter (@code{.}) or arbitrary expression aligned
6791 to the next @var{align} boundary. The single operand @code{ALIGN}
6792 doesn't change the value of the location counter---it just does
6793 arithmetic on it. The two operand @code{ALIGN} allows an arbitrary
6794 expression to be aligned upwards (@code{ALIGN(@var{align})} is
6795 equivalent to @code{ALIGN(ABSOLUTE(.), @var{align})}).
6797 Here is an example which aligns the output @code{.data} section to the
6798 next @code{0x2000} byte boundary after the preceding section and sets a
6799 variable within the section to the next @code{0x8000} boundary after the
6804 .data ALIGN(0x2000): @{
6806 variable = ALIGN(0x8000);
6812 The first use of @code{ALIGN} in this example specifies the location of
6813 a section because it is used as the optional @var{address} attribute of
6814 a section definition (@pxref{Output Section Address}). The second use
6815 of @code{ALIGN} is used to defines the value of a symbol.
6817 The builtin function @code{NEXT} is closely related to @code{ALIGN}.
6819 @item ALIGNOF(@var{section})
6820 @kindex ALIGNOF(@var{section})
6821 @cindex section alignment
6822 Return the alignment in bytes of the named @var{section}, if that section has
6823 been allocated. If the section has not been allocated when this is
6824 evaluated, the linker will report an error. In the following example,
6825 the alignment of the @code{.output} section is stored as the first
6826 value in that section.
6831 LONG (ALIGNOF (.output))
6838 @item BLOCK(@var{exp})
6839 @kindex BLOCK(@var{exp})
6840 This is a synonym for @code{ALIGN}, for compatibility with older linker
6841 scripts. It is most often seen when setting the address of an output
6844 @item DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6845 @kindex DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6846 This is equivalent to either
6848 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - 1)))
6852 (ALIGN(@var{maxpagesize})
6853 + ((. + @var{commonpagesize} - 1) & (@var{maxpagesize} - @var{commonpagesize})))
6856 depending on whether the latter uses fewer @var{commonpagesize} sized pages
6857 for the data segment (area between the result of this expression and
6858 @code{DATA_SEGMENT_END}) than the former or not.
6859 If the latter form is used, it means @var{commonpagesize} bytes of runtime
6860 memory will be saved at the expense of up to @var{commonpagesize} wasted
6861 bytes in the on-disk file.
6863 This expression can only be used directly in @code{SECTIONS} commands, not in
6864 any output section descriptions and only once in the linker script.
6865 @var{commonpagesize} should be less or equal to @var{maxpagesize} and should
6866 be the system page size the object wants to be optimized for while still
6867 running on system page sizes up to @var{maxpagesize}. Note however
6868 that @samp{-z relro} protection will not be effective if the system
6869 page size is larger than @var{commonpagesize}.
6874 . = DATA_SEGMENT_ALIGN(0x10000, 0x2000);
6877 @item DATA_SEGMENT_END(@var{exp})
6878 @kindex DATA_SEGMENT_END(@var{exp})
6879 This defines the end of data segment for @code{DATA_SEGMENT_ALIGN}
6880 evaluation purposes.
6883 . = DATA_SEGMENT_END(.);
6886 @item DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6887 @kindex DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6888 This defines the end of the @code{PT_GNU_RELRO} segment when
6889 @samp{-z relro} option is used.
6890 When @samp{-z relro} option is not present, @code{DATA_SEGMENT_RELRO_END}
6891 does nothing, otherwise @code{DATA_SEGMENT_ALIGN} is padded so that
6892 @var{exp} + @var{offset} is aligned to the @var{commonpagesize}
6893 argument given to @code{DATA_SEGMENT_ALIGN}. If present in the linker
6894 script, it must be placed between @code{DATA_SEGMENT_ALIGN} and
6895 @code{DATA_SEGMENT_END}. Evaluates to the second argument plus any
6896 padding needed at the end of the @code{PT_GNU_RELRO} segment due to
6900 . = DATA_SEGMENT_RELRO_END(24, .);
6903 @item DEFINED(@var{symbol})
6904 @kindex DEFINED(@var{symbol})
6905 @cindex symbol defaults
6906 Return 1 if @var{symbol} is in the linker global symbol table and is
6907 defined before the statement using DEFINED in the script, otherwise
6908 return 0. You can use this function to provide
6909 default values for symbols. For example, the following script fragment
6910 shows how to set a global symbol @samp{begin} to the first location in
6911 the @samp{.text} section---but if a symbol called @samp{begin} already
6912 existed, its value is preserved:
6918 begin = DEFINED(begin) ? begin : . ;
6926 @item LENGTH(@var{memory})
6927 @kindex LENGTH(@var{memory})
6928 Return the length of the memory region named @var{memory}.
6930 @item LOADADDR(@var{section})
6931 @kindex LOADADDR(@var{section})
6932 @cindex section load address in expression
6933 Return the absolute LMA of the named @var{section}. (@pxref{Output
6936 @item LOG2CEIL(@var{exp})
6937 @kindex LOG2CEIL(@var{exp})
6938 Return the binary logarithm of @var{exp} rounded towards infinity.
6939 @code{LOG2CEIL(0)} returns 0.
6942 @item MAX(@var{exp1}, @var{exp2})
6943 Returns the maximum of @var{exp1} and @var{exp2}.
6946 @item MIN(@var{exp1}, @var{exp2})
6947 Returns the minimum of @var{exp1} and @var{exp2}.
6949 @item NEXT(@var{exp})
6950 @kindex NEXT(@var{exp})
6951 @cindex unallocated address, next
6952 Return the next unallocated address that is a multiple of @var{exp}.
6953 This function is closely related to @code{ALIGN(@var{exp})}; unless you
6954 use the @code{MEMORY} command to define discontinuous memory for the
6955 output file, the two functions are equivalent.
6957 @item ORIGIN(@var{memory})
6958 @kindex ORIGIN(@var{memory})
6959 Return the origin of the memory region named @var{memory}.
6961 @item SEGMENT_START(@var{segment}, @var{default})
6962 @kindex SEGMENT_START(@var{segment}, @var{default})
6963 Return the base address of the named @var{segment}. If an explicit
6964 value has already been given for this segment (with a command-line
6965 @samp{-T} option) then that value will be returned otherwise the value
6966 will be @var{default}. At present, the @samp{-T} command-line option
6967 can only be used to set the base address for the ``text'', ``data'', and
6968 ``bss'' sections, but you can use @code{SEGMENT_START} with any segment
6971 @item SIZEOF(@var{section})
6972 @kindex SIZEOF(@var{section})
6973 @cindex section size
6974 Return the size in bytes of the named @var{section}, if that section has
6975 been allocated. If the section has not been allocated when this is
6976 evaluated, the linker will report an error. In the following example,
6977 @code{symbol_1} and @code{symbol_2} are assigned identical values:
6986 symbol_1 = .end - .start ;
6987 symbol_2 = SIZEOF(.output);
6992 @item SIZEOF_HEADERS
6993 @itemx sizeof_headers
6994 @kindex SIZEOF_HEADERS
6996 Return the size in bytes of the output file's headers. This is
6997 information which appears at the start of the output file. You can use
6998 this number when setting the start address of the first section, if you
6999 choose, to facilitate paging.
7001 @cindex not enough room for program headers
7002 @cindex program headers, not enough room
7003 When producing an ELF output file, if the linker script uses the
7004 @code{SIZEOF_HEADERS} builtin function, the linker must compute the
7005 number of program headers before it has determined all the section
7006 addresses and sizes. If the linker later discovers that it needs
7007 additional program headers, it will report an error @samp{not enough
7008 room for program headers}. To avoid this error, you must avoid using
7009 the @code{SIZEOF_HEADERS} function, or you must rework your linker
7010 script to avoid forcing the linker to use additional program headers, or
7011 you must define the program headers yourself using the @code{PHDRS}
7012 command (@pxref{PHDRS}).
7015 @node Implicit Linker Scripts
7016 @section Implicit Linker Scripts
7017 @cindex implicit linker scripts
7018 If you specify a linker input file which the linker can not recognize as
7019 an object file or an archive file, it will try to read the file as a
7020 linker script. If the file can not be parsed as a linker script, the
7021 linker will report an error.
7023 An implicit linker script will not replace the default linker script.
7025 Typically an implicit linker script would contain only symbol
7026 assignments, or the @code{INPUT}, @code{GROUP}, or @code{VERSION}
7029 Any input files read because of an implicit linker script will be read
7030 at the position in the command line where the implicit linker script was
7031 read. This can affect archive searching.
7034 @chapter Linker Plugins
7037 @cindex linker plugins
7038 The linker can use dynamically loaded plugins to modify its behavior.
7039 For example, the link-time optimization feature that some compilers
7040 support is implemented with a linker plugin.
7042 Currently there is only one plugin shipped by default, but more may
7043 be added here later.
7046 * libdep Plugin:: Static Library Dependencies Plugin
7050 @section Static Library Dependencies Plugin
7051 @cindex static library dependencies
7052 Originally, static libraries were contained in an archive file consisting
7053 just of a collection of relocatable object files. Later they evolved to
7054 optionally include a symbol table, to assist in finding the needed objects
7055 within a library. There their evolution ended, and dynamic libraries
7058 One useful feature of dynamic libraries was that, more than just collecting
7059 multiple objects into a single file, they also included a list of their
7060 dependencies, such that one could specify just the name of a single dynamic
7061 library at link time, and all of its dependencies would be implicitly
7062 referenced as well. But static libraries lacked this feature, so if a
7063 link invocation was switched from using dynamic libraries to static
7064 libraries, the link command would usually fail unless it was rewritten to
7065 explicitly list the dependencies of the static library.
7067 The GNU @command{ar} utility now supports a @option{--record-libdeps} option
7068 to embed dependency lists into static libraries as well, and the @file{libdep}
7069 plugin may be used to read this dependency information at link time. The
7070 dependency information is stored as a single string, carrying @option{-l}
7071 and @option{-L} arguments as they would normally appear in a linker
7072 command line. As such, the information can be written with any text
7073 utility and stored into any archive, even if GNU @command{ar} is not
7074 being used to create the archive. The information is stored in an
7075 archive member named @samp{__.LIBDEP}.
7077 For example, given a library @file{libssl.a} that depends on another
7078 library @file{libcrypto.a} which may be found in @file{/usr/local/lib},
7079 the @samp{__.LIBDEP} member of @file{libssl.a} would contain
7082 -L/usr/local/lib -lcrypto
7086 @node Machine Dependent
7087 @chapter Machine Dependent Features
7089 @cindex machine dependencies
7090 @command{ld} has additional features on some platforms; the following
7091 sections describe them. Machines where @command{ld} has no additional
7092 functionality are not listed.
7096 * H8/300:: @command{ld} and the H8/300
7099 * M68HC11/68HC12:: @code{ld} and the Motorola 68HC11 and 68HC12 families
7102 * ARM:: @command{ld} and the ARM family
7105 * HPPA ELF32:: @command{ld} and HPPA 32-bit ELF
7108 * M68K:: @command{ld} and the Motorola 68K family
7111 * MIPS:: @command{ld} and the MIPS family
7114 * MMIX:: @command{ld} and MMIX
7117 * MSP430:: @command{ld} and MSP430
7120 * NDS32:: @command{ld} and NDS32
7123 * Nios II:: @command{ld} and the Altera Nios II
7126 * PowerPC ELF32:: @command{ld} and PowerPC 32-bit ELF Support
7129 * PowerPC64 ELF64:: @command{ld} and PowerPC64 64-bit ELF Support
7132 * S/390 ELF:: @command{ld} and S/390 ELF Support
7135 * SPU ELF:: @command{ld} and SPU ELF Support
7138 * TI COFF:: @command{ld} and TI COFF
7141 * WIN32:: @command{ld} and WIN32 (cygwin/mingw)
7144 * Xtensa:: @command{ld} and Xtensa Processors
7155 @section @command{ld} and the H8/300
7157 @cindex H8/300 support
7158 For the H8/300, @command{ld} can perform these global optimizations when
7159 you specify the @samp{--relax} command-line option.
7162 @cindex relaxing on H8/300
7163 @item relaxing address modes
7164 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
7165 targets are within eight bits, and turns them into eight-bit
7166 program-counter relative @code{bsr} and @code{bra} instructions,
7169 @cindex synthesizing on H8/300
7170 @item synthesizing instructions
7171 @c FIXME: specifically mov.b, or any mov instructions really? -> mov.b only, at least on H8, H8H, H8S
7172 @command{ld} finds all @code{mov.b} instructions which use the
7173 sixteen-bit absolute address form, but refer to the top
7174 page of memory, and changes them to use the eight-bit address form.
7175 (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into
7176 @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the
7177 top page of memory).
7179 @command{ld} finds all @code{mov} instructions which use the register
7180 indirect with 32-bit displacement addressing mode, but use a small
7181 displacement inside 16-bit displacement range, and changes them to use
7182 the 16-bit displacement form. (That is: the linker turns @samp{mov.b
7183 @code{@@}@var{d}:32,ERx} into @samp{mov.b @code{@@}@var{d}:16,ERx}
7184 whenever the displacement @var{d} is in the 16 bit signed integer
7185 range. Only implemented in ELF-format ld).
7187 @item bit manipulation instructions
7188 @command{ld} finds all bit manipulation instructions like @code{band, bclr,
7189 biand, bild, bior, bist, bixor, bld, bnot, bor, bset, bst, btst, bxor}
7190 which use 32 bit and 16 bit absolute address form, but refer to the top
7191 page of memory, and changes them to use the 8 bit address form.
7192 (That is: the linker turns @samp{bset #xx:3,@code{@@}@var{aa}:32} into
7193 @samp{bset #xx:3,@code{@@}@var{aa}:8} whenever the address @var{aa} is in
7194 the top page of memory).
7196 @item system control instructions
7197 @command{ld} finds all @code{ldc.w, stc.w} instructions which use the
7198 32 bit absolute address form, but refer to the top page of memory, and
7199 changes them to use 16 bit address form.
7200 (That is: the linker turns @samp{ldc.w @code{@@}@var{aa}:32,ccr} into
7201 @samp{ldc.w @code{@@}@var{aa}:16,ccr} whenever the address @var{aa} is in
7202 the top page of memory).
7212 @c This stuff is pointless to say unless you're especially concerned
7213 @c with Renesas chips; don't enable it for generic case, please.
7215 @chapter @command{ld} and Other Renesas Chips
7217 @command{ld} also supports the Renesas (formerly Hitachi) H8/300H,
7218 H8/500, and SH chips. No special features, commands, or command-line
7219 options are required for these chips.
7233 @node M68HC11/68HC12
7234 @section @command{ld} and the Motorola 68HC11 and 68HC12 families
7236 @cindex M68HC11 and 68HC12 support
7238 @subsection Linker Relaxation
7240 For the Motorola 68HC11, @command{ld} can perform these global
7241 optimizations when you specify the @samp{--relax} command-line option.
7244 @cindex relaxing on M68HC11
7245 @item relaxing address modes
7246 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
7247 targets are within eight bits, and turns them into eight-bit
7248 program-counter relative @code{bsr} and @code{bra} instructions,
7251 @command{ld} also looks at all 16-bit extended addressing modes and
7252 transforms them in a direct addressing mode when the address is in
7253 page 0 (between 0 and 0x0ff).
7255 @item relaxing gcc instruction group
7256 When @command{gcc} is called with @option{-mrelax}, it can emit group
7257 of instructions that the linker can optimize to use a 68HC11 direct
7258 addressing mode. These instructions consists of @code{bclr} or
7259 @code{bset} instructions.
7263 @subsection Trampoline Generation
7265 @cindex trampoline generation on M68HC11
7266 @cindex trampoline generation on M68HC12
7267 For 68HC11 and 68HC12, @command{ld} can generate trampoline code to
7268 call a far function using a normal @code{jsr} instruction. The linker
7269 will also change the relocation to some far function to use the
7270 trampoline address instead of the function address. This is typically the
7271 case when a pointer to a function is taken. The pointer will in fact
7272 point to the function trampoline.
7280 @section @command{ld} and the ARM family
7282 @cindex ARM interworking support
7283 @kindex --support-old-code
7284 For the ARM, @command{ld} will generate code stubs to allow functions calls
7285 between ARM and Thumb code. These stubs only work with code that has
7286 been compiled and assembled with the @samp{-mthumb-interwork} command
7287 line option. If it is necessary to link with old ARM object files or
7288 libraries, which have not been compiled with the -mthumb-interwork
7289 option then the @samp{--support-old-code} command-line switch should be
7290 given to the linker. This will make it generate larger stub functions
7291 which will work with non-interworking aware ARM code. Note, however,
7292 the linker does not support generating stubs for function calls to
7293 non-interworking aware Thumb code.
7295 @cindex thumb entry point
7296 @cindex entry point, thumb
7297 @kindex --thumb-entry=@var{entry}
7298 The @samp{--thumb-entry} switch is a duplicate of the generic
7299 @samp{--entry} switch, in that it sets the program's starting address.
7300 But it also sets the bottom bit of the address, so that it can be
7301 branched to using a BX instruction, and the program will start
7302 executing in Thumb mode straight away.
7304 @cindex PE import table prefixing
7305 @kindex --use-nul-prefixed-import-tables
7306 The @samp{--use-nul-prefixed-import-tables} switch is specifying, that
7307 the import tables idata4 and idata5 have to be generated with a zero
7308 element prefix for import libraries. This is the old style to generate
7309 import tables. By default this option is turned off.
7313 The @samp{--be8} switch instructs @command{ld} to generate BE8 format
7314 executables. This option is only valid when linking big-endian
7315 objects - ie ones which have been assembled with the @option{-EB}
7316 option. The resulting image will contain big-endian data and
7320 @kindex --target1-rel
7321 @kindex --target1-abs
7322 The @samp{R_ARM_TARGET1} relocation is typically used for entries in the
7323 @samp{.init_array} section. It is interpreted as either @samp{R_ARM_REL32}
7324 or @samp{R_ARM_ABS32}, depending on the target. The @samp{--target1-rel}
7325 and @samp{--target1-abs} switches override the default.
7328 @kindex --target2=@var{type}
7329 The @samp{--target2=type} switch overrides the default definition of the
7330 @samp{R_ARM_TARGET2} relocation. Valid values for @samp{type}, their
7331 meanings, and target defaults are as follows:
7334 @samp{R_ARM_REL32} (arm*-*-elf, arm*-*-eabi)
7338 @samp{R_ARM_GOT_PREL} (arm*-*-linux, arm*-*-*bsd)
7343 The @samp{R_ARM_V4BX} relocation (defined by the ARM AAELF
7344 specification) enables objects compiled for the ARMv4 architecture to be
7345 interworking-safe when linked with other objects compiled for ARMv4t, but
7346 also allows pure ARMv4 binaries to be built from the same ARMv4 objects.
7348 In the latter case, the switch @option{--fix-v4bx} must be passed to the
7349 linker, which causes v4t @code{BX rM} instructions to be rewritten as
7350 @code{MOV PC,rM}, since v4 processors do not have a @code{BX} instruction.
7352 In the former case, the switch should not be used, and @samp{R_ARM_V4BX}
7353 relocations are ignored.
7355 @cindex FIX_V4BX_INTERWORKING
7356 @kindex --fix-v4bx-interworking
7357 Replace @code{BX rM} instructions identified by @samp{R_ARM_V4BX}
7358 relocations with a branch to the following veneer:
7366 This allows generation of libraries/applications that work on ARMv4 cores
7367 and are still interworking safe. Note that the above veneer clobbers the
7368 condition flags, so may cause incorrect program behavior in rare cases.
7372 The @samp{--use-blx} switch enables the linker to use ARM/Thumb
7373 BLX instructions (available on ARMv5t and above) in various
7374 situations. Currently it is used to perform calls via the PLT from Thumb
7375 code using BLX rather than using BX and a mode-switching stub before
7376 each PLT entry. This should lead to such calls executing slightly faster.
7378 @cindex VFP11_DENORM_FIX
7379 @kindex --vfp11-denorm-fix
7380 The @samp{--vfp11-denorm-fix} switch enables a link-time workaround for a
7381 bug in certain VFP11 coprocessor hardware, which sometimes allows
7382 instructions with denorm operands (which must be handled by support code)
7383 to have those operands overwritten by subsequent instructions before
7384 the support code can read the intended values.
7386 The bug may be avoided in scalar mode if you allow at least one
7387 intervening instruction between a VFP11 instruction which uses a register
7388 and another instruction which writes to the same register, or at least two
7389 intervening instructions if vector mode is in use. The bug only affects
7390 full-compliance floating-point mode: you do not need this workaround if
7391 you are using "runfast" mode. Please contact ARM for further details.
7393 If you know you are using buggy VFP11 hardware, you can
7394 enable this workaround by specifying the linker option
7395 @samp{--vfp-denorm-fix=scalar} if you are using the VFP11 scalar
7396 mode only, or @samp{--vfp-denorm-fix=vector} if you are using
7397 vector mode (the latter also works for scalar code). The default is
7398 @samp{--vfp-denorm-fix=none}.
7400 If the workaround is enabled, instructions are scanned for
7401 potentially-troublesome sequences, and a veneer is created for each
7402 such sequence which may trigger the erratum. The veneer consists of the
7403 first instruction of the sequence and a branch back to the subsequent
7404 instruction. The original instruction is then replaced with a branch to
7405 the veneer. The extra cycles required to call and return from the veneer
7406 are sufficient to avoid the erratum in both the scalar and vector cases.
7408 @cindex ARM1176 erratum workaround
7409 @kindex --fix-arm1176
7410 @kindex --no-fix-arm1176
7411 The @samp{--fix-arm1176} switch enables a link-time workaround for an erratum
7412 in certain ARM1176 processors. The workaround is enabled by default if you
7413 are targeting ARM v6 (excluding ARM v6T2) or earlier. It can be disabled
7414 unconditionally by specifying @samp{--no-fix-arm1176}.
7416 Further information is available in the ``ARM1176JZ-S and ARM1176JZF-S
7417 Programmer Advice Notice'' available on the ARM documentation website at:
7418 http://infocenter.arm.com/.
7420 @cindex STM32L4xx erratum workaround
7421 @kindex --fix-stm32l4xx-629360
7423 The @samp{--fix-stm32l4xx-629360} switch enables a link-time
7424 workaround for a bug in the bus matrix / memory controller for some of
7425 the STM32 Cortex-M4 based products (STM32L4xx). When accessing
7426 off-chip memory via the affected bus for bus reads of 9 words or more,
7427 the bus can generate corrupt data and/or abort. These are only
7428 core-initiated accesses (not DMA), and might affect any access:
7429 integer loads such as LDM, POP and floating-point loads such as VLDM,
7430 VPOP. Stores are not affected.
7432 The bug can be avoided by splitting memory accesses into the
7433 necessary chunks to keep bus reads below 8 words.
7435 The workaround is not enabled by default, this is equivalent to use
7436 @samp{--fix-stm32l4xx-629360=none}. If you know you are using buggy
7437 STM32L4xx hardware, you can enable the workaround by specifying the
7438 linker option @samp{--fix-stm32l4xx-629360}, or the equivalent
7439 @samp{--fix-stm32l4xx-629360=default}.
7441 If the workaround is enabled, instructions are scanned for
7442 potentially-troublesome sequences, and a veneer is created for each
7443 such sequence which may trigger the erratum. The veneer consists in a
7444 replacement sequence emulating the behaviour of the original one and a
7445 branch back to the subsequent instruction. The original instruction is
7446 then replaced with a branch to the veneer.
7448 The workaround does not always preserve the memory access order for
7449 the LDMDB instruction, when the instruction loads the PC.
7451 The workaround is not able to handle problematic instructions when
7452 they are in the middle of an IT block, since a branch is not allowed
7453 there. In that case, the linker reports a warning and no replacement
7456 The workaround is not able to replace problematic instructions with a
7457 PC-relative branch instruction if the @samp{.text} section is too
7458 large. In that case, when the branch that replaces the original code
7459 cannot be encoded, the linker reports a warning and no replacement
7462 @cindex NO_ENUM_SIZE_WARNING
7463 @kindex --no-enum-size-warning
7464 The @option{--no-enum-size-warning} switch prevents the linker from
7465 warning when linking object files that specify incompatible EABI
7466 enumeration size attributes. For example, with this switch enabled,
7467 linking of an object file using 32-bit enumeration values with another
7468 using enumeration values fitted into the smallest possible space will
7471 @cindex NO_WCHAR_SIZE_WARNING
7472 @kindex --no-wchar-size-warning
7473 The @option{--no-wchar-size-warning} switch prevents the linker from
7474 warning when linking object files that specify incompatible EABI
7475 @code{wchar_t} size attributes. For example, with this switch enabled,
7476 linking of an object file using 32-bit @code{wchar_t} values with another
7477 using 16-bit @code{wchar_t} values will not be diagnosed.
7480 @kindex --pic-veneer
7481 The @samp{--pic-veneer} switch makes the linker use PIC sequences for
7482 ARM/Thumb interworking veneers, even if the rest of the binary
7483 is not PIC. This avoids problems on uClinux targets where
7484 @samp{--emit-relocs} is used to generate relocatable binaries.
7486 @cindex STUB_GROUP_SIZE
7487 @kindex --stub-group-size=@var{N}
7488 The linker will automatically generate and insert small sequences of
7489 code into a linked ARM ELF executable whenever an attempt is made to
7490 perform a function call to a symbol that is too far away. The
7491 placement of these sequences of instructions - called stubs - is
7492 controlled by the command-line option @option{--stub-group-size=N}.
7493 The placement is important because a poor choice can create a need for
7494 duplicate stubs, increasing the code size. The linker will try to
7495 group stubs together in order to reduce interruptions to the flow of
7496 code, but it needs guidance as to how big these groups should be and
7497 where they should be placed.
7499 The value of @samp{N}, the parameter to the
7500 @option{--stub-group-size=} option controls where the stub groups are
7501 placed. If it is negative then all stubs are placed after the first
7502 branch that needs them. If it is positive then the stubs can be
7503 placed either before or after the branches that need them. If the
7504 value of @samp{N} is 1 (either +1 or -1) then the linker will choose
7505 exactly where to place groups of stubs, using its built in heuristics.
7506 A value of @samp{N} greater than 1 (or smaller than -1) tells the
7507 linker that a single group of stubs can service at most @samp{N} bytes
7508 from the input sections.
7510 The default, if @option{--stub-group-size=} is not specified, is
7513 Farcalls stubs insertion is fully supported for the ARM-EABI target
7514 only, because it relies on object files properties not present
7517 @cindex Cortex-A8 erratum workaround
7518 @kindex --fix-cortex-a8
7519 @kindex --no-fix-cortex-a8
7520 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}.
7522 The erratum only affects Thumb-2 code. Please contact ARM for further details.
7524 @cindex Cortex-A53 erratum 835769 workaround
7525 @kindex --fix-cortex-a53-835769
7526 @kindex --no-fix-cortex-a53-835769
7527 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}.
7529 Please contact ARM for further details.
7531 @kindex --merge-exidx-entries
7532 @kindex --no-merge-exidx-entries
7533 @cindex Merging exidx entries
7534 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent exidx entries in debuginfo.
7537 @cindex 32-bit PLT entries
7538 The @samp{--long-plt} option enables the use of 16 byte PLT entries
7539 which support up to 4Gb of code. The default is to use 12 byte PLT
7540 entries which only support 512Mb of code.
7542 @kindex --no-apply-dynamic-relocs
7543 @cindex AArch64 rela addend
7544 The @samp{--no-apply-dynamic-relocs} option makes AArch64 linker do not apply
7545 link-time values for dynamic relocations.
7547 @cindex Placement of SG veneers
7548 All SG veneers are placed in the special output section @code{.gnu.sgstubs}.
7549 Its start address must be set, either with the command-line option
7550 @samp{--section-start} or in a linker script, to indicate where to place these
7553 @kindex --cmse-implib
7554 @cindex Secure gateway import library
7555 The @samp{--cmse-implib} option requests that the import libraries
7556 specified by the @samp{--out-implib} and @samp{--in-implib} options are
7557 secure gateway import libraries, suitable for linking a non-secure
7558 executable against secure code as per ARMv8-M Security Extensions.
7560 @kindex --in-implib=@var{file}
7561 @cindex Input import library
7562 The @samp{--in-implib=file} specifies an input import library whose symbols
7563 must keep the same address in the executable being produced. A warning is
7564 given if no @samp{--out-implib} is given but new symbols have been introduced
7565 in the executable that should be listed in its import library. Otherwise, if
7566 @samp{--out-implib} is specified, the symbols are added to the output import
7567 library. A warning is also given if some symbols present in the input import
7568 library have disappeared from the executable. This option is only effective
7569 for Secure Gateway import libraries, ie. when @samp{--cmse-implib} is
7583 @section @command{ld} and HPPA 32-bit ELF Support
7584 @cindex HPPA multiple sub-space stubs
7585 @kindex --multi-subspace
7586 When generating a shared library, @command{ld} will by default generate
7587 import stubs suitable for use with a single sub-space application.
7588 The @samp{--multi-subspace} switch causes @command{ld} to generate export
7589 stubs, and different (larger) import stubs suitable for use with
7590 multiple sub-spaces.
7592 @cindex HPPA stub grouping
7593 @kindex --stub-group-size=@var{N}
7594 Long branch stubs and import/export stubs are placed by @command{ld} in
7595 stub sections located between groups of input sections.
7596 @samp{--stub-group-size} specifies the maximum size of a group of input
7597 sections handled by one stub section. Since branch offsets are signed,
7598 a stub section may serve two groups of input sections, one group before
7599 the stub section, and one group after it. However, when using
7600 conditional branches that require stubs, it may be better (for branch
7601 prediction) that stub sections only serve one group of input sections.
7602 A negative value for @samp{N} chooses this scheme, ensuring that
7603 branches to stubs always use a negative offset. Two special values of
7604 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
7605 @command{ld} to automatically size input section groups for the branch types
7606 detected, with the same behaviour regarding stub placement as other
7607 positive or negative values of @samp{N} respectively.
7609 Note that @samp{--stub-group-size} does not split input sections. A
7610 single input section larger than the group size specified will of course
7611 create a larger group (of one section). If input sections are too
7612 large, it may not be possible for a branch to reach its stub.
7625 @section @command{ld} and the Motorola 68K family
7627 @cindex Motorola 68K GOT generation
7628 @kindex --got=@var{type}
7629 The @samp{--got=@var{type}} option lets you choose the GOT generation scheme.
7630 The choices are @samp{single}, @samp{negative}, @samp{multigot} and
7631 @samp{target}. When @samp{target} is selected the linker chooses
7632 the default GOT generation scheme for the current target.
7633 @samp{single} tells the linker to generate a single GOT with
7634 entries only at non-negative offsets.
7635 @samp{negative} instructs the linker to generate a single GOT with
7636 entries at both negative and positive offsets. Not all environments
7638 @samp{multigot} allows the linker to generate several GOTs in the
7639 output file. All GOT references from a single input object
7640 file access the same GOT, but references from different input object
7641 files might access different GOTs. Not all environments support such GOTs.
7654 @section @command{ld} and the MIPS family
7656 @cindex MIPS microMIPS instruction choice selection
7659 The @samp{--insn32} and @samp{--no-insn32} options control the choice of
7660 microMIPS instructions used in code generated by the linker, such as that
7661 in the PLT or lazy binding stubs, or in relaxation. If @samp{--insn32} is
7662 used, then the linker only uses 32-bit instruction encodings. By default
7663 or if @samp{--no-insn32} is used, all instruction encodings are used,
7664 including 16-bit ones where possible.
7666 @cindex MIPS branch relocation check control
7667 @kindex --ignore-branch-isa
7668 @kindex --no-ignore-branch-isa
7669 The @samp{--ignore-branch-isa} and @samp{--no-ignore-branch-isa} options
7670 control branch relocation checks for invalid ISA mode transitions. If
7671 @samp{--ignore-branch-isa} is used, then the linker accepts any branch
7672 relocations and any ISA mode transition required is lost in relocation
7673 calculation, except for some cases of @code{BAL} instructions which meet
7674 relaxation conditions and are converted to equivalent @code{JALX}
7675 instructions as the associated relocation is calculated. By default
7676 or if @samp{--no-ignore-branch-isa} is used a check is made causing
7677 the loss of an ISA mode transition to produce an error.
7690 @section @code{ld} and MMIX
7691 For MMIX, there is a choice of generating @code{ELF} object files or
7692 @code{mmo} object files when linking. The simulator @code{mmix}
7693 understands the @code{mmo} format. The binutils @code{objcopy} utility
7694 can translate between the two formats.
7696 There is one special section, the @samp{.MMIX.reg_contents} section.
7697 Contents in this section is assumed to correspond to that of global
7698 registers, and symbols referring to it are translated to special symbols,
7699 equal to registers. In a final link, the start address of the
7700 @samp{.MMIX.reg_contents} section corresponds to the first allocated
7701 global register multiplied by 8. Register @code{$255} is not included in
7702 this section; it is always set to the program entry, which is at the
7703 symbol @code{Main} for @code{mmo} files.
7705 Global symbols with the prefix @code{__.MMIX.start.}, for example
7706 @code{__.MMIX.start..text} and @code{__.MMIX.start..data} are special.
7707 The default linker script uses these to set the default start address
7710 Initial and trailing multiples of zero-valued 32-bit words in a section,
7711 are left out from an mmo file.
7724 @section @code{ld} and MSP430
7725 For the MSP430 it is possible to select the MPU architecture. The flag @samp{-m [mpu type]}
7726 will select an appropriate linker script for selected MPU type. (To get a list of known MPUs
7727 just pass @samp{-m help} option to the linker).
7729 @cindex MSP430 extra sections
7730 The linker will recognize some extra sections which are MSP430 specific:
7733 @item @samp{.vectors}
7734 Defines a portion of ROM where interrupt vectors located.
7736 @item @samp{.bootloader}
7737 Defines the bootloader portion of the ROM (if applicable). Any code
7738 in this section will be uploaded to the MPU.
7740 @item @samp{.infomem}
7741 Defines an information memory section (if applicable). Any code in
7742 this section will be uploaded to the MPU.
7744 @item @samp{.infomemnobits}
7745 This is the same as the @samp{.infomem} section except that any code
7746 in this section will not be uploaded to the MPU.
7748 @item @samp{.noinit}
7749 Denotes a portion of RAM located above @samp{.bss} section.
7751 The last two sections are used by gcc.
7755 @cindex MSP430 Options
7756 @kindex --code-region
7757 @item --code-region=[either,lower,upper,none]
7758 This will transform .text* sections to [either,lower,upper].text* sections. The
7759 argument passed to GCC for -mcode-region is propagated to the linker
7762 @kindex --data-region
7763 @item --data-region=[either,lower,upper,none]
7764 This will transform .data*, .bss* and .rodata* sections to
7765 [either,lower,upper].[data,bss,rodata]* sections. The argument passed to GCC
7766 for -mdata-region is propagated to the linker using this option.
7768 @kindex --disable-sec-transformation
7769 @item --disable-sec-transformation
7770 Prevent the transformation of sections as specified by the @code{--code-region}
7771 and @code{--data-region} options.
7772 This is useful if you are compiling and linking using a single call to the GCC
7773 wrapper, and want to compile the source files using -m[code,data]-region but
7774 not transform the sections for prebuilt libraries and objects.
7788 @section @code{ld} and NDS32
7789 @kindex relaxing on NDS32
7790 For NDS32, there are some options to select relaxation behavior. The linker
7791 relaxes objects according to these options.
7794 @item @samp{--m[no-]fp-as-gp}
7795 Disable/enable fp-as-gp relaxation.
7797 @item @samp{--mexport-symbols=FILE}
7798 Exporting symbols and their address into FILE as linker script.
7800 @item @samp{--m[no-]ex9}
7801 Disable/enable link-time EX9 relaxation.
7803 @item @samp{--mexport-ex9=FILE}
7804 Export the EX9 table after linking.
7806 @item @samp{--mimport-ex9=FILE}
7807 Import the Ex9 table for EX9 relaxation.
7809 @item @samp{--mupdate-ex9}
7810 Update the existing EX9 table.
7812 @item @samp{--mex9-limit=NUM}
7813 Maximum number of entries in the ex9 table.
7815 @item @samp{--mex9-loop-aware}
7816 Avoid generating the EX9 instruction inside the loop.
7818 @item @samp{--m[no-]ifc}
7819 Disable/enable the link-time IFC optimization.
7821 @item @samp{--mifc-loop-aware}
7822 Avoid generating the IFC instruction inside the loop.
7836 @section @command{ld} and the Altera Nios II
7837 @cindex Nios II call relaxation
7838 @kindex --relax on Nios II
7840 Call and immediate jump instructions on Nios II processors are limited to
7841 transferring control to addresses in the same 256MB memory segment,
7842 which may result in @command{ld} giving
7843 @samp{relocation truncated to fit} errors with very large programs.
7844 The command-line option @option{--relax} enables the generation of
7845 trampolines that can access the entire 32-bit address space for calls
7846 outside the normal @code{call} and @code{jmpi} address range. These
7847 trampolines are inserted at section boundaries, so may not themselves
7848 be reachable if an input section and its associated call trampolines are
7851 The @option{--relax} option is enabled by default unless @option{-r}
7852 is also specified. You can disable trampoline generation by using the
7853 @option{--no-relax} linker option. You can also disable this optimization
7854 locally by using the @samp{set .noat} directive in assembly-language
7855 source files, as the linker-inserted trampolines use the @code{at}
7856 register as a temporary.
7858 Note that the linker @option{--relax} option is independent of assembler
7859 relaxation options, and that using the GNU assembler's @option{-relax-all}
7860 option interferes with the linker's more selective call instruction relaxation.
7873 @section @command{ld} and PowerPC 32-bit ELF Support
7874 @cindex PowerPC long branches
7875 @kindex --relax on PowerPC
7876 Branches on PowerPC processors are limited to a signed 26-bit
7877 displacement, which may result in @command{ld} giving
7878 @samp{relocation truncated to fit} errors with very large programs.
7879 @samp{--relax} enables the generation of trampolines that can access
7880 the entire 32-bit address space. These trampolines are inserted at
7881 section boundaries, so may not themselves be reachable if an input
7882 section exceeds 33M in size. You may combine @samp{-r} and
7883 @samp{--relax} to add trampolines in a partial link. In that case
7884 both branches to undefined symbols and inter-section branches are also
7885 considered potentially out of range, and trampolines inserted.
7887 @cindex PowerPC ELF32 options
7892 Current PowerPC GCC accepts a @samp{-msecure-plt} option that
7893 generates code capable of using a newer PLT and GOT layout that has
7894 the security advantage of no executable section ever needing to be
7895 writable and no writable section ever being executable. PowerPC
7896 @command{ld} will generate this layout, including stubs to access the
7897 PLT, if all input files (including startup and static libraries) were
7898 compiled with @samp{-msecure-plt}. @samp{--bss-plt} forces the old
7899 BSS PLT (and GOT layout) which can give slightly better performance.
7901 @kindex --secure-plt
7903 @command{ld} will use the new PLT and GOT layout if it is linking new
7904 @samp{-fpic} or @samp{-fPIC} code, but does not do so automatically
7905 when linking non-PIC code. This option requests the new PLT and GOT
7906 layout. A warning will be given if some object file requires the old
7912 The new secure PLT and GOT are placed differently relative to other
7913 sections compared to older BSS PLT and GOT placement. The location of
7914 @code{.plt} must change because the new secure PLT is an initialized
7915 section while the old PLT is uninitialized. The reason for the
7916 @code{.got} change is more subtle: The new placement allows
7917 @code{.got} to be read-only in applications linked with
7918 @samp{-z relro -z now}. However, this placement means that
7919 @code{.sdata} cannot always be used in shared libraries, because the
7920 PowerPC ABI accesses @code{.sdata} in shared libraries from the GOT
7921 pointer. @samp{--sdata-got} forces the old GOT placement. PowerPC
7922 GCC doesn't use @code{.sdata} in shared libraries, so this option is
7923 really only useful for other compilers that may do so.
7925 @cindex PowerPC stub symbols
7926 @kindex --emit-stub-syms
7927 @item --emit-stub-syms
7928 This option causes @command{ld} to label linker stubs with a local
7929 symbol that encodes the stub type and destination.
7931 @cindex PowerPC TLS optimization
7932 @kindex --no-tls-optimize
7933 @item --no-tls-optimize
7934 PowerPC @command{ld} normally performs some optimization of code
7935 sequences used to access Thread-Local Storage. Use this option to
7936 disable the optimization.
7949 @node PowerPC64 ELF64
7950 @section @command{ld} and PowerPC64 64-bit ELF Support
7952 @cindex PowerPC64 ELF64 options
7954 @cindex PowerPC64 stub grouping
7955 @kindex --stub-group-size
7956 @item --stub-group-size
7957 Long branch stubs, PLT call stubs and TOC adjusting stubs are placed
7958 by @command{ld} in stub sections located between groups of input sections.
7959 @samp{--stub-group-size} specifies the maximum size of a group of input
7960 sections handled by one stub section. Since branch offsets are signed,
7961 a stub section may serve two groups of input sections, one group before
7962 the stub section, and one group after it. However, when using
7963 conditional branches that require stubs, it may be better (for branch
7964 prediction) that stub sections only serve one group of input sections.
7965 A negative value for @samp{N} chooses this scheme, ensuring that
7966 branches to stubs always use a negative offset. Two special values of
7967 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
7968 @command{ld} to automatically size input section groups for the branch types
7969 detected, with the same behaviour regarding stub placement as other
7970 positive or negative values of @samp{N} respectively.
7972 Note that @samp{--stub-group-size} does not split input sections. A
7973 single input section larger than the group size specified will of course
7974 create a larger group (of one section). If input sections are too
7975 large, it may not be possible for a branch to reach its stub.
7977 @cindex PowerPC64 stub symbols
7978 @kindex --emit-stub-syms
7979 @item --emit-stub-syms
7980 This option causes @command{ld} to label linker stubs with a local
7981 symbol that encodes the stub type and destination.
7983 @cindex PowerPC64 dot symbols
7985 @kindex --no-dotsyms
7988 These two options control how @command{ld} interprets version patterns
7989 in a version script. Older PowerPC64 compilers emitted both a
7990 function descriptor symbol with the same name as the function, and a
7991 code entry symbol with the name prefixed by a dot (@samp{.}). To
7992 properly version a function @samp{foo}, the version script thus needs
7993 to control both @samp{foo} and @samp{.foo}. The option
7994 @samp{--dotsyms}, on by default, automatically adds the required
7995 dot-prefixed patterns. Use @samp{--no-dotsyms} to disable this
7998 @cindex PowerPC64 register save/restore functions
7999 @kindex --save-restore-funcs
8000 @kindex --no-save-restore-funcs
8001 @item --save-restore-funcs
8002 @itemx --no-save-restore-funcs
8003 These two options control whether PowerPC64 @command{ld} automatically
8004 provides out-of-line register save and restore functions used by
8005 @samp{-Os} code. The default is to provide any such referenced
8006 function for a normal final link, and to not do so for a relocatable
8009 @cindex PowerPC64 TLS optimization
8010 @kindex --no-tls-optimize
8011 @item --no-tls-optimize
8012 PowerPC64 @command{ld} normally performs some optimization of code
8013 sequences used to access Thread-Local Storage. Use this option to
8014 disable the optimization.
8016 @cindex PowerPC64 __tls_get_addr optimization
8017 @kindex --tls-get-addr-optimize
8018 @kindex --no-tls-get-addr-optimize
8019 @kindex --tls-get-addr-regsave
8020 @kindex --no-tls-get-addr-regsave
8021 @item --tls-get-addr-optimize
8022 @itemx --no-tls-get-addr-optimize
8023 These options control how PowerPC64 @command{ld} uses a special
8024 stub to call __tls_get_addr. PowerPC64 glibc 2.22 and later support
8025 an optimization that allows the second and subsequent calls to
8026 @code{__tls_get_addr} for a given symbol to be resolved by the special
8027 stub without calling in to glibc. By default the linker enables
8028 generation of the stub when glibc advertises the availability of
8030 Using @option{--tls-get-addr-optimize} with an older glibc won't do
8031 much besides slow down your applications, but may be useful if linking
8032 an application against an older glibc with the expectation that it
8033 will normally be used on systems having a newer glibc.
8034 @option{--tls-get-addr-regsave} forces generation of a stub that saves
8035 and restores volatile registers around the call into glibc. Normally,
8036 this is done when the linker detects a call to __tls_get_addr_desc.
8037 Such calls then go via the register saving stub to __tls_get_addr_opt.
8038 @option{--no-tls-get-addr-regsave} disables generation of the
8041 @cindex PowerPC64 OPD optimization
8042 @kindex --no-opd-optimize
8043 @item --no-opd-optimize
8044 PowerPC64 @command{ld} normally removes @code{.opd} section entries
8045 corresponding to deleted link-once functions, or functions removed by
8046 the action of @samp{--gc-sections} or linker script @code{/DISCARD/}.
8047 Use this option to disable @code{.opd} optimization.
8049 @cindex PowerPC64 OPD spacing
8050 @kindex --non-overlapping-opd
8051 @item --non-overlapping-opd
8052 Some PowerPC64 compilers have an option to generate compressed
8053 @code{.opd} entries spaced 16 bytes apart, overlapping the third word,
8054 the static chain pointer (unused in C) with the first word of the next
8055 entry. This option expands such entries to the full 24 bytes.
8057 @cindex PowerPC64 TOC optimization
8058 @kindex --no-toc-optimize
8059 @item --no-toc-optimize
8060 PowerPC64 @command{ld} normally removes unused @code{.toc} section
8061 entries. Such entries are detected by examining relocations that
8062 reference the TOC in code sections. A reloc in a deleted code section
8063 marks a TOC word as unneeded, while a reloc in a kept code section
8064 marks a TOC word as needed. Since the TOC may reference itself, TOC
8065 relocs are also examined. TOC words marked as both needed and
8066 unneeded will of course be kept. TOC words without any referencing
8067 reloc are assumed to be part of a multi-word entry, and are kept or
8068 discarded as per the nearest marked preceding word. This works
8069 reliably for compiler generated code, but may be incorrect if assembly
8070 code is used to insert TOC entries. Use this option to disable the
8073 @cindex PowerPC64 inline PLT call optimization
8074 @kindex --no-inline-optimize
8075 @item --no-inline-optimize
8076 PowerPC64 @command{ld} normally replaces inline PLT call sequences
8077 marked with @code{R_PPC64_PLTSEQ}, @code{R_PPC64_PLTCALL},
8078 @code{R_PPC64_PLT16_HA} and @code{R_PPC64_PLT16_LO_DS} relocations by
8079 a number of @code{nop}s and a direct call when the function is defined
8080 locally and can't be overridden by some other definition. This option
8081 disables that optimization.
8083 @cindex PowerPC64 multi-TOC
8084 @kindex --no-multi-toc
8085 @item --no-multi-toc
8086 If given any toc option besides @code{-mcmodel=medium} or
8087 @code{-mcmodel=large}, PowerPC64 GCC generates code for a TOC model
8089 entries are accessed with a 16-bit offset from r2. This limits the
8090 total TOC size to 64K. PowerPC64 @command{ld} extends this limit by
8091 grouping code sections such that each group uses less than 64K for its
8092 TOC entries, then inserts r2 adjusting stubs between inter-group
8093 calls. @command{ld} does not split apart input sections, so cannot
8094 help if a single input file has a @code{.toc} section that exceeds
8095 64K, most likely from linking multiple files with @command{ld -r}.
8096 Use this option to turn off this feature.
8098 @cindex PowerPC64 TOC sorting
8099 @kindex --no-toc-sort
8101 By default, @command{ld} sorts TOC sections so that those whose file
8102 happens to have a section called @code{.init} or @code{.fini} are
8103 placed first, followed by TOC sections referenced by code generated
8104 with PowerPC64 gcc's @code{-mcmodel=small}, and lastly TOC sections
8105 referenced only by code generated with PowerPC64 gcc's
8106 @code{-mcmodel=medium} or @code{-mcmodel=large} options. Doing this
8107 results in better TOC grouping for multi-TOC. Use this option to turn
8110 @cindex PowerPC64 PLT stub alignment
8112 @kindex --no-plt-align
8114 @itemx --no-plt-align
8115 Use these options to control whether individual PLT call stubs are
8116 aligned to a 32-byte boundary, or to the specified power of two
8117 boundary when using @code{--plt-align=}. A negative value may be
8118 specified to pad PLT call stubs so that they do not cross the
8119 specified power of two boundary (or the minimum number of boundaries
8120 if a PLT stub is so large that it must cross a boundary). By default
8121 PLT call stubs are aligned to 32-byte boundaries.
8123 @cindex PowerPC64 PLT call stub static chain
8124 @kindex --plt-static-chain
8125 @kindex --no-plt-static-chain
8126 @item --plt-static-chain
8127 @itemx --no-plt-static-chain
8128 Use these options to control whether PLT call stubs load the static
8129 chain pointer (r11). @code{ld} defaults to not loading the static
8130 chain since there is never any need to do so on a PLT call.
8132 @cindex PowerPC64 PLT call stub thread safety
8133 @kindex --plt-thread-safe
8134 @kindex --no-plt-thread-safe
8135 @item --plt-thread-safe
8136 @itemx --no-plt-thread-safe
8137 With power7's weakly ordered memory model, it is possible when using
8138 lazy binding for ld.so to update a plt entry in one thread and have
8139 another thread see the individual plt entry words update in the wrong
8140 order, despite ld.so carefully writing in the correct order and using
8141 memory write barriers. To avoid this we need some sort of read
8142 barrier in the call stub, or use LD_BIND_NOW=1. By default, @code{ld}
8143 looks for calls to commonly used functions that create threads, and if
8144 seen, adds the necessary barriers. Use these options to change the
8147 @cindex PowerPC64 ELFv2 PLT localentry optimization
8148 @kindex --plt-localentry
8149 @kindex --no-plt-localentry
8150 @item --plt-localentry
8151 @itemx --no-localentry
8152 ELFv2 functions with localentry:0 are those with a single entry point,
8153 ie. global entry == local entry, and that have no requirement on r2
8154 (the TOC/GOT pointer) or r12, and guarantee r2 is unchanged on return.
8155 Such an external function can be called via the PLT without saving r2
8156 or restoring it on return, avoiding a common load-hit-store for small
8157 functions. The optimization is attractive, with up to 40% reduction
8158 in execution time for a small function, but can result in symbol
8159 interposition failures. Also, minor changes in a shared library,
8160 including system libraries, can cause a function that was localentry:0
8161 to become localentry:8. This will result in a dynamic loader
8162 complaint and failure to run. The option is experimental, use with
8163 care. @option{--no-plt-localentry} is the default.
8165 @cindex PowerPC64 Power10 stubs
8166 @kindex --power10-stubs
8167 @kindex --no-power10-stubs
8168 @item --power10-stubs
8169 @itemx --no-power10-stubs
8170 When PowerPC64 @command{ld} links input object files containing
8171 relocations used on power10 prefixed instructions it normally creates
8172 linkage stubs (PLT call and long branch) using power10 instructions
8173 for @code{@@notoc} PLT calls where @code{r2} is not known. The
8174 power10 notoc stubs are smaller and faster, so are preferred for
8175 power10. @option{--power10-stubs} and @option{--no-power10-stubs}
8176 allow you to override the linker's selection of stub instructions.
8177 @option{--power10-stubs=auto} allows the user to select the default
8192 @section @command{ld} and S/390 ELF Support
8194 @cindex S/390 ELF options
8198 @kindex --s390-pgste
8200 This option marks the result file with a @code{PT_S390_PGSTE}
8201 segment. The Linux kernel is supposed to allocate 4k page tables for
8202 binaries marked that way.
8216 @section @command{ld} and SPU ELF Support
8218 @cindex SPU ELF options
8224 This option marks an executable as a PIC plugin module.
8226 @cindex SPU overlays
8227 @kindex --no-overlays
8229 Normally, @command{ld} recognizes calls to functions within overlay
8230 regions, and redirects such calls to an overlay manager via a stub.
8231 @command{ld} also provides a built-in overlay manager. This option
8232 turns off all this special overlay handling.
8234 @cindex SPU overlay stub symbols
8235 @kindex --emit-stub-syms
8236 @item --emit-stub-syms
8237 This option causes @command{ld} to label overlay stubs with a local
8238 symbol that encodes the stub type and destination.
8240 @cindex SPU extra overlay stubs
8241 @kindex --extra-overlay-stubs
8242 @item --extra-overlay-stubs
8243 This option causes @command{ld} to add overlay call stubs on all
8244 function calls out of overlay regions. Normally stubs are not added
8245 on calls to non-overlay regions.
8247 @cindex SPU local store size
8248 @kindex --local-store=lo:hi
8249 @item --local-store=lo:hi
8250 @command{ld} usually checks that a final executable for SPU fits in
8251 the address range 0 to 256k. This option may be used to change the
8252 range. Disable the check entirely with @option{--local-store=0:0}.
8255 @kindex --stack-analysis
8256 @item --stack-analysis
8257 SPU local store space is limited. Over-allocation of stack space
8258 unnecessarily limits space available for code and data, while
8259 under-allocation results in runtime failures. If given this option,
8260 @command{ld} will provide an estimate of maximum stack usage.
8261 @command{ld} does this by examining symbols in code sections to
8262 determine the extents of functions, and looking at function prologues
8263 for stack adjusting instructions. A call-graph is created by looking
8264 for relocations on branch instructions. The graph is then searched
8265 for the maximum stack usage path. Note that this analysis does not
8266 find calls made via function pointers, and does not handle recursion
8267 and other cycles in the call graph. Stack usage may be
8268 under-estimated if your code makes such calls. Also, stack usage for
8269 dynamic allocation, e.g. alloca, will not be detected. If a link map
8270 is requested, detailed information about each function's stack usage
8271 and calls will be given.
8274 @kindex --emit-stack-syms
8275 @item --emit-stack-syms
8276 This option, if given along with @option{--stack-analysis} will result
8277 in @command{ld} emitting stack sizing symbols for each function.
8278 These take the form @code{__stack_<function_name>} for global
8279 functions, and @code{__stack_<number>_<function_name>} for static
8280 functions. @code{<number>} is the section id in hex. The value of
8281 such symbols is the stack requirement for the corresponding function.
8282 The symbol size will be zero, type @code{STT_NOTYPE}, binding
8283 @code{STB_LOCAL}, and section @code{SHN_ABS}.
8297 @section @command{ld}'s Support for Various TI COFF Versions
8298 @cindex TI COFF versions
8299 @kindex --format=@var{version}
8300 The @samp{--format} switch allows selection of one of the various
8301 TI COFF versions. The latest of this writing is 2; versions 0 and 1 are
8302 also supported. The TI COFF versions also vary in header byte-order
8303 format; @command{ld} will read any version or byte order, but the output
8304 header format depends on the default specified by the specific target.
8317 @section @command{ld} and WIN32 (cygwin/mingw)
8319 This section describes some of the win32 specific @command{ld} issues.
8320 See @ref{Options,,Command-line Options} for detailed description of the
8321 command-line options mentioned here.
8324 @cindex import libraries
8325 @item import libraries
8326 The standard Windows linker creates and uses so-called import
8327 libraries, which contains information for linking to dll's. They are
8328 regular static archives and are handled as any other static
8329 archive. The cygwin and mingw ports of @command{ld} have specific
8330 support for creating such libraries provided with the
8331 @samp{--out-implib} command-line option.
8333 @item exporting DLL symbols
8334 @cindex exporting DLL symbols
8335 The cygwin/mingw @command{ld} has several ways to export symbols for dll's.
8338 @item using auto-export functionality
8339 @cindex using auto-export functionality
8340 By default @command{ld} exports symbols with the auto-export functionality,
8341 which is controlled by the following command-line options:
8344 @item --export-all-symbols [This is the default]
8345 @item --exclude-symbols
8346 @item --exclude-libs
8347 @item --exclude-modules-for-implib
8348 @item --version-script
8351 When auto-export is in operation, @command{ld} will export all the non-local
8352 (global and common) symbols it finds in a DLL, with the exception of a few
8353 symbols known to belong to the system's runtime and libraries. As it will
8354 often not be desirable to export all of a DLL's symbols, which may include
8355 private functions that are not part of any public interface, the command-line
8356 options listed above may be used to filter symbols out from the list for
8357 exporting. The @samp{--output-def} option can be used in order to see the
8358 final list of exported symbols with all exclusions taken into effect.
8360 If @samp{--export-all-symbols} is not given explicitly on the
8361 command line, then the default auto-export behavior will be @emph{disabled}
8362 if either of the following are true:
8365 @item A DEF file is used.
8366 @item Any symbol in any object file was marked with the __declspec(dllexport) attribute.
8369 @item using a DEF file
8370 @cindex using a DEF file
8371 Another way of exporting symbols is using a DEF file. A DEF file is
8372 an ASCII file containing definitions of symbols which should be
8373 exported when a dll is created. Usually it is named @samp{<dll
8374 name>.def} and is added as any other object file to the linker's
8375 command line. The file's name must end in @samp{.def} or @samp{.DEF}.
8378 gcc -o <output> <objectfiles> <dll name>.def
8381 Using a DEF file turns off the normal auto-export behavior, unless the
8382 @samp{--export-all-symbols} option is also used.
8384 Here is an example of a DEF file for a shared library called @samp{xyz.dll}:
8387 LIBRARY "xyz.dll" BASE=0x20000000
8393 another_foo = abc.dll.afoo
8399 This example defines a DLL with a non-default base address and seven
8400 symbols in the export table. The third exported symbol @code{_bar} is an
8401 alias for the second. The fourth symbol, @code{another_foo} is resolved
8402 by "forwarding" to another module and treating it as an alias for
8403 @code{afoo} exported from the DLL @samp{abc.dll}. The final symbol
8404 @code{var1} is declared to be a data object. The @samp{doo} symbol in
8405 export library is an alias of @samp{foo}, which gets the string name
8406 in export table @samp{foo2}. The @samp{eoo} symbol is an data export
8407 symbol, which gets in export table the name @samp{var1}.
8409 The optional @code{LIBRARY <name>} command indicates the @emph{internal}
8410 name of the output DLL. If @samp{<name>} does not include a suffix,
8411 the default library suffix, @samp{.DLL} is appended.
8413 When the .DEF file is used to build an application, rather than a
8414 library, the @code{NAME <name>} command should be used instead of
8415 @code{LIBRARY}. If @samp{<name>} does not include a suffix, the default
8416 executable suffix, @samp{.EXE} is appended.
8418 With either @code{LIBRARY <name>} or @code{NAME <name>} the optional
8419 specification @code{BASE = <number>} may be used to specify a
8420 non-default base address for the image.
8422 If neither @code{LIBRARY <name>} nor @code{NAME <name>} is specified,
8423 or they specify an empty string, the internal name is the same as the
8424 filename specified on the command line.
8426 The complete specification of an export symbol is:
8430 ( ( ( <name1> [ = <name2> ] )
8431 | ( <name1> = <module-name> . <external-name>))
8432 [ @@ <integer> ] [NONAME] [DATA] [CONSTANT] [PRIVATE] [== <name3>] ) *
8435 Declares @samp{<name1>} as an exported symbol from the DLL, or declares
8436 @samp{<name1>} as an exported alias for @samp{<name2>}; or declares
8437 @samp{<name1>} as a "forward" alias for the symbol
8438 @samp{<external-name>} in the DLL @samp{<module-name>}.
8439 Optionally, the symbol may be exported by the specified ordinal
8440 @samp{<integer>} alias. The optional @samp{<name3>} is the to be used
8441 string in import/export table for the symbol.
8443 The optional keywords that follow the declaration indicate:
8445 @code{NONAME}: Do not put the symbol name in the DLL's export table. It
8446 will still be exported by its ordinal alias (either the value specified
8447 by the .def specification or, otherwise, the value assigned by the
8448 linker). The symbol name, however, does remain visible in the import
8449 library (if any), unless @code{PRIVATE} is also specified.
8451 @code{DATA}: The symbol is a variable or object, rather than a function.
8452 The import lib will export only an indirect reference to @code{foo} as
8453 the symbol @code{_imp__foo} (ie, @code{foo} must be resolved as
8456 @code{CONSTANT}: Like @code{DATA}, but put the undecorated @code{foo} as
8457 well as @code{_imp__foo} into the import library. Both refer to the
8458 read-only import address table's pointer to the variable, not to the
8459 variable itself. This can be dangerous. If the user code fails to add
8460 the @code{dllimport} attribute and also fails to explicitly add the
8461 extra indirection that the use of the attribute enforces, the
8462 application will behave unexpectedly.
8464 @code{PRIVATE}: Put the symbol in the DLL's export table, but do not put
8465 it into the static import library used to resolve imports at link time. The
8466 symbol can still be imported using the @code{LoadLibrary/GetProcAddress}
8467 API at runtime or by using the GNU ld extension of linking directly to
8468 the DLL without an import library.
8470 See ld/deffilep.y in the binutils sources for the full specification of
8471 other DEF file statements
8473 @cindex creating a DEF file
8474 While linking a shared dll, @command{ld} is able to create a DEF file
8475 with the @samp{--output-def <file>} command-line option.
8477 @item Using decorations
8478 @cindex Using decorations
8479 Another way of marking symbols for export is to modify the source code
8480 itself, so that when building the DLL each symbol to be exported is
8484 __declspec(dllexport) int a_variable
8485 __declspec(dllexport) void a_function(int with_args)
8488 All such symbols will be exported from the DLL. If, however,
8489 any of the object files in the DLL contain symbols decorated in
8490 this way, then the normal auto-export behavior is disabled, unless
8491 the @samp{--export-all-symbols} option is also used.
8493 Note that object files that wish to access these symbols must @emph{not}
8494 decorate them with dllexport. Instead, they should use dllimport,
8498 __declspec(dllimport) int a_variable
8499 __declspec(dllimport) void a_function(int with_args)
8502 This complicates the structure of library header files, because
8503 when included by the library itself the header must declare the
8504 variables and functions as dllexport, but when included by client
8505 code the header must declare them as dllimport. There are a number
8506 of idioms that are typically used to do this; often client code can
8507 omit the __declspec() declaration completely. See
8508 @samp{--enable-auto-import} and @samp{automatic data imports} for more
8512 @cindex automatic data imports
8513 @item automatic data imports
8514 The standard Windows dll format supports data imports from dlls only
8515 by adding special decorations (dllimport/dllexport), which let the
8516 compiler produce specific assembler instructions to deal with this
8517 issue. This increases the effort necessary to port existing Un*x
8518 code to these platforms, especially for large
8519 c++ libraries and applications. The auto-import feature, which was
8520 initially provided by Paul Sokolovsky, allows one to omit the
8521 decorations to achieve a behavior that conforms to that on POSIX/Un*x
8522 platforms. This feature is enabled with the @samp{--enable-auto-import}
8523 command-line option, although it is enabled by default on cygwin/mingw.
8524 The @samp{--enable-auto-import} option itself now serves mainly to
8525 suppress any warnings that are ordinarily emitted when linked objects
8526 trigger the feature's use.
8528 auto-import of variables does not always work flawlessly without
8529 additional assistance. Sometimes, you will see this message
8531 "variable '<var>' can't be auto-imported. Please read the
8532 documentation for ld's @code{--enable-auto-import} for details."
8534 The @samp{--enable-auto-import} documentation explains why this error
8535 occurs, and several methods that can be used to overcome this difficulty.
8536 One of these methods is the @emph{runtime pseudo-relocs} feature, described
8539 @cindex runtime pseudo-relocation
8540 For complex variables imported from DLLs (such as structs or classes),
8541 object files typically contain a base address for the variable and an
8542 offset (@emph{addend}) within the variable--to specify a particular
8543 field or public member, for instance. Unfortunately, the runtime loader used
8544 in win32 environments is incapable of fixing these references at runtime
8545 without the additional information supplied by dllimport/dllexport decorations.
8546 The standard auto-import feature described above is unable to resolve these
8549 The @samp{--enable-runtime-pseudo-relocs} switch allows these references to
8550 be resolved without error, while leaving the task of adjusting the references
8551 themselves (with their non-zero addends) to specialized code provided by the
8552 runtime environment. Recent versions of the cygwin and mingw environments and
8553 compilers provide this runtime support; older versions do not. However, the
8554 support is only necessary on the developer's platform; the compiled result will
8555 run without error on an older system.
8557 @samp{--enable-runtime-pseudo-relocs} is not the default; it must be explicitly
8560 @cindex direct linking to a dll
8561 @item direct linking to a dll
8562 The cygwin/mingw ports of @command{ld} support the direct linking,
8563 including data symbols, to a dll without the usage of any import
8564 libraries. This is much faster and uses much less memory than does the
8565 traditional import library method, especially when linking large
8566 libraries or applications. When @command{ld} creates an import lib, each
8567 function or variable exported from the dll is stored in its own bfd, even
8568 though a single bfd could contain many exports. The overhead involved in
8569 storing, loading, and processing so many bfd's is quite large, and explains the
8570 tremendous time, memory, and storage needed to link against particularly
8571 large or complex libraries when using import libs.
8573 Linking directly to a dll uses no extra command-line switches other than
8574 @samp{-L} and @samp{-l}, because @command{ld} already searches for a number
8575 of names to match each library. All that is needed from the developer's
8576 perspective is an understanding of this search, in order to force ld to
8577 select the dll instead of an import library.
8580 For instance, when ld is called with the argument @samp{-lxxx} it will attempt
8581 to find, in the first directory of its search path,
8594 before moving on to the next directory in the search path.
8596 (*) Actually, this is not @samp{cygxxx.dll} but in fact is @samp{<prefix>xxx.dll},
8597 where @samp{<prefix>} is set by the @command{ld} option
8598 @samp{--dll-search-prefix=<prefix>}. In the case of cygwin, the standard gcc spec
8599 file includes @samp{--dll-search-prefix=cyg}, so in effect we actually search for
8602 Other win32-based unix environments, such as mingw or pw32, may use other
8603 @samp{<prefix>}es, although at present only cygwin makes use of this feature. It
8604 was originally intended to help avoid name conflicts among dll's built for the
8605 various win32/un*x environments, so that (for example) two versions of a zlib dll
8606 could coexist on the same machine.
8608 The generic cygwin/mingw path layout uses a @samp{bin} directory for
8609 applications and dll's and a @samp{lib} directory for the import
8610 libraries (using cygwin nomenclature):
8616 libxxx.dll.a (in case of dll's)
8617 libxxx.a (in case of static archive)
8620 Linking directly to a dll without using the import library can be
8623 1. Use the dll directly by adding the @samp{bin} path to the link line
8625 gcc -Wl,-verbose -o a.exe -L../bin/ -lxxx
8628 However, as the dll's often have version numbers appended to their names
8629 (@samp{cygncurses-5.dll}) this will often fail, unless one specifies
8630 @samp{-L../bin -lncurses-5} to include the version. Import libs are generally
8631 not versioned, and do not have this difficulty.
8633 2. Create a symbolic link from the dll to a file in the @samp{lib}
8634 directory according to the above mentioned search pattern. This
8635 should be used to avoid unwanted changes in the tools needed for
8639 ln -s bin/cygxxx.dll lib/[cyg|lib|]xxx.dll[.a]
8642 Then you can link without any make environment changes.
8645 gcc -Wl,-verbose -o a.exe -L../lib/ -lxxx
8648 This technique also avoids the version number problems, because the following is
8655 libxxx.dll.a -> ../bin/cygxxx-5.dll
8658 Linking directly to a dll without using an import lib will work
8659 even when auto-import features are exercised, and even when
8660 @samp{--enable-runtime-pseudo-relocs} is used.
8662 Given the improvements in speed and memory usage, one might justifiably
8663 wonder why import libraries are used at all. There are three reasons:
8665 1. Until recently, the link-directly-to-dll functionality did @emph{not}
8666 work with auto-imported data.
8668 2. Sometimes it is necessary to include pure static objects within the
8669 import library (which otherwise contains only bfd's for indirection
8670 symbols that point to the exports of a dll). Again, the import lib
8671 for the cygwin kernel makes use of this ability, and it is not
8672 possible to do this without an import lib.
8674 3. Symbol aliases can only be resolved using an import lib. This is
8675 critical when linking against OS-supplied dll's (eg, the win32 API)
8676 in which symbols are usually exported as undecorated aliases of their
8677 stdcall-decorated assembly names.
8679 So, import libs are not going away. But the ability to replace
8680 true import libs with a simple symbolic link to (or a copy of)
8681 a dll, in many cases, is a useful addition to the suite of tools
8682 binutils makes available to the win32 developer. Given the
8683 massive improvements in memory requirements during linking, storage
8684 requirements, and linking speed, we expect that many developers
8685 will soon begin to use this feature whenever possible.
8687 @item symbol aliasing
8689 @item adding additional names
8690 Sometimes, it is useful to export symbols with additional names.
8691 A symbol @samp{foo} will be exported as @samp{foo}, but it can also be
8692 exported as @samp{_foo} by using special directives in the DEF file
8693 when creating the dll. This will affect also the optional created
8694 import library. Consider the following DEF file:
8697 LIBRARY "xyz.dll" BASE=0x61000000
8704 The line @samp{_foo = foo} maps the symbol @samp{foo} to @samp{_foo}.
8706 Another method for creating a symbol alias is to create it in the
8707 source code using the "weak" attribute:
8710 void foo () @{ /* Do something. */; @}
8711 void _foo () __attribute__ ((weak, alias ("foo")));
8714 See the gcc manual for more information about attributes and weak
8717 @item renaming symbols
8718 Sometimes it is useful to rename exports. For instance, the cygwin
8719 kernel does this regularly. A symbol @samp{_foo} can be exported as
8720 @samp{foo} but not as @samp{_foo} by using special directives in the
8721 DEF file. (This will also affect the import library, if it is
8722 created). In the following example:
8725 LIBRARY "xyz.dll" BASE=0x61000000
8731 The line @samp{_foo = foo} maps the exported symbol @samp{foo} to
8735 Note: using a DEF file disables the default auto-export behavior,
8736 unless the @samp{--export-all-symbols} command-line option is used.
8737 If, however, you are trying to rename symbols, then you should list
8738 @emph{all} desired exports in the DEF file, including the symbols
8739 that are not being renamed, and do @emph{not} use the
8740 @samp{--export-all-symbols} option. If you list only the
8741 renamed symbols in the DEF file, and use @samp{--export-all-symbols}
8742 to handle the other symbols, then the both the new names @emph{and}
8743 the original names for the renamed symbols will be exported.
8744 In effect, you'd be aliasing those symbols, not renaming them,
8745 which is probably not what you wanted.
8747 @cindex weak externals
8748 @item weak externals
8749 The Windows object format, PE, specifies a form of weak symbols called
8750 weak externals. When a weak symbol is linked and the symbol is not
8751 defined, the weak symbol becomes an alias for some other symbol. There
8752 are three variants of weak externals:
8754 @item Definition is searched for in objects and libraries, historically
8755 called lazy externals.
8756 @item Definition is searched for only in other objects, not in libraries.
8757 This form is not presently implemented.
8758 @item No search; the symbol is an alias. This form is not presently
8761 As a GNU extension, weak symbols that do not specify an alternate symbol
8762 are supported. If the symbol is undefined when linking, the symbol
8763 uses a default value.
8765 @cindex aligned common symbols
8766 @item aligned common symbols
8767 As a GNU extension to the PE file format, it is possible to specify the
8768 desired alignment for a common symbol. This information is conveyed from
8769 the assembler or compiler to the linker by means of GNU-specific commands
8770 carried in the object file's @samp{.drectve} section, which are recognized
8771 by @command{ld} and respected when laying out the common symbols. Native
8772 tools will be able to process object files employing this GNU extension,
8773 but will fail to respect the alignment instructions, and may issue noisy
8774 warnings about unknown linker directives.
8789 @section @code{ld} and Xtensa Processors
8791 @cindex Xtensa processors
8792 The default @command{ld} behavior for Xtensa processors is to interpret
8793 @code{SECTIONS} commands so that lists of explicitly named sections in a
8794 specification with a wildcard file will be interleaved when necessary to
8795 keep literal pools within the range of PC-relative load offsets. For
8796 example, with the command:
8808 @command{ld} may interleave some of the @code{.literal}
8809 and @code{.text} sections from different object files to ensure that the
8810 literal pools are within the range of PC-relative load offsets. A valid
8811 interleaving might place the @code{.literal} sections from an initial
8812 group of files followed by the @code{.text} sections of that group of
8813 files. Then, the @code{.literal} sections from the rest of the files
8814 and the @code{.text} sections from the rest of the files would follow.
8816 @cindex @option{--relax} on Xtensa
8817 @cindex relaxing on Xtensa
8818 Relaxation is enabled by default for the Xtensa version of @command{ld} and
8819 provides two important link-time optimizations. The first optimization
8820 is to combine identical literal values to reduce code size. A redundant
8821 literal will be removed and all the @code{L32R} instructions that use it
8822 will be changed to reference an identical literal, as long as the
8823 location of the replacement literal is within the offset range of all
8824 the @code{L32R} instructions. The second optimization is to remove
8825 unnecessary overhead from assembler-generated ``longcall'' sequences of
8826 @code{L32R}/@code{CALLX@var{n}} when the target functions are within
8827 range of direct @code{CALL@var{n}} instructions.
8829 For each of these cases where an indirect call sequence can be optimized
8830 to a direct call, the linker will change the @code{CALLX@var{n}}
8831 instruction to a @code{CALL@var{n}} instruction, remove the @code{L32R}
8832 instruction, and remove the literal referenced by the @code{L32R}
8833 instruction if it is not used for anything else. Removing the
8834 @code{L32R} instruction always reduces code size but can potentially
8835 hurt performance by changing the alignment of subsequent branch targets.
8836 By default, the linker will always preserve alignments, either by
8837 switching some instructions between 24-bit encodings and the equivalent
8838 density instructions or by inserting a no-op in place of the @code{L32R}
8839 instruction that was removed. If code size is more important than
8840 performance, the @option{--size-opt} option can be used to prevent the
8841 linker from widening density instructions or inserting no-ops, except in
8842 a few cases where no-ops are required for correctness.
8844 The following Xtensa-specific command-line options can be used to
8847 @cindex Xtensa options
8850 When optimizing indirect calls to direct calls, optimize for code size
8851 more than performance. With this option, the linker will not insert
8852 no-ops or widen density instructions to preserve branch target
8853 alignment. There may still be some cases where no-ops are required to
8854 preserve the correctness of the code.
8856 @item --abi-windowed
8858 Choose ABI for the output object and for the generated PLT code.
8859 PLT code inserted by the linker must match ABI of the output object
8860 because windowed and call0 ABI use incompatible function call
8862 Default ABI is chosen by the ABI tag in the @code{.xtensa.info} section
8863 of the first input object.
8864 A warning is issued if ABI tags of input objects do not match each other
8865 or the chosen output object ABI.
8873 @ifclear SingleFormat
8878 @cindex object file management
8879 @cindex object formats available
8881 The linker accesses object and archive files using the BFD libraries.
8882 These libraries allow the linker to use the same routines to operate on
8883 object files whatever the object file format. A different object file
8884 format can be supported simply by creating a new BFD back end and adding
8885 it to the library. To conserve runtime memory, however, the linker and
8886 associated tools are usually configured to support only a subset of the
8887 object file formats available. You can use @code{objdump -i}
8888 (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to
8889 list all the formats available for your configuration.
8891 @cindex BFD requirements
8892 @cindex requirements for BFD
8893 As with most implementations, BFD is a compromise between
8894 several conflicting requirements. The major factor influencing
8895 BFD design was efficiency: any time used converting between
8896 formats is time which would not have been spent had BFD not
8897 been involved. This is partly offset by abstraction payback; since
8898 BFD simplifies applications and back ends, more time and care
8899 may be spent optimizing algorithms for a greater speed.
8901 One minor artifact of the BFD solution which you should bear in
8902 mind is the potential for information loss. There are two places where
8903 useful information can be lost using the BFD mechanism: during
8904 conversion and during output. @xref{BFD information loss}.
8907 * BFD outline:: How it works: an outline of BFD
8911 @section How It Works: An Outline of BFD
8912 @cindex opening object files
8913 @include bfdsumm.texi
8916 @node Reporting Bugs
8917 @chapter Reporting Bugs
8918 @cindex bugs in @command{ld}
8919 @cindex reporting bugs in @command{ld}
8921 Your bug reports play an essential role in making @command{ld} reliable.
8923 Reporting a bug may help you by bringing a solution to your problem, or
8924 it may not. But in any case the principal function of a bug report is
8925 to help the entire community by making the next version of @command{ld}
8926 work better. Bug reports are your contribution to the maintenance of
8929 In order for a bug report to serve its purpose, you must include the
8930 information that enables us to fix the bug.
8933 * Bug Criteria:: Have you found a bug?
8934 * Bug Reporting:: How to report bugs
8938 @section Have You Found a Bug?
8939 @cindex bug criteria
8941 If you are not sure whether you have found a bug, here are some guidelines:
8944 @cindex fatal signal
8945 @cindex linker crash
8946 @cindex crash of linker
8948 If the linker gets a fatal signal, for any input whatever, that is a
8949 @command{ld} bug. Reliable linkers never crash.
8951 @cindex error on valid input
8953 If @command{ld} produces an error message for valid input, that is a bug.
8955 @cindex invalid input
8957 If @command{ld} does not produce an error message for invalid input, that
8958 may be a bug. In the general case, the linker can not verify that
8959 object files are correct.
8962 If you are an experienced user of linkers, your suggestions for
8963 improvement of @command{ld} are welcome in any case.
8967 @section How to Report Bugs
8969 @cindex @command{ld} bugs, reporting
8971 A number of companies and individuals offer support for @sc{gnu}
8972 products. If you obtained @command{ld} from a support organization, we
8973 recommend you contact that organization first.
8975 You can find contact information for many support companies and
8976 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
8980 Otherwise, send bug reports for @command{ld} to
8984 The fundamental principle of reporting bugs usefully is this:
8985 @strong{report all the facts}. If you are not sure whether to state a
8986 fact or leave it out, state it!
8988 Often people omit facts because they think they know what causes the
8989 problem and assume that some details do not matter. Thus, you might
8990 assume that the name of a symbol you use in an example does not
8991 matter. Well, probably it does not, but one cannot be sure. Perhaps
8992 the bug is a stray memory reference which happens to fetch from the
8993 location where that name is stored in memory; perhaps, if the name
8994 were different, the contents of that location would fool the linker
8995 into doing the right thing despite the bug. Play it safe and give a
8996 specific, complete example. That is the easiest thing for you to do,
8997 and the most helpful.
8999 Keep in mind that the purpose of a bug report is to enable us to fix
9000 the bug if it is new to us. Therefore, always write your bug reports
9001 on the assumption that the bug has not been reported previously.
9003 Sometimes people give a few sketchy facts and ask, ``Does this ring a
9004 bell?'' This cannot help us fix a bug, so it is basically useless. We
9005 respond by asking for enough details to enable us to investigate.
9006 You might as well expedite matters by sending them to begin with.
9008 To enable us to fix the bug, you should include all these things:
9012 The version of @command{ld}. @command{ld} announces it if you start it with
9013 the @samp{--version} argument.
9015 Without this, we will not know whether there is any point in looking for
9016 the bug in the current version of @command{ld}.
9019 Any patches you may have applied to the @command{ld} source, including any
9020 patches made to the @code{BFD} library.
9023 The type of machine you are using, and the operating system name and
9027 What compiler (and its version) was used to compile @command{ld}---e.g.
9031 The command arguments you gave the linker to link your example and
9032 observe the bug. To guarantee you will not omit something important,
9033 list them all. A copy of the Makefile (or the output from make) is
9036 If we were to try to guess the arguments, we would probably guess wrong
9037 and then we might not encounter the bug.
9040 A complete input file, or set of input files, that will reproduce the
9041 bug. It is generally most helpful to send the actual object files
9042 provided that they are reasonably small. Say no more than 10K. For
9043 bigger files you can either make them available by FTP or HTTP or else
9044 state that you are willing to send the object file(s) to whomever
9045 requests them. (Note - your email will be going to a mailing list, so
9046 we do not want to clog it up with large attachments). But small
9047 attachments are best.
9049 If the source files were assembled using @code{gas} or compiled using
9050 @code{gcc}, then it may be OK to send the source files rather than the
9051 object files. In this case, be sure to say exactly what version of
9052 @code{gas} or @code{gcc} was used to produce the object files. Also say
9053 how @code{gas} or @code{gcc} were configured.
9056 A description of what behavior you observe that you believe is
9057 incorrect. For example, ``It gets a fatal signal.''
9059 Of course, if the bug is that @command{ld} gets a fatal signal, then we
9060 will certainly notice it. But if the bug is incorrect output, we might
9061 not notice unless it is glaringly wrong. You might as well not give us
9062 a chance to make a mistake.
9064 Even if the problem you experience is a fatal signal, you should still
9065 say so explicitly. Suppose something strange is going on, such as, your
9066 copy of @command{ld} is out of sync, or you have encountered a bug in the
9067 C library on your system. (This has happened!) Your copy might crash
9068 and ours would not. If you told us to expect a crash, then when ours
9069 fails to crash, we would know that the bug was not happening for us. If
9070 you had not told us to expect a crash, then we would not be able to draw
9071 any conclusion from our observations.
9074 If you wish to suggest changes to the @command{ld} source, send us context
9075 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or
9076 @samp{-p} option. Always send diffs from the old file to the new file.
9077 If you even discuss something in the @command{ld} source, refer to it by
9078 context, not by line number.
9080 The line numbers in our development sources will not match those in your
9081 sources. Your line numbers would convey no useful information to us.
9084 Here are some things that are not necessary:
9088 A description of the envelope of the bug.
9090 Often people who encounter a bug spend a lot of time investigating
9091 which changes to the input file will make the bug go away and which
9092 changes will not affect it.
9094 This is often time consuming and not very useful, because the way we
9095 will find the bug is by running a single example under the debugger
9096 with breakpoints, not by pure deduction from a series of examples.
9097 We recommend that you save your time for something else.
9099 Of course, if you can find a simpler example to report @emph{instead}
9100 of the original one, that is a convenience for us. Errors in the
9101 output will be easier to spot, running under the debugger will take
9102 less time, and so on.
9104 However, simplification is not vital; if you do not want to do this,
9105 report the bug anyway and send us the entire test case you used.
9108 A patch for the bug.
9110 A patch for the bug does help us if it is a good one. But do not omit
9111 the necessary information, such as the test case, on the assumption that
9112 a patch is all we need. We might see problems with your patch and decide
9113 to fix the problem another way, or we might not understand it at all.
9115 Sometimes with a program as complicated as @command{ld} it is very hard to
9116 construct an example that will make the program follow a certain path
9117 through the code. If you do not send us the example, we will not be
9118 able to construct one, so we will not be able to verify that the bug is
9121 And if we cannot understand what bug you are trying to fix, or why your
9122 patch should be an improvement, we will not install it. A test case will
9123 help us to understand.
9126 A guess about what the bug is or what it depends on.
9128 Such guesses are usually wrong. Even we cannot guess right about such
9129 things without first using the debugger to find the facts.
9133 @appendix MRI Compatible Script Files
9134 @cindex MRI compatibility
9135 To aid users making the transition to @sc{gnu} @command{ld} from the MRI
9136 linker, @command{ld} can use MRI compatible linker scripts as an
9137 alternative to the more general-purpose linker scripting language
9138 described in @ref{Scripts}. MRI compatible linker scripts have a much
9139 simpler command set than the scripting language otherwise used with
9140 @command{ld}. @sc{gnu} @command{ld} supports the most commonly used MRI
9141 linker commands; these commands are described here.
9143 In general, MRI scripts aren't of much use with the @code{a.out} object
9144 file format, since it only has three sections and MRI scripts lack some
9145 features to make use of them.
9147 You can specify a file containing an MRI-compatible script using the
9148 @samp{-c} command-line option.
9150 Each command in an MRI-compatible script occupies its own line; each
9151 command line starts with the keyword that identifies the command (though
9152 blank lines are also allowed for punctuation). If a line of an
9153 MRI-compatible script begins with an unrecognized keyword, @command{ld}
9154 issues a warning message, but continues processing the script.
9156 Lines beginning with @samp{*} are comments.
9158 You can write these commands using all upper-case letters, or all
9159 lower case; for example, @samp{chip} is the same as @samp{CHIP}.
9160 The following list shows only the upper-case form of each command.
9163 @cindex @code{ABSOLUTE} (MRI)
9164 @item ABSOLUTE @var{secname}
9165 @itemx ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname}
9166 Normally, @command{ld} includes in the output file all sections from all
9167 the input files. However, in an MRI-compatible script, you can use the
9168 @code{ABSOLUTE} command to restrict the sections that will be present in
9169 your output program. If the @code{ABSOLUTE} command is used at all in a
9170 script, then only the sections named explicitly in @code{ABSOLUTE}
9171 commands will appear in the linker output. You can still use other
9172 input sections (whatever you select on the command line, or using
9173 @code{LOAD}) to resolve addresses in the output file.
9175 @cindex @code{ALIAS} (MRI)
9176 @item ALIAS @var{out-secname}, @var{in-secname}
9177 Use this command to place the data from input section @var{in-secname}
9178 in a section called @var{out-secname} in the linker output file.
9180 @var{in-secname} may be an integer.
9182 @cindex @code{ALIGN} (MRI)
9183 @item ALIGN @var{secname} = @var{expression}
9184 Align the section called @var{secname} to @var{expression}. The
9185 @var{expression} should be a power of two.
9187 @cindex @code{BASE} (MRI)
9188 @item BASE @var{expression}
9189 Use the value of @var{expression} as the lowest address (other than
9190 absolute addresses) in the output file.
9192 @cindex @code{CHIP} (MRI)
9193 @item CHIP @var{expression}
9194 @itemx CHIP @var{expression}, @var{expression}
9195 This command does nothing; it is accepted only for compatibility.
9197 @cindex @code{END} (MRI)
9199 This command does nothing whatever; it's only accepted for compatibility.
9201 @cindex @code{FORMAT} (MRI)
9202 @item FORMAT @var{output-format}
9203 Similar to the @code{OUTPUT_FORMAT} command in the more general linker
9204 language, but restricted to S-records, if @var{output-format} is @samp{S}
9206 @cindex @code{LIST} (MRI)
9207 @item LIST @var{anything}@dots{}
9208 Print (to the standard output file) a link map, as produced by the
9209 @command{ld} command-line option @samp{-M}.
9211 The keyword @code{LIST} may be followed by anything on the
9212 same line, with no change in its effect.
9214 @cindex @code{LOAD} (MRI)
9215 @item LOAD @var{filename}
9216 @itemx LOAD @var{filename}, @var{filename}, @dots{} @var{filename}
9217 Include one or more object file @var{filename} in the link; this has the
9218 same effect as specifying @var{filename} directly on the @command{ld}
9221 @cindex @code{NAME} (MRI)
9222 @item NAME @var{output-name}
9223 @var{output-name} is the name for the program produced by @command{ld}; the
9224 MRI-compatible command @code{NAME} is equivalent to the command-line
9225 option @samp{-o} or the general script language command @code{OUTPUT}.
9227 @cindex @code{ORDER} (MRI)
9228 @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname}
9229 @itemx ORDER @var{secname} @var{secname} @var{secname}
9230 Normally, @command{ld} orders the sections in its output file in the
9231 order in which they first appear in the input files. In an MRI-compatible
9232 script, you can override this ordering with the @code{ORDER} command. The
9233 sections you list with @code{ORDER} will appear first in your output
9234 file, in the order specified.
9236 @cindex @code{PUBLIC} (MRI)
9237 @item PUBLIC @var{name}=@var{expression}
9238 @itemx PUBLIC @var{name},@var{expression}
9239 @itemx PUBLIC @var{name} @var{expression}
9240 Supply a value (@var{expression}) for external symbol
9241 @var{name} used in the linker input files.
9243 @cindex @code{SECT} (MRI)
9244 @item SECT @var{secname}, @var{expression}
9245 @itemx SECT @var{secname}=@var{expression}
9246 @itemx SECT @var{secname} @var{expression}
9247 You can use any of these three forms of the @code{SECT} command to
9248 specify the start address (@var{expression}) for section @var{secname}.
9249 If you have more than one @code{SECT} statement for the same
9250 @var{secname}, only the @emph{first} sets the start address.
9253 @node GNU Free Documentation License
9254 @appendix GNU Free Documentation License
9258 @unnumbered LD Index
9263 % I think something like @@colophon should be in texinfo. In the
9265 \long\def\colophon{\hbox to0pt{}\vfill
9266 \centerline{The body of this manual is set in}
9267 \centerline{\fontname\tenrm,}
9268 \centerline{with headings in {\bf\fontname\tenbf}}
9269 \centerline{and examples in {\tt\fontname\tentt}.}
9270 \centerline{{\it\fontname\tenit\/} and}
9271 \centerline{{\sl\fontname\tensl\/}}
9272 \centerline{are used for emphasis.}\vfill}
9274 % Blame: doc@@cygnus.com, 28mar91.