3 @c Copyright (C) 1991-2020 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{} 1991-2020 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{} 1991-2020 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 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 When generating an executable or shared library, mark it to tell the
1321 dynamic linker to defer function call resolution to the point when
1322 the function is called (lazy binding), rather than at load time.
1323 Lazy binding is the default.
1326 Specify that the object's filters be processed immediately at runtime.
1328 @item max-page-size=@var{value}
1329 Set the maximum memory page size supported to @var{value}.
1332 Allow multiple definitions.
1335 Disable linker generated .dynbss variables used in place of variables
1336 defined in shared libraries. May result in dynamic text relocations.
1339 Specify that the dynamic loader search for dependencies of this object
1340 should ignore any default library search paths.
1343 Specify that the object shouldn't be unloaded at runtime.
1346 Specify that the object is not available to @code{dlopen}.
1349 Specify that the object can not be dumped by @code{dldump}.
1352 Marks the object as not requiring executable stack.
1354 @item noextern-protected-data
1355 Don't treat protected data symbols as external when building a shared
1356 library. This option overrides the linker backend default. It can be
1357 used to work around incorrect relocations against protected data symbols
1358 generated by compiler. Updates on protected data symbols by another
1359 module aren't visible to the resulting shared library. Supported for
1362 @item noreloc-overflow
1363 Disable relocation overflow check. This can be used to disable
1364 relocation overflow check if there will be no dynamic relocation
1365 overflow at run-time. Supported for x86_64.
1368 When generating an executable or shared library, mark it to tell the
1369 dynamic linker to resolve all symbols when the program is started, or
1370 when the shared library is loaded by dlopen, instead of deferring
1371 function call resolution to the point when the function is first
1375 Specify that the object requires @samp{$ORIGIN} handling in paths.
1379 Create an ELF @code{PT_GNU_RELRO} segment header in the object. This
1380 specifies a memory segment that should be made read-only after
1381 relocation, if supported. Specifying @samp{common-page-size} smaller
1382 than the system page size will render this protection ineffective.
1383 Don't create an ELF @code{PT_GNU_RELRO} segment if @samp{norelro}.
1386 @itemx noseparate-code
1387 Create separate code @code{PT_LOAD} segment header in the object. This
1388 specifies a memory segment that should contain only instructions and must
1389 be in wholly disjoint pages from any other data. Don't create separate
1390 code @code{PT_LOAD} segment if @samp{noseparate-code} is used.
1393 @itemx nounique-symbol
1394 Avoid duplicated local symbol names in the symbol string table. Append
1395 ".@code{number}" to duplicated local symbol names if @samp{unique-symbol}
1396 is used. @option{nounique-symbol} is the default.
1399 Generate GNU_PROPERTY_X86_FEATURE_1_SHSTK in .note.gnu.property section
1400 to indicate compatibility with Intel Shadow Stack. Supported for
1401 Linux/i386 and Linux/x86_64.
1403 @item stack-size=@var{value}
1404 Specify a stack size for an ELF @code{PT_GNU_STACK} segment.
1405 Specifying zero will override any default non-zero sized
1406 @code{PT_GNU_STACK} segment creation.
1408 @item start-stop-visibility=@var{value}
1410 @cindex ELF symbol visibility
1411 Specify the ELF symbol visibility for synthesized
1412 @code{__start_SECNAME} and @code{__stop_SECNAME} symbols (@pxref{Input
1413 Section Example}). @var{value} must be exactly @samp{default},
1414 @samp{internal}, @samp{hidden}, or @samp{protected}. If no @samp{-z
1415 start-stop-visibility} option is given, @samp{protected} is used for
1416 compatibility with historical practice. However, it's highly
1417 recommended to use @samp{-z start-stop-visibility=hidden} in new
1418 programs and shared libraries so that these symbols are not exported
1419 between shared objects, which is not usually what's intended.
1424 Report an error if DT_TEXTREL is set, i.e., if the position-independent
1425 or shared object has dynamic relocations in read-only sections. Don't
1426 report an error if @samp{notext} or @samp{textoff}.
1429 Do not report unresolved symbol references from regular object files,
1430 either when creating an executable, or when creating a shared library.
1431 This option is the inverse of @samp{-z defs}.
1433 @item x86-64-baseline
1437 Specify the x86-64 ISA level needed in .note.gnu.property section.
1438 @option{x86-64-baseline} generates @code{GNU_PROPERTY_X86_ISA_1_BASELINE}.
1439 @option{x86-64-v2} generates @code{GNU_PROPERTY_X86_ISA_1_V2}.
1440 @option{x86-64-v3} generates @code{GNU_PROPERTY_X86_ISA_1_V3}.
1441 @option{x86-64-v4} generates @code{GNU_PROPERTY_X86_ISA_1_V4}.
1442 Supported for Linux/i386 and Linux/x86_64.
1446 Other keywords are ignored for Solaris compatibility.
1449 @cindex groups of archives
1450 @item -( @var{archives} -)
1451 @itemx --start-group @var{archives} --end-group
1452 The @var{archives} should be a list of archive files. They may be
1453 either explicit file names, or @samp{-l} options.
1455 The specified archives are searched repeatedly until no new undefined
1456 references are created. Normally, an archive is searched only once in
1457 the order that it is specified on the command line. If a symbol in that
1458 archive is needed to resolve an undefined symbol referred to by an
1459 object in an archive that appears later on the command line, the linker
1460 would not be able to resolve that reference. By grouping the archives,
1461 they will all be searched repeatedly until all possible references are
1464 Using this option has a significant performance cost. It is best to use
1465 it only when there are unavoidable circular references between two or
1468 @kindex --accept-unknown-input-arch
1469 @kindex --no-accept-unknown-input-arch
1470 @item --accept-unknown-input-arch
1471 @itemx --no-accept-unknown-input-arch
1472 Tells the linker to accept input files whose architecture cannot be
1473 recognised. The assumption is that the user knows what they are doing
1474 and deliberately wants to link in these unknown input files. This was
1475 the default behaviour of the linker, before release 2.14. The default
1476 behaviour from release 2.14 onwards is to reject such input files, and
1477 so the @samp{--accept-unknown-input-arch} option has been added to
1478 restore the old behaviour.
1481 @kindex --no-as-needed
1483 @itemx --no-as-needed
1484 This option affects ELF DT_NEEDED tags for dynamic libraries mentioned
1485 on the command line after the @option{--as-needed} option. Normally
1486 the linker will add a DT_NEEDED tag for each dynamic library mentioned
1487 on the command line, regardless of whether the library is actually
1488 needed or not. @option{--as-needed} causes a DT_NEEDED tag to only be
1489 emitted for a library that @emph{at that point in the link} satisfies a
1490 non-weak undefined symbol reference from a regular object file or, if
1491 the library is not found in the DT_NEEDED lists of other needed libraries, a
1492 non-weak undefined symbol reference from another needed dynamic library.
1493 Object files or libraries appearing on the command line @emph{after}
1494 the library in question do not affect whether the library is seen as
1495 needed. This is similar to the rules for extraction of object files
1496 from archives. @option{--no-as-needed} restores the default behaviour.
1498 @kindex --add-needed
1499 @kindex --no-add-needed
1501 @itemx --no-add-needed
1502 These two options have been deprecated because of the similarity of
1503 their names to the @option{--as-needed} and @option{--no-as-needed}
1504 options. They have been replaced by @option{--copy-dt-needed-entries}
1505 and @option{--no-copy-dt-needed-entries}.
1507 @kindex -assert @var{keyword}
1508 @item -assert @var{keyword}
1509 This option is ignored for SunOS compatibility.
1513 @kindex -call_shared
1517 Link against dynamic libraries. This is only meaningful on platforms
1518 for which shared libraries are supported. This option is normally the
1519 default on such platforms. The different variants of this option are
1520 for compatibility with various systems. You may use this option
1521 multiple times on the command line: it affects library searching for
1522 @option{-l} options which follow it.
1526 Set the @code{DF_1_GROUP} flag in the @code{DT_FLAGS_1} entry in the dynamic
1527 section. This causes the runtime linker to handle lookups in this
1528 object and its dependencies to be performed only inside the group.
1529 @option{--unresolved-symbols=report-all} is implied. This option is
1530 only meaningful on ELF platforms which support shared libraries.
1540 Do not link against shared libraries. This is only meaningful on
1541 platforms for which shared libraries are supported. The different
1542 variants of this option are for compatibility with various systems. You
1543 may use this option multiple times on the command line: it affects
1544 library searching for @option{-l} options which follow it. This
1545 option also implies @option{--unresolved-symbols=report-all}. This
1546 option can be used with @option{-shared}. Doing so means that a
1547 shared library is being created but that all of the library's external
1548 references must be resolved by pulling in entries from static
1553 When creating a shared library, bind references to global symbols to the
1554 definition within the shared library, if any. Normally, it is possible
1555 for a program linked against a shared library to override the definition
1556 within the shared library. This option is only meaningful on ELF
1557 platforms which support shared libraries.
1559 @kindex -Bsymbolic-functions
1560 @item -Bsymbolic-functions
1561 When creating a shared library, bind references to global function
1562 symbols to the definition within the shared library, if any.
1563 This option is only meaningful on ELF platforms which support shared
1566 @kindex --dynamic-list=@var{dynamic-list-file}
1567 @item --dynamic-list=@var{dynamic-list-file}
1568 Specify the name of a dynamic list file to the linker. This is
1569 typically used when creating shared libraries to specify a list of
1570 global symbols whose references shouldn't be bound to the definition
1571 within the shared library, or creating dynamically linked executables
1572 to specify a list of symbols which should be added to the symbol table
1573 in the executable. This option is only meaningful on ELF platforms
1574 which support shared libraries.
1576 The format of the dynamic list is the same as the version node without
1577 scope and node name. See @ref{VERSION} for more information.
1579 @kindex --dynamic-list-data
1580 @item --dynamic-list-data
1581 Include all global data symbols to the dynamic list.
1583 @kindex --dynamic-list-cpp-new
1584 @item --dynamic-list-cpp-new
1585 Provide the builtin dynamic list for C++ operator new and delete. It
1586 is mainly useful for building shared libstdc++.
1588 @kindex --dynamic-list-cpp-typeinfo
1589 @item --dynamic-list-cpp-typeinfo
1590 Provide the builtin dynamic list for C++ runtime type identification.
1592 @kindex --check-sections
1593 @kindex --no-check-sections
1594 @item --check-sections
1595 @itemx --no-check-sections
1596 Asks the linker @emph{not} to check section addresses after they have
1597 been assigned to see if there are any overlaps. Normally the linker will
1598 perform this check, and if it finds any overlaps it will produce
1599 suitable error messages. The linker does know about, and does make
1600 allowances for sections in overlays. The default behaviour can be
1601 restored by using the command-line switch @option{--check-sections}.
1602 Section overlap is not usually checked for relocatable links. You can
1603 force checking in that case by using the @option{--check-sections}
1606 @kindex --copy-dt-needed-entries
1607 @kindex --no-copy-dt-needed-entries
1608 @item --copy-dt-needed-entries
1609 @itemx --no-copy-dt-needed-entries
1610 This option affects the treatment of dynamic libraries referred to
1611 by DT_NEEDED tags @emph{inside} ELF dynamic libraries mentioned on the
1612 command line. Normally the linker won't add a DT_NEEDED tag to the
1613 output binary for each library mentioned in a DT_NEEDED tag in an
1614 input dynamic library. With @option{--copy-dt-needed-entries}
1615 specified on the command line however any dynamic libraries that
1616 follow it will have their DT_NEEDED entries added. The default
1617 behaviour can be restored with @option{--no-copy-dt-needed-entries}.
1619 This option also has an effect on the resolution of symbols in dynamic
1620 libraries. With @option{--copy-dt-needed-entries} dynamic libraries
1621 mentioned on the command line will be recursively searched, following
1622 their DT_NEEDED tags to other libraries, in order to resolve symbols
1623 required by the output binary. With the default setting however
1624 the searching of dynamic libraries that follow it will stop with the
1625 dynamic library itself. No DT_NEEDED links will be traversed to resolve
1628 @cindex cross reference table
1631 Output a cross reference table. If a linker map file is being
1632 generated, the cross reference table is printed to the map file.
1633 Otherwise, it is printed on the standard output.
1635 The format of the table is intentionally simple, so that it may be
1636 easily processed by a script if necessary. The symbols are printed out,
1637 sorted by name. For each symbol, a list of file names is given. If the
1638 symbol is defined, the first file listed is the location of the
1639 definition. If the symbol is defined as a common value then any files
1640 where this happens appear next. Finally any files that reference the
1643 @cindex ctf variables
1644 @kindex --ctf-variables
1645 @kindex --no-ctf-variables
1646 @item --ctf-variables
1647 @item --no-ctf-variables
1648 The CTF debuginfo format supports a section which encodes the names and
1649 types of variables found in the program which do not appear in any symbol
1650 table. These variables clearly cannot be looked up by address by
1651 conventional debuggers, so the space used for their types and names is
1652 usually wasted: the types are usually small but the names are often not.
1653 @option{--ctf-variables} causes the generation of such a section.
1654 The default behaviour can be restored with @option{--no-ctf-variables}.
1656 @cindex ctf type sharing
1657 @kindex --ctf-share-types
1658 @item --ctf-share-types=@var{method}
1659 Adjust the method used to share types between translation units in CTF.
1662 @item share-unconflicted
1663 Put all types that do not have ambiguous definitions into the shared dictionary,
1664 where debuggers can easily access them, even if they only occur in one
1665 translation unit. This is the default.
1667 @item share-duplicated
1668 Put only types that occur in multiple translation units into the shared
1669 dictionary: types with only one definition go into per-translation-unit
1670 dictionaries. Types with ambiguous definitions in multiple translation units
1671 always go into per-translation-unit dictionaries. This tends to make the CTF
1672 larger, but may reduce the amount of CTF in the shared dictionary. For very
1673 large projects this may speed up opening the CTF and save memory in the CTF
1674 consumer at runtime.
1677 @cindex common allocation
1678 @kindex --no-define-common
1679 @item --no-define-common
1680 This option inhibits the assignment of addresses to common symbols.
1681 The script command @code{INHIBIT_COMMON_ALLOCATION} has the same effect.
1682 @xref{Miscellaneous Commands}.
1684 The @samp{--no-define-common} option allows decoupling
1685 the decision to assign addresses to Common symbols from the choice
1686 of the output file type; otherwise a non-Relocatable output type
1687 forces assigning addresses to Common symbols.
1688 Using @samp{--no-define-common} allows Common symbols that are referenced
1689 from a shared library to be assigned addresses only in the main program.
1690 This eliminates the unused duplicate space in the shared library,
1691 and also prevents any possible confusion over resolving to the wrong
1692 duplicate when there are many dynamic modules with specialized search
1693 paths for runtime symbol resolution.
1695 @cindex group allocation in linker script
1696 @cindex section groups
1698 @kindex --force-group-allocation
1699 @item --force-group-allocation
1700 This option causes the linker to place section group members like
1701 normal input sections, and to delete the section groups. This is the
1702 default behaviour for a final link but this option can be used to
1703 change the behaviour of a relocatable link (@samp{-r}). The script
1704 command @code{FORCE_GROUP_ALLOCATION} has the same
1705 effect. @xref{Miscellaneous Commands}.
1707 @cindex symbols, from command line
1708 @kindex --defsym=@var{symbol}=@var{exp}
1709 @item --defsym=@var{symbol}=@var{expression}
1710 Create a global symbol in the output file, containing the absolute
1711 address given by @var{expression}. You may use this option as many
1712 times as necessary to define multiple symbols in the command line. A
1713 limited form of arithmetic is supported for the @var{expression} in this
1714 context: you may give a hexadecimal constant or the name of an existing
1715 symbol, or use @code{+} and @code{-} to add or subtract hexadecimal
1716 constants or symbols. If you need more elaborate expressions, consider
1717 using the linker command language from a script (@pxref{Assignments}).
1718 @emph{Note:} there should be no white space between @var{symbol}, the
1719 equals sign (``@key{=}''), and @var{expression}.
1721 The linker processes @samp{--defsym} arguments and @samp{-T} arguments
1722 in order, placing @samp{--defsym} before @samp{-T} will define the
1723 symbol before the linker script from @samp{-T} is processed, while
1724 placing @samp{--defsym} after @samp{-T} will define the symbol after
1725 the linker script has been processed. This difference has
1726 consequences for expressions within the linker script that use the
1727 @samp{--defsym} symbols, which order is correct will depend on what
1728 you are trying to achieve.
1730 @cindex demangling, from command line
1731 @kindex --demangle[=@var{style}]
1732 @kindex --no-demangle
1733 @item --demangle[=@var{style}]
1734 @itemx --no-demangle
1735 These options control whether to demangle symbol names in error messages
1736 and other output. When the linker is told to demangle, it tries to
1737 present symbol names in a readable fashion: it strips leading
1738 underscores if they are used by the object file format, and converts C++
1739 mangled symbol names into user readable names. Different compilers have
1740 different mangling styles. The optional demangling style argument can be used
1741 to choose an appropriate demangling style for your compiler. The linker will
1742 demangle by default unless the environment variable @samp{COLLECT_NO_DEMANGLE}
1743 is set. These options may be used to override the default.
1745 @cindex dynamic linker, from command line
1746 @kindex -I@var{file}
1747 @kindex --dynamic-linker=@var{file}
1749 @itemx --dynamic-linker=@var{file}
1750 Set the name of the dynamic linker. This is only meaningful when
1751 generating dynamically linked ELF executables. The default dynamic
1752 linker is normally correct; don't use this unless you know what you are
1755 @kindex --no-dynamic-linker
1756 @item --no-dynamic-linker
1757 When producing an executable file, omit the request for a dynamic
1758 linker to be used at load-time. This is only meaningful for ELF
1759 executables that contain dynamic relocations, and usually requires
1760 entry point code that is capable of processing these relocations.
1762 @kindex --embedded-relocs
1763 @item --embedded-relocs
1764 This option is similar to the @option{--emit-relocs} option except
1765 that the relocs are stored in a target-specific section. This option
1766 is only supported by the @samp{BFIN}, @samp{CR16} and @emph{M68K}
1769 @kindex --disable-multiple-abs-defs
1770 @item --disable-multiple-abs-defs
1771 Do not allow multiple definitions with symbols included
1772 in filename invoked by -R or --just-symbols
1774 @kindex --fatal-warnings
1775 @kindex --no-fatal-warnings
1776 @item --fatal-warnings
1777 @itemx --no-fatal-warnings
1778 Treat all warnings as errors. The default behaviour can be restored
1779 with the option @option{--no-fatal-warnings}.
1781 @kindex --force-exe-suffix
1782 @item --force-exe-suffix
1783 Make sure that an output file has a .exe suffix.
1785 If a successfully built fully linked output file does not have a
1786 @code{.exe} or @code{.dll} suffix, this option forces the linker to copy
1787 the output file to one of the same name with a @code{.exe} suffix. This
1788 option is useful when using unmodified Unix makefiles on a Microsoft
1789 Windows host, since some versions of Windows won't run an image unless
1790 it ends in a @code{.exe} suffix.
1792 @kindex --gc-sections
1793 @kindex --no-gc-sections
1794 @cindex garbage collection
1796 @itemx --no-gc-sections
1797 Enable garbage collection of unused input sections. It is ignored on
1798 targets that do not support this option. The default behaviour (of not
1799 performing this garbage collection) can be restored by specifying
1800 @samp{--no-gc-sections} on the command line. Note that garbage
1801 collection for COFF and PE format targets is supported, but the
1802 implementation is currently considered to be experimental.
1804 @samp{--gc-sections} decides which input sections are used by
1805 examining symbols and relocations. The section containing the entry
1806 symbol and all sections containing symbols undefined on the
1807 command-line will be kept, as will sections containing symbols
1808 referenced by dynamic objects. Note that when building shared
1809 libraries, the linker must assume that any visible symbol is
1810 referenced. Once this initial set of sections has been determined,
1811 the linker recursively marks as used any section referenced by their
1812 relocations. See @samp{--entry}, @samp{--undefined}, and
1813 @samp{--gc-keep-exported}.
1815 This option can be set when doing a partial link (enabled with option
1816 @samp{-r}). In this case the root of symbols kept must be explicitly
1817 specified either by one of the options @samp{--entry},
1818 @samp{--undefined}, or @samp{--gc-keep-exported} or by a @code{ENTRY}
1819 command in the linker script.
1821 As a GNU extension, ELF input sections marked with the
1822 @code{SHF_GNU_RETAIN} flag will not be garbage collected.
1824 @kindex --print-gc-sections
1825 @kindex --no-print-gc-sections
1826 @cindex garbage collection
1827 @item --print-gc-sections
1828 @itemx --no-print-gc-sections
1829 List all sections removed by garbage collection. The listing is
1830 printed on stderr. This option is only effective if garbage
1831 collection has been enabled via the @samp{--gc-sections}) option. The
1832 default behaviour (of not listing the sections that are removed) can
1833 be restored by specifying @samp{--no-print-gc-sections} on the command
1836 @kindex --gc-keep-exported
1837 @cindex garbage collection
1838 @item --gc-keep-exported
1839 When @samp{--gc-sections} is enabled, this option prevents garbage
1840 collection of unused input sections that contain global symbols having
1841 default or protected visibility. This option is intended to be used for
1842 executables where unreferenced sections would otherwise be garbage
1843 collected regardless of the external visibility of contained symbols.
1844 Note that this option has no effect when linking shared objects since
1845 it is already the default behaviour. This option is only supported for
1848 @kindex --print-output-format
1849 @cindex output format
1850 @item --print-output-format
1851 Print the name of the default output format (perhaps influenced by
1852 other command-line options). This is the string that would appear
1853 in an @code{OUTPUT_FORMAT} linker script command (@pxref{File Commands}).
1855 @kindex --print-memory-usage
1856 @cindex memory usage
1857 @item --print-memory-usage
1858 Print used size, total size and used size of memory regions created with
1859 the @ref{MEMORY} command. This is useful on embedded targets to have a
1860 quick view of amount of free memory. The format of the output has one
1861 headline and one line per region. It is both human readable and easily
1862 parsable by tools. Here is an example of an output:
1865 Memory region Used Size Region Size %age Used
1866 ROM: 256 KB 1 MB 25.00%
1867 RAM: 32 B 2 GB 0.00%
1874 Print a summary of the command-line options on the standard output and exit.
1876 @kindex --target-help
1878 Print a summary of all target-specific options on the standard output and exit.
1880 @kindex -Map=@var{mapfile}
1881 @item -Map=@var{mapfile}
1882 Print a link map to the file @var{mapfile}. See the description of the
1883 @option{-M} option, above. If @var{mapfile} is just the character
1884 @code{-} then the map will be written to stdout.
1886 Specifying a directory as @var{mapfile} causes the linker map to be
1887 written as a file inside the directory. Normally name of the file
1888 inside the directory is computed as the basename of the @var{output}
1889 file with @code{.map} appended. If however the special character
1890 @code{%} is used then this will be replaced by the full path of the
1891 output file. Additionally if there are any characters after the
1892 @var{%} symbol then @code{.map} will no longer be appended.
1895 -o foo.exe -Map=bar [Creates ./bar]
1896 -o ../dir/foo.exe -Map=bar [Creates ./bar]
1897 -o foo.exe -Map=../dir [Creates ../dir/foo.exe.map]
1898 -o ../dir2/foo.exe -Map=../dir [Creates ../dir/foo.exe.map]
1899 -o foo.exe -Map=% [Creates ./foo.exe.map]
1900 -o ../dir/foo.exe -Map=% [Creates ../dir/foo.exe.map]
1901 -o foo.exe -Map=%.bar [Creates ./foo.exe.bar]
1902 -o ../dir/foo.exe -Map=%.bar [Creates ../dir/foo.exe.bar]
1903 -o ../dir2/foo.exe -Map=../dir/% [Creates ../dir/../dir2/foo.exe.map]
1904 -o ../dir2/foo.exe -Map=../dir/%.bar [Creates ../dir/../dir2/foo.exe.bar]
1907 It is an error to specify more than one @code{%} character.
1909 If the map file already exists then it will be overwritten by this
1912 @cindex memory usage
1913 @kindex --no-keep-memory
1914 @item --no-keep-memory
1915 @command{ld} normally optimizes for speed over memory usage by caching the
1916 symbol tables of input files in memory. This option tells @command{ld} to
1917 instead optimize for memory usage, by rereading the symbol tables as
1918 necessary. This may be required if @command{ld} runs out of memory space
1919 while linking a large executable.
1921 @kindex --no-undefined
1924 @item --no-undefined
1926 Report unresolved symbol references from regular object files. This
1927 is done even if the linker is creating a non-symbolic shared library.
1928 The switch @option{--[no-]allow-shlib-undefined} controls the
1929 behaviour for reporting unresolved references found in shared
1930 libraries being linked in.
1932 The effects of this option can be reverted by using @code{-z undefs}.
1934 @kindex --allow-multiple-definition
1936 @item --allow-multiple-definition
1938 Normally when a symbol is defined multiple times, the linker will
1939 report a fatal error. These options allow multiple definitions and the
1940 first definition will be used.
1942 @kindex --allow-shlib-undefined
1943 @kindex --no-allow-shlib-undefined
1944 @item --allow-shlib-undefined
1945 @itemx --no-allow-shlib-undefined
1946 Allows or disallows undefined symbols in shared libraries.
1947 This switch is similar to @option{--no-undefined} except that it
1948 determines the behaviour when the undefined symbols are in a
1949 shared library rather than a regular object file. It does not affect
1950 how undefined symbols in regular object files are handled.
1952 The default behaviour is to report errors for any undefined symbols
1953 referenced in shared libraries if the linker is being used to create
1954 an executable, but to allow them if the linker is being used to create
1957 The reasons for allowing undefined symbol references in shared
1958 libraries specified at link time are that:
1962 A shared library specified at link time may not be the same as the one
1963 that is available at load time, so the symbol might actually be
1964 resolvable at load time.
1966 There are some operating systems, eg BeOS and HPPA, where undefined
1967 symbols in shared libraries are normal.
1969 The BeOS kernel for example patches shared libraries at load time to
1970 select whichever function is most appropriate for the current
1971 architecture. This is used, for example, to dynamically select an
1972 appropriate memset function.
1975 @kindex --error-handling-script=@var{scriptname}
1976 @item --error-handling-script=@var{scriptname}
1977 If this option is provided then the linker will invoke
1978 @var{scriptname} whenever an error is encountered. Currently however
1979 only two kinds of error are supported: missing symbols and missing
1980 libraries. Two arguments will be passed to script: the keyword
1981 ``undefined-symbol'' or `missing-lib'' and the @var{name} of the
1982 undefined symbol or missing library. The intention is that the script
1983 will provide suggestions to the user as to where the symbol or library
1984 might be found. After the script has finished then the normal linker
1985 error message will be displayed.
1987 The availability of this option is controlled by a configure time
1988 switch, so it may not be present in specific implementations.
1990 @kindex --no-undefined-version
1991 @item --no-undefined-version
1992 Normally when a symbol has an undefined version, the linker will ignore
1993 it. This option disallows symbols with undefined version and a fatal error
1994 will be issued instead.
1996 @kindex --default-symver
1997 @item --default-symver
1998 Create and use a default symbol version (the soname) for unversioned
2001 @kindex --default-imported-symver
2002 @item --default-imported-symver
2003 Create and use a default symbol version (the soname) for unversioned
2006 @kindex --no-warn-mismatch
2007 @item --no-warn-mismatch
2008 Normally @command{ld} will give an error if you try to link together input
2009 files that are mismatched for some reason, perhaps because they have
2010 been compiled for different processors or for different endiannesses.
2011 This option tells @command{ld} that it should silently permit such possible
2012 errors. This option should only be used with care, in cases when you
2013 have taken some special action that ensures that the linker errors are
2016 @kindex --no-warn-search-mismatch
2017 @item --no-warn-search-mismatch
2018 Normally @command{ld} will give a warning if it finds an incompatible
2019 library during a library search. This option silences the warning.
2021 @kindex --no-whole-archive
2022 @item --no-whole-archive
2023 Turn off the effect of the @option{--whole-archive} option for subsequent
2026 @cindex output file after errors
2027 @kindex --noinhibit-exec
2028 @item --noinhibit-exec
2029 Retain the executable output file whenever it is still usable.
2030 Normally, the linker will not produce an output file if it encounters
2031 errors during the link process; it exits without writing an output file
2032 when it issues any error whatsoever.
2036 Only search library directories explicitly specified on the
2037 command line. Library directories specified in linker scripts
2038 (including linker scripts specified on the command line) are ignored.
2040 @ifclear SingleFormat
2041 @kindex --oformat=@var{output-format}
2042 @item --oformat=@var{output-format}
2043 @command{ld} may be configured to support more than one kind of object
2044 file. If your @command{ld} is configured this way, you can use the
2045 @samp{--oformat} option to specify the binary format for the output
2046 object file. Even when @command{ld} is configured to support alternative
2047 object formats, you don't usually need to specify this, as @command{ld}
2048 should be configured to produce as a default output format the most
2049 usual format on each machine. @var{output-format} is a text string, the
2050 name of a particular format supported by the BFD libraries. (You can
2051 list the available binary formats with @samp{objdump -i}.) The script
2052 command @code{OUTPUT_FORMAT} can also specify the output format, but
2053 this option overrides it. @xref{BFD}.
2056 @kindex --out-implib
2057 @item --out-implib @var{file}
2058 Create an import library in @var{file} corresponding to the executable
2059 the linker is generating (eg. a DLL or ELF program). This import
2060 library (which should be called @code{*.dll.a} or @code{*.a} for DLLs)
2061 may be used to link clients against the generated executable; this
2062 behaviour makes it possible to skip a separate import library creation
2063 step (eg. @code{dlltool} for DLLs). This option is only available for
2064 the i386 PE and ELF targetted ports of the linker.
2067 @kindex --pic-executable
2069 @itemx --pic-executable
2070 @cindex position independent executables
2071 Create a position independent executable. This is currently only supported on
2072 ELF platforms. Position independent executables are similar to shared
2073 libraries in that they are relocated by the dynamic linker to the virtual
2074 address the OS chooses for them (which can vary between invocations). Like
2075 normal dynamically linked executables they can be executed and symbols
2076 defined in the executable cannot be overridden by shared libraries.
2080 This option is ignored for Linux compatibility.
2084 This option is ignored for SVR4 compatibility.
2087 @cindex synthesizing linker
2088 @cindex relaxing addressing modes
2092 An option with machine dependent effects.
2094 This option is only supported on a few targets.
2097 @xref{H8/300,,@command{ld} and the H8/300}.
2100 @xref{Xtensa,, @command{ld} and Xtensa Processors}.
2103 @xref{M68HC11/68HC12,,@command{ld} and the 68HC11 and 68HC12}.
2106 @xref{Nios II,,@command{ld} and the Altera Nios II}.
2109 @xref{PowerPC ELF32,,@command{ld} and PowerPC 32-bit ELF Support}.
2112 On some platforms the @option{--relax} option performs target specific,
2113 global optimizations that become possible when the linker resolves
2114 addressing in the program, such as relaxing address modes,
2115 synthesizing new instructions, selecting shorter version of current
2116 instructions, and combining constant values.
2118 On some platforms these link time global optimizations may make symbolic
2119 debugging of the resulting executable impossible.
2121 This is known to be the case for the Matsushita MN10200 and MN10300
2122 family of processors.
2125 On platforms where the feature is supported, the option
2126 @option{--no-relax} will disable it.
2128 On platforms where the feature is not supported, both @option{--relax}
2129 and @option{--no-relax} are accepted, but ignored.
2131 @cindex retaining specified symbols
2132 @cindex stripping all but some symbols
2133 @cindex symbols, retaining selectively
2134 @kindex --retain-symbols-file=@var{filename}
2135 @item --retain-symbols-file=@var{filename}
2136 Retain @emph{only} the symbols listed in the file @var{filename},
2137 discarding all others. @var{filename} is simply a flat file, with one
2138 symbol name per line. This option is especially useful in environments
2142 where a large global symbol table is accumulated gradually, to conserve
2145 @samp{--retain-symbols-file} does @emph{not} discard undefined symbols,
2146 or symbols needed for relocations.
2148 You may only specify @samp{--retain-symbols-file} once in the command
2149 line. It overrides @samp{-s} and @samp{-S}.
2152 @item -rpath=@var{dir}
2153 @cindex runtime library search path
2154 @kindex -rpath=@var{dir}
2155 Add a directory to the runtime library search path. This is used when
2156 linking an ELF executable with shared objects. All @option{-rpath}
2157 arguments are concatenated and passed to the runtime linker, which uses
2158 them to locate shared objects at runtime.
2160 The @option{-rpath} option is also used when locating shared objects which
2161 are needed by shared objects explicitly included in the link; see the
2162 description of the @option{-rpath-link} option. Searching @option{-rpath}
2163 in this way is only supported by native linkers and cross linkers which
2164 have been configured with the @option{--with-sysroot} option.
2166 If @option{-rpath} is not used when linking an ELF executable, the
2167 contents of the environment variable @code{LD_RUN_PATH} will be used if it
2170 The @option{-rpath} option may also be used on SunOS. By default, on
2171 SunOS, the linker will form a runtime search path out of all the
2172 @option{-L} options it is given. If a @option{-rpath} option is used, the
2173 runtime search path will be formed exclusively using the @option{-rpath}
2174 options, ignoring the @option{-L} options. This can be useful when using
2175 gcc, which adds many @option{-L} options which may be on NFS mounted
2178 For compatibility with other ELF linkers, if the @option{-R} option is
2179 followed by a directory name, rather than a file name, it is treated as
2180 the @option{-rpath} option.
2184 @cindex link-time runtime library search path
2185 @kindex -rpath-link=@var{dir}
2186 @item -rpath-link=@var{dir}
2187 When using ELF or SunOS, one shared library may require another. This
2188 happens when an @code{ld -shared} link includes a shared library as one
2191 When the linker encounters such a dependency when doing a non-shared,
2192 non-relocatable link, it will automatically try to locate the required
2193 shared library and include it in the link, if it is not included
2194 explicitly. In such a case, the @option{-rpath-link} option
2195 specifies the first set of directories to search. The
2196 @option{-rpath-link} option may specify a sequence of directory names
2197 either by specifying a list of names separated by colons, or by
2198 appearing multiple times.
2200 The tokens @var{$ORIGIN} and @var{$LIB} can appear in these search
2201 directories. They will be replaced by the full path to the directory
2202 containing the program or shared object in the case of @var{$ORIGIN}
2203 and either @samp{lib} - for 32-bit binaries - or @samp{lib64} - for
2204 64-bit binaries - in the case of @var{$LIB}.
2206 The alternative form of these tokens - @var{$@{ORIGIN@}} and
2207 @var{$@{LIB@}} can also be used. The token @var{$PLATFORM} is not
2210 This option should be used with caution as it overrides the search path
2211 that may have been hard compiled into a shared library. In such a case it
2212 is possible to use unintentionally a different search path than the
2213 runtime linker would do.
2215 The linker uses the following search paths to locate required shared
2220 Any directories specified by @option{-rpath-link} options.
2222 Any directories specified by @option{-rpath} options. The difference
2223 between @option{-rpath} and @option{-rpath-link} is that directories
2224 specified by @option{-rpath} options are included in the executable and
2225 used at runtime, whereas the @option{-rpath-link} option is only effective
2226 at link time. Searching @option{-rpath} in this way is only supported
2227 by native linkers and cross linkers which have been configured with
2228 the @option{--with-sysroot} option.
2230 On an ELF system, for native linkers, if the @option{-rpath} and
2231 @option{-rpath-link} options were not used, search the contents of the
2232 environment variable @code{LD_RUN_PATH}.
2234 On SunOS, if the @option{-rpath} option was not used, search any
2235 directories specified using @option{-L} options.
2237 For a native linker, search the contents of the environment
2238 variable @code{LD_LIBRARY_PATH}.
2240 For a native ELF linker, the directories in @code{DT_RUNPATH} or
2241 @code{DT_RPATH} of a shared library are searched for shared
2242 libraries needed by it. The @code{DT_RPATH} entries are ignored if
2243 @code{DT_RUNPATH} entries exist.
2245 The default directories, normally @file{/lib} and @file{/usr/lib}.
2247 For a linker for a Linux system, if the file @file{/etc/ld.so.conf}
2248 exists, the list of directories found in that file. Note: the path
2249 to this file is prefixed with the @code{sysroot} value, if that is
2250 defined, and then any @code{prefix} string if the linker was
2251 configured with the @command{--prefix=<path>} option.
2253 For a native linker on a FreeBSD system, any directories specified by
2254 the @code{_PATH_ELF_HINTS} macro defined in the @file{elf-hints.h}
2257 Any directories specifed by a @code{SEARCH_DIR} command in the
2258 linker script being used.
2261 If the required shared library is not found, the linker will issue a
2262 warning and continue with the link.
2269 @cindex shared libraries
2270 Create a shared library. This is currently only supported on ELF, XCOFF
2271 and SunOS platforms. On SunOS, the linker will automatically create a
2272 shared library if the @option{-e} option is not used and there are
2273 undefined symbols in the link.
2275 @kindex --sort-common
2277 @itemx --sort-common=ascending
2278 @itemx --sort-common=descending
2279 This option tells @command{ld} to sort the common symbols by alignment in
2280 ascending or descending order when it places them in the appropriate output
2281 sections. The symbol alignments considered are sixteen-byte or larger,
2282 eight-byte, four-byte, two-byte, and one-byte. This is to prevent gaps
2283 between symbols due to alignment constraints. If no sorting order is
2284 specified, then descending order is assumed.
2286 @kindex --sort-section=name
2287 @item --sort-section=name
2288 This option will apply @code{SORT_BY_NAME} to all wildcard section
2289 patterns in the linker script.
2291 @kindex --sort-section=alignment
2292 @item --sort-section=alignment
2293 This option will apply @code{SORT_BY_ALIGNMENT} to all wildcard section
2294 patterns in the linker script.
2296 @kindex --spare-dynamic-tags
2297 @item --spare-dynamic-tags=@var{count}
2298 This option specifies the number of empty slots to leave in the
2299 .dynamic section of ELF shared objects. Empty slots may be needed by
2300 post processing tools, such as the prelinker. The default is 5.
2302 @kindex --split-by-file
2303 @item --split-by-file[=@var{size}]
2304 Similar to @option{--split-by-reloc} but creates a new output section for
2305 each input file when @var{size} is reached. @var{size} defaults to a
2306 size of 1 if not given.
2308 @kindex --split-by-reloc
2309 @item --split-by-reloc[=@var{count}]
2310 Tries to creates extra sections in the output file so that no single
2311 output section in the file contains more than @var{count} relocations.
2312 This is useful when generating huge relocatable files for downloading into
2313 certain real time kernels with the COFF object file format; since COFF
2314 cannot represent more than 65535 relocations in a single section. Note
2315 that this will fail to work with object file formats which do not
2316 support arbitrary sections. The linker will not split up individual
2317 input sections for redistribution, so if a single input section contains
2318 more than @var{count} relocations one output section will contain that
2319 many relocations. @var{count} defaults to a value of 32768.
2323 Compute and display statistics about the operation of the linker, such
2324 as execution time and memory usage.
2326 @kindex --sysroot=@var{directory}
2327 @item --sysroot=@var{directory}
2328 Use @var{directory} as the location of the sysroot, overriding the
2329 configure-time default. This option is only supported by linkers
2330 that were configured using @option{--with-sysroot}.
2334 This is used by COFF/PE based targets to create a task-linked object
2335 file where all of the global symbols have been converted to statics.
2337 @kindex --traditional-format
2338 @cindex traditional format
2339 @item --traditional-format
2340 For some targets, the output of @command{ld} is different in some ways from
2341 the output of some existing linker. This switch requests @command{ld} to
2342 use the traditional format instead.
2345 For example, on SunOS, @command{ld} combines duplicate entries in the
2346 symbol string table. This can reduce the size of an output file with
2347 full debugging information by over 30 percent. Unfortunately, the SunOS
2348 @code{dbx} program can not read the resulting program (@code{gdb} has no
2349 trouble). The @samp{--traditional-format} switch tells @command{ld} to not
2350 combine duplicate entries.
2352 @kindex --section-start=@var{sectionname}=@var{org}
2353 @item --section-start=@var{sectionname}=@var{org}
2354 Locate a section in the output file at the absolute
2355 address given by @var{org}. You may use this option as many
2356 times as necessary to locate multiple sections in the command
2358 @var{org} must be a single hexadecimal integer;
2359 for compatibility with other linkers, you may omit the leading
2360 @samp{0x} usually associated with hexadecimal values. @emph{Note:} there
2361 should be no white space between @var{sectionname}, the equals
2362 sign (``@key{=}''), and @var{org}.
2364 @kindex -Tbss=@var{org}
2365 @kindex -Tdata=@var{org}
2366 @kindex -Ttext=@var{org}
2367 @cindex segment origins, cmd line
2368 @item -Tbss=@var{org}
2369 @itemx -Tdata=@var{org}
2370 @itemx -Ttext=@var{org}
2371 Same as @option{--section-start}, with @code{.bss}, @code{.data} or
2372 @code{.text} as the @var{sectionname}.
2374 @kindex -Ttext-segment=@var{org}
2375 @item -Ttext-segment=@var{org}
2376 @cindex text segment origin, cmd line
2377 When creating an ELF executable, it will set the address of the first
2378 byte of the text segment.
2380 @kindex -Trodata-segment=@var{org}
2381 @item -Trodata-segment=@var{org}
2382 @cindex rodata segment origin, cmd line
2383 When creating an ELF executable or shared object for a target where
2384 the read-only data is in its own segment separate from the executable
2385 text, it will set the address of the first byte of the read-only data segment.
2387 @kindex -Tldata-segment=@var{org}
2388 @item -Tldata-segment=@var{org}
2389 @cindex ldata segment origin, cmd line
2390 When creating an ELF executable or shared object for x86-64 medium memory
2391 model, it will set the address of the first byte of the ldata segment.
2393 @kindex --unresolved-symbols
2394 @item --unresolved-symbols=@var{method}
2395 Determine how to handle unresolved symbols. There are four possible
2396 values for @samp{method}:
2400 Do not report any unresolved symbols.
2403 Report all unresolved symbols. This is the default.
2405 @item ignore-in-object-files
2406 Report unresolved symbols that are contained in shared libraries, but
2407 ignore them if they come from regular object files.
2409 @item ignore-in-shared-libs
2410 Report unresolved symbols that come from regular object files, but
2411 ignore them if they come from shared libraries. This can be useful
2412 when creating a dynamic binary and it is known that all the shared
2413 libraries that it should be referencing are included on the linker's
2417 The behaviour for shared libraries on their own can also be controlled
2418 by the @option{--[no-]allow-shlib-undefined} option.
2420 Normally the linker will generate an error message for each reported
2421 unresolved symbol but the option @option{--warn-unresolved-symbols}
2422 can change this to a warning.
2424 @kindex --verbose[=@var{NUMBER}]
2425 @cindex verbose[=@var{NUMBER}]
2427 @itemx --verbose[=@var{NUMBER}]
2428 Display the version number for @command{ld} and list the linker emulations
2429 supported. Display which input files can and cannot be opened. Display
2430 the linker script being used by the linker. If the optional @var{NUMBER}
2431 argument > 1, plugin symbol status will also be displayed.
2433 @kindex --version-script=@var{version-scriptfile}
2434 @cindex version script, symbol versions
2435 @item --version-script=@var{version-scriptfile}
2436 Specify the name of a version script to the linker. This is typically
2437 used when creating shared libraries to specify additional information
2438 about the version hierarchy for the library being created. This option
2439 is only fully supported on ELF platforms which support shared libraries;
2440 see @ref{VERSION}. It is partially supported on PE platforms, which can
2441 use version scripts to filter symbol visibility in auto-export mode: any
2442 symbols marked @samp{local} in the version script will not be exported.
2445 @kindex --warn-common
2446 @cindex warnings, on combining symbols
2447 @cindex combining symbols, warnings on
2449 Warn when a common symbol is combined with another common symbol or with
2450 a symbol definition. Unix linkers allow this somewhat sloppy practice,
2451 but linkers on some other operating systems do not. This option allows
2452 you to find potential problems from combining global symbols.
2453 Unfortunately, some C libraries use this practice, so you may get some
2454 warnings about symbols in the libraries as well as in your programs.
2456 There are three kinds of global symbols, illustrated here by C examples:
2460 A definition, which goes in the initialized data section of the output
2464 An undefined reference, which does not allocate space.
2465 There must be either a definition or a common symbol for the
2469 A common symbol. If there are only (one or more) common symbols for a
2470 variable, it goes in the uninitialized data area of the output file.
2471 The linker merges multiple common symbols for the same variable into a
2472 single symbol. If they are of different sizes, it picks the largest
2473 size. The linker turns a common symbol into a declaration, if there is
2474 a definition of the same variable.
2477 The @samp{--warn-common} option can produce five kinds of warnings.
2478 Each warning consists of a pair of lines: the first describes the symbol
2479 just encountered, and the second describes the previous symbol
2480 encountered with the same name. One or both of the two symbols will be
2485 Turning a common symbol into a reference, because there is already a
2486 definition for the symbol.
2488 @var{file}(@var{section}): warning: common of `@var{symbol}'
2489 overridden by definition
2490 @var{file}(@var{section}): warning: defined here
2494 Turning a common symbol into a reference, because a later definition for
2495 the symbol is encountered. This is the same as the previous case,
2496 except that the symbols are encountered in a different order.
2498 @var{file}(@var{section}): warning: definition of `@var{symbol}'
2500 @var{file}(@var{section}): warning: common is here
2504 Merging a common symbol with a previous same-sized common symbol.
2506 @var{file}(@var{section}): warning: multiple common
2508 @var{file}(@var{section}): warning: previous common is here
2512 Merging a common symbol with a previous larger common symbol.
2514 @var{file}(@var{section}): warning: common of `@var{symbol}'
2515 overridden by larger common
2516 @var{file}(@var{section}): warning: larger common is here
2520 Merging a common symbol with a previous smaller common symbol. This is
2521 the same as the previous case, except that the symbols are
2522 encountered in a different order.
2524 @var{file}(@var{section}): warning: common of `@var{symbol}'
2525 overriding smaller common
2526 @var{file}(@var{section}): warning: smaller common is here
2530 @kindex --warn-constructors
2531 @item --warn-constructors
2532 Warn if any global constructors are used. This is only useful for a few
2533 object file formats. For formats like COFF or ELF, the linker can not
2534 detect the use of global constructors.
2536 @kindex --warn-multiple-gp
2537 @item --warn-multiple-gp
2538 Warn if multiple global pointer values are required in the output file.
2539 This is only meaningful for certain processors, such as the Alpha.
2540 Specifically, some processors put large-valued constants in a special
2541 section. A special register (the global pointer) points into the middle
2542 of this section, so that constants can be loaded efficiently via a
2543 base-register relative addressing mode. Since the offset in
2544 base-register relative mode is fixed and relatively small (e.g., 16
2545 bits), this limits the maximum size of the constant pool. Thus, in
2546 large programs, it is often necessary to use multiple global pointer
2547 values in order to be able to address all possible constants. This
2548 option causes a warning to be issued whenever this case occurs.
2551 @cindex warnings, on undefined symbols
2552 @cindex undefined symbols, warnings on
2554 Only warn once for each undefined symbol, rather than once per module
2557 @kindex --warn-section-align
2558 @cindex warnings, on section alignment
2559 @cindex section alignment, warnings on
2560 @item --warn-section-align
2561 Warn if the address of an output section is changed because of
2562 alignment. Typically, the alignment will be set by an input section.
2563 The address will only be changed if it not explicitly specified; that
2564 is, if the @code{SECTIONS} command does not specify a start address for
2565 the section (@pxref{SECTIONS}).
2567 @kindex --warn-textrel
2568 @item --warn-textrel
2569 Warn if the linker adds DT_TEXTREL to a position-independent executable
2572 @kindex --warn-alternate-em
2573 @item --warn-alternate-em
2574 Warn if an object has alternate ELF machine code.
2576 @kindex --warn-unresolved-symbols
2577 @item --warn-unresolved-symbols
2578 If the linker is going to report an unresolved symbol (see the option
2579 @option{--unresolved-symbols}) it will normally generate an error.
2580 This option makes it generate a warning instead.
2582 @kindex --error-unresolved-symbols
2583 @item --error-unresolved-symbols
2584 This restores the linker's default behaviour of generating errors when
2585 it is reporting unresolved symbols.
2587 @kindex --whole-archive
2588 @cindex including an entire archive
2589 @item --whole-archive
2590 For each archive mentioned on the command line after the
2591 @option{--whole-archive} option, include every object file in the archive
2592 in the link, rather than searching the archive for the required object
2593 files. This is normally used to turn an archive file into a shared
2594 library, forcing every object to be included in the resulting shared
2595 library. This option may be used more than once.
2597 Two notes when using this option from gcc: First, gcc doesn't know
2598 about this option, so you have to use @option{-Wl,-whole-archive}.
2599 Second, don't forget to use @option{-Wl,-no-whole-archive} after your
2600 list of archives, because gcc will add its own list of archives to
2601 your link and you may not want this flag to affect those as well.
2603 @kindex --wrap=@var{symbol}
2604 @item --wrap=@var{symbol}
2605 Use a wrapper function for @var{symbol}. Any undefined reference to
2606 @var{symbol} will be resolved to @code{__wrap_@var{symbol}}. Any
2607 undefined reference to @code{__real_@var{symbol}} will be resolved to
2610 This can be used to provide a wrapper for a system function. The
2611 wrapper function should be called @code{__wrap_@var{symbol}}. If it
2612 wishes to call the system function, it should call
2613 @code{__real_@var{symbol}}.
2615 Here is a trivial example:
2619 __wrap_malloc (size_t c)
2621 printf ("malloc called with %zu\n", c);
2622 return __real_malloc (c);
2626 If you link other code with this file using @option{--wrap malloc}, then
2627 all calls to @code{malloc} will call the function @code{__wrap_malloc}
2628 instead. The call to @code{__real_malloc} in @code{__wrap_malloc} will
2629 call the real @code{malloc} function.
2631 You may wish to provide a @code{__real_malloc} function as well, so that
2632 links without the @option{--wrap} option will succeed. If you do this,
2633 you should not put the definition of @code{__real_malloc} in the same
2634 file as @code{__wrap_malloc}; if you do, the assembler may resolve the
2635 call before the linker has a chance to wrap it to @code{malloc}.
2637 Only undefined references are replaced by the linker. So, translation unit
2638 internal references to @var{symbol} are not resolved to
2639 @code{__wrap_@var{symbol}}. In the next example, the call to @code{f} in
2640 @code{g} is not resolved to @code{__wrap_f}.
2656 @kindex --eh-frame-hdr
2657 @kindex --no-eh-frame-hdr
2658 @item --eh-frame-hdr
2659 @itemx --no-eh-frame-hdr
2660 Request (@option{--eh-frame-hdr}) or suppress
2661 (@option{--no-eh-frame-hdr}) the creation of @code{.eh_frame_hdr}
2662 section and ELF @code{PT_GNU_EH_FRAME} segment header.
2664 @kindex --ld-generated-unwind-info
2665 @item --no-ld-generated-unwind-info
2666 Request creation of @code{.eh_frame} unwind info for linker
2667 generated code sections like PLT. This option is on by default
2668 if linker generated unwind info is supported.
2670 @kindex --enable-new-dtags
2671 @kindex --disable-new-dtags
2672 @item --enable-new-dtags
2673 @itemx --disable-new-dtags
2674 This linker can create the new dynamic tags in ELF. But the older ELF
2675 systems may not understand them. If you specify
2676 @option{--enable-new-dtags}, the new dynamic tags will be created as needed
2677 and older dynamic tags will be omitted.
2678 If you specify @option{--disable-new-dtags}, no new dynamic tags will be
2679 created. By default, the new dynamic tags are not created. Note that
2680 those options are only available for ELF systems.
2682 @kindex --hash-size=@var{number}
2683 @item --hash-size=@var{number}
2684 Set the default size of the linker's hash tables to a prime number
2685 close to @var{number}. Increasing this value can reduce the length of
2686 time it takes the linker to perform its tasks, at the expense of
2687 increasing the linker's memory requirements. Similarly reducing this
2688 value can reduce the memory requirements at the expense of speed.
2690 @kindex --hash-style=@var{style}
2691 @item --hash-style=@var{style}
2692 Set the type of linker's hash table(s). @var{style} can be either
2693 @code{sysv} for classic ELF @code{.hash} section, @code{gnu} for
2694 new style GNU @code{.gnu.hash} section or @code{both} for both
2695 the classic ELF @code{.hash} and new style GNU @code{.gnu.hash}
2696 hash tables. The default depends upon how the linker was configured,
2697 but for most Linux based systems it will be @code{both}.
2699 @kindex --compress-debug-sections=none
2700 @kindex --compress-debug-sections=zlib
2701 @kindex --compress-debug-sections=zlib-gnu
2702 @kindex --compress-debug-sections=zlib-gabi
2703 @item --compress-debug-sections=none
2704 @itemx --compress-debug-sections=zlib
2705 @itemx --compress-debug-sections=zlib-gnu
2706 @itemx --compress-debug-sections=zlib-gabi
2707 On ELF platforms, these options control how DWARF debug sections are
2708 compressed using zlib.
2710 @option{--compress-debug-sections=none} doesn't compress DWARF debug
2711 sections. @option{--compress-debug-sections=zlib-gnu} compresses
2712 DWARF debug sections and renames them to begin with @samp{.zdebug}
2713 instead of @samp{.debug}. @option{--compress-debug-sections=zlib-gabi}
2714 also compresses DWARF debug sections, but rather than renaming them it
2715 sets the SHF_COMPRESSED flag in the sections' headers.
2717 The @option{--compress-debug-sections=zlib} option is an alias for
2718 @option{--compress-debug-sections=zlib-gabi}.
2720 Note that this option overrides any compression in input debug
2721 sections, so if a binary is linked with @option{--compress-debug-sections=none}
2722 for example, then any compressed debug sections in input files will be
2723 uncompressed before they are copied into the output binary.
2725 The default compression behaviour varies depending upon the target
2726 involved and the configure options used to build the toolchain. The
2727 default can be determined by examining the output from the linker's
2728 @option{--help} option.
2730 @kindex --reduce-memory-overheads
2731 @item --reduce-memory-overheads
2732 This option reduces memory requirements at ld runtime, at the expense of
2733 linking speed. This was introduced to select the old O(n^2) algorithm
2734 for link map file generation, rather than the new O(n) algorithm which uses
2735 about 40% more memory for symbol storage.
2737 Another effect of the switch is to set the default hash table size to
2738 1021, which again saves memory at the cost of lengthening the linker's
2739 run time. This is not done however if the @option{--hash-size} switch
2742 The @option{--reduce-memory-overheads} switch may be also be used to
2743 enable other tradeoffs in future versions of the linker.
2746 @kindex --build-id=@var{style}
2748 @itemx --build-id=@var{style}
2749 Request the creation of a @code{.note.gnu.build-id} ELF note section
2750 or a @code{.buildid} COFF section. The contents of the note are
2751 unique bits identifying this linked file. @var{style} can be
2752 @code{uuid} to use 128 random bits, @code{sha1} to use a 160-bit
2753 @sc{SHA1} hash on the normative parts of the output contents,
2754 @code{md5} to use a 128-bit @sc{MD5} hash on the normative parts of
2755 the output contents, or @code{0x@var{hexstring}} to use a chosen bit
2756 string specified as an even number of hexadecimal digits (@code{-} and
2757 @code{:} characters between digit pairs are ignored). If @var{style}
2758 is omitted, @code{sha1} is used.
2760 The @code{md5} and @code{sha1} styles produces an identifier
2761 that is always the same in an identical output file, but will be
2762 unique among all nonidentical output files. It is not intended
2763 to be compared as a checksum for the file's contents. A linked
2764 file may be changed later by other tools, but the build ID bit
2765 string identifying the original linked file does not change.
2767 Passing @code{none} for @var{style} disables the setting from any
2768 @code{--build-id} options earlier on the command line.
2773 @subsection Options Specific to i386 PE Targets
2775 @c man begin OPTIONS
2777 The i386 PE linker supports the @option{-shared} option, which causes
2778 the output to be a dynamically linked library (DLL) instead of a
2779 normal executable. You should name the output @code{*.dll} when you
2780 use this option. In addition, the linker fully supports the standard
2781 @code{*.def} files, which may be specified on the linker command line
2782 like an object file (in fact, it should precede archives it exports
2783 symbols from, to ensure that they get linked in, just like a normal
2786 In addition to the options common to all targets, the i386 PE linker
2787 support additional command-line options that are specific to the i386
2788 PE target. Options that take values may be separated from their
2789 values by either a space or an equals sign.
2793 @kindex --add-stdcall-alias
2794 @item --add-stdcall-alias
2795 If given, symbols with a stdcall suffix (@@@var{nn}) will be exported
2796 as-is and also with the suffix stripped.
2797 [This option is specific to the i386 PE targeted port of the linker]
2800 @item --base-file @var{file}
2801 Use @var{file} as the name of a file in which to save the base
2802 addresses of all the relocations needed for generating DLLs with
2804 [This is an i386 PE specific option]
2808 Create a DLL instead of a regular executable. You may also use
2809 @option{-shared} or specify a @code{LIBRARY} in a given @code{.def}
2811 [This option is specific to the i386 PE targeted port of the linker]
2813 @kindex --enable-long-section-names
2814 @kindex --disable-long-section-names
2815 @item --enable-long-section-names
2816 @itemx --disable-long-section-names
2817 The PE variants of the COFF object format add an extension that permits
2818 the use of section names longer than eight characters, the normal limit
2819 for COFF. By default, these names are only allowed in object files, as
2820 fully-linked executable images do not carry the COFF string table required
2821 to support the longer names. As a GNU extension, it is possible to
2822 allow their use in executable images as well, or to (probably pointlessly!)
2823 disallow it in object files, by using these two options. Executable images
2824 generated with these long section names are slightly non-standard, carrying
2825 as they do a string table, and may generate confusing output when examined
2826 with non-GNU PE-aware tools, such as file viewers and dumpers. However,
2827 GDB relies on the use of PE long section names to find Dwarf-2 debug
2828 information sections in an executable image at runtime, and so if neither
2829 option is specified on the command-line, @command{ld} will enable long
2830 section names, overriding the default and technically correct behaviour,
2831 when it finds the presence of debug information while linking an executable
2832 image and not stripping symbols.
2833 [This option is valid for all PE targeted ports of the linker]
2835 @kindex --enable-stdcall-fixup
2836 @kindex --disable-stdcall-fixup
2837 @item --enable-stdcall-fixup
2838 @itemx --disable-stdcall-fixup
2839 If the link finds a symbol that it cannot resolve, it will attempt to
2840 do ``fuzzy linking'' by looking for another defined symbol that differs
2841 only in the format of the symbol name (cdecl vs stdcall) and will
2842 resolve that symbol by linking to the match. For example, the
2843 undefined symbol @code{_foo} might be linked to the function
2844 @code{_foo@@12}, or the undefined symbol @code{_bar@@16} might be linked
2845 to the function @code{_bar}. When the linker does this, it prints a
2846 warning, since it normally should have failed to link, but sometimes
2847 import libraries generated from third-party dlls may need this feature
2848 to be usable. If you specify @option{--enable-stdcall-fixup}, this
2849 feature is fully enabled and warnings are not printed. If you specify
2850 @option{--disable-stdcall-fixup}, this feature is disabled and such
2851 mismatches are considered to be errors.
2852 [This option is specific to the i386 PE targeted port of the linker]
2854 @kindex --leading-underscore
2855 @kindex --no-leading-underscore
2856 @item --leading-underscore
2857 @itemx --no-leading-underscore
2858 For most targets default symbol-prefix is an underscore and is defined
2859 in target's description. By this option it is possible to
2860 disable/enable the default underscore symbol-prefix.
2862 @cindex DLLs, creating
2863 @kindex --export-all-symbols
2864 @item --export-all-symbols
2865 If given, all global symbols in the objects used to build a DLL will
2866 be exported by the DLL. Note that this is the default if there
2867 otherwise wouldn't be any exported symbols. When symbols are
2868 explicitly exported via DEF files or implicitly exported via function
2869 attributes, the default is to not export anything else unless this
2870 option is given. Note that the symbols @code{DllMain@@12},
2871 @code{DllEntryPoint@@0}, @code{DllMainCRTStartup@@12}, and
2872 @code{impure_ptr} will not be automatically
2873 exported. Also, symbols imported from other DLLs will not be
2874 re-exported, nor will symbols specifying the DLL's internal layout
2875 such as those beginning with @code{_head_} or ending with
2876 @code{_iname}. In addition, no symbols from @code{libgcc},
2877 @code{libstd++}, @code{libmingw32}, or @code{crtX.o} will be exported.
2878 Symbols whose names begin with @code{__rtti_} or @code{__builtin_} will
2879 not be exported, to help with C++ DLLs. Finally, there is an
2880 extensive list of cygwin-private symbols that are not exported
2881 (obviously, this applies on when building DLLs for cygwin targets).
2882 These cygwin-excludes are: @code{_cygwin_dll_entry@@12},
2883 @code{_cygwin_crt0_common@@8}, @code{_cygwin_noncygwin_dll_entry@@12},
2884 @code{_fmode}, @code{_impure_ptr}, @code{cygwin_attach_dll},
2885 @code{cygwin_premain0}, @code{cygwin_premain1}, @code{cygwin_premain2},
2886 @code{cygwin_premain3}, and @code{environ}.
2887 [This option is specific to the i386 PE targeted port of the linker]
2889 @kindex --exclude-symbols
2890 @item --exclude-symbols @var{symbol},@var{symbol},...
2891 Specifies a list of symbols which should not be automatically
2892 exported. The symbol names may be delimited by commas or colons.
2893 [This option is specific to the i386 PE targeted port of the linker]
2895 @kindex --exclude-all-symbols
2896 @item --exclude-all-symbols
2897 Specifies no symbols should be automatically exported.
2898 [This option is specific to the i386 PE targeted port of the linker]
2900 @kindex --file-alignment
2901 @item --file-alignment
2902 Specify the file alignment. Sections in the file will always begin at
2903 file offsets which are multiples of this number. This defaults to
2905 [This option is specific to the i386 PE targeted port of the linker]
2909 @item --heap @var{reserve}
2910 @itemx --heap @var{reserve},@var{commit}
2911 Specify the number of bytes of memory to reserve (and optionally commit)
2912 to be used as heap for this program. The default is 1MB reserved, 4K
2914 [This option is specific to the i386 PE targeted port of the linker]
2917 @kindex --image-base
2918 @item --image-base @var{value}
2919 Use @var{value} as the base address of your program or dll. This is
2920 the lowest memory location that will be used when your program or dll
2921 is loaded. To reduce the need to relocate and improve performance of
2922 your dlls, each should have a unique base address and not overlap any
2923 other dlls. The default is 0x400000 for executables, and 0x10000000
2925 [This option is specific to the i386 PE targeted port of the linker]
2929 If given, the stdcall suffixes (@@@var{nn}) will be stripped from
2930 symbols before they are exported.
2931 [This option is specific to the i386 PE targeted port of the linker]
2933 @kindex --large-address-aware
2934 @item --large-address-aware
2935 If given, the appropriate bit in the ``Characteristics'' field of the COFF
2936 header is set to indicate that this executable supports virtual addresses
2937 greater than 2 gigabytes. This should be used in conjunction with the /3GB
2938 or /USERVA=@var{value} megabytes switch in the ``[operating systems]''
2939 section of the BOOT.INI. Otherwise, this bit has no effect.
2940 [This option is specific to PE targeted ports of the linker]
2942 @kindex --disable-large-address-aware
2943 @item --disable-large-address-aware
2944 Reverts the effect of a previous @samp{--large-address-aware} option.
2945 This is useful if @samp{--large-address-aware} is always set by the compiler
2946 driver (e.g. Cygwin gcc) and the executable does not support virtual
2947 addresses greater than 2 gigabytes.
2948 [This option is specific to PE targeted ports of the linker]
2950 @kindex --major-image-version
2951 @item --major-image-version @var{value}
2952 Sets the major number of the ``image version''. Defaults to 1.
2953 [This option is specific to the i386 PE targeted port of the linker]
2955 @kindex --major-os-version
2956 @item --major-os-version @var{value}
2957 Sets the major number of the ``os version''. Defaults to 4.
2958 [This option is specific to the i386 PE targeted port of the linker]
2960 @kindex --major-subsystem-version
2961 @item --major-subsystem-version @var{value}
2962 Sets the major number of the ``subsystem version''. Defaults to 4.
2963 [This option is specific to the i386 PE targeted port of the linker]
2965 @kindex --minor-image-version
2966 @item --minor-image-version @var{value}
2967 Sets the minor number of the ``image version''. Defaults to 0.
2968 [This option is specific to the i386 PE targeted port of the linker]
2970 @kindex --minor-os-version
2971 @item --minor-os-version @var{value}
2972 Sets the minor number of the ``os version''. Defaults to 0.
2973 [This option is specific to the i386 PE targeted port of the linker]
2975 @kindex --minor-subsystem-version
2976 @item --minor-subsystem-version @var{value}
2977 Sets the minor number of the ``subsystem version''. Defaults to 0.
2978 [This option is specific to the i386 PE targeted port of the linker]
2980 @cindex DEF files, creating
2981 @cindex DLLs, creating
2982 @kindex --output-def
2983 @item --output-def @var{file}
2984 The linker will create the file @var{file} which will contain a DEF
2985 file corresponding to the DLL the linker is generating. This DEF file
2986 (which should be called @code{*.def}) may be used to create an import
2987 library with @code{dlltool} or may be used as a reference to
2988 automatically or implicitly exported symbols.
2989 [This option is specific to the i386 PE targeted port of the linker]
2991 @cindex DLLs, creating
2992 @kindex --enable-auto-image-base
2993 @item --enable-auto-image-base
2994 @itemx --enable-auto-image-base=@var{value}
2995 Automatically choose the image base for DLLs, optionally starting with base
2996 @var{value}, unless one is specified using the @code{--image-base} argument.
2997 By using a hash generated from the dllname to create unique image bases
2998 for each DLL, in-memory collisions and relocations which can delay program
2999 execution are avoided.
3000 [This option is specific to the i386 PE targeted port of the linker]
3002 @kindex --disable-auto-image-base
3003 @item --disable-auto-image-base
3004 Do not automatically generate a unique image base. If there is no
3005 user-specified image base (@code{--image-base}) then use the platform
3007 [This option is specific to the i386 PE targeted port of the linker]
3009 @cindex DLLs, linking to
3010 @kindex --dll-search-prefix
3011 @item --dll-search-prefix @var{string}
3012 When linking dynamically to a dll without an import library,
3013 search for @code{<string><basename>.dll} in preference to
3014 @code{lib<basename>.dll}. This behaviour allows easy distinction
3015 between DLLs built for the various "subplatforms": native, cygwin,
3016 uwin, pw, etc. For instance, cygwin DLLs typically use
3017 @code{--dll-search-prefix=cyg}.
3018 [This option is specific to the i386 PE targeted port of the linker]
3020 @kindex --enable-auto-import
3021 @item --enable-auto-import
3022 Do sophisticated linking of @code{_symbol} to @code{__imp__symbol} for
3023 DATA imports from DLLs, thus making it possible to bypass the dllimport
3024 mechanism on the user side and to reference unmangled symbol names.
3025 [This option is specific to the i386 PE targeted port of the linker]
3027 The following remarks pertain to the original implementation of the
3028 feature and are obsolete nowadays for Cygwin and MinGW targets.
3030 Note: Use of the 'auto-import' extension will cause the text section
3031 of the image file to be made writable. This does not conform to the
3032 PE-COFF format specification published by Microsoft.
3034 Note - use of the 'auto-import' extension will also cause read only
3035 data which would normally be placed into the .rdata section to be
3036 placed into the .data section instead. This is in order to work
3037 around a problem with consts that is described here:
3038 http://www.cygwin.com/ml/cygwin/2004-09/msg01101.html
3040 Using 'auto-import' generally will 'just work' -- but sometimes you may
3043 "variable '<var>' can't be auto-imported. Please read the
3044 documentation for ld's @code{--enable-auto-import} for details."
3046 This message occurs when some (sub)expression accesses an address
3047 ultimately given by the sum of two constants (Win32 import tables only
3048 allow one). Instances where this may occur include accesses to member
3049 fields of struct variables imported from a DLL, as well as using a
3050 constant index into an array variable imported from a DLL. Any
3051 multiword variable (arrays, structs, long long, etc) may trigger
3052 this error condition. However, regardless of the exact data type
3053 of the offending exported variable, ld will always detect it, issue
3054 the warning, and exit.
3056 There are several ways to address this difficulty, regardless of the
3057 data type of the exported variable:
3059 One way is to use --enable-runtime-pseudo-reloc switch. This leaves the task
3060 of adjusting references in your client code for runtime environment, so
3061 this method works only when runtime environment supports this feature.
3063 A second solution is to force one of the 'constants' to be a variable --
3064 that is, unknown and un-optimizable at compile time. For arrays,
3065 there are two possibilities: a) make the indexee (the array's address)
3066 a variable, or b) make the 'constant' index a variable. Thus:
3069 extern type extern_array[];
3071 @{ volatile type *t=extern_array; t[1] @}
3077 extern type extern_array[];
3079 @{ volatile int t=1; extern_array[t] @}
3082 For structs (and most other multiword data types) the only option
3083 is to make the struct itself (or the long long, or the ...) variable:
3086 extern struct s extern_struct;
3087 extern_struct.field -->
3088 @{ volatile struct s *t=&extern_struct; t->field @}
3094 extern long long extern_ll;
3096 @{ volatile long long * local_ll=&extern_ll; *local_ll @}
3099 A third method of dealing with this difficulty is to abandon
3100 'auto-import' for the offending symbol and mark it with
3101 @code{__declspec(dllimport)}. However, in practice that
3102 requires using compile-time #defines to indicate whether you are
3103 building a DLL, building client code that will link to the DLL, or
3104 merely building/linking to a static library. In making the choice
3105 between the various methods of resolving the 'direct address with
3106 constant offset' problem, you should consider typical real-world usage:
3114 void main(int argc, char **argv)@{
3115 printf("%d\n",arr[1]);
3125 void main(int argc, char **argv)@{
3126 /* This workaround is for win32 and cygwin; do not "optimize" */
3127 volatile int *parr = arr;
3128 printf("%d\n",parr[1]);
3135 /* Note: auto-export is assumed (no __declspec(dllexport)) */
3136 #if (defined(_WIN32) || defined(__CYGWIN__)) && \
3137 !(defined(FOO_BUILD_DLL) || defined(FOO_STATIC))
3138 #define FOO_IMPORT __declspec(dllimport)
3142 extern FOO_IMPORT int arr[];
3145 void main(int argc, char **argv)@{
3146 printf("%d\n",arr[1]);
3150 A fourth way to avoid this problem is to re-code your
3151 library to use a functional interface rather than a data interface
3152 for the offending variables (e.g. set_foo() and get_foo() accessor
3155 @kindex --disable-auto-import
3156 @item --disable-auto-import
3157 Do not attempt to do sophisticated linking of @code{_symbol} to
3158 @code{__imp__symbol} for DATA imports from DLLs.
3159 [This option is specific to the i386 PE targeted port of the linker]
3161 @kindex --enable-runtime-pseudo-reloc
3162 @item --enable-runtime-pseudo-reloc
3163 If your code contains expressions described in --enable-auto-import section,
3164 that is, DATA imports from DLL with non-zero offset, this switch will create
3165 a vector of 'runtime pseudo relocations' which can be used by runtime
3166 environment to adjust references to such data in your client code.
3167 [This option is specific to the i386 PE targeted port of the linker]
3169 @kindex --disable-runtime-pseudo-reloc
3170 @item --disable-runtime-pseudo-reloc
3171 Do not create pseudo relocations for non-zero offset DATA imports from DLLs.
3172 [This option is specific to the i386 PE targeted port of the linker]
3174 @kindex --enable-extra-pe-debug
3175 @item --enable-extra-pe-debug
3176 Show additional debug info related to auto-import symbol thunking.
3177 [This option is specific to the i386 PE targeted port of the linker]
3179 @kindex --section-alignment
3180 @item --section-alignment
3181 Sets the section alignment. Sections in memory will always begin at
3182 addresses which are a multiple of this number. Defaults to 0x1000.
3183 [This option is specific to the i386 PE targeted port of the linker]
3187 @item --stack @var{reserve}
3188 @itemx --stack @var{reserve},@var{commit}
3189 Specify the number of bytes of memory to reserve (and optionally commit)
3190 to be used as stack for this program. The default is 2MB reserved, 4K
3192 [This option is specific to the i386 PE targeted port of the linker]
3195 @item --subsystem @var{which}
3196 @itemx --subsystem @var{which}:@var{major}
3197 @itemx --subsystem @var{which}:@var{major}.@var{minor}
3198 Specifies the subsystem under which your program will execute. The
3199 legal values for @var{which} are @code{native}, @code{windows},
3200 @code{console}, @code{posix}, and @code{xbox}. You may optionally set
3201 the subsystem version also. Numeric values are also accepted for
3203 [This option is specific to the i386 PE targeted port of the linker]
3205 The following options set flags in the @code{DllCharacteristics} field
3206 of the PE file header:
3207 [These options are specific to PE targeted ports of the linker]
3209 @kindex --high-entropy-va
3210 @item --high-entropy-va
3211 @itemx --disable-high-entropy-va
3212 Image is compatible with 64-bit address space layout randomization
3213 (ASLR). This option is enabled by default for 64-bit PE images.
3215 This option also implies @option{--dynamicbase} and
3216 @option{--enable-reloc-section}.
3218 @kindex --dynamicbase
3220 @itemx --disable-dynamicbase
3221 The image base address may be relocated using address space layout
3222 randomization (ASLR). This feature was introduced with MS Windows
3223 Vista for i386 PE targets. This option is enabled by default but
3224 can be disabled via the @option{--disable-dynamicbase} option.
3225 This option also implies @option{--enable-reloc-section}.
3227 @kindex --forceinteg
3229 @itemx --disable-forceinteg
3230 Code integrity checks are enforced. This option is disabled by
3235 @item --disable-nxcompat
3236 The image is compatible with the Data Execution Prevention.
3237 This feature was introduced with MS Windows XP SP2 for i386 PE
3238 targets. The option is enabled by default.
3240 @kindex --no-isolation
3241 @item --no-isolation
3242 @itemx --disable-no-isolation
3243 Although the image understands isolation, do not isolate the image.
3244 This option is disabled by default.
3248 @itemx --disable-no-seh
3249 The image does not use SEH. No SE handler may be called from
3250 this image. This option is disabled by default.
3254 @itemx --disable-no-bind
3255 Do not bind this image. This option is disabled by default.
3259 @itemx --disable-wdmdriver
3260 The driver uses the MS Windows Driver Model. This option is disabled
3265 @itemx --disable-tsaware
3266 The image is Terminal Server aware. This option is disabled by
3269 @kindex --insert-timestamp
3270 @item --insert-timestamp
3271 @itemx --no-insert-timestamp
3272 Insert a real timestamp into the image. This is the default behaviour
3273 as it matches legacy code and it means that the image will work with
3274 other, proprietary tools. The problem with this default is that it
3275 will result in slightly different images being produced each time the
3276 same sources are linked. The option @option{--no-insert-timestamp}
3277 can be used to insert a zero value for the timestamp, this ensuring
3278 that binaries produced from identical sources will compare
3281 @kindex --enable-reloc-section
3282 @item --enable-reloc-section
3283 @itemx --disable-reloc-section
3284 Create the base relocation table, which is necessary if the image
3285 is loaded at a different image base than specified in the PE header.
3286 This option is enabled by default.
3292 @subsection Options specific to C6X uClinux targets
3294 @c man begin OPTIONS
3296 The C6X uClinux target uses a binary format called DSBT to support shared
3297 libraries. Each shared library in the system needs to have a unique index;
3298 all executables use an index of 0.
3303 @item --dsbt-size @var{size}
3304 This option sets the number of entries in the DSBT of the current executable
3305 or shared library to @var{size}. The default is to create a table with 64
3308 @kindex --dsbt-index
3309 @item --dsbt-index @var{index}
3310 This option sets the DSBT index of the current executable or shared library
3311 to @var{index}. The default is 0, which is appropriate for generating
3312 executables. If a shared library is generated with a DSBT index of 0, the
3313 @code{R_C6000_DSBT_INDEX} relocs are copied into the output file.
3315 @kindex --no-merge-exidx-entries
3316 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent
3317 exidx entries in frame unwind info.
3325 @subsection Options specific to C-SKY targets
3327 @c man begin OPTIONS
3331 @kindex --branch-stub on C-SKY
3333 This option enables linker branch relaxation by inserting branch stub
3334 sections when needed to extend the range of branches. This option is
3335 usually not required since C-SKY supports branch and call instructions that
3336 can access the full memory range and branch relaxation is normally handled by
3337 the compiler or assembler.
3339 @kindex --stub-group-size on C-SKY
3340 @item --stub-group-size=@var{N}
3341 This option allows finer control of linker branch stub creation.
3342 It sets the maximum size of a group of input sections that can
3343 be handled by one stub section. A negative value of @var{N} locates
3344 stub sections after their branches, while a positive value allows stub
3345 sections to appear either before or after the branches. Values of
3346 @samp{1} or @samp{-1} indicate that the
3347 linker should choose suitable defaults.
3355 @subsection Options specific to Motorola 68HC11 and 68HC12 targets
3357 @c man begin OPTIONS
3359 The 68HC11 and 68HC12 linkers support specific options to control the
3360 memory bank switching mapping and trampoline code generation.
3364 @kindex --no-trampoline
3365 @item --no-trampoline
3366 This option disables the generation of trampoline. By default a trampoline
3367 is generated for each far function which is called using a @code{jsr}
3368 instruction (this happens when a pointer to a far function is taken).
3370 @kindex --bank-window
3371 @item --bank-window @var{name}
3372 This option indicates to the linker the name of the memory region in
3373 the @samp{MEMORY} specification that describes the memory bank window.
3374 The definition of such region is then used by the linker to compute
3375 paging and addresses within the memory window.
3383 @subsection Options specific to Motorola 68K target
3385 @c man begin OPTIONS
3387 The following options are supported to control handling of GOT generation
3388 when linking for 68K targets.
3393 @item --got=@var{type}
3394 This option tells the linker which GOT generation scheme to use.
3395 @var{type} should be one of @samp{single}, @samp{negative},
3396 @samp{multigot} or @samp{target}. For more information refer to the
3397 Info entry for @file{ld}.
3405 @subsection Options specific to MIPS targets
3407 @c man begin OPTIONS
3409 The following options are supported to control microMIPS instruction
3410 generation and branch relocation checks for ISA mode transitions when
3411 linking for MIPS targets.
3419 These options control the choice of microMIPS instructions used in code
3420 generated by the linker, such as that in the PLT or lazy binding stubs,
3421 or in relaxation. If @samp{--insn32} is used, then the linker only uses
3422 32-bit instruction encodings. By default or if @samp{--no-insn32} is
3423 used, all instruction encodings are used, including 16-bit ones where
3426 @kindex --ignore-branch-isa
3427 @item --ignore-branch-isa
3428 @kindex --no-ignore-branch-isa
3429 @itemx --no-ignore-branch-isa
3430 These options control branch relocation checks for invalid ISA mode
3431 transitions. If @samp{--ignore-branch-isa} is used, then the linker
3432 accepts any branch relocations and any ISA mode transition required
3433 is lost in relocation calculation, except for some cases of @code{BAL}
3434 instructions which meet relaxation conditions and are converted to
3435 equivalent @code{JALX} instructions as the associated relocation is
3436 calculated. By default or if @samp{--no-ignore-branch-isa} is used
3437 a check is made causing the loss of an ISA mode transition to produce
3440 @kindex --compact-branches
3441 @item --compact-branches
3442 @kindex --no-compact-branches
3443 @itemx --no-compact-branches
3444 These options control the generation of compact instructions by the linker
3445 in the PLT entries for MIPS R6.
3454 @subsection Options specific to PDP11 targets
3456 @c man begin OPTIONS
3458 For the pdp11-aout target, three variants of the output format can be
3459 produced as selected by the following options. The default variant
3460 for pdp11-aout is the @samp{--omagic} option, whereas for other
3461 targets @samp{--nmagic} is the default. The @samp{--imagic} option is
3462 defined only for the pdp11-aout target, while the others are described
3463 here as they apply to the pdp11-aout target.
3472 Mark the output as @code{OMAGIC} (0407) in the @file{a.out} header to
3473 indicate that the text segment is not to be write-protected and
3474 shared. Since the text and data sections are both readable and
3475 writable, the data section is allocated immediately contiguous after
3476 the text segment. This is the oldest format for PDP11 executable
3477 programs and is the default for @command{ld} on PDP11 Unix systems
3478 from the beginning through 2.11BSD.
3485 Mark the output as @code{NMAGIC} (0410) in the @file{a.out} header to
3486 indicate that when the output file is executed, the text portion will
3487 be read-only and shareable among all processes executing the same
3488 file. This involves moving the data areas up to the first possible 8K
3489 byte page boundary following the end of the text. This option creates
3490 a @emph{pure executable} format.
3497 Mark the output as @code{IMAGIC} (0411) in the @file{a.out} header to
3498 indicate that when the output file is executed, the program text and
3499 data areas will be loaded into separate address spaces using the split
3500 instruction and data space feature of the memory management unit in
3501 larger models of the PDP11. This doubles the address space available
3502 to the program. The text segment is again pure, write-protected, and
3503 shareable. The only difference in the output format between this
3504 option and the others, besides the magic number, is that both the text
3505 and data sections start at location 0. The @samp{-z} option selected
3506 this format in 2.11BSD. This option creates a @emph{separate
3512 Equivalent to @samp{--nmagic} for pdp11-aout.
3521 @section Environment Variables
3523 @c man begin ENVIRONMENT
3525 You can change the behaviour of @command{ld} with the environment variables
3526 @ifclear SingleFormat
3529 @code{LDEMULATION} and @code{COLLECT_NO_DEMANGLE}.
3531 @ifclear SingleFormat
3533 @cindex default input format
3534 @code{GNUTARGET} determines the input-file object format if you don't
3535 use @samp{-b} (or its synonym @samp{--format}). Its value should be one
3536 of the BFD names for an input format (@pxref{BFD}). If there is no
3537 @code{GNUTARGET} in the environment, @command{ld} uses the natural format
3538 of the target. If @code{GNUTARGET} is set to @code{default} then BFD
3539 attempts to discover the input format by examining binary input files;
3540 this method often succeeds, but there are potential ambiguities, since
3541 there is no method of ensuring that the magic number used to specify
3542 object-file formats is unique. However, the configuration procedure for
3543 BFD on each system places the conventional format for that system first
3544 in the search-list, so ambiguities are resolved in favor of convention.
3548 @cindex default emulation
3549 @cindex emulation, default
3550 @code{LDEMULATION} determines the default emulation if you don't use the
3551 @samp{-m} option. The emulation can affect various aspects of linker
3552 behaviour, particularly the default linker script. You can list the
3553 available emulations with the @samp{--verbose} or @samp{-V} options. If
3554 the @samp{-m} option is not used, and the @code{LDEMULATION} environment
3555 variable is not defined, the default emulation depends upon how the
3556 linker was configured.
3558 @kindex COLLECT_NO_DEMANGLE
3559 @cindex demangling, default
3560 Normally, the linker will default to demangling symbols. However, if
3561 @code{COLLECT_NO_DEMANGLE} is set in the environment, then it will
3562 default to not demangling symbols. This environment variable is used in
3563 a similar fashion by the @code{gcc} linker wrapper program. The default
3564 may be overridden by the @samp{--demangle} and @samp{--no-demangle}
3571 @chapter Linker Scripts
3574 @cindex linker scripts
3575 @cindex command files
3576 Every link is controlled by a @dfn{linker script}. This script is
3577 written in the linker command language.
3579 The main purpose of the linker script is to describe how the sections in
3580 the input files should be mapped into the output file, and to control
3581 the memory layout of the output file. Most linker scripts do nothing
3582 more than this. However, when necessary, the linker script can also
3583 direct the linker to perform many other operations, using the commands
3586 The linker always uses a linker script. If you do not supply one
3587 yourself, the linker will use a default script that is compiled into the
3588 linker executable. You can use the @samp{--verbose} command-line option
3589 to display the default linker script. Certain command-line options,
3590 such as @samp{-r} or @samp{-N}, will affect the default linker script.
3592 You may supply your own linker script by using the @samp{-T} command
3593 line option. When you do this, your linker script will replace the
3594 default linker script.
3596 You may also use linker scripts implicitly by naming them as input files
3597 to the linker, as though they were files to be linked. @xref{Implicit
3601 * Basic Script Concepts:: Basic Linker Script Concepts
3602 * Script Format:: Linker Script Format
3603 * Simple Example:: Simple Linker Script Example
3604 * Simple Commands:: Simple Linker Script Commands
3605 * Assignments:: Assigning Values to Symbols
3606 * SECTIONS:: SECTIONS Command
3607 * MEMORY:: MEMORY Command
3608 * PHDRS:: PHDRS Command
3609 * VERSION:: VERSION Command
3610 * Expressions:: Expressions in Linker Scripts
3611 * Implicit Linker Scripts:: Implicit Linker Scripts
3614 @node Basic Script Concepts
3615 @section Basic Linker Script Concepts
3616 @cindex linker script concepts
3617 We need to define some basic concepts and vocabulary in order to
3618 describe the linker script language.
3620 The linker combines input files into a single output file. The output
3621 file and each input file are in a special data format known as an
3622 @dfn{object file format}. Each file is called an @dfn{object file}.
3623 The output file is often called an @dfn{executable}, but for our
3624 purposes we will also call it an object file. Each object file has,
3625 among other things, a list of @dfn{sections}. We sometimes refer to a
3626 section in an input file as an @dfn{input section}; similarly, a section
3627 in the output file is an @dfn{output section}.
3629 Each section in an object file has a name and a size. Most sections
3630 also have an associated block of data, known as the @dfn{section
3631 contents}. A section may be marked as @dfn{loadable}, which means that
3632 the contents should be loaded into memory when the output file is run.
3633 A section with no contents may be @dfn{allocatable}, which means that an
3634 area in memory should be set aside, but nothing in particular should be
3635 loaded there (in some cases this memory must be zeroed out). A section
3636 which is neither loadable nor allocatable typically contains some sort
3637 of debugging information.
3639 Every loadable or allocatable output section has two addresses. The
3640 first is the @dfn{VMA}, or virtual memory address. This is the address
3641 the section will have when the output file is run. The second is the
3642 @dfn{LMA}, or load memory address. This is the address at which the
3643 section will be loaded. In most cases the two addresses will be the
3644 same. An example of when they might be different is when a data section
3645 is loaded into ROM, and then copied into RAM when the program starts up
3646 (this technique is often used to initialize global variables in a ROM
3647 based system). In this case the ROM address would be the LMA, and the
3648 RAM address would be the VMA.
3650 You can see the sections in an object file by using the @code{objdump}
3651 program with the @samp{-h} option.
3653 Every object file also has a list of @dfn{symbols}, known as the
3654 @dfn{symbol table}. A symbol may be defined or undefined. Each symbol
3655 has a name, and each defined symbol has an address, among other
3656 information. If you compile a C or C++ program into an object file, you
3657 will get a defined symbol for every defined function and global or
3658 static variable. Every undefined function or global variable which is
3659 referenced in the input file will become an undefined symbol.
3661 You can see the symbols in an object file by using the @code{nm}
3662 program, or by using the @code{objdump} program with the @samp{-t}
3666 @section Linker Script Format
3667 @cindex linker script format
3668 Linker scripts are text files.
3670 You write a linker script as a series of commands. Each command is
3671 either a keyword, possibly followed by arguments, or an assignment to a
3672 symbol. You may separate commands using semicolons. Whitespace is
3675 Strings such as file or format names can normally be entered directly.
3676 If the file name contains a character such as a comma which would
3677 otherwise serve to separate file names, you may put the file name in
3678 double quotes. There is no way to use a double quote character in a
3681 You may include comments in linker scripts just as in C, delimited by
3682 @samp{/*} and @samp{*/}. As in C, comments are syntactically equivalent
3685 @node Simple Example
3686 @section Simple Linker Script Example
3687 @cindex linker script example
3688 @cindex example of linker script
3689 Many linker scripts are fairly simple.
3691 The simplest possible linker script has just one command:
3692 @samp{SECTIONS}. You use the @samp{SECTIONS} command to describe the
3693 memory layout of the output file.
3695 The @samp{SECTIONS} command is a powerful command. Here we will
3696 describe a simple use of it. Let's assume your program consists only of
3697 code, initialized data, and uninitialized data. These will be in the
3698 @samp{.text}, @samp{.data}, and @samp{.bss} sections, respectively.
3699 Let's assume further that these are the only sections which appear in
3702 For this example, let's say that the code should be loaded at address
3703 0x10000, and that the data should start at address 0x8000000. Here is a
3704 linker script which will do that:
3709 .text : @{ *(.text) @}
3711 .data : @{ *(.data) @}
3712 .bss : @{ *(.bss) @}
3716 You write the @samp{SECTIONS} command as the keyword @samp{SECTIONS},
3717 followed by a series of symbol assignments and output section
3718 descriptions enclosed in curly braces.
3720 The first line inside the @samp{SECTIONS} command of the above example
3721 sets the value of the special symbol @samp{.}, which is the location
3722 counter. If you do not specify the address of an output section in some
3723 other way (other ways are described later), the address is set from the
3724 current value of the location counter. The location counter is then
3725 incremented by the size of the output section. At the start of the
3726 @samp{SECTIONS} command, the location counter has the value @samp{0}.
3728 The second line defines an output section, @samp{.text}. The colon is
3729 required syntax which may be ignored for now. Within the curly braces
3730 after the output section name, you list the names of the input sections
3731 which should be placed into this output section. The @samp{*} is a
3732 wildcard which matches any file name. The expression @samp{*(.text)}
3733 means all @samp{.text} input sections in all input files.
3735 Since the location counter is @samp{0x10000} when the output section
3736 @samp{.text} is defined, the linker will set the address of the
3737 @samp{.text} section in the output file to be @samp{0x10000}.
3739 The remaining lines define the @samp{.data} and @samp{.bss} sections in
3740 the output file. The linker will place the @samp{.data} output section
3741 at address @samp{0x8000000}. After the linker places the @samp{.data}
3742 output section, the value of the location counter will be
3743 @samp{0x8000000} plus the size of the @samp{.data} output section. The
3744 effect is that the linker will place the @samp{.bss} output section
3745 immediately after the @samp{.data} output section in memory.
3747 The linker will ensure that each output section has the required
3748 alignment, by increasing the location counter if necessary. In this
3749 example, the specified addresses for the @samp{.text} and @samp{.data}
3750 sections will probably satisfy any alignment constraints, but the linker
3751 may have to create a small gap between the @samp{.data} and @samp{.bss}
3754 That's it! That's a simple and complete linker script.
3756 @node Simple Commands
3757 @section Simple Linker Script Commands
3758 @cindex linker script simple commands
3759 In this section we describe the simple linker script commands.
3762 * Entry Point:: Setting the entry point
3763 * File Commands:: Commands dealing with files
3764 @ifclear SingleFormat
3765 * Format Commands:: Commands dealing with object file formats
3768 * REGION_ALIAS:: Assign alias names to memory regions
3769 * Miscellaneous Commands:: Other linker script commands
3773 @subsection Setting the Entry Point
3774 @kindex ENTRY(@var{symbol})
3775 @cindex start of execution
3776 @cindex first instruction
3778 The first instruction to execute in a program is called the @dfn{entry
3779 point}. You can use the @code{ENTRY} linker script command to set the
3780 entry point. The argument is a symbol name:
3785 There are several ways to set the entry point. The linker will set the
3786 entry point by trying each of the following methods in order, and
3787 stopping when one of them succeeds:
3790 the @samp{-e} @var{entry} command-line option;
3792 the @code{ENTRY(@var{symbol})} command in a linker script;
3794 the value of a target-specific symbol, if it is defined; For many
3795 targets this is @code{start}, but PE- and BeOS-based systems for example
3796 check a list of possible entry symbols, matching the first one found.
3798 the address of the first byte of the @samp{.text} section, if present;
3800 The address @code{0}.
3804 @subsection Commands Dealing with Files
3805 @cindex linker script file commands
3806 Several linker script commands deal with files.
3809 @item INCLUDE @var{filename}
3810 @kindex INCLUDE @var{filename}
3811 @cindex including a linker script
3812 Include the linker script @var{filename} at this point. The file will
3813 be searched for in the current directory, and in any directory specified
3814 with the @option{-L} option. You can nest calls to @code{INCLUDE} up to
3817 You can place @code{INCLUDE} directives at the top level, in @code{MEMORY} or
3818 @code{SECTIONS} commands, or in output section descriptions.
3820 @item INPUT(@var{file}, @var{file}, @dots{})
3821 @itemx INPUT(@var{file} @var{file} @dots{})
3822 @kindex INPUT(@var{files})
3823 @cindex input files in linker scripts
3824 @cindex input object files in linker scripts
3825 @cindex linker script input object files
3826 The @code{INPUT} command directs the linker to include the named files
3827 in the link, as though they were named on the command line.
3829 For example, if you always want to include @file{subr.o} any time you do
3830 a link, but you can't be bothered to put it on every link command line,
3831 then you can put @samp{INPUT (subr.o)} in your linker script.
3833 In fact, if you like, you can list all of your input files in the linker
3834 script, and then invoke the linker with nothing but a @samp{-T} option.
3836 In case a @dfn{sysroot prefix} is configured, and the filename starts
3837 with the @samp{/} character, and the script being processed was
3838 located inside the @dfn{sysroot prefix}, the filename will be looked
3839 for in the @dfn{sysroot prefix}. The @dfn{sysroot prefix} can also be forced by specifying
3840 @code{=} as the first character in the filename path, or prefixing the
3841 filename path with @code{$SYSROOT}. See also the description of
3842 @samp{-L} in @ref{Options,,Command-line Options}.
3844 If a @dfn{sysroot prefix} is not used then the linker will try to open
3845 the file in the directory containing the linker script. If it is not
3846 found the linker will then search the current directory. If it is still
3847 not found the linker will search through the archive library search
3850 If you use @samp{INPUT (-l@var{file})}, @command{ld} will transform the
3851 name to @code{lib@var{file}.a}, as with the command-line argument
3854 When you use the @code{INPUT} command in an implicit linker script, the
3855 files will be included in the link at the point at which the linker
3856 script file is included. This can affect archive searching.
3858 @item GROUP(@var{file}, @var{file}, @dots{})
3859 @itemx GROUP(@var{file} @var{file} @dots{})
3860 @kindex GROUP(@var{files})
3861 @cindex grouping input files
3862 The @code{GROUP} command is like @code{INPUT}, except that the named
3863 files should all be archives, and they are searched repeatedly until no
3864 new undefined references are created. See the description of @samp{-(}
3865 in @ref{Options,,Command-line Options}.
3867 @item AS_NEEDED(@var{file}, @var{file}, @dots{})
3868 @itemx AS_NEEDED(@var{file} @var{file} @dots{})
3869 @kindex AS_NEEDED(@var{files})
3870 This construct can appear only inside of the @code{INPUT} or @code{GROUP}
3871 commands, among other filenames. The files listed will be handled
3872 as if they appear directly in the @code{INPUT} or @code{GROUP} commands,
3873 with the exception of ELF shared libraries, that will be added only
3874 when they are actually needed. This construct essentially enables
3875 @option{--as-needed} option for all the files listed inside of it
3876 and restores previous @option{--as-needed} resp. @option{--no-as-needed}
3879 @item OUTPUT(@var{filename})
3880 @kindex OUTPUT(@var{filename})
3881 @cindex output file name in linker script
3882 The @code{OUTPUT} command names the output file. Using
3883 @code{OUTPUT(@var{filename})} in the linker script is exactly like using
3884 @samp{-o @var{filename}} on the command line (@pxref{Options,,Command
3885 Line Options}). If both are used, the command-line option takes
3888 You can use the @code{OUTPUT} command to define a default name for the
3889 output file other than the usual default of @file{a.out}.
3891 @item SEARCH_DIR(@var{path})
3892 @kindex SEARCH_DIR(@var{path})
3893 @cindex library search path in linker script
3894 @cindex archive search path in linker script
3895 @cindex search path in linker script
3896 The @code{SEARCH_DIR} command adds @var{path} to the list of paths where
3897 @command{ld} looks for archive libraries. Using
3898 @code{SEARCH_DIR(@var{path})} is exactly like using @samp{-L @var{path}}
3899 on the command line (@pxref{Options,,Command-line Options}). If both
3900 are used, then the linker will search both paths. Paths specified using
3901 the command-line option are searched first.
3903 @item STARTUP(@var{filename})
3904 @kindex STARTUP(@var{filename})
3905 @cindex first input file
3906 The @code{STARTUP} command is just like the @code{INPUT} command, except
3907 that @var{filename} will become the first input file to be linked, as
3908 though it were specified first on the command line. This may be useful
3909 when using a system in which the entry point is always the start of the
3913 @ifclear SingleFormat
3914 @node Format Commands
3915 @subsection Commands Dealing with Object File Formats
3916 A couple of linker script commands deal with object file formats.
3919 @item OUTPUT_FORMAT(@var{bfdname})
3920 @itemx OUTPUT_FORMAT(@var{default}, @var{big}, @var{little})
3921 @kindex OUTPUT_FORMAT(@var{bfdname})
3922 @cindex output file format in linker script
3923 The @code{OUTPUT_FORMAT} command names the BFD format to use for the
3924 output file (@pxref{BFD}). Using @code{OUTPUT_FORMAT(@var{bfdname})} is
3925 exactly like using @samp{--oformat @var{bfdname}} on the command line
3926 (@pxref{Options,,Command-line Options}). If both are used, the command
3927 line option takes precedence.
3929 You can use @code{OUTPUT_FORMAT} with three arguments to use different
3930 formats based on the @samp{-EB} and @samp{-EL} command-line options.
3931 This permits the linker script to set the output format based on the
3934 If neither @samp{-EB} nor @samp{-EL} are used, then the output format
3935 will be the first argument, @var{default}. If @samp{-EB} is used, the
3936 output format will be the second argument, @var{big}. If @samp{-EL} is
3937 used, the output format will be the third argument, @var{little}.
3939 For example, the default linker script for the MIPS ELF target uses this
3942 OUTPUT_FORMAT(elf32-bigmips, elf32-bigmips, elf32-littlemips)
3944 This says that the default format for the output file is
3945 @samp{elf32-bigmips}, but if the user uses the @samp{-EL} command-line
3946 option, the output file will be created in the @samp{elf32-littlemips}
3949 @item TARGET(@var{bfdname})
3950 @kindex TARGET(@var{bfdname})
3951 @cindex input file format in linker script
3952 The @code{TARGET} command names the BFD format to use when reading input
3953 files. It affects subsequent @code{INPUT} and @code{GROUP} commands.
3954 This command is like using @samp{-b @var{bfdname}} on the command line
3955 (@pxref{Options,,Command-line Options}). If the @code{TARGET} command
3956 is used but @code{OUTPUT_FORMAT} is not, then the last @code{TARGET}
3957 command is also used to set the format for the output file. @xref{BFD}.
3962 @subsection Assign alias names to memory regions
3963 @kindex REGION_ALIAS(@var{alias}, @var{region})
3964 @cindex region alias
3965 @cindex region names
3967 Alias names can be added to existing memory regions created with the
3968 @ref{MEMORY} command. Each name corresponds to at most one memory region.
3971 REGION_ALIAS(@var{alias}, @var{region})
3974 The @code{REGION_ALIAS} function creates an alias name @var{alias} for the
3975 memory region @var{region}. This allows a flexible mapping of output sections
3976 to memory regions. An example follows.
3978 Suppose we have an application for embedded systems which come with various
3979 memory storage devices. All have a general purpose, volatile memory @code{RAM}
3980 that allows code execution or data storage. Some may have a read-only,
3981 non-volatile memory @code{ROM} that allows code execution and read-only data
3982 access. The last variant is a read-only, non-volatile memory @code{ROM2} with
3983 read-only data access and no code execution capability. We have four output
3988 @code{.text} program code;
3990 @code{.rodata} read-only data;
3992 @code{.data} read-write initialized data;
3994 @code{.bss} read-write zero initialized data.
3997 The goal is to provide a linker command file that contains a system independent
3998 part defining the output sections and a system dependent part mapping the
3999 output sections to the memory regions available on the system. Our embedded
4000 systems come with three different memory setups @code{A}, @code{B} and
4002 @multitable @columnfractions .25 .25 .25 .25
4003 @item Section @tab Variant A @tab Variant B @tab Variant C
4004 @item .text @tab RAM @tab ROM @tab ROM
4005 @item .rodata @tab RAM @tab ROM @tab ROM2
4006 @item .data @tab RAM @tab RAM/ROM @tab RAM/ROM2
4007 @item .bss @tab RAM @tab RAM @tab RAM
4009 The notation @code{RAM/ROM} or @code{RAM/ROM2} means that this section is
4010 loaded into region @code{ROM} or @code{ROM2} respectively. Please note that
4011 the load address of the @code{.data} section starts in all three variants at
4012 the end of the @code{.rodata} section.
4014 The base linker script that deals with the output sections follows. It
4015 includes the system dependent @code{linkcmds.memory} file that describes the
4018 INCLUDE linkcmds.memory
4031 .data : AT (rodata_end)
4036 data_size = SIZEOF(.data);
4037 data_load_start = LOADADDR(.data);
4045 Now we need three different @code{linkcmds.memory} files to define memory
4046 regions and alias names. The content of @code{linkcmds.memory} for the three
4047 variants @code{A}, @code{B} and @code{C}:
4050 Here everything goes into the @code{RAM}.
4054 RAM : ORIGIN = 0, LENGTH = 4M
4057 REGION_ALIAS("REGION_TEXT", RAM);
4058 REGION_ALIAS("REGION_RODATA", RAM);
4059 REGION_ALIAS("REGION_DATA", RAM);
4060 REGION_ALIAS("REGION_BSS", RAM);
4063 Program code and read-only data go into the @code{ROM}. Read-write data goes
4064 into the @code{RAM}. An image of the initialized data is loaded into the
4065 @code{ROM} and will be copied during system start into the @code{RAM}.
4069 ROM : ORIGIN = 0, LENGTH = 3M
4070 RAM : ORIGIN = 0x10000000, LENGTH = 1M
4073 REGION_ALIAS("REGION_TEXT", ROM);
4074 REGION_ALIAS("REGION_RODATA", ROM);
4075 REGION_ALIAS("REGION_DATA", RAM);
4076 REGION_ALIAS("REGION_BSS", RAM);
4079 Program code goes into the @code{ROM}. Read-only data goes into the
4080 @code{ROM2}. Read-write data goes into the @code{RAM}. An image of the
4081 initialized data is loaded into the @code{ROM2} and will be copied during
4082 system start into the @code{RAM}.
4086 ROM : ORIGIN = 0, LENGTH = 2M
4087 ROM2 : ORIGIN = 0x10000000, LENGTH = 1M
4088 RAM : ORIGIN = 0x20000000, LENGTH = 1M
4091 REGION_ALIAS("REGION_TEXT", ROM);
4092 REGION_ALIAS("REGION_RODATA", ROM2);
4093 REGION_ALIAS("REGION_DATA", RAM);
4094 REGION_ALIAS("REGION_BSS", RAM);
4098 It is possible to write a common system initialization routine to copy the
4099 @code{.data} section from @code{ROM} or @code{ROM2} into the @code{RAM} if
4104 extern char data_start [];
4105 extern char data_size [];
4106 extern char data_load_start [];
4108 void copy_data(void)
4110 if (data_start != data_load_start)
4112 memcpy(data_start, data_load_start, (size_t) data_size);
4117 @node Miscellaneous Commands
4118 @subsection Other Linker Script Commands
4119 There are a few other linker scripts commands.
4122 @item ASSERT(@var{exp}, @var{message})
4124 @cindex assertion in linker script
4125 Ensure that @var{exp} is non-zero. If it is zero, then exit the linker
4126 with an error code, and print @var{message}.
4128 Note that assertions are checked before the final stages of linking
4129 take place. This means that expressions involving symbols PROVIDEd
4130 inside section definitions will fail if the user has not set values
4131 for those symbols. The only exception to this rule is PROVIDEd
4132 symbols that just reference dot. Thus an assertion like this:
4137 PROVIDE (__stack = .);
4138 PROVIDE (__stack_size = 0x100);
4139 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
4143 will fail if @code{__stack_size} is not defined elsewhere. Symbols
4144 PROVIDEd outside of section definitions are evaluated earlier, so they
4145 can be used inside ASSERTions. Thus:
4148 PROVIDE (__stack_size = 0x100);
4151 PROVIDE (__stack = .);
4152 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
4158 @item EXTERN(@var{symbol} @var{symbol} @dots{})
4160 @cindex undefined symbol in linker script
4161 Force @var{symbol} to be entered in the output file as an undefined
4162 symbol. Doing this may, for example, trigger linking of additional
4163 modules from standard libraries. You may list several @var{symbol}s for
4164 each @code{EXTERN}, and you may use @code{EXTERN} multiple times. This
4165 command has the same effect as the @samp{-u} command-line option.
4167 @item FORCE_COMMON_ALLOCATION
4168 @kindex FORCE_COMMON_ALLOCATION
4169 @cindex common allocation in linker script
4170 This command has the same effect as the @samp{-d} command-line option:
4171 to make @command{ld} assign space to common symbols even if a relocatable
4172 output file is specified (@samp{-r}).
4174 @item INHIBIT_COMMON_ALLOCATION
4175 @kindex INHIBIT_COMMON_ALLOCATION
4176 @cindex common allocation in linker script
4177 This command has the same effect as the @samp{--no-define-common}
4178 command-line option: to make @code{ld} omit the assignment of addresses
4179 to common symbols even for a non-relocatable output file.
4181 @item FORCE_GROUP_ALLOCATION
4182 @kindex FORCE_GROUP_ALLOCATION
4183 @cindex group allocation in linker script
4184 @cindex section groups
4186 This command has the same effect as the
4187 @samp{--force-group-allocation} command-line option: to make
4188 @command{ld} place section group members like normal input sections,
4189 and to delete the section groups even if a relocatable output file is
4190 specified (@samp{-r}).
4192 @item INSERT [ AFTER | BEFORE ] @var{output_section}
4194 @cindex insert user script into default script
4195 This command is typically used in a script specified by @samp{-T} to
4196 augment the default @code{SECTIONS} with, for example, overlays. It
4197 inserts all prior linker script statements after (or before)
4198 @var{output_section}, and also causes @samp{-T} to not override the
4199 default linker script. The exact insertion point is as for orphan
4200 sections. @xref{Location Counter}. The insertion happens after the
4201 linker has mapped input sections to output sections. Prior to the
4202 insertion, since @samp{-T} scripts are parsed before the default
4203 linker script, statements in the @samp{-T} script occur before the
4204 default linker script statements in the internal linker representation
4205 of the script. In particular, input section assignments will be made
4206 to @samp{-T} output sections before those in the default script. Here
4207 is an example of how a @samp{-T} script using @code{INSERT} might look:
4214 .ov1 @{ ov1*(.text) @}
4215 .ov2 @{ ov2*(.text) @}
4221 @item NOCROSSREFS(@var{section} @var{section} @dots{})
4222 @kindex NOCROSSREFS(@var{sections})
4223 @cindex cross references
4224 This command may be used to tell @command{ld} to issue an error about any
4225 references among certain output sections.
4227 In certain types of programs, particularly on embedded systems when
4228 using overlays, when one section is loaded into memory, another section
4229 will not be. Any direct references between the two sections would be
4230 errors. For example, it would be an error if code in one section called
4231 a function defined in the other section.
4233 The @code{NOCROSSREFS} command takes a list of output section names. If
4234 @command{ld} detects any cross references between the sections, it reports
4235 an error and returns a non-zero exit status. Note that the
4236 @code{NOCROSSREFS} command uses output section names, not input section
4239 @item NOCROSSREFS_TO(@var{tosection} @var{fromsection} @dots{})
4240 @kindex NOCROSSREFS_TO(@var{tosection} @var{fromsections})
4241 @cindex cross references
4242 This command may be used to tell @command{ld} to issue an error about any
4243 references to one section from a list of other sections.
4245 The @code{NOCROSSREFS} command is useful when ensuring that two or more
4246 output sections are entirely independent but there are situations where
4247 a one-way dependency is needed. For example, in a multi-core application
4248 there may be shared code that can be called from each core but for safety
4249 must never call back.
4251 The @code{NOCROSSREFS_TO} command takes a list of output section names.
4252 The first section can not be referenced from any of the other sections.
4253 If @command{ld} detects any references to the first section from any of
4254 the other sections, it reports an error and returns a non-zero exit
4255 status. Note that the @code{NOCROSSREFS_TO} command uses output section
4256 names, not input section names.
4258 @ifclear SingleFormat
4259 @item OUTPUT_ARCH(@var{bfdarch})
4260 @kindex OUTPUT_ARCH(@var{bfdarch})
4261 @cindex machine architecture
4262 @cindex architecture
4263 Specify a particular output machine architecture. The argument is one
4264 of the names used by the BFD library (@pxref{BFD}). You can see the
4265 architecture of an object file by using the @code{objdump} program with
4266 the @samp{-f} option.
4269 @item LD_FEATURE(@var{string})
4270 @kindex LD_FEATURE(@var{string})
4271 This command may be used to modify @command{ld} behavior. If
4272 @var{string} is @code{"SANE_EXPR"} then absolute symbols and numbers
4273 in a script are simply treated as numbers everywhere.
4274 @xref{Expression Section}.
4278 @section Assigning Values to Symbols
4279 @cindex assignment in scripts
4280 @cindex symbol definition, scripts
4281 @cindex variables, defining
4282 You may assign a value to a symbol in a linker script. This will define
4283 the symbol and place it into the symbol table with a global scope.
4286 * Simple Assignments:: Simple Assignments
4289 * PROVIDE_HIDDEN:: PROVIDE_HIDDEN
4290 * Source Code Reference:: How to use a linker script defined symbol in source code
4293 @node Simple Assignments
4294 @subsection Simple Assignments
4296 You may assign to a symbol using any of the C assignment operators:
4299 @item @var{symbol} = @var{expression} ;
4300 @itemx @var{symbol} += @var{expression} ;
4301 @itemx @var{symbol} -= @var{expression} ;
4302 @itemx @var{symbol} *= @var{expression} ;
4303 @itemx @var{symbol} /= @var{expression} ;
4304 @itemx @var{symbol} <<= @var{expression} ;
4305 @itemx @var{symbol} >>= @var{expression} ;
4306 @itemx @var{symbol} &= @var{expression} ;
4307 @itemx @var{symbol} |= @var{expression} ;
4310 The first case will define @var{symbol} to the value of
4311 @var{expression}. In the other cases, @var{symbol} must already be
4312 defined, and the value will be adjusted accordingly.
4314 The special symbol name @samp{.} indicates the location counter. You
4315 may only use this within a @code{SECTIONS} command. @xref{Location Counter}.
4317 The semicolon after @var{expression} is required.
4319 Expressions are defined below; see @ref{Expressions}.
4321 You may write symbol assignments as commands in their own right, or as
4322 statements within a @code{SECTIONS} command, or as part of an output
4323 section description in a @code{SECTIONS} command.
4325 The section of the symbol will be set from the section of the
4326 expression; for more information, see @ref{Expression Section}.
4328 Here is an example showing the three different places that symbol
4329 assignments may be used:
4340 _bdata = (. + 3) & ~ 3;
4341 .data : @{ *(.data) @}
4345 In this example, the symbol @samp{floating_point} will be defined as
4346 zero. The symbol @samp{_etext} will be defined as the address following
4347 the last @samp{.text} input section. The symbol @samp{_bdata} will be
4348 defined as the address following the @samp{.text} output section aligned
4349 upward to a 4 byte boundary.
4354 For ELF targeted ports, define a symbol that will be hidden and won't be
4355 exported. The syntax is @code{HIDDEN(@var{symbol} = @var{expression})}.
4357 Here is the example from @ref{Simple Assignments}, rewritten to use
4361 HIDDEN(floating_point = 0);
4369 HIDDEN(_bdata = (. + 3) & ~ 3);
4370 .data : @{ *(.data) @}
4374 In this case none of the three symbols will be visible outside this module.
4379 In some cases, it is desirable for a linker script to define a symbol
4380 only if it is referenced and is not defined by any object included in
4381 the link. For example, traditional linkers defined the symbol
4382 @samp{etext}. However, ANSI C requires that the user be able to use
4383 @samp{etext} as a function name without encountering an error. The
4384 @code{PROVIDE} keyword may be used to define a symbol, such as
4385 @samp{etext}, only if it is referenced but not defined. The syntax is
4386 @code{PROVIDE(@var{symbol} = @var{expression})}.
4388 Here is an example of using @code{PROVIDE} to define @samp{etext}:
4401 In this example, if the program defines @samp{_etext} (with a leading
4402 underscore), the linker will give a multiple definition error. If, on
4403 the other hand, the program defines @samp{etext} (with no leading
4404 underscore), the linker will silently use the definition in the program.
4405 If the program references @samp{etext} but does not define it, the
4406 linker will use the definition in the linker script.
4408 Note - the @code{PROVIDE} directive considers a common symbol to be
4409 defined, even though such a symbol could be combined with the symbol
4410 that the @code{PROVIDE} would create. This is particularly important
4411 when considering constructor and destructor list symbols such as
4412 @samp{__CTOR_LIST__} as these are often defined as common symbols.
4414 @node PROVIDE_HIDDEN
4415 @subsection PROVIDE_HIDDEN
4416 @cindex PROVIDE_HIDDEN
4417 Similar to @code{PROVIDE}. For ELF targeted ports, the symbol will be
4418 hidden and won't be exported.
4420 @node Source Code Reference
4421 @subsection Source Code Reference
4423 Accessing a linker script defined variable from source code is not
4424 intuitive. In particular a linker script symbol is not equivalent to
4425 a variable declaration in a high level language, it is instead a
4426 symbol that does not have a value.
4428 Before going further, it is important to note that compilers often
4429 transform names in the source code into different names when they are
4430 stored in the symbol table. For example, Fortran compilers commonly
4431 prepend or append an underscore, and C++ performs extensive @samp{name
4432 mangling}. Therefore there might be a discrepancy between the name
4433 of a variable as it is used in source code and the name of the same
4434 variable as it is defined in a linker script. For example in C a
4435 linker script variable might be referred to as:
4441 But in the linker script it might be defined as:
4447 In the remaining examples however it is assumed that no name
4448 transformation has taken place.
4450 When a symbol is declared in a high level language such as C, two
4451 things happen. The first is that the compiler reserves enough space
4452 in the program's memory to hold the @emph{value} of the symbol. The
4453 second is that the compiler creates an entry in the program's symbol
4454 table which holds the symbol's @emph{address}. ie the symbol table
4455 contains the address of the block of memory holding the symbol's
4456 value. So for example the following C declaration, at file scope:
4462 creates an entry called @samp{foo} in the symbol table. This entry
4463 holds the address of an @samp{int} sized block of memory where the
4464 number 1000 is initially stored.
4466 When a program references a symbol the compiler generates code that
4467 first accesses the symbol table to find the address of the symbol's
4468 memory block and then code to read the value from that memory block.
4475 looks up the symbol @samp{foo} in the symbol table, gets the address
4476 associated with this symbol and then writes the value 1 into that
4483 looks up the symbol @samp{foo} in the symbol table, gets its address
4484 and then copies this address into the block of memory associated with
4485 the variable @samp{a}.
4487 Linker scripts symbol declarations, by contrast, create an entry in
4488 the symbol table but do not assign any memory to them. Thus they are
4489 an address without a value. So for example the linker script definition:
4495 creates an entry in the symbol table called @samp{foo} which holds
4496 the address of memory location 1000, but nothing special is stored at
4497 address 1000. This means that you cannot access the @emph{value} of a
4498 linker script defined symbol - it has no value - all you can do is
4499 access the @emph{address} of a linker script defined symbol.
4501 Hence when you are using a linker script defined symbol in source code
4502 you should always take the address of the symbol, and never attempt to
4503 use its value. For example suppose you want to copy the contents of a
4504 section of memory called .ROM into a section called .FLASH and the
4505 linker script contains these declarations:
4509 start_of_ROM = .ROM;
4510 end_of_ROM = .ROM + sizeof (.ROM);
4511 start_of_FLASH = .FLASH;
4515 Then the C source code to perform the copy would be:
4519 extern char start_of_ROM, end_of_ROM, start_of_FLASH;
4521 memcpy (& start_of_FLASH, & start_of_ROM, & end_of_ROM - & start_of_ROM);
4525 Note the use of the @samp{&} operators. These are correct.
4526 Alternatively the symbols can be treated as the names of vectors or
4527 arrays and then the code will again work as expected:
4531 extern char start_of_ROM[], end_of_ROM[], start_of_FLASH[];
4533 memcpy (start_of_FLASH, start_of_ROM, end_of_ROM - start_of_ROM);
4537 Note how using this method does not require the use of @samp{&}
4541 @section SECTIONS Command
4543 The @code{SECTIONS} command tells the linker how to map input sections
4544 into output sections, and how to place the output sections in memory.
4546 The format of the @code{SECTIONS} command is:
4550 @var{sections-command}
4551 @var{sections-command}
4556 Each @var{sections-command} may of be one of the following:
4560 an @code{ENTRY} command (@pxref{Entry Point,,Entry command})
4562 a symbol assignment (@pxref{Assignments})
4564 an output section description
4566 an overlay description
4569 The @code{ENTRY} command and symbol assignments are permitted inside the
4570 @code{SECTIONS} command for convenience in using the location counter in
4571 those commands. This can also make the linker script easier to
4572 understand because you can use those commands at meaningful points in
4573 the layout of the output file.
4575 Output section descriptions and overlay descriptions are described
4578 If you do not use a @code{SECTIONS} command in your linker script, the
4579 linker will place each input section into an identically named output
4580 section in the order that the sections are first encountered in the
4581 input files. If all input sections are present in the first file, for
4582 example, the order of sections in the output file will match the order
4583 in the first input file. The first section will be at address zero.
4586 * Output Section Description:: Output section description
4587 * Output Section Name:: Output section name
4588 * Output Section Address:: Output section address
4589 * Input Section:: Input section description
4590 * Output Section Data:: Output section data
4591 * Output Section Keywords:: Output section keywords
4592 * Output Section Discarding:: Output section discarding
4593 * Output Section Attributes:: Output section attributes
4594 * Overlay Description:: Overlay description
4597 @node Output Section Description
4598 @subsection Output Section Description
4599 The full description of an output section looks like this:
4602 @var{section} [@var{address}] [(@var{type})] :
4604 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
4605 [SUBALIGN(@var{subsection_align})]
4608 @var{output-section-command}
4609 @var{output-section-command}
4611 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}] [,]
4615 Most output sections do not use most of the optional section attributes.
4617 The whitespace around @var{section} is required, so that the section
4618 name is unambiguous. The colon and the curly braces are also required.
4619 The comma at the end may be required if a @var{fillexp} is used and
4620 the next @var{sections-command} looks like a continuation of the expression.
4621 The line breaks and other white space are optional.
4623 Each @var{output-section-command} may be one of the following:
4627 a symbol assignment (@pxref{Assignments})
4629 an input section description (@pxref{Input Section})
4631 data values to include directly (@pxref{Output Section Data})
4633 a special output section keyword (@pxref{Output Section Keywords})
4636 @node Output Section Name
4637 @subsection Output Section Name
4638 @cindex name, section
4639 @cindex section name
4640 The name of the output section is @var{section}. @var{section} must
4641 meet the constraints of your output format. In formats which only
4642 support a limited number of sections, such as @code{a.out}, the name
4643 must be one of the names supported by the format (@code{a.out}, for
4644 example, allows only @samp{.text}, @samp{.data} or @samp{.bss}). If the
4645 output format supports any number of sections, but with numbers and not
4646 names (as is the case for Oasys), the name should be supplied as a
4647 quoted numeric string. A section name may consist of any sequence of
4648 characters, but a name which contains any unusual characters such as
4649 commas must be quoted.
4651 The output section name @samp{/DISCARD/} is special; @ref{Output Section
4654 @node Output Section Address
4655 @subsection Output Section Address
4656 @cindex address, section
4657 @cindex section address
4658 The @var{address} is an expression for the VMA (the virtual memory
4659 address) of the output section. This address is optional, but if it
4660 is provided then the output address will be set exactly as specified.
4662 If the output address is not specified then one will be chosen for the
4663 section, based on the heuristic below. This address will be adjusted
4664 to fit the alignment requirement of the output section. The
4665 alignment requirement is the strictest alignment of any input section
4666 contained within the output section.
4668 The output section address heuristic is as follows:
4672 If an output memory @var{region} is set for the section then it
4673 is added to this region and its address will be the next free address
4677 If the MEMORY command has been used to create a list of memory
4678 regions then the first region which has attributes compatible with the
4679 section is selected to contain it. The section's output address will
4680 be the next free address in that region; @ref{MEMORY}.
4683 If no memory regions were specified, or none match the section then
4684 the output address will be based on the current value of the location
4692 .text . : @{ *(.text) @}
4699 .text : @{ *(.text) @}
4703 are subtly different. The first will set the address of the
4704 @samp{.text} output section to the current value of the location
4705 counter. The second will set it to the current value of the location
4706 counter aligned to the strictest alignment of any of the @samp{.text}
4709 The @var{address} may be an arbitrary expression; @ref{Expressions}.
4710 For example, if you want to align the section on a 0x10 byte boundary,
4711 so that the lowest four bits of the section address are zero, you could
4712 do something like this:
4714 .text ALIGN(0x10) : @{ *(.text) @}
4717 This works because @code{ALIGN} returns the current location counter
4718 aligned upward to the specified value.
4720 Specifying @var{address} for a section will change the value of the
4721 location counter, provided that the section is non-empty. (Empty
4722 sections are ignored).
4725 @subsection Input Section Description
4726 @cindex input sections
4727 @cindex mapping input sections to output sections
4728 The most common output section command is an input section description.
4730 The input section description is the most basic linker script operation.
4731 You use output sections to tell the linker how to lay out your program
4732 in memory. You use input section descriptions to tell the linker how to
4733 map the input files into your memory layout.
4736 * Input Section Basics:: Input section basics
4737 * Input Section Wildcards:: Input section wildcard patterns
4738 * Input Section Common:: Input section for common symbols
4739 * Input Section Keep:: Input section and garbage collection
4740 * Input Section Example:: Input section example
4743 @node Input Section Basics
4744 @subsubsection Input Section Basics
4745 @cindex input section basics
4746 An input section description consists of a file name optionally followed
4747 by a list of section names in parentheses.
4749 The file name and the section name may be wildcard patterns, which we
4750 describe further below (@pxref{Input Section Wildcards}).
4752 The most common input section description is to include all input
4753 sections with a particular name in the output section. For example, to
4754 include all input @samp{.text} sections, you would write:
4759 Here the @samp{*} is a wildcard which matches any file name. To exclude a list
4760 @cindex EXCLUDE_FILE
4761 of files from matching the file name wildcard, EXCLUDE_FILE may be used to
4762 match all files except the ones specified in the EXCLUDE_FILE list. For
4765 EXCLUDE_FILE (*crtend.o *otherfile.o) *(.ctors)
4768 will cause all .ctors sections from all files except @file{crtend.o}
4769 and @file{otherfile.o} to be included. The EXCLUDE_FILE can also be
4770 placed inside the section list, for example:
4772 *(EXCLUDE_FILE (*crtend.o *otherfile.o) .ctors)
4775 The result of this is identically to the previous example. Supporting
4776 two syntaxes for EXCLUDE_FILE is useful if the section list contains
4777 more than one section, as described below.
4779 There are two ways to include more than one section:
4785 The difference between these is the order in which the @samp{.text} and
4786 @samp{.rdata} input sections will appear in the output section. In the
4787 first example, they will be intermingled, appearing in the same order as
4788 they are found in the linker input. In the second example, all
4789 @samp{.text} input sections will appear first, followed by all
4790 @samp{.rdata} input sections.
4792 When using EXCLUDE_FILE with more than one section, if the exclusion
4793 is within the section list then the exclusion only applies to the
4794 immediately following section, for example:
4796 *(EXCLUDE_FILE (*somefile.o) .text .rdata)
4799 will cause all @samp{.text} sections from all files except
4800 @file{somefile.o} to be included, while all @samp{.rdata} sections
4801 from all files, including @file{somefile.o}, will be included. To
4802 exclude the @samp{.rdata} sections from @file{somefile.o} the example
4803 could be modified to:
4805 *(EXCLUDE_FILE (*somefile.o) .text EXCLUDE_FILE (*somefile.o) .rdata)
4808 Alternatively, placing the EXCLUDE_FILE outside of the section list,
4809 before the input file selection, will cause the exclusion to apply for
4810 all sections. Thus the previous example can be rewritten as:
4812 EXCLUDE_FILE (*somefile.o) *(.text .rdata)
4815 You can specify a file name to include sections from a particular file.
4816 You would do this if one or more of your files contain special data that
4817 needs to be at a particular location in memory. For example:
4822 To refine the sections that are included based on the section flags
4823 of an input section, INPUT_SECTION_FLAGS may be used.
4825 Here is a simple example for using Section header flags for ELF sections:
4830 .text : @{ INPUT_SECTION_FLAGS (SHF_MERGE & SHF_STRINGS) *(.text) @}
4831 .text2 : @{ INPUT_SECTION_FLAGS (!SHF_WRITE) *(.text) @}
4836 In this example, the output section @samp{.text} will be comprised of any
4837 input section matching the name *(.text) whose section header flags
4838 @code{SHF_MERGE} and @code{SHF_STRINGS} are set. The output section
4839 @samp{.text2} will be comprised of any input section matching the name *(.text)
4840 whose section header flag @code{SHF_WRITE} is clear.
4842 You can also specify files within archives by writing a pattern
4843 matching the archive, a colon, then the pattern matching the file,
4844 with no whitespace around the colon.
4848 matches file within archive
4850 matches the whole archive
4852 matches file but not one in an archive
4855 Either one or both of @samp{archive} and @samp{file} can contain shell
4856 wildcards. On DOS based file systems, the linker will assume that a
4857 single letter followed by a colon is a drive specifier, so
4858 @samp{c:myfile.o} is a simple file specification, not @samp{myfile.o}
4859 within an archive called @samp{c}. @samp{archive:file} filespecs may
4860 also be used within an @code{EXCLUDE_FILE} list, but may not appear in
4861 other linker script contexts. For instance, you cannot extract a file
4862 from an archive by using @samp{archive:file} in an @code{INPUT}
4865 If you use a file name without a list of sections, then all sections in
4866 the input file will be included in the output section. This is not
4867 commonly done, but it may by useful on occasion. For example:
4872 When you use a file name which is not an @samp{archive:file} specifier
4873 and does not contain any wild card
4874 characters, the linker will first see if you also specified the file
4875 name on the linker command line or in an @code{INPUT} command. If you
4876 did not, the linker will attempt to open the file as an input file, as
4877 though it appeared on the command line. Note that this differs from an
4878 @code{INPUT} command, because the linker will not search for the file in
4879 the archive search path.
4881 @node Input Section Wildcards
4882 @subsubsection Input Section Wildcard Patterns
4883 @cindex input section wildcards
4884 @cindex wildcard file name patterns
4885 @cindex file name wildcard patterns
4886 @cindex section name wildcard patterns
4887 In an input section description, either the file name or the section
4888 name or both may be wildcard patterns.
4890 The file name of @samp{*} seen in many examples is a simple wildcard
4891 pattern for the file name.
4893 The wildcard patterns are like those used by the Unix shell.
4897 matches any number of characters
4899 matches any single character
4901 matches a single instance of any of the @var{chars}; the @samp{-}
4902 character may be used to specify a range of characters, as in
4903 @samp{[a-z]} to match any lower case letter
4905 quotes the following character
4908 When a file name is matched with a wildcard, the wildcard characters
4909 will not match a @samp{/} character (used to separate directory names on
4910 Unix). A pattern consisting of a single @samp{*} character is an
4911 exception; it will always match any file name, whether it contains a
4912 @samp{/} or not. In a section name, the wildcard characters will match
4913 a @samp{/} character.
4915 File name wildcard patterns only match files which are explicitly
4916 specified on the command line or in an @code{INPUT} command. The linker
4917 does not search directories to expand wildcards.
4919 If a file name matches more than one wildcard pattern, or if a file name
4920 appears explicitly and is also matched by a wildcard pattern, the linker
4921 will use the first match in the linker script. For example, this
4922 sequence of input section descriptions is probably in error, because the
4923 @file{data.o} rule will not be used:
4925 .data : @{ *(.data) @}
4926 .data1 : @{ data.o(.data) @}
4929 @cindex SORT_BY_NAME
4930 Normally, the linker will place files and sections matched by wildcards
4931 in the order in which they are seen during the link. You can change
4932 this by using the @code{SORT_BY_NAME} keyword, which appears before a wildcard
4933 pattern in parentheses (e.g., @code{SORT_BY_NAME(.text*)}). When the
4934 @code{SORT_BY_NAME} keyword is used, the linker will sort the files or sections
4935 into ascending order by name before placing them in the output file.
4937 @cindex SORT_BY_ALIGNMENT
4938 @code{SORT_BY_ALIGNMENT} is similar to @code{SORT_BY_NAME}.
4939 @code{SORT_BY_ALIGNMENT} will sort sections into descending order of
4940 alignment before placing them in the output file. Placing larger
4941 alignments before smaller alignments can reduce the amount of padding
4944 @cindex SORT_BY_INIT_PRIORITY
4945 @code{SORT_BY_INIT_PRIORITY} is also similar to @code{SORT_BY_NAME}.
4946 @code{SORT_BY_INIT_PRIORITY} will sort sections into ascending
4947 numerical order of the GCC init_priority attribute encoded in the
4948 section name before placing them in the output file. In
4949 @code{.init_array.NNNNN} and @code{.fini_array.NNNNN}, @code{NNNNN} is
4950 the init_priority. In @code{.ctors.NNNNN} and @code{.dtors.NNNNN},
4951 @code{NNNNN} is 65535 minus the init_priority.
4954 @code{SORT} is an alias for @code{SORT_BY_NAME}.
4956 When there are nested section sorting commands in linker script, there
4957 can be at most 1 level of nesting for section sorting commands.
4961 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
4962 It will sort the input sections by name first, then by alignment if two
4963 sections have the same name.
4965 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
4966 It will sort the input sections by alignment first, then by name if two
4967 sections have the same alignment.
4969 @code{SORT_BY_NAME} (@code{SORT_BY_NAME} (wildcard section pattern)) is
4970 treated the same as @code{SORT_BY_NAME} (wildcard section pattern).
4972 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern))
4973 is treated the same as @code{SORT_BY_ALIGNMENT} (wildcard section pattern).
4975 All other nested section sorting commands are invalid.
4978 When both command-line section sorting option and linker script
4979 section sorting command are used, section sorting command always
4980 takes precedence over the command-line option.
4982 If the section sorting command in linker script isn't nested, the
4983 command-line option will make the section sorting command to be
4984 treated as nested sorting command.
4988 @code{SORT_BY_NAME} (wildcard section pattern ) with
4989 @option{--sort-sections alignment} is equivalent to
4990 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
4992 @code{SORT_BY_ALIGNMENT} (wildcard section pattern) with
4993 @option{--sort-section name} is equivalent to
4994 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
4997 If the section sorting command in linker script is nested, the
4998 command-line option will be ignored.
5001 @code{SORT_NONE} disables section sorting by ignoring the command-line
5002 section sorting option.
5004 If you ever get confused about where input sections are going, use the
5005 @samp{-M} linker option to generate a map file. The map file shows
5006 precisely how input sections are mapped to output sections.
5008 This example shows how wildcard patterns might be used to partition
5009 files. This linker script directs the linker to place all @samp{.text}
5010 sections in @samp{.text} and all @samp{.bss} sections in @samp{.bss}.
5011 The linker will place the @samp{.data} section from all files beginning
5012 with an upper case character in @samp{.DATA}; for all other files, the
5013 linker will place the @samp{.data} section in @samp{.data}.
5017 .text : @{ *(.text) @}
5018 .DATA : @{ [A-Z]*(.data) @}
5019 .data : @{ *(.data) @}
5020 .bss : @{ *(.bss) @}
5025 @node Input Section Common
5026 @subsubsection Input Section for Common Symbols
5027 @cindex common symbol placement
5028 @cindex uninitialized data placement
5029 A special notation is needed for common symbols, because in many object
5030 file formats common symbols do not have a particular input section. The
5031 linker treats common symbols as though they are in an input section
5032 named @samp{COMMON}.
5034 You may use file names with the @samp{COMMON} section just as with any
5035 other input sections. You can use this to place common symbols from a
5036 particular input file in one section while common symbols from other
5037 input files are placed in another section.
5039 In most cases, common symbols in input files will be placed in the
5040 @samp{.bss} section in the output file. For example:
5042 .bss @{ *(.bss) *(COMMON) @}
5045 @cindex scommon section
5046 @cindex small common symbols
5047 Some object file formats have more than one type of common symbol. For
5048 example, the MIPS ELF object file format distinguishes standard common
5049 symbols and small common symbols. In this case, the linker will use a
5050 different special section name for other types of common symbols. In
5051 the case of MIPS ELF, the linker uses @samp{COMMON} for standard common
5052 symbols and @samp{.scommon} for small common symbols. This permits you
5053 to map the different types of common symbols into memory at different
5057 You will sometimes see @samp{[COMMON]} in old linker scripts. This
5058 notation is now considered obsolete. It is equivalent to
5061 @node Input Section Keep
5062 @subsubsection Input Section and Garbage Collection
5064 @cindex garbage collection
5065 When link-time garbage collection is in use (@samp{--gc-sections}),
5066 it is often useful to mark sections that should not be eliminated.
5067 This is accomplished by surrounding an input section's wildcard entry
5068 with @code{KEEP()}, as in @code{KEEP(*(.init))} or
5069 @code{KEEP(SORT_BY_NAME(*)(.ctors))}.
5071 @node Input Section Example
5072 @subsubsection Input Section Example
5073 The following example is a complete linker script. It tells the linker
5074 to read all of the sections from file @file{all.o} and place them at the
5075 start of output section @samp{outputa} which starts at location
5076 @samp{0x10000}. All of section @samp{.input1} from file @file{foo.o}
5077 follows immediately, in the same output section. All of section
5078 @samp{.input2} from @file{foo.o} goes into output section
5079 @samp{outputb}, followed by section @samp{.input1} from @file{foo1.o}.
5080 All of the remaining @samp{.input1} and @samp{.input2} sections from any
5081 files are written to output section @samp{outputc}.
5109 If an output section's name is the same as the input section's name
5110 and is representable as a C identifier, then the linker will
5111 automatically @pxref{PROVIDE} two symbols: __start_SECNAME and
5112 __stop_SECNAME, where SECNAME is the name of the section. These
5113 indicate the start address and end address of the output section
5114 respectively. Note: most section names are not representable as
5115 C identifiers because they contain a @samp{.} character.
5117 @node Output Section Data
5118 @subsection Output Section Data
5120 @cindex section data
5121 @cindex output section data
5122 @kindex BYTE(@var{expression})
5123 @kindex SHORT(@var{expression})
5124 @kindex LONG(@var{expression})
5125 @kindex QUAD(@var{expression})
5126 @kindex SQUAD(@var{expression})
5127 You can include explicit bytes of data in an output section by using
5128 @code{BYTE}, @code{SHORT}, @code{LONG}, @code{QUAD}, or @code{SQUAD} as
5129 an output section command. Each keyword is followed by an expression in
5130 parentheses providing the value to store (@pxref{Expressions}). The
5131 value of the expression is stored at the current value of the location
5134 The @code{BYTE}, @code{SHORT}, @code{LONG}, and @code{QUAD} commands
5135 store one, two, four, and eight bytes (respectively). After storing the
5136 bytes, the location counter is incremented by the number of bytes
5139 For example, this will store the byte 1 followed by the four byte value
5140 of the symbol @samp{addr}:
5146 When using a 64 bit host or target, @code{QUAD} and @code{SQUAD} are the
5147 same; they both store an 8 byte, or 64 bit, value. When both host and
5148 target are 32 bits, an expression is computed as 32 bits. In this case
5149 @code{QUAD} stores a 32 bit value zero extended to 64 bits, and
5150 @code{SQUAD} stores a 32 bit value sign extended to 64 bits.
5152 If the object file format of the output file has an explicit endianness,
5153 which is the normal case, the value will be stored in that endianness.
5154 When the object file format does not have an explicit endianness, as is
5155 true of, for example, S-records, the value will be stored in the
5156 endianness of the first input object file.
5158 Note---these commands only work inside a section description and not
5159 between them, so the following will produce an error from the linker:
5161 SECTIONS @{@ .text : @{@ *(.text) @}@ LONG(1) .data : @{@ *(.data) @}@ @}@
5163 whereas this will work:
5165 SECTIONS @{@ .text : @{@ *(.text) ; LONG(1) @}@ .data : @{@ *(.data) @}@ @}@
5168 @kindex FILL(@var{expression})
5169 @cindex holes, filling
5170 @cindex unspecified memory
5171 You may use the @code{FILL} command to set the fill pattern for the
5172 current section. It is followed by an expression in parentheses. Any
5173 otherwise unspecified regions of memory within the section (for example,
5174 gaps left due to the required alignment of input sections) are filled
5175 with the value of the expression, repeated as
5176 necessary. A @code{FILL} statement covers memory locations after the
5177 point at which it occurs in the section definition; by including more
5178 than one @code{FILL} statement, you can have different fill patterns in
5179 different parts of an output section.
5181 This example shows how to fill unspecified regions of memory with the
5187 The @code{FILL} command is similar to the @samp{=@var{fillexp}} output
5188 section attribute, but it only affects the
5189 part of the section following the @code{FILL} command, rather than the
5190 entire section. If both are used, the @code{FILL} command takes
5191 precedence. @xref{Output Section Fill}, for details on the fill
5194 @node Output Section Keywords
5195 @subsection Output Section Keywords
5196 There are a couple of keywords which can appear as output section
5200 @kindex CREATE_OBJECT_SYMBOLS
5201 @cindex input filename symbols
5202 @cindex filename symbols
5203 @item CREATE_OBJECT_SYMBOLS
5204 The command tells the linker to create a symbol for each input file.
5205 The name of each symbol will be the name of the corresponding input
5206 file. The section of each symbol will be the output section in which
5207 the @code{CREATE_OBJECT_SYMBOLS} command appears.
5209 This is conventional for the a.out object file format. It is not
5210 normally used for any other object file format.
5212 @kindex CONSTRUCTORS
5213 @cindex C++ constructors, arranging in link
5214 @cindex constructors, arranging in link
5216 When linking using the a.out object file format, the linker uses an
5217 unusual set construct to support C++ global constructors and
5218 destructors. When linking object file formats which do not support
5219 arbitrary sections, such as ECOFF and XCOFF, the linker will
5220 automatically recognize C++ global constructors and destructors by name.
5221 For these object file formats, the @code{CONSTRUCTORS} command tells the
5222 linker to place constructor information in the output section where the
5223 @code{CONSTRUCTORS} command appears. The @code{CONSTRUCTORS} command is
5224 ignored for other object file formats.
5226 The symbol @w{@code{__CTOR_LIST__}} marks the start of the global
5227 constructors, and the symbol @w{@code{__CTOR_END__}} marks the end.
5228 Similarly, @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_END__}} mark
5229 the start and end of the global destructors. The
5230 first word in the list is the number of entries, followed by the address
5231 of each constructor or destructor, followed by a zero word. The
5232 compiler must arrange to actually run the code. For these object file
5233 formats @sc{gnu} C++ normally calls constructors from a subroutine
5234 @code{__main}; a call to @code{__main} is automatically inserted into
5235 the startup code for @code{main}. @sc{gnu} C++ normally runs
5236 destructors either by using @code{atexit}, or directly from the function
5239 For object file formats such as @code{COFF} or @code{ELF} which support
5240 arbitrary section names, @sc{gnu} C++ will normally arrange to put the
5241 addresses of global constructors and destructors into the @code{.ctors}
5242 and @code{.dtors} sections. Placing the following sequence into your
5243 linker script will build the sort of table which the @sc{gnu} C++
5244 runtime code expects to see.
5248 LONG((__CTOR_END__ - __CTOR_LIST__) / 4 - 2)
5253 LONG((__DTOR_END__ - __DTOR_LIST__) / 4 - 2)
5259 If you are using the @sc{gnu} C++ support for initialization priority,
5260 which provides some control over the order in which global constructors
5261 are run, you must sort the constructors at link time to ensure that they
5262 are executed in the correct order. When using the @code{CONSTRUCTORS}
5263 command, use @samp{SORT_BY_NAME(CONSTRUCTORS)} instead. When using the
5264 @code{.ctors} and @code{.dtors} sections, use @samp{*(SORT_BY_NAME(.ctors))} and
5265 @samp{*(SORT_BY_NAME(.dtors))} instead of just @samp{*(.ctors)} and
5268 Normally the compiler and linker will handle these issues automatically,
5269 and you will not need to concern yourself with them. However, you may
5270 need to consider this if you are using C++ and writing your own linker
5275 @node Output Section Discarding
5276 @subsection Output Section Discarding
5277 @cindex discarding sections
5278 @cindex sections, discarding
5279 @cindex removing sections
5280 The linker will not normally create output sections with no contents.
5281 This is for convenience when referring to input sections that may or
5282 may not be present in any of the input files. For example:
5284 .foo : @{ *(.foo) @}
5287 will only create a @samp{.foo} section in the output file if there is a
5288 @samp{.foo} section in at least one input file, and if the input
5289 sections are not all empty. Other link script directives that allocate
5290 space in an output section will also create the output section. So
5291 too will assignments to dot even if the assignment does not create
5292 space, except for @samp{. = 0}, @samp{. = . + 0}, @samp{. = sym},
5293 @samp{. = . + sym} and @samp{. = ALIGN (. != 0, expr, 1)} when
5294 @samp{sym} is an absolute symbol of value 0 defined in the script.
5295 This allows you to force output of an empty section with @samp{. = .}.
5297 The linker will ignore address assignments (@pxref{Output Section Address})
5298 on discarded output sections, except when the linker script defines
5299 symbols in the output section. In that case the linker will obey
5300 the address assignments, possibly advancing dot even though the
5301 section is discarded.
5304 The special output section name @samp{/DISCARD/} may be used to discard
5305 input sections. Any input sections which are assigned to an output
5306 section named @samp{/DISCARD/} are not included in the output file.
5308 This can be used to discard input sections marked with the ELF flag
5309 @code{SHF_GNU_RETAIN}, which would otherwise have been saved from linker
5312 Note, sections that match the @samp{/DISCARD/} output section will be
5313 discarded even if they are in an ELF section group which has other
5314 members which are not being discarded. This is deliberate.
5315 Discarding takes precedence over grouping.
5317 @node Output Section Attributes
5318 @subsection Output Section Attributes
5319 @cindex output section attributes
5320 We showed above that the full description of an output section looked
5325 @var{section} [@var{address}] [(@var{type})] :
5327 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
5328 [SUBALIGN(@var{subsection_align})]
5331 @var{output-section-command}
5332 @var{output-section-command}
5334 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}]
5338 We've already described @var{section}, @var{address}, and
5339 @var{output-section-command}. In this section we will describe the
5340 remaining section attributes.
5343 * Output Section Type:: Output section type
5344 * Output Section LMA:: Output section LMA
5345 * Forced Output Alignment:: Forced Output Alignment
5346 * Forced Input Alignment:: Forced Input Alignment
5347 * Output Section Constraint:: Output section constraint
5348 * Output Section Region:: Output section region
5349 * Output Section Phdr:: Output section phdr
5350 * Output Section Fill:: Output section fill
5353 @node Output Section Type
5354 @subsubsection Output Section Type
5355 Each output section may have a type. The type is a keyword in
5356 parentheses. The following types are defined:
5360 The section should be marked as not loadable, so that it will not be
5361 loaded into memory when the program is run.
5366 These type names are supported for backward compatibility, and are
5367 rarely used. They all have the same effect: the section should be
5368 marked as not allocatable, so that no memory is allocated for the
5369 section when the program is run.
5373 @cindex prevent unnecessary loading
5374 @cindex loading, preventing
5375 The linker normally sets the attributes of an output section based on
5376 the input sections which map into it. You can override this by using
5377 the section type. For example, in the script sample below, the
5378 @samp{ROM} section is addressed at memory location @samp{0} and does not
5379 need to be loaded when the program is run.
5383 ROM 0 (NOLOAD) : @{ @dots{} @}
5389 @node Output Section LMA
5390 @subsubsection Output Section LMA
5391 @kindex AT>@var{lma_region}
5392 @kindex AT(@var{lma})
5393 @cindex load address
5394 @cindex section load address
5395 Every section has a virtual address (VMA) and a load address (LMA); see
5396 @ref{Basic Script Concepts}. The virtual address is specified by the
5397 @pxref{Output Section Address} described earlier. The load address is
5398 specified by the @code{AT} or @code{AT>} keywords. Specifying a load
5399 address is optional.
5401 The @code{AT} keyword takes an expression as an argument. This
5402 specifies the exact load address of the section. The @code{AT>} keyword
5403 takes the name of a memory region as an argument. @xref{MEMORY}. The
5404 load address of the section is set to the next free address in the
5405 region, aligned to the section's alignment requirements.
5407 If neither @code{AT} nor @code{AT>} is specified for an allocatable
5408 section, the linker will use the following heuristic to determine the
5413 If the section has a specific VMA address, then this is used as
5414 the LMA address as well.
5417 If the section is not allocatable then its LMA is set to its VMA.
5420 Otherwise if a memory region can be found that is compatible
5421 with the current section, and this region contains at least one
5422 section, then the LMA is set so the difference between the
5423 VMA and LMA is the same as the difference between the VMA and LMA of
5424 the last section in the located region.
5427 If no memory regions have been declared then a default region
5428 that covers the entire address space is used in the previous step.
5431 If no suitable region could be found, or there was no previous
5432 section then the LMA is set equal to the VMA.
5435 @cindex ROM initialized data
5436 @cindex initialized data in ROM
5437 This feature is designed to make it easy to build a ROM image. For
5438 example, the following linker script creates three output sections: one
5439 called @samp{.text}, which starts at @code{0x1000}, one called
5440 @samp{.mdata}, which is loaded at the end of the @samp{.text} section
5441 even though its VMA is @code{0x2000}, and one called @samp{.bss} to hold
5442 uninitialized data at address @code{0x3000}. The symbol @code{_data} is
5443 defined with the value @code{0x2000}, which shows that the location
5444 counter holds the VMA value, not the LMA value.
5450 .text 0x1000 : @{ *(.text) _etext = . ; @}
5452 AT ( ADDR (.text) + SIZEOF (.text) )
5453 @{ _data = . ; *(.data); _edata = . ; @}
5455 @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@}
5460 The run-time initialization code for use with a program generated with
5461 this linker script would include something like the following, to copy
5462 the initialized data from the ROM image to its runtime address. Notice
5463 how this code takes advantage of the symbols defined by the linker
5468 extern char _etext, _data, _edata, _bstart, _bend;
5469 char *src = &_etext;
5472 /* ROM has data at end of text; copy it. */
5473 while (dst < &_edata)
5477 for (dst = &_bstart; dst< &_bend; dst++)
5482 @node Forced Output Alignment
5483 @subsubsection Forced Output Alignment
5484 @kindex ALIGN(@var{section_align})
5485 @cindex forcing output section alignment
5486 @cindex output section alignment
5487 You can increase an output section's alignment by using ALIGN. As an
5488 alternative you can enforce that the difference between the VMA and LMA remains
5489 intact throughout this output section with the ALIGN_WITH_INPUT attribute.
5491 @node Forced Input Alignment
5492 @subsubsection Forced Input Alignment
5493 @kindex SUBALIGN(@var{subsection_align})
5494 @cindex forcing input section alignment
5495 @cindex input section alignment
5496 You can force input section alignment within an output section by using
5497 SUBALIGN. The value specified overrides any alignment given by input
5498 sections, whether larger or smaller.
5500 @node Output Section Constraint
5501 @subsubsection Output Section Constraint
5504 @cindex constraints on output sections
5505 You can specify that an output section should only be created if all
5506 of its input sections are read-only or all of its input sections are
5507 read-write by using the keyword @code{ONLY_IF_RO} and
5508 @code{ONLY_IF_RW} respectively.
5510 @node Output Section Region
5511 @subsubsection Output Section Region
5512 @kindex >@var{region}
5513 @cindex section, assigning to memory region
5514 @cindex memory regions and sections
5515 You can assign a section to a previously defined region of memory by
5516 using @samp{>@var{region}}. @xref{MEMORY}.
5518 Here is a simple example:
5521 MEMORY @{ rom : ORIGIN = 0x1000, LENGTH = 0x1000 @}
5522 SECTIONS @{ ROM : @{ *(.text) @} >rom @}
5526 @node Output Section Phdr
5527 @subsubsection Output Section Phdr
5529 @cindex section, assigning to program header
5530 @cindex program headers and sections
5531 You can assign a section to a previously defined program segment by
5532 using @samp{:@var{phdr}}. @xref{PHDRS}. If a section is assigned to
5533 one or more segments, then all subsequent allocated sections will be
5534 assigned to those segments as well, unless they use an explicitly
5535 @code{:@var{phdr}} modifier. You can use @code{:NONE} to tell the
5536 linker to not put the section in any segment at all.
5538 Here is a simple example:
5541 PHDRS @{ text PT_LOAD ; @}
5542 SECTIONS @{ .text : @{ *(.text) @} :text @}
5546 @node Output Section Fill
5547 @subsubsection Output Section Fill
5548 @kindex =@var{fillexp}
5549 @cindex section fill pattern
5550 @cindex fill pattern, entire section
5551 You can set the fill pattern for an entire section by using
5552 @samp{=@var{fillexp}}. @var{fillexp} is an expression
5553 (@pxref{Expressions}). Any otherwise unspecified regions of memory
5554 within the output section (for example, gaps left due to the required
5555 alignment of input sections) will be filled with the value, repeated as
5556 necessary. If the fill expression is a simple hex number, ie. a string
5557 of hex digit starting with @samp{0x} and without a trailing @samp{k} or @samp{M}, then
5558 an arbitrarily long sequence of hex digits can be used to specify the
5559 fill pattern; Leading zeros become part of the pattern too. For all
5560 other cases, including extra parentheses or a unary @code{+}, the fill
5561 pattern is the four least significant bytes of the value of the
5562 expression. In all cases, the number is big-endian.
5564 You can also change the fill value with a @code{FILL} command in the
5565 output section commands; (@pxref{Output Section Data}).
5567 Here is a simple example:
5570 SECTIONS @{ .text : @{ *(.text) @} =0x90909090 @}
5574 @node Overlay Description
5575 @subsection Overlay Description
5578 An overlay description provides an easy way to describe sections which
5579 are to be loaded as part of a single memory image but are to be run at
5580 the same memory address. At run time, some sort of overlay manager will
5581 copy the overlaid sections in and out of the runtime memory address as
5582 required, perhaps by simply manipulating addressing bits. This approach
5583 can be useful, for example, when a certain region of memory is faster
5586 Overlays are described using the @code{OVERLAY} command. The
5587 @code{OVERLAY} command is used within a @code{SECTIONS} command, like an
5588 output section description. The full syntax of the @code{OVERLAY}
5589 command is as follows:
5592 OVERLAY [@var{start}] : [NOCROSSREFS] [AT ( @var{ldaddr} )]
5596 @var{output-section-command}
5597 @var{output-section-command}
5599 @} [:@var{phdr}@dots{}] [=@var{fill}]
5602 @var{output-section-command}
5603 @var{output-section-command}
5605 @} [:@var{phdr}@dots{}] [=@var{fill}]
5607 @} [>@var{region}] [:@var{phdr}@dots{}] [=@var{fill}] [,]
5611 Everything is optional except @code{OVERLAY} (a keyword), and each
5612 section must have a name (@var{secname1} and @var{secname2} above). The
5613 section definitions within the @code{OVERLAY} construct are identical to
5614 those within the general @code{SECTIONS} construct (@pxref{SECTIONS}),
5615 except that no addresses and no memory regions may be defined for
5616 sections within an @code{OVERLAY}.
5618 The comma at the end may be required if a @var{fill} is used and
5619 the next @var{sections-command} looks like a continuation of the expression.
5621 The sections are all defined with the same starting address. The load
5622 addresses of the sections are arranged such that they are consecutive in
5623 memory starting at the load address used for the @code{OVERLAY} as a
5624 whole (as with normal section definitions, the load address is optional,
5625 and defaults to the start address; the start address is also optional,
5626 and defaults to the current value of the location counter).
5628 If the @code{NOCROSSREFS} keyword is used, and there are any
5629 references among the sections, the linker will report an error. Since
5630 the sections all run at the same address, it normally does not make
5631 sense for one section to refer directly to another.
5632 @xref{Miscellaneous Commands, NOCROSSREFS}.
5634 For each section within the @code{OVERLAY}, the linker automatically
5635 provides two symbols. The symbol @code{__load_start_@var{secname}} is
5636 defined as the starting load address of the section. The symbol
5637 @code{__load_stop_@var{secname}} is defined as the final load address of
5638 the section. Any characters within @var{secname} which are not legal
5639 within C identifiers are removed. C (or assembler) code may use these
5640 symbols to move the overlaid sections around as necessary.
5642 At the end of the overlay, the value of the location counter is set to
5643 the start address of the overlay plus the size of the largest section.
5645 Here is an example. Remember that this would appear inside a
5646 @code{SECTIONS} construct.
5649 OVERLAY 0x1000 : AT (0x4000)
5651 .text0 @{ o1/*.o(.text) @}
5652 .text1 @{ o2/*.o(.text) @}
5657 This will define both @samp{.text0} and @samp{.text1} to start at
5658 address 0x1000. @samp{.text0} will be loaded at address 0x4000, and
5659 @samp{.text1} will be loaded immediately after @samp{.text0}. The
5660 following symbols will be defined if referenced: @code{__load_start_text0},
5661 @code{__load_stop_text0}, @code{__load_start_text1},
5662 @code{__load_stop_text1}.
5664 C code to copy overlay @code{.text1} into the overlay area might look
5669 extern char __load_start_text1, __load_stop_text1;
5670 memcpy ((char *) 0x1000, &__load_start_text1,
5671 &__load_stop_text1 - &__load_start_text1);
5675 Note that the @code{OVERLAY} command is just syntactic sugar, since
5676 everything it does can be done using the more basic commands. The above
5677 example could have been written identically as follows.
5681 .text0 0x1000 : AT (0x4000) @{ o1/*.o(.text) @}
5682 PROVIDE (__load_start_text0 = LOADADDR (.text0));
5683 PROVIDE (__load_stop_text0 = LOADADDR (.text0) + SIZEOF (.text0));
5684 .text1 0x1000 : AT (0x4000 + SIZEOF (.text0)) @{ o2/*.o(.text) @}
5685 PROVIDE (__load_start_text1 = LOADADDR (.text1));
5686 PROVIDE (__load_stop_text1 = LOADADDR (.text1) + SIZEOF (.text1));
5687 . = 0x1000 + MAX (SIZEOF (.text0), SIZEOF (.text1));
5692 @section MEMORY Command
5694 @cindex memory regions
5695 @cindex regions of memory
5696 @cindex allocating memory
5697 @cindex discontinuous memory
5698 The linker's default configuration permits allocation of all available
5699 memory. You can override this by using the @code{MEMORY} command.
5701 The @code{MEMORY} command describes the location and size of blocks of
5702 memory in the target. You can use it to describe which memory regions
5703 may be used by the linker, and which memory regions it must avoid. You
5704 can then assign sections to particular memory regions. The linker will
5705 set section addresses based on the memory regions, and will warn about
5706 regions that become too full. The linker will not shuffle sections
5707 around to fit into the available regions.
5709 A linker script may contain many uses of the @code{MEMORY} command,
5710 however, all memory blocks defined are treated as if they were
5711 specified inside a single @code{MEMORY} command. The syntax for
5717 @var{name} [(@var{attr})] : ORIGIN = @var{origin}, LENGTH = @var{len}
5723 The @var{name} is a name used in the linker script to refer to the
5724 region. The region name has no meaning outside of the linker script.
5725 Region names are stored in a separate name space, and will not conflict
5726 with symbol names, file names, or section names. Each memory region
5727 must have a distinct name within the @code{MEMORY} command. However you can
5728 add later alias names to existing memory regions with the @ref{REGION_ALIAS}
5731 @cindex memory region attributes
5732 The @var{attr} string is an optional list of attributes that specify
5733 whether to use a particular memory region for an input section which is
5734 not explicitly mapped in the linker script. As described in
5735 @ref{SECTIONS}, if you do not specify an output section for some input
5736 section, the linker will create an output section with the same name as
5737 the input section. If you define region attributes, the linker will use
5738 them to select the memory region for the output section that it creates.
5740 The @var{attr} string must consist only of the following characters:
5755 Invert the sense of any of the attributes that follow
5758 If an unmapped section matches any of the listed attributes other than
5759 @samp{!}, it will be placed in the memory region. The @samp{!}
5760 attribute reverses the test for the characters that follow, so that an
5761 unmapped section will be placed in the memory region only if it does
5762 not match any of the attributes listed afterwards. Thus an attribute
5763 string of @samp{RW!X} will match any unmapped section that has either
5764 or both of the @samp{R} and @samp{W} attributes, but only as long as
5765 the section does not also have the @samp{X} attribute.
5770 The @var{origin} is an numerical expression for the start address of
5771 the memory region. The expression must evaluate to a constant and it
5772 cannot involve any symbols. The keyword @code{ORIGIN} may be
5773 abbreviated to @code{org} or @code{o} (but not, for example,
5779 The @var{len} is an expression for the size in bytes of the memory
5780 region. As with the @var{origin} expression, the expression must
5781 be numerical only and must evaluate to a constant. The keyword
5782 @code{LENGTH} may be abbreviated to @code{len} or @code{l}.
5784 In the following example, we specify that there are two memory regions
5785 available for allocation: one starting at @samp{0} for 256 kilobytes,
5786 and the other starting at @samp{0x40000000} for four megabytes. The
5787 linker will place into the @samp{rom} memory region every section which
5788 is not explicitly mapped into a memory region, and is either read-only
5789 or executable. The linker will place other sections which are not
5790 explicitly mapped into a memory region into the @samp{ram} memory
5797 rom (rx) : ORIGIN = 0, LENGTH = 256K
5798 ram (!rx) : org = 0x40000000, l = 4M
5803 Once you define a memory region, you can direct the linker to place
5804 specific output sections into that memory region by using the
5805 @samp{>@var{region}} output section attribute. For example, if you have
5806 a memory region named @samp{mem}, you would use @samp{>mem} in the
5807 output section definition. @xref{Output Section Region}. If no address
5808 was specified for the output section, the linker will set the address to
5809 the next available address within the memory region. If the combined
5810 output sections directed to a memory region are too large for the
5811 region, the linker will issue an error message.
5813 It is possible to access the origin and length of a memory in an
5814 expression via the @code{ORIGIN(@var{memory})} and
5815 @code{LENGTH(@var{memory})} functions:
5819 _fstack = ORIGIN(ram) + LENGTH(ram) - 4;
5824 @section PHDRS Command
5826 @cindex program headers
5827 @cindex ELF program headers
5828 @cindex program segments
5829 @cindex segments, ELF
5830 The ELF object file format uses @dfn{program headers}, also knows as
5831 @dfn{segments}. The program headers describe how the program should be
5832 loaded into memory. You can print them out by using the @code{objdump}
5833 program with the @samp{-p} option.
5835 When you run an ELF program on a native ELF system, the system loader
5836 reads the program headers in order to figure out how to load the
5837 program. This will only work if the program headers are set correctly.
5838 This manual does not describe the details of how the system loader
5839 interprets program headers; for more information, see the ELF ABI.
5841 The linker will create reasonable program headers by default. However,
5842 in some cases, you may need to specify the program headers more
5843 precisely. You may use the @code{PHDRS} command for this purpose. When
5844 the linker sees the @code{PHDRS} command in the linker script, it will
5845 not create any program headers other than the ones specified.
5847 The linker only pays attention to the @code{PHDRS} command when
5848 generating an ELF output file. In other cases, the linker will simply
5849 ignore @code{PHDRS}.
5851 This is the syntax of the @code{PHDRS} command. The words @code{PHDRS},
5852 @code{FILEHDR}, @code{AT}, and @code{FLAGS} are keywords.
5858 @var{name} @var{type} [ FILEHDR ] [ PHDRS ] [ AT ( @var{address} ) ]
5859 [ FLAGS ( @var{flags} ) ] ;
5864 The @var{name} is used only for reference in the @code{SECTIONS} command
5865 of the linker script. It is not put into the output file. Program
5866 header names are stored in a separate name space, and will not conflict
5867 with symbol names, file names, or section names. Each program header
5868 must have a distinct name. The headers are processed in order and it
5869 is usual for them to map to sections in ascending load address order.
5871 Certain program header types describe segments of memory which the
5872 system loader will load from the file. In the linker script, you
5873 specify the contents of these segments by placing allocatable output
5874 sections in the segments. You use the @samp{:@var{phdr}} output section
5875 attribute to place a section in a particular segment. @xref{Output
5878 It is normal to put certain sections in more than one segment. This
5879 merely implies that one segment of memory contains another. You may
5880 repeat @samp{:@var{phdr}}, using it once for each segment which should
5881 contain the section.
5883 If you place a section in one or more segments using @samp{:@var{phdr}},
5884 then the linker will place all subsequent allocatable sections which do
5885 not specify @samp{:@var{phdr}} in the same segments. This is for
5886 convenience, since generally a whole set of contiguous sections will be
5887 placed in a single segment. You can use @code{:NONE} to override the
5888 default segment and tell the linker to not put the section in any
5893 You may use the @code{FILEHDR} and @code{PHDRS} keywords after
5894 the program header type to further describe the contents of the segment.
5895 The @code{FILEHDR} keyword means that the segment should include the ELF
5896 file header. The @code{PHDRS} keyword means that the segment should
5897 include the ELF program headers themselves. If applied to a loadable
5898 segment (@code{PT_LOAD}), all prior loadable segments must have one of
5901 The @var{type} may be one of the following. The numbers indicate the
5902 value of the keyword.
5905 @item @code{PT_NULL} (0)
5906 Indicates an unused program header.
5908 @item @code{PT_LOAD} (1)
5909 Indicates that this program header describes a segment to be loaded from
5912 @item @code{PT_DYNAMIC} (2)
5913 Indicates a segment where dynamic linking information can be found.
5915 @item @code{PT_INTERP} (3)
5916 Indicates a segment where the name of the program interpreter may be
5919 @item @code{PT_NOTE} (4)
5920 Indicates a segment holding note information.
5922 @item @code{PT_SHLIB} (5)
5923 A reserved program header type, defined but not specified by the ELF
5926 @item @code{PT_PHDR} (6)
5927 Indicates a segment where the program headers may be found.
5929 @item @code{PT_TLS} (7)
5930 Indicates a segment containing thread local storage.
5932 @item @var{expression}
5933 An expression giving the numeric type of the program header. This may
5934 be used for types not defined above.
5937 You can specify that a segment should be loaded at a particular address
5938 in memory by using an @code{AT} expression. This is identical to the
5939 @code{AT} command used as an output section attribute (@pxref{Output
5940 Section LMA}). The @code{AT} command for a program header overrides the
5941 output section attribute.
5943 The linker will normally set the segment flags based on the sections
5944 which comprise the segment. You may use the @code{FLAGS} keyword to
5945 explicitly specify the segment flags. The value of @var{flags} must be
5946 an integer. It is used to set the @code{p_flags} field of the program
5949 Here is an example of @code{PHDRS}. This shows a typical set of program
5950 headers used on a native ELF system.
5956 headers PT_PHDR PHDRS ;
5958 text PT_LOAD FILEHDR PHDRS ;
5960 dynamic PT_DYNAMIC ;
5966 .interp : @{ *(.interp) @} :text :interp
5967 .text : @{ *(.text) @} :text
5968 .rodata : @{ *(.rodata) @} /* defaults to :text */
5970 . = . + 0x1000; /* move to a new page in memory */
5971 .data : @{ *(.data) @} :data
5972 .dynamic : @{ *(.dynamic) @} :data :dynamic
5979 @section VERSION Command
5980 @kindex VERSION @{script text@}
5981 @cindex symbol versions
5982 @cindex version script
5983 @cindex versions of symbols
5984 The linker supports symbol versions when using ELF. Symbol versions are
5985 only useful when using shared libraries. The dynamic linker can use
5986 symbol versions to select a specific version of a function when it runs
5987 a program that may have been linked against an earlier version of the
5990 You can include a version script directly in the main linker script, or
5991 you can supply the version script as an implicit linker script. You can
5992 also use the @samp{--version-script} linker option.
5994 The syntax of the @code{VERSION} command is simply
5996 VERSION @{ version-script-commands @}
5999 The format of the version script commands is identical to that used by
6000 Sun's linker in Solaris 2.5. The version script defines a tree of
6001 version nodes. You specify the node names and interdependencies in the
6002 version script. You can specify which symbols are bound to which
6003 version nodes, and you can reduce a specified set of symbols to local
6004 scope so that they are not globally visible outside of the shared
6007 The easiest way to demonstrate the version script language is with a few
6033 This example version script defines three version nodes. The first
6034 version node defined is @samp{VERS_1.1}; it has no other dependencies.
6035 The script binds the symbol @samp{foo1} to @samp{VERS_1.1}. It reduces
6036 a number of symbols to local scope so that they are not visible outside
6037 of the shared library; this is done using wildcard patterns, so that any
6038 symbol whose name begins with @samp{old}, @samp{original}, or @samp{new}
6039 is matched. The wildcard patterns available are the same as those used
6040 in the shell when matching filenames (also known as ``globbing'').
6041 However, if you specify the symbol name inside double quotes, then the
6042 name is treated as literal, rather than as a glob pattern.
6044 Next, the version script defines node @samp{VERS_1.2}. This node
6045 depends upon @samp{VERS_1.1}. The script binds the symbol @samp{foo2}
6046 to the version node @samp{VERS_1.2}.
6048 Finally, the version script defines node @samp{VERS_2.0}. This node
6049 depends upon @samp{VERS_1.2}. The scripts binds the symbols @samp{bar1}
6050 and @samp{bar2} are bound to the version node @samp{VERS_2.0}.
6052 When the linker finds a symbol defined in a library which is not
6053 specifically bound to a version node, it will effectively bind it to an
6054 unspecified base version of the library. You can bind all otherwise
6055 unspecified symbols to a given version node by using @samp{global: *;}
6056 somewhere in the version script. Note that it's slightly crazy to use
6057 wildcards in a global spec except on the last version node. Global
6058 wildcards elsewhere run the risk of accidentally adding symbols to the
6059 set exported for an old version. That's wrong since older versions
6060 ought to have a fixed set of symbols.
6062 The names of the version nodes have no specific meaning other than what
6063 they might suggest to the person reading them. The @samp{2.0} version
6064 could just as well have appeared in between @samp{1.1} and @samp{1.2}.
6065 However, this would be a confusing way to write a version script.
6067 Node name can be omitted, provided it is the only version node
6068 in the version script. Such version script doesn't assign any versions to
6069 symbols, only selects which symbols will be globally visible out and which
6073 @{ global: foo; bar; local: *; @};
6076 When you link an application against a shared library that has versioned
6077 symbols, the application itself knows which version of each symbol it
6078 requires, and it also knows which version nodes it needs from each
6079 shared library it is linked against. Thus at runtime, the dynamic
6080 loader can make a quick check to make sure that the libraries you have
6081 linked against do in fact supply all of the version nodes that the
6082 application will need to resolve all of the dynamic symbols. In this
6083 way it is possible for the dynamic linker to know with certainty that
6084 all external symbols that it needs will be resolvable without having to
6085 search for each symbol reference.
6087 The symbol versioning is in effect a much more sophisticated way of
6088 doing minor version checking that SunOS does. The fundamental problem
6089 that is being addressed here is that typically references to external
6090 functions are bound on an as-needed basis, and are not all bound when
6091 the application starts up. If a shared library is out of date, a
6092 required interface may be missing; when the application tries to use
6093 that interface, it may suddenly and unexpectedly fail. With symbol
6094 versioning, the user will get a warning when they start their program if
6095 the libraries being used with the application are too old.
6097 There are several GNU extensions to Sun's versioning approach. The
6098 first of these is the ability to bind a symbol to a version node in the
6099 source file where the symbol is defined instead of in the versioning
6100 script. This was done mainly to reduce the burden on the library
6101 maintainer. You can do this by putting something like:
6103 __asm__(".symver original_foo,foo@@VERS_1.1");
6106 in the C source file. This renames the function @samp{original_foo} to
6107 be an alias for @samp{foo} bound to the version node @samp{VERS_1.1}.
6108 The @samp{local:} directive can be used to prevent the symbol
6109 @samp{original_foo} from being exported. A @samp{.symver} directive
6110 takes precedence over a version script.
6112 The second GNU extension is to allow multiple versions of the same
6113 function to appear in a given shared library. In this way you can make
6114 an incompatible change to an interface without increasing the major
6115 version number of the shared library, while still allowing applications
6116 linked against the old interface to continue to function.
6118 To do this, you must use multiple @samp{.symver} directives in the
6119 source file. Here is an example:
6122 __asm__(".symver original_foo,foo@@");
6123 __asm__(".symver old_foo,foo@@VERS_1.1");
6124 __asm__(".symver old_foo1,foo@@VERS_1.2");
6125 __asm__(".symver new_foo,foo@@@@VERS_2.0");
6128 In this example, @samp{foo@@} represents the symbol @samp{foo} bound to the
6129 unspecified base version of the symbol. The source file that contains this
6130 example would define 4 C functions: @samp{original_foo}, @samp{old_foo},
6131 @samp{old_foo1}, and @samp{new_foo}.
6133 When you have multiple definitions of a given symbol, there needs to be
6134 some way to specify a default version to which external references to
6135 this symbol will be bound. You can do this with the
6136 @samp{foo@@@@VERS_2.0} type of @samp{.symver} directive. You can only
6137 declare one version of a symbol as the default in this manner; otherwise
6138 you would effectively have multiple definitions of the same symbol.
6140 If you wish to bind a reference to a specific version of the symbol
6141 within the shared library, you can use the aliases of convenience
6142 (i.e., @samp{old_foo}), or you can use the @samp{.symver} directive to
6143 specifically bind to an external version of the function in question.
6145 You can also specify the language in the version script:
6148 VERSION extern "lang" @{ version-script-commands @}
6151 The supported @samp{lang}s are @samp{C}, @samp{C++}, and @samp{Java}.
6152 The linker will iterate over the list of symbols at the link time and
6153 demangle them according to @samp{lang} before matching them to the
6154 patterns specified in @samp{version-script-commands}. The default
6155 @samp{lang} is @samp{C}.
6157 Demangled names may contains spaces and other special characters. As
6158 described above, you can use a glob pattern to match demangled names,
6159 or you can use a double-quoted string to match the string exactly. In
6160 the latter case, be aware that minor differences (such as differing
6161 whitespace) between the version script and the demangler output will
6162 cause a mismatch. As the exact string generated by the demangler
6163 might change in the future, even if the mangled name does not, you
6164 should check that all of your version directives are behaving as you
6165 expect when you upgrade.
6168 @section Expressions in Linker Scripts
6171 The syntax for expressions in the linker script language is identical to
6172 that of C expressions. All expressions are evaluated as integers. All
6173 expressions are evaluated in the same size, which is 32 bits if both the
6174 host and target are 32 bits, and is otherwise 64 bits.
6176 You can use and set symbol values in expressions.
6178 The linker defines several special purpose builtin functions for use in
6182 * Constants:: Constants
6183 * Symbolic Constants:: Symbolic constants
6184 * Symbols:: Symbol Names
6185 * Orphan Sections:: Orphan Sections
6186 * Location Counter:: The Location Counter
6187 * Operators:: Operators
6188 * Evaluation:: Evaluation
6189 * Expression Section:: The Section of an Expression
6190 * Builtin Functions:: Builtin Functions
6194 @subsection Constants
6195 @cindex integer notation
6196 @cindex constants in linker scripts
6197 All constants are integers.
6199 As in C, the linker considers an integer beginning with @samp{0} to be
6200 octal, and an integer beginning with @samp{0x} or @samp{0X} to be
6201 hexadecimal. Alternatively the linker accepts suffixes of @samp{h} or
6202 @samp{H} for hexadecimal, @samp{o} or @samp{O} for octal, @samp{b} or
6203 @samp{B} for binary and @samp{d} or @samp{D} for decimal. Any integer
6204 value without a prefix or a suffix is considered to be decimal.
6206 @cindex scaled integers
6207 @cindex K and M integer suffixes
6208 @cindex M and K integer suffixes
6209 @cindex suffixes for integers
6210 @cindex integer suffixes
6211 In addition, you can use the suffixes @code{K} and @code{M} to scale a
6215 @c END TEXI2ROFF-KILL
6216 @code{1024} or @code{1024*1024}
6220 ${\rm 1024}$ or ${\rm 1024}^2$
6222 @c END TEXI2ROFF-KILL
6223 respectively. For example, the following
6224 all refer to the same quantity:
6233 Note - the @code{K} and @code{M} suffixes cannot be used in
6234 conjunction with the base suffixes mentioned above.
6236 @node Symbolic Constants
6237 @subsection Symbolic Constants
6238 @cindex symbolic constants
6240 It is possible to refer to target-specific constants via the use of
6241 the @code{CONSTANT(@var{name})} operator, where @var{name} is one of:
6246 The target's maximum page size.
6248 @item COMMONPAGESIZE
6249 @kindex COMMONPAGESIZE
6250 The target's default page size.
6256 .text ALIGN (CONSTANT (MAXPAGESIZE)) : @{ *(.text) @}
6259 will create a text section aligned to the largest page boundary
6260 supported by the target.
6263 @subsection Symbol Names
6264 @cindex symbol names
6266 @cindex quoted symbol names
6268 Unless quoted, symbol names start with a letter, underscore, or period
6269 and may include letters, digits, underscores, periods, and hyphens.
6270 Unquoted symbol names must not conflict with any keywords. You can
6271 specify a symbol which contains odd characters or has the same name as a
6272 keyword by surrounding the symbol name in double quotes:
6275 "with a space" = "also with a space" + 10;
6278 Since symbols can contain many non-alphabetic characters, it is safest
6279 to delimit symbols with spaces. For example, @samp{A-B} is one symbol,
6280 whereas @samp{A - B} is an expression involving subtraction.
6282 @node Orphan Sections
6283 @subsection Orphan Sections
6285 Orphan sections are sections present in the input files which
6286 are not explicitly placed into the output file by the linker
6287 script. The linker will still copy these sections into the
6288 output file by either finding, or creating a suitable output section
6289 in which to place the orphaned input section.
6291 If the name of an orphaned input section exactly matches the name of
6292 an existing output section, then the orphaned input section will be
6293 placed at the end of that output section.
6295 If there is no output section with a matching name then new output
6296 sections will be created. Each new output section will have the same
6297 name as the orphan section placed within it. If there are multiple
6298 orphan sections with the same name, these will all be combined into
6299 one new output section.
6301 If new output sections are created to hold orphaned input sections,
6302 then the linker must decide where to place these new output sections
6303 in relation to existing output sections. On most modern targets, the
6304 linker attempts to place orphan sections after sections of the same
6305 attribute, such as code vs data, loadable vs non-loadable, etc. If no
6306 sections with matching attributes are found, or your target lacks this
6307 support, the orphan section is placed at the end of the file.
6309 The command-line options @samp{--orphan-handling} and @samp{--unique}
6310 (@pxref{Options,,Command-line Options}) can be used to control which
6311 output sections an orphan is placed in.
6313 @node Location Counter
6314 @subsection The Location Counter
6317 @cindex location counter
6318 @cindex current output location
6319 The special linker variable @dfn{dot} @samp{.} always contains the
6320 current output location counter. Since the @code{.} always refers to a
6321 location in an output section, it may only appear in an expression
6322 within a @code{SECTIONS} command. The @code{.} symbol may appear
6323 anywhere that an ordinary symbol is allowed in an expression.
6326 Assigning a value to @code{.} will cause the location counter to be
6327 moved. This may be used to create holes in the output section. The
6328 location counter may not be moved backwards inside an output section,
6329 and may not be moved backwards outside of an output section if so
6330 doing creates areas with overlapping LMAs.
6346 In the previous example, the @samp{.text} section from @file{file1} is
6347 located at the beginning of the output section @samp{output}. It is
6348 followed by a 1000 byte gap. Then the @samp{.text} section from
6349 @file{file2} appears, also with a 1000 byte gap following before the
6350 @samp{.text} section from @file{file3}. The notation @samp{= 0x12345678}
6351 specifies what data to write in the gaps (@pxref{Output Section Fill}).
6353 @cindex dot inside sections
6354 Note: @code{.} actually refers to the byte offset from the start of the
6355 current containing object. Normally this is the @code{SECTIONS}
6356 statement, whose start address is 0, hence @code{.} can be used as an
6357 absolute address. If @code{.} is used inside a section description
6358 however, it refers to the byte offset from the start of that section,
6359 not an absolute address. Thus in a script like this:
6377 The @samp{.text} section will be assigned a starting address of 0x100
6378 and a size of exactly 0x200 bytes, even if there is not enough data in
6379 the @samp{.text} input sections to fill this area. (If there is too
6380 much data, an error will be produced because this would be an attempt to
6381 move @code{.} backwards). The @samp{.data} section will start at 0x500
6382 and it will have an extra 0x600 bytes worth of space after the end of
6383 the values from the @samp{.data} input sections and before the end of
6384 the @samp{.data} output section itself.
6386 @cindex dot outside sections
6387 Setting symbols to the value of the location counter outside of an
6388 output section statement can result in unexpected values if the linker
6389 needs to place orphan sections. For example, given the following:
6395 .text: @{ *(.text) @}
6399 .data: @{ *(.data) @}
6404 If the linker needs to place some input section, e.g. @code{.rodata},
6405 not mentioned in the script, it might choose to place that section
6406 between @code{.text} and @code{.data}. You might think the linker
6407 should place @code{.rodata} on the blank line in the above script, but
6408 blank lines are of no particular significance to the linker. As well,
6409 the linker doesn't associate the above symbol names with their
6410 sections. Instead, it assumes that all assignments or other
6411 statements belong to the previous output section, except for the
6412 special case of an assignment to @code{.}. I.e., the linker will
6413 place the orphan @code{.rodata} section as if the script was written
6420 .text: @{ *(.text) @}
6424 .rodata: @{ *(.rodata) @}
6425 .data: @{ *(.data) @}
6430 This may or may not be the script author's intention for the value of
6431 @code{start_of_data}. One way to influence the orphan section
6432 placement is to assign the location counter to itself, as the linker
6433 assumes that an assignment to @code{.} is setting the start address of
6434 a following output section and thus should be grouped with that
6435 section. So you could write:
6441 .text: @{ *(.text) @}
6446 .data: @{ *(.data) @}
6451 Now, the orphan @code{.rodata} section will be placed between
6452 @code{end_of_text} and @code{start_of_data}.
6456 @subsection Operators
6457 @cindex operators for arithmetic
6458 @cindex arithmetic operators
6459 @cindex precedence in expressions
6460 The linker recognizes the standard C set of arithmetic operators, with
6461 the standard bindings and precedence levels:
6464 @c END TEXI2ROFF-KILL
6466 precedence associativity Operators Notes
6472 5 left == != > < <= >=
6478 11 right &= += -= *= /= (2)
6482 (1) Prefix operators
6483 (2) @xref{Assignments}.
6487 \vskip \baselineskip
6488 %"lispnarrowing" is the extra indent used generally for smallexample
6489 \hskip\lispnarrowing\vbox{\offinterlineskip
6492 {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr
6493 height2pt&\omit&&\omit&&\omit&\cr
6494 &Precedence&& Associativity &&{\rm Operators}&\cr
6495 height2pt&\omit&&\omit&&\omit&\cr
6497 height2pt&\omit&&\omit&&\omit&\cr
6499 % '176 is tilde, '~' in tt font
6500 &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr
6501 &2&&left&&* / \%&\cr
6504 &5&&left&&== != > < <= >=&\cr
6507 &8&&left&&{\&\&}&\cr
6510 &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr
6512 height2pt&\omit&&\omit&&\omit&\cr}
6517 @obeylines@parskip=0pt@parindent=0pt
6518 @dag@quad Prefix operators.
6519 @ddag@quad @xref{Assignments}.
6522 @c END TEXI2ROFF-KILL
6525 @subsection Evaluation
6526 @cindex lazy evaluation
6527 @cindex expression evaluation order
6528 The linker evaluates expressions lazily. It only computes the value of
6529 an expression when absolutely necessary.
6531 The linker needs some information, such as the value of the start
6532 address of the first section, and the origins and lengths of memory
6533 regions, in order to do any linking at all. These values are computed
6534 as soon as possible when the linker reads in the linker script.
6536 However, other values (such as symbol values) are not known or needed
6537 until after storage allocation. Such values are evaluated later, when
6538 other information (such as the sizes of output sections) is available
6539 for use in the symbol assignment expression.
6541 The sizes of sections cannot be known until after allocation, so
6542 assignments dependent upon these are not performed until after
6545 Some expressions, such as those depending upon the location counter
6546 @samp{.}, must be evaluated during section allocation.
6548 If the result of an expression is required, but the value is not
6549 available, then an error results. For example, a script like the
6555 .text 9+this_isnt_constant :
6561 will cause the error message @samp{non constant expression for initial
6564 @node Expression Section
6565 @subsection The Section of an Expression
6566 @cindex expression sections
6567 @cindex absolute expressions
6568 @cindex relative expressions
6569 @cindex absolute and relocatable symbols
6570 @cindex relocatable and absolute symbols
6571 @cindex symbols, relocatable and absolute
6572 Addresses and symbols may be section relative, or absolute. A section
6573 relative symbol is relocatable. If you request relocatable output
6574 using the @samp{-r} option, a further link operation may change the
6575 value of a section relative symbol. On the other hand, an absolute
6576 symbol will retain the same value throughout any further link
6579 Some terms in linker expressions are addresses. This is true of
6580 section relative symbols and for builtin functions that return an
6581 address, such as @code{ADDR}, @code{LOADADDR}, @code{ORIGIN} and
6582 @code{SEGMENT_START}. Other terms are simply numbers, or are builtin
6583 functions that return a non-address value, such as @code{LENGTH}.
6584 One complication is that unless you set @code{LD_FEATURE ("SANE_EXPR")}
6585 (@pxref{Miscellaneous Commands}), numbers and absolute symbols are treated
6586 differently depending on their location, for compatibility with older
6587 versions of @code{ld}. Expressions appearing outside an output
6588 section definition treat all numbers as absolute addresses.
6589 Expressions appearing inside an output section definition treat
6590 absolute symbols as numbers. If @code{LD_FEATURE ("SANE_EXPR")} is
6591 given, then absolute symbols and numbers are simply treated as numbers
6594 In the following simple example,
6601 __executable_start = 0x100;
6605 __data_start = 0x10;
6613 both @code{.} and @code{__executable_start} are set to the absolute
6614 address 0x100 in the first two assignments, then both @code{.} and
6615 @code{__data_start} are set to 0x10 relative to the @code{.data}
6616 section in the second two assignments.
6618 For expressions involving numbers, relative addresses and absolute
6619 addresses, ld follows these rules to evaluate terms:
6623 Unary operations on an absolute address or number, and binary
6624 operations on two absolute addresses or two numbers, or between one
6625 absolute address and a number, apply the operator to the value(s).
6627 Unary operations on a relative address, and binary operations on two
6628 relative addresses in the same section or between one relative address
6629 and a number, apply the operator to the offset part of the address(es).
6631 Other binary operations, that is, between two relative addresses not
6632 in the same section, or between a relative address and an absolute
6633 address, first convert any non-absolute term to an absolute address
6634 before applying the operator.
6637 The result section of each sub-expression is as follows:
6641 An operation involving only numbers results in a number.
6643 The result of comparisons, @samp{&&} and @samp{||} is also a number.
6645 The result of other binary arithmetic and logical operations on two
6646 relative addresses in the same section or two absolute addresses
6647 (after above conversions) is also a number when
6648 @code{LD_FEATURE ("SANE_EXPR")} or inside an output section definition
6649 but an absolute address otherwise.
6651 The result of other operations on relative addresses or one
6652 relative address and a number, is a relative address in the same
6653 section as the relative operand(s).
6655 The result of other operations on absolute addresses (after above
6656 conversions) is an absolute address.
6659 You can use the builtin function @code{ABSOLUTE} to force an expression
6660 to be absolute when it would otherwise be relative. For example, to
6661 create an absolute symbol set to the address of the end of the output
6662 section @samp{.data}:
6666 .data : @{ *(.data) _edata = ABSOLUTE(.); @}
6670 If @samp{ABSOLUTE} were not used, @samp{_edata} would be relative to the
6671 @samp{.data} section.
6673 Using @code{LOADADDR} also forces an expression absolute, since this
6674 particular builtin function returns an absolute address.
6676 @node Builtin Functions
6677 @subsection Builtin Functions
6678 @cindex functions in expressions
6679 The linker script language includes a number of builtin functions for
6680 use in linker script expressions.
6683 @item ABSOLUTE(@var{exp})
6684 @kindex ABSOLUTE(@var{exp})
6685 @cindex expression, absolute
6686 Return the absolute (non-relocatable, as opposed to non-negative) value
6687 of the expression @var{exp}. Primarily useful to assign an absolute
6688 value to a symbol within a section definition, where symbol values are
6689 normally section relative. @xref{Expression Section}.
6691 @item ADDR(@var{section})
6692 @kindex ADDR(@var{section})
6693 @cindex section address in expression
6694 Return the address (VMA) of the named @var{section}. Your
6695 script must previously have defined the location of that section. In
6696 the following example, @code{start_of_output_1}, @code{symbol_1} and
6697 @code{symbol_2} are assigned equivalent values, except that
6698 @code{symbol_1} will be relative to the @code{.output1} section while
6699 the other two will be absolute:
6705 start_of_output_1 = ABSOLUTE(.);
6710 symbol_1 = ADDR(.output1);
6711 symbol_2 = start_of_output_1;
6717 @item ALIGN(@var{align})
6718 @itemx ALIGN(@var{exp},@var{align})
6719 @kindex ALIGN(@var{align})
6720 @kindex ALIGN(@var{exp},@var{align})
6721 @cindex round up location counter
6722 @cindex align location counter
6723 @cindex round up expression
6724 @cindex align expression
6725 Return the location counter (@code{.}) or arbitrary expression aligned
6726 to the next @var{align} boundary. The single operand @code{ALIGN}
6727 doesn't change the value of the location counter---it just does
6728 arithmetic on it. The two operand @code{ALIGN} allows an arbitrary
6729 expression to be aligned upwards (@code{ALIGN(@var{align})} is
6730 equivalent to @code{ALIGN(ABSOLUTE(.), @var{align})}).
6732 Here is an example which aligns the output @code{.data} section to the
6733 next @code{0x2000} byte boundary after the preceding section and sets a
6734 variable within the section to the next @code{0x8000} boundary after the
6739 .data ALIGN(0x2000): @{
6741 variable = ALIGN(0x8000);
6747 The first use of @code{ALIGN} in this example specifies the location of
6748 a section because it is used as the optional @var{address} attribute of
6749 a section definition (@pxref{Output Section Address}). The second use
6750 of @code{ALIGN} is used to defines the value of a symbol.
6752 The builtin function @code{NEXT} is closely related to @code{ALIGN}.
6754 @item ALIGNOF(@var{section})
6755 @kindex ALIGNOF(@var{section})
6756 @cindex section alignment
6757 Return the alignment in bytes of the named @var{section}, if that section has
6758 been allocated. If the section has not been allocated when this is
6759 evaluated, the linker will report an error. In the following example,
6760 the alignment of the @code{.output} section is stored as the first
6761 value in that section.
6766 LONG (ALIGNOF (.output))
6773 @item BLOCK(@var{exp})
6774 @kindex BLOCK(@var{exp})
6775 This is a synonym for @code{ALIGN}, for compatibility with older linker
6776 scripts. It is most often seen when setting the address of an output
6779 @item DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6780 @kindex DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6781 This is equivalent to either
6783 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - 1)))
6787 (ALIGN(@var{maxpagesize})
6788 + ((. + @var{commonpagesize} - 1) & (@var{maxpagesize} - @var{commonpagesize})))
6791 depending on whether the latter uses fewer @var{commonpagesize} sized pages
6792 for the data segment (area between the result of this expression and
6793 @code{DATA_SEGMENT_END}) than the former or not.
6794 If the latter form is used, it means @var{commonpagesize} bytes of runtime
6795 memory will be saved at the expense of up to @var{commonpagesize} wasted
6796 bytes in the on-disk file.
6798 This expression can only be used directly in @code{SECTIONS} commands, not in
6799 any output section descriptions and only once in the linker script.
6800 @var{commonpagesize} should be less or equal to @var{maxpagesize} and should
6801 be the system page size the object wants to be optimized for while still
6802 running on system page sizes up to @var{maxpagesize}. Note however
6803 that @samp{-z relro} protection will not be effective if the system
6804 page size is larger than @var{commonpagesize}.
6809 . = DATA_SEGMENT_ALIGN(0x10000, 0x2000);
6812 @item DATA_SEGMENT_END(@var{exp})
6813 @kindex DATA_SEGMENT_END(@var{exp})
6814 This defines the end of data segment for @code{DATA_SEGMENT_ALIGN}
6815 evaluation purposes.
6818 . = DATA_SEGMENT_END(.);
6821 @item DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6822 @kindex DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6823 This defines the end of the @code{PT_GNU_RELRO} segment when
6824 @samp{-z relro} option is used.
6825 When @samp{-z relro} option is not present, @code{DATA_SEGMENT_RELRO_END}
6826 does nothing, otherwise @code{DATA_SEGMENT_ALIGN} is padded so that
6827 @var{exp} + @var{offset} is aligned to the @var{commonpagesize}
6828 argument given to @code{DATA_SEGMENT_ALIGN}. If present in the linker
6829 script, it must be placed between @code{DATA_SEGMENT_ALIGN} and
6830 @code{DATA_SEGMENT_END}. Evaluates to the second argument plus any
6831 padding needed at the end of the @code{PT_GNU_RELRO} segment due to
6835 . = DATA_SEGMENT_RELRO_END(24, .);
6838 @item DEFINED(@var{symbol})
6839 @kindex DEFINED(@var{symbol})
6840 @cindex symbol defaults
6841 Return 1 if @var{symbol} is in the linker global symbol table and is
6842 defined before the statement using DEFINED in the script, otherwise
6843 return 0. You can use this function to provide
6844 default values for symbols. For example, the following script fragment
6845 shows how to set a global symbol @samp{begin} to the first location in
6846 the @samp{.text} section---but if a symbol called @samp{begin} already
6847 existed, its value is preserved:
6853 begin = DEFINED(begin) ? begin : . ;
6861 @item LENGTH(@var{memory})
6862 @kindex LENGTH(@var{memory})
6863 Return the length of the memory region named @var{memory}.
6865 @item LOADADDR(@var{section})
6866 @kindex LOADADDR(@var{section})
6867 @cindex section load address in expression
6868 Return the absolute LMA of the named @var{section}. (@pxref{Output
6871 @item LOG2CEIL(@var{exp})
6872 @kindex LOG2CEIL(@var{exp})
6873 Return the binary logarithm of @var{exp} rounded towards infinity.
6874 @code{LOG2CEIL(0)} returns 0.
6877 @item MAX(@var{exp1}, @var{exp2})
6878 Returns the maximum of @var{exp1} and @var{exp2}.
6881 @item MIN(@var{exp1}, @var{exp2})
6882 Returns the minimum of @var{exp1} and @var{exp2}.
6884 @item NEXT(@var{exp})
6885 @kindex NEXT(@var{exp})
6886 @cindex unallocated address, next
6887 Return the next unallocated address that is a multiple of @var{exp}.
6888 This function is closely related to @code{ALIGN(@var{exp})}; unless you
6889 use the @code{MEMORY} command to define discontinuous memory for the
6890 output file, the two functions are equivalent.
6892 @item ORIGIN(@var{memory})
6893 @kindex ORIGIN(@var{memory})
6894 Return the origin of the memory region named @var{memory}.
6896 @item SEGMENT_START(@var{segment}, @var{default})
6897 @kindex SEGMENT_START(@var{segment}, @var{default})
6898 Return the base address of the named @var{segment}. If an explicit
6899 value has already been given for this segment (with a command-line
6900 @samp{-T} option) then that value will be returned otherwise the value
6901 will be @var{default}. At present, the @samp{-T} command-line option
6902 can only be used to set the base address for the ``text'', ``data'', and
6903 ``bss'' sections, but you can use @code{SEGMENT_START} with any segment
6906 @item SIZEOF(@var{section})
6907 @kindex SIZEOF(@var{section})
6908 @cindex section size
6909 Return the size in bytes of the named @var{section}, if that section has
6910 been allocated. If the section has not been allocated when this is
6911 evaluated, the linker will report an error. In the following example,
6912 @code{symbol_1} and @code{symbol_2} are assigned identical values:
6921 symbol_1 = .end - .start ;
6922 symbol_2 = SIZEOF(.output);
6927 @item SIZEOF_HEADERS
6928 @itemx sizeof_headers
6929 @kindex SIZEOF_HEADERS
6931 Return the size in bytes of the output file's headers. This is
6932 information which appears at the start of the output file. You can use
6933 this number when setting the start address of the first section, if you
6934 choose, to facilitate paging.
6936 @cindex not enough room for program headers
6937 @cindex program headers, not enough room
6938 When producing an ELF output file, if the linker script uses the
6939 @code{SIZEOF_HEADERS} builtin function, the linker must compute the
6940 number of program headers before it has determined all the section
6941 addresses and sizes. If the linker later discovers that it needs
6942 additional program headers, it will report an error @samp{not enough
6943 room for program headers}. To avoid this error, you must avoid using
6944 the @code{SIZEOF_HEADERS} function, or you must rework your linker
6945 script to avoid forcing the linker to use additional program headers, or
6946 you must define the program headers yourself using the @code{PHDRS}
6947 command (@pxref{PHDRS}).
6950 @node Implicit Linker Scripts
6951 @section Implicit Linker Scripts
6952 @cindex implicit linker scripts
6953 If you specify a linker input file which the linker can not recognize as
6954 an object file or an archive file, it will try to read the file as a
6955 linker script. If the file can not be parsed as a linker script, the
6956 linker will report an error.
6958 An implicit linker script will not replace the default linker script.
6960 Typically an implicit linker script would contain only symbol
6961 assignments, or the @code{INPUT}, @code{GROUP}, or @code{VERSION}
6964 Any input files read because of an implicit linker script will be read
6965 at the position in the command line where the implicit linker script was
6966 read. This can affect archive searching.
6969 @chapter Linker Plugins
6972 @cindex linker plugins
6973 The linker can use dynamically loaded plugins to modify its behavior.
6974 For example, the link-time optimization feature that some compilers
6975 support is implemented with a linker plugin.
6977 Currently there is only one plugin shipped by default, but more may
6978 be added here later.
6981 * libdep Plugin:: Static Library Dependencies Plugin
6985 @section Static Library Dependencies Plugin
6986 @cindex static library dependencies
6987 Originally, static libraries were contained in an archive file consisting
6988 just of a collection of relocatable object files. Later they evolved to
6989 optionally include a symbol table, to assist in finding the needed objects
6990 within a library. There their evolution ended, and dynamic libraries
6993 One useful feature of dynamic libraries was that, more than just collecting
6994 multiple objects into a single file, they also included a list of their
6995 dependencies, such that one could specify just the name of a single dynamic
6996 library at link time, and all of its dependencies would be implicitly
6997 referenced as well. But static libraries lacked this feature, so if a
6998 link invocation was switched from using dynamic libraries to static
6999 libraries, the link command would usually fail unless it was rewritten to
7000 explicitly list the dependencies of the static library.
7002 The GNU @command{ar} utility now supports a @option{--record-libdeps} option
7003 to embed dependency lists into static libraries as well, and the @file{libdep}
7004 plugin may be used to read this dependency information at link time. The
7005 dependency information is stored as a single string, carrying @option{-l}
7006 and @option{-L} arguments as they would normally appear in a linker
7007 command line. As such, the information can be written with any text
7008 utility and stored into any archive, even if GNU @command{ar} is not
7009 being used to create the archive. The information is stored in an
7010 archive member named @samp{__.LIBDEP}.
7012 For example, given a library @file{libssl.a} that depends on another
7013 library @file{libcrypto.a} which may be found in @file{/usr/local/lib},
7014 the @samp{__.LIBDEP} member of @file{libssl.a} would contain
7017 -L/usr/local/lib -lcrypto
7021 @node Machine Dependent
7022 @chapter Machine Dependent Features
7024 @cindex machine dependencies
7025 @command{ld} has additional features on some platforms; the following
7026 sections describe them. Machines where @command{ld} has no additional
7027 functionality are not listed.
7031 * H8/300:: @command{ld} and the H8/300
7034 * M68HC11/68HC12:: @code{ld} and the Motorola 68HC11 and 68HC12 families
7037 * ARM:: @command{ld} and the ARM family
7040 * HPPA ELF32:: @command{ld} and HPPA 32-bit ELF
7043 * M68K:: @command{ld} and the Motorola 68K family
7046 * MIPS:: @command{ld} and the MIPS family
7049 * MMIX:: @command{ld} and MMIX
7052 * MSP430:: @command{ld} and MSP430
7055 * NDS32:: @command{ld} and NDS32
7058 * Nios II:: @command{ld} and the Altera Nios II
7061 * PowerPC ELF32:: @command{ld} and PowerPC 32-bit ELF Support
7064 * PowerPC64 ELF64:: @command{ld} and PowerPC64 64-bit ELF Support
7067 * S/390 ELF:: @command{ld} and S/390 ELF Support
7070 * SPU ELF:: @command{ld} and SPU ELF Support
7073 * TI COFF:: @command{ld} and TI COFF
7076 * WIN32:: @command{ld} and WIN32 (cygwin/mingw)
7079 * Xtensa:: @command{ld} and Xtensa Processors
7090 @section @command{ld} and the H8/300
7092 @cindex H8/300 support
7093 For the H8/300, @command{ld} can perform these global optimizations when
7094 you specify the @samp{--relax} command-line option.
7097 @cindex relaxing on H8/300
7098 @item relaxing address modes
7099 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
7100 targets are within eight bits, and turns them into eight-bit
7101 program-counter relative @code{bsr} and @code{bra} instructions,
7104 @cindex synthesizing on H8/300
7105 @item synthesizing instructions
7106 @c FIXME: specifically mov.b, or any mov instructions really? -> mov.b only, at least on H8, H8H, H8S
7107 @command{ld} finds all @code{mov.b} instructions which use the
7108 sixteen-bit absolute address form, but refer to the top
7109 page of memory, and changes them to use the eight-bit address form.
7110 (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into
7111 @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the
7112 top page of memory).
7114 @command{ld} finds all @code{mov} instructions which use the register
7115 indirect with 32-bit displacement addressing mode, but use a small
7116 displacement inside 16-bit displacement range, and changes them to use
7117 the 16-bit displacement form. (That is: the linker turns @samp{mov.b
7118 @code{@@}@var{d}:32,ERx} into @samp{mov.b @code{@@}@var{d}:16,ERx}
7119 whenever the displacement @var{d} is in the 16 bit signed integer
7120 range. Only implemented in ELF-format ld).
7122 @item bit manipulation instructions
7123 @command{ld} finds all bit manipulation instructions like @code{band, bclr,
7124 biand, bild, bior, bist, bixor, bld, bnot, bor, bset, bst, btst, bxor}
7125 which use 32 bit and 16 bit absolute address form, but refer to the top
7126 page of memory, and changes them to use the 8 bit address form.
7127 (That is: the linker turns @samp{bset #xx:3,@code{@@}@var{aa}:32} into
7128 @samp{bset #xx:3,@code{@@}@var{aa}:8} whenever the address @var{aa} is in
7129 the top page of memory).
7131 @item system control instructions
7132 @command{ld} finds all @code{ldc.w, stc.w} instructions which use the
7133 32 bit absolute address form, but refer to the top page of memory, and
7134 changes them to use 16 bit address form.
7135 (That is: the linker turns @samp{ldc.w @code{@@}@var{aa}:32,ccr} into
7136 @samp{ldc.w @code{@@}@var{aa}:16,ccr} whenever the address @var{aa} is in
7137 the top page of memory).
7147 @c This stuff is pointless to say unless you're especially concerned
7148 @c with Renesas chips; don't enable it for generic case, please.
7150 @chapter @command{ld} and Other Renesas Chips
7152 @command{ld} also supports the Renesas (formerly Hitachi) H8/300H,
7153 H8/500, and SH chips. No special features, commands, or command-line
7154 options are required for these chips.
7168 @node M68HC11/68HC12
7169 @section @command{ld} and the Motorola 68HC11 and 68HC12 families
7171 @cindex M68HC11 and 68HC12 support
7173 @subsection Linker Relaxation
7175 For the Motorola 68HC11, @command{ld} can perform these global
7176 optimizations when you specify the @samp{--relax} command-line option.
7179 @cindex relaxing on M68HC11
7180 @item relaxing address modes
7181 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
7182 targets are within eight bits, and turns them into eight-bit
7183 program-counter relative @code{bsr} and @code{bra} instructions,
7186 @command{ld} also looks at all 16-bit extended addressing modes and
7187 transforms them in a direct addressing mode when the address is in
7188 page 0 (between 0 and 0x0ff).
7190 @item relaxing gcc instruction group
7191 When @command{gcc} is called with @option{-mrelax}, it can emit group
7192 of instructions that the linker can optimize to use a 68HC11 direct
7193 addressing mode. These instructions consists of @code{bclr} or
7194 @code{bset} instructions.
7198 @subsection Trampoline Generation
7200 @cindex trampoline generation on M68HC11
7201 @cindex trampoline generation on M68HC12
7202 For 68HC11 and 68HC12, @command{ld} can generate trampoline code to
7203 call a far function using a normal @code{jsr} instruction. The linker
7204 will also change the relocation to some far function to use the
7205 trampoline address instead of the function address. This is typically the
7206 case when a pointer to a function is taken. The pointer will in fact
7207 point to the function trampoline.
7215 @section @command{ld} and the ARM family
7217 @cindex ARM interworking support
7218 @kindex --support-old-code
7219 For the ARM, @command{ld} will generate code stubs to allow functions calls
7220 between ARM and Thumb code. These stubs only work with code that has
7221 been compiled and assembled with the @samp{-mthumb-interwork} command
7222 line option. If it is necessary to link with old ARM object files or
7223 libraries, which have not been compiled with the -mthumb-interwork
7224 option then the @samp{--support-old-code} command-line switch should be
7225 given to the linker. This will make it generate larger stub functions
7226 which will work with non-interworking aware ARM code. Note, however,
7227 the linker does not support generating stubs for function calls to
7228 non-interworking aware Thumb code.
7230 @cindex thumb entry point
7231 @cindex entry point, thumb
7232 @kindex --thumb-entry=@var{entry}
7233 The @samp{--thumb-entry} switch is a duplicate of the generic
7234 @samp{--entry} switch, in that it sets the program's starting address.
7235 But it also sets the bottom bit of the address, so that it can be
7236 branched to using a BX instruction, and the program will start
7237 executing in Thumb mode straight away.
7239 @cindex PE import table prefixing
7240 @kindex --use-nul-prefixed-import-tables
7241 The @samp{--use-nul-prefixed-import-tables} switch is specifying, that
7242 the import tables idata4 and idata5 have to be generated with a zero
7243 element prefix for import libraries. This is the old style to generate
7244 import tables. By default this option is turned off.
7248 The @samp{--be8} switch instructs @command{ld} to generate BE8 format
7249 executables. This option is only valid when linking big-endian
7250 objects - ie ones which have been assembled with the @option{-EB}
7251 option. The resulting image will contain big-endian data and
7255 @kindex --target1-rel
7256 @kindex --target1-abs
7257 The @samp{R_ARM_TARGET1} relocation is typically used for entries in the
7258 @samp{.init_array} section. It is interpreted as either @samp{R_ARM_REL32}
7259 or @samp{R_ARM_ABS32}, depending on the target. The @samp{--target1-rel}
7260 and @samp{--target1-abs} switches override the default.
7263 @kindex --target2=@var{type}
7264 The @samp{--target2=type} switch overrides the default definition of the
7265 @samp{R_ARM_TARGET2} relocation. Valid values for @samp{type}, their
7266 meanings, and target defaults are as follows:
7269 @samp{R_ARM_REL32} (arm*-*-elf, arm*-*-eabi)
7271 @samp{R_ARM_ABS32} (arm*-*-symbianelf)
7273 @samp{R_ARM_GOT_PREL} (arm*-*-linux, arm*-*-*bsd)
7278 The @samp{R_ARM_V4BX} relocation (defined by the ARM AAELF
7279 specification) enables objects compiled for the ARMv4 architecture to be
7280 interworking-safe when linked with other objects compiled for ARMv4t, but
7281 also allows pure ARMv4 binaries to be built from the same ARMv4 objects.
7283 In the latter case, the switch @option{--fix-v4bx} must be passed to the
7284 linker, which causes v4t @code{BX rM} instructions to be rewritten as
7285 @code{MOV PC,rM}, since v4 processors do not have a @code{BX} instruction.
7287 In the former case, the switch should not be used, and @samp{R_ARM_V4BX}
7288 relocations are ignored.
7290 @cindex FIX_V4BX_INTERWORKING
7291 @kindex --fix-v4bx-interworking
7292 Replace @code{BX rM} instructions identified by @samp{R_ARM_V4BX}
7293 relocations with a branch to the following veneer:
7301 This allows generation of libraries/applications that work on ARMv4 cores
7302 and are still interworking safe. Note that the above veneer clobbers the
7303 condition flags, so may cause incorrect program behavior in rare cases.
7307 The @samp{--use-blx} switch enables the linker to use ARM/Thumb
7308 BLX instructions (available on ARMv5t and above) in various
7309 situations. Currently it is used to perform calls via the PLT from Thumb
7310 code using BLX rather than using BX and a mode-switching stub before
7311 each PLT entry. This should lead to such calls executing slightly faster.
7313 This option is enabled implicitly for SymbianOS, so there is no need to
7314 specify it if you are using that target.
7316 @cindex VFP11_DENORM_FIX
7317 @kindex --vfp11-denorm-fix
7318 The @samp{--vfp11-denorm-fix} switch enables a link-time workaround for a
7319 bug in certain VFP11 coprocessor hardware, which sometimes allows
7320 instructions with denorm operands (which must be handled by support code)
7321 to have those operands overwritten by subsequent instructions before
7322 the support code can read the intended values.
7324 The bug may be avoided in scalar mode if you allow at least one
7325 intervening instruction between a VFP11 instruction which uses a register
7326 and another instruction which writes to the same register, or at least two
7327 intervening instructions if vector mode is in use. The bug only affects
7328 full-compliance floating-point mode: you do not need this workaround if
7329 you are using "runfast" mode. Please contact ARM for further details.
7331 If you know you are using buggy VFP11 hardware, you can
7332 enable this workaround by specifying the linker option
7333 @samp{--vfp-denorm-fix=scalar} if you are using the VFP11 scalar
7334 mode only, or @samp{--vfp-denorm-fix=vector} if you are using
7335 vector mode (the latter also works for scalar code). The default is
7336 @samp{--vfp-denorm-fix=none}.
7338 If the workaround is enabled, instructions are scanned for
7339 potentially-troublesome sequences, and a veneer is created for each
7340 such sequence which may trigger the erratum. The veneer consists of the
7341 first instruction of the sequence and a branch back to the subsequent
7342 instruction. The original instruction is then replaced with a branch to
7343 the veneer. The extra cycles required to call and return from the veneer
7344 are sufficient to avoid the erratum in both the scalar and vector cases.
7346 @cindex ARM1176 erratum workaround
7347 @kindex --fix-arm1176
7348 @kindex --no-fix-arm1176
7349 The @samp{--fix-arm1176} switch enables a link-time workaround for an erratum
7350 in certain ARM1176 processors. The workaround is enabled by default if you
7351 are targeting ARM v6 (excluding ARM v6T2) or earlier. It can be disabled
7352 unconditionally by specifying @samp{--no-fix-arm1176}.
7354 Further information is available in the ``ARM1176JZ-S and ARM1176JZF-S
7355 Programmer Advice Notice'' available on the ARM documentation website at:
7356 http://infocenter.arm.com/.
7358 @cindex STM32L4xx erratum workaround
7359 @kindex --fix-stm32l4xx-629360
7361 The @samp{--fix-stm32l4xx-629360} switch enables a link-time
7362 workaround for a bug in the bus matrix / memory controller for some of
7363 the STM32 Cortex-M4 based products (STM32L4xx). When accessing
7364 off-chip memory via the affected bus for bus reads of 9 words or more,
7365 the bus can generate corrupt data and/or abort. These are only
7366 core-initiated accesses (not DMA), and might affect any access:
7367 integer loads such as LDM, POP and floating-point loads such as VLDM,
7368 VPOP. Stores are not affected.
7370 The bug can be avoided by splitting memory accesses into the
7371 necessary chunks to keep bus reads below 8 words.
7373 The workaround is not enabled by default, this is equivalent to use
7374 @samp{--fix-stm32l4xx-629360=none}. If you know you are using buggy
7375 STM32L4xx hardware, you can enable the workaround by specifying the
7376 linker option @samp{--fix-stm32l4xx-629360}, or the equivalent
7377 @samp{--fix-stm32l4xx-629360=default}.
7379 If the workaround is enabled, instructions are scanned for
7380 potentially-troublesome sequences, and a veneer is created for each
7381 such sequence which may trigger the erratum. The veneer consists in a
7382 replacement sequence emulating the behaviour of the original one and a
7383 branch back to the subsequent instruction. The original instruction is
7384 then replaced with a branch to the veneer.
7386 The workaround does not always preserve the memory access order for
7387 the LDMDB instruction, when the instruction loads the PC.
7389 The workaround is not able to handle problematic instructions when
7390 they are in the middle of an IT block, since a branch is not allowed
7391 there. In that case, the linker reports a warning and no replacement
7394 The workaround is not able to replace problematic instructions with a
7395 PC-relative branch instruction if the @samp{.text} section is too
7396 large. In that case, when the branch that replaces the original code
7397 cannot be encoded, the linker reports a warning and no replacement
7400 @cindex NO_ENUM_SIZE_WARNING
7401 @kindex --no-enum-size-warning
7402 The @option{--no-enum-size-warning} switch prevents the linker from
7403 warning when linking object files that specify incompatible EABI
7404 enumeration size attributes. For example, with this switch enabled,
7405 linking of an object file using 32-bit enumeration values with another
7406 using enumeration values fitted into the smallest possible space will
7409 @cindex NO_WCHAR_SIZE_WARNING
7410 @kindex --no-wchar-size-warning
7411 The @option{--no-wchar-size-warning} switch prevents the linker from
7412 warning when linking object files that specify incompatible EABI
7413 @code{wchar_t} size attributes. For example, with this switch enabled,
7414 linking of an object file using 32-bit @code{wchar_t} values with another
7415 using 16-bit @code{wchar_t} values will not be diagnosed.
7418 @kindex --pic-veneer
7419 The @samp{--pic-veneer} switch makes the linker use PIC sequences for
7420 ARM/Thumb interworking veneers, even if the rest of the binary
7421 is not PIC. This avoids problems on uClinux targets where
7422 @samp{--emit-relocs} is used to generate relocatable binaries.
7424 @cindex STUB_GROUP_SIZE
7425 @kindex --stub-group-size=@var{N}
7426 The linker will automatically generate and insert small sequences of
7427 code into a linked ARM ELF executable whenever an attempt is made to
7428 perform a function call to a symbol that is too far away. The
7429 placement of these sequences of instructions - called stubs - is
7430 controlled by the command-line option @option{--stub-group-size=N}.
7431 The placement is important because a poor choice can create a need for
7432 duplicate stubs, increasing the code size. The linker will try to
7433 group stubs together in order to reduce interruptions to the flow of
7434 code, but it needs guidance as to how big these groups should be and
7435 where they should be placed.
7437 The value of @samp{N}, the parameter to the
7438 @option{--stub-group-size=} option controls where the stub groups are
7439 placed. If it is negative then all stubs are placed after the first
7440 branch that needs them. If it is positive then the stubs can be
7441 placed either before or after the branches that need them. If the
7442 value of @samp{N} is 1 (either +1 or -1) then the linker will choose
7443 exactly where to place groups of stubs, using its built in heuristics.
7444 A value of @samp{N} greater than 1 (or smaller than -1) tells the
7445 linker that a single group of stubs can service at most @samp{N} bytes
7446 from the input sections.
7448 The default, if @option{--stub-group-size=} is not specified, is
7451 Farcalls stubs insertion is fully supported for the ARM-EABI target
7452 only, because it relies on object files properties not present
7455 @cindex Cortex-A8 erratum workaround
7456 @kindex --fix-cortex-a8
7457 @kindex --no-fix-cortex-a8
7458 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}.
7460 The erratum only affects Thumb-2 code. Please contact ARM for further details.
7462 @cindex Cortex-A53 erratum 835769 workaround
7463 @kindex --fix-cortex-a53-835769
7464 @kindex --no-fix-cortex-a53-835769
7465 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}.
7467 Please contact ARM for further details.
7469 @kindex --merge-exidx-entries
7470 @kindex --no-merge-exidx-entries
7471 @cindex Merging exidx entries
7472 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent exidx entries in debuginfo.
7475 @cindex 32-bit PLT entries
7476 The @samp{--long-plt} option enables the use of 16 byte PLT entries
7477 which support up to 4Gb of code. The default is to use 12 byte PLT
7478 entries which only support 512Mb of code.
7480 @kindex --no-apply-dynamic-relocs
7481 @cindex AArch64 rela addend
7482 The @samp{--no-apply-dynamic-relocs} option makes AArch64 linker do not apply
7483 link-time values for dynamic relocations.
7485 @cindex Placement of SG veneers
7486 All SG veneers are placed in the special output section @code{.gnu.sgstubs}.
7487 Its start address must be set, either with the command-line option
7488 @samp{--section-start} or in a linker script, to indicate where to place these
7491 @kindex --cmse-implib
7492 @cindex Secure gateway import library
7493 The @samp{--cmse-implib} option requests that the import libraries
7494 specified by the @samp{--out-implib} and @samp{--in-implib} options are
7495 secure gateway import libraries, suitable for linking a non-secure
7496 executable against secure code as per ARMv8-M Security Extensions.
7498 @kindex --in-implib=@var{file}
7499 @cindex Input import library
7500 The @samp{--in-implib=file} specifies an input import library whose symbols
7501 must keep the same address in the executable being produced. A warning is
7502 given if no @samp{--out-implib} is given but new symbols have been introduced
7503 in the executable that should be listed in its import library. Otherwise, if
7504 @samp{--out-implib} is specified, the symbols are added to the output import
7505 library. A warning is also given if some symbols present in the input import
7506 library have disappeared from the executable. This option is only effective
7507 for Secure Gateway import libraries, ie. when @samp{--cmse-implib} is
7521 @section @command{ld} and HPPA 32-bit ELF Support
7522 @cindex HPPA multiple sub-space stubs
7523 @kindex --multi-subspace
7524 When generating a shared library, @command{ld} will by default generate
7525 import stubs suitable for use with a single sub-space application.
7526 The @samp{--multi-subspace} switch causes @command{ld} to generate export
7527 stubs, and different (larger) import stubs suitable for use with
7528 multiple sub-spaces.
7530 @cindex HPPA stub grouping
7531 @kindex --stub-group-size=@var{N}
7532 Long branch stubs and import/export stubs are placed by @command{ld} in
7533 stub sections located between groups of input sections.
7534 @samp{--stub-group-size} specifies the maximum size of a group of input
7535 sections handled by one stub section. Since branch offsets are signed,
7536 a stub section may serve two groups of input sections, one group before
7537 the stub section, and one group after it. However, when using
7538 conditional branches that require stubs, it may be better (for branch
7539 prediction) that stub sections only serve one group of input sections.
7540 A negative value for @samp{N} chooses this scheme, ensuring that
7541 branches to stubs always use a negative offset. Two special values of
7542 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
7543 @command{ld} to automatically size input section groups for the branch types
7544 detected, with the same behaviour regarding stub placement as other
7545 positive or negative values of @samp{N} respectively.
7547 Note that @samp{--stub-group-size} does not split input sections. A
7548 single input section larger than the group size specified will of course
7549 create a larger group (of one section). If input sections are too
7550 large, it may not be possible for a branch to reach its stub.
7563 @section @command{ld} and the Motorola 68K family
7565 @cindex Motorola 68K GOT generation
7566 @kindex --got=@var{type}
7567 The @samp{--got=@var{type}} option lets you choose the GOT generation scheme.
7568 The choices are @samp{single}, @samp{negative}, @samp{multigot} and
7569 @samp{target}. When @samp{target} is selected the linker chooses
7570 the default GOT generation scheme for the current target.
7571 @samp{single} tells the linker to generate a single GOT with
7572 entries only at non-negative offsets.
7573 @samp{negative} instructs the linker to generate a single GOT with
7574 entries at both negative and positive offsets. Not all environments
7576 @samp{multigot} allows the linker to generate several GOTs in the
7577 output file. All GOT references from a single input object
7578 file access the same GOT, but references from different input object
7579 files might access different GOTs. Not all environments support such GOTs.
7592 @section @command{ld} and the MIPS family
7594 @cindex MIPS microMIPS instruction choice selection
7597 The @samp{--insn32} and @samp{--no-insn32} options control the choice of
7598 microMIPS instructions used in code generated by the linker, such as that
7599 in the PLT or lazy binding stubs, or in relaxation. If @samp{--insn32} is
7600 used, then the linker only uses 32-bit instruction encodings. By default
7601 or if @samp{--no-insn32} is used, all instruction encodings are used,
7602 including 16-bit ones where possible.
7604 @cindex MIPS branch relocation check control
7605 @kindex --ignore-branch-isa
7606 @kindex --no-ignore-branch-isa
7607 The @samp{--ignore-branch-isa} and @samp{--no-ignore-branch-isa} options
7608 control branch relocation checks for invalid ISA mode transitions. If
7609 @samp{--ignore-branch-isa} is used, then the linker accepts any branch
7610 relocations and any ISA mode transition required is lost in relocation
7611 calculation, except for some cases of @code{BAL} instructions which meet
7612 relaxation conditions and are converted to equivalent @code{JALX}
7613 instructions as the associated relocation is calculated. By default
7614 or if @samp{--no-ignore-branch-isa} is used a check is made causing
7615 the loss of an ISA mode transition to produce an error.
7628 @section @code{ld} and MMIX
7629 For MMIX, there is a choice of generating @code{ELF} object files or
7630 @code{mmo} object files when linking. The simulator @code{mmix}
7631 understands the @code{mmo} format. The binutils @code{objcopy} utility
7632 can translate between the two formats.
7634 There is one special section, the @samp{.MMIX.reg_contents} section.
7635 Contents in this section is assumed to correspond to that of global
7636 registers, and symbols referring to it are translated to special symbols,
7637 equal to registers. In a final link, the start address of the
7638 @samp{.MMIX.reg_contents} section corresponds to the first allocated
7639 global register multiplied by 8. Register @code{$255} is not included in
7640 this section; it is always set to the program entry, which is at the
7641 symbol @code{Main} for @code{mmo} files.
7643 Global symbols with the prefix @code{__.MMIX.start.}, for example
7644 @code{__.MMIX.start..text} and @code{__.MMIX.start..data} are special.
7645 The default linker script uses these to set the default start address
7648 Initial and trailing multiples of zero-valued 32-bit words in a section,
7649 are left out from an mmo file.
7662 @section @code{ld} and MSP430
7663 For the MSP430 it is possible to select the MPU architecture. The flag @samp{-m [mpu type]}
7664 will select an appropriate linker script for selected MPU type. (To get a list of known MPUs
7665 just pass @samp{-m help} option to the linker).
7667 @cindex MSP430 extra sections
7668 The linker will recognize some extra sections which are MSP430 specific:
7671 @item @samp{.vectors}
7672 Defines a portion of ROM where interrupt vectors located.
7674 @item @samp{.bootloader}
7675 Defines the bootloader portion of the ROM (if applicable). Any code
7676 in this section will be uploaded to the MPU.
7678 @item @samp{.infomem}
7679 Defines an information memory section (if applicable). Any code in
7680 this section will be uploaded to the MPU.
7682 @item @samp{.infomemnobits}
7683 This is the same as the @samp{.infomem} section except that any code
7684 in this section will not be uploaded to the MPU.
7686 @item @samp{.noinit}
7687 Denotes a portion of RAM located above @samp{.bss} section.
7689 The last two sections are used by gcc.
7693 @cindex MSP430 Options
7694 @kindex --code-region
7695 @item --code-region=[either,lower,upper,none]
7696 This will transform .text* sections to [either,lower,upper].text* sections. The
7697 argument passed to GCC for -mcode-region is propagated to the linker
7700 @kindex --data-region
7701 @item --data-region=[either,lower,upper,none]
7702 This will transform .data*, .bss* and .rodata* sections to
7703 [either,lower,upper].[data,bss,rodata]* sections. The argument passed to GCC
7704 for -mdata-region is propagated to the linker using this option.
7706 @kindex --disable-sec-transformation
7707 @item --disable-sec-transformation
7708 Prevent the transformation of sections as specified by the @code{--code-region}
7709 and @code{--data-region} options.
7710 This is useful if you are compiling and linking using a single call to the GCC
7711 wrapper, and want to compile the source files using -m[code,data]-region but
7712 not transform the sections for prebuilt libraries and objects.
7726 @section @code{ld} and NDS32
7727 @kindex relaxing on NDS32
7728 For NDS32, there are some options to select relaxation behavior. The linker
7729 relaxes objects according to these options.
7732 @item @samp{--m[no-]fp-as-gp}
7733 Disable/enable fp-as-gp relaxation.
7735 @item @samp{--mexport-symbols=FILE}
7736 Exporting symbols and their address into FILE as linker script.
7738 @item @samp{--m[no-]ex9}
7739 Disable/enable link-time EX9 relaxation.
7741 @item @samp{--mexport-ex9=FILE}
7742 Export the EX9 table after linking.
7744 @item @samp{--mimport-ex9=FILE}
7745 Import the Ex9 table for EX9 relaxation.
7747 @item @samp{--mupdate-ex9}
7748 Update the existing EX9 table.
7750 @item @samp{--mex9-limit=NUM}
7751 Maximum number of entries in the ex9 table.
7753 @item @samp{--mex9-loop-aware}
7754 Avoid generating the EX9 instruction inside the loop.
7756 @item @samp{--m[no-]ifc}
7757 Disable/enable the link-time IFC optimization.
7759 @item @samp{--mifc-loop-aware}
7760 Avoid generating the IFC instruction inside the loop.
7774 @section @command{ld} and the Altera Nios II
7775 @cindex Nios II call relaxation
7776 @kindex --relax on Nios II
7778 Call and immediate jump instructions on Nios II processors are limited to
7779 transferring control to addresses in the same 256MB memory segment,
7780 which may result in @command{ld} giving
7781 @samp{relocation truncated to fit} errors with very large programs.
7782 The command-line option @option{--relax} enables the generation of
7783 trampolines that can access the entire 32-bit address space for calls
7784 outside the normal @code{call} and @code{jmpi} address range. These
7785 trampolines are inserted at section boundaries, so may not themselves
7786 be reachable if an input section and its associated call trampolines are
7789 The @option{--relax} option is enabled by default unless @option{-r}
7790 is also specified. You can disable trampoline generation by using the
7791 @option{--no-relax} linker option. You can also disable this optimization
7792 locally by using the @samp{set .noat} directive in assembly-language
7793 source files, as the linker-inserted trampolines use the @code{at}
7794 register as a temporary.
7796 Note that the linker @option{--relax} option is independent of assembler
7797 relaxation options, and that using the GNU assembler's @option{-relax-all}
7798 option interferes with the linker's more selective call instruction relaxation.
7811 @section @command{ld} and PowerPC 32-bit ELF Support
7812 @cindex PowerPC long branches
7813 @kindex --relax on PowerPC
7814 Branches on PowerPC processors are limited to a signed 26-bit
7815 displacement, which may result in @command{ld} giving
7816 @samp{relocation truncated to fit} errors with very large programs.
7817 @samp{--relax} enables the generation of trampolines that can access
7818 the entire 32-bit address space. These trampolines are inserted at
7819 section boundaries, so may not themselves be reachable if an input
7820 section exceeds 33M in size. You may combine @samp{-r} and
7821 @samp{--relax} to add trampolines in a partial link. In that case
7822 both branches to undefined symbols and inter-section branches are also
7823 considered potentially out of range, and trampolines inserted.
7825 @cindex PowerPC ELF32 options
7830 Current PowerPC GCC accepts a @samp{-msecure-plt} option that
7831 generates code capable of using a newer PLT and GOT layout that has
7832 the security advantage of no executable section ever needing to be
7833 writable and no writable section ever being executable. PowerPC
7834 @command{ld} will generate this layout, including stubs to access the
7835 PLT, if all input files (including startup and static libraries) were
7836 compiled with @samp{-msecure-plt}. @samp{--bss-plt} forces the old
7837 BSS PLT (and GOT layout) which can give slightly better performance.
7839 @kindex --secure-plt
7841 @command{ld} will use the new PLT and GOT layout if it is linking new
7842 @samp{-fpic} or @samp{-fPIC} code, but does not do so automatically
7843 when linking non-PIC code. This option requests the new PLT and GOT
7844 layout. A warning will be given if some object file requires the old
7850 The new secure PLT and GOT are placed differently relative to other
7851 sections compared to older BSS PLT and GOT placement. The location of
7852 @code{.plt} must change because the new secure PLT is an initialized
7853 section while the old PLT is uninitialized. The reason for the
7854 @code{.got} change is more subtle: The new placement allows
7855 @code{.got} to be read-only in applications linked with
7856 @samp{-z relro -z now}. However, this placement means that
7857 @code{.sdata} cannot always be used in shared libraries, because the
7858 PowerPC ABI accesses @code{.sdata} in shared libraries from the GOT
7859 pointer. @samp{--sdata-got} forces the old GOT placement. PowerPC
7860 GCC doesn't use @code{.sdata} in shared libraries, so this option is
7861 really only useful for other compilers that may do so.
7863 @cindex PowerPC stub symbols
7864 @kindex --emit-stub-syms
7865 @item --emit-stub-syms
7866 This option causes @command{ld} to label linker stubs with a local
7867 symbol that encodes the stub type and destination.
7869 @cindex PowerPC TLS optimization
7870 @kindex --no-tls-optimize
7871 @item --no-tls-optimize
7872 PowerPC @command{ld} normally performs some optimization of code
7873 sequences used to access Thread-Local Storage. Use this option to
7874 disable the optimization.
7887 @node PowerPC64 ELF64
7888 @section @command{ld} and PowerPC64 64-bit ELF Support
7890 @cindex PowerPC64 ELF64 options
7892 @cindex PowerPC64 stub grouping
7893 @kindex --stub-group-size
7894 @item --stub-group-size
7895 Long branch stubs, PLT call stubs and TOC adjusting stubs are placed
7896 by @command{ld} in stub sections located between groups of input sections.
7897 @samp{--stub-group-size} specifies the maximum size of a group of input
7898 sections handled by one stub section. Since branch offsets are signed,
7899 a stub section may serve two groups of input sections, one group before
7900 the stub section, and one group after it. However, when using
7901 conditional branches that require stubs, it may be better (for branch
7902 prediction) that stub sections only serve one group of input sections.
7903 A negative value for @samp{N} chooses this scheme, ensuring that
7904 branches to stubs always use a negative offset. Two special values of
7905 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
7906 @command{ld} to automatically size input section groups for the branch types
7907 detected, with the same behaviour regarding stub placement as other
7908 positive or negative values of @samp{N} respectively.
7910 Note that @samp{--stub-group-size} does not split input sections. A
7911 single input section larger than the group size specified will of course
7912 create a larger group (of one section). If input sections are too
7913 large, it may not be possible for a branch to reach its stub.
7915 @cindex PowerPC64 stub symbols
7916 @kindex --emit-stub-syms
7917 @item --emit-stub-syms
7918 This option causes @command{ld} to label linker stubs with a local
7919 symbol that encodes the stub type and destination.
7921 @cindex PowerPC64 dot symbols
7923 @kindex --no-dotsyms
7926 These two options control how @command{ld} interprets version patterns
7927 in a version script. Older PowerPC64 compilers emitted both a
7928 function descriptor symbol with the same name as the function, and a
7929 code entry symbol with the name prefixed by a dot (@samp{.}). To
7930 properly version a function @samp{foo}, the version script thus needs
7931 to control both @samp{foo} and @samp{.foo}. The option
7932 @samp{--dotsyms}, on by default, automatically adds the required
7933 dot-prefixed patterns. Use @samp{--no-dotsyms} to disable this
7936 @cindex PowerPC64 register save/restore functions
7937 @kindex --save-restore-funcs
7938 @kindex --no-save-restore-funcs
7939 @item --save-restore-funcs
7940 @itemx --no-save-restore-funcs
7941 These two options control whether PowerPC64 @command{ld} automatically
7942 provides out-of-line register save and restore functions used by
7943 @samp{-Os} code. The default is to provide any such referenced
7944 function for a normal final link, and to not do so for a relocatable
7947 @cindex PowerPC64 TLS optimization
7948 @kindex --no-tls-optimize
7949 @item --no-tls-optimize
7950 PowerPC64 @command{ld} normally performs some optimization of code
7951 sequences used to access Thread-Local Storage. Use this option to
7952 disable the optimization.
7954 @cindex PowerPC64 __tls_get_addr optimization
7955 @kindex --tls-get-addr-optimize
7956 @kindex --no-tls-get-addr-optimize
7957 @kindex --tls-get-addr-regsave
7958 @kindex --no-tls-get-addr-regsave
7959 @item --tls-get-addr-optimize
7960 @itemx --no-tls-get-addr-optimize
7961 These options control how PowerPC64 @command{ld} uses a special
7962 stub to call __tls_get_addr. PowerPC64 glibc 2.22 and later support
7963 an optimization that allows the second and subsequent calls to
7964 @code{__tls_get_addr} for a given symbol to be resolved by the special
7965 stub without calling in to glibc. By default the linker enables
7966 generation of the stub when glibc advertises the availability of
7968 Using @option{--tls-get-addr-optimize} with an older glibc won't do
7969 much besides slow down your applications, but may be useful if linking
7970 an application against an older glibc with the expectation that it
7971 will normally be used on systems having a newer glibc.
7972 @option{--tls-get-addr-regsave} forces generation of a stub that saves
7973 and restores volatile registers around the call into glibc. Normally,
7974 this is done when the linker detects a call to __tls_get_addr_desc.
7975 Such calls then go via the register saving stub to __tls_get_addr_opt.
7976 @option{--no-tls-get-addr-regsave} disables generation of the
7979 @cindex PowerPC64 OPD optimization
7980 @kindex --no-opd-optimize
7981 @item --no-opd-optimize
7982 PowerPC64 @command{ld} normally removes @code{.opd} section entries
7983 corresponding to deleted link-once functions, or functions removed by
7984 the action of @samp{--gc-sections} or linker script @code{/DISCARD/}.
7985 Use this option to disable @code{.opd} optimization.
7987 @cindex PowerPC64 OPD spacing
7988 @kindex --non-overlapping-opd
7989 @item --non-overlapping-opd
7990 Some PowerPC64 compilers have an option to generate compressed
7991 @code{.opd} entries spaced 16 bytes apart, overlapping the third word,
7992 the static chain pointer (unused in C) with the first word of the next
7993 entry. This option expands such entries to the full 24 bytes.
7995 @cindex PowerPC64 TOC optimization
7996 @kindex --no-toc-optimize
7997 @item --no-toc-optimize
7998 PowerPC64 @command{ld} normally removes unused @code{.toc} section
7999 entries. Such entries are detected by examining relocations that
8000 reference the TOC in code sections. A reloc in a deleted code section
8001 marks a TOC word as unneeded, while a reloc in a kept code section
8002 marks a TOC word as needed. Since the TOC may reference itself, TOC
8003 relocs are also examined. TOC words marked as both needed and
8004 unneeded will of course be kept. TOC words without any referencing
8005 reloc are assumed to be part of a multi-word entry, and are kept or
8006 discarded as per the nearest marked preceding word. This works
8007 reliably for compiler generated code, but may be incorrect if assembly
8008 code is used to insert TOC entries. Use this option to disable the
8011 @cindex PowerPC64 inline PLT call optimization
8012 @kindex --no-inline-optimize
8013 @item --no-inline-optimize
8014 PowerPC64 @command{ld} normally replaces inline PLT call sequences
8015 marked with @code{R_PPC64_PLTSEQ}, @code{R_PPC64_PLTCALL},
8016 @code{R_PPC64_PLT16_HA} and @code{R_PPC64_PLT16_LO_DS} relocations by
8017 a number of @code{nop}s and a direct call when the function is defined
8018 locally and can't be overridden by some other definition. This option
8019 disables that optimization.
8021 @cindex PowerPC64 multi-TOC
8022 @kindex --no-multi-toc
8023 @item --no-multi-toc
8024 If given any toc option besides @code{-mcmodel=medium} or
8025 @code{-mcmodel=large}, PowerPC64 GCC generates code for a TOC model
8027 entries are accessed with a 16-bit offset from r2. This limits the
8028 total TOC size to 64K. PowerPC64 @command{ld} extends this limit by
8029 grouping code sections such that each group uses less than 64K for its
8030 TOC entries, then inserts r2 adjusting stubs between inter-group
8031 calls. @command{ld} does not split apart input sections, so cannot
8032 help if a single input file has a @code{.toc} section that exceeds
8033 64K, most likely from linking multiple files with @command{ld -r}.
8034 Use this option to turn off this feature.
8036 @cindex PowerPC64 TOC sorting
8037 @kindex --no-toc-sort
8039 By default, @command{ld} sorts TOC sections so that those whose file
8040 happens to have a section called @code{.init} or @code{.fini} are
8041 placed first, followed by TOC sections referenced by code generated
8042 with PowerPC64 gcc's @code{-mcmodel=small}, and lastly TOC sections
8043 referenced only by code generated with PowerPC64 gcc's
8044 @code{-mcmodel=medium} or @code{-mcmodel=large} options. Doing this
8045 results in better TOC grouping for multi-TOC. Use this option to turn
8048 @cindex PowerPC64 PLT stub alignment
8050 @kindex --no-plt-align
8052 @itemx --no-plt-align
8053 Use these options to control whether individual PLT call stubs are
8054 aligned to a 32-byte boundary, or to the specified power of two
8055 boundary when using @code{--plt-align=}. A negative value may be
8056 specified to pad PLT call stubs so that they do not cross the
8057 specified power of two boundary (or the minimum number of boundaries
8058 if a PLT stub is so large that it must cross a boundary). By default
8059 PLT call stubs are aligned to 32-byte boundaries.
8061 @cindex PowerPC64 PLT call stub static chain
8062 @kindex --plt-static-chain
8063 @kindex --no-plt-static-chain
8064 @item --plt-static-chain
8065 @itemx --no-plt-static-chain
8066 Use these options to control whether PLT call stubs load the static
8067 chain pointer (r11). @code{ld} defaults to not loading the static
8068 chain since there is never any need to do so on a PLT call.
8070 @cindex PowerPC64 PLT call stub thread safety
8071 @kindex --plt-thread-safe
8072 @kindex --no-plt-thread-safe
8073 @item --plt-thread-safe
8074 @itemx --no-plt-thread-safe
8075 With power7's weakly ordered memory model, it is possible when using
8076 lazy binding for ld.so to update a plt entry in one thread and have
8077 another thread see the individual plt entry words update in the wrong
8078 order, despite ld.so carefully writing in the correct order and using
8079 memory write barriers. To avoid this we need some sort of read
8080 barrier in the call stub, or use LD_BIND_NOW=1. By default, @code{ld}
8081 looks for calls to commonly used functions that create threads, and if
8082 seen, adds the necessary barriers. Use these options to change the
8085 @cindex PowerPC64 ELFv2 PLT localentry optimization
8086 @kindex --plt-localentry
8087 @kindex --no-plt-localentry
8088 @item --plt-localentry
8089 @itemx --no-localentry
8090 ELFv2 functions with localentry:0 are those with a single entry point,
8091 ie. global entry == local entry, and that have no requirement on r2
8092 (the TOC/GOT pointer) or r12, and guarantee r2 is unchanged on return.
8093 Such an external function can be called via the PLT without saving r2
8094 or restoring it on return, avoiding a common load-hit-store for small
8095 functions. The optimization is attractive, with up to 40% reduction
8096 in execution time for a small function, but can result in symbol
8097 interposition failures. Also, minor changes in a shared library,
8098 including system libraries, can cause a function that was localentry:0
8099 to become localentry:8. This will result in a dynamic loader
8100 complaint and failure to run. The option is experimental, use with
8101 care. @option{--no-plt-localentry} is the default.
8103 @cindex PowerPC64 Power10 stubs
8104 @kindex --power10-stubs
8105 @kindex --no-power10-stubs
8106 @item --power10-stubs
8107 @itemx --no-power10-stubs
8108 When PowerPC64 @command{ld} links input object files containing
8109 relocations used on power10 prefixed instructions it normally creates
8110 linkage stubs (PLT call and long branch) using power10 instructions
8111 for @code{@@notoc} PLT calls where @code{r2} is not known. The
8112 power10 notoc stubs are smaller and faster, so are preferred for
8113 power10. @option{--power10-stubs} and @option{--no-power10-stubs}
8114 allow you to override the linker's selection of stub instructions.
8115 @option{--power10-stubs=auto} allows the user to select the default
8130 @section @command{ld} and S/390 ELF Support
8132 @cindex S/390 ELF options
8136 @kindex --s390-pgste
8138 This option marks the result file with a @code{PT_S390_PGSTE}
8139 segment. The Linux kernel is supposed to allocate 4k page tables for
8140 binaries marked that way.
8154 @section @command{ld} and SPU ELF Support
8156 @cindex SPU ELF options
8162 This option marks an executable as a PIC plugin module.
8164 @cindex SPU overlays
8165 @kindex --no-overlays
8167 Normally, @command{ld} recognizes calls to functions within overlay
8168 regions, and redirects such calls to an overlay manager via a stub.
8169 @command{ld} also provides a built-in overlay manager. This option
8170 turns off all this special overlay handling.
8172 @cindex SPU overlay stub symbols
8173 @kindex --emit-stub-syms
8174 @item --emit-stub-syms
8175 This option causes @command{ld} to label overlay stubs with a local
8176 symbol that encodes the stub type and destination.
8178 @cindex SPU extra overlay stubs
8179 @kindex --extra-overlay-stubs
8180 @item --extra-overlay-stubs
8181 This option causes @command{ld} to add overlay call stubs on all
8182 function calls out of overlay regions. Normally stubs are not added
8183 on calls to non-overlay regions.
8185 @cindex SPU local store size
8186 @kindex --local-store=lo:hi
8187 @item --local-store=lo:hi
8188 @command{ld} usually checks that a final executable for SPU fits in
8189 the address range 0 to 256k. This option may be used to change the
8190 range. Disable the check entirely with @option{--local-store=0:0}.
8193 @kindex --stack-analysis
8194 @item --stack-analysis
8195 SPU local store space is limited. Over-allocation of stack space
8196 unnecessarily limits space available for code and data, while
8197 under-allocation results in runtime failures. If given this option,
8198 @command{ld} will provide an estimate of maximum stack usage.
8199 @command{ld} does this by examining symbols in code sections to
8200 determine the extents of functions, and looking at function prologues
8201 for stack adjusting instructions. A call-graph is created by looking
8202 for relocations on branch instructions. The graph is then searched
8203 for the maximum stack usage path. Note that this analysis does not
8204 find calls made via function pointers, and does not handle recursion
8205 and other cycles in the call graph. Stack usage may be
8206 under-estimated if your code makes such calls. Also, stack usage for
8207 dynamic allocation, e.g. alloca, will not be detected. If a link map
8208 is requested, detailed information about each function's stack usage
8209 and calls will be given.
8212 @kindex --emit-stack-syms
8213 @item --emit-stack-syms
8214 This option, if given along with @option{--stack-analysis} will result
8215 in @command{ld} emitting stack sizing symbols for each function.
8216 These take the form @code{__stack_<function_name>} for global
8217 functions, and @code{__stack_<number>_<function_name>} for static
8218 functions. @code{<number>} is the section id in hex. The value of
8219 such symbols is the stack requirement for the corresponding function.
8220 The symbol size will be zero, type @code{STT_NOTYPE}, binding
8221 @code{STB_LOCAL}, and section @code{SHN_ABS}.
8235 @section @command{ld}'s Support for Various TI COFF Versions
8236 @cindex TI COFF versions
8237 @kindex --format=@var{version}
8238 The @samp{--format} switch allows selection of one of the various
8239 TI COFF versions. The latest of this writing is 2; versions 0 and 1 are
8240 also supported. The TI COFF versions also vary in header byte-order
8241 format; @command{ld} will read any version or byte order, but the output
8242 header format depends on the default specified by the specific target.
8255 @section @command{ld} and WIN32 (cygwin/mingw)
8257 This section describes some of the win32 specific @command{ld} issues.
8258 See @ref{Options,,Command-line Options} for detailed description of the
8259 command-line options mentioned here.
8262 @cindex import libraries
8263 @item import libraries
8264 The standard Windows linker creates and uses so-called import
8265 libraries, which contains information for linking to dll's. They are
8266 regular static archives and are handled as any other static
8267 archive. The cygwin and mingw ports of @command{ld} have specific
8268 support for creating such libraries provided with the
8269 @samp{--out-implib} command-line option.
8271 @item exporting DLL symbols
8272 @cindex exporting DLL symbols
8273 The cygwin/mingw @command{ld} has several ways to export symbols for dll's.
8276 @item using auto-export functionality
8277 @cindex using auto-export functionality
8278 By default @command{ld} exports symbols with the auto-export functionality,
8279 which is controlled by the following command-line options:
8282 @item --export-all-symbols [This is the default]
8283 @item --exclude-symbols
8284 @item --exclude-libs
8285 @item --exclude-modules-for-implib
8286 @item --version-script
8289 When auto-export is in operation, @command{ld} will export all the non-local
8290 (global and common) symbols it finds in a DLL, with the exception of a few
8291 symbols known to belong to the system's runtime and libraries. As it will
8292 often not be desirable to export all of a DLL's symbols, which may include
8293 private functions that are not part of any public interface, the command-line
8294 options listed above may be used to filter symbols out from the list for
8295 exporting. The @samp{--output-def} option can be used in order to see the
8296 final list of exported symbols with all exclusions taken into effect.
8298 If @samp{--export-all-symbols} is not given explicitly on the
8299 command line, then the default auto-export behavior will be @emph{disabled}
8300 if either of the following are true:
8303 @item A DEF file is used.
8304 @item Any symbol in any object file was marked with the __declspec(dllexport) attribute.
8307 @item using a DEF file
8308 @cindex using a DEF file
8309 Another way of exporting symbols is using a DEF file. A DEF file is
8310 an ASCII file containing definitions of symbols which should be
8311 exported when a dll is created. Usually it is named @samp{<dll
8312 name>.def} and is added as any other object file to the linker's
8313 command line. The file's name must end in @samp{.def} or @samp{.DEF}.
8316 gcc -o <output> <objectfiles> <dll name>.def
8319 Using a DEF file turns off the normal auto-export behavior, unless the
8320 @samp{--export-all-symbols} option is also used.
8322 Here is an example of a DEF file for a shared library called @samp{xyz.dll}:
8325 LIBRARY "xyz.dll" BASE=0x20000000
8331 another_foo = abc.dll.afoo
8337 This example defines a DLL with a non-default base address and seven
8338 symbols in the export table. The third exported symbol @code{_bar} is an
8339 alias for the second. The fourth symbol, @code{another_foo} is resolved
8340 by "forwarding" to another module and treating it as an alias for
8341 @code{afoo} exported from the DLL @samp{abc.dll}. The final symbol
8342 @code{var1} is declared to be a data object. The @samp{doo} symbol in
8343 export library is an alias of @samp{foo}, which gets the string name
8344 in export table @samp{foo2}. The @samp{eoo} symbol is an data export
8345 symbol, which gets in export table the name @samp{var1}.
8347 The optional @code{LIBRARY <name>} command indicates the @emph{internal}
8348 name of the output DLL. If @samp{<name>} does not include a suffix,
8349 the default library suffix, @samp{.DLL} is appended.
8351 When the .DEF file is used to build an application, rather than a
8352 library, the @code{NAME <name>} command should be used instead of
8353 @code{LIBRARY}. If @samp{<name>} does not include a suffix, the default
8354 executable suffix, @samp{.EXE} is appended.
8356 With either @code{LIBRARY <name>} or @code{NAME <name>} the optional
8357 specification @code{BASE = <number>} may be used to specify a
8358 non-default base address for the image.
8360 If neither @code{LIBRARY <name>} nor @code{NAME <name>} is specified,
8361 or they specify an empty string, the internal name is the same as the
8362 filename specified on the command line.
8364 The complete specification of an export symbol is:
8368 ( ( ( <name1> [ = <name2> ] )
8369 | ( <name1> = <module-name> . <external-name>))
8370 [ @@ <integer> ] [NONAME] [DATA] [CONSTANT] [PRIVATE] [== <name3>] ) *
8373 Declares @samp{<name1>} as an exported symbol from the DLL, or declares
8374 @samp{<name1>} as an exported alias for @samp{<name2>}; or declares
8375 @samp{<name1>} as a "forward" alias for the symbol
8376 @samp{<external-name>} in the DLL @samp{<module-name>}.
8377 Optionally, the symbol may be exported by the specified ordinal
8378 @samp{<integer>} alias. The optional @samp{<name3>} is the to be used
8379 string in import/export table for the symbol.
8381 The optional keywords that follow the declaration indicate:
8383 @code{NONAME}: Do not put the symbol name in the DLL's export table. It
8384 will still be exported by its ordinal alias (either the value specified
8385 by the .def specification or, otherwise, the value assigned by the
8386 linker). The symbol name, however, does remain visible in the import
8387 library (if any), unless @code{PRIVATE} is also specified.
8389 @code{DATA}: The symbol is a variable or object, rather than a function.
8390 The import lib will export only an indirect reference to @code{foo} as
8391 the symbol @code{_imp__foo} (ie, @code{foo} must be resolved as
8394 @code{CONSTANT}: Like @code{DATA}, but put the undecorated @code{foo} as
8395 well as @code{_imp__foo} into the import library. Both refer to the
8396 read-only import address table's pointer to the variable, not to the
8397 variable itself. This can be dangerous. If the user code fails to add
8398 the @code{dllimport} attribute and also fails to explicitly add the
8399 extra indirection that the use of the attribute enforces, the
8400 application will behave unexpectedly.
8402 @code{PRIVATE}: Put the symbol in the DLL's export table, but do not put
8403 it into the static import library used to resolve imports at link time. The
8404 symbol can still be imported using the @code{LoadLibrary/GetProcAddress}
8405 API at runtime or by using the GNU ld extension of linking directly to
8406 the DLL without an import library.
8408 See ld/deffilep.y in the binutils sources for the full specification of
8409 other DEF file statements
8411 @cindex creating a DEF file
8412 While linking a shared dll, @command{ld} is able to create a DEF file
8413 with the @samp{--output-def <file>} command-line option.
8415 @item Using decorations
8416 @cindex Using decorations
8417 Another way of marking symbols for export is to modify the source code
8418 itself, so that when building the DLL each symbol to be exported is
8422 __declspec(dllexport) int a_variable
8423 __declspec(dllexport) void a_function(int with_args)
8426 All such symbols will be exported from the DLL. If, however,
8427 any of the object files in the DLL contain symbols decorated in
8428 this way, then the normal auto-export behavior is disabled, unless
8429 the @samp{--export-all-symbols} option is also used.
8431 Note that object files that wish to access these symbols must @emph{not}
8432 decorate them with dllexport. Instead, they should use dllimport,
8436 __declspec(dllimport) int a_variable
8437 __declspec(dllimport) void a_function(int with_args)
8440 This complicates the structure of library header files, because
8441 when included by the library itself the header must declare the
8442 variables and functions as dllexport, but when included by client
8443 code the header must declare them as dllimport. There are a number
8444 of idioms that are typically used to do this; often client code can
8445 omit the __declspec() declaration completely. See
8446 @samp{--enable-auto-import} and @samp{automatic data imports} for more
8450 @cindex automatic data imports
8451 @item automatic data imports
8452 The standard Windows dll format supports data imports from dlls only
8453 by adding special decorations (dllimport/dllexport), which let the
8454 compiler produce specific assembler instructions to deal with this
8455 issue. This increases the effort necessary to port existing Un*x
8456 code to these platforms, especially for large
8457 c++ libraries and applications. The auto-import feature, which was
8458 initially provided by Paul Sokolovsky, allows one to omit the
8459 decorations to achieve a behavior that conforms to that on POSIX/Un*x
8460 platforms. This feature is enabled with the @samp{--enable-auto-import}
8461 command-line option, although it is enabled by default on cygwin/mingw.
8462 The @samp{--enable-auto-import} option itself now serves mainly to
8463 suppress any warnings that are ordinarily emitted when linked objects
8464 trigger the feature's use.
8466 auto-import of variables does not always work flawlessly without
8467 additional assistance. Sometimes, you will see this message
8469 "variable '<var>' can't be auto-imported. Please read the
8470 documentation for ld's @code{--enable-auto-import} for details."
8472 The @samp{--enable-auto-import} documentation explains why this error
8473 occurs, and several methods that can be used to overcome this difficulty.
8474 One of these methods is the @emph{runtime pseudo-relocs} feature, described
8477 @cindex runtime pseudo-relocation
8478 For complex variables imported from DLLs (such as structs or classes),
8479 object files typically contain a base address for the variable and an
8480 offset (@emph{addend}) within the variable--to specify a particular
8481 field or public member, for instance. Unfortunately, the runtime loader used
8482 in win32 environments is incapable of fixing these references at runtime
8483 without the additional information supplied by dllimport/dllexport decorations.
8484 The standard auto-import feature described above is unable to resolve these
8487 The @samp{--enable-runtime-pseudo-relocs} switch allows these references to
8488 be resolved without error, while leaving the task of adjusting the references
8489 themselves (with their non-zero addends) to specialized code provided by the
8490 runtime environment. Recent versions of the cygwin and mingw environments and
8491 compilers provide this runtime support; older versions do not. However, the
8492 support is only necessary on the developer's platform; the compiled result will
8493 run without error on an older system.
8495 @samp{--enable-runtime-pseudo-relocs} is not the default; it must be explicitly
8498 @cindex direct linking to a dll
8499 @item direct linking to a dll
8500 The cygwin/mingw ports of @command{ld} support the direct linking,
8501 including data symbols, to a dll without the usage of any import
8502 libraries. This is much faster and uses much less memory than does the
8503 traditional import library method, especially when linking large
8504 libraries or applications. When @command{ld} creates an import lib, each
8505 function or variable exported from the dll is stored in its own bfd, even
8506 though a single bfd could contain many exports. The overhead involved in
8507 storing, loading, and processing so many bfd's is quite large, and explains the
8508 tremendous time, memory, and storage needed to link against particularly
8509 large or complex libraries when using import libs.
8511 Linking directly to a dll uses no extra command-line switches other than
8512 @samp{-L} and @samp{-l}, because @command{ld} already searches for a number
8513 of names to match each library. All that is needed from the developer's
8514 perspective is an understanding of this search, in order to force ld to
8515 select the dll instead of an import library.
8518 For instance, when ld is called with the argument @samp{-lxxx} it will attempt
8519 to find, in the first directory of its search path,
8532 before moving on to the next directory in the search path.
8534 (*) Actually, this is not @samp{cygxxx.dll} but in fact is @samp{<prefix>xxx.dll},
8535 where @samp{<prefix>} is set by the @command{ld} option
8536 @samp{--dll-search-prefix=<prefix>}. In the case of cygwin, the standard gcc spec
8537 file includes @samp{--dll-search-prefix=cyg}, so in effect we actually search for
8540 Other win32-based unix environments, such as mingw or pw32, may use other
8541 @samp{<prefix>}es, although at present only cygwin makes use of this feature. It
8542 was originally intended to help avoid name conflicts among dll's built for the
8543 various win32/un*x environments, so that (for example) two versions of a zlib dll
8544 could coexist on the same machine.
8546 The generic cygwin/mingw path layout uses a @samp{bin} directory for
8547 applications and dll's and a @samp{lib} directory for the import
8548 libraries (using cygwin nomenclature):
8554 libxxx.dll.a (in case of dll's)
8555 libxxx.a (in case of static archive)
8558 Linking directly to a dll without using the import library can be
8561 1. Use the dll directly by adding the @samp{bin} path to the link line
8563 gcc -Wl,-verbose -o a.exe -L../bin/ -lxxx
8566 However, as the dll's often have version numbers appended to their names
8567 (@samp{cygncurses-5.dll}) this will often fail, unless one specifies
8568 @samp{-L../bin -lncurses-5} to include the version. Import libs are generally
8569 not versioned, and do not have this difficulty.
8571 2. Create a symbolic link from the dll to a file in the @samp{lib}
8572 directory according to the above mentioned search pattern. This
8573 should be used to avoid unwanted changes in the tools needed for
8577 ln -s bin/cygxxx.dll lib/[cyg|lib|]xxx.dll[.a]
8580 Then you can link without any make environment changes.
8583 gcc -Wl,-verbose -o a.exe -L../lib/ -lxxx
8586 This technique also avoids the version number problems, because the following is
8593 libxxx.dll.a -> ../bin/cygxxx-5.dll
8596 Linking directly to a dll without using an import lib will work
8597 even when auto-import features are exercised, and even when
8598 @samp{--enable-runtime-pseudo-relocs} is used.
8600 Given the improvements in speed and memory usage, one might justifiably
8601 wonder why import libraries are used at all. There are three reasons:
8603 1. Until recently, the link-directly-to-dll functionality did @emph{not}
8604 work with auto-imported data.
8606 2. Sometimes it is necessary to include pure static objects within the
8607 import library (which otherwise contains only bfd's for indirection
8608 symbols that point to the exports of a dll). Again, the import lib
8609 for the cygwin kernel makes use of this ability, and it is not
8610 possible to do this without an import lib.
8612 3. Symbol aliases can only be resolved using an import lib. This is
8613 critical when linking against OS-supplied dll's (eg, the win32 API)
8614 in which symbols are usually exported as undecorated aliases of their
8615 stdcall-decorated assembly names.
8617 So, import libs are not going away. But the ability to replace
8618 true import libs with a simple symbolic link to (or a copy of)
8619 a dll, in many cases, is a useful addition to the suite of tools
8620 binutils makes available to the win32 developer. Given the
8621 massive improvements in memory requirements during linking, storage
8622 requirements, and linking speed, we expect that many developers
8623 will soon begin to use this feature whenever possible.
8625 @item symbol aliasing
8627 @item adding additional names
8628 Sometimes, it is useful to export symbols with additional names.
8629 A symbol @samp{foo} will be exported as @samp{foo}, but it can also be
8630 exported as @samp{_foo} by using special directives in the DEF file
8631 when creating the dll. This will affect also the optional created
8632 import library. Consider the following DEF file:
8635 LIBRARY "xyz.dll" BASE=0x61000000
8642 The line @samp{_foo = foo} maps the symbol @samp{foo} to @samp{_foo}.
8644 Another method for creating a symbol alias is to create it in the
8645 source code using the "weak" attribute:
8648 void foo () @{ /* Do something. */; @}
8649 void _foo () __attribute__ ((weak, alias ("foo")));
8652 See the gcc manual for more information about attributes and weak
8655 @item renaming symbols
8656 Sometimes it is useful to rename exports. For instance, the cygwin
8657 kernel does this regularly. A symbol @samp{_foo} can be exported as
8658 @samp{foo} but not as @samp{_foo} by using special directives in the
8659 DEF file. (This will also affect the import library, if it is
8660 created). In the following example:
8663 LIBRARY "xyz.dll" BASE=0x61000000
8669 The line @samp{_foo = foo} maps the exported symbol @samp{foo} to
8673 Note: using a DEF file disables the default auto-export behavior,
8674 unless the @samp{--export-all-symbols} command-line option is used.
8675 If, however, you are trying to rename symbols, then you should list
8676 @emph{all} desired exports in the DEF file, including the symbols
8677 that are not being renamed, and do @emph{not} use the
8678 @samp{--export-all-symbols} option. If you list only the
8679 renamed symbols in the DEF file, and use @samp{--export-all-symbols}
8680 to handle the other symbols, then the both the new names @emph{and}
8681 the original names for the renamed symbols will be exported.
8682 In effect, you'd be aliasing those symbols, not renaming them,
8683 which is probably not what you wanted.
8685 @cindex weak externals
8686 @item weak externals
8687 The Windows object format, PE, specifies a form of weak symbols called
8688 weak externals. When a weak symbol is linked and the symbol is not
8689 defined, the weak symbol becomes an alias for some other symbol. There
8690 are three variants of weak externals:
8692 @item Definition is searched for in objects and libraries, historically
8693 called lazy externals.
8694 @item Definition is searched for only in other objects, not in libraries.
8695 This form is not presently implemented.
8696 @item No search; the symbol is an alias. This form is not presently
8699 As a GNU extension, weak symbols that do not specify an alternate symbol
8700 are supported. If the symbol is undefined when linking, the symbol
8701 uses a default value.
8703 @cindex aligned common symbols
8704 @item aligned common symbols
8705 As a GNU extension to the PE file format, it is possible to specify the
8706 desired alignment for a common symbol. This information is conveyed from
8707 the assembler or compiler to the linker by means of GNU-specific commands
8708 carried in the object file's @samp{.drectve} section, which are recognized
8709 by @command{ld} and respected when laying out the common symbols. Native
8710 tools will be able to process object files employing this GNU extension,
8711 but will fail to respect the alignment instructions, and may issue noisy
8712 warnings about unknown linker directives.
8727 @section @code{ld} and Xtensa Processors
8729 @cindex Xtensa processors
8730 The default @command{ld} behavior for Xtensa processors is to interpret
8731 @code{SECTIONS} commands so that lists of explicitly named sections in a
8732 specification with a wildcard file will be interleaved when necessary to
8733 keep literal pools within the range of PC-relative load offsets. For
8734 example, with the command:
8746 @command{ld} may interleave some of the @code{.literal}
8747 and @code{.text} sections from different object files to ensure that the
8748 literal pools are within the range of PC-relative load offsets. A valid
8749 interleaving might place the @code{.literal} sections from an initial
8750 group of files followed by the @code{.text} sections of that group of
8751 files. Then, the @code{.literal} sections from the rest of the files
8752 and the @code{.text} sections from the rest of the files would follow.
8754 @cindex @option{--relax} on Xtensa
8755 @cindex relaxing on Xtensa
8756 Relaxation is enabled by default for the Xtensa version of @command{ld} and
8757 provides two important link-time optimizations. The first optimization
8758 is to combine identical literal values to reduce code size. A redundant
8759 literal will be removed and all the @code{L32R} instructions that use it
8760 will be changed to reference an identical literal, as long as the
8761 location of the replacement literal is within the offset range of all
8762 the @code{L32R} instructions. The second optimization is to remove
8763 unnecessary overhead from assembler-generated ``longcall'' sequences of
8764 @code{L32R}/@code{CALLX@var{n}} when the target functions are within
8765 range of direct @code{CALL@var{n}} instructions.
8767 For each of these cases where an indirect call sequence can be optimized
8768 to a direct call, the linker will change the @code{CALLX@var{n}}
8769 instruction to a @code{CALL@var{n}} instruction, remove the @code{L32R}
8770 instruction, and remove the literal referenced by the @code{L32R}
8771 instruction if it is not used for anything else. Removing the
8772 @code{L32R} instruction always reduces code size but can potentially
8773 hurt performance by changing the alignment of subsequent branch targets.
8774 By default, the linker will always preserve alignments, either by
8775 switching some instructions between 24-bit encodings and the equivalent
8776 density instructions or by inserting a no-op in place of the @code{L32R}
8777 instruction that was removed. If code size is more important than
8778 performance, the @option{--size-opt} option can be used to prevent the
8779 linker from widening density instructions or inserting no-ops, except in
8780 a few cases where no-ops are required for correctness.
8782 The following Xtensa-specific command-line options can be used to
8785 @cindex Xtensa options
8788 When optimizing indirect calls to direct calls, optimize for code size
8789 more than performance. With this option, the linker will not insert
8790 no-ops or widen density instructions to preserve branch target
8791 alignment. There may still be some cases where no-ops are required to
8792 preserve the correctness of the code.
8794 @item --abi-windowed
8796 Choose ABI for the output object and for the generated PLT code.
8797 PLT code inserted by the linker must match ABI of the output object
8798 because windowed and call0 ABI use incompatible function call
8800 Default ABI is chosen by the ABI tag in the @code{.xtensa.info} section
8801 of the first input object.
8802 A warning is issued if ABI tags of input objects do not match each other
8803 or the chosen output object ABI.
8811 @ifclear SingleFormat
8816 @cindex object file management
8817 @cindex object formats available
8819 The linker accesses object and archive files using the BFD libraries.
8820 These libraries allow the linker to use the same routines to operate on
8821 object files whatever the object file format. A different object file
8822 format can be supported simply by creating a new BFD back end and adding
8823 it to the library. To conserve runtime memory, however, the linker and
8824 associated tools are usually configured to support only a subset of the
8825 object file formats available. You can use @code{objdump -i}
8826 (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to
8827 list all the formats available for your configuration.
8829 @cindex BFD requirements
8830 @cindex requirements for BFD
8831 As with most implementations, BFD is a compromise between
8832 several conflicting requirements. The major factor influencing
8833 BFD design was efficiency: any time used converting between
8834 formats is time which would not have been spent had BFD not
8835 been involved. This is partly offset by abstraction payback; since
8836 BFD simplifies applications and back ends, more time and care
8837 may be spent optimizing algorithms for a greater speed.
8839 One minor artifact of the BFD solution which you should bear in
8840 mind is the potential for information loss. There are two places where
8841 useful information can be lost using the BFD mechanism: during
8842 conversion and during output. @xref{BFD information loss}.
8845 * BFD outline:: How it works: an outline of BFD
8849 @section How It Works: An Outline of BFD
8850 @cindex opening object files
8851 @include bfdsumm.texi
8854 @node Reporting Bugs
8855 @chapter Reporting Bugs
8856 @cindex bugs in @command{ld}
8857 @cindex reporting bugs in @command{ld}
8859 Your bug reports play an essential role in making @command{ld} reliable.
8861 Reporting a bug may help you by bringing a solution to your problem, or
8862 it may not. But in any case the principal function of a bug report is
8863 to help the entire community by making the next version of @command{ld}
8864 work better. Bug reports are your contribution to the maintenance of
8867 In order for a bug report to serve its purpose, you must include the
8868 information that enables us to fix the bug.
8871 * Bug Criteria:: Have you found a bug?
8872 * Bug Reporting:: How to report bugs
8876 @section Have You Found a Bug?
8877 @cindex bug criteria
8879 If you are not sure whether you have found a bug, here are some guidelines:
8882 @cindex fatal signal
8883 @cindex linker crash
8884 @cindex crash of linker
8886 If the linker gets a fatal signal, for any input whatever, that is a
8887 @command{ld} bug. Reliable linkers never crash.
8889 @cindex error on valid input
8891 If @command{ld} produces an error message for valid input, that is a bug.
8893 @cindex invalid input
8895 If @command{ld} does not produce an error message for invalid input, that
8896 may be a bug. In the general case, the linker can not verify that
8897 object files are correct.
8900 If you are an experienced user of linkers, your suggestions for
8901 improvement of @command{ld} are welcome in any case.
8905 @section How to Report Bugs
8907 @cindex @command{ld} bugs, reporting
8909 A number of companies and individuals offer support for @sc{gnu}
8910 products. If you obtained @command{ld} from a support organization, we
8911 recommend you contact that organization first.
8913 You can find contact information for many support companies and
8914 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
8918 Otherwise, send bug reports for @command{ld} to
8922 The fundamental principle of reporting bugs usefully is this:
8923 @strong{report all the facts}. If you are not sure whether to state a
8924 fact or leave it out, state it!
8926 Often people omit facts because they think they know what causes the
8927 problem and assume that some details do not matter. Thus, you might
8928 assume that the name of a symbol you use in an example does not
8929 matter. Well, probably it does not, but one cannot be sure. Perhaps
8930 the bug is a stray memory reference which happens to fetch from the
8931 location where that name is stored in memory; perhaps, if the name
8932 were different, the contents of that location would fool the linker
8933 into doing the right thing despite the bug. Play it safe and give a
8934 specific, complete example. That is the easiest thing for you to do,
8935 and the most helpful.
8937 Keep in mind that the purpose of a bug report is to enable us to fix
8938 the bug if it is new to us. Therefore, always write your bug reports
8939 on the assumption that the bug has not been reported previously.
8941 Sometimes people give a few sketchy facts and ask, ``Does this ring a
8942 bell?'' This cannot help us fix a bug, so it is basically useless. We
8943 respond by asking for enough details to enable us to investigate.
8944 You might as well expedite matters by sending them to begin with.
8946 To enable us to fix the bug, you should include all these things:
8950 The version of @command{ld}. @command{ld} announces it if you start it with
8951 the @samp{--version} argument.
8953 Without this, we will not know whether there is any point in looking for
8954 the bug in the current version of @command{ld}.
8957 Any patches you may have applied to the @command{ld} source, including any
8958 patches made to the @code{BFD} library.
8961 The type of machine you are using, and the operating system name and
8965 What compiler (and its version) was used to compile @command{ld}---e.g.
8969 The command arguments you gave the linker to link your example and
8970 observe the bug. To guarantee you will not omit something important,
8971 list them all. A copy of the Makefile (or the output from make) is
8974 If we were to try to guess the arguments, we would probably guess wrong
8975 and then we might not encounter the bug.
8978 A complete input file, or set of input files, that will reproduce the
8979 bug. It is generally most helpful to send the actual object files
8980 provided that they are reasonably small. Say no more than 10K. For
8981 bigger files you can either make them available by FTP or HTTP or else
8982 state that you are willing to send the object file(s) to whomever
8983 requests them. (Note - your email will be going to a mailing list, so
8984 we do not want to clog it up with large attachments). But small
8985 attachments are best.
8987 If the source files were assembled using @code{gas} or compiled using
8988 @code{gcc}, then it may be OK to send the source files rather than the
8989 object files. In this case, be sure to say exactly what version of
8990 @code{gas} or @code{gcc} was used to produce the object files. Also say
8991 how @code{gas} or @code{gcc} were configured.
8994 A description of what behavior you observe that you believe is
8995 incorrect. For example, ``It gets a fatal signal.''
8997 Of course, if the bug is that @command{ld} gets a fatal signal, then we
8998 will certainly notice it. But if the bug is incorrect output, we might
8999 not notice unless it is glaringly wrong. You might as well not give us
9000 a chance to make a mistake.
9002 Even if the problem you experience is a fatal signal, you should still
9003 say so explicitly. Suppose something strange is going on, such as, your
9004 copy of @command{ld} is out of sync, or you have encountered a bug in the
9005 C library on your system. (This has happened!) Your copy might crash
9006 and ours would not. If you told us to expect a crash, then when ours
9007 fails to crash, we would know that the bug was not happening for us. If
9008 you had not told us to expect a crash, then we would not be able to draw
9009 any conclusion from our observations.
9012 If you wish to suggest changes to the @command{ld} source, send us context
9013 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or
9014 @samp{-p} option. Always send diffs from the old file to the new file.
9015 If you even discuss something in the @command{ld} source, refer to it by
9016 context, not by line number.
9018 The line numbers in our development sources will not match those in your
9019 sources. Your line numbers would convey no useful information to us.
9022 Here are some things that are not necessary:
9026 A description of the envelope of the bug.
9028 Often people who encounter a bug spend a lot of time investigating
9029 which changes to the input file will make the bug go away and which
9030 changes will not affect it.
9032 This is often time consuming and not very useful, because the way we
9033 will find the bug is by running a single example under the debugger
9034 with breakpoints, not by pure deduction from a series of examples.
9035 We recommend that you save your time for something else.
9037 Of course, if you can find a simpler example to report @emph{instead}
9038 of the original one, that is a convenience for us. Errors in the
9039 output will be easier to spot, running under the debugger will take
9040 less time, and so on.
9042 However, simplification is not vital; if you do not want to do this,
9043 report the bug anyway and send us the entire test case you used.
9046 A patch for the bug.
9048 A patch for the bug does help us if it is a good one. But do not omit
9049 the necessary information, such as the test case, on the assumption that
9050 a patch is all we need. We might see problems with your patch and decide
9051 to fix the problem another way, or we might not understand it at all.
9053 Sometimes with a program as complicated as @command{ld} it is very hard to
9054 construct an example that will make the program follow a certain path
9055 through the code. If you do not send us the example, we will not be
9056 able to construct one, so we will not be able to verify that the bug is
9059 And if we cannot understand what bug you are trying to fix, or why your
9060 patch should be an improvement, we will not install it. A test case will
9061 help us to understand.
9064 A guess about what the bug is or what it depends on.
9066 Such guesses are usually wrong. Even we cannot guess right about such
9067 things without first using the debugger to find the facts.
9071 @appendix MRI Compatible Script Files
9072 @cindex MRI compatibility
9073 To aid users making the transition to @sc{gnu} @command{ld} from the MRI
9074 linker, @command{ld} can use MRI compatible linker scripts as an
9075 alternative to the more general-purpose linker scripting language
9076 described in @ref{Scripts}. MRI compatible linker scripts have a much
9077 simpler command set than the scripting language otherwise used with
9078 @command{ld}. @sc{gnu} @command{ld} supports the most commonly used MRI
9079 linker commands; these commands are described here.
9081 In general, MRI scripts aren't of much use with the @code{a.out} object
9082 file format, since it only has three sections and MRI scripts lack some
9083 features to make use of them.
9085 You can specify a file containing an MRI-compatible script using the
9086 @samp{-c} command-line option.
9088 Each command in an MRI-compatible script occupies its own line; each
9089 command line starts with the keyword that identifies the command (though
9090 blank lines are also allowed for punctuation). If a line of an
9091 MRI-compatible script begins with an unrecognized keyword, @command{ld}
9092 issues a warning message, but continues processing the script.
9094 Lines beginning with @samp{*} are comments.
9096 You can write these commands using all upper-case letters, or all
9097 lower case; for example, @samp{chip} is the same as @samp{CHIP}.
9098 The following list shows only the upper-case form of each command.
9101 @cindex @code{ABSOLUTE} (MRI)
9102 @item ABSOLUTE @var{secname}
9103 @itemx ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname}
9104 Normally, @command{ld} includes in the output file all sections from all
9105 the input files. However, in an MRI-compatible script, you can use the
9106 @code{ABSOLUTE} command to restrict the sections that will be present in
9107 your output program. If the @code{ABSOLUTE} command is used at all in a
9108 script, then only the sections named explicitly in @code{ABSOLUTE}
9109 commands will appear in the linker output. You can still use other
9110 input sections (whatever you select on the command line, or using
9111 @code{LOAD}) to resolve addresses in the output file.
9113 @cindex @code{ALIAS} (MRI)
9114 @item ALIAS @var{out-secname}, @var{in-secname}
9115 Use this command to place the data from input section @var{in-secname}
9116 in a section called @var{out-secname} in the linker output file.
9118 @var{in-secname} may be an integer.
9120 @cindex @code{ALIGN} (MRI)
9121 @item ALIGN @var{secname} = @var{expression}
9122 Align the section called @var{secname} to @var{expression}. The
9123 @var{expression} should be a power of two.
9125 @cindex @code{BASE} (MRI)
9126 @item BASE @var{expression}
9127 Use the value of @var{expression} as the lowest address (other than
9128 absolute addresses) in the output file.
9130 @cindex @code{CHIP} (MRI)
9131 @item CHIP @var{expression}
9132 @itemx CHIP @var{expression}, @var{expression}
9133 This command does nothing; it is accepted only for compatibility.
9135 @cindex @code{END} (MRI)
9137 This command does nothing whatever; it's only accepted for compatibility.
9139 @cindex @code{FORMAT} (MRI)
9140 @item FORMAT @var{output-format}
9141 Similar to the @code{OUTPUT_FORMAT} command in the more general linker
9142 language, but restricted to S-records, if @var{output-format} is @samp{S}
9144 @cindex @code{LIST} (MRI)
9145 @item LIST @var{anything}@dots{}
9146 Print (to the standard output file) a link map, as produced by the
9147 @command{ld} command-line option @samp{-M}.
9149 The keyword @code{LIST} may be followed by anything on the
9150 same line, with no change in its effect.
9152 @cindex @code{LOAD} (MRI)
9153 @item LOAD @var{filename}
9154 @itemx LOAD @var{filename}, @var{filename}, @dots{} @var{filename}
9155 Include one or more object file @var{filename} in the link; this has the
9156 same effect as specifying @var{filename} directly on the @command{ld}
9159 @cindex @code{NAME} (MRI)
9160 @item NAME @var{output-name}
9161 @var{output-name} is the name for the program produced by @command{ld}; the
9162 MRI-compatible command @code{NAME} is equivalent to the command-line
9163 option @samp{-o} or the general script language command @code{OUTPUT}.
9165 @cindex @code{ORDER} (MRI)
9166 @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname}
9167 @itemx ORDER @var{secname} @var{secname} @var{secname}
9168 Normally, @command{ld} orders the sections in its output file in the
9169 order in which they first appear in the input files. In an MRI-compatible
9170 script, you can override this ordering with the @code{ORDER} command. The
9171 sections you list with @code{ORDER} will appear first in your output
9172 file, in the order specified.
9174 @cindex @code{PUBLIC} (MRI)
9175 @item PUBLIC @var{name}=@var{expression}
9176 @itemx PUBLIC @var{name},@var{expression}
9177 @itemx PUBLIC @var{name} @var{expression}
9178 Supply a value (@var{expression}) for external symbol
9179 @var{name} used in the linker input files.
9181 @cindex @code{SECT} (MRI)
9182 @item SECT @var{secname}, @var{expression}
9183 @itemx SECT @var{secname}=@var{expression}
9184 @itemx SECT @var{secname} @var{expression}
9185 You can use any of these three forms of the @code{SECT} command to
9186 specify the start address (@var{expression}) for section @var{secname}.
9187 If you have more than one @code{SECT} statement for the same
9188 @var{secname}, only the @emph{first} sets the start address.
9191 @node GNU Free Documentation License
9192 @appendix GNU Free Documentation License
9196 @unnumbered LD Index
9201 % I think something like @@colophon should be in texinfo. In the
9203 \long\def\colophon{\hbox to0pt{}\vfill
9204 \centerline{The body of this manual is set in}
9205 \centerline{\fontname\tenrm,}
9206 \centerline{with headings in {\bf\fontname\tenbf}}
9207 \centerline{and examples in {\tt\fontname\tentt}.}
9208 \centerline{{\it\fontname\tenit\/} and}
9209 \centerline{{\sl\fontname\tensl\/}}
9210 \centerline{are used for emphasis.}\vfill}
9212 % Blame: doc@@cygnus.com, 28mar91.