3 @c Copyright (C) 1991-2015 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
45 @dircategory Software development
47 * Ld: (ld). The GNU linker.
52 This file documents the @sc{gnu} linker LD
53 @ifset VERSION_PACKAGE
54 @value{VERSION_PACKAGE}
56 version @value{VERSION}.
58 Copyright @copyright{} 1991-2015 Free Software Foundation, Inc.
60 Permission is granted to copy, distribute and/or modify this document
61 under the terms of the GNU Free Documentation License, Version 1.3
62 or any later version published by the Free Software Foundation;
63 with no Invariant Sections, with no Front-Cover Texts, and with no
64 Back-Cover Texts. A copy of the license is included in the
65 section entitled ``GNU Free Documentation License''.
69 @setchapternewpage odd
70 @settitle The GNU linker
75 @ifset VERSION_PACKAGE
76 @subtitle @value{VERSION_PACKAGE}
78 @subtitle Version @value{VERSION}
79 @author Steve Chamberlain
80 @author Ian Lance Taylor
85 \hfill Red Hat Inc\par
86 \hfill nickc\@credhat.com, doc\@redhat.com\par
87 \hfill {\it The GNU linker}\par
88 \hfill Edited by Jeffrey Osier (jeffrey\@cygnus.com)\par
90 \global\parindent=0pt % Steve likes it this way.
93 @vskip 0pt plus 1filll
94 @c man begin COPYRIGHT
95 Copyright @copyright{} 1991-2015 Free Software Foundation, Inc.
97 Permission is granted to copy, distribute and/or modify this document
98 under the terms of the GNU Free Documentation License, Version 1.3
99 or any later version published by the Free Software Foundation;
100 with no Invariant Sections, with no Front-Cover Texts, and with no
101 Back-Cover Texts. A copy of the license is included in the
102 section entitled ``GNU Free Documentation License''.
108 @c FIXME: Talk about importance of *order* of args, cmds to linker!
113 This file documents the @sc{gnu} linker ld
114 @ifset VERSION_PACKAGE
115 @value{VERSION_PACKAGE}
117 version @value{VERSION}.
119 This document is distributed under the terms of the GNU Free
120 Documentation License version 1.3. A copy of the license is included
121 in the section entitled ``GNU Free Documentation License''.
124 * Overview:: Overview
125 * Invocation:: Invocation
126 * Scripts:: Linker Scripts
128 * Machine Dependent:: Machine Dependent Features
132 * H8/300:: ld and the H8/300
135 * Renesas:: ld and other Renesas micros
138 * i960:: ld and the Intel 960 family
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 * SPU ELF:: ld and SPU ELF Support
165 * TI COFF:: ld and the TI COFF
168 * Win32:: ld and WIN32 (cygwin/mingw)
171 * Xtensa:: ld and Xtensa Processors
174 @ifclear SingleFormat
177 @c Following blank line required for remaining bug in makeinfo conds/menus
179 * Reporting Bugs:: Reporting Bugs
180 * MRI:: MRI Compatible Script Files
181 * GNU Free Documentation License:: GNU Free Documentation License
182 * LD Index:: LD Index
189 @cindex @sc{gnu} linker
190 @cindex what is this?
193 @c man begin SYNOPSIS
194 ld [@b{options}] @var{objfile} @dots{}
198 ar(1), nm(1), objcopy(1), objdump(1), readelf(1) and
199 the Info entries for @file{binutils} and
204 @c man begin DESCRIPTION
206 @command{ld} combines a number of object and archive files, relocates
207 their data and ties up symbol references. Usually the last step in
208 compiling a program is to run @command{ld}.
210 @command{ld} accepts Linker Command Language files written in
211 a superset of AT&T's Link Editor Command Language syntax,
212 to provide explicit and total control over the linking process.
216 This man page does not describe the command language; see the
217 @command{ld} entry in @code{info} for full details on the command
218 language and on other aspects of the GNU linker.
221 @ifclear SingleFormat
222 This version of @command{ld} uses the general purpose BFD libraries
223 to operate on object files. This allows @command{ld} to read, combine, and
224 write object files in many different formats---for example, COFF or
225 @code{a.out}. Different formats may be linked together to produce any
226 available kind of object file. @xref{BFD}, for more information.
229 Aside from its flexibility, the @sc{gnu} linker is more helpful than other
230 linkers in providing diagnostic information. Many linkers abandon
231 execution immediately upon encountering an error; whenever possible,
232 @command{ld} continues executing, allowing you to identify other errors
233 (or, in some cases, to get an output file in spite of the error).
240 @c man begin DESCRIPTION
242 The @sc{gnu} linker @command{ld} is meant to cover a broad range of situations,
243 and to be as compatible as possible with other linkers. As a result,
244 you have many choices to control its behavior.
250 * Options:: Command Line Options
251 * Environment:: Environment Variables
255 @section Command Line Options
263 The linker supports a plethora of command-line options, but in actual
264 practice few of them are used in any particular context.
265 @cindex standard Unix system
266 For instance, a frequent use of @command{ld} is to link standard Unix
267 object files on a standard, supported Unix system. On such a system, to
268 link a file @code{hello.o}:
271 ld -o @var{output} /lib/crt0.o hello.o -lc
274 This tells @command{ld} to produce a file called @var{output} as the
275 result of linking the file @code{/lib/crt0.o} with @code{hello.o} and
276 the library @code{libc.a}, which will come from the standard search
277 directories. (See the discussion of the @samp{-l} option below.)
279 Some of the command-line options to @command{ld} may be specified at any
280 point in the command line. However, options which refer to files, such
281 as @samp{-l} or @samp{-T}, cause the file to be read at the point at
282 which the option appears in the command line, relative to the object
283 files and other file options. Repeating non-file options with a
284 different argument will either have no further effect, or override prior
285 occurrences (those further to the left on the command line) of that
286 option. Options which may be meaningfully specified more than once are
287 noted in the descriptions below.
290 Non-option arguments are object files or archives which are to be linked
291 together. They may follow, precede, or be mixed in with command-line
292 options, except that an object file argument may not be placed between
293 an option and its argument.
295 Usually the linker is invoked with at least one object file, but you can
296 specify other forms of binary input files using @samp{-l}, @samp{-R},
297 and the script command language. If @emph{no} binary input files at all
298 are specified, the linker does not produce any output, and issues the
299 message @samp{No input files}.
301 If the linker cannot recognize the format of an object file, it will
302 assume that it is a linker script. A script specified in this way
303 augments the main linker script used for the link (either the default
304 linker script or the one specified by using @samp{-T}). This feature
305 permits the linker to link against a file which appears to be an object
306 or an archive, but actually merely defines some symbol values, or uses
307 @code{INPUT} or @code{GROUP} to load other objects. Specifying a
308 script in this way merely augments the main linker script, with the
309 extra commands placed after the main script; use the @samp{-T} option
310 to replace the default linker script entirely, but note the effect of
311 the @code{INSERT} command. @xref{Scripts}.
313 For options whose names are a single letter,
314 option arguments must either follow the option letter without intervening
315 whitespace, or be given as separate arguments immediately following the
316 option that requires them.
318 For options whose names are multiple letters, either one dash or two can
319 precede the option name; for example, @samp{-trace-symbol} and
320 @samp{--trace-symbol} are equivalent. Note---there is one exception to
321 this rule. Multiple letter options that start with a lower case 'o' can
322 only be preceded by two dashes. This is to reduce confusion with the
323 @samp{-o} option. So for example @samp{-omagic} sets the output file
324 name to @samp{magic} whereas @samp{--omagic} sets the NMAGIC flag on the
327 Arguments to multiple-letter options must either be separated from the
328 option name by an equals sign, or be given as separate arguments
329 immediately following the option that requires them. For example,
330 @samp{--trace-symbol foo} and @samp{--trace-symbol=foo} are equivalent.
331 Unique abbreviations of the names of multiple-letter options are
334 Note---if the linker is being invoked indirectly, via a compiler driver
335 (e.g. @samp{gcc}) then all the linker command line options should be
336 prefixed by @samp{-Wl,} (or whatever is appropriate for the particular
337 compiler driver) like this:
340 gcc -Wl,--start-group foo.o bar.o -Wl,--end-group
343 This is important, because otherwise the compiler driver program may
344 silently drop the linker options, resulting in a bad link. Confusion
345 may also arise when passing options that require values through a
346 driver, as the use of a space between option and argument acts as
347 a separator, and causes the driver to pass only the option to the linker
348 and the argument to the compiler. In this case, it is simplest to use
349 the joined forms of both single- and multiple-letter options, such as:
352 gcc foo.o bar.o -Wl,-eENTRY -Wl,-Map=a.map
355 Here is a table of the generic command line switches accepted by the GNU
359 @include at-file.texi
361 @kindex -a @var{keyword}
362 @item -a @var{keyword}
363 This option is supported for HP/UX compatibility. The @var{keyword}
364 argument must be one of the strings @samp{archive}, @samp{shared}, or
365 @samp{default}. @samp{-aarchive} is functionally equivalent to
366 @samp{-Bstatic}, and the other two keywords are functionally equivalent
367 to @samp{-Bdynamic}. This option may be used any number of times.
369 @kindex --audit @var{AUDITLIB}
370 @item --audit @var{AUDITLIB}
371 Adds @var{AUDITLIB} to the @code{DT_AUDIT} entry of the dynamic section.
372 @var{AUDITLIB} is not checked for existence, nor will it use the DT_SONAME
373 specified in the library. If specified multiple times @code{DT_AUDIT}
374 will contain a colon separated list of audit interfaces to use. If the linker
375 finds an object with an audit entry while searching for shared libraries,
376 it will add a corresponding @code{DT_DEPAUDIT} entry in the output file.
377 This option is only meaningful on ELF platforms supporting the rtld-audit
381 @cindex architectures
382 @kindex -A @var{arch}
383 @item -A @var{architecture}
384 @kindex --architecture=@var{arch}
385 @itemx --architecture=@var{architecture}
386 In the current release of @command{ld}, this option is useful only for the
387 Intel 960 family of architectures. In that @command{ld} configuration, the
388 @var{architecture} argument identifies the particular architecture in
389 the 960 family, enabling some safeguards and modifying the
390 archive-library search path. @xref{i960,,@command{ld} and the Intel 960
391 family}, for details.
393 Future releases of @command{ld} may support similar functionality for
394 other architecture families.
397 @ifclear SingleFormat
398 @cindex binary input format
399 @kindex -b @var{format}
400 @kindex --format=@var{format}
403 @item -b @var{input-format}
404 @itemx --format=@var{input-format}
405 @command{ld} may be configured to support more than one kind of object
406 file. If your @command{ld} is configured this way, you can use the
407 @samp{-b} option to specify the binary format for input object files
408 that follow this option on the command line. Even when @command{ld} is
409 configured to support alternative object formats, you don't usually need
410 to specify this, as @command{ld} should be configured to expect as a
411 default input format the most usual format on each machine.
412 @var{input-format} is a text string, the name of a particular format
413 supported by the BFD libraries. (You can list the available binary
414 formats with @samp{objdump -i}.)
417 You may want to use this option if you are linking files with an unusual
418 binary format. You can also use @samp{-b} to switch formats explicitly (when
419 linking object files of different formats), by including
420 @samp{-b @var{input-format}} before each group of object files in a
423 The default format is taken from the environment variable
428 You can also define the input format from a script, using the command
431 see @ref{Format Commands}.
435 @kindex -c @var{MRI-cmdfile}
436 @kindex --mri-script=@var{MRI-cmdfile}
437 @cindex compatibility, MRI
438 @item -c @var{MRI-commandfile}
439 @itemx --mri-script=@var{MRI-commandfile}
440 For compatibility with linkers produced by MRI, @command{ld} accepts script
441 files written in an alternate, restricted command language, described in
443 @ref{MRI,,MRI Compatible Script Files}.
446 the MRI Compatible Script Files section of GNU ld documentation.
448 Introduce MRI script files with
449 the option @samp{-c}; use the @samp{-T} option to run linker
450 scripts written in the general-purpose @command{ld} scripting language.
451 If @var{MRI-cmdfile} does not exist, @command{ld} looks for it in the directories
452 specified by any @samp{-L} options.
454 @cindex common allocation
461 These three options are equivalent; multiple forms are supported for
462 compatibility with other linkers. They assign space to common symbols
463 even if a relocatable output file is specified (with @samp{-r}). The
464 script command @code{FORCE_COMMON_ALLOCATION} has the same effect.
465 @xref{Miscellaneous Commands}.
467 @kindex --depaudit @var{AUDITLIB}
468 @kindex -P @var{AUDITLIB}
469 @item --depaudit @var{AUDITLIB}
470 @itemx -P @var{AUDITLIB}
471 Adds @var{AUDITLIB} to the @code{DT_DEPAUDIT} entry of the dynamic section.
472 @var{AUDITLIB} is not checked for existence, nor will it use the DT_SONAME
473 specified in the library. If specified multiple times @code{DT_DEPAUDIT}
474 will contain a colon separated list of audit interfaces to use. This
475 option is only meaningful on ELF platforms supporting the rtld-audit interface.
476 The -P option is provided for Solaris compatibility.
478 @cindex entry point, from command line
479 @kindex -e @var{entry}
480 @kindex --entry=@var{entry}
482 @itemx --entry=@var{entry}
483 Use @var{entry} as the explicit symbol for beginning execution of your
484 program, rather than the default entry point. If there is no symbol
485 named @var{entry}, the linker will try to parse @var{entry} as a number,
486 and use that as the entry address (the number will be interpreted in
487 base 10; you may use a leading @samp{0x} for base 16, or a leading
488 @samp{0} for base 8). @xref{Entry Point}, for a discussion of defaults
489 and other ways of specifying the entry point.
491 @kindex --exclude-libs
492 @item --exclude-libs @var{lib},@var{lib},...
493 Specifies a list of archive libraries from which symbols should not be automatically
494 exported. The library names may be delimited by commas or colons. Specifying
495 @code{--exclude-libs ALL} excludes symbols in all archive libraries from
496 automatic export. This option is available only for the i386 PE targeted
497 port of the linker and for ELF targeted ports. For i386 PE, symbols
498 explicitly listed in a .def file are still exported, regardless of this
499 option. For ELF targeted ports, symbols affected by this option will
500 be treated as hidden.
502 @kindex --exclude-modules-for-implib
503 @item --exclude-modules-for-implib @var{module},@var{module},...
504 Specifies a list of object files or archive members, from which symbols
505 should not be automatically exported, but which should be copied wholesale
506 into the import library being generated during the link. The module names
507 may be delimited by commas or colons, and must match exactly the filenames
508 used by @command{ld} to open the files; for archive members, this is simply
509 the member name, but for object files the name listed must include and
510 match precisely any path used to specify the input file on the linker's
511 command-line. This option is available only for the i386 PE targeted port
512 of the linker. Symbols explicitly listed in a .def file are still exported,
513 regardless of this option.
515 @cindex dynamic symbol table
517 @kindex --export-dynamic
518 @kindex --no-export-dynamic
520 @itemx --export-dynamic
521 @itemx --no-export-dynamic
522 When creating a dynamically linked executable, using the @option{-E}
523 option or the @option{--export-dynamic} option causes the linker to add
524 all symbols to the dynamic symbol table. The dynamic symbol table is the
525 set of symbols which are visible from dynamic objects at run time.
527 If you do not use either of these options (or use the
528 @option{--no-export-dynamic} option to restore the default behavior), the
529 dynamic symbol table will normally contain only those symbols which are
530 referenced by some dynamic object mentioned in the link.
532 If you use @code{dlopen} to load a dynamic object which needs to refer
533 back to the symbols defined by the program, rather than some other
534 dynamic object, then you will probably need to use this option when
535 linking the program itself.
537 You can also use the dynamic list to control what symbols should
538 be added to the dynamic symbol table if the output format supports it.
539 See the description of @samp{--dynamic-list}.
541 Note that this option is specific to ELF targeted ports. PE targets
542 support a similar function to export all symbols from a DLL or EXE; see
543 the description of @samp{--export-all-symbols} below.
545 @ifclear SingleFormat
546 @cindex big-endian objects
550 Link big-endian objects. This affects the default output format.
552 @cindex little-endian objects
555 Link little-endian objects. This affects the default output format.
558 @kindex -f @var{name}
559 @kindex --auxiliary=@var{name}
561 @itemx --auxiliary=@var{name}
562 When creating an ELF shared object, set the internal DT_AUXILIARY field
563 to the specified name. This tells the dynamic linker that the symbol
564 table of the shared object should be used as an auxiliary filter on the
565 symbol table of the shared object @var{name}.
567 If you later link a program against this filter object, then, when you
568 run the program, the dynamic linker will see the DT_AUXILIARY field. If
569 the dynamic linker resolves any symbols from the filter object, it will
570 first check whether there is a definition in the shared object
571 @var{name}. If there is one, it will be used instead of the definition
572 in the filter object. The shared object @var{name} need not exist.
573 Thus the shared object @var{name} may be used to provide an alternative
574 implementation of certain functions, perhaps for debugging or for
575 machine specific performance.
577 This option may be specified more than once. The DT_AUXILIARY entries
578 will be created in the order in which they appear on the command line.
580 @kindex -F @var{name}
581 @kindex --filter=@var{name}
583 @itemx --filter=@var{name}
584 When creating an ELF shared object, set the internal DT_FILTER field to
585 the specified name. This tells the dynamic linker that the symbol table
586 of the shared object which is being created should be used as a filter
587 on the symbol table of the shared object @var{name}.
589 If you later link a program against this filter object, then, when you
590 run the program, the dynamic linker will see the DT_FILTER field. The
591 dynamic linker will resolve symbols according to the symbol table of the
592 filter object as usual, but it will actually link to the definitions
593 found in the shared object @var{name}. Thus the filter object can be
594 used to select a subset of the symbols provided by the object
597 Some older linkers used the @option{-F} option throughout a compilation
598 toolchain for specifying object-file format for both input and output
600 @ifclear SingleFormat
601 The @sc{gnu} linker uses other mechanisms for this purpose: the
602 @option{-b}, @option{--format}, @option{--oformat} options, the
603 @code{TARGET} command in linker scripts, and the @code{GNUTARGET}
604 environment variable.
606 The @sc{gnu} linker will ignore the @option{-F} option when not
607 creating an ELF shared object.
609 @cindex finalization function
610 @kindex -fini=@var{name}
611 @item -fini=@var{name}
612 When creating an ELF executable or shared object, call NAME when the
613 executable or shared object is unloaded, by setting DT_FINI to the
614 address of the function. By default, the linker uses @code{_fini} as
615 the function to call.
619 Ignored. Provided for compatibility with other tools.
621 @kindex -G @var{value}
622 @kindex --gpsize=@var{value}
625 @itemx --gpsize=@var{value}
626 Set the maximum size of objects to be optimized using the GP register to
627 @var{size}. This is only meaningful for object file formats such as
628 MIPS ELF that support putting large and small objects into different
629 sections. This is ignored for other object file formats.
631 @cindex runtime library name
632 @kindex -h @var{name}
633 @kindex -soname=@var{name}
635 @itemx -soname=@var{name}
636 When creating an ELF shared object, set the internal DT_SONAME field to
637 the specified name. When an executable is linked with a shared object
638 which has a DT_SONAME field, then when the executable is run the dynamic
639 linker will attempt to load the shared object specified by the DT_SONAME
640 field rather than the using the file name given to the linker.
643 @cindex incremental link
645 Perform an incremental link (same as option @samp{-r}).
647 @cindex initialization function
648 @kindex -init=@var{name}
649 @item -init=@var{name}
650 When creating an ELF executable or shared object, call NAME when the
651 executable or shared object is loaded, by setting DT_INIT to the address
652 of the function. By default, the linker uses @code{_init} as the
655 @cindex archive files, from cmd line
656 @kindex -l @var{namespec}
657 @kindex --library=@var{namespec}
658 @item -l @var{namespec}
659 @itemx --library=@var{namespec}
660 Add the archive or object file specified by @var{namespec} to the
661 list of files to link. This option may be used any number of times.
662 If @var{namespec} is of the form @file{:@var{filename}}, @command{ld}
663 will search the library path for a file called @var{filename}, otherwise it
664 will search the library path for a file called @file{lib@var{namespec}.a}.
666 On systems which support shared libraries, @command{ld} may also search for
667 files other than @file{lib@var{namespec}.a}. Specifically, on ELF
668 and SunOS systems, @command{ld} will search a directory for a library
669 called @file{lib@var{namespec}.so} before searching for one called
670 @file{lib@var{namespec}.a}. (By convention, a @code{.so} extension
671 indicates a shared library.) Note that this behavior does not apply
672 to @file{:@var{filename}}, which always specifies a file called
675 The linker will search an archive only once, at the location where it is
676 specified on the command line. If the archive defines a symbol which
677 was undefined in some object which appeared before the archive on the
678 command line, the linker will include the appropriate file(s) from the
679 archive. However, an undefined symbol in an object appearing later on
680 the command line will not cause the linker to search the archive again.
682 See the @option{-(} option for a way to force the linker to search
683 archives multiple times.
685 You may list the same archive multiple times on the command line.
688 This type of archive searching is standard for Unix linkers. However,
689 if you are using @command{ld} on AIX, note that it is different from the
690 behaviour of the AIX linker.
693 @cindex search directory, from cmd line
695 @kindex --library-path=@var{dir}
696 @item -L @var{searchdir}
697 @itemx --library-path=@var{searchdir}
698 Add path @var{searchdir} to the list of paths that @command{ld} will search
699 for archive libraries and @command{ld} control scripts. You may use this
700 option any number of times. The directories are searched in the order
701 in which they are specified on the command line. Directories specified
702 on the command line are searched before the default directories. All
703 @option{-L} options apply to all @option{-l} options, regardless of the
704 order in which the options appear. @option{-L} options do not affect
705 how @command{ld} searches for a linker script unless @option{-T}
708 If @var{searchdir} begins with @code{=}, then the @code{=} will be replaced
709 by the @dfn{sysroot prefix}, controlled by the @samp{--sysroot} option, or
710 specified when the linker is configured.
713 The default set of paths searched (without being specified with
714 @samp{-L}) depends on which emulation mode @command{ld} is using, and in
715 some cases also on how it was configured. @xref{Environment}.
718 The paths can also be specified in a link script with the
719 @code{SEARCH_DIR} command. Directories specified this way are searched
720 at the point in which the linker script appears in the command line.
723 @kindex -m @var{emulation}
724 @item -m @var{emulation}
725 Emulate the @var{emulation} linker. You can list the available
726 emulations with the @samp{--verbose} or @samp{-V} options.
728 If the @samp{-m} option is not used, the emulation is taken from the
729 @code{LDEMULATION} environment variable, if that is defined.
731 Otherwise, the default emulation depends upon how the linker was
739 Print a link map to the standard output. A link map provides
740 information about the link, including the following:
744 Where object files are mapped into memory.
746 How common symbols are allocated.
748 All archive members included in the link, with a mention of the symbol
749 which caused the archive member to be brought in.
751 The values assigned to symbols.
753 Note - symbols whose values are computed by an expression which
754 involves a reference to a previous value of the same symbol may not
755 have correct result displayed in the link map. This is because the
756 linker discards intermediate results and only retains the final value
757 of an expression. Under such circumstances the linker will display
758 the final value enclosed by square brackets. Thus for example a
759 linker script containing:
767 will produce the following output in the link map if the @option{-M}
772 [0x0000000c] foo = (foo * 0x4)
773 [0x0000000c] foo = (foo + 0x8)
776 See @ref{Expressions} for more information about expressions in linker
781 @cindex read-only text
786 Turn off page alignment of sections, and disable linking against shared
787 libraries. If the output format supports Unix style magic numbers,
788 mark the output as @code{NMAGIC}.
792 @cindex read/write from cmd line
796 Set the text and data sections to be readable and writable. Also, do
797 not page-align the data segment, and disable linking against shared
798 libraries. If the output format supports Unix style magic numbers,
799 mark the output as @code{OMAGIC}. Note: Although a writable text section
800 is allowed for PE-COFF targets, it does not conform to the format
801 specification published by Microsoft.
806 This option negates most of the effects of the @option{-N} option. It
807 sets the text section to be read-only, and forces the data segment to
808 be page-aligned. Note - this option does not enable linking against
809 shared libraries. Use @option{-Bdynamic} for this.
811 @kindex -o @var{output}
812 @kindex --output=@var{output}
813 @cindex naming the output file
814 @item -o @var{output}
815 @itemx --output=@var{output}
816 Use @var{output} as the name for the program produced by @command{ld}; if this
817 option is not specified, the name @file{a.out} is used by default. The
818 script command @code{OUTPUT} can also specify the output file name.
820 @kindex -O @var{level}
821 @cindex generating optimized output
823 If @var{level} is a numeric values greater than zero @command{ld} optimizes
824 the output. This might take significantly longer and therefore probably
825 should only be enabled for the final binary. At the moment this
826 option only affects ELF shared library generation. Future releases of
827 the linker may make more use of this option. Also currently there is
828 no difference in the linker's behaviour for different non-zero values
829 of this option. Again this may change with future releases.
832 @cindex push state governing input file handling
834 The @option{--push-state} allows to preserve the current state of the
835 flags which govern the input file handling so that they can all be
836 restored with one corresponding @option{--pop-state} option.
838 The option which are covered are: @option{-Bdynamic}, @option{-Bstatic},
839 @option{-dn}, @option{-dy}, @option{-call_shared}, @option{-non_shared},
840 @option{-static}, @option{-N}, @option{-n}, @option{--whole-archive},
841 @option{--no-whole-archive}, @option{-r}, @option{-Ur},
842 @option{--copy-dt-needed-entries}, @option{--no-copy-dt-needed-entries},
843 @option{--as-needed}, @option{--no-as-needed}, and @option{-a}.
845 One target for this option are specifications for @file{pkg-config}. When
846 used with the @option{--libs} option all possibly needed libraries are
847 listed and then possibly linked with all the time. It is better to return
848 something as follows:
851 -Wl,--push-state,--as-needed -libone -libtwo -Wl,--pop-state
855 @cindex pop state governing input file handling
856 Undoes the effect of --push-state, restores the previous values of the
857 flags governing input file handling.
860 @kindex --emit-relocs
861 @cindex retain relocations in final executable
864 Leave relocation sections and contents in fully linked executables.
865 Post link analysis and optimization tools may need this information in
866 order to perform correct modifications of executables. This results
867 in larger executables.
869 This option is currently only supported on ELF platforms.
871 @kindex --force-dynamic
872 @cindex forcing the creation of dynamic sections
873 @item --force-dynamic
874 Force the output file to have dynamic sections. This option is specific
878 @cindex relocatable output
880 @kindex --relocatable
883 Generate relocatable output---i.e., generate an output file that can in
884 turn serve as input to @command{ld}. This is often called @dfn{partial
885 linking}. As a side effect, in environments that support standard Unix
886 magic numbers, this option also sets the output file's magic number to
888 @c ; see @option{-N}.
889 If this option is not specified, an absolute file is produced. When
890 linking C++ programs, this option @emph{will not} resolve references to
891 constructors; to do that, use @samp{-Ur}.
893 When an input file does not have the same format as the output file,
894 partial linking is only supported if that input file does not contain any
895 relocations. Different output formats can have further restrictions; for
896 example some @code{a.out}-based formats do not support partial linking
897 with input files in other formats at all.
899 This option does the same thing as @samp{-i}.
901 @kindex -R @var{file}
902 @kindex --just-symbols=@var{file}
903 @cindex symbol-only input
904 @item -R @var{filename}
905 @itemx --just-symbols=@var{filename}
906 Read symbol names and their addresses from @var{filename}, but do not
907 relocate it or include it in the output. This allows your output file
908 to refer symbolically to absolute locations of memory defined in other
909 programs. You may use this option more than once.
911 For compatibility with other ELF linkers, if the @option{-R} option is
912 followed by a directory name, rather than a file name, it is treated as
913 the @option{-rpath} option.
917 @cindex strip all symbols
920 Omit all symbol information from the output file.
923 @kindex --strip-debug
924 @cindex strip debugger symbols
927 Omit debugger symbol information (but not all symbols) from the output file.
931 @cindex input files, displaying
934 Print the names of the input files as @command{ld} processes them.
936 @kindex -T @var{script}
937 @kindex --script=@var{script}
939 @item -T @var{scriptfile}
940 @itemx --script=@var{scriptfile}
941 Use @var{scriptfile} as the linker script. This script replaces
942 @command{ld}'s default linker script (rather than adding to it), so
943 @var{commandfile} must specify everything necessary to describe the
944 output file. @xref{Scripts}. If @var{scriptfile} does not exist in
945 the current directory, @code{ld} looks for it in the directories
946 specified by any preceding @samp{-L} options. Multiple @samp{-T}
949 @kindex -dT @var{script}
950 @kindex --default-script=@var{script}
952 @item -dT @var{scriptfile}
953 @itemx --default-script=@var{scriptfile}
954 Use @var{scriptfile} as the default linker script. @xref{Scripts}.
956 This option is similar to the @option{--script} option except that
957 processing of the script is delayed until after the rest of the
958 command line has been processed. This allows options placed after the
959 @option{--default-script} option on the command line to affect the
960 behaviour of the linker script, which can be important when the linker
961 command line cannot be directly controlled by the user. (eg because
962 the command line is being constructed by another tool, such as
965 @kindex -u @var{symbol}
966 @kindex --undefined=@var{symbol}
967 @cindex undefined symbol
968 @item -u @var{symbol}
969 @itemx --undefined=@var{symbol}
970 Force @var{symbol} to be entered in the output file as an undefined
971 symbol. Doing this may, for example, trigger linking of additional
972 modules from standard libraries. @samp{-u} may be repeated with
973 different option arguments to enter additional undefined symbols. This
974 option is equivalent to the @code{EXTERN} linker script command.
979 For anything other than C++ programs, this option is equivalent to
980 @samp{-r}: it generates relocatable output---i.e., an output file that can in
981 turn serve as input to @command{ld}. When linking C++ programs, @samp{-Ur}
982 @emph{does} resolve references to constructors, unlike @samp{-r}.
983 It does not work to use @samp{-Ur} on files that were themselves linked
984 with @samp{-Ur}; once the constructor table has been built, it cannot
985 be added to. Use @samp{-Ur} only for the last partial link, and
986 @samp{-r} for the others.
988 @kindex --unique[=@var{SECTION}]
989 @item --unique[=@var{SECTION}]
990 Creates a separate output section for every input section matching
991 @var{SECTION}, or if the optional wildcard @var{SECTION} argument is
992 missing, for every orphan input section. An orphan section is one not
993 specifically mentioned in a linker script. You may use this option
994 multiple times on the command line; It prevents the normal merging of
995 input sections with the same name, overriding output section assignments
1005 Display the version number for @command{ld}. The @option{-V} option also
1006 lists the supported emulations.
1009 @kindex --discard-all
1010 @cindex deleting local symbols
1012 @itemx --discard-all
1013 Delete all local symbols.
1016 @kindex --discard-locals
1017 @cindex local symbols, deleting
1019 @itemx --discard-locals
1020 Delete all temporary local symbols. (These symbols start with
1021 system-specific local label prefixes, typically @samp{.L} for ELF systems
1022 or @samp{L} for traditional a.out systems.)
1024 @kindex -y @var{symbol}
1025 @kindex --trace-symbol=@var{symbol}
1026 @cindex symbol tracing
1027 @item -y @var{symbol}
1028 @itemx --trace-symbol=@var{symbol}
1029 Print the name of each linked file in which @var{symbol} appears. This
1030 option may be given any number of times. On many systems it is necessary
1031 to prepend an underscore.
1033 This option is useful when you have an undefined symbol in your link but
1034 don't know where the reference is coming from.
1036 @kindex -Y @var{path}
1038 Add @var{path} to the default library search path. This option exists
1039 for Solaris compatibility.
1041 @kindex -z @var{keyword}
1042 @item -z @var{keyword}
1043 The recognized keywords are:
1047 Combines multiple reloc sections and sorts them to make dynamic symbol
1048 lookup caching possible.
1051 Disallows undefined symbols in object files. Undefined symbols in
1052 shared libraries are still allowed.
1055 Marks the object as requiring executable stack.
1058 This option is only meaningful when building a shared object. It makes
1059 the symbols defined by this shared object available for symbol resolution
1060 of subsequently loaded libraries.
1063 This option is only meaningful when building a shared object.
1064 It marks the object so that its runtime initialization will occur
1065 before the runtime initialization of any other objects brought into
1066 the process at the same time. Similarly the runtime finalization of
1067 the object will occur after the runtime finalization of any other
1071 Marks the object that its symbol table interposes before all symbols
1072 but the primary executable.
1075 When generating an executable or shared library, mark it to tell the
1076 dynamic linker to defer function call resolution to the point when
1077 the function is called (lazy binding), rather than at load time.
1078 Lazy binding is the default.
1081 Marks the object that its filters be processed immediately at
1085 Allows multiple definitions.
1088 Disables multiple reloc sections combining.
1091 Disable linker generated .dynbss variables used in place of variables
1092 defined in shared libraries. May result in dynamic text relocations.
1095 Marks the object that the search for dependencies of this object will
1096 ignore any default library search paths.
1099 Marks the object shouldn't be unloaded at runtime.
1102 Marks the object not available to @code{dlopen}.
1105 Marks the object can not be dumped by @code{dldump}.
1108 Marks the object as not requiring executable stack.
1111 Treat DT_TEXTREL in shared object as error.
1114 Don't treat DT_TEXTREL in shared object as error.
1117 Don't treat DT_TEXTREL in shared object as error.
1120 Don't create an ELF @code{PT_GNU_RELRO} segment header in the object.
1123 When generating an executable or shared library, mark it to tell the
1124 dynamic linker to resolve all symbols when the program is started, or
1125 when the shared library is linked to using dlopen, instead of
1126 deferring function call resolution to the point when the function is
1130 Marks the object may contain $ORIGIN.
1133 Create an ELF @code{PT_GNU_RELRO} segment header in the object.
1135 @item max-page-size=@var{value}
1136 Set the emulation maximum page size to @var{value}.
1138 @item common-page-size=@var{value}
1139 Set the emulation common page size to @var{value}.
1141 @item stack-size=@var{value}
1142 Specify a stack size for in an ELF @code{PT_GNU_STACK} segment.
1143 Specifying zero will override any default non-zero sized
1144 @code{PT_GNU_STACK} segment creation.
1147 Always generate BND prefix in PLT entries. Supported for Linux/x86_64.
1149 @item noextern-protected-data
1150 Don't treat protected data symbol as external when building shared
1151 library. This option overrides linker backend default. It can be used
1152 to workaround incorrect relocations against protected data symbols
1153 generated by compiler. Updates on protected data symbols by another
1154 module aren't visibile to the resulting shared library. Supported for
1159 Other keywords are ignored for Solaris compatibility.
1162 @cindex groups of archives
1163 @item -( @var{archives} -)
1164 @itemx --start-group @var{archives} --end-group
1165 The @var{archives} should be a list of archive files. They may be
1166 either explicit file names, or @samp{-l} options.
1168 The specified archives are searched repeatedly until no new undefined
1169 references are created. Normally, an archive is searched only once in
1170 the order that it is specified on the command line. If a symbol in that
1171 archive is needed to resolve an undefined symbol referred to by an
1172 object in an archive that appears later on the command line, the linker
1173 would not be able to resolve that reference. By grouping the archives,
1174 they all be searched repeatedly until all possible references are
1177 Using this option has a significant performance cost. It is best to use
1178 it only when there are unavoidable circular references between two or
1181 @kindex --accept-unknown-input-arch
1182 @kindex --no-accept-unknown-input-arch
1183 @item --accept-unknown-input-arch
1184 @itemx --no-accept-unknown-input-arch
1185 Tells the linker to accept input files whose architecture cannot be
1186 recognised. The assumption is that the user knows what they are doing
1187 and deliberately wants to link in these unknown input files. This was
1188 the default behaviour of the linker, before release 2.14. The default
1189 behaviour from release 2.14 onwards is to reject such input files, and
1190 so the @samp{--accept-unknown-input-arch} option has been added to
1191 restore the old behaviour.
1194 @kindex --no-as-needed
1196 @itemx --no-as-needed
1197 This option affects ELF DT_NEEDED tags for dynamic libraries mentioned
1198 on the command line after the @option{--as-needed} option. Normally
1199 the linker will add a DT_NEEDED tag for each dynamic library mentioned
1200 on the command line, regardless of whether the library is actually
1201 needed or not. @option{--as-needed} causes a DT_NEEDED tag to only be
1202 emitted for a library that @emph{at that point in the link} satisfies a
1203 non-weak undefined symbol reference from a regular object file or, if
1204 the library is not found in the DT_NEEDED lists of other needed libraries, a
1205 non-weak undefined symbol reference from another needed dynamic library.
1206 Object files or libraries appearing on the command line @emph{after}
1207 the library in question do not affect whether the library is seen as
1208 needed. This is similar to the rules for extraction of object files
1209 from archives. @option{--no-as-needed} restores the default behaviour.
1211 @kindex --add-needed
1212 @kindex --no-add-needed
1214 @itemx --no-add-needed
1215 These two options have been deprecated because of the similarity of
1216 their names to the @option{--as-needed} and @option{--no-as-needed}
1217 options. They have been replaced by @option{--copy-dt-needed-entries}
1218 and @option{--no-copy-dt-needed-entries}.
1220 @kindex -assert @var{keyword}
1221 @item -assert @var{keyword}
1222 This option is ignored for SunOS compatibility.
1226 @kindex -call_shared
1230 Link against dynamic libraries. This is only meaningful on platforms
1231 for which shared libraries are supported. This option is normally the
1232 default on such platforms. The different variants of this option are
1233 for compatibility with various systems. You may use this option
1234 multiple times on the command line: it affects library searching for
1235 @option{-l} options which follow it.
1239 Set the @code{DF_1_GROUP} flag in the @code{DT_FLAGS_1} entry in the dynamic
1240 section. This causes the runtime linker to handle lookups in this
1241 object and its dependencies to be performed only inside the group.
1242 @option{--unresolved-symbols=report-all} is implied. This option is
1243 only meaningful on ELF platforms which support shared libraries.
1253 Do not link against shared libraries. This is only meaningful on
1254 platforms for which shared libraries are supported. The different
1255 variants of this option are for compatibility with various systems. You
1256 may use this option multiple times on the command line: it affects
1257 library searching for @option{-l} options which follow it. This
1258 option also implies @option{--unresolved-symbols=report-all}. This
1259 option can be used with @option{-shared}. Doing so means that a
1260 shared library is being created but that all of the library's external
1261 references must be resolved by pulling in entries from static
1266 When creating a shared library, bind references to global symbols to the
1267 definition within the shared library, if any. Normally, it is possible
1268 for a program linked against a shared library to override the definition
1269 within the shared library. This option is only meaningful on ELF
1270 platforms which support shared libraries.
1272 @kindex -Bsymbolic-functions
1273 @item -Bsymbolic-functions
1274 When creating a shared library, bind references to global function
1275 symbols to the definition within the shared library, if any.
1276 This option is only meaningful on ELF platforms which support shared
1279 @kindex --dynamic-list=@var{dynamic-list-file}
1280 @item --dynamic-list=@var{dynamic-list-file}
1281 Specify the name of a dynamic list file to the linker. This is
1282 typically used when creating shared libraries to specify a list of
1283 global symbols whose references shouldn't be bound to the definition
1284 within the shared library, or creating dynamically linked executables
1285 to specify a list of symbols which should be added to the symbol table
1286 in the executable. This option is only meaningful on ELF platforms
1287 which support shared libraries.
1289 The format of the dynamic list is the same as the version node without
1290 scope and node name. See @ref{VERSION} for more information.
1292 @kindex --dynamic-list-data
1293 @item --dynamic-list-data
1294 Include all global data symbols to the dynamic list.
1296 @kindex --dynamic-list-cpp-new
1297 @item --dynamic-list-cpp-new
1298 Provide the builtin dynamic list for C++ operator new and delete. It
1299 is mainly useful for building shared libstdc++.
1301 @kindex --dynamic-list-cpp-typeinfo
1302 @item --dynamic-list-cpp-typeinfo
1303 Provide the builtin dynamic list for C++ runtime type identification.
1305 @kindex --check-sections
1306 @kindex --no-check-sections
1307 @item --check-sections
1308 @itemx --no-check-sections
1309 Asks the linker @emph{not} to check section addresses after they have
1310 been assigned to see if there are any overlaps. Normally the linker will
1311 perform this check, and if it finds any overlaps it will produce
1312 suitable error messages. The linker does know about, and does make
1313 allowances for sections in overlays. The default behaviour can be
1314 restored by using the command line switch @option{--check-sections}.
1315 Section overlap is not usually checked for relocatable links. You can
1316 force checking in that case by using the @option{--check-sections}
1319 @kindex --copy-dt-needed-entries
1320 @kindex --no-copy-dt-needed-entries
1321 @item --copy-dt-needed-entries
1322 @itemx --no-copy-dt-needed-entries
1323 This option affects the treatment of dynamic libraries referred to
1324 by DT_NEEDED tags @emph{inside} ELF dynamic libraries mentioned on the
1325 command line. Normally the linker won't add a DT_NEEDED tag to the
1326 output binary for each library mentioned in a DT_NEEDED tag in an
1327 input dynamic library. With @option{--copy-dt-needed-entries}
1328 specified on the command line however any dynamic libraries that
1329 follow it will have their DT_NEEDED entries added. The default
1330 behaviour can be restored with @option{--no-copy-dt-needed-entries}.
1332 This option also has an effect on the resolution of symbols in dynamic
1333 libraries. With @option{--copy-dt-needed-entries} dynamic libraries
1334 mentioned on the command line will be recursively searched, following
1335 their DT_NEEDED tags to other libraries, in order to resolve symbols
1336 required by the output binary. With the default setting however
1337 the searching of dynamic libraries that follow it will stop with the
1338 dynamic library itself. No DT_NEEDED links will be traversed to resolve
1341 @cindex cross reference table
1344 Output a cross reference table. If a linker map file is being
1345 generated, the cross reference table is printed to the map file.
1346 Otherwise, it is printed on the standard output.
1348 The format of the table is intentionally simple, so that it may be
1349 easily processed by a script if necessary. The symbols are printed out,
1350 sorted by name. For each symbol, a list of file names is given. If the
1351 symbol is defined, the first file listed is the location of the
1352 definition. If the symbol is defined as a common value then any files
1353 where this happens appear next. Finally any files that reference the
1356 @cindex common allocation
1357 @kindex --no-define-common
1358 @item --no-define-common
1359 This option inhibits the assignment of addresses to common symbols.
1360 The script command @code{INHIBIT_COMMON_ALLOCATION} has the same effect.
1361 @xref{Miscellaneous Commands}.
1363 The @samp{--no-define-common} option allows decoupling
1364 the decision to assign addresses to Common symbols from the choice
1365 of the output file type; otherwise a non-Relocatable output type
1366 forces assigning addresses to Common symbols.
1367 Using @samp{--no-define-common} allows Common symbols that are referenced
1368 from a shared library to be assigned addresses only in the main program.
1369 This eliminates the unused duplicate space in the shared library,
1370 and also prevents any possible confusion over resolving to the wrong
1371 duplicate when there are many dynamic modules with specialized search
1372 paths for runtime symbol resolution.
1374 @cindex symbols, from command line
1375 @kindex --defsym=@var{symbol}=@var{exp}
1376 @item --defsym=@var{symbol}=@var{expression}
1377 Create a global symbol in the output file, containing the absolute
1378 address given by @var{expression}. You may use this option as many
1379 times as necessary to define multiple symbols in the command line. A
1380 limited form of arithmetic is supported for the @var{expression} in this
1381 context: you may give a hexadecimal constant or the name of an existing
1382 symbol, or use @code{+} and @code{-} to add or subtract hexadecimal
1383 constants or symbols. If you need more elaborate expressions, consider
1384 using the linker command language from a script (@pxref{Assignments}).
1385 @emph{Note:} there should be no white space between @var{symbol}, the
1386 equals sign (``@key{=}''), and @var{expression}.
1388 @cindex demangling, from command line
1389 @kindex --demangle[=@var{style}]
1390 @kindex --no-demangle
1391 @item --demangle[=@var{style}]
1392 @itemx --no-demangle
1393 These options control whether to demangle symbol names in error messages
1394 and other output. When the linker is told to demangle, it tries to
1395 present symbol names in a readable fashion: it strips leading
1396 underscores if they are used by the object file format, and converts C++
1397 mangled symbol names into user readable names. Different compilers have
1398 different mangling styles. The optional demangling style argument can be used
1399 to choose an appropriate demangling style for your compiler. The linker will
1400 demangle by default unless the environment variable @samp{COLLECT_NO_DEMANGLE}
1401 is set. These options may be used to override the default.
1403 @cindex dynamic linker, from command line
1404 @kindex -I@var{file}
1405 @kindex --dynamic-linker=@var{file}
1407 @itemx --dynamic-linker=@var{file}
1408 Set the name of the dynamic linker. This is only meaningful when
1409 generating dynamically linked ELF executables. The default dynamic
1410 linker is normally correct; don't use this unless you know what you are
1413 @kindex --fatal-warnings
1414 @kindex --no-fatal-warnings
1415 @item --fatal-warnings
1416 @itemx --no-fatal-warnings
1417 Treat all warnings as errors. The default behaviour can be restored
1418 with the option @option{--no-fatal-warnings}.
1420 @kindex --force-exe-suffix
1421 @item --force-exe-suffix
1422 Make sure that an output file has a .exe suffix.
1424 If a successfully built fully linked output file does not have a
1425 @code{.exe} or @code{.dll} suffix, this option forces the linker to copy
1426 the output file to one of the same name with a @code{.exe} suffix. This
1427 option is useful when using unmodified Unix makefiles on a Microsoft
1428 Windows host, since some versions of Windows won't run an image unless
1429 it ends in a @code{.exe} suffix.
1431 @kindex --gc-sections
1432 @kindex --no-gc-sections
1433 @cindex garbage collection
1435 @itemx --no-gc-sections
1436 Enable garbage collection of unused input sections. It is ignored on
1437 targets that do not support this option. The default behaviour (of not
1438 performing this garbage collection) can be restored by specifying
1439 @samp{--no-gc-sections} on the command line.
1441 @samp{--gc-sections} decides which input sections are used by
1442 examining symbols and relocations. The section containing the entry
1443 symbol and all sections containing symbols undefined on the
1444 command-line will be kept, as will sections containing symbols
1445 referenced by dynamic objects. Note that when building shared
1446 libraries, the linker must assume that any visible symbol is
1447 referenced. Once this initial set of sections has been determined,
1448 the linker recursively marks as used any section referenced by their
1449 relocations. See @samp{--entry} and @samp{--undefined}.
1451 This option can be set when doing a partial link (enabled with option
1452 @samp{-r}). In this case the root of symbols kept must be explicitly
1453 specified either by an @samp{--entry} or @samp{--undefined} option or by
1454 a @code{ENTRY} command in the linker script.
1456 @kindex --print-gc-sections
1457 @kindex --no-print-gc-sections
1458 @cindex garbage collection
1459 @item --print-gc-sections
1460 @itemx --no-print-gc-sections
1461 List all sections removed by garbage collection. The listing is
1462 printed on stderr. This option is only effective if garbage
1463 collection has been enabled via the @samp{--gc-sections}) option. The
1464 default behaviour (of not listing the sections that are removed) can
1465 be restored by specifying @samp{--no-print-gc-sections} on the command
1468 @kindex --print-output-format
1469 @cindex output format
1470 @item --print-output-format
1471 Print the name of the default output format (perhaps influenced by
1472 other command-line options). This is the string that would appear
1473 in an @code{OUTPUT_FORMAT} linker script command (@pxref{File Commands}).
1479 Print a summary of the command-line options on the standard output and exit.
1481 @kindex --target-help
1483 Print a summary of all target specific options on the standard output and exit.
1485 @kindex -Map=@var{mapfile}
1486 @item -Map=@var{mapfile}
1487 Print a link map to the file @var{mapfile}. See the description of the
1488 @option{-M} option, above.
1490 @cindex memory usage
1491 @kindex --no-keep-memory
1492 @item --no-keep-memory
1493 @command{ld} normally optimizes for speed over memory usage by caching the
1494 symbol tables of input files in memory. This option tells @command{ld} to
1495 instead optimize for memory usage, by rereading the symbol tables as
1496 necessary. This may be required if @command{ld} runs out of memory space
1497 while linking a large executable.
1499 @kindex --no-undefined
1501 @item --no-undefined
1503 Report unresolved symbol references from regular object files. This
1504 is done even if the linker is creating a non-symbolic shared library.
1505 The switch @option{--[no-]allow-shlib-undefined} controls the
1506 behaviour for reporting unresolved references found in shared
1507 libraries being linked in.
1509 @kindex --allow-multiple-definition
1511 @item --allow-multiple-definition
1513 Normally when a symbol is defined multiple times, the linker will
1514 report a fatal error. These options allow multiple definitions and the
1515 first definition will be used.
1517 @kindex --allow-shlib-undefined
1518 @kindex --no-allow-shlib-undefined
1519 @item --allow-shlib-undefined
1520 @itemx --no-allow-shlib-undefined
1521 Allows or disallows undefined symbols in shared libraries.
1522 This switch is similar to @option{--no-undefined} except that it
1523 determines the behaviour when the undefined symbols are in a
1524 shared library rather than a regular object file. It does not affect
1525 how undefined symbols in regular object files are handled.
1527 The default behaviour is to report errors for any undefined symbols
1528 referenced in shared libraries if the linker is being used to create
1529 an executable, but to allow them if the linker is being used to create
1532 The reasons for allowing undefined symbol references in shared
1533 libraries specified at link time are that:
1537 A shared library specified at link time may not be the same as the one
1538 that is available at load time, so the symbol might actually be
1539 resolvable at load time.
1541 There are some operating systems, eg BeOS and HPPA, where undefined
1542 symbols in shared libraries are normal.
1544 The BeOS kernel for example patches shared libraries at load time to
1545 select whichever function is most appropriate for the current
1546 architecture. This is used, for example, to dynamically select an
1547 appropriate memset function.
1550 @kindex --no-undefined-version
1551 @item --no-undefined-version
1552 Normally when a symbol has an undefined version, the linker will ignore
1553 it. This option disallows symbols with undefined version and a fatal error
1554 will be issued instead.
1556 @kindex --default-symver
1557 @item --default-symver
1558 Create and use a default symbol version (the soname) for unversioned
1561 @kindex --default-imported-symver
1562 @item --default-imported-symver
1563 Create and use a default symbol version (the soname) for unversioned
1566 @kindex --no-warn-mismatch
1567 @item --no-warn-mismatch
1568 Normally @command{ld} will give an error if you try to link together input
1569 files that are mismatched for some reason, perhaps because they have
1570 been compiled for different processors or for different endiannesses.
1571 This option tells @command{ld} that it should silently permit such possible
1572 errors. This option should only be used with care, in cases when you
1573 have taken some special action that ensures that the linker errors are
1576 @kindex --no-warn-search-mismatch
1577 @item --no-warn-search-mismatch
1578 Normally @command{ld} will give a warning if it finds an incompatible
1579 library during a library search. This option silences the warning.
1581 @kindex --no-whole-archive
1582 @item --no-whole-archive
1583 Turn off the effect of the @option{--whole-archive} option for subsequent
1586 @cindex output file after errors
1587 @kindex --noinhibit-exec
1588 @item --noinhibit-exec
1589 Retain the executable output file whenever it is still usable.
1590 Normally, the linker will not produce an output file if it encounters
1591 errors during the link process; it exits without writing an output file
1592 when it issues any error whatsoever.
1596 Only search library directories explicitly specified on the
1597 command line. Library directories specified in linker scripts
1598 (including linker scripts specified on the command line) are ignored.
1600 @ifclear SingleFormat
1601 @kindex --oformat=@var{output-format}
1602 @item --oformat=@var{output-format}
1603 @command{ld} may be configured to support more than one kind of object
1604 file. If your @command{ld} is configured this way, you can use the
1605 @samp{--oformat} option to specify the binary format for the output
1606 object file. Even when @command{ld} is configured to support alternative
1607 object formats, you don't usually need to specify this, as @command{ld}
1608 should be configured to produce as a default output format the most
1609 usual format on each machine. @var{output-format} is a text string, the
1610 name of a particular format supported by the BFD libraries. (You can
1611 list the available binary formats with @samp{objdump -i}.) The script
1612 command @code{OUTPUT_FORMAT} can also specify the output format, but
1613 this option overrides it. @xref{BFD}.
1617 @kindex --pic-executable
1619 @itemx --pic-executable
1620 @cindex position independent executables
1621 Create a position independent executable. This is currently only supported on
1622 ELF platforms. Position independent executables are similar to shared
1623 libraries in that they are relocated by the dynamic linker to the virtual
1624 address the OS chooses for them (which can vary between invocations). Like
1625 normal dynamically linked executables they can be executed and symbols
1626 defined in the executable cannot be overridden by shared libraries.
1630 This option is ignored for Linux compatibility.
1634 This option is ignored for SVR4 compatibility.
1637 @cindex synthesizing linker
1638 @cindex relaxing addressing modes
1642 An option with machine dependent effects.
1644 This option is only supported on a few targets.
1647 @xref{H8/300,,@command{ld} and the H8/300}.
1650 @xref{i960,, @command{ld} and the Intel 960 family}.
1653 @xref{Xtensa,, @command{ld} and Xtensa Processors}.
1656 @xref{M68HC11/68HC12,,@command{ld} and the 68HC11 and 68HC12}.
1659 @xref{Nios II,,@command{ld} and the Altera Nios II}.
1662 @xref{PowerPC ELF32,,@command{ld} and PowerPC 32-bit ELF Support}.
1665 On some platforms the @samp{--relax} option performs target specific,
1666 global optimizations that become possible when the linker resolves
1667 addressing in the program, such as relaxing address modes,
1668 synthesizing new instructions, selecting shorter version of current
1669 instructions, and combining constant values.
1671 On some platforms these link time global optimizations may make symbolic
1672 debugging of the resulting executable impossible.
1674 This is known to be the case for the Matsushita MN10200 and MN10300
1675 family of processors.
1679 On platforms where this is not supported, @samp{--relax} is accepted,
1683 On platforms where @samp{--relax} is accepted the option
1684 @samp{--no-relax} can be used to disable the feature.
1686 @cindex retaining specified symbols
1687 @cindex stripping all but some symbols
1688 @cindex symbols, retaining selectively
1689 @kindex --retain-symbols-file=@var{filename}
1690 @item --retain-symbols-file=@var{filename}
1691 Retain @emph{only} the symbols listed in the file @var{filename},
1692 discarding all others. @var{filename} is simply a flat file, with one
1693 symbol name per line. This option is especially useful in environments
1697 where a large global symbol table is accumulated gradually, to conserve
1700 @samp{--retain-symbols-file} does @emph{not} discard undefined symbols,
1701 or symbols needed for relocations.
1703 You may only specify @samp{--retain-symbols-file} once in the command
1704 line. It overrides @samp{-s} and @samp{-S}.
1707 @item -rpath=@var{dir}
1708 @cindex runtime library search path
1709 @kindex -rpath=@var{dir}
1710 Add a directory to the runtime library search path. This is used when
1711 linking an ELF executable with shared objects. All @option{-rpath}
1712 arguments are concatenated and passed to the runtime linker, which uses
1713 them to locate shared objects at runtime. The @option{-rpath} option is
1714 also used when locating shared objects which are needed by shared
1715 objects explicitly included in the link; see the description of the
1716 @option{-rpath-link} option. If @option{-rpath} is not used when linking an
1717 ELF executable, the contents of the environment variable
1718 @code{LD_RUN_PATH} will be used if it is defined.
1720 The @option{-rpath} option may also be used on SunOS. By default, on
1721 SunOS, the linker will form a runtime search patch out of all the
1722 @option{-L} options it is given. If a @option{-rpath} option is used, the
1723 runtime search path will be formed exclusively using the @option{-rpath}
1724 options, ignoring the @option{-L} options. This can be useful when using
1725 gcc, which adds many @option{-L} options which may be on NFS mounted
1728 For compatibility with other ELF linkers, if the @option{-R} option is
1729 followed by a directory name, rather than a file name, it is treated as
1730 the @option{-rpath} option.
1734 @cindex link-time runtime library search path
1735 @kindex -rpath-link=@var{dir}
1736 @item -rpath-link=@var{dir}
1737 When using ELF or SunOS, one shared library may require another. This
1738 happens when an @code{ld -shared} link includes a shared library as one
1741 When the linker encounters such a dependency when doing a non-shared,
1742 non-relocatable link, it will automatically try to locate the required
1743 shared library and include it in the link, if it is not included
1744 explicitly. In such a case, the @option{-rpath-link} option
1745 specifies the first set of directories to search. The
1746 @option{-rpath-link} option may specify a sequence of directory names
1747 either by specifying a list of names separated by colons, or by
1748 appearing multiple times.
1750 This option should be used with caution as it overrides the search path
1751 that may have been hard compiled into a shared library. In such a case it
1752 is possible to use unintentionally a different search path than the
1753 runtime linker would do.
1755 The linker uses the following search paths to locate required shared
1759 Any directories specified by @option{-rpath-link} options.
1761 Any directories specified by @option{-rpath} options. The difference
1762 between @option{-rpath} and @option{-rpath-link} is that directories
1763 specified by @option{-rpath} options are included in the executable and
1764 used at runtime, whereas the @option{-rpath-link} option is only effective
1765 at link time. Searching @option{-rpath} in this way is only supported
1766 by native linkers and cross linkers which have been configured with
1767 the @option{--with-sysroot} option.
1769 On an ELF system, for native linkers, if the @option{-rpath} and
1770 @option{-rpath-link} options were not used, search the contents of the
1771 environment variable @code{LD_RUN_PATH}.
1773 On SunOS, if the @option{-rpath} option was not used, search any
1774 directories specified using @option{-L} options.
1776 For a native linker, search the contents of the environment
1777 variable @code{LD_LIBRARY_PATH}.
1779 For a native ELF linker, the directories in @code{DT_RUNPATH} or
1780 @code{DT_RPATH} of a shared library are searched for shared
1781 libraries needed by it. The @code{DT_RPATH} entries are ignored if
1782 @code{DT_RUNPATH} entries exist.
1784 The default directories, normally @file{/lib} and @file{/usr/lib}.
1786 For a native linker on an ELF system, if the file @file{/etc/ld.so.conf}
1787 exists, the list of directories found in that file.
1790 If the required shared library is not found, the linker will issue a
1791 warning and continue with the link.
1798 @cindex shared libraries
1799 Create a shared library. This is currently only supported on ELF, XCOFF
1800 and SunOS platforms. On SunOS, the linker will automatically create a
1801 shared library if the @option{-e} option is not used and there are
1802 undefined symbols in the link.
1804 @kindex --sort-common
1806 @itemx --sort-common=ascending
1807 @itemx --sort-common=descending
1808 This option tells @command{ld} to sort the common symbols by alignment in
1809 ascending or descending order when it places them in the appropriate output
1810 sections. The symbol alignments considered are sixteen-byte or larger,
1811 eight-byte, four-byte, two-byte, and one-byte. This is to prevent gaps
1812 between symbols due to alignment constraints. If no sorting order is
1813 specified, then descending order is assumed.
1815 @kindex --sort-section=name
1816 @item --sort-section=name
1817 This option will apply @code{SORT_BY_NAME} to all wildcard section
1818 patterns in the linker script.
1820 @kindex --sort-section=alignment
1821 @item --sort-section=alignment
1822 This option will apply @code{SORT_BY_ALIGNMENT} to all wildcard section
1823 patterns in the linker script.
1825 @kindex --split-by-file
1826 @item --split-by-file[=@var{size}]
1827 Similar to @option{--split-by-reloc} but creates a new output section for
1828 each input file when @var{size} is reached. @var{size} defaults to a
1829 size of 1 if not given.
1831 @kindex --split-by-reloc
1832 @item --split-by-reloc[=@var{count}]
1833 Tries to creates extra sections in the output file so that no single
1834 output section in the file contains more than @var{count} relocations.
1835 This is useful when generating huge relocatable files for downloading into
1836 certain real time kernels with the COFF object file format; since COFF
1837 cannot represent more than 65535 relocations in a single section. Note
1838 that this will fail to work with object file formats which do not
1839 support arbitrary sections. The linker will not split up individual
1840 input sections for redistribution, so if a single input section contains
1841 more than @var{count} relocations one output section will contain that
1842 many relocations. @var{count} defaults to a value of 32768.
1846 Compute and display statistics about the operation of the linker, such
1847 as execution time and memory usage.
1849 @kindex --sysroot=@var{directory}
1850 @item --sysroot=@var{directory}
1851 Use @var{directory} as the location of the sysroot, overriding the
1852 configure-time default. This option is only supported by linkers
1853 that were configured using @option{--with-sysroot}.
1855 @kindex --traditional-format
1856 @cindex traditional format
1857 @item --traditional-format
1858 For some targets, the output of @command{ld} is different in some ways from
1859 the output of some existing linker. This switch requests @command{ld} to
1860 use the traditional format instead.
1863 For example, on SunOS, @command{ld} combines duplicate entries in the
1864 symbol string table. This can reduce the size of an output file with
1865 full debugging information by over 30 percent. Unfortunately, the SunOS
1866 @code{dbx} program can not read the resulting program (@code{gdb} has no
1867 trouble). The @samp{--traditional-format} switch tells @command{ld} to not
1868 combine duplicate entries.
1870 @kindex --section-start=@var{sectionname}=@var{org}
1871 @item --section-start=@var{sectionname}=@var{org}
1872 Locate a section in the output file at the absolute
1873 address given by @var{org}. You may use this option as many
1874 times as necessary to locate multiple sections in the command
1876 @var{org} must be a single hexadecimal integer;
1877 for compatibility with other linkers, you may omit the leading
1878 @samp{0x} usually associated with hexadecimal values. @emph{Note:} there
1879 should be no white space between @var{sectionname}, the equals
1880 sign (``@key{=}''), and @var{org}.
1882 @kindex -Tbss=@var{org}
1883 @kindex -Tdata=@var{org}
1884 @kindex -Ttext=@var{org}
1885 @cindex segment origins, cmd line
1886 @item -Tbss=@var{org}
1887 @itemx -Tdata=@var{org}
1888 @itemx -Ttext=@var{org}
1889 Same as @option{--section-start}, with @code{.bss}, @code{.data} or
1890 @code{.text} as the @var{sectionname}.
1892 @kindex -Ttext-segment=@var{org}
1893 @item -Ttext-segment=@var{org}
1894 @cindex text segment origin, cmd line
1895 When creating an ELF executable, it will set the address of the first
1896 byte of the text segment.
1898 @kindex -Trodata-segment=@var{org}
1899 @item -Trodata-segment=@var{org}
1900 @cindex rodata segment origin, cmd line
1901 When creating an ELF executable or shared object for a target where
1902 the read-only data is in its own segment separate from the executable
1903 text, it will set the address of the first byte of the read-only data segment.
1905 @kindex -Tldata-segment=@var{org}
1906 @item -Tldata-segment=@var{org}
1907 @cindex ldata segment origin, cmd line
1908 When creating an ELF executable or shared object for x86-64 medium memory
1909 model, it will set the address of the first byte of the ldata segment.
1911 @kindex --unresolved-symbols
1912 @item --unresolved-symbols=@var{method}
1913 Determine how to handle unresolved symbols. There are four possible
1914 values for @samp{method}:
1918 Do not report any unresolved symbols.
1921 Report all unresolved symbols. This is the default.
1923 @item ignore-in-object-files
1924 Report unresolved symbols that are contained in shared libraries, but
1925 ignore them if they come from regular object files.
1927 @item ignore-in-shared-libs
1928 Report unresolved symbols that come from regular object files, but
1929 ignore them if they come from shared libraries. This can be useful
1930 when creating a dynamic binary and it is known that all the shared
1931 libraries that it should be referencing are included on the linker's
1935 The behaviour for shared libraries on their own can also be controlled
1936 by the @option{--[no-]allow-shlib-undefined} option.
1938 Normally the linker will generate an error message for each reported
1939 unresolved symbol but the option @option{--warn-unresolved-symbols}
1940 can change this to a warning.
1942 @kindex --verbose[=@var{NUMBER}]
1943 @cindex verbose[=@var{NUMBER}]
1945 @itemx --verbose[=@var{NUMBER}]
1946 Display the version number for @command{ld} and list the linker emulations
1947 supported. Display which input files can and cannot be opened. Display
1948 the linker script being used by the linker. If the optional @var{NUMBER}
1949 argument > 1, plugin symbol status will also be displayed.
1951 @kindex --version-script=@var{version-scriptfile}
1952 @cindex version script, symbol versions
1953 @item --version-script=@var{version-scriptfile}
1954 Specify the name of a version script to the linker. This is typically
1955 used when creating shared libraries to specify additional information
1956 about the version hierarchy for the library being created. This option
1957 is only fully supported on ELF platforms which support shared libraries;
1958 see @ref{VERSION}. It is partially supported on PE platforms, which can
1959 use version scripts to filter symbol visibility in auto-export mode: any
1960 symbols marked @samp{local} in the version script will not be exported.
1963 @kindex --warn-common
1964 @cindex warnings, on combining symbols
1965 @cindex combining symbols, warnings on
1967 Warn when a common symbol is combined with another common symbol or with
1968 a symbol definition. Unix linkers allow this somewhat sloppy practice,
1969 but linkers on some other operating systems do not. This option allows
1970 you to find potential problems from combining global symbols.
1971 Unfortunately, some C libraries use this practice, so you may get some
1972 warnings about symbols in the libraries as well as in your programs.
1974 There are three kinds of global symbols, illustrated here by C examples:
1978 A definition, which goes in the initialized data section of the output
1982 An undefined reference, which does not allocate space.
1983 There must be either a definition or a common symbol for the
1987 A common symbol. If there are only (one or more) common symbols for a
1988 variable, it goes in the uninitialized data area of the output file.
1989 The linker merges multiple common symbols for the same variable into a
1990 single symbol. If they are of different sizes, it picks the largest
1991 size. The linker turns a common symbol into a declaration, if there is
1992 a definition of the same variable.
1995 The @samp{--warn-common} option can produce five kinds of warnings.
1996 Each warning consists of a pair of lines: the first describes the symbol
1997 just encountered, and the second describes the previous symbol
1998 encountered with the same name. One or both of the two symbols will be
2003 Turning a common symbol into a reference, because there is already a
2004 definition for the symbol.
2006 @var{file}(@var{section}): warning: common of `@var{symbol}'
2007 overridden by definition
2008 @var{file}(@var{section}): warning: defined here
2012 Turning a common symbol into a reference, because a later definition for
2013 the symbol is encountered. This is the same as the previous case,
2014 except that the symbols are encountered in a different order.
2016 @var{file}(@var{section}): warning: definition of `@var{symbol}'
2018 @var{file}(@var{section}): warning: common is here
2022 Merging a common symbol with a previous same-sized common symbol.
2024 @var{file}(@var{section}): warning: multiple common
2026 @var{file}(@var{section}): warning: previous common is here
2030 Merging a common symbol with a previous larger common symbol.
2032 @var{file}(@var{section}): warning: common of `@var{symbol}'
2033 overridden by larger common
2034 @var{file}(@var{section}): warning: larger common is here
2038 Merging a common symbol with a previous smaller common symbol. This is
2039 the same as the previous case, except that the symbols are
2040 encountered in a different order.
2042 @var{file}(@var{section}): warning: common of `@var{symbol}'
2043 overriding smaller common
2044 @var{file}(@var{section}): warning: smaller common is here
2048 @kindex --warn-constructors
2049 @item --warn-constructors
2050 Warn if any global constructors are used. This is only useful for a few
2051 object file formats. For formats like COFF or ELF, the linker can not
2052 detect the use of global constructors.
2054 @kindex --warn-multiple-gp
2055 @item --warn-multiple-gp
2056 Warn if multiple global pointer values are required in the output file.
2057 This is only meaningful for certain processors, such as the Alpha.
2058 Specifically, some processors put large-valued constants in a special
2059 section. A special register (the global pointer) points into the middle
2060 of this section, so that constants can be loaded efficiently via a
2061 base-register relative addressing mode. Since the offset in
2062 base-register relative mode is fixed and relatively small (e.g., 16
2063 bits), this limits the maximum size of the constant pool. Thus, in
2064 large programs, it is often necessary to use multiple global pointer
2065 values in order to be able to address all possible constants. This
2066 option causes a warning to be issued whenever this case occurs.
2069 @cindex warnings, on undefined symbols
2070 @cindex undefined symbols, warnings on
2072 Only warn once for each undefined symbol, rather than once per module
2075 @kindex --warn-orphan
2076 @kindex --no-warn-orphan
2077 @cindex warnings, on orphan sections
2078 @cindex orphan sections, warnings on
2080 The @option{--warn-orphan} option tells the linker to generate a
2081 warning message whenever it has to place an orphan section into the
2082 output file. @xref{Orphan Sections} The @option{--no-warn-orphan}
2083 option restores the default behaviour of just silently placing these
2086 @kindex --warn-section-align
2087 @cindex warnings, on section alignment
2088 @cindex section alignment, warnings on
2089 @item --warn-section-align
2090 Warn if the address of an output section is changed because of
2091 alignment. Typically, the alignment will be set by an input section.
2092 The address will only be changed if it not explicitly specified; that
2093 is, if the @code{SECTIONS} command does not specify a start address for
2094 the section (@pxref{SECTIONS}).
2096 @kindex --warn-shared-textrel
2097 @item --warn-shared-textrel
2098 Warn if the linker adds a DT_TEXTREL to a shared object.
2100 @kindex --warn-alternate-em
2101 @item --warn-alternate-em
2102 Warn if an object has alternate ELF machine code.
2104 @kindex --warn-unresolved-symbols
2105 @item --warn-unresolved-symbols
2106 If the linker is going to report an unresolved symbol (see the option
2107 @option{--unresolved-symbols}) it will normally generate an error.
2108 This option makes it generate a warning instead.
2110 @kindex --error-unresolved-symbols
2111 @item --error-unresolved-symbols
2112 This restores the linker's default behaviour of generating errors when
2113 it is reporting unresolved symbols.
2115 @kindex --whole-archive
2116 @cindex including an entire archive
2117 @item --whole-archive
2118 For each archive mentioned on the command line after the
2119 @option{--whole-archive} option, include every object file in the archive
2120 in the link, rather than searching the archive for the required object
2121 files. This is normally used to turn an archive file into a shared
2122 library, forcing every object to be included in the resulting shared
2123 library. This option may be used more than once.
2125 Two notes when using this option from gcc: First, gcc doesn't know
2126 about this option, so you have to use @option{-Wl,-whole-archive}.
2127 Second, don't forget to use @option{-Wl,-no-whole-archive} after your
2128 list of archives, because gcc will add its own list of archives to
2129 your link and you may not want this flag to affect those as well.
2131 @kindex --wrap=@var{symbol}
2132 @item --wrap=@var{symbol}
2133 Use a wrapper function for @var{symbol}. Any undefined reference to
2134 @var{symbol} will be resolved to @code{__wrap_@var{symbol}}. Any
2135 undefined reference to @code{__real_@var{symbol}} will be resolved to
2138 This can be used to provide a wrapper for a system function. The
2139 wrapper function should be called @code{__wrap_@var{symbol}}. If it
2140 wishes to call the system function, it should call
2141 @code{__real_@var{symbol}}.
2143 Here is a trivial example:
2147 __wrap_malloc (size_t c)
2149 printf ("malloc called with %zu\n", c);
2150 return __real_malloc (c);
2154 If you link other code with this file using @option{--wrap malloc}, then
2155 all calls to @code{malloc} will call the function @code{__wrap_malloc}
2156 instead. The call to @code{__real_malloc} in @code{__wrap_malloc} will
2157 call the real @code{malloc} function.
2159 You may wish to provide a @code{__real_malloc} function as well, so that
2160 links without the @option{--wrap} option will succeed. If you do this,
2161 you should not put the definition of @code{__real_malloc} in the same
2162 file as @code{__wrap_malloc}; if you do, the assembler may resolve the
2163 call before the linker has a chance to wrap it to @code{malloc}.
2165 @kindex --eh-frame-hdr
2166 @item --eh-frame-hdr
2167 Request creation of @code{.eh_frame_hdr} section and ELF
2168 @code{PT_GNU_EH_FRAME} segment header.
2170 @kindex --ld-generated-unwind-info
2171 @item --no-ld-generated-unwind-info
2172 Request creation of @code{.eh_frame} unwind info for linker
2173 generated code sections like PLT. This option is on by default
2174 if linker generated unwind info is supported.
2176 @kindex --enable-new-dtags
2177 @kindex --disable-new-dtags
2178 @item --enable-new-dtags
2179 @itemx --disable-new-dtags
2180 This linker can create the new dynamic tags in ELF. But the older ELF
2181 systems may not understand them. If you specify
2182 @option{--enable-new-dtags}, the new dynamic tags will be created as needed
2183 and older dynamic tags will be omitted.
2184 If you specify @option{--disable-new-dtags}, no new dynamic tags will be
2185 created. By default, the new dynamic tags are not created. Note that
2186 those options are only available for ELF systems.
2188 @kindex --hash-size=@var{number}
2189 @item --hash-size=@var{number}
2190 Set the default size of the linker's hash tables to a prime number
2191 close to @var{number}. Increasing this value can reduce the length of
2192 time it takes the linker to perform its tasks, at the expense of
2193 increasing the linker's memory requirements. Similarly reducing this
2194 value can reduce the memory requirements at the expense of speed.
2196 @kindex --hash-style=@var{style}
2197 @item --hash-style=@var{style}
2198 Set the type of linker's hash table(s). @var{style} can be either
2199 @code{sysv} for classic ELF @code{.hash} section, @code{gnu} for
2200 new style GNU @code{.gnu.hash} section or @code{both} for both
2201 the classic ELF @code{.hash} and new style GNU @code{.gnu.hash}
2202 hash tables. The default is @code{sysv}.
2204 @kindex --compress-debug-sections=none
2205 @kindex --compress-debug-sections=zlib
2206 @kindex --compress-debug-sections=zlib-gnu
2207 @kindex --compress-debug-sections=zlib-gabi
2208 @item --compress-debug-sections=none
2209 @itemx --compress-debug-sections=zlib
2210 @itemx --compress-debug-sections=zlib-gnu
2211 @itemx --compress-debug-sections=zlib-gabi
2212 On ELF platforms , these options control how DWARF debug sections are
2213 compressed using zlib. @option{--compress-debug-sections=none} doesn't
2214 compress DWARF debug sections. @option{--compress-debug-sections=zlib}
2215 and @option{--compress-debug-sections=zlib-gnu} compress DWARF debug
2216 sections and rename debug section names to begin with @samp{.zdebug}
2217 instead of @samp{.debug}. @option{--compress-debug-sections=zlib-gabi}
2218 compresses DWARF debug sections with SHF_COMPRESSED from the ELF ABI.
2220 @kindex --reduce-memory-overheads
2221 @item --reduce-memory-overheads
2222 This option reduces memory requirements at ld runtime, at the expense of
2223 linking speed. This was introduced to select the old O(n^2) algorithm
2224 for link map file generation, rather than the new O(n) algorithm which uses
2225 about 40% more memory for symbol storage.
2227 Another effect of the switch is to set the default hash table size to
2228 1021, which again saves memory at the cost of lengthening the linker's
2229 run time. This is not done however if the @option{--hash-size} switch
2232 The @option{--reduce-memory-overheads} switch may be also be used to
2233 enable other tradeoffs in future versions of the linker.
2236 @kindex --build-id=@var{style}
2238 @itemx --build-id=@var{style}
2239 Request the creation of a @code{.note.gnu.build-id} ELF note section
2240 or a @code{.build-id} COFF section. The contents of the note are
2241 unique bits identifying this linked file. @var{style} can be
2242 @code{uuid} to use 128 random bits, @code{sha1} to use a 160-bit
2243 @sc{SHA1} hash on the normative parts of the output contents,
2244 @code{md5} to use a 128-bit @sc{MD5} hash on the normative parts of
2245 the output contents, or @code{0x@var{hexstring}} to use a chosen bit
2246 string specified as an even number of hexadecimal digits (@code{-} and
2247 @code{:} characters between digit pairs are ignored). If @var{style}
2248 is omitted, @code{sha1} is used.
2250 The @code{md5} and @code{sha1} styles produces an identifier
2251 that is always the same in an identical output file, but will be
2252 unique among all nonidentical output files. It is not intended
2253 to be compared as a checksum for the file's contents. A linked
2254 file may be changed later by other tools, but the build ID bit
2255 string identifying the original linked file does not change.
2257 Passing @code{none} for @var{style} disables the setting from any
2258 @code{--build-id} options earlier on the command line.
2263 @subsection Options Specific to i386 PE Targets
2265 @c man begin OPTIONS
2267 The i386 PE linker supports the @option{-shared} option, which causes
2268 the output to be a dynamically linked library (DLL) instead of a
2269 normal executable. You should name the output @code{*.dll} when you
2270 use this option. In addition, the linker fully supports the standard
2271 @code{*.def} files, which may be specified on the linker command line
2272 like an object file (in fact, it should precede archives it exports
2273 symbols from, to ensure that they get linked in, just like a normal
2276 In addition to the options common to all targets, the i386 PE linker
2277 support additional command line options that are specific to the i386
2278 PE target. Options that take values may be separated from their
2279 values by either a space or an equals sign.
2283 @kindex --add-stdcall-alias
2284 @item --add-stdcall-alias
2285 If given, symbols with a stdcall suffix (@@@var{nn}) will be exported
2286 as-is and also with the suffix stripped.
2287 [This option is specific to the i386 PE targeted port of the linker]
2290 @item --base-file @var{file}
2291 Use @var{file} as the name of a file in which to save the base
2292 addresses of all the relocations needed for generating DLLs with
2294 [This is an i386 PE specific option]
2298 Create a DLL instead of a regular executable. You may also use
2299 @option{-shared} or specify a @code{LIBRARY} in a given @code{.def}
2301 [This option is specific to the i386 PE targeted port of the linker]
2303 @kindex --enable-long-section-names
2304 @kindex --disable-long-section-names
2305 @item --enable-long-section-names
2306 @itemx --disable-long-section-names
2307 The PE variants of the Coff object format add an extension that permits
2308 the use of section names longer than eight characters, the normal limit
2309 for Coff. By default, these names are only allowed in object files, as
2310 fully-linked executable images do not carry the Coff string table required
2311 to support the longer names. As a GNU extension, it is possible to
2312 allow their use in executable images as well, or to (probably pointlessly!)
2313 disallow it in object files, by using these two options. Executable images
2314 generated with these long section names are slightly non-standard, carrying
2315 as they do a string table, and may generate confusing output when examined
2316 with non-GNU PE-aware tools, such as file viewers and dumpers. However,
2317 GDB relies on the use of PE long section names to find Dwarf-2 debug
2318 information sections in an executable image at runtime, and so if neither
2319 option is specified on the command-line, @command{ld} will enable long
2320 section names, overriding the default and technically correct behaviour,
2321 when it finds the presence of debug information while linking an executable
2322 image and not stripping symbols.
2323 [This option is valid for all PE targeted ports of the linker]
2325 @kindex --enable-stdcall-fixup
2326 @kindex --disable-stdcall-fixup
2327 @item --enable-stdcall-fixup
2328 @itemx --disable-stdcall-fixup
2329 If the link finds a symbol that it cannot resolve, it will attempt to
2330 do ``fuzzy linking'' by looking for another defined symbol that differs
2331 only in the format of the symbol name (cdecl vs stdcall) and will
2332 resolve that symbol by linking to the match. For example, the
2333 undefined symbol @code{_foo} might be linked to the function
2334 @code{_foo@@12}, or the undefined symbol @code{_bar@@16} might be linked
2335 to the function @code{_bar}. When the linker does this, it prints a
2336 warning, since it normally should have failed to link, but sometimes
2337 import libraries generated from third-party dlls may need this feature
2338 to be usable. If you specify @option{--enable-stdcall-fixup}, this
2339 feature is fully enabled and warnings are not printed. If you specify
2340 @option{--disable-stdcall-fixup}, this feature is disabled and such
2341 mismatches are considered to be errors.
2342 [This option is specific to the i386 PE targeted port of the linker]
2344 @kindex --leading-underscore
2345 @kindex --no-leading-underscore
2346 @item --leading-underscore
2347 @itemx --no-leading-underscore
2348 For most targets default symbol-prefix is an underscore and is defined
2349 in target's description. By this option it is possible to
2350 disable/enable the default underscore symbol-prefix.
2352 @cindex DLLs, creating
2353 @kindex --export-all-symbols
2354 @item --export-all-symbols
2355 If given, all global symbols in the objects used to build a DLL will
2356 be exported by the DLL. Note that this is the default if there
2357 otherwise wouldn't be any exported symbols. When symbols are
2358 explicitly exported via DEF files or implicitly exported via function
2359 attributes, the default is to not export anything else unless this
2360 option is given. Note that the symbols @code{DllMain@@12},
2361 @code{DllEntryPoint@@0}, @code{DllMainCRTStartup@@12}, and
2362 @code{impure_ptr} will not be automatically
2363 exported. Also, symbols imported from other DLLs will not be
2364 re-exported, nor will symbols specifying the DLL's internal layout
2365 such as those beginning with @code{_head_} or ending with
2366 @code{_iname}. In addition, no symbols from @code{libgcc},
2367 @code{libstd++}, @code{libmingw32}, or @code{crtX.o} will be exported.
2368 Symbols whose names begin with @code{__rtti_} or @code{__builtin_} will
2369 not be exported, to help with C++ DLLs. Finally, there is an
2370 extensive list of cygwin-private symbols that are not exported
2371 (obviously, this applies on when building DLLs for cygwin targets).
2372 These cygwin-excludes are: @code{_cygwin_dll_entry@@12},
2373 @code{_cygwin_crt0_common@@8}, @code{_cygwin_noncygwin_dll_entry@@12},
2374 @code{_fmode}, @code{_impure_ptr}, @code{cygwin_attach_dll},
2375 @code{cygwin_premain0}, @code{cygwin_premain1}, @code{cygwin_premain2},
2376 @code{cygwin_premain3}, and @code{environ}.
2377 [This option is specific to the i386 PE targeted port of the linker]
2379 @kindex --exclude-symbols
2380 @item --exclude-symbols @var{symbol},@var{symbol},...
2381 Specifies a list of symbols which should not be automatically
2382 exported. The symbol names may be delimited by commas or colons.
2383 [This option is specific to the i386 PE targeted port of the linker]
2385 @kindex --exclude-all-symbols
2386 @item --exclude-all-symbols
2387 Specifies no symbols should be automatically exported.
2388 [This option is specific to the i386 PE targeted port of the linker]
2390 @kindex --file-alignment
2391 @item --file-alignment
2392 Specify the file alignment. Sections in the file will always begin at
2393 file offsets which are multiples of this number. This defaults to
2395 [This option is specific to the i386 PE targeted port of the linker]
2399 @item --heap @var{reserve}
2400 @itemx --heap @var{reserve},@var{commit}
2401 Specify the number of bytes of memory to reserve (and optionally commit)
2402 to be used as heap for this program. The default is 1MB reserved, 4K
2404 [This option is specific to the i386 PE targeted port of the linker]
2407 @kindex --image-base
2408 @item --image-base @var{value}
2409 Use @var{value} as the base address of your program or dll. This is
2410 the lowest memory location that will be used when your program or dll
2411 is loaded. To reduce the need to relocate and improve performance of
2412 your dlls, each should have a unique base address and not overlap any
2413 other dlls. The default is 0x400000 for executables, and 0x10000000
2415 [This option is specific to the i386 PE targeted port of the linker]
2419 If given, the stdcall suffixes (@@@var{nn}) will be stripped from
2420 symbols before they are exported.
2421 [This option is specific to the i386 PE targeted port of the linker]
2423 @kindex --large-address-aware
2424 @item --large-address-aware
2425 If given, the appropriate bit in the ``Characteristics'' field of the COFF
2426 header is set to indicate that this executable supports virtual addresses
2427 greater than 2 gigabytes. This should be used in conjunction with the /3GB
2428 or /USERVA=@var{value} megabytes switch in the ``[operating systems]''
2429 section of the BOOT.INI. Otherwise, this bit has no effect.
2430 [This option is specific to PE targeted ports of the linker]
2432 @kindex --disable-large-address-aware
2433 @item --disable-large-address-aware
2434 Reverts the effect of a previous @samp{--large-address-aware} option.
2435 This is useful if @samp{--large-address-aware} is always set by the compiler
2436 driver (e.g. Cygwin gcc) and the executable does not support virtual
2437 addresses greater than 2 gigabytes.
2438 [This option is specific to PE targeted ports of the linker]
2440 @kindex --major-image-version
2441 @item --major-image-version @var{value}
2442 Sets the major number of the ``image version''. Defaults to 1.
2443 [This option is specific to the i386 PE targeted port of the linker]
2445 @kindex --major-os-version
2446 @item --major-os-version @var{value}
2447 Sets the major number of the ``os version''. Defaults to 4.
2448 [This option is specific to the i386 PE targeted port of the linker]
2450 @kindex --major-subsystem-version
2451 @item --major-subsystem-version @var{value}
2452 Sets the major number of the ``subsystem version''. Defaults to 4.
2453 [This option is specific to the i386 PE targeted port of the linker]
2455 @kindex --minor-image-version
2456 @item --minor-image-version @var{value}
2457 Sets the minor number of the ``image version''. Defaults to 0.
2458 [This option is specific to the i386 PE targeted port of the linker]
2460 @kindex --minor-os-version
2461 @item --minor-os-version @var{value}
2462 Sets the minor number of the ``os version''. Defaults to 0.
2463 [This option is specific to the i386 PE targeted port of the linker]
2465 @kindex --minor-subsystem-version
2466 @item --minor-subsystem-version @var{value}
2467 Sets the minor number of the ``subsystem version''. Defaults to 0.
2468 [This option is specific to the i386 PE targeted port of the linker]
2470 @cindex DEF files, creating
2471 @cindex DLLs, creating
2472 @kindex --output-def
2473 @item --output-def @var{file}
2474 The linker will create the file @var{file} which will contain a DEF
2475 file corresponding to the DLL the linker is generating. This DEF file
2476 (which should be called @code{*.def}) may be used to create an import
2477 library with @code{dlltool} or may be used as a reference to
2478 automatically or implicitly exported symbols.
2479 [This option is specific to the i386 PE targeted port of the linker]
2481 @cindex DLLs, creating
2482 @kindex --out-implib
2483 @item --out-implib @var{file}
2484 The linker will create the file @var{file} which will contain an
2485 import lib corresponding to the DLL the linker is generating. This
2486 import lib (which should be called @code{*.dll.a} or @code{*.a}
2487 may be used to link clients against the generated DLL; this behaviour
2488 makes it possible to skip a separate @code{dlltool} import library
2490 [This option is specific to the i386 PE targeted port of the linker]
2492 @kindex --enable-auto-image-base
2493 @item --enable-auto-image-base
2494 @itemx --enable-auto-image-base=@var{value}
2495 Automatically choose the image base for DLLs, optionally starting with base
2496 @var{value}, unless one is specified using the @code{--image-base} argument.
2497 By using a hash generated from the dllname to create unique image bases
2498 for each DLL, in-memory collisions and relocations which can delay program
2499 execution are avoided.
2500 [This option is specific to the i386 PE targeted port of the linker]
2502 @kindex --disable-auto-image-base
2503 @item --disable-auto-image-base
2504 Do not automatically generate a unique image base. If there is no
2505 user-specified image base (@code{--image-base}) then use the platform
2507 [This option is specific to the i386 PE targeted port of the linker]
2509 @cindex DLLs, linking to
2510 @kindex --dll-search-prefix
2511 @item --dll-search-prefix @var{string}
2512 When linking dynamically to a dll without an import library,
2513 search for @code{<string><basename>.dll} in preference to
2514 @code{lib<basename>.dll}. This behaviour allows easy distinction
2515 between DLLs built for the various "subplatforms": native, cygwin,
2516 uwin, pw, etc. For instance, cygwin DLLs typically use
2517 @code{--dll-search-prefix=cyg}.
2518 [This option is specific to the i386 PE targeted port of the linker]
2520 @kindex --enable-auto-import
2521 @item --enable-auto-import
2522 Do sophisticated linking of @code{_symbol} to @code{__imp__symbol} for
2523 DATA imports from DLLs, and create the necessary thunking symbols when
2524 building the import libraries with those DATA exports. Note: Use of the
2525 'auto-import' extension will cause the text section of the image file
2526 to be made writable. This does not conform to the PE-COFF format
2527 specification published by Microsoft.
2529 Note - use of the 'auto-import' extension will also cause read only
2530 data which would normally be placed into the .rdata section to be
2531 placed into the .data section instead. This is in order to work
2532 around a problem with consts that is described here:
2533 http://www.cygwin.com/ml/cygwin/2004-09/msg01101.html
2535 Using 'auto-import' generally will 'just work' -- but sometimes you may
2538 "variable '<var>' can't be auto-imported. Please read the
2539 documentation for ld's @code{--enable-auto-import} for details."
2541 This message occurs when some (sub)expression accesses an address
2542 ultimately given by the sum of two constants (Win32 import tables only
2543 allow one). Instances where this may occur include accesses to member
2544 fields of struct variables imported from a DLL, as well as using a
2545 constant index into an array variable imported from a DLL. Any
2546 multiword variable (arrays, structs, long long, etc) may trigger
2547 this error condition. However, regardless of the exact data type
2548 of the offending exported variable, ld will always detect it, issue
2549 the warning, and exit.
2551 There are several ways to address this difficulty, regardless of the
2552 data type of the exported variable:
2554 One way is to use --enable-runtime-pseudo-reloc switch. This leaves the task
2555 of adjusting references in your client code for runtime environment, so
2556 this method works only when runtime environment supports this feature.
2558 A second solution is to force one of the 'constants' to be a variable --
2559 that is, unknown and un-optimizable at compile time. For arrays,
2560 there are two possibilities: a) make the indexee (the array's address)
2561 a variable, or b) make the 'constant' index a variable. Thus:
2564 extern type extern_array[];
2566 @{ volatile type *t=extern_array; t[1] @}
2572 extern type extern_array[];
2574 @{ volatile int t=1; extern_array[t] @}
2577 For structs (and most other multiword data types) the only option
2578 is to make the struct itself (or the long long, or the ...) variable:
2581 extern struct s extern_struct;
2582 extern_struct.field -->
2583 @{ volatile struct s *t=&extern_struct; t->field @}
2589 extern long long extern_ll;
2591 @{ volatile long long * local_ll=&extern_ll; *local_ll @}
2594 A third method of dealing with this difficulty is to abandon
2595 'auto-import' for the offending symbol and mark it with
2596 @code{__declspec(dllimport)}. However, in practice that
2597 requires using compile-time #defines to indicate whether you are
2598 building a DLL, building client code that will link to the DLL, or
2599 merely building/linking to a static library. In making the choice
2600 between the various methods of resolving the 'direct address with
2601 constant offset' problem, you should consider typical real-world usage:
2609 void main(int argc, char **argv)@{
2610 printf("%d\n",arr[1]);
2620 void main(int argc, char **argv)@{
2621 /* This workaround is for win32 and cygwin; do not "optimize" */
2622 volatile int *parr = arr;
2623 printf("%d\n",parr[1]);
2630 /* Note: auto-export is assumed (no __declspec(dllexport)) */
2631 #if (defined(_WIN32) || defined(__CYGWIN__)) && \
2632 !(defined(FOO_BUILD_DLL) || defined(FOO_STATIC))
2633 #define FOO_IMPORT __declspec(dllimport)
2637 extern FOO_IMPORT int arr[];
2640 void main(int argc, char **argv)@{
2641 printf("%d\n",arr[1]);
2645 A fourth way to avoid this problem is to re-code your
2646 library to use a functional interface rather than a data interface
2647 for the offending variables (e.g. set_foo() and get_foo() accessor
2649 [This option is specific to the i386 PE targeted port of the linker]
2651 @kindex --disable-auto-import
2652 @item --disable-auto-import
2653 Do not attempt to do sophisticated linking of @code{_symbol} to
2654 @code{__imp__symbol} for DATA imports from DLLs.
2655 [This option is specific to the i386 PE targeted port of the linker]
2657 @kindex --enable-runtime-pseudo-reloc
2658 @item --enable-runtime-pseudo-reloc
2659 If your code contains expressions described in --enable-auto-import section,
2660 that is, DATA imports from DLL with non-zero offset, this switch will create
2661 a vector of 'runtime pseudo relocations' which can be used by runtime
2662 environment to adjust references to such data in your client code.
2663 [This option is specific to the i386 PE targeted port of the linker]
2665 @kindex --disable-runtime-pseudo-reloc
2666 @item --disable-runtime-pseudo-reloc
2667 Do not create pseudo relocations for non-zero offset DATA imports from
2669 [This option is specific to the i386 PE targeted port of the linker]
2671 @kindex --enable-extra-pe-debug
2672 @item --enable-extra-pe-debug
2673 Show additional debug info related to auto-import symbol thunking.
2674 [This option is specific to the i386 PE targeted port of the linker]
2676 @kindex --section-alignment
2677 @item --section-alignment
2678 Sets the section alignment. Sections in memory will always begin at
2679 addresses which are a multiple of this number. Defaults to 0x1000.
2680 [This option is specific to the i386 PE targeted port of the linker]
2684 @item --stack @var{reserve}
2685 @itemx --stack @var{reserve},@var{commit}
2686 Specify the number of bytes of memory to reserve (and optionally commit)
2687 to be used as stack for this program. The default is 2MB reserved, 4K
2689 [This option is specific to the i386 PE targeted port of the linker]
2692 @item --subsystem @var{which}
2693 @itemx --subsystem @var{which}:@var{major}
2694 @itemx --subsystem @var{which}:@var{major}.@var{minor}
2695 Specifies the subsystem under which your program will execute. The
2696 legal values for @var{which} are @code{native}, @code{windows},
2697 @code{console}, @code{posix}, and @code{xbox}. You may optionally set
2698 the subsystem version also. Numeric values are also accepted for
2700 [This option is specific to the i386 PE targeted port of the linker]
2702 The following options set flags in the @code{DllCharacteristics} field
2703 of the PE file header:
2704 [These options are specific to PE targeted ports of the linker]
2706 @kindex --high-entropy-va
2707 @item --high-entropy-va
2708 Image is compatible with 64-bit address space layout randomization
2711 @kindex --dynamicbase
2713 The image base address may be relocated using address space layout
2714 randomization (ASLR). This feature was introduced with MS Windows
2715 Vista for i386 PE targets.
2717 @kindex --forceinteg
2719 Code integrity checks are enforced.
2723 The image is compatible with the Data Execution Prevention.
2724 This feature was introduced with MS Windows XP SP2 for i386 PE targets.
2726 @kindex --no-isolation
2727 @item --no-isolation
2728 Although the image understands isolation, do not isolate the image.
2732 The image does not use SEH. No SE handler may be called from
2737 Do not bind this image.
2741 The driver uses the MS Windows Driver Model.
2745 The image is Terminal Server aware.
2747 @kindex --insert-timestamp
2748 @item --insert-timestamp
2749 @itemx --no-insert-timestamp
2750 Insert a real timestamp into the image. This is the default behaviour
2751 as it matches legacy code and it means that the image will work with
2752 other, proprietary tools. The problem with this default is that it
2753 will result in slightly different images being produced each tiem the
2754 same sources are linked. The option @option{--no-insert-timestamp}
2755 can be used to insert a zero value for the timestamp, this ensuring
2756 that binaries produced from indentical sources will compare
2763 @subsection Options specific to C6X uClinux targets
2765 @c man begin OPTIONS
2767 The C6X uClinux target uses a binary format called DSBT to support shared
2768 libraries. Each shared library in the system needs to have a unique index;
2769 all executables use an index of 0.
2774 @item --dsbt-size @var{size}
2775 This option sets the number of entires in the DSBT of the current executable
2776 or shared library to @var{size}. The default is to create a table with 64
2779 @kindex --dsbt-index
2780 @item --dsbt-index @var{index}
2781 This option sets the DSBT index of the current executable or shared library
2782 to @var{index}. The default is 0, which is appropriate for generating
2783 executables. If a shared library is generated with a DSBT index of 0, the
2784 @code{R_C6000_DSBT_INDEX} relocs are copied into the output file.
2786 @kindex --no-merge-exidx-entries
2787 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent
2788 exidx entries in frame unwind info.
2796 @subsection Options specific to Motorola 68HC11 and 68HC12 targets
2798 @c man begin OPTIONS
2800 The 68HC11 and 68HC12 linkers support specific options to control the
2801 memory bank switching mapping and trampoline code generation.
2805 @kindex --no-trampoline
2806 @item --no-trampoline
2807 This option disables the generation of trampoline. By default a trampoline
2808 is generated for each far function which is called using a @code{jsr}
2809 instruction (this happens when a pointer to a far function is taken).
2811 @kindex --bank-window
2812 @item --bank-window @var{name}
2813 This option indicates to the linker the name of the memory region in
2814 the @samp{MEMORY} specification that describes the memory bank window.
2815 The definition of such region is then used by the linker to compute
2816 paging and addresses within the memory window.
2824 @subsection Options specific to Motorola 68K target
2826 @c man begin OPTIONS
2828 The following options are supported to control handling of GOT generation
2829 when linking for 68K targets.
2834 @item --got=@var{type}
2835 This option tells the linker which GOT generation scheme to use.
2836 @var{type} should be one of @samp{single}, @samp{negative},
2837 @samp{multigot} or @samp{target}. For more information refer to the
2838 Info entry for @file{ld}.
2846 @subsection Options specific to MIPS targets
2848 @c man begin OPTIONS
2850 The following options are supported to control microMIPS instruction
2851 generation when linking for MIPS targets.
2859 These options control the choice of microMIPS instructions used in code
2860 generated by the linker, such as that in the PLT or lazy binding stubs,
2861 or in relaxation. If @samp{--insn32} is used, then the linker only uses
2862 32-bit instruction encodings. By default or if @samp{--no-insn32} is
2863 used, all instruction encodings are used, including 16-bit ones where
2873 @section Environment Variables
2875 @c man begin ENVIRONMENT
2877 You can change the behaviour of @command{ld} with the environment variables
2878 @ifclear SingleFormat
2881 @code{LDEMULATION} and @code{COLLECT_NO_DEMANGLE}.
2883 @ifclear SingleFormat
2885 @cindex default input format
2886 @code{GNUTARGET} determines the input-file object format if you don't
2887 use @samp{-b} (or its synonym @samp{--format}). Its value should be one
2888 of the BFD names for an input format (@pxref{BFD}). If there is no
2889 @code{GNUTARGET} in the environment, @command{ld} uses the natural format
2890 of the target. If @code{GNUTARGET} is set to @code{default} then BFD
2891 attempts to discover the input format by examining binary input files;
2892 this method often succeeds, but there are potential ambiguities, since
2893 there is no method of ensuring that the magic number used to specify
2894 object-file formats is unique. However, the configuration procedure for
2895 BFD on each system places the conventional format for that system first
2896 in the search-list, so ambiguities are resolved in favor of convention.
2900 @cindex default emulation
2901 @cindex emulation, default
2902 @code{LDEMULATION} determines the default emulation if you don't use the
2903 @samp{-m} option. The emulation can affect various aspects of linker
2904 behaviour, particularly the default linker script. You can list the
2905 available emulations with the @samp{--verbose} or @samp{-V} options. If
2906 the @samp{-m} option is not used, and the @code{LDEMULATION} environment
2907 variable is not defined, the default emulation depends upon how the
2908 linker was configured.
2910 @kindex COLLECT_NO_DEMANGLE
2911 @cindex demangling, default
2912 Normally, the linker will default to demangling symbols. However, if
2913 @code{COLLECT_NO_DEMANGLE} is set in the environment, then it will
2914 default to not demangling symbols. This environment variable is used in
2915 a similar fashion by the @code{gcc} linker wrapper program. The default
2916 may be overridden by the @samp{--demangle} and @samp{--no-demangle}
2923 @chapter Linker Scripts
2926 @cindex linker scripts
2927 @cindex command files
2928 Every link is controlled by a @dfn{linker script}. This script is
2929 written in the linker command language.
2931 The main purpose of the linker script is to describe how the sections in
2932 the input files should be mapped into the output file, and to control
2933 the memory layout of the output file. Most linker scripts do nothing
2934 more than this. However, when necessary, the linker script can also
2935 direct the linker to perform many other operations, using the commands
2938 The linker always uses a linker script. If you do not supply one
2939 yourself, the linker will use a default script that is compiled into the
2940 linker executable. You can use the @samp{--verbose} command line option
2941 to display the default linker script. Certain command line options,
2942 such as @samp{-r} or @samp{-N}, will affect the default linker script.
2944 You may supply your own linker script by using the @samp{-T} command
2945 line option. When you do this, your linker script will replace the
2946 default linker script.
2948 You may also use linker scripts implicitly by naming them as input files
2949 to the linker, as though they were files to be linked. @xref{Implicit
2953 * Basic Script Concepts:: Basic Linker Script Concepts
2954 * Script Format:: Linker Script Format
2955 * Simple Example:: Simple Linker Script Example
2956 * Simple Commands:: Simple Linker Script Commands
2957 * Assignments:: Assigning Values to Symbols
2958 * SECTIONS:: SECTIONS Command
2959 * MEMORY:: MEMORY Command
2960 * PHDRS:: PHDRS Command
2961 * VERSION:: VERSION Command
2962 * Expressions:: Expressions in Linker Scripts
2963 * Implicit Linker Scripts:: Implicit Linker Scripts
2966 @node Basic Script Concepts
2967 @section Basic Linker Script Concepts
2968 @cindex linker script concepts
2969 We need to define some basic concepts and vocabulary in order to
2970 describe the linker script language.
2972 The linker combines input files into a single output file. The output
2973 file and each input file are in a special data format known as an
2974 @dfn{object file format}. Each file is called an @dfn{object file}.
2975 The output file is often called an @dfn{executable}, but for our
2976 purposes we will also call it an object file. Each object file has,
2977 among other things, a list of @dfn{sections}. We sometimes refer to a
2978 section in an input file as an @dfn{input section}; similarly, a section
2979 in the output file is an @dfn{output section}.
2981 Each section in an object file has a name and a size. Most sections
2982 also have an associated block of data, known as the @dfn{section
2983 contents}. A section may be marked as @dfn{loadable}, which means that
2984 the contents should be loaded into memory when the output file is run.
2985 A section with no contents may be @dfn{allocatable}, which means that an
2986 area in memory should be set aside, but nothing in particular should be
2987 loaded there (in some cases this memory must be zeroed out). A section
2988 which is neither loadable nor allocatable typically contains some sort
2989 of debugging information.
2991 Every loadable or allocatable output section has two addresses. The
2992 first is the @dfn{VMA}, or virtual memory address. This is the address
2993 the section will have when the output file is run. The second is the
2994 @dfn{LMA}, or load memory address. This is the address at which the
2995 section will be loaded. In most cases the two addresses will be the
2996 same. An example of when they might be different is when a data section
2997 is loaded into ROM, and then copied into RAM when the program starts up
2998 (this technique is often used to initialize global variables in a ROM
2999 based system). In this case the ROM address would be the LMA, and the
3000 RAM address would be the VMA.
3002 You can see the sections in an object file by using the @code{objdump}
3003 program with the @samp{-h} option.
3005 Every object file also has a list of @dfn{symbols}, known as the
3006 @dfn{symbol table}. A symbol may be defined or undefined. Each symbol
3007 has a name, and each defined symbol has an address, among other
3008 information. If you compile a C or C++ program into an object file, you
3009 will get a defined symbol for every defined function and global or
3010 static variable. Every undefined function or global variable which is
3011 referenced in the input file will become an undefined symbol.
3013 You can see the symbols in an object file by using the @code{nm}
3014 program, or by using the @code{objdump} program with the @samp{-t}
3018 @section Linker Script Format
3019 @cindex linker script format
3020 Linker scripts are text files.
3022 You write a linker script as a series of commands. Each command is
3023 either a keyword, possibly followed by arguments, or an assignment to a
3024 symbol. You may separate commands using semicolons. Whitespace is
3027 Strings such as file or format names can normally be entered directly.
3028 If the file name contains a character such as a comma which would
3029 otherwise serve to separate file names, you may put the file name in
3030 double quotes. There is no way to use a double quote character in a
3033 You may include comments in linker scripts just as in C, delimited by
3034 @samp{/*} and @samp{*/}. As in C, comments are syntactically equivalent
3037 @node Simple Example
3038 @section Simple Linker Script Example
3039 @cindex linker script example
3040 @cindex example of linker script
3041 Many linker scripts are fairly simple.
3043 The simplest possible linker script has just one command:
3044 @samp{SECTIONS}. You use the @samp{SECTIONS} command to describe the
3045 memory layout of the output file.
3047 The @samp{SECTIONS} command is a powerful command. Here we will
3048 describe a simple use of it. Let's assume your program consists only of
3049 code, initialized data, and uninitialized data. These will be in the
3050 @samp{.text}, @samp{.data}, and @samp{.bss} sections, respectively.
3051 Let's assume further that these are the only sections which appear in
3054 For this example, let's say that the code should be loaded at address
3055 0x10000, and that the data should start at address 0x8000000. Here is a
3056 linker script which will do that:
3061 .text : @{ *(.text) @}
3063 .data : @{ *(.data) @}
3064 .bss : @{ *(.bss) @}
3068 You write the @samp{SECTIONS} command as the keyword @samp{SECTIONS},
3069 followed by a series of symbol assignments and output section
3070 descriptions enclosed in curly braces.
3072 The first line inside the @samp{SECTIONS} command of the above example
3073 sets the value of the special symbol @samp{.}, which is the location
3074 counter. If you do not specify the address of an output section in some
3075 other way (other ways are described later), the address is set from the
3076 current value of the location counter. The location counter is then
3077 incremented by the size of the output section. At the start of the
3078 @samp{SECTIONS} command, the location counter has the value @samp{0}.
3080 The second line defines an output section, @samp{.text}. The colon is
3081 required syntax which may be ignored for now. Within the curly braces
3082 after the output section name, you list the names of the input sections
3083 which should be placed into this output section. The @samp{*} is a
3084 wildcard which matches any file name. The expression @samp{*(.text)}
3085 means all @samp{.text} input sections in all input files.
3087 Since the location counter is @samp{0x10000} when the output section
3088 @samp{.text} is defined, the linker will set the address of the
3089 @samp{.text} section in the output file to be @samp{0x10000}.
3091 The remaining lines define the @samp{.data} and @samp{.bss} sections in
3092 the output file. The linker will place the @samp{.data} output section
3093 at address @samp{0x8000000}. After the linker places the @samp{.data}
3094 output section, the value of the location counter will be
3095 @samp{0x8000000} plus the size of the @samp{.data} output section. The
3096 effect is that the linker will place the @samp{.bss} output section
3097 immediately after the @samp{.data} output section in memory.
3099 The linker will ensure that each output section has the required
3100 alignment, by increasing the location counter if necessary. In this
3101 example, the specified addresses for the @samp{.text} and @samp{.data}
3102 sections will probably satisfy any alignment constraints, but the linker
3103 may have to create a small gap between the @samp{.data} and @samp{.bss}
3106 That's it! That's a simple and complete linker script.
3108 @node Simple Commands
3109 @section Simple Linker Script Commands
3110 @cindex linker script simple commands
3111 In this section we describe the simple linker script commands.
3114 * Entry Point:: Setting the entry point
3115 * File Commands:: Commands dealing with files
3116 @ifclear SingleFormat
3117 * Format Commands:: Commands dealing with object file formats
3120 * REGION_ALIAS:: Assign alias names to memory regions
3121 * Miscellaneous Commands:: Other linker script commands
3125 @subsection Setting the Entry Point
3126 @kindex ENTRY(@var{symbol})
3127 @cindex start of execution
3128 @cindex first instruction
3130 The first instruction to execute in a program is called the @dfn{entry
3131 point}. You can use the @code{ENTRY} linker script command to set the
3132 entry point. The argument is a symbol name:
3137 There are several ways to set the entry point. The linker will set the
3138 entry point by trying each of the following methods in order, and
3139 stopping when one of them succeeds:
3142 the @samp{-e} @var{entry} command-line option;
3144 the @code{ENTRY(@var{symbol})} command in a linker script;
3146 the value of a target specific symbol, if it is defined; For many
3147 targets this is @code{start}, but PE and BeOS based systems for example
3148 check a list of possible entry symbols, matching the first one found.
3150 the address of the first byte of the @samp{.text} section, if present;
3152 The address @code{0}.
3156 @subsection Commands Dealing with Files
3157 @cindex linker script file commands
3158 Several linker script commands deal with files.
3161 @item INCLUDE @var{filename}
3162 @kindex INCLUDE @var{filename}
3163 @cindex including a linker script
3164 Include the linker script @var{filename} at this point. The file will
3165 be searched for in the current directory, and in any directory specified
3166 with the @option{-L} option. You can nest calls to @code{INCLUDE} up to
3169 You can place @code{INCLUDE} directives at the top level, in @code{MEMORY} or
3170 @code{SECTIONS} commands, or in output section descriptions.
3172 @item INPUT(@var{file}, @var{file}, @dots{})
3173 @itemx INPUT(@var{file} @var{file} @dots{})
3174 @kindex INPUT(@var{files})
3175 @cindex input files in linker scripts
3176 @cindex input object files in linker scripts
3177 @cindex linker script input object files
3178 The @code{INPUT} command directs the linker to include the named files
3179 in the link, as though they were named on the command line.
3181 For example, if you always want to include @file{subr.o} any time you do
3182 a link, but you can't be bothered to put it on every link command line,
3183 then you can put @samp{INPUT (subr.o)} in your linker script.
3185 In fact, if you like, you can list all of your input files in the linker
3186 script, and then invoke the linker with nothing but a @samp{-T} option.
3188 In case a @dfn{sysroot prefix} is configured, and the filename starts
3189 with the @samp{/} character, and the script being processed was
3190 located inside the @dfn{sysroot prefix}, the filename will be looked
3191 for in the @dfn{sysroot prefix}. Otherwise, the linker will try to
3192 open the file in the current directory. If it is not found, the
3193 linker will search through the archive library search path.
3194 The @dfn{sysroot prefix} can also be forced by specifying @code{=}
3195 as the first character in the filename path. See also the
3196 description of @samp{-L} in @ref{Options,,Command Line Options}.
3198 If you use @samp{INPUT (-l@var{file})}, @command{ld} will transform the
3199 name to @code{lib@var{file}.a}, as with the command line argument
3202 When you use the @code{INPUT} command in an implicit linker script, the
3203 files will be included in the link at the point at which the linker
3204 script file is included. This can affect archive searching.
3206 @item GROUP(@var{file}, @var{file}, @dots{})
3207 @itemx GROUP(@var{file} @var{file} @dots{})
3208 @kindex GROUP(@var{files})
3209 @cindex grouping input files
3210 The @code{GROUP} command is like @code{INPUT}, except that the named
3211 files should all be archives, and they are searched repeatedly until no
3212 new undefined references are created. See the description of @samp{-(}
3213 in @ref{Options,,Command Line Options}.
3215 @item AS_NEEDED(@var{file}, @var{file}, @dots{})
3216 @itemx AS_NEEDED(@var{file} @var{file} @dots{})
3217 @kindex AS_NEEDED(@var{files})
3218 This construct can appear only inside of the @code{INPUT} or @code{GROUP}
3219 commands, among other filenames. The files listed will be handled
3220 as if they appear directly in the @code{INPUT} or @code{GROUP} commands,
3221 with the exception of ELF shared libraries, that will be added only
3222 when they are actually needed. This construct essentially enables
3223 @option{--as-needed} option for all the files listed inside of it
3224 and restores previous @option{--as-needed} resp. @option{--no-as-needed}
3227 @item OUTPUT(@var{filename})
3228 @kindex OUTPUT(@var{filename})
3229 @cindex output file name in linker script
3230 The @code{OUTPUT} command names the output file. Using
3231 @code{OUTPUT(@var{filename})} in the linker script is exactly like using
3232 @samp{-o @var{filename}} on the command line (@pxref{Options,,Command
3233 Line Options}). If both are used, the command line option takes
3236 You can use the @code{OUTPUT} command to define a default name for the
3237 output file other than the usual default of @file{a.out}.
3239 @item SEARCH_DIR(@var{path})
3240 @kindex SEARCH_DIR(@var{path})
3241 @cindex library search path in linker script
3242 @cindex archive search path in linker script
3243 @cindex search path in linker script
3244 The @code{SEARCH_DIR} command adds @var{path} to the list of paths where
3245 @command{ld} looks for archive libraries. Using
3246 @code{SEARCH_DIR(@var{path})} is exactly like using @samp{-L @var{path}}
3247 on the command line (@pxref{Options,,Command Line Options}). If both
3248 are used, then the linker will search both paths. Paths specified using
3249 the command line option are searched first.
3251 @item STARTUP(@var{filename})
3252 @kindex STARTUP(@var{filename})
3253 @cindex first input file
3254 The @code{STARTUP} command is just like the @code{INPUT} command, except
3255 that @var{filename} will become the first input file to be linked, as
3256 though it were specified first on the command line. This may be useful
3257 when using a system in which the entry point is always the start of the
3261 @ifclear SingleFormat
3262 @node Format Commands
3263 @subsection Commands Dealing with Object File Formats
3264 A couple of linker script commands deal with object file formats.
3267 @item OUTPUT_FORMAT(@var{bfdname})
3268 @itemx OUTPUT_FORMAT(@var{default}, @var{big}, @var{little})
3269 @kindex OUTPUT_FORMAT(@var{bfdname})
3270 @cindex output file format in linker script
3271 The @code{OUTPUT_FORMAT} command names the BFD format to use for the
3272 output file (@pxref{BFD}). Using @code{OUTPUT_FORMAT(@var{bfdname})} is
3273 exactly like using @samp{--oformat @var{bfdname}} on the command line
3274 (@pxref{Options,,Command Line Options}). If both are used, the command
3275 line option takes precedence.
3277 You can use @code{OUTPUT_FORMAT} with three arguments to use different
3278 formats based on the @samp{-EB} and @samp{-EL} command line options.
3279 This permits the linker script to set the output format based on the
3282 If neither @samp{-EB} nor @samp{-EL} are used, then the output format
3283 will be the first argument, @var{default}. If @samp{-EB} is used, the
3284 output format will be the second argument, @var{big}. If @samp{-EL} is
3285 used, the output format will be the third argument, @var{little}.
3287 For example, the default linker script for the MIPS ELF target uses this
3290 OUTPUT_FORMAT(elf32-bigmips, elf32-bigmips, elf32-littlemips)
3292 This says that the default format for the output file is
3293 @samp{elf32-bigmips}, but if the user uses the @samp{-EL} command line
3294 option, the output file will be created in the @samp{elf32-littlemips}
3297 @item TARGET(@var{bfdname})
3298 @kindex TARGET(@var{bfdname})
3299 @cindex input file format in linker script
3300 The @code{TARGET} command names the BFD format to use when reading input
3301 files. It affects subsequent @code{INPUT} and @code{GROUP} commands.
3302 This command is like using @samp{-b @var{bfdname}} on the command line
3303 (@pxref{Options,,Command Line Options}). If the @code{TARGET} command
3304 is used but @code{OUTPUT_FORMAT} is not, then the last @code{TARGET}
3305 command is also used to set the format for the output file. @xref{BFD}.
3310 @subsection Assign alias names to memory regions
3311 @kindex REGION_ALIAS(@var{alias}, @var{region})
3312 @cindex region alias
3313 @cindex region names
3315 Alias names can be added to existing memory regions created with the
3316 @ref{MEMORY} command. Each name corresponds to at most one memory region.
3319 REGION_ALIAS(@var{alias}, @var{region})
3322 The @code{REGION_ALIAS} function creates an alias name @var{alias} for the
3323 memory region @var{region}. This allows a flexible mapping of output sections
3324 to memory regions. An example follows.
3326 Suppose we have an application for embedded systems which come with various
3327 memory storage devices. All have a general purpose, volatile memory @code{RAM}
3328 that allows code execution or data storage. Some may have a read-only,
3329 non-volatile memory @code{ROM} that allows code execution and read-only data
3330 access. The last variant is a read-only, non-volatile memory @code{ROM2} with
3331 read-only data access and no code execution capability. We have four output
3336 @code{.text} program code;
3338 @code{.rodata} read-only data;
3340 @code{.data} read-write initialized data;
3342 @code{.bss} read-write zero initialized data.
3345 The goal is to provide a linker command file that contains a system independent
3346 part defining the output sections and a system dependent part mapping the
3347 output sections to the memory regions available on the system. Our embedded
3348 systems come with three different memory setups @code{A}, @code{B} and
3350 @multitable @columnfractions .25 .25 .25 .25
3351 @item Section @tab Variant A @tab Variant B @tab Variant C
3352 @item .text @tab RAM @tab ROM @tab ROM
3353 @item .rodata @tab RAM @tab ROM @tab ROM2
3354 @item .data @tab RAM @tab RAM/ROM @tab RAM/ROM2
3355 @item .bss @tab RAM @tab RAM @tab RAM
3357 The notation @code{RAM/ROM} or @code{RAM/ROM2} means that this section is
3358 loaded into region @code{ROM} or @code{ROM2} respectively. Please note that
3359 the load address of the @code{.data} section starts in all three variants at
3360 the end of the @code{.rodata} section.
3362 The base linker script that deals with the output sections follows. It
3363 includes the system dependent @code{linkcmds.memory} file that describes the
3366 INCLUDE linkcmds.memory
3379 .data : AT (rodata_end)
3384 data_size = SIZEOF(.data);
3385 data_load_start = LOADADDR(.data);
3393 Now we need three different @code{linkcmds.memory} files to define memory
3394 regions and alias names. The content of @code{linkcmds.memory} for the three
3395 variants @code{A}, @code{B} and @code{C}:
3398 Here everything goes into the @code{RAM}.
3402 RAM : ORIGIN = 0, LENGTH = 4M
3405 REGION_ALIAS("REGION_TEXT", RAM);
3406 REGION_ALIAS("REGION_RODATA", RAM);
3407 REGION_ALIAS("REGION_DATA", RAM);
3408 REGION_ALIAS("REGION_BSS", RAM);
3411 Program code and read-only data go into the @code{ROM}. Read-write data goes
3412 into the @code{RAM}. An image of the initialized data is loaded into the
3413 @code{ROM} and will be copied during system start into the @code{RAM}.
3417 ROM : ORIGIN = 0, LENGTH = 3M
3418 RAM : ORIGIN = 0x10000000, LENGTH = 1M
3421 REGION_ALIAS("REGION_TEXT", ROM);
3422 REGION_ALIAS("REGION_RODATA", ROM);
3423 REGION_ALIAS("REGION_DATA", RAM);
3424 REGION_ALIAS("REGION_BSS", RAM);
3427 Program code goes into the @code{ROM}. Read-only data goes into the
3428 @code{ROM2}. Read-write data goes into the @code{RAM}. An image of the
3429 initialized data is loaded into the @code{ROM2} and will be copied during
3430 system start into the @code{RAM}.
3434 ROM : ORIGIN = 0, LENGTH = 2M
3435 ROM2 : ORIGIN = 0x10000000, LENGTH = 1M
3436 RAM : ORIGIN = 0x20000000, LENGTH = 1M
3439 REGION_ALIAS("REGION_TEXT", ROM);
3440 REGION_ALIAS("REGION_RODATA", ROM2);
3441 REGION_ALIAS("REGION_DATA", RAM);
3442 REGION_ALIAS("REGION_BSS", RAM);
3446 It is possible to write a common system initialization routine to copy the
3447 @code{.data} section from @code{ROM} or @code{ROM2} into the @code{RAM} if
3452 extern char data_start [];
3453 extern char data_size [];
3454 extern char data_load_start [];
3456 void copy_data(void)
3458 if (data_start != data_load_start)
3460 memcpy(data_start, data_load_start, (size_t) data_size);
3465 @node Miscellaneous Commands
3466 @subsection Other Linker Script Commands
3467 There are a few other linker scripts commands.
3470 @item ASSERT(@var{exp}, @var{message})
3472 @cindex assertion in linker script
3473 Ensure that @var{exp} is non-zero. If it is zero, then exit the linker
3474 with an error code, and print @var{message}.
3476 Note that assertions are checked before the final stages of linking
3477 take place. This means that expressions involving symbols PROVIDEd
3478 inside section definitions will fail if the user has not set values
3479 for those symbols. The only exception to this rule is PROVIDEd
3480 symbols that just reference dot. Thus an assertion like this:
3485 PROVIDE (__stack = .);
3486 PROVIDE (__stack_size = 0x100);
3487 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
3491 will fail if @code{__stack_size} is not defined elsewhere. Symbols
3492 PROVIDEd outside of section definitions are evaluated earlier, so they
3493 can be used inside ASSERTions. Thus:
3496 PROVIDE (__stack_size = 0x100);
3499 PROVIDE (__stack = .);
3500 ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack");
3506 @item EXTERN(@var{symbol} @var{symbol} @dots{})
3508 @cindex undefined symbol in linker script
3509 Force @var{symbol} to be entered in the output file as an undefined
3510 symbol. Doing this may, for example, trigger linking of additional
3511 modules from standard libraries. You may list several @var{symbol}s for
3512 each @code{EXTERN}, and you may use @code{EXTERN} multiple times. This
3513 command has the same effect as the @samp{-u} command-line option.
3515 @item FORCE_COMMON_ALLOCATION
3516 @kindex FORCE_COMMON_ALLOCATION
3517 @cindex common allocation in linker script
3518 This command has the same effect as the @samp{-d} command-line option:
3519 to make @command{ld} assign space to common symbols even if a relocatable
3520 output file is specified (@samp{-r}).
3522 @item INHIBIT_COMMON_ALLOCATION
3523 @kindex INHIBIT_COMMON_ALLOCATION
3524 @cindex common allocation in linker script
3525 This command has the same effect as the @samp{--no-define-common}
3526 command-line option: to make @code{ld} omit the assignment of addresses
3527 to common symbols even for a non-relocatable output file.
3529 @item INSERT [ AFTER | BEFORE ] @var{output_section}
3531 @cindex insert user script into default script
3532 This command is typically used in a script specified by @samp{-T} to
3533 augment the default @code{SECTIONS} with, for example, overlays. It
3534 inserts all prior linker script statements after (or before)
3535 @var{output_section}, and also causes @samp{-T} to not override the
3536 default linker script. The exact insertion point is as for orphan
3537 sections. @xref{Location Counter}. The insertion happens after the
3538 linker has mapped input sections to output sections. Prior to the
3539 insertion, since @samp{-T} scripts are parsed before the default
3540 linker script, statements in the @samp{-T} script occur before the
3541 default linker script statements in the internal linker representation
3542 of the script. In particular, input section assignments will be made
3543 to @samp{-T} output sections before those in the default script. Here
3544 is an example of how a @samp{-T} script using @code{INSERT} might look:
3551 .ov1 @{ ov1*(.text) @}
3552 .ov2 @{ ov2*(.text) @}
3558 @item NOCROSSREFS(@var{section} @var{section} @dots{})
3559 @kindex NOCROSSREFS(@var{sections})
3560 @cindex cross references
3561 This command may be used to tell @command{ld} to issue an error about any
3562 references among certain output sections.
3564 In certain types of programs, particularly on embedded systems when
3565 using overlays, when one section is loaded into memory, another section
3566 will not be. Any direct references between the two sections would be
3567 errors. For example, it would be an error if code in one section called
3568 a function defined in the other section.
3570 The @code{NOCROSSREFS} command takes a list of output section names. If
3571 @command{ld} detects any cross references between the sections, it reports
3572 an error and returns a non-zero exit status. Note that the
3573 @code{NOCROSSREFS} command uses output section names, not input section
3576 @ifclear SingleFormat
3577 @item OUTPUT_ARCH(@var{bfdarch})
3578 @kindex OUTPUT_ARCH(@var{bfdarch})
3579 @cindex machine architecture
3580 @cindex architecture
3581 Specify a particular output machine architecture. The argument is one
3582 of the names used by the BFD library (@pxref{BFD}). You can see the
3583 architecture of an object file by using the @code{objdump} program with
3584 the @samp{-f} option.
3587 @item LD_FEATURE(@var{string})
3588 @kindex LD_FEATURE(@var{string})
3589 This command may be used to modify @command{ld} behavior. If
3590 @var{string} is @code{"SANE_EXPR"} then absolute symbols and numbers
3591 in a script are simply treated as numbers everywhere.
3592 @xref{Expression Section}.
3596 @section Assigning Values to Symbols
3597 @cindex assignment in scripts
3598 @cindex symbol definition, scripts
3599 @cindex variables, defining
3600 You may assign a value to a symbol in a linker script. This will define
3601 the symbol and place it into the symbol table with a global scope.
3604 * Simple Assignments:: Simple Assignments
3607 * PROVIDE_HIDDEN:: PROVIDE_HIDDEN
3608 * Source Code Reference:: How to use a linker script defined symbol in source code
3611 @node Simple Assignments
3612 @subsection Simple Assignments
3614 You may assign to a symbol using any of the C assignment operators:
3617 @item @var{symbol} = @var{expression} ;
3618 @itemx @var{symbol} += @var{expression} ;
3619 @itemx @var{symbol} -= @var{expression} ;
3620 @itemx @var{symbol} *= @var{expression} ;
3621 @itemx @var{symbol} /= @var{expression} ;
3622 @itemx @var{symbol} <<= @var{expression} ;
3623 @itemx @var{symbol} >>= @var{expression} ;
3624 @itemx @var{symbol} &= @var{expression} ;
3625 @itemx @var{symbol} |= @var{expression} ;
3628 The first case will define @var{symbol} to the value of
3629 @var{expression}. In the other cases, @var{symbol} must already be
3630 defined, and the value will be adjusted accordingly.
3632 The special symbol name @samp{.} indicates the location counter. You
3633 may only use this within a @code{SECTIONS} command. @xref{Location Counter}.
3635 The semicolon after @var{expression} is required.
3637 Expressions are defined below; see @ref{Expressions}.
3639 You may write symbol assignments as commands in their own right, or as
3640 statements within a @code{SECTIONS} command, or as part of an output
3641 section description in a @code{SECTIONS} command.
3643 The section of the symbol will be set from the section of the
3644 expression; for more information, see @ref{Expression Section}.
3646 Here is an example showing the three different places that symbol
3647 assignments may be used:
3658 _bdata = (. + 3) & ~ 3;
3659 .data : @{ *(.data) @}
3663 In this example, the symbol @samp{floating_point} will be defined as
3664 zero. The symbol @samp{_etext} will be defined as the address following
3665 the last @samp{.text} input section. The symbol @samp{_bdata} will be
3666 defined as the address following the @samp{.text} output section aligned
3667 upward to a 4 byte boundary.
3672 For ELF targeted ports, define a symbol that will be hidden and won't be
3673 exported. The syntax is @code{HIDDEN(@var{symbol} = @var{expression})}.
3675 Here is the example from @ref{Simple Assignments}, rewritten to use
3679 HIDDEN(floating_point = 0);
3687 HIDDEN(_bdata = (. + 3) & ~ 3);
3688 .data : @{ *(.data) @}
3692 In this case none of the three symbols will be visible outside this module.
3697 In some cases, it is desirable for a linker script to define a symbol
3698 only if it is referenced and is not defined by any object included in
3699 the link. For example, traditional linkers defined the symbol
3700 @samp{etext}. However, ANSI C requires that the user be able to use
3701 @samp{etext} as a function name without encountering an error. The
3702 @code{PROVIDE} keyword may be used to define a symbol, such as
3703 @samp{etext}, only if it is referenced but not defined. The syntax is
3704 @code{PROVIDE(@var{symbol} = @var{expression})}.
3706 Here is an example of using @code{PROVIDE} to define @samp{etext}:
3719 In this example, if the program defines @samp{_etext} (with a leading
3720 underscore), the linker will give a multiple definition error. If, on
3721 the other hand, the program defines @samp{etext} (with no leading
3722 underscore), the linker will silently use the definition in the program.
3723 If the program references @samp{etext} but does not define it, the
3724 linker will use the definition in the linker script.
3726 @node PROVIDE_HIDDEN
3727 @subsection PROVIDE_HIDDEN
3728 @cindex PROVIDE_HIDDEN
3729 Similar to @code{PROVIDE}. For ELF targeted ports, the symbol will be
3730 hidden and won't be exported.
3732 @node Source Code Reference
3733 @subsection Source Code Reference
3735 Accessing a linker script defined variable from source code is not
3736 intuitive. In particular a linker script symbol is not equivalent to
3737 a variable declaration in a high level language, it is instead a
3738 symbol that does not have a value.
3740 Before going further, it is important to note that compilers often
3741 transform names in the source code into different names when they are
3742 stored in the symbol table. For example, Fortran compilers commonly
3743 prepend or append an underscore, and C++ performs extensive @samp{name
3744 mangling}. Therefore there might be a discrepancy between the name
3745 of a variable as it is used in source code and the name of the same
3746 variable as it is defined in a linker script. For example in C a
3747 linker script variable might be referred to as:
3753 But in the linker script it might be defined as:
3759 In the remaining examples however it is assumed that no name
3760 transformation has taken place.
3762 When a symbol is declared in a high level language such as C, two
3763 things happen. The first is that the compiler reserves enough space
3764 in the program's memory to hold the @emph{value} of the symbol. The
3765 second is that the compiler creates an entry in the program's symbol
3766 table which holds the symbol's @emph{address}. ie the symbol table
3767 contains the address of the block of memory holding the symbol's
3768 value. So for example the following C declaration, at file scope:
3774 creates an entry called @samp{foo} in the symbol table. This entry
3775 holds the address of an @samp{int} sized block of memory where the
3776 number 1000 is initially stored.
3778 When a program references a symbol the compiler generates code that
3779 first accesses the symbol table to find the address of the symbol's
3780 memory block and then code to read the value from that memory block.
3787 looks up the symbol @samp{foo} in the symbol table, gets the address
3788 associated with this symbol and then writes the value 1 into that
3795 looks up the symbol @samp{foo} in the symbol table, gets its address
3796 and then copies this address into the block of memory associated with
3797 the variable @samp{a}.
3799 Linker scripts symbol declarations, by contrast, create an entry in
3800 the symbol table but do not assign any memory to them. Thus they are
3801 an address without a value. So for example the linker script definition:
3807 creates an entry in the symbol table called @samp{foo} which holds
3808 the address of memory location 1000, but nothing special is stored at
3809 address 1000. This means that you cannot access the @emph{value} of a
3810 linker script defined symbol - it has no value - all you can do is
3811 access the @emph{address} of a linker script defined symbol.
3813 Hence when you are using a linker script defined symbol in source code
3814 you should always take the address of the symbol, and never attempt to
3815 use its value. For example suppose you want to copy the contents of a
3816 section of memory called .ROM into a section called .FLASH and the
3817 linker script contains these declarations:
3821 start_of_ROM = .ROM;
3822 end_of_ROM = .ROM + sizeof (.ROM) - 1;
3823 start_of_FLASH = .FLASH;
3827 Then the C source code to perform the copy would be:
3831 extern char start_of_ROM, end_of_ROM, start_of_FLASH;
3833 memcpy (& start_of_FLASH, & start_of_ROM, & end_of_ROM - & start_of_ROM);
3837 Note the use of the @samp{&} operators. These are correct.
3840 @section SECTIONS Command
3842 The @code{SECTIONS} command tells the linker how to map input sections
3843 into output sections, and how to place the output sections in memory.
3845 The format of the @code{SECTIONS} command is:
3849 @var{sections-command}
3850 @var{sections-command}
3855 Each @var{sections-command} may of be one of the following:
3859 an @code{ENTRY} command (@pxref{Entry Point,,Entry command})
3861 a symbol assignment (@pxref{Assignments})
3863 an output section description
3865 an overlay description
3868 The @code{ENTRY} command and symbol assignments are permitted inside the
3869 @code{SECTIONS} command for convenience in using the location counter in
3870 those commands. This can also make the linker script easier to
3871 understand because you can use those commands at meaningful points in
3872 the layout of the output file.
3874 Output section descriptions and overlay descriptions are described
3877 If you do not use a @code{SECTIONS} command in your linker script, the
3878 linker will place each input section into an identically named output
3879 section in the order that the sections are first encountered in the
3880 input files. If all input sections are present in the first file, for
3881 example, the order of sections in the output file will match the order
3882 in the first input file. The first section will be at address zero.
3885 * Output Section Description:: Output section description
3886 * Output Section Name:: Output section name
3887 * Output Section Address:: Output section address
3888 * Input Section:: Input section description
3889 * Output Section Data:: Output section data
3890 * Output Section Keywords:: Output section keywords
3891 * Output Section Discarding:: Output section discarding
3892 * Output Section Attributes:: Output section attributes
3893 * Overlay Description:: Overlay description
3896 @node Output Section Description
3897 @subsection Output Section Description
3898 The full description of an output section looks like this:
3901 @var{section} [@var{address}] [(@var{type})] :
3903 [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT]
3904 [SUBALIGN(@var{subsection_align})]
3907 @var{output-section-command}
3908 @var{output-section-command}
3910 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}] [,]
3914 Most output sections do not use most of the optional section attributes.
3916 The whitespace around @var{section} is required, so that the section
3917 name is unambiguous. The colon and the curly braces are also required.
3918 The comma at the end may be required if a @var{fillexp} is used and
3919 the next @var{sections-command} looks like a continuation of the expression.
3920 The line breaks and other white space are optional.
3922 Each @var{output-section-command} may be one of the following:
3926 a symbol assignment (@pxref{Assignments})
3928 an input section description (@pxref{Input Section})
3930 data values to include directly (@pxref{Output Section Data})
3932 a special output section keyword (@pxref{Output Section Keywords})
3935 @node Output Section Name
3936 @subsection Output Section Name
3937 @cindex name, section
3938 @cindex section name
3939 The name of the output section is @var{section}. @var{section} must
3940 meet the constraints of your output format. In formats which only
3941 support a limited number of sections, such as @code{a.out}, the name
3942 must be one of the names supported by the format (@code{a.out}, for
3943 example, allows only @samp{.text}, @samp{.data} or @samp{.bss}). If the
3944 output format supports any number of sections, but with numbers and not
3945 names (as is the case for Oasys), the name should be supplied as a
3946 quoted numeric string. A section name may consist of any sequence of
3947 characters, but a name which contains any unusual characters such as
3948 commas must be quoted.
3950 The output section name @samp{/DISCARD/} is special; @ref{Output Section
3953 @node Output Section Address
3954 @subsection Output Section Address
3955 @cindex address, section
3956 @cindex section address
3957 The @var{address} is an expression for the VMA (the virtual memory
3958 address) of the output section. This address is optional, but if it
3959 is provided then the output address will be set exactly as specified.
3961 If the output address is not specified then one will be chosen for the
3962 section, based on the heuristic below. This address will be adjusted
3963 to fit the alignment requirement of the output section. The
3964 alignment requirement is the strictest alignment of any input section
3965 contained within the output section.
3967 The output section address heuristic is as follows:
3971 If an output memory @var{region} is set for the section then it
3972 is added to this region and its address will be the next free address
3976 If the MEMORY command has been used to create a list of memory
3977 regions then the first region which has attributes compatible with the
3978 section is selected to contain it. The section's output address will
3979 be the next free address in that region; @ref{MEMORY}.
3982 If no memory regions were specified, or none match the section then
3983 the output address will be based on the current value of the location
3991 .text . : @{ *(.text) @}
3998 .text : @{ *(.text) @}
4002 are subtly different. The first will set the address of the
4003 @samp{.text} output section to the current value of the location
4004 counter. The second will set it to the current value of the location
4005 counter aligned to the strictest alignment of any of the @samp{.text}
4008 The @var{address} may be an arbitrary expression; @ref{Expressions}.
4009 For example, if you want to align the section on a 0x10 byte boundary,
4010 so that the lowest four bits of the section address are zero, you could
4011 do something like this:
4013 .text ALIGN(0x10) : @{ *(.text) @}
4016 This works because @code{ALIGN} returns the current location counter
4017 aligned upward to the specified value.
4019 Specifying @var{address} for a section will change the value of the
4020 location counter, provided that the section is non-empty. (Empty
4021 sections are ignored).
4024 @subsection Input Section Description
4025 @cindex input sections
4026 @cindex mapping input sections to output sections
4027 The most common output section command is an input section description.
4029 The input section description is the most basic linker script operation.
4030 You use output sections to tell the linker how to lay out your program
4031 in memory. You use input section descriptions to tell the linker how to
4032 map the input files into your memory layout.
4035 * Input Section Basics:: Input section basics
4036 * Input Section Wildcards:: Input section wildcard patterns
4037 * Input Section Common:: Input section for common symbols
4038 * Input Section Keep:: Input section and garbage collection
4039 * Input Section Example:: Input section example
4042 @node Input Section Basics
4043 @subsubsection Input Section Basics
4044 @cindex input section basics
4045 An input section description consists of a file name optionally followed
4046 by a list of section names in parentheses.
4048 The file name and the section name may be wildcard patterns, which we
4049 describe further below (@pxref{Input Section Wildcards}).
4051 The most common input section description is to include all input
4052 sections with a particular name in the output section. For example, to
4053 include all input @samp{.text} sections, you would write:
4058 Here the @samp{*} is a wildcard which matches any file name. To exclude a list
4059 of files from matching the file name wildcard, EXCLUDE_FILE may be used to
4060 match all files except the ones specified in the EXCLUDE_FILE list. For
4063 *(EXCLUDE_FILE (*crtend.o *otherfile.o) .ctors)
4065 will cause all .ctors sections from all files except @file{crtend.o} and
4066 @file{otherfile.o} to be included.
4068 There are two ways to include more than one section:
4074 The difference between these is the order in which the @samp{.text} and
4075 @samp{.rdata} input sections will appear in the output section. In the
4076 first example, they will be intermingled, appearing in the same order as
4077 they are found in the linker input. In the second example, all
4078 @samp{.text} input sections will appear first, followed by all
4079 @samp{.rdata} input sections.
4081 You can specify a file name to include sections from a particular file.
4082 You would do this if one or more of your files contain special data that
4083 needs to be at a particular location in memory. For example:
4088 To refine the sections that are included based on the section flags
4089 of an input section, INPUT_SECTION_FLAGS may be used.
4091 Here is a simple example for using Section header flags for ELF sections:
4096 .text : @{ INPUT_SECTION_FLAGS (SHF_MERGE & SHF_STRINGS) *(.text) @}
4097 .text2 : @{ INPUT_SECTION_FLAGS (!SHF_WRITE) *(.text) @}
4102 In this example, the output section @samp{.text} will be comprised of any
4103 input section matching the name *(.text) whose section header flags
4104 @code{SHF_MERGE} and @code{SHF_STRINGS} are set. The output section
4105 @samp{.text2} will be comprised of any input section matching the name *(.text)
4106 whose section header flag @code{SHF_WRITE} is clear.
4108 You can also specify files within archives by writing a pattern
4109 matching the archive, a colon, then the pattern matching the file,
4110 with no whitespace around the colon.
4114 matches file within archive
4116 matches the whole archive
4118 matches file but not one in an archive
4121 Either one or both of @samp{archive} and @samp{file} can contain shell
4122 wildcards. On DOS based file systems, the linker will assume that a
4123 single letter followed by a colon is a drive specifier, so
4124 @samp{c:myfile.o} is a simple file specification, not @samp{myfile.o}
4125 within an archive called @samp{c}. @samp{archive:file} filespecs may
4126 also be used within an @code{EXCLUDE_FILE} list, but may not appear in
4127 other linker script contexts. For instance, you cannot extract a file
4128 from an archive by using @samp{archive:file} in an @code{INPUT}
4131 If you use a file name without a list of sections, then all sections in
4132 the input file will be included in the output section. This is not
4133 commonly done, but it may by useful on occasion. For example:
4138 When you use a file name which is not an @samp{archive:file} specifier
4139 and does not contain any wild card
4140 characters, the linker will first see if you also specified the file
4141 name on the linker command line or in an @code{INPUT} command. If you
4142 did not, the linker will attempt to open the file as an input file, as
4143 though it appeared on the command line. Note that this differs from an
4144 @code{INPUT} command, because the linker will not search for the file in
4145 the archive search path.
4147 @node Input Section Wildcards
4148 @subsubsection Input Section Wildcard Patterns
4149 @cindex input section wildcards
4150 @cindex wildcard file name patterns
4151 @cindex file name wildcard patterns
4152 @cindex section name wildcard patterns
4153 In an input section description, either the file name or the section
4154 name or both may be wildcard patterns.
4156 The file name of @samp{*} seen in many examples is a simple wildcard
4157 pattern for the file name.
4159 The wildcard patterns are like those used by the Unix shell.
4163 matches any number of characters
4165 matches any single character
4167 matches a single instance of any of the @var{chars}; the @samp{-}
4168 character may be used to specify a range of characters, as in
4169 @samp{[a-z]} to match any lower case letter
4171 quotes the following character
4174 When a file name is matched with a wildcard, the wildcard characters
4175 will not match a @samp{/} character (used to separate directory names on
4176 Unix). A pattern consisting of a single @samp{*} character is an
4177 exception; it will always match any file name, whether it contains a
4178 @samp{/} or not. In a section name, the wildcard characters will match
4179 a @samp{/} character.
4181 File name wildcard patterns only match files which are explicitly
4182 specified on the command line or in an @code{INPUT} command. The linker
4183 does not search directories to expand wildcards.
4185 If a file name matches more than one wildcard pattern, or if a file name
4186 appears explicitly and is also matched by a wildcard pattern, the linker
4187 will use the first match in the linker script. For example, this
4188 sequence of input section descriptions is probably in error, because the
4189 @file{data.o} rule will not be used:
4191 .data : @{ *(.data) @}
4192 .data1 : @{ data.o(.data) @}
4195 @cindex SORT_BY_NAME
4196 Normally, the linker will place files and sections matched by wildcards
4197 in the order in which they are seen during the link. You can change
4198 this by using the @code{SORT_BY_NAME} keyword, which appears before a wildcard
4199 pattern in parentheses (e.g., @code{SORT_BY_NAME(.text*)}). When the
4200 @code{SORT_BY_NAME} keyword is used, the linker will sort the files or sections
4201 into ascending order by name before placing them in the output file.
4203 @cindex SORT_BY_ALIGNMENT
4204 @code{SORT_BY_ALIGNMENT} is very similar to @code{SORT_BY_NAME}. The
4205 difference is @code{SORT_BY_ALIGNMENT} will sort sections into
4206 descending order by alignment before placing them in the output file.
4207 Larger alignments are placed before smaller alignments in order to
4208 reduce the amount of padding necessary.
4210 @cindex SORT_BY_INIT_PRIORITY
4211 @code{SORT_BY_INIT_PRIORITY} is very similar to @code{SORT_BY_NAME}. The
4212 difference is @code{SORT_BY_INIT_PRIORITY} will sort sections into
4213 ascending order by numerical value of the GCC init_priority attribute
4214 encoded in the section name before placing them in the output file.
4217 @code{SORT} is an alias for @code{SORT_BY_NAME}.
4219 When there are nested section sorting commands in linker script, there
4220 can be at most 1 level of nesting for section sorting commands.
4224 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
4225 It will sort the input sections by name first, then by alignment if two
4226 sections have the same name.
4228 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
4229 It will sort the input sections by alignment first, then by name if two
4230 sections have the same alignment.
4232 @code{SORT_BY_NAME} (@code{SORT_BY_NAME} (wildcard section pattern)) is
4233 treated the same as @code{SORT_BY_NAME} (wildcard section pattern).
4235 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern))
4236 is treated the same as @code{SORT_BY_ALIGNMENT} (wildcard section pattern).
4238 All other nested section sorting commands are invalid.
4241 When both command line section sorting option and linker script
4242 section sorting command are used, section sorting command always
4243 takes precedence over the command line option.
4245 If the section sorting command in linker script isn't nested, the
4246 command line option will make the section sorting command to be
4247 treated as nested sorting command.
4251 @code{SORT_BY_NAME} (wildcard section pattern ) with
4252 @option{--sort-sections alignment} is equivalent to
4253 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
4255 @code{SORT_BY_ALIGNMENT} (wildcard section pattern) with
4256 @option{--sort-section name} is equivalent to
4257 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
4260 If the section sorting command in linker script is nested, the
4261 command line option will be ignored.
4264 @code{SORT_NONE} disables section sorting by ignoring the command line
4265 section sorting option.
4267 If you ever get confused about where input sections are going, use the
4268 @samp{-M} linker option to generate a map file. The map file shows
4269 precisely how input sections are mapped to output sections.
4271 This example shows how wildcard patterns might be used to partition
4272 files. This linker script directs the linker to place all @samp{.text}
4273 sections in @samp{.text} and all @samp{.bss} sections in @samp{.bss}.
4274 The linker will place the @samp{.data} section from all files beginning
4275 with an upper case character in @samp{.DATA}; for all other files, the
4276 linker will place the @samp{.data} section in @samp{.data}.
4280 .text : @{ *(.text) @}
4281 .DATA : @{ [A-Z]*(.data) @}
4282 .data : @{ *(.data) @}
4283 .bss : @{ *(.bss) @}
4288 @node Input Section Common
4289 @subsubsection Input Section for Common Symbols
4290 @cindex common symbol placement
4291 @cindex uninitialized data placement
4292 A special notation is needed for common symbols, because in many object
4293 file formats common symbols do not have a particular input section. The
4294 linker treats common symbols as though they are in an input section
4295 named @samp{COMMON}.
4297 You may use file names with the @samp{COMMON} section just as with any
4298 other input sections. You can use this to place common symbols from a
4299 particular input file in one section while common symbols from other
4300 input files are placed in another section.
4302 In most cases, common symbols in input files will be placed in the
4303 @samp{.bss} section in the output file. For example:
4305 .bss @{ *(.bss) *(COMMON) @}
4308 @cindex scommon section
4309 @cindex small common symbols
4310 Some object file formats have more than one type of common symbol. For
4311 example, the MIPS ELF object file format distinguishes standard common
4312 symbols and small common symbols. In this case, the linker will use a
4313 different special section name for other types of common symbols. In
4314 the case of MIPS ELF, the linker uses @samp{COMMON} for standard common
4315 symbols and @samp{.scommon} for small common symbols. This permits you
4316 to map the different types of common symbols into memory at different
4320 You will sometimes see @samp{[COMMON]} in old linker scripts. This
4321 notation is now considered obsolete. It is equivalent to
4324 @node Input Section Keep
4325 @subsubsection Input Section and Garbage Collection
4327 @cindex garbage collection
4328 When link-time garbage collection is in use (@samp{--gc-sections}),
4329 it is often useful to mark sections that should not be eliminated.
4330 This is accomplished by surrounding an input section's wildcard entry
4331 with @code{KEEP()}, as in @code{KEEP(*(.init))} or
4332 @code{KEEP(SORT_BY_NAME(*)(.ctors))}.
4334 @node Input Section Example
4335 @subsubsection Input Section Example
4336 The following example is a complete linker script. It tells the linker
4337 to read all of the sections from file @file{all.o} and place them at the
4338 start of output section @samp{outputa} which starts at location
4339 @samp{0x10000}. All of section @samp{.input1} from file @file{foo.o}
4340 follows immediately, in the same output section. All of section
4341 @samp{.input2} from @file{foo.o} goes into output section
4342 @samp{outputb}, followed by section @samp{.input1} from @file{foo1.o}.
4343 All of the remaining @samp{.input1} and @samp{.input2} sections from any
4344 files are written to output section @samp{outputc}.
4372 @node Output Section Data
4373 @subsection Output Section Data
4375 @cindex section data
4376 @cindex output section data
4377 @kindex BYTE(@var{expression})
4378 @kindex SHORT(@var{expression})
4379 @kindex LONG(@var{expression})
4380 @kindex QUAD(@var{expression})
4381 @kindex SQUAD(@var{expression})
4382 You can include explicit bytes of data in an output section by using
4383 @code{BYTE}, @code{SHORT}, @code{LONG}, @code{QUAD}, or @code{SQUAD} as
4384 an output section command. Each keyword is followed by an expression in
4385 parentheses providing the value to store (@pxref{Expressions}). The
4386 value of the expression is stored at the current value of the location
4389 The @code{BYTE}, @code{SHORT}, @code{LONG}, and @code{QUAD} commands
4390 store one, two, four, and eight bytes (respectively). After storing the
4391 bytes, the location counter is incremented by the number of bytes
4394 For example, this will store the byte 1 followed by the four byte value
4395 of the symbol @samp{addr}:
4401 When using a 64 bit host or target, @code{QUAD} and @code{SQUAD} are the
4402 same; they both store an 8 byte, or 64 bit, value. When both host and
4403 target are 32 bits, an expression is computed as 32 bits. In this case
4404 @code{QUAD} stores a 32 bit value zero extended to 64 bits, and
4405 @code{SQUAD} stores a 32 bit value sign extended to 64 bits.
4407 If the object file format of the output file has an explicit endianness,
4408 which is the normal case, the value will be stored in that endianness.
4409 When the object file format does not have an explicit endianness, as is
4410 true of, for example, S-records, the value will be stored in the
4411 endianness of the first input object file.
4413 Note---these commands only work inside a section description and not
4414 between them, so the following will produce an error from the linker:
4416 SECTIONS @{@ .text : @{@ *(.text) @}@ LONG(1) .data : @{@ *(.data) @}@ @}@
4418 whereas this will work:
4420 SECTIONS @{@ .text : @{@ *(.text) ; LONG(1) @}@ .data : @{@ *(.data) @}@ @}@
4423 @kindex FILL(@var{expression})
4424 @cindex holes, filling
4425 @cindex unspecified memory
4426 You may use the @code{FILL} command to set the fill pattern for the
4427 current section. It is followed by an expression in parentheses. Any
4428 otherwise unspecified regions of memory within the section (for example,
4429 gaps left due to the required alignment of input sections) are filled
4430 with the value of the expression, repeated as
4431 necessary. A @code{FILL} statement covers memory locations after the
4432 point at which it occurs in the section definition; by including more
4433 than one @code{FILL} statement, you can have different fill patterns in
4434 different parts of an output section.
4436 This example shows how to fill unspecified regions of memory with the
4442 The @code{FILL} command is similar to the @samp{=@var{fillexp}} output
4443 section attribute, but it only affects the
4444 part of the section following the @code{FILL} command, rather than the
4445 entire section. If both are used, the @code{FILL} command takes
4446 precedence. @xref{Output Section Fill}, for details on the fill
4449 @node Output Section Keywords
4450 @subsection Output Section Keywords
4451 There are a couple of keywords which can appear as output section
4455 @kindex CREATE_OBJECT_SYMBOLS
4456 @cindex input filename symbols
4457 @cindex filename symbols
4458 @item CREATE_OBJECT_SYMBOLS
4459 The command tells the linker to create a symbol for each input file.
4460 The name of each symbol will be the name of the corresponding input
4461 file. The section of each symbol will be the output section in which
4462 the @code{CREATE_OBJECT_SYMBOLS} command appears.
4464 This is conventional for the a.out object file format. It is not
4465 normally used for any other object file format.
4467 @kindex CONSTRUCTORS
4468 @cindex C++ constructors, arranging in link
4469 @cindex constructors, arranging in link
4471 When linking using the a.out object file format, the linker uses an
4472 unusual set construct to support C++ global constructors and
4473 destructors. When linking object file formats which do not support
4474 arbitrary sections, such as ECOFF and XCOFF, the linker will
4475 automatically recognize C++ global constructors and destructors by name.
4476 For these object file formats, the @code{CONSTRUCTORS} command tells the
4477 linker to place constructor information in the output section where the
4478 @code{CONSTRUCTORS} command appears. The @code{CONSTRUCTORS} command is
4479 ignored for other object file formats.
4481 The symbol @w{@code{__CTOR_LIST__}} marks the start of the global
4482 constructors, and the symbol @w{@code{__CTOR_END__}} marks the end.
4483 Similarly, @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_END__}} mark
4484 the start and end of the global destructors. The
4485 first word in the list is the number of entries, followed by the address
4486 of each constructor or destructor, followed by a zero word. The
4487 compiler must arrange to actually run the code. For these object file
4488 formats @sc{gnu} C++ normally calls constructors from a subroutine
4489 @code{__main}; a call to @code{__main} is automatically inserted into
4490 the startup code for @code{main}. @sc{gnu} C++ normally runs
4491 destructors either by using @code{atexit}, or directly from the function
4494 For object file formats such as @code{COFF} or @code{ELF} which support
4495 arbitrary section names, @sc{gnu} C++ will normally arrange to put the
4496 addresses of global constructors and destructors into the @code{.ctors}
4497 and @code{.dtors} sections. Placing the following sequence into your
4498 linker script will build the sort of table which the @sc{gnu} C++
4499 runtime code expects to see.
4503 LONG((__CTOR_END__ - __CTOR_LIST__) / 4 - 2)
4508 LONG((__DTOR_END__ - __DTOR_LIST__) / 4 - 2)
4514 If you are using the @sc{gnu} C++ support for initialization priority,
4515 which provides some control over the order in which global constructors
4516 are run, you must sort the constructors at link time to ensure that they
4517 are executed in the correct order. When using the @code{CONSTRUCTORS}
4518 command, use @samp{SORT_BY_NAME(CONSTRUCTORS)} instead. When using the
4519 @code{.ctors} and @code{.dtors} sections, use @samp{*(SORT_BY_NAME(.ctors))} and
4520 @samp{*(SORT_BY_NAME(.dtors))} instead of just @samp{*(.ctors)} and
4523 Normally the compiler and linker will handle these issues automatically,
4524 and you will not need to concern yourself with them. However, you may
4525 need to consider this if you are using C++ and writing your own linker
4530 @node Output Section Discarding
4531 @subsection Output Section Discarding
4532 @cindex discarding sections
4533 @cindex sections, discarding
4534 @cindex removing sections
4535 The linker will not normally create output sections with no contents.
4536 This is for convenience when referring to input sections that may or
4537 may not be present in any of the input files. For example:
4539 .foo : @{ *(.foo) @}
4542 will only create a @samp{.foo} section in the output file if there is a
4543 @samp{.foo} section in at least one input file, and if the input
4544 sections are not all empty. Other link script directives that allocate
4545 space in an output section will also create the output section. So
4546 too will assignments to dot even if the assignment does not create
4547 space, except for @samp{. = 0}, @samp{. = . + 0}, @samp{. = sym},
4548 @samp{. = . + sym} and @samp{. = ALIGN (. != 0, expr, 1)} when
4549 @samp{sym} is an absolute symbol of value 0 defined in the script.
4550 This allows you to force output of an empty section with @samp{. = .}.
4552 The linker will ignore address assignments (@pxref{Output Section Address})
4553 on discarded output sections, except when the linker script defines
4554 symbols in the output section. In that case the linker will obey
4555 the address assignments, possibly advancing dot even though the
4556 section is discarded.
4559 The special output section name @samp{/DISCARD/} may be used to discard
4560 input sections. Any input sections which are assigned to an output
4561 section named @samp{/DISCARD/} are not included in the output file.
4563 @node Output Section Attributes
4564 @subsection Output Section Attributes
4565 @cindex output section attributes
4566 We showed above that the full description of an output section looked
4571 @var{section} [@var{address}] [(@var{type})] :
4573 [ALIGN(@var{section_align})]
4574 [SUBALIGN(@var{subsection_align})]
4577 @var{output-section-command}
4578 @var{output-section-command}
4580 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}]
4584 We've already described @var{section}, @var{address}, and
4585 @var{output-section-command}. In this section we will describe the
4586 remaining section attributes.
4589 * Output Section Type:: Output section type
4590 * Output Section LMA:: Output section LMA
4591 * Forced Output Alignment:: Forced Output Alignment
4592 * Forced Input Alignment:: Forced Input Alignment
4593 * Output Section Constraint:: Output section constraint
4594 * Output Section Region:: Output section region
4595 * Output Section Phdr:: Output section phdr
4596 * Output Section Fill:: Output section fill
4599 @node Output Section Type
4600 @subsubsection Output Section Type
4601 Each output section may have a type. The type is a keyword in
4602 parentheses. The following types are defined:
4606 The section should be marked as not loadable, so that it will not be
4607 loaded into memory when the program is run.
4612 These type names are supported for backward compatibility, and are
4613 rarely used. They all have the same effect: the section should be
4614 marked as not allocatable, so that no memory is allocated for the
4615 section when the program is run.
4619 @cindex prevent unnecessary loading
4620 @cindex loading, preventing
4621 The linker normally sets the attributes of an output section based on
4622 the input sections which map into it. You can override this by using
4623 the section type. For example, in the script sample below, the
4624 @samp{ROM} section is addressed at memory location @samp{0} and does not
4625 need to be loaded when the program is run.
4629 ROM 0 (NOLOAD) : @{ @dots{} @}
4635 @node Output Section LMA
4636 @subsubsection Output Section LMA
4637 @kindex AT>@var{lma_region}
4638 @kindex AT(@var{lma})
4639 @cindex load address
4640 @cindex section load address
4641 Every section has a virtual address (VMA) and a load address (LMA); see
4642 @ref{Basic Script Concepts}. The virtual address is specified by the
4643 @pxref{Output Section Address} described earlier. The load address is
4644 specified by the @code{AT} or @code{AT>} keywords. Specifying a load
4645 address is optional.
4647 The @code{AT} keyword takes an expression as an argument. This
4648 specifies the exact load address of the section. The @code{AT>} keyword
4649 takes the name of a memory region as an argument. @xref{MEMORY}. The
4650 load address of the section is set to the next free address in the
4651 region, aligned to the section's alignment requirements.
4653 If neither @code{AT} nor @code{AT>} is specified for an allocatable
4654 section, the linker will use the following heuristic to determine the
4659 If the section has a specific VMA address, then this is used as
4660 the LMA address as well.
4663 If the section is not allocatable then its LMA is set to its VMA.
4666 Otherwise if a memory region can be found that is compatible
4667 with the current section, and this region contains at least one
4668 section, then the LMA is set so the difference between the
4669 VMA and LMA is the same as the difference between the VMA and LMA of
4670 the last section in the located region.
4673 If no memory regions have been declared then a default region
4674 that covers the entire address space is used in the previous step.
4677 If no suitable region could be found, or there was no previous
4678 section then the LMA is set equal to the VMA.
4681 @cindex ROM initialized data
4682 @cindex initialized data in ROM
4683 This feature is designed to make it easy to build a ROM image. For
4684 example, the following linker script creates three output sections: one
4685 called @samp{.text}, which starts at @code{0x1000}, one called
4686 @samp{.mdata}, which is loaded at the end of the @samp{.text} section
4687 even though its VMA is @code{0x2000}, and one called @samp{.bss} to hold
4688 uninitialized data at address @code{0x3000}. The symbol @code{_data} is
4689 defined with the value @code{0x2000}, which shows that the location
4690 counter holds the VMA value, not the LMA value.
4696 .text 0x1000 : @{ *(.text) _etext = . ; @}
4698 AT ( ADDR (.text) + SIZEOF (.text) )
4699 @{ _data = . ; *(.data); _edata = . ; @}
4701 @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@}
4706 The run-time initialization code for use with a program generated with
4707 this linker script would include something like the following, to copy
4708 the initialized data from the ROM image to its runtime address. Notice
4709 how this code takes advantage of the symbols defined by the linker
4714 extern char _etext, _data, _edata, _bstart, _bend;
4715 char *src = &_etext;
4718 /* ROM has data at end of text; copy it. */
4719 while (dst < &_edata)
4723 for (dst = &_bstart; dst< &_bend; dst++)
4728 @node Forced Output Alignment
4729 @subsubsection Forced Output Alignment
4730 @kindex ALIGN(@var{section_align})
4731 @cindex forcing output section alignment
4732 @cindex output section alignment
4733 You can increase an output section's alignment by using ALIGN. As an
4734 alternative you can enforce that the difference between the VMA and LMA remains
4735 intact throughout this output section with the ALIGN_WITH_INPUT attribute.
4737 @node Forced Input Alignment
4738 @subsubsection Forced Input Alignment
4739 @kindex SUBALIGN(@var{subsection_align})
4740 @cindex forcing input section alignment
4741 @cindex input section alignment
4742 You can force input section alignment within an output section by using
4743 SUBALIGN. The value specified overrides any alignment given by input
4744 sections, whether larger or smaller.
4746 @node Output Section Constraint
4747 @subsubsection Output Section Constraint
4750 @cindex constraints on output sections
4751 You can specify that an output section should only be created if all
4752 of its input sections are read-only or all of its input sections are
4753 read-write by using the keyword @code{ONLY_IF_RO} and
4754 @code{ONLY_IF_RW} respectively.
4756 @node Output Section Region
4757 @subsubsection Output Section Region
4758 @kindex >@var{region}
4759 @cindex section, assigning to memory region
4760 @cindex memory regions and sections
4761 You can assign a section to a previously defined region of memory by
4762 using @samp{>@var{region}}. @xref{MEMORY}.
4764 Here is a simple example:
4767 MEMORY @{ rom : ORIGIN = 0x1000, LENGTH = 0x1000 @}
4768 SECTIONS @{ ROM : @{ *(.text) @} >rom @}
4772 @node Output Section Phdr
4773 @subsubsection Output Section Phdr
4775 @cindex section, assigning to program header
4776 @cindex program headers and sections
4777 You can assign a section to a previously defined program segment by
4778 using @samp{:@var{phdr}}. @xref{PHDRS}. If a section is assigned to
4779 one or more segments, then all subsequent allocated sections will be
4780 assigned to those segments as well, unless they use an explicitly
4781 @code{:@var{phdr}} modifier. You can use @code{:NONE} to tell the
4782 linker to not put the section in any segment at all.
4784 Here is a simple example:
4787 PHDRS @{ text PT_LOAD ; @}
4788 SECTIONS @{ .text : @{ *(.text) @} :text @}
4792 @node Output Section Fill
4793 @subsubsection Output Section Fill
4794 @kindex =@var{fillexp}
4795 @cindex section fill pattern
4796 @cindex fill pattern, entire section
4797 You can set the fill pattern for an entire section by using
4798 @samp{=@var{fillexp}}. @var{fillexp} is an expression
4799 (@pxref{Expressions}). Any otherwise unspecified regions of memory
4800 within the output section (for example, gaps left due to the required
4801 alignment of input sections) will be filled with the value, repeated as
4802 necessary. If the fill expression is a simple hex number, ie. a string
4803 of hex digit starting with @samp{0x} and without a trailing @samp{k} or @samp{M}, then
4804 an arbitrarily long sequence of hex digits can be used to specify the
4805 fill pattern; Leading zeros become part of the pattern too. For all
4806 other cases, including extra parentheses or a unary @code{+}, the fill
4807 pattern is the four least significant bytes of the value of the
4808 expression. In all cases, the number is big-endian.
4810 You can also change the fill value with a @code{FILL} command in the
4811 output section commands; (@pxref{Output Section Data}).
4813 Here is a simple example:
4816 SECTIONS @{ .text : @{ *(.text) @} =0x90909090 @}
4820 @node Overlay Description
4821 @subsection Overlay Description
4824 An overlay description provides an easy way to describe sections which
4825 are to be loaded as part of a single memory image but are to be run at
4826 the same memory address. At run time, some sort of overlay manager will
4827 copy the overlaid sections in and out of the runtime memory address as
4828 required, perhaps by simply manipulating addressing bits. This approach
4829 can be useful, for example, when a certain region of memory is faster
4832 Overlays are described using the @code{OVERLAY} command. The
4833 @code{OVERLAY} command is used within a @code{SECTIONS} command, like an
4834 output section description. The full syntax of the @code{OVERLAY}
4835 command is as follows:
4838 OVERLAY [@var{start}] : [NOCROSSREFS] [AT ( @var{ldaddr} )]
4842 @var{output-section-command}
4843 @var{output-section-command}
4845 @} [:@var{phdr}@dots{}] [=@var{fill}]
4848 @var{output-section-command}
4849 @var{output-section-command}
4851 @} [:@var{phdr}@dots{}] [=@var{fill}]
4853 @} [>@var{region}] [:@var{phdr}@dots{}] [=@var{fill}] [,]
4857 Everything is optional except @code{OVERLAY} (a keyword), and each
4858 section must have a name (@var{secname1} and @var{secname2} above). The
4859 section definitions within the @code{OVERLAY} construct are identical to
4860 those within the general @code{SECTIONS} construct (@pxref{SECTIONS}),
4861 except that no addresses and no memory regions may be defined for
4862 sections within an @code{OVERLAY}.
4864 The comma at the end may be required if a @var{fill} is used and
4865 the next @var{sections-command} looks like a continuation of the expression.
4867 The sections are all defined with the same starting address. The load
4868 addresses of the sections are arranged such that they are consecutive in
4869 memory starting at the load address used for the @code{OVERLAY} as a
4870 whole (as with normal section definitions, the load address is optional,
4871 and defaults to the start address; the start address is also optional,
4872 and defaults to the current value of the location counter).
4874 If the @code{NOCROSSREFS} keyword is used, and there are any
4875 references among the sections, the linker will report an error. Since
4876 the sections all run at the same address, it normally does not make
4877 sense for one section to refer directly to another.
4878 @xref{Miscellaneous Commands, NOCROSSREFS}.
4880 For each section within the @code{OVERLAY}, the linker automatically
4881 provides two symbols. The symbol @code{__load_start_@var{secname}} is
4882 defined as the starting load address of the section. The symbol
4883 @code{__load_stop_@var{secname}} is defined as the final load address of
4884 the section. Any characters within @var{secname} which are not legal
4885 within C identifiers are removed. C (or assembler) code may use these
4886 symbols to move the overlaid sections around as necessary.
4888 At the end of the overlay, the value of the location counter is set to
4889 the start address of the overlay plus the size of the largest section.
4891 Here is an example. Remember that this would appear inside a
4892 @code{SECTIONS} construct.
4895 OVERLAY 0x1000 : AT (0x4000)
4897 .text0 @{ o1/*.o(.text) @}
4898 .text1 @{ o2/*.o(.text) @}
4903 This will define both @samp{.text0} and @samp{.text1} to start at
4904 address 0x1000. @samp{.text0} will be loaded at address 0x4000, and
4905 @samp{.text1} will be loaded immediately after @samp{.text0}. The
4906 following symbols will be defined if referenced: @code{__load_start_text0},
4907 @code{__load_stop_text0}, @code{__load_start_text1},
4908 @code{__load_stop_text1}.
4910 C code to copy overlay @code{.text1} into the overlay area might look
4915 extern char __load_start_text1, __load_stop_text1;
4916 memcpy ((char *) 0x1000, &__load_start_text1,
4917 &__load_stop_text1 - &__load_start_text1);
4921 Note that the @code{OVERLAY} command is just syntactic sugar, since
4922 everything it does can be done using the more basic commands. The above
4923 example could have been written identically as follows.
4927 .text0 0x1000 : AT (0x4000) @{ o1/*.o(.text) @}
4928 PROVIDE (__load_start_text0 = LOADADDR (.text0));
4929 PROVIDE (__load_stop_text0 = LOADADDR (.text0) + SIZEOF (.text0));
4930 .text1 0x1000 : AT (0x4000 + SIZEOF (.text0)) @{ o2/*.o(.text) @}
4931 PROVIDE (__load_start_text1 = LOADADDR (.text1));
4932 PROVIDE (__load_stop_text1 = LOADADDR (.text1) + SIZEOF (.text1));
4933 . = 0x1000 + MAX (SIZEOF (.text0), SIZEOF (.text1));
4938 @section MEMORY Command
4940 @cindex memory regions
4941 @cindex regions of memory
4942 @cindex allocating memory
4943 @cindex discontinuous memory
4944 The linker's default configuration permits allocation of all available
4945 memory. You can override this by using the @code{MEMORY} command.
4947 The @code{MEMORY} command describes the location and size of blocks of
4948 memory in the target. You can use it to describe which memory regions
4949 may be used by the linker, and which memory regions it must avoid. You
4950 can then assign sections to particular memory regions. The linker will
4951 set section addresses based on the memory regions, and will warn about
4952 regions that become too full. The linker will not shuffle sections
4953 around to fit into the available regions.
4955 A linker script may contain at most one use of the @code{MEMORY}
4956 command. However, you can define as many blocks of memory within it as
4957 you wish. The syntax is:
4962 @var{name} [(@var{attr})] : ORIGIN = @var{origin}, LENGTH = @var{len}
4968 The @var{name} is a name used in the linker script to refer to the
4969 region. The region name has no meaning outside of the linker script.
4970 Region names are stored in a separate name space, and will not conflict
4971 with symbol names, file names, or section names. Each memory region
4972 must have a distinct name within the @code{MEMORY} command. However you can
4973 add later alias names to existing memory regions with the @ref{REGION_ALIAS}
4976 @cindex memory region attributes
4977 The @var{attr} string is an optional list of attributes that specify
4978 whether to use a particular memory region for an input section which is
4979 not explicitly mapped in the linker script. As described in
4980 @ref{SECTIONS}, if you do not specify an output section for some input
4981 section, the linker will create an output section with the same name as
4982 the input section. If you define region attributes, the linker will use
4983 them to select the memory region for the output section that it creates.
4985 The @var{attr} string must consist only of the following characters:
5000 Invert the sense of any of the attributes that follow
5003 If a unmapped section matches any of the listed attributes other than
5004 @samp{!}, it will be placed in the memory region. The @samp{!}
5005 attribute reverses this test, so that an unmapped section will be placed
5006 in the memory region only if it does not match any of the listed
5012 The @var{origin} is an numerical expression for the start address of
5013 the memory region. The expression must evaluate to a constant and it
5014 cannot involve any symbols. The keyword @code{ORIGIN} may be
5015 abbreviated to @code{org} or @code{o} (but not, for example,
5021 The @var{len} is an expression for the size in bytes of the memory
5022 region. As with the @var{origin} expression, the expression must
5023 be numerical only and must evaluate to a constant. The keyword
5024 @code{LENGTH} may be abbreviated to @code{len} or @code{l}.
5026 In the following example, we specify that there are two memory regions
5027 available for allocation: one starting at @samp{0} for 256 kilobytes,
5028 and the other starting at @samp{0x40000000} for four megabytes. The
5029 linker will place into the @samp{rom} memory region every section which
5030 is not explicitly mapped into a memory region, and is either read-only
5031 or executable. The linker will place other sections which are not
5032 explicitly mapped into a memory region into the @samp{ram} memory
5039 rom (rx) : ORIGIN = 0, LENGTH = 256K
5040 ram (!rx) : org = 0x40000000, l = 4M
5045 Once you define a memory region, you can direct the linker to place
5046 specific output sections into that memory region by using the
5047 @samp{>@var{region}} output section attribute. For example, if you have
5048 a memory region named @samp{mem}, you would use @samp{>mem} in the
5049 output section definition. @xref{Output Section Region}. If no address
5050 was specified for the output section, the linker will set the address to
5051 the next available address within the memory region. If the combined
5052 output sections directed to a memory region are too large for the
5053 region, the linker will issue an error message.
5055 It is possible to access the origin and length of a memory in an
5056 expression via the @code{ORIGIN(@var{memory})} and
5057 @code{LENGTH(@var{memory})} functions:
5061 _fstack = ORIGIN(ram) + LENGTH(ram) - 4;
5066 @section PHDRS Command
5068 @cindex program headers
5069 @cindex ELF program headers
5070 @cindex program segments
5071 @cindex segments, ELF
5072 The ELF object file format uses @dfn{program headers}, also knows as
5073 @dfn{segments}. The program headers describe how the program should be
5074 loaded into memory. You can print them out by using the @code{objdump}
5075 program with the @samp{-p} option.
5077 When you run an ELF program on a native ELF system, the system loader
5078 reads the program headers in order to figure out how to load the
5079 program. This will only work if the program headers are set correctly.
5080 This manual does not describe the details of how the system loader
5081 interprets program headers; for more information, see the ELF ABI.
5083 The linker will create reasonable program headers by default. However,
5084 in some cases, you may need to specify the program headers more
5085 precisely. You may use the @code{PHDRS} command for this purpose. When
5086 the linker sees the @code{PHDRS} command in the linker script, it will
5087 not create any program headers other than the ones specified.
5089 The linker only pays attention to the @code{PHDRS} command when
5090 generating an ELF output file. In other cases, the linker will simply
5091 ignore @code{PHDRS}.
5093 This is the syntax of the @code{PHDRS} command. The words @code{PHDRS},
5094 @code{FILEHDR}, @code{AT}, and @code{FLAGS} are keywords.
5100 @var{name} @var{type} [ FILEHDR ] [ PHDRS ] [ AT ( @var{address} ) ]
5101 [ FLAGS ( @var{flags} ) ] ;
5106 The @var{name} is used only for reference in the @code{SECTIONS} command
5107 of the linker script. It is not put into the output file. Program
5108 header names are stored in a separate name space, and will not conflict
5109 with symbol names, file names, or section names. Each program header
5110 must have a distinct name. The headers are processed in order and it
5111 is usual for them to map to sections in ascending load address order.
5113 Certain program header types describe segments of memory which the
5114 system loader will load from the file. In the linker script, you
5115 specify the contents of these segments by placing allocatable output
5116 sections in the segments. You use the @samp{:@var{phdr}} output section
5117 attribute to place a section in a particular segment. @xref{Output
5120 It is normal to put certain sections in more than one segment. This
5121 merely implies that one segment of memory contains another. You may
5122 repeat @samp{:@var{phdr}}, using it once for each segment which should
5123 contain the section.
5125 If you place a section in one or more segments using @samp{:@var{phdr}},
5126 then the linker will place all subsequent allocatable sections which do
5127 not specify @samp{:@var{phdr}} in the same segments. This is for
5128 convenience, since generally a whole set of contiguous sections will be
5129 placed in a single segment. You can use @code{:NONE} to override the
5130 default segment and tell the linker to not put the section in any
5135 You may use the @code{FILEHDR} and @code{PHDRS} keywords after
5136 the program header type to further describe the contents of the segment.
5137 The @code{FILEHDR} keyword means that the segment should include the ELF
5138 file header. The @code{PHDRS} keyword means that the segment should
5139 include the ELF program headers themselves. If applied to a loadable
5140 segment (@code{PT_LOAD}), all prior loadable segments must have one of
5143 The @var{type} may be one of the following. The numbers indicate the
5144 value of the keyword.
5147 @item @code{PT_NULL} (0)
5148 Indicates an unused program header.
5150 @item @code{PT_LOAD} (1)
5151 Indicates that this program header describes a segment to be loaded from
5154 @item @code{PT_DYNAMIC} (2)
5155 Indicates a segment where dynamic linking information can be found.
5157 @item @code{PT_INTERP} (3)
5158 Indicates a segment where the name of the program interpreter may be
5161 @item @code{PT_NOTE} (4)
5162 Indicates a segment holding note information.
5164 @item @code{PT_SHLIB} (5)
5165 A reserved program header type, defined but not specified by the ELF
5168 @item @code{PT_PHDR} (6)
5169 Indicates a segment where the program headers may be found.
5171 @item @var{expression}
5172 An expression giving the numeric type of the program header. This may
5173 be used for types not defined above.
5176 You can specify that a segment should be loaded at a particular address
5177 in memory by using an @code{AT} expression. This is identical to the
5178 @code{AT} command used as an output section attribute (@pxref{Output
5179 Section LMA}). The @code{AT} command for a program header overrides the
5180 output section attribute.
5182 The linker will normally set the segment flags based on the sections
5183 which comprise the segment. You may use the @code{FLAGS} keyword to
5184 explicitly specify the segment flags. The value of @var{flags} must be
5185 an integer. It is used to set the @code{p_flags} field of the program
5188 Here is an example of @code{PHDRS}. This shows a typical set of program
5189 headers used on a native ELF system.
5195 headers PT_PHDR PHDRS ;
5197 text PT_LOAD FILEHDR PHDRS ;
5199 dynamic PT_DYNAMIC ;
5205 .interp : @{ *(.interp) @} :text :interp
5206 .text : @{ *(.text) @} :text
5207 .rodata : @{ *(.rodata) @} /* defaults to :text */
5209 . = . + 0x1000; /* move to a new page in memory */
5210 .data : @{ *(.data) @} :data
5211 .dynamic : @{ *(.dynamic) @} :data :dynamic
5218 @section VERSION Command
5219 @kindex VERSION @{script text@}
5220 @cindex symbol versions
5221 @cindex version script
5222 @cindex versions of symbols
5223 The linker supports symbol versions when using ELF. Symbol versions are
5224 only useful when using shared libraries. The dynamic linker can use
5225 symbol versions to select a specific version of a function when it runs
5226 a program that may have been linked against an earlier version of the
5229 You can include a version script directly in the main linker script, or
5230 you can supply the version script as an implicit linker script. You can
5231 also use the @samp{--version-script} linker option.
5233 The syntax of the @code{VERSION} command is simply
5235 VERSION @{ version-script-commands @}
5238 The format of the version script commands is identical to that used by
5239 Sun's linker in Solaris 2.5. The version script defines a tree of
5240 version nodes. You specify the node names and interdependencies in the
5241 version script. You can specify which symbols are bound to which
5242 version nodes, and you can reduce a specified set of symbols to local
5243 scope so that they are not globally visible outside of the shared
5246 The easiest way to demonstrate the version script language is with a few
5272 This example version script defines three version nodes. The first
5273 version node defined is @samp{VERS_1.1}; it has no other dependencies.
5274 The script binds the symbol @samp{foo1} to @samp{VERS_1.1}. It reduces
5275 a number of symbols to local scope so that they are not visible outside
5276 of the shared library; this is done using wildcard patterns, so that any
5277 symbol whose name begins with @samp{old}, @samp{original}, or @samp{new}
5278 is matched. The wildcard patterns available are the same as those used
5279 in the shell when matching filenames (also known as ``globbing'').
5280 However, if you specify the symbol name inside double quotes, then the
5281 name is treated as literal, rather than as a glob pattern.
5283 Next, the version script defines node @samp{VERS_1.2}. This node
5284 depends upon @samp{VERS_1.1}. The script binds the symbol @samp{foo2}
5285 to the version node @samp{VERS_1.2}.
5287 Finally, the version script defines node @samp{VERS_2.0}. This node
5288 depends upon @samp{VERS_1.2}. The scripts binds the symbols @samp{bar1}
5289 and @samp{bar2} are bound to the version node @samp{VERS_2.0}.
5291 When the linker finds a symbol defined in a library which is not
5292 specifically bound to a version node, it will effectively bind it to an
5293 unspecified base version of the library. You can bind all otherwise
5294 unspecified symbols to a given version node by using @samp{global: *;}
5295 somewhere in the version script. Note that it's slightly crazy to use
5296 wildcards in a global spec except on the last version node. Global
5297 wildcards elsewhere run the risk of accidentally adding symbols to the
5298 set exported for an old version. That's wrong since older versions
5299 ought to have a fixed set of symbols.
5301 The names of the version nodes have no specific meaning other than what
5302 they might suggest to the person reading them. The @samp{2.0} version
5303 could just as well have appeared in between @samp{1.1} and @samp{1.2}.
5304 However, this would be a confusing way to write a version script.
5306 Node name can be omitted, provided it is the only version node
5307 in the version script. Such version script doesn't assign any versions to
5308 symbols, only selects which symbols will be globally visible out and which
5312 @{ global: foo; bar; local: *; @};
5315 When you link an application against a shared library that has versioned
5316 symbols, the application itself knows which version of each symbol it
5317 requires, and it also knows which version nodes it needs from each
5318 shared library it is linked against. Thus at runtime, the dynamic
5319 loader can make a quick check to make sure that the libraries you have
5320 linked against do in fact supply all of the version nodes that the
5321 application will need to resolve all of the dynamic symbols. In this
5322 way it is possible for the dynamic linker to know with certainty that
5323 all external symbols that it needs will be resolvable without having to
5324 search for each symbol reference.
5326 The symbol versioning is in effect a much more sophisticated way of
5327 doing minor version checking that SunOS does. The fundamental problem
5328 that is being addressed here is that typically references to external
5329 functions are bound on an as-needed basis, and are not all bound when
5330 the application starts up. If a shared library is out of date, a
5331 required interface may be missing; when the application tries to use
5332 that interface, it may suddenly and unexpectedly fail. With symbol
5333 versioning, the user will get a warning when they start their program if
5334 the libraries being used with the application are too old.
5336 There are several GNU extensions to Sun's versioning approach. The
5337 first of these is the ability to bind a symbol to a version node in the
5338 source file where the symbol is defined instead of in the versioning
5339 script. This was done mainly to reduce the burden on the library
5340 maintainer. You can do this by putting something like:
5342 __asm__(".symver original_foo,foo@@VERS_1.1");
5345 in the C source file. This renames the function @samp{original_foo} to
5346 be an alias for @samp{foo} bound to the version node @samp{VERS_1.1}.
5347 The @samp{local:} directive can be used to prevent the symbol
5348 @samp{original_foo} from being exported. A @samp{.symver} directive
5349 takes precedence over a version script.
5351 The second GNU extension is to allow multiple versions of the same
5352 function to appear in a given shared library. In this way you can make
5353 an incompatible change to an interface without increasing the major
5354 version number of the shared library, while still allowing applications
5355 linked against the old interface to continue to function.
5357 To do this, you must use multiple @samp{.symver} directives in the
5358 source file. Here is an example:
5361 __asm__(".symver original_foo,foo@@");
5362 __asm__(".symver old_foo,foo@@VERS_1.1");
5363 __asm__(".symver old_foo1,foo@@VERS_1.2");
5364 __asm__(".symver new_foo,foo@@@@VERS_2.0");
5367 In this example, @samp{foo@@} represents the symbol @samp{foo} bound to the
5368 unspecified base version of the symbol. The source file that contains this
5369 example would define 4 C functions: @samp{original_foo}, @samp{old_foo},
5370 @samp{old_foo1}, and @samp{new_foo}.
5372 When you have multiple definitions of a given symbol, there needs to be
5373 some way to specify a default version to which external references to
5374 this symbol will be bound. You can do this with the
5375 @samp{foo@@@@VERS_2.0} type of @samp{.symver} directive. You can only
5376 declare one version of a symbol as the default in this manner; otherwise
5377 you would effectively have multiple definitions of the same symbol.
5379 If you wish to bind a reference to a specific version of the symbol
5380 within the shared library, you can use the aliases of convenience
5381 (i.e., @samp{old_foo}), or you can use the @samp{.symver} directive to
5382 specifically bind to an external version of the function in question.
5384 You can also specify the language in the version script:
5387 VERSION extern "lang" @{ version-script-commands @}
5390 The supported @samp{lang}s are @samp{C}, @samp{C++}, and @samp{Java}.
5391 The linker will iterate over the list of symbols at the link time and
5392 demangle them according to @samp{lang} before matching them to the
5393 patterns specified in @samp{version-script-commands}. The default
5394 @samp{lang} is @samp{C}.
5396 Demangled names may contains spaces and other special characters. As
5397 described above, you can use a glob pattern to match demangled names,
5398 or you can use a double-quoted string to match the string exactly. In
5399 the latter case, be aware that minor differences (such as differing
5400 whitespace) between the version script and the demangler output will
5401 cause a mismatch. As the exact string generated by the demangler
5402 might change in the future, even if the mangled name does not, you
5403 should check that all of your version directives are behaving as you
5404 expect when you upgrade.
5407 @section Expressions in Linker Scripts
5410 The syntax for expressions in the linker script language is identical to
5411 that of C expressions. All expressions are evaluated as integers. All
5412 expressions are evaluated in the same size, which is 32 bits if both the
5413 host and target are 32 bits, and is otherwise 64 bits.
5415 You can use and set symbol values in expressions.
5417 The linker defines several special purpose builtin functions for use in
5421 * Constants:: Constants
5422 * Symbolic Constants:: Symbolic constants
5423 * Symbols:: Symbol Names
5424 * Orphan Sections:: Orphan Sections
5425 * Location Counter:: The Location Counter
5426 * Operators:: Operators
5427 * Evaluation:: Evaluation
5428 * Expression Section:: The Section of an Expression
5429 * Builtin Functions:: Builtin Functions
5433 @subsection Constants
5434 @cindex integer notation
5435 @cindex constants in linker scripts
5436 All constants are integers.
5438 As in C, the linker considers an integer beginning with @samp{0} to be
5439 octal, and an integer beginning with @samp{0x} or @samp{0X} to be
5440 hexadecimal. Alternatively the linker accepts suffixes of @samp{h} or
5441 @samp{H} for hexadecimal, @samp{o} or @samp{O} for octal, @samp{b} or
5442 @samp{B} for binary and @samp{d} or @samp{D} for decimal. Any integer
5443 value without a prefix or a suffix is considered to be decimal.
5445 @cindex scaled integers
5446 @cindex K and M integer suffixes
5447 @cindex M and K integer suffixes
5448 @cindex suffixes for integers
5449 @cindex integer suffixes
5450 In addition, you can use the suffixes @code{K} and @code{M} to scale a
5454 @c END TEXI2ROFF-KILL
5455 @code{1024} or @code{1024*1024}
5459 ${\rm 1024}$ or ${\rm 1024}^2$
5461 @c END TEXI2ROFF-KILL
5462 respectively. For example, the following
5463 all refer to the same quantity:
5472 Note - the @code{K} and @code{M} suffixes cannot be used in
5473 conjunction with the base suffixes mentioned above.
5475 @node Symbolic Constants
5476 @subsection Symbolic Constants
5477 @cindex symbolic constants
5479 It is possible to refer to target specific constants via the use of
5480 the @code{CONSTANT(@var{name})} operator, where @var{name} is one of:
5485 The target's maximum page size.
5487 @item COMMONPAGESIZE
5488 @kindex COMMONPAGESIZE
5489 The target's default page size.
5495 .text ALIGN (CONSTANT (MAXPAGESIZE)) : @{ *(.text) @}
5498 will create a text section aligned to the largest page boundary
5499 supported by the target.
5502 @subsection Symbol Names
5503 @cindex symbol names
5505 @cindex quoted symbol names
5507 Unless quoted, symbol names start with a letter, underscore, or period
5508 and may include letters, digits, underscores, periods, and hyphens.
5509 Unquoted symbol names must not conflict with any keywords. You can
5510 specify a symbol which contains odd characters or has the same name as a
5511 keyword by surrounding the symbol name in double quotes:
5514 "with a space" = "also with a space" + 10;
5517 Since symbols can contain many non-alphabetic characters, it is safest
5518 to delimit symbols with spaces. For example, @samp{A-B} is one symbol,
5519 whereas @samp{A - B} is an expression involving subtraction.
5521 @node Orphan Sections
5522 @subsection Orphan Sections
5524 Orphan sections are sections present in the input files which
5525 are not explicitly placed into the output file by the linker
5526 script. The linker will still copy these sections into the
5527 output file, but it has to guess as to where they should be
5528 placed. The linker uses a simple heuristic to do this. It
5529 attempts to place orphan sections after non-orphan sections of the
5530 same attribute, such as code vs data, loadable vs non-loadable, etc.
5531 If there is not enough room to do this then it places
5532 at the end of the file.
5534 For ELF targets, the attribute of the section includes section type as
5535 well as section flag.
5537 If an orphaned section's name is representable as a C identifier then
5538 the linker will automatically @pxref{PROVIDE} two symbols:
5539 __start_SECNAME and __stop_SECNAME, where SECNAME is the name of the
5540 section. These indicate the start address and end address of the
5541 orphaned section respectively. Note: most section names are not
5542 representable as C identifiers because they contain a @samp{.}
5545 @node Location Counter
5546 @subsection The Location Counter
5549 @cindex location counter
5550 @cindex current output location
5551 The special linker variable @dfn{dot} @samp{.} always contains the
5552 current output location counter. Since the @code{.} always refers to a
5553 location in an output section, it may only appear in an expression
5554 within a @code{SECTIONS} command. The @code{.} symbol may appear
5555 anywhere that an ordinary symbol is allowed in an expression.
5558 Assigning a value to @code{.} will cause the location counter to be
5559 moved. This may be used to create holes in the output section. The
5560 location counter may not be moved backwards inside an output section,
5561 and may not be moved backwards outside of an output section if so
5562 doing creates areas with overlapping LMAs.
5578 In the previous example, the @samp{.text} section from @file{file1} is
5579 located at the beginning of the output section @samp{output}. It is
5580 followed by a 1000 byte gap. Then the @samp{.text} section from
5581 @file{file2} appears, also with a 1000 byte gap following before the
5582 @samp{.text} section from @file{file3}. The notation @samp{= 0x12345678}
5583 specifies what data to write in the gaps (@pxref{Output Section Fill}).
5585 @cindex dot inside sections
5586 Note: @code{.} actually refers to the byte offset from the start of the
5587 current containing object. Normally this is the @code{SECTIONS}
5588 statement, whose start address is 0, hence @code{.} can be used as an
5589 absolute address. If @code{.} is used inside a section description
5590 however, it refers to the byte offset from the start of that section,
5591 not an absolute address. Thus in a script like this:
5609 The @samp{.text} section will be assigned a starting address of 0x100
5610 and a size of exactly 0x200 bytes, even if there is not enough data in
5611 the @samp{.text} input sections to fill this area. (If there is too
5612 much data, an error will be produced because this would be an attempt to
5613 move @code{.} backwards). The @samp{.data} section will start at 0x500
5614 and it will have an extra 0x600 bytes worth of space after the end of
5615 the values from the @samp{.data} input sections and before the end of
5616 the @samp{.data} output section itself.
5618 @cindex dot outside sections
5619 Setting symbols to the value of the location counter outside of an
5620 output section statement can result in unexpected values if the linker
5621 needs to place orphan sections. For example, given the following:
5627 .text: @{ *(.text) @}
5631 .data: @{ *(.data) @}
5636 If the linker needs to place some input section, e.g. @code{.rodata},
5637 not mentioned in the script, it might choose to place that section
5638 between @code{.text} and @code{.data}. You might think the linker
5639 should place @code{.rodata} on the blank line in the above script, but
5640 blank lines are of no particular significance to the linker. As well,
5641 the linker doesn't associate the above symbol names with their
5642 sections. Instead, it assumes that all assignments or other
5643 statements belong to the previous output section, except for the
5644 special case of an assignment to @code{.}. I.e., the linker will
5645 place the orphan @code{.rodata} section as if the script was written
5652 .text: @{ *(.text) @}
5656 .rodata: @{ *(.rodata) @}
5657 .data: @{ *(.data) @}
5662 This may or may not be the script author's intention for the value of
5663 @code{start_of_data}. One way to influence the orphan section
5664 placement is to assign the location counter to itself, as the linker
5665 assumes that an assignment to @code{.} is setting the start address of
5666 a following output section and thus should be grouped with that
5667 section. So you could write:
5673 .text: @{ *(.text) @}
5678 .data: @{ *(.data) @}
5683 Now, the orphan @code{.rodata} section will be placed between
5684 @code{end_of_text} and @code{start_of_data}.
5688 @subsection Operators
5689 @cindex operators for arithmetic
5690 @cindex arithmetic operators
5691 @cindex precedence in expressions
5692 The linker recognizes the standard C set of arithmetic operators, with
5693 the standard bindings and precedence levels:
5696 @c END TEXI2ROFF-KILL
5698 precedence associativity Operators Notes
5704 5 left == != > < <= >=
5710 11 right &= += -= *= /= (2)
5714 (1) Prefix operators
5715 (2) @xref{Assignments}.
5719 \vskip \baselineskip
5720 %"lispnarrowing" is the extra indent used generally for smallexample
5721 \hskip\lispnarrowing\vbox{\offinterlineskip
5724 {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr
5725 height2pt&\omit&&\omit&&\omit&\cr
5726 &Precedence&& Associativity &&{\rm Operators}&\cr
5727 height2pt&\omit&&\omit&&\omit&\cr
5729 height2pt&\omit&&\omit&&\omit&\cr
5731 % '176 is tilde, '~' in tt font
5732 &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr
5733 &2&&left&&* / \%&\cr
5736 &5&&left&&== != > < <= >=&\cr
5739 &8&&left&&{\&\&}&\cr
5742 &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr
5744 height2pt&\omit&&\omit&&\omit&\cr}
5749 @obeylines@parskip=0pt@parindent=0pt
5750 @dag@quad Prefix operators.
5751 @ddag@quad @xref{Assignments}.
5754 @c END TEXI2ROFF-KILL
5757 @subsection Evaluation
5758 @cindex lazy evaluation
5759 @cindex expression evaluation order
5760 The linker evaluates expressions lazily. It only computes the value of
5761 an expression when absolutely necessary.
5763 The linker needs some information, such as the value of the start
5764 address of the first section, and the origins and lengths of memory
5765 regions, in order to do any linking at all. These values are computed
5766 as soon as possible when the linker reads in the linker script.
5768 However, other values (such as symbol values) are not known or needed
5769 until after storage allocation. Such values are evaluated later, when
5770 other information (such as the sizes of output sections) is available
5771 for use in the symbol assignment expression.
5773 The sizes of sections cannot be known until after allocation, so
5774 assignments dependent upon these are not performed until after
5777 Some expressions, such as those depending upon the location counter
5778 @samp{.}, must be evaluated during section allocation.
5780 If the result of an expression is required, but the value is not
5781 available, then an error results. For example, a script like the
5787 .text 9+this_isnt_constant :
5793 will cause the error message @samp{non constant expression for initial
5796 @node Expression Section
5797 @subsection The Section of an Expression
5798 @cindex expression sections
5799 @cindex absolute expressions
5800 @cindex relative expressions
5801 @cindex absolute and relocatable symbols
5802 @cindex relocatable and absolute symbols
5803 @cindex symbols, relocatable and absolute
5804 Addresses and symbols may be section relative, or absolute. A section
5805 relative symbol is relocatable. If you request relocatable output
5806 using the @samp{-r} option, a further link operation may change the
5807 value of a section relative symbol. On the other hand, an absolute
5808 symbol will retain the same value throughout any further link
5811 Some terms in linker expressions are addresses. This is true of
5812 section relative symbols and for builtin functions that return an
5813 address, such as @code{ADDR}, @code{LOADADDR}, @code{ORIGIN} and
5814 @code{SEGMENT_START}. Other terms are simply numbers, or are builtin
5815 functions that return a non-address value, such as @code{LENGTH}.
5816 One complication is that unless you set @code{LD_FEATURE ("SANE_EXPR")}
5817 (@pxref{Miscellaneous Commands}), numbers and absolute symbols are treated
5818 differently depending on their location, for compatibility with older
5819 versions of @code{ld}. Expressions appearing outside an output
5820 section definition treat all numbers as absolute addresses.
5821 Expressions appearing inside an output section definition treat
5822 absolute symbols as numbers. If @code{LD_FEATURE ("SANE_EXPR")} is
5823 given, then absolute symbols and numbers are simply treated as numbers
5826 In the following simple example,
5833 __executable_start = 0x100;
5837 __data_start = 0x10;
5845 both @code{.} and @code{__executable_start} are set to the absolute
5846 address 0x100 in the first two assignments, then both @code{.} and
5847 @code{__data_start} are set to 0x10 relative to the @code{.data}
5848 section in the second two assignments.
5850 For expressions involving numbers, relative addresses and absolute
5851 addresses, ld follows these rules to evaluate terms:
5855 Unary operations on an absolute address or number, and binary
5856 operations on two absolute addresses or two numbers, or between one
5857 absolute address and a number, apply the operator to the value(s).
5859 Unary operations on a relative address, and binary operations on two
5860 relative addresses in the same section or between one relative address
5861 and a number, apply the operator to the offset part of the address(es).
5863 Other binary operations, that is, between two relative addresses not
5864 in the same section, or between a relative address and an absolute
5865 address, first convert any non-absolute term to an absolute address
5866 before applying the operator.
5869 The result section of each sub-expression is as follows:
5873 An operation involving only numbers results in a number.
5875 The result of comparisons, @samp{&&} and @samp{||} is also a number.
5877 The result of other binary arithmetic and logical operations on two
5878 relative addresses in the same section or two absolute addresses
5879 (after above conversions) is also a number.
5881 The result of other operations on relative addresses or one
5882 relative address and a number, is a relative address in the same
5883 section as the relative operand(s).
5885 The result of other operations on absolute addresses (after above
5886 conversions) is an absolute address.
5889 You can use the builtin function @code{ABSOLUTE} to force an expression
5890 to be absolute when it would otherwise be relative. For example, to
5891 create an absolute symbol set to the address of the end of the output
5892 section @samp{.data}:
5896 .data : @{ *(.data) _edata = ABSOLUTE(.); @}
5900 If @samp{ABSOLUTE} were not used, @samp{_edata} would be relative to the
5901 @samp{.data} section.
5903 Using @code{LOADADDR} also forces an expression absolute, since this
5904 particular builtin function returns an absolute address.
5906 @node Builtin Functions
5907 @subsection Builtin Functions
5908 @cindex functions in expressions
5909 The linker script language includes a number of builtin functions for
5910 use in linker script expressions.
5913 @item ABSOLUTE(@var{exp})
5914 @kindex ABSOLUTE(@var{exp})
5915 @cindex expression, absolute
5916 Return the absolute (non-relocatable, as opposed to non-negative) value
5917 of the expression @var{exp}. Primarily useful to assign an absolute
5918 value to a symbol within a section definition, where symbol values are
5919 normally section relative. @xref{Expression Section}.
5921 @item ADDR(@var{section})
5922 @kindex ADDR(@var{section})
5923 @cindex section address in expression
5924 Return the address (VMA) of the named @var{section}. Your
5925 script must previously have defined the location of that section. In
5926 the following example, @code{start_of_output_1}, @code{symbol_1} and
5927 @code{symbol_2} are assigned equivalent values, except that
5928 @code{symbol_1} will be relative to the @code{.output1} section while
5929 the other two will be absolute:
5935 start_of_output_1 = ABSOLUTE(.);
5940 symbol_1 = ADDR(.output1);
5941 symbol_2 = start_of_output_1;
5947 @item ALIGN(@var{align})
5948 @itemx ALIGN(@var{exp},@var{align})
5949 @kindex ALIGN(@var{align})
5950 @kindex ALIGN(@var{exp},@var{align})
5951 @cindex round up location counter
5952 @cindex align location counter
5953 @cindex round up expression
5954 @cindex align expression
5955 Return the location counter (@code{.}) or arbitrary expression aligned
5956 to the next @var{align} boundary. The single operand @code{ALIGN}
5957 doesn't change the value of the location counter---it just does
5958 arithmetic on it. The two operand @code{ALIGN} allows an arbitrary
5959 expression to be aligned upwards (@code{ALIGN(@var{align})} is
5960 equivalent to @code{ALIGN(., @var{align})}).
5962 Here is an example which aligns the output @code{.data} section to the
5963 next @code{0x2000} byte boundary after the preceding section and sets a
5964 variable within the section to the next @code{0x8000} boundary after the
5969 .data ALIGN(0x2000): @{
5971 variable = ALIGN(0x8000);
5977 The first use of @code{ALIGN} in this example specifies the location of
5978 a section because it is used as the optional @var{address} attribute of
5979 a section definition (@pxref{Output Section Address}). The second use
5980 of @code{ALIGN} is used to defines the value of a symbol.
5982 The builtin function @code{NEXT} is closely related to @code{ALIGN}.
5984 @item ALIGNOF(@var{section})
5985 @kindex ALIGNOF(@var{section})
5986 @cindex section alignment
5987 Return the alignment in bytes of the named @var{section}, if that section has
5988 been allocated. If the section has not been allocated when this is
5989 evaluated, the linker will report an error. In the following example,
5990 the alignment of the @code{.output} section is stored as the first
5991 value in that section.
5996 LONG (ALIGNOF (.output))
6003 @item BLOCK(@var{exp})
6004 @kindex BLOCK(@var{exp})
6005 This is a synonym for @code{ALIGN}, for compatibility with older linker
6006 scripts. It is most often seen when setting the address of an output
6009 @item DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6010 @kindex DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
6011 This is equivalent to either
6013 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - 1)))
6017 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - @var{commonpagesize})))
6020 depending on whether the latter uses fewer @var{commonpagesize} sized pages
6021 for the data segment (area between the result of this expression and
6022 @code{DATA_SEGMENT_END}) than the former or not.
6023 If the latter form is used, it means @var{commonpagesize} bytes of runtime
6024 memory will be saved at the expense of up to @var{commonpagesize} wasted
6025 bytes in the on-disk file.
6027 This expression can only be used directly in @code{SECTIONS} commands, not in
6028 any output section descriptions and only once in the linker script.
6029 @var{commonpagesize} should be less or equal to @var{maxpagesize} and should
6030 be the system page size the object wants to be optimized for (while still
6031 working on system page sizes up to @var{maxpagesize}).
6036 . = DATA_SEGMENT_ALIGN(0x10000, 0x2000);
6039 @item DATA_SEGMENT_END(@var{exp})
6040 @kindex DATA_SEGMENT_END(@var{exp})
6041 This defines the end of data segment for @code{DATA_SEGMENT_ALIGN}
6042 evaluation purposes.
6045 . = DATA_SEGMENT_END(.);
6048 @item DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6049 @kindex DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
6050 This defines the end of the @code{PT_GNU_RELRO} segment when
6051 @samp{-z relro} option is used.
6052 When @samp{-z relro} option is not present, @code{DATA_SEGMENT_RELRO_END}
6053 does nothing, otherwise @code{DATA_SEGMENT_ALIGN} is padded so that
6054 @var{exp} + @var{offset} is aligned to the most commonly used page
6055 boundary for particular target. If present in the linker script,
6056 it must always come in between @code{DATA_SEGMENT_ALIGN} and
6057 @code{DATA_SEGMENT_END}. Evaluates to the second argument plus any
6058 padding needed at the end of the @code{PT_GNU_RELRO} segment due to
6062 . = DATA_SEGMENT_RELRO_END(24, .);
6065 @item DEFINED(@var{symbol})
6066 @kindex DEFINED(@var{symbol})
6067 @cindex symbol defaults
6068 Return 1 if @var{symbol} is in the linker global symbol table and is
6069 defined before the statement using DEFINED in the script, otherwise
6070 return 0. You can use this function to provide
6071 default values for symbols. For example, the following script fragment
6072 shows how to set a global symbol @samp{begin} to the first location in
6073 the @samp{.text} section---but if a symbol called @samp{begin} already
6074 existed, its value is preserved:
6080 begin = DEFINED(begin) ? begin : . ;
6088 @item LENGTH(@var{memory})
6089 @kindex LENGTH(@var{memory})
6090 Return the length of the memory region named @var{memory}.
6092 @item LOADADDR(@var{section})
6093 @kindex LOADADDR(@var{section})
6094 @cindex section load address in expression
6095 Return the absolute LMA of the named @var{section}. (@pxref{Output
6098 @item LOG2CEIL(@var{exp})
6099 @kindex LOG2CEIL(@var{exp})
6100 Return the binary logarithm of @var{exp} rounded towards infinity.
6101 @code{LOG2CEIL(0)} returns 0.
6104 @item MAX(@var{exp1}, @var{exp2})
6105 Returns the maximum of @var{exp1} and @var{exp2}.
6108 @item MIN(@var{exp1}, @var{exp2})
6109 Returns the minimum of @var{exp1} and @var{exp2}.
6111 @item NEXT(@var{exp})
6112 @kindex NEXT(@var{exp})
6113 @cindex unallocated address, next
6114 Return the next unallocated address that is a multiple of @var{exp}.
6115 This function is closely related to @code{ALIGN(@var{exp})}; unless you
6116 use the @code{MEMORY} command to define discontinuous memory for the
6117 output file, the two functions are equivalent.
6119 @item ORIGIN(@var{memory})
6120 @kindex ORIGIN(@var{memory})
6121 Return the origin of the memory region named @var{memory}.
6123 @item SEGMENT_START(@var{segment}, @var{default})
6124 @kindex SEGMENT_START(@var{segment}, @var{default})
6125 Return the base address of the named @var{segment}. If an explicit
6126 value has already been given for this segment (with a command-line
6127 @samp{-T} option) then that value will be returned otherwise the value
6128 will be @var{default}. At present, the @samp{-T} command-line option
6129 can only be used to set the base address for the ``text'', ``data'', and
6130 ``bss'' sections, but you can use @code{SEGMENT_START} with any segment
6133 @item SIZEOF(@var{section})
6134 @kindex SIZEOF(@var{section})
6135 @cindex section size
6136 Return the size in bytes of the named @var{section}, if that section has
6137 been allocated. If the section has not been allocated when this is
6138 evaluated, the linker will report an error. In the following example,
6139 @code{symbol_1} and @code{symbol_2} are assigned identical values:
6148 symbol_1 = .end - .start ;
6149 symbol_2 = SIZEOF(.output);
6154 @item SIZEOF_HEADERS
6155 @itemx sizeof_headers
6156 @kindex SIZEOF_HEADERS
6158 Return the size in bytes of the output file's headers. This is
6159 information which appears at the start of the output file. You can use
6160 this number when setting the start address of the first section, if you
6161 choose, to facilitate paging.
6163 @cindex not enough room for program headers
6164 @cindex program headers, not enough room
6165 When producing an ELF output file, if the linker script uses the
6166 @code{SIZEOF_HEADERS} builtin function, the linker must compute the
6167 number of program headers before it has determined all the section
6168 addresses and sizes. If the linker later discovers that it needs
6169 additional program headers, it will report an error @samp{not enough
6170 room for program headers}. To avoid this error, you must avoid using
6171 the @code{SIZEOF_HEADERS} function, or you must rework your linker
6172 script to avoid forcing the linker to use additional program headers, or
6173 you must define the program headers yourself using the @code{PHDRS}
6174 command (@pxref{PHDRS}).
6177 @node Implicit Linker Scripts
6178 @section Implicit Linker Scripts
6179 @cindex implicit linker scripts
6180 If you specify a linker input file which the linker can not recognize as
6181 an object file or an archive file, it will try to read the file as a
6182 linker script. If the file can not be parsed as a linker script, the
6183 linker will report an error.
6185 An implicit linker script will not replace the default linker script.
6187 Typically an implicit linker script would contain only symbol
6188 assignments, or the @code{INPUT}, @code{GROUP}, or @code{VERSION}
6191 Any input files read because of an implicit linker script will be read
6192 at the position in the command line where the implicit linker script was
6193 read. This can affect archive searching.
6196 @node Machine Dependent
6197 @chapter Machine Dependent Features
6199 @cindex machine dependencies
6200 @command{ld} has additional features on some platforms; the following
6201 sections describe them. Machines where @command{ld} has no additional
6202 functionality are not listed.
6206 * H8/300:: @command{ld} and the H8/300
6209 * i960:: @command{ld} and the Intel 960 family
6212 * M68HC11/68HC12:: @code{ld} and the Motorola 68HC11 and 68HC12 families
6215 * ARM:: @command{ld} and the ARM family
6218 * HPPA ELF32:: @command{ld} and HPPA 32-bit ELF
6221 * M68K:: @command{ld} and the Motorola 68K family
6224 * MIPS:: @command{ld} and the MIPS family
6227 * MMIX:: @command{ld} and MMIX
6230 * MSP430:: @command{ld} and MSP430
6233 * NDS32:: @command{ld} and NDS32
6236 * Nios II:: @command{ld} and the Altera Nios II
6239 * PowerPC ELF32:: @command{ld} and PowerPC 32-bit ELF Support
6242 * PowerPC64 ELF64:: @command{ld} and PowerPC64 64-bit ELF Support
6245 * SPU ELF:: @command{ld} and SPU ELF Support
6248 * TI COFF:: @command{ld} and TI COFF
6251 * WIN32:: @command{ld} and WIN32 (cygwin/mingw)
6254 * Xtensa:: @command{ld} and Xtensa Processors
6265 @section @command{ld} and the H8/300
6267 @cindex H8/300 support
6268 For the H8/300, @command{ld} can perform these global optimizations when
6269 you specify the @samp{--relax} command-line option.
6272 @cindex relaxing on H8/300
6273 @item relaxing address modes
6274 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
6275 targets are within eight bits, and turns them into eight-bit
6276 program-counter relative @code{bsr} and @code{bra} instructions,
6279 @cindex synthesizing on H8/300
6280 @item synthesizing instructions
6281 @c FIXME: specifically mov.b, or any mov instructions really? -> mov.b only, at least on H8, H8H, H8S
6282 @command{ld} finds all @code{mov.b} instructions which use the
6283 sixteen-bit absolute address form, but refer to the top
6284 page of memory, and changes them to use the eight-bit address form.
6285 (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into
6286 @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the
6287 top page of memory).
6289 @command{ld} finds all @code{mov} instructions which use the register
6290 indirect with 32-bit displacement addressing mode, but use a small
6291 displacement inside 16-bit displacement range, and changes them to use
6292 the 16-bit displacement form. (That is: the linker turns @samp{mov.b
6293 @code{@@}@var{d}:32,ERx} into @samp{mov.b @code{@@}@var{d}:16,ERx}
6294 whenever the displacement @var{d} is in the 16 bit signed integer
6295 range. Only implemented in ELF-format ld).
6297 @item bit manipulation instructions
6298 @command{ld} finds all bit manipulation instructions like @code{band, bclr,
6299 biand, bild, bior, bist, bixor, bld, bnot, bor, bset, bst, btst, bxor}
6300 which use 32 bit and 16 bit absolute address form, but refer to the top
6301 page of memory, and changes them to use the 8 bit address form.
6302 (That is: the linker turns @samp{bset #xx:3,@code{@@}@var{aa}:32} into
6303 @samp{bset #xx:3,@code{@@}@var{aa}:8} whenever the address @var{aa} is in
6304 the top page of memory).
6306 @item system control instructions
6307 @command{ld} finds all @code{ldc.w, stc.w} instructions which use the
6308 32 bit absolute address form, but refer to the top page of memory, and
6309 changes them to use 16 bit address form.
6310 (That is: the linker turns @samp{ldc.w @code{@@}@var{aa}:32,ccr} into
6311 @samp{ldc.w @code{@@}@var{aa}:16,ccr} whenever the address @var{aa} is in
6312 the top page of memory).
6322 @c This stuff is pointless to say unless you're especially concerned
6323 @c with Renesas chips; don't enable it for generic case, please.
6325 @chapter @command{ld} and Other Renesas Chips
6327 @command{ld} also supports the Renesas (formerly Hitachi) H8/300H,
6328 H8/500, and SH chips. No special features, commands, or command-line
6329 options are required for these chips.
6339 @section @command{ld} and the Intel 960 Family
6341 @cindex i960 support
6343 You can use the @samp{-A@var{architecture}} command line option to
6344 specify one of the two-letter names identifying members of the 960
6345 family; the option specifies the desired output target, and warns of any
6346 incompatible instructions in the input files. It also modifies the
6347 linker's search strategy for archive libraries, to support the use of
6348 libraries specific to each particular architecture, by including in the
6349 search loop names suffixed with the string identifying the architecture.
6351 For example, if your @command{ld} command line included @w{@samp{-ACA}} as
6352 well as @w{@samp{-ltry}}, the linker would look (in its built-in search
6353 paths, and in any paths you specify with @samp{-L}) for a library with
6366 The first two possibilities would be considered in any event; the last
6367 two are due to the use of @w{@samp{-ACA}}.
6369 You can meaningfully use @samp{-A} more than once on a command line, since
6370 the 960 architecture family allows combination of target architectures; each
6371 use will add another pair of name variants to search for when @w{@samp{-l}}
6372 specifies a library.
6374 @cindex @option{--relax} on i960
6375 @cindex relaxing on i960
6376 @command{ld} supports the @samp{--relax} option for the i960 family. If
6377 you specify @samp{--relax}, @command{ld} finds all @code{balx} and
6378 @code{calx} instructions whose targets are within 24 bits, and turns
6379 them into 24-bit program-counter relative @code{bal} and @code{cal}
6380 instructions, respectively. @command{ld} also turns @code{cal}
6381 instructions into @code{bal} instructions when it determines that the
6382 target subroutine is a leaf routine (that is, the target subroutine does
6383 not itself call any subroutines).
6400 @node M68HC11/68HC12
6401 @section @command{ld} and the Motorola 68HC11 and 68HC12 families
6403 @cindex M68HC11 and 68HC12 support
6405 @subsection Linker Relaxation
6407 For the Motorola 68HC11, @command{ld} can perform these global
6408 optimizations when you specify the @samp{--relax} command-line option.
6411 @cindex relaxing on M68HC11
6412 @item relaxing address modes
6413 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
6414 targets are within eight bits, and turns them into eight-bit
6415 program-counter relative @code{bsr} and @code{bra} instructions,
6418 @command{ld} also looks at all 16-bit extended addressing modes and
6419 transforms them in a direct addressing mode when the address is in
6420 page 0 (between 0 and 0x0ff).
6422 @item relaxing gcc instruction group
6423 When @command{gcc} is called with @option{-mrelax}, it can emit group
6424 of instructions that the linker can optimize to use a 68HC11 direct
6425 addressing mode. These instructions consists of @code{bclr} or
6426 @code{bset} instructions.
6430 @subsection Trampoline Generation
6432 @cindex trampoline generation on M68HC11
6433 @cindex trampoline generation on M68HC12
6434 For 68HC11 and 68HC12, @command{ld} can generate trampoline code to
6435 call a far function using a normal @code{jsr} instruction. The linker
6436 will also change the relocation to some far function to use the
6437 trampoline address instead of the function address. This is typically the
6438 case when a pointer to a function is taken. The pointer will in fact
6439 point to the function trampoline.
6447 @section @command{ld} and the ARM family
6449 @cindex ARM interworking support
6450 @kindex --support-old-code
6451 For the ARM, @command{ld} will generate code stubs to allow functions calls
6452 between ARM and Thumb code. These stubs only work with code that has
6453 been compiled and assembled with the @samp{-mthumb-interwork} command
6454 line option. If it is necessary to link with old ARM object files or
6455 libraries, which have not been compiled with the -mthumb-interwork
6456 option then the @samp{--support-old-code} command line switch should be
6457 given to the linker. This will make it generate larger stub functions
6458 which will work with non-interworking aware ARM code. Note, however,
6459 the linker does not support generating stubs for function calls to
6460 non-interworking aware Thumb code.
6462 @cindex thumb entry point
6463 @cindex entry point, thumb
6464 @kindex --thumb-entry=@var{entry}
6465 The @samp{--thumb-entry} switch is a duplicate of the generic
6466 @samp{--entry} switch, in that it sets the program's starting address.
6467 But it also sets the bottom bit of the address, so that it can be
6468 branched to using a BX instruction, and the program will start
6469 executing in Thumb mode straight away.
6471 @cindex PE import table prefixing
6472 @kindex --use-nul-prefixed-import-tables
6473 The @samp{--use-nul-prefixed-import-tables} switch is specifying, that
6474 the import tables idata4 and idata5 have to be generated with a zero
6475 element prefix for import libraries. This is the old style to generate
6476 import tables. By default this option is turned off.
6480 The @samp{--be8} switch instructs @command{ld} to generate BE8 format
6481 executables. This option is only valid when linking big-endian
6482 objects - ie ones which have been assembled with the @option{-EB}
6483 option. The resulting image will contain big-endian data and
6487 @kindex --target1-rel
6488 @kindex --target1-abs
6489 The @samp{R_ARM_TARGET1} relocation is typically used for entries in the
6490 @samp{.init_array} section. It is interpreted as either @samp{R_ARM_REL32}
6491 or @samp{R_ARM_ABS32}, depending on the target. The @samp{--target1-rel}
6492 and @samp{--target1-abs} switches override the default.
6495 @kindex --target2=@var{type}
6496 The @samp{--target2=type} switch overrides the default definition of the
6497 @samp{R_ARM_TARGET2} relocation. Valid values for @samp{type}, their
6498 meanings, and target defaults are as follows:
6501 @samp{R_ARM_REL32} (arm*-*-elf, arm*-*-eabi)
6503 @samp{R_ARM_ABS32} (arm*-*-symbianelf)
6505 @samp{R_ARM_GOT_PREL} (arm*-*-linux, arm*-*-*bsd)
6510 The @samp{R_ARM_V4BX} relocation (defined by the ARM AAELF
6511 specification) enables objects compiled for the ARMv4 architecture to be
6512 interworking-safe when linked with other objects compiled for ARMv4t, but
6513 also allows pure ARMv4 binaries to be built from the same ARMv4 objects.
6515 In the latter case, the switch @option{--fix-v4bx} must be passed to the
6516 linker, which causes v4t @code{BX rM} instructions to be rewritten as
6517 @code{MOV PC,rM}, since v4 processors do not have a @code{BX} instruction.
6519 In the former case, the switch should not be used, and @samp{R_ARM_V4BX}
6520 relocations are ignored.
6522 @cindex FIX_V4BX_INTERWORKING
6523 @kindex --fix-v4bx-interworking
6524 Replace @code{BX rM} instructions identified by @samp{R_ARM_V4BX}
6525 relocations with a branch to the following veneer:
6533 This allows generation of libraries/applications that work on ARMv4 cores
6534 and are still interworking safe. Note that the above veneer clobbers the
6535 condition flags, so may cause incorrect program behavior in rare cases.
6539 The @samp{--use-blx} switch enables the linker to use ARM/Thumb
6540 BLX instructions (available on ARMv5t and above) in various
6541 situations. Currently it is used to perform calls via the PLT from Thumb
6542 code using BLX rather than using BX and a mode-switching stub before
6543 each PLT entry. This should lead to such calls executing slightly faster.
6545 This option is enabled implicitly for SymbianOS, so there is no need to
6546 specify it if you are using that target.
6548 @cindex VFP11_DENORM_FIX
6549 @kindex --vfp11-denorm-fix
6550 The @samp{--vfp11-denorm-fix} switch enables a link-time workaround for a
6551 bug in certain VFP11 coprocessor hardware, which sometimes allows
6552 instructions with denorm operands (which must be handled by support code)
6553 to have those operands overwritten by subsequent instructions before
6554 the support code can read the intended values.
6556 The bug may be avoided in scalar mode if you allow at least one
6557 intervening instruction between a VFP11 instruction which uses a register
6558 and another instruction which writes to the same register, or at least two
6559 intervening instructions if vector mode is in use. The bug only affects
6560 full-compliance floating-point mode: you do not need this workaround if
6561 you are using "runfast" mode. Please contact ARM for further details.
6563 If you know you are using buggy VFP11 hardware, you can
6564 enable this workaround by specifying the linker option
6565 @samp{--vfp-denorm-fix=scalar} if you are using the VFP11 scalar
6566 mode only, or @samp{--vfp-denorm-fix=vector} if you are using
6567 vector mode (the latter also works for scalar code). The default is
6568 @samp{--vfp-denorm-fix=none}.
6570 If the workaround is enabled, instructions are scanned for
6571 potentially-troublesome sequences, and a veneer is created for each
6572 such sequence which may trigger the erratum. The veneer consists of the
6573 first instruction of the sequence and a branch back to the subsequent
6574 instruction. The original instruction is then replaced with a branch to
6575 the veneer. The extra cycles required to call and return from the veneer
6576 are sufficient to avoid the erratum in both the scalar and vector cases.
6578 @cindex ARM1176 erratum workaround
6579 @kindex --fix-arm1176
6580 @kindex --no-fix-arm1176
6581 The @samp{--fix-arm1176} switch enables a link-time workaround for an erratum
6582 in certain ARM1176 processors. The workaround is enabled by default if you
6583 are targeting ARM v6 (excluding ARM v6T2) or earlier. It can be disabled
6584 unconditionally by specifying @samp{--no-fix-arm1176}.
6586 Further information is available in the ``ARM1176JZ-S and ARM1176JZF-S
6587 Programmer Advice Notice'' available on the ARM documentation website at:
6588 http://infocenter.arm.com/.
6590 @cindex NO_ENUM_SIZE_WARNING
6591 @kindex --no-enum-size-warning
6592 The @option{--no-enum-size-warning} switch prevents the linker from
6593 warning when linking object files that specify incompatible EABI
6594 enumeration size attributes. For example, with this switch enabled,
6595 linking of an object file using 32-bit enumeration values with another
6596 using enumeration values fitted into the smallest possible space will
6599 @cindex NO_WCHAR_SIZE_WARNING
6600 @kindex --no-wchar-size-warning
6601 The @option{--no-wchar-size-warning} switch prevents the linker from
6602 warning when linking object files that specify incompatible EABI
6603 @code{wchar_t} size attributes. For example, with this switch enabled,
6604 linking of an object file using 32-bit @code{wchar_t} values with another
6605 using 16-bit @code{wchar_t} values will not be diagnosed.
6608 @kindex --pic-veneer
6609 The @samp{--pic-veneer} switch makes the linker use PIC sequences for
6610 ARM/Thumb interworking veneers, even if the rest of the binary
6611 is not PIC. This avoids problems on uClinux targets where
6612 @samp{--emit-relocs} is used to generate relocatable binaries.
6614 @cindex STUB_GROUP_SIZE
6615 @kindex --stub-group-size=@var{N}
6616 The linker will automatically generate and insert small sequences of
6617 code into a linked ARM ELF executable whenever an attempt is made to
6618 perform a function call to a symbol that is too far away. The
6619 placement of these sequences of instructions - called stubs - is
6620 controlled by the command line option @option{--stub-group-size=N}.
6621 The placement is important because a poor choice can create a need for
6622 duplicate stubs, increasing the code size. The linker will try to
6623 group stubs together in order to reduce interruptions to the flow of
6624 code, but it needs guidance as to how big these groups should be and
6625 where they should be placed.
6627 The value of @samp{N}, the parameter to the
6628 @option{--stub-group-size=} option controls where the stub groups are
6629 placed. If it is negative then all stubs are placed after the first
6630 branch that needs them. If it is positive then the stubs can be
6631 placed either before or after the branches that need them. If the
6632 value of @samp{N} is 1 (either +1 or -1) then the linker will choose
6633 exactly where to place groups of stubs, using its built in heuristics.
6634 A value of @samp{N} greater than 1 (or smaller than -1) tells the
6635 linker that a single group of stubs can service at most @samp{N} bytes
6636 from the input sections.
6638 The default, if @option{--stub-group-size=} is not specified, is
6641 Farcalls stubs insertion is fully supported for the ARM-EABI target
6642 only, because it relies on object files properties not present
6645 @cindex Cortex-A8 erratum workaround
6646 @kindex --fix-cortex-a8
6647 @kindex --no-fix-cortex-a8
6648 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}.
6650 The erratum only affects Thumb-2 code. Please contact ARM for further details.
6652 @cindex Cortex-A53 erratum 835769 workaround
6653 @kindex --fix-cortex-a53-835769
6654 @kindex --no-fix-cortex-a53-835769
6655 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}.
6657 Please contact ARM for further details.
6659 @kindex --merge-exidx-entries
6660 @kindex --no-merge-exidx-entries
6661 @cindex Merging exidx entries
6662 The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent exidx entries in debuginfo.
6665 @cindex 32-bit PLT entries
6666 The @samp{--long-plt} option enables the use of 16 byte PLT entries
6667 which support up to 4Gb of code. The default is to use 12 byte PLT
6668 entries which only support 512Mb of code.
6681 @section @command{ld} and HPPA 32-bit ELF Support
6682 @cindex HPPA multiple sub-space stubs
6683 @kindex --multi-subspace
6684 When generating a shared library, @command{ld} will by default generate
6685 import stubs suitable for use with a single sub-space application.
6686 The @samp{--multi-subspace} switch causes @command{ld} to generate export
6687 stubs, and different (larger) import stubs suitable for use with
6688 multiple sub-spaces.
6690 @cindex HPPA stub grouping
6691 @kindex --stub-group-size=@var{N}
6692 Long branch stubs and import/export stubs are placed by @command{ld} in
6693 stub sections located between groups of input sections.
6694 @samp{--stub-group-size} specifies the maximum size of a group of input
6695 sections handled by one stub section. Since branch offsets are signed,
6696 a stub section may serve two groups of input sections, one group before
6697 the stub section, and one group after it. However, when using
6698 conditional branches that require stubs, it may be better (for branch
6699 prediction) that stub sections only serve one group of input sections.
6700 A negative value for @samp{N} chooses this scheme, ensuring that
6701 branches to stubs always use a negative offset. Two special values of
6702 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
6703 @command{ld} to automatically size input section groups for the branch types
6704 detected, with the same behaviour regarding stub placement as other
6705 positive or negative values of @samp{N} respectively.
6707 Note that @samp{--stub-group-size} does not split input sections. A
6708 single input section larger than the group size specified will of course
6709 create a larger group (of one section). If input sections are too
6710 large, it may not be possible for a branch to reach its stub.
6723 @section @command{ld} and the Motorola 68K family
6725 @cindex Motorola 68K GOT generation
6726 @kindex --got=@var{type}
6727 The @samp{--got=@var{type}} option lets you choose the GOT generation scheme.
6728 The choices are @samp{single}, @samp{negative}, @samp{multigot} and
6729 @samp{target}. When @samp{target} is selected the linker chooses
6730 the default GOT generation scheme for the current target.
6731 @samp{single} tells the linker to generate a single GOT with
6732 entries only at non-negative offsets.
6733 @samp{negative} instructs the linker to generate a single GOT with
6734 entries at both negative and positive offsets. Not all environments
6736 @samp{multigot} allows the linker to generate several GOTs in the
6737 output file. All GOT references from a single input object
6738 file access the same GOT, but references from different input object
6739 files might access different GOTs. Not all environments support such GOTs.
6752 @section @command{ld} and the MIPS family
6754 @cindex MIPS microMIPS instruction choice selection
6757 The @samp{--insn32} and @samp{--no-insn32} options control the choice of
6758 microMIPS instructions used in code generated by the linker, such as that
6759 in the PLT or lazy binding stubs, or in relaxation. If @samp{--insn32} is
6760 used, then the linker only uses 32-bit instruction encodings. By default
6761 or if @samp{--no-insn32} is used, all instruction encodings are used,
6762 including 16-bit ones where possible.
6775 @section @code{ld} and MMIX
6776 For MMIX, there is a choice of generating @code{ELF} object files or
6777 @code{mmo} object files when linking. The simulator @code{mmix}
6778 understands the @code{mmo} format. The binutils @code{objcopy} utility
6779 can translate between the two formats.
6781 There is one special section, the @samp{.MMIX.reg_contents} section.
6782 Contents in this section is assumed to correspond to that of global
6783 registers, and symbols referring to it are translated to special symbols,
6784 equal to registers. In a final link, the start address of the
6785 @samp{.MMIX.reg_contents} section corresponds to the first allocated
6786 global register multiplied by 8. Register @code{$255} is not included in
6787 this section; it is always set to the program entry, which is at the
6788 symbol @code{Main} for @code{mmo} files.
6790 Global symbols with the prefix @code{__.MMIX.start.}, for example
6791 @code{__.MMIX.start..text} and @code{__.MMIX.start..data} are special.
6792 The default linker script uses these to set the default start address
6795 Initial and trailing multiples of zero-valued 32-bit words in a section,
6796 are left out from an mmo file.
6809 @section @code{ld} and MSP430
6810 For the MSP430 it is possible to select the MPU architecture. The flag @samp{-m [mpu type]}
6811 will select an appropriate linker script for selected MPU type. (To get a list of known MPUs
6812 just pass @samp{-m help} option to the linker).
6814 @cindex MSP430 extra sections
6815 The linker will recognize some extra sections which are MSP430 specific:
6818 @item @samp{.vectors}
6819 Defines a portion of ROM where interrupt vectors located.
6821 @item @samp{.bootloader}
6822 Defines the bootloader portion of the ROM (if applicable). Any code
6823 in this section will be uploaded to the MPU.
6825 @item @samp{.infomem}
6826 Defines an information memory section (if applicable). Any code in
6827 this section will be uploaded to the MPU.
6829 @item @samp{.infomemnobits}
6830 This is the same as the @samp{.infomem} section except that any code
6831 in this section will not be uploaded to the MPU.
6833 @item @samp{.noinit}
6834 Denotes a portion of RAM located above @samp{.bss} section.
6836 The last two sections are used by gcc.
6850 @section @code{ld} and NDS32
6851 @kindex relaxing on NDS32
6852 For NDS32, there are some options to select relaxation behavior. The linker
6853 relaxes objects according to these options.
6856 @item @samp{--m[no-]fp-as-gp}
6857 Disable/enable fp-as-gp relaxation.
6859 @item @samp{--mexport-symbols=FILE}
6860 Exporting symbols and their address into FILE as linker script.
6862 @item @samp{--m[no-]ex9}
6863 Disable/enable link-time EX9 relaxation.
6865 @item @samp{--mexport-ex9=FILE}
6866 Export the EX9 table after linking.
6868 @item @samp{--mimport-ex9=FILE}
6869 Import the Ex9 table for EX9 relaxation.
6871 @item @samp{--mupdate-ex9}
6872 Update the existing EX9 table.
6874 @item @samp{--mex9-limit=NUM}
6875 Maximum number of entries in the ex9 table.
6877 @item @samp{--mex9-loop-aware}
6878 Avoid generating the EX9 instruction inside the loop.
6880 @item @samp{--m[no-]ifc}
6881 Disable/enable the link-time IFC optimization.
6883 @item @samp{--mifc-loop-aware}
6884 Avoid generating the IFC instruction inside the loop.
6898 @section @command{ld} and the Altera Nios II
6899 @cindex Nios II call relaxation
6900 @kindex --relax on Nios II
6902 Call and immediate jump instructions on Nios II processors are limited to
6903 transferring control to addresses in the same 256MB memory segment,
6904 which may result in @command{ld} giving
6905 @samp{relocation truncated to fit} errors with very large programs.
6906 The command-line option @option{--relax} enables the generation of
6907 trampolines that can access the entire 32-bit address space for calls
6908 outside the normal @code{call} and @code{jmpi} address range. These
6909 trampolines are inserted at section boundaries, so may not themselves
6910 be reachable if an input section and its associated call trampolines are
6913 The @option{--relax} option is enabled by default unless @option{-r}
6914 is also specified. You can disable trampoline generation by using the
6915 @option{--no-relax} linker option. You can also disable this optimization
6916 locally by using the @samp{set .noat} directive in assembly-language
6917 source files, as the linker-inserted trampolines use the @code{at}
6918 register as a temporary.
6920 Note that the linker @option{--relax} option is independent of assembler
6921 relaxation options, and that using the GNU assembler's @option{-relax-all}
6922 option interferes with the linker's more selective call instruction relaxation.
6935 @section @command{ld} and PowerPC 32-bit ELF Support
6936 @cindex PowerPC long branches
6937 @kindex --relax on PowerPC
6938 Branches on PowerPC processors are limited to a signed 26-bit
6939 displacement, which may result in @command{ld} giving
6940 @samp{relocation truncated to fit} errors with very large programs.
6941 @samp{--relax} enables the generation of trampolines that can access
6942 the entire 32-bit address space. These trampolines are inserted at
6943 section boundaries, so may not themselves be reachable if an input
6944 section exceeds 33M in size. You may combine @samp{-r} and
6945 @samp{--relax} to add trampolines in a partial link. In that case
6946 both branches to undefined symbols and inter-section branches are also
6947 considered potentially out of range, and trampolines inserted.
6949 @cindex PowerPC ELF32 options
6954 Current PowerPC GCC accepts a @samp{-msecure-plt} option that
6955 generates code capable of using a newer PLT and GOT layout that has
6956 the security advantage of no executable section ever needing to be
6957 writable and no writable section ever being executable. PowerPC
6958 @command{ld} will generate this layout, including stubs to access the
6959 PLT, if all input files (including startup and static libraries) were
6960 compiled with @samp{-msecure-plt}. @samp{--bss-plt} forces the old
6961 BSS PLT (and GOT layout) which can give slightly better performance.
6963 @kindex --secure-plt
6965 @command{ld} will use the new PLT and GOT layout if it is linking new
6966 @samp{-fpic} or @samp{-fPIC} code, but does not do so automatically
6967 when linking non-PIC code. This option requests the new PLT and GOT
6968 layout. A warning will be given if some object file requires the old
6974 The new secure PLT and GOT are placed differently relative to other
6975 sections compared to older BSS PLT and GOT placement. The location of
6976 @code{.plt} must change because the new secure PLT is an initialized
6977 section while the old PLT is uninitialized. The reason for the
6978 @code{.got} change is more subtle: The new placement allows
6979 @code{.got} to be read-only in applications linked with
6980 @samp{-z relro -z now}. However, this placement means that
6981 @code{.sdata} cannot always be used in shared libraries, because the
6982 PowerPC ABI accesses @code{.sdata} in shared libraries from the GOT
6983 pointer. @samp{--sdata-got} forces the old GOT placement. PowerPC
6984 GCC doesn't use @code{.sdata} in shared libraries, so this option is
6985 really only useful for other compilers that may do so.
6987 @cindex PowerPC stub symbols
6988 @kindex --emit-stub-syms
6989 @item --emit-stub-syms
6990 This option causes @command{ld} to label linker stubs with a local
6991 symbol that encodes the stub type and destination.
6993 @cindex PowerPC TLS optimization
6994 @kindex --no-tls-optimize
6995 @item --no-tls-optimize
6996 PowerPC @command{ld} normally performs some optimization of code
6997 sequences used to access Thread-Local Storage. Use this option to
6998 disable the optimization.
7011 @node PowerPC64 ELF64
7012 @section @command{ld} and PowerPC64 64-bit ELF Support
7014 @cindex PowerPC64 ELF64 options
7016 @cindex PowerPC64 stub grouping
7017 @kindex --stub-group-size
7018 @item --stub-group-size
7019 Long branch stubs, PLT call stubs and TOC adjusting stubs are placed
7020 by @command{ld} in stub sections located between groups of input sections.
7021 @samp{--stub-group-size} specifies the maximum size of a group of input
7022 sections handled by one stub section. Since branch offsets are signed,
7023 a stub section may serve two groups of input sections, one group before
7024 the stub section, and one group after it. However, when using
7025 conditional branches that require stubs, it may be better (for branch
7026 prediction) that stub sections only serve one group of input sections.
7027 A negative value for @samp{N} chooses this scheme, ensuring that
7028 branches to stubs always use a negative offset. Two special values of
7029 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
7030 @command{ld} to automatically size input section groups for the branch types
7031 detected, with the same behaviour regarding stub placement as other
7032 positive or negative values of @samp{N} respectively.
7034 Note that @samp{--stub-group-size} does not split input sections. A
7035 single input section larger than the group size specified will of course
7036 create a larger group (of one section). If input sections are too
7037 large, it may not be possible for a branch to reach its stub.
7039 @cindex PowerPC64 stub symbols
7040 @kindex --emit-stub-syms
7041 @item --emit-stub-syms
7042 This option causes @command{ld} to label linker stubs with a local
7043 symbol that encodes the stub type and destination.
7045 @cindex PowerPC64 dot symbols
7047 @kindex --no-dotsyms
7048 @item --dotsyms, --no-dotsyms
7049 These two options control how @command{ld} interprets version patterns
7050 in a version script. Older PowerPC64 compilers emitted both a
7051 function descriptor symbol with the same name as the function, and a
7052 code entry symbol with the name prefixed by a dot (@samp{.}). To
7053 properly version a function @samp{foo}, the version script thus needs
7054 to control both @samp{foo} and @samp{.foo}. The option
7055 @samp{--dotsyms}, on by default, automatically adds the required
7056 dot-prefixed patterns. Use @samp{--no-dotsyms} to disable this
7059 @cindex PowerPC64 TLS optimization
7060 @kindex --no-tls-optimize
7061 @item --no-tls-optimize
7062 PowerPC64 @command{ld} normally performs some optimization of code
7063 sequences used to access Thread-Local Storage. Use this option to
7064 disable the optimization.
7066 @cindex PowerPC64 OPD optimization
7067 @kindex --no-opd-optimize
7068 @item --no-opd-optimize
7069 PowerPC64 @command{ld} normally removes @code{.opd} section entries
7070 corresponding to deleted link-once functions, or functions removed by
7071 the action of @samp{--gc-sections} or linker script @code{/DISCARD/}.
7072 Use this option to disable @code{.opd} optimization.
7074 @cindex PowerPC64 OPD spacing
7075 @kindex --non-overlapping-opd
7076 @item --non-overlapping-opd
7077 Some PowerPC64 compilers have an option to generate compressed
7078 @code{.opd} entries spaced 16 bytes apart, overlapping the third word,
7079 the static chain pointer (unused in C) with the first word of the next
7080 entry. This option expands such entries to the full 24 bytes.
7082 @cindex PowerPC64 TOC optimization
7083 @kindex --no-toc-optimize
7084 @item --no-toc-optimize
7085 PowerPC64 @command{ld} normally removes unused @code{.toc} section
7086 entries. Such entries are detected by examining relocations that
7087 reference the TOC in code sections. A reloc in a deleted code section
7088 marks a TOC word as unneeded, while a reloc in a kept code section
7089 marks a TOC word as needed. Since the TOC may reference itself, TOC
7090 relocs are also examined. TOC words marked as both needed and
7091 unneeded will of course be kept. TOC words without any referencing
7092 reloc are assumed to be part of a multi-word entry, and are kept or
7093 discarded as per the nearest marked preceding word. This works
7094 reliably for compiler generated code, but may be incorrect if assembly
7095 code is used to insert TOC entries. Use this option to disable the
7098 @cindex PowerPC64 multi-TOC
7099 @kindex --no-multi-toc
7100 @item --no-multi-toc
7101 If given any toc option besides @code{-mcmodel=medium} or
7102 @code{-mcmodel=large}, PowerPC64 GCC generates code for a TOC model
7104 entries are accessed with a 16-bit offset from r2. This limits the
7105 total TOC size to 64K. PowerPC64 @command{ld} extends this limit by
7106 grouping code sections such that each group uses less than 64K for its
7107 TOC entries, then inserts r2 adjusting stubs between inter-group
7108 calls. @command{ld} does not split apart input sections, so cannot
7109 help if a single input file has a @code{.toc} section that exceeds
7110 64K, most likely from linking multiple files with @command{ld -r}.
7111 Use this option to turn off this feature.
7113 @cindex PowerPC64 TOC sorting
7114 @kindex --no-toc-sort
7116 By default, @command{ld} sorts TOC sections so that those whose file
7117 happens to have a section called @code{.init} or @code{.fini} are
7118 placed first, followed by TOC sections referenced by code generated
7119 with PowerPC64 gcc's @code{-mcmodel=small}, and lastly TOC sections
7120 referenced only by code generated with PowerPC64 gcc's
7121 @code{-mcmodel=medium} or @code{-mcmodel=large} options. Doing this
7122 results in better TOC grouping for multi-TOC. Use this option to turn
7125 @cindex PowerPC64 PLT stub alignment
7127 @kindex --no-plt-align
7129 @itemx --no-plt-align
7130 Use these options to control whether individual PLT call stubs are
7131 padded so that they don't cross a 32-byte boundary, or to the
7132 specified power of two boundary when using @code{--plt-align=}. Note
7133 that this isn't alignment in the usual sense. By default PLT call
7134 stubs are packed tightly.
7136 @cindex PowerPC64 PLT call stub static chain
7137 @kindex --plt-static-chain
7138 @kindex --no-plt-static-chain
7139 @item --plt-static-chain
7140 @itemx --no-plt-static-chain
7141 Use these options to control whether PLT call stubs load the static
7142 chain pointer (r11). @code{ld} defaults to not loading the static
7143 chain since there is never any need to do so on a PLT call.
7145 @cindex PowerPC64 PLT call stub thread safety
7146 @kindex --plt-thread-safe
7147 @kindex --no-plt-thread-safe
7148 @item --plt-thread-safe
7149 @itemx --no-thread-safe
7150 With power7's weakly ordered memory model, it is possible when using
7151 lazy binding for ld.so to update a plt entry in one thread and have
7152 another thread see the individual plt entry words update in the wrong
7153 order, despite ld.so carefully writing in the correct order and using
7154 memory write barriers. To avoid this we need some sort of read
7155 barrier in the call stub, or use LD_BIND_NOW=1. By default, @code{ld}
7156 looks for calls to commonly used functions that create threads, and if
7157 seen, adds the necessary barriers. Use these options to change the
7172 @section @command{ld} and SPU ELF Support
7174 @cindex SPU ELF options
7180 This option marks an executable as a PIC plugin module.
7182 @cindex SPU overlays
7183 @kindex --no-overlays
7185 Normally, @command{ld} recognizes calls to functions within overlay
7186 regions, and redirects such calls to an overlay manager via a stub.
7187 @command{ld} also provides a built-in overlay manager. This option
7188 turns off all this special overlay handling.
7190 @cindex SPU overlay stub symbols
7191 @kindex --emit-stub-syms
7192 @item --emit-stub-syms
7193 This option causes @command{ld} to label overlay stubs with a local
7194 symbol that encodes the stub type and destination.
7196 @cindex SPU extra overlay stubs
7197 @kindex --extra-overlay-stubs
7198 @item --extra-overlay-stubs
7199 This option causes @command{ld} to add overlay call stubs on all
7200 function calls out of overlay regions. Normally stubs are not added
7201 on calls to non-overlay regions.
7203 @cindex SPU local store size
7204 @kindex --local-store=lo:hi
7205 @item --local-store=lo:hi
7206 @command{ld} usually checks that a final executable for SPU fits in
7207 the address range 0 to 256k. This option may be used to change the
7208 range. Disable the check entirely with @option{--local-store=0:0}.
7211 @kindex --stack-analysis
7212 @item --stack-analysis
7213 SPU local store space is limited. Over-allocation of stack space
7214 unnecessarily limits space available for code and data, while
7215 under-allocation results in runtime failures. If given this option,
7216 @command{ld} will provide an estimate of maximum stack usage.
7217 @command{ld} does this by examining symbols in code sections to
7218 determine the extents of functions, and looking at function prologues
7219 for stack adjusting instructions. A call-graph is created by looking
7220 for relocations on branch instructions. The graph is then searched
7221 for the maximum stack usage path. Note that this analysis does not
7222 find calls made via function pointers, and does not handle recursion
7223 and other cycles in the call graph. Stack usage may be
7224 under-estimated if your code makes such calls. Also, stack usage for
7225 dynamic allocation, e.g. alloca, will not be detected. If a link map
7226 is requested, detailed information about each function's stack usage
7227 and calls will be given.
7230 @kindex --emit-stack-syms
7231 @item --emit-stack-syms
7232 This option, if given along with @option{--stack-analysis} will result
7233 in @command{ld} emitting stack sizing symbols for each function.
7234 These take the form @code{__stack_<function_name>} for global
7235 functions, and @code{__stack_<number>_<function_name>} for static
7236 functions. @code{<number>} is the section id in hex. The value of
7237 such symbols is the stack requirement for the corresponding function.
7238 The symbol size will be zero, type @code{STT_NOTYPE}, binding
7239 @code{STB_LOCAL}, and section @code{SHN_ABS}.
7253 @section @command{ld}'s Support for Various TI COFF Versions
7254 @cindex TI COFF versions
7255 @kindex --format=@var{version}
7256 The @samp{--format} switch allows selection of one of the various
7257 TI COFF versions. The latest of this writing is 2; versions 0 and 1 are
7258 also supported. The TI COFF versions also vary in header byte-order
7259 format; @command{ld} will read any version or byte order, but the output
7260 header format depends on the default specified by the specific target.
7273 @section @command{ld} and WIN32 (cygwin/mingw)
7275 This section describes some of the win32 specific @command{ld} issues.
7276 See @ref{Options,,Command Line Options} for detailed description of the
7277 command line options mentioned here.
7280 @cindex import libraries
7281 @item import libraries
7282 The standard Windows linker creates and uses so-called import
7283 libraries, which contains information for linking to dll's. They are
7284 regular static archives and are handled as any other static
7285 archive. The cygwin and mingw ports of @command{ld} have specific
7286 support for creating such libraries provided with the
7287 @samp{--out-implib} command line option.
7289 @item exporting DLL symbols
7290 @cindex exporting DLL symbols
7291 The cygwin/mingw @command{ld} has several ways to export symbols for dll's.
7294 @item using auto-export functionality
7295 @cindex using auto-export functionality
7296 By default @command{ld} exports symbols with the auto-export functionality,
7297 which is controlled by the following command line options:
7300 @item --export-all-symbols [This is the default]
7301 @item --exclude-symbols
7302 @item --exclude-libs
7303 @item --exclude-modules-for-implib
7304 @item --version-script
7307 When auto-export is in operation, @command{ld} will export all the non-local
7308 (global and common) symbols it finds in a DLL, with the exception of a few
7309 symbols known to belong to the system's runtime and libraries. As it will
7310 often not be desirable to export all of a DLL's symbols, which may include
7311 private functions that are not part of any public interface, the command-line
7312 options listed above may be used to filter symbols out from the list for
7313 exporting. The @samp{--output-def} option can be used in order to see the
7314 final list of exported symbols with all exclusions taken into effect.
7316 If @samp{--export-all-symbols} is not given explicitly on the
7317 command line, then the default auto-export behavior will be @emph{disabled}
7318 if either of the following are true:
7321 @item A DEF file is used.
7322 @item Any symbol in any object file was marked with the __declspec(dllexport) attribute.
7325 @item using a DEF file
7326 @cindex using a DEF file
7327 Another way of exporting symbols is using a DEF file. A DEF file is
7328 an ASCII file containing definitions of symbols which should be
7329 exported when a dll is created. Usually it is named @samp{<dll
7330 name>.def} and is added as any other object file to the linker's
7331 command line. The file's name must end in @samp{.def} or @samp{.DEF}.
7334 gcc -o <output> <objectfiles> <dll name>.def
7337 Using a DEF file turns off the normal auto-export behavior, unless the
7338 @samp{--export-all-symbols} option is also used.
7340 Here is an example of a DEF file for a shared library called @samp{xyz.dll}:
7343 LIBRARY "xyz.dll" BASE=0x20000000
7349 another_foo = abc.dll.afoo
7355 This example defines a DLL with a non-default base address and seven
7356 symbols in the export table. The third exported symbol @code{_bar} is an
7357 alias for the second. The fourth symbol, @code{another_foo} is resolved
7358 by "forwarding" to another module and treating it as an alias for
7359 @code{afoo} exported from the DLL @samp{abc.dll}. The final symbol
7360 @code{var1} is declared to be a data object. The @samp{doo} symbol in
7361 export library is an alias of @samp{foo}, which gets the string name
7362 in export table @samp{foo2}. The @samp{eoo} symbol is an data export
7363 symbol, which gets in export table the name @samp{var1}.
7365 The optional @code{LIBRARY <name>} command indicates the @emph{internal}
7366 name of the output DLL. If @samp{<name>} does not include a suffix,
7367 the default library suffix, @samp{.DLL} is appended.
7369 When the .DEF file is used to build an application, rather than a
7370 library, the @code{NAME <name>} command should be used instead of
7371 @code{LIBRARY}. If @samp{<name>} does not include a suffix, the default
7372 executable suffix, @samp{.EXE} is appended.
7374 With either @code{LIBRARY <name>} or @code{NAME <name>} the optional
7375 specification @code{BASE = <number>} may be used to specify a
7376 non-default base address for the image.
7378 If neither @code{LIBRARY <name>} nor @code{NAME <name>} is specified,
7379 or they specify an empty string, the internal name is the same as the
7380 filename specified on the command line.
7382 The complete specification of an export symbol is:
7386 ( ( ( <name1> [ = <name2> ] )
7387 | ( <name1> = <module-name> . <external-name>))
7388 [ @@ <integer> ] [NONAME] [DATA] [CONSTANT] [PRIVATE] [== <name3>] ) *
7391 Declares @samp{<name1>} as an exported symbol from the DLL, or declares
7392 @samp{<name1>} as an exported alias for @samp{<name2>}; or declares
7393 @samp{<name1>} as a "forward" alias for the symbol
7394 @samp{<external-name>} in the DLL @samp{<module-name>}.
7395 Optionally, the symbol may be exported by the specified ordinal
7396 @samp{<integer>} alias. The optional @samp{<name3>} is the to be used
7397 string in import/export table for the symbol.
7399 The optional keywords that follow the declaration indicate:
7401 @code{NONAME}: Do not put the symbol name in the DLL's export table. It
7402 will still be exported by its ordinal alias (either the value specified
7403 by the .def specification or, otherwise, the value assigned by the
7404 linker). The symbol name, however, does remain visible in the import
7405 library (if any), unless @code{PRIVATE} is also specified.
7407 @code{DATA}: The symbol is a variable or object, rather than a function.
7408 The import lib will export only an indirect reference to @code{foo} as
7409 the symbol @code{_imp__foo} (ie, @code{foo} must be resolved as
7412 @code{CONSTANT}: Like @code{DATA}, but put the undecorated @code{foo} as
7413 well as @code{_imp__foo} into the import library. Both refer to the
7414 read-only import address table's pointer to the variable, not to the
7415 variable itself. This can be dangerous. If the user code fails to add
7416 the @code{dllimport} attribute and also fails to explicitly add the
7417 extra indirection that the use of the attribute enforces, the
7418 application will behave unexpectedly.
7420 @code{PRIVATE}: Put the symbol in the DLL's export table, but do not put
7421 it into the static import library used to resolve imports at link time. The
7422 symbol can still be imported using the @code{LoadLibrary/GetProcAddress}
7423 API at runtime or by by using the GNU ld extension of linking directly to
7424 the DLL without an import library.
7426 See ld/deffilep.y in the binutils sources for the full specification of
7427 other DEF file statements
7429 @cindex creating a DEF file
7430 While linking a shared dll, @command{ld} is able to create a DEF file
7431 with the @samp{--output-def <file>} command line option.
7433 @item Using decorations
7434 @cindex Using decorations
7435 Another way of marking symbols for export is to modify the source code
7436 itself, so that when building the DLL each symbol to be exported is
7440 __declspec(dllexport) int a_variable
7441 __declspec(dllexport) void a_function(int with_args)
7444 All such symbols will be exported from the DLL. If, however,
7445 any of the object files in the DLL contain symbols decorated in
7446 this way, then the normal auto-export behavior is disabled, unless
7447 the @samp{--export-all-symbols} option is also used.
7449 Note that object files that wish to access these symbols must @emph{not}
7450 decorate them with dllexport. Instead, they should use dllimport,
7454 __declspec(dllimport) int a_variable
7455 __declspec(dllimport) void a_function(int with_args)
7458 This complicates the structure of library header files, because
7459 when included by the library itself the header must declare the
7460 variables and functions as dllexport, but when included by client
7461 code the header must declare them as dllimport. There are a number
7462 of idioms that are typically used to do this; often client code can
7463 omit the __declspec() declaration completely. See
7464 @samp{--enable-auto-import} and @samp{automatic data imports} for more
7468 @cindex automatic data imports
7469 @item automatic data imports
7470 The standard Windows dll format supports data imports from dlls only
7471 by adding special decorations (dllimport/dllexport), which let the
7472 compiler produce specific assembler instructions to deal with this
7473 issue. This increases the effort necessary to port existing Un*x
7474 code to these platforms, especially for large
7475 c++ libraries and applications. The auto-import feature, which was
7476 initially provided by Paul Sokolovsky, allows one to omit the
7477 decorations to achieve a behavior that conforms to that on POSIX/Un*x
7478 platforms. This feature is enabled with the @samp{--enable-auto-import}
7479 command-line option, although it is enabled by default on cygwin/mingw.
7480 The @samp{--enable-auto-import} option itself now serves mainly to
7481 suppress any warnings that are ordinarily emitted when linked objects
7482 trigger the feature's use.
7484 auto-import of variables does not always work flawlessly without
7485 additional assistance. Sometimes, you will see this message
7487 "variable '<var>' can't be auto-imported. Please read the
7488 documentation for ld's @code{--enable-auto-import} for details."
7490 The @samp{--enable-auto-import} documentation explains why this error
7491 occurs, and several methods that can be used to overcome this difficulty.
7492 One of these methods is the @emph{runtime pseudo-relocs} feature, described
7495 @cindex runtime pseudo-relocation
7496 For complex variables imported from DLLs (such as structs or classes),
7497 object files typically contain a base address for the variable and an
7498 offset (@emph{addend}) within the variable--to specify a particular
7499 field or public member, for instance. Unfortunately, the runtime loader used
7500 in win32 environments is incapable of fixing these references at runtime
7501 without the additional information supplied by dllimport/dllexport decorations.
7502 The standard auto-import feature described above is unable to resolve these
7505 The @samp{--enable-runtime-pseudo-relocs} switch allows these references to
7506 be resolved without error, while leaving the task of adjusting the references
7507 themselves (with their non-zero addends) to specialized code provided by the
7508 runtime environment. Recent versions of the cygwin and mingw environments and
7509 compilers provide this runtime support; older versions do not. However, the
7510 support is only necessary on the developer's platform; the compiled result will
7511 run without error on an older system.
7513 @samp{--enable-runtime-pseudo-relocs} is not the default; it must be explicitly
7516 @cindex direct linking to a dll
7517 @item direct linking to a dll
7518 The cygwin/mingw ports of @command{ld} support the direct linking,
7519 including data symbols, to a dll without the usage of any import
7520 libraries. This is much faster and uses much less memory than does the
7521 traditional import library method, especially when linking large
7522 libraries or applications. When @command{ld} creates an import lib, each
7523 function or variable exported from the dll is stored in its own bfd, even
7524 though a single bfd could contain many exports. The overhead involved in
7525 storing, loading, and processing so many bfd's is quite large, and explains the
7526 tremendous time, memory, and storage needed to link against particularly
7527 large or complex libraries when using import libs.
7529 Linking directly to a dll uses no extra command-line switches other than
7530 @samp{-L} and @samp{-l}, because @command{ld} already searches for a number
7531 of names to match each library. All that is needed from the developer's
7532 perspective is an understanding of this search, in order to force ld to
7533 select the dll instead of an import library.
7536 For instance, when ld is called with the argument @samp{-lxxx} it will attempt
7537 to find, in the first directory of its search path,
7549 before moving on to the next directory in the search path.
7551 (*) Actually, this is not @samp{cygxxx.dll} but in fact is @samp{<prefix>xxx.dll},
7552 where @samp{<prefix>} is set by the @command{ld} option
7553 @samp{--dll-search-prefix=<prefix>}. In the case of cygwin, the standard gcc spec
7554 file includes @samp{--dll-search-prefix=cyg}, so in effect we actually search for
7557 Other win32-based unix environments, such as mingw or pw32, may use other
7558 @samp{<prefix>}es, although at present only cygwin makes use of this feature. It
7559 was originally intended to help avoid name conflicts among dll's built for the
7560 various win32/un*x environments, so that (for example) two versions of a zlib dll
7561 could coexist on the same machine.
7563 The generic cygwin/mingw path layout uses a @samp{bin} directory for
7564 applications and dll's and a @samp{lib} directory for the import
7565 libraries (using cygwin nomenclature):
7571 libxxx.dll.a (in case of dll's)
7572 libxxx.a (in case of static archive)
7575 Linking directly to a dll without using the import library can be
7578 1. Use the dll directly by adding the @samp{bin} path to the link line
7580 gcc -Wl,-verbose -o a.exe -L../bin/ -lxxx
7583 However, as the dll's often have version numbers appended to their names
7584 (@samp{cygncurses-5.dll}) this will often fail, unless one specifies
7585 @samp{-L../bin -lncurses-5} to include the version. Import libs are generally
7586 not versioned, and do not have this difficulty.
7588 2. Create a symbolic link from the dll to a file in the @samp{lib}
7589 directory according to the above mentioned search pattern. This
7590 should be used to avoid unwanted changes in the tools needed for
7594 ln -s bin/cygxxx.dll lib/[cyg|lib|]xxx.dll[.a]
7597 Then you can link without any make environment changes.
7600 gcc -Wl,-verbose -o a.exe -L../lib/ -lxxx
7603 This technique also avoids the version number problems, because the following is
7610 libxxx.dll.a -> ../bin/cygxxx-5.dll
7613 Linking directly to a dll without using an import lib will work
7614 even when auto-import features are exercised, and even when
7615 @samp{--enable-runtime-pseudo-relocs} is used.
7617 Given the improvements in speed and memory usage, one might justifiably
7618 wonder why import libraries are used at all. There are three reasons:
7620 1. Until recently, the link-directly-to-dll functionality did @emph{not}
7621 work with auto-imported data.
7623 2. Sometimes it is necessary to include pure static objects within the
7624 import library (which otherwise contains only bfd's for indirection
7625 symbols that point to the exports of a dll). Again, the import lib
7626 for the cygwin kernel makes use of this ability, and it is not
7627 possible to do this without an import lib.
7629 3. Symbol aliases can only be resolved using an import lib. This is
7630 critical when linking against OS-supplied dll's (eg, the win32 API)
7631 in which symbols are usually exported as undecorated aliases of their
7632 stdcall-decorated assembly names.
7634 So, import libs are not going away. But the ability to replace
7635 true import libs with a simple symbolic link to (or a copy of)
7636 a dll, in many cases, is a useful addition to the suite of tools
7637 binutils makes available to the win32 developer. Given the
7638 massive improvements in memory requirements during linking, storage
7639 requirements, and linking speed, we expect that many developers
7640 will soon begin to use this feature whenever possible.
7642 @item symbol aliasing
7644 @item adding additional names
7645 Sometimes, it is useful to export symbols with additional names.
7646 A symbol @samp{foo} will be exported as @samp{foo}, but it can also be
7647 exported as @samp{_foo} by using special directives in the DEF file
7648 when creating the dll. This will affect also the optional created
7649 import library. Consider the following DEF file:
7652 LIBRARY "xyz.dll" BASE=0x61000000
7659 The line @samp{_foo = foo} maps the symbol @samp{foo} to @samp{_foo}.
7661 Another method for creating a symbol alias is to create it in the
7662 source code using the "weak" attribute:
7665 void foo () @{ /* Do something. */; @}
7666 void _foo () __attribute__ ((weak, alias ("foo")));
7669 See the gcc manual for more information about attributes and weak
7672 @item renaming symbols
7673 Sometimes it is useful to rename exports. For instance, the cygwin
7674 kernel does this regularly. A symbol @samp{_foo} can be exported as
7675 @samp{foo} but not as @samp{_foo} by using special directives in the
7676 DEF file. (This will also affect the import library, if it is
7677 created). In the following example:
7680 LIBRARY "xyz.dll" BASE=0x61000000
7686 The line @samp{_foo = foo} maps the exported symbol @samp{foo} to
7690 Note: using a DEF file disables the default auto-export behavior,
7691 unless the @samp{--export-all-symbols} command line option is used.
7692 If, however, you are trying to rename symbols, then you should list
7693 @emph{all} desired exports in the DEF file, including the symbols
7694 that are not being renamed, and do @emph{not} use the
7695 @samp{--export-all-symbols} option. If you list only the
7696 renamed symbols in the DEF file, and use @samp{--export-all-symbols}
7697 to handle the other symbols, then the both the new names @emph{and}
7698 the original names for the renamed symbols will be exported.
7699 In effect, you'd be aliasing those symbols, not renaming them,
7700 which is probably not what you wanted.
7702 @cindex weak externals
7703 @item weak externals
7704 The Windows object format, PE, specifies a form of weak symbols called
7705 weak externals. When a weak symbol is linked and the symbol is not
7706 defined, the weak symbol becomes an alias for some other symbol. There
7707 are three variants of weak externals:
7709 @item Definition is searched for in objects and libraries, historically
7710 called lazy externals.
7711 @item Definition is searched for only in other objects, not in libraries.
7712 This form is not presently implemented.
7713 @item No search; the symbol is an alias. This form is not presently
7716 As a GNU extension, weak symbols that do not specify an alternate symbol
7717 are supported. If the symbol is undefined when linking, the symbol
7718 uses a default value.
7720 @cindex aligned common symbols
7721 @item aligned common symbols
7722 As a GNU extension to the PE file format, it is possible to specify the
7723 desired alignment for a common symbol. This information is conveyed from
7724 the assembler or compiler to the linker by means of GNU-specific commands
7725 carried in the object file's @samp{.drectve} section, which are recognized
7726 by @command{ld} and respected when laying out the common symbols. Native
7727 tools will be able to process object files employing this GNU extension,
7728 but will fail to respect the alignment instructions, and may issue noisy
7729 warnings about unknown linker directives.
7744 @section @code{ld} and Xtensa Processors
7746 @cindex Xtensa processors
7747 The default @command{ld} behavior for Xtensa processors is to interpret
7748 @code{SECTIONS} commands so that lists of explicitly named sections in a
7749 specification with a wildcard file will be interleaved when necessary to
7750 keep literal pools within the range of PC-relative load offsets. For
7751 example, with the command:
7763 @command{ld} may interleave some of the @code{.literal}
7764 and @code{.text} sections from different object files to ensure that the
7765 literal pools are within the range of PC-relative load offsets. A valid
7766 interleaving might place the @code{.literal} sections from an initial
7767 group of files followed by the @code{.text} sections of that group of
7768 files. Then, the @code{.literal} sections from the rest of the files
7769 and the @code{.text} sections from the rest of the files would follow.
7771 @cindex @option{--relax} on Xtensa
7772 @cindex relaxing on Xtensa
7773 Relaxation is enabled by default for the Xtensa version of @command{ld} and
7774 provides two important link-time optimizations. The first optimization
7775 is to combine identical literal values to reduce code size. A redundant
7776 literal will be removed and all the @code{L32R} instructions that use it
7777 will be changed to reference an identical literal, as long as the
7778 location of the replacement literal is within the offset range of all
7779 the @code{L32R} instructions. The second optimization is to remove
7780 unnecessary overhead from assembler-generated ``longcall'' sequences of
7781 @code{L32R}/@code{CALLX@var{n}} when the target functions are within
7782 range of direct @code{CALL@var{n}} instructions.
7784 For each of these cases where an indirect call sequence can be optimized
7785 to a direct call, the linker will change the @code{CALLX@var{n}}
7786 instruction to a @code{CALL@var{n}} instruction, remove the @code{L32R}
7787 instruction, and remove the literal referenced by the @code{L32R}
7788 instruction if it is not used for anything else. Removing the
7789 @code{L32R} instruction always reduces code size but can potentially
7790 hurt performance by changing the alignment of subsequent branch targets.
7791 By default, the linker will always preserve alignments, either by
7792 switching some instructions between 24-bit encodings and the equivalent
7793 density instructions or by inserting a no-op in place of the @code{L32R}
7794 instruction that was removed. If code size is more important than
7795 performance, the @option{--size-opt} option can be used to prevent the
7796 linker from widening density instructions or inserting no-ops, except in
7797 a few cases where no-ops are required for correctness.
7799 The following Xtensa-specific command-line options can be used to
7802 @cindex Xtensa options
7805 When optimizing indirect calls to direct calls, optimize for code size
7806 more than performance. With this option, the linker will not insert
7807 no-ops or widen density instructions to preserve branch target
7808 alignment. There may still be some cases where no-ops are required to
7809 preserve the correctness of the code.
7817 @ifclear SingleFormat
7822 @cindex object file management
7823 @cindex object formats available
7825 The linker accesses object and archive files using the BFD libraries.
7826 These libraries allow the linker to use the same routines to operate on
7827 object files whatever the object file format. A different object file
7828 format can be supported simply by creating a new BFD back end and adding
7829 it to the library. To conserve runtime memory, however, the linker and
7830 associated tools are usually configured to support only a subset of the
7831 object file formats available. You can use @code{objdump -i}
7832 (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to
7833 list all the formats available for your configuration.
7835 @cindex BFD requirements
7836 @cindex requirements for BFD
7837 As with most implementations, BFD is a compromise between
7838 several conflicting requirements. The major factor influencing
7839 BFD design was efficiency: any time used converting between
7840 formats is time which would not have been spent had BFD not
7841 been involved. This is partly offset by abstraction payback; since
7842 BFD simplifies applications and back ends, more time and care
7843 may be spent optimizing algorithms for a greater speed.
7845 One minor artifact of the BFD solution which you should bear in
7846 mind is the potential for information loss. There are two places where
7847 useful information can be lost using the BFD mechanism: during
7848 conversion and during output. @xref{BFD information loss}.
7851 * BFD outline:: How it works: an outline of BFD
7855 @section How It Works: An Outline of BFD
7856 @cindex opening object files
7857 @include bfdsumm.texi
7860 @node Reporting Bugs
7861 @chapter Reporting Bugs
7862 @cindex bugs in @command{ld}
7863 @cindex reporting bugs in @command{ld}
7865 Your bug reports play an essential role in making @command{ld} reliable.
7867 Reporting a bug may help you by bringing a solution to your problem, or
7868 it may not. But in any case the principal function of a bug report is
7869 to help the entire community by making the next version of @command{ld}
7870 work better. Bug reports are your contribution to the maintenance of
7873 In order for a bug report to serve its purpose, you must include the
7874 information that enables us to fix the bug.
7877 * Bug Criteria:: Have you found a bug?
7878 * Bug Reporting:: How to report bugs
7882 @section Have You Found a Bug?
7883 @cindex bug criteria
7885 If you are not sure whether you have found a bug, here are some guidelines:
7888 @cindex fatal signal
7889 @cindex linker crash
7890 @cindex crash of linker
7892 If the linker gets a fatal signal, for any input whatever, that is a
7893 @command{ld} bug. Reliable linkers never crash.
7895 @cindex error on valid input
7897 If @command{ld} produces an error message for valid input, that is a bug.
7899 @cindex invalid input
7901 If @command{ld} does not produce an error message for invalid input, that
7902 may be a bug. In the general case, the linker can not verify that
7903 object files are correct.
7906 If you are an experienced user of linkers, your suggestions for
7907 improvement of @command{ld} are welcome in any case.
7911 @section How to Report Bugs
7913 @cindex @command{ld} bugs, reporting
7915 A number of companies and individuals offer support for @sc{gnu}
7916 products. If you obtained @command{ld} from a support organization, we
7917 recommend you contact that organization first.
7919 You can find contact information for many support companies and
7920 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7924 Otherwise, send bug reports for @command{ld} to
7928 The fundamental principle of reporting bugs usefully is this:
7929 @strong{report all the facts}. If you are not sure whether to state a
7930 fact or leave it out, state it!
7932 Often people omit facts because they think they know what causes the
7933 problem and assume that some details do not matter. Thus, you might
7934 assume that the name of a symbol you use in an example does not
7935 matter. Well, probably it does not, but one cannot be sure. Perhaps
7936 the bug is a stray memory reference which happens to fetch from the
7937 location where that name is stored in memory; perhaps, if the name
7938 were different, the contents of that location would fool the linker
7939 into doing the right thing despite the bug. Play it safe and give a
7940 specific, complete example. That is the easiest thing for you to do,
7941 and the most helpful.
7943 Keep in mind that the purpose of a bug report is to enable us to fix
7944 the bug if it is new to us. Therefore, always write your bug reports
7945 on the assumption that the bug has not been reported previously.
7947 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7948 bell?'' This cannot help us fix a bug, so it is basically useless. We
7949 respond by asking for enough details to enable us to investigate.
7950 You might as well expedite matters by sending them to begin with.
7952 To enable us to fix the bug, you should include all these things:
7956 The version of @command{ld}. @command{ld} announces it if you start it with
7957 the @samp{--version} argument.
7959 Without this, we will not know whether there is any point in looking for
7960 the bug in the current version of @command{ld}.
7963 Any patches you may have applied to the @command{ld} source, including any
7964 patches made to the @code{BFD} library.
7967 The type of machine you are using, and the operating system name and
7971 What compiler (and its version) was used to compile @command{ld}---e.g.
7975 The command arguments you gave the linker to link your example and
7976 observe the bug. To guarantee you will not omit something important,
7977 list them all. A copy of the Makefile (or the output from make) is
7980 If we were to try to guess the arguments, we would probably guess wrong
7981 and then we might not encounter the bug.
7984 A complete input file, or set of input files, that will reproduce the
7985 bug. It is generally most helpful to send the actual object files
7986 provided that they are reasonably small. Say no more than 10K. For
7987 bigger files you can either make them available by FTP or HTTP or else
7988 state that you are willing to send the object file(s) to whomever
7989 requests them. (Note - your email will be going to a mailing list, so
7990 we do not want to clog it up with large attachments). But small
7991 attachments are best.
7993 If the source files were assembled using @code{gas} or compiled using
7994 @code{gcc}, then it may be OK to send the source files rather than the
7995 object files. In this case, be sure to say exactly what version of
7996 @code{gas} or @code{gcc} was used to produce the object files. Also say
7997 how @code{gas} or @code{gcc} were configured.
8000 A description of what behavior you observe that you believe is
8001 incorrect. For example, ``It gets a fatal signal.''
8003 Of course, if the bug is that @command{ld} gets a fatal signal, then we
8004 will certainly notice it. But if the bug is incorrect output, we might
8005 not notice unless it is glaringly wrong. You might as well not give us
8006 a chance to make a mistake.
8008 Even if the problem you experience is a fatal signal, you should still
8009 say so explicitly. Suppose something strange is going on, such as, your
8010 copy of @command{ld} is out of sync, or you have encountered a bug in the
8011 C library on your system. (This has happened!) Your copy might crash
8012 and ours would not. If you told us to expect a crash, then when ours
8013 fails to crash, we would know that the bug was not happening for us. If
8014 you had not told us to expect a crash, then we would not be able to draw
8015 any conclusion from our observations.
8018 If you wish to suggest changes to the @command{ld} source, send us context
8019 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or
8020 @samp{-p} option. Always send diffs from the old file to the new file.
8021 If you even discuss something in the @command{ld} source, refer to it by
8022 context, not by line number.
8024 The line numbers in our development sources will not match those in your
8025 sources. Your line numbers would convey no useful information to us.
8028 Here are some things that are not necessary:
8032 A description of the envelope of the bug.
8034 Often people who encounter a bug spend a lot of time investigating
8035 which changes to the input file will make the bug go away and which
8036 changes will not affect it.
8038 This is often time consuming and not very useful, because the way we
8039 will find the bug is by running a single example under the debugger
8040 with breakpoints, not by pure deduction from a series of examples.
8041 We recommend that you save your time for something else.
8043 Of course, if you can find a simpler example to report @emph{instead}
8044 of the original one, that is a convenience for us. Errors in the
8045 output will be easier to spot, running under the debugger will take
8046 less time, and so on.
8048 However, simplification is not vital; if you do not want to do this,
8049 report the bug anyway and send us the entire test case you used.
8052 A patch for the bug.
8054 A patch for the bug does help us if it is a good one. But do not omit
8055 the necessary information, such as the test case, on the assumption that
8056 a patch is all we need. We might see problems with your patch and decide
8057 to fix the problem another way, or we might not understand it at all.
8059 Sometimes with a program as complicated as @command{ld} it is very hard to
8060 construct an example that will make the program follow a certain path
8061 through the code. If you do not send us the example, we will not be
8062 able to construct one, so we will not be able to verify that the bug is
8065 And if we cannot understand what bug you are trying to fix, or why your
8066 patch should be an improvement, we will not install it. A test case will
8067 help us to understand.
8070 A guess about what the bug is or what it depends on.
8072 Such guesses are usually wrong. Even we cannot guess right about such
8073 things without first using the debugger to find the facts.
8077 @appendix MRI Compatible Script Files
8078 @cindex MRI compatibility
8079 To aid users making the transition to @sc{gnu} @command{ld} from the MRI
8080 linker, @command{ld} can use MRI compatible linker scripts as an
8081 alternative to the more general-purpose linker scripting language
8082 described in @ref{Scripts}. MRI compatible linker scripts have a much
8083 simpler command set than the scripting language otherwise used with
8084 @command{ld}. @sc{gnu} @command{ld} supports the most commonly used MRI
8085 linker commands; these commands are described here.
8087 In general, MRI scripts aren't of much use with the @code{a.out} object
8088 file format, since it only has three sections and MRI scripts lack some
8089 features to make use of them.
8091 You can specify a file containing an MRI-compatible script using the
8092 @samp{-c} command-line option.
8094 Each command in an MRI-compatible script occupies its own line; each
8095 command line starts with the keyword that identifies the command (though
8096 blank lines are also allowed for punctuation). If a line of an
8097 MRI-compatible script begins with an unrecognized keyword, @command{ld}
8098 issues a warning message, but continues processing the script.
8100 Lines beginning with @samp{*} are comments.
8102 You can write these commands using all upper-case letters, or all
8103 lower case; for example, @samp{chip} is the same as @samp{CHIP}.
8104 The following list shows only the upper-case form of each command.
8107 @cindex @code{ABSOLUTE} (MRI)
8108 @item ABSOLUTE @var{secname}
8109 @itemx ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname}
8110 Normally, @command{ld} includes in the output file all sections from all
8111 the input files. However, in an MRI-compatible script, you can use the
8112 @code{ABSOLUTE} command to restrict the sections that will be present in
8113 your output program. If the @code{ABSOLUTE} command is used at all in a
8114 script, then only the sections named explicitly in @code{ABSOLUTE}
8115 commands will appear in the linker output. You can still use other
8116 input sections (whatever you select on the command line, or using
8117 @code{LOAD}) to resolve addresses in the output file.
8119 @cindex @code{ALIAS} (MRI)
8120 @item ALIAS @var{out-secname}, @var{in-secname}
8121 Use this command to place the data from input section @var{in-secname}
8122 in a section called @var{out-secname} in the linker output file.
8124 @var{in-secname} may be an integer.
8126 @cindex @code{ALIGN} (MRI)
8127 @item ALIGN @var{secname} = @var{expression}
8128 Align the section called @var{secname} to @var{expression}. The
8129 @var{expression} should be a power of two.
8131 @cindex @code{BASE} (MRI)
8132 @item BASE @var{expression}
8133 Use the value of @var{expression} as the lowest address (other than
8134 absolute addresses) in the output file.
8136 @cindex @code{CHIP} (MRI)
8137 @item CHIP @var{expression}
8138 @itemx CHIP @var{expression}, @var{expression}
8139 This command does nothing; it is accepted only for compatibility.
8141 @cindex @code{END} (MRI)
8143 This command does nothing whatever; it's only accepted for compatibility.
8145 @cindex @code{FORMAT} (MRI)
8146 @item FORMAT @var{output-format}
8147 Similar to the @code{OUTPUT_FORMAT} command in the more general linker
8148 language, but restricted to one of these output formats:
8152 S-records, if @var{output-format} is @samp{S}
8155 IEEE, if @var{output-format} is @samp{IEEE}
8158 COFF (the @samp{coff-m68k} variant in BFD), if @var{output-format} is
8162 @cindex @code{LIST} (MRI)
8163 @item LIST @var{anything}@dots{}
8164 Print (to the standard output file) a link map, as produced by the
8165 @command{ld} command-line option @samp{-M}.
8167 The keyword @code{LIST} may be followed by anything on the
8168 same line, with no change in its effect.
8170 @cindex @code{LOAD} (MRI)
8171 @item LOAD @var{filename}
8172 @itemx LOAD @var{filename}, @var{filename}, @dots{} @var{filename}
8173 Include one or more object file @var{filename} in the link; this has the
8174 same effect as specifying @var{filename} directly on the @command{ld}
8177 @cindex @code{NAME} (MRI)
8178 @item NAME @var{output-name}
8179 @var{output-name} is the name for the program produced by @command{ld}; the
8180 MRI-compatible command @code{NAME} is equivalent to the command-line
8181 option @samp{-o} or the general script language command @code{OUTPUT}.
8183 @cindex @code{ORDER} (MRI)
8184 @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname}
8185 @itemx ORDER @var{secname} @var{secname} @var{secname}
8186 Normally, @command{ld} orders the sections in its output file in the
8187 order in which they first appear in the input files. In an MRI-compatible
8188 script, you can override this ordering with the @code{ORDER} command. The
8189 sections you list with @code{ORDER} will appear first in your output
8190 file, in the order specified.
8192 @cindex @code{PUBLIC} (MRI)
8193 @item PUBLIC @var{name}=@var{expression}
8194 @itemx PUBLIC @var{name},@var{expression}
8195 @itemx PUBLIC @var{name} @var{expression}
8196 Supply a value (@var{expression}) for external symbol
8197 @var{name} used in the linker input files.
8199 @cindex @code{SECT} (MRI)
8200 @item SECT @var{secname}, @var{expression}
8201 @itemx SECT @var{secname}=@var{expression}
8202 @itemx SECT @var{secname} @var{expression}
8203 You can use any of these three forms of the @code{SECT} command to
8204 specify the start address (@var{expression}) for section @var{secname}.
8205 If you have more than one @code{SECT} statement for the same
8206 @var{secname}, only the @emph{first} sets the start address.
8209 @node GNU Free Documentation License
8210 @appendix GNU Free Documentation License
8214 @unnumbered LD Index
8219 % I think something like @@colophon should be in texinfo. In the
8221 \long\def\colophon{\hbox to0pt{}\vfill
8222 \centerline{The body of this manual is set in}
8223 \centerline{\fontname\tenrm,}
8224 \centerline{with headings in {\bf\fontname\tenbf}}
8225 \centerline{and examples in {\tt\fontname\tentt}.}
8226 \centerline{{\it\fontname\tenit\/} and}
8227 \centerline{{\sl\fontname\tensl\/}}
8228 \centerline{are used for emphasis.}\vfill}
8230 % Blame: doc@@cygnus.com, 28mar91.