3 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
4 @c 2001, 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
7 @include configdoc.texi
8 @c (configdoc.texi is generated by the Makefile)
14 @macro gcctabopt{body}
20 @c Configure for the generation of man pages
58 * Ld: (ld). The GNU linker.
64 This file documents the @sc{gnu} linker LD version @value{VERSION}.
66 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000,
67 2001, 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
71 Permission is granted to copy, distribute and/or modify this document
72 under the terms of the GNU Free Documentation License, Version 1.1
73 or any later version published by the Free Software Foundation;
74 with no Invariant Sections, with no Front-Cover Texts, and with no
75 Back-Cover Texts. A copy of the license is included in the
76 section entitled ``GNU Free Documentation License''.
78 Permission is granted to process this file through Tex and print the
79 results, provided the printed document carries copying permission
80 notice identical to this one except for the removal of this paragraph
81 (this paragraph not being relevant to the printed manual).
87 @setchapternewpage odd
88 @settitle The GNU linker
92 @subtitle @code{ld} version 2
93 @subtitle Version @value{VERSION}
94 @author Steve Chamberlain
95 @author Ian Lance Taylor
100 \hfill Red Hat Inc\par
101 \hfill nickc\@credhat.com, doc\@redhat.com\par
102 \hfill {\it The GNU linker}\par
103 \hfill Edited by Jeffrey Osier (jeffrey\@cygnus.com)\par
105 \global\parindent=0pt % Steve likes it this way.
108 @vskip 0pt plus 1filll
109 @c man begin COPYRIGHT
110 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001,
111 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
113 Permission is granted to copy, distribute and/or modify this document
114 under the terms of the GNU Free Documentation License, Version 1.1
115 or any later version published by the Free Software Foundation;
116 with no Invariant Sections, with no Front-Cover Texts, and with no
117 Back-Cover Texts. A copy of the license is included in the
118 section entitled ``GNU Free Documentation License''.
123 @c FIXME: Talk about importance of *order* of args, cmds to linker!
128 This file documents the @sc{gnu} linker ld version @value{VERSION}.
130 This document is distributed under the terms of the GNU Free
131 Documentation License. A copy of the license is included in the
132 section entitled ``GNU Free Documentation License''.
135 * Overview:: Overview
136 * Invocation:: Invocation
137 * Scripts:: Linker Scripts
139 * Machine Dependent:: Machine Dependent Features
143 * H8/300:: ld and the H8/300
146 * Renesas:: ld and other Renesas micros
149 * i960:: ld and the Intel 960 family
152 * ARM:: ld and the ARM family
155 * HPPA ELF32:: ld and HPPA 32-bit ELF
158 * M68HC11/68HC12:: ld and the Motorola 68HC11 and 68HC12 families
161 * PowerPC ELF32:: ld and PowerPC 32-bit ELF Support
164 * PowerPC64 ELF64:: ld and PowerPC64 64-bit ELF Support
167 * TI COFF:: ld and the TI COFF
170 * Win32:: ld and WIN32 (cygwin/mingw)
173 * Xtensa:: ld and Xtensa Processors
176 @ifclear SingleFormat
179 @c Following blank line required for remaining bug in makeinfo conds/menus
181 * Reporting Bugs:: Reporting Bugs
182 * MRI:: MRI Compatible Script Files
183 * GNU Free Documentation License:: GNU Free Documentation License
184 * LD Index:: LD Index
191 @cindex @sc{gnu} linker
192 @cindex what is this?
195 @c man begin SYNOPSIS
196 ld [@b{options}] @var{objfile} @dots{}
200 ar(1), nm(1), objcopy(1), objdump(1), readelf(1) and
201 the Info entries for @file{binutils} and
206 @c man begin DESCRIPTION
208 @command{ld} combines a number of object and archive files, relocates
209 their data and ties up symbol references. Usually the last step in
210 compiling a program is to run @command{ld}.
212 @command{ld} accepts Linker Command Language files written in
213 a superset of AT&T's Link Editor Command Language syntax,
214 to provide explicit and total control over the linking process.
218 This man page does not describe the command language; see the
219 @command{ld} entry in @code{info} for full details on the command
220 language and on other aspects of the GNU linker.
223 @ifclear SingleFormat
224 This version of @command{ld} uses the general purpose BFD libraries
225 to operate on object files. This allows @command{ld} to read, combine, and
226 write object files in many different formats---for example, COFF or
227 @code{a.out}. Different formats may be linked together to produce any
228 available kind of object file. @xref{BFD}, for more information.
231 Aside from its flexibility, the @sc{gnu} linker is more helpful than other
232 linkers in providing diagnostic information. Many linkers abandon
233 execution immediately upon encountering an error; whenever possible,
234 @command{ld} continues executing, allowing you to identify other errors
235 (or, in some cases, to get an output file in spite of the error).
242 @c man begin DESCRIPTION
244 The @sc{gnu} linker @command{ld} is meant to cover a broad range of situations,
245 and to be as compatible as possible with other linkers. As a result,
246 you have many choices to control its behavior.
252 * Options:: Command Line Options
253 * Environment:: Environment Variables
257 @section Command Line Options
265 The linker supports a plethora of command-line options, but in actual
266 practice few of them are used in any particular context.
267 @cindex standard Unix system
268 For instance, a frequent use of @command{ld} is to link standard Unix
269 object files on a standard, supported Unix system. On such a system, to
270 link a file @code{hello.o}:
273 ld -o @var{output} /lib/crt0.o hello.o -lc
276 This tells @command{ld} to produce a file called @var{output} as the
277 result of linking the file @code{/lib/crt0.o} with @code{hello.o} and
278 the library @code{libc.a}, which will come from the standard search
279 directories. (See the discussion of the @samp{-l} option below.)
281 Some of the command-line options to @command{ld} may be specified at any
282 point in the command line. However, options which refer to files, such
283 as @samp{-l} or @samp{-T}, cause the file to be read at the point at
284 which the option appears in the command line, relative to the object
285 files and other file options. Repeating non-file options with a
286 different argument will either have no further effect, or override prior
287 occurrences (those further to the left on the command line) of that
288 option. Options which may be meaningfully specified more than once are
289 noted in the descriptions below.
292 Non-option arguments are object files or archives which are to be linked
293 together. They may follow, precede, or be mixed in with command-line
294 options, except that an object file argument may not be placed between
295 an option and its argument.
297 Usually the linker is invoked with at least one object file, but you can
298 specify other forms of binary input files using @samp{-l}, @samp{-R},
299 and the script command language. If @emph{no} binary input files at all
300 are specified, the linker does not produce any output, and issues the
301 message @samp{No input files}.
303 If the linker cannot recognize the format of an object file, it will
304 assume that it is a linker script. A script specified in this way
305 augments the main linker script used for the link (either the default
306 linker script or the one specified by using @samp{-T}). This feature
307 permits the linker to link against a file which appears to be an object
308 or an archive, but actually merely defines some symbol values, or uses
309 @code{INPUT} or @code{GROUP} to load other objects. Note that
310 specifying a script in this way merely augments the main linker script;
311 use the @samp{-T} option to replace the default linker script entirely.
314 For options whose names are a single letter,
315 option arguments must either follow the option letter without intervening
316 whitespace, or be given as separate arguments immediately following the
317 option that requires them.
319 For options whose names are multiple letters, either one dash or two can
320 precede the option name; for example, @samp{-trace-symbol} and
321 @samp{--trace-symbol} are equivalent. Note---there is one exception to
322 this rule. Multiple letter options that start with a lower case 'o' can
323 only be preceded by two dashes. This is to reduce confusion with the
324 @samp{-o} option. So for example @samp{-omagic} sets the output file
325 name to @samp{magic} whereas @samp{--omagic} sets the NMAGIC flag on the
328 Arguments to multiple-letter options must either be separated from the
329 option name by an equals sign, or be given as separate arguments
330 immediately following the option that requires them. For example,
331 @samp{--trace-symbol foo} and @samp{--trace-symbol=foo} are equivalent.
332 Unique abbreviations of the names of multiple-letter options are
335 Note---if the linker is being invoked indirectly, via a compiler driver
336 (e.g. @samp{gcc}) then all the linker command line options should be
337 prefixed by @samp{-Wl,} (or whatever is appropriate for the particular
338 compiler driver) like this:
341 gcc -Wl,--startgroup foo.o bar.o -Wl,--endgroup
344 This is important, because otherwise the compiler driver program may
345 silently drop the linker options, resulting in a bad link.
347 Here is a table of the generic command line switches accepted by the GNU
351 @include at-file.texi
353 @kindex -a@var{keyword}
354 @item -a@var{keyword}
355 This option is supported for HP/UX compatibility. The @var{keyword}
356 argument must be one of the strings @samp{archive}, @samp{shared}, or
357 @samp{default}. @samp{-aarchive} is functionally equivalent to
358 @samp{-Bstatic}, and the other two keywords are functionally equivalent
359 to @samp{-Bdynamic}. This option may be used any number of times.
362 @cindex architectures
364 @item -A@var{architecture}
365 @kindex --architecture=@var{arch}
366 @itemx --architecture=@var{architecture}
367 In the current release of @command{ld}, this option is useful only for the
368 Intel 960 family of architectures. In that @command{ld} configuration, the
369 @var{architecture} argument identifies the particular architecture in
370 the 960 family, enabling some safeguards and modifying the
371 archive-library search path. @xref{i960,,@command{ld} and the Intel 960
372 family}, for details.
374 Future releases of @command{ld} may support similar functionality for
375 other architecture families.
378 @ifclear SingleFormat
379 @cindex binary input format
380 @kindex -b @var{format}
381 @kindex --format=@var{format}
384 @item -b @var{input-format}
385 @itemx --format=@var{input-format}
386 @command{ld} may be configured to support more than one kind of object
387 file. If your @command{ld} is configured this way, you can use the
388 @samp{-b} option to specify the binary format for input object files
389 that follow this option on the command line. Even when @command{ld} is
390 configured to support alternative object formats, you don't usually need
391 to specify this, as @command{ld} should be configured to expect as a
392 default input format the most usual format on each machine.
393 @var{input-format} is a text string, the name of a particular format
394 supported by the BFD libraries. (You can list the available binary
395 formats with @samp{objdump -i}.)
398 You may want to use this option if you are linking files with an unusual
399 binary format. You can also use @samp{-b} to switch formats explicitly (when
400 linking object files of different formats), by including
401 @samp{-b @var{input-format}} before each group of object files in a
404 The default format is taken from the environment variable
409 You can also define the input format from a script, using the command
412 see @ref{Format Commands}.
416 @kindex -c @var{MRI-cmdfile}
417 @kindex --mri-script=@var{MRI-cmdfile}
418 @cindex compatibility, MRI
419 @item -c @var{MRI-commandfile}
420 @itemx --mri-script=@var{MRI-commandfile}
421 For compatibility with linkers produced by MRI, @command{ld} accepts script
422 files written in an alternate, restricted command language, described in
424 @ref{MRI,,MRI Compatible Script Files}.
427 the MRI Compatible Script Files section of GNU ld documentation.
429 Introduce MRI script files with
430 the option @samp{-c}; use the @samp{-T} option to run linker
431 scripts written in the general-purpose @command{ld} scripting language.
432 If @var{MRI-cmdfile} does not exist, @command{ld} looks for it in the directories
433 specified by any @samp{-L} options.
435 @cindex common allocation
442 These three options are equivalent; multiple forms are supported for
443 compatibility with other linkers. They assign space to common symbols
444 even if a relocatable output file is specified (with @samp{-r}). The
445 script command @code{FORCE_COMMON_ALLOCATION} has the same effect.
446 @xref{Miscellaneous Commands}.
448 @cindex entry point, from command line
449 @kindex -e @var{entry}
450 @kindex --entry=@var{entry}
452 @itemx --entry=@var{entry}
453 Use @var{entry} as the explicit symbol for beginning execution of your
454 program, rather than the default entry point. If there is no symbol
455 named @var{entry}, the linker will try to parse @var{entry} as a number,
456 and use that as the entry address (the number will be interpreted in
457 base 10; you may use a leading @samp{0x} for base 16, or a leading
458 @samp{0} for base 8). @xref{Entry Point}, for a discussion of defaults
459 and other ways of specifying the entry point.
461 @kindex --exclude-libs
462 @item --exclude-libs @var{lib},@var{lib},...
463 Specifies a list of archive libraries from which symbols should not be automatically
464 exported. The library names may be delimited by commas or colons. Specifying
465 @code{--exclude-libs ALL} excludes symbols in all archive libraries from
466 automatic export. This option is available only for the i386 PE targeted
467 port of the linker and for ELF targeted ports. For i386 PE, symbols
468 explicitly listed in a .def file are still exported, regardless of this
469 option. For ELF targeted ports, symbols affected by this option will
470 be treated as hidden.
472 @cindex dynamic symbol table
474 @kindex --export-dynamic
476 @itemx --export-dynamic
477 When creating a dynamically linked executable, add all symbols to the
478 dynamic symbol table. The dynamic symbol table is the set of symbols
479 which are visible from dynamic objects at run time.
481 If you do not use this option, the dynamic symbol table will normally
482 contain only those symbols which are referenced by some dynamic object
483 mentioned in the link.
485 If you use @code{dlopen} to load a dynamic object which needs to refer
486 back to the symbols defined by the program, rather than some other
487 dynamic object, then you will probably need to use this option when
488 linking the program itself.
490 You can also use the dynamic list to control what symbols should
491 be added to the dynamic symbol table if the output format supports it.
492 See the description of @samp{--dynamic-list}.
494 @ifclear SingleFormat
495 @cindex big-endian objects
499 Link big-endian objects. This affects the default output format.
501 @cindex little-endian objects
504 Link little-endian objects. This affects the default output format.
510 @itemx --auxiliary @var{name}
511 When creating an ELF shared object, set the internal DT_AUXILIARY field
512 to the specified name. This tells the dynamic linker that the symbol
513 table of the shared object should be used as an auxiliary filter on the
514 symbol table of the shared object @var{name}.
516 If you later link a program against this filter object, then, when you
517 run the program, the dynamic linker will see the DT_AUXILIARY field. If
518 the dynamic linker resolves any symbols from the filter object, it will
519 first check whether there is a definition in the shared object
520 @var{name}. If there is one, it will be used instead of the definition
521 in the filter object. The shared object @var{name} need not exist.
522 Thus the shared object @var{name} may be used to provide an alternative
523 implementation of certain functions, perhaps for debugging or for
524 machine specific performance.
526 This option may be specified more than once. The DT_AUXILIARY entries
527 will be created in the order in which they appear on the command line.
532 @itemx --filter @var{name}
533 When creating an ELF shared object, set the internal DT_FILTER field to
534 the specified name. This tells the dynamic linker that the symbol table
535 of the shared object which is being created should be used as a filter
536 on the symbol table of the shared object @var{name}.
538 If you later link a program against this filter object, then, when you
539 run the program, the dynamic linker will see the DT_FILTER field. The
540 dynamic linker will resolve symbols according to the symbol table of the
541 filter object as usual, but it will actually link to the definitions
542 found in the shared object @var{name}. Thus the filter object can be
543 used to select a subset of the symbols provided by the object
546 Some older linkers used the @option{-F} option throughout a compilation
547 toolchain for specifying object-file format for both input and output
549 @ifclear SingleFormat
550 The @sc{gnu} linker uses other mechanisms for this purpose: the
551 @option{-b}, @option{--format}, @option{--oformat} options, the
552 @code{TARGET} command in linker scripts, and the @code{GNUTARGET}
553 environment variable.
555 The @sc{gnu} linker will ignore the @option{-F} option when not
556 creating an ELF shared object.
558 @cindex finalization function
560 @item -fini @var{name}
561 When creating an ELF executable or shared object, call NAME when the
562 executable or shared object is unloaded, by setting DT_FINI to the
563 address of the function. By default, the linker uses @code{_fini} as
564 the function to call.
568 Ignored. Provided for compatibility with other tools.
574 @itemx --gpsize=@var{value}
575 Set the maximum size of objects to be optimized using the GP register to
576 @var{size}. This is only meaningful for object file formats such as
577 MIPS ECOFF which supports putting large and small objects into different
578 sections. This is ignored for other object file formats.
580 @cindex runtime library name
582 @kindex -soname=@var{name}
584 @itemx -soname=@var{name}
585 When creating an ELF shared object, set the internal DT_SONAME field to
586 the specified name. When an executable is linked with a shared object
587 which has a DT_SONAME field, then when the executable is run the dynamic
588 linker will attempt to load the shared object specified by the DT_SONAME
589 field rather than the using the file name given to the linker.
592 @cindex incremental link
594 Perform an incremental link (same as option @samp{-r}).
596 @cindex initialization function
598 @item -init @var{name}
599 When creating an ELF executable or shared object, call NAME when the
600 executable or shared object is loaded, by setting DT_INIT to the address
601 of the function. By default, the linker uses @code{_init} as the
604 @cindex archive files, from cmd line
605 @kindex -l@var{archive}
606 @kindex --library=@var{archive}
607 @item -l@var{archive}
608 @itemx --library=@var{archive}
609 Add archive file @var{archive} to the list of files to link. This
610 option may be used any number of times. @command{ld} will search its
611 path-list for occurrences of @code{lib@var{archive}.a} for every
612 @var{archive} specified.
614 On systems which support shared libraries, @command{ld} may also search for
615 libraries with extensions other than @code{.a}. Specifically, on ELF
616 and SunOS systems, @command{ld} will search a directory for a library with
617 an extension of @code{.so} before searching for one with an extension of
618 @code{.a}. By convention, a @code{.so} extension indicates a shared
621 The linker will search an archive only once, at the location where it is
622 specified on the command line. If the archive defines a symbol which
623 was undefined in some object which appeared before the archive on the
624 command line, the linker will include the appropriate file(s) from the
625 archive. However, an undefined symbol in an object appearing later on
626 the command line will not cause the linker to search the archive again.
628 See the @option{-(} option for a way to force the linker to search
629 archives multiple times.
631 You may list the same archive multiple times on the command line.
634 This type of archive searching is standard for Unix linkers. However,
635 if you are using @command{ld} on AIX, note that it is different from the
636 behaviour of the AIX linker.
639 @cindex search directory, from cmd line
641 @kindex --library-path=@var{dir}
642 @item -L@var{searchdir}
643 @itemx --library-path=@var{searchdir}
644 Add path @var{searchdir} to the list of paths that @command{ld} will search
645 for archive libraries and @command{ld} control scripts. You may use this
646 option any number of times. The directories are searched in the order
647 in which they are specified on the command line. Directories specified
648 on the command line are searched before the default directories. All
649 @option{-L} options apply to all @option{-l} options, regardless of the
650 order in which the options appear.
652 If @var{searchdir} begins with @code{=}, then the @code{=} will be replaced
653 by the @dfn{sysroot prefix}, a path specified when the linker is configured.
656 The default set of paths searched (without being specified with
657 @samp{-L}) depends on which emulation mode @command{ld} is using, and in
658 some cases also on how it was configured. @xref{Environment}.
661 The paths can also be specified in a link script with the
662 @code{SEARCH_DIR} command. Directories specified this way are searched
663 at the point in which the linker script appears in the command line.
666 @kindex -m @var{emulation}
667 @item -m@var{emulation}
668 Emulate the @var{emulation} linker. You can list the available
669 emulations with the @samp{--verbose} or @samp{-V} options.
671 If the @samp{-m} option is not used, the emulation is taken from the
672 @code{LDEMULATION} environment variable, if that is defined.
674 Otherwise, the default emulation depends upon how the linker was
682 Print a link map to the standard output. A link map provides
683 information about the link, including the following:
687 Where object files are mapped into memory.
689 How common symbols are allocated.
691 All archive members included in the link, with a mention of the symbol
692 which caused the archive member to be brought in.
694 The values assigned to symbols.
696 Note - symbols whose values are computed by an expression which
697 involves a reference to a previous value of the same symbol may not
698 have correct result displayed in the link map. This is because the
699 linker discards intermediate results and only retains the final value
700 of an expression. Under such circumstances the linker will display
701 the final value enclosed by square brackets. Thus for example a
702 linker script containing:
710 will produce the following output in the link map if the @option{-M}
715 [0x0000000c] foo = (foo * 0x4)
716 [0x0000000c] foo = (foo + 0x8)
719 See @ref{Expressions} for more information about expressions in linker
724 @cindex read-only text
729 Turn off page alignment of sections, and mark the output as
730 @code{NMAGIC} if possible.
734 @cindex read/write from cmd line
738 Set the text and data sections to be readable and writable. Also, do
739 not page-align the data segment, and disable linking against shared
740 libraries. If the output format supports Unix style magic numbers,
741 mark the output as @code{OMAGIC}. Note: Although a writable text section
742 is allowed for PE-COFF targets, it does not conform to the format
743 specification published by Microsoft.
748 This option negates most of the effects of the @option{-N} option. It
749 sets the text section to be read-only, and forces the data segment to
750 be page-aligned. Note - this option does not enable linking against
751 shared libraries. Use @option{-Bdynamic} for this.
753 @kindex -o @var{output}
754 @kindex --output=@var{output}
755 @cindex naming the output file
756 @item -o @var{output}
757 @itemx --output=@var{output}
758 Use @var{output} as the name for the program produced by @command{ld}; if this
759 option is not specified, the name @file{a.out} is used by default. The
760 script command @code{OUTPUT} can also specify the output file name.
762 @kindex -O @var{level}
763 @cindex generating optimized output
765 If @var{level} is a numeric values greater than zero @command{ld} optimizes
766 the output. This might take significantly longer and therefore probably
767 should only be enabled for the final binary.
770 @kindex --emit-relocs
771 @cindex retain relocations in final executable
774 Leave relocation sections and contents in fully linked executables.
775 Post link analysis and optimization tools may need this information in
776 order to perform correct modifications of executables. This results
777 in larger executables.
779 This option is currently only supported on ELF platforms.
781 @kindex --force-dynamic
782 @cindex forcing the creation of dynamic sections
783 @item --force-dynamic
784 Force the output file to have dynamic sections. This option is specific
788 @cindex relocatable output
790 @kindex --relocatable
793 Generate relocatable output---i.e., generate an output file that can in
794 turn serve as input to @command{ld}. This is often called @dfn{partial
795 linking}. As a side effect, in environments that support standard Unix
796 magic numbers, this option also sets the output file's magic number to
798 @c ; see @option{-N}.
799 If this option is not specified, an absolute file is produced. When
800 linking C++ programs, this option @emph{will not} resolve references to
801 constructors; to do that, use @samp{-Ur}.
803 When an input file does not have the same format as the output file,
804 partial linking is only supported if that input file does not contain any
805 relocations. Different output formats can have further restrictions; for
806 example some @code{a.out}-based formats do not support partial linking
807 with input files in other formats at all.
809 This option does the same thing as @samp{-i}.
811 @kindex -R @var{file}
812 @kindex --just-symbols=@var{file}
813 @cindex symbol-only input
814 @item -R @var{filename}
815 @itemx --just-symbols=@var{filename}
816 Read symbol names and their addresses from @var{filename}, but do not
817 relocate it or include it in the output. This allows your output file
818 to refer symbolically to absolute locations of memory defined in other
819 programs. You may use this option more than once.
821 For compatibility with other ELF linkers, if the @option{-R} option is
822 followed by a directory name, rather than a file name, it is treated as
823 the @option{-rpath} option.
827 @cindex strip all symbols
830 Omit all symbol information from the output file.
833 @kindex --strip-debug
834 @cindex strip debugger symbols
837 Omit debugger symbol information (but not all symbols) from the output file.
841 @cindex input files, displaying
844 Print the names of the input files as @command{ld} processes them.
846 @kindex -T @var{script}
847 @kindex --script=@var{script}
849 @item -T @var{scriptfile}
850 @itemx --script=@var{scriptfile}
851 Use @var{scriptfile} as the linker script. This script replaces
852 @command{ld}'s default linker script (rather than adding to it), so
853 @var{commandfile} must specify everything necessary to describe the
854 output file. @xref{Scripts}. If @var{scriptfile} does not exist in
855 the current directory, @code{ld} looks for it in the directories
856 specified by any preceding @samp{-L} options. Multiple @samp{-T}
859 @kindex -u @var{symbol}
860 @kindex --undefined=@var{symbol}
861 @cindex undefined symbol
862 @item -u @var{symbol}
863 @itemx --undefined=@var{symbol}
864 Force @var{symbol} to be entered in the output file as an undefined
865 symbol. Doing this may, for example, trigger linking of additional
866 modules from standard libraries. @samp{-u} may be repeated with
867 different option arguments to enter additional undefined symbols. This
868 option is equivalent to the @code{EXTERN} linker script command.
873 For anything other than C++ programs, this option is equivalent to
874 @samp{-r}: it generates relocatable output---i.e., an output file that can in
875 turn serve as input to @command{ld}. When linking C++ programs, @samp{-Ur}
876 @emph{does} resolve references to constructors, unlike @samp{-r}.
877 It does not work to use @samp{-Ur} on files that were themselves linked
878 with @samp{-Ur}; once the constructor table has been built, it cannot
879 be added to. Use @samp{-Ur} only for the last partial link, and
880 @samp{-r} for the others.
882 @kindex --unique[=@var{SECTION}]
883 @item --unique[=@var{SECTION}]
884 Creates a separate output section for every input section matching
885 @var{SECTION}, or if the optional wildcard @var{SECTION} argument is
886 missing, for every orphan input section. An orphan section is one not
887 specifically mentioned in a linker script. You may use this option
888 multiple times on the command line; It prevents the normal merging of
889 input sections with the same name, overriding output section assignments
899 Display the version number for @command{ld}. The @option{-V} option also
900 lists the supported emulations.
903 @kindex --discard-all
904 @cindex deleting local symbols
907 Delete all local symbols.
910 @kindex --discard-locals
911 @cindex local symbols, deleting
913 @itemx --discard-locals
914 Delete all temporary local symbols. (These symbols start with
915 system-specific local label prefixes, typically @samp{.L} for ELF systems
916 or @samp{L} for traditional a.out systems.)
918 @kindex -y @var{symbol}
919 @kindex --trace-symbol=@var{symbol}
920 @cindex symbol tracing
921 @item -y @var{symbol}
922 @itemx --trace-symbol=@var{symbol}
923 Print the name of each linked file in which @var{symbol} appears. This
924 option may be given any number of times. On many systems it is necessary
925 to prepend an underscore.
927 This option is useful when you have an undefined symbol in your link but
928 don't know where the reference is coming from.
930 @kindex -Y @var{path}
932 Add @var{path} to the default library search path. This option exists
933 for Solaris compatibility.
935 @kindex -z @var{keyword}
936 @item -z @var{keyword}
937 The recognized keywords are:
941 Combines multiple reloc sections and sorts them to make dynamic symbol
942 lookup caching possible.
945 Disallows undefined symbols in object files. Undefined symbols in
946 shared libraries are still allowed.
949 Marks the object as requiring executable stack.
952 This option is only meaningful when building a shared object.
953 It marks the object so that its runtime initialization will occur
954 before the runtime initialization of any other objects brought into
955 the process at the same time. Similarly the runtime finalization of
956 the object will occur after the runtime finalization of any other
960 Marks the object that its symbol table interposes before all symbols
961 but the primary executable.
964 When generating an executable or shared library, mark it to tell the
965 dynamic linker to defer function call resolution to the point when
966 the function is called (lazy binding), rather than at load time.
967 Lazy binding is the default.
970 Marks the object that its filters be processed immediately at
974 Allows multiple definitions.
977 Disables multiple reloc sections combining.
980 Disables production of copy relocs.
983 Marks the object that the search for dependencies of this object will
984 ignore any default library search paths.
987 Marks the object shouldn't be unloaded at runtime.
990 Marks the object not available to @code{dlopen}.
993 Marks the object can not be dumped by @code{dldump}.
996 Marks the object as not requiring executable stack.
999 Don't create an ELF @code{PT_GNU_RELRO} segment header in the object.
1002 When generating an executable or shared library, mark it to tell the
1003 dynamic linker to resolve all symbols when the program is started, or
1004 when the shared library is linked to using dlopen, instead of
1005 deferring function call resolution to the point when the function is
1009 Marks the object may contain $ORIGIN.
1012 Create an ELF @code{PT_GNU_RELRO} segment header in the object.
1014 @item max-page-size=@var{value}
1015 Set the emulation maximum page size to @var{value}.
1017 @item common-page-size=@var{value}
1018 Set the emulation common page size to @var{value}.
1022 Other keywords are ignored for Solaris compatibility.
1025 @cindex groups of archives
1026 @item -( @var{archives} -)
1027 @itemx --start-group @var{archives} --end-group
1028 The @var{archives} should be a list of archive files. They may be
1029 either explicit file names, or @samp{-l} options.
1031 The specified archives are searched repeatedly until no new undefined
1032 references are created. Normally, an archive is searched only once in
1033 the order that it is specified on the command line. If a symbol in that
1034 archive is needed to resolve an undefined symbol referred to by an
1035 object in an archive that appears later on the command line, the linker
1036 would not be able to resolve that reference. By grouping the archives,
1037 they all be searched repeatedly until all possible references are
1040 Using this option has a significant performance cost. It is best to use
1041 it only when there are unavoidable circular references between two or
1044 @kindex --accept-unknown-input-arch
1045 @kindex --no-accept-unknown-input-arch
1046 @item --accept-unknown-input-arch
1047 @itemx --no-accept-unknown-input-arch
1048 Tells the linker to accept input files whose architecture cannot be
1049 recognised. The assumption is that the user knows what they are doing
1050 and deliberately wants to link in these unknown input files. This was
1051 the default behaviour of the linker, before release 2.14. The default
1052 behaviour from release 2.14 onwards is to reject such input files, and
1053 so the @samp{--accept-unknown-input-arch} option has been added to
1054 restore the old behaviour.
1057 @kindex --no-as-needed
1059 @itemx --no-as-needed
1060 This option affects ELF DT_NEEDED tags for dynamic libraries mentioned
1061 on the command line after the @option{--as-needed} option. Normally,
1062 the linker will add a DT_NEEDED tag for each dynamic library mentioned
1063 on the command line, regardless of whether the library is actually
1064 needed. @option{--as-needed} causes DT_NEEDED tags to only be emitted
1065 for libraries that satisfy some symbol reference from regular objects
1066 which is undefined at the point that the library was linked.
1067 @option{--no-as-needed} restores the default behaviour.
1069 @kindex --add-needed
1070 @kindex --no-add-needed
1072 @itemx --no-add-needed
1073 This option affects the treatment of dynamic libraries from ELF
1074 DT_NEEDED tags in dynamic libraries mentioned on the command line after
1075 the @option{--no-add-needed} option. Normally, the linker will add
1076 a DT_NEEDED tag for each dynamic library from DT_NEEDED tags.
1077 @option{--no-add-needed} causes DT_NEEDED tags will never be emitted
1078 for those libraries from DT_NEEDED tags. @option{--add-needed} restores
1079 the default behaviour.
1081 @kindex -assert @var{keyword}
1082 @item -assert @var{keyword}
1083 This option is ignored for SunOS compatibility.
1087 @kindex -call_shared
1091 Link against dynamic libraries. This is only meaningful on platforms
1092 for which shared libraries are supported. This option is normally the
1093 default on such platforms. The different variants of this option are
1094 for compatibility with various systems. You may use this option
1095 multiple times on the command line: it affects library searching for
1096 @option{-l} options which follow it.
1100 Set the @code{DF_1_GROUP} flag in the @code{DT_FLAGS_1} entry in the dynamic
1101 section. This causes the runtime linker to handle lookups in this
1102 object and its dependencies to be performed only inside the group.
1103 @option{--unresolved-symbols=report-all} is implied. This option is
1104 only meaningful on ELF platforms which support shared libraries.
1114 Do not link against shared libraries. This is only meaningful on
1115 platforms for which shared libraries are supported. The different
1116 variants of this option are for compatibility with various systems. You
1117 may use this option multiple times on the command line: it affects
1118 library searching for @option{-l} options which follow it. This
1119 option also implies @option{--unresolved-symbols=report-all}. This
1120 option can be used with @option{-shared}. Doing so means that a
1121 shared library is being created but that all of the library's external
1122 references must be resolved by pulling in entries from static
1127 When creating a shared library, bind references to global symbols to the
1128 definition within the shared library, if any. Normally, it is possible
1129 for a program linked against a shared library to override the definition
1130 within the shared library. This option is only meaningful on ELF
1131 platforms which support shared libraries.
1133 @kindex -Bsymbolic-functions
1134 @item -Bsymbolic-functions
1135 When creating a shared library, bind references to global function
1136 symbols to the definition within the shared library, if any.
1137 @option{-Bsymbolic-functions} is an alias for @option{--dynamic-list-data}.
1138 This option is only meaningful on ELF platforms which support shared
1141 @kindex --dynamic-list=@var{dynamic-list-file}
1142 @item --dynamic-list=@var{dynamic-list-file}
1143 Specify the name of a dynamic list file to the linker. This is
1144 typically used when creating shared libraries to specify a list of
1145 global symbols whose references shouldn't be bound to the definition
1146 within the shared library, or creating dynamically linked executables
1147 to specify a list of symbols which should be added to the symbol table
1148 in the executable. This option is only meaningful on ELF platforms
1149 which support shared libraries.
1151 The format of the dynamic list is the same as the version node without
1152 scope and node name. See @ref{VERSION} for more information.
1154 @kindex --dynamic-list-data
1155 @item --dynamic-list-data
1156 Include all global data symbols to the dynamic list.
1158 @kindex --dynamic-list-cpp-new
1159 @item --dynamic-list-cpp-new
1160 Provide the builtin dynamic list for C++ operator new and delete. It
1161 is mainly useful for building shared libstdc++.
1163 @kindex --dynamic-list-cpp-typeinfo
1164 @item --dynamic-list-cpp-typeinfo
1165 Provide the builtin dynamic list for C++ runtime type identification.
1167 @kindex --check-sections
1168 @kindex --no-check-sections
1169 @item --check-sections
1170 @itemx --no-check-sections
1171 Asks the linker @emph{not} to check section addresses after they have
1172 been assigned to see if there are any overlaps. Normally the linker will
1173 perform this check, and if it finds any overlaps it will produce
1174 suitable error messages. The linker does know about, and does make
1175 allowances for sections in overlays. The default behaviour can be
1176 restored by using the command line switch @option{--check-sections}.
1178 @cindex cross reference table
1181 Output a cross reference table. If a linker map file is being
1182 generated, the cross reference table is printed to the map file.
1183 Otherwise, it is printed on the standard output.
1185 The format of the table is intentionally simple, so that it may be
1186 easily processed by a script if necessary. The symbols are printed out,
1187 sorted by name. For each symbol, a list of file names is given. If the
1188 symbol is defined, the first file listed is the location of the
1189 definition. The remaining files contain references to the symbol.
1191 @cindex common allocation
1192 @kindex --no-define-common
1193 @item --no-define-common
1194 This option inhibits the assignment of addresses to common symbols.
1195 The script command @code{INHIBIT_COMMON_ALLOCATION} has the same effect.
1196 @xref{Miscellaneous Commands}.
1198 The @samp{--no-define-common} option allows decoupling
1199 the decision to assign addresses to Common symbols from the choice
1200 of the output file type; otherwise a non-Relocatable output type
1201 forces assigning addresses to Common symbols.
1202 Using @samp{--no-define-common} allows Common symbols that are referenced
1203 from a shared library to be assigned addresses only in the main program.
1204 This eliminates the unused duplicate space in the shared library,
1205 and also prevents any possible confusion over resolving to the wrong
1206 duplicate when there are many dynamic modules with specialized search
1207 paths for runtime symbol resolution.
1209 @cindex symbols, from command line
1210 @kindex --defsym @var{symbol}=@var{exp}
1211 @item --defsym @var{symbol}=@var{expression}
1212 Create a global symbol in the output file, containing the absolute
1213 address given by @var{expression}. You may use this option as many
1214 times as necessary to define multiple symbols in the command line. A
1215 limited form of arithmetic is supported for the @var{expression} in this
1216 context: you may give a hexadecimal constant or the name of an existing
1217 symbol, or use @code{+} and @code{-} to add or subtract hexadecimal
1218 constants or symbols. If you need more elaborate expressions, consider
1219 using the linker command language from a script (@pxref{Assignments,,
1220 Assignment: Symbol Definitions}). @emph{Note:} there should be no white
1221 space between @var{symbol}, the equals sign (``@key{=}''), and
1224 @cindex demangling, from command line
1225 @kindex --demangle[=@var{style}]
1226 @kindex --no-demangle
1227 @item --demangle[=@var{style}]
1228 @itemx --no-demangle
1229 These options control whether to demangle symbol names in error messages
1230 and other output. When the linker is told to demangle, it tries to
1231 present symbol names in a readable fashion: it strips leading
1232 underscores if they are used by the object file format, and converts C++
1233 mangled symbol names into user readable names. Different compilers have
1234 different mangling styles. The optional demangling style argument can be used
1235 to choose an appropriate demangling style for your compiler. The linker will
1236 demangle by default unless the environment variable @samp{COLLECT_NO_DEMANGLE}
1237 is set. These options may be used to override the default.
1239 @cindex dynamic linker, from command line
1240 @kindex -I@var{file}
1241 @kindex --dynamic-linker @var{file}
1242 @item --dynamic-linker @var{file}
1243 Set the name of the dynamic linker. This is only meaningful when
1244 generating dynamically linked ELF executables. The default dynamic
1245 linker is normally correct; don't use this unless you know what you are
1249 @kindex --fatal-warnings
1250 @item --fatal-warnings
1251 Treat all warnings as errors.
1253 @kindex --force-exe-suffix
1254 @item --force-exe-suffix
1255 Make sure that an output file has a .exe suffix.
1257 If a successfully built fully linked output file does not have a
1258 @code{.exe} or @code{.dll} suffix, this option forces the linker to copy
1259 the output file to one of the same name with a @code{.exe} suffix. This
1260 option is useful when using unmodified Unix makefiles on a Microsoft
1261 Windows host, since some versions of Windows won't run an image unless
1262 it ends in a @code{.exe} suffix.
1264 @kindex --gc-sections
1265 @kindex --no-gc-sections
1266 @cindex garbage collection
1268 @itemx --no-gc-sections
1269 Enable garbage collection of unused input sections. It is ignored on
1270 targets that do not support this option. This option is not compatible
1271 with @samp{-r} or @samp{--emit-relocs}. The default behaviour (of not
1272 performing this garbage collection) can be restored by specifying
1273 @samp{--no-gc-sections} on the command line.
1275 @kindex --print-gc-sections
1276 @kindex --no-print-gc-sections
1277 @cindex garbage collection
1278 @item --print-gc-sections
1279 @itemx --no-print-gc-sections
1280 List all sections removed by garbage collection. The listing is
1281 printed on stderr. This option is only effective if garbage
1282 collection has been enabled via the @samp{--gc-sections}) option. The
1283 default behaviour (of not listing the sections that are removed) can
1284 be restored by specifying @samp{--no-print-gc-sections} on the command
1291 Print a summary of the command-line options on the standard output and exit.
1293 @kindex --target-help
1295 Print a summary of all target specific options on the standard output and exit.
1298 @item -Map @var{mapfile}
1299 Print a link map to the file @var{mapfile}. See the description of the
1300 @option{-M} option, above.
1302 @cindex memory usage
1303 @kindex --no-keep-memory
1304 @item --no-keep-memory
1305 @command{ld} normally optimizes for speed over memory usage by caching the
1306 symbol tables of input files in memory. This option tells @command{ld} to
1307 instead optimize for memory usage, by rereading the symbol tables as
1308 necessary. This may be required if @command{ld} runs out of memory space
1309 while linking a large executable.
1311 @kindex --no-undefined
1313 @item --no-undefined
1315 Report unresolved symbol references from regular object files. This
1316 is done even if the linker is creating a non-symbolic shared library.
1317 The switch @option{--[no-]allow-shlib-undefined} controls the
1318 behaviour for reporting unresolved references found in shared
1319 libraries being linked in.
1321 @kindex --allow-multiple-definition
1323 @item --allow-multiple-definition
1325 Normally when a symbol is defined multiple times, the linker will
1326 report a fatal error. These options allow multiple definitions and the
1327 first definition will be used.
1329 @kindex --allow-shlib-undefined
1330 @kindex --no-allow-shlib-undefined
1331 @item --allow-shlib-undefined
1332 @itemx --no-allow-shlib-undefined
1333 Allows (the default) or disallows undefined symbols in shared libraries.
1334 This switch is similar to @option{--no-undefined} except that it
1335 determines the behaviour when the undefined symbols are in a
1336 shared library rather than a regular object file. It does not affect
1337 how undefined symbols in regular object files are handled.
1339 The reason that @option{--allow-shlib-undefined} is the default is that
1340 the shared library being specified at link time may not be the same as
1341 the one that is available at load time, so the symbols might actually be
1342 resolvable at load time. Plus there are some systems, (eg BeOS) where
1343 undefined symbols in shared libraries is normal. (The kernel patches
1344 them at load time to select which function is most appropriate
1345 for the current architecture. This is used for example to dynamically
1346 select an appropriate memset function). Apparently it is also normal
1347 for HPPA shared libraries to have undefined symbols.
1349 @kindex --no-undefined-version
1350 @item --no-undefined-version
1351 Normally when a symbol has an undefined version, the linker will ignore
1352 it. This option disallows symbols with undefined version and a fatal error
1353 will be issued instead.
1355 @kindex --default-symver
1356 @item --default-symver
1357 Create and use a default symbol version (the soname) for unversioned
1360 @kindex --default-imported-symver
1361 @item --default-imported-symver
1362 Create and use a default symbol version (the soname) for unversioned
1365 @kindex --no-warn-mismatch
1366 @item --no-warn-mismatch
1367 Normally @command{ld} will give an error if you try to link together input
1368 files that are mismatched for some reason, perhaps because they have
1369 been compiled for different processors or for different endiannesses.
1370 This option tells @command{ld} that it should silently permit such possible
1371 errors. This option should only be used with care, in cases when you
1372 have taken some special action that ensures that the linker errors are
1375 @kindex --no-whole-archive
1376 @item --no-whole-archive
1377 Turn off the effect of the @option{--whole-archive} option for subsequent
1380 @cindex output file after errors
1381 @kindex --noinhibit-exec
1382 @item --noinhibit-exec
1383 Retain the executable output file whenever it is still usable.
1384 Normally, the linker will not produce an output file if it encounters
1385 errors during the link process; it exits without writing an output file
1386 when it issues any error whatsoever.
1390 Only search library directories explicitly specified on the
1391 command line. Library directories specified in linker scripts
1392 (including linker scripts specified on the command line) are ignored.
1394 @ifclear SingleFormat
1396 @item --oformat @var{output-format}
1397 @command{ld} may be configured to support more than one kind of object
1398 file. If your @command{ld} is configured this way, you can use the
1399 @samp{--oformat} option to specify the binary format for the output
1400 object file. Even when @command{ld} is configured to support alternative
1401 object formats, you don't usually need to specify this, as @command{ld}
1402 should be configured to produce as a default output format the most
1403 usual format on each machine. @var{output-format} is a text string, the
1404 name of a particular format supported by the BFD libraries. (You can
1405 list the available binary formats with @samp{objdump -i}.) The script
1406 command @code{OUTPUT_FORMAT} can also specify the output format, but
1407 this option overrides it. @xref{BFD}.
1411 @kindex --pic-executable
1413 @itemx --pic-executable
1414 @cindex position independent executables
1415 Create a position independent executable. This is currently only supported on
1416 ELF platforms. Position independent executables are similar to shared
1417 libraries in that they are relocated by the dynamic linker to the virtual
1418 address the OS chooses for them (which can vary between invocations). Like
1419 normal dynamically linked executables they can be executed and symbols
1420 defined in the executable cannot be overridden by shared libraries.
1424 This option is ignored for Linux compatibility.
1428 This option is ignored for SVR4 compatibility.
1431 @cindex synthesizing linker
1432 @cindex relaxing addressing modes
1434 An option with machine dependent effects.
1436 This option is only supported on a few targets.
1439 @xref{H8/300,,@command{ld} and the H8/300}.
1442 @xref{i960,, @command{ld} and the Intel 960 family}.
1445 @xref{Xtensa,, @command{ld} and Xtensa Processors}.
1448 @xref{M68HC11/68HC12,,@command{ld} and the 68HC11 and 68HC12}.
1451 @xref{PowerPC ELF32,,@command{ld} and PowerPC 32-bit ELF Support}.
1454 On some platforms, the @samp{--relax} option performs global
1455 optimizations that become possible when the linker resolves addressing
1456 in the program, such as relaxing address modes and synthesizing new
1457 instructions in the output object file.
1459 On some platforms these link time global optimizations may make symbolic
1460 debugging of the resulting executable impossible.
1463 the case for the Matsushita MN10200 and MN10300 family of processors.
1467 On platforms where this is not supported, @samp{--relax} is accepted,
1471 @cindex retaining specified symbols
1472 @cindex stripping all but some symbols
1473 @cindex symbols, retaining selectively
1474 @item --retain-symbols-file @var{filename}
1475 Retain @emph{only} the symbols listed in the file @var{filename},
1476 discarding all others. @var{filename} is simply a flat file, with one
1477 symbol name per line. This option is especially useful in environments
1481 where a large global symbol table is accumulated gradually, to conserve
1484 @samp{--retain-symbols-file} does @emph{not} discard undefined symbols,
1485 or symbols needed for relocations.
1487 You may only specify @samp{--retain-symbols-file} once in the command
1488 line. It overrides @samp{-s} and @samp{-S}.
1491 @item -rpath @var{dir}
1492 @cindex runtime library search path
1494 Add a directory to the runtime library search path. This is used when
1495 linking an ELF executable with shared objects. All @option{-rpath}
1496 arguments are concatenated and passed to the runtime linker, which uses
1497 them to locate shared objects at runtime. The @option{-rpath} option is
1498 also used when locating shared objects which are needed by shared
1499 objects explicitly included in the link; see the description of the
1500 @option{-rpath-link} option. If @option{-rpath} is not used when linking an
1501 ELF executable, the contents of the environment variable
1502 @code{LD_RUN_PATH} will be used if it is defined.
1504 The @option{-rpath} option may also be used on SunOS. By default, on
1505 SunOS, the linker will form a runtime search patch out of all the
1506 @option{-L} options it is given. If a @option{-rpath} option is used, the
1507 runtime search path will be formed exclusively using the @option{-rpath}
1508 options, ignoring the @option{-L} options. This can be useful when using
1509 gcc, which adds many @option{-L} options which may be on NFS mounted
1512 For compatibility with other ELF linkers, if the @option{-R} option is
1513 followed by a directory name, rather than a file name, it is treated as
1514 the @option{-rpath} option.
1518 @cindex link-time runtime library search path
1520 @item -rpath-link @var{DIR}
1521 When using ELF or SunOS, one shared library may require another. This
1522 happens when an @code{ld -shared} link includes a shared library as one
1525 When the linker encounters such a dependency when doing a non-shared,
1526 non-relocatable link, it will automatically try to locate the required
1527 shared library and include it in the link, if it is not included
1528 explicitly. In such a case, the @option{-rpath-link} option
1529 specifies the first set of directories to search. The
1530 @option{-rpath-link} option may specify a sequence of directory names
1531 either by specifying a list of names separated by colons, or by
1532 appearing multiple times.
1534 This option should be used with caution as it overrides the search path
1535 that may have been hard compiled into a shared library. In such a case it
1536 is possible to use unintentionally a different search path than the
1537 runtime linker would do.
1539 The linker uses the following search paths to locate required shared
1543 Any directories specified by @option{-rpath-link} options.
1545 Any directories specified by @option{-rpath} options. The difference
1546 between @option{-rpath} and @option{-rpath-link} is that directories
1547 specified by @option{-rpath} options are included in the executable and
1548 used at runtime, whereas the @option{-rpath-link} option is only effective
1549 at link time. Searching @option{-rpath} in this way is only supported
1550 by native linkers and cross linkers which have been configured with
1551 the @option{--with-sysroot} option.
1553 On an ELF system, if the @option{-rpath} and @code{rpath-link} options
1554 were not used, search the contents of the environment variable
1555 @code{LD_RUN_PATH}. It is for the native linker only.
1557 On SunOS, if the @option{-rpath} option was not used, search any
1558 directories specified using @option{-L} options.
1560 For a native linker, the contents of the environment variable
1561 @code{LD_LIBRARY_PATH}.
1563 For a native ELF linker, the directories in @code{DT_RUNPATH} or
1564 @code{DT_RPATH} of a shared library are searched for shared
1565 libraries needed by it. The @code{DT_RPATH} entries are ignored if
1566 @code{DT_RUNPATH} entries exist.
1568 The default directories, normally @file{/lib} and @file{/usr/lib}.
1570 For a native linker on an ELF system, if the file @file{/etc/ld.so.conf}
1571 exists, the list of directories found in that file.
1574 If the required shared library is not found, the linker will issue a
1575 warning and continue with the link.
1582 @cindex shared libraries
1583 Create a shared library. This is currently only supported on ELF, XCOFF
1584 and SunOS platforms. On SunOS, the linker will automatically create a
1585 shared library if the @option{-e} option is not used and there are
1586 undefined symbols in the link.
1589 @kindex --sort-common
1590 This option tells @command{ld} to sort the common symbols by size when it
1591 places them in the appropriate output sections. First come all the one
1592 byte symbols, then all the two byte, then all the four byte, and then
1593 everything else. This is to prevent gaps between symbols due to
1594 alignment constraints.
1596 @kindex --sort-section name
1597 @item --sort-section name
1598 This option will apply @code{SORT_BY_NAME} to all wildcard section
1599 patterns in the linker script.
1601 @kindex --sort-section alignment
1602 @item --sort-section alignment
1603 This option will apply @code{SORT_BY_ALIGNMENT} to all wildcard section
1604 patterns in the linker script.
1606 @kindex --split-by-file
1607 @item --split-by-file [@var{size}]
1608 Similar to @option{--split-by-reloc} but creates a new output section for
1609 each input file when @var{size} is reached. @var{size} defaults to a
1610 size of 1 if not given.
1612 @kindex --split-by-reloc
1613 @item --split-by-reloc [@var{count}]
1614 Tries to creates extra sections in the output file so that no single
1615 output section in the file contains more than @var{count} relocations.
1616 This is useful when generating huge relocatable files for downloading into
1617 certain real time kernels with the COFF object file format; since COFF
1618 cannot represent more than 65535 relocations in a single section. Note
1619 that this will fail to work with object file formats which do not
1620 support arbitrary sections. The linker will not split up individual
1621 input sections for redistribution, so if a single input section contains
1622 more than @var{count} relocations one output section will contain that
1623 many relocations. @var{count} defaults to a value of 32768.
1627 Compute and display statistics about the operation of the linker, such
1628 as execution time and memory usage.
1631 @item --sysroot=@var{directory}
1632 Use @var{directory} as the location of the sysroot, overriding the
1633 configure-time default. This option is only supported by linkers
1634 that were configured using @option{--with-sysroot}.
1636 @kindex --traditional-format
1637 @cindex traditional format
1638 @item --traditional-format
1639 For some targets, the output of @command{ld} is different in some ways from
1640 the output of some existing linker. This switch requests @command{ld} to
1641 use the traditional format instead.
1644 For example, on SunOS, @command{ld} combines duplicate entries in the
1645 symbol string table. This can reduce the size of an output file with
1646 full debugging information by over 30 percent. Unfortunately, the SunOS
1647 @code{dbx} program can not read the resulting program (@code{gdb} has no
1648 trouble). The @samp{--traditional-format} switch tells @command{ld} to not
1649 combine duplicate entries.
1651 @kindex --section-start @var{sectionname}=@var{org}
1652 @item --section-start @var{sectionname}=@var{org}
1653 Locate a section in the output file at the absolute
1654 address given by @var{org}. You may use this option as many
1655 times as necessary to locate multiple sections in the command
1657 @var{org} must be a single hexadecimal integer;
1658 for compatibility with other linkers, you may omit the leading
1659 @samp{0x} usually associated with hexadecimal values. @emph{Note:} there
1660 should be no white space between @var{sectionname}, the equals
1661 sign (``@key{=}''), and @var{org}.
1663 @kindex -Tbss @var{org}
1664 @kindex -Tdata @var{org}
1665 @kindex -Ttext @var{org}
1666 @cindex segment origins, cmd line
1667 @item -Tbss @var{org}
1668 @itemx -Tdata @var{org}
1669 @itemx -Ttext @var{org}
1670 Same as --section-start, with @code{.bss}, @code{.data} or
1671 @code{.text} as the @var{sectionname}.
1673 @kindex --unresolved-symbols
1674 @item --unresolved-symbols=@var{method}
1675 Determine how to handle unresolved symbols. There are four possible
1676 values for @samp{method}:
1680 Do not report any unresolved symbols.
1683 Report all unresolved symbols. This is the default.
1685 @item ignore-in-object-files
1686 Report unresolved symbols that are contained in shared libraries, but
1687 ignore them if they come from regular object files.
1689 @item ignore-in-shared-libs
1690 Report unresolved symbols that come from regular object files, but
1691 ignore them if they come from shared libraries. This can be useful
1692 when creating a dynamic binary and it is known that all the shared
1693 libraries that it should be referencing are included on the linker's
1697 The behaviour for shared libraries on their own can also be controlled
1698 by the @option{--[no-]allow-shlib-undefined} option.
1700 Normally the linker will generate an error message for each reported
1701 unresolved symbol but the option @option{--warn-unresolved-symbols}
1702 can change this to a warning.
1708 Display the version number for @command{ld} and list the linker emulations
1709 supported. Display which input files can and cannot be opened. Display
1710 the linker script being used by the linker.
1712 @kindex --version-script=@var{version-scriptfile}
1713 @cindex version script, symbol versions
1714 @itemx --version-script=@var{version-scriptfile}
1715 Specify the name of a version script to the linker. This is typically
1716 used when creating shared libraries to specify additional information
1717 about the version hierarchy for the library being created. This option
1718 is only meaningful on ELF platforms which support shared libraries.
1721 @kindex --warn-common
1722 @cindex warnings, on combining symbols
1723 @cindex combining symbols, warnings on
1725 Warn when a common symbol is combined with another common symbol or with
1726 a symbol definition. Unix linkers allow this somewhat sloppy practise,
1727 but linkers on some other operating systems do not. This option allows
1728 you to find potential problems from combining global symbols.
1729 Unfortunately, some C libraries use this practise, so you may get some
1730 warnings about symbols in the libraries as well as in your programs.
1732 There are three kinds of global symbols, illustrated here by C examples:
1736 A definition, which goes in the initialized data section of the output
1740 An undefined reference, which does not allocate space.
1741 There must be either a definition or a common symbol for the
1745 A common symbol. If there are only (one or more) common symbols for a
1746 variable, it goes in the uninitialized data area of the output file.
1747 The linker merges multiple common symbols for the same variable into a
1748 single symbol. If they are of different sizes, it picks the largest
1749 size. The linker turns a common symbol into a declaration, if there is
1750 a definition of the same variable.
1753 The @samp{--warn-common} option can produce five kinds of warnings.
1754 Each warning consists of a pair of lines: the first describes the symbol
1755 just encountered, and the second describes the previous symbol
1756 encountered with the same name. One or both of the two symbols will be
1761 Turning a common symbol into a reference, because there is already a
1762 definition for the symbol.
1764 @var{file}(@var{section}): warning: common of `@var{symbol}'
1765 overridden by definition
1766 @var{file}(@var{section}): warning: defined here
1770 Turning a common symbol into a reference, because a later definition for
1771 the symbol is encountered. This is the same as the previous case,
1772 except that the symbols are encountered in a different order.
1774 @var{file}(@var{section}): warning: definition of `@var{symbol}'
1776 @var{file}(@var{section}): warning: common is here
1780 Merging a common symbol with a previous same-sized common symbol.
1782 @var{file}(@var{section}): warning: multiple common
1784 @var{file}(@var{section}): warning: previous common is here
1788 Merging a common symbol with a previous larger common symbol.
1790 @var{file}(@var{section}): warning: common of `@var{symbol}'
1791 overridden by larger common
1792 @var{file}(@var{section}): warning: larger common is here
1796 Merging a common symbol with a previous smaller common symbol. This is
1797 the same as the previous case, except that the symbols are
1798 encountered in a different order.
1800 @var{file}(@var{section}): warning: common of `@var{symbol}'
1801 overriding smaller common
1802 @var{file}(@var{section}): warning: smaller common is here
1806 @kindex --warn-constructors
1807 @item --warn-constructors
1808 Warn if any global constructors are used. This is only useful for a few
1809 object file formats. For formats like COFF or ELF, the linker can not
1810 detect the use of global constructors.
1812 @kindex --warn-multiple-gp
1813 @item --warn-multiple-gp
1814 Warn if multiple global pointer values are required in the output file.
1815 This is only meaningful for certain processors, such as the Alpha.
1816 Specifically, some processors put large-valued constants in a special
1817 section. A special register (the global pointer) points into the middle
1818 of this section, so that constants can be loaded efficiently via a
1819 base-register relative addressing mode. Since the offset in
1820 base-register relative mode is fixed and relatively small (e.g., 16
1821 bits), this limits the maximum size of the constant pool. Thus, in
1822 large programs, it is often necessary to use multiple global pointer
1823 values in order to be able to address all possible constants. This
1824 option causes a warning to be issued whenever this case occurs.
1827 @cindex warnings, on undefined symbols
1828 @cindex undefined symbols, warnings on
1830 Only warn once for each undefined symbol, rather than once per module
1833 @kindex --warn-section-align
1834 @cindex warnings, on section alignment
1835 @cindex section alignment, warnings on
1836 @item --warn-section-align
1837 Warn if the address of an output section is changed because of
1838 alignment. Typically, the alignment will be set by an input section.
1839 The address will only be changed if it not explicitly specified; that
1840 is, if the @code{SECTIONS} command does not specify a start address for
1841 the section (@pxref{SECTIONS}).
1843 @kindex --warn-shared-textrel
1844 @item --warn-shared-textrel
1845 Warn if the linker adds a DT_TEXTREL to a shared object.
1847 @kindex --warn-unresolved-symbols
1848 @item --warn-unresolved-symbols
1849 If the linker is going to report an unresolved symbol (see the option
1850 @option{--unresolved-symbols}) it will normally generate an error.
1851 This option makes it generate a warning instead.
1853 @kindex --error-unresolved-symbols
1854 @item --error-unresolved-symbols
1855 This restores the linker's default behaviour of generating errors when
1856 it is reporting unresolved symbols.
1858 @kindex --whole-archive
1859 @cindex including an entire archive
1860 @item --whole-archive
1861 For each archive mentioned on the command line after the
1862 @option{--whole-archive} option, include every object file in the archive
1863 in the link, rather than searching the archive for the required object
1864 files. This is normally used to turn an archive file into a shared
1865 library, forcing every object to be included in the resulting shared
1866 library. This option may be used more than once.
1868 Two notes when using this option from gcc: First, gcc doesn't know
1869 about this option, so you have to use @option{-Wl,-whole-archive}.
1870 Second, don't forget to use @option{-Wl,-no-whole-archive} after your
1871 list of archives, because gcc will add its own list of archives to
1872 your link and you may not want this flag to affect those as well.
1875 @item --wrap @var{symbol}
1876 Use a wrapper function for @var{symbol}. Any undefined reference to
1877 @var{symbol} will be resolved to @code{__wrap_@var{symbol}}. Any
1878 undefined reference to @code{__real_@var{symbol}} will be resolved to
1881 This can be used to provide a wrapper for a system function. The
1882 wrapper function should be called @code{__wrap_@var{symbol}}. If it
1883 wishes to call the system function, it should call
1884 @code{__real_@var{symbol}}.
1886 Here is a trivial example:
1890 __wrap_malloc (size_t c)
1892 printf ("malloc called with %zu\n", c);
1893 return __real_malloc (c);
1897 If you link other code with this file using @option{--wrap malloc}, then
1898 all calls to @code{malloc} will call the function @code{__wrap_malloc}
1899 instead. The call to @code{__real_malloc} in @code{__wrap_malloc} will
1900 call the real @code{malloc} function.
1902 You may wish to provide a @code{__real_malloc} function as well, so that
1903 links without the @option{--wrap} option will succeed. If you do this,
1904 you should not put the definition of @code{__real_malloc} in the same
1905 file as @code{__wrap_malloc}; if you do, the assembler may resolve the
1906 call before the linker has a chance to wrap it to @code{malloc}.
1908 @kindex --eh-frame-hdr
1909 @item --eh-frame-hdr
1910 Request creation of @code{.eh_frame_hdr} section and ELF
1911 @code{PT_GNU_EH_FRAME} segment header.
1913 @kindex --enable-new-dtags
1914 @kindex --disable-new-dtags
1915 @item --enable-new-dtags
1916 @itemx --disable-new-dtags
1917 This linker can create the new dynamic tags in ELF. But the older ELF
1918 systems may not understand them. If you specify
1919 @option{--enable-new-dtags}, the dynamic tags will be created as needed.
1920 If you specify @option{--disable-new-dtags}, no new dynamic tags will be
1921 created. By default, the new dynamic tags are not created. Note that
1922 those options are only available for ELF systems.
1924 @kindex --hash-size=@var{number}
1925 @item --hash-size=@var{number}
1926 Set the default size of the linker's hash tables to a prime number
1927 close to @var{number}. Increasing this value can reduce the length of
1928 time it takes the linker to perform its tasks, at the expense of
1929 increasing the linker's memory requirements. Similarly reducing this
1930 value can reduce the memory requirements at the expense of speed.
1932 @kindex --hash-style=@var{style}
1933 @item --hash-style=@var{style}
1934 Set the type of linker's hash table(s). @var{style} can be either
1935 @code{sysv} for classic ELF @code{.hash} section, @code{gnu} for
1936 new style GNU @code{.gnu.hash} section or @code{both} for both
1937 the classic ELF @code{.hash} and new style GNU @code{.gnu.hash}
1938 hash tables. The default is @code{sysv}.
1940 @kindex --reduce-memory-overheads
1941 @item --reduce-memory-overheads
1942 This option reduces memory requirements at ld runtime, at the expense of
1943 linking speed. This was introduced to select the old O(n^2) algorithm
1944 for link map file generation, rather than the new O(n) algorithm which uses
1945 about 40% more memory for symbol storage.
1947 Another effect of the switch is to set the default hash table size to
1948 1021, which again saves memory at the cost of lengthening the linker's
1949 run time. This is not done however if the @option{--hash-size} switch
1952 The @option{--reduce-memory-overheads} switch may be also be used to
1953 enable other tradeoffs in future versions of the linker.
1959 @subsection Options Specific to i386 PE Targets
1961 @c man begin OPTIONS
1963 The i386 PE linker supports the @option{-shared} option, which causes
1964 the output to be a dynamically linked library (DLL) instead of a
1965 normal executable. You should name the output @code{*.dll} when you
1966 use this option. In addition, the linker fully supports the standard
1967 @code{*.def} files, which may be specified on the linker command line
1968 like an object file (in fact, it should precede archives it exports
1969 symbols from, to ensure that they get linked in, just like a normal
1972 In addition to the options common to all targets, the i386 PE linker
1973 support additional command line options that are specific to the i386
1974 PE target. Options that take values may be separated from their
1975 values by either a space or an equals sign.
1979 @kindex --add-stdcall-alias
1980 @item --add-stdcall-alias
1981 If given, symbols with a stdcall suffix (@@@var{nn}) will be exported
1982 as-is and also with the suffix stripped.
1983 [This option is specific to the i386 PE targeted port of the linker]
1986 @item --base-file @var{file}
1987 Use @var{file} as the name of a file in which to save the base
1988 addresses of all the relocations needed for generating DLLs with
1990 [This is an i386 PE specific option]
1994 Create a DLL instead of a regular executable. You may also use
1995 @option{-shared} or specify a @code{LIBRARY} in a given @code{.def}
1997 [This option is specific to the i386 PE targeted port of the linker]
1999 @kindex --enable-stdcall-fixup
2000 @kindex --disable-stdcall-fixup
2001 @item --enable-stdcall-fixup
2002 @itemx --disable-stdcall-fixup
2003 If the link finds a symbol that it cannot resolve, it will attempt to
2004 do ``fuzzy linking'' by looking for another defined symbol that differs
2005 only in the format of the symbol name (cdecl vs stdcall) and will
2006 resolve that symbol by linking to the match. For example, the
2007 undefined symbol @code{_foo} might be linked to the function
2008 @code{_foo@@12}, or the undefined symbol @code{_bar@@16} might be linked
2009 to the function @code{_bar}. When the linker does this, it prints a
2010 warning, since it normally should have failed to link, but sometimes
2011 import libraries generated from third-party dlls may need this feature
2012 to be usable. If you specify @option{--enable-stdcall-fixup}, this
2013 feature is fully enabled and warnings are not printed. If you specify
2014 @option{--disable-stdcall-fixup}, this feature is disabled and such
2015 mismatches are considered to be errors.
2016 [This option is specific to the i386 PE targeted port of the linker]
2018 @cindex DLLs, creating
2019 @kindex --export-all-symbols
2020 @item --export-all-symbols
2021 If given, all global symbols in the objects used to build a DLL will
2022 be exported by the DLL. Note that this is the default if there
2023 otherwise wouldn't be any exported symbols. When symbols are
2024 explicitly exported via DEF files or implicitly exported via function
2025 attributes, the default is to not export anything else unless this
2026 option is given. Note that the symbols @code{DllMain@@12},
2027 @code{DllEntryPoint@@0}, @code{DllMainCRTStartup@@12}, and
2028 @code{impure_ptr} will not be automatically
2029 exported. Also, symbols imported from other DLLs will not be
2030 re-exported, nor will symbols specifying the DLL's internal layout
2031 such as those beginning with @code{_head_} or ending with
2032 @code{_iname}. In addition, no symbols from @code{libgcc},
2033 @code{libstd++}, @code{libmingw32}, or @code{crtX.o} will be exported.
2034 Symbols whose names begin with @code{__rtti_} or @code{__builtin_} will
2035 not be exported, to help with C++ DLLs. Finally, there is an
2036 extensive list of cygwin-private symbols that are not exported
2037 (obviously, this applies on when building DLLs for cygwin targets).
2038 These cygwin-excludes are: @code{_cygwin_dll_entry@@12},
2039 @code{_cygwin_crt0_common@@8}, @code{_cygwin_noncygwin_dll_entry@@12},
2040 @code{_fmode}, @code{_impure_ptr}, @code{cygwin_attach_dll},
2041 @code{cygwin_premain0}, @code{cygwin_premain1}, @code{cygwin_premain2},
2042 @code{cygwin_premain3}, and @code{environ}.
2043 [This option is specific to the i386 PE targeted port of the linker]
2045 @kindex --exclude-symbols
2046 @item --exclude-symbols @var{symbol},@var{symbol},...
2047 Specifies a list of symbols which should not be automatically
2048 exported. The symbol names may be delimited by commas or colons.
2049 [This option is specific to the i386 PE targeted port of the linker]
2051 @kindex --file-alignment
2052 @item --file-alignment
2053 Specify the file alignment. Sections in the file will always begin at
2054 file offsets which are multiples of this number. This defaults to
2056 [This option is specific to the i386 PE targeted port of the linker]
2060 @item --heap @var{reserve}
2061 @itemx --heap @var{reserve},@var{commit}
2062 Specify the amount of memory to reserve (and optionally commit) to be
2063 used as heap for this program. The default is 1Mb reserved, 4K
2065 [This option is specific to the i386 PE targeted port of the linker]
2068 @kindex --image-base
2069 @item --image-base @var{value}
2070 Use @var{value} as the base address of your program or dll. This is
2071 the lowest memory location that will be used when your program or dll
2072 is loaded. To reduce the need to relocate and improve performance of
2073 your dlls, each should have a unique base address and not overlap any
2074 other dlls. The default is 0x400000 for executables, and 0x10000000
2076 [This option is specific to the i386 PE targeted port of the linker]
2080 If given, the stdcall suffixes (@@@var{nn}) will be stripped from
2081 symbols before they are exported.
2082 [This option is specific to the i386 PE targeted port of the linker]
2084 @kindex --large-address-aware
2085 @item --large-address-aware
2086 If given, the appropriate bit in the ``Characteristics'' field of the COFF
2087 header is set to indicate that this executable supports virtual addresses
2088 greater than 2 gigabytes. This should be used in conjunction with the /3GB
2089 or /USERVA=@var{value} megabytes switch in the ``[operating systems]''
2090 section of the BOOT.INI. Otherwise, this bit has no effect.
2091 [This option is specific to PE targeted ports of the linker]
2093 @kindex --major-image-version
2094 @item --major-image-version @var{value}
2095 Sets the major number of the ``image version''. Defaults to 1.
2096 [This option is specific to the i386 PE targeted port of the linker]
2098 @kindex --major-os-version
2099 @item --major-os-version @var{value}
2100 Sets the major number of the ``os version''. Defaults to 4.
2101 [This option is specific to the i386 PE targeted port of the linker]
2103 @kindex --major-subsystem-version
2104 @item --major-subsystem-version @var{value}
2105 Sets the major number of the ``subsystem version''. Defaults to 4.
2106 [This option is specific to the i386 PE targeted port of the linker]
2108 @kindex --minor-image-version
2109 @item --minor-image-version @var{value}
2110 Sets the minor number of the ``image version''. Defaults to 0.
2111 [This option is specific to the i386 PE targeted port of the linker]
2113 @kindex --minor-os-version
2114 @item --minor-os-version @var{value}
2115 Sets the minor number of the ``os version''. Defaults to 0.
2116 [This option is specific to the i386 PE targeted port of the linker]
2118 @kindex --minor-subsystem-version
2119 @item --minor-subsystem-version @var{value}
2120 Sets the minor number of the ``subsystem version''. Defaults to 0.
2121 [This option is specific to the i386 PE targeted port of the linker]
2123 @cindex DEF files, creating
2124 @cindex DLLs, creating
2125 @kindex --output-def
2126 @item --output-def @var{file}
2127 The linker will create the file @var{file} which will contain a DEF
2128 file corresponding to the DLL the linker is generating. This DEF file
2129 (which should be called @code{*.def}) may be used to create an import
2130 library with @code{dlltool} or may be used as a reference to
2131 automatically or implicitly exported symbols.
2132 [This option is specific to the i386 PE targeted port of the linker]
2134 @cindex DLLs, creating
2135 @kindex --out-implib
2136 @item --out-implib @var{file}
2137 The linker will create the file @var{file} which will contain an
2138 import lib corresponding to the DLL the linker is generating. This
2139 import lib (which should be called @code{*.dll.a} or @code{*.a}
2140 may be used to link clients against the generated DLL; this behaviour
2141 makes it possible to skip a separate @code{dlltool} import library
2143 [This option is specific to the i386 PE targeted port of the linker]
2145 @kindex --enable-auto-image-base
2146 @item --enable-auto-image-base
2147 Automatically choose the image base for DLLs, unless one is specified
2148 using the @code{--image-base} argument. By using a hash generated
2149 from the dllname to create unique image bases for each DLL, in-memory
2150 collisions and relocations which can delay program execution are
2152 [This option is specific to the i386 PE targeted port of the linker]
2154 @kindex --disable-auto-image-base
2155 @item --disable-auto-image-base
2156 Do not automatically generate a unique image base. If there is no
2157 user-specified image base (@code{--image-base}) then use the platform
2159 [This option is specific to the i386 PE targeted port of the linker]
2161 @cindex DLLs, linking to
2162 @kindex --dll-search-prefix
2163 @item --dll-search-prefix @var{string}
2164 When linking dynamically to a dll without an import library,
2165 search for @code{<string><basename>.dll} in preference to
2166 @code{lib<basename>.dll}. This behaviour allows easy distinction
2167 between DLLs built for the various "subplatforms": native, cygwin,
2168 uwin, pw, etc. For instance, cygwin DLLs typically use
2169 @code{--dll-search-prefix=cyg}.
2170 [This option is specific to the i386 PE targeted port of the linker]
2172 @kindex --enable-auto-import
2173 @item --enable-auto-import
2174 Do sophisticated linking of @code{_symbol} to @code{__imp__symbol} for
2175 DATA imports from DLLs, and create the necessary thunking symbols when
2176 building the import libraries with those DATA exports. Note: Use of the
2177 'auto-import' extension will cause the text section of the image file
2178 to be made writable. This does not conform to the PE-COFF format
2179 specification published by Microsoft.
2181 Using 'auto-import' generally will 'just work' -- but sometimes you may
2184 "variable '<var>' can't be auto-imported. Please read the
2185 documentation for ld's @code{--enable-auto-import} for details."
2187 This message occurs when some (sub)expression accesses an address
2188 ultimately given by the sum of two constants (Win32 import tables only
2189 allow one). Instances where this may occur include accesses to member
2190 fields of struct variables imported from a DLL, as well as using a
2191 constant index into an array variable imported from a DLL. Any
2192 multiword variable (arrays, structs, long long, etc) may trigger
2193 this error condition. However, regardless of the exact data type
2194 of the offending exported variable, ld will always detect it, issue
2195 the warning, and exit.
2197 There are several ways to address this difficulty, regardless of the
2198 data type of the exported variable:
2200 One way is to use --enable-runtime-pseudo-reloc switch. This leaves the task
2201 of adjusting references in your client code for runtime environment, so
2202 this method works only when runtime environment supports this feature.
2204 A second solution is to force one of the 'constants' to be a variable --
2205 that is, unknown and un-optimizable at compile time. For arrays,
2206 there are two possibilities: a) make the indexee (the array's address)
2207 a variable, or b) make the 'constant' index a variable. Thus:
2210 extern type extern_array[];
2212 @{ volatile type *t=extern_array; t[1] @}
2218 extern type extern_array[];
2220 @{ volatile int t=1; extern_array[t] @}
2223 For structs (and most other multiword data types) the only option
2224 is to make the struct itself (or the long long, or the ...) variable:
2227 extern struct s extern_struct;
2228 extern_struct.field -->
2229 @{ volatile struct s *t=&extern_struct; t->field @}
2235 extern long long extern_ll;
2237 @{ volatile long long * local_ll=&extern_ll; *local_ll @}
2240 A third method of dealing with this difficulty is to abandon
2241 'auto-import' for the offending symbol and mark it with
2242 @code{__declspec(dllimport)}. However, in practise that
2243 requires using compile-time #defines to indicate whether you are
2244 building a DLL, building client code that will link to the DLL, or
2245 merely building/linking to a static library. In making the choice
2246 between the various methods of resolving the 'direct address with
2247 constant offset' problem, you should consider typical real-world usage:
2255 void main(int argc, char **argv)@{
2256 printf("%d\n",arr[1]);
2266 void main(int argc, char **argv)@{
2267 /* This workaround is for win32 and cygwin; do not "optimize" */
2268 volatile int *parr = arr;
2269 printf("%d\n",parr[1]);
2276 /* Note: auto-export is assumed (no __declspec(dllexport)) */
2277 #if (defined(_WIN32) || defined(__CYGWIN__)) && \
2278 !(defined(FOO_BUILD_DLL) || defined(FOO_STATIC))
2279 #define FOO_IMPORT __declspec(dllimport)
2283 extern FOO_IMPORT int arr[];
2286 void main(int argc, char **argv)@{
2287 printf("%d\n",arr[1]);
2291 A fourth way to avoid this problem is to re-code your
2292 library to use a functional interface rather than a data interface
2293 for the offending variables (e.g. set_foo() and get_foo() accessor
2295 [This option is specific to the i386 PE targeted port of the linker]
2297 @kindex --disable-auto-import
2298 @item --disable-auto-import
2299 Do not attempt to do sophisticated linking of @code{_symbol} to
2300 @code{__imp__symbol} for DATA imports from DLLs.
2301 [This option is specific to the i386 PE targeted port of the linker]
2303 @kindex --enable-runtime-pseudo-reloc
2304 @item --enable-runtime-pseudo-reloc
2305 If your code contains expressions described in --enable-auto-import section,
2306 that is, DATA imports from DLL with non-zero offset, this switch will create
2307 a vector of 'runtime pseudo relocations' which can be used by runtime
2308 environment to adjust references to such data in your client code.
2309 [This option is specific to the i386 PE targeted port of the linker]
2311 @kindex --disable-runtime-pseudo-reloc
2312 @item --disable-runtime-pseudo-reloc
2313 Do not create pseudo relocations for non-zero offset DATA imports from
2314 DLLs. This is the default.
2315 [This option is specific to the i386 PE targeted port of the linker]
2317 @kindex --enable-extra-pe-debug
2318 @item --enable-extra-pe-debug
2319 Show additional debug info related to auto-import symbol thunking.
2320 [This option is specific to the i386 PE targeted port of the linker]
2322 @kindex --section-alignment
2323 @item --section-alignment
2324 Sets the section alignment. Sections in memory will always begin at
2325 addresses which are a multiple of this number. Defaults to 0x1000.
2326 [This option is specific to the i386 PE targeted port of the linker]
2330 @item --stack @var{reserve}
2331 @itemx --stack @var{reserve},@var{commit}
2332 Specify the amount of memory to reserve (and optionally commit) to be
2333 used as stack for this program. The default is 2Mb reserved, 4K
2335 [This option is specific to the i386 PE targeted port of the linker]
2338 @item --subsystem @var{which}
2339 @itemx --subsystem @var{which}:@var{major}
2340 @itemx --subsystem @var{which}:@var{major}.@var{minor}
2341 Specifies the subsystem under which your program will execute. The
2342 legal values for @var{which} are @code{native}, @code{windows},
2343 @code{console}, @code{posix}, and @code{xbox}. You may optionally set
2344 the subsystem version also. Numeric values are also accepted for
2346 [This option is specific to the i386 PE targeted port of the linker]
2353 @subsection Options specific to Motorola 68HC11 and 68HC12 targets
2355 @c man begin OPTIONS
2357 The 68HC11 and 68HC12 linkers support specific options to control the
2358 memory bank switching mapping and trampoline code generation.
2362 @kindex --no-trampoline
2363 @item --no-trampoline
2364 This option disables the generation of trampoline. By default a trampoline
2365 is generated for each far function which is called using a @code{jsr}
2366 instruction (this happens when a pointer to a far function is taken).
2368 @kindex --bank-window
2369 @item --bank-window @var{name}
2370 This option indicates to the linker the name of the memory region in
2371 the @samp{MEMORY} specification that describes the memory bank window.
2372 The definition of such region is then used by the linker to compute
2373 paging and addresses within the memory window.
2382 @section Environment Variables
2384 @c man begin ENVIRONMENT
2386 You can change the behaviour of @command{ld} with the environment variables
2387 @ifclear SingleFormat
2390 @code{LDEMULATION} and @code{COLLECT_NO_DEMANGLE}.
2392 @ifclear SingleFormat
2394 @cindex default input format
2395 @code{GNUTARGET} determines the input-file object format if you don't
2396 use @samp{-b} (or its synonym @samp{--format}). Its value should be one
2397 of the BFD names for an input format (@pxref{BFD}). If there is no
2398 @code{GNUTARGET} in the environment, @command{ld} uses the natural format
2399 of the target. If @code{GNUTARGET} is set to @code{default} then BFD
2400 attempts to discover the input format by examining binary input files;
2401 this method often succeeds, but there are potential ambiguities, since
2402 there is no method of ensuring that the magic number used to specify
2403 object-file formats is unique. However, the configuration procedure for
2404 BFD on each system places the conventional format for that system first
2405 in the search-list, so ambiguities are resolved in favor of convention.
2409 @cindex default emulation
2410 @cindex emulation, default
2411 @code{LDEMULATION} determines the default emulation if you don't use the
2412 @samp{-m} option. The emulation can affect various aspects of linker
2413 behaviour, particularly the default linker script. You can list the
2414 available emulations with the @samp{--verbose} or @samp{-V} options. If
2415 the @samp{-m} option is not used, and the @code{LDEMULATION} environment
2416 variable is not defined, the default emulation depends upon how the
2417 linker was configured.
2419 @kindex COLLECT_NO_DEMANGLE
2420 @cindex demangling, default
2421 Normally, the linker will default to demangling symbols. However, if
2422 @code{COLLECT_NO_DEMANGLE} is set in the environment, then it will
2423 default to not demangling symbols. This environment variable is used in
2424 a similar fashion by the @code{gcc} linker wrapper program. The default
2425 may be overridden by the @samp{--demangle} and @samp{--no-demangle}
2432 @chapter Linker Scripts
2435 @cindex linker scripts
2436 @cindex command files
2437 Every link is controlled by a @dfn{linker script}. This script is
2438 written in the linker command language.
2440 The main purpose of the linker script is to describe how the sections in
2441 the input files should be mapped into the output file, and to control
2442 the memory layout of the output file. Most linker scripts do nothing
2443 more than this. However, when necessary, the linker script can also
2444 direct the linker to perform many other operations, using the commands
2447 The linker always uses a linker script. If you do not supply one
2448 yourself, the linker will use a default script that is compiled into the
2449 linker executable. You can use the @samp{--verbose} command line option
2450 to display the default linker script. Certain command line options,
2451 such as @samp{-r} or @samp{-N}, will affect the default linker script.
2453 You may supply your own linker script by using the @samp{-T} command
2454 line option. When you do this, your linker script will replace the
2455 default linker script.
2457 You may also use linker scripts implicitly by naming them as input files
2458 to the linker, as though they were files to be linked. @xref{Implicit
2462 * Basic Script Concepts:: Basic Linker Script Concepts
2463 * Script Format:: Linker Script Format
2464 * Simple Example:: Simple Linker Script Example
2465 * Simple Commands:: Simple Linker Script Commands
2466 * Assignments:: Assigning Values to Symbols
2467 * SECTIONS:: SECTIONS Command
2468 * MEMORY:: MEMORY Command
2469 * PHDRS:: PHDRS Command
2470 * VERSION:: VERSION Command
2471 * Expressions:: Expressions in Linker Scripts
2472 * Implicit Linker Scripts:: Implicit Linker Scripts
2475 @node Basic Script Concepts
2476 @section Basic Linker Script Concepts
2477 @cindex linker script concepts
2478 We need to define some basic concepts and vocabulary in order to
2479 describe the linker script language.
2481 The linker combines input files into a single output file. The output
2482 file and each input file are in a special data format known as an
2483 @dfn{object file format}. Each file is called an @dfn{object file}.
2484 The output file is often called an @dfn{executable}, but for our
2485 purposes we will also call it an object file. Each object file has,
2486 among other things, a list of @dfn{sections}. We sometimes refer to a
2487 section in an input file as an @dfn{input section}; similarly, a section
2488 in the output file is an @dfn{output section}.
2490 Each section in an object file has a name and a size. Most sections
2491 also have an associated block of data, known as the @dfn{section
2492 contents}. A section may be marked as @dfn{loadable}, which mean that
2493 the contents should be loaded into memory when the output file is run.
2494 A section with no contents may be @dfn{allocatable}, which means that an
2495 area in memory should be set aside, but nothing in particular should be
2496 loaded there (in some cases this memory must be zeroed out). A section
2497 which is neither loadable nor allocatable typically contains some sort
2498 of debugging information.
2500 Every loadable or allocatable output section has two addresses. The
2501 first is the @dfn{VMA}, or virtual memory address. This is the address
2502 the section will have when the output file is run. The second is the
2503 @dfn{LMA}, or load memory address. This is the address at which the
2504 section will be loaded. In most cases the two addresses will be the
2505 same. An example of when they might be different is when a data section
2506 is loaded into ROM, and then copied into RAM when the program starts up
2507 (this technique is often used to initialize global variables in a ROM
2508 based system). In this case the ROM address would be the LMA, and the
2509 RAM address would be the VMA.
2511 You can see the sections in an object file by using the @code{objdump}
2512 program with the @samp{-h} option.
2514 Every object file also has a list of @dfn{symbols}, known as the
2515 @dfn{symbol table}. A symbol may be defined or undefined. Each symbol
2516 has a name, and each defined symbol has an address, among other
2517 information. If you compile a C or C++ program into an object file, you
2518 will get a defined symbol for every defined function and global or
2519 static variable. Every undefined function or global variable which is
2520 referenced in the input file will become an undefined symbol.
2522 You can see the symbols in an object file by using the @code{nm}
2523 program, or by using the @code{objdump} program with the @samp{-t}
2527 @section Linker Script Format
2528 @cindex linker script format
2529 Linker scripts are text files.
2531 You write a linker script as a series of commands. Each command is
2532 either a keyword, possibly followed by arguments, or an assignment to a
2533 symbol. You may separate commands using semicolons. Whitespace is
2536 Strings such as file or format names can normally be entered directly.
2537 If the file name contains a character such as a comma which would
2538 otherwise serve to separate file names, you may put the file name in
2539 double quotes. There is no way to use a double quote character in a
2542 You may include comments in linker scripts just as in C, delimited by
2543 @samp{/*} and @samp{*/}. As in C, comments are syntactically equivalent
2546 @node Simple Example
2547 @section Simple Linker Script Example
2548 @cindex linker script example
2549 @cindex example of linker script
2550 Many linker scripts are fairly simple.
2552 The simplest possible linker script has just one command:
2553 @samp{SECTIONS}. You use the @samp{SECTIONS} command to describe the
2554 memory layout of the output file.
2556 The @samp{SECTIONS} command is a powerful command. Here we will
2557 describe a simple use of it. Let's assume your program consists only of
2558 code, initialized data, and uninitialized data. These will be in the
2559 @samp{.text}, @samp{.data}, and @samp{.bss} sections, respectively.
2560 Let's assume further that these are the only sections which appear in
2563 For this example, let's say that the code should be loaded at address
2564 0x10000, and that the data should start at address 0x8000000. Here is a
2565 linker script which will do that:
2570 .text : @{ *(.text) @}
2572 .data : @{ *(.data) @}
2573 .bss : @{ *(.bss) @}
2577 You write the @samp{SECTIONS} command as the keyword @samp{SECTIONS},
2578 followed by a series of symbol assignments and output section
2579 descriptions enclosed in curly braces.
2581 The first line inside the @samp{SECTIONS} command of the above example
2582 sets the value of the special symbol @samp{.}, which is the location
2583 counter. If you do not specify the address of an output section in some
2584 other way (other ways are described later), the address is set from the
2585 current value of the location counter. The location counter is then
2586 incremented by the size of the output section. At the start of the
2587 @samp{SECTIONS} command, the location counter has the value @samp{0}.
2589 The second line defines an output section, @samp{.text}. The colon is
2590 required syntax which may be ignored for now. Within the curly braces
2591 after the output section name, you list the names of the input sections
2592 which should be placed into this output section. The @samp{*} is a
2593 wildcard which matches any file name. The expression @samp{*(.text)}
2594 means all @samp{.text} input sections in all input files.
2596 Since the location counter is @samp{0x10000} when the output section
2597 @samp{.text} is defined, the linker will set the address of the
2598 @samp{.text} section in the output file to be @samp{0x10000}.
2600 The remaining lines define the @samp{.data} and @samp{.bss} sections in
2601 the output file. The linker will place the @samp{.data} output section
2602 at address @samp{0x8000000}. After the linker places the @samp{.data}
2603 output section, the value of the location counter will be
2604 @samp{0x8000000} plus the size of the @samp{.data} output section. The
2605 effect is that the linker will place the @samp{.bss} output section
2606 immediately after the @samp{.data} output section in memory.
2608 The linker will ensure that each output section has the required
2609 alignment, by increasing the location counter if necessary. In this
2610 example, the specified addresses for the @samp{.text} and @samp{.data}
2611 sections will probably satisfy any alignment constraints, but the linker
2612 may have to create a small gap between the @samp{.data} and @samp{.bss}
2615 That's it! That's a simple and complete linker script.
2617 @node Simple Commands
2618 @section Simple Linker Script Commands
2619 @cindex linker script simple commands
2620 In this section we describe the simple linker script commands.
2623 * Entry Point:: Setting the entry point
2624 * File Commands:: Commands dealing with files
2625 @ifclear SingleFormat
2626 * Format Commands:: Commands dealing with object file formats
2629 * Miscellaneous Commands:: Other linker script commands
2633 @subsection Setting the Entry Point
2634 @kindex ENTRY(@var{symbol})
2635 @cindex start of execution
2636 @cindex first instruction
2638 The first instruction to execute in a program is called the @dfn{entry
2639 point}. You can use the @code{ENTRY} linker script command to set the
2640 entry point. The argument is a symbol name:
2645 There are several ways to set the entry point. The linker will set the
2646 entry point by trying each of the following methods in order, and
2647 stopping when one of them succeeds:
2650 the @samp{-e} @var{entry} command-line option;
2652 the @code{ENTRY(@var{symbol})} command in a linker script;
2654 the value of the symbol @code{start}, if defined;
2656 the address of the first byte of the @samp{.text} section, if present;
2658 The address @code{0}.
2662 @subsection Commands Dealing with Files
2663 @cindex linker script file commands
2664 Several linker script commands deal with files.
2667 @item INCLUDE @var{filename}
2668 @kindex INCLUDE @var{filename}
2669 @cindex including a linker script
2670 Include the linker script @var{filename} at this point. The file will
2671 be searched for in the current directory, and in any directory specified
2672 with the @option{-L} option. You can nest calls to @code{INCLUDE} up to
2675 @item INPUT(@var{file}, @var{file}, @dots{})
2676 @itemx INPUT(@var{file} @var{file} @dots{})
2677 @kindex INPUT(@var{files})
2678 @cindex input files in linker scripts
2679 @cindex input object files in linker scripts
2680 @cindex linker script input object files
2681 The @code{INPUT} command directs the linker to include the named files
2682 in the link, as though they were named on the command line.
2684 For example, if you always want to include @file{subr.o} any time you do
2685 a link, but you can't be bothered to put it on every link command line,
2686 then you can put @samp{INPUT (subr.o)} in your linker script.
2688 In fact, if you like, you can list all of your input files in the linker
2689 script, and then invoke the linker with nothing but a @samp{-T} option.
2691 In case a @dfn{sysroot prefix} is configured, and the filename starts
2692 with the @samp{/} character, and the script being processed was
2693 located inside the @dfn{sysroot prefix}, the filename will be looked
2694 for in the @dfn{sysroot prefix}. Otherwise, the linker will try to
2695 open the file in the current directory. If it is not found, the
2696 linker will search through the archive library search path. See the
2697 description of @samp{-L} in @ref{Options,,Command Line Options}.
2699 If you use @samp{INPUT (-l@var{file})}, @command{ld} will transform the
2700 name to @code{lib@var{file}.a}, as with the command line argument
2703 When you use the @code{INPUT} command in an implicit linker script, the
2704 files will be included in the link at the point at which the linker
2705 script file is included. This can affect archive searching.
2707 @item GROUP(@var{file}, @var{file}, @dots{})
2708 @itemx GROUP(@var{file} @var{file} @dots{})
2709 @kindex GROUP(@var{files})
2710 @cindex grouping input files
2711 The @code{GROUP} command is like @code{INPUT}, except that the named
2712 files should all be archives, and they are searched repeatedly until no
2713 new undefined references are created. See the description of @samp{-(}
2714 in @ref{Options,,Command Line Options}.
2716 @item AS_NEEDED(@var{file}, @var{file}, @dots{})
2717 @itemx AS_NEEDED(@var{file} @var{file} @dots{})
2718 @kindex AS_NEEDED(@var{files})
2719 This construct can appear only inside of the @code{INPUT} or @code{GROUP}
2720 commands, among other filenames. The files listed will be handled
2721 as if they appear directly in the @code{INPUT} or @code{GROUP} commands,
2722 with the exception of ELF shared libraries, that will be added only
2723 when they are actually needed. This construct essentially enables
2724 @option{--as-needed} option for all the files listed inside of it
2725 and restores previous @option{--as-needed} resp. @option{--no-as-needed}
2728 @item OUTPUT(@var{filename})
2729 @kindex OUTPUT(@var{filename})
2730 @cindex output file name in linker script
2731 The @code{OUTPUT} command names the output file. Using
2732 @code{OUTPUT(@var{filename})} in the linker script is exactly like using
2733 @samp{-o @var{filename}} on the command line (@pxref{Options,,Command
2734 Line Options}). If both are used, the command line option takes
2737 You can use the @code{OUTPUT} command to define a default name for the
2738 output file other than the usual default of @file{a.out}.
2740 @item SEARCH_DIR(@var{path})
2741 @kindex SEARCH_DIR(@var{path})
2742 @cindex library search path in linker script
2743 @cindex archive search path in linker script
2744 @cindex search path in linker script
2745 The @code{SEARCH_DIR} command adds @var{path} to the list of paths where
2746 @command{ld} looks for archive libraries. Using
2747 @code{SEARCH_DIR(@var{path})} is exactly like using @samp{-L @var{path}}
2748 on the command line (@pxref{Options,,Command Line Options}). If both
2749 are used, then the linker will search both paths. Paths specified using
2750 the command line option are searched first.
2752 @item STARTUP(@var{filename})
2753 @kindex STARTUP(@var{filename})
2754 @cindex first input file
2755 The @code{STARTUP} command is just like the @code{INPUT} command, except
2756 that @var{filename} will become the first input file to be linked, as
2757 though it were specified first on the command line. This may be useful
2758 when using a system in which the entry point is always the start of the
2762 @ifclear SingleFormat
2763 @node Format Commands
2764 @subsection Commands Dealing with Object File Formats
2765 A couple of linker script commands deal with object file formats.
2768 @item OUTPUT_FORMAT(@var{bfdname})
2769 @itemx OUTPUT_FORMAT(@var{default}, @var{big}, @var{little})
2770 @kindex OUTPUT_FORMAT(@var{bfdname})
2771 @cindex output file format in linker script
2772 The @code{OUTPUT_FORMAT} command names the BFD format to use for the
2773 output file (@pxref{BFD}). Using @code{OUTPUT_FORMAT(@var{bfdname})} is
2774 exactly like using @samp{--oformat @var{bfdname}} on the command line
2775 (@pxref{Options,,Command Line Options}). If both are used, the command
2776 line option takes precedence.
2778 You can use @code{OUTPUT_FORMAT} with three arguments to use different
2779 formats based on the @samp{-EB} and @samp{-EL} command line options.
2780 This permits the linker script to set the output format based on the
2783 If neither @samp{-EB} nor @samp{-EL} are used, then the output format
2784 will be the first argument, @var{default}. If @samp{-EB} is used, the
2785 output format will be the second argument, @var{big}. If @samp{-EL} is
2786 used, the output format will be the third argument, @var{little}.
2788 For example, the default linker script for the MIPS ELF target uses this
2791 OUTPUT_FORMAT(elf32-bigmips, elf32-bigmips, elf32-littlemips)
2793 This says that the default format for the output file is
2794 @samp{elf32-bigmips}, but if the user uses the @samp{-EL} command line
2795 option, the output file will be created in the @samp{elf32-littlemips}
2798 @item TARGET(@var{bfdname})
2799 @kindex TARGET(@var{bfdname})
2800 @cindex input file format in linker script
2801 The @code{TARGET} command names the BFD format to use when reading input
2802 files. It affects subsequent @code{INPUT} and @code{GROUP} commands.
2803 This command is like using @samp{-b @var{bfdname}} on the command line
2804 (@pxref{Options,,Command Line Options}). If the @code{TARGET} command
2805 is used but @code{OUTPUT_FORMAT} is not, then the last @code{TARGET}
2806 command is also used to set the format for the output file. @xref{BFD}.
2810 @node Miscellaneous Commands
2811 @subsection Other Linker Script Commands
2812 There are a few other linker scripts commands.
2815 @item ASSERT(@var{exp}, @var{message})
2817 @cindex assertion in linker script
2818 Ensure that @var{exp} is non-zero. If it is zero, then exit the linker
2819 with an error code, and print @var{message}.
2821 @item EXTERN(@var{symbol} @var{symbol} @dots{})
2823 @cindex undefined symbol in linker script
2824 Force @var{symbol} to be entered in the output file as an undefined
2825 symbol. Doing this may, for example, trigger linking of additional
2826 modules from standard libraries. You may list several @var{symbol}s for
2827 each @code{EXTERN}, and you may use @code{EXTERN} multiple times. This
2828 command has the same effect as the @samp{-u} command-line option.
2830 @item FORCE_COMMON_ALLOCATION
2831 @kindex FORCE_COMMON_ALLOCATION
2832 @cindex common allocation in linker script
2833 This command has the same effect as the @samp{-d} command-line option:
2834 to make @command{ld} assign space to common symbols even if a relocatable
2835 output file is specified (@samp{-r}).
2837 @item INHIBIT_COMMON_ALLOCATION
2838 @kindex INHIBIT_COMMON_ALLOCATION
2839 @cindex common allocation in linker script
2840 This command has the same effect as the @samp{--no-define-common}
2841 command-line option: to make @code{ld} omit the assignment of addresses
2842 to common symbols even for a non-relocatable output file.
2844 @item NOCROSSREFS(@var{section} @var{section} @dots{})
2845 @kindex NOCROSSREFS(@var{sections})
2846 @cindex cross references
2847 This command may be used to tell @command{ld} to issue an error about any
2848 references among certain output sections.
2850 In certain types of programs, particularly on embedded systems when
2851 using overlays, when one section is loaded into memory, another section
2852 will not be. Any direct references between the two sections would be
2853 errors. For example, it would be an error if code in one section called
2854 a function defined in the other section.
2856 The @code{NOCROSSREFS} command takes a list of output section names. If
2857 @command{ld} detects any cross references between the sections, it reports
2858 an error and returns a non-zero exit status. Note that the
2859 @code{NOCROSSREFS} command uses output section names, not input section
2862 @ifclear SingleFormat
2863 @item OUTPUT_ARCH(@var{bfdarch})
2864 @kindex OUTPUT_ARCH(@var{bfdarch})
2865 @cindex machine architecture
2866 @cindex architecture
2867 Specify a particular output machine architecture. The argument is one
2868 of the names used by the BFD library (@pxref{BFD}). You can see the
2869 architecture of an object file by using the @code{objdump} program with
2870 the @samp{-f} option.
2875 @section Assigning Values to Symbols
2876 @cindex assignment in scripts
2877 @cindex symbol definition, scripts
2878 @cindex variables, defining
2879 You may assign a value to a symbol in a linker script. This will define
2880 the symbol and place it into the symbol table with a global scope.
2883 * Simple Assignments:: Simple Assignments
2885 * PROVIDE_HIDDEN:: PROVIDE_HIDDEN
2886 * Source Code Reference:: How to use a linker script defined symbol in source code
2889 @node Simple Assignments
2890 @subsection Simple Assignments
2892 You may assign to a symbol using any of the C assignment operators:
2895 @item @var{symbol} = @var{expression} ;
2896 @itemx @var{symbol} += @var{expression} ;
2897 @itemx @var{symbol} -= @var{expression} ;
2898 @itemx @var{symbol} *= @var{expression} ;
2899 @itemx @var{symbol} /= @var{expression} ;
2900 @itemx @var{symbol} <<= @var{expression} ;
2901 @itemx @var{symbol} >>= @var{expression} ;
2902 @itemx @var{symbol} &= @var{expression} ;
2903 @itemx @var{symbol} |= @var{expression} ;
2906 The first case will define @var{symbol} to the value of
2907 @var{expression}. In the other cases, @var{symbol} must already be
2908 defined, and the value will be adjusted accordingly.
2910 The special symbol name @samp{.} indicates the location counter. You
2911 may only use this within a @code{SECTIONS} command. @xref{Location Counter}.
2913 The semicolon after @var{expression} is required.
2915 Expressions are defined below; see @ref{Expressions}.
2917 You may write symbol assignments as commands in their own right, or as
2918 statements within a @code{SECTIONS} command, or as part of an output
2919 section description in a @code{SECTIONS} command.
2921 The section of the symbol will be set from the section of the
2922 expression; for more information, see @ref{Expression Section}.
2924 Here is an example showing the three different places that symbol
2925 assignments may be used:
2936 _bdata = (. + 3) & ~ 3;
2937 .data : @{ *(.data) @}
2941 In this example, the symbol @samp{floating_point} will be defined as
2942 zero. The symbol @samp{_etext} will be defined as the address following
2943 the last @samp{.text} input section. The symbol @samp{_bdata} will be
2944 defined as the address following the @samp{.text} output section aligned
2945 upward to a 4 byte boundary.
2950 In some cases, it is desirable for a linker script to define a symbol
2951 only if it is referenced and is not defined by any object included in
2952 the link. For example, traditional linkers defined the symbol
2953 @samp{etext}. However, ANSI C requires that the user be able to use
2954 @samp{etext} as a function name without encountering an error. The
2955 @code{PROVIDE} keyword may be used to define a symbol, such as
2956 @samp{etext}, only if it is referenced but not defined. The syntax is
2957 @code{PROVIDE(@var{symbol} = @var{expression})}.
2959 Here is an example of using @code{PROVIDE} to define @samp{etext}:
2972 In this example, if the program defines @samp{_etext} (with a leading
2973 underscore), the linker will give a multiple definition error. If, on
2974 the other hand, the program defines @samp{etext} (with no leading
2975 underscore), the linker will silently use the definition in the program.
2976 If the program references @samp{etext} but does not define it, the
2977 linker will use the definition in the linker script.
2979 @node PROVIDE_HIDDEN
2980 @subsection PROVIDE_HIDDEN
2981 @cindex PROVIDE_HIDDEN
2982 Similar to @code{PROVIDE}. For ELF targeted ports, the symbol will be
2983 hidden and won't be exported.
2985 @node Source Code Reference
2986 @subsection Source Code Reference
2988 Accessing a linker script defined variable from source code is not
2989 intuitive. In particular a linker script symbol is not equivalent to
2990 a variable declaration in a high level language, it is instead a
2991 symbol that does not have a value.
2993 Before going further, it is important to note that compilers often
2994 transform names in the source code into different names when they are
2995 stored in the symbol table. For example, Fortran compilers commonly
2996 prepend or append an underscore, and C++ performs extensive @samp{name
2997 mangling}. Therefore there might be a discrepancy between the name
2998 of a variable as it is used in source code and the name of the same
2999 variable as it is defined in a linker script. For example in C a
3000 linker script variable might be referred to as:
3006 But in the linker script it might be defined as:
3012 In the remaining examples however it is assumed that no name
3013 transformation has taken place.
3015 When a symbol is declared in a high level language such as C, two
3016 things happen. The first is that the compiler reserves enough space
3017 in the program's memory to hold the @emph{value} of the symbol. The
3018 second is that the compiler creates an entry in the program's symbol
3019 table which holds the symbol's @emph{address}. ie the symbol table
3020 contains the address of the block of memory holding the symbol's
3021 value. So for example the following C declaration, at file scope:
3027 creates a entry called @samp{foo} in the symbol table. This entry
3028 holds the address of an @samp{int} sized block of memory where the
3029 number 1000 is initially stored.
3031 When a program references a symbol the compiler generates code that
3032 first accesses the symbol table to find the address of the symbol's
3033 memory block and then code to read the value from that memory block.
3040 looks up the symbol @samp{foo} in the symbol table, gets the address
3041 associated with this symbol and then writes the value 1 into that
3048 looks up the symbol @samp{foo} in the symbol table, gets it address
3049 and then copies this address into the block of memory associated with
3050 the variable @samp{a}.
3052 Linker scripts symbol declarations, by contrast, create an entry in
3053 the symbol table but do not assign any memory to them. Thus they are
3054 an address without a value. So for example the linker script definition:
3060 creates an entry in the symbol table called @samp{foo} which holds
3061 the address of memory location 1000, but nothing special is stored at
3062 address 1000. This means that you cannot access the @emph{value} of a
3063 linker script defined symbol - it has no value - all you can do is
3064 access the @emph{address} of a linker script defined symbol.
3066 Hence when you are using a linker script defined symbol in source code
3067 you should always take the address of the symbol, and never attempt to
3068 use its value. For example suppose you want to copy the contents of a
3069 section of memory called .ROM into a section called .FLASH and the
3070 linker script contains these declarations:
3074 start_of_ROM = .ROM;
3075 end_of_ROM = .ROM + sizeof (.ROM) - 1;
3076 start_of_FLASH = .FLASH;
3080 Then the C source code to perform the copy would be:
3084 extern char start_of_ROM, end_of_ROM, start_of_FLASH;
3086 memcpy (& start_of_FLASH, & start_of_ROM, & end_of_ROM - & start_of_ROM);
3090 Note the use of the @samp{&} operators. These are correct.
3093 @section SECTIONS Command
3095 The @code{SECTIONS} command tells the linker how to map input sections
3096 into output sections, and how to place the output sections in memory.
3098 The format of the @code{SECTIONS} command is:
3102 @var{sections-command}
3103 @var{sections-command}
3108 Each @var{sections-command} may of be one of the following:
3112 an @code{ENTRY} command (@pxref{Entry Point,,Entry command})
3114 a symbol assignment (@pxref{Assignments})
3116 an output section description
3118 an overlay description
3121 The @code{ENTRY} command and symbol assignments are permitted inside the
3122 @code{SECTIONS} command for convenience in using the location counter in
3123 those commands. This can also make the linker script easier to
3124 understand because you can use those commands at meaningful points in
3125 the layout of the output file.
3127 Output section descriptions and overlay descriptions are described
3130 If you do not use a @code{SECTIONS} command in your linker script, the
3131 linker will place each input section into an identically named output
3132 section in the order that the sections are first encountered in the
3133 input files. If all input sections are present in the first file, for
3134 example, the order of sections in the output file will match the order
3135 in the first input file. The first section will be at address zero.
3138 * Output Section Description:: Output section description
3139 * Output Section Name:: Output section name
3140 * Output Section Address:: Output section address
3141 * Input Section:: Input section description
3142 * Output Section Data:: Output section data
3143 * Output Section Keywords:: Output section keywords
3144 * Output Section Discarding:: Output section discarding
3145 * Output Section Attributes:: Output section attributes
3146 * Overlay Description:: Overlay description
3149 @node Output Section Description
3150 @subsection Output Section Description
3151 The full description of an output section looks like this:
3154 @var{section} [@var{address}] [(@var{type})] :
3155 [AT(@var{lma})] [ALIGN(@var{section_align})] [SUBALIGN(@var{subsection_align})]
3157 @var{output-section-command}
3158 @var{output-section-command}
3160 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}]
3164 Most output sections do not use most of the optional section attributes.
3166 The whitespace around @var{section} is required, so that the section
3167 name is unambiguous. The colon and the curly braces are also required.
3168 The line breaks and other white space are optional.
3170 Each @var{output-section-command} may be one of the following:
3174 a symbol assignment (@pxref{Assignments})
3176 an input section description (@pxref{Input Section})
3178 data values to include directly (@pxref{Output Section Data})
3180 a special output section keyword (@pxref{Output Section Keywords})
3183 @node Output Section Name
3184 @subsection Output Section Name
3185 @cindex name, section
3186 @cindex section name
3187 The name of the output section is @var{section}. @var{section} must
3188 meet the constraints of your output format. In formats which only
3189 support a limited number of sections, such as @code{a.out}, the name
3190 must be one of the names supported by the format (@code{a.out}, for
3191 example, allows only @samp{.text}, @samp{.data} or @samp{.bss}). If the
3192 output format supports any number of sections, but with numbers and not
3193 names (as is the case for Oasys), the name should be supplied as a
3194 quoted numeric string. A section name may consist of any sequence of
3195 characters, but a name which contains any unusual characters such as
3196 commas must be quoted.
3198 The output section name @samp{/DISCARD/} is special; @ref{Output Section
3201 @node Output Section Address
3202 @subsection Output Section Address
3203 @cindex address, section
3204 @cindex section address
3205 The @var{address} is an expression for the VMA (the virtual memory
3206 address) of the output section. If you do not provide @var{address},
3207 the linker will set it based on @var{region} if present, or otherwise
3208 based on the current value of the location counter.
3210 If you provide @var{address}, the address of the output section will be
3211 set to precisely that. If you provide neither @var{address} nor
3212 @var{region}, then the address of the output section will be set to the
3213 current value of the location counter aligned to the alignment
3214 requirements of the output section. The alignment requirement of the
3215 output section is the strictest alignment of any input section contained
3216 within the output section.
3220 .text . : @{ *(.text) @}
3225 .text : @{ *(.text) @}
3228 are subtly different. The first will set the address of the
3229 @samp{.text} output section to the current value of the location
3230 counter. The second will set it to the current value of the location
3231 counter aligned to the strictest alignment of a @samp{.text} input
3234 The @var{address} may be an arbitrary expression; @ref{Expressions}.
3235 For example, if you want to align the section on a 0x10 byte boundary,
3236 so that the lowest four bits of the section address are zero, you could
3237 do something like this:
3239 .text ALIGN(0x10) : @{ *(.text) @}
3242 This works because @code{ALIGN} returns the current location counter
3243 aligned upward to the specified value.
3245 Specifying @var{address} for a section will change the value of the
3249 @subsection Input Section Description
3250 @cindex input sections
3251 @cindex mapping input sections to output sections
3252 The most common output section command is an input section description.
3254 The input section description is the most basic linker script operation.
3255 You use output sections to tell the linker how to lay out your program
3256 in memory. You use input section descriptions to tell the linker how to
3257 map the input files into your memory layout.
3260 * Input Section Basics:: Input section basics
3261 * Input Section Wildcards:: Input section wildcard patterns
3262 * Input Section Common:: Input section for common symbols
3263 * Input Section Keep:: Input section and garbage collection
3264 * Input Section Example:: Input section example
3267 @node Input Section Basics
3268 @subsubsection Input Section Basics
3269 @cindex input section basics
3270 An input section description consists of a file name optionally followed
3271 by a list of section names in parentheses.
3273 The file name and the section name may be wildcard patterns, which we
3274 describe further below (@pxref{Input Section Wildcards}).
3276 The most common input section description is to include all input
3277 sections with a particular name in the output section. For example, to
3278 include all input @samp{.text} sections, you would write:
3283 Here the @samp{*} is a wildcard which matches any file name. To exclude a list
3284 of files from matching the file name wildcard, EXCLUDE_FILE may be used to
3285 match all files except the ones specified in the EXCLUDE_FILE list. For
3288 (*(EXCLUDE_FILE (*crtend.o *otherfile.o) .ctors))
3290 will cause all .ctors sections from all files except @file{crtend.o} and
3291 @file{otherfile.o} to be included.
3293 There are two ways to include more than one section:
3299 The difference between these is the order in which the @samp{.text} and
3300 @samp{.rdata} input sections will appear in the output section. In the
3301 first example, they will be intermingled, appearing in the same order as
3302 they are found in the linker input. In the second example, all
3303 @samp{.text} input sections will appear first, followed by all
3304 @samp{.rdata} input sections.
3306 You can specify a file name to include sections from a particular file.
3307 You would do this if one or more of your files contain special data that
3308 needs to be at a particular location in memory. For example:
3313 If you use a file name without a list of sections, then all sections in
3314 the input file will be included in the output section. This is not
3315 commonly done, but it may by useful on occasion. For example:
3320 When you use a file name which does not contain any wild card
3321 characters, the linker will first see if you also specified the file
3322 name on the linker command line or in an @code{INPUT} command. If you
3323 did not, the linker will attempt to open the file as an input file, as
3324 though it appeared on the command line. Note that this differs from an
3325 @code{INPUT} command, because the linker will not search for the file in
3326 the archive search path.
3328 @node Input Section Wildcards
3329 @subsubsection Input Section Wildcard Patterns
3330 @cindex input section wildcards
3331 @cindex wildcard file name patterns
3332 @cindex file name wildcard patterns
3333 @cindex section name wildcard patterns
3334 In an input section description, either the file name or the section
3335 name or both may be wildcard patterns.
3337 The file name of @samp{*} seen in many examples is a simple wildcard
3338 pattern for the file name.
3340 The wildcard patterns are like those used by the Unix shell.
3344 matches any number of characters
3346 matches any single character
3348 matches a single instance of any of the @var{chars}; the @samp{-}
3349 character may be used to specify a range of characters, as in
3350 @samp{[a-z]} to match any lower case letter
3352 quotes the following character
3355 When a file name is matched with a wildcard, the wildcard characters
3356 will not match a @samp{/} character (used to separate directory names on
3357 Unix). A pattern consisting of a single @samp{*} character is an
3358 exception; it will always match any file name, whether it contains a
3359 @samp{/} or not. In a section name, the wildcard characters will match
3360 a @samp{/} character.
3362 File name wildcard patterns only match files which are explicitly
3363 specified on the command line or in an @code{INPUT} command. The linker
3364 does not search directories to expand wildcards.
3366 If a file name matches more than one wildcard pattern, or if a file name
3367 appears explicitly and is also matched by a wildcard pattern, the linker
3368 will use the first match in the linker script. For example, this
3369 sequence of input section descriptions is probably in error, because the
3370 @file{data.o} rule will not be used:
3372 .data : @{ *(.data) @}
3373 .data1 : @{ data.o(.data) @}
3376 @cindex SORT_BY_NAME
3377 Normally, the linker will place files and sections matched by wildcards
3378 in the order in which they are seen during the link. You can change
3379 this by using the @code{SORT_BY_NAME} keyword, which appears before a wildcard
3380 pattern in parentheses (e.g., @code{SORT_BY_NAME(.text*)}). When the
3381 @code{SORT_BY_NAME} keyword is used, the linker will sort the files or sections
3382 into ascending order by name before placing them in the output file.
3384 @cindex SORT_BY_ALIGNMENT
3385 @code{SORT_BY_ALIGNMENT} is very similar to @code{SORT_BY_NAME}. The
3386 difference is @code{SORT_BY_ALIGNMENT} will sort sections into
3387 ascending order by alignment before placing them in the output file.
3390 @code{SORT} is an alias for @code{SORT_BY_NAME}.
3392 When there are nested section sorting commands in linker script, there
3393 can be at most 1 level of nesting for section sorting commands.
3397 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
3398 It will sort the input sections by name first, then by alignment if 2
3399 sections have the same name.
3401 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
3402 It will sort the input sections by alignment first, then by name if 2
3403 sections have the same alignment.
3405 @code{SORT_BY_NAME} (@code{SORT_BY_NAME} (wildcard section pattern)) is
3406 treated the same as @code{SORT_BY_NAME} (wildcard section pattern).
3408 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern))
3409 is treated the same as @code{SORT_BY_ALIGNMENT} (wildcard section pattern).
3411 All other nested section sorting commands are invalid.
3414 When both command line section sorting option and linker script
3415 section sorting command are used, section sorting command always
3416 takes precedence over the command line option.
3418 If the section sorting command in linker script isn't nested, the
3419 command line option will make the section sorting command to be
3420 treated as nested sorting command.
3424 @code{SORT_BY_NAME} (wildcard section pattern ) with
3425 @option{--sort-sections alignment} is equivalent to
3426 @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)).
3428 @code{SORT_BY_ALIGNMENT} (wildcard section pattern) with
3429 @option{--sort-section name} is equivalent to
3430 @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)).
3433 If the section sorting command in linker script is nested, the
3434 command line option will be ignored.
3436 If you ever get confused about where input sections are going, use the
3437 @samp{-M} linker option to generate a map file. The map file shows
3438 precisely how input sections are mapped to output sections.
3440 This example shows how wildcard patterns might be used to partition
3441 files. This linker script directs the linker to place all @samp{.text}
3442 sections in @samp{.text} and all @samp{.bss} sections in @samp{.bss}.
3443 The linker will place the @samp{.data} section from all files beginning
3444 with an upper case character in @samp{.DATA}; for all other files, the
3445 linker will place the @samp{.data} section in @samp{.data}.
3449 .text : @{ *(.text) @}
3450 .DATA : @{ [A-Z]*(.data) @}
3451 .data : @{ *(.data) @}
3452 .bss : @{ *(.bss) @}
3457 @node Input Section Common
3458 @subsubsection Input Section for Common Symbols
3459 @cindex common symbol placement
3460 @cindex uninitialized data placement
3461 A special notation is needed for common symbols, because in many object
3462 file formats common symbols do not have a particular input section. The
3463 linker treats common symbols as though they are in an input section
3464 named @samp{COMMON}.
3466 You may use file names with the @samp{COMMON} section just as with any
3467 other input sections. You can use this to place common symbols from a
3468 particular input file in one section while common symbols from other
3469 input files are placed in another section.
3471 In most cases, common symbols in input files will be placed in the
3472 @samp{.bss} section in the output file. For example:
3474 .bss @{ *(.bss) *(COMMON) @}
3477 @cindex scommon section
3478 @cindex small common symbols
3479 Some object file formats have more than one type of common symbol. For
3480 example, the MIPS ELF object file format distinguishes standard common
3481 symbols and small common symbols. In this case, the linker will use a
3482 different special section name for other types of common symbols. In
3483 the case of MIPS ELF, the linker uses @samp{COMMON} for standard common
3484 symbols and @samp{.scommon} for small common symbols. This permits you
3485 to map the different types of common symbols into memory at different
3489 You will sometimes see @samp{[COMMON]} in old linker scripts. This
3490 notation is now considered obsolete. It is equivalent to
3493 @node Input Section Keep
3494 @subsubsection Input Section and Garbage Collection
3496 @cindex garbage collection
3497 When link-time garbage collection is in use (@samp{--gc-sections}),
3498 it is often useful to mark sections that should not be eliminated.
3499 This is accomplished by surrounding an input section's wildcard entry
3500 with @code{KEEP()}, as in @code{KEEP(*(.init))} or
3501 @code{KEEP(SORT_BY_NAME(*)(.ctors))}.
3503 @node Input Section Example
3504 @subsubsection Input Section Example
3505 The following example is a complete linker script. It tells the linker
3506 to read all of the sections from file @file{all.o} and place them at the
3507 start of output section @samp{outputa} which starts at location
3508 @samp{0x10000}. All of section @samp{.input1} from file @file{foo.o}
3509 follows immediately, in the same output section. All of section
3510 @samp{.input2} from @file{foo.o} goes into output section
3511 @samp{outputb}, followed by section @samp{.input1} from @file{foo1.o}.
3512 All of the remaining @samp{.input1} and @samp{.input2} sections from any
3513 files are written to output section @samp{outputc}.
3541 @node Output Section Data
3542 @subsection Output Section Data
3544 @cindex section data
3545 @cindex output section data
3546 @kindex BYTE(@var{expression})
3547 @kindex SHORT(@var{expression})
3548 @kindex LONG(@var{expression})
3549 @kindex QUAD(@var{expression})
3550 @kindex SQUAD(@var{expression})
3551 You can include explicit bytes of data in an output section by using
3552 @code{BYTE}, @code{SHORT}, @code{LONG}, @code{QUAD}, or @code{SQUAD} as
3553 an output section command. Each keyword is followed by an expression in
3554 parentheses providing the value to store (@pxref{Expressions}). The
3555 value of the expression is stored at the current value of the location
3558 The @code{BYTE}, @code{SHORT}, @code{LONG}, and @code{QUAD} commands
3559 store one, two, four, and eight bytes (respectively). After storing the
3560 bytes, the location counter is incremented by the number of bytes
3563 For example, this will store the byte 1 followed by the four byte value
3564 of the symbol @samp{addr}:
3570 When using a 64 bit host or target, @code{QUAD} and @code{SQUAD} are the
3571 same; they both store an 8 byte, or 64 bit, value. When both host and
3572 target are 32 bits, an expression is computed as 32 bits. In this case
3573 @code{QUAD} stores a 32 bit value zero extended to 64 bits, and
3574 @code{SQUAD} stores a 32 bit value sign extended to 64 bits.
3576 If the object file format of the output file has an explicit endianness,
3577 which is the normal case, the value will be stored in that endianness.
3578 When the object file format does not have an explicit endianness, as is
3579 true of, for example, S-records, the value will be stored in the
3580 endianness of the first input object file.
3582 Note---these commands only work inside a section description and not
3583 between them, so the following will produce an error from the linker:
3585 SECTIONS @{@ .text : @{@ *(.text) @}@ LONG(1) .data : @{@ *(.data) @}@ @}@
3587 whereas this will work:
3589 SECTIONS @{@ .text : @{@ *(.text) ; LONG(1) @}@ .data : @{@ *(.data) @}@ @}@
3592 @kindex FILL(@var{expression})
3593 @cindex holes, filling
3594 @cindex unspecified memory
3595 You may use the @code{FILL} command to set the fill pattern for the
3596 current section. It is followed by an expression in parentheses. Any
3597 otherwise unspecified regions of memory within the section (for example,
3598 gaps left due to the required alignment of input sections) are filled
3599 with the value of the expression, repeated as
3600 necessary. A @code{FILL} statement covers memory locations after the
3601 point at which it occurs in the section definition; by including more
3602 than one @code{FILL} statement, you can have different fill patterns in
3603 different parts of an output section.
3605 This example shows how to fill unspecified regions of memory with the
3611 The @code{FILL} command is similar to the @samp{=@var{fillexp}} output
3612 section attribute, but it only affects the
3613 part of the section following the @code{FILL} command, rather than the
3614 entire section. If both are used, the @code{FILL} command takes
3615 precedence. @xref{Output Section Fill}, for details on the fill
3618 @node Output Section Keywords
3619 @subsection Output Section Keywords
3620 There are a couple of keywords which can appear as output section
3624 @kindex CREATE_OBJECT_SYMBOLS
3625 @cindex input filename symbols
3626 @cindex filename symbols
3627 @item CREATE_OBJECT_SYMBOLS
3628 The command tells the linker to create a symbol for each input file.
3629 The name of each symbol will be the name of the corresponding input
3630 file. The section of each symbol will be the output section in which
3631 the @code{CREATE_OBJECT_SYMBOLS} command appears.
3633 This is conventional for the a.out object file format. It is not
3634 normally used for any other object file format.
3636 @kindex CONSTRUCTORS
3637 @cindex C++ constructors, arranging in link
3638 @cindex constructors, arranging in link
3640 When linking using the a.out object file format, the linker uses an
3641 unusual set construct to support C++ global constructors and
3642 destructors. When linking object file formats which do not support
3643 arbitrary sections, such as ECOFF and XCOFF, the linker will
3644 automatically recognize C++ global constructors and destructors by name.
3645 For these object file formats, the @code{CONSTRUCTORS} command tells the
3646 linker to place constructor information in the output section where the
3647 @code{CONSTRUCTORS} command appears. The @code{CONSTRUCTORS} command is
3648 ignored for other object file formats.
3650 The symbol @w{@code{__CTOR_LIST__}} marks the start of the global
3651 constructors, and the symbol @w{@code{__CTOR_END__}} marks the end.
3652 Similarly, @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_END__}} mark
3653 the start and end of the global destructors. The
3654 first word in the list is the number of entries, followed by the address
3655 of each constructor or destructor, followed by a zero word. The
3656 compiler must arrange to actually run the code. For these object file
3657 formats @sc{gnu} C++ normally calls constructors from a subroutine
3658 @code{__main}; a call to @code{__main} is automatically inserted into
3659 the startup code for @code{main}. @sc{gnu} C++ normally runs
3660 destructors either by using @code{atexit}, or directly from the function
3663 For object file formats such as @code{COFF} or @code{ELF} which support
3664 arbitrary section names, @sc{gnu} C++ will normally arrange to put the
3665 addresses of global constructors and destructors into the @code{.ctors}
3666 and @code{.dtors} sections. Placing the following sequence into your
3667 linker script will build the sort of table which the @sc{gnu} C++
3668 runtime code expects to see.
3672 LONG((__CTOR_END__ - __CTOR_LIST__) / 4 - 2)
3677 LONG((__DTOR_END__ - __DTOR_LIST__) / 4 - 2)
3683 If you are using the @sc{gnu} C++ support for initialization priority,
3684 which provides some control over the order in which global constructors
3685 are run, you must sort the constructors at link time to ensure that they
3686 are executed in the correct order. When using the @code{CONSTRUCTORS}
3687 command, use @samp{SORT_BY_NAME(CONSTRUCTORS)} instead. When using the
3688 @code{.ctors} and @code{.dtors} sections, use @samp{*(SORT_BY_NAME(.ctors))} and
3689 @samp{*(SORT_BY_NAME(.dtors))} instead of just @samp{*(.ctors)} and
3692 Normally the compiler and linker will handle these issues automatically,
3693 and you will not need to concern yourself with them. However, you may
3694 need to consider this if you are using C++ and writing your own linker
3699 @node Output Section Discarding
3700 @subsection Output Section Discarding
3701 @cindex discarding sections
3702 @cindex sections, discarding
3703 @cindex removing sections
3704 The linker will not create output sections with no contents. This is
3705 for convenience when referring to input sections that may or may not
3706 be present in any of the input files. For example:
3708 .foo : @{ *(.foo) @}
3711 will only create a @samp{.foo} section in the output file if there is a
3712 @samp{.foo} section in at least one input file, and if the input
3713 sections are not all empty. Other link script directives that allocate
3714 space in an output section will also create the output section.
3716 The linker will ignore address assignments (@pxref{Output Section Address})
3717 on discarded output sections, except when the linker script defines
3718 symbols in the output section. In that case the linker will obey
3719 the address assignments, possibly advancing dot even though the
3720 section is discarded.
3723 The special output section name @samp{/DISCARD/} may be used to discard
3724 input sections. Any input sections which are assigned to an output
3725 section named @samp{/DISCARD/} are not included in the output file.
3727 @node Output Section Attributes
3728 @subsection Output Section Attributes
3729 @cindex output section attributes
3730 We showed above that the full description of an output section looked
3734 @var{section} [@var{address}] [(@var{type})] :
3735 [AT(@var{lma})] [ALIGN(@var{section_align})] [SUBALIGN(@var{subsection_align})]
3737 @var{output-section-command}
3738 @var{output-section-command}
3740 @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}]
3743 We've already described @var{section}, @var{address}, and
3744 @var{output-section-command}. In this section we will describe the
3745 remaining section attributes.
3748 * Output Section Type:: Output section type
3749 * Output Section LMA:: Output section LMA
3750 * Forced Output Alignment:: Forced Output Alignment
3751 * Forced Input Alignment:: Forced Input Alignment
3752 * Output Section Region:: Output section region
3753 * Output Section Phdr:: Output section phdr
3754 * Output Section Fill:: Output section fill
3757 @node Output Section Type
3758 @subsubsection Output Section Type
3759 Each output section may have a type. The type is a keyword in
3760 parentheses. The following types are defined:
3764 The section should be marked as not loadable, so that it will not be
3765 loaded into memory when the program is run.
3770 These type names are supported for backward compatibility, and are
3771 rarely used. They all have the same effect: the section should be
3772 marked as not allocatable, so that no memory is allocated for the
3773 section when the program is run.
3777 @cindex prevent unnecessary loading
3778 @cindex loading, preventing
3779 The linker normally sets the attributes of an output section based on
3780 the input sections which map into it. You can override this by using
3781 the section type. For example, in the script sample below, the
3782 @samp{ROM} section is addressed at memory location @samp{0} and does not
3783 need to be loaded when the program is run. The contents of the
3784 @samp{ROM} section will appear in the linker output file as usual.
3788 ROM 0 (NOLOAD) : @{ @dots{} @}
3794 @node Output Section LMA
3795 @subsubsection Output Section LMA
3796 @kindex AT>@var{lma_region}
3797 @kindex AT(@var{lma})
3798 @cindex load address
3799 @cindex section load address
3800 Every section has a virtual address (VMA) and a load address (LMA); see
3801 @ref{Basic Script Concepts}. The address expression which may appear in
3802 an output section description sets the VMA (@pxref{Output Section
3805 The expression @var{lma} that follows the @code{AT} keyword specifies
3806 the load address of the section.
3808 Alternatively, with @samp{AT>@var{lma_region}} expression, you may
3809 specify a memory region for the section's load address. @xref{MEMORY}.
3810 Note that if the section has not had a VMA assigned to it then the
3811 linker will use the @var{lma_region} as the VMA region as well.
3813 If neither @code{AT} nor @code{AT>} is specified for an allocatable
3814 section, the linker will set the LMA such that the difference between
3815 VMA and LMA for the section is the same as the preceding output
3816 section in the same region. If there is no preceding output section
3817 or the section is not allocatable, the linker will set the LMA equal
3819 @xref{Output Section Region}.
3821 @cindex ROM initialized data
3822 @cindex initialized data in ROM
3823 This feature is designed to make it easy to build a ROM image. For
3824 example, the following linker script creates three output sections: one
3825 called @samp{.text}, which starts at @code{0x1000}, one called
3826 @samp{.mdata}, which is loaded at the end of the @samp{.text} section
3827 even though its VMA is @code{0x2000}, and one called @samp{.bss} to hold
3828 uninitialized data at address @code{0x3000}. The symbol @code{_data} is
3829 defined with the value @code{0x2000}, which shows that the location
3830 counter holds the VMA value, not the LMA value.
3836 .text 0x1000 : @{ *(.text) _etext = . ; @}
3838 AT ( ADDR (.text) + SIZEOF (.text) )
3839 @{ _data = . ; *(.data); _edata = . ; @}
3841 @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@}
3846 The run-time initialization code for use with a program generated with
3847 this linker script would include something like the following, to copy
3848 the initialized data from the ROM image to its runtime address. Notice
3849 how this code takes advantage of the symbols defined by the linker
3854 extern char _etext, _data, _edata, _bstart, _bend;
3855 char *src = &_etext;
3858 /* ROM has data at end of text; copy it. */
3859 while (dst < &_edata) @{
3864 for (dst = &_bstart; dst< &_bend; dst++)
3869 @node Forced Output Alignment
3870 @subsubsection Forced Output Alignment
3871 @kindex ALIGN(@var{section_align})
3872 @cindex forcing output section alignment
3873 @cindex output section alignment
3874 You can increase an output section's alignment by using ALIGN.
3876 @node Forced Input Alignment
3877 @subsubsection Forced Input Alignment
3878 @kindex SUBALIGN(@var{subsection_align})
3879 @cindex forcing input section alignment
3880 @cindex input section alignment
3881 You can force input section alignment within an output section by using
3882 SUBALIGN. The value specified overrides any alignment given by input
3883 sections, whether larger or smaller.
3885 @node Output Section Region
3886 @subsubsection Output Section Region
3887 @kindex >@var{region}
3888 @cindex section, assigning to memory region
3889 @cindex memory regions and sections
3890 You can assign a section to a previously defined region of memory by
3891 using @samp{>@var{region}}. @xref{MEMORY}.
3893 Here is a simple example:
3896 MEMORY @{ rom : ORIGIN = 0x1000, LENGTH = 0x1000 @}
3897 SECTIONS @{ ROM : @{ *(.text) @} >rom @}
3901 @node Output Section Phdr
3902 @subsubsection Output Section Phdr
3904 @cindex section, assigning to program header
3905 @cindex program headers and sections
3906 You can assign a section to a previously defined program segment by
3907 using @samp{:@var{phdr}}. @xref{PHDRS}. If a section is assigned to
3908 one or more segments, then all subsequent allocated sections will be
3909 assigned to those segments as well, unless they use an explicitly
3910 @code{:@var{phdr}} modifier. You can use @code{:NONE} to tell the
3911 linker to not put the section in any segment at all.
3913 Here is a simple example:
3916 PHDRS @{ text PT_LOAD ; @}
3917 SECTIONS @{ .text : @{ *(.text) @} :text @}
3921 @node Output Section Fill
3922 @subsubsection Output Section Fill
3923 @kindex =@var{fillexp}
3924 @cindex section fill pattern
3925 @cindex fill pattern, entire section
3926 You can set the fill pattern for an entire section by using
3927 @samp{=@var{fillexp}}. @var{fillexp} is an expression
3928 (@pxref{Expressions}). Any otherwise unspecified regions of memory
3929 within the output section (for example, gaps left due to the required
3930 alignment of input sections) will be filled with the value, repeated as
3931 necessary. If the fill expression is a simple hex number, ie. a string
3932 of hex digit starting with @samp{0x} and without a trailing @samp{k} or @samp{M}, then
3933 an arbitrarily long sequence of hex digits can be used to specify the
3934 fill pattern; Leading zeros become part of the pattern too. For all
3935 other cases, including extra parentheses or a unary @code{+}, the fill
3936 pattern is the four least significant bytes of the value of the
3937 expression. In all cases, the number is big-endian.
3939 You can also change the fill value with a @code{FILL} command in the
3940 output section commands; (@pxref{Output Section Data}).
3942 Here is a simple example:
3945 SECTIONS @{ .text : @{ *(.text) @} =0x90909090 @}
3949 @node Overlay Description
3950 @subsection Overlay Description
3953 An overlay description provides an easy way to describe sections which
3954 are to be loaded as part of a single memory image but are to be run at
3955 the same memory address. At run time, some sort of overlay manager will
3956 copy the overlaid sections in and out of the runtime memory address as
3957 required, perhaps by simply manipulating addressing bits. This approach
3958 can be useful, for example, when a certain region of memory is faster
3961 Overlays are described using the @code{OVERLAY} command. The
3962 @code{OVERLAY} command is used within a @code{SECTIONS} command, like an
3963 output section description. The full syntax of the @code{OVERLAY}
3964 command is as follows:
3967 OVERLAY [@var{start}] : [NOCROSSREFS] [AT ( @var{ldaddr} )]
3971 @var{output-section-command}
3972 @var{output-section-command}
3974 @} [:@var{phdr}@dots{}] [=@var{fill}]
3977 @var{output-section-command}
3978 @var{output-section-command}
3980 @} [:@var{phdr}@dots{}] [=@var{fill}]
3982 @} [>@var{region}] [:@var{phdr}@dots{}] [=@var{fill}]
3986 Everything is optional except @code{OVERLAY} (a keyword), and each
3987 section must have a name (@var{secname1} and @var{secname2} above). The
3988 section definitions within the @code{OVERLAY} construct are identical to
3989 those within the general @code{SECTIONS} contruct (@pxref{SECTIONS}),
3990 except that no addresses and no memory regions may be defined for
3991 sections within an @code{OVERLAY}.
3993 The sections are all defined with the same starting address. The load
3994 addresses of the sections are arranged such that they are consecutive in
3995 memory starting at the load address used for the @code{OVERLAY} as a
3996 whole (as with normal section definitions, the load address is optional,
3997 and defaults to the start address; the start address is also optional,
3998 and defaults to the current value of the location counter).
4000 If the @code{NOCROSSREFS} keyword is used, and there any references
4001 among the sections, the linker will report an error. Since the sections
4002 all run at the same address, it normally does not make sense for one
4003 section to refer directly to another. @xref{Miscellaneous Commands,
4006 For each section within the @code{OVERLAY}, the linker automatically
4007 defines two symbols. The symbol @code{__load_start_@var{secname}} is
4008 defined as the starting load address of the section. The symbol
4009 @code{__load_stop_@var{secname}} is defined as the final load address of
4010 the section. Any characters within @var{secname} which are not legal
4011 within C identifiers are removed. C (or assembler) code may use these
4012 symbols to move the overlaid sections around as necessary.
4014 At the end of the overlay, the value of the location counter is set to
4015 the start address of the overlay plus the size of the largest section.
4017 Here is an example. Remember that this would appear inside a
4018 @code{SECTIONS} construct.
4021 OVERLAY 0x1000 : AT (0x4000)
4023 .text0 @{ o1/*.o(.text) @}
4024 .text1 @{ o2/*.o(.text) @}
4029 This will define both @samp{.text0} and @samp{.text1} to start at
4030 address 0x1000. @samp{.text0} will be loaded at address 0x4000, and
4031 @samp{.text1} will be loaded immediately after @samp{.text0}. The
4032 following symbols will be defined: @code{__load_start_text0},
4033 @code{__load_stop_text0}, @code{__load_start_text1},
4034 @code{__load_stop_text1}.
4036 C code to copy overlay @code{.text1} into the overlay area might look
4041 extern char __load_start_text1, __load_stop_text1;
4042 memcpy ((char *) 0x1000, &__load_start_text1,
4043 &__load_stop_text1 - &__load_start_text1);
4047 Note that the @code{OVERLAY} command is just syntactic sugar, since
4048 everything it does can be done using the more basic commands. The above
4049 example could have been written identically as follows.
4053 .text0 0x1000 : AT (0x4000) @{ o1/*.o(.text) @}
4054 __load_start_text0 = LOADADDR (.text0);
4055 __load_stop_text0 = LOADADDR (.text0) + SIZEOF (.text0);
4056 .text1 0x1000 : AT (0x4000 + SIZEOF (.text0)) @{ o2/*.o(.text) @}
4057 __load_start_text1 = LOADADDR (.text1);
4058 __load_stop_text1 = LOADADDR (.text1) + SIZEOF (.text1);
4059 . = 0x1000 + MAX (SIZEOF (.text0), SIZEOF (.text1));
4064 @section MEMORY Command
4066 @cindex memory regions
4067 @cindex regions of memory
4068 @cindex allocating memory
4069 @cindex discontinuous memory
4070 The linker's default configuration permits allocation of all available
4071 memory. You can override this by using the @code{MEMORY} command.
4073 The @code{MEMORY} command describes the location and size of blocks of
4074 memory in the target. You can use it to describe which memory regions
4075 may be used by the linker, and which memory regions it must avoid. You
4076 can then assign sections to particular memory regions. The linker will
4077 set section addresses based on the memory regions, and will warn about
4078 regions that become too full. The linker will not shuffle sections
4079 around to fit into the available regions.
4081 A linker script may contain at most one use of the @code{MEMORY}
4082 command. However, you can define as many blocks of memory within it as
4083 you wish. The syntax is:
4088 @var{name} [(@var{attr})] : ORIGIN = @var{origin}, LENGTH = @var{len}
4094 The @var{name} is a name used in the linker script to refer to the
4095 region. The region name has no meaning outside of the linker script.
4096 Region names are stored in a separate name space, and will not conflict
4097 with symbol names, file names, or section names. Each memory region
4098 must have a distinct name.
4100 @cindex memory region attributes
4101 The @var{attr} string is an optional list of attributes that specify
4102 whether to use a particular memory region for an input section which is
4103 not explicitly mapped in the linker script. As described in
4104 @ref{SECTIONS}, if you do not specify an output section for some input
4105 section, the linker will create an output section with the same name as
4106 the input section. If you define region attributes, the linker will use
4107 them to select the memory region for the output section that it creates.
4109 The @var{attr} string must consist only of the following characters:
4124 Invert the sense of any of the preceding attributes
4127 If a unmapped section matches any of the listed attributes other than
4128 @samp{!}, it will be placed in the memory region. The @samp{!}
4129 attribute reverses this test, so that an unmapped section will be placed
4130 in the memory region only if it does not match any of the listed
4136 The @var{origin} is an numerical expression for the start address of
4137 the memory region. The expression must evaluate to a constant and it
4138 cannot involve any symbols. The keyword @code{ORIGIN} may be
4139 abbreviated to @code{org} or @code{o} (but not, for example,
4145 The @var{len} is an expression for the size in bytes of the memory
4146 region. As with the @var{origin} expression, the expression must
4147 be numerical only and must evaluate to a constant. The keyword
4148 @code{LENGTH} may be abbreviated to @code{len} or @code{l}.
4150 In the following example, we specify that there are two memory regions
4151 available for allocation: one starting at @samp{0} for 256 kilobytes,
4152 and the other starting at @samp{0x40000000} for four megabytes. The
4153 linker will place into the @samp{rom} memory region every section which
4154 is not explicitly mapped into a memory region, and is either read-only
4155 or executable. The linker will place other sections which are not
4156 explicitly mapped into a memory region into the @samp{ram} memory
4163 rom (rx) : ORIGIN = 0, LENGTH = 256K
4164 ram (!rx) : org = 0x40000000, l = 4M
4169 Once you define a memory region, you can direct the linker to place
4170 specific output sections into that memory region by using the
4171 @samp{>@var{region}} output section attribute. For example, if you have
4172 a memory region named @samp{mem}, you would use @samp{>mem} in the
4173 output section definition. @xref{Output Section Region}. If no address
4174 was specified for the output section, the linker will set the address to
4175 the next available address within the memory region. If the combined
4176 output sections directed to a memory region are too large for the
4177 region, the linker will issue an error message.
4179 It is possible to access the origin and length of a memory in an
4180 expression via the @code{ORIGIN(@var{memory})} and
4181 @code{LENGTH(@var{memory})} functions:
4185 _fstack = ORIGIN(ram) + LENGTH(ram) - 4;
4190 @section PHDRS Command
4192 @cindex program headers
4193 @cindex ELF program headers
4194 @cindex program segments
4195 @cindex segments, ELF
4196 The ELF object file format uses @dfn{program headers}, also knows as
4197 @dfn{segments}. The program headers describe how the program should be
4198 loaded into memory. You can print them out by using the @code{objdump}
4199 program with the @samp{-p} option.
4201 When you run an ELF program on a native ELF system, the system loader
4202 reads the program headers in order to figure out how to load the
4203 program. This will only work if the program headers are set correctly.
4204 This manual does not describe the details of how the system loader
4205 interprets program headers; for more information, see the ELF ABI.
4207 The linker will create reasonable program headers by default. However,
4208 in some cases, you may need to specify the program headers more
4209 precisely. You may use the @code{PHDRS} command for this purpose. When
4210 the linker sees the @code{PHDRS} command in the linker script, it will
4211 not create any program headers other than the ones specified.
4213 The linker only pays attention to the @code{PHDRS} command when
4214 generating an ELF output file. In other cases, the linker will simply
4215 ignore @code{PHDRS}.
4217 This is the syntax of the @code{PHDRS} command. The words @code{PHDRS},
4218 @code{FILEHDR}, @code{AT}, and @code{FLAGS} are keywords.
4224 @var{name} @var{type} [ FILEHDR ] [ PHDRS ] [ AT ( @var{address} ) ]
4225 [ FLAGS ( @var{flags} ) ] ;
4230 The @var{name} is used only for reference in the @code{SECTIONS} command
4231 of the linker script. It is not put into the output file. Program
4232 header names are stored in a separate name space, and will not conflict
4233 with symbol names, file names, or section names. Each program header
4234 must have a distinct name.
4236 Certain program header types describe segments of memory which the
4237 system loader will load from the file. In the linker script, you
4238 specify the contents of these segments by placing allocatable output
4239 sections in the segments. You use the @samp{:@var{phdr}} output section
4240 attribute to place a section in a particular segment. @xref{Output
4243 It is normal to put certain sections in more than one segment. This
4244 merely implies that one segment of memory contains another. You may
4245 repeat @samp{:@var{phdr}}, using it once for each segment which should
4246 contain the section.
4248 If you place a section in one or more segments using @samp{:@var{phdr}},
4249 then the linker will place all subsequent allocatable sections which do
4250 not specify @samp{:@var{phdr}} in the same segments. This is for
4251 convenience, since generally a whole set of contiguous sections will be
4252 placed in a single segment. You can use @code{:NONE} to override the
4253 default segment and tell the linker to not put the section in any
4258 You may use the @code{FILEHDR} and @code{PHDRS} keywords appear after
4259 the program header type to further describe the contents of the segment.
4260 The @code{FILEHDR} keyword means that the segment should include the ELF
4261 file header. The @code{PHDRS} keyword means that the segment should
4262 include the ELF program headers themselves.
4264 The @var{type} may be one of the following. The numbers indicate the
4265 value of the keyword.
4268 @item @code{PT_NULL} (0)
4269 Indicates an unused program header.
4271 @item @code{PT_LOAD} (1)
4272 Indicates that this program header describes a segment to be loaded from
4275 @item @code{PT_DYNAMIC} (2)
4276 Indicates a segment where dynamic linking information can be found.
4278 @item @code{PT_INTERP} (3)
4279 Indicates a segment where the name of the program interpreter may be
4282 @item @code{PT_NOTE} (4)
4283 Indicates a segment holding note information.
4285 @item @code{PT_SHLIB} (5)
4286 A reserved program header type, defined but not specified by the ELF
4289 @item @code{PT_PHDR} (6)
4290 Indicates a segment where the program headers may be found.
4292 @item @var{expression}
4293 An expression giving the numeric type of the program header. This may
4294 be used for types not defined above.
4297 You can specify that a segment should be loaded at a particular address
4298 in memory by using an @code{AT} expression. This is identical to the
4299 @code{AT} command used as an output section attribute (@pxref{Output
4300 Section LMA}). The @code{AT} command for a program header overrides the
4301 output section attribute.
4303 The linker will normally set the segment flags based on the sections
4304 which comprise the segment. You may use the @code{FLAGS} keyword to
4305 explicitly specify the segment flags. The value of @var{flags} must be
4306 an integer. It is used to set the @code{p_flags} field of the program
4309 Here is an example of @code{PHDRS}. This shows a typical set of program
4310 headers used on a native ELF system.
4316 headers PT_PHDR PHDRS ;
4318 text PT_LOAD FILEHDR PHDRS ;
4320 dynamic PT_DYNAMIC ;
4326 .interp : @{ *(.interp) @} :text :interp
4327 .text : @{ *(.text) @} :text
4328 .rodata : @{ *(.rodata) @} /* defaults to :text */
4330 . = . + 0x1000; /* move to a new page in memory */
4331 .data : @{ *(.data) @} :data
4332 .dynamic : @{ *(.dynamic) @} :data :dynamic
4339 @section VERSION Command
4340 @kindex VERSION @{script text@}
4341 @cindex symbol versions
4342 @cindex version script
4343 @cindex versions of symbols
4344 The linker supports symbol versions when using ELF. Symbol versions are
4345 only useful when using shared libraries. The dynamic linker can use
4346 symbol versions to select a specific version of a function when it runs
4347 a program that may have been linked against an earlier version of the
4350 You can include a version script directly in the main linker script, or
4351 you can supply the version script as an implicit linker script. You can
4352 also use the @samp{--version-script} linker option.
4354 The syntax of the @code{VERSION} command is simply
4356 VERSION @{ version-script-commands @}
4359 The format of the version script commands is identical to that used by
4360 Sun's linker in Solaris 2.5. The version script defines a tree of
4361 version nodes. You specify the node names and interdependencies in the
4362 version script. You can specify which symbols are bound to which
4363 version nodes, and you can reduce a specified set of symbols to local
4364 scope so that they are not globally visible outside of the shared
4367 The easiest way to demonstrate the version script language is with a few
4388 "int f(int, double)";
4393 This example version script defines three version nodes. The first
4394 version node defined is @samp{VERS_1.1}; it has no other dependencies.
4395 The script binds the symbol @samp{foo1} to @samp{VERS_1.1}. It reduces
4396 a number of symbols to local scope so that they are not visible outside
4397 of the shared library; this is done using wildcard patterns, so that any
4398 symbol whose name begins with @samp{old}, @samp{original}, or @samp{new}
4399 is matched. The wildcard patterns available are the same as those used
4400 in the shell when matching filenames (also known as ``globbing'').
4401 However, if you specify the symbol name inside double quotes, then the
4402 name is treated as literal, rather than as a glob pattern.
4404 Next, the version script defines node @samp{VERS_1.2}. This node
4405 depends upon @samp{VERS_1.1}. The script binds the symbol @samp{foo2}
4406 to the version node @samp{VERS_1.2}.
4408 Finally, the version script defines node @samp{VERS_2.0}. This node
4409 depends upon @samp{VERS_1.2}. The scripts binds the symbols @samp{bar1}
4410 and @samp{bar2} are bound to the version node @samp{VERS_2.0}.
4412 When the linker finds a symbol defined in a library which is not
4413 specifically bound to a version node, it will effectively bind it to an
4414 unspecified base version of the library. You can bind all otherwise
4415 unspecified symbols to a given version node by using @samp{global: *;}
4416 somewhere in the version script.
4418 The names of the version nodes have no specific meaning other than what
4419 they might suggest to the person reading them. The @samp{2.0} version
4420 could just as well have appeared in between @samp{1.1} and @samp{1.2}.
4421 However, this would be a confusing way to write a version script.
4423 Node name can be omitted, provided it is the only version node
4424 in the version script. Such version script doesn't assign any versions to
4425 symbols, only selects which symbols will be globally visible out and which
4429 @{ global: foo; bar; local: *; @};
4432 When you link an application against a shared library that has versioned
4433 symbols, the application itself knows which version of each symbol it
4434 requires, and it also knows which version nodes it needs from each
4435 shared library it is linked against. Thus at runtime, the dynamic
4436 loader can make a quick check to make sure that the libraries you have
4437 linked against do in fact supply all of the version nodes that the
4438 application will need to resolve all of the dynamic symbols. In this
4439 way it is possible for the dynamic linker to know with certainty that
4440 all external symbols that it needs will be resolvable without having to
4441 search for each symbol reference.
4443 The symbol versioning is in effect a much more sophisticated way of
4444 doing minor version checking that SunOS does. The fundamental problem
4445 that is being addressed here is that typically references to external
4446 functions are bound on an as-needed basis, and are not all bound when
4447 the application starts up. If a shared library is out of date, a
4448 required interface may be missing; when the application tries to use
4449 that interface, it may suddenly and unexpectedly fail. With symbol
4450 versioning, the user will get a warning when they start their program if
4451 the libraries being used with the application are too old.
4453 There are several GNU extensions to Sun's versioning approach. The
4454 first of these is the ability to bind a symbol to a version node in the
4455 source file where the symbol is defined instead of in the versioning
4456 script. This was done mainly to reduce the burden on the library
4457 maintainer. You can do this by putting something like:
4459 __asm__(".symver original_foo,foo@@VERS_1.1");
4462 in the C source file. This renames the function @samp{original_foo} to
4463 be an alias for @samp{foo} bound to the version node @samp{VERS_1.1}.
4464 The @samp{local:} directive can be used to prevent the symbol
4465 @samp{original_foo} from being exported. A @samp{.symver} directive
4466 takes precedence over a version script.
4468 The second GNU extension is to allow multiple versions of the same
4469 function to appear in a given shared library. In this way you can make
4470 an incompatible change to an interface without increasing the major
4471 version number of the shared library, while still allowing applications
4472 linked against the old interface to continue to function.
4474 To do this, you must use multiple @samp{.symver} directives in the
4475 source file. Here is an example:
4478 __asm__(".symver original_foo,foo@@");
4479 __asm__(".symver old_foo,foo@@VERS_1.1");
4480 __asm__(".symver old_foo1,foo@@VERS_1.2");
4481 __asm__(".symver new_foo,foo@@@@VERS_2.0");
4484 In this example, @samp{foo@@} represents the symbol @samp{foo} bound to the
4485 unspecified base version of the symbol. The source file that contains this
4486 example would define 4 C functions: @samp{original_foo}, @samp{old_foo},
4487 @samp{old_foo1}, and @samp{new_foo}.
4489 When you have multiple definitions of a given symbol, there needs to be
4490 some way to specify a default version to which external references to
4491 this symbol will be bound. You can do this with the
4492 @samp{foo@@@@VERS_2.0} type of @samp{.symver} directive. You can only
4493 declare one version of a symbol as the default in this manner; otherwise
4494 you would effectively have multiple definitions of the same symbol.
4496 If you wish to bind a reference to a specific version of the symbol
4497 within the shared library, you can use the aliases of convenience
4498 (i.e., @samp{old_foo}), or you can use the @samp{.symver} directive to
4499 specifically bind to an external version of the function in question.
4501 You can also specify the language in the version script:
4504 VERSION extern "lang" @{ version-script-commands @}
4507 The supported @samp{lang}s are @samp{C}, @samp{C++}, and @samp{Java}.
4508 The linker will iterate over the list of symbols at the link time and
4509 demangle them according to @samp{lang} before matching them to the
4510 patterns specified in @samp{version-script-commands}.
4512 Demangled names may contains spaces and other special characters. As
4513 described above, you can use a glob pattern to match demangled names,
4514 or you can use a double-quoted string to match the string exactly. In
4515 the latter case, be aware that minor differences (such as differing
4516 whitespace) between the version script and the demangler output will
4517 cause a mismatch. As the exact string generated by the demangler
4518 might change in the future, even if the mangled name does not, you
4519 should check that all of your version directives are behaving as you
4520 expect when you upgrade.
4523 @section Expressions in Linker Scripts
4526 The syntax for expressions in the linker script language is identical to
4527 that of C expressions. All expressions are evaluated as integers. All
4528 expressions are evaluated in the same size, which is 32 bits if both the
4529 host and target are 32 bits, and is otherwise 64 bits.
4531 You can use and set symbol values in expressions.
4533 The linker defines several special purpose builtin functions for use in
4537 * Constants:: Constants
4538 * Symbols:: Symbol Names
4539 * Orphan Sections:: Orphan Sections
4540 * Location Counter:: The Location Counter
4541 * Operators:: Operators
4542 * Evaluation:: Evaluation
4543 * Expression Section:: The Section of an Expression
4544 * Builtin Functions:: Builtin Functions
4548 @subsection Constants
4549 @cindex integer notation
4550 @cindex constants in linker scripts
4551 All constants are integers.
4553 As in C, the linker considers an integer beginning with @samp{0} to be
4554 octal, and an integer beginning with @samp{0x} or @samp{0X} to be
4555 hexadecimal. The linker considers other integers to be decimal.
4557 @cindex scaled integers
4558 @cindex K and M integer suffixes
4559 @cindex M and K integer suffixes
4560 @cindex suffixes for integers
4561 @cindex integer suffixes
4562 In addition, you can use the suffixes @code{K} and @code{M} to scale a
4566 @c END TEXI2ROFF-KILL
4567 @code{1024} or @code{1024*1024}
4571 ${\rm 1024}$ or ${\rm 1024}^2$
4573 @c END TEXI2ROFF-KILL
4574 respectively. For example, the following all refer to the same quantity:
4582 @subsection Symbol Names
4583 @cindex symbol names
4585 @cindex quoted symbol names
4587 Unless quoted, symbol names start with a letter, underscore, or period
4588 and may include letters, digits, underscores, periods, and hyphens.
4589 Unquoted symbol names must not conflict with any keywords. You can
4590 specify a symbol which contains odd characters or has the same name as a
4591 keyword by surrounding the symbol name in double quotes:
4594 "with a space" = "also with a space" + 10;
4597 Since symbols can contain many non-alphabetic characters, it is safest
4598 to delimit symbols with spaces. For example, @samp{A-B} is one symbol,
4599 whereas @samp{A - B} is an expression involving subtraction.
4601 @node Orphan Sections
4602 @subsection Orphan Sections
4604 Orphan sections are sections present in the input files which
4605 are not explicitly placed into the output file by the linker
4606 script. The linker will still copy these sections into the
4607 output file, but it has to guess as to where they should be
4608 placed. The linker uses a simple heuristic to do this. It
4609 attempts to place orphan sections after non-orphan sections of the
4610 same attribute, such as code vs data, loadable vs non-loadable, etc.
4611 If there is not enough room to do this then it places
4612 at the end of the file.
4614 For ELF targets, the attribute of the section includes section type as
4615 well as section flag.
4617 @node Location Counter
4618 @subsection The Location Counter
4621 @cindex location counter
4622 @cindex current output location
4623 The special linker variable @dfn{dot} @samp{.} always contains the
4624 current output location counter. Since the @code{.} always refers to a
4625 location in an output section, it may only appear in an expression
4626 within a @code{SECTIONS} command. The @code{.} symbol may appear
4627 anywhere that an ordinary symbol is allowed in an expression.
4630 Assigning a value to @code{.} will cause the location counter to be
4631 moved. This may be used to create holes in the output section. The
4632 location counter may not be moved backwards inside an output section,
4633 and may not be moved backwards outside of an output section if so
4634 doing creates areas with overlapping LMAs.
4650 In the previous example, the @samp{.text} section from @file{file1} is
4651 located at the beginning of the output section @samp{output}. It is
4652 followed by a 1000 byte gap. Then the @samp{.text} section from
4653 @file{file2} appears, also with a 1000 byte gap following before the
4654 @samp{.text} section from @file{file3}. The notation @samp{= 0x12345678}
4655 specifies what data to write in the gaps (@pxref{Output Section Fill}).
4657 @cindex dot inside sections
4658 Note: @code{.} actually refers to the byte offset from the start of the
4659 current containing object. Normally this is the @code{SECTIONS}
4660 statement, whose start address is 0, hence @code{.} can be used as an
4661 absolute address. If @code{.} is used inside a section description
4662 however, it refers to the byte offset from the start of that section,
4663 not an absolute address. Thus in a script like this:
4681 The @samp{.text} section will be assigned a starting address of 0x100
4682 and a size of exactly 0x200 bytes, even if there is not enough data in
4683 the @samp{.text} input sections to fill this area. (If there is too
4684 much data, an error will be produced because this would be an attempt to
4685 move @code{.} backwards). The @samp{.data} section will start at 0x500
4686 and it will have an extra 0x600 bytes worth of space after the end of
4687 the values from the @samp{.data} input sections and before the end of
4688 the @samp{.data} output section itself.
4690 @cindex dot outside sections
4691 Setting symbols to the value of the location counter outside of an
4692 output section statement can result in unexpected values if the linker
4693 needs to place orphan sections. For example, given the following:
4699 .text: @{ *(.text) @}
4703 .data: @{ *(.data) @}
4708 If the linker needs to place some input section, e.g. @code{.rodata},
4709 not mentioned in the script, it might choose to place that section
4710 between @code{.text} and @code{.data}. You might think the linker
4711 should place @code{.rodata} on the blank line in the above script, but
4712 blank lines are of no particular significance to the linker. As well,
4713 the linker doesn't associate the above symbol names with their
4714 sections. Instead, it assumes that all assignments or other
4715 statements belong to the previous output section, except for the
4716 special case of an assignment to @code{.}. I.e., the linker will
4717 place the orphan @code{.rodata} section as if the script was written
4724 .text: @{ *(.text) @}
4728 .rodata: @{ *(.rodata) @}
4729 .data: @{ *(.data) @}
4734 This may or may not be the script author's intention for the value of
4735 @code{start_of_data}. One way to influence the orphan section
4736 placement is to assign the location counter to itself, as the linker
4737 assumes that an assignment to @code{.} is setting the start address of
4738 a following output section and thus should be grouped with that
4739 section. So you could write:
4745 .text: @{ *(.text) @}
4750 .data: @{ *(.data) @}
4755 Now, the orphan @code{.rodata} section will be placed between
4756 @code{end_of_text} and @code{start_of_data}.
4760 @subsection Operators
4761 @cindex operators for arithmetic
4762 @cindex arithmetic operators
4763 @cindex precedence in expressions
4764 The linker recognizes the standard C set of arithmetic operators, with
4765 the standard bindings and precedence levels:
4768 @c END TEXI2ROFF-KILL
4770 precedence associativity Operators Notes
4776 5 left == != > < <= >=
4782 11 right &= += -= *= /= (2)
4786 (1) Prefix operators
4787 (2) @xref{Assignments}.
4791 \vskip \baselineskip
4792 %"lispnarrowing" is the extra indent used generally for smallexample
4793 \hskip\lispnarrowing\vbox{\offinterlineskip
4796 {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr
4797 height2pt&\omit&&\omit&&\omit&\cr
4798 &Precedence&& Associativity &&{\rm Operators}&\cr
4799 height2pt&\omit&&\omit&&\omit&\cr
4801 height2pt&\omit&&\omit&&\omit&\cr
4803 % '176 is tilde, '~' in tt font
4804 &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr
4805 &2&&left&&* / \%&\cr
4808 &5&&left&&== != > < <= >=&\cr
4811 &8&&left&&{\&\&}&\cr
4814 &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr
4816 height2pt&\omit&&\omit&&\omit&\cr}
4821 @obeylines@parskip=0pt@parindent=0pt
4822 @dag@quad Prefix operators.
4823 @ddag@quad @xref{Assignments}.
4826 @c END TEXI2ROFF-KILL
4829 @subsection Evaluation
4830 @cindex lazy evaluation
4831 @cindex expression evaluation order
4832 The linker evaluates expressions lazily. It only computes the value of
4833 an expression when absolutely necessary.
4835 The linker needs some information, such as the value of the start
4836 address of the first section, and the origins and lengths of memory
4837 regions, in order to do any linking at all. These values are computed
4838 as soon as possible when the linker reads in the linker script.
4840 However, other values (such as symbol values) are not known or needed
4841 until after storage allocation. Such values are evaluated later, when
4842 other information (such as the sizes of output sections) is available
4843 for use in the symbol assignment expression.
4845 The sizes of sections cannot be known until after allocation, so
4846 assignments dependent upon these are not performed until after
4849 Some expressions, such as those depending upon the location counter
4850 @samp{.}, must be evaluated during section allocation.
4852 If the result of an expression is required, but the value is not
4853 available, then an error results. For example, a script like the
4859 .text 9+this_isnt_constant :
4865 will cause the error message @samp{non constant expression for initial
4868 @node Expression Section
4869 @subsection The Section of an Expression
4870 @cindex expression sections
4871 @cindex absolute expressions
4872 @cindex relative expressions
4873 @cindex absolute and relocatable symbols
4874 @cindex relocatable and absolute symbols
4875 @cindex symbols, relocatable and absolute
4876 When the linker evaluates an expression, the result is either absolute
4877 or relative to some section. A relative expression is expressed as a
4878 fixed offset from the base of a section.
4880 The position of the expression within the linker script determines
4881 whether it is absolute or relative. An expression which appears within
4882 an output section definition is relative to the base of the output
4883 section. An expression which appears elsewhere will be absolute.
4885 A symbol set to a relative expression will be relocatable if you request
4886 relocatable output using the @samp{-r} option. That means that a
4887 further link operation may change the value of the symbol. The symbol's
4888 section will be the section of the relative expression.
4890 A symbol set to an absolute expression will retain the same value
4891 through any further link operation. The symbol will be absolute, and
4892 will not have any particular associated section.
4894 You can use the builtin function @code{ABSOLUTE} to force an expression
4895 to be absolute when it would otherwise be relative. For example, to
4896 create an absolute symbol set to the address of the end of the output
4897 section @samp{.data}:
4901 .data : @{ *(.data) _edata = ABSOLUTE(.); @}
4905 If @samp{ABSOLUTE} were not used, @samp{_edata} would be relative to the
4906 @samp{.data} section.
4908 @node Builtin Functions
4909 @subsection Builtin Functions
4910 @cindex functions in expressions
4911 The linker script language includes a number of builtin functions for
4912 use in linker script expressions.
4915 @item ABSOLUTE(@var{exp})
4916 @kindex ABSOLUTE(@var{exp})
4917 @cindex expression, absolute
4918 Return the absolute (non-relocatable, as opposed to non-negative) value
4919 of the expression @var{exp}. Primarily useful to assign an absolute
4920 value to a symbol within a section definition, where symbol values are
4921 normally section relative. @xref{Expression Section}.
4923 @item ADDR(@var{section})
4924 @kindex ADDR(@var{section})
4925 @cindex section address in expression
4926 Return the absolute address (the VMA) of the named @var{section}. Your
4927 script must previously have defined the location of that section. In
4928 the following example, @code{symbol_1} and @code{symbol_2} are assigned
4935 start_of_output_1 = ABSOLUTE(.);
4940 symbol_1 = ADDR(.output1);
4941 symbol_2 = start_of_output_1;
4947 @item ALIGN(@var{align})
4948 @itemx ALIGN(@var{exp},@var{align})
4949 @kindex ALIGN(@var{align})
4950 @kindex ALIGN(@var{exp},@var{align})
4951 @cindex round up location counter
4952 @cindex align location counter
4953 @cindex round up expression
4954 @cindex align expression
4955 Return the location counter (@code{.}) or arbitrary expression aligned
4956 to the next @var{align} boundary. The single operand @code{ALIGN}
4957 doesn't change the value of the location counter---it just does
4958 arithmetic on it. The two operand @code{ALIGN} allows an arbitrary
4959 expression to be aligned upwards (@code{ALIGN(@var{align})} is
4960 equivalent to @code{ALIGN(., @var{align})}).
4962 Here is an example which aligns the output @code{.data} section to the
4963 next @code{0x2000} byte boundary after the preceding section and sets a
4964 variable within the section to the next @code{0x8000} boundary after the
4969 .data ALIGN(0x2000): @{
4971 variable = ALIGN(0x8000);
4977 The first use of @code{ALIGN} in this example specifies the location of
4978 a section because it is used as the optional @var{address} attribute of
4979 a section definition (@pxref{Output Section Address}). The second use
4980 of @code{ALIGN} is used to defines the value of a symbol.
4982 The builtin function @code{NEXT} is closely related to @code{ALIGN}.
4984 @item BLOCK(@var{exp})
4985 @kindex BLOCK(@var{exp})
4986 This is a synonym for @code{ALIGN}, for compatibility with older linker
4987 scripts. It is most often seen when setting the address of an output
4990 @item DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
4991 @kindex DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize})
4992 This is equivalent to either
4994 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - 1)))
4998 (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - @var{commonpagesize})))
5001 depending on whether the latter uses fewer @var{commonpagesize} sized pages
5002 for the data segment (area between the result of this expression and
5003 @code{DATA_SEGMENT_END}) than the former or not.
5004 If the latter form is used, it means @var{commonpagesize} bytes of runtime
5005 memory will be saved at the expense of up to @var{commonpagesize} wasted
5006 bytes in the on-disk file.
5008 This expression can only be used directly in @code{SECTIONS} commands, not in
5009 any output section descriptions and only once in the linker script.
5010 @var{commonpagesize} should be less or equal to @var{maxpagesize} and should
5011 be the system page size the object wants to be optimized for (while still
5012 working on system page sizes up to @var{maxpagesize}).
5017 . = DATA_SEGMENT_ALIGN(0x10000, 0x2000);
5020 @item DATA_SEGMENT_END(@var{exp})
5021 @kindex DATA_SEGMENT_END(@var{exp})
5022 This defines the end of data segment for @code{DATA_SEGMENT_ALIGN}
5023 evaluation purposes.
5026 . = DATA_SEGMENT_END(.);
5029 @item DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
5030 @kindex DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp})
5031 This defines the end of the @code{PT_GNU_RELRO} segment when
5032 @samp{-z relro} option is used. Second argument is returned.
5033 When @samp{-z relro} option is not present, @code{DATA_SEGMENT_RELRO_END}
5034 does nothing, otherwise @code{DATA_SEGMENT_ALIGN} is padded so that
5035 @var{exp} + @var{offset} is aligned to the most commonly used page
5036 boundary for particular target. If present in the linker script,
5037 it must always come in between @code{DATA_SEGMENT_ALIGN} and
5038 @code{DATA_SEGMENT_END}.
5041 . = DATA_SEGMENT_RELRO_END(24, .);
5044 @item DEFINED(@var{symbol})
5045 @kindex DEFINED(@var{symbol})
5046 @cindex symbol defaults
5047 Return 1 if @var{symbol} is in the linker global symbol table and is
5048 defined before the statement using DEFINED in the script, otherwise
5049 return 0. You can use this function to provide
5050 default values for symbols. For example, the following script fragment
5051 shows how to set a global symbol @samp{begin} to the first location in
5052 the @samp{.text} section---but if a symbol called @samp{begin} already
5053 existed, its value is preserved:
5059 begin = DEFINED(begin) ? begin : . ;
5067 @item LENGTH(@var{memory})
5068 @kindex LENGTH(@var{memory})
5069 Return the length of the memory region named @var{memory}.
5071 @item LOADADDR(@var{section})
5072 @kindex LOADADDR(@var{section})
5073 @cindex section load address in expression
5074 Return the absolute LMA of the named @var{section}. This is normally
5075 the same as @code{ADDR}, but it may be different if the @code{AT}
5076 attribute is used in the output section definition (@pxref{Output
5080 @item MAX(@var{exp1}, @var{exp2})
5081 Returns the maximum of @var{exp1} and @var{exp2}.
5084 @item MIN(@var{exp1}, @var{exp2})
5085 Returns the minimum of @var{exp1} and @var{exp2}.
5087 @item NEXT(@var{exp})
5088 @kindex NEXT(@var{exp})
5089 @cindex unallocated address, next
5090 Return the next unallocated address that is a multiple of @var{exp}.
5091 This function is closely related to @code{ALIGN(@var{exp})}; unless you
5092 use the @code{MEMORY} command to define discontinuous memory for the
5093 output file, the two functions are equivalent.
5095 @item ORIGIN(@var{memory})
5096 @kindex ORIGIN(@var{memory})
5097 Return the origin of the memory region named @var{memory}.
5099 @item SEGMENT_START(@var{segment}, @var{default})
5100 @kindex SEGMENT_START(@var{segment}, @var{default})
5101 Return the base address of the named @var{segment}. If an explicit
5102 value has been given for this segment (with a command-line @samp{-T}
5103 option) that value will be returned; otherwise the value will be
5104 @var{default}. At present, the @samp{-T} command-line option can only
5105 be used to set the base address for the ``text'', ``data'', and
5106 ``bss'' sections, but you use @code{SEGMENT_START} with any segment
5109 @item SIZEOF(@var{section})
5110 @kindex SIZEOF(@var{section})
5111 @cindex section size
5112 Return the size in bytes of the named @var{section}, if that section has
5113 been allocated. If the section has not been allocated when this is
5114 evaluated, the linker will report an error. In the following example,
5115 @code{symbol_1} and @code{symbol_2} are assigned identical values:
5124 symbol_1 = .end - .start ;
5125 symbol_2 = SIZEOF(.output);
5130 @item SIZEOF_HEADERS
5131 @itemx sizeof_headers
5132 @kindex SIZEOF_HEADERS
5134 Return the size in bytes of the output file's headers. This is
5135 information which appears at the start of the output file. You can use
5136 this number when setting the start address of the first section, if you
5137 choose, to facilitate paging.
5139 @cindex not enough room for program headers
5140 @cindex program headers, not enough room
5141 When producing an ELF output file, if the linker script uses the
5142 @code{SIZEOF_HEADERS} builtin function, the linker must compute the
5143 number of program headers before it has determined all the section
5144 addresses and sizes. If the linker later discovers that it needs
5145 additional program headers, it will report an error @samp{not enough
5146 room for program headers}. To avoid this error, you must avoid using
5147 the @code{SIZEOF_HEADERS} function, or you must rework your linker
5148 script to avoid forcing the linker to use additional program headers, or
5149 you must define the program headers yourself using the @code{PHDRS}
5150 command (@pxref{PHDRS}).
5153 @node Implicit Linker Scripts
5154 @section Implicit Linker Scripts
5155 @cindex implicit linker scripts
5156 If you specify a linker input file which the linker can not recognize as
5157 an object file or an archive file, it will try to read the file as a
5158 linker script. If the file can not be parsed as a linker script, the
5159 linker will report an error.
5161 An implicit linker script will not replace the default linker script.
5163 Typically an implicit linker script would contain only symbol
5164 assignments, or the @code{INPUT}, @code{GROUP}, or @code{VERSION}
5167 Any input files read because of an implicit linker script will be read
5168 at the position in the command line where the implicit linker script was
5169 read. This can affect archive searching.
5172 @node Machine Dependent
5173 @chapter Machine Dependent Features
5175 @cindex machine dependencies
5176 @command{ld} has additional features on some platforms; the following
5177 sections describe them. Machines where @command{ld} has no additional
5178 functionality are not listed.
5182 * H8/300:: @command{ld} and the H8/300
5185 * i960:: @command{ld} and the Intel 960 family
5188 * ARM:: @command{ld} and the ARM family
5191 * HPPA ELF32:: @command{ld} and HPPA 32-bit ELF
5194 * MMIX:: @command{ld} and MMIX
5197 * MSP430:: @command{ld} and MSP430
5200 * M68HC11/68HC12:: @code{ld} and the Motorola 68HC11 and 68HC12 families
5203 * PowerPC ELF32:: @command{ld} and PowerPC 32-bit ELF Support
5206 * PowerPC64 ELF64:: @command{ld} and PowerPC64 64-bit ELF Support
5209 * TI COFF:: @command{ld} and TI COFF
5212 * WIN32:: @command{ld} and WIN32 (cygwin/mingw)
5215 * Xtensa:: @command{ld} and Xtensa Processors
5226 @section @command{ld} and the H8/300
5228 @cindex H8/300 support
5229 For the H8/300, @command{ld} can perform these global optimizations when
5230 you specify the @samp{--relax} command-line option.
5233 @cindex relaxing on H8/300
5234 @item relaxing address modes
5235 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
5236 targets are within eight bits, and turns them into eight-bit
5237 program-counter relative @code{bsr} and @code{bra} instructions,
5240 @cindex synthesizing on H8/300
5241 @item synthesizing instructions
5242 @c FIXME: specifically mov.b, or any mov instructions really?
5243 @command{ld} finds all @code{mov.b} instructions which use the
5244 sixteen-bit absolute address form, but refer to the top
5245 page of memory, and changes them to use the eight-bit address form.
5246 (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into
5247 @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the
5248 top page of memory).
5250 @item bit manipulation instructions
5251 @command{ld} finds all bit manipulation instructions like @code{band, bclr,
5252 biand, bild, bior, bist, bixor, bld, bnot, bor, bset, bst, btst, bxor}
5253 which use 32 bit and 16 bit absolute address form, but refer to the top
5254 page of memory, and changes them to use the 8 bit address form.
5255 (That is: the linker turns @samp{bset #xx:3,@code{@@}@var{aa}:32} into
5256 @samp{bset #xx:3,@code{@@}@var{aa}:8} whenever the address @var{aa} is in
5257 the top page of memory).
5259 @item system control instructions
5260 @command{ld} finds all @code{ldc.w, stc.w} instructions which use the
5261 32 bit absolute address form, but refer to the top page of memory, and
5262 changes them to use 16 bit address form.
5263 (That is: the linker turns @samp{ldc.w @code{@@}@var{aa}:32,ccr} into
5264 @samp{ldc.w @code{@@}@var{aa}:16,ccr} whenever the address @var{aa} is in
5265 the top page of memory).
5275 @c This stuff is pointless to say unless you're especially concerned
5276 @c with Renesas chips; don't enable it for generic case, please.
5278 @chapter @command{ld} and Other Renesas Chips
5280 @command{ld} also supports the Renesas (formerly Hitachi) H8/300H,
5281 H8/500, and SH chips. No special features, commands, or command-line
5282 options are required for these chips.
5292 @section @command{ld} and the Intel 960 Family
5294 @cindex i960 support
5296 You can use the @samp{-A@var{architecture}} command line option to
5297 specify one of the two-letter names identifying members of the 960
5298 family; the option specifies the desired output target, and warns of any
5299 incompatible instructions in the input files. It also modifies the
5300 linker's search strategy for archive libraries, to support the use of
5301 libraries specific to each particular architecture, by including in the
5302 search loop names suffixed with the string identifying the architecture.
5304 For example, if your @command{ld} command line included @w{@samp{-ACA}} as
5305 well as @w{@samp{-ltry}}, the linker would look (in its built-in search
5306 paths, and in any paths you specify with @samp{-L}) for a library with
5319 The first two possibilities would be considered in any event; the last
5320 two are due to the use of @w{@samp{-ACA}}.
5322 You can meaningfully use @samp{-A} more than once on a command line, since
5323 the 960 architecture family allows combination of target architectures; each
5324 use will add another pair of name variants to search for when @w{@samp{-l}}
5325 specifies a library.
5327 @cindex @option{--relax} on i960
5328 @cindex relaxing on i960
5329 @command{ld} supports the @samp{--relax} option for the i960 family. If
5330 you specify @samp{--relax}, @command{ld} finds all @code{balx} and
5331 @code{calx} instructions whose targets are within 24 bits, and turns
5332 them into 24-bit program-counter relative @code{bal} and @code{cal}
5333 instructions, respectively. @command{ld} also turns @code{cal}
5334 instructions into @code{bal} instructions when it determines that the
5335 target subroutine is a leaf routine (that is, the target subroutine does
5336 not itself call any subroutines).
5353 @node M68HC11/68HC12
5354 @section @command{ld} and the Motorola 68HC11 and 68HC12 families
5356 @cindex M68HC11 and 68HC12 support
5358 @subsection Linker Relaxation
5360 For the Motorola 68HC11, @command{ld} can perform these global
5361 optimizations when you specify the @samp{--relax} command-line option.
5364 @cindex relaxing on M68HC11
5365 @item relaxing address modes
5366 @command{ld} finds all @code{jsr} and @code{jmp} instructions whose
5367 targets are within eight bits, and turns them into eight-bit
5368 program-counter relative @code{bsr} and @code{bra} instructions,
5371 @command{ld} also looks at all 16-bit extended addressing modes and
5372 transforms them in a direct addressing mode when the address is in
5373 page 0 (between 0 and 0x0ff).
5375 @item relaxing gcc instruction group
5376 When @command{gcc} is called with @option{-mrelax}, it can emit group
5377 of instructions that the linker can optimize to use a 68HC11 direct
5378 addressing mode. These instructions consists of @code{bclr} or
5379 @code{bset} instructions.
5383 @subsection Trampoline Generation
5385 @cindex trampoline generation on M68HC11
5386 @cindex trampoline generation on M68HC12
5387 For 68HC11 and 68HC12, @command{ld} can generate trampoline code to
5388 call a far function using a normal @code{jsr} instruction. The linker
5389 will also change the relocation to some far function to use the
5390 trampoline address instead of the function address. This is typically the
5391 case when a pointer to a function is taken. The pointer will in fact
5392 point to the function trampoline.
5400 @section @command{ld} and the ARM family
5402 @cindex ARM interworking support
5403 @kindex --support-old-code
5404 For the ARM, @command{ld} will generate code stubs to allow functions calls
5405 between ARM and Thumb code. These stubs only work with code that has
5406 been compiled and assembled with the @samp{-mthumb-interwork} command
5407 line option. If it is necessary to link with old ARM object files or
5408 libraries, which have not been compiled with the -mthumb-interwork
5409 option then the @samp{--support-old-code} command line switch should be
5410 given to the linker. This will make it generate larger stub functions
5411 which will work with non-interworking aware ARM code. Note, however,
5412 the linker does not support generating stubs for function calls to
5413 non-interworking aware Thumb code.
5415 @cindex thumb entry point
5416 @cindex entry point, thumb
5417 @kindex --thumb-entry=@var{entry}
5418 The @samp{--thumb-entry} switch is a duplicate of the generic
5419 @samp{--entry} switch, in that it sets the program's starting address.
5420 But it also sets the bottom bit of the address, so that it can be
5421 branched to using a BX instruction, and the program will start
5422 executing in Thumb mode straight away.
5426 The @samp{--be8} switch instructs @command{ld} to generate BE8 format
5427 executables. This option is only valid when linking big-endian objects.
5428 The resulting image will contain big-endian data and little-endian code.
5431 @kindex --target1-rel
5432 @kindex --target1-abs
5433 The @samp{R_ARM_TARGET1} relocation is typically used for entries in the
5434 @samp{.init_array} section. It is interpreted as either @samp{R_ARM_REL32}
5435 or @samp{R_ARM_ABS32}, depending on the target. The @samp{--target1-rel}
5436 and @samp{--target1-abs} switches override the default.
5439 @kindex --target2=@var{type}
5440 The @samp{--target2=type} switch overrides the default definition of the
5441 @samp{R_ARM_TARGET2} relocation. Valid values for @samp{type}, their
5442 meanings, and target defaults are as follows:
5445 @samp{R_ARM_REL32} (arm*-*-elf, arm*-*-eabi)
5447 @samp{R_ARM_ABS32} (arm*-*-symbianelf)
5449 @samp{R_ARM_GOT_PREL} (arm*-*-linux, arm*-*-*bsd)
5454 The @samp{R_ARM_V4BX} relocation (defined by the ARM AAELF
5455 specification) enables objects compiled for the ARMv4 architecture to be
5456 interworking-safe when linked with other objects compiled for ARMv4t, but
5457 also allows pure ARMv4 binaries to be built from the same ARMv4 objects.
5459 In the latter case, the switch @option{--fix-v4bx} must be passed to the
5460 linker, which causes v4t @code{BX rM} instructions to be rewritten as
5461 @code{MOV PC,rM}, since v4 processors do not have a @code{BX} instruction.
5463 In the former case, the switch should not be used, and @samp{R_ARM_V4BX}
5464 relocations are ignored.
5468 The @samp{--use-blx} switch enables the linker to use ARM/Thumb
5469 BLX instructions (available on ARMv5t and above) in various
5470 situations. Currently it is used to perform calls via the PLT from Thumb
5471 code using BLX rather than using BX and a mode-switching stub before
5472 each PLT entry. This should lead to such calls executing slightly faster.
5474 This option is enabled implicitly for SymbianOS, so there is no need to
5475 specify it if you are using that target.
5488 @section @command{ld} and HPPA 32-bit ELF Support
5489 @cindex HPPA multiple sub-space stubs
5490 @kindex --multi-subspace
5491 When generating a shared library, @command{ld} will by default generate
5492 import stubs suitable for use with a single sub-space application.
5493 The @samp{--multi-subspace} switch causes @command{ld} to generate export
5494 stubs, and different (larger) import stubs suitable for use with
5495 multiple sub-spaces.
5497 @cindex HPPA stub grouping
5498 @kindex --stub-group-size=@var{N}
5499 Long branch stubs and import/export stubs are placed by @command{ld} in
5500 stub sections located between groups of input sections.
5501 @samp{--stub-group-size} specifies the maximum size of a group of input
5502 sections handled by one stub section. Since branch offsets are signed,
5503 a stub section may serve two groups of input sections, one group before
5504 the stub section, and one group after it. However, when using
5505 conditional branches that require stubs, it may be better (for branch
5506 prediction) that stub sections only serve one group of input sections.
5507 A negative value for @samp{N} chooses this scheme, ensuring that
5508 branches to stubs always use a negative offset. Two special values of
5509 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
5510 @command{ld} to automatically size input section groups for the branch types
5511 detected, with the same behaviour regarding stub placement as other
5512 positive or negative values of @samp{N} respectively.
5514 Note that @samp{--stub-group-size} does not split input sections. A
5515 single input section larger than the group size specified will of course
5516 create a larger group (of one section). If input sections are too
5517 large, it may not be possible for a branch to reach its stub.
5530 @section @code{ld} and MMIX
5531 For MMIX, there is a choice of generating @code{ELF} object files or
5532 @code{mmo} object files when linking. The simulator @code{mmix}
5533 understands the @code{mmo} format. The binutils @code{objcopy} utility
5534 can translate between the two formats.
5536 There is one special section, the @samp{.MMIX.reg_contents} section.
5537 Contents in this section is assumed to correspond to that of global
5538 registers, and symbols referring to it are translated to special symbols,
5539 equal to registers. In a final link, the start address of the
5540 @samp{.MMIX.reg_contents} section corresponds to the first allocated
5541 global register multiplied by 8. Register @code{$255} is not included in
5542 this section; it is always set to the program entry, which is at the
5543 symbol @code{Main} for @code{mmo} files.
5545 Symbols with the prefix @code{__.MMIX.start.}, for example
5546 @code{__.MMIX.start..text} and @code{__.MMIX.start..data} are special;
5547 there must be only one each, even if they are local. The default linker
5548 script uses these to set the default start address of a section.
5550 Initial and trailing multiples of zero-valued 32-bit words in a section,
5551 are left out from an mmo file.
5564 @section @code{ld} and MSP430
5565 For the MSP430 it is possible to select the MPU architecture. The flag @samp{-m [mpu type]}
5566 will select an appropriate linker script for selected MPU type. (To get a list of known MPUs
5567 just pass @samp{-m help} option to the linker).
5569 @cindex MSP430 extra sections
5570 The linker will recognize some extra sections which are MSP430 specific:
5573 @item @samp{.vectors}
5574 Defines a portion of ROM where interrupt vectors located.
5576 @item @samp{.bootloader}
5577 Defines the bootloader portion of the ROM (if applicable). Any code
5578 in this section will be uploaded to the MPU.
5580 @item @samp{.infomem}
5581 Defines an information memory section (if applicable). Any code in
5582 this section will be uploaded to the MPU.
5584 @item @samp{.infomemnobits}
5585 This is the same as the @samp{.infomem} section except that any code
5586 in this section will not be uploaded to the MPU.
5588 @item @samp{.noinit}
5589 Denotes a portion of RAM located above @samp{.bss} section.
5591 The last two sections are used by gcc.
5605 @section @command{ld} and PowerPC 32-bit ELF Support
5606 @cindex PowerPC long branches
5607 @kindex --relax on PowerPC
5608 Branches on PowerPC processors are limited to a signed 26-bit
5609 displacement, which may result in @command{ld} giving
5610 @samp{relocation truncated to fit} errors with very large programs.
5611 @samp{--relax} enables the generation of trampolines that can access
5612 the entire 32-bit address space. These trampolines are inserted at
5613 section boundaries, so may not themselves be reachable if an input
5614 section exceeds 33M in size.
5616 @cindex PowerPC ELF32 options
5621 Current PowerPC GCC accepts a @samp{-msecure-plt} option that
5622 generates code capable of using a newer PLT and GOT layout that has
5623 the security advantage of no executable section ever needing to be
5624 writable and no writable section ever being executable. PowerPC
5625 @command{ld} will generate this layout, including stubs to access the
5626 PLT, if all input files (including startup and static libraries) were
5627 compiled with @samp{-msecure-plt}. @samp{--bss-plt} forces the old
5628 BSS PLT (and GOT layout) which can give slightly better performance.
5633 The new secure PLT and GOT are placed differently relative to other
5634 sections compared to older BSS PLT and GOT placement. The location of
5635 @code{.plt} must change because the new secure PLT is an initialized
5636 section while the old PLT is uninitialized. The reason for the
5637 @code{.got} change is more subtle: The new placement allows
5638 @code{.got} to be read-only in applications linked with
5639 @samp{-z relro -z now}. However, this placement means that
5640 @code{.sdata} cannot always be used in shared libraries, because the
5641 PowerPC ABI accesses @code{.sdata} in shared libraries from the GOT
5642 pointer. @samp{--sdata-got} forces the old GOT placement. PowerPC
5643 GCC doesn't use @code{.sdata} in shared libraries, so this option is
5644 really only useful for other compilers that may do so.
5646 @cindex PowerPC stub symbols
5647 @kindex --emit-stub-syms
5648 @item --emit-stub-syms
5649 This option causes @command{ld} to label linker stubs with a local
5650 symbol that encodes the stub type and destination.
5652 @cindex PowerPC TLS optimization
5653 @kindex --no-tls-optimize
5654 @item --no-tls-optimize
5655 PowerPC @command{ld} normally performs some optimization of code
5656 sequences used to access Thread-Local Storage. Use this option to
5657 disable the optimization.
5670 @node PowerPC64 ELF64
5671 @section @command{ld} and PowerPC64 64-bit ELF Support
5673 @cindex PowerPC64 ELF64 options
5675 @cindex PowerPC64 stub grouping
5676 @kindex --stub-group-size
5677 @item --stub-group-size
5678 Long branch stubs, PLT call stubs and TOC adjusting stubs are placed
5679 by @command{ld} in stub sections located between groups of input sections.
5680 @samp{--stub-group-size} specifies the maximum size of a group of input
5681 sections handled by one stub section. Since branch offsets are signed,
5682 a stub section may serve two groups of input sections, one group before
5683 the stub section, and one group after it. However, when using
5684 conditional branches that require stubs, it may be better (for branch
5685 prediction) that stub sections only serve one group of input sections.
5686 A negative value for @samp{N} chooses this scheme, ensuring that
5687 branches to stubs always use a negative offset. Two special values of
5688 @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct
5689 @command{ld} to automatically size input section groups for the branch types
5690 detected, with the same behaviour regarding stub placement as other
5691 positive or negative values of @samp{N} respectively.
5693 Note that @samp{--stub-group-size} does not split input sections. A
5694 single input section larger than the group size specified will of course
5695 create a larger group (of one section). If input sections are too
5696 large, it may not be possible for a branch to reach its stub.
5698 @cindex PowerPC64 stub symbols
5699 @kindex --emit-stub-syms
5700 @item --emit-stub-syms
5701 This option causes @command{ld} to label linker stubs with a local
5702 symbol that encodes the stub type and destination.
5704 @cindex PowerPC64 dot symbols
5706 @kindex --no-dotsyms
5707 @item --dotsyms, --no-dotsyms
5708 These two options control how @command{ld} interprets version patterns
5709 in a version script. Older PowerPC64 compilers emitted both a
5710 function descriptor symbol with the same name as the function, and a
5711 code entry symbol with the name prefixed by a dot (@samp{.}). To
5712 properly version a function @samp{foo}, the version script thus needs
5713 to control both @samp{foo} and @samp{.foo}. The option
5714 @samp{--dotsyms}, on by default, automatically adds the required
5715 dot-prefixed patterns. Use @samp{--no-dotsyms} to disable this
5718 @cindex PowerPC64 TLS optimization
5719 @kindex --no-tls-optimize
5720 @item --no-tls-optimize
5721 PowerPC64 @command{ld} normally performs some optimization of code
5722 sequences used to access Thread-Local Storage. Use this option to
5723 disable the optimization.
5725 @cindex PowerPC64 OPD optimization
5726 @kindex --no-opd-optimize
5727 @item --no-opd-optimize
5728 PowerPC64 @command{ld} normally removes @code{.opd} section entries
5729 corresponding to deleted link-once functions, or functions removed by
5730 the action of @samp{--gc-sections} or linker scrip @code{/DISCARD/}.
5731 Use this option to disable @code{.opd} optimization.
5733 @cindex PowerPC64 OPD spacing
5734 @kindex --non-overlapping-opd
5735 @item --non-overlapping-opd
5736 Some PowerPC64 compilers have an option to generate compressed
5737 @code{.opd} entries spaced 16 bytes apart, overlapping the third word,
5738 the static chain pointer (unused in C) with the first word of the next
5739 entry. This option expands such entries to the full 24 bytes.
5741 @cindex PowerPC64 TOC optimization
5742 @kindex --no-toc-optimize
5743 @item --no-toc-optimize
5744 PowerPC64 @command{ld} normally removes unused @code{.toc} section
5745 entries. Such entries are detected by examining relocations that
5746 reference the TOC in code sections. A reloc in a deleted code section
5747 marks a TOC word as unneeded, while a reloc in a kept code section
5748 marks a TOC word as needed. Since the TOC may reference itself, TOC
5749 relocs are also examined. TOC words marked as both needed and
5750 unneeded will of course be kept. TOC words without any referencing
5751 reloc are assumed to be part of a multi-word entry, and are kept or
5752 discarded as per the nearest marked preceding word. This works
5753 reliably for compiler generated code, but may be incorrect if assembly
5754 code is used to insert TOC entries. Use this option to disable the
5757 @cindex PowerPC64 multi-TOC
5758 @kindex --no-multi-toc
5759 @item --no-multi-toc
5760 By default, PowerPC64 GCC generates code for a TOC model where TOC
5761 entries are accessed with a 16-bit offset from r2. This limits the
5762 total TOC size to 64K. PowerPC64 @command{ld} extends this limit by
5763 grouping code sections such that each group uses less than 64K for its
5764 TOC entries, then inserts r2 adjusting stubs between inter-group
5765 calls. @command{ld} does not split apart input sections, so cannot
5766 help if a single input file has a @code{.toc} section that exceeds
5767 64K, most likely from linking multiple files with @command{ld -r}.
5768 Use this option to turn off this feature.
5782 @section @command{ld}'s Support for Various TI COFF Versions
5783 @cindex TI COFF versions
5784 @kindex --format=@var{version}
5785 The @samp{--format} switch allows selection of one of the various
5786 TI COFF versions. The latest of this writing is 2; versions 0 and 1 are
5787 also supported. The TI COFF versions also vary in header byte-order
5788 format; @command{ld} will read any version or byte order, but the output
5789 header format depends on the default specified by the specific target.
5802 @section @command{ld} and WIN32 (cygwin/mingw)
5804 This section describes some of the win32 specific @command{ld} issues.
5805 See @ref{Options,,Command Line Options} for detailed description of the
5806 command line options mentioned here.
5809 @cindex import libraries
5810 @item import libraries
5811 The standard Windows linker creates and uses so-called import
5812 libraries, which contains information for linking to dll's. They are
5813 regular static archives and are handled as any other static
5814 archive. The cygwin and mingw ports of @command{ld} have specific
5815 support for creating such libraries provided with the
5816 @samp{--out-implib} command line option.
5818 @item exporting DLL symbols
5819 @cindex exporting DLL symbols
5820 The cygwin/mingw @command{ld} has several ways to export symbols for dll's.
5823 @item using auto-export functionality
5824 @cindex using auto-export functionality
5825 By default @command{ld} exports symbols with the auto-export functionality,
5826 which is controlled by the following command line options:
5829 @item --export-all-symbols [This is the default]
5830 @item --exclude-symbols
5831 @item --exclude-libs
5834 If, however, @samp{--export-all-symbols} is not given explicitly on the
5835 command line, then the default auto-export behavior will be @emph{disabled}
5836 if either of the following are true:
5839 @item A DEF file is used.
5840 @item Any symbol in any object file was marked with the __declspec(dllexport) attribute.
5843 @item using a DEF file
5844 @cindex using a DEF file
5845 Another way of exporting symbols is using a DEF file. A DEF file is
5846 an ASCII file containing definitions of symbols which should be
5847 exported when a dll is created. Usually it is named @samp{<dll
5848 name>.def} and is added as any other object file to the linker's
5849 command line. The file's name must end in @samp{.def} or @samp{.DEF}.
5852 gcc -o <output> <objectfiles> <dll name>.def
5855 Using a DEF file turns off the normal auto-export behavior, unless the
5856 @samp{--export-all-symbols} option is also used.
5858 Here is an example of a DEF file for a shared library called @samp{xyz.dll}:
5861 LIBRARY "xyz.dll" BASE=0x20000000
5867 another_foo = abc.dll.afoo
5871 This example defines a DLL with a non-default base address and five
5872 symbols in the export table. The third exported symbol @code{_bar} is an
5873 alias for the second. The fourth symbol, @code{another_foo} is resolved
5874 by "forwarding" to another module and treating it as an alias for
5875 @code{afoo} exported from the DLL @samp{abc.dll}. The final symbol
5876 @code{var1} is declared to be a data object.
5878 The optional @code{LIBRARY <name>} command indicates the @emph{internal}
5879 name of the output DLL. If @samp{<name>} does not include a suffix,
5880 the default library suffix, @samp{.DLL} is appended.
5882 When the .DEF file is used to build an application, rather than a
5883 library, the @code{NAME <name>} command should be used instead of
5884 @code{LIBRARY}. If @samp{<name>} does not include a suffix, the default
5885 executable suffix, @samp{.EXE} is appended.
5887 With either @code{LIBRARY <name>} or @code{NAME <name>} the optional
5888 specification @code{BASE = <number>} may be used to specify a
5889 non-default base address for the image.
5891 If neither @code{LIBRARY <name>} nor @code{NAME <name>} is specified,
5892 or they specify an empty string, the internal name is the same as the
5893 filename specified on the command line.
5895 The complete specification of an export symbol is:
5899 ( ( ( <name1> [ = <name2> ] )
5900 | ( <name1> = <module-name> . <external-name>))
5901 [ @@ <integer> ] [NONAME] [DATA] [CONSTANT] [PRIVATE] ) *
5904 Declares @samp{<name1>} as an exported symbol from the DLL, or declares
5905 @samp{<name1>} as an exported alias for @samp{<name2>}; or declares
5906 @samp{<name1>} as a "forward" alias for the symbol
5907 @samp{<external-name>} in the DLL @samp{<module-name>}.
5908 Optionally, the symbol may be exported by the specified ordinal
5909 @samp{<integer>} alias.
5911 The optional keywords that follow the declaration indicate:
5913 @code{NONAME}: Do not put the symbol name in the DLL's export table. It
5914 will still be exported by its ordinal alias (either the value specified
5915 by the .def specification or, otherwise, the value assigned by the
5916 linker). The symbol name, however, does remain visible in the import
5917 library (if any), unless @code{PRIVATE} is also specified.
5919 @code{DATA}: The symbol is a variable or object, rather than a function.
5920 The import lib will export only an indirect reference to @code{foo} as
5921 the symbol @code{_imp__foo} (ie, @code{foo} must be resolved as
5924 @code{CONSTANT}: Like @code{DATA}, but put the undecorated @code{foo} as
5925 well as @code{_imp__foo} into the import library. Both refer to the
5926 read-only import address table's pointer to the variable, not to the
5927 variable itself. This can be dangerous. If the user code fails to add
5928 the @code{dllimport} attribute and also fails to explicitly add the
5929 extra indirection that the use of the attribute enforces, the
5930 application will behave unexpectedly.
5932 @code{PRIVATE}: Put the symbol in the DLL's export table, but do not put
5933 it into the static import library used to resolve imports at link time. The
5934 symbol can still be imported using the @code{LoadLibrary/GetProcAddress}
5935 API at runtime or by by using the GNU ld extension of linking directly to
5936 the DLL without an import library.
5938 See ld/deffilep.y in the binutils sources for the full specification of
5939 other DEF file statements
5941 @cindex creating a DEF file
5942 While linking a shared dll, @command{ld} is able to create a DEF file
5943 with the @samp{--output-def <file>} command line option.
5945 @item Using decorations
5946 @cindex Using decorations
5947 Another way of marking symbols for export is to modify the source code
5948 itself, so that when building the DLL each symbol to be exported is
5952 __declspec(dllexport) int a_variable
5953 __declspec(dllexport) void a_function(int with_args)
5956 All such symbols will be exported from the DLL. If, however,
5957 any of the object files in the DLL contain symbols decorated in
5958 this way, then the normal auto-export behavior is disabled, unless
5959 the @samp{--export-all-symbols} option is also used.
5961 Note that object files that wish to access these symbols must @emph{not}
5962 decorate them with dllexport. Instead, they should use dllimport,
5966 __declspec(dllimport) int a_variable
5967 __declspec(dllimport) void a_function(int with_args)
5970 This complicates the structure of library header files, because
5971 when included by the library itself the header must declare the
5972 variables and functions as dllexport, but when included by client
5973 code the header must declare them as dllimport. There are a number
5974 of idioms that are typically used to do this; often client code can
5975 omit the __declspec() declaration completely. See
5976 @samp{--enable-auto-import} and @samp{automatic data imports} for more
5980 @cindex automatic data imports
5981 @item automatic data imports
5982 The standard Windows dll format supports data imports from dlls only
5983 by adding special decorations (dllimport/dllexport), which let the
5984 compiler produce specific assembler instructions to deal with this
5985 issue. This increases the effort necessary to port existing Un*x
5986 code to these platforms, especially for large
5987 c++ libraries and applications. The auto-import feature, which was
5988 initially provided by Paul Sokolovsky, allows one to omit the
5989 decorations to achieve a behavior that conforms to that on POSIX/Un*x
5990 platforms. This feature is enabled with the @samp{--enable-auto-import}
5991 command-line option, although it is enabled by default on cygwin/mingw.
5992 The @samp{--enable-auto-import} option itself now serves mainly to
5993 suppress any warnings that are ordinarily emitted when linked objects
5994 trigger the feature's use.
5996 auto-import of variables does not always work flawlessly without
5997 additional assistance. Sometimes, you will see this message
5999 "variable '<var>' can't be auto-imported. Please read the
6000 documentation for ld's @code{--enable-auto-import} for details."
6002 The @samp{--enable-auto-import} documentation explains why this error
6003 occurs, and several methods that can be used to overcome this difficulty.
6004 One of these methods is the @emph{runtime pseudo-relocs} feature, described
6007 @cindex runtime pseudo-relocation
6008 For complex variables imported from DLLs (such as structs or classes),
6009 object files typically contain a base address for the variable and an
6010 offset (@emph{addend}) within the variable--to specify a particular
6011 field or public member, for instance. Unfortunately, the runtime loader used
6012 in win32 environments is incapable of fixing these references at runtime
6013 without the additional information supplied by dllimport/dllexport decorations.
6014 The standard auto-import feature described above is unable to resolve these
6017 The @samp{--enable-runtime-pseudo-relocs} switch allows these references to
6018 be resolved without error, while leaving the task of adjusting the references
6019 themselves (with their non-zero addends) to specialized code provided by the
6020 runtime environment. Recent versions of the cygwin and mingw environments and
6021 compilers provide this runtime support; older versions do not. However, the
6022 support is only necessary on the developer's platform; the compiled result will
6023 run without error on an older system.
6025 @samp{--enable-runtime-pseudo-relocs} is not the default; it must be explicitly
6028 @cindex direct linking to a dll
6029 @item direct linking to a dll
6030 The cygwin/mingw ports of @command{ld} support the direct linking,
6031 including data symbols, to a dll without the usage of any import
6032 libraries. This is much faster and uses much less memory than does the
6033 traditional import library method, especially when linking large
6034 libraries or applications. When @command{ld} creates an import lib, each
6035 function or variable exported from the dll is stored in its own bfd, even
6036 though a single bfd could contain many exports. The overhead involved in
6037 storing, loading, and processing so many bfd's is quite large, and explains the
6038 tremendous time, memory, and storage needed to link against particularly
6039 large or complex libraries when using import libs.
6041 Linking directly to a dll uses no extra command-line switches other than
6042 @samp{-L} and @samp{-l}, because @command{ld} already searches for a number
6043 of names to match each library. All that is needed from the developer's
6044 perspective is an understanding of this search, in order to force ld to
6045 select the dll instead of an import library.
6048 For instance, when ld is called with the argument @samp{-lxxx} it will attempt
6049 to find, in the first directory of its search path,
6061 before moving on to the next directory in the search path.
6063 (*) Actually, this is not @samp{cygxxx.dll} but in fact is @samp{<prefix>xxx.dll},
6064 where @samp{<prefix>} is set by the @command{ld} option
6065 @samp{--dll-search-prefix=<prefix>}. In the case of cygwin, the standard gcc spec
6066 file includes @samp{--dll-search-prefix=cyg}, so in effect we actually search for
6069 Other win32-based unix environments, such as mingw or pw32, may use other
6070 @samp{<prefix>}es, although at present only cygwin makes use of this feature. It
6071 was originally intended to help avoid name conflicts among dll's built for the
6072 various win32/un*x environments, so that (for example) two versions of a zlib dll
6073 could coexist on the same machine.
6075 The generic cygwin/mingw path layout uses a @samp{bin} directory for
6076 applications and dll's and a @samp{lib} directory for the import
6077 libraries (using cygwin nomenclature):
6083 libxxx.dll.a (in case of dll's)
6084 libxxx.a (in case of static archive)
6087 Linking directly to a dll without using the import library can be
6090 1. Use the dll directly by adding the @samp{bin} path to the link line
6092 gcc -Wl,-verbose -o a.exe -L../bin/ -lxxx
6095 However, as the dll's often have version numbers appended to their names
6096 (@samp{cygncurses-5.dll}) this will often fail, unless one specifies
6097 @samp{-L../bin -lncurses-5} to include the version. Import libs are generally
6098 not versioned, and do not have this difficulty.
6100 2. Create a symbolic link from the dll to a file in the @samp{lib}
6101 directory according to the above mentioned search pattern. This
6102 should be used to avoid unwanted changes in the tools needed for
6106 ln -s bin/cygxxx.dll lib/[cyg|lib|]xxx.dll[.a]
6109 Then you can link without any make environment changes.
6112 gcc -Wl,-verbose -o a.exe -L../lib/ -lxxx
6115 This technique also avoids the version number problems, because the following is
6122 libxxx.dll.a -> ../bin/cygxxx-5.dll
6125 Linking directly to a dll without using an import lib will work
6126 even when auto-import features are exercised, and even when
6127 @samp{--enable-runtime-pseudo-relocs} is used.
6129 Given the improvements in speed and memory usage, one might justifiably
6130 wonder why import libraries are used at all. There are three reasons:
6132 1. Until recently, the link-directly-to-dll functionality did @emph{not}
6133 work with auto-imported data.
6135 2. Sometimes it is necessary to include pure static objects within the
6136 import library (which otherwise contains only bfd's for indirection
6137 symbols that point to the exports of a dll). Again, the import lib
6138 for the cygwin kernel makes use of this ability, and it is not
6139 possible to do this without an import lib.
6141 3. Symbol aliases can only be resolved using an import lib. This is
6142 critical when linking against OS-supplied dll's (eg, the win32 API)
6143 in which symbols are usually exported as undecorated aliases of their
6144 stdcall-decorated assembly names.
6146 So, import libs are not going away. But the ability to replace
6147 true import libs with a simple symbolic link to (or a copy of)
6148 a dll, in many cases, is a useful addition to the suite of tools
6149 binutils makes available to the win32 developer. Given the
6150 massive improvements in memory requirements during linking, storage
6151 requirements, and linking speed, we expect that many developers
6152 will soon begin to use this feature whenever possible.
6154 @item symbol aliasing
6156 @item adding additional names
6157 Sometimes, it is useful to export symbols with additional names.
6158 A symbol @samp{foo} will be exported as @samp{foo}, but it can also be
6159 exported as @samp{_foo} by using special directives in the DEF file
6160 when creating the dll. This will affect also the optional created
6161 import library. Consider the following DEF file:
6164 LIBRARY "xyz.dll" BASE=0x61000000
6171 The line @samp{_foo = foo} maps the symbol @samp{foo} to @samp{_foo}.
6173 Another method for creating a symbol alias is to create it in the
6174 source code using the "weak" attribute:
6177 void foo () @{ /* Do something. */; @}
6178 void _foo () __attribute__ ((weak, alias ("foo")));
6181 See the gcc manual for more information about attributes and weak
6184 @item renaming symbols
6185 Sometimes it is useful to rename exports. For instance, the cygwin
6186 kernel does this regularly. A symbol @samp{_foo} can be exported as
6187 @samp{foo} but not as @samp{_foo} by using special directives in the
6188 DEF file. (This will also affect the import library, if it is
6189 created). In the following example:
6192 LIBRARY "xyz.dll" BASE=0x61000000
6198 The line @samp{_foo = foo} maps the exported symbol @samp{foo} to
6202 Note: using a DEF file disables the default auto-export behavior,
6203 unless the @samp{--export-all-symbols} command line option is used.
6204 If, however, you are trying to rename symbols, then you should list
6205 @emph{all} desired exports in the DEF file, including the symbols
6206 that are not being renamed, and do @emph{not} use the
6207 @samp{--export-all-symbols} option. If you list only the
6208 renamed symbols in the DEF file, and use @samp{--export-all-symbols}
6209 to handle the other symbols, then the both the new names @emph{and}
6210 the original names for the renamed symbols will be exported.
6211 In effect, you'd be aliasing those symbols, not renaming them,
6212 which is probably not what you wanted.
6214 @cindex weak externals
6215 @item weak externals
6216 The Windows object format, PE, specifies a form of weak symbols called
6217 weak externals. When a weak symbol is linked and the symbol is not
6218 defined, the weak symbol becomes an alias for some other symbol. There
6219 are three variants of weak externals:
6221 @item Definition is searched for in objects and libraries, historically
6222 called lazy externals.
6223 @item Definition is searched for only in other objects, not in libraries.
6224 This form is not presently implemented.
6225 @item No search; the symbol is an alias. This form is not presently
6228 As a GNU extension, weak symbols that do not specify an alternate symbol
6229 are supported. If the symbol is undefined when linking, the symbol
6230 uses a default value.
6244 @section @code{ld} and Xtensa Processors
6246 @cindex Xtensa processors
6247 The default @command{ld} behavior for Xtensa processors is to interpret
6248 @code{SECTIONS} commands so that lists of explicitly named sections in a
6249 specification with a wildcard file will be interleaved when necessary to
6250 keep literal pools within the range of PC-relative load offsets. For
6251 example, with the command:
6263 @command{ld} may interleave some of the @code{.literal}
6264 and @code{.text} sections from different object files to ensure that the
6265 literal pools are within the range of PC-relative load offsets. A valid
6266 interleaving might place the @code{.literal} sections from an initial
6267 group of files followed by the @code{.text} sections of that group of
6268 files. Then, the @code{.literal} sections from the rest of the files
6269 and the @code{.text} sections from the rest of the files would follow.
6271 @cindex @option{--relax} on Xtensa
6272 @cindex relaxing on Xtensa
6273 Relaxation is enabled by default for the Xtensa version of @command{ld} and
6274 provides two important link-time optimizations. The first optimization
6275 is to combine identical literal values to reduce code size. A redundant
6276 literal will be removed and all the @code{L32R} instructions that use it
6277 will be changed to reference an identical literal, as long as the
6278 location of the replacement literal is within the offset range of all
6279 the @code{L32R} instructions. The second optimization is to remove
6280 unnecessary overhead from assembler-generated ``longcall'' sequences of
6281 @code{L32R}/@code{CALLX@var{n}} when the target functions are within
6282 range of direct @code{CALL@var{n}} instructions.
6284 For each of these cases where an indirect call sequence can be optimized
6285 to a direct call, the linker will change the @code{CALLX@var{n}}
6286 instruction to a @code{CALL@var{n}} instruction, remove the @code{L32R}
6287 instruction, and remove the literal referenced by the @code{L32R}
6288 instruction if it is not used for anything else. Removing the
6289 @code{L32R} instruction always reduces code size but can potentially
6290 hurt performance by changing the alignment of subsequent branch targets.
6291 By default, the linker will always preserve alignments, either by
6292 switching some instructions between 24-bit encodings and the equivalent
6293 density instructions or by inserting a no-op in place of the @code{L32R}
6294 instruction that was removed. If code size is more important than
6295 performance, the @option{--size-opt} option can be used to prevent the
6296 linker from widening density instructions or inserting no-ops, except in
6297 a few cases where no-ops are required for correctness.
6299 The following Xtensa-specific command-line options can be used to
6302 @cindex Xtensa options
6306 Since the Xtensa version of @code{ld} enables the @option{--relax} option
6307 by default, the @option{--no-relax} option is provided to disable
6311 When optimizing indirect calls to direct calls, optimize for code size
6312 more than performance. With this option, the linker will not insert
6313 no-ops or widen density instructions to preserve branch target
6314 alignment. There may still be some cases where no-ops are required to
6315 preserve the correctness of the code.
6323 @ifclear SingleFormat
6328 @cindex object file management
6329 @cindex object formats available
6331 The linker accesses object and archive files using the BFD libraries.
6332 These libraries allow the linker to use the same routines to operate on
6333 object files whatever the object file format. A different object file
6334 format can be supported simply by creating a new BFD back end and adding
6335 it to the library. To conserve runtime memory, however, the linker and
6336 associated tools are usually configured to support only a subset of the
6337 object file formats available. You can use @code{objdump -i}
6338 (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to
6339 list all the formats available for your configuration.
6341 @cindex BFD requirements
6342 @cindex requirements for BFD
6343 As with most implementations, BFD is a compromise between
6344 several conflicting requirements. The major factor influencing
6345 BFD design was efficiency: any time used converting between
6346 formats is time which would not have been spent had BFD not
6347 been involved. This is partly offset by abstraction payback; since
6348 BFD simplifies applications and back ends, more time and care
6349 may be spent optimizing algorithms for a greater speed.
6351 One minor artifact of the BFD solution which you should bear in
6352 mind is the potential for information loss. There are two places where
6353 useful information can be lost using the BFD mechanism: during
6354 conversion and during output. @xref{BFD information loss}.
6357 * BFD outline:: How it works: an outline of BFD
6361 @section How It Works: An Outline of BFD
6362 @cindex opening object files
6363 @include bfdsumm.texi
6366 @node Reporting Bugs
6367 @chapter Reporting Bugs
6368 @cindex bugs in @command{ld}
6369 @cindex reporting bugs in @command{ld}
6371 Your bug reports play an essential role in making @command{ld} reliable.
6373 Reporting a bug may help you by bringing a solution to your problem, or
6374 it may not. But in any case the principal function of a bug report is
6375 to help the entire community by making the next version of @command{ld}
6376 work better. Bug reports are your contribution to the maintenance of
6379 In order for a bug report to serve its purpose, you must include the
6380 information that enables us to fix the bug.
6383 * Bug Criteria:: Have you found a bug?
6384 * Bug Reporting:: How to report bugs
6388 @section Have You Found a Bug?
6389 @cindex bug criteria
6391 If you are not sure whether you have found a bug, here are some guidelines:
6394 @cindex fatal signal
6395 @cindex linker crash
6396 @cindex crash of linker
6398 If the linker gets a fatal signal, for any input whatever, that is a
6399 @command{ld} bug. Reliable linkers never crash.
6401 @cindex error on valid input
6403 If @command{ld} produces an error message for valid input, that is a bug.
6405 @cindex invalid input
6407 If @command{ld} does not produce an error message for invalid input, that
6408 may be a bug. In the general case, the linker can not verify that
6409 object files are correct.
6412 If you are an experienced user of linkers, your suggestions for
6413 improvement of @command{ld} are welcome in any case.
6417 @section How to Report Bugs
6419 @cindex @command{ld} bugs, reporting
6421 A number of companies and individuals offer support for @sc{gnu}
6422 products. If you obtained @command{ld} from a support organization, we
6423 recommend you contact that organization first.
6425 You can find contact information for many support companies and
6426 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6429 Otherwise, send bug reports for @command{ld} to
6430 @samp{bug-binutils@@gnu.org}.
6432 The fundamental principle of reporting bugs usefully is this:
6433 @strong{report all the facts}. If you are not sure whether to state a
6434 fact or leave it out, state it!
6436 Often people omit facts because they think they know what causes the
6437 problem and assume that some details do not matter. Thus, you might
6438 assume that the name of a symbol you use in an example does not
6439 matter. Well, probably it does not, but one cannot be sure. Perhaps
6440 the bug is a stray memory reference which happens to fetch from the
6441 location where that name is stored in memory; perhaps, if the name
6442 were different, the contents of that location would fool the linker
6443 into doing the right thing despite the bug. Play it safe and give a
6444 specific, complete example. That is the easiest thing for you to do,
6445 and the most helpful.
6447 Keep in mind that the purpose of a bug report is to enable us to fix
6448 the bug if it is new to us. Therefore, always write your bug reports
6449 on the assumption that the bug has not been reported previously.
6451 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6452 bell?'' This cannot help us fix a bug, so it is basically useless. We
6453 respond by asking for enough details to enable us to investigate.
6454 You might as well expedite matters by sending them to begin with.
6456 To enable us to fix the bug, you should include all these things:
6460 The version of @command{ld}. @command{ld} announces it if you start it with
6461 the @samp{--version} argument.
6463 Without this, we will not know whether there is any point in looking for
6464 the bug in the current version of @command{ld}.
6467 Any patches you may have applied to the @command{ld} source, including any
6468 patches made to the @code{BFD} library.
6471 The type of machine you are using, and the operating system name and
6475 What compiler (and its version) was used to compile @command{ld}---e.g.
6479 The command arguments you gave the linker to link your example and
6480 observe the bug. To guarantee you will not omit something important,
6481 list them all. A copy of the Makefile (or the output from make) is
6484 If we were to try to guess the arguments, we would probably guess wrong
6485 and then we might not encounter the bug.
6488 A complete input file, or set of input files, that will reproduce the
6489 bug. It is generally most helpful to send the actual object files
6490 provided that they are reasonably small. Say no more than 10K. For
6491 bigger files you can either make them available by FTP or HTTP or else
6492 state that you are willing to send the object file(s) to whomever
6493 requests them. (Note - your email will be going to a mailing list, so
6494 we do not want to clog it up with large attachments). But small
6495 attachments are best.
6497 If the source files were assembled using @code{gas} or compiled using
6498 @code{gcc}, then it may be OK to send the source files rather than the
6499 object files. In this case, be sure to say exactly what version of
6500 @code{gas} or @code{gcc} was used to produce the object files. Also say
6501 how @code{gas} or @code{gcc} were configured.
6504 A description of what behavior you observe that you believe is
6505 incorrect. For example, ``It gets a fatal signal.''
6507 Of course, if the bug is that @command{ld} gets a fatal signal, then we
6508 will certainly notice it. But if the bug is incorrect output, we might
6509 not notice unless it is glaringly wrong. You might as well not give us
6510 a chance to make a mistake.
6512 Even if the problem you experience is a fatal signal, you should still
6513 say so explicitly. Suppose something strange is going on, such as, your
6514 copy of @command{ld} is out of sync, or you have encountered a bug in the
6515 C library on your system. (This has happened!) Your copy might crash
6516 and ours would not. If you told us to expect a crash, then when ours
6517 fails to crash, we would know that the bug was not happening for us. If
6518 you had not told us to expect a crash, then we would not be able to draw
6519 any conclusion from our observations.
6522 If you wish to suggest changes to the @command{ld} source, send us context
6523 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or
6524 @samp{-p} option. Always send diffs from the old file to the new file.
6525 If you even discuss something in the @command{ld} source, refer to it by
6526 context, not by line number.
6528 The line numbers in our development sources will not match those in your
6529 sources. Your line numbers would convey no useful information to us.
6532 Here are some things that are not necessary:
6536 A description of the envelope of the bug.
6538 Often people who encounter a bug spend a lot of time investigating
6539 which changes to the input file will make the bug go away and which
6540 changes will not affect it.
6542 This is often time consuming and not very useful, because the way we
6543 will find the bug is by running a single example under the debugger
6544 with breakpoints, not by pure deduction from a series of examples.
6545 We recommend that you save your time for something else.
6547 Of course, if you can find a simpler example to report @emph{instead}
6548 of the original one, that is a convenience for us. Errors in the
6549 output will be easier to spot, running under the debugger will take
6550 less time, and so on.
6552 However, simplification is not vital; if you do not want to do this,
6553 report the bug anyway and send us the entire test case you used.
6556 A patch for the bug.
6558 A patch for the bug does help us if it is a good one. But do not omit
6559 the necessary information, such as the test case, on the assumption that
6560 a patch is all we need. We might see problems with your patch and decide
6561 to fix the problem another way, or we might not understand it at all.
6563 Sometimes with a program as complicated as @command{ld} it is very hard to
6564 construct an example that will make the program follow a certain path
6565 through the code. If you do not send us the example, we will not be
6566 able to construct one, so we will not be able to verify that the bug is
6569 And if we cannot understand what bug you are trying to fix, or why your
6570 patch should be an improvement, we will not install it. A test case will
6571 help us to understand.
6574 A guess about what the bug is or what it depends on.
6576 Such guesses are usually wrong. Even we cannot guess right about such
6577 things without first using the debugger to find the facts.
6581 @appendix MRI Compatible Script Files
6582 @cindex MRI compatibility
6583 To aid users making the transition to @sc{gnu} @command{ld} from the MRI
6584 linker, @command{ld} can use MRI compatible linker scripts as an
6585 alternative to the more general-purpose linker scripting language
6586 described in @ref{Scripts}. MRI compatible linker scripts have a much
6587 simpler command set than the scripting language otherwise used with
6588 @command{ld}. @sc{gnu} @command{ld} supports the most commonly used MRI
6589 linker commands; these commands are described here.
6591 In general, MRI scripts aren't of much use with the @code{a.out} object
6592 file format, since it only has three sections and MRI scripts lack some
6593 features to make use of them.
6595 You can specify a file containing an MRI-compatible script using the
6596 @samp{-c} command-line option.
6598 Each command in an MRI-compatible script occupies its own line; each
6599 command line starts with the keyword that identifies the command (though
6600 blank lines are also allowed for punctuation). If a line of an
6601 MRI-compatible script begins with an unrecognized keyword, @command{ld}
6602 issues a warning message, but continues processing the script.
6604 Lines beginning with @samp{*} are comments.
6606 You can write these commands using all upper-case letters, or all
6607 lower case; for example, @samp{chip} is the same as @samp{CHIP}.
6608 The following list shows only the upper-case form of each command.
6611 @cindex @code{ABSOLUTE} (MRI)
6612 @item ABSOLUTE @var{secname}
6613 @itemx ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname}
6614 Normally, @command{ld} includes in the output file all sections from all
6615 the input files. However, in an MRI-compatible script, you can use the
6616 @code{ABSOLUTE} command to restrict the sections that will be present in
6617 your output program. If the @code{ABSOLUTE} command is used at all in a
6618 script, then only the sections named explicitly in @code{ABSOLUTE}
6619 commands will appear in the linker output. You can still use other
6620 input sections (whatever you select on the command line, or using
6621 @code{LOAD}) to resolve addresses in the output file.
6623 @cindex @code{ALIAS} (MRI)
6624 @item ALIAS @var{out-secname}, @var{in-secname}
6625 Use this command to place the data from input section @var{in-secname}
6626 in a section called @var{out-secname} in the linker output file.
6628 @var{in-secname} may be an integer.
6630 @cindex @code{ALIGN} (MRI)
6631 @item ALIGN @var{secname} = @var{expression}
6632 Align the section called @var{secname} to @var{expression}. The
6633 @var{expression} should be a power of two.
6635 @cindex @code{BASE} (MRI)
6636 @item BASE @var{expression}
6637 Use the value of @var{expression} as the lowest address (other than
6638 absolute addresses) in the output file.
6640 @cindex @code{CHIP} (MRI)
6641 @item CHIP @var{expression}
6642 @itemx CHIP @var{expression}, @var{expression}
6643 This command does nothing; it is accepted only for compatibility.
6645 @cindex @code{END} (MRI)
6647 This command does nothing whatever; it's only accepted for compatibility.
6649 @cindex @code{FORMAT} (MRI)
6650 @item FORMAT @var{output-format}
6651 Similar to the @code{OUTPUT_FORMAT} command in the more general linker
6652 language, but restricted to one of these output formats:
6656 S-records, if @var{output-format} is @samp{S}
6659 IEEE, if @var{output-format} is @samp{IEEE}
6662 COFF (the @samp{coff-m68k} variant in BFD), if @var{output-format} is
6666 @cindex @code{LIST} (MRI)
6667 @item LIST @var{anything}@dots{}
6668 Print (to the standard output file) a link map, as produced by the
6669 @command{ld} command-line option @samp{-M}.
6671 The keyword @code{LIST} may be followed by anything on the
6672 same line, with no change in its effect.
6674 @cindex @code{LOAD} (MRI)
6675 @item LOAD @var{filename}
6676 @itemx LOAD @var{filename}, @var{filename}, @dots{} @var{filename}
6677 Include one or more object file @var{filename} in the link; this has the
6678 same effect as specifying @var{filename} directly on the @command{ld}
6681 @cindex @code{NAME} (MRI)
6682 @item NAME @var{output-name}
6683 @var{output-name} is the name for the program produced by @command{ld}; the
6684 MRI-compatible command @code{NAME} is equivalent to the command-line
6685 option @samp{-o} or the general script language command @code{OUTPUT}.
6687 @cindex @code{ORDER} (MRI)
6688 @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname}
6689 @itemx ORDER @var{secname} @var{secname} @var{secname}
6690 Normally, @command{ld} orders the sections in its output file in the
6691 order in which they first appear in the input files. In an MRI-compatible
6692 script, you can override this ordering with the @code{ORDER} command. The
6693 sections you list with @code{ORDER} will appear first in your output
6694 file, in the order specified.
6696 @cindex @code{PUBLIC} (MRI)
6697 @item PUBLIC @var{name}=@var{expression}
6698 @itemx PUBLIC @var{name},@var{expression}
6699 @itemx PUBLIC @var{name} @var{expression}
6700 Supply a value (@var{expression}) for external symbol
6701 @var{name} used in the linker input files.
6703 @cindex @code{SECT} (MRI)
6704 @item SECT @var{secname}, @var{expression}
6705 @itemx SECT @var{secname}=@var{expression}
6706 @itemx SECT @var{secname} @var{expression}
6707 You can use any of these three forms of the @code{SECT} command to
6708 specify the start address (@var{expression}) for section @var{secname}.
6709 If you have more than one @code{SECT} statement for the same
6710 @var{secname}, only the @emph{first} sets the start address.
6716 @unnumbered LD Index
6721 % I think something like @colophon should be in texinfo. In the
6723 \long\def\colophon{\hbox to0pt{}\vfill
6724 \centerline{The body of this manual is set in}
6725 \centerline{\fontname\tenrm,}
6726 \centerline{with headings in {\bf\fontname\tenbf}}
6727 \centerline{and examples in {\tt\fontname\tentt}.}
6728 \centerline{{\it\fontname\tenit\/} and}
6729 \centerline{{\sl\fontname\tensl\/}}
6730 \centerline{are used for emphasis.}\vfill}
6732 % Blame: doc@cygnus.com, 28mar91.