Commit | Line | Data |
---|---|---|
f22eee08 | 1 | \input texinfo |
c8072296 | 2 | @setfilename ld.info |
b4d4e8e3 | 3 | @syncodeindex ky cp |
7f9ae73e | 4 | @include configdoc.texi |
8de26d62 | 5 | @c (configdoc.texi is generated by the Makefile) |
ec40bbb8 DM |
6 | |
7 | @c @smallbook | |
1c48127e RP |
8 | |
9 | @ifinfo | |
10 | @format | |
11 | START-INFO-DIR-ENTRY | |
f9d3d71a | 12 | * Ld: (ld). The GNU linker. |
1c48127e RP |
13 | END-INFO-DIR-ENTRY |
14 | @end format | |
15 | @end ifinfo | |
16 | ||
b4d4e8e3 | 17 | @ifinfo |
246504a5 | 18 | This file documents the GNU linker LD. |
b4d4e8e3 | 19 | |
0b0642d6 | 20 | Copyright (C) 1991, 92, 93, 94, 1995 Free Software Foundation, Inc. |
b4d4e8e3 RP |
21 | |
22 | Permission is granted to make and distribute verbatim copies of | |
23 | this manual provided the copyright notice and this permission notice | |
24 | are preserved on all copies. | |
25 | ||
d76ae847 RP |
26 | Permission is granted to copy and distribute modified versions of this |
27 | manual under the conditions for verbatim copying, provided also that | |
28 | the entire resulting derived work is distributed under the terms of a | |
29 | permission notice identical to this one. | |
30 | ||
31 | Permission is granted to copy and distribute translations of this manual | |
32 | into another language, under the above conditions for modified versions. | |
33 | ||
b4d4e8e3 RP |
34 | @ignore |
35 | Permission is granted to process this file through Tex and print the | |
36 | results, provided the printed document carries copying permission | |
37 | notice identical to this one except for the removal of this paragraph | |
38 | (this paragraph not being relevant to the printed manual). | |
39 | ||
40 | @end ignore | |
b4d4e8e3 | 41 | @end ifinfo |
2c5c0674 RP |
42 | @iftex |
43 | @finalout | |
b4d4e8e3 | 44 | @setchapternewpage odd |
246504a5 | 45 | @settitle Using LD, the GNU linker |
f22eee08 | 46 | @titlepage |
246504a5 | 47 | @title Using ld |
c8072296 | 48 | @subtitle The GNU linker |
f22eee08 | 49 | @sp 1 |
cb70c872 | 50 | @subtitle @code{ld} version 2 |
d4e5e3c3 | 51 | @subtitle January 1994 |
c8072296 RP |
52 | @author Steve Chamberlain and Roland Pesch |
53 | @author Cygnus Support | |
b4d4e8e3 RP |
54 | @page |
55 | ||
56 | @tex | |
b4d4e8e3 RP |
57 | {\parskip=0pt |
58 | \hfill Cygnus Support\par | |
2c5c0674 | 59 | \hfill steve\@cygnus.com, pesch\@cygnus.com\par |
ec40bbb8 | 60 | \hfill {\it Using LD, the GNU linker}\par |
d4e5e3c3 DM |
61 | \hfill Edited by Jeffrey Osier (jeffrey\@cygnus.com)\par |
62 | \hfill and Roland Pesch (pesch\@cygnus.com)\par | |
b4d4e8e3 RP |
63 | } |
64 | \global\parindent=0pt % Steve likes it this way. | |
65 | @end tex | |
66 | ||
f22eee08 | 67 | @vskip 0pt plus 1filll |
0b0642d6 | 68 | Copyright @copyright{} 1991, 92, 93, 94, 1995 Free Software Foundation, Inc. |
f22eee08 RP |
69 | |
70 | Permission is granted to make and distribute verbatim copies of | |
71 | this manual provided the copyright notice and this permission notice | |
72 | are preserved on all copies. | |
73 | ||
74 | Permission is granted to copy and distribute modified versions of this | |
75 | manual under the conditions for verbatim copying, provided also that | |
76 | the entire resulting derived work is distributed under the terms of a | |
77 | permission notice identical to this one. | |
78 | ||
79 | Permission is granted to copy and distribute translations of this manual | |
80 | into another language, under the above conditions for modified versions. | |
f22eee08 | 81 | @end titlepage |
2c5c0674 | 82 | @end iftex |
b4d4e8e3 | 83 | @c FIXME: Talk about importance of *order* of args, cmds to linker! |
f22eee08 | 84 | |
f22eee08 | 85 | @ifinfo |
ec40bbb8 DM |
86 | @node Top |
87 | @top Using ld | |
246504a5 | 88 | This file documents the GNU linker ld. |
f22eee08 | 89 | |
2c5c0674 | 90 | @menu |
2d59b2c3 RP |
91 | * Overview:: Overview |
92 | * Invocation:: Invocation | |
93 | * Commands:: Command Language | |
ec40bbb8 | 94 | @ifset GENERIC |
2d59b2c3 | 95 | * Machine Dependent:: Machine Dependent Features |
ec40bbb8 DM |
96 | @end ifset |
97 | @ifclear GENERIC | |
98 | @ifset H8300 | |
99 | * H8/300:: ld and the H8/300 | |
100 | @end ifset | |
f9d3d71a ILT |
101 | @ifset Hitachi |
102 | * Hitachi:: ld and other Hitachi micros | |
103 | @end ifset | |
ec40bbb8 DM |
104 | @ifset I960 |
105 | * i960:: ld and the Intel 960 family | |
106 | @end ifset | |
107 | @end ifclear | |
108 | @ifclear SingleFormat | |
2d59b2c3 | 109 | * BFD:: BFD |
ec40bbb8 DM |
110 | @end ifclear |
111 | @c Following blank line required for remaining bug in makeinfo conds/menus | |
112 | ||
2d59b2c3 RP |
113 | * MRI:: MRI Compatible Script Files |
114 | * Index:: Index | |
2c5c0674 | 115 | @end menu |
ec40bbb8 | 116 | @end ifinfo |
2c5c0674 | 117 | |
ec40bbb8 | 118 | @node Overview |
f22eee08 RP |
119 | @chapter Overview |
120 | ||
2c5c0674 RP |
121 | @cindex GNU linker |
122 | @cindex what is this? | |
246504a5 | 123 | @code{ld} combines a number of object and archive files, relocates |
ec40bbb8 DM |
124 | their data and ties up symbol references. Usually the last step in |
125 | compiling a program is to run @code{ld}. | |
f22eee08 | 126 | |
246504a5 | 127 | @code{ld} accepts Linker Command Language files written in |
2c5c0674 | 128 | a superset of AT&T's Link Editor Command Language syntax, |
b4d4e8e3 | 129 | to provide explicit and total control over the linking process. |
f22eee08 | 130 | |
ec40bbb8 | 131 | @ifclear SingleFormat |
246504a5 RP |
132 | This version of @code{ld} uses the general purpose BFD libraries |
133 | to operate on object files. This allows @code{ld} to read, combine, and | |
b4d4e8e3 RP |
134 | write object files in many different formats---for example, COFF or |
135 | @code{a.out}. Different formats may be linked together to produce any | |
d4e5e3c3 | 136 | available kind of object file. @xref{BFD}, for more information. |
ec40bbb8 | 137 | @end ifclear |
f22eee08 | 138 | |
2c5c0674 RP |
139 | Aside from its flexibility, the GNU linker is more helpful than other |
140 | linkers in providing diagnostic information. Many linkers abandon | |
141 | execution immediately upon encountering an error; whenever possible, | |
246504a5 | 142 | @code{ld} continues executing, allowing you to identify other errors |
2c5c0674 RP |
143 | (or, in some cases, to get an output file in spite of the error). |
144 | ||
ec40bbb8 | 145 | @node Invocation |
2c5c0674 RP |
146 | @chapter Invocation |
147 | ||
246504a5 | 148 | The GNU linker @code{ld} is meant to cover a broad range of situations, |
2c5c0674 | 149 | and to be as compatible as possible with other linkers. As a result, |
ec40bbb8 | 150 | you have many choices to control its behavior. |
2c5c0674 | 151 | |
ec40bbb8 | 152 | @ifset UsesEnvVars |
2c5c0674 | 153 | @menu |
2d59b2c3 RP |
154 | * Options:: Command Line Options |
155 | * Environment:: Environment Variables | |
2c5c0674 | 156 | @end menu |
f22eee08 | 157 | |
ec40bbb8 | 158 | @node Options |
2c5c0674 | 159 | @section Command Line Options |
ec40bbb8 | 160 | @end ifset |
2c5c0674 RP |
161 | |
162 | @cindex command line | |
163 | @cindex options | |
ec40bbb8 | 164 | Here is a summary of the options you can use on the @code{ld} command |
2c5c0674 | 165 | line: |
f22eee08 | 166 | |
ec40bbb8 | 167 | @c FIXME! -relax only avail h8/300, i960. Conditionals screwed in examples. |
c8072296 | 168 | @smallexample |
de87cdb4 | 169 | ld [ -o @var{output} ] @var{objfile}@dots{} |
4551e108 ILT |
170 | [ -A@var{architecture} ] [ -b @var{input-format} ] |
171 | [ -Bstatic ] [ -Bdynamic ] [ -Bsymbolic ] | |
cb70c872 | 172 | [ -c @var{MRI-commandfile} ] [ -d | -dc | -dp ] |
d76ae847 | 173 | [ -defsym @var{symbol}=@var{expression} ] |
01bc8f35 | 174 | [ -dynamic-linker @var{file} ] [ -embedded-relocs ] |
cb70c872 | 175 | [ -e @var{entry} ] [ -F ] [ -F @var{format} ] |
d4e5e3c3 DM |
176 | [ -format @var{input-format} ] [ -g ] [ -G @var{size} ] [ -help ] |
177 | [ -i ] [ -l@var{archive} ] [ -L@var{searchdir} ] [ -M ] | |
178 | [ -Map @var{mapfile} ] [ -m @var{emulation} ] [ -N | -n ] | |
4551e108 | 179 | [ -noinhibit-exec ] [ -no-keep-memory ] [ -oformat @var{output-format} ] |
d4e5e3c3 | 180 | [ -R @var{filename} ] [ -relax ] [ -retain-symbols-file @var{filename} ] |
e54bf1c1 ILT |
181 | [ -r | -Ur ] [ -rpath @var{dir} ] [-rpath-link @var{dir} ] |
182 | [ -S ] [ -s ] [ -soname @var{name} ] [ -shared ] | |
183 | [ -sort-common ] [ -stats ] [ -T @var{commandfile} ] | |
867a1b8a | 184 | [ -Ttext @var{org} ] [ -Tdata @var{org} ] |
c96386c4 | 185 | [ -Tbss @var{org} ] [ -t ] [ -traditional-format ] |
01bc8f35 | 186 | [ -u @var{symbol}] [-V] [-v] [ -verbose] [ -version ] |
0b0642d6 ILT |
187 | [ -warn-common ] [ -warn-constructors] [ -warn-once ] |
188 | [ -y @var{symbol} ] [ -X ] [-x ] | |
a1ad915d | 189 | [ -( [ archives ] -) ] [ --start-group [ archives ] --end-group ] |
4551e108 | 190 | [ -split-by-reloc @var{count} ] [ -split-by-file ] [ --whole-archive ] |
c8072296 | 191 | @end smallexample |
b4d4e8e3 RP |
192 | |
193 | This plethora of command-line options may seem intimidating, but in | |
194 | actual practice few of them are used in any particular context. | |
2c5c0674 | 195 | @cindex standard Unix system |
246504a5 | 196 | For instance, a frequent use of @code{ld} is to link standard Unix |
b4d4e8e3 RP |
197 | object files on a standard, supported Unix system. On such a system, to |
198 | link a file @code{hello.o}: | |
ec40bbb8 | 199 | |
f22eee08 | 200 | @example |
ec40bbb8 | 201 | ld -o @var{output} /lib/crt0.o hello.o -lc |
f22eee08 | 202 | @end example |
ec40bbb8 | 203 | |
d76ae847 | 204 | This tells @code{ld} to produce a file called @var{output} as the |
b4d4e8e3 | 205 | result of linking the file @code{/lib/crt0.o} with @code{hello.o} and |
ec40bbb8 DM |
206 | the library @code{libc.a}, which will come from the standard search |
207 | directories. (See the discussion of the @samp{-l} option below.) | |
f22eee08 | 208 | |
246504a5 | 209 | The command-line options to @code{ld} may be specified in any order, and |
ec40bbb8 | 210 | may be repeated at will. Repeating most options with a |
f22eee08 | 211 | different argument will either have no further effect, or override prior |
ec40bbb8 | 212 | occurrences (those further to the left on the command line) of that |
f22eee08 RP |
213 | option. |
214 | ||
ec40bbb8 | 215 | @ifclear SingleFormat |
2c5c0674 | 216 | The exceptions---which may meaningfully be used more than once---are |
ec40bbb8 | 217 | @samp{-A}, @samp{-b} (or its synonym @samp{-format}), @samp{-defsym}, |
0b3499f6 ILT |
218 | @samp{-L}, @samp{-l}, @samp{-R}, @samp{-u}, and @samp{-(} (or its |
219 | synonym @samp{--start-group}).. | |
ec40bbb8 DM |
220 | @end ifclear |
221 | @ifset SingleFormat | |
222 | The exceptions---which may meaningfully be used more than once---are | |
0b3499f6 ILT |
223 | @samp{-A}, @samp{-defsym}, @samp{-L}, @samp{-l}, @samp{-R}, @samp{-u}, |
224 | and @samp{-(} (or its synonym @samp{--start-group}). | |
ec40bbb8 | 225 | @end ifset |
f22eee08 | 226 | |
2c5c0674 | 227 | @cindex object files |
8ddef552 | 228 | The list of object files to be linked together, shown as @var{objfile}@dots{}, |
ec40bbb8 DM |
229 | may follow, precede, or be mixed in with command-line options, except that |
230 | an @var{objfile} argument may not be placed between an option and | |
b4d4e8e3 | 231 | its argument. |
f22eee08 | 232 | |
7f9ae73e RP |
233 | Usually the linker is invoked with at least one object file, but you can |
234 | specify other forms of binary input files using @samp{-l}, @samp{-R}, | |
235 | and the script command language. If @emph{no} binary input files at all | |
236 | are specified, the linker does not produce any output, and issues the | |
237 | message @samp{No input files}. | |
2c5c0674 | 238 | |
0b3499f6 ILT |
239 | If the linker can not recognize the format of an object file, it will |
240 | assume that it is a linker script. A script specified in this way | |
241 | augments the main linker script used for the link (either the default | |
242 | linker script or the one specified by using @samp{-T}). This feature | |
243 | permits the linker to link against a file which appears to be an object | |
244 | or an archive, but actually merely defines some symbol values, or uses | |
245 | @code{INPUT} or @code{GROUP} to load other objects. @xref{Commands}. | |
246 | ||
d4e5e3c3 DM |
247 | For options whose names are a single letter, |
248 | option arguments must either follow the option letter without intervening | |
f22eee08 RP |
249 | whitespace, or be given as separate arguments immediately following the |
250 | option that requires them. | |
251 | ||
d4e5e3c3 DM |
252 | For options whose names are multiple letters, either one dash or two can |
253 | precede the option name; for example, @samp{--oformat} and | |
254 | @samp{-oformat} are equivalent. Arguments to multiple-letter options | |
255 | must either be separated from the option name by an equals sign, or be | |
256 | given as separate arguments immediately following the option that | |
257 | requires them. For example, @samp{--oformat srec} and | |
258 | @samp{--oformat=srec} are equivalent. Unique abbreviations of the names | |
259 | of multiple-letter options are accepted. | |
260 | ||
f22eee08 | 261 | @table @code |
ec40bbb8 | 262 | @ifset I960 |
2c5c0674 RP |
263 | @cindex architectures |
264 | @kindex -A@var{arch} | |
b4d4e8e3 | 265 | @item -A@var{architecture} |
246504a5 RP |
266 | In the current release of @code{ld}, this option is useful only for the |
267 | Intel 960 family of architectures. In that @code{ld} configuration, the | |
1c48127e RP |
268 | @var{architecture} argument identifies the particular architecture in |
269 | the 960 family, enabling some safeguards and modifying the | |
d76ae847 RP |
270 | archive-library search path. @xref{i960,,@code{ld} and the Intel 960 |
271 | family}, for details. | |
b4d4e8e3 | 272 | |
246504a5 | 273 | Future releases of @code{ld} may support similar functionality for |
b4d4e8e3 | 274 | other architecture families. |
ec40bbb8 | 275 | @end ifset |
b4d4e8e3 | 276 | |
ec40bbb8 | 277 | @ifclear SingleFormat |
2c5c0674 RP |
278 | @cindex binary input format |
279 | @kindex -b @var{format} | |
280 | @cindex input format | |
281 | @item -b @var{input-format} | |
282 | @cindex input format | |
1fb57a5d RP |
283 | @code{ld} may be configured to support more than one kind of object |
284 | file. If your @code{ld} is configured this way, you can use the | |
285 | @samp{-b} option to specify the binary format for input object files | |
286 | that follow this option on the command line. Even when @code{ld} is | |
287 | configured to support alternative object formats, you don't usually need | |
288 | to specify this, as @code{ld} should be configured to expect as a | |
289 | default input format the most usual format on each machine. | |
290 | @var{input-format} is a text string, the name of a particular format | |
291 | supported by the BFD libraries. (You can list the available binary | |
292 | formats with @samp{objdump -i}.) @w{@samp{-format @var{input-format}}} | |
293 | has the same effect, as does the script command @code{TARGET}. | |
294 | @xref{BFD}. | |
2c5c0674 RP |
295 | |
296 | You may want to use this option if you are linking files with an unusual | |
ec40bbb8 | 297 | binary format. You can also use @samp{-b} to switch formats explicitly (when |
2c5c0674 | 298 | linking object files of different formats), by including |
ec40bbb8 | 299 | @samp{-b @var{input-format}} before each group of object files in a |
2c5c0674 RP |
300 | particular format. |
301 | ||
302 | The default format is taken from the environment variable | |
ec40bbb8 DM |
303 | @code{GNUTARGET}. |
304 | @ifset UsesEnvVars | |
305 | @xref{Environment}. | |
306 | @end ifset | |
307 | You can also define the input | |
867a1b8a | 308 | format from a script, using the command @code{TARGET}; see @ref{Option |
d76ae847 | 309 | Commands}. |
ec40bbb8 | 310 | @end ifclear |
2c5c0674 RP |
311 | |
312 | @kindex -Bstatic | |
f22eee08 | 313 | @item -Bstatic |
4551e108 ILT |
314 | Do not link against shared libraries. This is only meaningful on |
315 | platforms for which shared libraries are supported. | |
316 | ||
317 | @kindex -Bdynamic | |
318 | @item -Bdynamic | |
319 | Link against dynamic libraries. This is only meaningful on platforms | |
320 | for which shared libraries are supported. This option is normally the | |
321 | default on such platforms. | |
322 | ||
323 | @kindex -Bsymbolic | |
324 | @item -Bsymbolic | |
325 | When creating a shared library, bind references to global symbols to the | |
326 | definition within the shared library, if any. Normally, it is possible | |
327 | for a program linked against a shared library to override the definition | |
328 | within the shared library. This option is only meaningful on ELF | |
329 | platforms which support shared libraries. | |
f22eee08 | 330 | |
2d59b2c3 RP |
331 | @kindex -c @var{MRI-cmdfile} |
332 | @cindex compatibility, MRI | |
333 | @item -c @var{MRI-commandfile} | |
334 | For compatibility with linkers produced by MRI, @code{ld} accepts script | |
335 | files written in an alternate, restricted command language, described in | |
d76ae847 | 336 | @ref{MRI,,MRI Compatible Script Files}. Introduce MRI script files with |
ec40bbb8 | 337 | the option @samp{-c}; use the @samp{-T} option to run linker |
d76ae847 | 338 | scripts written in the general-purpose @code{ld} scripting language. |
8ddef552 DM |
339 | If @var{MRI-cmdfile} does not exist, @code{ld} looks for it in the directories |
340 | specified by any @samp{-L} options. | |
b4d4e8e3 | 341 | |
2c5c0674 RP |
342 | @cindex common allocation |
343 | @kindex -d | |
b4d4e8e3 | 344 | @item -d |
2c5c0674 | 345 | @kindex -dc |
b4d4e8e3 | 346 | @itemx -dc |
2c5c0674 | 347 | @kindex -dp |
b4d4e8e3 RP |
348 | @itemx -dp |
349 | These three options are equivalent; multiple forms are supported for | |
ec40bbb8 | 350 | compatibility with other linkers. They |
2c5c0674 | 351 | assign space to common symbols even if a relocatable output file is |
ec40bbb8 | 352 | specified (with @samp{-r}). The script command |
867a1b8a | 353 | @code{FORCE_COMMON_ALLOCATION} has the same effect. @xref{Option |
d76ae847 | 354 | Commands}. |
b4d4e8e3 | 355 | |
2c5c0674 | 356 | @cindex symbols, from command line |
d76ae847 RP |
357 | @kindex -defsym @var{symbol}=@var{exp} |
358 | @item -defsym @var{symbol}=@var{expression} | |
2c5c0674 RP |
359 | Create a global symbol in the output file, containing the absolute |
360 | address given by @var{expression}. You may use this option as many | |
361 | times as necessary to define multiple symbols in the command line. A | |
362 | limited form of arithmetic is supported for the @var{expression} in this | |
363 | context: you may give a hexadecimal constant or the name of an existing | |
364 | symbol, or use @code{+} and @code{-} to add or subtract hexadecimal | |
365 | constants or symbols. If you need more elaborate expressions, consider | |
d76ae847 RP |
366 | using the linker command language from a script (@pxref{Assignment, , |
367 | Assignment: Symbol Definitions}). @emph{Note:} there should be no | |
368 | white space between @var{symbol}, the equals sign (``@key{=}''), and | |
369 | @var{expression}. | |
2c5c0674 | 370 | |
a1ad915d ILT |
371 | @ifset GENERIC |
372 | @cindex dynamic linker, from command line | |
373 | @kindex -dynamic-linker @var{file} | |
374 | @item -dynamic-linker @var{file} | |
375 | Set the name of the dynamic linker. This is only meaningful when | |
376 | generating dynamically linked ELF executables. The default dynamic | |
377 | linker is normally correct; don't use this unless you know what you are | |
378 | doing. | |
379 | @end ifset | |
380 | ||
01bc8f35 ILT |
381 | @cindex MIPS embedded PIC code |
382 | @kindex -embedded-relocs | |
383 | @item -embedded-relocs | |
384 | This option is only meaningful when linking MIPS embedded PIC code, | |
385 | generated by the -membedded-pic option to the GNU compiler and | |
386 | assembler. It causes the linker to create a table which may be used at | |
387 | runtime to relocate any data which was statically initialized to pointer | |
388 | values. See the code in testsuite/ld-empic for details. | |
389 | ||
2c5c0674 RP |
390 | @cindex entry point, from command line |
391 | @kindex -e @var{entry} | |
f22eee08 RP |
392 | @item -e @var{entry} |
393 | Use @var{entry} as the explicit symbol for beginning execution of your | |
b4d4e8e3 RP |
394 | program, rather than the default entry point. @xref{Entry Point}, for a |
395 | discussion of defaults and other ways of specifying the | |
396 | entry point. | |
f22eee08 | 397 | |
ec40bbb8 | 398 | @ifclear SingleFormat |
2c5c0674 | 399 | @kindex -F |
b4d4e8e3 | 400 | @item -F |
2c5c0674 | 401 | @itemx -F@var{format} |
ec40bbb8 DM |
402 | Ignored. Some older linkers used this option throughout a compilation |
403 | toolchain for specifying object-file format for both input and output | |
404 | object files. The mechanisms @code{ld} uses for this purpose (the | |
867a1b8a DM |
405 | @samp{-b} or @samp{-format} options for input files, @samp{-oformat} |
406 | option or the @code{TARGET} command in linker scripts for output files, | |
407 | the @code{GNUTARGET} environment variable) are more flexible, but | |
408 | @code{ld} accepts the @samp{-F} option for compatibility with scripts | |
409 | written to call the old linker. | |
2c5c0674 RP |
410 | |
411 | @kindex -format | |
412 | @item -format @var{input-format} | |
ec40bbb8 DM |
413 | Synonym for @samp{-b @var{input-format}}. |
414 | @end ifclear | |
2c5c0674 RP |
415 | |
416 | @kindex -g | |
b4d4e8e3 | 417 | @item -g |
ec40bbb8 | 418 | Ignored. Provided for compatibility with other tools. |
b4d4e8e3 | 419 | |
8ddef552 DM |
420 | @kindex -G |
421 | @cindex object size | |
422 | @item -G@var{value} | |
423 | @itemx -G @var{value} | |
424 | Set the maximum size of objects to be optimized using the GP register to | |
425 | @var{size} under MIPS ECOFF. Ignored for other object file formats. | |
426 | ||
de87cdb4 DM |
427 | @cindex help |
428 | @cindex usage | |
d4e5e3c3 DM |
429 | @kindex -help |
430 | @item -help | |
de87cdb4 | 431 | Print a summary of the command-line options on the standard output and exit. |
de87cdb4 | 432 | |
2c5c0674 RP |
433 | @kindex -i |
434 | @cindex incremental link | |
f22eee08 | 435 | @item -i |
ec40bbb8 | 436 | Perform an incremental link (same as option @samp{-r}). |
f22eee08 | 437 | |
2c5c0674 | 438 | @cindex archive files, from cmd line |
de87cdb4 | 439 | @kindex -l@var{archive} |
b4d4e8e3 | 440 | @item -l@var{ar} |
de87cdb4 | 441 | Add archive file @var{archive} to the list of files to link. This |
f22eee08 | 442 | option may be used any number of times. @code{ld} will search its |
de87cdb4 | 443 | path-list for occurrences of @code{lib@var{ar}.a} for every @var{archive} |
f22eee08 RP |
444 | specified. |
445 | ||
2c5c0674 RP |
446 | @cindex search directory, from cmd line |
447 | @kindex -L@var{dir} | |
b4d4e8e3 | 448 | @item -L@var{searchdir} |
836a5ee4 | 449 | @itemx -L @var{searchdir} |
ec40bbb8 | 450 | Add path @var{searchdir} to the list of paths that @code{ld} will search |
8ddef552 | 451 | for archive libraries and @code{ld} control scripts. You may use this |
0b0642d6 ILT |
452 | option any number of times. The directories are searched in the order |
453 | in which they are specified on the command line. Directories specified | |
454 | on the command line are searched before the default directories. All | |
455 | @code{-L} options apply to all @code{-l} options, regardless of the | |
456 | order in which the options appear. | |
f22eee08 | 457 | |
ec40bbb8 | 458 | @ifset UsesEnvVars |
2c5c0674 | 459 | The default set of paths searched (without being specified with |
ec40bbb8 DM |
460 | @samp{-L}) depends on which emulation mode @code{ld} is using, and in |
461 | some cases also on how it was configured. @xref{Environment}. | |
462 | @end ifset | |
463 | ||
464 | The paths can also be specified in a link script with the | |
0b0642d6 ILT |
465 | @code{SEARCH_DIR} command. Directories specified this way are searched |
466 | at the point in which the linker script appears in the command line. | |
f22eee08 | 467 | |
2c5c0674 RP |
468 | @cindex link map |
469 | @kindex -M | |
f22eee08 | 470 | @item -M |
e54bf1c1 ILT |
471 | Print (to the standard output) a link map---diagnostic information about |
472 | where symbols are mapped by @code{ld}, and information on global common | |
473 | storage allocation. | |
f22eee08 | 474 | |
d18a4527 DM |
475 | @cindex link map |
476 | @kindex -Map | |
477 | @item -Map @var{mapfile} | |
478 | Print to the file @var{mapfile} a link map---diagnostic information | |
479 | about where symbols are mapped by @code{ld}, and information on global | |
480 | common storage allocation. | |
481 | ||
8ddef552 DM |
482 | @cindex emulation |
483 | @kindex -m @var{emulation} | |
484 | @item -m@var{emulation} | |
485 | @itemx -m @var{emulation} | |
486 | Emulate the @var{emulation} linker. You can list the available | |
01bc8f35 ILT |
487 | emulations with the @samp{--verbose} or @samp{-V} options. The default |
488 | depends on how your @code{ld} was configured. | |
8ddef552 | 489 | |
2c5c0674 RP |
490 | @kindex -N |
491 | @cindex read/write from cmd line | |
492 | @kindex OMAGIC | |
f22eee08 | 493 | @item -N |
ec40bbb8 DM |
494 | Set the text and data sections to be readable and writable. Also, do |
495 | not page-align the data segment. If the output format supports Unix | |
496 | style magic numbers, mark the output as @code{OMAGIC}. | |
f22eee08 | 497 | |
2c5c0674 RP |
498 | @kindex -n |
499 | @cindex read-only text | |
500 | @kindex NMAGIC | |
d4e5e3c3 | 501 | @item -n |
ec40bbb8 DM |
502 | Set the text segment to be read only, and mark the output as |
503 | @code{NMAGIC} if possible. | |
f22eee08 | 504 | |
b4d4e8e3 | 505 | @item -noinhibit-exec |
2c5c0674 RP |
506 | @cindex output file after errors |
507 | @kindex -noinhibit-exec | |
ec40bbb8 | 508 | Retain the executable output file whenever it is still usable. |
b4d4e8e3 | 509 | Normally, the linker will not produce an output file if it encounters |
ec40bbb8 DM |
510 | errors during the link process; it exits without writing an output file |
511 | when it issues any error whatsoever. | |
b4d4e8e3 | 512 | |
4551e108 ILT |
513 | @item -no-keep-memory |
514 | @cindex memory usage | |
515 | @kindex -no-keep-memory | |
516 | @code{ld} normally optimizes for speed over memory usage by caching the | |
517 | symbol tables of input files in memory. This option tells @code{ld} to | |
518 | instead optimize for memory usage, by rereading the symbol tables as | |
519 | necessary. This may be required if @code{ld} runs out of memory space | |
520 | while linking a large executable. | |
521 | ||
f22eee08 | 522 | @item -o @var{output} |
2c5c0674 RP |
523 | @kindex -o @var{output} |
524 | @cindex naming the output file | |
ec40bbb8 DM |
525 | Use @var{output} as the name for the program produced by @code{ld}; if this |
526 | option is not specified, the name @file{a.out} is used by default. The | |
2c5c0674 | 527 | script command @code{OUTPUT} can also specify the output file name. |
f22eee08 | 528 | |
7f9ae73e | 529 | @ifclear SingleFormat |
346535cc DM |
530 | @kindex -oformat |
531 | @item -oformat @var{output-format} | |
1fb57a5d RP |
532 | @code{ld} may be configured to support more than one kind of object |
533 | file. If your @code{ld} is configured this way, you can use the | |
534 | @samp{-oformat} option to specify the binary format for the output | |
535 | object file. Even when @code{ld} is configured to support alternative | |
536 | object formats, you don't usually need to specify this, as @code{ld} | |
537 | should be configured to produce as a default output format the most | |
538 | usual format on each machine. @var{output-format} is a text string, the | |
539 | name of a particular format supported by the BFD libraries. (You can | |
540 | list the available binary formats with @samp{objdump -i}.) The script | |
541 | command @code{OUTPUT_FORMAT} can also specify the output format, but | |
542 | this option overrides it. @xref{BFD}. | |
7f9ae73e | 543 | @end ifclear |
346535cc | 544 | |
b4d4e8e3 | 545 | @item -R @var{filename} |
2c5c0674 RP |
546 | @kindex -R @var{file} |
547 | @cindex symbol-only input | |
867a1b8a DM |
548 | Read symbol names and their addresses from @var{filename}, but do not |
549 | relocate it or include it in the output. This allows your output file | |
550 | to refer symbolically to absolute locations of memory defined in other | |
551 | programs. | |
ec40bbb8 | 552 | |
1c48127e RP |
553 | @kindex -relax |
554 | @cindex synthesizing linker | |
555 | @cindex relaxing addressing modes | |
d4e5e3c3 | 556 | @item -relax |
1fb57a5d RP |
557 | An option with machine dependent effects. |
558 | @ifset GENERIC | |
559 | Currently this option is only supported on the H8/300 and the Intel 960. | |
560 | @end ifset | |
ec40bbb8 DM |
561 | @ifset H8300 |
562 | @xref{H8/300,,@code{ld} and the H8/300}. | |
563 | @end ifset | |
1fb57a5d RP |
564 | @ifset I960 |
565 | @xref{i960,, @code{ld} and the Intel 960 family}. | |
566 | @end ifset | |
1c48127e | 567 | |
1fb57a5d | 568 | On some platforms, the @samp{-relax} option performs global optimizations that |
ec40bbb8 | 569 | become possible when the linker resolves addressing in the program, such |
1c48127e RP |
570 | as relaxing address modes and synthesizing new instructions in the |
571 | output object file. | |
572 | ||
1fb57a5d | 573 | @ifset GENERIC |
1c48127e | 574 | On platforms where this is not supported, @samp{-relax} is accepted, but |
ec40bbb8 | 575 | ignored. |
1fb57a5d | 576 | @end ifset |
1c48127e | 577 | |
7c8fab26 RP |
578 | @item -retain-symbols-file @var{filename} |
579 | @cindex retaining specified symbols | |
580 | @cindex stripping all but some symbols | |
581 | @cindex symbols, retaining selectively | |
582 | Retain @emph{only} the symbols listed in the file @var{filename}, | |
583 | discarding all others. @var{filename} is simply a flat file, with one | |
584 | symbol name per line. This option is especially useful in environments | |
585 | @ifset GENERIC | |
586 | (such as VxWorks) | |
587 | @end ifset | |
588 | where a large global symbol table is accumulated gradually, to conserve | |
589 | run-time memory. | |
590 | ||
591 | @samp{-retain-symbols-file} does @emph{not} discard undefined symbols, | |
592 | or symbols needed for relocations. | |
593 | ||
594 | You may only specify @samp{-retain-symbols-file} once in the command | |
595 | line. It overrides @samp{-s} and @samp{-S}. | |
596 | ||
a1ad915d ILT |
597 | @ifset GENERIC |
598 | @item -rpath @var{dir} | |
599 | @cindex runtime library search path | |
600 | @kindex -rpath | |
0b0642d6 ILT |
601 | Add a directory to the runtime library search path. This is used when |
602 | linking an ELF executable with shared objects. All @code{-rpath} | |
603 | arguments are concatenated and passed to the runtime linker, which uses | |
e54bf1c1 ILT |
604 | them to locate shared objects at runtime. The @code{-rpath} option is |
605 | also used when locating shared objects which are needed by shared | |
606 | objects explicitly included in the link; see the description of the | |
607 | @code{-rpath-link} option. | |
0b0642d6 ILT |
608 | |
609 | The @code{-rpath} option may also be used on SunOS. By default, on | |
610 | SunOS, the linker will form a runtime search patch out of all the | |
e54bf1c1 | 611 | @code{-L} options it is given. If a @code{-rpath} option is used, the |
0b0642d6 ILT |
612 | runtime search path will be formed exclusively using the @code{-rpath} |
613 | options, ignoring the @code{-L} options. This can be useful when using | |
614 | gcc, which adds many @code{-L} options which may be on NFS mounted | |
615 | filesystems. | |
a1ad915d ILT |
616 | @end ifset |
617 | ||
e54bf1c1 ILT |
618 | @ifset GENERIC |
619 | @cindex link-time runtime library search path | |
620 | @kindex -rpath-link | |
621 | @item -rpath-link @var{DIR} | |
622 | When using ELF or SunOS, one shared library may require another. This | |
623 | happens when an @code{ld -shared} link includes a shared library as one | |
624 | of the input files. | |
625 | ||
626 | When the linker encounters such a dependency when doing a non-shared, | |
627 | non-relocateable link, it will automatically try to locate the required | |
628 | shared library and include it in the link, if it is not included | |
629 | explicitly. In such a case, the @code{-rpath-link} option | |
630 | specifies the first set of directories to search. The | |
631 | @code{-rpath-link} option may specify a sequence of directory names | |
632 | either by specifying a list of names separated by colons, or by | |
633 | appearing multiple times. | |
634 | ||
635 | The linker uses the following search paths to locate required shared | |
636 | libraries. | |
637 | @enumerate | |
638 | @item | |
639 | Any directories specified by @code{-rpath-link} options. | |
640 | @item | |
641 | Any directories specified by @code{-rpath} options. The difference | |
642 | between @code{-rpath} and @code{-rpath-link} is that directories | |
643 | specified by @code{-rpath} options are included in the executable and | |
644 | used at runtime, whereas the @code{-rpath-link} option is only effective | |
645 | at link time. | |
646 | @item | |
647 | On SunOS, if the @code{-rpath} option was not used, search any | |
648 | directories specified using @code{-L} options. | |
649 | @item | |
650 | For a native linker, the contents of the environment variable | |
651 | @code{LD_LIBRARY_PATH}. | |
652 | @item | |
653 | The default directories, normally @file{/lib} and @file{/usr/lib}. | |
654 | @end enumerate | |
655 | ||
656 | If the required shared library is not found, the linker will issue a | |
657 | warning and continue with the link. | |
658 | @end ifset | |
659 | ||
b4d4e8e3 | 660 | @cindex partial link |
2c5c0674 RP |
661 | @cindex relocatable output |
662 | @kindex -r | |
d4e5e3c3 | 663 | @item -r |
ec40bbb8 | 664 | Generate relocatable output---i.e., generate an output file that can in |
246504a5 | 665 | turn serve as input to @code{ld}. This is often called @dfn{partial |
2c5c0674 RP |
666 | linking}. As a side effect, in environments that support standard Unix |
667 | magic numbers, this option also sets the output file's magic number to | |
668 | @code{OMAGIC}. | |
669 | @c ; see @code{-N}. | |
670 | If this option is not specified, an absolute file is produced. When | |
671 | linking C++ programs, this option @emph{will not} resolve references to | |
ec40bbb8 | 672 | constructors; to do that, use @samp{-Ur}. |
2c5c0674 | 673 | |
867a1b8a | 674 | This option does the same thing as @samp{-i}. |
f22eee08 | 675 | |
2c5c0674 RP |
676 | @kindex -S |
677 | @cindex strip debugger symbols | |
d4e5e3c3 | 678 | @item -S |
ec40bbb8 | 679 | Omit debugger symbol information (but not all symbols) from the output file. |
f22eee08 | 680 | |
2c5c0674 RP |
681 | @kindex -s |
682 | @cindex strip all symbols | |
d4e5e3c3 | 683 | @item -s |
ec40bbb8 | 684 | Omit all symbol information from the output file. |
f22eee08 | 685 | |
0b3499f6 ILT |
686 | @ifset GENERIC |
687 | @item -soname @var{name} | |
688 | @cindex runtime library name | |
689 | @kindex -soname | |
690 | When creating an ELF shared object, set the internal DT_SONAME field to | |
691 | the specified name. When an executable is linked with a shared object | |
692 | which has a DT_SONAME field, then when the executable is run the dynamic | |
693 | linker will attempt to load the shared object specified by the DT_SONAME | |
694 | field rather than the using the file name given to the linker. | |
695 | @end ifset | |
696 | ||
4551e108 ILT |
697 | @item -shared |
698 | @cindex shared libraries | |
699 | @kindex -shared | |
e54bf1c1 ILT |
700 | Create a shared library. This is currently only supported on ELF and |
701 | SunOS platforms. On SunOS, the linker will automatically create a | |
702 | shared library if the @code{-e} option is not used and there are | |
703 | undefined symbols in the link. | |
4551e108 | 704 | |
2a28d8b0 | 705 | @item -sort-common |
4551e108 | 706 | @kindex -sort-common |
2a28d8b0 DM |
707 | Normally, when @code{ld} places the global common symbols in the |
708 | appropriate output sections, it sorts them by size. First come all the | |
709 | one byte symbols, then all the two bytes, then all the four bytes, and | |
836a5ee4 DM |
710 | then everything else. This is to prevent gaps between symbols due to |
711 | alignment constraints. This option disables that sorting. | |
2a28d8b0 | 712 | |
01bc8f35 ILT |
713 | @item -split-by-reloc @var{count} |
714 | @kindex split | |
715 | Trys to creates extra sections in the output file so that no single output section | |
716 | in the file contains more than @var{count} relocations. This | |
717 | is useful when generating huge relocatable for downloading into | |
718 | certain real time kernels with the COFF object file format; since | |
719 | COFF cannot represent more than 65535 relocations in a single section. | |
720 | Note that this will fail to work with object file formats which do not | |
721 | support arbitrary sections. The linker will not split up individual input | |
722 | sections for redistribution, so if a single input section contains | |
723 | more than @var{count} relocations one output section will contain that | |
724 | many relocations. | |
725 | ||
726 | @item -split-by-file | |
727 | @kindex split | |
728 | Similar to -split-by-reloc but creates a new output section for each | |
729 | input file. | |
730 | ||
8594f568 SS |
731 | @item -stats |
732 | Compute and display statistics about the operation of the linker, | |
733 | such as execution time and memory usage. | |
734 | ||
867a1b8a DM |
735 | @item -Tbss @var{org} |
736 | @kindex -Tbss @var{org} | |
737 | @itemx -Tdata @var{org} | |
738 | @kindex -Tdata @var{org} | |
739 | @itemx -Ttext @var{org} | |
740 | @kindex -Ttext @var{org} | |
2c5c0674 RP |
741 | @cindex segment origins, cmd line |
742 | Use @var{org} as the starting address for---respectively---the | |
b4d4e8e3 | 743 | @code{bss}, @code{data}, or the @code{text} segment of the output file. |
ec40bbb8 DM |
744 | @var{org} must be a single hexadecimal integer; |
745 | for compatibility with other linkers, you may omit the leading | |
d76ae847 | 746 | @samp{0x} usually associated with hexadecimal values. |
f22eee08 | 747 | |
b4d4e8e3 RP |
748 | @item -T @var{commandfile} |
749 | @itemx -T@var{commandfile} | |
2c5c0674 | 750 | @kindex -T @var{script} |
2d59b2c3 | 751 | @cindex script files |
de87cdb4 | 752 | Read link commands from the file @var{commandfile}. These commands |
8de26d62 DM |
753 | replace @code{ld}'s default link script (rather than adding |
754 | to it), so @var{commandfile} must specify everything necessary to describe | |
de87cdb4 DM |
755 | the target format. @xref{Commands}. If @var{commandfile} does not |
756 | exist, @code{ld} looks for it in the directories specified by any | |
757 | preceding @samp{-L} options. Multiple @samp{-T} options accumulate. | |
f22eee08 | 758 | |
2c5c0674 RP |
759 | @kindex -t |
760 | @cindex verbose | |
761 | @cindex input files, displaying | |
d4e5e3c3 | 762 | @item -t |
ec40bbb8 | 763 | Print the names of the input files as @code{ld} processes them. |
f22eee08 | 764 | |
c96386c4 ILT |
765 | @kindex -traditional-format |
766 | @cindex traditional format | |
767 | @item -traditional-format | |
768 | For some targets, the output of @code{ld} is different in some ways from | |
769 | the output of some existing linker. This switch requests @code{ld} to | |
770 | use the traditional format instead. | |
771 | ||
772 | @cindex dbx | |
773 | For example, on SunOS, @code{ld} combines duplicate entries in the | |
774 | symbol string table. This can reduce the size of an output file with | |
775 | full debugging information by over 30 percent. Unfortunately, the SunOS | |
776 | @code{dbx} program can not read the resulting program (@code{gdb} has no | |
777 | trouble). The @samp{-traditional-format} switch tells @code{ld} to not | |
778 | combine duplicate entries. | |
779 | ||
de87cdb4 DM |
780 | @item -u @var{symbol} |
781 | @kindex -u @var{symbol} | |
2c5c0674 | 782 | @cindex undefined symbol |
de87cdb4 | 783 | Force @var{symbol} to be entered in the output file as an undefined symbol. |
ec40bbb8 DM |
784 | Doing this may, for example, trigger linking of additional modules from |
785 | standard libraries. @samp{-u} may be repeated with different option | |
2c5c0674 RP |
786 | arguments to enter additional undefined symbols. |
787 | @c Nice idea, but no such command: This option is equivalent | |
788 | @c to the @code{EXTERN} linker command. | |
f22eee08 | 789 | |
2c5c0674 | 790 | @kindex -Ur |
b4d4e8e3 | 791 | @cindex constructors |
d4e5e3c3 | 792 | @item -Ur |
b4d4e8e3 | 793 | For anything other than C++ programs, this option is equivalent to |
ec40bbb8 DM |
794 | @samp{-r}: it generates relocatable output---i.e., an output file that can in |
795 | turn serve as input to @code{ld}. When linking C++ programs, @samp{-Ur} | |
1fb57a5d | 796 | @emph{does} resolve references to constructors, unlike @samp{-r}. |
3e27cc11 | 797 | It does not work to use @samp{-Ur} on files that were themselves linked |
1fb57a5d | 798 | with @samp{-Ur}; once the constructor table has been built, it cannot |
3e27cc11 DM |
799 | be added to. Use @samp{-Ur} only for the last partial link, and |
800 | @samp{-r} for the others. | |
b4d4e8e3 | 801 | |
01bc8f35 | 802 | @kindex --verbose |
8ddef552 | 803 | @cindex version |
01bc8f35 | 804 | @item --verbose |
1fb57a5d RP |
805 | Display the version number for @code{ld} and list the linker emulations |
806 | supported. Display which input files can and cannot be opened. | |
8ddef552 | 807 | |
2c5c0674 | 808 | @kindex -v |
01bc8f35 | 809 | @kindex -V |
b4d4e8e3 | 810 | @cindex version |
d4e5e3c3 | 811 | @item -v |
01bc8f35 ILT |
812 | @itemx -V |
813 | Display the version number for @code{ld}. The @code{-V} option also | |
814 | lists the supported emulations. | |
de87cdb4 | 815 | |
d4e5e3c3 DM |
816 | @item -version |
817 | @kindex -version | |
de87cdb4 | 818 | Display the version number for @code{ld} and exit. |
b4d4e8e3 | 819 | |
2a28d8b0 | 820 | @item -warn-common |
7f9ae73e RP |
821 | @kindex -warn-comon |
822 | @cindex warnings, on combining symbols | |
823 | @cindex combining symbols, warnings on | |
2a28d8b0 DM |
824 | Warn when a common symbol is combined with another common symbol or with |
825 | a symbol definition. Unix linkers allow this somewhat sloppy practice, | |
826 | but linkers on some other operating systems do not. This option allows | |
827 | you to find potential problems from combining global symbols. | |
1cd4cca9 DM |
828 | Unfortunately, some C libraries use this practice, so you may get some |
829 | warnings about symbols in the libraries as well as in your programs. | |
2a28d8b0 DM |
830 | |
831 | There are three kinds of global symbols, illustrated here by C examples: | |
832 | ||
833 | @table @samp | |
834 | @item int i = 1; | |
835 | A definition, which goes in the initialized data section of the output | |
836 | file. | |
837 | ||
838 | @item extern int i; | |
839 | An undefined reference, which does not allocate space. | |
840 | There must be either a definition or a common symbol for the | |
841 | variable somewhere. | |
842 | ||
843 | @item int i; | |
844 | A common symbol. If there are only (one or more) common symbols for a | |
845 | variable, it goes in the uninitialized data area of the output file. | |
846 | The linker merges multiple common symbols for the same variable into a | |
847 | single symbol. If they are of different sizes, it picks the largest | |
848 | size. The linker turns a common symbol into a declaration, if there is | |
849 | a definition of the same variable. | |
850 | @end table | |
851 | ||
852 | The @samp{-warn-common} option can produce five kinds of warnings. Each | |
853 | warning consists of a pair of lines: the first describes the symbol just | |
854 | encountered, and the second describes the previous symbol encountered | |
855 | with the same name. One or both of the two symbols will be a common | |
856 | symbol. | |
857 | ||
858 | @enumerate | |
859 | @item | |
860 | Turning a common symbol into a reference, because there is already a | |
861 | definition for the symbol. | |
862 | @smallexample | |
8920addc RP |
863 | @var{file}(@var{section}): warning: common of `@var{symbol}' |
864 | overridden by definition | |
2a28d8b0 DM |
865 | @var{file}(@var{section}): warning: defined here |
866 | @end smallexample | |
867 | ||
868 | @item | |
869 | Turning a common symbol into a reference, because a later definition for | |
870 | the symbol is encountered. This is the same as the previous case, | |
871 | except that the symbols are encountered in a different order. | |
872 | @smallexample | |
8920addc RP |
873 | @var{file}(@var{section}): warning: definition of `@var{symbol}' |
874 | overriding common | |
2a28d8b0 DM |
875 | @var{file}(@var{section}): warning: common is here |
876 | @end smallexample | |
877 | ||
878 | @item | |
879 | Merging a common symbol with a previous same-sized common symbol. | |
880 | @smallexample | |
8920addc RP |
881 | @var{file}(@var{section}): warning: multiple common |
882 | of `@var{symbol}' | |
2a28d8b0 DM |
883 | @var{file}(@var{section}): warning: previous common is here |
884 | @end smallexample | |
885 | ||
886 | @item | |
887 | Merging a common symbol with a previous larger common symbol. | |
888 | @smallexample | |
8920addc RP |
889 | @var{file}(@var{section}): warning: common of `@var{symbol}' |
890 | overridden by larger common | |
2a28d8b0 DM |
891 | @var{file}(@var{section}): warning: larger common is here |
892 | @end smallexample | |
893 | ||
894 | @item | |
895 | Merging a common symbol with a previous smaller common symbol. This is | |
896 | the same as the previous case, except that the symbols are | |
897 | encountered in a different order. | |
898 | @smallexample | |
8920addc RP |
899 | @var{file}(@var{section}): warning: common of `@var{symbol}' |
900 | overriding smaller common | |
2a28d8b0 DM |
901 | @var{file}(@var{section}): warning: smaller common is here |
902 | @end smallexample | |
903 | @end enumerate | |
904 | ||
0b0642d6 ILT |
905 | @kindex -warn-constructors |
906 | @item -warn-constructors | |
907 | Warn if any global constructors are used. This is only useful for a few | |
908 | object file formats. For formats like COFF or ELF, the linker can not | |
909 | detect the use of global constructors. | |
910 | ||
01bc8f35 ILT |
911 | @kindex -warn-once |
912 | @cindex warnings, on undefined symbols | |
913 | @cindex undefined symbols, warnings on | |
914 | @item -warn-once | |
915 | Only warn once for each undefined symbol, rather than once per module | |
916 | which refers to it. | |
917 | ||
4551e108 ILT |
918 | @kindex --whole-archive |
919 | @cindex including an entire archive | |
920 | For each archive mentioned on the command line, include every object | |
921 | file in the archive in the link, rather than searching the archive for | |
922 | the required object files. This is normally used to turn an archive | |
923 | file into a shared library, forcing every object to be included in the | |
924 | resulting shared library. | |
925 | ||
2c5c0674 RP |
926 | @kindex -X |
927 | @cindex local symbols, deleting | |
928 | @cindex L, deleting symbols beginning | |
d4e5e3c3 | 929 | @item -X |
01bc8f35 ILT |
930 | Delete all temporary local symbols. For most targets, this is all local |
931 | symbols whose names begin with @samp{L}. | |
f22eee08 | 932 | |
2c5c0674 RP |
933 | @kindex -x |
934 | @cindex deleting local symbols | |
d4e5e3c3 | 935 | @item -x |
01bc8f35 | 936 | Delete all local symbols. |
b4d4e8e3 | 937 | |
1fb57a5d RP |
938 | @item -y @var{symbol} |
939 | @kindex -y @var{symbol} | |
d76ae847 | 940 | @cindex symbol tracing |
ec40bbb8 | 941 | Print the name of each linked file in which @var{symbol} appears. This |
d76ae847 RP |
942 | option may be given any number of times. On many systems it is necessary |
943 | to prepend an underscore. | |
944 | ||
945 | This option is useful when you have an undefined symbol in your link but | |
946 | don't know where the reference is coming from. | |
a1ad915d ILT |
947 | |
948 | @item -( @var{archives} -) | |
949 | @itemx --start-group @var{archives} --end-group | |
950 | @kindex -( | |
951 | @cindex groups of archives | |
952 | The @var{archives} should be a list of archive files. They may be | |
953 | either explicit file names, or @samp{-l} options. | |
954 | ||
955 | The specified archives are searched repeatedly until no new undefined | |
956 | references are created. Normally, an archive is searched only once in | |
957 | the order that it is specified on the command line. If a symbol in that | |
958 | archive is needed to resolve an undefined symbol referred to by an | |
959 | object in an archive that appears later on the command line, the linker | |
960 | would not be able to resolve that reference. By grouping the archives, | |
961 | they all be searched repeatedly until all possible references are | |
962 | resolved. | |
963 | ||
964 | Using this option has a significant performance cost. It is best to use | |
965 | it only when there are unavoidable circular references between two or | |
966 | more archives. | |
f22eee08 | 967 | @end table |
b4d4e8e3 | 968 | |
ec40bbb8 DM |
969 | @ifset UsesEnvVars |
970 | @node Environment | |
2c5c0674 RP |
971 | @section Environment Variables |
972 | ||
d76ae847 RP |
973 | You can change the behavior of @code{ld} with the environment |
974 | variable @code{GNUTARGET}. | |
2c5c0674 RP |
975 | |
976 | @kindex GNUTARGET | |
977 | @cindex default input format | |
978 | @code{GNUTARGET} determines the input-file object format if you don't | |
ec40bbb8 | 979 | use @samp{-b} (or its synonym @samp{-format}). Its value should be one |
2c5c0674 | 980 | of the BFD names for an input format (@pxref{BFD}). If there is no |
246504a5 | 981 | @code{GNUTARGET} in the environment, @code{ld} uses the natural format |
8920addc | 982 | of the target. If @code{GNUTARGET} is set to @code{default} then BFD attempts to discover the |
2c5c0674 RP |
983 | input format by examining binary input files; this method often |
984 | succeeds, but there are potential ambiguities, since there is no method | |
ec40bbb8 | 985 | of ensuring that the magic number used to specify object-file formats is |
2c5c0674 RP |
986 | unique. However, the configuration procedure for BFD on each system |
987 | places the conventional format for that system first in the search-list, | |
988 | so ambiguities are resolved in favor of convention. | |
ec40bbb8 | 989 | @end ifset |
2c5c0674 | 990 | |
ec40bbb8 | 991 | @node Commands |
2c5c0674 | 992 | @chapter Command Language |
f22eee08 | 993 | |
2c5c0674 | 994 | @cindex command files |
ec40bbb8 | 995 | The command language provides explicit control over the link process, |
b4d4e8e3 | 996 | allowing complete specification of the mapping between the linker's |
ec40bbb8 | 997 | input files and its output. It controls: |
b4d4e8e3 | 998 | @itemize @bullet |
2c5c0674 RP |
999 | @item |
1000 | input files | |
1001 | @item | |
1002 | file formats | |
1003 | @item | |
867a1b8a | 1004 | output file layout |
2c5c0674 RP |
1005 | @item |
1006 | addresses of sections | |
1007 | @item | |
1008 | placement of common blocks | |
b4d4e8e3 | 1009 | @end itemize |
f22eee08 | 1010 | |
2c5c0674 | 1011 | You may supply a command file (also known as a link script) to the |
ec40bbb8 | 1012 | linker either explicitly through the @samp{-T} option, or implicitly as |
2c5c0674 | 1013 | an ordinary file. If the linker opens a file which it cannot recognize |
867a1b8a | 1014 | as a supported object or archive format, it reports an error. |
2c5c0674 | 1015 | |
2c5c0674 | 1016 | @menu |
2d59b2c3 RP |
1017 | * Scripts:: Linker Scripts |
1018 | * Expressions:: Expressions | |
1019 | * MEMORY:: MEMORY Command | |
1020 | * SECTIONS:: SECTIONS Command | |
1021 | * Entry Point:: The Entry Point | |
867a1b8a | 1022 | * Option Commands:: Option Commands |
2c5c0674 RP |
1023 | @end menu |
1024 | ||
ec40bbb8 | 1025 | @node Scripts |
b4d4e8e3 | 1026 | @section Linker Scripts |
246504a5 | 1027 | The @code{ld} command language is a collection of statements; some are |
ec40bbb8 DM |
1028 | simple keywords setting a particular option, some are used to select and |
1029 | group input files or name output files; and two statement | |
b4d4e8e3 RP |
1030 | types have a fundamental and pervasive impact on the linking process. |
1031 | ||
2c5c0674 RP |
1032 | @cindex fundamental script commands |
1033 | @cindex commands, fundamental | |
1034 | @cindex output file layout | |
1035 | @cindex layout of output file | |
246504a5 | 1036 | The most fundamental command of the @code{ld} command language is the |
b4d4e8e3 RP |
1037 | @code{SECTIONS} command (@pxref{SECTIONS}). Every meaningful command |
1038 | script must have a @code{SECTIONS} command: it specifies a | |
1039 | ``picture'' of the output file's layout, in varying degrees of detail. | |
1040 | No other command is required in all cases. | |
1041 | ||
1042 | The @code{MEMORY} command complements @code{SECTIONS} by describing the | |
2c5c0674 | 1043 | available memory in the target architecture. This command is optional; |
246504a5 | 1044 | if you don't use a @code{MEMORY} command, @code{ld} assumes sufficient |
2c5c0674 RP |
1045 | memory is available in a contiguous block for all output. |
1046 | @xref{MEMORY}. | |
b4d4e8e3 | 1047 | |
2c5c0674 RP |
1048 | @cindex comments |
1049 | You may include comments in linker scripts just as in C: delimited | |
1050 | by @samp{/*} and @samp{*/}. As in C, comments are syntactically | |
1051 | equivalent to whitespace. | |
1052 | ||
ec40bbb8 | 1053 | @node Expressions |
f22eee08 | 1054 | @section Expressions |
2c5c0674 RP |
1055 | @cindex expression syntax |
1056 | @cindex arithmetic | |
b4d4e8e3 RP |
1057 | Many useful commands involve arithmetic expressions. The syntax for |
1058 | expressions in the command language is identical to that of C | |
1059 | expressions, with the following features: | |
1060 | @itemize @bullet | |
2c5c0674 RP |
1061 | @item |
1062 | All expressions evaluated as integers and | |
f22eee08 | 1063 | are of ``long'' or ``unsigned long'' type. |
2c5c0674 RP |
1064 | @item |
1065 | All constants are integers. | |
1066 | @item | |
1067 | All of the C arithmetic operators are provided. | |
1068 | @item | |
1069 | You may reference, define, and create global variables. | |
1070 | @item | |
1071 | You may call special purpose built-in functions. | |
b4d4e8e3 | 1072 | @end itemize |
f22eee08 | 1073 | |
2c5c0674 | 1074 | @menu |
2d59b2c3 RP |
1075 | * Integers:: Integers |
1076 | * Symbols:: Symbol Names | |
1077 | * Location Counter:: The Location Counter | |
1078 | * Operators:: Operators | |
1079 | * Evaluation:: Evaluation | |
1080 | * Assignment:: Assignment: Defining Symbols | |
867a1b8a | 1081 | * Arithmetic Functions:: Built-In Functions |
2c5c0674 RP |
1082 | @end menu |
1083 | ||
ec40bbb8 | 1084 | @node Integers |
f22eee08 | 1085 | @subsection Integers |
2c5c0674 RP |
1086 | @cindex integer notation |
1087 | @cindex octal integers | |
f22eee08 RP |
1088 | An octal integer is @samp{0} followed by zero or more of the octal |
1089 | digits (@samp{01234567}). | |
b4d4e8e3 | 1090 | @example |
2c5c0674 | 1091 | _as_octal = 0157255; |
b4d4e8e3 | 1092 | @end example |
f22eee08 | 1093 | |
2c5c0674 | 1094 | @cindex decimal integers |
f22eee08 RP |
1095 | A decimal integer starts with a non-zero digit followed by zero or |
1096 | more digits (@samp{0123456789}). | |
b4d4e8e3 | 1097 | @example |
2c5c0674 | 1098 | _as_decimal = 57005; |
b4d4e8e3 | 1099 | @end example |
f22eee08 | 1100 | |
2c5c0674 RP |
1101 | @cindex hexadecimal integers |
1102 | @kindex 0x | |
f22eee08 RP |
1103 | A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or |
1104 | more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}. | |
b4d4e8e3 RP |
1105 | @example |
1106 | _as_hex = 0xdead; | |
1107 | @end example | |
f22eee08 | 1108 | |
2c5c0674 | 1109 | @cindex negative integers |
ec40bbb8 | 1110 | To write a negative integer, use |
b4d4e8e3 RP |
1111 | the prefix operator @samp{-}; @pxref{Operators}. |
1112 | @example | |
b4d4e8e3 RP |
1113 | _as_neg = -57005; |
1114 | @end example | |
f22eee08 | 1115 | |
2c5c0674 RP |
1116 | @cindex scaled integers |
1117 | @cindex K and M integer suffixes | |
1118 | @cindex M and K integer suffixes | |
1119 | @cindex suffixes for integers | |
1120 | @cindex integer suffixes | |
b4d4e8e3 RP |
1121 | Additionally the suffixes @code{K} and @code{M} may be used to scale a |
1122 | constant by | |
c8072296 RP |
1123 | @c TEXI2ROFF-KILL |
1124 | @ifinfo | |
1125 | @c END TEXI2ROFF-KILL | |
1126 | @code{1024} or @code{1024*1024} | |
1127 | @c TEXI2ROFF-KILL | |
1128 | @end ifinfo | |
f22eee08 | 1129 | @tex |
b4d4e8e3 | 1130 | ${\rm 1024}$ or ${\rm 1024}^2$ |
f22eee08 | 1131 | @end tex |
c8072296 | 1132 | @c END TEXI2ROFF-KILL |
ec40bbb8 | 1133 | respectively. For example, the following all refer to the same quantity: |
f22eee08 RP |
1134 | |
1135 | @example | |
2c5c0674 RP |
1136 | _fourk_1 = 4K; |
1137 | _fourk_2 = 4096; | |
1138 | _fourk_3 = 0x1000; | |
f22eee08 | 1139 | @end example |
b4d4e8e3 | 1140 | |
ec40bbb8 | 1141 | @node Symbols |
b4d4e8e3 | 1142 | @subsection Symbol Names |
2c5c0674 RP |
1143 | @cindex symbol names |
1144 | @cindex names | |
1145 | @cindex quoted symbol names | |
1146 | @kindex " | |
1fb57a5d RP |
1147 | Unless quoted, symbol names start with a letter, underscore, or point |
1148 | and may include any letters, underscores, digits, points, | |
1149 | and hyphens. Unquoted symbol names must not conflict with any | |
b4d4e8e3 RP |
1150 | keywords. You can specify a symbol which contains odd characters or has |
1151 | the same name as a keyword, by surrounding the symbol name in double quotes: | |
1152 | @example | |
1153 | "SECTION" = 9; | |
1154 | "with a space" = "also with a space" + 10; | |
1155 | @end example | |
1156 | ||
1fb57a5d RP |
1157 | Since symbols can contain many non-alphabetic characters, it is safest |
1158 | to delimit symbols with spaces. For example, @samp{A-B} is one symbol, | |
1159 | whereas @samp{A - B} is an expression involving subtraction. | |
1160 | ||
ec40bbb8 | 1161 | @node Location Counter |
b4d4e8e3 | 1162 | @subsection The Location Counter |
2c5c0674 RP |
1163 | @kindex . |
1164 | @cindex dot | |
1165 | @cindex location counter | |
1166 | @cindex current output location | |
b4d4e8e3 RP |
1167 | The special linker variable @dfn{dot} @samp{.} always contains the |
1168 | current output location counter. Since the @code{.} always refers to | |
1169 | a location in an output section, it must always appear in an | |
1170 | expression within a @code{SECTIONS} command. The @code{.} symbol | |
1171 | may appear anywhere that an ordinary symbol is allowed in an | |
1172 | expression, but its assignments have a side effect. Assigning a value | |
1173 | to the @code{.} symbol will cause the location counter to be moved. | |
2c5c0674 | 1174 | @cindex holes |
b4d4e8e3 RP |
1175 | This may be used to create holes in the output section. The location |
1176 | counter may never be moved backwards. | |
1177 | @example | |
2c5c0674 RP |
1178 | SECTIONS |
1179 | @{ | |
d4e5e3c3 DM |
1180 | output : |
1181 | @{ | |
1182 | file1(.text) | |
1183 | . = . + 1000; | |
1184 | file2(.text) | |
1185 | . += 1000; | |
1186 | file3(.text) | |
1187 | @} = 0x1234; | |
2c5c0674 | 1188 | @} |
b4d4e8e3 | 1189 | @end example |
2c5c0674 RP |
1190 | @noindent |
1191 | In the previous example, @code{file1} is located at the beginning of the | |
1192 | output section, then there is a 1000 byte gap. Then @code{file2} | |
1193 | appears, also with a 1000 byte gap following before @code{file3} is | |
1194 | loaded. The notation @samp{= 0x1234} specifies what data to write in | |
1195 | the gaps (@pxref{Section Options}). | |
b4d4e8e3 | 1196 | |
d4e5e3c3 DM |
1197 | @iftex |
1198 | @vfill | |
1199 | @end iftex | |
1200 | ||
1201 | @need 5000 | |
ec40bbb8 | 1202 | @node Operators |
f22eee08 | 1203 | @subsection Operators |
2c5c0674 RP |
1204 | @cindex Operators for arithmetic |
1205 | @cindex arithmetic operators | |
1206 | @cindex precedence in expressions | |
b4d4e8e3 | 1207 | The linker recognizes the standard C set of arithmetic operators, with |
f22eee08 | 1208 | the standard bindings and precedence levels: |
c8072296 | 1209 | @c TEXI2ROFF-KILL |
b4d4e8e3 | 1210 | @ifinfo |
c8072296 | 1211 | @c END TEXI2ROFF-KILL |
f22eee08 | 1212 | @example |
c8072296 | 1213 | precedence associativity Operators Notes |
b4d4e8e3 | 1214 | (highest) |
c8072296 RP |
1215 | 1 left ! - ~ (1) |
1216 | 2 left * / % | |
1217 | 3 left + - | |
1218 | 4 left >> << | |
1219 | 5 left == != > < <= >= | |
1220 | 6 left & | |
1221 | 7 left | | |
1222 | 8 left && | |
1223 | 9 left || | |
1224 | 10 right ? : | |
1225 | 11 right &= += -= *= /= (2) | |
b4d4e8e3 | 1226 | (lowest) |
f22eee08 | 1227 | @end example |
2c5c0674 RP |
1228 | Notes: |
1229 | (1) Prefix operators | |
1230 | (2) @xref{Assignment} | |
c8072296 | 1231 | @c TEXI2ROFF-KILL |
b4d4e8e3 | 1232 | @end ifinfo |
f22eee08 | 1233 | @tex |
2c5c0674 RP |
1234 | \vskip \baselineskip |
1235 | %"lispnarrowing" is the extra indent used generally for @example | |
1236 | \hskip\lispnarrowing\vbox{\offinterlineskip | |
f22eee08 RP |
1237 | \hrule |
1238 | \halign | |
2c5c0674 RP |
1239 | {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr |
1240 | height2pt&\omit&&\omit&&\omit&\cr | |
1241 | &Precedence&& Associativity &&{\rm Operators}&\cr | |
1242 | height2pt&\omit&&\omit&&\omit&\cr | |
f22eee08 | 1243 | \noalign{\hrule} |
2c5c0674 | 1244 | height2pt&\omit&&\omit&&\omit&\cr |
b4d4e8e3 | 1245 | &highest&&&&&\cr |
2c5c0674 RP |
1246 | % '176 is tilde, '~' in tt font |
1247 | &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr | |
1248 | &2&&left&&* / \%&\cr | |
1249 | &3&&left&&+ -&\cr | |
1250 | &4&&left&&>> <<&\cr | |
1251 | &5&&left&&== != > < <= >=&\cr | |
f22eee08 | 1252 | &6&&left&&\&&\cr |
f22eee08 | 1253 | &7&&left&&|&\cr |
f22eee08 | 1254 | &8&&left&&{\&\&}&\cr |
f22eee08 | 1255 | &9&&left&&||&\cr |
2c5c0674 RP |
1256 | &10&&right&&? :&\cr |
1257 | &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr | |
b4d4e8e3 | 1258 | &lowest&&&&&\cr |
2c5c0674 | 1259 | height2pt&\omit&&\omit&&\omit&\cr} |
f22eee08 RP |
1260 | \hrule} |
1261 | @end tex | |
2c5c0674 RP |
1262 | @iftex |
1263 | { | |
1264 | @obeylines@parskip=0pt@parindent=0pt | |
1265 | @dag@quad Prefix operators. | |
1266 | @ddag@quad @xref{Assignment}. | |
1267 | } | |
1268 | @end iftex | |
c8072296 | 1269 | @c END TEXI2ROFF-KILL |
f22eee08 | 1270 | |
ec40bbb8 | 1271 | @node Evaluation |
b4d4e8e3 RP |
1272 | @subsection Evaluation |
1273 | ||
2c5c0674 RP |
1274 | @cindex lazy evaluation |
1275 | @cindex expression evaluation order | |
b4d4e8e3 RP |
1276 | The linker uses ``lazy evaluation'' for expressions; it only calculates |
1277 | an expression when absolutely necessary. The linker needs the value of | |
1278 | the start address, and the lengths of memory regions, in order to do any | |
1279 | linking at all; these values are computed as soon as possible when the | |
1280 | linker reads in the command file. However, other values (such as symbol | |
1281 | values) are not known or needed until after storage allocation. Such | |
1282 | values are evaluated later, when other information (such as the sizes of | |
1283 | output sections) is available for use in the symbol assignment | |
1284 | expression. | |
1285 | ||
ec40bbb8 | 1286 | @node Assignment |
b4d4e8e3 | 1287 | @subsection Assignment: Defining Symbols |
2c5c0674 RP |
1288 | @cindex assignment in scripts |
1289 | @cindex symbol definition, scripts | |
1290 | @cindex variables, defining | |
b4d4e8e3 RP |
1291 | You may create global symbols, and assign values (addresses) to global |
1292 | symbols, using any of the C assignment operators: | |
1293 | ||
1294 | @table @code | |
1295 | @item @var{symbol} = @var{expression} ; | |
2c5c0674 | 1296 | @itemx @var{symbol} &= @var{expression} ; |
b4d4e8e3 RP |
1297 | @itemx @var{symbol} += @var{expression} ; |
1298 | @itemx @var{symbol} -= @var{expression} ; | |
1299 | @itemx @var{symbol} *= @var{expression} ; | |
1300 | @itemx @var{symbol} /= @var{expression} ; | |
1301 | @end table | |
1302 | ||
246504a5 | 1303 | Two things distinguish assignment from other operators in @code{ld} |
b4d4e8e3 RP |
1304 | expressions. |
1305 | @itemize @bullet | |
2c5c0674 RP |
1306 | @item |
1307 | Assignment may only be used at the root of an expression; | |
b4d4e8e3 | 1308 | @samp{a=b+3;} is allowed, but @samp{a+b=3;} is an error. |
2c5c0674 RP |
1309 | |
1310 | @kindex ; | |
1311 | @cindex semicolon | |
1312 | @item | |
d76ae847 RP |
1313 | You must place a trailing semicolon (``@key{;}'') at the end of an |
1314 | assignment statement. | |
b4d4e8e3 RP |
1315 | @end itemize |
1316 | ||
1317 | Assignment statements may appear: | |
1318 | @itemize @bullet | |
2c5c0674 | 1319 | @item |
246504a5 | 1320 | as commands in their own right in an @code{ld} script; or |
2c5c0674 RP |
1321 | @item |
1322 | as independent statements within a @code{SECTIONS} command; or | |
1323 | @item | |
1324 | as part of the contents of a section definition in a | |
b4d4e8e3 RP |
1325 | @code{SECTIONS} command. |
1326 | @end itemize | |
1327 | ||
1328 | The first two cases are equivalent in effect---both define a symbol with | |
ec40bbb8 | 1329 | an absolute address. The last case defines a symbol whose address is |
b4d4e8e3 RP |
1330 | relative to a particular section (@pxref{SECTIONS}). |
1331 | ||
2c5c0674 RP |
1332 | @cindex absolute and relocatable symbols |
1333 | @cindex relocatable and absolute symbols | |
1334 | @cindex symbols, relocatable and absolute | |
1335 | When a linker expression is evaluated and assigned to a variable, it is | |
1336 | given either an absolute or a relocatable type. An absolute expression | |
1337 | type is one in which the symbol contains the value that it will have in | |
867a1b8a | 1338 | the output file; a relocatable expression type is one in which the |
2c5c0674 | 1339 | value is expressed as a fixed offset from the base of a section. |
b4d4e8e3 RP |
1340 | |
1341 | The type of the expression is controlled by its position in the script | |
2c5c0674 RP |
1342 | file. A symbol assigned within a section definition is created relative |
1343 | to the base of the section; a symbol assigned in any other place is | |
1344 | created as an absolute symbol. Since a symbol created within a | |
1345 | section definition is relative to the base of the section, it | |
1346 | will remain relocatable if relocatable output is requested. A symbol | |
1347 | may be created with an absolute value even when assigned to within a | |
1348 | section definition by using the absolute assignment function | |
1349 | @code{ABSOLUTE}. For example, to create an absolute symbol whose address | |
1350 | is the last byte of an output section named @code{.data}: | |
b4d4e8e3 | 1351 | @example |
2c5c0674 | 1352 | SECTIONS@{ @dots{} |
d4e5e3c3 DM |
1353 | .data : |
1354 | @{ | |
1355 | *(.data) | |
1356 | _edata = ABSOLUTE(.) ; | |
1357 | @} | |
2c5c0674 | 1358 | @dots{} @} |
b4d4e8e3 RP |
1359 | @end example |
1360 | ||
2c5c0674 RP |
1361 | The linker tries to put off the evaluation of an assignment until all |
1362 | the terms in the source expression are known (@pxref{Evaluation}). For | |
ec40bbb8 | 1363 | instance, the sizes of sections cannot be known until after allocation, |
2c5c0674 RP |
1364 | so assignments dependent upon these are not performed until after |
1365 | allocation. Some expressions, such as those depending upon the location | |
1366 | counter @dfn{dot}, @samp{.} must be evaluated during allocation. If the | |
1367 | result of an expression is required, but the value is not available, | |
1368 | then an error results. For example, a script like the following | |
b4d4e8e3 | 1369 | @example |
2c5c0674 | 1370 | SECTIONS @{ @dots{} |
d4e5e3c3 DM |
1371 | text 9+this_isnt_constant : |
1372 | @{ @dots{} | |
1373 | @} | |
2c5c0674 | 1374 | @dots{} @} |
b4d4e8e3 | 1375 | @end example |
2c5c0674 RP |
1376 | @kindex Non constant expression |
1377 | @noindent | |
1378 | will cause the error message ``@code{Non constant expression for initial | |
0b3499f6 ILT |
1379 | address}''. |
1380 | ||
1381 | @cindex provide | |
1382 | In some cases, it is desirable for a linker script to define a symbol | |
1383 | only if it is referenced, and only if it is not defined by any object | |
1384 | included in the link. For example, traditional linkers defined the | |
1385 | symbol @samp{etext}. However, ANSI C requires that the user be able to | |
1386 | use @samp{etext} as a function name without encountering an error. | |
1387 | The @code{PROVIDE} keyword may be used to define a symbol, such as | |
1388 | @samp{etext}, only if it is referenced but not defined. The syntax is | |
1389 | @code{PROVIDE(@var{symbol} = @var{expression})}. | |
b4d4e8e3 | 1390 | |
867a1b8a DM |
1391 | @node Arithmetic Functions |
1392 | @subsection Arithmetic Functions | |
2c5c0674 | 1393 | @cindex functions in expression language |
ec40bbb8 | 1394 | The command language includes a number of built-in |
2c5c0674 RP |
1395 | functions for use in link script expressions. |
1396 | @table @code | |
1397 | @item ABSOLUTE(@var{exp}) | |
1398 | @kindex ABSOLUTE(@var{exp}) | |
1399 | @cindex expression, absolute | |
ec40bbb8 DM |
1400 | Return the absolute (non-relocatable, as opposed to non-negative) value |
1401 | of the expression @var{exp}. Primarily useful to assign an absolute | |
1402 | value to a symbol within a section definition, where symbol values are | |
1403 | normally section-relative. | |
2c5c0674 RP |
1404 | |
1405 | @item ADDR(@var{section}) | |
1406 | @kindex ADDR(@var{section}) | |
1407 | @cindex section address | |
ec40bbb8 | 1408 | Return the absolute address of the named @var{section}. Your script must |
b4d4e8e3 | 1409 | previously have defined the location of that section. In the following |
ec40bbb8 | 1410 | example, @code{symbol_1} and @code{symbol_2} are assigned identical |
b4d4e8e3 | 1411 | values: |
f22eee08 | 1412 | @example |
2c5c0674 | 1413 | SECTIONS@{ @dots{} |
d4e5e3c3 DM |
1414 | .output1 : |
1415 | @{ | |
1416 | start_of_output_1 = ABSOLUTE(.); | |
1417 | @dots{} | |
1418 | @} | |
1419 | .output : | |
1420 | @{ | |
1421 | symbol_1 = ADDR(.output1); | |
1422 | symbol_2 = start_of_output_1; | |
1423 | @} | |
2c5c0674 RP |
1424 | @dots{} @} |
1425 | @end example | |
1426 | ||
1427 | @item ALIGN(@var{exp}) | |
1428 | @kindex ALIGN(@var{exp}) | |
1429 | @cindex rounding up location counter | |
ec40bbb8 | 1430 | Return the result of the current location counter (@code{.}) aligned to |
2c5c0674 RP |
1431 | the next @var{exp} boundary. @var{exp} must be an expression whose |
1432 | value is a power of two. This is equivalent to | |
1433 | @example | |
cb70c872 | 1434 | (. + @var{exp} - 1) & ~(@var{exp} - 1) |
2c5c0674 RP |
1435 | @end example |
1436 | ||
1437 | @code{ALIGN} doesn't change the value of the location counter---it just | |
1438 | does arithmetic on it. As an example, to align the output @code{.data} | |
1439 | section to the next @code{0x2000} byte boundary after the preceding | |
1440 | section and to set a variable within the section to the next | |
1441 | @code{0x8000} boundary after the input sections: | |
1442 | @example | |
1443 | SECTIONS@{ @dots{} | |
d4e5e3c3 DM |
1444 | .data ALIGN(0x2000): @{ |
1445 | *(.data) | |
1446 | variable = ALIGN(0x8000); | |
1447 | @} | |
2c5c0674 RP |
1448 | @dots{} @} |
1449 | @end example | |
1450 | @noindent | |
1451 | The first use of @code{ALIGN} in this example specifies the location of | |
1452 | a section because it is used as the optional @var{start} attribute of a | |
1453 | section definition (@pxref{Section Options}). The second use simply | |
1454 | defines the value of a variable. | |
1455 | ||
1456 | The built-in @code{NEXT} is closely related to @code{ALIGN}. | |
1457 | ||
1458 | @item DEFINED(@var{symbol}) | |
1459 | @kindex DEFINED(@var{symbol}) | |
1460 | @cindex symbol defaults | |
ec40bbb8 DM |
1461 | Return 1 if @var{symbol} is in the linker global symbol table and is |
1462 | defined, otherwise return 0. You can use this function to provide default | |
1463 | values for symbols. For example, the following command-file fragment shows how | |
2c5c0674 RP |
1464 | to set a global symbol @code{begin} to the first location in the |
1465 | @code{.text} section---but if a symbol called @code{begin} already | |
1466 | existed, its value is preserved: | |
d4e5e3c3 | 1467 | |
c8072296 | 1468 | @smallexample |
2c5c0674 | 1469 | SECTIONS@{ @dots{} |
d4e5e3c3 DM |
1470 | .text : @{ |
1471 | begin = DEFINED(begin) ? begin : . ; | |
1472 | @dots{} | |
1473 | @} | |
2c5c0674 | 1474 | @dots{} @} |
c8072296 | 1475 | @end smallexample |
f22eee08 | 1476 | |
2c5c0674 RP |
1477 | @item NEXT(@var{exp}) |
1478 | @kindex NEXT(@var{exp}) | |
1479 | @cindex unallocated address, next | |
ec40bbb8 DM |
1480 | Return the next unallocated address that is a multiple of @var{exp}. |
1481 | This function is closely related to @code{ALIGN(@var{exp})}; unless you | |
2c5c0674 | 1482 | use the @code{MEMORY} command to define discontinuous memory for the |
ec40bbb8 | 1483 | output file, the two functions are equivalent. |
2c5c0674 RP |
1484 | |
1485 | @item SIZEOF(@var{section}) | |
1486 | @kindex SIZEOF(@var{section}) | |
1487 | @cindex section size | |
ec40bbb8 DM |
1488 | Return the size in bytes of the named @var{section}, if that section has |
1489 | been allocated. In the following example, @code{symbol_1} and | |
f22eee08 | 1490 | @code{symbol_2} are assigned identical values: |
ec40bbb8 | 1491 | @c What does it return if the section hasn't been allocated? 0? |
f22eee08 | 1492 | @example |
2c5c0674 | 1493 | SECTIONS@{ @dots{} |
d4e5e3c3 DM |
1494 | .output @{ |
1495 | .start = . ; | |
1496 | @dots{} | |
1497 | .end = . ; | |
1498 | @} | |
1499 | symbol_1 = .end - .start ; | |
1500 | symbol_2 = SIZEOF(.output); | |
2c5c0674 | 1501 | @dots{} @} |
f22eee08 | 1502 | |
f22eee08 | 1503 | @end example |
b4d4e8e3 | 1504 | |
2c5c0674 RP |
1505 | @item SIZEOF_HEADERS |
1506 | @kindex SIZEOF_HEADERS | |
1507 | @cindex header size | |
1508 | @itemx sizeof_headers | |
1509 | @kindex sizeof_headers | |
ec40bbb8 | 1510 | Return the size in bytes of the output file's headers. You can use this number |
2c5c0674 RP |
1511 | as the start address of the first section, if you choose, to facilitate |
1512 | paging. | |
1513 | ||
1514 | @end table | |
1515 | ||
ec40bbb8 | 1516 | @node MEMORY |
867a1b8a | 1517 | @section Memory Layout |
2c5c0674 RP |
1518 | @kindex MEMORY |
1519 | @cindex regions of memory | |
1520 | @cindex discontinuous memory | |
1521 | @cindex allocating memory | |
ec40bbb8 DM |
1522 | The linker's default configuration permits allocation of all available memory. |
1523 | You can override this configuration by using the @code{MEMORY} command. The | |
b4d4e8e3 RP |
1524 | @code{MEMORY} command describes the location and size of blocks of |
1525 | memory in the target. By using it carefully, you can describe which | |
1526 | memory regions may be used by the linker, and which memory regions it | |
1527 | must avoid. The linker does not shuffle sections to fit into the | |
1528 | available regions, but does move the requested sections into the correct | |
1529 | regions and issue errors when the regions become too full. | |
1530 | ||
867a1b8a | 1531 | A command file may contain at most one use of the @code{MEMORY} |
b4d4e8e3 RP |
1532 | command; however, you can define as many blocks of memory within it as |
1533 | you wish. The syntax is: | |
c8072296 | 1534 | |
f22eee08 | 1535 | @example |
b4d4e8e3 | 1536 | MEMORY |
d4e5e3c3 DM |
1537 | @{ |
1538 | @var{name} (@var{attr}) : ORIGIN = @var{origin}, LENGTH = @var{len} | |
1539 | @dots{} | |
1540 | @} | |
f22eee08 RP |
1541 | @end example |
1542 | @table @code | |
2c5c0674 | 1543 | @cindex naming memory regions |
d4e5e3c3 | 1544 | @item @var{name} |
f22eee08 RP |
1545 | is a name used internally by the linker to refer to the region. Any |
1546 | symbol name may be used. The region names are stored in a separate | |
ec40bbb8 | 1547 | name space, and will not conflict with symbols, file names or section |
b4d4e8e3 | 1548 | names. Use distinct names to specify multiple regions. |
d4e5e3c3 | 1549 | |
2c5c0674 | 1550 | @cindex memory region attributes |
d4e5e3c3 | 1551 | @item (@var{attr}) |
2c5c0674 | 1552 | is an optional list of attributes, permitted for compatibility with the |
246504a5 | 1553 | AT&T linker but not used by @code{ld} beyond checking that the |
2c5c0674 RP |
1554 | attribute list is valid. Valid attribute lists must be made up of the |
1555 | characters ``@code{LIRWX}''. If you omit the attribute list, you may | |
1556 | omit the parentheses around it as well. | |
d4e5e3c3 | 1557 | |
cb70c872 RP |
1558 | @kindex ORIGIN = |
1559 | @kindex o = | |
1560 | @kindex org = | |
d4e5e3c3 | 1561 | @item @var{origin} |
ec40bbb8 DM |
1562 | is the start address of the region in physical memory. It is |
1563 | an expression that must evaluate to a constant before | |
f22eee08 | 1564 | memory allocation is performed. The keyword @code{ORIGIN} may be |
867a1b8a | 1565 | abbreviated to @code{org} or @code{o} (but not, for example, @samp{ORG}). |
d4e5e3c3 | 1566 | |
cb70c872 RP |
1567 | @kindex LENGTH = |
1568 | @kindex len = | |
1569 | @kindex l = | |
d4e5e3c3 | 1570 | @item @var{len} |
b4d4e8e3 | 1571 | is the size in bytes of the region (an expression). |
2c5c0674 | 1572 | The keyword @code{LENGTH} may be abbreviated to @code{len} or @code{l}. |
f22eee08 RP |
1573 | @end table |
1574 | ||
1575 | For example, to specify that memory has two regions available for | |
ec40bbb8 | 1576 | allocation---one starting at 0 for 256 kilobytes, and the other |
2c5c0674 | 1577 | starting at @code{0x40000000} for four megabytes: |
f22eee08 RP |
1578 | |
1579 | @example | |
b4d4e8e3 | 1580 | MEMORY |
d4e5e3c3 DM |
1581 | @{ |
1582 | rom : ORIGIN = 0, LENGTH = 256K | |
1583 | ram : org = 0x40000000, l = 4M | |
1584 | @} | |
f22eee08 RP |
1585 | @end example |
1586 | ||
b4d4e8e3 | 1587 | Once you have defined a region of memory named @var{mem}, you can direct |
2c5c0674 RP |
1588 | specific output sections there by using a command ending in |
1589 | @samp{>@var{mem}} within the @code{SECTIONS} command (@pxref{Section | |
1590 | Options}). If the combined output sections directed to a region are too | |
1591 | big for the region, the linker will issue an error message. | |
b4d4e8e3 | 1592 | |
ec40bbb8 | 1593 | @node SECTIONS |
867a1b8a | 1594 | @section Specifying Output Sections |
67c4333b | 1595 | |
2c5c0674 | 1596 | @kindex SECTIONS |
b4d4e8e3 | 1597 | The @code{SECTIONS} command controls exactly where input sections are |
867a1b8a DM |
1598 | placed into output sections, their order in the output file, and to |
1599 | which output sections they are allocated. | |
b4d4e8e3 | 1600 | |
867a1b8a | 1601 | You may use at most one @code{SECTIONS} command in a script file, |
b4d4e8e3 RP |
1602 | but you can have as many statements within it as you wish. Statements |
1603 | within the @code{SECTIONS} command can do one of three things: | |
67c4333b | 1604 | |
b4d4e8e3 RP |
1605 | @itemize @bullet |
1606 | @item | |
1607 | define the entry point; | |
67c4333b | 1608 | |
b4d4e8e3 RP |
1609 | @item |
1610 | assign a value to a symbol; | |
67c4333b | 1611 | |
b4d4e8e3 | 1612 | @item |
867a1b8a DM |
1613 | describe the placement of a named output section, and which input |
1614 | sections go into it. | |
b4d4e8e3 RP |
1615 | @end itemize |
1616 | ||
8920addc RP |
1617 | You can also use the first two operations---defining the entry point and |
1618 | defining symbols---outside the @code{SECTIONS} command: @pxref{Entry | |
1619 | Point}, and @pxref{Assignment}. They are permitted here as well for | |
1620 | your convenience in reading the script, so that symbols and the entry | |
1621 | point can be defined at meaningful points in your output-file layout. | |
f22eee08 | 1622 | |
67c4333b | 1623 | If you do not use a @code{SECTIONS} command, the linker places each input |
867a1b8a DM |
1624 | section into an identically named output section in the order that the |
1625 | sections are first encountered in the input files. If all input sections | |
1626 | are present in the first file, for example, the order of sections in the | |
1627 | output file will match the order in the first input file. | |
b4d4e8e3 | 1628 | |
2c5c0674 | 1629 | @menu |
2d59b2c3 | 1630 | * Section Definition:: Section Definitions |
867a1b8a DM |
1631 | * Section Placement:: Section Placement |
1632 | * Section Data Expressions:: Section Data Expressions | |
2d59b2c3 | 1633 | * Section Options:: Optional Section Attributes |
2c5c0674 RP |
1634 | @end menu |
1635 | ||
ec40bbb8 | 1636 | @node Section Definition |
b4d4e8e3 | 1637 | @subsection Section Definitions |
2c5c0674 | 1638 | @cindex section definition |
b4d4e8e3 | 1639 | The most frequently used statement in the @code{SECTIONS} command is |
867a1b8a | 1640 | the @dfn{section definition}, which specifies the |
b4d4e8e3 | 1641 | properties of an output section: its location, alignment, contents, |
ec40bbb8 | 1642 | fill pattern, and target memory region. Most of |
b4d4e8e3 RP |
1643 | these specifications are optional; the simplest form of a section |
1644 | definition is | |
1645 | @example | |
2c5c0674 | 1646 | SECTIONS @{ @dots{} |
d4e5e3c3 DM |
1647 | @var{secname} : @{ |
1648 | @var{contents} | |
1649 | @} | |
2c5c0674 | 1650 | @dots{} @} |
b4d4e8e3 | 1651 | @end example |
2c5c0674 | 1652 | @cindex naming output sections |
b4d4e8e3 RP |
1653 | @noindent |
1654 | @var{secname} is the name of the output section, and @var{contents} a | |
ec40bbb8 | 1655 | specification of what goes there---for example, a list of input files or |
867a1b8a DM |
1656 | sections of input files (@pxref{Section Placement}). As you might |
1657 | assume, the whitespace shown is optional. You do need the colon | |
1658 | @samp{:} and the braces @samp{@{@}}, however. | |
b4d4e8e3 RP |
1659 | |
1660 | @var{secname} must meet the constraints of your output format. In | |
1661 | formats which only support a limited number of sections, such as | |
1662 | @code{a.out}, the name must be one of the names supported by the format | |
2c5c0674 RP |
1663 | (@code{a.out}, for example, allows only @code{.text}, @code{.data} or |
1664 | @code{.bss}). If the output format supports any number of sections, but | |
1665 | with numbers and not names (as is the case for Oasys), the name should be | |
1666 | supplied as a quoted numeric string. A section name may consist of any | |
867a1b8a | 1667 | sequence of characters, but any name which does not conform to the standard |
246504a5 | 1668 | @code{ld} symbol name syntax must be quoted. |
ec40bbb8 | 1669 | @xref{Symbols, , Symbol Names}. |
2c5c0674 | 1670 | |
f9d3d71a ILT |
1671 | The linker will not create output sections which do not have any |
1672 | contents. This is for convenience when referring to input sections that | |
1673 | may or may not exist. For example, | |
1674 | @example | |
1675 | .foo @{ *(.foo @} | |
1676 | @end example | |
1677 | will only create a @samp{.foo} section in the output file if there is a | |
1678 | @samp{.foo} section in at least one input file. | |
1679 | ||
867a1b8a DM |
1680 | @node Section Placement |
1681 | @subsection Section Placement | |
67c4333b | 1682 | |
2c5c0674 | 1683 | @cindex contents of a section |
67c4333b RP |
1684 | In a section definition, you can specify the contents of an output |
1685 | section by listing particular input files, by listing particular | |
1686 | input-file sections, or by a combination of the two. You can also place | |
1687 | arbitrary data in the section, and define symbols relative to the | |
1688 | beginning of the section. | |
b4d4e8e3 RP |
1689 | |
1690 | The @var{contents} of a section definition may include any of the | |
1691 | following kinds of statement. You can include as many of these as you | |
1692 | like in a single section definition, separated from one another by | |
1693 | whitespace. | |
f22eee08 | 1694 | |
b4d4e8e3 | 1695 | @table @code |
2c5c0674 RP |
1696 | @kindex @var{filename} |
1697 | @cindex input files, section defn | |
1698 | @cindex files, including in output sections | |
d4e5e3c3 | 1699 | @item @var{filename} |
b4d4e8e3 | 1700 | You may simply name a particular input file to be placed in the current |
2c5c0674 | 1701 | output section; @emph{all} sections from that file are placed in the |
867a1b8a DM |
1702 | current section definition. If the file name has already been mentioned |
1703 | in another section definition, with an explicit section name list, then | |
1704 | only those sections which have not yet been allocated are used. | |
1705 | ||
1706 | To specify a list of particular files by name: | |
f22eee08 | 1707 | @example |
cb70c872 | 1708 | .data : @{ afile.o bfile.o cfile.o @} |
f22eee08 | 1709 | @end example |
2c5c0674 RP |
1710 | @noindent |
1711 | The example also illustrates that multiple statements can be included in | |
ec40bbb8 | 1712 | the contents of a section definition, since each file name is a separate |
2c5c0674 | 1713 | statement. |
f22eee08 | 1714 | |
d4e5e3c3 DM |
1715 | @kindex @var{filename}(@var{section}) |
1716 | @cindex files and sections, section defn | |
2c5c0674 RP |
1717 | @item @var{filename}( @var{section} ) |
1718 | @itemx @var{filename}( @var{section}, @var{section}, @dots{} ) | |
1719 | @itemx @var{filename}( @var{section} @var{section} @dots{} ) | |
2c5c0674 RP |
1720 | You can name one or more sections from your input files, for |
1721 | insertion in the current output section. If you wish to specify a list | |
1722 | of input-file sections inside the parentheses, you may separate the | |
1723 | section names by either commas or whitespace. | |
1724 | ||
d4e5e3c3 DM |
1725 | @cindex input sections to output section |
1726 | @kindex *(@var{section}) | |
b4d4e8e3 RP |
1727 | @item * (@var{section}) |
1728 | @itemx * (@var{section}, @var{section}, @dots{}) | |
836a5ee4 | 1729 | @itemx * (@var{section} @var{section} @dots{}) |
b4d4e8e3 | 1730 | Instead of explicitly naming particular input files in a link control |
246504a5 | 1731 | script, you can refer to @emph{all} files from the @code{ld} command |
ec40bbb8 | 1732 | line: use @samp{*} instead of a particular file name before the |
b4d4e8e3 RP |
1733 | parenthesized input-file section list. |
1734 | ||
867a1b8a DM |
1735 | If you have already explicitly included some files by name, @samp{*} |
1736 | refers to all @emph{remaining} files---those whose places in the output | |
1737 | file have not yet been defined. | |
1738 | ||
ec40bbb8 | 1739 | For example, to copy sections @code{1} through @code{4} from an Oasys file |
b4d4e8e3 RP |
1740 | into the @code{.text} section of an @code{a.out} file, and sections @code{13} |
1741 | and @code{14} into the @code{.data} section: | |
1742 | @example | |
2c5c0674 | 1743 | SECTIONS @{ |
d4e5e3c3 DM |
1744 | .text :@{ |
1745 | *("1" "2" "3" "4") | |
1746 | @} | |
1747 | ||
1748 | .data :@{ | |
1749 | *("13" "14") | |
1750 | @} | |
b4d4e8e3 | 1751 | @} |
f22eee08 RP |
1752 | @end example |
1753 | ||
d4e5e3c3 | 1754 | @cindex @code{[@var{section}@dots{}]}, not supported |
836a5ee4 DM |
1755 | @samp{[ @var{section} @dots{} ]} used to be accepted as an alternate way |
1756 | to specify named sections from all unallocated input files. Because | |
1757 | some operating systems (VMS) allow brackets in file names, that notation | |
1758 | is no longer supported. | |
1759 | ||
2c5c0674 RP |
1760 | @cindex uninitialized data |
1761 | @cindex commons in output | |
d4e5e3c3 DM |
1762 | @kindex *( COMMON ) |
1763 | @item @var{filename}@code{( COMMON )} | |
1764 | @itemx *( COMMON ) | |
b4d4e8e3 | 1765 | Specify where in your output file to place uninitialized data |
d76ae847 | 1766 | with this notation. @code{*(COMMON)} by itself refers to all |
b4d4e8e3 RP |
1767 | uninitialized data from all input files (so far as it is not yet |
1768 | allocated); @var{filename}@code{(COMMON)} refers to uninitialized data | |
1769 | from a particular file. Both are special cases of the general | |
1770 | mechanisms for specifying where to place input-file sections: | |
246504a5 | 1771 | @code{ld} permits you to refer to uninitialized data as if it |
b4d4e8e3 RP |
1772 | were in an input-file section named @code{COMMON}, regardless of the |
1773 | input file's format. | |
1774 | @end table | |
1775 | ||
2c5c0674 | 1776 | For example, the following command script arranges the output file into |
b4d4e8e3 RP |
1777 | three consecutive sections, named @code{.text}, @code{.data}, and |
1778 | @code{.bss}, taking the input for each from the correspondingly named | |
1779 | sections of all the input files: | |
d4e5e3c3 | 1780 | |
f22eee08 | 1781 | @example |
2c5c0674 | 1782 | SECTIONS @{ |
d76ae847 RP |
1783 | .text : @{ *(.text) @} |
1784 | .data : @{ *(.data) @} | |
1785 | .bss : @{ *(.bss) *(COMMON) @} | |
2c5c0674 | 1786 | @} |
f22eee08 | 1787 | @end example |
b4d4e8e3 RP |
1788 | |
1789 | The following example reads all of the sections from file @code{all.o} | |
1790 | and places them at the start of output section @code{outputa} which | |
1791 | starts at location @code{0x10000}. All of section @code{.input1} from | |
1792 | file @code{foo.o} follows immediately, in the same output section. All | |
1793 | of section @code{.input2} from @code{foo.o} goes into output section | |
1794 | @code{outputb}, followed by section @code{.input1} from @code{foo1.o}. | |
1795 | All of the remaining @code{.input1} and @code{.input2} sections from any | |
1796 | files are written to output section @code{outputc}. | |
1797 | ||
1798 | @example | |
2c5c0674 | 1799 | SECTIONS @{ |
d4e5e3c3 DM |
1800 | outputa 0x10000 : |
1801 | @{ | |
1802 | all.o | |
1803 | foo.o (.input1) | |
1804 | @} | |
1805 | outputb : | |
1806 | @{ | |
1807 | foo.o (.input2) | |
1808 | foo1.o (.input1) | |
1809 | @} | |
1810 | outputc : | |
1811 | @{ | |
1812 | *(.input1) | |
1813 | *(.input2) | |
1814 | @} | |
2c5c0674 | 1815 | @} |
b4d4e8e3 RP |
1816 | @end example |
1817 | ||
867a1b8a DM |
1818 | @node Section Data Expressions |
1819 | @subsection Section Data Expressions | |
67c4333b | 1820 | |
867a1b8a | 1821 | @cindex expressions in a section |
67c4333b RP |
1822 | The foregoing statements arrange, in your output file, data originating |
1823 | from your input files. You can also place data directly in an output | |
1824 | section from the link command script. Most of these additional | |
1825 | statements involve expressions; @pxref{Expressions}. Although these | |
1826 | statements are shown separately here for ease of presentation, no such | |
1827 | segregation is needed within a section definition in the @code{SECTIONS} | |
1828 | command; you can intermix them freely with any of the statements we've | |
1829 | just described. | |
f22eee08 | 1830 | |
b4d4e8e3 | 1831 | @table @code |
2c5c0674 RP |
1832 | @cindex input filename symbols |
1833 | @cindex filename symbols | |
d4e5e3c3 DM |
1834 | @kindex CREATE_OBJECT_SYMBOLS |
1835 | @item CREATE_OBJECT_SYMBOLS | |
ec40bbb8 DM |
1836 | Create a symbol for each input file |
1837 | in the current section, set to the address of the first byte of | |
867a1b8a | 1838 | data written from that input file. For instance, with @code{a.out} |
b4d4e8e3 RP |
1839 | files it is conventional to have a symbol for each input file. You can |
1840 | accomplish this by defining the output @code{.text} section as follows: | |
1841 | @example | |
1842 | SECTIONS @{ | |
d4e5e3c3 DM |
1843 | .text 0x2020 : |
1844 | @{ | |
1845 | CREATE_OBJECT_SYMBOLS | |
1846 | *(.text) | |
1847 | _etext = ALIGN(0x2000); | |
1848 | @} | |
1849 | @dots{} | |
2c5c0674 | 1850 | @} |
f22eee08 | 1851 | @end example |
b4d4e8e3 | 1852 | |
867a1b8a | 1853 | If @code{sample.ld} is a file containing this script, and @code{a.o}, |
b4d4e8e3 RP |
1854 | @code{b.o}, @code{c.o}, and @code{d.o} are four input files with |
1855 | contents like the following--- | |
f22eee08 | 1856 | @example |
b4d4e8e3 RP |
1857 | /* a.c */ |
1858 | ||
2c5c0674 | 1859 | afunction() @{ @} |
b4d4e8e3 RP |
1860 | int adata=1; |
1861 | int abss; | |
1862 | @end example | |
f22eee08 | 1863 | |
b4d4e8e3 | 1864 | @noindent |
867a1b8a | 1865 | @samp{ld -M -T sample.ld a.o b.o c.o d.o} would create a map like this, |
b4d4e8e3 RP |
1866 | containing symbols matching the object file names: |
1867 | @example | |
f22eee08 RP |
1868 | 00000000 A __DYNAMIC |
1869 | 00004020 B _abss | |
1870 | 00004000 D _adata | |
1871 | 00002020 T _afunction | |
1872 | 00004024 B _bbss | |
1873 | 00004008 D _bdata | |
1874 | 00002038 T _bfunction | |
1875 | 00004028 B _cbss | |
1876 | 00004010 D _cdata | |
1877 | 00002050 T _cfunction | |
1878 | 0000402c B _dbss | |
1879 | 00004018 D _ddata | |
1880 | 00002068 T _dfunction | |
1881 | 00004020 D _edata | |
1882 | 00004030 B _end | |
1883 | 00004000 T _etext | |
1884 | 00002020 t a.o | |
1885 | 00002038 t b.o | |
1886 | 00002050 t c.o | |
1887 | 00002068 t d.o | |
f22eee08 RP |
1888 | @end example |
1889 | ||
2c5c0674 | 1890 | @kindex @var{symbol} = @var{expression} ; |
2c5c0674 | 1891 | @kindex @var{symbol} @var{f}= @var{expression} ; |
d4e5e3c3 DM |
1892 | @item @var{symbol} = @var{expression} ; |
1893 | @itemx @var{symbol} @var{f}= @var{expression} ; | |
2c5c0674 RP |
1894 | @var{symbol} is any symbol name (@pxref{Symbols}). ``@var{f}='' |
1895 | refers to any of the operators @code{&= += -= *= /=} which combine | |
1896 | arithmetic and assignment. | |
1897 | ||
1898 | @cindex assignment, in section defn | |
1899 | When you assign a value to a symbol within a particular section | |
1900 | definition, the value is relative to the beginning of the section | |
1901 | (@pxref{Assignment}). If you write | |
d4e5e3c3 | 1902 | |
b4d4e8e3 | 1903 | @example |
2c5c0674 | 1904 | SECTIONS @{ |
b4d4e8e3 | 1905 | abs = 14 ; |
2c5c0674 | 1906 | @dots{} |
cb70c872 | 1907 | .data : @{ @dots{} rel = 14 ; @dots{} @} |
b4d4e8e3 | 1908 | abs2 = 14 + ADDR(.data); |
2c5c0674 RP |
1909 | @dots{} |
1910 | @} | |
f22eee08 | 1911 | @end example |
d4e5e3c3 | 1912 | |
2c5c0674 | 1913 | @c FIXME: Try above example! |
b4d4e8e3 | 1914 | @noindent |
ec40bbb8 | 1915 | @code{abs} and @code{rel} do not have the same value; @code{rel} has the |
b4d4e8e3 RP |
1916 | same value as @code{abs2}. |
1917 | ||
2c5c0674 | 1918 | @kindex BYTE(@var{expression}) |
2c5c0674 | 1919 | @kindex SHORT(@var{expression}) |
2c5c0674 | 1920 | @kindex LONG(@var{expression}) |
c477527c | 1921 | @kindex QUAD(@var{expression}) |
2c5c0674 | 1922 | @cindex direct output |
d4e5e3c3 DM |
1923 | @item BYTE(@var{expression}) |
1924 | @itemx SHORT(@var{expression}) | |
1925 | @itemx LONG(@var{expression}) | |
1926 | @itemx QUAD(@var{expression}) | |
c477527c ILT |
1927 | By including one of these four statements in a section definition, you |
1928 | can explicitly place one, two, four, or eight bytes (respectively) at | |
1929 | the current address of that section. @code{QUAD} is only supported when | |
1930 | using a 64 bit host or target. | |
ec40bbb8 DM |
1931 | |
1932 | @ifclear SingleFormat | |
1933 | Multiple-byte quantities are represented in whatever byte order is | |
1934 | appropriate for the output file format (@pxref{BFD}). | |
1935 | @end ifclear | |
b4d4e8e3 RP |
1936 | |
1937 | @item FILL(@var{expression}) | |
2c5c0674 RP |
1938 | @kindex FILL(@var{expression}) |
1939 | @cindex holes, filling | |
1940 | @cindex unspecified memory | |
867a1b8a | 1941 | Specify the ``fill pattern'' for the current section. Any otherwise |
b4d4e8e3 RP |
1942 | unspecified regions of memory within the section (for example, regions |
1943 | you skip over by assigning a new value to the location counter @samp{.}) | |
1944 | are filled with the two least significant bytes from the | |
1945 | @var{expression} argument. A @code{FILL} statement covers memory | |
1946 | locations @emph{after} the point it occurs in the section definition; by | |
1947 | including more than one @code{FILL} statement, you can have different | |
1948 | fill patterns in different parts of an output section. | |
1949 | @end table | |
1950 | ||
ec40bbb8 | 1951 | @node Section Options |
b4d4e8e3 | 1952 | @subsection Optional Section Attributes |
2c5c0674 | 1953 | @cindex section defn, full syntax |
b4d4e8e3 RP |
1954 | Here is the full syntax of a section definition, including all the |
1955 | optional portions: | |
1956 | ||
d76ae847 | 1957 | @smallexample |
2c5c0674 RP |
1958 | SECTIONS @{ |
1959 | @dots{} | |
67c4333b | 1960 | @var{secname} @var{start} BLOCK(@var{align}) (NOLOAD) : AT ( @var{ldadr} ) |
f9d3d71a | 1961 | @{ @var{contents} @} >@var{region} =@var{fill} |
2c5c0674 | 1962 | @dots{} |
b4d4e8e3 | 1963 | @} |
d76ae847 | 1964 | @end smallexample |
b4d4e8e3 RP |
1965 | |
1966 | @var{secname} and @var{contents} are required. @xref{Section | |
67c4333b RP |
1967 | Definition}, and @pxref{Section Placement} for details on |
1968 | @var{contents}. The remaining elements---@var{start}, | |
1969 | @code{BLOCK(@var{align)}}, @code{(NOLOAD)}, @code{AT ( @var{ldadr} )}, | |
f9d3d71a | 1970 | @code{>@var{region}}, and @code{=@var{fill}}---are all optional. |
f22eee08 | 1971 | |
b4d4e8e3 | 1972 | @table @code |
2c5c0674 RP |
1973 | @cindex start address, section |
1974 | @cindex section start | |
1975 | @cindex section address | |
d4e5e3c3 | 1976 | @item @var{start} |
b4d4e8e3 RP |
1977 | You can force the output section to be loaded at a specified address by |
1978 | specifying @var{start} immediately following the section name. | |
1979 | @var{start} can be represented as any expression. The following | |
1980 | example generates section @var{output} at location | |
1981 | @code{0x40000000}: | |
d4e5e3c3 | 1982 | |
b4d4e8e3 RP |
1983 | @example |
1984 | SECTIONS @{ | |
d4e5e3c3 DM |
1985 | @dots{} |
1986 | output 0x40000000: @{ | |
1987 | @dots{} | |
1988 | @} | |
1989 | @dots{} | |
b4d4e8e3 | 1990 | @} |
f22eee08 | 1991 | @end example |
f22eee08 | 1992 | |
2c5c0674 RP |
1993 | @kindex BLOCK(@var{align}) |
1994 | @cindex section alignment | |
1995 | @cindex aligning sections | |
d4e5e3c3 | 1996 | @item BLOCK(@var{align}) |
ec40bbb8 | 1997 | You can include @code{BLOCK()} specification to advance |
2c5c0674 RP |
1998 | the location counter @code{.} prior to the beginning of the section, so |
1999 | that the section will begin at the specified alignment. @var{align} is | |
2000 | an expression. | |
f22eee08 | 2001 | |
d76ae847 RP |
2002 | @kindex NOLOAD |
2003 | @cindex prevent unnecessary loading | |
67c4333b RP |
2004 | @cindex loading, preventing |
2005 | @item (NOLOAD) | |
d76ae847 RP |
2006 | Use @samp{(NOLOAD)} to prevent a section from being loaded into memory |
2007 | each time it is accessed. For example, in the script sample below, the | |
2008 | @code{ROM} segment is addressed at memory location @samp{0} and does not | |
2009 | need to be loaded into each object file: | |
67c4333b | 2010 | |
d76ae847 RP |
2011 | @example |
2012 | SECTIONS @{ | |
d4e5e3c3 DM |
2013 | ROM 0 (NOLOAD) : @{ @dots{} @} |
2014 | @dots{} | |
d76ae847 RP |
2015 | @} |
2016 | @end example | |
2017 | ||
67c4333b RP |
2018 | @kindex AT ( @var{ldadr} ) |
2019 | @cindex specify load address | |
2020 | @cindex load address, specifying | |
2021 | @item AT ( @var{ldadr} ) | |
2022 | The expression @var{ldadr} that follows the @code{AT} keyword specifies | |
2023 | the load address of the section. The default (if you do not use the | |
2024 | @code{AT} keyword) is to make the load address the same as the | |
2025 | relocation address. This feature is designed to make it easy to build a | |
2026 | ROM image. For example, this @code{SECTIONS} definition creates two | |
2027 | output sections: one called @samp{.text}, which starts at @code{0x1000}, | |
2028 | and one called @samp{.mdata}, which is loaded at the end of the | |
2029 | @samp{.text} section even though its relocation address is | |
2030 | @code{0x2000}. The symbol @code{_data} is defined with the value | |
2031 | @code{0x2000}: | |
2032 | ||
2033 | @smallexample | |
2034 | SECTIONS | |
139c8857 RP |
2035 | @{ |
2036 | .text 0x1000 : @{ *(.text) _etext = . ; @} | |
2037 | .mdata 0x2000 : | |
2038 | AT ( ADDR(.text) + SIZEOF ( .text ) ) | |
2039 | @{ _data = . ; *(.data); _edata = . ; @} | |
2040 | .bss 0x3000 : | |
2041 | @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@} | |
67c4333b RP |
2042 | @} |
2043 | @end smallexample | |
2044 | ||
2045 | The run-time initialization code (for C programs, usually @code{crt0}) | |
2046 | for use with a ROM generated this way has to include something like | |
2047 | the following, to copy the initialized data from the ROM image to its runtime | |
2048 | address: | |
2049 | ||
139c8857 | 2050 | @smallexample |
67c4333b RP |
2051 | char *src = _etext; |
2052 | char *dst = _data; | |
2053 | ||
139c8857 | 2054 | /* ROM has data at end of text; copy it. */ |
67c4333b | 2055 | while (dst < _edata) @{ |
139c8857 | 2056 | *dst++ = *src++; |
67c4333b RP |
2057 | @} |
2058 | ||
2059 | /* Zero bss */ | |
2060 | for (dst = _bstart; dst< _bend; dst++) | |
139c8857 RP |
2061 | *dst = 0; |
2062 | @end smallexample | |
67c4333b | 2063 | |
f9d3d71a ILT |
2064 | @kindex >@var{region} |
2065 | @cindex section, assigning to memory region | |
2066 | @cindex memory regions and sections | |
2067 | @item >@var{region} | |
2068 | Assign this section to a previously defined region of memory. | |
2069 | @xref{MEMORY}. | |
2070 | ||
2c5c0674 RP |
2071 | @kindex =@var{fill} |
2072 | @cindex section fill pattern | |
2073 | @cindex fill pattern, entire section | |
d4e5e3c3 DM |
2074 | @item =@var{fill} |
2075 | Including @code{=@var{fill}} in a section definition specifies the | |
2076 | initial fill value for that section. You may use any expression to | |
2077 | specify @var{fill}. Any unallocated holes in the current output section | |
2078 | when written to the output file will be filled with the two least | |
2079 | significant bytes of the value, repeated as necessary. You can also | |
2080 | change the fill value with a @code{FILL} statement in the @var{contents} | |
2081 | of a section definition. | |
f22eee08 | 2082 | |
f22eee08 | 2083 | @end table |
b4d4e8e3 | 2084 | |
ec40bbb8 | 2085 | @node Entry Point |
b4d4e8e3 | 2086 | @section The Entry Point |
2c5c0674 RP |
2087 | @kindex ENTRY(@var{symbol}) |
2088 | @cindex start of execution | |
2089 | @cindex first instruction | |
b4d4e8e3 RP |
2090 | The linker command language includes a command specifically for |
2091 | defining the first executable instruction in an output file (its | |
2092 | @dfn{entry point}). Its argument is a symbol name: | |
f22eee08 | 2093 | @example |
b4d4e8e3 | 2094 | ENTRY(@var{symbol}) |
f22eee08 | 2095 | @end example |
b4d4e8e3 RP |
2096 | |
2097 | Like symbol assignments, the @code{ENTRY} command may be placed either | |
2098 | as an independent command in the command file, or among the section | |
2099 | definitions within the @code{SECTIONS} command---whatever makes the most | |
2100 | sense for your layout. | |
2101 | ||
2c5c0674 | 2102 | @cindex entry point, defaults |
b4d4e8e3 RP |
2103 | @code{ENTRY} is only one of several ways of choosing the entry point. |
2104 | You may indicate it in any of the following ways (shown in descending | |
2105 | order of priority: methods higher in the list override methods lower down). | |
f22eee08 RP |
2106 | @itemize @bullet |
2107 | @item | |
ec40bbb8 | 2108 | the @samp{-e} @var{entry} command-line option; |
f22eee08 | 2109 | @item |
8de26d62 | 2110 | the @code{ENTRY(@var{symbol})} command in a linker control script; |
f22eee08 | 2111 | @item |
b4d4e8e3 | 2112 | the value of the symbol @code{start}, if present; |
f22eee08 | 2113 | @item |
b4d4e8e3 | 2114 | the address of the first byte of the @code{.text} section, if present; |
f22eee08 | 2115 | @item |
b4d4e8e3 | 2116 | The address @code{0}. |
f22eee08 | 2117 | @end itemize |
b4d4e8e3 | 2118 | |
2c5c0674 RP |
2119 | For example, you can use these rules to generate an entry point with an |
2120 | assignment statement: if no symbol @code{start} is defined within your | |
2121 | input files, you can simply define it, assigning it an appropriate | |
2122 | value--- | |
d4e5e3c3 | 2123 | |
f22eee08 | 2124 | @example |
b4d4e8e3 | 2125 | start = 0x2020; |
f22eee08 | 2126 | @end example |
b4d4e8e3 RP |
2127 | |
2128 | @noindent | |
2129 | The example shows an absolute address, but you can use any expression. | |
2130 | For example, if your input object files use some other symbol-name | |
2131 | convention for the entry point, you can just assign the value of | |
2132 | whatever symbol contains the start address to @code{start}: | |
d4e5e3c3 | 2133 | |
f22eee08 | 2134 | @example |
cb70c872 | 2135 | start = other_symbol ; |
f22eee08 | 2136 | @end example |
f22eee08 | 2137 | |
867a1b8a DM |
2138 | @node Option Commands |
2139 | @section Option Commands | |
b4d4e8e3 RP |
2140 | The command language includes a number of other commands that you can |
2141 | use for specialized purposes. They are similar in purpose to | |
2142 | command-line options. | |
2143 | ||
2144 | @table @code | |
1fb57a5d RP |
2145 | @kindex CONSTRUCTORS |
2146 | @cindex C++ constructors, arranging in link | |
2147 | @cindex constructors, arranging in link | |
2148 | @item CONSTRUCTORS | |
2149 | This command ties up C++ style constructor and destructor records. The | |
2150 | details of the constructor representation vary from one object format to | |
2151 | another, but usually lists of constructors and destructors appear as | |
2152 | special sections. The @code{CONSTRUCTORS} command specifies where the | |
2153 | linker is to place the data from these sections, relative to the rest of | |
2154 | the linked output. Constructor data is marked by the symbol | |
2155 | @w{@code{__CTOR_LIST__}} at the start, and @w{@code{__CTOR_LIST_END}} at | |
2156 | the end; destructor data is bracketed similarly, between | |
2157 | @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_LIST_END}}. (The compiler | |
2158 | must arrange to actually run this code; GNU C++ calls constructors from | |
2159 | a subroutine @code{__main}, which it inserts automatically into the | |
2160 | startup code for @code{main}, and destructors from @code{_exit}.) | |
2161 | ||
d4e5e3c3 | 2162 | @need 1000 |
2c5c0674 | 2163 | @kindex FLOAT |
2c5c0674 | 2164 | @kindex NOFLOAT |
1fb57a5d RP |
2165 | @item FLOAT |
2166 | @itemx NOFLOAT | |
2c5c0674 | 2167 | These keywords were used in some older linkers to request a particular |
246504a5 | 2168 | math subroutine library. @code{ld} doesn't use the keywords, assuming |
2c5c0674 RP |
2169 | instead that any necessary subroutines are in libraries specified using |
2170 | the general mechanisms for linking to archives; but to permit the use of | |
2171 | scripts that were written for the older linkers, the keywords | |
2172 | @code{FLOAT} and @code{NOFLOAT} are accepted and ignored. | |
2173 | ||
2c5c0674 RP |
2174 | @kindex FORCE_COMMON_ALLOCATION |
2175 | @cindex common allocation | |
1fb57a5d | 2176 | @item FORCE_COMMON_ALLOCATION |
ec40bbb8 | 2177 | This command has the same effect as the @samp{-d} command-line option: |
2c5c0674 | 2178 | to make @code{ld} assign space to common symbols even if a relocatable |
ec40bbb8 | 2179 | output file is specified (@samp{-r}). |
b4d4e8e3 | 2180 | |
2c5c0674 | 2181 | @kindex INPUT ( @var{files} ) |
2c5c0674 | 2182 | @cindex binary input files |
1fb57a5d RP |
2183 | @item INPUT ( @var{file}, @var{file}, @dots{} ) |
2184 | @itemx INPUT ( @var{file} @var{file} @dots{} ) | |
2c5c0674 | 2185 | Use this command to include binary input files in the link, without |
7f9ae73e RP |
2186 | including them in a particular section definition. |
2187 | Specify the full name for each @var{file}, including @samp{.a} if | |
2188 | required. | |
2189 | ||
2190 | @code{ld} searches for each @var{file} through the archive-library | |
2191 | search path, just as for files you specify on the command line. | |
2192 | See the description of @samp{-L} in @ref{Options,,Command Line | |
2193 | Options}. | |
b4d4e8e3 | 2194 | |
01bc8f35 ILT |
2195 | If you use @samp{-l@var{file}}, @code{ld} will transform the name to |
2196 | @code{lib@var{file}.a} as with the command line argument @samp{-l}. | |
2197 | ||
0b3499f6 ILT |
2198 | @kindex GROUP ( @var{files} ) |
2199 | @cindex grouping input files | |
2200 | @item GROUP ( @var{file}, @var{file}, @dots{} ) | |
2201 | @itemx GROUP ( @var{file} @var{file} @dots{} ) | |
2202 | This command is like @code{INPUT}, except that the named files should | |
2203 | all be archives, and they are searched repeatedly until no new undefined | |
2204 | references are created. See the description of @samp{-(} in | |
2205 | @ref{Options,,Command Line Options}. | |
2206 | ||
2c5c0674 | 2207 | @ignore |
b4d4e8e3 | 2208 | @item MAP ( @var{name} ) |
2c5c0674 RP |
2209 | @kindex MAP ( @var{name} ) |
2210 | @c MAP(...) appears to look for an F in the arg, ignoring all other | |
2211 | @c chars; if it finds one, it sets "map_option_f" to true. But nothing | |
2212 | @c checks map_option_f. Apparently a stub for the future... | |
2213 | @end ignore | |
b4d4e8e3 RP |
2214 | |
2215 | @item OUTPUT ( @var{filename} ) | |
2c5c0674 RP |
2216 | @kindex OUTPUT ( @var{filename} ) |
2217 | @cindex naming the output file | |
d76ae847 RP |
2218 | Use this command to name the link output file @var{filename}. The |
2219 | effect of @code{OUTPUT(@var{filename})} is identical to the effect of | |
c477527c ILT |
2220 | @w{@samp{-o @var{filename}}}, which overrides it. You can use this |
2221 | command to supply a default output-file name other than @code{a.out}. | |
2c5c0674 | 2222 | |
ec40bbb8 | 2223 | @ifclear SingleFormat |
2c5c0674 RP |
2224 | @item OUTPUT_ARCH ( @var{bfdname} ) |
2225 | @kindex OUTPUT_ARCH ( @var{bfdname} ) | |
2226 | @cindex machine architecture, output | |
2227 | Specify a particular output machine architecture, with one of the names | |
2228 | used by the BFD back-end routines (@pxref{BFD}). This command is often | |
2229 | unnecessary; the architecture is most often set implicitly by either the | |
2230 | system BFD configuration or as a side effect of the @code{OUTPUT_FORMAT} | |
ec40bbb8 | 2231 | command. |
2c5c0674 RP |
2232 | |
2233 | @item OUTPUT_FORMAT ( @var{bfdname} ) | |
2234 | @kindex OUTPUT_FORMAT ( @var{bfdname} ) | |
2235 | @cindex format, output file | |
1fb57a5d RP |
2236 | When @code{ld} is configured to support multiple object code formats, |
2237 | you can use this command to specify a particular output format. | |
2238 | @var{bfdname} is one of the names used by the BFD back-end routines | |
2239 | (@pxref{BFD}). The effect is identical to the effect of the | |
2240 | @samp{-oformat} command-line option. This selection affects only | |
2c5c0674 | 2241 | the output file; the related command @code{TARGET} affects primarily |
ec40bbb8 DM |
2242 | input files. |
2243 | @end ifclear | |
2c5c0674 RP |
2244 | |
2245 | @item SEARCH_DIR ( @var{path} ) | |
2246 | @kindex SEARCH_DIR ( @var{path} ) | |
2247 | @cindex path for libraries | |
2248 | @cindex search path, libraries | |
246504a5 | 2249 | Add @var{path} to the list of paths where @code{ld} looks for |
2c5c0674 | 2250 | archive libraries. @code{SEARCH_DIR(@var{path})} has the same |
ec40bbb8 | 2251 | effect as @samp{-L@var{path}} on the command line. |
2c5c0674 RP |
2252 | |
2253 | @item STARTUP ( @var{filename} ) | |
2254 | @kindex STARTUP ( @var{filename} ) | |
2255 | @cindex first input file | |
2256 | Ensure that @var{filename} is the first input file used in the link | |
2257 | process. | |
b4d4e8e3 | 2258 | |
ec40bbb8 | 2259 | @ifclear SingleFormat |
b4d4e8e3 | 2260 | @item TARGET ( @var{format} ) |
2c5c0674 RP |
2261 | @cindex input file format |
2262 | @kindex TARGET ( @var{format} ) | |
1fb57a5d RP |
2263 | When @code{ld} is configured to support multiple object code formats, |
2264 | you can use this command to change the input-file object code format | |
2265 | (like the command-line option @samp{-b} or its synonym @samp{-format}). | |
2266 | The argument @var{format} is one of the strings used by BFD to name | |
2267 | binary formats. If @code{TARGET} is specified but @code{OUTPUT_FORMAT} | |
2268 | is not, the last @code{TARGET} argument is also used as the default | |
2269 | format for the @code{ld} output file. @xref{BFD}. | |
2c5c0674 RP |
2270 | |
2271 | @kindex GNUTARGET | |
246504a5 | 2272 | If you don't use the @code{TARGET} command, @code{ld} uses the value of |
2c5c0674 | 2273 | the environment variable @code{GNUTARGET}, if available, to select the |
246504a5 | 2274 | output file format. If that variable is also absent, @code{ld} uses |
2c5c0674 | 2275 | the default format configured for your machine in the BFD libraries. |
ec40bbb8 | 2276 | @end ifclear |
b4d4e8e3 RP |
2277 | @end table |
2278 | ||
ec40bbb8 DM |
2279 | @ifset GENERIC |
2280 | @node Machine Dependent | |
1c48127e RP |
2281 | @chapter Machine Dependent Features |
2282 | ||
2283 | @cindex machine dependencies | |
246504a5 RP |
2284 | @code{ld} has additional features on some platforms; the following |
2285 | sections describe them. Machines where @code{ld} has no additional | |
1c48127e RP |
2286 | functionality are not listed. |
2287 | ||
2288 | @menu | |
246504a5 RP |
2289 | * H8/300:: @code{ld} and the H8/300 |
2290 | * i960:: @code{ld} and the Intel 960 family | |
1c48127e | 2291 | @end menu |
ec40bbb8 DM |
2292 | @end ifset |
2293 | ||
7f9ae73e | 2294 | @c FIXME! This could use @raisesections/@lowersections, but there seems to be a conflict |
ec40bbb8 DM |
2295 | @c between those and node-defaulting. |
2296 | @ifset H8300 | |
2297 | @ifclear GENERIC | |
7f9ae73e | 2298 | @raisesections |
ec40bbb8 DM |
2299 | @end ifclear |
2300 | @node H8/300 | |
246504a5 | 2301 | @section @code{ld} and the H8/300 |
1c48127e RP |
2302 | |
2303 | @cindex H8/300 support | |
246504a5 | 2304 | For the H8/300, @code{ld} can perform these global optimizations when |
1c48127e RP |
2305 | you specify the @samp{-relax} command-line option. |
2306 | ||
2307 | @table @emph | |
2308 | @item relaxing address modes | |
d76ae847 | 2309 | @cindex relaxing on H8/300 |
246504a5 | 2310 | @code{ld} finds all @code{jsr} and @code{jmp} instructions whose |
1c48127e RP |
2311 | targets are within eight bits, and turns them into eight-bit |
2312 | program-counter relative @code{bsr} and @code{bra} instructions, | |
2313 | respectively. | |
2314 | ||
2315 | @item synthesizing instructions | |
d76ae847 | 2316 | @cindex synthesizing on H8/300 |
1c48127e | 2317 | @c FIXME: specifically mov.b, or any mov instructions really? |
246504a5 | 2318 | @code{ld} finds all @code{mov.b} instructions which use the |
1c48127e RP |
2319 | sixteen-bit absolute address form, but refer to the top |
2320 | page of memory, and changes them to use the eight-bit address form. | |
2321 | (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into | |
2322 | @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the | |
2323 | top page of memory). | |
2324 | @end table | |
ec40bbb8 | 2325 | @ifclear GENERIC |
7f9ae73e | 2326 | @lowersections |
ec40bbb8 DM |
2327 | @end ifclear |
2328 | @end ifset | |
2329 | ||
f9d3d71a ILT |
2330 | @ifclear GENERIC |
2331 | @ifset Hitachi | |
2332 | @c This stuff is pointless to say unless you're especially concerned | |
2333 | @c with Hitachi chips; don't enable it for generic case, please. | |
2334 | @node Hitachi | |
2335 | @chapter @code{ld} and other Hitachi chips | |
2336 | ||
2337 | @code{ld} also supports the H8/300H, the H8/500, and the Hitachi SH. No | |
2338 | special features, commands, or command-line options are required for | |
2339 | these chips. | |
2340 | @end ifset | |
2341 | @end ifclear | |
2342 | ||
ec40bbb8 DM |
2343 | @ifset I960 |
2344 | @ifclear GENERIC | |
7f9ae73e | 2345 | @raisesections |
ec40bbb8 DM |
2346 | @end ifclear |
2347 | @node i960 | |
246504a5 | 2348 | @section @code{ld} and the Intel 960 family |
1c48127e RP |
2349 | |
2350 | @cindex i960 support | |
d76ae847 | 2351 | |
1c48127e RP |
2352 | You can use the @samp{-A@var{architecture}} command line option to |
2353 | specify one of the two-letter names identifying members of the 960 | |
2354 | family; the option specifies the desired output target, and warns of any | |
2355 | incompatible instructions in the input files. It also modifies the | |
2356 | linker's search strategy for archive libraries, to support the use of | |
2357 | libraries specific to each particular architecture, by including in the | |
2358 | search loop names suffixed with the string identifying the architecture. | |
2359 | ||
246504a5 | 2360 | For example, if your @code{ld} command line included @w{@samp{-ACA}} as |
1c48127e | 2361 | well as @w{@samp{-ltry}}, the linker would look (in its built-in search |
ec40bbb8 | 2362 | paths, and in any paths you specify with @samp{-L}) for a library with |
1c48127e | 2363 | the names |
ec40bbb8 | 2364 | |
1c48127e RP |
2365 | @example |
2366 | try | |
2367 | libtry.a | |
2368 | tryca | |
2369 | libtryca.a | |
2370 | @end example | |
ec40bbb8 | 2371 | |
1c48127e RP |
2372 | @noindent |
2373 | The first two possibilities would be considered in any event; the last | |
2374 | two are due to the use of @w{@samp{-ACA}}. | |
2375 | ||
ec40bbb8 | 2376 | You can meaningfully use @samp{-A} more than once on a command line, since |
1c48127e | 2377 | the 960 architecture family allows combination of target architectures; each |
ec40bbb8 | 2378 | use will add another pair of name variants to search for when @w{@samp{-l}} |
1c48127e | 2379 | specifies a library. |
1fb57a5d RP |
2380 | |
2381 | @cindex @code{-relax} on i960 | |
2382 | @cindex relaxing on i960 | |
2383 | @code{ld} supports the @samp{-relax} option for the i960 family. If you | |
2384 | specify @samp{-relax}, @code{ld} finds all @code{balx} and @code{calx} | |
2385 | instructions whose targets are within 24 bits, and turns them into | |
2386 | 24-bit program-counter relative @code{bal} and @code{cal} | |
2387 | instructions, respectively. @code{ld} also turns @code{cal} | |
2388 | instructions into @code{bal} instructions when it determines that the | |
2389 | target subroutine is a leaf routine (that is, the target subroutine does | |
2390 | not itself call any subroutines). | |
2391 | ||
ec40bbb8 | 2392 | @ifclear GENERIC |
7f9ae73e | 2393 | @lowersections |
ec40bbb8 DM |
2394 | @end ifclear |
2395 | @end ifset | |
1c48127e | 2396 | |
ec40bbb8 DM |
2397 | @ifclear SingleFormat |
2398 | @node BFD | |
f22eee08 RP |
2399 | @chapter BFD |
2400 | ||
2c5c0674 RP |
2401 | @cindex back end |
2402 | @cindex object file management | |
d4e5e3c3 DM |
2403 | @cindex object formats available |
2404 | @kindex objdump -i | |
2c5c0674 RP |
2405 | The linker accesses object and archive files using the BFD libraries. |
2406 | These libraries allow the linker to use the same routines to operate on | |
2407 | object files whatever the object file format. A different object file | |
2408 | format can be supported simply by creating a new BFD back end and adding | |
d4e5e3c3 DM |
2409 | it to the library. To conserve runtime memory, however, the linker and |
2410 | associated tools are usually configured to support only a subset of the | |
2411 | object file formats available. You can use @code{objdump -i} | |
1c48127e | 2412 | (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to |
d4e5e3c3 | 2413 | list all the formats available for your configuration. |
f22eee08 | 2414 | |
2c5c0674 RP |
2415 | @cindex BFD requirements |
2416 | @cindex requirements for BFD | |
2417 | As with most implementations, BFD is a compromise between | |
f22eee08 | 2418 | several conflicting requirements. The major factor influencing |
2c5c0674 RP |
2419 | BFD design was efficiency: any time used converting between |
2420 | formats is time which would not have been spent had BFD not | |
f22eee08 | 2421 | been involved. This is partly offset by abstraction payback; since |
2c5c0674 | 2422 | BFD simplifies applications and back ends, more time and care |
f22eee08 RP |
2423 | may be spent optimizing algorithms for a greater speed. |
2424 | ||
2c5c0674 RP |
2425 | One minor artifact of the BFD solution which you should bear in |
2426 | mind is the potential for information loss. There are two places where | |
ec40bbb8 | 2427 | useful information can be lost using the BFD mechanism: during |
2c5c0674 RP |
2428 | conversion and during output. @xref{BFD information loss}. |
2429 | ||
2430 | @menu | |
2d59b2c3 | 2431 | * BFD outline:: How it works: an outline of BFD |
2c5c0674 | 2432 | @end menu |
f22eee08 | 2433 | |
ec40bbb8 | 2434 | @node BFD outline |
b4d4e8e3 | 2435 | @section How it works: an outline of BFD |
2c5c0674 | 2436 | @cindex opening object files |
3e27cc11 | 2437 | @include bfdsumm.texi |
ec40bbb8 | 2438 | @end ifclear |
f22eee08 | 2439 | |
ec40bbb8 | 2440 | @node MRI |
2d59b2c3 RP |
2441 | @appendix MRI Compatible Script Files |
2442 | @cindex MRI compatibility | |
2443 | To aid users making the transition to @sc{gnu} @code{ld} from the MRI | |
2444 | linker, @code{ld} can use MRI compatible linker scripts as an | |
2445 | alternative to the more general-purpose linker scripting language | |
2446 | described in @ref{Commands,,Command Language}. MRI compatible linker | |
2447 | scripts have a much simpler command set than the scripting language | |
2448 | otherwise used with @code{ld}. @sc{gnu} @code{ld} supports the most | |
2449 | commonly used MRI linker commands; these commands are described here. | |
2450 | ||
867a1b8a DM |
2451 | In general, MRI scripts aren't of much use with the @code{a.out} object |
2452 | file format, since it only has three sections and MRI scripts lack some | |
2453 | features to make use of them. | |
2454 | ||
2d59b2c3 RP |
2455 | You can specify a file containing an MRI-compatible script using the |
2456 | @samp{-c} command-line option. | |
2457 | ||
2458 | Each command in an MRI-compatible script occupies its own line; each | |
2459 | command line starts with the keyword that identifies the command (though | |
2460 | blank lines are also allowed for punctuation). If a line of an | |
2461 | MRI-compatible script begins with an unrecognized keyword, @code{ld} | |
2462 | issues a warning message, but continues processing the script. | |
2463 | ||
2464 | Lines beginning with @samp{*} are comments. | |
2465 | ||
2466 | You can write these commands using all upper-case letters, or all | |
2467 | lower case; for example, @samp{chip} is the same as @samp{CHIP}. | |
2468 | The following list shows only the upper-case form of each command. | |
2469 | ||
2470 | @table @code | |
d4e5e3c3 | 2471 | @cindex @code{ABSOLUTE} (MRI) |
2d59b2c3 RP |
2472 | @item ABSOLUTE @var{secname} |
2473 | @item ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname} | |
2d59b2c3 RP |
2474 | Normally, @code{ld} includes in the output file all sections from all |
2475 | the input files. However, in an MRI-compatible script, you can use the | |
2476 | @code{ABSOLUTE} command to restrict the sections that will be present in | |
2477 | your output program. If the @code{ABSOLUTE} command is used at all in a | |
2478 | script, then only the sections named explicitly in @code{ABSOLUTE} | |
2479 | commands will appear in the linker output. You can still use other | |
2480 | input sections (whatever you select on the command line, or using | |
7b015547 | 2481 | @code{LOAD}) to resolve addresses in the output file. |
2d59b2c3 | 2482 | |
2d59b2c3 | 2483 | @cindex @code{ALIAS} (MRI) |
d4e5e3c3 | 2484 | @item ALIAS @var{out-secname}, @var{in-secname} |
2d59b2c3 RP |
2485 | Use this command to place the data from input section @var{in-secname} |
2486 | in a section called @var{out-secname} in the linker output file. | |
2487 | ||
2488 | @var{in-secname} may be an integer. | |
2489 | ||
e54bf1c1 ILT |
2490 | @cindex @code{ALIGN} (MRI) |
2491 | @item ALIGN @var{secname} = @var{expression} | |
2492 | Align the section called @var{secname} to @var{expression}. The | |
2493 | @var{expression} should be a power of two. | |
2494 | ||
2d59b2c3 | 2495 | @cindex @code{BASE} (MRI) |
d4e5e3c3 | 2496 | @item BASE @var{expression} |
2d59b2c3 RP |
2497 | Use the value of @var{expression} as the lowest address (other than |
2498 | absolute addresses) in the output file. | |
2499 | ||
d4e5e3c3 | 2500 | @cindex @code{CHIP} (MRI) |
2d59b2c3 RP |
2501 | @item CHIP @var{expression} |
2502 | @itemx CHIP @var{expression}, @var{expression} | |
ec40bbb8 | 2503 | This command does nothing; it is accepted only for compatibility. |
2d59b2c3 | 2504 | |
2d59b2c3 | 2505 | @cindex @code{END} (MRI) |
d4e5e3c3 | 2506 | @item END |
2d59b2c3 RP |
2507 | This command does nothing whatever; it's only accepted for compatibility. |
2508 | ||
2d59b2c3 | 2509 | @cindex @code{FORMAT} (MRI) |
d4e5e3c3 | 2510 | @item FORMAT @var{output-format} |
2d59b2c3 RP |
2511 | Similar to the @code{OUTPUT_FORMAT} command in the more general linker |
2512 | language, but restricted to one of these output formats: | |
d4e5e3c3 | 2513 | |
2d59b2c3 RP |
2514 | @enumerate |
2515 | @item | |
2516 | S-records, if @var{output-format} is @samp{S} | |
2517 | ||
2518 | @item | |
2519 | IEEE, if @var{output-format} is @samp{IEEE} | |
2520 | ||
2521 | @item | |
2522 | COFF (the @samp{coff-m68k} variant in BFD), if @var{output-format} is | |
2523 | @samp{COFF} | |
2524 | @end enumerate | |
2525 | ||
2d59b2c3 | 2526 | @cindex @code{LIST} (MRI) |
d4e5e3c3 | 2527 | @item LIST @var{anything}@dots{} |
2d59b2c3 RP |
2528 | Print (to the standard output file) a link map, as produced by the |
2529 | @code{ld} command-line option @samp{-M}. | |
2530 | ||
ec40bbb8 DM |
2531 | The keyword @code{LIST} may be followed by anything on the |
2532 | same line, with no change in its effect. | |
2d59b2c3 | 2533 | |
d4e5e3c3 | 2534 | @cindex @code{LOAD} (MRI) |
2d59b2c3 RP |
2535 | @item LOAD @var{filename} |
2536 | @item LOAD @var{filename}, @var{filename}, @dots{} @var{filename} | |
2d59b2c3 RP |
2537 | Include one or more object file @var{filename} in the link; this has the |
2538 | same effect as specifying @var{filename} directly on the @code{ld} | |
2539 | command line. | |
2540 | ||
2d59b2c3 | 2541 | @cindex @code{NAME} (MRI) |
d4e5e3c3 | 2542 | @item NAME @var{output-name} |
2d59b2c3 RP |
2543 | @var{output-name} is the name for the program produced by @code{ld}; the |
2544 | MRI-compatible command @code{NAME} is equivalent to the command-line | |
2545 | option @samp{-o} or the general script language command @code{OUTPUT}. | |
2546 | ||
d4e5e3c3 | 2547 | @cindex @code{ORDER} (MRI) |
2d59b2c3 RP |
2548 | @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname} |
2549 | @itemx ORDER @var{secname} @var{secname} @var{secname} | |
ec40bbb8 DM |
2550 | Normally, @code{ld} orders the sections in its output file in the |
2551 | order in which they first appear in the input files. In an MRI-compatible | |
2552 | script, you can override this ordering with the @code{ORDER} command. The | |
2d59b2c3 RP |
2553 | sections you list with @code{ORDER} will appear first in your output |
2554 | file, in the order specified. | |
2555 | ||
d4e5e3c3 | 2556 | @cindex @code{PUBLIC} (MRI) |
2d59b2c3 RP |
2557 | @item PUBLIC @var{name}=@var{expression} |
2558 | @itemx PUBLIC @var{name},@var{expression} | |
2559 | @itemx PUBLIC @var{name} @var{expression} | |
ec40bbb8 | 2560 | Supply a value (@var{expression}) for external symbol |
2d59b2c3 RP |
2561 | @var{name} used in the linker input files. |
2562 | ||
d4e5e3c3 | 2563 | @cindex @code{SECT} (MRI) |
2d59b2c3 RP |
2564 | @item SECT @var{secname}, @var{expression} |
2565 | @itemx SECT @var{secname}=@var{expression} | |
2566 | @itemx SECT @var{secname} @var{expression} | |
2d59b2c3 RP |
2567 | You can use any of these three forms of the @code{SECT} command to |
2568 | specify the start address (@var{expression}) for section @var{secname}. | |
2569 | If you have more than one @code{SECT} statement for the same | |
2570 | @var{secname}, only the @emph{first} sets the start address. | |
2571 | @end table | |
2572 | ||
2573 | ||
ec40bbb8 | 2574 | @node Index |
2c5c0674 RP |
2575 | @unnumbered Index |
2576 | ||
2577 | @printindex cp | |
2578 | ||
2579 | @tex | |
2580 | % I think something like @colophon should be in texinfo. In the | |
2581 | % meantime: | |
2582 | \long\def\colophon{\hbox to0pt{}\vfill | |
2583 | \centerline{The body of this manual is set in} | |
2584 | \centerline{\fontname\tenrm,} | |
2585 | \centerline{with headings in {\bf\fontname\tenbf}} | |
2586 | \centerline{and examples in {\tt\fontname\tentt}.} | |
2587 | \centerline{{\it\fontname\tenit\/} and} | |
2588 | \centerline{{\sl\fontname\tensl\/}} | |
2589 | \centerline{are used for emphasis.}\vfill} | |
2590 | \page\colophon | |
2591 | % Blame: pesch@cygnus.com, 28mar91. | |
2592 | @end tex | |
2593 | ||
2594 | ||
b4d4e8e3 | 2595 | @contents |
f22eee08 RP |
2596 | @bye |
2597 | ||
2598 |