Commit | Line | Data |
---|---|---|
252b5132 RH |
1 | \input texinfo |
2 | @setfilename ld.info | |
250d07de | 3 | @c Copyright (C) 1991-2021 Free Software Foundation, Inc. |
252b5132 | 4 | @syncodeindex ky cp |
dff70155 | 5 | @c man begin INCLUDE |
252b5132 RH |
6 | @include configdoc.texi |
7 | @c (configdoc.texi is generated by the Makefile) | |
c428fa83 | 8 | @include bfdver.texi |
dff70155 | 9 | @c man end |
252b5132 RH |
10 | |
11 | @c @smallbook | |
12 | ||
ff5dcc92 SC |
13 | @macro gcctabopt{body} |
14 | @code{\body\} | |
15 | @end macro | |
16 | ||
0285c67d NC |
17 | @c man begin NAME |
18 | @ifset man | |
19 | @c Configure for the generation of man pages | |
20 | @set UsesEnvVars | |
21 | @set GENERIC | |
0285c67d | 22 | @set ARM |
ac145307 | 23 | @set C6X |
b8891f8d | 24 | @set CSKY |
49fa1e15 | 25 | @set H8300 |
0285c67d | 26 | @set HPPA |
0285c67d | 27 | @set M68HC11 |
7fb9f789 | 28 | @set M68K |
833794fc | 29 | @set MIPS |
3c3bdf30 | 30 | @set MMIX |
2469cfa2 | 31 | @set MSP430 |
35c08157 | 32 | @set NDS32 |
78058a5e | 33 | @set NIOSII |
fa1477dc | 34 | @set PDP11 |
2a60a7a8 AM |
35 | @set POWERPC |
36 | @set POWERPC64 | |
49fa1e15 | 37 | @set Renesas |
b4cbbe8f | 38 | @set S/390 |
49fa1e15 AM |
39 | @set SPU |
40 | @set TICOFF | |
2ca22b03 | 41 | @set WIN32 |
e0001a05 | 42 | @set XTENSA |
0285c67d NC |
43 | @end ifset |
44 | @c man end | |
45 | ||
9160ea82 AM |
46 | @ifnottex |
47 | @dircategory Software development | |
48 | @direntry | |
252b5132 | 49 | * Ld: (ld). The GNU linker. |
9160ea82 AM |
50 | @end direntry |
51 | @end ifnottex | |
252b5132 | 52 | |
0e9517a9 | 53 | @copying |
e49e529d JM |
54 | This file documents the @sc{gnu} linker LD |
55 | @ifset VERSION_PACKAGE | |
56 | @value{VERSION_PACKAGE} | |
57 | @end ifset | |
58 | version @value{VERSION}. | |
252b5132 | 59 | |
250d07de | 60 | Copyright @copyright{} 1991-2021 Free Software Foundation, Inc. |
252b5132 | 61 | |
cf055d54 | 62 | Permission is granted to copy, distribute and/or modify this document |
793c5807 | 63 | under the terms of the GNU Free Documentation License, Version 1.3 |
cf055d54 NC |
64 | or any later version published by the Free Software Foundation; |
65 | with no Invariant Sections, with no Front-Cover Texts, and with no | |
66 | Back-Cover Texts. A copy of the license is included in the | |
36f63dca | 67 | section entitled ``GNU Free Documentation License''. |
0e9517a9 | 68 | @end copying |
252b5132 RH |
69 | @iftex |
70 | @finalout | |
71 | @setchapternewpage odd | |
71ba23f6 | 72 | @settitle The GNU linker |
252b5132 | 73 | @titlepage |
71ba23f6 | 74 | @title The GNU linker |
252b5132 | 75 | @sp 1 |
e49e529d JM |
76 | @subtitle @code{ld} |
77 | @ifset VERSION_PACKAGE | |
78 | @subtitle @value{VERSION_PACKAGE} | |
79 | @end ifset | |
252b5132 RH |
80 | @subtitle Version @value{VERSION} |
81 | @author Steve Chamberlain | |
82 | @author Ian Lance Taylor | |
252b5132 RH |
83 | @page |
84 | ||
85 | @tex | |
86 | {\parskip=0pt | |
704c465c NC |
87 | \hfill Red Hat Inc\par |
88 | \hfill nickc\@credhat.com, doc\@redhat.com\par | |
71ba23f6 | 89 | \hfill {\it The GNU linker}\par |
252b5132 RH |
90 | \hfill Edited by Jeffrey Osier (jeffrey\@cygnus.com)\par |
91 | } | |
92 | \global\parindent=0pt % Steve likes it this way. | |
93 | @end tex | |
94 | ||
95 | @vskip 0pt plus 1filll | |
0285c67d | 96 | @c man begin COPYRIGHT |
250d07de | 97 | Copyright @copyright{} 1991-2021 Free Software Foundation, Inc. |
252b5132 | 98 | |
0285c67d | 99 | Permission is granted to copy, distribute and/or modify this document |
793c5807 | 100 | under the terms of the GNU Free Documentation License, Version 1.3 |
0285c67d NC |
101 | or any later version published by the Free Software Foundation; |
102 | with no Invariant Sections, with no Front-Cover Texts, and with no | |
103 | Back-Cover Texts. A copy of the license is included in the | |
36f63dca | 104 | section entitled ``GNU Free Documentation License''. |
0285c67d | 105 | @c man end |
252b5132 | 106 | |
252b5132 RH |
107 | @end titlepage |
108 | @end iftex | |
4ecceb71 | 109 | @contents |
252b5132 RH |
110 | @c FIXME: Talk about importance of *order* of args, cmds to linker! |
111 | ||
84ec0e6d | 112 | @ifnottex |
252b5132 | 113 | @node Top |
71ba23f6 | 114 | @top LD |
e49e529d JM |
115 | This file documents the @sc{gnu} linker ld |
116 | @ifset VERSION_PACKAGE | |
117 | @value{VERSION_PACKAGE} | |
118 | @end ifset | |
119 | version @value{VERSION}. | |
252b5132 | 120 | |
cf055d54 | 121 | This document is distributed under the terms of the GNU Free |
793c5807 NC |
122 | Documentation License version 1.3. A copy of the license is included |
123 | in the section entitled ``GNU Free Documentation License''. | |
cf055d54 | 124 | |
252b5132 RH |
125 | @menu |
126 | * Overview:: Overview | |
127 | * Invocation:: Invocation | |
128 | * Scripts:: Linker Scripts | |
bf6d8037 | 129 | * Plugins:: Linker Plugins |
252b5132 RH |
130 | @ifset GENERIC |
131 | * Machine Dependent:: Machine Dependent Features | |
132 | @end ifset | |
133 | @ifclear GENERIC | |
134 | @ifset H8300 | |
135 | * H8/300:: ld and the H8/300 | |
136 | @end ifset | |
c2dcd04e NC |
137 | @ifset Renesas |
138 | * Renesas:: ld and other Renesas micros | |
252b5132 | 139 | @end ifset |
36f63dca NC |
140 | @ifset ARM |
141 | * ARM:: ld and the ARM family | |
142 | @end ifset | |
93fd0973 SC |
143 | @ifset M68HC11 |
144 | * M68HC11/68HC12:: ld and the Motorola 68HC11 and 68HC12 families | |
145 | @end ifset | |
7ca01ed9 NC |
146 | @ifset HPPA |
147 | * HPPA ELF32:: ld and HPPA 32-bit ELF | |
148 | @end ifset | |
7fb9f789 NC |
149 | @ifset M68K |
150 | * M68K:: ld and Motorola 68K family | |
151 | @end ifset | |
833794fc MR |
152 | @ifset MIPS |
153 | * MIPS:: ld and MIPS family | |
154 | @end ifset | |
2a60a7a8 AM |
155 | @ifset POWERPC |
156 | * PowerPC ELF32:: ld and PowerPC 32-bit ELF Support | |
157 | @end ifset | |
158 | @ifset POWERPC64 | |
159 | * PowerPC64 ELF64:: ld and PowerPC64 64-bit ELF Support | |
160 | @end ifset | |
b4cbbe8f AK |
161 | @ifset S/390 |
162 | * S/390 ELF:: ld and S/390 ELF Support | |
163 | @end ifset | |
49fa1e15 AM |
164 | @ifset SPU |
165 | * SPU ELF:: ld and SPU ELF Support | |
166 | @end ifset | |
74459f0e TW |
167 | @ifset TICOFF |
168 | * TI COFF:: ld and the TI COFF | |
169 | @end ifset | |
2ca22b03 NC |
170 | @ifset WIN32 |
171 | * Win32:: ld and WIN32 (cygwin/mingw) | |
172 | @end ifset | |
e0001a05 NC |
173 | @ifset XTENSA |
174 | * Xtensa:: ld and Xtensa Processors | |
175 | @end ifset | |
252b5132 RH |
176 | @end ifclear |
177 | @ifclear SingleFormat | |
178 | * BFD:: BFD | |
179 | @end ifclear | |
180 | @c Following blank line required for remaining bug in makeinfo conds/menus | |
181 | ||
182 | * Reporting Bugs:: Reporting Bugs | |
183 | * MRI:: MRI Compatible Script Files | |
704c465c | 184 | * GNU Free Documentation License:: GNU Free Documentation License |
370b66a1 | 185 | * LD Index:: LD Index |
252b5132 | 186 | @end menu |
84ec0e6d | 187 | @end ifnottex |
252b5132 RH |
188 | |
189 | @node Overview | |
190 | @chapter Overview | |
191 | ||
192 | @cindex @sc{gnu} linker | |
193 | @cindex what is this? | |
0285c67d | 194 | |
0879a67a | 195 | @ifset man |
0285c67d | 196 | @c man begin SYNOPSIS |
ff5dcc92 | 197 | ld [@b{options}] @var{objfile} @dots{} |
0285c67d NC |
198 | @c man end |
199 | ||
200 | @c man begin SEEALSO | |
201 | ar(1), nm(1), objcopy(1), objdump(1), readelf(1) and | |
202 | the Info entries for @file{binutils} and | |
203 | @file{ld}. | |
204 | @c man end | |
205 | @end ifset | |
206 | ||
207 | @c man begin DESCRIPTION | |
208 | ||
ff5dcc92 | 209 | @command{ld} combines a number of object and archive files, relocates |
252b5132 | 210 | their data and ties up symbol references. Usually the last step in |
ff5dcc92 | 211 | compiling a program is to run @command{ld}. |
252b5132 | 212 | |
ff5dcc92 | 213 | @command{ld} accepts Linker Command Language files written in |
252b5132 RH |
214 | a superset of AT&T's Link Editor Command Language syntax, |
215 | to provide explicit and total control over the linking process. | |
216 | ||
0285c67d NC |
217 | @ifset man |
218 | @c For the man only | |
ece2d90e | 219 | This man page does not describe the command language; see the |
71ba23f6 NC |
220 | @command{ld} entry in @code{info} for full details on the command |
221 | language and on other aspects of the GNU linker. | |
0285c67d NC |
222 | @end ifset |
223 | ||
252b5132 | 224 | @ifclear SingleFormat |
ff5dcc92 SC |
225 | This version of @command{ld} uses the general purpose BFD libraries |
226 | to operate on object files. This allows @command{ld} to read, combine, and | |
252b5132 RH |
227 | write object files in many different formats---for example, COFF or |
228 | @code{a.out}. Different formats may be linked together to produce any | |
229 | available kind of object file. @xref{BFD}, for more information. | |
230 | @end ifclear | |
231 | ||
232 | Aside from its flexibility, the @sc{gnu} linker is more helpful than other | |
233 | linkers in providing diagnostic information. Many linkers abandon | |
234 | execution immediately upon encountering an error; whenever possible, | |
ff5dcc92 | 235 | @command{ld} continues executing, allowing you to identify other errors |
252b5132 RH |
236 | (or, in some cases, to get an output file in spite of the error). |
237 | ||
0285c67d NC |
238 | @c man end |
239 | ||
252b5132 RH |
240 | @node Invocation |
241 | @chapter Invocation | |
242 | ||
0285c67d NC |
243 | @c man begin DESCRIPTION |
244 | ||
ff5dcc92 | 245 | The @sc{gnu} linker @command{ld} is meant to cover a broad range of situations, |
252b5132 RH |
246 | and to be as compatible as possible with other linkers. As a result, |
247 | you have many choices to control its behavior. | |
248 | ||
0285c67d NC |
249 | @c man end |
250 | ||
252b5132 RH |
251 | @ifset UsesEnvVars |
252 | @menu | |
a05a5b64 | 253 | * Options:: Command-line Options |
252b5132 RH |
254 | * Environment:: Environment Variables |
255 | @end menu | |
256 | ||
257 | @node Options | |
a05a5b64 | 258 | @section Command-line Options |
252b5132 RH |
259 | @end ifset |
260 | ||
261 | @cindex command line | |
262 | @cindex options | |
0285c67d NC |
263 | |
264 | @c man begin OPTIONS | |
265 | ||
252b5132 RH |
266 | The linker supports a plethora of command-line options, but in actual |
267 | practice few of them are used in any particular context. | |
268 | @cindex standard Unix system | |
ff5dcc92 | 269 | For instance, a frequent use of @command{ld} is to link standard Unix |
252b5132 RH |
270 | object files on a standard, supported Unix system. On such a system, to |
271 | link a file @code{hello.o}: | |
272 | ||
273 | @smallexample | |
274 | ld -o @var{output} /lib/crt0.o hello.o -lc | |
275 | @end smallexample | |
276 | ||
ff5dcc92 | 277 | This tells @command{ld} to produce a file called @var{output} as the |
252b5132 RH |
278 | result of linking the file @code{/lib/crt0.o} with @code{hello.o} and |
279 | the library @code{libc.a}, which will come from the standard search | |
280 | directories. (See the discussion of the @samp{-l} option below.) | |
281 | ||
ff5dcc92 | 282 | Some of the command-line options to @command{ld} may be specified at any |
511ab9e9 ILT |
283 | point in the command line. However, options which refer to files, such |
284 | as @samp{-l} or @samp{-T}, cause the file to be read at the point at | |
285 | which the option appears in the command line, relative to the object | |
286 | files and other file options. Repeating non-file options with a | |
287 | different argument will either have no further effect, or override prior | |
252b5132 RH |
288 | occurrences (those further to the left on the command line) of that |
289 | option. Options which may be meaningfully specified more than once are | |
290 | noted in the descriptions below. | |
291 | ||
292 | @cindex object files | |
511ab9e9 ILT |
293 | Non-option arguments are object files or archives which are to be linked |
294 | together. They may follow, precede, or be mixed in with command-line | |
295 | options, except that an object file argument may not be placed between | |
296 | an option and its argument. | |
252b5132 RH |
297 | |
298 | Usually the linker is invoked with at least one object file, but you can | |
299 | specify other forms of binary input files using @samp{-l}, @samp{-R}, | |
300 | and the script command language. If @emph{no} binary input files at all | |
301 | are specified, the linker does not produce any output, and issues the | |
302 | message @samp{No input files}. | |
303 | ||
36f63dca | 304 | If the linker cannot recognize the format of an object file, it will |
252b5132 RH |
305 | assume that it is a linker script. A script specified in this way |
306 | augments the main linker script used for the link (either the default | |
307 | linker script or the one specified by using @samp{-T}). This feature | |
308 | permits the linker to link against a file which appears to be an object | |
309 | or an archive, but actually merely defines some symbol values, or uses | |
53d25da6 AM |
310 | @code{INPUT} or @code{GROUP} to load other objects. Specifying a |
311 | script in this way merely augments the main linker script, with the | |
312 | extra commands placed after the main script; use the @samp{-T} option | |
313 | to replace the default linker script entirely, but note the effect of | |
314 | the @code{INSERT} command. @xref{Scripts}. | |
252b5132 RH |
315 | |
316 | For options whose names are a single letter, | |
317 | option arguments must either follow the option letter without intervening | |
318 | whitespace, or be given as separate arguments immediately following the | |
319 | option that requires them. | |
320 | ||
321 | For options whose names are multiple letters, either one dash or two can | |
e4897a32 | 322 | precede the option name; for example, @samp{-trace-symbol} and |
36f63dca | 323 | @samp{--trace-symbol} are equivalent. Note---there is one exception to |
e4897a32 | 324 | this rule. Multiple letter options that start with a lower case 'o' can |
ba1be17e | 325 | only be preceded by two dashes. This is to reduce confusion with the |
e4897a32 NC |
326 | @samp{-o} option. So for example @samp{-omagic} sets the output file |
327 | name to @samp{magic} whereas @samp{--omagic} sets the NMAGIC flag on the | |
328 | output. | |
329 | ||
330 | Arguments to multiple-letter options must either be separated from the | |
331 | option name by an equals sign, or be given as separate arguments | |
332 | immediately following the option that requires them. For example, | |
333 | @samp{--trace-symbol foo} and @samp{--trace-symbol=foo} are equivalent. | |
334 | Unique abbreviations of the names of multiple-letter options are | |
335 | accepted. | |
252b5132 | 336 | |
36f63dca | 337 | Note---if the linker is being invoked indirectly, via a compiler driver |
a05a5b64 | 338 | (e.g. @samp{gcc}) then all the linker command-line options should be |
fa19fce0 NC |
339 | prefixed by @samp{-Wl,} (or whatever is appropriate for the particular |
340 | compiler driver) like this: | |
4e53152f NC |
341 | |
342 | @smallexample | |
2509a395 | 343 | gcc -Wl,--start-group foo.o bar.o -Wl,--end-group |
4e53152f NC |
344 | @end smallexample |
345 | ||
346 | This is important, because otherwise the compiler driver program may | |
2509a395 SL |
347 | silently drop the linker options, resulting in a bad link. Confusion |
348 | may also arise when passing options that require values through a | |
349 | driver, as the use of a space between option and argument acts as | |
350 | a separator, and causes the driver to pass only the option to the linker | |
351 | and the argument to the compiler. In this case, it is simplest to use | |
352 | the joined forms of both single- and multiple-letter options, such as: | |
353 | ||
354 | @smallexample | |
355 | gcc foo.o bar.o -Wl,-eENTRY -Wl,-Map=a.map | |
356 | @end smallexample | |
4e53152f | 357 | |
a05a5b64 | 358 | Here is a table of the generic command-line switches accepted by the GNU |
4e53152f NC |
359 | linker: |
360 | ||
ff5dcc92 | 361 | @table @gcctabopt |
38fc1cb1 | 362 | @include at-file.texi |
dff70155 | 363 | |
2509a395 SL |
364 | @kindex -a @var{keyword} |
365 | @item -a @var{keyword} | |
252b5132 RH |
366 | This option is supported for HP/UX compatibility. The @var{keyword} |
367 | argument must be one of the strings @samp{archive}, @samp{shared}, or | |
368 | @samp{default}. @samp{-aarchive} is functionally equivalent to | |
369 | @samp{-Bstatic}, and the other two keywords are functionally equivalent | |
370 | to @samp{-Bdynamic}. This option may be used any number of times. | |
371 | ||
7ee314fa AM |
372 | @kindex --audit @var{AUDITLIB} |
373 | @item --audit @var{AUDITLIB} | |
374 | Adds @var{AUDITLIB} to the @code{DT_AUDIT} entry of the dynamic section. | |
375 | @var{AUDITLIB} is not checked for existence, nor will it use the DT_SONAME | |
376 | specified in the library. If specified multiple times @code{DT_AUDIT} | |
377 | will contain a colon separated list of audit interfaces to use. If the linker | |
378 | finds an object with an audit entry while searching for shared libraries, | |
9d5777a3 | 379 | it will add a corresponding @code{DT_DEPAUDIT} entry in the output file. |
7ee314fa | 380 | This option is only meaningful on ELF platforms supporting the rtld-audit |
9d5777a3 | 381 | interface. |
7ee314fa | 382 | |
252b5132 RH |
383 | @ifclear SingleFormat |
384 | @cindex binary input format | |
385 | @kindex -b @var{format} | |
386 | @kindex --format=@var{format} | |
387 | @cindex input format | |
388 | @cindex input format | |
389 | @item -b @var{input-format} | |
390 | @itemx --format=@var{input-format} | |
ff5dcc92 SC |
391 | @command{ld} may be configured to support more than one kind of object |
392 | file. If your @command{ld} is configured this way, you can use the | |
252b5132 | 393 | @samp{-b} option to specify the binary format for input object files |
ff5dcc92 | 394 | that follow this option on the command line. Even when @command{ld} is |
252b5132 | 395 | configured to support alternative object formats, you don't usually need |
ff5dcc92 | 396 | to specify this, as @command{ld} should be configured to expect as a |
252b5132 RH |
397 | default input format the most usual format on each machine. |
398 | @var{input-format} is a text string, the name of a particular format | |
399 | supported by the BFD libraries. (You can list the available binary | |
400 | formats with @samp{objdump -i}.) | |
401 | @xref{BFD}. | |
402 | ||
403 | You may want to use this option if you are linking files with an unusual | |
404 | binary format. You can also use @samp{-b} to switch formats explicitly (when | |
405 | linking object files of different formats), by including | |
406 | @samp{-b @var{input-format}} before each group of object files in a | |
a1ab1d2a | 407 | particular format. |
252b5132 RH |
408 | |
409 | The default format is taken from the environment variable | |
410 | @code{GNUTARGET}. | |
411 | @ifset UsesEnvVars | |
412 | @xref{Environment}. | |
413 | @end ifset | |
414 | You can also define the input format from a script, using the command | |
0285c67d NC |
415 | @code{TARGET}; |
416 | @ifclear man | |
417 | see @ref{Format Commands}. | |
418 | @end ifclear | |
252b5132 RH |
419 | @end ifclear |
420 | ||
421 | @kindex -c @var{MRI-cmdfile} | |
422 | @kindex --mri-script=@var{MRI-cmdfile} | |
423 | @cindex compatibility, MRI | |
424 | @item -c @var{MRI-commandfile} | |
425 | @itemx --mri-script=@var{MRI-commandfile} | |
ff5dcc92 | 426 | For compatibility with linkers produced by MRI, @command{ld} accepts script |
252b5132 | 427 | files written in an alternate, restricted command language, described in |
0285c67d NC |
428 | @ifclear man |
429 | @ref{MRI,,MRI Compatible Script Files}. | |
430 | @end ifclear | |
431 | @ifset man | |
432 | the MRI Compatible Script Files section of GNU ld documentation. | |
433 | @end ifset | |
434 | Introduce MRI script files with | |
252b5132 | 435 | the option @samp{-c}; use the @samp{-T} option to run linker |
ff5dcc92 SC |
436 | scripts written in the general-purpose @command{ld} scripting language. |
437 | If @var{MRI-cmdfile} does not exist, @command{ld} looks for it in the directories | |
252b5132 RH |
438 | specified by any @samp{-L} options. |
439 | ||
440 | @cindex common allocation | |
441 | @kindex -d | |
442 | @kindex -dc | |
443 | @kindex -dp | |
a1ab1d2a | 444 | @item -d |
252b5132 RH |
445 | @itemx -dc |
446 | @itemx -dp | |
447 | These three options are equivalent; multiple forms are supported for | |
448 | compatibility with other linkers. They assign space to common symbols | |
449 | even if a relocatable output file is specified (with @samp{-r}). The | |
450 | script command @code{FORCE_COMMON_ALLOCATION} has the same effect. | |
451 | @xref{Miscellaneous Commands}. | |
452 | ||
7ee314fa AM |
453 | @kindex --depaudit @var{AUDITLIB} |
454 | @kindex -P @var{AUDITLIB} | |
455 | @item --depaudit @var{AUDITLIB} | |
456 | @itemx -P @var{AUDITLIB} | |
457 | Adds @var{AUDITLIB} to the @code{DT_DEPAUDIT} entry of the dynamic section. | |
458 | @var{AUDITLIB} is not checked for existence, nor will it use the DT_SONAME | |
459 | specified in the library. If specified multiple times @code{DT_DEPAUDIT} | |
460 | will contain a colon separated list of audit interfaces to use. This | |
461 | option is only meaningful on ELF platforms supporting the rtld-audit interface. | |
9d5777a3 | 462 | The -P option is provided for Solaris compatibility. |
7ee314fa | 463 | |
abf874aa CL |
464 | @kindex --enable-non-contiguous-regions |
465 | @item --enable-non-contiguous-regions | |
466 | This option avoids generating an error if an input section does not | |
467 | fit a matching output section. The linker tries to allocate the input | |
468 | section to subseque nt matching output sections, and generates an | |
469 | error only if no output section is large enough. This is useful when | |
470 | several non-contiguous memory regions are available and the input | |
471 | section does not require a particular one. The order in which input | |
472 | sections are evaluated does not change, for instance: | |
473 | ||
474 | @smallexample | |
475 | MEMORY @{ | |
476 | MEM1 (rwx) : ORIGIN : 0x1000, LENGTH = 0x14 | |
477 | MEM2 (rwx) : ORIGIN : 0x1000, LENGTH = 0x40 | |
478 | MEM3 (rwx) : ORIGIN : 0x2000, LENGTH = 0x40 | |
479 | @} | |
480 | SECTIONS @{ | |
481 | mem1 : @{ *(.data.*); @} > MEM1 | |
482 | mem2 : @{ *(.data.*); @} > MEM2 | |
483 | mem3 : @{ *(.data.*); @} > MEM2 | |
484 | @} | |
485 | ||
486 | with input sections: | |
487 | .data.1: size 8 | |
488 | .data.2: size 0x10 | |
489 | .data.3: size 4 | |
490 | ||
491 | results in .data.1 affected to mem1, and .data.2 and .data.3 | |
492 | affected to mem2, even though .data.3 would fit in mem3. | |
493 | @end smallexample | |
494 | ||
495 | This option is incompatible with INSERT statements because it changes | |
496 | the way input sections are mapped to output sections. | |
497 | ||
498 | @kindex --enable-non-contiguous-regions-warnings | |
499 | @item --enable-non-contiguous-regions-warnings | |
500 | This option enables warnings when | |
501 | @code{--enable-non-contiguous-regions} allows possibly unexpected | |
502 | matches in sections mapping, potentially leading to silently | |
503 | discarding a section instead of failing because it does not fit any | |
504 | output region. | |
505 | ||
252b5132 RH |
506 | @cindex entry point, from command line |
507 | @kindex -e @var{entry} | |
508 | @kindex --entry=@var{entry} | |
a1ab1d2a | 509 | @item -e @var{entry} |
252b5132 RH |
510 | @itemx --entry=@var{entry} |
511 | Use @var{entry} as the explicit symbol for beginning execution of your | |
512 | program, rather than the default entry point. If there is no symbol | |
513 | named @var{entry}, the linker will try to parse @var{entry} as a number, | |
514 | and use that as the entry address (the number will be interpreted in | |
515 | base 10; you may use a leading @samp{0x} for base 16, or a leading | |
516 | @samp{0} for base 8). @xref{Entry Point}, for a discussion of defaults | |
517 | and other ways of specifying the entry point. | |
518 | ||
b58f81ae DJ |
519 | @kindex --exclude-libs |
520 | @item --exclude-libs @var{lib},@var{lib},... | |
521 | Specifies a list of archive libraries from which symbols should not be automatically | |
e1c37eb5 | 522 | exported. The library names may be delimited by commas or colons. Specifying |
b58f81ae DJ |
523 | @code{--exclude-libs ALL} excludes symbols in all archive libraries from |
524 | automatic export. This option is available only for the i386 PE targeted | |
525 | port of the linker and for ELF targeted ports. For i386 PE, symbols | |
526 | explicitly listed in a .def file are still exported, regardless of this | |
527 | option. For ELF targeted ports, symbols affected by this option will | |
528 | be treated as hidden. | |
529 | ||
e1c37eb5 DK |
530 | @kindex --exclude-modules-for-implib |
531 | @item --exclude-modules-for-implib @var{module},@var{module},... | |
532 | Specifies a list of object files or archive members, from which symbols | |
533 | should not be automatically exported, but which should be copied wholesale | |
534 | into the import library being generated during the link. The module names | |
535 | may be delimited by commas or colons, and must match exactly the filenames | |
536 | used by @command{ld} to open the files; for archive members, this is simply | |
537 | the member name, but for object files the name listed must include and | |
538 | match precisely any path used to specify the input file on the linker's | |
539 | command-line. This option is available only for the i386 PE targeted port | |
540 | of the linker. Symbols explicitly listed in a .def file are still exported, | |
541 | regardless of this option. | |
542 | ||
252b5132 RH |
543 | @cindex dynamic symbol table |
544 | @kindex -E | |
545 | @kindex --export-dynamic | |
267e2722 | 546 | @kindex --no-export-dynamic |
252b5132 RH |
547 | @item -E |
548 | @itemx --export-dynamic | |
267e2722 CD |
549 | @itemx --no-export-dynamic |
550 | When creating a dynamically linked executable, using the @option{-E} | |
551 | option or the @option{--export-dynamic} option causes the linker to add | |
552 | all symbols to the dynamic symbol table. The dynamic symbol table is the | |
553 | set of symbols which are visible from dynamic objects at run time. | |
554 | ||
555 | If you do not use either of these options (or use the | |
556 | @option{--no-export-dynamic} option to restore the default behavior), the | |
557 | dynamic symbol table will normally contain only those symbols which are | |
558 | referenced by some dynamic object mentioned in the link. | |
252b5132 RH |
559 | |
560 | If you use @code{dlopen} to load a dynamic object which needs to refer | |
561 | back to the symbols defined by the program, rather than some other | |
562 | dynamic object, then you will probably need to use this option when | |
563 | linking the program itself. | |
564 | ||
55255dae | 565 | You can also use the dynamic list to control what symbols should |
cb840a31 | 566 | be added to the dynamic symbol table if the output format supports it. |
55255dae | 567 | See the description of @samp{--dynamic-list}. |
cb840a31 | 568 | |
8b747e1a DK |
569 | Note that this option is specific to ELF targeted ports. PE targets |
570 | support a similar function to export all symbols from a DLL or EXE; see | |
571 | the description of @samp{--export-all-symbols} below. | |
572 | ||
37a141bf FS |
573 | @kindex --export-dynamic-symbol=@var{glob} |
574 | @cindex export dynamic symbol | |
575 | @item --export-dynamic-symbol=@var{glob} | |
576 | When creating a dynamically linked executable, symbols matching | |
577 | @var{glob} will be added to the dynamic symbol table. When creating a | |
578 | shared library, references to symbols matching @var{glob} will not be | |
579 | bound to the definitions within the shared library. This option is a | |
580 | no-op when creating a shared library and @samp{-Bsymbolic} or | |
581 | @samp{--dynamic-list} are not specified. This option is only meaningful | |
582 | on ELF platforms which support shared libraries. | |
583 | ||
584 | @kindex --export-dynamic-symbol-list=@var{file} | |
585 | @cindex export dynamic symbol list | |
586 | @item --export-dynamic-symbol-list=@var{file} | |
587 | Specify a @samp{--export-dynamic-symbol} for each pattern in the file. | |
588 | The format of the file is the same as the version node without | |
589 | scope and node name. See @ref{VERSION} for more information. | |
590 | ||
36f63dca | 591 | @ifclear SingleFormat |
252b5132 RH |
592 | @cindex big-endian objects |
593 | @cindex endianness | |
594 | @kindex -EB | |
595 | @item -EB | |
596 | Link big-endian objects. This affects the default output format. | |
597 | ||
598 | @cindex little-endian objects | |
599 | @kindex -EL | |
600 | @item -EL | |
601 | Link little-endian objects. This affects the default output format. | |
36f63dca | 602 | @end ifclear |
252b5132 | 603 | |
2509a395 SL |
604 | @kindex -f @var{name} |
605 | @kindex --auxiliary=@var{name} | |
606 | @item -f @var{name} | |
607 | @itemx --auxiliary=@var{name} | |
252b5132 RH |
608 | When creating an ELF shared object, set the internal DT_AUXILIARY field |
609 | to the specified name. This tells the dynamic linker that the symbol | |
610 | table of the shared object should be used as an auxiliary filter on the | |
611 | symbol table of the shared object @var{name}. | |
612 | ||
613 | If you later link a program against this filter object, then, when you | |
614 | run the program, the dynamic linker will see the DT_AUXILIARY field. If | |
615 | the dynamic linker resolves any symbols from the filter object, it will | |
616 | first check whether there is a definition in the shared object | |
617 | @var{name}. If there is one, it will be used instead of the definition | |
618 | in the filter object. The shared object @var{name} need not exist. | |
619 | Thus the shared object @var{name} may be used to provide an alternative | |
620 | implementation of certain functions, perhaps for debugging or for | |
a094d01f | 621 | machine-specific performance. |
252b5132 RH |
622 | |
623 | This option may be specified more than once. The DT_AUXILIARY entries | |
624 | will be created in the order in which they appear on the command line. | |
625 | ||
2509a395 SL |
626 | @kindex -F @var{name} |
627 | @kindex --filter=@var{name} | |
252b5132 | 628 | @item -F @var{name} |
2509a395 | 629 | @itemx --filter=@var{name} |
252b5132 RH |
630 | When creating an ELF shared object, set the internal DT_FILTER field to |
631 | the specified name. This tells the dynamic linker that the symbol table | |
632 | of the shared object which is being created should be used as a filter | |
633 | on the symbol table of the shared object @var{name}. | |
634 | ||
635 | If you later link a program against this filter object, then, when you | |
636 | run the program, the dynamic linker will see the DT_FILTER field. The | |
637 | dynamic linker will resolve symbols according to the symbol table of the | |
638 | filter object as usual, but it will actually link to the definitions | |
639 | found in the shared object @var{name}. Thus the filter object can be | |
640 | used to select a subset of the symbols provided by the object | |
641 | @var{name}. | |
642 | ||
ff5dcc92 | 643 | Some older linkers used the @option{-F} option throughout a compilation |
252b5132 | 644 | toolchain for specifying object-file format for both input and output |
36f63dca NC |
645 | object files. |
646 | @ifclear SingleFormat | |
647 | The @sc{gnu} linker uses other mechanisms for this purpose: the | |
ece2d90e | 648 | @option{-b}, @option{--format}, @option{--oformat} options, the |
252b5132 | 649 | @code{TARGET} command in linker scripts, and the @code{GNUTARGET} |
36f63dca NC |
650 | environment variable. |
651 | @end ifclear | |
652 | The @sc{gnu} linker will ignore the @option{-F} option when not | |
653 | creating an ELF shared object. | |
252b5132 | 654 | |
3dbf70a2 | 655 | @cindex finalization function |
2509a395 SL |
656 | @kindex -fini=@var{name} |
657 | @item -fini=@var{name} | |
3dbf70a2 MM |
658 | When creating an ELF executable or shared object, call NAME when the |
659 | executable or shared object is unloaded, by setting DT_FINI to the | |
660 | address of the function. By default, the linker uses @code{_fini} as | |
661 | the function to call. | |
662 | ||
252b5132 RH |
663 | @kindex -g |
664 | @item -g | |
665 | Ignored. Provided for compatibility with other tools. | |
666 | ||
2509a395 SL |
667 | @kindex -G @var{value} |
668 | @kindex --gpsize=@var{value} | |
252b5132 | 669 | @cindex object size |
2509a395 | 670 | @item -G @var{value} |
252b5132 RH |
671 | @itemx --gpsize=@var{value} |
672 | Set the maximum size of objects to be optimized using the GP register to | |
673 | @var{size}. This is only meaningful for object file formats such as | |
e8044f35 | 674 | MIPS ELF that support putting large and small objects into different |
252b5132 RH |
675 | sections. This is ignored for other object file formats. |
676 | ||
677 | @cindex runtime library name | |
2509a395 | 678 | @kindex -h @var{name} |
252b5132 | 679 | @kindex -soname=@var{name} |
2509a395 | 680 | @item -h @var{name} |
252b5132 RH |
681 | @itemx -soname=@var{name} |
682 | When creating an ELF shared object, set the internal DT_SONAME field to | |
683 | the specified name. When an executable is linked with a shared object | |
684 | which has a DT_SONAME field, then when the executable is run the dynamic | |
685 | linker will attempt to load the shared object specified by the DT_SONAME | |
686 | field rather than the using the file name given to the linker. | |
687 | ||
688 | @kindex -i | |
689 | @cindex incremental link | |
690 | @item -i | |
691 | Perform an incremental link (same as option @samp{-r}). | |
692 | ||
3dbf70a2 | 693 | @cindex initialization function |
2509a395 SL |
694 | @kindex -init=@var{name} |
695 | @item -init=@var{name} | |
3dbf70a2 MM |
696 | When creating an ELF executable or shared object, call NAME when the |
697 | executable or shared object is loaded, by setting DT_INIT to the address | |
698 | of the function. By default, the linker uses @code{_init} as the | |
699 | function to call. | |
700 | ||
252b5132 | 701 | @cindex archive files, from cmd line |
2509a395 | 702 | @kindex -l @var{namespec} |
bcb674cf | 703 | @kindex --library=@var{namespec} |
2509a395 | 704 | @item -l @var{namespec} |
bcb674cf RS |
705 | @itemx --library=@var{namespec} |
706 | Add the archive or object file specified by @var{namespec} to the | |
707 | list of files to link. This option may be used any number of times. | |
708 | If @var{namespec} is of the form @file{:@var{filename}}, @command{ld} | |
07d8eb55 | 709 | will search the library path for a file called @var{filename}, otherwise it |
bcb674cf | 710 | will search the library path for a file called @file{lib@var{namespec}.a}. |
252b5132 | 711 | |
ff5dcc92 | 712 | On systems which support shared libraries, @command{ld} may also search for |
bcb674cf RS |
713 | files other than @file{lib@var{namespec}.a}. Specifically, on ELF |
714 | and SunOS systems, @command{ld} will search a directory for a library | |
715 | called @file{lib@var{namespec}.so} before searching for one called | |
716 | @file{lib@var{namespec}.a}. (By convention, a @code{.so} extension | |
717 | indicates a shared library.) Note that this behavior does not apply | |
718 | to @file{:@var{filename}}, which always specifies a file called | |
719 | @var{filename}. | |
252b5132 RH |
720 | |
721 | The linker will search an archive only once, at the location where it is | |
722 | specified on the command line. If the archive defines a symbol which | |
723 | was undefined in some object which appeared before the archive on the | |
724 | command line, the linker will include the appropriate file(s) from the | |
725 | archive. However, an undefined symbol in an object appearing later on | |
726 | the command line will not cause the linker to search the archive again. | |
727 | ||
ff5dcc92 | 728 | See the @option{-(} option for a way to force the linker to search |
252b5132 RH |
729 | archives multiple times. |
730 | ||
731 | You may list the same archive multiple times on the command line. | |
732 | ||
733 | @ifset GENERIC | |
734 | This type of archive searching is standard for Unix linkers. However, | |
ff5dcc92 | 735 | if you are using @command{ld} on AIX, note that it is different from the |
252b5132 RH |
736 | behaviour of the AIX linker. |
737 | @end ifset | |
738 | ||
739 | @cindex search directory, from cmd line | |
2509a395 | 740 | @kindex -L @var{dir} |
252b5132 | 741 | @kindex --library-path=@var{dir} |
2509a395 | 742 | @item -L @var{searchdir} |
252b5132 | 743 | @itemx --library-path=@var{searchdir} |
ff5dcc92 SC |
744 | Add path @var{searchdir} to the list of paths that @command{ld} will search |
745 | for archive libraries and @command{ld} control scripts. You may use this | |
252b5132 RH |
746 | option any number of times. The directories are searched in the order |
747 | in which they are specified on the command line. Directories specified | |
748 | on the command line are searched before the default directories. All | |
ff5dcc92 | 749 | @option{-L} options apply to all @option{-l} options, regardless of the |
7d24f02c KH |
750 | order in which the options appear. @option{-L} options do not affect |
751 | how @command{ld} searches for a linker script unless @option{-T} | |
752 | option is specified. | |
252b5132 | 753 | |
3aa2d05a NC |
754 | If @var{searchdir} begins with @code{=} or @code{$SYSROOT}, then this |
755 | prefix will be replaced by the @dfn{sysroot prefix}, controlled by the | |
756 | @samp{--sysroot} option, or specified when the linker is configured. | |
9c8ebd6a | 757 | |
252b5132 RH |
758 | @ifset UsesEnvVars |
759 | The default set of paths searched (without being specified with | |
ff5dcc92 | 760 | @samp{-L}) depends on which emulation mode @command{ld} is using, and in |
252b5132 RH |
761 | some cases also on how it was configured. @xref{Environment}. |
762 | @end ifset | |
763 | ||
764 | The paths can also be specified in a link script with the | |
765 | @code{SEARCH_DIR} command. Directories specified this way are searched | |
766 | at the point in which the linker script appears in the command line. | |
767 | ||
768 | @cindex emulation | |
769 | @kindex -m @var{emulation} | |
2509a395 | 770 | @item -m @var{emulation} |
252b5132 RH |
771 | Emulate the @var{emulation} linker. You can list the available |
772 | emulations with the @samp{--verbose} or @samp{-V} options. | |
773 | ||
774 | If the @samp{-m} option is not used, the emulation is taken from the | |
775 | @code{LDEMULATION} environment variable, if that is defined. | |
776 | ||
777 | Otherwise, the default emulation depends upon how the linker was | |
778 | configured. | |
779 | ||
780 | @cindex link map | |
781 | @kindex -M | |
782 | @kindex --print-map | |
783 | @item -M | |
784 | @itemx --print-map | |
785 | Print a link map to the standard output. A link map provides | |
786 | information about the link, including the following: | |
787 | ||
788 | @itemize @bullet | |
789 | @item | |
3b83e13a | 790 | Where object files are mapped into memory. |
252b5132 RH |
791 | @item |
792 | How common symbols are allocated. | |
793 | @item | |
794 | All archive members included in the link, with a mention of the symbol | |
795 | which caused the archive member to be brought in. | |
3b83e13a NC |
796 | @item |
797 | The values assigned to symbols. | |
798 | ||
799 | Note - symbols whose values are computed by an expression which | |
800 | involves a reference to a previous value of the same symbol may not | |
801 | have correct result displayed in the link map. This is because the | |
802 | linker discards intermediate results and only retains the final value | |
803 | of an expression. Under such circumstances the linker will display | |
804 | the final value enclosed by square brackets. Thus for example a | |
805 | linker script containing: | |
806 | ||
807 | @smallexample | |
808 | foo = 1 | |
809 | foo = foo * 4 | |
810 | foo = foo + 8 | |
811 | @end smallexample | |
812 | ||
813 | will produce the following output in the link map if the @option{-M} | |
814 | option is used: | |
815 | ||
816 | @smallexample | |
817 | 0x00000001 foo = 0x1 | |
818 | [0x0000000c] foo = (foo * 0x4) | |
819 | [0x0000000c] foo = (foo + 0x8) | |
820 | @end smallexample | |
821 | ||
822 | See @ref{Expressions} for more information about expressions in linker | |
823 | scripts. | |
d2ef37eb | 824 | |
06ddcada AM |
825 | @item |
826 | How GNU properties are merged. | |
d2ef37eb | 827 | |
06ddcada AM |
828 | When the linker merges input .note.gnu.property sections into one output |
829 | .note.gnu.property section, some properties are removed or updated. | |
830 | These actions are reported in the link map. For example: | |
d2ef37eb L |
831 | |
832 | @smallexample | |
833 | Removed property 0xc0000002 to merge foo.o (0x1) and bar.o (not found) | |
834 | @end smallexample | |
835 | ||
06ddcada | 836 | This indicates that property 0xc0000002 is removed from output when |
d2ef37eb L |
837 | merging properties in @file{foo.o}, whose property 0xc0000002 value |
838 | is 0x1, and @file{bar.o}, which doesn't have property 0xc0000002. | |
839 | ||
840 | @smallexample | |
06ddcada | 841 | Updated property 0xc0010001 (0x1) to merge foo.o (0x1) and bar.o (0x1) |
d2ef37eb L |
842 | @end smallexample |
843 | ||
06ddcada | 844 | This indicates that property 0xc0010001 value is updated to 0x1 in output |
d2ef37eb L |
845 | when merging properties in @file{foo.o}, whose 0xc0010001 property value |
846 | is 0x1, and @file{bar.o}, whose 0xc0010001 property value is 0x1. | |
252b5132 RH |
847 | @end itemize |
848 | ||
035801ce FS |
849 | @cindex link map discarded |
850 | @kindex --print-map-discarded | |
851 | @kindex --no-print-map-discarded | |
852 | @item --print-map-discarded | |
853 | @itemx --no-print-map-discarded | |
854 | Print (or do not print) the list of discarded and garbage collected sections | |
855 | in the link map. Enabled by default. | |
856 | ||
252b5132 RH |
857 | @kindex -n |
858 | @cindex read-only text | |
859 | @cindex NMAGIC | |
860 | @kindex --nmagic | |
861 | @item -n | |
862 | @itemx --nmagic | |
90f5d9d9 JZ |
863 | Turn off page alignment of sections, and disable linking against shared |
864 | libraries. If the output format supports Unix style magic numbers, | |
865 | mark the output as @code{NMAGIC}. | |
252b5132 RH |
866 | |
867 | @kindex -N | |
868 | @kindex --omagic | |
869 | @cindex read/write from cmd line | |
870 | @cindex OMAGIC | |
a1ab1d2a | 871 | @item -N |
252b5132 RH |
872 | @itemx --omagic |
873 | Set the text and data sections to be readable and writable. Also, do | |
63fd3b82 NC |
874 | not page-align the data segment, and disable linking against shared |
875 | libraries. If the output format supports Unix style magic numbers, | |
4d8907ac DS |
876 | mark the output as @code{OMAGIC}. Note: Although a writable text section |
877 | is allowed for PE-COFF targets, it does not conform to the format | |
878 | specification published by Microsoft. | |
63fd3b82 NC |
879 | |
880 | @kindex --no-omagic | |
881 | @cindex OMAGIC | |
882 | @item --no-omagic | |
883 | This option negates most of the effects of the @option{-N} option. It | |
884 | sets the text section to be read-only, and forces the data segment to | |
885 | be page-aligned. Note - this option does not enable linking against | |
886 | shared libraries. Use @option{-Bdynamic} for this. | |
252b5132 RH |
887 | |
888 | @kindex -o @var{output} | |
889 | @kindex --output=@var{output} | |
890 | @cindex naming the output file | |
891 | @item -o @var{output} | |
892 | @itemx --output=@var{output} | |
ff5dcc92 | 893 | Use @var{output} as the name for the program produced by @command{ld}; if this |
252b5132 RH |
894 | option is not specified, the name @file{a.out} is used by default. The |
895 | script command @code{OUTPUT} can also specify the output file name. | |
896 | ||
f37b21b4 RM |
897 | @kindex --dependency-file=@var{depfile} |
898 | @cindex dependency file | |
899 | @item --dependency-file=@var{depfile} | |
900 | Write a @dfn{dependency file} to @var{depfile}. This file contains a rule | |
901 | suitable for @code{make} describing the output file and all the input files | |
902 | that were read to produce it. The output is similar to the compiler's | |
903 | output with @samp{-M -MP} (@pxref{Preprocessor Options,, Options | |
904 | Controlling the Preprocessor, gcc.info, Using the GNU Compiler | |
905 | Collection}). Note that there is no option like the compiler's @samp{-MM}, | |
906 | to exclude ``system files'' (which is not a well-specified concept in the | |
907 | linker, unlike ``system headers'' in the compiler). So the output from | |
908 | @samp{--dependency-file} is always specific to the exact state of the | |
909 | installation where it was produced, and should not be copied into | |
910 | distributed makefiles without careful editing. | |
911 | ||
252b5132 RH |
912 | @kindex -O @var{level} |
913 | @cindex generating optimized output | |
914 | @item -O @var{level} | |
ff5dcc92 | 915 | If @var{level} is a numeric values greater than zero @command{ld} optimizes |
252b5132 | 916 | the output. This might take significantly longer and therefore probably |
98c503ac NC |
917 | should only be enabled for the final binary. At the moment this |
918 | option only affects ELF shared library generation. Future releases of | |
919 | the linker may make more use of this option. Also currently there is | |
920 | no difference in the linker's behaviour for different non-zero values | |
921 | of this option. Again this may change with future releases. | |
252b5132 | 922 | |
387dd777 DP |
923 | @kindex -plugin @var{name} |
924 | @item -plugin @var{name} | |
925 | Involve a plugin in the linking process. The @var{name} parameter is | |
926 | the absolute filename of the plugin. Usually this parameter is | |
927 | automatically added by the complier, when using link time | |
928 | optimization, but users can also add their own plugins if they so | |
929 | wish. | |
930 | ||
931 | Note that the location of the compiler originated plugins is different | |
932 | from the place where the @command{ar}, @command{nm} and | |
933 | @command{ranlib} programs search for their plugins. In order for | |
934 | those commands to make use of a compiler based plugin it must first be | |
41f37a6f | 935 | copied into the @file{$@{libdir@}/bfd-plugins} directory. All gcc |
387dd777 | 936 | based linker plugins are backward compatible, so it is sufficient to |
6c19b93b | 937 | just copy in the newest one. |
387dd777 | 938 | |
26278bb8 UD |
939 | @kindex --push-state |
940 | @cindex push state governing input file handling | |
941 | @item --push-state | |
f9a6a8f0 | 942 | The @option{--push-state} allows one to preserve the current state of the |
26278bb8 UD |
943 | flags which govern the input file handling so that they can all be |
944 | restored with one corresponding @option{--pop-state} option. | |
945 | ||
946 | The option which are covered are: @option{-Bdynamic}, @option{-Bstatic}, | |
947 | @option{-dn}, @option{-dy}, @option{-call_shared}, @option{-non_shared}, | |
948 | @option{-static}, @option{-N}, @option{-n}, @option{--whole-archive}, | |
949 | @option{--no-whole-archive}, @option{-r}, @option{-Ur}, | |
950 | @option{--copy-dt-needed-entries}, @option{--no-copy-dt-needed-entries}, | |
951 | @option{--as-needed}, @option{--no-as-needed}, and @option{-a}. | |
952 | ||
953 | One target for this option are specifications for @file{pkg-config}. When | |
954 | used with the @option{--libs} option all possibly needed libraries are | |
955 | listed and then possibly linked with all the time. It is better to return | |
956 | something as follows: | |
957 | ||
958 | @smallexample | |
959 | -Wl,--push-state,--as-needed -libone -libtwo -Wl,--pop-state | |
960 | @end smallexample | |
961 | ||
962 | @kindex --pop-state | |
963 | @cindex pop state governing input file handling | |
67cecaba | 964 | @item --pop-state |
26278bb8 UD |
965 | Undoes the effect of --push-state, restores the previous values of the |
966 | flags governing input file handling. | |
967 | ||
a712da20 NC |
968 | @kindex -q |
969 | @kindex --emit-relocs | |
970 | @cindex retain relocations in final executable | |
971 | @item -q | |
972 | @itemx --emit-relocs | |
ba1be17e | 973 | Leave relocation sections and contents in fully linked executables. |
a712da20 NC |
974 | Post link analysis and optimization tools may need this information in |
975 | order to perform correct modifications of executables. This results | |
976 | in larger executables. | |
977 | ||
dbab7a7b NC |
978 | This option is currently only supported on ELF platforms. |
979 | ||
4f471f39 RS |
980 | @kindex --force-dynamic |
981 | @cindex forcing the creation of dynamic sections | |
982 | @item --force-dynamic | |
983 | Force the output file to have dynamic sections. This option is specific | |
984 | to VxWorks targets. | |
985 | ||
252b5132 RH |
986 | @cindex partial link |
987 | @cindex relocatable output | |
988 | @kindex -r | |
1049f94e | 989 | @kindex --relocatable |
252b5132 | 990 | @item -r |
1049f94e | 991 | @itemx --relocatable |
252b5132 | 992 | Generate relocatable output---i.e., generate an output file that can in |
ff5dcc92 | 993 | turn serve as input to @command{ld}. This is often called @dfn{partial |
252b5132 RH |
994 | linking}. As a side effect, in environments that support standard Unix |
995 | magic numbers, this option also sets the output file's magic number to | |
996 | @code{OMAGIC}. | |
ff5dcc92 | 997 | @c ; see @option{-N}. |
252b5132 RH |
998 | If this option is not specified, an absolute file is produced. When |
999 | linking C++ programs, this option @emph{will not} resolve references to | |
1000 | constructors; to do that, use @samp{-Ur}. | |
1001 | ||
62bf86b4 HPN |
1002 | When an input file does not have the same format as the output file, |
1003 | partial linking is only supported if that input file does not contain any | |
1004 | relocations. Different output formats can have further restrictions; for | |
1005 | example some @code{a.out}-based formats do not support partial linking | |
1006 | with input files in other formats at all. | |
1007 | ||
252b5132 RH |
1008 | This option does the same thing as @samp{-i}. |
1009 | ||
1010 | @kindex -R @var{file} | |
1011 | @kindex --just-symbols=@var{file} | |
1012 | @cindex symbol-only input | |
1013 | @item -R @var{filename} | |
1014 | @itemx --just-symbols=@var{filename} | |
1015 | Read symbol names and their addresses from @var{filename}, but do not | |
1016 | relocate it or include it in the output. This allows your output file | |
1017 | to refer symbolically to absolute locations of memory defined in other | |
1018 | programs. You may use this option more than once. | |
1019 | ||
ff5dcc92 | 1020 | For compatibility with other ELF linkers, if the @option{-R} option is |
252b5132 | 1021 | followed by a directory name, rather than a file name, it is treated as |
ff5dcc92 | 1022 | the @option{-rpath} option. |
252b5132 RH |
1023 | |
1024 | @kindex -s | |
1025 | @kindex --strip-all | |
1026 | @cindex strip all symbols | |
a1ab1d2a | 1027 | @item -s |
252b5132 RH |
1028 | @itemx --strip-all |
1029 | Omit all symbol information from the output file. | |
1030 | ||
1031 | @kindex -S | |
1032 | @kindex --strip-debug | |
1033 | @cindex strip debugger symbols | |
a1ab1d2a | 1034 | @item -S |
252b5132 RH |
1035 | @itemx --strip-debug |
1036 | Omit debugger symbol information (but not all symbols) from the output file. | |
1037 | ||
a70f34c0 NC |
1038 | @kindex --strip-discarded |
1039 | @kindex --no-strip-discarded | |
1040 | @item --strip-discarded | |
1041 | @itemx --no-strip-discarded | |
1042 | Omit (or do not omit) global symbols defined in discarded sections. | |
1043 | Enabled by default. | |
1044 | ||
252b5132 RH |
1045 | @kindex -t |
1046 | @kindex --trace | |
1047 | @cindex input files, displaying | |
a1ab1d2a | 1048 | @item -t |
252b5132 | 1049 | @itemx --trace |
4f4690cd AM |
1050 | Print the names of the input files as @command{ld} processes them. If |
1051 | @samp{-t} is given twice then members within archives are also printed. | |
1052 | @samp{-t} output is useful to generate a list of all the object files | |
1053 | and scripts involved in linking, for example, when packaging files for | |
1054 | a linker bug report. | |
252b5132 RH |
1055 | |
1056 | @kindex -T @var{script} | |
1057 | @kindex --script=@var{script} | |
1058 | @cindex script files | |
1059 | @item -T @var{scriptfile} | |
1060 | @itemx --script=@var{scriptfile} | |
1061 | Use @var{scriptfile} as the linker script. This script replaces | |
ff5dcc92 | 1062 | @command{ld}'s default linker script (rather than adding to it), so |
252b5132 | 1063 | @var{commandfile} must specify everything necessary to describe the |
114283d8 NC |
1064 | output file. @xref{Scripts}. If @var{scriptfile} does not exist in |
1065 | the current directory, @code{ld} looks for it in the directories | |
1066 | specified by any preceding @samp{-L} options. Multiple @samp{-T} | |
1067 | options accumulate. | |
252b5132 | 1068 | |
14be8564 L |
1069 | @kindex -dT @var{script} |
1070 | @kindex --default-script=@var{script} | |
1071 | @cindex script files | |
1072 | @item -dT @var{scriptfile} | |
1073 | @itemx --default-script=@var{scriptfile} | |
1074 | Use @var{scriptfile} as the default linker script. @xref{Scripts}. | |
1075 | ||
1076 | This option is similar to the @option{--script} option except that | |
1077 | processing of the script is delayed until after the rest of the | |
1078 | command line has been processed. This allows options placed after the | |
1079 | @option{--default-script} option on the command line to affect the | |
1080 | behaviour of the linker script, which can be important when the linker | |
1081 | command line cannot be directly controlled by the user. (eg because | |
1082 | the command line is being constructed by another tool, such as | |
1083 | @samp{gcc}). | |
1084 | ||
252b5132 RH |
1085 | @kindex -u @var{symbol} |
1086 | @kindex --undefined=@var{symbol} | |
1087 | @cindex undefined symbol | |
1088 | @item -u @var{symbol} | |
1089 | @itemx --undefined=@var{symbol} | |
1090 | Force @var{symbol} to be entered in the output file as an undefined | |
1091 | symbol. Doing this may, for example, trigger linking of additional | |
1092 | modules from standard libraries. @samp{-u} may be repeated with | |
1093 | different option arguments to enter additional undefined symbols. This | |
1094 | option is equivalent to the @code{EXTERN} linker script command. | |
1095 | ||
0a618243 AB |
1096 | If this option is being used to force additional modules to be pulled |
1097 | into the link, and if it is an error for the symbol to remain | |
1098 | undefined, then the option @option{--require-defined} should be used | |
1099 | instead. | |
1100 | ||
1101 | @kindex --require-defined=@var{symbol} | |
1102 | @cindex symbols, require defined | |
1103 | @cindex defined symbol | |
1104 | @item --require-defined=@var{symbol} | |
1105 | Require that @var{symbol} is defined in the output file. This option | |
1106 | is the same as option @option{--undefined} except that if @var{symbol} | |
1107 | is not defined in the output file then the linker will issue an error | |
1108 | and exit. The same effect can be achieved in a linker script by using | |
1109 | @code{EXTERN}, @code{ASSERT} and @code{DEFINED} together. This option | |
1110 | can be used multiple times to require additional symbols. | |
1111 | ||
252b5132 RH |
1112 | @kindex -Ur |
1113 | @cindex constructors | |
a1ab1d2a | 1114 | @item -Ur |
252b5132 RH |
1115 | For anything other than C++ programs, this option is equivalent to |
1116 | @samp{-r}: it generates relocatable output---i.e., an output file that can in | |
ff5dcc92 | 1117 | turn serve as input to @command{ld}. When linking C++ programs, @samp{-Ur} |
252b5132 RH |
1118 | @emph{does} resolve references to constructors, unlike @samp{-r}. |
1119 | It does not work to use @samp{-Ur} on files that were themselves linked | |
1120 | with @samp{-Ur}; once the constructor table has been built, it cannot | |
1121 | be added to. Use @samp{-Ur} only for the last partial link, and | |
1122 | @samp{-r} for the others. | |
1123 | ||
c005eb9e AB |
1124 | @kindex --orphan-handling=@var{MODE} |
1125 | @cindex orphan sections | |
1126 | @cindex sections, orphan | |
1127 | @item --orphan-handling=@var{MODE} | |
1128 | Control how orphan sections are handled. An orphan section is one not | |
1129 | specifically mentioned in a linker script. @xref{Orphan Sections}. | |
1130 | ||
1131 | @var{MODE} can have any of the following values: | |
1132 | ||
1133 | @table @code | |
1134 | @item place | |
1135 | Orphan sections are placed into a suitable output section following | |
1136 | the strategy described in @ref{Orphan Sections}. The option | |
d2732b69 | 1137 | @samp{--unique} also affects how sections are placed. |
c005eb9e AB |
1138 | |
1139 | @item discard | |
1140 | All orphan sections are discarded, by placing them in the | |
1141 | @samp{/DISCARD/} section (@pxref{Output Section Discarding}). | |
1142 | ||
1143 | @item warn | |
1144 | The linker will place the orphan section as for @code{place} and also | |
1145 | issue a warning. | |
1146 | ||
1147 | @item error | |
1148 | The linker will exit with an error if any orphan section is found. | |
1149 | @end table | |
1150 | ||
1151 | The default if @samp{--orphan-handling} is not given is @code{place}. | |
1152 | ||
577a0623 AM |
1153 | @kindex --unique[=@var{SECTION}] |
1154 | @item --unique[=@var{SECTION}] | |
1155 | Creates a separate output section for every input section matching | |
1156 | @var{SECTION}, or if the optional wildcard @var{SECTION} argument is | |
1157 | missing, for every orphan input section. An orphan section is one not | |
1158 | specifically mentioned in a linker script. You may use this option | |
1159 | multiple times on the command line; It prevents the normal merging of | |
1160 | input sections with the same name, overriding output section assignments | |
1161 | in a linker script. | |
a854a4a7 | 1162 | |
252b5132 RH |
1163 | @kindex -v |
1164 | @kindex -V | |
1165 | @kindex --version | |
1166 | @cindex version | |
1167 | @item -v | |
1168 | @itemx --version | |
1169 | @itemx -V | |
ff5dcc92 | 1170 | Display the version number for @command{ld}. The @option{-V} option also |
252b5132 RH |
1171 | lists the supported emulations. |
1172 | ||
1173 | @kindex -x | |
1174 | @kindex --discard-all | |
1175 | @cindex deleting local symbols | |
1176 | @item -x | |
1177 | @itemx --discard-all | |
1178 | Delete all local symbols. | |
1179 | ||
1180 | @kindex -X | |
1181 | @kindex --discard-locals | |
1182 | @cindex local symbols, deleting | |
a1ab1d2a | 1183 | @item -X |
252b5132 | 1184 | @itemx --discard-locals |
3c68c38f BW |
1185 | Delete all temporary local symbols. (These symbols start with |
1186 | system-specific local label prefixes, typically @samp{.L} for ELF systems | |
1187 | or @samp{L} for traditional a.out systems.) | |
252b5132 RH |
1188 | |
1189 | @kindex -y @var{symbol} | |
1190 | @kindex --trace-symbol=@var{symbol} | |
1191 | @cindex symbol tracing | |
1192 | @item -y @var{symbol} | |
1193 | @itemx --trace-symbol=@var{symbol} | |
1194 | Print the name of each linked file in which @var{symbol} appears. This | |
1195 | option may be given any number of times. On many systems it is necessary | |
1196 | to prepend an underscore. | |
1197 | ||
1198 | This option is useful when you have an undefined symbol in your link but | |
1199 | don't know where the reference is coming from. | |
1200 | ||
1201 | @kindex -Y @var{path} | |
1202 | @item -Y @var{path} | |
1203 | Add @var{path} to the default library search path. This option exists | |
1204 | for Solaris compatibility. | |
1205 | ||
1206 | @kindex -z @var{keyword} | |
1207 | @item -z @var{keyword} | |
cd6d6c15 NC |
1208 | The recognized keywords are: |
1209 | @table @samp | |
1210 | ||
e6e2dfbd AM |
1211 | @item bndplt |
1212 | Always generate BND prefix in PLT entries. Supported for Linux/x86_64. | |
1213 | ||
1214 | @item call-nop=prefix-addr | |
1215 | @itemx call-nop=suffix-nop | |
1216 | @itemx call-nop=prefix-@var{byte} | |
1217 | @itemx call-nop=suffix-@var{byte} | |
1218 | Specify the 1-byte @code{NOP} padding when transforming indirect call | |
1219 | to a locally defined function, foo, via its GOT slot. | |
1220 | @option{call-nop=prefix-addr} generates @code{0x67 call foo}. | |
1221 | @option{call-nop=suffix-nop} generates @code{call foo 0x90}. | |
1222 | @option{call-nop=prefix-@var{byte}} generates @code{@var{byte} call foo}. | |
1223 | @option{call-nop=suffix-@var{byte}} generates @code{call foo @var{byte}}. | |
1224 | Supported for i386 and x86_64. | |
1225 | ||
233a0083 L |
1226 | @item cet-report=none |
1227 | @itemx cet-report=warning | |
1228 | @itemx cet-report=error | |
1229 | Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_IBT and | |
1230 | GNU_PROPERTY_X86_FEATURE_1_SHSTK properties in input .note.gnu.property | |
1231 | section. @option{cet-report=none}, which is the default, will make the | |
1232 | linker not report missing properties in input files. | |
1233 | @option{cet-report=warning} will make the linker issue a warning for | |
1234 | missing properties in input files. @option{cet-report=error} will make | |
1235 | the linker issue an error for missing properties in input files. | |
1236 | Note that @option{ibt} will turn off the missing | |
1237 | GNU_PROPERTY_X86_FEATURE_1_IBT property report and @option{shstk} will | |
1238 | turn off the missing GNU_PROPERTY_X86_FEATURE_1_SHSTK property report. | |
1239 | Supported for Linux/i386 and Linux/x86_64. | |
1240 | ||
cd6d6c15 | 1241 | @item combreloc |
e6e2dfbd AM |
1242 | @itemx nocombreloc |
1243 | Combine multiple dynamic relocation sections and sort to improve | |
1244 | dynamic symbol lookup caching. Do not do this if @samp{nocombreloc}. | |
cd6d6c15 | 1245 | |
b8871f35 | 1246 | @item common |
e6e2dfbd AM |
1247 | @itemx nocommon |
1248 | Generate common symbols with STT_COMMON type during a relocatable | |
1249 | link. Use STT_OBJECT type if @samp{nocommon}. | |
1250 | ||
1251 | @item common-page-size=@var{value} | |
1252 | Set the page size most commonly used to @var{value}. Memory image | |
1253 | layout will be optimized to minimize memory pages if the system is | |
1254 | using pages of this size. | |
b8871f35 | 1255 | |
cd6d6c15 | 1256 | @item defs |
97a232d7 NC |
1257 | Report unresolved symbol references from regular object files. This |
1258 | is done even if the linker is creating a non-symbolic shared library. | |
1259 | This option is the inverse of @samp{-z undefs}. | |
cd6d6c15 | 1260 | |
e6e2dfbd AM |
1261 | @item dynamic-undefined-weak |
1262 | @itemx nodynamic-undefined-weak | |
1263 | Make undefined weak symbols dynamic when building a dynamic object, | |
1264 | if they are referenced from a regular object file and not forced local | |
1265 | by symbol visibility or versioning. Do not make them dynamic if | |
1266 | @samp{nodynamic-undefined-weak}. If neither option is given, a target | |
1267 | may default to either option being in force, or make some other | |
1268 | selection of undefined weak symbols dynamic. Not all targets support | |
1269 | these options. | |
1270 | ||
6aa29e7b JJ |
1271 | @item execstack |
1272 | Marks the object as requiring executable stack. | |
1273 | ||
b039ef04 L |
1274 | @item global |
1275 | This option is only meaningful when building a shared object. It makes | |
1276 | the symbols defined by this shared object available for symbol resolution | |
1277 | of subsequently loaded libraries. | |
1278 | ||
93ab9c0d NC |
1279 | @item globalaudit |
1280 | This option is only meaningful when building a dynamic executable. | |
1281 | This option marks the executable as requiring global auditing by | |
1282 | setting the @code{DF_1_GLOBAUDIT} bit in the @code{DT_FLAGS_1} dynamic | |
1283 | tag. Global auditing requires that any auditing library defined via | |
a05a5b64 | 1284 | the @option{--depaudit} or @option{-P} command-line options be run for |
6c19b93b | 1285 | all dynamic objects loaded by the application. |
93ab9c0d | 1286 | |
e6e2dfbd AM |
1287 | @item ibtplt |
1288 | Generate Intel Indirect Branch Tracking (IBT) enabled PLT entries. | |
1289 | Supported for Linux/i386 and Linux/x86_64. | |
1290 | ||
8df93cb5 | 1291 | @item ibt |
e6e2dfbd AM |
1292 | Generate GNU_PROPERTY_X86_FEATURE_1_IBT in .note.gnu.property section |
1293 | to indicate compatibility with IBT. This also implies @option{ibtplt}. | |
1294 | Supported for Linux/i386 and Linux/x86_64. | |
1295 | ||
cd6d6c15 NC |
1296 | @item initfirst |
1297 | This option is only meaningful when building a shared object. | |
1298 | It marks the object so that its runtime initialization will occur | |
1299 | before the runtime initialization of any other objects brought into | |
1300 | the process at the same time. Similarly the runtime finalization of | |
1301 | the object will occur after the runtime finalization of any other | |
1302 | objects. | |
1303 | ||
1304 | @item interpose | |
e6e2dfbd AM |
1305 | Specify that the dynamic loader should modify its symbol search order |
1306 | so that symbols in this shared library interpose all other shared | |
1307 | libraries not so marked. | |
cd6d6c15 | 1308 | |
c3805e4c VDM |
1309 | @item unique |
1310 | @itemx nounique | |
1311 | When generating a shared library or other dynamically loadable ELF | |
1312 | object mark it as one that should (by default) only ever be loaded once, | |
1313 | and only in the main namespace (when using @code{dlmopen}). This is | |
1314 | primarily used to mark fundamental libraries such as libc, libpthread et | |
1315 | al which do not usually function correctly unless they are the sole instances | |
1316 | of themselves. This behaviour can be overridden by the @code{dlmopen} caller | |
1317 | and does not apply to certain loading mechanisms (such as audit libraries). | |
1318 | ||
279d901e L |
1319 | @item lam-u48 |
1320 | Generate GNU_PROPERTY_X86_FEATURE_1_LAM_U48 in .note.gnu.property section | |
1321 | to indicate compatibility with Intel LAM_U48. Supported for Linux/x86_64. | |
1322 | ||
1323 | @item lam-u57 | |
1324 | Generate GNU_PROPERTY_X86_FEATURE_1_LAM_U57 in .note.gnu.property section | |
1325 | to indicate compatibility with Intel LAM_U57. Supported for Linux/x86_64. | |
1326 | ||
1327 | @item lam-u48-report=none | |
1328 | @itemx lam-u48-report=warning | |
1329 | @itemx lam-u48-report=error | |
1330 | Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U48 | |
1331 | property in input .note.gnu.property section. | |
1332 | @option{lam-u48-report=none}, which is the default, will make the | |
1333 | linker not report missing properties in input files. | |
1334 | @option{lam-u48-report=warning} will make the linker issue a warning for | |
1335 | missing properties in input files. @option{lam-u48-report=error} will | |
1336 | make the linker issue an error for missing properties in input files. | |
1337 | Supported for Linux/x86_64. | |
1338 | ||
1339 | @item lam-u57-report=none | |
1340 | @itemx lam-u57-report=warning | |
1341 | @itemx lam-u57-report=error | |
1342 | Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U57 | |
1343 | property in input .note.gnu.property section. | |
1344 | @option{lam-u57-report=none}, which is the default, will make the | |
1345 | linker not report missing properties in input files. | |
1346 | @option{lam-u57-report=warning} will make the linker issue a warning for | |
1347 | missing properties in input files. @option{lam-u57-report=error} will | |
1348 | make the linker issue an error for missing properties in input files. | |
1349 | Supported for Linux/x86_64. | |
1350 | ||
1351 | @item lam-report=none | |
1352 | @itemx lam-report=warning | |
1353 | @itemx lam-report=error | |
1354 | Specify how to report the missing GNU_PROPERTY_X86_FEATURE_1_LAM_U48 and | |
1355 | GNU_PROPERTY_X86_FEATURE_1_LAM_U57 properties in input .note.gnu.property | |
1356 | section. @option{lam-report=none}, which is the default, will make the | |
1357 | linker not report missing properties in input files. | |
1358 | @option{lam-report=warning} will make the linker issue a warning for | |
1359 | missing properties in input files. @option{lam-report=error} will make | |
1360 | the linker issue an error for missing properties in input files. | |
1361 | Supported for Linux/x86_64. | |
1362 | ||
5fa222e4 AM |
1363 | @item lazy |
1364 | When generating an executable or shared library, mark it to tell the | |
1365 | dynamic linker to defer function call resolution to the point when | |
1366 | the function is called (lazy binding), rather than at load time. | |
1367 | Lazy binding is the default. | |
1368 | ||
cd6d6c15 | 1369 | @item loadfltr |
e6e2dfbd | 1370 | Specify that the object's filters be processed immediately at runtime. |
cd6d6c15 | 1371 | |
e6e2dfbd AM |
1372 | @item max-page-size=@var{value} |
1373 | Set the maximum memory page size supported to @var{value}. | |
cd6d6c15 | 1374 | |
e6e2dfbd AM |
1375 | @item muldefs |
1376 | Allow multiple definitions. | |
b8871f35 | 1377 | |
cd6d6c15 | 1378 | @item nocopyreloc |
daf220f0 AM |
1379 | Disable linker generated .dynbss variables used in place of variables |
1380 | defined in shared libraries. May result in dynamic text relocations. | |
cd6d6c15 NC |
1381 | |
1382 | @item nodefaultlib | |
e6e2dfbd AM |
1383 | Specify that the dynamic loader search for dependencies of this object |
1384 | should ignore any default library search paths. | |
cd6d6c15 NC |
1385 | |
1386 | @item nodelete | |
e6e2dfbd | 1387 | Specify that the object shouldn't be unloaded at runtime. |
cd6d6c15 NC |
1388 | |
1389 | @item nodlopen | |
e6e2dfbd | 1390 | Specify that the object is not available to @code{dlopen}. |
cd6d6c15 NC |
1391 | |
1392 | @item nodump | |
e6e2dfbd | 1393 | Specify that the object can not be dumped by @code{dldump}. |
cd6d6c15 | 1394 | |
6aa29e7b JJ |
1395 | @item noexecstack |
1396 | Marks the object as not requiring executable stack. | |
1397 | ||
e6e2dfbd AM |
1398 | @item noextern-protected-data |
1399 | Don't treat protected data symbols as external when building a shared | |
1400 | library. This option overrides the linker backend default. It can be | |
1401 | used to work around incorrect relocations against protected data symbols | |
1402 | generated by compiler. Updates on protected data symbols by another | |
1403 | module aren't visible to the resulting shared library. Supported for | |
1404 | i386 and x86-64. | |
8dfef1bd | 1405 | |
e6e2dfbd AM |
1406 | @item noreloc-overflow |
1407 | Disable relocation overflow check. This can be used to disable | |
1408 | relocation overflow check if there will be no dynamic relocation | |
1409 | overflow at run-time. Supported for x86_64. | |
6aa29e7b | 1410 | |
cd6d6c15 NC |
1411 | @item now |
1412 | When generating an executable or shared library, mark it to tell the | |
1413 | dynamic linker to resolve all symbols when the program is started, or | |
e6e2dfbd AM |
1414 | when the shared library is loaded by dlopen, instead of deferring |
1415 | function call resolution to the point when the function is first | |
1416 | called. | |
cd6d6c15 NC |
1417 | |
1418 | @item origin | |
e6e2dfbd | 1419 | Specify that the object requires @samp{$ORIGIN} handling in paths. |
cd6d6c15 | 1420 | |
6aa29e7b | 1421 | @item relro |
e6e2dfbd | 1422 | @itemx norelro |
def5c83c AM |
1423 | Create an ELF @code{PT_GNU_RELRO} segment header in the object. This |
1424 | specifies a memory segment that should be made read-only after | |
1425 | relocation, if supported. Specifying @samp{common-page-size} smaller | |
1426 | than the system page size will render this protection ineffective. | |
e6e2dfbd | 1427 | Don't create an ELF @code{PT_GNU_RELRO} segment if @samp{norelro}. |
6aa29e7b | 1428 | |
68b00778 L |
1429 | @item report-relative-reloc |
1430 | Report dynamic relative relocations generated by linker. Supported for | |
1431 | Linux/i386 and Linux/x86_64. | |
1432 | ||
47acac12 L |
1433 | @item separate-code |
1434 | @itemx noseparate-code | |
1435 | Create separate code @code{PT_LOAD} segment header in the object. This | |
1436 | specifies a memory segment that should contain only instructions and must | |
1437 | be in wholly disjoint pages from any other data. Don't create separate | |
1438 | code @code{PT_LOAD} segment if @samp{noseparate-code} is used. | |
1439 | ||
496afd17 L |
1440 | @item unique-symbol |
1441 | @itemx nounique-symbol | |
1442 | Avoid duplicated local symbol names in the symbol string table. Append | |
1443 | ".@code{number}" to duplicated local symbol names if @samp{unique-symbol} | |
1444 | is used. @option{nounique-symbol} is the default. | |
1445 | ||
e6e2dfbd AM |
1446 | @item shstk |
1447 | Generate GNU_PROPERTY_X86_FEATURE_1_SHSTK in .note.gnu.property section | |
1448 | to indicate compatibility with Intel Shadow Stack. Supported for | |
1449 | Linux/i386 and Linux/x86_64. | |
24718e3b | 1450 | |
04c3a755 | 1451 | @item stack-size=@var{value} |
e6e2dfbd | 1452 | Specify a stack size for an ELF @code{PT_GNU_STACK} segment. |
04c3a755 NS |
1453 | Specifying zero will override any default non-zero sized |
1454 | @code{PT_GNU_STACK} segment creation. | |
1455 | ||
8ee10e86 AM |
1456 | @item start-stop-gc |
1457 | @itemx nostart-stop-gc | |
1458 | @cindex start-stop-gc | |
1459 | When @samp{--gc-sections} is in effect, a reference from a retained | |
1460 | section to @code{__start_SECNAME} or @code{__stop_SECNAME} causes all | |
1461 | input sections named @code{SECNAME} to also be retained, if | |
1462 | @code{SECNAME} is representable as a C identifier and either | |
1463 | @code{__start_SECNAME} or @code{__stop_SECNAME} is synthesized by the | |
1464 | linker. @samp{-z start-stop-gc} disables this effect, allowing | |
1465 | sections to be garbage collected as if the special synthesized symbols | |
1466 | were not defined. @samp{-z start-stop-gc} has no effect on a | |
1467 | definition of @code{__start_SECNAME} or @code{__stop_SECNAME} in an | |
1468 | object file or linker script. Such a definition will prevent the | |
1469 | linker providing a synthesized @code{__start_SECNAME} or | |
1470 | @code{__stop_SECNAME} respectively, and therefore the special | |
1471 | treatment by garbage collection for those references. | |
1472 | ||
cae64165 RM |
1473 | @item start-stop-visibility=@var{value} |
1474 | @cindex visibility | |
1475 | @cindex ELF symbol visibility | |
1476 | Specify the ELF symbol visibility for synthesized | |
1477 | @code{__start_SECNAME} and @code{__stop_SECNAME} symbols (@pxref{Input | |
1478 | Section Example}). @var{value} must be exactly @samp{default}, | |
1479 | @samp{internal}, @samp{hidden}, or @samp{protected}. If no @samp{-z | |
1480 | start-stop-visibility} option is given, @samp{protected} is used for | |
1481 | compatibility with historical practice. However, it's highly | |
1482 | recommended to use @samp{-z start-stop-visibility=hidden} in new | |
1483 | programs and shared libraries so that these symbols are not exported | |
1484 | between shared objects, which is not usually what's intended. | |
1485 | ||
e6e2dfbd AM |
1486 | @item text |
1487 | @itemx notext | |
1488 | @itemx textoff | |
a6dbf402 L |
1489 | Report an error if DT_TEXTREL is set, i.e., if the position-independent |
1490 | or shared object has dynamic relocations in read-only sections. Don't | |
1491 | report an error if @samp{notext} or @samp{textoff}. | |
48580982 | 1492 | |
97a232d7 NC |
1493 | @item undefs |
1494 | Do not report unresolved symbol references from regular object files, | |
1495 | either when creating an executable, or when creating a shared library. | |
1496 | This option is the inverse of @samp{-z defs}. | |
1497 | ||
b0ab0693 | 1498 | @item x86-64-baseline |
32930e4e L |
1499 | @item x86-64-v2 |
1500 | @item x86-64-v3 | |
1501 | @itemx x86-64-v4 | |
1502 | Specify the x86-64 ISA level needed in .note.gnu.property section. | |
b0ab0693 | 1503 | @option{x86-64-baseline} generates @code{GNU_PROPERTY_X86_ISA_1_BASELINE}. |
32930e4e L |
1504 | @option{x86-64-v2} generates @code{GNU_PROPERTY_X86_ISA_1_V2}. |
1505 | @option{x86-64-v3} generates @code{GNU_PROPERTY_X86_ISA_1_V3}. | |
1506 | @option{x86-64-v4} generates @code{GNU_PROPERTY_X86_ISA_1_V4}. | |
1507 | Supported for Linux/i386 and Linux/x86_64. | |
1508 | ||
cd6d6c15 NC |
1509 | @end table |
1510 | ||
ece2d90e | 1511 | Other keywords are ignored for Solaris compatibility. |
252b5132 RH |
1512 | |
1513 | @kindex -( | |
1514 | @cindex groups of archives | |
1515 | @item -( @var{archives} -) | |
1516 | @itemx --start-group @var{archives} --end-group | |
1517 | The @var{archives} should be a list of archive files. They may be | |
1518 | either explicit file names, or @samp{-l} options. | |
1519 | ||
1520 | The specified archives are searched repeatedly until no new undefined | |
1521 | references are created. Normally, an archive is searched only once in | |
1522 | the order that it is specified on the command line. If a symbol in that | |
1523 | archive is needed to resolve an undefined symbol referred to by an | |
1524 | object in an archive that appears later on the command line, the linker | |
1525 | would not be able to resolve that reference. By grouping the archives, | |
5d3db3e2 | 1526 | they will all be searched repeatedly until all possible references are |
252b5132 RH |
1527 | resolved. |
1528 | ||
1529 | Using this option has a significant performance cost. It is best to use | |
1530 | it only when there are unavoidable circular references between two or | |
1531 | more archives. | |
1532 | ||
69da35b5 NC |
1533 | @kindex --accept-unknown-input-arch |
1534 | @kindex --no-accept-unknown-input-arch | |
1535 | @item --accept-unknown-input-arch | |
1536 | @itemx --no-accept-unknown-input-arch | |
1537 | Tells the linker to accept input files whose architecture cannot be | |
2ca22b03 | 1538 | recognised. The assumption is that the user knows what they are doing |
69da35b5 NC |
1539 | and deliberately wants to link in these unknown input files. This was |
1540 | the default behaviour of the linker, before release 2.14. The default | |
1541 | behaviour from release 2.14 onwards is to reject such input files, and | |
1542 | so the @samp{--accept-unknown-input-arch} option has been added to | |
1543 | restore the old behaviour. | |
2ca22b03 | 1544 | |
4a43e768 AM |
1545 | @kindex --as-needed |
1546 | @kindex --no-as-needed | |
1547 | @item --as-needed | |
1548 | @itemx --no-as-needed | |
1549 | This option affects ELF DT_NEEDED tags for dynamic libraries mentioned | |
ddbb8a31 | 1550 | on the command line after the @option{--as-needed} option. Normally |
4a43e768 AM |
1551 | the linker will add a DT_NEEDED tag for each dynamic library mentioned |
1552 | on the command line, regardless of whether the library is actually | |
ddbb8a31 | 1553 | needed or not. @option{--as-needed} causes a DT_NEEDED tag to only be |
ffa9430d AM |
1554 | emitted for a library that @emph{at that point in the link} satisfies a |
1555 | non-weak undefined symbol reference from a regular object file or, if | |
1240be6b AM |
1556 | the library is not found in the DT_NEEDED lists of other needed libraries, a |
1557 | non-weak undefined symbol reference from another needed dynamic library. | |
ffa9430d AM |
1558 | Object files or libraries appearing on the command line @emph{after} |
1559 | the library in question do not affect whether the library is seen as | |
1560 | needed. This is similar to the rules for extraction of object files | |
1561 | from archives. @option{--no-as-needed} restores the default behaviour. | |
4a43e768 | 1562 | |
6418520e NC |
1563 | Note: On Linux based systems the @option{--as-needed} option also has |
1564 | an affect on the behaviour of the @option{--rpath} and | |
1565 | @option{--rpath-link} options. See the description of | |
1566 | @option{--rpath-link} for more details. | |
1567 | ||
e56f61be L |
1568 | @kindex --add-needed |
1569 | @kindex --no-add-needed | |
1570 | @item --add-needed | |
1571 | @itemx --no-add-needed | |
ddbb8a31 NC |
1572 | These two options have been deprecated because of the similarity of |
1573 | their names to the @option{--as-needed} and @option{--no-as-needed} | |
1574 | options. They have been replaced by @option{--copy-dt-needed-entries} | |
1575 | and @option{--no-copy-dt-needed-entries}. | |
e56f61be | 1576 | |
252b5132 RH |
1577 | @kindex -assert @var{keyword} |
1578 | @item -assert @var{keyword} | |
1579 | This option is ignored for SunOS compatibility. | |
1580 | ||
1581 | @kindex -Bdynamic | |
1582 | @kindex -dy | |
1583 | @kindex -call_shared | |
1584 | @item -Bdynamic | |
1585 | @itemx -dy | |
1586 | @itemx -call_shared | |
1587 | Link against dynamic libraries. This is only meaningful on platforms | |
1588 | for which shared libraries are supported. This option is normally the | |
1589 | default on such platforms. The different variants of this option are | |
1590 | for compatibility with various systems. You may use this option | |
1591 | multiple times on the command line: it affects library searching for | |
da8bce14 | 1592 | @option{-l} options which follow it. |
252b5132 | 1593 | |
a1ab1d2a UD |
1594 | @kindex -Bgroup |
1595 | @item -Bgroup | |
1596 | Set the @code{DF_1_GROUP} flag in the @code{DT_FLAGS_1} entry in the dynamic | |
1597 | section. This causes the runtime linker to handle lookups in this | |
1598 | object and its dependencies to be performed only inside the group. | |
560e09e9 NC |
1599 | @option{--unresolved-symbols=report-all} is implied. This option is |
1600 | only meaningful on ELF platforms which support shared libraries. | |
a1ab1d2a | 1601 | |
252b5132 RH |
1602 | @kindex -Bstatic |
1603 | @kindex -dn | |
1604 | @kindex -non_shared | |
1605 | @kindex -static | |
a1ab1d2a | 1606 | @item -Bstatic |
252b5132 RH |
1607 | @itemx -dn |
1608 | @itemx -non_shared | |
1609 | @itemx -static | |
1610 | Do not link against shared libraries. This is only meaningful on | |
1611 | platforms for which shared libraries are supported. The different | |
1612 | variants of this option are for compatibility with various systems. You | |
1613 | may use this option multiple times on the command line: it affects | |
560e09e9 | 1614 | library searching for @option{-l} options which follow it. This |
e9156f74 NC |
1615 | option also implies @option{--unresolved-symbols=report-all}. This |
1616 | option can be used with @option{-shared}. Doing so means that a | |
1617 | shared library is being created but that all of the library's external | |
1618 | references must be resolved by pulling in entries from static | |
ece2d90e | 1619 | libraries. |
252b5132 RH |
1620 | |
1621 | @kindex -Bsymbolic | |
1622 | @item -Bsymbolic | |
1623 | When creating a shared library, bind references to global symbols to the | |
1624 | definition within the shared library, if any. Normally, it is possible | |
1625 | for a program linked against a shared library to override the definition | |
1914e264 AM |
1626 | within the shared library. This option is only meaningful on ELF |
1627 | platforms which support shared libraries. | |
252b5132 | 1628 | |
40b36307 L |
1629 | @kindex -Bsymbolic-functions |
1630 | @item -Bsymbolic-functions | |
1631 | When creating a shared library, bind references to global function | |
c0065db7 | 1632 | symbols to the definition within the shared library, if any. |
40b36307 | 1633 | This option is only meaningful on ELF platforms which support shared |
1914e264 | 1634 | libraries. |
40b36307 | 1635 | |
cf893b0e FS |
1636 | @kindex -Bno-symbolic |
1637 | @item -Bno-symbolic | |
1638 | This option can cancel previously specified @samp{-Bsymbolic} and | |
1639 | @samp{-Bsymbolic-functions}. | |
1640 | ||
55255dae L |
1641 | @kindex --dynamic-list=@var{dynamic-list-file} |
1642 | @item --dynamic-list=@var{dynamic-list-file} | |
1643 | Specify the name of a dynamic list file to the linker. This is | |
1644 | typically used when creating shared libraries to specify a list of | |
1645 | global symbols whose references shouldn't be bound to the definition | |
1646 | within the shared library, or creating dynamically linked executables | |
1647 | to specify a list of symbols which should be added to the symbol table | |
1648 | in the executable. This option is only meaningful on ELF platforms | |
1649 | which support shared libraries. | |
1650 | ||
1651 | The format of the dynamic list is the same as the version node without | |
1652 | scope and node name. See @ref{VERSION} for more information. | |
1653 | ||
40b36307 L |
1654 | @kindex --dynamic-list-data |
1655 | @item --dynamic-list-data | |
1656 | Include all global data symbols to the dynamic list. | |
1657 | ||
1658 | @kindex --dynamic-list-cpp-new | |
1659 | @item --dynamic-list-cpp-new | |
1660 | Provide the builtin dynamic list for C++ operator new and delete. It | |
1661 | is mainly useful for building shared libstdc++. | |
1662 | ||
0b8a70d9 L |
1663 | @kindex --dynamic-list-cpp-typeinfo |
1664 | @item --dynamic-list-cpp-typeinfo | |
1665 | Provide the builtin dynamic list for C++ runtime type identification. | |
1666 | ||
252b5132 RH |
1667 | @kindex --check-sections |
1668 | @kindex --no-check-sections | |
1669 | @item --check-sections | |
308b1ffd | 1670 | @itemx --no-check-sections |
252b5132 | 1671 | Asks the linker @emph{not} to check section addresses after they have |
7d816a17 | 1672 | been assigned to see if there are any overlaps. Normally the linker will |
252b5132 RH |
1673 | perform this check, and if it finds any overlaps it will produce |
1674 | suitable error messages. The linker does know about, and does make | |
1675 | allowances for sections in overlays. The default behaviour can be | |
a05a5b64 | 1676 | restored by using the command-line switch @option{--check-sections}. |
02b0b1aa NS |
1677 | Section overlap is not usually checked for relocatable links. You can |
1678 | force checking in that case by using the @option{--check-sections} | |
1679 | option. | |
252b5132 | 1680 | |
ddbb8a31 NC |
1681 | @kindex --copy-dt-needed-entries |
1682 | @kindex --no-copy-dt-needed-entries | |
1683 | @item --copy-dt-needed-entries | |
1684 | @itemx --no-copy-dt-needed-entries | |
9d5777a3 | 1685 | This option affects the treatment of dynamic libraries referred to |
ddbb8a31 | 1686 | by DT_NEEDED tags @emph{inside} ELF dynamic libraries mentioned on the |
08efffb8 | 1687 | command line. Normally the linker won't add a DT_NEEDED tag to the |
ddbb8a31 | 1688 | output binary for each library mentioned in a DT_NEEDED tag in an |
08efffb8 | 1689 | input dynamic library. With @option{--copy-dt-needed-entries} |
ddbb8a31 | 1690 | specified on the command line however any dynamic libraries that |
08efffb8 MM |
1691 | follow it will have their DT_NEEDED entries added. The default |
1692 | behaviour can be restored with @option{--no-copy-dt-needed-entries}. | |
ddbb8a31 NC |
1693 | |
1694 | This option also has an effect on the resolution of symbols in dynamic | |
08efffb8 MM |
1695 | libraries. With @option{--copy-dt-needed-entries} dynamic libraries |
1696 | mentioned on the command line will be recursively searched, following | |
1697 | their DT_NEEDED tags to other libraries, in order to resolve symbols | |
1698 | required by the output binary. With the default setting however | |
1699 | the searching of dynamic libraries that follow it will stop with the | |
1700 | dynamic library itself. No DT_NEEDED links will be traversed to resolve | |
ddbb8a31 NC |
1701 | symbols. |
1702 | ||
252b5132 RH |
1703 | @cindex cross reference table |
1704 | @kindex --cref | |
1705 | @item --cref | |
1706 | Output a cross reference table. If a linker map file is being | |
1707 | generated, the cross reference table is printed to the map file. | |
1708 | Otherwise, it is printed on the standard output. | |
1709 | ||
1710 | The format of the table is intentionally simple, so that it may be | |
1711 | easily processed by a script if necessary. The symbols are printed out, | |
1712 | sorted by name. For each symbol, a list of file names is given. If the | |
1713 | symbol is defined, the first file listed is the location of the | |
049c1c8e NC |
1714 | definition. If the symbol is defined as a common value then any files |
1715 | where this happens appear next. Finally any files that reference the | |
1716 | symbol are listed. | |
252b5132 | 1717 | |
5dba6f05 NA |
1718 | @cindex ctf variables |
1719 | @kindex --ctf-variables | |
1720 | @kindex --no-ctf-variables | |
1721 | @item --ctf-variables | |
1722 | @item --no-ctf-variables | |
1723 | The CTF debuginfo format supports a section which encodes the names and | |
1724 | types of variables found in the program which do not appear in any symbol | |
1725 | table. These variables clearly cannot be looked up by address by | |
1726 | conventional debuggers, so the space used for their types and names is | |
1727 | usually wasted: the types are usually small but the names are often not. | |
1728 | @option{--ctf-variables} causes the generation of such a section. | |
1729 | The default behaviour can be restored with @option{--no-ctf-variables}. | |
1730 | ||
1731 | @cindex ctf type sharing | |
1732 | @kindex --ctf-share-types | |
1733 | @item --ctf-share-types=@var{method} | |
1734 | Adjust the method used to share types between translation units in CTF. | |
1735 | ||
1736 | @table @samp | |
1737 | @item share-unconflicted | |
1738 | Put all types that do not have ambiguous definitions into the shared dictionary, | |
1739 | where debuggers can easily access them, even if they only occur in one | |
1740 | translation unit. This is the default. | |
1741 | ||
1742 | @item share-duplicated | |
1743 | Put only types that occur in multiple translation units into the shared | |
1744 | dictionary: types with only one definition go into per-translation-unit | |
1745 | dictionaries. Types with ambiguous definitions in multiple translation units | |
1746 | always go into per-translation-unit dictionaries. This tends to make the CTF | |
1747 | larger, but may reduce the amount of CTF in the shared dictionary. For very | |
1748 | large projects this may speed up opening the CTF and save memory in the CTF | |
1749 | consumer at runtime. | |
1750 | @end table | |
1751 | ||
4818e05f AM |
1752 | @cindex common allocation |
1753 | @kindex --no-define-common | |
1754 | @item --no-define-common | |
1755 | This option inhibits the assignment of addresses to common symbols. | |
1756 | The script command @code{INHIBIT_COMMON_ALLOCATION} has the same effect. | |
1757 | @xref{Miscellaneous Commands}. | |
1758 | ||
1759 | The @samp{--no-define-common} option allows decoupling | |
1760 | the decision to assign addresses to Common symbols from the choice | |
1761 | of the output file type; otherwise a non-Relocatable output type | |
1762 | forces assigning addresses to Common symbols. | |
1763 | Using @samp{--no-define-common} allows Common symbols that are referenced | |
1764 | from a shared library to be assigned addresses only in the main program. | |
1765 | This eliminates the unused duplicate space in the shared library, | |
1766 | and also prevents any possible confusion over resolving to the wrong | |
1767 | duplicate when there are many dynamic modules with specialized search | |
1768 | paths for runtime symbol resolution. | |
1769 | ||
7bdf4127 AB |
1770 | @cindex group allocation in linker script |
1771 | @cindex section groups | |
1772 | @cindex COMDAT | |
1773 | @kindex --force-group-allocation | |
1774 | @item --force-group-allocation | |
1775 | This option causes the linker to place section group members like | |
1776 | normal input sections, and to delete the section groups. This is the | |
1777 | default behaviour for a final link but this option can be used to | |
1778 | change the behaviour of a relocatable link (@samp{-r}). The script | |
1779 | command @code{FORCE_GROUP_ALLOCATION} has the same | |
1780 | effect. @xref{Miscellaneous Commands}. | |
1781 | ||
252b5132 | 1782 | @cindex symbols, from command line |
2509a395 SL |
1783 | @kindex --defsym=@var{symbol}=@var{exp} |
1784 | @item --defsym=@var{symbol}=@var{expression} | |
252b5132 RH |
1785 | Create a global symbol in the output file, containing the absolute |
1786 | address given by @var{expression}. You may use this option as many | |
1787 | times as necessary to define multiple symbols in the command line. A | |
1788 | limited form of arithmetic is supported for the @var{expression} in this | |
1789 | context: you may give a hexadecimal constant or the name of an existing | |
1790 | symbol, or use @code{+} and @code{-} to add or subtract hexadecimal | |
1791 | constants or symbols. If you need more elaborate expressions, consider | |
66bc8739 AM |
1792 | using the linker command language from a script (@pxref{Assignments}). |
1793 | @emph{Note:} there should be no white space between @var{symbol}, the | |
1794 | equals sign (``@key{=}''), and @var{expression}. | |
252b5132 | 1795 | |
54874444 AB |
1796 | The linker processes @samp{--defsym} arguments and @samp{-T} arguments |
1797 | in order, placing @samp{--defsym} before @samp{-T} will define the | |
1798 | symbol before the linker script from @samp{-T} is processed, while | |
1799 | placing @samp{--defsym} after @samp{-T} will define the symbol after | |
1800 | the linker script has been processed. This difference has | |
1801 | consequences for expressions within the linker script that use the | |
1802 | @samp{--defsym} symbols, which order is correct will depend on what | |
1803 | you are trying to achieve. | |
1804 | ||
252b5132 | 1805 | @cindex demangling, from command line |
28c309a2 | 1806 | @kindex --demangle[=@var{style}] |
252b5132 | 1807 | @kindex --no-demangle |
28c309a2 | 1808 | @item --demangle[=@var{style}] |
252b5132 RH |
1809 | @itemx --no-demangle |
1810 | These options control whether to demangle symbol names in error messages | |
1811 | and other output. When the linker is told to demangle, it tries to | |
1812 | present symbol names in a readable fashion: it strips leading | |
1813 | underscores if they are used by the object file format, and converts C++ | |
a1ab1d2a UD |
1814 | mangled symbol names into user readable names. Different compilers have |
1815 | different mangling styles. The optional demangling style argument can be used | |
1816 | to choose an appropriate demangling style for your compiler. The linker will | |
28c309a2 NC |
1817 | demangle by default unless the environment variable @samp{COLLECT_NO_DEMANGLE} |
1818 | is set. These options may be used to override the default. | |
252b5132 RH |
1819 | |
1820 | @cindex dynamic linker, from command line | |
506eee22 | 1821 | @kindex -I@var{file} |
2509a395 SL |
1822 | @kindex --dynamic-linker=@var{file} |
1823 | @item -I@var{file} | |
1824 | @itemx --dynamic-linker=@var{file} | |
252b5132 RH |
1825 | Set the name of the dynamic linker. This is only meaningful when |
1826 | generating dynamically linked ELF executables. The default dynamic | |
1827 | linker is normally correct; don't use this unless you know what you are | |
1828 | doing. | |
1829 | ||
9b8b325a RF |
1830 | @kindex --no-dynamic-linker |
1831 | @item --no-dynamic-linker | |
1832 | When producing an executable file, omit the request for a dynamic | |
1833 | linker to be used at load-time. This is only meaningful for ELF | |
1834 | executables that contain dynamic relocations, and usually requires | |
1835 | entry point code that is capable of processing these relocations. | |
1836 | ||
a70f34c0 NC |
1837 | @kindex --embedded-relocs |
1838 | @item --embedded-relocs | |
1839 | This option is similar to the @option{--emit-relocs} option except | |
a094d01f | 1840 | that the relocs are stored in a target-specific section. This option |
a70f34c0 NC |
1841 | is only supported by the @samp{BFIN}, @samp{CR16} and @emph{M68K} |
1842 | targets. | |
1843 | ||
3f0a5f17 ME |
1844 | @kindex --disable-multiple-abs-defs |
1845 | @item --disable-multiple-abs-defs | |
1846 | Do not allow multiple definitions with symbols included | |
1847 | in filename invoked by -R or --just-symbols | |
1848 | ||
7ce691ae | 1849 | @kindex --fatal-warnings |
0fe58ccd | 1850 | @kindex --no-fatal-warnings |
7ce691ae | 1851 | @item --fatal-warnings |
0fe58ccd NC |
1852 | @itemx --no-fatal-warnings |
1853 | Treat all warnings as errors. The default behaviour can be restored | |
1854 | with the option @option{--no-fatal-warnings}. | |
7ce691ae | 1855 | |
252b5132 RH |
1856 | @kindex --force-exe-suffix |
1857 | @item --force-exe-suffix | |
1858 | Make sure that an output file has a .exe suffix. | |
1859 | ||
1860 | If a successfully built fully linked output file does not have a | |
1861 | @code{.exe} or @code{.dll} suffix, this option forces the linker to copy | |
1862 | the output file to one of the same name with a @code{.exe} suffix. This | |
1863 | option is useful when using unmodified Unix makefiles on a Microsoft | |
1864 | Windows host, since some versions of Windows won't run an image unless | |
1865 | it ends in a @code{.exe} suffix. | |
1866 | ||
1867 | @kindex --gc-sections | |
1868 | @kindex --no-gc-sections | |
1869 | @cindex garbage collection | |
c17d87de NC |
1870 | @item --gc-sections |
1871 | @itemx --no-gc-sections | |
252b5132 | 1872 | Enable garbage collection of unused input sections. It is ignored on |
ac69cbc6 | 1873 | targets that do not support this option. The default behaviour (of not |
b3549761 | 1874 | performing this garbage collection) can be restored by specifying |
0f088b2a KT |
1875 | @samp{--no-gc-sections} on the command line. Note that garbage |
1876 | collection for COFF and PE format targets is supported, but the | |
1877 | implementation is currently considered to be experimental. | |
252b5132 | 1878 | |
d5465ba2 AM |
1879 | @samp{--gc-sections} decides which input sections are used by |
1880 | examining symbols and relocations. The section containing the entry | |
1881 | symbol and all sections containing symbols undefined on the | |
1882 | command-line will be kept, as will sections containing symbols | |
1883 | referenced by dynamic objects. Note that when building shared | |
1884 | libraries, the linker must assume that any visible symbol is | |
1885 | referenced. Once this initial set of sections has been determined, | |
1886 | the linker recursively marks as used any section referenced by their | |
b69e1ff3 CC |
1887 | relocations. See @samp{--entry}, @samp{--undefined}, and |
1888 | @samp{--gc-keep-exported}. | |
d5465ba2 | 1889 | |
ac69cbc6 | 1890 | This option can be set when doing a partial link (enabled with option |
9d5777a3 | 1891 | @samp{-r}). In this case the root of symbols kept must be explicitly |
b69e1ff3 CC |
1892 | specified either by one of the options @samp{--entry}, |
1893 | @samp{--undefined}, or @samp{--gc-keep-exported} or by a @code{ENTRY} | |
1894 | command in the linker script. | |
ac69cbc6 | 1895 | |
99fabbc9 JL |
1896 | As a GNU extension, ELF input sections marked with the |
1897 | @code{SHF_GNU_RETAIN} flag will not be garbage collected. | |
1898 | ||
c17d87de NC |
1899 | @kindex --print-gc-sections |
1900 | @kindex --no-print-gc-sections | |
1901 | @cindex garbage collection | |
1902 | @item --print-gc-sections | |
1903 | @itemx --no-print-gc-sections | |
1904 | List all sections removed by garbage collection. The listing is | |
1905 | printed on stderr. This option is only effective if garbage | |
1906 | collection has been enabled via the @samp{--gc-sections}) option. The | |
1907 | default behaviour (of not listing the sections that are removed) can | |
1908 | be restored by specifying @samp{--no-print-gc-sections} on the command | |
1909 | line. | |
1910 | ||
22185505 | 1911 | @kindex --gc-keep-exported |
1912 | @cindex garbage collection | |
1913 | @item --gc-keep-exported | |
1914 | When @samp{--gc-sections} is enabled, this option prevents garbage | |
1915 | collection of unused input sections that contain global symbols having | |
1916 | default or protected visibility. This option is intended to be used for | |
1917 | executables where unreferenced sections would otherwise be garbage | |
1918 | collected regardless of the external visibility of contained symbols. | |
1919 | Note that this option has no effect when linking shared objects since | |
1920 | it is already the default behaviour. This option is only supported for | |
1921 | ELF format targets. | |
1922 | ||
30824704 RM |
1923 | @kindex --print-output-format |
1924 | @cindex output format | |
1925 | @item --print-output-format | |
1926 | Print the name of the default output format (perhaps influenced by | |
1927 | other command-line options). This is the string that would appear | |
1928 | in an @code{OUTPUT_FORMAT} linker script command (@pxref{File Commands}). | |
1929 | ||
3604cb1f TG |
1930 | @kindex --print-memory-usage |
1931 | @cindex memory usage | |
1932 | @item --print-memory-usage | |
1933 | Print used size, total size and used size of memory regions created with | |
1934 | the @ref{MEMORY} command. This is useful on embedded targets to have a | |
1935 | quick view of amount of free memory. The format of the output has one | |
1936 | headline and one line per region. It is both human readable and easily | |
1937 | parsable by tools. Here is an example of an output: | |
1938 | ||
1939 | @smallexample | |
1940 | Memory region Used Size Region Size %age Used | |
1941 | ROM: 256 KB 1 MB 25.00% | |
1942 | RAM: 32 B 2 GB 0.00% | |
1943 | @end smallexample | |
1944 | ||
252b5132 RH |
1945 | @cindex help |
1946 | @cindex usage | |
1947 | @kindex --help | |
1948 | @item --help | |
1949 | Print a summary of the command-line options on the standard output and exit. | |
1950 | ||
ea20a7da CC |
1951 | @kindex --target-help |
1952 | @item --target-help | |
a094d01f | 1953 | Print a summary of all target-specific options on the standard output and exit. |
ea20a7da | 1954 | |
2509a395 SL |
1955 | @kindex -Map=@var{mapfile} |
1956 | @item -Map=@var{mapfile} | |
252b5132 | 1957 | Print a link map to the file @var{mapfile}. See the description of the |
2c72361c NC |
1958 | @option{-M} option, above. If @var{mapfile} is just the character |
1959 | @code{-} then the map will be written to stdout. | |
1960 | ||
1961 | Specifying a directory as @var{mapfile} causes the linker map to be | |
1962 | written as a file inside the directory. Normally name of the file | |
1963 | inside the directory is computed as the basename of the @var{output} | |
1964 | file with @code{.map} appended. If however the special character | |
1965 | @code{%} is used then this will be replaced by the full path of the | |
1966 | output file. Additionally if there are any characters after the | |
1967 | @var{%} symbol then @code{.map} will no longer be appended. | |
1968 | ||
1969 | @smallexample | |
1970 | -o foo.exe -Map=bar [Creates ./bar] | |
1971 | -o ../dir/foo.exe -Map=bar [Creates ./bar] | |
1972 | -o foo.exe -Map=../dir [Creates ../dir/foo.exe.map] | |
1973 | -o ../dir2/foo.exe -Map=../dir [Creates ../dir/foo.exe.map] | |
1974 | -o foo.exe -Map=% [Creates ./foo.exe.map] | |
1975 | -o ../dir/foo.exe -Map=% [Creates ../dir/foo.exe.map] | |
1976 | -o foo.exe -Map=%.bar [Creates ./foo.exe.bar] | |
1977 | -o ../dir/foo.exe -Map=%.bar [Creates ../dir/foo.exe.bar] | |
1978 | -o ../dir2/foo.exe -Map=../dir/% [Creates ../dir/../dir2/foo.exe.map] | |
1979 | -o ../dir2/foo.exe -Map=../dir/%.bar [Creates ../dir/../dir2/foo.exe.bar] | |
1980 | @end smallexample | |
1981 | ||
1982 | It is an error to specify more than one @code{%} character. | |
1983 | ||
1984 | If the map file already exists then it will be overwritten by this | |
1985 | operation. | |
252b5132 RH |
1986 | |
1987 | @cindex memory usage | |
1988 | @kindex --no-keep-memory | |
1989 | @item --no-keep-memory | |
ff5dcc92 SC |
1990 | @command{ld} normally optimizes for speed over memory usage by caching the |
1991 | symbol tables of input files in memory. This option tells @command{ld} to | |
252b5132 | 1992 | instead optimize for memory usage, by rereading the symbol tables as |
ff5dcc92 | 1993 | necessary. This may be required if @command{ld} runs out of memory space |
252b5132 RH |
1994 | while linking a large executable. |
1995 | ||
1996 | @kindex --no-undefined | |
a1ab1d2a | 1997 | @kindex -z defs |
97a232d7 | 1998 | @kindex -z undefs |
252b5132 | 1999 | @item --no-undefined |
a1ab1d2a | 2000 | @itemx -z defs |
560e09e9 NC |
2001 | Report unresolved symbol references from regular object files. This |
2002 | is done even if the linker is creating a non-symbolic shared library. | |
2003 | The switch @option{--[no-]allow-shlib-undefined} controls the | |
2004 | behaviour for reporting unresolved references found in shared | |
ece2d90e | 2005 | libraries being linked in. |
252b5132 | 2006 | |
97a232d7 NC |
2007 | The effects of this option can be reverted by using @code{-z undefs}. |
2008 | ||
aa713662 L |
2009 | @kindex --allow-multiple-definition |
2010 | @kindex -z muldefs | |
2011 | @item --allow-multiple-definition | |
2012 | @itemx -z muldefs | |
2013 | Normally when a symbol is defined multiple times, the linker will | |
2014 | report a fatal error. These options allow multiple definitions and the | |
2015 | first definition will be used. | |
2016 | ||
b79e8c78 | 2017 | @kindex --allow-shlib-undefined |
ae9a127f | 2018 | @kindex --no-allow-shlib-undefined |
b79e8c78 | 2019 | @item --allow-shlib-undefined |
ae9a127f | 2020 | @itemx --no-allow-shlib-undefined |
903249d7 | 2021 | Allows or disallows undefined symbols in shared libraries. |
560e09e9 NC |
2022 | This switch is similar to @option{--no-undefined} except that it |
2023 | determines the behaviour when the undefined symbols are in a | |
2024 | shared library rather than a regular object file. It does not affect | |
2025 | how undefined symbols in regular object files are handled. | |
2026 | ||
903249d7 NC |
2027 | The default behaviour is to report errors for any undefined symbols |
2028 | referenced in shared libraries if the linker is being used to create | |
2029 | an executable, but to allow them if the linker is being used to create | |
2030 | a shared library. | |
2031 | ||
2032 | The reasons for allowing undefined symbol references in shared | |
2033 | libraries specified at link time are that: | |
2034 | ||
2035 | @itemize @bullet | |
2036 | @item | |
2037 | A shared library specified at link time may not be the same as the one | |
2038 | that is available at load time, so the symbol might actually be | |
2039 | resolvable at load time. | |
2040 | @item | |
2041 | There are some operating systems, eg BeOS and HPPA, where undefined | |
2042 | symbols in shared libraries are normal. | |
2043 | ||
2044 | The BeOS kernel for example patches shared libraries at load time to | |
2045 | select whichever function is most appropriate for the current | |
2046 | architecture. This is used, for example, to dynamically select an | |
2047 | appropriate memset function. | |
2048 | @end itemize | |
b79e8c78 | 2049 | |
23ae20f5 NC |
2050 | @kindex --error-handling-script=@var{scriptname} |
2051 | @item --error-handling-script=@var{scriptname} | |
2052 | If this option is provided then the linker will invoke | |
2053 | @var{scriptname} whenever an error is encountered. Currently however | |
2054 | only two kinds of error are supported: missing symbols and missing | |
2055 | libraries. Two arguments will be passed to script: the keyword | |
93cf38c0 NC |
2056 | ``undefined-symbol'' or `missing-lib'' and the @var{name} of the |
2057 | undefined symbol or missing library. The intention is that the script | |
2058 | will provide suggestions to the user as to where the symbol or library | |
23ae20f5 NC |
2059 | might be found. After the script has finished then the normal linker |
2060 | error message will be displayed. | |
2061 | ||
2062 | The availability of this option is controlled by a configure time | |
2063 | switch, so it may not be present in specific implementations. | |
2064 | ||
31941635 L |
2065 | @kindex --no-undefined-version |
2066 | @item --no-undefined-version | |
2067 | Normally when a symbol has an undefined version, the linker will ignore | |
2068 | it. This option disallows symbols with undefined version and a fatal error | |
2069 | will be issued instead. | |
2070 | ||
3e3b46e5 PB |
2071 | @kindex --default-symver |
2072 | @item --default-symver | |
2073 | Create and use a default symbol version (the soname) for unversioned | |
fc0e6df6 PB |
2074 | exported symbols. |
2075 | ||
2076 | @kindex --default-imported-symver | |
2077 | @item --default-imported-symver | |
2078 | Create and use a default symbol version (the soname) for unversioned | |
2079 | imported symbols. | |
3e3b46e5 | 2080 | |
252b5132 RH |
2081 | @kindex --no-warn-mismatch |
2082 | @item --no-warn-mismatch | |
ff5dcc92 | 2083 | Normally @command{ld} will give an error if you try to link together input |
252b5132 RH |
2084 | files that are mismatched for some reason, perhaps because they have |
2085 | been compiled for different processors or for different endiannesses. | |
ff5dcc92 | 2086 | This option tells @command{ld} that it should silently permit such possible |
252b5132 RH |
2087 | errors. This option should only be used with care, in cases when you |
2088 | have taken some special action that ensures that the linker errors are | |
2089 | inappropriate. | |
2090 | ||
fe7929ce AM |
2091 | @kindex --no-warn-search-mismatch |
2092 | @item --no-warn-search-mismatch | |
2093 | Normally @command{ld} will give a warning if it finds an incompatible | |
2094 | library during a library search. This option silences the warning. | |
2095 | ||
252b5132 RH |
2096 | @kindex --no-whole-archive |
2097 | @item --no-whole-archive | |
ff5dcc92 | 2098 | Turn off the effect of the @option{--whole-archive} option for subsequent |
252b5132 RH |
2099 | archive files. |
2100 | ||
2101 | @cindex output file after errors | |
2102 | @kindex --noinhibit-exec | |
2103 | @item --noinhibit-exec | |
2104 | Retain the executable output file whenever it is still usable. | |
2105 | Normally, the linker will not produce an output file if it encounters | |
2106 | errors during the link process; it exits without writing an output file | |
2107 | when it issues any error whatsoever. | |
2108 | ||
0a9c1c8e CD |
2109 | @kindex -nostdlib |
2110 | @item -nostdlib | |
2111 | Only search library directories explicitly specified on the | |
2112 | command line. Library directories specified in linker scripts | |
2113 | (including linker scripts specified on the command line) are ignored. | |
2114 | ||
252b5132 | 2115 | @ifclear SingleFormat |
2509a395 SL |
2116 | @kindex --oformat=@var{output-format} |
2117 | @item --oformat=@var{output-format} | |
ff5dcc92 SC |
2118 | @command{ld} may be configured to support more than one kind of object |
2119 | file. If your @command{ld} is configured this way, you can use the | |
252b5132 | 2120 | @samp{--oformat} option to specify the binary format for the output |
ff5dcc92 SC |
2121 | object file. Even when @command{ld} is configured to support alternative |
2122 | object formats, you don't usually need to specify this, as @command{ld} | |
252b5132 RH |
2123 | should be configured to produce as a default output format the most |
2124 | usual format on each machine. @var{output-format} is a text string, the | |
2125 | name of a particular format supported by the BFD libraries. (You can | |
2126 | list the available binary formats with @samp{objdump -i}.) The script | |
2127 | command @code{OUTPUT_FORMAT} can also specify the output format, but | |
2128 | this option overrides it. @xref{BFD}. | |
2129 | @end ifclear | |
2130 | ||
76359541 TP |
2131 | @kindex --out-implib |
2132 | @item --out-implib @var{file} | |
2133 | Create an import library in @var{file} corresponding to the executable | |
2134 | the linker is generating (eg. a DLL or ELF program). This import | |
2135 | library (which should be called @code{*.dll.a} or @code{*.a} for DLLs) | |
2136 | may be used to link clients against the generated executable; this | |
2137 | behaviour makes it possible to skip a separate import library creation | |
2138 | step (eg. @code{dlltool} for DLLs). This option is only available for | |
2139 | the i386 PE and ELF targetted ports of the linker. | |
2140 | ||
36af4a4e JJ |
2141 | @kindex -pie |
2142 | @kindex --pic-executable | |
2143 | @item -pie | |
2144 | @itemx --pic-executable | |
2145 | @cindex position independent executables | |
2146 | Create a position independent executable. This is currently only supported on | |
2147 | ELF platforms. Position independent executables are similar to shared | |
2148 | libraries in that they are relocated by the dynamic linker to the virtual | |
7e7d5768 | 2149 | address the OS chooses for them (which can vary between invocations). Like |
36af4a4e JJ |
2150 | normal dynamically linked executables they can be executed and symbols |
2151 | defined in the executable cannot be overridden by shared libraries. | |
2152 | ||
252b5132 RH |
2153 | @kindex -qmagic |
2154 | @item -qmagic | |
2155 | This option is ignored for Linux compatibility. | |
2156 | ||
2157 | @kindex -Qy | |
2158 | @item -Qy | |
2159 | This option is ignored for SVR4 compatibility. | |
2160 | ||
2161 | @kindex --relax | |
2162 | @cindex synthesizing linker | |
2163 | @cindex relaxing addressing modes | |
28d5f677 | 2164 | @cindex --no-relax |
252b5132 | 2165 | @item --relax |
28d5f677 | 2166 | @itemx --no-relax |
a1ab1d2a | 2167 | An option with machine dependent effects. |
252b5132 RH |
2168 | @ifset GENERIC |
2169 | This option is only supported on a few targets. | |
2170 | @end ifset | |
2171 | @ifset H8300 | |
ff5dcc92 | 2172 | @xref{H8/300,,@command{ld} and the H8/300}. |
252b5132 | 2173 | @end ifset |
e0001a05 NC |
2174 | @ifset XTENSA |
2175 | @xref{Xtensa,, @command{ld} and Xtensa Processors}. | |
2176 | @end ifset | |
93fd0973 SC |
2177 | @ifset M68HC11 |
2178 | @xref{M68HC11/68HC12,,@command{ld} and the 68HC11 and 68HC12}. | |
2179 | @end ifset | |
78058a5e SL |
2180 | @ifset NIOSII |
2181 | @xref{Nios II,,@command{ld} and the Altera Nios II}. | |
2182 | @end ifset | |
2a60a7a8 AM |
2183 | @ifset POWERPC |
2184 | @xref{PowerPC ELF32,,@command{ld} and PowerPC 32-bit ELF Support}. | |
2185 | @end ifset | |
252b5132 | 2186 | |
ccd9fae5 | 2187 | On some platforms the @option{--relax} option performs target specific, |
28d5f677 NC |
2188 | global optimizations that become possible when the linker resolves |
2189 | addressing in the program, such as relaxing address modes, | |
2190 | synthesizing new instructions, selecting shorter version of current | |
11e7fd74 | 2191 | instructions, and combining constant values. |
252b5132 RH |
2192 | |
2193 | On some platforms these link time global optimizations may make symbolic | |
2194 | debugging of the resulting executable impossible. | |
2195 | @ifset GENERIC | |
28d5f677 NC |
2196 | This is known to be the case for the Matsushita MN10200 and MN10300 |
2197 | family of processors. | |
252b5132 RH |
2198 | @end ifset |
2199 | ||
ccd9fae5 NC |
2200 | On platforms where the feature is supported, the option |
2201 | @option{--no-relax} will disable it. | |
28d5f677 | 2202 | |
ccd9fae5 NC |
2203 | On platforms where the feature is not supported, both @option{--relax} |
2204 | and @option{--no-relax} are accepted, but ignored. | |
2205 | ||
252b5132 RH |
2206 | @cindex retaining specified symbols |
2207 | @cindex stripping all but some symbols | |
2208 | @cindex symbols, retaining selectively | |
2509a395 SL |
2209 | @kindex --retain-symbols-file=@var{filename} |
2210 | @item --retain-symbols-file=@var{filename} | |
252b5132 RH |
2211 | Retain @emph{only} the symbols listed in the file @var{filename}, |
2212 | discarding all others. @var{filename} is simply a flat file, with one | |
2213 | symbol name per line. This option is especially useful in environments | |
2214 | @ifset GENERIC | |
2215 | (such as VxWorks) | |
2216 | @end ifset | |
2217 | where a large global symbol table is accumulated gradually, to conserve | |
2218 | run-time memory. | |
2219 | ||
2220 | @samp{--retain-symbols-file} does @emph{not} discard undefined symbols, | |
2221 | or symbols needed for relocations. | |
2222 | ||
2223 | You may only specify @samp{--retain-symbols-file} once in the command | |
2224 | line. It overrides @samp{-s} and @samp{-S}. | |
2225 | ||
2226 | @ifset GENERIC | |
2509a395 | 2227 | @item -rpath=@var{dir} |
252b5132 | 2228 | @cindex runtime library search path |
2509a395 | 2229 | @kindex -rpath=@var{dir} |
252b5132 | 2230 | Add a directory to the runtime library search path. This is used when |
ff5dcc92 | 2231 | linking an ELF executable with shared objects. All @option{-rpath} |
252b5132 | 2232 | arguments are concatenated and passed to the runtime linker, which uses |
c1b00498 MR |
2233 | them to locate shared objects at runtime. |
2234 | ||
2235 | The @option{-rpath} option is also used when locating shared objects which | |
2236 | are needed by shared objects explicitly included in the link; see the | |
2237 | description of the @option{-rpath-link} option. Searching @option{-rpath} | |
2238 | in this way is only supported by native linkers and cross linkers which | |
2239 | have been configured with the @option{--with-sysroot} option. | |
2240 | ||
2241 | If @option{-rpath} is not used when linking an ELF executable, the | |
2242 | contents of the environment variable @code{LD_RUN_PATH} will be used if it | |
2243 | is defined. | |
252b5132 | 2244 | |
ff5dcc92 | 2245 | The @option{-rpath} option may also be used on SunOS. By default, on |
d8506323 | 2246 | SunOS, the linker will form a runtime search path out of all the |
ff5dcc92 SC |
2247 | @option{-L} options it is given. If a @option{-rpath} option is used, the |
2248 | runtime search path will be formed exclusively using the @option{-rpath} | |
2249 | options, ignoring the @option{-L} options. This can be useful when using | |
2250 | gcc, which adds many @option{-L} options which may be on NFS mounted | |
b45619c0 | 2251 | file systems. |
252b5132 | 2252 | |
ff5dcc92 | 2253 | For compatibility with other ELF linkers, if the @option{-R} option is |
252b5132 | 2254 | followed by a directory name, rather than a file name, it is treated as |
ff5dcc92 | 2255 | the @option{-rpath} option. |
252b5132 RH |
2256 | @end ifset |
2257 | ||
2258 | @ifset GENERIC | |
2259 | @cindex link-time runtime library search path | |
2509a395 SL |
2260 | @kindex -rpath-link=@var{dir} |
2261 | @item -rpath-link=@var{dir} | |
252b5132 RH |
2262 | When using ELF or SunOS, one shared library may require another. This |
2263 | happens when an @code{ld -shared} link includes a shared library as one | |
2264 | of the input files. | |
2265 | ||
2266 | When the linker encounters such a dependency when doing a non-shared, | |
2267 | non-relocatable link, it will automatically try to locate the required | |
2268 | shared library and include it in the link, if it is not included | |
ff5dcc92 | 2269 | explicitly. In such a case, the @option{-rpath-link} option |
252b5132 | 2270 | specifies the first set of directories to search. The |
ff5dcc92 | 2271 | @option{-rpath-link} option may specify a sequence of directory names |
252b5132 RH |
2272 | either by specifying a list of names separated by colons, or by |
2273 | appearing multiple times. | |
2274 | ||
e680a6b6 NC |
2275 | The tokens @var{$ORIGIN} and @var{$LIB} can appear in these search |
2276 | directories. They will be replaced by the full path to the directory | |
2277 | containing the program or shared object in the case of @var{$ORIGIN} | |
2278 | and either @samp{lib} - for 32-bit binaries - or @samp{lib64} - for | |
2279 | 64-bit binaries - in the case of @var{$LIB}. | |
2280 | ||
2281 | The alternative form of these tokens - @var{$@{ORIGIN@}} and | |
2282 | @var{$@{LIB@}} can also be used. The token @var{$PLATFORM} is not | |
2283 | supported. | |
2284 | ||
28c309a2 NC |
2285 | This option should be used with caution as it overrides the search path |
2286 | that may have been hard compiled into a shared library. In such a case it | |
2287 | is possible to use unintentionally a different search path than the | |
2288 | runtime linker would do. | |
2289 | ||
252b5132 | 2290 | The linker uses the following search paths to locate required shared |
ece2d90e | 2291 | libraries: |
d8e4137b | 2292 | |
252b5132 RH |
2293 | @enumerate |
2294 | @item | |
ff5dcc92 | 2295 | Any directories specified by @option{-rpath-link} options. |
252b5132 | 2296 | @item |
ff5dcc92 SC |
2297 | Any directories specified by @option{-rpath} options. The difference |
2298 | between @option{-rpath} and @option{-rpath-link} is that directories | |
2299 | specified by @option{-rpath} options are included in the executable and | |
2300 | used at runtime, whereas the @option{-rpath-link} option is only effective | |
ece2d90e NC |
2301 | at link time. Searching @option{-rpath} in this way is only supported |
2302 | by native linkers and cross linkers which have been configured with | |
2303 | the @option{--with-sysroot} option. | |
252b5132 | 2304 | @item |
e2a83dd0 NC |
2305 | On an ELF system, for native linkers, if the @option{-rpath} and |
2306 | @option{-rpath-link} options were not used, search the contents of the | |
2307 | environment variable @code{LD_RUN_PATH}. | |
252b5132 | 2308 | @item |
ff5dcc92 SC |
2309 | On SunOS, if the @option{-rpath} option was not used, search any |
2310 | directories specified using @option{-L} options. | |
252b5132 | 2311 | @item |
a1b8d843 | 2312 | For a native linker, search the contents of the environment |
e2a83dd0 | 2313 | variable @code{LD_LIBRARY_PATH}. |
252b5132 | 2314 | @item |
ec4eb78a L |
2315 | For a native ELF linker, the directories in @code{DT_RUNPATH} or |
2316 | @code{DT_RPATH} of a shared library are searched for shared | |
2317 | libraries needed by it. The @code{DT_RPATH} entries are ignored if | |
2318 | @code{DT_RUNPATH} entries exist. | |
2319 | @item | |
252b5132 RH |
2320 | The default directories, normally @file{/lib} and @file{/usr/lib}. |
2321 | @item | |
d8e4137b NC |
2322 | For a linker for a Linux system, if the file @file{/etc/ld.so.conf} |
2323 | exists, the list of directories found in that file. Note: the path | |
2324 | to this file is prefixed with the @code{sysroot} value, if that is | |
2325 | defined, and then any @code{prefix} string if the linker was | |
2326 | configured with the @command{--prefix=<path>} option. | |
2327 | @item | |
2328 | For a native linker on a FreeBSD system, any directories specified by | |
2329 | the @code{_PATH_ELF_HINTS} macro defined in the @file{elf-hints.h} | |
2330 | header file. | |
2331 | @item | |
2332 | Any directories specifed by a @code{SEARCH_DIR} command in the | |
2333 | linker script being used. | |
252b5132 RH |
2334 | @end enumerate |
2335 | ||
6418520e NC |
2336 | Note however on Linux based systems there is an additional caveat: If |
2337 | the @option{--as-needed} option is active @emph{and} a shared library | |
2338 | is located which would normally satisfy the search @emph{and} this | |
2339 | library does not have DT_NEEDED tag for @file{libc.so} | |
2340 | @emph{and} there is a shared library later on in the set of search | |
2341 | directories which also satisfies the search @emph{and} | |
2342 | this second shared library does have a DT_NEEDED tag for | |
2343 | @file{libc.so} @emph{then} the second library will be selected instead | |
2344 | of the first. | |
2345 | ||
252b5132 RH |
2346 | If the required shared library is not found, the linker will issue a |
2347 | warning and continue with the link. | |
6418520e | 2348 | |
252b5132 RH |
2349 | @end ifset |
2350 | ||
2351 | @kindex -shared | |
2352 | @kindex -Bshareable | |
2353 | @item -shared | |
2354 | @itemx -Bshareable | |
2355 | @cindex shared libraries | |
2356 | Create a shared library. This is currently only supported on ELF, XCOFF | |
2357 | and SunOS platforms. On SunOS, the linker will automatically create a | |
ff5dcc92 | 2358 | shared library if the @option{-e} option is not used and there are |
252b5132 RH |
2359 | undefined symbols in the link. |
2360 | ||
252b5132 | 2361 | @kindex --sort-common |
2509a395 SL |
2362 | @item --sort-common |
2363 | @itemx --sort-common=ascending | |
2364 | @itemx --sort-common=descending | |
de7dd2bd NC |
2365 | This option tells @command{ld} to sort the common symbols by alignment in |
2366 | ascending or descending order when it places them in the appropriate output | |
2367 | sections. The symbol alignments considered are sixteen-byte or larger, | |
2368 | eight-byte, four-byte, two-byte, and one-byte. This is to prevent gaps | |
2369 | between symbols due to alignment constraints. If no sorting order is | |
2370 | specified, then descending order is assumed. | |
252b5132 | 2371 | |
2509a395 SL |
2372 | @kindex --sort-section=name |
2373 | @item --sort-section=name | |
bcaa7b3e L |
2374 | This option will apply @code{SORT_BY_NAME} to all wildcard section |
2375 | patterns in the linker script. | |
2376 | ||
2509a395 SL |
2377 | @kindex --sort-section=alignment |
2378 | @item --sort-section=alignment | |
bcaa7b3e L |
2379 | This option will apply @code{SORT_BY_ALIGNMENT} to all wildcard section |
2380 | patterns in the linker script. | |
2381 | ||
a70f34c0 NC |
2382 | @kindex --spare-dynamic-tags |
2383 | @item --spare-dynamic-tags=@var{count} | |
2384 | This option specifies the number of empty slots to leave in the | |
2385 | .dynamic section of ELF shared objects. Empty slots may be needed by | |
2386 | post processing tools, such as the prelinker. The default is 5. | |
2387 | ||
252b5132 | 2388 | @kindex --split-by-file |
2509a395 | 2389 | @item --split-by-file[=@var{size}] |
ff5dcc92 | 2390 | Similar to @option{--split-by-reloc} but creates a new output section for |
a854a4a7 AM |
2391 | each input file when @var{size} is reached. @var{size} defaults to a |
2392 | size of 1 if not given. | |
252b5132 RH |
2393 | |
2394 | @kindex --split-by-reloc | |
2509a395 | 2395 | @item --split-by-reloc[=@var{count}] |
a854a4a7 | 2396 | Tries to creates extra sections in the output file so that no single |
252b5132 | 2397 | output section in the file contains more than @var{count} relocations. |
a854a4a7 | 2398 | This is useful when generating huge relocatable files for downloading into |
252b5132 RH |
2399 | certain real time kernels with the COFF object file format; since COFF |
2400 | cannot represent more than 65535 relocations in a single section. Note | |
2401 | that this will fail to work with object file formats which do not | |
2402 | support arbitrary sections. The linker will not split up individual | |
2403 | input sections for redistribution, so if a single input section contains | |
2404 | more than @var{count} relocations one output section will contain that | |
a854a4a7 | 2405 | many relocations. @var{count} defaults to a value of 32768. |
252b5132 RH |
2406 | |
2407 | @kindex --stats | |
2408 | @item --stats | |
2409 | Compute and display statistics about the operation of the linker, such | |
2410 | as execution time and memory usage. | |
2411 | ||
2509a395 | 2412 | @kindex --sysroot=@var{directory} |
e2243057 RS |
2413 | @item --sysroot=@var{directory} |
2414 | Use @var{directory} as the location of the sysroot, overriding the | |
2415 | configure-time default. This option is only supported by linkers | |
2416 | that were configured using @option{--with-sysroot}. | |
2417 | ||
a70f34c0 NC |
2418 | @kindex --task-link |
2419 | @item --task-link | |
2420 | This is used by COFF/PE based targets to create a task-linked object | |
2421 | file where all of the global symbols have been converted to statics. | |
2422 | ||
252b5132 RH |
2423 | @kindex --traditional-format |
2424 | @cindex traditional format | |
2425 | @item --traditional-format | |
ff5dcc92 SC |
2426 | For some targets, the output of @command{ld} is different in some ways from |
2427 | the output of some existing linker. This switch requests @command{ld} to | |
252b5132 RH |
2428 | use the traditional format instead. |
2429 | ||
2430 | @cindex dbx | |
ff5dcc92 | 2431 | For example, on SunOS, @command{ld} combines duplicate entries in the |
252b5132 RH |
2432 | symbol string table. This can reduce the size of an output file with |
2433 | full debugging information by over 30 percent. Unfortunately, the SunOS | |
2434 | @code{dbx} program can not read the resulting program (@code{gdb} has no | |
ff5dcc92 | 2435 | trouble). The @samp{--traditional-format} switch tells @command{ld} to not |
252b5132 RH |
2436 | combine duplicate entries. |
2437 | ||
2509a395 SL |
2438 | @kindex --section-start=@var{sectionname}=@var{org} |
2439 | @item --section-start=@var{sectionname}=@var{org} | |
176355da NC |
2440 | Locate a section in the output file at the absolute |
2441 | address given by @var{org}. You may use this option as many | |
2442 | times as necessary to locate multiple sections in the command | |
2443 | line. | |
2444 | @var{org} must be a single hexadecimal integer; | |
2445 | for compatibility with other linkers, you may omit the leading | |
2446 | @samp{0x} usually associated with hexadecimal values. @emph{Note:} there | |
2447 | should be no white space between @var{sectionname}, the equals | |
2448 | sign (``@key{=}''), and @var{org}. | |
2449 | ||
2509a395 SL |
2450 | @kindex -Tbss=@var{org} |
2451 | @kindex -Tdata=@var{org} | |
2452 | @kindex -Ttext=@var{org} | |
252b5132 | 2453 | @cindex segment origins, cmd line |
2509a395 SL |
2454 | @item -Tbss=@var{org} |
2455 | @itemx -Tdata=@var{org} | |
2456 | @itemx -Ttext=@var{org} | |
2457 | Same as @option{--section-start}, with @code{.bss}, @code{.data} or | |
a6e02871 | 2458 | @code{.text} as the @var{sectionname}. |
252b5132 | 2459 | |
2509a395 SL |
2460 | @kindex -Ttext-segment=@var{org} |
2461 | @item -Ttext-segment=@var{org} | |
258795f5 | 2462 | @cindex text segment origin, cmd line |
2b8c06a3 L |
2463 | When creating an ELF executable, it will set the address of the first |
2464 | byte of the text segment. | |
258795f5 | 2465 | |
9d5777a3 RM |
2466 | @kindex -Trodata-segment=@var{org} |
2467 | @item -Trodata-segment=@var{org} | |
2468 | @cindex rodata segment origin, cmd line | |
2469 | When creating an ELF executable or shared object for a target where | |
2470 | the read-only data is in its own segment separate from the executable | |
2471 | text, it will set the address of the first byte of the read-only data segment. | |
2472 | ||
0d705e9f AM |
2473 | @kindex -Tldata-segment=@var{org} |
2474 | @item -Tldata-segment=@var{org} | |
2475 | @cindex ldata segment origin, cmd line | |
2476 | When creating an ELF executable or shared object for x86-64 medium memory | |
2477 | model, it will set the address of the first byte of the ldata segment. | |
2478 | ||
560e09e9 NC |
2479 | @kindex --unresolved-symbols |
2480 | @item --unresolved-symbols=@var{method} | |
2481 | Determine how to handle unresolved symbols. There are four possible | |
2482 | values for @samp{method}: | |
2483 | ||
2484 | @table @samp | |
2485 | @item ignore-all | |
da8bce14 | 2486 | Do not report any unresolved symbols. |
560e09e9 NC |
2487 | |
2488 | @item report-all | |
da8bce14 | 2489 | Report all unresolved symbols. This is the default. |
560e09e9 NC |
2490 | |
2491 | @item ignore-in-object-files | |
2492 | Report unresolved symbols that are contained in shared libraries, but | |
2493 | ignore them if they come from regular object files. | |
2494 | ||
2495 | @item ignore-in-shared-libs | |
2496 | Report unresolved symbols that come from regular object files, but | |
2497 | ignore them if they come from shared libraries. This can be useful | |
2498 | when creating a dynamic binary and it is known that all the shared | |
2499 | libraries that it should be referencing are included on the linker's | |
2500 | command line. | |
2501 | @end table | |
2502 | ||
2503 | The behaviour for shared libraries on their own can also be controlled | |
2504 | by the @option{--[no-]allow-shlib-undefined} option. | |
2505 | ||
2506 | Normally the linker will generate an error message for each reported | |
2507 | unresolved symbol but the option @option{--warn-unresolved-symbols} | |
2508 | can change this to a warning. | |
2509 | ||
1715a13c L |
2510 | @kindex --verbose[=@var{NUMBER}] |
2511 | @cindex verbose[=@var{NUMBER}] | |
252b5132 | 2512 | @item --dll-verbose |
1715a13c | 2513 | @itemx --verbose[=@var{NUMBER}] |
ff5dcc92 | 2514 | Display the version number for @command{ld} and list the linker emulations |
252b5132 | 2515 | supported. Display which input files can and cannot be opened. Display |
1715a13c L |
2516 | the linker script being used by the linker. If the optional @var{NUMBER} |
2517 | argument > 1, plugin symbol status will also be displayed. | |
252b5132 RH |
2518 | |
2519 | @kindex --version-script=@var{version-scriptfile} | |
2520 | @cindex version script, symbol versions | |
2509a395 | 2521 | @item --version-script=@var{version-scriptfile} |
252b5132 RH |
2522 | Specify the name of a version script to the linker. This is typically |
2523 | used when creating shared libraries to specify additional information | |
36f63dca | 2524 | about the version hierarchy for the library being created. This option |
09e2aba4 DK |
2525 | is only fully supported on ELF platforms which support shared libraries; |
2526 | see @ref{VERSION}. It is partially supported on PE platforms, which can | |
2527 | use version scripts to filter symbol visibility in auto-export mode: any | |
2528 | symbols marked @samp{local} in the version script will not be exported. | |
2529 | @xref{WIN32}. | |
252b5132 | 2530 | |
7ce691ae | 2531 | @kindex --warn-common |
252b5132 RH |
2532 | @cindex warnings, on combining symbols |
2533 | @cindex combining symbols, warnings on | |
2534 | @item --warn-common | |
2535 | Warn when a common symbol is combined with another common symbol or with | |
11e7fd74 | 2536 | a symbol definition. Unix linkers allow this somewhat sloppy practice, |
252b5132 RH |
2537 | but linkers on some other operating systems do not. This option allows |
2538 | you to find potential problems from combining global symbols. | |
11e7fd74 | 2539 | Unfortunately, some C libraries use this practice, so you may get some |
252b5132 RH |
2540 | warnings about symbols in the libraries as well as in your programs. |
2541 | ||
2542 | There are three kinds of global symbols, illustrated here by C examples: | |
2543 | ||
2544 | @table @samp | |
2545 | @item int i = 1; | |
2546 | A definition, which goes in the initialized data section of the output | |
2547 | file. | |
2548 | ||
2549 | @item extern int i; | |
2550 | An undefined reference, which does not allocate space. | |
2551 | There must be either a definition or a common symbol for the | |
2552 | variable somewhere. | |
2553 | ||
2554 | @item int i; | |
2555 | A common symbol. If there are only (one or more) common symbols for a | |
2556 | variable, it goes in the uninitialized data area of the output file. | |
2557 | The linker merges multiple common symbols for the same variable into a | |
2558 | single symbol. If they are of different sizes, it picks the largest | |
2559 | size. The linker turns a common symbol into a declaration, if there is | |
2560 | a definition of the same variable. | |
2561 | @end table | |
2562 | ||
2563 | The @samp{--warn-common} option can produce five kinds of warnings. | |
2564 | Each warning consists of a pair of lines: the first describes the symbol | |
2565 | just encountered, and the second describes the previous symbol | |
2566 | encountered with the same name. One or both of the two symbols will be | |
2567 | a common symbol. | |
2568 | ||
2569 | @enumerate | |
2570 | @item | |
2571 | Turning a common symbol into a reference, because there is already a | |
2572 | definition for the symbol. | |
2573 | @smallexample | |
2574 | @var{file}(@var{section}): warning: common of `@var{symbol}' | |
2575 | overridden by definition | |
2576 | @var{file}(@var{section}): warning: defined here | |
2577 | @end smallexample | |
2578 | ||
2579 | @item | |
2580 | Turning a common symbol into a reference, because a later definition for | |
2581 | the symbol is encountered. This is the same as the previous case, | |
2582 | except that the symbols are encountered in a different order. | |
2583 | @smallexample | |
2584 | @var{file}(@var{section}): warning: definition of `@var{symbol}' | |
2585 | overriding common | |
2586 | @var{file}(@var{section}): warning: common is here | |
2587 | @end smallexample | |
2588 | ||
2589 | @item | |
2590 | Merging a common symbol with a previous same-sized common symbol. | |
2591 | @smallexample | |
2592 | @var{file}(@var{section}): warning: multiple common | |
2593 | of `@var{symbol}' | |
2594 | @var{file}(@var{section}): warning: previous common is here | |
2595 | @end smallexample | |
2596 | ||
2597 | @item | |
2598 | Merging a common symbol with a previous larger common symbol. | |
2599 | @smallexample | |
2600 | @var{file}(@var{section}): warning: common of `@var{symbol}' | |
2601 | overridden by larger common | |
2602 | @var{file}(@var{section}): warning: larger common is here | |
2603 | @end smallexample | |
2604 | ||
2605 | @item | |
2606 | Merging a common symbol with a previous smaller common symbol. This is | |
2607 | the same as the previous case, except that the symbols are | |
2608 | encountered in a different order. | |
2609 | @smallexample | |
2610 | @var{file}(@var{section}): warning: common of `@var{symbol}' | |
2611 | overriding smaller common | |
2612 | @var{file}(@var{section}): warning: smaller common is here | |
2613 | @end smallexample | |
2614 | @end enumerate | |
2615 | ||
2616 | @kindex --warn-constructors | |
2617 | @item --warn-constructors | |
2618 | Warn if any global constructors are used. This is only useful for a few | |
2619 | object file formats. For formats like COFF or ELF, the linker can not | |
2620 | detect the use of global constructors. | |
2621 | ||
2622 | @kindex --warn-multiple-gp | |
2623 | @item --warn-multiple-gp | |
2624 | Warn if multiple global pointer values are required in the output file. | |
2625 | This is only meaningful for certain processors, such as the Alpha. | |
2626 | Specifically, some processors put large-valued constants in a special | |
2627 | section. A special register (the global pointer) points into the middle | |
2628 | of this section, so that constants can be loaded efficiently via a | |
2629 | base-register relative addressing mode. Since the offset in | |
2630 | base-register relative mode is fixed and relatively small (e.g., 16 | |
2631 | bits), this limits the maximum size of the constant pool. Thus, in | |
2632 | large programs, it is often necessary to use multiple global pointer | |
2633 | values in order to be able to address all possible constants. This | |
2634 | option causes a warning to be issued whenever this case occurs. | |
2635 | ||
2636 | @kindex --warn-once | |
2637 | @cindex warnings, on undefined symbols | |
2638 | @cindex undefined symbols, warnings on | |
2639 | @item --warn-once | |
2640 | Only warn once for each undefined symbol, rather than once per module | |
2641 | which refers to it. | |
2642 | ||
2643 | @kindex --warn-section-align | |
2644 | @cindex warnings, on section alignment | |
2645 | @cindex section alignment, warnings on | |
2646 | @item --warn-section-align | |
2647 | Warn if the address of an output section is changed because of | |
2648 | alignment. Typically, the alignment will be set by an input section. | |
2649 | The address will only be changed if it not explicitly specified; that | |
2650 | is, if the @code{SECTIONS} command does not specify a start address for | |
2651 | the section (@pxref{SECTIONS}). | |
2652 | ||
a6dbf402 L |
2653 | @kindex --warn-textrel |
2654 | @item --warn-textrel | |
2655 | Warn if the linker adds DT_TEXTREL to a position-independent executable | |
2656 | or shared object. | |
8fdd7217 | 2657 | |
a0c402a5 L |
2658 | @kindex --warn-alternate-em |
2659 | @item --warn-alternate-em | |
2660 | Warn if an object has alternate ELF machine code. | |
2661 | ||
560e09e9 NC |
2662 | @kindex --warn-unresolved-symbols |
2663 | @item --warn-unresolved-symbols | |
2664 | If the linker is going to report an unresolved symbol (see the option | |
2665 | @option{--unresolved-symbols}) it will normally generate an error. | |
2666 | This option makes it generate a warning instead. | |
2667 | ||
2668 | @kindex --error-unresolved-symbols | |
2669 | @item --error-unresolved-symbols | |
2670 | This restores the linker's default behaviour of generating errors when | |
2671 | it is reporting unresolved symbols. | |
2672 | ||
252b5132 RH |
2673 | @kindex --whole-archive |
2674 | @cindex including an entire archive | |
2675 | @item --whole-archive | |
2676 | For each archive mentioned on the command line after the | |
ff5dcc92 | 2677 | @option{--whole-archive} option, include every object file in the archive |
252b5132 RH |
2678 | in the link, rather than searching the archive for the required object |
2679 | files. This is normally used to turn an archive file into a shared | |
2680 | library, forcing every object to be included in the resulting shared | |
2681 | library. This option may be used more than once. | |
2682 | ||
7ec229ce | 2683 | Two notes when using this option from gcc: First, gcc doesn't know |
ff5dcc92 SC |
2684 | about this option, so you have to use @option{-Wl,-whole-archive}. |
2685 | Second, don't forget to use @option{-Wl,-no-whole-archive} after your | |
7ec229ce DD |
2686 | list of archives, because gcc will add its own list of archives to |
2687 | your link and you may not want this flag to affect those as well. | |
2688 | ||
2509a395 SL |
2689 | @kindex --wrap=@var{symbol} |
2690 | @item --wrap=@var{symbol} | |
252b5132 RH |
2691 | Use a wrapper function for @var{symbol}. Any undefined reference to |
2692 | @var{symbol} will be resolved to @code{__wrap_@var{symbol}}. Any | |
2693 | undefined reference to @code{__real_@var{symbol}} will be resolved to | |
2694 | @var{symbol}. | |
2695 | ||
2696 | This can be used to provide a wrapper for a system function. The | |
2697 | wrapper function should be called @code{__wrap_@var{symbol}}. If it | |
2698 | wishes to call the system function, it should call | |
2699 | @code{__real_@var{symbol}}. | |
2700 | ||
2701 | Here is a trivial example: | |
2702 | ||
2703 | @smallexample | |
2704 | void * | |
cc2f008e | 2705 | __wrap_malloc (size_t c) |
252b5132 | 2706 | @{ |
cc2f008e | 2707 | printf ("malloc called with %zu\n", c); |
252b5132 RH |
2708 | return __real_malloc (c); |
2709 | @} | |
2710 | @end smallexample | |
2711 | ||
ff5dcc92 | 2712 | If you link other code with this file using @option{--wrap malloc}, then |
252b5132 RH |
2713 | all calls to @code{malloc} will call the function @code{__wrap_malloc} |
2714 | instead. The call to @code{__real_malloc} in @code{__wrap_malloc} will | |
2715 | call the real @code{malloc} function. | |
2716 | ||
2717 | You may wish to provide a @code{__real_malloc} function as well, so that | |
ff5dcc92 | 2718 | links without the @option{--wrap} option will succeed. If you do this, |
252b5132 RH |
2719 | you should not put the definition of @code{__real_malloc} in the same |
2720 | file as @code{__wrap_malloc}; if you do, the assembler may resolve the | |
2721 | call before the linker has a chance to wrap it to @code{malloc}. | |
2722 | ||
4ea904ed SH |
2723 | Only undefined references are replaced by the linker. So, translation unit |
2724 | internal references to @var{symbol} are not resolved to | |
2725 | @code{__wrap_@var{symbol}}. In the next example, the call to @code{f} in | |
2726 | @code{g} is not resolved to @code{__wrap_f}. | |
2727 | ||
2728 | @smallexample | |
2729 | int | |
2730 | f (void) | |
2731 | @{ | |
2732 | return 123; | |
2733 | @} | |
2734 | ||
2735 | int | |
2736 | g (void) | |
2737 | @{ | |
2738 | return f(); | |
2739 | @} | |
2740 | @end smallexample | |
2741 | ||
6aa29e7b | 2742 | @kindex --eh-frame-hdr |
29063f8b | 2743 | @kindex --no-eh-frame-hdr |
6aa29e7b | 2744 | @item --eh-frame-hdr |
29063f8b NC |
2745 | @itemx --no-eh-frame-hdr |
2746 | Request (@option{--eh-frame-hdr}) or suppress | |
2747 | (@option{--no-eh-frame-hdr}) the creation of @code{.eh_frame_hdr} | |
2748 | section and ELF @code{PT_GNU_EH_FRAME} segment header. | |
6aa29e7b | 2749 | |
e41b3a13 JJ |
2750 | @kindex --ld-generated-unwind-info |
2751 | @item --no-ld-generated-unwind-info | |
2752 | Request creation of @code{.eh_frame} unwind info for linker | |
2753 | generated code sections like PLT. This option is on by default | |
2754 | if linker generated unwind info is supported. | |
2755 | ||
6c1439be L |
2756 | @kindex --enable-new-dtags |
2757 | @kindex --disable-new-dtags | |
2758 | @item --enable-new-dtags | |
2759 | @itemx --disable-new-dtags | |
2760 | This linker can create the new dynamic tags in ELF. But the older ELF | |
2761 | systems may not understand them. If you specify | |
b1b00fcc MF |
2762 | @option{--enable-new-dtags}, the new dynamic tags will be created as needed |
2763 | and older dynamic tags will be omitted. | |
ff5dcc92 | 2764 | If you specify @option{--disable-new-dtags}, no new dynamic tags will be |
6c1439be L |
2765 | created. By default, the new dynamic tags are not created. Note that |
2766 | those options are only available for ELF systems. | |
2767 | ||
2d643429 | 2768 | @kindex --hash-size=@var{number} |
e185dd51 | 2769 | @item --hash-size=@var{number} |
2d643429 NC |
2770 | Set the default size of the linker's hash tables to a prime number |
2771 | close to @var{number}. Increasing this value can reduce the length of | |
2772 | time it takes the linker to perform its tasks, at the expense of | |
2773 | increasing the linker's memory requirements. Similarly reducing this | |
2774 | value can reduce the memory requirements at the expense of speed. | |
2775 | ||
fdc90cb4 JJ |
2776 | @kindex --hash-style=@var{style} |
2777 | @item --hash-style=@var{style} | |
2778 | Set the type of linker's hash table(s). @var{style} can be either | |
2779 | @code{sysv} for classic ELF @code{.hash} section, @code{gnu} for | |
2780 | new style GNU @code{.gnu.hash} section or @code{both} for both | |
2781 | the classic ELF @code{.hash} and new style GNU @code{.gnu.hash} | |
c8455dc9 NC |
2782 | hash tables. The default depends upon how the linker was configured, |
2783 | but for most Linux based systems it will be @code{both}. | |
fdc90cb4 | 2784 | |
0ce398f1 L |
2785 | @kindex --compress-debug-sections=none |
2786 | @kindex --compress-debug-sections=zlib | |
2787 | @kindex --compress-debug-sections=zlib-gnu | |
2788 | @kindex --compress-debug-sections=zlib-gabi | |
2789 | @item --compress-debug-sections=none | |
2790 | @itemx --compress-debug-sections=zlib | |
2791 | @itemx --compress-debug-sections=zlib-gnu | |
2792 | @itemx --compress-debug-sections=zlib-gabi | |
9af89fba NC |
2793 | On ELF platforms, these options control how DWARF debug sections are |
2794 | compressed using zlib. | |
2795 | ||
2796 | @option{--compress-debug-sections=none} doesn't compress DWARF debug | |
2797 | sections. @option{--compress-debug-sections=zlib-gnu} compresses | |
2798 | DWARF debug sections and renames them to begin with @samp{.zdebug} | |
2799 | instead of @samp{.debug}. @option{--compress-debug-sections=zlib-gabi} | |
2800 | also compresses DWARF debug sections, but rather than renaming them it | |
2801 | sets the SHF_COMPRESSED flag in the sections' headers. | |
2802 | ||
2803 | The @option{--compress-debug-sections=zlib} option is an alias for | |
2804 | @option{--compress-debug-sections=zlib-gabi}. | |
2805 | ||
2806 | Note that this option overrides any compression in input debug | |
2807 | sections, so if a binary is linked with @option{--compress-debug-sections=none} | |
2808 | for example, then any compressed debug sections in input files will be | |
2809 | uncompressed before they are copied into the output binary. | |
2810 | ||
2811 | The default compression behaviour varies depending upon the target | |
2812 | involved and the configure options used to build the toolchain. The | |
2813 | default can be determined by examining the output from the linker's | |
2814 | @option{--help} option. | |
0ce398f1 | 2815 | |
35835446 JR |
2816 | @kindex --reduce-memory-overheads |
2817 | @item --reduce-memory-overheads | |
2818 | This option reduces memory requirements at ld runtime, at the expense of | |
f2a8f148 | 2819 | linking speed. This was introduced to select the old O(n^2) algorithm |
35835446 | 2820 | for link map file generation, rather than the new O(n) algorithm which uses |
2d643429 NC |
2821 | about 40% more memory for symbol storage. |
2822 | ||
4f9c04f7 | 2823 | Another effect of the switch is to set the default hash table size to |
2d643429 | 2824 | 1021, which again saves memory at the cost of lengthening the linker's |
a85785bc | 2825 | run time. This is not done however if the @option{--hash-size} switch |
2d643429 NC |
2826 | has been used. |
2827 | ||
2828 | The @option{--reduce-memory-overheads} switch may be also be used to | |
2829 | enable other tradeoffs in future versions of the linker. | |
35835446 | 2830 | |
c0065db7 RM |
2831 | @kindex --build-id |
2832 | @kindex --build-id=@var{style} | |
2833 | @item --build-id | |
2834 | @itemx --build-id=@var{style} | |
61e2488c | 2835 | Request the creation of a @code{.note.gnu.build-id} ELF note section |
6033bf41 | 2836 | or a @code{.buildid} COFF section. The contents of the note are |
61e2488c JT |
2837 | unique bits identifying this linked file. @var{style} can be |
2838 | @code{uuid} to use 128 random bits, @code{sha1} to use a 160-bit | |
2839 | @sc{SHA1} hash on the normative parts of the output contents, | |
2840 | @code{md5} to use a 128-bit @sc{MD5} hash on the normative parts of | |
2841 | the output contents, or @code{0x@var{hexstring}} to use a chosen bit | |
2842 | string specified as an even number of hexadecimal digits (@code{-} and | |
2843 | @code{:} characters between digit pairs are ignored). If @var{style} | |
2844 | is omitted, @code{sha1} is used. | |
24382dca RM |
2845 | |
2846 | The @code{md5} and @code{sha1} styles produces an identifier | |
2847 | that is always the same in an identical output file, but will be | |
2848 | unique among all nonidentical output files. It is not intended | |
2849 | to be compared as a checksum for the file's contents. A linked | |
2850 | file may be changed later by other tools, but the build ID bit | |
2851 | string identifying the original linked file does not change. | |
c0065db7 RM |
2852 | |
2853 | Passing @code{none} for @var{style} disables the setting from any | |
2854 | @code{--build-id} options earlier on the command line. | |
252b5132 RH |
2855 | @end table |
2856 | ||
0285c67d NC |
2857 | @c man end |
2858 | ||
36f63dca | 2859 | @subsection Options Specific to i386 PE Targets |
252b5132 | 2860 | |
0285c67d NC |
2861 | @c man begin OPTIONS |
2862 | ||
ff5dcc92 | 2863 | The i386 PE linker supports the @option{-shared} option, which causes |
252b5132 RH |
2864 | the output to be a dynamically linked library (DLL) instead of a |
2865 | normal executable. You should name the output @code{*.dll} when you | |
2866 | use this option. In addition, the linker fully supports the standard | |
2867 | @code{*.def} files, which may be specified on the linker command line | |
2868 | like an object file (in fact, it should precede archives it exports | |
2869 | symbols from, to ensure that they get linked in, just like a normal | |
2870 | object file). | |
2871 | ||
2872 | In addition to the options common to all targets, the i386 PE linker | |
a05a5b64 | 2873 | support additional command-line options that are specific to the i386 |
252b5132 RH |
2874 | PE target. Options that take values may be separated from their |
2875 | values by either a space or an equals sign. | |
2876 | ||
ff5dcc92 | 2877 | @table @gcctabopt |
252b5132 RH |
2878 | |
2879 | @kindex --add-stdcall-alias | |
2880 | @item --add-stdcall-alias | |
2881 | If given, symbols with a stdcall suffix (@@@var{nn}) will be exported | |
2882 | as-is and also with the suffix stripped. | |
bb10df36 | 2883 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
2884 | |
2885 | @kindex --base-file | |
2886 | @item --base-file @var{file} | |
2887 | Use @var{file} as the name of a file in which to save the base | |
2888 | addresses of all the relocations needed for generating DLLs with | |
2889 | @file{dlltool}. | |
bb10df36 | 2890 | [This is an i386 PE specific option] |
252b5132 RH |
2891 | |
2892 | @kindex --dll | |
2893 | @item --dll | |
2894 | Create a DLL instead of a regular executable. You may also use | |
ff5dcc92 | 2895 | @option{-shared} or specify a @code{LIBRARY} in a given @code{.def} |
252b5132 | 2896 | file. |
bb10df36 | 2897 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 | 2898 | |
88183869 DK |
2899 | @kindex --enable-long-section-names |
2900 | @kindex --disable-long-section-names | |
2901 | @item --enable-long-section-names | |
2902 | @itemx --disable-long-section-names | |
56e6cf80 | 2903 | The PE variants of the COFF object format add an extension that permits |
88183869 | 2904 | the use of section names longer than eight characters, the normal limit |
56e6cf80 NC |
2905 | for COFF. By default, these names are only allowed in object files, as |
2906 | fully-linked executable images do not carry the COFF string table required | |
88183869 DK |
2907 | to support the longer names. As a GNU extension, it is possible to |
2908 | allow their use in executable images as well, or to (probably pointlessly!) | |
2909 | disallow it in object files, by using these two options. Executable images | |
2910 | generated with these long section names are slightly non-standard, carrying | |
2911 | as they do a string table, and may generate confusing output when examined | |
9d5777a3 RM |
2912 | with non-GNU PE-aware tools, such as file viewers and dumpers. However, |
2913 | GDB relies on the use of PE long section names to find Dwarf-2 debug | |
3efd345c DK |
2914 | information sections in an executable image at runtime, and so if neither |
2915 | option is specified on the command-line, @command{ld} will enable long | |
2916 | section names, overriding the default and technically correct behaviour, | |
2917 | when it finds the presence of debug information while linking an executable | |
2918 | image and not stripping symbols. | |
88183869 DK |
2919 | [This option is valid for all PE targeted ports of the linker] |
2920 | ||
252b5132 RH |
2921 | @kindex --enable-stdcall-fixup |
2922 | @kindex --disable-stdcall-fixup | |
2923 | @item --enable-stdcall-fixup | |
2924 | @itemx --disable-stdcall-fixup | |
2925 | If the link finds a symbol that it cannot resolve, it will attempt to | |
36f63dca | 2926 | do ``fuzzy linking'' by looking for another defined symbol that differs |
252b5132 RH |
2927 | only in the format of the symbol name (cdecl vs stdcall) and will |
2928 | resolve that symbol by linking to the match. For example, the | |
2929 | undefined symbol @code{_foo} might be linked to the function | |
2930 | @code{_foo@@12}, or the undefined symbol @code{_bar@@16} might be linked | |
2931 | to the function @code{_bar}. When the linker does this, it prints a | |
2932 | warning, since it normally should have failed to link, but sometimes | |
2933 | import libraries generated from third-party dlls may need this feature | |
ff5dcc92 | 2934 | to be usable. If you specify @option{--enable-stdcall-fixup}, this |
252b5132 | 2935 | feature is fully enabled and warnings are not printed. If you specify |
ff5dcc92 | 2936 | @option{--disable-stdcall-fixup}, this feature is disabled and such |
252b5132 | 2937 | mismatches are considered to be errors. |
bb10df36 | 2938 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 | 2939 | |
522f09cd KT |
2940 | @kindex --leading-underscore |
2941 | @kindex --no-leading-underscore | |
2942 | @item --leading-underscore | |
2943 | @itemx --no-leading-underscore | |
2944 | For most targets default symbol-prefix is an underscore and is defined | |
2945 | in target's description. By this option it is possible to | |
2946 | disable/enable the default underscore symbol-prefix. | |
2947 | ||
252b5132 RH |
2948 | @cindex DLLs, creating |
2949 | @kindex --export-all-symbols | |
2950 | @item --export-all-symbols | |
2951 | If given, all global symbols in the objects used to build a DLL will | |
2952 | be exported by the DLL. Note that this is the default if there | |
2953 | otherwise wouldn't be any exported symbols. When symbols are | |
2954 | explicitly exported via DEF files or implicitly exported via function | |
2955 | attributes, the default is to not export anything else unless this | |
2956 | option is given. Note that the symbols @code{DllMain@@12}, | |
ece2d90e | 2957 | @code{DllEntryPoint@@0}, @code{DllMainCRTStartup@@12}, and |
b044cda1 | 2958 | @code{impure_ptr} will not be automatically |
ece2d90e NC |
2959 | exported. Also, symbols imported from other DLLs will not be |
2960 | re-exported, nor will symbols specifying the DLL's internal layout | |
2961 | such as those beginning with @code{_head_} or ending with | |
2962 | @code{_iname}. In addition, no symbols from @code{libgcc}, | |
b044cda1 CW |
2963 | @code{libstd++}, @code{libmingw32}, or @code{crtX.o} will be exported. |
2964 | Symbols whose names begin with @code{__rtti_} or @code{__builtin_} will | |
2965 | not be exported, to help with C++ DLLs. Finally, there is an | |
ece2d90e | 2966 | extensive list of cygwin-private symbols that are not exported |
b044cda1 | 2967 | (obviously, this applies on when building DLLs for cygwin targets). |
ece2d90e | 2968 | These cygwin-excludes are: @code{_cygwin_dll_entry@@12}, |
b044cda1 | 2969 | @code{_cygwin_crt0_common@@8}, @code{_cygwin_noncygwin_dll_entry@@12}, |
ece2d90e | 2970 | @code{_fmode}, @code{_impure_ptr}, @code{cygwin_attach_dll}, |
b044cda1 | 2971 | @code{cygwin_premain0}, @code{cygwin_premain1}, @code{cygwin_premain2}, |
ece2d90e | 2972 | @code{cygwin_premain3}, and @code{environ}. |
bb10df36 | 2973 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
2974 | |
2975 | @kindex --exclude-symbols | |
1d0a3c9c | 2976 | @item --exclude-symbols @var{symbol},@var{symbol},... |
252b5132 RH |
2977 | Specifies a list of symbols which should not be automatically |
2978 | exported. The symbol names may be delimited by commas or colons. | |
bb10df36 | 2979 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 | 2980 | |
2927aaca NC |
2981 | @kindex --exclude-all-symbols |
2982 | @item --exclude-all-symbols | |
2983 | Specifies no symbols should be automatically exported. | |
2984 | [This option is specific to the i386 PE targeted port of the linker] | |
2985 | ||
252b5132 RH |
2986 | @kindex --file-alignment |
2987 | @item --file-alignment | |
2988 | Specify the file alignment. Sections in the file will always begin at | |
2989 | file offsets which are multiples of this number. This defaults to | |
2990 | 512. | |
bb10df36 | 2991 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
2992 | |
2993 | @cindex heap size | |
2994 | @kindex --heap | |
2995 | @item --heap @var{reserve} | |
2996 | @itemx --heap @var{reserve},@var{commit} | |
a00b50c5 | 2997 | Specify the number of bytes of memory to reserve (and optionally commit) |
fe6d7d6a | 2998 | to be used as heap for this program. The default is 1MB reserved, 4K |
252b5132 | 2999 | committed. |
bb10df36 | 3000 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3001 | |
3002 | @cindex image base | |
3003 | @kindex --image-base | |
3004 | @item --image-base @var{value} | |
3005 | Use @var{value} as the base address of your program or dll. This is | |
3006 | the lowest memory location that will be used when your program or dll | |
3007 | is loaded. To reduce the need to relocate and improve performance of | |
3008 | your dlls, each should have a unique base address and not overlap any | |
3009 | other dlls. The default is 0x400000 for executables, and 0x10000000 | |
3010 | for dlls. | |
bb10df36 | 3011 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3012 | |
3013 | @kindex --kill-at | |
3014 | @item --kill-at | |
3015 | If given, the stdcall suffixes (@@@var{nn}) will be stripped from | |
3016 | symbols before they are exported. | |
bb10df36 | 3017 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 | 3018 | |
26d2d8a2 BF |
3019 | @kindex --large-address-aware |
3020 | @item --large-address-aware | |
b45619c0 | 3021 | If given, the appropriate bit in the ``Characteristics'' field of the COFF |
26d2d8a2 | 3022 | header is set to indicate that this executable supports virtual addresses |
b45619c0 | 3023 | greater than 2 gigabytes. This should be used in conjunction with the /3GB |
26d2d8a2 BF |
3024 | or /USERVA=@var{value} megabytes switch in the ``[operating systems]'' |
3025 | section of the BOOT.INI. Otherwise, this bit has no effect. | |
3026 | [This option is specific to PE targeted ports of the linker] | |
3027 | ||
f69a2f97 NC |
3028 | @kindex --disable-large-address-aware |
3029 | @item --disable-large-address-aware | |
3030 | Reverts the effect of a previous @samp{--large-address-aware} option. | |
3031 | This is useful if @samp{--large-address-aware} is always set by the compiler | |
3032 | driver (e.g. Cygwin gcc) and the executable does not support virtual | |
3033 | addresses greater than 2 gigabytes. | |
3034 | [This option is specific to PE targeted ports of the linker] | |
3035 | ||
252b5132 RH |
3036 | @kindex --major-image-version |
3037 | @item --major-image-version @var{value} | |
36f63dca | 3038 | Sets the major number of the ``image version''. Defaults to 1. |
bb10df36 | 3039 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3040 | |
3041 | @kindex --major-os-version | |
3042 | @item --major-os-version @var{value} | |
36f63dca | 3043 | Sets the major number of the ``os version''. Defaults to 4. |
bb10df36 | 3044 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3045 | |
3046 | @kindex --major-subsystem-version | |
3047 | @item --major-subsystem-version @var{value} | |
36f63dca | 3048 | Sets the major number of the ``subsystem version''. Defaults to 4. |
bb10df36 | 3049 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3050 | |
3051 | @kindex --minor-image-version | |
3052 | @item --minor-image-version @var{value} | |
36f63dca | 3053 | Sets the minor number of the ``image version''. Defaults to 0. |
bb10df36 | 3054 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3055 | |
3056 | @kindex --minor-os-version | |
3057 | @item --minor-os-version @var{value} | |
36f63dca | 3058 | Sets the minor number of the ``os version''. Defaults to 0. |
bb10df36 | 3059 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3060 | |
3061 | @kindex --minor-subsystem-version | |
3062 | @item --minor-subsystem-version @var{value} | |
36f63dca | 3063 | Sets the minor number of the ``subsystem version''. Defaults to 0. |
bb10df36 | 3064 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3065 | |
3066 | @cindex DEF files, creating | |
3067 | @cindex DLLs, creating | |
3068 | @kindex --output-def | |
3069 | @item --output-def @var{file} | |
3070 | The linker will create the file @var{file} which will contain a DEF | |
3071 | file corresponding to the DLL the linker is generating. This DEF file | |
3072 | (which should be called @code{*.def}) may be used to create an import | |
3073 | library with @code{dlltool} or may be used as a reference to | |
3074 | automatically or implicitly exported symbols. | |
bb10df36 | 3075 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 | 3076 | |
b044cda1 | 3077 | @cindex DLLs, creating |
b044cda1 CW |
3078 | @kindex --enable-auto-image-base |
3079 | @item --enable-auto-image-base | |
d0e6d77b CF |
3080 | @itemx --enable-auto-image-base=@var{value} |
3081 | Automatically choose the image base for DLLs, optionally starting with base | |
3082 | @var{value}, unless one is specified using the @code{--image-base} argument. | |
3083 | By using a hash generated from the dllname to create unique image bases | |
3084 | for each DLL, in-memory collisions and relocations which can delay program | |
3085 | execution are avoided. | |
bb10df36 | 3086 | [This option is specific to the i386 PE targeted port of the linker] |
b044cda1 CW |
3087 | |
3088 | @kindex --disable-auto-image-base | |
3089 | @item --disable-auto-image-base | |
3090 | Do not automatically generate a unique image base. If there is no | |
3091 | user-specified image base (@code{--image-base}) then use the platform | |
3092 | default. | |
bb10df36 | 3093 | [This option is specific to the i386 PE targeted port of the linker] |
b044cda1 CW |
3094 | |
3095 | @cindex DLLs, linking to | |
3096 | @kindex --dll-search-prefix | |
3097 | @item --dll-search-prefix @var{string} | |
489d0400 | 3098 | When linking dynamically to a dll without an import library, |
ece2d90e | 3099 | search for @code{<string><basename>.dll} in preference to |
560e09e9 | 3100 | @code{lib<basename>.dll}. This behaviour allows easy distinction |
b044cda1 CW |
3101 | between DLLs built for the various "subplatforms": native, cygwin, |
3102 | uwin, pw, etc. For instance, cygwin DLLs typically use | |
ece2d90e | 3103 | @code{--dll-search-prefix=cyg}. |
bb10df36 | 3104 | [This option is specific to the i386 PE targeted port of the linker] |
b044cda1 CW |
3105 | |
3106 | @kindex --enable-auto-import | |
3107 | @item --enable-auto-import | |
ece2d90e | 3108 | Do sophisticated linking of @code{_symbol} to @code{__imp__symbol} for |
317ff008 EB |
3109 | DATA imports from DLLs, thus making it possible to bypass the dllimport |
3110 | mechanism on the user side and to reference unmangled symbol names. | |
3111 | [This option is specific to the i386 PE targeted port of the linker] | |
3112 | ||
3113 | The following remarks pertain to the original implementation of the | |
3114 | feature and are obsolete nowadays for Cygwin and MinGW targets. | |
3115 | ||
3116 | Note: Use of the 'auto-import' extension will cause the text section | |
3117 | of the image file to be made writable. This does not conform to the | |
3118 | PE-COFF format specification published by Microsoft. | |
4d8907ac | 3119 | |
e2a83dd0 NC |
3120 | Note - use of the 'auto-import' extension will also cause read only |
3121 | data which would normally be placed into the .rdata section to be | |
3122 | placed into the .data section instead. This is in order to work | |
3123 | around a problem with consts that is described here: | |
3124 | http://www.cygwin.com/ml/cygwin/2004-09/msg01101.html | |
3125 | ||
4d8907ac DS |
3126 | Using 'auto-import' generally will 'just work' -- but sometimes you may |
3127 | see this message: | |
0d888aac | 3128 | |
ece2d90e | 3129 | "variable '<var>' can't be auto-imported. Please read the |
0d888aac CW |
3130 | documentation for ld's @code{--enable-auto-import} for details." |
3131 | ||
ece2d90e NC |
3132 | This message occurs when some (sub)expression accesses an address |
3133 | ultimately given by the sum of two constants (Win32 import tables only | |
c0065db7 RM |
3134 | allow one). Instances where this may occur include accesses to member |
3135 | fields of struct variables imported from a DLL, as well as using a | |
3136 | constant index into an array variable imported from a DLL. Any | |
2f8d8971 NC |
3137 | multiword variable (arrays, structs, long long, etc) may trigger |
3138 | this error condition. However, regardless of the exact data type | |
3139 | of the offending exported variable, ld will always detect it, issue | |
3140 | the warning, and exit. | |
3141 | ||
3142 | There are several ways to address this difficulty, regardless of the | |
3143 | data type of the exported variable: | |
0d888aac | 3144 | |
2fa9fc65 NC |
3145 | One way is to use --enable-runtime-pseudo-reloc switch. This leaves the task |
3146 | of adjusting references in your client code for runtime environment, so | |
560e09e9 | 3147 | this method works only when runtime environment supports this feature. |
2fa9fc65 | 3148 | |
c0065db7 RM |
3149 | A second solution is to force one of the 'constants' to be a variable -- |
3150 | that is, unknown and un-optimizable at compile time. For arrays, | |
3151 | there are two possibilities: a) make the indexee (the array's address) | |
0d888aac CW |
3152 | a variable, or b) make the 'constant' index a variable. Thus: |
3153 | ||
3154 | @example | |
3155 | extern type extern_array[]; | |
c0065db7 | 3156 | extern_array[1] --> |
0d888aac CW |
3157 | @{ volatile type *t=extern_array; t[1] @} |
3158 | @end example | |
3159 | ||
3160 | or | |
3161 | ||
3162 | @example | |
3163 | extern type extern_array[]; | |
c0065db7 | 3164 | extern_array[1] --> |
0d888aac CW |
3165 | @{ volatile int t=1; extern_array[t] @} |
3166 | @end example | |
3167 | ||
c0065db7 | 3168 | For structs (and most other multiword data types) the only option |
2f8d8971 | 3169 | is to make the struct itself (or the long long, or the ...) variable: |
0d888aac CW |
3170 | |
3171 | @example | |
3172 | extern struct s extern_struct; | |
c0065db7 | 3173 | extern_struct.field --> |
0d888aac CW |
3174 | @{ volatile struct s *t=&extern_struct; t->field @} |
3175 | @end example | |
3176 | ||
c406afaf NC |
3177 | or |
3178 | ||
3179 | @example | |
3180 | extern long long extern_ll; | |
3181 | extern_ll --> | |
3182 | @{ volatile long long * local_ll=&extern_ll; *local_ll @} | |
3183 | @end example | |
3184 | ||
2fa9fc65 | 3185 | A third method of dealing with this difficulty is to abandon |
c0065db7 | 3186 | 'auto-import' for the offending symbol and mark it with |
11e7fd74 | 3187 | @code{__declspec(dllimport)}. However, in practice that |
0d888aac | 3188 | requires using compile-time #defines to indicate whether you are |
c0065db7 RM |
3189 | building a DLL, building client code that will link to the DLL, or |
3190 | merely building/linking to a static library. In making the choice | |
3191 | between the various methods of resolving the 'direct address with | |
0d888aac CW |
3192 | constant offset' problem, you should consider typical real-world usage: |
3193 | ||
3194 | Original: | |
3195 | @example | |
3196 | --foo.h | |
3197 | extern int arr[]; | |
3198 | --foo.c | |
3199 | #include "foo.h" | |
3200 | void main(int argc, char **argv)@{ | |
3201 | printf("%d\n",arr[1]); | |
3202 | @} | |
3203 | @end example | |
3204 | ||
3205 | Solution 1: | |
3206 | @example | |
3207 | --foo.h | |
3208 | extern int arr[]; | |
3209 | --foo.c | |
3210 | #include "foo.h" | |
3211 | void main(int argc, char **argv)@{ | |
3212 | /* This workaround is for win32 and cygwin; do not "optimize" */ | |
3213 | volatile int *parr = arr; | |
3214 | printf("%d\n",parr[1]); | |
3215 | @} | |
3216 | @end example | |
3217 | ||
3218 | Solution 2: | |
3219 | @example | |
3220 | --foo.h | |
3221 | /* Note: auto-export is assumed (no __declspec(dllexport)) */ | |
3222 | #if (defined(_WIN32) || defined(__CYGWIN__)) && \ | |
3223 | !(defined(FOO_BUILD_DLL) || defined(FOO_STATIC)) | |
3224 | #define FOO_IMPORT __declspec(dllimport) | |
3225 | #else | |
3226 | #define FOO_IMPORT | |
3227 | #endif | |
3228 | extern FOO_IMPORT int arr[]; | |
3229 | --foo.c | |
3230 | #include "foo.h" | |
3231 | void main(int argc, char **argv)@{ | |
3232 | printf("%d\n",arr[1]); | |
3233 | @} | |
3234 | @end example | |
3235 | ||
c0065db7 | 3236 | A fourth way to avoid this problem is to re-code your |
0d888aac CW |
3237 | library to use a functional interface rather than a data interface |
3238 | for the offending variables (e.g. set_foo() and get_foo() accessor | |
3239 | functions). | |
b044cda1 CW |
3240 | |
3241 | @kindex --disable-auto-import | |
3242 | @item --disable-auto-import | |
c0065db7 | 3243 | Do not attempt to do sophisticated linking of @code{_symbol} to |
b044cda1 | 3244 | @code{__imp__symbol} for DATA imports from DLLs. |
bb10df36 | 3245 | [This option is specific to the i386 PE targeted port of the linker] |
b044cda1 | 3246 | |
2fa9fc65 NC |
3247 | @kindex --enable-runtime-pseudo-reloc |
3248 | @item --enable-runtime-pseudo-reloc | |
3249 | If your code contains expressions described in --enable-auto-import section, | |
3250 | that is, DATA imports from DLL with non-zero offset, this switch will create | |
3251 | a vector of 'runtime pseudo relocations' which can be used by runtime | |
c0065db7 | 3252 | environment to adjust references to such data in your client code. |
bb10df36 | 3253 | [This option is specific to the i386 PE targeted port of the linker] |
2fa9fc65 NC |
3254 | |
3255 | @kindex --disable-runtime-pseudo-reloc | |
3256 | @item --disable-runtime-pseudo-reloc | |
317ff008 | 3257 | Do not create pseudo relocations for non-zero offset DATA imports from DLLs. |
bb10df36 | 3258 | [This option is specific to the i386 PE targeted port of the linker] |
2fa9fc65 | 3259 | |
b044cda1 CW |
3260 | @kindex --enable-extra-pe-debug |
3261 | @item --enable-extra-pe-debug | |
3262 | Show additional debug info related to auto-import symbol thunking. | |
bb10df36 | 3263 | [This option is specific to the i386 PE targeted port of the linker] |
b044cda1 | 3264 | |
252b5132 RH |
3265 | @kindex --section-alignment |
3266 | @item --section-alignment | |
3267 | Sets the section alignment. Sections in memory will always begin at | |
3268 | addresses which are a multiple of this number. Defaults to 0x1000. | |
bb10df36 | 3269 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3270 | |
3271 | @cindex stack size | |
3272 | @kindex --stack | |
3273 | @item --stack @var{reserve} | |
3274 | @itemx --stack @var{reserve},@var{commit} | |
a00b50c5 | 3275 | Specify the number of bytes of memory to reserve (and optionally commit) |
fe6d7d6a | 3276 | to be used as stack for this program. The default is 2MB reserved, 4K |
252b5132 | 3277 | committed. |
bb10df36 | 3278 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 RH |
3279 | |
3280 | @kindex --subsystem | |
3281 | @item --subsystem @var{which} | |
3282 | @itemx --subsystem @var{which}:@var{major} | |
3283 | @itemx --subsystem @var{which}:@var{major}.@var{minor} | |
3284 | Specifies the subsystem under which your program will execute. The | |
3285 | legal values for @var{which} are @code{native}, @code{windows}, | |
33f362e1 NC |
3286 | @code{console}, @code{posix}, and @code{xbox}. You may optionally set |
3287 | the subsystem version also. Numeric values are also accepted for | |
3288 | @var{which}. | |
bb10df36 | 3289 | [This option is specific to the i386 PE targeted port of the linker] |
252b5132 | 3290 | |
2f563b51 DK |
3291 | The following options set flags in the @code{DllCharacteristics} field |
3292 | of the PE file header: | |
3293 | [These options are specific to PE targeted ports of the linker] | |
3294 | ||
2d5c3743 NC |
3295 | @kindex --high-entropy-va |
3296 | @item --high-entropy-va | |
514b4e19 | 3297 | @itemx --disable-high-entropy-va |
2d5c3743 | 3298 | Image is compatible with 64-bit address space layout randomization |
514b4e19 JD |
3299 | (ASLR). This option is enabled by default for 64-bit PE images. |
3300 | ||
dc9bd8c9 HD |
3301 | This option also implies @option{--dynamicbase} and |
3302 | @option{--enable-reloc-section}. | |
2d5c3743 | 3303 | |
2f563b51 DK |
3304 | @kindex --dynamicbase |
3305 | @item --dynamicbase | |
514b4e19 | 3306 | @itemx --disable-dynamicbase |
2f563b51 DK |
3307 | The image base address may be relocated using address space layout |
3308 | randomization (ASLR). This feature was introduced with MS Windows | |
514b4e19 JD |
3309 | Vista for i386 PE targets. This option is enabled by default but |
3310 | can be disabled via the @option{--disable-dynamicbase} option. | |
dc9bd8c9 | 3311 | This option also implies @option{--enable-reloc-section}. |
2f563b51 DK |
3312 | |
3313 | @kindex --forceinteg | |
3314 | @item --forceinteg | |
514b4e19 JD |
3315 | @itemx --disable-forceinteg |
3316 | Code integrity checks are enforced. This option is disabled by | |
3317 | default. | |
2f563b51 DK |
3318 | |
3319 | @kindex --nxcompat | |
3320 | @item --nxcompat | |
514b4e19 | 3321 | @item --disable-nxcompat |
2f563b51 | 3322 | The image is compatible with the Data Execution Prevention. |
514b4e19 JD |
3323 | This feature was introduced with MS Windows XP SP2 for i386 PE |
3324 | targets. The option is enabled by default. | |
2f563b51 DK |
3325 | |
3326 | @kindex --no-isolation | |
3327 | @item --no-isolation | |
514b4e19 | 3328 | @itemx --disable-no-isolation |
2f563b51 | 3329 | Although the image understands isolation, do not isolate the image. |
514b4e19 | 3330 | This option is disabled by default. |
2f563b51 DK |
3331 | |
3332 | @kindex --no-seh | |
3333 | @item --no-seh | |
514b4e19 | 3334 | @itemx --disable-no-seh |
2f563b51 | 3335 | The image does not use SEH. No SE handler may be called from |
514b4e19 | 3336 | this image. This option is disabled by default. |
2f563b51 DK |
3337 | |
3338 | @kindex --no-bind | |
3339 | @item --no-bind | |
514b4e19 JD |
3340 | @itemx --disable-no-bind |
3341 | Do not bind this image. This option is disabled by default. | |
2f563b51 DK |
3342 | |
3343 | @kindex --wdmdriver | |
3344 | @item --wdmdriver | |
514b4e19 JD |
3345 | @itemx --disable-wdmdriver |
3346 | The driver uses the MS Windows Driver Model. This option is disabled | |
3347 | by default. | |
9d5777a3 | 3348 | |
2f563b51 DK |
3349 | @kindex --tsaware |
3350 | @item --tsaware | |
514b4e19 JD |
3351 | @itemx --disable-tsaware |
3352 | The image is Terminal Server aware. This option is disabled by | |
3353 | default. | |
2f563b51 | 3354 | |
0cb112f7 CF |
3355 | @kindex --insert-timestamp |
3356 | @item --insert-timestamp | |
eeb14e5a LZ |
3357 | @itemx --no-insert-timestamp |
3358 | Insert a real timestamp into the image. This is the default behaviour | |
3359 | as it matches legacy code and it means that the image will work with | |
3360 | other, proprietary tools. The problem with this default is that it | |
56e6cf80 | 3361 | will result in slightly different images being produced each time the |
eeb14e5a LZ |
3362 | same sources are linked. The option @option{--no-insert-timestamp} |
3363 | can be used to insert a zero value for the timestamp, this ensuring | |
56e6cf80 | 3364 | that binaries produced from identical sources will compare |
eeb14e5a | 3365 | identically. |
dc9bd8c9 HD |
3366 | |
3367 | @kindex --enable-reloc-section | |
3368 | @item --enable-reloc-section | |
514b4e19 | 3369 | @itemx --disable-reloc-section |
dc9bd8c9 HD |
3370 | Create the base relocation table, which is necessary if the image |
3371 | is loaded at a different image base than specified in the PE header. | |
514b4e19 | 3372 | This option is enabled by default. |
252b5132 RH |
3373 | @end table |
3374 | ||
0285c67d NC |
3375 | @c man end |
3376 | ||
ac145307 BS |
3377 | @ifset C6X |
3378 | @subsection Options specific to C6X uClinux targets | |
3379 | ||
3380 | @c man begin OPTIONS | |
3381 | ||
3382 | The C6X uClinux target uses a binary format called DSBT to support shared | |
3383 | libraries. Each shared library in the system needs to have a unique index; | |
3384 | all executables use an index of 0. | |
3385 | ||
3386 | @table @gcctabopt | |
3387 | ||
3388 | @kindex --dsbt-size | |
3389 | @item --dsbt-size @var{size} | |
56e6cf80 | 3390 | This option sets the number of entries in the DSBT of the current executable |
ac145307 BS |
3391 | or shared library to @var{size}. The default is to create a table with 64 |
3392 | entries. | |
3393 | ||
3394 | @kindex --dsbt-index | |
3395 | @item --dsbt-index @var{index} | |
3396 | This option sets the DSBT index of the current executable or shared library | |
3397 | to @var{index}. The default is 0, which is appropriate for generating | |
3398 | executables. If a shared library is generated with a DSBT index of 0, the | |
3399 | @code{R_C6000_DSBT_INDEX} relocs are copied into the output file. | |
3400 | ||
fbd9ad90 PB |
3401 | @kindex --no-merge-exidx-entries |
3402 | The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent | |
3403 | exidx entries in frame unwind info. | |
3404 | ||
ac145307 BS |
3405 | @end table |
3406 | ||
3407 | @c man end | |
3408 | @end ifset | |
3409 | ||
b8891f8d AJ |
3410 | @ifset CSKY |
3411 | @subsection Options specific to C-SKY targets | |
3412 | ||
3413 | @c man begin OPTIONS | |
3414 | ||
3415 | @table @gcctabopt | |
3416 | ||
3417 | @kindex --branch-stub on C-SKY | |
3418 | @item --branch-stub | |
3419 | This option enables linker branch relaxation by inserting branch stub | |
3420 | sections when needed to extend the range of branches. This option is | |
3421 | usually not required since C-SKY supports branch and call instructions that | |
3422 | can access the full memory range and branch relaxation is normally handled by | |
3423 | the compiler or assembler. | |
3424 | ||
3425 | @kindex --stub-group-size on C-SKY | |
3426 | @item --stub-group-size=@var{N} | |
3427 | This option allows finer control of linker branch stub creation. | |
3428 | It sets the maximum size of a group of input sections that can | |
3429 | be handled by one stub section. A negative value of @var{N} locates | |
3430 | stub sections after their branches, while a positive value allows stub | |
3431 | sections to appear either before or after the branches. Values of | |
3432 | @samp{1} or @samp{-1} indicate that the | |
3433 | linker should choose suitable defaults. | |
3434 | ||
3435 | @end table | |
3436 | ||
3437 | @c man end | |
3438 | @end ifset | |
3439 | ||
93fd0973 SC |
3440 | @ifset M68HC11 |
3441 | @subsection Options specific to Motorola 68HC11 and 68HC12 targets | |
3442 | ||
3443 | @c man begin OPTIONS | |
3444 | ||
3445 | The 68HC11 and 68HC12 linkers support specific options to control the | |
3446 | memory bank switching mapping and trampoline code generation. | |
3447 | ||
3448 | @table @gcctabopt | |
3449 | ||
3450 | @kindex --no-trampoline | |
3451 | @item --no-trampoline | |
3452 | This option disables the generation of trampoline. By default a trampoline | |
3453 | is generated for each far function which is called using a @code{jsr} | |
3454 | instruction (this happens when a pointer to a far function is taken). | |
3455 | ||
3456 | @kindex --bank-window | |
3457 | @item --bank-window @var{name} | |
3458 | This option indicates to the linker the name of the memory region in | |
3459 | the @samp{MEMORY} specification that describes the memory bank window. | |
3460 | The definition of such region is then used by the linker to compute | |
3461 | paging and addresses within the memory window. | |
3462 | ||
3463 | @end table | |
3464 | ||
3465 | @c man end | |
3466 | @end ifset | |
3467 | ||
7fb9f789 NC |
3468 | @ifset M68K |
3469 | @subsection Options specific to Motorola 68K target | |
3470 | ||
3471 | @c man begin OPTIONS | |
3472 | ||
3473 | The following options are supported to control handling of GOT generation | |
3474 | when linking for 68K targets. | |
3475 | ||
3476 | @table @gcctabopt | |
3477 | ||
3478 | @kindex --got | |
3479 | @item --got=@var{type} | |
3480 | This option tells the linker which GOT generation scheme to use. | |
3481 | @var{type} should be one of @samp{single}, @samp{negative}, | |
3482 | @samp{multigot} or @samp{target}. For more information refer to the | |
3483 | Info entry for @file{ld}. | |
3484 | ||
3485 | @end table | |
3486 | ||
3487 | @c man end | |
3488 | @end ifset | |
3489 | ||
833794fc MR |
3490 | @ifset MIPS |
3491 | @subsection Options specific to MIPS targets | |
3492 | ||
3493 | @c man begin OPTIONS | |
3494 | ||
3495 | The following options are supported to control microMIPS instruction | |
8b10b0b3 MR |
3496 | generation and branch relocation checks for ISA mode transitions when |
3497 | linking for MIPS targets. | |
833794fc MR |
3498 | |
3499 | @table @gcctabopt | |
3500 | ||
3501 | @kindex --insn32 | |
3502 | @item --insn32 | |
3503 | @kindex --no-insn32 | |
3504 | @itemx --no-insn32 | |
3505 | These options control the choice of microMIPS instructions used in code | |
3506 | generated by the linker, such as that in the PLT or lazy binding stubs, | |
3507 | or in relaxation. If @samp{--insn32} is used, then the linker only uses | |
3508 | 32-bit instruction encodings. By default or if @samp{--no-insn32} is | |
3509 | used, all instruction encodings are used, including 16-bit ones where | |
3510 | possible. | |
3511 | ||
8b10b0b3 MR |
3512 | @kindex --ignore-branch-isa |
3513 | @item --ignore-branch-isa | |
3514 | @kindex --no-ignore-branch-isa | |
3515 | @itemx --no-ignore-branch-isa | |
3516 | These options control branch relocation checks for invalid ISA mode | |
3517 | transitions. If @samp{--ignore-branch-isa} is used, then the linker | |
3518 | accepts any branch relocations and any ISA mode transition required | |
3519 | is lost in relocation calculation, except for some cases of @code{BAL} | |
3520 | instructions which meet relaxation conditions and are converted to | |
3521 | equivalent @code{JALX} instructions as the associated relocation is | |
3522 | calculated. By default or if @samp{--no-ignore-branch-isa} is used | |
3523 | a check is made causing the loss of an ISA mode transition to produce | |
3524 | an error. | |
3525 | ||
3734320d MF |
3526 | @kindex --compact-branches |
3527 | @item --compact-branches | |
3528 | @kindex --no-compact-branches | |
fa1477dc | 3529 | @itemx --no-compact-branches |
3734320d MF |
3530 | These options control the generation of compact instructions by the linker |
3531 | in the PLT entries for MIPS R6. | |
3532 | ||
833794fc MR |
3533 | @end table |
3534 | ||
3535 | @c man end | |
3536 | @end ifset | |
3537 | ||
fa1477dc SC |
3538 | |
3539 | @ifset PDP11 | |
3540 | @subsection Options specific to PDP11 targets | |
3541 | ||
3542 | @c man begin OPTIONS | |
3543 | ||
3544 | For the pdp11-aout target, three variants of the output format can be | |
3545 | produced as selected by the following options. The default variant | |
3546 | for pdp11-aout is the @samp{--omagic} option, whereas for other | |
3547 | targets @samp{--nmagic} is the default. The @samp{--imagic} option is | |
3548 | defined only for the pdp11-aout target, while the others are described | |
3549 | here as they apply to the pdp11-aout target. | |
3550 | ||
3551 | @table @gcctabopt | |
3552 | ||
3553 | @kindex -N | |
3554 | @item -N | |
3555 | @kindex --omagic | |
3556 | @itemx --omagic | |
3557 | ||
3558 | Mark the output as @code{OMAGIC} (0407) in the @file{a.out} header to | |
3559 | indicate that the text segment is not to be write-protected and | |
3560 | shared. Since the text and data sections are both readable and | |
3561 | writable, the data section is allocated immediately contiguous after | |
3562 | the text segment. This is the oldest format for PDP11 executable | |
3563 | programs and is the default for @command{ld} on PDP11 Unix systems | |
3564 | from the beginning through 2.11BSD. | |
3565 | ||
3566 | @kindex -n | |
3567 | @item -n | |
3568 | @kindex --nmagic | |
3569 | @itemx --nmagic | |
3570 | ||
3571 | Mark the output as @code{NMAGIC} (0410) in the @file{a.out} header to | |
3572 | indicate that when the output file is executed, the text portion will | |
3573 | be read-only and shareable among all processes executing the same | |
3574 | file. This involves moving the data areas up to the first possible 8K | |
3575 | byte page boundary following the end of the text. This option creates | |
3576 | a @emph{pure executable} format. | |
3577 | ||
3578 | @kindex -z | |
3579 | @item -z | |
3580 | @kindex --imagic | |
3581 | @itemx --imagic | |
3582 | ||
3583 | Mark the output as @code{IMAGIC} (0411) in the @file{a.out} header to | |
3584 | indicate that when the output file is executed, the program text and | |
3585 | data areas will be loaded into separate address spaces using the split | |
3586 | instruction and data space feature of the memory management unit in | |
3587 | larger models of the PDP11. This doubles the address space available | |
3588 | to the program. The text segment is again pure, write-protected, and | |
3589 | shareable. The only difference in the output format between this | |
3590 | option and the others, besides the magic number, is that both the text | |
3591 | and data sections start at location 0. The @samp{-z} option selected | |
3592 | this format in 2.11BSD. This option creates a @emph{separate | |
3593 | executable} format. | |
3594 | ||
3595 | @kindex --no-omagic | |
3596 | @item --no-omagic | |
3597 | ||
3598 | Equivalent to @samp{--nmagic} for pdp11-aout. | |
3599 | ||
3600 | @end table | |
3601 | ||
3602 | @c man end | |
3603 | @end ifset | |
3604 | ||
252b5132 RH |
3605 | @ifset UsesEnvVars |
3606 | @node Environment | |
3607 | @section Environment Variables | |
3608 | ||
0285c67d NC |
3609 | @c man begin ENVIRONMENT |
3610 | ||
560e09e9 | 3611 | You can change the behaviour of @command{ld} with the environment variables |
36f63dca NC |
3612 | @ifclear SingleFormat |
3613 | @code{GNUTARGET}, | |
3614 | @end ifclear | |
3615 | @code{LDEMULATION} and @code{COLLECT_NO_DEMANGLE}. | |
252b5132 | 3616 | |
36f63dca | 3617 | @ifclear SingleFormat |
252b5132 RH |
3618 | @kindex GNUTARGET |
3619 | @cindex default input format | |
3620 | @code{GNUTARGET} determines the input-file object format if you don't | |
3621 | use @samp{-b} (or its synonym @samp{--format}). Its value should be one | |
3622 | of the BFD names for an input format (@pxref{BFD}). If there is no | |
ff5dcc92 | 3623 | @code{GNUTARGET} in the environment, @command{ld} uses the natural format |
252b5132 RH |
3624 | of the target. If @code{GNUTARGET} is set to @code{default} then BFD |
3625 | attempts to discover the input format by examining binary input files; | |
3626 | this method often succeeds, but there are potential ambiguities, since | |
3627 | there is no method of ensuring that the magic number used to specify | |
3628 | object-file formats is unique. However, the configuration procedure for | |
3629 | BFD on each system places the conventional format for that system first | |
3630 | in the search-list, so ambiguities are resolved in favor of convention. | |
36f63dca | 3631 | @end ifclear |
252b5132 RH |
3632 | |
3633 | @kindex LDEMULATION | |
3634 | @cindex default emulation | |
3635 | @cindex emulation, default | |
3636 | @code{LDEMULATION} determines the default emulation if you don't use the | |
3637 | @samp{-m} option. The emulation can affect various aspects of linker | |
3638 | behaviour, particularly the default linker script. You can list the | |
3639 | available emulations with the @samp{--verbose} or @samp{-V} options. If | |
3640 | the @samp{-m} option is not used, and the @code{LDEMULATION} environment | |
3641 | variable is not defined, the default emulation depends upon how the | |
3642 | linker was configured. | |
252b5132 RH |
3643 | |
3644 | @kindex COLLECT_NO_DEMANGLE | |
3645 | @cindex demangling, default | |
3646 | Normally, the linker will default to demangling symbols. However, if | |
3647 | @code{COLLECT_NO_DEMANGLE} is set in the environment, then it will | |
3648 | default to not demangling symbols. This environment variable is used in | |
3649 | a similar fashion by the @code{gcc} linker wrapper program. The default | |
3650 | may be overridden by the @samp{--demangle} and @samp{--no-demangle} | |
3651 | options. | |
3652 | ||
0285c67d NC |
3653 | @c man end |
3654 | @end ifset | |
3655 | ||
252b5132 RH |
3656 | @node Scripts |
3657 | @chapter Linker Scripts | |
3658 | ||
3659 | @cindex scripts | |
3660 | @cindex linker scripts | |
3661 | @cindex command files | |
3662 | Every link is controlled by a @dfn{linker script}. This script is | |
3663 | written in the linker command language. | |
3664 | ||
3665 | The main purpose of the linker script is to describe how the sections in | |
3666 | the input files should be mapped into the output file, and to control | |
3667 | the memory layout of the output file. Most linker scripts do nothing | |
3668 | more than this. However, when necessary, the linker script can also | |
3669 | direct the linker to perform many other operations, using the commands | |
3670 | described below. | |
3671 | ||
3672 | The linker always uses a linker script. If you do not supply one | |
3673 | yourself, the linker will use a default script that is compiled into the | |
a05a5b64 TP |
3674 | linker executable. You can use the @samp{--verbose} command-line option |
3675 | to display the default linker script. Certain command-line options, | |
252b5132 RH |
3676 | such as @samp{-r} or @samp{-N}, will affect the default linker script. |
3677 | ||
3678 | You may supply your own linker script by using the @samp{-T} command | |
3679 | line option. When you do this, your linker script will replace the | |
3680 | default linker script. | |
3681 | ||
3682 | You may also use linker scripts implicitly by naming them as input files | |
3683 | to the linker, as though they were files to be linked. @xref{Implicit | |
3684 | Linker Scripts}. | |
3685 | ||
3686 | @menu | |
3687 | * Basic Script Concepts:: Basic Linker Script Concepts | |
3688 | * Script Format:: Linker Script Format | |
3689 | * Simple Example:: Simple Linker Script Example | |
3690 | * Simple Commands:: Simple Linker Script Commands | |
3691 | * Assignments:: Assigning Values to Symbols | |
3692 | * SECTIONS:: SECTIONS Command | |
3693 | * MEMORY:: MEMORY Command | |
3694 | * PHDRS:: PHDRS Command | |
3695 | * VERSION:: VERSION Command | |
3696 | * Expressions:: Expressions in Linker Scripts | |
3697 | * Implicit Linker Scripts:: Implicit Linker Scripts | |
3698 | @end menu | |
3699 | ||
3700 | @node Basic Script Concepts | |
3701 | @section Basic Linker Script Concepts | |
3702 | @cindex linker script concepts | |
3703 | We need to define some basic concepts and vocabulary in order to | |
3704 | describe the linker script language. | |
3705 | ||
3706 | The linker combines input files into a single output file. The output | |
3707 | file and each input file are in a special data format known as an | |
3708 | @dfn{object file format}. Each file is called an @dfn{object file}. | |
3709 | The output file is often called an @dfn{executable}, but for our | |
3710 | purposes we will also call it an object file. Each object file has, | |
3711 | among other things, a list of @dfn{sections}. We sometimes refer to a | |
3712 | section in an input file as an @dfn{input section}; similarly, a section | |
3713 | in the output file is an @dfn{output section}. | |
3714 | ||
3715 | Each section in an object file has a name and a size. Most sections | |
3716 | also have an associated block of data, known as the @dfn{section | |
56dd11f0 | 3717 | contents}. A section may be marked as @dfn{loadable}, which means that |
252b5132 RH |
3718 | the contents should be loaded into memory when the output file is run. |
3719 | A section with no contents may be @dfn{allocatable}, which means that an | |
3720 | area in memory should be set aside, but nothing in particular should be | |
3721 | loaded there (in some cases this memory must be zeroed out). A section | |
3722 | which is neither loadable nor allocatable typically contains some sort | |
3723 | of debugging information. | |
3724 | ||
3725 | Every loadable or allocatable output section has two addresses. The | |
3726 | first is the @dfn{VMA}, or virtual memory address. This is the address | |
3727 | the section will have when the output file is run. The second is the | |
3728 | @dfn{LMA}, or load memory address. This is the address at which the | |
3729 | section will be loaded. In most cases the two addresses will be the | |
3730 | same. An example of when they might be different is when a data section | |
3731 | is loaded into ROM, and then copied into RAM when the program starts up | |
3732 | (this technique is often used to initialize global variables in a ROM | |
3733 | based system). In this case the ROM address would be the LMA, and the | |
3734 | RAM address would be the VMA. | |
3735 | ||
3736 | You can see the sections in an object file by using the @code{objdump} | |
3737 | program with the @samp{-h} option. | |
3738 | ||
3739 | Every object file also has a list of @dfn{symbols}, known as the | |
3740 | @dfn{symbol table}. A symbol may be defined or undefined. Each symbol | |
3741 | has a name, and each defined symbol has an address, among other | |
3742 | information. If you compile a C or C++ program into an object file, you | |
3743 | will get a defined symbol for every defined function and global or | |
3744 | static variable. Every undefined function or global variable which is | |
3745 | referenced in the input file will become an undefined symbol. | |
3746 | ||
3747 | You can see the symbols in an object file by using the @code{nm} | |
3748 | program, or by using the @code{objdump} program with the @samp{-t} | |
3749 | option. | |
3750 | ||
3751 | @node Script Format | |
3752 | @section Linker Script Format | |
3753 | @cindex linker script format | |
3754 | Linker scripts are text files. | |
3755 | ||
3756 | You write a linker script as a series of commands. Each command is | |
3757 | either a keyword, possibly followed by arguments, or an assignment to a | |
3758 | symbol. You may separate commands using semicolons. Whitespace is | |
3759 | generally ignored. | |
3760 | ||
3761 | Strings such as file or format names can normally be entered directly. | |
3762 | If the file name contains a character such as a comma which would | |
3763 | otherwise serve to separate file names, you may put the file name in | |
3764 | double quotes. There is no way to use a double quote character in a | |
3765 | file name. | |
3766 | ||
3767 | You may include comments in linker scripts just as in C, delimited by | |
3768 | @samp{/*} and @samp{*/}. As in C, comments are syntactically equivalent | |
3769 | to whitespace. | |
3770 | ||
3771 | @node Simple Example | |
3772 | @section Simple Linker Script Example | |
3773 | @cindex linker script example | |
3774 | @cindex example of linker script | |
3775 | Many linker scripts are fairly simple. | |
3776 | ||
3777 | The simplest possible linker script has just one command: | |
3778 | @samp{SECTIONS}. You use the @samp{SECTIONS} command to describe the | |
3779 | memory layout of the output file. | |
3780 | ||
3781 | The @samp{SECTIONS} command is a powerful command. Here we will | |
3782 | describe a simple use of it. Let's assume your program consists only of | |
3783 | code, initialized data, and uninitialized data. These will be in the | |
3784 | @samp{.text}, @samp{.data}, and @samp{.bss} sections, respectively. | |
3785 | Let's assume further that these are the only sections which appear in | |
3786 | your input files. | |
3787 | ||
3788 | For this example, let's say that the code should be loaded at address | |
3789 | 0x10000, and that the data should start at address 0x8000000. Here is a | |
3790 | linker script which will do that: | |
3791 | @smallexample | |
3792 | SECTIONS | |
3793 | @{ | |
3794 | . = 0x10000; | |
3795 | .text : @{ *(.text) @} | |
3796 | . = 0x8000000; | |
3797 | .data : @{ *(.data) @} | |
3798 | .bss : @{ *(.bss) @} | |
3799 | @} | |
3800 | @end smallexample | |
3801 | ||
3802 | You write the @samp{SECTIONS} command as the keyword @samp{SECTIONS}, | |
3803 | followed by a series of symbol assignments and output section | |
3804 | descriptions enclosed in curly braces. | |
3805 | ||
252b5132 RH |
3806 | The first line inside the @samp{SECTIONS} command of the above example |
3807 | sets the value of the special symbol @samp{.}, which is the location | |
3808 | counter. If you do not specify the address of an output section in some | |
3809 | other way (other ways are described later), the address is set from the | |
3810 | current value of the location counter. The location counter is then | |
3811 | incremented by the size of the output section. At the start of the | |
3812 | @samp{SECTIONS} command, the location counter has the value @samp{0}. | |
3813 | ||
3814 | The second line defines an output section, @samp{.text}. The colon is | |
3815 | required syntax which may be ignored for now. Within the curly braces | |
3816 | after the output section name, you list the names of the input sections | |
3817 | which should be placed into this output section. The @samp{*} is a | |
3818 | wildcard which matches any file name. The expression @samp{*(.text)} | |
3819 | means all @samp{.text} input sections in all input files. | |
3820 | ||
3821 | Since the location counter is @samp{0x10000} when the output section | |
3822 | @samp{.text} is defined, the linker will set the address of the | |
3823 | @samp{.text} section in the output file to be @samp{0x10000}. | |
3824 | ||
3825 | The remaining lines define the @samp{.data} and @samp{.bss} sections in | |
3826 | the output file. The linker will place the @samp{.data} output section | |
3827 | at address @samp{0x8000000}. After the linker places the @samp{.data} | |
3828 | output section, the value of the location counter will be | |
3829 | @samp{0x8000000} plus the size of the @samp{.data} output section. The | |
3830 | effect is that the linker will place the @samp{.bss} output section | |
58434bc1 | 3831 | immediately after the @samp{.data} output section in memory. |
252b5132 RH |
3832 | |
3833 | The linker will ensure that each output section has the required | |
3834 | alignment, by increasing the location counter if necessary. In this | |
3835 | example, the specified addresses for the @samp{.text} and @samp{.data} | |
3836 | sections will probably satisfy any alignment constraints, but the linker | |
3837 | may have to create a small gap between the @samp{.data} and @samp{.bss} | |
3838 | sections. | |
3839 | ||
3840 | That's it! That's a simple and complete linker script. | |
3841 | ||
3842 | @node Simple Commands | |
3843 | @section Simple Linker Script Commands | |
3844 | @cindex linker script simple commands | |
3845 | In this section we describe the simple linker script commands. | |
3846 | ||
3847 | @menu | |
3848 | * Entry Point:: Setting the entry point | |
3849 | * File Commands:: Commands dealing with files | |
3850 | @ifclear SingleFormat | |
3851 | * Format Commands:: Commands dealing with object file formats | |
3852 | @end ifclear | |
3853 | ||
4a93e180 | 3854 | * REGION_ALIAS:: Assign alias names to memory regions |
252b5132 RH |
3855 | * Miscellaneous Commands:: Other linker script commands |
3856 | @end menu | |
3857 | ||
3858 | @node Entry Point | |
36f63dca | 3859 | @subsection Setting the Entry Point |
252b5132 RH |
3860 | @kindex ENTRY(@var{symbol}) |
3861 | @cindex start of execution | |
3862 | @cindex first instruction | |
3863 | @cindex entry point | |
3864 | The first instruction to execute in a program is called the @dfn{entry | |
3865 | point}. You can use the @code{ENTRY} linker script command to set the | |
3866 | entry point. The argument is a symbol name: | |
3867 | @smallexample | |
3868 | ENTRY(@var{symbol}) | |
3869 | @end smallexample | |
3870 | ||
3871 | There are several ways to set the entry point. The linker will set the | |
3872 | entry point by trying each of the following methods in order, and | |
3873 | stopping when one of them succeeds: | |
3874 | @itemize @bullet | |
a1ab1d2a | 3875 | @item |
252b5132 | 3876 | the @samp{-e} @var{entry} command-line option; |
a1ab1d2a | 3877 | @item |
252b5132 | 3878 | the @code{ENTRY(@var{symbol})} command in a linker script; |
a1ab1d2a | 3879 | @item |
a094d01f | 3880 | the value of a target-specific symbol, if it is defined; For many |
8a758655 | 3881 | targets this is @code{start}, but PE- and BeOS-based systems for example |
3ab904c4 | 3882 | check a list of possible entry symbols, matching the first one found. |
a1ab1d2a | 3883 | @item |
252b5132 | 3884 | the address of the first byte of the @samp{.text} section, if present; |
a1ab1d2a | 3885 | @item |
252b5132 RH |
3886 | The address @code{0}. |
3887 | @end itemize | |
3888 | ||
3889 | @node File Commands | |
36f63dca | 3890 | @subsection Commands Dealing with Files |
252b5132 RH |
3891 | @cindex linker script file commands |
3892 | Several linker script commands deal with files. | |
3893 | ||
3894 | @table @code | |
3895 | @item INCLUDE @var{filename} | |
3896 | @kindex INCLUDE @var{filename} | |
3897 | @cindex including a linker script | |
3898 | Include the linker script @var{filename} at this point. The file will | |
3899 | be searched for in the current directory, and in any directory specified | |
ff5dcc92 | 3900 | with the @option{-L} option. You can nest calls to @code{INCLUDE} up to |
252b5132 RH |
3901 | 10 levels deep. |
3902 | ||
4006703d NS |
3903 | You can place @code{INCLUDE} directives at the top level, in @code{MEMORY} or |
3904 | @code{SECTIONS} commands, or in output section descriptions. | |
3905 | ||
252b5132 RH |
3906 | @item INPUT(@var{file}, @var{file}, @dots{}) |
3907 | @itemx INPUT(@var{file} @var{file} @dots{}) | |
3908 | @kindex INPUT(@var{files}) | |
3909 | @cindex input files in linker scripts | |
3910 | @cindex input object files in linker scripts | |
3911 | @cindex linker script input object files | |
3912 | The @code{INPUT} command directs the linker to include the named files | |
3913 | in the link, as though they were named on the command line. | |
3914 | ||
3915 | For example, if you always want to include @file{subr.o} any time you do | |
3916 | a link, but you can't be bothered to put it on every link command line, | |
3917 | then you can put @samp{INPUT (subr.o)} in your linker script. | |
3918 | ||
3919 | In fact, if you like, you can list all of your input files in the linker | |
3920 | script, and then invoke the linker with nothing but a @samp{-T} option. | |
3921 | ||
e3f2db7f AO |
3922 | In case a @dfn{sysroot prefix} is configured, and the filename starts |
3923 | with the @samp{/} character, and the script being processed was | |
3924 | located inside the @dfn{sysroot prefix}, the filename will be looked | |
16171946 FS |
3925 | for in the @dfn{sysroot prefix}. The @dfn{sysroot prefix} can also be forced by specifying |
3926 | @code{=} as the first character in the filename path, or prefixing the | |
3927 | filename path with @code{$SYSROOT}. See also the description of | |
3928 | @samp{-L} in @ref{Options,,Command-line Options}. | |
3929 | ||
3930 | If a @dfn{sysroot prefix} is not used then the linker will try to open | |
3931 | the file in the directory containing the linker script. If it is not | |
3932 | found the linker will then search the current directory. If it is still | |
3933 | not found the linker will search through the archive library search | |
3934 | path. | |
252b5132 | 3935 | |
ff5dcc92 | 3936 | If you use @samp{INPUT (-l@var{file})}, @command{ld} will transform the |
a05a5b64 | 3937 | name to @code{lib@var{file}.a}, as with the command-line argument |
252b5132 RH |
3938 | @samp{-l}. |
3939 | ||
3940 | When you use the @code{INPUT} command in an implicit linker script, the | |
3941 | files will be included in the link at the point at which the linker | |
3942 | script file is included. This can affect archive searching. | |
3943 | ||
3944 | @item GROUP(@var{file}, @var{file}, @dots{}) | |
3945 | @itemx GROUP(@var{file} @var{file} @dots{}) | |
3946 | @kindex GROUP(@var{files}) | |
3947 | @cindex grouping input files | |
3948 | The @code{GROUP} command is like @code{INPUT}, except that the named | |
3949 | files should all be archives, and they are searched repeatedly until no | |
3950 | new undefined references are created. See the description of @samp{-(} | |
a05a5b64 | 3951 | in @ref{Options,,Command-line Options}. |
252b5132 | 3952 | |
b717d30e JJ |
3953 | @item AS_NEEDED(@var{file}, @var{file}, @dots{}) |
3954 | @itemx AS_NEEDED(@var{file} @var{file} @dots{}) | |
3955 | @kindex AS_NEEDED(@var{files}) | |
3956 | This construct can appear only inside of the @code{INPUT} or @code{GROUP} | |
3957 | commands, among other filenames. The files listed will be handled | |
3958 | as if they appear directly in the @code{INPUT} or @code{GROUP} commands, | |
3959 | with the exception of ELF shared libraries, that will be added only | |
3960 | when they are actually needed. This construct essentially enables | |
3961 | @option{--as-needed} option for all the files listed inside of it | |
3962 | and restores previous @option{--as-needed} resp. @option{--no-as-needed} | |
3963 | setting afterwards. | |
3964 | ||
252b5132 RH |
3965 | @item OUTPUT(@var{filename}) |
3966 | @kindex OUTPUT(@var{filename}) | |
b45619c0 | 3967 | @cindex output file name in linker script |
252b5132 RH |
3968 | The @code{OUTPUT} command names the output file. Using |
3969 | @code{OUTPUT(@var{filename})} in the linker script is exactly like using | |
3970 | @samp{-o @var{filename}} on the command line (@pxref{Options,,Command | |
a05a5b64 | 3971 | Line Options}). If both are used, the command-line option takes |
252b5132 RH |
3972 | precedence. |
3973 | ||
3974 | You can use the @code{OUTPUT} command to define a default name for the | |
3975 | output file other than the usual default of @file{a.out}. | |
3976 | ||
3977 | @item SEARCH_DIR(@var{path}) | |
3978 | @kindex SEARCH_DIR(@var{path}) | |
3979 | @cindex library search path in linker script | |
3980 | @cindex archive search path in linker script | |
3981 | @cindex search path in linker script | |
3982 | The @code{SEARCH_DIR} command adds @var{path} to the list of paths where | |
ff5dcc92 | 3983 | @command{ld} looks for archive libraries. Using |
252b5132 | 3984 | @code{SEARCH_DIR(@var{path})} is exactly like using @samp{-L @var{path}} |
a05a5b64 | 3985 | on the command line (@pxref{Options,,Command-line Options}). If both |
252b5132 | 3986 | are used, then the linker will search both paths. Paths specified using |
a05a5b64 | 3987 | the command-line option are searched first. |
252b5132 RH |
3988 | |
3989 | @item STARTUP(@var{filename}) | |
3990 | @kindex STARTUP(@var{filename}) | |
3991 | @cindex first input file | |
3992 | The @code{STARTUP} command is just like the @code{INPUT} command, except | |
3993 | that @var{filename} will become the first input file to be linked, as | |
3994 | though it were specified first on the command line. This may be useful | |
3995 | when using a system in which the entry point is always the start of the | |
3996 | first file. | |
3997 | @end table | |
3998 | ||
3999 | @ifclear SingleFormat | |
4000 | @node Format Commands | |
36f63dca | 4001 | @subsection Commands Dealing with Object File Formats |
252b5132 RH |
4002 | A couple of linker script commands deal with object file formats. |
4003 | ||
4004 | @table @code | |
4005 | @item OUTPUT_FORMAT(@var{bfdname}) | |
4006 | @itemx OUTPUT_FORMAT(@var{default}, @var{big}, @var{little}) | |
4007 | @kindex OUTPUT_FORMAT(@var{bfdname}) | |
4008 | @cindex output file format in linker script | |
4009 | The @code{OUTPUT_FORMAT} command names the BFD format to use for the | |
4010 | output file (@pxref{BFD}). Using @code{OUTPUT_FORMAT(@var{bfdname})} is | |
024531e2 | 4011 | exactly like using @samp{--oformat @var{bfdname}} on the command line |
a05a5b64 | 4012 | (@pxref{Options,,Command-line Options}). If both are used, the command |
252b5132 RH |
4013 | line option takes precedence. |
4014 | ||
4015 | You can use @code{OUTPUT_FORMAT} with three arguments to use different | |
a05a5b64 | 4016 | formats based on the @samp{-EB} and @samp{-EL} command-line options. |
252b5132 RH |
4017 | This permits the linker script to set the output format based on the |
4018 | desired endianness. | |
4019 | ||
4020 | If neither @samp{-EB} nor @samp{-EL} are used, then the output format | |
4021 | will be the first argument, @var{default}. If @samp{-EB} is used, the | |
4022 | output format will be the second argument, @var{big}. If @samp{-EL} is | |
4023 | used, the output format will be the third argument, @var{little}. | |
4024 | ||
4025 | For example, the default linker script for the MIPS ELF target uses this | |
4026 | command: | |
4027 | @smallexample | |
4028 | OUTPUT_FORMAT(elf32-bigmips, elf32-bigmips, elf32-littlemips) | |
4029 | @end smallexample | |
4030 | This says that the default format for the output file is | |
a05a5b64 | 4031 | @samp{elf32-bigmips}, but if the user uses the @samp{-EL} command-line |
252b5132 RH |
4032 | option, the output file will be created in the @samp{elf32-littlemips} |
4033 | format. | |
4034 | ||
4035 | @item TARGET(@var{bfdname}) | |
4036 | @kindex TARGET(@var{bfdname}) | |
4037 | @cindex input file format in linker script | |
4038 | The @code{TARGET} command names the BFD format to use when reading input | |
4039 | files. It affects subsequent @code{INPUT} and @code{GROUP} commands. | |
4040 | This command is like using @samp{-b @var{bfdname}} on the command line | |
a05a5b64 | 4041 | (@pxref{Options,,Command-line Options}). If the @code{TARGET} command |
252b5132 RH |
4042 | is used but @code{OUTPUT_FORMAT} is not, then the last @code{TARGET} |
4043 | command is also used to set the format for the output file. @xref{BFD}. | |
4044 | @end table | |
4045 | @end ifclear | |
4046 | ||
4a93e180 NC |
4047 | @node REGION_ALIAS |
4048 | @subsection Assign alias names to memory regions | |
4049 | @kindex REGION_ALIAS(@var{alias}, @var{region}) | |
4050 | @cindex region alias | |
4051 | @cindex region names | |
4052 | ||
4053 | Alias names can be added to existing memory regions created with the | |
4054 | @ref{MEMORY} command. Each name corresponds to at most one memory region. | |
4055 | ||
4056 | @smallexample | |
4057 | REGION_ALIAS(@var{alias}, @var{region}) | |
4058 | @end smallexample | |
4059 | ||
4060 | The @code{REGION_ALIAS} function creates an alias name @var{alias} for the | |
4061 | memory region @var{region}. This allows a flexible mapping of output sections | |
4062 | to memory regions. An example follows. | |
4063 | ||
4064 | Suppose we have an application for embedded systems which come with various | |
4065 | memory storage devices. All have a general purpose, volatile memory @code{RAM} | |
4066 | that allows code execution or data storage. Some may have a read-only, | |
4067 | non-volatile memory @code{ROM} that allows code execution and read-only data | |
4068 | access. The last variant is a read-only, non-volatile memory @code{ROM2} with | |
4069 | read-only data access and no code execution capability. We have four output | |
4070 | sections: | |
4071 | ||
4072 | @itemize @bullet | |
4073 | @item | |
4074 | @code{.text} program code; | |
4075 | @item | |
4076 | @code{.rodata} read-only data; | |
4077 | @item | |
4078 | @code{.data} read-write initialized data; | |
4079 | @item | |
4080 | @code{.bss} read-write zero initialized data. | |
4081 | @end itemize | |
4082 | ||
4083 | The goal is to provide a linker command file that contains a system independent | |
4084 | part defining the output sections and a system dependent part mapping the | |
4085 | output sections to the memory regions available on the system. Our embedded | |
4086 | systems come with three different memory setups @code{A}, @code{B} and | |
4087 | @code{C}: | |
4088 | @multitable @columnfractions .25 .25 .25 .25 | |
4089 | @item Section @tab Variant A @tab Variant B @tab Variant C | |
9d5777a3 | 4090 | @item .text @tab RAM @tab ROM @tab ROM |
4a93e180 NC |
4091 | @item .rodata @tab RAM @tab ROM @tab ROM2 |
4092 | @item .data @tab RAM @tab RAM/ROM @tab RAM/ROM2 | |
4093 | @item .bss @tab RAM @tab RAM @tab RAM | |
4094 | @end multitable | |
4095 | The notation @code{RAM/ROM} or @code{RAM/ROM2} means that this section is | |
4096 | loaded into region @code{ROM} or @code{ROM2} respectively. Please note that | |
4097 | the load address of the @code{.data} section starts in all three variants at | |
4098 | the end of the @code{.rodata} section. | |
4099 | ||
4100 | The base linker script that deals with the output sections follows. It | |
4101 | includes the system dependent @code{linkcmds.memory} file that describes the | |
4102 | memory layout: | |
4103 | @smallexample | |
4104 | INCLUDE linkcmds.memory | |
4105 | ||
4106 | SECTIONS | |
4107 | @{ | |
4108 | .text : | |
4109 | @{ | |
4110 | *(.text) | |
4111 | @} > REGION_TEXT | |
4112 | .rodata : | |
4113 | @{ | |
4114 | *(.rodata) | |
4115 | rodata_end = .; | |
4116 | @} > REGION_RODATA | |
4117 | .data : AT (rodata_end) | |
4118 | @{ | |
4119 | data_start = .; | |
4120 | *(.data) | |
4121 | @} > REGION_DATA | |
4122 | data_size = SIZEOF(.data); | |
4123 | data_load_start = LOADADDR(.data); | |
4124 | .bss : | |
4125 | @{ | |
4126 | *(.bss) | |
4127 | @} > REGION_BSS | |
4128 | @} | |
4129 | @end smallexample | |
4130 | ||
4131 | Now we need three different @code{linkcmds.memory} files to define memory | |
4132 | regions and alias names. The content of @code{linkcmds.memory} for the three | |
4133 | variants @code{A}, @code{B} and @code{C}: | |
4134 | @table @code | |
4135 | @item A | |
4136 | Here everything goes into the @code{RAM}. | |
4137 | @smallexample | |
4138 | MEMORY | |
4139 | @{ | |
4140 | RAM : ORIGIN = 0, LENGTH = 4M | |
4141 | @} | |
4142 | ||
4143 | REGION_ALIAS("REGION_TEXT", RAM); | |
4144 | REGION_ALIAS("REGION_RODATA", RAM); | |
4145 | REGION_ALIAS("REGION_DATA", RAM); | |
4146 | REGION_ALIAS("REGION_BSS", RAM); | |
4147 | @end smallexample | |
4148 | @item B | |
4149 | Program code and read-only data go into the @code{ROM}. Read-write data goes | |
4150 | into the @code{RAM}. An image of the initialized data is loaded into the | |
4151 | @code{ROM} and will be copied during system start into the @code{RAM}. | |
4152 | @smallexample | |
4153 | MEMORY | |
4154 | @{ | |
4155 | ROM : ORIGIN = 0, LENGTH = 3M | |
4156 | RAM : ORIGIN = 0x10000000, LENGTH = 1M | |
4157 | @} | |
4158 | ||
4159 | REGION_ALIAS("REGION_TEXT", ROM); | |
4160 | REGION_ALIAS("REGION_RODATA", ROM); | |
4161 | REGION_ALIAS("REGION_DATA", RAM); | |
4162 | REGION_ALIAS("REGION_BSS", RAM); | |
4163 | @end smallexample | |
4164 | @item C | |
4165 | Program code goes into the @code{ROM}. Read-only data goes into the | |
4166 | @code{ROM2}. Read-write data goes into the @code{RAM}. An image of the | |
4167 | initialized data is loaded into the @code{ROM2} and will be copied during | |
4168 | system start into the @code{RAM}. | |
4169 | @smallexample | |
4170 | MEMORY | |
4171 | @{ | |
4172 | ROM : ORIGIN = 0, LENGTH = 2M | |
4173 | ROM2 : ORIGIN = 0x10000000, LENGTH = 1M | |
4174 | RAM : ORIGIN = 0x20000000, LENGTH = 1M | |
4175 | @} | |
4176 | ||
4177 | REGION_ALIAS("REGION_TEXT", ROM); | |
4178 | REGION_ALIAS("REGION_RODATA", ROM2); | |
4179 | REGION_ALIAS("REGION_DATA", RAM); | |
4180 | REGION_ALIAS("REGION_BSS", RAM); | |
4181 | @end smallexample | |
4182 | @end table | |
4183 | ||
4184 | It is possible to write a common system initialization routine to copy the | |
4185 | @code{.data} section from @code{ROM} or @code{ROM2} into the @code{RAM} if | |
4186 | necessary: | |
4187 | @smallexample | |
4188 | #include <string.h> | |
4189 | ||
4190 | extern char data_start []; | |
4191 | extern char data_size []; | |
4192 | extern char data_load_start []; | |
4193 | ||
4194 | void copy_data(void) | |
4195 | @{ | |
4196 | if (data_start != data_load_start) | |
4197 | @{ | |
4198 | memcpy(data_start, data_load_start, (size_t) data_size); | |
4199 | @} | |
4200 | @} | |
4201 | @end smallexample | |
4202 | ||
252b5132 | 4203 | @node Miscellaneous Commands |
36f63dca | 4204 | @subsection Other Linker Script Commands |
252b5132 RH |
4205 | There are a few other linker scripts commands. |
4206 | ||
4207 | @table @code | |
4208 | @item ASSERT(@var{exp}, @var{message}) | |
4209 | @kindex ASSERT | |
4210 | @cindex assertion in linker script | |
4211 | Ensure that @var{exp} is non-zero. If it is zero, then exit the linker | |
4212 | with an error code, and print @var{message}. | |
4213 | ||
fd1c4238 NC |
4214 | Note that assertions are checked before the final stages of linking |
4215 | take place. This means that expressions involving symbols PROVIDEd | |
4216 | inside section definitions will fail if the user has not set values | |
4217 | for those symbols. The only exception to this rule is PROVIDEd | |
4218 | symbols that just reference dot. Thus an assertion like this: | |
4219 | ||
4220 | @smallexample | |
4221 | .stack : | |
4222 | @{ | |
4223 | PROVIDE (__stack = .); | |
4224 | PROVIDE (__stack_size = 0x100); | |
4225 | ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack"); | |
4226 | @} | |
4227 | @end smallexample | |
4228 | ||
4229 | will fail if @code{__stack_size} is not defined elsewhere. Symbols | |
4230 | PROVIDEd outside of section definitions are evaluated earlier, so they | |
4231 | can be used inside ASSERTions. Thus: | |
4232 | ||
4233 | @smallexample | |
4234 | PROVIDE (__stack_size = 0x100); | |
4235 | .stack : | |
4236 | @{ | |
4237 | PROVIDE (__stack = .); | |
4238 | ASSERT ((__stack > (_end + __stack_size)), "Error: No room left for the stack"); | |
4239 | @} | |
4240 | @end smallexample | |
4241 | ||
4242 | will work. | |
4243 | ||
252b5132 RH |
4244 | @item EXTERN(@var{symbol} @var{symbol} @dots{}) |
4245 | @kindex EXTERN | |
4246 | @cindex undefined symbol in linker script | |
4247 | Force @var{symbol} to be entered in the output file as an undefined | |
4248 | symbol. Doing this may, for example, trigger linking of additional | |
4249 | modules from standard libraries. You may list several @var{symbol}s for | |
4250 | each @code{EXTERN}, and you may use @code{EXTERN} multiple times. This | |
4251 | command has the same effect as the @samp{-u} command-line option. | |
4252 | ||
4253 | @item FORCE_COMMON_ALLOCATION | |
4254 | @kindex FORCE_COMMON_ALLOCATION | |
4255 | @cindex common allocation in linker script | |
4256 | This command has the same effect as the @samp{-d} command-line option: | |
ff5dcc92 | 4257 | to make @command{ld} assign space to common symbols even if a relocatable |
252b5132 RH |
4258 | output file is specified (@samp{-r}). |
4259 | ||
4818e05f AM |
4260 | @item INHIBIT_COMMON_ALLOCATION |
4261 | @kindex INHIBIT_COMMON_ALLOCATION | |
4262 | @cindex common allocation in linker script | |
4263 | This command has the same effect as the @samp{--no-define-common} | |
4264 | command-line option: to make @code{ld} omit the assignment of addresses | |
4265 | to common symbols even for a non-relocatable output file. | |
4266 | ||
7bdf4127 AB |
4267 | @item FORCE_GROUP_ALLOCATION |
4268 | @kindex FORCE_GROUP_ALLOCATION | |
4269 | @cindex group allocation in linker script | |
4270 | @cindex section groups | |
4271 | @cindex COMDAT | |
4272 | This command has the same effect as the | |
4273 | @samp{--force-group-allocation} command-line option: to make | |
4274 | @command{ld} place section group members like normal input sections, | |
4275 | and to delete the section groups even if a relocatable output file is | |
4276 | specified (@samp{-r}). | |
4277 | ||
53d25da6 AM |
4278 | @item INSERT [ AFTER | BEFORE ] @var{output_section} |
4279 | @kindex INSERT | |
4280 | @cindex insert user script into default script | |
4281 | This command is typically used in a script specified by @samp{-T} to | |
4282 | augment the default @code{SECTIONS} with, for example, overlays. It | |
4283 | inserts all prior linker script statements after (or before) | |
4284 | @var{output_section}, and also causes @samp{-T} to not override the | |
4285 | default linker script. The exact insertion point is as for orphan | |
4286 | sections. @xref{Location Counter}. The insertion happens after the | |
4287 | linker has mapped input sections to output sections. Prior to the | |
4288 | insertion, since @samp{-T} scripts are parsed before the default | |
4289 | linker script, statements in the @samp{-T} script occur before the | |
4290 | default linker script statements in the internal linker representation | |
4291 | of the script. In particular, input section assignments will be made | |
4292 | to @samp{-T} output sections before those in the default script. Here | |
4293 | is an example of how a @samp{-T} script using @code{INSERT} might look: | |
4294 | ||
4295 | @smallexample | |
4296 | SECTIONS | |
4297 | @{ | |
4298 | OVERLAY : | |
4299 | @{ | |
4300 | .ov1 @{ ov1*(.text) @} | |
4301 | .ov2 @{ ov2*(.text) @} | |
4302 | @} | |
4303 | @} | |
4304 | INSERT AFTER .text; | |
4305 | @end smallexample | |
4306 | ||
252b5132 RH |
4307 | @item NOCROSSREFS(@var{section} @var{section} @dots{}) |
4308 | @kindex NOCROSSREFS(@var{sections}) | |
4309 | @cindex cross references | |
ff5dcc92 | 4310 | This command may be used to tell @command{ld} to issue an error about any |
252b5132 RH |
4311 | references among certain output sections. |
4312 | ||
4313 | In certain types of programs, particularly on embedded systems when | |
4314 | using overlays, when one section is loaded into memory, another section | |
4315 | will not be. Any direct references between the two sections would be | |
4316 | errors. For example, it would be an error if code in one section called | |
4317 | a function defined in the other section. | |
4318 | ||
4319 | The @code{NOCROSSREFS} command takes a list of output section names. If | |
ff5dcc92 | 4320 | @command{ld} detects any cross references between the sections, it reports |
252b5132 RH |
4321 | an error and returns a non-zero exit status. Note that the |
4322 | @code{NOCROSSREFS} command uses output section names, not input section | |
4323 | names. | |
4324 | ||
cdf96953 MF |
4325 | @item NOCROSSREFS_TO(@var{tosection} @var{fromsection} @dots{}) |
4326 | @kindex NOCROSSREFS_TO(@var{tosection} @var{fromsections}) | |
4327 | @cindex cross references | |
4328 | This command may be used to tell @command{ld} to issue an error about any | |
4329 | references to one section from a list of other sections. | |
4330 | ||
4331 | The @code{NOCROSSREFS} command is useful when ensuring that two or more | |
4332 | output sections are entirely independent but there are situations where | |
4333 | a one-way dependency is needed. For example, in a multi-core application | |
4334 | there may be shared code that can be called from each core but for safety | |
4335 | must never call back. | |
4336 | ||
4337 | The @code{NOCROSSREFS_TO} command takes a list of output section names. | |
4338 | The first section can not be referenced from any of the other sections. | |
4339 | If @command{ld} detects any references to the first section from any of | |
4340 | the other sections, it reports an error and returns a non-zero exit | |
4341 | status. Note that the @code{NOCROSSREFS_TO} command uses output section | |
4342 | names, not input section names. | |
4343 | ||
252b5132 RH |
4344 | @ifclear SingleFormat |
4345 | @item OUTPUT_ARCH(@var{bfdarch}) | |
4346 | @kindex OUTPUT_ARCH(@var{bfdarch}) | |
4347 | @cindex machine architecture | |
4348 | @cindex architecture | |
4349 | Specify a particular output machine architecture. The argument is one | |
4350 | of the names used by the BFD library (@pxref{BFD}). You can see the | |
4351 | architecture of an object file by using the @code{objdump} program with | |
4352 | the @samp{-f} option. | |
4353 | @end ifclear | |
01554a74 AM |
4354 | |
4355 | @item LD_FEATURE(@var{string}) | |
4356 | @kindex LD_FEATURE(@var{string}) | |
4357 | This command may be used to modify @command{ld} behavior. If | |
4358 | @var{string} is @code{"SANE_EXPR"} then absolute symbols and numbers | |
4359 | in a script are simply treated as numbers everywhere. | |
4360 | @xref{Expression Section}. | |
252b5132 RH |
4361 | @end table |
4362 | ||
4363 | @node Assignments | |
4364 | @section Assigning Values to Symbols | |
4365 | @cindex assignment in scripts | |
4366 | @cindex symbol definition, scripts | |
4367 | @cindex variables, defining | |
4368 | You may assign a value to a symbol in a linker script. This will define | |
73ae6183 | 4369 | the symbol and place it into the symbol table with a global scope. |
252b5132 RH |
4370 | |
4371 | @menu | |
4372 | * Simple Assignments:: Simple Assignments | |
eb8476a6 | 4373 | * HIDDEN:: HIDDEN |
252b5132 | 4374 | * PROVIDE:: PROVIDE |
7af8e998 | 4375 | * PROVIDE_HIDDEN:: PROVIDE_HIDDEN |
73ae6183 | 4376 | * Source Code Reference:: How to use a linker script defined symbol in source code |
252b5132 RH |
4377 | @end menu |
4378 | ||
4379 | @node Simple Assignments | |
4380 | @subsection Simple Assignments | |
4381 | ||
4382 | You may assign to a symbol using any of the C assignment operators: | |
4383 | ||
4384 | @table @code | |
4385 | @item @var{symbol} = @var{expression} ; | |
4386 | @itemx @var{symbol} += @var{expression} ; | |
4387 | @itemx @var{symbol} -= @var{expression} ; | |
4388 | @itemx @var{symbol} *= @var{expression} ; | |
4389 | @itemx @var{symbol} /= @var{expression} ; | |
4390 | @itemx @var{symbol} <<= @var{expression} ; | |
4391 | @itemx @var{symbol} >>= @var{expression} ; | |
4392 | @itemx @var{symbol} &= @var{expression} ; | |
4393 | @itemx @var{symbol} |= @var{expression} ; | |
4394 | @end table | |
4395 | ||
4396 | The first case will define @var{symbol} to the value of | |
4397 | @var{expression}. In the other cases, @var{symbol} must already be | |
4398 | defined, and the value will be adjusted accordingly. | |
4399 | ||
4400 | The special symbol name @samp{.} indicates the location counter. You | |
b5666f2f | 4401 | may only use this within a @code{SECTIONS} command. @xref{Location Counter}. |
252b5132 RH |
4402 | |
4403 | The semicolon after @var{expression} is required. | |
4404 | ||
4405 | Expressions are defined below; see @ref{Expressions}. | |
4406 | ||
4407 | You may write symbol assignments as commands in their own right, or as | |
4408 | statements within a @code{SECTIONS} command, or as part of an output | |
4409 | section description in a @code{SECTIONS} command. | |
4410 | ||
4411 | The section of the symbol will be set from the section of the | |
4412 | expression; for more information, see @ref{Expression Section}. | |
4413 | ||
4414 | Here is an example showing the three different places that symbol | |
4415 | assignments may be used: | |
4416 | ||
4417 | @smallexample | |
4418 | floating_point = 0; | |
4419 | SECTIONS | |
4420 | @{ | |
4421 | .text : | |
4422 | @{ | |
4423 | *(.text) | |
4424 | _etext = .; | |
4425 | @} | |
156e34dd | 4426 | _bdata = (. + 3) & ~ 3; |
252b5132 RH |
4427 | .data : @{ *(.data) @} |
4428 | @} | |
4429 | @end smallexample | |
4430 | @noindent | |
4431 | In this example, the symbol @samp{floating_point} will be defined as | |
4432 | zero. The symbol @samp{_etext} will be defined as the address following | |
4433 | the last @samp{.text} input section. The symbol @samp{_bdata} will be | |
4434 | defined as the address following the @samp{.text} output section aligned | |
4435 | upward to a 4 byte boundary. | |
4436 | ||
eb8476a6 MR |
4437 | @node HIDDEN |
4438 | @subsection HIDDEN | |
4439 | @cindex HIDDEN | |
4440 | For ELF targeted ports, define a symbol that will be hidden and won't be | |
4441 | exported. The syntax is @code{HIDDEN(@var{symbol} = @var{expression})}. | |
4442 | ||
4443 | Here is the example from @ref{Simple Assignments}, rewritten to use | |
4444 | @code{HIDDEN}: | |
4445 | ||
4446 | @smallexample | |
4447 | HIDDEN(floating_point = 0); | |
4448 | SECTIONS | |
4449 | @{ | |
4450 | .text : | |
4451 | @{ | |
4452 | *(.text) | |
4453 | HIDDEN(_etext = .); | |
4454 | @} | |
4455 | HIDDEN(_bdata = (. + 3) & ~ 3); | |
4456 | .data : @{ *(.data) @} | |
4457 | @} | |
4458 | @end smallexample | |
4459 | @noindent | |
4460 | In this case none of the three symbols will be visible outside this module. | |
4461 | ||
252b5132 RH |
4462 | @node PROVIDE |
4463 | @subsection PROVIDE | |
4464 | @cindex PROVIDE | |
4465 | In some cases, it is desirable for a linker script to define a symbol | |
4466 | only if it is referenced and is not defined by any object included in | |
4467 | the link. For example, traditional linkers defined the symbol | |
4468 | @samp{etext}. However, ANSI C requires that the user be able to use | |
4469 | @samp{etext} as a function name without encountering an error. The | |
4470 | @code{PROVIDE} keyword may be used to define a symbol, such as | |
4471 | @samp{etext}, only if it is referenced but not defined. The syntax is | |
4472 | @code{PROVIDE(@var{symbol} = @var{expression})}. | |
4473 | ||
4474 | Here is an example of using @code{PROVIDE} to define @samp{etext}: | |
4475 | @smallexample | |
4476 | SECTIONS | |
4477 | @{ | |
4478 | .text : | |
4479 | @{ | |
4480 | *(.text) | |
4481 | _etext = .; | |
4482 | PROVIDE(etext = .); | |
4483 | @} | |
4484 | @} | |
4485 | @end smallexample | |
4486 | ||
4487 | In this example, if the program defines @samp{_etext} (with a leading | |
3c27360b JB |
4488 | underscore), the linker will give a multiple definition diagnostic. If, |
4489 | on the other hand, the program defines @samp{etext} (with no leading | |
252b5132 RH |
4490 | underscore), the linker will silently use the definition in the program. |
4491 | If the program references @samp{etext} but does not define it, the | |
4492 | linker will use the definition in the linker script. | |
4493 | ||
b0daac83 NC |
4494 | Note - the @code{PROVIDE} directive considers a common symbol to be |
4495 | defined, even though such a symbol could be combined with the symbol | |
4496 | that the @code{PROVIDE} would create. This is particularly important | |
4497 | when considering constructor and destructor list symbols such as | |
4498 | @samp{__CTOR_LIST__} as these are often defined as common symbols. | |
4499 | ||
7af8e998 L |
4500 | @node PROVIDE_HIDDEN |
4501 | @subsection PROVIDE_HIDDEN | |
4502 | @cindex PROVIDE_HIDDEN | |
4503 | Similar to @code{PROVIDE}. For ELF targeted ports, the symbol will be | |
4504 | hidden and won't be exported. | |
4505 | ||
73ae6183 NC |
4506 | @node Source Code Reference |
4507 | @subsection Source Code Reference | |
4508 | ||
4509 | Accessing a linker script defined variable from source code is not | |
4510 | intuitive. In particular a linker script symbol is not equivalent to | |
4511 | a variable declaration in a high level language, it is instead a | |
4512 | symbol that does not have a value. | |
4513 | ||
4514 | Before going further, it is important to note that compilers often | |
4515 | transform names in the source code into different names when they are | |
4516 | stored in the symbol table. For example, Fortran compilers commonly | |
4517 | prepend or append an underscore, and C++ performs extensive @samp{name | |
4518 | mangling}. Therefore there might be a discrepancy between the name | |
4519 | of a variable as it is used in source code and the name of the same | |
4520 | variable as it is defined in a linker script. For example in C a | |
4521 | linker script variable might be referred to as: | |
4522 | ||
4523 | @smallexample | |
4524 | extern int foo; | |
4525 | @end smallexample | |
4526 | ||
4527 | But in the linker script it might be defined as: | |
4528 | ||
4529 | @smallexample | |
4530 | _foo = 1000; | |
4531 | @end smallexample | |
4532 | ||
4533 | In the remaining examples however it is assumed that no name | |
4534 | transformation has taken place. | |
4535 | ||
4536 | When a symbol is declared in a high level language such as C, two | |
4537 | things happen. The first is that the compiler reserves enough space | |
4538 | in the program's memory to hold the @emph{value} of the symbol. The | |
4539 | second is that the compiler creates an entry in the program's symbol | |
4540 | table which holds the symbol's @emph{address}. ie the symbol table | |
4541 | contains the address of the block of memory holding the symbol's | |
4542 | value. So for example the following C declaration, at file scope: | |
4543 | ||
4544 | @smallexample | |
4545 | int foo = 1000; | |
4546 | @end smallexample | |
4547 | ||
10bf6894 | 4548 | creates an entry called @samp{foo} in the symbol table. This entry |
73ae6183 NC |
4549 | holds the address of an @samp{int} sized block of memory where the |
4550 | number 1000 is initially stored. | |
4551 | ||
4552 | When a program references a symbol the compiler generates code that | |
4553 | first accesses the symbol table to find the address of the symbol's | |
4554 | memory block and then code to read the value from that memory block. | |
4555 | So: | |
4556 | ||
4557 | @smallexample | |
4558 | foo = 1; | |
4559 | @end smallexample | |
4560 | ||
4561 | looks up the symbol @samp{foo} in the symbol table, gets the address | |
4562 | associated with this symbol and then writes the value 1 into that | |
4563 | address. Whereas: | |
4564 | ||
4565 | @smallexample | |
4566 | int * a = & foo; | |
4567 | @end smallexample | |
4568 | ||
10bf6894 | 4569 | looks up the symbol @samp{foo} in the symbol table, gets its address |
73ae6183 NC |
4570 | and then copies this address into the block of memory associated with |
4571 | the variable @samp{a}. | |
4572 | ||
4573 | Linker scripts symbol declarations, by contrast, create an entry in | |
4574 | the symbol table but do not assign any memory to them. Thus they are | |
4575 | an address without a value. So for example the linker script definition: | |
4576 | ||
4577 | @smallexample | |
4578 | foo = 1000; | |
4579 | @end smallexample | |
4580 | ||
4581 | creates an entry in the symbol table called @samp{foo} which holds | |
4582 | the address of memory location 1000, but nothing special is stored at | |
4583 | address 1000. This means that you cannot access the @emph{value} of a | |
4584 | linker script defined symbol - it has no value - all you can do is | |
4585 | access the @emph{address} of a linker script defined symbol. | |
4586 | ||
4587 | Hence when you are using a linker script defined symbol in source code | |
4588 | you should always take the address of the symbol, and never attempt to | |
4589 | use its value. For example suppose you want to copy the contents of a | |
4590 | section of memory called .ROM into a section called .FLASH and the | |
4591 | linker script contains these declarations: | |
4592 | ||
4593 | @smallexample | |
4594 | @group | |
4595 | start_of_ROM = .ROM; | |
a5e406b5 | 4596 | end_of_ROM = .ROM + sizeof (.ROM); |
73ae6183 NC |
4597 | start_of_FLASH = .FLASH; |
4598 | @end group | |
4599 | @end smallexample | |
4600 | ||
4601 | Then the C source code to perform the copy would be: | |
4602 | ||
4603 | @smallexample | |
4604 | @group | |
4605 | extern char start_of_ROM, end_of_ROM, start_of_FLASH; | |
c0065db7 | 4606 | |
73ae6183 NC |
4607 | memcpy (& start_of_FLASH, & start_of_ROM, & end_of_ROM - & start_of_ROM); |
4608 | @end group | |
4609 | @end smallexample | |
4610 | ||
4611 | Note the use of the @samp{&} operators. These are correct. | |
5707d2ad NC |
4612 | Alternatively the symbols can be treated as the names of vectors or |
4613 | arrays and then the code will again work as expected: | |
4614 | ||
4615 | @smallexample | |
4616 | @group | |
4617 | extern char start_of_ROM[], end_of_ROM[], start_of_FLASH[]; | |
4618 | ||
4619 | memcpy (start_of_FLASH, start_of_ROM, end_of_ROM - start_of_ROM); | |
4620 | @end group | |
4621 | @end smallexample | |
4622 | ||
4623 | Note how using this method does not require the use of @samp{&} | |
4624 | operators. | |
73ae6183 | 4625 | |
252b5132 | 4626 | @node SECTIONS |
36f63dca | 4627 | @section SECTIONS Command |
252b5132 RH |
4628 | @kindex SECTIONS |
4629 | The @code{SECTIONS} command tells the linker how to map input sections | |
4630 | into output sections, and how to place the output sections in memory. | |
4631 | ||
4632 | The format of the @code{SECTIONS} command is: | |
4633 | @smallexample | |
4634 | SECTIONS | |
4635 | @{ | |
4636 | @var{sections-command} | |
4637 | @var{sections-command} | |
4638 | @dots{} | |
4639 | @} | |
4640 | @end smallexample | |
4641 | ||
4642 | Each @var{sections-command} may of be one of the following: | |
4643 | ||
4644 | @itemize @bullet | |
4645 | @item | |
4646 | an @code{ENTRY} command (@pxref{Entry Point,,Entry command}) | |
4647 | @item | |
4648 | a symbol assignment (@pxref{Assignments}) | |
4649 | @item | |
4650 | an output section description | |
4651 | @item | |
4652 | an overlay description | |
4653 | @end itemize | |
4654 | ||
4655 | The @code{ENTRY} command and symbol assignments are permitted inside the | |
4656 | @code{SECTIONS} command for convenience in using the location counter in | |
4657 | those commands. This can also make the linker script easier to | |
4658 | understand because you can use those commands at meaningful points in | |
4659 | the layout of the output file. | |
4660 | ||
4661 | Output section descriptions and overlay descriptions are described | |
4662 | below. | |
4663 | ||
4664 | If you do not use a @code{SECTIONS} command in your linker script, the | |
4665 | linker will place each input section into an identically named output | |
4666 | section in the order that the sections are first encountered in the | |
4667 | input files. If all input sections are present in the first file, for | |
4668 | example, the order of sections in the output file will match the order | |
4669 | in the first input file. The first section will be at address zero. | |
4670 | ||
4671 | @menu | |
4672 | * Output Section Description:: Output section description | |
4673 | * Output Section Name:: Output section name | |
4674 | * Output Section Address:: Output section address | |
4675 | * Input Section:: Input section description | |
4676 | * Output Section Data:: Output section data | |
4677 | * Output Section Keywords:: Output section keywords | |
4678 | * Output Section Discarding:: Output section discarding | |
4679 | * Output Section Attributes:: Output section attributes | |
4680 | * Overlay Description:: Overlay description | |
4681 | @end menu | |
4682 | ||
4683 | @node Output Section Description | |
36f63dca | 4684 | @subsection Output Section Description |
252b5132 RH |
4685 | The full description of an output section looks like this: |
4686 | @smallexample | |
a1ab1d2a | 4687 | @group |
7e7d5768 | 4688 | @var{section} [@var{address}] [(@var{type})] : |
0c71d759 | 4689 | [AT(@var{lma})] |
1eec346e | 4690 | [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT] |
0c71d759 NC |
4691 | [SUBALIGN(@var{subsection_align})] |
4692 | [@var{constraint}] | |
252b5132 RH |
4693 | @{ |
4694 | @var{output-section-command} | |
4695 | @var{output-section-command} | |
4696 | @dots{} | |
abc9061b | 4697 | @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}] [,] |
252b5132 RH |
4698 | @end group |
4699 | @end smallexample | |
4700 | ||
4701 | Most output sections do not use most of the optional section attributes. | |
4702 | ||
4703 | The whitespace around @var{section} is required, so that the section | |
4704 | name is unambiguous. The colon and the curly braces are also required. | |
abc9061b CC |
4705 | The comma at the end may be required if a @var{fillexp} is used and |
4706 | the next @var{sections-command} looks like a continuation of the expression. | |
252b5132 RH |
4707 | The line breaks and other white space are optional. |
4708 | ||
4709 | Each @var{output-section-command} may be one of the following: | |
4710 | ||
4711 | @itemize @bullet | |
4712 | @item | |
4713 | a symbol assignment (@pxref{Assignments}) | |
4714 | @item | |
4715 | an input section description (@pxref{Input Section}) | |
4716 | @item | |
4717 | data values to include directly (@pxref{Output Section Data}) | |
4718 | @item | |
4719 | a special output section keyword (@pxref{Output Section Keywords}) | |
4720 | @end itemize | |
4721 | ||
4722 | @node Output Section Name | |
36f63dca | 4723 | @subsection Output Section Name |
252b5132 RH |
4724 | @cindex name, section |
4725 | @cindex section name | |
4726 | The name of the output section is @var{section}. @var{section} must | |
4727 | meet the constraints of your output format. In formats which only | |
4728 | support a limited number of sections, such as @code{a.out}, the name | |
4729 | must be one of the names supported by the format (@code{a.out}, for | |
4730 | example, allows only @samp{.text}, @samp{.data} or @samp{.bss}). If the | |
4731 | output format supports any number of sections, but with numbers and not | |
4732 | names (as is the case for Oasys), the name should be supplied as a | |
4733 | quoted numeric string. A section name may consist of any sequence of | |
4734 | characters, but a name which contains any unusual characters such as | |
4735 | commas must be quoted. | |
4736 | ||
4737 | The output section name @samp{/DISCARD/} is special; @ref{Output Section | |
4738 | Discarding}. | |
4739 | ||
4740 | @node Output Section Address | |
2a16d82a | 4741 | @subsection Output Section Address |
252b5132 RH |
4742 | @cindex address, section |
4743 | @cindex section address | |
4744 | The @var{address} is an expression for the VMA (the virtual memory | |
ea5cae92 NC |
4745 | address) of the output section. This address is optional, but if it |
4746 | is provided then the output address will be set exactly as specified. | |
4747 | ||
4748 | If the output address is not specified then one will be chosen for the | |
4749 | section, based on the heuristic below. This address will be adjusted | |
4750 | to fit the alignment requirement of the output section. The | |
4751 | alignment requirement is the strictest alignment of any input section | |
4752 | contained within the output section. | |
4753 | ||
4754 | The output section address heuristic is as follows: | |
4755 | ||
4756 | @itemize @bullet | |
4757 | @item | |
4758 | If an output memory @var{region} is set for the section then it | |
4759 | is added to this region and its address will be the next free address | |
4760 | in that region. | |
4761 | ||
4762 | @item | |
4763 | If the MEMORY command has been used to create a list of memory | |
4764 | regions then the first region which has attributes compatible with the | |
4765 | section is selected to contain it. The section's output address will | |
4766 | be the next free address in that region; @ref{MEMORY}. | |
4767 | ||
4768 | @item | |
4769 | If no memory regions were specified, or none match the section then | |
4770 | the output address will be based on the current value of the location | |
4771 | counter. | |
4772 | @end itemize | |
4773 | ||
4774 | @noindent | |
4775 | For example: | |
4776 | ||
252b5132 RH |
4777 | @smallexample |
4778 | .text . : @{ *(.text) @} | |
4779 | @end smallexample | |
ea5cae92 | 4780 | |
252b5132 RH |
4781 | @noindent |
4782 | and | |
ea5cae92 | 4783 | |
252b5132 RH |
4784 | @smallexample |
4785 | .text : @{ *(.text) @} | |
4786 | @end smallexample | |
ea5cae92 | 4787 | |
252b5132 RH |
4788 | @noindent |
4789 | are subtly different. The first will set the address of the | |
4790 | @samp{.text} output section to the current value of the location | |
4791 | counter. The second will set it to the current value of the location | |
ea5cae92 NC |
4792 | counter aligned to the strictest alignment of any of the @samp{.text} |
4793 | input sections. | |
252b5132 RH |
4794 | |
4795 | The @var{address} may be an arbitrary expression; @ref{Expressions}. | |
4796 | For example, if you want to align the section on a 0x10 byte boundary, | |
4797 | so that the lowest four bits of the section address are zero, you could | |
4798 | do something like this: | |
4799 | @smallexample | |
4800 | .text ALIGN(0x10) : @{ *(.text) @} | |
4801 | @end smallexample | |
4802 | @noindent | |
4803 | This works because @code{ALIGN} returns the current location counter | |
4804 | aligned upward to the specified value. | |
4805 | ||
4806 | Specifying @var{address} for a section will change the value of the | |
6ce340f1 NC |
4807 | location counter, provided that the section is non-empty. (Empty |
4808 | sections are ignored). | |
252b5132 RH |
4809 | |
4810 | @node Input Section | |
36f63dca | 4811 | @subsection Input Section Description |
252b5132 RH |
4812 | @cindex input sections |
4813 | @cindex mapping input sections to output sections | |
4814 | The most common output section command is an input section description. | |
4815 | ||
4816 | The input section description is the most basic linker script operation. | |
4817 | You use output sections to tell the linker how to lay out your program | |
4818 | in memory. You use input section descriptions to tell the linker how to | |
4819 | map the input files into your memory layout. | |
4820 | ||
4821 | @menu | |
4822 | * Input Section Basics:: Input section basics | |
4823 | * Input Section Wildcards:: Input section wildcard patterns | |
4824 | * Input Section Common:: Input section for common symbols | |
4825 | * Input Section Keep:: Input section and garbage collection | |
4826 | * Input Section Example:: Input section example | |
4827 | @end menu | |
4828 | ||
4829 | @node Input Section Basics | |
36f63dca | 4830 | @subsubsection Input Section Basics |
252b5132 RH |
4831 | @cindex input section basics |
4832 | An input section description consists of a file name optionally followed | |
4833 | by a list of section names in parentheses. | |
4834 | ||
4835 | The file name and the section name may be wildcard patterns, which we | |
4836 | describe further below (@pxref{Input Section Wildcards}). | |
4837 | ||
4838 | The most common input section description is to include all input | |
4839 | sections with a particular name in the output section. For example, to | |
4840 | include all input @samp{.text} sections, you would write: | |
4841 | @smallexample | |
4842 | *(.text) | |
4843 | @end smallexample | |
4844 | @noindent | |
18625d54 | 4845 | Here the @samp{*} is a wildcard which matches any file name. To exclude a list |
8f1732fc | 4846 | @cindex EXCLUDE_FILE |
18625d54 CM |
4847 | of files from matching the file name wildcard, EXCLUDE_FILE may be used to |
4848 | match all files except the ones specified in the EXCLUDE_FILE list. For | |
4849 | example: | |
252b5132 | 4850 | @smallexample |
8f1732fc AB |
4851 | EXCLUDE_FILE (*crtend.o *otherfile.o) *(.ctors) |
4852 | @end smallexample | |
4853 | @noindent | |
4854 | will cause all .ctors sections from all files except @file{crtend.o} | |
4855 | and @file{otherfile.o} to be included. The EXCLUDE_FILE can also be | |
4856 | placed inside the section list, for example: | |
4857 | @smallexample | |
b4346c09 | 4858 | *(EXCLUDE_FILE (*crtend.o *otherfile.o) .ctors) |
252b5132 | 4859 | @end smallexample |
8f1732fc AB |
4860 | @noindent |
4861 | The result of this is identically to the previous example. Supporting | |
4862 | two syntaxes for EXCLUDE_FILE is useful if the section list contains | |
4863 | more than one section, as described below. | |
252b5132 RH |
4864 | |
4865 | There are two ways to include more than one section: | |
4866 | @smallexample | |
4867 | *(.text .rdata) | |
4868 | *(.text) *(.rdata) | |
4869 | @end smallexample | |
4870 | @noindent | |
4871 | The difference between these is the order in which the @samp{.text} and | |
4872 | @samp{.rdata} input sections will appear in the output section. In the | |
b6bf44ba AM |
4873 | first example, they will be intermingled, appearing in the same order as |
4874 | they are found in the linker input. In the second example, all | |
252b5132 RH |
4875 | @samp{.text} input sections will appear first, followed by all |
4876 | @samp{.rdata} input sections. | |
4877 | ||
8f1732fc AB |
4878 | When using EXCLUDE_FILE with more than one section, if the exclusion |
4879 | is within the section list then the exclusion only applies to the | |
4880 | immediately following section, for example: | |
a5bf7d4f AB |
4881 | @smallexample |
4882 | *(EXCLUDE_FILE (*somefile.o) .text .rdata) | |
4883 | @end smallexample | |
4884 | @noindent | |
4885 | will cause all @samp{.text} sections from all files except | |
4886 | @file{somefile.o} to be included, while all @samp{.rdata} sections | |
4887 | from all files, including @file{somefile.o}, will be included. To | |
4888 | exclude the @samp{.rdata} sections from @file{somefile.o} the example | |
8f1732fc | 4889 | could be modified to: |
a5bf7d4f AB |
4890 | @smallexample |
4891 | *(EXCLUDE_FILE (*somefile.o) .text EXCLUDE_FILE (*somefile.o) .rdata) | |
4892 | @end smallexample | |
8f1732fc AB |
4893 | @noindent |
4894 | Alternatively, placing the EXCLUDE_FILE outside of the section list, | |
4895 | before the input file selection, will cause the exclusion to apply for | |
4896 | all sections. Thus the previous example can be rewritten as: | |
4897 | @smallexample | |
4898 | EXCLUDE_FILE (*somefile.o) *(.text .rdata) | |
4899 | @end smallexample | |
a5bf7d4f | 4900 | |
252b5132 RH |
4901 | You can specify a file name to include sections from a particular file. |
4902 | You would do this if one or more of your files contain special data that | |
4903 | needs to be at a particular location in memory. For example: | |
4904 | @smallexample | |
4905 | data.o(.data) | |
4906 | @end smallexample | |
4907 | ||
ae17ab41 CM |
4908 | To refine the sections that are included based on the section flags |
4909 | of an input section, INPUT_SECTION_FLAGS may be used. | |
4910 | ||
4911 | Here is a simple example for using Section header flags for ELF sections: | |
4912 | ||
4913 | @smallexample | |
4914 | @group | |
4915 | SECTIONS @{ | |
4916 | .text : @{ INPUT_SECTION_FLAGS (SHF_MERGE & SHF_STRINGS) *(.text) @} | |
4917 | .text2 : @{ INPUT_SECTION_FLAGS (!SHF_WRITE) *(.text) @} | |
4918 | @} | |
4919 | @end group | |
4920 | @end smallexample | |
4921 | ||
4922 | In this example, the output section @samp{.text} will be comprised of any | |
4923 | input section matching the name *(.text) whose section header flags | |
4924 | @code{SHF_MERGE} and @code{SHF_STRINGS} are set. The output section | |
4925 | @samp{.text2} will be comprised of any input section matching the name *(.text) | |
4926 | whose section header flag @code{SHF_WRITE} is clear. | |
4927 | ||
967928e9 AM |
4928 | You can also specify files within archives by writing a pattern |
4929 | matching the archive, a colon, then the pattern matching the file, | |
4930 | with no whitespace around the colon. | |
4931 | ||
4932 | @table @samp | |
4933 | @item archive:file | |
4934 | matches file within archive | |
4935 | @item archive: | |
4936 | matches the whole archive | |
4937 | @item :file | |
4938 | matches file but not one in an archive | |
4939 | @end table | |
4940 | ||
4941 | Either one or both of @samp{archive} and @samp{file} can contain shell | |
4942 | wildcards. On DOS based file systems, the linker will assume that a | |
4943 | single letter followed by a colon is a drive specifier, so | |
4944 | @samp{c:myfile.o} is a simple file specification, not @samp{myfile.o} | |
4945 | within an archive called @samp{c}. @samp{archive:file} filespecs may | |
4946 | also be used within an @code{EXCLUDE_FILE} list, but may not appear in | |
4947 | other linker script contexts. For instance, you cannot extract a file | |
4948 | from an archive by using @samp{archive:file} in an @code{INPUT} | |
4949 | command. | |
4950 | ||
252b5132 RH |
4951 | If you use a file name without a list of sections, then all sections in |
4952 | the input file will be included in the output section. This is not | |
4953 | commonly done, but it may by useful on occasion. For example: | |
4954 | @smallexample | |
4955 | data.o | |
4956 | @end smallexample | |
4957 | ||
967928e9 AM |
4958 | When you use a file name which is not an @samp{archive:file} specifier |
4959 | and does not contain any wild card | |
252b5132 RH |
4960 | characters, the linker will first see if you also specified the file |
4961 | name on the linker command line or in an @code{INPUT} command. If you | |
4962 | did not, the linker will attempt to open the file as an input file, as | |
4963 | though it appeared on the command line. Note that this differs from an | |
4964 | @code{INPUT} command, because the linker will not search for the file in | |
4965 | the archive search path. | |
4966 | ||
4967 | @node Input Section Wildcards | |
36f63dca | 4968 | @subsubsection Input Section Wildcard Patterns |
252b5132 RH |
4969 | @cindex input section wildcards |
4970 | @cindex wildcard file name patterns | |
4971 | @cindex file name wildcard patterns | |
4972 | @cindex section name wildcard patterns | |
4973 | In an input section description, either the file name or the section | |
4974 | name or both may be wildcard patterns. | |
4975 | ||
4976 | The file name of @samp{*} seen in many examples is a simple wildcard | |
4977 | pattern for the file name. | |
4978 | ||
4979 | The wildcard patterns are like those used by the Unix shell. | |
4980 | ||
4981 | @table @samp | |
4982 | @item * | |
4983 | matches any number of characters | |
4984 | @item ? | |
4985 | matches any single character | |
4986 | @item [@var{chars}] | |
4987 | matches a single instance of any of the @var{chars}; the @samp{-} | |
4988 | character may be used to specify a range of characters, as in | |
4989 | @samp{[a-z]} to match any lower case letter | |
4990 | @item \ | |
4991 | quotes the following character | |
4992 | @end table | |
4993 | ||
4994 | When a file name is matched with a wildcard, the wildcard characters | |
4995 | will not match a @samp{/} character (used to separate directory names on | |
4996 | Unix). A pattern consisting of a single @samp{*} character is an | |
4997 | exception; it will always match any file name, whether it contains a | |
4998 | @samp{/} or not. In a section name, the wildcard characters will match | |
4999 | a @samp{/} character. | |
5000 | ||
5001 | File name wildcard patterns only match files which are explicitly | |
5002 | specified on the command line or in an @code{INPUT} command. The linker | |
5003 | does not search directories to expand wildcards. | |
5004 | ||
5005 | If a file name matches more than one wildcard pattern, or if a file name | |
5006 | appears explicitly and is also matched by a wildcard pattern, the linker | |
5007 | will use the first match in the linker script. For example, this | |
5008 | sequence of input section descriptions is probably in error, because the | |
5009 | @file{data.o} rule will not be used: | |
5010 | @smallexample | |
5011 | .data : @{ *(.data) @} | |
5012 | .data1 : @{ data.o(.data) @} | |
5013 | @end smallexample | |
5014 | ||
bcaa7b3e | 5015 | @cindex SORT_BY_NAME |
252b5132 RH |
5016 | Normally, the linker will place files and sections matched by wildcards |
5017 | in the order in which they are seen during the link. You can change | |
bcaa7b3e L |
5018 | this by using the @code{SORT_BY_NAME} keyword, which appears before a wildcard |
5019 | pattern in parentheses (e.g., @code{SORT_BY_NAME(.text*)}). When the | |
5020 | @code{SORT_BY_NAME} keyword is used, the linker will sort the files or sections | |
252b5132 RH |
5021 | into ascending order by name before placing them in the output file. |
5022 | ||
bcaa7b3e | 5023 | @cindex SORT_BY_ALIGNMENT |
9a24a276 AM |
5024 | @code{SORT_BY_ALIGNMENT} is similar to @code{SORT_BY_NAME}. |
5025 | @code{SORT_BY_ALIGNMENT} will sort sections into descending order of | |
5026 | alignment before placing them in the output file. Placing larger | |
5027 | alignments before smaller alignments can reduce the amount of padding | |
5028 | needed. | |
bcaa7b3e | 5029 | |
02ecc8e9 | 5030 | @cindex SORT_BY_INIT_PRIORITY |
9a24a276 AM |
5031 | @code{SORT_BY_INIT_PRIORITY} is also similar to @code{SORT_BY_NAME}. |
5032 | @code{SORT_BY_INIT_PRIORITY} will sort sections into ascending | |
5033 | numerical order of the GCC init_priority attribute encoded in the | |
5034 | section name before placing them in the output file. In | |
5035 | @code{.init_array.NNNNN} and @code{.fini_array.NNNNN}, @code{NNNNN} is | |
5036 | the init_priority. In @code{.ctors.NNNNN} and @code{.dtors.NNNNN}, | |
5037 | @code{NNNNN} is 65535 minus the init_priority. | |
02ecc8e9 | 5038 | |
bcaa7b3e L |
5039 | @cindex SORT |
5040 | @code{SORT} is an alias for @code{SORT_BY_NAME}. | |
5041 | ||
5042 | When there are nested section sorting commands in linker script, there | |
5043 | can be at most 1 level of nesting for section sorting commands. | |
5044 | ||
5045 | @enumerate | |
5046 | @item | |
5047 | @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)). | |
1ae5c3ae | 5048 | It will sort the input sections by name first, then by alignment if two |
bcaa7b3e L |
5049 | sections have the same name. |
5050 | @item | |
5051 | @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)). | |
1ae5c3ae | 5052 | It will sort the input sections by alignment first, then by name if two |
bcaa7b3e L |
5053 | sections have the same alignment. |
5054 | @item | |
c0065db7 | 5055 | @code{SORT_BY_NAME} (@code{SORT_BY_NAME} (wildcard section pattern)) is |
bcaa7b3e L |
5056 | treated the same as @code{SORT_BY_NAME} (wildcard section pattern). |
5057 | @item | |
5058 | @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)) | |
5059 | is treated the same as @code{SORT_BY_ALIGNMENT} (wildcard section pattern). | |
5060 | @item | |
5061 | All other nested section sorting commands are invalid. | |
5062 | @end enumerate | |
5063 | ||
a05a5b64 | 5064 | When both command-line section sorting option and linker script |
bcaa7b3e | 5065 | section sorting command are used, section sorting command always |
a05a5b64 | 5066 | takes precedence over the command-line option. |
bcaa7b3e L |
5067 | |
5068 | If the section sorting command in linker script isn't nested, the | |
a05a5b64 | 5069 | command-line option will make the section sorting command to be |
bcaa7b3e L |
5070 | treated as nested sorting command. |
5071 | ||
5072 | @enumerate | |
5073 | @item | |
5074 | @code{SORT_BY_NAME} (wildcard section pattern ) with | |
5075 | @option{--sort-sections alignment} is equivalent to | |
5076 | @code{SORT_BY_NAME} (@code{SORT_BY_ALIGNMENT} (wildcard section pattern)). | |
5077 | @item | |
5078 | @code{SORT_BY_ALIGNMENT} (wildcard section pattern) with | |
5079 | @option{--sort-section name} is equivalent to | |
5080 | @code{SORT_BY_ALIGNMENT} (@code{SORT_BY_NAME} (wildcard section pattern)). | |
5081 | @end enumerate | |
5082 | ||
5083 | If the section sorting command in linker script is nested, the | |
a05a5b64 | 5084 | command-line option will be ignored. |
bcaa7b3e | 5085 | |
eda680f8 | 5086 | @cindex SORT_NONE |
a05a5b64 | 5087 | @code{SORT_NONE} disables section sorting by ignoring the command-line |
eda680f8 L |
5088 | section sorting option. |
5089 | ||
252b5132 RH |
5090 | If you ever get confused about where input sections are going, use the |
5091 | @samp{-M} linker option to generate a map file. The map file shows | |
5092 | precisely how input sections are mapped to output sections. | |
5093 | ||
5094 | This example shows how wildcard patterns might be used to partition | |
5095 | files. This linker script directs the linker to place all @samp{.text} | |
5096 | sections in @samp{.text} and all @samp{.bss} sections in @samp{.bss}. | |
5097 | The linker will place the @samp{.data} section from all files beginning | |
5098 | with an upper case character in @samp{.DATA}; for all other files, the | |
5099 | linker will place the @samp{.data} section in @samp{.data}. | |
5100 | @smallexample | |
5101 | @group | |
5102 | SECTIONS @{ | |
5103 | .text : @{ *(.text) @} | |
5104 | .DATA : @{ [A-Z]*(.data) @} | |
5105 | .data : @{ *(.data) @} | |
5106 | .bss : @{ *(.bss) @} | |
5107 | @} | |
5108 | @end group | |
5109 | @end smallexample | |
5110 | ||
5111 | @node Input Section Common | |
36f63dca | 5112 | @subsubsection Input Section for Common Symbols |
252b5132 RH |
5113 | @cindex common symbol placement |
5114 | @cindex uninitialized data placement | |
5115 | A special notation is needed for common symbols, because in many object | |
5116 | file formats common symbols do not have a particular input section. The | |
5117 | linker treats common symbols as though they are in an input section | |
5118 | named @samp{COMMON}. | |
5119 | ||
5120 | You may use file names with the @samp{COMMON} section just as with any | |
5121 | other input sections. You can use this to place common symbols from a | |
5122 | particular input file in one section while common symbols from other | |
5123 | input files are placed in another section. | |
5124 | ||
5125 | In most cases, common symbols in input files will be placed in the | |
5126 | @samp{.bss} section in the output file. For example: | |
5127 | @smallexample | |
5128 | .bss @{ *(.bss) *(COMMON) @} | |
5129 | @end smallexample | |
5130 | ||
5131 | @cindex scommon section | |
5132 | @cindex small common symbols | |
5133 | Some object file formats have more than one type of common symbol. For | |
5134 | example, the MIPS ELF object file format distinguishes standard common | |
5135 | symbols and small common symbols. In this case, the linker will use a | |
5136 | different special section name for other types of common symbols. In | |
5137 | the case of MIPS ELF, the linker uses @samp{COMMON} for standard common | |
5138 | symbols and @samp{.scommon} for small common symbols. This permits you | |
5139 | to map the different types of common symbols into memory at different | |
5140 | locations. | |
5141 | ||
5142 | @cindex [COMMON] | |
5143 | You will sometimes see @samp{[COMMON]} in old linker scripts. This | |
5144 | notation is now considered obsolete. It is equivalent to | |
5145 | @samp{*(COMMON)}. | |
5146 | ||
5147 | @node Input Section Keep | |
36f63dca | 5148 | @subsubsection Input Section and Garbage Collection |
252b5132 RH |
5149 | @cindex KEEP |
5150 | @cindex garbage collection | |
5151 | When link-time garbage collection is in use (@samp{--gc-sections}), | |
a1ab1d2a | 5152 | it is often useful to mark sections that should not be eliminated. |
252b5132 RH |
5153 | This is accomplished by surrounding an input section's wildcard entry |
5154 | with @code{KEEP()}, as in @code{KEEP(*(.init))} or | |
bcaa7b3e | 5155 | @code{KEEP(SORT_BY_NAME(*)(.ctors))}. |
252b5132 RH |
5156 | |
5157 | @node Input Section Example | |
36f63dca | 5158 | @subsubsection Input Section Example |
252b5132 RH |
5159 | The following example is a complete linker script. It tells the linker |
5160 | to read all of the sections from file @file{all.o} and place them at the | |
5161 | start of output section @samp{outputa} which starts at location | |
5162 | @samp{0x10000}. All of section @samp{.input1} from file @file{foo.o} | |
5163 | follows immediately, in the same output section. All of section | |
5164 | @samp{.input2} from @file{foo.o} goes into output section | |
5165 | @samp{outputb}, followed by section @samp{.input1} from @file{foo1.o}. | |
5166 | All of the remaining @samp{.input1} and @samp{.input2} sections from any | |
5167 | files are written to output section @samp{outputc}. | |
5168 | ||
5169 | @smallexample | |
5170 | @group | |
5171 | SECTIONS @{ | |
5172 | outputa 0x10000 : | |
5173 | @{ | |
5174 | all.o | |
5175 | foo.o (.input1) | |
5176 | @} | |
36f63dca NC |
5177 | @end group |
5178 | @group | |
252b5132 RH |
5179 | outputb : |
5180 | @{ | |
5181 | foo.o (.input2) | |
5182 | foo1.o (.input1) | |
5183 | @} | |
36f63dca NC |
5184 | @end group |
5185 | @group | |
252b5132 RH |
5186 | outputc : |
5187 | @{ | |
5188 | *(.input1) | |
5189 | *(.input2) | |
5190 | @} | |
5191 | @} | |
5192 | @end group | |
a1ab1d2a | 5193 | @end smallexample |
252b5132 | 5194 | |
cbd0eecf L |
5195 | If an output section's name is the same as the input section's name |
5196 | and is representable as a C identifier, then the linker will | |
5197 | automatically @pxref{PROVIDE} two symbols: __start_SECNAME and | |
5198 | __stop_SECNAME, where SECNAME is the name of the section. These | |
5199 | indicate the start address and end address of the output section | |
5200 | respectively. Note: most section names are not representable as | |
5201 | C identifiers because they contain a @samp{.} character. | |
5202 | ||
252b5132 | 5203 | @node Output Section Data |
36f63dca | 5204 | @subsection Output Section Data |
252b5132 RH |
5205 | @cindex data |
5206 | @cindex section data | |
5207 | @cindex output section data | |
5208 | @kindex BYTE(@var{expression}) | |
5209 | @kindex SHORT(@var{expression}) | |
5210 | @kindex LONG(@var{expression}) | |
5211 | @kindex QUAD(@var{expression}) | |
5212 | @kindex SQUAD(@var{expression}) | |
5213 | You can include explicit bytes of data in an output section by using | |
5214 | @code{BYTE}, @code{SHORT}, @code{LONG}, @code{QUAD}, or @code{SQUAD} as | |
5215 | an output section command. Each keyword is followed by an expression in | |
5216 | parentheses providing the value to store (@pxref{Expressions}). The | |
5217 | value of the expression is stored at the current value of the location | |
5218 | counter. | |
5219 | ||
5220 | The @code{BYTE}, @code{SHORT}, @code{LONG}, and @code{QUAD} commands | |
5221 | store one, two, four, and eight bytes (respectively). After storing the | |
5222 | bytes, the location counter is incremented by the number of bytes | |
5223 | stored. | |
5224 | ||
5225 | For example, this will store the byte 1 followed by the four byte value | |
5226 | of the symbol @samp{addr}: | |
5227 | @smallexample | |
5228 | BYTE(1) | |
5229 | LONG(addr) | |
5230 | @end smallexample | |
5231 | ||
5232 | When using a 64 bit host or target, @code{QUAD} and @code{SQUAD} are the | |
5233 | same; they both store an 8 byte, or 64 bit, value. When both host and | |
5234 | target are 32 bits, an expression is computed as 32 bits. In this case | |
5235 | @code{QUAD} stores a 32 bit value zero extended to 64 bits, and | |
5236 | @code{SQUAD} stores a 32 bit value sign extended to 64 bits. | |
5237 | ||
5238 | If the object file format of the output file has an explicit endianness, | |
5239 | which is the normal case, the value will be stored in that endianness. | |
5240 | When the object file format does not have an explicit endianness, as is | |
5241 | true of, for example, S-records, the value will be stored in the | |
5242 | endianness of the first input object file. | |
5243 | ||
36f63dca | 5244 | Note---these commands only work inside a section description and not |
2b5fc1f5 NC |
5245 | between them, so the following will produce an error from the linker: |
5246 | @smallexample | |
5247 | SECTIONS @{@ .text : @{@ *(.text) @}@ LONG(1) .data : @{@ *(.data) @}@ @}@ | |
5248 | @end smallexample | |
5249 | whereas this will work: | |
5250 | @smallexample | |
5251 | SECTIONS @{@ .text : @{@ *(.text) ; LONG(1) @}@ .data : @{@ *(.data) @}@ @}@ | |
5252 | @end smallexample | |
5253 | ||
252b5132 RH |
5254 | @kindex FILL(@var{expression}) |
5255 | @cindex holes, filling | |
5256 | @cindex unspecified memory | |
5257 | You may use the @code{FILL} command to set the fill pattern for the | |
5258 | current section. It is followed by an expression in parentheses. Any | |
5259 | otherwise unspecified regions of memory within the section (for example, | |
5260 | gaps left due to the required alignment of input sections) are filled | |
a139d329 | 5261 | with the value of the expression, repeated as |
252b5132 RH |
5262 | necessary. A @code{FILL} statement covers memory locations after the |
5263 | point at which it occurs in the section definition; by including more | |
5264 | than one @code{FILL} statement, you can have different fill patterns in | |
5265 | different parts of an output section. | |
5266 | ||
5267 | This example shows how to fill unspecified regions of memory with the | |
563e308f | 5268 | value @samp{0x90}: |
252b5132 | 5269 | @smallexample |
563e308f | 5270 | FILL(0x90909090) |
252b5132 RH |
5271 | @end smallexample |
5272 | ||
5273 | The @code{FILL} command is similar to the @samp{=@var{fillexp}} output | |
9673c93c | 5274 | section attribute, but it only affects the |
252b5132 RH |
5275 | part of the section following the @code{FILL} command, rather than the |
5276 | entire section. If both are used, the @code{FILL} command takes | |
9673c93c | 5277 | precedence. @xref{Output Section Fill}, for details on the fill |
a139d329 | 5278 | expression. |
252b5132 RH |
5279 | |
5280 | @node Output Section Keywords | |
36f63dca | 5281 | @subsection Output Section Keywords |
252b5132 RH |
5282 | There are a couple of keywords which can appear as output section |
5283 | commands. | |
5284 | ||
5285 | @table @code | |
5286 | @kindex CREATE_OBJECT_SYMBOLS | |
5287 | @cindex input filename symbols | |
5288 | @cindex filename symbols | |
5289 | @item CREATE_OBJECT_SYMBOLS | |
5290 | The command tells the linker to create a symbol for each input file. | |
5291 | The name of each symbol will be the name of the corresponding input | |
5292 | file. The section of each symbol will be the output section in which | |
5293 | the @code{CREATE_OBJECT_SYMBOLS} command appears. | |
5294 | ||
5295 | This is conventional for the a.out object file format. It is not | |
5296 | normally used for any other object file format. | |
5297 | ||
5298 | @kindex CONSTRUCTORS | |
5299 | @cindex C++ constructors, arranging in link | |
5300 | @cindex constructors, arranging in link | |
5301 | @item CONSTRUCTORS | |
5302 | When linking using the a.out object file format, the linker uses an | |
5303 | unusual set construct to support C++ global constructors and | |
5304 | destructors. When linking object file formats which do not support | |
5305 | arbitrary sections, such as ECOFF and XCOFF, the linker will | |
5306 | automatically recognize C++ global constructors and destructors by name. | |
5307 | For these object file formats, the @code{CONSTRUCTORS} command tells the | |
5308 | linker to place constructor information in the output section where the | |
5309 | @code{CONSTRUCTORS} command appears. The @code{CONSTRUCTORS} command is | |
5310 | ignored for other object file formats. | |
5311 | ||
5312 | The symbol @w{@code{__CTOR_LIST__}} marks the start of the global | |
7e69709c AM |
5313 | constructors, and the symbol @w{@code{__CTOR_END__}} marks the end. |
5314 | Similarly, @w{@code{__DTOR_LIST__}} and @w{@code{__DTOR_END__}} mark | |
5315 | the start and end of the global destructors. The | |
252b5132 RH |
5316 | first word in the list is the number of entries, followed by the address |
5317 | of each constructor or destructor, followed by a zero word. The | |
5318 | compiler must arrange to actually run the code. For these object file | |
5319 | formats @sc{gnu} C++ normally calls constructors from a subroutine | |
5320 | @code{__main}; a call to @code{__main} is automatically inserted into | |
5321 | the startup code for @code{main}. @sc{gnu} C++ normally runs | |
5322 | destructors either by using @code{atexit}, or directly from the function | |
5323 | @code{exit}. | |
5324 | ||
5325 | For object file formats such as @code{COFF} or @code{ELF} which support | |
5326 | arbitrary section names, @sc{gnu} C++ will normally arrange to put the | |
5327 | addresses of global constructors and destructors into the @code{.ctors} | |
5328 | and @code{.dtors} sections. Placing the following sequence into your | |
5329 | linker script will build the sort of table which the @sc{gnu} C++ | |
5330 | runtime code expects to see. | |
5331 | ||
5332 | @smallexample | |
5333 | __CTOR_LIST__ = .; | |
5334 | LONG((__CTOR_END__ - __CTOR_LIST__) / 4 - 2) | |
5335 | *(.ctors) | |
5336 | LONG(0) | |
5337 | __CTOR_END__ = .; | |
5338 | __DTOR_LIST__ = .; | |
5339 | LONG((__DTOR_END__ - __DTOR_LIST__) / 4 - 2) | |
5340 | *(.dtors) | |
5341 | LONG(0) | |
5342 | __DTOR_END__ = .; | |
5343 | @end smallexample | |
5344 | ||
5345 | If you are using the @sc{gnu} C++ support for initialization priority, | |
5346 | which provides some control over the order in which global constructors | |
5347 | are run, you must sort the constructors at link time to ensure that they | |
5348 | are executed in the correct order. When using the @code{CONSTRUCTORS} | |
bcaa7b3e L |
5349 | command, use @samp{SORT_BY_NAME(CONSTRUCTORS)} instead. When using the |
5350 | @code{.ctors} and @code{.dtors} sections, use @samp{*(SORT_BY_NAME(.ctors))} and | |
5351 | @samp{*(SORT_BY_NAME(.dtors))} instead of just @samp{*(.ctors)} and | |
252b5132 RH |
5352 | @samp{*(.dtors)}. |
5353 | ||
5354 | Normally the compiler and linker will handle these issues automatically, | |
5355 | and you will not need to concern yourself with them. However, you may | |
5356 | need to consider this if you are using C++ and writing your own linker | |
5357 | scripts. | |
5358 | ||
5359 | @end table | |
5360 | ||
5361 | @node Output Section Discarding | |
36f63dca | 5362 | @subsection Output Section Discarding |
252b5132 RH |
5363 | @cindex discarding sections |
5364 | @cindex sections, discarding | |
5365 | @cindex removing sections | |
2edab91c AM |
5366 | The linker will not normally create output sections with no contents. |
5367 | This is for convenience when referring to input sections that may or | |
5368 | may not be present in any of the input files. For example: | |
252b5132 | 5369 | @smallexample |
49c13adb | 5370 | .foo : @{ *(.foo) @} |
252b5132 RH |
5371 | @end smallexample |
5372 | @noindent | |
5373 | will only create a @samp{.foo} section in the output file if there is a | |
74541ad4 AM |
5374 | @samp{.foo} section in at least one input file, and if the input |
5375 | sections are not all empty. Other link script directives that allocate | |
2edab91c AM |
5376 | space in an output section will also create the output section. So |
5377 | too will assignments to dot even if the assignment does not create | |
5378 | space, except for @samp{. = 0}, @samp{. = . + 0}, @samp{. = sym}, | |
5379 | @samp{. = . + sym} and @samp{. = ALIGN (. != 0, expr, 1)} when | |
5380 | @samp{sym} is an absolute symbol of value 0 defined in the script. | |
5381 | This allows you to force output of an empty section with @samp{. = .}. | |
74541ad4 | 5382 | |
a0976ea4 | 5383 | The linker will ignore address assignments (@pxref{Output Section Address}) |
74541ad4 AM |
5384 | on discarded output sections, except when the linker script defines |
5385 | symbols in the output section. In that case the linker will obey | |
a0976ea4 AM |
5386 | the address assignments, possibly advancing dot even though the |
5387 | section is discarded. | |
252b5132 RH |
5388 | |
5389 | @cindex /DISCARD/ | |
5390 | The special output section name @samp{/DISCARD/} may be used to discard | |
5391 | input sections. Any input sections which are assigned to an output | |
5392 | section named @samp{/DISCARD/} are not included in the output file. | |
5393 | ||
99fabbc9 JL |
5394 | This can be used to discard input sections marked with the ELF flag |
5395 | @code{SHF_GNU_RETAIN}, which would otherwise have been saved from linker | |
5396 | garbage collection. | |
5397 | ||
a2e098c3 NC |
5398 | Note, sections that match the @samp{/DISCARD/} output section will be |
5399 | discarded even if they are in an ELF section group which has other | |
5400 | members which are not being discarded. This is deliberate. | |
5401 | Discarding takes precedence over grouping. | |
5402 | ||
252b5132 | 5403 | @node Output Section Attributes |
36f63dca | 5404 | @subsection Output Section Attributes |
252b5132 RH |
5405 | @cindex output section attributes |
5406 | We showed above that the full description of an output section looked | |
5407 | like this: | |
0c71d759 | 5408 | |
252b5132 | 5409 | @smallexample |
a1ab1d2a | 5410 | @group |
7e7d5768 | 5411 | @var{section} [@var{address}] [(@var{type})] : |
0c71d759 | 5412 | [AT(@var{lma})] |
3bbec4bd | 5413 | [ALIGN(@var{section_align}) | ALIGN_WITH_INPUT] |
0c71d759 NC |
5414 | [SUBALIGN(@var{subsection_align})] |
5415 | [@var{constraint}] | |
252b5132 RH |
5416 | @{ |
5417 | @var{output-section-command} | |
5418 | @var{output-section-command} | |
5419 | @dots{} | |
562d3460 | 5420 | @} [>@var{region}] [AT>@var{lma_region}] [:@var{phdr} :@var{phdr} @dots{}] [=@var{fillexp}] |
252b5132 RH |
5421 | @end group |
5422 | @end smallexample | |
0c71d759 | 5423 | |
252b5132 RH |
5424 | We've already described @var{section}, @var{address}, and |
5425 | @var{output-section-command}. In this section we will describe the | |
5426 | remaining section attributes. | |
5427 | ||
a1ab1d2a | 5428 | @menu |
252b5132 RH |
5429 | * Output Section Type:: Output section type |
5430 | * Output Section LMA:: Output section LMA | |
bbf115d3 | 5431 | * Forced Output Alignment:: Forced Output Alignment |
7e7d5768 | 5432 | * Forced Input Alignment:: Forced Input Alignment |
0c71d759 | 5433 | * Output Section Constraint:: Output section constraint |
252b5132 RH |
5434 | * Output Section Region:: Output section region |
5435 | * Output Section Phdr:: Output section phdr | |
5436 | * Output Section Fill:: Output section fill | |
5437 | @end menu | |
5438 | ||
5439 | @node Output Section Type | |
36f63dca | 5440 | @subsubsection Output Section Type |
252b5132 RH |
5441 | Each output section may have a type. The type is a keyword in |
5442 | parentheses. The following types are defined: | |
5443 | ||
5444 | @table @code | |
5445 | @item NOLOAD | |
5446 | The section should be marked as not loadable, so that it will not be | |
5447 | loaded into memory when the program is run. | |
5448 | @item DSECT | |
5449 | @itemx COPY | |
5450 | @itemx INFO | |
5451 | @itemx OVERLAY | |
5452 | These type names are supported for backward compatibility, and are | |
5453 | rarely used. They all have the same effect: the section should be | |
5454 | marked as not allocatable, so that no memory is allocated for the | |
5455 | section when the program is run. | |
5456 | @end table | |
5457 | ||
5458 | @kindex NOLOAD | |
5459 | @cindex prevent unnecessary loading | |
5460 | @cindex loading, preventing | |
5461 | The linker normally sets the attributes of an output section based on | |
5462 | the input sections which map into it. You can override this by using | |
5463 | the section type. For example, in the script sample below, the | |
5464 | @samp{ROM} section is addressed at memory location @samp{0} and does not | |
2e76e85a | 5465 | need to be loaded when the program is run. |
252b5132 RH |
5466 | @smallexample |
5467 | @group | |
5468 | SECTIONS @{ | |
5469 | ROM 0 (NOLOAD) : @{ @dots{} @} | |
5470 | @dots{} | |
5471 | @} | |
5472 | @end group | |
5473 | @end smallexample | |
5474 | ||
5475 | @node Output Section LMA | |
36f63dca | 5476 | @subsubsection Output Section LMA |
562d3460 | 5477 | @kindex AT>@var{lma_region} |
252b5132 RH |
5478 | @kindex AT(@var{lma}) |
5479 | @cindex load address | |
5480 | @cindex section load address | |
5481 | Every section has a virtual address (VMA) and a load address (LMA); see | |
ea5cae92 NC |
5482 | @ref{Basic Script Concepts}. The virtual address is specified by the |
5483 | @pxref{Output Section Address} described earlier. The load address is | |
5484 | specified by the @code{AT} or @code{AT>} keywords. Specifying a load | |
5485 | address is optional. | |
6bdafbeb | 5486 | |
ea5cae92 NC |
5487 | The @code{AT} keyword takes an expression as an argument. This |
5488 | specifies the exact load address of the section. The @code{AT>} keyword | |
5489 | takes the name of a memory region as an argument. @xref{MEMORY}. The | |
5490 | load address of the section is set to the next free address in the | |
5491 | region, aligned to the section's alignment requirements. | |
dc0b6aa0 AM |
5492 | |
5493 | If neither @code{AT} nor @code{AT>} is specified for an allocatable | |
ea5cae92 NC |
5494 | section, the linker will use the following heuristic to determine the |
5495 | load address: | |
5496 | ||
5497 | @itemize @bullet | |
5498 | @item | |
5499 | If the section has a specific VMA address, then this is used as | |
5500 | the LMA address as well. | |
5501 | ||
5502 | @item | |
5503 | If the section is not allocatable then its LMA is set to its VMA. | |
5504 | ||
5505 | @item | |
5506 | Otherwise if a memory region can be found that is compatible | |
5507 | with the current section, and this region contains at least one | |
5508 | section, then the LMA is set so the difference between the | |
5509 | VMA and LMA is the same as the difference between the VMA and LMA of | |
5510 | the last section in the located region. | |
5511 | ||
5512 | @item | |
5513 | If no memory regions have been declared then a default region | |
5514 | that covers the entire address space is used in the previous step. | |
5515 | ||
5516 | @item | |
5517 | If no suitable region could be found, or there was no previous | |
5518 | section then the LMA is set equal to the VMA. | |
5519 | @end itemize | |
252b5132 RH |
5520 | |
5521 | @cindex ROM initialized data | |
5522 | @cindex initialized data in ROM | |
5523 | This feature is designed to make it easy to build a ROM image. For | |
5524 | example, the following linker script creates three output sections: one | |
5525 | called @samp{.text}, which starts at @code{0x1000}, one called | |
5526 | @samp{.mdata}, which is loaded at the end of the @samp{.text} section | |
5527 | even though its VMA is @code{0x2000}, and one called @samp{.bss} to hold | |
5528 | uninitialized data at address @code{0x3000}. The symbol @code{_data} is | |
5529 | defined with the value @code{0x2000}, which shows that the location | |
5530 | counter holds the VMA value, not the LMA value. | |
5531 | ||
5532 | @smallexample | |
5533 | @group | |
5534 | SECTIONS | |
5535 | @{ | |
5536 | .text 0x1000 : @{ *(.text) _etext = . ; @} | |
a1ab1d2a | 5537 | .mdata 0x2000 : |
252b5132 RH |
5538 | AT ( ADDR (.text) + SIZEOF (.text) ) |
5539 | @{ _data = . ; *(.data); _edata = . ; @} | |
5540 | .bss 0x3000 : | |
5541 | @{ _bstart = . ; *(.bss) *(COMMON) ; _bend = . ;@} | |
5542 | @} | |
5543 | @end group | |
5544 | @end smallexample | |
5545 | ||
5546 | The run-time initialization code for use with a program generated with | |
5547 | this linker script would include something like the following, to copy | |
5548 | the initialized data from the ROM image to its runtime address. Notice | |
5549 | how this code takes advantage of the symbols defined by the linker | |
5550 | script. | |
5551 | ||
5552 | @smallexample | |
5553 | @group | |
5554 | extern char _etext, _data, _edata, _bstart, _bend; | |
5555 | char *src = &_etext; | |
5556 | char *dst = &_data; | |
5557 | ||
ea5cae92 NC |
5558 | /* ROM has data at end of text; copy it. */ |
5559 | while (dst < &_edata) | |
252b5132 | 5560 | *dst++ = *src++; |
252b5132 | 5561 | |
ea5cae92 | 5562 | /* Zero bss. */ |
252b5132 RH |
5563 | for (dst = &_bstart; dst< &_bend; dst++) |
5564 | *dst = 0; | |
5565 | @end group | |
5566 | @end smallexample | |
5567 | ||
bbf115d3 L |
5568 | @node Forced Output Alignment |
5569 | @subsubsection Forced Output Alignment | |
5570 | @kindex ALIGN(@var{section_align}) | |
5571 | @cindex forcing output section alignment | |
5572 | @cindex output section alignment | |
1eec346e | 5573 | You can increase an output section's alignment by using ALIGN. As an |
13075d04 SH |
5574 | alternative you can enforce that the difference between the VMA and LMA remains |
5575 | intact throughout this output section with the ALIGN_WITH_INPUT attribute. | |
bbf115d3 | 5576 | |
7e7d5768 AM |
5577 | @node Forced Input Alignment |
5578 | @subsubsection Forced Input Alignment | |
5579 | @kindex SUBALIGN(@var{subsection_align}) | |
5580 | @cindex forcing input section alignment | |
5581 | @cindex input section alignment | |
5582 | You can force input section alignment within an output section by using | |
5583 | SUBALIGN. The value specified overrides any alignment given by input | |
5584 | sections, whether larger or smaller. | |
5585 | ||
0c71d759 NC |
5586 | @node Output Section Constraint |
5587 | @subsubsection Output Section Constraint | |
5588 | @kindex ONLY_IF_RO | |
5589 | @kindex ONLY_IF_RW | |
5590 | @cindex constraints on output sections | |
5591 | You can specify that an output section should only be created if all | |
5592 | of its input sections are read-only or all of its input sections are | |
5593 | read-write by using the keyword @code{ONLY_IF_RO} and | |
5594 | @code{ONLY_IF_RW} respectively. | |
5595 | ||
252b5132 | 5596 | @node Output Section Region |
36f63dca | 5597 | @subsubsection Output Section Region |
252b5132 RH |
5598 | @kindex >@var{region} |
5599 | @cindex section, assigning to memory region | |
5600 | @cindex memory regions and sections | |
5601 | You can assign a section to a previously defined region of memory by | |
5602 | using @samp{>@var{region}}. @xref{MEMORY}. | |
5603 | ||
5604 | Here is a simple example: | |
5605 | @smallexample | |
5606 | @group | |
5607 | MEMORY @{ rom : ORIGIN = 0x1000, LENGTH = 0x1000 @} | |
5608 | SECTIONS @{ ROM : @{ *(.text) @} >rom @} | |
5609 | @end group | |
5610 | @end smallexample | |
5611 | ||
5612 | @node Output Section Phdr | |
36f63dca | 5613 | @subsubsection Output Section Phdr |
252b5132 RH |
5614 | @kindex :@var{phdr} |
5615 | @cindex section, assigning to program header | |
5616 | @cindex program headers and sections | |
5617 | You can assign a section to a previously defined program segment by | |
5618 | using @samp{:@var{phdr}}. @xref{PHDRS}. If a section is assigned to | |
5619 | one or more segments, then all subsequent allocated sections will be | |
5620 | assigned to those segments as well, unless they use an explicitly | |
5621 | @code{:@var{phdr}} modifier. You can use @code{:NONE} to tell the | |
5622 | linker to not put the section in any segment at all. | |
5623 | ||
5624 | Here is a simple example: | |
5625 | @smallexample | |
5626 | @group | |
5627 | PHDRS @{ text PT_LOAD ; @} | |
5628 | SECTIONS @{ .text : @{ *(.text) @} :text @} | |
5629 | @end group | |
5630 | @end smallexample | |
5631 | ||
5632 | @node Output Section Fill | |
36f63dca | 5633 | @subsubsection Output Section Fill |
252b5132 RH |
5634 | @kindex =@var{fillexp} |
5635 | @cindex section fill pattern | |
5636 | @cindex fill pattern, entire section | |
5637 | You can set the fill pattern for an entire section by using | |
5638 | @samp{=@var{fillexp}}. @var{fillexp} is an expression | |
5639 | (@pxref{Expressions}). Any otherwise unspecified regions of memory | |
5640 | within the output section (for example, gaps left due to the required | |
a139d329 AM |
5641 | alignment of input sections) will be filled with the value, repeated as |
5642 | necessary. If the fill expression is a simple hex number, ie. a string | |
9673c93c | 5643 | of hex digit starting with @samp{0x} and without a trailing @samp{k} or @samp{M}, then |
a139d329 AM |
5644 | an arbitrarily long sequence of hex digits can be used to specify the |
5645 | fill pattern; Leading zeros become part of the pattern too. For all | |
9673c93c | 5646 | other cases, including extra parentheses or a unary @code{+}, the fill |
a139d329 AM |
5647 | pattern is the four least significant bytes of the value of the |
5648 | expression. In all cases, the number is big-endian. | |
252b5132 RH |
5649 | |
5650 | You can also change the fill value with a @code{FILL} command in the | |
9673c93c | 5651 | output section commands; (@pxref{Output Section Data}). |
252b5132 RH |
5652 | |
5653 | Here is a simple example: | |
5654 | @smallexample | |
5655 | @group | |
563e308f | 5656 | SECTIONS @{ .text : @{ *(.text) @} =0x90909090 @} |
252b5132 RH |
5657 | @end group |
5658 | @end smallexample | |
5659 | ||
5660 | @node Overlay Description | |
36f63dca | 5661 | @subsection Overlay Description |
252b5132 RH |
5662 | @kindex OVERLAY |
5663 | @cindex overlays | |
5664 | An overlay description provides an easy way to describe sections which | |
5665 | are to be loaded as part of a single memory image but are to be run at | |
5666 | the same memory address. At run time, some sort of overlay manager will | |
5667 | copy the overlaid sections in and out of the runtime memory address as | |
5668 | required, perhaps by simply manipulating addressing bits. This approach | |
5669 | can be useful, for example, when a certain region of memory is faster | |
5670 | than another. | |
5671 | ||
5672 | Overlays are described using the @code{OVERLAY} command. The | |
5673 | @code{OVERLAY} command is used within a @code{SECTIONS} command, like an | |
5674 | output section description. The full syntax of the @code{OVERLAY} | |
5675 | command is as follows: | |
5676 | @smallexample | |
5677 | @group | |
5678 | OVERLAY [@var{start}] : [NOCROSSREFS] [AT ( @var{ldaddr} )] | |
5679 | @{ | |
5680 | @var{secname1} | |
5681 | @{ | |
5682 | @var{output-section-command} | |
5683 | @var{output-section-command} | |
5684 | @dots{} | |
5685 | @} [:@var{phdr}@dots{}] [=@var{fill}] | |
5686 | @var{secname2} | |
5687 | @{ | |
5688 | @var{output-section-command} | |
5689 | @var{output-section-command} | |
5690 | @dots{} | |
5691 | @} [:@var{phdr}@dots{}] [=@var{fill}] | |
5692 | @dots{} | |
abc9061b | 5693 | @} [>@var{region}] [:@var{phdr}@dots{}] [=@var{fill}] [,] |
252b5132 RH |
5694 | @end group |
5695 | @end smallexample | |
5696 | ||
5697 | Everything is optional except @code{OVERLAY} (a keyword), and each | |
5698 | section must have a name (@var{secname1} and @var{secname2} above). The | |
5699 | section definitions within the @code{OVERLAY} construct are identical to | |
11e7fd74 | 5700 | those within the general @code{SECTIONS} construct (@pxref{SECTIONS}), |
252b5132 RH |
5701 | except that no addresses and no memory regions may be defined for |
5702 | sections within an @code{OVERLAY}. | |
5703 | ||
abc9061b CC |
5704 | The comma at the end may be required if a @var{fill} is used and |
5705 | the next @var{sections-command} looks like a continuation of the expression. | |
5706 | ||
252b5132 RH |
5707 | The sections are all defined with the same starting address. The load |
5708 | addresses of the sections are arranged such that they are consecutive in | |
5709 | memory starting at the load address used for the @code{OVERLAY} as a | |
5710 | whole (as with normal section definitions, the load address is optional, | |
5711 | and defaults to the start address; the start address is also optional, | |
5712 | and defaults to the current value of the location counter). | |
5713 | ||
56dd11f0 NC |
5714 | If the @code{NOCROSSREFS} keyword is used, and there are any |
5715 | references among the sections, the linker will report an error. Since | |
5716 | the sections all run at the same address, it normally does not make | |
5717 | sense for one section to refer directly to another. | |
5718 | @xref{Miscellaneous Commands, NOCROSSREFS}. | |
252b5132 RH |
5719 | |
5720 | For each section within the @code{OVERLAY}, the linker automatically | |
34711ca3 | 5721 | provides two symbols. The symbol @code{__load_start_@var{secname}} is |
252b5132 RH |
5722 | defined as the starting load address of the section. The symbol |
5723 | @code{__load_stop_@var{secname}} is defined as the final load address of | |
5724 | the section. Any characters within @var{secname} which are not legal | |
5725 | within C identifiers are removed. C (or assembler) code may use these | |
5726 | symbols to move the overlaid sections around as necessary. | |
5727 | ||
5728 | At the end of the overlay, the value of the location counter is set to | |
5729 | the start address of the overlay plus the size of the largest section. | |
5730 | ||
5731 | Here is an example. Remember that this would appear inside a | |
5732 | @code{SECTIONS} construct. | |
5733 | @smallexample | |
5734 | @group | |
5735 | OVERLAY 0x1000 : AT (0x4000) | |
5736 | @{ | |
5737 | .text0 @{ o1/*.o(.text) @} | |
5738 | .text1 @{ o2/*.o(.text) @} | |
5739 | @} | |
5740 | @end group | |
5741 | @end smallexample | |
5742 | @noindent | |
5743 | This will define both @samp{.text0} and @samp{.text1} to start at | |
5744 | address 0x1000. @samp{.text0} will be loaded at address 0x4000, and | |
5745 | @samp{.text1} will be loaded immediately after @samp{.text0}. The | |
34711ca3 | 5746 | following symbols will be defined if referenced: @code{__load_start_text0}, |
252b5132 RH |
5747 | @code{__load_stop_text0}, @code{__load_start_text1}, |
5748 | @code{__load_stop_text1}. | |
5749 | ||
5750 | C code to copy overlay @code{.text1} into the overlay area might look | |
5751 | like the following. | |
5752 | ||
5753 | @smallexample | |
5754 | @group | |
5755 | extern char __load_start_text1, __load_stop_text1; | |
5756 | memcpy ((char *) 0x1000, &__load_start_text1, | |
5757 | &__load_stop_text1 - &__load_start_text1); | |
5758 | @end group | |
5759 | @end smallexample | |
5760 | ||
5761 | Note that the @code{OVERLAY} command is just syntactic sugar, since | |
5762 | everything it does can be done using the more basic commands. The above | |
5763 | example could have been written identically as follows. | |
5764 | ||
5765 | @smallexample | |
5766 | @group | |
5767 | .text0 0x1000 : AT (0x4000) @{ o1/*.o(.text) @} | |
34711ca3 AM |
5768 | PROVIDE (__load_start_text0 = LOADADDR (.text0)); |
5769 | PROVIDE (__load_stop_text0 = LOADADDR (.text0) + SIZEOF (.text0)); | |
252b5132 | 5770 | .text1 0x1000 : AT (0x4000 + SIZEOF (.text0)) @{ o2/*.o(.text) @} |
34711ca3 AM |
5771 | PROVIDE (__load_start_text1 = LOADADDR (.text1)); |
5772 | PROVIDE (__load_stop_text1 = LOADADDR (.text1) + SIZEOF (.text1)); | |
252b5132 RH |
5773 | . = 0x1000 + MAX (SIZEOF (.text0), SIZEOF (.text1)); |
5774 | @end group | |
5775 | @end smallexample | |
5776 | ||
5777 | @node MEMORY | |
36f63dca | 5778 | @section MEMORY Command |
252b5132 RH |
5779 | @kindex MEMORY |
5780 | @cindex memory regions | |
5781 | @cindex regions of memory | |
5782 | @cindex allocating memory | |
5783 | @cindex discontinuous memory | |
5784 | The linker's default configuration permits allocation of all available | |
5785 | memory. You can override this by using the @code{MEMORY} command. | |
5786 | ||
5787 | The @code{MEMORY} command describes the location and size of blocks of | |
5788 | memory in the target. You can use it to describe which memory regions | |
5789 | may be used by the linker, and which memory regions it must avoid. You | |
5790 | can then assign sections to particular memory regions. The linker will | |
5791 | set section addresses based on the memory regions, and will warn about | |
5792 | regions that become too full. The linker will not shuffle sections | |
5793 | around to fit into the available regions. | |
5794 | ||
127fcdff AB |
5795 | A linker script may contain many uses of the @code{MEMORY} command, |
5796 | however, all memory blocks defined are treated as if they were | |
5797 | specified inside a single @code{MEMORY} command. The syntax for | |
5798 | @code{MEMORY} is: | |
252b5132 RH |
5799 | @smallexample |
5800 | @group | |
a1ab1d2a | 5801 | MEMORY |
252b5132 RH |
5802 | @{ |
5803 | @var{name} [(@var{attr})] : ORIGIN = @var{origin}, LENGTH = @var{len} | |
5804 | @dots{} | |
5805 | @} | |
5806 | @end group | |
5807 | @end smallexample | |
5808 | ||
5809 | The @var{name} is a name used in the linker script to refer to the | |
5810 | region. The region name has no meaning outside of the linker script. | |
5811 | Region names are stored in a separate name space, and will not conflict | |
5812 | with symbol names, file names, or section names. Each memory region | |
4a93e180 NC |
5813 | must have a distinct name within the @code{MEMORY} command. However you can |
5814 | add later alias names to existing memory regions with the @ref{REGION_ALIAS} | |
9d5777a3 | 5815 | command. |
252b5132 RH |
5816 | |
5817 | @cindex memory region attributes | |
5818 | The @var{attr} string is an optional list of attributes that specify | |
5819 | whether to use a particular memory region for an input section which is | |
5820 | not explicitly mapped in the linker script. As described in | |
5821 | @ref{SECTIONS}, if you do not specify an output section for some input | |
5822 | section, the linker will create an output section with the same name as | |
5823 | the input section. If you define region attributes, the linker will use | |
5824 | them to select the memory region for the output section that it creates. | |
5825 | ||
5826 | The @var{attr} string must consist only of the following characters: | |
5827 | @table @samp | |
5828 | @item R | |
5829 | Read-only section | |
5830 | @item W | |
5831 | Read/write section | |
5832 | @item X | |
5833 | Executable section | |
5834 | @item A | |
5835 | Allocatable section | |
5836 | @item I | |
5837 | Initialized section | |
5838 | @item L | |
5839 | Same as @samp{I} | |
5840 | @item ! | |
c09e9a8c | 5841 | Invert the sense of any of the attributes that follow |
252b5132 RH |
5842 | @end table |
5843 | ||
81c688d5 | 5844 | If an unmapped section matches any of the listed attributes other than |
252b5132 | 5845 | @samp{!}, it will be placed in the memory region. The @samp{!} |
81c688d5 NC |
5846 | attribute reverses the test for the characters that follow, so that an |
5847 | unmapped section will be placed in the memory region only if it does | |
5848 | not match any of the attributes listed afterwards. Thus an attribute | |
5849 | string of @samp{RW!X} will match any unmapped section that has either | |
5850 | or both of the @samp{R} and @samp{W} attributes, but only as long as | |
5851 | the section does not also have the @samp{X} attribute. | |
252b5132 RH |
5852 | |
5853 | @kindex ORIGIN = | |
5854 | @kindex o = | |
5855 | @kindex org = | |
9cd6d51a NC |
5856 | The @var{origin} is an numerical expression for the start address of |
5857 | the memory region. The expression must evaluate to a constant and it | |
5858 | cannot involve any symbols. The keyword @code{ORIGIN} may be | |
5859 | abbreviated to @code{org} or @code{o} (but not, for example, | |
5860 | @code{ORG}). | |
252b5132 RH |
5861 | |
5862 | @kindex LENGTH = | |
5863 | @kindex len = | |
5864 | @kindex l = | |
5865 | The @var{len} is an expression for the size in bytes of the memory | |
5866 | region. As with the @var{origin} expression, the expression must | |
9cd6d51a NC |
5867 | be numerical only and must evaluate to a constant. The keyword |
5868 | @code{LENGTH} may be abbreviated to @code{len} or @code{l}. | |
252b5132 RH |
5869 | |
5870 | In the following example, we specify that there are two memory regions | |
5871 | available for allocation: one starting at @samp{0} for 256 kilobytes, | |
5872 | and the other starting at @samp{0x40000000} for four megabytes. The | |
5873 | linker will place into the @samp{rom} memory region every section which | |
5874 | is not explicitly mapped into a memory region, and is either read-only | |
5875 | or executable. The linker will place other sections which are not | |
5876 | explicitly mapped into a memory region into the @samp{ram} memory | |
5877 | region. | |
5878 | ||
5879 | @smallexample | |
5880 | @group | |
a1ab1d2a | 5881 | MEMORY |
252b5132 RH |
5882 | @{ |
5883 | rom (rx) : ORIGIN = 0, LENGTH = 256K | |
5884 | ram (!rx) : org = 0x40000000, l = 4M | |
5885 | @} | |
5886 | @end group | |
5887 | @end smallexample | |
5888 | ||
5889 | Once you define a memory region, you can direct the linker to place | |
5890 | specific output sections into that memory region by using the | |
5891 | @samp{>@var{region}} output section attribute. For example, if you have | |
5892 | a memory region named @samp{mem}, you would use @samp{>mem} in the | |
5893 | output section definition. @xref{Output Section Region}. If no address | |
5894 | was specified for the output section, the linker will set the address to | |
5895 | the next available address within the memory region. If the combined | |
5896 | output sections directed to a memory region are too large for the | |
5897 | region, the linker will issue an error message. | |
5898 | ||
3ec57632 | 5899 | It is possible to access the origin and length of a memory in an |
c0065db7 | 5900 | expression via the @code{ORIGIN(@var{memory})} and |
3ec57632 NC |
5901 | @code{LENGTH(@var{memory})} functions: |
5902 | ||
5903 | @smallexample | |
5904 | @group | |
c0065db7 | 5905 | _fstack = ORIGIN(ram) + LENGTH(ram) - 4; |
3ec57632 NC |
5906 | @end group |
5907 | @end smallexample | |
5908 | ||
252b5132 RH |
5909 | @node PHDRS |
5910 | @section PHDRS Command | |
5911 | @kindex PHDRS | |
5912 | @cindex program headers | |
5913 | @cindex ELF program headers | |
5914 | @cindex program segments | |
5915 | @cindex segments, ELF | |
5916 | The ELF object file format uses @dfn{program headers}, also knows as | |
5917 | @dfn{segments}. The program headers describe how the program should be | |
5918 | loaded into memory. You can print them out by using the @code{objdump} | |
5919 | program with the @samp{-p} option. | |
5920 | ||
5921 | When you run an ELF program on a native ELF system, the system loader | |
5922 | reads the program headers in order to figure out how to load the | |
5923 | program. This will only work if the program headers are set correctly. | |
5924 | This manual does not describe the details of how the system loader | |
5925 | interprets program headers; for more information, see the ELF ABI. | |
5926 | ||
5927 | The linker will create reasonable program headers by default. However, | |
5928 | in some cases, you may need to specify the program headers more | |
5929 | precisely. You may use the @code{PHDRS} command for this purpose. When | |
5930 | the linker sees the @code{PHDRS} command in the linker script, it will | |
5931 | not create any program headers other than the ones specified. | |
5932 | ||
5933 | The linker only pays attention to the @code{PHDRS} command when | |
5934 | generating an ELF output file. In other cases, the linker will simply | |
5935 | ignore @code{PHDRS}. | |
5936 | ||
5937 | This is the syntax of the @code{PHDRS} command. The words @code{PHDRS}, | |
5938 | @code{FILEHDR}, @code{AT}, and @code{FLAGS} are keywords. | |
5939 | ||
5940 | @smallexample | |
5941 | @group | |
5942 | PHDRS | |
5943 | @{ | |
5944 | @var{name} @var{type} [ FILEHDR ] [ PHDRS ] [ AT ( @var{address} ) ] | |
5945 | [ FLAGS ( @var{flags} ) ] ; | |
5946 | @} | |
5947 | @end group | |
5948 | @end smallexample | |
5949 | ||
5950 | The @var{name} is used only for reference in the @code{SECTIONS} command | |
5951 | of the linker script. It is not put into the output file. Program | |
5952 | header names are stored in a separate name space, and will not conflict | |
5953 | with symbol names, file names, or section names. Each program header | |
5c1a3f0f NS |
5954 | must have a distinct name. The headers are processed in order and it |
5955 | is usual for them to map to sections in ascending load address order. | |
252b5132 RH |
5956 | |
5957 | Certain program header types describe segments of memory which the | |
5958 | system loader will load from the file. In the linker script, you | |
5959 | specify the contents of these segments by placing allocatable output | |
5960 | sections in the segments. You use the @samp{:@var{phdr}} output section | |
5961 | attribute to place a section in a particular segment. @xref{Output | |
5962 | Section Phdr}. | |
5963 | ||
5964 | It is normal to put certain sections in more than one segment. This | |
5965 | merely implies that one segment of memory contains another. You may | |
5966 | repeat @samp{:@var{phdr}}, using it once for each segment which should | |
5967 | contain the section. | |
5968 | ||
5969 | If you place a section in one or more segments using @samp{:@var{phdr}}, | |
5970 | then the linker will place all subsequent allocatable sections which do | |
5971 | not specify @samp{:@var{phdr}} in the same segments. This is for | |
5972 | convenience, since generally a whole set of contiguous sections will be | |
5973 | placed in a single segment. You can use @code{:NONE} to override the | |
5974 | default segment and tell the linker to not put the section in any | |
5975 | segment at all. | |
5976 | ||
5977 | @kindex FILEHDR | |
5978 | @kindex PHDRS | |
5c1a3f0f | 5979 | You may use the @code{FILEHDR} and @code{PHDRS} keywords after |
252b5132 RH |
5980 | the program header type to further describe the contents of the segment. |
5981 | The @code{FILEHDR} keyword means that the segment should include the ELF | |
5982 | file header. The @code{PHDRS} keyword means that the segment should | |
5c1a3f0f | 5983 | include the ELF program headers themselves. If applied to a loadable |
4100cea3 AM |
5984 | segment (@code{PT_LOAD}), all prior loadable segments must have one of |
5985 | these keywords. | |
252b5132 RH |
5986 | |
5987 | The @var{type} may be one of the following. The numbers indicate the | |
5988 | value of the keyword. | |
5989 | ||
5990 | @table @asis | |
5991 | @item @code{PT_NULL} (0) | |
5992 | Indicates an unused program header. | |
5993 | ||
5994 | @item @code{PT_LOAD} (1) | |
5995 | Indicates that this program header describes a segment to be loaded from | |
5996 | the file. | |
5997 | ||
5998 | @item @code{PT_DYNAMIC} (2) | |
5999 | Indicates a segment where dynamic linking information can be found. | |
6000 | ||
6001 | @item @code{PT_INTERP} (3) | |
6002 | Indicates a segment where the name of the program interpreter may be | |
6003 | found. | |
6004 | ||
6005 | @item @code{PT_NOTE} (4) | |
6006 | Indicates a segment holding note information. | |
6007 | ||
6008 | @item @code{PT_SHLIB} (5) | |
6009 | A reserved program header type, defined but not specified by the ELF | |
6010 | ABI. | |
6011 | ||
6012 | @item @code{PT_PHDR} (6) | |
6013 | Indicates a segment where the program headers may be found. | |
6014 | ||
1a9ccd70 NC |
6015 | @item @code{PT_TLS} (7) |
6016 | Indicates a segment containing thread local storage. | |
6017 | ||
252b5132 RH |
6018 | @item @var{expression} |
6019 | An expression giving the numeric type of the program header. This may | |
6020 | be used for types not defined above. | |
6021 | @end table | |
6022 | ||
6023 | You can specify that a segment should be loaded at a particular address | |
6024 | in memory by using an @code{AT} expression. This is identical to the | |
6025 | @code{AT} command used as an output section attribute (@pxref{Output | |
6026 | Section LMA}). The @code{AT} command for a program header overrides the | |
6027 | output section attribute. | |
6028 | ||
6029 | The linker will normally set the segment flags based on the sections | |
6030 | which comprise the segment. You may use the @code{FLAGS} keyword to | |
6031 | explicitly specify the segment flags. The value of @var{flags} must be | |
6032 | an integer. It is used to set the @code{p_flags} field of the program | |
6033 | header. | |
6034 | ||
6035 | Here is an example of @code{PHDRS}. This shows a typical set of program | |
6036 | headers used on a native ELF system. | |
6037 | ||
6038 | @example | |
6039 | @group | |
6040 | PHDRS | |
6041 | @{ | |
6042 | headers PT_PHDR PHDRS ; | |
6043 | interp PT_INTERP ; | |
6044 | text PT_LOAD FILEHDR PHDRS ; | |
6045 | data PT_LOAD ; | |
6046 | dynamic PT_DYNAMIC ; | |
6047 | @} | |
6048 | ||
6049 | SECTIONS | |
6050 | @{ | |
6051 | . = SIZEOF_HEADERS; | |
6052 | .interp : @{ *(.interp) @} :text :interp | |
6053 | .text : @{ *(.text) @} :text | |
6054 | .rodata : @{ *(.rodata) @} /* defaults to :text */ | |
6055 | @dots{} | |
6056 | . = . + 0x1000; /* move to a new page in memory */ | |
6057 | .data : @{ *(.data) @} :data | |
6058 | .dynamic : @{ *(.dynamic) @} :data :dynamic | |
6059 | @dots{} | |
6060 | @} | |
6061 | @end group | |
6062 | @end example | |
6063 | ||
6064 | @node VERSION | |
6065 | @section VERSION Command | |
6066 | @kindex VERSION @{script text@} | |
6067 | @cindex symbol versions | |
6068 | @cindex version script | |
6069 | @cindex versions of symbols | |
6070 | The linker supports symbol versions when using ELF. Symbol versions are | |
6071 | only useful when using shared libraries. The dynamic linker can use | |
6072 | symbol versions to select a specific version of a function when it runs | |
6073 | a program that may have been linked against an earlier version of the | |
6074 | shared library. | |
6075 | ||
6076 | You can include a version script directly in the main linker script, or | |
6077 | you can supply the version script as an implicit linker script. You can | |
6078 | also use the @samp{--version-script} linker option. | |
6079 | ||
6080 | The syntax of the @code{VERSION} command is simply | |
6081 | @smallexample | |
6082 | VERSION @{ version-script-commands @} | |
6083 | @end smallexample | |
6084 | ||
6085 | The format of the version script commands is identical to that used by | |
6086 | Sun's linker in Solaris 2.5. The version script defines a tree of | |
6087 | version nodes. You specify the node names and interdependencies in the | |
6088 | version script. You can specify which symbols are bound to which | |
6089 | version nodes, and you can reduce a specified set of symbols to local | |
6090 | scope so that they are not globally visible outside of the shared | |
6091 | library. | |
6092 | ||
6093 | The easiest way to demonstrate the version script language is with a few | |
6094 | examples. | |
6095 | ||
6096 | @smallexample | |
6097 | VERS_1.1 @{ | |
6098 | global: | |
6099 | foo1; | |
6100 | local: | |
a1ab1d2a UD |
6101 | old*; |
6102 | original*; | |
6103 | new*; | |
252b5132 RH |
6104 | @}; |
6105 | ||
6106 | VERS_1.2 @{ | |
6107 | foo2; | |
6108 | @} VERS_1.1; | |
6109 | ||
6110 | VERS_2.0 @{ | |
6111 | bar1; bar2; | |
c0065db7 | 6112 | extern "C++" @{ |
86043bbb | 6113 | ns::*; |
bb1515f2 MF |
6114 | "f(int, double)"; |
6115 | @}; | |
252b5132 RH |
6116 | @} VERS_1.2; |
6117 | @end smallexample | |
6118 | ||
6119 | This example version script defines three version nodes. The first | |
6120 | version node defined is @samp{VERS_1.1}; it has no other dependencies. | |
6121 | The script binds the symbol @samp{foo1} to @samp{VERS_1.1}. It reduces | |
6122 | a number of symbols to local scope so that they are not visible outside | |
313e35ee AM |
6123 | of the shared library; this is done using wildcard patterns, so that any |
6124 | symbol whose name begins with @samp{old}, @samp{original}, or @samp{new} | |
6125 | is matched. The wildcard patterns available are the same as those used | |
6126 | in the shell when matching filenames (also known as ``globbing''). | |
86043bbb MM |
6127 | However, if you specify the symbol name inside double quotes, then the |
6128 | name is treated as literal, rather than as a glob pattern. | |
252b5132 RH |
6129 | |
6130 | Next, the version script defines node @samp{VERS_1.2}. This node | |
6131 | depends upon @samp{VERS_1.1}. The script binds the symbol @samp{foo2} | |
6132 | to the version node @samp{VERS_1.2}. | |
6133 | ||
6134 | Finally, the version script defines node @samp{VERS_2.0}. This node | |
6135 | depends upon @samp{VERS_1.2}. The scripts binds the symbols @samp{bar1} | |
6136 | and @samp{bar2} are bound to the version node @samp{VERS_2.0}. | |
6137 | ||
6138 | When the linker finds a symbol defined in a library which is not | |
6139 | specifically bound to a version node, it will effectively bind it to an | |
6140 | unspecified base version of the library. You can bind all otherwise | |
a981ed6f | 6141 | unspecified symbols to a given version node by using @samp{global: *;} |
ae5a3597 AM |
6142 | somewhere in the version script. Note that it's slightly crazy to use |
6143 | wildcards in a global spec except on the last version node. Global | |
6144 | wildcards elsewhere run the risk of accidentally adding symbols to the | |
6145 | set exported for an old version. That's wrong since older versions | |
6146 | ought to have a fixed set of symbols. | |
252b5132 RH |
6147 | |
6148 | The names of the version nodes have no specific meaning other than what | |
6149 | they might suggest to the person reading them. The @samp{2.0} version | |
6150 | could just as well have appeared in between @samp{1.1} and @samp{1.2}. | |
6151 | However, this would be a confusing way to write a version script. | |
6152 | ||
0f6bf451 | 6153 | Node name can be omitted, provided it is the only version node |
6b9b879a JJ |
6154 | in the version script. Such version script doesn't assign any versions to |
6155 | symbols, only selects which symbols will be globally visible out and which | |
6156 | won't. | |
6157 | ||
6158 | @smallexample | |
7c9c73be | 6159 | @{ global: foo; bar; local: *; @}; |
9d201f2f | 6160 | @end smallexample |
6b9b879a | 6161 | |
252b5132 RH |
6162 | When you link an application against a shared library that has versioned |
6163 | symbols, the application itself knows which version of each symbol it | |
6164 | requires, and it also knows which version nodes it needs from each | |
6165 | shared library it is linked against. Thus at runtime, the dynamic | |
6166 | loader can make a quick check to make sure that the libraries you have | |
6167 | linked against do in fact supply all of the version nodes that the | |
6168 | application will need to resolve all of the dynamic symbols. In this | |
6169 | way it is possible for the dynamic linker to know with certainty that | |
6170 | all external symbols that it needs will be resolvable without having to | |
6171 | search for each symbol reference. | |
6172 | ||
6173 | The symbol versioning is in effect a much more sophisticated way of | |
6174 | doing minor version checking that SunOS does. The fundamental problem | |
6175 | that is being addressed here is that typically references to external | |
6176 | functions are bound on an as-needed basis, and are not all bound when | |
6177 | the application starts up. If a shared library is out of date, a | |
6178 | required interface may be missing; when the application tries to use | |
6179 | that interface, it may suddenly and unexpectedly fail. With symbol | |
6180 | versioning, the user will get a warning when they start their program if | |
6181 | the libraries being used with the application are too old. | |
6182 | ||
6183 | There are several GNU extensions to Sun's versioning approach. The | |
6184 | first of these is the ability to bind a symbol to a version node in the | |
6185 | source file where the symbol is defined instead of in the versioning | |
6186 | script. This was done mainly to reduce the burden on the library | |
6187 | maintainer. You can do this by putting something like: | |
6188 | @smallexample | |
6189 | __asm__(".symver original_foo,foo@@VERS_1.1"); | |
6190 | @end smallexample | |
6191 | @noindent | |
6192 | in the C source file. This renames the function @samp{original_foo} to | |
6193 | be an alias for @samp{foo} bound to the version node @samp{VERS_1.1}. | |
6194 | The @samp{local:} directive can be used to prevent the symbol | |
96a94295 L |
6195 | @samp{original_foo} from being exported. A @samp{.symver} directive |
6196 | takes precedence over a version script. | |
252b5132 RH |
6197 | |
6198 | The second GNU extension is to allow multiple versions of the same | |
6199 | function to appear in a given shared library. In this way you can make | |
6200 | an incompatible change to an interface without increasing the major | |
6201 | version number of the shared library, while still allowing applications | |
6202 | linked against the old interface to continue to function. | |
6203 | ||
6204 | To do this, you must use multiple @samp{.symver} directives in the | |
6205 | source file. Here is an example: | |
6206 | ||
6207 | @smallexample | |
6208 | __asm__(".symver original_foo,foo@@"); | |
6209 | __asm__(".symver old_foo,foo@@VERS_1.1"); | |
6210 | __asm__(".symver old_foo1,foo@@VERS_1.2"); | |
6211 | __asm__(".symver new_foo,foo@@@@VERS_2.0"); | |
6212 | @end smallexample | |
6213 | ||
6214 | In this example, @samp{foo@@} represents the symbol @samp{foo} bound to the | |
6215 | unspecified base version of the symbol. The source file that contains this | |
6216 | example would define 4 C functions: @samp{original_foo}, @samp{old_foo}, | |
6217 | @samp{old_foo1}, and @samp{new_foo}. | |
6218 | ||
6219 | When you have multiple definitions of a given symbol, there needs to be | |
6220 | some way to specify a default version to which external references to | |
6221 | this symbol will be bound. You can do this with the | |
6222 | @samp{foo@@@@VERS_2.0} type of @samp{.symver} directive. You can only | |
6223 | declare one version of a symbol as the default in this manner; otherwise | |
6224 | you would effectively have multiple definitions of the same symbol. | |
6225 | ||
6226 | If you wish to bind a reference to a specific version of the symbol | |
6227 | within the shared library, you can use the aliases of convenience | |
36f63dca | 6228 | (i.e., @samp{old_foo}), or you can use the @samp{.symver} directive to |
252b5132 RH |
6229 | specifically bind to an external version of the function in question. |
6230 | ||
cb840a31 L |
6231 | You can also specify the language in the version script: |
6232 | ||
6233 | @smallexample | |
6234 | VERSION extern "lang" @{ version-script-commands @} | |
6235 | @end smallexample | |
6236 | ||
c0065db7 | 6237 | The supported @samp{lang}s are @samp{C}, @samp{C++}, and @samp{Java}. |
cb840a31 L |
6238 | The linker will iterate over the list of symbols at the link time and |
6239 | demangle them according to @samp{lang} before matching them to the | |
bb1515f2 MF |
6240 | patterns specified in @samp{version-script-commands}. The default |
6241 | @samp{lang} is @samp{C}. | |
cb840a31 | 6242 | |
86043bbb MM |
6243 | Demangled names may contains spaces and other special characters. As |
6244 | described above, you can use a glob pattern to match demangled names, | |
6245 | or you can use a double-quoted string to match the string exactly. In | |
6246 | the latter case, be aware that minor differences (such as differing | |
6247 | whitespace) between the version script and the demangler output will | |
6248 | cause a mismatch. As the exact string generated by the demangler | |
6249 | might change in the future, even if the mangled name does not, you | |
6250 | should check that all of your version directives are behaving as you | |
6251 | expect when you upgrade. | |
6252 | ||
252b5132 RH |
6253 | @node Expressions |
6254 | @section Expressions in Linker Scripts | |
6255 | @cindex expressions | |
6256 | @cindex arithmetic | |
6257 | The syntax for expressions in the linker script language is identical to | |
6258 | that of C expressions. All expressions are evaluated as integers. All | |
6259 | expressions are evaluated in the same size, which is 32 bits if both the | |
6260 | host and target are 32 bits, and is otherwise 64 bits. | |
6261 | ||
6262 | You can use and set symbol values in expressions. | |
6263 | ||
6264 | The linker defines several special purpose builtin functions for use in | |
6265 | expressions. | |
6266 | ||
6267 | @menu | |
6268 | * Constants:: Constants | |
0c71d759 | 6269 | * Symbolic Constants:: Symbolic constants |
252b5132 | 6270 | * Symbols:: Symbol Names |
ecca9871 | 6271 | * Orphan Sections:: Orphan Sections |
252b5132 RH |
6272 | * Location Counter:: The Location Counter |
6273 | * Operators:: Operators | |
6274 | * Evaluation:: Evaluation | |
6275 | * Expression Section:: The Section of an Expression | |
6276 | * Builtin Functions:: Builtin Functions | |
6277 | @end menu | |
6278 | ||
6279 | @node Constants | |
6280 | @subsection Constants | |
6281 | @cindex integer notation | |
6282 | @cindex constants in linker scripts | |
6283 | All constants are integers. | |
6284 | ||
6285 | As in C, the linker considers an integer beginning with @samp{0} to be | |
6286 | octal, and an integer beginning with @samp{0x} or @samp{0X} to be | |
8a308ae8 | 6287 | hexadecimal. Alternatively the linker accepts suffixes of @samp{h} or |
11e7fd74 | 6288 | @samp{H} for hexadecimal, @samp{o} or @samp{O} for octal, @samp{b} or |
8a308ae8 NC |
6289 | @samp{B} for binary and @samp{d} or @samp{D} for decimal. Any integer |
6290 | value without a prefix or a suffix is considered to be decimal. | |
252b5132 RH |
6291 | |
6292 | @cindex scaled integers | |
6293 | @cindex K and M integer suffixes | |
6294 | @cindex M and K integer suffixes | |
6295 | @cindex suffixes for integers | |
6296 | @cindex integer suffixes | |
6297 | In addition, you can use the suffixes @code{K} and @code{M} to scale a | |
6298 | constant by | |
6299 | @c TEXI2ROFF-KILL | |
36f63dca | 6300 | @ifnottex |
252b5132 RH |
6301 | @c END TEXI2ROFF-KILL |
6302 | @code{1024} or @code{1024*1024} | |
6303 | @c TEXI2ROFF-KILL | |
36f63dca | 6304 | @end ifnottex |
252b5132 RH |
6305 | @tex |
6306 | ${\rm 1024}$ or ${\rm 1024}^2$ | |
6307 | @end tex | |
6308 | @c END TEXI2ROFF-KILL | |
8a308ae8 NC |
6309 | respectively. For example, the following |
6310 | all refer to the same quantity: | |
6311 | ||
252b5132 | 6312 | @smallexample |
36f63dca NC |
6313 | _fourk_1 = 4K; |
6314 | _fourk_2 = 4096; | |
6315 | _fourk_3 = 0x1000; | |
8a308ae8 | 6316 | _fourk_4 = 10000o; |
252b5132 RH |
6317 | @end smallexample |
6318 | ||
8a308ae8 NC |
6319 | Note - the @code{K} and @code{M} suffixes cannot be used in |
6320 | conjunction with the base suffixes mentioned above. | |
6321 | ||
0c71d759 NC |
6322 | @node Symbolic Constants |
6323 | @subsection Symbolic Constants | |
6324 | @cindex symbolic constants | |
6325 | @kindex CONSTANT | |
a094d01f | 6326 | It is possible to refer to target-specific constants via the use of |
0c71d759 NC |
6327 | the @code{CONSTANT(@var{name})} operator, where @var{name} is one of: |
6328 | ||
6329 | @table @code | |
6330 | @item MAXPAGESIZE | |
6331 | @kindex MAXPAGESIZE | |
6332 | The target's maximum page size. | |
6333 | ||
6334 | @item COMMONPAGESIZE | |
6335 | @kindex COMMONPAGESIZE | |
6336 | The target's default page size. | |
6337 | @end table | |
6338 | ||
6339 | So for example: | |
6340 | ||
6341 | @smallexample | |
9d5777a3 | 6342 | .text ALIGN (CONSTANT (MAXPAGESIZE)) : @{ *(.text) @} |
0c71d759 NC |
6343 | @end smallexample |
6344 | ||
6345 | will create a text section aligned to the largest page boundary | |
6346 | supported by the target. | |
6347 | ||
252b5132 RH |
6348 | @node Symbols |
6349 | @subsection Symbol Names | |
6350 | @cindex symbol names | |
6351 | @cindex names | |
6352 | @cindex quoted symbol names | |
6353 | @kindex " | |
6354 | Unless quoted, symbol names start with a letter, underscore, or period | |
6355 | and may include letters, digits, underscores, periods, and hyphens. | |
6356 | Unquoted symbol names must not conflict with any keywords. You can | |
6357 | specify a symbol which contains odd characters or has the same name as a | |
6358 | keyword by surrounding the symbol name in double quotes: | |
6359 | @smallexample | |
36f63dca NC |
6360 | "SECTION" = 9; |
6361 | "with a space" = "also with a space" + 10; | |
252b5132 RH |
6362 | @end smallexample |
6363 | ||
6364 | Since symbols can contain many non-alphabetic characters, it is safest | |
6365 | to delimit symbols with spaces. For example, @samp{A-B} is one symbol, | |
6366 | whereas @samp{A - B} is an expression involving subtraction. | |
6367 | ||
ecca9871 L |
6368 | @node Orphan Sections |
6369 | @subsection Orphan Sections | |
6370 | @cindex orphan | |
6371 | Orphan sections are sections present in the input files which | |
6372 | are not explicitly placed into the output file by the linker | |
6373 | script. The linker will still copy these sections into the | |
a87ded7b AB |
6374 | output file by either finding, or creating a suitable output section |
6375 | in which to place the orphaned input section. | |
6376 | ||
6377 | If the name of an orphaned input section exactly matches the name of | |
6378 | an existing output section, then the orphaned input section will be | |
6379 | placed at the end of that output section. | |
6380 | ||
6381 | If there is no output section with a matching name then new output | |
6382 | sections will be created. Each new output section will have the same | |
6383 | name as the orphan section placed within it. If there are multiple | |
6384 | orphan sections with the same name, these will all be combined into | |
6385 | one new output section. | |
6386 | ||
6387 | If new output sections are created to hold orphaned input sections, | |
6388 | then the linker must decide where to place these new output sections | |
e299b355 AM |
6389 | in relation to existing output sections. On most modern targets, the |
6390 | linker attempts to place orphan sections after sections of the same | |
6391 | attribute, such as code vs data, loadable vs non-loadable, etc. If no | |
6392 | sections with matching attributes are found, or your target lacks this | |
6393 | support, the orphan section is placed at the end of the file. | |
ecca9871 | 6394 | |
a05a5b64 TP |
6395 | The command-line options @samp{--orphan-handling} and @samp{--unique} |
6396 | (@pxref{Options,,Command-line Options}) can be used to control which | |
c005eb9e AB |
6397 | output sections an orphan is placed in. |
6398 | ||
252b5132 RH |
6399 | @node Location Counter |
6400 | @subsection The Location Counter | |
6401 | @kindex . | |
6402 | @cindex dot | |
6403 | @cindex location counter | |
6404 | @cindex current output location | |
6405 | The special linker variable @dfn{dot} @samp{.} always contains the | |
6406 | current output location counter. Since the @code{.} always refers to a | |
6407 | location in an output section, it may only appear in an expression | |
6408 | within a @code{SECTIONS} command. The @code{.} symbol may appear | |
6409 | anywhere that an ordinary symbol is allowed in an expression. | |
6410 | ||
6411 | @cindex holes | |
6412 | Assigning a value to @code{.} will cause the location counter to be | |
6413 | moved. This may be used to create holes in the output section. The | |
dc0b6aa0 AM |
6414 | location counter may not be moved backwards inside an output section, |
6415 | and may not be moved backwards outside of an output section if so | |
6416 | doing creates areas with overlapping LMAs. | |
252b5132 RH |
6417 | |
6418 | @smallexample | |
6419 | SECTIONS | |
6420 | @{ | |
6421 | output : | |
6422 | @{ | |
6423 | file1(.text) | |
6424 | . = . + 1000; | |
6425 | file2(.text) | |
6426 | . += 1000; | |
6427 | file3(.text) | |
563e308f | 6428 | @} = 0x12345678; |
252b5132 RH |
6429 | @} |
6430 | @end smallexample | |
6431 | @noindent | |
6432 | In the previous example, the @samp{.text} section from @file{file1} is | |
6433 | located at the beginning of the output section @samp{output}. It is | |
6434 | followed by a 1000 byte gap. Then the @samp{.text} section from | |
6435 | @file{file2} appears, also with a 1000 byte gap following before the | |
563e308f | 6436 | @samp{.text} section from @file{file3}. The notation @samp{= 0x12345678} |
252b5132 RH |
6437 | specifies what data to write in the gaps (@pxref{Output Section Fill}). |
6438 | ||
5c6bbab8 NC |
6439 | @cindex dot inside sections |
6440 | Note: @code{.} actually refers to the byte offset from the start of the | |
6441 | current containing object. Normally this is the @code{SECTIONS} | |
69da35b5 | 6442 | statement, whose start address is 0, hence @code{.} can be used as an |
5c6bbab8 NC |
6443 | absolute address. If @code{.} is used inside a section description |
6444 | however, it refers to the byte offset from the start of that section, | |
6445 | not an absolute address. Thus in a script like this: | |
6446 | ||
6447 | @smallexample | |
6448 | SECTIONS | |
6449 | @{ | |
6450 | . = 0x100 | |
6451 | .text: @{ | |
6452 | *(.text) | |
6453 | . = 0x200 | |
6454 | @} | |
6455 | . = 0x500 | |
6456 | .data: @{ | |
6457 | *(.data) | |
6458 | . += 0x600 | |
6459 | @} | |
6460 | @} | |
6461 | @end smallexample | |
6462 | ||
6463 | The @samp{.text} section will be assigned a starting address of 0x100 | |
6464 | and a size of exactly 0x200 bytes, even if there is not enough data in | |
6465 | the @samp{.text} input sections to fill this area. (If there is too | |
6466 | much data, an error will be produced because this would be an attempt to | |
6467 | move @code{.} backwards). The @samp{.data} section will start at 0x500 | |
6468 | and it will have an extra 0x600 bytes worth of space after the end of | |
6469 | the values from the @samp{.data} input sections and before the end of | |
6470 | the @samp{.data} output section itself. | |
6471 | ||
b5666f2f AM |
6472 | @cindex dot outside sections |
6473 | Setting symbols to the value of the location counter outside of an | |
6474 | output section statement can result in unexpected values if the linker | |
6475 | needs to place orphan sections. For example, given the following: | |
6476 | ||
6477 | @smallexample | |
6478 | SECTIONS | |
6479 | @{ | |
6480 | start_of_text = . ; | |
6481 | .text: @{ *(.text) @} | |
6482 | end_of_text = . ; | |
6483 | ||
6484 | start_of_data = . ; | |
6485 | .data: @{ *(.data) @} | |
6486 | end_of_data = . ; | |
6487 | @} | |
6488 | @end smallexample | |
6489 | ||
6490 | If the linker needs to place some input section, e.g. @code{.rodata}, | |
6491 | not mentioned in the script, it might choose to place that section | |
6492 | between @code{.text} and @code{.data}. You might think the linker | |
6493 | should place @code{.rodata} on the blank line in the above script, but | |
6494 | blank lines are of no particular significance to the linker. As well, | |
6495 | the linker doesn't associate the above symbol names with their | |
6496 | sections. Instead, it assumes that all assignments or other | |
6497 | statements belong to the previous output section, except for the | |
6498 | special case of an assignment to @code{.}. I.e., the linker will | |
6499 | place the orphan @code{.rodata} section as if the script was written | |
6500 | as follows: | |
6501 | ||
6502 | @smallexample | |
6503 | SECTIONS | |
6504 | @{ | |
6505 | start_of_text = . ; | |
6506 | .text: @{ *(.text) @} | |
6507 | end_of_text = . ; | |
6508 | ||
6509 | start_of_data = . ; | |
6510 | .rodata: @{ *(.rodata) @} | |
6511 | .data: @{ *(.data) @} | |
6512 | end_of_data = . ; | |
6513 | @} | |
6514 | @end smallexample | |
6515 | ||
6516 | This may or may not be the script author's intention for the value of | |
6517 | @code{start_of_data}. One way to influence the orphan section | |
6518 | placement is to assign the location counter to itself, as the linker | |
6519 | assumes that an assignment to @code{.} is setting the start address of | |
6520 | a following output section and thus should be grouped with that | |
6521 | section. So you could write: | |
6522 | ||
6523 | @smallexample | |
6524 | SECTIONS | |
6525 | @{ | |
6526 | start_of_text = . ; | |
6527 | .text: @{ *(.text) @} | |
6528 | end_of_text = . ; | |
6529 | ||
6530 | . = . ; | |
6531 | start_of_data = . ; | |
6532 | .data: @{ *(.data) @} | |
6533 | end_of_data = . ; | |
6534 | @} | |
6535 | @end smallexample | |
6536 | ||
6537 | Now, the orphan @code{.rodata} section will be placed between | |
6538 | @code{end_of_text} and @code{start_of_data}. | |
6539 | ||
252b5132 RH |
6540 | @need 2000 |
6541 | @node Operators | |
6542 | @subsection Operators | |
6543 | @cindex operators for arithmetic | |
6544 | @cindex arithmetic operators | |
6545 | @cindex precedence in expressions | |
6546 | The linker recognizes the standard C set of arithmetic operators, with | |
6547 | the standard bindings and precedence levels: | |
6548 | @c TEXI2ROFF-KILL | |
36f63dca | 6549 | @ifnottex |
252b5132 RH |
6550 | @c END TEXI2ROFF-KILL |
6551 | @smallexample | |
6552 | precedence associativity Operators Notes | |
6553 | (highest) | |
6554 | 1 left ! - ~ (1) | |
6555 | 2 left * / % | |
6556 | 3 left + - | |
6557 | 4 left >> << | |
6558 | 5 left == != > < <= >= | |
6559 | 6 left & | |
6560 | 7 left | | |
6561 | 8 left && | |
6562 | 9 left || | |
6563 | 10 right ? : | |
6564 | 11 right &= += -= *= /= (2) | |
6565 | (lowest) | |
6566 | @end smallexample | |
6567 | Notes: | |
a1ab1d2a | 6568 | (1) Prefix operators |
252b5132 RH |
6569 | (2) @xref{Assignments}. |
6570 | @c TEXI2ROFF-KILL | |
36f63dca | 6571 | @end ifnottex |
252b5132 RH |
6572 | @tex |
6573 | \vskip \baselineskip | |
6574 | %"lispnarrowing" is the extra indent used generally for smallexample | |
6575 | \hskip\lispnarrowing\vbox{\offinterlineskip | |
6576 | \hrule | |
6577 | \halign | |
6578 | {\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ #\ \hfil&\vrule#&\strut\hfil\ {\tt #}\ \hfil&\vrule#\cr | |
6579 | height2pt&\omit&&\omit&&\omit&\cr | |
6580 | &Precedence&& Associativity &&{\rm Operators}&\cr | |
6581 | height2pt&\omit&&\omit&&\omit&\cr | |
6582 | \noalign{\hrule} | |
6583 | height2pt&\omit&&\omit&&\omit&\cr | |
6584 | &highest&&&&&\cr | |
6585 | % '176 is tilde, '~' in tt font | |
a1ab1d2a | 6586 | &1&&left&&\qquad- \char'176\ !\qquad\dag&\cr |
252b5132 RH |
6587 | &2&&left&&* / \%&\cr |
6588 | &3&&left&&+ -&\cr | |
6589 | &4&&left&&>> <<&\cr | |
6590 | &5&&left&&== != > < <= >=&\cr | |
6591 | &6&&left&&\&&\cr | |
6592 | &7&&left&&|&\cr | |
6593 | &8&&left&&{\&\&}&\cr | |
6594 | &9&&left&&||&\cr | |
6595 | &10&&right&&? :&\cr | |
6596 | &11&&right&&\qquad\&= += -= *= /=\qquad\ddag&\cr | |
6597 | &lowest&&&&&\cr | |
6598 | height2pt&\omit&&\omit&&\omit&\cr} | |
6599 | \hrule} | |
6600 | @end tex | |
6601 | @iftex | |
6602 | { | |
6603 | @obeylines@parskip=0pt@parindent=0pt | |
6604 | @dag@quad Prefix operators. | |
6605 | @ddag@quad @xref{Assignments}. | |
6606 | } | |
6607 | @end iftex | |
6608 | @c END TEXI2ROFF-KILL | |
6609 | ||
6610 | @node Evaluation | |
6611 | @subsection Evaluation | |
6612 | @cindex lazy evaluation | |
6613 | @cindex expression evaluation order | |
6614 | The linker evaluates expressions lazily. It only computes the value of | |
6615 | an expression when absolutely necessary. | |
6616 | ||
6617 | The linker needs some information, such as the value of the start | |
6618 | address of the first section, and the origins and lengths of memory | |
6619 | regions, in order to do any linking at all. These values are computed | |
6620 | as soon as possible when the linker reads in the linker script. | |
6621 | ||
6622 | However, other values (such as symbol values) are not known or needed | |
6623 | until after storage allocation. Such values are evaluated later, when | |
6624 | other information (such as the sizes of output sections) is available | |
6625 | for use in the symbol assignment expression. | |
6626 | ||
6627 | The sizes of sections cannot be known until after allocation, so | |
6628 | assignments dependent upon these are not performed until after | |
6629 | allocation. | |
6630 | ||
6631 | Some expressions, such as those depending upon the location counter | |
6632 | @samp{.}, must be evaluated during section allocation. | |
6633 | ||
6634 | If the result of an expression is required, but the value is not | |
6635 | available, then an error results. For example, a script like the | |
6636 | following | |
6637 | @smallexample | |
6638 | @group | |
6639 | SECTIONS | |
6640 | @{ | |
a1ab1d2a | 6641 | .text 9+this_isnt_constant : |
252b5132 RH |
6642 | @{ *(.text) @} |
6643 | @} | |
6644 | @end group | |
6645 | @end smallexample | |
6646 | @noindent | |
6647 | will cause the error message @samp{non constant expression for initial | |
6648 | address}. | |
6649 | ||
6650 | @node Expression Section | |
6651 | @subsection The Section of an Expression | |
6652 | @cindex expression sections | |
6653 | @cindex absolute expressions | |
6654 | @cindex relative expressions | |
6655 | @cindex absolute and relocatable symbols | |
6656 | @cindex relocatable and absolute symbols | |
6657 | @cindex symbols, relocatable and absolute | |
7542af2a AM |
6658 | Addresses and symbols may be section relative, or absolute. A section |
6659 | relative symbol is relocatable. If you request relocatable output | |
6660 | using the @samp{-r} option, a further link operation may change the | |
6661 | value of a section relative symbol. On the other hand, an absolute | |
6662 | symbol will retain the same value throughout any further link | |
6663 | operations. | |
6664 | ||
abf4be64 AM |
6665 | Some terms in linker expressions are addresses. This is true of |
6666 | section relative symbols and for builtin functions that return an | |
6667 | address, such as @code{ADDR}, @code{LOADADDR}, @code{ORIGIN} and | |
6668 | @code{SEGMENT_START}. Other terms are simply numbers, or are builtin | |
6669 | functions that return a non-address value, such as @code{LENGTH}. | |
01554a74 AM |
6670 | One complication is that unless you set @code{LD_FEATURE ("SANE_EXPR")} |
6671 | (@pxref{Miscellaneous Commands}), numbers and absolute symbols are treated | |
5c3049d2 AM |
6672 | differently depending on their location, for compatibility with older |
6673 | versions of @code{ld}. Expressions appearing outside an output | |
6674 | section definition treat all numbers as absolute addresses. | |
6675 | Expressions appearing inside an output section definition treat | |
01554a74 AM |
6676 | absolute symbols as numbers. If @code{LD_FEATURE ("SANE_EXPR")} is |
6677 | given, then absolute symbols and numbers are simply treated as numbers | |
6678 | everywhere. | |
5c3049d2 AM |
6679 | |
6680 | In the following simple example, | |
252b5132 | 6681 | |
7542af2a AM |
6682 | @smallexample |
6683 | @group | |
6684 | SECTIONS | |
6685 | @{ | |
6686 | . = 0x100; | |
6687 | __executable_start = 0x100; | |
6688 | .data : | |
6689 | @{ | |
6690 | . = 0x10; | |
6691 | __data_start = 0x10; | |
6692 | *(.data) | |
6693 | @} | |
6694 | @dots{} | |
6695 | @} | |
6696 | @end group | |
6697 | @end smallexample | |
252b5132 | 6698 | |
7542af2a AM |
6699 | both @code{.} and @code{__executable_start} are set to the absolute |
6700 | address 0x100 in the first two assignments, then both @code{.} and | |
6701 | @code{__data_start} are set to 0x10 relative to the @code{.data} | |
6702 | section in the second two assignments. | |
252b5132 | 6703 | |
5c3049d2 AM |
6704 | For expressions involving numbers, relative addresses and absolute |
6705 | addresses, ld follows these rules to evaluate terms: | |
7542af2a AM |
6706 | |
6707 | @itemize @bullet | |
6708 | @item | |
c05f749e AM |
6709 | Unary operations on an absolute address or number, and binary |
6710 | operations on two absolute addresses or two numbers, or between one | |
6711 | absolute address and a number, apply the operator to the value(s). | |
6712 | @item | |
7542af2a AM |
6713 | Unary operations on a relative address, and binary operations on two |
6714 | relative addresses in the same section or between one relative address | |
6715 | and a number, apply the operator to the offset part of the address(es). | |
6716 | @item | |
c05f749e AM |
6717 | Other binary operations, that is, between two relative addresses not |
6718 | in the same section, or between a relative address and an absolute | |
6719 | address, first convert any non-absolute term to an absolute address | |
6720 | before applying the operator. | |
7542af2a AM |
6721 | @end itemize |
6722 | ||
6723 | The result section of each sub-expression is as follows: | |
6724 | ||
6725 | @itemize @bullet | |
6726 | @item | |
6727 | An operation involving only numbers results in a number. | |
6728 | @item | |
6729 | The result of comparisons, @samp{&&} and @samp{||} is also a number. | |
6730 | @item | |
9bc8bb33 | 6731 | The result of other binary arithmetic and logical operations on two |
11e7fd74 | 6732 | relative addresses in the same section or two absolute addresses |
94b41882 AM |
6733 | (after above conversions) is also a number when |
6734 | @code{LD_FEATURE ("SANE_EXPR")} or inside an output section definition | |
6735 | but an absolute address otherwise. | |
9bc8bb33 AM |
6736 | @item |
6737 | The result of other operations on relative addresses or one | |
6738 | relative address and a number, is a relative address in the same | |
6739 | section as the relative operand(s). | |
7542af2a AM |
6740 | @item |
6741 | The result of other operations on absolute addresses (after above | |
6742 | conversions) is an absolute address. | |
6743 | @end itemize | |
252b5132 RH |
6744 | |
6745 | You can use the builtin function @code{ABSOLUTE} to force an expression | |
6746 | to be absolute when it would otherwise be relative. For example, to | |
6747 | create an absolute symbol set to the address of the end of the output | |
6748 | section @samp{.data}: | |
6749 | @smallexample | |
6750 | SECTIONS | |
6751 | @{ | |
6752 | .data : @{ *(.data) _edata = ABSOLUTE(.); @} | |
6753 | @} | |
6754 | @end smallexample | |
6755 | @noindent | |
6756 | If @samp{ABSOLUTE} were not used, @samp{_edata} would be relative to the | |
6757 | @samp{.data} section. | |
6758 | ||
7542af2a AM |
6759 | Using @code{LOADADDR} also forces an expression absolute, since this |
6760 | particular builtin function returns an absolute address. | |
6761 | ||
252b5132 RH |
6762 | @node Builtin Functions |
6763 | @subsection Builtin Functions | |
6764 | @cindex functions in expressions | |
6765 | The linker script language includes a number of builtin functions for | |
6766 | use in linker script expressions. | |
6767 | ||
6768 | @table @code | |
6769 | @item ABSOLUTE(@var{exp}) | |
6770 | @kindex ABSOLUTE(@var{exp}) | |
6771 | @cindex expression, absolute | |
6772 | Return the absolute (non-relocatable, as opposed to non-negative) value | |
6773 | of the expression @var{exp}. Primarily useful to assign an absolute | |
6774 | value to a symbol within a section definition, where symbol values are | |
6775 | normally section relative. @xref{Expression Section}. | |
6776 | ||
6777 | @item ADDR(@var{section}) | |
6778 | @kindex ADDR(@var{section}) | |
6779 | @cindex section address in expression | |
7542af2a | 6780 | Return the address (VMA) of the named @var{section}. Your |
252b5132 | 6781 | script must previously have defined the location of that section. In |
7542af2a AM |
6782 | the following example, @code{start_of_output_1}, @code{symbol_1} and |
6783 | @code{symbol_2} are assigned equivalent values, except that | |
6784 | @code{symbol_1} will be relative to the @code{.output1} section while | |
6785 | the other two will be absolute: | |
252b5132 RH |
6786 | @smallexample |
6787 | @group | |
6788 | SECTIONS @{ @dots{} | |
6789 | .output1 : | |
a1ab1d2a | 6790 | @{ |
252b5132 RH |
6791 | start_of_output_1 = ABSOLUTE(.); |
6792 | @dots{} | |
6793 | @} | |
6794 | .output : | |
6795 | @{ | |
6796 | symbol_1 = ADDR(.output1); | |
6797 | symbol_2 = start_of_output_1; | |
6798 | @} | |
6799 | @dots{} @} | |
6800 | @end group | |
6801 | @end smallexample | |
6802 | ||
876f4090 NS |
6803 | @item ALIGN(@var{align}) |
6804 | @itemx ALIGN(@var{exp},@var{align}) | |
6805 | @kindex ALIGN(@var{align}) | |
6806 | @kindex ALIGN(@var{exp},@var{align}) | |
252b5132 RH |
6807 | @cindex round up location counter |
6808 | @cindex align location counter | |
876f4090 NS |
6809 | @cindex round up expression |
6810 | @cindex align expression | |
6811 | Return the location counter (@code{.}) or arbitrary expression aligned | |
6812 | to the next @var{align} boundary. The single operand @code{ALIGN} | |
6813 | doesn't change the value of the location counter---it just does | |
6814 | arithmetic on it. The two operand @code{ALIGN} allows an arbitrary | |
6815 | expression to be aligned upwards (@code{ALIGN(@var{align})} is | |
e0a3af22 | 6816 | equivalent to @code{ALIGN(ABSOLUTE(.), @var{align})}). |
876f4090 NS |
6817 | |
6818 | Here is an example which aligns the output @code{.data} section to the | |
6819 | next @code{0x2000} byte boundary after the preceding section and sets a | |
6820 | variable within the section to the next @code{0x8000} boundary after the | |
6821 | input sections: | |
252b5132 RH |
6822 | @smallexample |
6823 | @group | |
6824 | SECTIONS @{ @dots{} | |
6825 | .data ALIGN(0x2000): @{ | |
6826 | *(.data) | |
6827 | variable = ALIGN(0x8000); | |
6828 | @} | |
6829 | @dots{} @} | |
6830 | @end group | |
6831 | @end smallexample | |
6832 | @noindent | |
6833 | The first use of @code{ALIGN} in this example specifies the location of | |
6834 | a section because it is used as the optional @var{address} attribute of | |
6835 | a section definition (@pxref{Output Section Address}). The second use | |
6836 | of @code{ALIGN} is used to defines the value of a symbol. | |
6837 | ||
6838 | The builtin function @code{NEXT} is closely related to @code{ALIGN}. | |
6839 | ||
362c1d1a NS |
6840 | @item ALIGNOF(@var{section}) |
6841 | @kindex ALIGNOF(@var{section}) | |
6842 | @cindex section alignment | |
6843 | Return the alignment in bytes of the named @var{section}, if that section has | |
6844 | been allocated. If the section has not been allocated when this is | |
6845 | evaluated, the linker will report an error. In the following example, | |
6846 | the alignment of the @code{.output} section is stored as the first | |
6847 | value in that section. | |
6848 | @smallexample | |
6849 | @group | |
6850 | SECTIONS@{ @dots{} | |
6851 | .output @{ | |
6852 | LONG (ALIGNOF (.output)) | |
6853 | @dots{} | |
6854 | @} | |
6855 | @dots{} @} | |
6856 | @end group | |
6857 | @end smallexample | |
6858 | ||
252b5132 RH |
6859 | @item BLOCK(@var{exp}) |
6860 | @kindex BLOCK(@var{exp}) | |
6861 | This is a synonym for @code{ALIGN}, for compatibility with older linker | |
6862 | scripts. It is most often seen when setting the address of an output | |
6863 | section. | |
6864 | ||
2d20f7bf JJ |
6865 | @item DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize}) |
6866 | @kindex DATA_SEGMENT_ALIGN(@var{maxpagesize}, @var{commonpagesize}) | |
6867 | This is equivalent to either | |
6868 | @smallexample | |
6869 | (ALIGN(@var{maxpagesize}) + (. & (@var{maxpagesize} - 1))) | |
6870 | @end smallexample | |
6871 | or | |
6872 | @smallexample | |
fe6052e1 AM |
6873 | (ALIGN(@var{maxpagesize}) |
6874 | + ((. + @var{commonpagesize} - 1) & (@var{maxpagesize} - @var{commonpagesize}))) | |
2d20f7bf JJ |
6875 | @end smallexample |
6876 | @noindent | |
6877 | depending on whether the latter uses fewer @var{commonpagesize} sized pages | |
6878 | for the data segment (area between the result of this expression and | |
6879 | @code{DATA_SEGMENT_END}) than the former or not. | |
6880 | If the latter form is used, it means @var{commonpagesize} bytes of runtime | |
6881 | memory will be saved at the expense of up to @var{commonpagesize} wasted | |
6882 | bytes in the on-disk file. | |
6883 | ||
6884 | This expression can only be used directly in @code{SECTIONS} commands, not in | |
6885 | any output section descriptions and only once in the linker script. | |
6886 | @var{commonpagesize} should be less or equal to @var{maxpagesize} and should | |
def5c83c AM |
6887 | be the system page size the object wants to be optimized for while still |
6888 | running on system page sizes up to @var{maxpagesize}. Note however | |
6889 | that @samp{-z relro} protection will not be effective if the system | |
6890 | page size is larger than @var{commonpagesize}. | |
2d20f7bf JJ |
6891 | |
6892 | @noindent | |
6893 | Example: | |
6894 | @smallexample | |
6895 | . = DATA_SEGMENT_ALIGN(0x10000, 0x2000); | |
6896 | @end smallexample | |
6897 | ||
6898 | @item DATA_SEGMENT_END(@var{exp}) | |
6899 | @kindex DATA_SEGMENT_END(@var{exp}) | |
6900 | This defines the end of data segment for @code{DATA_SEGMENT_ALIGN} | |
6901 | evaluation purposes. | |
6902 | ||
6903 | @smallexample | |
6904 | . = DATA_SEGMENT_END(.); | |
6905 | @end smallexample | |
6906 | ||
a4f5ad88 JJ |
6907 | @item DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp}) |
6908 | @kindex DATA_SEGMENT_RELRO_END(@var{offset}, @var{exp}) | |
6909 | This defines the end of the @code{PT_GNU_RELRO} segment when | |
eec2f3ed | 6910 | @samp{-z relro} option is used. |
a4f5ad88 JJ |
6911 | When @samp{-z relro} option is not present, @code{DATA_SEGMENT_RELRO_END} |
6912 | does nothing, otherwise @code{DATA_SEGMENT_ALIGN} is padded so that | |
def5c83c AM |
6913 | @var{exp} + @var{offset} is aligned to the @var{commonpagesize} |
6914 | argument given to @code{DATA_SEGMENT_ALIGN}. If present in the linker | |
6915 | script, it must be placed between @code{DATA_SEGMENT_ALIGN} and | |
eec2f3ed AM |
6916 | @code{DATA_SEGMENT_END}. Evaluates to the second argument plus any |
6917 | padding needed at the end of the @code{PT_GNU_RELRO} segment due to | |
6918 | section alignment. | |
a4f5ad88 JJ |
6919 | |
6920 | @smallexample | |
6921 | . = DATA_SEGMENT_RELRO_END(24, .); | |
6922 | @end smallexample | |
6923 | ||
252b5132 RH |
6924 | @item DEFINED(@var{symbol}) |
6925 | @kindex DEFINED(@var{symbol}) | |
6926 | @cindex symbol defaults | |
6927 | Return 1 if @var{symbol} is in the linker global symbol table and is | |
420e579c HPN |
6928 | defined before the statement using DEFINED in the script, otherwise |
6929 | return 0. You can use this function to provide | |
252b5132 RH |
6930 | default values for symbols. For example, the following script fragment |
6931 | shows how to set a global symbol @samp{begin} to the first location in | |
6932 | the @samp{.text} section---but if a symbol called @samp{begin} already | |
6933 | existed, its value is preserved: | |
6934 | ||
6935 | @smallexample | |
6936 | @group | |
6937 | SECTIONS @{ @dots{} | |
6938 | .text : @{ | |
6939 | begin = DEFINED(begin) ? begin : . ; | |
6940 | @dots{} | |
6941 | @} | |
6942 | @dots{} | |
6943 | @} | |
6944 | @end group | |
6945 | @end smallexample | |
6946 | ||
3ec57632 NC |
6947 | @item LENGTH(@var{memory}) |
6948 | @kindex LENGTH(@var{memory}) | |
6949 | Return the length of the memory region named @var{memory}. | |
6950 | ||
252b5132 RH |
6951 | @item LOADADDR(@var{section}) |
6952 | @kindex LOADADDR(@var{section}) | |
6953 | @cindex section load address in expression | |
7542af2a | 6954 | Return the absolute LMA of the named @var{section}. (@pxref{Output |
252b5132 RH |
6955 | Section LMA}). |
6956 | ||
2e53f7d6 NC |
6957 | @item LOG2CEIL(@var{exp}) |
6958 | @kindex LOG2CEIL(@var{exp}) | |
6959 | Return the binary logarithm of @var{exp} rounded towards infinity. | |
6960 | @code{LOG2CEIL(0)} returns 0. | |
6961 | ||
252b5132 RH |
6962 | @kindex MAX |
6963 | @item MAX(@var{exp1}, @var{exp2}) | |
6964 | Returns the maximum of @var{exp1} and @var{exp2}. | |
6965 | ||
6966 | @kindex MIN | |
6967 | @item MIN(@var{exp1}, @var{exp2}) | |
6968 | Returns the minimum of @var{exp1} and @var{exp2}. | |
6969 | ||
6970 | @item NEXT(@var{exp}) | |
6971 | @kindex NEXT(@var{exp}) | |
6972 | @cindex unallocated address, next | |
6973 | Return the next unallocated address that is a multiple of @var{exp}. | |
6974 | This function is closely related to @code{ALIGN(@var{exp})}; unless you | |
6975 | use the @code{MEMORY} command to define discontinuous memory for the | |
6976 | output file, the two functions are equivalent. | |
6977 | ||
3ec57632 NC |
6978 | @item ORIGIN(@var{memory}) |
6979 | @kindex ORIGIN(@var{memory}) | |
6980 | Return the origin of the memory region named @var{memory}. | |
6981 | ||
ba916c8a MM |
6982 | @item SEGMENT_START(@var{segment}, @var{default}) |
6983 | @kindex SEGMENT_START(@var{segment}, @var{default}) | |
6984 | Return the base address of the named @var{segment}. If an explicit | |
c5da8c7d NC |
6985 | value has already been given for this segment (with a command-line |
6986 | @samp{-T} option) then that value will be returned otherwise the value | |
6987 | will be @var{default}. At present, the @samp{-T} command-line option | |
6988 | can only be used to set the base address for the ``text'', ``data'', and | |
7542af2a | 6989 | ``bss'' sections, but you can use @code{SEGMENT_START} with any segment |
ba916c8a MM |
6990 | name. |
6991 | ||
252b5132 RH |
6992 | @item SIZEOF(@var{section}) |
6993 | @kindex SIZEOF(@var{section}) | |
6994 | @cindex section size | |
6995 | Return the size in bytes of the named @var{section}, if that section has | |
6996 | been allocated. If the section has not been allocated when this is | |
6997 | evaluated, the linker will report an error. In the following example, | |
6998 | @code{symbol_1} and @code{symbol_2} are assigned identical values: | |
6999 | @smallexample | |
7000 | @group | |
7001 | SECTIONS@{ @dots{} | |
7002 | .output @{ | |
7003 | .start = . ; | |
7004 | @dots{} | |
7005 | .end = . ; | |
7006 | @} | |
7007 | symbol_1 = .end - .start ; | |
7008 | symbol_2 = SIZEOF(.output); | |
7009 | @dots{} @} | |
7010 | @end group | |
7011 | @end smallexample | |
7012 | ||
7013 | @item SIZEOF_HEADERS | |
7014 | @itemx sizeof_headers | |
7015 | @kindex SIZEOF_HEADERS | |
7016 | @cindex header size | |
7017 | Return the size in bytes of the output file's headers. This is | |
7018 | information which appears at the start of the output file. You can use | |
7019 | this number when setting the start address of the first section, if you | |
7020 | choose, to facilitate paging. | |
7021 | ||
7022 | @cindex not enough room for program headers | |
7023 | @cindex program headers, not enough room | |
7024 | When producing an ELF output file, if the linker script uses the | |
7025 | @code{SIZEOF_HEADERS} builtin function, the linker must compute the | |
7026 | number of program headers before it has determined all the section | |
7027 | addresses and sizes. If the linker later discovers that it needs | |
7028 | additional program headers, it will report an error @samp{not enough | |
7029 | room for program headers}. To avoid this error, you must avoid using | |
7030 | the @code{SIZEOF_HEADERS} function, or you must rework your linker | |
7031 | script to avoid forcing the linker to use additional program headers, or | |
7032 | you must define the program headers yourself using the @code{PHDRS} | |
7033 | command (@pxref{PHDRS}). | |
7034 | @end table | |
7035 | ||
7036 | @node Implicit Linker Scripts | |
7037 | @section Implicit Linker Scripts | |
7038 | @cindex implicit linker scripts | |
7039 | If you specify a linker input file which the linker can not recognize as | |
7040 | an object file or an archive file, it will try to read the file as a | |
7041 | linker script. If the file can not be parsed as a linker script, the | |
7042 | linker will report an error. | |
7043 | ||
7044 | An implicit linker script will not replace the default linker script. | |
7045 | ||
7046 | Typically an implicit linker script would contain only symbol | |
7047 | assignments, or the @code{INPUT}, @code{GROUP}, or @code{VERSION} | |
7048 | commands. | |
7049 | ||
7050 | Any input files read because of an implicit linker script will be read | |
7051 | at the position in the command line where the implicit linker script was | |
7052 | read. This can affect archive searching. | |
7053 | ||
bf6d8037 HC |
7054 | @node Plugins |
7055 | @chapter Linker Plugins | |
7056 | ||
7057 | @cindex plugins | |
7058 | @cindex linker plugins | |
7059 | The linker can use dynamically loaded plugins to modify its behavior. | |
7060 | For example, the link-time optimization feature that some compilers | |
7061 | support is implemented with a linker plugin. | |
7062 | ||
7063 | Currently there is only one plugin shipped by default, but more may | |
7064 | be added here later. | |
7065 | ||
7066 | @menu | |
7067 | * libdep Plugin:: Static Library Dependencies Plugin | |
7068 | @end menu | |
7069 | ||
7070 | @node libdep Plugin | |
7071 | @section Static Library Dependencies Plugin | |
7072 | @cindex static library dependencies | |
7073 | Originally, static libraries were contained in an archive file consisting | |
7074 | just of a collection of relocatable object files. Later they evolved to | |
7075 | optionally include a symbol table, to assist in finding the needed objects | |
7076 | within a library. There their evolution ended, and dynamic libraries | |
7077 | rose to ascendance. | |
7078 | ||
7079 | One useful feature of dynamic libraries was that, more than just collecting | |
7080 | multiple objects into a single file, they also included a list of their | |
7081 | dependencies, such that one could specify just the name of a single dynamic | |
7082 | library at link time, and all of its dependencies would be implicitly | |
7083 | referenced as well. But static libraries lacked this feature, so if a | |
7084 | link invocation was switched from using dynamic libraries to static | |
7085 | libraries, the link command would usually fail unless it was rewritten to | |
7086 | explicitly list the dependencies of the static library. | |
7087 | ||
7088 | The GNU @command{ar} utility now supports a @option{--record-libdeps} option | |
7089 | to embed dependency lists into static libraries as well, and the @file{libdep} | |
7090 | plugin may be used to read this dependency information at link time. The | |
7091 | dependency information is stored as a single string, carrying @option{-l} | |
7092 | and @option{-L} arguments as they would normally appear in a linker | |
7093 | command line. As such, the information can be written with any text | |
7094 | utility and stored into any archive, even if GNU @command{ar} is not | |
7095 | being used to create the archive. The information is stored in an | |
7096 | archive member named @samp{__.LIBDEP}. | |
7097 | ||
7098 | For example, given a library @file{libssl.a} that depends on another | |
7099 | library @file{libcrypto.a} which may be found in @file{/usr/local/lib}, | |
7100 | the @samp{__.LIBDEP} member of @file{libssl.a} would contain | |
7101 | ||
7102 | @smallexample | |
7103 | -L/usr/local/lib -lcrypto | |
7104 | @end smallexample | |
7105 | ||
252b5132 RH |
7106 | @ifset GENERIC |
7107 | @node Machine Dependent | |
7108 | @chapter Machine Dependent Features | |
7109 | ||
7110 | @cindex machine dependencies | |
ff5dcc92 SC |
7111 | @command{ld} has additional features on some platforms; the following |
7112 | sections describe them. Machines where @command{ld} has no additional | |
252b5132 RH |
7113 | functionality are not listed. |
7114 | ||
7115 | @menu | |
36f63dca NC |
7116 | @ifset H8300 |
7117 | * H8/300:: @command{ld} and the H8/300 | |
7118 | @end ifset | |
7ca01ed9 NC |
7119 | @ifset M68HC11 |
7120 | * M68HC11/68HC12:: @code{ld} and the Motorola 68HC11 and 68HC12 families | |
7121 | @end ifset | |
36f63dca NC |
7122 | @ifset ARM |
7123 | * ARM:: @command{ld} and the ARM family | |
7124 | @end ifset | |
7125 | @ifset HPPA | |
7126 | * HPPA ELF32:: @command{ld} and HPPA 32-bit ELF | |
7127 | @end ifset | |
7fb9f789 NC |
7128 | @ifset M68K |
7129 | * M68K:: @command{ld} and the Motorola 68K family | |
7130 | @end ifset | |
833794fc MR |
7131 | @ifset MIPS |
7132 | * MIPS:: @command{ld} and the MIPS family | |
7133 | @end ifset | |
3c3bdf30 | 7134 | @ifset MMIX |
36f63dca | 7135 | * MMIX:: @command{ld} and MMIX |
3c3bdf30 | 7136 | @end ifset |
2469cfa2 | 7137 | @ifset MSP430 |
36f63dca | 7138 | * MSP430:: @command{ld} and MSP430 |
2469cfa2 | 7139 | @end ifset |
35c08157 KLC |
7140 | @ifset NDS32 |
7141 | * NDS32:: @command{ld} and NDS32 | |
7142 | @end ifset | |
78058a5e SL |
7143 | @ifset NIOSII |
7144 | * Nios II:: @command{ld} and the Altera Nios II | |
7145 | @end ifset | |
2a60a7a8 AM |
7146 | @ifset POWERPC |
7147 | * PowerPC ELF32:: @command{ld} and PowerPC 32-bit ELF Support | |
7148 | @end ifset | |
7149 | @ifset POWERPC64 | |
7150 | * PowerPC64 ELF64:: @command{ld} and PowerPC64 64-bit ELF Support | |
7151 | @end ifset | |
b4cbbe8f AK |
7152 | @ifset S/390 |
7153 | * S/390 ELF:: @command{ld} and S/390 ELF Support | |
7154 | @end ifset | |
49fa1e15 AM |
7155 | @ifset SPU |
7156 | * SPU ELF:: @command{ld} and SPU ELF Support | |
7157 | @end ifset | |
74459f0e | 7158 | @ifset TICOFF |
ff5dcc92 | 7159 | * TI COFF:: @command{ld} and TI COFF |
74459f0e | 7160 | @end ifset |
2ca22b03 NC |
7161 | @ifset WIN32 |
7162 | * WIN32:: @command{ld} and WIN32 (cygwin/mingw) | |
7163 | @end ifset | |
e0001a05 NC |
7164 | @ifset XTENSA |
7165 | * Xtensa:: @command{ld} and Xtensa Processors | |
7166 | @end ifset | |
252b5132 RH |
7167 | @end menu |
7168 | @end ifset | |
7169 | ||
252b5132 RH |
7170 | @ifset H8300 |
7171 | @ifclear GENERIC | |
7172 | @raisesections | |
7173 | @end ifclear | |
7174 | ||
7175 | @node H8/300 | |
ff5dcc92 | 7176 | @section @command{ld} and the H8/300 |
252b5132 RH |
7177 | |
7178 | @cindex H8/300 support | |
ff5dcc92 | 7179 | For the H8/300, @command{ld} can perform these global optimizations when |
252b5132 RH |
7180 | you specify the @samp{--relax} command-line option. |
7181 | ||
7182 | @table @emph | |
7183 | @cindex relaxing on H8/300 | |
7184 | @item relaxing address modes | |
ff5dcc92 | 7185 | @command{ld} finds all @code{jsr} and @code{jmp} instructions whose |
252b5132 RH |
7186 | targets are within eight bits, and turns them into eight-bit |
7187 | program-counter relative @code{bsr} and @code{bra} instructions, | |
7188 | respectively. | |
7189 | ||
7190 | @cindex synthesizing on H8/300 | |
7191 | @item synthesizing instructions | |
81f5558e | 7192 | @c FIXME: specifically mov.b, or any mov instructions really? -> mov.b only, at least on H8, H8H, H8S |
ff5dcc92 | 7193 | @command{ld} finds all @code{mov.b} instructions which use the |
252b5132 RH |
7194 | sixteen-bit absolute address form, but refer to the top |
7195 | page of memory, and changes them to use the eight-bit address form. | |
7196 | (That is: the linker turns @samp{mov.b @code{@@}@var{aa}:16} into | |
7197 | @samp{mov.b @code{@@}@var{aa}:8} whenever the address @var{aa} is in the | |
7198 | top page of memory). | |
1502569c | 7199 | |
81f5558e NC |
7200 | @command{ld} finds all @code{mov} instructions which use the register |
7201 | indirect with 32-bit displacement addressing mode, but use a small | |
7202 | displacement inside 16-bit displacement range, and changes them to use | |
7203 | the 16-bit displacement form. (That is: the linker turns @samp{mov.b | |
7204 | @code{@@}@var{d}:32,ERx} into @samp{mov.b @code{@@}@var{d}:16,ERx} | |
7205 | whenever the displacement @var{d} is in the 16 bit signed integer | |
7206 | range. Only implemented in ELF-format ld). | |
7207 | ||
1502569c | 7208 | @item bit manipulation instructions |
c0065db7 | 7209 | @command{ld} finds all bit manipulation instructions like @code{band, bclr, |
1502569c | 7210 | biand, bild, bior, bist, bixor, bld, bnot, bor, bset, bst, btst, bxor} |
c0065db7 | 7211 | which use 32 bit and 16 bit absolute address form, but refer to the top |
1502569c NC |
7212 | page of memory, and changes them to use the 8 bit address form. |
7213 | (That is: the linker turns @samp{bset #xx:3,@code{@@}@var{aa}:32} into | |
c0065db7 | 7214 | @samp{bset #xx:3,@code{@@}@var{aa}:8} whenever the address @var{aa} is in |
1502569c NC |
7215 | the top page of memory). |
7216 | ||
7217 | @item system control instructions | |
c0065db7 RM |
7218 | @command{ld} finds all @code{ldc.w, stc.w} instructions which use the |
7219 | 32 bit absolute address form, but refer to the top page of memory, and | |
1502569c NC |
7220 | changes them to use 16 bit address form. |
7221 | (That is: the linker turns @samp{ldc.w @code{@@}@var{aa}:32,ccr} into | |
c0065db7 | 7222 | @samp{ldc.w @code{@@}@var{aa}:16,ccr} whenever the address @var{aa} is in |
1502569c | 7223 | the top page of memory). |
252b5132 RH |
7224 | @end table |
7225 | ||
7226 | @ifclear GENERIC | |
7227 | @lowersections | |
7228 | @end ifclear | |
7229 | @end ifset | |
7230 | ||
36f63dca | 7231 | @ifclear GENERIC |
c2dcd04e | 7232 | @ifset Renesas |
36f63dca | 7233 | @c This stuff is pointless to say unless you're especially concerned |
c2dcd04e NC |
7234 | @c with Renesas chips; don't enable it for generic case, please. |
7235 | @node Renesas | |
7236 | @chapter @command{ld} and Other Renesas Chips | |
36f63dca | 7237 | |
c2dcd04e NC |
7238 | @command{ld} also supports the Renesas (formerly Hitachi) H8/300H, |
7239 | H8/500, and SH chips. No special features, commands, or command-line | |
7240 | options are required for these chips. | |
36f63dca NC |
7241 | @end ifset |
7242 | @end ifclear | |
7243 | ||
36f63dca NC |
7244 | @ifset ARM |
7245 | @ifclear GENERIC | |
7246 | @raisesections | |
7247 | @end ifclear | |
7248 | ||
93fd0973 SC |
7249 | @ifset M68HC11 |
7250 | @ifclear GENERIC | |
7251 | @raisesections | |
7252 | @end ifclear | |
7253 | ||
7254 | @node M68HC11/68HC12 | |
7255 | @section @command{ld} and the Motorola 68HC11 and 68HC12 families | |
7256 | ||
7257 | @cindex M68HC11 and 68HC12 support | |
7258 | ||
7259 | @subsection Linker Relaxation | |
7260 | ||
7261 | For the Motorola 68HC11, @command{ld} can perform these global | |
7262 | optimizations when you specify the @samp{--relax} command-line option. | |
7263 | ||
7264 | @table @emph | |
7265 | @cindex relaxing on M68HC11 | |
7266 | @item relaxing address modes | |
7267 | @command{ld} finds all @code{jsr} and @code{jmp} instructions whose | |
7268 | targets are within eight bits, and turns them into eight-bit | |
7269 | program-counter relative @code{bsr} and @code{bra} instructions, | |
7270 | respectively. | |
7271 | ||
7272 | @command{ld} also looks at all 16-bit extended addressing modes and | |
7273 | transforms them in a direct addressing mode when the address is in | |
7274 | page 0 (between 0 and 0x0ff). | |
7275 | ||
7276 | @item relaxing gcc instruction group | |
7277 | When @command{gcc} is called with @option{-mrelax}, it can emit group | |
7278 | of instructions that the linker can optimize to use a 68HC11 direct | |
7279 | addressing mode. These instructions consists of @code{bclr} or | |
7280 | @code{bset} instructions. | |
7281 | ||
7282 | @end table | |
7283 | ||
7284 | @subsection Trampoline Generation | |
7285 | ||
7286 | @cindex trampoline generation on M68HC11 | |
7287 | @cindex trampoline generation on M68HC12 | |
7288 | For 68HC11 and 68HC12, @command{ld} can generate trampoline code to | |
7289 | call a far function using a normal @code{jsr} instruction. The linker | |
c0065db7 | 7290 | will also change the relocation to some far function to use the |
93fd0973 SC |
7291 | trampoline address instead of the function address. This is typically the |
7292 | case when a pointer to a function is taken. The pointer will in fact | |
7293 | point to the function trampoline. | |
7294 | ||
7295 | @ifclear GENERIC | |
7296 | @lowersections | |
7297 | @end ifclear | |
7298 | @end ifset | |
7299 | ||
36f63dca | 7300 | @node ARM |
3674e28a | 7301 | @section @command{ld} and the ARM family |
36f63dca NC |
7302 | |
7303 | @cindex ARM interworking support | |
7304 | @kindex --support-old-code | |
7305 | For the ARM, @command{ld} will generate code stubs to allow functions calls | |
b45619c0 | 7306 | between ARM and Thumb code. These stubs only work with code that has |
36f63dca NC |
7307 | been compiled and assembled with the @samp{-mthumb-interwork} command |
7308 | line option. If it is necessary to link with old ARM object files or | |
7309 | libraries, which have not been compiled with the -mthumb-interwork | |
a05a5b64 | 7310 | option then the @samp{--support-old-code} command-line switch should be |
36f63dca NC |
7311 | given to the linker. This will make it generate larger stub functions |
7312 | which will work with non-interworking aware ARM code. Note, however, | |
7313 | the linker does not support generating stubs for function calls to | |
7314 | non-interworking aware Thumb code. | |
7315 | ||
7316 | @cindex thumb entry point | |
7317 | @cindex entry point, thumb | |
7318 | @kindex --thumb-entry=@var{entry} | |
7319 | The @samp{--thumb-entry} switch is a duplicate of the generic | |
7320 | @samp{--entry} switch, in that it sets the program's starting address. | |
7321 | But it also sets the bottom bit of the address, so that it can be | |
7322 | branched to using a BX instruction, and the program will start | |
7323 | executing in Thumb mode straight away. | |
7324 | ||
ce11ba6c KT |
7325 | @cindex PE import table prefixing |
7326 | @kindex --use-nul-prefixed-import-tables | |
7327 | The @samp{--use-nul-prefixed-import-tables} switch is specifying, that | |
7328 | the import tables idata4 and idata5 have to be generated with a zero | |
11e7fd74 | 7329 | element prefix for import libraries. This is the old style to generate |
ce11ba6c KT |
7330 | import tables. By default this option is turned off. |
7331 | ||
e489d0ae PB |
7332 | @cindex BE8 |
7333 | @kindex --be8 | |
7334 | The @samp{--be8} switch instructs @command{ld} to generate BE8 format | |
080bb7bb NC |
7335 | executables. This option is only valid when linking big-endian |
7336 | objects - ie ones which have been assembled with the @option{-EB} | |
7337 | option. The resulting image will contain big-endian data and | |
7338 | little-endian code. | |
e489d0ae | 7339 | |
3674e28a PB |
7340 | @cindex TARGET1 |
7341 | @kindex --target1-rel | |
7342 | @kindex --target1-abs | |
7343 | The @samp{R_ARM_TARGET1} relocation is typically used for entries in the | |
7344 | @samp{.init_array} section. It is interpreted as either @samp{R_ARM_REL32} | |
7345 | or @samp{R_ARM_ABS32}, depending on the target. The @samp{--target1-rel} | |
7346 | and @samp{--target1-abs} switches override the default. | |
7347 | ||
7348 | @cindex TARGET2 | |
7349 | @kindex --target2=@var{type} | |
7350 | The @samp{--target2=type} switch overrides the default definition of the | |
7351 | @samp{R_ARM_TARGET2} relocation. Valid values for @samp{type}, their | |
7352 | meanings, and target defaults are as follows: | |
7353 | @table @samp | |
7354 | @item rel | |
eeac373a PB |
7355 | @samp{R_ARM_REL32} (arm*-*-elf, arm*-*-eabi) |
7356 | @item abs | |
a57d1773 | 7357 | @samp{R_ARM_ABS32} |
3674e28a PB |
7358 | @item got-rel |
7359 | @samp{R_ARM_GOT_PREL} (arm*-*-linux, arm*-*-*bsd) | |
7360 | @end table | |
7361 | ||
319850b4 JB |
7362 | @cindex FIX_V4BX |
7363 | @kindex --fix-v4bx | |
7364 | The @samp{R_ARM_V4BX} relocation (defined by the ARM AAELF | |
7365 | specification) enables objects compiled for the ARMv4 architecture to be | |
7366 | interworking-safe when linked with other objects compiled for ARMv4t, but | |
7367 | also allows pure ARMv4 binaries to be built from the same ARMv4 objects. | |
7368 | ||
7369 | In the latter case, the switch @option{--fix-v4bx} must be passed to the | |
7370 | linker, which causes v4t @code{BX rM} instructions to be rewritten as | |
7371 | @code{MOV PC,rM}, since v4 processors do not have a @code{BX} instruction. | |
7372 | ||
7373 | In the former case, the switch should not be used, and @samp{R_ARM_V4BX} | |
7374 | relocations are ignored. | |
7375 | ||
845b51d6 PB |
7376 | @cindex FIX_V4BX_INTERWORKING |
7377 | @kindex --fix-v4bx-interworking | |
7378 | Replace @code{BX rM} instructions identified by @samp{R_ARM_V4BX} | |
7379 | relocations with a branch to the following veneer: | |
7380 | ||
7381 | @smallexample | |
7382 | TST rM, #1 | |
7383 | MOVEQ PC, rM | |
7384 | BX Rn | |
7385 | @end smallexample | |
7386 | ||
7387 | This allows generation of libraries/applications that work on ARMv4 cores | |
7388 | and are still interworking safe. Note that the above veneer clobbers the | |
11e7fd74 | 7389 | condition flags, so may cause incorrect program behavior in rare cases. |
845b51d6 | 7390 | |
33bfe774 JB |
7391 | @cindex USE_BLX |
7392 | @kindex --use-blx | |
7393 | The @samp{--use-blx} switch enables the linker to use ARM/Thumb | |
7394 | BLX instructions (available on ARMv5t and above) in various | |
7395 | situations. Currently it is used to perform calls via the PLT from Thumb | |
7396 | code using BLX rather than using BX and a mode-switching stub before | |
7397 | each PLT entry. This should lead to such calls executing slightly faster. | |
7398 | ||
c6dd86c6 JB |
7399 | @cindex VFP11_DENORM_FIX |
7400 | @kindex --vfp11-denorm-fix | |
7401 | The @samp{--vfp11-denorm-fix} switch enables a link-time workaround for a | |
7402 | bug in certain VFP11 coprocessor hardware, which sometimes allows | |
7403 | instructions with denorm operands (which must be handled by support code) | |
7404 | to have those operands overwritten by subsequent instructions before | |
7405 | the support code can read the intended values. | |
7406 | ||
7407 | The bug may be avoided in scalar mode if you allow at least one | |
7408 | intervening instruction between a VFP11 instruction which uses a register | |
7409 | and another instruction which writes to the same register, or at least two | |
7410 | intervening instructions if vector mode is in use. The bug only affects | |
7411 | full-compliance floating-point mode: you do not need this workaround if | |
7412 | you are using "runfast" mode. Please contact ARM for further details. | |
7413 | ||
7414 | If you know you are using buggy VFP11 hardware, you can | |
7415 | enable this workaround by specifying the linker option | |
7416 | @samp{--vfp-denorm-fix=scalar} if you are using the VFP11 scalar | |
7417 | mode only, or @samp{--vfp-denorm-fix=vector} if you are using | |
7418 | vector mode (the latter also works for scalar code). The default is | |
7419 | @samp{--vfp-denorm-fix=none}. | |
7420 | ||
7421 | If the workaround is enabled, instructions are scanned for | |
7422 | potentially-troublesome sequences, and a veneer is created for each | |
7423 | such sequence which may trigger the erratum. The veneer consists of the | |
7424 | first instruction of the sequence and a branch back to the subsequent | |
7425 | instruction. The original instruction is then replaced with a branch to | |
7426 | the veneer. The extra cycles required to call and return from the veneer | |
7427 | are sufficient to avoid the erratum in both the scalar and vector cases. | |
7428 | ||
2de70689 MGD |
7429 | @cindex ARM1176 erratum workaround |
7430 | @kindex --fix-arm1176 | |
7431 | @kindex --no-fix-arm1176 | |
9d5777a3 RM |
7432 | The @samp{--fix-arm1176} switch enables a link-time workaround for an erratum |
7433 | in certain ARM1176 processors. The workaround is enabled by default if you | |
7434 | are targeting ARM v6 (excluding ARM v6T2) or earlier. It can be disabled | |
2de70689 MGD |
7435 | unconditionally by specifying @samp{--no-fix-arm1176}. |
7436 | ||
9d5777a3 | 7437 | Further information is available in the ``ARM1176JZ-S and ARM1176JZF-S |
11e7fd74 | 7438 | Programmer Advice Notice'' available on the ARM documentation website at: |
2de70689 MGD |
7439 | http://infocenter.arm.com/. |
7440 | ||
a504d23a LA |
7441 | @cindex STM32L4xx erratum workaround |
7442 | @kindex --fix-stm32l4xx-629360 | |
7443 | ||
7444 | The @samp{--fix-stm32l4xx-629360} switch enables a link-time | |
7445 | workaround for a bug in the bus matrix / memory controller for some of | |
7446 | the STM32 Cortex-M4 based products (STM32L4xx). When accessing | |
7447 | off-chip memory via the affected bus for bus reads of 9 words or more, | |
7448 | the bus can generate corrupt data and/or abort. These are only | |
7449 | core-initiated accesses (not DMA), and might affect any access: | |
7450 | integer loads such as LDM, POP and floating-point loads such as VLDM, | |
7451 | VPOP. Stores are not affected. | |
7452 | ||
7453 | The bug can be avoided by splitting memory accesses into the | |
7454 | necessary chunks to keep bus reads below 8 words. | |
7455 | ||
7456 | The workaround is not enabled by default, this is equivalent to use | |
7457 | @samp{--fix-stm32l4xx-629360=none}. If you know you are using buggy | |
7458 | STM32L4xx hardware, you can enable the workaround by specifying the | |
7459 | linker option @samp{--fix-stm32l4xx-629360}, or the equivalent | |
7460 | @samp{--fix-stm32l4xx-629360=default}. | |
7461 | ||
7462 | If the workaround is enabled, instructions are scanned for | |
7463 | potentially-troublesome sequences, and a veneer is created for each | |
7464 | such sequence which may trigger the erratum. The veneer consists in a | |
7465 | replacement sequence emulating the behaviour of the original one and a | |
7466 | branch back to the subsequent instruction. The original instruction is | |
7467 | then replaced with a branch to the veneer. | |
7468 | ||
7469 | The workaround does not always preserve the memory access order for | |
7470 | the LDMDB instruction, when the instruction loads the PC. | |
7471 | ||
7472 | The workaround is not able to handle problematic instructions when | |
7473 | they are in the middle of an IT block, since a branch is not allowed | |
7474 | there. In that case, the linker reports a warning and no replacement | |
7475 | occurs. | |
7476 | ||
7477 | The workaround is not able to replace problematic instructions with a | |
7478 | PC-relative branch instruction if the @samp{.text} section is too | |
7479 | large. In that case, when the branch that replaces the original code | |
7480 | cannot be encoded, the linker reports a warning and no replacement | |
7481 | occurs. | |
7482 | ||
bf21ed78 MS |
7483 | @cindex NO_ENUM_SIZE_WARNING |
7484 | @kindex --no-enum-size-warning | |
726150b7 | 7485 | The @option{--no-enum-size-warning} switch prevents the linker from |
bf21ed78 MS |
7486 | warning when linking object files that specify incompatible EABI |
7487 | enumeration size attributes. For example, with this switch enabled, | |
7488 | linking of an object file using 32-bit enumeration values with another | |
7489 | using enumeration values fitted into the smallest possible space will | |
7490 | not be diagnosed. | |
a9dc9481 JM |
7491 | |
7492 | @cindex NO_WCHAR_SIZE_WARNING | |
7493 | @kindex --no-wchar-size-warning | |
7494 | The @option{--no-wchar-size-warning} switch prevents the linker from | |
7495 | warning when linking object files that specify incompatible EABI | |
7496 | @code{wchar_t} size attributes. For example, with this switch enabled, | |
7497 | linking of an object file using 32-bit @code{wchar_t} values with another | |
7498 | using 16-bit @code{wchar_t} values will not be diagnosed. | |
bf21ed78 | 7499 | |
726150b7 NC |
7500 | @cindex PIC_VENEER |
7501 | @kindex --pic-veneer | |
7502 | The @samp{--pic-veneer} switch makes the linker use PIC sequences for | |
7503 | ARM/Thumb interworking veneers, even if the rest of the binary | |
7504 | is not PIC. This avoids problems on uClinux targets where | |
7505 | @samp{--emit-relocs} is used to generate relocatable binaries. | |
7506 | ||
7507 | @cindex STUB_GROUP_SIZE | |
7508 | @kindex --stub-group-size=@var{N} | |
7509 | The linker will automatically generate and insert small sequences of | |
7510 | code into a linked ARM ELF executable whenever an attempt is made to | |
7511 | perform a function call to a symbol that is too far away. The | |
7512 | placement of these sequences of instructions - called stubs - is | |
a05a5b64 | 7513 | controlled by the command-line option @option{--stub-group-size=N}. |
726150b7 | 7514 | The placement is important because a poor choice can create a need for |
11e7fd74 | 7515 | duplicate stubs, increasing the code size. The linker will try to |
726150b7 NC |
7516 | group stubs together in order to reduce interruptions to the flow of |
7517 | code, but it needs guidance as to how big these groups should be and | |
7518 | where they should be placed. | |
7519 | ||
7520 | The value of @samp{N}, the parameter to the | |
7521 | @option{--stub-group-size=} option controls where the stub groups are | |
07d72278 | 7522 | placed. If it is negative then all stubs are placed after the first |
726150b7 NC |
7523 | branch that needs them. If it is positive then the stubs can be |
7524 | placed either before or after the branches that need them. If the | |
7525 | value of @samp{N} is 1 (either +1 or -1) then the linker will choose | |
7526 | exactly where to place groups of stubs, using its built in heuristics. | |
7527 | A value of @samp{N} greater than 1 (or smaller than -1) tells the | |
7528 | linker that a single group of stubs can service at most @samp{N} bytes | |
7529 | from the input sections. | |
7530 | ||
7531 | The default, if @option{--stub-group-size=} is not specified, is | |
7532 | @samp{N = +1}. | |
7533 | ||
1a51c1a4 NC |
7534 | Farcalls stubs insertion is fully supported for the ARM-EABI target |
7535 | only, because it relies on object files properties not present | |
7536 | otherwise. | |
7537 | ||
1db37fe6 YG |
7538 | @cindex Cortex-A8 erratum workaround |
7539 | @kindex --fix-cortex-a8 | |
7540 | @kindex --no-fix-cortex-a8 | |
7541 | The @samp{--fix-cortex-a8} switch enables a link-time workaround for an erratum in certain Cortex-A8 processors. The workaround is enabled by default if you are targeting the ARM v7-A architecture profile. It can be enabled otherwise by specifying @samp{--fix-cortex-a8}, or disabled unconditionally by specifying @samp{--no-fix-cortex-a8}. | |
7542 | ||
7543 | The erratum only affects Thumb-2 code. Please contact ARM for further details. | |
7544 | ||
68fcca92 JW |
7545 | @cindex Cortex-A53 erratum 835769 workaround |
7546 | @kindex --fix-cortex-a53-835769 | |
7547 | @kindex --no-fix-cortex-a53-835769 | |
7548 | The @samp{--fix-cortex-a53-835769} switch enables a link-time workaround for erratum 835769 present on certain early revisions of Cortex-A53 processors. The workaround is disabled by default. It can be enabled by specifying @samp{--fix-cortex-a53-835769}, or disabled unconditionally by specifying @samp{--no-fix-cortex-a53-835769}. | |
7549 | ||
7550 | Please contact ARM for further details. | |
7551 | ||
1db37fe6 YG |
7552 | @kindex --merge-exidx-entries |
7553 | @kindex --no-merge-exidx-entries | |
7554 | @cindex Merging exidx entries | |
7555 | The @samp{--no-merge-exidx-entries} switch disables the merging of adjacent exidx entries in debuginfo. | |
7556 | ||
7557 | @kindex --long-plt | |
7558 | @cindex 32-bit PLT entries | |
7559 | The @samp{--long-plt} option enables the use of 16 byte PLT entries | |
7560 | which support up to 4Gb of code. The default is to use 12 byte PLT | |
7561 | entries which only support 512Mb of code. | |
7562 | ||
1f56df9d JW |
7563 | @kindex --no-apply-dynamic-relocs |
7564 | @cindex AArch64 rela addend | |
7565 | The @samp{--no-apply-dynamic-relocs} option makes AArch64 linker do not apply | |
7566 | link-time values for dynamic relocations. | |
7567 | ||
4ba2ef8f TP |
7568 | @cindex Placement of SG veneers |
7569 | All SG veneers are placed in the special output section @code{.gnu.sgstubs}. | |
a05a5b64 | 7570 | Its start address must be set, either with the command-line option |
4ba2ef8f TP |
7571 | @samp{--section-start} or in a linker script, to indicate where to place these |
7572 | veneers in memory. | |
7573 | ||
54ddd295 TP |
7574 | @kindex --cmse-implib |
7575 | @cindex Secure gateway import library | |
7576 | The @samp{--cmse-implib} option requests that the import libraries | |
7577 | specified by the @samp{--out-implib} and @samp{--in-implib} options are | |
7578 | secure gateway import libraries, suitable for linking a non-secure | |
7579 | executable against secure code as per ARMv8-M Security Extensions. | |
7580 | ||
0955507f TP |
7581 | @kindex --in-implib=@var{file} |
7582 | @cindex Input import library | |
7583 | The @samp{--in-implib=file} specifies an input import library whose symbols | |
7584 | must keep the same address in the executable being produced. A warning is | |
7585 | given if no @samp{--out-implib} is given but new symbols have been introduced | |
7586 | in the executable that should be listed in its import library. Otherwise, if | |
7587 | @samp{--out-implib} is specified, the symbols are added to the output import | |
7588 | library. A warning is also given if some symbols present in the input import | |
7589 | library have disappeared from the executable. This option is only effective | |
7590 | for Secure Gateway import libraries, ie. when @samp{--cmse-implib} is | |
7591 | specified. | |
7592 | ||
36f63dca NC |
7593 | @ifclear GENERIC |
7594 | @lowersections | |
7595 | @end ifclear | |
7596 | @end ifset | |
7597 | ||
7598 | @ifset HPPA | |
7599 | @ifclear GENERIC | |
7600 | @raisesections | |
7601 | @end ifclear | |
7602 | ||
7603 | @node HPPA ELF32 | |
7604 | @section @command{ld} and HPPA 32-bit ELF Support | |
7605 | @cindex HPPA multiple sub-space stubs | |
7606 | @kindex --multi-subspace | |
7607 | When generating a shared library, @command{ld} will by default generate | |
7608 | import stubs suitable for use with a single sub-space application. | |
7609 | The @samp{--multi-subspace} switch causes @command{ld} to generate export | |
7610 | stubs, and different (larger) import stubs suitable for use with | |
7611 | multiple sub-spaces. | |
7612 | ||
7613 | @cindex HPPA stub grouping | |
7614 | @kindex --stub-group-size=@var{N} | |
7615 | Long branch stubs and import/export stubs are placed by @command{ld} in | |
7616 | stub sections located between groups of input sections. | |
7617 | @samp{--stub-group-size} specifies the maximum size of a group of input | |
7618 | sections handled by one stub section. Since branch offsets are signed, | |
7619 | a stub section may serve two groups of input sections, one group before | |
7620 | the stub section, and one group after it. However, when using | |
7621 | conditional branches that require stubs, it may be better (for branch | |
7622 | prediction) that stub sections only serve one group of input sections. | |
7623 | A negative value for @samp{N} chooses this scheme, ensuring that | |
7624 | branches to stubs always use a negative offset. Two special values of | |
7625 | @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct | |
7626 | @command{ld} to automatically size input section groups for the branch types | |
7627 | detected, with the same behaviour regarding stub placement as other | |
7628 | positive or negative values of @samp{N} respectively. | |
7629 | ||
7630 | Note that @samp{--stub-group-size} does not split input sections. A | |
7631 | single input section larger than the group size specified will of course | |
7632 | create a larger group (of one section). If input sections are too | |
7633 | large, it may not be possible for a branch to reach its stub. | |
7634 | ||
7635 | @ifclear GENERIC | |
7636 | @lowersections | |
7637 | @end ifclear | |
7638 | @end ifset | |
7639 | ||
7fb9f789 NC |
7640 | @ifset M68K |
7641 | @ifclear GENERIC | |
7642 | @raisesections | |
7643 | @end ifclear | |
7644 | ||
7645 | @node M68K | |
7646 | @section @command{ld} and the Motorola 68K family | |
7647 | ||
7648 | @cindex Motorola 68K GOT generation | |
7649 | @kindex --got=@var{type} | |
7650 | The @samp{--got=@var{type}} option lets you choose the GOT generation scheme. | |
7651 | The choices are @samp{single}, @samp{negative}, @samp{multigot} and | |
7652 | @samp{target}. When @samp{target} is selected the linker chooses | |
7653 | the default GOT generation scheme for the current target. | |
7654 | @samp{single} tells the linker to generate a single GOT with | |
7655 | entries only at non-negative offsets. | |
7656 | @samp{negative} instructs the linker to generate a single GOT with | |
7657 | entries at both negative and positive offsets. Not all environments | |
7658 | support such GOTs. | |
7659 | @samp{multigot} allows the linker to generate several GOTs in the | |
7660 | output file. All GOT references from a single input object | |
7661 | file access the same GOT, but references from different input object | |
7662 | files might access different GOTs. Not all environments support such GOTs. | |
7663 | ||
7664 | @ifclear GENERIC | |
7665 | @lowersections | |
7666 | @end ifclear | |
7667 | @end ifset | |
7668 | ||
833794fc MR |
7669 | @ifset MIPS |
7670 | @ifclear GENERIC | |
7671 | @raisesections | |
7672 | @end ifclear | |
7673 | ||
7674 | @node MIPS | |
7675 | @section @command{ld} and the MIPS family | |
7676 | ||
7677 | @cindex MIPS microMIPS instruction choice selection | |
7678 | @kindex --insn32 | |
7679 | @kindex --no-insn32 | |
7680 | The @samp{--insn32} and @samp{--no-insn32} options control the choice of | |
7681 | microMIPS instructions used in code generated by the linker, such as that | |
7682 | in the PLT or lazy binding stubs, or in relaxation. If @samp{--insn32} is | |
7683 | used, then the linker only uses 32-bit instruction encodings. By default | |
7684 | or if @samp{--no-insn32} is used, all instruction encodings are used, | |
7685 | including 16-bit ones where possible. | |
7686 | ||
8b10b0b3 MR |
7687 | @cindex MIPS branch relocation check control |
7688 | @kindex --ignore-branch-isa | |
7689 | @kindex --no-ignore-branch-isa | |
7690 | The @samp{--ignore-branch-isa} and @samp{--no-ignore-branch-isa} options | |
7691 | control branch relocation checks for invalid ISA mode transitions. If | |
7692 | @samp{--ignore-branch-isa} is used, then the linker accepts any branch | |
7693 | relocations and any ISA mode transition required is lost in relocation | |
7694 | calculation, except for some cases of @code{BAL} instructions which meet | |
7695 | relaxation conditions and are converted to equivalent @code{JALX} | |
7696 | instructions as the associated relocation is calculated. By default | |
7697 | or if @samp{--no-ignore-branch-isa} is used a check is made causing | |
7698 | the loss of an ISA mode transition to produce an error. | |
7699 | ||
833794fc MR |
7700 | @ifclear GENERIC |
7701 | @lowersections | |
7702 | @end ifclear | |
7703 | @end ifset | |
7704 | ||
36f63dca NC |
7705 | @ifset MMIX |
7706 | @ifclear GENERIC | |
7707 | @raisesections | |
7708 | @end ifclear | |
7709 | ||
7710 | @node MMIX | |
7711 | @section @code{ld} and MMIX | |
7712 | For MMIX, there is a choice of generating @code{ELF} object files or | |
7713 | @code{mmo} object files when linking. The simulator @code{mmix} | |
7714 | understands the @code{mmo} format. The binutils @code{objcopy} utility | |
7715 | can translate between the two formats. | |
7716 | ||
7717 | There is one special section, the @samp{.MMIX.reg_contents} section. | |
7718 | Contents in this section is assumed to correspond to that of global | |
7719 | registers, and symbols referring to it are translated to special symbols, | |
7720 | equal to registers. In a final link, the start address of the | |
7721 | @samp{.MMIX.reg_contents} section corresponds to the first allocated | |
7722 | global register multiplied by 8. Register @code{$255} is not included in | |
7723 | this section; it is always set to the program entry, which is at the | |
7724 | symbol @code{Main} for @code{mmo} files. | |
7725 | ||
7a2de473 HPN |
7726 | Global symbols with the prefix @code{__.MMIX.start.}, for example |
7727 | @code{__.MMIX.start..text} and @code{__.MMIX.start..data} are special. | |
7728 | The default linker script uses these to set the default start address | |
7729 | of a section. | |
36f63dca NC |
7730 | |
7731 | Initial and trailing multiples of zero-valued 32-bit words in a section, | |
7732 | are left out from an mmo file. | |
7733 | ||
7734 | @ifclear GENERIC | |
7735 | @lowersections | |
7736 | @end ifclear | |
7737 | @end ifset | |
7738 | ||
7739 | @ifset MSP430 | |
7740 | @ifclear GENERIC | |
7741 | @raisesections | |
7742 | @end ifclear | |
7743 | ||
7744 | @node MSP430 | |
7745 | @section @code{ld} and MSP430 | |
7746 | For the MSP430 it is possible to select the MPU architecture. The flag @samp{-m [mpu type]} | |
7747 | will select an appropriate linker script for selected MPU type. (To get a list of known MPUs | |
7748 | just pass @samp{-m help} option to the linker). | |
7749 | ||
7750 | @cindex MSP430 extra sections | |
7751 | The linker will recognize some extra sections which are MSP430 specific: | |
7752 | ||
7753 | @table @code | |
7754 | @item @samp{.vectors} | |
7755 | Defines a portion of ROM where interrupt vectors located. | |
7756 | ||
7757 | @item @samp{.bootloader} | |
7758 | Defines the bootloader portion of the ROM (if applicable). Any code | |
7759 | in this section will be uploaded to the MPU. | |
7760 | ||
7761 | @item @samp{.infomem} | |
7762 | Defines an information memory section (if applicable). Any code in | |
7763 | this section will be uploaded to the MPU. | |
7764 | ||
c0065db7 | 7765 | @item @samp{.infomemnobits} |
36f63dca NC |
7766 | This is the same as the @samp{.infomem} section except that any code |
7767 | in this section will not be uploaded to the MPU. | |
7768 | ||
7769 | @item @samp{.noinit} | |
7770 | Denotes a portion of RAM located above @samp{.bss} section. | |
7771 | ||
c0065db7 | 7772 | The last two sections are used by gcc. |
36f63dca NC |
7773 | @end table |
7774 | ||
7ef3addb JL |
7775 | @table @option |
7776 | @cindex MSP430 Options | |
7777 | @kindex --code-region | |
7778 | @item --code-region=[either,lower,upper,none] | |
7779 | This will transform .text* sections to [either,lower,upper].text* sections. The | |
7780 | argument passed to GCC for -mcode-region is propagated to the linker | |
7781 | using this option. | |
7782 | ||
7783 | @kindex --data-region | |
7784 | @item --data-region=[either,lower,upper,none] | |
7785 | This will transform .data*, .bss* and .rodata* sections to | |
7786 | [either,lower,upper].[data,bss,rodata]* sections. The argument passed to GCC | |
7787 | for -mdata-region is propagated to the linker using this option. | |
7788 | ||
7789 | @kindex --disable-sec-transformation | |
7790 | @item --disable-sec-transformation | |
7791 | Prevent the transformation of sections as specified by the @code{--code-region} | |
7792 | and @code{--data-region} options. | |
7793 | This is useful if you are compiling and linking using a single call to the GCC | |
7794 | wrapper, and want to compile the source files using -m[code,data]-region but | |
7795 | not transform the sections for prebuilt libraries and objects. | |
7796 | @end table | |
7797 | ||
36f63dca NC |
7798 | @ifclear GENERIC |
7799 | @lowersections | |
7800 | @end ifclear | |
7801 | @end ifset | |
7802 | ||
35c08157 KLC |
7803 | @ifset NDS32 |
7804 | @ifclear GENERIC | |
7805 | @raisesections | |
7806 | @end ifclear | |
7807 | ||
7808 | @node NDS32 | |
7809 | @section @code{ld} and NDS32 | |
7810 | @kindex relaxing on NDS32 | |
7811 | For NDS32, there are some options to select relaxation behavior. The linker | |
7812 | relaxes objects according to these options. | |
7813 | ||
7814 | @table @code | |
7815 | @item @samp{--m[no-]fp-as-gp} | |
7816 | Disable/enable fp-as-gp relaxation. | |
7817 | ||
7818 | @item @samp{--mexport-symbols=FILE} | |
7819 | Exporting symbols and their address into FILE as linker script. | |
7820 | ||
7821 | @item @samp{--m[no-]ex9} | |
7822 | Disable/enable link-time EX9 relaxation. | |
7823 | ||
7824 | @item @samp{--mexport-ex9=FILE} | |
7825 | Export the EX9 table after linking. | |
7826 | ||
7827 | @item @samp{--mimport-ex9=FILE} | |
7828 | Import the Ex9 table for EX9 relaxation. | |
7829 | ||
7830 | @item @samp{--mupdate-ex9} | |
7831 | Update the existing EX9 table. | |
7832 | ||
7833 | @item @samp{--mex9-limit=NUM} | |
7834 | Maximum number of entries in the ex9 table. | |
7835 | ||
7836 | @item @samp{--mex9-loop-aware} | |
7837 | Avoid generating the EX9 instruction inside the loop. | |
7838 | ||
7839 | @item @samp{--m[no-]ifc} | |
7840 | Disable/enable the link-time IFC optimization. | |
7841 | ||
7842 | @item @samp{--mifc-loop-aware} | |
7843 | Avoid generating the IFC instruction inside the loop. | |
7844 | @end table | |
7845 | ||
7846 | @ifclear GENERIC | |
7847 | @lowersections | |
7848 | @end ifclear | |
7849 | @end ifset | |
7850 | ||
78058a5e SL |
7851 | @ifset NIOSII |
7852 | @ifclear GENERIC | |
7853 | @raisesections | |
7854 | @end ifclear | |
7855 | ||
7856 | @node Nios II | |
7857 | @section @command{ld} and the Altera Nios II | |
7858 | @cindex Nios II call relaxation | |
7859 | @kindex --relax on Nios II | |
7860 | ||
7861 | Call and immediate jump instructions on Nios II processors are limited to | |
7862 | transferring control to addresses in the same 256MB memory segment, | |
7863 | which may result in @command{ld} giving | |
7864 | @samp{relocation truncated to fit} errors with very large programs. | |
7865 | The command-line option @option{--relax} enables the generation of | |
7866 | trampolines that can access the entire 32-bit address space for calls | |
7867 | outside the normal @code{call} and @code{jmpi} address range. These | |
7868 | trampolines are inserted at section boundaries, so may not themselves | |
7869 | be reachable if an input section and its associated call trampolines are | |
7870 | larger than 256MB. | |
7871 | ||
7872 | The @option{--relax} option is enabled by default unless @option{-r} | |
7873 | is also specified. You can disable trampoline generation by using the | |
7874 | @option{--no-relax} linker option. You can also disable this optimization | |
7875 | locally by using the @samp{set .noat} directive in assembly-language | |
7876 | source files, as the linker-inserted trampolines use the @code{at} | |
7877 | register as a temporary. | |
7878 | ||
7879 | Note that the linker @option{--relax} option is independent of assembler | |
7880 | relaxation options, and that using the GNU assembler's @option{-relax-all} | |
7881 | option interferes with the linker's more selective call instruction relaxation. | |
7882 | ||
7883 | @ifclear GENERIC | |
7884 | @lowersections | |
7885 | @end ifclear | |
7886 | @end ifset | |
7887 | ||
2a60a7a8 AM |
7888 | @ifset POWERPC |
7889 | @ifclear GENERIC | |
7890 | @raisesections | |
7891 | @end ifclear | |
7892 | ||
7893 | @node PowerPC ELF32 | |
7894 | @section @command{ld} and PowerPC 32-bit ELF Support | |
7895 | @cindex PowerPC long branches | |
7896 | @kindex --relax on PowerPC | |
7897 | Branches on PowerPC processors are limited to a signed 26-bit | |
7898 | displacement, which may result in @command{ld} giving | |
7899 | @samp{relocation truncated to fit} errors with very large programs. | |
7900 | @samp{--relax} enables the generation of trampolines that can access | |
7901 | the entire 32-bit address space. These trampolines are inserted at | |
7902 | section boundaries, so may not themselves be reachable if an input | |
c8a1f254 NS |
7903 | section exceeds 33M in size. You may combine @samp{-r} and |
7904 | @samp{--relax} to add trampolines in a partial link. In that case | |
7905 | both branches to undefined symbols and inter-section branches are also | |
7906 | considered potentially out of range, and trampolines inserted. | |
2a60a7a8 AM |
7907 | |
7908 | @cindex PowerPC ELF32 options | |
7909 | @table @option | |
7910 | @cindex PowerPC PLT | |
7911 | @kindex --bss-plt | |
7912 | @item --bss-plt | |
7913 | Current PowerPC GCC accepts a @samp{-msecure-plt} option that | |
7914 | generates code capable of using a newer PLT and GOT layout that has | |
7915 | the security advantage of no executable section ever needing to be | |
7916 | writable and no writable section ever being executable. PowerPC | |
7917 | @command{ld} will generate this layout, including stubs to access the | |
7918 | PLT, if all input files (including startup and static libraries) were | |
7919 | compiled with @samp{-msecure-plt}. @samp{--bss-plt} forces the old | |
7920 | BSS PLT (and GOT layout) which can give slightly better performance. | |
7921 | ||
016687f8 AM |
7922 | @kindex --secure-plt |
7923 | @item --secure-plt | |
7924 | @command{ld} will use the new PLT and GOT layout if it is linking new | |
7925 | @samp{-fpic} or @samp{-fPIC} code, but does not do so automatically | |
7926 | when linking non-PIC code. This option requests the new PLT and GOT | |
7927 | layout. A warning will be given if some object file requires the old | |
7928 | style BSS PLT. | |
7929 | ||
2a60a7a8 AM |
7930 | @cindex PowerPC GOT |
7931 | @kindex --sdata-got | |
7932 | @item --sdata-got | |
7933 | The new secure PLT and GOT are placed differently relative to other | |
7934 | sections compared to older BSS PLT and GOT placement. The location of | |
7935 | @code{.plt} must change because the new secure PLT is an initialized | |
7936 | section while the old PLT is uninitialized. The reason for the | |
7937 | @code{.got} change is more subtle: The new placement allows | |
7938 | @code{.got} to be read-only in applications linked with | |
7939 | @samp{-z relro -z now}. However, this placement means that | |
7940 | @code{.sdata} cannot always be used in shared libraries, because the | |
7941 | PowerPC ABI accesses @code{.sdata} in shared libraries from the GOT | |
7942 | pointer. @samp{--sdata-got} forces the old GOT placement. PowerPC | |
7943 | GCC doesn't use @code{.sdata} in shared libraries, so this option is | |
7944 | really only useful for other compilers that may do so. | |
7945 | ||
7946 | @cindex PowerPC stub symbols | |
7947 | @kindex --emit-stub-syms | |
7948 | @item --emit-stub-syms | |
7949 | This option causes @command{ld} to label linker stubs with a local | |
7950 | symbol that encodes the stub type and destination. | |
7951 | ||
7952 | @cindex PowerPC TLS optimization | |
7953 | @kindex --no-tls-optimize | |
7954 | @item --no-tls-optimize | |
7955 | PowerPC @command{ld} normally performs some optimization of code | |
7956 | sequences used to access Thread-Local Storage. Use this option to | |
7957 | disable the optimization. | |
7958 | @end table | |
7959 | ||
7960 | @ifclear GENERIC | |
7961 | @lowersections | |
7962 | @end ifclear | |
7963 | @end ifset | |
7964 | ||
7965 | @ifset POWERPC64 | |
7966 | @ifclear GENERIC | |
7967 | @raisesections | |
7968 | @end ifclear | |
7969 | ||
7970 | @node PowerPC64 ELF64 | |
7971 | @section @command{ld} and PowerPC64 64-bit ELF Support | |
7972 | ||
7973 | @cindex PowerPC64 ELF64 options | |
7974 | @table @option | |
7975 | @cindex PowerPC64 stub grouping | |
7976 | @kindex --stub-group-size | |
7977 | @item --stub-group-size | |
7978 | Long branch stubs, PLT call stubs and TOC adjusting stubs are placed | |
7979 | by @command{ld} in stub sections located between groups of input sections. | |
7980 | @samp{--stub-group-size} specifies the maximum size of a group of input | |
7981 | sections handled by one stub section. Since branch offsets are signed, | |
7982 | a stub section may serve two groups of input sections, one group before | |
7983 | the stub section, and one group after it. However, when using | |
7984 | conditional branches that require stubs, it may be better (for branch | |
7985 | prediction) that stub sections only serve one group of input sections. | |
7986 | A negative value for @samp{N} chooses this scheme, ensuring that | |
7987 | branches to stubs always use a negative offset. Two special values of | |
7988 | @samp{N} are recognized, @samp{1} and @samp{-1}. These both instruct | |
7989 | @command{ld} to automatically size input section groups for the branch types | |
7990 | detected, with the same behaviour regarding stub placement as other | |
7991 | positive or negative values of @samp{N} respectively. | |
7992 | ||
7993 | Note that @samp{--stub-group-size} does not split input sections. A | |
7994 | single input section larger than the group size specified will of course | |
7995 | create a larger group (of one section). If input sections are too | |
7996 | large, it may not be possible for a branch to reach its stub. | |
7997 | ||
7998 | @cindex PowerPC64 stub symbols | |
7999 | @kindex --emit-stub-syms | |
8000 | @item --emit-stub-syms | |
8001 | This option causes @command{ld} to label linker stubs with a local | |
8002 | symbol that encodes the stub type and destination. | |
8003 | ||
8004 | @cindex PowerPC64 dot symbols | |
8005 | @kindex --dotsyms | |
8006 | @kindex --no-dotsyms | |
95421fb9 AM |
8007 | @item --dotsyms |
8008 | @itemx --no-dotsyms | |
2a60a7a8 AM |
8009 | These two options control how @command{ld} interprets version patterns |
8010 | in a version script. Older PowerPC64 compilers emitted both a | |
8011 | function descriptor symbol with the same name as the function, and a | |
8012 | code entry symbol with the name prefixed by a dot (@samp{.}). To | |
8013 | properly version a function @samp{foo}, the version script thus needs | |
8014 | to control both @samp{foo} and @samp{.foo}. The option | |
8015 | @samp{--dotsyms}, on by default, automatically adds the required | |
8016 | dot-prefixed patterns. Use @samp{--no-dotsyms} to disable this | |
8017 | feature. | |
8018 | ||
7ae4ea7d AM |
8019 | @cindex PowerPC64 register save/restore functions |
8020 | @kindex --save-restore-funcs | |
8021 | @kindex --no-save-restore-funcs | |
95421fb9 AM |
8022 | @item --save-restore-funcs |
8023 | @itemx --no-save-restore-funcs | |
7ae4ea7d AM |
8024 | These two options control whether PowerPC64 @command{ld} automatically |
8025 | provides out-of-line register save and restore functions used by | |
8026 | @samp{-Os} code. The default is to provide any such referenced | |
8027 | function for a normal final link, and to not do so for a relocatable | |
8028 | link. | |
8029 | ||
2a60a7a8 AM |
8030 | @cindex PowerPC64 TLS optimization |
8031 | @kindex --no-tls-optimize | |
8032 | @item --no-tls-optimize | |
8033 | PowerPC64 @command{ld} normally performs some optimization of code | |
8034 | sequences used to access Thread-Local Storage. Use this option to | |
8035 | disable the optimization. | |
8036 | ||
7c9cf415 AM |
8037 | @cindex PowerPC64 __tls_get_addr optimization |
8038 | @kindex --tls-get-addr-optimize | |
8039 | @kindex --no-tls-get-addr-optimize | |
9e7028aa AM |
8040 | @kindex --tls-get-addr-regsave |
8041 | @kindex --no-tls-get-addr-regsave | |
95421fb9 AM |
8042 | @item --tls-get-addr-optimize |
8043 | @itemx --no-tls-get-addr-optimize | |
9e7028aa | 8044 | These options control how PowerPC64 @command{ld} uses a special |
7c9cf415 AM |
8045 | stub to call __tls_get_addr. PowerPC64 glibc 2.22 and later support |
8046 | an optimization that allows the second and subsequent calls to | |
8047 | @code{__tls_get_addr} for a given symbol to be resolved by the special | |
9e7028aa AM |
8048 | stub without calling in to glibc. By default the linker enables |
8049 | generation of the stub when glibc advertises the availability of | |
8050 | __tls_get_addr_opt. | |
8051 | Using @option{--tls-get-addr-optimize} with an older glibc won't do | |
8052 | much besides slow down your applications, but may be useful if linking | |
8053 | an application against an older glibc with the expectation that it | |
8054 | will normally be used on systems having a newer glibc. | |
8055 | @option{--tls-get-addr-regsave} forces generation of a stub that saves | |
8056 | and restores volatile registers around the call into glibc. Normally, | |
8057 | this is done when the linker detects a call to __tls_get_addr_desc. | |
8058 | Such calls then go via the register saving stub to __tls_get_addr_opt. | |
5cebc931 | 8059 | @option{--no-tls-get-addr-regsave} disables generation of the |
9e7028aa | 8060 | register saves. |
7c9cf415 | 8061 | |
2a60a7a8 AM |
8062 | @cindex PowerPC64 OPD optimization |
8063 | @kindex --no-opd-optimize | |
8064 | @item --no-opd-optimize | |
8065 | PowerPC64 @command{ld} normally removes @code{.opd} section entries | |
8066 | corresponding to deleted link-once functions, or functions removed by | |
e7fc76dd | 8067 | the action of @samp{--gc-sections} or linker script @code{/DISCARD/}. |
2a60a7a8 AM |
8068 | Use this option to disable @code{.opd} optimization. |
8069 | ||
8070 | @cindex PowerPC64 OPD spacing | |
8071 | @kindex --non-overlapping-opd | |
8072 | @item --non-overlapping-opd | |
8073 | Some PowerPC64 compilers have an option to generate compressed | |
8074 | @code{.opd} entries spaced 16 bytes apart, overlapping the third word, | |
8075 | the static chain pointer (unused in C) with the first word of the next | |
8076 | entry. This option expands such entries to the full 24 bytes. | |
8077 | ||
8078 | @cindex PowerPC64 TOC optimization | |
8079 | @kindex --no-toc-optimize | |
8080 | @item --no-toc-optimize | |
8081 | PowerPC64 @command{ld} normally removes unused @code{.toc} section | |
8082 | entries. Such entries are detected by examining relocations that | |
8083 | reference the TOC in code sections. A reloc in a deleted code section | |
8084 | marks a TOC word as unneeded, while a reloc in a kept code section | |
8085 | marks a TOC word as needed. Since the TOC may reference itself, TOC | |
8086 | relocs are also examined. TOC words marked as both needed and | |
8087 | unneeded will of course be kept. TOC words without any referencing | |
8088 | reloc are assumed to be part of a multi-word entry, and are kept or | |
8089 | discarded as per the nearest marked preceding word. This works | |
8090 | reliably for compiler generated code, but may be incorrect if assembly | |
8091 | code is used to insert TOC entries. Use this option to disable the | |
8092 | optimization. | |
8093 | ||
d882c988 AM |
8094 | @cindex PowerPC64 inline PLT call optimization |
8095 | @kindex --no-inline-optimize | |
8096 | @item --no-inline-optimize | |
8097 | PowerPC64 @command{ld} normally replaces inline PLT call sequences | |
8098 | marked with @code{R_PPC64_PLTSEQ}, @code{R_PPC64_PLTCALL}, | |
8099 | @code{R_PPC64_PLT16_HA} and @code{R_PPC64_PLT16_LO_DS} relocations by | |
8100 | a number of @code{nop}s and a direct call when the function is defined | |
8101 | locally and can't be overridden by some other definition. This option | |
8102 | disables that optimization. | |
8103 | ||
2a60a7a8 AM |
8104 | @cindex PowerPC64 multi-TOC |
8105 | @kindex --no-multi-toc | |
8106 | @item --no-multi-toc | |
794e51c0 AM |
8107 | If given any toc option besides @code{-mcmodel=medium} or |
8108 | @code{-mcmodel=large}, PowerPC64 GCC generates code for a TOC model | |
8109 | where TOC | |
2a60a7a8 AM |
8110 | entries are accessed with a 16-bit offset from r2. This limits the |
8111 | total TOC size to 64K. PowerPC64 @command{ld} extends this limit by | |
8112 | grouping code sections such that each group uses less than 64K for its | |
8113 | TOC entries, then inserts r2 adjusting stubs between inter-group | |
8114 | calls. @command{ld} does not split apart input sections, so cannot | |
8115 | help if a single input file has a @code{.toc} section that exceeds | |
8116 | 64K, most likely from linking multiple files with @command{ld -r}. | |
8117 | Use this option to turn off this feature. | |
794e51c0 AM |
8118 | |
8119 | @cindex PowerPC64 TOC sorting | |
8120 | @kindex --no-toc-sort | |
8121 | @item --no-toc-sort | |
8122 | By default, @command{ld} sorts TOC sections so that those whose file | |
8123 | happens to have a section called @code{.init} or @code{.fini} are | |
8124 | placed first, followed by TOC sections referenced by code generated | |
8125 | with PowerPC64 gcc's @code{-mcmodel=small}, and lastly TOC sections | |
8126 | referenced only by code generated with PowerPC64 gcc's | |
8127 | @code{-mcmodel=medium} or @code{-mcmodel=large} options. Doing this | |
8128 | results in better TOC grouping for multi-TOC. Use this option to turn | |
8129 | off this feature. | |
8130 | ||
8131 | @cindex PowerPC64 PLT stub alignment | |
8132 | @kindex --plt-align | |
8133 | @kindex --no-plt-align | |
8134 | @item --plt-align | |
8135 | @itemx --no-plt-align | |
8136 | Use these options to control whether individual PLT call stubs are | |
2420fff6 AM |
8137 | aligned to a 32-byte boundary, or to the specified power of two |
8138 | boundary when using @code{--plt-align=}. A negative value may be | |
8139 | specified to pad PLT call stubs so that they do not cross the | |
8140 | specified power of two boundary (or the minimum number of boundaries | |
8141 | if a PLT stub is so large that it must cross a boundary). By default | |
8142 | PLT call stubs are aligned to 32-byte boundaries. | |
794e51c0 AM |
8143 | |
8144 | @cindex PowerPC64 PLT call stub static chain | |
8145 | @kindex --plt-static-chain | |
8146 | @kindex --no-plt-static-chain | |
8147 | @item --plt-static-chain | |
8148 | @itemx --no-plt-static-chain | |
8149 | Use these options to control whether PLT call stubs load the static | |
8150 | chain pointer (r11). @code{ld} defaults to not loading the static | |
8151 | chain since there is never any need to do so on a PLT call. | |
8152 | ||
8153 | @cindex PowerPC64 PLT call stub thread safety | |
8154 | @kindex --plt-thread-safe | |
8155 | @kindex --no-plt-thread-safe | |
8156 | @item --plt-thread-safe | |
1be5d8d3 | 8157 | @itemx --no-plt-thread-safe |
794e51c0 AM |
8158 | With power7's weakly ordered memory model, it is possible when using |
8159 | lazy binding for ld.so to update a plt entry in one thread and have | |
8160 | another thread see the individual plt entry words update in the wrong | |
8161 | order, despite ld.so carefully writing in the correct order and using | |
8162 | memory write barriers. To avoid this we need some sort of read | |
8163 | barrier in the call stub, or use LD_BIND_NOW=1. By default, @code{ld} | |
8164 | looks for calls to commonly used functions that create threads, and if | |
8165 | seen, adds the necessary barriers. Use these options to change the | |
8166 | default behaviour. | |
8b5f1ed8 AM |
8167 | |
8168 | @cindex PowerPC64 ELFv2 PLT localentry optimization | |
8169 | @kindex --plt-localentry | |
8170 | @kindex --no-plt-localentry | |
8171 | @item --plt-localentry | |
8172 | @itemx --no-localentry | |
8173 | ELFv2 functions with localentry:0 are those with a single entry point, | |
8174 | ie. global entry == local entry, and that have no requirement on r2 | |
8175 | (the TOC/GOT pointer) or r12, and guarantee r2 is unchanged on return. | |
8176 | Such an external function can be called via the PLT without saving r2 | |
8177 | or restoring it on return, avoiding a common load-hit-store for small | |
8178 | functions. The optimization is attractive, with up to 40% reduction | |
8179 | in execution time for a small function, but can result in symbol | |
d44c746a AM |
8180 | interposition failures. Also, minor changes in a shared library, |
8181 | including system libraries, can cause a function that was localentry:0 | |
8182 | to become localentry:8. This will result in a dynamic loader | |
8183 | complaint and failure to run. The option is experimental, use with | |
8184 | care. @option{--no-plt-localentry} is the default. | |
d882c988 AM |
8185 | |
8186 | @cindex PowerPC64 Power10 stubs | |
8187 | @kindex --power10-stubs | |
8188 | @kindex --no-power10-stubs | |
8189 | @item --power10-stubs | |
8190 | @itemx --no-power10-stubs | |
8191 | When PowerPC64 @command{ld} links input object files containing | |
8192 | relocations used on power10 prefixed instructions it normally creates | |
e10a07b3 AM |
8193 | linkage stubs (PLT call and long branch) using power10 instructions |
8194 | for @code{@@notoc} PLT calls where @code{r2} is not known. The | |
8195 | power10 notoc stubs are smaller and faster, so are preferred for | |
d882c988 AM |
8196 | power10. @option{--power10-stubs} and @option{--no-power10-stubs} |
8197 | allow you to override the linker's selection of stub instructions. | |
e10a07b3 AM |
8198 | @option{--power10-stubs=auto} allows the user to select the default |
8199 | auto mode. | |
2a60a7a8 AM |
8200 | @end table |
8201 | ||
8202 | @ifclear GENERIC | |
8203 | @lowersections | |
8204 | @end ifclear | |
8205 | @end ifset | |
8206 | ||
b4cbbe8f AK |
8207 | @ifset S/390 |
8208 | @ifclear GENERIC | |
8209 | @raisesections | |
8210 | @end ifclear | |
8211 | ||
8212 | @node S/390 ELF | |
8213 | @section @command{ld} and S/390 ELF Support | |
8214 | ||
8215 | @cindex S/390 ELF options | |
8216 | @table @option | |
8217 | ||
8218 | @cindex S/390 | |
8219 | @kindex --s390-pgste | |
8220 | @item --s390-pgste | |
8221 | This option marks the result file with a @code{PT_S390_PGSTE} | |
8222 | segment. The Linux kernel is supposed to allocate 4k page tables for | |
8223 | binaries marked that way. | |
8224 | @end table | |
8225 | ||
8226 | @ifclear GENERIC | |
8227 | @lowersections | |
8228 | @end ifclear | |
8229 | @end ifset | |
8230 | ||
49fa1e15 AM |
8231 | @ifset SPU |
8232 | @ifclear GENERIC | |
8233 | @raisesections | |
8234 | @end ifclear | |
8235 | ||
8236 | @node SPU ELF | |
8237 | @section @command{ld} and SPU ELF Support | |
8238 | ||
8239 | @cindex SPU ELF options | |
8240 | @table @option | |
8241 | ||
8242 | @cindex SPU plugins | |
8243 | @kindex --plugin | |
8244 | @item --plugin | |
8245 | This option marks an executable as a PIC plugin module. | |
8246 | ||
8247 | @cindex SPU overlays | |
8248 | @kindex --no-overlays | |
8249 | @item --no-overlays | |
8250 | Normally, @command{ld} recognizes calls to functions within overlay | |
8251 | regions, and redirects such calls to an overlay manager via a stub. | |
8252 | @command{ld} also provides a built-in overlay manager. This option | |
8253 | turns off all this special overlay handling. | |
8254 | ||
8255 | @cindex SPU overlay stub symbols | |
8256 | @kindex --emit-stub-syms | |
8257 | @item --emit-stub-syms | |
8258 | This option causes @command{ld} to label overlay stubs with a local | |
8259 | symbol that encodes the stub type and destination. | |
8260 | ||
8261 | @cindex SPU extra overlay stubs | |
8262 | @kindex --extra-overlay-stubs | |
8263 | @item --extra-overlay-stubs | |
8264 | This option causes @command{ld} to add overlay call stubs on all | |
8265 | function calls out of overlay regions. Normally stubs are not added | |
8266 | on calls to non-overlay regions. | |
8267 | ||
8268 | @cindex SPU local store size | |
8269 | @kindex --local-store=lo:hi | |
8270 | @item --local-store=lo:hi | |
8271 | @command{ld} usually checks that a final executable for SPU fits in | |
8272 | the address range 0 to 256k. This option may be used to change the | |
8273 | range. Disable the check entirely with @option{--local-store=0:0}. | |
8274 | ||
c0065db7 | 8275 | @cindex SPU |
49fa1e15 AM |
8276 | @kindex --stack-analysis |
8277 | @item --stack-analysis | |
8278 | SPU local store space is limited. Over-allocation of stack space | |
8279 | unnecessarily limits space available for code and data, while | |
8280 | under-allocation results in runtime failures. If given this option, | |
8281 | @command{ld} will provide an estimate of maximum stack usage. | |
8282 | @command{ld} does this by examining symbols in code sections to | |
8283 | determine the extents of functions, and looking at function prologues | |
8284 | for stack adjusting instructions. A call-graph is created by looking | |
8285 | for relocations on branch instructions. The graph is then searched | |
8286 | for the maximum stack usage path. Note that this analysis does not | |
8287 | find calls made via function pointers, and does not handle recursion | |
8288 | and other cycles in the call graph. Stack usage may be | |
8289 | under-estimated if your code makes such calls. Also, stack usage for | |
8290 | dynamic allocation, e.g. alloca, will not be detected. If a link map | |
8291 | is requested, detailed information about each function's stack usage | |
8292 | and calls will be given. | |
8293 | ||
c0065db7 | 8294 | @cindex SPU |
49fa1e15 AM |
8295 | @kindex --emit-stack-syms |
8296 | @item --emit-stack-syms | |
8297 | This option, if given along with @option{--stack-analysis} will result | |
8298 | in @command{ld} emitting stack sizing symbols for each function. | |
8299 | These take the form @code{__stack_<function_name>} for global | |
8300 | functions, and @code{__stack_<number>_<function_name>} for static | |
8301 | functions. @code{<number>} is the section id in hex. The value of | |
8302 | such symbols is the stack requirement for the corresponding function. | |
8303 | The symbol size will be zero, type @code{STT_NOTYPE}, binding | |
c0065db7 | 8304 | @code{STB_LOCAL}, and section @code{SHN_ABS}. |
49fa1e15 AM |
8305 | @end table |
8306 | ||
8307 | @ifclear GENERIC | |
8308 | @lowersections | |
8309 | @end ifclear | |
8310 | @end ifset | |
8311 | ||
36f63dca NC |
8312 | @ifset TICOFF |
8313 | @ifclear GENERIC | |
8314 | @raisesections | |
8315 | @end ifclear | |
8316 | ||
8317 | @node TI COFF | |
8318 | @section @command{ld}'s Support for Various TI COFF Versions | |
8319 | @cindex TI COFF versions | |
8320 | @kindex --format=@var{version} | |
8321 | The @samp{--format} switch allows selection of one of the various | |
8322 | TI COFF versions. The latest of this writing is 2; versions 0 and 1 are | |
8323 | also supported. The TI COFF versions also vary in header byte-order | |
8324 | format; @command{ld} will read any version or byte order, but the output | |
8325 | header format depends on the default specified by the specific target. | |
8326 | ||
8327 | @ifclear GENERIC | |
8328 | @lowersections | |
8329 | @end ifclear | |
8330 | @end ifset | |
8331 | ||
2ca22b03 NC |
8332 | @ifset WIN32 |
8333 | @ifclear GENERIC | |
8334 | @raisesections | |
8335 | @end ifclear | |
8336 | ||
8337 | @node WIN32 | |
8338 | @section @command{ld} and WIN32 (cygwin/mingw) | |
8339 | ||
c0065db7 | 8340 | This section describes some of the win32 specific @command{ld} issues. |
a05a5b64 TP |
8341 | See @ref{Options,,Command-line Options} for detailed description of the |
8342 | command-line options mentioned here. | |
2ca22b03 NC |
8343 | |
8344 | @table @emph | |
c0065db7 RM |
8345 | @cindex import libraries |
8346 | @item import libraries | |
69da35b5 | 8347 | The standard Windows linker creates and uses so-called import |
2ca22b03 | 8348 | libraries, which contains information for linking to dll's. They are |
69da35b5 NC |
8349 | regular static archives and are handled as any other static |
8350 | archive. The cygwin and mingw ports of @command{ld} have specific | |
2ca22b03 | 8351 | support for creating such libraries provided with the |
a05a5b64 | 8352 | @samp{--out-implib} command-line option. |
2ca22b03 | 8353 | |
c0065db7 RM |
8354 | @item exporting DLL symbols |
8355 | @cindex exporting DLL symbols | |
dc8465bf NC |
8356 | The cygwin/mingw @command{ld} has several ways to export symbols for dll's. |
8357 | ||
8358 | @table @emph | |
8359 | @item using auto-export functionality | |
8360 | @cindex using auto-export functionality | |
8361 | By default @command{ld} exports symbols with the auto-export functionality, | |
a05a5b64 | 8362 | which is controlled by the following command-line options: |
dc8465bf | 8363 | |
0a5d968e NC |
8364 | @itemize |
8365 | @item --export-all-symbols [This is the default] | |
8366 | @item --exclude-symbols | |
8367 | @item --exclude-libs | |
e1c37eb5 | 8368 | @item --exclude-modules-for-implib |
09e2aba4 | 8369 | @item --version-script |
0a5d968e NC |
8370 | @end itemize |
8371 | ||
09e2aba4 DK |
8372 | When auto-export is in operation, @command{ld} will export all the non-local |
8373 | (global and common) symbols it finds in a DLL, with the exception of a few | |
8374 | symbols known to belong to the system's runtime and libraries. As it will | |
8375 | often not be desirable to export all of a DLL's symbols, which may include | |
8376 | private functions that are not part of any public interface, the command-line | |
9d5777a3 | 8377 | options listed above may be used to filter symbols out from the list for |
09e2aba4 DK |
8378 | exporting. The @samp{--output-def} option can be used in order to see the |
8379 | final list of exported symbols with all exclusions taken into effect. | |
8380 | ||
8381 | If @samp{--export-all-symbols} is not given explicitly on the | |
0a5d968e NC |
8382 | command line, then the default auto-export behavior will be @emph{disabled} |
8383 | if either of the following are true: | |
8384 | ||
8385 | @itemize | |
8386 | @item A DEF file is used. | |
8387 | @item Any symbol in any object file was marked with the __declspec(dllexport) attribute. | |
8388 | @end itemize | |
dc8465bf | 8389 | |
c0065db7 RM |
8390 | @item using a DEF file |
8391 | @cindex using a DEF file | |
dc8465bf NC |
8392 | Another way of exporting symbols is using a DEF file. A DEF file is |
8393 | an ASCII file containing definitions of symbols which should be | |
8394 | exported when a dll is created. Usually it is named @samp{<dll | |
8395 | name>.def} and is added as any other object file to the linker's | |
0a5d968e | 8396 | command line. The file's name must end in @samp{.def} or @samp{.DEF}. |
dc8465bf NC |
8397 | |
8398 | @example | |
8399 | gcc -o <output> <objectfiles> <dll name>.def | |
8400 | @end example | |
8401 | ||
0a5d968e NC |
8402 | Using a DEF file turns off the normal auto-export behavior, unless the |
8403 | @samp{--export-all-symbols} option is also used. | |
8404 | ||
dc8465bf NC |
8405 | Here is an example of a DEF file for a shared library called @samp{xyz.dll}: |
8406 | ||
8407 | @example | |
4b5bd4e7 | 8408 | LIBRARY "xyz.dll" BASE=0x20000000 |
dc8465bf NC |
8409 | |
8410 | EXPORTS | |
8411 | foo | |
8412 | bar | |
8413 | _bar = bar | |
4b5bd4e7 DS |
8414 | another_foo = abc.dll.afoo |
8415 | var1 DATA | |
7fcab871 KT |
8416 | doo = foo == foo2 |
8417 | eoo DATA == var1 | |
c0065db7 | 8418 | @end example |
dc8465bf | 8419 | |
7fcab871 | 8420 | This example defines a DLL with a non-default base address and seven |
4b5bd4e7 DS |
8421 | symbols in the export table. The third exported symbol @code{_bar} is an |
8422 | alias for the second. The fourth symbol, @code{another_foo} is resolved | |
8423 | by "forwarding" to another module and treating it as an alias for | |
8424 | @code{afoo} exported from the DLL @samp{abc.dll}. The final symbol | |
7fcab871 KT |
8425 | @code{var1} is declared to be a data object. The @samp{doo} symbol in |
8426 | export library is an alias of @samp{foo}, which gets the string name | |
8427 | in export table @samp{foo2}. The @samp{eoo} symbol is an data export | |
8428 | symbol, which gets in export table the name @samp{var1}. | |
4b5bd4e7 | 8429 | |
6b31ad16 DS |
8430 | The optional @code{LIBRARY <name>} command indicates the @emph{internal} |
8431 | name of the output DLL. If @samp{<name>} does not include a suffix, | |
8432 | the default library suffix, @samp{.DLL} is appended. | |
8433 | ||
b45619c0 NC |
8434 | When the .DEF file is used to build an application, rather than a |
8435 | library, the @code{NAME <name>} command should be used instead of | |
6b31ad16 | 8436 | @code{LIBRARY}. If @samp{<name>} does not include a suffix, the default |
c0065db7 | 8437 | executable suffix, @samp{.EXE} is appended. |
6b31ad16 DS |
8438 | |
8439 | With either @code{LIBRARY <name>} or @code{NAME <name>} the optional | |
8440 | specification @code{BASE = <number>} may be used to specify a | |
c0065db7 | 8441 | non-default base address for the image. |
6b31ad16 DS |
8442 | |
8443 | If neither @code{LIBRARY <name>} nor @code{NAME <name>} is specified, | |
a2877985 DS |
8444 | or they specify an empty string, the internal name is the same as the |
8445 | filename specified on the command line. | |
6b31ad16 | 8446 | |
4b5bd4e7 DS |
8447 | The complete specification of an export symbol is: |
8448 | ||
8449 | @example | |
8450 | EXPORTS | |
8451 | ( ( ( <name1> [ = <name2> ] ) | |
8452 | | ( <name1> = <module-name> . <external-name>)) | |
7fcab871 | 8453 | [ @@ <integer> ] [NONAME] [DATA] [CONSTANT] [PRIVATE] [== <name3>] ) * |
c0065db7 | 8454 | @end example |
4b5bd4e7 DS |
8455 | |
8456 | Declares @samp{<name1>} as an exported symbol from the DLL, or declares | |
8457 | @samp{<name1>} as an exported alias for @samp{<name2>}; or declares | |
8458 | @samp{<name1>} as a "forward" alias for the symbol | |
8459 | @samp{<external-name>} in the DLL @samp{<module-name>}. | |
8460 | Optionally, the symbol may be exported by the specified ordinal | |
7fcab871 KT |
8461 | @samp{<integer>} alias. The optional @samp{<name3>} is the to be used |
8462 | string in import/export table for the symbol. | |
4b5bd4e7 DS |
8463 | |
8464 | The optional keywords that follow the declaration indicate: | |
8465 | ||
8466 | @code{NONAME}: Do not put the symbol name in the DLL's export table. It | |
8467 | will still be exported by its ordinal alias (either the value specified | |
8468 | by the .def specification or, otherwise, the value assigned by the | |
8469 | linker). The symbol name, however, does remain visible in the import | |
8470 | library (if any), unless @code{PRIVATE} is also specified. | |
8471 | ||
8472 | @code{DATA}: The symbol is a variable or object, rather than a function. | |
8473 | The import lib will export only an indirect reference to @code{foo} as | |
8474 | the symbol @code{_imp__foo} (ie, @code{foo} must be resolved as | |
8475 | @code{*_imp__foo}). | |
8476 | ||
8477 | @code{CONSTANT}: Like @code{DATA}, but put the undecorated @code{foo} as | |
8478 | well as @code{_imp__foo} into the import library. Both refer to the | |
8479 | read-only import address table's pointer to the variable, not to the | |
8480 | variable itself. This can be dangerous. If the user code fails to add | |
8481 | the @code{dllimport} attribute and also fails to explicitly add the | |
8482 | extra indirection that the use of the attribute enforces, the | |
8483 | application will behave unexpectedly. | |
8484 | ||
8485 | @code{PRIVATE}: Put the symbol in the DLL's export table, but do not put | |
8486 | it into the static import library used to resolve imports at link time. The | |
8487 | symbol can still be imported using the @code{LoadLibrary/GetProcAddress} | |
de194d85 | 8488 | API at runtime or by using the GNU ld extension of linking directly to |
4b5bd4e7 | 8489 | the DLL without an import library. |
c0065db7 | 8490 | |
4b5bd4e7 DS |
8491 | See ld/deffilep.y in the binutils sources for the full specification of |
8492 | other DEF file statements | |
dc8465bf NC |
8493 | |
8494 | @cindex creating a DEF file | |
8495 | While linking a shared dll, @command{ld} is able to create a DEF file | |
a05a5b64 | 8496 | with the @samp{--output-def <file>} command-line option. |
0a5d968e NC |
8497 | |
8498 | @item Using decorations | |
8499 | @cindex Using decorations | |
8500 | Another way of marking symbols for export is to modify the source code | |
8501 | itself, so that when building the DLL each symbol to be exported is | |
8502 | declared as: | |
8503 | ||
8504 | @example | |
8505 | __declspec(dllexport) int a_variable | |
8506 | __declspec(dllexport) void a_function(int with_args) | |
8507 | @end example | |
8508 | ||
8509 | All such symbols will be exported from the DLL. If, however, | |
8510 | any of the object files in the DLL contain symbols decorated in | |
8511 | this way, then the normal auto-export behavior is disabled, unless | |
8512 | the @samp{--export-all-symbols} option is also used. | |
8513 | ||
8514 | Note that object files that wish to access these symbols must @emph{not} | |
c0065db7 | 8515 | decorate them with dllexport. Instead, they should use dllimport, |
0a5d968e NC |
8516 | instead: |
8517 | ||
8518 | @example | |
8519 | __declspec(dllimport) int a_variable | |
8520 | __declspec(dllimport) void a_function(int with_args) | |
8521 | @end example | |
8522 | ||
c0065db7 RM |
8523 | This complicates the structure of library header files, because |
8524 | when included by the library itself the header must declare the | |
0a5d968e NC |
8525 | variables and functions as dllexport, but when included by client |
8526 | code the header must declare them as dllimport. There are a number | |
c0065db7 | 8527 | of idioms that are typically used to do this; often client code can |
0a5d968e NC |
8528 | omit the __declspec() declaration completely. See |
8529 | @samp{--enable-auto-import} and @samp{automatic data imports} for more | |
b45619c0 | 8530 | information. |
c0065db7 | 8531 | @end table |
dc8465bf | 8532 | |
2ca22b03 NC |
8533 | @cindex automatic data imports |
8534 | @item automatic data imports | |
8535 | The standard Windows dll format supports data imports from dlls only | |
69da35b5 | 8536 | by adding special decorations (dllimport/dllexport), which let the |
2ca22b03 | 8537 | compiler produce specific assembler instructions to deal with this |
c0065db7 | 8538 | issue. This increases the effort necessary to port existing Un*x |
69da35b5 | 8539 | code to these platforms, especially for large |
2ca22b03 | 8540 | c++ libraries and applications. The auto-import feature, which was |
c0065db7 | 8541 | initially provided by Paul Sokolovsky, allows one to omit the |
b45619c0 | 8542 | decorations to achieve a behavior that conforms to that on POSIX/Un*x |
c0065db7 | 8543 | platforms. This feature is enabled with the @samp{--enable-auto-import} |
69da35b5 NC |
8544 | command-line option, although it is enabled by default on cygwin/mingw. |
8545 | The @samp{--enable-auto-import} option itself now serves mainly to | |
8546 | suppress any warnings that are ordinarily emitted when linked objects | |
8547 | trigger the feature's use. | |
8548 | ||
c0065db7 | 8549 | auto-import of variables does not always work flawlessly without |
69da35b5 NC |
8550 | additional assistance. Sometimes, you will see this message |
8551 | ||
c0065db7 | 8552 | "variable '<var>' can't be auto-imported. Please read the |
69da35b5 NC |
8553 | documentation for ld's @code{--enable-auto-import} for details." |
8554 | ||
c0065db7 RM |
8555 | The @samp{--enable-auto-import} documentation explains why this error |
8556 | occurs, and several methods that can be used to overcome this difficulty. | |
8557 | One of these methods is the @emph{runtime pseudo-relocs} feature, described | |
69da35b5 NC |
8558 | below. |
8559 | ||
8560 | @cindex runtime pseudo-relocation | |
c0065db7 RM |
8561 | For complex variables imported from DLLs (such as structs or classes), |
8562 | object files typically contain a base address for the variable and an | |
8563 | offset (@emph{addend}) within the variable--to specify a particular | |
8564 | field or public member, for instance. Unfortunately, the runtime loader used | |
8565 | in win32 environments is incapable of fixing these references at runtime | |
69da35b5 | 8566 | without the additional information supplied by dllimport/dllexport decorations. |
c0065db7 | 8567 | The standard auto-import feature described above is unable to resolve these |
69da35b5 NC |
8568 | references. |
8569 | ||
c0065db7 RM |
8570 | The @samp{--enable-runtime-pseudo-relocs} switch allows these references to |
8571 | be resolved without error, while leaving the task of adjusting the references | |
8572 | themselves (with their non-zero addends) to specialized code provided by the | |
8573 | runtime environment. Recent versions of the cygwin and mingw environments and | |
8574 | compilers provide this runtime support; older versions do not. However, the | |
8575 | support is only necessary on the developer's platform; the compiled result will | |
69da35b5 NC |
8576 | run without error on an older system. |
8577 | ||
c0065db7 RM |
8578 | @samp{--enable-runtime-pseudo-relocs} is not the default; it must be explicitly |
8579 | enabled as needed. | |
2ca22b03 NC |
8580 | |
8581 | @cindex direct linking to a dll | |
8582 | @item direct linking to a dll | |
8583 | The cygwin/mingw ports of @command{ld} support the direct linking, | |
8584 | including data symbols, to a dll without the usage of any import | |
69da35b5 | 8585 | libraries. This is much faster and uses much less memory than does the |
b45619c0 | 8586 | traditional import library method, especially when linking large |
c0065db7 RM |
8587 | libraries or applications. When @command{ld} creates an import lib, each |
8588 | function or variable exported from the dll is stored in its own bfd, even | |
8589 | though a single bfd could contain many exports. The overhead involved in | |
69da35b5 | 8590 | storing, loading, and processing so many bfd's is quite large, and explains the |
c0065db7 | 8591 | tremendous time, memory, and storage needed to link against particularly |
69da35b5 NC |
8592 | large or complex libraries when using import libs. |
8593 | ||
c0065db7 | 8594 | Linking directly to a dll uses no extra command-line switches other than |
69da35b5 | 8595 | @samp{-L} and @samp{-l}, because @command{ld} already searches for a number |
c0065db7 | 8596 | of names to match each library. All that is needed from the developer's |
69da35b5 NC |
8597 | perspective is an understanding of this search, in order to force ld to |
8598 | select the dll instead of an import library. | |
8599 | ||
2ca22b03 | 8600 | |
69da35b5 NC |
8601 | For instance, when ld is called with the argument @samp{-lxxx} it will attempt |
8602 | to find, in the first directory of its search path, | |
2ca22b03 NC |
8603 | |
8604 | @example | |
45e948fe NC |
8605 | libxxx.dll.a |
8606 | xxx.dll.a | |
8607 | libxxx.a | |
8608 | xxx.lib | |
f6c4d4b1 | 8609 | libxxx.lib |
69da35b5 | 8610 | cygxxx.dll (*) |
45e948fe NC |
8611 | libxxx.dll |
8612 | xxx.dll | |
2ca22b03 NC |
8613 | @end example |
8614 | ||
69da35b5 NC |
8615 | before moving on to the next directory in the search path. |
8616 | ||
c0065db7 RM |
8617 | (*) Actually, this is not @samp{cygxxx.dll} but in fact is @samp{<prefix>xxx.dll}, |
8618 | where @samp{<prefix>} is set by the @command{ld} option | |
8619 | @samp{--dll-search-prefix=<prefix>}. In the case of cygwin, the standard gcc spec | |
8620 | file includes @samp{--dll-search-prefix=cyg}, so in effect we actually search for | |
69da35b5 NC |
8621 | @samp{cygxxx.dll}. |
8622 | ||
c0065db7 RM |
8623 | Other win32-based unix environments, such as mingw or pw32, may use other |
8624 | @samp{<prefix>}es, although at present only cygwin makes use of this feature. It | |
69da35b5 NC |
8625 | was originally intended to help avoid name conflicts among dll's built for the |
8626 | various win32/un*x environments, so that (for example) two versions of a zlib dll | |
8627 | could coexist on the same machine. | |
8628 | ||
2ca22b03 NC |
8629 | The generic cygwin/mingw path layout uses a @samp{bin} directory for |
8630 | applications and dll's and a @samp{lib} directory for the import | |
69da35b5 | 8631 | libraries (using cygwin nomenclature): |
2ca22b03 NC |
8632 | |
8633 | @example | |
8634 | bin/ | |
8635 | cygxxx.dll | |
8636 | lib/ | |
8637 | libxxx.dll.a (in case of dll's) | |
c0065db7 | 8638 | libxxx.a (in case of static archive) |
2ca22b03 NC |
8639 | @end example |
8640 | ||
c0065db7 RM |
8641 | Linking directly to a dll without using the import library can be |
8642 | done two ways: | |
2ca22b03 NC |
8643 | |
8644 | 1. Use the dll directly by adding the @samp{bin} path to the link line | |
8645 | @example | |
8646 | gcc -Wl,-verbose -o a.exe -L../bin/ -lxxx | |
c0065db7 | 8647 | @end example |
2ca22b03 | 8648 | |
69da35b5 NC |
8649 | However, as the dll's often have version numbers appended to their names |
8650 | (@samp{cygncurses-5.dll}) this will often fail, unless one specifies | |
8651 | @samp{-L../bin -lncurses-5} to include the version. Import libs are generally | |
8652 | not versioned, and do not have this difficulty. | |
8653 | ||
2ca22b03 NC |
8654 | 2. Create a symbolic link from the dll to a file in the @samp{lib} |
8655 | directory according to the above mentioned search pattern. This | |
8656 | should be used to avoid unwanted changes in the tools needed for | |
8657 | making the app/dll. | |
8658 | ||
8659 | @example | |
8660 | ln -s bin/cygxxx.dll lib/[cyg|lib|]xxx.dll[.a] | |
c0065db7 | 8661 | @end example |
2ca22b03 NC |
8662 | |
8663 | Then you can link without any make environment changes. | |
8664 | ||
8665 | @example | |
8666 | gcc -Wl,-verbose -o a.exe -L../lib/ -lxxx | |
c0065db7 | 8667 | @end example |
69da35b5 NC |
8668 | |
8669 | This technique also avoids the version number problems, because the following is | |
8670 | perfectly legal | |
8671 | ||
8672 | @example | |
8673 | bin/ | |
8674 | cygxxx-5.dll | |
8675 | lib/ | |
c0065db7 | 8676 | libxxx.dll.a -> ../bin/cygxxx-5.dll |
69da35b5 NC |
8677 | @end example |
8678 | ||
dc8465bf | 8679 | Linking directly to a dll without using an import lib will work |
69da35b5 NC |
8680 | even when auto-import features are exercised, and even when |
8681 | @samp{--enable-runtime-pseudo-relocs} is used. | |
8682 | ||
8683 | Given the improvements in speed and memory usage, one might justifiably | |
45e948fe | 8684 | wonder why import libraries are used at all. There are three reasons: |
69da35b5 NC |
8685 | |
8686 | 1. Until recently, the link-directly-to-dll functionality did @emph{not} | |
8687 | work with auto-imported data. | |
8688 | ||
dc8465bf NC |
8689 | 2. Sometimes it is necessary to include pure static objects within the |
8690 | import library (which otherwise contains only bfd's for indirection | |
8691 | symbols that point to the exports of a dll). Again, the import lib | |
8692 | for the cygwin kernel makes use of this ability, and it is not | |
8693 | possible to do this without an import lib. | |
69da35b5 | 8694 | |
45e948fe NC |
8695 | 3. Symbol aliases can only be resolved using an import lib. This is |
8696 | critical when linking against OS-supplied dll's (eg, the win32 API) | |
8697 | in which symbols are usually exported as undecorated aliases of their | |
8698 | stdcall-decorated assembly names. | |
8699 | ||
69da35b5 | 8700 | So, import libs are not going away. But the ability to replace |
c0065db7 RM |
8701 | true import libs with a simple symbolic link to (or a copy of) |
8702 | a dll, in many cases, is a useful addition to the suite of tools | |
8703 | binutils makes available to the win32 developer. Given the | |
69da35b5 NC |
8704 | massive improvements in memory requirements during linking, storage |
8705 | requirements, and linking speed, we expect that many developers | |
8706 | will soon begin to use this feature whenever possible. | |
dc8465bf | 8707 | |
c0065db7 | 8708 | @item symbol aliasing |
dc8465bf | 8709 | @table @emph |
c0065db7 RM |
8710 | @item adding additional names |
8711 | Sometimes, it is useful to export symbols with additional names. | |
dc8465bf NC |
8712 | A symbol @samp{foo} will be exported as @samp{foo}, but it can also be |
8713 | exported as @samp{_foo} by using special directives in the DEF file | |
8714 | when creating the dll. This will affect also the optional created | |
c0065db7 | 8715 | import library. Consider the following DEF file: |
dc8465bf | 8716 | |
c0065db7 | 8717 | @example |
dc8465bf NC |
8718 | LIBRARY "xyz.dll" BASE=0x61000000 |
8719 | ||
8720 | EXPORTS | |
c0065db7 | 8721 | foo |
dc8465bf | 8722 | _foo = foo |
c0065db7 | 8723 | @end example |
dc8465bf NC |
8724 | |
8725 | The line @samp{_foo = foo} maps the symbol @samp{foo} to @samp{_foo}. | |
8726 | ||
8727 | Another method for creating a symbol alias is to create it in the | |
8728 | source code using the "weak" attribute: | |
8729 | ||
c0065db7 RM |
8730 | @example |
8731 | void foo () @{ /* Do something. */; @} | |
dc8465bf | 8732 | void _foo () __attribute__ ((weak, alias ("foo"))); |
c0065db7 | 8733 | @end example |
dc8465bf NC |
8734 | |
8735 | See the gcc manual for more information about attributes and weak | |
8736 | symbols. | |
8737 | ||
8738 | @item renaming symbols | |
8739 | Sometimes it is useful to rename exports. For instance, the cygwin | |
c0065db7 | 8740 | kernel does this regularly. A symbol @samp{_foo} can be exported as |
dc8465bf NC |
8741 | @samp{foo} but not as @samp{_foo} by using special directives in the |
8742 | DEF file. (This will also affect the import library, if it is | |
c0065db7 | 8743 | created). In the following example: |
dc8465bf | 8744 | |
c0065db7 | 8745 | @example |
dc8465bf NC |
8746 | LIBRARY "xyz.dll" BASE=0x61000000 |
8747 | ||
8748 | EXPORTS | |
8749 | _foo = foo | |
c0065db7 | 8750 | @end example |
dc8465bf NC |
8751 | |
8752 | The line @samp{_foo = foo} maps the exported symbol @samp{foo} to | |
8753 | @samp{_foo}. | |
c0065db7 | 8754 | @end table |
dc8465bf | 8755 | |
0a5d968e | 8756 | Note: using a DEF file disables the default auto-export behavior, |
a05a5b64 | 8757 | unless the @samp{--export-all-symbols} command-line option is used. |
0a5d968e | 8758 | If, however, you are trying to rename symbols, then you should list |
c0065db7 RM |
8759 | @emph{all} desired exports in the DEF file, including the symbols |
8760 | that are not being renamed, and do @emph{not} use the | |
8761 | @samp{--export-all-symbols} option. If you list only the | |
8762 | renamed symbols in the DEF file, and use @samp{--export-all-symbols} | |
8763 | to handle the other symbols, then the both the new names @emph{and} | |
8764 | the original names for the renamed symbols will be exported. | |
8765 | In effect, you'd be aliasing those symbols, not renaming them, | |
0a5d968e | 8766 | which is probably not what you wanted. |
c87db184 CF |
8767 | |
8768 | @cindex weak externals | |
8769 | @item weak externals | |
8770 | The Windows object format, PE, specifies a form of weak symbols called | |
8771 | weak externals. When a weak symbol is linked and the symbol is not | |
8772 | defined, the weak symbol becomes an alias for some other symbol. There | |
8773 | are three variants of weak externals: | |
8774 | @itemize | |
8775 | @item Definition is searched for in objects and libraries, historically | |
8776 | called lazy externals. | |
8777 | @item Definition is searched for only in other objects, not in libraries. | |
8778 | This form is not presently implemented. | |
8779 | @item No search; the symbol is an alias. This form is not presently | |
8780 | implemented. | |
8781 | @end itemize | |
8782 | As a GNU extension, weak symbols that do not specify an alternate symbol | |
8783 | are supported. If the symbol is undefined when linking, the symbol | |
8784 | uses a default value. | |
c1711530 DK |
8785 | |
8786 | @cindex aligned common symbols | |
8787 | @item aligned common symbols | |
8788 | As a GNU extension to the PE file format, it is possible to specify the | |
8789 | desired alignment for a common symbol. This information is conveyed from | |
8790 | the assembler or compiler to the linker by means of GNU-specific commands | |
8791 | carried in the object file's @samp{.drectve} section, which are recognized | |
8792 | by @command{ld} and respected when laying out the common symbols. Native | |
8793 | tools will be able to process object files employing this GNU extension, | |
8794 | but will fail to respect the alignment instructions, and may issue noisy | |
8795 | warnings about unknown linker directives. | |
5063daf7 | 8796 | |
2ca22b03 NC |
8797 | @end table |
8798 | ||
8799 | @ifclear GENERIC | |
8800 | @lowersections | |
8801 | @end ifclear | |
8802 | @end ifset | |
8803 | ||
e0001a05 NC |
8804 | @ifset XTENSA |
8805 | @ifclear GENERIC | |
8806 | @raisesections | |
8807 | @end ifclear | |
8808 | ||
8809 | @node Xtensa | |
8810 | @section @code{ld} and Xtensa Processors | |
8811 | ||
8812 | @cindex Xtensa processors | |
8813 | The default @command{ld} behavior for Xtensa processors is to interpret | |
8814 | @code{SECTIONS} commands so that lists of explicitly named sections in a | |
8815 | specification with a wildcard file will be interleaved when necessary to | |
8816 | keep literal pools within the range of PC-relative load offsets. For | |
8817 | example, with the command: | |
8818 | ||
8819 | @smallexample | |
8820 | SECTIONS | |
8821 | @{ | |
8822 | .text : @{ | |
8823 | *(.literal .text) | |
8824 | @} | |
8825 | @} | |
8826 | @end smallexample | |
8827 | ||
8828 | @noindent | |
8829 | @command{ld} may interleave some of the @code{.literal} | |
8830 | and @code{.text} sections from different object files to ensure that the | |
8831 | literal pools are within the range of PC-relative load offsets. A valid | |
8832 | interleaving might place the @code{.literal} sections from an initial | |
8833 | group of files followed by the @code{.text} sections of that group of | |
8834 | files. Then, the @code{.literal} sections from the rest of the files | |
8835 | and the @code{.text} sections from the rest of the files would follow. | |
e0001a05 | 8836 | |
43cd72b9 | 8837 | @cindex @option{--relax} on Xtensa |
e0001a05 | 8838 | @cindex relaxing on Xtensa |
43cd72b9 BW |
8839 | Relaxation is enabled by default for the Xtensa version of @command{ld} and |
8840 | provides two important link-time optimizations. The first optimization | |
8841 | is to combine identical literal values to reduce code size. A redundant | |
8842 | literal will be removed and all the @code{L32R} instructions that use it | |
8843 | will be changed to reference an identical literal, as long as the | |
8844 | location of the replacement literal is within the offset range of all | |
8845 | the @code{L32R} instructions. The second optimization is to remove | |
8846 | unnecessary overhead from assembler-generated ``longcall'' sequences of | |
8847 | @code{L32R}/@code{CALLX@var{n}} when the target functions are within | |
8848 | range of direct @code{CALL@var{n}} instructions. | |
8849 | ||
8850 | For each of these cases where an indirect call sequence can be optimized | |
8851 | to a direct call, the linker will change the @code{CALLX@var{n}} | |
8852 | instruction to a @code{CALL@var{n}} instruction, remove the @code{L32R} | |
8853 | instruction, and remove the literal referenced by the @code{L32R} | |
8854 | instruction if it is not used for anything else. Removing the | |
8855 | @code{L32R} instruction always reduces code size but can potentially | |
8856 | hurt performance by changing the alignment of subsequent branch targets. | |
8857 | By default, the linker will always preserve alignments, either by | |
8858 | switching some instructions between 24-bit encodings and the equivalent | |
8859 | density instructions or by inserting a no-op in place of the @code{L32R} | |
8860 | instruction that was removed. If code size is more important than | |
8861 | performance, the @option{--size-opt} option can be used to prevent the | |
8862 | linker from widening density instructions or inserting no-ops, except in | |
8863 | a few cases where no-ops are required for correctness. | |
8864 | ||
8865 | The following Xtensa-specific command-line options can be used to | |
8866 | control the linker: | |
8867 | ||
8868 | @cindex Xtensa options | |
8869 | @table @option | |
43cd72b9 BW |
8870 | @item --size-opt |
8871 | When optimizing indirect calls to direct calls, optimize for code size | |
8872 | more than performance. With this option, the linker will not insert | |
8873 | no-ops or widen density instructions to preserve branch target | |
8874 | alignment. There may still be some cases where no-ops are required to | |
8875 | preserve the correctness of the code. | |
7a77f1ac MF |
8876 | |
8877 | @item --abi-windowed | |
8878 | @itemx --abi-call0 | |
8879 | Choose ABI for the output object and for the generated PLT code. | |
8880 | PLT code inserted by the linker must match ABI of the output object | |
8881 | because windowed and call0 ABI use incompatible function call | |
8882 | conventions. | |
8883 | Default ABI is chosen by the ABI tag in the @code{.xtensa.info} section | |
8884 | of the first input object. | |
8885 | A warning is issued if ABI tags of input objects do not match each other | |
8886 | or the chosen output object ABI. | |
43cd72b9 | 8887 | @end table |
e0001a05 NC |
8888 | |
8889 | @ifclear GENERIC | |
8890 | @lowersections | |
8891 | @end ifclear | |
8892 | @end ifset | |
8893 | ||
252b5132 RH |
8894 | @ifclear SingleFormat |
8895 | @node BFD | |
8896 | @chapter BFD | |
8897 | ||
8898 | @cindex back end | |
8899 | @cindex object file management | |
8900 | @cindex object formats available | |
8901 | @kindex objdump -i | |
8902 | The linker accesses object and archive files using the BFD libraries. | |
8903 | These libraries allow the linker to use the same routines to operate on | |
8904 | object files whatever the object file format. A different object file | |
8905 | format can be supported simply by creating a new BFD back end and adding | |
8906 | it to the library. To conserve runtime memory, however, the linker and | |
8907 | associated tools are usually configured to support only a subset of the | |
8908 | object file formats available. You can use @code{objdump -i} | |
8909 | (@pxref{objdump,,objdump,binutils.info,The GNU Binary Utilities}) to | |
8910 | list all the formats available for your configuration. | |
8911 | ||
8912 | @cindex BFD requirements | |
8913 | @cindex requirements for BFD | |
8914 | As with most implementations, BFD is a compromise between | |
8915 | several conflicting requirements. The major factor influencing | |
8916 | BFD design was efficiency: any time used converting between | |
8917 | formats is time which would not have been spent had BFD not | |
8918 | been involved. This is partly offset by abstraction payback; since | |
8919 | BFD simplifies applications and back ends, more time and care | |
8920 | may be spent optimizing algorithms for a greater speed. | |
8921 | ||
8922 | One minor artifact of the BFD solution which you should bear in | |
8923 | mind is the potential for information loss. There are two places where | |
8924 | useful information can be lost using the BFD mechanism: during | |
8925 | conversion and during output. @xref{BFD information loss}. | |
8926 | ||
8927 | @menu | |
8928 | * BFD outline:: How it works: an outline of BFD | |
8929 | @end menu | |
8930 | ||
8931 | @node BFD outline | |
36f63dca | 8932 | @section How It Works: An Outline of BFD |
252b5132 RH |
8933 | @cindex opening object files |
8934 | @include bfdsumm.texi | |
8935 | @end ifclear | |
8936 | ||
8937 | @node Reporting Bugs | |
8938 | @chapter Reporting Bugs | |
ff5dcc92 SC |
8939 | @cindex bugs in @command{ld} |
8940 | @cindex reporting bugs in @command{ld} | |
252b5132 | 8941 | |
ff5dcc92 | 8942 | Your bug reports play an essential role in making @command{ld} reliable. |
252b5132 RH |
8943 | |
8944 | Reporting a bug may help you by bringing a solution to your problem, or | |
8945 | it may not. But in any case the principal function of a bug report is | |
ff5dcc92 | 8946 | to help the entire community by making the next version of @command{ld} |
252b5132 | 8947 | work better. Bug reports are your contribution to the maintenance of |
ff5dcc92 | 8948 | @command{ld}. |
252b5132 RH |
8949 | |
8950 | In order for a bug report to serve its purpose, you must include the | |
8951 | information that enables us to fix the bug. | |
8952 | ||
8953 | @menu | |
8954 | * Bug Criteria:: Have you found a bug? | |
8955 | * Bug Reporting:: How to report bugs | |
8956 | @end menu | |
8957 | ||
8958 | @node Bug Criteria | |
36f63dca | 8959 | @section Have You Found a Bug? |
252b5132 RH |
8960 | @cindex bug criteria |
8961 | ||
8962 | If you are not sure whether you have found a bug, here are some guidelines: | |
8963 | ||
8964 | @itemize @bullet | |
8965 | @cindex fatal signal | |
8966 | @cindex linker crash | |
8967 | @cindex crash of linker | |
8968 | @item | |
8969 | If the linker gets a fatal signal, for any input whatever, that is a | |
ff5dcc92 | 8970 | @command{ld} bug. Reliable linkers never crash. |
252b5132 RH |
8971 | |
8972 | @cindex error on valid input | |
8973 | @item | |
ff5dcc92 | 8974 | If @command{ld} produces an error message for valid input, that is a bug. |
252b5132 RH |
8975 | |
8976 | @cindex invalid input | |
8977 | @item | |
ff5dcc92 | 8978 | If @command{ld} does not produce an error message for invalid input, that |
252b5132 RH |
8979 | may be a bug. In the general case, the linker can not verify that |
8980 | object files are correct. | |
8981 | ||
8982 | @item | |
8983 | If you are an experienced user of linkers, your suggestions for | |
ff5dcc92 | 8984 | improvement of @command{ld} are welcome in any case. |
252b5132 RH |
8985 | @end itemize |
8986 | ||
8987 | @node Bug Reporting | |
36f63dca | 8988 | @section How to Report Bugs |
252b5132 | 8989 | @cindex bug reports |
ff5dcc92 | 8990 | @cindex @command{ld} bugs, reporting |
252b5132 RH |
8991 | |
8992 | A number of companies and individuals offer support for @sc{gnu} | |
ff5dcc92 | 8993 | products. If you obtained @command{ld} from a support organization, we |
252b5132 RH |
8994 | recommend you contact that organization first. |
8995 | ||
8996 | You can find contact information for many support companies and | |
8997 | individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs | |
8998 | distribution. | |
8999 | ||
ad22bfe8 | 9000 | @ifset BUGURL |
ff5dcc92 | 9001 | Otherwise, send bug reports for @command{ld} to |
ad22bfe8 JM |
9002 | @value{BUGURL}. |
9003 | @end ifset | |
252b5132 RH |
9004 | |
9005 | The fundamental principle of reporting bugs usefully is this: | |
9006 | @strong{report all the facts}. If you are not sure whether to state a | |
9007 | fact or leave it out, state it! | |
9008 | ||
9009 | Often people omit facts because they think they know what causes the | |
9010 | problem and assume that some details do not matter. Thus, you might | |
b553b183 NC |
9011 | assume that the name of a symbol you use in an example does not |
9012 | matter. Well, probably it does not, but one cannot be sure. Perhaps | |
9013 | the bug is a stray memory reference which happens to fetch from the | |
9014 | location where that name is stored in memory; perhaps, if the name | |
9015 | were different, the contents of that location would fool the linker | |
9016 | into doing the right thing despite the bug. Play it safe and give a | |
9017 | specific, complete example. That is the easiest thing for you to do, | |
c0065db7 | 9018 | and the most helpful. |
b553b183 NC |
9019 | |
9020 | Keep in mind that the purpose of a bug report is to enable us to fix | |
9021 | the bug if it is new to us. Therefore, always write your bug reports | |
9022 | on the assumption that the bug has not been reported previously. | |
252b5132 RH |
9023 | |
9024 | Sometimes people give a few sketchy facts and ask, ``Does this ring a | |
36f63dca NC |
9025 | bell?'' This cannot help us fix a bug, so it is basically useless. We |
9026 | respond by asking for enough details to enable us to investigate. | |
9027 | You might as well expedite matters by sending them to begin with. | |
252b5132 RH |
9028 | |
9029 | To enable us to fix the bug, you should include all these things: | |
9030 | ||
9031 | @itemize @bullet | |
9032 | @item | |
ff5dcc92 | 9033 | The version of @command{ld}. @command{ld} announces it if you start it with |
252b5132 RH |
9034 | the @samp{--version} argument. |
9035 | ||
9036 | Without this, we will not know whether there is any point in looking for | |
ff5dcc92 | 9037 | the bug in the current version of @command{ld}. |
252b5132 RH |
9038 | |
9039 | @item | |
ff5dcc92 | 9040 | Any patches you may have applied to the @command{ld} source, including any |
252b5132 RH |
9041 | patches made to the @code{BFD} library. |
9042 | ||
9043 | @item | |
9044 | The type of machine you are using, and the operating system name and | |
9045 | version number. | |
9046 | ||
9047 | @item | |
ff5dcc92 | 9048 | What compiler (and its version) was used to compile @command{ld}---e.g. |
252b5132 RH |
9049 | ``@code{gcc-2.7}''. |
9050 | ||
9051 | @item | |
9052 | The command arguments you gave the linker to link your example and | |
9053 | observe the bug. To guarantee you will not omit something important, | |
9054 | list them all. A copy of the Makefile (or the output from make) is | |
9055 | sufficient. | |
9056 | ||
9057 | If we were to try to guess the arguments, we would probably guess wrong | |
9058 | and then we might not encounter the bug. | |
9059 | ||
9060 | @item | |
9061 | A complete input file, or set of input files, that will reproduce the | |
b553b183 NC |
9062 | bug. It is generally most helpful to send the actual object files |
9063 | provided that they are reasonably small. Say no more than 10K. For | |
9064 | bigger files you can either make them available by FTP or HTTP or else | |
9065 | state that you are willing to send the object file(s) to whomever | |
9066 | requests them. (Note - your email will be going to a mailing list, so | |
9067 | we do not want to clog it up with large attachments). But small | |
9068 | attachments are best. | |
252b5132 RH |
9069 | |
9070 | If the source files were assembled using @code{gas} or compiled using | |
9071 | @code{gcc}, then it may be OK to send the source files rather than the | |
9072 | object files. In this case, be sure to say exactly what version of | |
9073 | @code{gas} or @code{gcc} was used to produce the object files. Also say | |
9074 | how @code{gas} or @code{gcc} were configured. | |
9075 | ||
9076 | @item | |
9077 | A description of what behavior you observe that you believe is | |
9078 | incorrect. For example, ``It gets a fatal signal.'' | |
9079 | ||
ff5dcc92 | 9080 | Of course, if the bug is that @command{ld} gets a fatal signal, then we |
252b5132 RH |
9081 | will certainly notice it. But if the bug is incorrect output, we might |
9082 | not notice unless it is glaringly wrong. You might as well not give us | |
9083 | a chance to make a mistake. | |
9084 | ||
9085 | Even if the problem you experience is a fatal signal, you should still | |
9086 | say so explicitly. Suppose something strange is going on, such as, your | |
b45619c0 | 9087 | copy of @command{ld} is out of sync, or you have encountered a bug in the |
252b5132 RH |
9088 | C library on your system. (This has happened!) Your copy might crash |
9089 | and ours would not. If you told us to expect a crash, then when ours | |
9090 | fails to crash, we would know that the bug was not happening for us. If | |
9091 | you had not told us to expect a crash, then we would not be able to draw | |
9092 | any conclusion from our observations. | |
9093 | ||
9094 | @item | |
ff5dcc92 | 9095 | If you wish to suggest changes to the @command{ld} source, send us context |
252b5132 RH |
9096 | diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or |
9097 | @samp{-p} option. Always send diffs from the old file to the new file. | |
ff5dcc92 | 9098 | If you even discuss something in the @command{ld} source, refer to it by |
252b5132 RH |
9099 | context, not by line number. |
9100 | ||
9101 | The line numbers in our development sources will not match those in your | |
9102 | sources. Your line numbers would convey no useful information to us. | |
9103 | @end itemize | |
9104 | ||
9105 | Here are some things that are not necessary: | |
9106 | ||
9107 | @itemize @bullet | |
9108 | @item | |
9109 | A description of the envelope of the bug. | |
9110 | ||
9111 | Often people who encounter a bug spend a lot of time investigating | |
9112 | which changes to the input file will make the bug go away and which | |
9113 | changes will not affect it. | |
9114 | ||
9115 | This is often time consuming and not very useful, because the way we | |
9116 | will find the bug is by running a single example under the debugger | |
9117 | with breakpoints, not by pure deduction from a series of examples. | |
9118 | We recommend that you save your time for something else. | |
9119 | ||
9120 | Of course, if you can find a simpler example to report @emph{instead} | |
9121 | of the original one, that is a convenience for us. Errors in the | |
9122 | output will be easier to spot, running under the debugger will take | |
9123 | less time, and so on. | |
9124 | ||
9125 | However, simplification is not vital; if you do not want to do this, | |
9126 | report the bug anyway and send us the entire test case you used. | |
9127 | ||
9128 | @item | |
9129 | A patch for the bug. | |
9130 | ||
9131 | A patch for the bug does help us if it is a good one. But do not omit | |
9132 | the necessary information, such as the test case, on the assumption that | |
9133 | a patch is all we need. We might see problems with your patch and decide | |
9134 | to fix the problem another way, or we might not understand it at all. | |
9135 | ||
ff5dcc92 | 9136 | Sometimes with a program as complicated as @command{ld} it is very hard to |
252b5132 RH |
9137 | construct an example that will make the program follow a certain path |
9138 | through the code. If you do not send us the example, we will not be | |
9139 | able to construct one, so we will not be able to verify that the bug is | |
9140 | fixed. | |
9141 | ||
9142 | And if we cannot understand what bug you are trying to fix, or why your | |
9143 | patch should be an improvement, we will not install it. A test case will | |
9144 | help us to understand. | |
9145 | ||
9146 | @item | |
9147 | A guess about what the bug is or what it depends on. | |
9148 | ||
9149 | Such guesses are usually wrong. Even we cannot guess right about such | |
9150 | things without first using the debugger to find the facts. | |
9151 | @end itemize | |
9152 | ||
9153 | @node MRI | |
9154 | @appendix MRI Compatible Script Files | |
9155 | @cindex MRI compatibility | |
ff5dcc92 SC |
9156 | To aid users making the transition to @sc{gnu} @command{ld} from the MRI |
9157 | linker, @command{ld} can use MRI compatible linker scripts as an | |
252b5132 RH |
9158 | alternative to the more general-purpose linker scripting language |
9159 | described in @ref{Scripts}. MRI compatible linker scripts have a much | |
9160 | simpler command set than the scripting language otherwise used with | |
ff5dcc92 | 9161 | @command{ld}. @sc{gnu} @command{ld} supports the most commonly used MRI |
252b5132 RH |
9162 | linker commands; these commands are described here. |
9163 | ||
9164 | In general, MRI scripts aren't of much use with the @code{a.out} object | |
9165 | file format, since it only has three sections and MRI scripts lack some | |
9166 | features to make use of them. | |
9167 | ||
9168 | You can specify a file containing an MRI-compatible script using the | |
9169 | @samp{-c} command-line option. | |
9170 | ||
9171 | Each command in an MRI-compatible script occupies its own line; each | |
9172 | command line starts with the keyword that identifies the command (though | |
9173 | blank lines are also allowed for punctuation). If a line of an | |
ff5dcc92 | 9174 | MRI-compatible script begins with an unrecognized keyword, @command{ld} |
252b5132 RH |
9175 | issues a warning message, but continues processing the script. |
9176 | ||
9177 | Lines beginning with @samp{*} are comments. | |
9178 | ||
9179 | You can write these commands using all upper-case letters, or all | |
9180 | lower case; for example, @samp{chip} is the same as @samp{CHIP}. | |
9181 | The following list shows only the upper-case form of each command. | |
9182 | ||
9183 | @table @code | |
9184 | @cindex @code{ABSOLUTE} (MRI) | |
9185 | @item ABSOLUTE @var{secname} | |
9186 | @itemx ABSOLUTE @var{secname}, @var{secname}, @dots{} @var{secname} | |
ff5dcc92 | 9187 | Normally, @command{ld} includes in the output file all sections from all |
252b5132 RH |
9188 | the input files. However, in an MRI-compatible script, you can use the |
9189 | @code{ABSOLUTE} command to restrict the sections that will be present in | |
9190 | your output program. If the @code{ABSOLUTE} command is used at all in a | |
9191 | script, then only the sections named explicitly in @code{ABSOLUTE} | |
9192 | commands will appear in the linker output. You can still use other | |
9193 | input sections (whatever you select on the command line, or using | |
9194 | @code{LOAD}) to resolve addresses in the output file. | |
9195 | ||
9196 | @cindex @code{ALIAS} (MRI) | |
9197 | @item ALIAS @var{out-secname}, @var{in-secname} | |
9198 | Use this command to place the data from input section @var{in-secname} | |
9199 | in a section called @var{out-secname} in the linker output file. | |
9200 | ||
9201 | @var{in-secname} may be an integer. | |
9202 | ||
9203 | @cindex @code{ALIGN} (MRI) | |
9204 | @item ALIGN @var{secname} = @var{expression} | |
9205 | Align the section called @var{secname} to @var{expression}. The | |
9206 | @var{expression} should be a power of two. | |
9207 | ||
9208 | @cindex @code{BASE} (MRI) | |
9209 | @item BASE @var{expression} | |
9210 | Use the value of @var{expression} as the lowest address (other than | |
9211 | absolute addresses) in the output file. | |
9212 | ||
9213 | @cindex @code{CHIP} (MRI) | |
9214 | @item CHIP @var{expression} | |
9215 | @itemx CHIP @var{expression}, @var{expression} | |
9216 | This command does nothing; it is accepted only for compatibility. | |
9217 | ||
9218 | @cindex @code{END} (MRI) | |
9219 | @item END | |
9220 | This command does nothing whatever; it's only accepted for compatibility. | |
9221 | ||
9222 | @cindex @code{FORMAT} (MRI) | |
9223 | @item FORMAT @var{output-format} | |
9224 | Similar to the @code{OUTPUT_FORMAT} command in the more general linker | |
dc12032b | 9225 | language, but restricted to S-records, if @var{output-format} is @samp{S} |
252b5132 RH |
9226 | |
9227 | @cindex @code{LIST} (MRI) | |
9228 | @item LIST @var{anything}@dots{} | |
9229 | Print (to the standard output file) a link map, as produced by the | |
ff5dcc92 | 9230 | @command{ld} command-line option @samp{-M}. |
252b5132 RH |
9231 | |
9232 | The keyword @code{LIST} may be followed by anything on the | |
9233 | same line, with no change in its effect. | |
9234 | ||
9235 | @cindex @code{LOAD} (MRI) | |
9236 | @item LOAD @var{filename} | |
9237 | @itemx LOAD @var{filename}, @var{filename}, @dots{} @var{filename} | |
9238 | Include one or more object file @var{filename} in the link; this has the | |
ff5dcc92 | 9239 | same effect as specifying @var{filename} directly on the @command{ld} |
252b5132 RH |
9240 | command line. |
9241 | ||
9242 | @cindex @code{NAME} (MRI) | |
9243 | @item NAME @var{output-name} | |
ff5dcc92 | 9244 | @var{output-name} is the name for the program produced by @command{ld}; the |
252b5132 RH |
9245 | MRI-compatible command @code{NAME} is equivalent to the command-line |
9246 | option @samp{-o} or the general script language command @code{OUTPUT}. | |
9247 | ||
9248 | @cindex @code{ORDER} (MRI) | |
9249 | @item ORDER @var{secname}, @var{secname}, @dots{} @var{secname} | |
9250 | @itemx ORDER @var{secname} @var{secname} @var{secname} | |
ff5dcc92 | 9251 | Normally, @command{ld} orders the sections in its output file in the |
252b5132 RH |
9252 | order in which they first appear in the input files. In an MRI-compatible |
9253 | script, you can override this ordering with the @code{ORDER} command. The | |
9254 | sections you list with @code{ORDER} will appear first in your output | |
9255 | file, in the order specified. | |
9256 | ||
9257 | @cindex @code{PUBLIC} (MRI) | |
9258 | @item PUBLIC @var{name}=@var{expression} | |
9259 | @itemx PUBLIC @var{name},@var{expression} | |
9260 | @itemx PUBLIC @var{name} @var{expression} | |
9261 | Supply a value (@var{expression}) for external symbol | |
9262 | @var{name} used in the linker input files. | |
9263 | ||
9264 | @cindex @code{SECT} (MRI) | |
9265 | @item SECT @var{secname}, @var{expression} | |
9266 | @itemx SECT @var{secname}=@var{expression} | |
9267 | @itemx SECT @var{secname} @var{expression} | |
9268 | You can use any of these three forms of the @code{SECT} command to | |
9269 | specify the start address (@var{expression}) for section @var{secname}. | |
9270 | If you have more than one @code{SECT} statement for the same | |
9271 | @var{secname}, only the @emph{first} sets the start address. | |
9272 | @end table | |
9273 | ||
793c5807 NC |
9274 | @node GNU Free Documentation License |
9275 | @appendix GNU Free Documentation License | |
36f63dca | 9276 | @include fdl.texi |
704c465c | 9277 | |
370b66a1 CD |
9278 | @node LD Index |
9279 | @unnumbered LD Index | |
252b5132 RH |
9280 | |
9281 | @printindex cp | |
9282 | ||
9283 | @tex | |
7ca01ed9 | 9284 | % I think something like @@colophon should be in texinfo. In the |
252b5132 RH |
9285 | % meantime: |
9286 | \long\def\colophon{\hbox to0pt{}\vfill | |
9287 | \centerline{The body of this manual is set in} | |
9288 | \centerline{\fontname\tenrm,} | |
9289 | \centerline{with headings in {\bf\fontname\tenbf}} | |
9290 | \centerline{and examples in {\tt\fontname\tentt}.} | |
9291 | \centerline{{\it\fontname\tenit\/} and} | |
9292 | \centerline{{\sl\fontname\tensl\/}} | |
9293 | \centerline{are used for emphasis.}\vfill} | |
9294 | \page\colophon | |
7ca01ed9 | 9295 | % Blame: doc@@cygnus.com, 28mar91. |
252b5132 RH |
9296 | @end tex |
9297 | ||
252b5132 | 9298 | @bye |