Commit | Line | Data |
---|---|---|
a933dad1 | 1 | This file describes various problems that have been encountered |
0a4dd4e4 | 2 | in compiling, installing and running GNU Emacs. Try doing Ctl-C Ctl-t |
9dc15871 | 3 | and browsing through the outline headers. |
a933dad1 | 4 | |
9dc15871 | 5 | * Emacs startup failures |
32364f49 | 6 | |
9dc15871 | 7 | ** Emacs fails to start, complaining about missing fonts. |
32364f49 | 8 | |
9dc15871 | 9 | A typical error message might be something like |
32364f49 | 10 | |
9dc15871 | 11 | No fonts match `-*-fixed-medium-r-*--6-*-*-*-*-*-iso8859-1' |
984002eb | 12 | |
9dc15871 EZ |
13 | This happens because some X resource specifies a bad font family for |
14 | Emacs to use. The possible places where this specification might be | |
15 | are: | |
984002eb | 16 | |
9dc15871 | 17 | - in your ~/.Xdefaults file |
984002eb | 18 | |
9dc15871 EZ |
19 | - client-side X resource file, such as ~/Emacs or |
20 | /usr/X11R6/lib/app-defaults/Emacs or | |
21 | /usr/X11R6/lib/X11/app-defaults/Emacs | |
984002eb | 22 | |
9dc15871 EZ |
23 | One of these files might have bad or malformed specification of a |
24 | fontset that Emacs should use. To fix the problem, you need to find | |
25 | the problematic line(s) and correct them. | |
984002eb | 26 | |
9dc15871 | 27 | ** Emacs aborts while starting up, only when run without X. |
984002eb | 28 | |
9dc15871 EZ |
29 | This problem often results from compiling Emacs with GCC when GCC was |
30 | installed incorrectly. The usual error in installing GCC is to | |
31 | specify --includedir=/usr/include. Installation of GCC makes | |
32 | corrected copies of the system header files. GCC is supposed to use | |
33 | the corrected copies in preference to the original system headers. | |
34 | Specifying --includedir=/usr/include causes the original system header | |
35 | files to be used. On some systems, the definition of ioctl in the | |
36 | original system header files is invalid for ANSI C and causes Emacs | |
37 | not to work. | |
984002eb | 38 | |
9dc15871 EZ |
39 | The fix is to reinstall GCC, and this time do not specify --includedir |
40 | when you configure it. Then recompile Emacs. Specifying --includedir | |
41 | is appropriate only in very special cases and it should *never* be the | |
42 | same directory where system header files are kept. | |
984002eb | 43 | |
9dc15871 | 44 | ** Emacs does not start, complaining that it cannot open termcap database file. |
984002eb | 45 | |
9dc15871 EZ |
46 | If your system uses Terminfo rather than termcap (most modern |
47 | systems do), this could happen if the proper version of | |
48 | ncurses is not visible to the Emacs configure script (i.e. it | |
49 | cannot be found along the usual path the linker looks for | |
50 | libraries). It can happen because your version of ncurses is | |
51 | obsolete, or is available only in form of binaries. | |
f16a1bf6 | 52 | |
9dc15871 EZ |
53 | The solution is to install an up-to-date version of ncurses in |
54 | the developer's form (header files, static libraries and | |
55 | symbolic links); in some GNU/Linux distributions (e.g. Debian) | |
56 | it constitutes a separate package. | |
f16a1bf6 | 57 | |
9dc15871 | 58 | ** Emacs 20 and later fails to load Lisp files at startup. |
5b4ffca2 | 59 | |
9dc15871 | 60 | The typical error message might be like this: |
5b4ffca2 | 61 | |
9dc15871 | 62 | "Cannot open load file: fontset" |
24efe898 | 63 | |
9dc15871 EZ |
64 | This could happen if you compress the file lisp/subdirs.el. That file |
65 | tells Emacs what are the directories where it should look for Lisp | |
66 | files. Emacs cannot work with subdirs.el compressed, since the | |
67 | Auto-compress mode it needs for this will not be loaded until later, | |
68 | when your .emacs file is processed. (The package `fontset.el' is | |
69 | required to set up fonts used to display text on window systems, and | |
70 | it's loaded very early in the startup procedure.) | |
24efe898 | 71 | |
9dc15871 EZ |
72 | Similarly, any other .el file for which there's no corresponding .elc |
73 | file could fail to load if it is compressed. | |
24efe898 | 74 | |
9dc15871 EZ |
75 | The solution is to uncompress all .el files which don't have a .elc |
76 | file. | |
f0f62f71 | 77 | |
9dc15871 EZ |
78 | Another possible reason for such failures is stale *.elc files |
79 | lurking somewhere on your load-path. The following command will | |
80 | print any duplicate Lisp files that are present in load-path: | |
f0f62f71 | 81 | |
9dc15871 | 82 | emacs -q -batch -f list-load-path-shadows |
f0f62f71 | 83 | |
9dc15871 EZ |
84 | If this command prints any file names, some of these files are stale, |
85 | and should be deleted or their directories removed from your | |
86 | load-path. | |
fc1bfc2a | 87 | |
9dc15871 | 88 | ** Emacs prints an error at startup after upgrading from an earlier version. |
fc1bfc2a | 89 | |
9dc15871 | 90 | An example of such an error is: |
fc1bfc2a | 91 | |
9dc15871 | 92 | x-complement-fontset-spec: "Wrong type argument: stringp, nil" |
60f553d2 | 93 | |
9dc15871 EZ |
94 | This can be another symptom of stale *.elc files in your load-path. |
95 | The following command will print any duplicate Lisp files that are | |
96 | present in load-path: | |
3f82efb4 | 97 | |
9dc15871 | 98 | emacs -q -batch -f list-load-path-shadows |
3f82efb4 | 99 | |
9dc15871 EZ |
100 | If this command prints any file names, some of these files are stale, |
101 | and should be deleted or their directories removed from your | |
102 | load-path. | |
3f82efb4 | 103 | |
9dc15871 | 104 | ** With X11R6.4, public-patch-3, Emacs crashes at startup. |
3f82efb4 | 105 | |
9dc15871 | 106 | Reportedly this patch in X fixes the problem. |
cc305a60 | 107 | |
9dc15871 EZ |
108 | --- xc/lib/X11/imInt.c~ Wed Jun 30 13:31:56 1999 |
109 | +++ xc/lib/X11/imInt.c Thu Jul 1 15:10:27 1999 | |
110 | @@ -1,4 +1,4 @@ | |
111 | -/* $TOG: imInt.c /main/5 1998/05/30 21:11:16 kaleb $ */ | |
112 | +/* $TOG: imInt.c /main/5 1998/05/30 21:11:16 kaleb $ */ | |
113 | /****************************************************************** | |
cc305a60 | 114 | |
9dc15871 EZ |
115 | Copyright 1992, 1993, 1994 by FUJITSU LIMITED |
116 | @@ -166,8 +166,8 @@ | |
117 | _XimMakeImName(lcd) | |
118 | XLCd lcd; | |
119 | { | |
120 | - char* begin; | |
121 | - char* end; | |
122 | + char* begin = NULL; | |
123 | + char* end = NULL; | |
124 | char* ret; | |
125 | int i = 0; | |
126 | char* ximmodifier = XIMMODIFIER; | |
127 | @@ -182,7 +182,11 @@ | |
128 | } | |
129 | ret = Xmalloc(end - begin + 2); | |
130 | if (ret != NULL) { | |
131 | - (void)strncpy(ret, begin, end - begin + 1); | |
132 | + if (begin != NULL) { | |
133 | + (void)strncpy(ret, begin, end - begin + 1); | |
134 | + } else { | |
135 | + ret[0] = '\0'; | |
136 | + } | |
137 | ret[end - begin + 1] = '\0'; | |
138 | } | |
139 | return ret; | |
fc2938d1 | 140 | |
9dc15871 | 141 | * Crash bugs |
fc2938d1 | 142 | |
9dc15871 | 143 | ** Emacs crashes in x-popup-dialog. |
177c0ea7 | 144 | |
9dc15871 EZ |
145 | This can happen if the dialog widget cannot find the font it wants to |
146 | use. You can work around the problem by specifying another font with | |
147 | an X resource--for example, `Emacs.dialog*.font: 9x15' (or any font that | |
148 | happens to exist on your X server). | |
fc2938d1 | 149 | |
9dc15871 | 150 | ** Emacs crashes when you use Bibtex mode. |
fc2938d1 | 151 | |
9dc15871 EZ |
152 | This happens if your system puts a small limit on stack size. You can |
153 | prevent the problem by using a suitable shell command (often `ulimit') | |
154 | to raise the stack size limit before you run Emacs. | |
2aa82bcf | 155 | |
9dc15871 EZ |
156 | Patches to raise the stack size limit automatically in `main' |
157 | (src/emacs.c) on various systems would be greatly appreciated. | |
2aa82bcf | 158 | |
9dc15871 EZ |
159 | ** Error message `Symbol's value as variable is void: x', followed by |
160 | a segmentation fault and core dump. | |
c93bdf05 | 161 | |
9dc15871 EZ |
162 | This has been tracked to a bug in tar! People report that tar erroneously |
163 | added a line like this at the beginning of files of Lisp code: | |
4593687f | 164 | |
9dc15871 | 165 | x FILENAME, N bytes, B tape blocks |
9272ccfc | 166 | |
9dc15871 EZ |
167 | If your tar has this problem, install GNU tar--if you can manage to |
168 | untar it :-). | |
9272ccfc | 169 | |
9dc15871 EZ |
170 | ** Crashes when displaying GIF images in Emacs built with version |
171 | libungif-4.1.0 are resolved by using version libungif-4.1.0b1. | |
172 | Configure checks for the correct version, but this problem could occur | |
173 | if a binary built against a shared libungif is run on a system with an | |
174 | older version. | |
9272ccfc | 175 | |
9dc15871 | 176 | ** Emacs aborts inside the function `tparam1'. |
7aa70236 | 177 | |
9dc15871 EZ |
178 | This can happen if Emacs was built without terminfo support, but the |
179 | terminal's capabilities use format that is only supported by terminfo. | |
180 | If your system has ncurses installed, this might happen if your | |
181 | version of ncurses is broken; upgrading to a newer version of ncurses | |
182 | and reconfiguring and rebuilding Emacs should solve this. | |
7aa70236 | 183 | |
9dc15871 EZ |
184 | All modern systems support terminfo, so even if ncurses is not the |
185 | problem, you should look for a way to configure Emacs so that it uses | |
186 | terminfo when built. | |
7c22dc9d | 187 | |
9dc15871 | 188 | ** Emacs crashes when using the Exceed 6.0 X server. |
7c22dc9d | 189 | |
9dc15871 EZ |
190 | If you are using Exceed 6.1, upgrade to a later version. This was |
191 | reported to prevent the crashes. | |
7c22dc9d | 192 | |
9dc15871 | 193 | ** Emacs crashes with SIGSEGV in XtInitializeWidgetClass. |
7c22dc9d | 194 | |
9dc15871 | 195 | It crashes on X, but runs fine when called with option "-nw". |
7c22dc9d | 196 | |
9dc15871 EZ |
197 | This has been observed when Emacs is linked with GNU ld but without passing |
198 | the -z nocombreloc flag. Emacs normally knows to pass the -z nocombreloc | |
199 | flag when needed, so if you come across a situation where the flag is | |
200 | necessary but missing, please report it via M-x report-emacs-bug. | |
677e7496 | 201 | |
9dc15871 EZ |
202 | On platforms such as Solaris, you can also work around this problem by |
203 | configuring your compiler to use the native linker instead of GNU ld. | |
677e7496 | 204 | |
9dc15871 | 205 | * General runtime problems |
677e7496 | 206 | |
9dc15871 | 207 | ** Lisp problems |
677e7496 | 208 | |
9dc15871 | 209 | *** Changes made to .el files do not take effect. |
677e7496 | 210 | |
9dc15871 EZ |
211 | You may have forgotten to recompile them into .elc files. |
212 | Then the old .elc files will be loaded, and your changes | |
213 | will not be seen. To fix this, do M-x byte-recompile-directory | |
214 | and specify the directory that contains the Lisp files. | |
677e7496 | 215 | |
9dc15871 EZ |
216 | Emacs should print a warning when loading a .elc file which is older |
217 | than the corresponding .el file. | |
9ed04369 | 218 | |
9dc15871 | 219 | *** Watch out for .emacs files and EMACSLOADPATH environment vars. |
9ed04369 | 220 | |
9dc15871 EZ |
221 | These control the actions of Emacs. |
222 | ~/.emacs is your Emacs init file. | |
223 | EMACSLOADPATH overrides which directories the function | |
224 | "load" will search. | |
b87207a0 | 225 | |
9dc15871 EZ |
226 | If you observe strange problems, check for these and get rid |
227 | of them, then try again. | |
b87207a0 | 228 | |
9dc15871 | 229 | *** Using epop3.el package causes Emacs to signal an error. |
177c0ea7 | 230 | |
9dc15871 | 231 | The error message might be something like this: |
177c0ea7 | 232 | |
9dc15871 | 233 | "Lisp nesting exceeds max-lisp-eval-depth" |
177c0ea7 | 234 | |
9dc15871 EZ |
235 | This happens because epop3 redefines the function gethash, which is a |
236 | built-in primitive beginning with Emacs 21.1. We don't have a patch | |
237 | for epop3 that fixes this, but perhaps a newer version of epop3 | |
238 | corrects that. | |
177c0ea7 | 239 | |
9dc15871 | 240 | *** Buffers from `with-output-to-temp-buffer' get set up in Help mode. |
b87207a0 | 241 | |
9dc15871 EZ |
242 | Changes in Emacs 20.4 to the hooks used by that function cause |
243 | problems for some packages, specifically BBDB. See the function's | |
244 | documentation for the hooks involved. BBDB 2.00.06 fixes the problem. | |
61638355 | 245 | |
9dc15871 EZ |
246 | *** The Hyperbole package causes *Help* buffers not to be displayed in |
247 | Help mode due to setting `temp-buffer-show-hook' rather than using | |
248 | `add-hook'. Using `(add-hook 'temp-buffer-show-hook | |
249 | 'help-mode-maybe)' after loading Hyperbole should fix this. | |
61638355 | 250 | |
9dc15871 | 251 | ** Keyboard problems |
a47a639f | 252 | |
9dc15871 | 253 | *** "Compose Character" key does strange things when used as a Meta key. |
a47a639f | 254 | |
9dc15871 EZ |
255 | If you define one key to serve as both Meta and Compose Character, you |
256 | will get strange results. In previous Emacs versions, this "worked" | |
257 | in that the key acted as Meta--that's because the older Emacs versions | |
258 | did not try to support Compose Character. Now Emacs tries to do | |
259 | character composition in the standard X way. This means that you | |
260 | must pick one meaning or the other for any given key. | |
a47a639f | 261 | |
9dc15871 EZ |
262 | You can use both functions (Meta, and Compose Character) if you assign |
263 | them to two different keys. | |
a47a639f | 264 | |
9dc15871 | 265 | *** C-z just refreshes the screen instead of suspending Emacs. |
61638355 | 266 | |
9dc15871 EZ |
267 | You are probably using a shell that doesn't support job control, even |
268 | though the system itself is capable of it. Either use a different shell, | |
269 | or set the variable `cannot-suspend' to a non-nil value. | |
61638355 | 270 | |
9dc15871 EZ |
271 | *** With M-x enable-flow-control, you need to type C-\ twice |
272 | to do incremental search--a single C-\ gets no response. | |
61638355 | 273 | |
9dc15871 EZ |
274 | This has been traced to communicating with your machine via kermit, |
275 | with C-\ as the kermit escape character. One solution is to use | |
276 | another escape character in kermit. One user did | |
61638355 | 277 | |
9dc15871 | 278 | set escape-character 17 |
61638355 | 279 | |
9dc15871 | 280 | in his .kermrc file, to make C-q the kermit escape character. |
61638355 | 281 | |
9dc15871 | 282 | ** Mailers and other helper programs |
61638355 | 283 | |
9dc15871 | 284 | *** movemail compiled with POP support can't connect to the POP server. |
61638355 | 285 | |
9dc15871 EZ |
286 | Make sure that the `pop' entry in /etc/services, or in the services |
287 | NIS map if your machine uses NIS, has the same port number as the | |
288 | entry on the POP server. A common error is for the POP server to be | |
289 | listening on port 110, the assigned port for the POP3 protocol, while | |
290 | the client is trying to connect on port 109, the assigned port for the | |
291 | old POP protocol. | |
61638355 | 292 | |
9dc15871 | 293 | *** RMAIL gets error getting new mail. |
61638355 | 294 | |
9dc15871 EZ |
295 | RMAIL gets new mail from /usr/spool/mail/$USER using a program |
296 | called `movemail'. This program interlocks with /bin/mail using | |
297 | the protocol defined by /bin/mail. | |
61638355 | 298 | |
9dc15871 EZ |
299 | There are two different protocols in general use. One of them uses |
300 | the `flock' system call. The other involves creating a lock file; | |
301 | `movemail' must be able to write in /usr/spool/mail in order to do | |
302 | this. You control which one is used by defining, or not defining, | |
303 | the macro MAIL_USE_FLOCK in config.h or the m- or s- file it includes. | |
304 | IF YOU DON'T USE THE FORM OF INTERLOCKING THAT IS NORMAL ON YOUR | |
305 | SYSTEM, YOU CAN LOSE MAIL! | |
61638355 | 306 | |
9dc15871 EZ |
307 | If your system uses the lock file protocol, and fascist restrictions |
308 | prevent ordinary users from writing the lock files in /usr/spool/mail, | |
309 | you may need to make `movemail' setgid to a suitable group such as | |
310 | `mail'. You can use these commands (as root): | |
61638355 | 311 | |
9dc15871 EZ |
312 | chgrp mail movemail |
313 | chmod 2755 movemail | |
61638355 | 314 | |
9dc15871 EZ |
315 | If your system uses the lock file protocol, and fascist restrictions |
316 | prevent ordinary users from writing the lock files in /usr/spool/mail, | |
317 | you may need to make `movemail' setgid to a suitable group such as | |
318 | `mail'. To do this, use the following commands (as root) after doing the | |
319 | make install. | |
61638355 | 320 | |
9dc15871 EZ |
321 | chgrp mail movemail |
322 | chmod 2755 movemail | |
61638355 | 323 | |
9dc15871 EZ |
324 | Installation normally copies movemail from the build directory to an |
325 | installation directory which is usually under /usr/local/lib. The | |
326 | installed copy of movemail is usually in the directory | |
327 | /usr/local/lib/emacs/VERSION/TARGET. You must change the group and | |
328 | mode of the installed copy; changing the group and mode of the build | |
329 | directory copy is ineffective. | |
61638355 | 330 | |
9dc15871 | 331 | *** rcs2log gives you the awk error message "too many fields". |
61638355 | 332 | |
9dc15871 EZ |
333 | This is due to an arbitrary limit in certain versions of awk. |
334 | The solution is to use gawk (GNU awk). | |
556a9fad | 335 | |
9dc15871 | 336 | ** Problems with hostname resolution |
61638355 | 337 | |
9dc15871 EZ |
338 | *** Emacs fails to understand most Internet host names, even though |
339 | the names work properly with other programs on the same system. | |
340 | *** Emacs won't work with X-windows if the value of DISPLAY is HOSTNAME:0. | |
341 | *** GNUs can't make contact with the specified host for nntp. | |
49172314 | 342 | |
9dc15871 EZ |
343 | This typically happens on Suns and other systems that use shared |
344 | libraries. The cause is that the site has installed a version of the | |
345 | shared library which uses a name server--but has not installed a | |
346 | similar version of the unshared library which Emacs uses. | |
96bde66b | 347 | |
9dc15871 EZ |
348 | The result is that most programs, using the shared library, work with |
349 | the nameserver, but Emacs does not. | |
ed0d1d91 | 350 | |
9dc15871 EZ |
351 | The fix is to install an unshared library that corresponds to what you |
352 | installed in the shared library, and then relink Emacs. | |
6e1a66dc | 353 | |
9dc15871 EZ |
354 | On SunOS 4.1, simply define HAVE_RES_INIT. |
355 | ||
356 | If you have already installed the name resolver in the file libresolv.a, | |
357 | then you need to compile Emacs to use that library. The easiest way to | |
358 | do this is to add to config.h a definition of LIBS_SYSTEM, LIBS_MACHINE | |
359 | or LIB_STANDARD which uses -lresolv. Watch out! If you redefine a macro | |
360 | that is already in use in your configuration to supply some other libraries, | |
361 | be careful not to lose the others. | |
6e1a66dc | 362 | |
9dc15871 | 363 | Thus, you could start by adding this to config.h: |
3c418e54 | 364 | |
9dc15871 | 365 | #define LIBS_SYSTEM -lresolv |
3c418e54 | 366 | |
9dc15871 EZ |
367 | Then if this gives you an error for redefining a macro, and you see that |
368 | the s- file defines LIBS_SYSTEM as -lfoo -lbar, you could change config.h | |
369 | again to say this: | |
3c418e54 | 370 | |
9dc15871 | 371 | #define LIBS_SYSTEM -lresolv -lfoo -lbar |
f9130829 | 372 | |
9dc15871 EZ |
373 | *** Emacs does not know your host's fully-qualified domain name. |
374 | ||
375 | You need to configure your machine with a fully qualified domain name, | |
376 | either in /etc/hosts, /etc/hostname, the NIS, or wherever your system | |
377 | calls for specifying this. | |
378 | ||
379 | If you cannot fix the configuration, you can set the Lisp variable | |
380 | mail-host-address to the value you want. | |
381 | ||
382 | ** NFS and RFS | |
383 | ||
384 | *** Emacs says it has saved a file, but the file does not actually | |
385 | appear on disk. | |
386 | ||
387 | This can happen on certain systems when you are using NFS, if the | |
388 | remote disk is full. It is due to a bug in NFS (or certain NFS | |
389 | implementations), and there is apparently nothing Emacs can do to | |
390 | detect the problem. Emacs checks the failure codes of all the system | |
391 | calls involved in writing a file, including `close'; but in the case | |
392 | where the problem occurs, none of those system calls fails. | |
393 | ||
394 | *** Editing files through RFS gives spurious "file has changed" warnings. | |
395 | It is possible that a change in Emacs 18.37 gets around this problem, | |
396 | but in case not, here is a description of how to fix the RFS bug that | |
397 | causes it. | |
398 | ||
399 | There was a serious pair of bugs in the handling of the fsync() system | |
400 | call in the RFS server. | |
401 | ||
402 | The first is that the fsync() call is handled as another name for the | |
403 | close() system call (!!). It appears that fsync() is not used by very | |
404 | many programs; Emacs version 18 does an fsync() before closing files | |
405 | to make sure that the bits are on the disk. | |
406 | ||
407 | This is fixed by the enclosed patch to the RFS server. | |
408 | ||
409 | The second, more serious problem, is that fsync() is treated as a | |
410 | non-blocking system call (i.e., it's implemented as a message that | |
411 | gets sent to the remote system without waiting for a reply). Fsync is | |
412 | a useful tool for building atomic file transactions. Implementing it | |
413 | as a non-blocking RPC call (when the local call blocks until the sync | |
414 | is done) is a bad idea; unfortunately, changing it will break the RFS | |
415 | protocol. No fix was supplied for this problem. | |
416 | ||
417 | (as always, your line numbers may vary) | |
418 | ||
419 | % rcsdiff -c -r1.2 serversyscall.c | |
420 | RCS file: RCS/serversyscall.c,v | |
421 | retrieving revision 1.2 | |
422 | diff -c -r1.2 serversyscall.c | |
423 | *** /tmp/,RCSt1003677 Wed Jan 28 15:15:02 1987 | |
424 | --- serversyscall.c Wed Jan 28 15:14:48 1987 | |
425 | *************** | |
426 | *** 163,169 **** | |
427 | /* | |
428 | * No return sent for close or fsync! | |
429 | */ | |
430 | ! if (syscall == RSYS_close || syscall == RSYS_fsync) | |
431 | proc->p_returnval = deallocate_fd(proc, msg->m_args[0]); | |
432 | else | |
433 | { | |
434 | --- 166,172 ---- | |
435 | /* | |
436 | * No return sent for close or fsync! | |
437 | */ | |
438 | ! if (syscall == RSYS_close) | |
439 | proc->p_returnval = deallocate_fd(proc, msg->m_args[0]); | |
440 | else | |
441 | { | |
442 | ||
443 | ** PSGML | |
444 | ||
445 | *** Old versions of the PSGML package use the obsolete variables | |
446 | `before-change-function' and `after-change-function', which are no | |
447 | longer used by Emacs. Please use PSGML 1.2.3 or later. | |
448 | ||
449 | *** PSGML conflicts with sgml-mode. | |
450 | ||
451 | PSGML package uses the same names of some variables (like keymap) | |
452 | as built-in sgml-mode.el because it was created as a replacement | |
453 | of that package. The conflict will be shown if you load | |
454 | sgml-mode.el before psgml.el. E.g. this could happen if you edit | |
455 | HTML page and then start to work with SGML or XML file. html-mode | |
456 | (from sgml-mode.el) is used for HTML file and loading of psgml.el | |
457 | (for sgml-mode or xml-mode) will cause an error. | |
458 | ||
459 | *** Versions of the PSGML package earlier than 1.0.3 (stable) or 1.1.2 | |
460 | (alpha) fail to parse DTD files correctly in Emacs 20.3 and later. | |
461 | Here is a patch for psgml-parse.el from PSGML 1.0.1 and, probably, | |
462 | earlier versions. | |
463 | ||
464 | --- psgml-parse.el 1998/08/21 19:18:18 1.1 | |
465 | +++ psgml-parse.el 1998/08/21 19:20:00 | |
466 | @@ -2383,7 +2383,7 @@ (defun sgml-push-to-entity (entity &opti | |
467 | (setq sgml-buffer-parse-state nil)) | |
468 | (cond | |
469 | ((stringp entity) ; a file name | |
470 | - (save-excursion (insert-file-contents entity)) | |
471 | + (insert-file-contents entity) | |
472 | (setq default-directory (file-name-directory entity))) | |
473 | ((consp (sgml-entity-text entity)) ; external id? | |
474 | (let* ((extid (sgml-entity-text entity)) | |
475 | ||
90a02640 | 476 | ** AUCTeX |
9dc15871 | 477 | |
90a02640 DK |
478 | You should not be using a version older than 11.52 if you can avoid |
479 | it. | |
9dc15871 | 480 | |
90a02640 DK |
481 | *** Emacs 21 freezes when visiting a TeX file with AUCTeX installed. |
482 | ||
483 | Emacs 21 needs version 10 or later of AUCTeX; upgrading should solve | |
9dc15871 EZ |
484 | these problems. |
485 | ||
90a02640 | 486 | *** No colors in AUCTeX with Emacs 21. |
9dc15871 EZ |
487 | |
488 | Upgrade to AUC TeX version 10 or later, and make sure it is | |
489 | byte-compiled with Emacs 21. | |
490 | ||
9dc15871 EZ |
491 | ** Miscellaneous problems |
492 | ||
493 | *** Self-documentation messages are garbled. | |
494 | ||
495 | This means that the file `etc/DOC-...' doesn't properly correspond | |
496 | with the Emacs executable. Redumping Emacs and then installing the | |
497 | corresponding pair of files should fix the problem. | |
498 | ||
499 | *** Programs running under terminal emulator do not recognize `emacs' | |
500 | terminal type. | |
501 | ||
502 | The cause of this is a shell startup file that sets the TERMCAP | |
503 | environment variable. The terminal emulator uses that variable to | |
504 | provide the information on the special terminal type that Emacs | |
505 | emulates. | |
506 | ||
507 | Rewrite your shell startup file so that it does not change TERMCAP | |
508 | in such a case. You could use the following conditional which sets | |
509 | it only if it is undefined. | |
510 | ||
511 | if ( ! ${?TERMCAP} ) setenv TERMCAP ~/my-termcap-file | |
512 | ||
513 | Or you could set TERMCAP only when you set TERM--which should not | |
514 | happen in a non-login shell. | |
515 | ||
516 | *** In Shell mode, you get a ^M at the end of every line. | |
517 | ||
518 | This happens to people who use tcsh, because it is trying to be too | |
519 | smart. It sees that the Shell uses terminal type `unknown' and turns | |
520 | on the flag to output ^M at the end of each line. You can fix the | |
521 | problem by adding this to your .cshrc file: | |
522 | ||
523 | if ($?EMACS) then | |
524 | if ($EMACS == "t") then | |
525 | unset edit | |
526 | stty -icrnl -onlcr -echo susp ^Z | |
527 | endif | |
528 | endif | |
529 | ||
530 | *** Emacs startup on GNU/Linux systems (and possibly other systems) is slow. | |
531 | ||
532 | This can happen if the system is misconfigured and Emacs can't get the | |
533 | full qualified domain name, FQDN. You should have your FQDN in the | |
534 | /etc/hosts file, something like this: | |
535 | ||
536 | 127.0.0.1 localhost | |
537 | 129.187.137.82 nuc04.t30.physik.tu-muenchen.de nuc04 | |
538 | ||
539 | The way to set this up may vary on non-GNU systems. | |
540 | ||
541 | *** Attempting to visit remote files via ange-ftp fails. | |
542 | ||
543 | If the error message is "ange-ftp-file-modtime: Specified time is not | |
544 | representable", then this could happen when `lukemftp' is used as the | |
545 | ftp client. This was reported to happen on Debian GNU/Linux, kernel | |
546 | version 2.4.3, with `lukemftp' 1.5-5, but might happen on other | |
547 | systems as well. To avoid this problem, switch to using the standard | |
548 | ftp client. On a Debian system, type | |
549 | ||
550 | update-alternatives --config ftp | |
551 | ||
552 | and then choose /usr/bin/netkit-ftp. | |
553 | ||
554 | *** JPEG images aren't displayed. | |
555 | ||
556 | This has been reported when Emacs is built with jpeg-6a library. | |
557 | Upgrading to jpeg-6b solves the problem. Configure checks for the | |
558 | correct version, but this problem could occur if a binary built | |
559 | against a shared libjpeg is run on a system with an older version. | |
560 | ||
561 | *** Dired is very slow. | |
562 | ||
563 | This could happen if invocation of the `df' program takes a long | |
564 | time. Possible reasons for this include: | |
565 | ||
566 | - ClearCase mounted filesystems (VOBs) that sometimes make `df' | |
567 | response time extremely slow (dozens of seconds); | |
568 | ||
569 | - slow automounters on some old versions of Unix; | |
570 | ||
571 | - slow operation of some versions of `df'. | |
572 | ||
573 | To work around the problem, you could either (a) set the variable | |
574 | `directory-free-space-program' to nil, and thus prevent Emacs from | |
575 | invoking `df'; (b) use `df' from the GNU Fileutils package; or | |
576 | (c) use CVS, which is Free Software, instead of ClearCase. | |
577 | ||
578 | *** Versions of the W3 package released before Emacs 21.1 don't run | |
579 | under Emacs 21. This fixed in W3 version 4.0pre.47. | |
580 | ||
581 | *** The LDAP support rely on ldapsearch program from OpenLDAP version 2. | |
582 | ||
583 | It can fail to work with ldapsearch program from OpenLDAP version 1. | |
584 | Version 1 of OpenLDAP is now deprecated. If you are still using it, | |
585 | please upgrade to version 2. As a temporary workaround, remove | |
586 | argument "-x" from the variable `ldap-ldapsearch-args'. | |
587 | ||
588 | *** ps-print commands fail to find prologue files ps-prin*.ps. | |
589 | ||
590 | This can happen if you use an old version of X-Symbol package: it | |
591 | defines compatibility functions which trick ps-print into thinking it | |
592 | runs in XEmacs, and look for the prologue files in a wrong directory. | |
593 | ||
594 | The solution is to upgrade X-Symbol to a later version. | |
595 | ||
596 | *** On systems with shared libraries you might encounter run-time errors | |
597 | from the dynamic linker telling you that it is unable to find some | |
598 | shared libraries, for instance those for Xaw3d or image support. | |
599 | These errors mean Emacs has been linked with a library whose shared | |
600 | library is not in the default search path of the dynamic linker. | |
601 | ||
602 | Similar problems could prevent Emacs from building, since the build | |
603 | process invokes Emacs several times. | |
604 | ||
605 | On many systems, it is possible to set LD_LIBRARY_PATH in your | |
606 | environment to specify additional directories where shared libraries | |
607 | can be found. | |
608 | ||
609 | Other systems allow to set LD_RUN_PATH in a similar way, but before | |
610 | Emacs is linked. With LD_RUN_PATH set, the linker will include a | |
611 | specified run-time search path in the executable. | |
612 | ||
613 | On some systems, Emacs can crash due to problems with dynamic | |
614 | linking. Specifically, on SGI Irix 6.5, crashes were reported with | |
615 | backtraces like this: | |
616 | ||
617 | (dbx) where | |
618 | 0 strcmp(0xf49239d, 0x4031184, 0x40302b4, 0x12, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2) ["/xlv22/ficus-jan23/work/irix/lib/libc/libc_n32_M3_ns/strings/strcmp.s":35, 0xfb7e480] | |
619 | 1 general_find_symbol(0xf49239d, 0x0, 0x0, 0x0, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2) | |
620 | ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":2140, 0xfb65a98] | |
621 | 2 resolve_symbol(0xf49239d, 0x4031184, 0x0, 0xfbdd438, 0x0, 0xf4923aa, 0x0, 0x492ddb2) | |
622 | ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":1947, 0xfb657e4] | |
623 | 3 lazy_text_resolve(0xd18, 0x1a3, 0x40302b4, 0x12, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2) | |
624 | ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":997, 0xfb64d44] | |
625 | 4 _rld_text_resolve(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0) | |
626 | ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld_bridge.s":175, 0xfb6032c] | |
627 | ||
628 | (`rld' is the dynamic linker.) We don't know yet why this | |
629 | happens, but setting the environment variable LD_BIND_NOW to 1 (which | |
630 | forces the dynamic linker to bind all shared objects early on) seems | |
631 | to work around the problem. | |
632 | ||
633 | Please refer to the documentation of your dynamic linker for details. | |
634 | ||
635 | *** You request inverse video, and the first Emacs frame is in inverse | |
636 | video, but later frames are not in inverse video. | |
637 | ||
638 | This can happen if you have an old version of the custom library in | |
639 | your search path for Lisp packages. Use M-x list-load-path-shadows to | |
640 | check whether this is true. If it is, delete the old custom library. | |
641 | ||
642 | *** When you run Ispell from Emacs, it reports a "misalignment" error. | |
643 | ||
644 | This can happen if you compiled the Ispell program to use ASCII | |
645 | characters only and then try to use it from Emacs with non-ASCII | |
646 | characters, like Latin-1. The solution is to recompile Ispell with | |
647 | support for 8-bit characters. | |
648 | ||
649 | To see whether your Ispell program supports 8-bit characters, type | |
650 | this at your shell's prompt: | |
651 | ||
652 | ispell -vv | |
653 | ||
654 | and look in the output for the string "NO8BIT". If Ispell says | |
655 | "!NO8BIT (8BIT)", your speller supports 8-bit characters; otherwise it | |
656 | does not. | |
657 | ||
658 | To rebuild Ispell with 8-bit character support, edit the local.h file | |
659 | in the Ispell distribution and make sure it does _not_ define NO8BIT. | |
660 | Then rebuild the speller. | |
661 | ||
662 | Another possible cause for "misalignment" error messages is that the | |
663 | version of Ispell installed on your machine is old. Upgrade. | |
664 | ||
665 | Yet another possibility is that you are trying to spell-check a word | |
666 | in a language that doesn't fit the dictionary you choose for use by | |
667 | Ispell. (Ispell can only spell-check one language at a time, because | |
668 | it uses a single dictionary.) Make sure that the text you are | |
669 | spelling and the dictionary used by Ispell conform to each other. | |
670 | ||
671 | If your spell-checking program is Aspell, it has been reported that if | |
672 | you have a personal configuration file (normally ~/.aspell.conf), it | |
673 | can cause this error. Remove that file, execute `ispell-kill-ispell' | |
674 | in Emacs, and then try spell-checking again. | |
675 | ||
676 | * Runtime problems related to font handling | |
677 | ||
678 | ** Under X11, some characters appear as hollow boxes. | |
679 | ||
680 | Each X11 font covers just a fraction of the characters that Emacs | |
681 | supports. To display the whole range of Emacs characters requires | |
682 | many different fonts, collected into a fontset. | |
683 | ||
684 | If some of the fonts called for in your fontset do not exist on your X | |
685 | server, then the characters that have no font appear as hollow boxes. | |
686 | You can remedy the problem by installing additional fonts. | |
f9130829 | 687 | |
9dc15871 EZ |
688 | The intlfonts distribution includes a full spectrum of fonts that can |
689 | display all the characters Emacs supports. | |
f9130829 | 690 | |
9dc15871 EZ |
691 | Another cause of this for specific characters is fonts which have a |
692 | missing glyph and no default character. This is known to occur for | |
693 | character number 160 (no-break space) in some fonts, such as Lucida | |
694 | but Emacs sets the display table for the unibyte and Latin-1 version | |
695 | of this character to display a space. | |
f9130829 | 696 | |
9dc15871 | 697 | ** Under X11, some characters appear improperly aligned in their lines. |
f9130829 | 698 | |
9dc15871 | 699 | You may have bad X11 fonts; try installing the intlfonts distribution. |
f9130829 | 700 | |
9dc15871 | 701 | ** Certain fonts make each line take one pixel more than it "should". |
f9130829 | 702 | |
9dc15871 EZ |
703 | This is because these fonts contain characters a little taller |
704 | than the font's nominal height. Emacs needs to make sure that | |
705 | lines do not overlap. | |
ed85f61d | 706 | |
9dc15871 | 707 | ** Loading fonts is very slow. |
b300fd77 | 708 | |
9dc15871 EZ |
709 | You might be getting scalable fonts instead of precomputed bitmaps. |
710 | Known scalable font directories are "Type1" and "Speedo". A font | |
711 | directory contains scalable fonts if it contains the file | |
712 | "fonts.scale". | |
b300fd77 | 713 | |
9dc15871 EZ |
714 | If this is so, re-order your X windows font path to put the scalable |
715 | font directories last. See the documentation of `xset' for details. | |
ed85f61d | 716 | |
9dc15871 EZ |
717 | With some X servers, it may be necessary to take the scalable font |
718 | directories out of your path entirely, at least for Emacs 19.26. | |
719 | Changes in the future may make this unnecessary. | |
ed85f61d | 720 | |
9dc15871 | 721 | ** Font Lock displays portions of the buffer in incorrect faces. |
fa2301bf | 722 | |
9dc15871 EZ |
723 | By far the most frequent cause of this is a parenthesis `(' or a brace |
724 | `{' in column zero. Font Lock assumes that such a paren is outside of | |
725 | any comment or string. This is of course not true in general, but the | |
726 | vast majority of well-formatted program source files don't have such | |
727 | parens, and therefore this assumption is used to allow optimizations | |
728 | in Font Lock's syntactical analysis. These optimizations avoid some | |
729 | pathological cases where jit-lock, the Just-in-Time fontification | |
730 | introduced with Emacs 21.1, could significantly slow down scrolling | |
731 | through the buffer, especially scrolling backwards, and also jumping | |
732 | to the end of a very large buffer. | |
ed85f61d | 733 | |
bf247b6e | 734 | Beginning with version 22.1, a parenthesis or a brace in column zero |
9dc15871 EZ |
735 | is highlighted in bold-red face if it is inside a string or a comment, |
736 | to indicate that it could interfere with Font Lock (and also with | |
737 | indentation) and should be moved or escaped with a backslash. | |
d0cf6c7d | 738 | |
9dc15871 EZ |
739 | If you don't use large buffers, or have a very fast machine which |
740 | makes the delays insignificant, you can avoid the incorrect | |
741 | fontification by setting the variable | |
742 | `font-lock-beginning-of-syntax-function' to a nil value. (This must | |
743 | be done _after_ turning on Font Lock.) | |
d0cf6c7d | 744 | |
9dc15871 EZ |
745 | Another alternative is to avoid a paren in column zero. For example, |
746 | in a Lisp string you could precede the paren with a backslash. | |
d0cf6c7d | 747 | |
9dc15871 EZ |
748 | ** With certain fonts, when the cursor appears on a character, the |
749 | character doesn't appear--you get a solid box instead. | |
d0cf6c7d | 750 | |
9dc15871 EZ |
751 | One user on a Linux-based GNU system reported that this problem went |
752 | away with installation of a new X server. The failing server was | |
753 | XFree86 3.1.1. XFree86 3.1.2 works. | |
c289e7f0 | 754 | |
9dc15871 | 755 | ** Characters are displayed as empty boxes or with wrong font under X. |
339b95d5 | 756 | |
9dc15871 EZ |
757 | This can occur when two different versions of FontConfig are used. |
758 | For example, XFree86 4.3.0 has one version and Gnome usually comes | |
759 | with a newer version. Emacs compiled with --with-gtk will then use | |
760 | the newer version. In most cases the problem can be temporarily | |
761 | fixed by stopping the application that has the error (it can be | |
762 | Emacs or any other application), removing ~/.fonts.cache-1, | |
763 | and then start the application again. | |
764 | If removing ~/.fonts.cache-1 and restarting doesn't help, the | |
765 | application with problem must be recompiled with the same version | |
766 | of FontConfig as the rest of the system uses. For KDE, it is | |
767 | sufficient to recompile Qt. | |
339b95d5 | 768 | |
9dc15871 | 769 | ** Emacs pauses for several seconds when changing the default font. |
339b95d5 | 770 | |
9dc15871 EZ |
771 | This has been reported for fvwm 2.2.5 and the window manager of KDE |
772 | 2.1. The reason for the pause is Xt waiting for a ConfigureNotify | |
773 | event from the window manager, which the window manager doesn't send. | |
774 | Xt stops waiting after a default timeout of usually 5 seconds. | |
339b95d5 | 775 | |
9dc15871 | 776 | A workaround for this is to add something like |
b87207a0 | 777 | |
9dc15871 | 778 | emacs.waitForWM: false |
f936978f | 779 | |
9dc15871 EZ |
780 | to your X resources. Alternatively, add `(wait-for-wm . nil)' to a |
781 | frame's parameter list, like this: | |
6fb6f3ac | 782 | |
9dc15871 | 783 | (modify-frame-parameters nil '((wait-for-wm . nil))) |
b87207a0 | 784 | |
9dc15871 | 785 | (this should go into your `.emacs' file). |
b87207a0 | 786 | |
9dc15871 | 787 | ** Underlines appear at the wrong position. |
b87207a0 | 788 | |
9dc15871 EZ |
789 | This is caused by fonts having a wrong UNDERLINE_POSITION property. |
790 | Examples are the font 7x13 on XFree prior to version 4.1, or the jmk | |
791 | neep font from the Debian xfonts-jmk package. To circumvent this | |
792 | problem, set x-use-underline-position-properties to nil in your | |
793 | `.emacs'. | |
b87207a0 | 794 | |
9dc15871 EZ |
795 | To see what is the value of UNDERLINE_POSITION defined by the font, |
796 | type `xlsfonts -lll FONT' and look at the font's UNDERLINE_POSITION | |
797 | property. | |
e085efdb | 798 | |
9dc15871 | 799 | ** When using Exceed, fonts sometimes appear too tall. |
fa99e2a4 | 800 | |
9dc15871 EZ |
801 | When the display is set to an Exceed X-server and fonts are specified |
802 | (either explicitly with the -fn option or implicitly with X resources) | |
803 | then the fonts may appear "too tall". The actual character sizes are | |
804 | correct but there is too much vertical spacing between rows, which | |
805 | gives the appearance of "double spacing". | |
c8d9b4ee | 806 | |
9dc15871 EZ |
807 | To prevent this, turn off the Exceed's "automatic font substitution" |
808 | feature (in the font part of the configuration window). | |
fe445893 | 809 | |
9dc15871 | 810 | * Internationalization problems |
c8d9b4ee | 811 | |
9dc15871 | 812 | ** Characters from the mule-unicode charsets aren't displayed under X. |
d9810886 | 813 | |
9dc15871 EZ |
814 | XFree86 4 contains many fonts in iso10646-1 encoding which have |
815 | minimal character repertoires (whereas the encoding part of the font | |
816 | name is meant to be a reasonable indication of the repertoire | |
817 | according to the XLFD spec). Emacs may choose one of these to display | |
818 | characters from the mule-unicode charsets and then typically won't be | |
819 | able to find the glyphs to display many characters. (Check with C-u | |
820 | C-x = .) To avoid this, you may need to use a fontset which sets the | |
821 | font for the mule-unicode sets explicitly. E.g. to use GNU unifont, | |
822 | include in the fontset spec: | |
d9810886 | 823 | |
9dc15871 EZ |
824 | mule-unicode-2500-33ff:-gnu-unifont-*-iso10646-1,\ |
825 | mule-unicode-e000-ffff:-gnu-unifont-*-iso10646-1,\ | |
826 | mule-unicode-0100-24ff:-gnu-unifont-*-iso10646-1 | |
d04b2e49 | 827 | |
9dc15871 | 828 | ** The UTF-8/16/7 coding systems don't encode CJK (Far Eastern) characters. |
d9810886 | 829 | |
9dc15871 EZ |
830 | Emacs by default only supports the parts of the Unicode BMP whose code |
831 | points are in the ranges 0000-33ff and e000-ffff. This excludes: most | |
832 | of CJK, Yi and Hangul, as well as everything outside the BMP. | |
d9810886 | 833 | |
9dc15871 EZ |
834 | If you read UTF-8 data with code points outside these ranges, the |
835 | characters appear in the buffer as raw bytes of the original UTF-8 | |
836 | (composed into a single quasi-character) and they will be written back | |
837 | correctly as UTF-8, assuming you don't break the composed sequences. | |
838 | If you read such characters from UTF-16 or UTF-7 data, they are | |
839 | substituted with the Unicode `replacement character', and you lose | |
840 | information. | |
aa99760e | 841 | |
9dc15871 EZ |
842 | To edit such UTF data, turn on Utf-Translate-Cjk mode, which makes |
843 | many common CJK characters available for encoding and decoding and can | |
844 | be extended by updating the tables it uses. This also allows you to | |
845 | save as UTF buffers containing characters decoded by the chinese-, | |
846 | japanese- and korean- coding systems, e.g. cut and pasted from | |
847 | elsewhere. | |
aa99760e | 848 | |
9dc15871 | 849 | ** Mule-UCS loads very slowly. |
aa99760e | 850 | |
9dc15871 EZ |
851 | Changes to Emacs internals interact badly with Mule-UCS's `un-define' |
852 | library, which is the usual interface to Mule-UCS. Apply the | |
853 | following patch to Mule-UCS 0.84 and rebuild it. That will help, | |
854 | though loading will still be slower than in Emacs 20. (Some | |
855 | distributions, such as Debian, may already have applied such a patch.) | |
086b25d3 | 856 | |
9dc15871 EZ |
857 | --- lisp/un-define.el 6 Mar 2001 22:41:38 -0000 1.30 |
858 | +++ lisp/un-define.el 19 Apr 2002 18:34:26 -0000 | |
859 | @@ -610,13 +624,21 @@ by calling post-read-conversion and pre- | |
086b25d3 | 860 | |
9dc15871 EZ |
861 | (mapcar |
862 | (lambda (x) | |
863 | - (mapcar | |
864 | - (lambda (y) | |
865 | - (mucs-define-coding-system | |
866 | - (nth 0 y) (nth 1 y) (nth 2 y) | |
867 | - (nth 3 y) (nth 4 y) (nth 5 y) (nth 6 y)) | |
868 | - (coding-system-put (car y) 'alias-coding-systems (list (car x)))) | |
869 | - (cdr x))) | |
870 | + (if (fboundp 'register-char-codings) | |
871 | + ;; Mule 5, where we don't need the eol-type specified and | |
872 | + ;; register-char-codings may be very slow for these coding | |
873 | + ;; system definitions. | |
874 | + (let ((y (cadr x))) | |
875 | + (mucs-define-coding-system | |
876 | + (car x) (nth 1 y) (nth 2 y) | |
877 | + (nth 3 y) (nth 4 y) (nth 5 y))) | |
878 | + (mapcar | |
879 | + (lambda (y) | |
880 | + (mucs-define-coding-system | |
881 | + (nth 0 y) (nth 1 y) (nth 2 y) | |
882 | + (nth 3 y) (nth 4 y) (nth 5 y) (nth 6 y)) | |
883 | + (coding-system-put (car y) 'alias-coding-systems (list (car x))))) | |
884 | + (cdr x))) | |
885 | `((utf-8 | |
886 | (utf-8-unix | |
887 | ?u "UTF-8 coding system" | |
086b25d3 | 888 | |
9dc15871 EZ |
889 | Note that Emacs has native support for Unicode, roughly equivalent to |
890 | Mule-UCS's, so you may not need it. | |
086b25d3 | 891 | |
9dc15871 | 892 | ** Accented ISO-8859-1 characters are displayed as | or _. |
d6b7de9b EZ |
893 | |
894 | Try other font set sizes (S-mouse-1). If the problem persists with | |
895 | other sizes as well, your text is corrupted, probably through software | |
896 | that is not 8-bit clean. If the problem goes away with another font | |
897 | size, it's probably because some fonts pretend to be ISO-8859-1 fonts | |
898 | when they are really ASCII fonts. In particular the schumacher-clean | |
899 | fonts have this bug in some versions of X. | |
900 | ||
901 | To see what glyphs are included in a font, use `xfd', like this: | |
902 | ||
903 | xfd -fn -schumacher-clean-medium-r-normal--12-120-75-75-c-60-iso8859-1 | |
904 | ||
905 | If this shows only ASCII glyphs, the font is indeed the source of the | |
906 | problem. | |
907 | ||
908 | The solution is to remove the corresponding lines from the appropriate | |
909 | `fonts.alias' file, then run `mkfontdir' in that directory, and then run | |
910 | `xset fp rehash'. | |
911 | ||
9dc15871 | 912 | ** The `oc-unicode' package doesn't work with Emacs 21. |
b87207a0 | 913 | |
9dc15871 EZ |
914 | This package tries to define more private charsets than there are free |
915 | slots now. The current built-in Unicode support is actually more | |
916 | flexible. (Use option `utf-translate-cjk-mode' if you need CJK | |
917 | support.) Files encoded as emacs-mule using oc-unicode aren't | |
918 | generally read correctly by Emacs 21. | |
b87207a0 | 919 | |
9dc15871 | 920 | ** After a while, Emacs slips into unibyte mode. |
4e0bd469 | 921 | |
9dc15871 EZ |
922 | The VM mail package, which is not part of Emacs, sometimes does |
923 | (standard-display-european t) | |
924 | That should be changed to | |
925 | (standard-display-european 1 t) | |
4e0bd469 | 926 | |
9dc15871 | 927 | * X runtime problems |
4e0bd469 | 928 | |
9dc15871 | 929 | ** X keyboard problems |
4e0bd469 | 930 | |
9dc15871 | 931 | *** You "lose characters" after typing Compose Character key. |
4e0bd469 | 932 | |
9dc15871 EZ |
933 | This is because the Compose Character key is defined as the keysym |
934 | Multi_key, and Emacs (seeing that) does the proper X11 | |
935 | character-composition processing. If you don't want your Compose key | |
936 | to do that, you can redefine it with xmodmap. | |
4e0bd469 | 937 | |
9dc15871 | 938 | For example, here's one way to turn it into a Meta key: |
61638355 | 939 | |
9dc15871 | 940 | xmodmap -e "keysym Multi_key = Meta_L" |
61638355 | 941 | |
9dc15871 EZ |
942 | If all users at your site of a particular keyboard prefer Meta to |
943 | Compose, you can make the remapping happen automatically by adding the | |
944 | xmodmap command to the xdm setup script for that display. | |
61638355 | 945 | |
9dc15871 | 946 | *** Using X Windows, control-shift-leftbutton makes Emacs hang. |
61638355 | 947 | |
9dc15871 | 948 | Use the shell command `xset bc' to make the old X Menu package work. |
61638355 | 949 | |
9dc15871 | 950 | *** M-SPC seems to be ignored as input. |
61638355 | 951 | |
9dc15871 EZ |
952 | See if your X server is set up to use this as a command |
953 | for character composition. | |
a953a8d3 | 954 | |
9dc15871 | 955 | *** The S-C-t key combination doesn't get passed to Emacs on X. |
2ebf6139 | 956 | |
9dc15871 EZ |
957 | This happens because some X configurations assign the Ctrl-Shift-t |
958 | combination the same meaning as the Multi_key. The offending | |
959 | definition is in the file `...lib/X11/locale/iso8859-1/Compose'; there | |
960 | might be other similar combinations which are grabbed by X for similar | |
961 | purposes. | |
ec383c7d | 962 | |
9dc15871 EZ |
963 | We think that this can be countermanded with the `xmodmap' utility, if |
964 | you want to be able to bind one of these key sequences within Emacs. | |
cc2f2825 | 965 | |
9dc15871 | 966 | *** Under X, C-v and/or other keys don't work. |
1d297d9b | 967 | |
9dc15871 EZ |
968 | These may have been intercepted by your window manager. In |
969 | particular, AfterStep 1.6 is reported to steal C-v in its default | |
970 | configuration. Various Meta keys are also likely to be taken by the | |
971 | configuration of the `feel'. See the WM's documentation for how to | |
972 | change this. | |
8bd90f0a | 973 | |
9dc15871 | 974 | *** Clicking C-mouse-2 in the scroll bar doesn't split the window. |
8bd90f0a | 975 | |
9dc15871 EZ |
976 | This currently doesn't work with scroll-bar widgets (and we don't know |
977 | a good way of implementing it with widgets). If Emacs is configured | |
978 | --without-toolkit-scroll-bars, C-mouse-2 on the scroll bar does work. | |
8f4df059 | 979 | |
9dc15871 EZ |
980 | *** Inability to send an Alt-modified key, when Emacs is communicating |
981 | directly with an X server. | |
8f4df059 | 982 | |
9dc15871 EZ |
983 | If you have tried to bind an Alt-modified key as a command, and it |
984 | does not work to type the command, the first thing you should check is | |
985 | whether the key is getting through to Emacs. To do this, type C-h c | |
986 | followed by the Alt-modified key. C-h c should say what kind of event | |
987 | it read. If it says it read an Alt-modified key, then make sure you | |
988 | have made the key binding correctly. | |
b098c23c | 989 | |
9dc15871 EZ |
990 | If C-h c reports an event that doesn't have the Alt modifier, it may |
991 | be because your X server has no key for the Alt modifier. The X | |
992 | server that comes from MIT does not set up the Alt modifier by | |
993 | default. | |
224a0b4d | 994 | |
9dc15871 | 995 | If your keyboard has keys named Alt, you can enable them as follows: |
6343352f | 996 | |
9dc15871 EZ |
997 | xmodmap -e 'add mod2 = Alt_L' |
998 | xmodmap -e 'add mod2 = Alt_R' | |
6343352f | 999 | |
9dc15871 EZ |
1000 | If the keyboard has just one key named Alt, then only one of those |
1001 | commands is needed. The modifier `mod2' is a reasonable choice if you | |
1002 | are using an unmodified MIT version of X. Otherwise, choose any | |
1003 | modifier bit not otherwise used. | |
6343352f | 1004 | |
9dc15871 EZ |
1005 | If your keyboard does not have keys named Alt, you can use some other |
1006 | keys. Use the keysym command in xmodmap to turn a function key (or | |
1007 | some other 'spare' key) into Alt_L or into Alt_R, and then use the | |
1008 | commands show above to make them modifier keys. | |
6343352f | 1009 | |
9dc15871 EZ |
1010 | Note that if you have Alt keys but no Meta keys, Emacs translates Alt |
1011 | into Meta. This is because of the great importance of Meta in Emacs. | |
cc2f2825 | 1012 | |
9dc15871 | 1013 | ** Window-manager and toolkit-related problems |
cc2f2825 | 1014 | |
9dc15871 | 1015 | *** Gnome: Emacs' xterm-mouse-mode doesn't work on the Gnome terminal. |
cc2f2825 | 1016 | |
9dc15871 EZ |
1017 | A symptom of this bug is that double-clicks insert a control sequence |
1018 | into the buffer. The reason this happens is an apparent | |
1019 | incompatibility of the Gnome terminal with Xterm, which also affects | |
1020 | other programs using the Xterm mouse interface. A problem report has | |
1021 | been filed. | |
1f42cc71 | 1022 | |
9dc15871 EZ |
1023 | *** KDE: When running on KDE, colors or fonts are not as specified for Emacs, |
1024 | or messed up. | |
1f42cc71 | 1025 | |
9dc15871 EZ |
1026 | For example, you could see background you set for Emacs only in the |
1027 | empty portions of the Emacs display, while characters have some other | |
1028 | background. | |
1f42cc71 | 1029 | |
9dc15871 EZ |
1030 | This happens because KDE's defaults apply its color and font |
1031 | definitions even to applications that weren't compiled for KDE. The | |
1032 | solution is to uncheck the "Apply fonts and colors to non-KDE apps" | |
1033 | option in Preferences->Look&Feel->Style (KDE 2). In KDE 3, this option | |
1034 | is in the "Colors" section, rather than "Style". | |
1f42cc71 | 1035 | |
9dc15871 EZ |
1036 | Alternatively, if you do want the KDE defaults to apply to other |
1037 | applications, but not to Emacs, you could modify the file `Emacs.ad' | |
1038 | (should be in the `/usr/share/apps/kdisplay/app-defaults/' directory) | |
1039 | so that it doesn't set the default background and foreground only for | |
1040 | Emacs. For example, make sure the following resources are either not | |
1041 | present or commented out: | |
f4f4ee4d | 1042 | |
9dc15871 EZ |
1043 | Emacs.default.attributeForeground |
1044 | Emacs.default.attributeBackground | |
1045 | Emacs*Foreground | |
1046 | Emacs*Background | |
0cb26e21 | 1047 | |
9dc15871 | 1048 | *** KDE: Emacs hangs on KDE when a large portion of text is killed. |
f4f4ee4d | 1049 | |
9dc15871 EZ |
1050 | This is caused by a bug in the KDE applet `klipper' which periodically |
1051 | requests the X clipboard contents from applications. Early versions | |
1052 | of klipper don't implement the ICCM protocol for large selections, | |
1053 | which leads to Emacs being flooded with selection requests. After a | |
b11e8823 | 1054 | while, Emacs may print a message: |
f4f4ee4d | 1055 | |
9dc15871 | 1056 | Timed out waiting for property-notify event |
c31138a1 | 1057 | |
b11e8823 JD |
1058 | A workaround is to not use `klipper'. An upgrade to the `klipper' that |
1059 | comes with KDE 3.3 or later also solves the problem. | |
c31138a1 | 1060 | |
9dc15871 | 1061 | *** CDE: Frames may cover dialogs they created when using CDE. |
c31138a1 | 1062 | |
9dc15871 EZ |
1063 | This can happen if you have "Allow Primary Windows On Top" enabled which |
1064 | seems to be the default in the Common Desktop Environment. | |
1065 | To change, go in to "Desktop Controls" -> "Window Style Manager" | |
1066 | and uncheck "Allow Primary Windows On Top". | |
f4f4ee4d | 1067 | |
9dc15871 EZ |
1068 | *** Xaw3d : When using Xaw3d scroll bars without arrows, the very first mouse |
1069 | click in a scroll bar might be ignored by the scroll bar widget. This | |
1070 | is probably a bug in Xaw3d; when Xaw3d is compiled with arrows, the | |
1071 | problem disappears. | |
0c6456ad | 1072 | |
9dc15871 EZ |
1073 | *** Xaw: There are known binary incompatibilities between Xaw, Xaw3d, neXtaw, |
1074 | XawM and the few other derivatives of Xaw. So when you compile with | |
1075 | one of these, it may not work to dynamically link with another one. | |
1076 | For example, strange problems, such as Emacs exiting when you type | |
1077 | "C-x 1", were reported when Emacs compiled with Xaw3d and libXaw was | |
1078 | used with neXtaw at run time. | |
b1739b51 | 1079 | |
9dc15871 EZ |
1080 | The solution is to rebuild Emacs with the toolkit version you actually |
1081 | want to use, or set LD_PRELOAD to preload the same toolkit version you | |
1082 | built Emacs with. | |
b1739b51 | 1083 | |
9dc15871 | 1084 | *** Open Motif: Problems with file dialogs in Emacs built with Open Motif. |
b1739b51 | 1085 | |
9dc15871 EZ |
1086 | When Emacs 21 is built with Open Motif 2.1, it can happen that the |
1087 | graphical file dialog boxes do not work properly. The "OK", "Filter" | |
1088 | and "Cancel" buttons do not respond to mouse clicks. Dragging the | |
1089 | file dialog window usually causes the buttons to work again. | |
b1739b51 | 1090 | |
9dc15871 EZ |
1091 | The solution is to use LessTif instead. LessTif is a free replacement |
1092 | for Motif. See the file INSTALL for information on how to do this. | |
b1739b51 | 1093 | |
9dc15871 EZ |
1094 | Another workaround is not to use the mouse to trigger file prompts, |
1095 | but to use the keyboard. This way, you will be prompted for a file in | |
1096 | the minibuffer instead of a graphical file dialog. | |
b1739b51 | 1097 | |
9dc15871 | 1098 | *** LessTif: Problems in Emacs built with LessTif. |
e9a52cfe | 1099 | |
9dc15871 EZ |
1100 | The problems seem to depend on the version of LessTif and the Motif |
1101 | emulation for which it is set up. | |
e9a52cfe | 1102 | |
9dc15871 EZ |
1103 | Only the Motif 1.2 emulation seems to be stable enough in LessTif. |
1104 | Lesstif 0.92-17's Motif 1.2 emulation seems to work okay on FreeBSD. | |
1105 | On GNU/Linux systems, lesstif-0.92.6 configured with "./configure | |
1106 | --enable-build-12 --enable-default-12" is reported to be the most | |
1107 | successful. The binary GNU/Linux package | |
1108 | lesstif-devel-0.92.0-1.i386.rpm was reported to have problems with | |
1109 | menu placement. | |
e9a52cfe | 1110 | |
9dc15871 EZ |
1111 | On some systems, even with Motif 1.2 emulation, Emacs occasionally |
1112 | locks up, grabbing all mouse and keyboard events. We still don't know | |
1113 | what causes these problems; they are not reproducible by Emacs | |
1114 | developers. | |
e9a52cfe | 1115 | |
9dc15871 | 1116 | *** Motif: The Motif version of Emacs paints the screen a solid color. |
e9a52cfe | 1117 | |
9dc15871 | 1118 | This has been observed to result from the following X resource: |
e9a52cfe | 1119 | |
9dc15871 | 1120 | Emacs*default.attributeFont: -*-courier-medium-r-*-*-*-140-*-*-*-*-iso8859-* |
e9a52cfe | 1121 | |
9dc15871 EZ |
1122 | That the resource has this effect indicates a bug in something, but we |
1123 | do not yet know what. If it is an Emacs bug, we hope someone can | |
1124 | explain what the bug is so we can fix it. In the mean time, removing | |
1125 | the resource prevents the problem. | |
f25eb4f7 | 1126 | |
9dc15871 | 1127 | ** General X problems |
f25eb4f7 | 1128 | |
9dc15871 | 1129 | *** Redisplay using X11 is much slower than previous Emacs versions. |
f25eb4f7 | 1130 | |
9dc15871 EZ |
1131 | We've noticed that certain X servers draw the text much slower when |
1132 | scroll bars are on the left. We don't know why this happens. If this | |
1133 | happens to you, you can work around it by putting the scroll bars | |
1134 | on the right (as they were in Emacs 19). | |
f25eb4f7 | 1135 | |
9dc15871 | 1136 | Here's how to do this: |
f25eb4f7 | 1137 | |
9dc15871 | 1138 | (set-scroll-bar-mode 'right) |
f25eb4f7 | 1139 | |
9dc15871 EZ |
1140 | If you're not sure whether (or how much) this problem affects you, |
1141 | try that and see how much difference it makes. To set things back | |
1142 | to normal, do | |
edd7d3be | 1143 | |
9dc15871 | 1144 | (set-scroll-bar-mode 'left) |
edd7d3be | 1145 | |
9dc15871 | 1146 | *** Error messages about undefined colors on X. |
edd7d3be | 1147 | |
9dc15871 | 1148 | The messages might say something like this: |
42303132 | 1149 | |
9dc15871 | 1150 | Unable to load color "grey95" |
42303132 | 1151 | |
9dc15871 | 1152 | (typically, in the `*Messages*' buffer), or something like this: |
42303132 | 1153 | |
9dc15871 | 1154 | Error while displaying tooltip: (error Undefined color lightyellow) |
42303132 | 1155 | |
9dc15871 EZ |
1156 | These problems could happen if some other X program has used up too |
1157 | many colors of the X palette, leaving Emacs with insufficient system | |
1158 | resources to load all the colors it needs. | |
42303132 | 1159 | |
9dc15871 | 1160 | A solution is to exit the offending X programs before starting Emacs. |
42303132 | 1161 | |
9dc15871 | 1162 | *** Improving performance with slow X connections. |
f3d6f4ee | 1163 | |
9dc15871 EZ |
1164 | There are several ways to improve this performance, any subset of which can |
1165 | be carried out at the same time: | |
f3d6f4ee | 1166 | |
9dc15871 EZ |
1167 | 1) If you don't need X Input Methods (XIM) for entering text in some |
1168 | language you use, you can improve performance on WAN links by using | |
1169 | the X resource useXIM to turn off use of XIM. This does not affect | |
1170 | the use of Emacs' own input methods, which are part of the Leim | |
1171 | package. | |
f3d6f4ee | 1172 | |
9dc15871 EZ |
1173 | 2) If the connection is very slow, you might also want to consider |
1174 | switching off scroll bars, menu bar, and tool bar. | |
f3d6f4ee | 1175 | |
9dc15871 EZ |
1176 | 3) Use ssh to forward the X connection, and enable compression on this |
1177 | forwarded X connection (ssh -XC remotehostname emacs ...). | |
9f83d8b3 | 1178 | |
9dc15871 EZ |
1179 | 4) Use lbxproxy on the remote end of the connection. This is an interface |
1180 | to the low bandwidth X extension in most modern X servers, which | |
1181 | improves performance dramatically, at the slight expense of correctness | |
1182 | of the X protocol. lbxproxy acheives the performance gain by grouping | |
1183 | several X requests in one TCP packet and sending them off together, | |
1184 | instead of requiring a round-trip for each X request in a seperate | |
1185 | packet. The switches that seem to work best for emacs are: | |
1186 | -noatomsfile -nowinattr -cheaterrors -cheatevents | |
1187 | Note that the -nograbcmap option is known to cause problems. | |
1188 | For more about lbxproxy, see: | |
1189 | http://www.xfree86.org/4.3.0/lbxproxy.1.html | |
9f83d8b3 | 1190 | |
9dc15871 | 1191 | *** Emacs gives the error, Couldn't find per display information. |
9f83d8b3 | 1192 | |
9dc15871 EZ |
1193 | This can result if the X server runs out of memory because Emacs uses |
1194 | a large number of fonts. On systems where this happens, C-h h is | |
1195 | likely to cause it. | |
f29d1e75 | 1196 | |
9dc15871 | 1197 | We do not know of a way to prevent the problem. |
f29d1e75 | 1198 | |
9dc15871 | 1199 | *** Emacs does not notice when you release the mouse. |
c24be289 | 1200 | |
9dc15871 EZ |
1201 | There are reports that this happened with (some) Microsoft mice and |
1202 | that replacing the mouse made it stop. | |
c24be289 | 1203 | |
9dc15871 | 1204 | *** You can't select from submenus (in the X toolkit version). |
b35319bf | 1205 | |
9dc15871 EZ |
1206 | On certain systems, mouse-tracking and selection in top-level menus |
1207 | works properly with the X toolkit, but neither of them works when you | |
1208 | bring up a submenu (such as Bookmarks or Compare or Apply Patch, in | |
1209 | the Files menu). | |
b35319bf | 1210 | |
9dc15871 EZ |
1211 | This works on most systems. There is speculation that the failure is |
1212 | due to bugs in old versions of X toolkit libraries, but no one really | |
1213 | knows. If someone debugs this and finds the precise cause, perhaps a | |
1214 | workaround can be found. | |
b35319bf | 1215 | |
9dc15871 EZ |
1216 | *** An error message such as `X protocol error: BadMatch (invalid |
1217 | parameter attributes) on protocol request 93'. | |
b35319bf | 1218 | |
9dc15871 EZ |
1219 | This comes from having an invalid X resource, such as |
1220 | emacs*Cursor: black | |
1221 | (which is invalid because it specifies a color name for something | |
1222 | that isn't a color.) | |
787994b7 | 1223 | |
9dc15871 | 1224 | The fix is to correct your X resources. |
0a2eeca1 | 1225 | |
9dc15871 | 1226 | *** Slow startup on X11R6 with X windows. |
0a2eeca1 | 1227 | |
9dc15871 EZ |
1228 | If Emacs takes two minutes to start up on X11R6, see if your X |
1229 | resources specify any Adobe fonts. That causes the type-1 font | |
1230 | renderer to start up, even if the font you asked for is not a type-1 | |
1231 | font. | |
0a2eeca1 | 1232 | |
9dc15871 EZ |
1233 | One way to avoid this problem is to eliminate the type-1 fonts from |
1234 | your font path, like this: | |
0a2eeca1 | 1235 | |
9dc15871 | 1236 | xset -fp /usr/X11R6/lib/X11/fonts/Type1/ |
0a2eeca1 | 1237 | |
9dc15871 | 1238 | *** Pull-down menus appear in the wrong place, in the toolkit version of Emacs. |
0a2eeca1 | 1239 | |
9dc15871 | 1240 | An X resource of this form can cause the problem: |
0a2eeca1 | 1241 | |
9dc15871 | 1242 | Emacs*geometry: 80x55+0+0 |
0a2eeca1 | 1243 | |
9dc15871 EZ |
1244 | This resource is supposed to apply, and does apply, to the menus |
1245 | individually as well as to Emacs frames. If that is not what you | |
1246 | want, rewrite the resource. | |
119d3665 | 1247 | |
9dc15871 EZ |
1248 | To check thoroughly for such resource specifications, use `xrdb |
1249 | -query' to see what resources the X server records, and also look at | |
1250 | the user's ~/.Xdefaults and ~/.Xdefaults-* files. | |
119d3665 | 1251 | |
9dc15871 EZ |
1252 | *** Emacs running under X Windows does not handle mouse clicks. |
1253 | *** `emacs -geometry 80x20' finds a file named `80x20'. | |
0de9f9a8 | 1254 | |
9dc15871 EZ |
1255 | One cause of such problems is having (setq term-file-prefix nil) in |
1256 | your .emacs file. Another cause is a bad value of EMACSLOADPATH in | |
1257 | the environment. | |
0de9f9a8 | 1258 | |
9dc15871 | 1259 | *** Emacs fails to get default settings from X Windows server. |
0de9f9a8 | 1260 | |
9dc15871 EZ |
1261 | The X library in X11R4 has a bug; it interchanges the 2nd and 3rd |
1262 | arguments to XGetDefaults. Define the macro XBACKWARDS in config.h to | |
1263 | tell Emacs to compensate for this. | |
0de9f9a8 | 1264 | |
9dc15871 EZ |
1265 | I don't believe there is any way Emacs can determine for itself |
1266 | whether this problem is present on a given system. | |
0de9f9a8 | 1267 | |
9dc15871 | 1268 | *** X Windows doesn't work if DISPLAY uses a hostname. |
0de9f9a8 | 1269 | |
9dc15871 EZ |
1270 | People have reported kernel bugs in certain systems that cause Emacs |
1271 | not to work with X Windows if DISPLAY is set using a host name. But | |
1272 | the problem does not occur if DISPLAY is set to `unix:0.0'. I think | |
1273 | the bug has to do with SIGIO or FIONREAD. | |
a933dad1 | 1274 | |
9dc15871 EZ |
1275 | You may be able to compensate for the bug by doing (set-input-mode nil nil). |
1276 | However, that has the disadvantage of turning off interrupts, so that | |
1277 | you are unable to quit out of a Lisp program by typing C-g. | |
a933dad1 | 1278 | |
9dc15871 | 1279 | The easy way to do this is to put |
a933dad1 | 1280 | |
9dc15871 | 1281 | (setq x-sigio-bug t) |
a933dad1 | 1282 | |
9dc15871 | 1283 | in your site-init.el file. |
a933dad1 | 1284 | |
9dc15871 | 1285 | * Runtime problems on character termunals |
a933dad1 | 1286 | |
9dc15871 | 1287 | ** Emacs spontaneously displays "I-search: " at the bottom of the screen. |
a933dad1 | 1288 | |
9dc15871 EZ |
1289 | This means that Control-S/Control-Q (XON/XOFF) "flow control" is being |
1290 | used. C-s/C-q flow control is bad for Emacs editors because it takes | |
1291 | away C-s and C-q as user commands. Since editors do not output long | |
1292 | streams of text without user commands, there is no need for a | |
1293 | user-issuable "stop output" command in an editor; therefore, a | |
1294 | properly designed flow control mechanism would transmit all possible | |
1295 | input characters without interference. Designing such a mechanism is | |
1296 | easy, for a person with at least half a brain. | |
177c0ea7 | 1297 | |
9dc15871 | 1298 | There are three possible reasons why flow control could be taking place: |
a933dad1 | 1299 | |
9dc15871 EZ |
1300 | 1) Terminal has not been told to disable flow control |
1301 | 2) Insufficient padding for the terminal in use | |
1302 | 3) Some sort of terminal concentrator or line switch is responsible | |
a933dad1 | 1303 | |
9dc15871 EZ |
1304 | First of all, many terminals have a set-up mode which controls whether |
1305 | they generate XON/XOFF flow control characters. This must be set to | |
1306 | "no XON/XOFF" in order for Emacs to work. Sometimes there is an | |
1307 | escape sequence that the computer can send to turn flow control off | |
1308 | and on. If so, perhaps the termcap `ti' string should turn flow | |
1309 | control off, and the `te' string should turn it on. | |
a933dad1 | 1310 | |
9dc15871 EZ |
1311 | Once the terminal has been told "no flow control", you may find it |
1312 | needs more padding. The amount of padding Emacs sends is controlled | |
1313 | by the termcap entry for the terminal in use, and by the output baud | |
1314 | rate as known by the kernel. The shell command `stty' will print | |
1315 | your output baud rate; `stty' with suitable arguments will set it if | |
1316 | it is wrong. Setting to a higher speed causes increased padding. If | |
1317 | the results are wrong for the correct speed, there is probably a | |
1318 | problem in the termcap entry. You must speak to a local Unix wizard | |
1319 | to fix this. Perhaps you are just using the wrong terminal type. | |
de121241 | 1320 | |
9dc15871 EZ |
1321 | For terminals that lack a "no flow control" mode, sometimes just |
1322 | giving lots of padding will prevent actual generation of flow control | |
1323 | codes. You might as well try it. | |
de121241 | 1324 | |
9dc15871 EZ |
1325 | If you are really unlucky, your terminal is connected to the computer |
1326 | through a concentrator which sends XON/XOFF flow control to the | |
1327 | computer, or it insists on sending flow control itself no matter how | |
1328 | much padding you give it. Unless you can figure out how to turn flow | |
1329 | control off on this concentrator (again, refer to your local wizard), | |
1330 | you are screwed! You should have the terminal or concentrator | |
1331 | replaced with a properly designed one. In the mean time, some drastic | |
1332 | measures can make Emacs semi-work. | |
de121241 | 1333 | |
9dc15871 EZ |
1334 | You can make Emacs ignore C-s and C-q and let the operating system |
1335 | handle them. To do this on a per-session basis, just type M-x | |
1336 | enable-flow-control RET. You will see a message that C-\ and C-^ are | |
1337 | now translated to C-s and C-q. (Use the same command M-x | |
1338 | enable-flow-control to turn *off* this special mode. It toggles flow | |
1339 | control handling.) | |
a933dad1 | 1340 | |
9dc15871 EZ |
1341 | If C-\ and C-^ are inconvenient for you (for example, if one of them |
1342 | is the escape character of your terminal concentrator), you can choose | |
1343 | other characters by setting the variables flow-control-c-s-replacement | |
1344 | and flow-control-c-q-replacement. But choose carefully, since all | |
1345 | other control characters are already used by emacs. | |
a933dad1 | 1346 | |
9dc15871 EZ |
1347 | IMPORTANT: if you type C-s by accident while flow control is enabled, |
1348 | Emacs output will freeze, and you will have to remember to type C-q in | |
1349 | order to continue. | |
177c0ea7 | 1350 | |
9dc15871 EZ |
1351 | If you work in an environment where a majority of terminals of a |
1352 | certain type are flow control hobbled, you can use the function | |
1353 | `enable-flow-control-on' to turn on this flow control avoidance scheme | |
1354 | automatically. Here is an example: | |
a933dad1 | 1355 | |
9dc15871 | 1356 | (enable-flow-control-on "vt200" "vt300" "vt101" "vt131") |
a933dad1 | 1357 | |
9dc15871 EZ |
1358 | If this isn't quite correct (e.g. you have a mixture of flow-control hobbled |
1359 | and good vt200 terminals), you can still run enable-flow-control | |
1360 | manually. | |
a933dad1 | 1361 | |
9dc15871 EZ |
1362 | I have no intention of ever redesigning the Emacs command set for the |
1363 | assumption that terminals use C-s/C-q flow control. XON/XOFF flow | |
1364 | control technique is a bad design, and terminals that need it are bad | |
1365 | merchandise and should not be purchased. Now that X is becoming | |
1366 | widespread, XON/XOFF seems to be on the way out. If you can get some | |
1367 | use out of GNU Emacs on inferior terminals, more power to you, but I | |
1368 | will not make Emacs worse for properly designed systems for the sake | |
1369 | of inferior systems. | |
a933dad1 | 1370 | |
9dc15871 | 1371 | ** Control-S and Control-Q commands are ignored completely. |
a933dad1 | 1372 | |
9dc15871 EZ |
1373 | For some reason, your system is using brain-damaged C-s/C-q flow |
1374 | control despite Emacs's attempts to turn it off. Perhaps your | |
1375 | terminal is connected to the computer through a concentrator | |
1376 | that wants to use flow control. | |
a933dad1 | 1377 | |
9dc15871 EZ |
1378 | You should first try to tell the concentrator not to use flow control. |
1379 | If you succeed in this, try making the terminal work without | |
1380 | flow control, as described in the preceding section. | |
a933dad1 | 1381 | |
9dc15871 EZ |
1382 | If that line of approach is not successful, map some other characters |
1383 | into C-s and C-q using keyboard-translate-table. The example above | |
1384 | shows how to do this with C-^ and C-\. | |
a933dad1 | 1385 | |
9dc15871 | 1386 | ** Screen is updated wrong, but only on one kind of terminal. |
a933dad1 | 1387 | |
9dc15871 EZ |
1388 | This could mean that the termcap entry you are using for that |
1389 | terminal is wrong, or it could mean that Emacs has a bug handing | |
1390 | the combination of features specified for that terminal. | |
a933dad1 | 1391 | |
9dc15871 EZ |
1392 | The first step in tracking this down is to record what characters |
1393 | Emacs is sending to the terminal. Execute the Lisp expression | |
1394 | (open-termscript "./emacs-script") to make Emacs write all | |
1395 | terminal output into the file ~/emacs-script as well; then do | |
1396 | what makes the screen update wrong, and look at the file | |
1397 | and decode the characters using the manual for the terminal. | |
1398 | There are several possibilities: | |
a933dad1 | 1399 | |
9dc15871 | 1400 | 1) The characters sent are correct, according to the terminal manual. |
a933dad1 | 1401 | |
9dc15871 EZ |
1402 | In this case, there is no obvious bug in Emacs, and most likely you |
1403 | need more padding, or possibly the terminal manual is wrong. | |
a933dad1 | 1404 | |
9dc15871 EZ |
1405 | 2) The characters sent are incorrect, due to an obscure aspect |
1406 | of the terminal behavior not described in an obvious way | |
1407 | by termcap. | |
a933dad1 | 1408 | |
9dc15871 EZ |
1409 | This case is hard. It will be necessary to think of a way for |
1410 | Emacs to distinguish between terminals with this kind of behavior | |
1411 | and other terminals that behave subtly differently but are | |
1412 | classified the same by termcap; or else find an algorithm for | |
1413 | Emacs to use that avoids the difference. Such changes must be | |
1414 | tested on many kinds of terminals. | |
a933dad1 | 1415 | |
9dc15871 | 1416 | 3) The termcap entry is wrong. |
a933dad1 | 1417 | |
9dc15871 EZ |
1418 | See the file etc/TERMS for information on changes |
1419 | that are known to be needed in commonly used termcap entries | |
1420 | for certain terminals. | |
a933dad1 | 1421 | |
9dc15871 EZ |
1422 | 4) The characters sent are incorrect, and clearly cannot be |
1423 | right for any terminal with the termcap entry you were using. | |
a933dad1 | 1424 | |
9dc15871 EZ |
1425 | This is unambiguously an Emacs bug, and can probably be fixed |
1426 | in termcap.c, tparam.c, term.c, scroll.c, cm.c or dispnew.c. | |
a933dad1 | 1427 | |
9dc15871 | 1428 | ** Control-S and Control-Q commands are ignored completely on a net connection. |
a933dad1 | 1429 | |
9dc15871 EZ |
1430 | Some versions of rlogin (and possibly telnet) do not pass flow |
1431 | control characters to the remote system to which they connect. | |
1432 | On such systems, emacs on the remote system cannot disable flow | |
1433 | control on the local system. | |
a933dad1 | 1434 | |
9dc15871 EZ |
1435 | One way to cure this is to disable flow control on the local host |
1436 | (the one running rlogin, not the one running rlogind) using the | |
1437 | stty command, before starting the rlogin process. On many systems, | |
1438 | "stty start u stop u" will do this. | |
a933dad1 | 1439 | |
9dc15871 EZ |
1440 | Some versions of tcsh will prevent even this from working. One way |
1441 | around this is to start another shell before starting rlogin, and | |
1442 | issue the stty command to disable flow control from that shell. | |
a933dad1 | 1443 | |
9dc15871 EZ |
1444 | If none of these methods work, the best solution is to type |
1445 | M-x enable-flow-control at the beginning of your emacs session, or | |
1446 | if you expect the problem to continue, add a line such as the | |
1447 | following to your .emacs (on the host running rlogind): | |
a933dad1 | 1448 | |
9dc15871 | 1449 | (enable-flow-control-on "vt200" "vt300" "vt101" "vt131") |
a933dad1 | 1450 | |
9dc15871 EZ |
1451 | See the entry about spontaneous display of I-search (above) for more |
1452 | info. | |
a933dad1 | 1453 | |
9dc15871 | 1454 | ** Output from Control-V is slow. |
a933dad1 | 1455 | |
9dc15871 EZ |
1456 | On many bit-map terminals, scrolling operations are fairly slow. |
1457 | Often the termcap entry for the type of terminal in use fails | |
1458 | to inform Emacs of this. The two lines at the bottom of the screen | |
1459 | before a Control-V command are supposed to appear at the top after | |
1460 | the Control-V command. If Emacs thinks scrolling the lines is fast, | |
1461 | it will scroll them to the top of the screen. | |
a933dad1 | 1462 | |
9dc15871 EZ |
1463 | If scrolling is slow but Emacs thinks it is fast, the usual reason is |
1464 | that the termcap entry for the terminal you are using does not | |
1465 | specify any padding time for the `al' and `dl' strings. Emacs | |
1466 | concludes that these operations take only as much time as it takes to | |
1467 | send the commands at whatever line speed you are using. You must | |
1468 | fix the termcap entry to specify, for the `al' and `dl', as much | |
1469 | time as the operations really take. | |
a933dad1 | 1470 | |
9dc15871 EZ |
1471 | Currently Emacs thinks in terms of serial lines which send characters |
1472 | at a fixed rate, so that any operation which takes time for the | |
1473 | terminal to execute must also be padded. With bit-map terminals | |
1474 | operated across networks, often the network provides some sort of | |
1475 | flow control so that padding is never needed no matter how slow | |
1476 | an operation is. You must still specify a padding time if you want | |
1477 | Emacs to realize that the operation takes a long time. This will | |
1478 | cause padding characters to be sent unnecessarily, but they do | |
1479 | not really cost much. They will be transmitted while the scrolling | |
1480 | is happening and then discarded quickly by the terminal. | |
a933dad1 | 1481 | |
9dc15871 EZ |
1482 | Most bit-map terminals provide commands for inserting or deleting |
1483 | multiple lines at once. Define the `AL' and `DL' strings in the | |
1484 | termcap entry to say how to do these things, and you will have | |
1485 | fast output without wasted padding characters. These strings should | |
1486 | each contain a single %-spec saying how to send the number of lines | |
1487 | to be scrolled. These %-specs are like those in the termcap | |
1488 | `cm' string. | |
a933dad1 | 1489 | |
9dc15871 EZ |
1490 | You should also define the `IC' and `DC' strings if your terminal |
1491 | has a command to insert or delete multiple characters. These | |
1492 | take the number of positions to insert or delete as an argument. | |
a933dad1 | 1493 | |
9dc15871 EZ |
1494 | A `cs' string to set the scrolling region will reduce the amount |
1495 | of motion you see on the screen when part of the screen is scrolled. | |
a933dad1 | 1496 | |
9dc15871 | 1497 | ** You type Control-H (Backspace) expecting to delete characters. |
a933dad1 | 1498 | |
9dc15871 EZ |
1499 | Put `stty dec' in your .login file and your problems will disappear |
1500 | after a day or two. | |
d238f982 | 1501 | |
9dc15871 EZ |
1502 | The choice of Backspace for erasure was based on confusion, caused by |
1503 | the fact that backspacing causes erasure (later, when you type another | |
1504 | character) on most display terminals. But it is a mistake. Deletion | |
1505 | of text is not the same thing as backspacing followed by failure to | |
1506 | overprint. I do not wish to propagate this confusion by conforming | |
1507 | to it. | |
d238f982 | 1508 | |
9dc15871 EZ |
1509 | For this reason, I believe `stty dec' is the right mode to use, |
1510 | and I have designed Emacs to go with that. If there were a thousand | |
1511 | other control characters, I would define Control-h to delete as well; | |
1512 | but there are not very many other control characters, and I think | |
1513 | that providing the most mnemonic possible Help character is more | |
1514 | important than adapting to people who don't use `stty dec'. | |
d7185f9d | 1515 | |
9dc15871 EZ |
1516 | If you are obstinate about confusing buggy overprinting with deletion, |
1517 | you can redefine Backspace in your .emacs file: | |
1518 | (global-set-key "\b" 'delete-backward-char) | |
1519 | You can probably access help-command via f1. | |
d7185f9d | 1520 | |
9dc15871 | 1521 | ** Colors are not available on a tty or in xterm. |
a933dad1 | 1522 | |
9dc15871 EZ |
1523 | Emacs 21 supports colors on character terminals and terminal |
1524 | emulators, but this support relies on the terminfo or termcap database | |
1525 | entry to specify that the display supports color. Emacs looks at the | |
1526 | "Co" capability for the terminal to find out how many colors are | |
1527 | supported; it should be non-zero to activate the color support within | |
1528 | Emacs. (Most color terminals support 8 or 16 colors.) If your system | |
1529 | uses terminfo, the name of the capability equivalent to "Co" is | |
1530 | "colors". | |
a933dad1 | 1531 | |
9dc15871 EZ |
1532 | In addition to the "Co" capability, Emacs needs the "op" (for |
1533 | ``original pair'') capability, which tells how to switch the terminal | |
1534 | back to the default foreground and background colors. Emacs will not | |
1535 | use colors if this capability is not defined. If your terminal entry | |
1536 | doesn't provide such a capability, try using the ANSI standard escape | |
1537 | sequence \E[00m (that is, define a new termcap/terminfo entry and make | |
1538 | it use your current terminal's entry plus \E[00m for the "op" | |
1539 | capability). | |
a933dad1 | 1540 | |
9dc15871 EZ |
1541 | Finally, the "NC" capability (terminfo name: "ncv") tells Emacs which |
1542 | attributes cannot be used with colors. Setting this capability | |
1543 | incorrectly might have the effect of disabling colors; try setting | |
1544 | this capability to `0' (zero) and see if that helps. | |
a933dad1 | 1545 | |
9dc15871 EZ |
1546 | Emacs uses the database entry for the terminal whose name is the value |
1547 | of the environment variable TERM. With `xterm', a common terminal | |
1548 | entry that supports color is `xterm-color', so setting TERM's value to | |
1549 | `xterm-color' might activate the color support on an xterm-compatible | |
1550 | emulator. | |
a933dad1 | 1551 | |
bf247b6e | 1552 | Beginning with version 22.1, Emacs supports the --color command-line |
9dc15871 EZ |
1553 | option which may be used to force Emacs to use one of a few popular |
1554 | modes for getting colors on a tty. For example, --color=ansi8 sets up | |
1555 | for using the ANSI-standard escape sequences that support 8 colors. | |
a933dad1 | 1556 | |
9dc15871 EZ |
1557 | Some modes do not use colors unless you turn on the Font-lock mode. |
1558 | Some people have long ago set their `~/.emacs' files to turn on | |
1559 | Font-lock on X only, so they won't see colors on a tty. The | |
1560 | recommended way of turning on Font-lock is by typing "M-x | |
1561 | global-font-lock-mode RET" or by customizing the variable | |
1562 | `global-font-lock-mode'. | |
a933dad1 | 1563 | |
9dc15871 | 1564 | * Runtime problems specific to individual Unix variants |
a933dad1 | 1565 | |
9dc15871 | 1566 | ** GNU/Linux |
a933dad1 | 1567 | |
9dc15871 EZ |
1568 | *** GNU/Linux: On Linux-based GNU systems using libc versions 5.4.19 through |
1569 | 5.4.22, Emacs crashes at startup with a segmentation fault. | |
a933dad1 | 1570 | |
9dc15871 EZ |
1571 | This problem happens if libc defines the symbol __malloc_initialized. |
1572 | One known solution is to upgrade to a newer libc version. 5.4.33 is | |
1573 | known to work. | |
a933dad1 | 1574 | |
9dc15871 EZ |
1575 | *** GNU/Linux: After upgrading to a newer version of Emacs, |
1576 | the Meta key stops working. | |
a933dad1 | 1577 | |
9dc15871 EZ |
1578 | This was reported to happen on a GNU/Linux system distributed by |
1579 | Mandrake. The reason is that the previous version of Emacs was | |
1580 | modified by Mandrake to make the Alt key act as the Meta key, on a | |
1581 | keyboard where the Windows key is the one which produces the Meta | |
1582 | modifier. A user who started using a newer version of Emacs, which | |
1583 | was not hacked by Mandrake, expected the Alt key to continue to act as | |
1584 | Meta, and was astonished when that didn't happen. | |
a933dad1 | 1585 | |
9dc15871 EZ |
1586 | The solution is to find out what key on your keyboard produces the Meta |
1587 | modifier, and use that key instead. Try all of the keys to the left | |
1588 | and to the right of the space bar, together with the `x' key, and see | |
1589 | which combination produces "M-x" in the echo area. You can also use | |
1590 | the `xmodmap' utility to show all the keys which produce a Meta | |
1591 | modifier: | |
a933dad1 | 1592 | |
9dc15871 | 1593 | xmodmap -pk | egrep -i "meta|alt" |
17a37d87 | 1594 | |
9dc15871 EZ |
1595 | A more convenient way of finding out which keys produce a Meta modifier |
1596 | is to use the `xkbprint' utility, if it's available on your system: | |
17a37d87 | 1597 | |
9dc15871 | 1598 | xkbprint 0:0 /tmp/k.ps |
17a37d87 | 1599 | |
9dc15871 EZ |
1600 | This produces a PostScript file `/tmp/k.ps' with a picture of your |
1601 | keyboard; printing that file on a PostScript printer will show what | |
1602 | keys can serve as Meta. | |
17a37d87 | 1603 | |
9dc15871 EZ |
1604 | The `xkeycaps' also shows a visual representation of the current |
1605 | keyboard settings. It also allows to modify them. | |
a933dad1 | 1606 | |
9dc15871 | 1607 | *** GNU/Linux: low startup on Linux-based GNU systems. |
a933dad1 | 1608 | |
9dc15871 EZ |
1609 | People using systems based on the Linux kernel sometimes report that |
1610 | startup takes 10 to 15 seconds longer than `usual'. | |
a933dad1 | 1611 | |
9dc15871 EZ |
1612 | This is because Emacs looks up the host name when it starts. |
1613 | Normally, this takes negligible time; the extra delay is due to | |
1614 | improper system configuration. This problem can occur for both | |
1615 | networked and non-networked machines. | |
a933dad1 | 1616 | |
9dc15871 | 1617 | Here is how to fix the configuration. It requires being root. |
a933dad1 | 1618 | |
9dc15871 | 1619 | **** Networked Case. |
a933dad1 | 1620 | |
9dc15871 EZ |
1621 | First, make sure the files `/etc/hosts' and `/etc/host.conf' both |
1622 | exist. The first line in the `/etc/hosts' file should look like this | |
1623 | (replace HOSTNAME with your host name): | |
a933dad1 | 1624 | |
9dc15871 | 1625 | 127.0.0.1 HOSTNAME |
a933dad1 | 1626 | |
9dc15871 EZ |
1627 | Also make sure that the `/etc/host.conf' files contains the following |
1628 | lines: | |
a933dad1 | 1629 | |
9dc15871 EZ |
1630 | order hosts, bind |
1631 | multi on | |
a933dad1 | 1632 | |
9dc15871 EZ |
1633 | Any changes, permanent and temporary, to the host name should be |
1634 | indicated in the `/etc/hosts' file, since it acts a limited local | |
1635 | database of addresses and names (e.g., some SLIP connections | |
1636 | dynamically allocate ip addresses). | |
a933dad1 | 1637 | |
9dc15871 | 1638 | **** Non-Networked Case. |
a933dad1 | 1639 | |
9dc15871 EZ |
1640 | The solution described in the networked case applies here as well. |
1641 | However, if you never intend to network your machine, you can use a | |
1642 | simpler solution: create an empty `/etc/host.conf' file. The command | |
1643 | `touch /etc/host.conf' suffices to create the file. The `/etc/hosts' | |
1644 | file is not necessary with this approach. | |
3d00585e | 1645 | |
9dc15871 | 1646 | *** GNU/Linux: Emacs on a tty switches the cursor to large blinking block. |
3d00585e | 1647 | |
9dc15871 EZ |
1648 | This was reported to happen on some GNU/Linux systems which use |
1649 | ncurses version 5.0, but could be relevant for other versions as well. | |
1650 | These versions of ncurses come with a `linux' terminfo entry, where | |
1651 | the "cvvis" capability (termcap "vs") is defined as "\E[?25h\E[?8c" | |
1652 | (show cursor, change size). This escape sequence switches on a | |
1653 | blinking hardware text-mode cursor whose size is a full character | |
1654 | cell. This blinking cannot be stopped, since a hardware cursor | |
1655 | always blinks. | |
3d00585e | 1656 | |
9dc15871 EZ |
1657 | A work-around is to redefine the "cvvis" capability so that it |
1658 | enables a *software* cursor. The software cursor works by inverting | |
1659 | the colors of the character at point, so what you see is a block | |
1660 | cursor that doesn't blink. For this to work, you need to redefine | |
1661 | the "cnorm" capability as well, so that it operates on the software | |
1662 | cursor instead of the hardware cursor. | |
3d00585e | 1663 | |
9dc15871 EZ |
1664 | To this end, run "infocmp linux > linux-term", edit the file |
1665 | `linux-term' to make both the "cnorm" and "cvvis" capabilities send | |
1666 | the sequence "\E[?25h\E[?17;0;64c", and then run "tic linux-term" to | |
1667 | produce a modified terminfo entry. | |
3d00585e | 1668 | |
9dc15871 EZ |
1669 | Alternatively, if you want a blinking underscore as your Emacs cursor, |
1670 | change the "cvvis" capability to send the "\E[?25h\E[?0c" command. | |
a933dad1 | 1671 | |
9dc15871 | 1672 | *** GNU/Linux: Error messages `internal facep []' happen on GNU/Linux systems. |
7838ea1b | 1673 | |
9dc15871 EZ |
1674 | There is a report that replacing libc.so.5.0.9 with libc.so.5.2.16 |
1675 | caused this to start happening. People are not sure why, but the | |
1676 | problem seems unlikely to be in Emacs itself. Some suspect that it | |
1677 | is actually Xlib which won't work with libc.so.5.2.16. | |
a933dad1 | 1678 | |
9dc15871 | 1679 | Using the old library version is a workaround. |
a933dad1 | 1680 | |
9dc15871 | 1681 | ** Mac OS X |
a933dad1 | 1682 | |
9dc15871 | 1683 | *** Mac OS X (Carbon): Environment Variables from dotfiles are ignored. |
a933dad1 | 1684 | |
9dc15871 EZ |
1685 | When starting Emacs from the Dock or the Finder on Mac OS X, the |
1686 | environment variables that are set up in dotfiles, such as .cshrc or | |
1687 | .profile, are ignored. This is because the Finder and Dock are not | |
1688 | started from a shell, but instead from the Window Manager itself. | |
a933dad1 | 1689 | |
9dc15871 EZ |
1690 | The workaround for this is to create a .MacOSX/environment.plist file to |
1691 | setup these environment variables. These environment variables will | |
1692 | apply to all processes regardless of where they are started. | |
1693 | For me information, see http://developer.apple.com/qa/qa2001/qa1067.html. | |
b5cb4652 | 1694 | |
9dc15871 | 1695 | *** Mac OS X (Carbon): Process output truncated when using ptys. |
b5cb4652 | 1696 | |
9dc15871 EZ |
1697 | There appears to be a problem with the implementation of pty's on the |
1698 | Mac OS X that causes process output to be truncated. To avoid this, | |
1699 | leave process-connection-type set to its default value of nil. | |
a933dad1 | 1700 | |
9dc15871 | 1701 | ** FreeBSD |
a933dad1 | 1702 | |
9dc15871 EZ |
1703 | *** FreeBSD 2.1.5: useless symbolic links remain in /tmp or other |
1704 | directories that have the +t bit. | |
a933dad1 | 1705 | |
9dc15871 EZ |
1706 | This is because of a kernel bug in FreeBSD 2.1.5 (fixed in 2.2). |
1707 | Emacs uses symbolic links to implement file locks. In a directory | |
1708 | with +t bit, the directory owner becomes the owner of the symbolic | |
1709 | link, so that it cannot be removed by anyone else. | |
a933dad1 | 1710 | |
9dc15871 EZ |
1711 | If you don't like those useless links, you can let Emacs not to using |
1712 | file lock by adding #undef CLASH_DETECTION to config.h. | |
a933dad1 | 1713 | |
9dc15871 | 1714 | *** FreeBSD: Getting a Meta key on the console. |
a933dad1 | 1715 | |
9dc15871 EZ |
1716 | By default, neither Alt nor any other key acts as a Meta key on |
1717 | FreeBSD, but this can be changed using kbdcontrol(1). Dump the | |
1718 | current keymap to a file with the command | |
a933dad1 | 1719 | |
9dc15871 | 1720 | $ kbdcontrol -d >emacs.kbd |
a933dad1 | 1721 | |
9dc15871 EZ |
1722 | Edit emacs.kbd, and give the key you want to be the Meta key the |
1723 | definition `meta'. For instance, if your keyboard has a ``Windows'' | |
1724 | key with scan code 105, change the line for scan code 105 in emacs.kbd | |
1725 | to look like this | |
3156909f | 1726 | |
9dc15871 | 1727 | 105 meta meta meta meta meta meta meta meta O |
a933dad1 | 1728 | |
9dc15871 | 1729 | to make the Windows key the Meta key. Load the new keymap with |
a933dad1 | 1730 | |
9dc15871 | 1731 | $ kbdcontrol -l emacs.kbd |
a933dad1 | 1732 | |
9dc15871 | 1733 | ** HP-UX |
e96c5c69 | 1734 | |
9dc15871 | 1735 | *** HP/UX : Shell mode gives the message, "`tty`: Ambiguous". |
e96c5c69 | 1736 | |
9dc15871 | 1737 | christos@theory.tn.cornell.edu says: |
a933dad1 | 1738 | |
9dc15871 EZ |
1739 | The problem is that in your .cshrc you have something that tries to |
1740 | execute `tty`. If you are not running the shell on a real tty then | |
1741 | tty will print "not a tty". Csh expects one word in some places, | |
1742 | but tty is giving it back 3. | |
a933dad1 | 1743 | |
9dc15871 EZ |
1744 | The solution is to add a pair of quotes around `tty` to make it a single |
1745 | word: | |
a933dad1 | 1746 | |
9dc15871 | 1747 | if (`tty` == "/dev/console") |
a933dad1 | 1748 | |
9dc15871 | 1749 | should be changed to: |
a933dad1 | 1750 | |
9dc15871 | 1751 | if ("`tty`" == "/dev/console") |
a933dad1 | 1752 | |
9dc15871 EZ |
1753 | Even better, move things that set up terminal sections out of .cshrc |
1754 | and into .login. | |
a933dad1 | 1755 | |
9dc15871 | 1756 | *** HP/UX: `Pid xxx killed due to text modification or page I/O error'. |
a933dad1 | 1757 | |
9dc15871 EZ |
1758 | On HP/UX, you can get that error when the Emacs executable is on an NFS |
1759 | file system. HP/UX responds this way if it tries to swap in a page and | |
1760 | does not get a response from the server within a timeout whose default | |
1761 | value is just ten seconds. | |
a933dad1 | 1762 | |
9dc15871 | 1763 | If this happens to you, extend the timeout period. |
a933dad1 | 1764 | |
9dc15871 EZ |
1765 | *** HP/UX: The right Alt key works wrong on German HP keyboards (and perhaps |
1766 | other non-English HP keyboards too). | |
a933dad1 | 1767 | |
9dc15871 EZ |
1768 | This is because HP-UX defines the modifiers wrong in X. Here is a |
1769 | shell script to fix the problem; be sure that it is run after VUE | |
1770 | configures the X server. | |
a933dad1 | 1771 | |
9dc15871 EZ |
1772 | xmodmap 2> /dev/null - << EOF |
1773 | keysym Alt_L = Meta_L | |
1774 | keysym Alt_R = Meta_R | |
1775 | EOF | |
a933dad1 | 1776 | |
9dc15871 EZ |
1777 | xmodmap - << EOF |
1778 | clear mod1 | |
1779 | keysym Mode_switch = NoSymbol | |
1780 | add mod1 = Meta_L | |
1781 | keysym Meta_R = Mode_switch | |
1782 | add mod2 = Mode_switch | |
1783 | EOF | |
a933dad1 | 1784 | |
9dc15871 EZ |
1785 | *** HP/UX: "Cannot find callback list" messages from dialog boxes in |
1786 | Emacs built with Motif. | |
a933dad1 | 1787 | |
9dc15871 EZ |
1788 | This problem resulted from a bug in GCC 2.4.5. Newer GCC versions |
1789 | such as 2.7.0 fix the problem. | |
a933dad1 | 1790 | |
9dc15871 | 1791 | *** HP/UX: Emacs does not recognize the AltGr key. |
a933dad1 | 1792 | |
9dc15871 EZ |
1793 | To fix this, set up a file ~/.dt/sessions/sessionetc with executable |
1794 | rights, containing this text: | |
4c635a29 | 1795 | |
9dc15871 EZ |
1796 | -------------------------------- |
1797 | xmodmap 2> /dev/null - << EOF | |
1798 | keysym Alt_L = Meta_L | |
1799 | keysym Alt_R = Meta_R | |
1800 | EOF | |
a933dad1 | 1801 | |
9dc15871 EZ |
1802 | xmodmap - << EOF |
1803 | clear mod1 | |
1804 | keysym Mode_switch = NoSymbol | |
1805 | add mod1 = Meta_L | |
1806 | keysym Meta_R = Mode_switch | |
1807 | add mod2 = Mode_switch | |
1808 | EOF | |
1809 | -------------------------------- | |
a933dad1 | 1810 | |
9dc15871 | 1811 | *** HP/UX 11.0: Emacs makes HP/UX 11.0 crash. |
a933dad1 | 1812 | |
9dc15871 | 1813 | This is a bug in HPUX; HPUX patch PHKL_16260 is said to fix it. |
a933dad1 | 1814 | |
9dc15871 | 1815 | ** AIX |
a933dad1 | 1816 | |
9dc15871 | 1817 | *** AIX: Trouble using ptys. |
a933dad1 | 1818 | |
9dc15871 EZ |
1819 | People often install the pty devices on AIX incorrectly. |
1820 | Use `smit pty' to reinstall them properly. | |
a933dad1 | 1821 | |
9dc15871 | 1822 | *** AIXterm: Your Delete key sends a Backspace to the terminal. |
a933dad1 | 1823 | |
9dc15871 | 1824 | The solution is to include in your .Xdefaults the lines: |
a933dad1 | 1825 | |
9dc15871 EZ |
1826 | *aixterm.Translations: #override <Key>BackSpace: string(0x7f) |
1827 | aixterm*ttyModes: erase ^? | |
a933dad1 | 1828 | |
9dc15871 | 1829 | This makes your Backspace key send DEL (ASCII 127). |
a933dad1 | 1830 | |
9dc15871 EZ |
1831 | *** AIX: If linking fails because libXbsd isn't found, check if you |
1832 | are compiling with the system's `cc' and CFLAGS containing `-O5'. If | |
1833 | so, you have hit a compiler bug. Please make sure to re-configure | |
1834 | Emacs so that it isn't compiled with `-O5'. | |
a933dad1 | 1835 | |
9dc15871 | 1836 | *** AIX 4.3.x or 4.4: Compiling fails. |
a933dad1 | 1837 | |
9dc15871 EZ |
1838 | This could happen if you use /bin/c89 as your compiler, instead of |
1839 | the default `cc'. /bin/c89 treats certain warnings, such as benign | |
1840 | redefinitions of macros, as errors, and fails the build. A solution | |
1841 | is to use the default compiler `cc'. | |
a933dad1 | 1842 | |
9dc15871 EZ |
1843 | *** AIX 4: Some programs fail when run in a Shell buffer |
1844 | with an error message like No terminfo entry for "unknown". | |
a933dad1 | 1845 | |
9dc15871 EZ |
1846 | On AIX, many terminal type definitions are not installed by default. |
1847 | `unknown' is one of them. Install the "Special Generic Terminal | |
1848 | Definitions" to make them defined. | |
a933dad1 | 1849 | |
9dc15871 | 1850 | ** Solaris |
a933dad1 | 1851 | |
9dc15871 EZ |
1852 | We list bugs in current versions here. Solaris 2.x and 4.x are covered in the |
1853 | section on legacy systems. | |
a933dad1 | 1854 | |
9dc15871 | 1855 | *** On Solaris, C-x doesn't get through to Emacs when you use the console. |
a933dad1 | 1856 | |
9dc15871 EZ |
1857 | This is a Solaris feature (at least on Intel x86 cpus). Type C-r |
1858 | C-r C-t, to toggle whether C-x gets through to Emacs. | |
a933dad1 | 1859 | |
9dc15871 | 1860 | *** Problem with remote X server on Suns. |
a933dad1 | 1861 | |
9dc15871 EZ |
1862 | On a Sun, running Emacs on one machine with the X server on another |
1863 | may not work if you have used the unshared system libraries. This | |
1864 | is because the unshared libraries fail to use YP for host name lookup. | |
1865 | As a result, the host name you specify may not be recognized. | |
a933dad1 | 1866 | |
0a4dd4e4 | 1867 | *** Solaris 2,6: Emacs crashes with SIGBUS or SIGSEGV on Solaris after you delete a frame. |
a933dad1 | 1868 | |
0a4dd4e4 EZ |
1869 | We suspect that this is a bug in the X libraries provided by |
1870 | Sun. There is a report that one of these patches fixes the bug and | |
1871 | makes the problem stop: | |
a933dad1 | 1872 | |
0a4dd4e4 EZ |
1873 | 105216-01 105393-01 105518-01 105621-01 105665-01 105615-02 105216-02 |
1874 | 105667-01 105401-08 105615-03 105621-02 105686-02 105736-01 105755-03 | |
1875 | 106033-01 105379-01 105786-01 105181-04 105379-03 105786-04 105845-01 | |
1876 | 105284-05 105669-02 105837-01 105837-02 105558-01 106125-02 105407-01 | |
a933dad1 | 1877 | |
0a4dd4e4 EZ |
1878 | Another person using a newer system (kernel patch level Generic_105181-06) |
1879 | suspects that the bug was fixed by one of these more recent patches: | |
a933dad1 | 1880 | |
0a4dd4e4 EZ |
1881 | 106040-07 SunOS 5.6: X Input & Output Method patch |
1882 | 106222-01 OpenWindows 3.6: filemgr (ff.core) fixes | |
1883 | 105284-12 Motif 1.2.7: sparc Runtime library patch | |
a933dad1 | 1884 | |
0a4dd4e4 | 1885 | *** Solaris 7 or 8: Emacs reports a BadAtom error (from X) |
a933dad1 | 1886 | |
0a4dd4e4 EZ |
1887 | This happens when Emacs was built on some other version of Solaris. |
1888 | Rebuild it on Solaris 8. | |
a933dad1 | 1889 | |
9dc15871 EZ |
1890 | *** When using M-x dbx with the SparcWorks debugger, the `up' and `down' |
1891 | commands do not move the arrow in Emacs. | |
a933dad1 | 1892 | |
9dc15871 | 1893 | You can fix this by adding the following line to `~/.dbxinit': |
a933dad1 | 1894 | |
9dc15871 | 1895 | dbxenv output_short_file_name off |
a933dad1 | 1896 | |
0a4dd4e4 EZ |
1897 | *** On Solaris, CTRL-t is ignored by Emacs when you use |
1898 | the fr.ISO-8859-15 locale (and maybe other related locales). | |
a933dad1 | 1899 | |
0a4dd4e4 | 1900 | You can fix this by editing the file: |
a933dad1 | 1901 | |
0a4dd4e4 | 1902 | /usr/openwin/lib/locale/iso8859-15/Compose |
a01325b8 | 1903 | |
0a4dd4e4 | 1904 | Near the bottom there is a line that reads: |
a933dad1 | 1905 | |
0a4dd4e4 | 1906 | Ctrl<t> <quotedbl> <Y> : "\276" threequarters |
a933dad1 | 1907 | |
0a4dd4e4 | 1908 | that should read: |
a933dad1 | 1909 | |
0a4dd4e4 | 1910 | Ctrl<T> <quotedbl> <Y> : "\276" threequarters |
a933dad1 | 1911 | |
0a4dd4e4 | 1912 | Note the lower case <t>. Changing this line should make C-t work. |
a933dad1 | 1913 | |
0a4dd4e4 | 1914 | ** Irix |
a933dad1 | 1915 | |
9dc15871 | 1916 | *** Irix 6.5: Emacs crashes on the SGI R10K, when compiled with GCC. |
a933dad1 | 1917 | |
9dc15871 | 1918 | This seems to be fixed in GCC 2.95. |
a933dad1 | 1919 | |
0a4dd4e4 | 1920 | *** Irix: Trouble using ptys, or running out of ptys. |
a933dad1 | 1921 | |
9dc15871 EZ |
1922 | The program mkpts (which may be in `/usr/adm' or `/usr/sbin') needs to |
1923 | be set-UID to root, or non-root programs like Emacs will not be able | |
1924 | to allocate ptys reliably. | |
a933dad1 | 1925 | |
9dc15871 | 1926 | * Runtime problems specific to MS-Windows |
a933dad1 | 1927 | |
9dc15871 | 1928 | ** Emacs exits with "X protocol error" when run with an X server for MS-Windows. |
a933dad1 | 1929 | |
9dc15871 EZ |
1930 | A certain X server for Windows had a bug which caused this. |
1931 | Supposedly the newer 32-bit version of this server doesn't have the | |
1932 | problem. | |
a933dad1 | 1933 | |
9dc15871 | 1934 | ** Known problems with the MS-Windows port of Emacs 21.2. |
a933dad1 | 1935 | |
9dc15871 EZ |
1936 | Frames are not refreshed while the File or Font dialog or a pop-up menu |
1937 | is displayed. This also means help text for pop-up menus is not | |
1938 | displayed at all. This is because message handling under Windows is | |
1939 | synchronous, so we cannot handle repaint (or any other) messages while | |
1940 | waiting for a system function to return the result of the dialog or | |
1941 | pop-up menu interaction. | |
a933dad1 | 1942 | |
9dc15871 EZ |
1943 | Windows 95 and Windows NT up to version 4.0 do not support help text |
1944 | for menus. Help text is only available in later versions of Windows. | |
a933dad1 | 1945 | |
9dc15871 EZ |
1946 | There are problems with display if mouse-tracking is enabled and the |
1947 | mouse is moved off a frame, over another frame then back over the first | |
1948 | frame. A workaround is to click the left mouse button inside the frame | |
1949 | after moving back into it. | |
a933dad1 | 1950 | |
9dc15871 EZ |
1951 | Some minor flickering still persists during mouse-tracking, although |
1952 | not as severely as in 21.1. | |
a933dad1 | 1953 | |
9dc15871 EZ |
1954 | Emacs can sometimes abort when non-ASCII text, possibly with null |
1955 | characters, is copied and pasted into a buffer. | |
a933dad1 | 1956 | |
9dc15871 EZ |
1957 | An inactive cursor remains in an active window after the Windows |
1958 | Manager driven switch of the focus, until a key is pressed. | |
a933dad1 | 1959 | |
9dc15871 EZ |
1960 | Windows input methods are not recognized by Emacs (as of v21.2). Some |
1961 | of these input methods cause the keyboard to send characters encoded | |
1962 | in the appropriate coding system (e.g., ISO 8859-1 for Latin-1 | |
1963 | characters, ISO 8859-8 for Hebrew characters, etc.). To make this | |
1964 | work, set the keyboard coding system to the appropriate value after | |
1965 | you activate the Windows input method. For example, if you activate | |
1966 | the Hebrew input method, type "C-x RET k iso-8859-8 RET". (Emacs | |
1967 | ought to recognize the Windows language-change event and set up the | |
1968 | appropriate keyboard encoding automatically, but it doesn't do that | |
1969 | yet.) | |
a933dad1 | 1970 | |
9dc15871 EZ |
1971 | The %b specifier for format-time-string does not produce abbreviated |
1972 | month names with consistent widths for some locales on some versions | |
1973 | of Windows. This is caused by a deficiency in the underlying system | |
1974 | library function. | |
a933dad1 | 1975 | |
0a4dd4e4 | 1976 | ** Typing Alt-Shift has strange effects on MS-Windows. |
a933dad1 | 1977 | |
0a4dd4e4 EZ |
1978 | This combination of keys is a command to change keyboard layout. If |
1979 | you proceed to type another non-modifier key before you let go of Alt | |
1980 | and Shift, the Alt and Shift act as modifiers in the usual way. A | |
1981 | more permanent work around is to change it to another key combination, | |
1982 | or disable it in the keyboard control panel. | |
a933dad1 | 1983 | |
0a4dd4e4 | 1984 | ** Interrupting Cygwin port of Bash from Emacs doesn't work. |
a933dad1 | 1985 | |
0a4dd4e4 EZ |
1986 | Cygwin 1.x builds of the ported Bash cannot be interrupted from the |
1987 | MS-Windows version of Emacs. This is due to some change in the Bash | |
1988 | port or in the Cygwin library which apparently make Bash ignore the | |
1989 | keyboard interrupt event sent by Emacs to Bash. (Older Cygwin ports | |
1990 | of Bash, up to b20.1, did receive SIGINT from Emacs.) | |
a933dad1 | 1991 | |
0a4dd4e4 | 1992 | ** Accessing remote files with ange-ftp hangs the MS-Windows version of Emacs. |
a933dad1 | 1993 | |
0a4dd4e4 EZ |
1994 | If the FTP client is the Cygwin port of GNU `ftp', this appears to be |
1995 | due to some bug in the Cygwin DLL or some incompatibility between it | |
1996 | and the implementation of asynchronous subprocesses in the Windows | |
1997 | port of Emacs. Specifically, some parts of the FTP server responses | |
1998 | are not flushed out, apparently due to buffering issues, which | |
1999 | confuses ange-ftp. | |
a933dad1 | 2000 | |
0a4dd4e4 EZ |
2001 | The solution is to downgrade to an older version of the Cygwin DLL |
2002 | (version 1.3.2 was reported to solve the problem), or use the stock | |
2003 | Windows FTP client, usually found in the `C:\WINDOWS' or 'C:\WINNT' | |
2004 | directory. To force ange-ftp use the stock Windows client, set the | |
2005 | variable `ange-ftp-ftp-program-name' to the absolute file name of the | |
2006 | client's executable. For example: | |
a933dad1 | 2007 | |
9dc15871 | 2008 | (setq ange-ftp-ftp-program-name "c:/windows/ftp.exe") |
a933dad1 | 2009 | |
9dc15871 EZ |
2010 | If you want to stick with the Cygwin FTP client, you can work around |
2011 | this problem by putting this in your `.emacs' file: | |
a933dad1 | 2012 | |
9dc15871 | 2013 | (setq ange-ftp-ftp-program-args '("-i" "-n" "-g" "-v" "--prompt" "") |
a933dad1 | 2014 | |
9dc15871 | 2015 | ** lpr commands don't work on MS-Windows with some cheap printers. |
a933dad1 | 2016 | |
9dc15871 EZ |
2017 | This problem may also strike other platforms, but the solution is |
2018 | likely to be a global one, and not Emacs specific. | |
a933dad1 | 2019 | |
9dc15871 EZ |
2020 | Many cheap inkjet, and even some cheap laser printers, do not |
2021 | print plain text anymore, they will only print through graphical | |
2022 | printer drivers. A workaround on MS-Windows is to use Windows' basic | |
2023 | built in editor to print (this is possibly the only useful purpose it | |
2024 | has): | |
a933dad1 | 2025 | |
9dc15871 EZ |
2026 | (setq printer-name "") ;; notepad takes the default |
2027 | (setq lpr-command "notepad") ;; notepad | |
2028 | (setq lpr-switches nil) ;; not needed | |
2029 | (setq lpr-printer-switch "/P") ;; run notepad as batch printer | |
a933dad1 | 2030 | |
9dc15871 | 2031 | ** Antivirus software interacts badly with the MS-Windows version of Emacs. |
a933dad1 | 2032 | |
9dc15871 EZ |
2033 | The usual manifestation of these problems is that subprocesses don't |
2034 | work or even wedge the entire system. In particular, "M-x shell RET" | |
2035 | was reported to fail to work. But other commands also sometimes don't | |
2036 | work when an antivirus package is installed. | |
a933dad1 | 2037 | |
9dc15871 EZ |
2038 | The solution is to switch the antivirus software to a less aggressive |
2039 | mode (e.g., disable the ``auto-protect'' feature), or even uninstall | |
2040 | or disable it entirely. | |
a933dad1 | 2041 | |
9dc15871 | 2042 | ** Pressing the mouse button on MS-Windows does not give a mouse-2 event. |
a933dad1 | 2043 | |
9dc15871 EZ |
2044 | This is usually a problem with the mouse driver. Because most Windows |
2045 | programs do not do anything useful with the middle mouse button, many | |
2046 | mouse drivers allow you to define the wheel press to do something | |
2047 | different. Some drivers do not even have the option to generate a | |
2048 | middle button press. In such cases, setting the wheel press to | |
2049 | "scroll" sometimes works if you press the button twice. Trying a | |
2050 | generic mouse driver might help. | |
a933dad1 | 2051 | |
9dc15871 | 2052 | ** Scrolling the mouse wheel on MS-Windows always scrolls the top window. |
a933dad1 | 2053 | |
9dc15871 EZ |
2054 | This is another common problem with mouse drivers. Instead of |
2055 | generating scroll events, some mouse drivers try to fake scroll bar | |
2056 | movement. But they are not intelligent enough to handle multiple | |
2057 | scroll bars within a frame. Trying a generic mouse driver might help. | |
a933dad1 | 2058 | |
9dc15871 EZ |
2059 | ** Mail sent through Microsoft Exchange in some encodings appears to be |
2060 | mangled and is not seen correctly in Rmail or Gnus. We don't know | |
2061 | exactly what happens, but it isn't an Emacs problem in cases we've | |
2062 | seen. | |
a933dad1 | 2063 | |
9dc15871 EZ |
2064 | ** On MS-Windows, you cannot use the right-hand ALT key and the left-hand |
2065 | CTRL key together to type a Control-Meta character. | |
a933dad1 | 2066 | |
9dc15871 | 2067 | This is a consequence of a misfeature beyond Emacs's control. |
a933dad1 | 2068 | |
9dc15871 EZ |
2069 | Under Windows, the AltGr key on international keyboards generates key |
2070 | events with the modifiers Right-Alt and Left-Ctrl. Since Emacs cannot | |
2071 | distinguish AltGr from an explicit Right-Alt and Left-Ctrl | |
2072 | combination, whenever it sees Right-Alt and Left-Ctrl it assumes that | |
2073 | AltGr has been pressed. The variable `w32-recognize-altgr' can be set | |
2074 | to nil to tell Emacs that AltGr is really Ctrl and Alt. | |
a933dad1 | 2075 | |
9dc15871 | 2076 | ** Under some X-servers running on MS-Windows, Emacs' display is incorrect. |
a933dad1 | 2077 | |
9dc15871 EZ |
2078 | The symptoms are that Emacs does not completely erase blank areas of the |
2079 | screen during scrolling or some other screen operations (e.g., selective | |
2080 | display or when killing a region). M-x recenter will cause the screen | |
2081 | to be completely redisplayed and the "extra" characters will disappear. | |
a933dad1 | 2082 | |
9dc15871 EZ |
2083 | This is known to occur under Exceed 6, and possibly earlier versions |
2084 | as well; it is reportedly solved in version 6.2.0.16 and later. The | |
2085 | problem lies in the X-server settings. | |
a933dad1 | 2086 | |
9dc15871 EZ |
2087 | There are reports that you can solve the problem with Exceed by |
2088 | running `Xconfig' from within NT, choosing "X selection", then | |
2089 | un-checking the boxes "auto-copy X selection" and "auto-paste to X | |
2090 | selection". | |
a933dad1 | 2091 | |
9dc15871 EZ |
2092 | Of this does not work, please inform bug-gnu-emacs@gnu.org. Then |
2093 | please call support for your X-server and see if you can get a fix. | |
2094 | If you do, please send it to bug-gnu-emacs@gnu.org so we can list it | |
2095 | here. | |
a933dad1 | 2096 | |
9dc15871 | 2097 | * Build-time problems |
a933dad1 | 2098 | |
9dc15871 | 2099 | ** Configuration |
a933dad1 | 2100 | |
9dc15871 | 2101 | *** The `configure' script doesn't find the jpeg library. |
a933dad1 | 2102 | |
9dc15871 EZ |
2103 | There are reports that this happens on some systems because the linker |
2104 | by default only looks for shared libraries, but jpeg distribution by | |
2105 | default only installs a nonshared version of the library, `libjpeg.a'. | |
a933dad1 | 2106 | |
9dc15871 EZ |
2107 | If this is the problem, you can configure the jpeg library with the |
2108 | `--enable-shared' option and then rebuild libjpeg. This produces a | |
2109 | shared version of libjpeg, which you need to install. Finally, rerun | |
2110 | the Emacs configure script, which should now find the jpeg library. | |
2111 | Alternatively, modify the generated src/Makefile to link the .a file | |
2112 | explicitly, and edit src/config.h to define HAVE_JPEG. | |
a933dad1 | 2113 | |
9dc15871 | 2114 | ** Compilation |
a933dad1 | 2115 | |
9dc15871 | 2116 | *** Building Emacs over NFS fails with ``Text file busy''. |
a933dad1 | 2117 | |
9dc15871 EZ |
2118 | This was reported to happen when building Emacs on a GNU/Linux system |
2119 | (RedHat Linux 6.2) using a build directory automounted from Solaris | |
2120 | (SunOS 5.6) file server, but it might not be limited to that | |
2121 | configuration alone. Presumably, the NFS server doesn't commit the | |
2122 | files' data to disk quickly enough, and the Emacs executable file is | |
2123 | left ``busy'' for several seconds after Emacs has finished dumping | |
2124 | itself. This causes the subsequent commands which invoke the dumped | |
2125 | Emacs executable to fail with the above message. | |
a933dad1 | 2126 | |
9dc15871 EZ |
2127 | In some of these cases, a time skew between the NFS server and the |
2128 | machine where Emacs is built is detected and reported by GNU Make | |
2129 | (it says that some of the files have modification time in the future). | |
2130 | This might be a symptom of NFS-related problems. | |
a933dad1 | 2131 | |
9dc15871 EZ |
2132 | If the NFS server runs on Solaris, apply the Solaris patch 105379-05 |
2133 | (Sunos 5.6: /kernel/misc/nfssrv patch). If that doesn't work, or if | |
2134 | you have a different version of the OS or the NFS server, you can | |
2135 | force the NFS server to use 1KB blocks, which was reported to fix the | |
2136 | problem albeit at a price of slowing down file I/O. You can force 1KB | |
2137 | blocks by specifying the "-o rsize=1024,wsize=1024" options to the | |
2138 | `mount' command, or by adding ",rsize=1024,wsize=1024" to the mount | |
2139 | options in the appropriate system configuration file, such as | |
2140 | `/etc/auto.home'. | |
a933dad1 | 2141 | |
9dc15871 EZ |
2142 | Alternatively, when Make fails due to this problem, you could wait for |
2143 | a few seconds and then invoke Make again. In one particular case, | |
2144 | waiting for 10 or more seconds between the two Make invocations seemed | |
2145 | to work around the problem. | |
a933dad1 | 2146 | |
9dc15871 EZ |
2147 | Similar problems can happen if your machine NFS-mounts a directory |
2148 | onto itself. Suppose the Emacs sources live in `/usr/local/src' and | |
2149 | you are working on the host called `marvin'. Then an entry in the | |
2150 | `/etc/fstab' file like the following is asking for trouble: | |
a933dad1 | 2151 | |
9dc15871 | 2152 | marvin:/usr/local/src /usr/local/src ...options.omitted... |
a933dad1 | 2153 | |
9dc15871 | 2154 | The solution is to remove this line from `etc/fstab'. |
a933dad1 | 2155 | |
9dc15871 | 2156 | *** Building Emacs with GCC 2.9x fails in the `src' directory. |
a933dad1 | 2157 | |
9dc15871 EZ |
2158 | This may happen if you use a development version of GNU `cpp' from one |
2159 | of the GCC snapshots between Oct 2000 and Feb 2001, or from a released | |
2160 | version of GCC newer than 2.95.2 which was prepared around those | |
2161 | dates; similar problems were reported with some snapshots of GCC 3.1 | |
2162 | around Sep 30 2001. The preprocessor in those versions is | |
2163 | incompatible with a traditional Unix cpp (e.g., it expands ".." into | |
2164 | ". .", which breaks relative file names that reference the parent | |
2165 | directory; or inserts TAB characters before lines that set Make | |
2166 | variables). | |
a933dad1 | 2167 | |
9dc15871 EZ |
2168 | The solution is to make sure the preprocessor is run with the |
2169 | `-traditional' option. The `configure' script does that automatically | |
2170 | when it detects the known problems in your cpp, but you might hit some | |
2171 | unknown ones. To force the `configure' script to use `-traditional', | |
2172 | run the script like this: | |
a933dad1 | 2173 | |
9dc15871 | 2174 | CPP='gcc -E -traditional' ./configure ... |
a933dad1 | 2175 | |
9dc15871 EZ |
2176 | (replace the ellipsis "..." with any additional arguments you pass to |
2177 | the script). | |
a933dad1 | 2178 | |
9dc15871 EZ |
2179 | Note that this problem does not pertain to the MS-Windows port of |
2180 | Emacs, since it doesn't use the preprocessor to generate Makefiles. | |
a933dad1 | 2181 | |
9dc15871 EZ |
2182 | *** src/Makefile and lib-src/Makefile are truncated--most of the file missing. |
2183 | *** Compiling wakeup, in lib-src, says it can't make wakeup.c. | |
a933dad1 | 2184 | |
9dc15871 EZ |
2185 | This can happen if configure uses GNU sed version 2.03. That version |
2186 | had a bug. GNU sed version 2.05 works properly.To solve the | |
2187 | problem, install the current version of GNU Sed, then rerun Emacs's | |
2188 | configure script. | |
a933dad1 | 2189 | |
9dc15871 | 2190 | *** Compiling lib-src says there is no rule to make test-distrib.c. |
a933dad1 | 2191 | |
9dc15871 EZ |
2192 | This results from a bug in a VERY old version of GNU Sed. To solve |
2193 | the problem, install the current version of GNU Sed, then rerun | |
2194 | Emacs's configure script. | |
a933dad1 | 2195 | |
9dc15871 | 2196 | *** Building the MS-Windows port with Cygwin GCC can fail. |
a933dad1 | 2197 | |
9dc15871 EZ |
2198 | Emacs may not build using recent Cygwin builds of GCC, such as Cygwin |
2199 | version 1.1.8, using the default configure settings. It appears to be | |
2200 | necessary to specify the -mwin32 flag when compiling, and define | |
2201 | __MSVCRT__, like so: | |
a933dad1 | 2202 | |
9dc15871 | 2203 | configure --with-gcc --cflags -mwin32 --cflags -D__MSVCRT__ |
a933dad1 | 2204 | |
9dc15871 | 2205 | *** Building the MS-Windows port fails with a CreateProcess failure. |
a933dad1 | 2206 | |
9dc15871 EZ |
2207 | Some versions of mingw32 make on some versions of Windows do not seem |
2208 | to detect the shell correctly. Try "make SHELL=cmd.exe", or if that | |
2209 | fails, try running make from Cygwin bash instead. | |
a933dad1 | 2210 | |
9dc15871 | 2211 | *** Building the MS-Windows port with Leim fails in the `leim' directory. |
a933dad1 | 2212 | |
9dc15871 | 2213 | The error message might be something like this: |
a933dad1 | 2214 | |
9dc15871 EZ |
2215 | Converting d:/emacs-21.3/leim/CXTERM-DIC/4Corner.tit to quail-package... |
2216 | Invalid ENCODE: value in TIT dictionary | |
2217 | NMAKE : fatal error U1077: '"../src/obj-spd/i386/emacs.exe"' : return code | |
2218 | '0xffffffff' | |
2219 | Stop. | |
a933dad1 | 2220 | |
9dc15871 EZ |
2221 | This can happen if the Leim distribution is unpacked with a program |
2222 | which converts the `*.tit' files to DOS-style CR-LF text format. The | |
2223 | `*.tit' files in the leim/CXTERM-DIC directory require Unix-style line | |
2224 | endings to compile properly, because Emacs reads them without any code | |
2225 | or EOL conversions. | |
a933dad1 | 2226 | |
9dc15871 EZ |
2227 | The solution is to make sure the program used to unpack Leim does not |
2228 | change the files' line endings behind your back. The GNU FTP site has | |
2229 | in the `/gnu/emacs/windows' directory a program called `djtarnt.exe' | |
2230 | which can be used to unpack `.tar.gz' and `.zip' archives without | |
2231 | mangling them. | |
a933dad1 | 2232 | |
9dc15871 | 2233 | *** Building `ctags' for MS-Windows with the MinGW port of GCC fails. |
a933dad1 | 2234 | |
9dc15871 EZ |
2235 | This might happen due to a bug in the MinGW header assert.h, which |
2236 | defines the `assert' macro with a trailing semi-colon. The following | |
2237 | patch to assert.h should solve this: | |
a933dad1 | 2238 | |
9dc15871 EZ |
2239 | *** include/assert.h.orig Sun Nov 7 02:41:36 1999 |
2240 | --- include/assert.h Mon Jan 29 11:49:10 2001 | |
2241 | *************** | |
2242 | *** 41,47 **** | |
2243 | /* | |
2244 | * If not debugging, assert does nothing. | |
2245 | */ | |
2246 | ! #define assert(x) ((void)0); | |
a933dad1 | 2247 | |
9dc15871 | 2248 | #else /* debugging enabled */ |
a933dad1 | 2249 | |
9dc15871 EZ |
2250 | --- 41,47 ---- |
2251 | /* | |
2252 | * If not debugging, assert does nothing. | |
2253 | */ | |
2254 | ! #define assert(x) ((void)0) | |
a933dad1 | 2255 | |
9dc15871 | 2256 | #else /* debugging enabled */ |
a933dad1 | 2257 | |
a933dad1 | 2258 | |
9dc15871 | 2259 | ** Linking |
a933dad1 | 2260 | |
9dc15871 EZ |
2261 | *** Building Emacs with a system compiler fails to link because of an |
2262 | undefined symbol such as __eprintf which does not appear in Emacs. | |
a933dad1 | 2263 | |
9dc15871 EZ |
2264 | This can happen if some of the libraries linked into Emacs were built |
2265 | with GCC, but Emacs itself is being linked with a compiler other than | |
2266 | GCC. Object files compiled with GCC might need some helper functions | |
2267 | from libgcc.a, the library which comes with GCC, but the system | |
2268 | compiler does not instruct the linker to search libgcc.a during the | |
2269 | link stage. | |
a933dad1 | 2270 | |
9dc15871 | 2271 | A solution is to link with GCC, like this: |
a933dad1 | 2272 | |
9dc15871 | 2273 | make CC=gcc |
a933dad1 | 2274 | |
9dc15871 EZ |
2275 | Since the .o object files already exist, this will not recompile Emacs |
2276 | with GCC, but just restart by trying again to link temacs. | |
a933dad1 | 2277 | |
9dc15871 | 2278 | *** AIX 1.3 ptf 0013: Link failure. |
a933dad1 | 2279 | |
9dc15871 EZ |
2280 | There is a real duplicate definition of the function `_slibc_free' in |
2281 | the library /lib/libc_s.a (just do nm on it to verify). The | |
2282 | workaround/fix is: | |
a933dad1 | 2283 | |
9dc15871 EZ |
2284 | cd /lib |
2285 | ar xv libc_s.a NLtmtime.o | |
2286 | ar dv libc_s.a NLtmtime.o | |
a933dad1 | 2287 | |
9dc15871 EZ |
2288 | *** AIX 4.1.2: Linker error messages such as |
2289 | ld: 0711-212 SEVERE ERROR: Symbol .__quous, found in the global symbol table | |
2290 | of archive /usr/lib/libIM.a, was not defined in archive member shr.o. | |
a933dad1 | 2291 | |
9dc15871 EZ |
2292 | This is a problem in libIM.a. You can work around it by executing |
2293 | these shell commands in the src subdirectory of the directory where | |
2294 | you build Emacs: | |
a933dad1 | 2295 | |
9dc15871 EZ |
2296 | cp /usr/lib/libIM.a . |
2297 | chmod 664 libIM.a | |
2298 | ranlib libIM.a | |
a933dad1 | 2299 | |
9dc15871 EZ |
2300 | Then change -lIM to ./libIM.a in the command to link temacs (in |
2301 | Makefile). | |
a933dad1 | 2302 | |
9dc15871 | 2303 | *** Sun with acc: Link failure when using acc on a Sun. |
a933dad1 | 2304 | |
9dc15871 | 2305 | To use acc, you need additional options just before the libraries, such as |
a933dad1 | 2306 | |
9dc15871 | 2307 | /usr/lang/SC2.0.1/values-Xt.o -L/usr/lang/SC2.0.1/cg87 -L/usr/lang/SC2.0.1 |
a933dad1 | 2308 | |
9dc15871 | 2309 | and you need to add -lansi just before -lc. |
a933dad1 | 2310 | |
9dc15871 EZ |
2311 | The precise file names depend on the compiler version, so we |
2312 | cannot easily arrange to supply them. | |
a933dad1 | 2313 | |
9dc15871 | 2314 | *** Linking says that the functions insque and remque are undefined. |
a933dad1 | 2315 | |
9dc15871 | 2316 | Change oldXMenu/Makefile by adding insque.o to the variable OBJS. |
a933dad1 | 2317 | |
9dc15871 | 2318 | *** `tparam' reported as a multiply-defined symbol when linking with ncurses. |
a933dad1 | 2319 | |
9dc15871 EZ |
2320 | This problem results from an incompatible change in ncurses, in |
2321 | version 1.9.9e approximately. This version is unable to provide a | |
2322 | definition of tparm without also defining tparam. This is also | |
2323 | incompatible with Terminfo; as a result, the Emacs Terminfo support | |
2324 | does not work with this version of ncurses. | |
a933dad1 | 2325 | |
9dc15871 | 2326 | The fix is to install a newer version of ncurses, such as version 4.2. |
a933dad1 | 2327 | |
9dc15871 | 2328 | ** Dumping |
a933dad1 | 2329 | |
9dc15871 | 2330 | *** Linux: Segfault during `make bootstrap' under certain recent versions of the Linux kernel. |
a933dad1 | 2331 | |
9dc15871 | 2332 | With certain recent Linux kernels (like the one of Redhat Fedora Core |
1f02a4ba | 2333 | 1 and 2), the new "Exec-shield" functionality is enabled by default, which |
9dc15871 | 2334 | creates a different memory layout that breaks the emacs dumper. |
a933dad1 | 2335 | |
1f02a4ba JD |
2336 | Configure can overcome the problem of exec-shield if the architecture is |
2337 | x86 and the program setarch is present. On other architectures no | |
2338 | workaround is known. | |
2339 | ||
9dc15871 | 2340 | You can check the Exec-shield state like this: |
a933dad1 | 2341 | |
9dc15871 | 2342 | cat /proc/sys/kernel/exec-shield |
a933dad1 | 2343 | |
1f02a4ba | 2344 | It returns non-zero when Exec-shield is enabled, 0 otherwise. Please |
9dc15871 | 2345 | read your system documentation for more details on Exec-shield and |
1f02a4ba JD |
2346 | associated commands. Exec-shield can be turned off with this command: |
2347 | ||
2348 | echo "0" > /proc/sys/kernel/exec-shield | |
a933dad1 | 2349 | |
9dc15871 EZ |
2350 | When Exec-shield is enabled, building Emacs will segfault during the |
2351 | execution of this command: | |
a933dad1 | 2352 | |
1f02a4ba | 2353 | ./temacs --batch --load loadup [dump|bootstrap] |
177c0ea7 | 2354 | |
9dc15871 | 2355 | To work around this problem, it is necessary to temporarily disable |
1f02a4ba JD |
2356 | Exec-shield while building Emacs, or, on x86, by using the `setarch' |
2357 | command when running temacs like this: | |
2358 | ||
2359 | setarch i386 ./temacs --batch --load loadup [dump|bootstrap] | |
a933dad1 | 2360 | |
a933dad1 | 2361 | |
9dc15871 | 2362 | *** Fatal signal in the command temacs -l loadup inc dump. |
a933dad1 | 2363 | |
9dc15871 EZ |
2364 | This command is the final stage of building Emacs. It is run by the |
2365 | Makefile in the src subdirectory, or by build.com on VMS. | |
a933dad1 | 2366 | |
9dc15871 EZ |
2367 | It has been known to get fatal errors due to insufficient swapping |
2368 | space available on the machine. | |
a933dad1 | 2369 | |
9dc15871 EZ |
2370 | On 68000s, it has also happened because of bugs in the |
2371 | subroutine `alloca'. Verify that `alloca' works right, even | |
2372 | for large blocks (many pages). | |
a933dad1 | 2373 | |
9dc15871 EZ |
2374 | *** test-distrib says that the distribution has been clobbered. |
2375 | *** or, temacs prints "Command key out of range 0-127". | |
2376 | *** or, temacs runs and dumps emacs, but emacs totally fails to work. | |
2377 | *** or, temacs gets errors dumping emacs. | |
a933dad1 | 2378 | |
9dc15871 EZ |
2379 | This can be because the .elc files have been garbled. Do not be |
2380 | fooled by the fact that most of a .elc file is text: these are | |
2381 | binary files and can contain all 256 byte values. | |
a933dad1 | 2382 | |
9dc15871 EZ |
2383 | In particular `shar' cannot be used for transmitting GNU Emacs. |
2384 | It typically truncates "lines". What appear to be "lines" in | |
2385 | a binary file can of course be of any length. Even once `shar' | |
2386 | itself is made to work correctly, `sh' discards null characters | |
2387 | when unpacking the shell archive. | |
a933dad1 | 2388 | |
9dc15871 EZ |
2389 | I have also seen character \177 changed into \377. I do not know |
2390 | what transfer means caused this problem. Various network | |
2391 | file transfer programs are suspected of clobbering the high bit. | |
a933dad1 | 2392 | |
9dc15871 EZ |
2393 | If you have a copy of Emacs that has been damaged in its |
2394 | nonprinting characters, you can fix them: | |
a933dad1 | 2395 | |
9dc15871 EZ |
2396 | 1) Record the names of all the .elc files. |
2397 | 2) Delete all the .elc files. | |
2398 | 3) Recompile alloc.c with a value of PURESIZE twice as large. | |
2399 | (See puresize.h.) You might as well save the old alloc.o. | |
2400 | 4) Remake emacs. It should work now. | |
2401 | 5) Running emacs, do Meta-x byte-compile-file repeatedly | |
2402 | to recreate all the .elc files that used to exist. | |
2403 | You may need to increase the value of the variable | |
2404 | max-lisp-eval-depth to succeed in running the compiler interpreted | |
2405 | on certain .el files. 400 was sufficient as of last report. | |
2406 | 6) Reinstall the old alloc.o (undoing changes to alloc.c if any) | |
2407 | and remake temacs. | |
2408 | 7) Remake emacs. It should work now, with valid .elc files. | |
a933dad1 | 2409 | |
9dc15871 | 2410 | *** temacs prints "Pure Lisp storage exhausted". |
a933dad1 | 2411 | |
9dc15871 EZ |
2412 | This means that the Lisp code loaded from the .elc and .el |
2413 | files during temacs -l loadup inc dump took up more | |
2414 | space than was allocated. | |
a933dad1 | 2415 | |
9dc15871 EZ |
2416 | This could be caused by |
2417 | 1) adding code to the preloaded Lisp files | |
2418 | 2) adding more preloaded files in loadup.el | |
2419 | 3) having a site-init.el or site-load.el which loads files. | |
2420 | Note that ANY site-init.el or site-load.el is nonstandard; | |
2421 | if you have received Emacs from some other site | |
2422 | and it contains a site-init.el or site-load.el file, consider | |
2423 | deleting that file. | |
2424 | 4) getting the wrong .el or .elc files | |
2425 | (not from the directory you expected). | |
2426 | 5) deleting some .elc files that are supposed to exist. | |
2427 | This would cause the source files (.el files) to be | |
2428 | loaded instead. They take up more room, so you lose. | |
2429 | 6) a bug in the Emacs distribution which underestimates | |
2430 | the space required. | |
a933dad1 | 2431 | |
9dc15871 EZ |
2432 | If the need for more space is legitimate, change the definition |
2433 | of PURESIZE in puresize.h. | |
a933dad1 | 2434 | |
9dc15871 EZ |
2435 | But in some of the cases listed above, this problem is a consequence |
2436 | of something else that is wrong. Be sure to check and fix the real | |
2437 | problem. | |
a933dad1 | 2438 | |
9dc15871 | 2439 | *** Linux: Emacs crashes when dumping itself on Mac PPC running Yellow Dog GNU/Linux. |
a933dad1 | 2440 | |
9dc15871 EZ |
2441 | The crashes happen inside the function Fmake_symbol; here's a typical |
2442 | C backtrace printed by GDB: | |
a933dad1 | 2443 | |
9dc15871 EZ |
2444 | 0x190c0c0 in Fmake_symbol () |
2445 | (gdb) where | |
2446 | #0 0x190c0c0 in Fmake_symbol () | |
2447 | #1 0x1942ca4 in init_obarray () | |
2448 | #2 0x18b3500 in main () | |
2449 | #3 0x114371c in __libc_start_main (argc=5, argv=0x7ffff5b4, envp=0x7ffff5cc, | |
a933dad1 | 2450 | |
9dc15871 EZ |
2451 | This could happen because GCC version 2.95 and later changed the base |
2452 | of the load address to 0x10000000. Emacs needs to be told about this, | |
2453 | but we currently cannot do that automatically, because that breaks | |
2454 | other versions of GNU/Linux on the MacPPC. Until we find a way to | |
2455 | distinguish between the Yellow Dog and the other varieties of | |
2456 | GNU/Linux systems on the PPC, you will have to manually uncomment the | |
2457 | following section near the end of the file src/m/macppc.h in the Emacs | |
2458 | distribution: | |
a933dad1 | 2459 | |
9dc15871 EZ |
2460 | #if 0 /* This breaks things on PPC GNU/Linux except for Yellowdog, |
2461 | even with identical GCC, as, ld. Let's take it out until we | |
2462 | know what's really going on here. */ | |
2463 | /* GCC 2.95 and newer on GNU/Linux PPC changed the load address to | |
2464 | 0x10000000. */ | |
2465 | #if defined __linux__ | |
2466 | #if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 95) | |
2467 | #define DATA_SEG_BITS 0x10000000 | |
2468 | #endif | |
2469 | #endif | |
2470 | #endif /* 0 */ | |
a933dad1 | 2471 | |
9dc15871 EZ |
2472 | Remove the "#if 0" and "#endif" directives which surround this, save |
2473 | the file, and then reconfigure and rebuild Emacs. The dumping process | |
2474 | should now succeed. | |
a933dad1 | 2475 | |
9dc15871 | 2476 | ** Installation |
a933dad1 | 2477 | |
9dc15871 | 2478 | *** Installing Emacs gets an error running `install-info'. |
a933dad1 | 2479 | |
9dc15871 EZ |
2480 | You need to install a recent version of Texinfo; that package |
2481 | supplies the `install-info' command. | |
a933dad1 | 2482 | |
9dc15871 | 2483 | ** First execution |
a933dad1 | 2484 | |
9dc15871 | 2485 | *** Emacs binary is not in executable format, and cannot be run. |
a933dad1 | 2486 | |
9dc15871 EZ |
2487 | This was reported to happen when Emacs is built in a directory mounted |
2488 | via NFS, for some combinations of NFS client and NFS server. | |
2489 | Usually, the file `emacs' produced in these cases is full of | |
2490 | binary null characters, and the `file' utility says: | |
a933dad1 | 2491 | |
9dc15871 | 2492 | emacs: ASCII text, with no line terminators |
a933dad1 | 2493 | |
9dc15871 EZ |
2494 | We don't know what exactly causes this failure. A work-around is to |
2495 | build Emacs in a directory on a local disk. | |
a933dad1 | 2496 | |
9dc15871 | 2497 | *** The dumped Emacs crashes when run, trying to write pure data. |
a933dad1 | 2498 | |
9dc15871 | 2499 | Two causes have been seen for such problems. |
a933dad1 | 2500 | |
9dc15871 EZ |
2501 | 1) On a system where getpagesize is not a system call, it is defined |
2502 | as a macro. If the definition (in both unexec.c and malloc.c) is wrong, | |
2503 | it can cause problems like this. You might be able to find the correct | |
2504 | value in the man page for a.out (5). | |
a933dad1 | 2505 | |
9dc15871 EZ |
2506 | 2) Some systems allocate variables declared static among the |
2507 | initialized variables. Emacs makes all initialized variables in most | |
2508 | of its files pure after dumping, but the variables declared static and | |
2509 | not initialized are not supposed to be pure. On these systems you | |
2510 | may need to add "#define static" to the m- or the s- file. | |
a933dad1 | 2511 | |
9dc15871 | 2512 | * Emacs 19 problems |
a933dad1 | 2513 | |
9dc15871 | 2514 | ** Error messages `Wrong number of arguments: #<subr where-is-internal>, 5'. |
a933dad1 | 2515 | |
9dc15871 EZ |
2516 | This typically results from having the powerkey library loaded. |
2517 | Powerkey was designed for Emacs 19.22. It is obsolete now because | |
2518 | Emacs 19 now has this feature built in; and powerkey also calls | |
2519 | where-is-internal in an obsolete way. | |
a933dad1 | 2520 | |
9dc15871 | 2521 | So the fix is to arrange not to load powerkey. |
a933dad1 | 2522 | |
9dc15871 | 2523 | * Runtime problems on legacy systems |
a933dad1 | 2524 | |
9dc15871 EZ |
2525 | This section covers bugs reported on very old hardware or software. |
2526 | If you are using hardware and an operating system shipped after 2000, | |
2527 | it is unlikely you will see any of these. | |
a933dad1 | 2528 | |
9dc15871 | 2529 | ** Ancient operating systems |
a933dad1 | 2530 | |
0a4dd4e4 EZ |
2531 | AIX 4.2 was end-of-lifed on Dec 31st, 1999. |
2532 | ||
2533 | *** AIX: You get this compiler error message: | |
2534 | ||
2535 | Processing include file ./XMenuInt.h | |
2536 | 1501-106: (S) Include file X11/Xlib.h not found. | |
2537 | ||
2538 | This means your system was installed with only the X11 runtime i.d | |
2539 | libraries. You have to find your sipo (bootable tape) and install | |
2540 | X11Dev... with smit. | |
2541 | ||
2542 | (This report must be ancient. Bootable tapes are long dead.) | |
2543 | ||
2544 | *** AIX 3.2.4: Releasing Ctrl/Act key has no effect, if Shift is down. | |
2545 | ||
2546 | Due to a feature of AIX, pressing or releasing the Ctrl/Act key is | |
2547 | ignored when the Shift, Alt or AltGr keys are held down. This can | |
2548 | lead to the keyboard being "control-locked"--ordinary letters are | |
2549 | treated as control characters. | |
2550 | ||
2551 | You can get out of this "control-locked" state by pressing and | |
2552 | releasing Ctrl/Act while not pressing or holding any other keys. | |
2553 | ||
2554 | *** AIX 3.2.5: You get this message when running Emacs: | |
2555 | ||
2556 | Could not load program emacs | |
2557 | Symbol smtcheckinit in csh is undefined | |
2558 | Error was: Exec format error | |
2559 | ||
2560 | or this one: | |
2561 | ||
2562 | Could not load program .emacs | |
2563 | Symbol _system_con in csh is undefined | |
2564 | Symbol _fp_trapsta in csh is undefined | |
2565 | Error was: Exec format error | |
2566 | ||
2567 | These can happen when you try to run on AIX 3.2.5 a program that was | |
2568 | compiled with 3.2.4. The fix is to recompile. | |
2569 | ||
2570 | *** AIX 4.2: Emacs gets a segmentation fault at startup. | |
2571 | ||
2572 | If you are using IBM's xlc compiler, compile emacs.c | |
2573 | without optimization; that should avoid the problem. | |
2574 | ||
9dc15871 | 2575 | *** ISC Unix |
a933dad1 | 2576 | |
9dc15871 | 2577 | **** ISC: display-time causes kernel problems on ISC systems. |
a933dad1 | 2578 | |
9dc15871 EZ |
2579 | Under Interactive Unix versions 3.0.1 and 4.0 (and probably other |
2580 | versions), display-time causes the loss of large numbers of STREVENT | |
2581 | cells. Eventually the kernel's supply of these cells is exhausted. | |
2582 | This makes emacs and the whole system run slow, and can make other | |
2583 | processes die, in particular pcnfsd. | |
a933dad1 | 2584 | |
9dc15871 EZ |
2585 | Other emacs functions that communicate with remote processes may have |
2586 | the same problem. Display-time seems to be far the worst. | |
a933dad1 | 2587 | |
9dc15871 | 2588 | The only known fix: Don't run display-time. |
a933dad1 | 2589 | |
9dc15871 | 2590 | *** SunOS |
a933dad1 | 2591 | |
0a4dd4e4 EZ |
2592 | SunOS 4.1.4 stopped shipping on Sep 30 1998. |
2593 | ||
2594 | **** SunOS: You get linker errors | |
2595 | ld: Undefined symbol | |
2596 | _get_wmShellWidgetClass | |
2597 | _get_applicationShellWidgetClass | |
2598 | ||
9dc15871 | 2599 | **** Sun 4.0.x: M-x shell persistently reports "Process shell exited abnormally with code 1". |
a933dad1 | 2600 | |
9dc15871 EZ |
2601 | This happened on Suns as a result of what is said to be a bug in Sunos |
2602 | version 4.0.x. The only fix was to reboot the machine. | |
a933dad1 | 2603 | |
9dc15871 | 2604 | **** SunOS4.1.1 and SunOS4.1.3: Mail is lost when sent to local aliases. |
a933dad1 DL |
2605 | |
2606 | Many emacs mail user agents (VM and rmail, for instance) use the | |
2607 | sendmail.el library. This library can arrange for mail to be | |
2608 | delivered by passing messages to the /usr/lib/sendmail (usually) | |
2609 | program . In doing so, it passes the '-t' flag to sendmail, which | |
2610 | means that the name of the recipient of the message is not on the | |
2611 | command line and, therefore, that sendmail must parse the message to | |
2612 | obtain the destination address. | |
2613 | ||
2614 | There is a bug in the SunOS4.1.1 and SunOS4.1.3 versions of sendmail. | |
2615 | In short, when given the -t flag, the SunOS sendmail won't recognize | |
2616 | non-local (i.e. NIS) aliases. It has been reported that the Solaris | |
2617 | 2.x versions of sendmail do not have this bug. For those using SunOS | |
2618 | 4.1, the best fix is to install sendmail V8 or IDA sendmail (which | |
2619 | have other advantages over the regular sendmail as well). At the time | |
2620 | of this writing, these official versions are available: | |
2621 | ||
2622 | Sendmail V8 on ftp.cs.berkeley.edu in /ucb/sendmail: | |
2623 | sendmail.8.6.9.base.tar.Z (the base system source & documentation) | |
2624 | sendmail.8.6.9.cf.tar.Z (configuration files) | |
2625 | sendmail.8.6.9.misc.tar.Z (miscellaneous support programs) | |
2626 | sendmail.8.6.9.xdoc.tar.Z (extended documentation, with postscript) | |
2627 | ||
2628 | IDA sendmail on vixen.cso.uiuc.edu in /pub: | |
2629 | sendmail-5.67b+IDA-1.5.tar.gz | |
2630 | ||
9dc15871 | 2631 | **** Sunos 4: You get the error ld: Undefined symbol __lib_version. |
a933dad1 | 2632 | |
9dc15871 EZ |
2633 | This is the result of using cc or gcc with the shared library meant |
2634 | for acc (the Sunpro compiler). Check your LD_LIBRARY_PATH and delete | |
2635 | /usr/lang/SC2.0.1 or some similar directory. | |
a933dad1 | 2636 | |
9dc15871 | 2637 | **** SunOS 4.1.3: Emacs unpredictably crashes in _yp_dobind_soft. |
a933dad1 | 2638 | |
9dc15871 EZ |
2639 | This happens if you configure Emacs specifying just `sparc-sun-sunos4' |
2640 | on a system that is version 4.1.3. You must specify the precise | |
2641 | version number (or let configure figure out the configuration, which | |
2642 | it can do perfectly well for SunOS). | |
a933dad1 | 2643 | |
9dc15871 | 2644 | **** Sunos 4.1.3: Emacs gets hung shortly after startup. |
a933dad1 | 2645 | |
9dc15871 EZ |
2646 | We think this is due to a bug in Sunos. The word is that |
2647 | one of these Sunos patches fixes the bug: | |
a933dad1 | 2648 | |
9dc15871 EZ |
2649 | 100075-11 100224-06 100347-03 100482-05 100557-02 100623-03 100804-03 101080-01 |
2650 | 100103-12 100249-09 100496-02 100564-07 100630-02 100891-10 101134-01 | |
2651 | 100170-09 100296-04 100377-09 100507-04 100567-04 100650-02 101070-01 101145-01 | |
2652 | 100173-10 100305-15 100383-06 100513-04 100570-05 100689-01 101071-03 101200-02 | |
2653 | 100178-09 100338-05 100421-03 100536-02 100584-05 100784-01 101072-01 101207-01 | |
a933dad1 | 2654 | |
9dc15871 EZ |
2655 | We don't know which of these patches really matter. If you find out |
2656 | which ones, please inform bug-gnu-emacs@gnu.org. | |
a933dad1 | 2657 | |
9dc15871 EZ |
2658 | **** SunOS 4: Emacs processes keep going after you kill the X server |
2659 | (or log out, if you logged in using X). | |
a933dad1 | 2660 | |
9dc15871 | 2661 | Someone reported that recompiling with GCC 2.7.0 fixed this problem. |
a933dad1 | 2662 | |
9dc15871 EZ |
2663 | The fix to this is to install patch 100573 for OpenWindows 3.0 |
2664 | or link libXmu statically. | |
a933dad1 | 2665 | |
0a4dd4e4 EZ |
2666 | **** Sunos 5.3: Subprocesses remain, hanging but not zombies. |
2667 | ||
2668 | A bug in Sunos 5.3 causes Emacs subprocesses to remain after Emacs | |
2669 | exits. Sun patch # 101415-02 is part of the fix for this, but it only | |
2670 | applies to ptys, and doesn't fix the problem with subprocesses | |
2671 | communicating through pipes. | |
2672 | ||
9dc15871 | 2673 | *** Apollo Domain |
a933dad1 | 2674 | |
9dc15871 | 2675 | **** Shell mode ignores interrupts on Apollo Domain. |
a933dad1 | 2676 | |
9dc15871 | 2677 | You may find that M-x shell prints the following message: |
a933dad1 | 2678 | |
9dc15871 | 2679 | Warning: no access to tty; thus no job control in this shell... |
a933dad1 | 2680 | |
9dc15871 EZ |
2681 | This can happen if there are not enough ptys on your system. |
2682 | Here is how to make more of them. | |
a933dad1 | 2683 | |
9dc15871 EZ |
2684 | % cd /dev |
2685 | % ls pty* | |
2686 | # shows how many pty's you have. I had 8, named pty0 to pty7) | |
2687 | % /etc/crpty 8 | |
2688 | # creates eight new pty's | |
a933dad1 | 2689 | |
9dc15871 | 2690 | *** Irix |
a933dad1 | 2691 | |
9dc15871 | 2692 | *** Irix 6.2: No visible display on mips-sgi-irix6.2 when compiling with GCC 2.8.1. |
a933dad1 | 2693 | |
9dc15871 EZ |
2694 | This problem went away after installing the latest IRIX patches |
2695 | as of 8 Dec 1998. | |
a933dad1 | 2696 | |
9dc15871 | 2697 | The same problem has been reported on Irix 6.3. |
a933dad1 | 2698 | |
9dc15871 EZ |
2699 | *** Irix 6.3: substituting environment variables in file names |
2700 | in the minibuffer gives peculiar error messages such as | |
a933dad1 | 2701 | |
9dc15871 | 2702 | Substituting nonexistent environment variable "" |
a933dad1 | 2703 | |
9dc15871 EZ |
2704 | This is not an Emacs bug; it is caused by something in SGI patch |
2705 | 003082 August 11, 1998. | |
a933dad1 | 2706 | |
9dc15871 | 2707 | *** OPENSTEP |
a933dad1 | 2708 | |
9dc15871 | 2709 | **** OPENSTEP 4.2: Compiling syntax.c with gcc 2.7.2.1 fails. |
a933dad1 | 2710 | |
9dc15871 EZ |
2711 | The compiler was reported to crash while compiling syntax.c with the |
2712 | following message: | |
a933dad1 | 2713 | |
9dc15871 | 2714 | cc: Internal compiler error: program cc1obj got fatal signal 11 |
a933dad1 | 2715 | |
9dc15871 EZ |
2716 | To work around this, replace the macros UPDATE_SYNTAX_TABLE_FORWARD, |
2717 | INC_BOTH, and INC_FROM with functions. To this end, first define 3 | |
2718 | functions, one each for every macro. Here's an example: | |
a933dad1 | 2719 | |
9dc15871 EZ |
2720 | static int update_syntax_table_forward(int from) |
2721 | { | |
2722 | return(UPDATE_SYNTAX_TABLE_FORWARD(from)); | |
2723 | }/*update_syntax_table_forward*/ | |
a933dad1 | 2724 | |
9dc15871 EZ |
2725 | Then replace all references to UPDATE_SYNTAX_TABLE_FORWARD in syntax.c |
2726 | with a call to the function update_syntax_table_forward. | |
a933dad1 | 2727 | |
9dc15871 | 2728 | *** Solaris 2.x |
a933dad1 | 2729 | |
9dc15871 | 2730 | **** Strange results from format %d in a few cases, on a Sun. |
a933dad1 | 2731 | |
9dc15871 EZ |
2732 | Sun compiler version SC3.0 has been found to miscompile part of |
2733 | editfns.c. The workaround is to compile with some other compiler such | |
2734 | as GCC. | |
a933dad1 | 2735 | |
9dc15871 | 2736 | **** On Solaris, Emacs dumps core if lisp-complete-symbol is called. |
a933dad1 | 2737 | |
9dc15871 EZ |
2738 | If you compile Emacs with the -fast or -xO4 option with version 3.0.2 |
2739 | of the Sun C compiler, Emacs dumps core when lisp-complete-symbol is | |
2740 | called. The problem does not happen if you compile with GCC. | |
a933dad1 | 2741 | |
9dc15871 | 2742 | **** On Solaris, Emacs crashes if you use (display-time). |
a933dad1 | 2743 | |
9dc15871 EZ |
2744 | This can happen if you configure Emacs without specifying the precise |
2745 | version of Solaris that you are using. | |
a933dad1 | 2746 | |
9dc15871 | 2747 | **** Solaris 2.3 and 2.4: Unpredictable segmentation faults. |
a933dad1 | 2748 | |
9dc15871 EZ |
2749 | A user reported that this happened in 19.29 when it was compiled with |
2750 | the Sun compiler, but not when he recompiled with GCC 2.7.0. | |
a933dad1 | 2751 | |
9dc15871 | 2752 | We do not know whether something in Emacs is partly to blame for this. |
a933dad1 | 2753 | |
9dc15871 | 2754 | **** Solaris 2.4: Emacs dumps core on startup. |
a933dad1 | 2755 | |
9dc15871 EZ |
2756 | Bill Sebok says that the cause of this is Solaris 2.4 vendor patch |
2757 | 102303-05, which extends the Solaris linker to deal with the Solaris | |
2758 | Common Desktop Environment's linking needs. You can fix the problem | |
2759 | by removing this patch and installing patch 102049-02 instead. | |
2760 | However, that linker version won't work with CDE. | |
a933dad1 | 2761 | |
9dc15871 EZ |
2762 | Solaris 2.5 comes with a linker that has this bug. It is reported that if |
2763 | you install all the latest patches (as of June 1996), the bug is fixed. | |
2764 | We suspect the crucial patch is one of these, but we don't know | |
2765 | for certain. | |
a933dad1 | 2766 | |
9dc15871 EZ |
2767 | 103093-03: [README] SunOS 5.5: kernel patch (2140557 bytes) |
2768 | 102832-01: [README] OpenWindows 3.5: Xview Jumbo Patch (4181613 bytes) | |
2769 | 103242-04: [README] SunOS 5.5: linker patch (595363 bytes) | |
a933dad1 | 2770 | |
9dc15871 EZ |
2771 | (One user reports that the bug was fixed by those patches together |
2772 | with patches 102980-04, 103279-01, 103300-02, and 103468-01.) | |
a933dad1 | 2773 | |
9dc15871 EZ |
2774 | If you can determine which patch does fix the bug, please tell |
2775 | bug-gnu-emacs@gnu.org. | |
a933dad1 | 2776 | |
9dc15871 EZ |
2777 | Meanwhile, the GNU linker links Emacs properly on both Solaris 2.4 and |
2778 | Solaris 2.5. | |
a933dad1 | 2779 | |
9dc15871 EZ |
2780 | **** Solaris 2.4: Dired hangs and C-g does not work. Or Emacs hangs |
2781 | forever waiting for termination of a subprocess that is a zombie. | |
a933dad1 | 2782 | |
9dc15871 EZ |
2783 | casper@fwi.uva.nl says the problem is in X11R6. Rebuild libX11.so |
2784 | after changing the file xc/config/cf/sunLib.tmpl. Change the lines | |
a933dad1 | 2785 | |
9dc15871 EZ |
2786 | #if ThreadedX |
2787 | #define SharedX11Reqs -lthread | |
2788 | #endif | |
a933dad1 | 2789 | |
9dc15871 | 2790 | to: |
a933dad1 | 2791 | |
9dc15871 EZ |
2792 | #if OSMinorVersion < 4 |
2793 | #if ThreadedX | |
2794 | #define SharedX11Reqs -lthread | |
2795 | #endif | |
2796 | #endif | |
2797 | ||
2798 | Be sure also to edit x/config/cf/sun.cf so that OSMinorVersion is 4 | |
2799 | (as it should be for Solaris 2.4). The file has three definitions for | |
2800 | OSMinorVersion: the first is for x86, the second for SPARC under | |
2801 | Solaris, and the third for SunOS 4. Make sure to update the | |
2802 | definition for your type of machine and system. | |
a933dad1 | 2803 | |
9dc15871 EZ |
2804 | Then do `make Everything' in the top directory of X11R6, to rebuild |
2805 | the makefiles and rebuild X. The X built this way work only on | |
2806 | Solaris 2.4, not on 2.3. | |
a933dad1 | 2807 | |
9dc15871 EZ |
2808 | For multithreaded X to work it is necessary to install patch |
2809 | 101925-02 to fix problems in header files [2.4]. You need | |
2810 | to reinstall gcc or re-run just-fixinc after installing that | |
2811 | patch. | |
a933dad1 | 2812 | |
9dc15871 EZ |
2813 | However, Frank Rust <frust@iti.cs.tu-bs.de> used a simpler solution: |
2814 | he changed | |
2815 | #define ThreadedX YES | |
2816 | to | |
2817 | #define ThreadedX NO | |
2818 | in sun.cf and did `make World' to rebuild X11R6. Removing all | |
2819 | `-DXTHREAD*' flags and `-lthread' entries from lib/X11/Makefile and | |
2820 | typing 'make install' in that directory also seemed to work. | |
a933dad1 | 2821 | |
9dc15871 | 2822 | **** Solaris 2.x: GCC complains "64 bit integer types not supported". |
a933dad1 | 2823 | |
9dc15871 EZ |
2824 | This suggests that GCC is not installed correctly. Most likely you |
2825 | are using GCC 2.7.2.3 (or earlier) on Solaris 2.6 (or later); this | |
2826 | does not work without patching. To run GCC 2.7.2.3 on Solaris 2.6 or | |
2827 | later, you must patch fixinc.svr4 and reinstall GCC from scratch as | |
2828 | described in the Solaris FAQ | |
2829 | <http://www.wins.uva.nl/pub/solaris/solaris2.html>. A better fix is | |
2830 | to upgrade to GCC 2.8.1 or later. | |
a933dad1 | 2831 | |
9dc15871 EZ |
2832 | **** Solaris 2.7: Building Emacs with WorkShop Compilers 5.0 98/12/15 |
2833 | C 5.0 failed, apparently with non-default CFLAGS, most probably due to | |
2834 | compiler bugs. Using Sun Solaris 2.7 Sun WorkShop 6 update 1 C | |
2835 | release was reported to work without problems. It worked OK on | |
2836 | another system with Solaris 8 using apparently the same 5.0 compiler | |
2837 | and the default CFLAGS. | |
a933dad1 | 2838 | |
9dc15871 | 2839 | **** Solaris 2.x: Emacs dumps core when built with Motif. |
a933dad1 | 2840 | |
9dc15871 EZ |
2841 | The Solaris Motif libraries are buggy, at least up through Solaris 2.5.1. |
2842 | Install the current Motif runtime library patch appropriate for your host. | |
2843 | (Make sure the patch is current; some older patch versions still have the bug.) | |
2844 | You should install the other patches recommended by Sun for your host, too. | |
2845 | You can obtain Sun patches from ftp://sunsolve.sun.com/pub/patches/; | |
2846 | look for files with names ending in `.PatchReport' to see which patches | |
2847 | are currently recommended for your host. | |
a933dad1 | 2848 | |
9dc15871 EZ |
2849 | On Solaris 2.6, Emacs is said to work with Motif when Solaris patch |
2850 | 105284-12 is installed, but fail when 105284-15 is installed. | |
2851 | 105284-18 might fix it again. | |
a933dad1 | 2852 | |
0a4dd4e4 | 2853 | **** Solaris 2.6 and 7: the Compose key does not work. |
a933dad1 | 2854 | |
9dc15871 EZ |
2855 | This is a bug in Motif in Solaris. Supposedly it has been fixed for |
2856 | the next major release of Solaris. However, if someone with Sun | |
2857 | support complains to Sun about the bug, they may release a patch. | |
2858 | If you do this, mention Sun bug #4188711. | |
a933dad1 | 2859 | |
9dc15871 EZ |
2860 | One workaround is to use a locale that allows non-ASCII characters. |
2861 | For example, before invoking emacs, set the LC_ALL environment | |
2862 | variable to "en_US" (American English). The directory /usr/lib/locale | |
2863 | lists the supported locales; any locale other than "C" or "POSIX" | |
2864 | should do. | |
a933dad1 | 2865 | |
9dc15871 EZ |
2866 | pen@lysator.liu.se says (Feb 1998) that the Compose key does work |
2867 | if you link with the MIT X11 libraries instead of the Solaris X11 | |
2868 | libraries. | |
a933dad1 | 2869 | |
0a4dd4e4 EZ |
2870 | *** HP/UX versions before 11.0 |
2871 | ||
bf247b6e | 2872 | HP/UX 9 was end-of-lifed in December 1998. |
0a4dd4e4 EZ |
2873 | HP/UX 10 was end-of-lifed in May 1999. |
2874 | ||
2875 | **** HP/UX 9: Emacs crashes with SIGBUS or SIGSEGV after you delete a frame. | |
2876 | ||
2877 | We think this is due to a bug in the X libraries provided by HP. With | |
2878 | the alternative X libraries in /usr/contrib/mitX11R5/lib, the problem | |
2879 | does not happen. | |
2880 | ||
2881 | *** HP/UX 10: Large file support is disabled. | |
2882 | ||
2883 | See the comments in src/s/hpux10.h. | |
2884 | ||
2885 | *** HP/UX: Emacs is slow using X11R5. | |
2886 | ||
2887 | This happens if you use the MIT versions of the X libraries--it | |
2888 | doesn't run as fast as HP's version. People sometimes use the version | |
2889 | because they see the HP version doesn't have the libraries libXaw.a, | |
2890 | libXmu.a, libXext.a and others. HP/UX normally doesn't come with | |
2891 | those libraries installed. To get good performance, you need to | |
2892 | install them and rebuild Emacs. | |
2893 | ||
9dc15871 | 2894 | *** Ultrix and Digital Unix |
a933dad1 | 2895 | |
9dc15871 | 2896 | **** Ultrix 4.2: `make install' fails on install-doc with `Error 141'. |
a933dad1 | 2897 | |
9dc15871 EZ |
2898 | This happens on Ultrix 4.2 due to failure of a pipeline of tar |
2899 | commands. We don't know why they fail, but the bug seems not to be in | |
2900 | Emacs. The workaround is to run the shell command in install-doc by | |
2901 | hand. | |
a933dad1 | 2902 | |
9dc15871 | 2903 | **** Digital Unix 4.0: Garbled display on non-X terminals when Emacs runs. |
a933dad1 | 2904 | |
9dc15871 EZ |
2905 | So far it appears that running `tset' triggers this problem (when TERM |
2906 | is vt100, at least). If you do not run `tset', then Emacs displays | |
2907 | properly. If someone can tell us precisely which effect of running | |
2908 | `tset' actually causes the problem, we may be able to implement a fix | |
2909 | in Emacs. | |
a933dad1 | 2910 | |
9dc15871 | 2911 | **** Ultrix: `expand-file-name' fails to work on any but the machine you dumped Emacs on. |
a933dad1 | 2912 | |
9dc15871 EZ |
2913 | On Ultrix, if you use any of the functions which look up information |
2914 | in the passwd database before dumping Emacs (say, by using | |
2915 | expand-file-name in site-init.el), then those functions will not work | |
2916 | in the dumped Emacs on any host but the one Emacs was dumped on. | |
a933dad1 | 2917 | |
9dc15871 EZ |
2918 | The solution? Don't use expand-file-name in site-init.el, or in |
2919 | anything it loads. Yuck - some solution. | |
a933dad1 | 2920 | |
9dc15871 EZ |
2921 | I'm not sure why this happens; if you can find out exactly what is |
2922 | going on, and perhaps find a fix or a workaround, please let us know. | |
2923 | Perhaps the YP functions cache some information, the cache is included | |
2924 | in the dumped Emacs, and is then inaccurate on any other host. | |
a933dad1 | 2925 | |
9dc15871 | 2926 | *** SVr4 |
a933dad1 | 2927 | |
9dc15871 | 2928 | **** SVr4: On some variants of SVR4, Emacs does not work at all with X. |
a933dad1 | 2929 | |
9dc15871 EZ |
2930 | Try defining BROKEN_FIONREAD in your config.h file. If this solves |
2931 | the problem, please send a bug report to tell us this is needed; be | |
2932 | sure to say exactly what type of machine and system you are using. | |
a933dad1 | 2933 | |
9dc15871 | 2934 | **** SVr4: After running emacs once, subsequent invocations crash. |
a933dad1 | 2935 | |
9dc15871 EZ |
2936 | Some versions of SVR4 have a serious bug in the implementation of the |
2937 | mmap () system call in the kernel; this causes emacs to run correctly | |
2938 | the first time, and then crash when run a second time. | |
a933dad1 | 2939 | |
9dc15871 EZ |
2940 | Contact your vendor and ask for the mmap bug fix; in the mean time, |
2941 | you may be able to work around the problem by adding a line to your | |
2942 | operating system description file (whose name is reported by the | |
2943 | configure script) that reads: | |
2944 | #define SYSTEM_MALLOC | |
2945 | This makes Emacs use memory less efficiently, but seems to work around | |
2946 | the kernel bug. | |
a933dad1 | 2947 | |
0a4dd4e4 EZ |
2948 | *** Irix 5 and earlier |
2949 | ||
bf247b6e | 2950 | Exactly when Irix-5 end-of-lifed is obscure. But since Irix 6.0 |
0a4dd4e4 EZ |
2951 | shipped in 1994, it has been some years. |
2952 | ||
2953 | **** Irix 5.2: unexelfsgi.c can't find cmplrs/stsupport.h. | |
2954 | ||
2955 | The file cmplrs/stsupport.h was included in the wrong file set in the | |
2956 | Irix 5.2 distribution. You can find it in the optional fileset | |
2957 | compiler_dev, or copy it from some other Irix 5.2 system. A kludgy | |
2958 | workaround is to change unexelfsgi.c to include sym.h instead of | |
2959 | syms.h. | |
2960 | ||
2961 | **** Irix 5.3: "out of virtual swap space". | |
2962 | ||
2963 | This message occurs when the system runs out of swap space due to too | |
2964 | many large programs running. The solution is either to provide more | |
2965 | swap space or to reduce the number of large programs being run. You | |
2966 | can check the current status of the swap space by executing the | |
2967 | command `swap -l'. | |
2968 | ||
2969 | You can increase swap space by changing the file /etc/fstab. Adding a | |
2970 | line like this: | |
2971 | ||
2972 | /usr/swap/swap.more swap swap pri=3 0 0 | |
2973 | ||
2974 | where /usr/swap/swap.more is a file previously created (for instance | |
2975 | by using /etc/mkfile), will increase the swap space by the size of | |
2976 | that file. Execute `swap -m' or reboot the machine to activate the | |
2977 | new swap area. See the manpages for `swap' and `fstab' for further | |
2978 | information. | |
2979 | ||
2980 | The objectserver daemon can use up lots of memory because it can be | |
2981 | swamped with NIS information. It collects information about all users | |
2982 | on the network that can log on to the host. | |
2983 | ||
2984 | If you want to disable the objectserver completely, you can execute | |
2985 | the command `chkconfig objectserver off' and reboot. That may disable | |
2986 | some of the window system functionality, such as responding CDROM | |
2987 | icons. | |
2988 | ||
2989 | You can also remove NIS support from the objectserver. The SGI `admin' | |
2990 | FAQ has a detailed description on how to do that; see question 35 | |
2991 | ("Why isn't the objectserver working?"). The admin FAQ can be found at | |
2992 | ftp://viz.tamu.edu/pub/sgi/faq/. | |
2993 | ||
2994 | **** Irix 5.3: Emacs crashes in utmpname. | |
2995 | ||
2996 | This problem is fixed in Patch 3175 for Irix 5.3. | |
2997 | It is also fixed in Irix versions 6.2 and up. | |
2998 | ||
2999 | **** Irix 6.0: Make tries (and fails) to build a program named unexelfsgi. | |
3000 | ||
3001 | A compiler bug inserts spaces into the string "unexelfsgi . o" | |
3002 | in src/Makefile. Edit src/Makefile, after configure is run, | |
3003 | find that string, and take out the spaces. | |
3004 | ||
3005 | Compiler fixes in Irix 6.0.1 should eliminate this problem. | |
3006 | ||
3007 | *** SCO Unix and UnixWare | |
3008 | ||
3009 | **** SCO 3.2v4: Unusable default font. | |
3010 | ||
3011 | The Open Desktop environment comes with default X resource settings | |
3012 | that tell Emacs to use a variable-width font. Emacs cannot use such | |
3013 | fonts, so it does not work. | |
3014 | ||
3015 | This is caused by the file /usr/lib/X11/app-defaults/ScoTerm, which is | |
3016 | the application-specific resource file for the `scoterm' terminal | |
3017 | emulator program. It contains several extremely general X resources | |
3018 | that affect other programs besides `scoterm'. In particular, these | |
3019 | resources affect Emacs also: | |
3020 | ||
3021 | *Font: -*-helvetica-medium-r-*--12-*-p-* | |
3022 | *Background: scoBackground | |
3023 | *Foreground: scoForeground | |
3024 | ||
3025 | The best solution is to create an application-specific resource file for | |
3026 | Emacs, /usr/lib/X11/sco/startup/Emacs, with the following contents: | |
3027 | ||
3028 | Emacs*Font: -*-courier-medium-r-*-*-*-120-*-*-*-*-iso8859-1 | |
3029 | Emacs*Background: white | |
3030 | Emacs*Foreground: black | |
3031 | ||
3032 | (These settings mimic the Emacs defaults, but you can change them to | |
3033 | suit your needs.) This resource file is only read when the X server | |
3034 | starts up, so you should restart it by logging out of the Open Desktop | |
3035 | environment or by running `scologin stop; scologin start` from the shell | |
3036 | as root. Alternatively, you can put these settings in the | |
3037 | /usr/lib/X11/app-defaults/Emacs resource file and simply restart Emacs, | |
3038 | but then they will not affect remote invocations of Emacs that use the | |
3039 | Open Desktop display. | |
3040 | ||
3041 | These resource files are not normally shared across a network of SCO | |
3042 | machines; you must create the file on each machine individually. | |
3043 | ||
3044 | **** SCO 4.2.0: Regular expressions matching bugs on SCO systems. | |
3045 | ||
3046 | On SCO, there are problems in regexp matching when Emacs is compiled | |
3047 | with the system compiler. The compiler version is "Microsoft C | |
3048 | version 6", SCO 4.2.0h Dev Sys Maintenance Supplement 01/06/93; Quick | |
3049 | C Compiler Version 1.00.46 (Beta). The solution is to compile with | |
3050 | GCC. | |
3051 | ||
3052 | **** UnixWare 2.1: Error 12 (virtual memory exceeded) when dumping Emacs. | |
3053 | ||
3054 | Paul Abrahams (abrahams@acm.org) reports that with the installed | |
3055 | virtual memory settings for UnixWare 2.1.2, an Error 12 occurs during | |
3056 | the "make" that builds Emacs, when running temacs to dump emacs. That | |
3057 | error indicates that the per-process virtual memory limit has been | |
3058 | exceeded. The default limit is probably 32MB. Raising the virtual | |
3059 | memory limit to 40MB should make it possible to finish building Emacs. | |
3060 | ||
3061 | You can do this with the command `ulimit' (sh) or `limit' (csh). | |
3062 | But you have to be root to do it. | |
3063 | ||
3064 | According to Martin Sohnius, you can also retune this in the kernel: | |
3065 | ||
3066 | # /etc/conf/bin/idtune SDATLIM 33554432 ## soft data size limit | |
3067 | # /etc/conf/bin/idtune HDATLIM 33554432 ## hard " | |
3068 | # /etc/conf/bin/idtune SVMMSIZE unlimited ## soft process size limit | |
3069 | # /etc/conf/bin/idtune HVMMSIZE unlimited ## hard " | |
3070 | # /etc/conf/bin/idbuild -B | |
3071 | ||
3072 | (He recommends you not change the stack limit, though.) | |
3073 | These changes take effect when you reboot. | |
3074 | ||
9dc15871 | 3075 | *** Linux 1.x |
a933dad1 | 3076 | |
9dc15871 | 3077 | **** Linux 1.0-1.04: Typing C-c C-c in Shell mode kills your X server. |
a933dad1 | 3078 | |
9dc15871 EZ |
3079 | This happens with Linux kernel 1.0 thru 1.04, approximately. The workaround is |
3080 | to define SIGNALS_VIA_CHARACTERS in config.h and recompile Emacs. | |
3081 | Newer Linux kernel versions don't have this problem. | |
a933dad1 | 3082 | |
9dc15871 EZ |
3083 | **** Linux 1.3: Output from subprocess (such as man or diff) is randomly |
3084 | truncated on GNU/Linux systems. | |
a933dad1 | 3085 | |
9dc15871 EZ |
3086 | This is due to a kernel bug which seems to be fixed in Linux version |
3087 | 1.3.75. | |
a933dad1 | 3088 | |
0a4dd4e4 EZ |
3089 | ** Windows 3.1, 95, 98, and ME |
3090 | ||
3091 | *** MS-Windows NT/95: Problems running Perl under Emacs | |
3092 | ||
3093 | `perl -de 0' just hangs when executed in an Emacs subshell. | |
3094 | The fault lies with Perl (indirectly with Windows NT/95). | |
3095 | ||
3096 | The problem is that the Perl debugger explicitly opens a connection to | |
3097 | "CON", which is the DOS/NT equivalent of "/dev/tty", for interacting | |
3098 | with the user. | |
3099 | ||
3100 | On Unix, this is okay, because Emacs (or the shell?) creates a | |
3101 | pseudo-tty so that /dev/tty is really the pipe Emacs is using to | |
3102 | communicate with the subprocess. | |
3103 | ||
3104 | On NT, this fails because CON always refers to the handle for the | |
3105 | relevant console (approximately equivalent to a tty), and cannot be | |
3106 | redirected to refer to the pipe Emacs assigned to the subprocess as | |
3107 | stdin. | |
3108 | ||
3109 | A workaround is to modify perldb.pl to use STDIN/STDOUT instead of CON. | |
3110 | ||
3111 | For Perl 4: | |
3112 | ||
3113 | *** PERL/LIB/PERLDB.PL.orig Wed May 26 08:24:18 1993 | |
3114 | --- PERL/LIB/PERLDB.PL Mon Jul 01 15:28:16 1996 | |
3115 | *************** | |
3116 | *** 68,74 **** | |
3117 | $rcfile=".perldb"; | |
3118 | } | |
3119 | else { | |
3120 | ! $console = "con"; | |
3121 | $rcfile="perldb.ini"; | |
3122 | } | |
3123 | ||
3124 | --- 68,74 ---- | |
3125 | $rcfile=".perldb"; | |
3126 | } | |
3127 | else { | |
3128 | ! $console = ""; | |
3129 | $rcfile="perldb.ini"; | |
3130 | } | |
3131 | ||
3132 | ||
3133 | For Perl 5: | |
3134 | *** perl/5.001/lib/perl5db.pl.orig Sun Jun 04 21:13:40 1995 | |
3135 | --- perl/5.001/lib/perl5db.pl Mon Jul 01 17:00:08 1996 | |
3136 | *************** | |
3137 | *** 22,28 **** | |
3138 | $rcfile=".perldb"; | |
3139 | } | |
3140 | elsif (-e "con") { | |
3141 | ! $console = "con"; | |
3142 | $rcfile="perldb.ini"; | |
3143 | } | |
3144 | else { | |
3145 | --- 22,28 ---- | |
3146 | $rcfile=".perldb"; | |
3147 | } | |
3148 | elsif (-e "con") { | |
3149 | ! $console = ""; | |
3150 | $rcfile="perldb.ini"; | |
3151 | } | |
3152 | else { | |
3153 | ||
3154 | *** MS-Windows 95: Alt-f6 does not get through to Emacs. | |
3155 | ||
3156 | This character seems to be trapped by the kernel in Windows 95. | |
3157 | You can enter M-f6 by typing ESC f6. | |
3158 | ||
3159 | *** MS-Windows 95/98/ME: subprocesses do not terminate properly. | |
3160 | ||
3161 | This is a limitation of the Operating System, and can cause problems | |
3162 | when shutting down Windows. Ensure that all subprocesses are exited | |
3163 | cleanly before exiting Emacs. For more details, see the FAQ at | |
3164 | http://www.gnu.org/software/emacs/windows/. | |
3165 | ||
3166 | *** MS-Windows 95/98/ME: crashes when Emacs invokes non-existent programs. | |
3167 | ||
3168 | When a program you are trying to run is not found on the PATH, | |
3169 | Windows might respond by crashing or locking up your system. In | |
3170 | particular, this has been reported when trying to compile a Java | |
3171 | program in JDEE when javac.exe is installed, but not on the system | |
3172 | PATH. | |
3173 | ||
9dc15871 | 3174 | ** MS-DOS |
a933dad1 | 3175 | |
9dc15871 | 3176 | *** When compiling with DJGPP on MS-Windows NT, "config msdos" fails. |
a933dad1 | 3177 | |
9dc15871 EZ |
3178 | If the error message is "VDM has been already loaded", this is because |
3179 | Windows has a program called `redir.exe' that is incompatible with a | |
3180 | program by the same name supplied with DJGPP, which is used by | |
3181 | config.bat. To resolve this, move the DJGPP's `bin' subdirectory to | |
3182 | the front of your PATH environment variable. | |
a933dad1 | 3183 | |
9dc15871 EZ |
3184 | *** When compiling with DJGPP on MS-Windows 95, Make fails for some targets |
3185 | like make-docfile. | |
a933dad1 | 3186 | |
9dc15871 EZ |
3187 | This can happen if long file name support (the setting of environment |
3188 | variable LFN) when Emacs distribution was unpacked and during | |
3189 | compilation are not the same. See the MSDOG section of INSTALL for | |
3190 | the explanation of how to avoid this problem. | |
a933dad1 | 3191 | |
9dc15871 | 3192 | *** Emacs compiled with DJGPP complains at startup: |
a933dad1 | 3193 | |
9dc15871 | 3194 | "Wrong type of argument: internal-facep, msdos-menu-active-face" |
a933dad1 | 3195 | |
9dc15871 EZ |
3196 | This can happen if you define an environment variable `TERM'. Emacs |
3197 | on MSDOS uses an internal terminal emulator which is disabled if the | |
3198 | value of `TERM' is anything but the string "internal". Emacs then | |
3199 | works as if its terminal were a dumb glass teletype that doesn't | |
3200 | support faces. To work around this, arrange for `TERM' to be | |
3201 | undefined when Emacs runs. The best way to do that is to add an | |
3202 | [emacs] section to the DJGPP.ENV file which defines an empty value for | |
3203 | `TERM'; this way, only Emacs gets the empty value, while the rest of | |
3204 | your system works as before. | |
a933dad1 | 3205 | |
9dc15871 | 3206 | *** MS-DOS: Emacs crashes at startup. |
a933dad1 | 3207 | |
9dc15871 EZ |
3208 | Some users report that Emacs 19.29 requires dpmi memory management, |
3209 | and crashes on startup if the system does not have it. We don't yet | |
3210 | know why this happens--perhaps these machines don't have enough real | |
3211 | memory, or perhaps something is wrong in Emacs or the compiler. | |
3212 | However, arranging to use dpmi support is a workaround. | |
a933dad1 | 3213 | |
9dc15871 EZ |
3214 | You can find out if you have a dpmi host by running go32 without |
3215 | arguments; it will tell you if it uses dpmi memory. For more | |
3216 | information about dpmi memory, consult the djgpp FAQ. (djgpp | |
3217 | is the GNU C compiler as packaged for MSDOS.) | |
a933dad1 | 3218 | |
9dc15871 EZ |
3219 | Compiling Emacs under MSDOS is extremely sensitive for proper memory |
3220 | configuration. If you experience problems during compilation, consider | |
3221 | removing some or all memory resident programs (notably disk caches) | |
3222 | and make sure that your memory managers are properly configured. See | |
3223 | the djgpp faq for configuration hints. | |
a933dad1 | 3224 | |
9dc15871 EZ |
3225 | *** Emacs compiled with DJGPP for MS-DOS/MS-Windows cannot access files |
3226 | in the directory with the special name `dev' under the root of any | |
3227 | drive, e.g. `c:/dev'. | |
a933dad1 | 3228 | |
9dc15871 EZ |
3229 | This is an unfortunate side-effect of the support for Unix-style |
3230 | device names such as /dev/null in the DJGPP runtime library. A | |
3231 | work-around is to rename the problem directory to another name. | |
a933dad1 | 3232 | |
9dc15871 | 3233 | *** MS-DOS+DJGPP: Problems on MS-DOG if DJGPP v2.0 is used to compile Emacs. |
a933dad1 | 3234 | |
9dc15871 | 3235 | There are two DJGPP library bugs which cause problems: |
a933dad1 | 3236 | |
9dc15871 EZ |
3237 | * Running `shell-command' (or `compile', or `grep') you get |
3238 | `Searching for program: permission denied (EACCES), c:/command.com'; | |
3239 | * After you shell to DOS, Ctrl-Break kills Emacs. | |
a933dad1 | 3240 | |
9dc15871 EZ |
3241 | To work around these bugs, you can use two files in the msdos |
3242 | subdirectory: `is_exec.c' and `sigaction.c'. Compile them and link | |
3243 | them into the Emacs executable `temacs'; then they will replace the | |
3244 | incorrect library functions. | |
a933dad1 | 3245 | |
9dc15871 EZ |
3246 | *** MS-DOS: Emacs compiled for MSDOS cannot find some Lisp files, or other |
3247 | run-time support files, when long filename support is enabled. | |
a933dad1 | 3248 | |
9dc15871 EZ |
3249 | Usually, this problem will manifest itself when Emacs exits |
3250 | immediately after flashing the startup screen, because it cannot find | |
3251 | the Lisp files it needs to load at startup. Redirect Emacs stdout | |
3252 | and stderr to a file to see the error message printed by Emacs. | |
a933dad1 | 3253 | |
9dc15871 EZ |
3254 | Another manifestation of this problem is that Emacs is unable to load |
3255 | the support for editing program sources in languages such as C and | |
3256 | Lisp. | |
a933dad1 | 3257 | |
9dc15871 EZ |
3258 | This can happen if the Emacs distribution was unzipped without LFN |
3259 | support, thus causing long filenames to be truncated to the first 6 | |
3260 | characters and a numeric tail that Windows 95 normally attaches to it. | |
3261 | You should unzip the files again with a utility that supports long | |
3262 | filenames (such as djtar from DJGPP or InfoZip's UnZip program | |
3263 | compiled with DJGPP v2). The MSDOG section of the file INSTALL | |
3264 | explains this issue in more detail. | |
a933dad1 | 3265 | |
9dc15871 EZ |
3266 | Another possible reason for such failures is that Emacs compiled for |
3267 | MSDOS is used on Windows NT, where long file names are not supported | |
3268 | by this version of Emacs, but the distribution was unpacked by an | |
3269 | unzip program that preserved the long file names instead of truncating | |
3270 | them to DOS 8+3 limits. To be useful on NT, the MSDOS port of Emacs | |
3271 | must be unzipped by a DOS utility, so that long file names are | |
3272 | properly truncated. | |
a933dad1 | 3273 | |
9dc15871 | 3274 | ** Archaic window managers and toolkits |
a933dad1 | 3275 | |
9dc15871 | 3276 | *** OpenLook: Under OpenLook, the Emacs window disappears when you type M-q. |
a933dad1 | 3277 | |
9dc15871 EZ |
3278 | Some versions of the Open Look window manager interpret M-q as a quit |
3279 | command for whatever window you are typing at. If you want to use | |
3280 | Emacs with that window manager, you should try to configure the window | |
3281 | manager to use some other command. You can disable the | |
3282 | shortcut keys entirely by adding this line to ~/.OWdefaults: | |
a933dad1 | 3283 | |
9dc15871 | 3284 | OpenWindows.WindowMenuAccelerators: False |
a933dad1 | 3285 | |
9dc15871 | 3286 | **** twm: A position you specified in .Xdefaults is ignored, using twm. |
a933dad1 | 3287 | |
9dc15871 EZ |
3288 | twm normally ignores "program-specified" positions. |
3289 | You can tell it to obey them with this command in your `.twmrc' file: | |
a933dad1 | 3290 | |
9dc15871 | 3291 | UsePPosition "on" #allow clients to request a position |
a933dad1 | 3292 | |
9dc15871 | 3293 | ** Bugs related to old DEC hardware |
a933dad1 | 3294 | |
9dc15871 | 3295 | *** The Compose key on a DEC keyboard does not work as Meta key. |
a933dad1 | 3296 | |
9dc15871 | 3297 | This shell command should fix it: |
a933dad1 | 3298 | |
9dc15871 | 3299 | xmodmap -e 'keycode 0xb1 = Meta_L' |
a933dad1 | 3300 | |
9dc15871 EZ |
3301 | *** Keyboard input gets confused after a beep when using a DECserver |
3302 | as a concentrator. | |
a933dad1 | 3303 | |
9dc15871 EZ |
3304 | This problem seems to be a matter of configuring the DECserver to use |
3305 | 7 bit characters rather than 8 bit characters. | |
a933dad1 | 3306 | |
9dc15871 | 3307 | * Build problems on legacy systems |
a933dad1 | 3308 | |
9dc15871 | 3309 | ** BSD/386 1.0: --with-x-toolkit option configures wrong. |
a933dad1 | 3310 | |
9dc15871 EZ |
3311 | This problem is due to bugs in the shell in version 1.0 of BSD/386. |
3312 | The workaround is to edit the configure file to use some other shell, | |
3313 | such as bash. | |
a933dad1 | 3314 | |
9dc15871 EZ |
3315 | ** Digital Unix 4.0: Emacs fails to build, giving error message |
3316 | Invalid dimension for the charset-ID 160 | |
a933dad1 | 3317 | |
9dc15871 EZ |
3318 | This is due to a bug or an installation problem in GCC 2.8.0. |
3319 | Installing a more recent version of GCC fixes the problem. | |
a933dad1 | 3320 | |
9dc15871 | 3321 | ** Digital Unix 4.0: Failure in unexec while dumping emacs. |
a933dad1 | 3322 | |
9dc15871 | 3323 | This problem manifests itself as an error message |
a933dad1 | 3324 | |
9dc15871 | 3325 | unexec: Bad address, writing data section to ... |
a933dad1 | 3326 | |
9dc15871 EZ |
3327 | The user suspects that this happened because his X libraries |
3328 | were built for an older system version, | |
a933dad1 | 3329 | |
9dc15871 | 3330 | ./configure --x-includes=/usr/include --x-libraries=/usr/shlib |
a933dad1 | 3331 | |
9dc15871 | 3332 | made the problem go away. |
a933dad1 | 3333 | |
9dc15871 | 3334 | ** Sunos 4.1.1: there are errors compiling sysdep.c. |
a933dad1 | 3335 | |
9dc15871 | 3336 | If you get errors such as |
a933dad1 | 3337 | |
9dc15871 EZ |
3338 | "sysdep.c", line 2017: undefined structure or union |
3339 | "sysdep.c", line 2017: undefined structure or union | |
3340 | "sysdep.c", line 2019: nodename undefined | |
a933dad1 | 3341 | |
9dc15871 EZ |
3342 | This can result from defining LD_LIBRARY_PATH. It is very tricky |
3343 | to use that environment variable with Emacs. The Emacs configure | |
3344 | script links many test programs with the system libraries; you must | |
3345 | make sure that the libraries available to configure are the same | |
3346 | ones available when you build Emacs. | |
a933dad1 | 3347 | |
9dc15871 | 3348 | ** SunOS 4.1.1: You get this error message from GNU ld: |
a933dad1 | 3349 | |
9dc15871 | 3350 | /lib/libc.a(_Q_sub.o): Undefined symbol __Q_get_rp_rd referenced from text segment |
a933dad1 | 3351 | |
9dc15871 | 3352 | The problem is in the Sun shared C library, not in GNU ld. |
a933dad1 | 3353 | |
9dc15871 | 3354 | The solution is to install Patch-ID# 100267-03 from Sun. |
a933dad1 | 3355 | |
9dc15871 | 3356 | ** Sunos 4.1: Undefined symbols when linking using --with-x-toolkit. |
a933dad1 | 3357 | |
9dc15871 EZ |
3358 | If you get the undefined symbols _atowc _wcslen, _iswprint, _iswspace, |
3359 | _iswcntrl, _wcscpy, and _wcsncpy, then you need to add -lXwchar after | |
3360 | -lXaw in the command that links temacs. | |
a933dad1 | 3361 | |
9dc15871 EZ |
3362 | This problem seems to arise only when the international language |
3363 | extensions to X11R5 are installed. | |
a933dad1 | 3364 | |
9dc15871 | 3365 | ** SunOS: Emacs gets error message from linker on Sun. |
a933dad1 | 3366 | |
9dc15871 EZ |
3367 | If the error message says that a symbol such as `f68881_used' or |
3368 | `ffpa_used' or `start_float' is undefined, this probably indicates | |
3369 | that you have compiled some libraries, such as the X libraries, | |
3370 | with a floating point option other than the default. | |
a933dad1 | 3371 | |
9dc15871 EZ |
3372 | It's not terribly hard to make this work with small changes in |
3373 | crt0.c together with linking with Fcrt1.o, Wcrt1.o or Mcrt1.o. | |
3374 | However, the easiest approach is to build Xlib with the default | |
3375 | floating point option: -fsoft. | |
a933dad1 | 3376 | |
9dc15871 | 3377 | ** SunOS: Undefined symbols _dlopen, _dlsym and/or _dlclose. |
a933dad1 | 3378 | |
9dc15871 EZ |
3379 | If you see undefined symbols _dlopen, _dlsym, or _dlclose when linking |
3380 | with -lX11, compile and link against the file mit/util/misc/dlsym.c in | |
3381 | the MIT X11R5 distribution. Alternatively, link temacs using shared | |
3382 | libraries with s/sunos4shr.h. (This doesn't work if you use the X | |
3383 | toolkit.) | |
a933dad1 | 3384 | |
9dc15871 EZ |
3385 | If you get the additional error that the linker could not find |
3386 | lib_version.o, try extracting it from X11/usr/lib/X11/libvim.a in | |
3387 | X11R4, then use it in the link. | |
a933dad1 | 3388 | |
0a4dd4e4 EZ |
3389 | ** SunOS4, DGUX 5.4.2: --with-x-toolkit version crashes when used with shared libraries. |
3390 | ||
3391 | On some systems, including Sunos 4 and DGUX 5.4.2 and perhaps others, | |
3392 | unexec doesn't work properly with the shared library for the X | |
3393 | toolkit. You might be able to work around this by using a nonshared | |
3394 | libXt.a library. The real fix is to upgrade the various versions of | |
3395 | unexec and/or ralloc. We think this has been fixed on Sunos 4 | |
3396 | and Solaris in version 19.29. | |
3397 | ||
3398 | ** HPUX 10.20: Emacs crashes during dumping on the HPPA machine. | |
3399 | ||
3400 | This seems to be due to a GCC bug; it is fixed in GCC 2.8.1. | |
3401 | ||
9dc15871 | 3402 | ** VMS: Compilation errors on VMS. |
a933dad1 | 3403 | |
9dc15871 EZ |
3404 | You will get warnings when compiling on VMS because there are |
3405 | variable names longer than 32 (or whatever it is) characters. | |
3406 | This is not an error. Ignore it. | |
a933dad1 | 3407 | |
9dc15871 EZ |
3408 | VAX C does not support #if defined(foo). Uses of this construct |
3409 | were removed, but some may have crept back in. They must be rewritten. | |
a933dad1 | 3410 | |
9dc15871 EZ |
3411 | There is a bug in the C compiler which fails to sign extend characters |
3412 | in conditional expressions. The bug is: | |
3413 | char c = -1, d = 1; | |
3414 | int i; | |
a933dad1 | 3415 | |
9dc15871 EZ |
3416 | i = d ? c : d; |
3417 | The result is i == 255; the fix is to typecast the char in the | |
3418 | conditional expression as an (int). Known occurrences of such | |
3419 | constructs in Emacs have been fixed. | |
a933dad1 | 3420 | |
9dc15871 | 3421 | ** Vax C compiler bugs affecting Emacs. |
a933dad1 DL |
3422 | |
3423 | You may get one of these problems compiling Emacs: | |
3424 | ||
3425 | foo.c line nnn: compiler error: no table entry for op STASG | |
3426 | foo.c: fatal error in /lib/ccom | |
3427 | ||
3428 | These are due to bugs in the C compiler; the code is valid C. | |
3429 | Unfortunately, the bugs are unpredictable: the same construct | |
3430 | may compile properly or trigger one of these bugs, depending | |
3431 | on what else is in the source file being compiled. Even changes | |
3432 | in header files that should not affect the file being compiled | |
3433 | can affect whether the bug happens. In addition, sometimes files | |
3434 | that compile correctly on one machine get this bug on another machine. | |
3435 | ||
3436 | As a result, it is hard for me to make sure this bug will not affect | |
3437 | you. I have attempted to find and alter these constructs, but more | |
3438 | can always appear. However, I can tell you how to deal with it if it | |
3439 | should happen. The bug comes from having an indexed reference to an | |
3440 | array of Lisp_Objects, as an argument in a function call: | |
3441 | Lisp_Object *args; | |
3442 | ... | |
3443 | ... foo (5, args[i], ...)... | |
3444 | putting the argument into a temporary variable first, as in | |
3445 | Lisp_Object *args; | |
3446 | Lisp_Object tem; | |
3447 | ... | |
3448 | tem = args[i]; | |
3449 | ... foo (r, tem, ...)... | |
3450 | causes the problem to go away. | |
3451 | The `contents' field of a Lisp vector is an array of Lisp_Objects, | |
3452 | so you may see the problem happening with indexed references to that. | |
3453 | ||
9dc15871 | 3454 | ** 68000 C compiler problems |
a933dad1 DL |
3455 | |
3456 | Various 68000 compilers have different problems. | |
3457 | These are some that have been observed. | |
3458 | ||
9dc15871 | 3459 | *** Using value of assignment expression on union type loses. |
a933dad1 DL |
3460 | This means that x = y = z; or foo (x = z); does not work |
3461 | if x is of type Lisp_Object. | |
3462 | ||
9dc15871 | 3463 | *** "cannot reclaim" error. |
a933dad1 DL |
3464 | |
3465 | This means that an expression is too complicated. You get the correct | |
3466 | line number in the error message. The code must be rewritten with | |
3467 | simpler expressions. | |
3468 | ||
9dc15871 | 3469 | *** XCONS, XSTRING, etc macros produce incorrect code. |
a933dad1 DL |
3470 | |
3471 | If temacs fails to run at all, this may be the cause. | |
3472 | Compile this test program and look at the assembler code: | |
3473 | ||
3474 | struct foo { char x; unsigned int y : 24; }; | |
3475 | ||
3476 | lose (arg) | |
3477 | struct foo arg; | |
3478 | { | |
3479 | test ((int *) arg.y); | |
3480 | } | |
3481 | ||
3482 | If the code is incorrect, your compiler has this problem. | |
3483 | In the XCONS, etc., macros in lisp.h you must replace (a).u.val with | |
3484 | ((a).u.val + coercedummy) where coercedummy is declared as int. | |
3485 | ||
3486 | This problem will not happen if the m-...h file for your type | |
3487 | of machine defines NO_UNION_TYPE. That is the recommended setting now. | |
3488 | ||
9dc15871 | 3489 | *** C compilers lose on returning unions. |
a933dad1 DL |
3490 | |
3491 | I hear that some C compilers cannot handle returning a union type. | |
3492 | Most of the functions in GNU Emacs return type Lisp_Object, which is | |
3493 | defined as a union on some rare architectures. | |
3494 | ||
3495 | This problem will not happen if the m-...h file for your type | |
3496 | of machine defines NO_UNION_TYPE. | |
3497 | ||
53854552 | 3498 | \f |
9dc15871 | 3499 | Copyright 1987,88,89,93,94,95,96,97,98,1999,2001,2002,2004 |
fe6b4873 RS |
3500 | Free Software Foundation, Inc. |
3501 | ||
3502 | Copying and redistribution of this file with or without modification | |
3503 | are permitted without royalty provided this notice is preserved. | |
3504 | ||
53854552 EZ |
3505 | Local variables: |
3506 | mode: outline | |
3507 | paragraph-separate: "[ \f]*$" | |
3508 | end: | |
ab5796a9 MB |
3509 | |
3510 | arch-tag: 49fc0d95-88cb-4715-b21c-f27fb5a4764a |