(xg_frame_resized): Remove check if rows/columns have
[bpt/emacs.git] / etc / PROBLEMS
CommitLineData
5e14abf8
KS
1Known Problems with GNU Emacs
2
5b87ad55 3Copyright (C) 1987, 1988, 1989, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
8cabe764 4 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
5b87ad55
GM
5 Free Software Foundation, Inc.
6See the end of the file for license conditions.
7
8
a933dad1 9This file describes various problems that have been encountered
0a4dd4e4 10in compiling, installing and running GNU Emacs. Try doing Ctl-C Ctl-t
9dc15871 11and browsing through the outline headers.
a933dad1 12
8589dc17 13* Mule-UCS doesn't work in Emacs 23.
2c311b39
DL
14
15It's completely redundant now, as far as we know.
16
9dc15871 17* Emacs startup failures
32364f49 18
9dc15871 19** Emacs fails to start, complaining about missing fonts.
32364f49 20
9dc15871 21A typical error message might be something like
32364f49 22
9dc15871 23 No fonts match `-*-fixed-medium-r-*--6-*-*-*-*-*-iso8859-1'
6b61353c 24
9dc15871
EZ
25This happens because some X resource specifies a bad font family for
26Emacs to use. The possible places where this specification might be
27are:
6b61353c 28
9dc15871 29 - in your ~/.Xdefaults file
6b61353c 30
9dc15871
EZ
31 - client-side X resource file, such as ~/Emacs or
32 /usr/X11R6/lib/app-defaults/Emacs or
33 /usr/X11R6/lib/X11/app-defaults/Emacs
6b61353c 34
9dc15871
EZ
35One of these files might have bad or malformed specification of a
36fontset that Emacs should use. To fix the problem, you need to find
37the problematic line(s) and correct them.
6b61353c 38
9dc15871 39** Emacs aborts while starting up, only when run without X.
6b61353c 40
9dc15871
EZ
41This problem often results from compiling Emacs with GCC when GCC was
42installed incorrectly. The usual error in installing GCC is to
43specify --includedir=/usr/include. Installation of GCC makes
44corrected copies of the system header files. GCC is supposed to use
45the corrected copies in preference to the original system headers.
46Specifying --includedir=/usr/include causes the original system header
47files to be used. On some systems, the definition of ioctl in the
48original system header files is invalid for ANSI C and causes Emacs
49not to work.
6b61353c 50
9dc15871
EZ
51The fix is to reinstall GCC, and this time do not specify --includedir
52when you configure it. Then recompile Emacs. Specifying --includedir
53is appropriate only in very special cases and it should *never* be the
54same directory where system header files are kept.
6b61353c 55
9dc15871 56** Emacs does not start, complaining that it cannot open termcap database file.
6b61353c 57
9dc15871
EZ
58If your system uses Terminfo rather than termcap (most modern
59systems do), this could happen if the proper version of
60ncurses is not visible to the Emacs configure script (i.e. it
61cannot be found along the usual path the linker looks for
62libraries). It can happen because your version of ncurses is
63obsolete, or is available only in form of binaries.
6b61353c 64
9dc15871
EZ
65The solution is to install an up-to-date version of ncurses in
66the developer's form (header files, static libraries and
67symbolic links); in some GNU/Linux distributions (e.g. Debian)
68it constitutes a separate package.
6b61353c 69
9dc15871 70** Emacs 20 and later fails to load Lisp files at startup.
5b4ffca2 71
9dc15871 72The typical error message might be like this:
5b4ffca2 73
9dc15871 74 "Cannot open load file: fontset"
c763d515 75
9dc15871
EZ
76This could happen if you compress the file lisp/subdirs.el. That file
77tells Emacs what are the directories where it should look for Lisp
78files. Emacs cannot work with subdirs.el compressed, since the
79Auto-compress mode it needs for this will not be loaded until later,
80when your .emacs file is processed. (The package `fontset.el' is
81required to set up fonts used to display text on window systems, and
82it's loaded very early in the startup procedure.)
f1c231c4 83
9dc15871
EZ
84Similarly, any other .el file for which there's no corresponding .elc
85file could fail to load if it is compressed.
fc2938d1 86
9dc15871
EZ
87The solution is to uncompress all .el files which don't have a .elc
88file.
6b61353c 89
9dc15871
EZ
90Another possible reason for such failures is stale *.elc files
91lurking somewhere on your load-path. The following command will
92print any duplicate Lisp files that are present in load-path:
6b61353c 93
9dc15871 94 emacs -q -batch -f list-load-path-shadows
6b61353c 95
9dc15871
EZ
96If this command prints any file names, some of these files are stale,
97and should be deleted or their directories removed from your
98load-path.
f0f62f71 99
9dc15871 100** Emacs prints an error at startup after upgrading from an earlier version.
f0f62f71 101
9dc15871 102An example of such an error is:
f0f62f71 103
9dc15871 104 x-complement-fontset-spec: "Wrong type argument: stringp, nil"
fc1bfc2a 105
9dc15871
EZ
106This can be another symptom of stale *.elc files in your load-path.
107The following command will print any duplicate Lisp files that are
108present in load-path:
fc1bfc2a 109
9dc15871 110 emacs -q -batch -f list-load-path-shadows
fc1bfc2a 111
9dc15871
EZ
112If this command prints any file names, some of these files are stale,
113and should be deleted or their directories removed from your
114load-path.
60f553d2 115
9dc15871 116** With X11R6.4, public-patch-3, Emacs crashes at startup.
3f82efb4 117
9dc15871 118Reportedly this patch in X fixes the problem.
3f82efb4 119
9dc15871
EZ
120 --- xc/lib/X11/imInt.c~ Wed Jun 30 13:31:56 1999
121 +++ xc/lib/X11/imInt.c Thu Jul 1 15:10:27 1999
122 @@ -1,4 +1,4 @@
123 -/* $TOG: imInt.c /main/5 1998/05/30 21:11:16 kaleb $ */
124 +/* $TOG: imInt.c /main/5 1998/05/30 21:11:16 kaleb $ */
125 /******************************************************************
3f82efb4 126
3256a475 127 Copyright 1992, 1993, 1994 by FUJITSU LIMITED
9dc15871
EZ
128 @@ -166,8 +166,8 @@
129 _XimMakeImName(lcd)
3256a475 130 XLCd lcd;
9dc15871
EZ
131 {
132 - char* begin;
133 - char* end;
134 + char* begin = NULL;
135 + char* end = NULL;
3256a475
RC
136 char* ret;
137 int i = 0;
138 char* ximmodifier = XIMMODIFIER;
9dc15871 139 @@ -182,7 +182,11 @@
3256a475
RC
140 }
141 ret = Xmalloc(end - begin + 2);
142 if (ret != NULL) {
143 - (void)strncpy(ret, begin, end - begin + 1);
9dc15871 144 + if (begin != NULL) {
3256a475 145 + (void)strncpy(ret, begin, end - begin + 1);
9dc15871 146 + } else {
3256a475 147 + ret[0] = '\0';
9dc15871 148 + }
3256a475
RC
149 ret[end - begin + 1] = '\0';
150 }
151 return ret;
3f82efb4 152
13673952
CY
153** Emacs crashes on startup after a glibc upgrade.
154
155This is caused by a binary incompatible change to the malloc
156implementation in glibc 2.5.90-22. As a result, Emacs binaries built
157using prior versions of glibc crash when run under 2.5.90-22.
158
0a7aab83 159This problem was first seen in pre-release versions of Fedora 7, and
13673952
CY
160may be fixed in the final Fedora 7 release. To stop the crash from
161happening, first try upgrading to the newest version of glibc; if this
162does not work, rebuild Emacs with the same version of glibc that you
163will run it under. For details, see
164
165https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=239344
166
9dc15871 167* Crash bugs
cc305a60 168
9dc15871 169** Emacs crashes in x-popup-dialog.
cc305a60 170
9dc15871
EZ
171This can happen if the dialog widget cannot find the font it wants to
172use. You can work around the problem by specifying another font with
173an X resource--for example, `Emacs.dialog*.font: 9x15' (or any font that
174happens to exist on your X server).
fc2938d1 175
9dc15871 176** Emacs crashes when you use Bibtex mode.
fc2938d1 177
9dc15871
EZ
178This happens if your system puts a small limit on stack size. You can
179prevent the problem by using a suitable shell command (often `ulimit')
180to raise the stack size limit before you run Emacs.
177c0ea7 181
9dc15871
EZ
182Patches to raise the stack size limit automatically in `main'
183(src/emacs.c) on various systems would be greatly appreciated.
fc2938d1 184
9dc15871
EZ
185** Error message `Symbol's value as variable is void: x', followed by
186a segmentation fault and core dump.
c93bdf05 187
9dc15871
EZ
188This has been tracked to a bug in tar! People report that tar erroneously
189added a line like this at the beginning of files of Lisp code:
c93bdf05 190
9dc15871 191 x FILENAME, N bytes, B tape blocks
c93bdf05 192
9dc15871
EZ
193If your tar has this problem, install GNU tar--if you can manage to
194untar it :-).
c93bdf05 195
9dc15871
EZ
196** Crashes when displaying GIF images in Emacs built with version
197libungif-4.1.0 are resolved by using version libungif-4.1.0b1.
198Configure checks for the correct version, but this problem could occur
199if a binary built against a shared libungif is run on a system with an
200older version.
4593687f 201
9dc15871 202** Emacs aborts inside the function `tparam1'.
9272ccfc 203
9dc15871
EZ
204This can happen if Emacs was built without terminfo support, but the
205terminal's capabilities use format that is only supported by terminfo.
206If your system has ncurses installed, this might happen if your
207version of ncurses is broken; upgrading to a newer version of ncurses
208and reconfiguring and rebuilding Emacs should solve this.
9272ccfc 209
9dc15871
EZ
210All modern systems support terminfo, so even if ncurses is not the
211problem, you should look for a way to configure Emacs so that it uses
212terminfo when built.
9272ccfc 213
9dc15871 214** Emacs crashes when using the Exceed 6.0 X server.
7aa70236 215
9dc15871
EZ
216If you are using Exceed 6.1, upgrade to a later version. This was
217reported to prevent the crashes.
7aa70236 218
9dc15871 219** Emacs crashes with SIGSEGV in XtInitializeWidgetClass.
7c22dc9d 220
9dc15871 221It crashes on X, but runs fine when called with option "-nw".
7c22dc9d 222
9dc15871
EZ
223This has been observed when Emacs is linked with GNU ld but without passing
224the -z nocombreloc flag. Emacs normally knows to pass the -z nocombreloc
225flag when needed, so if you come across a situation where the flag is
226necessary but missing, please report it via M-x report-emacs-bug.
7c22dc9d 227
9dc15871
EZ
228On platforms such as Solaris, you can also work around this problem by
229configuring your compiler to use the native linker instead of GNU ld.
7c22dc9d 230
1437ec2b
JD
231** Emacs compiled with Gtk+ crashes when closing a display (x-close-connection).
232
233This happens because of bugs in Gtk+. Gtk+ 2.10 seems to be OK. See bug
234http://bugzilla.gnome.org/show_bug.cgi?id=85715.
235
1020d879 236** Emacs compiled with Gtk+ crashes on startup on Cygwin.
6df79155
JD
237
238A typical error message is
dc13f3f5 239 ***MEMORY-ERROR***: emacs[5172]: GSlice: failed to allocate 504 bytes
6df79155
JD
240 (alignment: 512): Function not implemented
241
242Emacs supplies its own malloc, but glib (part of Gtk+) calls memalign and on
1020d879
EZ
243Cygwin, that becomes the Cygwin supplied memalign. As malloc is not the
244Cygwin malloc, the Cygwin memalign always returns ENOSYS. A fix for this
6df79155
JD
245problem would be welcome.
246
9dc15871 247* General runtime problems
7c22dc9d 248
9dc15871 249** Lisp problems
677e7496 250
9dc15871 251*** Changes made to .el files do not take effect.
677e7496 252
9dc15871
EZ
253You may have forgotten to recompile them into .elc files.
254Then the old .elc files will be loaded, and your changes
255will not be seen. To fix this, do M-x byte-recompile-directory
256and specify the directory that contains the Lisp files.
677e7496 257
9dc15871
EZ
258Emacs should print a warning when loading a .elc file which is older
259than the corresponding .el file.
677e7496 260
9dc15871 261*** Watch out for .emacs files and EMACSLOADPATH environment vars.
677e7496 262
9dc15871
EZ
263These control the actions of Emacs.
264~/.emacs is your Emacs init file.
265EMACSLOADPATH overrides which directories the function
266"load" will search.
677e7496 267
9dc15871
EZ
268If you observe strange problems, check for these and get rid
269of them, then try again.
9ed04369 270
9dc15871 271*** Using epop3.el package causes Emacs to signal an error.
9ed04369 272
9dc15871 273The error message might be something like this:
b87207a0 274
9dc15871 275 "Lisp nesting exceeds max-lisp-eval-depth"
b87207a0 276
9dc15871
EZ
277This happens because epop3 redefines the function gethash, which is a
278built-in primitive beginning with Emacs 21.1. We don't have a patch
279for epop3 that fixes this, but perhaps a newer version of epop3
280corrects that.
177c0ea7 281
9dc15871 282*** Buffers from `with-output-to-temp-buffer' get set up in Help mode.
177c0ea7 283
9dc15871
EZ
284Changes in Emacs 20.4 to the hooks used by that function cause
285problems for some packages, specifically BBDB. See the function's
286documentation for the hooks involved. BBDB 2.00.06 fixes the problem.
177c0ea7 287
9dc15871
EZ
288*** The Hyperbole package causes *Help* buffers not to be displayed in
289Help mode due to setting `temp-buffer-show-hook' rather than using
290`add-hook'. Using `(add-hook 'temp-buffer-show-hook
291'help-mode-maybe)' after loading Hyperbole should fix this.
177c0ea7 292
9dc15871 293** Keyboard problems
b87207a0 294
9dc15871 295*** "Compose Character" key does strange things when used as a Meta key.
61638355 296
9dc15871
EZ
297If you define one key to serve as both Meta and Compose Character, you
298will get strange results. In previous Emacs versions, this "worked"
299in that the key acted as Meta--that's because the older Emacs versions
300did not try to support Compose Character. Now Emacs tries to do
301character composition in the standard X way. This means that you
302must pick one meaning or the other for any given key.
61638355 303
9dc15871
EZ
304You can use both functions (Meta, and Compose Character) if you assign
305them to two different keys.
a47a639f 306
9dc15871 307*** C-z just refreshes the screen instead of suspending Emacs.
a47a639f 308
9dc15871
EZ
309You are probably using a shell that doesn't support job control, even
310though the system itself is capable of it. Either use a different shell,
311or set the variable `cannot-suspend' to a non-nil value.
a47a639f 312
9dc15871
EZ
313*** With M-x enable-flow-control, you need to type C-\ twice
314to do incremental search--a single C-\ gets no response.
a47a639f 315
9dc15871
EZ
316This has been traced to communicating with your machine via kermit,
317with C-\ as the kermit escape character. One solution is to use
318another escape character in kermit. One user did
61638355 319
9dc15871 320 set escape-character 17
61638355 321
9dc15871 322in his .kermrc file, to make C-q the kermit escape character.
61638355 323
9dc15871 324** Mailers and other helper programs
61638355 325
9dc15871 326*** movemail compiled with POP support can't connect to the POP server.
61638355 327
9dc15871
EZ
328Make sure that the `pop' entry in /etc/services, or in the services
329NIS map if your machine uses NIS, has the same port number as the
330entry on the POP server. A common error is for the POP server to be
331listening on port 110, the assigned port for the POP3 protocol, while
332the client is trying to connect on port 109, the assigned port for the
333old POP protocol.
61638355 334
9dc15871 335*** RMAIL gets error getting new mail.
61638355 336
9dc15871
EZ
337RMAIL gets new mail from /usr/spool/mail/$USER using a program
338called `movemail'. This program interlocks with /bin/mail using
339the protocol defined by /bin/mail.
61638355 340
9dc15871
EZ
341There are two different protocols in general use. One of them uses
342the `flock' system call. The other involves creating a lock file;
343`movemail' must be able to write in /usr/spool/mail in order to do
344this. You control which one is used by defining, or not defining,
345the macro MAIL_USE_FLOCK in config.h or the m- or s- file it includes.
346IF YOU DON'T USE THE FORM OF INTERLOCKING THAT IS NORMAL ON YOUR
347SYSTEM, YOU CAN LOSE MAIL!
61638355 348
9dc15871
EZ
349If your system uses the lock file protocol, and fascist restrictions
350prevent ordinary users from writing the lock files in /usr/spool/mail,
351you may need to make `movemail' setgid to a suitable group such as
352`mail'. To do this, use the following commands (as root) after doing the
353make install.
61638355 354
3256a475
RC
355 chgrp mail movemail
356 chmod 2755 movemail
61638355 357
9dc15871
EZ
358Installation normally copies movemail from the build directory to an
359installation directory which is usually under /usr/local/lib. The
360installed copy of movemail is usually in the directory
361/usr/local/lib/emacs/VERSION/TARGET. You must change the group and
362mode of the installed copy; changing the group and mode of the build
363directory copy is ineffective.
61638355 364
9dc15871 365*** rcs2log gives you the awk error message "too many fields".
61638355 366
9dc15871
EZ
367This is due to an arbitrary limit in certain versions of awk.
368The solution is to use gawk (GNU awk).
61638355 369
9dc15871 370** Problems with hostname resolution
61638355 371
9dc15871
EZ
372*** Emacs fails to understand most Internet host names, even though
373the names work properly with other programs on the same system.
374*** Emacs won't work with X-windows if the value of DISPLAY is HOSTNAME:0.
20dc2215 375*** Gnus can't make contact with the specified host for nntp.
61638355 376
9dc15871
EZ
377This typically happens on Suns and other systems that use shared
378libraries. The cause is that the site has installed a version of the
379shared library which uses a name server--but has not installed a
380similar version of the unshared library which Emacs uses.
556a9fad 381
9dc15871
EZ
382The result is that most programs, using the shared library, work with
383the nameserver, but Emacs does not.
61638355 384
9dc15871
EZ
385The fix is to install an unshared library that corresponds to what you
386installed in the shared library, and then relink Emacs.
49172314 387
9dc15871
EZ
388If you have already installed the name resolver in the file libresolv.a,
389then you need to compile Emacs to use that library. The easiest way to
390do this is to add to config.h a definition of LIBS_SYSTEM, LIBS_MACHINE
391or LIB_STANDARD which uses -lresolv. Watch out! If you redefine a macro
392that is already in use in your configuration to supply some other libraries,
393be careful not to lose the others.
ed0d1d91 394
9dc15871 395Thus, you could start by adding this to config.h:
6e1a66dc 396
9dc15871 397#define LIBS_SYSTEM -lresolv
6e1a66dc 398
9dc15871
EZ
399Then if this gives you an error for redefining a macro, and you see that
400the s- file defines LIBS_SYSTEM as -lfoo -lbar, you could change config.h
401again to say this:
3c418e54 402
9dc15871 403#define LIBS_SYSTEM -lresolv -lfoo -lbar
3c418e54 404
9dc15871 405*** Emacs does not know your host's fully-qualified domain name.
3c418e54 406
53b30c38
GM
407For example, (system-name) returns some variation on
408"localhost.localdomain", rather the name you were expecting.
409
9dc15871 410You need to configure your machine with a fully qualified domain name,
53b30c38
GM
411(i.e. a name with at least one ".") either in /etc/hosts,
412/etc/hostname, the NIS, or wherever your system calls for specifying
413this.
f9130829 414
9dc15871
EZ
415If you cannot fix the configuration, you can set the Lisp variable
416mail-host-address to the value you want.
f9130829 417
9dc15871 418** NFS and RFS
f9130829 419
9dc15871
EZ
420*** Emacs says it has saved a file, but the file does not actually
421appear on disk.
f9130829 422
9dc15871
EZ
423This can happen on certain systems when you are using NFS, if the
424remote disk is full. It is due to a bug in NFS (or certain NFS
425implementations), and there is apparently nothing Emacs can do to
426detect the problem. Emacs checks the failure codes of all the system
427calls involved in writing a file, including `close'; but in the case
428where the problem occurs, none of those system calls fails.
f9130829 429
9dc15871
EZ
430*** Editing files through RFS gives spurious "file has changed" warnings.
431It is possible that a change in Emacs 18.37 gets around this problem,
432but in case not, here is a description of how to fix the RFS bug that
433causes it.
f9130829 434
9dc15871
EZ
435 There was a serious pair of bugs in the handling of the fsync() system
436 call in the RFS server.
f9130829 437
9dc15871
EZ
438 The first is that the fsync() call is handled as another name for the
439 close() system call (!!). It appears that fsync() is not used by very
440 many programs; Emacs version 18 does an fsync() before closing files
441 to make sure that the bits are on the disk.
ed85f61d 442
9dc15871 443 This is fixed by the enclosed patch to the RFS server.
b300fd77 444
9dc15871
EZ
445 The second, more serious problem, is that fsync() is treated as a
446 non-blocking system call (i.e., it's implemented as a message that
447 gets sent to the remote system without waiting for a reply). Fsync is
448 a useful tool for building atomic file transactions. Implementing it
449 as a non-blocking RPC call (when the local call blocks until the sync
450 is done) is a bad idea; unfortunately, changing it will break the RFS
451 protocol. No fix was supplied for this problem.
b300fd77 452
9dc15871 453 (as always, your line numbers may vary)
ed85f61d 454
9dc15871
EZ
455 % rcsdiff -c -r1.2 serversyscall.c
456 RCS file: RCS/serversyscall.c,v
457 retrieving revision 1.2
458 diff -c -r1.2 serversyscall.c
459 *** /tmp/,RCSt1003677 Wed Jan 28 15:15:02 1987
460 --- serversyscall.c Wed Jan 28 15:14:48 1987
461 ***************
462 *** 163,169 ****
3256a475
RC
463 /*
464 * No return sent for close or fsync!
465 */
9dc15871 466 ! if (syscall == RSYS_close || syscall == RSYS_fsync)
3256a475
RC
467 proc->p_returnval = deallocate_fd(proc, msg->m_args[0]);
468 else
469 {
9dc15871 470 --- 166,172 ----
3256a475
RC
471 /*
472 * No return sent for close or fsync!
473 */
9dc15871 474 ! if (syscall == RSYS_close)
3256a475
RC
475 proc->p_returnval = deallocate_fd(proc, msg->m_args[0]);
476 else
477 {
ed85f61d 478
9dc15871 479** PSGML
fa2301bf 480
9dc15871
EZ
481*** Old versions of the PSGML package use the obsolete variables
482`before-change-function' and `after-change-function', which are no
483longer used by Emacs. Please use PSGML 1.2.3 or later.
ed85f61d 484
9dc15871 485*** PSGML conflicts with sgml-mode.
d0cf6c7d 486
9dc15871
EZ
487PSGML package uses the same names of some variables (like keymap)
488as built-in sgml-mode.el because it was created as a replacement
489of that package. The conflict will be shown if you load
490sgml-mode.el before psgml.el. E.g. this could happen if you edit
491HTML page and then start to work with SGML or XML file. html-mode
492(from sgml-mode.el) is used for HTML file and loading of psgml.el
493(for sgml-mode or xml-mode) will cause an error.
d0cf6c7d 494
9dc15871
EZ
495*** Versions of the PSGML package earlier than 1.0.3 (stable) or 1.1.2
496(alpha) fail to parse DTD files correctly in Emacs 20.3 and later.
497Here is a patch for psgml-parse.el from PSGML 1.0.1 and, probably,
498earlier versions.
d0cf6c7d 499
9dc15871
EZ
500--- psgml-parse.el 1998/08/21 19:18:18 1.1
501+++ psgml-parse.el 1998/08/21 19:20:00
502@@ -2383,7 +2383,7 @@ (defun sgml-push-to-entity (entity &opti
503 (setq sgml-buffer-parse-state nil))
504 (cond
505 ((stringp entity) ; a file name
506- (save-excursion (insert-file-contents entity))
507+ (insert-file-contents entity)
508 (setq default-directory (file-name-directory entity)))
509 ((consp (sgml-entity-text entity)) ; external id?
510 (let* ((extid (sgml-entity-text entity))
d0cf6c7d 511
90a02640 512** AUCTeX
c289e7f0 513
90a02640
DK
514You should not be using a version older than 11.52 if you can avoid
515it.
339b95d5 516
90a02640 517*** Emacs 21 freezes when visiting a TeX file with AUCTeX installed.
339b95d5 518
90a02640 519Emacs 21 needs version 10 or later of AUCTeX; upgrading should solve
9dc15871 520these problems.
339b95d5 521
90a02640 522*** No colors in AUCTeX with Emacs 21.
339b95d5 523
9dc15871
EZ
524Upgrade to AUC TeX version 10 or later, and make sure it is
525byte-compiled with Emacs 21.
b87207a0 526
73639601
EZ
527** PCL-CVS
528
529*** Lines are not updated or new lines are added in the buffer upon commit.
530
531When committing files located higher in the hierarchy than the examined
532directory, some versions of the CVS program return an ambiguous message
533from which PCL-CVS cannot extract the full location of the committed
534files. As a result, the corresponding lines in the PCL-CVS buffer are
535not updated with the new revision of these files, and new lines are
536added to the top-level directory.
537
538This can happen with CVS versions 1.12.8 and 1.12.9. Upgrade to CVS
5391.12.10 or newer to fix this problem.
540
9dc15871 541** Miscellaneous problems
f936978f 542
f5578c7f
EZ
543*** Emacs uses 100% of CPU time
544
545This is a known problem with some versions of the Semantic package.
3f37e604
GM
546The solution is to upgrade Semantic to version 2.0pre4 (distributed
547with CEDET 1.0pre4) or later.
f5578c7f 548
9dc15871 549*** Self-documentation messages are garbled.
6fb6f3ac 550
9dc15871
EZ
551This means that the file `etc/DOC-...' doesn't properly correspond
552with the Emacs executable. Redumping Emacs and then installing the
553corresponding pair of files should fix the problem.
b87207a0 554
9dc15871
EZ
555*** Programs running under terminal emulator do not recognize `emacs'
556terminal type.
b87207a0 557
9dc15871
EZ
558The cause of this is a shell startup file that sets the TERMCAP
559environment variable. The terminal emulator uses that variable to
560provide the information on the special terminal type that Emacs
561emulates.
b87207a0 562
9dc15871
EZ
563Rewrite your shell startup file so that it does not change TERMCAP
564in such a case. You could use the following conditional which sets
565it only if it is undefined.
b87207a0 566
9dc15871 567 if ( ! ${?TERMCAP} ) setenv TERMCAP ~/my-termcap-file
e085efdb 568
9dc15871
EZ
569Or you could set TERMCAP only when you set TERM--which should not
570happen in a non-login shell.
fa99e2a4 571
9dc15871 572*** In Shell mode, you get a ^M at the end of every line.
c8d9b4ee 573
9dc15871
EZ
574This happens to people who use tcsh, because it is trying to be too
575smart. It sees that the Shell uses terminal type `unknown' and turns
576on the flag to output ^M at the end of each line. You can fix the
577problem by adding this to your .cshrc file:
fe445893 578
9dc15871 579 if ($?EMACS) then
4b1aaa8b 580 if ("$EMACS" =~ /*) then
9dc15871
EZ
581 unset edit
582 stty -icrnl -onlcr -echo susp ^Z
583 endif
584 endif
c8d9b4ee 585
9dc15871 586*** Emacs startup on GNU/Linux systems (and possibly other systems) is slow.
d9810886 587
9dc15871
EZ
588This can happen if the system is misconfigured and Emacs can't get the
589full qualified domain name, FQDN. You should have your FQDN in the
590/etc/hosts file, something like this:
a408ce18 591
9dc15871
EZ
592127.0.0.1 localhost
593129.187.137.82 nuc04.t30.physik.tu-muenchen.de nuc04
a408ce18 594
9dc15871 595The way to set this up may vary on non-GNU systems.
a38f41c4 596
9dc15871 597*** Attempting to visit remote files via ange-ftp fails.
a38f41c4 598
9dc15871
EZ
599If the error message is "ange-ftp-file-modtime: Specified time is not
600representable", then this could happen when `lukemftp' is used as the
601ftp client. This was reported to happen on Debian GNU/Linux, kernel
602version 2.4.3, with `lukemftp' 1.5-5, but might happen on other
603systems as well. To avoid this problem, switch to using the standard
604ftp client. On a Debian system, type
a38f41c4 605
9dc15871 606 update-alternatives --config ftp
a38f41c4 607
9dc15871 608and then choose /usr/bin/netkit-ftp.
a38f41c4 609
9dc15871 610*** JPEG images aren't displayed.
b87207a0 611
9dc15871
EZ
612This has been reported when Emacs is built with jpeg-6a library.
613Upgrading to jpeg-6b solves the problem. Configure checks for the
614correct version, but this problem could occur if a binary built
615against a shared libjpeg is run on a system with an older version.
b87207a0 616
9dc15871 617*** Dired is very slow.
4e0bd469
EZ
618
619This could happen if invocation of the `df' program takes a long
620time. Possible reasons for this include:
621
622 - ClearCase mounted filesystems (VOBs) that sometimes make `df'
623 response time extremely slow (dozens of seconds);
624
625 - slow automounters on some old versions of Unix;
626
627 - slow operation of some versions of `df'.
628
629To work around the problem, you could either (a) set the variable
630`directory-free-space-program' to nil, and thus prevent Emacs from
631invoking `df'; (b) use `df' from the GNU Fileutils package; or
632(c) use CVS, which is Free Software, instead of ClearCase.
633
9dc15871 634*** Versions of the W3 package released before Emacs 21.1 don't run
fc2938d1 635under Emacs 21. This fixed in W3 version 4.0pre.47.
a953a8d3 636
9dc15871 637*** The LDAP support rely on ldapsearch program from OpenLDAP version 2.
8f4df059
PJ
638
639It can fail to work with ldapsearch program from OpenLDAP version 1.
640Version 1 of OpenLDAP is now deprecated. If you are still using it,
641please upgrade to version 2. As a temporary workaround, remove
642argument "-x" from the variable `ldap-ldapsearch-args'.
643
9dc15871 644*** ps-print commands fail to find prologue files ps-prin*.ps.
cc2f2825
EZ
645
646This can happen if you use an old version of X-Symbol package: it
647defines compatibility functions which trick ps-print into thinking it
648runs in XEmacs, and look for the prologue files in a wrong directory.
649
650The solution is to upgrade X-Symbol to a later version.
651
9dc15871 652*** On systems with shared libraries you might encounter run-time errors
f4f4ee4d
GM
653from the dynamic linker telling you that it is unable to find some
654shared libraries, for instance those for Xaw3d or image support.
655These errors mean Emacs has been linked with a library whose shared
656library is not in the default search path of the dynamic linker.
657
0cb26e21
EZ
658Similar problems could prevent Emacs from building, since the build
659process invokes Emacs several times.
660
f4f4ee4d
GM
661On many systems, it is possible to set LD_LIBRARY_PATH in your
662environment to specify additional directories where shared libraries
663can be found.
664
665Other systems allow to set LD_RUN_PATH in a similar way, but before
666Emacs is linked. With LD_RUN_PATH set, the linker will include a
667specified run-time search path in the executable.
668
8643647c 669On some systems, Emacs can crash due to problems with dynamic
c31138a1
EZ
670linking. Specifically, on SGI Irix 6.5, crashes were reported with
671backtraces like this:
672
673 (dbx) where
674 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]
675 1 general_find_symbol(0xf49239d, 0x0, 0x0, 0x0, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2)
676 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":2140, 0xfb65a98]
677 2 resolve_symbol(0xf49239d, 0x4031184, 0x0, 0xfbdd438, 0x0, 0xf4923aa, 0x0, 0x492ddb2)
678 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":1947, 0xfb657e4]
679 3 lazy_text_resolve(0xd18, 0x1a3, 0x40302b4, 0x12, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2)
680 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":997, 0xfb64d44]
681 4 _rld_text_resolve(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0)
682 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld_bridge.s":175, 0xfb6032c]
683
8643647c
RS
684(`rld' is the dynamic linker.) We don't know yet why this
685happens, but setting the environment variable LD_BIND_NOW to 1 (which
c31138a1
EZ
686forces the dynamic linker to bind all shared objects early on) seems
687to work around the problem.
688
f4f4ee4d
GM
689Please refer to the documentation of your dynamic linker for details.
690
9dc15871
EZ
691*** You request inverse video, and the first Emacs frame is in inverse
692video, but later frames are not in inverse video.
b1739b51 693
9dc15871
EZ
694This can happen if you have an old version of the custom library in
695your search path for Lisp packages. Use M-x list-load-path-shadows to
696check whether this is true. If it is, delete the old custom library.
b1739b51 697
9dc15871 698*** When you run Ispell from Emacs, it reports a "misalignment" error.
b1739b51 699
9dc15871
EZ
700This can happen if you compiled the Ispell program to use ASCII
701characters only and then try to use it from Emacs with non-ASCII
702characters, like Latin-1. The solution is to recompile Ispell with
703support for 8-bit characters.
b1739b51 704
9dc15871
EZ
705To see whether your Ispell program supports 8-bit characters, type
706this at your shell's prompt:
b1739b51 707
9dc15871 708 ispell -vv
b1739b51 709
9dc15871
EZ
710and look in the output for the string "NO8BIT". If Ispell says
711"!NO8BIT (8BIT)", your speller supports 8-bit characters; otherwise it
712does not.
e9a52cfe 713
9dc15871
EZ
714To rebuild Ispell with 8-bit character support, edit the local.h file
715in the Ispell distribution and make sure it does _not_ define NO8BIT.
716Then rebuild the speller.
e9a52cfe 717
9dc15871
EZ
718Another possible cause for "misalignment" error messages is that the
719version of Ispell installed on your machine is old. Upgrade.
e9a52cfe 720
9dc15871
EZ
721Yet another possibility is that you are trying to spell-check a word
722in a language that doesn't fit the dictionary you choose for use by
723Ispell. (Ispell can only spell-check one language at a time, because
724it uses a single dictionary.) Make sure that the text you are
725spelling and the dictionary used by Ispell conform to each other.
e9a52cfe 726
9dc15871
EZ
727If your spell-checking program is Aspell, it has been reported that if
728you have a personal configuration file (normally ~/.aspell.conf), it
729can cause this error. Remove that file, execute `ispell-kill-ispell'
730in Emacs, and then try spell-checking again.
e9a52cfe 731
9dc15871 732* Runtime problems related to font handling
e9a52cfe 733
b1446261
CY
734** Characters are displayed as empty boxes or with wrong font under X.
735
736*** This can occur when two different versions of FontConfig are used.
737For example, XFree86 4.3.0 has one version and Gnome usually comes
738with a newer version. Emacs compiled with Gtk+ will then use the
739newer version. In most cases the problem can be temporarily fixed by
740stopping the application that has the error (it can be Emacs or any
741other application), removing ~/.fonts.cache-1, and then start the
742application again. If removing ~/.fonts.cache-1 and restarting
743doesn't help, the application with problem must be recompiled with the
744same version of FontConfig as the rest of the system uses. For KDE,
745it is sufficient to recompile Qt.
746
747*** Some fonts have a missing glyph and no default character. This is
748known to occur for character number 160 (no-break space) in some
749fonts, such as Lucida but Emacs sets the display table for the unibyte
750and Latin-1 version of this character to display a space.
751
752*** Some of the fonts called for in your fontset may not exist on your
753X server.
e9a52cfe 754
9dc15871
EZ
755Each X11 font covers just a fraction of the characters that Emacs
756supports. To display the whole range of Emacs characters requires
b1446261
CY
757many different fonts, collected into a fontset. You can remedy the
758problem by installing additional fonts.
f25eb4f7 759
9dc15871 760The intlfonts distribution includes a full spectrum of fonts that can
9222ba5e
EZ
761display all the characters Emacs supports. The etl-unicode collection
762of fonts (available from <URL:ftp://ftp.x.org/contrib/fonts/> and
763<URL:ftp://ftp.xfree86.org/pub/mirror/X.Org/contrib/fonts/>) includes
764fonts that can display many Unicode characters; they can also be used
765by ps-print and ps-mule to print Unicode characters.
f25eb4f7 766
9dc15871 767** Under X11, some characters appear improperly aligned in their lines.
f25eb4f7 768
9222ba5e 769You may have bad X11 fonts; try installing the intlfonts distribution
b1446261
CY
770or the etl-unicode collection (see above).
771
772** Under X, an unexpected monospace font is used as the default font.
773
774When compiled with XFT, Emacs tries to use a default font named
775"monospace". This is a "virtual font", which the operating system
776(Fontconfig) redirects to a suitable font such as DejaVu Sans Mono.
777On some systems, there exists a font that is actually named Monospace,
778which takes over the virtual font. This is considered an operating
779system bug; see
f25eb4f7 780
b1446261 781http://lists.gnu.org/archive/html/emacs-devel/2008-10/msg00696.html
edd7d3be 782
b1446261
CY
783If you encounter this problem, set the default font to a specific font
784in your .Xresources or initialization file. For instance, you can put
785the following in your .Xresources:
786
787Emacs.font: DejaVu Sans Mono 12
788
789** Certain fonts make each line take one pixel more than it should.
790
791This is because these fonts contain characters a little taller than
792the font's nominal height. Emacs needs to make sure that lines do not
793overlap.
edd7d3be 794
9dc15871 795** Loading fonts is very slow.
edd7d3be 796
9dc15871
EZ
797You might be getting scalable fonts instead of precomputed bitmaps.
798Known scalable font directories are "Type1" and "Speedo". A font
799directory contains scalable fonts if it contains the file
800"fonts.scale".
42303132 801
9dc15871
EZ
802If this is so, re-order your X windows font path to put the scalable
803font directories last. See the documentation of `xset' for details.
42303132 804
9dc15871
EZ
805With some X servers, it may be necessary to take the scalable font
806directories out of your path entirely, at least for Emacs 19.26.
807Changes in the future may make this unnecessary.
42303132 808
9dc15871 809** Font Lock displays portions of the buffer in incorrect faces.
42303132 810
9dc15871
EZ
811By far the most frequent cause of this is a parenthesis `(' or a brace
812`{' in column zero. Font Lock assumes that such a paren is outside of
813any comment or string. This is of course not true in general, but the
814vast majority of well-formatted program source files don't have such
815parens, and therefore this assumption is used to allow optimizations
816in Font Lock's syntactical analysis. These optimizations avoid some
817pathological cases where jit-lock, the Just-in-Time fontification
818introduced with Emacs 21.1, could significantly slow down scrolling
819through the buffer, especially scrolling backwards, and also jumping
820to the end of a very large buffer.
42303132 821
bf247b6e 822Beginning with version 22.1, a parenthesis or a brace in column zero
9dc15871
EZ
823is highlighted in bold-red face if it is inside a string or a comment,
824to indicate that it could interfere with Font Lock (and also with
825indentation) and should be moved or escaped with a backslash.
42303132 826
9dc15871
EZ
827If you don't use large buffers, or have a very fast machine which
828makes the delays insignificant, you can avoid the incorrect
829fontification by setting the variable
830`font-lock-beginning-of-syntax-function' to a nil value. (This must
831be done _after_ turning on Font Lock.)
f3d6f4ee 832
9dc15871
EZ
833Another alternative is to avoid a paren in column zero. For example,
834in a Lisp string you could precede the paren with a backslash.
f3d6f4ee 835
9dc15871
EZ
836** With certain fonts, when the cursor appears on a character, the
837character doesn't appear--you get a solid box instead.
f3d6f4ee 838
9dc15871
EZ
839One user on a Linux-based GNU system reported that this problem went
840away with installation of a new X server. The failing server was
841XFree86 3.1.1. XFree86 3.1.2 works.
f3d6f4ee 842
9dc15871 843** Emacs pauses for several seconds when changing the default font.
9f83d8b3 844
9dc15871
EZ
845This has been reported for fvwm 2.2.5 and the window manager of KDE
8462.1. The reason for the pause is Xt waiting for a ConfigureNotify
847event from the window manager, which the window manager doesn't send.
848Xt stops waiting after a default timeout of usually 5 seconds.
f29d1e75 849
9dc15871 850A workaround for this is to add something like
f29d1e75 851
9dc15871 852emacs.waitForWM: false
c24be289 853
9dc15871
EZ
854to your X resources. Alternatively, add `(wait-for-wm . nil)' to a
855frame's parameter list, like this:
c24be289 856
9dc15871 857 (modify-frame-parameters nil '((wait-for-wm . nil)))
b35319bf 858
9dc15871 859(this should go into your `.emacs' file).
b35319bf 860
9dc15871 861** Underlines appear at the wrong position.
b35319bf 862
9dc15871
EZ
863This is caused by fonts having a wrong UNDERLINE_POSITION property.
864Examples are the font 7x13 on XFree prior to version 4.1, or the jmk
6fc3871e
GM
865neep font from the Debian xfonts-jmk package prior to version 3.0.17.
866To circumvent this problem, set x-use-underline-position-properties
867to nil in your `.emacs'.
b35319bf 868
9dc15871
EZ
869To see what is the value of UNDERLINE_POSITION defined by the font,
870type `xlsfonts -lll FONT' and look at the font's UNDERLINE_POSITION
871property.
787994b7 872
9dc15871 873** When using Exceed, fonts sometimes appear too tall.
0a2eeca1 874
9dc15871
EZ
875When the display is set to an Exceed X-server and fonts are specified
876(either explicitly with the -fn option or implicitly with X resources)
877then the fonts may appear "too tall". The actual character sizes are
878correct but there is too much vertical spacing between rows, which
879gives the appearance of "double spacing".
0a2eeca1 880
9dc15871
EZ
881To prevent this, turn off the Exceed's "automatic font substitution"
882feature (in the font part of the configuration window).
0a2eeca1 883
405b495f
GM
884** Subscript/superscript text in TeX is hard to read.
885
886If `tex-fontify-script' is non-nil, tex-mode displays
887subscript/superscript text in the faces subscript/superscript, which
888are smaller than the normal font and lowered/raised. With some fonts,
889nested superscripts (say) can be hard to read. Switching to a
890different font, or changing your antialiasing setting (on an LCD
891screen), can both make the problem disappear. Alternatively, customize
892the following variables: tex-font-script-display (how much to
893lower/raise); tex-suscript-height-ratio (how much smaller than
894normal); tex-suscript-height-minimum (minimum height).
895
9dc15871 896* Internationalization problems
0a2eeca1 897
de25ebb8
RS
898** M-{ does not work on a Spanish PC keyboard.
899
900Many Spanish keyboards seem to ignore that combination. Emacs can't
901do anything about it.
902
9dc15871 903** Characters from the mule-unicode charsets aren't displayed under X.
0a2eeca1 904
9dc15871
EZ
905XFree86 4 contains many fonts in iso10646-1 encoding which have
906minimal character repertoires (whereas the encoding part of the font
907name is meant to be a reasonable indication of the repertoire
908according to the XLFD spec). Emacs may choose one of these to display
909characters from the mule-unicode charsets and then typically won't be
910able to find the glyphs to display many characters. (Check with C-u
911C-x = .) To avoid this, you may need to use a fontset which sets the
912font for the mule-unicode sets explicitly. E.g. to use GNU unifont,
913include in the fontset spec:
0a2eeca1 914
9dc15871
EZ
915mule-unicode-2500-33ff:-gnu-unifont-*-iso10646-1,\
916mule-unicode-e000-ffff:-gnu-unifont-*-iso10646-1,\
917mule-unicode-0100-24ff:-gnu-unifont-*-iso10646-1
0a2eeca1 918
9dc15871 919** The UTF-8/16/7 coding systems don't encode CJK (Far Eastern) characters.
0a2eeca1 920
ce9b56fe
KH
921Emacs directly supports the Unicode BMP whose code points are in the
922ranges 0000-33ff and e000-ffff, and indirectly supports the parts of
923CJK characters belonging to these legacy charsets:
924
925 GB2312, Big5, JISX0208, JISX0212, JISX0213-1, JISX0213-2, KSC5601
926
927The latter support is done in Utf-Translate-Cjk mode (turned on by
928default). Which Unicode CJK characters are decoded into which Emacs
929charset is decided by the current language environment. For instance,
930in Chinese-GB, most of them are decoded into chinese-gb2312.
119d3665 931
9dc15871
EZ
932If you read UTF-8 data with code points outside these ranges, the
933characters appear in the buffer as raw bytes of the original UTF-8
934(composed into a single quasi-character) and they will be written back
935correctly as UTF-8, assuming you don't break the composed sequences.
936If you read such characters from UTF-16 or UTF-7 data, they are
937substituted with the Unicode `replacement character', and you lose
938information.
119d3665 939
9dc15871 940** Mule-UCS loads very slowly.
a933dad1 941
9dc15871
EZ
942Changes to Emacs internals interact badly with Mule-UCS's `un-define'
943library, which is the usual interface to Mule-UCS. Apply the
944following patch to Mule-UCS 0.84 and rebuild it. That will help,
945though loading will still be slower than in Emacs 20. (Some
946distributions, such as Debian, may already have applied such a patch.)
0de9f9a8 947
9dc15871
EZ
948--- lisp/un-define.el 6 Mar 2001 22:41:38 -0000 1.30
949+++ lisp/un-define.el 19 Apr 2002 18:34:26 -0000
950@@ -610,13 +624,21 @@ by calling post-read-conversion and pre-
0de9f9a8 951
9dc15871
EZ
952 (mapcar
953 (lambda (x)
954- (mapcar
955- (lambda (y)
956- (mucs-define-coding-system
957- (nth 0 y) (nth 1 y) (nth 2 y)
958- (nth 3 y) (nth 4 y) (nth 5 y) (nth 6 y))
959- (coding-system-put (car y) 'alias-coding-systems (list (car x))))
960- (cdr x)))
961+ (if (fboundp 'register-char-codings)
962+ ;; Mule 5, where we don't need the eol-type specified and
963+ ;; register-char-codings may be very slow for these coding
964+ ;; system definitions.
965+ (let ((y (cadr x)))
3256a475
RC
966+ (mucs-define-coding-system
967+ (car x) (nth 1 y) (nth 2 y)
968+ (nth 3 y) (nth 4 y) (nth 5 y)))
9dc15871
EZ
969+ (mapcar
970+ (lambda (y)
3256a475
RC
971+ (mucs-define-coding-system
972+ (nth 0 y) (nth 1 y) (nth 2 y)
973+ (nth 3 y) (nth 4 y) (nth 5 y) (nth 6 y))
974+ (coding-system-put (car y) 'alias-coding-systems (list (car x)))))
9dc15871
EZ
975+ (cdr x)))
976 `((utf-8
977 (utf-8-unix
978 ?u "UTF-8 coding system"
0de9f9a8 979
9dc15871
EZ
980Note that Emacs has native support for Unicode, roughly equivalent to
981Mule-UCS's, so you may not need it.
a933dad1 982
d87ceee0
KH
983** Mule-UCS compilation problem.
984
985Emacs of old versions and XEmacs byte-compile the form `(progn progn
986...)' the same way as `(progn ...)', but Emacs of version 21.3 and the
987later process that form just as interpreter does, that is, as `progn'
988variable reference. Apply the following patch to Mule-UCS 0.84 to
989make it compiled by the latest Emacs.
990
991--- mucs-ccl.el 2 Sep 2005 00:42:23 -0000 1.1.1.1
992+++ mucs-ccl.el 2 Sep 2005 01:31:51 -0000 1.3
993@@ -639,10 +639,14 @@
994 (mucs-notify-embedment 'mucs-ccl-required name)
995 (setq ccl-pgm-list (cdr ccl-pgm-list)))
996 ; (message "MCCLREGFIN:%S" result)
997- `(progn
998- (setq mucs-ccl-facility-alist
3256a475 999- (quote ,mucs-ccl-facility-alist))
d87ceee0
KH
1000- ,@result)))
1001+ ;; The only way the function is used in this package is included
1002+ ;; in `mucs-package-definition-end-hook' value, where it must
1003+ ;; return (possibly empty) *list* of forms. Do this. Do not rely
ade79051 1004+ ;; on byte compiler to remove extra `progn's in `(progn ...)'
d87ceee0
KH
1005+ ;; form.
1006+ `((setq mucs-ccl-facility-alist
3256a475 1007+ (quote ,mucs-ccl-facility-alist))
d87ceee0 1008+ ,@result)))
ade79051 1009
d87ceee0
KH
1010 ;;; Add hook for embedding translation informations to a package.
1011 (add-hook 'mucs-package-definition-end-hook
1012
9dc15871 1013** Accented ISO-8859-1 characters are displayed as | or _.
a933dad1 1014
d6b7de9b
EZ
1015Try other font set sizes (S-mouse-1). If the problem persists with
1016other sizes as well, your text is corrupted, probably through software
1017that is not 8-bit clean. If the problem goes away with another font
1018size, it's probably because some fonts pretend to be ISO-8859-1 fonts
1019when they are really ASCII fonts. In particular the schumacher-clean
1020fonts have this bug in some versions of X.
a933dad1 1021
d6b7de9b 1022To see what glyphs are included in a font, use `xfd', like this:
a933dad1 1023
d6b7de9b 1024 xfd -fn -schumacher-clean-medium-r-normal--12-120-75-75-c-60-iso8859-1
a933dad1 1025
d6b7de9b
EZ
1026If this shows only ASCII glyphs, the font is indeed the source of the
1027problem.
a933dad1 1028
d6b7de9b
EZ
1029The solution is to remove the corresponding lines from the appropriate
1030`fonts.alias' file, then run `mkfontdir' in that directory, and then run
1031`xset fp rehash'.
177c0ea7 1032
9dc15871 1033** The `oc-unicode' package doesn't work with Emacs 21.
a933dad1 1034
9dc15871
EZ
1035This package tries to define more private charsets than there are free
1036slots now. The current built-in Unicode support is actually more
1037flexible. (Use option `utf-translate-cjk-mode' if you need CJK
1038support.) Files encoded as emacs-mule using oc-unicode aren't
1039generally read correctly by Emacs 21.
a933dad1 1040
9dc15871 1041** After a while, Emacs slips into unibyte mode.
a933dad1 1042
9dc15871
EZ
1043The VM mail package, which is not part of Emacs, sometimes does
1044 (standard-display-european t)
1045That should be changed to
1046 (standard-display-european 1 t)
de121241 1047
9dc15871 1048* X runtime problems
de121241 1049
9dc15871 1050** X keyboard problems
de121241 1051
9dc15871 1052*** You "lose characters" after typing Compose Character key.
a933dad1 1053
9dc15871
EZ
1054This is because the Compose Character key is defined as the keysym
1055Multi_key, and Emacs (seeing that) does the proper X11
1056character-composition processing. If you don't want your Compose key
1057to do that, you can redefine it with xmodmap.
a933dad1 1058
9dc15871 1059For example, here's one way to turn it into a Meta key:
177c0ea7 1060
9dc15871 1061 xmodmap -e "keysym Multi_key = Meta_L"
a933dad1 1062
9dc15871
EZ
1063If all users at your site of a particular keyboard prefer Meta to
1064Compose, you can make the remapping happen automatically by adding the
1065xmodmap command to the xdm setup script for that display.
a933dad1 1066
9dc15871 1067*** Using X Windows, control-shift-leftbutton makes Emacs hang.
a933dad1 1068
9dc15871 1069Use the shell command `xset bc' to make the old X Menu package work.
a933dad1 1070
9f4f9273 1071*** C-SPC fails to work on Fedora GNU/Linux (or with fcitx input method).
09352e8f
RS
1072
1073Fedora Core 4 steals the C-SPC key by default for the `iiimx' program
1074which is the input method for some languages. It blocks Emacs users
1075from using the C-SPC key for `set-mark-command'.
1076
1077One solutions is to remove the `<Ctrl>space' from the `Iiimx' file
1078which can be found in the `/usr/lib/X11/app-defaults' directory.
1079However, that requires root access.
1080
1081Another is to specify `Emacs*useXIM: false' in your X resources.
1082
1083Another is to build Emacs with the `--without-xim' configure option.
1084
2fb18d13
KH
1085The same problem happens on any other system if you are using fcitx
1086(Chinese input method) which by default use C-SPC for toggling. If
1087you want to use fcitx with Emacs, you have two choices. Toggle fcitx
1088by another key (e.g. C-\) by modifying ~/.fcitx/config, or be
1089accustomed to use C-@ for `set-mark-command'.
1090
9dc15871 1091*** M-SPC seems to be ignored as input.
a933dad1 1092
9dc15871
EZ
1093See if your X server is set up to use this as a command
1094for character composition.
a933dad1 1095
9dc15871 1096*** The S-C-t key combination doesn't get passed to Emacs on X.
a933dad1 1097
9dc15871
EZ
1098This happens because some X configurations assign the Ctrl-Shift-t
1099combination the same meaning as the Multi_key. The offending
1100definition is in the file `...lib/X11/locale/iso8859-1/Compose'; there
1101might be other similar combinations which are grabbed by X for similar
1102purposes.
a933dad1 1103
9dc15871
EZ
1104We think that this can be countermanded with the `xmodmap' utility, if
1105you want to be able to bind one of these key sequences within Emacs.
a933dad1 1106
9dc15871 1107*** Under X, C-v and/or other keys don't work.
a933dad1
DL
1108
1109These may have been intercepted by your window manager. In
1110particular, AfterStep 1.6 is reported to steal C-v in its default
1111configuration. Various Meta keys are also likely to be taken by the
1112configuration of the `feel'. See the WM's documentation for how to
1113change this.
1114
9dc15871 1115*** Clicking C-mouse-2 in the scroll bar doesn't split the window.
a933dad1 1116
9dc15871
EZ
1117This currently doesn't work with scroll-bar widgets (and we don't know
1118a good way of implementing it with widgets). If Emacs is configured
1119--without-toolkit-scroll-bars, C-mouse-2 on the scroll bar does work.
a933dad1 1120
9dc15871
EZ
1121*** Inability to send an Alt-modified key, when Emacs is communicating
1122directly with an X server.
a933dad1 1123
9dc15871
EZ
1124If you have tried to bind an Alt-modified key as a command, and it
1125does not work to type the command, the first thing you should check is
1126whether the key is getting through to Emacs. To do this, type C-h c
1127followed by the Alt-modified key. C-h c should say what kind of event
1128it read. If it says it read an Alt-modified key, then make sure you
1129have made the key binding correctly.
a933dad1 1130
9dc15871
EZ
1131If C-h c reports an event that doesn't have the Alt modifier, it may
1132be because your X server has no key for the Alt modifier. The X
1133server that comes from MIT does not set up the Alt modifier by
1134default.
a933dad1 1135
9dc15871 1136If your keyboard has keys named Alt, you can enable them as follows:
a933dad1 1137
9dc15871
EZ
1138 xmodmap -e 'add mod2 = Alt_L'
1139 xmodmap -e 'add mod2 = Alt_R'
a933dad1 1140
9dc15871
EZ
1141If the keyboard has just one key named Alt, then only one of those
1142commands is needed. The modifier `mod2' is a reasonable choice if you
1143are using an unmodified MIT version of X. Otherwise, choose any
1144modifier bit not otherwise used.
a933dad1 1145
9dc15871
EZ
1146If your keyboard does not have keys named Alt, you can use some other
1147keys. Use the keysym command in xmodmap to turn a function key (or
1148some other 'spare' key) into Alt_L or into Alt_R, and then use the
1149commands show above to make them modifier keys.
a933dad1 1150
9dc15871
EZ
1151Note that if you have Alt keys but no Meta keys, Emacs translates Alt
1152into Meta. This is because of the great importance of Meta in Emacs.
a933dad1 1153
9dc15871 1154** Window-manager and toolkit-related problems
a933dad1 1155
0e71e4a8
CY
1156*** Gnome: Emacs receives input directly from the keyboard, bypassing XIM.
1157
1158This seems to happen when gnome-settings-daemon version 2.12 or later
1159is running. If gnome-settings-daemon is not running, Emacs receives
1160input through XIM without any problem. Furthermore, this seems only
1161to happen in *.UTF-8 locales; zh_CN.GB2312 and zh_CN.GBK locales, for
1162example, work fine. A bug report has been filed in the Gnome
1163bugzilla: http://bugzilla.gnome.org/show_bug.cgi?id=357032
1164
9dc15871 1165*** Gnome: Emacs' xterm-mouse-mode doesn't work on the Gnome terminal.
a933dad1 1166
9dc15871
EZ
1167A symptom of this bug is that double-clicks insert a control sequence
1168into the buffer. The reason this happens is an apparent
1169incompatibility of the Gnome terminal with Xterm, which also affects
1170other programs using the Xterm mouse interface. A problem report has
1171been filed.
a933dad1 1172
9dc15871
EZ
1173*** KDE: When running on KDE, colors or fonts are not as specified for Emacs,
1174or messed up.
a933dad1 1175
9dc15871
EZ
1176For example, you could see background you set for Emacs only in the
1177empty portions of the Emacs display, while characters have some other
1178background.
a933dad1 1179
9dc15871
EZ
1180This happens because KDE's defaults apply its color and font
1181definitions even to applications that weren't compiled for KDE. The
1182solution is to uncheck the "Apply fonts and colors to non-KDE apps"
1183option in Preferences->Look&Feel->Style (KDE 2). In KDE 3, this option
1184is in the "Colors" section, rather than "Style".
a933dad1 1185
9dc15871
EZ
1186Alternatively, if you do want the KDE defaults to apply to other
1187applications, but not to Emacs, you could modify the file `Emacs.ad'
1188(should be in the `/usr/share/apps/kdisplay/app-defaults/' directory)
1189so that it doesn't set the default background and foreground only for
1190Emacs. For example, make sure the following resources are either not
1191present or commented out:
a933dad1 1192
9dc15871
EZ
1193 Emacs.default.attributeForeground
1194 Emacs.default.attributeBackground
1195 Emacs*Foreground
1196 Emacs*Background
a933dad1 1197
a3475659
JD
1198It is also reported that a bug in the gtk-engines-qt engine can cause this if
1199Emacs is compiled with Gtk+.
1200The bug is fixed in version 0.7 or newer of gtk-engines-qt.
1201
9dc15871 1202*** KDE: Emacs hangs on KDE when a large portion of text is killed.
a933dad1 1203
9dc15871
EZ
1204This is caused by a bug in the KDE applet `klipper' which periodically
1205requests the X clipboard contents from applications. Early versions
8576f724 1206of klipper don't implement the ICCCM protocol for large selections,
9dc15871 1207which leads to Emacs being flooded with selection requests. After a
b11e8823 1208while, Emacs may print a message:
a933dad1 1209
9dc15871 1210 Timed out waiting for property-notify event
a933dad1 1211
b11e8823
JD
1212A workaround is to not use `klipper'. An upgrade to the `klipper' that
1213comes with KDE 3.3 or later also solves the problem.
a933dad1 1214
9dc15871 1215*** CDE: Frames may cover dialogs they created when using CDE.
a933dad1 1216
9dc15871
EZ
1217This can happen if you have "Allow Primary Windows On Top" enabled which
1218seems to be the default in the Common Desktop Environment.
1219To change, go in to "Desktop Controls" -> "Window Style Manager"
1220and uncheck "Allow Primary Windows On Top".
d238f982 1221
9dc15871
EZ
1222*** Xaw3d : When using Xaw3d scroll bars without arrows, the very first mouse
1223click in a scroll bar might be ignored by the scroll bar widget. This
1224is probably a bug in Xaw3d; when Xaw3d is compiled with arrows, the
1225problem disappears.
d238f982 1226
9dc15871
EZ
1227*** Xaw: There are known binary incompatibilities between Xaw, Xaw3d, neXtaw,
1228XawM and the few other derivatives of Xaw. So when you compile with
1229one of these, it may not work to dynamically link with another one.
1230For example, strange problems, such as Emacs exiting when you type
1231"C-x 1", were reported when Emacs compiled with Xaw3d and libXaw was
1232used with neXtaw at run time.
d7185f9d 1233
9dc15871
EZ
1234The solution is to rebuild Emacs with the toolkit version you actually
1235want to use, or set LD_PRELOAD to preload the same toolkit version you
1236built Emacs with.
d7185f9d 1237
9dc15871 1238*** Open Motif: Problems with file dialogs in Emacs built with Open Motif.
a933dad1 1239
9dc15871
EZ
1240When Emacs 21 is built with Open Motif 2.1, it can happen that the
1241graphical file dialog boxes do not work properly. The "OK", "Filter"
1242and "Cancel" buttons do not respond to mouse clicks. Dragging the
1243file dialog window usually causes the buttons to work again.
a933dad1 1244
9dc15871
EZ
1245The solution is to use LessTif instead. LessTif is a free replacement
1246for Motif. See the file INSTALL for information on how to do this.
a933dad1 1247
9dc15871
EZ
1248Another workaround is not to use the mouse to trigger file prompts,
1249but to use the keyboard. This way, you will be prompted for a file in
1250the minibuffer instead of a graphical file dialog.
a933dad1 1251
9dc15871 1252*** LessTif: Problems in Emacs built with LessTif.
a933dad1 1253
9dc15871
EZ
1254The problems seem to depend on the version of LessTif and the Motif
1255emulation for which it is set up.
a933dad1 1256
9dc15871 1257Only the Motif 1.2 emulation seems to be stable enough in LessTif.
880ea925 1258LessTif 0.92-17's Motif 1.2 emulation seems to work okay on FreeBSD.
9dc15871
EZ
1259On GNU/Linux systems, lesstif-0.92.6 configured with "./configure
1260--enable-build-12 --enable-default-12" is reported to be the most
1261successful. The binary GNU/Linux package
1262lesstif-devel-0.92.0-1.i386.rpm was reported to have problems with
1263menu placement.
a933dad1 1264
9dc15871
EZ
1265On some systems, even with Motif 1.2 emulation, Emacs occasionally
1266locks up, grabbing all mouse and keyboard events. We still don't know
1267what causes these problems; they are not reproducible by Emacs
1268developers.
a933dad1 1269
9dc15871 1270*** Motif: The Motif version of Emacs paints the screen a solid color.
a933dad1 1271
9dc15871 1272This has been observed to result from the following X resource:
a933dad1 1273
9dc15871 1274 Emacs*default.attributeFont: -*-courier-medium-r-*-*-*-140-*-*-*-*-iso8859-*
a933dad1 1275
9dc15871
EZ
1276That the resource has this effect indicates a bug in something, but we
1277do not yet know what. If it is an Emacs bug, we hope someone can
1278explain what the bug is so we can fix it. In the mean time, removing
1279the resource prevents the problem.
a933dad1 1280
9dc15871 1281** General X problems
17a37d87 1282
9dc15871 1283*** Redisplay using X11 is much slower than previous Emacs versions.
17a37d87 1284
9dc15871
EZ
1285We've noticed that certain X servers draw the text much slower when
1286scroll bars are on the left. We don't know why this happens. If this
1287happens to you, you can work around it by putting the scroll bars
1288on the right (as they were in Emacs 19).
17a37d87 1289
9dc15871 1290Here's how to do this:
17a37d87 1291
9dc15871 1292 (set-scroll-bar-mode 'right)
a933dad1 1293
9dc15871
EZ
1294If you're not sure whether (or how much) this problem affects you,
1295try that and see how much difference it makes. To set things back
1296to normal, do
a933dad1 1297
9dc15871 1298 (set-scroll-bar-mode 'left)
a933dad1 1299
9dc15871 1300*** Error messages about undefined colors on X.
a933dad1 1301
9dc15871 1302The messages might say something like this:
a933dad1 1303
9dc15871 1304 Unable to load color "grey95"
a933dad1 1305
9dc15871 1306(typically, in the `*Messages*' buffer), or something like this:
a933dad1 1307
9dc15871 1308 Error while displaying tooltip: (error Undefined color lightyellow)
a933dad1 1309
9dc15871
EZ
1310These problems could happen if some other X program has used up too
1311many colors of the X palette, leaving Emacs with insufficient system
1312resources to load all the colors it needs.
a933dad1 1313
9dc15871 1314A solution is to exit the offending X programs before starting Emacs.
a933dad1 1315
9257b627
EZ
1316"undefined color" messages can also occur if the RgbPath entry in the
1317X configuration file is incorrect, or the rgb.txt file is not where
1318X expects to find it.
1319
9dc15871 1320*** Improving performance with slow X connections.
a933dad1 1321
9dc15871
EZ
1322There are several ways to improve this performance, any subset of which can
1323be carried out at the same time:
a933dad1 1324
9dc15871
EZ
13251) If you don't need X Input Methods (XIM) for entering text in some
1326 language you use, you can improve performance on WAN links by using
1327 the X resource useXIM to turn off use of XIM. This does not affect
1328 the use of Emacs' own input methods, which are part of the Leim
1329 package.
3d00585e 1330
9dc15871 13312) If the connection is very slow, you might also want to consider
634e516b
EZ
1332 switching off scroll bars, menu bar, and tool bar. Adding the
1333 following forms to your .emacs file will accomplish that, but only
1334 after the the initial frame is displayed:
1335
1336 (scroll-bar-mode -1)
1337 (menu-bar-mode -1)
1338 (tool-bar-mode -1)
1339
1340 For still quicker startup, put these X resources in your .Xdefaults
1341 file:
1342
1343 Emacs.verticalScrollBars: off
1344 Emacs.menuBar: off
1345 Emacs.toolBar: off
3d00585e 1346
9dc15871
EZ
13473) Use ssh to forward the X connection, and enable compression on this
1348 forwarded X connection (ssh -XC remotehostname emacs ...).
3d00585e 1349
9dc15871
EZ
13504) Use lbxproxy on the remote end of the connection. This is an interface
1351 to the low bandwidth X extension in most modern X servers, which
1352 improves performance dramatically, at the slight expense of correctness
1353 of the X protocol. lbxproxy acheives the performance gain by grouping
1354 several X requests in one TCP packet and sending them off together,
880ea925 1355 instead of requiring a round-trip for each X request in a separate
9dc15871
EZ
1356 packet. The switches that seem to work best for emacs are:
1357 -noatomsfile -nowinattr -cheaterrors -cheatevents
1358 Note that the -nograbcmap option is known to cause problems.
1359 For more about lbxproxy, see:
1360 http://www.xfree86.org/4.3.0/lbxproxy.1.html
3d00585e 1361
34431988
KS
13625) If copying and killing is slow, try to disable the interaction with the
1363 native system's clipboard by adding these lines to your .emacs file:
1364 (setq interprogram-cut-function nil)
1365 (setq interprogram-paste-function nil)
1366
9dc15871 1367*** Emacs gives the error, Couldn't find per display information.
3d00585e 1368
9dc15871
EZ
1369This can result if the X server runs out of memory because Emacs uses
1370a large number of fonts. On systems where this happens, C-h h is
1371likely to cause it.
a933dad1 1372
9dc15871 1373We do not know of a way to prevent the problem.
7838ea1b 1374
9dc15871 1375*** Emacs does not notice when you release the mouse.
a933dad1 1376
9dc15871
EZ
1377There are reports that this happened with (some) Microsoft mice and
1378that replacing the mouse made it stop.
a933dad1 1379
9dc15871 1380*** You can't select from submenus (in the X toolkit version).
a933dad1 1381
9dc15871
EZ
1382On certain systems, mouse-tracking and selection in top-level menus
1383works properly with the X toolkit, but neither of them works when you
1384bring up a submenu (such as Bookmarks or Compare or Apply Patch, in
1385the Files menu).
a933dad1 1386
9dc15871
EZ
1387This works on most systems. There is speculation that the failure is
1388due to bugs in old versions of X toolkit libraries, but no one really
1389knows. If someone debugs this and finds the precise cause, perhaps a
1390workaround can be found.
a933dad1 1391
9dc15871
EZ
1392*** An error message such as `X protocol error: BadMatch (invalid
1393parameter attributes) on protocol request 93'.
b5cb4652 1394
9dc15871
EZ
1395This comes from having an invalid X resource, such as
1396 emacs*Cursor: black
1397(which is invalid because it specifies a color name for something
1398that isn't a color.)
b5cb4652 1399
9dc15871 1400The fix is to correct your X resources.
a933dad1 1401
9dc15871 1402*** Slow startup on X11R6 with X windows.
a933dad1 1403
9dc15871
EZ
1404If Emacs takes two minutes to start up on X11R6, see if your X
1405resources specify any Adobe fonts. That causes the type-1 font
1406renderer to start up, even if the font you asked for is not a type-1
1407font.
a933dad1 1408
9dc15871
EZ
1409One way to avoid this problem is to eliminate the type-1 fonts from
1410your font path, like this:
a933dad1 1411
3256a475 1412 xset -fp /usr/X11R6/lib/X11/fonts/Type1/
a933dad1 1413
9dc15871 1414*** Pull-down menus appear in the wrong place, in the toolkit version of Emacs.
a933dad1 1415
9dc15871 1416An X resource of this form can cause the problem:
a933dad1 1417
9dc15871 1418 Emacs*geometry: 80x55+0+0
a933dad1 1419
9dc15871
EZ
1420This resource is supposed to apply, and does apply, to the menus
1421individually as well as to Emacs frames. If that is not what you
1422want, rewrite the resource.
3156909f 1423
9dc15871
EZ
1424To check thoroughly for such resource specifications, use `xrdb
1425-query' to see what resources the X server records, and also look at
1426the user's ~/.Xdefaults and ~/.Xdefaults-* files.
a933dad1 1427
9dc15871
EZ
1428*** Emacs running under X Windows does not handle mouse clicks.
1429*** `emacs -geometry 80x20' finds a file named `80x20'.
e96c5c69 1430
9dc15871
EZ
1431One cause of such problems is having (setq term-file-prefix nil) in
1432your .emacs file. Another cause is a bad value of EMACSLOADPATH in
1433the environment.
e96c5c69 1434
9dc15871 1435*** Emacs fails to get default settings from X Windows server.
a933dad1 1436
9dc15871
EZ
1437The X library in X11R4 has a bug; it interchanges the 2nd and 3rd
1438arguments to XGetDefaults. Define the macro XBACKWARDS in config.h to
1439tell Emacs to compensate for this.
a933dad1 1440
9dc15871
EZ
1441I don't believe there is any way Emacs can determine for itself
1442whether this problem is present on a given system.
a933dad1 1443
9dc15871 1444*** X Windows doesn't work if DISPLAY uses a hostname.
a933dad1 1445
9dc15871
EZ
1446People have reported kernel bugs in certain systems that cause Emacs
1447not to work with X Windows if DISPLAY is set using a host name. But
1448the problem does not occur if DISPLAY is set to `unix:0.0'. I think
1449the bug has to do with SIGIO or FIONREAD.
a933dad1 1450
9dc15871
EZ
1451You may be able to compensate for the bug by doing (set-input-mode nil nil).
1452However, that has the disadvantage of turning off interrupts, so that
1453you are unable to quit out of a Lisp program by typing C-g.
a933dad1 1454
9dc15871 1455The easy way to do this is to put
a933dad1 1456
9dc15871 1457 (setq x-sigio-bug t)
a933dad1 1458
9dc15871 1459in your site-init.el file.
a933dad1 1460
3256a475
RC
1461*** Prevent double pastes in X
1462
1463The problem: a region, such as a command, is pasted twice when you copy
1464it with your mouse from GNU Emacs to an xterm or an RXVT shell in X.
1465The solution: try the following in your X configuration file,
1466/etc/X11/xorg.conf This should enable both PS/2 and USB mice for
1467single copies. You do not need any other drivers or options.
1468
1469 Section "InputDevice"
1470 Identifier "Generic Mouse"
1471 Driver "mousedev"
1472 Option "Device" "/dev/input/mice"
1473 EndSection
1474
880ea925 1475* Runtime problems on character terminals
a933dad1 1476
9dc15871 1477** Emacs spontaneously displays "I-search: " at the bottom of the screen.
a933dad1 1478
9dc15871
EZ
1479This means that Control-S/Control-Q (XON/XOFF) "flow control" is being
1480used. C-s/C-q flow control is bad for Emacs editors because it takes
1481away C-s and C-q as user commands. Since editors do not output long
1482streams of text without user commands, there is no need for a
1483user-issuable "stop output" command in an editor; therefore, a
1484properly designed flow control mechanism would transmit all possible
1485input characters without interference. Designing such a mechanism is
1486easy, for a person with at least half a brain.
a933dad1 1487
9dc15871 1488There are three possible reasons why flow control could be taking place:
a933dad1 1489
9dc15871
EZ
1490 1) Terminal has not been told to disable flow control
1491 2) Insufficient padding for the terminal in use
1492 3) Some sort of terminal concentrator or line switch is responsible
a933dad1 1493
9dc15871
EZ
1494First of all, many terminals have a set-up mode which controls whether
1495they generate XON/XOFF flow control characters. This must be set to
1496"no XON/XOFF" in order for Emacs to work. Sometimes there is an
1497escape sequence that the computer can send to turn flow control off
1498and on. If so, perhaps the termcap `ti' string should turn flow
1499control off, and the `te' string should turn it on.
a933dad1 1500
9dc15871
EZ
1501Once the terminal has been told "no flow control", you may find it
1502needs more padding. The amount of padding Emacs sends is controlled
1503by the termcap entry for the terminal in use, and by the output baud
1504rate as known by the kernel. The shell command `stty' will print
1505your output baud rate; `stty' with suitable arguments will set it if
1506it is wrong. Setting to a higher speed causes increased padding. If
1507the results are wrong for the correct speed, there is probably a
1508problem in the termcap entry. You must speak to a local Unix wizard
1509to fix this. Perhaps you are just using the wrong terminal type.
a933dad1 1510
9dc15871
EZ
1511For terminals that lack a "no flow control" mode, sometimes just
1512giving lots of padding will prevent actual generation of flow control
1513codes. You might as well try it.
a933dad1 1514
9dc15871
EZ
1515If you are really unlucky, your terminal is connected to the computer
1516through a concentrator which sends XON/XOFF flow control to the
1517computer, or it insists on sending flow control itself no matter how
1518much padding you give it. Unless you can figure out how to turn flow
1519control off on this concentrator (again, refer to your local wizard),
1520you are screwed! You should have the terminal or concentrator
1521replaced with a properly designed one. In the mean time, some drastic
1522measures can make Emacs semi-work.
4c635a29 1523
9dc15871
EZ
1524You can make Emacs ignore C-s and C-q and let the operating system
1525handle them. To do this on a per-session basis, just type M-x
1526enable-flow-control RET. You will see a message that C-\ and C-^ are
1527now translated to C-s and C-q. (Use the same command M-x
1528enable-flow-control to turn *off* this special mode. It toggles flow
1529control handling.)
a933dad1 1530
9dc15871
EZ
1531If C-\ and C-^ are inconvenient for you (for example, if one of them
1532is the escape character of your terminal concentrator), you can choose
1533other characters by setting the variables flow-control-c-s-replacement
1534and flow-control-c-q-replacement. But choose carefully, since all
1535other control characters are already used by emacs.
a933dad1 1536
9dc15871
EZ
1537IMPORTANT: if you type C-s by accident while flow control is enabled,
1538Emacs output will freeze, and you will have to remember to type C-q in
1539order to continue.
a933dad1 1540
9dc15871
EZ
1541If you work in an environment where a majority of terminals of a
1542certain type are flow control hobbled, you can use the function
1543`enable-flow-control-on' to turn on this flow control avoidance scheme
1544automatically. Here is an example:
a933dad1 1545
9dc15871 1546(enable-flow-control-on "vt200" "vt300" "vt101" "vt131")
a933dad1 1547
9dc15871
EZ
1548If this isn't quite correct (e.g. you have a mixture of flow-control hobbled
1549and good vt200 terminals), you can still run enable-flow-control
1550manually.
a933dad1 1551
9dc15871
EZ
1552I have no intention of ever redesigning the Emacs command set for the
1553assumption that terminals use C-s/C-q flow control. XON/XOFF flow
1554control technique is a bad design, and terminals that need it are bad
1555merchandise and should not be purchased. Now that X is becoming
1556widespread, XON/XOFF seems to be on the way out. If you can get some
1557use out of GNU Emacs on inferior terminals, more power to you, but I
1558will not make Emacs worse for properly designed systems for the sake
1559of inferior systems.
a933dad1 1560
9dc15871 1561** Control-S and Control-Q commands are ignored completely.
a933dad1 1562
9dc15871
EZ
1563For some reason, your system is using brain-damaged C-s/C-q flow
1564control despite Emacs's attempts to turn it off. Perhaps your
1565terminal is connected to the computer through a concentrator
1566that wants to use flow control.
a933dad1 1567
9dc15871
EZ
1568You should first try to tell the concentrator not to use flow control.
1569If you succeed in this, try making the terminal work without
1570flow control, as described in the preceding section.
a933dad1 1571
9dc15871
EZ
1572If that line of approach is not successful, map some other characters
1573into C-s and C-q using keyboard-translate-table. The example above
1574shows how to do this with C-^ and C-\.
a933dad1 1575
9dc15871 1576** Screen is updated wrong, but only on one kind of terminal.
a933dad1 1577
9dc15871
EZ
1578This could mean that the termcap entry you are using for that
1579terminal is wrong, or it could mean that Emacs has a bug handing
1580the combination of features specified for that terminal.
a933dad1 1581
9dc15871
EZ
1582The first step in tracking this down is to record what characters
1583Emacs is sending to the terminal. Execute the Lisp expression
1584(open-termscript "./emacs-script") to make Emacs write all
1585terminal output into the file ~/emacs-script as well; then do
1586what makes the screen update wrong, and look at the file
1587and decode the characters using the manual for the terminal.
1588There are several possibilities:
a933dad1 1589
9dc15871 15901) The characters sent are correct, according to the terminal manual.
a933dad1 1591
9dc15871
EZ
1592In this case, there is no obvious bug in Emacs, and most likely you
1593need more padding, or possibly the terminal manual is wrong.
a933dad1 1594
9dc15871
EZ
15952) The characters sent are incorrect, due to an obscure aspect
1596 of the terminal behavior not described in an obvious way
1597 by termcap.
a933dad1 1598
9dc15871
EZ
1599This case is hard. It will be necessary to think of a way for
1600Emacs to distinguish between terminals with this kind of behavior
1601and other terminals that behave subtly differently but are
1602classified the same by termcap; or else find an algorithm for
1603Emacs to use that avoids the difference. Such changes must be
1604tested on many kinds of terminals.
a933dad1 1605
9dc15871 16063) The termcap entry is wrong.
a933dad1 1607
9dc15871
EZ
1608See the file etc/TERMS for information on changes
1609that are known to be needed in commonly used termcap entries
1610for certain terminals.
a933dad1 1611
9dc15871
EZ
16124) The characters sent are incorrect, and clearly cannot be
1613 right for any terminal with the termcap entry you were using.
a933dad1 1614
9dc15871
EZ
1615This is unambiguously an Emacs bug, and can probably be fixed
1616in termcap.c, tparam.c, term.c, scroll.c, cm.c or dispnew.c.
a933dad1 1617
9dc15871 1618** Control-S and Control-Q commands are ignored completely on a net connection.
a933dad1 1619
9dc15871
EZ
1620Some versions of rlogin (and possibly telnet) do not pass flow
1621control characters to the remote system to which they connect.
1622On such systems, emacs on the remote system cannot disable flow
1623control on the local system.
a933dad1 1624
9dc15871
EZ
1625One way to cure this is to disable flow control on the local host
1626(the one running rlogin, not the one running rlogind) using the
1627stty command, before starting the rlogin process. On many systems,
1628"stty start u stop u" will do this.
a933dad1 1629
9dc15871
EZ
1630Some versions of tcsh will prevent even this from working. One way
1631around this is to start another shell before starting rlogin, and
1632issue the stty command to disable flow control from that shell.
a933dad1 1633
9dc15871
EZ
1634If none of these methods work, the best solution is to type
1635M-x enable-flow-control at the beginning of your emacs session, or
1636if you expect the problem to continue, add a line such as the
1637following to your .emacs (on the host running rlogind):
a933dad1 1638
9dc15871 1639(enable-flow-control-on "vt200" "vt300" "vt101" "vt131")
a933dad1 1640
9dc15871
EZ
1641See the entry about spontaneous display of I-search (above) for more
1642info.
a933dad1 1643
9dc15871 1644** Output from Control-V is slow.
a933dad1 1645
9dc15871
EZ
1646On many bit-map terminals, scrolling operations are fairly slow.
1647Often the termcap entry for the type of terminal in use fails
1648to inform Emacs of this. The two lines at the bottom of the screen
1649before a Control-V command are supposed to appear at the top after
1650the Control-V command. If Emacs thinks scrolling the lines is fast,
1651it will scroll them to the top of the screen.
a933dad1 1652
9dc15871
EZ
1653If scrolling is slow but Emacs thinks it is fast, the usual reason is
1654that the termcap entry for the terminal you are using does not
1655specify any padding time for the `al' and `dl' strings. Emacs
1656concludes that these operations take only as much time as it takes to
1657send the commands at whatever line speed you are using. You must
1658fix the termcap entry to specify, for the `al' and `dl', as much
1659time as the operations really take.
a933dad1 1660
9dc15871
EZ
1661Currently Emacs thinks in terms of serial lines which send characters
1662at a fixed rate, so that any operation which takes time for the
1663terminal to execute must also be padded. With bit-map terminals
1664operated across networks, often the network provides some sort of
1665flow control so that padding is never needed no matter how slow
1666an operation is. You must still specify a padding time if you want
1667Emacs to realize that the operation takes a long time. This will
1668cause padding characters to be sent unnecessarily, but they do
1669not really cost much. They will be transmitted while the scrolling
1670is happening and then discarded quickly by the terminal.
a933dad1 1671
9dc15871
EZ
1672Most bit-map terminals provide commands for inserting or deleting
1673multiple lines at once. Define the `AL' and `DL' strings in the
1674termcap entry to say how to do these things, and you will have
1675fast output without wasted padding characters. These strings should
1676each contain a single %-spec saying how to send the number of lines
1677to be scrolled. These %-specs are like those in the termcap
1678`cm' string.
a933dad1 1679
9dc15871
EZ
1680You should also define the `IC' and `DC' strings if your terminal
1681has a command to insert or delete multiple characters. These
1682take the number of positions to insert or delete as an argument.
a933dad1 1683
9dc15871
EZ
1684A `cs' string to set the scrolling region will reduce the amount
1685of motion you see on the screen when part of the screen is scrolled.
a933dad1 1686
9dc15871 1687** You type Control-H (Backspace) expecting to delete characters.
a933dad1 1688
9dc15871
EZ
1689Put `stty dec' in your .login file and your problems will disappear
1690after a day or two.
a933dad1 1691
9dc15871
EZ
1692The choice of Backspace for erasure was based on confusion, caused by
1693the fact that backspacing causes erasure (later, when you type another
1694character) on most display terminals. But it is a mistake. Deletion
1695of text is not the same thing as backspacing followed by failure to
1696overprint. I do not wish to propagate this confusion by conforming
1697to it.
a933dad1 1698
9dc15871
EZ
1699For this reason, I believe `stty dec' is the right mode to use,
1700and I have designed Emacs to go with that. If there were a thousand
1701other control characters, I would define Control-h to delete as well;
1702but there are not very many other control characters, and I think
1703that providing the most mnemonic possible Help character is more
1704important than adapting to people who don't use `stty dec'.
a933dad1 1705
9dc15871
EZ
1706If you are obstinate about confusing buggy overprinting with deletion,
1707you can redefine Backspace in your .emacs file:
1708 (global-set-key "\b" 'delete-backward-char)
1709You can probably access help-command via f1.
a933dad1 1710
9dc15871 1711** Colors are not available on a tty or in xterm.
a933dad1 1712
9dc15871
EZ
1713Emacs 21 supports colors on character terminals and terminal
1714emulators, but this support relies on the terminfo or termcap database
1715entry to specify that the display supports color. Emacs looks at the
1716"Co" capability for the terminal to find out how many colors are
1717supported; it should be non-zero to activate the color support within
1718Emacs. (Most color terminals support 8 or 16 colors.) If your system
1719uses terminfo, the name of the capability equivalent to "Co" is
1720"colors".
a933dad1 1721
9dc15871
EZ
1722In addition to the "Co" capability, Emacs needs the "op" (for
1723``original pair'') capability, which tells how to switch the terminal
1724back to the default foreground and background colors. Emacs will not
1725use colors if this capability is not defined. If your terminal entry
1726doesn't provide such a capability, try using the ANSI standard escape
1727sequence \E[00m (that is, define a new termcap/terminfo entry and make
1728it use your current terminal's entry plus \E[00m for the "op"
1729capability).
a933dad1 1730
9dc15871
EZ
1731Finally, the "NC" capability (terminfo name: "ncv") tells Emacs which
1732attributes cannot be used with colors. Setting this capability
1733incorrectly might have the effect of disabling colors; try setting
1734this capability to `0' (zero) and see if that helps.
a933dad1 1735
9dc15871
EZ
1736Emacs uses the database entry for the terminal whose name is the value
1737of the environment variable TERM. With `xterm', a common terminal
1738entry that supports color is `xterm-color', so setting TERM's value to
1739`xterm-color' might activate the color support on an xterm-compatible
1740emulator.
a933dad1 1741
bf247b6e 1742Beginning with version 22.1, Emacs supports the --color command-line
9dc15871
EZ
1743option which may be used to force Emacs to use one of a few popular
1744modes for getting colors on a tty. For example, --color=ansi8 sets up
1745for using the ANSI-standard escape sequences that support 8 colors.
a933dad1 1746
9dc15871
EZ
1747Some modes do not use colors unless you turn on the Font-lock mode.
1748Some people have long ago set their `~/.emacs' files to turn on
1749Font-lock on X only, so they won't see colors on a tty. The
1750recommended way of turning on Font-lock is by typing "M-x
1751global-font-lock-mode RET" or by customizing the variable
1752`global-font-lock-mode'.
a933dad1 1753
9dc15871 1754* Runtime problems specific to individual Unix variants
f1e54ce1 1755
9dc15871 1756** GNU/Linux
f1e54ce1 1757
f77e4514
KS
1758*** GNU/Linux: Process output is corrupted.
1759
1760There is a bug in Linux kernel 2.6.10 PTYs that can cause emacs to
1761read corrupted process output.
1762
1763*** GNU/Linux: Remote access to CVS with SSH causes file corruption.
1764
1765If you access a remote CVS repository via SSH, files may be corrupted
1766due to bad interaction between CVS, SSH, and libc.
1767
1768To fix the problem, save the following script into a file, make it
1769executable, and set CVS_RSH environment variable to the file name of
1770the script:
1771
1772#!/bin/bash
1773exec 2> >(exec cat >&2 2>/dev/null)
1774exec ssh "$@"
1775
9dc15871
EZ
1776*** GNU/Linux: On Linux-based GNU systems using libc versions 5.4.19 through
17775.4.22, Emacs crashes at startup with a segmentation fault.
a933dad1 1778
9dc15871
EZ
1779This problem happens if libc defines the symbol __malloc_initialized.
1780One known solution is to upgrade to a newer libc version. 5.4.33 is
1781known to work.
a933dad1 1782
9dc15871
EZ
1783*** GNU/Linux: After upgrading to a newer version of Emacs,
1784the Meta key stops working.
a01325b8 1785
9dc15871
EZ
1786This was reported to happen on a GNU/Linux system distributed by
1787Mandrake. The reason is that the previous version of Emacs was
1788modified by Mandrake to make the Alt key act as the Meta key, on a
1789keyboard where the Windows key is the one which produces the Meta
1790modifier. A user who started using a newer version of Emacs, which
1791was not hacked by Mandrake, expected the Alt key to continue to act as
1792Meta, and was astonished when that didn't happen.
a01325b8 1793
9dc15871
EZ
1794The solution is to find out what key on your keyboard produces the Meta
1795modifier, and use that key instead. Try all of the keys to the left
1796and to the right of the space bar, together with the `x' key, and see
1797which combination produces "M-x" in the echo area. You can also use
1798the `xmodmap' utility to show all the keys which produce a Meta
1799modifier:
a933dad1 1800
9dc15871 1801 xmodmap -pk | egrep -i "meta|alt"
a933dad1 1802
9dc15871
EZ
1803A more convenient way of finding out which keys produce a Meta modifier
1804is to use the `xkbprint' utility, if it's available on your system:
a01325b8 1805
9dc15871 1806 xkbprint 0:0 /tmp/k.ps
a933dad1 1807
9dc15871
EZ
1808This produces a PostScript file `/tmp/k.ps' with a picture of your
1809keyboard; printing that file on a PostScript printer will show what
1810keys can serve as Meta.
a933dad1 1811
9dc15871
EZ
1812The `xkeycaps' also shows a visual representation of the current
1813keyboard settings. It also allows to modify them.
a933dad1 1814
ff3e9dbc 1815*** GNU/Linux: slow startup on Linux-based GNU systems.
a933dad1 1816
9dc15871
EZ
1817People using systems based on the Linux kernel sometimes report that
1818startup takes 10 to 15 seconds longer than `usual'.
a933dad1 1819
9dc15871
EZ
1820This is because Emacs looks up the host name when it starts.
1821Normally, this takes negligible time; the extra delay is due to
1822improper system configuration. This problem can occur for both
1823networked and non-networked machines.
a933dad1 1824
9dc15871 1825Here is how to fix the configuration. It requires being root.
a933dad1 1826
9dc15871 1827**** Networked Case.
a933dad1 1828
9dc15871
EZ
1829First, make sure the files `/etc/hosts' and `/etc/host.conf' both
1830exist. The first line in the `/etc/hosts' file should look like this
1831(replace HOSTNAME with your host name):
a933dad1 1832
9dc15871 1833 127.0.0.1 HOSTNAME
a933dad1 1834
9dc15871
EZ
1835Also make sure that the `/etc/host.conf' files contains the following
1836lines:
1dd8b979 1837
9dc15871
EZ
1838 order hosts, bind
1839 multi on
10a763e5 1840
9dc15871
EZ
1841Any changes, permanent and temporary, to the host name should be
1842indicated in the `/etc/hosts' file, since it acts a limited local
1843database of addresses and names (e.g., some SLIP connections
1844dynamically allocate ip addresses).
1dd8b979 1845
9dc15871 1846**** Non-Networked Case.
a933dad1 1847
9dc15871
EZ
1848The solution described in the networked case applies here as well.
1849However, if you never intend to network your machine, you can use a
1850simpler solution: create an empty `/etc/host.conf' file. The command
1851`touch /etc/host.conf' suffices to create the file. The `/etc/hosts'
1852file is not necessary with this approach.
a933dad1 1853
9dc15871 1854*** GNU/Linux: Emacs on a tty switches the cursor to large blinking block.
a933dad1 1855
9dc15871
EZ
1856This was reported to happen on some GNU/Linux systems which use
1857ncurses version 5.0, but could be relevant for other versions as well.
1858These versions of ncurses come with a `linux' terminfo entry, where
1859the "cvvis" capability (termcap "vs") is defined as "\E[?25h\E[?8c"
1860(show cursor, change size). This escape sequence switches on a
1861blinking hardware text-mode cursor whose size is a full character
1862cell. This blinking cannot be stopped, since a hardware cursor
1863always blinks.
3d00585e 1864
9dc15871
EZ
1865A work-around is to redefine the "cvvis" capability so that it
1866enables a *software* cursor. The software cursor works by inverting
1867the colors of the character at point, so what you see is a block
1868cursor that doesn't blink. For this to work, you need to redefine
1869the "cnorm" capability as well, so that it operates on the software
1870cursor instead of the hardware cursor.
3d00585e 1871
9dc15871
EZ
1872To this end, run "infocmp linux > linux-term", edit the file
1873`linux-term' to make both the "cnorm" and "cvvis" capabilities send
1874the sequence "\E[?25h\E[?17;0;64c", and then run "tic linux-term" to
1875produce a modified terminfo entry.
3d00585e 1876
9dc15871
EZ
1877Alternatively, if you want a blinking underscore as your Emacs cursor,
1878change the "cvvis" capability to send the "\E[?25h\E[?0c" command.
a933dad1 1879
9dc15871 1880*** GNU/Linux: Error messages `internal facep []' happen on GNU/Linux systems.
a933dad1
DL
1881
1882There is a report that replacing libc.so.5.0.9 with libc.so.5.2.16
1883caused this to start happening. People are not sure why, but the
1884problem seems unlikely to be in Emacs itself. Some suspect that it
1885is actually Xlib which won't work with libc.so.5.2.16.
1886
1887Using the old library version is a workaround.
1888
9dc15871 1889** FreeBSD
a933dad1 1890
9dc15871
EZ
1891*** FreeBSD 2.1.5: useless symbolic links remain in /tmp or other
1892directories that have the +t bit.
a933dad1 1893
9dc15871
EZ
1894This is because of a kernel bug in FreeBSD 2.1.5 (fixed in 2.2).
1895Emacs uses symbolic links to implement file locks. In a directory
1896with +t bit, the directory owner becomes the owner of the symbolic
1897link, so that it cannot be removed by anyone else.
a933dad1 1898
9dc15871
EZ
1899If you don't like those useless links, you can let Emacs not to using
1900file lock by adding #undef CLASH_DETECTION to config.h.
a933dad1 1901
9dc15871 1902*** FreeBSD: Getting a Meta key on the console.
a933dad1 1903
9dc15871
EZ
1904By default, neither Alt nor any other key acts as a Meta key on
1905FreeBSD, but this can be changed using kbdcontrol(1). Dump the
1906current keymap to a file with the command
a933dad1 1907
9dc15871 1908 $ kbdcontrol -d >emacs.kbd
a933dad1 1909
9dc15871
EZ
1910Edit emacs.kbd, and give the key you want to be the Meta key the
1911definition `meta'. For instance, if your keyboard has a ``Windows''
1912key with scan code 105, change the line for scan code 105 in emacs.kbd
1913to look like this
a933dad1 1914
9dc15871 1915 105 meta meta meta meta meta meta meta meta O
a933dad1 1916
9dc15871 1917to make the Windows key the Meta key. Load the new keymap with
a933dad1 1918
9dc15871 1919 $ kbdcontrol -l emacs.kbd
a933dad1 1920
9dc15871 1921** HP-UX
a933dad1 1922
9dc15871 1923*** HP/UX : Shell mode gives the message, "`tty`: Ambiguous".
a933dad1 1924
9dc15871 1925christos@theory.tn.cornell.edu says:
a933dad1 1926
9dc15871
EZ
1927The problem is that in your .cshrc you have something that tries to
1928execute `tty`. If you are not running the shell on a real tty then
1929tty will print "not a tty". Csh expects one word in some places,
1930but tty is giving it back 3.
a933dad1 1931
9dc15871
EZ
1932The solution is to add a pair of quotes around `tty` to make it a single
1933word:
a933dad1 1934
9dc15871 1935if (`tty` == "/dev/console")
a933dad1 1936
9dc15871 1937should be changed to:
a933dad1 1938
9dc15871 1939if ("`tty`" == "/dev/console")
a933dad1 1940
9dc15871
EZ
1941Even better, move things that set up terminal sections out of .cshrc
1942and into .login.
a933dad1 1943
9dc15871 1944*** HP/UX: `Pid xxx killed due to text modification or page I/O error'.
a933dad1 1945
9dc15871
EZ
1946On HP/UX, you can get that error when the Emacs executable is on an NFS
1947file system. HP/UX responds this way if it tries to swap in a page and
1948does not get a response from the server within a timeout whose default
1949value is just ten seconds.
a933dad1 1950
9dc15871 1951If this happens to you, extend the timeout period.
a933dad1 1952
9dc15871
EZ
1953*** HP/UX: The right Alt key works wrong on German HP keyboards (and perhaps
1954other non-English HP keyboards too).
a933dad1 1955
9dc15871
EZ
1956This is because HP-UX defines the modifiers wrong in X. Here is a
1957shell script to fix the problem; be sure that it is run after VUE
1958configures the X server.
a933dad1 1959
9dc15871
EZ
1960 xmodmap 2> /dev/null - << EOF
1961 keysym Alt_L = Meta_L
1962 keysym Alt_R = Meta_R
1963 EOF
a933dad1 1964
9dc15871
EZ
1965 xmodmap - << EOF
1966 clear mod1
1967 keysym Mode_switch = NoSymbol
1968 add mod1 = Meta_L
1969 keysym Meta_R = Mode_switch
1970 add mod2 = Mode_switch
1971 EOF
a933dad1 1972
9dc15871
EZ
1973*** HP/UX: "Cannot find callback list" messages from dialog boxes in
1974Emacs built with Motif.
a933dad1 1975
9dc15871
EZ
1976This problem resulted from a bug in GCC 2.4.5. Newer GCC versions
1977such as 2.7.0 fix the problem.
a933dad1 1978
9dc15871 1979*** HP/UX: Emacs does not recognize the AltGr key.
a933dad1 1980
9dc15871
EZ
1981To fix this, set up a file ~/.dt/sessions/sessionetc with executable
1982rights, containing this text:
a933dad1 1983
9dc15871
EZ
1984--------------------------------
1985xmodmap 2> /dev/null - << EOF
1986keysym Alt_L = Meta_L
1987keysym Alt_R = Meta_R
1988EOF
a933dad1 1989
9dc15871
EZ
1990xmodmap - << EOF
1991clear mod1
1992keysym Mode_switch = NoSymbol
1993add mod1 = Meta_L
1994keysym Meta_R = Mode_switch
1995add mod2 = Mode_switch
1996EOF
1997--------------------------------
a933dad1 1998
9dc15871 1999*** HP/UX 11.0: Emacs makes HP/UX 11.0 crash.
a933dad1 2000
9dc15871 2001This is a bug in HPUX; HPUX patch PHKL_16260 is said to fix it.
a933dad1 2002
9dc15871 2003** AIX
a933dad1 2004
9dc15871 2005*** AIX: Trouble using ptys.
a933dad1 2006
9dc15871
EZ
2007People often install the pty devices on AIX incorrectly.
2008Use `smit pty' to reinstall them properly.
a933dad1 2009
9dc15871 2010*** AIXterm: Your Delete key sends a Backspace to the terminal.
a933dad1 2011
9dc15871 2012The solution is to include in your .Xdefaults the lines:
a933dad1 2013
9dc15871
EZ
2014 *aixterm.Translations: #override <Key>BackSpace: string(0x7f)
2015 aixterm*ttyModes: erase ^?
a933dad1 2016
9dc15871 2017This makes your Backspace key send DEL (ASCII 127).
a933dad1 2018
9dc15871
EZ
2019*** AIX: If linking fails because libXbsd isn't found, check if you
2020are compiling with the system's `cc' and CFLAGS containing `-O5'. If
2021so, you have hit a compiler bug. Please make sure to re-configure
2022Emacs so that it isn't compiled with `-O5'.
a933dad1 2023
9dc15871 2024*** AIX 4.3.x or 4.4: Compiling fails.
a933dad1 2025
9dc15871
EZ
2026This could happen if you use /bin/c89 as your compiler, instead of
2027the default `cc'. /bin/c89 treats certain warnings, such as benign
2028redefinitions of macros, as errors, and fails the build. A solution
2029is to use the default compiler `cc'.
a933dad1 2030
9dc15871
EZ
2031*** AIX 4: Some programs fail when run in a Shell buffer
2032with an error message like No terminfo entry for "unknown".
a933dad1 2033
9dc15871
EZ
2034On AIX, many terminal type definitions are not installed by default.
2035`unknown' is one of them. Install the "Special Generic Terminal
2036Definitions" to make them defined.
a933dad1 2037
9dc15871 2038** Solaris
a933dad1 2039
9dc15871
EZ
2040We list bugs in current versions here. Solaris 2.x and 4.x are covered in the
2041section on legacy systems.
a933dad1 2042
9dc15871 2043*** On Solaris, C-x doesn't get through to Emacs when you use the console.
a933dad1 2044
9dc15871
EZ
2045This is a Solaris feature (at least on Intel x86 cpus). Type C-r
2046C-r C-t, to toggle whether C-x gets through to Emacs.
a933dad1 2047
9dc15871 2048*** Problem with remote X server on Suns.
a933dad1 2049
9dc15871
EZ
2050On a Sun, running Emacs on one machine with the X server on another
2051may not work if you have used the unshared system libraries. This
2052is because the unshared libraries fail to use YP for host name lookup.
2053As a result, the host name you specify may not be recognized.
a933dad1 2054
0a4dd4e4 2055*** Solaris 2,6: Emacs crashes with SIGBUS or SIGSEGV on Solaris after you delete a frame.
a933dad1 2056
0a4dd4e4
EZ
2057We suspect that this is a bug in the X libraries provided by
2058Sun. There is a report that one of these patches fixes the bug and
2059makes the problem stop:
a933dad1 2060
0a4dd4e4
EZ
2061105216-01 105393-01 105518-01 105621-01 105665-01 105615-02 105216-02
2062105667-01 105401-08 105615-03 105621-02 105686-02 105736-01 105755-03
2063106033-01 105379-01 105786-01 105181-04 105379-03 105786-04 105845-01
2064105284-05 105669-02 105837-01 105837-02 105558-01 106125-02 105407-01
a933dad1 2065
0a4dd4e4
EZ
2066Another person using a newer system (kernel patch level Generic_105181-06)
2067suspects that the bug was fixed by one of these more recent patches:
a933dad1 2068
0a4dd4e4
EZ
2069106040-07 SunOS 5.6: X Input & Output Method patch
2070106222-01 OpenWindows 3.6: filemgr (ff.core) fixes
2071105284-12 Motif 1.2.7: sparc Runtime library patch
a933dad1 2072
0a4dd4e4 2073*** Solaris 7 or 8: Emacs reports a BadAtom error (from X)
a933dad1 2074
0a4dd4e4
EZ
2075This happens when Emacs was built on some other version of Solaris.
2076Rebuild it on Solaris 8.
a933dad1 2077
9dc15871
EZ
2078*** When using M-x dbx with the SparcWorks debugger, the `up' and `down'
2079commands do not move the arrow in Emacs.
a933dad1 2080
9dc15871 2081You can fix this by adding the following line to `~/.dbxinit':
a933dad1 2082
9dc15871 2083 dbxenv output_short_file_name off
a933dad1 2084
0a4dd4e4
EZ
2085*** On Solaris, CTRL-t is ignored by Emacs when you use
2086the fr.ISO-8859-15 locale (and maybe other related locales).
a933dad1 2087
0a4dd4e4 2088You can fix this by editing the file:
a933dad1 2089
3256a475 2090 /usr/openwin/lib/locale/iso8859-15/Compose
a933dad1 2091
0a4dd4e4 2092Near the bottom there is a line that reads:
a933dad1 2093
3256a475 2094 Ctrl<t> <quotedbl> <Y> : "\276" threequarters
a933dad1 2095
0a4dd4e4 2096that should read:
a933dad1 2097
3256a475 2098 Ctrl<T> <quotedbl> <Y> : "\276" threequarters
a933dad1 2099
0a4dd4e4 2100Note the lower case <t>. Changing this line should make C-t work.
a933dad1 2101
1b6406b3
CY
2102*** On Solaris, Emacs fails to set menu-bar-update-hook on startup, with error
2103"Error in menu-bar-update-hook: (error Point before start of properties)".
2104This seems to be a GCC optimization bug that occurs for GCC 4.1.2 (-g
2105and -g -O2) and GCC 4.2.3 (-g -O and -g -O2). You can fix this by
2106compiling with GCC 4.2.3 or CC 5.7, with no optimizations.
2107
0a4dd4e4 2108** Irix
a933dad1 2109
9dc15871 2110*** Irix 6.5: Emacs crashes on the SGI R10K, when compiled with GCC.
a933dad1 2111
9dc15871 2112This seems to be fixed in GCC 2.95.
a933dad1 2113
0a4dd4e4 2114*** Irix: Trouble using ptys, or running out of ptys.
a933dad1 2115
9dc15871
EZ
2116The program mkpts (which may be in `/usr/adm' or `/usr/sbin') needs to
2117be set-UID to root, or non-root programs like Emacs will not be able
2118to allocate ptys reliably.
a933dad1 2119
9dc15871 2120* Runtime problems specific to MS-Windows
a933dad1 2121
ade79051
KS
2122** Windows 95 and networking.
2123
2124To support server sockets, Emacs 22.1 loads ws2_32.dll. If this file
2125is missing, all Emacs networking features are disabled.
2126
2127Old versions of Windows 95 may not have the required DLL. To use
2128Emacs' networking features on Windows 95, you must install the
2129"Windows Socket 2" update available from MicroSoft's support Web.
2130
9dc15871 2131** Emacs exits with "X protocol error" when run with an X server for MS-Windows.
a933dad1 2132
9dc15871
EZ
2133A certain X server for Windows had a bug which caused this.
2134Supposedly the newer 32-bit version of this server doesn't have the
2135problem.
177c0ea7 2136
b6ec0fa0
JR
2137** Emacs crashes when opening a file with a UNC path and rails-mode is loaded.
2138
2139Loading rails-mode seems to interfere with UNC path handling. This has been
2140reported as a bug against both Emacs and rails-mode, so look for an updated
2141rails-mode that avoids this crash, or avoid using UNC paths if using
2142rails-mode.
2143
2144** Known problems with the MS-Windows port of Emacs 22.3
de66e883 2145
117402b8
JR
2146M-x term does not work on MS-Windows. TTY emulation on Windows is
2147undocumented, and programs such as stty which are used on posix platforms
2148to control tty emulation do not exist for native windows terminals.
2149
de66e883
JR
2150Using create-fontset-from-ascii-font or the --font startup parameter
2151with a Chinese, Japanese or Korean font leads to display problems.
2152Use a Latin-only font as your default font. If you want control over
2153which font is used to display Chinese, Japanese or Korean character,
2154use create-fontset-from-fontset-spec to define a fontset.
a933dad1 2155
9dc15871
EZ
2156Frames are not refreshed while the File or Font dialog or a pop-up menu
2157is displayed. This also means help text for pop-up menus is not
2158displayed at all. This is because message handling under Windows is
2159synchronous, so we cannot handle repaint (or any other) messages while
2160waiting for a system function to return the result of the dialog or
2161pop-up menu interaction.
a933dad1 2162
9dc15871
EZ
2163Windows 95 and Windows NT up to version 4.0 do not support help text
2164for menus. Help text is only available in later versions of Windows.
a933dad1 2165
af71f3ce
EZ
2166When "ClearType" method is selected as the "method to smooth edges of
2167screen fonts" (in Display Properties, Appearance tab, under
2168"Effects"), there are various problems related to display of
f396bf16
JR
2169characters: Bold fonts can be hard to read, small portions of some
2170characters could appear chopped, etc. This happens because, under
2171ClearType, characters are drawn outside their advertised bounding box.
2172Emacs 21 disabled the use of ClearType, whereas Emacs 22 allows it and
2173has some code to enlarge the width of the bounding box. Apparently,
2174this display feature needs more changes to get it 100% right. A
2175workaround is to disable ClearType.
af71f3ce 2176
9dc15871
EZ
2177There are problems with display if mouse-tracking is enabled and the
2178mouse is moved off a frame, over another frame then back over the first
2179frame. A workaround is to click the left mouse button inside the frame
2180after moving back into it.
a933dad1 2181
9dc15871
EZ
2182Some minor flickering still persists during mouse-tracking, although
2183not as severely as in 21.1.
a933dad1 2184
9dc15871
EZ
2185An inactive cursor remains in an active window after the Windows
2186Manager driven switch of the focus, until a key is pressed.
a933dad1 2187
4ed1bce5 2188Windows input methods are not recognized by Emacs. However, some
9dc15871
EZ
2189of these input methods cause the keyboard to send characters encoded
2190in the appropriate coding system (e.g., ISO 8859-1 for Latin-1
4ed1bce5
EZ
2191characters, ISO 8859-8 for Hebrew characters, etc.). To make these
2192input methods work with Emacs, set the keyboard coding system to the
2193appropriate value after you activate the Windows input method. For
2194example, if you activate the Hebrew input method, type this:
2195
2196 C-x RET k hebrew-iso-8bit RET
2197
2198(Emacs ought to recognize the Windows language-change event and set up
2199the appropriate keyboard encoding automatically, but it doesn't do
2200that yet.) In addition, to use these Windows input methods, you
2201should set your "Language for non-Unicode programs" (on Windows XP,
2202this is on the Advanced tab of Regional Settings) to the language of
2203the input method.
a933dad1 2204
868c31fe
EZ
2205To bind keys that produce non-ASCII characters with modifiers, you
2206must specify raw byte codes. For instance, if you want to bind
2207META-a-grave to a command, you need to specify this in your `~/.emacs':
2208
2209 (global-set-key [?\M-\340] ...)
2210
2211The above example is for the Latin-1 environment where the byte code
2212of the encoded a-grave is 340 octal. For other environments, use the
2213encoding appropriate to that environment.
a933dad1 2214
9dc15871
EZ
2215The %b specifier for format-time-string does not produce abbreviated
2216month names with consistent widths for some locales on some versions
2217of Windows. This is caused by a deficiency in the underlying system
2218library function.
a933dad1 2219
cdc9f5c2
JR
2220The function set-time-zone-rule gives incorrect results for many
2221non-US timezones. This is due to over-simplistic handling of
2222daylight savings switchovers by the Windows libraries.
9f1bc31f 2223
365b9257
EZ
2224Files larger than 4GB cause overflow in the size (represented as a
222532-bit integer) reported by `file-attributes'. This affects Dired as
2226well, since the Windows port uses a Lisp emulation of `ls' that relies
2227on `file-attributes'.
2228
56dc0646
EZ
2229Sound playing is not supported with the `:data DATA' key-value pair.
2230You _must_ use the `:file FILE' method.
2231
0a4dd4e4 2232** Typing Alt-Shift has strange effects on MS-Windows.
a933dad1 2233
0a4dd4e4
EZ
2234This combination of keys is a command to change keyboard layout. If
2235you proceed to type another non-modifier key before you let go of Alt
2236and Shift, the Alt and Shift act as modifiers in the usual way. A
2237more permanent work around is to change it to another key combination,
d169ccbd
EZ
2238or disable it in the "Regional and Language Options" applet of the
2239Control Panel. (The exact sequence of mouse clicks in the "Regional
2240and Language Options" applet needed to find the key combination that
2241changes the keyboard layout depends on your Windows version; for XP,
2242in the Languages tab, click "Details" and then "Key Settings".)
a933dad1 2243
651f8172
EZ
2244** Cygwin build of Emacs hangs after rebasing Cygwin DLLs
2245
2246Usually, on Cygwin, one needs to rebase the DLLs if an application
2247aborts with a message like this:
2248
2249 C:\cygwin\bin\python.exe: *** unable to remap C:\cygwin\bin\cygssl.dll to
2250 same address as parent(0xDF0000) != 0xE00000
2251
2252However, since Cygwin DLL 1.5.17 was released, after such rebasing,
2253Emacs hangs.
2254
2255This was reported to happen for Emacs 21.2 and also for the pretest of
2256Emacs 22.1 on Cygwin.
2257
2258To work around this, build Emacs like this:
2259
2260 LDFLAGS='-Wl,--enable-auto-import -Wl,--enable-auto-image-base' ./configure
2261 make LD='$(CC)'
2262 make LD='$(CC)' install
2263
2264This produces an Emacs binary that is independent of rebasing.
2265
2266Note that you _must_ use LD='$(CC)' in the last two commands above, to
2267prevent GCC from passing the "--image-base 0x20000000" option to the
2268linker, which is what it does by default. That option produces an
2269Emacs binary with the base address 0x20000000, which will cause Emacs
2270to hang after Cygwin DLLs are rebased.
2271
0a4dd4e4 2272** Interrupting Cygwin port of Bash from Emacs doesn't work.
a933dad1 2273
0a4dd4e4
EZ
2274Cygwin 1.x builds of the ported Bash cannot be interrupted from the
2275MS-Windows version of Emacs. This is due to some change in the Bash
2276port or in the Cygwin library which apparently make Bash ignore the
2277keyboard interrupt event sent by Emacs to Bash. (Older Cygwin ports
2278of Bash, up to b20.1, did receive SIGINT from Emacs.)
a933dad1 2279
0a4dd4e4 2280** Accessing remote files with ange-ftp hangs the MS-Windows version of Emacs.
a933dad1 2281
0a4dd4e4
EZ
2282If the FTP client is the Cygwin port of GNU `ftp', this appears to be
2283due to some bug in the Cygwin DLL or some incompatibility between it
2284and the implementation of asynchronous subprocesses in the Windows
2285port of Emacs. Specifically, some parts of the FTP server responses
2286are not flushed out, apparently due to buffering issues, which
2287confuses ange-ftp.
a933dad1 2288
0a4dd4e4
EZ
2289The solution is to downgrade to an older version of the Cygwin DLL
2290(version 1.3.2 was reported to solve the problem), or use the stock
2291Windows FTP client, usually found in the `C:\WINDOWS' or 'C:\WINNT'
2292directory. To force ange-ftp use the stock Windows client, set the
2293variable `ange-ftp-ftp-program-name' to the absolute file name of the
2294client's executable. For example:
a933dad1 2295
9dc15871 2296 (setq ange-ftp-ftp-program-name "c:/windows/ftp.exe")
a933dad1 2297
9dc15871
EZ
2298If you want to stick with the Cygwin FTP client, you can work around
2299this problem by putting this in your `.emacs' file:
a933dad1 2300
9dc15871 2301 (setq ange-ftp-ftp-program-args '("-i" "-n" "-g" "-v" "--prompt" "")
a933dad1 2302
9dc15871 2303** lpr commands don't work on MS-Windows with some cheap printers.
a933dad1 2304
9dc15871
EZ
2305This problem may also strike other platforms, but the solution is
2306likely to be a global one, and not Emacs specific.
a933dad1 2307
9dc15871
EZ
2308Many cheap inkjet, and even some cheap laser printers, do not
2309print plain text anymore, they will only print through graphical
2310printer drivers. A workaround on MS-Windows is to use Windows' basic
2311built in editor to print (this is possibly the only useful purpose it
2312has):
a933dad1 2313
9dc15871
EZ
2314(setq printer-name "") ;; notepad takes the default
2315(setq lpr-command "notepad") ;; notepad
2316(setq lpr-switches nil) ;; not needed
2317(setq lpr-printer-switch "/P") ;; run notepad as batch printer
a933dad1 2318
9dc15871 2319** Antivirus software interacts badly with the MS-Windows version of Emacs.
a933dad1 2320
9dc15871
EZ
2321The usual manifestation of these problems is that subprocesses don't
2322work or even wedge the entire system. In particular, "M-x shell RET"
2323was reported to fail to work. But other commands also sometimes don't
2324work when an antivirus package is installed.
a933dad1 2325
9dc15871
EZ
2326The solution is to switch the antivirus software to a less aggressive
2327mode (e.g., disable the ``auto-protect'' feature), or even uninstall
2328or disable it entirely.
a933dad1 2329
9dc15871 2330** Pressing the mouse button on MS-Windows does not give a mouse-2 event.
a933dad1 2331
9dc15871
EZ
2332This is usually a problem with the mouse driver. Because most Windows
2333programs do not do anything useful with the middle mouse button, many
2334mouse drivers allow you to define the wheel press to do something
2335different. Some drivers do not even have the option to generate a
2336middle button press. In such cases, setting the wheel press to
2337"scroll" sometimes works if you press the button twice. Trying a
2338generic mouse driver might help.
a933dad1 2339
9dc15871 2340** Scrolling the mouse wheel on MS-Windows always scrolls the top window.
a933dad1 2341
9dc15871
EZ
2342This is another common problem with mouse drivers. Instead of
2343generating scroll events, some mouse drivers try to fake scroll bar
2344movement. But they are not intelligent enough to handle multiple
2345scroll bars within a frame. Trying a generic mouse driver might help.
a933dad1 2346
9dc15871
EZ
2347** Mail sent through Microsoft Exchange in some encodings appears to be
2348mangled and is not seen correctly in Rmail or Gnus. We don't know
2349exactly what happens, but it isn't an Emacs problem in cases we've
2350seen.
a933dad1 2351
9dc15871
EZ
2352** On MS-Windows, you cannot use the right-hand ALT key and the left-hand
2353CTRL key together to type a Control-Meta character.
a933dad1 2354
9dc15871 2355This is a consequence of a misfeature beyond Emacs's control.
a933dad1 2356
9dc15871
EZ
2357Under Windows, the AltGr key on international keyboards generates key
2358events with the modifiers Right-Alt and Left-Ctrl. Since Emacs cannot
2359distinguish AltGr from an explicit Right-Alt and Left-Ctrl
2360combination, whenever it sees Right-Alt and Left-Ctrl it assumes that
2361AltGr has been pressed. The variable `w32-recognize-altgr' can be set
2362to nil to tell Emacs that AltGr is really Ctrl and Alt.
a933dad1 2363
9dc15871 2364** Under some X-servers running on MS-Windows, Emacs' display is incorrect.
a933dad1 2365
9dc15871
EZ
2366The symptoms are that Emacs does not completely erase blank areas of the
2367screen during scrolling or some other screen operations (e.g., selective
2368display or when killing a region). M-x recenter will cause the screen
2369to be completely redisplayed and the "extra" characters will disappear.
a933dad1 2370
9dc15871
EZ
2371This is known to occur under Exceed 6, and possibly earlier versions
2372as well; it is reportedly solved in version 6.2.0.16 and later. The
2373problem lies in the X-server settings.
a933dad1 2374
9dc15871
EZ
2375There are reports that you can solve the problem with Exceed by
2376running `Xconfig' from within NT, choosing "X selection", then
2377un-checking the boxes "auto-copy X selection" and "auto-paste to X
2378selection".
a933dad1 2379
9dc15871
EZ
2380Of this does not work, please inform bug-gnu-emacs@gnu.org. Then
2381please call support for your X-server and see if you can get a fix.
2382If you do, please send it to bug-gnu-emacs@gnu.org so we can list it
2383here.
a933dad1 2384
9dc15871 2385* Build-time problems
a933dad1 2386
9dc15871 2387** Configuration
a933dad1 2388
9dc15871 2389*** The `configure' script doesn't find the jpeg library.
a933dad1 2390
9dc15871
EZ
2391There are reports that this happens on some systems because the linker
2392by default only looks for shared libraries, but jpeg distribution by
2393default only installs a nonshared version of the library, `libjpeg.a'.
a933dad1 2394
9dc15871
EZ
2395If this is the problem, you can configure the jpeg library with the
2396`--enable-shared' option and then rebuild libjpeg. This produces a
2397shared version of libjpeg, which you need to install. Finally, rerun
2398the Emacs configure script, which should now find the jpeg library.
2399Alternatively, modify the generated src/Makefile to link the .a file
2400explicitly, and edit src/config.h to define HAVE_JPEG.
a933dad1 2401
f4b84ef4
GM
2402*** `configure' warns ``accepted by the compiler, rejected by the preprocessor''.
2403
2404This indicates a mismatch between the C compiler and preprocessor that
2405configure is using. For example, on Solaris 10 trying to use
2406CC=/opt/SUNWspro/bin/cc (the Sun Studio compiler) together with
2407CPP=/usr/ccs/lib/cpp can result in errors of this form (you may also
2408see the error ``"/usr/include/sys/isa_defs.h", line 500: undefined control'').
2409
2410The solution is to tell configure to use the correct C preprocessor
2411for your C compiler (CPP="/opt/SUNWspro/bin/cc -E" in the above
2412example).
2413
2414*** `configure' fails with ``"junk.c", line 660: invalid input token: 8.elc''
2415
2416The final stage of the Emacs configure process uses the C preprocessor
2417to generate the Makefiles. Errors of this form can occur if the C
2418preprocessor inserts extra whitespace into its output. The solution
2419is to find the switches that stop your preprocessor from inserting extra
2420whitespace, add them to CPPFLAGS, and re-run configure. For example,
2421this error can occur on Solaris 10 when using the Sun Studio compiler
2422``Sun C 5.8'' with its preprocessor CPP="/opt/SUNWspro/bin/cc -E".
2423The relevant switch in this case is "-Xs" (``compile assuming
2424(pre-ANSI) K & R C style code'').
2425
9dc15871 2426** Compilation
a933dad1 2427
9dc15871 2428*** Building Emacs over NFS fails with ``Text file busy''.
a933dad1 2429
9dc15871 2430This was reported to happen when building Emacs on a GNU/Linux system
880ea925 2431(Red Hat Linux 6.2) using a build directory automounted from Solaris
9dc15871
EZ
2432(SunOS 5.6) file server, but it might not be limited to that
2433configuration alone. Presumably, the NFS server doesn't commit the
2434files' data to disk quickly enough, and the Emacs executable file is
2435left ``busy'' for several seconds after Emacs has finished dumping
2436itself. This causes the subsequent commands which invoke the dumped
2437Emacs executable to fail with the above message.
a933dad1 2438
9dc15871
EZ
2439In some of these cases, a time skew between the NFS server and the
2440machine where Emacs is built is detected and reported by GNU Make
2441(it says that some of the files have modification time in the future).
2442This might be a symptom of NFS-related problems.
a933dad1 2443
9dc15871
EZ
2444If the NFS server runs on Solaris, apply the Solaris patch 105379-05
2445(Sunos 5.6: /kernel/misc/nfssrv patch). If that doesn't work, or if
2446you have a different version of the OS or the NFS server, you can
2447force the NFS server to use 1KB blocks, which was reported to fix the
2448problem albeit at a price of slowing down file I/O. You can force 1KB
2449blocks by specifying the "-o rsize=1024,wsize=1024" options to the
2450`mount' command, or by adding ",rsize=1024,wsize=1024" to the mount
2451options in the appropriate system configuration file, such as
2452`/etc/auto.home'.
a933dad1 2453
9dc15871
EZ
2454Alternatively, when Make fails due to this problem, you could wait for
2455a few seconds and then invoke Make again. In one particular case,
2456waiting for 10 or more seconds between the two Make invocations seemed
2457to work around the problem.
a933dad1 2458
9dc15871
EZ
2459Similar problems can happen if your machine NFS-mounts a directory
2460onto itself. Suppose the Emacs sources live in `/usr/local/src' and
2461you are working on the host called `marvin'. Then an entry in the
2462`/etc/fstab' file like the following is asking for trouble:
a933dad1 2463
9dc15871 2464 marvin:/usr/local/src /usr/local/src ...options.omitted...
a933dad1 2465
9dc15871 2466The solution is to remove this line from `etc/fstab'.
a933dad1 2467
9dc15871 2468*** Building Emacs with GCC 2.9x fails in the `src' directory.
a933dad1 2469
9dc15871
EZ
2470This may happen if you use a development version of GNU `cpp' from one
2471of the GCC snapshots between Oct 2000 and Feb 2001, or from a released
2472version of GCC newer than 2.95.2 which was prepared around those
2473dates; similar problems were reported with some snapshots of GCC 3.1
2474around Sep 30 2001. The preprocessor in those versions is
2475incompatible with a traditional Unix cpp (e.g., it expands ".." into
2476". .", which breaks relative file names that reference the parent
2477directory; or inserts TAB characters before lines that set Make
2478variables).
a933dad1 2479
9dc15871
EZ
2480The solution is to make sure the preprocessor is run with the
2481`-traditional' option. The `configure' script does that automatically
2482when it detects the known problems in your cpp, but you might hit some
2483unknown ones. To force the `configure' script to use `-traditional',
2484run the script like this:
a933dad1 2485
9dc15871 2486 CPP='gcc -E -traditional' ./configure ...
a933dad1 2487
9dc15871
EZ
2488(replace the ellipsis "..." with any additional arguments you pass to
2489the script).
a933dad1 2490
9dc15871
EZ
2491Note that this problem does not pertain to the MS-Windows port of
2492Emacs, since it doesn't use the preprocessor to generate Makefiles.
a933dad1 2493
9dc15871
EZ
2494*** src/Makefile and lib-src/Makefile are truncated--most of the file missing.
2495*** Compiling wakeup, in lib-src, says it can't make wakeup.c.
a933dad1 2496
9dc15871
EZ
2497This can happen if configure uses GNU sed version 2.03. That version
2498had a bug. GNU sed version 2.05 works properly.To solve the
2499problem, install the current version of GNU Sed, then rerun Emacs's
2500configure script.
a933dad1 2501
9dc15871 2502*** Compiling lib-src says there is no rule to make test-distrib.c.
a933dad1 2503
9dc15871
EZ
2504This results from a bug in a VERY old version of GNU Sed. To solve
2505the problem, install the current version of GNU Sed, then rerun
2506Emacs's configure script.
a933dad1 2507
745377e8
GM
2508*** Building a 32-bit executable on a 64-bit GNU/Linux architecture.
2509
2510First ensure that the necessary 32-bit system libraries and include
2511files are installed. Then use:
2512
2513 env CC="gcc -m32" ./configure --build=i386-linux-gnu \
2514 --x-libraries=/usr/X11R6/lib
2515
2516(using the location of the 32-bit X libraries on your system).
2517
1020d879 2518*** Building the Cygwin port for MS-Windows can fail with some GCC versions
9c9f0081
EZ
2519
2520Building Emacs 22 with Cygwin builds of GCC 3.4.4-1 and 3.4.4-2 is
2521reported to either fail or cause Emacs to segfault at run time. In
2522addition, the Cygwin GCC 3.4.4-2 has problems with generating debug
2523info. Cygwin users are advised not to use these versions of GCC for
485a6c18
EZ
2524compiling Emacs. GCC versions 4.0.3, 4.0.4, 4.1.1, and 4.1.2
2525reportedly build a working Cygwin binary of Emacs, so we recommend
2526these GCC versions. Note that these versions of GCC, 4.0.3, 4.0.4,
25274.1.1, and 4.1.2, are currently the _only_ versions known to succeed
2528in building Emacs (as of v22.1).
9c9f0081
EZ
2529
2530*** Building the native MS-Windows port with Cygwin GCC can fail.
a933dad1 2531
ad05a5de 2532Emacs may not build using some Cygwin builds of GCC, such as Cygwin
9dc15871
EZ
2533version 1.1.8, using the default configure settings. It appears to be
2534necessary to specify the -mwin32 flag when compiling, and define
2535__MSVCRT__, like so:
a933dad1 2536
9dc15871 2537 configure --with-gcc --cflags -mwin32 --cflags -D__MSVCRT__
a933dad1 2538
9dc15871 2539*** Building the MS-Windows port fails with a CreateProcess failure.
a933dad1 2540
9dc15871
EZ
2541Some versions of mingw32 make on some versions of Windows do not seem
2542to detect the shell correctly. Try "make SHELL=cmd.exe", or if that
2543fails, try running make from Cygwin bash instead.
a933dad1 2544
9dc15871 2545*** Building `ctags' for MS-Windows with the MinGW port of GCC fails.
a933dad1 2546
9dc15871
EZ
2547This might happen due to a bug in the MinGW header assert.h, which
2548defines the `assert' macro with a trailing semi-colon. The following
2549patch to assert.h should solve this:
a933dad1 2550
0cc69e7d
EZ
2551 *** include/assert.h.orig Sun Nov 7 02:41:36 1999
2552 --- include/assert.h Mon Jan 29 11:49:10 2001
2553 ***************
2554 *** 41,47 ****
2555 /*
2556 * If not debugging, assert does nothing.
2557 */
2558 ! #define assert(x) ((void)0);
2559
2560 #else /* debugging enabled */
2561
2562 --- 41,47 ----
2563 /*
2564 * If not debugging, assert does nothing.
2565 */
2566 ! #define assert(x) ((void)0)
2567
2568 #else /* debugging enabled */
a933dad1 2569
a933dad1 2570
3e7c244e 2571*** Building the MS-Windows port with Visual Studio 2005 fails.
a933dad1 2572
3e7c244e
JR
2573Microsoft no longer ships the single threaded version of the C library
2574with their compiler, and the multithreaded static library is missing
0cc69e7d 2575some functions that Microsoft have deemed non-threadsafe. The
3e7c244e
JR
2576dynamically linked C library has all the functions, but there is a
2577conflict between the versions of malloc in the DLL and in Emacs, which
2578is not resolvable due to the way Windows does dynamic linking.
a933dad1 2579
3e7c244e
JR
2580We recommend the use of the MingW port of GCC for compiling Emacs, as
2581not only does it not suffer these problems, but it is also Free
2582software like Emacs.
a933dad1 2583
9dc15871 2584** Linking
a933dad1 2585
9dc15871
EZ
2586*** Building Emacs with a system compiler fails to link because of an
2587undefined symbol such as __eprintf which does not appear in Emacs.
a933dad1 2588
9dc15871
EZ
2589This can happen if some of the libraries linked into Emacs were built
2590with GCC, but Emacs itself is being linked with a compiler other than
2591GCC. Object files compiled with GCC might need some helper functions
2592from libgcc.a, the library which comes with GCC, but the system
2593compiler does not instruct the linker to search libgcc.a during the
2594link stage.
a933dad1 2595
9dc15871 2596A solution is to link with GCC, like this:
a933dad1 2597
3256a475 2598 make CC=gcc
a933dad1 2599
9dc15871
EZ
2600Since the .o object files already exist, this will not recompile Emacs
2601with GCC, but just restart by trying again to link temacs.
a933dad1 2602
9dc15871 2603*** AIX 1.3 ptf 0013: Link failure.
a933dad1 2604
9dc15871
EZ
2605There is a real duplicate definition of the function `_slibc_free' in
2606the library /lib/libc_s.a (just do nm on it to verify). The
2607workaround/fix is:
a933dad1 2608
9dc15871
EZ
2609 cd /lib
2610 ar xv libc_s.a NLtmtime.o
2611 ar dv libc_s.a NLtmtime.o
a933dad1 2612
9dc15871
EZ
2613*** AIX 4.1.2: Linker error messages such as
2614 ld: 0711-212 SEVERE ERROR: Symbol .__quous, found in the global symbol table
3256a475 2615 of archive /usr/lib/libIM.a, was not defined in archive member shr.o.
a933dad1 2616
9dc15871
EZ
2617This is a problem in libIM.a. You can work around it by executing
2618these shell commands in the src subdirectory of the directory where
2619you build Emacs:
a933dad1 2620
9dc15871
EZ
2621 cp /usr/lib/libIM.a .
2622 chmod 664 libIM.a
2623 ranlib libIM.a
a933dad1 2624
9dc15871
EZ
2625Then change -lIM to ./libIM.a in the command to link temacs (in
2626Makefile).
a933dad1 2627
9dc15871 2628*** Sun with acc: Link failure when using acc on a Sun.
a933dad1 2629
9dc15871 2630To use acc, you need additional options just before the libraries, such as
a933dad1 2631
9dc15871 2632 /usr/lang/SC2.0.1/values-Xt.o -L/usr/lang/SC2.0.1/cg87 -L/usr/lang/SC2.0.1
a933dad1 2633
9dc15871 2634and you need to add -lansi just before -lc.
a933dad1 2635
9dc15871
EZ
2636The precise file names depend on the compiler version, so we
2637cannot easily arrange to supply them.
a933dad1 2638
9dc15871 2639*** Linking says that the functions insque and remque are undefined.
a933dad1 2640
9dc15871 2641Change oldXMenu/Makefile by adding insque.o to the variable OBJS.
a933dad1 2642
9dc15871 2643*** `tparam' reported as a multiply-defined symbol when linking with ncurses.
a933dad1 2644
9dc15871
EZ
2645This problem results from an incompatible change in ncurses, in
2646version 1.9.9e approximately. This version is unable to provide a
2647definition of tparm without also defining tparam. This is also
2648incompatible with Terminfo; as a result, the Emacs Terminfo support
2649does not work with this version of ncurses.
a933dad1 2650
9dc15871 2651The fix is to install a newer version of ncurses, such as version 4.2.
a933dad1 2652
d7ef7cd4
GM
2653** Bootstrapping
2654
2655Bootstrapping (compiling the .el files) is normally only necessary
2656with CVS builds, since they the .elc files are pre-compiled in releases.
2657
2658*** "No rule to make target" with Ubuntu 8.04 make 3.81-3build1
2659
2660Compiling the lisp files fails at random places, complaining:
2661"No rule to make target `/path/to/some/lisp.elc'".
2662The causes of this problem are not understood. Using GNU make 3.81 compiled
2663from source, rather than the Ubuntu version, worked. See Bug#327,821.
2664
9dc15871 2665** Dumping
a933dad1 2666
9dc15871 2667*** Linux: Segfault during `make bootstrap' under certain recent versions of the Linux kernel.
a933dad1 2668
880ea925 2669With certain recent Linux kernels (like the one of Red Hat Fedora Core
cf14a51c 26701 and newer), the new "Exec-shield" functionality is enabled by default, which
ed214edf
JD
2671creates a different memory layout that breaks the emacs dumper. Emacs tries
2672to handle this at build time, but if the workaround used fails, these
2673instructions can be useful.
cf14a51c
JD
2674The work-around explained here is not enough on Fedora Core 4 (and possible
2675newer). Read the next item.
a933dad1 2676
1f02a4ba
JD
2677Configure can overcome the problem of exec-shield if the architecture is
2678x86 and the program setarch is present. On other architectures no
2679workaround is known.
2680
9dc15871 2681You can check the Exec-shield state like this:
a933dad1 2682
9dc15871 2683 cat /proc/sys/kernel/exec-shield
a933dad1 2684
1f02a4ba 2685It returns non-zero when Exec-shield is enabled, 0 otherwise. Please
9dc15871 2686read your system documentation for more details on Exec-shield and
1f02a4ba
JD
2687associated commands. Exec-shield can be turned off with this command:
2688
2689 echo "0" > /proc/sys/kernel/exec-shield
a933dad1 2690
9dc15871
EZ
2691When Exec-shield is enabled, building Emacs will segfault during the
2692execution of this command:
a933dad1 2693
1f02a4ba 2694 ./temacs --batch --load loadup [dump|bootstrap]
a933dad1 2695
9dc15871 2696To work around this problem, it is necessary to temporarily disable
1f02a4ba
JD
2697Exec-shield while building Emacs, or, on x86, by using the `setarch'
2698command when running temacs like this:
2699
2700 setarch i386 ./temacs --batch --load loadup [dump|bootstrap]
a933dad1 2701
ade79051 2702
ff0ab406
RS
2703*** Fedora Core 4 GNU/Linux: Segfault during dumping.
2704
2705In addition to exec-shield explained above "Linux: Segfault during
2706`make bootstrap' under certain recent versions of the Linux kernel"
2707item, Linux kernel shipped with Fedora Core 4 randomizes the virtual
cf14a51c
JD
2708address space of a process. As the result dumping may fail even if
2709you turn off exec-shield. In this case, use the -R option to the setarch
ff0ab406
RS
2710command:
2711
25fd144d 2712 setarch i386 -R ./temacs --batch --load loadup [dump|bootstrap]
ff0ab406
RS
2713
2714or
2715
ade79051 2716 setarch i386 -R make bootstrap
a933dad1 2717
9dc15871 2718*** Fatal signal in the command temacs -l loadup inc dump.
a933dad1
DL
2719
2720This command is the final stage of building Emacs. It is run by the
7c2fb837 2721Makefile in the src subdirectory.
a933dad1
DL
2722
2723It has been known to get fatal errors due to insufficient swapping
2724space available on the machine.
2725
9dc15871 2726On 68000s, it has also happened because of bugs in the
a933dad1
DL
2727subroutine `alloca'. Verify that `alloca' works right, even
2728for large blocks (many pages).
2729
9dc15871
EZ
2730*** test-distrib says that the distribution has been clobbered.
2731*** or, temacs prints "Command key out of range 0-127".
2732*** or, temacs runs and dumps emacs, but emacs totally fails to work.
2733*** or, temacs gets errors dumping emacs.
a933dad1
DL
2734
2735This can be because the .elc files have been garbled. Do not be
2736fooled by the fact that most of a .elc file is text: these are
2737binary files and can contain all 256 byte values.
2738
2739In particular `shar' cannot be used for transmitting GNU Emacs.
2740It typically truncates "lines". What appear to be "lines" in
2741a binary file can of course be of any length. Even once `shar'
2742itself is made to work correctly, `sh' discards null characters
2743when unpacking the shell archive.
2744
2745I have also seen character \177 changed into \377. I do not know
2746what transfer means caused this problem. Various network
2747file transfer programs are suspected of clobbering the high bit.
2748
2749If you have a copy of Emacs that has been damaged in its
2750nonprinting characters, you can fix them:
2751
2752 1) Record the names of all the .elc files.
2753 2) Delete all the .elc files.
2754 3) Recompile alloc.c with a value of PURESIZE twice as large.
2755 (See puresize.h.) You might as well save the old alloc.o.
2756 4) Remake emacs. It should work now.
2757 5) Running emacs, do Meta-x byte-compile-file repeatedly
2758 to recreate all the .elc files that used to exist.
2759 You may need to increase the value of the variable
2760 max-lisp-eval-depth to succeed in running the compiler interpreted
2761 on certain .el files. 400 was sufficient as of last report.
2762 6) Reinstall the old alloc.o (undoing changes to alloc.c if any)
2763 and remake temacs.
2764 7) Remake emacs. It should work now, with valid .elc files.
2765
9dc15871 2766*** temacs prints "Pure Lisp storage exhausted".
a933dad1
DL
2767
2768This means that the Lisp code loaded from the .elc and .el
2769files during temacs -l loadup inc dump took up more
2770space than was allocated.
2771
2772This could be caused by
2773 1) adding code to the preloaded Lisp files
2774 2) adding more preloaded files in loadup.el
2775 3) having a site-init.el or site-load.el which loads files.
2776 Note that ANY site-init.el or site-load.el is nonstandard;
2777 if you have received Emacs from some other site
2778 and it contains a site-init.el or site-load.el file, consider
2779 deleting that file.
2780 4) getting the wrong .el or .elc files
2781 (not from the directory you expected).
2782 5) deleting some .elc files that are supposed to exist.
2783 This would cause the source files (.el files) to be
2784 loaded instead. They take up more room, so you lose.
2785 6) a bug in the Emacs distribution which underestimates
2786 the space required.
2787
2788If the need for more space is legitimate, change the definition
2789of PURESIZE in puresize.h.
2790
2791But in some of the cases listed above, this problem is a consequence
2792of something else that is wrong. Be sure to check and fix the real
2793problem.
2794
9dc15871 2795*** Linux: Emacs crashes when dumping itself on Mac PPC running Yellow Dog GNU/Linux.
a933dad1 2796
9dc15871
EZ
2797The crashes happen inside the function Fmake_symbol; here's a typical
2798C backtrace printed by GDB:
a933dad1 2799
9dc15871
EZ
2800 0x190c0c0 in Fmake_symbol ()
2801 (gdb) where
2802 #0 0x190c0c0 in Fmake_symbol ()
2803 #1 0x1942ca4 in init_obarray ()
2804 #2 0x18b3500 in main ()
2805 #3 0x114371c in __libc_start_main (argc=5, argv=0x7ffff5b4, envp=0x7ffff5cc,
a933dad1 2806
9dc15871
EZ
2807This could happen because GCC version 2.95 and later changed the base
2808of the load address to 0x10000000. Emacs needs to be told about this,
2809but we currently cannot do that automatically, because that breaks
2810other versions of GNU/Linux on the MacPPC. Until we find a way to
2811distinguish between the Yellow Dog and the other varieties of
2812GNU/Linux systems on the PPC, you will have to manually uncomment the
2813following section near the end of the file src/m/macppc.h in the Emacs
2814distribution:
a933dad1 2815
9dc15871 2816 #if 0 /* This breaks things on PPC GNU/Linux except for Yellowdog,
3256a475
RC
2817 even with identical GCC, as, ld. Let's take it out until we
2818 know what's really going on here. */
9dc15871
EZ
2819 /* GCC 2.95 and newer on GNU/Linux PPC changed the load address to
2820 0x10000000. */
2821 #if defined __linux__
2822 #if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 95)
2823 #define DATA_SEG_BITS 0x10000000
2824 #endif
2825 #endif
2826 #endif /* 0 */
a933dad1 2827
9dc15871
EZ
2828Remove the "#if 0" and "#endif" directives which surround this, save
2829the file, and then reconfigure and rebuild Emacs. The dumping process
2830should now succeed.
a933dad1 2831
0bce976c
GM
2832*** OpenBSD 4.0 macppc: Segfault during dumping.
2833
2834The build aborts with signal 11 when the command `./temacs --batch
14395431 2835--load loadup bootstrap' tries to load files.el. A workaround seems
464df798 2836to be to reduce the level of compiler optimization used during the
14395431
GM
2837build (from -O2 to -O1). It is possible this is an OpenBSD
2838GCC problem specific to the macppc architecture, possibly only
2839occurring with older versions of GCC (e.g. 3.3.5).
0bce976c 2840
ce46543c
GM
2841*** openSUSE 10.3: Segfault in bcopy during dumping.
2842
2843This is due to a bug in the bcopy implementation in openSUSE 10.3.
2844It is/will be fixed in an openSUSE update.
2845
9dc15871 2846** Installation
a933dad1 2847
9dc15871 2848*** Installing Emacs gets an error running `install-info'.
a933dad1 2849
9dc15871
EZ
2850You need to install a recent version of Texinfo; that package
2851supplies the `install-info' command.
a933dad1 2852
6afdff63
GM
2853*** Installing to a directory with spaces in the name fails.
2854
2855For example, if you call configure with a directory-related option
2856with spaces in the value, eg --enable-locallisppath='/path/with\ spaces'.
2857Using directory paths with spaces is not supported at this time: you
2858must re-configure without using spaces.
2859
6c06422c
CY
2860*** Installing to a directory with non-ASCII characters in the name fails.
2861
2862Installation may fail, or the Emacs executable may not start
2863correctly, if a directory name containing non-ASCII characters is used
2864as a `configure' argument (e.g. `--prefix'). The problem can also
2865occur if a non-ASCII directory is specified in the EMACSLOADPATH
2866envvar.
2867
2f6e7d3f
GM
2868*** On Solaris, use GNU Make when installing an out-of-tree build
2869
2870The Emacs configuration process allows you to configure the
2871build environment so that you can build emacs in a directory
2872outside of the distribution tree. When installing Emacs from an
2873out-of-tree build directory on Solaris, you may need to use GNU
2874make. The make programs bundled with Solaris support the VPATH
2875macro but use it differently from the way the VPATH macro is
2876used by GNU make. The differences will cause the "make install"
2877step to fail, leaving you with an incomplete emacs
2878installation. GNU make is available in /usr/sfw/bin on Solaris
287910 and can be installed as /opt/sfw/bin/gmake from the Solaris 9
2880Software Companion CDROM.
2881
2882The problems due to the VPATH processing differences affect only
2883out of tree builds so, if you are on a Solaris installation
2884without GNU make, you can install Emacs completely by installing
2885from a build environment using the original emacs distribution tree.
2886
9dc15871 2887** First execution
a933dad1 2888
9dc15871 2889*** Emacs binary is not in executable format, and cannot be run.
a933dad1 2890
9dc15871
EZ
2891This was reported to happen when Emacs is built in a directory mounted
2892via NFS, for some combinations of NFS client and NFS server.
2893Usually, the file `emacs' produced in these cases is full of
2894binary null characters, and the `file' utility says:
a933dad1 2895
9dc15871 2896 emacs: ASCII text, with no line terminators
a933dad1 2897
9dc15871
EZ
2898We don't know what exactly causes this failure. A work-around is to
2899build Emacs in a directory on a local disk.
a933dad1 2900
9dc15871 2901*** The dumped Emacs crashes when run, trying to write pure data.
a933dad1
DL
2902
2903Two causes have been seen for such problems.
2904
29051) On a system where getpagesize is not a system call, it is defined
2906as a macro. If the definition (in both unexec.c and malloc.c) is wrong,
2907it can cause problems like this. You might be able to find the correct
2908value in the man page for a.out (5).
2909
29102) Some systems allocate variables declared static among the
2911initialized variables. Emacs makes all initialized variables in most
2912of its files pure after dumping, but the variables declared static and
2913not initialized are not supposed to be pure. On these systems you
2914may need to add "#define static" to the m- or the s- file.
2915
9dc15871 2916* Emacs 19 problems
a933dad1 2917
9dc15871 2918** Error messages `Wrong number of arguments: #<subr where-is-internal>, 5'.
a933dad1 2919
9dc15871
EZ
2920This typically results from having the powerkey library loaded.
2921Powerkey was designed for Emacs 19.22. It is obsolete now because
2922Emacs 19 now has this feature built in; and powerkey also calls
2923where-is-internal in an obsolete way.
a933dad1 2924
9dc15871 2925So the fix is to arrange not to load powerkey.
a933dad1 2926
9dc15871 2927* Runtime problems on legacy systems
a933dad1 2928
9dc15871
EZ
2929This section covers bugs reported on very old hardware or software.
2930If you are using hardware and an operating system shipped after 2000,
2931it is unlikely you will see any of these.
a933dad1 2932
9dc15871 2933** Ancient operating systems
a933dad1 2934
0a4dd4e4
EZ
2935AIX 4.2 was end-of-lifed on Dec 31st, 1999.
2936
2937*** AIX: You get this compiler error message:
2938
2939 Processing include file ./XMenuInt.h
2940 1501-106: (S) Include file X11/Xlib.h not found.
2941
2942This means your system was installed with only the X11 runtime i.d
2943libraries. You have to find your sipo (bootable tape) and install
2944X11Dev... with smit.
2945
2946(This report must be ancient. Bootable tapes are long dead.)
2947
2948*** AIX 3.2.4: Releasing Ctrl/Act key has no effect, if Shift is down.
2949
2950Due to a feature of AIX, pressing or releasing the Ctrl/Act key is
2951ignored when the Shift, Alt or AltGr keys are held down. This can
2952lead to the keyboard being "control-locked"--ordinary letters are
2953treated as control characters.
2954
2955You can get out of this "control-locked" state by pressing and
2956releasing Ctrl/Act while not pressing or holding any other keys.
2957
2958*** AIX 3.2.5: You get this message when running Emacs:
2959
2960 Could not load program emacs
2961 Symbol smtcheckinit in csh is undefined
2962 Error was: Exec format error
2963
2964or this one:
2965
2966 Could not load program .emacs
2967 Symbol _system_con in csh is undefined
2968 Symbol _fp_trapsta in csh is undefined
2969 Error was: Exec format error
2970
2971These can happen when you try to run on AIX 3.2.5 a program that was
2972compiled with 3.2.4. The fix is to recompile.
2973
2974*** AIX 4.2: Emacs gets a segmentation fault at startup.
2975
2976If you are using IBM's xlc compiler, compile emacs.c
2977without optimization; that should avoid the problem.
2978
9dc15871 2979*** ISC Unix
a933dad1 2980
9dc15871 2981**** ISC: display-time causes kernel problems on ISC systems.
a933dad1 2982
9dc15871
EZ
2983Under Interactive Unix versions 3.0.1 and 4.0 (and probably other
2984versions), display-time causes the loss of large numbers of STREVENT
2985cells. Eventually the kernel's supply of these cells is exhausted.
2986This makes emacs and the whole system run slow, and can make other
2987processes die, in particular pcnfsd.
a933dad1 2988
9dc15871
EZ
2989Other emacs functions that communicate with remote processes may have
2990the same problem. Display-time seems to be far the worst.
a933dad1 2991
9dc15871 2992The only known fix: Don't run display-time.
a933dad1 2993
0a4dd4e4
EZ
2994**** Sunos 5.3: Subprocesses remain, hanging but not zombies.
2995
2996A bug in Sunos 5.3 causes Emacs subprocesses to remain after Emacs
2997exits. Sun patch # 101415-02 is part of the fix for this, but it only
2998applies to ptys, and doesn't fix the problem with subprocesses
2999communicating through pipes.
3000
9dc15871 3001*** Irix
a933dad1 3002
9dc15871 3003*** Irix 6.2: No visible display on mips-sgi-irix6.2 when compiling with GCC 2.8.1.
a933dad1 3004
9dc15871
EZ
3005This problem went away after installing the latest IRIX patches
3006as of 8 Dec 1998.
a933dad1 3007
9dc15871 3008The same problem has been reported on Irix 6.3.
a933dad1 3009
9dc15871
EZ
3010*** Irix 6.3: substituting environment variables in file names
3011in the minibuffer gives peculiar error messages such as
a933dad1 3012
9dc15871 3013 Substituting nonexistent environment variable ""
a933dad1 3014
9dc15871
EZ
3015This is not an Emacs bug; it is caused by something in SGI patch
3016003082 August 11, 1998.
a933dad1 3017
9dc15871 3018*** OPENSTEP
a933dad1 3019
9dc15871 3020**** OPENSTEP 4.2: Compiling syntax.c with gcc 2.7.2.1 fails.
a933dad1 3021
9dc15871
EZ
3022The compiler was reported to crash while compiling syntax.c with the
3023following message:
a933dad1 3024
9dc15871 3025 cc: Internal compiler error: program cc1obj got fatal signal 11
a933dad1 3026
9dc15871
EZ
3027To work around this, replace the macros UPDATE_SYNTAX_TABLE_FORWARD,
3028INC_BOTH, and INC_FROM with functions. To this end, first define 3
3029functions, one each for every macro. Here's an example:
a933dad1 3030
9dc15871
EZ
3031 static int update_syntax_table_forward(int from)
3032 {
3033 return(UPDATE_SYNTAX_TABLE_FORWARD(from));
3034 }/*update_syntax_table_forward*/
a933dad1 3035
9dc15871
EZ
3036Then replace all references to UPDATE_SYNTAX_TABLE_FORWARD in syntax.c
3037with a call to the function update_syntax_table_forward.
a933dad1 3038
9dc15871 3039*** Solaris 2.x
a933dad1 3040
9dc15871 3041**** Strange results from format %d in a few cases, on a Sun.
a933dad1 3042
9dc15871
EZ
3043Sun compiler version SC3.0 has been found to miscompile part of
3044editfns.c. The workaround is to compile with some other compiler such
3045as GCC.
a933dad1 3046
9dc15871 3047**** On Solaris, Emacs dumps core if lisp-complete-symbol is called.
a933dad1 3048
9dc15871
EZ
3049If you compile Emacs with the -fast or -xO4 option with version 3.0.2
3050of the Sun C compiler, Emacs dumps core when lisp-complete-symbol is
3051called. The problem does not happen if you compile with GCC.
a933dad1 3052
9dc15871 3053**** On Solaris, Emacs crashes if you use (display-time).
a933dad1 3054
9dc15871
EZ
3055This can happen if you configure Emacs without specifying the precise
3056version of Solaris that you are using.
a933dad1 3057
9dc15871 3058**** Solaris 2.3 and 2.4: Unpredictable segmentation faults.
a933dad1 3059
9dc15871
EZ
3060A user reported that this happened in 19.29 when it was compiled with
3061the Sun compiler, but not when he recompiled with GCC 2.7.0.
a933dad1 3062
9dc15871 3063We do not know whether something in Emacs is partly to blame for this.
a933dad1 3064
9dc15871 3065**** Solaris 2.4: Emacs dumps core on startup.
a933dad1 3066
9dc15871
EZ
3067Bill Sebok says that the cause of this is Solaris 2.4 vendor patch
3068102303-05, which extends the Solaris linker to deal with the Solaris
3069Common Desktop Environment's linking needs. You can fix the problem
3070by removing this patch and installing patch 102049-02 instead.
3071However, that linker version won't work with CDE.
a933dad1 3072
9dc15871
EZ
3073Solaris 2.5 comes with a linker that has this bug. It is reported that if
3074you install all the latest patches (as of June 1996), the bug is fixed.
3075We suspect the crucial patch is one of these, but we don't know
3076for certain.
a933dad1 3077
9dc15871
EZ
3078 103093-03: [README] SunOS 5.5: kernel patch (2140557 bytes)
3079 102832-01: [README] OpenWindows 3.5: Xview Jumbo Patch (4181613 bytes)
3256a475 3080 103242-04: [README] SunOS 5.5: linker patch (595363 bytes)
a933dad1 3081
9dc15871
EZ
3082(One user reports that the bug was fixed by those patches together
3083with patches 102980-04, 103279-01, 103300-02, and 103468-01.)
a933dad1 3084
9dc15871
EZ
3085If you can determine which patch does fix the bug, please tell
3086bug-gnu-emacs@gnu.org.
a933dad1 3087
9dc15871
EZ
3088Meanwhile, the GNU linker links Emacs properly on both Solaris 2.4 and
3089Solaris 2.5.
a933dad1 3090
9dc15871
EZ
3091**** Solaris 2.4: Dired hangs and C-g does not work. Or Emacs hangs
3092forever waiting for termination of a subprocess that is a zombie.
a933dad1 3093
9dc15871
EZ
3094casper@fwi.uva.nl says the problem is in X11R6. Rebuild libX11.so
3095after changing the file xc/config/cf/sunLib.tmpl. Change the lines
a933dad1 3096
9dc15871
EZ
3097 #if ThreadedX
3098 #define SharedX11Reqs -lthread
3099 #endif
a933dad1 3100
9dc15871 3101to:
a933dad1 3102
9dc15871
EZ
3103 #if OSMinorVersion < 4
3104 #if ThreadedX
3105 #define SharedX11Reqs -lthread
3106 #endif
3107 #endif
3108
3109Be sure also to edit x/config/cf/sun.cf so that OSMinorVersion is 4
3110(as it should be for Solaris 2.4). The file has three definitions for
3111OSMinorVersion: the first is for x86, the second for SPARC under
3112Solaris, and the third for SunOS 4. Make sure to update the
3113definition for your type of machine and system.
a933dad1 3114
9dc15871
EZ
3115Then do `make Everything' in the top directory of X11R6, to rebuild
3116the makefiles and rebuild X. The X built this way work only on
3117Solaris 2.4, not on 2.3.
a933dad1 3118
9dc15871
EZ
3119For multithreaded X to work it is necessary to install patch
3120101925-02 to fix problems in header files [2.4]. You need
3121to reinstall gcc or re-run just-fixinc after installing that
3122patch.
a933dad1 3123
9dc15871
EZ
3124However, Frank Rust <frust@iti.cs.tu-bs.de> used a simpler solution:
3125he changed
3126 #define ThreadedX YES
3127to
3128 #define ThreadedX NO
3129in sun.cf and did `make World' to rebuild X11R6. Removing all
3130`-DXTHREAD*' flags and `-lthread' entries from lib/X11/Makefile and
3131typing 'make install' in that directory also seemed to work.
a933dad1 3132
9dc15871 3133**** Solaris 2.x: GCC complains "64 bit integer types not supported".
a933dad1 3134
9dc15871
EZ
3135This suggests that GCC is not installed correctly. Most likely you
3136are using GCC 2.7.2.3 (or earlier) on Solaris 2.6 (or later); this
3137does not work without patching. To run GCC 2.7.2.3 on Solaris 2.6 or
3138later, you must patch fixinc.svr4 and reinstall GCC from scratch as
3139described in the Solaris FAQ
3140<http://www.wins.uva.nl/pub/solaris/solaris2.html>. A better fix is
3141to upgrade to GCC 2.8.1 or later.
a933dad1 3142
9dc15871
EZ
3143**** Solaris 2.7: Building Emacs with WorkShop Compilers 5.0 98/12/15
3144C 5.0 failed, apparently with non-default CFLAGS, most probably due to
3145compiler bugs. Using Sun Solaris 2.7 Sun WorkShop 6 update 1 C
3146release was reported to work without problems. It worked OK on
3147another system with Solaris 8 using apparently the same 5.0 compiler
3148and the default CFLAGS.
a933dad1 3149
9dc15871 3150**** Solaris 2.x: Emacs dumps core when built with Motif.
a933dad1 3151
9dc15871
EZ
3152The Solaris Motif libraries are buggy, at least up through Solaris 2.5.1.
3153Install the current Motif runtime library patch appropriate for your host.
3154(Make sure the patch is current; some older patch versions still have the bug.)
3155You should install the other patches recommended by Sun for your host, too.
3156You can obtain Sun patches from ftp://sunsolve.sun.com/pub/patches/;
3157look for files with names ending in `.PatchReport' to see which patches
3158are currently recommended for your host.
a933dad1 3159
9dc15871
EZ
3160On Solaris 2.6, Emacs is said to work with Motif when Solaris patch
3161105284-12 is installed, but fail when 105284-15 is installed.
3162105284-18 might fix it again.
a933dad1 3163
0a4dd4e4 3164**** Solaris 2.6 and 7: the Compose key does not work.
a933dad1 3165
9dc15871
EZ
3166This is a bug in Motif in Solaris. Supposedly it has been fixed for
3167the next major release of Solaris. However, if someone with Sun
3168support complains to Sun about the bug, they may release a patch.
3169If you do this, mention Sun bug #4188711.
a933dad1 3170
9dc15871
EZ
3171One workaround is to use a locale that allows non-ASCII characters.
3172For example, before invoking emacs, set the LC_ALL environment
3173variable to "en_US" (American English). The directory /usr/lib/locale
3174lists the supported locales; any locale other than "C" or "POSIX"
3175should do.
a933dad1 3176
9dc15871
EZ
3177pen@lysator.liu.se says (Feb 1998) that the Compose key does work
3178if you link with the MIT X11 libraries instead of the Solaris X11
3179libraries.
a933dad1 3180
0a4dd4e4
EZ
3181*** HP/UX versions before 11.0
3182
bf247b6e 3183HP/UX 9 was end-of-lifed in December 1998.
0a4dd4e4
EZ
3184HP/UX 10 was end-of-lifed in May 1999.
3185
3186**** HP/UX 9: Emacs crashes with SIGBUS or SIGSEGV after you delete a frame.
3187
3188We think this is due to a bug in the X libraries provided by HP. With
3189the alternative X libraries in /usr/contrib/mitX11R5/lib, the problem
3190does not happen.
3191
3192*** HP/UX 10: Large file support is disabled.
3193
3194See the comments in src/s/hpux10.h.
3195
3196*** HP/UX: Emacs is slow using X11R5.
3197
3198This happens if you use the MIT versions of the X libraries--it
3199doesn't run as fast as HP's version. People sometimes use the version
3200because they see the HP version doesn't have the libraries libXaw.a,
3201libXmu.a, libXext.a and others. HP/UX normally doesn't come with
3202those libraries installed. To get good performance, you need to
3203install them and rebuild Emacs.
3204
9dc15871 3205*** Ultrix and Digital Unix
a933dad1 3206
9dc15871 3207**** Ultrix 4.2: `make install' fails on install-doc with `Error 141'.
a933dad1 3208
9dc15871
EZ
3209This happens on Ultrix 4.2 due to failure of a pipeline of tar
3210commands. We don't know why they fail, but the bug seems not to be in
3211Emacs. The workaround is to run the shell command in install-doc by
3212hand.
a933dad1 3213
9dc15871 3214**** Digital Unix 4.0: Garbled display on non-X terminals when Emacs runs.
a933dad1 3215
9dc15871
EZ
3216So far it appears that running `tset' triggers this problem (when TERM
3217is vt100, at least). If you do not run `tset', then Emacs displays
3218properly. If someone can tell us precisely which effect of running
3219`tset' actually causes the problem, we may be able to implement a fix
3220in Emacs.
a933dad1 3221
9dc15871 3222**** Ultrix: `expand-file-name' fails to work on any but the machine you dumped Emacs on.
a933dad1 3223
9dc15871
EZ
3224On Ultrix, if you use any of the functions which look up information
3225in the passwd database before dumping Emacs (say, by using
3226expand-file-name in site-init.el), then those functions will not work
3227in the dumped Emacs on any host but the one Emacs was dumped on.
a933dad1 3228
9dc15871
EZ
3229The solution? Don't use expand-file-name in site-init.el, or in
3230anything it loads. Yuck - some solution.
a933dad1 3231
9dc15871
EZ
3232I'm not sure why this happens; if you can find out exactly what is
3233going on, and perhaps find a fix or a workaround, please let us know.
3234Perhaps the YP functions cache some information, the cache is included
3235in the dumped Emacs, and is then inaccurate on any other host.
a933dad1 3236
9dc15871 3237*** SVr4
a933dad1 3238
9dc15871 3239**** SVr4: On some variants of SVR4, Emacs does not work at all with X.
a933dad1 3240
9dc15871
EZ
3241Try defining BROKEN_FIONREAD in your config.h file. If this solves
3242the problem, please send a bug report to tell us this is needed; be
3243sure to say exactly what type of machine and system you are using.
a933dad1 3244
9dc15871 3245**** SVr4: After running emacs once, subsequent invocations crash.
a933dad1 3246
9dc15871
EZ
3247Some versions of SVR4 have a serious bug in the implementation of the
3248mmap () system call in the kernel; this causes emacs to run correctly
3249the first time, and then crash when run a second time.
a933dad1 3250
9dc15871
EZ
3251Contact your vendor and ask for the mmap bug fix; in the mean time,
3252you may be able to work around the problem by adding a line to your
3253operating system description file (whose name is reported by the
3254configure script) that reads:
3255#define SYSTEM_MALLOC
3256This makes Emacs use memory less efficiently, but seems to work around
3257the kernel bug.
a933dad1 3258
0a4dd4e4
EZ
3259*** Irix 5 and earlier
3260
bf247b6e 3261Exactly when Irix-5 end-of-lifed is obscure. But since Irix 6.0
0a4dd4e4
EZ
3262shipped in 1994, it has been some years.
3263
3264**** Irix 5.2: unexelfsgi.c can't find cmplrs/stsupport.h.
3265
3266The file cmplrs/stsupport.h was included in the wrong file set in the
3267Irix 5.2 distribution. You can find it in the optional fileset
3268compiler_dev, or copy it from some other Irix 5.2 system. A kludgy
3269workaround is to change unexelfsgi.c to include sym.h instead of
3270syms.h.
3271
3272**** Irix 5.3: "out of virtual swap space".
3273
3274This message occurs when the system runs out of swap space due to too
3275many large programs running. The solution is either to provide more
3276swap space or to reduce the number of large programs being run. You
3277can check the current status of the swap space by executing the
3278command `swap -l'.
3279
3280You can increase swap space by changing the file /etc/fstab. Adding a
3281line like this:
3282
3283/usr/swap/swap.more swap swap pri=3 0 0
3284
3285where /usr/swap/swap.more is a file previously created (for instance
3286by using /etc/mkfile), will increase the swap space by the size of
3287that file. Execute `swap -m' or reboot the machine to activate the
3288new swap area. See the manpages for `swap' and `fstab' for further
3289information.
3290
3291The objectserver daemon can use up lots of memory because it can be
3292swamped with NIS information. It collects information about all users
3293on the network that can log on to the host.
3294
3295If you want to disable the objectserver completely, you can execute
3296the command `chkconfig objectserver off' and reboot. That may disable
3297some of the window system functionality, such as responding CDROM
3298icons.
3299
3300You can also remove NIS support from the objectserver. The SGI `admin'
3301FAQ has a detailed description on how to do that; see question 35
3302("Why isn't the objectserver working?"). The admin FAQ can be found at
3303ftp://viz.tamu.edu/pub/sgi/faq/.
3304
3305**** Irix 5.3: Emacs crashes in utmpname.
3306
3307This problem is fixed in Patch 3175 for Irix 5.3.
3308It is also fixed in Irix versions 6.2 and up.
3309
3310**** Irix 6.0: Make tries (and fails) to build a program named unexelfsgi.
3311
3312A compiler bug inserts spaces into the string "unexelfsgi . o"
3313in src/Makefile. Edit src/Makefile, after configure is run,
3314find that string, and take out the spaces.
3315
3316Compiler fixes in Irix 6.0.1 should eliminate this problem.
3317
3318*** SCO Unix and UnixWare
3319
3320**** SCO 3.2v4: Unusable default font.
3321
3322The Open Desktop environment comes with default X resource settings
3323that tell Emacs to use a variable-width font. Emacs cannot use such
3324fonts, so it does not work.
3325
3326This is caused by the file /usr/lib/X11/app-defaults/ScoTerm, which is
3327the application-specific resource file for the `scoterm' terminal
3328emulator program. It contains several extremely general X resources
3329that affect other programs besides `scoterm'. In particular, these
3330resources affect Emacs also:
3331
3256a475
RC
3332 *Font: -*-helvetica-medium-r-*--12-*-p-*
3333 *Background: scoBackground
3334 *Foreground: scoForeground
0a4dd4e4
EZ
3335
3336The best solution is to create an application-specific resource file for
3337Emacs, /usr/lib/X11/sco/startup/Emacs, with the following contents:
3338
3256a475
RC
3339 Emacs*Font: -*-courier-medium-r-*-*-*-120-*-*-*-*-iso8859-1
3340 Emacs*Background: white
3341 Emacs*Foreground: black
0a4dd4e4
EZ
3342
3343(These settings mimic the Emacs defaults, but you can change them to
3344suit your needs.) This resource file is only read when the X server
3345starts up, so you should restart it by logging out of the Open Desktop
3346environment or by running `scologin stop; scologin start` from the shell
3347as root. Alternatively, you can put these settings in the
3348/usr/lib/X11/app-defaults/Emacs resource file and simply restart Emacs,
3349but then they will not affect remote invocations of Emacs that use the
3350Open Desktop display.
3351
3352These resource files are not normally shared across a network of SCO
3353machines; you must create the file on each machine individually.
3354
3355**** SCO 4.2.0: Regular expressions matching bugs on SCO systems.
3356
3357On SCO, there are problems in regexp matching when Emacs is compiled
3358with the system compiler. The compiler version is "Microsoft C
3359version 6", SCO 4.2.0h Dev Sys Maintenance Supplement 01/06/93; Quick
3360C Compiler Version 1.00.46 (Beta). The solution is to compile with
3361GCC.
3362
3363**** UnixWare 2.1: Error 12 (virtual memory exceeded) when dumping Emacs.
3364
3365Paul Abrahams (abrahams@acm.org) reports that with the installed
3366virtual memory settings for UnixWare 2.1.2, an Error 12 occurs during
3367the "make" that builds Emacs, when running temacs to dump emacs. That
3368error indicates that the per-process virtual memory limit has been
3369exceeded. The default limit is probably 32MB. Raising the virtual
3370memory limit to 40MB should make it possible to finish building Emacs.
3371
3372You can do this with the command `ulimit' (sh) or `limit' (csh).
3373But you have to be root to do it.
3374
3375According to Martin Sohnius, you can also retune this in the kernel:
3376
3377 # /etc/conf/bin/idtune SDATLIM 33554432 ## soft data size limit
3378 # /etc/conf/bin/idtune HDATLIM 33554432 ## hard "
3379 # /etc/conf/bin/idtune SVMMSIZE unlimited ## soft process size limit
3380 # /etc/conf/bin/idtune HVMMSIZE unlimited ## hard "
3381 # /etc/conf/bin/idbuild -B
3382
3383(He recommends you not change the stack limit, though.)
3384These changes take effect when you reboot.
3385
9dc15871 3386*** Linux 1.x
a933dad1 3387
9dc15871 3388**** Linux 1.0-1.04: Typing C-c C-c in Shell mode kills your X server.
a933dad1 3389
9dc15871
EZ
3390This happens with Linux kernel 1.0 thru 1.04, approximately. The workaround is
3391to define SIGNALS_VIA_CHARACTERS in config.h and recompile Emacs.
3392Newer Linux kernel versions don't have this problem.
a933dad1 3393
9dc15871
EZ
3394**** Linux 1.3: Output from subprocess (such as man or diff) is randomly
3395truncated on GNU/Linux systems.
a933dad1 3396
9dc15871
EZ
3397This is due to a kernel bug which seems to be fixed in Linux version
33981.3.75.
a933dad1 3399
0a4dd4e4
EZ
3400** Windows 3.1, 95, 98, and ME
3401
3402*** MS-Windows NT/95: Problems running Perl under Emacs
3403
3404`perl -de 0' just hangs when executed in an Emacs subshell.
3405The fault lies with Perl (indirectly with Windows NT/95).
3406
3407The problem is that the Perl debugger explicitly opens a connection to
3408"CON", which is the DOS/NT equivalent of "/dev/tty", for interacting
3409with the user.
3410
3411On Unix, this is okay, because Emacs (or the shell?) creates a
3412pseudo-tty so that /dev/tty is really the pipe Emacs is using to
3413communicate with the subprocess.
3414
3415On NT, this fails because CON always refers to the handle for the
3416relevant console (approximately equivalent to a tty), and cannot be
3417redirected to refer to the pipe Emacs assigned to the subprocess as
3418stdin.
3419
3420A workaround is to modify perldb.pl to use STDIN/STDOUT instead of CON.
3421
3422For Perl 4:
3423
3424 *** PERL/LIB/PERLDB.PL.orig Wed May 26 08:24:18 1993
3425 --- PERL/LIB/PERLDB.PL Mon Jul 01 15:28:16 1996
3426 ***************
3427 *** 68,74 ****
3256a475 3428 $rcfile=".perldb";
0a4dd4e4
EZ
3429 }
3430 else {
3431 ! $console = "con";
3256a475 3432 $rcfile="perldb.ini";
0a4dd4e4
EZ
3433 }
3434
3435 --- 68,74 ----
3256a475 3436 $rcfile=".perldb";
0a4dd4e4
EZ
3437 }
3438 else {
3439 ! $console = "";
3256a475 3440 $rcfile="perldb.ini";
0a4dd4e4
EZ
3441 }
3442
3443
3444 For Perl 5:
3445 *** perl/5.001/lib/perl5db.pl.orig Sun Jun 04 21:13:40 1995
3446 --- perl/5.001/lib/perl5db.pl Mon Jul 01 17:00:08 1996
3447 ***************
3448 *** 22,28 ****
3256a475 3449 $rcfile=".perldb";
0a4dd4e4
EZ
3450 }
3451 elsif (-e "con") {
3452 ! $console = "con";
3256a475 3453 $rcfile="perldb.ini";
0a4dd4e4
EZ
3454 }
3455 else {
3456 --- 22,28 ----
3256a475 3457 $rcfile=".perldb";
0a4dd4e4
EZ
3458 }
3459 elsif (-e "con") {
3460 ! $console = "";
3256a475 3461 $rcfile="perldb.ini";
0a4dd4e4
EZ
3462 }
3463 else {
3464
3465*** MS-Windows 95: Alt-f6 does not get through to Emacs.
3466
3467This character seems to be trapped by the kernel in Windows 95.
3468You can enter M-f6 by typing ESC f6.
3469
3470*** MS-Windows 95/98/ME: subprocesses do not terminate properly.
3471
3472This is a limitation of the Operating System, and can cause problems
3473when shutting down Windows. Ensure that all subprocesses are exited
3474cleanly before exiting Emacs. For more details, see the FAQ at
3475http://www.gnu.org/software/emacs/windows/.
3476
3477*** MS-Windows 95/98/ME: crashes when Emacs invokes non-existent programs.
3478
3479When a program you are trying to run is not found on the PATH,
3480Windows might respond by crashing or locking up your system. In
3481particular, this has been reported when trying to compile a Java
3482program in JDEE when javac.exe is installed, but not on the system
3483PATH.
3484
9dc15871 3485** MS-DOS
a933dad1 3486
9dc15871 3487*** When compiling with DJGPP on MS-Windows NT, "config msdos" fails.
a933dad1 3488
9dc15871
EZ
3489If the error message is "VDM has been already loaded", this is because
3490Windows has a program called `redir.exe' that is incompatible with a
3491program by the same name supplied with DJGPP, which is used by
3492config.bat. To resolve this, move the DJGPP's `bin' subdirectory to
3493the front of your PATH environment variable.
a933dad1 3494
9dc15871
EZ
3495*** When compiling with DJGPP on MS-Windows 95, Make fails for some targets
3496like make-docfile.
a933dad1 3497
9dc15871
EZ
3498This can happen if long file name support (the setting of environment
3499variable LFN) when Emacs distribution was unpacked and during
3500compilation are not the same. See the MSDOG section of INSTALL for
3501the explanation of how to avoid this problem.
a933dad1 3502
9dc15871 3503*** Emacs compiled with DJGPP complains at startup:
a933dad1 3504
9dc15871 3505 "Wrong type of argument: internal-facep, msdos-menu-active-face"
a933dad1 3506
9dc15871
EZ
3507This can happen if you define an environment variable `TERM'. Emacs
3508on MSDOS uses an internal terminal emulator which is disabled if the
3509value of `TERM' is anything but the string "internal". Emacs then
3510works as if its terminal were a dumb glass teletype that doesn't
3511support faces. To work around this, arrange for `TERM' to be
3512undefined when Emacs runs. The best way to do that is to add an
3513[emacs] section to the DJGPP.ENV file which defines an empty value for
3514`TERM'; this way, only Emacs gets the empty value, while the rest of
3515your system works as before.
a933dad1 3516
9dc15871 3517*** MS-DOS: Emacs crashes at startup.
a933dad1 3518
9dc15871
EZ
3519Some users report that Emacs 19.29 requires dpmi memory management,
3520and crashes on startup if the system does not have it. We don't yet
3521know why this happens--perhaps these machines don't have enough real
3522memory, or perhaps something is wrong in Emacs or the compiler.
3523However, arranging to use dpmi support is a workaround.
a933dad1 3524
9dc15871
EZ
3525You can find out if you have a dpmi host by running go32 without
3526arguments; it will tell you if it uses dpmi memory. For more
3527information about dpmi memory, consult the djgpp FAQ. (djgpp
3528is the GNU C compiler as packaged for MSDOS.)
a933dad1 3529
9dc15871
EZ
3530Compiling Emacs under MSDOS is extremely sensitive for proper memory
3531configuration. If you experience problems during compilation, consider
3532removing some or all memory resident programs (notably disk caches)
3533and make sure that your memory managers are properly configured. See
3534the djgpp faq for configuration hints.
a933dad1 3535
9dc15871
EZ
3536*** Emacs compiled with DJGPP for MS-DOS/MS-Windows cannot access files
3537in the directory with the special name `dev' under the root of any
3538drive, e.g. `c:/dev'.
a933dad1 3539
9dc15871
EZ
3540This is an unfortunate side-effect of the support for Unix-style
3541device names such as /dev/null in the DJGPP runtime library. A
3542work-around is to rename the problem directory to another name.
a933dad1 3543
9dc15871 3544*** MS-DOS+DJGPP: Problems on MS-DOG if DJGPP v2.0 is used to compile Emacs.
a933dad1 3545
9dc15871 3546There are two DJGPP library bugs which cause problems:
a933dad1 3547
9dc15871
EZ
3548 * Running `shell-command' (or `compile', or `grep') you get
3549 `Searching for program: permission denied (EACCES), c:/command.com';
3550 * After you shell to DOS, Ctrl-Break kills Emacs.
a933dad1 3551
9dc15871
EZ
3552To work around these bugs, you can use two files in the msdos
3553subdirectory: `is_exec.c' and `sigaction.c'. Compile them and link
3554them into the Emacs executable `temacs'; then they will replace the
3555incorrect library functions.
a933dad1 3556
9dc15871
EZ
3557*** MS-DOS: Emacs compiled for MSDOS cannot find some Lisp files, or other
3558run-time support files, when long filename support is enabled.
a933dad1 3559
9dc15871
EZ
3560Usually, this problem will manifest itself when Emacs exits
3561immediately after flashing the startup screen, because it cannot find
3562the Lisp files it needs to load at startup. Redirect Emacs stdout
3563and stderr to a file to see the error message printed by Emacs.
a933dad1 3564
9dc15871
EZ
3565Another manifestation of this problem is that Emacs is unable to load
3566the support for editing program sources in languages such as C and
3567Lisp.
a933dad1 3568
9dc15871
EZ
3569This can happen if the Emacs distribution was unzipped without LFN
3570support, thus causing long filenames to be truncated to the first 6
3571characters and a numeric tail that Windows 95 normally attaches to it.
3572You should unzip the files again with a utility that supports long
3573filenames (such as djtar from DJGPP or InfoZip's UnZip program
3574compiled with DJGPP v2). The MSDOG section of the file INSTALL
3575explains this issue in more detail.
a933dad1 3576
9dc15871
EZ
3577Another possible reason for such failures is that Emacs compiled for
3578MSDOS is used on Windows NT, where long file names are not supported
3579by this version of Emacs, but the distribution was unpacked by an
3580unzip program that preserved the long file names instead of truncating
3581them to DOS 8+3 limits. To be useful on NT, the MSDOS port of Emacs
3582must be unzipped by a DOS utility, so that long file names are
3583properly truncated.
a933dad1 3584
9dc15871 3585** Archaic window managers and toolkits
a933dad1 3586
9dc15871 3587*** OpenLook: Under OpenLook, the Emacs window disappears when you type M-q.
a933dad1 3588
9dc15871
EZ
3589Some versions of the Open Look window manager interpret M-q as a quit
3590command for whatever window you are typing at. If you want to use
3591Emacs with that window manager, you should try to configure the window
3592manager to use some other command. You can disable the
3593shortcut keys entirely by adding this line to ~/.OWdefaults:
a933dad1 3594
9dc15871 3595 OpenWindows.WindowMenuAccelerators: False
a933dad1 3596
9dc15871 3597**** twm: A position you specified in .Xdefaults is ignored, using twm.
a933dad1 3598
9dc15871
EZ
3599twm normally ignores "program-specified" positions.
3600You can tell it to obey them with this command in your `.twmrc' file:
a933dad1 3601
9dc15871 3602 UsePPosition "on" #allow clients to request a position
a933dad1 3603
9dc15871 3604** Bugs related to old DEC hardware
a933dad1 3605
9dc15871 3606*** The Compose key on a DEC keyboard does not work as Meta key.
a933dad1 3607
9dc15871 3608This shell command should fix it:
a933dad1 3609
9dc15871 3610 xmodmap -e 'keycode 0xb1 = Meta_L'
a933dad1 3611
9dc15871
EZ
3612*** Keyboard input gets confused after a beep when using a DECserver
3613as a concentrator.
a933dad1 3614
9dc15871
EZ
3615This problem seems to be a matter of configuring the DECserver to use
36167 bit characters rather than 8 bit characters.
a933dad1 3617
9dc15871 3618* Build problems on legacy systems
a933dad1 3619
9dc15871 3620** BSD/386 1.0: --with-x-toolkit option configures wrong.
a933dad1 3621
9dc15871
EZ
3622This problem is due to bugs in the shell in version 1.0 of BSD/386.
3623The workaround is to edit the configure file to use some other shell,
3624such as bash.
a933dad1 3625
9dc15871
EZ
3626** Digital Unix 4.0: Emacs fails to build, giving error message
3627 Invalid dimension for the charset-ID 160
a933dad1 3628
9dc15871
EZ
3629This is due to a bug or an installation problem in GCC 2.8.0.
3630Installing a more recent version of GCC fixes the problem.
a933dad1 3631
9dc15871 3632** Digital Unix 4.0: Failure in unexec while dumping emacs.
a933dad1 3633
9dc15871 3634This problem manifests itself as an error message
a933dad1 3635
9dc15871 3636 unexec: Bad address, writing data section to ...
a933dad1 3637
9dc15871
EZ
3638The user suspects that this happened because his X libraries
3639were built for an older system version,
a933dad1 3640
9dc15871 3641 ./configure --x-includes=/usr/include --x-libraries=/usr/shlib
a933dad1 3642
9dc15871 3643made the problem go away.
a933dad1 3644
9dc15871 3645** Sunos 4.1.1: there are errors compiling sysdep.c.
a933dad1 3646
9dc15871 3647If you get errors such as
a933dad1 3648
9dc15871
EZ
3649 "sysdep.c", line 2017: undefined structure or union
3650 "sysdep.c", line 2017: undefined structure or union
3651 "sysdep.c", line 2019: nodename undefined
a933dad1 3652
9dc15871
EZ
3653This can result from defining LD_LIBRARY_PATH. It is very tricky
3654to use that environment variable with Emacs. The Emacs configure
3655script links many test programs with the system libraries; you must
3656make sure that the libraries available to configure are the same
3657ones available when you build Emacs.
a933dad1 3658
9dc15871 3659** SunOS 4.1.1: You get this error message from GNU ld:
a933dad1 3660
9dc15871 3661 /lib/libc.a(_Q_sub.o): Undefined symbol __Q_get_rp_rd referenced from text segment
a933dad1 3662
9dc15871 3663The problem is in the Sun shared C library, not in GNU ld.
a933dad1 3664
9dc15871 3665The solution is to install Patch-ID# 100267-03 from Sun.
a933dad1 3666
9dc15871 3667** Sunos 4.1: Undefined symbols when linking using --with-x-toolkit.
a933dad1 3668
9dc15871
EZ
3669If you get the undefined symbols _atowc _wcslen, _iswprint, _iswspace,
3670_iswcntrl, _wcscpy, and _wcsncpy, then you need to add -lXwchar after
3671-lXaw in the command that links temacs.
a933dad1 3672
9dc15871
EZ
3673This problem seems to arise only when the international language
3674extensions to X11R5 are installed.
a933dad1 3675
9dc15871 3676** SunOS: Emacs gets error message from linker on Sun.
a933dad1 3677
9dc15871
EZ
3678If the error message says that a symbol such as `f68881_used' or
3679`ffpa_used' or `start_float' is undefined, this probably indicates
3680that you have compiled some libraries, such as the X libraries,
3681with a floating point option other than the default.
a933dad1 3682
9dc15871
EZ
3683It's not terribly hard to make this work with small changes in
3684crt0.c together with linking with Fcrt1.o, Wcrt1.o or Mcrt1.o.
3685However, the easiest approach is to build Xlib with the default
3686floating point option: -fsoft.
a933dad1 3687
0a4dd4e4
EZ
3688** HPUX 10.20: Emacs crashes during dumping on the HPPA machine.
3689
3690This seems to be due to a GCC bug; it is fixed in GCC 2.8.1.
3691
9dc15871 3692** Vax C compiler bugs affecting Emacs.
a933dad1
DL
3693
3694You may get one of these problems compiling Emacs:
3695
3696 foo.c line nnn: compiler error: no table entry for op STASG
3697 foo.c: fatal error in /lib/ccom
3698
3699These are due to bugs in the C compiler; the code is valid C.
3700Unfortunately, the bugs are unpredictable: the same construct
3701may compile properly or trigger one of these bugs, depending
3702on what else is in the source file being compiled. Even changes
3703in header files that should not affect the file being compiled
3704can affect whether the bug happens. In addition, sometimes files
3705that compile correctly on one machine get this bug on another machine.
3706
3707As a result, it is hard for me to make sure this bug will not affect
3708you. I have attempted to find and alter these constructs, but more
3709can always appear. However, I can tell you how to deal with it if it
3710should happen. The bug comes from having an indexed reference to an
3711array of Lisp_Objects, as an argument in a function call:
3712 Lisp_Object *args;
3713 ...
3714 ... foo (5, args[i], ...)...
3715putting the argument into a temporary variable first, as in
3716 Lisp_Object *args;
3717 Lisp_Object tem;
3718 ...
3719 tem = args[i];
3720 ... foo (r, tem, ...)...
3721causes the problem to go away.
3722The `contents' field of a Lisp vector is an array of Lisp_Objects,
3723so you may see the problem happening with indexed references to that.
3724
9dc15871 3725** 68000 C compiler problems
a933dad1
DL
3726
3727Various 68000 compilers have different problems.
3728These are some that have been observed.
3729
9dc15871 3730*** Using value of assignment expression on union type loses.
a933dad1
DL
3731This means that x = y = z; or foo (x = z); does not work
3732if x is of type Lisp_Object.
3733
9dc15871 3734*** "cannot reclaim" error.
a933dad1
DL
3735
3736This means that an expression is too complicated. You get the correct
3737line number in the error message. The code must be rewritten with
3738simpler expressions.
3739
9dc15871 3740*** XCONS, XSTRING, etc macros produce incorrect code.
a933dad1
DL
3741
3742If temacs fails to run at all, this may be the cause.
3743Compile this test program and look at the assembler code:
3744
3745struct foo { char x; unsigned int y : 24; };
3746
3747lose (arg)
3748 struct foo arg;
3749{
3750 test ((int *) arg.y);
3751}
3752
3753If the code is incorrect, your compiler has this problem.
3754In the XCONS, etc., macros in lisp.h you must replace (a).u.val with
3755((a).u.val + coercedummy) where coercedummy is declared as int.
3756
6e5cb96f
DN
3757This problem will only happen if USE_LISP_UNION_TYPE is manually
3758defined in lisp.h.
a933dad1 3759
9dc15871 3760*** C compilers lose on returning unions.
a933dad1
DL
3761
3762I hear that some C compilers cannot handle returning a union type.
3763Most of the functions in GNU Emacs return type Lisp_Object, which is
3764defined as a union on some rare architectures.
3765
6e5cb96f
DN
3766This problem will only happen if USE_LISP_UNION_TYPE is manually
3767defined in lisp.h.
a933dad1 3768
53854552 3769\f
5b87ad55 3770This file is part of GNU Emacs.
fe6b4873 3771
ab73e885 3772GNU Emacs is free software: you can redistribute it and/or modify
5b87ad55 3773it under the terms of the GNU General Public License as published by
ab73e885
GM
3774the Free Software Foundation, either version 3 of the License, or
3775(at your option) any later version.
fe6b4873 3776
5b87ad55
GM
3777GNU Emacs is distributed in the hope that it will be useful,
3778but WITHOUT ANY WARRANTY; without even the implied warranty of
3779MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
3780GNU General Public License for more details.
3781
3782You should have received a copy of the GNU General Public License
ab73e885 3783along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
5b87ad55
GM
3784
3785\f
53854552
EZ
3786Local variables:
3787mode: outline
3256a475 3788paragraph-separate: "[ \f]*$"
53854552 3789end:
6b61353c
KH
3790
3791arch-tag: 49fc0d95-88cb-4715-b21c-f27fb5a4764a