1Copyright (C) 1987, 1988, 1989, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2 2001, 2002, 2003, 2004, 2005, 2006, 2007 3 Free Software Foundation, Inc. 4See the end of the file for license conditions. 5 6 7This file describes various problems that have been encountered 8in compiling, installing and running GNU Emacs. Try doing Ctl-C Ctl-t 9and browsing through the outline headers. 10 11* Emacs startup failures 12 13** Emacs fails to start, complaining about missing fonts. 14 15A typical error message might be something like 16 17 No fonts match `-*-fixed-medium-r-*--6-*-*-*-*-*-iso8859-1' 18 19This happens because some X resource specifies a bad font family for 20Emacs to use. The possible places where this specification might be 21are: 22 23 - in your ~/.Xdefaults file 24 25 - client-side X resource file, such as ~/Emacs or 26 /usr/X11R6/lib/app-defaults/Emacs or 27 /usr/X11R6/lib/X11/app-defaults/Emacs 28 29One of these files might have bad or malformed specification of a 30fontset that Emacs should use. To fix the problem, you need to find 31the problematic line(s) and correct them. 32 33** Emacs aborts while starting up, only when run without X. 34 35This problem often results from compiling Emacs with GCC when GCC was 36installed incorrectly. The usual error in installing GCC is to 37specify --includedir=/usr/include. Installation of GCC makes 38corrected copies of the system header files. GCC is supposed to use 39the corrected copies in preference to the original system headers. 40Specifying --includedir=/usr/include causes the original system header 41files to be used. On some systems, the definition of ioctl in the 42original system header files is invalid for ANSI C and causes Emacs 43not to work. 44 45The fix is to reinstall GCC, and this time do not specify --includedir 46when you configure it. Then recompile Emacs. Specifying --includedir 47is appropriate only in very special cases and it should *never* be the 48same directory where system header files are kept. 49 50** Emacs does not start, complaining that it cannot open termcap database file. 51 52If your system uses Terminfo rather than termcap (most modern 53systems do), this could happen if the proper version of 54ncurses is not visible to the Emacs configure script (i.e. it 55cannot be found along the usual path the linker looks for 56libraries). It can happen because your version of ncurses is 57obsolete, or is available only in form of binaries. 58 59The solution is to install an up-to-date version of ncurses in 60the developer's form (header files, static libraries and 61symbolic links); in some GNU/Linux distributions (e.g. Debian) 62it constitutes a separate package. 63 64** Emacs 20 and later fails to load Lisp files at startup. 65 66The typical error message might be like this: 67 68 "Cannot open load file: fontset" 69 70This could happen if you compress the file lisp/subdirs.el. That file 71tells Emacs what are the directories where it should look for Lisp 72files. Emacs cannot work with subdirs.el compressed, since the 73Auto-compress mode it needs for this will not be loaded until later, 74when your .emacs file is processed. (The package `fontset.el' is 75required to set up fonts used to display text on window systems, and 76it's loaded very early in the startup procedure.) 77 78Similarly, any other .el file for which there's no corresponding .elc 79file could fail to load if it is compressed. 80 81The solution is to uncompress all .el files which don't have a .elc 82file. 83 84Another possible reason for such failures is stale *.elc files 85lurking somewhere on your load-path. The following command will 86print any duplicate Lisp files that are present in load-path: 87 88 emacs -q -batch -f list-load-path-shadows 89 90If this command prints any file names, some of these files are stale, 91and should be deleted or their directories removed from your 92load-path. 93 94** Emacs prints an error at startup after upgrading from an earlier version. 95 96An example of such an error is: 97 98 x-complement-fontset-spec: "Wrong type argument: stringp, nil" 99 100This can be another symptom of stale *.elc files in your load-path. 101The following command will print any duplicate Lisp files that are 102present in load-path: 103 104 emacs -q -batch -f list-load-path-shadows 105 106If this command prints any file names, some of these files are stale, 107and should be deleted or their directories removed from your 108load-path. 109 110** With X11R6.4, public-patch-3, Emacs crashes at startup. 111 112Reportedly this patch in X fixes the problem. 113 114 --- xc/lib/X11/imInt.c~ Wed Jun 30 13:31:56 1999 115 +++ xc/lib/X11/imInt.c Thu Jul 1 15:10:27 1999 116 @@ -1,4 +1,4 @@ 117 -/* $TOG: imInt.c /main/5 1998/05/30 21:11:16 kaleb $ */ 118 +/* $TOG: imInt.c /main/5 1998/05/30 21:11:16 kaleb $ */ 119 /****************************************************************** 120 121 Copyright 1992, 1993, 1994 by FUJITSU LIMITED 122 @@ -166,8 +166,8 @@ 123 _XimMakeImName(lcd) 124 XLCd lcd; 125 { 126 - char* begin; 127 - char* end; 128 + char* begin = NULL; 129 + char* end = NULL; 130 char* ret; 131 int i = 0; 132 char* ximmodifier = XIMMODIFIER; 133 @@ -182,7 +182,11 @@ 134 } 135 ret = Xmalloc(end - begin + 2); 136 if (ret != NULL) { 137 - (void)strncpy(ret, begin, end - begin + 1); 138 + if (begin != NULL) { 139 + (void)strncpy(ret, begin, end - begin + 1); 140 + } else { 141 + ret[0] = '\0'; 142 + } 143 ret[end - begin + 1] = '\0'; 144 } 145 return ret; 146 147** Emacs crashes on startup after a glibc upgrade. 148 149This is caused by a binary incompatible change to the malloc 150implementation in glibc 2.5.90-22. As a result, Emacs binaries built 151using prior versions of glibc crash when run under 2.5.90-22. 152 153This problem was first seen in pre-release versions of Fedora 7, and 154may be fixed in the final Fedora 7 release. To stop the crash from 155happening, first try upgrading to the newest version of glibc; if this 156does not work, rebuild Emacs with the same version of glibc that you 157will run it under. For details, see 158 159https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=239344 160 161* Crash bugs 162 163** Emacs crashes in x-popup-dialog. 164 165This can happen if the dialog widget cannot find the font it wants to 166use. You can work around the problem by specifying another font with 167an X resource--for example, `Emacs.dialog*.font: 9x15' (or any font that 168happens to exist on your X server). 169 170** Emacs crashes when you use Bibtex mode. 171 172This happens if your system puts a small limit on stack size. You can 173prevent the problem by using a suitable shell command (often `ulimit') 174to raise the stack size limit before you run Emacs. 175 176Patches to raise the stack size limit automatically in `main' 177(src/emacs.c) on various systems would be greatly appreciated. 178 179** Error message `Symbol's value as variable is void: x', followed by 180a segmentation fault and core dump. 181 182This has been tracked to a bug in tar! People report that tar erroneously 183added a line like this at the beginning of files of Lisp code: 184 185 x FILENAME, N bytes, B tape blocks 186 187If your tar has this problem, install GNU tar--if you can manage to 188untar it :-). 189 190** Crashes when displaying GIF images in Emacs built with version 191libungif-4.1.0 are resolved by using version libungif-4.1.0b1. 192Configure checks for the correct version, but this problem could occur 193if a binary built against a shared libungif is run on a system with an 194older version. 195 196** Emacs aborts inside the function `tparam1'. 197 198This can happen if Emacs was built without terminfo support, but the 199terminal's capabilities use format that is only supported by terminfo. 200If your system has ncurses installed, this might happen if your 201version of ncurses is broken; upgrading to a newer version of ncurses 202and reconfiguring and rebuilding Emacs should solve this. 203 204All modern systems support terminfo, so even if ncurses is not the 205problem, you should look for a way to configure Emacs so that it uses 206terminfo when built. 207 208** Emacs crashes when using the Exceed 6.0 X server. 209 210If you are using Exceed 6.1, upgrade to a later version. This was 211reported to prevent the crashes. 212 213** Emacs crashes with SIGSEGV in XtInitializeWidgetClass. 214 215It crashes on X, but runs fine when called with option "-nw". 216 217This has been observed when Emacs is linked with GNU ld but without passing 218the -z nocombreloc flag. Emacs normally knows to pass the -z nocombreloc 219flag when needed, so if you come across a situation where the flag is 220necessary but missing, please report it via M-x report-emacs-bug. 221 222On platforms such as Solaris, you can also work around this problem by 223configuring your compiler to use the native linker instead of GNU ld. 224 225** Emacs compiled with Gtk+ crashes when closing a display (x-close-connection). 226 227This happens because of bugs in Gtk+. Gtk+ 2.10 seems to be OK. See bug 228http://bugzilla.gnome.org/show_bug.cgi?id=85715. 229 230** Emacs compiled with Gtk+ crashes on startup on Cygwin. 231 232A typical error message is 233 ***MEMORY-ERROR***: emacs[5172]: GSlice: failed to allocate 504 bytes 234 (alignment: 512): Function not implemented 235 236Emacs supplies its own malloc, but glib (part of Gtk+) calls memalign and on 237Cygwin, that becomes the Cygwin supplied memalign. As malloc is not the 238Cygwin malloc, the Cygwin memalign always returns ENOSYS. A fix for this 239problem would be welcome. 240 241* General runtime problems 242 243** Lisp problems 244 245*** Changes made to .el files do not take effect. 246 247You may have forgotten to recompile them into .elc files. 248Then the old .elc files will be loaded, and your changes 249will not be seen. To fix this, do M-x byte-recompile-directory 250and specify the directory that contains the Lisp files. 251 252Emacs should print a warning when loading a .elc file which is older 253than the corresponding .el file. 254 255*** Watch out for .emacs files and EMACSLOADPATH environment vars. 256 257These control the actions of Emacs. 258~/.emacs is your Emacs init file. 259EMACSLOADPATH overrides which directories the function 260"load" will search. 261 262If you observe strange problems, check for these and get rid 263of them, then try again. 264 265*** Using epop3.el package causes Emacs to signal an error. 266 267The error message might be something like this: 268 269 "Lisp nesting exceeds max-lisp-eval-depth" 270 271This happens because epop3 redefines the function gethash, which is a 272built-in primitive beginning with Emacs 21.1. We don't have a patch 273for epop3 that fixes this, but perhaps a newer version of epop3 274corrects that. 275 276*** Buffers from `with-output-to-temp-buffer' get set up in Help mode. 277 278Changes in Emacs 20.4 to the hooks used by that function cause 279problems for some packages, specifically BBDB. See the function's 280documentation for the hooks involved. BBDB 2.00.06 fixes the problem. 281 282*** The Hyperbole package causes *Help* buffers not to be displayed in 283Help mode due to setting `temp-buffer-show-hook' rather than using 284`add-hook'. Using `(add-hook 'temp-buffer-show-hook 285'help-mode-maybe)' after loading Hyperbole should fix this. 286 287** Keyboard problems 288 289*** "Compose Character" key does strange things when used as a Meta key. 290 291If you define one key to serve as both Meta and Compose Character, you 292will get strange results. In previous Emacs versions, this "worked" 293in that the key acted as Meta--that's because the older Emacs versions 294did not try to support Compose Character. Now Emacs tries to do 295character composition in the standard X way. This means that you 296must pick one meaning or the other for any given key. 297 298You can use both functions (Meta, and Compose Character) if you assign 299them to two different keys. 300 301*** C-z just refreshes the screen instead of suspending Emacs. 302 303You are probably using a shell that doesn't support job control, even 304though the system itself is capable of it. Either use a different shell, 305or set the variable `cannot-suspend' to a non-nil value. 306 307*** With M-x enable-flow-control, you need to type C-\ twice 308to do incremental search--a single C-\ gets no response. 309 310This has been traced to communicating with your machine via kermit, 311with C-\ as the kermit escape character. One solution is to use 312another escape character in kermit. One user did 313 314 set escape-character 17 315 316in his .kermrc file, to make C-q the kermit escape character. 317 318** Mailers and other helper programs 319 320*** movemail compiled with POP support can't connect to the POP server. 321 322Make sure that the `pop' entry in /etc/services, or in the services 323NIS map if your machine uses NIS, has the same port number as the 324entry on the POP server. A common error is for the POP server to be 325listening on port 110, the assigned port for the POP3 protocol, while 326the client is trying to connect on port 109, the assigned port for the 327old POP protocol. 328 329*** RMAIL gets error getting new mail. 330 331RMAIL gets new mail from /usr/spool/mail/$USER using a program 332called `movemail'. This program interlocks with /bin/mail using 333the protocol defined by /bin/mail. 334 335There are two different protocols in general use. One of them uses 336the `flock' system call. The other involves creating a lock file; 337`movemail' must be able to write in /usr/spool/mail in order to do 338this. You control which one is used by defining, or not defining, 339the macro MAIL_USE_FLOCK in config.h or the m- or s- file it includes. 340IF YOU DON'T USE THE FORM OF INTERLOCKING THAT IS NORMAL ON YOUR 341SYSTEM, YOU CAN LOSE MAIL! 342 343If your system uses the lock file protocol, and fascist restrictions 344prevent ordinary users from writing the lock files in /usr/spool/mail, 345you may need to make `movemail' setgid to a suitable group such as 346`mail'. To do this, use the following commands (as root) after doing the 347make install. 348 349 chgrp mail movemail 350 chmod 2755 movemail 351 352Installation normally copies movemail from the build directory to an 353installation directory which is usually under /usr/local/lib. The 354installed copy of movemail is usually in the directory 355/usr/local/lib/emacs/VERSION/TARGET. You must change the group and 356mode of the installed copy; changing the group and mode of the build 357directory copy is ineffective. 358 359*** rcs2log gives you the awk error message "too many fields". 360 361This is due to an arbitrary limit in certain versions of awk. 362The solution is to use gawk (GNU awk). 363 364** Problems with hostname resolution 365 366*** Emacs fails to understand most Internet host names, even though 367the names work properly with other programs on the same system. 368*** Emacs won't work with X-windows if the value of DISPLAY is HOSTNAME:0. 369*** Gnus can't make contact with the specified host for nntp. 370 371This typically happens on Suns and other systems that use shared 372libraries. The cause is that the site has installed a version of the 373shared library which uses a name server--but has not installed a 374similar version of the unshared library which Emacs uses. 375 376The result is that most programs, using the shared library, work with 377the nameserver, but Emacs does not. 378 379The fix is to install an unshared library that corresponds to what you 380installed in the shared library, and then relink Emacs. 381 382On SunOS 4.1, simply define HAVE_RES_INIT. 383 384If you have already installed the name resolver in the file libresolv.a, 385then you need to compile Emacs to use that library. The easiest way to 386do this is to add to config.h a definition of LIBS_SYSTEM, LIBS_MACHINE 387or LIB_STANDARD which uses -lresolv. Watch out! If you redefine a macro 388that is already in use in your configuration to supply some other libraries, 389be careful not to lose the others. 390 391Thus, you could start by adding this to config.h: 392 393#define LIBS_SYSTEM -lresolv 394 395Then if this gives you an error for redefining a macro, and you see that 396the s- file defines LIBS_SYSTEM as -lfoo -lbar, you could change config.h 397again to say this: 398 399#define LIBS_SYSTEM -lresolv -lfoo -lbar 400 401*** Emacs does not know your host's fully-qualified domain name. 402 403For example, (system-name) returns some variation on 404"localhost.localdomain", rather the name you were expecting. 405 406You need to configure your machine with a fully qualified domain name, 407(i.e. a name with at least one ".") either in /etc/hosts, 408/etc/hostname, the NIS, or wherever your system calls for specifying 409this. 410 411If you cannot fix the configuration, you can set the Lisp variable 412mail-host-address to the value you want. 413 414** NFS and RFS 415 416*** Emacs says it has saved a file, but the file does not actually 417appear on disk. 418 419This can happen on certain systems when you are using NFS, if the 420remote disk is full. It is due to a bug in NFS (or certain NFS 421implementations), and there is apparently nothing Emacs can do to 422detect the problem. Emacs checks the failure codes of all the system 423calls involved in writing a file, including `close'; but in the case 424where the problem occurs, none of those system calls fails. 425 426*** Editing files through RFS gives spurious "file has changed" warnings. 427It is possible that a change in Emacs 18.37 gets around this problem, 428but in case not, here is a description of how to fix the RFS bug that 429causes it. 430 431 There was a serious pair of bugs in the handling of the fsync() system 432 call in the RFS server. 433 434 The first is that the fsync() call is handled as another name for the 435 close() system call (!!). It appears that fsync() is not used by very 436 many programs; Emacs version 18 does an fsync() before closing files 437 to make sure that the bits are on the disk. 438 439 This is fixed by the enclosed patch to the RFS server. 440 441 The second, more serious problem, is that fsync() is treated as a 442 non-blocking system call (i.e., it's implemented as a message that 443 gets sent to the remote system without waiting for a reply). Fsync is 444 a useful tool for building atomic file transactions. Implementing it 445 as a non-blocking RPC call (when the local call blocks until the sync 446 is done) is a bad idea; unfortunately, changing it will break the RFS 447 protocol. No fix was supplied for this problem. 448 449 (as always, your line numbers may vary) 450 451 % rcsdiff -c -r1.2 serversyscall.c 452 RCS file: RCS/serversyscall.c,v 453 retrieving revision 1.2 454 diff -c -r1.2 serversyscall.c 455 *** /tmp/,RCSt1003677 Wed Jan 28 15:15:02 1987 456 --- serversyscall.c Wed Jan 28 15:14:48 1987 457 *************** 458 *** 163,169 **** 459 /* 460 * No return sent for close or fsync! 461 */ 462 ! if (syscall == RSYS_close || syscall == RSYS_fsync) 463 proc->p_returnval = deallocate_fd(proc, msg->m_args[0]); 464 else 465 { 466 --- 166,172 ---- 467 /* 468 * No return sent for close or fsync! 469 */ 470 ! if (syscall == RSYS_close) 471 proc->p_returnval = deallocate_fd(proc, msg->m_args[0]); 472 else 473 { 474 475** PSGML 476 477*** Old versions of the PSGML package use the obsolete variables 478`before-change-function' and `after-change-function', which are no 479longer used by Emacs. Please use PSGML 1.2.3 or later. 480 481*** PSGML conflicts with sgml-mode. 482 483PSGML package uses the same names of some variables (like keymap) 484as built-in sgml-mode.el because it was created as a replacement 485of that package. The conflict will be shown if you load 486sgml-mode.el before psgml.el. E.g. this could happen if you edit 487HTML page and then start to work with SGML or XML file. html-mode 488(from sgml-mode.el) is used for HTML file and loading of psgml.el 489(for sgml-mode or xml-mode) will cause an error. 490 491*** Versions of the PSGML package earlier than 1.0.3 (stable) or 1.1.2 492(alpha) fail to parse DTD files correctly in Emacs 20.3 and later. 493Here is a patch for psgml-parse.el from PSGML 1.0.1 and, probably, 494earlier versions. 495 496--- psgml-parse.el 1998/08/21 19:18:18 1.1 497+++ psgml-parse.el 1998/08/21 19:20:00 498@@ -2383,7 +2383,7 @@ (defun sgml-push-to-entity (entity &opti 499 (setq sgml-buffer-parse-state nil)) 500 (cond 501 ((stringp entity) ; a file name 502- (save-excursion (insert-file-contents entity)) 503+ (insert-file-contents entity) 504 (setq default-directory (file-name-directory entity))) 505 ((consp (sgml-entity-text entity)) ; external id? 506 (let* ((extid (sgml-entity-text entity)) 507 508** AUCTeX 509 510You should not be using a version older than 11.52 if you can avoid 511it. 512 513*** Emacs 21 freezes when visiting a TeX file with AUCTeX installed. 514 515Emacs 21 needs version 10 or later of AUCTeX; upgrading should solve 516these problems. 517 518*** No colors in AUCTeX with Emacs 21. 519 520Upgrade to AUC TeX version 10 or later, and make sure it is 521byte-compiled with Emacs 21. 522 523** PCL-CVS 524 525*** Lines are not updated or new lines are added in the buffer upon commit. 526 527When committing files located higher in the hierarchy than the examined 528directory, some versions of the CVS program return an ambiguous message 529from which PCL-CVS cannot extract the full location of the committed 530files. As a result, the corresponding lines in the PCL-CVS buffer are 531not updated with the new revision of these files, and new lines are 532added to the top-level directory. 533 534This can happen with CVS versions 1.12.8 and 1.12.9. Upgrade to CVS 5351.12.10 or newer to fix this problem. 536 537** Miscellaneous problems 538 539*** Self-documentation messages are garbled. 540 541This means that the file `etc/DOC-...' doesn't properly correspond 542with the Emacs executable. Redumping Emacs and then installing the 543corresponding pair of files should fix the problem. 544 545*** Programs running under terminal emulator do not recognize `emacs' 546terminal type. 547 548The cause of this is a shell startup file that sets the TERMCAP 549environment variable. The terminal emulator uses that variable to 550provide the information on the special terminal type that Emacs 551emulates. 552 553Rewrite your shell startup file so that it does not change TERMCAP 554in such a case. You could use the following conditional which sets 555it only if it is undefined. 556 557 if ( ! ${?TERMCAP} ) setenv TERMCAP ~/my-termcap-file 558 559Or you could set TERMCAP only when you set TERM--which should not 560happen in a non-login shell. 561 562*** In Shell mode, you get a ^M at the end of every line. 563 564This happens to people who use tcsh, because it is trying to be too 565smart. It sees that the Shell uses terminal type `unknown' and turns 566on the flag to output ^M at the end of each line. You can fix the 567problem by adding this to your .cshrc file: 568 569 if ($?EMACS) then 570 if ("$EMACS" =~ /*) then 571 unset edit 572 stty -icrnl -onlcr -echo susp ^Z 573 endif 574 endif 575 576*** Emacs startup on GNU/Linux systems (and possibly other systems) is slow. 577 578This can happen if the system is misconfigured and Emacs can't get the 579full qualified domain name, FQDN. You should have your FQDN in the 580/etc/hosts file, something like this: 581 582127.0.0.1 localhost 583129.187.137.82 nuc04.t30.physik.tu-muenchen.de nuc04 584 585The way to set this up may vary on non-GNU systems. 586 587*** Attempting to visit remote files via ange-ftp fails. 588 589If the error message is "ange-ftp-file-modtime: Specified time is not 590representable", then this could happen when `lukemftp' is used as the 591ftp client. This was reported to happen on Debian GNU/Linux, kernel 592version 2.4.3, with `lukemftp' 1.5-5, but might happen on other 593systems as well. To avoid this problem, switch to using the standard 594ftp client. On a Debian system, type 595 596 update-alternatives --config ftp 597 598and then choose /usr/bin/netkit-ftp. 599 600*** JPEG images aren't displayed. 601 602This has been reported when Emacs is built with jpeg-6a library. 603Upgrading to jpeg-6b solves the problem. Configure checks for the 604correct version, but this problem could occur if a binary built 605against a shared libjpeg is run on a system with an older version. 606 607*** Dired is very slow. 608 609This could happen if invocation of the `df' program takes a long 610time. Possible reasons for this include: 611 612 - ClearCase mounted filesystems (VOBs) that sometimes make `df' 613 response time extremely slow (dozens of seconds); 614 615 - slow automounters on some old versions of Unix; 616 617 - slow operation of some versions of `df'. 618 619To work around the problem, you could either (a) set the variable 620`directory-free-space-program' to nil, and thus prevent Emacs from 621invoking `df'; (b) use `df' from the GNU Fileutils package; or 622(c) use CVS, which is Free Software, instead of ClearCase. 623 624*** Versions of the W3 package released before Emacs 21.1 don't run 625under Emacs 21. This fixed in W3 version 4.0pre.47. 626 627*** The LDAP support rely on ldapsearch program from OpenLDAP version 2. 628 629It can fail to work with ldapsearch program from OpenLDAP version 1. 630Version 1 of OpenLDAP is now deprecated. If you are still using it, 631please upgrade to version 2. As a temporary workaround, remove 632argument "-x" from the variable `ldap-ldapsearch-args'. 633 634*** ps-print commands fail to find prologue files ps-prin*.ps. 635 636This can happen if you use an old version of X-Symbol package: it 637defines compatibility functions which trick ps-print into thinking it 638runs in XEmacs, and look for the prologue files in a wrong directory. 639 640The solution is to upgrade X-Symbol to a later version. 641 642*** On systems with shared libraries you might encounter run-time errors 643from the dynamic linker telling you that it is unable to find some 644shared libraries, for instance those for Xaw3d or image support. 645These errors mean Emacs has been linked with a library whose shared 646library is not in the default search path of the dynamic linker. 647 648Similar problems could prevent Emacs from building, since the build 649process invokes Emacs several times. 650 651On many systems, it is possible to set LD_LIBRARY_PATH in your 652environment to specify additional directories where shared libraries 653can be found. 654 655Other systems allow to set LD_RUN_PATH in a similar way, but before 656Emacs is linked. With LD_RUN_PATH set, the linker will include a 657specified run-time search path in the executable. 658 659On some systems, Emacs can crash due to problems with dynamic 660linking. Specifically, on SGI Irix 6.5, crashes were reported with 661backtraces like this: 662 663 (dbx) where 664 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] 665 1 general_find_symbol(0xf49239d, 0x0, 0x0, 0x0, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2) 666 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":2140, 0xfb65a98] 667 2 resolve_symbol(0xf49239d, 0x4031184, 0x0, 0xfbdd438, 0x0, 0xf4923aa, 0x0, 0x492ddb2) 668 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":1947, 0xfb657e4] 669 3 lazy_text_resolve(0xd18, 0x1a3, 0x40302b4, 0x12, 0xf0000000, 0xf4923aa, 0x0, 0x492ddb2) 670 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld.c":997, 0xfb64d44] 671 4 _rld_text_resolve(0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0) 672 ["/comp2/mtibuild/v73/workarea/v7.3/rld/rld_bridge.s":175, 0xfb6032c] 673 674(`rld' is the dynamic linker.) We don't know yet why this 675happens, but setting the environment variable LD_BIND_NOW to 1 (which 676forces the dynamic linker to bind all shared objects early on) seems 677to work around the problem. 678 679Please refer to the documentation of your dynamic linker for details. 680 681*** You request inverse video, and the first Emacs frame is in inverse 682video, but later frames are not in inverse video. 683 684This can happen if you have an old version of the custom library in 685your search path for Lisp packages. Use M-x list-load-path-shadows to 686check whether this is true. If it is, delete the old custom library. 687 688*** When you run Ispell from Emacs, it reports a "misalignment" error. 689 690This can happen if you compiled the Ispell program to use ASCII 691characters only and then try to use it from Emacs with non-ASCII 692characters, like Latin-1. The solution is to recompile Ispell with 693support for 8-bit characters. 694 695To see whether your Ispell program supports 8-bit characters, type 696this at your shell's prompt: 697 698 ispell -vv 699 700and look in the output for the string "NO8BIT". If Ispell says 701"!NO8BIT (8BIT)", your speller supports 8-bit characters; otherwise it 702does not. 703 704To rebuild Ispell with 8-bit character support, edit the local.h file 705in the Ispell distribution and make sure it does _not_ define NO8BIT. 706Then rebuild the speller. 707 708Another possible cause for "misalignment" error messages is that the 709version of Ispell installed on your machine is old. Upgrade. 710 711Yet another possibility is that you are trying to spell-check a word 712in a language that doesn't fit the dictionary you choose for use by 713Ispell. (Ispell can only spell-check one language at a time, because 714it uses a single dictionary.) Make sure that the text you are 715spelling and the dictionary used by Ispell conform to each other. 716 717If your spell-checking program is Aspell, it has been reported that if 718you have a personal configuration file (normally ~/.aspell.conf), it 719can cause this error. Remove that file, execute `ispell-kill-ispell' 720in Emacs, and then try spell-checking again. 721 722* Runtime problems related to font handling 723 724** Under X11, some characters appear as hollow boxes. 725 726Each X11 font covers just a fraction of the characters that Emacs 727supports. To display the whole range of Emacs characters requires 728many different fonts, collected into a fontset. 729 730If some of the fonts called for in your fontset do not exist on your X 731server, then the characters that have no font appear as hollow boxes. 732You can remedy the problem by installing additional fonts. 733 734The intlfonts distribution includes a full spectrum of fonts that can 735display all the characters Emacs supports. The etl-unicode collection 736of fonts (available from <URL:ftp://ftp.x.org/contrib/fonts/> and 737<URL:ftp://ftp.xfree86.org/pub/mirror/X.Org/contrib/fonts/>) includes 738fonts that can display many Unicode characters; they can also be used 739by ps-print and ps-mule to print Unicode characters. 740 741Another cause of this for specific characters is fonts which have a 742missing glyph and no default character. This is known to occur for 743character number 160 (no-break space) in some fonts, such as Lucida 744but Emacs sets the display table for the unibyte and Latin-1 version 745of this character to display a space. 746 747** Under X11, some characters appear improperly aligned in their lines. 748 749You may have bad X11 fonts; try installing the intlfonts distribution 750or the etl-unicode collection (see the previous entry). 751 752** Certain fonts make each line take one pixel more than it "should". 753 754This is because these fonts contain characters a little taller 755than the font's nominal height. Emacs needs to make sure that 756lines do not overlap. 757 758** Loading fonts is very slow. 759 760You might be getting scalable fonts instead of precomputed bitmaps. 761Known scalable font directories are "Type1" and "Speedo". A font 762directory contains scalable fonts if it contains the file 763"fonts.scale". 764 765If this is so, re-order your X windows font path to put the scalable 766font directories last. See the documentation of `xset' for details. 767 768With some X servers, it may be necessary to take the scalable font 769directories out of your path entirely, at least for Emacs 19.26. 770Changes in the future may make this unnecessary. 771 772** Font Lock displays portions of the buffer in incorrect faces. 773 774By far the most frequent cause of this is a parenthesis `(' or a brace 775`{' in column zero. Font Lock assumes that such a paren is outside of 776any comment or string. This is of course not true in general, but the 777vast majority of well-formatted program source files don't have such 778parens, and therefore this assumption is used to allow optimizations 779in Font Lock's syntactical analysis. These optimizations avoid some 780pathological cases where jit-lock, the Just-in-Time fontification 781introduced with Emacs 21.1, could significantly slow down scrolling 782through the buffer, especially scrolling backwards, and also jumping 783to the end of a very large buffer. 784 785Beginning with version 22.1, a parenthesis or a brace in column zero 786is highlighted in bold-red face if it is inside a string or a comment, 787to indicate that it could interfere with Font Lock (and also with 788indentation) and should be moved or escaped with a backslash. 789 790If you don't use large buffers, or have a very fast machine which 791makes the delays insignificant, you can avoid the incorrect 792fontification by setting the variable 793`font-lock-beginning-of-syntax-function' to a nil value. (This must 794be done _after_ turning on Font Lock.) 795 796Another alternative is to avoid a paren in column zero. For example, 797in a Lisp string you could precede the paren with a backslash. 798 799** With certain fonts, when the cursor appears on a character, the 800character doesn't appear--you get a solid box instead. 801 802One user on a Linux-based GNU system reported that this problem went 803away with installation of a new X server. The failing server was 804XFree86 3.1.1. XFree86 3.1.2 works. 805 806** Characters are displayed as empty boxes or with wrong font under X. 807 808This can occur when two different versions of FontConfig are used. 809For example, XFree86 4.3.0 has one version and Gnome usually comes 810with a newer version. Emacs compiled with --with-gtk will then use 811the newer version. In most cases the problem can be temporarily 812fixed by stopping the application that has the error (it can be 813Emacs or any other application), removing ~/.fonts.cache-1, 814and then start the application again. 815If removing ~/.fonts.cache-1 and restarting doesn't help, the 816application with problem must be recompiled with the same version 817of FontConfig as the rest of the system uses. For KDE, it is 818sufficient to recompile Qt. 819 820** Emacs pauses for several seconds when changing the default font. 821 822This has been reported for fvwm 2.2.5 and the window manager of KDE 8232.1. The reason for the pause is Xt waiting for a ConfigureNotify 824event from the window manager, which the window manager doesn't send. 825Xt stops waiting after a default timeout of usually 5 seconds. 826 827A workaround for this is to add something like 828 829emacs.waitForWM: false 830 831to your X resources. Alternatively, add `(wait-for-wm . nil)' to a 832frame's parameter list, like this: 833 834 (modify-frame-parameters nil '((wait-for-wm . nil))) 835 836(this should go into your `.emacs' file). 837 838** Underlines appear at the wrong position. 839 840This is caused by fonts having a wrong UNDERLINE_POSITION property. 841Examples are the font 7x13 on XFree prior to version 4.1, or the jmk 842neep font from the Debian xfonts-jmk package. To circumvent this 843problem, set x-use-underline-position-properties to nil in your 844`.emacs'. 845 846To see what is the value of UNDERLINE_POSITION defined by the font, 847type `xlsfonts -lll FONT' and look at the font's UNDERLINE_POSITION 848property. 849 850** When using Exceed, fonts sometimes appear too tall. 851 852When the display is set to an Exceed X-server and fonts are specified 853(either explicitly with the -fn option or implicitly with X resources) 854then the fonts may appear "too tall". The actual character sizes are 855correct but there is too much vertical spacing between rows, which 856gives the appearance of "double spacing". 857 858To prevent this, turn off the Exceed's "automatic font substitution" 859feature (in the font part of the configuration window). 860 861* Internationalization problems 862 863** M-{ does not work on a Spanish PC keyboard. 864 865Many Spanish keyboards seem to ignore that combination. Emacs can't 866do anything about it. 867 868** Characters from the mule-unicode charsets aren't displayed under X. 869 870XFree86 4 contains many fonts in iso10646-1 encoding which have 871minimal character repertoires (whereas the encoding part of the font 872name is meant to be a reasonable indication of the repertoire 873according to the XLFD spec). Emacs may choose one of these to display 874characters from the mule-unicode charsets and then typically won't be 875able to find the glyphs to display many characters. (Check with C-u 876C-x = .) To avoid this, you may need to use a fontset which sets the 877font for the mule-unicode sets explicitly. E.g. to use GNU unifont, 878include in the fontset spec: 879 880mule-unicode-2500-33ff:-gnu-unifont-*-iso10646-1,\ 881mule-unicode-e000-ffff:-gnu-unifont-*-iso10646-1,\ 882mule-unicode-0100-24ff:-gnu-unifont-*-iso10646-1 883 884** The UTF-8/16/7 coding systems don't encode CJK (Far Eastern) characters. 885 886Emacs directly supports the Unicode BMP whose code points are in the 887ranges 0000-33ff and e000-ffff, and indirectly supports the parts of 888CJK characters belonging to these legacy charsets: 889 890 GB2312, Big5, JISX0208, JISX0212, JISX0213-1, JISX0213-2, KSC5601 891 892The latter support is done in Utf-Translate-Cjk mode (turned on by 893default). Which Unicode CJK characters are decoded into which Emacs 894charset is decided by the current language environment. For instance, 895in Chinese-GB, most of them are decoded into chinese-gb2312. 896 897If you read UTF-8 data with code points outside these ranges, the 898characters appear in the buffer as raw bytes of the original UTF-8 899(composed into a single quasi-character) and they will be written back 900correctly as UTF-8, assuming you don't break the composed sequences. 901If you read such characters from UTF-16 or UTF-7 data, they are 902substituted with the Unicode `replacement character', and you lose 903information. 904 905** Mule-UCS loads very slowly. 906 907Changes to Emacs internals interact badly with Mule-UCS's `un-define' 908library, which is the usual interface to Mule-UCS. Apply the 909following patch to Mule-UCS 0.84 and rebuild it. That will help, 910though loading will still be slower than in Emacs 20. (Some 911distributions, such as Debian, may already have applied such a patch.) 912 913--- lisp/un-define.el 6 Mar 2001 22:41:38 -0000 1.30 914+++ lisp/un-define.el 19 Apr 2002 18:34:26 -0000 915@@ -610,13 +624,21 @@ by calling post-read-conversion and pre- 916 917 (mapcar 918 (lambda (x) 919- (mapcar 920- (lambda (y) 921- (mucs-define-coding-system 922- (nth 0 y) (nth 1 y) (nth 2 y) 923- (nth 3 y) (nth 4 y) (nth 5 y) (nth 6 y)) 924- (coding-system-put (car y) 'alias-coding-systems (list (car x)))) 925- (cdr x))) 926+ (if (fboundp 'register-char-codings) 927+ ;; Mule 5, where we don't need the eol-type specified and 928+ ;; register-char-codings may be very slow for these coding 929+ ;; system definitions. 930+ (let ((y (cadr x))) 931+ (mucs-define-coding-system 932+ (car x) (nth 1 y) (nth 2 y) 933+ (nth 3 y) (nth 4 y) (nth 5 y))) 934+ (mapcar 935+ (lambda (y) 936+ (mucs-define-coding-system 937+ (nth 0 y) (nth 1 y) (nth 2 y) 938+ (nth 3 y) (nth 4 y) (nth 5 y) (nth 6 y)) 939+ (coding-system-put (car y) 'alias-coding-systems (list (car x))))) 940+ (cdr x))) 941 `((utf-8 942 (utf-8-unix 943 ?u "UTF-8 coding system" 944 945Note that Emacs has native support for Unicode, roughly equivalent to 946Mule-UCS's, so you may not need it. 947 948** Mule-UCS compilation problem. 949 950Emacs of old versions and XEmacs byte-compile the form `(progn progn 951...)' the same way as `(progn ...)', but Emacs of version 21.3 and the 952later process that form just as interpreter does, that is, as `progn' 953variable reference. Apply the following patch to Mule-UCS 0.84 to 954make it compiled by the latest Emacs. 955 956--- mucs-ccl.el 2 Sep 2005 00:42:23 -0000 1.1.1.1 957+++ mucs-ccl.el 2 Sep 2005 01:31:51 -0000 1.3 958@@ -639,10 +639,14 @@ 959 (mucs-notify-embedment 'mucs-ccl-required name) 960 (setq ccl-pgm-list (cdr ccl-pgm-list))) 961 ; (message "MCCLREGFIN:%S" result) 962- `(progn 963- (setq mucs-ccl-facility-alist 964- (quote ,mucs-ccl-facility-alist)) 965- ,@result))) 966+ ;; The only way the function is used in this package is included 967+ ;; in `mucs-package-definition-end-hook' value, where it must 968+ ;; return (possibly empty) *list* of forms. Do this. Do not rely 969+ ;; on byte compiler to remove extra `progn's in `(progn ...)' 970+ ;; form. 971+ `((setq mucs-ccl-facility-alist 972+ (quote ,mucs-ccl-facility-alist)) 973+ ,@result))) 974 975 ;;; Add hook for embedding translation informations to a package. 976 (add-hook 'mucs-package-definition-end-hook 977 978** Accented ISO-8859-1 characters are displayed as | or _. 979 980Try other font set sizes (S-mouse-1). If the problem persists with 981other sizes as well, your text is corrupted, probably through software 982that is not 8-bit clean. If the problem goes away with another font 983size, it's probably because some fonts pretend to be ISO-8859-1 fonts 984when they are really ASCII fonts. In particular the schumacher-clean 985fonts have this bug in some versions of X. 986 987To see what glyphs are included in a font, use `xfd', like this: 988 989 xfd -fn -schumacher-clean-medium-r-normal--12-120-75-75-c-60-iso8859-1 990 991If this shows only ASCII glyphs, the font is indeed the source of the 992problem. 993 994The solution is to remove the corresponding lines from the appropriate 995`fonts.alias' file, then run `mkfontdir' in that directory, and then run 996`xset fp rehash'. 997 998** The `oc-unicode' package doesn't work with Emacs 21. 999 1000This package tries to define more private charsets than there are free 1001slots now. The current built-in Unicode support is actually more 1002flexible. (Use option `utf-translate-cjk-mode' if you need CJK 1003support.) Files encoded as emacs-mule using oc-unicode aren't 1004generally read correctly by Emacs 21. 1005 1006** After a while, Emacs slips into unibyte mode. 1007 1008The VM mail package, which is not part of Emacs, sometimes does 1009 (standard-display-european t) 1010That should be changed to 1011 (standard-display-european 1 t) 1012 1013* X runtime problems 1014 1015** X keyboard problems 1016 1017*** You "lose characters" after typing Compose Character key. 1018 1019This is because the Compose Character key is defined as the keysym 1020Multi_key, and Emacs (seeing that) does the proper X11 1021character-composition processing. If you don't want your Compose key 1022to do that, you can redefine it with xmodmap. 1023 1024For example, here's one way to turn it into a Meta key: 1025 1026 xmodmap -e "keysym Multi_key = Meta_L" 1027 1028If all users at your site of a particular keyboard prefer Meta to 1029Compose, you can make the remapping happen automatically by adding the 1030xmodmap command to the xdm setup script for that display. 1031 1032*** Using X Windows, control-shift-leftbutton makes Emacs hang. 1033 1034Use the shell command `xset bc' to make the old X Menu package work. 1035 1036*** C-SPC fails to work on Fedora GNU/Linux (or with fcitx input method). 1037 1038Fedora Core 4 steals the C-SPC key by default for the `iiimx' program 1039which is the input method for some languages. It blocks Emacs users 1040from using the C-SPC key for `set-mark-command'. 1041 1042One solutions is to remove the `<Ctrl>space' from the `Iiimx' file 1043which can be found in the `/usr/lib/X11/app-defaults' directory. 1044However, that requires root access. 1045 1046Another is to specify `Emacs*useXIM: false' in your X resources. 1047 1048Another is to build Emacs with the `--without-xim' configure option. 1049 1050The same problem happens on any other system if you are using fcitx 1051(Chinese input method) which by default use C-SPC for toggling. If 1052you want to use fcitx with Emacs, you have two choices. Toggle fcitx 1053by another key (e.g. C-\) by modifying ~/.fcitx/config, or be 1054accustomed to use C-@ for `set-mark-command'. 1055 1056*** M-SPC seems to be ignored as input. 1057 1058See if your X server is set up to use this as a command 1059for character composition. 1060 1061*** The S-C-t key combination doesn't get passed to Emacs on X. 1062 1063This happens because some X configurations assign the Ctrl-Shift-t 1064combination the same meaning as the Multi_key. The offending 1065definition is in the file `...lib/X11/locale/iso8859-1/Compose'; there 1066might be other similar combinations which are grabbed by X for similar 1067purposes. 1068 1069We think that this can be countermanded with the `xmodmap' utility, if 1070you want to be able to bind one of these key sequences within Emacs. 1071 1072*** Under X, C-v and/or other keys don't work. 1073 1074These may have been intercepted by your window manager. In 1075particular, AfterStep 1.6 is reported to steal C-v in its default 1076configuration. Various Meta keys are also likely to be taken by the 1077configuration of the `feel'. See the WM's documentation for how to 1078change this. 1079 1080*** Clicking C-mouse-2 in the scroll bar doesn't split the window. 1081 1082This currently doesn't work with scroll-bar widgets (and we don't know 1083a good way of implementing it with widgets). If Emacs is configured 1084--without-toolkit-scroll-bars, C-mouse-2 on the scroll bar does work. 1085 1086*** Inability to send an Alt-modified key, when Emacs is communicating 1087directly with an X server. 1088 1089If you have tried to bind an Alt-modified key as a command, and it 1090does not work to type the command, the first thing you should check is 1091whether the key is getting through to Emacs. To do this, type C-h c 1092followed by the Alt-modified key. C-h c should say what kind of event 1093it read. If it says it read an Alt-modified key, then make sure you 1094have made the key binding correctly. 1095 1096If C-h c reports an event that doesn't have the Alt modifier, it may 1097be because your X server has no key for the Alt modifier. The X 1098server that comes from MIT does not set up the Alt modifier by 1099default. 1100 1101If your keyboard has keys named Alt, you can enable them as follows: 1102 1103 xmodmap -e 'add mod2 = Alt_L' 1104 xmodmap -e 'add mod2 = Alt_R' 1105 1106If the keyboard has just one key named Alt, then only one of those 1107commands is needed. The modifier `mod2' is a reasonable choice if you 1108are using an unmodified MIT version of X. Otherwise, choose any 1109modifier bit not otherwise used. 1110 1111If your keyboard does not have keys named Alt, you can use some other 1112keys. Use the keysym command in xmodmap to turn a function key (or 1113some other 'spare' key) into Alt_L or into Alt_R, and then use the 1114commands show above to make them modifier keys. 1115 1116Note that if you have Alt keys but no Meta keys, Emacs translates Alt 1117into Meta. This is because of the great importance of Meta in Emacs. 1118 1119** Window-manager and toolkit-related problems 1120 1121*** Gnome: Emacs receives input directly from the keyboard, bypassing XIM. 1122 1123This seems to happen when gnome-settings-daemon version 2.12 or later 1124is running. If gnome-settings-daemon is not running, Emacs receives 1125input through XIM without any problem. Furthermore, this seems only 1126to happen in *.UTF-8 locales; zh_CN.GB2312 and zh_CN.GBK locales, for 1127example, work fine. A bug report has been filed in the Gnome 1128bugzilla: http://bugzilla.gnome.org/show_bug.cgi?id=357032 1129 1130*** Gnome: Emacs' xterm-mouse-mode doesn't work on the Gnome terminal. 1131 1132A symptom of this bug is that double-clicks insert a control sequence 1133into the buffer. The reason this happens is an apparent 1134incompatibility of the Gnome terminal with Xterm, which also affects 1135other programs using the Xterm mouse interface. A problem report has 1136been filed. 1137 1138*** KDE: When running on KDE, colors or fonts are not as specified for Emacs, 1139or messed up. 1140 1141For example, you could see background you set for Emacs only in the 1142empty portions of the Emacs display, while characters have some other 1143background. 1144 1145This happens because KDE's defaults apply its color and font 1146definitions even to applications that weren't compiled for KDE. The 1147solution is to uncheck the "Apply fonts and colors to non-KDE apps" 1148option in Preferences->Look&Feel->Style (KDE 2). In KDE 3, this option 1149is in the "Colors" section, rather than "Style". 1150 1151Alternatively, if you do want the KDE defaults to apply to other 1152applications, but not to Emacs, you could modify the file `Emacs.ad' 1153(should be in the `/usr/share/apps/kdisplay/app-defaults/' directory) 1154so that it doesn't set the default background and foreground only for 1155Emacs. For example, make sure the following resources are either not 1156present or commented out: 1157 1158 Emacs.default.attributeForeground 1159 Emacs.default.attributeBackground 1160 Emacs*Foreground 1161 Emacs*Background 1162 1163*** KDE: Emacs hangs on KDE when a large portion of text is killed. 1164 1165This is caused by a bug in the KDE applet `klipper' which periodically 1166requests the X clipboard contents from applications. Early versions 1167of klipper don't implement the ICCCM protocol for large selections, 1168which leads to Emacs being flooded with selection requests. After a 1169while, Emacs may print a message: 1170 1171 Timed out waiting for property-notify event 1172 1173A workaround is to not use `klipper'. An upgrade to the `klipper' that 1174comes with KDE 3.3 or later also solves the problem. 1175 1176*** CDE: Frames may cover dialogs they created when using CDE. 1177 1178This can happen if you have "Allow Primary Windows On Top" enabled which 1179seems to be the default in the Common Desktop Environment. 1180To change, go in to "Desktop Controls" -> "Window Style Manager" 1181and uncheck "Allow Primary Windows On Top". 1182 1183*** Xaw3d : When using Xaw3d scroll bars without arrows, the very first mouse 1184click in a scroll bar might be ignored by the scroll bar widget. This 1185is probably a bug in Xaw3d; when Xaw3d is compiled with arrows, the 1186problem disappears. 1187 1188*** Xaw: There are known binary incompatibilities between Xaw, Xaw3d, neXtaw, 1189XawM and the few other derivatives of Xaw. So when you compile with 1190one of these, it may not work to dynamically link with another one. 1191For example, strange problems, such as Emacs exiting when you type 1192"C-x 1", were reported when Emacs compiled with Xaw3d and libXaw was 1193used with neXtaw at run time. 1194 1195The solution is to rebuild Emacs with the toolkit version you actually 1196want to use, or set LD_PRELOAD to preload the same toolkit version you 1197built Emacs with. 1198 1199*** Open Motif: Problems with file dialogs in Emacs built with Open Motif. 1200 1201When Emacs 21 is built with Open Motif 2.1, it can happen that the 1202graphical file dialog boxes do not work properly. The "OK", "Filter" 1203and "Cancel" buttons do not respond to mouse clicks. Dragging the 1204file dialog window usually causes the buttons to work again. 1205 1206The solution is to use LessTif instead. LessTif is a free replacement 1207for Motif. See the file INSTALL for information on how to do this. 1208 1209Another workaround is not to use the mouse to trigger file prompts, 1210but to use the keyboard. This way, you will be prompted for a file in 1211the minibuffer instead of a graphical file dialog. 1212 1213*** LessTif: Problems in Emacs built with LessTif. 1214 1215The problems seem to depend on the version of LessTif and the Motif 1216emulation for which it is set up. 1217 1218Only the Motif 1.2 emulation seems to be stable enough in LessTif. 1219LessTif 0.92-17's Motif 1.2 emulation seems to work okay on FreeBSD. 1220On GNU/Linux systems, lesstif-0.92.6 configured with "./configure 1221--enable-build-12 --enable-default-12" is reported to be the most 1222successful. The binary GNU/Linux package 1223lesstif-devel-0.92.0-1.i386.rpm was reported to have problems with 1224menu placement. 1225 1226On some systems, even with Motif 1.2 emulation, Emacs occasionally 1227locks up, grabbing all mouse and keyboard events. We still don't know 1228what causes these problems; they are not reproducible by Emacs 1229developers. 1230 1231*** Motif: The Motif version of Emacs paints the screen a solid color. 1232 1233This has been observed to result from the following X resource: 1234 1235 Emacs*default.attributeFont: -*-courier-medium-r-*-*-*-140-*-*-*-*-iso8859-* 1236 1237That the resource has this effect indicates a bug in something, but we 1238do not yet know what. If it is an Emacs bug, we hope someone can 1239explain what the bug is so we can fix it. In the mean time, removing 1240the resource prevents the problem. 1241 1242** General X problems 1243 1244*** Redisplay using X11 is much slower than previous Emacs versions. 1245 1246We've noticed that certain X servers draw the text much slower when 1247scroll bars are on the left. We don't know why this happens. If this 1248happens to you, you can work around it by putting the scroll bars 1249on the right (as they were in Emacs 19). 1250 1251Here's how to do this: 1252 1253 (set-scroll-bar-mode 'right) 1254 1255If you're not sure whether (or how much) this problem affects you, 1256try that and see how much difference it makes. To set things back 1257to normal, do 1258 1259 (set-scroll-bar-mode 'left) 1260 1261*** Error messages about undefined colors on X. 1262 1263The messages might say something like this: 1264 1265 Unable to load color "grey95" 1266 1267(typically, in the `*Messages*' buffer), or something like this: 1268 1269 Error while displaying tooltip: (error Undefined color lightyellow) 1270 1271These problems could happen if some other X program has used up too 1272many colors of the X palette, leaving Emacs with insufficient system 1273resources to load all the colors it needs. 1274 1275A solution is to exit the offending X programs before starting Emacs. 1276 1277"undefined color" messages can also occur if the RgbPath entry in the 1278X configuration file is incorrect, or the rgb.txt file is not where 1279X expects to find it. 1280 1281*** Improving performance with slow X connections. 1282 1283There are several ways to improve this performance, any subset of which can 1284be carried out at the same time: 1285 12861) If you don't need X Input Methods (XIM) for entering text in some 1287 language you use, you can improve performance on WAN links by using 1288 the X resource useXIM to turn off use of XIM. This does not affect 1289 the use of Emacs' own input methods, which are part of the Leim 1290 package. 1291 12922) If the connection is very slow, you might also want to consider 1293 switching off scroll bars, menu bar, and tool bar. Adding the 1294 following forms to your .emacs file will accomplish that, but only 1295 after the the initial frame is displayed: 1296 1297 (scroll-bar-mode -1) 1298 (menu-bar-mode -1) 1299 (tool-bar-mode -1) 1300 1301 For still quicker startup, put these X resources in your .Xdefaults 1302 file: 1303 1304 Emacs.verticalScrollBars: off 1305 Emacs.menuBar: off 1306 Emacs.toolBar: off 1307 13083) Use ssh to forward the X connection, and enable compression on this 1309 forwarded X connection (ssh -XC remotehostname emacs ...). 1310 13114) Use lbxproxy on the remote end of the connection. This is an interface 1312 to the low bandwidth X extension in most modern X servers, which 1313 improves performance dramatically, at the slight expense of correctness 1314 of the X protocol. lbxproxy acheives the performance gain by grouping 1315 several X requests in one TCP packet and sending them off together, 1316 instead of requiring a round-trip for each X request in a separate 1317 packet. The switches that seem to work best for emacs are: 1318 -noatomsfile -nowinattr -cheaterrors -cheatevents 1319 Note that the -nograbcmap option is known to cause problems. 1320 For more about lbxproxy, see: 1321 http://www.xfree86.org/4.3.0/lbxproxy.1.html 1322 13235) If copying and killing is slow, try to disable the interaction with the 1324 native system's clipboard by adding these lines to your .emacs file: 1325 (setq interprogram-cut-function nil) 1326 (setq interprogram-paste-function nil) 1327 1328*** Emacs gives the error, Couldn't find per display information. 1329 1330This can result if the X server runs out of memory because Emacs uses 1331a large number of fonts. On systems where this happens, C-h h is 1332likely to cause it. 1333 1334We do not know of a way to prevent the problem. 1335 1336*** Emacs does not notice when you release the mouse. 1337 1338There are reports that this happened with (some) Microsoft mice and 1339that replacing the mouse made it stop. 1340 1341*** You can't select from submenus (in the X toolkit version). 1342 1343On certain systems, mouse-tracking and selection in top-level menus 1344works properly with the X toolkit, but neither of them works when you 1345bring up a submenu (such as Bookmarks or Compare or Apply Patch, in 1346the Files menu). 1347 1348This works on most systems. There is speculation that the failure is 1349due to bugs in old versions of X toolkit libraries, but no one really 1350knows. If someone debugs this and finds the precise cause, perhaps a 1351workaround can be found. 1352 1353*** An error message such as `X protocol error: BadMatch (invalid 1354parameter attributes) on protocol request 93'. 1355 1356This comes from having an invalid X resource, such as 1357 emacs*Cursor: black 1358(which is invalid because it specifies a color name for something 1359that isn't a color.) 1360 1361The fix is to correct your X resources. 1362 1363*** Slow startup on X11R6 with X windows. 1364 1365If Emacs takes two minutes to start up on X11R6, see if your X 1366resources specify any Adobe fonts. That causes the type-1 font 1367renderer to start up, even if the font you asked for is not a type-1 1368font. 1369 1370One way to avoid this problem is to eliminate the type-1 fonts from 1371your font path, like this: 1372 1373 xset -fp /usr/X11R6/lib/X11/fonts/Type1/ 1374 1375*** Pull-down menus appear in the wrong place, in the toolkit version of Emacs. 1376 1377An X resource of this form can cause the problem: 1378 1379 Emacs*geometry: 80x55+0+0 1380 1381This resource is supposed to apply, and does apply, to the menus 1382individually as well as to Emacs frames. If that is not what you 1383want, rewrite the resource. 1384 1385To check thoroughly for such resource specifications, use `xrdb 1386-query' to see what resources the X server records, and also look at 1387the user's ~/.Xdefaults and ~/.Xdefaults-* files. 1388 1389*** Emacs running under X Windows does not handle mouse clicks. 1390*** `emacs -geometry 80x20' finds a file named `80x20'. 1391 1392One cause of such problems is having (setq term-file-prefix nil) in 1393your .emacs file. Another cause is a bad value of EMACSLOADPATH in 1394the environment. 1395 1396*** Emacs fails to get default settings from X Windows server. 1397 1398The X library in X11R4 has a bug; it interchanges the 2nd and 3rd 1399arguments to XGetDefaults. Define the macro XBACKWARDS in config.h to 1400tell Emacs to compensate for this. 1401 1402I don't believe there is any way Emacs can determine for itself 1403whether this problem is present on a given system. 1404 1405*** X Windows doesn't work if DISPLAY uses a hostname. 1406 1407People have reported kernel bugs in certain systems that cause Emacs 1408not to work with X Windows if DISPLAY is set using a host name. But 1409the problem does not occur if DISPLAY is set to `unix:0.0'. I think 1410the bug has to do with SIGIO or FIONREAD. 1411 1412You may be able to compensate for the bug by doing (set-input-mode nil nil). 1413However, that has the disadvantage of turning off interrupts, so that 1414you are unable to quit out of a Lisp program by typing C-g. 1415 1416The easy way to do this is to put 1417 1418 (setq x-sigio-bug t) 1419 1420in your site-init.el file. 1421 1422* Runtime problems on character terminals 1423 1424** Emacs spontaneously displays "I-search: " at the bottom of the screen. 1425 1426This means that Control-S/Control-Q (XON/XOFF) "flow control" is being 1427used. C-s/C-q flow control is bad for Emacs editors because it takes 1428away C-s and C-q as user commands. Since editors do not output long 1429streams of text without user commands, there is no need for a 1430user-issuable "stop output" command in an editor; therefore, a 1431properly designed flow control mechanism would transmit all possible 1432input characters without interference. Designing such a mechanism is 1433easy, for a person with at least half a brain. 1434 1435There are three possible reasons why flow control could be taking place: 1436 1437 1) Terminal has not been told to disable flow control 1438 2) Insufficient padding for the terminal in use 1439 3) Some sort of terminal concentrator or line switch is responsible 1440 1441First of all, many terminals have a set-up mode which controls whether 1442they generate XON/XOFF flow control characters. This must be set to 1443"no XON/XOFF" in order for Emacs to work. Sometimes there is an 1444escape sequence that the computer can send to turn flow control off 1445and on. If so, perhaps the termcap `ti' string should turn flow 1446control off, and the `te' string should turn it on. 1447 1448Once the terminal has been told "no flow control", you may find it 1449needs more padding. The amount of padding Emacs sends is controlled 1450by the termcap entry for the terminal in use, and by the output baud 1451rate as known by the kernel. The shell command `stty' will print 1452your output baud rate; `stty' with suitable arguments will set it if 1453it is wrong. Setting to a higher speed causes increased padding. If 1454the results are wrong for the correct speed, there is probably a 1455problem in the termcap entry. You must speak to a local Unix wizard 1456to fix this. Perhaps you are just using the wrong terminal type. 1457 1458For terminals that lack a "no flow control" mode, sometimes just 1459giving lots of padding will prevent actual generation of flow control 1460codes. You might as well try it. 1461 1462If you are really unlucky, your terminal is connected to the computer 1463through a concentrator which sends XON/XOFF flow control to the 1464computer, or it insists on sending flow control itself no matter how 1465much padding you give it. Unless you can figure out how to turn flow 1466control off on this concentrator (again, refer to your local wizard), 1467you are screwed! You should have the terminal or concentrator 1468replaced with a properly designed one. In the mean time, some drastic 1469measures can make Emacs semi-work. 1470 1471You can make Emacs ignore C-s and C-q and let the operating system 1472handle them. To do this on a per-session basis, just type M-x 1473enable-flow-control RET. You will see a message that C-\ and C-^ are 1474now translated to C-s and C-q. (Use the same command M-x 1475enable-flow-control to turn *off* this special mode. It toggles flow 1476control handling.) 1477 1478If C-\ and C-^ are inconvenient for you (for example, if one of them 1479is the escape character of your terminal concentrator), you can choose 1480other characters by setting the variables flow-control-c-s-replacement 1481and flow-control-c-q-replacement. But choose carefully, since all 1482other control characters are already used by emacs. 1483 1484IMPORTANT: if you type C-s by accident while flow control is enabled, 1485Emacs output will freeze, and you will have to remember to type C-q in 1486order to continue. 1487 1488If you work in an environment where a majority of terminals of a 1489certain type are flow control hobbled, you can use the function 1490`enable-flow-control-on' to turn on this flow control avoidance scheme 1491automatically. Here is an example: 1492 1493(enable-flow-control-on "vt200" "vt300" "vt101" "vt131") 1494 1495If this isn't quite correct (e.g. you have a mixture of flow-control hobbled 1496and good vt200 terminals), you can still run enable-flow-control 1497manually. 1498 1499I have no intention of ever redesigning the Emacs command set for the 1500assumption that terminals use C-s/C-q flow control. XON/XOFF flow 1501control technique is a bad design, and terminals that need it are bad 1502merchandise and should not be purchased. Now that X is becoming 1503widespread, XON/XOFF seems to be on the way out. If you can get some 1504use out of GNU Emacs on inferior terminals, more power to you, but I 1505will not make Emacs worse for properly designed systems for the sake 1506of inferior systems. 1507 1508** Control-S and Control-Q commands are ignored completely. 1509 1510For some reason, your system is using brain-damaged C-s/C-q flow 1511control despite Emacs's attempts to turn it off. Perhaps your 1512terminal is connected to the computer through a concentrator 1513that wants to use flow control. 1514 1515You should first try to tell the concentrator not to use flow control. 1516If you succeed in this, try making the terminal work without 1517flow control, as described in the preceding section. 1518 1519If that line of approach is not successful, map some other characters 1520into C-s and C-q using keyboard-translate-table. The example above 1521shows how to do this with C-^ and C-\. 1522 1523** Screen is updated wrong, but only on one kind of terminal. 1524 1525This could mean that the termcap entry you are using for that 1526terminal is wrong, or it could mean that Emacs has a bug handing 1527the combination of features specified for that terminal. 1528 1529The first step in tracking this down is to record what characters 1530Emacs is sending to the terminal. Execute the Lisp expression 1531(open-termscript "./emacs-script") to make Emacs write all 1532terminal output into the file ~/emacs-script as well; then do 1533what makes the screen update wrong, and look at the file 1534and decode the characters using the manual for the terminal. 1535There are several possibilities: 1536 15371) The characters sent are correct, according to the terminal manual. 1538 1539In this case, there is no obvious bug in Emacs, and most likely you 1540need more padding, or possibly the terminal manual is wrong. 1541 15422) The characters sent are incorrect, due to an obscure aspect 1543 of the terminal behavior not described in an obvious way 1544 by termcap. 1545 1546This case is hard. It will be necessary to think of a way for 1547Emacs to distinguish between terminals with this kind of behavior 1548and other terminals that behave subtly differently but are 1549classified the same by termcap; or else find an algorithm for 1550Emacs to use that avoids the difference. Such changes must be 1551tested on many kinds of terminals. 1552 15533) The termcap entry is wrong. 1554 1555See the file etc/TERMS for information on changes 1556that are known to be needed in commonly used termcap entries 1557for certain terminals. 1558 15594) The characters sent are incorrect, and clearly cannot be 1560 right for any terminal with the termcap entry you were using. 1561 1562This is unambiguously an Emacs bug, and can probably be fixed 1563in termcap.c, tparam.c, term.c, scroll.c, cm.c or dispnew.c. 1564 1565** Control-S and Control-Q commands are ignored completely on a net connection. 1566 1567Some versions of rlogin (and possibly telnet) do not pass flow 1568control characters to the remote system to which they connect. 1569On such systems, emacs on the remote system cannot disable flow 1570control on the local system. 1571 1572One way to cure this is to disable flow control on the local host 1573(the one running rlogin, not the one running rlogind) using the 1574stty command, before starting the rlogin process. On many systems, 1575"stty start u stop u" will do this. 1576 1577Some versions of tcsh will prevent even this from working. One way 1578around this is to start another shell before starting rlogin, and 1579issue the stty command to disable flow control from that shell. 1580 1581If none of these methods work, the best solution is to type 1582M-x enable-flow-control at the beginning of your emacs session, or 1583if you expect the problem to continue, add a line such as the 1584following to your .emacs (on the host running rlogind): 1585 1586(enable-flow-control-on "vt200" "vt300" "vt101" "vt131") 1587 1588See the entry about spontaneous display of I-search (above) for more 1589info. 1590 1591** Output from Control-V is slow. 1592 1593On many bit-map terminals, scrolling operations are fairly slow. 1594Often the termcap entry for the type of terminal in use fails 1595to inform Emacs of this. The two lines at the bottom of the screen 1596before a Control-V command are supposed to appear at the top after 1597the Control-V command. If Emacs thinks scrolling the lines is fast, 1598it will scroll them to the top of the screen. 1599 1600If scrolling is slow but Emacs thinks it is fast, the usual reason is 1601that the termcap entry for the terminal you are using does not 1602specify any padding time for the `al' and `dl' strings. Emacs 1603concludes that these operations take only as much time as it takes to 1604send the commands at whatever line speed you are using. You must 1605fix the termcap entry to specify, for the `al' and `dl', as much 1606time as the operations really take. 1607 1608Currently Emacs thinks in terms of serial lines which send characters 1609at a fixed rate, so that any operation which takes time for the 1610terminal to execute must also be padded. With bit-map terminals 1611operated across networks, often the network provides some sort of 1612flow control so that padding is never needed no matter how slow 1613an operation is. You must still specify a padding time if you want 1614Emacs to realize that the operation takes a long time. This will 1615cause padding characters to be sent unnecessarily, but they do 1616not really cost much. They will be transmitted while the scrolling 1617is happening and then discarded quickly by the terminal. 1618 1619Most bit-map terminals provide commands for inserting or deleting 1620multiple lines at once. Define the `AL' and `DL' strings in the 1621termcap entry to say how to do these things, and you will have 1622fast output without wasted padding characters. These strings should 1623each contain a single %-spec saying how to send the number of lines 1624to be scrolled. These %-specs are like those in the termcap 1625`cm' string. 1626 1627You should also define the `IC' and `DC' strings if your terminal 1628has a command to insert or delete multiple characters. These 1629take the number of positions to insert or delete as an argument. 1630 1631A `cs' string to set the scrolling region will reduce the amount 1632of motion you see on the screen when part of the screen is scrolled. 1633 1634** You type Control-H (Backspace) expecting to delete characters. 1635 1636Put `stty dec' in your .login file and your problems will disappear 1637after a day or two. 1638 1639The choice of Backspace for erasure was based on confusion, caused by 1640the fact that backspacing causes erasure (later, when you type another 1641character) on most display terminals. But it is a mistake. Deletion 1642of text is not the same thing as backspacing followed by failure to 1643overprint. I do not wish to propagate this confusion by conforming 1644to it. 1645 1646For this reason, I believe `stty dec' is the right mode to use, 1647and I have designed Emacs to go with that. If there were a thousand 1648other control characters, I would define Control-h to delete as well; 1649but there are not very many other control characters, and I think 1650that providing the most mnemonic possible Help character is more 1651important than adapting to people who don't use `stty dec'. 1652 1653If you are obstinate about confusing buggy overprinting with deletion, 1654you can redefine Backspace in your .emacs file: 1655 (global-set-key "\b" 'delete-backward-char) 1656You can probably access help-command via f1. 1657 1658** Colors are not available on a tty or in xterm. 1659 1660Emacs 21 supports colors on character terminals and terminal 1661emulators, but this support relies on the terminfo or termcap database 1662entry to specify that the display supports color. Emacs looks at the 1663"Co" capability for the terminal to find out how many colors are 1664supported; it should be non-zero to activate the color support within 1665Emacs. (Most color terminals support 8 or 16 colors.) If your system 1666uses terminfo, the name of the capability equivalent to "Co" is 1667"colors". 1668 1669In addition to the "Co" capability, Emacs needs the "op" (for 1670``original pair'') capability, which tells how to switch the terminal 1671back to the default foreground and background colors. Emacs will not 1672use colors if this capability is not defined. If your terminal entry 1673doesn't provide such a capability, try using the ANSI standard escape 1674sequence \E[00m (that is, define a new termcap/terminfo entry and make 1675it use your current terminal's entry plus \E[00m for the "op" 1676capability). 1677 1678Finally, the "NC" capability (terminfo name: "ncv") tells Emacs which 1679attributes cannot be used with colors. Setting this capability 1680incorrectly might have the effect of disabling colors; try setting 1681this capability to `0' (zero) and see if that helps. 1682 1683Emacs uses the database entry for the terminal whose name is the value 1684of the environment variable TERM. With `xterm', a common terminal 1685entry that supports color is `xterm-color', so setting TERM's value to 1686`xterm-color' might activate the color support on an xterm-compatible 1687emulator. 1688 1689Beginning with version 22.1, Emacs supports the --color command-line 1690option which may be used to force Emacs to use one of a few popular 1691modes for getting colors on a tty. For example, --color=ansi8 sets up 1692for using the ANSI-standard escape sequences that support 8 colors. 1693 1694Some modes do not use colors unless you turn on the Font-lock mode. 1695Some people have long ago set their `~/.emacs' files to turn on 1696Font-lock on X only, so they won't see colors on a tty. The 1697recommended way of turning on Font-lock is by typing "M-x 1698global-font-lock-mode RET" or by customizing the variable 1699`global-font-lock-mode'. 1700 1701* Runtime problems specific to individual Unix variants 1702 1703** GNU/Linux 1704 1705*** GNU/Linux: Process output is corrupted. 1706 1707There is a bug in Linux kernel 2.6.10 PTYs that can cause emacs to 1708read corrupted process output. 1709 1710*** GNU/Linux: Remote access to CVS with SSH causes file corruption. 1711 1712If you access a remote CVS repository via SSH, files may be corrupted 1713due to bad interaction between CVS, SSH, and libc. 1714 1715To fix the problem, save the following script into a file, make it 1716executable, and set CVS_RSH environment variable to the file name of 1717the script: 1718 1719#!/bin/bash 1720exec 2> >(exec cat >&2 2>/dev/null) 1721exec ssh "$@" 1722 1723*** GNU/Linux: On Linux-based GNU systems using libc versions 5.4.19 through 17245.4.22, Emacs crashes at startup with a segmentation fault. 1725 1726This problem happens if libc defines the symbol __malloc_initialized. 1727One known solution is to upgrade to a newer libc version. 5.4.33 is 1728known to work. 1729 1730*** GNU/Linux: After upgrading to a newer version of Emacs, 1731the Meta key stops working. 1732 1733This was reported to happen on a GNU/Linux system distributed by 1734Mandrake. The reason is that the previous version of Emacs was 1735modified by Mandrake to make the Alt key act as the Meta key, on a 1736keyboard where the Windows key is the one which produces the Meta 1737modifier. A user who started using a newer version of Emacs, which 1738was not hacked by Mandrake, expected the Alt key to continue to act as 1739Meta, and was astonished when that didn't happen. 1740 1741The solution is to find out what key on your keyboard produces the Meta 1742modifier, and use that key instead. Try all of the keys to the left 1743and to the right of the space bar, together with the `x' key, and see 1744which combination produces "M-x" in the echo area. You can also use 1745the `xmodmap' utility to show all the keys which produce a Meta 1746modifier: 1747 1748 xmodmap -pk | egrep -i "meta|alt" 1749 1750A more convenient way of finding out which keys produce a Meta modifier 1751is to use the `xkbprint' utility, if it's available on your system: 1752 1753 xkbprint 0:0 /tmp/k.ps 1754 1755This produces a PostScript file `/tmp/k.ps' with a picture of your 1756keyboard; printing that file on a PostScript printer will show what 1757keys can serve as Meta. 1758 1759The `xkeycaps' also shows a visual representation of the current 1760keyboard settings. It also allows to modify them. 1761 1762*** GNU/Linux: slow startup on Linux-based GNU systems. 1763 1764People using systems based on the Linux kernel sometimes report that 1765startup takes 10 to 15 seconds longer than `usual'. 1766 1767This is because Emacs looks up the host name when it starts. 1768Normally, this takes negligible time; the extra delay is due to 1769improper system configuration. This problem can occur for both 1770networked and non-networked machines. 1771 1772Here is how to fix the configuration. It requires being root. 1773 1774**** Networked Case. 1775 1776First, make sure the files `/etc/hosts' and `/etc/host.conf' both 1777exist. The first line in the `/etc/hosts' file should look like this 1778(replace HOSTNAME with your host name): 1779 1780 127.0.0.1 HOSTNAME 1781 1782Also make sure that the `/etc/host.conf' files contains the following 1783lines: 1784 1785 order hosts, bind 1786 multi on 1787 1788Any changes, permanent and temporary, to the host name should be 1789indicated in the `/etc/hosts' file, since it acts a limited local 1790database of addresses and names (e.g., some SLIP connections 1791dynamically allocate ip addresses). 1792 1793**** Non-Networked Case. 1794 1795The solution described in the networked case applies here as well. 1796However, if you never intend to network your machine, you can use a 1797simpler solution: create an empty `/etc/host.conf' file. The command 1798`touch /etc/host.conf' suffices to create the file. The `/etc/hosts' 1799file is not necessary with this approach. 1800 1801*** GNU/Linux: Emacs on a tty switches the cursor to large blinking block. 1802 1803This was reported to happen on some GNU/Linux systems which use 1804ncurses version 5.0, but could be relevant for other versions as well. 1805These versions of ncurses come with a `linux' terminfo entry, where 1806the "cvvis" capability (termcap "vs") is defined as "\E[?25h\E[?8c" 1807(show cursor, change size). This escape sequence switches on a 1808blinking hardware text-mode cursor whose size is a full character 1809cell. This blinking cannot be stopped, since a hardware cursor 1810always blinks. 1811 1812A work-around is to redefine the "cvvis" capability so that it 1813enables a *software* cursor. The software cursor works by inverting 1814the colors of the character at point, so what you see is a block 1815cursor that doesn't blink. For this to work, you need to redefine 1816the "cnorm" capability as well, so that it operates on the software 1817cursor instead of the hardware cursor. 1818 1819To this end, run "infocmp linux > linux-term", edit the file 1820`linux-term' to make both the "cnorm" and "cvvis" capabilities send 1821the sequence "\E[?25h\E[?17;0;64c", and then run "tic linux-term" to 1822produce a modified terminfo entry. 1823 1824Alternatively, if you want a blinking underscore as your Emacs cursor, 1825change the "cvvis" capability to send the "\E[?25h\E[?0c" command. 1826 1827*** GNU/Linux: Error messages `internal facep []' happen on GNU/Linux systems. 1828 1829There is a report that replacing libc.so.5.0.9 with libc.so.5.2.16 1830caused this to start happening. People are not sure why, but the 1831problem seems unlikely to be in Emacs itself. Some suspect that it 1832is actually Xlib which won't work with libc.so.5.2.16. 1833 1834Using the old library version is a workaround. 1835 1836** Mac OS X 1837 1838*** Mac OS X (Carbon): Environment Variables from dotfiles are ignored. 1839 1840When starting Emacs from the Dock or the Finder on Mac OS X, the 1841environment variables that are set up in dotfiles, such as .cshrc or 1842.profile, are ignored. This is because the Finder and Dock are not 1843started from a shell, but instead from the Window Manager itself. 1844 1845The workaround for this is to create a .MacOSX/environment.plist file to 1846setup these environment variables. These environment variables will 1847apply to all processes regardless of where they are started. 1848For me information, see http://developer.apple.com/qa/qa2001/qa1067.html. 1849 1850*** Mac OS X (Carbon): Process output truncated when using ptys. 1851 1852There appears to be a problem with the implementation of pty's on the 1853Mac OS X that causes process output to be truncated. To avoid this, 1854leave process-connection-type set to its default value of nil. 1855 1856*** Mac OS X 10.3.9 (Carbon): QuickTime 7.0.4 updater breaks build. 1857 1858On the above environment, build fails at the link stage with the 1859message like "Undefined symbols: _HICopyAccessibilityActionDescription 1860referenced from QuickTime expected to be defined in Carbon". A 1861workaround is to use QuickTime 7.0.1 reinstaller. 1862 1863** FreeBSD 1864 1865*** FreeBSD 2.1.5: useless symbolic links remain in /tmp or other 1866directories that have the +t bit. 1867 1868This is because of a kernel bug in FreeBSD 2.1.5 (fixed in 2.2). 1869Emacs uses symbolic links to implement file locks. In a directory 1870with +t bit, the directory owner becomes the owner of the symbolic 1871link, so that it cannot be removed by anyone else. 1872 1873If you don't like those useless links, you can let Emacs not to using 1874file lock by adding #undef CLASH_DETECTION to config.h. 1875 1876*** FreeBSD: Getting a Meta key on the console. 1877 1878By default, neither Alt nor any other key acts as a Meta key on 1879FreeBSD, but this can be changed using kbdcontrol(1). Dump the 1880current keymap to a file with the command 1881 1882 $ kbdcontrol -d >emacs.kbd 1883 1884Edit emacs.kbd, and give the key you want to be the Meta key the 1885definition `meta'. For instance, if your keyboard has a ``Windows'' 1886key with scan code 105, change the line for scan code 105 in emacs.kbd 1887to look like this 1888 1889 105 meta meta meta meta meta meta meta meta O 1890 1891to make the Windows key the Meta key. Load the new keymap with 1892 1893 $ kbdcontrol -l emacs.kbd 1894 1895** HP-UX 1896 1897*** HP/UX : Shell mode gives the message, "`tty`: Ambiguous". 1898 1899christos@theory.tn.cornell.edu says: 1900 1901The problem is that in your .cshrc you have something that tries to 1902execute `tty`. If you are not running the shell on a real tty then 1903tty will print "not a tty". Csh expects one word in some places, 1904but tty is giving it back 3. 1905 1906The solution is to add a pair of quotes around `tty` to make it a single 1907word: 1908 1909if (`tty` == "/dev/console") 1910 1911should be changed to: 1912 1913if ("`tty`" == "/dev/console") 1914 1915Even better, move things that set up terminal sections out of .cshrc 1916and into .login. 1917 1918*** HP/UX: `Pid xxx killed due to text modification or page I/O error'. 1919 1920On HP/UX, you can get that error when the Emacs executable is on an NFS 1921file system. HP/UX responds this way if it tries to swap in a page and 1922does not get a response from the server within a timeout whose default 1923value is just ten seconds. 1924 1925If this happens to you, extend the timeout period. 1926 1927*** HP/UX: The right Alt key works wrong on German HP keyboards (and perhaps 1928other non-English HP keyboards too). 1929 1930This is because HP-UX defines the modifiers wrong in X. Here is a 1931shell script to fix the problem; be sure that it is run after VUE 1932configures the X server. 1933 1934 xmodmap 2> /dev/null - << EOF 1935 keysym Alt_L = Meta_L 1936 keysym Alt_R = Meta_R 1937 EOF 1938 1939 xmodmap - << EOF 1940 clear mod1 1941 keysym Mode_switch = NoSymbol 1942 add mod1 = Meta_L 1943 keysym Meta_R = Mode_switch 1944 add mod2 = Mode_switch 1945 EOF 1946 1947*** HP/UX: "Cannot find callback list" messages from dialog boxes in 1948Emacs built with Motif. 1949 1950This problem resulted from a bug in GCC 2.4.5. Newer GCC versions 1951such as 2.7.0 fix the problem. 1952 1953*** HP/UX: Emacs does not recognize the AltGr key. 1954 1955To fix this, set up a file ~/.dt/sessions/sessionetc with executable 1956rights, containing this text: 1957 1958-------------------------------- 1959xmodmap 2> /dev/null - << EOF 1960keysym Alt_L = Meta_L 1961keysym Alt_R = Meta_R 1962EOF 1963 1964xmodmap - << EOF 1965clear mod1 1966keysym Mode_switch = NoSymbol 1967add mod1 = Meta_L 1968keysym Meta_R = Mode_switch 1969add mod2 = Mode_switch 1970EOF 1971-------------------------------- 1972 1973*** HP/UX 11.0: Emacs makes HP/UX 11.0 crash. 1974 1975This is a bug in HPUX; HPUX patch PHKL_16260 is said to fix it. 1976 1977** AIX 1978 1979*** AIX: Trouble using ptys. 1980 1981People often install the pty devices on AIX incorrectly. 1982Use `smit pty' to reinstall them properly. 1983 1984*** AIXterm: Your Delete key sends a Backspace to the terminal. 1985 1986The solution is to include in your .Xdefaults the lines: 1987 1988 *aixterm.Translations: #override <Key>BackSpace: string(0x7f) 1989 aixterm*ttyModes: erase ^? 1990 1991This makes your Backspace key send DEL (ASCII 127). 1992 1993*** AIX: If linking fails because libXbsd isn't found, check if you 1994are compiling with the system's `cc' and CFLAGS containing `-O5'. If 1995so, you have hit a compiler bug. Please make sure to re-configure 1996Emacs so that it isn't compiled with `-O5'. 1997 1998*** AIX 4.3.x or 4.4: Compiling fails. 1999 2000This could happen if you use /bin/c89 as your compiler, instead of 2001the default `cc'. /bin/c89 treats certain warnings, such as benign 2002redefinitions of macros, as errors, and fails the build. A solution 2003is to use the default compiler `cc'. 2004 2005*** AIX 4: Some programs fail when run in a Shell buffer 2006with an error message like No terminfo entry for "unknown". 2007 2008On AIX, many terminal type definitions are not installed by default. 2009`unknown' is one of them. Install the "Special Generic Terminal 2010Definitions" to make them defined. 2011 2012** Solaris 2013 2014We list bugs in current versions here. Solaris 2.x and 4.x are covered in the 2015section on legacy systems. 2016 2017*** On Solaris, C-x doesn't get through to Emacs when you use the console. 2018 2019This is a Solaris feature (at least on Intel x86 cpus). Type C-r 2020C-r C-t, to toggle whether C-x gets through to Emacs. 2021 2022*** Problem with remote X server on Suns. 2023 2024On a Sun, running Emacs on one machine with the X server on another 2025may not work if you have used the unshared system libraries. This 2026is because the unshared libraries fail to use YP for host name lookup. 2027As a result, the host name you specify may not be recognized. 2028 2029*** Solaris 2,6: Emacs crashes with SIGBUS or SIGSEGV on Solaris after you delete a frame. 2030 2031We suspect that this is a bug in the X libraries provided by 2032Sun. There is a report that one of these patches fixes the bug and 2033makes the problem stop: 2034 2035105216-01 105393-01 105518-01 105621-01 105665-01 105615-02 105216-02 2036105667-01 105401-08 105615-03 105621-02 105686-02 105736-01 105755-03 2037106033-01 105379-01 105786-01 105181-04 105379-03 105786-04 105845-01 2038105284-05 105669-02 105837-01 105837-02 105558-01 106125-02 105407-01 2039 2040Another person using a newer system (kernel patch level Generic_105181-06) 2041suspects that the bug was fixed by one of these more recent patches: 2042 2043106040-07 SunOS 5.6: X Input & Output Method patch 2044106222-01 OpenWindows 3.6: filemgr (ff.core) fixes 2045105284-12 Motif 1.2.7: sparc Runtime library patch 2046 2047*** Solaris 7 or 8: Emacs reports a BadAtom error (from X) 2048 2049This happens when Emacs was built on some other version of Solaris. 2050Rebuild it on Solaris 8. 2051 2052*** When using M-x dbx with the SparcWorks debugger, the `up' and `down' 2053commands do not move the arrow in Emacs. 2054 2055You can fix this by adding the following line to `~/.dbxinit': 2056 2057 dbxenv output_short_file_name off 2058 2059*** On Solaris, CTRL-t is ignored by Emacs when you use 2060the fr.ISO-8859-15 locale (and maybe other related locales). 2061 2062You can fix this by editing the file: 2063 2064 /usr/openwin/lib/locale/iso8859-15/Compose 2065 2066Near the bottom there is a line that reads: 2067 2068 Ctrl<t> <quotedbl> <Y> : "\276" threequarters 2069 2070that should read: 2071 2072 Ctrl<T> <quotedbl> <Y> : "\276" threequarters 2073 2074Note the lower case <t>. Changing this line should make C-t work. 2075 2076** Irix 2077 2078*** Irix 6.5: Emacs crashes on the SGI R10K, when compiled with GCC. 2079 2080This seems to be fixed in GCC 2.95. 2081 2082*** Irix: Trouble using ptys, or running out of ptys. 2083 2084The program mkpts (which may be in `/usr/adm' or `/usr/sbin') needs to 2085be set-UID to root, or non-root programs like Emacs will not be able 2086to allocate ptys reliably. 2087 2088* Runtime problems specific to MS-Windows 2089 2090** Windows 95 and networking. 2091 2092To support server sockets, Emacs 22.1 loads ws2_32.dll. If this file 2093is missing, all Emacs networking features are disabled. 2094 2095Old versions of Windows 95 may not have the required DLL. To use 2096Emacs' networking features on Windows 95, you must install the 2097"Windows Socket 2" update available from MicroSoft's support Web. 2098 2099** Emacs exits with "X protocol error" when run with an X server for MS-Windows. 2100 2101A certain X server for Windows had a bug which caused this. 2102Supposedly the newer 32-bit version of this server doesn't have the 2103problem. 2104 2105** Known problems with the MS-Windows port of Emacs 22.1 2106 2107Using create-fontset-from-ascii-font or the --font startup parameter 2108with a Chinese, Japanese or Korean font leads to display problems. 2109Use a Latin-only font as your default font. If you want control over 2110which font is used to display Chinese, Japanese or Korean character, 2111use create-fontset-from-fontset-spec to define a fontset. 2112 2113Frames are not refreshed while the File or Font dialog or a pop-up menu 2114is displayed. This also means help text for pop-up menus is not 2115displayed at all. This is because message handling under Windows is 2116synchronous, so we cannot handle repaint (or any other) messages while 2117waiting for a system function to return the result of the dialog or 2118pop-up menu interaction. 2119 2120Windows 95 and Windows NT up to version 4.0 do not support help text 2121for menus. Help text is only available in later versions of Windows. 2122 2123When "ClearType" method is selected as the "method to smooth edges of 2124screen fonts" (in Display Properties, Appearance tab, under 2125"Effects"), there are various problems related to display of 2126characters: 2-pixel trace is left behind when moving overlays, bold 2127fonts can be hard to read, small portions of some characters could 2128appear chopped, etc. This happens because, under ClearType, 2129characters are drawn outside their advertised bounding box. Emacs 21 2130disabled the use of ClearType, whereas Emacs 22 allows it and has some 2131code to enlarge the width of the bounding box. Apparently, this 2132display feature needs more changes to get it 100% right. A workaround 2133is to disable ClearType. 2134 2135There are problems with display if mouse-tracking is enabled and the 2136mouse is moved off a frame, over another frame then back over the first 2137frame. A workaround is to click the left mouse button inside the frame 2138after moving back into it. 2139 2140Some minor flickering still persists during mouse-tracking, although 2141not as severely as in 21.1. 2142 2143An inactive cursor remains in an active window after the Windows 2144Manager driven switch of the focus, until a key is pressed. 2145 2146Windows input methods are not recognized by Emacs. However, some 2147of these input methods cause the keyboard to send characters encoded 2148in the appropriate coding system (e.g., ISO 8859-1 for Latin-1 2149characters, ISO 8859-8 for Hebrew characters, etc.). To make these 2150input methods work with Emacs, set the keyboard coding system to the 2151appropriate value after you activate the Windows input method. For 2152example, if you activate the Hebrew input method, type this: 2153 2154 C-x RET k hebrew-iso-8bit RET 2155 2156(Emacs ought to recognize the Windows language-change event and set up 2157the appropriate keyboard encoding automatically, but it doesn't do 2158that yet.) In addition, to use these Windows input methods, you 2159should set your "Language for non-Unicode programs" (on Windows XP, 2160this is on the Advanced tab of Regional Settings) to the language of 2161the input method. 2162 2163To bind keys that produce non-ASCII characters with modifiers, you 2164must specify raw byte codes. For instance, if you want to bind 2165META-a-grave to a command, you need to specify this in your `~/.emacs': 2166 2167 (global-set-key [?\M-\340] ...) 2168 2169The above example is for the Latin-1 environment where the byte code 2170of the encoded a-grave is 340 octal. For other environments, use the 2171encoding appropriate to that environment. 2172 2173The %b specifier for format-time-string does not produce abbreviated 2174month names with consistent widths for some locales on some versions 2175of Windows. This is caused by a deficiency in the underlying system 2176library function. 2177 2178Files larger than 4GB cause overflow in the size (represented as a 217932-bit integer) reported by `file-attributes'. This affects Dired as 2180well, since the Windows port uses a Lisp emulation of `ls' that relies 2181on `file-attributes'. 2182 2183** Typing Alt-Shift has strange effects on MS-Windows. 2184 2185This combination of keys is a command to change keyboard layout. If 2186you proceed to type another non-modifier key before you let go of Alt 2187and Shift, the Alt and Shift act as modifiers in the usual way. A 2188more permanent work around is to change it to another key combination, 2189or disable it in the keyboard control panel. 2190 2191** Cygwin build of Emacs hangs after rebasing Cygwin DLLs 2192 2193Usually, on Cygwin, one needs to rebase the DLLs if an application 2194aborts with a message like this: 2195 2196 C:\cygwin\bin\python.exe: *** unable to remap C:\cygwin\bin\cygssl.dll to 2197 same address as parent(0xDF0000) != 0xE00000 2198 2199However, since Cygwin DLL 1.5.17 was released, after such rebasing, 2200Emacs hangs. 2201 2202This was reported to happen for Emacs 21.2 and also for the pretest of 2203Emacs 22.1 on Cygwin. 2204 2205To work around this, build Emacs like this: 2206 2207 LDFLAGS='-Wl,--enable-auto-import -Wl,--enable-auto-image-base' ./configure 2208 make LD='$(CC)' 2209 make LD='$(CC)' install 2210 2211This produces an Emacs binary that is independent of rebasing. 2212 2213Note that you _must_ use LD='$(CC)' in the last two commands above, to 2214prevent GCC from passing the "--image-base 0x20000000" option to the 2215linker, which is what it does by default. That option produces an 2216Emacs binary with the base address 0x20000000, which will cause Emacs 2217to hang after Cygwin DLLs are rebased. 2218 2219** Interrupting Cygwin port of Bash from Emacs doesn't work. 2220 2221Cygwin 1.x builds of the ported Bash cannot be interrupted from the 2222MS-Windows version of Emacs. This is due to some change in the Bash 2223port or in the Cygwin library which apparently make Bash ignore the 2224keyboard interrupt event sent by Emacs to Bash. (Older Cygwin ports 2225of Bash, up to b20.1, did receive SIGINT from Emacs.) 2226 2227** Accessing remote files with ange-ftp hangs the MS-Windows version of Emacs. 2228 2229If the FTP client is the Cygwin port of GNU `ftp', this appears to be 2230due to some bug in the Cygwin DLL or some incompatibility between it 2231and the implementation of asynchronous subprocesses in the Windows 2232port of Emacs. Specifically, some parts of the FTP server responses 2233are not flushed out, apparently due to buffering issues, which 2234confuses ange-ftp. 2235 2236The solution is to downgrade to an older version of the Cygwin DLL 2237(version 1.3.2 was reported to solve the problem), or use the stock 2238Windows FTP client, usually found in the `C:\WINDOWS' or 'C:\WINNT' 2239directory. To force ange-ftp use the stock Windows client, set the 2240variable `ange-ftp-ftp-program-name' to the absolute file name of the 2241client's executable. For example: 2242 2243 (setq ange-ftp-ftp-program-name "c:/windows/ftp.exe") 2244 2245If you want to stick with the Cygwin FTP client, you can work around 2246this problem by putting this in your `.emacs' file: 2247 2248 (setq ange-ftp-ftp-program-args '("-i" "-n" "-g" "-v" "--prompt" "") 2249 2250** lpr commands don't work on MS-Windows with some cheap printers. 2251 2252This problem may also strike other platforms, but the solution is 2253likely to be a global one, and not Emacs specific. 2254 2255Many cheap inkjet, and even some cheap laser printers, do not 2256print plain text anymore, they will only print through graphical 2257printer drivers. A workaround on MS-Windows is to use Windows' basic 2258built in editor to print (this is possibly the only useful purpose it 2259has): 2260 2261(setq printer-name "") ;; notepad takes the default 2262(setq lpr-command "notepad") ;; notepad 2263(setq lpr-switches nil) ;; not needed 2264(setq lpr-printer-switch "/P") ;; run notepad as batch printer 2265 2266** Antivirus software interacts badly with the MS-Windows version of Emacs. 2267 2268The usual manifestation of these problems is that subprocesses don't 2269work or even wedge the entire system. In particular, "M-x shell RET" 2270was reported to fail to work. But other commands also sometimes don't 2271work when an antivirus package is installed. 2272 2273The solution is to switch the antivirus software to a less aggressive 2274mode (e.g., disable the ``auto-protect'' feature), or even uninstall 2275or disable it entirely. 2276 2277** Pressing the mouse button on MS-Windows does not give a mouse-2 event. 2278 2279This is usually a problem with the mouse driver. Because most Windows 2280programs do not do anything useful with the middle mouse button, many 2281mouse drivers allow you to define the wheel press to do something 2282different. Some drivers do not even have the option to generate a 2283middle button press. In such cases, setting the wheel press to 2284"scroll" sometimes works if you press the button twice. Trying a 2285generic mouse driver might help. 2286 2287** Scrolling the mouse wheel on MS-Windows always scrolls the top window. 2288 2289This is another common problem with mouse drivers. Instead of 2290generating scroll events, some mouse drivers try to fake scroll bar 2291movement. But they are not intelligent enough to handle multiple 2292scroll bars within a frame. Trying a generic mouse driver might help. 2293 2294** Mail sent through Microsoft Exchange in some encodings appears to be 2295mangled and is not seen correctly in Rmail or Gnus. We don't know 2296exactly what happens, but it isn't an Emacs problem in cases we've 2297seen. 2298 2299** On MS-Windows, you cannot use the right-hand ALT key and the left-hand 2300CTRL key together to type a Control-Meta character. 2301 2302This is a consequence of a misfeature beyond Emacs's control. 2303 2304Under Windows, the AltGr key on international keyboards generates key 2305events with the modifiers Right-Alt and Left-Ctrl. Since Emacs cannot 2306distinguish AltGr from an explicit Right-Alt and Left-Ctrl 2307combination, whenever it sees Right-Alt and Left-Ctrl it assumes that 2308AltGr has been pressed. The variable `w32-recognize-altgr' can be set 2309to nil to tell Emacs that AltGr is really Ctrl and Alt. 2310 2311** Under some X-servers running on MS-Windows, Emacs' display is incorrect. 2312 2313The symptoms are that Emacs does not completely erase blank areas of the 2314screen during scrolling or some other screen operations (e.g., selective 2315display or when killing a region). M-x recenter will cause the screen 2316to be completely redisplayed and the "extra" characters will disappear. 2317 2318This is known to occur under Exceed 6, and possibly earlier versions 2319as well; it is reportedly solved in version 6.2.0.16 and later. The 2320problem lies in the X-server settings. 2321 2322There are reports that you can solve the problem with Exceed by 2323running `Xconfig' from within NT, choosing "X selection", then 2324un-checking the boxes "auto-copy X selection" and "auto-paste to X 2325selection". 2326 2327Of this does not work, please inform bug-gnu-emacs@gnu.org. Then 2328please call support for your X-server and see if you can get a fix. 2329If you do, please send it to bug-gnu-emacs@gnu.org so we can list it 2330here. 2331 2332* Build-time problems 2333 2334** Configuration 2335 2336*** The `configure' script doesn't find the jpeg library. 2337 2338There are reports that this happens on some systems because the linker 2339by default only looks for shared libraries, but jpeg distribution by 2340default only installs a nonshared version of the library, `libjpeg.a'. 2341 2342If this is the problem, you can configure the jpeg library with the 2343`--enable-shared' option and then rebuild libjpeg. This produces a 2344shared version of libjpeg, which you need to install. Finally, rerun 2345the Emacs configure script, which should now find the jpeg library. 2346Alternatively, modify the generated src/Makefile to link the .a file 2347explicitly, and edit src/config.h to define HAVE_JPEG. 2348 2349*** `configure' warns ``accepted by the compiler, rejected by the preprocessor''. 2350 2351This indicates a mismatch between the C compiler and preprocessor that 2352configure is using. For example, on Solaris 10 trying to use 2353CC=/opt/SUNWspro/bin/cc (the Sun Studio compiler) together with 2354CPP=/usr/ccs/lib/cpp can result in errors of this form (you may also 2355see the error ``"/usr/include/sys/isa_defs.h", line 500: undefined control''). 2356 2357The solution is to tell configure to use the correct C preprocessor 2358for your C compiler (CPP="/opt/SUNWspro/bin/cc -E" in the above 2359example). 2360 2361*** `configure' fails with ``"junk.c", line 660: invalid input token: 8.elc'' 2362 2363The final stage of the Emacs configure process uses the C preprocessor 2364to generate the Makefiles. Errors of this form can occur if the C 2365preprocessor inserts extra whitespace into its output. The solution 2366is to find the switches that stop your preprocessor from inserting extra 2367whitespace, add them to CPPFLAGS, and re-run configure. For example, 2368this error can occur on Solaris 10 when using the Sun Studio compiler 2369``Sun C 5.8'' with its preprocessor CPP="/opt/SUNWspro/bin/cc -E". 2370The relevant switch in this case is "-Xs" (``compile assuming 2371(pre-ANSI) K & R C style code''). 2372 2373** Compilation 2374 2375*** Building Emacs over NFS fails with ``Text file busy''. 2376 2377This was reported to happen when building Emacs on a GNU/Linux system 2378(Red Hat Linux 6.2) using a build directory automounted from Solaris 2379(SunOS 5.6) file server, but it might not be limited to that 2380configuration alone. Presumably, the NFS server doesn't commit the 2381files' data to disk quickly enough, and the Emacs executable file is 2382left ``busy'' for several seconds after Emacs has finished dumping 2383itself. This causes the subsequent commands which invoke the dumped 2384Emacs executable to fail with the above message. 2385 2386In some of these cases, a time skew between the NFS server and the 2387machine where Emacs is built is detected and reported by GNU Make 2388(it says that some of the files have modification time in the future). 2389This might be a symptom of NFS-related problems. 2390 2391If the NFS server runs on Solaris, apply the Solaris patch 105379-05 2392(Sunos 5.6: /kernel/misc/nfssrv patch). If that doesn't work, or if 2393you have a different version of the OS or the NFS server, you can 2394force the NFS server to use 1KB blocks, which was reported to fix the 2395problem albeit at a price of slowing down file I/O. You can force 1KB 2396blocks by specifying the "-o rsize=1024,wsize=1024" options to the 2397`mount' command, or by adding ",rsize=1024,wsize=1024" to the mount 2398options in the appropriate system configuration file, such as 2399`/etc/auto.home'. 2400 2401Alternatively, when Make fails due to this problem, you could wait for 2402a few seconds and then invoke Make again. In one particular case, 2403waiting for 10 or more seconds between the two Make invocations seemed 2404to work around the problem. 2405 2406Similar problems can happen if your machine NFS-mounts a directory 2407onto itself. Suppose the Emacs sources live in `/usr/local/src' and 2408you are working on the host called `marvin'. Then an entry in the 2409`/etc/fstab' file like the following is asking for trouble: 2410 2411 marvin:/usr/local/src /usr/local/src ...options.omitted... 2412 2413The solution is to remove this line from `etc/fstab'. 2414 2415*** Building Emacs with GCC 2.9x fails in the `src' directory. 2416 2417This may happen if you use a development version of GNU `cpp' from one 2418of the GCC snapshots between Oct 2000 and Feb 2001, or from a released 2419version of GCC newer than 2.95.2 which was prepared around those 2420dates; similar problems were reported with some snapshots of GCC 3.1 2421around Sep 30 2001. The preprocessor in those versions is 2422incompatible with a traditional Unix cpp (e.g., it expands ".." into 2423". .", which breaks relative file names that reference the parent 2424directory; or inserts TAB characters before lines that set Make 2425variables). 2426 2427The solution is to make sure the preprocessor is run with the 2428`-traditional' option. The `configure' script does that automatically 2429when it detects the known problems in your cpp, but you might hit some 2430unknown ones. To force the `configure' script to use `-traditional', 2431run the script like this: 2432 2433 CPP='gcc -E -traditional' ./configure ... 2434 2435(replace the ellipsis "..." with any additional arguments you pass to 2436the script). 2437 2438Note that this problem does not pertain to the MS-Windows port of 2439Emacs, since it doesn't use the preprocessor to generate Makefiles. 2440 2441*** src/Makefile and lib-src/Makefile are truncated--most of the file missing. 2442*** Compiling wakeup, in lib-src, says it can't make wakeup.c. 2443 2444This can happen if configure uses GNU sed version 2.03. That version 2445had a bug. GNU sed version 2.05 works properly.To solve the 2446problem, install the current version of GNU Sed, then rerun Emacs's 2447configure script. 2448 2449*** Compiling lib-src says there is no rule to make test-distrib.c. 2450 2451This results from a bug in a VERY old version of GNU Sed. To solve 2452the problem, install the current version of GNU Sed, then rerun 2453Emacs's configure script. 2454 2455*** Building a 32-bit executable on a 64-bit GNU/Linux architecture. 2456 2457First ensure that the necessary 32-bit system libraries and include 2458files are installed. Then use: 2459 2460 env CC="gcc -m32" ./configure --build=i386-linux-gnu \ 2461 --x-libraries=/usr/X11R6/lib 2462 2463(using the location of the 32-bit X libraries on your system). 2464 2465*** Building the Cygwin port for MS-Windows can fail with some GCC versions 2466 2467Building Emacs 22 with Cygwin builds of GCC 3.4.4-1 and 3.4.4-2 is 2468reported to either fail or cause Emacs to segfault at run time. In 2469addition, the Cygwin GCC 3.4.4-2 has problems with generating debug 2470info. Cygwin users are advised not to use these versions of GCC for 2471compiling Emacs. GCC versions 4.0.3, 4.0.4, 4.1.1, and 4.1.2 2472reportedly build a working Cygwin binary of Emacs, so we recommend 2473these GCC versions. Note that these versions of GCC, 4.0.3, 4.0.4, 24744.1.1, and 4.1.2, are currently the _only_ versions known to succeed 2475in building Emacs (as of v22.1). 2476 2477*** Building the native MS-Windows port with Cygwin GCC can fail. 2478 2479Emacs may not build using recent Cygwin builds of GCC, such as Cygwin 2480version 1.1.8, using the default configure settings. It appears to be 2481necessary to specify the -mwin32 flag when compiling, and define 2482__MSVCRT__, like so: 2483 2484 configure --with-gcc --cflags -mwin32 --cflags -D__MSVCRT__ 2485 2486*** Building the MS-Windows port fails with a CreateProcess failure. 2487 2488Some versions of mingw32 make on some versions of Windows do not seem 2489to detect the shell correctly. Try "make SHELL=cmd.exe", or if that 2490fails, try running make from Cygwin bash instead. 2491 2492*** Building the MS-Windows port with Leim fails in the `leim' directory. 2493 2494The error message might be something like this: 2495 2496 Converting d:/emacs-21.3/leim/CXTERM-DIC/4Corner.tit to quail-package... 2497 Invalid ENCODE: value in TIT dictionary 2498 NMAKE : fatal error U1077: '"../src/obj-spd/i386/emacs.exe"' : return code 2499 '0xffffffff' 2500 Stop. 2501 2502This can happen if the Leim distribution is unpacked with a program 2503which converts the `*.tit' files to DOS-style CR-LF text format. The 2504`*.tit' files in the leim/CXTERM-DIC directory require Unix-style line 2505endings to compile properly, because Emacs reads them without any code 2506or EOL conversions. 2507 2508The solution is to make sure the program used to unpack Leim does not 2509change the files' line endings behind your back. The GNU FTP site has 2510in the `/gnu/emacs/windows' directory a program called `djtarnt.exe' 2511which can be used to unpack `.tar.gz' and `.zip' archives without 2512mangling them. 2513 2514*** Building `ctags' for MS-Windows with the MinGW port of GCC fails. 2515 2516This might happen due to a bug in the MinGW header assert.h, which 2517defines the `assert' macro with a trailing semi-colon. The following 2518patch to assert.h should solve this: 2519 2520 *** include/assert.h.orig Sun Nov 7 02:41:36 1999 2521 --- include/assert.h Mon Jan 29 11:49:10 2001 2522 *************** 2523 *** 41,47 **** 2524 /* 2525 * If not debugging, assert does nothing. 2526 */ 2527 ! #define assert(x) ((void)0); 2528 2529 #else /* debugging enabled */ 2530 2531 --- 41,47 ---- 2532 /* 2533 * If not debugging, assert does nothing. 2534 */ 2535 ! #define assert(x) ((void)0) 2536 2537 #else /* debugging enabled */ 2538 2539 2540*** Building the MS-Windows port with Visual Studio 2005 fails. 2541 2542Microsoft no longer ships the single threaded version of the C library 2543with their compiler, and the multithreaded static library is missing 2544some functions that Microsoft have deemed non-threadsafe. The 2545dynamically linked C library has all the functions, but there is a 2546conflict between the versions of malloc in the DLL and in Emacs, which 2547is not resolvable due to the way Windows does dynamic linking. 2548 2549We recommend the use of the MingW port of GCC for compiling Emacs, as 2550not only does it not suffer these problems, but it is also Free 2551software like Emacs. 2552 2553** Linking 2554 2555*** Building Emacs with a system compiler fails to link because of an 2556undefined symbol such as __eprintf which does not appear in Emacs. 2557 2558This can happen if some of the libraries linked into Emacs were built 2559with GCC, but Emacs itself is being linked with a compiler other than 2560GCC. Object files compiled with GCC might need some helper functions 2561from libgcc.a, the library which comes with GCC, but the system 2562compiler does not instruct the linker to search libgcc.a during the 2563link stage. 2564 2565A solution is to link with GCC, like this: 2566 2567 make CC=gcc 2568 2569Since the .o object files already exist, this will not recompile Emacs 2570with GCC, but just restart by trying again to link temacs. 2571 2572*** AIX 1.3 ptf 0013: Link failure. 2573 2574There is a real duplicate definition of the function `_slibc_free' in 2575the library /lib/libc_s.a (just do nm on it to verify). The 2576workaround/fix is: 2577 2578 cd /lib 2579 ar xv libc_s.a NLtmtime.o 2580 ar dv libc_s.a NLtmtime.o 2581 2582*** AIX 4.1.2: Linker error messages such as 2583 ld: 0711-212 SEVERE ERROR: Symbol .__quous, found in the global symbol table 2584 of archive /usr/lib/libIM.a, was not defined in archive member shr.o. 2585 2586This is a problem in libIM.a. You can work around it by executing 2587these shell commands in the src subdirectory of the directory where 2588you build Emacs: 2589 2590 cp /usr/lib/libIM.a . 2591 chmod 664 libIM.a 2592 ranlib libIM.a 2593 2594Then change -lIM to ./libIM.a in the command to link temacs (in 2595Makefile). 2596 2597*** Sun with acc: Link failure when using acc on a Sun. 2598 2599To use acc, you need additional options just before the libraries, such as 2600 2601 /usr/lang/SC2.0.1/values-Xt.o -L/usr/lang/SC2.0.1/cg87 -L/usr/lang/SC2.0.1 2602 2603and you need to add -lansi just before -lc. 2604 2605The precise file names depend on the compiler version, so we 2606cannot easily arrange to supply them. 2607 2608*** Linking says that the functions insque and remque are undefined. 2609 2610Change oldXMenu/Makefile by adding insque.o to the variable OBJS. 2611 2612*** `tparam' reported as a multiply-defined symbol when linking with ncurses. 2613 2614This problem results from an incompatible change in ncurses, in 2615version 1.9.9e approximately. This version is unable to provide a 2616definition of tparm without also defining tparam. This is also 2617incompatible with Terminfo; as a result, the Emacs Terminfo support 2618does not work with this version of ncurses. 2619 2620The fix is to install a newer version of ncurses, such as version 4.2. 2621 2622** Dumping 2623 2624*** Linux: Segfault during `make bootstrap' under certain recent versions of the Linux kernel. 2625 2626With certain recent Linux kernels (like the one of Red Hat Fedora Core 26271 and newer), the new "Exec-shield" functionality is enabled by default, which 2628creates a different memory layout that breaks the emacs dumper. Emacs tries 2629to handle this at build time, but if the workaround used fails, these 2630instructions can be useful. 2631The work-around explained here is not enough on Fedora Core 4 (and possible 2632newer). Read the next item. 2633 2634Configure can overcome the problem of exec-shield if the architecture is 2635x86 and the program setarch is present. On other architectures no 2636workaround is known. 2637 2638You can check the Exec-shield state like this: 2639 2640 cat /proc/sys/kernel/exec-shield 2641 2642It returns non-zero when Exec-shield is enabled, 0 otherwise. Please 2643read your system documentation for more details on Exec-shield and 2644associated commands. Exec-shield can be turned off with this command: 2645 2646 echo "0" > /proc/sys/kernel/exec-shield 2647 2648When Exec-shield is enabled, building Emacs will segfault during the 2649execution of this command: 2650 2651 ./temacs --batch --load loadup [dump|bootstrap] 2652 2653To work around this problem, it is necessary to temporarily disable 2654Exec-shield while building Emacs, or, on x86, by using the `setarch' 2655command when running temacs like this: 2656 2657 setarch i386 ./temacs --batch --load loadup [dump|bootstrap] 2658 2659 2660*** Fedora Core 4 GNU/Linux: Segfault during dumping. 2661 2662In addition to exec-shield explained above "Linux: Segfault during 2663`make bootstrap' under certain recent versions of the Linux kernel" 2664item, Linux kernel shipped with Fedora Core 4 randomizes the virtual 2665address space of a process. As the result dumping may fail even if 2666you turn off exec-shield. In this case, use the -R option to the setarch 2667command: 2668 2669 setarch i386 -R ./temacs --batch --load loadup [dump|bootstrap] 2670 2671or 2672 2673 setarch i386 -R make bootstrap 2674 2675*** Fatal signal in the command temacs -l loadup inc dump. 2676 2677This command is the final stage of building Emacs. It is run by the 2678Makefile in the src subdirectory, or by build.com on VMS. 2679 2680It has been known to get fatal errors due to insufficient swapping 2681space available on the machine. 2682 2683On 68000s, it has also happened because of bugs in the 2684subroutine `alloca'. Verify that `alloca' works right, even 2685for large blocks (many pages). 2686 2687*** test-distrib says that the distribution has been clobbered. 2688*** or, temacs prints "Command key out of range 0-127". 2689*** or, temacs runs and dumps emacs, but emacs totally fails to work. 2690*** or, temacs gets errors dumping emacs. 2691 2692This can be because the .elc files have been garbled. Do not be 2693fooled by the fact that most of a .elc file is text: these are 2694binary files and can contain all 256 byte values. 2695 2696In particular `shar' cannot be used for transmitting GNU Emacs. 2697It typically truncates "lines". What appear to be "lines" in 2698a binary file can of course be of any length. Even once `shar' 2699itself is made to work correctly, `sh' discards null characters 2700when unpacking the shell archive. 2701 2702I have also seen character \177 changed into \377. I do not know 2703what transfer means caused this problem. Various network 2704file transfer programs are suspected of clobbering the high bit. 2705 2706If you have a copy of Emacs that has been damaged in its 2707nonprinting characters, you can fix them: 2708 2709 1) Record the names of all the .elc files. 2710 2) Delete all the .elc files. 2711 3) Recompile alloc.c with a value of PURESIZE twice as large. 2712 (See puresize.h.) You might as well save the old alloc.o. 2713 4) Remake emacs. It should work now. 2714 5) Running emacs, do Meta-x byte-compile-file repeatedly 2715 to recreate all the .elc files that used to exist. 2716 You may need to increase the value of the variable 2717 max-lisp-eval-depth to succeed in running the compiler interpreted 2718 on certain .el files. 400 was sufficient as of last report. 2719 6) Reinstall the old alloc.o (undoing changes to alloc.c if any) 2720 and remake temacs. 2721 7) Remake emacs. It should work now, with valid .elc files. 2722 2723*** temacs prints "Pure Lisp storage exhausted". 2724 2725This means that the Lisp code loaded from the .elc and .el 2726files during temacs -l loadup inc dump took up more 2727space than was allocated. 2728 2729This could be caused by 2730 1) adding code to the preloaded Lisp files 2731 2) adding more preloaded files in loadup.el 2732 3) having a site-init.el or site-load.el which loads files. 2733 Note that ANY site-init.el or site-load.el is nonstandard; 2734 if you have received Emacs from some other site 2735 and it contains a site-init.el or site-load.el file, consider 2736 deleting that file. 2737 4) getting the wrong .el or .elc files 2738 (not from the directory you expected). 2739 5) deleting some .elc files that are supposed to exist. 2740 This would cause the source files (.el files) to be 2741 loaded instead. They take up more room, so you lose. 2742 6) a bug in the Emacs distribution which underestimates 2743 the space required. 2744 2745If the need for more space is legitimate, change the definition 2746of PURESIZE in puresize.h. 2747 2748But in some of the cases listed above, this problem is a consequence 2749of something else that is wrong. Be sure to check and fix the real 2750problem. 2751 2752*** Linux: Emacs crashes when dumping itself on Mac PPC running Yellow Dog GNU/Linux. 2753 2754The crashes happen inside the function Fmake_symbol; here's a typical 2755C backtrace printed by GDB: 2756 2757 0x190c0c0 in Fmake_symbol () 2758 (gdb) where 2759 #0 0x190c0c0 in Fmake_symbol () 2760 #1 0x1942ca4 in init_obarray () 2761 #2 0x18b3500 in main () 2762 #3 0x114371c in __libc_start_main (argc=5, argv=0x7ffff5b4, envp=0x7ffff5cc, 2763 2764This could happen because GCC version 2.95 and later changed the base 2765of the load address to 0x10000000. Emacs needs to be told about this, 2766but we currently cannot do that automatically, because that breaks 2767other versions of GNU/Linux on the MacPPC. Until we find a way to 2768distinguish between the Yellow Dog and the other varieties of 2769GNU/Linux systems on the PPC, you will have to manually uncomment the 2770following section near the end of the file src/m/macppc.h in the Emacs 2771distribution: 2772 2773 #if 0 /* This breaks things on PPC GNU/Linux except for Yellowdog, 2774 even with identical GCC, as, ld. Let's take it out until we 2775 know what's really going on here. */ 2776 /* GCC 2.95 and newer on GNU/Linux PPC changed the load address to 2777 0x10000000. */ 2778 #if defined __linux__ 2779 #if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 95) 2780 #define DATA_SEG_BITS 0x10000000 2781 #endif 2782 #endif 2783 #endif /* 0 */ 2784 2785Remove the "#if 0" and "#endif" directives which surround this, save 2786the file, and then reconfigure and rebuild Emacs. The dumping process 2787should now succeed. 2788 2789*** OpenBSD 4.0 macppc: Segfault during dumping. 2790 2791The build aborts with signal 11 when the command `./temacs --batch 2792--load loadup bootstrap' tries to load files.el. A workaround seems 2793to be to reduce the level of compiler optimization used during the 2794build (from -O2 to -O1). It is possible this is an OpenBSD 2795GCC problem specific to the macppc architecture, possibly only 2796occurring with older versions of GCC (e.g. 3.3.5). 2797 2798** Installation 2799 2800*** Installing Emacs gets an error running `install-info'. 2801 2802You need to install a recent version of Texinfo; that package 2803supplies the `install-info' command. 2804 2805*** Installing to a directory with spaces in the name fails. 2806 2807For example, if you call configure with a directory-related option 2808with spaces in the value, eg --enable-locallisppath='/path/with\ spaces'. 2809Using directory paths with spaces is not supported at this time: you 2810must re-configure without using spaces. 2811 2812*** Installing to a directory with non-ASCII characters in the name fails. 2813 2814Installation may fail, or the Emacs executable may not start 2815correctly, if a directory name containing non-ASCII characters is used 2816as a `configure' argument (e.g. `--prefix'). The problem can also 2817occur if a non-ASCII directory is specified in the EMACSLOADPATH 2818envvar. 2819 2820*** On Solaris, use GNU Make when installing an out-of-tree build 2821 2822The Emacs configuration process allows you to configure the 2823build environment so that you can build emacs in a directory 2824outside of the distribution tree. When installing Emacs from an 2825out-of-tree build directory on Solaris, you may need to use GNU 2826make. The make programs bundled with Solaris support the VPATH 2827macro but use it differently from the way the VPATH macro is 2828used by GNU make. The differences will cause the "make install" 2829step to fail, leaving you with an incomplete emacs 2830installation. GNU make is available in /usr/sfw/bin on Solaris 283110 and can be installed as /opt/sfw/bin/gmake from the Solaris 9 2832Software Companion CDROM. 2833 2834The problems due to the VPATH processing differences affect only 2835out of tree builds so, if you are on a Solaris installation 2836without GNU make, you can install Emacs completely by installing 2837from a build environment using the original emacs distribution tree. 2838 2839** First execution 2840 2841*** Emacs binary is not in executable format, and cannot be run. 2842 2843This was reported to happen when Emacs is built in a directory mounted 2844via NFS, for some combinations of NFS client and NFS server. 2845Usually, the file `emacs' produced in these cases is full of 2846binary null characters, and the `file' utility says: 2847 2848 emacs: ASCII text, with no line terminators 2849 2850We don't know what exactly causes this failure. A work-around is to 2851build Emacs in a directory on a local disk. 2852 2853*** The dumped Emacs crashes when run, trying to write pure data. 2854 2855Two causes have been seen for such problems. 2856 28571) On a system where getpagesize is not a system call, it is defined 2858as a macro. If the definition (in both unexec.c and malloc.c) is wrong, 2859it can cause problems like this. You might be able to find the correct 2860value in the man page for a.out (5). 2861 28622) Some systems allocate variables declared static among the 2863initialized variables. Emacs makes all initialized variables in most 2864of its files pure after dumping, but the variables declared static and 2865not initialized are not supposed to be pure. On these systems you 2866may need to add "#define static" to the m- or the s- file. 2867 2868* Emacs 19 problems 2869 2870** Error messages `Wrong number of arguments: #<subr where-is-internal>, 5'. 2871 2872This typically results from having the powerkey library loaded. 2873Powerkey was designed for Emacs 19.22. It is obsolete now because 2874Emacs 19 now has this feature built in; and powerkey also calls 2875where-is-internal in an obsolete way. 2876 2877So the fix is to arrange not to load powerkey. 2878 2879* Runtime problems on legacy systems 2880 2881This section covers bugs reported on very old hardware or software. 2882If you are using hardware and an operating system shipped after 2000, 2883it is unlikely you will see any of these. 2884 2885** Ancient operating systems 2886 2887AIX 4.2 was end-of-lifed on Dec 31st, 1999. 2888 2889*** AIX: You get this compiler error message: 2890 2891 Processing include file ./XMenuInt.h 2892 1501-106: (S) Include file X11/Xlib.h not found. 2893 2894This means your system was installed with only the X11 runtime i.d 2895libraries. You have to find your sipo (bootable tape) and install 2896X11Dev... with smit. 2897 2898(This report must be ancient. Bootable tapes are long dead.) 2899 2900*** AIX 3.2.4: Releasing Ctrl/Act key has no effect, if Shift is down. 2901 2902Due to a feature of AIX, pressing or releasing the Ctrl/Act key is 2903ignored when the Shift, Alt or AltGr keys are held down. This can 2904lead to the keyboard being "control-locked"--ordinary letters are 2905treated as control characters. 2906 2907You can get out of this "control-locked" state by pressing and 2908releasing Ctrl/Act while not pressing or holding any other keys. 2909 2910*** AIX 3.2.5: You get this message when running Emacs: 2911 2912 Could not load program emacs 2913 Symbol smtcheckinit in csh is undefined 2914 Error was: Exec format error 2915 2916or this one: 2917 2918 Could not load program .emacs 2919 Symbol _system_con in csh is undefined 2920 Symbol _fp_trapsta in csh is undefined 2921 Error was: Exec format error 2922 2923These can happen when you try to run on AIX 3.2.5 a program that was 2924compiled with 3.2.4. The fix is to recompile. 2925 2926*** AIX 4.2: Emacs gets a segmentation fault at startup. 2927 2928If you are using IBM's xlc compiler, compile emacs.c 2929without optimization; that should avoid the problem. 2930 2931*** ISC Unix 2932 2933**** ISC: display-time causes kernel problems on ISC systems. 2934 2935Under Interactive Unix versions 3.0.1 and 4.0 (and probably other 2936versions), display-time causes the loss of large numbers of STREVENT 2937cells. Eventually the kernel's supply of these cells is exhausted. 2938This makes emacs and the whole system run slow, and can make other 2939processes die, in particular pcnfsd. 2940 2941Other emacs functions that communicate with remote processes may have 2942the same problem. Display-time seems to be far the worst. 2943 2944The only known fix: Don't run display-time. 2945 2946*** SunOS 2947 2948SunOS 4.1.4 stopped shipping on Sep 30 1998. 2949 2950**** SunOS: You get linker errors 2951 ld: Undefined symbol 2952 _get_wmShellWidgetClass 2953 _get_applicationShellWidgetClass 2954 2955**** Sun 4.0.x: M-x shell persistently reports "Process shell exited abnormally with code 1". 2956 2957This happened on Suns as a result of what is said to be a bug in Sunos 2958version 4.0.x. The only fix was to reboot the machine. 2959 2960**** SunOS4.1.1 and SunOS4.1.3: Mail is lost when sent to local aliases. 2961 2962Many emacs mail user agents (VM and rmail, for instance) use the 2963sendmail.el library. This library can arrange for mail to be 2964delivered by passing messages to the /usr/lib/sendmail (usually) 2965program . In doing so, it passes the '-t' flag to sendmail, which 2966means that the name of the recipient of the message is not on the 2967command line and, therefore, that sendmail must parse the message to 2968obtain the destination address. 2969 2970There is a bug in the SunOS4.1.1 and SunOS4.1.3 versions of sendmail. 2971In short, when given the -t flag, the SunOS sendmail won't recognize 2972non-local (i.e. NIS) aliases. It has been reported that the Solaris 29732.x versions of sendmail do not have this bug. For those using SunOS 29744.1, the best fix is to install sendmail V8 or IDA sendmail (which 2975have other advantages over the regular sendmail as well). At the time 2976of this writing, these official versions are available: 2977 2978 Sendmail V8 on ftp.cs.berkeley.edu in /ucb/sendmail: 2979 sendmail.8.6.9.base.tar.Z (the base system source & documentation) 2980 sendmail.8.6.9.cf.tar.Z (configuration files) 2981 sendmail.8.6.9.misc.tar.Z (miscellaneous support programs) 2982 sendmail.8.6.9.xdoc.tar.Z (extended documentation, with postscript) 2983 2984 IDA sendmail on vixen.cso.uiuc.edu in /pub: 2985 sendmail-5.67b+IDA-1.5.tar.gz 2986 2987**** Sunos 4: You get the error ld: Undefined symbol __lib_version. 2988 2989This is the result of using cc or gcc with the shared library meant 2990for acc (the Sunpro compiler). Check your LD_LIBRARY_PATH and delete 2991/usr/lang/SC2.0.1 or some similar directory. 2992 2993**** SunOS 4.1.3: Emacs unpredictably crashes in _yp_dobind_soft. 2994 2995This happens if you configure Emacs specifying just `sparc-sun-sunos4' 2996on a system that is version 4.1.3. You must specify the precise 2997version number (or let configure figure out the configuration, which 2998it can do perfectly well for SunOS). 2999 3000**** Sunos 4.1.3: Emacs gets hung shortly after startup. 3001 3002We think this is due to a bug in Sunos. The word is that 3003one of these Sunos patches fixes the bug: 3004 3005100075-11 100224-06 100347-03 100482-05 100557-02 100623-03 100804-03 101080-01 3006100103-12 100249-09 100496-02 100564-07 100630-02 100891-10 101134-01 3007100170-09 100296-04 100377-09 100507-04 100567-04 100650-02 101070-01 101145-01 3008100173-10 100305-15 100383-06 100513-04 100570-05 100689-01 101071-03 101200-02 3009100178-09 100338-05 100421-03 100536-02 100584-05 100784-01 101072-01 101207-01 3010 3011We don't know which of these patches really matter. If you find out 3012which ones, please inform bug-gnu-emacs@gnu.org. 3013 3014**** SunOS 4: Emacs processes keep going after you kill the X server 3015(or log out, if you logged in using X). 3016 3017Someone reported that recompiling with GCC 2.7.0 fixed this problem. 3018 3019The fix to this is to install patch 100573 for OpenWindows 3.0 3020or link libXmu statically. 3021 3022**** Sunos 5.3: Subprocesses remain, hanging but not zombies. 3023 3024A bug in Sunos 5.3 causes Emacs subprocesses to remain after Emacs 3025exits. Sun patch # 101415-02 is part of the fix for this, but it only 3026applies to ptys, and doesn't fix the problem with subprocesses 3027communicating through pipes. 3028 3029*** Apollo Domain 3030 3031**** Shell mode ignores interrupts on Apollo Domain. 3032 3033You may find that M-x shell prints the following message: 3034 3035 Warning: no access to tty; thus no job control in this shell... 3036 3037This can happen if there are not enough ptys on your system. 3038Here is how to make more of them. 3039 3040 % cd /dev 3041 % ls pty* 3042 # shows how many pty's you have. I had 8, named pty0 to pty7) 3043 % /etc/crpty 8 3044 # creates eight new pty's 3045 3046*** Irix 3047 3048*** Irix 6.2: No visible display on mips-sgi-irix6.2 when compiling with GCC 2.8.1. 3049 3050This problem went away after installing the latest IRIX patches 3051as of 8 Dec 1998. 3052 3053The same problem has been reported on Irix 6.3. 3054 3055*** Irix 6.3: substituting environment variables in file names 3056in the minibuffer gives peculiar error messages such as 3057 3058 Substituting nonexistent environment variable "" 3059 3060This is not an Emacs bug; it is caused by something in SGI patch 3061003082 August 11, 1998. 3062 3063*** OPENSTEP 3064 3065**** OPENSTEP 4.2: Compiling syntax.c with gcc 2.7.2.1 fails. 3066 3067The compiler was reported to crash while compiling syntax.c with the 3068following message: 3069 3070 cc: Internal compiler error: program cc1obj got fatal signal 11 3071 3072To work around this, replace the macros UPDATE_SYNTAX_TABLE_FORWARD, 3073INC_BOTH, and INC_FROM with functions. To this end, first define 3 3074functions, one each for every macro. Here's an example: 3075 3076 static int update_syntax_table_forward(int from) 3077 { 3078 return(UPDATE_SYNTAX_TABLE_FORWARD(from)); 3079 }/*update_syntax_table_forward*/ 3080 3081Then replace all references to UPDATE_SYNTAX_TABLE_FORWARD in syntax.c 3082with a call to the function update_syntax_table_forward. 3083 3084*** Solaris 2.x 3085 3086**** Strange results from format %d in a few cases, on a Sun. 3087 3088Sun compiler version SC3.0 has been found to miscompile part of 3089editfns.c. The workaround is to compile with some other compiler such 3090as GCC. 3091 3092**** On Solaris, Emacs dumps core if lisp-complete-symbol is called. 3093 3094If you compile Emacs with the -fast or -xO4 option with version 3.0.2 3095of the Sun C compiler, Emacs dumps core when lisp-complete-symbol is 3096called. The problem does not happen if you compile with GCC. 3097 3098**** On Solaris, Emacs crashes if you use (display-time). 3099 3100This can happen if you configure Emacs without specifying the precise 3101version of Solaris that you are using. 3102 3103**** Solaris 2.3 and 2.4: Unpredictable segmentation faults. 3104 3105A user reported that this happened in 19.29 when it was compiled with 3106the Sun compiler, but not when he recompiled with GCC 2.7.0. 3107 3108We do not know whether something in Emacs is partly to blame for this. 3109 3110**** Solaris 2.4: Emacs dumps core on startup. 3111 3112Bill Sebok says that the cause of this is Solaris 2.4 vendor patch 3113102303-05, which extends the Solaris linker to deal with the Solaris 3114Common Desktop Environment's linking needs. You can fix the problem 3115by removing this patch and installing patch 102049-02 instead. 3116However, that linker version won't work with CDE. 3117 3118Solaris 2.5 comes with a linker that has this bug. It is reported that if 3119you install all the latest patches (as of June 1996), the bug is fixed. 3120We suspect the crucial patch is one of these, but we don't know 3121for certain. 3122 3123 103093-03: [README] SunOS 5.5: kernel patch (2140557 bytes) 3124 102832-01: [README] OpenWindows 3.5: Xview Jumbo Patch (4181613 bytes) 3125 103242-04: [README] SunOS 5.5: linker patch (595363 bytes) 3126 3127(One user reports that the bug was fixed by those patches together 3128with patches 102980-04, 103279-01, 103300-02, and 103468-01.) 3129 3130If you can determine which patch does fix the bug, please tell 3131bug-gnu-emacs@gnu.org. 3132 3133Meanwhile, the GNU linker links Emacs properly on both Solaris 2.4 and 3134Solaris 2.5. 3135 3136**** Solaris 2.4: Dired hangs and C-g does not work. Or Emacs hangs 3137forever waiting for termination of a subprocess that is a zombie. 3138 3139casper@fwi.uva.nl says the problem is in X11R6. Rebuild libX11.so 3140after changing the file xc/config/cf/sunLib.tmpl. Change the lines 3141 3142 #if ThreadedX 3143 #define SharedX11Reqs -lthread 3144 #endif 3145 3146to: 3147 3148 #if OSMinorVersion < 4 3149 #if ThreadedX 3150 #define SharedX11Reqs -lthread 3151 #endif 3152 #endif 3153 3154Be sure also to edit x/config/cf/sun.cf so that OSMinorVersion is 4 3155(as it should be for Solaris 2.4). The file has three definitions for 3156OSMinorVersion: the first is for x86, the second for SPARC under 3157Solaris, and the third for SunOS 4. Make sure to update the 3158definition for your type of machine and system. 3159 3160Then do `make Everything' in the top directory of X11R6, to rebuild 3161the makefiles and rebuild X. The X built this way work only on 3162Solaris 2.4, not on 2.3. 3163 3164For multithreaded X to work it is necessary to install patch 3165101925-02 to fix problems in header files [2.4]. You need 3166to reinstall gcc or re-run just-fixinc after installing that 3167patch. 3168 3169However, Frank Rust <frust@iti.cs.tu-bs.de> used a simpler solution: 3170he changed 3171 #define ThreadedX YES 3172to 3173 #define ThreadedX NO 3174in sun.cf and did `make World' to rebuild X11R6. Removing all 3175`-DXTHREAD*' flags and `-lthread' entries from lib/X11/Makefile and 3176typing 'make install' in that directory also seemed to work. 3177 3178**** Solaris 2.x: GCC complains "64 bit integer types not supported". 3179 3180This suggests that GCC is not installed correctly. Most likely you 3181are using GCC 2.7.2.3 (or earlier) on Solaris 2.6 (or later); this 3182does not work without patching. To run GCC 2.7.2.3 on Solaris 2.6 or 3183later, you must patch fixinc.svr4 and reinstall GCC from scratch as 3184described in the Solaris FAQ 3185<http://www.wins.uva.nl/pub/solaris/solaris2.html>. A better fix is 3186to upgrade to GCC 2.8.1 or later. 3187 3188**** Solaris 2.7: Building Emacs with WorkShop Compilers 5.0 98/12/15 3189C 5.0 failed, apparently with non-default CFLAGS, most probably due to 3190compiler bugs. Using Sun Solaris 2.7 Sun WorkShop 6 update 1 C 3191release was reported to work without problems. It worked OK on 3192another system with Solaris 8 using apparently the same 5.0 compiler 3193and the default CFLAGS. 3194 3195**** Solaris 2.x: Emacs dumps core when built with Motif. 3196 3197The Solaris Motif libraries are buggy, at least up through Solaris 2.5.1. 3198Install the current Motif runtime library patch appropriate for your host. 3199(Make sure the patch is current; some older patch versions still have the bug.) 3200You should install the other patches recommended by Sun for your host, too. 3201You can obtain Sun patches from ftp://sunsolve.sun.com/pub/patches/; 3202look for files with names ending in `.PatchReport' to see which patches 3203are currently recommended for your host. 3204 3205On Solaris 2.6, Emacs is said to work with Motif when Solaris patch 3206105284-12 is installed, but fail when 105284-15 is installed. 3207105284-18 might fix it again. 3208 3209**** Solaris 2.6 and 7: the Compose key does not work. 3210 3211This is a bug in Motif in Solaris. Supposedly it has been fixed for 3212the next major release of Solaris. However, if someone with Sun 3213support complains to Sun about the bug, they may release a patch. 3214If you do this, mention Sun bug #4188711. 3215 3216One workaround is to use a locale that allows non-ASCII characters. 3217For example, before invoking emacs, set the LC_ALL environment 3218variable to "en_US" (American English). The directory /usr/lib/locale 3219lists the supported locales; any locale other than "C" or "POSIX" 3220should do. 3221 3222pen@lysator.liu.se says (Feb 1998) that the Compose key does work 3223if you link with the MIT X11 libraries instead of the Solaris X11 3224libraries. 3225 3226*** HP/UX versions before 11.0 3227 3228HP/UX 9 was end-of-lifed in December 1998. 3229HP/UX 10 was end-of-lifed in May 1999. 3230 3231**** HP/UX 9: Emacs crashes with SIGBUS or SIGSEGV after you delete a frame. 3232 3233We think this is due to a bug in the X libraries provided by HP. With 3234the alternative X libraries in /usr/contrib/mitX11R5/lib, the problem 3235does not happen. 3236 3237*** HP/UX 10: Large file support is disabled. 3238 3239See the comments in src/s/hpux10.h. 3240 3241*** HP/UX: Emacs is slow using X11R5. 3242 3243This happens if you use the MIT versions of the X libraries--it 3244doesn't run as fast as HP's version. People sometimes use the version 3245because they see the HP version doesn't have the libraries libXaw.a, 3246libXmu.a, libXext.a and others. HP/UX normally doesn't come with 3247those libraries installed. To get good performance, you need to 3248install them and rebuild Emacs. 3249 3250*** Ultrix and Digital Unix 3251 3252**** Ultrix 4.2: `make install' fails on install-doc with `Error 141'. 3253 3254This happens on Ultrix 4.2 due to failure of a pipeline of tar 3255commands. We don't know why they fail, but the bug seems not to be in 3256Emacs. The workaround is to run the shell command in install-doc by 3257hand. 3258 3259**** Digital Unix 4.0: Garbled display on non-X terminals when Emacs runs. 3260 3261So far it appears that running `tset' triggers this problem (when TERM 3262is vt100, at least). If you do not run `tset', then Emacs displays 3263properly. If someone can tell us precisely which effect of running 3264`tset' actually causes the problem, we may be able to implement a fix 3265in Emacs. 3266 3267**** Ultrix: `expand-file-name' fails to work on any but the machine you dumped Emacs on. 3268 3269On Ultrix, if you use any of the functions which look up information 3270in the passwd database before dumping Emacs (say, by using 3271expand-file-name in site-init.el), then those functions will not work 3272in the dumped Emacs on any host but the one Emacs was dumped on. 3273 3274The solution? Don't use expand-file-name in site-init.el, or in 3275anything it loads. Yuck - some solution. 3276 3277I'm not sure why this happens; if you can find out exactly what is 3278going on, and perhaps find a fix or a workaround, please let us know. 3279Perhaps the YP functions cache some information, the cache is included 3280in the dumped Emacs, and is then inaccurate on any other host. 3281 3282*** SVr4 3283 3284**** SVr4: On some variants of SVR4, Emacs does not work at all with X. 3285 3286Try defining BROKEN_FIONREAD in your config.h file. If this solves 3287the problem, please send a bug report to tell us this is needed; be 3288sure to say exactly what type of machine and system you are using. 3289 3290**** SVr4: After running emacs once, subsequent invocations crash. 3291 3292Some versions of SVR4 have a serious bug in the implementation of the 3293mmap () system call in the kernel; this causes emacs to run correctly 3294the first time, and then crash when run a second time. 3295 3296Contact your vendor and ask for the mmap bug fix; in the mean time, 3297you may be able to work around the problem by adding a line to your 3298operating system description file (whose name is reported by the 3299configure script) that reads: 3300#define SYSTEM_MALLOC 3301This makes Emacs use memory less efficiently, but seems to work around 3302the kernel bug. 3303 3304*** Irix 5 and earlier 3305 3306Exactly when Irix-5 end-of-lifed is obscure. But since Irix 6.0 3307shipped in 1994, it has been some years. 3308 3309**** Irix 5.2: unexelfsgi.c can't find cmplrs/stsupport.h. 3310 3311The file cmplrs/stsupport.h was included in the wrong file set in the 3312Irix 5.2 distribution. You can find it in the optional fileset 3313compiler_dev, or copy it from some other Irix 5.2 system. A kludgy 3314workaround is to change unexelfsgi.c to include sym.h instead of 3315syms.h. 3316 3317**** Irix 5.3: "out of virtual swap space". 3318 3319This message occurs when the system runs out of swap space due to too 3320many large programs running. The solution is either to provide more 3321swap space or to reduce the number of large programs being run. You 3322can check the current status of the swap space by executing the 3323command `swap -l'. 3324 3325You can increase swap space by changing the file /etc/fstab. Adding a 3326line like this: 3327 3328/usr/swap/swap.more swap swap pri=3 0 0 3329 3330where /usr/swap/swap.more is a file previously created (for instance 3331by using /etc/mkfile), will increase the swap space by the size of 3332that file. Execute `swap -m' or reboot the machine to activate the 3333new swap area. See the manpages for `swap' and `fstab' for further 3334information. 3335 3336The objectserver daemon can use up lots of memory because it can be 3337swamped with NIS information. It collects information about all users 3338on the network that can log on to the host. 3339 3340If you want to disable the objectserver completely, you can execute 3341the command `chkconfig objectserver off' and reboot. That may disable 3342some of the window system functionality, such as responding CDROM 3343icons. 3344 3345You can also remove NIS support from the objectserver. The SGI `admin' 3346FAQ has a detailed description on how to do that; see question 35 3347("Why isn't the objectserver working?"). The admin FAQ can be found at 3348ftp://viz.tamu.edu/pub/sgi/faq/. 3349 3350**** Irix 5.3: Emacs crashes in utmpname. 3351 3352This problem is fixed in Patch 3175 for Irix 5.3. 3353It is also fixed in Irix versions 6.2 and up. 3354 3355**** Irix 6.0: Make tries (and fails) to build a program named unexelfsgi. 3356 3357A compiler bug inserts spaces into the string "unexelfsgi . o" 3358in src/Makefile. Edit src/Makefile, after configure is run, 3359find that string, and take out the spaces. 3360 3361Compiler fixes in Irix 6.0.1 should eliminate this problem. 3362 3363*** SCO Unix and UnixWare 3364 3365**** SCO 3.2v4: Unusable default font. 3366 3367The Open Desktop environment comes with default X resource settings 3368that tell Emacs to use a variable-width font. Emacs cannot use such 3369fonts, so it does not work. 3370 3371This is caused by the file /usr/lib/X11/app-defaults/ScoTerm, which is 3372the application-specific resource file for the `scoterm' terminal 3373emulator program. It contains several extremely general X resources 3374that affect other programs besides `scoterm'. In particular, these 3375resources affect Emacs also: 3376 3377 *Font: -*-helvetica-medium-r-*--12-*-p-* 3378 *Background: scoBackground 3379 *Foreground: scoForeground 3380 3381The best solution is to create an application-specific resource file for 3382Emacs, /usr/lib/X11/sco/startup/Emacs, with the following contents: 3383 3384 Emacs*Font: -*-courier-medium-r-*-*-*-120-*-*-*-*-iso8859-1 3385 Emacs*Background: white 3386 Emacs*Foreground: black 3387 3388(These settings mimic the Emacs defaults, but you can change them to 3389suit your needs.) This resource file is only read when the X server 3390starts up, so you should restart it by logging out of the Open Desktop 3391environment or by running `scologin stop; scologin start` from the shell 3392as root. Alternatively, you can put these settings in the 3393/usr/lib/X11/app-defaults/Emacs resource file and simply restart Emacs, 3394but then they will not affect remote invocations of Emacs that use the 3395Open Desktop display. 3396 3397These resource files are not normally shared across a network of SCO 3398machines; you must create the file on each machine individually. 3399 3400**** SCO 4.2.0: Regular expressions matching bugs on SCO systems. 3401 3402On SCO, there are problems in regexp matching when Emacs is compiled 3403with the system compiler. The compiler version is "Microsoft C 3404version 6", SCO 4.2.0h Dev Sys Maintenance Supplement 01/06/93; Quick 3405C Compiler Version 1.00.46 (Beta). The solution is to compile with 3406GCC. 3407 3408**** UnixWare 2.1: Error 12 (virtual memory exceeded) when dumping Emacs. 3409 3410Paul Abrahams (abrahams@acm.org) reports that with the installed 3411virtual memory settings for UnixWare 2.1.2, an Error 12 occurs during 3412the "make" that builds Emacs, when running temacs to dump emacs. That 3413error indicates that the per-process virtual memory limit has been 3414exceeded. The default limit is probably 32MB. Raising the virtual 3415memory limit to 40MB should make it possible to finish building Emacs. 3416 3417You can do this with the command `ulimit' (sh) or `limit' (csh). 3418But you have to be root to do it. 3419 3420According to Martin Sohnius, you can also retune this in the kernel: 3421 3422 # /etc/conf/bin/idtune SDATLIM 33554432 ## soft data size limit 3423 # /etc/conf/bin/idtune HDATLIM 33554432 ## hard " 3424 # /etc/conf/bin/idtune SVMMSIZE unlimited ## soft process size limit 3425 # /etc/conf/bin/idtune HVMMSIZE unlimited ## hard " 3426 # /etc/conf/bin/idbuild -B 3427 3428(He recommends you not change the stack limit, though.) 3429These changes take effect when you reboot. 3430 3431*** Linux 1.x 3432 3433**** Linux 1.0-1.04: Typing C-c C-c in Shell mode kills your X server. 3434 3435This happens with Linux kernel 1.0 thru 1.04, approximately. The workaround is 3436to define SIGNALS_VIA_CHARACTERS in config.h and recompile Emacs. 3437Newer Linux kernel versions don't have this problem. 3438 3439**** Linux 1.3: Output from subprocess (such as man or diff) is randomly 3440truncated on GNU/Linux systems. 3441 3442This is due to a kernel bug which seems to be fixed in Linux version 34431.3.75. 3444 3445** Windows 3.1, 95, 98, and ME 3446 3447*** MS-Windows NT/95: Problems running Perl under Emacs 3448 3449`perl -de 0' just hangs when executed in an Emacs subshell. 3450The fault lies with Perl (indirectly with Windows NT/95). 3451 3452The problem is that the Perl debugger explicitly opens a connection to 3453"CON", which is the DOS/NT equivalent of "/dev/tty", for interacting 3454with the user. 3455 3456On Unix, this is okay, because Emacs (or the shell?) creates a 3457pseudo-tty so that /dev/tty is really the pipe Emacs is using to 3458communicate with the subprocess. 3459 3460On NT, this fails because CON always refers to the handle for the 3461relevant console (approximately equivalent to a tty), and cannot be 3462redirected to refer to the pipe Emacs assigned to the subprocess as 3463stdin. 3464 3465A workaround is to modify perldb.pl to use STDIN/STDOUT instead of CON. 3466 3467For Perl 4: 3468 3469 *** PERL/LIB/PERLDB.PL.orig Wed May 26 08:24:18 1993 3470 --- PERL/LIB/PERLDB.PL Mon Jul 01 15:28:16 1996 3471 *************** 3472 *** 68,74 **** 3473 $rcfile=".perldb"; 3474 } 3475 else { 3476 ! $console = "con"; 3477 $rcfile="perldb.ini"; 3478 } 3479 3480 --- 68,74 ---- 3481 $rcfile=".perldb"; 3482 } 3483 else { 3484 ! $console = ""; 3485 $rcfile="perldb.ini"; 3486 } 3487 3488 3489 For Perl 5: 3490 *** perl/5.001/lib/perl5db.pl.orig Sun Jun 04 21:13:40 1995 3491 --- perl/5.001/lib/perl5db.pl Mon Jul 01 17:00:08 1996 3492 *************** 3493 *** 22,28 **** 3494 $rcfile=".perldb"; 3495 } 3496 elsif (-e "con") { 3497 ! $console = "con"; 3498 $rcfile="perldb.ini"; 3499 } 3500 else { 3501 --- 22,28 ---- 3502 $rcfile=".perldb"; 3503 } 3504 elsif (-e "con") { 3505 ! $console = ""; 3506 $rcfile="perldb.ini"; 3507 } 3508 else { 3509 3510*** MS-Windows 95: Alt-f6 does not get through to Emacs. 3511 3512This character seems to be trapped by the kernel in Windows 95. 3513You can enter M-f6 by typing ESC f6. 3514 3515*** MS-Windows 95/98/ME: subprocesses do not terminate properly. 3516 3517This is a limitation of the Operating System, and can cause problems 3518when shutting down Windows. Ensure that all subprocesses are exited 3519cleanly before exiting Emacs. For more details, see the FAQ at 3520http://www.gnu.org/software/emacs/windows/. 3521 3522*** MS-Windows 95/98/ME: crashes when Emacs invokes non-existent programs. 3523 3524When a program you are trying to run is not found on the PATH, 3525Windows might respond by crashing or locking up your system. In 3526particular, this has been reported when trying to compile a Java 3527program in JDEE when javac.exe is installed, but not on the system 3528PATH. 3529 3530** MS-DOS 3531 3532*** When compiling with DJGPP on MS-Windows NT, "config msdos" fails. 3533 3534If the error message is "VDM has been already loaded", this is because 3535Windows has a program called `redir.exe' that is incompatible with a 3536program by the same name supplied with DJGPP, which is used by 3537config.bat. To resolve this, move the DJGPP's `bin' subdirectory to 3538the front of your PATH environment variable. 3539 3540*** When compiling with DJGPP on MS-Windows 95, Make fails for some targets 3541like make-docfile. 3542 3543This can happen if long file name support (the setting of environment 3544variable LFN) when Emacs distribution was unpacked and during 3545compilation are not the same. See the MSDOG section of INSTALL for 3546the explanation of how to avoid this problem. 3547 3548*** Emacs compiled with DJGPP complains at startup: 3549 3550 "Wrong type of argument: internal-facep, msdos-menu-active-face" 3551 3552This can happen if you define an environment variable `TERM'. Emacs 3553on MSDOS uses an internal terminal emulator which is disabled if the 3554value of `TERM' is anything but the string "internal". Emacs then 3555works as if its terminal were a dumb glass teletype that doesn't 3556support faces. To work around this, arrange for `TERM' to be 3557undefined when Emacs runs. The best way to do that is to add an 3558[emacs] section to the DJGPP.ENV file which defines an empty value for 3559`TERM'; this way, only Emacs gets the empty value, while the rest of 3560your system works as before. 3561 3562*** MS-DOS: Emacs crashes at startup. 3563 3564Some users report that Emacs 19.29 requires dpmi memory management, 3565and crashes on startup if the system does not have it. We don't yet 3566know why this happens--perhaps these machines don't have enough real 3567memory, or perhaps something is wrong in Emacs or the compiler. 3568However, arranging to use dpmi support is a workaround. 3569 3570You can find out if you have a dpmi host by running go32 without 3571arguments; it will tell you if it uses dpmi memory. For more 3572information about dpmi memory, consult the djgpp FAQ. (djgpp 3573is the GNU C compiler as packaged for MSDOS.) 3574 3575Compiling Emacs under MSDOS is extremely sensitive for proper memory 3576configuration. If you experience problems during compilation, consider 3577removing some or all memory resident programs (notably disk caches) 3578and make sure that your memory managers are properly configured. See 3579the djgpp faq for configuration hints. 3580 3581*** Emacs compiled with DJGPP for MS-DOS/MS-Windows cannot access files 3582in the directory with the special name `dev' under the root of any 3583drive, e.g. `c:/dev'. 3584 3585This is an unfortunate side-effect of the support for Unix-style 3586device names such as /dev/null in the DJGPP runtime library. A 3587work-around is to rename the problem directory to another name. 3588 3589*** MS-DOS+DJGPP: Problems on MS-DOG if DJGPP v2.0 is used to compile Emacs. 3590 3591There are two DJGPP library bugs which cause problems: 3592 3593 * Running `shell-command' (or `compile', or `grep') you get 3594 `Searching for program: permission denied (EACCES), c:/command.com'; 3595 * After you shell to DOS, Ctrl-Break kills Emacs. 3596 3597To work around these bugs, you can use two files in the msdos 3598subdirectory: `is_exec.c' and `sigaction.c'. Compile them and link 3599them into the Emacs executable `temacs'; then they will replace the 3600incorrect library functions. 3601 3602*** MS-DOS: Emacs compiled for MSDOS cannot find some Lisp files, or other 3603run-time support files, when long filename support is enabled. 3604 3605Usually, this problem will manifest itself when Emacs exits 3606immediately after flashing the startup screen, because it cannot find 3607the Lisp files it needs to load at startup. Redirect Emacs stdout 3608and stderr to a file to see the error message printed by Emacs. 3609 3610Another manifestation of this problem is that Emacs is unable to load 3611the support for editing program sources in languages such as C and 3612Lisp. 3613 3614This can happen if the Emacs distribution was unzipped without LFN 3615support, thus causing long filenames to be truncated to the first 6 3616characters and a numeric tail that Windows 95 normally attaches to it. 3617You should unzip the files again with a utility that supports long 3618filenames (such as djtar from DJGPP or InfoZip's UnZip program 3619compiled with DJGPP v2). The MSDOG section of the file INSTALL 3620explains this issue in more detail. 3621 3622Another possible reason for such failures is that Emacs compiled for 3623MSDOS is used on Windows NT, where long file names are not supported 3624by this version of Emacs, but the distribution was unpacked by an 3625unzip program that preserved the long file names instead of truncating 3626them to DOS 8+3 limits. To be useful on NT, the MSDOS port of Emacs 3627must be unzipped by a DOS utility, so that long file names are 3628properly truncated. 3629 3630** Archaic window managers and toolkits 3631 3632*** OpenLook: Under OpenLook, the Emacs window disappears when you type M-q. 3633 3634Some versions of the Open Look window manager interpret M-q as a quit 3635command for whatever window you are typing at. If you want to use 3636Emacs with that window manager, you should try to configure the window 3637manager to use some other command. You can disable the 3638shortcut keys entirely by adding this line to ~/.OWdefaults: 3639 3640 OpenWindows.WindowMenuAccelerators: False 3641 3642**** twm: A position you specified in .Xdefaults is ignored, using twm. 3643 3644twm normally ignores "program-specified" positions. 3645You can tell it to obey them with this command in your `.twmrc' file: 3646 3647 UsePPosition "on" #allow clients to request a position 3648 3649** Bugs related to old DEC hardware 3650 3651*** The Compose key on a DEC keyboard does not work as Meta key. 3652 3653This shell command should fix it: 3654 3655 xmodmap -e 'keycode 0xb1 = Meta_L' 3656 3657*** Keyboard input gets confused after a beep when using a DECserver 3658as a concentrator. 3659 3660This problem seems to be a matter of configuring the DECserver to use 36617 bit characters rather than 8 bit characters. 3662 3663* Build problems on legacy systems 3664 3665** BSD/386 1.0: --with-x-toolkit option configures wrong. 3666 3667This problem is due to bugs in the shell in version 1.0 of BSD/386. 3668The workaround is to edit the configure file to use some other shell, 3669such as bash. 3670 3671** Digital Unix 4.0: Emacs fails to build, giving error message 3672 Invalid dimension for the charset-ID 160 3673 3674This is due to a bug or an installation problem in GCC 2.8.0. 3675Installing a more recent version of GCC fixes the problem. 3676 3677** Digital Unix 4.0: Failure in unexec while dumping emacs. 3678 3679This problem manifests itself as an error message 3680 3681 unexec: Bad address, writing data section to ... 3682 3683The user suspects that this happened because his X libraries 3684were built for an older system version, 3685 3686 ./configure --x-includes=/usr/include --x-libraries=/usr/shlib 3687 3688made the problem go away. 3689 3690** Sunos 4.1.1: there are errors compiling sysdep.c. 3691 3692If you get errors such as 3693 3694 "sysdep.c", line 2017: undefined structure or union 3695 "sysdep.c", line 2017: undefined structure or union 3696 "sysdep.c", line 2019: nodename undefined 3697 3698This can result from defining LD_LIBRARY_PATH. It is very tricky 3699to use that environment variable with Emacs. The Emacs configure 3700script links many test programs with the system libraries; you must 3701make sure that the libraries available to configure are the same 3702ones available when you build Emacs. 3703 3704** SunOS 4.1.1: You get this error message from GNU ld: 3705 3706 /lib/libc.a(_Q_sub.o): Undefined symbol __Q_get_rp_rd referenced from text segment 3707 3708The problem is in the Sun shared C library, not in GNU ld. 3709 3710The solution is to install Patch-ID# 100267-03 from Sun. 3711 3712** Sunos 4.1: Undefined symbols when linking using --with-x-toolkit. 3713 3714If you get the undefined symbols _atowc _wcslen, _iswprint, _iswspace, 3715_iswcntrl, _wcscpy, and _wcsncpy, then you need to add -lXwchar after 3716-lXaw in the command that links temacs. 3717 3718This problem seems to arise only when the international language 3719extensions to X11R5 are installed. 3720 3721** SunOS: Emacs gets error message from linker on Sun. 3722 3723If the error message says that a symbol such as `f68881_used' or 3724`ffpa_used' or `start_float' is undefined, this probably indicates 3725that you have compiled some libraries, such as the X libraries, 3726with a floating point option other than the default. 3727 3728It's not terribly hard to make this work with small changes in 3729crt0.c together with linking with Fcrt1.o, Wcrt1.o or Mcrt1.o. 3730However, the easiest approach is to build Xlib with the default 3731floating point option: -fsoft. 3732 3733** SunOS: Undefined symbols _dlopen, _dlsym and/or _dlclose. 3734 3735If you see undefined symbols _dlopen, _dlsym, or _dlclose when linking 3736with -lX11, compile and link against the file mit/util/misc/dlsym.c in 3737the MIT X11R5 distribution. Alternatively, link temacs using shared 3738libraries with s/sunos4shr.h. (This doesn't work if you use the X 3739toolkit.) 3740 3741If you get the additional error that the linker could not find 3742lib_version.o, try extracting it from X11/usr/lib/X11/libvim.a in 3743X11R4, then use it in the link. 3744 3745** SunOS4, DGUX 5.4.2: --with-x-toolkit version crashes when used with shared libraries. 3746 3747On some systems, including Sunos 4 and DGUX 5.4.2 and perhaps others, 3748unexec doesn't work properly with the shared library for the X 3749toolkit. You might be able to work around this by using a nonshared 3750libXt.a library. The real fix is to upgrade the various versions of 3751unexec and/or ralloc. We think this has been fixed on Sunos 4 3752and Solaris in version 19.29. 3753 3754** HPUX 10.20: Emacs crashes during dumping on the HPPA machine. 3755 3756This seems to be due to a GCC bug; it is fixed in GCC 2.8.1. 3757 3758** VMS: Compilation errors on VMS. 3759 3760You will get warnings when compiling on VMS because there are 3761variable names longer than 32 (or whatever it is) characters. 3762This is not an error. Ignore it. 3763 3764VAX C does not support #if defined(foo). Uses of this construct 3765were removed, but some may have crept back in. They must be rewritten. 3766 3767There is a bug in the C compiler which fails to sign extend characters 3768in conditional expressions. The bug is: 3769 char c = -1, d = 1; 3770 int i; 3771 3772 i = d ? c : d; 3773The result is i == 255; the fix is to typecast the char in the 3774conditional expression as an (int). Known occurrences of such 3775constructs in Emacs have been fixed. 3776 3777** Vax C compiler bugs affecting Emacs. 3778 3779You may get one of these problems compiling Emacs: 3780 3781 foo.c line nnn: compiler error: no table entry for op STASG 3782 foo.c: fatal error in /lib/ccom 3783 3784These are due to bugs in the C compiler; the code is valid C. 3785Unfortunately, the bugs are unpredictable: the same construct 3786may compile properly or trigger one of these bugs, depending 3787on what else is in the source file being compiled. Even changes 3788in header files that should not affect the file being compiled 3789can affect whether the bug happens. In addition, sometimes files 3790that compile correctly on one machine get this bug on another machine. 3791 3792As a result, it is hard for me to make sure this bug will not affect 3793you. I have attempted to find and alter these constructs, but more 3794can always appear. However, I can tell you how to deal with it if it 3795should happen. The bug comes from having an indexed reference to an 3796array of Lisp_Objects, as an argument in a function call: 3797 Lisp_Object *args; 3798 ... 3799 ... foo (5, args[i], ...)... 3800putting the argument into a temporary variable first, as in 3801 Lisp_Object *args; 3802 Lisp_Object tem; 3803 ... 3804 tem = args[i]; 3805 ... foo (r, tem, ...)... 3806causes the problem to go away. 3807The `contents' field of a Lisp vector is an array of Lisp_Objects, 3808so you may see the problem happening with indexed references to that. 3809 3810** 68000 C compiler problems 3811 3812Various 68000 compilers have different problems. 3813These are some that have been observed. 3814 3815*** Using value of assignment expression on union type loses. 3816This means that x = y = z; or foo (x = z); does not work 3817if x is of type Lisp_Object. 3818 3819*** "cannot reclaim" error. 3820 3821This means that an expression is too complicated. You get the correct 3822line number in the error message. The code must be rewritten with 3823simpler expressions. 3824 3825*** XCONS, XSTRING, etc macros produce incorrect code. 3826 3827If temacs fails to run at all, this may be the cause. 3828Compile this test program and look at the assembler code: 3829 3830struct foo { char x; unsigned int y : 24; }; 3831 3832lose (arg) 3833 struct foo arg; 3834{ 3835 test ((int *) arg.y); 3836} 3837 3838If the code is incorrect, your compiler has this problem. 3839In the XCONS, etc., macros in lisp.h you must replace (a).u.val with 3840((a).u.val + coercedummy) where coercedummy is declared as int. 3841 3842This problem will not happen if the m-...h file for your type 3843of machine defines NO_UNION_TYPE. That is the recommended setting now. 3844 3845*** C compilers lose on returning unions. 3846 3847I hear that some C compilers cannot handle returning a union type. 3848Most of the functions in GNU Emacs return type Lisp_Object, which is 3849defined as a union on some rare architectures. 3850 3851This problem will not happen if the m-...h file for your type 3852of machine defines NO_UNION_TYPE. 3853 3854 3855This file is part of GNU Emacs. 3856 3857GNU Emacs is free software; you can redistribute it and/or modify 3858it under the terms of the GNU General Public License as published by 3859the Free Software Foundation; either version 2, or (at your option) 3860any later version. 3861 3862GNU Emacs is distributed in the hope that it will be useful, 3863but WITHOUT ANY WARRANTY; without even the implied warranty of 3864MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 3865GNU General Public License for more details. 3866 3867You should have received a copy of the GNU General Public License 3868along with GNU Emacs; see the file COPYING. If not, write to the 3869Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, 3870Boston, MA 02110-1301, USA. 3871 3872 3873Local variables: 3874mode: outline 3875paragraph-separate: "[ ]*$" 3876end: 3877 3878arch-tag: 49fc0d95-88cb-4715-b21c-f27fb5a4764a 3879