#
[bpt/emacs.git] / man / trouble.texi
1 @c This is part of the Emacs manual.
2 @c Copyright (C) 1985, 86, 87, 93, 94, 95, 1997 Free Software Foundation, Inc.
3 @c See file emacs.texi for copying conditions.
4 @iftex
5 @chapter Dealing with Common Problems
6
7 If you type an Emacs command you did not intend, the results are often
8 mysterious. This chapter tells what you can do to cancel your mistake or
9 recover from a mysterious situation. Emacs bugs and system crashes are
10 also considered.
11 @end iftex
12
13 @node Quitting, Lossage, Customization, Top
14 @section Quitting and Aborting
15 @cindex quitting
16
17 @table @kbd
18 @item C-g
19 @itemx C-@key{BREAK} (MS-DOS)
20 Quit. Cancel running or partially typed command.
21 @item C-]
22 Abort innermost recursive editing level and cancel the command which
23 invoked it (@code{abort-recursive-edit}).
24 @item @key{ESC} @key{ESC} @key{ESC}
25 Either quit or abort, whichever makes sense (@code{keyboard-escape-quit}).
26 @item M-x top-level
27 Abort all recursive editing levels that are currently executing.
28 @item C-x u
29 Cancel a previously made change in the buffer contents (@code{undo}).
30 @end table
31
32 There are two ways of canceling commands which are not finished
33 executing: @dfn{quitting} with @kbd{C-g}, and @dfn{aborting} with
34 @kbd{C-]} or @kbd{M-x top-level}. Quitting cancels a partially typed
35 command or one which is already running. Aborting exits a recursive
36 editing level and cancels the command that invoked the recursive edit.
37 (@xref{Recursive Edit}.)
38
39 @cindex quitting
40 @kindex C-g
41 Quitting with @kbd{C-g} is used for getting rid of a partially typed
42 command, or a numeric argument that you don't want. It also stops a
43 running command in the middle in a relatively safe way, so you can use
44 it if you accidentally give a command which takes a long time. In
45 particular, it is safe to quit out of killing; either your text will
46 @emph{all} still be in the buffer, or it will @emph{all} be in the kill
47 ring (or maybe both). Quitting an incremental search does special
48 things documented under searching; in general, it may take two
49 successive @kbd{C-g} characters to get out of a search
50 (@pxref{Incremental Search}).
51
52 On MS-DOS, the character @kbd{C-@key{BREAK}} serves as a quit character
53 like @kbd{C-g}. The reason is that it is not feasible, on MS-DOS, to
54 recognize @kbd{C-g} while a command is running, between interactions
55 with the user. By contrast, it @emph{is} feasible to recognize
56 @kbd{C-@key{BREAK}} at all times. @xref{MS-DOS Input}.
57
58 @kbd{C-g} works by setting the variable @code{quit-flag} to @code{t}
59 the instant @kbd{C-g} is typed; Emacs Lisp checks this variable
60 frequently and quits if it is non-@code{nil}. @kbd{C-g} is only
61 actually executed as a command if you type it while Emacs is waiting for
62 input.
63
64 If you quit with @kbd{C-g} a second time before the first @kbd{C-g} is
65 recognized, you activate the ``emergency escape'' feature and return to
66 the shell. @xref{Emergency Escape}.
67
68 @cindex NFS and quitting
69 There may be times when you cannot quit. When Emacs is waiting for
70 the operating system to do something, quitting is impossible unless
71 special pains are taken for the particular system call within Emacs
72 where the waiting occurs. We have done this for the system calls that
73 users are likely to want to quit from, but it's possible you will find
74 another. In one very common case---waiting for file input or output
75 using NFS---Emacs itself knows how to quit, but most NFS implementations
76 simply do not allow user programs to stop waiting for NFS when the NFS
77 server is hung.
78
79 @cindex aborting recursive edit
80 @findex abort-recursive-edit
81 @kindex C-]
82 Aborting with @kbd{C-]} (@code{abort-recursive-edit}) is used to get
83 out of a recursive editing level and cancel the command which invoked
84 it. Quitting with @kbd{C-g} does not do this, and could not do this,
85 because it is used to cancel a partially typed command @emph{within} the
86 recursive editing level. Both operations are useful. For example, if
87 you are in a recursive edit and type @kbd{C-u 8} to enter a numeric
88 argument, you can cancel that argument with @kbd{C-g} and remain in the
89 recursive edit.
90
91 @findex keyboard-escape-quit
92 @kindex ESC ESC ESC
93 The command @kbd{@key{ESC} @key{ESC} @key{ESC}}
94 (@code{keyboard-escape-quit}) can either quit or abort. This key was
95 defined because @key{ESC} is used to ``get out'' in many PC programs.
96 It can cancel a prefix argument, clear a selected region, or get out of
97 a Query Replace, like @kbd{C-g}. It can get out of the minibuffer or a
98 recursive edit, like @kbd{C-]}. It can also get out of splitting the
99 frame into multiple windows, like @kbd{C-x 1}. One thing it cannot do,
100 however, is stop a command that is running. That's because it executes
101 as an ordinary command, and Emacs doesn't notice it until it is ready
102 for a command.
103
104 @findex top-level
105 The command @kbd{M-x top-level} is equivalent to ``enough'' @kbd{C-]}
106 commands to get you out of all the levels of recursive edits that you
107 are in. @kbd{C-]} gets you out one level at a time, but @kbd{M-x
108 top-level} goes out all levels at once. Both @kbd{C-]} and @kbd{M-x
109 top-level} are like all other commands, and unlike @kbd{C-g}, in that
110 they take effect only when Emacs is ready for a command. @kbd{C-]} is
111 an ordinary key and has its meaning only because of its binding in the
112 keymap. @xref{Recursive Edit}.
113
114 @kbd{C-x u} (@code{undo}) is not strictly speaking a way of canceling
115 a command, but you can think of it as canceling a command that already
116 finished executing. @xref{Undo}.
117
118 @node Lossage, Bugs, Quitting, Top
119 @section Dealing with Emacs Trouble
120
121 This section describes various conditions in which Emacs fails to work
122 normally, and how to recognize them and correct them.
123
124 @menu
125 * DEL Gets Help:: What to do if @key{DEL} doesn't delete.
126 * Stuck Recursive:: `[...]' in mode line around the parentheses.
127 * Screen Garbled:: Garbage on the screen.
128 * Text Garbled:: Garbage in the text.
129 * Unasked-for Search:: Spontaneous entry to incremental search.
130 * Memory Full:: How to cope when you run out of memory.
131 * After a Crash:: Recovering editing in an Emacs session that crashed.
132 * Emergency Escape:: Emergency escape---
133 What to do if Emacs stops responding.
134 * Total Frustration:: When you are at your wits' end.
135 @end menu
136
137 @node DEL Gets Help
138 @subsection If @key{DEL} Fails to Delete
139
140 If you find that @key{DEL} enters Help like @kbd{Control-h} instead of
141 deleting a character, your terminal is sending the wrong code for
142 @key{DEL}. You can work around this problem by changing the keyboard
143 translation table (@pxref{Keyboard Translations}).
144
145 @node Stuck Recursive
146 @subsection Recursive Editing Levels
147
148 Recursive editing levels are important and useful features of Emacs, but
149 they can seem like malfunctions to the user who does not understand them.
150
151 If the mode line has square brackets @samp{[@dots{}]} around the parentheses
152 that contain the names of the major and minor modes, you have entered a
153 recursive editing level. If you did not do this on purpose, or if you
154 don't understand what that means, you should just get out of the recursive
155 editing level. To do so, type @kbd{M-x top-level}. This is called getting
156 back to top level. @xref{Recursive Edit}.
157
158 @node Screen Garbled
159 @subsection Garbage on the Screen
160
161 If the data on the screen looks wrong, the first thing to do is see
162 whether the text is really wrong. Type @kbd{C-l}, to redisplay the
163 entire screen. If the screen appears correct after this, the problem
164 was entirely in the previous screen update. (Otherwise, see @ref{Text
165 Garbled}.)
166
167 Display updating problems often result from an incorrect termcap entry
168 for the terminal you are using. The file @file{etc/TERMS} in the Emacs
169 distribution gives the fixes for known problems of this sort.
170 @file{INSTALL} contains general advice for these problems in one of its
171 sections. Very likely there is simply insufficient padding for certain
172 display operations. To investigate the possibility that you have this sort
173 of problem, try Emacs on another terminal made by a different manufacturer.
174 If problems happen frequently on one kind of terminal but not another kind,
175 it is likely to be a bad termcap entry, though it could also be due to a
176 bug in Emacs that appears for terminals that have or that lack specific
177 features.
178
179 @node Text Garbled
180 @subsection Garbage in the Text
181
182 If @kbd{C-l} shows that the text is wrong, try undoing the changes to it
183 using @kbd{C-x u} until it gets back to a state you consider correct. Also
184 try @kbd{C-h l} to find out what command you typed to produce the observed
185 results.
186
187 If a large portion of text appears to be missing at the beginning or
188 end of the buffer, check for the word @samp{Narrow} in the mode line.
189 If it appears, the text you don't see is probably still present, but
190 temporarily off-limits. To make it accessible again, type @kbd{C-x n
191 w}. @xref{Narrowing}.
192
193 @node Unasked-for Search
194 @subsection Spontaneous Entry to Incremental Search
195
196 If Emacs spontaneously displays @samp{I-search:} at the bottom of the
197 screen, it means that the terminal is sending @kbd{C-s} and @kbd{C-q}
198 according to the poorly designed xon/xoff ``flow control'' protocol.
199
200 If this happens to you, your best recourse is to put the terminal in a
201 mode where it will not use flow control, or give it so much padding that
202 it will never send a @kbd{C-s}. (One way to increase the amount of
203 padding is to set the variable @code{baud-rate} to a larger value. Its
204 value is the terminal output speed, measured in the conventional units
205 of baud.)
206
207 @cindex flow control
208 @cindex xon-xoff
209 @findex enable-flow-control
210 If you don't succeed in turning off flow control, the next best thing
211 is to tell Emacs to cope with it. To do this, call the function
212 @code{enable-flow-control}.
213
214 @findex enable-flow-control-on
215 Typically there are particular terminal types with which you must use
216 flow control. You can conveniently ask for flow control on those
217 terminal types only, using @code{enable-flow-control-on}. For example,
218 if you find you must use flow control on VT-100 and H19 terminals, put
219 the following in your @file{.emacs} file:
220
221 @example
222 (enable-flow-control-on "vt100" "h19")
223 @end example
224
225 When flow control is enabled, you must type @kbd{C-\} to get the
226 effect of a @kbd{C-s}, and type @kbd{C-^} to get the effect of a
227 @kbd{C-q}. (These aliases work by means of keyboard translations; see
228 @ref{Keyboard Translations}.)
229
230 @node Memory Full
231 @subsection Running out of Memory
232 @cindex memory full
233 @cindex out of memory
234
235 If you get the error message @samp{Virtual memory exceeded}, save your
236 modified buffers with @kbd{C-x s}. This method of saving them has the
237 smallest need for additional memory. Emacs keeps a reserve of memory
238 which it makes available when this error happens; that should be enough
239 to enable @kbd{C-x s} to complete its work.
240
241 Once you have saved your modified buffers, you can exit this Emacs job
242 and start another, or you can use @kbd{M-x kill-some-buffers} to free
243 space in the current Emacs job. If you kill buffers containing a
244 substantial amount of text, you can safely go on editing. Emacs refills
245 its memory reserve automatically when it sees sufficient free space
246 available, in case you run out of memory another time.
247
248 Do not use @kbd{M-x buffer-menu} to save or kill buffers when you run
249 out of memory, because the buffer menu needs a fair amount memory
250 itself, and the reserve supply may not be enough.
251
252 @node After a Crash
253 @subsection Recovery After a Crash
254
255 If Emacs or the computer crashes, you can recover the files you were
256 editing at the time of the crash from their auto-save files. To do
257 this, start Emacs again and type the command @kbd{M-x recover-session}.
258
259 This command initially displays a buffer which lists interrupted
260 session files, each with its date. You must choose which session to
261 recover from. Typically the one you want is the most recent one. Move
262 point to the one you choose, and type @kbd{C-c C-c}.
263
264 Then @code{recover-session} asks about each of the files that you were
265 editing during that session; it asks whether to recover that file. If
266 you answer @kbd{y} for a file, it shows the dates of that file and its
267 auto-save file, then asks once again whether to recover that file. For
268 the second question, you must confirm with @kbd{yes}. If you do, Emacs
269 visits the file but gets the text from the auto-save file.
270
271 When @code{recover-session} is done, the files you've chosen to
272 recover are present in Emacs buffers. You should then save them. Only
273 this---saving them---updates the files themselves.
274
275 @node Emergency Escape
276 @subsection Emergency Escape
277
278 Because at times there have been bugs causing Emacs to loop without
279 checking @code{quit-flag}, a special feature causes Emacs to be suspended
280 immediately if you type a second @kbd{C-g} while the flag is already set,
281 so you can always get out of GNU Emacs. Normally Emacs recognizes and
282 clears @code{quit-flag} (and quits!) quickly enough to prevent this from
283 happening. (On MS-DOS and compatible systems, type @kbd{C-@key{BREAK}}
284 twice.)
285
286 When you resume Emacs after a suspension caused by multiple @kbd{C-g}, it
287 asks two questions before going back to what it had been doing:
288
289 @example
290 Auto-save? (y or n)
291 Abort (and dump core)? (y or n)
292 @end example
293
294 @noindent
295 Answer each one with @kbd{y} or @kbd{n} followed by @key{RET}.
296
297 Saying @kbd{y} to @samp{Auto-save?} causes immediate auto-saving of all
298 modified buffers in which auto-saving is enabled.
299
300 Saying @kbd{y} to @samp{Abort (and dump core)?} causes an illegal instruction to be
301 executed, dumping core. This is to enable a wizard to figure out why Emacs
302 was failing to quit in the first place. Execution does not continue
303 after a core dump. If you answer @kbd{n}, execution does continue. With
304 luck, GNU Emacs will ultimately check @code{quit-flag} and quit normally.
305 If not, and you type another @kbd{C-g}, it is suspended again.
306
307 If Emacs is not really hung, just slow, you may invoke the double
308 @kbd{C-g} feature without really meaning to. Then just resume and answer
309 @kbd{n} to both questions, and you will arrive at your former state.
310 Presumably the quit you requested will happen soon.
311
312 The double-@kbd{C-g} feature is turned off when Emacs is running under
313 the X Window System, since you can use the window manager to kill Emacs
314 or to create another window and run another program.
315
316 On MS-DOS and compatible systems, the emergency escape feature is
317 sometimes unavailable, even if you press @kbd{C-@key{BREAK}} twice, when
318 some system call (MS-DOS or BIOS) hangs, or when Emacs is stuck in a
319 very tight endless loop (in C code, @strong{not} in Lisp code).
320
321 @node Total Frustration
322 @subsection Help for Total Frustration
323 @cindex Eliza
324 @cindex doctor
325
326 If using Emacs (or something else) becomes terribly frustrating and none
327 of the techniques described above solve the problem, Emacs can still help
328 you.
329
330 First, if the Emacs you are using is not responding to commands, type
331 @kbd{C-g C-g} to get out of it and then start a new one.
332
333 @findex doctor
334 Second, type @kbd{M-x doctor @key{RET}}.
335
336 The doctor will help you feel better. Each time you say something to
337 the doctor, you must end it by typing @key{RET} @key{RET}. This lets
338 the doctor know you are finished.
339
340 @node Bugs, Contributing, Lossage, Top
341 @section Reporting Bugs
342
343 @cindex bugs
344 Sometimes you will encounter a bug in Emacs. Although we cannot
345 promise we can or will fix the bug, and we might not even agree that it
346 is a bug, we want to hear about problems you encounter. Often we agree
347 they are bugs and want to fix them.
348
349 To make it possible for us to fix a bug, you must report it. In order
350 to do so effectively, you must know when and how to do it.
351
352 @menu
353 * Criteria: Bug Criteria. Have you really found a bug?
354 * Understanding Bug Reporting:: How to report a bug effectively.
355 * Checklist:: Steps to follow for a good bug report.
356 * Sending Patches:: How to send a patch for GNU Emacs.
357 @end menu
358
359 @node Bug Criteria
360 @subsection When Is There a Bug
361
362 If Emacs executes an illegal instruction, or dies with an operating
363 system error message that indicates a problem in the program (as opposed to
364 something like ``disk full''), then it is certainly a bug.
365
366 If Emacs updates the display in a way that does not correspond to what is
367 in the buffer, then it is certainly a bug. If a command seems to do the
368 wrong thing but the problem corrects itself if you type @kbd{C-l}, it is a
369 case of incorrect display updating.
370
371 Taking forever to complete a command can be a bug, but you must make
372 certain that it was really Emacs's fault. Some commands simply take a
373 long time. Type @kbd{C-g} (@kbd{C-@key{BREAK}} on MS-DOS) and then @kbd{C-h l}
374 to see whether the input Emacs received was what you intended to type;
375 if the input was such that you @emph{know} it should have been processed
376 quickly, report a bug. If you don't know whether the command should
377 take a long time, find out by looking in the manual or by asking for
378 assistance.
379
380 If a command you are familiar with causes an Emacs error message in a
381 case where its usual definition ought to be reasonable, it is probably a
382 bug.
383
384 If a command does the wrong thing, that is a bug. But be sure you know
385 for certain what it ought to have done. If you aren't familiar with the
386 command, or don't know for certain how the command is supposed to work,
387 then it might actually be working right. Rather than jumping to
388 conclusions, show the problem to someone who knows for certain.
389
390 Finally, a command's intended definition may not be best for editing
391 with. This is a very important sort of problem, but it is also a matter of
392 judgment. Also, it is easy to come to such a conclusion out of ignorance
393 of some of the existing features. It is probably best not to complain
394 about such a problem until you have checked the documentation in the usual
395 ways, feel confident that you understand it, and know for certain that what
396 you want is not available. If you are not sure what the command is
397 supposed to do after a careful reading of the manual, check the index and
398 glossary for any terms that may be unclear.
399
400 If after careful rereading of the manual you still do not understand
401 what the command should do, that indicates a bug in the manual, which
402 you should report. The manual's job is to make everything clear to
403 people who are not Emacs experts---including you. It is just as
404 important to report documentation bugs as program bugs.
405
406 If the on-line documentation string of a function or variable disagrees
407 with the manual, one of them must be wrong; that is a bug.
408
409 @node Understanding Bug Reporting
410 @subsection Understanding Bug Reporting
411
412 @findex emacs-version
413 When you decide that there is a bug, it is important to report it and to
414 report it in a way which is useful. What is most useful is an exact
415 description of what commands you type, starting with the shell command to
416 run Emacs, until the problem happens.
417
418 The most important principle in reporting a bug is to report
419 @emph{facts}. Hypotheses and verbal descriptions are no substitute for
420 the detailed raw data. Reporting the facts is straightforward, but many
421 people strain to posit explanations and report them instead of the
422 facts. If the explanations are based on guesses about how Emacs is
423 implemented, they will be useless; meanwhile, lacking the facts, we will
424 have no real information about the bug.
425
426 For example, suppose that you type @kbd{C-x C-f /glorp/baz.ugh
427 @key{RET}}, visiting a file which (you know) happens to be rather large,
428 and Emacs displayed @samp{I feel pretty today}. The best way to report
429 the bug is with a sentence like the preceding one, because it gives all
430 the facts.
431
432 A bad way would be to assume that the problem is due to the size of
433 the file and say, ``I visited a large file, and Emacs displayed @samp{I
434 feel pretty today}.'' This is what we mean by ``guessing
435 explanations.'' The problem is just as likely to be due to the fact
436 that there is a @samp{z} in the file name. If this is so, then when we
437 got your report, we would try out the problem with some ``large file,''
438 probably with no @samp{z} in its name, and not see any problem. There
439 is no way in the world that we could guess that we should try visiting a
440 file with a @samp{z} in its name.
441
442 Alternatively, the problem might be due to the fact that the file starts
443 with exactly 25 spaces. For this reason, you should make sure that you
444 inform us of the exact contents of any file that is needed to reproduce the
445 bug. What if the problem only occurs when you have typed the @kbd{C-x C-a}
446 command previously? This is why we ask you to give the exact sequence of
447 characters you typed since starting the Emacs session.
448
449 You should not even say ``visit a file'' instead of @kbd{C-x C-f} unless
450 you @emph{know} that it makes no difference which visiting command is used.
451 Similarly, rather than saying ``if I have three characters on the line,''
452 say ``after I type @kbd{@key{RET} A B C @key{RET} C-p},'' if that is
453 the way you entered the text.@refill
454
455 So please don't guess any explanations when you report a bug. If you
456 want to actually @emph{debug} the problem, and report explanations that
457 are more than guesses, that is useful---but please include the facts as
458 well.
459
460 @node Checklist
461 @subsection Checklist for Bug Reports
462
463 @cindex reporting bugs
464 The best way to send a bug report is to mail it electronically to the
465 Emacs maintainers at @samp{bug-gnu-emacs@@gnu.org}. (If you
466 want to suggest a change as an improvement, use the same address.)
467
468 If you'd like to read the bug reports, you can find them on the
469 newsgroup @samp{gnu.emacs.bug}; keep in mind, however, that as a
470 spectator you should not criticize anything about what you see there.
471 The purpose of bug reports is to give information to the Emacs
472 maintainers. Spectators are welcome only as long as they do not
473 interfere with this. In particular, some bug reports contain large
474 amounts of data; spectators should not complain about this.
475
476 Please do not post bug reports using netnews; mail is more reliable
477 than netnews about reporting your correct address, which we may need in
478 order to ask you for more information.
479
480 If you can't send electronic mail, then mail the bug report on paper
481 or machine-readable media to this address:
482
483 @format
484 GNU Emacs Bugs
485 Free Software Foundation
486 59 Temple Place, Suite 330
487 Boston, MA 02111-1307 USA
488 @end format
489
490 We do not promise to fix the bug; but if the bug is serious,
491 or ugly, or easy to fix, chances are we will want to.
492
493 @findex report-emacs-bug
494 A convenient way to send a bug report for Emacs is to use the command
495 @kbd{M-x report-emacs-bug}. This sets up a mail buffer (@pxref{Sending
496 Mail}) and automatically inserts @emph{some} of the essential
497 information. However, it cannot supply all the necessary information;
498 you should still read and follow the guidelines below, so you can enter
499 the other crucial information by hand before you send the message.
500
501 To enable maintainers to investigate a bug, your report
502 should include all these things:
503
504 @itemize @bullet
505 @item
506 The version number of Emacs. Without this, we won't know whether there
507 is any point in looking for the bug in the current version of GNU
508 Emacs.
509
510 You can get the version number by typing @kbd{M-x emacs-version
511 @key{RET}}. If that command does not work, you probably have something
512 other than GNU Emacs, so you will have to report the bug somewhere
513 else.
514
515 @item
516 The type of machine you are using, and the operating system name and
517 version number. @kbd{M-x emacs-version @key{RET}} provides this
518 information too. Copy its output from the @samp{*Messages*} buffer, so
519 that you get it all and get it accurately.
520
521 @item
522 The operands given to the @code{configure} command when Emacs was
523 installed.
524
525 @item
526 A complete list of any modifications you have made to the Emacs source.
527 (We may not have time to investigate the bug unless it happens in an
528 unmodified Emacs. But if you've made modifications and you don't tell
529 us, you are sending us on a wild goose chase.)
530
531 Be precise about these changes. A description in English is not
532 enough---send a context diff for them.
533
534 Adding files of your own, or porting to another machine, is a
535 modification of the source.
536
537 @item
538 Details of any other deviations from the standard procedure for installing
539 GNU Emacs.
540
541 @item
542 The complete text of any files needed to reproduce the bug.
543
544 If you can tell us a way to cause the problem without visiting any files,
545 please do so. This makes it much easier to debug. If you do need files,
546 make sure you arrange for us to see their exact contents. For example, it
547 can often matter whether there are spaces at the ends of lines, or a
548 newline after the last line in the buffer (nothing ought to care whether
549 the last line is terminated, but try telling the bugs that).
550
551 @item
552 The precise commands we need to type to reproduce the bug.
553
554 @findex open-dribble-file
555 @cindex dribble file
556 The easy way to record the input to Emacs precisely is to write a
557 dribble file. To start the file, execute the Lisp expression
558
559 @example
560 (open-dribble-file "~/dribble")
561 @end example
562
563 @noindent
564 using @kbd{M-:} or from the @samp{*scratch*} buffer just after
565 starting Emacs. From then on, Emacs copies all your input to the
566 specified dribble file until the Emacs process is killed.
567
568 @item
569 @findex open-termscript
570 @cindex termscript file
571 @cindex @code{TERM} environment variable
572 For possible display bugs, the terminal type (the value of environment
573 variable @code{TERM}), the complete termcap entry for the terminal from
574 @file{/etc/termcap} (since that file is not identical on all machines),
575 and the output that Emacs actually sent to the terminal.
576
577 The way to collect the terminal output is to execute the Lisp expression
578
579 @example
580 (open-termscript "~/termscript")
581 @end example
582
583 @noindent
584 using @kbd{M-:} or from the @samp{*scratch*} buffer just after
585 starting Emacs. From then on, Emacs copies all terminal output to the
586 specified termscript file as well, until the Emacs process is killed.
587 If the problem happens when Emacs starts up, put this expression into
588 your @file{.emacs} file so that the termscript file will be open when
589 Emacs displays the screen for the first time.
590
591 Be warned: it is often difficult, and sometimes impossible, to fix a
592 terminal-dependent bug without access to a terminal of the type that
593 stimulates the bug.@refill
594
595 @item
596 A description of what behavior you observe that you believe is
597 incorrect. For example, ``The Emacs process gets a fatal signal,'' or,
598 ``The resulting text is as follows, which I think is wrong.''
599
600 Of course, if the bug is that Emacs gets a fatal signal, then one can't
601 miss it. But if the bug is incorrect text, the maintainer might fail to
602 notice what is wrong. Why leave it to chance?
603
604 Even if the problem you experience is a fatal signal, you should still
605 say so explicitly. Suppose something strange is going on, such as, your
606 copy of the source is out of sync, or you have encountered a bug in the
607 C library on your system. (This has happened!) Your copy might crash
608 and the copy here might not. If you @emph{said} to expect a crash, then
609 when Emacs here fails to crash, we would know that the bug was not
610 happening. If you don't say to expect a crash, then we would not know
611 whether the bug was happening---we would not be able to draw any
612 conclusion from our observations.
613
614 @item
615 If the manifestation of the bug is an Emacs error message, it is
616 important to report the precise text of the error message, and a
617 backtrace showing how the Lisp program in Emacs arrived at the error.
618
619 To get the error message text accurately, copy it from the
620 @samp{*Messages*} buffer into the bug report. Copy all of it, not just
621 part.
622
623 To make a backtrace for the error, evaluate the Lisp expression
624 @code{(setq @w{debug-on-error t})} before the error happens (that is to
625 say, you must execute that expression and then make the bug happen).
626 This causes the error to run the Lisp debugger, which shows you a
627 backtrace. Copy the text of the debugger's backtrace into the bug
628 report.
629
630 This use of the debugger is possible only if you know how to make the
631 bug happen again. If you can't make it happen again, at least copy
632 the whole error message.
633
634 @item
635 Check whether any programs you have loaded into the Lisp world,
636 including your @file{.emacs} file, set any variables that may affect the
637 functioning of Emacs. Also, see whether the problem happens in a
638 freshly started Emacs without loading your @file{.emacs} file (start
639 Emacs with the @code{-q} switch to prevent loading the init file). If
640 the problem does @emph{not} occur then, you must report the precise
641 contents of any programs that you must load into the Lisp world in order
642 to cause the problem to occur.
643
644 @item
645 If the problem does depend on an init file or other Lisp programs that
646 are not part of the standard Emacs system, then you should make sure it
647 is not a bug in those programs by complaining to their maintainers
648 first. After they verify that they are using Emacs in a way that is
649 supposed to work, they should report the bug.
650
651 @item
652 If you wish to mention something in the GNU Emacs source, show the line
653 of code with a few lines of context. Don't just give a line number.
654
655 The line numbers in the development sources don't match those in your
656 sources. It would take extra work for the maintainers to determine what
657 code is in your version at a given line number, and we could not be
658 certain.
659
660 @item
661 Additional information from a C debugger such as GDB might enable
662 someone to find a problem on a machine which he does not have available.
663 If you don't know how to use GDB, please read the GDB manual---it is not
664 very long, and using GDB is easy. You can find the GDB distribution,
665 including the GDB manual in online form, in most of the same places you
666 can find the Emacs distribution. To run Emacs under GDB, you should
667 switch to the @file{src} subdirectory in which Emacs was compiled, then
668 do @samp{gdb emacs}. It is important for the directory @file{src} to be
669 current so that GDB will read the @file{.gdbinit} file in this
670 directory.
671
672 However, you need to think when you collect the additional information
673 if you want it to show what causes the bug.
674
675 @cindex backtrace for bug reports
676 For example, many people send just a backtrace, but that is not very
677 useful by itself. A simple backtrace with arguments often conveys
678 little about what is happening inside GNU Emacs, because most of the
679 arguments listed in the backtrace are pointers to Lisp objects. The
680 numeric values of these pointers have no significance whatever; all that
681 matters is the contents of the objects they point to (and most of the
682 contents are themselves pointers).
683
684 @findex debug_print
685 To provide useful information, you need to show the values of Lisp
686 objects in Lisp notation. Do this for each variable which is a Lisp
687 object, in several stack frames near the bottom of the stack. Look at
688 the source to see which variables are Lisp objects, because the debugger
689 thinks of them as integers.
690
691 To show a variable's value in Lisp syntax, first print its value, then
692 use the user-defined GDB command @code{pr} to print the Lisp object in
693 Lisp syntax. (If you must use another debugger, call the function
694 @code{debug_print} with the object as an argument.) The @code{pr}
695 command is defined by the file @file{.gdbinit}, and it works only if you
696 are debugging a running process (not with a core dump).
697
698 To make Lisp errors stop Emacs and return to GDB, put a breakpoint at
699 @code{Fsignal}.
700
701 To find out which Lisp functions are running, using GDB, move up the
702 stack, and each time you get to a frame for the function
703 @code{Ffuncall}, type these GDB commands:
704
705 @example
706 p *args
707 pr
708 @end example
709
710 @noindent
711 To print the first argument that the function received, use these
712 commands:
713
714 @example
715 p args[1]
716 pr
717 @end example
718
719 @noindent
720 You can print the other arguments likewise. The argument @code{nargs}
721 of @code{Ffuncall} says how many arguments @code{Ffuncall} received;
722 these include the Lisp function itself and the arguments for that
723 function.
724
725 The file @file{.gdbinit} defines several other commands that are useful
726 for examining the data types and contents of Lisp objects. Their names
727 begin with @samp{x}. These commands work at a lower level than
728 @code{pr}, and are less convenient, but they may work even when
729 @code{pr} does not, such as when debugging a core dump or when Emacs has
730 had a fatal signal.
731
732 @item
733 If the symptom of the bug is that Emacs fails to respond, don't assume
734 Emacs is ``hung''---it may instead be in an infinite loop. To find out
735 which, make the problem happen under GDB and stop Emacs once it is not
736 responding. (If Emacs is using X Windows directly, you can stop Emacs
737 by typing @kbd{C-z} at the GDB job.) Then try stepping with
738 @samp{step}. If Emacs is hung, the @samp{step} command won't return.
739 If it is looping, @samp{step} will return.
740
741 If this shows Emacs is hung in a system call, stop it again and examine
742 the arguments of the call. In your bug report, state exactly where in
743 the source the system call is, and what the arguments are.
744
745 If Emacs is in an infinite loop, please determine where the loop starts
746 and ends. The easiest way to do this is to use the GDB command
747 @samp{finish}. Each time you use it, Emacs resumes execution until it
748 exits one stack frame. Keep typing @samp{finish} until it doesn't
749 return---that means the infinite loop is in the stack frame which you
750 just tried to finish.
751
752 Stop Emacs again, and use @samp{finish} repeatedly again until you get
753 @emph{back to} that frame. Then use @samp{next} to step through that
754 frame. By stepping, you will see where the loop starts and ends. Also
755 please examine the data being used in the loop and try to determine why
756 the loop does not exit when it should. Include all of this information
757 in your bug report.
758 @end itemize
759
760 Here are some things that are not necessary in a bug report:
761
762 @itemize @bullet
763 @item
764 A description of the envelope of the bug---this is not necessary for a
765 reproducible bug.
766
767 Often people who encounter a bug spend a lot of time investigating
768 which changes to the input file will make the bug go away and which
769 changes will not affect it.
770
771 This is often time-consuming and not very useful, because the way we
772 will find the bug is by running a single example under the debugger with
773 breakpoints, not by pure deduction from a series of examples. You might
774 as well save time by not searching for additional examples.
775
776 Of course, if you can find a simpler example to report @emph{instead} of
777 the original one, that is a convenience. Errors in the output will be
778 easier to spot, running under the debugger will take less time, etc.
779
780 However, simplification is not vital; if you can't do this or don't have
781 time to try, please report the bug with your original test case.
782
783 @item
784 A system-call trace of Emacs execution.
785
786 System-call traces are very useful for certain special kinds of
787 debugging, but in most cases they give little useful information. It is
788 therefore strange that many people seem to think that @emph{the} way to
789 report information about a crash is to send a system-call trace. Perhaps
790 this is a habit formed from experience debugging programs that don't
791 have source code or debugging symbols.
792
793 In most programs, a backtrace is normally far, far more informative than
794 a system-call trace. Even in Emacs, a simple backtrace is generally
795 more informative, though to give full information you should supplement
796 the backtrace by displaying variable values and printing them as Lisp
797 objects with @code{pr} (see above).
798
799 @item
800 A patch for the bug.
801
802 A patch for the bug is useful if it is a good one. But don't omit the
803 other information that a bug report needs, such as the test case, on the
804 assumption that a patch is sufficient. We might see problems with your
805 patch and decide to fix the problem another way, or we might not
806 understand it at all. And if we can't understand what bug you are
807 trying to fix, or why your patch should be an improvement, we mustn't
808 install it.
809
810 @ifinfo
811 @xref{Sending Patches}, for guidelines on how to make it easy for us to
812 understand and install your patches.
813 @end ifinfo
814
815 @item
816 A guess about what the bug is or what it depends on.
817
818 Such guesses are usually wrong. Even experts can't guess right about
819 such things without first using the debugger to find the facts.
820 @end itemize
821
822 @node Sending Patches
823 @subsection Sending Patches for GNU Emacs
824
825 @cindex sending patches for GNU Emacs
826 @cindex patches, sending
827 If you would like to write bug fixes or improvements for GNU Emacs,
828 that is very helpful. When you send your changes, please follow these
829 guidelines to make it easy for the maintainers to use them. If you
830 don't follow these guidelines, your information might still be useful,
831 but using it will take extra work. Maintaining GNU Emacs is a lot of
832 work in the best of circumstances, and we can't keep up unless you do
833 your best to help.
834
835 @itemize @bullet
836 @item
837 Send an explanation with your changes of what problem they fix or what
838 improvement they bring about. For a bug fix, just include a copy of the
839 bug report, and explain why the change fixes the bug.
840
841 (Referring to a bug report is not as good as including it, because then
842 we will have to look it up, and we have probably already deleted it if
843 we've already fixed the bug.)
844
845 @item
846 Always include a proper bug report for the problem you think you have
847 fixed. We need to convince ourselves that the change is right before
848 installing it. Even if it is correct, we might have trouble
849 understanding it if we don't have a way to reproduce the problem.
850
851 @item
852 Include all the comments that are appropriate to help people reading the
853 source in the future understand why this change was needed.
854
855 @item
856 Don't mix together changes made for different reasons.
857 Send them @emph{individually}.
858
859 If you make two changes for separate reasons, then we might not want to
860 install them both. We might want to install just one. If you send them
861 all jumbled together in a single set of diffs, we have to do extra work
862 to disentangle them---to figure out which parts of the change serve
863 which purpose. If we don't have time for this, we might have to ignore
864 your changes entirely.
865
866 If you send each change as soon as you have written it, with its own
867 explanation, then two changes never get tangled up, and we can consider
868 each one properly without any extra work to disentangle them.
869
870 @item
871 Send each change as soon as that change is finished. Sometimes people
872 think they are helping us by accumulating many changes to send them all
873 together. As explained above, this is absolutely the worst thing you
874 could do.
875
876 Since you should send each change separately, you might as well send it
877 right away. That gives us the option of installing it immediately if it
878 is important.
879
880 @item
881 Use @samp{diff -c} to make your diffs. Diffs without context are hard
882 to install reliably. More than that, they are hard to study; we must
883 always study a patch to decide whether we want to install it. Unidiff
884 format is better than contextless diffs, but not as easy to read as
885 @samp{-c} format.
886
887 If you have GNU diff, use @samp{diff -c -F'^[_a-zA-Z0-9$]+ *('} when
888 making diffs of C code. This shows the name of the function that each
889 change occurs in.
890
891 @item
892 Avoid any ambiguity as to which is the old version and which is the new.
893 Please make the old version the first argument to diff, and the new
894 version the second argument. And please give one version or the other a
895 name that indicates whether it is the old version or your new changed
896 one.
897
898 @item
899 Write the change log entries for your changes. This is both to save us
900 the extra work of writing them, and to help explain your changes so we
901 can understand them.
902
903 The purpose of the change log is to show people where to find what was
904 changed. So you need to be specific about what functions you changed;
905 in large functions, it's often helpful to indicate where within the
906 function the change was.
907
908 On the other hand, once you have shown people where to find the change,
909 you need not explain its purpose in the change log. Thus, if you add a
910 new function, all you need to say about it is that it is new. If you
911 feel that the purpose needs explaining, it probably does---but put the
912 explanation in comments in the code. It will be more useful there.
913
914 Please read the @file{ChangeLog} files in the @file{src} and @file{lisp}
915 directories to see what sorts of information to put in, and to learn the
916 style that we use. If you would like your name to appear in the header
917 line, showing who made the change, send us the header line.
918 @xref{Change Log}.
919
920 @item
921 When you write the fix, keep in mind that we can't install a change that
922 would break other systems. Please think about what effect your change
923 will have if compiled on another type of system.
924
925 Sometimes people send fixes that @emph{might} be an improvement in
926 general---but it is hard to be sure of this. It's hard to install
927 such changes because we have to study them very carefully. Of course,
928 a good explanation of the reasoning by which you concluded the change
929 was correct can help convince us.
930
931 The safest changes are changes to the configuration files for a
932 particular machine. These are safe because they can't create new bugs
933 on other machines.
934
935 Please help us keep up with the workload by designing the patch in a
936 form that is clearly safe to install.
937 @end itemize
938
939 @node Contributing, Service, Bugs, Top
940 @section Contributing to Emacs Development
941
942 If you would like to help pretest Emacs releases to assure they work
943 well, or if you would like to work on improving Emacs, please contact
944 the maintainers at @code{bug-gnu-emacs@@gnu.org}. A pretester
945 should be prepared to investigate bugs as well as report them. If you'd
946 like to work on improving Emacs, please ask for suggested projects or
947 suggest your own ideas.
948
949 If you have already written an improvement, please tell us about it. If
950 you have not yet started work, it is useful to contact
951 @code{bug-gnu-emacs@@gnu.org} before you start; it might be
952 possible to suggest ways to make your extension fit in better with the
953 rest of Emacs.
954
955 @node Service, Command Arguments, Contributing, Top
956 @section How To Get Help with GNU Emacs
957
958 If you need help installing, using or changing GNU Emacs, there are two
959 ways to find it:
960
961 @itemize @bullet
962 @item
963 Send a message to the mailing list
964 @code{help-gnu-emacs@@gnu.org}, or post your request on
965 newsgroup @code{gnu.emacs.help}. (This mailing list and newsgroup
966 interconnect, so it does not matter which one you use.)
967
968 @item
969 Look in the service directory for someone who might help you for a fee.
970 The service directory is found in the file named @file{etc/SERVICE} in the
971 Emacs distribution.
972 @end itemize