trouble.texi 59.3 KB
Newer Older
Dave Love's avatar
#  
Dave Love committed
1
@c This is part of the Emacs manual.
Paul Eggert's avatar
Paul Eggert committed
2
@c Copyright (C) 1985-1987, 1993-1995, 1997, 2001-2019 Free Software
3
@c Foundation, Inc.
Dave Love's avatar
#  
Dave Love committed
4 5 6 7 8 9 10 11 12 13
@c See file emacs.texi for copying conditions.
@iftex
@chapter Dealing with Common Problems

  If you type an Emacs command you did not intend, the results are often
mysterious.  This chapter tells what you can do to cancel your mistake or
recover from a mysterious situation.  Emacs bugs and system crashes are
also considered.
@end iftex

14 15 16 17
@ifnottex
@raisesections
@end ifnottex

18
@node Quitting
Dave Love's avatar
#  
Dave Love committed
19 20 21 22
@section Quitting and Aborting

@table @kbd
@item C-g
23
@itemx C-@key{Break} @r{(MS-DOS only)}
24
Quit: cancel running or partially typed command.
Dave Love's avatar
#  
Dave Love committed
25 26 27 28 29 30 31
@item C-]
Abort innermost recursive editing level and cancel the command which
invoked it (@code{abort-recursive-edit}).
@item @key{ESC} @key{ESC} @key{ESC}
Either quit or abort, whichever makes sense (@code{keyboard-escape-quit}).
@item M-x top-level
Abort all recursive editing levels that are currently executing.
32 33 34
@item C-/
@itemx C-x u
@itemx C-_
Dave Love's avatar
#  
Dave Love committed
35 36 37
Cancel a previously made change in the buffer contents (@code{undo}).
@end table

Richard M. Stallman's avatar
Richard M. Stallman committed
38 39 40 41
  There are two ways of canceling a command before it has finished:
@dfn{quitting} with @kbd{C-g}, and @dfn{aborting} with @kbd{C-]} or
@kbd{M-x top-level}.  Quitting cancels a partially typed command, or
one which is still running.  Aborting exits a recursive editing level
42 43
and cancels the command that invoked the recursive edit
(@pxref{Recursive Edit}).
Dave Love's avatar
#  
Dave Love committed
44 45 46

@cindex quitting
@kindex C-g
Richard M. Stallman's avatar
Richard M. Stallman committed
47
  Quitting with @kbd{C-g} is the way to get rid of a partially typed
48 49 50 51 52 53 54
command, or a numeric argument that you don't want.  Furthermore, if
you are in the middle of a command that is running, @kbd{C-g} stops
the command in a relatively safe way.  For example, if you quit out of
a kill command that is taking a long time, either your text will
@emph{all} still be in the buffer, or it will @emph{all} be in the
kill ring, or maybe both.  If the region is active, @kbd{C-g}
deactivates the mark, unless Transient Mark mode is off
55
(@pxref{Disabled Transient Mark}).  If you are in the middle of an
56
incremental search, @kbd{C-g} behaves specially; it may take two
57 58
successive @kbd{C-g} characters to get out of a search.
@xref{Incremental Search}, for details.
Dave Love's avatar
#  
Dave Love committed
59

60
  On MS-DOS, the character @kbd{C-@key{Break}} serves as a quit character
Dave Love's avatar
#  
Dave Love committed
61 62 63
like @kbd{C-g}.  The reason is that it is not feasible, on MS-DOS, to
recognize @kbd{C-g} while a command is running, between interactions
with the user.  By contrast, it @emph{is} feasible to recognize
64
@kbd{C-@key{Break}} at all times.
65 66 67 68 69 70 71
@iftex
@xref{MS-DOS Keyboard,,,emacs-xtra, Specialized Emacs Features}.
@end iftex
@ifnottex
@xref{MS-DOS Keyboard}.
@end ifnottex

72
@findex keyboard-quit
Dave Love's avatar
#  
Dave Love committed
73 74
  @kbd{C-g} works by setting the variable @code{quit-flag} to @code{t}
the instant @kbd{C-g} is typed; Emacs Lisp checks this variable
Richard M. Stallman's avatar
Richard M. Stallman committed
75
frequently, and quits if it is non-@code{nil}.  @kbd{C-g} is only
Dave Love's avatar
#  
Dave Love committed
76
actually executed as a command if you type it while Emacs is waiting for
77
input.  In that case, the command it runs is @code{keyboard-quit}.
Dave Love's avatar
#  
Dave Love committed
78

79
  On a text terminal, if you quit with @kbd{C-g} a second time before
Paul Eggert's avatar
Paul Eggert committed
80 81
the first @kbd{C-g} is recognized, you activate the emergency-escape
feature and return to the shell.  @xref{Emergency Escape}.
Dave Love's avatar
#  
Dave Love committed
82 83

@cindex NFS and quitting
Richard M. Stallman's avatar
Richard M. Stallman committed
84 85 86 87 88
  There are some situations where you cannot quit.  When Emacs is
waiting for the operating system to do something, quitting is
impossible unless special pains are taken for the particular system
call within Emacs where the waiting occurs.  We have done this for the
system calls that users are likely to want to quit from, but it's
Martin Rudalics's avatar
Martin Rudalics committed
89
possible you will encounter a case not handled.  In one very common
Richard M. Stallman's avatar
Richard M. Stallman committed
90 91 92
case---waiting for file input or output using NFS---Emacs itself knows
how to quit, but many NFS implementations simply do not allow user
programs to stop waiting for NFS when the NFS server is hung.
Dave Love's avatar
#  
Dave Love committed
93 94 95 96 97 98 99 100 101 102 103 104 105 106

@cindex aborting recursive edit
@findex abort-recursive-edit
@kindex C-]
  Aborting with @kbd{C-]} (@code{abort-recursive-edit}) is used to get
out of a recursive editing level and cancel the command which invoked
it.  Quitting with @kbd{C-g} does not do this, and could not do this,
because it is used to cancel a partially typed command @emph{within} the
recursive editing level.  Both operations are useful.  For example, if
you are in a recursive edit and type @kbd{C-u 8} to enter a numeric
argument, you can cancel that argument with @kbd{C-g} and remain in the
recursive edit.

@findex keyboard-escape-quit
107
@kindex ESC ESC ESC
Richard M. Stallman's avatar
Richard M. Stallman committed
108 109 110 111 112 113 114 115 116 117
  The sequence @kbd{@key{ESC} @key{ESC} @key{ESC}}
(@code{keyboard-escape-quit}) can either quit or abort.  (We defined
it this way because @key{ESC} means ``get out'' in many PC programs.)
It can cancel a prefix argument, clear a selected region, or get out
of a Query Replace, like @kbd{C-g}.  It can get out of the minibuffer
or a recursive edit, like @kbd{C-]}.  It can also get out of splitting
the frame into multiple windows, as with @kbd{C-x 1}.  One thing it
cannot do, however, is stop a command that is running.  That's because
it executes as an ordinary command, and Emacs doesn't notice it until
it is ready for the next command.
Dave Love's avatar
#  
Dave Love committed
118 119

@findex top-level
Paul Eggert's avatar
Paul Eggert committed
120
  The command @kbd{M-x top-level} is equivalent to enough
121 122 123 124 125 126 127
@kbd{C-]} commands to get you out of all the levels of recursive edits
that you are in; it also exits the minibuffer if it is active.
@kbd{C-]} gets you out one level at a time, but @kbd{M-x top-level}
goes out all levels at once.  Both @kbd{C-]} and @kbd{M-x top-level}
are like all other commands, and unlike @kbd{C-g}, in that they take
effect only when Emacs is ready for a command.  @kbd{C-]} is an
ordinary key and has its meaning only because of its binding in the
Dave Love's avatar
#  
Dave Love committed
128 129
keymap.  @xref{Recursive Edit}.

130
  @kbd{C-/} (@code{undo}) is not strictly speaking a way of canceling
Dave Love's avatar
#  
Dave Love committed
131
a command, but you can think of it as canceling a command that already
132 133
finished executing.  @xref{Undo}, for more information about the undo
facility.
Dave Love's avatar
#  
Dave Love committed
134

135
@node Lossage
Dave Love's avatar
#  
Dave Love committed
136
@section Dealing with Emacs Trouble
137
@cindex troubleshooting Emacs
Dave Love's avatar
#  
Dave Love committed
138

139 140 141 142 143 144
  This section describes how to recognize and deal with situations in
which Emacs does not work as you expect, such as keyboard code mixups,
garbled displays, running out of memory, and crashes and hangs.

  @xref{Bugs}, for what to do when you think you have found a bug in
Emacs.
Dave Love's avatar
#  
Dave Love committed
145 146

@menu
Richard M. Stallman's avatar
Richard M. Stallman committed
147
* DEL Does Not Delete::   What to do if @key{DEL} doesn't delete.
148
* Stuck Recursive::       '[...]' in mode line around the parentheses.
149 150 151
* Screen Garbled::        Garbage on the screen.
* Text Garbled::          Garbage in the text.
* Memory Full::           How to cope when you run out of memory.
152
* Crashing::              What Emacs does when it crashes.
153
* After a Crash::         Recovering editing in an Emacs session that crashed.
154
* Emergency Escape::      What to do if Emacs stops responding.
Dave Love's avatar
#  
Dave Love committed
155 156
@end menu

157
@node DEL Does Not Delete
Dave Love's avatar
#  
Dave Love committed
158
@subsection If @key{DEL} Fails to Delete
159 160
@cindex @key{DEL} vs @key{BACKSPACE}
@cindex @key{BACKSPACE} vs @key{DEL}
161
@cindex @key{DEL} does not delete
162

163
  Every keyboard has a large key, usually labeled @key{BACKSPACE},
164
which is ordinarily used to erase the last character that you typed.
165
In Emacs, this key is supposed to be equivalent to @key{DEL}.
166

167
  When Emacs starts up on a graphical display, it determines
168
automatically which key should be @key{DEL}.  In some unusual cases,
169
Emacs gets the wrong information from the system, and @key{BACKSPACE}
170
ends up deleting forwards instead of backwards.
171

172
  Some keyboards also have a @key{Delete} key, which is ordinarily
173 174 175
used to delete forwards.  If this key deletes backward in Emacs, that
too suggests Emacs got the wrong information---but in the opposite
sense.
176

177
  On a text terminal, if you find that @key{BACKSPACE} prompts for a
178
Help command, like @kbd{Control-h}, instead of deleting a character,
179
it means that key is actually sending the @samp{BS} character.  Emacs
180
ought to be treating @key{BS} as @key{DEL}, but it isn't.
181

182
@findex normal-erase-is-backspace-mode
183
  In all of those cases, the immediate remedy is the same: use the
184 185
command @kbd{M-x normal-erase-is-backspace-mode}.  This toggles
between the two modes that Emacs supports for handling @key{DEL}, so
Richard M. Stallman's avatar
Richard M. Stallman committed
186
if Emacs starts in the wrong mode, this should switch to the right
187
mode.  On a text terminal, if you want to ask for help when @key{BS}
188 189
is treated as @key{DEL}, use @key{F1} instead of @kbd{C-h}; @kbd{C-?}
may also work, if it sends character code 127.
190

191 192
  To fix the problem in every Emacs session, put one of the following
lines into your initialization file (@pxref{Init File}).  For the
193 194
first case above, where @key{BACKSPACE} deletes forwards instead of
backwards, use this line to make @key{BACKSPACE} act as @key{DEL}:
195 196 197 198 199 200

@lisp
(normal-erase-is-backspace-mode 0)
@end lisp

@noindent
201
For the other two cases, use this line:
202 203 204 205 206 207 208 209 210 211 212

@lisp
(normal-erase-is-backspace-mode 1)
@end lisp

@vindex normal-erase-is-backspace
  Another way to fix the problem for every Emacs session is to
customize the variable @code{normal-erase-is-backspace}: the value
@code{t} specifies the mode where @key{BS} or @key{BACKSPACE} is
@key{DEL}, and @code{nil} specifies the other mode.  @xref{Easy
Customization}.
Dave Love's avatar
#  
Dave Love committed
213 214 215

@node Stuck Recursive
@subsection Recursive Editing Levels
216 217
@cindex stuck in recursive editing
@cindex recursive editing, cannot exit
Dave Love's avatar
#  
Dave Love committed
218 219

  Recursive editing levels are important and useful features of Emacs, but
220
they can seem like malfunctions if you do not understand them.
Dave Love's avatar
#  
Dave Love committed
221

222 223 224 225 226 227
  If the mode line has square brackets @samp{[@dots{}]} around the
parentheses that contain the names of the major and minor modes, you
have entered a recursive editing level.  If you did not do this on
purpose, or if you don't understand what that means, you should just
get out of the recursive editing level.  To do so, type @kbd{M-x
top-level}.  @xref{Recursive Edit}.
Dave Love's avatar
#  
Dave Love committed
228 229 230

@node Screen Garbled
@subsection Garbage on the Screen
231 232 233
@cindex garbled display
@cindex display, incorrect
@cindex screen display, wrong
Dave Love's avatar
#  
Dave Love committed
234

235
  If the text on a text terminal looks wrong, the first thing to do is
236 237 238 239
see whether it is wrong in the buffer.  Type @kbd{C-l}
(@code{recenter-top-bottom}) to redisplay the entire screen.  If the
screen appears correct after this, the problem was entirely in the
previous screen update.  (Otherwise, see the following section.)
Dave Love's avatar
#  
Dave Love committed
240

Richard M. Stallman's avatar
Richard M. Stallman committed
241 242 243 244
  Display updating problems often result from an incorrect terminfo
entry for the terminal you are using.  The file @file{etc/TERMS} in
the Emacs distribution gives the fixes for known problems of this
sort.  @file{INSTALL} contains general advice for these problems in
245 246 247
one of its sections.  If you seem to be using the right terminfo
entry, it is possible that there is a bug in the terminfo entry, or a
bug in Emacs that appears for certain terminal types.
Dave Love's avatar
#  
Dave Love committed
248 249 250

@node Text Garbled
@subsection Garbage in the Text
251 252
@cindex garbled text
@cindex buffer text garbled
Dave Love's avatar
#  
Dave Love committed
253

254 255 256 257 258
  If @kbd{C-l} shows that the text is wrong, first type @kbd{C-h l}
(@code{view-lossage}) to see what commands you typed to produce the
observed results.  Then try undoing the changes step by step using
@kbd{C-x u} (@code{undo}), until it gets back to a state you consider
correct.
Dave Love's avatar
#  
Dave Love committed
259 260 261 262 263

  If a large portion of text appears to be missing at the beginning or
end of the buffer, check for the word @samp{Narrow} in the mode line.
If it appears, the text you don't see is probably still present, but
temporarily off-limits.  To make it accessible again, type @kbd{C-x n
264
w} (@code{widen}).  @xref{Narrowing}.
Dave Love's avatar
#  
Dave Love committed
265 266 267 268 269 270

@node Memory Full
@subsection Running out of Memory
@cindex memory full
@cindex out of memory

271
  If you get the error message @samp{Virtual memory exceeded}, save
272 273 274 275 276 277 278
your modified buffers with @kbd{C-x s} (@code{save-some-buffers}).
This method of saving them has the smallest need for additional
memory.  Emacs keeps a reserve of memory which it makes available when
this error happens; that should be enough to enable @kbd{C-x s} to
complete its work.  When the reserve has been used, @samp{!MEM FULL!}
appears at the beginning of the mode line, indicating there is no more
reserve.
279 280 281 282 283 284 285

  Once you have saved your modified buffers, you can exit this Emacs
session and start another, or you can use @kbd{M-x kill-some-buffers}
to free space in the current Emacs job.  If this frees up sufficient
space, Emacs will refill its memory reserve, and @samp{!MEM FULL!}
will disappear from the mode line.  That means you can safely go on
editing in the same Emacs session.
Dave Love's avatar
#  
Dave Love committed
286 287

  Do not use @kbd{M-x buffer-menu} to save or kill buffers when you run
288
out of memory, because the Buffer Menu needs a fair amount of memory
Dave Love's avatar
#  
Dave Love committed
289 290
itself, and the reserve supply may not be enough.

291 292 293
@node Crashing
@subsection When Emacs Crashes

294
@cindex crash report
295 296
@cindex backtrace
@cindex @file{emacs_backtrace.txt} file, MS-Windows
297 298 299
  Emacs is not supposed to crash, but if it does, it produces a
@dfn{crash report} prior to exiting.  The crash report is printed to
the standard error stream.  If Emacs was started from a graphical
300 301 302 303 304 305
desktop on a GNU or Unix system, the standard error stream is commonly
redirected to a file such as @file{~/.xsession-errors}, so you can
look for the crash report there.  On MS-Windows, the crash report is
written to a file named @file{emacs_backtrace.txt} in the current
directory of the Emacs process, in addition to the standard error
stream.
306 307 308 309 310

  The format of the crash report depends on the platform.  On some
platforms, such as those using the GNU C Library, the crash report
includes a @dfn{backtrace} describing the execution state prior to
crashing, which can be used to help debug the crash.  Here is an
311
example for a GNU system:
312 313 314 315 316 317 318 319 320 321 322

@example
Fatal error 11: Segmentation fault
Backtrace:
emacs[0x5094e4]
emacs[0x4ed3e6]
emacs[0x4ed504]
/lib64/libpthread.so.0[0x375220efe0]
/lib64/libpthread.so.0(read+0xe)[0x375220e08e]
emacs[0x509af6]
emacs[0x5acc26]
323
@dots{}
324 325 326
@end example

@noindent
327 328 329 330 331 332 333 334
The number @samp{11} is the system signal number corresponding to the
crash---in this case a segmentation fault.  The hexadecimal numbers
are program addresses, which can be associated with source code lines
using a debugging tool.  For example, the GDB command
@samp{list *0x509af6} prints the source-code lines corresponding to
the @samp{emacs[0x509af6]} entry.  If your system has the
@command{addr2line} utility, the following shell command outputs a
backtrace with source-code line numbers:
335 336 337

@example
sed -n 's/.*\[\(.*\)]$/\1/p' @var{backtrace} |
338
  addr2line -C -f -i -p -e @var{bindir}/@var{emacs-binary}
339 340 341 342
@end example

@noindent
Here, @var{backtrace} is the name of a text file containing a copy of
343 344 345
the backtrace, @var{bindir} is the name of the directory that
contains the Emacs executable, and @var{emacs-binary} is the name of
the Emacs executable file, normally @file{emacs} on GNU and Unix
Paul Eggert's avatar
Paul Eggert committed
346
systems and @file{emacs.exe} on MS-Windows and MS-DOS@.  Omit the
Paul Eggert's avatar
Paul Eggert committed
347 348
@option{-p} option if your version of @command{addr2line} is too old
to have it.
349

350
@cindex core dump
351 352 353
  Optionally, Emacs can generate a @dfn{core dump} when it crashes, on
systems that support core files.  A core dump is a file containing
voluminous data about the state of the program prior to the crash,
354
usually examined by loading it into a debugger such as GDB@.  On many
355 356
platforms, core dumps are disabled by default, and you must explicitly
enable them by running the shell command @samp{ulimit -c unlimited}
357
(e.g., in your shell startup script).
358

Dave Love's avatar
#  
Dave Love committed
359 360
@node After a Crash
@subsection Recovery After a Crash
361
@cindex recovering crashed session
Dave Love's avatar
#  
Dave Love committed
362 363 364 365 366 367 368 369 370 371

  If Emacs or the computer crashes, you can recover the files you were
editing at the time of the crash from their auto-save files.  To do
this, start Emacs again and type the command @kbd{M-x recover-session}.

  This command initially displays a buffer which lists interrupted
session files, each with its date.  You must choose which session to
recover from.  Typically the one you want is the most recent one.  Move
point to the one you choose, and type @kbd{C-c C-c}.

Richard M. Stallman's avatar
Richard M. Stallman committed
372 373 374 375 376 377 378
  Then @code{recover-session} considers each of the files that you
were editing during that session; for each such file, it asks whether
to recover that file.  If you answer @kbd{y} for a file, it shows the
dates of that file and its auto-save file, then asks once again
whether to recover that file.  For the second question, you must
confirm with @kbd{yes}.  If you do, Emacs visits the file but gets the
text from the auto-save file.
Dave Love's avatar
#  
Dave Love committed
379 380 381 382 383

  When @code{recover-session} is done, the files you've chosen to
recover are present in Emacs buffers.  You should then save them.  Only
this---saving them---updates the files themselves.

384 385 386
  As a last resort, if you had buffers with content which were not
associated with any files, or if the autosave was not recent enough to
have recorded important changes, you can use the
387 388 389 390 391
@file{etc/emacs-buffer.gdb} script with GDB (the GNU Debugger) to
retrieve them from a core dump--provided that a core dump was saved,
and that the Emacs executable was not stripped of its debugging
symbols.

Richard M. Stallman's avatar
Richard M. Stallman committed
392 393 394
  As soon as you get the core dump, rename it to another name such as
@file{core.emacs}, so that another crash won't overwrite it.

395
  To use this script, run @code{gdb} with the file name of your Emacs
396
executable and the file name of the core dump, e.g., @samp{gdb
397 398
/usr/bin/emacs core.emacs}.  At the @code{(gdb)} prompt, load the
recovery script: @samp{source /usr/src/emacs/etc/emacs-buffer.gdb}.
399 400 401 402 403
Then type the command @code{ybuffer-list} to see which buffers are
available.  For each buffer, it lists a buffer number.  To save a
buffer, use @code{ysave-buffer}; you specify the buffer number, and
the file name to write that buffer into.  You should use a file name
which does not already exist; if the file does exist, the script does
404
not make a backup of its old contents.
405

Dave Love's avatar
#  
Dave Love committed
406 407
@node Emergency Escape
@subsection Emergency Escape
408
@cindex emergency escape
Dave Love's avatar
#  
Dave Love committed
409

410 411
  On text terminals, the @dfn{emergency escape} feature suspends Emacs
immediately if you type @kbd{C-g} a second time before Emacs can
Richard M. Stallman's avatar
Richard M. Stallman committed
412 413 414 415 416 417 418
actually respond to the first one by quitting.  This is so you can
always get out of GNU Emacs no matter how badly it might be hung.
When things are working properly, Emacs recognizes and handles the
first @kbd{C-g} so fast that the second one won't trigger emergency
escape.  However, if some problem prevents Emacs from handling the
first @kbd{C-g} properly, then the second one will get you back to the
shell.
Dave Love's avatar
#  
Dave Love committed
419

Richard M. Stallman's avatar
Richard M. Stallman committed
420
  When you resume Emacs after a suspension caused by emergency escape,
421 422
it reports the resumption and asks a question or two before going back
to what it had been doing:
Dave Love's avatar
#  
Dave Love committed
423 424

@example
425
Emacs is resuming after an emergency escape.
Dave Love's avatar
#  
Dave Love committed
426 427 428 429 430
Auto-save? (y or n)
Abort (and dump core)? (y or n)
@end example

@noindent
431
Answer each question with @kbd{y} or @kbd{n} followed by @key{RET}.
Dave Love's avatar
#  
Dave Love committed
432

Richard M. Stallman's avatar
Richard M. Stallman committed
433 434
  Saying @kbd{y} to @samp{Auto-save?} causes immediate auto-saving of
all modified buffers in which auto-saving is enabled.  Saying @kbd{n}
435 436
skips this.  This question is omitted if Emacs is in a state where
auto-saving cannot be done safely.
Richard M. Stallman's avatar
Richard M. Stallman committed
437 438 439 440 441

  Saying @kbd{y} to @samp{Abort (and dump core)?} causes Emacs to
crash, dumping core.  This is to enable a wizard to figure out why
Emacs was failing to quit in the first place.  Execution does not
continue after a core dump.
Dave Love's avatar
#  
Dave Love committed
442

Richard M. Stallman's avatar
Richard M. Stallman committed
443 444 445
  If you answer this question @kbd{n}, Emacs execution resumes.  With
luck, Emacs will ultimately do the requested quit.  If not, each
subsequent @kbd{C-g} invokes emergency escape again.
Dave Love's avatar
#  
Dave Love committed
446 447

  If Emacs is not really hung, just slow, you may invoke the double
Richard M. Stallman's avatar
Richard M. Stallman committed
448 449 450
@kbd{C-g} feature without really meaning to.  Then just resume and
answer @kbd{n} to both questions, and you will get back to the former
state.  The quit you requested will happen by and by.
Dave Love's avatar
#  
Dave Love committed
451

452 453 454
  Emergency escape is active only for text terminals.  On graphical
displays, you can use the mouse to kill Emacs or switch to another
program.
Dave Love's avatar
#  
Dave Love committed
455

456
  On MS-DOS, you must type @kbd{C-@key{Break}} (twice) to cause
457
emergency escape---but there are cases where it won't work, when a
Richard M. Stallman's avatar
Richard M. Stallman committed
458
system call hangs or when Emacs is stuck in a tight loop in C code.
Dave Love's avatar
#  
Dave Love committed
459

460
@node Bugs
Dave Love's avatar
#  
Dave Love committed
461 462 463
@section Reporting Bugs

@cindex bugs
464 465 466 467 468
  If you think you have found a bug in Emacs, please report it.  We
cannot promise to fix it, or always to agree that it is a bug, but we
certainly want to hear about it.  The same applies for new features
you would like to see added.  The following sections will help you to
construct an effective bug report.
469

Dave Love's avatar
#  
Dave Love committed
470
@menu
471
* Known Problems::               How to read about known problems and bugs.
472 473 474 475
* Criteria:  Bug Criteria.       Have you really found a bug?
* Understanding Bug Reporting::  How to report a bug effectively.
* Checklist::                    Steps to follow for a good bug report.
* Sending Patches::              How to send a patch for GNU Emacs.
Dave Love's avatar
#  
Dave Love committed
476 477
@end menu

478 479 480 481 482 483 484 485 486 487
@node Known Problems
@subsection Reading Existing Bug Reports and Known Problems

  Before reporting a bug, if at all possible please check to see if it
is already known about.  Indeed, it may already have been fixed in a
later release of Emacs, or in the development version.  Here is a list
of the main places you can read about known issues:

@itemize
@item
488 489 490 491
The @file{etc/PROBLEMS} file; type @kbd{C-h C-p} to read it.  This
file contains a list of particularly well-known issues that have been
encountered in compiling, installing and running Emacs.  Often, there
are suggestions for workarounds and solutions.
492

493 494
@cindex bug tracker
@item
495
The GNU Bug Tracker at @url{https://debbugs.gnu.org}.  Emacs bugs are
496 497 498 499 500 501 502 503 504 505 506
filed in the tracker under the @samp{emacs} package.  The tracker
records information about the status of each bug, the initial bug
report, and the follow-up messages by the bug reporter and Emacs
developers.  You can search for bugs by subject, severity, and other
criteria.

@cindex debbugs package
Instead of browsing the bug tracker as a webpage, you can browse it
from Emacs using the @code{debbugs} package, which can be downloaded
via the Package Menu (@pxref{Packages}).  This package provides the
command @kbd{M-x debbugs-gnu} to list bugs, and @kbd{M-x
507 508
debbugs-gnu-search} to search for a specific bug.  User tags, applied
by the Emacs maintainers, are shown by @kbd{M-x debbugs-gnu-usertags}.
509

510 511
@item
The @samp{bug-gnu-emacs} mailing list (also available as the newsgroup
512
@samp{gnu.emacs.bug}).  You can read the list archives at
Paul Eggert's avatar
Paul Eggert committed
513
@url{https://lists.gnu.org/mailman/listinfo/bug-gnu-emacs}.  This list
Paul Eggert's avatar
Paul Eggert committed
514
works as a mirror of the Emacs bug reports and follow-up messages
515 516
which are sent to the bug tracker.  It also contains old bug reports
from before the bug tracker was introduced (in early 2008).
517

518 519 520 521
If you like, you can subscribe to the list.  Be aware that its purpose
is to provide the Emacs maintainers with information about bugs and
feature requests, so reports may contain fairly large amounts of data;
spectators should not complain about this.
522 523 524 525 526 527

@item
The @samp{emacs-pretest-bug} mailing list.  This list is no longer
used, and is mainly of historical interest.  At one time, it was used
for bug reports in development (i.e., not yet released) versions of
Emacs.  You can read the archives for 2003 to mid 2007 at
528
@url{https://lists.gnu.org/r/emacs-pretest-bug/}.  Nowadays,
529
it is an alias for @samp{bug-gnu-emacs}.
530 531 532 533 534 535 536 537 538 539

@item
The @samp{emacs-devel} mailing list.  Sometimes people report bugs to
this mailing list.  This is not the main purpose of the list, however,
and it is much better to send bug reports to the bug list.  You should
not feel obliged to read this list before reporting a bug.

@end itemize


Dave Love's avatar
#  
Dave Love committed
540 541
@node Bug Criteria
@subsection When Is There a Bug
542 543
@cindex bug criteria
@cindex what constitutes an Emacs bug
Dave Love's avatar
#  
Dave Love committed
544

545 546 547 548
  If Emacs accesses an invalid memory location (a.k.a.@:
``segmentation fault'') or exits with an operating system error
message that indicates a problem in the program (as opposed to
something like ``disk full''), then it is certainly a bug.
Dave Love's avatar
#  
Dave Love committed
549

550 551 552 553
  If the Emacs display does not correspond properly to the contents of
the buffer, then it is a bug.  But you should check that features like
buffer narrowing (@pxref{Narrowing}), which can hide parts of the
buffer or change how it is displayed, are not responsible.
Dave Love's avatar
#  
Dave Love committed
554 555

  Taking forever to complete a command can be a bug, but you must make
556
sure that it is really Emacs's fault.  Some commands simply take a
557
long time.  Type @kbd{C-g} (@kbd{C-@key{Break}} on MS-DOS) and then
558 559 560 561 562
@kbd{C-h l} to see whether the input Emacs received was what you
intended to type; if the input was such that you @emph{know} it should
have been processed quickly, report a bug.  If you don't know whether
the command should take a long time, find out by looking in the manual
or by asking for assistance.
Dave Love's avatar
#  
Dave Love committed
563 564 565 566 567

  If a command you are familiar with causes an Emacs error message in a
case where its usual definition ought to be reasonable, it is probably a
bug.

568 569 570 571
  If a command does the wrong thing, that is a bug.  But be sure you
know for certain what it ought to have done.  If you aren't familiar
with the command, it might actually be working right.  If in doubt,
read the command's documentation (@pxref{Name Help}).
Dave Love's avatar
#  
Dave Love committed
572

573 574 575
  A command's intended definition may not be the best possible
definition for editing with.  This is a very important sort of
problem, but it is also a matter of judgment.  Also, it is easy to
576 577 578 579
come to such a conclusion out of ignorance of some of the existing
features.  It is probably best not to complain about such a problem
until you have checked the documentation in the usual ways, feel
confident that you understand it, and know for certain that what you
Richard M. Stallman's avatar
Richard M. Stallman committed
580 581 582 583
want is not available.  Ask other Emacs users, too.  If you are not
sure what the command is supposed to do after a careful reading of the
manual, check the index and glossary for any terms that may be
unclear.
Dave Love's avatar
#  
Dave Love committed
584 585 586 587 588 589 590

  If after careful rereading of the manual you still do not understand
what the command should do, that indicates a bug in the manual, which
you should report.  The manual's job is to make everything clear to
people who are not Emacs experts---including you.  It is just as
important to report documentation bugs as program bugs.

591
  If the built-in documentation for a function or variable disagrees
Dave Love's avatar
#  
Dave Love committed
592 593
with the manual, one of them must be wrong; that is a bug.

594 595 596
  For problems with packages that are not part of Emacs, it is better
to begin by reporting them to the package developers.

Dave Love's avatar
#  
Dave Love committed
597 598
@node Understanding Bug Reporting
@subsection Understanding Bug Reporting
599 600
@cindex bug reporting
@cindex report an Emacs bug, how to
Dave Love's avatar
#  
Dave Love committed
601

602 603 604 605
  When you decide that there is a bug, it is important to report it
and to report it in a way which is useful.  What is most useful is an
exact description of what commands you type, starting with the shell
command to run Emacs, until the problem happens.
Dave Love's avatar
#  
Dave Love committed
606 607

  The most important principle in reporting a bug is to report
608 609 610 611 612 613 614 615 616
@emph{facts}.  Hypotheses and verbal descriptions are no substitute
for the detailed raw data.  Reporting the facts is straightforward,
but many people strain to posit explanations and report them instead
of the facts.  If the explanations are based on guesses about how
Emacs is implemented, they will be useless; meanwhile, lacking the
facts, we will have no real information about the bug.  If you want to
actually @emph{debug} the problem, and report explanations that are
more than guesses, that is useful---but please include the raw facts
as well.
Dave Love's avatar
#  
Dave Love committed
617 618

  For example, suppose that you type @kbd{C-x C-f /glorp/baz.ugh
Richard M. Stallman's avatar
Richard M. Stallman committed
619
@key{RET}}, visiting a file which (you know) happens to be rather
620 621 622 623
large, and Emacs displays @samp{I feel pretty today}.  The bug report
would need to provide all that information.  You should not assume
that the problem is due to the size of the file and say, ``I visited a
large file, and Emacs displayed @samp{I feel pretty today}.''  This is
624
what we mean by ``guessing explanations''.  The problem might be due
625 626
to the fact that there is a @samp{z} in the file name.  If this is so,
then when we got your report, we would try out the problem with some
Paul Eggert's avatar
Paul Eggert committed
627
large file, probably with no @samp{z} in its name, and not see any
628
problem.  There is no way we could guess that we should try visiting a
Dave Love's avatar
#  
Dave Love committed
629 630
file with a @samp{z} in its name.

631 632
  You should not even say ``visit a file'' instead of @kbd{C-x C-f}.
Similarly, rather than saying ``if I have three characters on the
633
line'', say ``after I type @kbd{@key{RET} A B C @key{RET} C-p}'', if
634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
that is the way you entered the text.

  If possible, try quickly to reproduce the bug by invoking Emacs with
@command{emacs -Q} (so that Emacs starts with no initial
customizations; @pxref{Initial Options}), and repeating the steps that
you took to trigger the bug.  If you can reproduce the bug this way,
that rules out bugs in your personal customizations.  Then your bug
report should begin by stating that you started Emacs with
@command{emacs -Q}, followed by the exact sequence of steps for
reproducing the bug.  If possible, inform us of the exact contents of
any file that is needed to reproduce the bug.

  Some bugs are not reproducible from @command{emacs -Q}; some are not
easily reproducible at all.  In that case, you should report what you
have---but, as before, please stick to the raw facts about what you
did to trigger the bug the first time.
Dave Love's avatar
#  
Dave Love committed
650

651 652
  If you have multiple issues that you want to report, please make a
separate bug report for each.
653

Dave Love's avatar
#  
Dave Love committed
654 655
@node Checklist
@subsection Checklist for Bug Reports
656 657
@cindex checklist before reporting a bug
@cindex bug reporting, checklist
658 659 660 661 662 663 664 665 666

  Before reporting a bug, first try to see if the problem has already
been reported (@pxref{Known Problems}).

If you are able to, try the latest release of Emacs to see if the
problem has already been fixed.  Even better is to try the latest
development version.  We recognize that this is not easy for some
people, so do not feel that you absolutely must do this before making
a report.
Dave Love's avatar
#  
Dave Love committed
667 668

@findex report-emacs-bug
669 670 671 672 673 674 675 676 677 678 679 680
  The best way to write a bug report for Emacs is to use the command
@kbd{M-x report-emacs-bug}.  This sets up a mail buffer
(@pxref{Sending Mail}) and automatically inserts @emph{some} of the
essential information.  However, it cannot supply all the necessary
information; you should still read and follow the guidelines below, so
you can enter the other crucial information by hand before you send
the message.  You may feel that some of the information inserted by
@kbd{M-x report-emacs-bug} is not relevant, but unless you are
absolutely sure it is best to leave it, so that the developers can
decide for themselves.

When you have finished writing your report, type @kbd{C-c C-c} and it
681 682 683 684 685
will be sent to the Emacs maintainers at
@ifnothtml
@email{bug-gnu-emacs@@gnu.org}.
@end ifnothtml
@ifhtml
Paul Eggert's avatar
Paul Eggert committed
686
@url{https://lists.gnu.org/mailman/listinfo/bug-gnu-emacs, bug-gnu-emacs}.
687
@end ifhtml
688 689
(If you want to suggest an improvement or new feature, use the same
address.)  If you cannot send mail from inside Emacs, you can copy the
690
text of your report to your normal mail client (if your system
691
supports it, you can type @kbd{C-c M-i} to have Emacs do this for you)
692 693
and send it to that address.  Or you can simply send an email to that
address describing the problem.
694 695

Your report will be sent to the @samp{bug-gnu-emacs} mailing list, and
696
stored in the GNU Bug Tracker at @url{https://debbugs.gnu.org}.  Please
697 698 699 700
include a valid reply email address, in case we need to ask you for
more information about your report.  Submissions are moderated, so
there may be a delay before your report appears.

Glenn Morris's avatar
Glenn Morris committed
701
You do not need to know how the GNU Bug Tracker works in order to
702 703
report a bug, but if you want to, you can read the tracker's online
documentation to see the various features you can use.
704 705

All mail sent to the @samp{bug-gnu-emacs} mailing list is also
706
gatewayed to the @samp{gnu.emacs.bug} newsgroup.  The reverse is also
707 708 709 710
true, but we ask you not to post bug reports (or replies) via the
newsgroup.  It can make it much harder to contact you if we need to ask
for more information, and it does not integrate well with the bug
tracker.
711 712 713

If your data is more than 500,000 bytes, please don't include it
directly in the bug report; instead, offer to send it on request, or
714 715
make it available online and say where.

Glenn Morris's avatar
Glenn Morris committed
716
The GNU Bug Tracker will assign a bug number to your report; please
717
use it in the following discussions.
Dave Love's avatar
#  
Dave Love committed
718 719 720 721 722 723

  To enable maintainers to investigate a bug, your report
should include all these things:

@itemize @bullet
@item
724 725
The version number of Emacs.  Without this, we won't know whether there is any
point in looking for the bug in the current version of GNU Emacs.
Dave Love's avatar
#  
Dave Love committed
726

727
@findex emacs-version
728 729 730 731 732
@kbd{M-x report-emacs-bug} includes this information automatically,
but if you are not using that command for your report you can get the
version number by typing @kbd{M-x emacs-version @key{RET}}.  If that
command does not work, you probably have something other than GNU
Emacs, so you will have to report the bug somewhere else.
Dave Love's avatar
#  
Dave Love committed
733 734 735

@item
The type of machine you are using, and the operating system name and
736 737
version number (again, automatically included by @kbd{M-x
report-emacs-bug}).  @kbd{M-x emacs-version @key{RET}} provides this
738
information too.  Copy its output from the @file{*Messages*} buffer,
739
so that you get it all and get it accurately.
Dave Love's avatar
#  
Dave Love committed
740 741 742

@item
The operands given to the @code{configure} command when Emacs was
743
installed (automatically included by @kbd{M-x report-emacs-bug}).
Dave Love's avatar
#  
Dave Love committed
744 745 746 747 748 749 750 751

@item
A complete list of any modifications you have made to the Emacs source.
(We may not have time to investigate the bug unless it happens in an
unmodified Emacs.  But if you've made modifications and you don't tell
us, you are sending us on a wild goose chase.)

Be precise about these changes.  A description in English is not
752
enough---send a unified context diff for them.
Dave Love's avatar
#  
Dave Love committed
753 754 755 756 757 758 759 760 761 762 763 764 765 766

Adding files of your own, or porting to another machine, is a
modification of the source.

@item
Details of any other deviations from the standard procedure for installing
GNU Emacs.

@item
The complete text of any files needed to reproduce the bug.

  If you can tell us a way to cause the problem without visiting any files,
please do so.  This makes it much easier to debug.  If you do need files,
make sure you arrange for us to see their exact contents.  For example, it
Richard M. Stallman's avatar
Richard M. Stallman committed
767
can matter whether there are spaces at the ends of lines, or a
Dave Love's avatar
#  
Dave Love committed
768 769 770 771
newline after the last line in the buffer (nothing ought to care whether
the last line is terminated, but try telling the bugs that).

@item
772 773 774 775
The precise commands we need to type to reproduce the bug.  If at all
possible, give a full recipe for an Emacs started with the @samp{-Q}
option (@pxref{Initial Options}).  This bypasses your personal
customizations.
Dave Love's avatar
#  
Dave Love committed
776 777 778

@findex open-dribble-file
@cindex dribble file
779
@cindex logging keystrokes
780
One way to record the input to Emacs precisely is to write a dribble
781 782 783 784 785
file.  To start the file, use the @kbd{M-x open-dribble-file} command.
From then on, Emacs copies all your input to the specified dribble
file until the Emacs process is killed.  Be aware that sensitive
information (such as passwords) may end up recorded in the dribble
file.
Dave Love's avatar
#  
Dave Love committed
786 787 788 789

@item
@findex open-termscript
@cindex termscript file
790
@vindex TERM@r{, environment variable, and display bugs}
791 792 793 794 795
For possible display bugs on text-mode terminals, the terminal type
(the value of environment variable @env{TERM}), the complete termcap
entry for the terminal from @file{/etc/termcap} (since that file is
not identical on all machines), and the output that Emacs actually
sent to the terminal.
Dave Love's avatar
#  
Dave Love committed
796 797 798 799 800 801 802 803

The way to collect the terminal output is to execute the Lisp expression

@example
(open-termscript "~/termscript")
@end example

@noindent
804
using @kbd{M-:} or from the @file{*scratch*} buffer just after
Dave Love's avatar
#  
Dave Love committed
805 806 807
starting Emacs.  From then on, Emacs copies all terminal output to the
specified termscript file as well, until the Emacs process is killed.
If the problem happens when Emacs starts up, put this expression into
808 809
your Emacs initialization file so that the termscript file will be
open when Emacs displays the screen for the first time.
Dave Love's avatar
#  
Dave Love committed
810 811 812

Be warned: it is often difficult, and sometimes impossible, to fix a
terminal-dependent bug without access to a terminal of the type that
Richard M. Stallman's avatar
Richard M. Stallman committed
813
stimulates the bug.
Dave Love's avatar
#  
Dave Love committed
814

815
@item
816
If non-@acronym{ASCII} text or internationalization is relevant, the locale that
817
was current when you started Emacs.  On GNU/Linux and Unix systems, or
Paul Eggert's avatar
Paul Eggert committed
818
if you use a POSIX-style shell such as Bash, you can use this shell
819
command to view the relevant values:
820

821
@smallexample
Richard M. Stallman's avatar
Richard M. Stallman committed
822
echo LC_ALL=$LC_ALL LC_COLLATE=$LC_COLLATE LC_CTYPE=$LC_CTYPE \
823
  LC_MESSAGES=$LC_MESSAGES LC_TIME=$LC_TIME LANG=$LANG
824
@end smallexample
825

826 827 828 829
Alternatively, use the @command{locale} command, if your system has it,
to display your locale settings.

You can use the @kbd{M-!} command to execute these commands from
830
Emacs, and then copy the output from the @file{*Messages*} buffer into
831
the bug report.  Alternatively, @kbd{M-x getenv @key{RET} LC_ALL
832
@key{RET}} will display the value of @code{LC_ALL} in the echo area, and
833
you can copy its output from the @file{*Messages*} buffer.
834

Dave Love's avatar
#  
Dave Love committed
835 836
@item
A description of what behavior you observe that you believe is
837
incorrect.  For example, ``The Emacs process gets a fatal signal'', or,
Dave Love's avatar
#  
Dave Love committed
838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853
``The resulting text is as follows, which I think is wrong.''

Of course, if the bug is that Emacs gets a fatal signal, then one can't
miss it.  But if the bug is incorrect text, the maintainer might fail to
notice what is wrong.  Why leave it to chance?

Even if the problem you experience is a fatal signal, you should still
say so explicitly.  Suppose something strange is going on, such as, your
copy of the source is out of sync, or you have encountered a bug in the
C library on your system.  (This has happened!)  Your copy might crash
and the copy here might not.  If you @emph{said} to expect a crash, then
when Emacs here fails to crash, we would know that the bug was not
happening.  If you don't say to expect a crash, then we would not know
whether the bug was happening---we would not be able to draw any
conclusion from our observations.

854 855 856
@item
If the bug is that the Emacs Manual or the Emacs Lisp Reference Manual
fails to describe the actual behavior of Emacs, or that the text is
857
confusing, copy in the text from the manual which you think is
858 859
at fault.  If the section is small, just the section name is enough.

Dave Love's avatar
#  
Dave Love committed
860 861 862 863 864 865
@item
If the manifestation of the bug is an Emacs error message, it is
important to report the precise text of the error message, and a
backtrace showing how the Lisp program in Emacs arrived at the error.

To get the error message text accurately, copy it from the
866
@file{*Messages*} buffer into the bug report.  Copy all of it, not just
Dave Love's avatar
#  
Dave Love committed
867 868
part.

869
@findex toggle-debug-on-error
Eli Zaretskii's avatar
Eli Zaretskii committed
870
@pindex Edebug
871 872
To make a backtrace for the error, use @kbd{M-x toggle-debug-on-error}
before the error happens (that is to say, you must give that command
Richard M. Stallman's avatar
Richard M. Stallman committed
873
and then make the bug happen).  This causes the error to start the Lisp
874
debugger, which shows you a backtrace.  Copy the text of the
875 876 877
debugger's backtrace into the bug report.  @xref{Edebug,, Edebug,
elisp, the Emacs Lisp Reference Manual}, for information on debugging
Emacs Lisp programs with the Edebug package.
Dave Love's avatar
#  
Dave Love committed
878 879 880 881 882

This use of the debugger is possible only if you know how to make the
bug happen again.  If you can't make it happen again, at least copy
the whole error message.

883 884 885 886 887 888 889
@vindex debug-on-quit
If Emacs appears to be stuck in an infinite loop or in a very long
operation, typing @kbd{C-g} with the variable @code{debug-on-quit}
non-@code{nil} will start the Lisp debugger and show a backtrace.
This backtrace is useful for debugging such long loops, so if you can
produce it, copy it into the bug report.

Glenn Morris's avatar
Glenn Morris committed
890 891 892 893 894 895
@vindex debug-on-event
If you cannot get Emacs to respond to @kbd{C-g} (e.g., because
@code{inhibit-quit} is set), then you can try sending the signal
specified by @code{debug-on-event} (default SIGUSR2) from outside
Emacs to cause it to enter the debugger.

Dave Love's avatar
#  
Dave Love committed
896 897
@item
Check whether any programs you have loaded into the Lisp world,
898 899 900 901 902 903 904
including your initialization file, set any variables that may affect
the functioning of Emacs.  Also, see whether the problem happens in a
freshly started Emacs without loading your initialization file (start
Emacs with the @code{-Q} switch to prevent loading the init files).
If the problem does @emph{not} occur then, you must report the precise
contents of any programs that you must load into the Lisp world in
order to cause the problem to occur.
Dave Love's avatar
#  
Dave Love committed
905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937

@item
If the problem does depend on an init file or other Lisp programs that
are not part of the standard Emacs system, then you should make sure it
is not a bug in those programs by complaining to their maintainers
first.  After they verify that they are using Emacs in a way that is
supposed to work, they should report the bug.

@item
If you wish to mention something in the GNU Emacs source, show the line
of code with a few lines of context.  Don't just give a line number.

The line numbers in the development sources don't match those in your
sources.  It would take extra work for the maintainers to determine what
code is in your version at a given line number, and we could not be
certain.

@item
Additional information from a C debugger such as GDB might enable
someone to find a problem on a machine which he does not have available.
If you don't know how to use GDB, please read the GDB manual---it is not
very long, and using GDB is easy.  You can find the GDB distribution,
including the GDB manual in online form, in most of the same places you
can find the Emacs distribution.  To run Emacs under GDB, you should
switch to the @file{src} subdirectory in which Emacs was compiled, then
do @samp{gdb emacs}.  It is important for the directory @file{src} to be
current so that GDB will read the @file{.gdbinit} file in this
directory.

However, you need to think when you collect the additional information
if you want it to show what causes the bug.

@cindex backtrace for bug reports
938 939 940 941 942 943 944
For example, many people send just a C-level backtrace, but that is
not very useful by itself.  A simple backtrace with arguments often
conveys little about what is happening inside GNU Emacs, because most
of the arguments listed in the backtrace are pointers to Lisp objects.
The numeric values of these pointers have no significance whatever;
all that matters is the contents of the objects they point to (and
most of the contents are themselves pointers).
Dave Love's avatar
#  
Dave Love committed
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962

@findex debug_print
To provide useful information, you need to show the values of Lisp
objects in Lisp notation.  Do this for each variable which is a Lisp
object, in several stack frames near the bottom of the stack.  Look at
the source to see which variables are Lisp objects, because the debugger
thinks of them as integers.

To show a variable's value in Lisp syntax, first print its value, then
use the user-defined GDB command @code{pr} to print the Lisp object in
Lisp syntax.  (If you must use another debugger, call the function
@code{debug_print} with the object as an argument.)  The @code{pr}
command is defined by the file @file{.gdbinit}, and it works only if you
are debugging a running process (not with a core dump).

To make Lisp errors stop Emacs and return to GDB, put a breakpoint at
@code{Fsignal}.

Gerd Moellmann's avatar
Gerd Moellmann committed
963
For a short listing of Lisp functions running, type the GDB
964
command @code{xbacktrace}.
Gerd Moellmann's avatar
Gerd Moellmann committed
965

Dave Love's avatar
#  
Dave Love committed
966 967 968 969 970 971 972
The file @file{.gdbinit} defines several other commands that are useful
for examining the data types and contents of Lisp objects.  Their names
begin with @samp{x}.  These commands work at a lower level than
@code{pr}, and are less convenient, but they may work even when
@code{pr} does not, such as when debugging a core dump or when Emacs has
had a fatal signal.

973 974 975 976 977
@cindex debugging Emacs, tricks and techniques
More detailed advice and other useful techniques for debugging Emacs
are available in the file @file{etc/DEBUG} in the Emacs distribution.
That file also includes instructions for investigating problems
whereby Emacs stops responding (many people assume that Emacs is
978
``hung'', whereas in fact it might be in an infinite loop).
979

980 981
To find the file @file{etc/DEBUG} in your Emacs installation, use the
directory name stored in the variable @code{data-directory}.
Dave Love's avatar
#  
Dave Love committed
982 983 984 985 986 987 988 989 990 991 992 993 994 995
@end itemize

Here are some things that are not necessary in a bug report:

@itemize @bullet
@item
A description of the envelope of the bug---this is not necessary for a
reproducible bug.

Often people who encounter a bug spend a lot of time investigating
which changes to the input file will make the bug go away and which
changes will not affect it.

This is often time-consuming and not very useful, because the way we
996 997 998 999 1000
will find the bug is by running a single example under the debugger
with breakpoints, not by pure deduction from a series of examples.
You might as well save time by not searching for additional examples.
It is better to send the bug report right away, go back to editing,
and find another bug to report.
Dave Love's avatar
#  
Dave Love committed
1001 1002 1003 1004 1005 1006 1007 1008

Of course, if you can find a simpler example to report @emph{instead} of
the original one, that is a convenience.  Errors in the output will be
easier to spot, running under the debugger will take less time, etc.

However, simplification is not vital; if you can't do this or don't have
time to try, please report the bug with your original test case.

1009 1010 1011 1012 1013 1014 1015 1016 1017
@item
A core dump file.

Debugging the core dump might be useful, but it can only be done on
your machine, with your Emacs executable.  Therefore, sending the core
dump file to the Emacs maintainers won't be useful.  Above all, don't
include the core file in an email bug report!  Such a large message
can be extremely inconvenient.

Dave Love's avatar
#  
Dave Love committed
1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
@item
A system-call trace of Emacs execution.

System-call traces are very useful for certain special kinds of
debugging, but in most cases they give little useful information.  It is
therefore strange that many people seem to think that @emph{the} way to
report information about a crash is to send a system-call trace.  Perhaps
this is a habit formed from experience debugging programs that don't
have source code or debugging symbols.

In most programs, a backtrace is normally far, far more informative than
a system-call trace.  Even in Emacs, a simple backtrace is generally
more informative, though to give full information you should supplement
the backtrace by displaying variable values and printing them as Lisp
objects with @code{pr} (see above).

@item
A patch for the bug.

A patch for the bug is useful if it is a good one.  But don't omit the
other information that a bug report needs, such as the test case, on the
assumption that a patch is sufficient.  We might see problems with your
patch and decide to fix the problem another way, or we might not
understand it at all.  And if we can't understand what bug you are
trying to fix, or why your patch should be an improvement, we mustn't
install it.

1045
@ifnottex
Dave Love's avatar
#  
Dave Love committed
1046 1047
@xref{Sending Patches}, for guidelines on how to make it easy for us to
understand and install your patches.
1048
@end ifnottex
Dave Love's avatar
#  
Dave Love committed
1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069

@item
A guess about what the bug is or what it depends on.

Such guesses are usually wrong.  Even experts can't guess right about
such things without first using the debugger to find the facts.
@end itemize

@node Sending Patches
@subsection Sending Patches for GNU Emacs

@cindex sending patches for GNU Emacs
@cindex patches, sending
  If you would like to write bug fixes or improvements for GNU Emacs,
that is very helpful.  When you send your changes, please follow these
guidelines to make it easy for the maintainers to use them.  If you
don't follow these guidelines, your information might still be useful,
but using it will take extra work.  Maintaining GNU Emacs is a lot of
work in the best of circumstances, and we can't keep up unless you do
your best to help.

1070 1071 1072 1073 1074 1075 1076 1077 1078
Every patch must have several pieces of information before we
can properly evaluate it.

When you have all these pieces, bundle them up in a mail message and
send it to the developers.  Sending it to
@email{bug-gnu-emacs@@gnu.org} (which is the bug/feature list) is
recommended, because that list is coupled to a tracking system that
makes it easier to locate patches.  If your patch is not complete and
you think it needs more discussion, you might want to send it to
Eli Zaretskii's avatar
Eli Zaretskii committed
1079
@email{emacs-devel@@gnu.org} instead.  If you revise your patch,
1080 1081 1082 1083 1084
send it as a followup to the initial topic.

We prefer to get the patches as plain text, either inline (be careful
your mail client does not change line breaks) or as MIME attachments.

Dave Love's avatar
#  
Dave Love committed
1085 1086
@itemize @bullet
@item
1087 1088 1089 1090 1091 1092
Include an explanation with your changes of what problem they fix or what
improvement they bring about.

@itemize
@item
For a fix for an existing bug, it is
1093
best to reply to the relevant discussion on the @samp{bug-gnu-emacs}
1094
list, or the bug entry in the GNU Bug Tracker at
1095
@url{https://debbugs.gnu.org}.  Explain why your change fixes the bug.
Dave Love's avatar
#  
Dave Love committed
1096 1097

@item
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
For a new feature, include a description of the feature and your
implementation.

@item
For a new bug, include a proper bug report for the problem you think
you have fixed.  We need to convince ourselves that the change is
right before installing it.  Even if it is correct, we might have
trouble understanding it if we don't have a way to reproduce the
problem.
@end itemize
Dave Love's avatar
#  
Dave Love committed
1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138

@item
Include all the comments that are appropriate to help people reading the
source in the future understand why this change was needed.

@item
Don't mix together changes made for different reasons.
Send them @emph{individually}.

If you make two changes for separate reasons, then we might not want to
install them both.  We might want to install just one.  If you send them
all jumbled together in a single set of diffs, we have to do extra work
to disentangle them---to figure out which parts of the change serve
which purpose.  If we don't have time for this, we might have to ignore
your changes entirely.

If you send each change as soon as you have written it, with its own
explanation, then two changes never get tangled up, and we can consider
each one properly without any extra work to disentangle them.

@item
Send each change as soon as that change is finished.  Sometimes people
think they are helping us by accumulating many changes to send them all
together.  As explained above, this is absolutely the worst thing you
could do.

Since you should send each change separately, you might as well send it
right away.  That gives us the option of installing it immediately if it
is important.

@item
1139 1140
The patch itself.

1141
Use @samp{diff -u} to make your diffs.  Diffs without context are hard
Dave Love's avatar
#  
Dave Love committed
1142
to install reliably.  More than that, they are hard to study; we must
1143
always study a patch to decide whether we want to install it.  Context
Dmitry Gutov's avatar
Dmitry Gutov committed
1144 1145
format is better than contextless diffs, but we prefer the unified
format.
Dave Love's avatar
#  
Dave Love committed
1146

1147
If you have GNU diff, use @samp{diff -u -F'^[_a-zA-Z0-9$]\+ *('} when
Dave Love's avatar
#  
Dave Love committed
1148 1149 1150
making diffs of C code.  This shows the name of the function that each
change occurs in.

1151
If you are using the Emacs repository, make sure your copy is
Paul Eggert's avatar
Paul Eggert committed
1152
up-to-date (e.g., with @code{git pull}).  You can commit your changes
1153 1154 1155 1156
to a private branch and generate a patch from the master version by
using @code{git format-patch master}. Or you can leave your changes
uncommitted and use @code{git diff}.

Dave Love's avatar
#  
Dave Love committed
1157 1158 1159 1160 1161 1162 1163 1164
@item
Avoid any ambiguity as to which is the old version and which is the new.
Please make the old version the first argument to diff, and the new
version the second argument.  And please give one version or the other a
name that indicates whether it is the old version or your new changed
one.

@item
1165
Write the commit log entries for your changes.  This is both to save us
Dave Love's avatar
#  
Dave Love committed
1166 1167 1168
the extra work of writing them, and to help explain your changes so we
can understand them.

1169
The purpose of the commit log is to show people where to find what was
Dave Love's avatar
#  
Dave Love committed
1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
changed.  So you need to be specific about what functions you changed;
in large functions, it's often helpful to indicate where within the
function the change was.

On the other hand, once you have shown people where to find the change,
you need not explain its purpose in the change log.  Thus, if you add a
new function, all you need to say about it is that it is new.  If you
feel that the purpose needs explaining, it probably does---but put the
explanation in comments in the code.  It will be more useful there.

1180 1181 1182
Please look at the commit log entries of recent commits to see what
sorts of information to put in, and to learn the style that we use.
Note that, unlike some other projects, we do require commit logs for
Paul Eggert's avatar
Paul Eggert committed
1183
documentation, i.e., Texinfo files.
1184 1185 1186
@xref{Change Log},
@ifset WWW_GNU_ORG
see
Paul Eggert's avatar