Richard Stallman | 1 Oct 2006 02:03
Picon
Picon

Re: Emacs uses 100% cpu time

    >From time to time emacs starts utilizing 100% of cpu time. Sometimes it falls
    back to normal usage for some seconds, but it always catches up again.

Try turning off global-font-lock-mode.  Does it still happen?
Richard Stallman | 1 Oct 2006 02:04
Picon
Picon

Re: Emacs uses 100% cpu time

The way to find out where a program is running during a long operation is
to stop the program several times, making a backtrace each time.  When
you see a pattern start to emerge, then please report it.

Use plain `bt', NOT `bt full' (the latter produces a lot of clutter
which tends to obscure the pattern).

If the pattern you see involves lots of calls to Ffuncall, that means
it is running Lisp code.  In that case only, you need to use
`xbacktrace' in order to look at the Lisp-level backtrace.  You should
do this over and over until you see a pattern in which Lisp functions
are running.

Once you see the pattern, you should follow the instructions in etc/DEBUG
to figure out which of the frames is actually looping.

Of course, if people think this is a bug that is already fixed,
it is worth checking that first before you spend the time to do
the debugging described above.
Richard Stallman | 1 Oct 2006 02:04
Picon
Picon

Re: Emacs uses 100% cpu time

    With Stefane's patch to jit-lock.el my idle loop problems where
    emacs appeared to lock up went away but it's still pegging the
    cpu at 100% when idle.

First, is that due to stealth fontification?

Second, if it is due to stealth fontification,
is it a bug?
Richard Stallman | 1 Oct 2006 02:04
Picon
Picon

Re: Quitting Emacs for buffer unassociated with file

    emacs -Q
    C-x f toto

Do you mean C-x C-f?

    You get this menu (numbered here):

    Save file toto?
    1 Yes
    2 No
    3 View This File
    4 View Changes In File
    5 Save This Buffer But No More
    6 No For All

    3 and 4 make no sense in this context - they should be dimmed.

#3 makes sense, but you actually view the buffer.
#4 makes sense, though it would be sort of trivial,
since the whole contents would appear as an insertion.

Maybe I should just change "file" to "buffer".

    What does "No For All" mean? Unclear. Reword to make clear that it
    means do not save this buffer and don't ask about saving any others.

It is hard to do this.
Kim F. Storm | 1 Oct 2006 02:15
Picon

Re: list-processes and process sentinels not run.

Matt Hodges <M.P.Hodges <at> rl.ac.uk> writes:

> In GNU Emacs 22.0.50.3 (i686-pc-linux-gnu, GTK+ Version 2.8.20)
>  of 2006-09-29 on saiga
> X server distributor `The X.Org Foundation', version 11.0.70101000
> configured using `configure '--with-gtk''
>
> If a process status changes to exit and list-processes is called
> without there being idle time, then the process is removed in
> list_processes_1:
>
>       if (EQ (symbol, Qsignal) || EQ (symbol, Qexit))
>         remove_process (proc);
>
> and the process sentinel isn't run.

Does the following patch solve the problem?

*** process.c	08 Sep 2006 15:51:27 +0200	1.488
--- process.c	01 Oct 2006 02:11:29 +0200	
***************
*** 1314,1319 ****
--- 1314,1320 ----
    register struct Lisp_Process *p;
    char tembuf[300];
    int w_proc, w_buffer, w_tty;
+   int exited = 0;
    Lisp_Object i_status, i_buffer, i_tty, i_command;

    w_proc = 4;    /* Proc   */
(Continue reading)

Drew Adams | 1 Oct 2006 02:32
Picon
Favicon

RE: Quitting Emacs for buffer unassociated with file

        emacs -Q
        C-x f toto

    Do you mean C-x C-f?

Yes.

        You get this menu (numbered here):

        Save file toto?
        1 Yes
        2 No
        3 View This File
        4 View Changes In File
        5 Save This Buffer But No More
        6 No For All

        3 and 4 make no sense in this context - they should be dimmed.

    #3 makes sense, but you actually view the buffer.
    #4 makes sense, though it would be sort of trivial,
    since the whole contents would appear as an insertion.

    Maybe I should just change "file" to "buffer".

That wouldn't hurt (buffers are what get saved, not files), but it misses
the point of the bug report.

If a buffer has never been saved, there is no file to compare with it (and,
a fortiori, no need for a comparison). The only options that make sense then
(Continue reading)

Drew Adams | 1 Oct 2006 04:23
Picon
Favicon

Wrong link destination for apropos-command

emacs -Q

C-h f apropos-command

In *Help*, click the link for the apropos file (upper right). It
should take you to the definition of command
`apropos-command'. Instead, it takes you to the definition of button
type `apropos-command':

(define-button-type 'apropos-command
  'apropos-label "Command"
  'help-echo "mouse-2, RET: Display more help on this command"
  'follow-link t
  'action (lambda (button)
	    (describe-function (button-get button 'apropos-symbol))))

In GNU Emacs 22.0.50.1 (i386-msvc-nt5.1.2600)
 of 2006-07-19 on BOS-CTHEWLAP2
X server distributor `Microsoft Corp.', version 5.1.2600
configured using `configure --with-msvc (12.00)'

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: ENU
(Continue reading)

T. V. Raman | 1 Oct 2006 04:28

Re: Emacs uses 100% cpu time


It still appears to happen with global-font-lock-mode turned off
---

I've done a make distclean and built from scratch to be sure.

Also, a freshly started emacs doesn't appear to exhibit the
problem, am still trying to figure out at what point the problem
starts showing up.

>>>>> "Richard" == Richard Stallman <rms <at> gnu.org> writes:
    >> From time to time emacs starts utilizing 100% of cpu
    >> time. Sometimes it falls
    Richard>     back to normal usage for some seconds, but it
    Richard> always catches up again.
    Richard> 
    Richard> Try turning off global-font-lock-mode.  Does it
    Richard> still happen?
    Richard> 
    Richard> 
    Richard> _______________________________________________
    Richard> emacs-pretest-bug mailing list
    Richard> emacs-pretest-bug <at> gnu.org
    Richard> http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug

--

-- 
Best Regards,
--raman

      
(Continue reading)

Stefan Monnier | 1 Oct 2006 05:03
Picon

Re: Emacs uses 100% cpu time

>> From time to time emacs starts utilizing 100% of cpu time. Sometimes it falls
> back to normal usage for some seconds, but it always catches up again.

> Emacs stays responsible and the only indication is that my fan is running all
> the time, thus I'm not sure since when I have this problem. And I don't find a
> way to reproduce the bug. The only thing I would say is that it subjectively
> happens earlier, if network apps like gnus and circe are running.

Does (setq jit-lock-stealth-time nil) fix it?

        Stefan
Tassilo Horn | 1 Oct 2006 11:53
Picon
Favicon

Re: Emacs uses 100% cpu time

Stefan Monnier <monnier <at> iro.umontreal.ca> writes:

Hello Stefan,

> Does (setq jit-lock-stealth-time nil) fix it?

No. If it has already started looping it has no effect at all, and if I
set in in my ~/.emacs and restart it, I can reproduce the bug again, but
it seems to happen not that fast.

If it's set to the default of 16 I just have to open my ~/.emacs, and
bam, 100% cpu utilization. If it's set to nil I had to jump arround the
buffer for some time.

Now I'll do what Richard suggested: Making normal backtraces and looking
for a common pattern...

Bye,
Tassilo

Gmane