bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Markus Triska]

When I do:

  $ emacs -Q --daemon
  $ emacsclient -c

and, there, evaluate the forms:

   (defun my-test ())
   (add-hook 'window-configuration-change-hook 'my-test)
   (debug-on-entry 'my-test)

and then press M-x, the debugger is entered, as expected. When I press
C-x k to close the backtrace window and then press "C-]" twice (i.e.,
C-] C-]) to abort the recursive edit, I get the message:

   No catch for tag: exit, t

and cannot abort the recursive edit. M-x top-level RET says "Back to
top level", and fails to return to the top level too. When I press
C-M-c instead of C-] above, I get:

   Terminal 0 is locked, cannot read from it

These problems only occur in client/server mode. Relatedly, from NEWS:

   Now, the default behavior is to open a new Emacs frame by default.
   Use the -c option to get the old behavior of opening files in the
   currently selected Emacs frame.

I find that not to be the case: When I do:

   $ emacs -Q --daemon
   $ emacsclient test.txt

I get:

   Waiting for Emacs...

and see no new frame. When I do "emacsclient -c", a new frame appears.


In GNU Emacs 23.0.60.10 (i386-apple-darwin8.11.1, GTK+ Version 2.12.9)
 of 2008-10-10 on mt-computer.local
Windowing system distributor `The XFree86 Project, Inc', version 11.0.40400000
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: en_GB.UTF-8
  value of $XMODIFIERS: nil
  locale-coding-system: nil
  default-enable-multibyte-characters: t

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Andrew Hyatt]

Hi Markus -

I can't replicate this in Emacs 25.  Do you still see this?

Markus Triska <markus.triska@gmx.at> writes:

> When I do:
>
>   $ emacs -Q --daemon
>   $ emacsclient -c
>
> and, there, evaluate the forms:
>
>    (defun my-test ())
>    (add-hook 'window-configuration-change-hook 'my-test)
>    (debug-on-entry 'my-test)
>
> and then press M-x, the debugger is entered, as expected. When I press
> C-x k to close the backtrace window and then press "C-]" twice (i.e.,
> C-] C-]) to abort the recursive edit, I get the message:
>
>    No catch for tag: exit, t
>
> and cannot abort the recursive edit. M-x top-level RET says "Back to
> top level", and fails to return to the top level too. When I press
> C-M-c instead of C-] above, I get:
>
>    Terminal 0 is locked, cannot read from it
>
> These problems only occur in client/server mode. Relatedly, from NEWS:
>
>    Now, the default behavior is to open a new Emacs frame by default.
>    Use the -c option to get the old behavior of opening files in the
>    currently selected Emacs frame.
>
> I find that not to be the case: When I do:
>
>    $ emacs -Q --daemon
>    $ emacsclient test.txt
>
> I get:
>
>    Waiting for Emacs...
>
> and see no new frame. When I do "emacsclient -c", a new frame appears.
>
>
> In GNU Emacs 23.0.60.10 (i386-apple-darwin8.11.1, GTK+ Version 2.12.9)
>  of 2008-10-10 on mt-computer.local
> Windowing system distributor `The XFree86 Project, Inc', version 11.0.40400000
> 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: en_GB.UTF-8
>   value of $XMODIFIERS: nil
>   locale-coding-system: nil
>   default-enable-multibyte-characters: t

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Markus Triska]

Hi Andrew,
 
> I can't replicate this in Emacs 25. Do you still see this?

I have stopped using emacsclient due to its memory leaks (#1504).

All the best,
Markus

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Andrew Hyatt]

[-- Attachment #1: Type: text/plain, Size: 411 bytes --]

OK, thanks for letting me know.  I'll mark this as "doneunreproducible",
but if you are able to reproduce it in the future, let me know and I'll
re-open it.

On Mon, Jan 4, 2016 at 3:04 PM Markus Triska <markus.triska@gmx.at> wrote:

> Hi Andrew,
>
> > I can't replicate this in Emacs 25. Do you still see this?
>
> I have stopped using emacsclient due to its memory leaks (#1504).
>
> All the best,
> Markus
>

[-- Attachment #2: Type: text/html, Size: 706 bytes --]

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Ivan Andrus]

On Jan 4, 2016, at 1:04 PM, Markus Triska <markus.triska@gmx.at> wrote:
> 
> Hi Andrew,
>  
>> I can't replicate this in Emacs 25. Do you still see this?
> 
> I have stopped using emacsclient due to its memory leaks (#1504).

That bug (#1504) has been closed, so if it still leaks memory you 
should open a new bug (or reopen that one).

-Ivan

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Markus Triska]

Hi Ivan,

Ivan Andrus writes:

> That bug (#1504) has been closed, so if it still leaks memory you
> should open a new bug (or reopen that one).

#1504 contains all the information for reproducing the problem, and
you can easily see that the problem still exists in the latest git
version of Emacs.

If possible, I would like to focus on filing the issues I find and
providing you with all the information you need to reproduce these
problems, so please classify the issues I file as you see fit.

Thank you and all the best!
Markus

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Andrew Hyatt]

[-- Attachment #1: Type: text/plain, Size: 945 bytes --]

On Tue, Jan 5, 2016 at 1:36 PM Markus Triska <markus.triska@gmx.at> wrote:

> Hi Ivan,
>
> Ivan Andrus writes:
>
> > That bug (#1504) has been closed, so if it still leaks memory you
> > should open a new bug (or reopen that one).
>
> #1504 contains all the information for reproducing the problem, and
> you can easily see that the problem still exists in the latest git
> version of Emacs.
>
> If possible, I would like to focus on filing the issues I find and
> providing you with all the information you need to reproduce these
> problems, so please classify the issues I file as you see fit.
>
> Thank you and all the best!
> Markus
>

FWIW, I can't reproduce that bug on my Mac.  Or, if it leaks memory, it
leaks it so slowly that even running the emacsclient loop for minutes
results in no clear difference.  It may be that there is some system level
difference, or difference in how we built or installed emacs that's causing
the issue.

[-- Attachment #2: Type: text/html, Size: 1277 bytes --]

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Markus Triska]

Hi Andrew,
 
Andrew Hyatt writes: 

> FWIW, I can't reproduce that bug on my Mac.  Or, if it leaks memory, it
> leaks it so slowly that even running the emacsclient loop for minutes
> results in no clear difference.  It may be that there is some system level
> difference, or difference in how we built or installed emacs that's causing the issue.

Here are the settings I use on OSX 10.10.2, and which allow me to reproduce #1504
with the latest git version today: 

 ./configure --without-ns CFLAGS=-I/opt/local/include LDFLAGS=-L/opt/local/lib

and the following features (note in particular X11 and LUCID):

XAW3D XPM JPEG TIFF GIF PNG IMAGEMAGICK GSETTINGS NOTIFY ACL GNUTLS
LIBXML2 FREETYPE XFT ZLIB TOOLKIT_SCROLL_BARS LUCID X11

Important settings:
  value of $LANG: en_US.UTF-8
  locale-coding-system: utf-8-unix

bug#1148: 23.0.60; recursive edits can sometimes not be stopped in emacsclient

[Andrew Hyatt]

Thanks!  Since you can reproduce it still, I'll definitely leave this
open, and remove the "unreproducible" tag.

"Markus Triska" <markus.triska@gmx.at> writes:

> Hi Andrew,
>  
> Andrew Hyatt writes: 
>
>> FWIW, I can't reproduce that bug on my Mac.  Or, if it leaks memory, it
>> leaks it so slowly that even running the emacsclient loop for minutes
>> results in no clear difference.  It may be that there is some system level
>> difference, or difference in how we built or installed emacs that's causing the issue.
>
> Here are the settings I use on OSX 10.10.2, and which allow me to reproduce #1504
> with the latest git version today: 
>
>  ./configure --without-ns CFLAGS=-I/opt/local/include LDFLAGS=-L/opt/local/lib
>
> and the following features (note in particular X11 and LUCID):
>
> XAW3D XPM JPEG TIFF GIF PNG IMAGEMAGICK GSETTINGS NOTIFY ACL GNUTLS
> LIBXML2 FREETYPE XFT ZLIB TOOLKIT_SCROLL_BARS LUCID X11
>
> Important settings:
>   value of $LANG: en_US.UTF-8
>   locale-coding-system: utf-8-unix