Emacs on Gtk3.8

Emacs on Gtk3.8

[Achim Gratz]

Since openSUSE Tumbleweed upgraded to Gtk3.8 process handling in Emacs
(both 24.3 and 24.3.50) is broken.  The issue manifests itself in a
failure to recognize that a process started from Emacs has in fact
terminated (it lingers as a zombie) and Emacs spinwaits.  Ctrl-G lets me
regain access to Emacs, but the high CPU utilization doesn't stop.

On current trunk the only way to stop the madness is to exit Emacs (it
will ask to kill the running process which actually doesn't run
anymore).  On Emacs 24.3 it is usually possible to kill the process
buffer to get the CPU back to normal, in a few cases I had to
kill-process the (already finished) process.  This at least allows to
continue working without having to restart Emacs.

Before I open a bugreport, is there anything I could/should do?


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Samples for the Waldorf Blofeld:
http://Synth.Stromeko.net/Downloads.html#BlofeldSamplesExtra

Re: Emacs on Gtk3.8

[Paul Eggert]

On 05/14/13 12:17, Achim Gratz wrote:
> Before I open a bugreport, is there anything I could/should do?

I suggest reading this bug report:

bugs.gnu.org/14297

You can send email to 14297@debbugs.gnu.org to add
further comments.

My current guess is that it's a Gtk bug, since
it appears that programs other than Emacs are affected.
But it's just a guess, and I could be wrong.

Re: Emacs on Gtk3.8

[Achim Gratz]

Paul Eggert writes:
> My current guess is that it's a Gtk bug, since
> it appears that programs other than Emacs are affected.
> But it's just a guess, and I could be wrong.

I concur about something fishy with Gtk.  However there's also something
that has changed between the release version and current trunk (only in
degrees, but still).  Emacs 24.2 and earlier don't even start, btw.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Waldorf MIDI Implementation & additional documentation:
http://Synth.Stromeko.net/Downloads.html#WaldorfDocs

Re: Emacs on Gtk3.8

[Eli Zaretskii]

> From: Achim Gratz <Stromeko@nexgo.de>
> Date: Tue, 14 May 2013 21:42:45 +0200
> 
> Paul Eggert writes:
> > My current guess is that it's a Gtk bug, since
> > it appears that programs other than Emacs are affected.
> > But it's just a guess, and I could be wrong.
> 
> I concur about something fishy with Gtk.  However there's also something
> that has changed between the release version and current trunk (only in
> degrees, but still).  Emacs 24.2 and earlier don't even start, btw.

Of course, "something" changed!  Take a look at process.c, and you
will see that the way Emacs waits for subprocesses has been completely
rewritten, using 'waitpid' instead of 'wait'.