GNU bug report logs - #14333
24.3.50; Emacs hangs when trying to exit

Previous Next

Full log

Message #44 received at 14333 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Dani Moncayo <dmoncayo <at> gmail.com>
Cc: 14333 <at> debbugs.gnu.org
Subject: Re: bug#14333: 24.3.50; Emacs hangs when trying to exit
Date: Thu, 16 May 2013 10:35:04 +0300

> Date: Thu, 16 May 2013 09:18:16 +0200
> From: Dani Moncayo <dmoncayo <at> gmail.com>
> Cc: 14333 <at> debbugs.gnu.org
> 
> > Grepping through Lisp files, it looks like these features are the ones
> > that use server sockets which could be involved in this issue:
> >
> >   Emacs server (server.el)
> >   ERC
> >   Gnus
> >   Tramp (unlikely, since it only uses server connections for some rare cases)
> >   org-irc.el
> >
> > If you use any of these, they are the potential culprits.
> 
> Of those, the only feature I think I use is Emacs server, because I
> have `(server-start)' in my init file.

That's what I thought.  So, if you start "emacs -Q" and then manually
start the server, does Emacs hang when exiting right after that?

> And FWIW, the hanged session I reported yesterday didn't have any
> client session connected to that server session (I rarely start client
> sessions).

That's expected, because this code near the place where the main
thread was stuck:

  if (winsock_lib != NULL && winsock_inuse == 0)
    {
      /* Not sure what would cause WSAENETDOWN, or even if it can happen
	 after WSAStartup returns successfully, but it seems reasonable
	 to allow unloading winsock anyway in that case. */
      if (pfn_WSACleanup () == 0 ||
	  pfn_WSAGetLastError () == WSAENETDOWN)
	{

clearly shows that winsock_inuse must be zero, which means we already
closed all the socket connections.  The question is, why the reader
thread of one of these connections is still waiting?

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.