bug#14683: 23.1; shell freezes during hunspell dictionary check

[Petr Hracek]

On 06/24/2013 12:59 PM, Agustin Martin wrote:
> [Sending also to the bug address, forgot this in my first reply. Please
> reply to the bug address]
>
> On Mon, Jun 24, 2013 at 10:01:01AM +0200, Petr Hracek wrote:
> > On 06/21/2013 04:53 PM, Agustin Martin wrote:
> > > On Fri, Jun 21, 2013 at 09:33:44AM +0200, Petr Hracek wrote:
> > > > Hi folks,
> > > >
> > > > In emacs version 23.1
> > > > I have a following problem with following scenario:
> > > >
> > > > 1.Press M-x shell
> > > > 2.In the shell input echo "Hello wolrd" then press Enter
> > > > 3.Press M-x ispell-buffer
> > > > 4.Press 0 twice
> > > > wolrd has been changed to world, this is fine
> > > > 5. input echo "heavon" then press Enter
> > > > 6. Press M-x ispell-buffer
> > > >
> > > > A message continuing spelling check using hunspell with default
> > > > dictionary appears but the window freezes
> > > Hi,
> > >
> > > I cannot reproduce this neither with current emacs-snapshot from FSF repo
> > > nor with Debian/sid emacs24 (using backported bleeding edge ispell.el).
> > >
> > > However, I see the problem with old Debian/sid emacs23 (also using
> > > backported bleeding edge ispell.el), so may be related to a bug in that
> > > old Emacs version.
> > >
> > > Can you reproduce the problem with recent Emacs?
> > this problem can be observed only in emacs23.
> > Could you please send me a bug in debian side and whether it has
> > been solved?
> Nothing special was done in Debian side about this, so I guess this is
> related to Emacs 23 itself. Are you using ispell.el as shipped with Emacs 23?
Yes I am using ispell.el shipped with emacs.
But how to debug that case? Any help?
I can use gdb but I would like to debug ispell.el firstly and afterwards
hunspell.cxx code.
> > In the emacs24 all is working properly.
> >
> > I would like to test whether problem is on emacs or in hunspell side but
> > I do not know how?
> I tested with the same hunspell and roughly similar ispell.el (pristine for
> emacs-snapshot and with the same backport changes for Emacs 23 and 24),
> finding the problem only in Emacs 23.
>
> I also tested with pristine Emacs 23 and I can also reproduce the problem,
> so I guess the problem is in the emacs side.
Hmm, is there any way how to find out the relevant commit.
I know that it is more or less pretty old version.
> Since Emacs 23 is an old version, unless somebody points to the Emacs commit
> addressing this, I do not expect this to be addressed.
>
> Regards,


--
Best regards / S pozdravem
Petr Hracek