emacs/lisp/ChangeLog CVS lock

emacs/lisp/ChangeLog CVS lock

[Simon Josefsson]

(Copied to both cvs-hackers and savannah-hackers since I'm not sure
which one removed the lock yesterday.)

It seems to have happened again, my "cvs commit ChangeLog" has now
been hanging for 5 minutes or so, I'm not pressing C-c this time but
leave it running.

Am I causing this somehow?  Again, sorry for the trouble.

bash-2.05a$ cvs --version

Concurrent Versions System (CVS) 1.11.2 (client/server)

Copyright (c) 1989-2001 Brian Berliner, david d `zoo' zuhn,
                        Jeff Polk, and other authors

CVS may be copied only under the terms of the GNU General Public License,
a copy of which can be found with the CVS distribution kit.

Specify the --help option for further information about CVS
bash-2.05a$ cvs commit ChangeLog

Log message unchanged or not specified
a)bort, c)ontinue, e)dit, !)reuse this message unchanged for remaining dirs
Action: (continue) c

<HANGING>

Re: emacs/lisp/ChangeLog CVS lock

[Juanma Barranquero]

On Wed, 10 Jul 2002 13:22:03 +0200, Simon Josefsson <jas@extundo.com> wrote:

> It seems to have happened again, my "cvs commit ChangeLog" has now
> been hanging for 5 minutes or so, I'm not pressing C-c this time but
> leave it running.
> 
> Am I causing this somehow?  Again, sorry for the trouble.

Hmm.

I think both times you and I were commiting at the same time.

I'm not using the standard CVS, but CVSNT (from www.cvsnt.org), because
I'm on Windows and that port offers better functionality.

My CVSNT is connecting to the Savannah server through the standard
pserver functionality, so I'm not sure how that could be causing trouble.
Still, it is a weird coincidence.


                                                           /L/e/k/t/u

Re: emacs/lisp/ChangeLog CVS lock

[Francesco Potorti`]

On a related but different topic, I haven't been able to log in on
subversions for three days.  This is what I get:

bash-2.05a$ kauth
pot@GNU.ORG's Password:   <-- Password RET, wait about ten seconds
kauth: Retry count exceeded (send_to_kdc)

I wrote to system-hackers@ about 40 hours ago, but I got no response
yet.  Any suggestions?

Re: emacs/lisp/ChangeLog CVS lock

[Kim F. Storm]

Francesco Potorti` <pot@gnu.org> writes:

> On a related but different topic, I haven't been able to log in on
> subversions for three days.  This is what I get:
> 
> bash-2.05a$ kauth
> pot@GNU.ORG's Password:   <-- Password RET, wait about ten seconds
> kauth: Retry count exceeded (send_to_kdc)
> 
> I wrote to system-hackers@ about 40 hours ago, but I got no response
> yet.  Any suggestions?

I experience the same problems...
Maybe the system hackers are locked out too :-)

BTW, the stale lock is still not removed after almost 24 hours.

Couldn't the cvs server be improved to remove stale locks after, say,
1 hour?

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Mathieu Roy]

> BTW, the stale lock is still not removed after almost 24 hours.
> 
> Couldn't the cvs server be improved to remove stale locks after, say,
> 1 hour?

If you can provide the enhancement, you're welcome.

Loic Dachary and Jaime Villate, lead savannah hackers, are this week at 
the LSM in Bordeaux. This explain why the delay to fix this kind of 
troubles are longer than usual.

-- 
Mathieu Roy
  http://savannah.gnu.org/users/yeupou
http://yeupou.coleumes.org
http://gpg.coleumes.org (GPG Key)

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Andreas Schwab]

Mathieu Roy <yeupou@gnu.org> writes:

|> > BTW, the stale lock is still not removed after almost 24 hours.
|> > Couldn't the cvs server be improved to remove stale locks after, say,
|> > 1 hour?
|> 
|> If you can provide the enhancement, you're welcome.

The GCC repository is using a cron job for this.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Mathieu Roy]

Le jeu 11 jui 2002 à 14h00, Andreas Schwab a écrit :
> Mathieu Roy <yeupou@gnu.org> writes:
> 
> |> > BTW, the stale lock is still not removed after almost 24 hours.
> |> > Couldn't the cvs server be improved to remove stale locks after,
> say,
> |> > 1 hour?
> |>
> |> If you can provide the enhancement, you're welcome.
> 
> The GCC repository is using a cron job for this.

Are you willing to take care of this or write us the way they do it 
precisely?
It would be nice.

Regards,


-- 
Mathieu Roy
  http://savannah.gnu.org/users/yeupou
http://yeupou.coleumes.org
http://gpg.coleumes.org (GPG Key)

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Andreas Schwab]

Mathieu Roy <yeupou@gnu.org> writes:

|> Le jeu 11 jui 2002 à 14h00, Andreas Schwab a écrit :
|> > Mathieu Roy <yeupou@gnu.org> writes:
|> > |> > BTW, the stale lock is still not removed after almost 24 hours.
|> > |> > Couldn't the cvs server be improved to remove stale locks after,
|> > say,
|> > |> > 1 hour?
|> > |>
|> > |> If you can provide the enhancement, you're welcome.
|> > The GCC repository is using a cron job for this.
|> 
|> Are you willing to take care of this or write us the way they do it
|> precisely?
|> It would be nice.

I'd suggest to ask overseers@gcc.gnu.org for their experience.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Jeff Bailey]

On Thu, Jul 11, 2002 at 02:00:32PM +0200, Andreas Schwab wrote:

> |> > BTW, the stale lock is still not removed after almost 24 hours.
> |> > Couldn't the cvs server be improved to remove stale locks after, say,
> |> > 1 hour?
> |> 
> |> If you can provide the enhancement, you're welcome.

> The GCC repository is using a cron job for this.

Can you provide the source to the cron job?

-- 
I reincarnated for this?

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Andreas Schwab]

Jeff Bailey <jbailey@nisa.net> writes:

|> On Thu, Jul 11, 2002 at 02:00:32PM +0200, Andreas Schwab wrote:
|> 
|> > |> > BTW, the stale lock is still not removed after almost 24 hours.
|> > |> > Couldn't the cvs server be improved to remove stale locks after, say,
|> > |> > 1 hour?
|> > |> 
|> > |> If you can provide the enhancement, you're welcome.
|> 
|> > The GCC repository is using a cron job for this.
|> 
|> Can you provide the source to the cron job?

I don't have it, please ask overseers@gcc.gnu.org for it.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Francesco Potorti`]

   > Couldn't the cvs server be improved to remove stale locks after, say,
   > 1 hour?
   
   If you can provide the enhancement, you're welcome.

I am *not* a CVS hacker.  But I suspect that the following line in a
crontab would do the trick:

MAILTO=cvs-hackers@gnu.org
*/5 * * * * find /cvsroot -name '#cvs.[trw]fl.*' -o -name '#cvs.lock' \
			  -mmin +20 -exec rm --force --verbose {} \;

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Mathieu Roy]

Thanks you, we'll see this and probably add this to the crontab.

Regards,


Le jeu 11 jui 2002 à 14h11, Francesco Potorti` a écrit :
>    > Couldn't the cvs server be improved to remove stale locks after,
> say,
>    > 1 hour?
> 
>    If you can provide the enhancement, you're welcome.
> 
> I am *not* a CVS hacker.  But I suspect that the following line in a
> crontab would do the trick:
> 
> MAILTO=cvs-hackers@gnu.org
> */5 * * * * find /cvsroot -name '#cvs.[trw]fl.*' -o -name '#cvs.lock'
> \
> 			  -mmin +20 -exec rm --force --verbose {} \;
> 
> _______________________________________________
> Savannah-hackers mailing list
> Savannah-hackers@gnu.org
> http://mail.gnu.org/mailman/listinfo/savannah-hackers
> 

-- 
Mathieu Roy
  http://savannah.gnu.org/users/yeupou
http://yeupou.coleumes.org
http://gpg.coleumes.org (GPG Key)

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Miles Bader]

Francesco Potorti` <pot@gnu.org> writes:
> MAILTO=cvs-hackers@gnu.org
> */5 * * * * find /cvsroot -name '#cvs.[trw]fl.*' -o -name '#cvs.lock' \
> 			  -mmin +20 -exec rm --force --verbose {} \;

Is `#cvs.lock' a per-file lock, a per-directory lock, or a per-session lock?

Remember that updating over a dialup can be _very slow_, so 20 minutes
is certainly too short a time for either of the latter two.

-Miles
-- 
o The existentialist, not having a pillow, goes everywhere with the book by
  Sullivan, _I am going to spit on your graves_.

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Mathieu Roy]

Well, one hour should be ok.


Le ven 12 jui 2002 à  3h57, Miles Bader a écrit :
> Francesco Potorti` <pot@gnu.org> writes:
> > MAILTO=cvs-hackers@gnu.org
> > */5 * * * * find /cvsroot -name '#cvs.[trw]fl.*' -o -name
> '#cvs.lock' \
> > 			  -mmin +20 -exec rm --force --verbose {} \;
> 
> Is `#cvs.lock' a per-file lock, a per-directory lock, or a per-session
> lock?
> 
> Remember that updating over a dialup can be _very slow_, so 20 minutes
> is certainly too short a time for either of the latter two.
> 
> -Miles
> --
> o The existentialist, not having a pillow, goes everywhere with the
> book by
>   Sullivan, _I am going to spit on your graves_.
> 
> _______________________________________________
> Savannah-hackers mailing list
> Savannah-hackers@gnu.org
> http://mail.gnu.org/mailman/listinfo/savannah-hackers
> 

-- 
Mathieu Roy
  http://savannah.gnu.org/users/yeupou
http://yeupou.coleumes.org
http://gpg.coleumes.org (GPG Key)

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Miles Bader]

Mathieu Roy <yeupou@gnu.org> writes:
> > Is `#cvs.lock' a per-file lock, a per-directory lock, or a
> > per-session lock?  Remember that updating over a dialup can be _very
> > slow_, so 20 minutes is certainly too short a time for either of the
> > latter two.
>
> Well, one hour should be ok.

When is the lock held?  Is it only while updating a single file?
Then it should be OK.

-Miles
-- 
P.S.  All information contained in the above letter is false,
      for reasons of military security.

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Mathieu Roy]

Le ven 12 jui 2002 à  8h58, Miles Bader a écrit :
> Mathieu Roy <yeupou@gnu.org> writes:
> > > Is `#cvs.lock' a per-file lock, a per-directory lock, or a
> > > per-session lock?  Remember that updating over a dialup can be
> _very
> > > slow_, so 20 minutes is certainly too short a time for either of
> the
> > > latter two.
> >
> > Well, one hour should be ok.
> 
> When is the lock held?  Is it only while updating a single file?
> Then it should be OK.

In fact, this is more complex than just removing a lock file. We need 
to remove the process that create the lock file too.


> 
> -Miles
> --
> P.S.  All information contained in the above letter is false,
>       for reasons of military security.
> 
> _______________________________________________
> Savannah-hackers mailing list
> Savannah-hackers@gnu.org
> http://mail.gnu.org/mailman/listinfo/savannah-hackers
> 

-- 
Mathieu Roy
  http://savannah.gnu.org/users/yeupou
http://yeupou.coleumes.org
http://gpg.coleumes.org (GPG Key)

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Kim F. Storm]

Mathieu Roy <yeupou@gnu.org> writes:

> 
> In fact, this is more complex than just removing a lock file. We need
> to remove the process that create the lock file too.

Which is why any sensible application writes its own PID in the
lock files so other applications can easily check whether the
lock is still valid by doing a check like

        while (file_exists(LOCK_FILE)) {
                pid = atoi(read_first_line_of_file(LOCK_FILE));
                if (pid <= 0) {
                        /* lock file is incomplete,
                           wait for locker to write its pid. */
                        sleep(1);
                        continue;
                }
                if (kill(pid, 0) == 0) {
                        /* lock is valid */
                        sleep(5);
                        continue;
                }
                /* remove stale lock and go on */
                unlink(LOCK_FILE);
        }
        /* Now it is our turn to create the file */
        
-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: [Savannah-hackers] Re: emacs/lisp/ChangeLog CVS lock

[Francesco Potorti`]

   > MAILTO=cvs-hackers@gnu.org
   > */5 * * * * find /cvsroot -name '#cvs.[trw]fl.*' -o -name '#cvs.lock' \
   > 			  -mmin +20 -exec rm --force --verbose {} \;
   
   Is `#cvs.lock' a per-file lock, a per-directory lock, or a per-session lock?

As I said, I am *not* a cvs expert.  I just read the cvs man page and
extracted all the names of lock files from the FILES section, then put
them together in a find command.  Sorry, I don't know more offhand.

GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Ken Raeburn]

Francesco Potorti` <pot@gnu.org> writes:

> On a related but different topic, I haven't been able to log in on
> subversions for three days.  This is what I get:
>
> bash-2.05a$ kauth
> pot@GNU.ORG's Password:   <-- Password RET, wait about ten seconds
> kauth: Retry count exceeded (send_to_kdc)
>
> I wrote to system-hackers@ about 40 hours ago, but I got no response
> yet.  Any suggestions?


The Kerberos server seems to be online now, though some (all?) of its
data has been trashed.  No more raeburn@GNU.ORG any more, and
cross-realm authentication from ATHENA.MIT.EDU is broken too; I can't
even get tickets from the GNU KDC.

I talked to a guy who knows some people etc, and he said one of the
FSF machines was broken into, one for general remote login access.
Not the Kerberos server, so I don't know why it's been changed.  And I
don't know *when* the breakin happened, but three days of down time
seems pretty lame.

It also appears that Kerberos rsh access to the CVS server has been
shut off.  That's how I'd been accessing it for years, since before
the FSF ran ssh servers.[1] So I had a whole bunch of problems in my
way when I was trying yesterday....

I haven't seen any hints as to whether/when they'll restore the
Kerberos database, or whether/when they'll restore Kerberos rsh
access.

Don't you just love how they keep us so well informed, and in such a
timely fashion?

</complain>

Ken

[1] Some of you may know I'm a Kerberos guy, and I've done some CVS
hacking in the past.  I can go into why I don't use Kerberos CVS
access to the FSF if you care, but I think it's a tangent.

Re: GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Miles Bader]

Ken Raeburn <raeburn@raeburn.org> writes:
> Don't you just love how they keep us so well informed, and in such a
> timely fashion?

They are working to restore things (I just got ssh access back yesterday).

There was an email message, but just to people who complained; it said:
> This is a preliminary message going to those who emailed to us today
> about fencepost.  A more thought-out message will go out tomorrow or
> Wednesday.

[I've not seen any wider message yet, though]

-Miles
-- 
A zen-buddhist walked into a pizza shop and
said, "Make me one with everything."

Re: GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Jeff Bailey]

On Thu, Jul 11, 2002 at 10:32:25PM -0400, Ken Raeburn wrote:

> > I wrote to system-hackers@ about 40 hours ago, but I got no response
> > yet.  Any suggestions?

Sorry, we've been a bit busy trying to make sure that we get
everything up and running.

> The Kerberos server seems to be online now, though some (all?) of its
> data has been trashed.  No more raeburn@GNU.ORG any more, and
> cross-realm authentication from ATHENA.MIT.EDU is broken too; I can't
> even get tickets from the GNU KDC.

I haven't seen a note to accounts or system-hackers from you yet.  Did
I miss one?

I don't know if cross-realm auth from ATHENA.MIT.EDU will be
preserved.  We don't really have useful contacts there these days, so
it's hard to maintain that as a trust relationship.

> I talked to a guy who knows some people etc, 

> and he said one of the FSF machines was broken into, one for general
> remote login access.  Not the Kerberos server, so I don't know why
> it's been changed.  And I don't know *when* the breakin happened,
> but three days of down time seems pretty lame.

I haven't seen an email from you offering help either.  Did I miss
that too?

> It also appears that Kerberos rsh access to the CVS server has been
> shut off.

How are you testing this if you can't get a ticket?

> That's how I'd been accessing it for years, since before the FSF ran
> ssh servers.[1] So I had a whole bunch of problems in my way when I
> was trying yesterday....

And you're welcome to keep accessing it that way.  Although we'll be
offering Kerberised SSHv2 shortly as well.

Any other questions?  The rest of your email appeared to just be a
rant.

-- 
I reincarnated for this?

Re: GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Ken Raeburn]

Jeff Bailey <jbailey@nisa.net> writes:
> On Thu, Jul 11, 2002 at 10:32:25PM -0400, Ken Raeburn wrote:
>
>> > I wrote to system-hackers@ about 40 hours ago, but I got no response
>> > yet.  Any suggestions?
>
> Sorry, we've been a bit busy trying to make sure that we get
> everything up and running.
>
>> The Kerberos server seems to be online now, though some (all?) of its
>> data has been trashed.  No more raeburn@GNU.ORG any more, and
>> cross-realm authentication from ATHENA.MIT.EDU is broken too; I can't
>> even get tickets from the GNU KDC.
>
> I haven't seen a note to accounts or system-hackers from you yet.  Did
> I miss one?

Well, there was the email you just responded to, to system-hackers and
Francesco among others, telling him what I've observed and heard about
the situation, since he said he hadn't gotten a response to his email
in nearly two days.

> I don't know if cross-realm auth from ATHENA.MIT.EDU will be
> preserved.  We don't really have useful contacts there these days, so
> it's hard to maintain that as a trust relationship.

Well, there's me... but my only useful contact on your end for
Kerberos stuff has been Joel.  He's been talking to me (and maybe Tom
Yu?) about Kerberos stuff on occasion, as needed.

Why did the entries get deleted, anyways?  Did the Kerberos server get
broken into after all?


> I haven't seen an email from you offering help either.  Did I miss
> that too?

If you need help on Kerberos related stuff, I'd be happy to give some.
I could even make half an argument that debugging Kerberos problems at
your site is kind of job-related, maybe.

Non-Kerberos stuff, well, that'd depend on my free time, and my coming
up to speed on what you're doing, etc, etc.

> How are you testing this if you can't get a ticket?

"telnet subversions.gnu.org 544" says "connection refused"; 544 is
Kerberos rsh.

> And you're welcome to keep accessing it that way.  Although we'll be
> offering Kerberised SSHv2 shortly as well.

That's good news, thank you.

> Any other questions?  The rest of your email appeared to just be a
> rant.

Yes, I'm sorry about the tone, but being shut out for a couple of
days, with no information I could find on why, and (from Francesco's
comments) no information coming from sysadmins, I was feeling pretty
frustrated.

Ken

Re: GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Joel N. Weber II]

   Well, there's me... but my only useful contact on your end for
   Kerberos stuff has been Joel.  He's been talking to me (and maybe Tom
   Yu?) about Kerberos stuff on occasion, as needed.

Sam Hartman also provided me a lot of assistance in initially setting
up kerberos.  I think I've only interacted with tlyu with actually
setting up the shared key.

Re: GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Jeff Bailey]

On Fri, Jul 12, 2002 at 02:00:17AM -0400, Ken Raeburn wrote:

> If you need help on Kerberos related stuff, I'd be happy to give
> some.  I could even make half an argument that debugging Kerberos
> problems at your site is kind of job-related, maybe.

We're having some weirdness with ticket passing and ssh.  Do you have
experience with this?

(Now I'm curious what your job is.  I know I've worked with you on
something in the last 7 or 8 years, but I can't pinpoint exactly
what).

> Yes, I'm sorry about the tone, but being shut out for a couple of
> days, with no information I could find on why, and (from Francesco's
> comments) no information coming from sysadmins, I was feeling pretty
> frustrated.

I totally understand.  I also probably came out stronger than I
expected - So many people have tended to choose the 'flame-first'
communication method. =)

-- 
I reincarnated for this?

Re: GNU Kerberos breakage (was Re: emacs/lisp/ChangeLog CVS lock)

[Ken Raeburn]

Jeff Bailey <jbailey@nisa.net> writes:
> We're having some weirdness with ticket passing and ssh.  Do you have
> experience with this?

Not much, but if you forward me the details (I don't think all these
lists need to be involved), I'll be glad to look into it.

> (Now I'm curious what your job is.  I know I've worked with you on
> something in the last 7 or 8 years, but I can't pinpoint exactly
> what).

I'm one of the Kerberos developers at MIT.  Haven't been there all
that long, but a few years ago while I was at Cygnus I was also
working part of the time on Kerberos, when I wasn't on gcc or binutils
work.

Ken

Re: emacs/lisp/ChangeLog CVS lock

[Simon Josefsson]

Ok, so it happened again.  Could someone please remove the lock?

Same file and same CVS version used as in the report below...

I'm terribly sorry for the trouble, but I think the frequency of this
suggest there is a real bug somewhere.  I use CVS extensively and this
is the only one server and only one file I have had these troubles
with.  Note that I can commit to other files in the Emacs CVS without
problems.  Any ideas?

This is on Debian Sid.

Simon Josefsson <jas@extundo.com> writes:

> (Copied to both cvs-hackers and savannah-hackers since I'm not sure
> which one removed the lock yesterday.)
>
> It seems to have happened again, my "cvs commit ChangeLog" has now
> been hanging for 5 minutes or so, I'm not pressing C-c this time but
> leave it running.
>
> Am I causing this somehow?  Again, sorry for the trouble.
>
> bash-2.05a$ cvs --version
>
> Concurrent Versions System (CVS) 1.11.2 (client/server)
>
> Copyright (c) 1989-2001 Brian Berliner, david d `zoo' zuhn,
>                         Jeff Polk, and other authors
>
> CVS may be copied only under the terms of the GNU General Public License,
> a copy of which can be found with the CVS distribution kit.
>
> Specify the --help option for further information about CVS
> bash-2.05a$ cvs commit ChangeLog
>
> Log message unchanged or not specified
> a)bort, c)ontinue, e)dit, !)reuse this message unchanged for remaining dirs
> Action: (continue) c
>
> <HANGING>
>
>
> _______________________________________________
> Emacs-devel mailing list
> Emacs-devel@gnu.org
> http://mail.gnu.org/mailman/listinfo/emacs-devel

Re: emacs/lisp/ChangeLog CVS lock

[Simon Josefsson]

Someone removed the lock, thanks.

FYI: I found a bug report a few days old about this, and it seems to
be a known problem with CVS from Debian testing and Debian unstable,
see:

http://bugs.debian.org/157881

CVS from Debian stable works.

Simon Josefsson <jas@extundo.com> writes:

> Ok, so it happened again.  Could someone please remove the lock?
>
> Same file and same CVS version used as in the report below...
>
> I'm terribly sorry for the trouble, but I think the frequency of this
> suggest there is a real bug somewhere.  I use CVS extensively and this
> is the only one server and only one file I have had these troubles
> with.  Note that I can commit to other files in the Emacs CVS without
> problems.  Any ideas?
>
> This is on Debian Sid.
>
> Simon Josefsson <jas@extundo.com> writes:
>
>> (Copied to both cvs-hackers and savannah-hackers since I'm not sure
>> which one removed the lock yesterday.)
>>
>> It seems to have happened again, my "cvs commit ChangeLog" has now
>> been hanging for 5 minutes or so, I'm not pressing C-c this time but
>> leave it running.
>>
>> Am I causing this somehow?  Again, sorry for the trouble.
>>
>> bash-2.05a$ cvs --version
>>
>> Concurrent Versions System (CVS) 1.11.2 (client/server)
>>
>> Copyright (c) 1989-2001 Brian Berliner, david d `zoo' zuhn,
>>                         Jeff Polk, and other authors
>>
>> CVS may be copied only under the terms of the GNU General Public License,
>> a copy of which can be found with the CVS distribution kit.
>>
>> Specify the --help option for further information about CVS
>> bash-2.05a$ cvs commit ChangeLog
>>
>> Log message unchanged or not specified
>> a)bort, c)ontinue, e)dit, !)reuse this message unchanged for remaining dirs
>> Action: (continue) c
>>
>> <HANGING>
>>
>>
>> _______________________________________________
>> Emacs-devel mailing list
>> Emacs-devel@gnu.org
>> http://mail.gnu.org/mailman/listinfo/emacs-devel
>
>
>
> _______________________________________________
> Emacs-devel mailing list
> Emacs-devel@gnu.org
> http://mail.gnu.org/mailman/listinfo/emacs-devel