On 06/07/2010 08:28 PM Timo Sirainen wrote:
> http://hg.dovecot.org/dovecot-2.0/rev/2f084e625f09 and hopefully
> http://hg.dovecot.org/dovecot-2.0/rev/fd447208ccb9
The latest commits from today solved the problem.
Regards,
Pascal
--
The trapper recommends today: beeffeed.1015...@localdomain.org
On la, 2010-06-05 at 01:07 +0200, Pascal Volk wrote:
> On 06/04/2010 05:16 PM Timo Sirainen wrote:
> > This really should fix it:
> > http://hg.dovecot.org/dovecot-2.0/rev/b578861bfa40
>
> Sorry, didn't help yet, looks worse now -> dbl segfault.
> Backtraces are attached.
http://hg.dovecot.org/do
On 06/04/2010 05:16 PM Timo Sirainen wrote:
> This really should fix it:
> http://hg.dovecot.org/dovecot-2.0/rev/b578861bfa40
Sorry, didn't help yet, looks worse now -> dbl segfault.
Backtraces are attached.
Regards,
Pascal
--
The trapper recommends today: beeffeed.1015...@localdomain.org
#0 0
On to, 2010-06-03 at 07:50 +0200, Pascal Volk wrote:
> On 06/02/2010 07:39 PM Pascal Volk wrote:
> > The 'fix', which made auth crashing, was intended to solve another
> > issue. But the other issue should be also fixed with that commit:
> > http://hg.dovecot.org/dovecot-2.0/rev/1e890076c4e9
>
> H
On 06/02/2010 07:39 PM Pascal Volk wrote:
> The 'fix', which made auth crashing, was intended to solve another
> issue. But the other issue should be also fixed with that commit:
> http://hg.dovecot.org/dovecot-2.0/rev/1e890076c4e9
Hm, seems not be fixed. Now my installation is up to date (2.0.bet
Am 02.06.2010 um 19:39 schrieb Pascal Volk:
> huh, didn't it solve your problem completely? Crashes the auth process
> still when your lmtp performs a userdb lookup?
No it doesn't crash anymore, but it sounded like the commit you listed was
reverting a partial fix, hence I considered it a rollba
On 06/02/2010 07:14 PM Thomas Leuxner wrote:
>
> I can confirm that above patch mitigates the situation. Would it still help
> if I'd debug the crash or has the source been tracked down?
huh, didn't it solve your problem completely? Crashes the auth process
still when your lmtp performs a userdb
Am 02.06.2010 um 02:35 schrieb Pascal Volk:
>
> Try this: http://hg.dovecot.org/dovecot-2.0/rev/dc94f9240ec1
I can confirm that above patch mitigates the situation. Would it still help if
I'd debug the crash or has the source been tracked down?
Thomas
On 06/01/2010 10:02 PM Thomas Leuxner wrote:
> ==> /var/log/dovecot.log <==
> Jun 1 21:52:30 spectre dovecot: lmtp(6788): Connect from local
> Jun 1 21:52:30 spectre dovecot: master: Error: service(auth): child 6797
> killed with signal 11 (core not dumped)
> Jun 1 21:52:30 spectre dovecot: lmt
On ti, 2010-06-01 at 22:02 +0200, Thomas Leuxner wrote:
> Jun 1 21:52:30 spectre dovecot: master: Error: service(auth): child 6797
> killed with signal 11 (core not dumped)
Could you get gdb backtrace? http://dovecot.org/bugreport.html#coredumps
Hi,
not sure which change caused it, but one of the latest adds between those two
snapshots breaks userdb lookup:
Jun 1 21:53:30 spectre dovecot: master: Dovecot v2.0.beta5 (1b45f9641924)
starting up
Jun 1 21:53:30 spectre dovecot: master: Warning: Killed with signal 15 (by
pid=6898 uid=0
11 matches
Mail list logo
