popa3d-users - Re: "failed or refused to load"

Openwall Project		/home Owl JtR Pro crypt pam_passwdqc tcb phpass scanlogd popa3d msulogin / Linux BIND / advisories presentations / services donations / wordlists passwords / news community lists wiki CVSweb mirrors signatures
bringing security into open environments

Password Recovery Resources on the Net

[<prev] [next>] [<thread-prev] [thread-next>] [month] [year] [list]

Date: Wed, 30 Oct 2002 09:24:16 +0100
From: Hatto von Hatzfeld <hatto@...esianer.de>
To: popa3d-users@...ts.openwall.com
Subject: Re: "failed or refused to load"

On Wed, Oct 30, 2002 at 05:06:02AM +0300, Solar Designer wrote:
> On Tue, Oct 29, 2002 at 07:03:27PM +0100, Hatto von Hatzfeld wrote:
> 
> > For some reason (a problem of procmail or formail, or of popa3 itself?)
> 
> Please mention the versions of everything involved, including the OS.

Sorry for having forgotten this. 
OS Linux SuSE 7.3, Kernel 2.4.10-4GB
sendmail 8.11.6
procmail/formail v3.15.1 2001/01/08
popa3d 0.5 (Oct 2001)
 
> > in my mailboxes there is sometimes (hard to verify under which
> > conditions) an initial empty line (ASCII 10), either before the first
> > "From " or as only contents of a mailbox.
> 
> I suggest that you start by making sure all of the software on your
> system that may access mailboxes uses compatible locking.

As far as I know all of them use flock(). Especially procmail certainly
does it.
 
> > When a pop3 client tries to fetch mails, popa3 crashes and sends a
> >    Failed or refused to load /var/spool/mail/somebox
> > to syslog.
> 
> Best is to locate and eliminate the problem such that mailboxes aren't
> corrupted in this way.

Since only procmail and popa3d access the mailboxes...

OK, it seems that the problem increased since I made procmail (resp.
formail) insert a header line in every message. So probably I have to
blame procmail or formail.
 
> > Or could someone tell me what to do to convince popa3d to delete
> > initial empty lines in a mailbox?
> 
> That would be a modification to the state machine implemented in
> mailbox.c: mailbox_parse(). Not too hard to do, but still some work.

Sigh; I finally should start learning a bit of C...

Thanks and bye,
Hatto v. Hatzfeld