4 messages in net.sourceforge.lists.courier-usersRe: [courier-users] imapd enters a de...
FromSent OnAttachments
hexDec 23, 2002 9:18 am 
Laurence MooreDec 23, 2002 10:23 am 
hexDec 23, 2002 3:57 pm 
Sam VarshavchikDec 23, 2002 7:11 pm 
Actions with this message:
Paste this link in email or IM:
Paste this link in email or IM:
Atom feed for this thread
Paste this URL into your reader:
Subject:Re: [courier-users] imapd enters a dead loop and loads CPUActions...
From:Laurence Moore (lmo@starwon.com.au)
Date:Dec 23, 2002 10:23:42 am
List:net.sourceforge.lists.courier-users

On Mon, 23 Dec 2002, hex wrote:

Hello there,

I'm running OpenBSD-3.1-stable-i386 with the latest version of Courier-IMAP installed from sources. I use The Bat! as client software. Checking a certain account for new messages results in an imapd process consuming all available CPU resources. The process keeps running even after the connection is dropped. I believe there is a bug in the code, but I haven't got much time to dig the sources, learn the IMAP protocol and investigate the bug on my own. I have archived the whole maildir at the moment the problem existed. I know which message file, if removed, heals the problem, but I believe that the problem might not be in that file. If this is a known problem, please point me to the resource which describes the solution, because trying to find one with Google didn't help. If there is no know solution, then I'd like to know who to contact to discuss this problem further, because I'd not like the situation to repeat. The maildir contains a number of messages which I can not make publicly available.

This appears to be a problem related to OpenBSD 2.9 and above with courier-imap 1.3.12 and above. The problem appears to be attributed to an 8-bit character in the header. The IMAP process gets in an infinite loop in rfc2047_encode_callback in rfc822/rfc2047.c.

I suspect the problem is in libc as staticaly linking imapd to libc from OpenBSD 2.7 does not reveal the problem.

The person to contact regarding this issue I believe would be the port maintainer.

Cheers,

Larry.