[exim] Exim 4.88 segfault on CentOS 6

Top Page
Delete this message
Reply to this message
Author: Kunszt Árpád
Date:  
To: exim-users
Subject: [exim] Exim 4.88 segfault on CentOS 6
Hello!

I'm using Exim from the EPEL repository on CentOS 6. After I upgraded to 4.88 I'm starting to have some segfaults in the logs.

2017-02-27T09:12:07+01:00 mail kernel: exim[30289]: segfault at 0 ip 00007fbaffd35790 sp 00007ffc17cac210 error 4 in exim[7fbaffc54000+120000]

I couldn't create a core file yet (I don't know why it's not working) but I could able to get a debug output, this is only the last few lines, if necessary I can post the whole output but I have to anonymize it first (LDAP passwords, internal e-mail addresses etc.)

09:12:07 30289 Data file name: /var/spool/exim//input//1ciGPf-0007sX-Id-D
09:12:07 30289 SMTP>> 250 538 byte chunk received
09:12:07 30289 chunking state 0
09:12:07 30289 SMTP<< BDAT 919448 LAST
09:12:07 30289 chunking state 2, 919448 bytes
09:12:07 30277 child 30289 ended: status=0xb

The client is another Exim - same version, same OS. The e-mails comes from a script I wrote and it uses the sendmail -t -oi command to create an e-mail.
After I disabled the chunking for this client there wasn't any segfaults but it doesn't happen everyday so I'm just hoping now.

chunking_advertise_hosts = ! 10.0.2.1

Is it a known issue? Is it worth to fight to create a core dump or do I wait for the upcoming 4.89 and try with that one?

Thanks,

Árpád Kunszt

PS: Sorry for my English.