Skip to content
Phil Pennock edited this page May 28, 2014 · 23 revisions

Exim Security

Much consideration of Exim's security is given in Chapter 54 - Security considerations of The Exim Specification. This includes suggested hardening steps.

Vulnerability History

Note that a "remote code execution as Exim run-time user" vulnerability can be combined with a privilege escalation attack to become even more serious.

  • CVE-2014-2957 fixed in 4.82.1, introduced in 4.82: used untrusted data when parsing the From header in Experimental DMARC code and allowed macro expansion. Details post
  • CVE-2012-5671 fixed in 4.80.1, introduced in 4.70: buffer overflow vulnerability in DKIM DNS response processing. Impact: remote code execution as Exim run-time user. Details post
  • CVE-2011-1764 fixed in 4.76, introduced in 4.70: format string attack in DKIM processing. Impact: remote code execution as Exim run-time user. Bugzilla 1106.
  • CVE-2011-1407 fixed in 4.76, introduced in 4.70: flaw in handling DKIM DNS records. Impact: remote code execution as Exim run-time user
  • CVE-2011-0017 fixed in 4.73: return values of setuid()/setgid() not checked; only an issue on Linux. Impact: privilege escalation from Exim run-time user to root
  • CVE-2010-4345 fixed in 4.73: Exim privilege escalation from Exim run-time user to root via configuration overrides
  • CVE-2010-2023 fixed in 4.72: Hardlink attack via sticky mbox directory. Impact: overwrite files of target user on same partition as mbox directory. Bugzilla 988.
  • CVE-2010-2024 fixed in 4.72: Symlink attack in /tmp for MBX locking algorithm. Bugzilla 989.
  • CVE-2010-4344 fixed in 4.70: buffer overflow in string_format(). Impact: remote code execution as Exim run-time user. Bugzilla 787.

See Also

  • [[Security Release Process][SecurityReleaseProcess]]
Clone this wiki locally