A Critical vulnerability recently addressed in the popular open-source email server Exim could lead to remote code execution.
Exim is an open source mail transfer agent (MTA) widely used in systems running Linux and macOS. At the moment, Exim powers over half of email servers out there.
Tracked as CVE-2019-16928, the newly addressed security vulnerability is a heap-based buffer overflow in string_vformat (string.c).
The issue, Exim explains in an advisory, is that an attacker could send a long EHLO string to crash the Exim process that receives the message.
“While at this mode of operation Exim already dropped its privileges, other paths to reach the vulnerable code may exist,” Exim notes.
By exploiting the vulnerability, the attacker could not only deny access to the Exim service, but could also potentially execute code remotely on the affected server, security researchers with the QAX A-TEAM explain in a post on the Exim bug tracker.
Proof-of-concept code has been published as well, to exemplify the use of an extraordinary long EHLO string to trigger the crash.
The security flaw was introduced with the release of Exim 4.92 and does not impact previous versions of the email server. However, versions 4.92, 4.92.1, and 4.92.2 are known to be vulnerable.
Exim addressed the vulnerability with the release of version 4.92.3 of the server.
In the beginning of September, Exim addressed a vulnerability (CVE-2019-15846) that allowed attackers to remotely execute arbitrary code with root privileges.
Earlier this year, the development team addressed a vulnerability (CVE-2019-10149) that allowed for the execution of arbitrary commands with execv(), as root. The first attacks targeting the flaw were observed within weeks, but the attacks haven’t stopped, and botnets also attempt to exploit it.