Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 12 Sep 2016 07:58:16 -0300
From: Dawid Golunski <>
Subject: Re: CVE-2016-6662 - MySQL Remote Root Code Execution /
 Privilege Escalation ( 0day )

Hi Alexander,

I was just going to reply to your email you sent earlier.
Thanks for the feedback. I actually updated the introduction after your email.
The advisory focuses on CVE-2016-6662 vulnerability which lets users
to modify/create my.cnf files. A fix would prevent users from writing
to my.cnf config.

And yes there's a typo in the last paragraph made after a few
sleepless nights ;) I've fixed it now.

The CVE-2016-6663 is not public yet. I refer to it in the advisory to
give some heads up in case someone wanted to discard this issue based
on reasoning that FILE privs are not common and that they will never
be pwned etc. It'll soon be published then it'll be clear what this
CVEID is about ;)


On Mon, Sep 12, 2016 at 7:35 AM, Solar Designer <> wrote:
> On Mon, Sep 12, 2016 at 06:09:10AM -0300, Dawid Golunski wrote:
>> Vulnerability: MySQL Remote Root Code Execution / Privilege Escalation 0day
>> CVE: CVE-2016-6662
>> Severity: Critical
>> Affected MySQL versions (including the latest):
>> <= 5.7.15
>> <= 5.6.33
>> <= 5.5.52
> Thank you for posting this.  For archival, and to comply with
> oss-security content guidelines, I am attaching a text/plain version of
> the above advisory (which includes a lot of detail not in your posting).
> Also, to add detail on the disclosure timeline: Dawid brought this to
> the distros list yesterday (Sunday).
> As I had pointed out in a reply on distros, it is not entirely clear
> what exact issue the CVE-2016-6662 identifier is for.  The advisory
> talks about multiple sysadmin practices, packaging issues, dangerous
> features of MySQL, and finally of safe_mysqld including the data
> directory in its search path for my.cnf.  I guess it would be most
> reasonable to have the CVE ID refer only to the latter aspect, but
> confirmation/clarification is needed.  As it is, it's unclear from the
> advisory what exact "vulnerabilities were patched by PerconaDB and
> MariaDB vendors" (the advisory says so), and it is unclear what Oracle
> and distros "fixing" CVE-2016-6662 would mean.
> Also, in this paragraph I guess the advisory wanted to refer to the
> upcoming CVE-2016-6663 (I have no idea what that issue is, beyond what
> the advisory says), like it does in a few other places:
> "It is worth to note that attackers could use one of the other vulnerabilities discovered
> by the author of this advisory which has been assigned a CVEID of CVE-2016-6662 and is
> pending disclosure. The undisclosed vulnerability makes it easy for certain attackers to
> create /var/lib/mysql/my.cnf file with arbitrary contents without the FILE privilege
> requirement."
> Alexander

Dawid Golunski

Powered by blists - more mailing lists

Please check out the Open Source Software Security Wiki, which is counterpart to this mailing list.

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.