Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Thu, 06 Nov 2014 19:34:37 -0800
From: Russ Allbery <eagle@...ie.org>
To: Andreas Barth <aba@...us.org>
Cc: oss-security@...ts.openwall.com,  742140@...s.debian.org
Subject: Re: Re: Bug#742140: libpam-oath: PAM module does not check whether strdup allocations succeeded

Andreas Barth <aba@...us.org> writes:

> we have the following debian bug report about an security isuse in
> libpam-oath (source oath-toolkit, upstream web page
> http://www.nongnu.org/oath-toolkit/ ).

> What is the appropriate process to get an CVE number on it? This issue
> is already public, as it is documented in the debian bug tracking
> system.

Is not checking memory allocations for failure in this fashion considered
CVE-worthy?  I'm probably missing something, but this seems difficult to
exploit: the first strdup is only trying to allocate a byte of memory, and
the second will not allocate more than MAX_OTP_LEN memory due to an
earlier check.  This means the attacker would have to have essentially
exhausted system memory already to force strdup to return NULL.

And, even if that happens, strdup returns NULL, which leads immediately to
a NULL pointer dereference and presumably a process crash.  But to create
this situation, the attacker has to nearly exhaust all process memory, and
could just go a step farther and exhaust all memory, which would almost
certainly result in a process crash anyway, or an OOM kill.

Am I overlooking something?

-- 
Russ Allbery (eagle@...ie.org)              <http://www.eyrie.org/~eagle/>

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.