Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 5 Apr 2011 10:35:47 -0400 (EDT)
From: "Steven M. Christey" <coley@...-smtp.mitre.org>
To: oss-security@...ts.openwall.com
cc: Josh Bressers <bressers@...hat.com>,
        "Steven M. Christey" <coley@...-smtp.mitre.org>,
        Eugene Teo <eugene@...hat.com>
Subject: Re: CVE request: kernel: multiple issues in ROSE


On Mon, 4 Apr 2011, Dan Rosenberg wrote:

> I think one makes sense, since all the problems were in a single
> protocol and were addressed at the same time.

These days, some of the main criteria for splitting into separate CVEs 
are:

(1) different bug types [where we can argue about whether 
memory-corruption/overflow variants are all "one type"]

(2) different affected versions [using best-available knowledge of 
affected versions at the time of assignment]

(3) different researchers [this being a relatively recent development in 
CVE-land, which ultimately creates more CVEs but is also much more useful 
in many scenarios]

When you get a large number of bugs/fixes at one time, these criteria are 
sometimes relaxed for the sake of usability to CVE consumers (by keeping 
the number of CVE's reasonably low), and keeping the amount of analysis 
time relatively low.  In the case of the ROSE patches, there's a fairly 
large number but it doesn't seem too bad.

CVE consistency can't be guaranteed all the time, but where reasonable, it 
makes sense to follow the guidelines.

- Steve



>
> -Dan
>
>> Thanks.
>>
>> --
>>    JB
>>
>> ----- Original Message -----
>>> Any update on this?
>>>
>>> Thanks,
>>> Dan
>>>
>>> On Mon, Mar 21, 2011 at 12:47 AM, Eugene Teo <eugene@...hat.com>
>>> wrote:
>>>> On 03/21/2011 03:40 AM, Dan Rosenberg wrote:
>>>>>
>>>>> I sent in a patch [1] resolving two issues in ROSE:
>>>>>
>>>>> "When parsing the FAC_NATIONAL_DIGIS facilities field, it's
>>>>> possible
>>>>> for a remote host to provide more digipeaters than expected,
>>>>> resulting
>>>>> in heap corruption. Check against ROSE_MAX_DIGIS to prevent
>>>>> overflows, and abort facilities parsing on failure.
>>>>>
>>>>> Additionally, when parsing the FAC_CCITT_DEST_NSAP and
>>>>> FAC_CCITT_SRC_NSAP facilities fields, a remote host can provide a
>>>>> length of less than 10, resulting in an underflow in a memcpy size,
>>>>> causing a kernel panic due to massive heap corruption. A length of
>>>>> greater than 20 results in a stack overflow of the callsign array.
>>>>> Abort facilities parsing on these invalid length values."
>>>>>
>>>>> These issues may both result in code execution. They may be
>>>>> triggered
>>>>> by a remote attacker if the victim has a listening ROSE socket, or
>>>>> by
>>>>> a local attacker (for privilege escalation) if a ROSE device exists
>>>>> (e.g. rose0).
>>>>>
>>>>> Ben Hutchings followed up with a patch [2] that resolves a number
>>>>> of
>>>>> other ROSE issues related to lack of size field validation, some of
>>>>> which may also result in heap corruption.
>>>>>
>>>>> Not sure about the proper CVE breakdown for all these issues, since
>>>>> the entire protocol was quite broken. Perhaps one is enough to
>>>>> cover
>>>>> everything.
>>>>
>>>> I am not sure. I would just assign one for the collection of issues
>>>> here but
>>>> I will let Steve decide instead.
>>>>
>>>>> [1] http://marc.info/?l=linux-netdev&m=130060344616926
>>>>> [2] http://marc.info/?l=linux-netdev&m=130063972406389&w=2
>>>>
>>>> Thanks, Eugene
>>>> --
>>>> main(i) { putchar(182623909 >> (i-1) * 5&31|!!(i<7)<<6) &&
>>>> main(++i); }
>>>>
>>
>
>

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.