reassign 384922 src:linux linux-2.6/2.6.12-1
found 384922 linux/3.2.32-1
tags 384922 - fixed-upstream + wontfix upstream
quit
Hi again,
In February, Paul Szabo wrote:
>> Do I understand correctly that you are requesting an export or mountd
>> option filter_gid, which would behave like --manage-g
Dear Jonathan,
>> NFSv4+krb is better only because ...
> Surely the ability to squash multiple uids is also a help. ;-)
Not when asking to squash groups. :-)
I thought that idmapd worked also with AUTH_SYS.
> Do I understand correctly that you are requesting an export or mountd
> option filter_
paul.sz...@sydney.edu.au wrote:
> NFSv4+krb is better only because it does not have a concept of groups.
> Remove groups from AUTH_SYS, ignoring all groups or in other words doing
> "manage primary group" similar to secondaries with -manage_gids, and
> issue might be solved.
Surely the ability to
> ... AUTH_SYS with untrusted root on clients is not a good fit ...
> NFSv4 with kerberos authentication would be less broken. root_squash
> is a simplistic and incomplete band-aid.
NFSv4+krb is better only because it does not have a concept of groups.
Remove groups from AUTH_SYS, ignoring all gr
Hi,
Paul Szabo wrote:
> I will re-phrase the problem, this may be clearer for some people:
>
> The root_squash option is to protect from an "evil root". Though group
> staff is root-equivalent, root_squash does not currently squash that group
> (for various reasons, the kernel not supportin
Dear Moritz,
Please see comments in
http://bugzilla.kernel.org/show_bug.cgi?id=14295
:
> This looks more like a feature request than a bug report to me. The right
> address for that kind of discussion would be on the linux-...@vger.kernel.org
> mailing list, not bugzilla.
> Right, a good first s
Dear Moritz,
> Please file an enhancement bug at bugzilla.kernel.org ...
Done:
http://bugzilla.kernel.org/show_bug.cgi?id=14295
Cheers, Paul
Paul Szabo p...@maths.usyd.edu.au http://www.maths.usyd.edu.au/u/psz/
School of Mathematics and Statistics University of SydneyAustralia
--
On Sat, Sep 02, 2006 at 11:57:03PM +1000, Paul Szabo wrote:
> I will re-phrase the problem, this may be clearer for some people:
>
> The root_squash option is to protect from an "evil root". Though group
> staff is root-equivalent, root_squash does not currently squash that group
> (for vari
I will re-phrase the problem, this may be clearer for some people:
The root_squash option is to protect from an "evil root". Though group
staff is root-equivalent, root_squash does not currently squash that group
(for various reasons, the kernel not supporting such options being one).
An "
retitle 384922 NFS root_squash broken without support for squashing multiple
groups
severity 384922 critical
thanks
Dear Steve,
> [root_squash is] often circumventable ...
References (CERT kb, securityfocus BID, secunia advisory)? I do not know of
any (other than this bug) instances of defeati
On Thu, Aug 31, 2006 at 10:49:10PM +1000, Paul Szabo wrote:
> I am confused: what is the use and intent of root_squash, why is it enabled
> by default, and why is there an option to turn it off?
It's a naive and often circumventable technique for limiting the impact of
an untrusted/compromised cl
severity 384922 critical
thanks
Dear Steve,
The issue is root compromise of an NFS server. If that is possible then
it is critical; if it is not possible then the bug is solved. It seems
logically impossible to downgrade this kind of bugs.
Cheers,
Paul Szabo [EMAIL PROTECTED] http://www.mat
severity 384922 critical
thanks
Dear Steve,
> It happens to be very dangerous to share a filesystem via NFS between
> systems that have different security contexts. This does not make it a
> critical bug ...
Is it acceptable for a root compromise of one system to easily propagate
onto another?
severity 384922 important
quit
On Thu, Aug 31, 2006 at 04:34:00PM +1000, Paul Szabo wrote:
> Sorry, I missed one:
> > ... only exploitable when
> > - you have a non-empty "staff" group on the client (+/- equivalent to
> > untrusted root users on the client, since any root user can simply add
>
On Thu, Aug 31, 2006 at 04:34:00PM +1000, Paul Szabo wrote:
>I am somewhat curious: who is Steinar, and who are you?
Steve Langasek is the Debian Release Manager
Steinar H. Gunderson is a Debian Developer, comaintainer of nfs-utils
>I had submitted a bug against nfs-kernel-server; the maintainer
On Thu, Aug 31, 2006 at 01:05:41PM +1000, Paul Szabo wrote:
> > The last two points are true by default on Debian, but the first three
> > points are configuration decisions on the part of the NFS server
> > administrator. I understand that you have reasons to export shares allowing
> > suid binar
severity 384922 critical
thanks
Dear Steve,
Sorry, I missed one:
> ... only exploitable when
>
> - you have a non-empty "staff" group on the client (+/- equivalent to
> untrusted root users on the client, since any root user can simply add
> users to this group)
> - you have NFS-shared files
Dear Steve,
Thanks for your response.
> The bug log indicates that it's only exploitable when
>
> - you have a non-empty "staff" group on the client (+/- equivalent to
> untrusted root users on the client, since any root user can simply add
> users to this group)
> - you have NFS-shared files
On Thu, Aug 31, 2006 at 11:46:15AM +1000, Paul Szabo wrote:
> You seem to think that this is "important" but not "critical".
> Don't you agree that it is a root security hole?
Indeed I do not agree that it's a root security hole. The bug log indicates
that it's only exploitable when
- you have
Dear Steve,
You seem to think that this is "important" but not "critical".
Don't you agree that it is a root security hole?
Thanks,
Paul Szabo [EMAIL PROTECTED] http://www.maths.usyd.edu.au/u/psz/
School of Mathematics and Statistics University of SydneyAustralia
--
To UNSUBSCRIBE,
20 matches
Mail list logo
