full-disclosure - Microsoft GhostBuster Opinions

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

From: dufresne at winternet.com (Ron DuFresne)
Subject: Microsoft GhostBuster Opinions

On Fri, 18 Mar 2005, Todd Towles wrote:

>
> Dave wrote:
>
> >     About Tripwire, I understand what it does.  It basically
> > runs a file integrity check on certain files and reports the
> > differences from the last (hopefully known good) scan.  Say
> > that Tripwire is running on a system that's been compromised
> > by a rootkit that's been designed to evade file integrity
> > checkers such as tripwire.  Since the rootkit has control of
> > the kernel it has control of all the low level functions,
> > like returning a file when asked for one.  So one way to
> > evade tripwire would be to return the real file when asked
> > for it in read-only mode and return the rootkit file when
> > asked for it in execution mode.  That way tripwire won't
> > think the file has changed, since it's being given the same
> > file as it checked before, but when the file is executed then
> > it's the malicious file.
>
> But could this not be bypassed by running Tripwire from a bootable CD?
> The modified keneral would be inactive and therefore you would see the
> two separate files are opposed to just one. This is the idea that this
> new Microsoft products uses, but as people have stated, this can be done
> now with a combination of open-source products.

Now days, with the price of drives and devices being hot swapable, this is
even easier to deal with, at least in suspected cases of system tampering;

where I work we tend to make dd or rsync copies of the running drive to a
backup dirve.  Most of the servers allow hot swapping out these secondary
drives.  If the main drive fails I can boot off the backup, if I suspect a
problem with the integrity of the main file system, I can place the drive
in another system, mount it and run a tripwire against that.

AS for that LKM that can keep two copies of a file on the system and send
one for reads and execute another, that;s pretty advanced, and would have
to evade the differences that the various tripwire signatures are going to
produce on a file or filesystem that has been moved and altered in this
fashion.  Or take advantage of those places, and there are many, that
don;t properly keep the tripwrie db on read only media and alter those
sigs contained therein.  At this point it;s perhaps mostly hypothetical as
I'm unaware of rootkits that can accomplish that.

The poster talking about alternate data streams on ntfs systems <long time
since that topic was hot>, that's a problem to get around still on those
platforms.

To the original reply and other poster, thanks for the links to alternate
URL's for the old paper on kernel hacking!

Thanks,

Ron DuFresne
-- 
"Sometimes you get the blues because your baby leaves you. Sometimes you get'em
'cause she comes back." --B.B. King
        ***testing, only testing, and damn good at it too!***

OK, so you're a Ph.D.  Just don't touch anything.