Re: Texas Bank Dumps Antivirus for Whitelisting

["Richard M. Smith" <rms () computerbytesman com>]

Since HA machines don't really exist in nature, it's kind of hard to say
what they can and cannot do. ;-)

The answer to your JIT question would depend on how our mythical HA CPU
allows the memory dedicated to instruction store to be loaded up.

Richard

-----Original Message-----
From: Larry Seltzer [mailto:larry () larryseltzer com] 
Sent: Thursday, July 17, 2008 7:47 PM
To: Richard M. Smith; Drsolly
Cc: funsec () linuxbox org; rMslade () shaw ca
Subject: RE: [funsec] Texas Bank Dumps Antivirus for Whitelisting

And since you bring up Java, I guess JITs are not possible on HA.

Larry Seltzer
eWEEK.com Security Center Editor
http://security.eweek.com/
http://blogs.pcmag.com/securitywatch/
Contributing Editor, PC Magazine
larry.seltzer () ziffdavisenterprise com


-----Original Message-----
From: Richard M. Smith [mailto:rms () computerbytesman com] 
Sent: Thursday, July 17, 2008 6:40 PM
To: 'Drsolly'; Larry Seltzer
Cc: funsec () linuxbox org; rMslade () shaw ca
Subject: RE: [funsec] Texas Bank Dumps Antivirus for Whitelisting

Code bytes only get security measures supported by the hardware.  Data
bytes
can be subjected to additional security checks.  For example, a JVM can
implement a security model of its choosing for P-code.  (From the
viewpoint
of the real CPU, P-Code is not instructions but just data bytes that
gets
processed like any other data.)

Richard

-----Original Message-----
From: funsec-bounces () linuxbox org [mailto:funsec-bounces () linuxbox org]
On
Behalf Of Drsolly
Sent: Thursday, July 17, 2008 6:21 PM
To: Larry Seltzer
Cc: funsec () linuxbox org; rMslade () shaw ca
Subject: Re: [funsec] Texas Bank Dumps Antivirus for Whitelisting

On Thu, 17 Jul 2008, Larry Seltzer wrote:

> > > Harvard architecture, unlike von Neumann architecture, had a strict
> separation of 
> progrma and data store and representation.  It would have been
> impossible for a 
> program to modify its own or other executable material.  Data was not
> executable, 
> so SQL injection and XSS would have been impossible.  (So would a lot
of
> other 
> things, but ...)
>
> I'm not a real computer scientist, I just play one online, but this
> isn't how I thought it worked. SQL isn't actually executable code,
it's
> just data that program code uses in order to decide what to execute. A
> program in a Harvard architecture is capable of going "if x==1 then
> do_this() else if x==2 then do_that(); etc(),etc(),etc()" - can't it?
>
> Things like buffer overflows would be impossible with a Harvard
> architecture, but I don't see why SQL injection or Trojan horse
programs
> or many other malicious items would be any less likely.
 
What's the difference between bytes that are executable, and bytes that 
are used by the computer to decide what to do?

_______________________________________________
Fun and Misc security discussion for OT posts.
https://linuxbox.org/cgi-bin/mailman/listinfo/funsec
Note: funsec is a public and open mailing list.



_______________________________________________
Fun and Misc security discussion for OT posts.
https://linuxbox.org/cgi-bin/mailman/listinfo/funsec
Note: funsec is a public and open mailing list.