[Code-Crunchers] Fwd: Re: [funsec] Description of the Intel CPU bugs

[Gadi Evron]

----- Forwarded message from Valdis.Kletnieks at vt.edu -----

To: Gadi Evron <ge at linuxbox.org>
Cc: Dude VanWinkle <dudevanwinkle at gmail.com>, funsec at linuxbox.org,
	Larry Seltzer <Larry at larryseltzer.com>
Subject: Re: [funsec] Description of the Intel CPU bugs
From: Valdis.Kletnieks at vt.edu
Date: Thu, 28 Jun 2007 12:36:37 -0400

On Thu, 28 Jun 2007 11:05:04 CDT, Gadi Evron said:

> These patches in most cases would be work-arounds for
> stability/replacement functionality. They can be bypassed. This is a
> real patch, band-aid like. Or am I wrong?

All the patch does is upload the microcode if your BIOS hasn't already.

So, for example, my laptop has a T7200 in it - one of the affected chipsets. At
BIOS release A04, the BIOS would upload the then-current release into the CPU,
which didn't address the problem.  

If I then install the Microsoft patch (or more likely, update the Linux
microcode_ctl to include Intel's April 27 microcode update), the system will
boot, the BIOS would upgrade the microcode with an older update, and then
microcode_ctl would load *newer* code.

Then I update to Dell's A06 BIOS released May 7, which included the April 27
microcode.  At boot, the BIOS would load that release, and then when the MS
patch or microcode_ctl run, it would say "Oh, that one's already loaded". (and
in fact, that's what my laptop is doing at the moment...)




----- End forwarded message -----

-- 
--
"beepbeep it, i leave work, stop reading sec lists and im still hearing
gadi"
- HD Moore to Gadi Evron on IM, on Gadi's interview on npr, March 2007.