Bugzilla Bug 257023

CVE-2009-0282 (http://nvd.nist.gov/nvd.cfm?cvename=CVE-2009-0282):
  Integer overflow in Ralink Technology USB wireless adapter (RT73)
  3.08 for Windows, and other wireless card drivers including rt2400,
  rt2500, rt2570, and rt61, allows remote attackers to cause a denial
  of service (crash) and possibly execute arbitrary code via a Probe
  Request packet with a long SSID, possibly related to an integer
  signedness error.

This might allow for remote root compromise. Debian already distributes
patches, are they upstream yet?

*** Bug 257631 has been marked as a duplicate of this bug. ***

Yes, patches are in their advisories:
http://www.debian.org/security/2009/dsa-1712
http://www.debian.org/security/2009/dsa-1713
http://www.debian.org/security/2009/dsa-1714

Changing to [ebuild].

I do not have the hardware to test, so I won't be touching this. Anyone else,
please bump (darkside?).

ping, this is a rather serious issue...

(In reply to comment #4)
> I do not have the hardware to test, so I won't be touching this. Anyone else,
> please bump (darkside?).
> 

Nope, I just failed at searching for the bug and saw the vuln in my blog radar.
Sorry.

I do apologize for totally missing this bug until today - is there a way to
maybe switch a security bug around to have something of note in the summary? 
CVE would be nice (assuming one exists) - so that they stand out somehow. 
Anyway - I've been using the in kernel rt2500usb since 2.26(I think) - I've
always stated that the out of kernel would be going away when it was stable
enough for use, however at the moment, I don't have any cards with me (I'm at
work) - so I can't test a fix.  Most of my cards that are rt2x00 based require
WEP or open access points, which I don't currently have access to.  I do have
one rt2x00 card that CAN do WPA, I have not had a chance to look in to it
though.  If you can find one near you, it is the
http://en.wikipedia.org/wiki/Nintendo_Wi-Fi_USB_Connector .  You'd have to find
it at a resell shop most likely.  To finish off the comment, I've always said
that Security can fix any issue that creeps up without me getting upset.  Same
in this case for any package I maintain.

(In reply to comment #7)
> I do apologize for totally missing this bug until today - is there a way to
> maybe switch a security bug around to have something of note in the summary? 

It's in the "Gentoo Security" component. Bugzilla queries allow filtering for
that, and it's in some X-header in bugmails if you use procmail or similar to
filter.


> Anyway - I've been using the in kernel rt2500usb since 2.26(I think) - I've
> always stated that the out of kernel would be going away when it was stable
> enough for use, however at the moment, I don't have any cards with me (I'm at
> work) - so I can't test a fix.  Most of my cards that are rt2x00 based require
> WEP or open access points, which I don't currently have access to.  I do have
> one rt2x00 card that CAN do WPA, I have not had a chance to look in to it
> though.  If you can find one near you, it is the
> http://en.wikipedia.org/wiki/Nintendo_Wi-Fi_USB_Connector .  You'd have to find
> it at a resell shop most likely.  To finish off the comment, I've always said
> that Security can fix any issue that creeps up without me getting upset.  Same
> in this case for any package I maintain.

Well, no worries if it's going to take a few more days, but if you find the
in-kernel support to be as complete as the extra modules, they sure can be
kicked out. 

GLSA 200907-08, this bug can be closed once they are gone.

all are gone

Looks like we are finally done here. Closing.