Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20120217063318.GA3330@p183.telecom.by>
Date: Fri, 17 Feb 2012 09:33:19 +0300
From: Alexey Dobriyan <adobriyan@...il.com>
To: David Windsor <dwindsor@...il.com>
Cc: Kees Cook <keescook@...omium.org>, Greg KH <gregkh@...uxfoundation.org>,
	pageexec@...email.hu,
	Ubuntu security discussion <ubuntu-hardened@...ts.ubuntu.com>,
	spender@...ecurity.net, linux-kernel@...r.kernel.org,
	kernel-hardening@...ts.openwall.com
Subject: Re: [ubuntu-hardened] Add overflow protection to kref

On Thu, Feb 16, 2012 at 09:48:38PM -0500, David Windsor wrote:
> <snip>
> 
> >>
> >> I have yet to see a patch, so why are we arguing about this?  :)
> >>
> >> Again, I don't know of any kref overflows that have ever happened, so
> >> trying to "protect" this type of thing, seems odd to me.
> >
> > Well, I think the issue was to protect counting things (which seems to
> > be what PaX was after originally), and that kref seemed like the place
> > to put it. I'll let David take it further.
> >
> 
> Patches are forthcoming that will first introduce overflow protection
> to kref.

Patches have already been posted:
http://marc.info/?l=linux-kernel&m=132337541830590&w=4
They were dropped for various (uninteresting) reasons, though.

> Once that's in place, I'll move a few refcount users from
> atomic_t to kref as a reference for other subsystems;

This sucks because dtor argument is mandatory.

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.