linux-kernel - Re: PATCH v6 0/6] livepatch: Atomic replace feature

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 for Android: free password hash cracker in your pocket

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

Date:   Thu, 1 Feb 2018 09:18:07 -0600
From:   Josh Poimboeuf <jpoimboe@...hat.com>
To:     Miroslav Benes <mbenes@...e.cz>
Cc:     Joe Lawrence <joe.lawrence@...hat.com>,
        Petr Mladek <pmladek@...e.com>, jikos@...nel.org,
        Jason Baron <jbaron@...mai.com>, jeyu@...nel.org,
        Evgenii Shatokhin <eshatokhin@...tuozzo.com>,
        linux-kernel@...r.kernel.org, live-patching@...r.kernel.org
Subject: Re: PATCH v6 0/6] livepatch: Atomic replace feature

On Thu, Feb 01, 2018 at 04:08:14PM +0100, Miroslav Benes wrote:
> On Thu, 1 Feb 2018, Joe Lawrence wrote:
> 
> > On 02/01/2018 08:49 AM, Miroslav Benes wrote:
> > > 
> > > Well, one more thing. I think there is a problem with shadow variables. 
> > > Similar to callbacks situation. Shadow variables cannot be destroyed the 
> > > way it is shown in our samples. Cumulative patches want to preserve 
> > > everything as much as possible. If I'm right, it should be mentioned in 
> > > the documentation.
> > 
> > Are you talking about using klp_shadow_free_all() call in a module_exit
> > routine?  Yeah, I think in this case, that responsibility would be
> > passed to the newly loaded cumulative patch, right?
> 
> Yes, but we haven't got an option not to call it here (as with callbacks, 
> where we can omit callbacks completely with atomic replace patches). A 
> live patch author must be aware of this and use shadow variables 
> appropriately.

So maybe we should recommend that shadow variables generally be freed
from a post-unpatch callback.

-- 
Josh