linux-kernel - Re: [patch 01/10] emm: mm_lock: Lock a process against reclaim

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <20080407193544.GH20587@duo.random>
Date:	Mon, 7 Apr 2008 21:35:44 +0200
From:	Andrea Arcangeli <andrea@...ranet.com>
To:	Jeremy Fitzhardinge <jeremy@...p.org>
Cc:	Christoph Lameter <clameter@....com>, Robin Holt <holt@....com>,
	kvm-devel@...ts.sourceforge.net,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	general@...ts.openfabrics.org, steiner@....com,
	linux-kernel@...r.kernel.org, linux-mm@...ck.org
Subject: Re: [patch 01/10] emm: mm_lock: Lock a process against reclaim

On Mon, Apr 07, 2008 at 12:02:53PM -0700, Jeremy Fitzhardinge wrote:
> It's per-mm though.  How many processes would need to have notifiers?

There can be up to hundreds of VM in a single system. Not sure to
understand the point of the question though.

> Well, its definitely going to need more comments then.  I assumed it would 
> end up locking everything, so unlocking everything would be sufficient.

After your comments, I'm writing an alternate version that will
guarantee a O(N) worst case to both sigkill and cond_resched but
frankly this is low priority. Without mmu notifiers /dev/kvm can't be
given to a normal luser without at least losing mlock ulimits, so lack
of a mmu notifiers is a bigger issue than whatever complexity in
mm_lock as far as /dev/kvm ownership is concerned.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/