linux-kernel - Re: [PATCH 2.6.20-rc4 1/4] futex priority based wakeup

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: <45A52D2A.2030300@redhat.com>
Date:	Wed, 10 Jan 2007 10:15:06 -0800
From:	Ulrich Drepper <drepper@...hat.com>
To:	Pierre Peiffer <pierre.peiffer@...l.net>
CC:	Jakub Jelinek <jakub@...hat.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Dinakar Guniguntala <dino@...ibm.com>,
	Jean-Pierre Dion <jean-pierre.dion@...l.net>,
	Ingo Molnar <mingo@...e.hu>, Darren Hart <dvhltc@...ibm.com>,
	Sebastien Dugue <sebastien.dugue@...l.net>
Subject: Re: [PATCH 2.6.20-rc4 1/4] futex priority based wakeup

Pierre Peiffer wrote:
> But there can be a performance impact when several processes use
> different futexes which have the same hash key.
> In fact, the plist contains all waiters _of_all_futexes_ having the same
> hash key, not only the waiters of a given futex. This can be more a
> problem,

s/can be/is/

There are systems with thousands of active futexes, maybe tens of
thousands.  Not only is hash collision likely, it's also a matter of
using and administering the plist.  We have to make futexes less
connected, not more.  Now I definitely want to see real world tests first.

-- 
➧ Ulrich Drepper ➧ Red Hat, Inc. ➧ 444 Castro St ➧ Mountain View, CA ❖


Download attachment "signature.asc" of type "application/pgp-signature" (252 bytes)