linux-kernel - Re: Improve spinlock performance by moving work to one core

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>] [day] [month] [year] [list]

Message-ID: <CAOGi=dM2Xypi3w-n=LJ8pxiBuNe2o6otFXODo-9Cknm8xsaAsg@mail.gmail.com>
Date:	Wed, 4 Nov 2015 20:44:46 +0800
From:	Ling Ma <ling.ma.program@...il.com>
To:	linux-kernel@...r.kernel.org, Ling <ling.ml@...baba-inc.com>
Subject: Re: Improve spinlock performance by moving work to one core

Hi All,

(send again for linux-kernel@...r.kernel.org)

Spinlock caused cache line ping-pong between cores,
we have to spend lots of time to get serialized execution.
However if we present the serialized work to one core,
it will help us save much time.

In the attachment we changed code based on queued spinlock
The data tell us the spinlock performance can be improved by over 2X on 68cores
over 3X on 2 cores (Intel 2699v3 2 sockets, COD off, HT on)

In the following time we will try to send out the formal patch

Appreciate your comments.

Thanks

Download attachment "adv_spinlock.tar.bz2" of type "application/x-bzip2" (5234 bytes)