[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20151021075456.GY2881@worktop.programming.kicks-ass.net>
Date: Wed, 21 Oct 2015 09:54:56 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Ling Ma <ling.ma.program@...il.com>
Cc: Ingo Molnar <mingo@...nel.org>, mingo@...hat.com,
linux-kernel@...r.kernel.org, Ma Ling <ling.ml@...baba-inc.com>,
Waiman Long <Waiman.Long@....com>
Subject: Re: [RFC PATCH] qspinlock: Improve performance by reducing load
instruction rollback
On Wed, Oct 21, 2015 at 01:28:04PM +0800, Ling Ma wrote:
> Ok, we will put the spinlock test into the perf bench.
The attached is what I used back when we were doing the initial
qspinlock stuff.
I've not looked at it in quite some time, so it might be out of sync
with the kernel sources.
Download attachment "spinlocks.tar.bz2" of type "application/octet-stream" (30648 bytes)
Powered by blists - more mailing lists