From mboxrd@z Thu Jan  1 00:00:00 1970
From: Thomas Monjalon <thomas.monjalon-pdR9zngts4EAvxtiuMwx3w@public.gmane.org>
Subject: Re: Bit spinlocks in DPDK
Date: Fri, 20 Dec 2013 16:39:05 +0100
Message-ID: <201312201639.05277.thomas.monjalon@6wind.com>
References: <6895EAE0CA8DEE40B92D7CA88BB521F332BA572E6B@HQ1-EXCH02.corp.brocade.com>
 <4656219.tgqzelRNOJ@x220> <01d001cef375$7167e300$5437a900$@com>
Mime-Version: 1.0
Content-Type: Text/Plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Cc: dev-VfR2kkLFssw@public.gmane.org
To: "=?iso-8859-1?q?Fran=E7ois-Fr=E9d=E9ric?= Ozog" <ff-SpBG2i6TTxU@public.gmane.org>
Return-path: <dev-bounces-VfR2kkLFssw@public.gmane.org>
In-Reply-To: <01d001cef375$7167e300$5437a900$@com>
List-Id: patches and discussions about DPDK <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request-VfR2kkLFssw@public.gmane.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev-VfR2kkLFssw@public.gmane.org>
List-Help: <mailto:dev-request-VfR2kkLFssw@public.gmane.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request-VfR2kkLFssw@public.gmane.org?subject=subscribe>
Errors-To: dev-bounces-VfR2kkLFssw@public.gmane.org
Sender: "dev" <dev-bounces-VfR2kkLFssw@public.gmane.org>

Hello,

07/12/2013 18:54, Fran=E7ois-Fr=E9d=E9ric Ozog :
> 1) If the critical section deals with weakly ordered loads then explicit
> fencing MUST be used: if not, out of order execution will just kill your
> idea of critical section.
[...]
> So use rte_mb() or rte_wmb() or rte_rmb() where appropriate. I recommend
> the rte_unlock code and documentation explains the out of order execution
> issues and the conditions they have to be mitigated with rte*mb(). I
> wonder if having an explicit mfence in rte_sinlock_unlock wouldn't be just
> necessary to avoid "hairy" bugs. In addition, we would have
> rte_sinlock_unlock_no_mb used internally for performance reasons, and
> usable externally by advanced users.

Using lock prefix is lighter than using memory barrier and have the same=20
effects.
But you're right about the bug in spinlocks.
I am going to send a patch for this.

=2D-=20
Thomas