From mboxrd@z Thu Jan 1 00:00:00 1970 From: Steffen Klassert Subject: Re: [PATCH net-next] xfrm: Namespacify xfrm_policy_sk_bundles Date: Tue, 17 Dec 2013 12:39:50 +0100 Message-ID: <20131217113950.GL31491@secunet.com> References: <1387189568-31769-1-git-send-email-fan.du@windriver.com> <20131216122325.GI31491@secunet.com> <52AFACCA.20107@windriver.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: davem@davemloft.net, netdev@vger.kernel.org To: Fan Du Return-path: Received: from a.mx.secunet.com ([195.81.216.161]:60734 "EHLO a.mx.secunet.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751072Ab3LQLjw (ORCPT ); Tue, 17 Dec 2013 06:39:52 -0500 Content-Disposition: inline In-Reply-To: <52AFACCA.20107@windriver.com> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, Dec 17, 2013 at 09:45:46AM +0800, Fan Du wrote: >=20 >=20 > On 2013=E5=B9=B412=E6=9C=8816=E6=97=A5 20:23, Steffen Klassert wrote: > >On Mon, Dec 16, 2013 at 06:26:08PM +0800, Fan Du wrote: > >> > >>And also since xfrm_policy_sk_bundles is only used in xfrm_lookup a= nd > >>__xfrm_garbage_collect, both in process context, no reason we shoul= d turn > >>BH off. > > > >Are you sure about that? > > > >__xfrm_garbage_collect() is called via dst_alloc() which can be call= ed > >from softirq and process context. >=20 > Thanks for pointing this out, you are correct! :) >=20 > IMO, xchg can still cover those two cases: > 1. xfrm_lookup(Process context) vs __xfrm_garbage_collect(softirq c= ontext) > 2. xfrm_lookup(Process context) vs __xfrm_garbage_collect(Process c= ontext when SPD change or dev down) >=20 > I will fix commit message properly on v2 if you are ok with above des= cription. Yes, please fix and resend.