From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759885AbbKTOVk (ORCPT ); Fri, 20 Nov 2015 09:21:40 -0500 Received: from mail-ig0-f196.google.com ([209.85.213.196]:35881 "EHLO mail-ig0-f196.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1162311AbbKTOVj (ORCPT ); Fri, 20 Nov 2015 09:21:39 -0500 Date: Fri, 20 Nov 2015 22:21:11 +0800 From: Boqun Feng To: Peter Zijlstra Cc: Paul Turner , Ingo Molnar , Oleg Nesterov , LKML , Paul McKenney , Jonathan Corbet , mhocko@kernel.org, dhowells@redhat.com, Linus Torvalds , will.deacon@arm.com Subject: Re: [PATCH 2/4] sched: Document Program-Order guarantees Message-ID: <20151120142111.GB19693@fixme-laptop.cn.ibm.com> References: <20151102132901.157178466@infradead.org> <20151102134940.883198067@infradead.org> <20151120100230.GA17308@twins.programming.kicks-ass.net> <20151120140850.GA19693@fixme-laptop.cn.ibm.com> <20151120141812.GF17308@twins.programming.kicks-ass.net> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="MW5yreqqjyrRcusr" Content-Disposition: inline In-Reply-To: <20151120141812.GF17308@twins.programming.kicks-ass.net> User-Agent: Mutt/1.5.24 (2015-08-30) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org --MW5yreqqjyrRcusr Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Nov 20, 2015 at 03:18:12PM +0100, Peter Zijlstra wrote: > On Fri, Nov 20, 2015 at 10:08:50PM +0800, Boqun Feng wrote: > > Hi Peter, > >=20 > > On Fri, Nov 20, 2015 at 11:02:30AM +0100, Peter Zijlstra wrote: > > [snip] > > > + * BLOCKING -- aka. SLEEP + WAKEUP > > > + * > > > + * For blocking we (obviously) need to provide the same guarantee as= for > > > + * migration. However the means are completely different as there is= no lock > > > + * chain to provide order. Instead we do: > > > + * > > > + * 1) smp_store_release(X->on_cpu, 0) > > > + * 2) smp_cond_acquire(!X->on_cpu) > > > + * > > > + * Example: > > > + * > > > + * CPU0 (schedule) CPU1 (try_to_wake_up) CPU2 (schedule) > > > + * > > > + * LOCK rq(0)->lock LOCK X->pi_lock > > > + * dequeue X > > > + * sched-out X > > > + * smp_store_release(X->on_cpu, 0); > > > + * > > > + * smp_cond_acquire(!X->on_cpu); > > > + * X->state =3D WAKING > > > + * set_task_cpu(X,2) > > > + * > > > + * LOCK rq(2)->lock > > > + * enqueue X > > > + * X->state =3D RUNNING > > > + * UNLOCK rq(2)->lock > > > + * > > > + * LOCK rq(2)->lock // orde= rs against CPU1 > > > + * sched-out Z > > > + * sched-in X > > > + * UNLOCK rq(1)->lock > > > + * > > > + * UNLOCK X->pi_lock > > > + * UNLOCK rq(0)->lock > > > + * > > > + * > > > + * However; for wakeups there is a second guarantee we must provide,= namely we > > > + * must observe the state that lead to our wakeup. That is, not only= must our > > > + * task observe its own prior state, it must also observe the stores= prior to > > > + * its wakeup. > > > + * > > > + * This means that any means of doing remote wakeups must order the = CPU doing > > > + * the wakeup against the CPU the task is going to end up running on= =2E This, > > > + * however, is already required for the regular Program-Order guaran= tee above, > > > + * since the waking CPU is the one issueing the ACQUIRE (2). > > > + * > >=20 > > Hope I'm the only one who got confused about the "2" in "ACQUIRE (2)", > > what does that refer? "2) smp_cond_acquire(!X->on_cpu)"? >=20 > Yes, exactly that. Would an unadorned 2 be clearer? How about "the one issueing the ACQUIRE (smp_cond_acquire)"? --MW5yreqqjyrRcusr Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAABCAAGBQJWTyxTAAoJEEl56MO1B/q4VfMH+wfnZjh4EC4VSO/o1R5aCCez 7hfYXvuvBjG2RsOCWo/OV2DX7JiJToEYKs2SBvb7ea234aJj4XbbnDBigp/fv1u5 nIKRSQL4X2x3fOor/C4pHBN7YB8vGlX6Vie0V3BaUk1F3LRYUEORpt3gdxFuJ2MG Qx6bA/eRU7EXt2DfBEfGBDXqN9tH/cVdbedHaDPYPuIKb3gb4hlCeJ2M8YafD8p9 LXdRCZIL3jHrmJJOQ+T5OGTGEyQfAn8GyYjXI6Rm3bY/SZGVig/X8QKgKPgao1HL tvbvef6o7JGJ3BL598oXRuQZ2ecYrUOkN2x8ndepyO3Em54HSRHHgyEJW8Cao30= =BjKP -----END PGP SIGNATURE----- --MW5yreqqjyrRcusr--