From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754089AbaENJGk (ORCPT ); Wed, 14 May 2014 05:06:40 -0400 Received: from casper.infradead.org ([85.118.1.10]:48281 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751407AbaENJGf (ORCPT ); Wed, 14 May 2014 05:06:35 -0400 Date: Wed, 14 May 2014 11:06:29 +0200 From: Peter Zijlstra To: Frederic Weisbecker Cc: LKML , Andrew Morton , Ingo Molnar , Kevin Hilman , "Paul E. McKenney" , Thomas Gleixner , Viresh Kumar Subject: Re: [PATCH 1/3] irq_work: Implement remote queueing Message-ID: <20140514090629.GC30445@twins.programming.kicks-ass.net> References: <1400019956-25511-1-git-send-email-fweisbec@gmail.com> <1400019956-25511-2-git-send-email-fweisbec@gmail.com> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="Kz6eUT30PReluLpA" Content-Disposition: inline In-Reply-To: <1400019956-25511-2-git-send-email-fweisbec@gmail.com> User-Agent: Mutt/1.5.21 (2012-12-30) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org --Kz6eUT30PReluLpA Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Wed, May 14, 2014 at 12:25:54AM +0200, Frederic Weisbecker wrote: > irq work currently only supports local callbacks. However its code > is mostly ready to run remote callbacks and we have some potential user. >=20 > The full nohz subsystem currently open codes its own remote irq work > on top of the scheduler ipi when it wants a CPU to reevaluate its next > tick. However this ad hoc solution bloats the scheduler IPI. >=20 > Lets just extend the irq work subsystem to support remote queuing on top > of the generic SMP IPI to handle this kind of user. This shouldn't add > noticeable overhead. >=20 > Suggested-by: Peter Zijlstra > Cc: Andrew Morton > Cc: Ingo Molnar > Cc: Kevin Hilman > Cc: Paul E. McKenney > Cc: Peter Zijlstra > Cc: Thomas Gleixner > Cc: Viresh Kumar > Signed-off-by: Frederic Weisbecker > --- > include/linux/irq_work.h | 2 ++ > kernel/irq_work.c | 19 ++++++++++++++++++- > kernel/smp.c | 4 ++++ > 3 files changed, 24 insertions(+), 1 deletion(-) >=20 > diff --git a/include/linux/irq_work.h b/include/linux/irq_work.h > index 19ae05d..ae44aa2 100644 > --- a/include/linux/irq_work.h > +++ b/include/linux/irq_work.h > @@ -33,6 +33,8 @@ void init_irq_work(struct irq_work *work, void (*func)(= struct irq_work *)) > #define DEFINE_IRQ_WORK(name, _f) struct irq_work name =3D { .func =3D (= _f), } > =20 > bool irq_work_queue(struct irq_work *work); > +bool irq_work_queue_on(struct irq_work *work, int cpu); > + > void irq_work_run(void); > void irq_work_sync(struct irq_work *work); > =20 > diff --git a/kernel/irq_work.c b/kernel/irq_work.c > index a82170e..9f9be55 100644 > --- a/kernel/irq_work.c > +++ b/kernel/irq_work.c > @@ -56,11 +56,28 @@ void __weak arch_irq_work_raise(void) > } > =20 > /* > - * Enqueue the irq_work @entry unless it's already pending > + * Enqueue the irq_work @work on @cpu unless it's already pending > * somewhere. > * > * Can be re-enqueued while the callback is still in progress. > */ > +bool irq_work_queue_on(struct irq_work *work, int cpu) > +{ > + /* Only queue if not already pending */ > + if (!irq_work_claim(work)) > + return false; > + > + /* All work should have been flushed before going offline */ > + WARN_ON_ONCE(cpu_is_offline(cpu)); WARN_ON_ONCE(in_nmi()); > + > + llist_add(&work->llnode, &per_cpu(irq_work_list, cpu)); > + native_send_call_func_single_ipi(cpu); At the very leastestest make that: if (llist_add(&work->llnode, &per_cpu(irq_work_list, cpu))) native_send_call_func_single_ipi(cpu); But ideally, also test the IRQ_WORK_LAZY support, its weird to have that only be supported for the other queue. Hmm, why do we need that LAZY crap, that completely wrecks a perfectly simple thing. The changelog (bc6679aef673f), not the printk() usage make much sense, printk() can't cause an IPI storm... printk() isn't fast enough to storm anything. > + > + return true; > +} > +EXPORT_SYMBOL_GPL(irq_work_queue_on); --Kz6eUT30PReluLpA Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iQIcBAEBAgAGBQJTczIQAAoJEHZH4aRLwOS66sQQAJ6SVnxJRXxQtr+LoEwXNN/6 zlpPJ4To45XxuRMNtSyIuZXwnEeItJtdprZ9KiHAgfCLHoX+FQ/LAVpGOMb/9tc6 5hi+6saPF3VlqWJR6dH8AVBa292+qPQlhcP7Duvn0DYz6lL5ECLUPFqCKVqI+U/R JBnVFp69Y2RoxAmgjJfCAu/zVmpgpUZ92ev85RcfXoqlpHpopEkJYrd1NR1KhkMF V5oLVBVkqBxiYYe6RtBAP+bKWXTcOTmodSqBKCyUS271NmUFZBsdlII4eLfJrnpM Zn8UW1JVp+h3eJyP8QmLd+Heq6+3BZhh3fRbqwOwB/Tr9kMM6BgbCu5kUm9wu74J 9ruQydXrEoVbAntCZSfMlT4wm/s6jHDkp24gvEn0nX97dgrAlC/MBELiwHCfqg+q OFSsUypfDuIULDzk3VSjGeBD5TgSiv/L71MKSb4bGOeIV2L6li4qIZ7H34/9dTBf zAaRMwbtsFqdcl0i+YcYMr0YxKN4YkwPDVQBla3ODaMswHxPaxcQzoIQm8aJNMH9 c94iJeuom7xZ1pBl8cA1bpUF+tDDgUOwbCoXOIsRmmnO9jpFr/GPmWhNw3UzJNu/ KnnNlVBYrk9mVqtfUGt7uMNeCdzPaA8kKy3TL1bTDGDf4ZQE+hpAxuWhtQsjgtsw szZlDOOc0tUYT81ycnjf =x1Vz -----END PGP SIGNATURE----- --Kz6eUT30PReluLpA--