From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932565AbbGGR1a (ORCPT ); Tue, 7 Jul 2015 13:27:30 -0400 Received: from foss.arm.com ([217.140.101.70]:37152 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932229AbbGGR1W (ORCPT ); Tue, 7 Jul 2015 13:27:22 -0400 Date: Tue, 7 Jul 2015 18:27:18 +0100 From: Will Deacon To: Waiman Long Cc: Peter Zijlstra , Ingo Molnar , Arnd Bergmann , Thomas Gleixner , "linux-arch@vger.kernel.org" , "linux-kernel@vger.kernel.org" , Scott J Norton , Douglas Hatch Subject: Re: [PATCH 2/4] locking/qrwlock: Reduce reader/writer to reader lock transfer latency Message-ID: <20150707172718.GJ23879@arm.com> References: <1436197386-58635-1-git-send-email-Waiman.Long@hp.com> <1436197386-58635-3-git-send-email-Waiman.Long@hp.com> <20150706182353.GC1607@arm.com> <559ADBCD.6020803@hp.com> <20150707091711.GA23879@arm.com> <20150707111731.GQ3644@twins.programming.kicks-ass.net> <20150707114918.GG23879@arm.com> <559BE27E.6060901@hp.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <559BE27E.6060901@hp.com> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Jul 07, 2015 at 03:30:22PM +0100, Waiman Long wrote: > On 07/07/2015 07:49 AM, Will Deacon wrote: > > On Tue, Jul 07, 2015 at 12:17:31PM +0100, Peter Zijlstra wrote: > >> On Tue, Jul 07, 2015 at 10:17:11AM +0100, Will Deacon wrote: > >>>>> Thinking about it, can we kill _QW_WAITING altogether and set (cmpxchg > >>>>> from 0) wmode to _QW_LOCKED in the write_lock slowpath, polling (acquire) > >>>>> rmode until it hits zero? > >>>> No, this is how we make the lock fair so that an incoming streams of > >>>> later readers won't block a writer from getting the lock. > >>> But won't those readers effectively see that the lock is held for write > >>> (because we set wmode to _QW_LOCKED before the existing reader had drained) > >>> and therefore fall down the slow-path and get held up on the spinlock? > >> Yes, that's the entire point. Once there's a writer pending, new readers > >> should queue too. > > Agreed. My point was that we can achieve the same result without > > a separate _QW_WAITING flag afaict. > > _QW_WAITING and _QW_LOCKED has different semantics and are necessary for > the proper handshake between readers and writer. We set _QW_WAITING when > readers own the lock and the writer is waiting for the readers to go > away. The _QW_WAITING flag will force new readers to go to queuing while > the writer is waiting. We set _QW_LOCKED when a writer own the lock and > it can only be set atomically when no reader is present. Without the > intermediate _QW_WAITING step, a continuous stream of incoming readers > (which make the reader count never 0) could deny a writer from getting > the lock indefinitely. It's probably best if I try to implement something and we can either pick holes in the patch or I'll realise why I'm wrong in the process :) Will