From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bartlomiej Zolnierkiewicz Subject: Re: [PATCH] Correctly prevent IDE timer expiry function to run if request was already handled Date: Fri, 6 Apr 2007 23:35:26 +0200 Message-ID: <200704062335.26560.bzolnier@gmail.com> References: <20070406173603.GA82722@freefall.freebsd.org> <50660.65.57.245.11.1175892938.squirrel@ssl.mu.org> <4616B70B.3070501@ru.mvista.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from ug-out-1314.google.com ([66.249.92.173]:50924 "EHLO ug-out-1314.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933082AbXDFV0W (ORCPT ); Fri, 6 Apr 2007 17:26:22 -0400 Received: by ug-out-1314.google.com with SMTP id 44so1234532uga for ; Fri, 06 Apr 2007 14:26:20 -0700 (PDT) In-Reply-To: <4616B70B.3070501@ru.mvista.com> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Sergei Shtylyov Cc: Suleiman Souhlal , linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org On Friday 06 April 2007, Sergei Shtylyov wrote: > Hello. > > Suleiman Souhlal wrote: > > The reason the timer could have been active at that point, before applying > > this patch, is that we try to dispatch a new request after handling one. > > The new request will then have its own expiry timer, along with a handler. > > Since before this patch ide_timer_expiry() only looked at whether or not > > a handler was present, it would incorrectly think the request had not been > > handled already, and incorrectly tried to restart the timer. > > Hm, I'm still not sure why this happens at all, probably need to try > reproducing it (unless you post a stack trace :-). not a stack trace but should explain the issue a bit: ... ide_timer_expiry -> just before taking ide_lock IRQ happens ... ide_intr -> completes the command and queues the next one ... ide_timer_expiry (resumed) -> takes ide_lock -> the bug happens ;) should be even easier to trigger it on SMP since ide_timer_expiry may be running at the same time as ide_intr Bart