From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bartlomiej Zolnierkiewicz Subject: Re: ide: Bug fixes and improvements in the disk head parking code Date: Mon, 13 Oct 2008 21:04:38 +0200 Message-ID: <200810132104.39673.bzolnier@gmail.com> References: <87fxn3hw33.fsf@denkblock.local> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from gv-out-0910.google.com ([216.239.58.186]:40870 "EHLO gv-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759875AbYJMUM3 (ORCPT ); Mon, 13 Oct 2008 16:12:29 -0400 Received: by gv-out-0910.google.com with SMTP id e6so408246gvc.37 for ; Mon, 13 Oct 2008 13:12:28 -0700 (PDT) In-Reply-To: <87fxn3hw33.fsf@denkblock.local> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Elias Oltmanns Cc: linux-ide@vger.kernel.org On Saturday 11 October 2008, Elias Oltmanns wrote: > Fix a memory leak due to a missing blk_put_request() in > issue_park_cmd(). Additionally, we should plug the queue when enqueueing > the unpark request because there is no guarantee that the park timeout > has not expired by then. Even though the chance for that to happen is > very slim, the request might end up hanging in the queue until the next > I/O operation is queued up. While at it, clean up the code a little: > - make issue_park_cmd() a function of type void since nobody cares for > the return value anyway; > - use blk_start_queueing() instead of __blk_run_queue() since we don't > have to worry about recursion; > - remove a superfluous pointer deference in task_no_data_intr(). > > Signed-off-by: Elias Oltmanns > --- > Feel free to merge with original patch if Linus hasn't pulled your tree > yet. done, thanks