From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: can_queue modification within select_queue_depths Date: Thu, 19 Aug 2004 12:19:02 +0100 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <20040819121902.A8273@infradead.org> References: <41233AB1.7030605@il.marvell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from imladris.demon.co.uk ([193.237.130.41]:18695 "EHLO phoenix.infradead.org") by vger.kernel.org with ESMTP id S265492AbUHSLTE (ORCPT ); Thu, 19 Aug 2004 07:19:04 -0400 Content-Disposition: inline In-Reply-To: <41233AB1.7030605@il.marvell.com>; from Saeed.Bishara@il.marvell.com on Wed, Aug 18, 2004 at 02:17:05PM +0300 List-Id: linux-scsi@vger.kernel.org To: Saeed Bishara Cc: linux-scsi@vger.kernel.org On Wed, Aug 18, 2004 at 02:17:05PM +0300, Saeed Bishara wrote: > Hi, > is it allowed to modify the host can_queue with select_queue_depths > function? Yes. > if yes, is it needed to lock/unlock io_request_lock ? > thanks Dunno about 2.4 locking, but given this is at init time and thus implicitly serialized I wouldn't care too much about locking. BTW, if you happen to work on a driver for the marvell SATA part please talk to Jeff Garzik who is doing a libata driver for it. Also note that we're not going to accept pure SCSI LLDDs for (S)ATA hardware.