From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH 2/5] tcm: Unify INQUIRY subsystem plugin handling Date: Thu, 14 Oct 2010 12:06:09 +0200 Message-ID: <20101014100609.GA29611@lst.de> References: <1286959681-3445-1-git-send-email-nab@linux-iscsi.org> <20101013110653.GA26366@lst.de> <1287001189.7334.47.camel@haakon2.linux-iscsi.org> <20101014000421.GB15583@lst.de> <1287030937.9909.30.camel@haakon2.linux-iscsi.org> <20101014092601.GB28446@lst.de> <1287050087.9909.79.camel@haakon2.linux-iscsi.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <1287050087.9909.79.camel@haakon2.linux-iscsi.org> Sender: linux-kernel-owner@vger.kernel.org To: "Nicholas A. Bellinger" Cc: Christoph Hellwig , linux-scsi , linux-kernel , FUJITA Tomonori , Mike Christie , Hannes Reinecke , James Bottomley , Boaz Harrosh , Jens Axboe , "Martin K. Petersen" , Douglas Gilbert , Richard Sharpe List-Id: linux-scsi@vger.kernel.org On Thu, Oct 14, 2010 at 02:54:47AM -0700, Nicholas A. Bellinger wrote: > It would make the most sense to just go ahead and set a pre-configured > struct queue_limits into transport_add_device_to_core_hba() from > subsystem specific ->create_virtdevice() code, and just drop whatever > unnecessary struct se_subsystem_api ops are duplicate of what is being > registered as defaults with struct queue_limits. That sounds good. Note that by now it's not just queue_limits but also other paramters, so a rename of the structure might be a good idea.