From mboxrd@z Thu Jan 1 00:00:00 1970 From: Cornelia Huck Subject: Re: [PATCH] sysfs: add filter function to groups Date: Tue, 30 Oct 2007 09:55:08 +0100 Message-ID: <20071030095508.00a24b8f@gondolin.boeblingen.de.ibm.com> References: <1193671019.3383.13.camel@localhost.localdomain> <1193676852.2655.21.camel@lov.site> <1193677071.3383.56.camel@localhost.localdomain> <20071029181814.25410e00@gondolin.boeblingen.de.ibm.com> <1193678646.3383.76.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: Received: from mtagate7.uk.ibm.com ([195.212.29.140]:65223 "EHLO mtagate7.uk.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751826AbXJ3I4B (ORCPT ); Tue, 30 Oct 2007 04:56:01 -0400 In-Reply-To: <1193678646.3383.76.camel@localhost.localdomain> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: James Bottomley Cc: Kay Sievers , Greg KH , linux-scsi , linux-kernel On Mon, 29 Oct 2007 12:24:06 -0500, James Bottomley wrote: > > Can you determine which subset of the attributes you want just before > > actually creating the group? Then you could do something like: > > > > create_group(grp, kobj) > > { > > grp->update_creation_mask(kobj); > > actually_create_attrs(); > > } > > That's actually what we currently do (at least in hand coded form) in > the current transport classes. However, it leads to one separate group > for each attached class. With the filter approach, we only need one > constructed group for every transport class. I meant doing it in the core. You still have one group for all cases, but immediately before creating the attributes, the core checks back which ones it should create. (Of course, that doesn't solve your problems if you dynamically want to change availability of attributes later on. You would need a different mechanism for that.)