From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jon Mason Subject: Re: [RFC v2 1/2] PCI-Express Non-Transparent Bridge Support Date: Wed, 1 Aug 2012 18:49:11 -0700 Message-ID: <20120802014911.GC17548@jonmason-lab> References: <1343607994-32415-1-git-send-email-jon.mason@intel.com> <1343607994-32415-2-git-send-email-jon.mason@intel.com> <20120731222555.GB19134@kroah.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-kernel@vger.kernel.org, netdev@vger.kernel.org, linux-pci@vger.kernel.org, Dave Jiang To: Greg KH Return-path: Received: from mga01.intel.com ([192.55.52.88]:6153 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752655Ab2HBBtM (ORCPT ); Wed, 1 Aug 2012 21:49:12 -0400 Content-Disposition: inline In-Reply-To: <20120731222555.GB19134@kroah.com> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, Jul 31, 2012 at 03:25:55PM -0700, Greg KH wrote: > On Sun, Jul 29, 2012 at 05:26:33PM -0700, Jon Mason wrote: > > +struct ntb_transport_qp; > > + > > +struct ntb_client { > > + char *name; > > + int (*probe) (struct pci_dev *pdev); > > + void (*remove) (struct pci_dev *pdev); > > +}; > > Why isn't this tied into the driver model? That looks like you really > want to use a 'struct device' here, right? > > Then you have drivers that bind to them, and your devices show up in > sysfs, which is probably a requirement you have, right? I was trying to do something more light-weight than that and avoid the complexity, but this is probably the direction it needs to go. Thanks, Jon > > greg k-h