From mboxrd@z Thu Jan  1 00:00:00 1970
From: Thomas Monjalon <thomas.monjalon@6wind.com>
Subject: Re: [PATCH v11 09/13] pci: add bus driver
Date: Wed, 15 Feb 2017 12:20:37 +0100
Message-ID: <3443176.du41bGMhrB@xps13>
References: <1484748329-5418-1-git-send-email-shreyansh.jain@nxp.com>
 <1484801117-779-10-git-send-email-thomas.monjalon@6wind.com>
 <CALe+Z02OOPGAGtSJFLMoiKAw1ehUr6H443mT3eogL=uBf65vGw@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7Bit
Cc: Shreyansh Jain <shreyansh.jain@nxp.com>, dev@dpdk.org
To: Jan Blunck <jblunck@infradead.org>
Return-path: <dev-bounces@dpdk.org>
Received: from mail-wm0-f54.google.com (mail-wm0-f54.google.com [74.125.82.54])
 by dpdk.org (Postfix) with ESMTP id DD87C9E7
 for <dev@dpdk.org>; Wed, 15 Feb 2017 12:20:39 +0100 (CET)
Received: by mail-wm0-f54.google.com with SMTP id v186so39285531wmd.0
 for <dev@dpdk.org>; Wed, 15 Feb 2017 03:20:39 -0800 (PST)
In-Reply-To: <CALe+Z02OOPGAGtSJFLMoiKAw1ehUr6H443mT3eogL=uBf65vGw@mail.gmail.com>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

2017-02-15 11:42, Jan Blunck:
> >  /**
> > + * Add a PCI device to the PCI Bus (append to PCI Device list). This function
> > + * also updates the bus references of the PCI Device (and the generic device
> > + * object embedded within.
> > + *
> > + * @param pci_dev
> > + *     PCI device to add
> > + * @return void
> > + */
> > +void rte_eal_pci_add_device(struct rte_pci_device *pci_dev);
> > +
> 
> Who would be the user of this?
> 
> From my understanding a device will show up on the bus if:
> 1. bus->scan() finds it
> 2. bus->attach(devargs) explicitly adds it to the whitelist
> 
> Both methods shouldn't require us to expose the API outside of eal.

I agree

> > +struct rte_pci_bus rte_pci_bus = {
> > +       .bus = {
> > +               .scan = rte_eal_pci_scan,
> > +               .probe = rte_eal_pci_probe,
> > +       },
> > +       .device_list = TAILQ_HEAD_INITIALIZER(rte_pci_bus.device_list),
> > +       .driver_list = TAILQ_HEAD_INITIALIZER(rte_pci_bus.driver_list),
> 
> I don't see why this is necessary and I believe it makes people think
> we might want to model topology. It's better to keep it simple and
> have it locally in eal_common_pci.c.

I agree