From mboxrd@z Thu Jan  1 00:00:00 1970
From: Shreyansh Jain <shreyansh.jain@nxp.com>
Subject: Re: [PATCH v5 05/12] eal: add probe and remove support
	for rte_driver
Date: Fri, 6 Jan 2017 17:14:34 +0530
Message-ID: <7af56aaf-b4dc-4933-588d-d0c375d23844@nxp.com>
References: <1482756644-13726-1-git-send-email-shreyansh.jain@nxp.com>
 <1482758645-23057-1-git-send-email-shreyansh.jain@nxp.com>
 <1482758645-23057-6-git-send-email-shreyansh.jain@nxp.com>
 <5088350.GnaeIEgOPh@xps13>
Mime-Version: 1.0
Content-Type: text/plain; charset="windows-1252"; format=flowed
Content-Transfer-Encoding: 7bit
Cc: <david.marchand@6wind.com>, <dev@dpdk.org>
To: Thomas Monjalon <thomas.monjalon@6wind.com>
Return-path: <dev-bounces@dpdk.org>
Received: from NAM01-BY2-obe.outbound.protection.outlook.com
 (mail-by2nam01on0065.outbound.protection.outlook.com [104.47.34.65])
 by dpdk.org (Postfix) with ESMTP id A9DA82A58
 for <dev@dpdk.org>; Fri,  6 Jan 2017 12:41:07 +0100 (CET)
In-Reply-To: <5088350.GnaeIEgOPh@xps13>
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>

On Wednesday 04 January 2017 03:35 AM, Thomas Monjalon wrote:
> 2016-12-26 18:53, Shreyansh Jain:
>> --- a/lib/librte_eal/common/include/rte_dev.h
>> +++ b/lib/librte_eal/common/include/rte_dev.h
>> @@ -152,6 +162,8 @@ struct rte_driver {
>>  	struct rte_bus *bus;           /**< Bus serviced by this driver */
>>  	const char *name;                   /**< Driver name. */
>>  	const char *alias;              /**< Driver alias. */
>> +	driver_probe_t *probe;         /**< Probe the device */
>> +	driver_remove_t *remove;       /**< Remove/hotplugging the device */
>>  };
>
> If I understand well, this probe function does neither scan nor match.
> So it could be named init.

Current model is:

After scanning for devices and populating bus->device_list,
Bus probe does:
  `-> bus->match()
  `-> rte_driver->probe() for matched driver

For PCI drivers, '.probe = rte_eal_pci_probe'.

For example, igb_ethdev.c:

--->8---
static struct eth_driver rte_igb_pmd = {
         .pci_drv = {
                 .driver = {
                         .probe = rte_eal_pci_probe,
                         .remove = rte_eal_pci_remove,
                 },
...
--->8---

>
> I think the probe (init) and remove ops must be specific to the bus.
> We can have them in rte_bus, and as an example, the pci implementation
> would call the pci probe and remove ops of rte_pci_driver.

So,
---
After scanning for devices (bus->scan()):
Bus probe (rte_eal_bus_probe()):
  `-> bus->match()
  `-> bus->init() - a new fn rte_bus_pci_init()
      -> which calls rte_eal_pci_probe()
      -> and rte_pci_driver->probe()

and remove rte_driver probe and remove callbacks because they are now
redundant. (they were added in bus patches itself)
---

Is the above correct understanding of your statement?

Somehow I don't remember why I didn't do this in first place - it seems
to be better option than introducing a rte_driver->probe()/remove()
layer. I will change it (and think again why I rejected this idea in
first place). Thanks.

>
> Please use rte_ prefix in public headers.
>

I am assuming you are referring to driver_probe_t/driver_remove_t.