From mboxrd@z Thu Jan  1 00:00:00 1970
From: Thomas Monjalon <thomas@monjalon.net>
Subject: Re: [PATCH v2 2/4] ethdev: add siblings iterators
Date: Mon, 01 Apr 2019 03:59:45 +0200
Message-ID: <2318869.vaV4MP09YM@xps>
References: <20181130002716.27325-1-thomas@monjalon.net>
 <20190227100724.tjkxtuis5yfwvvrd@bidouze.vm.6wind.com>
 <5809024.HDbGbE6FtF@xps>
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Cc: dev@dpdk.org, Ferruh Yigit <ferruh.yigit@intel.com>,
 Andrew Rybchenko <arybchenko@solarflare.com>
To: =?ISO-8859-1?Q?Ga=EBtan?= Rivet <gaetan.rivet@6wind.com>
Return-path: <dev-bounces@dpdk.org>
Received: from out1-smtp.messagingengine.com (out1-smtp.messagingengine.com
 [66.111.4.25]) by dpdk.org (Postfix) with ESMTP id 266DF10A3
 for <dev@dpdk.org>; Mon,  1 Apr 2019 03:59:48 +0200 (CEST)
In-Reply-To: <5809024.HDbGbE6FtF@xps>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://mails.dpdk.org/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://mails.dpdk.org/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <https://mails.dpdk.org/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

27/02/2019 11:51, Thomas Monjalon:
> 27/02/2019 11:07, Ga=EBtan Rivet:
> > On Wed, Feb 20, 2019 at 11:10:49PM +0100, Thomas Monjalon wrote:
> > > +uint16_t __rte_experimental
> > > +rte_eth_find_next_of(uint16_t port_id, const struct rte_device *pare=
nt)
> > > +{
> > > +	while (port_id < RTE_MAX_ETHPORTS &&
> > > +			rte_eth_devices[port_id].state =3D=3D RTE_ETH_DEV_UNUSED &&
> > > +			rte_eth_devices[port_id].device !=3D parent)
> > > +		port_id++;
> >=20
> > Why not call rte_eth_find_next directly from this function, and
> > add your specific test on top of it?
> >=20
> > Something like:
> >=20
> >     	while (port_id < RTE_MAX_ETHPORTS &&
> >     	       rte_eth_devices[port_id].device !=3D parent)
> >     		port_id =3D rte_eth_find_next(port_id + 1);
> >=20
> > this way you won't have to rewrite the test on the device state. Having=
 the
> > logic expressed in several places would make reworking the device state=
s more
> > complicated than necessary if it were to happen (just as you did when s=
witching
> > the test from !(ATTACHED || REMOVED) to (UNUSED).
>=20
> About the intent, you are right.
> About the solution, it seems buggy. We can try to find another way
> of coding this loop by using rte_eth_find_next()
> and adding the parent condition.

Your proposal is correct if adding a first call before the loop:
	port_id =3D rte_eth_find_next(port_id);