From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Loftus, Ciara" <ciara.loftus@intel.com>
Subject: Re: [PATCH v13 2/2] vhost: Add VHOST PMD
Date: Tue, 22 Mar 2016 10:33:31 +0000
Message-ID: <74F120C019F4A64C9B78E802F6AD4CC24F847885@IRSMSX106.ger.corp.intel.com>
References: <1458030701-11487-3-git-send-email-mukawa@igel.co.jp>
 <1458539108-15686-3-git-send-email-mukawa@igel.co.jp>
 <74F120C019F4A64C9B78E802F6AD4CC24F84543E@IRSMSX106.ger.corp.intel.com>
 <56F0A5F7.2090205@igel.co.jp> <56F0B2DC.3010602@igel.co.jp>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Cc: "Richardson, Bruce" <bruce.richardson@intel.com>,
 "ann.zhuangyanying@huawei.com" <ann.zhuangyanying@huawei.com>,
 "thomas.monjalon@6wind.com" <thomas.monjalon@6wind.com>
To: Tetsuya Mukawa <mukawa@igel.co.jp>, "dev@dpdk.org" <dev@dpdk.org>
Return-path: <dev-bounces@dpdk.org>
Received: from mga03.intel.com (mga03.intel.com [134.134.136.65])
 by dpdk.org (Postfix) with ESMTP id 128772BD7
 for <dev@dpdk.org>; Tue, 22 Mar 2016 11:33:48 +0100 (CET)
In-Reply-To: <56F0B2DC.3010602@igel.co.jp>
Content-Language: en-US
List-Id: patches and discussions about DPDK <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>

>=20
> On 2016/03/22 10:55, Tetsuya Mukawa wrote:
> > On 2016/03/22 0:40, Loftus, Ciara wrote:
> >>> +
> >>> +static void
> >>> +eth_dev_info(struct rte_eth_dev *dev,
> >>> +	     struct rte_eth_dev_info *dev_info)
> >>> +{
> >>> +	dev_info->driver_name =3D drivername;
> >>> +	dev_info->max_mac_addrs =3D 1;
> >>> +	dev_info->max_rx_pktlen =3D (uint32_t)-1;
> >>> +	dev_info->max_rx_queues =3D dev->data->nb_rx_queues;
> >>> +	dev_info->max_tx_queues =3D dev->data->nb_tx_queues;
> >> I'm not entirely familiar with eth driver code so please correct me if=
 I am
> wrong.
> >>
> >> I'm wondering if assigning the max queue values to dev->data-
> >nb_*x_queues is correct.
> >> A user could change the value of nb_*x_queues with a call to
> rte_eth_dev_configure(n_queues) which in turn calls
> rte_eth_dev_*x_queue_config(n_queues) which will set dev->data-
> >nb_*x_queues to the value of n_queues which can be arbitrary and
> decided by the user. If this is the case, dev->data->nb_*x_queues will no
> longer reflect the max, rather the value the user chose in the call to
> rte_eth_dev_configure. And the max could potentially change with multiple
> calls to configure. Is this intended behaviour?
> > Hi Ciara,
> >
> > Thanks for reviewing it. Here is a part of rte_eth_dev_configure().
> >
> > int
> > rte_eth_dev_configure(uint8_t port_id, uint16_t nb_rx_q, uint16_t
> nb_tx_q,
> >                       const struct rte_eth_conf *dev_conf)
> > {
> >         <snip>
> >         /*
> >          * Check that the numbers of RX and TX queues are not greater
> >          * than the maximum number of RX and TX queues supported by the
> >          * configured device.
> >          */
> >         (*dev->dev_ops->dev_infos_get)(dev, &dev_info);
> >
> >         if (nb_rx_q =3D=3D 0 && nb_tx_q =3D=3D 0) {
> >                <snip>
> >                 return -EINVAL;
> >         }
> >
> >         if (nb_rx_q > dev_info.max_rx_queues) {
> >                <snip>
> >                 return -EINVAL;
> >         }
> >
> >         if (nb_tx_q > dev_info.max_tx_queues) {
> >                <snip>
> >                 return -EINVAL;
> >         }
> >
> >         <snip>
> >
> >         /*
> >          * Setup new number of RX/TX queues and reconfigure device.
> >          */
> >         diag =3D rte_eth_dev_rx_queue_config(dev, nb_rx_q);
> >         <snip>
> >         diag =3D rte_eth_dev_tx_queue_config(dev, nb_tx_q);
> >         <snip>
> > }
> >
> > Anyway, rte_eth_dev_tx/rx_queue_config() will be called only after
> > checking the current maximum number of queues.
> > So the user cannot set the number of queues greater than current
> maximum
> > number.
> >
> > Regards,
> > Tetsuya
>=20
> Hi Ciara,
>=20
> Now, I understand what you say.
> Probably you pointed out the case that the user specified a value
> smaller than current maximum value.
>=20
> For example, if we have 4 queues. Below code will be failed at last line.
> rte_eth_dev_configure(portid, 4, 4, ...);
> rte_eth_dev_configure(portid, 2, 2, ...);
> rte_eth_dev_configure(portid, 4, 4, ...);
>=20
> I will submit a patch to fix it. Could you please review and ack it?

Hi Tetsuya,

Correct, sorry for the initial confusion. Thanks for the patch so quickly.
I've reviewed the code - looks good. I just want to run some tests and will=
 give my Ack later today all going well.

Thanks,
Ciara

>=20
> Regards,
> Tetsuya