From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Eads, Gage" <gage.eads@intel.com>
Subject: Re: [PATCH 3/4] app/eventdev: add perf pipeline test
Date: Thu, 30 Nov 2017 17:15:14 +0000
Message-ID: <9184057F7FC11744A2107296B6B8EB1E2BB0182B@FMSMSX108.amr.corp.intel.com>
References: <20171130072406.15605-1-pbhagavatula@caviumnetworks.com>
 <20171130072406.15605-3-pbhagavatula@caviumnetworks.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Cc: "dev@dpdk.org" <dev@dpdk.org>
To: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>,
 "jerin.jacobkollanukkaran@cavium.com" <jerin.jacobkollanukkaran@cavium.com>,
 "Van Haaren, Harry" <harry.van.haaren@intel.com>, "Richardson, Bruce"
 <bruce.richardson@intel.com>, "hemant.agrawal@nxp.com"
 <hemant.agrawal@nxp.com>, "nipun.gupta@nxp.com" <nipun.gupta@nxp.com>, "Rao,
 Nikhil" <nikhil.rao@intel.com>
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 DB3547CB6
 for <dev@dpdk.org>; Thu, 30 Nov 2017 18:15:16 +0100 (CET)
In-Reply-To: <20171130072406.15605-3-pbhagavatula@caviumnetworks.com>
Content-Language: en-US
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>

Hi Pavan,

> -----Original Message-----
> From: Pavan Nikhilesh [mailto:pbhagavatula@caviumnetworks.com]
> Sent: Thursday, November 30, 2017 1:24 AM
> To: jerin.jacobkollanukkaran@cavium.com; Eads, Gage
> <gage.eads@intel.com>; Van Haaren, Harry <harry.van.haaren@intel.com>;
> Richardson, Bruce <bruce.richardson@intel.com>; hemant.agrawal@nxp.com;
> nipun.gupta@nxp.com; Rao, Nikhil <nikhil.rao@intel.com>
> Cc: dev@dpdk.org; Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
> Subject: [dpdk-dev] [PATCH 3/4] app/eventdev: add perf pipeline test
>=20
> This is a performance test case that aims at testing the following:
> 1. Measure the end-to-end performance of an event dev with a ethernet dev=
.
> 2. Maintain packet ordering from Rx to Tx.
>=20
> The perf pipeline test configures the eventdev with Q queues and P ports,=
 where
> Q is nb_ethdev * nb_stages and P is nb_workers.
>=20
> The user can choose the number of workers and number of stages through th=
e --
> wlcores and the --stlist application command line arguments respectively.
> The probed ethernet devices act as producer(s) for this application.
>=20
> The ethdevs are configured as event Rx adapters that enables them to inje=
cts
> events to eventdev based the first stage schedule type list requested by =
the user
> through --stlist the command line argument.
>=20
> Based on the number of stages to process(selected through --stlist), the
> application forwards the event to next upstream queue and when it reaches=
 last
> stage in the pipeline if the event type is ATOMIC it is enqueued onto eth=
dev Tx
> queue else to maintain ordering the event type is set to ATOMIC and enque=
ued
> onto the last stage queue.
> On packet Tx, application increments the number events processed and prin=
t
> periodically in one second to get the number of events processed in one s=
econd.
>=20
> Note: The --prod_type_ethdev is mandatory for running the application.
>=20
> Example command to run perf pipeline test:
> sudo build/app/dpdk-test-eventdev -c 0xf -s 0x8 --vdev=3Devent_sw0 -- \ -=
-
> test=3Dperf_pipeline --wlcore=3D1 --prod_type_ethdev --stlist=3Dao
>=20
> Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>

</snip>

> +static int
> +perf_pipeline_eventdev_setup(struct evt_test *test, struct evt_options
> +*opt) {
> +	int ret;
> +	int nb_ports;
> +	int nb_queues;
> +	int nb_stages =3D opt->nb_stages;
> +	uint8_t queue;
> +	uint8_t port;
> +	uint8_t atq =3D evt_has_all_types_queue(opt->dev_id);
> +	struct test_perf *t =3D evt_test_priv(test);
> +
> +	nb_ports =3D evt_nr_active_lcores(opt->wlcores);
> +	nb_queues =3D rte_eth_dev_count() * (nb_stages);
> +	nb_queues +=3D atq ? 0 : rte_eth_dev_count();
> +
> +	const struct rte_event_dev_config config =3D {
> +			.nb_event_queues =3D nb_queues,
> +			.nb_event_ports =3D nb_ports,
> +			.nb_events_limit  =3D 4096,
> +			.nb_event_queue_flows =3D opt->nb_flows,
> +			.nb_event_port_dequeue_depth =3D 128,
> +			.nb_event_port_enqueue_depth =3D 128,
> +	};
> +
> +	ret =3D rte_event_dev_configure(opt->dev_id, &config);
> +	if (ret) {
> +		evt_err("failed to configure eventdev %d", opt->dev_id);
> +		return ret;
> +	}
> +
> +	struct rte_event_queue_conf q_conf =3D {
> +			.priority =3D RTE_EVENT_DEV_PRIORITY_NORMAL,
> +			.nb_atomic_flows =3D opt->nb_flows,
> +			.nb_atomic_order_sequences =3D opt->nb_flows,
> +	};
> +	/* queue configurations */
> +	for (queue =3D 0; queue < nb_queues; queue++) {
> +		if (atq) {
> +			q_conf.event_queue_cfg =3D
> RTE_EVENT_QUEUE_CFG_ALL_TYPES;
> +		} else {
> +			uint8_t slot;
> +
> +			slot =3D queue % (nb_stages + 1);
> +			q_conf.schedule_type =3D slot =3D=3D nb_stages ?
> +				RTE_SCHED_TYPE_ATOMIC :
> +				opt->sched_type_list[slot];
> +		}
> +
> +		ret =3D rte_event_queue_setup(opt->dev_id, queue, &q_conf);
> +		if (ret) {
> +			evt_err("failed to setup queue=3D%d", queue);
> +			return ret;
> +		}
> +	}
> +
> +	/* port configuration */
> +	const struct rte_event_port_conf p_conf =3D {
> +			.dequeue_depth =3D opt->wkr_deq_dep,
> +			.enqueue_depth =3D 64,
> +			.new_event_threshold =3D 4096,
> +	};
> +

For the hard-coded event device (new_event_threshold, port_dequeue_depth, e=
nqueue_depth) and port configuration (enqueue_depth, new_event_threshold) v=
alues, it would be better to use the values from rte_event_dev_info_get() a=
nd rte_event_port_default_conf_get() instead.

Thanks,
Gage