From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Van Haaren, Harry" Subject: Re: [PATCH v4 1/2] eventdev: add device stop flush callback Date: Tue, 27 Mar 2018 08:20:28 +0000 Message-ID: References: <1521087130-20244-1-git-send-email-gage.eads@intel.com> <1521555187-25710-1-git-send-email-gage.eads@intel.com> <9184057F7FC11744A2107296B6B8EB1E3FA88B42@FMSMSX108.amr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Cc: "Richardson, Bruce" , "santosh.shukla@caviumnetworks.com" , "nipun.gupta@nxp.com" , "dev@dpdk.org" To: "Eads, Gage" , "jerin.jacob@caviumnetworks.com" , "hemant.agrawal@nxp.com" Return-path: Received: from mga14.intel.com (mga14.intel.com [192.55.52.115]) by dpdk.org (Postfix) with ESMTP id 03C117CE1 for ; Tue, 27 Mar 2018 10:20:31 +0200 (CEST) In-Reply-To: <9184057F7FC11744A2107296B6B8EB1E3FA88B42@FMSMSX108.amr.corp.intel.com> Content-Language: en-US List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" > From: Eads, Gage > Sent: Monday, March 26, 2018 10:59 PM > To: Van Haaren, Harry ; dev@dpdk.org > Cc: jerin.jacob@caviumnetworks.com; hemant.agrawal@nxp.com; Richardson, B= ruce > ; santosh.shukla@caviumnetworks.com; > nipun.gupta@nxp.com > Subject: RE: [PATCH v4 1/2] eventdev: add device stop flush callback >=20 >=20 >=20 > > -----Original Message----- > > From: Van Haaren, Harry > > Sent: Friday, March 23, 2018 11:57 AM > > To: Eads, Gage ; dev@dpdk.org > > Cc: jerin.jacob@caviumnetworks.com; hemant.agrawal@nxp.com; Richardson, > > Bruce ; santosh.shukla@caviumnetworks.com; > > nipun.gupta@nxp.com > > Subject: RE: [PATCH v4 1/2] eventdev: add device stop flush callback > > > > > From: Eads, Gage > > > Sent: Tuesday, March 20, 2018 2:13 PM > > > To: dev@dpdk.org > > > Cc: jerin.jacob@caviumnetworks.com; Van Haaren, Harry > > > ; hemant.agrawal@nxp.com; Richardson, > > > Bruce ; santosh.shukla@caviumnetworks.com= ; > > > nipun.gupta@nxp.com > > > Subject: [PATCH v4 1/2] eventdev: add device stop flush callback > > > > > > When an event device is stopped, it drains all event queues. These > > > events may contain pointers, so to prevent memory leaks eventdev now > > > supports a user-provided flush callback that is called during the que= ue > drain > > process. > > > This callback is stored in process memory, so the callback must be > > > registered by any process that may call rte_event_dev_stop(). > > > > > > This commit also clarifies the behavior of rte_event_dev_stop(). > > > > > > This follows this mailing list discussion: > > > http://dpdk.org/ml/archives/dev/2018-January/087484.html > > > > > > Signed-off-by: Gage Eads > > > > > > > > > /** > > > - * Stop an event device. The device can be restarted with a call to > > > - * rte_event_dev_start() > > > + * Stop an event device. > > > + * > > > + * This function causes all queued events to be drained. While > > > + draining > > > events > > > + * out of the device, this function calls the user-provided flush > > > + callback > > > + * (if one was registered) once per event. > > > + * > > > + * This function does not drain events from event ports; the > > > + application is > > > + * responsible for flushing events from all ports before stopping th= e > > > device. > > > > > > Question about how an application is expected to correctly cleanup all = the > > events here. Note in particular the last part: "application is responsi= ble > for > > flushing events from all ports **BEFORE** stopping the device". > > > > Given the event device is still running, how can the application be sur= e it > has > > flushed all the events (from the dequeue side in particular)? > > >=20 > Appreciate the feedback -- good points all around. >=20 > I was expecting that the application would unlink queues from the ports, = and > then dequeue until each port has no events. However, there are PMDs for w= hich > runtime port link/unlink is not supported, so I see that this is not a vi= able > approach. Plus, this adds the application burden that you describe below. +1. > > > > In order to drain all events from the ports, I was expecting the follow= ing: > > > > // stop scheduling new events to worker cores > > rte_event_dev_stop() > > ---> callback gets called for each event > > > > // to dequeue events from each port, and app cleans them up? > > FOR_EACH_PORT( rte_event_dev_dequeue(..., port_id, ...) ) > > > > > > I'd like to avoid the dequeue-each-port() approach in application, as i= t > adds extra > > burden to clean up correctly... >=20 > Agreed, but for a different reason: that approach means we'd have to chan= ge > the documented eventdev behavior. rte_eventdev.h states that the "schedul= e, > enqueue and dequeue functions should not be invoked when the device is > stopped," and this patch reiterates that in the rte_event_dev_stop() > documentation ("Threads that continue to enqueue/dequeue while the device= is > stopped, or being stopped, will result in undefined behavior"). Since a P= MD's > stop cleanup code could just be repeated calls to a PMD's dequeue code, > allowing applications to dequeue simultaneously could be troublesome. All +1 too, good point about the header stating it is undefined behavior. > > What if we say that dequeue() returns zero after stop() (leaving events > possibly > > in the port-dequeue side SW buffers), and these events which were about= to > be > > dequeued by the worker core are also passed to the dev_stop_flush callb= ack? >=20 > I'd prefer to have dequeue-while-stopped be unsupported, so we don't need= an > additional check or synchronization in the datapath, but passing the even= ts in > a port to the callback should work (for the sw PMD, at least). How does t= hat > sound? That's fine with me, both from design point of view, and SW PMD. @HW PMD maintainers, would the above approach work for you?