From: Noa Ezra <noae@mellanox.com>
To: Maxime Coquelin <maxime.coquelin@redhat.com>,
"tiwei.bie@intel.com" <tiwei.bie@intel.com>,
"zhihong.wang@intel.com" <zhihong.wang@intel.com>
Cc: Matan Azrad <matan@mellanox.com>, "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [Suspected-Phishing][PATCH] net/vhost: add an API for get queue status
Date: Thu, 12 Sep 2019 10:04:53 +0000 [thread overview]
Message-ID: <AM0PR05MB438805B910972E1A9C02BAFCA0B00@AM0PR05MB4388.eurprd05.prod.outlook.com> (raw)
In-Reply-To: <bfbddee4-ecb7-2850-8c51-e404d89d6c4c@redhat.com>
Hi Maxime,
For me, the solution you suggested is good enough:
"call rte_eth_vhost_get_queue_event() in loop at startup until you get -1 and build the states based on that"
Getting the state on a direct way can be good and easier than reading it in a loop, but I'm not sure that your suggestion will solve the problem of getting queue state for requested queue directly in the application.
Thanks,
Noa.
> -----Original Message-----
> From: Maxime Coquelin [mailto:maxime.coquelin@redhat.com]
> Sent: Friday, August 30, 2019 11:56 AM
> To: Noa Ezra <noae@mellanox.com>; tiwei.bie@intel.com;
> zhihong.wang@intel.com
> Cc: Matan Azrad <matan@mellanox.com>; dev@dpdk.org
> Subject: Re: [Suspected-Phishing][PATCH] net/vhost: add an API for get
> queue status
>
> Hi Noa,
>
> I was thinking about an alternative that would avoid adding an API.
> What about the Vhost-user library to replay the queue status for all
> configured queues when the device is ready (i.e. after it has called its
> .new_device() callback)?
>
> On 6/24/19 1:08 PM, Noa Ezra wrote:
> > Hi,
> > What do you say about this patch?
> >
> > Thanks,
> > Noa.
> >
> >> -----Original Message-----
> >> From: Noa Ezra [mailto:noae@mellanox.com]
> >> Sent: Wednesday, June 19, 2019 9:15 AM
> >> To: maxime.coquelin@redhat.com
> >> Cc: Matan Azrad <matan@mellanox.com>; dev@dpdk.org; Noa Ezra
> >> <noae@mellanox.com>
> >> Subject: [Suspected-Phishing][PATCH] net/vhost: add an API for get
> >> queue status
> >>
> >> Add an API that returns queue status for requested queue in the port.
> >> The queue's status can be changed before the user has signed for the
> >> queue state event interrupt. In this case the user can't know the
> >> current queue's status. This API returns the current status.
> >>
> >> Signed-off-by: Noa Ezra <noae@mellanox.com>
> >> Reviewed-by: Matan Azrad <matan@mellanox.com>
> >> ---
> >> drivers/net/vhost/rte_eth_vhost.c | 47
> >> +++++++++++++++++++++++++++++
> >> drivers/net/vhost/rte_eth_vhost.h | 18 +++++++++++
> >> drivers/net/vhost/rte_pmd_vhost_version.map | 6 ++++
> >> 3 files changed, 71 insertions(+)
> >>
> >> diff --git a/drivers/net/vhost/rte_eth_vhost.c
> >> b/drivers/net/vhost/rte_eth_vhost.c
> >> index 9a54020..cad1e5c 100644
> >> --- a/drivers/net/vhost/rte_eth_vhost.c
> >> +++ b/drivers/net/vhost/rte_eth_vhost.c
> >> @@ -855,6 +855,7 @@ struct vhost_xstats_name_off {
> >> /* won't be NULL */
> >> state = vring_states[eth_dev->data->port_id];
> >> rte_spinlock_lock(&state->lock);
> >> +
> >> state->cur[vring] = enable;
> >> state->max_vring = RTE_MAX(vring, state->max_vring);
> >> rte_spinlock_unlock(&state->lock);
> >> @@ -874,6 +875,52 @@ struct vhost_xstats_name_off { };
> >>
> >> int
> >> +rte_eth_vhost_get_queue_status(uint16_t port_id, bool rx, uint16_t
> >> queue_id,
> >> + bool *queue_status)
> >> +{
> >> + struct rte_vhost_vring_state *state;
> >> + struct internal_list *list;
> >> + struct rte_eth_dev *eth_dev;
> >> + int found = 0;
> >> + uint16_t nb_q = 0;
> >> +
> >> + if (port_id >= RTE_MAX_ETHPORTS) {
> >> + VHOST_LOG(ERR, "Invalid port id\n");
> >> + return -1;
> >> + }
> >> + TAILQ_FOREACH(list, &internal_list, next) {
> >> + eth_dev = list->eth_dev;
> >> + if (eth_dev->data->port_id == port_id) {
> >> + nb_q = rx ? eth_dev->data->nb_rx_queues :
> >> + eth_dev->data->nb_tx_queues;
> >> + found = 1;
> >> + break;
> >> + }
> >> + }
> >> + if (!found) {
> >> + VHOST_LOG(ERR, "No device found for port id %u\n",
> >> port_id);
> >> + return -1;
> >> + }
> >> + if (queue_id >= nb_q) {
> >> + VHOST_LOG(ERR, "Invalid queue id\n");
> >> + return -1;
> >> + }
> >> +
> >> + state = vring_states[port_id];
> >> + if (!state) {
> >> + VHOST_LOG(ERR, "Unused port\n");
> >> + return -1;
> >> + }
> >> +
> >> + rte_spinlock_lock(&state->lock);
> >> + *queue_status = rx ? state->cur[queue_id * 2 + 1] :
> >> + state->cur[queue_id * 2];
> >> + rte_spinlock_unlock(&state->lock);
> >> +
> >> + return 0;
> >> +}
> >> +
> >> +int
> >> rte_eth_vhost_get_queue_event(uint16_t port_id,
> >> struct rte_eth_vhost_queue_event *event) { diff --git
> >> a/drivers/net/vhost/rte_eth_vhost.h
> >> b/drivers/net/vhost/rte_eth_vhost.h
> >> index 0e68b9f..1e65c69 100644
> >> --- a/drivers/net/vhost/rte_eth_vhost.h
> >> +++ b/drivers/net/vhost/rte_eth_vhost.h
> >> @@ -44,6 +44,24 @@ int rte_eth_vhost_get_queue_event(uint16_t
> port_id,
> >> struct rte_eth_vhost_queue_event *event);
> >>
> >> /**
> >> + * Get queue status for specific queue in the port.
> >> + *
> >> + * @param[in] port_id
> >> + * Port id.
> >> + * @param[in] rx
> >> + * True is rx, False if tx
> >> + * @paran[in] queue_id
> >> + * Queue_id
> >> + * @param[out] queue_status
> >> + * Pointer to a boolean, True is enable, False if disable.
> >> + * @return
> >> + * - On success, zero, queue_status is updated.
> >> + * - On failure, a negative value, queue_status is not updated.
> >> + */
> >> +int rte_eth_vhost_get_queue_status(uint16_t port_id, bool rx,
> >> +uint16_t
> >> queue_id,
> >> + bool *queue_status);
> >> +
> >> +/**
> >> * Get the 'vid' value associated with the specified port.
> >> *
> >> * @return
> >> diff --git a/drivers/net/vhost/rte_pmd_vhost_version.map
> >> b/drivers/net/vhost/rte_pmd_vhost_version.map
> >> index 695db85..1eabfd2 100644
> >> --- a/drivers/net/vhost/rte_pmd_vhost_version.map
> >> +++ b/drivers/net/vhost/rte_pmd_vhost_version.map
> >> @@ -11,3 +11,9 @@ DPDK_16.11 {
> >>
> >> rte_eth_vhost_get_vid_from_port_id;
> >> };
> >> +
> >> +DPDK_19.08 {
> >> + global:
> >> +
> >> + rte_eth_vhost_get_queue_status;
> >> +};
> >> --
> >> 1.8.3.1
> >
prev parent reply other threads:[~2019-09-12 10:04 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-06-19 6:14 [dpdk-dev] [PATCH] net/vhost: add an API for get queue status Noa Ezra
2019-06-24 11:08 ` [dpdk-dev] [Suspected-Phishing][PATCH] " Noa Ezra
2019-06-24 16:47 ` Maxime Coquelin
2019-06-25 7:00 ` Noa Ezra
2019-06-25 7:36 ` Maxime Coquelin
2019-06-25 7:49 ` Maxime Coquelin
2019-08-30 8:55 ` Maxime Coquelin
2019-09-12 10:04 ` Noa Ezra [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=AM0PR05MB438805B910972E1A9C02BAFCA0B00@AM0PR05MB4388.eurprd05.prod.outlook.com \
--to=noae@mellanox.com \
--cc=dev@dpdk.org \
--cc=matan@mellanox.com \
--cc=maxime.coquelin@redhat.com \
--cc=tiwei.bie@intel.com \
--cc=zhihong.wang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).