DPDK patches and discussions
 help / color / mirror / Atom feed
From: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
To: Chengwen Feng <fengchengwen@huawei.com>,
	thomas@monjalon.net, ferruh.yigit@xilinx.com,
	ndabilpuram@marvell.com, kirankumark@marvell.com,
	skori@marvell.com, skoteshwar@marvell.com
Cc: mb@smartsharesystems.com, stephen@networkplumber.org, dev@dpdk.org
Subject: Re: [PATCH v3 08/10] ethdev: support auto-filled flag when telemetry stats
Date: Thu, 12 May 2022 13:26:48 +0300	[thread overview]
Message-ID: <74e9be9d-cf50-20fe-2d78-c92ca0e2fa13@oktetlabs.ru> (raw)
In-Reply-To: <20220505080233.12737-9-fengchengwen@huawei.com>

On 5/5/22 11:02, Chengwen Feng wrote:
> This patch supports auto-filled queue xstats when telemetry stats.
> 
> Fixes: f30e69b41f94 ("ethdev: add device flag to bypass auto-filled queue xstats")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
> ---
>   lib/ethdev/rte_ethdev.c | 14 +++++++++-----
>   1 file changed, 9 insertions(+), 5 deletions(-)
> 
> diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> index f26a9bac6d..76037e635b 100644
> --- a/lib/ethdev/rte_ethdev.c
> +++ b/lib/ethdev/rte_ethdev.c
> @@ -5499,6 +5499,7 @@ eth_dev_handle_port_stats(const char *cmd __rte_unused,
>   		struct rte_tel_data *d)
>   {
>   	struct rte_eth_stats stats;
> +	struct rte_eth_dev *dev;
>   	int port_id, ret;
>   
>   	if (params == NULL || strlen(params) == 0 || !isdigit(*params))
> @@ -5507,6 +5508,7 @@ eth_dev_handle_port_stats(const char *cmd __rte_unused,
>   	port_id = atoi(params);
>   	if (!rte_eth_dev_is_valid_port(port_id))
>   		return -1;
> +	dev = &rte_eth_devices[port_id];
>   
>   	ret = rte_eth_stats_get(port_id, &stats);
>   	if (ret < 0)
> @@ -5521,11 +5523,13 @@ eth_dev_handle_port_stats(const char *cmd __rte_unused,
>   	ADD_DICT_STAT(stats, ierrors);
>   	ADD_DICT_STAT(stats, oerrors);
>   	ADD_DICT_STAT(stats, rx_nombuf);
> -	eth_dev_add_port_queue_stats(d, stats.q_ipackets, "q_ipackets");
> -	eth_dev_add_port_queue_stats(d, stats.q_opackets, "q_opackets");
> -	eth_dev_add_port_queue_stats(d, stats.q_ibytes, "q_ibytes");
> -	eth_dev_add_port_queue_stats(d, stats.q_obytes, "q_obytes");
> -	eth_dev_add_port_queue_stats(d, stats.q_errors, "q_errors");
> +	if (dev->data->dev_flags & RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS) {

I don't understand it. RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS driver
flag means that driver asks ethdev layer to autofill per-queue
*xstats* (generate names and put values) using per-queue basic
statistics.

Drivers do no set the flag if provide per-queue xstats itself
(to avoid duplication) or do not fill in per-queue basic
xstats (to avoid meaningless/misleading zeros in xstats).

So, absence of the flag does not mean that per-queue basic
statistics are not filled in and should be used as the
guard to include corresponding values in telemetry.

> +		eth_dev_add_port_queue_stats(d, stats.q_ipackets, "q_ipackets");
> +		eth_dev_add_port_queue_stats(d, stats.q_opackets, "q_opackets");
> +		eth_dev_add_port_queue_stats(d, stats.q_ibytes, "q_ibytes");
> +		eth_dev_add_port_queue_stats(d, stats.q_obytes, "q_obytes");
> +		eth_dev_add_port_queue_stats(d, stats.q_errors, "q_errors");
> +	}
>   
>   	return 0;
>   }


  reply	other threads:[~2022-05-12 10:26 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-04-16  1:07 [PATCH 0/3] bugfix for ethdev telemetry Chengwen Feng
2022-04-16  1:07 ` [PATCH 1/3] ethdev: fix telemetry xstats return null with some PMDs Chengwen Feng
2022-04-16  1:38   ` Stephen Hemminger
2022-04-21  6:49     ` Andrew Rybchenko
2022-04-24  3:44       ` fengchengwen
2022-04-25 10:16         ` Morten Brørup
2022-04-28 13:38           ` fengchengwen
2022-04-28 13:53             ` Morten Brørup
2022-04-16  1:07 ` [PATCH 2/3] ethdev: fix memory leak when telemetry xstats Chengwen Feng
2022-04-21  6:51   ` Andrew Rybchenko
2022-04-21  8:09   ` David Marchand
2022-04-21  9:03     ` Bruce Richardson
2022-04-22  8:14       ` David Marchand
2022-04-16  1:07 ` [PATCH 3/3] net/cnxk: fix telemetry possible null pointer access Chengwen Feng
2022-04-28 13:15 ` [PATCH v2 0/9] bugfix for ethdev telemetry Chengwen Feng
2022-04-28 13:15   ` [PATCH v2 1/9] ethdev: define retval when xstats is null of get xstats Chengwen Feng
2022-05-04 10:32     ` Andrew Rybchenko
2022-04-28 13:15   ` [PATCH v2 2/9] net/hns3: adjust " Chengwen Feng
2022-05-06  7:56     ` Min Hu (Connor)
2022-04-28 13:15   ` [PATCH v2 3/9] net/ipn3ke: " Chengwen Feng
2022-04-28 13:15   ` [PATCH v2 4/9] net/mvpp2: " Chengwen Feng
2022-04-28 13:15   ` [PATCH v2 5/9] net/axgbe: " Chengwen Feng
2022-04-28 13:15   ` [PATCH v2 6/9] ethdev: fix memory leak when telemetry xstats Chengwen Feng
2022-04-28 13:15   ` [PATCH v2 7/9] ethdev: support auto-filled flag when telemetry stats Chengwen Feng
2022-04-28 13:15   ` [PATCH v2 8/9] ethdev: fix possible null pointer access Chengwen Feng
2022-04-28 13:16   ` [PATCH v2 9/9] net/cnxk: fix telemetry " Chengwen Feng
2022-05-05  8:02 ` [PATCH v3 00/10] bugfix for ethdev telemetry Chengwen Feng
2022-05-05  8:02   ` [PATCH v3 01/10] ethdev: define retval when xstats is null of get xstats Chengwen Feng
2022-05-05  8:02   ` [PATCH v3 02/10] ethdev: optimize xstats-get API's implementation Chengwen Feng
2022-05-12 10:31     ` Andrew Rybchenko
2022-05-05  8:02   ` [PATCH v3 03/10] net/hns3: adjust retval when xstats is null of get xstats Chengwen Feng
2022-05-12  9:51     ` Andrew Rybchenko
2022-05-05  8:02   ` [PATCH v3 04/10] net/ipn3ke: " Chengwen Feng
2022-05-05  8:02   ` [PATCH v3 05/10] net/mvpp2: " Chengwen Feng
2022-05-05  8:02   ` [PATCH v3 06/10] net/axgbe: " Chengwen Feng
2022-05-12 10:00     ` Andrew Rybchenko
2022-05-05  8:02   ` [PATCH v3 07/10] ethdev: fix memory leak when telemetry xstats Chengwen Feng
2022-05-05  8:02   ` [PATCH v3 08/10] ethdev: support auto-filled flag when telemetry stats Chengwen Feng
2022-05-12 10:26     ` Andrew Rybchenko [this message]
2022-05-05  8:02   ` [PATCH v3 09/10] ethdev: fix possible null pointer access Chengwen Feng
2022-05-12 10:29     ` Andrew Rybchenko
2022-05-05  8:02   ` [PATCH v3 10/10] net/cnxk: fix telemetry " Chengwen Feng
2022-05-12 10:30     ` Andrew Rybchenko
2022-05-13  2:53 ` [PATCH v4 0/9] bugfix for ethdev telemetry Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 1/9] ethdev: specify return value of xstats-get API Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 2/9] ethdev: optimize xstats-get API's implementation Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 3/9] net/hns3: adjust return value of xstats-get ops Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 4/9] net/ipn3ke: " Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 5/9] net/mvpp2: " Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 6/9] net/axgbe: " Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 7/9] ethdev: fix memory leak when telemetry xstats Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 8/9] ethdev: fix possible null pointer access Chengwen Feng
2022-05-13  2:53   ` [PATCH v4 9/9] net/cnxk: fix telemetry " Chengwen Feng
2022-05-16  8:43   ` [PATCH v4 0/9] bugfix for ethdev telemetry Morten Brørup
2022-05-20 14:50     ` Andrew Rybchenko

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=74e9be9d-cf50-20fe-2d78-c92ca0e2fa13@oktetlabs.ru \
    --to=andrew.rybchenko@oktetlabs.ru \
    --cc=dev@dpdk.org \
    --cc=fengchengwen@huawei.com \
    --cc=ferruh.yigit@xilinx.com \
    --cc=kirankumark@marvell.com \
    --cc=mb@smartsharesystems.com \
    --cc=ndabilpuram@marvell.com \
    --cc=skori@marvell.com \
    --cc=skoteshwar@marvell.com \
    --cc=stephen@networkplumber.org \
    --cc=thomas@monjalon.net \
    /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).