From: Ferruh Yigit <ferruh.yigit@amd.com>
To: Chengwen Feng <fengchengwen@huawei.com>,
thomas@monjalon.net, andrew.rybchenko@oktetlabs.ru
Cc: dev@dpdk.org, ciara.power@intel.com, bruce.richardson@intel.com
Subject: Re: [PATCH v2] ethdev: telemetry xstats support hide zero
Date: Mon, 13 Feb 2023 18:18:33 +0000 [thread overview]
Message-ID: <044c4354-15c0-17fb-0628-019a148d1804@amd.com> (raw)
In-Reply-To: <20230213023456.45972-1-fengchengwen@huawei.com>
On 2/13/2023 2:34 AM, Chengwen Feng wrote:
> The number of xstats may be large, after the hide zero option is added,
> only non-zero values can be displayed.
>
> So display xstats with hide zero:
> /ethdev/xstats,0,hide_zero=true
> and without hide zero (same as the original):
> /ethdev/xstats,0
>
> Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
>
> ---
> v2: using hide_zero=true optional parameter which address Ferruh's
> comments.
>
> ---
> lib/ethdev/rte_ethdev.c | 38 ++++++++++++++++++++++++++++++++------
> 1 file changed, 32 insertions(+), 6 deletions(-)
>
> diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> index d25db35f7f..7e70292fc6 100644
> --- a/lib/ethdev/rte_ethdev.c
> +++ b/lib/ethdev/rte_ethdev.c
> @@ -15,6 +15,7 @@
> #include <bus_driver.h>
> #include <rte_log.h>
> #include <rte_interrupts.h>
> +#include <rte_kvargs.h>
> #include <rte_memcpy.h>
> #include <rte_common.h>
> #include <rte_mempool.h>
> @@ -5863,6 +5864,18 @@ eth_dev_handle_port_stats(const char *cmd __rte_unused,
> return 0;
> }
>
> +#define TEL_XSTATS_HIDE_ZERO_KEY "hide_zero"
> +#define TEL_XSTATS_HIDE_ZERO_VAL "true"
> +
> +static int
> +eth_dev_parse_hide_zero(const char *key, const char *value, void *extra_args)
> +{
> + RTE_SET_USED(key);
> + if (strcmp(value, TEL_XSTATS_HIDE_ZERO_VAL) == 0)
kvargs allows to provide parameter without value, like
'/ethdev/xstats,0,hide_zero'
In this case 'value' will be NULL and this cause a crash in dpdk
application.
Better to check that 'value' is not NULL before using it.
> + *(bool *)extra_args = true;
> + return 0;
> +}
> +
> static int
> eth_dev_handle_port_xstats(const char *cmd __rte_unused,
> const char *params,
> @@ -5870,20 +5883,30 @@ eth_dev_handle_port_xstats(const char *cmd __rte_unused,
> {
> struct rte_eth_xstat *eth_xstats;
> struct rte_eth_xstat_name *xstat_names;
> + struct rte_kvargs *kvlist;
> int port_id, num_xstats;
> - int i, ret;
> + bool hide_zero = false;
> char *end_param;
> + int i, ret;
>
> if (params == NULL || strlen(params) == 0 || !isdigit(*params))
> return -1;
>
> port_id = strtoul(params, &end_param, 0);
> - if (*end_param != '\0')
> - RTE_ETHDEV_LOG(NOTICE,
> - "Extra parameters passed to ethdev telemetry command, ignoring");
> if (!rte_eth_dev_is_valid_port(port_id))
> return -1;
>
> + if (*end_param != '\0') {
> + kvlist = rte_kvargs_parse(end_param, NULL);
> + if (kvlist != NULL)
> + rte_kvargs_process(kvlist, TEL_XSTATS_HIDE_ZERO_KEY,
> + eth_dev_parse_hide_zero, &hide_zero);
> + if (kvlist == NULL || !hide_zero || kvlist->count > 1)
Instead of "kvlist->count > 1" can be better to use 'valid_keys'
argument of 'rte_kvargs_parse()' to detect extra parameters, this way
multiple 'hide_zero' parameters or future expansion of args won't be
blocked.
Similarly, '!hide_zero' check prevents "hide_zero=false" usage, it is
not documented but this is expected to be supported intuitively, perhaps
can be better to support "hide_zero=false" fully, document it and remove
'!hide_zero' from check?
> + RTE_ETHDEV_LOG(NOTICE,
> + "Unknown extra parameters passed to ethdev telemetry command, ignoring\n");
> + rte_kvargs_free(kvlist);
> + }
> +
> num_xstats = rte_eth_xstats_get(port_id, NULL, 0);
> if (num_xstats < 0)
> return -1;
> @@ -5908,9 +5931,12 @@ eth_dev_handle_port_xstats(const char *cmd __rte_unused,
> }
>
> rte_tel_data_start_dict(d);
> - for (i = 0; i < num_xstats; i++)
> + for (i = 0; i < num_xstats; i++) {
> + if (hide_zero && eth_xstats[i].value == 0)
> + continue;
> rte_tel_data_add_dict_uint(d, xstat_names[i].name,
> eth_xstats[i].value);
> + }
> free(eth_xstats);
> return 0;
> }
> @@ -6328,7 +6354,7 @@ RTE_INIT(ethdev_init_telemetry)
> rte_telemetry_register_cmd("/ethdev/stats", eth_dev_handle_port_stats,
> "Returns the common stats for a port. Parameters: int port_id");
> rte_telemetry_register_cmd("/ethdev/xstats", eth_dev_handle_port_xstats,
> - "Returns the extended stats for a port. Parameters: int port_id");
> + "Returns the extended stats for a port. Parameters: int port_id,hide_zero=true(Optional for indicates hide zero xstats)");
> #ifndef RTE_EXEC_ENV_WINDOWS
> rte_telemetry_register_cmd("/ethdev/dump_priv", eth_dev_handle_port_dump_priv,
> "Returns dump private information for a port. Parameters: int port_id");
next prev parent reply other threads:[~2023-02-13 18:18 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-09 3:07 [PATCH] " Chengwen Feng
2023-02-09 23:30 ` Ferruh Yigit
2023-02-10 1:23 ` fengchengwen
2023-02-10 11:55 ` Ferruh Yigit
2023-02-13 2:44 ` fengchengwen
2023-02-13 2:34 ` [PATCH v2] " Chengwen Feng
2023-02-13 18:18 ` Ferruh Yigit [this message]
2023-02-14 1:57 ` fengchengwen
2023-02-14 1:35 ` [PATCH v3] " Chengwen Feng
2023-02-14 12:59 ` Ferruh Yigit
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=044c4354-15c0-17fb-0628-019a148d1804@amd.com \
--to=ferruh.yigit@amd.com \
--cc=andrew.rybchenko@oktetlabs.ru \
--cc=bruce.richardson@intel.com \
--cc=ciara.power@intel.com \
--cc=dev@dpdk.org \
--cc=fengchengwen@huawei.com \
--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).