From: "Van Haaren, Harry" <harry.van.haaren@intel.com>
To: "Horton, Remy" <remy.horton@intel.com>, "dev@dpdk.org" <dev@dpdk.org>
Cc: Thomas Monjalon <thomas.monjalon@6wind.com>
Subject: Re: [dpdk-dev] [PATCH v7 1/6] lib: add information metrics library
Date: Tue, 17 Jan 2017 14:23:56 +0000 [thread overview]
Message-ID: <E923DB57A917B54B9182A2E928D00FA6129E3D80@IRSMSX102.ger.corp.intel.com> (raw)
In-Reply-To: <0d2e03a5-3eb0-ec3f-9fe7-82e905d3adb3@intel.com>
> -----Original Message-----
> From: Horton, Remy
> Sent: Tuesday, January 17, 2017 1:40 PM
> To: Van Haaren, Harry <harry.van.haaren@intel.com>; dev@dpdk.org
> Cc: Thomas Monjalon <thomas.monjalon@6wind.com>
> Subject: Re: [dpdk-dev] [PATCH v7 1/6] lib: add information metrics library
>
>
> On 17/01/2017 11:01, Van Haaren, Harry wrote:
> >> -----Original Message-----
> >> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Remy Horton
> [..]
> > Nit: why the _s in rte_metrics_meta_s? Is there a reason why struct
> > rte_metrics_meta { is not suitable? Same question for
> > rte_metrics_data_s.
>
> Think it was originally to avoid a namespace collision, and I left the
> suffix in because this is an internal data-structure.
OK.
> >> + if (memzone == NULL)
> >> + rte_exit(EXIT_FAILURE, "Unable to allocate stats memzone\n");
> >> + stats = memzone->addr;
> >> + memset(stats, 0, sizeof(struct rte_metrics_data_s));
> >> + rte_spinlock_init(&stats->lock);
> >
> > Should this spinlock be initialized before the creation of the
> > memzone? Creating the memzone first, and later initializing the
> > locking-structure for it feels racey to me. The lock should probably
> > be taken and unlocked again after init and memset.
>
> Problem is the lock being part of the reserved memzone allocation. It is
> in there so that secondary processes can use the lock.
Ah ok - apologies I missed that the lock is in the memory itself.
>
> > Nit: would rte_metrics_reg_name() be a better function name?
> [..]
> > Nit: would rte_metrics_reg_names() be a better function name?
>
> Agree. Done.
Great
> > Would rte_metrics_update_single() be a better function name?
> [..]
> > Would rte_metrics_update() be a better function name?
>
> Think rte_metrics_update_value & rte_metrics_update_values would make a
> better pair. My preference at the moment is not to nominate a preferred one.
>
>
> >> + if (port_id != RTE_METRICS_GLOBAL &&
> >> + (port_id < 0 || port_id > RTE_MAX_ETHPORTS))
> >> + return -EINVAL;
> >> +
> >> + rte_metrics_init();
> >
> > See above comments on rte_metrics_init() taking a socket_id parameter. Here any core
> could call update_metrics(), and if the library was not yet initialized, the memory for
> metrics would end up on the socket of this core. This should be avoided by
> > A) adding socket_id parameter to the metrics_init() function
> > B) Demanding that metrics_init() is called by the application before any core uses
> update_metrics()
>
> Done. Should also help alleviate the race.
Cool.
> > What is a "set border"? I think this ensures that one set of metrics
> > cannot overwrite the index's of another metrics set - correct?
>
> It is intended to stop things like:
>
> idx1 = rte_metrics_reg_names(some_names, 2);
> idx2 = rte_metrics_reg_names(more_names, 3);
> ...
> rte_metrics_update_values(port_id, idx1, &new_values, 5);
>
> as the above assumes idx2=idx1+2 which is not guaranteed in concurrent
> enviornments
That confirms my understanding - thanks.
next prev parent reply other threads:[~2017-01-17 14:23 UTC|newest]
Thread overview: 115+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-08-24 14:58 [dpdk-dev] [RFC PATCH v1] rte: add bit-rate metrics to xstats Remy Horton
2016-08-26 13:28 ` Pattan, Reshma
2016-08-29 10:01 ` Pattan, Reshma
2016-08-29 11:19 ` Remy Horton
2016-10-28 1:04 ` [dpdk-dev] [PATCH v2 0/3] expanded statistic reporting Remy Horton
2016-10-28 1:04 ` [dpdk-dev] [RFC PATCH v2 1/3] lib: add information metrics library Remy Horton
2016-10-28 1:04 ` [dpdk-dev] [RFC PATCH v2 2/3] lib: add bitrate statistics library Remy Horton
2016-10-28 1:12 ` Stephen Hemminger
2016-10-28 7:48 ` Remy Horton
2016-10-28 7:39 ` Morten Brørup
2016-11-01 1:53 ` Remy Horton
2016-10-28 1:04 ` [dpdk-dev] [RFC PATCH v2 3/3] app/test-pmd: add support for bitrate statistics Remy Horton
2016-11-04 3:36 ` [dpdk-dev] [PATCH v3 0/3] Expanded statistics reporting Remy Horton
2016-11-04 3:36 ` [dpdk-dev] [PATCH v3 1/3] lib: add information metrics library Remy Horton
2016-11-04 16:42 ` Pattan, Reshma
2016-11-07 15:25 ` Pattan, Reshma
2016-11-08 3:19 ` Remy Horton
2016-11-04 3:36 ` [dpdk-dev] [PATCH v3 2/3] lib: add bitrate statistics library Remy Horton
2016-11-04 3:36 ` [dpdk-dev] [PATCH v3 3/3] app/test-pmd: add support for bitrate statistics Remy Horton
2016-11-15 7:15 ` [dpdk-dev] [PATCH v4 0/3] Expanded statistics reporting Remy Horton
2016-11-15 7:15 ` [dpdk-dev] [PATCH v4 1/3] lib: add information metrics library Remy Horton
2016-11-15 7:15 ` [dpdk-dev] [PATCH v4 2/3] lib: add bitrate statistics library Remy Horton
2016-11-15 15:17 ` Pattan, Reshma
2016-11-15 7:15 ` [dpdk-dev] [PATCH v4 3/3] app/test-pmd: add support for bitrate statistics Remy Horton
2016-11-18 8:00 ` [dpdk-dev] [PATCH v5 0/4] Expanded statistics reporting Remy Horton
2016-11-18 8:00 ` [dpdk-dev] [PATCH v5 1/4] lib: add information metrics library Remy Horton
2016-11-18 8:00 ` [dpdk-dev] [PATCH v5 2/4] lib: add bitrate statistics library Remy Horton
2016-11-18 8:00 ` [dpdk-dev] [PATCH v5 3/4] app/test-pmd: add support for bitrate statistics Remy Horton
2016-11-18 8:00 ` [dpdk-dev] [PATCH v5 4/4] latencystats: added new library for latency stats Remy Horton
2017-01-11 16:03 ` [dpdk-dev] [PATCH v6 0/4] Expanded statistics reporting Remy Horton
2017-01-11 16:03 ` [dpdk-dev] [PATCH v6 1/4] lib: add information metrics library Remy Horton
2017-01-12 13:22 ` Thomas Monjalon
2017-01-12 15:30 ` Remy Horton
2017-01-12 19:05 ` Thomas Monjalon
2017-01-16 10:27 ` Remy Horton
2017-01-11 16:03 ` [dpdk-dev] [PATCH v6 2/4] lib: add bitrate statistics library Remy Horton
2017-01-11 16:15 ` Stephen Hemminger
2017-01-16 13:18 ` Remy Horton
2017-01-11 16:03 ` [dpdk-dev] [PATCH v6 3/4] app/test-pmd: add support for bitrate statistics Remy Horton
2017-01-12 13:32 ` Thomas Monjalon
2017-01-11 16:03 ` [dpdk-dev] [PATCH v6 4/4] latencystats: added new library for latency stats Remy Horton
2017-01-12 13:41 ` Thomas Monjalon
2017-01-12 14:44 ` Remy Horton
2017-01-13 9:45 ` Mcnamara, John
2017-01-13 9:53 ` Thomas Monjalon
2017-01-16 16:18 ` Mcnamara, John
2017-01-11 16:58 ` [dpdk-dev] [PATCH v6 0/4] Expanded statistics reporting Thomas Monjalon
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 0/6] " Remy Horton
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 1/6] lib: add information metrics library Remy Horton
2017-01-17 11:01 ` Van Haaren, Harry
2017-01-17 13:40 ` Remy Horton
2017-01-17 14:23 ` Van Haaren, Harry [this message]
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 2/6] app/proc_info: add metrics displaying Remy Horton
2017-01-17 11:08 ` Van Haaren, Harry
2017-01-17 14:27 ` Remy Horton
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 3/6] lib: add bitrate statistics library Remy Horton
2017-01-17 11:16 ` Van Haaren, Harry
2017-01-17 15:37 ` Remy Horton
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 4/6] app/test-pmd: add bitrate statistics calculation Remy Horton
2017-01-17 11:19 ` Van Haaren, Harry
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 5/6] lib: added new library for latency stats Remy Horton
2017-01-17 4:29 ` Jerin Jacob
2017-01-17 6:48 ` Remy Horton
2017-01-17 7:35 ` Jerin Jacob
2017-01-17 11:19 ` Mcnamara, John
2017-01-17 12:34 ` Jerin Jacob
2017-01-17 14:53 ` Mcnamara, John
2017-01-17 16:25 ` Jerin Jacob
2017-01-18 20:11 ` Olivier Matz
2017-01-24 15:24 ` Olivier MATZ
2017-01-17 11:41 ` Van Haaren, Harry
2017-01-16 16:19 ` [dpdk-dev] [PATCH v7 6/6] app/test-pmd: add latency statistics calculation Remy Horton
2017-01-17 11:45 ` Van Haaren, Harry
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 0/7] Expanded statistics reporting Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 1/7] lib: add information metrics library Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 2/7] app/proc_info: add metrics displaying Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 3/7] lib: add bitrate statistics library Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 4/7] app/test-pmd: add bitrate statistics calculation Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 5/7] mbuf: add a timestamp to the mbuf for latencystats Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 6/7] lib: added new library for latency stats Remy Horton
2017-01-17 23:24 ` [dpdk-dev] [PATCH v8 7/7] app/test-pmd: add latency statistics calculation Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 0/7] Expanded statistics reporting Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 1/7] lib: add information metrics library Remy Horton
2017-01-30 15:50 ` Thomas Monjalon
2017-01-30 21:44 ` Remy Horton
2017-01-31 13:13 ` Mcnamara, John
2017-01-31 13:28 ` Bruce Richardson
2017-02-02 17:22 ` Thomas Monjalon
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 2/7] app/proc_info: add metrics displaying Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 3/7] lib: add bitrate statistics library Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 4/7] app/test-pmd: add bitrate statistics calculation Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 5/7] mbuf: add a timestamp to the mbuf for latencystats Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 6/7] lib: added new library for latency stats Remy Horton
2017-01-18 15:05 ` [dpdk-dev] [PATCH v9 7/7] app/test-pmd: add latency statistics calculation Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 0/7] Expanded statistics reporting Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 1/7] lib: add information metrics library Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 2/7] app/proc_info: add metrics displaying Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 3/7] lib: add bitrate statistics library Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 4/7] app/test-pmd: add bitrate statistics calculation Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 5/7] mbuf: add a timestamp to the mbuf for latencystats Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 6/7] lib: added new library for latency stats Remy Horton
2017-02-03 10:33 ` [dpdk-dev] [PATCH v10 7/7] app/test-pmd: add latency statistics calculation Remy Horton
2017-02-16 10:53 ` [dpdk-dev] [PATCH v10 0/7] Expanded statistics reporting Thomas Monjalon
2017-02-23 7:09 ` Remy Horton
2017-02-23 8:45 ` Thomas Monjalon
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 " Remy Horton
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 1/7] lib: add information metrics library Remy Horton
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 2/7] app/proc_info: add metrics displaying Remy Horton
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 3/7] lib: add bitrate statistics library Remy Horton
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 4/7] app/test-pmd: add bitrate statistics calculation Remy Horton
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 5/7] mbuf: add a timestamp to the mbuf for latencystats Remy Horton
2017-03-09 19:02 ` Stephen Hemminger
2017-03-10 9:48 ` Van Haaren, Harry
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 6/7] lib: added new library for latency stats Remy Horton
2017-03-09 16:25 ` [dpdk-dev] [PATCH v11 7/7] app/test-pmd: add latency statistics calculation Remy Horton
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=E923DB57A917B54B9182A2E928D00FA6129E3D80@IRSMSX102.ger.corp.intel.com \
--to=harry.van.haaren@intel.com \
--cc=dev@dpdk.org \
--cc=remy.horton@intel.com \
--cc=thomas.monjalon@6wind.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).