DPDK patches and discussions
 help / color / mirror / Atom feed
From: Stephen Hemminger <stephen@networkplumber.org>
To: Remy Horton <remy.horton@intel.com>
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH v4 1/3] rte: add keep alive functionality
Date: Wed, 11 Nov 2015 08:28:01 -0800	[thread overview]
Message-ID: <20151111082801.3eedaf21@xeon-e3> (raw)
In-Reply-To: <1446723178-14876-2-git-send-email-remy.horton@intel.com>


In general, try not to introduce new thinks and avoid extra code.
Also Linux has more robust mechanism in the watchdog timers, why not use that?

Could you use RTE_PER_LCORE some how?




> +#ifdef KEEPALIVE_DEBUG_MSGS
Any #ifdef must have a config option to enable it.

> +static void
> +print_trace(const char *msg, struct rte_keepalive *keepcfg, int idx_core)
> +{
> +	printf("%sLast seen %" PRId64  "ms ago.\n",
> +		msg,
> +		((rte_rdtsc() - keepcfg->last_alive[idx_core])*1000)
> +		/ rte_get_tsc_hz()
> +	      );
> +}
> +#else
> +static void
> +print_trace(__attribute__((unused)) const char *msg,
> +	__attribute__((unused)) struct rte_keepalive *keepcfg,
> +	__attribute__((unused)) int idx_core)
> +{
> +}
> +#endif

Agree with others don't introduce not logging macros.


> +void
> +rte_keepalive_dispatch_pings(__attribute__((unused)) void *ptr_timer,

Use __rte_unused
> +	void *ptr_data)
> +{
> +	struct rte_keepalive *keepcfg = (struct rte_keepalive *)ptr_data;

Cast of void  * is unnecessary in C.

> +	int idx_core;
> +
> +	for (idx_core = 0; idx_core < RTE_KEEPALIVE_MAXCORES; idx_core++) {
> +		if (keepcfg->active_cores[idx_core] == 0)
> +			continue;
> +		switch (keepcfg->state_flags[idx_core]) {

My personal preference, prefer blank line after continue.

> +		case ALIVE: /* Alive */
> +			keepcfg->state_flags[idx_core] = 0;
> +			keepcfg->last_alive[idx_core] = rte_rdtsc();
> +			break;
> +		case MISSING: /* MIA */
> +			print_trace("Core MIA. ", keepcfg, idx_core);
> +			keepcfg->state_flags[idx_core] = 2;
> +			break;
> +		case DEAD: /* Dead */
> +			keepcfg->state_flags[idx_core] = 3;
> +			print_trace("Core died. ", keepcfg, idx_core);
> +			if (keepcfg->callback)
> +				keepcfg->callback(
> +					keepcfg->callback_data,
> +					idx_core
> +					);
> +			break;
> +		case GONE: /* Buried */
> +			break;
> +		}
> +	}
> +}
> +
> +
> +struct rte_keepalive *
> +rte_keepalive_create(rte_keepalive_failure_callback_t callback,
> +	void *data)
> +{
> +	int idx_core;
> +	struct rte_keepalive *keepcfg;
> +
> +	keepcfg = malloc(sizeof(struct rte_keepalive));

Why not use rte_zmalloc()?

> +	if (keepcfg != NULL) {
> +		for (idx_core = 0;
> +				idx_core < RTE_KEEPALIVE_MAXCORES;
> +				idx_core++) {
> +			keepcfg->state_flags[idx_core] = 0;
> +			keepcfg->active_cores[idx_core] = 0;
> +		}
> +		keepcfg->callback = callback;
> +		keepcfg->callback_data = data;
> +		keepcfg->tsc_initial = rte_rdtsc();
> +		keepcfg->tsc_mhz = rte_get_tsc_hz() / 1000;
> +	}
> +	return keepcfg;
> +}
> +
> +
> +void
> +rte_keepalive_register_core(struct rte_keepalive *keepcfg, const int id_core)
> +{
> +	if (id_core < RTE_KEEPALIVE_MAXCORES)
> +		keepcfg->active_cores[id_core] = 1;
> +}

  parent reply	other threads:[~2015-11-11 16:27 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-09-15 12:16 [dpdk-dev] [RFC PATCH v1] rte: LCore heartbeat example Remy Horton
2015-09-15 13:10 ` Thomas Monjalon
2015-09-23  9:00   ` Remy Horton
2015-10-05 11:16   ` Mcnamara, John
2015-09-30  9:04 ` [dpdk-dev] [PATCH v2 0/3] Keepalive monitoring & reporting Remy Horton
2015-09-30  9:04   ` [dpdk-dev] [PATCH v2 1/3] rte: add keep alive functionality Remy Horton
2015-10-23 11:40     ` Tahhan, Maryam
2015-10-23 14:27     ` Wiles, Keith
2015-10-26 16:36       ` Remy Horton
2015-09-30  9:04   ` [dpdk-dev] [PATCH v2 2/3] l2fwd: keep alive sample application Remy Horton
2015-10-23 14:23     ` Tahhan, Maryam
2015-10-27  8:48       ` Remy Horton
2015-10-27 10:10         ` Bruce Richardson
2015-09-30  9:04   ` [dpdk-dev] [PATCH v2 3/3] docs: add keep alive sample app guide Remy Horton
2015-10-28  8:52   ` [dpdk-dev] [PATCH v3 0/3] Keepalive monitoring & reporting Remy Horton
2015-10-28  8:52     ` [dpdk-dev] [PATCH v3 1/3] rte: add keep alive functionality Remy Horton
2015-10-28 12:25       ` Tahhan, Maryam
2015-11-04  1:48       ` Thomas Monjalon
2015-11-04  1:54       ` Thomas Monjalon
2015-10-28  8:52     ` [dpdk-dev] [PATCH v3 2/3] docs: add keep alive sample app guide Remy Horton
2015-10-28 11:22       ` Van Haaren, Harry
2015-10-28  8:52     ` [dpdk-dev] [PATCH v3 3/3] example: add keep alive sample application Remy Horton
2015-10-28 12:28       ` Tahhan, Maryam
2015-11-05 11:32     ` [dpdk-dev] [PATCH v4 0/3] Keepalive monitoring & reporting Remy Horton
2015-11-05 11:32       ` [dpdk-dev] [PATCH v4 1/3] rte: add keep alive functionality Remy Horton
2015-11-05 16:43         ` Tahhan, Maryam
2015-11-10 14:02         ` Thomas Monjalon
2015-11-11  9:21           ` Remy Horton
2015-11-11 16:28         ` Stephen Hemminger [this message]
2015-11-18 10:21           ` Remy Horton
2015-11-13 16:09         ` Thomas Monjalon
2015-11-05 11:32       ` [dpdk-dev] [PATCH v4 2/3] docs: add keep alive sample app guide & release notes Remy Horton
2015-11-05 16:40         ` Van Haaren, Harry
2015-11-05 11:32       ` [dpdk-dev] [PATCH v4 3/3] example: add keep alive sample application Remy Horton
2015-11-05 16:44         ` Tahhan, Maryam
2015-11-11  6:52       ` [dpdk-dev] [PATCH v4 0/3] Keepalive monitoring & reporting Cao, Min
2015-11-18 14:05       ` [dpdk-dev] [PATCH v5 " Remy Horton
2015-11-18 14:05         ` [dpdk-dev] [PATCH v5 1/3] rte: add keep alive functionality Remy Horton
2015-11-18 14:05         ` [dpdk-dev] [PATCH v5 2/3] docs: add keep alive sample app guide & release notes Remy Horton
2015-11-19 11:32           ` Thomas Monjalon
2015-11-19 11:43             ` Remy Horton
2015-11-18 14:05         ` [dpdk-dev] [PATCH v5 3/3] example: add keep alive sample application Remy Horton
2015-11-19 11:31           ` Thomas Monjalon
2015-11-19 14:49         ` [dpdk-dev] [PATCH v5 0/3] Keepalive monitoring & reporting Thomas Monjalon

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=20151111082801.3eedaf21@xeon-e3 \
    --to=stephen@networkplumber.org \
    --cc=dev@dpdk.org \
    --cc=remy.horton@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).