[dpdk-dev] [PATCH 2/2] app/testpmd: add command to resume a TM node

Singh, Jasvinder jasvinder.singh at intel.com
Fri Feb 16 14:10:39 CET 2018



> -----Original Message-----
> From: Tomasz Duszynski [mailto:tdu at semihalf.com]
> Sent: Thursday, February 1, 2018 9:37 AM
> To: dev at dpdk.org
> Cc: Lu, Wenzhuo <wenzhuo.lu at intel.com>; Wu, Jingjing
> <jingjing.wu at intel.com>; Singh, Jasvinder <jasvinder.singh at intel.com>;
> Tomasz Duszynski <tdu at semihalf.com>
> Subject: [PATCH 2/2] app/testpmd: add command to resume a TM node
> 
> Traffic manager provides an API for resuming an arbitrary node in a
> hierarchy.
> 
> This commit adds support for calling this API from testpmd.
> 
> Signed-off-by: Tomasz Duszynski <tdu at semihalf.com>
> ---
>  app/test-pmd/cmdline.c                      |  4 ++
>  app/test-pmd/cmdline_tm.c                   | 70
> +++++++++++++++++++++++++++++
>  app/test-pmd/cmdline_tm.h                   |  1 +
>  doc/guides/testpmd_app_ug/testpmd_funcs.rst |  5 +++
>  4 files changed, 80 insertions(+)
> 
> diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c index
> 6bbd606..f9827f6 100644
> --- a/app/test-pmd/cmdline.c
> +++ b/app/test-pmd/cmdline.c
> @@ -800,6 +800,9 @@ static void cmd_help_long_parsed(void
> *parsed_result,
>  			"suspend port tm node (port_id) (node_id)"
>  			"       Suspend tm node.\n\n"
> 
> +			"resume port tm node (port_id) (node_id)"
> +			"       Resume tm node.\n\n"
> +
>  			"port tm hierarchy commit (port_id)
> (clean_on_fail)\n"
>  			"	Commit tm hierarchy.\n\n"
> 
> @@ -16251,6 +16254,7 @@ cmdline_parse_ctx_t main_ctx[] = {
>  	(cmdline_parse_inst_t *)&cmd_del_port_tm_node,
>  	(cmdline_parse_inst_t *)&cmd_set_port_tm_node_parent,
>  	(cmdline_parse_inst_t *)&cmd_suspend_port_tm_node,
> +	(cmdline_parse_inst_t *)&cmd_resume_port_tm_node,
>  	(cmdline_parse_inst_t *)&cmd_port_tm_hierarchy_commit,
>  	NULL,
>  };
> diff --git a/app/test-pmd/cmdline_tm.c b/app/test-pmd/cmdline_tm.c index
> c9a18dd..807e724 100644
> --- a/app/test-pmd/cmdline_tm.c
> +++ b/app/test-pmd/cmdline_tm.c
> @@ -2036,6 +2036,76 @@ cmdline_parse_inst_t
> cmd_suspend_port_tm_node = {
>  	},
>  };
> 
> +/* *** Resume Port TM Node *** */
> +struct cmd_resume_port_tm_node_result {
> +	cmdline_fixed_string_t resume;
> +	cmdline_fixed_string_t port;
> +	cmdline_fixed_string_t tm;
> +	cmdline_fixed_string_t node;
> +	uint16_t port_id;
> +	uint32_t node_id;
> +};
> +
> +cmdline_parse_token_string_t cmd_resume_port_tm_node_resume =
> +	TOKEN_STRING_INITIALIZER(
> +		struct cmd_resume_port_tm_node_result, resume,
> "resume");
> +cmdline_parse_token_string_t cmd_resume_port_tm_node_port =
> +	TOKEN_STRING_INITIALIZER(
> +		struct cmd_resume_port_tm_node_result, port, "port");
> +cmdline_parse_token_string_t cmd_resume_port_tm_node_tm =
> +	TOKEN_STRING_INITIALIZER(
> +		struct cmd_resume_port_tm_node_result, tm, "tm");
> +cmdline_parse_token_string_t cmd_resume_port_tm_node_node =
> +	TOKEN_STRING_INITIALIZER(
> +		struct cmd_resume_port_tm_node_result, node, "node");
> +cmdline_parse_token_num_t cmd_resume_port_tm_node_port_id =
> +	TOKEN_NUM_INITIALIZER(
> +		struct cmd_resume_port_tm_node_result, port_id, UINT16);
> +cmdline_parse_token_num_t cmd_resume_port_tm_node_node_id =
> +	TOKEN_NUM_INITIALIZER(
> +		struct cmd_resume_port_tm_node_result, node_id, UINT32);
> +
> +static void cmd_resume_port_tm_node_parsed(void *parsed_result,
> +	__attribute__((unused)) struct cmdline *cl,
> +	__attribute__((unused)) void *data)
> +{
> +	struct cmd_resume_port_tm_node_result *res = parsed_result;
> +	struct rte_tm_error error;
> +	uint32_t node_id = res->node_id;
> +	portid_t port_id = res->port_id;
> +	int ret;
> +
> +	if (port_id_is_invalid(port_id, ENABLED_WARN))
> +		return;
> +
> +	/* Port status */
> +	if (!port_is_started(port_id)) {
> +		printf(" Port %u not started (error)\n", port_id);
> +		return;
> +	}

I suggest to remove the CLI layer restriction to check the port status for node suspend/resume.
The device can check at the driver layer whether it is ok suspend/resume node in started/stopped state depending upon the allowed configuration.

Besides this, 
Reviewed-by: Jasvinder Singh <jasvinder.singh at inel.com> 


More information about the dev mailing list