From mboxrd@z Thu Jan  1 00:00:00 1970
From: Remy Horton <remy.horton@intel.com>
Subject: Re: [PATCH v6 1/4] lib: add information metrics library
Date: Thu, 12 Jan 2017 15:30:06 +0000
Message-ID: <d68dcbe6-296f-aaed-3514-76a321db7c20@intel.com>
References: <1484150594-3758-1-git-send-email-remy.horton@intel.com>
 <1484150594-3758-2-git-send-email-remy.horton@intel.com>
 <1951093.Zqtj2Qm3TA@xps13>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Cc: dev@dpdk.org
To: Thomas Monjalon <thomas.monjalon@6wind.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mga14.intel.com (mga14.intel.com [192.55.52.115])
 by dpdk.org (Postfix) with ESMTP id BC7CF5424
 for <dev@dpdk.org>; Thu, 12 Jan 2017 16:30:08 +0100 (CET)
In-Reply-To: <1951093.Zqtj2Qm3TA@xps13>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>


On 12/01/2017 13:22, Thomas Monjalon wrote:
> 2017-01-12 00:03, Remy Horton:
[..]
>> --- /dev/null
>> +++ b/lib/librte_metrics/rte_metrics.h
>> +/** Used to indicate port-independent information */
>> +#define RTE_METRICS_NONPORT -1
>
> I do not understand this constant.
> Why using the word "port" to name any device?
> What means independent?

The metric library hold two sets of values: Ones specific to ports, and 
ones that are global/aggregate. Agree "non port" is not the clearest 
choice of name..


>> +/**
>> + * Metric name
>> + */
>> +struct rte_metric_name {
>> +	/** String describing metric */
>> +	char name[RTE_METRICS_MAX_NAME_LEN];
>> +};
>
> Why a struct for a simple string?

It is modelled after xstats, which does the same thing. In this case 
thinking was that using:

rte_metrics_get_names(struct rte_metric_name *names,
	uint16_t capacity)

would be tidier than (e.g.):

rte_metrics_get_names(char *names[RTE_METRICS_MAX_NAME_LEN],
	uint16_t capacity)


>> + */
>> +struct rte_metric_value {
>> +	/** Numeric identifier of metric */
>> +	uint16_t key;
>
> How the key is bound to the name?

Corresponds to array position in struct rte_metric_name. Will amend 
doxygen comments.


> Remember how the xstats comments were improved:
> 	http://dpdk.org/commit/6d52d1d

Will take account for them. I wrote this code back in October, so it 
slipped my mind to apply the same comments to this module.