From mboxrd@z Thu Jan  1 00:00:00 1970
From: Olivier MATZ <olivier.matz@6wind.com>
Subject: Re: [dpdk-dev,v5,1/3] mempool: support external handler
Date: Tue, 31 May 2016 22:40:59 +0200
Message-ID: <574DF6DB.4050905@6wind.com>
References: <1463665501-18325-2-git-send-email-david.hunt@intel.com>
 <20160523143511.7d30699b@pcviktorin.fit.vutbr.cz>
 <574D54D6.1080409@intel.com>
 <20160531140652.018a03de@pcviktorin.fit.vutbr.cz>
 <574D95E9.4020504@intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Cc: dev@dpdk.org, yuanhan.liu@linux.intel.com, pmatilai@redhat.com,
 jerin.jacob@caviumnetworks.com
To: "Hunt, David" <david.hunt@intel.com>,
 Jan Viktorin <viktorin@rehivetech.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mail-wm0-f53.google.com (mail-wm0-f53.google.com [74.125.82.53])
 by dpdk.org (Postfix) with ESMTP id F274838EB
 for <dev@dpdk.org>; Tue, 31 May 2016 22:41:00 +0200 (CEST)
Received: by mail-wm0-f53.google.com with SMTP id n184so868908wmn.1
 for <dev@dpdk.org>; Tue, 31 May 2016 13:41:00 -0700 (PDT)
In-Reply-To: <574D95E9.4020504@intel.com>
List-Id: patches and discussions about DPDK <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>

Hi,

On 05/31/2016 03:47 PM, Hunt, David wrote:
> On 5/31/2016 1:06 PM, Jan Viktorin wrote:
>> On Tue, 31 May 2016 10:09:42 +0100
>> "Hunt, David" <david.hunt@intel.com> wrote:
>>
>>> The *p pointer is the opaque data for a given mempool handler (ring,
>>> array, linked list, etc)
>> Again, doc comments...
>>
>> I don't like the obj_table representation to be an array of void *. I
>> could see
>> it already in DPDK for defining Ethernet driver queues, so, it's
>> probably not
>> an issue. I just say, I would prefer some basic type safety like
>>
>> struct rte_mempool_obj {
>>     void *p;
>> };
>>
>> Is there somebody with different opinions?
>>
>> [...]
> 
> Comments added. I've left as a void* for the moment.

Jan, could you please detail why you think having a
rte_mempool_obj structure brings more safety?

For now, I'm in favor of keeping the array of void *, because
that's what we use in other mempool or ring functions.


>>>>> +/** Structure defining a mempool handler. */
>>>> Later in the text, I suggested to rename rte_mempool_handler to
>>>> rte_mempool_ops.
>>>> I believe that it explains the purpose of this struct better. It
>>>> would improve
>>>> consistency in function names (the *_ext_* mark is very strange and
>>>> inconsistent).
>>> I agree. I've gone through all the code and renamed to
>>> rte_mempool_handler_ops.
>> Ok. I meant rte_mempool_ops because I find the word "handler" to be
>> redundant.
> 
> I prefer the use of the word handler, unless others also have opinions
> either way?

Well, I think rte_mempool_ops is clear enough, and shorter,
so I'd vote for it.


Regards,
Olivier