[dpdk-dev] [PATCH v13 1/3] mempool: support external mempool operations

Hunt, David david.hunt at intel.com
Fri Jun 17 11:24:49 CEST 2016



On 17/6/2016 10:09 AM, Thomas Monjalon wrote:
> 2016-06-17 09:42, Hunt, David:
>> On 17/6/2016 9:08 AM, Olivier Matz wrote:
>>> Hi David,
>>>
>>> On 06/17/2016 08:58 AM, Hunt, David wrote:
>>>> A comment below:
>>>>
>>>> On 16/6/2016 1:30 PM, David Hunt wrote:
>>>>> +/**
>>>>> + * Set the ops of a mempool.
>>>>> + *
>>>>> + * This can only be done on a mempool that is not populated, i.e.
>>>>> just after
>>>>> + * a call to rte_mempool_create_empty().
>>>>> + *
>>>>> + * @param mp
>>>>> + *   Pointer to the memory pool.
>>>>> + * @param name
>>>>> + *   Name of the ops structure to use for this mempool.
>>>> + * @param pool_config
>>>> + *   Opaque data that can be used by the ops functions.
>>>>> + * @return
>>>>> + *   - 0: Success; the mempool is now using the requested ops functions.
>>>>> + *   - -EINVAL - Invalid ops struct name provided.
>>>>> + *   - -EEXIST - mempool already has an ops struct assigned.
>>>>> + */
>>>>> +int
>>>>> +rte_mempool_set_ops_byname(struct rte_mempool *mp, const char *name,
>>>>> +        void *pool_config);
>>>>> +
>>> The changes related to the pool_config look good to me.
>>>
>>> If you plan to do a v14 for this API comment, I'm wondering if the
>>> documentation could be slightly modified too. I think "external mempool
>>> manager" was the legacy name for the feature, but maybe it could be
>>> changed in "alternative mempool handlers" or "changing the mempool
>>> handler". I mean the word "external" is probably not appropriate now,
>>> especially if we add other handlers in the mempool lib.
>>>
>>> My 2 cents,
>>> Olivier
>> I had not planned on doing another revision. And I think the term "External
>> Mempool Manager" accurately describes the functionality, so I'd really
>> prefer to leave it as it is.
> I think there is no manager, just a default handler which can be changed.
> I agree the documentation must be fixed.

OK, I have two suggestions to add into the mix.
1. mempool handler framework
or simply
2. mempool handlers. (the alternative is implied). "The mempool handler 
feature", etc.

Thoughts?
David.







More information about the dev mailing list