On 2020/8/6 23:25, Ferruh Yigit wrote:
> On 8/6/2020 5:00 AM, Chengchang Tang wrote:
>> Struct rte_eth_rxq_info will be modified to include a new field, indicating
>> the size of each buffer that could be used for hw to receive packets. Add
>> this field to rte_eth_rxq_info to expose relevant information to upper
>> layer users/application.
>>
>> For more details:
>> https://mails.dpdk.org/archives/dev/2020-July/176135.html
>>
>> Signed-off-by: Chengchang Tang <tangchengch...@huawei.com>
>> Acked-by: Andrew Rybchenko <arybche...@solarflare.com>
>> ---
>> doc/guides/rel_notes/deprecation.rst | 9 +++++++++
>> 1 file changed, 9 insertions(+)
>>
>> diff --git a/doc/guides/rel_notes/deprecation.rst
>> b/doc/guides/rel_notes/deprecation.rst
>> index ea4cfa7..f08b5f9 100644
>> --- a/doc/guides/rel_notes/deprecation.rst
>> +++ b/doc/guides/rel_notes/deprecation.rst
>> @@ -110,6 +110,15 @@ Deprecation Notices
>> break the ABI checks, that is why change is planned for 20.11.
>> The list of internal APIs are mainly ones listed in
>> ``rte_ethdev_driver.h``.
>>
>> +* ethdev: A new field will be added to the public data structure
>> + ``rte_eth_rxq_info`` to indicate the buffer size used in receiving packets
>> + for HW. When receive packets, HW DMA won't exceed this size.
>
> Overall +1 to provide this information.
>
> This field is only to report back the PMD configured Rx buffer size, it won't
> affect the configuration step, do you think should we highlight this?
I think it is not necessary because this structure is designed to report PMD
configuration. And it is only used in rte_eth_rx_queue_info_get.
>
> Also will this field be optional or mandatory, this matters for the scope of
> the
> work for 20.11. I think the intention is to provide an optional field, what do
> you think to documenting that it is optional?
Yes, it is optional. I will highlight this in v3.
>
>> And it will
>> + affect the number of fragments in receiving packets when scatter is
>> enabled.
>
> Is this detail required in the deprecation notice? I see it is relevant but
> the configured Rx buffer size affects the number of the fragments, but
> reporting
> this value does not.
> Do you want to mention above as motivation to have the field? If so I don't
> expect application to calculate the number of the fragments using this value.
> I am for dropping above sentences if I am not missing anything.
Thank you for this advice. I am not sure what information should be reflected in
a deprecation notice. I seem to have added some redundant and inappropriate
stuff.
I will drop these sentences in v3.
>
>> + So, add this field to ``rte_eth_rxq_info`` to expose relevant information
>> to
>> + upper layer user/application.
>
> And not sure above sentences says anything new, looks like duplication to me.
>
>> + This change is planned for 20.11. For more details:
>> + https://mails.dpdk.org/archives/dev/2020-July/176135.html
>> +
>> * traffic manager: All traffic manager API's in ``rte_tm.h`` were
>> mistakenly made
>> ABI stable in the v19.11 release. The TM maintainer and other
>> contributors have
>> agreed to keep the TM APIs as experimental in expectation of additional
>> spec
>>
>
>
> .
>
