[dpdk-dev] [PATCH 5/5] net/szedata2: add more supported firmwares

Ferruh Yigit ferruh.yigit at intel.com
Wed May 31 11:11:40 CEST 2017


On 5/31/2017 9:34 AM, Matej Vido wrote:
> On 29.05.2017 14:42, Ferruh Yigit wrote:
>> On 5/26/2017 7:05 PM, Matej Vido wrote:
>>> Add IBUF and OBUF offsets definitions for new firmwares.
>> just to double check, is it required to add new pci device ids to
>> support these new cards?
>>
>> Is dpdk supported nics (http://dpdk.org/doc/nics) needs to be updated?
>>
>> Can you please update release notes to announce new cards support?
> Hi Ferruh,
> 
> This patch set adds mainly the support of new firmwares for the cards 
> that are already supported.
> But you are right, there is also one new card which would need to add 
> new pci device id. We have to check if we want to add the card in this 
> patch set and I will update v2 accordingly.
> 
>>
>>> Signed-off-by: Matej Vido <vido at cesnet.cz>
>> <...>
>>
>>> +#elif RTE_LIBRTE_PMD_SZEDATA2_AS == 5
>>> +
>>> +const uint32_t szedata2_ibuf_base_table[] = {
>>> +	0x8000,
>>> +	0x8200,
>>> +	0x8400,
>>> +	0x8600,
>>> +	0x8800
>>> +};
>>> +const uint32_t szedata2_obuf_base_table[] = {
>>> +	0x9000,
>>> +	0x9200,
>>> +	0x9400,
>>> +	0x9600,
>>> +	0x9800
>>> +};
>> AS = 5 is not documented in both documentation and source code comment,
>> can you please add this one too?
> 5 is for future use and experimental firmwares. Would it make sense to 
> write this note in the comments but not in the documentation to avoid 
> confusion or should it be mentioned in the documentation too?

I think it is good to mention in documentation, since this will be part
of released code, but this can be brief.

> 
> Thanks,
> Matej
>>> +
>>>   #else
>>>   #error "RTE_LIBRTE_PMD_SZEDATA2_AS has wrong value, see comments in config file"
>>>   #endif
>>>
> 



More information about the dev mailing list