[dpdk-dev] [PATCH v5 2/5] ethdev: increase port_id range

Ferruh Yigit ferruh.yigit at intel.com
Wed Sep 27 20:44:42 CEST 2017


On 9/26/2017 8:01 AM, Yang, Zhiyong wrote:
> Hi Ferruh, Thomas,
> 
>>> @@ -40,7 +40,7 @@ LIB = librte_pmd_af_packet.a
>>>
>>>  EXPORT_MAP := rte_pmd_af_packet_version.map
>>>
>>> -LIBABIVER := 1
>>> +LIBABIVER := 2
>>
>> I have just recognized this one. This shouldn't be updated. Only LIABIVER of the
>> libraries that their API modified should be updated.
>>
>> The release notes "Shared Library Versions" updates and Makefile LIBABIVER
>> updates should be compatible. I mean either both needs to be updated or both
>> not updated.
>>
> 
> Ok.  I will cleanup the unnecessary updates. 
> 
> Here are the list I will update .
> 
> -     librte_bitratestats.so.1
> +     librte_bitratestats.so.2
> -     librte_ethdev.so.7
> +     librte_ethdev.so.8
> -     librte_pdump.so.1
> +     librte_pdump.so.2
> -     librte_pmd_bond.so.1
> +     librte_pmd_bnxt.so.2
> +     librte_pmd_bond.so.2
> +     librte_pmd_failsafe.so.2
> +     librte_pmd_i40e.so.2
> +     librte_pmd_ixgbe.so.2
> +     librte_pmd_vhost.so.2
> 
> Among them, the following libs are added newly.
> 
> +     librte_pmd_bnxt.so.2
> +     librte_pmd_failsafe.so.2
> +     librte_pmd_i40e.so.2
> +     librte_pmd_ixgbe.so.2
> +     librte_pmd_vhost.so.2

There is a patch to add missing ones [1], I think it would be better to
add them with a separate patch.

But I missed the "librte_pmd_vhost" one, I will send a new version soon.

Also failsafe don't have any PMD specific API, no need to add it.


[1]
http://dpdk.org/dev/patchwork/patch/28598/


> 
> One  LIBABIVER need to be fixed.
> Could you apply your patch  http://www.dpdk.org/dev/patchwork/patch/28738/  in the main tree firstly?
> 
> I hope the patchset can be merged into main tree.  How do you think about it?
> 
>> There is no way to split this patch more without breaking build right?
> 
> Yes. it is too hard to split it.
> 
> thanks
> Zhiyong
> 
>> It would be possible to do it if port_t used, but that has been decided not to...
>>
>> <...>



More information about the dev mailing list