Re: [PATCH 1/3] mbuf:add two TX offload flags and change three fields

[Olivier MATZ <olivier.matz-pdR9zngts4EAvxtiuMwx3w@public.gmane.org>]

Hi Jijiang,

On 11/27/2014 02:14 PM, Liu, Jijiang wrote:
>>> diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib/librte_mbuf/rte_mbuf.h
>>> index 367fc56..48cd8e1 100644
>>> --- a/lib/librte_mbuf/rte_mbuf.h
>>> +++ b/lib/librte_mbuf/rte_mbuf.h
>>> @@ -99,10 +99,9 @@ extern "C" {
>>>   #define PKT_RX_TUNNEL_IPV6_HDR (1ULL << 12) /**< RX tunnel packet with
>> IPv6 header. */
>>>   #define PKT_RX_FDIR_ID       (1ULL << 13) /**< FD id reported if FDIR match.
>> */
>>>   #define PKT_RX_FDIR_FLX      (1ULL << 14) /**< Flexible bytes reported if FDIR
>> match. */
>>> -/* add new RX flags here */
>>>
>>
>> We should probably not remove this line.
> 
> Why?
> There are two lines "/* add new RX flags here */" in rte_mbuf.h file.

No, one is RX, the other is TX.


>>> +/** Tell the NIC it's an outer IPv6 packet for tunneling packet.*/
>>> +#define PKT_TX_OUTER_IPV6    (1ULL << 59)
>>>
>>
>> This flag is not in the cover letter or commit log. What is its purpose?
>>
> With FVL, if outer L3 header is IPv6, to make HW TX checksum offload work , SW must be responsible to tell hardware this information.


What hardware checksum are you talking about?
I understand that outer L4 checksum is not supported from one of
Konstantin's mail.
And there is no L3 checksum in IPv6.


Regards,
Olivier