Re: [dpdk-dev] [RFC PATCH] mbuf: fix to update documentation of PKT_RX_QINQ_STRIPPED

[Andrew Rybchenko <arybchenko@solarflare.com>]

Somnath,

On 12/31/19 5:15 AM, Somnath Kotur wrote:
> Andrew,
>
> On Tue, Dec 24, 2019 at 3:23 PM Andrew Rybchenko
> <arybchenko@solarflare.com> wrote:
>> On 12/24/19 6:16 AM, Somnath Kotur wrote:
>>> Given that we haven't heard any objection from anyone in a while on
>>> this ...can we get this in please?
>> I'm sorry, but have you seen below?
>> It means that PKT_RX_QINQ_STRIPPED, PKT_RX_QINQ, PKT_RX_VLAN
>> and PKT_RX_VLAN_STRIPPED must be clarified.
>>
> OK, not sure I understood what is the next action here? Will you or someone
> from the main tree maintainers be sending out a patch with this clarification?

Please, send non-RCF version of the patch which fixes
PKT_RX_QINQ_STRIPPED and PKT_RX_QINQ description.
PKT_RX_QINQ must not claim that both VLAN headers
have been stripped in the case of PKT_RX_QINQ_STRIPPED.

I think that VLAN should be used instead of "vlan" in description
as well as TCI instead of "tci". Also vlans -> VLANs.

>> It sounds like change of semantics in order to resolve the
>> problem, but anyway it is still a small change of semantics.

May be dropped.

>> BTW, it is better to make summary human readable and avoid
>> PKT_RX_QINQ_STRIPPED (I guess check-git-log.sh yells on it).

Please, don't forget about it as well.

>> Also RFC patch cannot be applied, non-RFC version is required.
>>
>> CC main tree maintainers.
>>
>>> Thanks
>>>
>>> On Mon, Dec 16, 2019 at 2:43 PM Andrew Rybchenko
>>> <arybchenko@solarflare.com> wrote:
>>>> On 12/16/19 11:47 AM, Somnath Kotur wrote:
>>>>> On Mon, Dec 16, 2019 at 12:01 PM Andrew Rybchenko
>>>>> <arybchenko@solarflare.com> wrote:
>>>>>> On 12/16/19 6:16 AM, Somnath Kotur wrote:
>>>>>>> Certain hardware may be able to strip and/or save only the outermost
>>>>>>> VLAN instead of both the VLANs in the mbuf in a QinQ scenario.
>>>>>>> To handle such cases, we could re-interpret setting of just PKT_RX_QINQ_STRIPPED
>>>>>>> to indicate that only the outermost VLAN has been stripped by the hardware and
>>>>>>> saved in mbuf->vlan_tci_outer.
>>>>>>> Only When both PKT_RX_QINQ_STRIPPED and PKT_RX_VLAN_STRIPPED are set, the 2 vlans
>>>>>>> have been stripped by the hardware and their tci are saved in mbuf->vlan_tci (inner)
>>>>>>> and mbuf->vlan_tci_outer (outer).
>>>>>>>
>>>>>>> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
>>>>>>> Signed-off-by: JP Lee <jongpil.lee@broadcom.com>
>>>>>>> ---
>>>>>>>  lib/librte_mbuf/rte_mbuf_core.h | 15 +++++++++++----
>>>>>>>  1 file changed, 11 insertions(+), 4 deletions(-)
>>>>>>>
>>>>>>> diff --git a/lib/librte_mbuf/rte_mbuf_core.h b/lib/librte_mbuf/rte_mbuf_core.h
>>>>>>> index 9a8557d..db1070b 100644
>>>>>>> --- a/lib/librte_mbuf/rte_mbuf_core.h
>>>>>>> +++ b/lib/librte_mbuf/rte_mbuf_core.h
>>>>>>> @@ -124,12 +124,19 @@
>>>>>>>  #define PKT_RX_FDIR_FLX      (1ULL << 14)
>>>>>>>
>>>>>>>  /**
>>>>>>> - * The 2 vlans have been stripped by the hardware and their tci are
>>>>>>> - * saved in mbuf->vlan_tci (inner) and mbuf->vlan_tci_outer (outer).
>>>>>>> + * The outer vlan has been stripped by the hardware and their tci are
>>>>>>> + * saved in mbuf->vlan_tci_outer (outer).
>>>>>>>   * This can only happen if vlan stripping is enabled in the RX
>>>>>>>   * configuration of the PMD.
>>>>>>> - * When PKT_RX_QINQ_STRIPPED is set, the flags (PKT_RX_VLAN |
>>>>>>> - * PKT_RX_VLAN_STRIPPED | PKT_RX_QINQ) must also be set.
>>>>>>> + * When PKT_RX_QINQ_STRIPPED is set, the flags (PKT_RX_VLAN |  PKT_RX_QINQ)
>>>>>>> + * must also be set.
>>>>>>> + * When both PKT_RX_QINQ_STRIPPED and PKT_RX_VLAN_STRIPPED are set, the 2 vlans
>>>>>>> + * have been stripped by the hardware and their tci are saved in
>>>>>>> + * mbuf->vlan_tci (inner) and mbuf->vlan_tci_outer (outer).
>>>>>>> + * This can only happen if vlan stripping is enabled in the RX configuration
>>>>>>> + * of the PMD.
>>>>>>> + * When PKT_RX_QINQ_STRIPPED and PKT_RX_VLAN_STRIPPED are set,
>>>>>>> + * (PKT_RX_VLAN | PKT_RX_QINQ) must also be set.
>>>>>>>   */
>>>>>>>  #define PKT_RX_QINQ_STRIPPED (1ULL << 15)
>>>>>>>
>>>>>> I always thought that PKT_RX_VLAN_STRIPPED means *one* VLAN
>>>>>> stripped regardless if it is outer (if the packet is double
>>>>>> tagged) or inner (if only one VLAN tag was present).
>>>>>>
>>>>>> That's why PKT_RX_QINQ_STRIPPED description says that *two*
>>>>>> VLANs have been stripped.
>>>>>>
>>>>>> What is the problem with such approach?
>>>>> The problem is that RX_VLAN_STRIPPED implies that the stripped VLAN
>>>>> (outer or inner) is saved in mbuf->vlan_tci, correct?
>>>> Yes.
>>>>
>>>>> There is no way to convey that it is in QinQ mode and yet only outer
>>>>> VLAN has been stripped and saved in mbuf->vlan_tci_outer ?
>>>> Ah, it looks like I understand now that the problem is in
>>>> PKT_RX_QINQ description which claims that TCI is saved in
>>>> mbuf->vlan_tci_outer and PKT_RX_QINQ_STRIPPED means that
>>>> both VLAN tags are stripped regardless (PKT_RX_VLAN_STRIPPED).
>>>> Moreover PKT_RX_QINQ_STRIPPED requires PKT_RX_VLAN_STRIPPED.
>>>>
>>>> It means that PKT_RX_QINQ_STRIPPED, PKT_RX_QINQ, PKT_RX_VLAN
>>>> and PKT_RX_VLAN_STRIPPED must be clarified.
>>>>
>>>> I'm not sure, but it looks like it could affect net/dpaa2,
>>>> so I'm including driver maintainers in CC.