[PATCH v3] ethdev: deprecate header fields and metadata flow actions

[Ferruh Yigit]

On 11/25/2021 1:56 PM, Ori Kam wrote:
> 
> 
>> -----Original Message-----
>> From: Ferruh Yigit <ferruh.yigit at intel.com>
>> Sent: Thursday, November 25, 2021 2:47 PM
>> To: Ori Kam <orika at nvidia.com>; Slava Ovsiienko <viacheslavo at nvidia.com>
>> Cc: NBU-Contact-Thomas Monjalon (EXTERNAL) <thomas at monjalon.net>; David Marchand
>> <david.marchand at redhat.com>; Andrew Rybchenko <andrew.rybchenko at oktetlabs.ru>;
>> dev at dpdk.org; Ray Kinsella <mdr at ashroe.eu>
>> Subject: Re: [PATCH v3] ethdev: deprecate header fields and metadata flow actions
>>
>> On 11/25/2021 12:29 PM, Ori Kam wrote:
>>> Hi
>>>
>>>> -----Original Message-----
>>>> From: Ferruh Yigit <ferruh.yigit at intel.com>
>>>> Sent: Thursday, November 25, 2021 2:19 PM
>>>> Subject: Re: [PATCH v3] ethdev: deprecate header fields and metadata flow actions
>>>>
>>>> On 11/24/2021 3:37 PM, Viacheslav Ovsiienko wrote:
>>>>> @@ -3331,6 +3430,9 @@ struct rte_flow_action_set_tag {
>>>>>     };
>>>>>
>>>>>     /**
>>>>> + * @deprecated
>>>>> + * @see RTE_FLOW_ACTION_TYPE_MODIFY_FIELD
>>>>> + *
>>>>>      * @warning
>>>>>      * @b EXPERIMENTAL: this structure may change without prior notice
>>>>>      *
>>>>> @@ -3355,6 +3457,9 @@ struct rte_flow_action_set_meta {
>>>>>     };
>>>>>
>>>>>     /**
>>>>> + * @deprecated
>>>>> + * @see RTE_FLOW_ACTION_TYPE_MODIFY_FIELD
>>>>> + *
>>>>>      * RTE_FLOW_ACTION_TYPE_SET_IPV4_DSCP
>>>>>      * RTE_FLOW_ACTION_TYPE_SET_IPV6_DSCP
>>>>>      *
>>>>
>>>> Hi Viacheslav, Ori,
>>>>
>>>> Just recognized that some of the "struct rte_flow_*" added as experimental
>>>> (experimental note in the struct comment), some without experimental note.
>>>>
>>>> Is it because some forgotten the add the experimental note?
>>>>
>>>> What is the rule to add new "struct rte_flow_*"?
>>>>
>>>
>>> In the beginning struct were never experimental,
>>> and in the last few releases we started to put experimental also on structs
>>> which I think is much better since at the end this is a new API and just like any other
>>> new API it is not stable.
>>>
>>
>> Got it, thanks.
>>
>> So to record for archives, new "struct rte_flow_*" structs should be
>> experimental by default.
>>
>> Does it applies to enums too?
> 
> Yes, I think that anything new is experimental don't you?
> 

I just want to record some guidelines that we can reference later.

So we all can check that new struct & enums are marked as experimental
in rte_flow.h after this point.