Re: [PATCH v2 iproute2-next 00/10] Add tc-mqprio and tc-taprio support for preemptible traffic classes

[David Ahern <dsahern@kernel.org>]

On 4/25/23 6:55 AM, Vladimir Oltean wrote:
> On Mon, Apr 24, 2023 at 07:47:31PM -0600, David Ahern wrote:
>> On 4/22/23 10:59 AM, Vladimir Oltean wrote:
>>> Unless there are changes I need to make to the contents of the patches,
>>> could you take these from the lists, or is that a no-no?
>>
>> iproute2 follows the netdev dev model with a main tree for bug fixes and
>> -next tree for features. In the future please separate out the patches
>> and send with proper targets. If a merge is needed you can state that in
>> the cover letter of the set for -next.
> 
> I know that the trees are split and it is no coincidence that my patches
> were sorted in the correct order. I've been working for 10 months on
> this small feature and I was impatient to get it over with, so I wanted
> to eliminate one round-trip time if possible (send to "iproute2", ask
> for merge, send to "iproute2-next"). I requested this honestly thinking
> that there would be no difference to the end result, only less pretentious
> in terms of the process. If there is any automation (I didn't see any in
> Patchwork at least) or any other reason that would justify the more
> pretentious process, then again, my excuses, I plead ignorance and I
> will follow it more strictly next time, but I'd also like to know it :)

Maybe the word choice here is a language issue, but it is not a
'pretentious' process, it is "the" process for submitting patches to
both networking trees and iproute2 trees. You would not send a mixed
patch set to the netdev maintainers, so don't do it for iproute2.