Re: [PATCH wireless-next 14/35] wifi: mm81x: add mac.c

[Lachlan Hodges]

> > We expect to see some larger features - including monitor mode, and
> > mesh in the near to mid-term future within the driver itself, but the
> > core development will still remain in mac80211 & cfg80211 as we
> > extend the S1G implementation.
> 
> Sounds like there would be quite some co-development with cfg/mac and
> the driver

Thinking about this a bit further, the initial items on the todo
list are more or less as follows:

1. hostapd support (mostly unrelated, some minor tweaks to hwsim)
2. Rx reporting - mainly mac and cfg with some minor driver changes
   + usermode aswell
3. iw and iwinfo support (again, unrelated)
4. extend regulatory support - this will be a big challenge and once
   again mostly cfg and mac

So most driver work would just be plumbing through any required
changes for the above, no large features in the near future.

> which is probably simpler if I apply driver patches too,
> otherwise you have to synchronise pull requests to when I apply patches
> to my tree? OTOH, to pass the bot checks you already have to do that
> anyway, unless sending cfg/mac/driver patches in one series, which also
> isn't great since it ends to bury the cfg/mac patches.

Initially that works with us if you apply the patches :). Obviously,
once supports get a bit more mature and the work is deocupled the goal
would be to send you pull requests. Let's go with patches to you now,
and if that works out to not be ideal we can move to pull requests

> > As for workflows, we are still figuring that out for ourselves.
> 
> I'd prefer if you could create an account on patchwork.kernel.org, and
> then I can automatically delegate patches to this driver to you. Whether
> or not you then re-assign them to me in patchwork or collect them and
> send a pull request is somewhat secondary, but the latter obviously
> makes things a bit simpler for me. If you _are_ going to do that longer
> term than just the initial driver, probably should document a T: entry
> in the maintainers file too.

I have made an account with the user "lhodges" on patchwork - no
issues with that. We haven't got a public tree tracking wireless-next
yet. Nothing necessarily blocking that except internal processes, I
assume there's no issue adding it later?

lachlan