On Sun, Nov 27, 2016 at 12:38:21PM -0800, Olof Johansson wrote:
> On Wed, Nov 16, 2016 at 10:19 AM, Benson Leung <ble...@chromium.org> wrote:
> > I'll be taking over maintainership of platform/chrome from Olof,
> > so let's add me to the list of maintainers.
> > Signed-off-by: Benson Leung <ble...@chromium.org>
> Acked-by: Olof Johansson <o...@lixom.net>
> I talked to Darren at Plumbers about drivers/platform/chrome, and
> whether he was open to merge that contents through his
> drivers/platform tree, and he was open to that.
+Andy Shevchenko, we co-maintain platform-drivers-x86
> Darren, meet Benson -- Benson will be picking up maintainership of the
> chrome platform code since I no longer have good access to hardware to
> test new changes. We didn't talk about the method of merging, if you
> prefer to get patches sent directly to you or if you're OK with
> merging a git branch from Benson. The latter might be easier for
> Benson, the former might be easier for you.
It depends a bit on what Benson is looking for from us. Are you looking for
close code review and inspection? I would assume not, that drivers/chrome is
something you expect to entirely own and you would like to roll it up to Linus
through us because it is small enough that it doesn't make sense to roll it up
independently? If that's accurate, then:
We don't currently have anything that we manage via pull requests, however
platform/chrome is sufficiently independent from the rest of our tree that I'm
open to using pull requests.
Benson, what would you prefer?
Olof, any suggestions on working with pull requests and rolling them up to
- Do you ask your sub-maintainers to have their code in -next before a pull
request to you? Assuming not.
In general, please use signed tags:
Plan to have the bulk of your changes sent to us in a pull request a week or two
before the merge window opens so they can bake a bit in our for-next branch.
>From there, there is a rapid taper off period from rc1 to rc4 or so in terms of
the types of changes, by rc4 and on we would expect only serious bug fixes.
> In the past, the amount of code coming in through the platform/chrome
> tree has been relatively small, it's mostly been a matter of acking
> drivers going in through MFD or other trees. The main reason to split
> them out to a separate platform directory has been because it cuts
> across architectures and none of the previous ones were a good fit
> (i.e. drivers/platform/x86 was too narrow). I expect that to stay true
> as long as there are ARM Chromebooks, which is hopefully forever. :)
Yes, that makes sense.
Intel Open Source Technology Center