[GIT PULL] arm64: Second batch of DTS changes for X-Gene platforms queued for v4.4

[Olof Johansson]

Hi,

On Fri, Oct 30, 2015 at 7:15 PM, Duc Dang <dhdang at apm.com> wrote:
> Hi Olof,
>
> On Friday, October 30, 2015, Olof Johansson <olof at lixom.net> wrote:
>>
>> Hi Duc,
>>
>> On Fri, Oct 30, 2015 at 01:35:54PM -0700, Duc Dang wrote:
>> > On Wed, Oct 28, 2015 at 12:08 PM, Duc Dang <dhdang at apm.com> wrote:
>> > > Hi Arnd, Olof,
>> > >
>> > > The following changes since commit
>> > > d2d0a722df735de52015f0c8a01f27a9ec28622e:
>> > >
>> > >   arm64: dts: Add poweroff device nodes for APM X-Gene 2 platform
>> > > (2015-10-12 14:18:38 -0700)
>> > >
>> > > are available in the git repository at:
>> > >
>> > >   https://github.com/AppliedMicro/xgene-next.git
>> > > tags/xgene-dts-for-v4.4-2
>> > >
>> > > for you to fetch changes up to
>> > > fa535e2632f275f0d5d4080d60c8e1331ac37549:
>> > >
>> > >   arm64: dts: Add L2 cache topology for APM X-Gene SoC (2015-10-26
>> > > 02:31:43 -0700)
>> > >
>> > > ----------------------------------------------------------------
>> > > Second batch of DTS changes for X-Gene platforms queued for v4.4
>> > >
>> > > This patch set adds DTS entries to support various IPs
>> > > for X-Gene v1 and X-Gene v2 SoC:
>> > > - X-Gene v1: Enable support for MMC, USB, GPIO controllers,
>> > >   I2C controller, L2 Cache topology
>> > > - X-Gene v2: Enable support for MMC, USB, GPIO controller,
>> > >   I2C controller (with RTC), PCIe controller with GICv2m MSI,
>> > >   EDAC, L2 Cache topology, TRNG
>> > >
>> > > ----------------------------------------------------------------
>> > > Duc Dang (16):
>> > >       arm64: dts: Add the arasan mmc DTS entries for APm X-Gene v2 SoC
>> > >       arm64: dts: X-Gene: Do not reset or enable/disable clock for AHB
>> > > block
>> > >       arm64: dts: Add RNG device tree nodes for APM X-Gene v2 platform
>> > >       arm64: dts: Add v2m MSI frame nodes for APM X-Gene v2 platforms
>> > >       arm64: dts: Add PCIe node for APM X-Gene v2 platforms
>> > >       arm64: dts: Add USB nodes for APM X-Gene v1 platforms
>> > >       arm64: dts: Add USB nodes for APM X-Gene v2 platforms
>> > >       arm64: dts: Add APM X-Gene v1 SoC GFC GPIO controller DTS
>> > > entries
>> > >       arm64: dts: Add APM X-Gene v2 SoC GFC GPIO controller DTS entry
>> > >       arm64: dts: Add Designware GPIO dts binding for APM X-Gene v1
>> > > platform
>> > >       arm64: dts: Add APM X-Gene v2 SoC Designware GPIO controller DTS
>> > > entry
>> > >       arm64: dts: Add APM X-Gene v2 SoC EDAC DTS entries
>> > >       arm64: dts: Add Designware I2C controller DTS entries for X-Gene
>> > > v1 SoC
>> > >       arm64: dts: Add Designware I2C controller DTS entries for X-Gene
>> > > v2 SoC platform
>> > >       arm64: dts: Add RTC DTS entry for X-Gene v2 SoC platform
>> > >       arm64: dts: Add L2 cache topology for APM X-Gene SoC
>> > >
>> > > Suman Tripathi (1):
>> > >       arm64: dts: Add the arasan mmc DTS entries for APM X-Gene v1 SoC
>> > >
>> > >  arch/arm64/boot/dts/apm/apm-merlin.dts     |  12 +
>> > >  arch/arm64/boot/dts/apm/apm-mustang.dts    |   4 +
>> > >  arch/arm64/boot/dts/apm/apm-shadowcat.dtsi | 401
>> > > +++++++++++++++++++++++++++++
>> > >  arch/arm64/boot/dts/apm/apm-storm.dtsi     | 126 +++++++++
>> > >  4 files changed, 543 insertions(+)
>> >
>> > Hi Arnd, Olof,
>> >
>> > I am sorry that I sent this pull request late. But do you still have
>> > some cycle to have this change set pulled into your xgene/dt in
>> > next/dt for 4.4?
>>
>> This is indeed showing up quite late, since the merge window opens in just
>> a couple of days.
>>
>> Can I please ask that you send us a new pull request for this once -rc1 is
>> out,
>> and we'll queue it up for 4.5?
>
> Yes, I realized the merge window start this Sunday too late.
>
> Anyway, do you maintain a general rule like when You expect people send you
> a pull request for the next merge window? For example: the request must be
> before rc6?

The sooner the better. Ideally most code should show up sometime
between rc4 and rc6. After that we will usually take smaller
incremental branches that adds a few patches on top of previous
material.

If you have cross-dependencies with other maintainer trees, for
example due to shared introduced header files, then please try to get
it in early in case we need to coordinate something.

The community as a whole has a self-imposed limit that most material
should be in -next by -rc6, but it's not strictly enforced and is seen
as a guideline.

I think the ARM64 arch maintainers try to get most material in by
-rc4, but they need slightly more bake/test time on their code so
that's understandable.


Arnd gave a presentation about how we run the maintainer tree at this
year's ELC, and I gave one two years ago more around how we organize
our tree. The videos and slides are online and gives a bit of an
overview if you're interested in learning more.


-Olof