[PATCH v3 0/5] AHCI and SATA PHY support for Broadcom STB SoCs

[Brian Norris]

I can explain part of this, but I'm curious if anyone else has different
info.

On Thu, May 21, 2015 at 06:23:50PM -0400, Tejun Heo wrote:
> But the rules have never been clear to me.  If the subsystem
> maintainer is okay with it, I'm happy to take the patches.  I'm just
> kinda curious why this doesn't go through devicetree tree while some
> other devicetree patches go through there.

AFAIK, there is no official tree for device tree bindings. There's just
a mailing list and several reviewers, who usually try to help on the big
picture binding review. Note that there's no tree listed in MAINTAINERS
under:

  OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS

But you will see several MAINTAINERS entries for different subdirs of
Documentation/devicetree/bindings/. Maybe you should add one for .../ata
if you're going to continue taking patches?

It's possible you're confusing binding documentation with .dts source
files? The DTS files (arch/*/boot/dts/) go through arch trees. For
instance, the arm-soc maintainers have a structured process by which
sub-architecture maintainers track .dts(i) file updates for their
boards/chips and filter them up to Arnd, Olof, etc., via their separate
'dts' branches. That's why Florian took patch 5 to his tree.

> Can somebody explain the
> overall policy to me?  I'm not looking for some absolute rules and
> exceptions are fine but I do wanna have a general direction.

Brian