Re: New 'make dtbs_check W=1' warnings

[Bjorn Andersson]

On Mon 12 Apr 08:14 CDT 2021, Arnd Bergmann wrote:

> On Mon, Apr 12, 2021 at 1:32 PM Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
> > On Thu, Apr 8, 2021 at 5:08 PM Arnd Bergmann <arnd@xxxxxxxxxx> wrote:
> 
> > >
> > > For this merge window, I don't think any of them are show-stoppers (Rob, let me
> > > know if you disagree), but in the long run we may want to gradually enforce
> > > a rule about not merging changes that introduce any new warnings, in order to
> > > have a chance of cleaning up the existing ones.
> >
> > This may not be as simple as it sounds, as DT binding updates typically
> > follow a different path than DTS(i) updates.  DT bindings updates may be
> > picked up by a subsystem maintainer, by Rob, or by the platform
> > maintainer.
> 
> I checked out the bindings from linux-next, which seems to have covered
> most of these. Sometimes it pays off to be lazy and merge them after
> everyone else.
> 
> > For trivial updates (e.g. adding a compatible value, and sometimes
> > extending a limit), a DTS(i) update may be accepted by the platform
> > maintainer before the corresponding DT binding update.  The latter may
> > even be merged one or more kernel revisions later, especially when
> > involving subsystems that are not traditionally rooted into platforms
> > using DT.
> >
> > Of course we could mention any expected warning regressions in our pull
> > requests for soc.
> 
> Right, that would certainly help. Some maintainers send all binding
> updates both to the driver maintainers and to the soc tree, along
> with the other changes that only go into one tree. That is of course
> also more work on your side, but it solves the problem entirely.
> 

So the same binding patch is picked up both in the driver and soc tree?
I was expecting that to cause (harmless) conflicts when things arrive in
Linus' merge queue?

Or are you saying people go the length to create immutable branches for
each binding?


I'm curious because it's fairly often that we introduce clocks,
interconnects etc where the macros from the dt bindings includes aren't
available for another release (so we use numerical constants and then go
back and fix them up later).

Regards,
Bjorn

> > > renesas/r8a774a1-beacon-rzg2m-kit.dt.yaml: csi2@feaa0000: ports:
> > > 'port@0' is a required property
> >
> > [...]
> >
> > I've replied to these as a response to your PR reply, see
> > https://lore.kernel.org/linux-renesas-soc/CAMuHMdWHLnXgBSjP7VKUdx-YNr9rSKFkE5Ge5q_tarU6HP9Lhw@xxxxxxxxxxxxxx/
> 
> Ok, thanks.
> 
>       Arnd