From: Raymond Mao <raymond.mao@linaro.org>
To: Rob Herring <robh@kernel.org>
Cc: linux-doc@vger.kernel.org,
Krzysztof Kozlowski <krzk+dt@kernel.org>,
Conor Dooley <conor+dt@kernel.org>,
devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
devicetree-spec@vger.kernel.org
Subject: Re: [PATCH] docs: devicetree: overlay-notes: recommend top-level compatible in DTSO
Date: Mon, 14 Jul 2025 10:37:03 -0400 [thread overview]
Message-ID: <CAEfUkUKXskNxdsFG0htpmcgNnVeau7n4wtH2FOweEW7BxC3scA@mail.gmail.com> (raw)
In-Reply-To: <20250710005821.GA94507-robh@kernel.org>
Hi Rob,
On Wed, 9 Jul 2025 at 20:58, Rob Herring <robh@kernel.org> wrote:
>
> +devicetree-spec (because linux-doc doesn't really care)
>
> On Tue, Jun 24, 2025 at 11:13:20AM -0700, Raymond Mao wrote:
> > When managing multiple base device trees and overlays in a structured
> > way (e.g. bundled in firmware or tools), it is helpful to identify the
> > intended target base DT for each overlay, which can be done via a
> > top-level compatible string in the overlay.
>
> I think this should be more general and more specific at the same time.
>
> You might not want to match on a top-level board/soc compatible, but
> rather the compatible for a specific node. For example, you may have an
> overlay for a cape, hat, etc. that applies to a connector node and that
> connector node could be on any number of boards or even multiple
> connectors on 1 board. That's all under development, but so far in those
> cases we expect some sort of connector driver to apply the overlays. But
> I think you could have the same issue of identifying which overlay files
> are relevant. I don't think folks working on add-on boards have thought
> that far ahead.
>
The use case from the angle of a specific driver might be a completely
different scenario.
The motivation of this patch is to support handing over DTOs as
transfer entries across different boot stages via transfer list.
Unlike for a specific driver, a DTO within a transfer entry is usually
board-leveled and contains nodes for more than one driver.
The transfer list library in each boot stage does not have knowledge
on a single specific driver and which target base should be applied is
the only interest.
That is the common DTO use case in Firmware Handoff and why we need
the top-level compatible to be introduced.
For details, please see my PR for the Firmware Handoff spec update to
support DTO as transfer entries at:
https://github.com/FirmwareHandoff/firmware_handoff/pull/74
> And since we don't know the target-path up front, it is just left blank
> so far. It would be better if we expressed *something*. Perhaps
> 'target-compatible'? Something like that would work in your case I
> think.
>
> You'd have to be somewhat crazy, but you can bundle a bunch of
> mutually-exclusive or unrelated overlays within a single overlay file. I
> don't know that we want to prevent doing that. Someone might come up
> with some not crazy reason to do that...
>
> >
> > This patch updates the document with a note and example for this
> > practice.
> >
> > Signed-off-by: Raymond Mao <raymond.mao@linaro.org>
> > ---
> > Documentation/devicetree/overlay-notes.rst | 28 ++++++++++++++++++++++
> > 1 file changed, 28 insertions(+)
> >
> > diff --git a/Documentation/devicetree/overlay-notes.rst b/Documentation/devicetree/overlay-notes.rst
> > index 35e79242af9a..30b142d1b2ee 100644
> > --- a/Documentation/devicetree/overlay-notes.rst
> > +++ b/Documentation/devicetree/overlay-notes.rst
> > @@ -103,6 +103,34 @@ The above bar.dtso example modified to use target path syntax is::
> > ---- bar.dtso --------------------------------------------------------------
> >
> >
> > +Overlay identification
> > +----------------------
> > +
> > +When managing overlays dynamically or bundling multiple base device trees
> > +and overlays in a single system (e.g., in firmware, initramfs, or user-space
> > +tools), it becomes important to associate each overlay with its intended
> > +target base DT.
> > +
> > +To support this, overlays should include the top-level compatible string
> > +from its base DT.
>
> The base has multiple compatible strings, so which one? Has to match on
> any one or all of them?
Match only with base's top-level compatible, see above for the
explanation of the use case.
Regards,
Raymond
>
> > +This enables higher-level software or firmware to identify which base DT
> > +an overlay is compatible with and apply it accordingly.
> > +
> > +Example usage::
> > +
> > + ---- bar.dtso - overlay with top-level compatible string -------------------
> > + /dts-v1/;
> > + /plugin/;
> > + compatible = "corp,foo";
> > +
> > + ...
> > + ---- bar.dtso --------------------------------------------------------------
> > +
> > +This top-level compatible string is not required by the kernel overlay
> > +mechanism itself, but it is strongly recommended for managing overlays in
> > +scalable systems.
> > +
> > +
> > Overlay in-kernel API
> > --------------------------------
> >
> > --
> > 2.25.1
> >
prev parent reply other threads:[~2025-07-14 14:37 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-06-24 18:13 [PATCH] docs: devicetree: overlay-notes: recommend top-level compatible in DTSO Raymond Mao
2025-06-26 16:32 ` Conor Dooley
2025-07-01 15:15 ` Ilias Apalodimas
2025-07-10 0:58 ` Rob Herring
2025-07-14 14:37 ` Raymond Mao [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAEfUkUKXskNxdsFG0htpmcgNnVeau7n4wtH2FOweEW7BxC3scA@mail.gmail.com \
--to=raymond.mao@linaro.org \
--cc=conor+dt@kernel.org \
--cc=devicetree-spec@vger.kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=krzk+dt@kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=robh@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).