[PATCH v3 0/7] dt-bindings: mfd: syscon: Document more compatibles and require simpe-mfd description

[Krzysztof Kozlowski]

On 27/06/2024 10:18, Lee Jones wrote:
> On Wed, 26 Jun 2024, Krzysztof Kozlowski wrote:
> 
>> Hi,
>>
>> Dependency
>> ==========
>> Rebased on Lee's MFD tree, because dependency is there already:
>> https://lore.kernel.org/all/171828959006.2643902.8308227314531523435.b4-ty@kernel.org/
>>
>> Merging
>> =======
>> Preferrably everything via MFD tree (file/context dependencies).
>>
>> Changes in v3
>> =============
>> - Add tags
>> - intel,lgm-syscon: change maintainers (email bounce)
>> - syscon/Split: drop unneeded |, use const instead of enum in select:
>> - Link to v2: https://lore.kernel.org/r/20240616-dt-bindings-mfd-syscon-split-v2-0-571b5850174a@linaro.org
>>
>> Changes in v2
>> =============
>> - Add acks
>> - lgm-syscon: add ranges to binding and example
>> - syscon.yaml: add big select with all compatibles for older dtschema
>> - Link to v1: https://lore.kernel.org/r/20240519-dt-bindings-mfd-syscon-split-v1-0-aaf996e2313a@linaro.org
>>
>> Description/problem
>> ===================
>> Simple syscon nodes can be documented in common syscon.yaml, however
>> devices with simple-mfd compatible, thus some children, should have
>> their own schema listing these children.  Such listing makes the binding
>> specific, allows better validation (so the incorrect child would not
>> appear in the simple-mfd node) and actually enforces repeated rule for
>> simple-mfd devices:
>>
>>   "simple-mfd" is only for simple devices, where the children do not
>>   depend on the parent.
>>
>> Currently the syscon+simple-mfd binding is quite broad and allows
>> any child or property, thus above rule cannot be enforced.
>>
>> Solution
>> ========
>> 1. Split the syscon.yaml binding into common syscon properties, used
>>    potentially by many bindings, and only simple syscon devices (NO
>>    simple-mfd!).
>> 2. Move some known simple-mfd bindings from syscon.yaml to dedicated
>>    files.
>>
>> This patchset might introduce new dtbs_check warnings for devices having
>> simple-mfd and being part of syscon.yaml previously. I fixed some of
>> them, but probably not all.
>>
>> Best regards,
>> Krzysztof
>>
>> To: Lee Jones <lee at kernel.org>
>> To: Rob Herring <robh at kernel.org>
>> To: Krzysztof Kozlowski <krzk+dt at kernel.org>
>> To: Conor Dooley <conor+dt at kernel.org>
>> To: Lars Povlsen <lars.povlsen at microchip.com>
>> To: Steen Hegelund <Steen.Hegelund at microchip.com>
>> To: Daniel Machon <daniel.machon at microchip.com>
>> To: UNGLinuxDriver at microchip.com
>> To: Nishanth Menon <nm at ti.com>
>> To: Matthias Brugger <matthias.bgg at gmail.com>
>> To: AngeloGioacchino Del Regno <angelogioacchino.delregno at collabora.com>
>> Cc: devicetree at vger.kernel.org
>> Cc: linux-kernel at vger.kernel.org
>> Cc: linux-arm-kernel at lists.infradead.org
>> Cc: linux-mediatek at lists.infradead.org
>>
>> ---
>> Krzysztof Kozlowski (7):
>>       dt-bindings: mfd: syscon: Drop hwlocks
>>       dt-bindings: soc: sprd: sc9863a-glbregs: Document SC9863A syscon
>>       dt-bindings: soc: intel: lgm-syscon: Move to dedicated schema
>>       dt-bindings: soc: microchip: sparx5-cpu-syscon: Move to dedicated schema
>>       dt-bindings: soc: ti: am654-serdes-ctrl: Move to dedicated schema
>>       dt-bindings: mfd: syscon: Split and enforce documenting MFD children
>>       dt-bindings: mfd: syscon: Add APM poweroff mailbox
>>
>>  .../devicetree/bindings/mfd/syscon-common.yaml     |  71 +++++
>>  Documentation/devicetree/bindings/mfd/syscon.yaml  | 306 ++++++++++++---------
>>  .../bindings/soc/intel/intel,lgm-syscon.yaml       |  57 ++++
>>  .../soc/microchip/microchip,sparx5-cpu-syscon.yaml |  49 ++++
>>  .../bindings/soc/sprd/sprd,sc9863a-glbregs.yaml    |  55 ++++
>>  .../bindings/soc/ti/ti,am654-serdes-ctrl.yaml      |  42 +++
>>  6 files changed, 457 insertions(+), 123 deletions(-)
>> ---
>> base-commit: 8dc7c29f608649f3d9eca826e9d4fe4b8a32c472
>> change-id: 20240517-dt-bindings-mfd-syscon-split-37e23996523d
> 
> Okay, I tried to apply these whilst fixing up all the conflicts, but
> lost the will to live.  Please rebase and [RESEND].

This was based on your for-next 8dc7c29f608649f3d9ec "mfd: lm3533: Move
to new GPIO descriptor-based APIs". I'll rebase.

> 

Best regards,
Krzysztof