Re: [PATCH v3 1/3] dt-bindings: mfd: syscon: Disallow simple-bus with syscon
"Rob Herring (Arm)" <[email protected]>
| Newsgroups | org.kernel.vger.linux-omap,org.infradead.lists.linux-arm-kernel,org.infradead.lists.linux-mediatek,org.infradead.lists.linux-rockchip,org.kernel.vger.linux-devicetree,org.kernel.vger.linux-kernel,org.kernel.vger.linux-renesas-soc |
|---|---|
| Message-ID | <[email protected]> |
On Mon, 08 Jun 2026 22:44:24 +0200, Krzysztof Kozlowski wrote: > "syscon" is a system controller with registers having their own > functions, thus not really a trivial MMIO simple bus. "simple-bus" on > the other hand is just a bus on which multiple devices sit and the > "simple" means no functions are allowed here. > > Combination of both "syscon" and "simple-bus" is abuse of DT for easier > instantiating of Linux device drivers so add a schema to disallow that. > > Unfortunately there are a few old cases of that patterns, so add > exceptions: > > 1. "cznic,turris1x-cpld" and "img,pistachio-cr-periph" are already used > in upstream DTS. > > 2. TI has several DTSI with a child of SCM device (e.g. "ti,am3-scm") > using "syscon" and "simple-bus" but without a dedicated compatible > documented anywhere. Add new compatibles for such cases. > > Additionally, add comments around code enforcing two or three > compatibles: it is similar safeguard detecting incorrect bindings. > > Signed-off-by: Krzysztof Kozlowski <[email protected]> > > --- > > Changes in v3: > 1. s/ti,omap5-scm-conf/ti,omap5-sysc-padconf-global/ because it is more > appropriate (specific) > 2. Add comments, why simple-mfd+syscon has dedicated if:then: > > Changes in v2: > 1. Complete patch. I accidentally sent only part of it, built on top of > internal WIP which I forgot to squash. > I received Ack from Rob, but change is significant, so please kindly > re-review. > --- > .../devicetree/bindings/mfd/syscon-common.yaml | 34 ++++++++++++++++++++++ > 1 file changed, 34 insertions(+) > Acked-by: Rob Herring (Arm) <[email protected]>