Re: [PATCH 2/2] ARM: dts: aspeed: Add ASRock E3C256D4I BMC

[Zev Weiss]

On Wed, Nov 15, 2023 at 04:13:29AM PST, Krzysztof Kozlowski wrote:
> On 14/11/2023 23:37, Zev Weiss wrote:
>
> > > > +	};
> > > > +
> > > > +	memory@80000000 {
> > > > +		reg = <0x80000000 0x20000000>;
> > > > +	};
> > > > +
> > > > +	leds {
> > > > +		compatible = "gpio-leds";
> > > > +
> > > > +		heartbeat {
> > >
> > > It does not look like you tested the DTS against bindings. Please run
> > > `make dtbs_check W=1` (see
> > > Documentation/devicetree/bindings/writing-schema.rst or
> > > https://www.linaro.org/blog/tips-and-tricks-for-validating-devicetree-sources-with-the-devicetree-schema/
> > > for instructions).
>
> The node names are clearly wrong and you got output. It's easy to spot
> if your patch worked in the first place:
>
> fatal error: dt-bindings/watchdog/aspeed-wdt.h: No such file or directory

The patch series was based on Linus's tree at the time I sent it; that 
file was added in commit 9931be2cfca3 ("dt-bindings: watchdog: 
aspeed-wdt: Add aspeed,reset-mask property"), which was included in the 
6.7-rc1 tag, FWIW.

After some debugging I discovered that the reason I wasn't getting any 
output from dtbs_check was that I had neglected to specify ARCH=arm when 
invoking it, and when that defaulted to x86 it silently did a whole 
bunch of nothing.  It might be nice if something emitted some sort of 
warning when invoked with nonsensical parameters (especially if I've 
explicitly specified a dtstree=... argument pointing to arch/arm/...), 
but oh well.

After re-running it with ARCH=arm, it produced a fairly voluminous spew 
of output, though after combing through it all as best I can tell only 
one or two lines of it actually pertain to the .dts in question itself 
(the vast majority being from a .dtsi it #includes).  I've fixed the one 
remaining issue it reported (missing #address-cells and #size-cells on 
the FRU eeprom node) and will post a v2 in the next few days (and 
likewise on the spc621d8hm3 series).


Zev