Re: linux-next: build warnings after merge of the gpio tree

[Rob Herring]

On Thu, Nov 9, 2017 at 1:42 AM, Linus Walleij <linus.walleij@xxxxxxxxxx> wrote:
> On Thu, Nov 9, 2017 at 4:51 AM, Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> wrote:
>> On Fri, 3 Nov 2017 16:37:24 +1100 Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> wrote:
>>>
>>> After merging the gpio tree, yesterday's linux-next build (arm
>>> multi_v7_defconfig) produced these warnings:
>>>
>>> arch/arm/boot/dts/bcm2835-rpi-b.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2835-rpi-b-rev2.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2835-rpi-a.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2835-rpi-b-plus.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2835-rpi-a-plus.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2836-rpi-2-b.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2837-rpi-3-b.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2835-rpi-zero.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/bcm2835-rpi-zero-w.dtb: Warning (phys_property): Missing property '#phy-cells' in node /phy or bad phandle (referred from /soc/usb@7e980000:phys[0])
>>> arch/arm/boot/dts/exynos5410-odroidxu.dtb: Warning (interrupts_property): Missing interrupt-controller or interrupt-map property in /soc/system-controller@10040000
>>>
>>> and many, many more.
>>>
>>> I have no idea what caused this.
>>
>> I am still getting lots of these ...
>
> I have absolutely no clue either.
>
> What I know is that there is a device tree compiler warning that can
> be turned on, and
> it generates these warnings a lot. The actual problems have been in
> the DTS files
> forever. They just recently started to look into them.
>
> It has nothing to do with the GPIO tree whatsoever, so I wonder if it
> is a side effect of
> something else?

There's a dtc update in -next which adds these new checks. It's been
there for over a month now. Unlike other dtc checks, these are
generally actual errors, so they are not off by default (enabled with
W=1,2).

In reality, there's only about 50-100 warnings. The way dts includes
are structured and because the dtc checks don't have the original
source file lines, it just multiplies the warnings.

Rob