On Fri, Oct 6, 2023 at 7:03 AM Roger Quadros <rogerq@xxxxxxxxxx> wrote:
>
>
>
> On 05/10/2023 16:29, Nishanth Menon wrote:
> > On 16:12-20231005, Roger Quadros wrote:
> >> Hi,
> >>
> >> On 05/10/2023 11:25, Raghavendra, Vignesh wrote:
> >>> + Rob and DT list
> >>>
> >>> Hi Stephen
> >>>
> >>> On 10/5/2023 8:45 AM, Stephen Rothwell wrote:
> >>>> Hi all,
> >>>>
> >>>> [I may have missed this yesterday, sorry]
> >>>>
> >>>> After merging the ti tree, today's linux-next build (arm64 defconfig)
> >>>> produced these warnings:
> >>>>
> >>>> arch/arm64/boot/dts/ti/k3-am642-evm-nand.dtso:65.8-140.3: Warning (avoid_default_addr_size): /fragment@3/__overlay__: Relying on default #address-cells value
> >>>> arch/arm64/boot/dts/ti/k3-am642-evm-nand.dtso:65.8-140.3: Warning (avoid_default_addr_size): /fragment@3/__overlay__: Relying on default #size-cells value
> >>>>
> >>>> Introduced by commit
> >>>>
> >>>> 45a0c06571e1 ("arm64: dts: ti: am642-evm: Add overlay for NAND expansion card")
> >>>>
> >>>
> >>> Thanks for the report. I will drop the offending comment.
> >>>
> >>> Roger,
> >>>
> >>> Sorry, this would need to be fixed in dtc or need exception from DT
> >>> maintainers to ignore the warnings.
> >>
> >> Please don't drop this patch as the issue is not with the patch but with
> >> the dtc tool itself.
> >>
> >> As this is a DT overlay there is no way to specify address-cells/size-cells
> >> of parent here. This will be resolved only after merge with base tree.
> >>
> >> This will be fixed in next dtc sync.
> >> https://www.spinics.net/lists/devicetree-compiler/msg04036.html
> >>
> >> See further discussion here
> >> https://lore.kernel.org/all/CAL_JsqLmv904+_2EOmsQ__y1yLDvsT+_02i85phuh0cpe7X8NQ@xxxxxxxxxxxxxx/
> >>
> >
> > Roger, build warnings are a strict NO,NO for kernel. Lets bring in the
> > series *after* the dtc sync is complete.
> >
>
> Hi Rob,
>
> Following commit from dtc is required to resolve this issue.
>
> afbddcd418fb ("Suppress warnings on overlay fragments")
That is from 2018, so it's been in the kernel's dtc for a long time.
I believe what you need is the patch I referenced which hasn't been
applied upstream.
> FYI. Another patch of similar nature is on its way and will also require the
> above commit to DTC.
> https://lore.kernel.org/all/20231005093739.4071934-3-n-yadav@xxxxxx/
You'll need to fix dtc first.
Rob
