Samsung platforms are clean of W=1 dtc warnings, so enable the warnings
by default. This way submitters don't have to remember to run a W=1
build of the .dts files and the grumpiness of the maintainers can be
reduced.
Signed-off-by: Rob Herring <[email protected]>
---
Well, there's a couple of warnings on 32-bit, but they look fixable to
me.
There's a few other platforms we could do this to. Sadly, they are still
the minority. Otherwise, we could change the default and add a flag to
disable (I_STILL_HAVENT_FIXED_MY_PLATFORMS=1).
---
arch/arm/boot/dts/samsung/Makefile | 3 +++
arch/arm64/boot/dts/exynos/Makefile | 4 ++++
2 files changed, 7 insertions(+)
diff --git a/arch/arm/boot/dts/samsung/Makefile b/arch/arm/boot/dts/samsung/Makefile
index 7becf36656b1..b6a0e2beac00 100644
--- a/arch/arm/boot/dts/samsung/Makefile
+++ b/arch/arm/boot/dts/samsung/Makefile
@@ -1,4 +1,7 @@
# SPDX-License-Identifier: GPL-2.0
+
+KBUILD_EXTRA_WARN=1
+
dtb-$(CONFIG_ARCH_EXYNOS3) += \
exynos3250-artik5-eval.dtb \
exynos3250-monk.dtb \
diff --git a/arch/arm64/boot/dts/exynos/Makefile b/arch/arm64/boot/dts/exynos/Makefile
index 6e4ba69268e5..45c996fba5e4 100644
--- a/arch/arm64/boot/dts/exynos/Makefile
+++ b/arch/arm64/boot/dts/exynos/Makefile
@@ -1,4 +1,8 @@
# SPDX-License-Identifier: GPL-2.0
+
+KBUILD_EXTRA_WARN=1
+DTC_FLAGS += -Wno-graph_child_address
+
dtb-$(CONFIG_ARCH_EXYNOS) += \
exynos5433-tm2.dtb \
exynos5433-tm2e.dtb \
--
2.42.0
On 16/11/2023 22:17, Rob Herring wrote:
> Samsung platforms are clean of W=1 dtc warnings, so enable the warnings
> by default. This way submitters don't have to remember to run a W=1
> build of the .dts files and the grumpiness of the maintainers can be
> reduced.
>
> Signed-off-by: Rob Herring <[email protected]>
> ---
> Well, there's a couple of warnings on 32-bit, but they look fixable to
> me.
>
> There's a few other platforms we could do this to. Sadly, they are still
> the minority. Otherwise, we could change the default and add a flag to
> disable (I_STILL_HAVENT_FIXED_MY_PLATFORMS=1).
64-bit has still few warnings:
https://krzk.eu/#/builders/29/builds/3710/steps/26/logs/warnings__6_
I guess these could be fixed by changing ports to port in the DTS and
bindings. I assume drivers are ready for it.
32-bit I fixed in this cycle - last three commits - so all Samsung
32-bit platforms are free from W=1 dtc warnings (yay!).
Best regards,
Krzysztof
On Fri, Nov 17, 2023 at 3:19 AM Krzysztof Kozlowski
<[email protected]> wrote:
>
> On 16/11/2023 22:17, Rob Herring wrote:
> > Samsung platforms are clean of W=1 dtc warnings, so enable the warnings
> > by default. This way submitters don't have to remember to run a W=1
> > build of the .dts files and the grumpiness of the maintainers can be
> > reduced.
> >
> > Signed-off-by: Rob Herring <[email protected]>
> > ---
> > Well, there's a couple of warnings on 32-bit, but they look fixable to
> > me.
> >
> > There's a few other platforms we could do this to. Sadly, they are still
> > the minority. Otherwise, we could change the default and add a flag to
> > disable (I_STILL_HAVENT_FIXED_MY_PLATFORMS=1).
>
> 64-bit has still few warnings:
> https://krzk.eu/#/builders/29/builds/3710/steps/26/logs/warnings__6_
I may move that graph check to W=2. There's some cases where port@1 is
optional and it doesn't really make sense to fix these.
Also, Conor wants to do this for all of riscv, but this solution is
per directory. So I need to rework it to use a different variable that
can be per directory or global.
Rob
On 21/11/2023 15:58, Rob Herring wrote:
> On Fri, Nov 17, 2023 at 3:19 AM Krzysztof Kozlowski
> <[email protected]> wrote:
>>
>> On 16/11/2023 22:17, Rob Herring wrote:
>>> Samsung platforms are clean of W=1 dtc warnings, so enable the warnings
>>> by default. This way submitters don't have to remember to run a W=1
>>> build of the .dts files and the grumpiness of the maintainers can be
>>> reduced.
>>>
>>> Signed-off-by: Rob Herring <[email protected]>
>>> ---
>>> Well, there's a couple of warnings on 32-bit, but they look fixable to
>>> me.
>>>
>>> There's a few other platforms we could do this to. Sadly, they are still
>>> the minority. Otherwise, we could change the default and add a flag to
>>> disable (I_STILL_HAVENT_FIXED_MY_PLATFORMS=1).
>>
>> 64-bit has still few warnings:
>> https://krzk.eu/#/builders/29/builds/3710/steps/26/logs/warnings__6_
>
> I may move that graph check to W=2. There's some cases where port@1 is
> optional and it doesn't really make sense to fix these.
>
> Also, Conor wants to do this for all of riscv, but this solution is
> per directory. So I need to rework it to use a different variable that
> can be per directory or global.
Just to clarify, because I still have this patch in my inbox: I assume
you are going to send a v2 of this, so I drop it.
Best regards,
Krzysztof
On Thu, Jan 25, 2024 at 12:31:42PM +0100, Krzysztof Kozlowski wrote:
> On 21/11/2023 15:58, Rob Herring wrote:
> > On Fri, Nov 17, 2023 at 3:19 AM Krzysztof Kozlowski
> > <[email protected]> wrote:
> >>
> >> On 16/11/2023 22:17, Rob Herring wrote:
> >>> Samsung platforms are clean of W=1 dtc warnings, so enable the warnings
> >>> by default. This way submitters don't have to remember to run a W=1
> >>> build of the .dts files and the grumpiness of the maintainers can be
> >>> reduced.
> >>>
> >>> Signed-off-by: Rob Herring <[email protected]>
> >>> ---
> >>> Well, there's a couple of warnings on 32-bit, but they look fixable to
> >>> me.
> >>>
> >>> There's a few other platforms we could do this to. Sadly, they are still
> >>> the minority. Otherwise, we could change the default and add a flag to
> >>> disable (I_STILL_HAVENT_FIXED_MY_PLATFORMS=1).
> >>
> >> 64-bit has still few warnings:
> >> https://krzk.eu/#/builders/29/builds/3710/steps/26/logs/warnings__6_
> >
> > I may move that graph check to W=2. There's some cases where port@1 is
> > optional and it doesn't really make sense to fix these.
> >
> > Also, Conor wants to do this for all of riscv, but this solution is
> > per directory. So I need to rework it to use a different variable that
> > can be per directory or global.
>
> Just to clarify, because I still have this patch in my inbox: I assume
> you are going to send a v2 of this, so I drop it.
We got
https://lore.kernel.org/linux-devicetree/[email protected]/
out of this instead of a v2, did we not?
On 25/01/2024 18:51, Conor Dooley wrote:
>>>>
>>>> 64-bit has still few warnings:
>>>> https://krzk.eu/#/builders/29/builds/3710/steps/26/logs/warnings__6_
>>>
>>> I may move that graph check to W=2. There's some cases where port@1 is
>>> optional and it doesn't really make sense to fix these.
>>>
>>> Also, Conor wants to do this for all of riscv, but this solution is
>>> per directory. So I need to rework it to use a different variable that
>>> can be per directory or global.
>>
>> Just to clarify, because I still have this patch in my inbox: I assume
>> you are going to send a v2 of this, so I drop it.
>
> We got
> https://lore.kernel.org/linux-devicetree/[email protected]/
> out of this instead of a v2, did we not?
Indeed, I forgot.
Best regards,
Krzysztof