[PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[Uwe Kleine-König]

If the overlay's target is only created in a previous fragment, it
doesn't exist in the unmodified base device tree. For the phandle
overwrite check this can be ignored because in this case the base tree
doesn't contain a phandle that could be overwritten.

Adapt the corresponding check to not error out if that happens but just
continue with the next fragment.

This is currently triggered by
arch/arm64/boot/dts/renesas/salvator-panel-aa104xd12.dtso in the kernel
repository which creates /panel in its first fragment and modifies it in
its second.

Reported-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/all/CAL_JsqL9MPycDjqQfPNAuGfC6EMrdzUivr+fuOS7YgU3biGd4A@mail.gmail.com/
Fixes: 1fad065080e6 ("libfdt: overlay: ensure that existing phandles are not overwritten")
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
---
 libfdt/fdt_overlay.c                | 8 +++++++-
 tests/overlay_overlay_bypath.dts    | 4 ++++
 tests/overlay_overlay_no_fixups.dts | 8 ++++++++
 3 files changed, 19 insertions(+), 1 deletion(-)

diff --git a/libfdt/fdt_overlay.c b/libfdt/fdt_overlay.c
index e19506fb56a5..28b667ffc490 100644
--- a/libfdt/fdt_overlay.c
+++ b/libfdt/fdt_overlay.c
@@ -729,7 +729,13 @@ static int overlay_prevent_phandle_overwrite(void *fdt, void *fdto)
 			return overlay;
 
 		target = fdt_overlay_target_offset(fdt, fdto, fragment, NULL);
-		if (target < 0)
+		if (target == -FDT_ERR_NOTFOUND)
+			/*
+			 * The subtree doesn't exist in the base, so nothing
+			 * will be overwritten.
+			 */
+			continue;
+		else if (target < 0)
 			return target;
 
 		ret = overlay_prevent_phandle_overwrite_node(fdt, target,
diff --git a/tests/overlay_overlay_bypath.dts b/tests/overlay_overlay_bypath.dts
index f23e7b6035e2..dfcb7cdb25a4 100644
--- a/tests/overlay_overlay_bypath.dts
+++ b/tests/overlay_overlay_bypath.dts
@@ -46,3 +46,7 @@
 		new-sub-test-property;
 	};
 };
+
+&{/new-local-node} {
+	another-new-property;
+};
diff --git a/tests/overlay_overlay_no_fixups.dts b/tests/overlay_overlay_no_fixups.dts
index e8d0f96d889c..1dbdcdc2b40f 100644
--- a/tests/overlay_overlay_no_fixups.dts
+++ b/tests/overlay_overlay_no_fixups.dts
@@ -72,6 +72,14 @@
 		};
 	};
 
+	fragment@7 {
+		target-path = "/new-local-node";
+
+		__overlay__ {
+			another-new-property;
+		};
+	};
+
 	__local_fixups__ {
 		fragment@5 {
 			__overlay__ {

base-commit: ff4f17eb58650784ffb2e8a8fbefebce1038f80b
-- 
2.43.0

Re: [PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 1472 bytes --]

Hello David,

On Wed, Jun 26, 2024 at 09:55:52AM +0200, Uwe Kleine-König wrote:
> If the overlay's target is only created in a previous fragment, it
> doesn't exist in the unmodified base device tree. For the phandle
> overwrite check this can be ignored because in this case the base tree
> doesn't contain a phandle that could be overwritten.
> 
> Adapt the corresponding check to not error out if that happens but just
> continue with the next fragment.
> 
> This is currently triggered by
> arch/arm64/boot/dts/renesas/salvator-panel-aa104xd12.dtso in the kernel
> repository which creates /panel in its first fragment and modifies it in
> its second.
> 
> Reported-by: Rob Herring <robh@kernel.org>
> Link: https://lore.kernel.org/all/CAL_JsqL9MPycDjqQfPNAuGfC6EMrdzUivr+fuOS7YgU3biGd4A@mail.gmail.com/
> Fixes: 1fad065080e6 ("libfdt: overlay: ensure that existing phandles are not overwritten")
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>

I wonder about this patch's state. Does David wait for feedback by Rob
as he reported the issue? Does Rob expect Geert to comment as
salvator-panel-aa104xd12.dtso is in his maintainer area? Are the
responsible people just busy; or is this fix already hidden in their
backlog?

Otherwise I think the patch is fine to go in. It passes the test suite
and adds a new test that fails without the code change (but passes when
the patch is applied completely).

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[Rob Herring]

On Tue, Jul 2, 2024 at 7:44 AM Uwe Kleine-König
<u.kleine-koenig@baylibre.com> wrote:
>
> Hello David,
>
> On Wed, Jun 26, 2024 at 09:55:52AM +0200, Uwe Kleine-König wrote:
> > If the overlay's target is only created in a previous fragment, it
> > doesn't exist in the unmodified base device tree. For the phandle
> > overwrite check this can be ignored because in this case the base tree
> > doesn't contain a phandle that could be overwritten.
> >
> > Adapt the corresponding check to not error out if that happens but just
> > continue with the next fragment.
> >
> > This is currently triggered by
> > arch/arm64/boot/dts/renesas/salvator-panel-aa104xd12.dtso in the kernel
> > repository which creates /panel in its first fragment and modifies it in
> > its second.
> >
> > Reported-by: Rob Herring <robh@kernel.org>
> > Link: https://lore.kernel.org/all/CAL_JsqL9MPycDjqQfPNAuGfC6EMrdzUivr+fuOS7YgU3biGd4A@mail.gmail.com/
> > Fixes: 1fad065080e6 ("libfdt: overlay: ensure that existing phandles are not overwritten")
> > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
>
> I wonder about this patch's state. Does David wait for feedback by Rob
> as he reported the issue? Does Rob expect Geert to comment as
> salvator-panel-aa104xd12.dtso is in his maintainer area? Are the
> responsible people just busy; or is this fix already hidden in their
> backlog?

I think it's just waiting on David.

The patch looks good to me, but I haven't tested it nor am I that
familiar with this particular code.

Rob

Re: [PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 2157 bytes --]

On Wed, Jul 03, 2024 at 11:06:32AM -0600, Rob Herring wrote:
> On Tue, Jul 2, 2024 at 7:44 AM Uwe Kleine-König
> <u.kleine-koenig@baylibre.com> wrote:
> >
> > Hello David,
> >
> > On Wed, Jun 26, 2024 at 09:55:52AM +0200, Uwe Kleine-König wrote:
> > > If the overlay's target is only created in a previous fragment, it
> > > doesn't exist in the unmodified base device tree. For the phandle
> > > overwrite check this can be ignored because in this case the base tree
> > > doesn't contain a phandle that could be overwritten.
> > >
> > > Adapt the corresponding check to not error out if that happens but just
> > > continue with the next fragment.
> > >
> > > This is currently triggered by
> > > arch/arm64/boot/dts/renesas/salvator-panel-aa104xd12.dtso in the kernel
> > > repository which creates /panel in its first fragment and modifies it in
> > > its second.
> > >
> > > Reported-by: Rob Herring <robh@kernel.org>
> > > Link: https://lore.kernel.org/all/CAL_JsqL9MPycDjqQfPNAuGfC6EMrdzUivr+fuOS7YgU3biGd4A@mail.gmail.com/
> > > Fixes: 1fad065080e6 ("libfdt: overlay: ensure that existing phandles are not overwritten")
> > > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
> >
> > I wonder about this patch's state. Does David wait for feedback by Rob
> > as he reported the issue? Does Rob expect Geert to comment as
> > salvator-panel-aa104xd12.dtso is in his maintainer area? Are the
> > responsible people just busy; or is this fix already hidden in their
> > backlog?
> 
> I think it's just waiting on David.

Sorry, yes.  I'm always finding it hard to find time for dtc/libfdt
maintenance stuff, and I've been particularly busy lately.  I did
catch up on a bunch of trivial dtc/libfdt patches lately, but this
one's a bit more complex so I didn't get to it yet.

> The patch looks good to me, but I haven't tested it nor am I that
> familiar with this particular code.
> 
> Rob
> 

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 598 bytes --]

Hello David,

On Fri, Jul 05, 2024 at 09:49:17PM +1000, David Gibson wrote:
> On Wed, Jul 03, 2024 at 11:06:32AM -0600, Rob Herring wrote:
> > I think it's just waiting on David.
> 
> Sorry, yes.  I'm always finding it hard to find time for dtc/libfdt
> maintenance stuff, and I've been particularly busy lately.  I did
> catch up on a bunch of trivial dtc/libfdt patches lately, but this
> one's a bit more complex so I didn't get to it yet.

Good to know it's on your radar. If you have questions or other needs to
talk about this, don't hesitate to contact me.

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH] libfdt: overlay: Fix phandle overwrite check for new subtrees

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 2933 bytes --]

On Wed, Jun 26, 2024 at 09:55:52AM +0200, Uwe Kleine-König wrote:
> If the overlay's target is only created in a previous fragment, it
> doesn't exist in the unmodified base device tree. For the phandle
> overwrite check this can be ignored because in this case the base tree
> doesn't contain a phandle that could be overwritten.
> 
> Adapt the corresponding check to not error out if that happens but just
> continue with the next fragment.
> 
> This is currently triggered by
> arch/arm64/boot/dts/renesas/salvator-panel-aa104xd12.dtso in the kernel
> repository which creates /panel in its first fragment and modifies it in
> its second.
> 
> Reported-by: Rob Herring <robh@kernel.org>
> Link: https://lore.kernel.org/all/CAL_JsqL9MPycDjqQfPNAuGfC6EMrdzUivr+fuOS7YgU3biGd4A@mail.gmail.com/
> Fixes: 1fad065080e6 ("libfdt: overlay: ensure that existing phandles are not overwritten")
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>

Merged, thanks.

> ---
>  libfdt/fdt_overlay.c                | 8 +++++++-
>  tests/overlay_overlay_bypath.dts    | 4 ++++
>  tests/overlay_overlay_no_fixups.dts | 8 ++++++++
>  3 files changed, 19 insertions(+), 1 deletion(-)
> 
> diff --git a/libfdt/fdt_overlay.c b/libfdt/fdt_overlay.c
> index e19506fb56a5..28b667ffc490 100644
> --- a/libfdt/fdt_overlay.c
> +++ b/libfdt/fdt_overlay.c
> @@ -729,7 +729,13 @@ static int overlay_prevent_phandle_overwrite(void *fdt, void *fdto)
>  			return overlay;
>  
>  		target = fdt_overlay_target_offset(fdt, fdto, fragment, NULL);
> -		if (target < 0)
> +		if (target == -FDT_ERR_NOTFOUND)
> +			/*
> +			 * The subtree doesn't exist in the base, so nothing
> +			 * will be overwritten.
> +			 */
> +			continue;
> +		else if (target < 0)
>  			return target;
>  
>  		ret = overlay_prevent_phandle_overwrite_node(fdt, target,
> diff --git a/tests/overlay_overlay_bypath.dts b/tests/overlay_overlay_bypath.dts
> index f23e7b6035e2..dfcb7cdb25a4 100644
> --- a/tests/overlay_overlay_bypath.dts
> +++ b/tests/overlay_overlay_bypath.dts
> @@ -46,3 +46,7 @@
>  		new-sub-test-property;
>  	};
>  };
> +
> +&{/new-local-node} {
> +	another-new-property;
> +};
> diff --git a/tests/overlay_overlay_no_fixups.dts b/tests/overlay_overlay_no_fixups.dts
> index e8d0f96d889c..1dbdcdc2b40f 100644
> --- a/tests/overlay_overlay_no_fixups.dts
> +++ b/tests/overlay_overlay_no_fixups.dts
> @@ -72,6 +72,14 @@
>  		};
>  	};
>  
> +	fragment@7 {
> +		target-path = "/new-local-node";
> +
> +		__overlay__ {
> +			another-new-property;
> +		};
> +	};
> +
>  	__local_fixups__ {
>  		fragment@5 {
>  			__overlay__ {
> 
> base-commit: ff4f17eb58650784ffb2e8a8fbefebce1038f80b

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]