| Message ID | 20200224145821.262873-2-jbrunet@baylibre.com |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
| Series | ASoC: meson: convert dt-bindings to schema | expand |
| Context | Check | Description |
|---|---|---|
| robh/checkpatch | warning | "total: 0 errors, 1 warnings, 16 lines checked" |
On Mon, Feb 24, 2020 at 03:58:13PM +0100, Jerome Brunet wrote: > There is a fair amount of warnings when running 'make dtbs_check' with > amlogic,gx-sound-card.yaml. > > Ex: > arch/arm64/boot/dts/amlogic/meson-gxm-q200.dt.yaml: sound: dai-link-0:sound-dai:0:1: missing phandle tag in 0 > arch/arm64/boot/dts/amlogic/meson-gxm-q200.dt.yaml: sound: dai-link-0:sound-dai:0:2: missing phandle tag in 0 > arch/arm64/boot/dts/amlogic/meson-gxm-q200.dt.yaml: sound: dai-link-0:sound-dai:0: [66, 0, 0] is too long > > The reason is that the sound-dai phandle provided has cells, and in such > case the schema should use 'phandle-array' instead of 'phandle', even if > the property expects a single phandle. > > Fixes: fd00366b8e41 ("ASoC: meson: gx: add sound card dt-binding documentation") > Signed-off-by: Jerome Brunet <jbrunet@baylibre.com> > --- > Hi Mark, > > The statement above is based on this LKML discussion I found: > https://lkml.org/lkml/2019/9/30/382 > > To be honest, I don't really get why the consumer should know whether > the phandle will have cells or not. AFAIK, the consumer does not care > about this ... Yeah, I think another type definition is needed here to distinguish between a single phandle+args and an array of phandle+args. In any case, Acked-by: Rob Herring <robh@kernel.org> > > .../devicetree/bindings/sound/amlogic,gx-sound-card.yaml | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/Documentation/devicetree/bindings/sound/amlogic,gx-sound-card.yaml b/Documentation/devicetree/bindings/sound/amlogic,gx-sound-card.yaml index fb374c659be1..a48222e8cd08 100644 --- a/Documentation/devicetree/bindings/sound/amlogic,gx-sound-card.yaml +++ b/Documentation/devicetree/bindings/sound/amlogic,gx-sound-card.yaml @@ -57,7 +57,7 @@ patternProperties: rate sound-dai: - $ref: /schemas/types.yaml#/definitions/phandle + $ref: /schemas/types.yaml#/definitions/phandle-array description: phandle of the CPU DAI patternProperties: @@ -71,7 +71,7 @@ patternProperties: properties: sound-dai: - $ref: /schemas/types.yaml#/definitions/phandle + $ref: /schemas/types.yaml#/definitions/phandle-array description: phandle of the codec DAI required:
There is a fair amount of warnings when running 'make dtbs_check' with amlogic,gx-sound-card.yaml. Ex: arch/arm64/boot/dts/amlogic/meson-gxm-q200.dt.yaml: sound: dai-link-0:sound-dai:0:1: missing phandle tag in 0 arch/arm64/boot/dts/amlogic/meson-gxm-q200.dt.yaml: sound: dai-link-0:sound-dai:0:2: missing phandle tag in 0 arch/arm64/boot/dts/amlogic/meson-gxm-q200.dt.yaml: sound: dai-link-0:sound-dai:0: [66, 0, 0] is too long The reason is that the sound-dai phandle provided has cells, and in such case the schema should use 'phandle-array' instead of 'phandle', even if the property expects a single phandle. Fixes: fd00366b8e41 ("ASoC: meson: gx: add sound card dt-binding documentation") Signed-off-by: Jerome Brunet <jbrunet@baylibre.com> --- Hi Mark, The statement above is based on this LKML discussion I found: https://lkml.org/lkml/2019/9/30/382 To be honest, I don't really get why the consumer should know whether the phandle will have cells or not. AFAIK, the consumer does not care about this ... .../devicetree/bindings/sound/amlogic,gx-sound-card.yaml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)