On 3/13/22 18:50, Krzysztof Kozlowski wrote:
On 13/03/2022 18:09, Marek Vasut wrote:
On 3/13/22 16:47, Krzysztof Kozlowski wrote:
Hi,
[...]
diff --git a/Documentation/devicetree/bindings/display/bridge/nxp,ldb.yaml b/Documentation/devicetree/bindings/display/bridge/nxp,ldb.yaml
new file mode 100644
index 0000000000000..a05dd05547836
--- /dev/null
+++ b/Documentation/devicetree/bindings/display/bridge/nxp,ldb.yaml
@@ -0,0 +1,99 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/display/bridge/nxp,ldb.yaml#
In title, description and commit msg you point this is specific to
i.MX8M, so maybe reflect it in the file name as well.
That's how it is so far, however NXP does recycle their IPs quite a bit
so I don't want to encode the SoC type into the bindings file name. I do
expect them to re-use this bridge somewhere else sooner rather than later.
[...]
The rest is fixed.
[...]
+examples:
+ - |
+ #include <dt-bindings/clock/imx8mp-clock.h>
+
+ lvds-ldb {
Generic node name, so "bridge" or "display-bridge"
+ #address-cells = <0>;
+ #size-cells = <0>;
Why do you need address and size cells? This will complain if you test
your bindings with proper compatible.
Because the subnodes of this bridge have no dimension, so address/size
cells = 0.
I don't understand the second part about "proper compatible", can you
elaborate ?
You have wrong compatible in example and in bindings, so you do not see
the error of address/size cells.
They are also not required in your example, since you do not have unit
address. Otherwise your lvds-ldb node would be wrong (see its
address/size cells).
If you supply address-cells you should see a W=1 warning:
#address-cells/#size-cells without "ranges" or child "reg" property
Ah, right, both are fixed now. I'll wait a bit for more feedback esp. on
the bridge driver and send a V2.
