Hello Rob,

I just submitted patch v4 that addresses most of your feedback.

On Thu, Jan 20, 2022 at 12:07 PM,Rob Herring wrote:

From: Rob Herring <robh@kernel.org>

On Fri, Jan 14, 2022 at 06:31:33PM -0800, Dmitry Antipov wrote:

quoted

From: Dmitry Antipov <redacted>

Added documentation describes the required properties for implementing
Device Tree for a device supporting HID over SPI and also provides an
example.

Bindings are in DT schema format now.

quoted

Signed-off-by: Dmitry Antipov <redacted>
---
 .../bindings/input/hid-over-spi.txt           | 43 +++++++++++++++++++
 1 file changed, 43 insertions(+)
 create mode 100755
Documentation/devicetree/bindings/input/hid-over-spi.txt

diff --git a/Documentation/devicetree/bindings/input/hid-over-spi.txt

b/Documentation/devicetree/bindings/input/hid-over-spi.txt
new file mode 100755
index 000000000000..5eba95b5724e

--- /dev/null
+++ b/Documentation/devicetree/bindings/input/hid-over-spi.txt

@@ -0,0 +1,43 @@
+* HID over SPI Device-Tree bindings
+
+HID over SPI provides support for Human Interface Devices over the
+SPI bus. HID Over SPI Protocol Specification 1.0 was written by
+Microsoft and is available at

https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.m
icrosoft.com%2Fen-
us%2Fdownload%2Fdetails.aspx%3Fid%3D103325&data=04%7C01%7Cd
manti%40microsoft.com%7Cb0b84680fa2c45bc2f3a08d9dc5063f6%7C72f988
bf86f141af91ab2d7cd011db47%7C1%7C0%7C637783060144589082%7CUnkn
own%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1h
aWwiLCJXVCI6Mn0%3D%7C3000&sdata=bWGZDRYtVovy3XYlpDsZ8xR2AL
be5YoP0iSZ8xORSyA%3D&reserved=0.

quoted

+
+If this binding is used, the kernel module spi-hid will handle the
+communication

What's a kernel module? ;) Bindings are OS independent (or supposed to be), so
kernel details do not belong here.

quoted

+with the device and the generic hid core layer will handle the protocol.
+
+Required properties:
+- compatible: must be "hid-over-spi"

Bindings describe devices, not protocols. There is no such 'hid-over-spi' device.
Please see the existing hid-over-i2c binding.
It's fine to have this compatible, but only as a fallback compatible.

In the v4 patch the yaml file expects the device-specific compatible
string, but also 'hid-over-spi'.

quoted

+- interrupts: interrupt line
+- vdd-supply: phandle of the regulator that provides the supply
+voltage

What happens when the device has 2 supplies? And there's some timing
requirement between them?

I am not sure if there is a way to account for all possible scenarios.
Our device only has one power regulator and the binding is now
device-specific.

quoted

+- reset_gpio-gpios: gpio wired to the device reset line

'reset-gpios' is the standard name.

quoted

+- post-power-on-delay-ms: time required by the device after enabling
+its regulators or powering it on, before it is ready for
+communication
+- minimal-reset-delay-ms: minimum amount of time that device needs to
+be in reset state for the reset to take effect

These properties are what happens when we try to do generic bindings.
It's a never-ending addition of more properties to try to describe
(poorly) the power sequencing requirements.

I have made these optional in the v4 patch.

quoted

+- input-report-header-address: this property and the rest are
+described in HID Over SPI Protocol Spec 1.0
+- input-report-body-address
+- output-report-address
+- read-opcode
+- write-opcode
+- flags

A bit too generic. We generally want to avoid having a property name with 2
different types/meaning. It's not something we check for yet, but I plan to at
some point.

I've made "flags" more specific in the v4 patch.

quoted

+
+Example:
+	spi-hid-dev0 {
+		compatible = "hid-over-spi";
+		reg = <0>;
+		interrupts-extended = <&tlmm 42 IRQ_TYPE_EDGE_FALLING>;
+		vdd-supply = <&pm8350c_l3>;
+		input-report-header-address = <0x1000>;
+		input-report-body-address = <0x1004>;
+		output-report-address = <0x2000>;
+		read-opcode = <0x0b>;
+		write-opcode = <0x02>;
+		flags = <0x00>;
+		reset_gpio-gpios = <&tlmm 25 GPIO_ACTIVE_LOW>;
+		post-power-on-delay-ms = <5>;
+		minimal-reset-delay-ms = <5>;
+
+	};
--
2.25.1