diff mbox

media: add V4L2 DT binding documentation

Message ID Pine.LNX.4.64.1209111746420.22084@axis700.grange (mailing list archive)
State New, archived
Headers show

Commit Message

Guennadi Liakhovetski Sept. 11, 2012, 3:51 p.m. UTC
This patch adds a document, describing common V4L2 device tree bindings.

Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 Documentation/devicetree/bindings/media/v4l2.txt |  143 ++++++++++++++++++++++
 1 files changed, 143 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/media/v4l2.txt

Comments

Hi Guennadi,

On 09/11/2012 05:51 PM, Guennadi Liakhovetski wrote:
> This patch adds a document, describing common V4L2 device tree bindings.
> 
> Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> ---
>  Documentation/devicetree/bindings/media/v4l2.txt |  143 ++++++++++++++++++++++
>  1 files changed, 143 insertions(+), 0 deletions(-)
>  create mode 100644 Documentation/devicetree/bindings/media/v4l2.txt
> 
> diff --git a/Documentation/devicetree/bindings/media/v4l2.txt b/Documentation/devicetree/bindings/media/v4l2.txt
> new file mode 100644
> index 0000000..55da6de
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/media/v4l2.txt
> @@ -0,0 +1,143 @@
> +Video4Linux Version 2 (V4L2)
> +
> +General concept:
> +- video pipelines consist of external devices, e.g., camera sensors, controlled
> +  over an I2C bus, and SoC internal IP blocks, including video DMA engines and
> +  video data processors.
> +- this document describes common bindings of all video pipeline devices.
> +- SoC internal blocks are described by DT nodes, placed similarly to other SoC
> +  blocks.
> +- external devices are places on their respective control busses, e.g., I2C
> +- data interfaces on all video devices are described by "port" child DT nodes
> +- port configuration depends on other devices, participating in the data
> +  transfer, and is described by "link" DT nodes, specified as children of
> +  all "port" nodes, connected to this bus.
> +- if a port can be configured to work with more than one other device on the
> +  same bus, a "link" child DT node must be provided for each of them.
> +- if more than one port is present on a device or more than one link is
> +  connected to a port, a common scheme, using "#address-cells," "#size-cells"
> +  and "reg" properties is used.

How about rephrasing/rearranging this to:

This document describes common bindings of all video pipeline devices.

General concept
---------------

Video pipelines consist of external devices, e.g. camera sensors, controlled
over an I2C, SPI or UART bus, and SoC internal IP blocks, including video DMA
engines and video data processors.
SoC internal blocks are described by DT nodes, placed similarly to other SoC
blocks. External devices are represented as child nodes of their respective
bus controller nodes, e.g. I2C.

Data interfaces on all video devices are described by "port" child DT nodes.
Configuration of a port depends on other devices participating in the data
transfer and is described by "link" DT nodes, specified as children of the
"port" nodes:

/foo {
	port@0 {
		link@0 { ... };
		link@1 { ... };
	};
	port@1 { ... };
};

If a port can be configured to work with more than one other device on the
same bus, a "link" child DT node must be provided for each of them.
If more than one port is present on a device or more than one link is
connected to a port, a common scheme, using "#address-cells" "#size-cells"
and "reg" properties is used.

> +Optional link properties:
> +- remote: phandle to the other endpoint link DT node.
> +- data-shift: on parallel data busses, if data-width is used to specify the

s/busses/buses

> +  number of data lines, data-shift can be used to specify which data lines are
> +  used, e.g., "data-width=<10>; data-shift=<2>;" means, that lines 9:2 are used.
> +- hsync-active: 1 or 0 for active-high or -low HSYNC signal polarity
> +  respectively.
> +- vsync-active: ditto for VSYNC. Note, that if HSYNC and VSYNC polarities are
> +  not specified, embedded synchronisation may be required, where supported.
> +- pclk-sample: rising (1) or falling (0) edge to sample the pixel clock pin.
> +- immutable: used for SoC-internal links, if no configuration is required.
> +- data-lanes: array of serial, e.g., MIPI CSI-2, data hardware lane numbers in

nit: s/e.g.,/e.g. ?

> +  the ascending order, beginning with logical lane 0.
> +- clock-lanes: hardware lane number, used for the clock lane.

s/clock lane/clock ?

> +
> +Example:
> +
> +	ceu0: ceu@0xfe910000 {
> +		compatible = "renesas,sh-mobile-ceu";
> +		reg = <0xfe910000 0xa0>;
> +		interrupts = <0x880>;
> +
> +		mclk: master_clock {
> +			compatible = "renesas,ceu-clock";
> +			#clock-cells = <1>;
> +			clock-frequency = <50000000>;	/* max clock frequency */
> +			clock-output-names = "mclk";
> +		};
> +
> +		port {
> +			#address-cells = <1>;
> +			#size-cells = <0>;
> +
> +			ceu0_1: link@1 {
> +				reg = <1>;		/* local link # */
> +				remote = <&ov772x_1_1>;	/* remote phandle */
> +				bus-width = <8>;	/* used data lines */
> +				data-shift = <0>;	/* lines 7:0 are used */
> +	
> +				/* If [hv]sync-active are missing, embedded bt.605 sync is used */
> +				hsync-active = <1>;	/* active high */
> +				vsync-active = <1>;	/* active high */
> +				pclk-sample = <1>;	/* rising */
> +			};
> +
> +			ceu0_0: link@0 {
> +				reg = <0>;
> +				remote = <&csi2_2>;
> +				immutable;
> +			};
> +		};
> +	};
> +
> +	i2c0: i2c@0xfff20000 {
> +		...
> +		ov772x_1: camera@0x21 {
> +			compatible = "omnivision,ov772x";
> +			reg = <0x21>;
> +			vddio-supply = <&regulator1>;
> +			vddcore-supply = <&regulator2>;
> +
> +			clock-frequency = <20000000>;
> +			clocks = <&mclk 0>;
> +			clock-names = "xclk";
> +
> +			port {
> +				/* With 1 link per port no need in addresses */

s/in/for ?

> +				ov772x_1_1: link {
> +					bus-width = <8>;
> +					remote = <&ceu0_1>;
> +					hsync-active = <1>;
> +					hsync-active = <0>;	/* who came up with an inverter here?... */
> +					pclk-sample = <1>;
> +				};
> +			};
> +		};
> +
> +		imx074: camera@0x1a {
> +			compatible = "sony,imx074";
> +			reg = <0x1a>;
> +			vddio-supply = <&regulator1>;
> +			vddcore-supply = <&regulator2>;
> +
> +			clock-frequency = <30000000>;	/* shared clock with ov772x_1 */
> +			clocks = <&mclk 0>;
> +			clock-names = "sysclk";		/* assuming this is the name in the datasheet */
> +
> +			port {
> +				imx074_1: link {
> +					clock-lanes = <0>;
> +					data-lanes = <1>, <2>;
> +					remote = <&csi2_1>;
> +				};
> +			};
> +		};
> +	};
> +
> +	csi2: csi2@0xffc90000 {
> +		compatible = "renesas,sh-mobile-csi2";
> +		reg = <0xffc90000 0x1000>;
> +		interrupts = <0x17a0>;
> +		#address-cells = <1>;
> +		#size-cells = <0>;
> +
> +		port@1 {
> +			compatible = "renesas,csi2c";	/* one of CSI2I and CSI2C */
> +			reg = <1>;			/* CSI-2 PHY #1 of 2: PHY_S, PHY_M has port address 0, is unused */
> +
> +			csi2_1: link {
> +				clock-lanes = <0>;
> +				data-lanes = <2>, <1>;
> +				remote = <&imx074_1>;
> +			};
> +		};
> +		port@2 {
> +			reg = <2>;			/* port 2: link to the CEU */
> +
> +			csi2_2: link {
> +				immutable;
> +				remote = <&ceu0_0>;
> +			};
> +		};
> +	};

--

Regards,
Sylwester
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Stephen Warren Sept. 12, 2012, 6:53 p.m. UTC | #2
On 09/11/2012 09:51 AM, Guennadi Liakhovetski wrote:
> This patch adds a document, describing common V4L2 device tree bindings.
> 
> Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>

Overall, I think this looks pretty reasonable, so:

Acked-by: Stephen Warren <swarren@wwwdotorg.org>

Just a couple comments:

> +++ b/Documentation/devicetree/bindings/media/v4l2.txt

> +	ceu0: ceu@0xfe910000 {

> +		mclk: master_clock {
> +			compatible = "renesas,ceu-clock";
> +			#clock-cells = <1>;

Why 1? If there's only 1 clock output from this provider, I don't see a
need for any cells, unless there are some configuration flags?

> +			clock-frequency = <50000000>;	/* max clock frequency */
> +			clock-output-names = "mclk";
> +		};
> +
> +		port {
...
> +			ceu0_0: link@0 {
> +				reg = <0>;
> +				remote = <&csi2_2>;
> +				immutable;

Did we decide "immutable" was actually needed? Presumably the driver for
the HW in question knows the HW isn't configurable, and would simply not
attempt to apply any configuration even if the .dts author erroneously
provided some?

> +			};
> +		};
> +	};
> +
> +	i2c0: i2c@0xfff20000 {
...
> +		ov772x_1: camera@0x21 {
...
> +			clocks = <&mclk 0>;

So presumably that could just be "clocks = <&mclk>;"?
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Guennadi Liakhovetski Sept. 12, 2012, 7:28 p.m. UTC | #3
Hi Stephen

On Wed, 12 Sep 2012, Stephen Warren wrote:

> On 09/11/2012 09:51 AM, Guennadi Liakhovetski wrote:
> > This patch adds a document, describing common V4L2 device tree bindings.
> > 
> > Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
> > Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> 
> Overall, I think this looks pretty reasonable, so:

Good, thanks!

> Acked-by: Stephen Warren <swarren@wwwdotorg.org>
> 
> Just a couple comments:
> 
> > +++ b/Documentation/devicetree/bindings/media/v4l2.txt
> 
> > +	ceu0: ceu@0xfe910000 {
> 
> > +		mclk: master_clock {
> > +			compatible = "renesas,ceu-clock";
> > +			#clock-cells = <1>;
> 
> Why 1? If there's only 1 clock output from this provider, I don't see a
> need for any cells, unless there are some configuration flags?

Yes, indeed, that's also what's suggested in the clock bindings 
documentation, thanks for pointing out.

> 
> > +			clock-frequency = <50000000>;	/* max clock frequency */
> > +			clock-output-names = "mclk";
> > +		};
> > +
> > +		port {
> ...
> > +			ceu0_0: link@0 {
> > +				reg = <0>;
> > +				remote = <&csi2_2>;
> > +				immutable;
> 
> Did we decide "immutable" was actually needed? Presumably the driver for
> the HW in question knows the HW isn't configurable, and would simply not
> attempt to apply any configuration even if the .dts author erroneously
> provided some?

Well, I've been thinking about this today. I think, bridge drivers will 
call centrally provided helper functions to enumerate links inside ports. 
While doing that they will want to differentiate between links to external 
devices with explicit configuration, and links to internal devices, whose 
configuration drivers might be able to figure out themselves. How should a 
driver find out what device this link is pointing to? Should it follow the 
"remote" phandle and then check the "compatible" property? The word 
"immutable" is a hint, that this is a link to an internal device, but it 
might either be unneeded or be transformed into something more 
informative.

> > +			};
> > +		};
> > +	};
> > +
> > +	i2c0: i2c@0xfff20000 {
> ...
> > +		ov772x_1: camera@0x21 {
> ...
> > +			clocks = <&mclk 0>;
> 
> So presumably that could just be "clocks = <&mclk>;"?

Right.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Stephen Warren Sept. 12, 2012, 8:53 p.m. UTC | #4
On 09/12/2012 01:28 PM, Guennadi Liakhovetski wrote:
> Hi Stephen
> 
> On Wed, 12 Sep 2012, Stephen Warren wrote:
> 
>> On 09/11/2012 09:51 AM, Guennadi Liakhovetski wrote:
>>> This patch adds a document, describing common V4L2 device tree bindings.
>>>
>>> Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
>>> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
>>
>> Overall, I think this looks pretty reasonable, so:
...
>>
>>> +			clock-frequency = <50000000>;	/* max clock frequency */
>>> +			clock-output-names = "mclk";
>>> +		};
>>> +
>>> +		port {
>> ...
>>> +			ceu0_0: link@0 {
>>> +				reg = <0>;
>>> +				remote = <&csi2_2>;
>>> +				immutable;
>>
>> Did we decide "immutable" was actually needed? Presumably the driver for
>> the HW in question knows the HW isn't configurable, and would simply not
>> attempt to apply any configuration even if the .dts author erroneously
>> provided some?
> 
> Well, I've been thinking about this today. I think, bridge drivers will 

Sorry, I'm not sure what a "bridge" driver is; is it any v4l2 device?

> call centrally provided helper functions to enumerate links inside ports. 

Presumably a given driver would only parse the ports/links of its own DT
node, and hence would be able to provide a parameter to any helper
function that indicated the same information that "immutable" would?

> While doing that they will want to differentiate between links to external 
> devices with explicit configuration, and links to internal devices, whose 
> configuration drivers might be able to figure out themselves. How should a 
> driver find out what device this link is pointing to? Should it follow the 
> "remote" phandle and then check the "compatible" property? The word 
> "immutable" is a hint, that this is a link to an internal device, but it 
> might either be unneeded or be transformed into something more 
> informative.

I would imagine that a given driver would only ever parse its own DT
node; the far end of any link is purely the domain of the other driver.
I thought that each link node would contain whatever hsync-active,
data-lanes, ... properties that were needed to configure the local device?
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Guennadi Liakhovetski Sept. 12, 2012, 9:17 p.m. UTC | #5
On Wed, 12 Sep 2012, Stephen Warren wrote:

> On 09/12/2012 01:28 PM, Guennadi Liakhovetski wrote:
> > Hi Stephen
> > 
> > On Wed, 12 Sep 2012, Stephen Warren wrote:
> > 
> >> On 09/11/2012 09:51 AM, Guennadi Liakhovetski wrote:
> >>> This patch adds a document, describing common V4L2 device tree bindings.
> >>>
> >>> Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
> >>> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> >>
> >> Overall, I think this looks pretty reasonable, so:
> ...
> >>
> >>> +			clock-frequency = <50000000>;	/* max clock frequency */
> >>> +			clock-output-names = "mclk";
> >>> +		};
> >>> +
> >>> +		port {
> >> ...
> >>> +			ceu0_0: link@0 {
> >>> +				reg = <0>;
> >>> +				remote = <&csi2_2>;
> >>> +				immutable;
> >>
> >> Did we decide "immutable" was actually needed? Presumably the driver for
> >> the HW in question knows the HW isn't configurable, and would simply not
> >> attempt to apply any configuration even if the .dts author erroneously
> >> provided some?
> > 
> > Well, I've been thinking about this today. I think, bridge drivers will 
> 
> Sorry, I'm not sure what a "bridge" driver is; is it any v4l2 device?

Right, sorry for not explaining. We call a bridge a device, that's sitting 
on a bus like USB, PCI or - as in the SoC case - on an internal one and 
interfacing to, say, a video sensor or a TV decoder or encoder. In the SoC 
case most primitive bridges just take data from video sensors attached 
over, say, an 8-bit parallel bus and DMA it into RAM buffers. Some bridges 
can also perform some data processing.

> > call centrally provided helper functions to enumerate links inside ports. 
> 
> Presumably a given driver would only parse the ports/links of its own DT
> node, and hence would be able to provide a parameter to any helper
> function that indicated the same information that "immutable" would?

Well, let's drop "immutable" for now, since we don't have a good idea 
whether we need it or not. If we do need it, we'll add it later, removing 
a part of a standart is more difficult.

> > While doing that they will want to differentiate between links to external 
> > devices with explicit configuration, and links to internal devices, whose 
> > configuration drivers might be able to figure out themselves. How should a 
> > driver find out what device this link is pointing to? Should it follow the 
> > "remote" phandle and then check the "compatible" property? The word 
> > "immutable" is a hint, that this is a link to an internal device, but it 
> > might either be unneeded or be transformed into something more 
> > informative.
> 
> I would imagine that a given driver would only ever parse its own DT
> node; the far end of any link is purely the domain of the other driver.
> I thought that each link node would contain whatever hsync-active,
> data-lanes, ... properties that were needed to configure the local device?

I think, information needed to configure a bridge device to connect to a 
camera sensor, like sync polarities etc., might not be sufficient to 
configure it to interface to an SoC internal unit, like a MIPI CSI-2 
interface. I can see 2 possibilities to recognise, that this link is going 
to an internal device: (1) follow the remote phandle, (2) use a 
vendor-specific property to specify the remote device type. I guess, you'd 
prefer the latter?

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Stephen Warren Sept. 12, 2012, 10:59 p.m. UTC | #6
On 09/12/2012 03:17 PM, Guennadi Liakhovetski wrote:
> On Wed, 12 Sep 2012, Stephen Warren wrote:
> 
>> On 09/12/2012 01:28 PM, Guennadi Liakhovetski wrote:
>>> Hi Stephen
>>>
>>> On Wed, 12 Sep 2012, Stephen Warren wrote:
>>>
>>>> On 09/11/2012 09:51 AM, Guennadi Liakhovetski wrote:
>>>>> This patch adds a document, describing common V4L2 device tree bindings.
>>>>>
>>>>> Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
>>>>> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
>>>>
>>>> Overall, I think this looks pretty reasonable, so:
>> ...
>>>>
>>>>> +			clock-frequency = <50000000>;	/* max clock frequency */
>>>>> +			clock-output-names = "mclk";
>>>>> +		};
>>>>> +
>>>>> +		port {
>>>> ...
>>>>> +			ceu0_0: link@0 {
>>>>> +				reg = <0>;
>>>>> +				remote = <&csi2_2>;
>>>>> +				immutable;
>>>>
>>>> Did we decide "immutable" was actually needed? Presumably the driver for
>>>> the HW in question knows the HW isn't configurable, and would simply not
>>>> attempt to apply any configuration even if the .dts author erroneously
>>>> provided some?
>>>
>>> Well, I've been thinking about this today. I think, bridge drivers will 
>>
>> Sorry, I'm not sure what a "bridge" driver is; is it any v4l2 device?
> 
> Right, sorry for not explaining. We call a bridge a device, that's sitting 
> on a bus like USB, PCI or - as in the SoC case - on an internal one and 
> interfacing to, say, a video sensor or a TV decoder or encoder. In the SoC 
> case most primitive bridges just take data from video sensors attached 
> over, say, an 8-bit parallel bus and DMA it into RAM buffers. Some bridges 
> can also perform some data processing.
> 
>>> call centrally provided helper functions to enumerate links inside ports. 
>>
>> Presumably a given driver would only parse the ports/links of its own DT
>> node, and hence would be able to provide a parameter to any helper
>> function that indicated the same information that "immutable" would?
> 
> Well, let's drop "immutable" for now, since we don't have a good idea 
> whether we need it or not. If we do need it, we'll add it later, removing 
> a part of a standart is more difficult.
> 
>>> While doing that they will want to differentiate between links to external 
>>> devices with explicit configuration, and links to internal devices, whose 
>>> configuration drivers might be able to figure out themselves. How should a 
>>> driver find out what device this link is pointing to? Should it follow the 
>>> "remote" phandle and then check the "compatible" property? The word 
>>> "immutable" is a hint, that this is a link to an internal device, but it 
>>> might either be unneeded or be transformed into something more 
>>> informative.
>>
>> I would imagine that a given driver would only ever parse its own DT
>> node; the far end of any link is purely the domain of the other driver.
>> I thought that each link node would contain whatever hsync-active,
>> data-lanes, ... properties that were needed to configure the local device?
> 
> I think, information needed to configure a bridge device to connect to a 
> camera sensor, like sync polarities etc., might not be sufficient to 
> configure it to interface to an SoC internal unit, like a MIPI CSI-2 
> interface. I can see 2 possibilities to recognise, that this link is going 
> to an internal device: (1) follow the remote phandle, (2) use a 
> vendor-specific property to specify the remote device type. I guess, you'd 
> prefer the latter?

I guess I'm still not exactly clear on what/why the bridge device needs
to know; a concrete example might help.

Either way though, (2) sounds like it's probably best; all information
required to configure a given device is within that device's node.

Being pedantic, I'm not sure precisely what you mean by vendor-specific.
If you mean the DT property name would include a vendor prefix, then
yes, I imagine that's quite possible, although it'd be good to try and
create standardized properties that multiple vendors and devices can use
if it makes sense. If you mean something more than that, then I'd argue
that those properties aren't just vendor-specific, but rather specific
to the individual binding for the individual device (which does imply
the vendor prefix, but not the general applicability of the property to
all of a vendor's devices).
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Guennadi Liakhovetski Sept. 25, 2012, 2:59 p.m. UTC | #7
Hi Sylwester

Ok, looks like there are no more comments coming, so, we can submit a new 
version:-)

Thanks for your comments:

On Tue, 11 Sep 2012, Sylwester Nawrocki wrote:

> Hi Guennadi,
> 
> On 09/11/2012 05:51 PM, Guennadi Liakhovetski wrote:
> > This patch adds a document, describing common V4L2 device tree bindings.
> > 
> > Co-authored-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
> > Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> > ---
> >  Documentation/devicetree/bindings/media/v4l2.txt |  143 ++++++++++++++++++++++
> >  1 files changed, 143 insertions(+), 0 deletions(-)
> >  create mode 100644 Documentation/devicetree/bindings/media/v4l2.txt
> > 
> > diff --git a/Documentation/devicetree/bindings/media/v4l2.txt b/Documentation/devicetree/bindings/media/v4l2.txt
> > new file mode 100644
> > index 0000000..55da6de
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/media/v4l2.txt
> > @@ -0,0 +1,143 @@
> > +Video4Linux Version 2 (V4L2)
> > +
> > +General concept:
> > +- video pipelines consist of external devices, e.g., camera sensors, controlled
> > +  over an I2C bus, and SoC internal IP blocks, including video DMA engines and
> > +  video data processors.
> > +- this document describes common bindings of all video pipeline devices.
> > +- SoC internal blocks are described by DT nodes, placed similarly to other SoC
> > +  blocks.
> > +- external devices are places on their respective control busses, e.g., I2C
> > +- data interfaces on all video devices are described by "port" child DT nodes
> > +- port configuration depends on other devices, participating in the data
> > +  transfer, and is described by "link" DT nodes, specified as children of
> > +  all "port" nodes, connected to this bus.
> > +- if a port can be configured to work with more than one other device on the
> > +  same bus, a "link" child DT node must be provided for each of them.
> > +- if more than one port is present on a device or more than one link is
> > +  connected to a port, a common scheme, using "#address-cells," "#size-cells"
> > +  and "reg" properties is used.
> 
> How about rephrasing/rearranging this to:
> 
> This document describes common bindings of all video pipeline devices.
> 
> General concept
> ---------------
> 
> Video pipelines consist of external devices, e.g. camera sensors, controlled
> over an I2C, SPI or UART bus, and SoC internal IP blocks, including video DMA
> engines and video data processors.
> SoC internal blocks are described by DT nodes, placed similarly to other SoC
> blocks. External devices are represented as child nodes of their respective
> bus controller nodes, e.g. I2C.

Well, I don't mind. I looked at a random file under 
Documentation/devicetree/bindings and used the same style, but we can use 
a free-flowing text too.

> Data interfaces on all video devices are described by "port" child DT nodes.
> Configuration of a port depends on other devices participating in the data
> transfer and is described by "link" DT nodes, specified as children of the
> "port" nodes:
> 
> /foo {
> 	port@0 {
> 		link@0 { ... };
> 		link@1 { ... };
> 	};
> 	port@1 { ... };
> };
> 
> If a port can be configured to work with more than one other device on the
> same bus, a "link" child DT node must be provided for each of them.
> If more than one port is present on a device or more than one link is
> connected to a port, a common scheme, using "#address-cells" "#size-cells"
> and "reg" properties is used.
> 
> > +Optional link properties:
> > +- remote: phandle to the other endpoint link DT node.
> > +- data-shift: on parallel data busses, if data-width is used to specify the
> 
> s/busses/buses

Nnnnooo... I think, "busses" is British and "buses" is American English, 
and I'm usually trying to follow the former, as much as I can:-) Also see 
"drivers/i2c/busses/"

> > +  number of data lines, data-shift can be used to specify which data lines are
> > +  used, e.g., "data-width=<10>; data-shift=<2>;" means, that lines 9:2 are used.
> > +- hsync-active: 1 or 0 for active-high or -low HSYNC signal polarity
> > +  respectively.
> > +- vsync-active: ditto for VSYNC. Note, that if HSYNC and VSYNC polarities are
> > +  not specified, embedded synchronisation may be required, where supported.
> > +- pclk-sample: rising (1) or falling (0) edge to sample the pixel clock pin.
> > +- immutable: used for SoC-internal links, if no configuration is required.
> > +- data-lanes: array of serial, e.g., MIPI CSI-2, data hardware lane numbers in
> 
> nit: s/e.g.,/e.g. ?

This is a good question, actually. Until now I've been following a 
suggestion by some well-educated British friend of mine, who told me to 
use commas on either side of "e.g." and "i.e." But in most textx I read I 
indeed see only one comma - before those. So, I tried to search and got a 
couple of hits, which suggest, that putting a comma on both sides or at 
least _after_ those abbreviations is more American grammar, and only 
before is more British... So, I think I'll follow your advise on this too 
and switch to the no-comma-after style, thanks.

> > +  the ascending order, beginning with logical lane 0.
> > +- clock-lanes: hardware lane number, used for the clock lane.
> 
> s/clock lane/clock ?

Well, to me both make sense. The meanings are a bit different, but none 
seems wrong to me. My version seems a bit more specific - not just "some" 
clock, but the CSI-2 clock lane?

> > +
> > +Example:
> > +
> > +	ceu0: ceu@0xfe910000 {
> > +		compatible = "renesas,sh-mobile-ceu";
> > +		reg = <0xfe910000 0xa0>;
> > +		interrupts = <0x880>;
> > +
> > +		mclk: master_clock {
> > +			compatible = "renesas,ceu-clock";
> > +			#clock-cells = <1>;
> > +			clock-frequency = <50000000>;	/* max clock frequency */
> > +			clock-output-names = "mclk";
> > +		};
> > +
> > +		port {
> > +			#address-cells = <1>;
> > +			#size-cells = <0>;
> > +
> > +			ceu0_1: link@1 {
> > +				reg = <1>;		/* local link # */
> > +				remote = <&ov772x_1_1>;	/* remote phandle */
> > +				bus-width = <8>;	/* used data lines */
> > +				data-shift = <0>;	/* lines 7:0 are used */
> > +	
> > +				/* If [hv]sync-active are missing, embedded bt.605 sync is used */
> > +				hsync-active = <1>;	/* active high */
> > +				vsync-active = <1>;	/* active high */
> > +				pclk-sample = <1>;	/* rising */
> > +			};
> > +
> > +			ceu0_0: link@0 {
> > +				reg = <0>;
> > +				remote = <&csi2_2>;
> > +				immutable;
> > +			};
> > +		};
> > +	};
> > +
> > +	i2c0: i2c@0xfff20000 {
> > +		...
> > +		ov772x_1: camera@0x21 {
> > +			compatible = "omnivision,ov772x";
> > +			reg = <0x21>;
> > +			vddio-supply = <&regulator1>;
> > +			vddcore-supply = <&regulator2>;
> > +
> > +			clock-frequency = <20000000>;
> > +			clocks = <&mclk 0>;
> > +			clock-names = "xclk";
> > +
> > +			port {
> > +				/* With 1 link per port no need in addresses */
> 
> s/in/for ?

Ehm, no idea either. Both "no need in" and "no need for" produce enough 
hits in search engines:-)

> > +				ov772x_1_1: link {
> > +					bus-width = <8>;
> > +					remote = <&ceu0_1>;
> > +					hsync-active = <1>;
> > +					hsync-active = <0>;	/* who came up with an inverter here?... */
> > +					pclk-sample = <1>;
> > +				};
> > +			};
> > +		};
> > +
> > +		imx074: camera@0x1a {
> > +			compatible = "sony,imx074";
> > +			reg = <0x1a>;
> > +			vddio-supply = <&regulator1>;
> > +			vddcore-supply = <&regulator2>;
> > +
> > +			clock-frequency = <30000000>;	/* shared clock with ov772x_1 */
> > +			clocks = <&mclk 0>;
> > +			clock-names = "sysclk";		/* assuming this is the name in the datasheet */
> > +
> > +			port {
> > +				imx074_1: link {
> > +					clock-lanes = <0>;
> > +					data-lanes = <1>, <2>;
> > +					remote = <&csi2_1>;
> > +				};
> > +			};
> > +		};
> > +	};
> > +
> > +	csi2: csi2@0xffc90000 {
> > +		compatible = "renesas,sh-mobile-csi2";
> > +		reg = <0xffc90000 0x1000>;
> > +		interrupts = <0x17a0>;
> > +		#address-cells = <1>;
> > +		#size-cells = <0>;
> > +
> > +		port@1 {
> > +			compatible = "renesas,csi2c";	/* one of CSI2I and CSI2C */
> > +			reg = <1>;			/* CSI-2 PHY #1 of 2: PHY_S, PHY_M has port address 0, is unused */
> > +
> > +			csi2_1: link {
> > +				clock-lanes = <0>;
> > +				data-lanes = <2>, <1>;
> > +				remote = <&imx074_1>;
> > +			};
> > +		};
> > +		port@2 {
> > +			reg = <2>;			/* port 2: link to the CEU */
> > +
> > +			csi2_2: link {
> > +				immutable;
> > +				remote = <&ceu0_0>;
> > +			};
> > +		};
> > +	};
> 
> --
> 
> Regards,
> Sylwester
> 

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/Documentation/devicetree/bindings/media/v4l2.txt b/Documentation/devicetree/bindings/media/v4l2.txt
new file mode 100644
index 0000000..55da6de
--- /dev/null
+++ b/Documentation/devicetree/bindings/media/v4l2.txt
@@ -0,0 +1,143 @@ 
+Video4Linux Version 2 (V4L2)
+
+General concept:
+- video pipelines consist of external devices, e.g., camera sensors, controlled
+  over an I2C bus, and SoC internal IP blocks, including video DMA engines and
+  video data processors.
+- this document describes common bindings of all video pipeline devices.
+- SoC internal blocks are described by DT nodes, placed similarly to other SoC
+  blocks.
+- external devices are places on their respective control busses, e.g., I2C
+- data interfaces on all video devices are described by "port" child DT nodes
+- port configuration depends on other devices, participating in the data
+  transfer, and is described by "link" DT nodes, specified as children of
+  all "port" nodes, connected to this bus.
+- if a port can be configured to work with more than one other device on the
+  same bus, a "link" child DT node must be provided for each of them.
+- if more than one port is present on a device or more than one link is
+  connected to a port, a common scheme, using "#address-cells," "#size-cells"
+  and "reg" properties is used.
+
+Optional link properties:
+- remote: phandle to the other endpoint link DT node.
+- data-shift: on parallel data busses, if data-width is used to specify the
+  number of data lines, data-shift can be used to specify which data lines are
+  used, e.g., "data-width=<10>; data-shift=<2>;" means, that lines 9:2 are used.
+- hsync-active: 1 or 0 for active-high or -low HSYNC signal polarity
+  respectively.
+- vsync-active: ditto for VSYNC. Note, that if HSYNC and VSYNC polarities are
+  not specified, embedded synchronisation may be required, where supported.
+- pclk-sample: rising (1) or falling (0) edge to sample the pixel clock pin.
+- immutable: used for SoC-internal links, if no configuration is required.
+- data-lanes: array of serial, e.g., MIPI CSI-2, data hardware lane numbers in
+  the ascending order, beginning with logical lane 0.
+- clock-lanes: hardware lane number, used for the clock lane.
+
+Example:
+
+	ceu0: ceu@0xfe910000 {
+		compatible = "renesas,sh-mobile-ceu";
+		reg = <0xfe910000 0xa0>;
+		interrupts = <0x880>;
+
+		mclk: master_clock {
+			compatible = "renesas,ceu-clock";
+			#clock-cells = <1>;
+			clock-frequency = <50000000>;	/* max clock frequency */
+			clock-output-names = "mclk";
+		};
+
+		port {
+			#address-cells = <1>;
+			#size-cells = <0>;
+
+			ceu0_1: link@1 {
+				reg = <1>;		/* local link # */
+				remote = <&ov772x_1_1>;	/* remote phandle */
+				bus-width = <8>;	/* used data lines */
+				data-shift = <0>;	/* lines 7:0 are used */
+	
+				/* If [hv]sync-active are missing, embedded bt.605 sync is used */
+				hsync-active = <1>;	/* active high */
+				vsync-active = <1>;	/* active high */
+				pclk-sample = <1>;	/* rising */
+			};
+
+			ceu0_0: link@0 {
+				reg = <0>;
+				remote = <&csi2_2>;
+				immutable;
+			};
+		};
+	};
+
+	i2c0: i2c@0xfff20000 {
+		...
+		ov772x_1: camera@0x21 {
+			compatible = "omnivision,ov772x";
+			reg = <0x21>;
+			vddio-supply = <&regulator1>;
+			vddcore-supply = <&regulator2>;
+
+			clock-frequency = <20000000>;
+			clocks = <&mclk 0>;
+			clock-names = "xclk";
+
+			port {
+				/* With 1 link per port no need in addresses */
+				ov772x_1_1: link {
+					bus-width = <8>;
+					remote = <&ceu0_1>;
+					hsync-active = <1>;
+					hsync-active = <0>;	/* who came up with an inverter here?... */
+					pclk-sample = <1>;
+				};
+			};
+		};
+
+		imx074: camera@0x1a {
+			compatible = "sony,imx074";
+			reg = <0x1a>;
+			vddio-supply = <&regulator1>;
+			vddcore-supply = <&regulator2>;
+
+			clock-frequency = <30000000>;	/* shared clock with ov772x_1 */
+			clocks = <&mclk 0>;
+			clock-names = "sysclk";		/* assuming this is the name in the datasheet */
+
+			port {
+				imx074_1: link {
+					clock-lanes = <0>;
+					data-lanes = <1>, <2>;
+					remote = <&csi2_1>;
+				};
+			};
+		};
+	};
+
+	csi2: csi2@0xffc90000 {
+		compatible = "renesas,sh-mobile-csi2";
+		reg = <0xffc90000 0x1000>;
+		interrupts = <0x17a0>;
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		port@1 {
+			compatible = "renesas,csi2c";	/* one of CSI2I and CSI2C */
+			reg = <1>;			/* CSI-2 PHY #1 of 2: PHY_S, PHY_M has port address 0, is unused */
+
+			csi2_1: link {
+				clock-lanes = <0>;
+				data-lanes = <2>, <1>;
+				remote = <&imx074_1>;
+			};
+		};
+		port@2 {
+			reg = <2>;			/* port 2: link to the CEU */
+
+			csi2_2: link {
+				immutable;
+				remote = <&ceu0_0>;
+			};
+		};
+	};