Preliminary proposal, new APIs for HDMI and DVI control in v4l2

This is a preliminary proposal for an extension to the v4l2 api. To be discussed at the V4L2 'brainstorming' meeting in Warsaw, March 2011

Purpose: Provide basic controls for HDMI and DVI devices.

sink : HDMI/DVI/VGA receiver
source: HDMI/DVI/VGA transmitter

New Controls:
=============

These controls rely on control events to be able to notify the application of any change.

The idea is to create an event that will be triggered by the control framework whenever a control changes value.

TX source:
----------

Control: V4L2_CID_DV_TX_HOTPLUG
type: bitmask (output), read only
name: Hotplug
Hotplug is present on each output.

The hotplug is issued by the sink to indicate that an edid exists and should be read by the source. Not applicable for VGA connectors.

Control: V4L2_CID_DV_TX_RXSENSE
type: bitmask (output), read only
name: Rx Sense
Rx sense is present on each output.

Rx Sense is detection of pull-ups on the TMDS clock lines. Normally means that the sink has left/entered standby. Not applicable for VGA connectors.

Control: V4L2_CID_DV_TX_EDID_SEGMENT0_PRESENT
type: bitmask (output), read only
name: Edid segment 0

The source driver has read edid segment 0 from the sink.

Control: V4L2_CID_DV_TX_DVI_HDMI_MODE
type: menu, read and write
name: Hdmi/Dvi mode

Settings:
"HDMI"
"DVI-D"

RX sink:
--------

Control: V4L2_CID_DV_RX_5V
type: bitmask (input), read only
name: Rx 5V

The source must supply the sink with a +5v via the HDMI/DVI cable. This is often used to power an eeprom which contains edid information, such that the source can read the edid even if the sink
is in standby/power off.

Control: V4L2_CID_DV_CABLE_DETECT
type: bitmask (input), read only
name: Cable detect

This is not a part of the HDMI/DVI standards.
But many sinks (read monitors) uses a + 5v pull-up on a ground pin (typically pin 15 on dvi) to detect if a cable is is connected and terminated at the source end.

Control: V4L2_CID_DV_FORMAT_STATUS
type: menu, read only
name: Format status

Format status for the current selected input.
Statuses:
"No signal"
"Unknown"
"Unsupported"
"Encrypted"
"Valid/supported"

The input signal has changed format or status.
In the case of a valid signal use the
vidioc_query_dv_preset to read the new format.

Alternatively this could also be implemented as a bitmask with bits
for: 'signal present', 'encrypted', 'supported format'.

Or as three separate boolean controls.

New Events:
===========

V4L2_EVENT_TX_EDID

       struct v4l2_event_tx_edid {
              __u32 output;
              __u32 present;
              __u32 segment;
       }

The source driver has read an edid segment from the sink. One segment is 256 bytes and will contain one or two 128 bytes blocks. Most HDMI and DVI sinks will have only segment 0.

The driver should read segment 0 when hotplug is detected and report this to userspace. The application can then use the V4L2_G_EDID ioctl to get the actual edid data.

If the EDID data indicates the presence of further extension blocks the userspace application can use the V4L2_REQ_EDID ioctl to trigger a read. When the driver has read that extended segment it will generate a V4L2_EVENT_TX_EDID event.

New Ioctls:
===========

TX source:
----------

V4L2_REQ_EDID

       struct v4l2_req_edid {
              __u32 output;
              __u32 segment;
        }

Trigger a read of a 256 bytes segment. Normally the driver will read segment 0 by itself when hotplug is detected, But this ioctl may also be used to trigger a re-read of segment 0.

V4L2_G_EDID

       struct v4l2_g_edid {
              __u32 output;
              __u32 segment;
              __u8 edid[256];
        }

Get the edid segment indicated in the V4L2_EVENT_TX_EDID

RX sink:
--------

V4L2_S_EDID

       struct v4l2_s_edid {
              __u32 input;
              __u32 segment;
              __u8 _edid[256];
        }

Set the edid information in the source.

Additional Controls and Status:
===============================

Controls and statuses we want to add. Currently based on HDMI version 1.3a, must be updated

From Info frames

* Y0, Y1 RGB or YCBCR indicator.
* A0 Active Information Present. Indicates whether field R0...R3 is valid. See
CEA-861-D table 8 for details.
* B0, B1 Bar Info data valid. See CEA-861-D table 8 for details.
HDMI Licensing, LLC Page 113 of 156 High-Definition Multimedia Interface
Specification Version 1.3a
* S0, S1 Scan Information (i.e. overscan, underscan). See CEA-861-D table 8 for
details.
* C0, C1 Colorimetry (ITU BT.601, BT.709 etc.). See CEA-861-D table 9 for
details.
* EC0, EC1, EC2 Extended Colorimetry (IEC 61966-2-4 etc.). See CEA-861-D table 11 for
details.
* Q1, Q0 Quantization range (Full vs. Limited, etc.). See CEA-861-D table 11 for
details.

* ITC               IT Content. See CEA-861-D table 11 for details.
* M0, M1            Picture Aspect Ratio (4:3, 16:9). See CEA-861-D table 9 for details.
* R0...R3           Active Format Aspect Ratio. See CEA-861-D table 10 and Annex H for

details.
* VIC0...VIC6 Video Format Identification Code. When transmitting any video format in
section 6.2.4, above, an HDMI Source shall set
the VIC field to the Video
Code for that format. See CEA-861-D section 6.4
for details.
* PR0...PR3 Pixel Repetition factor. See CEA-861-D table 13 for details.
* SC1, SC0 Non-uniform Picture Scaling. See CEA-861-D table 11.

A lot of these handle colorspace information. It is not clear yet whether this should be set up using S_FMT/mediabus APIs, or by explicit controls.

Martin Bugge

reposting the query since the earlier post did not appear in mailing list.

Hi Martin,
We are also in requirement of these controls as described by you. I did a
search in the archives but could not find a suitable conclusion to the RFC. I could find that the dv_timings structure has been modified as a result of further discussions. But for many items like S_EDID, DV_CABLE_DETECT, Info frames etc , I could not find the logical conclusion to this RFC. Could please let me know the further updates on these requirements?

Thanks in Advance
Best Regards
Soby Mathew

Hi Soby

It has been on hold for a very long time, but just last week Hans Verkuil posted a RFC which is a follow up on this subject.

http://www.spinics.net/lists/linux-media/msg47671.html

So that thread has taken over.

Best regards
Martin Bugge

And expect to see a small patch series to be posted in the next few days containing a slightly changed version of the proposal, including DocBook documentation.

Please review those patches when they are posted and let us know if you are OK with it or if you have additional requirements.

Regards,

Hans