On 20/06/2020 06:56, Jeff Chase wrote:

quoted

quoted

quoted

quoted

This particular chip can actually be used both with DMI matching but also
on an ARM with device tree, but since you can't test this on an ARM board,
there is no point in adding support for that.

However, compared to the cros-ec and seco drivers you can do something a bit
different here: those drivers just return -ENODEV if there is no match, but
since this driver reads the EDID it can just continue as long as it does not
set the CEC_CAP_CONNECTOR_INFO capability.

Hi Hans,

After looking into this I don't think I want to actually register the
cec adapter with the hdmi port's cec notifier because then it will
start receiving physical address updates from the port. Since the
driver already updates the physical address itself it seems like that
would cause some issues.

I'm looking at just adding the connector to the adapter directly using
`cec_s_conn_info()`. But it doesn't look like I can get the conn_info
without registering with the notifier. Do you have any thoughts here
before I try tackling changes to the cec framework?

It's a very good point, I hadn't realized this complication.

The notifier support is definitely required for this, but there should be
a way to prevent the notifier from updating the physical address since it
is controlled by the CEC adapter instead.

Suggestion:

Add a 'bool adap_controls_phys_addr;' field to struct cec_adapter. If set to
true, then cec-notifier will skip the cec_s_phys_addr calls.

This driver then just has to set this field to true and it should work.

Regards,

	Hans