[PATCH v3 12/13] drm: bridge: dw_hdmi: Use dw_hdmi_connector_status_update()

[Jonas Karlman]

Hi Nicolas,

On 4/28/2026 1:53 PM, Nicolas Frattaroli wrote:
> On Friday, 3 April 2026 20:52:58 Central European Summer Time Jonas Karlman wrote:
>> Update connector EDID and CEC phys addr from detect and force funcs to
>> ensure that userspace always have access to latest read EDID after a
>> sink use a HPD low voltage pulse to indicate that EDID has changed.
>>
>> With EDID being updated in detect and force funcs, there should no
>> longer be a need to re-read EDID in get_modes funcs, so drop it.
>>
>> This change make the dw-hdmi connector work more closely like the bridge
>> connector does with a hdmi bridge.
>>
>> Signed-off-by: Jonas Karlman <jonas at kwiboo.se>
>> ---
>> v3: Reworked 'Update EDID during hotplug processing' patch
>> ---
>>  drivers/gpu/drm/bridge/synopsys/dw-hdmi.c | 24 +++++++++++++----------
>>  1 file changed, 14 insertions(+), 10 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>> index 0d42fdf9a386..5b5654ef6015 100644
>> --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>> +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>> @@ -2474,33 +2474,36 @@ dw_hdmi_connector_status_update(struct drm_connector *connector,
>>  	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi, connector);
>>  	const struct drm_edid *drm_edid;
>>  
>> +	if (status == connector_status_disconnected) {
>> +		drm_edid_connector_update(connector, NULL);
>> +		cec_notifier_phys_addr_invalidate(hdmi->cec_notifier);
>> +		return;
>> +	}
>> +
>>  	drm_edid = dw_hdmi_edid_read(hdmi, connector);
>>  	drm_edid_connector_update(connector, drm_edid);
>>  	drm_edid_free(drm_edid);
>>  
>> -	cec_notifier_set_phys_addr(hdmi->cec_notifier,
>> -				   connector->display_info.source_physical_address);
>> +	if (status == connector_status_connected)
>> +		cec_notifier_set_phys_addr(hdmi->cec_notifier,
>> +				connector->display_info.source_physical_address);
>>  }
>>  
>>  static enum drm_connector_status
>>  dw_hdmi_connector_detect(struct drm_connector *connector, bool force)
>>  {
>> -	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi,
>> -					     connector);
>> +	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi, connector);
>>  	enum drm_connector_status status;
>>  
>>  	status = dw_hdmi_detect(hdmi);
>>  
>> -	if (status == connector_status_disconnected)
>> -		cec_notifier_phys_addr_invalidate(hdmi->cec_notifier);
>> +	dw_hdmi_connector_status_update(connector, status);
>>  
>>  	return status;
>>  }
>>  
>>  static int dw_hdmi_connector_get_modes(struct drm_connector *connector)
>>  {
>> -	dw_hdmi_connector_status_update(connector, connector->status);
>> -
>>  	return drm_edid_connector_add_modes(connector);
>>  }
> 
> Are we absolutely sure status_update is called before this, so that the
> EDID doesn't need to be re-read? E.g. drm_connector_helper_get_modes
> does do an EDID read beforehand.

With the change to re-read EDID at .detect() we should not need to
re-read EDID at .get_modes(). This also closely match how the bridge
connector handle EDID update at .detect() and not .get_modes() for hdmi
bridges.

At least following call paths should ensure EDID have been re-read:

drm_helper_probe_single_connector_modes():
 -> connector->funcs->force()
    -> dw_hdmi_connector_force()
       -> dw_hdmi_connector_status_update()
          -> drm_edid_connector_update()             <<-- here
 -> drm_helper_probe_detect()
    -> detect_connector_status()
       -> connector->funcs->detect();
          -> dw_hdmi_connector_detect()
             -> dw_hdmi_connector_status_update()
                -> drm_edid_connector_update()       <<-- or here
 -> drm_helper_probe_get_modes()
    -> connector_funcs->get_modes()
       -> dw_hdmi_connector_get_modes()              <<-- now updated

drm_helper_hpd_irq_event():
 -> detect_connector_status()
    -> connector->funcs->detect();
       -> dw_hdmi_connector_detect()
          -> dw_hdmi_connector_status_update()
             -> drm_edid_connector_update()          <<-- also after HPD

> Speaking of which: if we had connector->ddc set, we could just use
> drm_connector_helper_get_modes for this callback. But maybe it's not
> worth spending the time refactoring the resource handling around
> hdmi->ddc when the eventual goal is to refactor this all into using
> a bridge connector anyway.

Correct, hopefully all connector related functions can be dropped from
dw-hdmi in a near future when dw-hdmi is converted to a hdmi bridge.

I currently have a work-in-progress snapshot at [1] where dw-hdmi have
been converted to a hdmi bridge and a bridge connector is used instead.

The changes in this series only aim to align dw-hdmi to the hdmi bridge
where possible to reduce any functional change in such future series.

I will likely keep this as-is in a v4 unless there is some objections.

[1] https://github.com/Kwiboo/linux-rockchip/commits/next-20260427-rk-hdmi-v2/

Regards,
Jonas

>>  
>> @@ -2530,14 +2533,15 @@ static int dw_hdmi_connector_atomic_check(struct drm_connector *connector,
>>  
>>  static void dw_hdmi_connector_force(struct drm_connector *connector)
>>  {
>> -	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi,
>> -					     connector);
>> +	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi, connector);
>>  
>>  	mutex_lock(&hdmi->mutex);
>>  	hdmi->force = connector->force;
>>  	hdmi->last_connector_result = connector->status;
>>  	dw_hdmi_update_phy_mask(hdmi);
>>  	mutex_unlock(&hdmi->mutex);
>> +
>> +	dw_hdmi_connector_status_update(connector, connector->status);
>>  }
>>  
>>  static const struct drm_connector_funcs dw_hdmi_connector_funcs = {
>>
> 
> 
> 
>