Hi,
On Sun, Dec 01, 2024 at 02:44:13AM +0200, Dmitry Baryshkov wrote:
> Use the helper function to update the connector's information. This
> makes sure that HDMI-related events are handled in a generic way.
> Currently it is limited to the HDMI state reporting to the sound system.
>
> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxx>
> ---
> drivers/gpu/drm/vc4/vc4_hdmi.c | 9 +++++++--
> 1 file changed, 7 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c
> index d0a9aff7ad43016647493263c00d593296a1e3ad..d83f587ab69f4b8f7d5c37a00777f11da8301bc1 100644
> --- a/drivers/gpu/drm/vc4/vc4_hdmi.c
> +++ b/drivers/gpu/drm/vc4/vc4_hdmi.c
> @@ -401,13 +401,16 @@ static void vc4_hdmi_handle_hotplug(struct vc4_hdmi *vc4_hdmi,
> */
>
> if (status == connector_status_disconnected) {
> + drm_atomic_helper_connector_hdmi_update_edid(connector, NULL);
> cec_phys_addr_invalidate(vc4_hdmi->cec_adap);
> return;
> }
>
> drm_edid = drm_edid_read_ddc(connector, vc4_hdmi->ddc);
>
> - drm_edid_connector_update(connector, drm_edid);
> + // TODO: use drm_atomic_helper_connector_hdmi_update() once it gains
> + // CEC support
> + drm_atomic_helper_connector_hdmi_update_edid(connector, drm_edid);
So, it's not just about EDID, and I think we shouldn't really focus on
that either.
As that patch points out, even if we only consider EDID's, we have
different path depending on the connector status. It shouldn't be up to
the drivers to get this right.
What I had in mind was something like a
drm_atomic_helper_connector_hdmi_hotplug function that would obviously
deal with EDID only here, but would expand to CEC, scrambling, etc.
later on.
And would cover both the connected/disconnected cases.
Maxime
Attachment:
signature.asc
Description: PGP signature
