On Wednesday, June 21st, 2023 at 14:26, Jani Nikula <jani.nikula@xxxxxxxxx> wrote: > On Wed, 21 Jun 2023, Simon Ser contact@xxxxxxxxxxx wrote: > > > On Wednesday, June 21st, 2023 at 14:11, Jani Nikula jani.nikula@xxxxxxxxx wrote: > > > > > On Wed, 21 Jun 2023, Simon Ser contact@xxxxxxxxxxx wrote: > > > > > > > On Wednesday, June 21st, 2023 at 14:05, Jani Nikula jani.nikula@xxxxxxxxx wrote: > > > > > > > > > > - if (changed) > > > > > > + if (hweight32(changed) == 1) > > > > > > + drm_kms_helper_connector_hotplug_event(first_changed_connector); > > > > > > > > > > What if more than one connector share the same hpd pin? > > > > > > > > Ah, I did not believe this could happen. I'll rework the patch to > > > > count the number of changed connectors instead. > > > > > > A DP++ port is probably the prime example of this, with both DP and HDMI > > > connectors. > > > > Wouldn't that be handled by the separate DP logic though? (The > > drm_dp_mst family of helpers.) > > It's not DP MST, though. See intel_ddi_init() initializing both DP and > HDMI. One encoder, one HPD pin, two connectors. Thanks for the explanation!
