On Thu, Dec 22, 2022 at 09:18:04PM +0100, Maarten Lankhorst wrote: > When intel_dp_init_connector fails, some power wells used in dp aux > communication may not be completely disabled yet. This may result in a > null pointer dereference when icl_aux_pw_to_phy() is called from > icl_combo_phy_aux_power_well_disable() after the encoder and connector > are already freed. > > Signed-off-by: Maarten Lankhorst <maarten.lankhorst@xxxxxxxxxxxxxxx> > Cc: Imre Deak <imre.deak@xxxxxxxxx> > Cc: Jani Nikula <jani.nikula@xxxxxxxxx> > --- > This approach better? Looks ok to me, thanks for fixing it. I think Jani's idea is good to refactor things wrt. the encoder hooks, but imo that could be done later keeping the fix simpler, so: Acked-by: Imre Deak <imre.deak@xxxxxxxxx> > > drivers/gpu/drm/i915/display/intel_dp.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c > index bf80f296a8fd..8cf1d6ca86f4 100644 > --- a/drivers/gpu/drm/i915/display/intel_dp.c > +++ b/drivers/gpu/drm/i915/display/intel_dp.c > @@ -5492,6 +5492,7 @@ intel_dp_init_connector(struct intel_digital_port *dig_port, > return true; > > fail: > + intel_display_power_flush_work(dev_priv); > drm_connector_cleanup(connector); > > return false; > -- > 2.37.2 >
