On Fri, Jul 25, 2025 at 10:33:58AM +0530, Suraj Kandpal wrote:
> Define function to destroy the drm_writbeack_connector and
> drm_connector associated with it.
>
> Signed-off-by: Suraj Kandpal <suraj.kand...@intel.com>
> ---
> drivers/gpu/drm/i915/display/intel_writeback.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_writeback.c
> b/drivers/gpu/drm/i915/display/intel_writeback.c
> index def33191a89e..9b2432d86d35 100644
> --- a/drivers/gpu/drm/i915/display/intel_writeback.c
> +++ b/drivers/gpu/drm/i915/display/intel_writeback.c
> @@ -180,6 +180,12 @@ intel_writeback_detect(struct drm_connector *connector,
> return connector_status_connected;
> }
>
> +static void intel_writeback_connector_destroy(struct drm_connector
> *connector)
> +{
> + drm_connector_cleanup(connector);
> + kfree(connector);
> +}
Nice example of what I've written in my response to the cover letter:
without this commit we have a memory leak here, don't we?
> +
> static struct drm_writeback_connector *
> intel_get_writeback_connector(struct drm_connector *connector)
> {
> @@ -208,6 +214,7 @@ const struct drm_connector_funcs conn_funcs = {
> .fill_modes = drm_helper_probe_single_connector_modes,
> .atomic_duplicate_state = intel_digital_connector_duplicate_state,
> .atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
> + .destroy = intel_writeback_connector_destroy,
> };
>
> static const struct drm_connector_helper_funcs conn_helper_funcs = {
> --
> 2.34.1
>
--
With best wishes
Dmitry
