Re: [PATCH v4 14/20] drm/hdmi: Rename __drm_atomic_helper_connector_hdmi_reset()

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Unchecked use of `connector->state` after `drm_atomic_helper_connector_reset()` leads to a NULL pointer dereference on state allocation failure.
- [Low] Extraneous space before the period in the kernel-doc block.
--

commit 4088843a1b581a7e32488e55861fe134c72198dc
Author: Maxime Ripard <mripard@kernel.org>
drm/hdmi: Rename __drm_atomic_helper_connector_hdmi_reset()

This commit renames the HDMI connector reset helper to reflect its new role
as a general state initialization function. It updates the callers in various
drivers and tests to use the new name.

> diff --git a/drivers/gpu/drm/display/drm_bridge_connector.c b/drivers/gpu/drm/display/drm_bridge_connector.c
> --- a/drivers/gpu/drm/display/drm_bridge_connector.c
> +++ b/drivers/gpu/drm/display/drm_bridge_connector.c
[ ... ]
> @@ -272,8 +272,8 @@ static void drm_bridge_connector_reset(struct drm_connector *connector)
>  
>  	drm_atomic_helper_connector_reset(connector);
>  	if (bridge_connector->bridge_hdmi)
> -		__drm_atomic_helper_connector_hdmi_reset(connector,
> -							 connector->state);
> +		__drm_atomic_helper_connector_hdmi_state_init(connector,
> +							      connector->state);

If drm_atomic_helper_connector_reset() fails to allocate memory, it sets
connector->state to NULL. Does passing connector->state immediately to the
initialization function lead to a NULL pointer dereference?

While a later patch in this series converts the bridge_connector code to safely
handle allocation failures via atomic_create_state, the sun4i and vc4 drivers
modified in this patch share this exact pattern. Could this leave those drivers
vulnerable to crashes under memory pressure?

> diff --git a/drivers/gpu/drm/display/drm_hdmi_state_helper.c b/drivers/gpu/drm/display/drm_hdmi_state_helper.c
> --- a/drivers/gpu/drm/display/drm_hdmi_state_helper.c
> +++ b/drivers/gpu/drm/display/drm_hdmi_state_helper.c
[ ... ]
>  /**
> - * __drm_atomic_helper_connector_hdmi_reset() - Initializes all HDMI @drm_connector_state resources
> + * __drm_atomic_helper_connector_hdmi_state_init() - Initialize all HDMI @drm_connector_state resources
>   * @connector: DRM connector
> - * @new_conn_state: connector state to reset
> + * @new_conn_state: connector state to initialize
>   *
>   * Initializes all HDMI resources from a @drm_connector_state without
>   * actually allocating it. This is useful for HDMI drivers, in
> - * combination with __drm_atomic_helper_connector_reset() or
> - * drm_atomic_helper_connector_reset().
> + * combination with __drm_atomic_helper_connector_state_init(),
> + * drm_atomic_helper_connector_reset(), or
> + * drm_atomic_helper_connector_create_state() .

This isn't a bug, but there is an extraneous space before the period.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260512-drm-mode-config-init-v4-0-591dfdcc1bf9@kernel.org?part=14