Re: [PATCH v2 3/7] drm/bridge: dw-mipi-dsi: stop using deprecated devm_drm_of_get_bridge()

["Luca Ceresoli" <luca.ceresoli@bootlin.com>]

On Tue Sep 29, 2026 at 3:55 PM CEST, sashiko-bot wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] drm: bridge reference leak on dw_mipi_dsi_host_detach early return
> --
>
> commit b7207f37872ee6fe7eb2903e38f89975b930c392
> Author: Luca Ceresoli <luca.ceresoli@bootlin.com>
>
> drm/bridge: dw-mipi-dsi: stop using deprecated devm_drm_of_get_bridge()
>
> This commit switches from the deprecated devm_drm_of_get_bridge() to
> of_drm_get_bridge_by_endpoint() to resolve a build failure when CONFIG_OF
> is disabled. Since of_drm_get_bridge_by_endpoint() returns a refcounted
> bridge, it adds the necessary put operations to release the reference.
>
> Link: https://lore.kernel.org/r/20260928161605.2271412-1-arnd@kernel.org
> Link: https://lore.kernel.org/oe-kbuild-all/202609291109.9RGna4XZ-lkp@intel.com/
>
>> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c b/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
>> index 5dd475fcaacf8..22d1b8cbe1d57 100644
>> --- a/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
>> +++ b/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
> [ ... ]
>> @@ -368,9 +368,8 @@ static int dw_mipi_dsi_host_detach(struct mipi_dsi_host *host,
>>  			return ret;
>
> [Severity: Medium]
> If pdata->host_ops->detach() returns a negative error code (e.g., during
> device unbind or PHY teardown errors) and takes this early return, does this
> leak the bridge reference acquired in the attach function?
>
> Since the bridge is now dynamically refcounted and we skip the newly added
> drm_bridge_clear_and_put(&dsi->bridge.next_bridge), it looks like the
> reference count will not be properly decremented on this error path.

The issue was there already for drm_of_panel_bridge_remove() (here deleted)
and drm_bridge_remove().

Id say it's orthogonal, and most likely nobody has this issue in real
life. So I took a note to fix this soon, and I'd let this patch be applied
as is because it fixes an issue bothering people.

Luca

--
Luca Ceresoli, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com