Hi Dario,
On Thu, Oct 29, 2026 at 03:41:28PM +0100, Dario Binacchi wrote:
> of_drm_find_and_get_bridge() returns NULL if the panel has not been
> registered yet. Returning -ENODEV in this case makes the probe fail
> instead of being retried, so the LDB is never bound and the display
> pipeline never comes up:
>
> platform 32e90000.display-controller: deferred probe pending: imx-lcdif:
> Failed to get bridge for endpoint0
>
> Return -EPROBE_DEFER, so that the probe is retried once the panel is
> available.
>
> Fixes: 802372cf316a ("drm/bridge: fsl-ldb: don't create a panel_bridge")
> Signed-off-by: Dario Binacchi <[email protected]>
>
> ---
>
> drivers/gpu/drm/bridge/fsl-ldb.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/bridge/fsl-ldb.c
> b/drivers/gpu/drm/bridge/fsl-ldb.c
> index 7ec84c81cd78..420b3e08130e 100644
> --- a/drivers/gpu/drm/bridge/fsl-ldb.c
> +++ b/drivers/gpu/drm/bridge/fsl-ldb.c
> @@ -342,7 +342,7 @@ static int fsl_ldb_probe(struct platform_device *pdev)
> fsl_ldb->bridge.next_bridge = of_drm_find_and_get_bridge(panel_node);
> of_node_put(panel_node);
> if (!fsl_ldb->bridge.next_bridge)
> - return -ENODEV;
> + return -EPROBE_DEFER;
Hmm, I have a patch [1] in the DCIF series that adds support for non-panel
bridges to fsl-ldb.c which, indirectly, should fix this as well. But, since I
don't see that patchset merged anytime soon:
Reviewed-by: Laurentiu Palcu <[email protected]>
[1]
https://lore.kernel.org/dri-devel/[email protected]/
--
Thanks,
Laurentiu
