Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [High] drm/bridge: dw-dp: permanent probe deferral in dw_dp_probe() on
  RK3588 boards
--

commit 75b0f7632b89aacf0c388d31b5dbe1c16aadde94
Author: Sebastian Reichel <[email protected]>

drm/bridge: synopsys: dw-dp: Add follow-up bridge support

This commit acquires the downstream DRM bridge connected to port 1 endpoint 0
during dw_dp_probe() and attaches it in dw_dp_bridge_attach() to support USB-C
DP altmode bridge chains.

> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-dp.c 
> b/drivers/gpu/drm/bridge/synopsys/dw-dp.c
> index 5db9063ee97e0..74ba1306caee6 100644
> --- a/drivers/gpu/drm/bridge/synopsys/dw-dp.c
> +++ b/drivers/gpu/drm/bridge/synopsys/dw-dp.c

[ ... ]

> @@ -2156,6 +2179,16 @@ struct dw_dp *dw_dp_probe(struct platform_device 
> *pdev, const struct dw_dp_plat_
>       if (ret)
>               return ERR_PTR(ret);
>  
> +     dp->next_bridge = of_drm_get_bridge_by_endpoint(dev->of_node, 1, 0);
> +     if (IS_ERR(dp->next_bridge)) {
> +             dev_err_probe(dev, PTR_ERR(dp->next_bridge), "failed to get 
> follow-up bridge\n");
> +             return ERR_CAST(dp->next_bridge);
> +     }

[Severity: High]
Does requiring dp->next_bridge cause dw_dp_probe() to defer forever on RK3588
boards where usbdp_phy0 uses a single port node?

In rk3588-lubancat-5io.dts and rk3588s-indiedroid-nova.dts, dp0_out_con
connects to usbdp_phy0_dp_in (endpoint@2) under a single port subnode on
usbdp_phy0 rather than a multi-port graph with port@3.  However,
rk_udphy_probe() only registers the auxiliary DRM bridge when port 3
endpoint 0 is present:

drivers/phy/rockchip/phy-rockchip-usbdp.c:rk_udphy_probe() {
    ...
        dp_aux_ep = fwnode_graph_get_endpoint_by_id(dev_fwnode(dev), 3, 0, 0);
        if (dp_aux_ep) {
                ret = drm_aux_bridge_register(dev);
    ...
}

On those boards, of_drm_get_bridge_by_endpoint() resolves the remote node to
usbdp_phy0, finds no registered drm_bridge, and returns -EPROBE_DEFER
indefinitely.  Because rockchip_drm_match_add() adds every enabled dw-dp
device as a required component, this prevents the entire rockchip-drm display
subsystem from binding.

-- 
Sashiko AI review ยท 
https://sashiko.dev/#/patchset/20261114-synopsys-dw-dp-improvements-v12-0-d42d3dc5b...@collabora.com?part=10

Reply via email to