Re: [PATCH] drm/msm/dp: Free resources after unregistering them

[Dmitry Baryshkov]

On 13/06/2023 01:02, Bjorn Andersson wrote:
> The DP component's unbind operation walks through the submodules to
> unregister and clean things up. But if the unbind happens because the DP
> controller itself is being removed, all the memory for those submodules
> has just been freed.
>
> Change the order of these operations to avoid the many use-after-free
> that otherwise happens in this code path.
>
> Fixes: c943b4948b58 ("drm/msm/dp: add displayPort driver support")
> Signed-off-by: Bjorn Andersson <quic_bjorande@xxxxxxxxxxx>
> ---
>   drivers/gpu/drm/msm/dp/dp_display.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/msm/dp/dp_display.c b/drivers/gpu/drm/msm/dp/dp_display.c
> index bbb0550a022b..ebc84b8fddf8 100644
> --- a/drivers/gpu/drm/msm/dp/dp_display.c
> +++ b/drivers/gpu/drm/msm/dp/dp_display.c
> @@ -1337,9 +1337,9 @@ static int dp_display_remove(struct platform_device *pdev)
>   {
>   	struct dp_display_private *dp = dev_get_dp_display_private(&pdev->dev);
>   
> +	component_del(&pdev->dev, &dp_display_comp_ops);
>   	dp_display_deinit_sub_modules(dp);
>   
> -	component_del(&pdev->dev, &dp_display_comp_ops);
>   	platform_set_drvdata(pdev, NULL);

This matches more or less the order in dp_display_probe().

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxx>

A note for the possible followup: the driver initializes DP debugfs from 
dpu_kms (ugh) by calling msm_dp_debugfs_init() -> dp_debug_get(). I 
think that dp_debug_put() in dp_display_deinit_sub_modules() does not 
look correct.

>   
>   	return 0;

--
With best wishes
Dmitry