[igt-dev] [PATCH 1/7] drm/leases: Drop object_id validation for negative ids

Boris Brezillon boris.brezillon at collabora.com
Thu Mar 14 07:54:24 UTC 2019


On Thu, 28 Feb 2019 15:49:04 +0100
Daniel Vetter <daniel.vetter at ffwll.ch> wrote:

> Not exactly sure what's the aim here, but the canonical nil object has
> id == 0, we don't use negative object ids for anything. Plus all
> object_id are valided by the object_idr, there's nothing we need to do
> on top of that ENOENT check a bit further down.
> 
> Spotted while typing exhaustive igt coverage for all these
> corner-cases.
> 
> Cc: Keith Packard <keithp at keithp.com>
> Signed-off-by: Daniel Vetter <daniel.vetter at intel.com>

Reviewed-by: Boris Brezillon <boris.brezillon at collabora.com>

> ---
>  drivers/gpu/drm/drm_lease.c | 5 -----
>  1 file changed, 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/drm_lease.c b/drivers/gpu/drm/drm_lease.c
> index 603b0bd9c5ce..1176d814cf7f 100644
> --- a/drivers/gpu/drm/drm_lease.c
> +++ b/drivers/gpu/drm/drm_lease.c
> @@ -403,11 +403,6 @@ static int fill_object_idr(struct drm_device *dev,
>  	/* step one - get references to all the mode objects
>  	   and check for validity. */
>  	for (o = 0; o < object_count; o++) {
> -		if ((int) object_ids[o] < 0) {
> -			ret = -EINVAL;
> -			goto out_free_objects;
> -		}
> -
>  		objects[o] = drm_mode_object_find(dev, lessor_priv,
>  						  object_ids[o],
>  						  DRM_MODE_OBJECT_ANY);



More information about the igt-dev mailing list