[bug report] drm/amd/display: If one stream full updates, full update all planes
Kazlauskas, Nicholas
Nicholas.Kazlauskas at amd.com
Tue Apr 23 14:16:53 UTC 2019
On 4/23/19 10:10 AM, Dan Carpenter wrote:
> Hello David Francis,
>
> This is a semi-automatic email about new static checker warnings.
>
> The patch c238bfe0be9e: "drm/amd/display: If one stream full updates,
> full update all planes" from Mar 29, 2019, leads to the following
> Smatch complaint:
>
> drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:1492 det_surface_update()
> warn: variable dereferenced before check 'u->surface' (see line 1467)
>
> drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c
> 1466
> 1467 if (u->surface->force_full_update) {
> ^^^^^^^^^^^^
> The patch adds a new dereference
>
> 1468 update_flags->bits.full_update = 1;
> 1469 return UPDATE_TYPE_FULL;
> 1470 }
> 1471
> 1472 type = get_plane_info_update_type(u);
> 1473 elevate_update_type(&overall_type, type);
> 1474
> 1475 type = get_scaling_info_update_type(u);
> 1476 elevate_update_type(&overall_type, type);
> 1477
> 1478 if (u->in_transfer_func)
> 1479 update_flags->bits.in_transfer_func_change = 1;
> 1480
> 1481 if (u->input_csc_color_matrix)
> 1482 update_flags->bits.input_csc_change = 1;
> 1483
> 1484 if (u->coeff_reduction_factor)
> 1485 update_flags->bits.coeff_reduction_change = 1;
> 1486
> 1487 if (u->gamma) {
> 1488 enum surface_pixel_format format = SURFACE_PIXEL_FORMAT_GRPH_BEGIN;
> 1489
> 1490 if (u->plane_info)
> 1491 format = u->plane_info->format;
> 1492 else if (u->surface)
> ^^^^^^^^^^
> But the old code assumed it could be NULL
This looks like a check that could probably be dropped. This function
shouldn't be called with a u->surface = NULL - there's even that other
deference at the top of the function with &u->surface->update_flags;
Nicholas Kazlauskas
>
> 1493 format = u->surface->format;
> 1494
>
> regards,
> dan carpenter
> _______________________________________________
> amd-gfx mailing list
> amd-gfx at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
>
More information about the amd-gfx
mailing list