Hi Ben,
On 26 July 2017 at 19:08, Ben Widawsky <ben@xxxxxxxxxxxx> wrote:
> +static bool intel_primary_plane_format_mod_supported(struct drm_plane *plane,
> + uint32_t format,
> + uint64_t modifier)
> +{
> + struct drm_i915_private *dev_priv = to_i915(plane->dev);
> +
> + if (WARN_ON(modifier == DRM_FORMAT_MOD_INVALID))
> + return false;
> +
> + if (!(modifier & DRM_FORMAT_MOD_VENDOR_INTEL) &&
> + modifier != DRM_FORMAT_MOD_LINEAR)
> + return false;
The vendor ID isn't shifted yet, so this comparison is wrong. Even if
it was shifted, they're sequential rather than bitmask, so
NV/QCOM/BROADCOM all match (vendor & INTEL). To fix both of these:
- if (!(modifier & DRM_FORMAT_MOD_VENDOR_INTEL) &&
- modifier != DRM_FORMAT_MOD_LINEAR)
+ if ((modifier >> 56) != DRM_FORMAT_MOD_VENDOR_INTEL &&
+ modifier != DRM_FORMAT_MOD_LINEAR)
The practical effect is that this knocked out MOD_Y_TILED and
MOD_Y_TILED_CCS from the list, since they're the only two Intel
modifiers without the low bit set.
Cheers,
Daniel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
