Il 24/07/23 19:58, Justin Green ha scritto:
Add a check to mtk_drm_mode_fb_create() that rejects any modifier that
is not the AFBC mode supported by MT8195's display overlays.
Tested by booting ChromeOS and verifying the UI works, and by running
the ChromeOS kms_addfb_basic binary, which has a test called
"addfb25-bad-modifier" that attempts to create a framebuffer with the
modifier DRM_FORMAT_MOD_INVALID and verifies the ADDFB2 ioctl returns
EINVAL.
Signed-off-by: Justin Green <greenjustin@xxxxxxxxxxxx>
---
drivers/gpu/drm/mediatek/mtk_drm_drv.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
index cd5b18ef7951..2096e8a794ad 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
@@ -51,6 +51,13 @@ mtk_drm_mode_fb_create(struct drm_device *dev,
if (info->num_planes != 1)
return ERR_PTR(-EINVAL);
+ if (cmd->modifier[0] &&
+ cmd->modifier[0] != DRM_FORMAT_MOD_ARM_AFBC(
+ AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 |
+ AFBC_FORMAT_MOD_SPLIT |
+ AFBC_FORMAT_MOD_SPARSE))
+ return ERR_PTR(-EINVAL);
Would it make more sense to commmonize function mtk_plane_format_mod_supported()
and call that one here instead?
Regards,
Angelo
+
return drm_gem_fb_create(dev, file, cmd);
}
