[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YyA9Y+Cs8ZCYHXAT@intel.com>
Date: Tue, 13 Sep 2022 11:20:51 +0300
From: Ville Syrjälä <ville.syrjala@...ux.intel.com>
To: butt3rflyh4ck <butterflyhuangxx@...il.com>
Cc: maarten.lankhorst@...ux.intel.com, mripard@...nel.org,
Thomas Zimmermann <tzimmermann@...e.de>,
Dave Airlie <airlied@...ux.ie>,
Daniel Vetter <daniel@...ll.ch>,
LKML <linux-kernel@...r.kernel.org>,
dri-devel <dri-devel@...ts.freedesktop.org>,
Ben Davis <ben.davis@....com>,
Brian Starkey <brian.starkey@....com>,
Liviu Dudau <Liviu.Dudau@....com>,
Jonas Karlman <jonas@...boo.se>
Subject: Re: A divide error bug in framebuffer_check
On Tue, Sep 13, 2022 at 01:49:40AM +0800, butt3rflyh4ck wrote:
> Hi, there is a divide error bug in framebuffer_check in
> drivers/gpu/drm/drm_framebuffer.c in the latest kernel.
> we can trigger it via drm_mode_addfb2 IOCTL.
> The call trace is drm_mode_addfb2 -> drm_internal_framebuffer_create
> -> framebuffer_check.
> let us see code below:
> ```
<snip>
> { .format = DRM_FORMAT_Q410, .depth = 0,
> .num_planes = 3, .char_per_block = { 2, 2, 2 },
> .block_w = { 1, 1, 1 }, .block_h = { 1, 1, 1 }, .hsub = 0,
> .vsub = 0, .is_yuv = true },
> { .format = DRM_FORMAT_Q401, .depth = 0,
> .num_planes = 3, .char_per_block = { 2, 2, 2 },
> .block_w = { 1, 1, 1 }, .block_h = { 1, 1, 1 }, .hsub = 0,
> .vsub = 0, .is_yuv = true },
> ```
> we can see format.hsub and format.vsub all are NULL.
Yeah, those look borked.
Added in commit 94b292b27734 ("drm: drm_fourcc: add NV15, Q410, Q401 YUV formats")
Adding the relevant people to cc...
--
Ville Syrjälä
Intel
Powered by blists - more mailing lists