diff options
| author | Dr. David Alan Gilbert <linux@treblig.org> | 2010-10-24 13:33:09 -0300 |
|---|---|---|
| committer | Mauro Carvalho Chehab <mchehab@redhat.com> | 2010-10-25 18:53:44 -0200 |
| commit | 910f5f05f99c1ffbb484c4e6eb2a460e7a08e2d7 (patch) | |
| tree | 51bafcd6627b43233ce0896ecae489b8a19ee00a /drivers/media | |
| parent | 52fae5eead706cb8d604ee634162f6694d75ccad (diff) | |
| download | kernel_samsung_crespo-910f5f05f99c1ffbb484c4e6eb2a460e7a08e2d7.zip kernel_samsung_crespo-910f5f05f99c1ffbb484c4e6eb2a460e7a08e2d7.tar.gz kernel_samsung_crespo-910f5f05f99c1ffbb484c4e6eb2a460e7a08e2d7.tar.bz2 | |
[media] Guard a divide in v4l1 compat layer
Hi,
I managed to trigger a divide by 0 in the v4l compat code
with the mem2mem test module; I suspect perhaps it shouldn't
have been returning a 0 pixel wide picture, but either way it seems
right to guard this divide by 0 in the compatibility layer.
Tested on 2.6.36 (ubuntu build, but the code in this is the same as upstream),
but ***not tested with a real video device***.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
cc: stable.kernel.org
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Diffstat (limited to 'drivers/media')
| -rw-r--r-- | drivers/media/video/v4l1-compat.c | 13 |
1 files changed, 10 insertions, 3 deletions
diff --git a/drivers/media/video/v4l1-compat.c b/drivers/media/video/v4l1-compat.c index 0c2105c..d4ac751 100644 --- a/drivers/media/video/v4l1-compat.c +++ b/drivers/media/video/v4l1-compat.c @@ -645,9 +645,16 @@ static noinline long v4l1_compat_get_picture( goto done; } - pict->depth = ((fmt->fmt.pix.bytesperline << 3) - + (fmt->fmt.pix.width - 1)) - / fmt->fmt.pix.width; + if (fmt->fmt.pix.width) + { + pict->depth = ((fmt->fmt.pix.bytesperline << 3) + + (fmt->fmt.pix.width - 1)) + / fmt->fmt.pix.width; + } else { + err = -EINVAL; + goto done; + } + pict->palette = pixelformat_to_palette( fmt->fmt.pix.pixelformat); done: |