Re: [V4L] [PATCH/RFC] videodev.[ch] redesign

[Mark McClelland <mark@alpha.dyndns.org>]

Gerd Knorr wrote:

>It also provides a ioctl wrapper function which handles copying the
>ioctl args from/to userspace, so we have this at one place can drop all
>the copy_from/to_user calls within the v4l device driver ioctl handlers.
>

Excellent work. I have no complaints, just a few questions:

1. Would it be better to memset the temp buffer in video_generic_ioctl() 
rather than in the driver? I've seen so many drivers forget to do this, 
and it's a potential (albeit very small) security hole.

2. In skeleton_open(), couldn't the device_data lookup code be replaced 
with:

    struct video_device *vdev = video_devdata(file);
    struct device_data *dev = vdev->priv;

3. In skeleton_initdev(), shouldn't...

    dev->vdev = skeleton_template;

...be...

    memcpy(&dev->vdev, &skeleton_template, sizeof(skeleton_template);

4. Is it safe to keep even 128 bytes on the stack in 
video_generic_ioctl()? Consider that devices might spend a relatively 
long time blocking on VIDIOCSYNC. With 32 devices in use at once, you'd 
be coming dangerously close to a stack overflow. IMHO it would be better 
to only allocate as much as MCAPTURE and SYNC need, and fall back on 
kmalloc for the less time-critical ones (if necessary).

Other than that, I extremely happy with what you've done!

-- 
Mark McClelland
mmcclell@bigfoot.com