This is the second version of this patch series.
Two changes:
- fixed the handling of a frequency of 0 in the radio-keene driver (thanks
to Laurent for spotting that special case)
- added a sixth patch that prevents the v4l2 core from printing more than the
size of the buffer containing the name. The data in that buffer may not be
0 terminated, either because that's how userspace passed the data or because
of driver bugs. So safe-guard against that.
If nothing else comes up, then I plan on posting a pull request on Friday.
Regards,
Hans
>From the first patch series:
This is the second and last phase of ensuring that the arguments of write-only
ioctls in V4L2 are const. The first phase was 4-5 months ago and added const
to s_crop, s_modulator, s_audio, s_audout, (un)subscribe_event, s_freq_hw_seek,
s_jpegcomp and s_fbuf.
This second phase adds const to s_frequency, s_tuner, s_std and s_register.
Actually, for s_std it doesn't add const but changes it to pass the std by
value which is more consistent in that particular case.
As a result drivers will be aware when they are implementing write-only ioctls
(and I saw a few drivers attempting to return data back to the user), and the
v4l2 core will know that drivers won't change the argument of a write-only
ioctls which simplifies the core debug code.
The changes have been compile-tested with the linux-media daily build but
I may have missed some more exotic architectures.
Ideally I would like to have this merged fairly early on so we have enough
time to shake out any remaining compile problems.
Regards,
Hans
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
