On 08/08/2016 09:30 PM, Robert Jarzmik wrote:
> Hi Hans,
>
> We're leaving the domain of the RFC to a proper submission.
>
> This is very alike to what you reviewed earlier, the code is very close, and :
> - the split between patches is done to better isolate cleanups from real code
> - start_streaming() was implemented
> - your remarks have been taken into account (please double-check if you're
> happy with it)
> - v4l2-compliance -f and v4l2-compliance -s were run without any error, and
> 6 warnings
> warn: v4l2-test-formats.cpp(713): TRY_FMT cannot handle an invalid
> pixelformat.
> warn: v4l2-test-formats.cpp(714): This may or may not be a problem. For
> more information see:
> warn: v4l2-test-formats.cpp(715):
> http://www.mail-archive.com/linux-media@vger.kernel.org/msg56550.html
> - soc_camera is not touched anymore
> - the driver is still functional from a capture point of view as before
> (ie. taking a real picture)
>
> I'm still relying on soc_mediabus, hence the not-so-nice Makefile diff hunk.
>
> The only architecture which will have its deconfigs impacted is pxa, under my
> maintainance, and once the review is finished and you have a landing cycle
> I'll
> complete with a simple serie on the pxa side (defconfig + platform data).
>
> I've also put the whole serie here if you want to fetch and review from git
> directly :
> - git fetch https://github.com/rjarzmik/linux.git work/v4l2
>
> Happy review.
As you can see from my replies I only have a few small review comments
remaining.
Can you fix it and post a v4 (rebased to the latest kernel)? Please also add the
v4l2-compliance -f output to the cover letter.
Unless I see something new I plan to make a pull request of v4 for 4.9.
I hope to post my conversion of the atmel-isi driver soon, and that might give
you some ideas on how to get rid of the soc-mbus stuff in a follow-up patch.
Regards,
Hans
