On 03/11/2016 09:55 PM, Niklas Söderlund wrote: > On 2016-03-11 12:16:30 +0100, Hans Verkuil wrote: >> On 03/11/2016 12:03 PM, Niklas Söderlund wrote: >>> Hi Hans, >>> >>> On 2016-03-11 09:26:48 +0100, Hans Verkuil wrote: >>>> Hi Niklas, >>>> >>>> On 02/29/2016 10:52 AM, Hans Verkuil wrote: >>>>> Hi Niklas, >>>>> >>>>> Thanks for your patch! Much appreciated. >>>>> >>>>> I have more comments for the v2, but nothing really big :-) >>>>> >>>> >>>> Just checking, you are working on a v3, right? I'd really like to get this in >>>> for kernel 4.7. >>> >>> Yes I had to switch focus for a bit but now I'm back working on this >>> again today. >>> >>> I have some trouble getting NV16 to work. I can't get it to work using >>> soc_camera driver either, or more accurate I get the same output broken >>> rendering in qv4l2. What would you say is better drop NV16 support form >>> the driver or keep it as is since it is compatible with the soc_camera >>> drivers implementation? I would like to keep it in the driver for now >>> since it at least works as good as in soc_camera. >> >> I would have the NV16 support as a separate patch so we can decide on this >> later. >> >> I don't really like having to support a broken format. >> >> Do you know in what way the format is broken? > > Turns out it was my fault all along for assuming qv4l2 could render NV16 > out-of-the-box. If one decodes the format correctly it works fine both > with this driver and the one in soc_camera. Is your qv4l2 compiled with openGL support? (See Help:About) The NV16 format is supported with openGL rendering, but not with software rendering. That's probably the reason why it wasn't reproduced correctly. Regards, Hans
