Re: Any update on em28xx on igevp2 ?
On Thu, 2010-03-18 at 13:26 +0100, Gert-Jan de Jonge wrote:
> Hi Devin,
> just some extra info:
> I have tested the difference on the arm board compared to the pc:
> int height=576;
> height >>= 576;
> printf("%d\n", height);
> on arm it gives a 0, on the pc it gives 576.
> on both i get a warning from the compiler, as in my test code the 576 is
> hardcoded in stead of a variable
> warning: right shift count >= width of type ( which is logical ;) )
> Gert-Jan de Jonge wrote:
> > Hi Devin,
> > I am a big step further, I can now get video from the device.
> > At this moment I am looking at the function em28xx_resolution_set on
> > arm the height is set to 0 by the following lines:
> > if (!dev->progressive)
> > height >>= norm_maxh(dev);
> > I am not sure what it should do, should it really shift the height
> > over the value of height ?
> > If I set the height to f.e. 576 ater this line, i can capture video (
> > it is 576 before this line )
> > should it shift by 1 if it is interlaced and the resolution is higher
> > than the interlaced height ?
> > regards,
> > Gert-Jan
Yeah this is how I fixed my problem also. Though mine was set as 0 but
144 on the pc. Strange since the resolution I then get out is certainly
Hardcoding the values completely fixed the problem for me (as long as I
ignore the compiler warnings :P ).
Looking to find out why the arm gives 0 to fix the cause.
John Banks - Head of Engineering
Noonan Media Ltd
MB: +44 779 62 64 707
video4linux-list mailing list