Uz.ytkownik Matteo Pampolini napisa?:
I'm really sorry - I'm not able to analyse your app. Only one remark -
you are going too test very out of date version of v4l2. The last ones
(kernels >= 2.4.20) you can find at http://bytesex.org
Hi everybody,
I'm quite now to Video4Linux and I'm experiencing something I can't
explain.
I'm currently working on a custom board running Linux 2.4.17 with
a V4L2 capture device, which, at boot, correctly registers itself to
videodevX module since I can see the writing:
V4L2: Registered "Capture device (0)" as char device 81, 0
V4L2: Registered "Capture device (1)" as char device 81, 1
that I found coming from videodevX itself.
Then I run a simple test app that does the following:
struct video_capability videoCaps;
//struct v4l2_capability videoCaps;
int videoFd;
int result;
videoFd = open("/dev/video1", O_RDWR);
if (videoFd < 0)
printf("Cannot open video input device\n");
result = ioctl(videoFd, VIDIOCGCAP, &videoCaps);
//result = ioctl(videoFd, VIDIOC_QUERYCAP, &videoCaps);
printf("result = %d\n", result);
As you can see I do the same thing in two different ways:
In V4L2 mode I got
result = 0
while in V4L mode I got:
ioctl 1 not currently supported
result = -1
Looking at videodevX code I put a printk inside
v4l_compat_translate_ioctl
that I assumed being called during user-mode ioctl call,
but I never saw my print, as if the translation is not performed
at all (tht's why I get the 'not supported' message).
Any idea? Am I missing something?
Thanks in advance,
Matteo.
--
Wies?aw Kierbedz'
gg 2128870; wieslaw_kierbedz@xxxxxxxxxx
www.v4l.prv.pl (video4linux - podstawy)