[linux-avc2210k] Re: Where's avcap, what's the latest driver version, and "device reboots"...
- From: Dr J A Gow <J.A.Gow@xxxxxxxxxxxxxxxxx>
- To: linux-avc2210k@xxxxxxxxxxxxx
- Date: Sun, 01 Apr 2007 21:17:29 +0100
Hi,
On Sun, 2007-04-01 at 19:02 +0100, j t wrote:
> Hi John,
>
> I've managed to get my newly-acquired avc-2210 working using the
> downloads available from 81.86.3.90. Could you please tell me:
>
> a) What's the recommended way to capture the data coming from
> /dev/avcap0? (I can't find the source for the user-space avcap capture
> program mentioned) I've been using "cat /dev/avcap0 > file.mpg", but I
> read somewhere that "cat" with a redirect isn't sufficiently buffered
> to be suitable. Also, would I be correct in saying that insufficient
> buffering might cause a/v sync problems?
>
cat /dev/avcap0 > file.mpg
The driver does its own buffering so cat should be fine.
> b) Are the versions available on 81.86.3.90 the most recent? Also, is
> development on these drivers still ongoing?
>
Apart from my unstable development tree, these are the most recent and
development is still very much ongoing, albeit slowly at the moment.
Next stage is to sort out the build system so there are fewer
distro-related problems.
> c) I've found that under certain circumstances, my avc-2210 "reboots".
> It seems to happen when I'm recording and the analog video input
> disappears. The symptoms are the same as in Wolfgang's mail:
> (//www.freelists.org/archives/linux-avc2210k/10-2006/msg00020.html)
> <quote>
It will. This is a problem with the firmware. Windows seems to work
around it but I have not found the trick yet. The code in my development
tree currently stops the stream if the video capture IC flags a sync
failure - this works a lot better but does not completely solve the
problem.
> One minor problem:
> Every now and then, when I'm only viewing the output of the device, xine
> or mplayer seem to have problems recognizing the stream type and after
> a while, the flashing green light on the device changes to blinking
> orange. After another while, the light becomes green again and after
> uploading the firmware, everything is working again. Ever heard of
> this?
> </quote>
>
Sounds like the same problem as above. Never seen this with a clean
video signal input.
> and I get the following in my syslog:
>
> Apr 1 18:40:43 server kernel: AVC2x10K : close index 0
> Apr 1 18:40:55 server kernel: AVC2x10K : buffer block is full -
> initiating stop handler
> Apr 1 18:40:55 server kernel: AVC2x10K : state -> DEV_STATE_STOPPING
> Apr 1 18:40:55 server kernel: AVC2x10K : stream stop handler entered
> Apr 1 18:40:55 server kernel: AVC2x10K : flushing buffers
>
This looks like a normal driver stream shutdown.
> I haven't yet tried using adaptec's windows s/w so I'm not sure
> whether the same thing happens under m$.
>
If you do, please let me know what happens. I used VMware and USB wire
dumps to work out what was being sent to the device - and the ability to
save and scan enormous dump files limited the length of time I could
capture for.
> Again, thank you for the work you've done on th
> ese drivers.
>
No problem, glad you find them useful.
John
Other related posts:
