I looks like the problem you are seeing is the same one people are
having with OV511 cameras on OHCI. They used to work with 2.4.3, and in
2.4.4 they either crash the kernel or produce corrupted images. The
camera initialization, however, works fine.
According to one user, it worked in 2.4.4-pre3 and broke in 2.4.4-pre7.
It appears that isochronous packets are being dropped, possibly when the
IRQ load gets too heavy, or when too much time is spent in the handler.
Not to jump to conclusions, but I wonder if the AMD 756 patch is what
caused this?
Sönke Ruempler wrote:
>still no one with a solution ?
>
>>ok - 2.4.4 final has the same problem. as i wrote, the cpia driver
>>worked fine with my terracam usb webcam in 2.4.3 until kernel oopsed
>>cause of the amd viper controller.
>>
>>and in kernel 2.4.4 usbcore has one unresolved symbol
>>
>>"_mmx_memcpy"
>>
>>that can be removed if i recompile the kernel without
>>"CONFIG_MODVERSIONS".
>>
>>the usb subsystem recognizes my webcam like in 2.4.3. I recompiled
>>w3cam/vidcat after upgrading to kernel 2.4.4 - and the same problem. so
>>i dont know what to do now...
>>
>>i can paste the kernel/dmesg-logs ... , if somebody has a solution ..
>>
>>cya, s0enke
>>
>>David Brownell schrieb:
>>
>>>>i am testing my cpia webcam (terracam usb) with kernel 2.4.4-pre8 and
>>>>amd 765.
>>>>
>>>>Apr 28 13:02:09 blah kernel: cpia.c:parse_picture(1947):EOL not found
>>>>giving up after 1350/81428 bytes
>>>>
>>>the evidence is as suggestive of a cpia or device problem as anything else...
>>>
--
Mark McClelland
[EMAIL PROTECTED]
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-devel
