[cpia] ezonics

[Johannes Erdfelt]

On Wed, Apr 12, 2000, George Talusan <gstalusan@math.uwaterloo.ca> wrote:
> 
> I'm running 2.3.99-pre4 and after using gqcam-0.7.2 for a few
> minutes the driver hangs.  gqcam goes into D state and the
> only way to get CPiA working again is by rebooting.
> 
> I'm using an Abit BP6 2x466MHz, and an EZonics EZcam USB.
> 
> Could this be a USB or CPiA problem?
> 
> Here's dmesg.
> 
> cpia data error: [9] len=768, status=FFFFFFAC
> cpia.c:parse_picture(1960):line length was not 1 but -2 after 3806/19886
> bytes
> cpia data error: [7] len=288, status=FFFFFFAC
> cpia.c:parse_picture(1960):line length was not 1 but 0 after 18483/20836
> bytes
> cpia data error: [6] len=128, status=FFFFFFAC
> cpia data error: [7] len=320, status=FFFFFFAC
> cpia data error: [8] len=128, status=FFFFFFAC
> cpia data error: [9] len=96, status=FFFFFFAC
> cpia.c:parse_picture(1960):line length was not 1 but -1 after 14574/18075
> bytes
> usb-uhci.c: interrupt, status 2, frame# 1071
> cpia data error: [0] len=416, status=FFFFFFAC
> cpia data error: [5] len=256, status=FFFFFFAC
> cpia.c:parse_picture(1960):line length was not 1 but -2 after 12901/25178
> bytes
> usb-uhci.c: interrupt, status 2, frame# 1449
> usb-uhci.c: interrupt, status 2, frame# 661
> usb-uhci.c: interrupt, status 2, frame# 1575
> usb-uhci.c: interrupt, status 2, frame# 295
> usb-uhci.c: interrupt, status 2, frame# 1541
> usb-uhci.c: interrupt, status 2, frame# 1673

Try using the Alternate UHCI driver I wrote, people have reported better
results.

But, I think there's a bug in the CPiA driver which is causing it to get
stuck in D. You will always get errors like you mentioned above, but you
shouldn't get too many of them (depending on cable length, etc). I think
if a couple of frames in a row get dropped things could go bad and it'll
get stuck.

JE