Hi Michael,
On Wednesday 04 June 2008, Michael Sparks wrote:
> Hi,
>
> I've got an Acer laptop with a Crystal Eye WebCam which identifies as
> follows:
>
> Bus 003 Device 002: ID 064e:a101 Suyin Corp.
>
> This is listed as a supported device here:
> http://linux-uvc.berlios.de/#devices
>
> But in practice using the current subversion checkout:
>
> URL: svn://svn.berlios.de/linux-uvc/linux-uvc/trunk
> Repository Root: svn://svn.berlios.de/linux-uvc
> Repository UUID: 790547d8-3c09-0410-a358-857946b093de
> Revision: 212
> Node Kind: directory
> Schedule: normal
> Last Changed Author: pinchartl
> Last Changed Rev: 212
> Last Changed Date: 2008-05-29 15:56:11 +0100 (Thu, 29 May 2008)
>
> When compiled and running (say), luvcview, I get this error:
> ~/Incoming> luvcview
> luvcview 0.2.4
>
> [snip]
> Unable to start capture: Protocol error
> Error grabbing
> Cleanup done. Exiting ...
*sigh*. I hate protocol errors. They are most of the time caused by buggy
devices, and buggy devices mean I have to find a software workaround :-S
> I get similar errors from other software (eg opencv, python v4l2 bindings,
> etc): ~/Incoming> ./WebCam.py
> Traceback (most recent call last):
> File "./WebCam.py", line 238, in <module>
> VideoCaptureSource(fps=32),
> File "./WebCam.py", line 139, in __init__
> self.camera.start()
> EnvironmentError: ioctl(VIDIOC_STREAMON) failure : 71, Protocol error
>
> It turns out that, for whatever reason, when the above error occurs,
> it switches on the webcam. (a little green light comes on). Rerunning
> results in the program then normally working (but green light goes out
> on exit, etc). Then failing when rerun again (leaving light on), etc.
>
> Anyway, I hunted down the problem - the "Protocol Error" is actually
> caused by this line:
> uvc_video.c
>
> if ((ret = usb_set_interface(video->dev->udev, intfnum, i))
> < 0) return ret;
>
> So it's *something* to do with the device presumably not always being
> switched "on" I guess. How that is supposed to be done, I don't know.
>
> HOWEVER, I have made a patch to uvc_video.c that works:
>
> ~/linux-uvc> svn diff
> Index: uvc_video.c
> ===================================================================
> --- uvc_video.c (revision 212)
> +++ uvc_video.c (working copy)
> @@ -750,8 +750,13 @@
> if (i >= intf->num_altsetting)
> return -EIO;
>
> - if ((ret = usb_set_interface(video->dev->udev, intfnum, i))
> < 0) - return ret;
> + if ((ret = usb_set_interface(video->dev->udev,
> intfnum, i)) < 0) {
> + if (ret == -71) { // If we get this error,
> give it another go & retry
> + if ((ret =
> usb_set_interface(video->dev->udev, intfnum, i)) < 0)
> + return ret;
> + } else
> + return ret;
> + }
>
> ret = uvc_init_video_isoc(video, ep);
> } else {
>
> You'll note that I just blindly give the same call another go. The
> first one manages to
> switch the green light on, the second causes it to be useful/usable.
>
> This rather surprisingly worked, and so this has fixed this issue with
> this webcam as used by the following software:
> * pygame.camera
> * opencv
> * luvcview
> * Skype
>
> If this isn't the "correct" way of doing this,please let me know what
> to test and I'll try a different patch :-)
Could you please try to blindly ignore the usb_set_interface return value ?
I'd like to know if the second call is really required.
What kernel version are you running ?
> BTW, many thanks for writing this module - it's great/really useful :-)
You're welcome. Now if we could convince webcam manufacturers to use the Linux
driver for testing purpose I'd be much happier :-)
Best regards,
Laurent Pinchart
_______________________________________________
Linux-uvc-devel mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/linux-uvc-devel
