On Sunday 26 August 2007, Felix Möller wrote: > Laurent Pinchart schrieb: > > Hello everybody. > > > > I received no answers to my recent call for testers for the > > suspend/resume patch. This either means nobody tried it, or it worked > > flawlessly for everybody. > > > > The patch is now in SVN. If you haven't tried it before, here's the time > > to check it out and complain about issues. > > As I am using an iSight I do not know how much this report is worth. But > r122 breaks suspend for me: > > Stopping tasks ... done. > Suspending console(s) > uvcvideo 5-4:1.0: suspend error -22 > suspend_device(): usb_suspend+0x0/0x17 [usbcore]() returns -22 > Could not suspend device 5-4: error -22 > Some devices failed to suspend > Restarting tasks ... done. > > This is on current openSUSE 10.3 factory.
Hmmm... I think I know where that comes from. The iSight report a vendor subclass in the interface descriptors (you already noticed the issue with revision 119). I'm not happy with that. The more I clean the driver, the more iSight specific quirks will be needed. I thought about writing an open-source firmware replacement that would be UVC compatible. That turned out to be difficult as the webcam doesn't have the necessary hardware to generate a UVC-compatible isochronous data stream. Patching the iSight firmware to report the right interface subclasses would be easy, but we can't make sure it won't have any side effect. Whatever direction we take, a dedicated iSight maintainer might be needed, as I'll probably break other things in the future (not intentionally of course :-)). Best regards, Laurent Pinchart _______________________________________________ Linux-uvc-devel mailing list [email protected] https://lists.berlios.de/mailman/listinfo/linux-uvc-devel
