On Sun, 8 Jul 2012, Laurent Pinchart wrote: > > this quirks also affect snd_usb_audio module. If for some reasons > > uvcvideo is not loaded, snd_usb_audio will fail - i mean haw mysterious > > sound problems. > > Good point. If we load the uvcvideo driver while the audio function is in > use, > I doubt resetting the device will lead to a good user experience. One could > argue whether this should happen in the first place though, as modules should > be auto-loaded. Could there be a race between audio and video probe, leading > to audio probe failure because we reset the device in the middle of the probe > sequence ?
Even if there's no race and uvcvideo never gets loaded, there still could be a problem. The device could get suspended while waiting for the snd-usb-audio module to load. Then when the audio driver is probed, it could be looking at a webcam that has already crashed because the RESET_RESUME quirk was never set. An alternative would be to modify the usb_detect_quirks() routine directly -- have it set RESET_RESUME whenever it sees a Logitech webcam. The difficulty there is how to recognize which devices are webcams. Look for a video interface? Any suggestions? Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
