Thanks Steve. I replied to the bugzilla reports you listed. A google search on that commit number reveals quite a few more similar bug reports and applications breaking because of it. I guess quite a few applications would not be aware of the existence of these metadata nodes. Anyway, I'll also work around it in the same way you did.
Op za 22 jun. 2019 om 02:30 schreef Steve Cotton <st...@s.cotton.clara.co.uk >: > On Fri, Jun 21, 2019 at 04:07:25PM +0200, Seba Kerckhof wrote: > > Under Debian 10 rc1, when I plug in my USB camera (the very common > logitech > > c930e), 2 devices are added (e.g. /dev/video0 and /dev/video1). > > > > v4l-info works on /dev/video0, but fails on /dev/video1 > > (VIDIOC_G_FMT(VIDEO_CAPTURE): Invalid argument). When I try to access the > > device using gstreamer it tells me that /dev/video1 is not a capture > device. > > This is the feature added in commit > 088ead25524583e2200aa99111bea2f66a86545a. > > media: uvcvideo: Add a metadata device node > > Some UVC video cameras contain metadata in their payload headers. This > patch extracts that data, adding more clock synchronisation > information, > on both bulk and isochronous endpoints and makes it available to the > user > space on a separate video node, using the V4L2_CAP_META_CAPTURE > capability > and the V4L2_BUF_TYPE_META_CAPTURE buffer queue type. By default, only > the > V4L2_META_FMT_UVC pixel format is available from those nodes. However, > cameras can be added to the device ID table to additionally specify > their > own metadata format, in which case that format will also become > available > from the metadata node. > > Guennadi, sorry but this feels more like a bug than a feature to me, and at > least three other people have reported it as a bug without working out the > cause. > > * https://bugs.debian.org/930860 > * https://bugzilla.kernel.org/show_bug.cgi?id=199193 > * https://bugzilla.kernel.org/show_bug.cgi?id=199575 > > Seba, thanks for reporting this. I met this feature while working on a > closed-source application, added a workaround to ignore cameras that don't > support normal video capture, and haven't yet got round to replying to the > Bugzilla reports above. I'd be happy if someone else took on that task. > > BR, > Steve >
