Guido Günther wrote: > On Thu, Aug 14, 2008 at 03:41:28PM -0500, Anthony Liguori wrote: >> Max Krasnyansky wrote: >>>>> Do you mind if we do something like that incrementally ? >>>>> >>>> Yeah, I don't really want to have QEMU depend on dbus so in this case, >>>> polling would be better. >>>> >>> I'm thinking that maybe we could use 'inotify' on /proc/bus/usb. >>> Would you be ok with 'inotify' ? >>> >> Yeah, but I would be a little surprised if /proc/bus/usb supports inotify... > What about /dev/bus/usb - it supports inotify fine on udev?
Yes it should since it's a regular filesystem. Now inotify based solution probably won't be pretty because we'd have to monitor each subdir. ie When new device get added top level /dev/bus/usb is not modified, what does get modified is /dev/bus/usb/<bus_num>/ directory so we'd have to monitor /dev/bus/usb and dynamically register/unregister monitors for each /dev/bus/usb/<bus_num>/. Maybe it won't be that bad. If I get a chance I'll give it a shot. btw Interface to HAL might still be useful in general to monitor other device classes that we may want to automatically assign to the VMs. So I'll play around with that too (some day :)). Max -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
