Hi, > This adds a race window. Between looking for /dev/video users and > invoking the ioctl another application might have opened the device.
True. However, as I understood the proposal, this race wasn't considered significant. > IMHO you can't fix that race with a pure userspace implementation. If you really wanted, I believe you could, using lockfiles. However, this probably becomes too complicated. > That scheme requires cooperation from all applications. Sure, but so does the proposed priority ioctl() ! Hmm -- you probably mean the default case, where _no_ priority is set by an app, but its channel change can still result in -EBUSY due to another app having used set up a higher priority. Ok, that is not achievable with the userspace solution, admitted. Regards, Wolfram. -- video4linux-list mailing list Unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/video4linux-list
