Hello White testing a driver I have stepped into some strange behaviour and I want to know if it is a feature or a bug. I am using yavta to test the system and I run this command: yavta /dev/video0 -c -n 0 to start a capture with 0 buffers (Even if I dont know where this can be useful) And I have found out that: 1) If the user does a streamon() and then close() the descriptor, streamoff is not called, this is because he has never been set as owner of the queue. (on vb2_fop_release, queue_release is only called if the owns the queue) Is this expected? Shouldn't we leave the stream stopped? I propose to set vdev->queue->owner to the current vdev on streamon if it does not have an owner. Or in vb2_fop_release set check for : if (!vdev->queue->owner || file->private_data == vdev->queue->owner) instead of if (file->private_data == vdev->queue->owner) Shall I post a patch? 2) the queue_setup handler of the driver is not called, this could be expected, since it is commented on the code. /* * In case of REQBUFS(0) return immediately without calling * driver's queue_setup() callback and allocating resources. */ But I find it strange, the driver could be doing some initialization there... Thanks! -- Ricardo Ribalda -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
