On Thu, Dec 19, 2019 at 6:26 PM Dmitry Sepp <dmitry.sepp@xxxxxxxxxxxxxxx> wrote: > > Hi Gerd, > > On Mittwoch, 18. Dezember 2019 14:40:37 CET Gerd Hoffmann wrote: > > Hi, > > > > > +The device MUST mark the last buffer with the > > > +VIRTIO_VIDEO_BUFFER_F_EOS flag to denote completion of the drain > > > +sequence. > > > > No, that would build a race condition into the protocol. The device > > could complete the last buffer after the driver has sent the drain > > command but before the device saw it. So the flag would not be > > reliable. > No, then it means the device was not in drain, but, for example, hit a > resolution change in the stream and tells us that this is the last buffer with > the old resolution. > > > > > I also can't see why the flag is needed in the first place. The driver > > should know which buffers are queued still and be able to figure > > whenever the drain is complete or not without depending on that flag. > > So I'd suggest to simply drop it. > This flag is used not for drain only. In marks the completion of whatever > specific buffer sequence, like a full end-of-stream, resolution change, drain > etc. We also need this to handle nested sequences. For instance, a resolution > change event might happen while in drain. Good point. The resolution change event is in a different virtqueue, so it's not serialized with the request completions.
