> > A simpler approach seems to be to only load the driver, > when streaming is > > required. Yes, it would add a (considerable) delay to > streaming begin, but > > you'd be completely honest to your user and privacy would > be guaranteed. > > The delay will likely be close to the one introduced by > flushing the shadow > registers: it will basically be the needed time to transfer > all registers data > via I2C. I would think the delay is potentially a lot worse, as the module has to be read from the filesystem (if the image sensor driver was built as a module). > > Another less secure approach would be to tie your LED to a > different > > function, to one, that only activates, when actual data is > transferred. > > Maybe you can tie it to one of sync or clock signals? Look > whether your > > sensor has any pins, that only get activated, when video data is > > transferred. > > I agree. If the hardware design is not finished, this is the > best way of > doing that. > > It is safer and more honest than attaching the led to the I2C bus, as > it will be monitoring the actual data transfer, an not the > control bus. I completely agree on this as well. Sadly, that doesn't seem to be an option at this point. ----------------------------------------------------------------------------------- This email message is for the sole use of the intended recipient(s) and may contain confidential information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. ----------------------------------------------------------------------------------- -- 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
