DongSoo(Nathaniel) Kim wrote:
Hi Tuukka, I understand that it is a huge thing to support VIDIOC_S_INPUT.
It might not be that much, in case it's supported by just one slave.
But without that, we don't have any proper "V4L2" api to get information about how many devices are attached to camera interface, and names of input devices...and so on. Because VIDIOC_ENUMINPUT and VIDIOC_G_INPUT needs VIDIOC_S_INPUT for prior. Of course we can refer to sysfs, but using only single set of APIs like V4L2 looks more decent. What do you think about this? If you think that it is a big burden, can I make a patch for this?
I'm just wondering the purpose --- as Tuukka explained, the ISP block doesn't make a camera, but a sensor essentially does. So for every sensor there's a video device, even if they are implemented by using just one ISP. How does this sound like?
But for video decoders at least it's definitely meaningful to support VIDIOC_S_INPUT.
-- Sakari Ailus sakari.ailus@xxxxxxxxxxxxxxxxxxxxxxxxxx -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
