Hans Verkuil <hverkuil@xxxxxxxxx> writes: > Now, I am not planning to merge that, but I will compare it to what Ezequiel has > and use that comparison as a starting point for further discussions. I'm not opposed to Ezequiel's changes in general - I only want him to present them as changes, and not as a "rewritten driver". Unless they are considered a rewritten driver, of course. > As I mentioned before, my preference is to merge a driver that supports both > frame and field modes (or whatever they are called). Well, I don't know about his, but my version presently only support frame mode (V4L2_FIELD_SEQ_*). It specifically doesn't support INTERLACED frame mode (SG DMA hw limitation) and it doesn't support field mode, except for FIELD_TOP and FIELD_BOTTOM (only specific fields) which I use to get a lower resolution image. I also plan to use an INTERLACED frame mode with a DMA (non-SG to CMA, it turned out it works well on my systems), but I would rather like to have the driver in the tree before submitting further (non-essential) patches. Field modes are of course possible, though I've never seen any interest for them, so I haven't bothered. Can add, it's a simple thing and it works fine (apart from increased QBUF/DQBUF rate). There is also a specific YUV420 mode (with a custom encoding) which I'd like to add (unfortunately unavailable in V4L2_FIELD_SEQ_*, but usable in INTERLACED and FIELD_*). This mode requires two separate line lengths - odd lines contain YUV data and are twice as long as even lines which only contain Y. How do I set up such a mode (fmt.pix.bytesperline)? This mode is useful on low-performance machines, e.g. feeding data to H.264 encoders. -- Krzysztof Halasa Industrial Research Institute for Automation and Measurements PIAP Al. Jerozolimskie 202, 02-486 Warsaw, Poland -- 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
