Hello again all, Previously I promised to comment further on progress with our work supporting HDMI input on Lager. After studying commit 4c28078 "[media] rcar_vin: Add scaling support" on Hans' subdev2 branch, I have come to the conclusion that the following is actually reasonable behaviour when no attempt to determine the resolution of the live stream (a non-VGA 720x576 in our case) has been made:
with our modifications to adv7604.c we have seen a) recovery of a 640x480 image which is cropped rather than scaled, and/or b) recovery of a 2048x2048 image with the stream content in the top left corner.
With our latest patchset, frame capture is essentially working and I believe a solution is close, but the scaling changes introduce behaviour that seems to be a regression for this particular test case - see final point below. We have: 1. ported code to subdev2 tree (with backward-compatibility for legacy driver removed) 2. adjusted adv7604.c to detect the ADV7612, adding calls to get_fmt/set_fmt/enum_mbus_code pad ops to suit 3. removed our attempt to query the driver's maximum resolution - the scaling patch above configures the pre-clip/post-clip rectangles sensibly without it 4. modified our attempt to query the live resolution by doing this in adv7604_fill_format() ... by reinstating adv761x_get_vid_info() as a lightweight means of recovering the resolution, then using query_dv_timings to update properly if state->timings is inconsistent with its results; this leads to the need for: 5. reverting the following, as not updating 'pix' with the (possibly updated) resolution set_fmt returns leads to userland requesting an inappropriately sized output buffer: [commit 4c28078, static int rcar_vin_try_fmt(...)] - pix->width = mf.width; - pix->height = mf.height; + /* Adjust only if VIN cannot scale */ + if (pix->width > mf.width * 2) + pix->width = mf.width * 2; + if (pix->height > mf.height * 3) + pix->height = mf.height * 3; Does this seem reasonable, or are there better ways of implementing any of the above steps? Cheers, Wills. -- 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
