On Sun, 22 Feb 2009 18:58:29 +0100, Hans Verkuil wrote: > Hi Mauro, > > Please pull from http://www.linuxtv.org/hg/~hverkuil/v4l-dvb-zoran for the > following: > > - saa7115: don't access reg 0x87 if it is not present. > - saa7185: add colorbar support. > - saa7115: add querystd and g_input_status support for zoran. > - zoran: convert to video_ioctl2 and remove 'ready_to_be_freed' hack. > - zoran: remove broken BIGPHYS_AREA and BUZ_HIMEM code, and allow for > kmallocs > 128 kB I had a local patch doing exactly the same ;) > - zoran: use slider flag with volume etc. controls. > - zoran: fix field typo. > - zoran: set bytesperline to 0 when using MJPEG. > - zoran: remove old V4L1 ioctls, use v4l1-compat instead. > - zoran: set correct parent of the video device. > - zoran: cleanups in an attempt to make the source a bit more readable. > - zoran: fix TRY_FMT support > - zoran: fix G_FMT > - zoran: if reqbufs is called with count == 0, do a streamoff. > - zoran et al: convert zoran i2c modules to V4L2. > - zoran: clean up some old V4L1 left-overs and remove the MAP_NR macro. > - zoran: change buffer defaults to something that works with tvtime > - zoran: TRY_FMT and S_FMT now do the same parameter checks. > - bt866: convert to v4l2_subdev. > - bt819: convert to v4l2_subdev. > - bt819: that delay include is needed after all. > - bt856: convert to v4l2_subdev. > - ks0127: convert to v4l2_subdev. > - ks0127: add supported ks0127 variants to the i2c device list. > - saa7110: convert to v4l2_subdev. > - saa7185: convert to v4l2_subdev. > - vpx3220: convert to v4l2_subdev. > - adv7170: convert to v4l2_subdev. > - adv7175: convert to v4l2-subdev. > - zoran: convert to v4l2_device/v4l2_subdev. > - v4l-dvb: cleanup obsolete references to v4l1 headers. > - zoran i2c modules: remove i2c autoprobing support. > - zoran: s_jpegcomp should return a proper result, not 0. > - zoran: increase bufsize to a value suitable for 768x576. > > This is a huge patch series that brings the zoran driver completely up to > date with the latest v4l2 framework. In particular it converts it to use > video_ioctl2, the communication with the i2c modules is converted to from > v4l1 to v4l2 and that made the conversion to v4l2_device/v4l2_subdev > possible. > > As a result of this the saa7111.c and saa7114.c drivers are removed and > instead the saa7115.c driver is used which can handle these older saa711x > devices as well. > > In addition the zoran driver contained experimental support for highmem > allocations which relied on a 'big phys area' kernel patch that was never > merged into the kernel, and the only reason this was needed at all is that > a long time ago kmalloc could only allocate up to 128 kB of contiguous > memory. These days those limitations are gone, and all this code could be > removed, thus making the driver a lot cleaner. > > Also the v4l1 ioctls have been removed and instead zoran relies on the v4l1 > compat layer. > > In addition I've fixed various zoran v4l2 bugs. I separated the fixes from > the big conversions in order to make reviews of the fixes possible. > > After all these changes the driver is actually much improved: various weird > behaviors with i2c (misdetections and not detecting decoders at all > sometimes) are now fixed. The v4l2 behavior has also been improved > considerably. > > What is not working is playback of video. However, it turns out that that > has been broken for some time now. The hope is that now that the driver has > been cleaned up someone can actually go in and try to discover what broke. > Note that video pass-through is working fine. > > I want to give special thanks to Jean Delvare for testing this driver on his > DC10+ and especially to Martin Samuelsson who tested it on his Buz, DC10+ > and 6Eyes cards, both with the vanilla kernel and with this tree. I am very > grateful for all the time he put into this! > > Together these cards use the saa7111, saa7110, bt866, adv7175, saa7185 and > ks0127 i2c devices, thus giving us a solid test coverage. We tested with > tvtime, xawtv, mplayer and lavtools. > > It took me several long weekends to get all this work done, but I think it's > been very worthwhile. Oh yes it is. The new driver looks so much better than the old piece of crap! I might even start working on it now. Mauro, please pull! > diffstat: > (...) > 37 files changed, 4570 insertions(+), 7404 deletions(-) Impressive :) -- Jean Delvare -- 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
