On Sun March 24 2013 17:13:40 Mauro Carvalho Chehab wrote: > Em Fri, 22 Mar 2013 15:36:35 +0100 > Hans Verkuil <hverkuil@xxxxxxxxx> escreveu: > > > Hi all, > > > > This pull request updates the staging go7007 driver to the latest > > V4L2 frameworks and actually makes it work reliably. > > > > This pull request assumes that pull request > > http://patchwork.linuxtv.org/patch/17568/ was merged first. > > > > Some highlights: > > > > - moved the custom i2c drivers to media/i2c. > > - replaced the s2250-loader by a common loader for all the supported > > devices. > > - replaced all MPEG-related custom ioctls by standard ioctls and FMT > > support. > > - added the saa7134-go7007 combination (similar to the saa7134-empress). > > - added support for ADS Tech DVD Xpress DX2. > > > > In addition I've made some V4L2 core and saa7115 changes (the first 7 > > patches): > > > > - eliminate false lockdep warnings when dealing with nested control > > handlers. This patch is a slightly modified version from the one Andy > > posted a long time ago. > > - add support to easily test if any subdevices support a particular operation. > > - fix a few bugs in the code that tests if an ioctl is available: it didn't > > take 'disabling of ioctls' into account. > > - added additional configuration flags to saa7115, needed by the go7007. > > - improved querystd support in saa7115. > > > > This driver now passes all v4l2-compliance tests. > > > > It has been tested with: > > > > - Plextor PX-TV402U (PAL model) > > - Sensoray S2250S (generously provided by Sensoray, all audio inputs > > now work!) > > - Sensoray Model 614 (saa7134+go7007 PCI board, generously provided by > > Sensoray) > > - WIS X-Men II sensor board (generously provided by Sensoray) > > - Adlink PCI-MPG24 surveillance board > > - ADS Tech DVD Xpress DX2 > > > > Everything seems to work OK, but for two things: > > > > - the WIS X-Men and tthe S2250 do not honor requested frameperiod changes > > using S_PARM. The others work fine, and I have no idea why these work > > differently. > > - the bttv part of the Adlink card doesn't work for me: I just get black > > with fuzzy lines. This doesn't work in 3.8 either, so I don't know > > what's going on here. It's not related to my patch series, that's for > > sure. > > > > What needs to be done to get this driver out of staging? The main thing > > is the motion detection support. Volokh has some additional code for that, > > and I want to experiment with motion detection for this card and the > > solo6x10 card and see if I can come up with a nice API for that. > > > > It would also be nice to get the s2250-board.c code make use of the already > > existing i2c devices, but it is hooked up somewhat strangely, so I need to > > look at that some day. > > > > Regarding the firmware: they are available here: > > > > http://git.linuxtv.org/hverkuil/linux-firmware.git/shortlog/refs/heads/go7007 > > > > All firmwares relating to this driver have been collected in the go7007 > > directory with correct licensing. Note that this means that the s2250 firmwares > > have been renamed. Should this be an issue I can change this back and leave > > those files where they are today, but since the go7007 firmware files were > > never included in linux-firmware (and therefor the driver never worked with > > just linux-firmware) and because it is still a staging driver I thought it > > cleaner to have all firmware files in one place. > > > > Mauro, when should I make a pull request for the linux-firmware changes? > > After you have merged this pull request? > > > > In the meantime, the firmware files are also available here: > > > > http://hverkuil.home.xs4all.nl/go7007-fw.tar.bz2 > > > > Just unpack in /lib/firmware. > > > > Regards, > > > > Hans > > > > The following changes since commit 8bf1a5a826d06a9b6f65b3e8dffb9be59d8937c3: > > > > v4l2-ioctl: add precision when printing names. (2013-03-22 11:59:21 +0100) > > > > are available in the git repository at: > > > > git://linuxtv.org/hverkuil/media_tree.git go7007 > > > > for you to fetch changes up to 651f19e2186eb92393296717afaa7fc0873d6c2f: > > > > go7007: add support for ADS Tech DVD Xpress DX2 (2013-03-22 15:20:46 +0100) > > > > ---------------------------------------------------------------- > > Andy Walls (1): > > v4l2-ctrls: eliminate lockdep false alarms for struct v4l2_ctrl_handler.lock > > > > Hans Verkuil (47): > > v4l2-core: add code to check for specific ops. > > v4l2-ioctl: check if an ioctl is valid. > > v4l2-ctrls: add V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER control > > saa7115: add config flag to change the IDQ polarity. > > saa7115: improve querystd handling for the saa7115. > > saa7115: add support for double-rate ASCLK > > go7007: fix i2c_xfer return codes. > > tuner: add Sony BTF tuners > > sony-btf-mpx: the MPX driver for the sony BTF PAL/SECAM tuner > > ov7640: add new ov7640 driver > > uda1342: add new uda1342 audio codec driver > > tw9903: add new tw9903 video decoder. > > tw2804: add support for the Techwell tw2804. > > go7007: switch to standard tuner/i2c subdevs. > > go7007: remove all wis* drivers. > > go7007: add audio input ioctls. > > s2250-loader: use usbv2_cypress_load_firmware > > go7007: go7007: add device_caps and bus_info support to querycap. > > go7007: remove current_norm. > > go7007: fix DMA related errors. > > go7007: remember boot firmware. > > go7007: fix unregister/disconnect handling. > > go7007: convert to the control framework and remove obsolete JPEGCOMP support. > > s2250: convert to the control framework. > > go7007: add prio and control event support. > > go7007: add log_status support. > > go7007: tuner/std related fixes. > > go7007: standardize MPEG handling support. > > go7007: simplify the PX-TV402U board ID handling. > > go7007: set up the saa7115 audio clock correctly. > > go7007: drop struct go7007_file > > go7007: convert to core locking and vb2. > > go7007: embed struct video_device > > go7007: remove cropping functions > > saa7134-go7007: add support for this combination. > > I won't apply this one yet. A non-staging driver should not try to load a > staging one without a notice. That change would be ok if you were also > moving go7007 out of staging. Fair enough. I will prepare a patch that at least updates the saa7134-go7007.c source with my changes. That only leaves the patch to saa7134 itself that will need to be applied once this driver goes out of staging. > > > s2250: add comment describing the hardware. > > go7007-loader: renamed from s2250-loader > > go7007-loader: add support for the other devices and move fw files > > go7007: update the README > > You need to add there: > - move cypress load firmware to drivers/media/common; > > And some note about saa7134 integration. Would it be OK if I add the saa7134 patch to the go7007 directory? Rather than keeping it around in my git tree? > > > > MAINTAINERS: add the go7007 driver. > > go7007: a small improvement to querystd handling. > > go7007: add back 'repeat sequence header' control. > > go7007: correct a header check: MPEG4 has a different GOP code. > > go7007: drop firmware name in board config, make configs const. > > go7007: don't continue if firmware can't be loaded. > > This one didn't apply. Maybe due to the lack of saa7134-go7007. > > Maybe it is just a trivial merging conflict, but better if you could > check it before forcing it. Will do. Regards, Hans -- 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
