Oops, I forgot to mention that patches 1-4 are unchanged from v1. Also, if there are no comments about this driver, then I intend to make a pull request on Monday. Regards, Hans On 05/13/15 09:22, Hans Verkuil wrote: > From: Hans Verkuil <hans.verkuil@xxxxxxxxx> > > Changes since v1: > > - Fixed a lockdep bug in the alsa driver > - Moved the DMA descriptor allocation/freeing to buf_init and buf_cleanup, > which is where it belongs. > - Rebased to the latest media_tree master. > > Hi all, > > This driver is for the Cisco Cobalt card, which is a PCIe device with four > HDMI inputs (adv7604) and optionally one fifth input (adv7842) or one output > (adv7511). > > This board is not for sale (sadly) but it is used internally for testing and > prototyping. Many of the HDMI/Digital Video related features that I have added > to V4L2 over the last few years have been prototyped using this driver and I > am planning more new features based on this board. > > During the ELC in San Jose a month back I discussed whether it would be OK to > upstream this driver, even though the hardware is not for sale. Mauro had no > problem with this and given the fact that this driver is a good starting > point for similar HDMI hardware, and that this allows me to upstream new > API additions showing them off in this driver (so ensuring that they are > actually used somewhere), I've decided to go ahead with this. > > This patch series starts off with a few improvements to other drivers: > > The adv7842 now makes the output pixel port format configurable, just like > its cousin adv7604. Note that there is one user of adv7842_platform_data: > arch/blackfin/mach-bf609/boards/ezkit.c. However, this board code has been > broken from the beginning and nobody noticed since gcc doesn't support the > bf609. You need a custom toolchain to compile this > > I can't do anything about this, someone (Scott Jiang?) will need to fix > this. > > The next patch makes it possible to requeue buffers in vb2 from the driver. > It's a very small change, but the cobalt driver uses that while it is > waiting for a stable video signal. > > The next patch is from Jean-Michel Hautbois which hasn't been merged yet > since no driver used that event, but in the next patch I implement it. > > The final patch is the cobalt driver itself. > > Note that the m00* headers are generated from our FPGA code (slightly > cleaned up by hand), which is why there are many lines > 80 columns. > It makes sense in this case and it does not affect the readability, and > I don't want to edit them too much since that would make it hard to handle > when they are regenerated due to FPGA changes. > > And there are also a lot of volatile __iomem pointers: the memory-mapped > registers are written and read directly using struct pointers, so you really > need volatile __iomem there. > > In the near future I plan on added CEC support (once the CEC framework has > been merged), colorspace conversion support, possibly deep color support and > more. But let's try to get this driver in first. > > Regards, > > Hans > > Hans Verkuil (4): > adv7842: Make output format configurable through pad format operations > vb2: allow requeuing buffers while streaming > adv7604/adv7842: replace FMT_CHANGED by V4L2_DEVICE_NOTIFY_EVENT > cobalt: add new driver > > jean-michel.hautbois@xxxxxxxxxxx (1): > v4l2-subdev: allow subdev to send an event to the v4l2_device notify > function > > Documentation/video4linux/v4l2-framework.txt | 4 + > MAINTAINERS | 8 + > drivers/media/i2c/adv7604.c | 12 +- > drivers/media/i2c/adv7842.c | 280 ++++- > drivers/media/pci/Kconfig | 1 + > drivers/media/pci/Makefile | 1 + > drivers/media/pci/cobalt/Kconfig | 18 + > drivers/media/pci/cobalt/Makefile | 5 + > drivers/media/pci/cobalt/cobalt-alsa-main.c | 162 +++ > drivers/media/pci/cobalt/cobalt-alsa-pcm.c | 603 ++++++++++ > drivers/media/pci/cobalt/cobalt-alsa-pcm.h | 22 + > drivers/media/pci/cobalt/cobalt-alsa.h | 41 + > drivers/media/pci/cobalt/cobalt-cpld.c | 341 ++++++ > drivers/media/pci/cobalt/cobalt-cpld.h | 29 + > drivers/media/pci/cobalt/cobalt-driver.c | 821 +++++++++++++ > drivers/media/pci/cobalt/cobalt-driver.h | 377 ++++++ > drivers/media/pci/cobalt/cobalt-flash.c | 132 ++ > drivers/media/pci/cobalt/cobalt-flash.h | 29 + > drivers/media/pci/cobalt/cobalt-i2c.c | 396 ++++++ > drivers/media/pci/cobalt/cobalt-i2c.h | 25 + > drivers/media/pci/cobalt/cobalt-irq.c | 254 ++++ > drivers/media/pci/cobalt/cobalt-irq.h | 25 + > drivers/media/pci/cobalt/cobalt-omnitek.c | 341 ++++++ > drivers/media/pci/cobalt/cobalt-omnitek.h | 62 + > drivers/media/pci/cobalt/cobalt-v4l2.c | 1260 ++++++++++++++++++++ > drivers/media/pci/cobalt/cobalt-v4l2.h | 22 + > .../cobalt/m00233_video_measure_memmap_package.h | 115 ++ > .../pci/cobalt/m00235_fdma_packer_memmap_package.h | 44 + > .../media/pci/cobalt/m00389_cvi_memmap_package.h | 59 + > .../media/pci/cobalt/m00460_evcnt_memmap_package.h | 44 + > .../pci/cobalt/m00473_freewheel_memmap_package.h | 57 + > .../m00479_clk_loss_detector_memmap_package.h | 53 + > .../m00514_syncgen_flow_evcnt_memmap_package.h | 88 ++ > drivers/media/v4l2-core/videobuf2-core.c | 11 +- > include/media/adv7604.h | 1 - > include/media/adv7842.h | 92 +- > include/media/v4l2-subdev.h | 2 + > 37 files changed, 5743 insertions(+), 94 deletions(-) > create mode 100644 drivers/media/pci/cobalt/Kconfig > create mode 100644 drivers/media/pci/cobalt/Makefile > create mode 100644 drivers/media/pci/cobalt/cobalt-alsa-main.c > create mode 100644 drivers/media/pci/cobalt/cobalt-alsa-pcm.c > create mode 100644 drivers/media/pci/cobalt/cobalt-alsa-pcm.h > create mode 100644 drivers/media/pci/cobalt/cobalt-alsa.h > create mode 100644 drivers/media/pci/cobalt/cobalt-cpld.c > create mode 100644 drivers/media/pci/cobalt/cobalt-cpld.h > create mode 100644 drivers/media/pci/cobalt/cobalt-driver.c > create mode 100644 drivers/media/pci/cobalt/cobalt-driver.h > create mode 100644 drivers/media/pci/cobalt/cobalt-flash.c > create mode 100644 drivers/media/pci/cobalt/cobalt-flash.h > create mode 100644 drivers/media/pci/cobalt/cobalt-i2c.c > create mode 100644 drivers/media/pci/cobalt/cobalt-i2c.h > create mode 100644 drivers/media/pci/cobalt/cobalt-irq.c > create mode 100644 drivers/media/pci/cobalt/cobalt-irq.h > create mode 100644 drivers/media/pci/cobalt/cobalt-omnitek.c > create mode 100644 drivers/media/pci/cobalt/cobalt-omnitek.h > create mode 100644 drivers/media/pci/cobalt/cobalt-v4l2.c > create mode 100644 drivers/media/pci/cobalt/cobalt-v4l2.h > create mode 100644 drivers/media/pci/cobalt/m00233_video_measure_memmap_package.h > create mode 100644 drivers/media/pci/cobalt/m00235_fdma_packer_memmap_package.h > create mode 100644 drivers/media/pci/cobalt/m00389_cvi_memmap_package.h > create mode 100644 drivers/media/pci/cobalt/m00460_evcnt_memmap_package.h > create mode 100644 drivers/media/pci/cobalt/m00473_freewheel_memmap_package.h > create mode 100644 drivers/media/pci/cobalt/m00479_clk_loss_detector_memmap_package.h > create mode 100644 drivers/media/pci/cobalt/m00514_syncgen_flow_evcnt_memmap_package.h > -- 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
