- 02 Aug, 2018 6 commits
-
-
Todor Tomov authored
of_graph_get_next_endpoint increases the refcount of the returned node and decreases the refcount of the passed node. Take this into account and use of_node_put properly. Signed-off-by:
Todor Tomov <todor.tomov@linaro.org> Signed-off-by:
Hans Verkuil <hansverk@cisco.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
-
Todor Tomov authored
Use SPDX license headers for all files of the Qualcomm CAMSS driver. Signed-off-by:
-
Todor Tomov authored
Support for camera subsystem on QComm MSM8996/APQ8096 is to be added so remove hardware version from CAMSS driver's path. Signed-off-by:
-
Todor Tomov authored
The new format will be called V4L2_PIX_FMT_Y10P. It is similar to the V4L2_PIX_FMT_SBGGR10P family formats but V4L2_PIX_FMT_Y10P is a grayscale format. Signed-off-by:
Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by:
-
Todor Tomov authored
The code will be called MEDIA_BUS_FMT_Y10_2X8_PADHI_LE. It is similar to MEDIA_BUS_FMT_SBGGR10_2X8_PADHI_LE but MEDIA_BUS_FMT_Y10_2X8_PADHI_LE describes grayscale data. Signed-off-by:
-
Sakari Ailus authored
These formats are compressed 14-bit raw bayer formats with four different pixel orders. They are similar to 10-bit variants. The formats added by this patch are V4L2_PIX_FMT_SBGGR14P V4L2_PIX_FMT_SGBRG14P V4L2_PIX_FMT_SGRBG14P V4L2_PIX_FMT_SRGGB14P Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
-
- 30 Jul, 2018 29 commits
-
-
Robert Schlabbach authored
This patch disables the null packet filter for the Hauppauge WinTV-dualHD. There are applications which require the unfiltered transport stream (e.g. DOCSIS segment load analyzers). Tests showed that the device is capable of delivering two unfiltered EuroDOCSIS 3.0 transport streams simultaneously, i.e. over 100 Mbit/s worth of data, without any losses. Signed-off-by:
Robert Schlabbach <Robert.Schlabbach@gmx.net> Signed-off-by:
-
Daniel Scheller authored
Add a break statement in set_params() for the SYS_DVBT(2). As reported by gcc: drivers/media/dvb-frontends/tda18271c2dd.c:1144:3: warning: this statement may fall through [-Wimplicit-fallthrough=] There is a nested switch() inside the code with sets the tuner to the right standard. Without the break, the code will always set to DVB-C mode, with can be sub-optimal for DVB-T. Signed-off-by:Daniel Scheller <d.scheller@gmx.net> Signed-off-by:
-
Daniel Scheller authored
Take note that the SX8 IQ mode is only available on a single tuner, and remove the MCI/SX8 DIAG CMD defines. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Allow for tuning to transponders with specific modulations in set_parameters(). Setting a specific modulation will also enable lower modulations. Picked up from the upstream dddvb GIT. Upstream also has support for APSK64/128/256 modulations which aren't supported yet by the DVB API, so comment them out until support for them is added. Signed-off-by:
-
Daniel Scheller authored
The SX8 cards by default do an automatic search for the PLS code. This is not necessarily wanted as this can eventually be detected wrong, so disable this. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
The MCI can report the command status more finegrained, so, add more status code defines and update the MCI_SUCCESS macro. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Split off all code specific to the MaxSX8 cards to a separate ddbridge-sx8 module and hook it up in the Makefile. This also adds evaluation of the mci_type to allow for using different attach handling for different cards. As different cards can implement things differently (ie. support differing frontend_ops, and have different base structs being put ontop of the common mci_base struct), this introduces the mci_cfg struct which is initially used to hold a few specifics to the -sx8 submodule. While at it, the handling of the i/q mode is adjusted slightly. Besides this and handling mci_base and sx8_base struct pointers where needed, all code is copied unmodified from ddbridge-mci.c. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
In preparation for splitting all MaxSX8 related code parts from the common MCI code, prefix both mci_cmd() and mci_config() functions with ddb_, remove the static marking and add matching function prototypes to ddbridge-mci.h so these functions can be reused from other files within the ddbridge driver. As this requires the mci-related structs to be defined in ddbridge-mci.h, move struct mci and struct mci_base there and clean them up. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
For better support for future MCI based cards, rename the mci struct member to mci_ports to carry the number of ports on the cards, and add a mci_type member to identify the card type to handle differing hardware. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Recent FPGA firmware reports more data and values in sent command responses. Adjust the mci_command and mci_result structs including it's unions to match these changes and add a few comments explaining things. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Adjustments to match the FPGA firmware, and the signal I/Q values are reported as s16 types from the card firmware. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Implement querying signal statistics from the MCI and report this data in read_status() as DVBv5 statistics. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Update the copyright year information in the MCI headers to 2017-2018. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Change the start of the MCI ID range (internally used only) to 48 and define an ID for the SX8 card type. Use this new ID to handle device attachment. This change is done in preparation for support of more MCI based cards. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Throughout the function, dev->link[l] is used several times. Unclutter this a bit by declaring a ddb_link var at the top of the function, assign the address of dev->link[l] to it and use that var to access the link[] struct member. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
ddbridge has a few MDIO related remainders (defines, hwinfo struct) which aren't of any use for the in-kernel driver at all (they're only used in conjunction with the OctoNet SAT>IP boxes which the kernel driver doesn't have any support for), so clean this up. Signed-off-by:
-
Daniel Scheller authored
The I2C_COMMAND response reports an error in the I2C bus communication using bit 17. Evaluate the response more thoroughly and log an error if an I2C problem was detected. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
Devices supporting dummy tuner operation can exist on any link, not only on link 0, so fix this accordingly. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Daniel Scheller authored
In demod_attach_stv0910(), the LNBH25 IC is being blindly attached and, if the result is bad, blindly attached on another possible I2C address. The LNBH25 uses it's set_voltage function to test for the IC and will print an error to the kernel log on failure. Prevent this by probing the possible I2C address and use this (and only this) to attach the LNBH25 I2C driver. This also allows the stv0910 attach function to be a bit cleaner. Picked up from the upstream dddvb GIT and adapted for the LNBH25 driver variant from the kernel tree. Signed-off-by:
-
Daniel Scheller authored
To avoid miscalculations related to the BER denominator, the shift expression needs to be casted as ULL. Picked up from the upstream dddvb GIT. Signed-off-by:
-
Mauro Carvalho Chehab authored
There are a number of other ioctls that aren't used anywhere inside the Kernel tree. Get rid of them. Signed-off-by: -
Mauro Carvalho Chehab authored
There are a number of other ioctls that aren't used anywhere inside the Kernel tree. Get rid of them. Signed-off-by: -
Mauro Carvalho Chehab authored
While this ioctl is there at least since Kernel 2.6.12-rc2, it was never used by any upstream driver. Get rid of it. Signed-off-by: -
Robert Schlabbach authored
The em28xx driver never touched the EM2874 register bits that control the transport stream packet filters, leaving them at whatever default the firmware has set. E.g. the Pinnacle 290e disables them by default, while the Hauppauge WinTV dualHD enables discarding NULL packets by default. However, some applications require NULL packets, e.g. to determine the load in DOCSIS segments, so discarding NULL packets is undesired for such applications. This patch simply extends the bit mask when starting or stopping the transport stream packet capture, so that the filter bits are cleared. It has been verified that this makes the Hauppauge WinTV dualHD pass an unfiltered DVB-C stream including NULL packets, which it didn't before. Signed-off-by:
-
Akihiro Tsukada authored
add entries for the following drivers: - earth_pt{1,3} DVB adapter drivers - mxl301rf DVB tuner drivers - qm1d1{b0004, c0042} DVB tuner drivers - tc90522 DVB demod driver Signed-off-by:Akihiro Tsukada <tskd08@gmail.com> Signed-off-by:
-
Akihiro Tsukada authored
Found and reported by kbuild test robot: Message ID: <201805052003.MC007f9h%fengguang.wu@intel.com> and holding an address of an empty struct in .driver.pm does no harm when CONFIG_PM_SLEEP is not defined. Signed-off-by:
-
Akihiro Tsukada authored
dvb-pll module was 'put' twice on exit: once by dvb_frontend_detach() and another by dvb_module_release(). Signed-off-by:
-
Akihiro Tsukada authored
Friio device contains "gl861" bridge and "tc90522" demod, for which the separate drivers are already in the kernel. But friio driver was monolithic and did not use them, practically copying those features. This patch decomposes friio driver into sub drivers and re-uses existing ones, thus reduces some code. It adds some features to gl861, to support the friio-specific init/config of the devices and implement i2c communications to the tuner via demod with USB vendor requests. [mchehab+samsung@kernel.org: fix merge conflicts] Signed-off-by:
-
Corentin Labbe authored
The include/linux/platform_data/media/sii9234.h header file is unused, let's remove it. Signed-off-by:
Corentin Labbe <clabbe@baylibre.com> Signed-off-by:
-
- 27 Jul, 2018 5 commits
-
-
Sylwester Nawrocki authored
Look up of buffers in s5p_mfc_handle_frame_new, s5p_mfc_handle_frame_copy_time functions is not working properly for DMA addresses above 2 GiB. As a result flags and timestamp of returned buffers are not set correctly and it breaks operation of GStreamer/OMX plugins which rely on the CAPTURE buffer queue flags. Due to improper return type of the get_dec_y_adr, get_dspl_y_adr callbacks and sign bit extension these callbacks return incorrect address values, e.g. 0xfffffffffefc0000 instead of 0x00000000fefc0000. Then the statement: "if (vb2_dma_contig_plane_dma_addr(&dst_buf->b->vb2_buf, 0) == dec_y_addr)" is always false, which breaks looking up capture queue buffers. To ensure proper matching by address u32 type is used for the DMA addresses. This should work on all related SoCs, since the MFC DMA address width is not larger than 32-bit. Changes done in this patch are minimal as there is a larger patch series pending refactoring the whole driver. Signed-off-by:
Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by:
-
Ezequiel Garcia authored
Commit 34dbb848 ("media: mem2mem: Remove excessive try_run call") removed a redundant call to v4l2_m2m_try_run but instead introduced a bug. Consider the following case: 1) Context A schedules, queues and runs job A. 2) While the m2m device is running, context B schedules and queues job B. Job B cannot run, because it has to wait for job A. 3) Job A completes, calls v4l2_m2m_job_finish, and tries to queue a job for context A, but since the context is empty it won't do anything. In this scenario, queued job B will never run. Fix this by calling v4l2_m2m_try_run from v4l2_m2m_try_schedule. While here, add more documentation to these functions. Fixes: 34dbb848 ("media: mem2mem: Remove excessive try_run call") Signed-off-by:
Ezequiel Garcia <ezequiel@collabora.com> Signed-off-by:
-
Luca Ceresoli authored
Currently this driver does not support cropping. The supported modes are the following, all capturing the entire area: - 3840x2160, 1:1 binning (native sensor resolution) - 1920x1080, 2:1 binning - 1280x720, 3:1 binning The VIDIOC_SUBDEV_S_FMT ioctl chooses among these 3 configurations the one that matches the requested format. Add cropping support via VIDIOC_SUBDEV_S_SELECTION: with target V4L2_SEL_TGT_CROP to choose the captured area, with V4L2_SEL_TGT_COMPOSE to choose the output resolution. To maintain backward compatibility we also allow setting the compose format via VIDIOC_SUBDEV_S_FMT. To obtain this, compose rect and output format are computed in the common helper function __imx274_change_compose(), which sets both to the same width/height values. Cropping also calls __imx274_change_compose() whenever cropping rect size changes in order to reset the compose rect (and output format size) for 1:1 binning. Also rename enum imx274_mode to imx274_binning (and its values from IMX274_MODE_BINNING_* to IMX274_BINNING_*). Without cropping, the two naming are equivalent. With cropping, the resolution could be different, e.g. using 2:1 binning mode to crop 1200x960 and output a 600x480 format. Using binning in the names avoids any misunderstanding. For the same reason, replace the 'size' field in struct imx274_frmfmt with 'bin_ratio'. [Sakari Ailus: Remove leftover condition in imx274_apply_trimming] Signed-off-by:
Luca Ceresoli <luca@lucaceresoli.net> Signed-off-by:
-
Luca Ceresoli authored
Currently 2-bytes and 3-bytes registers are set by very similar functions doing the needed shift & mask manipulation, followed by very similar for loops setting one byte at a time over I2C. Replace all of this code by a unique helper function that calls regmap_bulk_write(), which has two advantages: - sets all the bytes in a unique I2C transaction - removes lots of now unused code. Signed-off-by:
-
Rui Miguel Silva authored
This patch adds V4L2 sub-device driver for OV2680 image sensor. The OV2680 is a 1/5" CMOS color sensor from Omnivision. Supports output format: 10-bit Raw RGB. The OV2680 has a single lane MIPI interface. The driver exposes following V4L2 controls: - auto/manual exposure, - exposure, - auto/manual gain, - gain, - horizontal/vertical flip, - test pattern menu. Supported resolution are only: QUXGA, 720P, UXGA. [Sakari Ailus: Drop "-level" from Kconfig help text] Signed-off-by:
Rui Miguel Silva <rui.silva@linaro.org> Signed-off-by:
-
