- 23 Dec, 2014 40 commits
-
-
Benoit Parrot authored
This patch adds Video Processing Front End (VPFE) driver for AM437X family of devices Driver supports the following: - V4L2 API using MMAP buffer access based on videobuf2 api - Asynchronous sensor/decoder sub device registration - DT support Signed-off-by:
Benoit Parrot <bparrot@ti.com> Signed-off-by:
Darren Etheridge <detheridge@ti.com> Signed-off-by:
Lad, Prabhakar <prabhakar.csengg@gmail.com> [hans.verkuil@cisco.com: swapped two lines to fix vpfe_release() & add pinctrl include] Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@osg.samsung.com>
-
Sakari Ailus authored
This makes the macros a little bit safer. Signed-off-by:
Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by:
-
Sakari Ailus authored
The pll_flags quirk just returned the extra PLL flags the sensor required, but the init quirk is far more versatile. It can be used to perform any extra initialisation needed by the sensor, including allocating memory for sensor specific struct and creating sensor specific new controls. Signed-off-by:
-
Sakari Ailus authored
As the controls creation is separated in two sections, the available media bus codes and link frequencies can be enumerated later on. Signed-off-by:
-
Sakari Ailus authored
The default values and limits for certain controls need the knowledge of available media bus codes or link frequencies. Create such controls later on, so that most of the initialisation of the sensor has already been done when the init quirk is called. [mchehab@osg.samsung.com: make checkpatch.pl happier] Signed-off-by:
-
Sakari Ailus authored
Clean up smiapp_init_controls() by adding newlines to appropriate places and by removing superfluous error handling. The caller will clean up control handlers in any case if the function fails. Signed-off-by:
-
Sakari Ailus authored
Also call smiapp_cleanup() in smiapp_remove(), replacing code that did the same than the function. Signed-off-by:
-
Sakari Ailus authored
The flash capability register is already read as part of the limit registers. Do no access it separately; instead use the value from the limits. Signed-off-by:
-
Sakari Ailus authored
In the case of platform data, ISPs that provide clocks to the sensor must probe before the sensor does. Accessing the sensor does require the clocks, and thus, probe cannot access the sensor in such a system. This limitation does not exist in the case of the DT. Perform all initialisation except Media entity initialisation, link creation and sub-device registration in probe. Signed-off-by:
Sakari Ailus <sakari.ailus@iki.fi> Signed-off-by:
-
Sakari Ailus authored
The registered callback is called by the V4L2 async framework after the bound callback. This allows separating the functionality in the registered callback so that on DT based systems only sub-device registration is done there. Signed-off-by:
-
Sakari Ailus authored
Platform data support is retained. of_property_read_u64_array() isn't used yet as it's not in yet. Signed-off-by:
-
Sakari Ailus authored
Document the smiapp device tree properties. Signed-off-by:
-
Sakari Ailus authored
link-frequencies is a 64-bit unsigned integer array of allowed link frequencies. Signed-off-by:
-
Sakari Ailus authored
The SMIA compatible sensors only need a single clock. Signed-off-by:
-
Sakari Ailus authored
Register and unregister async sub-device for DT. Signed-off-by:
-
Sakari Ailus authored
The source sub-device's name will be overwritten shortly. Don't give it a name in the meantime. Signed-off-by:
-
Sakari Ailus authored
Signed-off-by:
-
Sakari Ailus authored
struct device has a forward declaration in the header already. The header is only needed in the .c file. Signed-off-by:
-
Sakari Ailus authored
This is part of the smiapp driver. Signed-off-by:
-
Sakari Ailus authored
Remove FSF's address information from the license header in the smiapp driver and the smiapp-pll PLL calculator. This should no longer be needed, and would be rendered outdated in case the FSF chooses to relocate its office. Signed-off-by:
-
Aviv Greenberg authored
These formats are just like 10-bit raw bayer formats that exist already, but the pixels are not padded to byte boundaries. Instead, the eight high order bits of four consecutive pixels are stored in four bytes, followed by a byte of two low order bits of each of the four pixels. Signed-off-by:
Aviv Greenberg <aviv.d.greenberg@intel.com> Signed-off-by:
-
Sakari Ailus authored
Rearrange 12-bit raw bayer format definitions after 10-bit ones. Also remove the comment related to 16-bit bayer formats, it was simply wrong. Signed-off-by:
-
Sakari Ailus authored
The documentation began with "The following four pixel formats"... but the format definitions preceded this sentence. Replace it with "These four pixel formats". Signed-off-by:
-
Geert Uytterhoeven authored
If VIDEO_CAFE_CCIC=y, but VIDEOBUF2_DMA_SG=m: drivers/built-in.o: In function `mcam_v4l_open': mcam-core.c:(.text+0x1c2e81): undefined reference to `vb2_dma_sg_memops' mcam-core.c:(.text+0x1c2eb0): undefined reference to `vb2_dma_sg_init_ctx' drivers/built-in.o: In function `mcam_v4l_release': mcam-core.c:(.text+0x1c34bf): undefined reference to `vb2_dma_sg_cleanup_ctx' Reported-by:
Fengguang Wu <fengguang.wu@intel.com> Signed-off-by:
Geert Uytterhoeven <geert@linux-m68k.org> Acked-by:
Jonathan Corbet <corbet@lwn.net> Signed-off-by:
-
Shuah Khan authored
This driver supports VBI and the comment "VBI support is not yet working" is inaccurate. Remove it. Signed-off-by:
Shuah Khan <shuahkh@osg.samsung.com> Signed-off-by:
-
Hans Verkuil authored
drivers/media/i2c/s5k5baf.c:1796:33: warning: duplicate const drivers/media/i2c/s5k5baf.c:379:24: warning: cast to restricted __le16 drivers/media/i2c/s5k5baf.c:437:11: warning: incorrect type in assignment (different base types) drivers/media/i2c/s5k5baf.c:445:16: warning: incorrect type in return expression (different base types) Signed-off-by:
Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by:
-
Hans Verkuil authored
drivers/media/i2c/s5k6aa.c:351:16: warning: cast to restricted __be16 drivers/media/i2c/s5k6aa.c:351:16: warning: cast to restricted __be16 drivers/media/i2c/s5k6aa.c:351:16: warning: cast to restricted __be16 drivers/media/i2c/s5k6aa.c:351:16: warning: cast to restricted __be16 Signed-off-by:
-
Hans Verkuil authored
drivers/media/i2c/s5k4ecgx.c:223:16: warning: cast to restricted __be16 drivers/media/i2c/s5k4ecgx.c:223:16: warning: cast to restricted __be16 drivers/media/i2c/s5k4ecgx.c:223:16: warning: cast to restricted __be16 drivers/media/i2c/s5k4ecgx.c:223:16: warning: cast to restricted __be16 drivers/media/i2c/s5k4ecgx.c:344:20: warning: cast to restricted __le32 drivers/media/i2c/s5k4ecgx.c:354:20: warning: cast to restricted __le32 drivers/media/i2c/s5k4ecgx.c:364:24: warning: cast to restricted __le32 drivers/media/i2c/s5k4ecgx.c:366:23: warning: cast to restricted __le16 The get_unaligned_le*() functions return the value using cpu endianness, so calling le*_to_cpu is wrong. It hasn't been not noticed because this code has only been run on little endian systems, so le*_to_cpu doesn't do anything. Signed-off-by:
-
Hans Verkuil authored
drivers/media/i2c/m5mols/m5mols_core.c:128:24: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:128:24: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:128:24: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:128:24: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:130:24: warning: cast to restricted __be32 drivers/media/i2c/m5mols/m5mols_core.c:130:24: warning: cast to restricted __be32 drivers/media/i2c/m5mols/m5mols_core.c:130:24: warning: cast to restricted __be32 drivers/media/i2c/m5mols/m5mols_core.c:130:24: warning: cast to restricted __be32 drivers/media/i2c/m5mols/m5mols_core.c:130:24: warning: cast to restricted __be32 drivers/media/i2c/m5mols/m5mols_core.c:130:24: warning: cast to restricted __be32 drivers/media/i2c/m5mols/m5mols_core.c:457:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:457:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:457:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:457:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:458:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:458:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:458:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:458:19: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:459:22: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:459:22: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:459:22: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:459:22: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:460:20: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:460:20: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:460:20: warning: cast to restricted __be16 drivers/media/i2c/m5mols/m5mols_core.c:460:20: warning: cast to restricted __be16 The be16_to_cpu conversions in m5mols_get_version() are not needed since the data is already using cpu endianness. This was never noticed since these version fields are never used. Signed-off-by:
-
Hans Verkuil authored
drivers/media/dvb-frontends/hd29l2.c:29:18: warning: Variable length array is used. drivers/media/dvb-frontends/hd29l2.c:34:32: error: cannot size expression drivers/media/dvb-frontends/hd29l2.c:125:5: warning: symbol 'hd29l2_rd_reg_mask' was not declared. Should it be static? Variable length arrays are frowned upon, so replace with a fixed length and check that there won't be a buffer overrun. Signed-off-by:
Antti Palosaari <crope@iki.fi> Signed-off-by:
-
Hans Verkuil authored
Fix this warning: drivers/media/v4l2-core/videobuf2-vmalloc.c:98:28: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/videobuf2-vmalloc.c:158:28: warning: incorrect type in argument 1 (different address spaces) The warning is correct, but we have no other choice here to forcibly cast. At least it is now explicit that such a cast is needed. Signed-off-by:
-
Hans Verkuil authored
Fix this warning: drivers/media/pci/ivtv/ivtv-irq.c:418:9: warning: context imbalance in 'ivtv_dma_stream_dec_prepare' - different lock contexts for basic block sparse didn't quite understand the locking scheme, so rewrite it to keep sparse happy. Signed-off-by:
-
Hans Verkuil authored
Fixes these sparse warnings. drivers/media/pci/ttpci/budget-core.c:250:17: warning: context imbalance in 'ttpci_budget_debiread' - different lock contexts for basic block drivers/media/pci/ttpci/budget-core.c:289:17: warning: context imbalance in 'ttpci_budget_debiwrite' - different lock contexts for basic block To be honest, the new code does look better than the old. Signed-off-by:
-
Hans Verkuil authored
drivers/media/pci/ttpci/av7110.c:1226:15: warning: memset with byte count of 192512 Instead of memsetting this in one go, loop over each line and memset each line separately. Signed-off-by:
-
Hans Verkuil authored
The debug attribute in /sys/class/video4linux/<devX>/debug was never documented. Add this. Signed-off-by:
-
Hans Verkuil authored
The old debug field is renamed to dev_debug to ensure that existing drivers (including out-of-tree drivers) that try to use the old name will no longer compile. A comment has also been added that makes it explicit that drivers shouldn't use this field. Additional bits have been added to the debug flag to be more fine-grained when debugging, especially when dealing with streaming ioctls and read, write and poll. You want to enable those explicitly to prevent flooding the log when streaming unless you actually want to do that. Signed-off-by:
-
Hans Verkuil authored
The debug field in struct video_device is for internal use only and drivers should mix that with their own debug module options. It is handled by the V4L2 core and users can set it using /sys/class/video4linux/<devX>/debug. It has been deprecated for some time now, so it is time to remove it completely from the drivers. Signed-off-by:
-
Andrey Utkin authored
[media] solo6x10: just pass frame motion flag from hardware, drop additional handling as complicated and unstable Dropping code (introduced in 316d9e84) which intends to make raising of motion events more "smooth"(?). It made motion event never appear in my installation. That code is complicated, so I couldn't figure out quickly how to fix it, so dropping it seems better to me. Another justification is that anyway application would implement "motion signal stabilization" if required, it is not necessarily kernel driver's job. Signed-off-by:
Andrey Utkin <andrey.utkin@corp.bluecherry.net> Signed-off-by:
-
Hans Verkuil authored
The video output functionality never worked for this driver. Now remove the creation of the output video nodes as well to prevent users from thinking that video output is available, when it isn't. To correctly implement this the video output should use vb2 as well, and that requires rewriting the output DMA setup. But without hardware to test I am not able to do that. Signed-off-by:
-
Hans Verkuil authored
Add support for the VIDIOC_CREATE_BUFS ioctl. This was missing in this driver and in vb2 it's trivial to add. Signed-off-by:
-
