- 19 Sep, 2009 40 commits
-
-
Muralidharan Karicheri authored
Adds Kconfig and Makefile changes required for vpif capture driver Resending to merge to V4L linux-next Reviewed-by:
Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by:
Muralidharan Karicheri <m-karicheri2@ti.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@redhat.com>
-
Muralidharan Karicheri authored
The structure name for vpif display driver changed since it was not unique. So this update is done to reflect the same. Also removed the code related to register address space iomap. Uses v4l2_i2c_new_subdev_board() instead of v4l2_i2c_new_probed_subdev() so that platform data can be added for subdevice configuration for polarities. This has incorporated comments against version v0 of the patch series. Resending the original patch for merge to V4L linux-next Reviewed-by:
-
Hans Verkuil authored
Warn when the desired device node number is already in use, except when the new video_register_device_no_warn function is called since in some use-cases that warning is not relevant. Signed-off-by:
-
Hans Verkuil authored
There are some subtle differences in the way the devnode numbers are handled depending on whether the FIXED_MINOR_RANGES config option is set. Add some simple wrapper functions to handle that correctly. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
The term 'kernel number' is very vague, so replace it with the somewhat more descriptive term 'device node number'. In one place the local variable 'nr' was used to create the device node number of the new device name. This has been replaced with the vdev->num field to more clearly mark this as being the device node number and not the minor number. Signed-off-by:
-
Hans Verkuil authored
video_register_device_index is never actually called, instead the stream index number is always calculated automatically. This patch removes this function and simplifies the internal get_index function since that can now always just return the first free index. Signed-off-by:
-
Hans Verkuil authored
Rewrite v4l2_i2c_new_subdev as a simplified version of v4l2_i2c_new_subdev_cfg and remove v4l2_i2c_new_probed_subdev and v4l2_i2c_new_probed_subdev_addr. This simplifies this API substantially. Signed-off-by:
-
Guennadi Liakhovetski authored
Since we are not using v4l2_device_call_* calls any more, we don't need to initialise subdevice .grp_id any more. This also fixes compiler warnings on 64-bit platforms. Signed-off-by:
Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by:
-
Guennadi Liakhovetski authored
This makes the soc-camera interface for V4L2 subdevices thinner yet. Handle gain and exposure internally in each driver just like all other controls. Signed-off-by:
-
Guennadi Liakhovetski authored
Remove unneeded soc-camera operations, this also makes the soc-camera API to v4l2 subdevices thinner. Signed-off-by:
-
Guennadi Liakhovetski authored
The initial soc-camera scaling and cropping implementation turned out to be incompliant with the V4L2 API, e.g., it expected the user to specify cropping in output window pixels, instead of input window pixels. This patch converts the soc-camera core and all drivers to comply with the standard. Signed-off-by:
-
Guennadi Liakhovetski authored
Signed-off-by:
-
Guennadi Liakhovetski authored
Signed-off-by:
-
Guennadi Liakhovetski authored
Signed-off-by:
-
Guennadi Liakhovetski authored
Use v4l2_subdev_call() instead of v4l2_device_call_until_err() in all host drivers and in soc-camera core. Signed-off-by:
-
Guennadi Liakhovetski authored
Remove set_crop soc-camera device method and switch to s_crop from v4l2-subdev video operations. Also extend non-i2c drivers to also hold a pointer to their v4l2-subdev instance in control device driver-data, i.e., in dev_get_drvdata((struct device *)to_soc_camera_control(icd)) Signed-off-by:
-
Guennadi Liakhovetski authored
Use host-side image scaling when the client fails to set the requested format. We also have to take scaling into account when performing host-side cropping. Similar to cropping we try to use client-side scaling as much as possible to preserve bus bandwidth and optimise the frame-rate. Signed-off-by:
-
Guennadi Liakhovetski authored
The current tw9910 driver does not implement cropping correctly. Therefore, and also because various rectangles in struct soc_camera_device are in user scale, we cannot and shall not use rect_current as window location. Signed-off-by:
-
Guennadi Liakhovetski authored
OV772x sensors cannot crop, they only support two fixed formats: VGA and QVGA. We should not change the format when requested to crop, only S_FMT can do this. Signed-off-by:
-
Guennadi Liakhovetski authored
Prohibit S_FMT and S_CROP with a different window width or height after video buffer initialisation. This simplifies the work to be done in specific host and client drivers, and it doesn't seem to make much sense to allow these changes. We do however allow S_CROP with equal width and height to just move the window, this doesn't affect video buffer management and is usually easy enough to implement. Signed-off-by:
-
Guennadi Liakhovetski authored
V4L2 API mandates, that drivers do not update the argument of the S_CROP ioctl() with the actual geometry. Comply. Signed-off-by:
-
Guennadi Liakhovetski authored
Signed-off-by:
-
Guennadi Liakhovetski authored
Not all video capture devices can configure arbitrary cropping, whereas the CEU module on SuperH CPUs can crop with pixel precision. However, we want to use camera cropping if possible to save bandwidth and increase the frame-rate. This patch verifies whether the camera managed to crop exactly the requested rectangle, and if not, uses host-side cropping. To be able to crop on CEU we have to preserve camera rectangle too, for which the host_priv member in struct soc_camera_device is used. We now allocate memory dynamically, thus we have to use the .put_formats() method from struct soc_camera_host_ops to free it. Signed-off-by:
-
Guennadi Liakhovetski authored
Remove .set_std() method from struct soc_camera_ops, use .s_std() from struct v4l2_subdev_core_ops instead. Signed-off-by:
-
Guennadi Liakhovetski authored
Patch [PATCH] video: use videobuf_waiton() in sh_mobile_ceu free_buffer() was not quite correct. It closed a race, but introduced a potential lock-up, if for some reason an interrupt does not come. This has been observed in tests with tw9910. This patch safely dequeues buffers without waiting for their completion. It also moves a buffer state assignment under a spinlock to make it atomic with queuing of the buffer. Signed-off-by:
-
Guennadi Liakhovetski authored
The move of format translation initialisation into soc_camera_open() was temporary for the soc-camera as platform driver intermediate step, put it back into soc_camera_probe(). Also add a .put_formats() method to soc_camera_host_ops to free any resources host driver might have allocated in .get_formats(). Signed-off-by:
-
Guennadi Liakhovetski authored
V4L2 drivers are allowed to configure a geometry different than what has been requested by the user with S_CROP, but then they have to adjust the input rectangle accordingly. Fix ov772x to comply with this requirement. Signed-off-by:
-
Guennadi Liakhovetski authored
1. soc-camera always requests a valid rectangle, when calling .s_fmt(), no need to check and adjust 2. in .s_crop(), if the rectangle exceeds sensor limits, push it to the respective border instead of centering 3. take into account left and top borders when checking Signed-off-by:
-
Guennadi Liakhovetski authored
Signed-off-by:
-
Guennadi Liakhovetski authored
Switch to using struct v4l2_rect in struct soc_camera_device for uniformity and simplicity. Signed-off-by:
-
Guennadi Liakhovetski authored
Use the V4L2_CID_SHARPNESS control to switch SH-mobile camera low-pass filter. Signed-off-by:
-
Guennadi Liakhovetski authored
Until now soc-camera only supported client (sensor) controls. This patch enables camera-host drivers to implement their own controls too. Signed-off-by:
-
Guennadi Liakhovetski authored
The V4L2_CID_BAND_STOP_FILTER control is used to switch the "Banding Filter" on OV772x cameras on and off and to set the minimum AEC value in BDBASE register. Signed-off-by:
-
Guennadi Liakhovetski authored
Add a new V4L2_CID_BAND_STOP_FILTER integer control, which either switches the band-stop filter off, or sets it to a certain strength. Signed-off-by:
-
Guennadi Liakhovetski authored
Convert the soc-camera framework to use the v4l2-(sub)dev API. Start using v4l2-subdev operations. Only a part of the interface between the soc_camera core, soc_camera host drivers on one side and soc_camera device drivers on the other side is replaced so far. The rest of the interface will be replaced in incremental steps, and will require extensions and, possibly, modifications to the v4l2-subdev code. Signed-off-by:
-
Guennadi Liakhovetski authored
We will be registering and unregistering the soc_camera_platform platform device multiple times, therefore we need a .release() method and have to nullify the kobj. Signed-off-by:
Paul Mundt <lethal@linux-sh.org> Signed-off-by:
-
Guennadi Liakhovetski authored
Signed-off-by:
-
Guennadi Liakhovetski authored
Update ap325rxa to specify .bus_id in struct soc_camera_link explicitly, remove unused .iface from struct soc_camera_platform_info. Signed-off-by:
-
Guennadi Liakhovetski authored
Convert soc-camera core and all drivers to platform device API. We already converted platforms to register a platform device for each soc-camera client, now we remove the compatibility code and switch completely to the new scheme. This is a preparatory step for the v4l2-subdev conversion. Signed-off-by:
-
