platform/kernel/linux-exynos.git
12 years ago[media] m5mols: Refactored controls handling
Sylwester Nawrocki [Mon, 9 Apr 2012 16:51:56 +0000 (13:51 -0300)]
[media] m5mols: Refactored controls handling

This patch is a prerequisite for the new controls addition. It consolidates
the control handling code, which is moved to m5mols_controls.c and
staticized. The controls initialization is reordered to better reflect
the control clusters and make the diffs smaller when new controls are added.
To make the code easier to follow when more controls is added use separate
set function for each control.

Rewrite the image effect registers handling.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] m5mols: Convert macros to inline functions
Sylwester Nawrocki [Tue, 6 Mar 2012 10:08:56 +0000 (07:08 -0300)]
[media] m5mols: Convert macros to inline functions

Make to_sd and to_m5mols macros static inline functions
for better type safety.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera auto focus controls
Sylwester Nawrocki [Fri, 11 May 2012 09:37:03 +0000 (06:37 -0300)]
[media] V4L: Add camera auto focus controls

Add following auto focus controls:

 - V4L2_CID_AUTO_FOCUS_START - single-shot auto focus start
 - V4L2_CID_AUTO_FOCUS_STOP -  single-shot auto focus stop
 - V4L2_CID_AUTO_FOCUS_STATUS - automatic focus status
 - V4L2_CID_AUTO_FOCUS_RANGE - automatic focus scan range selection

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera 3A lock control
Sylwester Nawrocki [Wed, 2 May 2012 09:24:33 +0000 (06:24 -0300)]
[media] V4L: Add camera 3A lock control

The V4L2_CID_3A_LOCK bitmask control allows applications to pause
or resume the automatic exposure, focus and wite balance adjustments.
It can be used, for example, to lock the 3A adjustments right before
a still image is captured, for pre-focus, etc.
The applications can control each of the algorithms independently,
through a corresponding control bit, if driver allows that.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera scene mode control
Sylwester Nawrocki [Mon, 16 Apr 2012 13:45:44 +0000 (10:45 -0300)]
[media] V4L: Add camera scene mode control

Add control for the scene mode feature available in image sensor
with more advanced ISP firmware. The V4L2_CID_SCENE_MODE menu
control allows to select a set of parameters or a specific image
processing and capture control algorithm optimized for common
image capture conditions.

Signed-off-by: HeungJun Kim <riverful.kim@samsung.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera exposure metering control
Sylwester Nawrocki [Mon, 30 Apr 2012 07:34:10 +0000 (04:34 -0300)]
[media] V4L: Add camera exposure metering control

The V4L2_CID_EXPOSURE_METERING control allows to determine
a method used by the camera for measuring the amount of light
available for automatic exposure.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera ISO sensitivity controls
Sylwester Nawrocki [Tue, 1 May 2012 20:39:45 +0000 (17:39 -0300)]
[media] V4L: Add camera ISO sensitivity controls

Add ISO sensitivity and ISO auto/manual controls. The sensitivity
values are related to level of amplification of the analog signal
between image sensor and ADC. These controls allow to support sensors
exposing an interface to accept the ISO values directly.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera image stabilization control
Sylwester Nawrocki [Tue, 1 May 2012 20:38:09 +0000 (17:38 -0300)]
[media] V4L: Add camera image stabilization control

Add V4L2_CID_IMAGE_STABILIZATION control for the camera image
stabilization feature. This control can be used to enable/disable
image stabilization. It might get converted to a menu control
in future if more options are needed.

Signed-off-by: HeungJun Kim <riverful.kim@samsung.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera wide dynamic range control
Sylwester Nawrocki [Tue, 6 Mar 2012 10:05:45 +0000 (07:05 -0300)]
[media] V4L: Add camera wide dynamic range control

Add V4L2_CID_WIDE_DYNAMIC_RANGE camera class control for the
camera wide dynamic range (WDR, HDR) feature. This control
can be used to enable/disable wide dynamic range. It might
get converted to a menu control in future if more options
are needed.

Signed-off-by: HeungJun Kim <riverful.kim@samsung.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add an extended camera white balance control
Sylwester Nawrocki [Tue, 6 Mar 2012 10:04:26 +0000 (07:04 -0300)]
[media] V4L: Add an extended camera white balance control

This patch adds V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE control which is
an extended version of the V4L2_CID_AUTO_WHITE_BALANCE control,
including white balance presets. The following presets are defined:

 - V4L2_WHITE_BALANCE_INCANDESCENT,
 - V4L2_WHITE_BALANCE_FLUORESCENT,
 - V4L2_WHITE_BALANCE_FLUORESCENT_H,
 - V4L2_WHITE_BALANCE_HORIZON,
 - V4L2_WHITE_BALANCE_DAYLIGHT,
 - V4L2_WHITE_BALANCE_FLASH,
 - V4L2_WHITE_BALANCE_CLOUDY,
 - V4L2_WHITE_BALANCE_SHADE.

Signed-off-by: HeungJun Kim <riverful.kim@samsung.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add camera exposure bias control
Sylwester Nawrocki [Tue, 6 Mar 2012 10:06:55 +0000 (07:06 -0300)]
[media] V4L: Add camera exposure bias control

The camera may in some conditions incorrectly determine the exposure,
and a manual automatic exposure correction may be needed. This patch
adds V4L2_CID_AUTO_EXPOSURE_BIAS control which allows to add some
offset in the automatic exposure control loop, to compensate for
frame under- or over-exposure.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Add helper function for standard integer menu controls
Sylwester Nawrocki [Sun, 6 May 2012 18:30:44 +0000 (15:30 -0300)]
[media] V4L: Add helper function for standard integer menu controls

This patch adds v4l2_ctrl_new_int_menu() helper function which can be used
in drivers for creating standard integer menu control with driver-specific
menu item list. It is similar to v4l2_ctrl_new_std_menu(), except it doesn't
have a mask parameter and an additional qmenu parameter allows passing
an array of signed 64-bit integers as the menu item list.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Sakari Ailus <sakari.ailus@iki.fi>
Tested-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] V4L: Extend V4L2_CID_COLORFX with more image effects
Sylwester Nawrocki [Mon, 2 Apr 2012 09:40:19 +0000 (06:40 -0300)]
[media] V4L: Extend V4L2_CID_COLORFX with more image effects

This patch adds definition of additional color effects:

 - V4L2_COLORFX_AQUA,
 - V4L2_COLORFX_ART_FREEZE,
 - V4L2_COLORFX_SILHOUETTE,
 - V4L2_COLORFX_SOLARIZATION,
 - V4L2_COLORFX_ANTIQUE,
 - V4L2_COLORFX_SET_CBCR.

The new V4L2_COLORFX_CBCR control is added to allow setting
the fixed Cb, Cr values that replace chroma Cb/Cr coefficients
in case of V4L2_COLORFX_SET_CBCR effect.

Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] radio-si470x-usb: remove autosuspend, implement suspend/resume
Hans Verkuil [Fri, 4 May 2012 12:42:29 +0000 (09:42 -0300)]
[media] radio-si470x-usb: remove autosuspend, implement suspend/resume

The radio-si470x-usb driver supported both autosuspend and it stopped the
radio the moment the last user of the radio device closed it. However, that
was very confusing since if you play the audio from the device (e.g. through
arecord -D ... | aplay) then no sound would play unless you had the radio
device open at the same time, even though there is no need to do anything
with that node.

On the other hand, the actual suspend/resume functions didn't do anything,
which would fail if you *did* have the radio node open at that time.

So:

- remove autosuspend (bad idea in general for USB radio devices)
- move the start/stop out of the open/release functions into the resume/suspend
  functions.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Tobias Lorenz <tobias.lorenz@gmx.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] radio-si470x-common.c: remove unnecessary kernel log spam
Hans Verkuil [Mon, 30 Apr 2012 22:49:27 +0000 (19:49 -0300)]
[media] radio-si470x-common.c: remove unnecessary kernel log spam

There is no need to report an error in the log, you are already returning
that error to userspace after all.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Tobias Lorenz <tobias.lorenz@gmx.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] si470x: add control event support and more v4l2 compliancy fixes
Hans Verkuil [Fri, 4 May 2012 12:20:53 +0000 (09:20 -0300)]
[media] si470x: add control event support and more v4l2 compliancy fixes

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Tobias Lorenz <tobias.lorenz@gmx.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] si470x: Clean up, introduce the control framework
Hans Verkuil [Fri, 4 May 2012 12:16:57 +0000 (09:16 -0300)]
[media] si470x: Clean up, introduce the control framework

This cleans up the code and si470x now uses the proper v4l2 frameworks
and passes most of the v4l2-compliance tests.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Tobias Lorenz <tobias.lorenz@gmx.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: v4l2-ctrls: moves the forward declaration of struct file
Laurent Pinchart [Mon, 23 Apr 2012 11:20:35 +0000 (08:20 -0300)]
[media] v4l: v4l2-ctrls: moves the forward declaration of struct file

This fixes the following warning:

In file included from drivers/media/video/v4l2-subdev.c:29:
include/media/v4l2-ctrls.h:501: warning: 'struct file' declared inside
parameter list
include/media/v4l2-ctrls.h:501: warning: its scope is only this
definition or declaration, which is probably not what you want
include/media/v4l2-ctrls.h:509: warning: 'struct file' declared inside
parameter list

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9v032: Correct the logic for the auto-exposure setting
Kartik Mohta [Wed, 2 May 2012 22:19:08 +0000 (19:19 -0300)]
[media] mt9v032: Correct the logic for the auto-exposure setting

The driver uses the ctrl value passed in as a bool to determine whether
to enable auto-exposure, but the auto-exposure setting is defined as an
enum where AUTO has a value of 0 and MANUAL has a value of 1. This leads
to a reversed logic where if you send in AUTO, it actually sets manual
exposure and vice-versa.

Signed-off-by: Kartik Mohta <kartikmohta@gmail.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9m032: use the available subdev pointer, don't re-calculate it
Guennadi Liakhovetski [Wed, 18 Apr 2012 08:00:52 +0000 (05:00 -0300)]
[media] mt9m032: use the available subdev pointer, don't re-calculate it

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9m032: fix compilation breakage
Guennadi Liakhovetski [Wed, 18 Apr 2012 07:59:01 +0000 (04:59 -0300)]
[media] mt9m032: fix compilation breakage

Fix the following compilation failure:

linux-2.6/drivers/media/video/mt9m032.c: In function '__mt9m032_get_pad_crop':
linux-2.6/drivers/media/video/mt9m032.c:337: error: implicit declaration of function 'v4l2_subdev_get_try_crop'
linux-2.6/drivers/media/video/mt9m032.c:337: warning: return makes pointer from integer without a cast
linux-2.6/drivers/media/video/mt9m032.c: In function '__mt9m032_get_pad_format':
linux-2.6/drivers/media/video/mt9m032.c:359: error: implicit declaration of function 'v4l2_subdev_get_try_format'
linux-2.6/drivers/media/video/mt9m032.c:359: warning: return makes pointer from integer without a cast
linux-2.6/drivers/media/video/mt9m032.c: In function 'mt9m032_probe':
linux-2.6/drivers/media/video/mt9m032.c:767: error: 'struct v4l2_subdev' has no member named 'entity'
linux-2.6/drivers/media/video/mt9m032.c:826: error: 'struct v4l2_subdev' has no member named 'entity'
linux-2.6/drivers/media/video/mt9m032.c: In function 'mt9m032_remove':
linux-2.6/drivers/media/video/mt9m032.c:842: error: 'struct v4l2_subdev' has no member named 'entity'
make[4]: *** [drivers/media/video/mt9m032.o] Error 1

by adding a dependency on VIDEO_V4L2_SUBDEV_API.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9m032: fix two dead-locks
Guennadi Liakhovetski [Sun, 8 Apr 2012 20:31:24 +0000 (17:31 -0300)]
[media] mt9m032: fix two dead-locks

Fix a copy-paste typo and a nested locking function call in mt9m032.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9p031: Implement black level compensation control
Laurent Pinchart [Sat, 10 Mar 2012 00:02:57 +0000 (21:02 -0300)]
[media] mt9p031: Implement black level compensation control

Add four new controls to configure black level compensation (BLC):

- V4L2_CID_BLC_AUTO selects between manual and auto BLC
- V4L2_CID_BLC_TARGET_LEVEL sets the target level for auto BLC
- V4L2_CID_BLC_ANALOG_OFFSET sets the analog offset for manual BLC
- V4L2_CID_BLC_DIGITAL_OFFSET sets the digital offset for manual BLC

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9p031: Replace the reset board callback by a GPIO number
Laurent Pinchart [Fri, 9 Mar 2012 13:59:41 +0000 (10:59 -0300)]
[media] mt9p031: Replace the reset board callback by a GPIO number

Use the GPIO from the sensor driver instead of calling back to board
code.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] mt9p031: Identify color/mono models using I2C device name
Laurent Pinchart [Fri, 9 Mar 2012 13:42:52 +0000 (10:42 -0300)]
[media] mt9p031: Identify color/mono models using I2C device name

Instead of passing a color/monochrome flag through platform data, rely
on the I2C device name to identify the chip model.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9035: add few new reference design USB IDs
Antti Palosaari [Mon, 7 May 2012 18:16:36 +0000 (15:16 -0300)]
[media] af9035: add few new reference design USB IDs

Add all known reference design USB IDs.
Rename two earlier reference design USB IDs.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9035: remove one config parameter
Antti Palosaari [Mon, 7 May 2012 17:59:55 +0000 (14:59 -0300)]
[media] af9035: remove one config parameter

We can use tuner ID instead of HW not supported flag.
Lesser code is better code.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9035: move device configuration to the state
Antti Palosaari [Mon, 7 May 2012 17:50:40 +0000 (14:50 -0300)]
[media] af9035: move device configuration to the state

Fixes most problems when having more than one device connected
as demod and tuner configurations are not shared.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9033: implement ber and ucb functions
Hans-Frieder Vogt [Sat, 7 Apr 2012 13:34:34 +0000 (10:34 -0300)]
[media] af9033: implement ber and ucb functions

af9033: implement read_ber and read_ucblocks functions. Version 2 of patch that
reflects my findings on the behaviour of abort_cnt, err_cnt and bit_cnt:

- bit_cnt is always 0x2710 (10000)
- abort_cnt is between 0 and 0x2710
- err_cnt is between 0 and 640000 (= 0x2710 * 8 * 8)

in the current implementation BER is calculated as the number of bit errors per
processed bits, ignoring those bits that are already discarded and counted in
abort_cnt, i.e. UCBLOCKS.

Signed-off-by: Hans-Frieder Vogt <hfvogt@gmx.net>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9035: add remote control support
Hans-Frieder Vogt [Sat, 21 Apr 2012 21:23:16 +0000 (18:23 -0300)]
[media] af9035: add remote control support

Signed-off-by: Hans-Frieder Vogt <hfvogt@gmx.net>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9035: remove unused struct
Antti Palosaari [Fri, 6 Apr 2012 22:05:11 +0000 (19:05 -0300)]
[media] af9035: remove unused struct

Not used anymore since new firmware downloader. I forget to remove
those earlier when changed firmware downloader.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] af9035: various small changes for af9035_ctrl_msg()
Antti Palosaari [Fri, 6 Apr 2012 19:32:30 +0000 (16:32 -0300)]
[media] af9035: various small changes for af9035_ctrl_msg()

Fix USB buffer len to maximum possible.
Various log writing fixes, remove extra new lines and excessive
type casts. Rename and type change some variables.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] pms: update to the latest V4L2 frameworks
Hans Verkuil [Sun, 6 May 2012 13:41:54 +0000 (10:41 -0300)]
[media] pms: update to the latest V4L2 frameworks

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] s5p-fimc: Fix locking in subdev set_crop op
Sylwester Nawrocki [Sat, 21 Apr 2012 21:46:30 +0000 (18:46 -0300)]
[media] s5p-fimc: Fix locking in subdev set_crop op

When setting TRY crop on the sub-device the mutex was erroneously acquired
rather than released on exit path. This bug is present in kernels starting
from v3.2.

Cc: stable@vger.kernel.org
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca - ov534: Add Hue control
Antonio Ospite [Mon, 14 May 2012 11:07:45 +0000 (08:07 -0300)]
[media] gspca - ov534: Add Hue control

Signed-off-by: Antonio Ospite <ospite@studenti.unina.it>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] Input: move drivers/input/fixp-arith.h to include/linux
Antonio Ospite [Mon, 14 May 2012 11:07:44 +0000 (08:07 -0300)]
[media] Input: move drivers/input/fixp-arith.h to include/linux

Move drivers/input/fixp-arith.h to include/linux so that the functions
defined there can be used by other subsystems, for instance some video
devices ISPs can control the output HUE value by setting registers for
sin(HUE) and cos(HUE).

Signed-off-by: Antonio Ospite <ospite@studenti.unina.it>
Acked-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca - ov534: Add Saturation control
Antonio Ospite [Mon, 14 May 2012 11:07:43 +0000 (08:07 -0300)]
[media] gspca - ov534: Add Saturation control

Also merge the "COLORS" control into it as it was V4L2_CID_SATURATION
anyway.

Signed-off-by: Antonio Ospite <ospite@studenti.unina.it>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_pac7311: Remove vflip control
Hans de Goede [Thu, 10 May 2012 15:13:16 +0000 (12:13 -0300)]
[media] gspca_pac7311: Remove vflip control

Enabling vflip leads to a much better image, with vflip disabled the
image looks washed out as if there is a too high brightness setting.

Since we don't know how to lower the brightness setting when not
vflipping, simply always vflip and tell userspace to flip the image back,
resulting in a much better (less washed out) image.

Since the image is now no longer too bright, also modify the luminance
level the auto-gain algorithm aims for.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_pac7311: Set register page at start of init
Hans de Goede [Thu, 10 May 2012 15:10:56 +0000 (12:10 -0300)]
[media] gspca_pac7311: Set register page at start of init

Our init sequence was not setting the register page to point to bank 1
before setting what should be the control reg. This causes the camera to
sometimes have its LED on after init. First selecting register bank 1,
rather then assuming the current register bank is bank 1, fixes this.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_pac7311: Convert to the control framework
Hans de Goede [Thu, 10 May 2012 13:52:54 +0000 (10:52 -0300)]
[media] gspca_pac7311: Convert to the control framework

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Remove gspca_auto_gain_n_exposure function
Hans de Goede [Thu, 10 May 2012 11:39:28 +0000 (08:39 -0300)]
[media] gspca: Remove gspca_auto_gain_n_exposure function

Now that the pac207 driver has been converted to the control framework, there
are no remaining users.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gscpa_pac207: Switch to coarse_grained_expo auto gain algorithm
Hans de Goede [Thu, 10 May 2012 11:52:37 +0000 (08:52 -0300)]
[media] gscpa_pac207: Switch to coarse_grained_expo auto gain algorithm

The pac207's exposure control is a clock-divider, so it goes with quite
big steps. So lets use an autogain algorithm optimised for that.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_pac207: Convert to the control framework
Hans de Goede [Thu, 10 May 2012 11:34:42 +0000 (08:34 -0300)]
[media] gspca_pac207: Convert to the control framework

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gscpa_pac207: use usb_err for error handling
Hans de Goede [Wed, 9 May 2012 15:13:29 +0000 (12:13 -0300)]
[media] gscpa_pac207: use usb_err for error handling

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gscpa: Move ctrl_handler to gspca_dev
Hans de Goede [Wed, 9 May 2012 14:19:00 +0000 (11:19 -0300)]
[media] gscpa: Move ctrl_handler to gspca_dev

We intend to eventually port all sub-drivers to the control-framework. At
which point it will make more sense to have the ctrl_handler in gspca_dev
then to have it in the subdrivers. Lets move it there now, to avoid a lot
of work to move it later.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_mars: Convert to the control framework
Hans Verkuil [Sun, 6 May 2012 12:28:31 +0000 (09:28 -0300)]
[media] gspca_mars: Convert to the control framework

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gscpa_stv06xx: Make sd_desc const
Hans de Goede [Wed, 9 May 2012 14:22:47 +0000 (11:22 -0300)]
[media] gscpa_stv06xx: Make sd_desc const

Now that stv06xx is using the control framework it is no longer necessary
to have a (non const) copy of sd_desc inside the sd specific data struct.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_stv06xx: Convert to the control framework
Hans Verkuil [Sun, 6 May 2012 12:28:30 +0000 (09:28 -0300)]
[media] gspca_stv06xx: Convert to the control framework

HdG:
1) Let the gspca-core cleanup the controls on control-init error, like
with the other converted sub drivers
2) Note this also fixes a bug in the hdcs1020 support which was wrongly
reporting an exposure range of 0-65535, even though the effective range was
only 0-255

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_sn9c20x: Whitespace fixes
Hans de Goede [Sat, 12 May 2012 08:43:49 +0000 (05:43 -0300)]
[media] gspca_sn9c20x: Whitespace fixes

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gcpca_sn9c20x: Convert to the control framework
Hans Verkuil [Sun, 6 May 2012 12:28:29 +0000 (09:28 -0300)]
[media] gcpca_sn9c20x: Convert to the control framework

HdG: Small fix: don't register some controls for sensors which don't
have an implementation for them.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_zc3xx: Convert to the control framework
Hans Verkuil [Sun, 6 May 2012 12:28:28 +0000 (09:28 -0300)]
[media] gspca_zc3xx: Convert to the control framework

The initial version was done by HV, corrections were made by HdG, and some
final small changes again by HV.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_zc3xx: Disable the highest quality setting as it is not usable
Hans de Goede [Sun, 6 May 2012 12:28:26 +0000 (09:28 -0300)]
[media] gspca_zc3xx: Disable the highest quality setting as it is not usable

Even with BRC the highest quality setting is not usable, BRC strips so
much data from each MCU that the quality becomes worse then using a lower
quality setting to begin with.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_zc3xx: Always automatically adjust BRC as needed
Hans de Goede [Sun, 6 May 2012 12:28:25 +0000 (09:28 -0300)]
[media] gspca_zc3xx: Always automatically adjust BRC as needed

Always automatically adjust the Bit Rate Control setting as needed, independent
of the sensor type. BRC is needed to not run out of bandwidth with higher
quality settings independent of the sensor.

Also only automatically adjust BRC, and don't adjust the JPEG quality control
automatically, as that is not needed and leads to ugly flashes when it is
changed. Note that before this patch-set the quality was never changed
either due to the bugs in the quality handling fixed in previous patches in
this set.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_zc3xx: Fix JPEG quality setting code
Hans de Goede [Sun, 6 May 2012 12:28:24 +0000 (09:28 -0300)]
[media] gspca_zc3xx: Fix JPEG quality setting code

The current code is using bits 0-1 of register 8 of the zc3xx controller
to set the JPEG quality, but the correct bits are bits 1-2. Bit 0 selects
between truncation or rounding in the quantization phase of the compression,
since rounding generally gives better results it should thus always be 1.

This patch also corrects the quality percentages which belong to the 4
different settings.

Last this patch removes the different reg 8 defaults depending on the sensor
type. Some of them where going for a default quality setting of 50%, which
generally is not necessary in any way and results in poor image quality.
75% is a good default to use for all scenarios.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_zc3xx: Fix setting of jpeg quality while streaming
Hans de Goede [Sun, 6 May 2012 12:28:23 +0000 (09:28 -0300)]
[media] gspca_zc3xx: Fix setting of jpeg quality while streaming

When the user changes the JPEG quality while the camera is streaming, the
driver should not only change the JPEG headers send to userspace, but also
actually tell the camera to use a different quantization table.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca_gl860: Add a present check to sd_stop0
Hans de Goede [Mon, 7 May 2012 09:53:27 +0000 (06:53 -0300)]
[media] gspca_gl860: Add a present check to sd_stop0

The sensor specific dev_post_unset_alt functions all try to write to the
bridge, and none free any memory, so they should be skipped if stop0
is called on disconnection.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Add autogain functions for use with control framework drivers
Hans de Goede [Wed, 9 May 2012 12:58:33 +0000 (09:58 -0300)]
[media] gspca: Add autogain functions for use with control framework drivers

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Set gspca_dev->usb_err to 0 at the begin of gspca_stream_off
Hans de Goede [Mon, 7 May 2012 09:44:21 +0000 (06:44 -0300)]
[media] gspca: Set gspca_dev->usb_err to 0 at the begin of gspca_stream_off

Just a small cleanup.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Call sd_stop0 on disconnect
Hans de Goede [Mon, 7 May 2012 09:25:30 +0000 (06:25 -0300)]
[media] gspca: Call sd_stop0 on disconnect

This is necessary to ensure that worker-threads accessing the device
are stopped before our disconnect handler returns.

This causes a problem with stream_off calling sd_stop0 a second time
when the device handle is closed. This is fixed by setting
gscpa_dev->streaming to 0 on disconnect.

Note that now stream_off will never be called on a disconnected device,
and the present check can thus be removed from stream_off.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Use req_events in poll
Hans de Goede [Mon, 7 May 2012 11:19:42 +0000 (08:19 -0300)]
[media] gspca: Use req_events in poll

So that we don't start a read stream when an app is only polling for control
events.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gscpa: Clear usb_err before calling sd methods from suspend/resume
Hans de Goede [Wed, 9 May 2012 15:11:12 +0000 (12:11 -0300)]
[media] gscpa: Clear usb_err before calling sd methods from suspend/resume

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Switch to V4L2 core locking, except for the buffer queuing ioctls
Hans Verkuil [Sun, 6 May 2012 12:28:27 +0000 (09:28 -0300)]
[media] gspca: Switch to V4L2 core locking, except for the buffer queuing ioctls

Due to latency concerns the VIDIOC_QBUF, DQBUF and QUERYBUF do not use the
core lock, instead they rely only on queue_lock.

Changes by HdG:
1) Change release from the video_device to the v4l2_device, to avoid a
race on disconnect.
2) Adjust for the V4L2 core changes which cause non ioctl fops to no longer
take the V4L2 core lock.

[mchehab@redhat.com: fix a merge conflict]
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Fix locking issues related to suspend/resume
Hans Verkuil [Sun, 6 May 2012 12:28:22 +0000 (09:28 -0300)]
[media] gspca: Fix locking issues related to suspend/resume

There are two bugs here: first the calls to stop0 (in gspca_suspend) and
gspca_init_transfer (in gspca_resume) need to be called with the usb_lock held.
That's true for the other places they are called and it is what subdrivers
expect. Quite a few will unlock the usb_lock in stop0 while waiting for a
worker thread to finish, and if usb_lock isn't held then that can cause a
kernel oops.

The other problem is that a worker thread needs to detect that it has to
halt due to a suspend. Otherwise it will just go on looping. So add tests
against gspca_dev->frozen in the worker threads that need it.

Hdg, 2 minor changes:
1) The finepix device is ok with stopping reading a frame halfway through,
   so add frozen checks in all places where we also check if we're still
   streaming
2) Use gspca_dev->dev instead of gspca_dev->present to check for disconnect
   in all touched drivers. I plan to do this everywhere in the future, and
   most relevant lines in the touched drivers are already modified by this
   patch.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Fix querycap and incorrect return codes
Hans Verkuil [Sun, 6 May 2012 12:28:21 +0000 (09:28 -0300)]
[media] gspca: Fix querycap and incorrect return codes

Add V4L2_CAP_DEVICE_CAPS support to querycap and replace -EINVAL by
-ENOTTY whenever an ioctl is not supported.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Add support for control events
Hans Verkuil [Sun, 6 May 2012 12:28:20 +0000 (09:28 -0300)]
[media] gspca: Add support for control events

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gscpa: Use v4l2_fh and add G/S_PRIORITY support
Hans Verkuil [Sun, 6 May 2012 12:28:19 +0000 (09:28 -0300)]
[media] gscpa: Use v4l2_fh and add G/S_PRIORITY support

In order to support control event gspca has to use struct v4l2_fh.
As a bonus feature this also gives priority handling for free.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Use video_drvdata(file) instead of file->private_data
Hans Verkuil [Sun, 6 May 2012 12:28:18 +0000 (09:28 -0300)]
[media] gspca: Use video_drvdata(file) instead of file->private_data

Prepare for control events: free up file->private_data by using
video_drvdata(file) to get to the gspca_dev struct.

[mchehab@redhat.com: fix a compile error: ‘file’ undeclared]
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] gspca: Allow subdrivers to use the control framework
Hans Verkuil [Sun, 6 May 2012 12:28:17 +0000 (09:28 -0300)]
[media] gspca: Allow subdrivers to use the control framework

Make the necessary changes to allow subdrivers to use the control framework.
This does not add control event support, that comes later.

It add a init_control cam_op that is called after init in probe that allows
the subdriver to set up the controls.

HdG: Call v4l2_ctrl_handler_setup from resume instead of
gspca_set_default_mode, as we just want to resend the current ctrl values to
the device.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] pwc: Fix locking
Hans de Goede [Wed, 9 May 2012 07:43:12 +0000 (04:43 -0300)]
[media] pwc: Fix locking

My last locking rework for pwc mistakenly assumed that videbuf2 does its
own locking, but it does not! This patch fixes the missing locking by
moving over the the video_device lock, and introducing a separate lock
for the videobuf2_queue.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] videobuf2: Fix a bug in fileio emulation error handling
Hans de Goede [Tue, 8 May 2012 17:47:39 +0000 (14:47 -0300)]
[media] videobuf2: Fix a bug in fileio emulation error handling

Various error paths in fileio_init where not setting the request-count
to 0 when unrequesting the buffers on error to init the fileio emulation.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l2-framework.txt: document v4l2_dont_use_cmd
Hans Verkuil [Thu, 10 May 2012 08:04:41 +0000 (05:04 -0300)]
[media] v4l2-framework.txt: document v4l2_dont_use_cmd

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l2-dev: add flag to have the core lock all file operations
Hans Verkuil [Thu, 10 May 2012 07:57:22 +0000 (04:57 -0300)]
[media] v4l2-dev: add flag to have the core lock all file operations

This used to be the default if the lock pointer was set, but now that lock is by
default only used for ioctl serialization. Those drivers that already used
core locking have this flag set explicitly, except for some drivers where
it was obvious that there was no need to serialize any file operations other
than ioctl.

The drivers that didn't need this flag were:

drivers/media/radio/dsbr100.c
drivers/media/radio/radio-isa.c
drivers/media/radio/radio-keene.c
drivers/media/radio/radio-miropcm20.c
drivers/media/radio/radio-mr800.c
drivers/media/radio/radio-tea5764.c
drivers/media/radio/radio-timb.c
drivers/media/video/vivi.c
sound/i2c/other/tea575x-tuner.c

The other drivers that use core locking and where it was not immediately
obvious that this flag wasn't needed were changed so that the flag is set
together with a comment that that driver needs work to avoid having to
set that flag. This will often involve taking the core lock in the fops
themselves.

Eventually this flag should go and it should not be used in new drivers.

There are a few reasons why we want to avoid core locking of non-ioctl
fops: in the case of mmap this can lead to a deadlock in rare situations
since when mmap is called the mmap_sem is held and it is possible for
other parts of the code to take that lock as well (copy_from_user()/copy_to_user()
perform a down_read(&mm->mmap_sem) when a page fault occurs).

It is very unlikely that that happens since the core lock serializes all
fops, but the kernel warns about it if lock validation is turned on.

For poll it is also undesirable to take the core lock as that can introduce
increased latency. The same is true for read/write.

While it was possible to make flags or something to turn on/off taking the
core lock for each file operation, in practice it is much simpler to just
not take it at all except for ioctl and leave it to the driver to take the
lock. There are only a handful fops compared to the zillion ioctls we have.

I also wanted to make it obvious which drivers still take the lock for all
fops, so that's why I chose to have drivers set it explicitly.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l2-ioctl: handle priority handling based on a table lookup
Hans Verkuil [Thu, 10 May 2012 08:40:50 +0000 (05:40 -0300)]
[media] v4l2-ioctl: handle priority handling based on a table lookup

Rather than checking the priority for each ioctl that needs to, just mark
such ioctls in the table and do it only once.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] tea575x-tuner: mark VIDIOC_S_HW_FREQ_SEEK as an invalid ioctl
Hans Verkuil [Thu, 10 May 2012 06:16:14 +0000 (03:16 -0300)]
[media] tea575x-tuner: mark VIDIOC_S_HW_FREQ_SEEK as an invalid ioctl

The tea575x-tuner framework can support the VIDIOC_S_HW_FREQ_SEEK for only
some of the tea575x-based boards. Mark this ioctl as invalid if the board
doesn't support it.

This fixes an issue with S_HW_FREQ_SEEK in combination with priority handling:
since the priority check is done first it could return -EBUSY, even though
calling the S_HW_FREQ_SEEK ioctl would return -ENOTTY. It should always return
ENOTTY in such a case.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l2-dev/ioctl: determine the valid ioctls upfront
Hans Verkuil [Thu, 10 May 2012 08:36:00 +0000 (05:36 -0300)]
[media] v4l2-dev/ioctl: determine the valid ioctls upfront

Rather than testing whether an ioctl is implemented in the driver or not
every time the ioctl is called, do it upfront when the device is registered.

This also allows a driver to disable certain ioctls based on the capabilities
of the detected board, something you can't do today without creating separate
v4l2_ioctl_ops structs for each new variation.

For the most part it is pretty straightforward, but for control ioctls a flag
is needed since it is possible that you have per-filehandle controls, and that
can't be determined upfront of course.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l2-dev: make it possible to skip locking for selected ioctls
Hans Verkuil [Thu, 10 May 2012 05:51:31 +0000 (02:51 -0300)]
[media] v4l2-dev: make it possible to skip locking for selected ioctls

Using the V4L2 core lock is a very robust method that is usually very good
at doing the right thing. But some drivers, particularly USB drivers, may
want to prevent the core from taking the lock for specific ioctls, particularly
buffer queuing ioctls.

The reason is that certain commands like S_CTRL can take a long time to process
over USB and all the time the core has the lock, preventing VIDIOC_DQBUF from
proceeding, even though a frame may be ready in the queue.

This introduces unwanted latency.

Since the buffer queuing commands often have their own internal lock it is
often not necessary to take the core lock. Drivers can now say that they don't
want the core to take the lock for specific ioctls.

As it is a specific opt-out it makes it clear to the reviewer that those
ioctls will need more care when reviewing.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] smiapp: Add driver
Sakari Ailus [Sat, 3 Mar 2012 20:19:52 +0000 (17:19 -0300)]
[media] smiapp: Add driver

Add driver for SMIA++/SMIA image sensors. The driver exposes the sensor as
three subdevs, pixel array, binner and scaler --- in case the device has a
scaler.

Currently it relies on the board code for external clock handling. There is
no fast way out of this dependency before the ISP drivers (omap3isp) among
others will be able to export that clock through the clock framework
instead.

Signed-off-by: Sakari Ailus <sakari.ailus@maxwell.research.nokia.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] smiapp: Generic SMIA++/SMIA PLL calculator
Sakari Ailus [Wed, 7 Dec 2011 16:45:25 +0000 (13:45 -0300)]
[media] smiapp: Generic SMIA++/SMIA PLL calculator

Calculate PLL configuration based on input data: sensor configuration, board
properties and sensor-specific limits.

[mchehab@redhat.com: Fix a Kconfig conflict affecting APTINA_PLL]
Signed-off-by: Sakari Ailus <sakari.ailus@maxwell.research.nokia.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Move CCDC link validation to ccdc_link_validate()
Sakari Ailus [Sat, 25 Feb 2012 23:42:07 +0000 (20:42 -0300)]
[media] omap3isp: Move CCDC link validation to ccdc_link_validate()

Perform CCDC link validation in ccdc_link_validate() instead of
isp_video_validate_pipeline(). Also perform maximum data rate check in
isp_video_check_external_subdevs().

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Default link validation for ccp2, csi2, preview and resizer
Sakari Ailus [Wed, 11 Jan 2012 16:27:02 +0000 (13:27 -0300)]
[media] omap3isp: Default link validation for ccp2, csi2, preview and resizer

Use default link validation for ccp2, csi2, preview and resizer. On ccp2,
csi2 and ccdc we also collect information on external subdevs as one may be
connected to those entities.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Use external rate instead of vpcfg
Sakari Ailus [Sun, 26 Feb 2012 00:13:41 +0000 (21:13 -0300)]
[media] omap3isp: Use external rate instead of vpcfg

Access pipe->external_rate instead of isp_ccdc.vpcfg.pixelclk. Also remove
means to set the value for isp_ccdc_vpcfg.pixelclk.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Introduce isp_video_check_external_subdevs()
Sakari Ailus [Tue, 17 Jan 2012 15:29:38 +0000 (12:29 -0300)]
[media] omap3isp: Introduce isp_video_check_external_subdevs()

isp_video_check_external_subdevs() will retrieve external subdev's
bits-per-pixel and pixel rate for the use of other ISP subdevs at streamon
time. isp_video_check_external_subdevs() is called after pipeline
validation.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Add information on external subdev to struct isp_pipeline
Sakari Ailus [Mon, 16 Jan 2012 21:58:01 +0000 (18:58 -0300)]
[media] omap3isp: Add information on external subdev to struct isp_pipeline

Add pointer to external subdev, pixel rate of the external subdev and bpp of
the format to struct isp_pipeline.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Refactor collecting information on entities in pipeline
Sakari Ailus [Mon, 5 Mar 2012 23:22:41 +0000 (20:22 -0300)]
[media] omap3isp: Refactor collecting information on entities in pipeline

Collect information on entities in pipeline in isp_video_far_end(), outside
pipeline validation. As this causes the function to have side effects,
rename the function accordingly.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Add lane configuration to platform data
Sakari Ailus [Mon, 10 Oct 2011 17:13:26 +0000 (14:13 -0300)]
[media] omap3isp: Add lane configuration to platform data

Add lane configuration (order of clock and data lane) to platform data on
both CCP2 and CSI-2.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Assume media_entity_pipeline_start may fail
Sakari Ailus [Mon, 10 Oct 2011 20:05:24 +0000 (17:05 -0300)]
[media] omap3isp: Assume media_entity_pipeline_start may fail

Since media_entity_pipeline_start() now does link validation, it may
actually fail. Perform the error handling.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Move setting constaints above media_entity_pipeline_start
Sakari Ailus [Mon, 16 Jan 2012 21:59:02 +0000 (18:59 -0300)]
[media] omap3isp: Move setting constaints above media_entity_pipeline_start

The clock rate for l3_ick will soon be read during pipeline validation which
is now part of media_entity_pipeline_start(). For that reason we set
constraints earlier on.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Move definitions required by board code under include/media
Sakari Ailus [Sun, 11 Dec 2011 12:57:51 +0000 (09:57 -0300)]
[media] omap3isp: Move definitions required by board code under include/media

XCLK definitions are often required by the board code. Move them to public
include file.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] omap3isp: Support additional in-memory compressed bayer formats
Sakari Ailus [Wed, 28 Dec 2011 09:28:41 +0000 (06:28 -0300)]
[media] omap3isp: Support additional in-memory compressed bayer formats

This also prevents accessing NULL pointer in csi2_try_format().

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Allow changing control handler lock
Sakari Ailus [Wed, 25 Jan 2012 00:05:34 +0000 (21:05 -0300)]
[media] v4l: Allow changing control handler lock

Allow choosing the lock used by the control handler. This may be handy
sometimes when a driver providing multiple subdevs does not want to use
several locks to serialise its functions.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Implement v4l2_subdev_link_validate()
Sakari Ailus [Mon, 10 Oct 2011 20:01:25 +0000 (17:01 -0300)]
[media] v4l: Implement v4l2_subdev_link_validate()

v4l2_subdev_link_validate() is the default op for validating a link. In V4L2
subdev context, it is used to call a pad op which performs the proper link
check without much extra work.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Improve sub-device documentation for pad ops
Sakari Ailus [Mon, 23 Jan 2012 06:03:00 +0000 (03:03 -0300)]
[media] v4l: Improve sub-device documentation for pad ops

Document that format related configuration is done through pad ops in case
the driver does use the media framework.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] media: Add link_validate() op to check links to the sink pad
Sakari Ailus [Wed, 11 Jan 2012 09:25:15 +0000 (06:25 -0300)]
[media] media: Add link_validate() op to check links to the sink pad

The purpose of the link_validate() op is to allow an entity driver to ensure
that the properties of the pads at the both ends of the link are suitable
for starting the pipeline. link_validate is called on sink pads on active
links which belong to the active part of the graph.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Add DPCM compressed raw bayer pixel formats
Sakari Ailus [Wed, 28 Dec 2011 09:17:26 +0000 (06:17 -0300)]
[media] v4l: Add DPCM compressed raw bayer pixel formats

Add three other colour orders for 10-bit to 8-bit DPCM compressed raw bayer
pixel formats.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Document raw bayer 4CC codes
Sakari Ailus [Tue, 17 Jan 2012 20:47:22 +0000 (17:47 -0300)]
[media] v4l: Document raw bayer 4CC codes

Document guidelines how 4CC codes should be named. Only raw bayer is
included currently. Other formats should be documented later on.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Image processing control class
Sakari Ailus [Thu, 2 Feb 2012 23:17:54 +0000 (20:17 -0300)]
[media] v4l: Image processing control class

Add control class for image processing controls. The control class deals
with controls processing image, for example digital gain or noise filtering,
which can be present in any part of the pipeline.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l: Image source control class
Sakari Ailus [Tue, 4 Oct 2011 11:20:05 +0000 (08:20 -0300)]
[media] v4l: Image source control class

Add image source control class. This control class is intended to contain
low level controls which deal with control of the image capture process ---
the A/D converter in image sensors, for example.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] v4l2: use __u32 rather than enums in ioctl() structs
Sakari Ailus [Thu, 10 May 2012 05:02:07 +0000 (02:02 -0300)]
[media] v4l2: use __u32 rather than enums in ioctl() structs

V4L2 uses the enum type in IOCTL arguments in IOCTLs that were defined until
the use of enum was considered less than ideal. Recently Rémi Denis-Courmont
brought up the issue by proposing a patch to convert the enums to unsigned:

<URL:http://www.spinics.net/lists/linux-media/msg46167.html>

This sparked a long discussion where another solution to the issue was
proposed: two sets of IOCTL structures, one with __u32 and the other with
enums, and conversion code between the two:

<URL:http://www.spinics.net/lists/linux-media/msg47168.html>

Both approaches implement a complete solution that resolves the problem. The
first one is simple but requires assuming enums and __u32 are the same in
size (so we won't break the ABI) while the second one is more complex and
less clean but does not require making that assumption.

The issue boils down to whether enums are fundamentally different from __u32
or not, and can the former be substituted by the latter. During the
discussion it was concluded that the __u32 has the same size as enums on all
archs Linux is supported: it has not been shown that replacing those enums
in IOCTL arguments would break neither source or binary compatibility. If no
such reason is found, just replacing the enums with __u32s is the way to go.

This is what this patch does. This patch is slightly different from Remi's
first RFC (link above): it uses __u32 instead of unsigned and also changes
the arguments of VIDIOC_G_PRIORITY and VIDIOC_S_PRIORITY.

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>
Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] media: videobuf2-dma-contig: quiet sparse noise about plain integer as NULL...
H Hartley Sweeten [Tue, 24 Apr 2012 22:12:48 +0000 (19:12 -0300)]
[media] media: videobuf2-dma-contig: quiet sparse noise about plain integer as NULL pointer

The function vb2_dma_contig_vaddr returns a void * not an integer.

Quiets the sparse noise:

warning: Using plain integer as NULL pointer

Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Pawel Osciak <pawel@osciak.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Acked-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
12 years ago[media] media: videobuf2-dma-contig: include header for exported symbols
H Hartley Sweeten [Tue, 24 Apr 2012 22:08:12 +0000 (19:08 -0300)]
[media] media: videobuf2-dma-contig: include header for exported symbols

Include the header to pickup the definitions of the exported symbols.

Quiets the following sparse warnings:

warning: symbol 'vb2_dma_contig_memops' was not declared. Should it be static?
warning: symbol 'vb2_dma_contig_init_ctx' was not declared. Should it be static?
warning: symbol 'vb2_dma_contig_cleanup_ctx' was not declared. Should it be static?

Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Pawel Osciak <pawel@osciak.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>