- 30 Jul, 2012 40 commits
-
-
Sachin Kamat authored
Replace custom err() macro with v4l2_err() macro. [s.nawrocki: added missing end-of-line at the log print] Signed-off-by:
Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by:
Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@redhat.com>
-
Sylwester Nawrocki authored
This patch adds locking for open(), close(), poll() and mmap() file operations in the driver as a follow up to the changes done in commit 5126f259 "v4l2-dev: add flag to have the core lock all file operations". Signed-off-by:
Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by:
-
Hans Verkuil authored
G_PARM should only be enabled if: - vidioc_g_parm is present - or: it is a video node and vidioc_g_std or tvnorms are set. Without this additional check v4l2-compliance would complain about being able to use g_parm when it didn't expect it. Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
v4l2_m2m_poll didn't support events, but that's essential if you want to be able to use control events for example. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Laurent Pinchart authored
The g_mbus_fmt operation only needs to return the current mbus frame format and doesn't need to configure the hardware to do so. Fix it to avoid requiring the chip to be powered on when calling the operation. Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by:
-
Laurent Pinchart authored
The g_mbus_fmt operation only needs to return the current mbus frame format and doesn't need to configure the hardware to do so. Fix it to avoid requiring the chip to be powered on when calling the operation. Signed-off-by:
-
Laurent Pinchart authored
The g_mbus_fmt operation only needs to return the current mbus frame format and doesn't need to configure the hardware to do so. Fix it to avoid requiring the chip to be powered on when calling the operation. Signed-off-by:
-
Laurent Pinchart authored
There will be no soc_camera_device instance with a soc-camera device is used with a non soc-camera host, so we won't be able to pass the soc_camera_device fake platform device to board code. Pass the physical device instead. The argument is currently not used by any board file so this is safe. Signed-off-by:
-
Laurent Pinchart authored
The soc-camera module exports functions that are needed by soc-camera client drivers even when not running in soc-camera mode. Replace the platform_driver_probe() with a platform_driver_register() call to avoid module load failures if no soc-camera device is present. Signed-off-by:
-
Hans Verkuil authored
Zero all fields after the first type field. Signed-off-by:
-
Hans Verkuil authored
The VIDIOC_DV_TIMINGS_CAP ioctl check wasn't added to determine_valid_ioctls(). This caused this ioctl to always return -ENOTTY. Signed-off-by:
-
Manjunath Hadli authored
The auto selection of pertinent helper chips (VIDEO_HELPER_CHIPS_AUTO) should select the ADV7343 and THS7303 driver, which is used by VPIF display driver. Signed-off-by:
Manjunath Hadli <manjunath.hadli@ti.com> Signed-off-by:
Lad, Prabhakar <prabhakar.lad@ti.com> Signed-off-by:
-
Manjunath Hadli authored
add generic build configuration for vpif capture and display drivers as it is common for DM6467/DA850/OMAP-L138. Signed-off-by:
-
Manjunath Hadli authored
add clock enable and disable in probe and remove functions. Probe will succeed only if the device clock is provided instead of assuming that the clock is always enabled. VPIF clock has to be dealt with during suspend and resume. Implement power management callbacks to VPIF driver to disable/enable clock on suspend/resume respectively. Signed-off-by:
-
Manjunath Hadli authored
Implement power management operations - suspend and resume as part of dev_pm_ops for VPIF capture driver. Signed-off-by:
-
Manjunath Hadli authored
Implement power management operations - suspend and resume as part of dev_pm_ops for VPIF display driver. Signed-off-by:
-
Manjunath Hadli authored
add hardware clipping support for VPIF output data. This is needed as it is possible that the external encoder might get confused between the FF or 00 which are a part of the data and that of the SAV or EAV codes. Signed-off-by:
-
Lad, Prabhakar authored
This patch migrates VPIF display driver to videobuf2 framework. Signed-off-by:
-
Lad, Prabhakar authored
This patch migrates VPIF capture driver to videobuf2 framework. Signed-off-by:
-
Manjunath Hadli authored
Size up the memory for the buffers from the buffer pool allocated in board file. Then adjust the reqbuf count depending the available memory. Signed-off-by:
-
Manjunath Hadli authored
Size up the memory for the buffers from the buffer pool allocated in board file. Then adjust the reqbuf count depending the available memory. Signed-off-by:
-
Manjunath Hadli authored
fix setting of data width in config_vpif_params() function, which was wrongly set. Signed-off-by:
-
Manjunath Hadli authored
omap-l138 shares the interrupt between capture and display. Make sure we are able to request for the same irq number by making a shared irq request. Signed-off-by:
-
Manjunath Hadli authored
change the dm646x specific strings in the driver to make them generic across platforms. In this case change all the strings which have a dm646x connotation to vpif which is a platform independent ip. Signed-off-by:
-
Manjunath Hadli authored
As the same interrupt is shared between capture and display devices, sometimes we get isr calls where the interrupt might not genuinely belong to capture or display. Hence, add a condition in the isr to check for interrupt ownership and channel number to make sure we do not service wrong interrupts. Signed-off-by:
-
Lad, Prabhakar authored
from commit a8f3c203 restore the mapping scheme for uncached buffers, which was changed in a common scheme for cached and uncached. This apparently was wrong, and was probably intended only for cached buffers. the fix fixes the crash observed while mapping uncached buffers. Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
-
Hans de Goede authored
This driver consists of 2 parts, a generic tea5777 driver and a driver for the Griffin radioSHARK v2 USB radio receiver, which is the only driver using the generic tea5777 for now. This first version only implements FM support, once the the new VIDIOC_ENUM_FREQ_BANDS API is upstream I'll also add AM support. Signed-off-by:
Hans de Goede <hdegoede@redhat.com> Signed-off-by:
-
Hans Verkuil authored
Commit efbceecd, adds a number of helper functions for ctrl related ioctls to v4l2-ioctl.c, these helpers assume that if file->private_data != NULL, it points to a v4l2_fh, which is only the case for drivers which actually use v4l2_fh. This breaks for example bttv which use the "filedata" pointer for its own uses, and now all the ctrl ioctls try to use whatever its filedata points to as v4l2_fh and think it has a ctrl_handler, leading to: [ 142.499214] BUG: unable to handle kernel NULL pointer dereference at 0000000000000021 [ 142.499270] IP: [<ffffffffa01cb959>] v4l2_queryctrl+0x29/0x230 [videodev] [ 142.514649] [<ffffffffa01c7a77>] v4l_queryctrl+0x47/0x90 [videodev] [ 142.517417] [<ffffffffa01c58b1>] __video_do_ioctl+0x2c1/0x420 [videodev] [ 142.520116] [<ffffffffa01c7ee6>] video_usercopy+0x1a6/0x470 [videodev] ... This patch adds the missing test_bit(V4L2_FL_USES_V4L2_FH, &vfd->flags) tests to the ctrl ioctl helpers v4l2_fh paths, fixing the issues with for example the bttv driver. Signed-off-by:
Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by:
-
Hans de Goede authored
And document V4L2_CID_AUTOBRIGHTNESS. Signed-off-by:
-
Hans de Goede authored
When the v4l2_device release handler gets called the kobject under vdev->dev has already been released, so we cannot use kobject_name on it (which video_device_node_name does). Signed-off-by:
-
Hans de Goede authored
As it cannot be changed while streaming. Signed-off-by:
-
Hans de Goede authored
These are broken with my test cam and I've been unable to fix them. Signed-off-by:
-
Hans de Goede authored
Signed-off-by:
-
Hans de Goede authored
Signed-off-by:
-
Hans de Goede authored
Signed-off-by:
-
Hans de Goede authored
The konica needs a freaking large time (circa 6.5 seconds) to "boot", and does not want to be bothered while doing so, so sleep for 6 seconds, and then query its status register at 100ms intervals until it becomes ready. This removes the "reg_w err: -32" messages shown in dmesg whenever a konica cam gets initialized, and also fixes the camera not working when an app tries to use it directly after it has been plugged in and after a suspend/resume cycle. Signed-off-by:
-
