| Age | Commit message (Collapse) | Author |
|---|
|
The format to be applied on the sensor is selected by two criteria: the
desired output size and the bit depth. As the selection depends on the
presence of a RAW stream and the streams configuration is handled in
validate() there is no need to re-compute the format in configure().
Centralize the computation of the sensor format in validate() and remove
it from configure().
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
|
|
The findBestFormat() helper operates on the list of sensor formats,
which is owned by the CameraData class. Move the function to that class
as well to:
1) Avoid passing the list of formats to the function
2) Remove a static helper in favour of a class function
3) Allow subclasses with access to CameraData to call the function
Move to the CameraData class the scoreFormat helper as well.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
|
|
populateSensorFormats() is a static helper that is called from a single
place and performs a simple loop over the sensor camera formats.
Remove it and in-line it in the caller to remove one static helper from
the pipeline_base.cpp file.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The commit trailers are checked as part of processing the commit message
with the newly introduced TrailersChecker.
This relies on the trailers property being correctly exposed by the
Commit object, and is implemented for the base Commit but not processed
for Amendment commits.
Refactor the trailer property handling to a helper function in the base
Commit class and make use of it with a newly added call to obtain the
existing Trailers from the most recent commit when using Amendment.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The Amendment commit class is derived from the StagedChanges class
(which in turn derives from the Commit base class), however there is no
code sharing between Amendment and StagedChanges other than the call to
initalise through the base Commit class.
Refactor the inheritance to make an Amendment derive directly from
Commit.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
There are no possible Trailers for staged changes as the commit message
has not yet been written.
Initialise the empty trailers when the commit object is initialised.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
If a Malformed trailer is identified the checkstyle script triggers a
RuntimeError and stops processing the rest of the commit.
A malformed trailer can be regarded as an issue in the commit and
reported as such using the same method as other faults identified by the
tool.
Convert the RuntimeError into a CommitIssue and continue processing
other trailers.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
ABI Compliance:
abi-compliance-checker reports 99% binary compatibility and 81.4% source
compatibilty between version 0.0.5 and 0.1.
- Binary compatibility: 99%
- Source compatibility: 81.4%
- Total binary compatibility problems: 4, warnings: 2
- Total source compatibility problems: 165, warnings: 7
Now that we have funtional ABI validation, the soname has been updated from
libcamera.so.0.0.5 to libcamera.so.0.1.
Future releases will now maintain the same soname until an ABI breakage is
detected to reduce the repackaging and rebuilding of other components that may
use libcamera. The abi-compatibility report can now be generated with the new
utility './utils/abi-compat.sh'.
Key differences here are due to preventing the installation of internal and
private headers in the packaging process, and most of this API compatibility
should not be visible to applications with the exception of:
- Camera::generateConfiguration ( StreamRoles const& roles )
The StreamRoles type definition has been removed. This is now handled as a
Span<StreamRole> to prevent unnecessary allocations and copies.
- CameraManager::addCamera ( std::shared_ptr<struct Camera> camera, std::vector<unsigned long>const& devnums )
- CameraManager::get ( dev_t devnum )
- CameraManager::removeCamera ( std::shared_ptr<struct Camera> camera )
These functions of the CameraManager were for use exclusively by the V4L2
adaptation layer, and not intended for use by applications. This has now been
reworked to use the new SystemDevices and the redundant functions removed.
The ABI/API breakages have been introduced by:
5ca0c9276f28 ("libcamera: CameraManager: Remove ::get(dev_t)")
26a4b83d1b94 ("libcamera: Remove `StreamRoles` alias")
63966ae587d7 ("libcamera: base: Do not install private headers")
Core:
- libcamera: v4l2_videodevice: Use O_CLOEXEC when exporting DMA buffers
- libcamera: ipa_manager: Allow IPA loading in nested subdirs
- meson: Fix usage of overwritten pipeline variable
- subprojects: Drop leftovers of pybind11
- camera_sensor: ipa: core: Add CFA pattern to IPACameraSensorInfo
- libcamera: controls: Define a default lens position behaviour
- libcamera: internal: request: convert to pragma once
- libcamera: camera_manager: Move private implementation to internal
- libcamera: camera_manager: Move {add,remove}Camera to internal
- libcamera: properties: Provide a Devices camera property
- libcamera: pipeline: Register device numbers with camera
- libcamera: camera_sensor: Adjust properties::Rotation
- libcamera: meson: Allow PH to change libcamera_deps
- libcamera: base: Move thread_annotations to private
- libcamera: base: Do not install private headers
- utils: ABI Compatibility checker
- meson: Use x.y soname versioning
- libcamera: camera: Take span of StreamRole instead of vector
- libcamera: Remove `StreamRoles` alias
- utils: ipu3-capture.sh: Fix the script to work with recent media-ctl versions
- Documentation: Add predefined macros from config.h to Doxyfile
- utils: checkstyle: Add __repr__ method to CommitFile class
- utils: checkstyle: Support running checkers selectively
- utils: checkstyle: Don't include commit ID in commit title
- utils: checkstyle: Add trailers checker
- tests: gstreamer: Test cameras' enumeration from GstDeviceProvider
- meson: Fix space around colon issues
- libcamera: Add option to configure udev support
- libcamera: CameraManager: Remove ::get(dev_t)
- tests: gstreamer: Fix compiler error with gcc 8.4.0
ipa:
- meson: ipa: Add mapping for pipeline handler to mojom interface file
- libcamera: ipa: Remove character restriction on the IPA name
- ipa: meson: Allow nested IPA directory structures
- ipa: raspberrypi: Introduce IpaBase class
- ipa: raspberrypi: agc: Move weights out of AGC
- ipa: rkisp1: agc: drop hard-coded analogue gain range
- libipa: camera_sensor_helper: Add IMX327 helper
- pipeline: rpi: ipa_base: Parse config files after platformRegister()
- libcamera: camera_sensor: Add Sony IMX327 sensor properties
- ipa: rpi: imx296: Small refinements to the IMX296 mono sensor tuning
- ipa: rpi: Handle controls for mono variant sensors
- libcamera: controls: Drop reference to IPA in controls documentation
- libipa: camera_sensor_helper: Restore alphabetical order
- ipa: ipu3: agc: Drop hard-codec analogue gain max
- ipa: rkisp1: agc: Restore minimum analogue gain limit
- pipeline: ipa: rpi: Return lens controls from ipa->configure()
- ipa: rpi: Set lens position to hyperfocal on startup
- ipa: rpi: tunings: Add missing short and long exposure profiles
- ipa: rpi: agc: Use std::string instead of char arrays
- ipa: rpi: agc: Gracefully handle missing agc modes
- ipa: rpi: agc: Do not switch to a default if a mode is unavailable
- ipa: rpi: imx708: Fix mode switch drop frame count
- ipa: rpi: imx296_mono: Disable all colour shading
pipeline:
- libcamera: pipeline: simple: Add support for ST's DCMIPP
- pipeline: meson: Allow nested pipeline handler directory structures
- pipeline: ipa: raspberrypi: Refactor and move the Raspberry Pi code
- pipeline: raspberrypi: rpi_stream: Set invalid buffer to id == 0
- pipeline: ipa: raspberrypi: Restructure the IPA mojom interface
- pipeline: raspberrypi: Make RPi::Stream::name() return const std::string &
- pipeline: raspberrypi: Introduce PipelineHandlerBase class
- pipeline: raspberrypi: Add stream flags to RPi::Stream
- libcamera: rkisp1: Generate config using main path
- libcamera: rkisp1: Crop on ISP before downscaling
- libcamera: rkisp1: Assign sizes to roles
- libcamera: rkisp1: Fix enumeration of RAW formats
- pipeline: rpi: Do not return an error from pipeline config file handling
- libcamera: pipeline: simple: Support TI CSI-RX
- pipeline: rpi: Account for Bayer packing when validating format
apps:
- apps: qcam: Remove redundant check
- v4l2: Move the v4l2 compat layer to libexec/libcamera
- android: mm: generic: use GRALLOC_HARDWARE_MODULE_ID
- android: mm: Stub libhardware for build tests
- py: Fix CameraManager.version property
- py: Move ColorSpace and Transform classes to separate files
- py: Use exceptions instead of returning error codes
- py: unittests.py: Add weakref helpers and use del
- py: Move to mainline pybind11 version
- py: Fix code formatting
- py: unittests.py: Fix type checker warnings
- gstreamer: Add enable_auto_focus option to the GStreamer plugin
- gstreamer: Drop libcamera_private dependency
- apps: Add ipa-verify application
- v4l2: v4l2_camera_proxy: Prevent ioctl sign-extensions
- v4l2: Use SystemDevices properties to identify cameras
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The provider g_autoptr variable introduced by commit adb1bbb748a1
("tests: gstreamer: Test cameras' enumeration from GstDeviceProvider")
is left uninitialized when declared. The cleanup function could thus get
called on an unitialized variable if the scope was exited before the
variable gets initialized. This can't occur here, but gcc 8.4.0 still
complains about it:
/usr/include/glib-2.0/glib/gmacros.h: In member function ‘virtual int GstreamerDeviceProviderTest::run()’:
/usr/include/glib-2.0/glib/gmacros.h:1049:27: error: ‘provider’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
{ if (_ptr) (cleanup) ((ParentName *) _ptr); } \
^
../test/gstreamer/gstreamer_device_provider_test.cpp:37:32: note: ‘provider’ was declared here
g_autoptr(GstDeviceProvider) provider;
Silence the error by initializing the variable to NULL at declaration
time. This is a good practice in any case, as later refactoring could
otherwise introduce a scope exit before initialization.
Fixes: adb1bbb748a1 ("tests: gstreamer: Test cameras' enumeration from GstDeviceProvider")
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The CameraManager::get(dev_t) implementation was provided only for the
V4L2 Adaptation layer. This has now been replaced with the use of the
public SystemDevices property.
Remove the deprecated function entirely, along with the camerasByDevnum_
map which was only used to support this functionality.
This is a clear (and intentional) breakage in both the API and ABI.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The CameraManager->get(dev_t) helper was implemented only to support the
V4L2 Adaptation layer, and has been deprecated now that a new camera
property - SystemDevices has been introduced.
Rework the implementation of getCameraIndex() to use the SystemDevices
property and remove reliance on the now deprecated call.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Handling ioctl's within applications is often wrapped with a helper such
as xioctl. Unfortunately, there are many instances of xioctl which
incorrectly handle the correct size of the ioctl request.
This leads to incorrect sign-extension of ioctl's which have bit-31 set,
and can cause values to be passed into the libcamera's v4l2 adaptation
layer which no longer represent the true IOCTL code.
Match the implementation of the Linux kernel and ensure that only 32
bits of the ioctl request are used by assigning to an unsigned int.
Link: https://github.com/Motion-Project/motion/discussions/1636
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Colour shading should be turned off for the mono variant of the IMX296
sensor. To do this, set all the colour gain coefficients to 1.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: David Plowman <daivd.plowman@raspberrypi.com>
Reviewed-by: Nick Hollinghurst <nick.hollinghurst@raspberrypi.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The imx708 must drop a single frame on startup - but only when in HDR
mode. Non-HDR modes do not need to drop frames. Fix the logic in
hideFramesModeSwitch() which currently unconditionally advertises to
drop one frame.
Unfortunately there is no clear way to tell if the sensor is in the HDR
mode. So for now, look the resolution and framerate to deduce this.
Additionally ensure we override hideFramesStartup() and return the same
number as hideFramesModeSwitch().
Bug: https://github.com/raspberrypi/libcamera-apps/issues/524
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: David Plowman <david.plowman@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
RPiCameraConfiguration::validate() did not account for user supplied
Bayer packing when validating the RAW stream configuration. Fix this
by seeing what packing (if any) has been specified selecting the output
RAW stream format.
Fixes: 6c71ee1f1530 ("pipeline: raspberrypi: Introduce PipelineHandlerBase class")
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: David Plowman <david.plowman@raspberrypi.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
New TI SoCs from J7 and AM62 family have a camera pipeline that receives
data using Cadence's DPHY-RX and CSI-RX bridge. A pixel-grabbing "shim"
IP routes this incoming stream of data to the DMA subsystem, to finally
store the frame data in memory.
The driver for this is not merged in mainline yet, but v7 was posted [1]
on linux-media list a few months ago.
With some minor fixes, the work-in-progress v8 series [2] works with the
simple pipeline handler, so we enable support for this device.
Link: https://lore.kernel.org/all/20230314115516.667-1-vaishnav.a@ti.com/ [1]
Link: https://github.com/jailuthra/linux/commits/6ff226ca13f34 [2]
Signed-off-by: Jai Luthra <j-luthra@ti.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
libcamera uses udev to support hotplug (and unplug) detection. This is
an optional feature, which gets selected automatically if udev is
available. Users may however want to build libcamera without hotplug
detection, even if udev is available on the system. Add a new feature
option to control udev support. The default value is auto, which retains
the existing behaviour.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The meson style, which libcamera follows, recommends a space before
colons in function parameters. Fix the style violations through the
project.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
Test the enumeration of the cameras through GstDeviceProvider against
the libcamera camera manager.
Signed-off-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
|
|
The libcamera git history contains numerous examples of incorrect commit
message trailers due to invalid trailer types (e.g. Change-Id), typos
and other small issues. Those went unnoticed through reviews, which
shows that an automated checker is required.
Add a trailers checker to checkstyle.py to catch invalid or malformed
trailers, with a set of supported trailers that match libcamera's commit
message practices. New trailer keys can easily be added later as new
needs arise.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The commit title and commit ID are two different pieces of information.
Don't include the latter in the former, to simplify code that only needs
the commit title. Constructing a string from the ID and title is easier
than splitting the combined string back into its elements.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
During development of the checkstyle.py script, it can be useful to run
only a subset of the checker. Add the ability to do so with a
'--checkers' command line argument.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Add a custom representation to the CommitFile class in order to
facilitate debugging.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
When packaging libcamera, distributions may break IPA module signatures
if the packaging process strips binaries. This can be fixed by resigning
the modules, but the process is error-prone.
Add a command line ipa-verify utility that tests the signature on an IPA
module to help packagers. The tool takes a single argument, the path to
an IPA module shared object, and expects the signature file (.sign) to
be in the same directory.
In order to access the public key needed for signature verification, add
a static function to the IPAManager class. As the class is internal to
libcamera, this doesn't affect the public API.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Tested-by: Javier Martinez Canillas <javierm@redhat.com>
|
|
libcamera creates a config.h file with predefined macros, and instructs
the compiler to include it implicitly with the -include argument.
Doxygen has no support for implicit inclusion of headers, but has a
PREDEFINED configuration option for its preprocessor that lists
predefined macros. Populate it with the values from the config_h
configuration data object that is used for generate the config.h file,
to ensure that documentation matches the configuration options libcamera
has been built with.
Bump the minimum meson version to 0.57 to use cfg_data.keys().
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Recent media-ctl versions include the framerate in the fmt property output:
- entity 37: ov5693 4-0036 (1 pad, 1 link)
type V4L2 subdev subtype Sensor flags 0
device node name /dev/v4l-subdev6
pad0: Source
[fmt:SBGGR10_1X10/2592x1944@1/30
crop.bounds:(16,6)/2592x1944
crop:(16,6)/2592x1944]
-> "ipu3-csi2 1":0 [ENABLED]
This resulted in $sensor_size getting set to: "2592x1944@1 30", which
causes the script to fail.
Fix this by:
1. Replacing the gsub() to remove the '/' between e.g. SBGGR10_1X10
and 2592x1944 with a sub() so that only that first '/' gets replaced
(resulting in a $sensor_size of "2592x1944@1/30" instead).
2. Adding a new sub() to remove the @1/30 suffix.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Now that `Camera::generateConfiguration()` takes a `libcamera::Span`
of `StreamRole`, remove the `StreamRoles` type, which was an alias
to `std::vector<libcamera::StreamRole>`.
The removal has two reasons:
- it is no longer strictly necessary,
- its presence may suggest that that is the preferred (or correct)
way to build/pass a list of `StreamRole`.
Signed-off-by: Barnabás Pőcze <pobrn@protonmail.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
[Kieran: Fix small checkstyle report on roles initialiser]
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Change the parameter type of `generateConfiguration()` from `const std::vector&`
to `libcamera::Span`. A span is almost always preferable to a const vector ref
because it does not force dynamic allocation when none are needed, and it allows
any contiguous container to be used.
A new overload is added that accepts an initializer list so that
cam->generateConfiguration({ ... })
keeps working.
There is no API break since a span can be constructed from a vector
and the initializer list overload takes care of the initializer lists,
but this change causes an ABI break.
Signed-off-by: Barnabás Pőcze <pobrn@protonmail.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
[Kieran: Apply checkstyle fixups]
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Now that we identify ABI breakages, provide incremental releases which
can support backwards compatible linkage across release points that have
a compatible ABI.
Introduction of this commit does not convey that libcamera now has a
stable API, but that patch releases with a common minor number
(0.minor.patch) may potentially be used by applications without
recompilation and linkage against new releases.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Provide support to compare ABI compatibility between any two git commits
or by a commit and the most recent ancestral tag of that commit.
Tested-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Split the public and private headers from the base library and stop installing
private headers as part of the install process.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The libcamera thread annotations are for internal use only and are
not part of any public API. Mark the header accordingly by including the
private.h header guard.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Drop libcamera_private dependency entirely as to avoid libcamerasrc
getting more dependent on it. In order to achieve that, one of the
mutex locks in GstLibcameraSrcState needs to be replaced with GMutex.
However doing so, this won't let us to use the clang's thread annotation
macros in libcamera (which should be fine as libcamerasrc would move
out of libcamera repo once matured).
Signed-off-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
|
|
Pipeline handlers can have custom dependencies.
Allow each pipeline handler subdir to add them to libcamera_deps[].
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
If a user provided pipeline config file is not present, or if the
version reported in the file is invalid, do not return with an error
when creating the pipeline handler. Instead, log a warning message and
return success with default pipeline config values used.
This now matches the behaviour when the pipeline config file could not
be parsed correctly, and we revert to default values.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: David Plowman <david.plowman@raspberrypi.com>
|
|
In commit 0ee9339331c6, a default metering/exposure/constraint mode is
used if a control sets a mode that is not listed in the camera tuning
file.
Setting a default mode may be undesirable in these cases, so instead
keep the agc mode unchanged. This also matches the behaviour for other
IPA controls where no changes are made in error conditions.
Fixes: 0ee9339331c6 ("ipa: rpi: agc: Gracefully handle missing agc modes")
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: David Plowman <david.plowman@raspberrypi.com>
|
|
As the CameraSensor::validateTransform() function compensate
for the sensor's mounting rotation, the properties::Rotation value
should be adjusted to make sure application that receive already
"corrected" images do not get confused by Rotation still reporting
a value.
Howerver, as an image sensor can only compensate rotations by applying
H/V flips, only correct Rotation when the mounting rotation is 180
degrees.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
Cameras such as the PiCam 3 support auto-focus, but the GStreamer plugin
for libcamera does not enable auto-focus. With this patch auto-focus can
be enabled for cameras that support it. By default it is disabled, which
means default behaviour remains unchanged. For cameras that do not
support auto-focus, an error message shows up if auto-focus is enabled.
This was tested on cameras that do not support auto-focus (e.g. PiCam2)
and was tested on a camera that does support auto-focus (PiCam3). The
test involved setting the focus to AfModeContinous and observing it.
However, by not setting "auto-focus-mode" or using AfModeManual as
the "auto-focus-mode" both resulting in auto-focus being disabled.
Bug: https://bugs.libcamera.org/show_bug.cgi?id=188
Signed-off-by: Cedric Nugteren <cedric@plumerai.com>
Reviewed-by: Maarten Lankhorst <dev@lankhorst.se>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Tested-by: Umang Jain <umang.jain@ideasonboard.com>
Signed-off-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
Register the identified device numbers with each camera as the
SystemDevices property.
This facilitates camera daemons or other systems to identify which
devices are being managed by libcamera, and can prevent duplication of
camera resources.
As the SystemDevices property now provides this list of devices, use it
directly from within the CameraManager when adding a Camera rather than
passing it through the internal API.
Tested-by: Ashok Sidipotu <ashok.sidipotu@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Provide a new Camera property that allows pipeline handlers to list any
kernel device used to operate the camera. This allows other frameworks
and daemons such as PipeWire to better understand the resources consumed
by a Camera and consider ignoring those resources when enumerating
camera devices on a system.
Tested-by: Ashok Sidipotu <ashok.sidipotu@collabora.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The CameraManager exposes addCamera and removeCamera as public API
calls, while they should never be called from an application. These
calls are only expected to be used by PipelineHandlers to update the
CameraManager that a new Camera has been created and allow the Camera
Manager to expose it to applications.
Remove the public calls and update the private implementations such that
they can be used directly by the PipelineHandler through the internal
CameraManager::Private provided by the Extensible class.
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Tested-by: Ashok Sidipotu <ashok.sidipotu@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The CameraManager makes use of the Extensible pattern to provide an
internal private implementation that is not exposed in the public API.
Move the Private declaration to an internal header to make it available
from other internal components in preperation for reducing the surface
area of the public interface of the CameraManager.
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Tested-by: Ashok Sidipotu <ashok.sidipotu@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Remove the verbose #ifndef/#define/#endif pattern for maintaining
header idempotency, and replace it with a simple #pragma once.
This simplifies the headers, and prevents redundant changes when
header files get moved.
The internal/request.h was missed at the time of the original
conversion. Update it.
Fixes: df131ad08893 ("libcamera: internal: Convert to pragma once")
Tested-by: Ashok Sidipotu <ashok.sidipotu@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
If a metering/exposure/constraint mode is not listed in the sensor
tuning file, and a control for the missing mode is set on the agc, we
terminate the application with a fatal log message.
Instead of this fatal termination, log a warning message and switch to
the appropriate default mode so that the application continues running.
Bug: https://github.com/raspberrypi/libcamera/issues/59
Bug: https://github.com/ayufan/camera-streamer/issues/67
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Replace the char array strings in struct AgcStatus with std::string
objects. This simplifies the string handling in the source code.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Some sensor tuning files were missing the short and/or long exposure
mode profiles. Add the missing items to the relevant files.
As a drive-by, rename the "sport" exposure profile to "short" for the
IMX290 tuning.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The current implementation enumerates a single RAW format (the sensor's
resolution) and does that regardless of what role the
CameraConfiguration has been generated for.
Fix this by:
- Enumerate RAW StreamFormats only when the requested role is
StreamRole::Raw
- Add all the sensor's provided resolutions that fit the video device
output maximum size
Before this patch, a single RAW size was enumerated in stream formats
* Pixelformat: SRGGB10 (4208x3120)-(4208x3120)/(+1,+1)
- 4208x3120
With this patch applied all sensor's supported resolutions are
enumerated but only when the stream role RAW is explicitly requested
* Pixelformat: SRGGB10 (1048x780)-(4208x3120)/(+0,+0)
- 1048x780
- 2104x1560
- 4032x3024
- 4208x3120
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
Currently each RkISP1 path (main and self) generate a configuration
by bounding the sensor's resolution to their respective maximum output
aspect ratio and size.
Size maxResolution = maxResolution_.boundedToAspectRatio(resolution)
.boundedTo(resolution);
In the case of self path, whose maximum size is 1920x1920, the generated
configuration could get assigned unusual sizes, as the result of the
above operation.
As an example, with the imx258 sensor whose resolution is 4208x3118, the
resulting size for the Viewfinder use case is an unusual 1920x1423.
Fix this by assigning to each role a desired output size:
- Use the sensor's resolution for StillCapture and Raw
- Use 1080p for Viewfinder and VideoRecording
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Crop on the resizer sink pad before downscaling to the aspect ratio
of the desired output size.
Cropping the input frame to the output aspect ratio allows to maintain
the correct picture proportions, as otherwise downscaling would change
the image geometry.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
The generateConfiguration() implementation in the Rockchip RkISP1
pipeline handler uses by default the self path (if available) for
the Viewfinder and VideoRecording StreamRoles.
The validate() implementation, at the contrary, prefers using the main
path, when available, for all streams.
As the self-path is limited in output resolution to 1920x1920,
generating a configuration using the self path limits the maximum
stream size to 1920x1920, while higher resolutions can be obtained by
using the main path.
Align the generateConfiguration() implementation to the validate() one
by using the main path by default if available.
Bug: https://bugs.libcamera.org/show_bug.cgi?id=180
Reported-by: libcamera@luigi311.com
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
