| Age | Commit message (Collapse) | Author |
|---|
|
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>
|
|
On the first ipa->configure() call, set the lens position (if a lens is
present) to the default position. Typically this would be the hyperfocal
position based on the tuning data.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
In addition to sensor controls, return lens controls from
IpaBase::configure() back to the pipeline handler. If there are lens
controls present in the ControlList, action them in the pipeline handler
as part of the configure routine.
Signed-off-by: Naushir Patuck <naush@raspberrypi.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>
|
|
Update the AfMode control description to explicitly define a
startup/default behaviour.
On startup, the camera will move the lens to the position given by the
default value of the LensPosition control if operating in manual focus
mode. Typically this would be the hyperfocal position of the lens.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Commit a3178dd0391f ("ipa: rkisp1: agc: drop hard-coded analogue gain range")
removed both minimum and maximum limits for the analogue gain value.
However, as some sensors can potentially have a minimum gain lower than
1.0, restore the check for the minimum limit.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
As the sensor's analogue gain range is known, drop the arbitrary
maximum limit for the sensor analogue gain.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Tested-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The addition of the CameraSensorHelperImx327 class was not correctly
inserted in alphabetical sort order.
Move it to the correct location.
Fixes: 7d5b38e2ef41 ("libipa: camera_sensor_helper: Add IMX327 helper")
Reported-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The controls relates to the public API, which doesn't expose IPA modules
as they are internal to libcamera. Replace the mention of "IPA" in the
controls documentation.
While at it, fix a small text reflow issue.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Do not advertise colour related controls (i.e. [A]WB, colour saturation)
in the ControlInfoMap of available controls returned out to the
application.
Silently ignore these controls in the control handler in case applications
don't use the advertised ControlInfoMap to validate controls.
As a drive-by fix, don't advertise controls::ColourCorrectionMatrix in
the ControlInfoMap as it is not handled by the IPA.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: David Plowman <david.plowman@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Add a new cfaPattern field to the IPACameraSensorInfo to pass the
CFA/Bayer pattern for the current sensor configuration to the IPA.
This field takes a value from properties::draft::ColorFilterArrangementEnum.
Populate cfaPattern in CameraSensor::sensorInfo(), called by the
pipeline handler before it calls ipa->init() and ipa->config().
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Fix type checker warnings by dropping unused imports and using _ for
variable names that are not used.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Fix code formatting.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The pybind11 subproject was removed in commit ad4719c10c6a ("py: Move to
mainline pybind11 version"), but two traces remained, one in .gitignore
and a second one as a meson.build for pybind11. Remove them.
Fixes: ad4719c10c6a ("py: Move to mainline pybind11 version")
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Update the noise profile and reference lux level with newly taken
calibration pictures. Remove GEQ/cross-talk processing block as this is
a mono sensor.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
We are using pybind11 'smart_holder' branch to solve the Camera
destructor issue (see the comment in this patch, or the commit
that originally added Python bindings support).
As it would be very nice to use the mainline pybind11 (which is packaged
in distributions), this patch adds a workaround allowing us to move to
the mainline pybind11 version.
The workaround is simply creating a custom holder class
(PyCameraSmartPtr), used only for the Camera, which wraps around the
shared_ptr. This makes the compiler happy.
Moving to mainline pybind11 is achieved with:
- Change the pybind11 wrap to point to the mainline pybdind11 version
- Tell pybind11 to always use shared_ptr<> as the holder for
PyCameraManager, as we use the singleton pattern for the
PyCameraManager, and using shared_ptr<> to manage it is a requirement
- Tell pybind11 to always use PyCameraSmartPtr<> as the holder for Camera
- Change the meson.build file to use a system-installed pybind11
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Add helpers to check if a weakref or a list of weakrefs is alive or
dead.
Also use 'del' for local variables instead of setting the variable to
None. This makes debugging the test easier as the locals will be gone
from locals() dict.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
We have multiple methods which return an error code, mimicking the C++
API. Using exceptions is more natural in the Python API, so change all
those methods to raise an Exception instead.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Move ColorSpace and Transform classes to separate files from the main
py_main.cpp, for clarity.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The current CameraManager.version doesn't work at all (raises a
TypeError), as that's not how you use expose C++ static methods as
Python class methods.
Fix it.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Commit 66c618f378aa ("android: mm: generic: use
GRALLOC_HARDWARE_MODULE_ID") made libhardware a required dependency for
the Android camera HAL on non-Chrome OS platforms. This isn't an issue
for real devices, as Android provides libhardware, but it prevents
compile-testing the camera HAL on traditional Linux systems.
To restore the compile-test coverage, stub the libhardware function used
by the camera HAL when libhardware isn't found.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Tested-by: Mattijs Korpershoek <mkorpershoek@baylibre.com> # meson build test
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
PlatformFrameBufferAllocator is an abstraction over gralloc.
Right now hardwareModule_ points towards a CAMERA_HARDWARE_MODULE_ID.
When gralloc_open() is called we observe:
libcamera: DEBUG HAL camera3_hal.cpp:75 Open camera gpu0
libcamera: ERROR Camera camera.cpp:524 Camera in Configured state trying acquire() requiring state Available
01-23 14:14:04.742 370 416 E libcamera: FATAL HAL generic_frame_buffer_allocator.cpp:105 gralloc_open() failed: -87
Which is wrong, gralloc_open() is attempting to re-open the camera HAL,
instead of the gralloc HAL.
Point to a GRALLOC_HARDWARE_MODULE_ID instead so that we can request
buffers from gralloc in android.
Note: this adds new dependencies on android's libhardware [1] and on libdl.
[1] https://android.googlesource.com/platform/hardware/libhardware
Fixes: c58662c5770e ("android: Introduce PlatformFrameBufferAllocator")
Signed-off-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Copy the sensor properties from Sony IMX290, as the IMX327 is a variant
of it.
Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
[Kieran: Adjust sort order]
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
If a user uses a pipeline handler config file where a custom timeout
value is specified, it would lead to a segmentation fault when
loadPipelineConfiguration() tried to access the as yet uninitialised
V4L2VideoDevice object.
To fix this, parse the pipeline handler config file only after fully
initialising and registering the pipeline handler and V4L2VideoDevice
objects.
Fixes: 6c71ee1f15305 ("pipeline: raspberrypi: Introduce PipelineHandlerBase class")
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: David Plowman <david.plowman@raspberrypi.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Add support for the Sony IMX327, which is added to the kernel with
commit 2d41947ec2c0 ("media: i2c: imx290: Add support for imx327
variant"). It is basically a derivate of the IMX290, therefore also
derive the helper.
Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
As the sensor's analogue gain range is known (read-out in
IPARkISP1::configure()), drop the limiting hard-coded range.
This enables better performance in low-light conditions for sensors with
a higher gain (e.g. the imx327).
Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
When iterating over enabled pipelines and IPA modules, libcamera
descends into subdirectories in a recursive manner, which involves
nested loops. Both the outer and inner loops use the same loop variable
named 'pipeline'. As the outer loop uses the variable after descending
into the inner directory, it ends up using an incorrect value. Fix it by
moving all use of the variable before the subdir() call, and add a
comment that warns about the issue to avoid reintroducing it.
Fixes: e8526c0c2bc6 ("ipa: meson: Allow nested IPA directory structures")
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
Now that we allow IPA modules to live in nested directories for the
RaspberryPi platform, it is required to allow parsing one more level
to be able to run libcamera from the source directory.
Without this patch the $(builddir)/src/ipa/rpi/vc4/ipa_rpi_vc4.so
IPA module cannot be loaded.
The issue is only present when running from the source directory
as when libcamera is installed all IPA modules are deployed to a single
$(prefix)/$(libcamera_libdir)/libamera/ location.
Fixes: 46aefed208fe ("pipeline: meson: Allow nested pipeline handler directory structures")
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
|
|
Files opened internally in libcamera without the O_CLOEXEC file will
remain open upon a call to one of the exec(3) functions. As exec()
doesn't destroy local or global objects, this can lead to various side
effects. Avoid this by opening file descriptors with O_CLOEXEC for all
internal files.
The O_CLOEXEC flag should also be set when obtaining file handles from
the V4L2 VIDIOC_EXPBUF operation, but was missed during the updates to
both d942bdc913c5 ("libcamera: v4l2_device: openat(2) with O_CLOEXEC to
cleanup after exec(3)") and 436b38fd89fe ("libcamera: Open files with
O_CLOEXEC").
Set the O_CLOEXEC flag on calls to ioctl(VIDIOC_EXPBUF).
Fixes: 436b38fd89fe ("libcamera: Open files with O_CLOEXEC")
Signed-off-by: Elias Naur <mail@eliasnaur.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Move the v4l2-compat.so shared library installation to the libcamera
directory under libexec. This is the same location that the proxy
workers live and will facilitate easier packaging of the V4L2
compatibility layer with distributions.
Create a new libcamera_libexecdir variable within meson to simplify
representation of this path and update the proxy worker meson.build
infrastructure to make use of it as well.
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Add a flags_ field to indicate stream state information in RPi::Stream.
This replaces the existing external_ and importOnly_ boolean flags.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Create a new PipelineHandlerBase class that handles general purpose
housekeeping duties for the Raspberry Pi pipeline handler. Code the
implementation of new class is essentially pulled from the existing
pipeline/rpi/vc4/raspberrypi.cpp file with a small amount of
refactoring.
Create a derived PipelineHandlerVc4 class from PipelineHandlerBase that
handles the VC4 pipeline specific tasks of the pipeline handler. Again,
code for this class implementation is taken from the existing
pipeline/rpi/vc4/raspberrypi.cpp with a small amount of
refactoring.
The goal of this change is to allow third parties to implement their own
pipeline handlers running on the Raspberry Pi without duplicating all of
the pipeline handler housekeeping tasks.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Return a const std::string reference from RPi::Stream::name() to avoid
copying a string when not needed.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The region weights for the the AGC zones are handled by the AGC
algorithm. Apply them directly in the IPA (vc4.cpp) to the statistics
that we pass to the AGC.
Signed-off-by: David Plowman <david.plowman@raspberrypi.com>
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Create a new IpaBase class that handles general purpose housekeeping
duties for the Raspberry Pi IPA. The implementation of the new class
is essentially pulled from the existing ipa/rpi/vc4/raspberrypi.cpp file
with a minimal amount of refactoring.
Create a derived IpaVc4 class from IpaBase that handles the VC4 pipeline
specific tasks of the IPA. Again, code for this class implementation is
taken from the existing ipa/rpi/vc4/raspberrypi.cpp with a
minimal amount of refactoring.
The goal of this change is to allow third parties to implement their own
IPA running on the Raspberry Pi without duplicating all of the IPA
housekeeping tasks.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Restructure the IPA mojom interface to be more consistent in the use
of the API. Function parameters are now grouped into *Params structures
and results are now returned in *Results structures.
The following pipeline -> IPA interfaces have been removed:
signalQueueRequest(libcamera.ControlList controls);
signalIspPrepare(ISPConfig data);
signalStatReady(uint32 bufferId, uint32 ipaContext);
and replaced with:
prepareIsp(PrepareParams params);
processStats(ProcessParams params);
signalQueueRequest() is now encompassed within prepareIsp().
The following IPA -> pipeline interfaces have been removed:
runIsp(uint32 bufferId);
embeddedComplete(uint32 bufferId);
statsMetadataComplete(uint32 bufferId, libcamera.ControlList controls);
and replaced with the following async calls:
prepareIspComplete(BufferIds buffers);
processStatsComplete(BufferIds buffers);
metadataReady(libcamera.ControlList metadata);
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
At present, the RPiStream buffer ids == -1 indicates an invalid value.
As a simplification, use id == 0 to indicate an invalid value. This
allows for better code readability.
As a consequence of this, use unsigned int for the buffer id values.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Split the Raspberry Pi pipeline handler and IPA source code into common
and VC4/BCM2835 specific file structures.
For the pipeline handler, the common code files now live in
src/libcamera/pipeline/rpi/common/
and the VC4-specific files in src/libcamera/pipeline/rpi/vc4/.
For the IPA, the common code files now live in
src/ipa/rpi/{cam_helper,controller}/
and the vc4 specific files in src/ipa/rpi/vc4/. With this change, the
camera tuning files are now installed under share/libcamera/ipa/rpi/vc4/.
To build the pipeline and IPA, the meson configuration options have now
changed from "raspberrypi" to "rpi/vc4":
meson setup build -Dipas=rpi/vc4 -Dpipelines=rpi/vc4
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The current pipeline handler build files require a flat directory
struture for each pipeline handler. Modify the build files to remove
this restriction and allow a directory structure such as:
src/libcamera/pipeline/
|- raspberrypi
|- common
|- vc4
|- rkisp1
|- ipu3
where each subdir (e.g. raspberrypi/common, raspberrypi/vc4) has its own
meson.build file. Such a directory structure will be introduced for the
Raspberry Pi pipeline handler in a future commit.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The current IPA build files require a flat directory structure for the
IPAs. Modify the build files to remove this restriction and allow a
directory structure such as:
src/ipa
|- raspberrypi
|- common
|- cam_helpers
|- controller
|- vc4
|- rkisp1
|- ipu3
where each subdir (e.g. raspberrypi/common, raspberrypi/cam_helper) has
its own meson.build file. Such a directory structure will be introduced
for the Raspberry Pi IPA in a future commit.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Remove the restriction on using the '/' character in the IPA name
string. This allows more flexibility in IPA directory structures where
different IPA modules might live in subdirectories under the usual
src/ipa/<platform>/ top level directory.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Allow an arbitrary mapping between the pipeline handler and IPA mojom
interface file in the build system. This removes the 1:1 mapping of
pipeline handler name to mojom filename, and allows more flexibility to
pipeline developers.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
The switch statement checks `roles.size()` with cases for 1 and 2,
so in the `default` branch, `role.size() > 2`, i.e. it is always
different from 1, so the check is unnecessary.
Signed-off-by: Barnabás Pőcze <pobrn@protonmail.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
|
|
The STM32 contains a camera pipeline known as the DCMIPP (Digital
Camera-Memory Interface Pixel Processor) which receives data from a
parallel interface and dumps the post-processed data to memory. The
pipeline is capable of some processing in the form of downscaling
captured data through cropping or skipping the sensor's output.
The simple pipeline handler is quite capable of handling the DCMIPP
given its operation is handled entirely through configuring the pads
of a media graph, so add support for the driver to the pipeline's
supportedDevices array.
Signed-off-by: Daniel Scally <dan.scally@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Bugfixes:
- meson: Ignore -Wredundant-move with gcc-13 and newer
- cam: file_sink: Workaround gcc-13 dangling-reference false positive
- libcamera: Open files with O_CLOEXEC
- meson: Fix git version parsing
- libcamera: camera_manager: Stop exponential explosive calls to createPipelineHandlers
- py: cam: Fix demosaic overflow issue
- libcamera: base: Make the registration of log categories atomic
Core:
- test: controls: control_info_map: Test default constructor
- controls: guard ControlInfoMap against nullptr idmap_
- device_enumerator_udev: Use std::string_view
- formats: Add 14-bits Bayer RAW formats
- device_enumerator_udev: Remove devnum from dependency map
- meson: libcamera: Report IPA signature method
- meson: Align handling of build metadata
- converter: Check converter validity
- Keep using syslog for isolated IPA modules
- utils: ipc: Update parser.py
- utils: checkstyle.py: Don't run commit title checker on staged commits
- framebuffer: Allow inheritance of FrameBuffer
- Remove transform from V4L2SubdeviceFormat
- libcamera: camera_sensor: Add tryFormat()
- camera_sensor: Fix typo in comment
- camera_sensor: Demote flip support message to Debug
- internal: tracepoints: Use correct include directive
- libcamera: camera: Ensure queued requests are invalid
ipa:
- doc: Clarify IPA acronym in its namespace description
- rkisp1: Add tuning files for PinephonePro sensors
- rkisp1: Add tuning files for Google DRU "Scarlet" sensors
- rkisp1: lsc: Fix integer division error
- libipa: Add OV2685 Camera Sensor Helper
- libipa: Add OV5647 Camera Sensor Helper
- raspberrypi: agc: Add "shadows" constraint mode
- raspberrypi: generalise algorithms
- raspberrypi: Fix default frame durations calculation
- raspberrypi: Ensure shutter speed and gain are clipped in the AGC
- raspberrypi: Use the new sensor limits fields in CameraMode
- raspberrypi: Add sensor mode limits to CameraMode
- raspberrypi: Fix crash under LTO
- raspberrypi: imx296: Minor tuning updates
- raspberrypi: agc: Fix overflow in Y value calculation
- raspberrypi: Better heuristics for calculating Unicam timeout
- raspberrypi: Generalise statistics
- raspberrypi: Change Unicam timeout handling
pipeline:
- simple: Validate transform
- imx8-isi: Remove mbusCode from formatsMap_
- imx8-isi: Automatically select media bus code
- imx8-isi: Split Bayer/YUV config generation
- imx8-isi: Break out RAW/YUV format selection
- raspberrypi: Iterate over all Unicam instances in match()
- raspberrypi: Add a Unicam timeout override config options
- raspberrypi: Always use ColorSpace::Raw for raw streams
- raspberrypi: Drop unused code
- ipu3: Drop unused code
apps:
- cam: kms_sink: Drop unique_ptr<> from DRM::AtomicRequest
- cam Return std::optional<> from StreamKeyValueParser::parseRole()
- android: jpeg: Add JEA implementation
ABI Compliance:
* abi-compliance-checker reports 98.6% ABI and 100% API compatibility with
with v0.0.4
Total binary compatibility problems: 1, warnings: 1
Total source compatibility problems: 0, warnings: 0
- First virtual method ~__dt ( ) has been added to this class.
1) The layout of type structure has been shifted by 8 bytes by the added v-table pointer.
2) Size of class has been increased by 8 bytes.
affected symbols: 6 (1.4%)
FrameBuffer::releaseFence ( )
FrameBuffer::setCookie ( uint64_t cookie )
FrameBuffer::cookie ( )
FrameBuffer::planes ( )
FrameBuffer::request ( )
FrameBuffer::metadata ( )
This ABI breakage was introduced by:
4843bfa66dc1 ("libcamera: framebuffer: Allow inheritance of FrameBuffer")
Acked-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
A new warning has been introduced to gcc-13 that produces a false
positive on the cam file sink object:
src/cam/file_sink.cpp:92:45: error: possibly dangling reference to a temporary [-Werror=dangling-reference]
92 | const FrameMetadata::Plane &meta = buffer->metadata().planes()[i];
| ^~~~
src/cam/file_sink.cpp:92:81: note: the temporary was destroyed at the end of the full expression '(& buffer->libcamera::FrameBuffer::metadata())->libcamera::FrameMetadata::planes().libcamera::Span<const libcamera::FrameMetadata::Plane>::operator[](i)'
92 | const FrameMetadata::Plane &meta = buffer->metadata().planes()[i];
| ^
cc1plus: all warnings being treated as errors
Workaround this issue by refactoring the code to take a local const
copy of the bytesused value, rather than a local const reference to the
plane.
Bug: https://bugs.libcamera.org/show_bug.cgi?id=185
Bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107532
Co-developed-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Eric Curtin <ecurtin@redhat.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
[Kieran: Commit and comment reworded prior to merge]
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Starting from 13.1, gcc implements the C++23 version of automatic move
from local variables in return statements (see
https://en.cppreference.com/w/cpp/language/return). As a result, some
previously required explicit `std::move()` in return statements generate
warnings. This is the case when a function returns an object whose type
is a class derived from the class type the function returns:
struct U { };
struct T : U { };
U f()
{
T t;
return t;
}
Up to C++20, the automatic move from local variables selects the move
constructor of class U, which is not the move constructor of the
expression. Overload resolution is then performed a second time, with t
considered as an lvalue. An explicit `std::move(t)` is needed in the
return statement to select the U move constructor.
Starting from C++23, `t` is treated as an xvalue, and the U move
constructor is selected without the need for an explicit `std::move(t)`.
An explicit `std:move()` then generates a redundant-move warning, as in
the valueOrTuple() function in src/py/libcamera/py_helpers.cpp.
Omitting the `std::move()` silences the warning, but selects the copy
constructor of U with older gcc versions and with clang, which
negatively impacts performance.
The easiest fix is to disable the warning. With -Wpessimizing-move
enabled, the compiler will still warn of pessimizing moves, only the
redundant but not pessimizing moves will be ignored.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Add to the RkISP1 data configuration files for the imx258 and ov8858
sensors found on Pine64 PinephonePro devices.
The tuning file contain LSC tables extracted from the Rockchip Android
BSP the PinephonePro ships with.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Tested-by: Robert Mader <robert.mader@collabora.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
