| Age | Commit message (Collapse) | Author |
|---|
|
DmaHeap class is useful outside the RPi pipeline handler too.
Move dma_heaps.h and dma_heaps.cpp to common directories. Update
the build files and RPi vc4 pipeline handler accordingly.
Tested-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> # sc8280xp Lenovo x13s
Tested-by: Pavel Machek <pavel@ucw.cz>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Pavel Machek <pavel@ucw.cz>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Andrey Konovalov <andrey.konovalov@linaro.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
SimpleCameraConfiguration::validate() adjusts the configuration of its
streams (if the size is not in the outputSizes) to the captureSize. But
the captureSize itself can be not in the outputSizes, and then the
adjusted configuration won't be valid resulting in camera configuration
failure.
Tested-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> # sc8280xp Lenovo x13s
Tested-by: Pavel Machek <pavel@ucw.cz>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
Reviewed-by: Pavel Machek <pavel@ucw.cz>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Andrey Konovalov <andrey.konovalov@linaro.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
When file output is requested from cam app, it simply dumps the
processed data and it must be converted to a readable image format
manually. Let's add support for PPM output file format to make files
produced by cam directly readable by image display and processing
software.
For now, only BGR888 output format, which is the simplest one to use, is
supported but nothing prevents adding support for other output formats
if needed. Nevertheless, they would typically need byte reordering or
other conversions for PPM file format. It may be better to find a way to
dump the image data in other output formats directly using some of the
already existing file formats or raw file format converters.
Signed-off-by: Milan Zamazal <mzamazal@redhat.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>
|
|
Due to my mistake, commit 705601781b6a ("v4l2: Unset _TIME_BITS in
addition to _FILE_OFFSET_BITS") got merged with incorrect sorting of
defines. Fix it.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
|
|
libcamera fails to build from source in Debian/Ubuntu on 32-bit
architectures under 64-bit time_t (to avoid the 'year 2038
problem'), because its v4l2 module legitimately un-sets
_FILE_OFFSET_BITS for building but this is not allowed without
also unsetting _TIME_BITS.
Having verified that nothing in this module is sensitive to 64-bit
time_t (none of the functions it intercepts handle time), we also
unset _TIME_BITS to allow this to build as before.
Signed-off-by: Steve Langasek <steve.langasek@canonical.com>
Reviewed-by: Dylan Aïssi <dylan.aissi@collabora.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>
|
|
Add a pipeline handler for the Mali-C55 ISP.
The pipeline doesn't currently support an IPA and does not run
any 3a algorithm but only handles the media graph topology and
formats/sizes configuration
Co-developed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Acked-by: Nayden Kanchev <nayden.kanchev@arm.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
As the Android framework sends to the camera device settings
incrementally (only the ones that change are updated), the CameraDevice
class in the Android camera HAL keeps a copy of the last received
settings to be able to apply controls to the libcamera Camera and to
populate metadata correctly.
When a valid 'camera3Request->settings' is provided, it gets saved to
'lastSettings_' but 'descriptor->settings_' is not initialized until
the next frame (assuming it does not contain more settings).
Fix this by assigning to 'descriptor->settings_' the last saved
settings unconditionally.
Signed-off-by: Fang Hui <hui.fang@nxp.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
|
|
Before adding more tests and more helper classes to lc-compliance,
reorganize the source tree to split test and helpers in two separate
directories.
While at it, rename the 'SimpleCapture' class and its derived classes to
just 'Capture'. Rename the source files accordingly.
Re-sort headers inclusions to please checkstyle.py too.
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
Tested-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
Fix a small typo in a comment.
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>
|
|
When a request fails to queue it is completed but with its status set to
RequestCancelled. Add a check in the requestComplete callback to make
sure that the request was completed successfully.
For the SimpleCaptureUnbalanced test we need to do this check only if
the capture isn't over yet, otherwise the few extra requests that get
cancelled at the end, which is the normal behavior, will make the test
fail.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
|
|
Pipeline handlers may need to know the Bayer order produced by the
sensor when a Transform is applied (horizontal or vertical flip). This
is currently implemented manually in the Raspberry Pi pipeline handler.
Move the implementation to the CameraSensor class to make it usable in
other pipeline handlers.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
|
|
The CameraSensor class tests if the sensor HBLANK control is read-only
by comparing the minimum and maximum values, and documents this as being
a workaround for the lack of a read-only control flag in V4L2. This is
incorrect, as the V4L2 API provides such a flag. Use it to replace the
workaround.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The CameraSensor class has grown a lot since its creation, with many
functions added for different types of purposes. They are not grouped by
categories in the class definition, generating confusion when reading
the header file. Improve readability by sorting functions by category:
- Getters for static data (model, entity, focus lens, ...)
- Format and sensor configuration accessors
- Properties and controls (including test pattern mode)
Update the .cpp file accordingly.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
The CameraSensor::updateControlInfo() function is a wrapper around the
same function of the V4L2Subdevice class. It was meant to be called by
pipeline handlers that modify the sensor configuration directly,
bypassing the CameraSensor::setFormat() function. This never happened,
and the function is called once only, internally to the CameraSensor
class. No external users are foreseen, drop the function and call
V4L2Subdevice::updateControlInfo() directly.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
In preparation for adding alternative implementations of the
CameraSensor class, move the code to a subdirectory to avoid cluttering
the main src/libcamera/ directory.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
The V4L2Subdevice class deals with streams in two places:
- In routing tables, streams as expressed as a pad number and a stream
number in a v4l2_subdev_route instance.
- In the format and selection get and set functions, streams as
expressed using the Stream structure, which binds the pad number and
stream number.
Expressing streams in different ways requires pipeline handlers and
other helpers to convert between the two representations. This isn't
much of an issue yet as libcamera has little stream-aware code, but it
is expected to increasingly become a burden.
To simplify the API, introduce a V4L2Subdevice::Route structure that
mimicks the kernel v4l2_subdev_route structure but represents streams as
V4L2Subdevice::Stream instances. This will improve seamless integration
of routes, formats and selection rectangles.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The main (and only at the moment) use case for the Routing::toString()
function is to print a representation of the routing table in a log
message. The function is implemented using an std::stringstream, and the
returned std::string is then inserted into an std::ostream. This is
inefficient. Replace the function with a specialization of the
operator<<() and use it in the caller.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Extend the V4L2Subdevice API with stream support for the functions that
get and set formats and selection rectangles. Add a Stream structure to
identify a subdev pad and stream, and use it to extend the V4L2Subdevice
functions that get and set formats and selection rectangles with stream
support.
To preserve the existing pad-based API, implement overloaded functions
that wrap the new stream-based API. This allows callers that are not
stream-aware to use a simpler pad-based API, instead of having to
explicitly set the stream number to 0 in all API calls.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The V4L2SubdeviceFormat::mbus_code member doesn't follow the libcamera
coding style as it should use camelCase. Fix it by renaming it to just
'code', to shorten lines in addition to fixing the coding style.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The V4L2SubdeviceFormat::bitsPerPixel() function is just a wrapper
around a MediaBusFormatInfo lookup. It made sense when the
MediaBusFormatInfo class was not exposed outside of the compilation
unit, but is now redundant. Drop it and use MediaBusFormatInfo in the
only caller.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Not all media bus formats describe image formats. Extend the
MediaBusFormatInfo class with a type member to indicate if the format
corresponds to image data or metadata, and add the (only) metatdata
format supported by the kernel to the known mediaBusFormatInfo.
The kernel doesn't (yet) have any metadata format specific to sensor
embedded data. This is being addressed in the V4L2 API. In preparation
for embedded data support, already introduce the EmbeddedData type here.
Corresponding formats will be added when available.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The V4L2SubdeviceFormatInfo structure, internal to the
v4l2_subdevice.cpp compilation unit, contains information about media
bus formats that will be useful in other parts of libcamera. To prepare
for this, expose the structure in the v4l2_subdevice.h header and turn
it into a class with a similar design as PixelFormatInfo.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
To prepare for exposing the MediaBusFormatInfo structure as an internal
API, add a code member to the structure to store the media bus code.
This makes MediaBusFormatInfo usable standalone, without having to
externally associate the code related to the info.
The entries in the mediaBusFormatInfo map are becoming too long, so
split them on multiple lines.
While at it, swap the order of the members to match the PixelFormatInfo
class for consistency.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The V4L2SubdeviceFormatInfo structure contains information about a media
bus format, not a V4L2 subdevice format. Rename it to MediaBusFormatInfo
to clarify its purpose. Rename the formatInfoMap map accordingly.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
This is useful in many cases although not included in the stl.
Note: This is an ABI incompatible change.
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
Remove the entry from the `buffers_` mapping if
`Camera::exportFrameBuffers()` fails, otherwise
there would be a stale entry in the map.
Signed-off-by: Barnabás Pőcze <pobrn@protonmail.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>
|
|
The compiler generated destructor works fine, so use that.
Signed-off-by: Barnabás Pőcze <pobrn@protonmail.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>
|
|
The single argument, of type `std::shared_ptr<Camera>`,
is passed by value, so it can simply be moved from in order to
avoid calling the copy constructor.
Signed-off-by: Barnabás Pőcze <pobrn@protonmail.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>
|
|
parseRectangles currently always parses Rectangle controls as an array
of Rectangles. This causes non-array Rectangle controls to not be parsed
correctly, as when the ControlValue is get()ed, the non-array assertion
will fail.
Set the ControlValue with a single Rectangle in case a single Rectangle
has been specified in the yaml capture script to fix that.
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Fix embedded data byte-skipping for 14-bit modes (4 out of 7 bytes
carry register data), and allow 14-bit modes in IMX708 PDAF parsing.
Signed-off-by: Nick Hollinghurst <nick.hollinghurst@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
Add MEDIA_BUS_FMT_SBGGR14_1X14 etc to formatInfoMap.
Not adding any Y14 for now.
Signed-off-by: Nick Hollinghurst <nick.hollinghurst@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The documentation for the SensorTemperature control has a spelling
mistake for thermal.
Fix it.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The camera pipeline for MediaTek MT8365 consists of the following:
Raw sensor (+ external ISP) --> SENINF --> CAMSV30 --> DRAM
SENINF (SENsor INterFace) is a CSI-2 receiver.
CAMSV30 (Camera Simple Version) is a DMA Engine which bypasses
ISP3.0 and writes directly to DRAM.
Both SENINF and CAMSV30 are supported by V4L2 drivers.
MT8365 platform consists of a hardware converter called MDP which
supports up to three streams.
Signed-off-by: Suhrid Subramaniam <suhrid.subramaniam@mediatek.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The plane validation in the stride helper incorrectly accepts the number
of planes as a plane index. Fix the off by one issue.
Fixes: e83727a194b5 ("libcamera: PixelFormatInfo: Add functions stride and frameSize")
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Umang Jain <umang.jain@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
|
|
The PixelFormatInfo documentation has typo. Fix two of them.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
|
|
Replace manual implementations of the utils::to_underlying() helper with
calls to the function.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
|
|
C++23 has a std::to_underlying() helper function that converts an
enumeration value to its underlying type. Add a compatible
implementation to the libcamera::utils namespace.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
|
|
Handle the RKISP1_V_IMX8MP version in the rkisp1 IPA. This enables basic
support of the i.MX8MP, excluding the processing blocks specific to the
new ISP version. Those will be supported later.
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
The histogram weights are initialized to hardcoded 1's for each
histogram grid cell. The code uses the wrong variable for the grid size,
resulting in some weights having a 0 value. Fix it.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
Versions of the ISP differ in the processing blocks they include, as
well as in the implementation of some of those blocks. In particular,
they have different numbers of histogram bins oe AE statistics cells.
The algorithms take these differences into account by checking the ISP
version reported by the driver.
These checks are currently scattered in multiple places. Centralize them
in the IPARkISP1::init() function, and store the version-dependent
hardware parameters in the IPA context, accessible by all algorithms.
While at it, drop the IPASessionConfiguration::hw member that stores the
revision number, unused by the algorithms. It can be added back laer to
the IPAHwSettings structure if needed.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
The RkISP1 statistics structure contains multiple arrays whose length
varies depending on the hardware revision. Accessing those arrays is
error-prone, wrap them in spans at the top level to reduce risks of
out-of-bound accesses.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Stefan Klug <stefan.klug@ideasonboard.com>
|
|
When validating a stream, and no valid packed pixel format can be
found, see if an unpacked format can be used instead.
This is particularly helpful for 8 (and 16) bit raw formats, where
asking for a packed format would previously have failed. Now the
configuration will be adjusted to give you a format (in fact, the only
format) that will work.
Signed-off-by: David Plowman <david.plowman@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
These formats are now defined in upstream GStreamer main branch, so it
is now safe to use their names. Note that libcamera only supports little
endian variants of these formats.
Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
This enables monochrome support in libcamerasrc.
Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
|
|
Several functions in libcamera classes are marked as thread-bound,
restricting the contexts in which those functions can be called. There
is no infrastructure to enforce these restrictions, causing difficult to
debug race conditions when they are not met by callers.
As a first step to solve this, add an assertThreadBound() protected
function to the Object class to test if the calling thread context is
valid, and use it in member functions of Object subclasses marked as
thread-bound. This replaces manual tests in a few locations.
The thread-bound member functions of classes that do not inherit from
Object are not checked, and neither are the functions of classes marked
as thread-bound at the class level. These issue should be addressed in
the future.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
|
|
Object instances are meant to be deleted from the thread they are bound
to, and this requirement is documented. There are however exceptions to
the rule, as threads may be stopped and restarted, with objects bound to
them not being deleted and recreated for every stop/restart cycle. Bound
objects may therefore need to be deleted after the thread has stopped,
making it impossible to use Object::deleteLater().
Document the lifetime requirements more precisely, and enforce them with
an assertion.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
|
|
Objects can be scheduled for deletion with Object::deleteLater(), which
queues a deferred deletion to the thread's event loop. As the
deleteLater() function is meant to be called from a different thread,
this may race with thread termination, and deferred deletions queued
just before calling Thread::exit() may not be processed by the event
loop. Make sure they get processed when finishing the thread, before
stopping.
This eliminates the race condition that occurs when calling
Object::deleteLater() followed by Thread::exit() from the same thread.
Calling deleteLater() from neither the thread the object is bound to or
the thread calling Thread::exit() is still inherently racy.
The change fixes a failure in the object-delete unit test.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
|
|
The signal.h header doesn't need to include object.h. Replace it with a
forward declaration, and instead include object.h in source files that
require it. It can speed up compilation a little bit, but more
importantly avoids unintended dependencies from the Signal class to the
Object class to be added later as the compiler will catch them.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
|
|
The Object::message() function is documented as thread-bound without
using the correct \threadbound reference. Fix it to ensure it gets
included in the thread safety context list.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Milan Zamazal <mzamazal@redhat.com>
|
|
Clearly all cases in the switch are already satisfied, but some
compilers fail to realize this and spit out an error:
Compiler version: gcc 11.2.0 "aarch64-buildroot-linux-gnu-gcc.br_real (Buildroot 2021.11) 11.2.0"
../../src/apps/common/dng_writer.cpp: In function ‘void thumbScanlineIPU3(const FormatInfo&, void*, const void*, unsigned int, unsigned int)’:
../../src/apps/common/dng_writer.cpp:277:55: error: ‘val4’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
277 | uint8_t value = (val1 + val2 + val3 + val4) >> 10;
| ^~~~
../../src/apps/common/dng_writer.cpp:277:48: error: ‘val3’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
277 | uint8_t value = (val1 + val2 + val3 + val4) >> 10;
| ^~~~
../../src/apps/common/dng_writer.cpp:277:41: error: ‘val2’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
277 | uint8_t value = (val1 + val2 + val3 + val4) >> 10;
| ^~~~
../../src/apps/common/dng_writer.cpp:277:34: error: ‘val1’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
277 | uint8_t value = (val1 + val2 + val3 + val4) >> 10;
| ^~~~
Add a default case for the switch-case on a modulo to silence this.
Bug: https://bugs.libcamera.org/show_bug.cgi?id=207
Signed-off-by: Paul Elder <paul.elder@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>
|
