summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)Author
2023-10-18ipa: rpi: contrast: Allow adaptive contrast enhancement to be disabledDavid Plowman
The enableCe() function enables or disables adaptive contrast enhancement and the restoreCe() function sets it back to its normal state (which is what was read from the tuning file). In future, algorithms like HDR might want to take over tonemapping functions, so any dynamic behaviour here would upset them. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: agc: Add an AGC stable regionDavid Plowman
Add a small "stable region" parameter (defaulting to 2%) within which the AGC will not adjust the exposure it requests. It allows applications to configure the AGC to avoid continual micro-adjustments of exposure values if they are somehow sensitive to it. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18pipeline: rpi: Make color space members in RPiCameraConfiguration publicNaushir Patuck
This allows them to be accessed by the pipeline handlers when needed. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18pipeline: rpi: Move flip handling validation codeNaushir Patuck
Move the handling of Bayer order changes due to flips to run before platformValidate(). This removes the need for this code to be split between platformValidate() and validate() as it is right now. Also add some validation to ensure the vc4 pipeline handler only supports CSI2 packing or no packing. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18pipeline: rpi: Add SW downscale status to RPi::StreamNaushir Patuck
Record if additional software downscaling is needed for a particular stream in the RPi::Stream class. Additional software downscaling may be needed if the user required downscale factor is greater than what the ISP hardware is capable of. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Add new algorithms for PiSPNaushir Patuck
Add new CAC, HDR, Saturation and Tonemapping algorithms. Add a new Denoise algorithm that handles spatial/temporal/colour denoise through one interface. With this change, the old SDN algorithm is now considered deprecated and a warning message will be displayed if it is enabled. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Prepare AWB for PiSP supportNaushir Patuck
Prepare the AWB algorithm to support the PiSP hardware. The key change is to factor in the LS correction in the AWB zone statistics. This is different from VC4 where the LS correction happens before statistics gathering. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Prepare ALSC for PiSP supportNaushir Patuck
Prepare the ALSC algorithm to support the PiSP hardware. The key change is to avoid factoring out the WB correction in the AWB zone statistics. Add the ALSC correction to the global metadata so that AWB can use it to factor the gains back in for the AWB calculations. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Add PiSP definitions to the Controller hardware descriptionNaushir Patuck
Add an entry to Controller::HardwareConfig describing the PiSP hardware for the IPA and controller algorithms to use. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Add statsInline to the Controller hardware descriptionNaushir Patuck
Add a new boolean field (statsInline) to Controller::HardwareConfigMap. This field indicates where the statistics are generated in the hardware ISP pipeline. For statsInline == true, statistics are generated before the frame is processed (e.g. the PiSP case), and statsInline == false indicates statistics are generated after the frame is processed (e.g. the VC4 case). Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Add try_lock() to RPiController::MetadataNaushir Patuck
Add the missing try_lock() member function to RPiController::Metadata. This will allow RPiController::Metadata to be used as a template parameter in std::scoped_lock. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18mojom: pipeline: ipa: rpi: Add fields for PiSP objectsNaushir Patuck
Add the PiSP Frontend and Backend SharedMemObject file descriptors to the ipa::init() call. This will allow the pipeline handler to pass these objects to the IPA after construction. Add a flag to indicate if buffer swaps are needed when starting the ISP for the stitch block. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18ipa: rpi: Add IpaBase::platformStart() member functionNaushir Patuck
Add a virtual IpaBase::platformStart() member function that is called at the end of IpaBase::start(). For the IpaVc4 derived class, this function does nothing, but will be used in the PiSP derived class to reset internal state on startup. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18pipeline: rpi: Add SharedMemObject classNaushir Patuck
Add new SharedMemObject class that wraps a memfd memory allocation and constructs a templated object in the memory. With appropriate locking, this object can then be shared across different processes using the associated allocation file handle. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18pipeline: rpi: Add Recurrent and Needs32bitConv flags to RPi::StreamNaushir Patuck
Add a new "Recurrent" stream flag. This flag indicates the stream buffer handling/management happend from the pipeline handler exclusively. This is used for TDN/Stitch and Config streams. Add a new Needs32bitConv stream flag to indicate that this stream needs a software postprocessing conversion run on it before returning out to the application. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-18pipeline: rpi: Add RequiresMmap flag to RPi::StreamNaushir Patuck
Add a new RequiresMmap flag to the RPi::Stream class indicating that buffers handled by the stream must be mmapped after allocation and cached internally. Add a new member function getBuffer(id) which can be used to obtain the mapped buffers for a given buffer id. Add a new member function acquireBuffer() which can be used to obtain any mapped buffer that has not already been acquired by the caller. As a drive-by, add the <algorithm> header to rpi_stream.cpp as it is needed for the std::find_if() function. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-16py: gen-py-controls: Remove SceneFlicker workaroundKieran Bingham
The python bindings layer has to parse the libcamera controls to ensure that they are converted to suitable names for the python layer. Part of this strips out common prefixes from control names, however the SceneFlicker control would end up using an illegal name if processed in the same way as the other controls. The SceneFlicker control has now been removed as part of the introduction of the AeFlickerMode and AeFlickerPeriod controls. Remove the workaround in the python layer. Fixes: 6fdbf3f38c31 ("libcamera: controls: Add controls for AEC/AGC flicker avoidance") Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-16py: Add the SensorConfiguration classDavid Plowman
We provide access to the various fields of the new SensorConfiguration class. The class also needs a constructor so that Python applications can make one and put it into the CameraConfiguration. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-10-06libcamera: converter: drop unnecessary #include directiveAndrey Konovalov
libcamera/internal/media_device.h includes linux/media.h already. Signed-off-by: Andrey Konovalov <andrey.konovalov@linaro.org> Suggested-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-10-06libcamera: converter: A few fixes to ConverterFactoryBase documentationAndrey Konovalov
The description of ConverterFactoryBase::registerType() referred to a converter factory as "converter class" and "converter". Fix that. Also make the descriptions of ConverterFactoryBase::compatibles() and ConverterFactoryBase::create() a bit more specific. Signed-off-by: Andrey Konovalov <andrey.konovalov@linaro.org> Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-10-05ipa: rpi: Avoid skipping IPAs on the first frame after the drop framesDavid Plowman
We avoid skipping the IPAs while frameCount_ is less than dropFrameCount_, indicating that these frames will not be sent to the application. This means that when these numbers are equal then this is the first frame the application will get, so again, we must avoid skipping the IPAs. Consequently the test here must avoid the case of equality. Fixes: 51533fecae8d ("ipa: rpi: Fix frame count logic when running algorithms") Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-27libcamera: rpi: Simplify validate() and configure() for YUV/RGB streamsNaushir Patuck
This commit simplifies the validate() and configure() calls in the pipeline handler in a number of ways: - Determine the V4L2DeviceFormat structure fields for all streams in validate(), cache them and reuse in configure() instead of re-generating this structure multiple times. - Remove setting a default pixel format in validate(), this code patch will not be used. - Use the recently added updateStreamConfig() and toV4L2DeviceFormat() helpers to populate fields in the V4L2DeviceFormat and StreamConfiguration structures to reduce code duplication. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: rpi: Change default stream formatsNaushir Patuck
Switch to XRGB8888 as a default Viewfinder role output format, this is a more correct description of the ISP hardware output, and what is accepted by the Raspberry Pi hardware. Switch to YUV420 as a default output format for everything else, as this format is best supported by encoding (e.g. H.264, JPEG) sinks on the Raspberry Pi platform. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> 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>
2023-09-27libcamera: rpi: Simplify validate() and configure() for RAW streamsNaushir Patuck
This commit simplifies the validate() and configure() calls in the pipeline handler in a number of ways: - Only pass the RPiCameraConfiguration structure into platformValidate() and platformConfigure(). - Determine the V4L2DeviceFormat structure fields for all streams in validate(), cache them and reuse in configure() instead of re-generating this structure multiple times. - Use the recently added updateStreamConfig() and toV4L2DeviceFormat() helpers to populate fields in the V4L2DeviceFormat and StreamConfiguration structures to reduce code duplication. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: rpi: Add some helpers to PipelineHandlerBaseNaushir Patuck
Add a helper updateStreamConfig() that updates the format related fields in a StreamConfiguration from a given V4L2DeviceFormat structure. Add and override to the toV4L2DeviceFormat() helper that returns a V4L2DeviceFormat structure populated from the format related fields in a StreamConfiguration. Both these helper functions will be used in a future commit to simplify the Raspberry Pi pipeline handler configuration/validation code. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: rpi: Cache rawStreams and outStreamsNaushir Patuck
Currently, the stream configuration is stored in two vectors, rawStreams and outStreams for convenience. However, these vectors are constructed in both platformValidate() and platformConfigure(). This change caches these vectors in the RPiCameraConfiguration class to construct them only once in platformValidate(). Pass a pointer to the current configuration to platformValidate() and platformConfigure() so that they can access the streams vectors. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: rpi: Make isRaw/isYuv/isRgb static functionsNaushir Patuck
Move the existing isRaw()/isYuv()/isRgb()into a static function of PipelineHandlerBase. This will allow them to be shared with the pipeline handler derived class. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.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>
2023-09-27libcamera: rpi: Fix wrong comment indentationJacopo Mondi
The closing line of a comment block was aligned with spaces and not tabs. Fix it. 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>
2023-09-27libcamera: rpi: Allow platformValidate() to adjust format stridesNaushir Patuck
Propagate any changes to the format stride done by platformValidate(). The stride value may be adjusted for performace reasons. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: rpi: Handle SensorConfigurationJacopo Mondi
Handle the SensorConfiguration provided by the application in the pipeline validate() and configure() call chains. During validation, first make sure SensorConfiguration is valid, then handle it to compute the sensor format. For the VC4 platform where the RAW stream follows the sensor's configuration adjust the RAW stream configuration to match the sensor configuration. Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: camera_sensor: Support SensorConfigurationJacopo Mondi
Add a class function to the CameraSensor class to apply a full configuration to the sensor. The configuration shall be fully populated and shall apply without modifications to the sensor. Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-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: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-27libcamera: camera: Introduce SensorConfigurationJacopo Mondi
Introduce SensorConfiguration in the libcamera API. The SensorConfiguration is part of the CameraConfiguration class and allows applications to control the sensor settings. Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-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: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
2023-09-25pipeline: rpi: Reset the frame lengths queue during configureNaushir Patuck
The IPA stores a list of the last 10 frame lengths applied to the sensor for determining the timeout to use. This list gets reset on start(), but there is a path through the code that accesses this list in configure() which happens earlier, causing a logical error. Fix this by constructing the list with 10 initial values of 0s. Bug: https://github.com/raspberrypi/libcamera/issues/64 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: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-25ipa: rpi: Fix frame count logic when running algorithmsNaushir Patuck
The frame counter test to determine if we run the IPA algorithms has a logic bug where it treats dropFrameCount_ and mistrustCount_ as frame numbers, not counts of frames (which it is). The implication is that startup convergence and initial settings take one extra frame to apply. Fix this. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Tested-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-25libcamera: formats: Fix typo in YV444 V4L2PixelFormat listingNaushir Patuck
This format was defined with the V4L2_PIX_FMT_YUV444M fourcc instead of the correct V4L2_PIX_FMT_YVU444M fourcc. Fixes: 3b9fe4ae996b ("libcamera: formats: Add YUV444 and YVU444 pixel formats") Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-21android: camera_device: Fix requestedStream handlingHarvey Yang
The Android CameraDevice class adds a sourceStream for each Mapped stream requested by the framework. When mapping multiple framework streams to the same sourceStream, the implementation of CameraDevice::processCaptureRequest wrongly erases the just added sourceStream from the list of streams to request to libcamera. Fix this by adding the stream instead of erasing it. Fixes: 7ea83eba0df6 ("android: camera_device: Postpone mapped streams handling") Signed-off-by: Harvey Yang <chenghaoyang@chromium.org> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-16ipa: rpi: Fix segfault when parsing invalid json fileNaushir Patuck
If the json file parsing failed due to a malformed file, the root pointer would be null. This was not tested and caused a segfault when trying to use the pointer to retrieve the version key. Fix this by bailing out early if the parser returns a null pointer. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-16ipa: rpi: agc: Use channel constraints in the AGC algorithmDavid Plowman
Whenever we run Agc::process(), we store the most recent total exposure requested for each channel. With these values we can apply the channel constraints after time-filtering the requested total exposure, but before working out how much digital gain is needed. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-16ipa: rpi: agc: Add AgcChannelConstraint classDavid Plowman
A channel constraint is somewhat similar to the upper/lower bound constraints that we use elsewhere, but these constraints apply between multiple AGC channels. For example, it lets you say things like "don't let the channel 1 total exposure be more than 8x that of channel 0", and so on. By using both an upper and lower bound constraint, you could fix one AGC channel always to be a fixed ratio of another. Also read a vector of them (if present) when loading the tuning file. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-16ipa: rpi: agc: Implementation of multi-channel AGCDavid Plowman
The switchMode, prepare and process methods are updated to implement multi-channel AGC correctly: * switchMode now invokes switchMode on all the channels (whether active or not). * prepare must find what channel the current frame is, and run on behalf of that channel. * process updates the most recent DeviceStatus and statistics for the channel of the frame that has just arrived, but generates updated values working through the active channels in round-robin fashion. One minor detail in process is that we don't want to change the DeviceStatus metadata of the current frame, so we now pass this to the AgcChannel's process method, rather than letting it find the DeviceStatus in the metadata. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-16ipa: rpi: agc: Reorganise code for multi-channel AGCDavid Plowman
This commit does the basic reorganisation of the code in order to implement multi-channel AGC. The main changes are: * The previous Agc class (in agc.cpp) has become the AgcChannel class in (agc_channel.cpp). * A new Agc class is introduced which is a wrapper round a number of AgcChannels. * The basic plumbing from ipa_base.cpp to Agc is updated to include a channel number. All the existing controls are hardwired to talk directly to channel 0. There are a couple of limitations which we expect to apply to multi-channel AGC. We're not allowing different frame durations to be applied to the channels, nor are we allowing separate metering modes. To be fair, supporting these things is not impossible, but there are reasons why it may be tricky so they remain "TBD" for now. This patch only includes the basic reorganisation and plumbing. It does not yet update the important methods (switchMode, prepare and process) to implement multi-channel AGC properly. This will appear in a subsequent commit. For now, these functions are hard-coded just to use channel 0, thereby preserving the existing behaviour. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-16ipa: rpi: histogram: Add interBinMean()Naushir Patuck
Add a new helper function Histogram::interBinMean() that essentially replaces the existing Histogram::interQuantileMean() logic but working on bins instead. Rework the interQuantileMean() to call into interBinMean() with the appropriate convertion from quatiles to bins. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
2023-09-04ipa: rpi: vc4: Make the StatisticsPtr construction clearerDavid Plowman
StatisticsPtr is a shared pointer, so the use of std::make_unique to create it was a bit confusing. Use std::make_shared instead. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04ipa: rpi: agc: Split AgcStatus into AgcStatus and AgcPrepareStatusDavid Plowman
The Agc::process() function returns an AgcStatus object in the metadata as before, but Agc::prepare() is changed to return the values it computes in a separate AgcPrepareStatus object (under the new tag "agc.prepare_status"). The "digitalGain" and "locked" fields are moved from AgcStatus to AgcPrepareStatus. This will be useful going forward as we can be more flexible about the order in which prepare() and process() are called, without them trampling on each other's results. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04ipa: rpi: agc: Filter exposures before dealing with digital gainDavid Plowman
We now time-filter the exposure before sorting out how much digital gain is required. This is actually a little more natural and simplifies the code. It also prepares us for some future work where this arrangement will be helpful. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrpyi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04ipa: rpi: agc: Fetch AWB status in process method, not prepareDavid Plowman
prepare() doesn't use the AWB status, so fetching it in process() is probably better. This change is preparatory to other changes, where we may find ourselves calling process() without having called prepare() previously. Signed-off-by: David Plowman <david.plowman@raspberrypi.com> Reviewed-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04pipeline: rpi: Simplify buffer id generationNaushir Patuck
Replace the buffer id generation in RPi::Stream with a simple integer counter since ids don't get recycled any more. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04pipeline: rpi: Rename RPi::Stream::setExternalBuffer()Naushir Patuck
Since we don't distinguish between externally and internally allocated dma bufs, rename this function to setExportedBuffer() to clearer on its function. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04pipeline: rpi: Remove additional external dma buf handling logicNaushir Patuck
There is no need to distinguish between dma bufs allocated outside of libcamera and internally allocated buffers. As such, remove all the special case handling of such buffers. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
2023-09-04pipeline: rpi: Increase buffer import count to 32Naushir Patuck
Hardcode the maximum number of buffers imported to the V4L2 video device to 32. This only has a minor disadvantage of over-allocating cache slots and V4L2 buffer indexes, but does allow more headroom for using dma buffers allocated from outside of libcamera. Signed-off-by: Naushir Patuck <naush@raspberrypi.com> Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>