When compiling with optimization, gcc 9 and newer throw an unitialized
variable warning:
../../src/libcamera/pipeline/ipu3/imgu.cpp: In function ‘void libcamera::{anonymous}::calculateBDSHeight(libcamera::ImgUDevice::Pipe*, const libcamera::Size&, const libcamera::Size&, unsigned int, float)’:
../../src/libcamera/pipeline/ipu3/imgu.cpp:172:17: error: ‘bdsHeight’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
172 | unsigned int bdsIntHeight = static_cast<unsigned int>(bdsHeight);
Neither clang not gcc versions older than 9 complain. This seems to be
a false positive.
However, there's an obvious error in the code. The second while () loop
in the first part of calculateBDSHeight() modifies the bdsHeight
variable set by the first loop even if the second loop doesn't find a
suitable height. This can result in an incorrect bdsHeight value. Fix
this, which also gets rid of the compiler warning.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Tested-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
This reverts commit 5b015e96cc.
The ImgU pipe configuration debug is useful to test the correctness
of the parameters computation against the Intel Python script.
However, the number of debug messages which is printed out by the
configuration procedure is so high it floods the logs, up to the point
that starting the Android camera3 HAL, which tests several configurations
at startup, becomes so slow it is barely usable.
Revert the patch that adds the excessive debug statements, which are mostly
useful only when testing the configuration procedure.
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Acked-by: Umang Jain <umang.jain@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
We need a separate control to report the nominal frame duration, but
it's also useful to report the min/max frame duration values that will
be used. Split the FrameDurations control into FrameDuration and
FrameDurationLimits respectively to support both of these.
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
The supportedDevices may contain entries which have the same driver
but different converters. For example, if we add these two entries:
{ "mtk-seninf", "mtk-mdp", 3 },
{ "mtk-seninf", "mtk-mdp3", 3 },
the simple pipeline handler will always take the first one where it
can acquire the driver and skip the rest.
So, make the changes to support this usecase.
Signed-off-by: Phi-Bang Nguyen <pnguyen@baylibre.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
IPU3CameraData stores requests that have been failed due to a
buffer shortage. The requests should be retried once enough
buffers are available. This sets the retry function as signal to
CIO2Device and IPU3Frame, and invokes it from
CIO2Device::tryReturnBuffer() and IPU3Frame::remove().
Signed-off-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
PipelineHandlerIPU3 returns -ENOBUFS and -ENOMEM on queueing a
request when there are not sufficient buffers for the request.
Since the request will be successful if it is queued later when
enough buffers are available. The requests failed due to a buffer
shortage should be stored and retried later in the FIFO order.
This introduces the queue in IPU3CameraData to do that.
Signed-off-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Update the revision of the ImgU Python configuration tool the
libcamera implementation is based on to commit 243d134
("Fix some bug for some resolutions").
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org
The IF rectangle height is iteratively computed by first subtracting
the scaling factor to the estimated height, then in a successive loop
by adding the same scaling factor until the maximum IF size is not
reached.
As the computed IF height is not cached in any variable, the second
loop over-writes the result of the first one, even if the BDS alignment
condition is not satisfied.
Fix this by caching the result of the two iterations, and use the one
that produced any result, with a preference for the one produced by the
second loop, as implemented in the reference python script.
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Tested-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
As reported in https://bugs.libcamera.org/show_bug.cgi?id=32
allowing resolutions < IF_CROP_MAX leads to a not manageable number
configurations to be tested, slowing down the ImgU pipe configuration
to a point which is not acceptable for production devices.
Filter all resolutions < IF_CROP_MAX to maintain the run-time complexity
acceptable and remove the safety check that was meant to avoid overflows
when computing the IF rectangle sizes.
Bug: https://bugs.libcamera.org/show_bug.cgi?id=32
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
With the recent change to merge existing Request metadata with the
ControlList provided by the IPA in commit fcfb1dc02a ("libcamera:
raspberry: Report sensor timestamp"), we can now write the
controls::ScalerCrop value at the start of the pipeline instead of at
the end.
This change simplifies the logic slightly, and allows us to write all
metadata items to the Request in one place.
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>
Write the controls::SensorTimestamp value in the Request metadata when
the request is popped from the queue ready to run the pipeline. This
ensures that the timestamp is written to the correct Request item,
which may not be at the top of the queue when the Unicam buffer dequeue
occurs.
Fixes: fcfb1dc02a ("libcamera: raspberry: Report sensor timestamp")
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>
The controller algorithms currently run on every frame provided to the
IPA by the pipeline handler. This may be undesirable for very fast fps
operating modes where it could significantly increase the computation
cycles (per unit time) without providing any significant changes to the
IQ parameters. The added latencies could also cause dropped frames.
Pass the FrameBuffer timestamp to the IPA through the controls. This
timestamp will be used to rate-limit the controller algorithms to run
with a minimum inter-frame time given by a compile time constant,
currently set to 16.66ms. On startup, we don't rate-limit the algorithms
until after the number of frames required for convergence.
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>
Report the sensor's timestamp in the Request metadata by using the
Unicam::Image buffer timestamp as an initial approximation.
The buffer's timestamp is recorded at DMA-transfer time, and it does not
theoretically matches the 'start of exposure' definition, but when used
to compare two consecutive frames it gives an acceptable estimation of
the sensor frame period duration.
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
When a Request is completed upon receiving the IPA produced metadata,
the metadata associated with the Request are over-written, deleting
the information set at output buffer completion, such as the
SensorTimestamp.
This commit applies to the RaspberryPi pipeline handler the same change
applied to IPU3 in commit 2ee3dd57e2fc ("libcamera: ipu3: Merge IPA
metadata controls").
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Naushir Patuck <naush@raspberrypi.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Report the sensor's timestamp in the Request metadata using the
completed buffer timestamp.
The buffer's timestamp reports the video capture buffer processing time,
and it does not theoretically matches the 'start of exposure'
definition.
VIMC being a testing platform and the test driver completes the buffers
for each media entity connected in the pipeline one after the other, the
current solution is acceptable for the pipeline.
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Report the sensor's timestamp in the Request metadata using the
completed buffer timestamp.
The buffer's timestamp is recorded at DMA-transfer time, and it does not
theoretically matches the 'start of exposure' definition. Record this with
a \todo entry.
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Report the sensor's timestamp in the Request metadata using the
completed buffer timestamp.
The buffer's timestamp is recorded at DMA-transfer time, and it does not
theoretically matches the 'start of exposure' definition. Record this with
a \todo entry.
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
When a Request is completed upon receiving the IPA produced metadata,
the metadata associated with the Request are over-written, deleting
the information set at output buffer completion, such as the
SensorTimestamp.
This commit applies to the RkISP1 pipeline handler the same change
applied to IPU3 in commit 2ee3dd57e2fc ("libcamera: ipu3: Merge IPA
metadata controls").
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Report the sensor's timestamp in the Request metadata using the
completed buffer timestamp.
The UVC driver reports timestamps of SOE event through metadata, for
which there is no support in the current pipeline implementation.
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
When a Request is completed upon receiving the IPA produced metadata,
they should be merged with the metadata set by the pipeline handler
before reporting them to applications.
Use ControlList::merge() to collect the IPA produced controls to the
request's metadata.
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Report the sensor's timestamp in the Request metadata by using the
CIO2 buffer timestamp as an initial approximation.
The buffer's timestamp is recorded at DMA-transfer time, and it does not
theoretically matches the 'start of exposure' definition, but when used
to compare two consecutive frames it gives an acceptable estimation of
the sensor frame period duration.
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo@jmondi.org>
Add a check, that makes sure that the given FrameBuffer is associated
to a RkISP1FrameInfo instance. The current code tries to access the
frame info without checking if it actually exists.
Fixes: 0eb65e14e ("libcamera: pipeline: rkisp1: Attach to an IPA")
Signed-off-by: Sebastian Fricke <sebastian.fricke@posteo.net>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
When the CIO2 returns a cancelled buffer, we will not queue buffers
to the IMGU.
These buffers should be explicitly marked as cancelled to ensure
the application knows there is no valid metadata or frame data
provided in the buffer.
Provide a cancel() method on the FrameBuffer to allow explicitly
cancelling a buffer.
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
The FrameInfo structure associates the data sent to the IPA
and is essential for handling events.
If it can not be found, this is a fatal error which must be fixed.
While we're changing the line, also correct a spelling error for
'Information'.
Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Previously when a frame got cancelled, the frameInfo flags
metadataProcessed and paramDequeued wouldn't get set, meaning that the
request wasn't able to be completed and cancelled.
Make sure that these flags are set when the frame gets cancelled so that
the request can be cancelled.
This issue happened while running lc-compliance.
Suggested-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Signed-off-by: Niklas Söderlund <niklas.soderlund@ragnatech.se>
When seting up the pipeline, the latest entity in the queue is
taken but the oldest one is poped. This is a mistake. Fix it.
Fixes: 4671911df0 ("pipeline: simple: Use breadth-first search to setup media pipeline")
Signed-off-by: Phi-Bang Nguyen <pnguyen@baylibre.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
For all frame indexes, use the same sequence number as generated
by the Request object.
This allows clear matching of what operations occurred to which request.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
A range-based for loop whose range expression is an array of char
pointers and range variable declaration is a const reference to a
std::string creates a temporary string from the char pointer and binds
the range variable reference to it. This creates a const reference to a
temporary, which is valid in C++, and extends the lifetime of the
temporary to the lifetime of the reference.
However, lifetime extension in range-based for loops is considered as a
sign of a potential issue, as a temporary is created for every
iteration, which can be costly, and the usage of a reference in the
range declaration doesn't make it obvious that the code isn't simply
binding a reference to an existing object. gcc 11, with the
-Wrange-loop-construct option, flags this:
uvcvideo.cpp:432:33: error: loop variable 'name' of type 'const string&' {aka 'const std::__cxx11::basic_string<cha
r>&'} binds to a temporary constructed from type 'const char* const' [-Werror=range-loop-construct]
| 432 | for (const std::string &name : { "idVendor", "idProduct" }) {
| | ^~~~
To please the compiler, make the range variable a const char *. This may
bring a tiny performance improvement, as the name is only used once, in
a location where the compiler can use
operator+(const std::string &, const char *)
instead of
operator+(const std::string &, const std::string &)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Use a const char * type instead of auto, and update the commit message
accordingly.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Ensure that the buffers are associated with the request even if they are
used internally to be able to correctly map back to the resources they
are being used to fulfil.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Ensure that checks on resource availability are handled first, and then
operate on the queues only after the resources are confirmed as
available.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
When the imguOutputBufferReady() detects a cancelled frame, it is
reporting that the metadata has been processed in order to be able to
complete the cancelled request.
This causes the FrameInfo to be completed and deleted early, but then an
active buffer on the IMGU can complete and be unable to find the
FrameInfo for it to complete correctly.
Do not mark metadataProcessed early on the event that a single buffer is
detected as cancelled. The stopping of the V4L2 devices will ensure
that all queued buffers are returned to us and we can follow the normal
and expected shutdown sequence.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Laurent Pinchart
Jacopo Mondi
Paul Elder
Umang Jain
Phi-Bang Nguyen
Hirokazu Honda
Naushir Patuck
David Plowman
Sebastian Fricke
Kieran Bingham
Nícolas F. R. A. Prado
Khem Raj