Reasons:
- DeviceMaster::_node_map does not need to be shared among instances,
because there is at most 1 instance per Flavor and different Flavors
have non-intersecting device paths.
- Keeping it static would also require a static lock
- DeviceMaster::_node_map was not locked at all when used from
uORB::Manager
So this fixes two synchronization issues:
- Different DeviceMaster objects could access the same static data in
parallel
- getDeviceNode() called from uORB::Manager did not use any locking at all
- priority field uses only the lower 8 bits, so we can merge with the
update_reported flag
- orb_set_interval is not used often, so make the necessary data an
optional pointer and alloc only when needed.
Memory savings:
- pixracer (w. ekf2): 7.3kB
- pixhawk: 5.3kB
This adds two uORB API calls:
- orb_advertise_queue
- orb_advertise_multi_queue
Both add a queue_size parameter to define a maximum number of buffered
item. The existing orb calls use all a queue size of one and thus their
behavior is unchanged. If a writer publishes too fast, the oldest elements
from the queue are silently dropped.
The returned timestamp is always the one from the latest message in the
queue.
Queue size can be set via ioctl during advertisement phase. After that it
cannot be changed anymore.
in detail:
- in the write method: the following are not necessarily atomic operations:
_last_update = hrt_absolute_time();
_generation++;
- appears_updated() was called with a lock held in some cases, but not
in ioctl
- use the SmartLock class, so that unlock() is not needed before every
return call. Makes it less error prone
This is necessary when using multiple instances of a topic. However it does
not free the underlying resources, as it is assumed they will be used again
at a later point.
This fixes the previously introduced unit test. It fixes the case where
orb_subscribe_multi is called multiple times with different instances,
and no publisher advertised the topic yet. In this case all subscribers
got the same instance 0.
Last fix for vdev_posix.cpp introduced a sleep from within
a HRT work item callback which blocks the HRT queue.
The code in uORBDevices_posix.cpp that handles message
throttling was commented out for posix. The code was re-enabled
and now seems to work.
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
Nuttx complains about an unresolved _impure_ptr at link time.
This is a known issue when using STL templates in NuttX on ARM.
Created new ORBMap and ORBSet classes for NuttX.
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
QuRT doesn't support unlink and does not provide getpid().
The DSPAL layer provides access to usleep so an implementation is
no longer needed.
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
This adds support for a dynamic build for QuRT and initial
Multi-uORB changes to enable communication between the DSP and
the application processor.
This part of the changes do not affect the POSIX build. This is
enablement for the QuRT build using Multi-uORB. The second part
of the changes will be added in a new module under src/modules.
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
Made the logging macros generic so they can be used for multiple targets.
Fixed toolchain_native.mk so err.h is included from src/systemlib for posix.
Reduced debug output for uORB.
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
uORB was refactored in order to support the MuORB changes required
for QURT. Those changes wil be added in a subsequent commit.
The files are split out by posix and nuttx so the changes are visible.
When this has been tested, the files can be re-merged and re-tested.
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
Beat Küng