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.
This is required for the logger, we just manually add the padding bytes
what would otherwise be done by the compiler. Additionally we reorder
the fields by type, so that padding is only necessary for nested types.
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 fixes a race condition: uORB::Manager::get_instance() is used in a
multi-thread context, but the singleton initialization was not thread-safe.
Further, this avoids having to check for nullptr every time the singleton
is accessed.
uORB::Manager::initialize() is called when uorb is started. No one else
accesses the singleton before that point, because it is only used in the
orb_* methods, and in muorb. Both require uorb to be started already when
they are used.
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 test failed on the pixracer because the subscriber thread was too slow
and thus orb messages got lost. This behavior is expected, but the test
should not fail because of that, so we increase the sleeping time.
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.
This fixes a subtle bug: the instance parameter of orb_advertise is an output
parameter and thus its value can be random. However in node_open this value
is accessed and thus the open(...) call could succeed even though it should
not. This can happen for example if a second advertiser of a topic calls
orb_advertise_multi with *instance=0.
The existing implementation worked only because *instance was initialized
with -1 in most cases.
Julian Oes