A previous PR merged (to remove runtime observers' dependency with heap)
added inconsistencies and compatibility breaks to the zbus. This commit
improves that by removing the inconsistencies and still attending to the
features requested by the community.
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
Add the option for a unique numeric identifiers to be attached to a
zbus channel. This identifier can then be used to lookup the channel
structure at runtime. This is useful in two situations (that I can
immediately think of).
Firstly for external interaction, i.e a text shell or remote procedure
calls. The current state of a channel can be queried by an ID that never
changes, as opposed to the external entity needing to know the exact
memory address of the channel for a given application binary.
Secondly for integrating with dynamically loaded extensions (llext).
These extensions can hook into the existing data streams without each
individual channel needing to be exported and visible to the loader.
Signed-off-by: Jordan Yates <jordan@embeint.com>
Add optional statistics around the channel publishing action. Store the
time a channel was last published to, and a total publish count.
This information can be used to determine how old a given channels data
is, and an average channel publishing frequency.
Signed-off-by: Jordan Yates <jordan@embeint.com>
ZBus currently does not have a `HEAP_MEM_POOL_ADD_SIZE_ZBUS` when using
message subscribers or runtime observers, forcing the developer to set that
manually. Adding the configuration option to zbus would improve its
usability and make zbus easier to use. With this feature, zbus aligns with
the other Zephyr subsystems' heap memory allocation approach.
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
Currently, zbus uses a single global `net_buf`pool to publish messages to
msg_subscribers. It would be good to have a way to separate the pools for
channels related to critical parts of the systems to avoid publication
failure on these particular channels. These channels will not use the
global pool. They can set an isolated pool by calling the
`zbus_chan_set_msg_sub_pool.`
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
Replace mutexes with semaphores to protect the channels in conjunction with
a priority boost algorithm based on the observers' priority.
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
Add a name for the Kconfig choice symbol
indicating the Zbus subscriber buffer allocation
and adjust the name of the existing choices.
Signed-off-by: Bartosz Bilas <bartosz.bilas@hotmail.com>
Zbus uses mutexes internally that are available
only when MULTITHREADING is enabled so add
it to fix the following error:
/opt/zephyr-sdk-0.16.3/arm-zephyr-eabi/bin/../lib/gcc/arm-zephyr-eabi/
12.2.0/../../../../arm-zephyr-eabi/bin/ld.bfd: zephyr/subsys/
zbus/libsubsys__zbus.a(zbus.c.obj): in function `k_mutex_init':
/builds/zephyr/mcuboot/zephyr/include/generated/syscalls/kernel.h:969:
undefined reference to `z_impl_k_mutex_init'
Signed-off-by: Bartosz Bilas <b.bilas@grinn-global.com>
Besides the changed channel reference, the message subscribers receive a
copy of the message during the VDED execution. ZBus guarantees message
delivery for `MSG_SUBSCRIBERS`.
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
ZBus stores observers in two ways: statically using a list and dynamically
using a memory slab. Both present limitations. Static observers work only
for channel definition. The dynamic observers rely on a memory slab that
forces the user to manage its size to avoid issues with adding
observers. This commit fixes the static allocation problem by using the
iterable sections for allocating observation data and replacing the VDED
execution sequence since now it is possible to prioritize static observer
execution. All the runtime observers are dynamically allocated on the heap
instead of a specific memory pool.
BREAK changes (only internal, not APIs):
* ZBus channel metadata changed. Remove the observers' static array
pointer. Rename the `runtime_observers` pointer to `observers`. Add
`observer_start_idx` and `observer_end_idx`;
* Change the VDED execution sequence. The position (on definition time),
the priority in conjunction with the lexical order, is considered for
static post-definition time observers. At last, the runtime observer
follows the adding sequence;
* Replace the `CONFIG_ZBUS_RUNTIME_OBSERVERS_POOL_SIZE` with
`CONFIG_ZBUS_RUNTIME_OBSERVERS`.
New APIs:
* New iterable section iterators (for channels and observers) can now
receive a user_data pointer to keep context between the function calls;
* New `ZBUS_LISTENER_DEFINE_WITH_ENABLE(_name, _cb, _enable)` and
`ZBUS_SUBSCRIBER_DEFINE_WITH_ENABLE(_name, _queue_size, enable)` that
enable developers define disabled observers. They need to be enabled
during runtime to receive notifications from the bus;
* `ZBUS_CHAN_ADD_OBS` macro for adding post-definition static observers of
a channel.
Important changes:
* Move the ZBus LD file content to the `common-ram.ld` LD file. That was
necessary to make ZBus compatible with some Xtensa and RISCV boards.
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
Add zbus message bus as a Zephyr subsystem. No message bus
or communication abstraction other than the usual (message queues,
mailboxes, etc.) enabled developers to implement event-driven systems in
Zephyr quickly. Zbus would fill that gap by providing the community with
a lightweight and flexible message bus. The implementation tries to be
closest as possible to the existing ones. We use the claim/finish
approach, and the API for publishing and reading channels are similar
in message queues. Zbus is about channels, messages, and observers.
Signed-off-by: Rodrigo Peixoto <rodrigopex@gmail.com>
