`bt_mesh_rpl_check` stops iterating `replay_list` if either it found an
entry with the requested source address or unassigned address. When IV
index updated, `bt_mesh_rpl_reset` is called. It will set `old_iv` to 1
for all entries with fresh IV index and remove entries with old IV index.
If the entries with old IV index are mixed with other entries, this will
cause fragmentation of `replay_list`. The next time `bt_mesh_rpl_check`
is called, it may stop iterating `replay_list` earlier than it should
because it will meet an empty entry before it iterates over all entries
in the list.
This commit does defragmentatino of `replay_list` on every
`bt_mesh_rpl_reset` by shiting existing entries to the vacated places.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
This reverts commit 7f51907fda.
The problem with setting the priority at the highest priority possible
is that when the IPC is under high traffic, the WQ could starve the
scheduler.
Move back to a more sane preemptive priority as default value.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Json library parser modify data so thats why we can't parse same
data again. Now Composite observation handler parse SenML Json or
CBOR resource path which is given to new API composite Read API
which not need any new data parser.
Signed-off-by: Juha Heiskanen <juha.heiskanen@nordicsemi.no>
Integrated updated JSON library to LwM2M 1.0 JSON.
Removed Old Json format default choice.
Signed-off-by: Juha Heiskanen <juha.heiskanen@nordicsemi.no>
Integrated updated JSON library to SenML-JSON.
This integrate affect that Coap Block wise transfer is not
supported.
Signed-off-by: Juha Heiskanen <juha.heiskanen@nordicsemi.no>
The commit fixes issue where image management would switch to using
heap, whether developer wanted or not, when CONFIG_HEAP_MEM_POOL
gets value greater than zero.
Now when heap is enabled the user can select whether image management
will keep on using static variable, taking static RAM, or will use
heap to allocate the flash image context only when needed.
For this purpose CONFIG_IMG_MGMT_USE_HEAP_FOR_FLASH_IMG_CONTEXT
has been added, which is available when CONFIG_HEAP_MEM_POOL is enabled.
Fixes#44214
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
The open source controller did not expose support for shutting it down
via the close() API in the HCI driver. Add support for it, which
completes the support for disabling and enabling the Bluetooth stack in
Zephyr.
Closes#3192.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Allow the controller to be deinitialized, adding the whole chain of
calls:
- ll_deinit()
- lll_deinit()
- lll_clock_deinit()
in order to be able to turn everything off, including the controller's
refcount of the LF clock.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
When bt_disable() was introduced, some of the global variables that
handle key aspects of Bluetooth were not converted to be also
re-initialized every time that the stack was disabled and the enabled
again. This meant that the FIFO and semaphore in the device structure
were not reset whenever the corresponding threads using them were shut
down and then restarted.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
The ipc_service_get_tx_buffer() has a timeout parameter that can be used
to wait a certain amount of time for a TX buffer to be available.
Unfortunately, for the static vrings backend, an asymmetry
between remote and host exists that makes the usage of this parameter
confusing when the user requests a buffer when no buffers are available
at that time.
When the remote endpoints requests a TX buffer specifying a certain
size and there are no TX buffers available, the function ignores the
parameter and ipc_service_get_tx_buffer() immediately returns -ENOMEM.
The same case on the host endpoint works correctly only when the
specified timeout is <= 15 seconds. All timeouts > 15 seconds simply
returns -EIO after 15 seconds.
This patch is reworking the timeout management trying to behave
correctly in all the cases.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
The current PHY is read only when automatic PHY
update procedure is enabled, 2M PHY is supported and
extended advertising is enabled and this is done after
application is notified about connection. This leads
to invalid connection info PHY data which always returns
1M PHY because when using extended advertising the connection
can be established on different PHY and host does not read
this value from the controller in most cases. In order to
have a current PHY value updated we need to read it in
connection complete event before the user application
is notified about connection.
Signed-off-by: Kamil Gawor <Kamil.Gawor@nordicsemi.no>
The commit replaces direct access to flash_img_context, for the
purpose of checking how much data has been written, with call
to the flash_img_bytes_written.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
Recently OpenAMP introduced the possibility to set the sizes for TX and
RX buffers per created instance. Expose this also to Zephyr users by
using a DT property "zephyr,buffer-size".
For the sake of simplicity use the same DT property to set the buffer
size for both TX and RX buffers.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
The automatic elevation of security and retry of ATT requests interferes
with some tests that expect authentication failures.
Affecting GATT/CL/GAR/BI-42-C
Signed-off-by: Herman Berget <herman.berget@nordicsemi.no>
EATT is not a requirement for the Multiple Variable Length Read
procedure, but previously one had to enable CONFIG_BT_EATT to enable
support for it.
Signed-off-by: Herman Berget <herman.berget@nordicsemi.no>
This fixes Kconfig that allowed to build PACS without support for sink
or source PACS. Few conditionals that no longer apply were removed.
Enabling ASCS support in specific direction enables related PACS
support.
Since PACS does `depends on` anything, it's safe to make use of `select`
statement.
Signed-off-by: Mariusz Skamra <mariusz.skamra@codecoup.pl>
This renames the UUID to follow the same naming scheme as for
BT_UUID_PACS_SUPPORTED_CONTEXT.
Signed-off-by: Mariusz Skamra <mariusz.skamra@codecoup.pl>
The NVS data lookup time grows linearly with the number of
allocation table entries to walk through, meaning that if
some data pair in the NVS changes frequently, access to
other data pairs that change rarely can take a lot of time.
It is particularly visible when the NVS is used as the
settings backend since the backend needs to perform multiple
NVS reads to find a requested key.
Implement a simple cache that stores an address of the most
recent ATE for all NVS IDs that fall into the given cache
position. CRC8/16 is used as a hash function used to
distribute NVS IDs across the cache entries.
The cache entries are only invalidated when an NVS sector
is erased as part of the garbage collector task.
Signed-off-by: Damian Krolik <damian.krolik@nordicsemi.no>
Currently a shallow clone of a packet will bump the reference count on
all the fragments. The net_pkt_unref() function, however, only drops the
reference count on the head fragment. Fix this by only bumping the ref
count on the head buf during shallow clone.
Only bumping the ref count of head is more in line with the idea that
head buf is not responsible for the fragments of its child.
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>
The commit adds check, to flash_area_open, whether there is any
device driver attached and returns -ENODEV if there isn't any.
This works around a problem where flash_area_open succeeds but
consecutive read/write causes crash.
It is enough to check the condition, and return error, here as
the flash_area_open has to precede, and be checked for success,
any read/write operations.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
Add a pseudo device diver with device tree bindings for coredump.
The device tree bindings exposes memory address/size values to be
included in any dump. And the driver exposes an API to add/remove
dump memory regions at runtime.
Signed-off-by: Mark Holden <mholden@fb.com>
This fixes a bug in the decoding of objlinks. Without this the object
instance id is not retrieved correctly as it tries to read the second id
starting from the colon.
Signed-off-by: Markus Rekdal <markus.rekdal@nordicsemi.no>
The unicast audio group can not have streams removed after the
CIG has been created as part of the QoS set procedure.
The bt_audio_unicast_group_remove_streams function may
leave the unicast group without streams, and thus the
check for "free" group was not correct, and has been replaced
with a simple boolean value.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The unicast client and server will attempt to use a
bidirectional CIS instead of two unidirectional CIS
whenever possible.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Change the return type of bt_audio_codec_qos_to_iso_qos to
void as it cannot fail, and this will make it easier to use.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add supporting for referercing either a sink endpoint,
a source endpoint or both in the bt_audio_iso struct.
This way a single bt_audio_iso can have references to up
to 2 endpoints.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Moves the bt_iso_chan and bt_iso_chan_qos structs
out from the bt_audio_ep struct, and instead use a
reference.
This is due to the fact that an endpoint can ever
only be unidirectional, and a CIS can be bidrectional.
The support for using a single bididrectional CIS
for two audio streams/endpoints is not implemented
in this commit.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Move the NULL-check for the iso_server to an earlier point.
There's no need to do anything on the device if no iso_server
has been registered.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The "iso connect" now also takes a security level option like
"iso listen", so that the central can easily encrypt
and connect a CIS in a single command.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The ISO security implementation works by verifying
against the acl (bt_conn) sec_level field. The
bt_conn sec_level field is only available
if CONFIG_BT_SMP is enabled, so this commit
adds guards for all ISO security checks as well.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
If the required_sec_level is lower than the
conn->sec_level, the central will now initialize the
security procecure to ensure that the CIS is encrypted
properly.
The algorithm implemented is as follows:
1) Check security levels for each (acl, iso) pair
2) For those with insufficient security,
call bt_conn_set_security
3) For those with sufficient security, connect the CIS
4) Once the ISO from 2) has been encrypted, connect the
CIS for the specific ACL
The idea behind this was to implement similar support
for autonomous encryption as we have for L2CAP.
It is more complex for ISO as we are dealing with
an array of (acl, iso) pairs, meaning more can go
wrong.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add check for connection and server security
levels, and reject the CIS if the ACL security
level is too low.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The API is documented as being blocking. Making it nonblocking was an
unintentional API change.
Signed-off-by: Herman Berget <herman.berget@nordicsemi.no>
The L2CAP channel ops->sent callback was required to call the
l2cap_chan_sdu_sent callback.
Previously, the only difference between the l2cap_chan_sdu_sent and
l2cap_chan_seg_sent callbacks was that l2cap_chan_sdu_sent called
ops->sent. This is no longer true and l2cap_chan_sdu_sent should always
be called when an SDU is sent.
Signed-off-by: Herman Berget <herman.berget@nordicsemi.no>
