According to spec we should ignore messages with incorrect msg size.
This patch adds a check to every opcode handler.
Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl>
```
3.7.3.4 Message error procedure
When receiving a message that is not understood by an element, it shall
ignore the message.
Note: A message can be falsely identified as a valid message, passing
the NetMIC and TransMIC authentication using a known network key and
application key even though that message was sent using different keys.
The decryption of that message using the wrong keys would result in a
message that is not understood by the element. The probability of such a
situation occurring is small but not insignificant.
A message that is not understood includes messages that have one or more
of the following conditions:
• The application opcode is unknown by the receiving element.
• The access message size for the application opcode is incorrect.
• The application parameters contain values that are currently
Prohibited.
Note: An element that sends an acknowledged message that is not
understood by a peer node will not receive any response message.
```
Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl>
Align the capitalization of the term "Bluetooth Mesh" to Bluetooth mesh"
in the documentation. This is done to to match the new updated naming
convention done in Bluetooth SIG. In the upcoming spec versions, it its
used "Bluetooth mesh" with the lower case convention.
Signed-off-by: Ingar Kulbrandstad <ingar.kulbrandstad@nordicsemi.no>
Bluetooth Mesh Vendor model hava company id field.
Accordin MeshPRFV1.0.1 3.7.3.1 Operation codes.
The 3-octet opcodes are used for manufacturer-specific opcodes.
The company identifiers are 16-bit values defined by the
Bluetooth SIG and are coded into the second and third octets of
the 3-octet opcodes.
Therefore, we can speed up the search process by checking whether
CID fields match, rather than comparing opcodes one by one.
Signed-off-by: Lingao Meng <menglingao@xiaomi.com>
This commit fixes a bug where incorrect pointer passed to publish_sent
in access.c caused bus fault.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
Periodic publication would previously build and send the first
publication inside the bt_mesh_model_pub() function, before cancelling
and rescheduling the next publication. The timer handler would only
handle retransmissions, and would abandon the rest of the publication
event if one of the packets failed to send.
This design has three issues:
- If the initial timer cancel fails, the publication would interfer with
the periodic publication management, which might skip an event or
send too many packets.
- If any of the messages fail to publish, the full publication event
would be abandoned. This is not predictable or expected from the API.
- bt_mesh_model_pub() required 384 bytes of stack to build the message,
which has to be factored into all calling threads.
This patch moves all transmission into the publication timer by
replacing k_work_cancel with a single k_work_reschedule(K_NO_WAIT). It
also changes the error recovery behavior to attempt to finish the full
publication event even if some of the transmissions fail.
Split out from #33782.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
The implementation of blocking calls is common for all the client
models.
This change reduces the code duplication by introducing new API that
helps to manage acknowledged messages.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
The Health Fast Period Divisor is stored within
the model publish parameters on the access layer.
The opposite part for divisor restoring has been missed.
Signed-off-by: Aleksandr Khromykh <aleksandr.khromykh@nordicsemi.no>
Moves the if !(pub) check to happen before it gets used in the
initialization of the message context in bt_mesh_model_publish.
Fixes#5116.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
The mesh settings.c module is a giant piece of code responsible for
storing the mesh stack configuration. Such approach makes it difficult
to control the data to be stored, breaks the stack modules'
encapsulation by forcing them to reveal the internal kitchen, which
leads to unpleasant issues such as #19799.
This commit moves the responsibility of storing the configuration
to corresponding modules while keeping control of the moment of storing
the configuration and of starting the stack after the settingss loading
is completed.
This doesn't introduce any abstraction between the mesh settings.c and
other modules as it will add more complexity and overhead than necessary
for the actual task.
Fixes#19850
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
Moves mesh feature configuration to a separate module, deprecating the
bt_mesh_cfg_srv structure. The initial values for the features should
now be enabled through KConfig, where new config entries have been added
for each feature.
This removes the upward dependency on the config server from the core
stack, and makes the config server a pure frontend for the configuration
states, as all spec mandated behavior around the feature states is now
encapsulated.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
Fixes bug where applications that disable model extensions end up in an
infinite loop, and adds support for walking model subtrees, as opposed
to forcing root to be unextended.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
`bt_mesh_model_tree_walk()` was too simplistic and did not track visited
nodes which caused it to fall into infinite loop. Moreover the double
next jump could skip a level causing depth value to be invalid.
Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl>
This is a major refactoring of the handling of the cryptographic
material of both the network and transport layers. The aim is to
encapsulate the key object manipulation, and improve overall modularity.
Pulls Applications and Subnets out of the bt_mesh and into separate
modules, with static storage types on the data. This has several
side-effects:
- The Config Server no longer operates directly on the bt_mesh.subs and
bt_mesh.apps lists, but goes through a public configuration interface,
following the pattern set in #27908.
- All iteration through the keys is done through iteration APIs
- Key resolution on RX and TX is centralized.
- Changes to the keys triggers events the other modules can register
handlers for.
- Friendship credentials are stored in the lpn and friend structures.
Part of #27842.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
Adds propagation of error returns from the model init callbacks in
Access, and removing any other checks for successful init in the
foundation models.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
CC: mesh/access.c
mesh/access.c: In function 'model_has_dst':
mesh/access.c:483:10:
warning: returning 'u16_t *' {aka 'short unsigned int *'} from
a function with return type '_Bool8' {aka 'unsigned char'} makes
integer from pointer without a cast [-Wint-conversion]
483 | return bt_mesh_model_find_group(&mod, dst);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: chao an <anchao@xiaomi.com>
Convert bluetooth mesh to using k_timeout struct. Many of the mesh
modules uses timeout calculations, so it is most practical to keep
the s32_t type and only initialize a k_timeout_t struct when
calling the kernel.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
According Mesh Profile 1.0.1. A application key shall
binding single network key. And Device key shall bind all
network key, and dev key only known by cfg_cli and node self,
only used by cfg_cli & cfg_srv.
Fixes: #21088
Signed-off-by: Lingao Meng <mengabc1086@gmail.com>
The publication context is checked for NULL in bt_mesh_model_publish()
however it was dereferenced before that. Move the assignment to
ctx.send_rel to the same place where other ctx members are set.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Until now the choice of reliable sending (segmented messages with
acks) was implicitly dependent on the size of the payload. Add a new
member to the bt_mesh_model_pub to force using segment acks even when
the payload would fit a single unsegmented message.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Allow models to skip a periodic publish interval by returning an error
from the publish update callback.
Previously, an error return from publish update would cancel periodic
publishing. This can't be recovered from, and as such, no valid model
implementation could return an error from this callback, and there was
no way to skip a periodic publish.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
Add comment that explains why a different byte order is used for the
3-byte opcode on the CID part of the opcode.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
If the duration to publish is roughly the same as the period, we might
end up with elapsed == period, which returns 0 and cancel the periodic
publication. Instead 1 should be returned, just like when the elapsed
time is greater than the period.
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Adds the model extension concept to the access layer, as described in
the Mesh Profile Specification, Section 2.3.6. Extensions are
implemented as a tree, using two pointers in each model:
The extends pointer points to the first extended model, and the next
pointer points to the next sibling or (if the NEXT_IS_PARENT flag is
set) the parent model in the tree, forming a cyclical "Left-child
right-sibling" (LCRS) tree. The tree root can be obtained by calling
bt_mesh_model_root_get(), and the extended models can be walked by
calling bt_mesh_model_tree_walk().
According to the Mesh Profile Specification Section 4.2.3, all models in
the same extension tree share one subscription list per element. This is
implemented by walking the model's extension tree, and pooling the
subscription lists of all models in the same element into one. If the
config server adds a subscription to a model, it may be stored in any of
the model tree's models' subscription lists. No two models in the same
extension tree and element will have duplicate groups listed. This
allows us to increase extended models' capacity for subscriptions
significantly.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
Make it possible to provision devices over advertising bearer (PB-ADV).
Many messages in the provisioning protocol are the same for provisioner
and device so much of the code could be reused by only changing when
they are expected to arrive.
This introduces to concept of local and remote device keys. The models
for cfg_cli and cfg_srv have been updated to reflect this concept. Both
the send and receive path in the transport layer have been updated to
support encrypting/decrypting with local and remote device keys.
When a node has been provisioned it is stored in bt_mesh_net.nodes. If
CONFIG_BT_SETTINGS is enabled, they are also saved to settings. If the
callback node_added in bt_mesh_prov has been set, it will be called for
every node that gets provisioned. This includes when they are retrieved
from settings.
The configuration CONFIG_BT_MESH_NODE_COUNT controls how many nodes that
can be provisioned.
Signed-off-by: Tobias Svehagen <tobias.svehagen@gmail.com>
Creates macros for determining model message lengths based on opcode,
payload length and MIC size. Also adds macro wrapping
NET_BUF_SIMPLE_DEFINE to serve the most common use case.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
Adds a structure of callbacks for each model instance. This allows for
more flexible model implementations, that can interact with the Mesh
stack without going through the application.
For now, only an init callback is added, replacing the init mechanism in
the foundation models. The init callback does not provide the primary
flag that used to be in the foundation model callbacks, but replaces
this with an inline function in access.h.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
The bt_mesh_fixed_group_match() function is intended to match the
various well-known group addresses, however it was never updated when
Proxy support was added.
Fixes#19015
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
As element addresses are sequential, there's no need for iterating
through the elements to find the one matching a unicast address.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
The period_start timestamp wasn't getting properly initialized in the
case that retransmission was not being used. In the case of
retransmission the timestamp was getting updated in the mod_publish()
delayed work callback. Add a send_start callback and do the
initialization there, since this covers both the retransmission as
well as the no-retransmission cases.
Fixes#17820
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This branch is inherited from old code that used to start the timer as
soon as a message was queued for transmission rather than when it was
actually transmitted (the case today). It'll also cause a race in case
the publication goes over the GATT layer since the "sent" callback
happens synchronously in that case.
Fixes#17821
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
We've already got GATT services in subsys/bluetooth/services so
subsys/bluetooth/mesh is a more natural place. Aditionally this aims
to fix the Kconfig dependencies to be able to use mesh together with
BT_CUSTOM (i.e. a custom, potentially non-HCI, host stack).
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2019-07-10 09:41:57 +03:00
Renamed from subsys/bluetooth/host/mesh/access.c (Browse further)
