Code was written using MBED_CONF_DEFAULT_* but the mbed_lib.json file
that created these values is prefixed with "nsapi" and therefore
final macros are MBED_CONF_NSAPI_DEFAULT_*
The first test focuse on the transport by testing that the reset sequence is
correctly sent to the controller and properly received.
The second test validates that the reset sequence fullfill the right state
of the stack and ensure that bluetooth initialization succeed.
If the automatic uplink is sent directly the call-stack becomes larger than 1K
which may cause serious problems in debug builds. Just to have a respite between
RX and TX we queue an event for the automatic uplink rather than directly undergoing
an automatic uplink.
A peer device that sends a connection parameter update may get block if it doesn't receive a response.
This patch accept the peer request unconditionnally. At some point this event will be managed by the application.
Depending on the role and the current state of the local device; pairing cancelation should be made with a call to a specific function. Normally the Nordic stack would reject invalid calls if the device is not in the correct state; therefore it was assumed that it was possible to detect the state from sd errors. Unfortunatelly this is not true with the latest softdevices as some calls succeed even if the device is not in the right state.
To solve that issue cancelation looks at the current state of the device first to select the right function that will trigger the pairing cancellation.
Note: the call to sd_ble_gap_authenticate was missing in the previous algorithm
LWIP stack is configured to be 1200 as default. Without debug enabled,
maximum stack size used for asynchronous DNS operations is 880 bytes. With
debug enabled maximum used stack size is 1248. Added configuration
to LWIP to increase stack size by 25 percent when debug is enabled on build.
Currently, if all TX descriptors are in use and IP stack calls K64F/K66F
ethernet driver link out, link out drops the packet. Added 10ms delay
to link out to wait for a descriptor to become available before dropping
the packet.
Changed K64F/K66F power up to return without waiting for link up i.e. for
the ethernet cable to be connected. This is needed for non-blocking use
of driver e.g. for using the driver from event queue.
RX1 and 2 delays needed to be more precise and aggregate tx time was
drifiting because of timing difference between actual tx interrupt and
our processing of that interrupt ever so slightly.
We now take a timestamp of the tx interrupt and take a time diff while
instantiating delay timers. The timestamp is then used to update the aggregate
tx time.
Two new methods are introduced in the LoRaMac class which provide current
timing and current receive slot. These functions are used by LoRaWANStack
for its processing.
Set correct SYSMPU register for proper USB operation. This bug was
introduced when the SYSMPU register names and defines were updated
in the commit:
"K64F: Updated the SYSMPU SDK driver"
93f8cfed05
Nanostack object constructor didn't actually initialise Nanostack.
Nanostack initialisation was deferred until an interface was actually
attached to the stack, which generally happened at first interface
connect.
Not normally a problem, unless you're trying to make direct Nanostack
setup calls prior to connect - some applications do this, and were
relying on ThreadInterface::initialise to do Nanostack
initialisation.
Unfortunately in 5.9 ThreadInterface::initialise no longer does
initialise Nanostack immediately, because the mesh interfaces were
aligned and integrated with the Ethernet interfaces, which did
initialisation on connect().
Make the Nanostack object constructor initialise Nanostack (as the LWIP
constructor does for lwIP), so calling Nanostack::get_instance() is the
5.9 API for Nanostack initialisation.
For future work, APIs like ns_file_system_set_root_path should be
exposed as methods of Nanostack, so everything happens with a single
Nanostack::get_instance().file_system_set_root_path().
- set ppp_active false if close fails in ppp disconnect.
- unset sigio in ppp disconnect
- take ownership of filehandle in CellularNetwork::disconnect even in case of failure
A friend class living in the namespace ble::vendor::cordio and named CordioHCITransportDriver can be added in applications requiring access to internal data of the HCI driver and HCI transport driver.
This is meant to be internal and not easily exploitable by application code.
Two cascading copy-paste errors stopped this working:
* PPPCellularInterface::get_gateway() called
nsapi_ppp_get_ip_address();
* nsapi_ppp_get_gateway() called the interface's get_netmask().
First bug has always been there - second one was introduced in 5.9.
AT_CellularNetwork currently lacks calls to get both netmask and gateway
- this patch would be needed for that when added.
Previously get_mac_address on a ThreadInterface returned the EUI-64
reported by the radio driver. This was required for commissioning, but
was inconsistent with other interfaces, and the API concept.
5.9.0 inadvertently changed this so that get_mac_address returned the
actual MAC address used by the radio, which is a hash result of the
EUI-64 for Thread.
The original "return the EUI-64" form was somewhat faulty, as
get_mac_address would not return the EUI-64 set by set_device_eui64() or
another mechanism before connect() was called.
Rather than revert to old behaviour, add a new API to get the device
EUI-64 to ThreadInterface, alongside the existing set API.
When all TX descriptors were reserved in a row so that TX buffer
reclaim interrupt did not happen during reservation sequence, after
the interrupt occurred, TX buffer reclaim did no longer free buffers.
This happened because when all descriptors were in use, last free
index pointed to consumed index.
TX pointer array was using RX ring length in its declaration.
Wasted memory if RX ring > TX ring, as is the default, but would
be broken if RX ring < TX ring.
16 RX buffers and 8 TX buffers is probably excessive. Nanostack
version of driver successfully used 4+4, and data pump should be
broadly equivalent.
This means that switching K64F devices from Nanostack to EMAC increases
base heap usage by 18K - observed in Nanostack border router builds.
Add a config option to make it possible to lower the number of buffers.
Defer consideration of lowering the default to later.
DNS resolver now counts how many sendto operations are successful
during one DNS server attempt cycle (attempting to contact DNS
servers from 0 to last index, including stack specific ones). If all
socket sendto operations fail, DNS resolution is aborted.
If one or more are successful, DNS resolution is continued until
total network attempts count is reached or all sendto operations
fails during a cycle.
If the value is an integer, the 4th bit is used for sign, so you can store
values upto 7 only whereas the datarate values could go upto 15. That's why
we need to turn this to an unsigned integer so that the last bit can also be used.
Travis astyle check pointed out some of the style mismatches in the code.
Not all of them are worth changing as they make the code unreadable and
some of them are semantically wrong.
So in this commit, we have attempted to pick the most important style
mismatches and rectify.
* CENTRAL_LINK_COUNT
* PERIPHERAL_LINK_COUNT
* gatts_enable_params.attr_tab_size
* gatts_enable_params.service_changed
* common_enable_params.vs_uuid_count
These settings control the range of functionality enabled in the softdevice as well as ram consumption.
In particular reducing these values is critical to enable usage of 16K nrf51 devices.
EMAC changes broke PPP glue - fix it up. In particular, PPP steals
the netif->state pointer, so we can't use it as a back pointer to
our LWIP::Interface.
There is a configuration problem with UBLOX_C027 cellular because of missing
LWIP memory configuration for the target.
Added configuration for LPC1768, LPC4088, LPC4088_DM targets and changed
NXP to more specific LPC546XX.
- Set network stack to store event queue so that mbed::mbed_event_queue() call
is not needed every time call_in() is called
- Added dns state variables and enum (states are: created, initiated and cancelled)
- Corrected DNS response handling so that if DNS server returns that host name is
unknown the DNS query is not tried again
- Reorder mutexes in nsapi_dns_query_multiple_async()
- Created nsapi_dns_query_async_initiate_next() function to initiate the next
DNS query from the queue after delete of previous query
- Added dsn_timer_running variable to supervise DNS timer start/stop
- Changed cancel function to only mark query as deleted and moved deletion
to timer function. This allows to run socket close on DNS thread
- Added new nsapi error NSAPI_ERROR_TIMEOUT for DNS (and other) timeouts
General stability improvements are performed.
A flag is added if a Class C RX2 window is open.
We shouldn't open it again if its already opened.
TX_CRYPTO_ERROR is renamed to CRYPTO_ERROR.
Keeping TX_CRYPTO_ERROR for backwards compatibility.
In Class C, rx timeout does not take place for RX2 windows, so if we have
not received anything, we would be retrying but if the no. of retries are
maxed out, and we have not recieved anything yet, we need a mechanism to
tell the upper layer that this has happened.
rx_abort() was sort of dead code as it was doing essentially nothing.
It might have actually meddled with the state machine if it was hit by
invoking on_ack_timeout_timer_event().
State machine and corresponding processors now take care of the ack timeout,
retries and all other bits, so we don't need abort_rx().
MCPS confirmation should be filled with the current number of retries
if the ack is not recieved for a CONFIRMED message.
Ack retry number needs to be incremented after the retry is made not before
that.
A few traces are added at the crucial junctions of the code just to tally with
the conceptual flow for debug purposes.
Application can use cancel_sending() API to stop any outstanding, outgoing
transmission (a TX which is not already queued for transmission). This can
potentially enable use cases where the application could cancel a transmission
and go to sleep if the backoff period is long enough rather than waiting for
the transmission to happen.
This API enables the application to get hold of remaining time after which
the transmission will take place. User can query the backoff time whenever
there is a packet in the TX pipe. If the event for the backoff expiry is
already queued, the stack does not provide backoff metadata.
An API is added to fetch any meta-data available after a succesful
transmission. The stack will make the meta data available after the
TX interrupt is processed. User can get the tx meta data after receiving
TX_DONE event.
Subtract 4 from the received packet length - the buffer contains the
CRC, which we shouldn't pass up.
Ensure we allocate receive buffers of a size corresponding to the
rounded-up size we tell the hardware - the hardware was overrunning the
allocation by a couple of bytes.
- Changed mutexes to platform mutexes
- Removed not needed null check from delete
- Changed nsapi_dns_call_in_set() to use call_in_callback_cb_t and
added prototype to header
- Serialized the sending of multiple async DNS queries since limits on event
message box sizes
- Added timer function that supervises the total time used to make DNS query
and triggers socket timeouts
- Changed nsapi_error_t to new nsapi_value_or_error_t on interface headers
- Corrected wording of gethostbyname_async return values
- Clarified .json options
- Added a new data type for socket callback that can be used from interrupts
- Corrected variable limits to use INT32_MAX etc. defines
- Changed mallocs to new
- Optimized variable sizes on DNS_QUERY definition
- Changed call_in/call methods of the stack to callback provided by the stack
- Specified what are limitations for operations that are made in callback
- Added pure virtual class DNS that defines DNS operations
- Added cancel operation and unique ID to DNS request used in cancel
- Added DNS configuration options to netsocket/mbed_lib.json for retries,
response wait time and cache size
- Changed host name to use dynamic memory in DNS query list and cache,
set maximum length for the name to 255 bytes.
- Added mutex to asynchronous DNS
- Reworked retries: there is now total retry count and a server specific count
- Ignores invalid incoming UDP socket messages (DNS header is not valid), and retries DNS query
- Reworked DNS module asynchronous operation functions
- Corrected other review issues (nothrow new, missing free, missing mutex unlock etc.)
- Added non-blocking DNS interface to network interface and
network stack.
- Added caching of DNS replies.
- Added a network stack function to get DNS addresses from stack.
- Added call and call_in hooks to onboard network stack to
allow calling functions from onboard stack context.
- Added support to call and call_in functions to LWIP and
Nanostack.
- Disabled LWIP DNS translator with the exception of DNS
address storage used in DNS address get.
Implementation of unified EMAC driver for Renesas mbed boards
Based on the driver so far, Renesas implemented the emac driver for GR-PEACH and VK-RZ/A1H.
The mainly changes is below.
- Add the connection part with LWIP according to the unified emac specification.
- Add three new multicast functions(add, remove, set_all).
The Greentea test netsocket and emac test passed.
Just checking "does the chip have an EMAC" doesn't work - there are
targets using those chips which do not have an Ethernet connector and
don't provide the necessary surrounding infrastructure (eg DISCO_F429ZI,
not providing the board emac config call, and HEXIWEAR not providing PHY
info).
Make the targets that actually do want EMAC define their own local
Freescale_EMAC and STM_EMAC labels, and move the drivers into
the corresponding TARGET_ directories, removing the #ifdefs.
Checking DEVICE_EMAC is problematic - particularly for the Odin W2 where
apps have been shutting this off to disable the Wi-fi interface.
Make drivers check a locally-relevant flag instead, pending new
thoughts on how to achieve application/test-relevant flagging for
XXX:get_default_instance() being provided by a system.
However that is achieved, drivers do require a flag set purely by the
target - they mustn't be tripped up by an add-on module providing itself
as the system's default EMAC.
Make Nanostack an OnboardNetworkInterface, implementing
add_ethernet_interface so it can use EMAC drivers.
Can now be used via EthernetInterface, and be the system's default
network stack.
Legacy support for NanostackEthernetInterface retained. Some
restructuring of mesh interface code to fit into the
OnboardNetworkStack:::Interface system.
As we've introduced virtual inheritance to support EMACInterface, we can
no longer use C-style casts or static_cast to downcast from
NetworkInterface to more specific types. RTTI is disabled in the
toolchains, so dynamic_cast is unavailables.
Add virtual downcast methods to permit conversions to the 6 derived
classes. Probably only needed for EMACInterface, WiFiInterface and
EthInterface, but handles the set.
The soft device is not consistent as it is required to force the connection to a resolved resolvable address so it should be known if the target is an identity address.
ns_event_loop_thread_start() is incorrectly used at connect() phase, the initial
setup is already done at init() phase and the eventloop thread is correctly initialized.
Also, the usage of ns_event_loop_thread_start() API should be behind MBED_CONF_NANOSTACK_HAL_EVENT_LOOP_DISPATCH_FROM_APPLICATION
flag as application can decide to use main thread for event loop, which will result in linker error for this API call in
case of ARMCC compiler.
This has been superceded by CellularBase. Name change occurred late
in review of https://github.com/ARMmbed/mbed-os/pull/4119 and
original unused CellularInterface was left behind.
Rather than let "EthernetInterface" be the base EMAC NetworkInterface,
insert an "EMACInterface" class.
EthernetInterface then derives from EMACInterface and EthInterface.
A Wi-Fi driver can derive from EMACInterface and WiFiInterface - this
will be more logical than deriving from EthernetInterface and
WiFiInterface.
This does mean adding a couple of virtual inheritances to avoid
duplicate NetworkInterfaces:
NetworkInterface
/ \
virtual / \ virtual
/ \
EMACInterface WiFiInterface
\ /
\ /
\ /
MyCustomWiFiInterface
Initial work by Bartek Szatkowski in https://github.com/ARMmbed/mbed-os/pull/4079,
reworked following review of https://github.com/ARMmbed/mbed-os/pull/5202 to
transform the entire system into C++, retaining the basic functionality.
Bartek's summary:
* Porting ethernet to EMAC
* Updating EMAC to enable multiple interfaces
* Untangling networking classes, making the abstractions a bit clearer to follow, etc
* General refactoring
* Removal of DEVICE_EMAC flag and introducing DEVICE_ETH and DEVICE_WIFI
Revisions since initial branch:
* Remove lwip depencies
* Correct doxygen warnings
* Remove emac_api.h, replace with C++ EMAC abstract class.
* Create OnboardNetworkInterface, and LWIP implementation.
* Mappings since #4079
lwip-interface/nsapi_stack_lwip.c -> LWIPStack.cpp
lwip-interface/ipstack_lwip.c -> LWIPInterface.cpp
netsocket/mbed_ipstack.h -> OnboardNetworkStack.h
hal/emac_api.h -> EMAC.h
* Reinstate use of EthInterface abstraction
* Correct and clarify HW address EMAC ops
* Restore MBED_MAC_ADDR implementation
* Integrate PPP support with LWIP::Interface.
* Convert K64F lwIP driver to K64F_EMAC.
To do:
* Convert emac_stack_mem.h to follow this pattern.
* Figure out DEVICE_ETH/EMAC
* Update all drivers to use EMAC
The overload of Gap::connect that accept peer_address_t has been added and gap connection and advertising report process have been updated to exploit peer_address_t in a backward compatible fashion.
Deprecation:
* Gap::AdvertisementCallback::addressType has been deprecated in favor of Gap::AdvertisementCallback::peerAddrType.
* Gap::ConnectionCallbackParams::peerAddrType has been deprecated in favor of Gap::ConnectionCallbackParams::peerAddressType.
* Gap::ConnectionCallbackParams::ownAddr has been deprecated in favor of nothing else as this information may be not available.
Overloads added to accept a peer_address_t:
* Gap::connect
* Gap::processConnectionEvent
* Gap::processAdvertisingReport
As this include is not actually needed. Having it will cause issues
with the bootloader, as this will cause a need to get the full
CMSIS/RTOS package etc., which would bloat the bootloader size.
The changes made to BLEProtocol::AddressType was not entirelly backward compatible as BLEProtocol::AddressType split random addresses in three category while the type RANDOM is a superset of these types.
If new events are signaled during processing then they should be processed when processEvent is called again. The goal is to let other processing happen and not process sollely ble events.
This change has been forced by a change in latest softdevice that requires all key pointers to not be NULL unlike what is indicated in the documentation.
Previously EBAD (invalid exchange), mapping the error CORRUPT to EILSEQ
(illegal byte sequence) makes more sense as a description of the type of
error.
Add the allocate_key API. This replaces the previously added set_alloc_key API
(which allocates a key and sets the value at the same time).
Reason for the change: Key allocation will typically be used by other storage
features (like StorageLite), keeping the allocated keys in another location.
Previous API created problems in the case key allocation and value setting
couldn't be done at the same time (for instance, if the set value was
derived from the allocated key, such as hash or CMAC).
Right now, many users are trying out many different filesystems.
Unfortunately, this can leave partially written filesystems on disk
in various states.
A very common pattern for using embedded filesystems is to attempt
a mount, and on failure, format the storage with the filesystem.
Unfortunately, this simply doesn't work if you try to change the
filesystem being used on a piece of storage. Filesystems don't always
use the same regions of storage, and can leave enough metadata lying
around from old filesystems to trick a different mount into thinking a
valid filesystem exists on disk. The filesystems we have were never
designed to check for malicious modification and can't protect against
arbitrary changes.
That being said, it's caused enough problems for users, so as a
workaround this patch adds a disk erase to the FAT filesystem format.
The most common error happens when you use LittleFS, followed by FAT,
followed again by LittleFS.
No other combination of filesystem usage has shown a similar failure,
but it is possible after extensive filesystem use, so it is still
suggested to force a format of the storage when changing filesystems.
In the reception data path, we needed to check the MCPS CONFIRMATION type
not the MCPS INDICATION type. Indication message type is for downlink message type
which can be UNCONFIRMED even if we have sent a CONFIRMED one, e.g., an ACK.
Add support for Alternative ECDSA and ECDH, on the higher level,
over CC310. Note that CC generates ECC keys according to FIPS 186,
while mbed TLS generates according to RFC 6979 and RFC 4754,
which causes test vectors for curve p521 to fail
Teppo Järvelin
Cruz Monrreal
paul-szczepanek-arm
Martin Kojtal
Seppo Takalo
Krzysztof Stachowiak
Ari Parkkila
Mika Leppänen
Anna Bridge
Vincent Coubard
Hasnain Virk
David Saada
Yoshihiro TSUBOI
Amanda Butler
Russ Butler
Kevin Bracey
Michael Schwarcz
Teemu Kultala
Yossi Levy
Joris Aerts
Kimmo Vaisanen
Antti Yli-Tokola
Andrew Leech
Donatien Garnier
Ron Eldor
Oren Cohen
Oren Cohen
Veijo Pesonen
Mirela Chirica
Bartek Szatkowski
Kari Haapalehto
Danny Shavit
Senthil Ramakrishnan
Asif Rizwan
Mahesh Mahadevan
TomoYamanaka
cyliangtw
Yogesh Pande
jeromecoutant
Christopher Haster
Janne Kiiskila
Juhani Puurula
Qinghao Shi