Do not clear block2 in subsequent block request.
When sending a request with block2 option, eg. indicating need
for response to be blockwised, copy the block2 option from the
sent_blockwise list item so that the block2 option will be added
to all requests. This fixes an issue where previously the block2
was only sent for the first blockwise request and not for the
subsequent ones, including the last request. This made the response
not follow the request block2 option.
Added check for valid interface ID to network scan timer start so that it
is not triggered after disconnect. Added variable for mac api so that it
is not tried to set again after re-connecting.
Removed main tasklet ID and used instead tasklet id. Added check
for valid interface ID to network scan timer start so that it
is not triggered after disconnect. Added small delay to re-connect
so that stack has time to send its status events to application
after disconnect.
When application requests link check request with add_link_check_request()
it was added to MAC command buffer. Again when actual TX was started,
stack cherry-picked the request into MAC buffer again resulting in a duplicate
entry in the buffer.
Functionality is changed now so that add_link_check_request() does not add
the request into MAC command buffer but it is cherry-picked into the buffer when
TX is issued.
set_next_channel() is the base function provided by LoRaPHY class and should be overridden
by the PHYs who behave differently as compared to EU868 like PHY layers.
CN470 PHY had been missing such an override.
In addition to that we have provided a parameter "fsb-mask-china" that can be used to
enforce a custom frequency sub-band of operation as most of the base stations in the market
may not support all 96 channels. Such a strategy will help in rapid network acquisition.
The code is written such that access to the data input/output happens word-by-word, and that means unaligned access is fine (though with a performance loss) on Cortex-M3/M4 devices.
Macro MBED_DEPRECATED_SINCE is defined in platform/mbed_toolchain.h which was not included.
If someone used member functions lock (which are prefixed with MBED_DEPRECATED_SINCE since some time), there would be a compile error instead of a warning.
Including mbed_toolchain.h fixes that.
Since `void lock(void)` member function has been added we need to remove default value from deprecated `lock(uint32_t millisec=osWaitForever)` member function since otherwise calling `lock()` would be ambiguous.
Add assertions for statuses returned by osMutexAcquire().
Add void Mutex::lock(void) member function - lock function which does not return status and waits for the mutex forever.
Make osStatus Mutex::lock(uint32_t millisec=osWaitForever) member function deprecated in favour of bool Mutex::trylock_for(uint32_t millisec).
Seppo Takalo
Antti Yli-Tokola
Lari-Matias Orjala
Juho Eskeli
Teppo Järvelin
Cruz Monrreal
Jimmy Brisson
Amanda Butler
Mika Leppänen
Jarno Lamsa
Kimmo Vaisanen
Hasnain Virk
Kevin Bracey
Martin Kojtal
Steven Cooreman
Steven
Przemyslaw Stekiel