It is quite possible that the user request for scheduling an uplink is deferred because of backoff or if it was a CONFIRMED message, a retry may take place on a different datarate and different channel.
We didn't have a hook for such deferred scheduling, telling the user whether the async rescheduling worked or not. This commit adds that capability and now we can tell the application if a scheduling failure took place after the original schedule request was accepted.
failed of its own accord) would prevent futher connect()'s due to the
CONN_IN_PROGRESS_FLAG remaining set. This change clears that flag in
the two cases described.
Ignore TEST_GROUP by AStyle. As this uses macro, is not visible to formatter.
By default, it assumes it is a function and treats methods as blocks inside a
function (inlined {}). We ignore it.
We might just not format UNITTESTS in the future if we face similar issues
in the future.
Although other venders implement this feature by using RTC, in my H/W(RZ_A1), I cannot use RTC because it does not satisfy the spec of LP Ticker (ms order and low frequency between 8 KHz and 64 KHz).
Therefore I implemented this feature by creating 1024 division by MTU2(Multi function Timer pulse Unit 2) in order to satisfy this spec.
As a result of investigating, the most unaffected channel among MTU2 placed on GR-PEACH and GR-LYCHEE was channel 3, so I use channel 3 for this feature.
- mbed_drv_cfg.h
I added a macro of MTU2 channel to this file for commonalizing code for GR-PEACH and GR-LYCHEE, and referenced it's macro at us_ticker.c.
- targets.json
I added a macro for enabling LP Ticker.
- mtu2.c mtu2.h
I defined fuction of MTU2's clock supply and stop.
Because MTU2 is utilized by pwm driver too, those function were referenced at lp_ticker driver and pwm driver.
- lp_ticker.c lp_ticker_init()
In order to satisfy the LP Ticker spec, I implemented by creating 1024 division by MTU2.
When an interrupt is required, it will be set with ticker_set_interrupt().
- lp_ticker.c lp_ticker_free()
This function stops the counting and powerdown the lp_ticker.
- lp_ticker.c lp_read()
This function returns the timer counter of MTU2.
- lp_ticker.c lp_ticker_set_interrupt()
In order to satisfy specifications, I implemented lp_ticker_set_interrupt() function.
- lp_ticker.c lp_ticker_fire_interrupt()
In order to satisfy spec, I implemented lp_ticker_fire_interrupt() function.
Also I added GIC_EnableIRQ for allowing the interrupt at end of function.
- lp_ticker.c lp_ticker_get_info()
To satisfy the spec, I implemented lp_ticker_get_info() function. The value of freq includes rounding off.
DEVICE_EMAC is defined only for boards with default Ethernet emac.
It is not defined for example for Realtek RTL8195AM board that
does not have Ethernet. Removed the check for device emac from emac
greentea tests.
i2c_api implementation for NRF52 does a disable() followed by uninit().
The uniinit() implementation in NRF drivers layer makes another call to
disable. This throws off the state of the I2C instance leading to an
assert. Since the disable is only invoked from the I2C API layer for
Nordic, remove this redundant call.
Signed-off-by: Naveen Kaje <Naveen.Kaje@arm.com>
We had a bug especially in the reception path. Our recv window opening
delays were being calculated on the premise that the radio has to capture
5 preamble symbols out of 8 transmitted by the base station. However, in PHY
layer while setting radio rc settings, we were setting preamble length to be 8.
Preamble length register needs to be configured differently for Uplink and Downlink.
For uplink, we wish to transmit 8 preamble symbols whereas in the reception path we need
to receive 5 preamble symbols at least out of 8.
Alongwith that the maximum range of timing error may vary from platform to platform as it
is based upon the crystal in the chip. We have now made these parameters configurable and
have loaded them with the most optimal defaults.
The issue rose up when using ARMC6. A test case didn't initialize NetID
parameter for ABP while using connect(params) API. NetID is the first 7 bits
of the Device Address. It makes sense to actually remove the net-id parameter
from ABP settings as the stack can deduce it from device address. However, the ABP
structure is exposed in public APIs, so we can't really do that at the moment.
Simpler fix is to move the mask that helps us to extract first 7 bits of the device address
is exposed in lorawan_types.h and the user can use it to deduce correct net-id.
Cruz Monrreal
Hasnain Virk
Martin Kojtal
Wajahat Abbas
Ari Parkkila
David Saada
ccli8
Jiwon Yune
bcostm
Brendan McDonnell
deepikabhavnani
Edd Inglis
Kari Haapalehto
TomoYamanaka
Mika Leppänen
Antti Yli-Tokola
cyliangtw
Jimmy Brisson
Naveen Kaje