* The following v14.2 files are removed from nrfx/legacy:
nrf_drv_clock.h
nrf_drv_common.h
nrf_drv_gpiote.h
nrf_drv_spi.h
nrf_drv_twi.h
nrf_drv_uarth.h
* Remove unneeded references to "nrfx_glue.h" and "nrf_drv_common.h"
* Remove the definition of PACKED from app_util_platform.h (already defined
in mbed_toolchain.h)
* Update all scatterfiles to define MBED_RAM_START and MBED_RAM_SIZE. These
get defined in mbed_lib.json when building with a SoftDevice but were missing
when no SoftDevice is brought in.
* Remove SDK 14.2 SoftDevice files.
* Add missing header file to qspi_api.c.
- Adjust memory for SoftDevice
- Enable PRIO=5 for interrupt priority check
- Change NRF_SD_BLE_API_VERSION to 6
- Add handle and buffer for advertising and scanning
- Remove guard for phy update
- Change scatter files and mbed_lib.json for PR #8607
* Update TARGET_NRF5x/README.md to improve "Changing SoftDevice" section
and added section on NRF52840 CryptoCell310 Support
* Update the file list in TARGET_SDK_15_0/TARGET_SOFTDEVICE_COMMON/README.md
* Add missing CR-LF to Nordic-provided SDK file
* Rename a header file in the TARGET_SOFTDEVICE_S112 tree
This reverts commit 3d2fa535a71b7bad08c6a91a31d74f4492c8948c.
This was a breaking change for the "MBR" and "NONE" builds.
After testing it was also determined that copying the vector
table a second time wasn't required for the "FULL" build.
- Add MBR, NONE and OTA SoftDevice build options for S132 and S140
- Add S112 SoftDevice (single build option)
- Some folder restructuring in TARGET_SOFTDEVICE_COMMON was required
The 15.0 SDK doesn't support the nRF51 so it must continue to use the legacy
NRF drivers. Thus the original common_rtc.c, gpio_api.c and us_ticker.c are
restored and placed under TARGET_NRF5x/TARGET_NRF51.
Likewise the modified common_rtc.c, gpio_api.c and us_ticker.c are moved to
TARGET_NRF5x/TARGET_NRF52 so they are excluded from nRF51 builds.
Ensure that vector table gets initialized properly. The table that we
initialize in startup_nrf52840.S gets wiped out as the section is
declared as noinit. Fix this by implementing the weak function mbed_sdk_init
that inits the vector table.
When the SoftDevice (SD) is enabled we need to set the app_offset
to 0x26000 to make room for the SoftDevice. If we let the build
tools self-manage this, MBED_APP_START is coming out at 0x25000
which is not correct for the Nordic 15.0 SDK.
The app_offset value is translated to MBED_APP_START by the build
infrastructure. We were hard coding MBED_APP_START in the scatter
and ld files as a temporary measure while testing. This now sets
it properly if the SD is being brought in.
- Added legacy nrf_drv_rng.c as there is no merit in fully adopting nrfx_rng.c
- Added nrf_queue library component
- Removed apply_old_config.h (unused for some time now)
- Updated sdk_config.h for queue and RNG support for nRF52832
- Brought back RNG into targets.json for nRF52832
Note: nRF52840 still uses CryptoCell 310 for TRNG
- Use new NRFX header file
- Fix nRF52832 linker script/ld files (hardcoded addresses for now)
- Temporarily remove DEVICE_TRNG for nRF52832 (which broke the build)
- Improve serial_putc() fix so we don't rely on "extra" functions
- Missed some NRFX defines that needed to changed
- Set PWM base clock to 125kHz (needs to be reverted back to 1 MHz)
- Updated sdk_config.h for nRF52_DK builds
- Brought in updates from PR7779 (fix for nRF52 PWM issues)
Also addressed:
- removed dependency on legacy config (excluded apply_old_config.h)
- removed legacy pwm and saadc headers
- Arm Compiler 5 linking issue (a band-aid for now... needs to
be properly addressed for peripheral sharing)
- added missing header in SoftDevice file
- top level files ported from TARGET_NORDIC/TARGET_NRF5x/
Also addressed:
- fixed linking issue for gcc
- added support for nRF52-DK builds, but reverted to using nRF52840 sdk_config.h (must be updated)
- introduced "RTC" to targets.json (might need to be removed eventually)
Reimplement atomic code in inline assembly. This can improve
optimisation, and avoids potential architectural problems with using
LDREX/STREX intrinsics.
API further extended:
* Bitwise operations (fetch_and/fetch_or/fetch_xor)
* fetch_add and fetch_sub (like incr/decr, but returning old value -
aligning with C++11)
* compare_exchange_weak
* Explicit memory order specification
* Basic freestanding template overloads for C++
This gives our existing C implementation essentially all the functionality
needed by C++11.
An actual Atomic<T> template based upon these C functions could follow.
The NRF52_DK and DELTA_DFBM_NQ620 have the SWO pin (p0_18) mapped to
LED2. This means that on startup LED2 turns on after the ITM is
initialized which is confusing. Since most users want LED2 usage
instead of SWO we remove the ITM for these targets.
The nRF52 readme is updated to instruct users how to get SWO support
if they need it.
Add the functions qspi_master_sclk_pinmap, qspi_master_ssel_pinmap and
qspi_master_data0_pinmap-qspi_master_data3_pinmap to all targets with
qspi support.
Use new atomics (exchange, load, store and bool types) to simplify
and improve the atomics in the nRF52 serial HAL.
* Ensure mutexes are released last and atomically when done
done inside a critical section.
* Compare-and-swap is not required for the spinlock - exchange is
sufficient. (Not clear a spinlock is needed anyway, but left in).
* Remove unneeded volatile, and make mutexes bool.
The DEVICE_FOO macros are always defined (either 0 or 1).
This patch replaces any instances of a define check on a DEVICE_FOO
macro with value test instead.
Signed-off-by: Alastair D'Silva <alastair@d-silva.org>
When using UARTSerial sending data over the uart follows the sequence
below:
<-TX done ISR runs and sets a software interrupt to pending
<-Software interrupt fires:
-disables TX done interrupt
-calls UARTSerial TX handler which sends bytes until the uart
buffer filled (writeable returns false). Sending a byte
re-enables the TX done interrupt continuing the cycle
Due to this sequence, if the UARTSerial TX handler does not send a byte
then the transmit state machine mentioned above will get stuck with
the TX done interrupt disabled. The events causing this failure:
<-TX done ISR runs and sets a software interrupt to pending
<-Software interrupt fires:
-disables TX done interrupt
-calls UARTSerial TX handler:
-checks writeable which is true and sends a byte
<- interrupted by a higher priority interrrupt
<- TX done ISR runs, setting software interrupt to
pending again
-checks writeable which is true and sends a second byte
-Software interrupt finishes
<-Software interrupt fires:
-disables TX done interrupt
-calls UARTSerial TX handler:
-checks writeable which is false and DOES NOT SEND A BYTE
-Software interrupt finishes, the TX interrupt is still disabled
*-Byte gets sent but TX done ISR does not fire
This patch prevents the TX lockup by removing the code in the
software interrupt which disables the TX done interrupt. Disabling the
TX done interrupt at this point is not necessary so this code is safe
to remove.
When initializing serial disable all interrupts as some of these may
have been enabled by a bootloader. This ensures that the NRF52
serial driver remains compatible with any bootloader version.
Make the following enhancement:
-Support serial port use without flow control
-Prevent dropped bytes by updating flow control handling
-Remove dead code
Serial port use without flow control:
In the prior implementation there was a window of time between
reloading DMA buffers after a timeout where bytes could be dropped.
This is because the uart needed to be turned off in order to flush the
bytes in the DMA buffer.
This change configures the DMA buffer to only receive one byte at a
time so there is no need to disable the uart to flush it. After each
byte is received the DMA transfer will be over so the transfer will
never be partially complete and need flushing. Since the uart is
always on it is safe to use it even without flow control.
Prevent dropped bytes by updating flow control handling:
To prevent dropped bytes due to high latency the flow control handling
of the RTS line was configured to be asserted automatically by
hardware after each byte. Once the CPU has read the byte and setup
the next receive buffer the RTS line is deasserted to the transfer can
continue. This ensure that when flow control is enabled data won't be
lost due to interrupt latency.
Remove dead code:
With the above changes there is a lot of dead code, such as the timer
handling code. This patch removes the code that is no longer used.
Busy-wait before sending a charecter instead of after. If
serial_writeable has been called first, the busy-wait loop will
be skipped.
Added initialization code to ensure NRF_UARTE_EVENT_TXDRDY is
armed correctly.
In nordic_nrf5_uart_event_handler if the events NRF_UARTE_EVENT_ENDRX
and NRF_UARTE_EVENT_RXSTARTED become pending after the check for
NRF_UARTE_EVENT_ENDRX but before the check for
NRF_UARTE_EVENT_RXSTARTED the RX DMA buffers will be setup incorrectly
by nordic_nrf5_uart_event_handler_rxstarted because active_bank hasn't
been updated. This cause dropped and incorrect data.
This patch fixes that problem by adding a second check for
NRF_UARTE_EVENT_ENDRX after checking for NRF_UARTE_EVENT_RXSTARTED
and skipping processing if NRF_UARTE_EVENT_ENDRX is set. The
subsequent interrupt will process both in the correct order. This
ensures that these events cannot be handled out of order and thus fixes
the corruption.
Due to buggy flow control logic in the UARTE, the stop signal
is not being set as it is supposed to when the the module is
not ready to receive data.
This commit signals the sender to halt transmitting when a DMA
buffer is full and only continue again when the atomic FIFO
buffer has been emptied. This allows platforms with hardware
flow control to minimize all buffers and rely on flow control
instead.
The preprocessor based macro check #if evaluates all
enums as 0 and hence the code does not get compiled.
Since move this to a runtime check where the pin variable
can be correctly evaluated.
Delete mbed_overrides.c as it has a target specific mbed_sdk_init() to
resolve linking problem.
This is a follow on patch to:
https://github.com/ARMmbed/mbed-os/pull/8046
While investigating the RX issue on NRF52_DK after SDK 14 updates,
it is observed that the RX FIFO doesn't get filled up, when the
flow control is disabled. Hence the readable never returns true.
If using Serial interface, the stdio file handles (0, 1, 2) get opened.
This results in configuring the flow control for STDIO, and it is observed
that the RX FIFO gets filled.
However, if RawSerial is used, the STDIO file handles
don't get opened. During the debug process it was observed that if the
flow control is configured once and then set to disabled, RX worked
as expected.
Alternative to this approach is that user application specifically
enables flow control as done in mbed's Greentea test suite. See https://goo.gl/r8nBYH
See https://goo.gl/8VB2qg step 14 for _initio's description.
See test code to reproduce the issue and test fix here: https://goo.gl/AQU1xG
Description
The change in behavior with NRF52's UART RX is documented here. #6891
This change is a fix for the above issue.
--legacyalign, --no_legacyalign are deprecated from ARMC6 compiler, in order to
remove deprecated flags all linker files (GCC and IAR as well to have uniformity)
should strictly align to 8-byte boundary
While investigating the RX issue on NRF52_DK after SDK 14 updates,
it is observed that the RX FIFO doesn't get filled up, when the
flow control is disabled. Hence the readable never returns true.
If using Serial interface, the stdio file handles (0, 1, 2) get opened.
This results in configuring the flow control for STDIO, and it is observed
that the RX FIFO gets filled.
However, if RawSerial is used, the STDIO file handles
don't get opened. During the debug process it was observed that if the
flow control is configured once and then set to disabled, RX worked
as expected.
Alternative to this approach is that user application specifically
enables flow control as done in mbed's Greentea test suite. See https://goo.gl/r8nBYH
See https://goo.gl/8VB2qg step 14 for _initio's description.
See test code to reproduce the issue and test fix here: https://goo.gl/AQU1xG
Description
The change in behavior with NRF52's UART RX is documented here. #6891
This change is a fix for the above issue.
There are scenarios where putc is called within a critical section, e.g
to log ASSERTs in early initialization code. The interrupts being
disabled here prevents the handlers for the UARTE from executing.
This breaks the tx_in_progress flag based approach. The tx_in_progress
never gets reset. Poll on the TXDRDY instead.
It can be recreated with a simple program as shown here:
*************** Current Behavior ****************
++ MbedOS Error Info ++
Error Status: 0x80FF0100 Code: 256 Module: 255
Error Message: F
************** With Fix *************************
++ MbedOS Error Info ++
Error Status: 0x80FF0100 Code: 256 Module: 255
Error Message: Fatal Run-time error
Location: 0x2C0A9
Error Value: 0x0
Current Thread: Id: 0x20005520 Entry: 0x30EBF StackSize: 0x1000 StackMem: 0x20004520 SP: 0x20005490
For more info, visit: https://armmbed.github.io/mbedos-error/?error=0x80FF0100
-- MbedOS Error Info --
nrf failure at .\main.cpp:22
***************************************************
Deleted lines that caused the problem. Note that, in nrf_drv_pwm_init there are lines that check if pwm instance is already running, so we don't even need to check it in nordic_pwm_init.
nrf_drv_uninit should be used in nordic_pwm_restart.
This PR provides implementation of ticker_free() function for the following boards:
ARCH_PRO
EV_COG_AD3029LZ
EV_COG_AD4050LZ
K22F
K64F
K82F
KW24D
KW41Z
LPC546XX
NRF51_DK
NRF52_DK
NUCLEO_F207ZG
NUCLEO_F401RE
NUCLEO_F429ZI
NUCLEO_F746ZG
REALTEK_RTL8195AM
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>
The ITM must be initialized before the SoftDevice, but due to the
lazy initialization in C++ on (at least) GCC the ITM init call
might happen too late.
This commit moves the initialization code into the NRF52 system
startup file.
* Elevate RTC2 interrupt priority to same level as UARTE to prevent
race condition on shared variables.
* Remove unused TXDRDY event code.
* Fix typo in macro.
It has been noticed that there is a problem with us ticker on NRF_52840 board after 32 bit counter has been used for us ticker on NRF52 family boards.
Currently NRF52 symbol is defined only for MCU_NRF52832 (not for MCU_NRF52840) and based on this symbol 16 or 32 bit counter is selected (16 bit for NRF51 family and 32 for NRF52 family).
Since MCU_NRF52840 defines NRF52840_XXAA and provides 32 bit counters, 32 bit counter should be selected also when NRF52840_XXAA symbol is defined.
New changes to Mbed error reporting in 5.9 exposed bug in SPI
driver where an instance was uninitialized twice which triggered
an ASSERT.
This fix keeps track of which instance has been initialized and
only calls uninit when it is safe.
Set the second highest user level, leaving the highest for UART (we are having constant overflows) and two levels below for everything else.
This should increase the timer accuracy.
RTC counter is 24-bit. Upper layer handles counter size and wraps ticks count when interrupt is to be fired before passing it to common_rtc_set_interrupt(), but for consistency and safety reasons we can wrap it again in the NRF driver.
The nrf section features has different implementations for each compilers supported by mbed-os. The header guard was ruling out compiler other than GCC by checking if __GNUC__ is defined. This check is not applicable on mbed os as the ARM compiler compile sources with gnu compatibility.
This patch makes sure that the right implementation is selected for the right compiler . The previous patch has been reverted as it is not reliable.
I decided to move these files to the targets/TARGET_NORDIC/TARGET_NRF5x/TARGET_NRF51 since us_ticker.h is for sure specific for NRF51_DK and common_rtc.c might be valid also for NRF52, but this needs to be checked while porting NRF52_DK board.
Provide the following modifications for lp ticker driver:
- According to NRF51_DK reference manual rtc interrupt cannot be controlled by rtc event. In the previous implementation interrupts were enabled permanently and specific interrupt was enabled/disabled by enabling/disabling the specific event. If event is enabled, then event signal is provided to Programmable Peripheral Interconnect (PPI). If interrupt is enabled, then interrupt signal is provided to Nested Vector Interrupt Controller (NVIC). Disable all events permanently. Enable lp ticker overflow interrupt permanently(needed for RTC), disable lp ticker capture/compare interrupt on init (lp_ticker_init) , enable lp ticker interrupt when lp ticker interrupt is set (lp_ticker_set_interrupt), disable lp ticker interrupt on disable request(lp_ticker_disable_interrupt).
- Provide lp ticker data for higher level (freq: 32kHz / len: 24 bits),
- Add the following features to init routine: disable lp ticker interrupt.
- Make ticker driver to operate on ticks instead of us.
- Simplify lp ticker read and set interrupt routines (upper layers handle conversion to us and interrupt scheduling).
According to new ticker standards the following requirements for us ticker are not met on RRF5 boards:
- has a frequency between 250KHz and 8MHz (currently is driven by 32kHz clock)
- ticker increments by 1 each tick (currently is scaled to 1 MHz by incrementing counter by ~31)
Since BLE softdevice uses TIMER0 the proposition is to use high speed TIMER1 for us ticker configured as follows:
- TIMER counter width: 16 bits (max)
- TIMER frequency: 1MHz
This solution also uses Timer's capture/compare register 0 to specify interrupt time and Timer's capture/compare register 1 to read current timer value.
New directory structure:
* TARGET_SOFTDEVICE_COMMON
* TARGET_SOFTDEVICE_S112
* TARGET_SOFTDEVICE_S132_FULL (MBR + SoftDevice, default)
* TARGET_SOFTDEVICE_S132_OTA (SoftDevice only, for firmware updates)
* TARGET_SOFTDEVICE_S132_MBR (MBR only, for bootloader builds)
* TARGET_SOFTDEVICE_S140_FULL (MBR + SoftDevice, default)
* TARGET_SOFTDEVICE_S140_OTA (SoftDevice only, for firmware updates)
* TARGET_SOFTDEVICE_S140_MBR (MBR only, for bootloader builds)
* TARGET_SOFTDEVICE_NONE
The X_OTA and X_MBR binaries are obtained from the original x_FULL SoftDevice
by splitting it in an MBR part and a SoftDevice part. The MBR is needed for
the bootloader and the SoftDevice for firmware updates.
Build application without SoftDevice:
"target_overrides": {
"*": {
"target.extra_labels_remove": ["SOFTDEVICE_COMMON", "SOFTDEVICE_X_FULL"],
"target.extra_labels_add": ["SOFTDEVICE_NONE"]
}
}
Build application for firmware update using SoftDevice X:
"target_overrides": {
"*": {
"target.extra_labels_remove": ["SOFTDEVICE_X_FULL"],
"target.extra_labels_add": ["SOFTDEVICE_X_OTA"]
}
}
Build bootloader without SoftDevice X:
"target_overrides": {
"*": {
"target.extra_labels_remove": ["SOFTDEVICE_COMMON", "SOFTDEVICE_X_FULL"],
"target.extra_labels_add": ["SOFTDEVICE_X_MBR"]
}
}
After rebase, the build target NRF52840_DK is using TARGET_NRF5x,
instead of TARGET_NRF5. Moved Cryptocell TRNG related code from
`targets/TARGET_NORDIC/TARGET_NRF5/` to `targets/TARGET_NORDIC/TARGET_NRF5x/`
The unified NRF51 target and feature BLE directories have been
reorganized to follow the naming and directory structure of the
NRF52 implementation.
This reorganization does not include TARGET_MCU_NRF51822 and
derived targets.
The MBR VTOR state depends on how the application is booted.
This makes it difficult to initialize the MBR correctly since a
bug prevents the MBR from being initialized more than once.
This commit resets the MBR and SoftDevice to a known state before
initializing the MBR and setting the VTOR through the SoftDevice.
Delayed initialization can cause problems when both UARTE instances
are in use. This change causes each UART object to initialize the
underlying UARTE instance immediately.
The vector table relocation sequence depends on:
1. Whether the SoftDevice is present.
2. The Application is a bootloader or not.
If the SoftDevice is present and the application is a bootloader
the MBR must be initialized to trap SoftDevice service calls
before setting the new vector table address.
The SCB->VTOR must be set to point at the MBR as well.
If the SoftDevice is not present the SCB->VTOR can point at the
new vector table directly.
SPI pins are not initialized correctly according to the current
SPI mode.
This commit changes how the SPI instance is configured and ensures
that:
1. SPI pins are initialized on each object initialization.
2. SPI pins are reinitialized on each mode or object change.
3. SPI instance is only initialized when a change has occured.
* Consolidated device_has and macros to the main MCU targets.
* Moved errata configuration to mbed_lib.json for HAL implementation.
* Moved clock configuration to mbed_lib.json for HAL implementation.
* Moved UART configuration to mbed_lib.json for HAL implementation.
Instance counter keeps track of how many objects have been
initialized and freed. On the first object the instance is
enabled and on the last object the instance is disabled.
Serial implementation uses UARTE instead of UART peripheral:
* EasyDMA is used for reading and writing data.
* Triple buffering for receiving data.
See README for full description.
When possible, the I2C HAL will now use the TWI driver in SDK 14.
The manual I2C API still maps onto the old TWI peripheral directly
because the TWI driver doesn't provide the needed low-level functionality.
George Beckstein
Chris Trowbridge
RFulchiero
Desmond Chen
Naveen Kaje
Ralph F
Kevin Bracey
j3hill
Martin Kojtal
Senthil Ramakrishnan
Cruz Monrreal
deepikabhavnani
Russ Butler
Przemyslaw Stekiel
Alastair D'Silva
Wilfried Chauveau
Russ Butler
Marcus Chang
Ashok Rao
David Saada
Andrew Leech
Yossi Levy
Jammu Kekkonen
Maciej Bocianski
Martin Kojtal
MateuszM
Jiwon Yune
Jiwon Yune
Bernhard Schuster
Melvin van den berg
Vincent Coubard
Anna Bridge
Bartek Szatkowski
Steven Cartmell
paul-szczepanek-arm
Ron Eldor
Jimmy Brisson