ARM Compiler 6.13 testing revealed linker errors pointing out
conflicting use of `__user_setup_stackheap` and
`__user_initial_stackheap` in some targets. Remove the unwanted
`__user_initial_stackheap` from the targets - the setup is
centralised in the common platform code.
Looking into this, a number of other issues were highlighted
* Almost all targets had `__initial_sp` hardcoded in assembler,
rather than getting it from the scatter file. This was behind
issue #11313. Fix this generally.
* A few targets' `__initial_sp` values did not match the scatter
file layout, in some cases meaning they were overlapping heap
space. They now all use the area reserved in the scatter file.
If any problems are seen, then there is an error in the
scatter file.
* A number of targets were reserving unneeded space for heap and
stack in their startup assembler, on top of the space reserved in
the scatter file, so wasting a few K. A couple were using that
space for the stack, rather than the space in the scatter file.
To clarify expected behaviour:
* Each scatter file contains empty regions `ARM_LIB_HEAP` and
`ARM_LIB_STACK` to reserve space. `ARM_LIB_STACK` is sized
by the macro `MBED_BOOT_STACK_SIZE`, which is set by the tools.
`ARM_LIB_HEAP` is generally the space left over after static
RAM and stack.
* The address of the end of `ARM_LIB_STACK` is written into the
vector table and on reset the CPU sets MSP to that address.
* The common platform code in Mbed OS provides `__user_setup_stackheap`
for the ARM library. The ARM library calls this during startup, and
it calls `__mbed_user_setup_stackheap`.
* The default weak definition of `__mbed_user_setup_stackheap` does not
modify SP, so we remain on the boot stack, and the heap is set to
the region described by `ARM_LIB_HEAP`. If `ARM_LIB_HEAP` doesn't
exist, then the heap is the space from the end of the used data in
`RW_IRAM1` to the start of `ARM_LIB_STACK`.
* Targets can override `__mbed_user_setup_stackheap` if they want.
Currently only Renesas (ARMv7-A class) devices do.
* If microlib is in use, then it doesn't call `__user_setup_stackheap`.
Instead it just finds and uses `ARM_LIB_STACK` and `ARM_LIB_HEAP`
itself.
The QSPI spec allows alt to be any size that is a multiple of the
number of data lines. For example, Micron's N25Q128A uses only a
single alt cycle for all read modes (1, 2, or 4 bits depending on
how many data lines are in use).
Instead of user defined symbols in assembly files or C files,
use linker scripts to add heap and stack - this is inconsistent
with ARM std linker scripts
Instead of user defined symbols in assembly files or C files,
use linker scripts to add heap and stack - this is inconsistent
with ARM std linker scripts
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.
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>
--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
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.
mbed RTC specifications now dictate that the RTC needs to retain and keep on counting through reset. On Silicon Labs parts, this means the RTC API can not be backed by the Silicon Labs RTC peripheral, since that doesn't provide retention functionality.
Therefore:
* On EFM32GG, EFM32WG, EFM32LG: mbed RTC API is now backed by BURTC.
* On EFM32PG, EFR32MG, EFM32PG12, EFR32MG12: mbed RTC API is now backed by RTCC.
* On EFM32ZG, EFM32HG: mbed RTC API is sadly no longer supported, since these chips don't have retained memory.
Re-implemented both us_ticker and lp_ticker to match the new API and specifications.
Details:
* On EFM32GG, EFM32WG, EFM32LG, EFM32HG, EFM32ZG: Use the RTC peripheral to back lp_ticker, and a TIMER to back us_ticker.
* On EFM32PG, EFR32MG, EFM32PG12, EFR32MG12: Use the RTCC peripheral to back lp_ticker (dual-purpose, also used to back RTC), and a TIMER to back us_ticker.
mbed RTC specifications now dictate that the RTC needs to retain and keep on counting through reset. On Silicon Labs parts, this means the RTC API can not be backed by the Silicon Labs RTC peripheral, since that doesn't provide retention functionality.
Therefore:
* On EFM32GG, EFM32WG, EFM32LG: mbed RTC API is now backed by BURTC.
* On EFM32PG, EFR32MG, EFM32PG12, EFR32MG12: mbed RTC API is now backed by RTCC.
* On EFM32ZG, EFM32HG: mbed RTC API is sadly no longer supported, since these chips don't have retained memory.
# Conflicts:
# targets/TARGET_Silicon_Labs/TARGET_EFM32/lp_ticker.c
# targets/TARGET_Silicon_Labs/TARGET_EFM32/rtc_api.c
# targets/targets.json
* Since mbed does not overwrite itself, make the flashing routines run out of flash by default
* Report a writeable size of 4 bytes (previously erroneously reported a full eraseable page as the minimum write size)
* IRQ handling got updated previously to a non-functional state when both callbacks were registered (it'd fire a fall callback for both rise and fall events). With this update, that faulty behaviour is corrected. Due to delays between the detection of the edge and the handling of the interrupt (and the fact that information about which edge you received on the pin is not stored anywhere), there is no way to be absolutely sure which edge got triggered on the pin. Therefore, we make a best-guess effort by looking at the pin state at the time of IRQ handling, and fire a callback as if that was the end state of the event. This will usually work out fine, except in cases were the signal is toggling faster than the IRQ handler's response time. In that case, a user won't get both callbacks (as expected for a pulse), but only the last event.
* Stripped some dead code.
* Updates driver library to v2.3.1 (2018q1) for bugfixes and convenience functions
* Provides library in correct format (2-byte wchar_t flag) for compiling with ARMCC (#6695 uncovered by #6577)
* Reverts to using a statically-allocated packet buffer since malloc is not thread-safe (and the asserts have been turned on)
* MCUs within a family like EFM32GG can omit some peripherals, e.g. EFM32GG230 doesn't have UART
* This commit adds a check to make them compilable, relevant mainly for custom boards
This commit fixes#5840. Fix verified by running mbed_hal-lp_ticker test suite with preloaded RTC counter such that it wrapped in the middle of the suite.
Also removes explicit sleep blocking from the us_ticker implementation, since sleep blocking for us tickers is done at mbed HAL level now. This was causing one of the lp_ticker tests to fail.
Timer code was written based on integer multiple HF clock frequencies. EFR32 doesn't conform to that (38.4), and so the timestamp ticks were off by 1%. Enough to trip up some CI tests on TB_SENSE_12 (#5496)
Kevin Bracey
Filip Jagodzinski
Matthew Macovsky
int_szyk
Arto Kinnunen
Steven Cooreman
Martin Kojtal
petroborys
Anna Bridge
petroborys
deepikabhavnani
Russ Butler
Przemyslaw Stekiel
Alastair D'Silva
David Saada
Steven
Cruz Monrreal
Yossi Levy
Michael Kaplan
Oren Cohen
PHST
Bartek Szatkowski
amq
Marcus Chang
Cruz Monrreal
Martin Kojtal
Alessandro Angelino