Add TF-M to Mbed OS, replacing the previous PSA implementation for
TF-M-capable targets. This commit adds files imported from TF-M, without
modification. The version of TF-M imported can be found in
`features/FEATURE_PSA/TARGET_TFM/VERSION.txt`.
These changes switch to TF-M as the sole PSA implementation for v8-M and
dual core targets, with TF-M running on the secure side and Mbed OS
running on the non-secure side. Single core v7-M targets will continue
to have PSA implemented via PSA emulation, implemented by Mbed OS.
Move or remove many PSA-implementing files, as PSA will be provided by
TF-M on non-single-v7-M targets. Delete any files that are not relevant
for PSA emulation mode.
- Remove imported TF-M SPM
- Remove Mbed SPM and tests
- Remove Mbed-implemented PSA services and tests
- Remove PSA_SRV_IMPL, PSA_SRV_IPC, PSA_SRV_EMUL and NSPE.
- Replace PSA_SRV_EMUL and PSA_SRV_IMPL with MBED_PSA_SRV
- Remove any files autogenerated by
"tools/psa/generate_partition_code.py", which no longer exists.
Add new feature `PSA` to support PSA in Mbed OS.
Move the Mbed OS implementation of PSA services for v7-M targets (which
employ PSA emulation, and don't yet use TF-M) to
features/FEATURE_PSA/TARGET_MBED_PSA_SRV. Update the `requires`
attribute in TESTS/configs/baremetal.json to avoid breaking baremetal
testing builds.
Update .astyleignore to match new directory structure
Update Mbed TLS importer to place files into FEATURE_PSA
Create the following generic PSA targets:
* `PSA_Target` (Root level PSA generic target)
* `PSA_V7_M` (Single v7-M PSA generic target)
* `PSA_DUAL_CORE` (Dual-core PSA generic target)
* `PSA_V8_M` (v8-M PSA generic target)
Flatten MUSCA_NS and private MUSCA targets into public MUSCA targets.
Move mcuboot.bin to flat location (removing prebuilt folder)
Signed-off-by: Devaraj Ranganna <devaraj.ranganna@arm.com>
Signed-off-by: Jaeden Amero <jaeden.amero@arm.com>
Prevent singleton lock if the RTOS is not yet ready.
lp_ticker is used during the RTOS initialization process.
ST lp_ticker implementation calls sleep functions
which in turn attempts to print to the console when sleep tracing
is enabled. Console initialization attempts to lock the singleton mutex.
`Mutex::unlock` was decrementing the `_count` member after releasing
the mutex, which meant it was unprotected, exposing a race that
could corrupt the count.
This could lead to an assert in `ConditionVariable::wait`, which
checks that the mutex count is one.
To line up with MemoryPool/Mail alloc, rework naming of get/put
Queue::get -> try_get, try_get_for
Queue::put -> try_put, try_put_for
Mail::get -> try_get, try_get_for
Mail::put (no change, but assert that it works)
In the future the names `get` and `put` can be used for untimed blocking
operations. In the interim, you have to use
`try_get_for(Kernel::wait_for_u32_forever)`.
`Mail::put` differs in that it has always been a non-blocking call, but
it can be assumed to always succeed when used correctly, because the
Queue has enough room to store a pointer to every block in the
MemoryPool. It could in future be made a `void` return, similar to the
change made to `Mutex::lock`.
alloc APIs were generally inconsistent - take the opportunity to align with
other APIs like Semaphore.
alloc -> try_alloc
alloc_for -> try_alloc_for
alloc_until -> try_alloc_until
In future the name `alloc` can be used for an untimed blocking
allocation.
RTE_Component can't be included in the assembly file, as it is application header file
(C/C++) therefore assembler fails in this case. If the symbol RTE_CMSIS_RTOS2_RTX5_ARMV8M_NS is
needed, just define it via config. Therefore I am leaving the rest of ifdef _RTE_ there.
CMSIS repo does not support pre-processor defines, hence multiple assembly
files are added for secure/non-secure and floating point tools.
Mbed OS tools support assembly file pre-processing, but the build system
does not support multiple assembly files for each target, hence updating
the assembly files.
1. Patch RTX so irq_cm4f.S files work with no FPU targets
2. Patch RTX so irq_armv8mml.S files to work with and without FPU
2. Patch RTX so irq_armv8mml.S and irq_armv8mbl.S files to work with secure and
non-secure builds
(cherry picked from commit 96e0689204)
In case of ARM compiler, idle and timer thread stack though assigned
to `.bss.os` section since not zero initialized are part of `data` section.
In this commit, we are moving stacks of idle and timer thread to bss
section and thereby saving ROM space.
(cherry picked from commit 9549fff786)
mbed OS used older RTX4 version and with osThreadDef accepting only 3
parameters, to preserve compatibility we hardcode the 'instances'
parameter to 1.
(cherry picked from commit 428acae1b2ac15c3ad523e8d40755a9301220822)
(cherry picked from commit 4360b7bbf8)
CMSIS repo does not support pre-processor defines, hence multiple assembly
files are added for secure/non-secure and floating point tools.
Mbed OS tools support assembly file pre-processing, but the build system
does not support multiple assembly files for each target, hence updating
the assembly files.
1. Patch RTX so irq_cm4f.S files work with no FPU targets
2. Patch RTX so irq_armv8mml.S files to work with and without FPU
2. Patch RTX so irq_armv8mml.S and irq_armv8mbl.S files to work with secure and
non-secure builds
In case of ARM compiler, idle and timer thread stack though assigned
to `.bss.os` section since not zero initialized are part of `data` section.
In this commit, we are moving stacks of idle and timer thread to bss
section and thereby saving ROM space.
mbed OS used older RTX4 version and with osThreadDef accepting only 3
parameters, to preserve compatibility we hardcode the 'instances'
parameter to 1.
(cherry picked from commit 428acae1b2ac15c3ad523e8d40755a9301220822)
(cherry picked from commit 4360b7bbf8)
Chrono conversions inadvertantly changed the core timed sleep routine
used by the RTOS idle to use `OsTimer::update_and_get_tick()` instead of
`OsTimer::get_tick()`.
Correct this, and expand/clarify documentation and naming to try to
prevent recurrence.
Another minor fix observed while inspecting code - `OsClock` can't just
use `milliseconds`, it should match the period of `OsTimer`, which
theoretically can be different.
Previously we overrode nRF targets to have a larger stack
due to memory required by SoftDevice. Having deprecated SoftDevice
in favour of Cordio for BLE (#12674), such requirement does not
apply anymore.
* Remove ambiguity in single-parameter Queue::put and get
* Fix type problems in RTC test - add missing include
* Don't attempt to use TimerEvent default constructor
* Remove references to Timer::read_duration
Lingkai Dong
Martin Kojtal
George Beckstein
Yuan Cao
Qinghao Shi
Evelyne Donnaes
Ashok Rao
Devaraj Ranganna
Hugues Kamba
Kevin Bracey
jeromecoutant
Bartek Szatkowski
deepikabhavnani
Rajkumar Kanagaraj