`thread_measure_stack_free()` previously assumed that reading past the
stack is safe. When the stack was indeed part of a thread, the
`thread_t` structure is put after the stack, increasing the odds of
this assumption to hold. However, `thread_measure_stack_free()` could
also be used on the ISR stack, which may be allocated at the end of
SRAM.
A second parameter had to be added to indicate the stack size, so that
reading past the stack can now be prevented.
This also makes valgrind happy on `native`/`native64`.
To allow a platform independent implementation of tread_arch for different ESP32x SoC variants, the platform specific code for Xtensa based ESP SoCs is moved to a separate module `esp_xtensa`.
cpu/esp_common/syscalls.c💯 error (memleak): Memory leak: mtx
cpu/esp_common/syscalls.c:131: error (memleak): Memory leak: rmtx
cpu/esp_common/syscalls.c:365: error (comparePointers): Subtracting pointers that point to different objects
cpu/esp_common/thread_arch.c:355: error (comparePointers): Comparing pointers that point to different objects
cpu/esp8266/startup.c:59: error (comparePointers): Subtracting pointers that point to different objects
- Add `WORD_ALIGNED` attribute to potentially unaligned allocations
- Use intermediate cast to `uintptr_t` to silence false positives of
`-Wcast-align`
Declare sched_active_thread and sched_active_pid locally in the ESP code for
now. Once the code is cleaned up to no longer tap into scheduler internals but
use the API instead, those can be dropped again.
