For RISCV, backtrace generation on device is not possible without
including and parsing DWARF sections. We extract the crash task stack
and let the host generate the backtrace
Core dump integrity check can now be parametrized through menuconfig.
It can be performed on boot or ignored. When core dump is activated
in the menuconfig, the user can still check the core dump at any time
with the function `esp_core_dump_image_check()`.
Fix a bug where `esp_core_dump_image_get()` was not accessible
when core dump was disabled.
Closes https://github.com/espressif/esp-idf/issues/6620
Add support to tasks stacks in RTC DRAM. Before this fix, any stack
in RTC DRAM would have been considered as corrupted, whichi is not
the case.
Fix a bug related to wrong parameters passed to esp_core_dump_get_stack.
Fix a bug reading fake stack memory, triggering a memory violation.
* Closes https://github.com/espressif/esp-idf/issues/6751
* Merges https://github.com/espressif/esp-idf/pull/6750
When `DIS_USB_JTAG` eFuse is NOT burned (`False`), it is not possible
to set pins 18 and 19 as GPIOs. This commit solves this by manually
disabling USB JTAG when using pins 18 or 19.
The functions shall use `gpio_hal_iomux_func_sel` instead of
`PIN_FUNC_SELELECT`.
This commit includes the refactoring of the core dump feature. Thanks to
this refactoring, it is easier to integrate the support of RISC-V
architecture for this feature.
Fixes ESP-1758
- elf.py: elf related structs
- gdb.py: gdb related functions
- loader.py: extract elf/bin from flash to elf format core file
- xtensa.py: xtensa arch target related structs and functions
Core dump can now be written in plain text if the partition itself is encrypted.
So it is important to modify the partition table if the user wants an encrypted
core dump partition as by default, it is not.
Added some functions in core_dump_port.c to abstract the checksum implementation
from core_dump_flash.c file.
Closes IDF-458
The coredump is now encrypted when saved on an encrypted flash.
The cache used for saving the file onto the flash is bigger,
making the proccess more efficient.
Relates to IDF-458
Closes https://github.com/espressif/esp-idf/issues/2932
* Target components pull in xtensa component directly
* Use CPU HAL where applicable
* Remove unnecessary xtensa headers
* Compilation changes necessary to support non-xtensa gcc types (ie int32_t/uint32_t is no
longer signed/unsigned int).
Changes come from internal branch commit a6723fc
GDB "-thread-info N" command (for a specific thread N) may fail if
GDB can not perform a backtrace. At the same time, "-thread-info"
command succeeds in this situation, returning information for all
threads.
Replace repeated calls to "-thread-info N" with a single call to
"-thread-info", also reducing the total execution time.
* changing dependencies from unity->cmock
* added component.mk and Makefile.projbuild
* ignore test dir in gen_esp_err_to_name.py
* added some brief introduction of CMock in IDF
