coredump_fault#

Overview#

The coredump_fault example demonstrates the basic usage of debug/coredump component.

In this example, a Usagefault is intentionally triggered using a division-by-zero operation. The fault is then captured in the UsageFault_Handler, where the cause of the fault, the current Exception Stack Frame (ESF), and partial RAM contents are saved to flash memory using the coredump component.

Afterward, a software reset is triggered. Upon reboot, the fault’s cause is analyzed using the API provided by the coredump component.

Please note that only flash targets are supported!

Running the demo#

Note: Once the program is download to the target board, please press Reset Button to reset program.

Expected log:

    Coredump Fault Example Start!
    Press any key to trigger usagefault...
    Coredump Fault Example Rebooted!!!
    Press any key to check coredump reason...
    [Info] Size: 3894.
    [Info] Reason: Usagefault caused by an integer divsion by zero!

How to debug#

Download the program and press SW1 to reset the program. Following log will be observed in terminal:
Coredump Fault Example Start! Press any key to trigger usagefault…
Attach to running target with debugger.
Press any key in terminal.

Use GDB server to examine core dump data.#

This feature is verified on armgcc toolchain.

Save the data with an offset of 0x10 from the coredump flash partition(Define in CONFIG_MFLASH_COREDUMP_REGION_START_ADDR) into a binary file (e.g., log.bin).

Note: The offset should be 0x10. Otherwise [ERROR][parser] Log header ID not found… will reported.
Note: Recommend to use Ozone.
Start the custom GDB server using the script coredump_gdbserver.py with the core dump binary log file(saved in step 1) and example elf file as pararmeters:
python ../components/debug/coredump/scripts/coredump_gdbserver.py coredump_fault.elf log.bin --port=2331
Start GDB:
arm-none-eabi-gdb.exe .\coredump_fault.elf

Inside GDB, connect to the GDB server via port 2331:
(gdb) target remote localhost:2331
Output from GDB Client:

Remote debugging using localhost:2331
0x080051d8 in DEMO_TriggetUseFault () at C:/Arepo/west2/sdk-next/mcuxsdk/examples/component_examples/coredump_fault/coredump_fault.c:152
152         x = x / y;

Examine the Core registers:
(gdb) info registers Output from GDB Client:

r0             0x61                97
r1             0x200033b8          536884152
r2             0x0                 0
r3             0x1                 1
r4             0x0                 0
r5             0x0                 0
r6             0x0                 0
r7             0x0                 0
r8             0x0                 0
r9             0x0                 0
r10            0x0                 0
r11            0x0                 0
r12            0x12121212          303174162
sp             0x20003398          0x20003398 <ucHeap+2136>
lr             0x80051f3           134238707
pc             0x80051d8           0x80051d8 <DEMO_TriggetUseFault+40>
xpsr           0x61000000          1627389952

Examine the backtrace:
(gdb) bt Output from GDB Client:

#0  0x080051d8 in DEMO_TriggetUseFault () at C:/Arepo/west2/sdk-next/mcuxsdk/examples/component_examples/coredump_fault/coredump_fault.c:152
#1  0x080051f2 in DEMO_FaultTriggerTask (parameters=<optimized out>) at C:/Arepo/west2/sdk-next/mcuxsdk/examples/component_examples/coredump_fault/coredump_fault.c:73
#2  0x00000000 in ?? ()
Backtrace stopped: previous frame identical to this frame (corrupt stack?)

Query thread info: (gdb) info threads Output from GDB Client:

Id   Target Id                                                                      Frame 
* 1    Thread 1 (name: )                                                              0x080051d8 in DEMO_TriggetUseFault ()
    at C:/Arepo/west2/sdk-next/mcuxsdk/examples/component_examples/coredump_fault/coredump_fault.c:152
2    Thread 2 (name: DEMO_MainTask, state: 0x80, user_options: 0x0, prio: 0x3)      <unavailable> in ?? ()
3    Thread 3 (name: DEMO_FaultTriggerTa, state: 0x0, user_options: 0x0, prio: 0x3) <unavailable> in ?? ()
4    Thread 4 (name: IDLE, state: 0x80, user_options: 0x0, prio: 0x0)               <unavailable> in ?? ()
5    Thread 5 (name: Tmr Svc, state: 0x2, user_options: 0x0, prio: 0x4)             <unavailable> in ?? ()

Update default settings#

This example defaults to dumping thread structure and stacks of all threads, and user can switch to other memory type by change “Memory dump” option in Kconfig. Switch Memory dump

coredump_fault

Contents