Fatal error awareness

[rodrigopex]

Is your feature request related to a problem? Please describe.
I'd like to understand when and why a fatal error occurs on Zephyr. This is difficult to see in some cases because the print cannot always print the message about the fatal error. I believe ZView is a perfect place to present that information.

Describe the solution you'd like
I would like to see a message on the TUI providing me the information that the sys_fatal_error_handler was called and the reason for the panic. In case of a failure, the developer would also receive the line that caused the issue (a dream).

Describe alternatives you've considered
There is a breakpoint in the debugger, but it is not good for daily use. Maybe a trace tool.

Additional context
None.

[wkhadgar]

Love this idea, and I think it's very doable.

Since ZView never halts the core and only reads RAM, I'd lean on Zephyr's own coredump for this. If you turn on CONFIG_DEBUG_COREDUMP with the in-memory backend, the fault already gets written to a known RAM buffer with the reason and the faulting PC inside. ZView would just poll that buffer, and since it already loads the ELF, map the PC back to a file and line and show it in its own little last panic view.

Does that sound like the direction you had in mind?

[wkhadgar]

Only downside is the coredump requirement. But considering this is something that would be enabled while debugging, it makes more sense. Could be opt-in if present, like heaps are today. Not globally available, but possible for those who may need it