Add new tool: Callstack Analyzer

[BenCrafterRED]

I developed a tool to help analyze the call stack of a running program inside RARS at the Univerity of Applied Sciences Zwickau. The intent is that this tool may be used by others to help explain and understand the call stack.

Features

• Choose between different strategies used to recognize different stack frames.
• Compare the results of different strategies.
• Live log of all events regarding call, return, and stack access.
• Dynamic stack diagram including:
  • Visual stack pointer and frame pointer
  • Name of the called symbol corresponding to a frame.
  • Resolution of return addresses to symbol names

Screenshot

---

NOTE: I have used Java 21 and the rest of the code worked just fine. If this were a problem, I could rewrite the Callstack Analyzer to use Java 11 only. However, I would prefer to use Java 21 as this would involve quite a bit additional work.

[ladyslipper-glade]

To what extent would changing to Java 21 cause hardship for the user community, some of whom might be "stuck" on an older version of Java for reasons outside their control? I don't have an answer to this question. Personally, I think moving to Java 21 sounds fine, but I can appreciate how it might be problematic.

[BenCrafterRED]

I just wanted to check if there is current interest in the inclusion of this tool. In case there are still adjustments to be done, I am happy to work on that.

[Shin-703]

Hello! Yes there is interest, the tool would be a nice addition I think. We are in discussion to update the version of Java.

For this PR, I get a lot of AddressErrorException :

Also, even when we disconnect the tool from the program after activating it once, it does not actually disconnect and continues running and slowing down everything.

[BenCrafterRED]

I'm happy there is interest in this tool.

For this PR, I get a lot of AddressErrorException :

I checked this with lw t0, 0xfffffff9 on the latest commit on master (814d573) and got the same exception. This seems to be part of RARS' RISC-V implementation rather than introduced by this PR. As far as I know, RISC-V allows implementations to not support unaligned memory access.

Also, even when we disconnect the tool from the program after activating it once, it does not actually disconnect and continues running and slowing down everything.

Thank you for pointing this out. I've fixed this now.