[lldb/Lua] Add Boilerplate for a Lua Script Interpreter

[JDevlieghere]

Summary

This adds the boilerplate necessary to support the Lua script interpreter. The interpreter itself is non-functional and just says it's not yet implemented.

Discussion on the mailing list:http://lists.llvm.org/pipermail/lldb-dev/2019-December/015812.html

[JDevlieghere]

Return the right name from GetPluginNameStatic.

[xiaobai]

I'm really excited to see where this goes! :D

lldb/source/Plugins/ScriptInterpreter/Lua/ScriptInterpreterLua.cpp

64

Why is this Null and not Lua? :P

[labath]

lldb/cmake/modules/LLDBConfig.cmake

28

I think this will tick off some bots (and people) because it means that the default configuration will not build unless one has (compatible?) lua installed. Though I don't really like that, the usual way to handle external dependencies in llvm is to detect their presence and automatically disable the relevant functionality.

Now, that's not how things work in lldb right now, so it _may_ make sense to do the same for lua (though it also may make sense to port everything to the llvm style). However, the current lldb behavior has been a source of friction in the past and I suspect a fresh build error might reignite some of that.

Anyway, you have been warned...

332

Can we replace this (and maybe python too, while at it) with a Host/Config.h entry? A global definition means that one has to recompile everything when these change in any way, whereas in practice only a handful of files need this..

[mgorny]

lldb/cmake/modules/LLDBConfig.cmake

28

Fixing this is one of the things at the far end of my todo. If you could look into replacing the disable logic with something better, a lot of people would really be grateful.

[JDevlieghere]

lldb/cmake/modules/LLDBConfig.cmake

28

I've changed the default for now and put up an "RFC" here: https://reviews.llvm.org/D71306

[JDevlieghere]

Reuse IOHandlerEditline for the Lua interpreter.

[JDevlieghere]

^ updated the wrong patch

[JDevlieghere]

Rebase

[labath]

Given the positive response to the RFC, I think we can start landing this stuff.

lldb/cmake/modules/LLDBConfig.cmake

28

That's cool. I think this can stay as off until we deal with the deal with the dependency problem

[labath]

lldb/cmake/modules/LLDBConfig.cmake

49

Umm... I guess this wanted to be LLDB_ENABLE_LUA ?