I’m not sure where to look or if it’s possible, if someone could point me in the right direction that would be great
There’s a few internal functions in my language like WriteLine. When I use lldb to ‘thread step-in’ (or vscode) the debugger will enter the function. Is it possible to let lldb know it should skip the function? I tried using clang to find some clues but puts and other function all allow me to step in and gets a missing source error.
Is there any llvm-ir code I can look at to show me how to skip a function? Or a language that generates the appropriate ir?
I don’t know of any DWARF functionality or extensions that are currently used to communicate “uninteresting” functions not to step into.
If there’s a specific example where you’ve seen this behavior work as you’d like in another language/frontend/etc - if you provide that example I/we might be able to explain why it works that way & how you could reproduce that behavior.
The first two options that come to mind are:
1. don't generate debug info for these functions
Many debuggers will avoid stepping into functions without debug info.
2. mark the function as as DW_AT_artifical to communicate to your debugger to skip it. I don't know if LLDB does this automatically, but you could easily make that the behavior for your language. For details it's best to ask on lldb-dev.
-- adrian