da-viper wrote: > I looked at the code some more and it seems like we return the address of the > symbol if there is one, and only use the actual address (i.e. the pc) when > there is none. I was trying to find the code that computes the line in that > case (i.e. `pc - symbol`) and I couldn't immediately find it.
we use the address to create a source reference. https://github.com/llvm/llvm-project/blob/8b79cd751c70d3e8aec78df0cc39e9779493adf1/lldb/tools/lldb-dap/DAP.cpp#L513-L530 > I think this PR is fine in the sense that it improves the status quo, but we > probably want to take a more wholistic approach at this problem at some > point. For example, if we don't have a symbol, it would be nice to have a > source reference for a given range (say the next 64 or 128 instructions) and > then as we step keep using that sourceReference but bump the "relative line > number" in that artificial source file. Yeah makes more sense, https://github.com/llvm/llvm-project/pull/144364 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
