mentlerd wrote: > A high-level comment: could we split up `FindLibCppStdFunctionCallableInfo`? > Exactly how I haven't fully thought through. But possibly a function for each > of the callable cases?
Sure! I think a natural split could be to make a version which returns a (possibly dummy) `ValueObjectSP` of wrapped callable in combination with a `CompilerType` of the function pointer type describing the kind of function this `std::function` implements. Other code then can either proceed to extracting the load address from the `ValueObjectSP` that we might land in when the `std::function` is called, or to just serve the `ValueObjectSP` as synthetic children. --- One open question that I have been thinking about is how to handle callable types with multiple `operator()(...)` function overloads. My thought process so far is that trying to figure out which overload `std::function` would end up calling is futile effort, instead we could try giving the thread plan all potential addresses. https://github.com/llvm/llvm-project/pull/111892 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
