labath wrote: Yeah, getting some numbers would be nice. Maybe just repeating the benchmark in that diff. It might be a good idea to do it with and without the accelerator tables, just to rule out any effect from those.
FWIW, I expect this to not matter these days. It looks like this change (https://reviews.llvm.org/D31451) predates our use of the ItaniumPartialDemangler (https://reviews.llvm.org/D50071), which means that we were getting the name information by demangling the name and then parsing the result. Now, we should get this straight from the demangler, so the parsing code should be less hot. https://github.com/llvm/llvm-project/pull/137072 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
