https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125773
--- Comment #3 from Illia Trukhanov <lahvuun at gmail dot com> --- >This might be an issue because GCC column info is just stored post tabs. Are you sure about that? I'm not familiar with the GCC codebase, but I looked around a bit before reporting, and sarif_builder in gcc/diagnostics/sarif-sink.cc has a `m_tabstop` initialized to `dc.get_column_options ().m_tabstop`. If I initialize it to a literal 1, startColumn gets the expected value. It's also a private member, only used for calculating startColumn and endColumn, so it shouldn't affect anything else other than SARIF output.
