https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117394
--- Comment #4 from Jonathan Wakely <redi at gcc dot gnu.org> --- (In reply to Nadav Har'El from comment #3) > I think this issue is more accurately a duplicate of > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86164. Which is the same problem as Bug 61601. > That bug is very sad, it has been open for seven years now and has led to a > mass exodus from std::regex to boost::regex which doesn't have this bug - > because it's a serious bug with serious symptoms (crash) that applications > have no way to avoid: There are other reasons to prefer boost::regex, so maybe it's good that people are using a better implementation. Our std::regex is unlikely to ever be fast or efficient, due to ABI compatibility reasons. Even if/when Bug 61601 is fixed, it will still be slow. > Issue 61601 was a less serious issue because it was about a very complex > regular expression, which users can simply avoid. But issue 86164 and the > one reported here is much more serious: Here the regular expression is > simple - but the user input is just longer, something which is hard for > applications to avoid. Even worse - the failure mode isn't some sort of > clean "error", it is an outright crash of the application. It's still the same underlying bug though.
