Hi, After spending countless hours between Ivan and myself fighting GCC's mysteriously-vanishing <type_traits> support for __int128_t (= qint128), it turns out that with -ansi/-std=c++NN, not even the most basic of <type_traits> work: std::is_signed_v<qint128> is _false_! I repeat: _FALSE_.
The following is /my/ professional opinion, Ivan may chime in himself: I think we need to mandate that if you want qint128 support, then you must compile with gnu++NN, which is actually the default on both GCC and Clang. We seem to switch that off (-ansi on). We cannot fight a broken stdlib. If the compiler says that these things don't exist as integral types, then who is Qt to disagree? It slows down development, and all those work-arounds slow down compilations, too: - std::is_integral_v<T> + std::disjunction_v<std::is_integral<T>, + std::is_same<T, qint128>, + std::is_same<T, quint128>> Unless someone can present a solution where <type_traits> and <limits> work for q(u)int128 under -ansi — even for users — I would undef QT_SUPPORTS_INT128 when __STRICT_ANSI__ is in effect, then drop the -ansi from Qt builds to make sure the symbols that contain qint128 are present in the library. User projects can then use whichever one they want. We have a similar issue with qfloat16::NativeFloat. On GCC-13, the traits only work (tested: <limits>) when compiling in C++23 mode, yet we insist on using the types before, eagerly grabbing _some_ definition from wherever we can (C20, platform-specifics, ...). And yes, that means we'll have BC issues from that, too. References: - https://bugreports.qt.io/browse/QTBUG-119901 - https://codereview.qt-project.org/c/qt/qtbase/+/478199/comment/bcbe8aca_d5d65018/ - https://codereview.qt-project.org/c/qt/qtbase/+/507383 Opinions? Comments? Thanks, Marc -- Marc Mutz <marc.m...@qt.io> Principal Software Engineer The Qt Company Erich-Thilo-Str. 10 12489 Berlin, Germany www.qt.io Geschäftsführer: Mika Pälsi, Juha Varelius, Jouni Lintunen Sitz der Gesellschaft: Berlin, Registergericht: Amtsgericht Charlottenburg, HRB 144331 B -- Development mailing list Development@qt-project.org https://lists.qt-project.org/listinfo/development
