On quinta-feira, 19 de janeiro de 2017 00:40:22 PST Prav wrote: > Hello, Thiago. > > > -no-rtti is not tested and clearly someone broke it. I've submitted a > > change to remove that option from configure starting with Qt 5.8.1. > > I was thinking that without RTTI compiled programs became lighter and > faster (but never saw values of how much)
Yes, the option causes the compiler not to emit the RTTI structures for polymorphic classes. That may save you a bit of read-only, sharable data in the Qt libraries. > As I can see Qt tries to be > lighter now to be used in embedded systems. You still think that total > remove of -no-rtti flag is OK in this case? The option only did anything on Windows, where a couple hundred bytes per library make much of a difference. It would make more sense on embedded systems, but it seems like it wasn't active for them. I don't know why. My guess is that it use to work for Qt for Embedded Linux back in Qt 4 days, but with the conversion to Qt 5 and QPA, the option was lost. And no one complained. > And why Qt now needs RTTI? Qt 5.7.0 was building fine without RTTI. What > features added to 5.8.0 made RTTI mandatory I wonder? Someone added a dynamic_cast. That's all. -- Thiago Macieira - thiago.macieira (AT) intel.com Software Architect - Intel Open Source Technology Center _______________________________________________ Interest mailing list Interest@qt-project.org http://lists.qt-project.org/mailman/listinfo/interest
