Le 21 juin 2011 à 15:06, David Chisnall a écrit : > I'm having a bit of difficulty working out exactly what is the bit that is > wrong in your test case. Can you try to put together a reduced test case? > My suspicion is that it's somehow related to the handling of unions. > NSMapTable uses unions of the OpenStep-style pointer functions structure and > the OS X 10.5-style NSPointerFunctions object in various places, and there's > a lot of potential for stuff to go wrong there due to weirdness in the Linux > ABI for unions.
I'm still trying to create a reduced test case that doesn't depend on Foundation, but it's quite hard… If I don't use Foundation, a simple standalone class roughly similar to NSConcretePointerFunctions doesn't exhibit the same issue. If I use Foundation and tweak -initWithOptions: in the original class, the bug can disappear by rearranging the statements inside -initWithOptions: or even a adding new statement such as 'assert(options == NSPointerFunctionsOpaqueMemoryOption)'. For example, the arguments are not corrupted if I add this assertion, this is quite weird. So I probably have to create a class that very closely mimics NSConcretePointerFunctions without depending on Foundation… unless you have a better suggestion? Thanks, Quentin. _______________________________________________ Etoile-dev mailing list Etoile-dev@gna.org https://mail.gna.org/listinfo/etoile-dev
