At Tue, 15 Apr 2014 10:36:02 +0200, Pasquale Dir wrote: > > Working for the other corrections but I disegree with these ones: > > 1. This is an "unsigned long long" in C, so it must be a long, not > NativeLong here. > 2. flags is an "unsigned int" in C, so it > must be an "int" in Java
That is a hard requirement when wrapping native libraries with JNA, it is an error to do what you suggest. See http://twall.github.io/jna/4.1.0/overview-summary.html#marshalling > As in java types are signed I'd avoid to allocate the exact same > memory, on the contrary I would allocate more space so that an high > positive value for an unsigned C type would stay an high positive > value in java (who would make it a negative value if space is not > enough). Impossible, you have to adhere to libvirt's ABI. Regards, Claudio -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
