On 18 Mar 2014 07:30, "Dan Olson" <zans.is.for.c...@yahoo.com> wrote: > > Johannes Pfau <nos...@example.com> writes: > > > iOS: > > I don't know about gdc/iOS but there seems to be some work on this > > for LDC: > > http://forum.dlang.org/thread/m2txc2kqxv....@comcast.net > > I tried adding iOS support to GDC a year ago, but the arm--darwin target > is missing in gcc since 4.2. It was turning into a lot of work to add > arm-darwin back into 4.8. > > LDC has an advantage for targeting iOS since Apple uses and contributes > to LLVM. Much of D works today on iOS (armv7) except TLS and issues
GDC has tls emulation routines in its druntime. Once it is properly hooked into the GC (all attempts I've tried currently get deadlocked) then the LDC guys can reuse parts of it to work their backend. > with getting correct "real" type when cross-compiling, though Kai has a > solution for that. Really? That is a problem? I'm not familiar with LDC or LLVM but I would at least expect the backend be able to provide that information on command. Or is this the fact that they still use native reals for CTFE? If so, then you can look at longdouble implementation in either DMD MSVC or GDC for that.