> It's pretty hard to debug those IR optimization loops.
> For now you can workaround them by disabling the OpenBSD4.x <fenv.h>
> workarounds in backend/evalu8.c but we should actually fix that some
> transformations loop if evalu8 fails.

Thanks for taking the time to found the problem! I'm going to try to
solve other challenges %)

_______________________________________________
dmd-internals mailing list
[email protected]
http://lists.puremagic.com/mailman/listinfo/dmd-internals

Reply via email to