On 23 June 2015 at 16:47, Charlie Turner <charlesturner...@gmail.com> wrote: > I would probably have to create a new suite of unit tests, ensuring > that for each ISO mode, only the correct set of library functions are > marked as isLibFunction(). I don't see any machinery in Clang that can > verify such things though, the builtins are lumped together.
I don't think this is worth it. > And you lose the contextual advantage distinct commits as well. I > think I'd prefer to just be more explicit about relationships between > my Phab reviews. In hindsight it was a mistake of mine not to market > this patch as a distinct entity. Can we pretend that I did? :) Yeah, Phab is not great in that sense. > // F -> this is a libc/libm function with a '__builtin_' prefix added. Of course! I read it in the first time I reviewed the patch, and now completely forgot. > I still feel like this patch is just cleaning up some omissions in the > C99 floating-point functions. I agree. I think we should just mark them all as F. cheers, --renato _______________________________________________ cfe-commits mailing list cfe-commits@cs.uiuc.edu http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits