Ah, so the cost isn't necessarily the same. I don't think this is worth it then. (Although, paradoxically, it opens up an avenue for testing, via a custom allocator which counts </!= comparisons.)
Any other ideas for improving the failure mode? > On 2016-Feb-09, at 16:14, Marshall Clow <mclow.li...@gmail.com> wrote: > > mclow.lists added a comment. > >> Since `__new_last` and `__end_` are raw pointers, < should be the same cost >> as !=. > > > `pointer` is an alias for `allocator_traits<_Alloc>::pointer`, which might > not be a raw pointer. > > > http://reviews.llvm.org/D17053 > > > _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
