mclow.lists added inline comments.
================
Comment at: include/algorithm:4431
value_type* __p = __buff;
- for (_BidirectionalIterator __i = __first; __i != __middle;
__d.__incr((value_type*)0), (void) ++__i, ++__p)
+ for (_BidirectionalIterator __i = __first; __i != __middle;
__d.__incr((value_type*)nullptr), (void) ++__i, ++__p)
::new(__p) value_type(_VSTD::move(*__i));
----------------
I'm not really a fan of casting `nullptr` to a `value_type *`. This change
doesn't seem to add anything to the code; it's just longer.
The call to `__incr` doesn't use the pointer at all. Maybe a better approach is
to change
```
template <class _Tp>
_LIBCPP_INLINE_VISIBILITY void __incr(_Tp*) _NOEXCEPT
{__incr(integral_constant<bool,
is_trivially_destructible<_Tp>::value>());}
```
to:
```
template <class _Tp>
_LIBCPP_INLINE_VISIBILITY void __incr() _NOEXCEPT
{__incr(integral_constant<bool,
is_trivially_destructible<_Tp>::value>());}
```
and remove the `nullptr`s from the calls in `<algorithm>`
Alternately, add an overload of `__incr` that takes a `nullptr_t`.
Repository:
rCXX libc++
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D43159/new/
https://reviews.llvm.org/D43159
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
