[Bug tree-optimization/78180] Poor optimization of std::array on gcc 4.8/5.4/6.2 as compared to simple raw array

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78180

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|middle-end                  |tree-optimization

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
r.idx_ is not optimized away inside the loop for some reason.

NonWorking case:

  _53 = r.idx_;
  _54 = (long unsigned int) _53;
  MEM[(value_type &)&r][_54] = _2;


Working case:

  _52 = r.idx_;
  r.times_[_52] = _2;


std::array operator[]:
  _5 = &MEM[(const long int[128] &)this_2(D)][__n_4(D)];

----------

      reference
      operator[](size_type __n) noexcept
      { return _AT_Type::_S_ref(_M_elems, __n); }




      static constexpr _Tp&
      _S_ref(const _Type& __t, std::size_t __n) noexcept
      { return const_cast<_Tp&>(__t[__n]); }



So SRA is not understanding the array being the first field .....