https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118120
--- Comment #5 from Slava Zakharin <szakharin at nvidia dot com> --- Right, this test deliberately introduces the aliasing to demonstrate that gfortran has too optimistic assumptions about aliasing of COMPLEX and REAL pointers. Thank you for the pointer to trans-array.cc! Might this particular statement be causing the lack of the temporary creation: https://github.com/gcc-mirror/gcc/blob/7eac34a6c793540606a50e20c4c56bd98476d3a1/gcc/fortran/trans-array.cc#L5347 ? It does not seem to be true for COMPLEX and REAL since F2008 introduced complex-part designators.
