https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99602
--- Comment #10 from Jürgen Reuter <juergen.reuter at desy dot de> --- (In reply to Jürgen Reuter from comment #9) > (In reply to Paul Thomas from comment #8) > > > > > Paul > > $ gfortran -fcheck=pointer repro.f90 > reuter@Manwe:~/local/packages/whizard/trunk/_build_flags/RT_20210315$ > ./a.out > At line 38 of file repro.f90 > Fortran runtime error: Pointer actual argument 'mm' is not associated > > Error termination. Backtrace: > #0 0x106ff6fde > #1 0x106ff7c85 > #2 0x106ff8256 > #3 0x106fe5a9d > #4 0x106fe5c8b > #5 0x106fe5cc6 So if I understand this correctly then gfortran does a false positive, it wrongly flags status of the pointer mm as non-associated and exits with a run-time error. gfortran before didn't do this, and also nagfor with -C=all doesn't flag this.
