Type, allocation of, and assignment to deferred-length character arrays

[KazuFurutaka]

@urbanjost , thanks for your comments.

@urbanjost

So the reproducer program has no problem accept it does not produce what you expect in the debugger, or is the reproducer failing in some way? Because you were using the debugger and saw the type not returning you may have identified a bug in the debugger/compiler interface and
not the cause of your program bug if that is the case.

No problem in the reproducer program, except that ptypeing in the debugger shows nothing (type = Type and then End Type).

@urbanjost

At the moment I do not have gdb available but I will later but in the meantime does this
replacement for the main program print the expected results, and if you make AHO() a contained
procedure does anything improve?

This replacement, with the aho() function separated into independent object and the original arraymod.o, shows the expected result, though the ptype of chararr() is empty! This probably indicates that there’s an issue in the debugger/compiler interface, right? (though it may not lead to resolve the issue of substitution into deferred-length character arrays in the old program).

Ah, I forgot to mention the compiler option; nothing for the reproducer, and -O -I$(INCDIR) for the original (and -g for debugging).

Kazuyoshi