Re: -frecord-marker

[Brooks Moses <bmoses at stanford dot edu>]

Steve Kargl wrote:
> Put these 2 subroutines into separate files.
>
>    subroutine foo1  ! Compile with 32-bit marker
>      open(unit=1, file='messy', status='append', ....)
>      ! Write a bunch of stuff
>      close(1)
>    end subroutine foo1
>
>    subroutine foo2  ! Compile with 64-bit marker
>      open(unit=1, file='messy', status='append', ....)
>      ! Write a bunch of stuff
>      close(1)
>    end subroutine foo2
>
> What happens when foo2 tries to write to 'messy' after foo1
> has already done so?  Do we reread the entire file and convert
> to 64-bit markers?  Do we issue a runtime error and abort?
> Do we mingle the 32-bit and 64-bit markers in the same file?

What currently happens if gfortran opens (as append) a file that has 
been written by an Intel-compiled program?  Or, for that matter, what 
happens if the two routines are in programs separately compiled in 
gfortran -- which is already allowed?

I should think that all of these conditions should be treated in exactly 
the same way.  It should not make any difference to foo2 whether the 
"messy" file was created by a different routine in the same program, or 
a separate program also compiled by gfortran, or a different program 
compiled by a different compiler.  In fact, I can't see how it _would_ 
make a difference.

Similarly, I can't see how it should make a difference to any of this 
whether or not the routine in question is compiled with gfortran's 
default record-marker format or with a different format.

- Brooks