(j3.2006) (SC22WG5.4900) [ukfortran] Comment on a comment on the WG5 letterballot on N1947

[N.M. Maclaren]

On Jan 16 2013, Bill Long wrote:
>
>In recent memory, I've only encountered one case of harmful integer 
>overflow, and that occurred in converting a large real value to integer 
>and was caught by the IEEE hardware traps.  Of course, there are 
>numerous cases, almost all related to user-written random number 
>generators, where overflow of integer multiply is intentional.  For 
>those cases, enabling a check would cause the program to stop working.

Do none of the codes you see ever include fragments like the following?

    SUBROUTINE weeble (arg)
    REAL(KIND=KIND(0.0D0)) :: arg(:,:)
    REAL(KIND=KIND(0.0D0)) :: scratch(UBOUND(arg,1)*(UBOUND(arg,1)+1)/2)

Being caught out by default integers being large enough for any single
dimension but not the total size of an array, and accidentally using a
calculation that creates the latter, is a very common mistake in all of
Fortran, C and C++.


Regards,
Nick Maclaren.