(j3.2006) MPI usage problems in Fortran

[Jim Xia]

j3-bounces at j3-fortran.org wrote on 03/19/2008 03:18:47 PM:

> > The user
> > is responsible for knowing within which program units the variable is
> > subject to volatile changes, and give the variable the volatile
> > attribute in those units.
> Unless it is a dummy argument, in which case the standard makes no 
sense. Yet 
> alone a VALUE dummy as Van said. If a dummy is volatile, it is volatile 
> inside the routine. Period. There is nothing special about it. It should 
not 
> even require an explicit rules. It should not in any way affect or have 
> anything to do with the actual (which can have its own volatile 
attribute if 
> needed).
> 
> 
> Best,
> Aleks

Just to comment on VOLATILE on VALUE: for structure with pointer 
components, the VOLATILE on VALUE is harmful in that the pointer 
component's values can be altered and therefore affects the actual.  The 
assumption that VALUE is completely separated from the actual-arg is 
illusive.

Cheers,

Jim Xia

RL Fortran Compiler Test
IBM Toronto Lab at 8200 Warden Ave, Markham, On, L6G 1C7
Phone (905) 413-3444  Tie-line 313-3444
email: jimxia at ca.ibm.com
D2/YF7/8200 /MKM
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://j3-fortran.org/pipermail/j3/attachments/20080319/418e1370/attachment-0001.html