Dear Balaji,
Thanks for your mail..I am sorry about my rather vague mails earlier
as I couldnt run totalview earlier (Thangam got an evaluation licence
for 15 days for totalview which let me run the debugger to pinpoint the
problem. We will process a purchase order for totalview but it may take
a few months before we get it. Before this I tried cvd and it took me
quite a while to figure out how to attach this to an mpi job. But it
didnt bring up the source window properly-I had done make localize- and
I had to use the traceback to find out the line). I have marked my
response to your suggestions below:
On Tue, 27 May 2003, V Balaji wrote:
> Hi Swathi: you may have got the 7.4 compiler along with 6.5.19. On the
> 7.4 compiler, the istat argument is no longer optional. Try using an
> earlier version of the compiler to see if the problem goes away.
>
When I run versions command I get
------------------------------------------------------------------------
I compiler_eoe 04/07/2003 IRIX Standard Execution Environment
(Base
Compiler Headers and Libraries, 7.4)
I compiler_dev 11/22/2001 Base Compiler Development Environment,
7.3.1.2m
I ftn90_fe 11/22/2001 Fortran 90 Front-end, 7.3.1.2m
-----------------------------------------------------------------------
Looks like I have a mix of 7.3.1.2m and 7.4 here. When we tried to
downgrade, there were so many conflicts we gave up.
> Failing which, correcting the FLUSH calls in mpp_io to add the second
> argument is the correct option. The next release of FMS will have this
> fix (which also works with older revisions of the compiler).
>
I am going to try this in a while after I "grep flush" in the full code.
I really dont like touching your code but I guess I have no other choice
at the moment. When is your next FMS public release scheduled?
Thanks for all your help.
Swathi
> P S Swathi writes:
>
> > Hello,
> >
> > The man pages (on IRIX 6.5.16m) of flush says the way to call flush is
> >
> > IRIX systems:
> > CALL FLUSH(iunit, istat)
> >
> > The istat argument is optional on UNICOS systems; the user program is
> > aborted if FLUSH is called without istat and an error condition is
> > encountered.
> >
> > But it used to work well without istat on IRIX until 6.5.16m. The mpp.F90
> > code does not have istat in any of the flush calls. I do not want to touch
> > Balaji's code in anyway. If this is indeed the problem with 6.5.19f, I
> > hope someone can give me a workaround without touching the code.
> >
> > Thanks
> >
> > Swathi
> >
>
> --
>
> Balaji 1 609.452.6516
> SGI/GFDL Princeton University
>
>
-------------------------------------------------------
This SF.net email is sponsored by: ObjectStore.
If flattening out C++ or Java code to make your application fit in a
relational database is painful, don't do it! Check out ObjectStore.
Now part of Progress Software. http://www.objectstore.net/sourceforge
_______________________________________________
FMS-mom4 mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/fms-mom4
