(j3.2006) exception message on termination

[Bill Long]

I believe it was an oversight that these requirements were not better 
integrated when parallel execution was added.  We  have a way to 
suppress the "FORTRAN STOP" message for a STOP statement with no stop 
code, since users were understandably unhappy with seeing 80,000 lines 
of this output at the end of a job with 80,000 images.  Similarly the 
"that image" text below could result in huge amounts of output about 
exceptions that are usually uninteresting. Fortunately, for most users 
of Linux, where disabling all fp traps is the default, this is mostly a 
non-issue, since they very rarely change the default. And for codes 
running on GPU's it is irrelevant since fp traps are not supported at 
all.   But if the current requirement started having the actual effect 
of generating output, the complaints would be deafening.  At a minimum, 
the recommendation needs to be modified to allow for a summary, such as 
"1453 images terminated reporting a DIVIDE_BY_ZERO exception, ...", to 
limit output explosion.

Perhaps instead of having the current requirement, we could add a 
IEEE_PRINT_EXCEPTION_SUMMARY intrinsic subroutine that the user could 
call if they wanted this sort of output, and disconnect the output from 
STOP.

Cheers,
Bill


On 6/11/13 10:59 PM, Robert Corbett wrote:
> Can we delete the second and third sentences of paragraph 2 in Clause 8.4 in the
> upcoming revision?  They state
>
>        If an exception (14) is signaling on that image,
>        the processor shall issue a warning indicating
>        which exceptions are signaling; this warning
>        shall be on the unit identified by the named
>        constant ERROR_UNIT (13.8.2.8).  It is recommended
>        that the stop code is made available by formatted
>        output to the same unit.
>
> Sun FORTRAN 77 produced such output.  It was a constant source of complaints
> from users.  As bad as Sun FORTRAN 77 was in this regard, it did not report
> inexact exceptions, as is required by the text of the standard.
>
> Robert Corbett
> _______________________________________________
> J3 mailing list
> J3 at mailman.j3-fortran.org
> http://mailman.j3-fortran.org/mailman/listinfo/j3
>

-- 
Bill Long                                           longb at cray.com
Fortran Technical Support    &                 voice: 651-605-9024
Bioinformatics Software Development            fax:   651-605-9142
Cray Inc./Cray Plaza, Suite 210/380 Jackson St./St. Paul, MN 55101