[J3] Question about the maximum event count of EVENT_TYPE variables

[Milan Curcic]

Thanks, Malcolm.

I will open a separate thread with ideas on how this could be improved in
F202y.

Milan


On Mon, Jun 7, 2021, 1:04 AM Malcolm Cohen via J3 <j3 at mailman.j3-fortran.org>
wrote:

> Hi,
>
>
>
> Integer overflow makes the program non-conforming. The standard should not
> say more about event count in particular, unless we want to **require**
> that overflow raise an error condition. That would be a new feature for
> F202y.
>
>
>
> Cheers,
>
> --
>
> ..............Malcolm Cohen, NAG Oxford/Tokyo.
>
>
>
> *From:* J3 <j3-bounces at mailman.j3-fortran.org> *On Behalf Of *Milan
> Curcic via J3
> *Sent:* Monday, June 7, 2021 3:49 PM
> *To:* General J3 interest list <j3 at mailman.j3-fortran.org>
> *Cc:* Milan Curcic <caomaco at gmail.com>
> *Subject:* [J3] Question about the maximum event count of EVENT_TYPE
> variables
>
>
>
> Section 16.10.2.10, paragraph 2, line 16 of 18-007r1 states:
>
>
>
> > The event count is of type integer with kind ATOMIC_INT_KIND from the
> intrinsic module ISO_FORTRAN_ENV
>
>
>
> and the ATOMIC_INT_KIND is defined as (16.10.2.2):
>
>
>
> > The value of the default integer scalar constant ATOMIC_INT_KIND is the
> kind type parameter value of type integer variables for which the processor
> supports atomic operations specified by atomic subroutines.
>
>
>
> My interpretation of this is that EVENT_TYPE variables have a maximum
> event count that is processor-dependent. With GFortran and ifort,
> that's 2147483647.
>
>
>
> However, I can't find any text that describes what happens to the event
> count when it exceeds its maximum value. With GFortran+OpenCoarrays, it
> overflows (i.e. -2147483648 follows 2147483647).
>
>
>
> Is this behavior:
>
>
>
> * Processor-dependent?
>
> * Undefined?
>
> * Non-conforming?
>
>
>
> I think that the Standard should at least say if the behavior is one of
> the above, if not also what should happen to the event count value. Do you
> agree? If yes, does this call for an interp?
>
>
>
> Thanks,
>
> Milan
>
>
>
>
>
>
>
> *Disclaimer*
>
> The Numerical Algorithms Group Ltd is a company registered in England and
> Wales with company number 1249803. The registered office is: 30 St. Giles,
> Oxford, OX1 3LE, United Kingdom. Please see our Privacy Notice
> <https://www.nag.co.uk/content/privacy-notice> for information on how we
> process personal data and for details of how to stop or limit
> communications from us.
>
> This e-mail has been scanned for all viruses and malware, and may have
> been automatically archived by Mimecast Ltd, an innovator in Software as a
> Service (SaaS) for business.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.j3-fortran.org/pipermail/j3/attachments/20210608/94d02ec0/attachment.htm>