[J3] Clarification on F18/017
Bill Long
longb at cray.com
Mon Jun 8 12:54:49 EDT 2020
> On Jun 8, 2020, at 11:31 AM, Ondřej Čertík via J3 <j3 at mailman.j3-fortran.org> wrote:
>
> Hi Malcolm,
>
> On Sat, Jun 6, 2020, at 2:29 PM, Ondřej Čertík via J3 wrote:
>> Hi Malcolm,
>>
>> On Fri, Jun 5, 2020, at 7:38 PM, Malcolm Cohen via J3 wrote:
>>> Hi folks,
>>>
>>>
>>> Okay, here’s the analysis of what happens with the example program,
>>> under the interp as it stands.
>>
>> Thank you very much for this detailed analysis. We really appreciate it.
>>
>> We'll discuss it, and if we have further questions or proposals, we
>> will post here.
>
> So we discussed it both inside LANL and outside, and at LANL we definitely do not want the CBPA order to be allowed, and it seems a lot of people in the wider community agree.
But not all.
>
> In your analysis you explained that due to automatic deallocation predating finalizers, it is natural for some compilers to allow CBPA.
Seemed like a reasonable explanation.
>
> Is there any other motivation to allow the CBPA order?
Some compilers might already produce that and it is not desirable to introduce incompatible changes.
>
> There do not seem to be any user codes that would depend on CBPA, simply because the CABP order has always been allowed and implemented by several compilers.
If “implemented by several compilers” were a criterion for standardization, then we would standardize REAL*8.
>
> From the user perspective, the CBPA order is unexpected and surprising.
Probably depends on the user. Users come to “expect” certain things based how their favorite compiler behaves. Rather than what the standard says.
Cheers,
Bill
> In our opinion based on the discussion so far, it should not be allowed by the Standard.
>
> Ondrej
Bill Long longb at cray.com
Principal Engineer, Fortran Technical Support & voice: 651-605-9024
Bioinformatics Software Development fax: 651-605-9143
Cray, a Hewlett Packard Enterprise company/ 2131 Lindau Lane/ Suite 1000/ Bloomington, MN 55425
More information about the J3
mailing list