[RFC] Native Coarrays (finally!) [Review part 2]

Nicolas König koenigni@student.ethz.ch
Tue Oct 13 12:42:33 GMT 2020


Hi Andre,

Thanks for the reviews! I'll incorporate the comments this weekend (I'll 
also implement stat & errmsg then).

Sorry about all of these #if 0, I completely forgot they were still in 
there :)

Kind regards

   Nicolas König

On 12/10/2020 15:48, Andre Vehreschild wrote:
> Hi Nicolas,
> 
> here is part two of the review of the compiler components. I will do the
> testsuite and library parts another day, because now I am already completely
> bonkers. (Yes, I know, that's normal for me :-)
> 
> Regards,
> 	Andre
> 
> On Mon, 5 Oct 2020 15:23:27 +0200
> Nicolas König <koenigni@student.ethz.ch> wrote:
> 
>> Hello Tobias,
>>
>> On 05/10/2020 11:54, Tobias Burnus wrote:
>>> Hi Nicolas,
>>>
>>> admittedly, I have not yet looked at your patch. However, I have to
>>> admit that I do not like the name. I understand that "native" refers
>>> to not needing an external library (libcaf.../libopencoarray...),
>>> but I still wonder whether something like "-fcoarray=shared" (i.e.
>>> working on a shared-memory system) would be better name from an end-user
>>> point of view.
>>
>> I think the name has been the most critized point of the entire patch up
>> till now. I'm going to change it to -fcoarray=shared, as you (and a few
>> other people) suggested :)
>>
>>>
>>> Tobias,
>>> who likes that coarray can be used without extra libs and thinks
>>> that this will help with users starting to use coarrays.
>>
>> That is the main reason I wrote the patch.
>>
>>>
>>> -----------------
>>> Mentor Graphics (Deutschland) GmbH, Arnulfstraße 201, 80634 München /
>>> Germany
>>> Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung,
>>> Alexander Walter
> 
> 


More information about the Fortran mailing list