[QE-users] [EXT] Re: QE-6.5-GPU and ENVIRON

[Michele Re Fiorentin]

Ok I had mistakenly removed also the line

USE environ_base, ONLY :                      &
>          e2_=> e2,                               &


which must be kept. Now the code compiles.
Sorry for the inconvenience.
Thank you again for this workarounìd, indeed I'm not planning to use cp.x.
Best,


M.


--
Michele Re Fiorentin, PhD

Istituto Italiano di Tecnologia (IIT)
Center for Sustainable Future Technologies @ PoliTO
via Livorno 60
10144 Torino (Italy)
tel: +39 0110904333
email:    michele.refiorentin at iit.it


Il giorno mer 24 giu 2020 alle ore 18:58 Michele Re Fiorentin <
michele.refiorentin at iit.it> ha scritto:

> Dear Oliviero,
>
> Thank you very much for your answer!!
> If I remove both e2 and e2_, the code successfully compiles. However, if I
> keep the line e2_ = 2.D0 the compiler rightfully complains that e2_ has not
> been explicitly declared. What declaration should be added?
> Thank you again,
>
>
> Michele
>
>
> --
> Michele Re Fiorentin, PhD
>
> Istituto Italiano di Tecnologia (IIT)
> Center for Sustainable Future Technologies @ PoliTO
> via Livorno 60
> 10144 Torino (Italy)
> tel: +39 0110904333
> email:    michele.refiorentin at iit.it
>
>
> Il giorno mer 24 giu 2020 alle ore 18:09 Andreussi, Oliviero <
> Oliviero.Andreussi at unt.edu> ha scritto:
>
>> Hi,
>>
>> Thank you for reporting the issue with the PGI compiler on Environ.
>>
>> I am really not sure why there should be an error for e2.  I don’t have
>> access to the PGI compiler and I am not sure I can debug this kind of
>> compilation error, I tried searching on google but with not great success
>> (and gfortran and intel fortran do not complain about this). I would try to
>> just change the INTENT(IN) to INTENT(INOUT), which should not be needed,
>> but may fix the compiler concerns.
>>
>> Note that if you only need to use pw.x and not cp.x (which is very often
>> the case, as I still have to find a good reason to use cp.x with Environ)
>> e2 is not really passed to environ_initbase(), so in fact e2 is not
>> present. I guess this may be the source of confusion of PGI, but the
>> argument is OPTIONAL and the compiler should know this. Anyway, another fix
>> is that you remove e2 from the arguments of the environ_initbase()
>> definition and only keep the line e2_ = 2.D0.
>>
>> If you find other instances or other problems, don’t hesitate to let me
>> know.
>>
>> Best,
>>
>> Oliviero
>>
>> On Jun 24, 2020, at 10:54 AM, Paolo Giannozzi <p.giannozzi at gmail.com>
>> wrote:
>>
>> Illegal attribute - intent specified for nondummy argumen
>>
>>
>> _______________________________________________
>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
>> users mailing list users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20200624/de50b586/attachment.html>