Hi Tiziano,<br><br>  Thank you for your answer. <br>  The compilation of the small test indeed fails with the PGI compiler 20.9. And it gives me the exact same error that I obtained when trying to compile CP2K directly. So I will not continue in this direction. <br><br>Best regards,<br>Conrad. <br><br><div class="gmail_quote"><div dir="auto" class="gmail_attr">On Friday, October 9, 2020 at 10:09:26 AM UTC+2 tiz...@chem.uzh.ch wrote:<br/></div><blockquote class="gmail_quote" style="margin: 0 0 0 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">Hi Conrad,
<br>
<br>On 10/9/20 9:33 AM, Conrad Hillairet wrote:
<br>> 
<br>> Good morning,
<br>> 
<br>>    I am facing some issues at compile time with the PGI compiler 20.9. I 
<br>> have a general question : is it possible to use the PGI compiler to 
<br>> compile CP2K ?
<br>
<br>I don't know about PGI 20.9, the last one I tested was 20.4: here F95(!) 
<br>support was still broken/incomplete.
<br>
<br>Try to build the following:
<br>
<br><a href="https://github.com/cp2k/dbcsr/blob/develop/cmake/compiler-tests/f95-reshape-order-allocatable.f90" target="_blank" rel="nofollow" data-saferedirecturl="https://www.google.com/url?hl=en&q=https://github.com/cp2k/dbcsr/blob/develop/cmake/compiler-tests/f95-reshape-order-allocatable.f90&source=gmail&ust=1602319469441000&usg=AFQjCNGmtbpe3ViDEs5FtWMgxmxg3qeKTQ">https://github.com/cp2k/dbcsr/blob/develop/cmake/compiler-tests/f95-reshape-order-allocatable.f90</a>
<br>
<br>.. or run CMake for DBCSR and it will tell you directly:
<br>
<br>   git clone <a href="https://github.com/cp2k/dbcsr.git" target="_blank" rel="nofollow" data-saferedirecturl="https://www.google.com/url?hl=en&q=https://github.com/cp2k/dbcsr.git&source=gmail&ust=1602319469441000&usg=AFQjCNE9KZ9alO7sDg9PAvLqG2yb85T_2Q">https://github.com/cp2k/dbcsr.git</a>
<br>   mkdir dbcsr.build
<br>   cd dbcsr.build
<br>   CC=pgcc CXX=pgc++ FC=pgfortran cmake -DUSE_MPI=OFF ../dbcsr
<br>
<br>If that succeeds but compilation fails later on, I will test again and 
<br>see whether I can come up with another test case.
<br>
<br>> From the CP2K webpage dedicated to compiler support I get 
<br>> the feeling that GNU and Intel are usually the common choice (on x86) 
<br>> but it is still unclear to me whether it should compile with PGI and 
<br>> which compiler is usually recommended.
<br>
<br>GNU and Intel (check versions on [1]) should work fine on x86, you can 
<br>also look at the dashboard [2] to check for system+compiler combinations.
<br>
<br>No, PGI is not yet supported. If you have a contact at PGI/NVIDIA, 
<br>please also report the issues directly with them.
<br>
<br>Best regards,
<br>Tiziano
<br>
<br>[1] <a href="https://www.cp2k.org/dev:compiler_support" target="_blank" rel="nofollow" data-saferedirecturl="https://www.google.com/url?hl=en&q=https://www.cp2k.org/dev:compiler_support&source=gmail&ust=1602319469442000&usg=AFQjCNFTUb1J9FN_bpiTpscUv743mOBwjQ">https://www.cp2k.org/dev:compiler_support</a>
<br>[2] <a href="https://dashboard.cp2k.org/" target="_blank" rel="nofollow" data-saferedirecturl="https://www.google.com/url?hl=en&q=https://dashboard.cp2k.org/&source=gmail&ust=1602319469442000&usg=AFQjCNHpF540Q_D8Rac5pm0IQ5iNxEAVPw">https://dashboard.cp2k.org/</a>
<br>
<br>-- 
<br>Tiziano Müller
<br>University of Zurich
<br>Department of Chemistry
<br>Winterthurerstrasse 190
<br>CH-8057 Zürich
<br>
<br>Tel: +41 44 63 54234
<br><a href="http://www.chem.uzh.ch" target="_blank" rel="nofollow" data-saferedirecturl="https://www.google.com/url?hl=en&q=http://www.chem.uzh.ch&source=gmail&ust=1602319469442000&usg=AFQjCNHZFGj9owdPcVdUHQIwRtR1l2wQSA">www.chem.uzh.ch</a>
<br><a href data-email-masked rel="nofollow">tiz...@chem.uzh.ch</a>
<br></blockquote></div>