Mpi installation and fortran compiler mismatch

[zuohen... at gmail.com]

OK, Thank you very much for your provided information. 
The compilation runs correctly now. I think it is just a warning.

Best,
Zuoheng

在 2018年8月29日星期三 UTC+2下午8:32:49，Travis写道：
>
> The format of module files produced by different Fortran compilers changes 
> pretty rapidly and the MPI .mod files may be unreadable by the version of 
> Fortran you are now using to compile CP2K. As the warning message implies, 
> you need to add '-D__HAS_NO_MPI_MOD' to your arch file along with your 
> other DFLAGS to resolve the conflict.
>
> More on that here, https://www.tcm.phy.cam.ac.uk/sw/gfortran.html
>
> I think Alfio is correct that you can ignore it if the compilation runs 
> correctly otherwise and generates a functioning executable. However, if you 
> see this warning and your compilation is breaking... it may be something to 
> consider.
>
>
> On Tuesday, August 28, 2018 at 6:40:58 AM UTC-4, zuoh... at gmail.com 
> wrote:
>
>> Hello everyone,
>>
>> When I compile my CP2K, I have the following Warning(or error?):
>>
>> USE mpi  ! compiler *errors* mean mpi installation and fortran compiler 
>> mismatch: see INSTALL (-D__HAS_NO_MPI_MOD)
>> Warning: USE statement at (1) has no ONLY qualifier [-Wuse-without-only]
>>
>> However, I could finish the compile and can use cp2k.popt or cp2k.psmp to 
>> do simulations. I am not sure if it will have some negtive effect on the 
>> speed of simulation.
>> I installed some dependence with 
>> tools/toolchain/install_cp2k_toolchain.sh and all is OK. I don't know why 
>> the mpi and fortran are mismatched.
>>
>> Thank you very much.
>>
>> Best wishes,
>>
>> Zuoheng
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20180830/9927ec28/attachment.htm>