vasp.5.2.12 installation error: metagga.o

Message

poorvasingh · #1 Post by **poorvasingh** » Thu Sep 08, 2011 8:43 am

We are getting following error during compilation of vasp.5.2.12:

fortcom: Error: metagga.f90, line 3051: Error in opening the compiled module file. Check INCLUDE paths. [PAW]
USE paw
----------^
compilation aborted for metagga.f90 (code 1)
make: *** [metagga.o] Error 1

Please suggest how to rectify this error.
<span class='smallblacktext'>[ Edited ]</span>

#2 Post by **admin** » Fri Sep 16, 2011 4:44 pm

please make sure that in the makefile you use paw_base is compiler BEFORE metagga, then paw should be available for the compilation of metagga

poorvasingh · #3 Post by **poorvasingh** » Tue Sep 20, 2011 4:09 pm

Thanks..makefile was incorrect..now it compiles correctly..

dagh · #4 Post by **dagh** » Tue Oct 18, 2011 7:12 pm

Hello!!

Thanks in advance for any reply...

We are getting problems with the compilation of VASP.5.2.12 in a SGI Altix machine (we have vasp 4.6 and 5.2* version working).

After solving similar problems with the makefile, as the reported in this thread, we are getting the following error in the compilation of the nmr.F module:

ifort -FR -cm -w95 -lowercase -w -safe_cray_ptr -fno-alias -assume byterecl -O3 -fno-alias -ip -ivdep_parallel -g -I/opt/sgi/mpt/mpt-1.25/include -c nmr.f90
fortcom: Error: nmr.F, line 4540: If the actual argument is scalar, the corresponding dummy argument shall be scalar unless the actual argument is an element of an array that is not an assumed-shape or pointer array, or a substring of such an element. [AVTOT_]
CALL SETDIJ_AVEC_(WDES, GRIDC_,GRIDUS,C_TO_US,LATT_CUR,P,T_INFO, LOVERL, &
--------------^
compilation aborted for nmr.f90 (code 1)
make: *** [nmr.o] Error 1

We are using the intel compilers (ifort, icc) and cmkl libraries (BLAS, LAPACK, SCALAPACK) and the previous modules seems to compile without problems.

Best regards.

dagh · #5 Post by **dagh** » Fri Oct 21, 2011 7:22 pm

Hello again!

It is just to say that we have got to solve our problem with the vasp.5.2.12 compilation in a SGI Altix machine:

compilation aborted for nmr.f90 (code 1)
make: *** [nmr.o] Error 1

It was solved by removing some flags for the intel fortran compiler in the makefile (we don't know which was exactly the source of the error, but we are working on). If somebody is interested in details, please send an email, it will be a pleasure to share ideas..

Regards..