[Pw_forum] problem compiling espresso-3.2 in SGI origin machine

Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu
Wed Jan 10 22:25:25 CET 2007


halim said wrote:
> Dear Axel,
> 
> Thank you for your reply and for the advices that you gave me, I would 
> like to say that I checked all what you said me,
> I checked Modules/make.depend
> contains the line:
> 
> bfgs_module.o : basic_algebra_routines.o  and it exists
> 
> But infortunately I don't have f95, I have just f90,  and the problem is 
> still the same, I could not run with success the code, is there please 
> other way in order to solve this problem,

the f95 suggestion was just a wild guess (it has worked for me on
an IBM AIX machine once).

for further debugging, i'd need access to such an SGI Origin
in order to narrow down the problem and ideally find a workaround
or a solution. to be frank, i don't think a machine like this
is worth the effort anymore.

one option you can try, is to install the g95 compiler from
http://ftp.g95.org, it is supposed to work with QE on x86-linux, but...

axel.


> 
> Thank you in advances for your help.
> 
> Regards,
> 
> Halim
> 
> */Axel Kohlmeyer <akohlmey at cmm.chem.upenn.edu>/* a écrit :
> 
>     On Tue, 9 Jan 2007, halim said wrote:
> 
>     HS> Dear Axel and Paolo,
>     HS>
>     HS> I am compiling espress-3.2 code in SGI origin machine, when I am
>     taping
> 
>     be prepared to suffer. SGI origins have inflicted a lot of
>     pain on me at times. good thing, they are more or less
>     obsolete these days...
> 
>     HS> ./configure
>     HS>
>     HS> every thing is fine.
> 
>     configure does next to no consistency checking on the fortran
>     side, so a completed configure does not have much meaning
>     except that it is required. simply using and adapting
>     an existing make.sys template is no longer sufficient.
> 
>     HS> but when I am taping gmake all,
>     HS>
>     HS> there are some errors appear, how to deal with this problem in order
>     HS> to success for compiling the code.
>     HS>
>     HS> Below I am attaching configure and the errors which appeared
>     after gmake
>     HS>
>     HS> Your help and support are highly appreciate.
>     HS>
>     HS> Regards,
>     HS>
>     HS> Halim
>     HS>
>     HS>
> 
>     [...]
> 
>     HS> --------------------------------------------------------------------
>     HS>
>     HS> The following libraries have been found:
>     HS> BLAS_LIBS=-lcomplib.sgimath
>     HS> LAPACK_LIBS=../flib/lapack.a
>     HS> FFT_LIBS=
>     HS> MPI_LIBS=-lmpi
>     HS> Parallel environment detected successfully.
>     HS> Configured for compilation of parallel executables.
>     HS> configure: success
> 
>     ok. looking normal up to here...
> 
>     HS> espresso.3-2> gmake all
>     HS> gmake[1]: Leaving directory `/h2/home/espresso-3.2/iotk'
>     HS> ( cd Modules ; if test "gmake" = "" ; then make TLDEPS= all ; \
>     HS> else gmake TLDEPS= all ; fi )
>     HS> gmake[1]: Entering directory `/h2/home/espresso-3.2/Modules'
>     HS> f90 -mips4 -64 -O2 -r10000 -r8 -ftpp -macro_expand -D__COMPLIB
>     -D__ORIGIN -D__FFTW -D__USE_INTERNAL_FFTW -D__MPI -D__PARA
>     -I../include -I./ -I../Modules -I../iotk/src -I../PW -I../PH
>     -I../CPV -c io_global.f90
> 
> 
>     HS> This product's license is for 2 CPU's; Your system has 7 CPU's
> 
> 
>     hmmmmmm.... looks like you need to pay SGI some more money. ;-)
>     but that should apply only at run-time.
> 
>     HS>
>     HS> MODULE bfgs_module
>     HS> ^
>     HS> f90-855 f90: ERROR BFGS_MODULE, File = bfgs_module.f90, Line =
>     9, Column = 8
>     HS> The compiler has detected errors in module "BFGS_MODULE". No
>     module information file will be created for this module.
>     HS>
>     HS> IF ( ( grad(:) .dot. step(:) ) > 0.D0 ) THEN
>     HS> ^
>     HS> f90-318 f90: ERROR GDIIS_STEP, File = bfgs_module.f90, Line =
>     464, Column = 26
>     HS> The operator "dot" has not been declared as a user defined operator.
> 
>     ok. this means that either the sgi compiler is not sufficiently
>     fortran 95 compliant or, that the prerequisite module
>     basic_algebra_routines.f90 has not been compiled correctly.
> 
>     please do a 'gmake clean ; gmake all' to make sure that you
>     are not continuing from a previous broken compile.
>     please also make sure that Modules/make.depend
>     contains the line:
> 
>     bfgs_module.o : basic_algebra_routines.o
> 
>     and if that does not help, i'd try to change
> 
>     F90 = f90
>     MPIF90 = f90
>     LD = f90
> 
>     to
> 
>     F90 = f95
>     MPIF90 = f95
>     LD = f95
> 
>     in make.sys
> 
> 
>     axel.
> 
>     HS>
>     HS> step(:) = - ( inv_hess(:,:) .times. grad(:) )
>     HS> ^
>     HS> f90-318 f90: ERROR GDIIS_STEP, File = bfgs_module.f90, Line =
>     469, Column = 42
>     HS>
>     HS> f90: MIPSpro Fortran 90 Version 7.3 (f52) Mon Jan 8, 2007 11:53:12
>     HS> f90: 1008 source lines
>     HS> f90: 3 Error(s), 0 Warning(s), 0 Other message(s), 0 ANSI(s)
>     HS> cf90: "explain cf90-message number" gives more information about
>     each message
>     HS> gmake[1]: *** [bfgs_module.o] Error 2
>     HS> gmake[1]: Leaving directory `/h2/home/espresso-3.2/Modules'
>     HS> gmake: *** [mods] Error 2
>     HS>
>     HS>
>     HS>
>     HS> __________________________________________________
>     HS> Do You Yahoo!?
>     HS> En finir avec le spam? Yahoo! Mail vous offre la meilleure
>     protection possible contre les messages non sollicités
>     HS> http://mail.yahoo.fr Yahoo! Mail
> 
>     -- 
>     =======================================================================
>     Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu
>     Center for Molecular Modeling -- University of Pennsylvania
>     Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
>     tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
>     =======================================================================
>     If you make something idiot-proof, the universe creates a better idiot.
> 
>     _______________________________________________
>     Pw_forum mailing list
>     Pw_forum at pwscf.org
>     http://www.democritos.it/mailman/listinfo/pw_forum
> 
> 
> __________________________________________________
> Do You Yahoo!?
> En finir avec le spam? Yahoo! Mail vous offre la meilleure protection 
> possible contre les messages non sollicités
> http://mail.yahoo.fr Yahoo! Mail
> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.



More information about the Pw_forum mailing list