[CPMD-list] Problems with Cray X1 build

Axel Kohlmeyer axel.kohlmeyer at theochem.ruhr-uni-bochum.de
Mon May 16 16:00:14 CEST 2005


On Mon, 16 May 2005, Chris Borchert wrote:

hi chris,

did you use the Configure.nocpp script to generate
your Makefile or mkconfig.sh? I think this has not
made it into the manual (can't check right now),
but Configure.nocpp is for Cray X1 platform _only_ 
but Cray X1 also _needs_ it.

regards,
	axel.

CB> I am trying to build CPMD 3.9.2 (for the first time
CB> ever) on a Cray X1 (UNICOS 2.5.31). I'm using the
CB> default cc, ftn, and cpp. On both the 32 and 64-bit
CB> builds the Makefile has to be modified to use "-c" to
CB> compile fortran files, instead of "-o". After several
CB> files successfully build, the 64-bit build dies on
CB> vpsi.f and the 32-bit dies on rnlsm2.f:
CB> ----------64 bit----------
CB> ftn -sdefault64 -dp -eZ -F -ra -sword_pointer 
CB> ./vpsi.f -c  ./vpsi.o
CB>  
CB> !CSD$ end parallel do
CB>  ^
CB> ftn-1749 ftn: ERROR VPSI, File = ./vpsi.f, Line = 245,
CB> Column = 2
CB>   A DO block must directly follow the CSD$ PARALLELDO
CB> directive.
CB>  
CB> !CSD$ end parallel do
CB>  ^
CB> ftn-1749 ftn: ERROR VPSI, File = ./vpsi.f, Line = 277,
CB> Column = 2
CB>   A DO block must directly follow the CSD$ PARALLELDO
CB> directive.
CB>  
CB> !CSD$ end parallel do
CB>  ^
CB> ftn-1749 ftn: ERROR VPSI, File = ./vpsi.f, Line = 294,
CB> Column = 2
CB>   A DO block must directly follow the CSD$ PARALLELDO
CB> directive.
CB>  
CB>       RETURN
CB>       ^
CB> ftn-549 ftn: ERROR VPSI, File = ./vpsi.f, Line = 508,
CB> Column = 7
CB>   A RETURN statement is not allowed within a parallel
CB> region.
CB>  
CB> Cray Fortran: Version 5.2.0.5
CB> (u3030f92057i10081p52116a52022e52009x9228)
CB> Cray Fortran: Thu May  5, 2005  11:17:35
CB> Cray Fortran: Compile time:  0.5600 seconds
CB> Cray Fortran: 2409 source lines
CB> Cray Fortran: 4 errors, 0 warnings, 0 other messages,
CB> 0 ansi
CB> Cray Fortran: "explain ftn-message number" gives more
CB> information about each message
CB> *** Error code 1 (bu21)
CB> ----------32 bit----------
CB> ftn    ./rnlsm2.f -c  ./rnlsm2.o
CB>  
CB> !CSD$ end parallel do
CB>  ^
CB> ftn-1751 ftn: ERROR RNLSM2, File = ./rnlsm2.f, Line =
CB> 137, Column = 2
CB>   The CSD$ ENDPARALLELDO directive must immediately
CB> follow the do loop.
CB>  
CB> !CSD$ end parallel do
CB>  ^
CB> ftn-1751 ftn: ERROR RNLSM2, File = ./rnlsm2.f, Line =
CB> 191, Column = 2
CB>   The CSD$ ENDPARALLELDO directive must immediately
CB> follow the do loop.
CB>  
CB> !CSD$ end parallel do
CB>  ^
CB> ftn-1751 ftn: ERROR RNLSM2, File = ./rnlsm2.f, Line =
CB> 216, Column = 2
CB>   The CSD$ ENDPARALLELDO directive must immediately
CB> follow the do loop.
CB>  
CB> Cray Fortran: Version 5.2.0.5
CB> (u3030f92057i10081p52116a52022e52009x9228)
CB> Cray Fortran: Thu May  5, 2005  11:46:27
CB> Cray Fortran: Compile time:  0.2100 seconds
CB> Cray Fortran: 1495 source lines
CB> Cray Fortran: 3 errors, 0 warnings, 0 other messages,
CB> 0 ansi
CB> Cray Fortran: "explain ftn-message number" gives more
CB> information about each message
CB> *** Error code 1 (bu21)
CB> ----------END----------
CB> 
CB> Any suggestions would be greatly appreciated.
CB> 
CB> Chris
CB> 
CB> 
CB> 
CB> 
CB> 
CB> 
CB> 		
CB> __________________________________ 
CB> Do you Yahoo!? 
CB> Read only the mail you want - Yahoo! Mail SpamGuard. 
CB> http://promotions.yahoo.com/new_mail 
CB> _______________________________________________
CB> CPMD-list mailing list
CB> CPMD-list at cpmd.org
CB> http://cpmd.org/mailman/listinfo/cpmd-list
CB> 
CB> 

-- 

=======================================================================
Dr. Axel Kohlmeyer   e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de
Lehrstuhl fuer Theoretische Chemie          Phone: ++49 (0)234/32-26673
Ruhr-Universitaet Bochum - NC 03/53         Fax:   ++49 (0)234/32-14045
D-44780 Bochum  http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.




More information about the CPMD-list mailing list