Dear Paolo Giannozzi Your suspicion is correct.
I'm having the same problem with quantum espresso in a alpha machine running a debian linux. I already trace the point that the code breaks down and I put some code to return the error message got by the fortran runtime : "Fortran runtime error: Writing more data than the record size (RECL)". That's in the file davcio.f90, the block of code is the following: 55: ELSE IF ( io > 0 ) THEN 56: ! 57: WRITE( UNIT = unit, REC = nrec, IOSTAT = ios ) vect ! rafael dias 58: IF ( ios /= 0 ) THEN 59: ..... Well, I change the #define DIRECT_IO_FACTOR to 8 and the code stop the don't break down more. thanks, Rafael Dias --- Em qua, 3/12/08, Paolo Giannozzi <xxxxxx at democritos.it> escreveu: > De: Paolo Giannozzi <xxxxxx at democritos.it> > ... > it might also be related to the value of parameter > DIRECT_IO_FACTOR, > used in PW/diropn.f90: > > #if defined(__ALPHA) > # define DIRECT_IO_FACTOR 2 > #else > # define DIRECT_IO_FACTOR 8 > #end > ! > ! the record length in direct-access I/O is given by > the number of > ! real*8 words times DIRECT_IO_FACTOR (may depend on the > compiler) > ! > > P. > -- > Paolo Giannozzi, Democritos and University of Udine, Italy > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum Veja quais s?o os assuntos do momento no Yahoo! +Buscados http://br.maisbuscados.yahoo.com
