lib-4001 : UNRECOVERABLE library error A READ operation tried to read past the end-of-file. Encountered during a sequential unformatted READ from unit 75 Fortran unit 75 is connected to a sequential unformatted file: "CHKPTCHI.002"You left a input_flags chkptchi in the input file when there was no checkpoint data for the nmr routine to restart from. (I should fix paratec so it doesn't crach when this happens)
MPI: hodgkin.hpcf.cam.ac.uk: 0x6bd4ffff0000099e: Exec of paratec.mpi failed MPI: hodgkin.hpcf.cam.ac.uk: 0x6bd4ffff0000099f: Exec of paratec.mpi failed UX:awk: ERROR: Cannot open OUT: No such file or directorysomething went wrong with the binary file paratec.mpi. Check it exists and has the correct architecture
*** MPI has run out of PER_PROC message headers. *** The current allocation levels are: *** MPI_MSGS_PER_HOST = 0 *** MPI_MSGS_PER_PROC = 1024 *** MPI_MSG_RETRIES = 500This has occured when reading the *_AEPS.DAT files. Basically due to braodcasting the data to the nodes in an inefficient fashion. This has now been fixed (Sept 2000)and shouldn't reoccur. If it does setenv MPI_MSGS_PER_PROC 16384 may work (for (t)csh).
When using either of the molecular NMR routines, Paratec fails with a segmentation fault just after printing MOLECULAR ALGORITHM (Eqn 3). The crystal method works fine.
We noticed this on the TCM alpha machines. It is not a paratec fault as such rather a problem with the default user environment. (not enough stack). The seg fault occurs in position_in_rspace.f90. Adding the following lines to your .tcshrc (and starting a new shell) fixes this
set lim=`limit -h datasize` limit $lim set lim=`limit -h stacksize` limit $lim unset lim