Re: AMBER: amber8 compile on Lemieux

From: David A. Case <case.scripps.edu>
Date: Thu, 3 Jun 2004 19:39:58 -0700

On Thu, Jun 03, 2004, Guanglei Cui wrote:
>
> I tried to compile amber8 with ./configure -mpi compaqf90, but am having
> problem with the compiled sander binary. Here is the error message from
> mdout and PBS. Did I compile it correctly? It seems there's a
> Machines/Machines.axp_f90_quadrics, but how should I choose it? Thanks.
>
>
> mdout--
> | Flags: MPI getting new box info from bottom of inpcrd|
> peek_ewald_inpcrd: Box info found
> |Largest sphere to fit in unit cell has radius = 41.169
> | INFO: Old style PARM file read
>
> PBS--
> forrtl: severe (64): input conversion error, unit 8, file
> /cluster/members/member1/local/268380/md10.r

Guanglei: does your program pass the parallel tests? From the above messages
it looks (?) like the "md10.r" file is being read as unit 8, which I think
is for a prmtop file; but the name looks like a coordinate or restart file.
Is there any chance you just have the command line arguments messed up?
Double check the file assignments listed at the top of the output file.

(This is obviously just a guess, but might be worth checking.)

....dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri Jun 04 2004 - 10:53:02 PDT
Custom Search