Re: Sufficiently large values for 'firsttimestep' can lead to immediate simulation abort.

From: Nikolaos Glykos (glykos_at_mbg.duth.gr)
Date: Sun Apr 08 2012 - 13:08:59 CDT

 Hi Axel,

> the symptoms you describe are that of a signed 32bit integer
> overflow. it is a little bit premature, since the maximum
> representable
> signed 32-bit integer would be 2,147,483,647. did you ask for
> 500,000,000 steps or more of additional MD?

 Ah, well spotted. Thanks. The overflow came from the addition. Changing
 'run 500000000' to 'run 400000000' did bypassed the issue. I can't
 say that I understand why the timestep should not be (at least
 internally) an unsigned long int but there I go ...
 [do I remember correctly that the DCD header contains total number
 of frames and not number of steps ?]

 Thanks,
 Nicholas

-- 
             Nicholas M. Glykos, Department of Molecular Biology
      and Genetics, Democritus University of Thrace, University Campus,
   Dragana, 68100 Alexandroupolis, Greece, Tel/Fax (office) 
 +302551030620,
     Ext.77620, Tel (lab) +302551030615, http://utopia.duth.gr/~glykos/

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:25 CST