From: Cesar Avila (cavila_at_fbqf.unt.edu.ar)
Date: Tue Jul 18 2006 - 09:46:02 CDT
It looks like I have solved my problem. I recompiled charm++ using
net-linux-amd64 with gcc3 and tcp options. After that, I recompiled NAMD.
The aleatory problems I was suffering during simulation dissapeared.
Indeed when I ran the same simulation within another binary/architecture
there was no problem with the simulation.
> Dear all,
> I am using NAMD 2.6B1 on AMD Atlhon Dual Core machines. I have recompiled
> from sources following the directions posted on the Wiki by Jim Philips
> "Load Balancer Hangs on dual-core Opteron clusters", ie moving the 1 to
> CMK_TIMER_USE_TIMES for net-linux-amd64. Now, when I run simulations no
> longer than 10000 steps, the simulation stops on some arbitrary point with
> the following error message:
> ------------- Processor 1 Exiting: Caught Signal ------------
> Signal: segmentation violation
> Suggestion: Try running with '++debug', or linking with '-memory
> paranoid'. Fatal error on PE 1> segmentation violation
> + Exit 1 charmrun +p2 ++local /usr/local/bin/namd2
> equilibrado.namd >equilibrado.log
> Has anyone experienced the same problem? What additional information
> should I provide in order to debug this problem? I am using kernel
> 22.214.171.124 SMP, gcc (GCC) 3.3.5 (Debian 1:3.3.5-13).
> I would deeply appreciate if someone could share with me binaries (or
> specific direction for building them) for running on AMD64 Dual Core
> clusters connected through gigabit ethernet.
> Best Regards
> Cesar Avila
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:22 CST