Re: Re: NAMD with Myrinet MX on MacOS X

From: Philip Blood (philb_at_hec.utah.edu)
Date: Wed Mar 01 2006 - 11:26:38 CST

Hi,

We tried to build and run this MX version on an AMD Opteron cluster, but
we encountered this error in the PBS output:

------------- Processor 61 Exiting: Caught Signal ------------
Signal: segmentation violation
Suggestion: Try running with '++debug', or linking with '-memory paranoid'.
Fatal error on PE 61> segmentation violation

and from the NAMD output:

Charm++: scheduler running in netpoll mode.
Info: NAMD 2.6b1 for Linux-amd64
Info:
Info: Please visit http://www.ks.uiuc.edu/Research/namd/
Info: and send feedback or bug reports to namd_at_ks.uiuc.edu
Info:
Info: Please cite Phillips et al., J. Comp. Chem. 26:1781-1802 (2005)
Info: in all publications reporting results obtained with NAMD.
Info:
Info: Based on Charm++/Converse 50911 for net-linux-amd64-mx
Info: Built Sun Feb 5 13:16:48 MST 2006 by mcuma on delicatearch1
Info: Sending usage information to NAMD developers via UDP. Sent data is:
Info: 1 NAMD 2.6b1 Linux-amd64 200 da189 philb
Info: Running on 200 processors.
Info: 9670 kB of memory in use.
Info: Configuration file is d.namd
TCL: Suspending until startup complete.
Stack Traceback:
  [0] /lib64/libpthread.so.0 [0x2aaaab5a8bde]
  [1] _int_malloc+0x280 [0x6b3338]
  [2] mm_malloc+0x9e [0x6b2192]
  [3] malloc+0x1f [0x6b5cec]
  [4] CkCreateLocalGroup+0x44 [0x6c2712]
  [5] _Z18_processBocInitMsgP11CkCoreStateP8envelope+0x34 [0x6c3120]
  [6] _Z15_processHandlerPvP11CkCoreState+0xc2 [0x6c32ce]
  [7] CmiHandleMessage+0x49 [0x71d123]
  [8] CsdScheduleForever+0x5f [0x71d35a]
  [9] CsdScheduler+0x16 [0x71d2d3]
  [10] _ZN7BackEnd4initEiPPc+0x1a0 [0x4c6740]

We are running NAMD with charmrun with the ++batch 10 option. Any ideas?

Thanks,
Phil

Gengbin Zheng wrote:

>
> Hi all,
>
> Based on the recent popular requests of porting NAMD/Charm to MX, we
> have added new support for both native MX library and MPICH/MX in
> charm++, and enhanced the build procedure of Charm++.
>
> I added detailed information on how to build charm++ at NAMD wiki page:
> http://www.ks.uiuc.edu/Research/namd/wiki/index.cgi?NamdOnMyrinet
>
> The native MX support (and so not using MPICH/MX) in Charm++ is very
> new, and more optimizations are coming up. But please let us know if
> you have any problems,
>
> Gengbin
>
> Goran Krilov wrote:
>
>> Hi,
>>
>> I am trying to build NAMD on Myrinet using the new MX Myrinet
>> protocols rather than GM. Has anyone accomplished this,
>> and what would be the correct strategy? I found only GM based
>> architecture
>> files for Charm builds. Many thanks!
>>
>> ------------------------------------------------------------------------
>> ----
>> Dr. Goran Krilov Tel. (617) 552-8927
>> Assistant Professor Fax (617) 552-2705
>> Department of Chemistry E-mail: goran.krilov_at_bc.edu
>> Boston College
>> Merkert Center, 2609 Beacon St
>> Chestnut Hill, MA 02467
>> ------------------------------------------------------------------------
>> -----
>>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:40 CST