On Wed, 2013-11-06 at 13:51 +0000, Hannes Loeffler wrote:
> On Wed, 6 Nov 2013 08:41:31 -0500
> Jason Swails <jason.swails.gmail.com> wrote:
>
> > On Wed, 2013-11-06 at 21:18 +0800, David Chen wrote:
> > > I got error message on terminal:
> > > mmpbsa_py_energy: malloc.c:2451: sYSMALLOc: Assertion `(old_top ==
> > > (((mbinptr) (((char *) &((av)->bins[((1) - 1) * 2])) -
> > > __builtin_offsetof (struct malloc_chunk, fd)))) && old_size == 0)
> > > || ((unsigned long) (old_size) >= (unsigned
> > > long)((((__builtin_offsetof (struct malloc_chunk, fd_nextsize))+((2
> > > * (sizeof(size_t))) - 1)) & ~((2 * (sizeof(size_t))) - 1))) &&
> > > ((old_top)->size & 0x1) && ((unsigned long)old_end & pagemask) ==
> > > 0)' failed.
> >
> > Well that's a strange error message. I'm not sure why malloc would
> > raise an error like this in a GB calculation. GB is not very
> > memory-intensive, so it's highly unlikely that your machine is running
> > out of RAM unless you are using a queueing system (torque/PBS, slurm,
> > etc.) which has a low memory limit on the job. If this is the case,
> > you can try to request more memory and see if that helps.
>
> I believe that's a hint to memory corruption in whatever code caused it.
It is being caused in a NAB program. There are a number of subtle ways
to break NAB in my experience (relocating it, copying it from one
machine to another, etc.). I'm certainly not ruling out a code problem,
but based on the available information and past experience, I'm inclined
to believe there is a different root cause of the memory corruption.
All the best,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 06 2013 - 06:30:03 PST
