On Sun, Jan 10, 2016 at 1:20 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
>
> Note there is this option --fno-underflow-warning. Jason perhaps we should
> include this in the build flags. Although that probably means checking
> compiler versions for compatibility with the sign. Sigh...
>
​Makes more sense to me to just avoid the divide-by-zero. If sander or
pmemd encounters a zero, it can either set the scaling factors to 0 (which
will 0-out the 1-4 even if the pair *is* mistakenly included in the 1-4
loop), or set it manually to something huge to make the simulation blow
up. Seems more portable and future-proof to me...
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 Mon Jan 11 2016 - 05:30:03 PST
