[gmx-developers] Fluctuating Charge and kernels

[Anton Feenstra]

On 24/02/14 16:59, Yuriy Khalak wrote:
> Hi.
> I'm trying to implement a fluctuating charge model into GROMACS (v 4.6.3).
> I've got the code for calculating the partial charges working (I think)
> and am updating the mdatoms->chargeA array each time step right before
> the Ewald / PME code in do_force_lowlevel(). However, after a few time
> steps the simulation crashes with a segmentation fault in
> nb_kernel_ElecEw_VdwLJ_GeomP1P1_F_sse2_single(). I'm baffled by what
> could be causing this. Is GROMACS using the chargeA array anywhere else
> besides the PME code? Are the kernels somehow buffering it between
> different time steps?

Long range forces are updated every nstlist steps;
could that be what bites you here?


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | IBIVU/Bioinformatics - Vrije Universiteit Amsterdam   |
|(   |   )| | | De Boelelaan 1081 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel +31 20 59 87783 - Fax +31 20 59 87653 - Room P136 |
|             | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/         |
|             | "Once Again Let Me Do This" (Urban Dance Squad)       |
|_____________|_______________________________________________________|