jianmu
/
lammps
forked from lijiext/lammps
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

kspace & dihedral can't do their own sync/modify 

because the verlet_kokkos system has
a "clever" optimization which will
alter the datamasks before calling sync/modify,
so the datamask framework must be
strictly obeyed for GPU correctness.
(the optimization is to concurrently
compute forces on the host and GPU,
and add them up at the end of an iteration.
calling your own sync will overwrite
the partial GPU forces with the
partial host forces).
This commit is contained in:
Dan Ibanez 2017-01-27 08:21:57 -07:00
parent f8506fee23
commit 09fc8b0bd7
2 changed files with 1 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/KOKKOS/dihedral_opls_kokkos.cpp
View File

@ -85,13 +85,10 @@ void DihedralOPLSKokkos<DeviceType>::compute(int eflag_in, int vflag_in)
d_vatom = k_vatom.d_view;
}
atomKK->sync(execution_space,datamask_read);
k_k1.template sync<DeviceType>();
k_k2.template sync<DeviceType>();
k_k3.template sync<DeviceType>();
k_k4.template sync<DeviceType>();
if (eflag || vflag) atomKK->modified(execution_space,datamask_modify);
else atomKK->modified(execution_space,F_MASK);
x = atomKK->k_x.view<DeviceType>();
f = atomKK->k_f.view<DeviceType>();

5
src/KOKKOS/pppm_kokkos.cpp
View File

@ -646,10 +646,7 @@ void PPPMKokkos<DeviceType>::compute(int eflag, int vflag)
x = atomKK->k_x.view<DeviceType>();
f = atomKK->k_f.view<DeviceType>();
q = atomKK->k_q.view<DeviceType>();
atomKK->sync(execution_space,datamask_read);
atomKK->modified(execution_space,datamask_modify);
//nlocal = atomKK->nlocal;
//nall = atomKK->nlocal + atomKK->nghost;
//newton_pair = force->newton_pair;