cu_copylist = new cCudaData<int, int, xx > (copylist, max_nsend);
}
void AtomVecAngleCuda::grow_send(int n,double** buf_send,int flag) //need to be able to grow the comm send_buffer since the array sahll be copied from the gpu in whole
void AtomVecAngleCuda::unpack_comm(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only positions are exchanged
void AtomVecAngleCuda::unpack_comm_vel(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only positions are exchanged
int AtomVecAngleCuda::pack_reverse(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only forces are exchanged
void AtomVecAngleCuda::unpack_reverse(int n, int *list, double *buf)//usually this should not be called since comm->communicate handles the communication if only forces are exchanged
cu_copylist = new cCudaData<int, int, xx > (copylist, max_nsend);
}
void AtomVecChargeCuda::grow_send(int n,double** buf_send,int flag) //need to be able to grow the comm send_buffer since the array sahll be copied from the gpu in whole
void AtomVecChargeCuda::unpack_comm(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only positions are exchanged
void AtomVecChargeCuda::unpack_comm_vel(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only positions are exchanged
int AtomVecChargeCuda::pack_reverse(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only forces are exchanged
void AtomVecChargeCuda::unpack_reverse(int n, int *list, double *buf)//usually this should not be called since comm->communicate handles the communication if only forces are exchanged
cu_copylist = new cCudaData<int, int, xx > (copylist, max_nsend);
}
void AtomVecFullCuda::grow_send(int n,double** buf_send,int flag) //need to be able to grow the comm send_buffer since the array sahll be copied from the gpu in whole
void AtomVecFullCuda::unpack_comm(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only positions are exchanged
void AtomVecFullCuda::unpack_comm_vel(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only positions are exchanged
int AtomVecFullCuda::pack_reverse(int n, int first, double *buf) //usually this should not be called since comm->communicate handles the communication if only forces are exchanged
void AtomVecFullCuda::unpack_reverse(int n, int *list, double *buf)//usually this should not be called since comm->communicate handles the communication if only forces are exchanged
Cuda_CommCuda_UnpackComm(&cuda->shared_data,recvnum[iswap],firstrecv[iswap],(void*)buf_recv,iswap); //Unpack for cpu exchange happens implicitely since buf==x[firstrecv]
if(not comm_x_only && not avec->cudable) cuda->downloadAll(); //if not comm_x_only the communication routine of the atom_vec style class is used
// exchange data with another proc
// if other proc is self, just copy
// if comm_x_only set, exchange or copy directly to x, don't unpack
for (int iswap = 0; iswap < nswap; iswap++) {
if (sendproc[iswap] != me)
{
if (comm_x_only)
{
//Cuda_CommCuda_UnpackComm(&cuda->shared_data,recvnum[iswap],firstrecv[iswap],cuda->shared_data.comm.buf_recv[iswap],iswap); //Unpack for cpu exchange happens implicitely since buf==x[firstrecv]
Cuda_CommCuda_UnpackComm(&cuda->shared_data,recvnum[iswap],firstrecv[iswap],buf_recv,iswap); //Unpack for cpu exchange happens implicitely since buf==x[firstrecv]
}
else if (ghost_velocity)
{
//Cuda_CommCuda_UnpackComm_Vel(&cuda->shared_data,recvnum[iswap],firstrecv[iswap],(void*)&buf_recv[iswap*maxrecv]); //Unpack for cpu exchange happens implicitely since buf==x[firstrecv]
if(shared_data.compile_settings.prec_glob!=sizeof(CUDA_FLOAT)/4) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: Global Precision: cuda %i cpp %i\n\n",shared_data.compile_settings.prec_glob, sizeof(CUDA_FLOAT)/4);
if(shared_data.compile_settings.prec_x!=sizeof(X_FLOAT)/4) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: X Precision: cuda %i cpp %i\n\n",shared_data.compile_settings.prec_x, sizeof(X_FLOAT)/4);
if(shared_data.compile_settings.prec_v!=sizeof(V_FLOAT)/4) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: V Precision: cuda %i cpp %i\n\n",shared_data.compile_settings.prec_v, sizeof(V_FLOAT)/4);
if(shared_data.compile_settings.prec_f!=sizeof(F_FLOAT)/4) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: F Precision: cuda %i cpp %i\n\n",shared_data.compile_settings.prec_f, sizeof(F_FLOAT)/4);
if(shared_data.compile_settings.prec_pppm!=sizeof(PPPM_FLOAT)/4) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: PPPM Precision: cuda %i cpp %i\n\n",shared_data.compile_settings.prec_pppm, sizeof(PPPM_FLOAT)/4);
if(shared_data.compile_settings.prec_fft!=sizeof(FFT_FLOAT)/4) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: FFT Precision: cuda %i cpp %i\n\n",shared_data.compile_settings.prec_fft, sizeof(FFT_FLOAT)/4);
#ifdef FFT_CUFFT
if(shared_data.compile_settings.cufft!=1) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: cufft: cuda %i cpp %i\n\n",shared_data.compile_settings.cufft, 1);
#else
if(shared_data.compile_settings.cufft!=0) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: cufft: cuda %i cpp %i\n\n",shared_data.compile_settings.cufft, 0);
#endif
if(shared_data.compile_settings.arch!=CUDA_ARCH) printf("\n\n # CUDA WARNING: Compile Settings of cuda and cpp code differ! \n # CUDA WARNING: arch: cuda %i cpp %i\n\n",shared_data.compile_settings.cufft, CUDA_ARCH);
std::map<NeighList*, CudaNeighList*>::iterator p = neigh_lists.begin();
while(p != neigh_lists.end())
{
delete p->second;
++p;
}
}
void Cuda::accelerator(int narg, char** arg)
{
if(device_set) return;
if(universe->me==0)
printf("# CUDA: Activate GPU \n");
int* devicelist=NULL;
int pppn=2;
for(int i=0;i<narg;i++)
{
if(strcmp(arg[i],"gpu/node")==0)
{
if(++i==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting a number or keyword 'special' after 'gpu/node' option.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting a number or keyword 'special' after 'gpu/node' option.");
if(strcmp(arg[i],"special")==0)
{
if(++i==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting number of GPUs to be used per node after keyword 'gpu/node special'.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting number of GPUs to be used per node after keyword 'gpu/node special'.");
pppn=atoi(arg[i]);
- if(pppn<1) error->all("Invalid Options for 'accelerator' command. Expecting number of GPUs to be used per node after keyword 'gpu/node special'.");
+ if(pppn<1) error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting number of GPUs to be used per node after keyword 'gpu/node special'.");
if(i+pppn==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting list of device ids after keyword 'gpu/node special'.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting list of device ids after keyword 'gpu/node special'.");
devicelist=new int[pppn];
for(int k=0;k<pppn;k++)
{i++;devicelist[k]=atoi(arg[i]);}
}
else
pppn=atoi(arg[i]);
}
if(strcmp(arg[i],"pinned")==0)
{
if(++i==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting a number after 'pinned' option.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting a number after 'pinned' option.");
pinned=atoi(arg[i])==0?false:true;
if((pinned==false)&&(universe->me==0)) printf(" #CUDA: Pinned memory is not used for communication\n");
}
if(strcmp(arg[i],"dotiming")==0)
{
dotiming=true;
}
if(strcmp(arg[i],"suffix")==0)
{
if(++i==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting a string after 'suffix' option.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting a string after 'suffix' option.");
strcpy(lmp->suffix,arg[i]);
}
if(strcmp(arg[i],"overlap_comm")==0)
{
shared_data.overlap_comm=1;
}
if(strcmp(arg[i],"dotest")==0)
{
if(++i==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting a number after 'dotest' option.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting a number after 'dotest' option.");
testatom=atof(arg[i]);
dotestatom=true;
}
if(strcmp(arg[i],"override_bpa")==0)
{
if(++i==narg)
- error->all("Invalid Options for 'accelerator' command. Expecting a number after 'override_bpa' option.");
+ error->all(FLERR,"Invalid Options for 'accelerator' command. Expecting a number after 'override_bpa' option.");
- //if(list->cuda_list->sneighlist.bin_nmax>512) error->all("To many atoms per bin. Likely cause is very long pair cutoff. This needs major rewrite of code and is not yet scheduled to be done.\n");
+ //if(list->cuda_list->sneighlist.bin_nmax>512) error->all(FLERR,"To many atoms per bin. Likely cause is very long pair cutoff. This needs major rewrite of code and is not yet scheduled to be done.\n");
- error->all("Pair style is incompatible with KSpace style");
+ error->all(FLERR,"Pair style is incompatible with KSpace style");
g_ewald = force->kspace->g_ewald;
cuda->shared_data.pair.g_ewald=g_ewald;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
- error->all("Pair style is incompatible with KSpace style");
+ error->all(FLERR,"Pair style is incompatible with KSpace style");
g_ewald = force->kspace->g_ewald;
cuda->shared_data.pair.g_ewald=g_ewald;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
- error->all("Pair style buck/coul/long requires atom attribute q");
+ error->all(FLERR,"Pair style buck/coul/long requires atom attribute q");
// request regular or rRESPA neighbor lists
int irequest;
- if (strstr(update->integrate_style,"respa")) error->all("Integrate Style Respa is not supported by pair style buck/coul/long/cuda");
+ if (strstr(update->integrate_style,"respa")) error->all(FLERR,"Integrate Style Respa is not supported by pair style buck/coul/long/cuda");
irequest = neighbor->request(this);
neighbor->requests[irequest]->full = 1;
neighbor->requests[irequest]->half = 0;
neighbor->requests[irequest]->cudable = 1;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
MYDBG(printf("# CUDA PairCGCMMCoulCutCuda::init_style start\n"); )
// request regular or rRESPA neighbor lists
int irequest;
if (update->whichflag == 0 && strstr(update->integrate_style,"respa")) {
}
else
{
irequest = neighbor->request(this);
neighbor->requests[irequest]->full = 1;
neighbor->requests[irequest]->half = 0;
neighbor->requests[irequest]->cudable = 1;
//neighbor->style=0; //0=NSQ neighboring
}
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
cut_respa=NULL;
- if (force->newton) error->warning("Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
+ if (force->newton) error->warning(FLERR,"Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
MYDBG(printf("# CUDA PairCGCMMCoulCutCuda::init_style end\n"); )
MYDBG(printf("# CUDA PairCGCMMCoulDebyeCuda::init_style start\n"); )
// request regular or rRESPA neighbor lists
int irequest;
if (update->whichflag == 0 && strstr(update->integrate_style,"respa")) {
}
else
{
irequest = neighbor->request(this);
neighbor->requests[irequest]->full = 1;
neighbor->requests[irequest]->half = 0;
neighbor->requests[irequest]->cudable = 1;
//neighbor->style=0; //0=NSQ neighboring
}
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
cut_respa=NULL;
- if (force->newton) error->warning("Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
+ if (force->newton) error->warning(FLERR,"Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
MYDBG(printf("# CUDA PairCGCMMCoulDebyeCuda::init_style end\n"); )
MYDBG(printf("# CUDA PairCGCMMCoulLongCuda::init_style start\n"); )
// request regular or rRESPA neighbor lists
int irequest;
if (update->whichflag == 0 && strstr(update->integrate_style,"respa")) {
}
else
{
irequest = neighbor->request(this);
neighbor->requests[irequest]->full = 1;
neighbor->requests[irequest]->half = 0;
neighbor->requests[irequest]->cudable = 1;
//neighbor->style=0; //0=NSQ neighboring
}
g_ewald = force->kspace->g_ewald;
cuda->shared_data.pair.g_ewald=g_ewald;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
cut_respa=NULL;
- if (force->newton) error->warning("Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
+ if (force->newton) error->warning(FLERR,"Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
MYDBG(printf("# CUDA PairCGCMMCoulLongCuda::init_style end\n"); )
- error->all("Pair style is incompatible with KSpace style");
+ error->all(FLERR,"Pair style is incompatible with KSpace style");
g_ewald = force->kspace->g_ewald;
cuda->shared_data.pair.g_ewald=g_ewald;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
- if (force->newton) error->warning("Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
+ if (force->newton) error->warning(FLERR,"Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
if (force->kspace == NULL)
- error->all("Pair style is incompatible with KSpace style");
+ error->all(FLERR,"Pair style is incompatible with KSpace style");
g_ewald = force->kspace->g_ewald;
cuda->shared_data.pair.g_ewald=g_ewald;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
- if (force->newton) error->warning("Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
+ if (force->newton) error->warning(FLERR,"Pair style uses does not use \"newton\" setting. You might test if \"newton off\" makes the simulation run faster.");
if (force->kspace == NULL)
- error->all("Pair style is incompatible with KSpace style");
+ error->all(FLERR,"Pair style is incompatible with KSpace style");
g_ewald = force->kspace->g_ewald;
cuda->shared_data.pair.g_ewald=g_ewald;
cuda->shared_data.pppm.qqrd2e=force->qqrd2e;
- if(ncoultablebits) error->warning("# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
+ if(ncoultablebits) error->warning(FLERR,"# CUDA: You asked for the useage of Coulomb Tables. This is not supported in CUDA Pair forces. Setting is ignored.\n");
fprintf(screen,"\n\npppm/cuda WARNING: \t\tSignificantly lower gridsize than requested.\n\t\t\t\t\tYou should most likely use '=' or '+' as precision modify option\n");
}
if (logfile) {
fprintf(logfile,"\n\npppm/cuda WARNING: \t\tSignificantly lower gridsize than requested.\n\t\t\t\t\tYou should most likely use '=' or '+' as precision modify option\n.");
MYDBG( printf("# CUDA VerletCuda::iterate: neighbor done\n"); )
//if bonded interactions are used (in this case collect_forces_later is true), transfer data which only changes upon exchange/border routines from GPU to CPU