git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@4789 f3b2605a-c512-4ea7-a41b-209d697bcdaa

This commit is contained in:
sjplimp 2010-09-14 23:47:34 +00:00
parent 45c17620cb
commit cd163c2dcd
2 changed files with 716 additions and 0 deletions

636
src/irregular.cpp Normal file
View File

@ -0,0 +1,636 @@
/* ----------------------------------------------------------------------
LAMMPS - Large-scale Atomic/Molecular Massively Parallel Simulator
http://lammps.sandia.gov, Sandia National Laboratories
Steve Plimpton, sjplimp@sandia.gov
Copyright (2003) Sandia Corporation. Under the terms of Contract
DE-AC04-94AL85000 with Sandia Corporation, the U.S. Government retains
certain rights in this software. This software is distributed under
the GNU General Public License.
See the README file in the top-level LAMMPS directory.
------------------------------------------------------------------------- */
#include "mpi.h"
#include "stdlib.h"
#include "string.h"
#include "irregular.h"
#include "atom.h"
#include "atom_vec.h"
#include "domain.h"
#include "comm.h"
#include "memory.h"
using namespace LAMMPS_NS;
#define BUFFACTOR 1.5
#define BUFMIN 1000
#define BUFEXTRA 1000
#define MIN(a,b) ((a) < (b) ? (a) : (b))
#define MAX(a,b) ((a) > (b) ? (a) : (b))
/* ---------------------------------------------------------------------- */
Irregular::Irregular(LAMMPS *lmp) : Pointers(lmp)
{
MPI_Comm_rank(world,&me);
MPI_Comm_size(world,&nprocs);
triclinic = domain->triclinic;
map_style = atom->map_style;
procgrid = comm->procgrid;
grid2proc = comm->grid2proc;
// initialize comm buffers
maxsend = BUFMIN;
buf_send = (double *)
memory->smalloc((maxsend+BUFEXTRA)*sizeof(double),"comm:buf_send");
maxrecv = BUFMIN;
buf_recv = (double *)
memory->smalloc(maxrecv*sizeof(double),"comm:buf_recv");
}
/* ---------------------------------------------------------------------- */
Irregular::~Irregular()
{
memory->sfree(buf_send);
memory->sfree(buf_recv);
}
/* ----------------------------------------------------------------------
communicate atoms to new owning procs via irregular communication
can be used in place of comm->exchange()
unlike exchange(), allows atoms to have moved arbitrarily long distances
first setup irregular comm pattern, then invoke it
atoms must be remapped to be inside simulation box before this is called
for triclinic: atoms must be in lamda coords (0-1) before this is called
------------------------------------------------------------------------- */
void Irregular::migrate_atoms()
{
// clear global->local map since atoms move to new procs
// zero out ghosts so map_set() at end will operate only on local atoms
// exchange() doesn't need to zero ghosts b/c borders()
// is called right after and it zeroes ghosts and calls map_set()
if (map_style) atom->map_clear();
atom->nghost = 0;
// subbox bounds for orthogonal or triclinic
double *sublo,*subhi;
if (triclinic == 0) {
sublo = domain->sublo;
subhi = domain->subhi;
} else {
sublo = domain->sublo_lamda;
subhi = domain->subhi_lamda;
}
// loop over atoms, flag any that are not in my sub-box
// fill buffer with atoms leaving my box, using < and >=
// assign which proc it belongs to via coord2proc()
// if coord2proc() returns me, due to round-off
// in triclinic x2lamda(), then keep atom and don't send
// when atom is deleted, fill it in with last atom
AtomVec *avec = atom->avec;
double **x = atom->x;
int nlocal = atom->nlocal;
int nsend = 0;
int nsendatom = 0;
int *sizes = new int[nlocal];
int *proclist = new int[nlocal];
int i = 0;
while (i < nlocal) {
if (x[i][0] < sublo[0] || x[i][0] >= subhi[0] ||
x[i][1] < sublo[1] || x[i][1] >= subhi[1] ||
x[i][2] < sublo[2] || x[i][2] >= subhi[2]) {
proclist[nsendatom] = coord2proc(x[i]);
if (proclist[nsendatom] != me) {
if (nsend > maxsend) grow_send(nsend,1);
sizes[nsendatom] = avec->pack_exchange(i,&buf_send[nsend]);
nsend += sizes[nsendatom];
nsendatom++;
avec->copy(nlocal-1,i);
nlocal--;
} else i++;
} else i++;
}
atom->nlocal = nlocal;
// create irregular communication plan, perform comm, destroy plan
// returned nrecv = size of buffer needed for incoming atoms
int nrecv;
PlanAtom *plan = create_atom(nsendatom,sizes,proclist,&nrecv);
if (nrecv > maxrecv) grow_recv(nrecv);
exchange_atom(plan,buf_send,sizes,buf_recv);
destroy_atom(plan);
delete [] sizes;
delete [] proclist;
// add received atoms to my list
int m = 0;
while (m < nrecv) m += avec->unpack_exchange(&buf_recv[m]);
// reset global->local map
if (map_style) atom->map_set();
}
/* ----------------------------------------------------------------------
create an irregular communication plan for atoms
n = # of atoms to send
sizes = # of doubles for each atom
proclist = proc to send each atom to (none to me)
return nrecvsize = total # of doubles I will recv
------------------------------------------------------------------------- */
Irregular::PlanAtom *Irregular::create_atom(int n, int *sizes,
int *proclist, int *nrecvsize)
{
int i;
// allocate plan and work vectors
PlanAtom *plan = (struct PlanAtom *)
memory->smalloc(sizeof(PlanAtom),"irregular:plan");
int *list = new int[nprocs];
int *count = new int[nprocs];
// nrecv = # of messages I receive
for (i = 0; i < nprocs; i++) {
list[i] = 0;
count[i] = 1;
}
for (i = 0; i < n; i++) list[proclist[i]] = 1;
int nrecv;
MPI_Reduce_scatter(list,&nrecv,count,MPI_INT,MPI_SUM,world);
// allocate receive arrays
int *proc_recv = new int[nrecv];
int *length_recv = new int[nrecv];
MPI_Request *request = new MPI_Request[nrecv];
MPI_Status *status = new MPI_Status[nrecv];
// nsend = # of messages I send
for (i = 0; i < nprocs; i++) list[i] = 0;
for (i = 0; i < n; i++) list[proclist[i]] += sizes[i];
int nsend = 0;
for (i = 0; i < nprocs; i++)
if (list[i]) nsend++;
// allocate send arrays
int *proc_send = new int[nsend];
int *length_send = new int[nsend];
int *num_send = new int[nsend];
int *index_send = new int[n];
int *offset_send = new int[n];
// list still stores size of message for procs I send to
// proc_send = procs I send to
// length_send = total size of message I send to each proc
// to balance pattern of send messages:
// each proc begins with iproc > me, continues until iproc = me
// reset list to store which send message each proc corresponds to
int iproc = me;
int isend = 0;
for (i = 0; i < nprocs; i++) {
iproc++;
if (iproc == nprocs) iproc = 0;
if (list[iproc] > 0) {
proc_send[isend] = iproc;
length_send[isend] = list[iproc];
list[iproc] = isend;
isend++;
}
}
// num_send = # of datums I send to each proc
for (i = 0; i < nsend; i++) num_send[i] = 0;
for (i = 0; i < n; i++) {
isend = list[proclist[i]];
num_send[isend]++;
}
// count = offsets into n-length index_send for each proc I send to
// index_send = list of which datums to send to each proc
// 1st N1 values are datum indices for 1st proc,
// next N2 values are datum indices for 2nd proc, etc
// offset_send = where each datum starts in send buffer
count[0] = 0;
for (i = 1; i < nsend; i++) count[i] = count[i-1] + num_send[i-1];
for (i = 0; i < n; i++) {
isend = list[proclist[i]];
index_send[count[isend]++] = i;
if (i) offset_send[i] = offset_send[i-1] + sizes[i-1];
else offset_send[i] = 0;
}
// tell receivers how much data I send
// sendmax = largest # of datums I send in a single message
int sendmax = 0;
for (i = 0; i < nsend; i++) {
MPI_Send(&length_send[i],1,MPI_INT,proc_send[i],0,world);
sendmax = MAX(sendmax,length_send[i]);
}
// receive incoming messages
// proc_recv = procs I recv from
// length_recv = total size of message each proc sends me
// nrecvsize = total size of data I recv
*nrecvsize = 0;
for (i = 0; i < nrecv; i++) {
MPI_Recv(&length_recv[i],1,MPI_INT,MPI_ANY_SOURCE,0,world,status);
proc_recv[i] = status->MPI_SOURCE;
*nrecvsize += length_recv[i];
}
// barrier to insure all MPI_ANY_SOURCE messages are received
// else another proc could proceed to irregular_perform() and send to me
MPI_Barrier(world);
// free work vectors
delete [] count;
delete [] list;
// initialize plan and return it
plan->nsend = nsend;
plan->nrecv = nrecv;
plan->sendmax = sendmax;
plan->proc_send = proc_send;
plan->length_send = length_send;
plan->num_send = num_send;
plan->index_send = index_send;
plan->offset_send = offset_send;
plan->proc_recv = proc_recv;
plan->length_recv = length_recv;
plan->request = request;
plan->status = status;
return plan;
}
/* ----------------------------------------------------------------------
create an irregular communication plan for datums
n = # of datums to send
proclist = proc to send each datum to (none to me)
return nrecvsize = total # of datums I will recv
------------------------------------------------------------------------- */
Irregular::PlanData *Irregular::create_data(int n, int *proclist,
int *nrecvsize)
{
int i;
// allocate plan and work vectors
PlanData *plan = (struct PlanData *)
memory->smalloc(sizeof(PlanData),"irregular:plan");
int *list = new int[nprocs];
int *count = new int[nprocs];
// nrecv = # of messages I receive
for (i = 0; i < nprocs; i++) {
list[i] = 0;
count[i] = 1;
}
for (i = 0; i < n; i++) list[proclist[i]] = 1;
int nrecv;
MPI_Reduce_scatter(list,&nrecv,count,MPI_INT,MPI_SUM,world);
// allocate receive arrays
int *proc_recv = new int[nrecv];
int *num_recv = new int[nrecv];
MPI_Request *request = new MPI_Request[nrecv];
MPI_Status *status = new MPI_Status[nrecv];
// nsend = # of messages I send
for (i = 0; i < nprocs; i++) list[i] = 0;
for (i = 0; i < n; i++) list[proclist[i]]++;
int nsend = 0;
for (i = 0; i < nprocs; i++)
if (list[i]) nsend++;
// allocate send arrays
int *proc_send = new int[nsend];
int *num_send = new int[nsend];
int *index_send = new int[n];
// proc_send = procs I send to
// num_send = # of datums I send to each proc
// to balance pattern of send messages:
// each proc begins with iproc > me, continues until iproc = me
int iproc = me;
int isend = 0;
for (i = 0; i < nprocs; i++) {
iproc++;
if (iproc == nprocs) iproc = 0;
if (list[iproc] > 0) {
proc_send[isend] = iproc;
num_send[isend] = list[iproc];
isend++;
}
}
// count = offsets into n-length index_send for each proc I send to
// index_send = list of which datums to send to each proc
// 1st N1 values are datum indices for 1st proc,
// next N2 values are datum indices for 2nd proc, etc
// offset_send = where each datum starts in send buffer
count[0] = 0;
for (i = 1; i < nsend; i++) count[i] = count[i-1] + num_send[i-1];
for (i = 0; i < n; i++) {
isend = list[proclist[i]];
index_send[count[isend]++] = i;
}
// tell receivers how much data I send
// sendmax = largest # of datums I send in a single message
int sendmax = 0;
for (i = 0; i < nsend; i++) {
MPI_Send(&num_send[i],1,MPI_INT,proc_send[i],0,world);
sendmax = MAX(sendmax,num_send[i]);
}
// receive incoming messages
// proc_recv = procs I recv from
// num_recv = total size of message each proc sends me
// nrecvsize = total size of data I recv
*nrecvsize = 0;
for (i = 0; i < nrecv; i++) {
MPI_Recv(&num_recv[i],1,MPI_INT,MPI_ANY_SOURCE,0,world,status);
proc_recv[i] = status->MPI_SOURCE;
*nrecvsize += num_recv[i];
}
// barrier to insure all MPI_ANY_SOURCE messages are received
// else another proc could proceed to irregular_perform() and send to me
MPI_Barrier(world);
// free work vectors
delete [] count;
delete [] list;
// initialize plan and return it
plan->nsend = nsend;
plan->nrecv = nrecv;
plan->sendmax = sendmax;
plan->proc_send = proc_send;
plan->num_send = num_send;
plan->index_send = index_send;
plan->proc_recv = proc_recv;
plan->num_recv = num_recv;
plan->request = request;
plan->status = status;
return plan;
}
/* ----------------------------------------------------------------------
perform irregular communication of atoms
plan = previouly computed PlanAtom via create_atom()
sendbuf = list of atoms to send
sizes = # of doubles for each atom
recvbuf = received atoms
------------------------------------------------------------------------- */
void Irregular::exchange_atom(PlanAtom *plan, double *sendbuf, int *sizes,
double *recvbuf)
{
int i,m,n,offset;
// post all receives
offset = 0;
for (int irecv = 0; irecv < plan->nrecv; irecv++) {
MPI_Irecv(&recvbuf[offset],plan->length_recv[irecv],MPI_DOUBLE,
plan->proc_recv[irecv],0,world,&plan->request[irecv]);
offset += plan->length_recv[irecv];
}
// allocate buf for largest send
double *buf = (double *) memory->smalloc(plan->sendmax*sizeof(double),
"irregular:buf");
// send each message
// pack buf with list of datums (datum = one atom)
// m = index of datum in sendbuf
n = 0;
for (int isend = 0; isend < plan->nsend; isend++) {
offset = 0;
for (i = 0; i < plan->num_send[isend]; i++) {
m = plan->index_send[n++];
memcpy(&buf[offset],&sendbuf[plan->offset_send[m]],
sizes[m]*sizeof(double));
offset += sizes[m];
}
MPI_Send(buf,plan->length_send[isend],MPI_DOUBLE,
plan->proc_send[isend],0,world);
}
// free temporary send buffer
memory->sfree(buf);
// wait on all incoming messages
if (plan->nrecv) MPI_Waitall(plan->nrecv,plan->request,plan->status);
}
/* ----------------------------------------------------------------------
perform irregular communication of datums
plan = previouly computed PlanData via create_data()
sendbuf = list of datums to send
nbytes = size of each datum
recvbuf = received datums
------------------------------------------------------------------------- */
void Irregular::exchange_data(PlanData *plan, char *sendbuf, int nbytes,
char *recvbuf)
{
int i,m,n,offset;
// post all receives
offset = 0;
for (int irecv = 0; irecv < plan->nrecv; irecv++) {
MPI_Irecv(&recvbuf[offset],nbytes*plan->num_recv[irecv],MPI_CHAR,
plan->proc_recv[irecv],0,world,&plan->request[irecv]);
offset += nbytes*plan->num_recv[irecv];
}
// allocate buf for largest send
char *buf = (char *) memory->smalloc(plan->sendmax*nbytes,"irregular:buf");
// send each message
// pack buf with list of datums (datum = one atom)
// m = index of datum in sendbuf
n = 0;
for (int isend = 0; isend < plan->nsend; isend++) {
for (i = 0; i < plan->num_send[isend]; i++) {
m = plan->index_send[n++];
memcpy(&buf[i*nbytes],&sendbuf[m*nbytes],nbytes);
}
MPI_Send(buf,nbytes*plan->num_send[isend],MPI_CHAR,
plan->proc_send[isend],0,world);
}
// free temporary send buffer
memory->sfree(buf);
// copy datums owned by self
for (i = 0; i < plan->num_self; i++) {
m = plan->index_self[i++];
memcpy(&recvbuf[i*nbytes],&sendbuf[m*nbytes],nbytes);
}
// wait on all incoming messages
if (plan->nrecv) MPI_Waitall(plan->nrecv,plan->request,plan->status);
}
/* ----------------------------------------------------------------------
destroy an irregular communication plan for atoms
------------------------------------------------------------------------- */
void Irregular::destroy_atom(PlanAtom *plan)
{
delete [] plan->proc_send;
delete [] plan->length_send;
delete [] plan->num_send;
delete [] plan->index_send;
delete [] plan->offset_send;
delete [] plan->proc_recv;
delete [] plan->length_recv;
delete [] plan->request;
delete [] plan->status;
memory->sfree(plan);
}
/* ----------------------------------------------------------------------
destroy an irregular communication plan for datums
------------------------------------------------------------------------- */
void Irregular::destroy_data(PlanData *plan)
{
delete [] plan->proc_send;
delete [] plan->num_send;
delete [] plan->index_send;
delete [] plan->proc_recv;
delete [] plan->num_recv;
delete [] plan->request;
delete [] plan->status;
memory->sfree(plan);
}
/* ----------------------------------------------------------------------
determine which proc owns atom with x coord
x will be in box (orthogonal) or lamda coords (triclinic)
------------------------------------------------------------------------- */
int Irregular::coord2proc(double *x)
{
int loc[3];
if (triclinic == 0) {
double *boxlo = domain->boxlo;
double *boxhi = domain->boxhi;
loc[0] = static_cast<int>
(procgrid[0] * (x[0]-boxlo[0]) / (boxhi[0]-boxlo[0]));
loc[1] = static_cast<int>
(procgrid[1] * (x[1]-boxlo[1]) / (boxhi[1]-boxlo[1]));
loc[2] = static_cast<int>
(procgrid[2] * (x[2]-boxlo[2]) / (boxhi[2]-boxlo[2]));
} else {
loc[0] = static_cast<int> (procgrid[0] * x[0]);
loc[1] = static_cast<int> (procgrid[1] * x[1]);
loc[2] = static_cast<int> (procgrid[2] * x[2]);
}
if (loc[0] < 0) loc[0] = 0;
if (loc[0] >= procgrid[0]) loc[0] = procgrid[0] - 1;
if (loc[1] < 0) loc[1] = 0;
if (loc[1] >= procgrid[1]) loc[1] = procgrid[1] - 1;
if (loc[2] < 0) loc[2] = 0;
if (loc[2] >= procgrid[2]) loc[2] = procgrid[2] - 1;
return grid2proc[loc[0]][loc[1]][loc[2]];
}
/* ----------------------------------------------------------------------
realloc the size of the send buffer as needed with BUFFACTOR & BUFEXTRA
if flag = 1, realloc
if flag = 0, don't need to realloc with copy, just free/malloc
------------------------------------------------------------------------- */
void Irregular::grow_send(int n, int flag)
{
maxsend = static_cast<int> (BUFFACTOR * n);
if (flag)
buf_send = (double *)
memory->srealloc(buf_send,(maxsend+BUFEXTRA)*sizeof(double),
"comm:buf_send");
else {
memory->sfree(buf_send);
buf_send = (double *) memory->smalloc((maxsend+BUFEXTRA)*sizeof(double),
"comm:buf_send");
}
}
/* ----------------------------------------------------------------------
free/malloc the size of the recv buffer as needed with BUFFACTOR
------------------------------------------------------------------------- */
void Irregular::grow_recv(int n)
{
maxrecv = static_cast<int> (BUFFACTOR * n);
memory->sfree(buf_recv);
buf_recv = (double *) memory->smalloc(maxrecv*sizeof(double),
"comm:buf_recv");
}

80
src/irregular.h Normal file
View File

@ -0,0 +1,80 @@
/* ----------------------------------------------------------------------
LAMMPS - Large-scale Atomic/Molecular Massively Parallel Simulator
http://lammps.sandia.gov, Sandia National Laboratories
Steve Plimpton, sjplimp@sandia.gov
Copyright (2003) Sandia Corporation. Under the terms of Contract
DE-AC04-94AL85000 with Sandia Corporation, the U.S. Government retains
certain rights in this software. This software is distributed under
the GNU General Public License.
See the README file in the top-level LAMMPS directory.
------------------------------------------------------------------------- */
#ifndef LMP_IRREGULAR_H
#define LMP_IRREGULAR_H
#include "pointers.h"
namespace LAMMPS_NS {
class Irregular : protected Pointers {
public:
struct PlanAtom { // plan for irregular communication of atoms
int nsend; // # of messages to send
int nrecv; // # of messages to recv
int sendmax; // # of doubles in largest send message
int *proc_send; // procs to send to
int *length_send; // # of doubles to send to each proc
int *num_send; // # of datums to send to each proc
int *index_send; // list of which datums to send to each proc
int *offset_send; // where each datum starts in send buffer
int *proc_recv; // procs to recv from
int *length_recv; // # of doubles to recv from each proc
MPI_Request *request; // MPI requests for posted recvs
MPI_Status *status; // MPI statuses for WaitAll
};
struct PlanData { // plan for irregular communication of data
int nsend; // # of messages to send
int nrecv; // # of messages to recv
int sendmax; // # of datums in largest send message
int *proc_send; // procs to send to
int *num_send; // # of datums to send to each proc
int *index_send; // list of which datums to send to each proc
int *proc_recv; // procs to recv from
int *num_recv; // # of doubles to recv from each proc
int num_self;
int *index_self;
MPI_Request *request; // MPI requests for posted recvs
MPI_Status *status; // MPI statuses for WaitAll
};
Irregular(class LAMMPS *);
~Irregular();
void migrate_atoms();
struct PlanData *create_data(int, int *, int *);
void exchange_data(PlanData *, char *, int, char *);
void destroy_data(PlanData *);
private:
int me,nprocs;
int triclinic;
int map_style;
int *procgrid;
int ***grid2proc;
int maxsend,maxrecv;
double *buf_send,*buf_recv;
struct PlanAtom *create_atom(int, int *, int *, int *);
void exchange_atom(PlanAtom *, double *, int *, double *);
void destroy_atom(PlanAtom *);
int coord2proc(double *);
void grow_send(int,int); // reallocate send buffer
void grow_recv(int); // free/allocate recv buffer
};
}
#endif