engine.c 69.3 KB
Newer Older
Pedro Gonnet's avatar
Pedro Gonnet committed
1
/*******************************************************************************
2
 * This file is part of SWIFT.
Pedro Gonnet's avatar
Pedro Gonnet committed
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
 * Coypright (c) 2012 Pedro Gonnet (pedro.gonnet@durham.ac.uk)
 * 
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU Lesser General Public License as published
 * by the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 * 
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 * 
 * You should have received a copy of the GNU Lesser General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 * 
 ******************************************************************************/

/* Config parameters. */
#include "../config.h"

/* Some standard headers. */
#include <stdio.h>
#include <stdlib.h>
26
#include <unistd.h>
Pedro Gonnet's avatar
Pedro Gonnet committed
27
28
29
30
31
32
33
34
#include <string.h>
#include <pthread.h>
#include <math.h>
#include <float.h>
#include <limits.h>
#include <omp.h>
#include <sched.h>

35
36
37
38
39
/* MPI headers. */
#ifdef WITH_MPI
    #include <mpi.h>
#endif

40
41
42
43
44
/* METIS headers. */
#ifdef HAVE_METIS
    #include <metis.h>
#endif

Pedro Gonnet's avatar
Pedro Gonnet committed
45
/* Local headers. */
46
#include "const.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
47
#include "cycle.h"
48
#include "atomic.h"
49
#include "timers.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
50
#include "const.h"
51
#include "vector.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
52
53
54
#include "lock.h"
#include "task.h"
#include "part.h"
55
#include "debug.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
56
#include "space.h"
57
#include "cell.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
58
#include "queue.h"
59
#include "scheduler.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
60
61
#include "engine.h"
#include "runner.h"
62
#include "proxy.h"
63
#include "error.h"
Pedro Gonnet's avatar
Pedro Gonnet committed
64

65
66
67
68
69
70
71
#ifdef LEGACY_GADGET2_SPH
#include "runner_iact_legacy.h"
#else
#include "runner_iact.h"
#endif


Pedro Gonnet's avatar
Pedro Gonnet committed
72
73
74
75
/* Convert cell location to ID. */
#define cell_getid( cdim , i , j , k ) ( (int)(k) + (cdim)[2]*( (int)(j) + (cdim)[1]*(int)(i) ) )


76
77
78
79
/** The rank of the engine as a global variable (for messages). */
int engine_rank;


80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
/**
 * @brief Redistribute the particles amongst the nodes accorind
 *      to their cell's node IDs.
 *
 * @param e The #engine.
 */
 
void engine_redistribute ( struct engine *e ) {

#ifdef WITH_MPI

    int i, j, k, cid;
    int nr_nodes = e->nr_nodes, nodeID = e->nodeID;
    struct space *s = e->s;
    int my_cells = 0;
    int *cdim = s->cdim;
    struct cell *cells = s->cells;
97
    int nr_cells = s->nr_cells;
98
99
100
101
102

    /* Start by sorting the particles according to their nodes and
       getting the counts. */
    int *counts, *dest;
    struct part *parts = s->parts;
Pedro Gonnet's avatar
Pedro Gonnet committed
103
    double ih[3];
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
    ih[0] = s->ih[0]; ih[1] = s->ih[1]; ih[2] = s->ih[2];
    if ( ( counts = (int *)malloc( sizeof(int) * nr_nodes * nr_nodes ) ) == NULL ||
         ( dest = (int *)malloc( sizeof(int) * s->nr_parts ) ) == NULL )
        error( "Failed to allocate count and dest buffers." );
    bzero( counts , sizeof(int) * nr_nodes * nr_nodes );
    for ( k = 0 ; k < s->nr_parts ; k++ ) {
        cid = cell_getid( cdim , parts[k].x[0]*ih[0] , parts[k].x[1]*ih[1] , parts[k].x[2]*ih[2] );
        dest[k] = cells[ cid ].nodeID;
        counts[ nodeID*nr_nodes + dest[k] ] += 1;
        }
    parts_sort( s->parts , s->xparts , dest , s->nr_parts , 0 , nr_nodes-1 );
    
    /* Get all the counts from all the nodes. */
    if ( MPI_Allreduce( MPI_IN_PLACE , counts , nr_nodes * nr_nodes , MPI_INT , MPI_SUM , MPI_COMM_WORLD ) != MPI_SUCCESS )
        error( "Failed to allreduce particle transfer counts." );
        
    /* Get the new number of parts for this node, be generous in allocating. */
    int nr_parts = 0;
    for ( k = 0 ; k < nr_nodes ; k++ )
        nr_parts += counts[ k*nr_nodes + nodeID ];
    struct part *parts_new;
    struct xpart *xparts_new, *xparts = s->xparts;
    if ( posix_memalign( (void **)&parts_new , part_align , sizeof(struct part) * nr_parts * 2 ) != 0 ||
         posix_memalign( (void **)&xparts_new , part_align , sizeof(struct xpart) * nr_parts * 2 ) != 0 )
        error( "Failed to allocate new part data." );
        
    /* Emit the sends and recvs for the particle data. */
    MPI_Request *reqs;
    if ( ( reqs = (MPI_Request *)malloc( sizeof(MPI_Request) * 4 * nr_nodes ) ) == NULL )
        error( "Failed to allocate MPI request list." );
    for ( k = 0 ; k < 4*nr_nodes ; k++ )
        reqs[k] = MPI_REQUEST_NULL;
    for ( i = 0 , j = 0 , k = 0 ; k < nr_nodes ; k++ ) {
        if ( k == nodeID && counts[ nodeID*nr_nodes + k ] > 0 ) {
            memcpy( &parts_new[j] , &parts[i] , sizeof(struct part) * counts[ k*nr_nodes + nodeID ] );
            memcpy( &xparts_new[j] , &xparts[i] , sizeof(struct xpart) * counts[ k*nr_nodes + nodeID ] );
            i += counts[ nodeID*nr_nodes + k ];
            j += counts[ k*nr_nodes + nodeID ];
            }
        if ( k != nodeID && counts[ nodeID*nr_nodes + k ] > 0 ) {
Pedro Gonnet's avatar
Pedro Gonnet committed
144
            if ( MPI_Isend( &parts[i] , sizeof(struct part) * counts[ nodeID*nr_nodes + k ] , MPI_BYTE , k , 2*(nodeID*nr_nodes + k) + 0 , MPI_COMM_WORLD , &reqs[4*k] ) != MPI_SUCCESS )
145
                error( "Failed to isend parts to node %i." , k );
Pedro Gonnet's avatar
Pedro Gonnet committed
146
            if ( MPI_Isend( &xparts[i] , sizeof(struct xpart) * counts[ nodeID*nr_nodes + k ] , MPI_BYTE , k , 2*(nodeID*nr_nodes + k) + 1 , MPI_COMM_WORLD , &reqs[4*k+1] ) != MPI_SUCCESS )
147
148
149
150
                error( "Failed to isend xparts to node %i." , k );
            i += counts[ nodeID*nr_nodes + k ];
            }
        if ( k != nodeID && counts[ k*nr_nodes + nodeID ] > 0 ) {
Pedro Gonnet's avatar
Pedro Gonnet committed
151
            if ( MPI_Irecv( &parts_new[j] , sizeof(struct part) * counts[ k*nr_nodes + nodeID ] , MPI_BYTE , k , 2*(k*nr_nodes + nodeID) + 0 , MPI_COMM_WORLD , &reqs[4*k+2] ) != MPI_SUCCESS )
152
                error( "Failed to emit irecv of parts from node %i." , k );
Pedro Gonnet's avatar
Pedro Gonnet committed
153
            if ( MPI_Irecv( &xparts_new[j] , sizeof(struct xpart) * counts[ k*nr_nodes + nodeID ] , MPI_BYTE , k , 2*(k*nr_nodes + nodeID) + 1 , MPI_COMM_WORLD , &reqs[4*k+3] ) != MPI_SUCCESS )
154
155
156
157
158
                error( "Failed to emit irecv of parts from node %i." , k );
            j += counts[ k*nr_nodes + nodeID ];
            }
        }
        
Pedro Gonnet's avatar
Pedro Gonnet committed
159
    /* Wait for all the sends and recvs to tumble in. */
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
    if ( MPI_Waitall( 4*nr_nodes , reqs , MPI_STATUSES_IGNORE ) != MPI_SUCCESS )
        error( "Failed during waitall for part data." );
        
    /* Verify that all parts are in the right place. */
    /* for ( k = 0 ; k < nr_parts ; k++ ) {
        cid = cell_getid( cdim , parts_new[k].x[0]*ih[0] , parts_new[k].x[1]*ih[1] , parts_new[k].x[2]*ih[2] );
        if ( cells[ cid ].nodeID != nodeID )
            error( "Received particle (%i) that does not belong here (nodeID=%i)." , k , cells[ cid ].nodeID );
        } */
        
    /* Set the new part data, free the old. */
    free( parts );
    free( xparts );
    s->parts = parts_new;
    s->xparts = xparts_new;
    s->nr_parts = nr_parts;
    s->size_parts = 2*nr_parts;
    
    /* Be verbose about what just happened. */
179
180
181
    for ( k = 0 ; k < nr_cells ; k++ )
        if ( cells[k].nodeID == nodeID )
            my_cells += 1;
182
183
184
185
186
187
188
189
190
191
192
193
194
195
    message( "node %i now has %i parts in %i cells." , nodeID , nr_parts , my_cells );
    
    /* Clean up other stuff. */
    free( reqs );
    free( counts );
    free( dest );
        
#else
    error( "SWIFT was not compiled with MPI and METIS support." );
#endif

    }


196
197
198
199
200
201
202
203
204
205
/**
 * @breif Repartition the cells amongst the nodes.
 *
 * @param e The #engine.
 */
 
void engine_repartition ( struct engine *e ) {

#if defined(WITH_MPI) && defined(HAVE_METIS)

206
    int i, j, k, l, cid, cjd, ii, jj, kk, res, w;
207
208
209
    idx_t *inds;
    idx_t *weights_v, *weights_e;
    struct space *s = e->s;
210
    int nr_cells = s->nr_cells, my_cells = 0;
211
212
213
214
215
    struct cell *cells = s->cells;
    int ind[3], *cdim = s->cdim;
    struct task *t, *tasks = e->sched.tasks;
    struct cell *ci, *cj;
    int nr_nodes = e->nr_nodes, nodeID = e->nodeID, *nodeIDs;
216
    float wscale = 0.01;
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
    
    /* Clear the repartition flag. */
    e->forcerepart = 0;
    
    /* Allocate the inds and weights. */
    if ( ( inds = (idx_t *)malloc( sizeof(idx_t) * 26*nr_cells ) ) == NULL ||
         ( weights_v = (idx_t *)malloc( sizeof(idx_t) * nr_cells ) ) == NULL ||
         ( weights_e = (idx_t *)malloc( sizeof(idx_t) * 26*nr_cells ) ) == NULL ||
         ( nodeIDs = (idx_t *)malloc( sizeof(idx_t) * nr_cells ) ) == NULL )
        error( "Failed to allocate inds and weights arrays." );
        
    /* Fill the inds array. */
    for ( cid = 0 ; cid < nr_cells ; cid++ ) {
        ind[0] = cells[cid].loc[0] / s->cells[cid].h[0] + 0.5;
        ind[1] = cells[cid].loc[1] / s->cells[cid].h[1] + 0.5;
        ind[2] = cells[cid].loc[2] / s->cells[cid].h[2] + 0.5;
        l = 0;
        for ( i = -1 ; i <= 1 ; i++ ) {
            ii = ind[0] + i;
            if ( ii < 0 ) ii += cdim[0];
            else if ( ii >= cdim[0] ) ii -= cdim[0];
            for ( j = -1 ; j <= 1 ; j++ ) {
                jj = ind[1] + j;
                if ( jj < 0 ) jj += cdim[1];
                else if ( jj >= cdim[1] ) jj -= cdim[1];
                for ( k = -1 ; k <= 1 ; k++ ) {
                    kk = ind[2] + k;
                    if ( kk < 0 ) kk += cdim[2];
                    else if ( kk >= cdim[2] ) kk -= cdim[2];
                    if ( i || j || k ) {
                        inds[ cid*26 + l ] = cell_getid( cdim , ii , jj , kk );
                        l += 1;
                        }
                    }
                }
            }
        }
        
    /* Init the weights arrays. */
256
257
    /* bzero( weights_e , sizeof(idx_t) * 26*nr_cells );
    bzero( weights_v , sizeof(idx_t) * nr_cells ); */
Pedro Gonnet's avatar
typo.    
Pedro Gonnet committed
258
    for ( k = 0 ; k < 26*nr_cells ; k++ )
259
        weights_e[k] = 1;
Pedro Gonnet's avatar
typo.    
Pedro Gonnet committed
260
    for ( k = 0 ; k < nr_cells ; k++ )
261
        weights_v[k] = 1;
262
263
264
265
266
267
268
269
270
271
    
    /* Loop over the tasks... */
    for ( j = 0 ; j < e->sched.nr_tasks ; j++ ) {
    
        /* Get a pointer to the kth task. */
        t = &tasks[j];
        
        /* Skip un-interesting tasks. */
        if ( t->type != task_type_self &&
             t->type != task_type_pair &&
272
273
274
275
             t->type != task_type_sub &&
             t->type != task_type_ghost &&
             t->type != task_type_kick1 &&
             t->type != task_type_kick2 )
276
            continue;
277
278
279
280
281
            
        /* Get the task weight. */
        w = ( t->toc - t->tic ) * wscale;
        if ( w < 0 )
            error( "Bad task weight (%i)." , w );
282
283
284
285
286
287
288
289
290
291
292
293
        
        /* Get the top-level cells involved. */
        for ( ci = t->ci ; ci->parent != NULL ; ci = ci->parent );
        if ( t->cj != NULL )
            for ( cj = t->cj ; cj->parent != NULL ; cj = cj->parent );
        else
            cj = NULL;
            
        /* Get the cell IDs. */
        cid = ci - cells;
            
        /* Different weights for different tasks. */
294
295
296
297
298
299
300
301
302
303
304
        if ( t->type == task_type_ghost ||
             t->type == task_type_kick1 || 
             t->type == task_type_kick2 ) {
             
            /* Particle updates add only to vertex weight. */
            weights_v[cid] += w;
            
            }
        
        /* Self interaction? */     
        else if ( ( t->type == task_type_self && ci->nodeID == nodeID ) ||
305
306
307
             ( t->type == task_type_sub && cj == NULL && ci->nodeID == nodeID ) ) {
        
            /* Self interactions add only to vertex weight. */
308
            weights_v[cid] += w;
309
310
311
312
313
314
315
316
317
318
319
            
            }
            
        /* Pair? */
        else if ( t->type == task_type_pair ||
                  ( t->type == task_type_sub && cj != NULL ) ) {
                  
            /* In-cell pair? */
            if ( ci == cj ) {
            
                /* Add weight to vertex for ci. */
320
                weights_v[cid] += w;
321
322
323
324
325
326
327
328
329
330
331
            
                }
                
            /* Distinct cells with local ci? */
            else if ( ci->nodeID == nodeID ) {
            
                /* Index of the jth cell. */
                cjd = cj - cells;
                
                /* Add half of weight to each cell. */
                if ( ci->nodeID == nodeID )
332
                    weights_v[cid] += 0.5 * w;
333
                if ( cj->nodeID == nodeID )
334
                    weights_v[cjd] += 0.5 * w;
335
336
337
                    
                /* Add Weight to edge. */
                for ( k = 26*cid ; inds[k] != cjd ; k++ );
338
                weights_e[ k ] += w;
339
                for ( k = 26*cjd ; inds[k] != cid ; k++ );
340
                weights_e[ k ] += w;
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
            
                }
                  
            }
    
        }
        
    /* Merge the weights arrays accross all nodes. */
    if ( ( res = MPI_Reduce( ( nodeID == 0 ) ? MPI_IN_PLACE : weights_v , weights_v , nr_cells , MPI_INT , MPI_SUM , 0 , MPI_COMM_WORLD ) ) != MPI_SUCCESS ) {
        char buff[ MPI_MAX_ERROR_STRING ];
        MPI_Error_string( res , buff , &i );
        error( "Failed to allreduce vertex weights (%s)." , buff );
        }
    if ( MPI_Reduce( ( nodeID == 0 ) ? MPI_IN_PLACE : weights_e , weights_e , 26*nr_cells , MPI_INT , MPI_SUM , 0 , MPI_COMM_WORLD ) != MPI_SUCCESS )
        error( "Failed to allreduce edge weights." );
        
    /* As of here, only one node needs to compute the partition. */
    if ( nodeID == 0 ) {
    
360
        /* Check that the edge weights are fully symmetric. */
361
        /* for ( cid = 0 ; cid < nr_cells ; cid++ )
362
363
364
365
366
            for ( k = 0 ; k < 26 ; k++ ) {
                cjd = inds[ cid*26 + k ];
                for ( j = 26*cjd ; inds[j] != cid ; j++ );
                if ( weights_e[ cid*26+k ] != weights_e[ j ] )
                    error( "Unsymmetric edge weights detected (%i vs %i)." , weights_e[ cid*26+k ] , weights_e[ j ] );
367
                } */
Pedro Gonnet's avatar
Pedro Gonnet committed
368
        /* int w_min = weights_e[0], w_max = weights_e[0];
369
370
371
372
373
374
375
376
377
378
379
380
        for ( k = 1 ; k < 26*nr_cells ; k++ )
            if ( weights_e[k] < w_min )
                w_min = weights_e[k];
            else if ( weights_e[k] > w_max )
                w_max = weights_e[k];
        message( "edge weights in [ %i , %i ]." , w_min , w_max );
        w_min = weights_e[0], w_max = weights_e[0];
        for ( k = 1 ; k < nr_cells ; k++ )
            if ( weights_v[k] < w_min )
                w_min = weights_v[k];
            else if ( weights_v[k] > w_max )
                w_max = weights_v[k];
Pedro Gonnet's avatar
Pedro Gonnet committed
381
        message( "vertex weights in [ %i , %i ]." , w_min , w_max ); */
382
383
384
385
                
        /* Make sure there are no zero weights. */
        for ( k = 0 ; k < 26*nr_cells ; k++ )
            if ( weights_e[k] == 0 )
Pedro Gonnet's avatar
typo.    
Pedro Gonnet committed
386
                weights_e[k] = 1;
387
    
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
        /* Allocate and fill the connection array. */
        idx_t *offsets;
        if ( ( offsets = (idx_t *)malloc( sizeof(idx_t) * (nr_cells + 1) ) ) == NULL )
            error( "Failed to allocate offsets buffer." );
        offsets[0] = 0;
        for ( k = 0 ; k < nr_cells ; k++ )
            offsets[k+1] = offsets[k] + 26;
            
        /* Set the METIS options. */
        idx_t options[METIS_NOPTIONS];
        METIS_SetDefaultOptions( options );
        options[ METIS_OPTION_OBJTYPE ] = METIS_OBJTYPE_CUT;
        options[ METIS_OPTION_NUMBERING ] = 0;
        options[ METIS_OPTION_CONTIG ] = 1;
        
403
404
405
406
        /* Set the initial partition, although this is probably ignored. */
        for ( k = 0 ; k < nr_cells ; k++ )
            nodeIDs[k] = cells[k].nodeID;
            
407
408
409
410
411
412
413
414
415
416
417
418
419
        /* Call METIS. */
        int one = 1;
        idx_t objval;
        if ( METIS_PartGraphKway( &nr_cells , &one , offsets , inds , weights_v , NULL , weights_e , &nr_nodes , NULL , NULL , options , &objval , nodeIDs ) != METIS_OK )
            error( "Call to METIS_PartGraphKway failed." );
    
        }
        
    /* Broadcast the result of the partition. */
    if ( MPI_Bcast( nodeIDs , nr_cells , MPI_INT , 0 , MPI_COMM_WORLD ) != MPI_SUCCESS )
        error( "Failed to bcast the node IDs." );
        
    /* Set the cell nodeIDs and clear any non-local parts. */
420
    for ( k = 0 ; k < nr_cells ; k++ ) {
421
        cells[k].nodeID = nodeIDs[k];
422
423
424
        if ( nodeIDs[k] == nodeID )
            my_cells += 1;
        }
425
426
427
428
429
430
431
432
433
434
435
436
437
        
    /* Clean up. */
    free( inds );
    free( weights_v );
    free( weights_e );
    free( nodeIDs );
        
    /* Now comes the tricky part: Exchange particles between all nodes.
       This is done in two steps, first allreducing a matrix of 
       how many particles go from where to where, then re-allocating
       the parts array, and emiting the sends and receives.
       Finally, the space, tasks, and proxies need to be rebuilt. */
       
438
439
    /* Redistribute the particles between the nodes. */
    engine_redistribute( e );
440
441
442
443
444
445
446
447
448
449
450
451
452
453
        
    /* Make the proxies. */
    engine_makeproxies( e );
        
    /* Tell the engine it should re-build whenever possible */
    e->forcerebuild = 1;
    
#else
    error( "SWIFT was not compiled with MPI and METIS support." );
#endif

    }


454
455
456
457
458
459
460
461
462
463
464
/**
 * @brief Add send tasks to a hierarchy of cells.
 *
 * @param e The #engine.
 * @param c The #cell.
 * @param t_xv The send_xv #task, if it has already been created.
 * @param t_rho The send_rho #task, if it has already been created.
 */

void engine_addtasks_send ( struct engine *e , struct cell *ci , struct cell *cj ) {

Pedro Gonnet's avatar
Pedro Gonnet committed
465
    int k;
466
467
468
469
470
471
472
473
474
475
476

    /* Check if any of the density tasks are for the target node. */
    for ( k = 0 ; k < ci->nr_density ; k++ )
        if ( ci->density[k]->ci->nodeID == cj->nodeID ||
             ( ci->density[k]->cj != NULL && ci->density[k]->cj->nodeID == cj->nodeID ) )
            break;

    /* If so, attach send tasks. */
    if ( k < ci->nr_density ) {

        /* Create the tasks. */
Pedro Gonnet's avatar
Pedro Gonnet committed
477
478
        struct task *t_xv = scheduler_addtask( &e->sched , task_type_send_xv , task_subtype_none , 2*ci->tag , 0 , ci , cj , 0 );
        struct task *t_rho = scheduler_addtask( &e->sched , task_type_send_rho , task_subtype_none , 2*ci->tag + 1 , 0 , ci , cj , 0 );
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510

        /* The send_rho task depends on the cell's ghost task. */
        task_addunlock( ci->ghost , t_rho );

        /* The send_rho task should unlock the super-cell's kick2 task. */
        task_addunlock( t_rho , ci->super->kick2 );

        /* The send_xv task should unlock the super-cell's ghost task. */
        task_addunlock( t_xv , ci->super->ghost );

        }
        
    /* Recurse? */
    else if ( ci->split )
        for ( k = 0 ; k < 8 ; k++ )
            if ( ci->progeny[k] != NULL )
                engine_addtasks_send( e , ci->progeny[k] , cj );

    }


/**
 * @brief Add recv tasks to a hierarchy of cells.
 *
 * @param e The #engine.
 * @param c The #cell.
 * @param t_xv The recv_xv #task, if it has already been created.
 * @param t_rho The recv_rho #task, if it has already been created.
 */

void engine_addtasks_recv ( struct engine *e , struct cell *c , struct task *t_xv , struct task *t_rho ) {

Pedro Gonnet's avatar
Pedro Gonnet committed
511
    int k;
512
513
514
515
516

    /* Do we need to construct a recv task? */
    if ( t_xv != NULL || c->nr_density > 0 ) {
    
        /* Create the tasks. */
Pedro Gonnet's avatar
Pedro Gonnet committed
517
518
        c->recv_xv = scheduler_addtask( &e->sched , task_type_recv_xv , task_subtype_none , 2*c->tag , 0 , c , NULL , 0 );
        c->recv_rho = scheduler_addtask( &e->sched , task_type_recv_rho , task_subtype_none , 2*c->tag + 1 , 0 , c , NULL , 0 );
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
        
        /* If there has been a higher-up recv task, then these tasks
           are implicit and depend on the higher-up task. */
        if ( t_xv != NULL ) {
            task_addunlock( c->parent->recv_xv , c->recv_xv );
            task_addunlock( c->parent->recv_rho , c->recv_rho );
            c->recv_xv->implicit = 1;
            c->recv_rho->implicit = 1;
            }
        else {
            t_xv = c->recv_xv;
            t_rho = c->recv_rho;
            }
        
        /* Add dependencies if there are density/force tasks. */
        for ( k = 0 ; k < c->nr_density ; k++ ) {
            task_addunlock( c->recv_xv , c->density[k] );
            task_addunlock( c->density[k] , t_rho );
            }
        for ( k = 0 ; k < c->nr_force ; k++ )
            task_addunlock( c->recv_rho , c->force[k] );
        if ( c->sorts != NULL )
            task_addunlock( c->recv_xv , c->sorts );
            
        }
        
    /* Recurse? */
    if ( c->split )
        for ( k = 0 ; k < 8 ; k++ )
            if ( c->progeny[k] != NULL )
                engine_addtasks_recv( e , c->progeny[k] , t_xv , t_rho );

    }


/**
 * @brief Exchange cell structures with other nodes.
 *
 * @param e The #engine.
 */
 
void engine_exchange_cells ( struct engine *e ) {

#ifdef WITH_MPI

    int j, k, pid, count = 0;
    struct pcell *pcells;
566
567
568
569
    struct space *s = e->s;
    struct cell *cells = s->cells;
    int nr_cells = s->nr_cells;
    int nr_proxies = e->nr_proxies;
570
    int offset[ nr_cells ];
571
572
    MPI_Request reqs_in[ engine_maxproxies ];
    MPI_Request reqs_out[ engine_maxproxies ];
573
    MPI_Status status;
574
    struct part *parts = &s->parts[ s->nr_parts ];
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
    
    /* Run through the cells and get the size of the ones that will be sent off. */
    for ( k = 0 ; k < nr_cells ; k++ ) {
        offset[k] = count;
        if ( cells[k].sendto )
            count += ( cells[k].pcell_size = cell_getsize( &cells[k] ) );
        }
        
    /* Allocate the pcells. */
    if ( ( pcells = (struct pcell *)malloc( sizeof(struct pcell) * count ) ) == NULL )
        error( "Failed to allocate pcell buffer." );
        
    /* Pack the cells. */
    for ( k = 0 ; k < nr_cells ; k++ )
        if ( cells[k].sendto ) {
            cell_pack( &cells[k] , &pcells[ offset[k] ] );
            cells[k].pcell = &pcells[ offset[k] ];
            }

    /* Launch the proxies. */
595
    for ( k = 0 ; k < nr_proxies ; k++ ) {
596
        proxy_cells_exch1( &e->proxies[k] );
597
        reqs_in[k] = e->proxies[k].req_cells_count_in;
Pedro Gonnet's avatar
Pedro Gonnet committed
598
        reqs_out[k] = e->proxies[k].req_cells_count_out;
599
600
601
        }
        
    /* Wait for each count to come in and start the recv. */
602
603
    for ( k = 0 ; k < nr_proxies ; k++ ) {
        if ( MPI_Waitany( nr_proxies , reqs_in , &pid , &status ) != MPI_SUCCESS ||
604
605
606
607
608
609
             pid == MPI_UNDEFINED )
            error( "MPI_Waitany failed." );
        // message( "request from proxy %i has arrived." , pid );
        proxy_cells_exch2( &e->proxies[pid] );
        }
        
Pedro Gonnet's avatar
Pedro Gonnet committed
610
611
612
613
    /* Wait for all the sends to have finnished too. */
    if ( MPI_Waitall( nr_proxies , reqs_out , MPI_STATUSES_IGNORE ) != MPI_SUCCESS )
        error( "MPI_Waitall on sends failed." );
        
614
    /* Set the requests for the cells. */
615
616
617
618
    for ( k = 0 ; k < nr_proxies ; k++ ) {
        reqs_in[k] = e->proxies[k].req_cells_in;
        reqs_out[k] = e->proxies[k].req_cells_out;
        }
619
620
    
    /* Wait for each pcell array to come in from the proxies. */
621
622
    for ( k = 0 ; k < nr_proxies ; k++ ) {
        if ( MPI_Waitany( nr_proxies , reqs_in , &pid , &status ) != MPI_SUCCESS ||
623
624
             pid == MPI_UNDEFINED )
            error( "MPI_Waitany failed." );
625
626
627
628
629
630
        // message( "cell data from proxy %i has arrived." , pid );
        for ( count = 0 , j = 0 ; j < e->proxies[pid].nr_cells_in ; j++ )
            count += cell_unpack( &e->proxies[pid].pcells_in[count] , e->proxies[pid].cells_in[j] , e->s );
        }
        
    /* Wait for all the sends to have finnished too. */
631
    if ( MPI_Waitall( nr_proxies , reqs_out , MPI_STATUSES_IGNORE ) != MPI_SUCCESS )
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
        error( "MPI_Waitall on sends failed." );
        
    /* Count the number of particles we need to import and re-allocate
       the buffer if needed. */
    for ( count = 0 , k = 0 ; k < nr_proxies ; k++ )
        for ( j = 0 ; j < e->proxies[k].nr_cells_in ; j++ )
            count += e->proxies[k].cells_in[j]->count;
    if ( count > s->size_parts_foreign ) {
        if ( s->parts_foreign != NULL )
            free( s->parts_foreign );
        s->size_parts_foreign = 1.1 * count;
        if ( posix_memalign( (void **)&s->parts_foreign , part_align , sizeof(struct part) * s->size_parts_foreign ) != 0 )
            error( "Failed to allocate foreign part data." );
        }
        
    /* Unpack the cells and link to the particle data. */
    parts = s->parts_foreign;
    for ( k = 0 ; k < nr_proxies ; k++ ) {
        for ( count = 0 , j = 0 ; j < e->proxies[k].nr_cells_in ; j++ ) {
            count += cell_link( e->proxies[k].cells_in[j] , parts );
            parts = &parts[ e->proxies[k].cells_in[j]->count ];
653
654
            }
        }
655
656
657
658
659
    s->nr_parts_foreign = parts - s->parts_foreign;
        
    /* Is the parts buffer large enough? */
    if ( s->nr_parts_foreign > s->size_parts_foreign )
        error( "Foreign parts buffer too small." );
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
        
    /* Free the pcell buffer. */
    free( pcells );
    
#else
    error( "SWIFT was not compiled with MPI support." );
#endif

    }


/**
 * @brief Exchange straying parts with other nodes.
 *
 * @param e The #engine.
 * @param parts An array of straying parts.
 * @param xparts The corresponding xparts.
 * @param ind The ID of the foreign #cell.
 * @param N The number of stray parts.
 *
 * @return The number of arrived parts copied to parts and xparts.
 */
 
int engine_exchange_strays ( struct engine *e , struct part *parts , struct xpart *xparts , int *ind , int N ) {

#ifdef WITH_MPI

687
    int k, pid, count = 0, nr_in = 0, nr_out = 0;
Pedro Gonnet's avatar
Pedro Gonnet committed
688
689
    MPI_Request reqs_in[ 2*engine_maxproxies ];
    MPI_Request reqs_out[ 2*engine_maxproxies ];
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
    MPI_Status status;
    struct proxy *p;

    /* Re-set the proxies. */
    for ( k = 0 ; k < e->nr_proxies ; k++ )
        e->proxies[k].nr_parts_out = 0;
    
    /* Put the parts into the corresponding proxies. */
    for ( k = 0 ; k < N ; k++ ) {
        pid = e->proxy_ind[ e->s->cells[ ind[k] ].nodeID ];
        if ( pid < 0 )
            error( "Do not have a proxy for the requested nodeID." );
        proxy_parts_load( &e->proxies[pid] , &parts[k] , &xparts[k] , 1 );
        }
    
    /* Launch the proxies. */
    for ( k = 0 ; k < e->nr_proxies ; k++ ) {
        proxy_parts_exch1( &e->proxies[k] );
708
        reqs_in[k] = e->proxies[k].req_parts_count_in;
Pedro Gonnet's avatar
Pedro Gonnet committed
709
        reqs_out[k] = e->proxies[k].req_parts_count_out;
710
711
712
713
        }
        
    /* Wait for each count to come in and start the recv. */
    for ( k = 0 ; k < e->nr_proxies ; k++ ) {
714
        if ( MPI_Waitany( e->nr_proxies , reqs_in , &pid , &status ) != MPI_SUCCESS ||
715
716
717
718
719
720
             pid == MPI_UNDEFINED )
            error( "MPI_Waitany failed." );
        // message( "request from proxy %i has arrived." , pid );
        proxy_parts_exch2( &e->proxies[pid] );
        }
        
Pedro Gonnet's avatar
Pedro Gonnet committed
721
722
723
724
725
    /* Wait for all the sends to have finnished too. */
    if ( MPI_Waitall( e->nr_proxies , reqs_out , MPI_STATUSES_IGNORE ) != MPI_SUCCESS )
        error( "MPI_Waitall on sends failed." );
        
    /* Return the number of harvested parts. */
726
    /* Set the requests for the particle data. */
727
    for ( k = 0 ; k < e->nr_proxies ; k++ ) {
728
        if ( e->proxies[k].nr_parts_in > 0 ) {
Pedro Gonnet's avatar
Pedro Gonnet committed
729
730
            reqs_in[2*k] = e->proxies[k].req_parts_in;
            reqs_in[2*k+1] = e->proxies[k].req_xparts_in;
731
732
733
            nr_in += 1;
            }
        else
Pedro Gonnet's avatar
Pedro Gonnet committed
734
            reqs_in[2*k] = reqs_in[2*k+1] = MPI_REQUEST_NULL;
735
        if ( e->proxies[k].nr_parts_out > 0 ) {
Pedro Gonnet's avatar
Pedro Gonnet committed
736
737
            reqs_out[2*k] = e->proxies[k].req_parts_out;
            reqs_out[2*k+1] = e->proxies[k].req_xparts_out;
738
739
740
            nr_out += 1;
            }
        else
Pedro Gonnet's avatar
Pedro Gonnet committed
741
            reqs_out[2*k] = reqs_out[2*k+1] = MPI_REQUEST_NULL;
742
        }
743
744
745
    
    /* Wait for each part array to come in and collect the new
       parts from the proxies. */
746
    for ( k = 0 ; k < 2*(nr_in + nr_out) ; k++ ) {
747
        if ( MPI_Waitany( 2*e->nr_proxies , reqs_in , &pid , &status ) != MPI_SUCCESS )
748
            error( "MPI_Waitany failed." );
749
750
        if ( pid == MPI_UNDEFINED )
            break;
751
        // message( "request from proxy %i has arrived." , pid );
Pedro Gonnet's avatar
Pedro Gonnet committed
752
753
754
755
756
757
758
759
760
761
762
        if ( reqs_in[pid & ~1] == MPI_REQUEST_NULL &&
             reqs_in[pid | 1 ] == MPI_REQUEST_NULL ) {
            p = &e->proxies[pid/2];
            memcpy( &parts[count] , p->parts_in , sizeof(struct part) * p->nr_parts_in );
            memcpy( &xparts[count] , p->xparts_in , sizeof(struct xpart) * p->nr_parts_in );
            count += p->nr_parts_in;
            /* for ( int k = 0 ; k < p->nr_parts_in ; k++ )
                message( "received particle %lli, x=[%.3e %.3e %.3e], h=%.3e, from node %i." ,
                    p->parts_in[k].id , p->parts_in[k].x[0] , p->parts_in[k].x[1] , p->parts_in[k].x[2] ,
                    p->parts_in[k].h , p->nodeID ); */
            }
763
764
        }
    
765
    /* Wait for all the sends to have finnished too. */
766
    if ( nr_out > 0 )
Pedro Gonnet's avatar
Pedro Gonnet committed
767
        if ( MPI_Waitall( 2*e->nr_proxies , reqs_out , MPI_STATUSES_IGNORE ) != MPI_SUCCESS )
768
            error( "MPI_Waitall on sends failed." );
769
        
770
771
772
773
774
775
776
777
778
779
780
    /* Return the number of harvested parts. */
    return count;
    
#else
    error( "SWIFT was not compiled with MPI support." );
    return 0;
#endif

    }


781
/**
782
 * @brief Fill the #space's task list.
783
 *
784
 * @param s The #space we are working in.
785
786
 */
 
787
void engine_maketasks ( struct engine *e ) {
788
789

    struct space *s = e->s;
790
791
792
793
794
795
796
    struct scheduler *sched = &e->sched;
    int i, j, k, ii, jj, kk, iii, jjj, kkk, cid, cjd, sid;
    int *cdim = s->cdim;
    struct task *t, *t2;
    struct cell *ci, *cj;

    /* Re-set the scheduler. */
797
    scheduler_reset( sched , s->tot_cells * engine_maxtaskspercell );
798
799
800
801
802
803
804
805
806
807
808
    
    /* Run through the highest level of cells and add pairs. */
    for ( i = 0 ; i < cdim[0] ; i++ )
        for ( j = 0 ; j < cdim[1] ; j++ )
            for ( k = 0 ; k < cdim[2] ; k++ ) {
                cid = cell_getid( cdim , i , j , k );
                if ( s->cells[cid].count == 0 )
                    continue;
                ci = &s->cells[cid];
                if ( ci->count == 0 )
                    continue;
809
810
                if ( ci->nodeID == e->nodeID )
                    scheduler_addtask( sched , task_type_self , task_subtype_density , 0 , 0 , ci , NULL , 0 );
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
                for ( ii = -1 ; ii < 2 ; ii++ ) {
                    iii = i + ii;
                    if ( !s->periodic && ( iii < 0 || iii >= cdim[0] ) )
                        continue;
                    iii = ( iii + cdim[0] ) % cdim[0];
                    for ( jj = -1 ; jj < 2 ; jj++ ) {
                        jjj = j + jj;
                        if ( !s->periodic && ( jjj < 0 || jjj >= cdim[1] ) )
                            continue;
                        jjj = ( jjj + cdim[1] ) % cdim[1];
                        for ( kk = -1 ; kk < 2 ; kk++ ) {
                            kkk = k + kk;
                            if ( !s->periodic && ( kkk < 0 || kkk >= cdim[2] ) )
                                continue;
                            kkk = ( kkk + cdim[2] ) % cdim[2];
                            cjd = cell_getid( cdim , iii , jjj , kkk );
                            cj = &s->cells[cjd];
828
829
                            if ( cid >= cjd || cj->count == 0 || 
                                 ( ci->nodeID != e->nodeID && cj->nodeID != e->nodeID ) )
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
                                continue;
                            sid = sortlistID[ (kk+1) + 3*( (jj+1) + 3*(ii+1) ) ];
                            t = scheduler_addtask( sched , task_type_pair , task_subtype_density , sid , 0 , ci , cj , 1 );
                            }
                        }
                    }
                }

    /* Split the tasks. */
    scheduler_splittasks( sched );
    
    /* Count the number of tasks associated with each cell and
       store the density tasks in each cell, and make each sort
       depend on the sorts of its progeny. */
    // #pragma omp parallel for private(t,j)
    for ( k = 0 ; k < sched->nr_tasks ; k++ ) {
        t = &sched->tasks[k];
        if ( t->skip )
            continue;
        if ( t->type == task_type_sort && t->ci->split )
850
851
            for ( j = 0 ; j < 8 ; j++ )
                if ( t->ci->progeny[j] != NULL && t->ci->progeny[j]->sorts != NULL ) {
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
                    t->ci->progeny[j]->sorts->skip = 0;
                    task_addunlock( t->ci->progeny[j]->sorts , t );
                    }
        if ( t->type == task_type_self ) {
            atomic_inc( &t->ci->nr_tasks );
            if ( t->subtype == task_subtype_density ) {
                t->ci->density[ atomic_inc( &t->ci->nr_density ) ] = t;
                }
            }
        else if ( t->type == task_type_pair ) {
            atomic_inc( &t->ci->nr_tasks );
            atomic_inc( &t->cj->nr_tasks );
            if ( t->subtype == task_subtype_density ) {
                t->ci->density[ atomic_inc( &t->ci->nr_density ) ] = t;
                t->cj->density[ atomic_inc( &t->cj->nr_density ) ] = t;
                }
            }
        else if ( t->type == task_type_sub ) {
            atomic_inc( &t->ci->nr_tasks );
            if ( t->cj != NULL )
                atomic_inc( &t->cj->nr_tasks );
            if ( t->subtype == task_subtype_density ) {
                t->ci->density[ atomic_inc( &t->ci->nr_density ) ] = t;
                if ( t->cj != NULL )
                    t->cj->density[ atomic_inc( &t->cj->nr_density ) ] = t;
                }
            }
        }
        
    /* Append a ghost task to each cell. */
882
883
884
885
    if ( e->policy & engine_policy_fixdt )
        space_map_cells_pre( s , 1 , &scheduler_map_mkghosts_nokick1 , sched );
    else
        space_map_cells_pre( s , 1 , &scheduler_map_mkghosts , sched );
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
    
    /* Run through the tasks and make force tasks for each density task.
       Each force task depends on the cell ghosts and unlocks the kick2 task
       of its super-cell. */
    kk = sched->nr_tasks;
    // #pragma omp parallel for private(t,t2)
    for ( k = 0 ; k < kk ; k++ ) {
    
        /* Get a pointer to the task. */
        t = &sched->tasks[k];
        
        /* Skip? */
        if ( t->skip )
            continue;
        
        /* Self-interaction? */
        if ( t->type == task_type_self && t->subtype == task_subtype_density ) {
            task_addunlock( t , t->ci->super->ghost );
            t2 = scheduler_addtask( sched , task_type_self , task_subtype_force , 0 , 0 , t->ci , NULL , 0 );
            task_addunlock( t->ci->ghost , t2 );
            task_addunlock( t2 , t->ci->super->kick2 );
907
            t->ci->force[ atomic_inc( &t->ci->nr_force ) ] = t2;
908
909
910
911
912
            }
            
        /* Otherwise, pair interaction? */
        else if ( t->type == task_type_pair && t->subtype == task_subtype_density ) {
            t2 = scheduler_addtask( sched , task_type_pair , task_subtype_force , 0 , 0 , t->ci , t->cj , 0 );
913
914
915
916
917
918
919
920
921
922
923
924
925
926
            if ( t->ci->nodeID == e->nodeID ) {
                task_addunlock( t->ci->ghost , t2 );
                task_addunlock( t , t->ci->super->ghost );
                task_addunlock( t2 , t->ci->super->kick2 );
                }
            if ( t->cj->nodeID == e->nodeID ) {
                task_addunlock( t->cj->ghost , t2 );
                if ( t->ci->super != t->cj->super ) {
                    task_addunlock( t , t->cj->super->ghost );
                    task_addunlock( t2 , t->cj->super->kick2 );
                    }
                }
            t->ci->force[ atomic_inc( &t->ci->nr_force ) ] = t2;
            t->cj->force[ atomic_inc( &t->cj->nr_force ) ] = t2;
927
928
929
930
931
            }
    
        /* Otherwise, sub interaction? */
        else if ( t->type == task_type_sub && t->subtype == task_subtype_density ) {
            t2 = scheduler_addtask( sched , task_type_sub , task_subtype_force , t->flags , 0 , t->ci , t->cj , 0 );
932
933
934
935
936
937
            if ( t->ci->nodeID == e->nodeID ) {
                task_addunlock( t , t->ci->super->ghost );
                task_addunlock( t->ci->ghost , t2 );
                task_addunlock( t2 , t->ci->super->kick2 );
                }
            if ( t->cj != NULL && t->cj->nodeID == e->nodeID ) {
938
                task_addunlock( t->cj->ghost , t2 );
939
940
941
942
943
944
945
946
                if ( t->ci->super != t->cj->super ) {
                    task_addunlock( t , t->cj->super->ghost );
                    task_addunlock( t2 , t->cj->super->kick2 );
                    }
                }
            t->ci->force[ atomic_inc( &t->ci->nr_force ) ] = t2;
            if ( t->cj != NULL )
                t->cj->force[ atomic_inc( &t->cj->nr_force ) ] = t2;
947
948
949
950
            }
            
        }
        
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
    /* Add the communication tasks if MPI is being used. */
    #ifdef WITH_MPI
        
        /* Loop over the proxies. */
        for ( int pid = 0 ; pid < e->nr_proxies ; pid++ ) {
        
            /* Get a handle on the proxy. */
            struct proxy *p = &e->proxies[pid];
            
            /* Loop through the proxy's incomming cells and add the
               recv tasks. */
            for ( k = 0 ; k < p->nr_cells_in ; k++ )
                engine_addtasks_recv( e , p->cells_in[k] , NULL , NULL );
            
            /* Loop through the proxy's outgoing cells and add the
               send tasks. */
967
            for ( k = 0 ; k < p->nr_cells_out ; k++ )
968
969
970
971
972
973
                engine_addtasks_send( e , p->cells_out[k] , p->cells_in[0] );
            
            }
        
    #endif
        
974
975
    /* Rank the tasks. */
    scheduler_ranktasks( sched );
Pedro Gonnet's avatar
Pedro Gonnet committed
976
    
977
978
979
    /* Weight the tasks. */
    scheduler_reweight( sched );
            
Pedro Gonnet's avatar
Pedro Gonnet committed
980
981
982
    /* Set the tasks age. */
    e->tasks_age = 0;
            
983
984
    }
    
985
    
986

987
988
989
990
991
992
993
994
995
996
997
998
999
/**
 * @brief Mark tasks to be skipped and set the sort flags accordingly.
 * 
 * @return 1 if the space has to be rebuilt, 0 otherwise.
 */
 
int engine_marktasks ( struct engine *e ) {

    struct scheduler *s = &e->sched;
    int k, nr_tasks = s->nr_tasks, *ind = s->tasks_ind;
    struct task *t, *tasks = s->tasks;
    float dt_step = e->dt_step;
    struct cell *ci, *cj;
1000
    // ticks tic = getticks();
For faster browsing, not all history is shown. View entire blame