Also added and updated the mdrun PME load balacing reporting.
Change-Id: Ifa633ee9e1fc6a22f7792bbb6af7ad7da6192d14
void nbnxn_cuda_init_atomdata(nbnxn_cuda_ptr_t cu_nb,
const nbnxn_atomdata_t *atomdata) FUNC_TERM
-/*! \brief Update parameters during PME auto-tuning. */
+/*! \brief Update parameters during PP-PME load balancing. */
FUNC_QUALIFIER
-void nbnxn_cuda_pmetune_update_param(nbnxn_cuda_ptr_t cu_nb,
- const interaction_const_t *ic) FUNC_TERM
+void nbnxn_cuda_pme_loadbal_update_param(nbnxn_cuda_ptr_t cu_nb,
+ const interaction_const_t *ic) FUNC_TERM
/*! Uploads shift vector to the GPU if the box is dynamic (otherwise just returns). */
FUNC_QUALIFIER
set(MDRUN_SOURCES
gctio.c ionize.c runner.c
- do_gct.c repl_ex.c xutils.c pme_switch.c
+ do_gct.c repl_ex.c xutils.c pme_loadbal.c
md.c mdrun.c genalg.c membed.c
md_openmm.c)
#include "sighandler.h"
#include "txtdump.h"
#include "string2.h"
-#include "pme_switch.h"
+#include "pme_loadbal.h"
#include "bondf.h"
#include "membed.h"
#include "types/nlistheuristics.h"
simulation stops. If equal to zero, don't
communicate any more between multisims.*/
/* PME load balancing data for GPU kernels */
- pme_switch_t pme_switch=NULL;
+ pme_load_balancing_t pme_loadbal=NULL;
double cycles_pmes;
gmx_bool bPMETuneTry=FALSE,bPMETuneRunning=FALSE;
(fr->nbv->bUseGPU || !(cr->duty & DUTY_PME)) &&
!bRerunMD)
{
- switch_pme_init(&pme_switch,ir,state->box,fr->ic,fr->pmedata);
+ pme_loadbal_init(&pme_loadbal,ir,state->box,fr->ic,fr->pmedata);
cycles_pmes = 0;
if (cr->duty & DUTY_PME)
{
* but the first cycle is always skipped anyhow.
*/
bPMETuneRunning =
- switch_pme(pme_switch,cr,
- (bVerbose && MASTER(cr)) ? stderr : NULL,
- fplog,
- ir,state,cycles_pmes,
- fr->ic,fr->nbv,&fr->pmedata,
- step);
+ pme_load_balance(pme_loadbal,cr,
+ (bVerbose && MASTER(cr)) ? stderr : NULL,
+ fplog,
+ ir,state,cycles_pmes,
+ fr->ic,fr->nbv,&fr->pmedata,
+ step);
fr->ewaldcoeff = fr->ic->ewaldcoeff;
}
fprintf(fplog,"Average neighborlist lifetime: %.1f steps, std.dev.: %.1f steps\n",nlh.s1/nlh.nns,sqrt(nlh.s2/nlh.nns - sqr(nlh.s1/nlh.nns)));
fprintf(fplog,"Average number of atoms that crossed the half buffer length: %.1f\n\n",nlh.ab/nlh.nns);
}
-
+
+ if (pme_loadbal != NULL)
+ {
+ pme_loadbal_done(pme_loadbal,fplog);
+ }
+
if (shellfc && fplog)
{
fprintf(fplog,"Fraction of iterations that converged: %.2f %%\n",
--- /dev/null
+/* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
+ *
+ *
+ * This source code is part of
+ *
+ * G R O M A C S
+ *
+ * GROningen MAchine for Chemical Simulations
+ *
+ * VERSION 4.6.0
+ * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
+ * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
+ * Copyright (c) 2001-2011, The GROMACS development team,
+ * check out http://www.gromacs.org for more information.
+
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * If you want to redistribute modifications, please consider that
+ * scientific software is very special. Version control is crucial -
+ * bugs must be traceable. We will be happy to consider code for
+ * inclusion in the official distribution, but derived work must not
+ * be called official GROMACS. Details are found in the README & COPYING
+ * files - if they are missing, get the official version at www.gromacs.org.
+ *
+ * To help us fund GROMACS development, we humbly ask that you cite
+ * the papers on the package - you can find them in the top README file.
+ *
+ * For more info, check our website at http://www.gromacs.org
+ *
+ * And Hey:
+ * Gallium Rubidium Oxygen Manganese Argon Carbon Silicon
+ */
+#ifdef HAVE_CONFIG_H
+#include <config.h>
+#endif
+
+#include "smalloc.h"
+#include "network.h"
+#include "calcgrid.h"
+#include "pme.h"
+#include "vec.h"
+#include "domdec.h"
+#include "nbnxn_cuda_data_mgmt.h"
+#include "force.h"
+#include "pme_loadbal.h"
+
+/* Parameters and setting for one PP-PME setup */
+typedef struct {
+ real rcut; /* Coulomb cut-off */
+ real rlist; /* pair-list cut-off */
+ real spacing; /* (largest) PME grid spacing */
+ ivec grid; /* the PME grid dimensions */
+ real grid_efficiency; /* ineffiency factor for non-uniform grids <= 1 */
+ real ewaldcoeff; /* the Ewald coefficient */
+ gmx_pme_t pmedata; /* the data structure used in the PME code */
+
+ int count; /* number of times this setup has been timed */
+ double cycles; /* the fastest time for this setup in cycles */
+} pme_setup_t;
+
+/* In the initial scan, step by grids that are at least a factor 0.8 coarser */
+#define PME_LB_GRID_SCALE_FAC 0.8
+/* In the initial scan, try to skip grids with uneven x/y/z spacing,
+ * checking if the "efficiency" is more than 5% worse than the previous grid.
+ */
+#define PME_LB_GRID_EFFICIENCY_REL_FAC 1.05
+/* Rerun up till 12% slower setups than the fastest up till now */
+#define PME_LB_SLOW_FAC 1.12
+/* If setups get more than 2% faster, do another round to avoid
+ * choosing a slower setup due to acceleration or fluctuations.
+ */
+#define PME_LB_ACCEL_TOL 1.02
+
+enum { epmelblimNO, epmelblimBOX, epmelblimDD, epmelblimNR };
+
+const char *pmelblim_str[epmelblimNR] =
+{ "no", "box size", "domain decompostion" };
+
+struct pme_load_balancing {
+ int nstage; /* the current maximum number of stages */
+
+ real cut_spacing; /* the minimum cutoff / PME grid spacing ratio */
+ real rbuf; /* the pairlist buffer size */
+ matrix box_start; /* the initial simulation box */
+ int n; /* the count of setup as well as the allocation size */
+ pme_setup_t *setup; /* the PME+cutoff setups */
+ int cur; /* the current setup */
+ int fastest; /* fastest setup up till now */
+ int start; /* start of setup range to consider in stage>0 */
+ int end; /* end of setup range to consider in stage>0 */
+ int elimited; /* was the balancing limited, uses enum above */
+
+ int stage; /* the current stage */
+};
+
+void pme_loadbal_init(pme_load_balancing_t *pme_lb_p,
+ const t_inputrec *ir,matrix box,
+ const interaction_const_t *ic,
+ gmx_pme_t pmedata)
+{
+ pme_load_balancing_t pme_lb;
+ real spm,sp;
+ int d;
+
+ snew(pme_lb,1);
+
+ /* Any number of stages >= 2 is supported */
+ pme_lb->nstage = 2;
+
+ pme_lb->rbuf = ic->rlist - ic->rcoulomb;
+
+ copy_mat(box,pme_lb->box_start);
+ if (ir->ePBC==epbcXY && ir->nwall==2)
+ {
+ svmul(ir->wall_ewald_zfac,pme_lb->box_start[ZZ],pme_lb->box_start[ZZ]);
+ }
+
+ pme_lb->n = 1;
+ snew(pme_lb->setup,pme_lb->n);
+
+ pme_lb->cur = 0;
+ pme_lb->setup[0].rcut = ic->rcoulomb;
+ pme_lb->setup[0].rlist = ic->rlist;
+ pme_lb->setup[0].grid[XX] = ir->nkx;
+ pme_lb->setup[0].grid[YY] = ir->nky;
+ pme_lb->setup[0].grid[ZZ] = ir->nkz;
+ pme_lb->setup[0].ewaldcoeff = ic->ewaldcoeff;
+
+ pme_lb->setup[0].pmedata = pmedata;
+
+ spm = 0;
+ for(d=0; d<DIM; d++)
+ {
+ sp = norm(pme_lb->box_start[d])/pme_lb->setup[0].grid[d];
+ if (sp > spm)
+ {
+ spm = sp;
+ }
+ }
+ pme_lb->setup[0].spacing = spm;
+
+ if (ir->fourier_spacing > 0)
+ {
+ pme_lb->cut_spacing = ir->rcoulomb/ir->fourier_spacing;
+ }
+ else
+ {
+ pme_lb->cut_spacing = ir->rcoulomb/pme_lb->setup[0].spacing;
+ }
+
+ pme_lb->stage = 0;
+
+ pme_lb->fastest = 0;
+ pme_lb->start = 0;
+ pme_lb->end = 0;
+ pme_lb->elimited = epmelblimNO;
+
+ *pme_lb_p = pme_lb;
+}
+
+static gmx_bool pme_loadbal_increase_cutoff(pme_load_balancing_t pme_lb,
+ int pme_order)
+{
+ pme_setup_t *set;
+ real fac,sp;
+ int d;
+
+ /* Try to add a new setup with next larger cut-off to the list */
+ pme_lb->n++;
+ srenew(pme_lb->setup,pme_lb->n);
+ set = &pme_lb->setup[pme_lb->n-1];
+ set->pmedata = NULL;
+
+ fac = 1;
+ do
+ {
+ fac *= 1.01;
+ clear_ivec(set->grid);
+ sp = calc_grid(NULL,pme_lb->box_start,
+ fac*pme_lb->setup[pme_lb->cur].spacing,
+ &set->grid[XX],
+ &set->grid[YY],
+ &set->grid[ZZ]);
+
+ /* In parallel we can't have grids smaller than 2*pme_order,
+ * and we would anyhow not gain much speed at these grid sizes.
+ */
+ for(d=0; d<DIM; d++)
+ {
+ if (set->grid[d] <= 2*pme_order)
+ {
+ pme_lb->n--;
+
+ return FALSE;
+ }
+ }
+ }
+ while (sp <= 1.001*pme_lb->setup[pme_lb->cur].spacing);
+
+ set->rcut = pme_lb->cut_spacing*sp;
+ set->rlist = set->rcut + pme_lb->rbuf;
+ set->spacing = sp;
+ /* The grid efficiency is the size wrt a grid with uniform x/y/z spacing */
+ set->grid_efficiency = 1;
+ for(d=0; d<DIM; d++)
+ {
+ set->grid_efficiency *= (set->grid[d]*sp)/norm(pme_lb->box_start[d]);
+ }
+ /* The Ewald coefficient is inversly proportional to the cut-off */
+ set->ewaldcoeff =
+ pme_lb->setup[0].ewaldcoeff*pme_lb->setup[0].rcut/set->rcut;
+
+ set->count = 0;
+ set->cycles = 0;
+
+ if (debug)
+ {
+ fprintf(debug,"PME loadbal: grid %d %d %d, cutoff %f\n",
+ set->grid[XX],set->grid[YY],set->grid[ZZ],set->rcut);
+ }
+
+ return TRUE;
+}
+
+static void print_grid(FILE *fp_err,FILE *fp_log,
+ const char *pre,
+ const char *desc,
+ const pme_setup_t *set,
+ double cycles)
+{
+ char buf[STRLEN],buft[STRLEN];
+
+ if (cycles >= 0)
+ {
+ sprintf(buft,": %.1f M-cycles",cycles*1e-6);
+ }
+ else
+ {
+ buft[0] = '\0';
+ }
+ sprintf(buf,"%-11s%10s pme grid %d %d %d, cutoff %.3f%s",
+ pre,
+ desc,set->grid[XX],set->grid[YY],set->grid[ZZ],set->rcut,
+ buft);
+ if (fp_err != NULL)
+ {
+ fprintf(fp_err,"\r%s\n",buf);
+ }
+ if (fp_log != NULL)
+ {
+ fprintf(fp_log,"%s\n",buf);
+ }
+}
+
+static int pme_loadbal_end(pme_load_balancing_t pme_lb)
+{
+ /* In the initial stage only n is set; end is not set yet */
+ if (pme_lb->end > 0)
+ {
+ return pme_lb->end;
+ }
+ else
+ {
+ return pme_lb->n;
+ }
+}
+
+static void print_loadbal_limited(FILE *fp_err,FILE *fp_log,
+ gmx_large_int_t step,
+ pme_load_balancing_t pme_lb)
+{
+ char buf[STRLEN],sbuf[22];
+
+ sprintf(buf,"step %4s: the %s limited the PME load balancing to a cut-off of %.3f",
+ gmx_step_str(step,sbuf),
+ pmelblim_str[pme_lb->elimited],
+ pme_lb->setup[pme_loadbal_end(pme_lb)].rcut);
+ if (fp_err != NULL)
+ {
+ fprintf(fp_err,"\r%s\n",buf);
+ }
+ if (fp_log != NULL)
+ {
+ fprintf(fp_log,"%s\n",buf);
+ }
+}
+
+static void switch_to_stage1(pme_load_balancing_t pme_lb)
+{
+ pme_lb->start = 0;
+ while (pme_lb->start+1 < pme_lb->n &&
+ (pme_lb->setup[pme_lb->start].count == 0 ||
+ pme_lb->setup[pme_lb->start].cycles >
+ pme_lb->setup[pme_lb->fastest].cycles*PME_LB_SLOW_FAC))
+ {
+ pme_lb->start++;
+ }
+ while (pme_lb->start > 0 && pme_lb->setup[pme_lb->start-1].cycles == 0)
+ {
+ pme_lb->start--;
+ }
+
+ pme_lb->end = pme_lb->n;
+ if (pme_lb->setup[pme_lb->end-1].count > 0 &&
+ pme_lb->setup[pme_lb->end-1].cycles >
+ pme_lb->setup[pme_lb->fastest].cycles*PME_LB_SLOW_FAC)
+ {
+ pme_lb->end--;
+ }
+
+ pme_lb->stage = 1;
+
+ /* Next we want to choose setup pme_lb->start, but as we will increase
+ * pme_ln->cur by one right after returning, we subtract 1 here.
+ */
+ pme_lb->cur = pme_lb->start - 1;
+}
+
+gmx_bool pme_load_balance(pme_load_balancing_t pme_lb,
+ t_commrec *cr,
+ FILE *fp_err,
+ FILE *fp_log,
+ t_inputrec *ir,
+ t_state *state,
+ double cycles,
+ interaction_const_t *ic,
+ nonbonded_verlet_t *nbv,
+ gmx_pme_t *pmedata,
+ gmx_large_int_t step)
+{
+ gmx_bool OK;
+ pme_setup_t *set;
+ double cycles_fast;
+ char buf[STRLEN],sbuf[22];
+
+ if (pme_lb->stage == pme_lb->nstage)
+ {
+ return FALSE;
+ }
+
+ if (PAR(cr))
+ {
+ gmx_sumd(1,&cycles,cr);
+ cycles /= cr->nnodes;
+ }
+
+ set = &pme_lb->setup[pme_lb->cur];
+
+ set->count++;
+ if (set->count % 2 == 1)
+ {
+ /* Skip the first cycle, because the first step after a switch
+ * is much slower due to allocation and/or caching effects.
+ */
+ return TRUE;
+ }
+
+ sprintf(buf, "step %4s: ", gmx_step_str(step,sbuf));
+ print_grid(fp_err,fp_log,buf,"timed with",set,cycles);
+
+ if (set->count <= 2)
+ {
+ set->cycles = cycles;
+ }
+ else
+ {
+ if (cycles*PME_LB_ACCEL_TOL < set->cycles &&
+ pme_lb->stage == pme_lb->nstage - 1)
+ {
+ /* The performance went up a lot (due to e.g. DD load balancing).
+ * Add a stage, keep the minima, but rescan all setups.
+ */
+ pme_lb->nstage++;
+
+ if (debug)
+ {
+ fprintf(debug,"The performance for grid %d %d %d went from %.3f to %.1f M-cycles, this is more than %f\n"
+ "Increased the number stages to %d"
+ " and ignoring the previous performance\n",
+ set->grid[XX],set->grid[YY],set->grid[ZZ],
+ cycles*1e-6,set->cycles*1e-6,PME_LB_ACCEL_TOL,
+ pme_lb->nstage);
+ }
+ }
+ set->cycles = min(set->cycles,cycles);
+ }
+
+ if (set->cycles < pme_lb->setup[pme_lb->fastest].cycles)
+ {
+ pme_lb->fastest = pme_lb->cur;
+ }
+ cycles_fast = pme_lb->setup[pme_lb->fastest].cycles;
+
+ /* Check in stage 0 if we should stop scanning grids.
+ * Stop when the time is more than SLOW_FAC longer than the fastest.
+ */
+ if (pme_lb->stage == 0 && pme_lb->cur > 0 &&
+ cycles > pme_lb->setup[pme_lb->fastest].cycles*PME_LB_SLOW_FAC)
+ {
+ pme_lb->n = pme_lb->cur + 1;
+ /* Done with scanning, go to stage 1 */
+ switch_to_stage1(pme_lb);
+ }
+
+ if (pme_lb->stage == 0)
+ {
+ int gridsize_start;
+
+ gridsize_start = set->grid[XX]*set->grid[YY]*set->grid[ZZ];
+
+ do
+ {
+ if (pme_lb->cur+1 < pme_lb->n)
+ {
+ /* We had already generated the next setup */
+ OK = TRUE;
+ }
+ else
+ {
+ /* Find the next setup */
+ OK = pme_loadbal_increase_cutoff(pme_lb,ir->pme_order);
+ }
+
+ if (OK && ir->ePBC != epbcNONE)
+ {
+ OK = (sqr(pme_lb->setup[pme_lb->cur+1].rlist)
+ <= max_cutoff2(ir->ePBC,state->box));
+ if (!OK)
+ {
+ pme_lb->elimited = epmelblimBOX;
+ }
+ }
+
+ if (OK)
+ {
+ pme_lb->cur++;
+
+ if (DOMAINDECOMP(cr))
+ {
+ OK = change_dd_cutoff(cr,state,ir,
+ pme_lb->setup[pme_lb->cur].rlist);
+ if (!OK)
+ {
+ /* Failed: do not use this setup */
+ pme_lb->cur--;
+ pme_lb->elimited = epmelblimDD;
+ }
+ }
+ }
+ if (!OK)
+ {
+ /* We hit the upper limit for the cut-off,
+ * the setup should not go further than cur.
+ */
+ pme_lb->n = pme_lb->cur + 1;
+ print_loadbal_limited(fp_err,fp_log,step,pme_lb);
+ /* Switch to the next stage */
+ switch_to_stage1(pme_lb);
+ }
+ }
+ while (OK &&
+ !(pme_lb->setup[pme_lb->cur].grid[XX]*
+ pme_lb->setup[pme_lb->cur].grid[YY]*
+ pme_lb->setup[pme_lb->cur].grid[ZZ] <
+ gridsize_start*PME_LB_GRID_SCALE_FAC
+ &&
+ pme_lb->setup[pme_lb->cur].grid_efficiency <
+ pme_lb->setup[pme_lb->cur-1].grid_efficiency*PME_LB_GRID_EFFICIENCY_REL_FAC));
+ }
+
+ if (pme_lb->stage > 0 && pme_lb->end == 1)
+ {
+ pme_lb->cur = 0;
+ pme_lb->stage = pme_lb->nstage;
+ }
+ else if (pme_lb->stage > 0 && pme_lb->end > 1)
+ {
+ /* If stage = nstage-1:
+ * scan over all setups, rerunning only those setups
+ * which are not much slower than the fastest
+ * else:
+ * use the next setup
+ */
+ do
+ {
+ pme_lb->cur++;
+ if (pme_lb->cur == pme_lb->end)
+ {
+ pme_lb->stage++;
+ pme_lb->cur = pme_lb->start;
+ }
+ }
+ while (pme_lb->stage == pme_lb->nstage - 1 &&
+ pme_lb->setup[pme_lb->cur].count > 0 &&
+ pme_lb->setup[pme_lb->cur].cycles > cycles_fast*PME_LB_SLOW_FAC);
+
+ if (pme_lb->stage == pme_lb->nstage)
+ {
+ /* We are done optimizing, use the fastest setup we found */
+ pme_lb->cur = pme_lb->fastest;
+ }
+ }
+
+ if (DOMAINDECOMP(cr) && pme_lb->stage > 0)
+ {
+ OK = change_dd_cutoff(cr,state,ir,pme_lb->setup[pme_lb->cur].rlist);
+ if (!OK)
+ {
+ /* Failsafe solution */
+ if (pme_lb->cur > 1 && pme_lb->stage == pme_lb->nstage)
+ {
+ pme_lb->stage--;
+ }
+ pme_lb->fastest = 0;
+ pme_lb->start = 0;
+ pme_lb->end = pme_lb->cur;
+ pme_lb->cur = pme_lb->start;
+ pme_lb->elimited = epmelblimDD;
+ print_loadbal_limited(fp_err,fp_log,step,pme_lb);
+ }
+ }
+
+ /* Change the Coulomb cut-off and the PME grid */
+
+ set = &pme_lb->setup[pme_lb->cur];
+
+ ic->rcoulomb = set->rcut;
+ ic->rlist = set->rlist;
+ ic->ewaldcoeff = set->ewaldcoeff;
+
+ if (nbv->grp[0].kernel_type == nbk8x8x8_CUDA)
+ {
+ nbnxn_cuda_pme_loadbal_update_param(nbv->cu_nbv,ic);
+ }
+ else
+ {
+ init_interaction_const_tables(NULL,ic,nbv->grp[0].kernel_type);
+ }
+
+ if (nbv->ngrp > 1)
+ {
+ init_interaction_const_tables(NULL,ic,nbv->grp[1].kernel_type);
+ }
+
+ if (cr->duty & DUTY_PME)
+ {
+ if (pme_lb->setup[pme_lb->cur].pmedata == NULL)
+ {
+ /* Generate a new PME data structure,
+ * copying part of the old pointers.
+ */
+ gmx_pme_reinit(&set->pmedata,
+ cr,pme_lb->setup[0].pmedata,ir,
+ set->grid);
+ }
+ *pmedata = set->pmedata;
+ }
+ else
+ {
+ /* Tell our PME-only node to switch grid */
+ gmx_pme_send_switch(cr, set->grid, set->ewaldcoeff);
+ }
+
+ if (debug)
+ {
+ print_grid(NULL,debug,"","switched to",set,-1);
+ }
+
+ if (pme_lb->stage == pme_lb->nstage)
+ {
+ print_grid(fp_err,fp_log,"","optimal",set,-1);
+ }
+
+ return TRUE;
+}
+
+void restart_pme_loadbal(pme_load_balancing_t pme_lb, int n)
+{
+ pme_lb->nstage += n;
+}
+
+static int pme_grid_points(const pme_setup_t *setup)
+{
+ return setup->grid[XX]*setup->grid[YY]*setup->grid[ZZ];
+}
+
+static void print_pme_loadbal_setting(FILE *fplog,
+ char *name,
+ const pme_setup_t *setup)
+{
+ fprintf(fplog,
+ " %-7s %6.3f nm %6.3f nm %3d %3d %3d %5.3f nm %5.3f nm\n",
+ name,
+ setup->rcut,setup->rlist,
+ setup->grid[XX],setup->grid[YY],setup->grid[ZZ],
+ setup->spacing,1/setup->ewaldcoeff);
+}
+
+static void print_pme_loadbal_settings(pme_load_balancing_t pme_lb,
+ FILE *fplog)
+{
+ double pp_ratio,grid_ratio;
+
+ pp_ratio = pow(pme_lb->setup[pme_lb->cur].rlist/pme_lb->setup[0].rlist,3.0);
+ grid_ratio = pme_grid_points(&pme_lb->setup[pme_lb->cur])/
+ (double)pme_grid_points(&pme_lb->setup[0]);
+
+ fprintf(fplog,"\n");
+ fprintf(fplog," P P - P M E L O A D B A L A N C I N G\n");
+ fprintf(fplog,"\n");
+ /* Here we only warn when the optimal setting is the last one */
+ if (pme_lb->elimited != epmelblimNO &&
+ pme_lb->cur == pme_loadbal_end(pme_lb)-1)
+ {
+ fprintf(fplog," NOTE: The PP/PME load balancing was limited by the %s,\n",
+ pmelblim_str[pme_lb->elimited]);
+ fprintf(fplog," you might not have reached a good load balance.\n");
+ if (pme_lb->elimited == epmelblimDD)
+ {
+ fprintf(fplog," Try different mdrun -dd settings or lower the -dds value.\n");
+ }
+ fprintf(fplog,"\n");
+ }
+ fprintf(fplog," PP/PME load balancing changed the cut-off and PME settings:\n");
+ fprintf(fplog," particle-particle PME\n");
+ fprintf(fplog," rcoulomb rlist grid spacing 1/beta\n");
+ print_pme_loadbal_setting(fplog,"initial",&pme_lb->setup[0]);
+ print_pme_loadbal_setting(fplog,"final" ,&pme_lb->setup[pme_lb->cur]);
+ fprintf(fplog," cost-ratio %4.2f %4.2f\n",
+ pp_ratio,grid_ratio);
+ fprintf(fplog," (note that these numbers concern only part of the total PP and PME load)\n");
+ fprintf(fplog,"\n");
+}
+
+void pme_loadbal_done(pme_load_balancing_t pme_lb, FILE *fplog)
+{
+ if (fplog != NULL && (pme_lb->cur > 0 || pme_lb->elimited != epmelblimNO))
+ {
+ print_pme_loadbal_settings(pme_lb,fplog);
+ }
+
+ /* TODO: Here we should free all pointers in pme_lb,
+ * but as it contains pme data structures,
+ * we need to first make pme.c free all data.
+ */
+}
--- /dev/null
+/* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
+ *
+ *
+ * This source code is part of
+ *
+ * G R O M A C S
+ *
+ * GROningen MAchine for Chemical Simulations
+ *
+ * VERSION 4.6.0
+ * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
+ * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
+ * Copyright (c) 2001-2011, The GROMACS development team,
+ * check out http://www.gromacs.org for more information.
+
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * If you want to redistribute modifications, please consider that
+ * scientific software is very special. Version control is crucial -
+ * bugs must be traceable. We will be happy to consider code for
+ * inclusion in the official distribution, but derived work must not
+ * be called official GROMACS. Details are found in the README & COPYING
+ * files - if they are missing, get the official version at www.gromacs.org.
+ *
+ * To help us fund GROMACS development, we humbly ask that you cite
+ * the papers on the package - you can find them in the top README file.
+ *
+ * For more info, check our website at http://www.gromacs.org
+ *
+ * And Hey:
+ * Gallium Rubidium Oxygen Manganese Argon Carbon Silicon
+ */
+
+#ifndef _pme_loadbal_h
+#define _pme_loadbal_h
+
+typedef struct pme_load_balancing *pme_load_balancing_t;
+
+/* Initialze the PP-PME load balacing data and infrastructure */
+void pme_loadbal_init(pme_load_balancing_t *pme_lb_p,
+ const t_inputrec *ir,matrix box,
+ const interaction_const_t *ic,
+ gmx_pme_t pmedata);
+
+/* Try to adjust the PME grid and Coulomb cut-off.
+ * The adjustment is done to generate a different non-bonded PP and PME load.
+ * With separate PME nodes (PP and PME on different processes) or with
+ * a GPU (PP on GPU, PME on CPU), PP and PME run on different resources
+ * and changing the load will affect the load balance and performance.
+ * The total time for a set of integration steps is monitored and a range
+ * of grid/cut-off setups is scanned. After calling pme_load_balance many
+ * times and acquiring enough statistics, the best performing setup is chosen.
+ * Here we try to take into account fluctuations and changes due to external
+ * factors as well as DD load balancing.
+ * Returns TRUE the load balancing continues, FALSE is the balancing is done.
+ */
+gmx_bool pme_load_balance(pme_load_balancing_t pme_lb,
+ t_commrec *cr,
+ FILE *fp_err,
+ FILE *fp_log,
+ t_inputrec *ir,
+ t_state *state,
+ double cycles,
+ interaction_const_t *ic,
+ nonbonded_verlet_t *nbv,
+ gmx_pme_t *pmedata,
+ gmx_large_int_t step);
+
+/* Restart the PME load balancing discarding all timings gathered up till now */
+void restart_pme_loadbal(pme_load_balancing_t pme_lb, int n);
+
+/* Finish the PME load balancing and print the settings when fplog!=NULL */
+void pme_loadbal_done(pme_load_balancing_t pme_lb, FILE *fplog);
+
+#endif /* _pme_loadbal_h */
+++ /dev/null
-/* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
- *
- *
- * This source code is part of
- *
- * G R O M A C S
- *
- * GROningen MAchine for Chemical Simulations
- *
- * VERSION 4.6.0
- * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
- * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
- * Copyright (c) 2001-2011, The GROMACS development team,
- * check out http://www.gromacs.org for more information.
-
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- *
- * If you want to redistribute modifications, please consider that
- * scientific software is very special. Version control is crucial -
- * bugs must be traceable. We will be happy to consider code for
- * inclusion in the official distribution, but derived work must not
- * be called official GROMACS. Details are found in the README & COPYING
- * files - if they are missing, get the official version at www.gromacs.org.
- *
- * To help us fund GROMACS development, we humbly ask that you cite
- * the papers on the package - you can find them in the top README file.
- *
- * For more info, check our website at http://www.gromacs.org
- *
- * And Hey:
- * Gallium Rubidium Oxygen Manganese Argon Carbon Silicon
- */
-#ifdef HAVE_CONFIG_H
-#include <config.h>
-#endif
-
-#include "smalloc.h"
-#include "network.h"
-#include "calcgrid.h"
-#include "pme.h"
-#include "vec.h"
-#include "domdec.h"
-#include "nbnxn_cuda_data_mgmt.h"
-#include "force.h"
-#include "pme_switch.h"
-
-typedef struct {
- real rcut;
- real rlist;
- real spacing;
- ivec grid;
- real grid_eff;
- real coeff;
- gmx_pme_t pmedata;
-
- int count;
- double cycles;
-} pme_setup_t;
-
-/* In the initial scan, step by grids that are at least a factor 0.8 coarser */
-#define PMES_GRID_SCALE_FAC 0.8
-/* In the initial scan, try to skip grids with uneven x/y/z spacing,
- * checking if the "efficiency" is more than 5% worse than the previous grid.
- */
-#define PMES_GRID_EFF_FAC 1.05
-/* Rerun up till 12% slower setups than the fastest up till now */
-#define PMES_SLOW_FAC 1.12
-/* If setups get more than 2% faster, do another round to avoid
- * choosing a slower setup due to acceleration or fluctuations.
- */
-#define PMES_ACCEL_TOL 1.02
-
-typedef struct pme_switch {
- int nstage; /* the current maximum number of stages */
-
- real cut_spacing; /* the minimum cutoff / PME grid spacing ratio */
- real rbuf; /* the pairlist buffer size */
- matrix box_start; /* the initial simulation box */
- int n; /* the count of setup as well as the allocation size */
- pme_setup_t *setup; /* the PME+cutoff setups */
- int cur; /* the current setup */
- int fastest; /* fastest setup up till now */
- int start; /* start of setup range to consider in stage>0 */
- int end; /* end of setup range to consider in stage>0 */
-
- int stage; /* the current stage */
-} t_pme_switch;
-
-void switch_pme_init(pme_switch_t *pmes_p,
- const t_inputrec *ir,matrix box,
- const interaction_const_t *ic,
- gmx_pme_t pmedata)
-{
- pme_switch_t pmes;
- real spm,sp;
- int d;
-
- snew(pmes,1);
-
- /* Any number of stages >= 2 is supported */
- pmes->nstage = 2;
-
- pmes->rbuf = ic->rlist - ic->rcoulomb;
-
- copy_mat(box,pmes->box_start);
- if (ir->ePBC==epbcXY && ir->nwall==2)
- {
- svmul(ir->wall_ewald_zfac,pmes->box_start[ZZ],pmes->box_start[ZZ]);
- }
-
- pmes->n = 1;
- snew(pmes->setup,pmes->n);
-
- pmes->cur = 0;
- pmes->setup[0].rcut = ic->rcoulomb;
- pmes->setup[0].rlist = ic->rlist;
- pmes->setup[0].grid[XX] = ir->nkx;
- pmes->setup[0].grid[YY] = ir->nky;
- pmes->setup[0].grid[ZZ] = ir->nkz;
- pmes->setup[0].coeff = ic->ewaldcoeff;
-
- pmes->setup[0].pmedata = pmedata;
-
- spm = 0;
- for(d=0; d<DIM; d++)
- {
- sp = norm(pmes->box_start[d])/pmes->setup[0].grid[d];
- if (sp > spm)
- {
- spm = sp;
- }
- }
- pmes->setup[0].spacing = spm;
-
- if (ir->fourier_spacing > 0)
- {
- pmes->cut_spacing = ir->rcoulomb/ir->fourier_spacing;
- }
- else
- {
- pmes->cut_spacing = ir->rcoulomb/pmes->setup[0].spacing;
- }
-
- pmes->stage = 0;
-
- pmes->fastest = 0;
- pmes->start = 0;
-
- *pmes_p = pmes;
-}
-
-static gmx_bool switch_pme_increase_cutoff(pme_switch_t pmes,int pme_order)
-{
- pme_setup_t *set;
- real fac,sp;
- int d;
-
- /* Try to add a new setup with next larger cut-off to the list */
- pmes->n++;
- srenew(pmes->setup,pmes->n);
- set = &pmes->setup[pmes->n-1];
- set->pmedata = NULL;
-
- fac = 1;
- do
- {
- fac *= 1.01;
- clear_ivec(set->grid);
- sp = calc_grid(NULL,pmes->box_start,
- fac*pmes->setup[pmes->cur].spacing,
- &set->grid[XX],
- &set->grid[YY],
- &set->grid[ZZ]);
-
- /* In parallel we can't have grids smaller than 2*pme_order,
- * and we would anyhow not gain much speed at these grid sizes.
- */
- for(d=0; d<DIM; d++)
- {
- if (set->grid[d] <= 2*pme_order)
- {
- pmes->n--;
-
- return FALSE;
- }
- }
- }
- while (sp <= 1.001*pmes->setup[pmes->cur].spacing);
-
- set->rcut = pmes->cut_spacing*sp;
- set->rlist = set->rcut + pmes->rbuf;
- set->spacing = sp;
- /* The grid efficiency is the size wrt a grid with uniform x/y/z spacing */
- set->grid_eff = 1;
- for(d=0; d<DIM; d++)
- {
- set->grid_eff *= (set->grid[d]*sp)/norm(pmes->box_start[d]);
- }
- /* The Ewald coefficient is inversly proportional to the cut-off */
- set->coeff = pmes->setup[0].coeff*pmes->setup[0].rcut/set->rcut;
-
- set->count = 0;
- set->cycles = 0;
-
- if (debug)
- {
- fprintf(debug,"PME switch grid %d %d %d, cutoff %f\n",
- set->grid[XX],set->grid[YY],set->grid[ZZ],set->rcut);
- }
-
- return TRUE;
-}
-
-static void print_grid(FILE *fp_err,FILE *fp_log,
- const char *pre,
- const char *desc,
- const pme_setup_t *set,
- double cycles)
-{
- char buf[STRLEN],buft[STRLEN];
-
- if (cycles >= 0)
- {
- sprintf(buft,": %.1f M-cycles",cycles*1e-6);
- }
- else
- {
- buft[0] = '\0';
- }
- sprintf(buf,"%-11s%10s pme grid %d %d %d, cutoff %.3f%s",
- pre,
- desc,set->grid[XX],set->grid[YY],set->grid[ZZ],set->rcut,
- buft);
- if (fp_err != NULL)
- {
- fprintf(fp_err,"%s\n",buf);
- }
- if (fp_log != NULL)
- {
- fprintf(fp_log,"%s\n",buf);
- }
-}
-
-static void switch_to_stage1(pme_switch_t pmes)
-{
- pmes->start = 0;
- while (pmes->start+1 < pmes->n &&
- (pmes->setup[pmes->start].count == 0 ||
- pmes->setup[pmes->start].cycles >
- pmes->setup[pmes->fastest].cycles*PMES_SLOW_FAC))
- {
- pmes->start++;
- }
- while (pmes->start > 0 && pmes->setup[pmes->start-1].cycles == 0)
- {
- pmes->start--;
- }
-
- pmes->end = pmes->n;
- if (pmes->setup[pmes->end-1].count > 0 &&
- pmes->setup[pmes->end-1].cycles >
- pmes->setup[pmes->fastest].cycles*PMES_SLOW_FAC)
- {
- pmes->end--;
- }
-
- pmes->stage = 1;
-
- /* Start add start, 1 will be added immediately after returning */
- pmes->cur = pmes->start - 1;
-}
-
-gmx_bool switch_pme(pme_switch_t pmes,
- t_commrec *cr,
- FILE *fp_err,
- FILE *fp_log,
- t_inputrec *ir,
- t_state *state,
- double cycles,
- interaction_const_t *ic,
- nonbonded_verlet_t *nbv,
- gmx_pme_t *pmedata,
- int step)
-{
- gmx_bool OK;
- pme_setup_t *set;
- double cycles_fast;
- char buf[STRLEN];
-
- if (pmes->stage == pmes->nstage)
- {
- return FALSE;
- }
-
- if (PAR(cr))
- {
- gmx_sumd(1,&cycles,cr);
- cycles /= cr->nnodes;
- }
-
- set = &pmes->setup[pmes->cur];
-
- set->count++;
- if (set->count % 2 == 1)
- {
- /* Skip the first cycle, because the first step after a switch
- * is much slower due to allocation and/or caching effects.
- */
- return TRUE;
- }
-
- sprintf(buf, "step %4d: ", step);
- print_grid(fp_err,fp_log,buf,"timed with",set,cycles);
-
- if (set->count <= 2)
- {
- set->cycles = cycles;
- }
- else
- {
- if (cycles*PMES_ACCEL_TOL < set->cycles &&
- pmes->stage == pmes->nstage - 1)
- {
- /* The performance went up a lot (due to e.g. DD load balancing).
- * Add a stage, keep the minima, but rescan all setups.
- */
- pmes->nstage++;
-
- if (debug)
- {
- fprintf(debug,"The performance for grid %d %d %d went from %.3f to %.1f M-cycles, this is more than %f\n"
- "Increased the number stages to %d"
- " and ignoring the previous performance\n",
- set->grid[XX],set->grid[YY],set->grid[ZZ],
- cycles*1e-6,set->cycles*1e-6,PMES_ACCEL_TOL,
- pmes->nstage);
- }
- }
- set->cycles = min(set->cycles,cycles);
- }
-
- if (set->cycles < pmes->setup[pmes->fastest].cycles)
- {
- pmes->fastest = pmes->cur;
- }
- cycles_fast = pmes->setup[pmes->fastest].cycles;
-
- /* Check in stage 0 if we should stop scanning grids.
- * Stop when the time is more than SLOW_FAC longer than the fastest.
- */
- if (pmes->stage == 0 && pmes->cur > 0 &&
- cycles > pmes->setup[pmes->fastest].cycles*PMES_SLOW_FAC)
- {
- pmes->n = pmes->cur + 1;
- /* Done with scanning, go to stage 1 */
- switch_to_stage1(pmes);
- }
-
- if (pmes->stage == 0)
- {
- int gridsize_start;
-
- gridsize_start = set->grid[XX]*set->grid[YY]*set->grid[ZZ];
-
- do
- {
- if (pmes->cur+1 < pmes->n)
- {
- /* We had already generated the next setup */
- OK = TRUE;
- }
- else
- {
- /* Find the next setup */
- OK = switch_pme_increase_cutoff(pmes,ir->pme_order);
- }
-
- if (OK && ir->ePBC != epbcNONE)
- {
- OK = (sqr(pmes->setup[pmes->cur+1].rlist)
- <= max_cutoff2(ir->ePBC,state->box));
- }
-
- if (OK)
- {
- pmes->cur++;
-
- if (DOMAINDECOMP(cr))
- {
- OK = change_dd_cutoff(cr,state,ir,
- pmes->setup[pmes->cur].rlist);
- if (!OK)
- {
- /* Failed: do not use this setup */
- pmes->cur--;
- }
- }
- }
- if (!OK)
- {
- /* We hit the upper limit for the cut-off,
- * the setup should not go further than cur.
- */
- pmes->n = pmes->cur + 1;
- /* Switch to the next stage */
- switch_to_stage1(pmes);
- }
- }
- while (OK &&
- !(pmes->setup[pmes->cur].grid[XX]*
- pmes->setup[pmes->cur].grid[YY]*
- pmes->setup[pmes->cur].grid[ZZ] <
- gridsize_start*PMES_GRID_SCALE_FAC
- &&
- pmes->setup[pmes->cur].grid_eff <
- pmes->setup[pmes->cur-1].grid_eff*PMES_GRID_EFF_FAC));
- }
-
- if (pmes->stage > 0 && pmes->end == 1)
- {
- pmes->cur = 0;
- pmes->stage = pmes->nstage;
- }
- else if (pmes->stage > 0 && pmes->end > 1)
- {
- /* If stage = nstage-1:
- * scan over all setups, rerunning only those setups
- * which are not much slower than the fastest
- * else:
- * use the next setup
- */
- do
- {
- pmes->cur++;
- if (pmes->cur == pmes->end)
- {
- pmes->stage++;
- pmes->cur = pmes->start;
- }
- }
- while (pmes->stage == pmes->nstage - 1 &&
- pmes->setup[pmes->cur].count > 0 &&
- pmes->setup[pmes->cur].cycles > cycles_fast*PMES_SLOW_FAC);
-
- if (pmes->stage == pmes->nstage)
- {
- /* We are done optiming, use the fastest setup we found */
- pmes->cur = pmes->fastest;
- }
- }
-
- if (DOMAINDECOMP(cr) && pmes->stage > 0)
- {
- OK = change_dd_cutoff(cr,state,ir,pmes->setup[pmes->cur].rlist);
- if (!OK)
- {
- /* Failsafe solution */
- if (pmes->cur > 1 && pmes->stage == pmes->nstage)
- {
- pmes->stage--;
- }
- pmes->fastest = 0;
- pmes->start = 0;
- pmes->end = pmes->cur;
- pmes->cur = pmes->start;
- }
- }
-
- /* Change the Coulomb cut-off and the PME grid */
-
- set = &pmes->setup[pmes->cur];
-
- ic->rcoulomb = set->rcut;
- ic->rlist = set->rlist;
- ic->ewaldcoeff = set->coeff;
-
- if (nbv->grp[0].kernel_type == nbk8x8x8_CUDA)
- {
- nbnxn_cuda_pmetune_update_param(nbv->cu_nbv,ic);
- }
- else
- {
- init_interaction_const_tables(NULL,ic,nbv->grp[0].kernel_type);
- }
-
- if (nbv->ngrp > 1)
- {
- init_interaction_const_tables(NULL,ic,nbv->grp[1].kernel_type);
- }
-
- if (cr->duty & DUTY_PME)
- {
- if (pmes->setup[pmes->cur].pmedata == NULL)
- {
- /* Generate a new PME data structure,
- * copying part of the old pointers.
- */
- gmx_pme_reinit(&set->pmedata,
- cr,pmes->setup[0].pmedata,ir,
- set->grid);
- }
- *pmedata = set->pmedata;
- }
- else
- {
- /* Tell our PME-only node to switch grid */
- gmx_pme_send_switch(cr, set->grid, set->coeff);
- }
-
- if (debug)
- {
- print_grid(NULL,debug,"","switched to",set,-1);
- }
-
- if (pmes->stage == pmes->nstage)
- {
- print_grid(fp_err,fp_log,"","optimal",set,-1);
- }
-
- return TRUE;
-}
-
-void restart_switch_pme(pme_switch_t pmes, int n)
-{
- pmes->nstage += n;
-}
+++ /dev/null
-/* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
- *
- *
- * This source code is part of
- *
- * G R O M A C S
- *
- * GROningen MAchine for Chemical Simulations
- *
- * VERSION 4.6.0
- * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
- * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
- * Copyright (c) 2001-2011, The GROMACS development team,
- * check out http://www.gromacs.org for more information.
-
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- *
- * If you want to redistribute modifications, please consider that
- * scientific software is very special. Version control is crucial -
- * bugs must be traceable. We will be happy to consider code for
- * inclusion in the official distribution, but derived work must not
- * be called official GROMACS. Details are found in the README & COPYING
- * files - if they are missing, get the official version at www.gromacs.org.
- *
- * To help us fund GROMACS development, we humbly ask that you cite
- * the papers on the package - you can find them in the top README file.
- *
- * For more info, check our website at http://www.gromacs.org
- *
- * And Hey:
- * Gallium Rubidium Oxygen Manganese Argon Carbon Silicon
- */
-
-#ifndef _pme_switch_h
-#define _pme_switch_h
-
-typedef struct pme_switch *pme_switch_t;
-
-/* Initialze the PME grid tuning data and infrastructure */
-void switch_pme_init(pme_switch_t *pmes_p,
- const t_inputrec *ir,matrix box,
- const interaction_const_t *ic,
- gmx_pme_t pmedata);
-
-/* Adjust the PME grid and Coulomb cut-off.
- * Returns TRUE the tuning continues, FALSE is the tuning is done.
- */
-gmx_bool switch_pme(pme_switch_t pmes,
- t_commrec *cr,
- FILE *fp_err,
- FILE *fp_log,
- t_inputrec *ir,
- t_state *state,
- double cycles,
- interaction_const_t *ic,
- nonbonded_verlet_t *nbv,
- gmx_pme_t *pmedata,
- int step);
-
-/* Restart the PME tuning, discarding all timings gathered up till now */
-void restart_switch_pme(pme_switch_t pmes, int n);
-
-#endif /* _pme_switch_h */
#include "gmx_cyclecounter.h"
#include "smalloc.h"
#include "gmx_fatal.h"
+#include "md_logging.h"
#ifdef GMX_LIB_MPI
#include <mpi.h>
* CPU-GPU load balancing is possible */
if (gpu_cpu_ratio < 0.75 || gpu_cpu_ratio > 1.2)
{
+ /* Only the sim master calls this function, so always print to stderr */
if (gpu_cpu_ratio < 0.75)
{
- sprintf(buf, "NOTE: The GPU has >25%% less load than the CPU. This imbalance causes\n"
- " performance loss, consider turning on PME tuning (-tunepme).");
+ if (npp > 1)
+ {
+ /* The user could have used -notunepme,
+ * but we currently can't check that here.
+ */
+ md_print_warn(NULL,fplog,
+ "NOTE: The GPU has >25%% less load than the CPU. This imbalance causes\n"
+ " performance loss. Maybe the domain decomposition limits the PME tuning.\n"
+ " In that case, try setting the DD grid manually (-dd) or lowering -dds.\n");
+ }
+ else
+ {
+ /* We should not end up here, unless the box is
+ * too small for increasing the cut-off for PME tuning.
+ */
+ md_print_warn(NULL,fplog,
+ "NOTE: The GPU has >25%% less load than the CPU. This imbalance causes\n"
+ " performance loss.\n");
+ }
}
if (gpu_cpu_ratio > 1.2)
{
- sprintf(buf, "NOTE: The GPU has >20%% more load than the CPU. This imbalance causes\n"
- " performance loss, consider using a shorter cut-off.");
+ md_print_warn(NULL,fplog,
+ "NOTE: The GPU has >20%% more load than the CPU. This imbalance causes\n"
+ " performance loss, consider using a shorter cut-off and a finer PME grid.\n");
}
- if (fplog)
- {
- fprintf(fplog,"\n%s\n",buf);
- }
- fprintf(stderr,"\n\n%s\n",buf);
}
}
}
(cycles[ewcDOMDEC] > tot*0.1 ||
cycles[ewcNS] > tot*0.1))
{
+ /* Only the sim master calls this function, so always print to stderr */
if (wc->wcc[ewcDOMDEC].n == 0)
{
- sprintf(buf,
- "NOTE: %d %% of the run time was spent in pair search,\n"
- " you might want to increase nstlist (this has no effect on accuracy)\n",
- (int)(100*cycles[ewcNS]/tot+0.5));
+ md_print_warn(NULL,fplog,
+ "NOTE: %d %% of the run time was spent in pair search,\n"
+ " you might want to increase nstlist (this has no effect on accuracy)\n",
+ (int)(100*cycles[ewcNS]/tot+0.5));
}
else
{
- sprintf(buf,
- "NOTE: %d %% of the run time was spent in domain decomposition,\n"
- " %d %% of the run time was spent in pair search,\n"
- " you might want to increase nstlist (this has no effect on accuracy)\n",
- (int)(100*cycles[ewcDOMDEC]/tot+0.5),
- (int)(100*cycles[ewcNS]/tot+0.5));
- }
- if (fplog)
- {
- fprintf(fplog,"\n%s\n",buf);
+ md_print_warn(NULL,fplog,
+ "NOTE: %d %% of the run time was spent in domain decomposition,\n"
+ " %d %% of the run time was spent in pair search,\n"
+ " you might want to increase nstlist (this has no effect on accuracy)\n",
+ (int)(100*cycles[ewcDOMDEC]/tot+0.5),
+ (int)(100*cycles[ewcNS]/tot+0.5));
}
- /* Only the sim master calls this function, so always print to stderr */
- fprintf(stderr,"\n%s\n",buf);
}
if (cycles[ewcMoveE] > tot*0.05)
{
- sprintf(buf,
- "NOTE: %d %% of the run time was spent communicating energies,\n"
- " you might want to use the -gcom option of mdrun\n",
- (int)(100*cycles[ewcMoveE]/tot+0.5));
- if (fplog)
- {
- fprintf(fplog,"\n%s\n",buf);
- }
/* Only the sim master calls this function, so always print to stderr */
- fprintf(stderr,"\n%s\n",buf);
+ md_print_warn(NULL,fplog,
+ "NOTE: %d %% of the run time was spent communicating energies,\n"
+ " you might want to use the -gcom option of mdrun\n",
+ (int)(100*cycles[ewcMoveE]/tot+0.5));
}
}
/*! Re-generate the GPU Ewald force table, resets rlist, and update the
* electrostatic type switching to twin cut-off (or back) if needed. */
-void nbnxn_cuda_pmetune_update_param(nbnxn_cuda_ptr_t cu_nb,
- const interaction_const_t *ic)
+void nbnxn_cuda_pme_loadbal_update_param(nbnxn_cuda_ptr_t cu_nb,
+ const interaction_const_t *ic)
{
cu_nbparam_t *nbp = cu_nb->nbparam;
biod.pnpi.spb.ru / alexxy / gromacs.git / commitdiff