-
Notifications
You must be signed in to change notification settings - Fork 4
/
main_master.f90
1086 lines (981 loc) · 37.4 KB
/
main_master.f90
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
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
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
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
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
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
511
512
513
514
515
516
517
518
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
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
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
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
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
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
!
! ParaGauss, a program package for high-performance computations of
! molecular systems
!
! Copyright (C) 2014 T. Belling, T. Grauschopf, S. Krüger,
! F. Nörtemann, M. Staufer, M. Mayer, V. A. Nasluzov, U. Birkenheuer,
! A. Hu, A. V. Matveev, A. V. Shor, M. S. K. Fuchs-Rohr, K. M. Neyman,
! D. I. Ganyushin, T. Kerdcharoen, A. Woiterski, A. B. Gordienko,
! S. Majumder, M. H. i Rotllant, R. Ramakrishnan, G. Dixit,
! A. Nikodem, T. Soini, M. Roderus, N. Rösch
!
! This program is free software; you can redistribute it and/or modify
! it under the terms of the GNU General Public License version 2 as
! published by the Free Software Foundation [1].
!
! 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.
!
! [1] http://www.gnu.org/licenses/gpl-2.0.html
!
! Please see the accompanying LICENSE file for further information.
!
!=====================================================================
! Public interface of module
!=====================================================================
subroutine main_master()
!
! This routine encodes the MASTER PLAN for all workers. Historically
! it was executed only by the rank-0 worker (usually called "master"
! processor), hence the name. By now it is executed in a parallel
! context.
!
! The routines which divide the LCGTO into its major parts are called
! from this level:
!
! (1) main_symm() -> symmetry_part
!
! (2) main_integral()
!
! (3) main_scf() -> does the SCF cycle including Hamiltonian,
! eigenvalue solution, Reoccupation of levels ...
!
! (4) post_scf_main() -> does the calculation of the final total
! xc-energy on the integration grid if operations_gradients is
! true, also the xc-contributions to the energy gradient will be
! calculated
!
! (5) main_gradient() -> calculation of the energy gradients (except
! xc-part)
!
! Subroutine called by: main()
!
!
! Author: TB, FN
! Date: 10/95
!
!===================================================================
! End of public interface of module
!===================================================================
!---------------------------------------------------------------------
! Modifications
!---------------------------------------------------------------------
!
! Modification (Please copy before editing)
! Author: TB
! Date: 12/95
! Description: added operations options for steering,
! debug output to be switched on and off
! and call to orbital_test.
! moved call to read_start to read_input.
!
! Modification (Please copy before editing)
! Author: TB
! Date: 5/96
! Description: added call to integral part
!
! Modification (Please copy before editing)
! Author: MS
! Date: 3/97
! Description: added call to main_gradient
!
! Modification (Please copy before editing)
! Author: HH
! Date: 10/97
! Description: added call to response_main()
!
! Modification (Please copy before editing)
! Author: AS
! Date: 7/98
! Description: pvm -> comm
!
! Modification (Please copy before editing)
! Author: AS
! Date: 11-12/99
! Description: added calls to calculating solvent effect
!
! Modification (Please copy before editing)
! Author: KN
! Date: 26/7/99
! Description: added call to main_gtensor
!
! Modification
! Author: TS
! Date: August 09
! Description: added call to empirical_methods (DFT-D)
!
! Modification (Please copy before editing)
! Author: ...
! Date: ...
! Description: ...
!---------------------------------------------------------------------
# include "def.h"
use type_module, only: i4_kind, r8_kind
use operations_module ! defines which operations are to be performed
use comm, only: comm_rank, comm_same
use filename_module, only: filesystem_is_parallel
use iounitadmin_module, only: output_unit, stdout_unit, &
write_to_output_units, write_to_trace_unit
use options_module, only: options_directaccess_integrals, &
options_integrals_on_file, update_hessian_iteration
use time_module, only: start_timer, stop_timer
use timer_module, only: timer_initialisation, timer_print_summary, &
timer_print_slavetiming
use output_module, only: output_timing_detailedsummary, &
output_timing_slaves, output_timing_summary
use potential_calc_module, only: charge_constr, esp_map, pdc, &
use_saved_densmatrix, V_electronic, calc_plane_grid, &
grid2space_2d, get_poten_and_shutdown_2d, calc_shell_grid, &
collect_poten_3d, calc_poten_derive_charges
use integralpar_module, only: integralpar_set, integralpar_cpksdervs, &
integralpar_int_part_name
use integralstore_module, only: integralstore_deallocate, &
integralstore_deallocate_pcm
use initialization, only: initialize_with_input, finalize_geometry
use xc_cntrl, only: xc_is_on=>is_on, xc_ANY
use post_scf_module, only: post_scf_main
use energy_calc_module, only: write_energies, get_energy
#ifdef FPP_DEBUG
use error_module, only: MyID
#endif
#ifdef WITH_RESPONSE
use response_module, only: response_main
#endif
use efield_module, only: efield_calculate_integrals, &
efield_applied, efield_intensity, efield_change
use unique_atom_module, only: unique_atom_iwork
use unique_atom_methods, only: unique_atom_make_gx
use occupation_module, only: occupation_symmetry_check
use convergence_module, only: convergence_max_geo_iteration
use properties_module, only: properties_main
#ifdef WITH_DFTPU
use dft_plus_u_module, only: dft_plus_u_output
#endif
#ifdef WITH_EPE
use epecom_module, only: get_epe_energies, &
epe_side_optimized_energy_prev, epe_side_optimized_energy
#ifdef NEW_EPE
use ewaldpc_module, only: epe_relaxation, get_qm_references, qm_ref_run
#else
use ewaldpc_module, only: epe_relaxation
#endif
#endif
#ifdef WITH_EFP
use efp_module, only: n_efp, read_gx_qm, def_efp_arrays, &
calc_X_points, calc_efield_points, print_id, qm_fixed
use efp_efp_module, only: efp_efp_energy
use efp_only_opt_module, only: geom_converged
use qmmm_interface_module, only: efp
#endif
use density_data_module, only: open_densmat
use solv_cavity_module, only: stop_solv
use potential_module, only: send_recv_space_point
! DONT use elec_static_field_module, and nothing breaks?
use symmetry, only: main_symm
use interfaces, only: main_integral
use interfaces, only: potential_calculate
use interfaces, only: main_molmech
#ifdef WITH_MOLMECH
use qmmm_interface_module, only: imomm, imomm_mm_large, &
imomm_mm_small, mm_run, qm_mm, qm_mm_1, qm_mm_run, &
qm_mm_1_task, qmmm_read_input, &
sum_up_grads_and_write_gx
use qmmm1_interface_module, only: def_qm_mm_1_tasks, &
read_gx_qmmm, qmmm2pc, qmfield_at_mm_points, write_gx_qmmm
#endif
use calc3c_switches, only: print_epe
#ifdef WITH_OPTIMIZER
use opt_data_module, only: filename_setup_opt
#endif
use induced_dipoles_module, only: calc_Pol_centers, dealloc_pol_center_inform
USE_MEMLOG
! --------------------------------------------------
implicit none
character (len=128) :: version = FPP_PARAGAUSS_VERS
integer (i4_kind) :: tasks=0 ! Number of tasks LEFT to be done
integer (i4_kind) :: loop, max_geo_loop ! loop -- Laufvariable fuer Runs
logical :: geometry_converged = .false. ! determines if Simol is converged or not
#ifdef WITH_EPE
real (r8_kind) :: energy
real (r8_kind) :: energy2, epe_latt_energy, cluster_regI
real (r8_kind) :: eshort
logical :: epe_side_energy_converged
#endif
logical :: use_dens_mat
DPRINT 'main_master: entered'
!
! NOTE: The code has been converted to SPMD (single program,
! multiple data) so that all workers execute this code. Some rarely
! used branches not covered by the testsuite have not been tested,
! though.
!
! If you are thinking about adding code, try making sure that can be
! executed on all workers. Usually it is sufficient to avoid
! writing to the same file and making sure the pre-requisites are
! available and the same on all workers. You may consider adding
! such code to main.f90 or other high-level procedure called from
! here, such as
!
! - initialize_with_input()
! - main_gradient()
! - properties_main()
! - finalize_geometry()
! - ...
!
!
! Print the version info and machine config, uses output_unit, so
! call it after initialize_environment():
!
call legal (version)
call write_to_trace_unit (" ")
call write_to_trace_unit (" -------------------------------------------")
call write_to_trace_unit (" ")
call write_to_trace_unit (" executing program : mainscf_" // trim (version))
call write_to_trace_unit (" ")
call write_to_trace_unit (" -------------------------------------------")
call write_to_trace_unit (" ")
#ifdef WITH_EPE
epe_side_optimized_energy_prev=0.0_r8_kind
#endif
#ifdef WITH_MOLMECH
qm_mm_1_task=0 !!!!!!!!!!!AS
#endif
![[=== MAIN LOOP OVER TASKS/GEOMETRIES ===================================
2001 CONTINUE ! an "entry point" for the task-loop
tasks = 1 ! will be eventualy incremented after read_input!
max_geo_loop = 1 ! will be eventualy reset to a higher value after read_input!
loop = 0
geometry_converged = .false.
geometry_loop: do while (tasks > 0)
loop = loop + 1
if (loop > max_geo_loop) then
! at loop 1 is not true anyway,
! after read_input max_geo_loop gets the proper value ...
call say ("maximum number of geo interations exceeded")
tasks = tasks - 1
cycle geometry_loop
endif
! reset memory couters, if using MEMLOG
MEMSET (0)
call start_timer (timer_initialisation)
! Otherwise there will be multiple copies printed:
if (output_unit > 0 .and. stdout_unit > 0) then ! yes, AND!
call write_to_output_units (" ------------------------------------")
call write_to_output_units (" - -")
call write_to_output_units (" - main_master: Run No. ", inte=loop)
call write_to_output_units (" - -")
call write_to_output_units (" ------------------------------------")
endif
! TODO: move call read_input() and
! subsequent control manupulations out of geometry loop!
DPRINT 'main_master: call read_input()'
call read_input (loop)
DPRINT 'main_master: .'
! 7 runs for intensity calculation by finite difference of forces
! in presence of electric field.
if (efield_intensity()) then
max_geo_loop = 7
tasks = tasks + 1
! change the direction of electric field in all 6 direction i.e +/-X, +/-Y and +/- Z
call efield_change (loop)
end if
#ifdef WITH_OPTIMIZER
call filename_setup_opt (optonly=.false.)
#endif
#ifdef WITH_MOLMECH
if (operations_qm_mm_new .and. qm_mm_1) then !!!!!!!!!!!AS
call def_qm_mm_1_tasks() !!!!!!!!!!!AS
end if
#endif
DPRINT 'main_master: options_directaccesa_integrals=', options_directaccess_integrals()
if (options_directaccess_integrals()) then
ASSERT (.not.filesystem_is_parallel)
endif
! ... and set the variable 'max_geo_loop' according to the input
if (operations_geo_opt .or. operations_optimizer_only) then
max_geo_loop = convergence_max_geo_iteration()
endif
![[=== decide wheather or not to do the hessian ===
select case (update_hessian_iteration)
case (0)
! do nothing, the default
case (1:)
! geometry optimization with regular updates of the hessian
if (MOD (loop-1, update_hessian_iteration) == 0) then
! at loop 1, update_hessian_iteration+1, 2*update_hessian_iteration+1, ...
call integralpar_set ('SecondDervs')
! sets module vars:
! integralpar_2dervs = .true.
! integralpar_cpksdervs = .true.
print *,'main_master: do second derivatives at loop', loop &
,'(every', update_hessian_iteration,'iterations)'
else
call integralpar_set ('NoSecondDervs')
! integralpar_2dervs = .false.
! integralpar_cpksdervs = .false.
endif
case (-1)
! frequencies after geometry optimization
if (loop == 1) then
! add a task of computing second derivatives after geometry optimization:
tasks = tasks + 1
endif
if (geometry_converged .or. loop == max_geo_loop) then
! enable second derivatives only after geometry is converged:
update_hessian_iteration = 1
call integralpar_set ('SecondDervs')
! sets module vars:
! integralpar_2dervs = .true.
! integralpar_cpksdervs = .true.
! nothing below seems to depend on it:
operations_geo_opt = .false.
print *,'main_master: do second derivatives at loop', loop &
,'(after geometry is converged)'
! mark the task completed (a bit early, of couse):
tasks = tasks - 1
endif
case default
print*,'ERROR: no such update_hessian_iteration=', update_hessian_iteration
ABORT ('no such case yet!')
end select
!]]================================================
#ifdef WITH_EPE
! EPE lattice calculations
if (operations_epe_lattice) then
call say ("calling EPE-lattice optimization")
call epe_lattice_optimization()
call stop_timer (timer_initialisation)
! DONT exit geometry_loop
tasks = tasks - 1
cycle geometry_loop
end if
#endif
if (operations_optimizer_only) then
call say ("calling opimizer")
geometry_converged = optimizer_step (1)
call stop_timer (timer_initialisation)
! DONT exit geometry_loop
tasks = tasks - 1
cycle geometry_loop
endif
#ifdef WITH_MOLMECH
if (operations_mol_mech) then !!!!!!!!!!!AS
call say ("calling molecular mechanics program") !!!!!!!!!!!AS
call main_molmech (mm_run, unique_atom_iwork, max_geo_loop) !!!!!!!!!!!AS
if (unique_atom_iwork > 0) then !!!!!!!!!!!AS
operations_geo_opt=.true. !!!!!!!!!!!AS
end if !!!!!!!!!!!AS
end if !!!!!!!!!!!AS
if (operations_qm_mm_new .and. (qm_mm .or. qm_mm_1)) then !!!!!!!!!!!AS
call say ("distributing data for QM+MM run")
call read_gx_qmmm()
call main_molmech (qmmm_read_input)
call qmmm2pc()
end if !!!!!!!!!!AS
#endif
#ifdef WITH_EFP
if (efp) then
call say ("EFP - reading QM atoms from GX")
call read_gx_qm()
end if
#endif
! call symmetry part
if (operations_symm .and. .not. operations_epe_lattice) then
call say ("calling symmetry part")
print_epe = loop .eq. 1
call main_symm()
endif
#ifdef WITH_EFP
!generation of arrays of external points for EFP calculation
if (efp) then
call say ("EFP - generation of arrays of external points")
call def_efp_arrays()
call calc_X_points()
if (calc_Pol_centers()) call calc_efield_points()
end if
#endif
! write input file if desired
if (operations_write_input .or. operations_get_input_out) then
call say ("write_input ")
call write_input()
call say ("write_input done")
endif
if (operations_get_input_out) then
call stop_timer (timer_initialisation)
! DONT exit geometry_loop
tasks = tasks - 1
cycle geometry_loop
end if
! generating suface charge distribution (solvation effect)
if (operations_solvation_effect) then
call say ("call build_mol_surfaces()")
! Subroutine with an implicit (unchecked) interface
! here. Though it runs on all workers, most of the work is
! done on master only:
call build_mol_surfaces() ! no comm?
if (stop_solv) then
call stop_timer (timer_initialisation)
! DONT exit geometry_loop
tasks = tasks - 1
cycle geometry_loop
end if
endif
#ifdef WITH_EFP
!calculating interfragment interactions (EFP)
if (efp .and. n_efp > 0) then
call say ("EFP - calculating interfragment interactions")
call efp_efp_energy (print_id)
end if
#endif
! write gx file
if (operations_make_gx) then
call say ("unique_atoms_make_gx")
#ifndef WITH_EPE
! FIXME: is it upto date?
call unique_atom_make_gx (iloop= 1)
#else
call unique_atom_make_epegx()
#endif
end if
!
! Initialize various modules in part by broadcasting
! initialisation information taken from input file. It is a
! convenient place to put the code to be executed on all workers
! ...
!
call say ("initialize_with_input")
call initialize_with_input()
call stop_timer (timer_initialisation)
!
! At this point all information should be in place that is needed
! to check the fixed occupation numbers.
!
! Check the 'n_nonempty_irreps'-specification in
! the namelist OCCUPATION with the variable 'symmetry_data_n_irreps'
! after the symmetry part has been run:
!
call occupation_symmetry_check()
use_dens_mat = .false.
if (operations_potential .and. esp_map) then
if (.not. V_electronic .or. use_saved_densmatrix) goto 1111
elseif (operations_potential .and. pdc) then
if (charge_constr .and. use_saved_densmatrix) then
use_dens_mat = .true.
goto 1111
endif
endif
! do integral part
if (operations_integral) then
call say ("Starting the Integral Part")
!
! Not all integrals are needed in a property run without prior
! SCF. On the other hand if SCF is performed anyway, no extra
! integrals are needed for properties (that is the theory):
!
if (operations_scf) then
call say ("call integralpar_set (Normal)")
call integralpar_set ('Normal')
elseif (operations_properties) then
call say ("call integralpar_set (Properties)")
call integralpar_set ('Properties')
else
ABORT ("ever happens?")
endif
call say ("done")
call main_integral ()
call say ("done with Integral Part")
end if
! calculate integrals of external electrical field if one is applied
if (efield_applied()) then
call say ("Calculating integrals of external electrical field")
call efield_calculate_integrals ()
endif
! calculate integrals of electrostatic potential
if (operations_solvation_effect .and. operations_integral) then
call say ("Calculate integrals of electrostatic potential ...")
call send_recv_space_point()
call potential_calculate ('Solvation')
call say ("Done with integrals of electrostatic potential.")
endif
#ifdef WITH_EFP
if (calc_Pol_centers() .and. operations_integral) then
call field_calculate ()
end if
#endif
! do scf part
if (operations_scf) then
MEMSET (0)
call say ("Starting the main SCF routine ...")
call main_scf()
call say ("Done with the main SCF routine.")
MEMSET (0)
endif
! Calculate and print dipole moments:
if (operations_dipole .or. operations_gtensor .or. operations_hfc) then
call say ("call main_dipole()")
call main_dipole()
call say ("done main_dipole()")
endif
! Potential derived charges here (not the same as solvation, even
! though the same buildign blocks are used). FIXME: too much
! logic/branches for a master plan. Leave here only the entry
! points, move logic to the respective modules:
1111 if (operations_potential) then
call say ("Starting the potential routines ...")
if (esp_map) then
if (comm_rank() == 0) then
call calc_plane_grid()
endif
call grid2space_2d()
if (V_electronic) then
call say ("call potential_calculate (Vel)")
if (use_saved_densmatrix) then
call open_densmat() ! no comm, reads disk
endif
call potential_calculate ('Potential')
endif
call get_poten_and_shutdown_2d()
if (.not. V_electronic) then
tasks = tasks - 1
cycle geometry_loop
endif
elseif (pdc) then
call say ("call potential_calculate (PDC)")
call calc_shell_grid()
if (use_dens_mat) then
call open_densmat() ! no comm, reads disk
endif
call potential_calculate ('Potential')
call collect_poten_3d()
if (comm_rank() == 0) then
call calc_poten_derive_charges() ! no comm
endif
endif
call say ("Done with the potential routines.")
endif
#ifdef WITH_DFTPU
!
! DFT+U output, does nothing if not in use:
!
call dft_plus_u_output ()
#endif
! Do properties before deallocating the integral storage.
! Otherwise a read_overlap() called from properties_main() will
! fail. Slaves execute properties_main() too:
if (operations_properties) then
call say ("call properties_main()")
call properties_main()
call say ("done properties_main()")
end if
! Kin and Nuc are now always allocated, dellaocate:
call integralstore_deallocate (deallocate_kin=.true., deallocate_nuc=.true.)
if (.not. options_integrals_on_file() .and. .not. integralpar_cpksdervs) then
! if integralpar_cpksdervs -- deallocate after
! cpks_g4constructs()
! Deallocate integral storage ...
call integralstore_deallocate()
! Deallocate PCM-integral storage (if been allocated) ...
call integralstore_deallocate_pcm()
end if
! FIXME: please specify the goal of GOTO by words as well,
! not just by the number!
! GO TO: .... because ...
if (operations_potential .and. use_saved_densmatrix) goto 1112
! do Post Scf calculation of Exchange Energy
if (operations_post_scf) then
! only if XC /= off:
if (xc_is_on (xc_ANY)) then
call say ("call post_scf_main()")
call post_scf_main()
call say ("done post_scf_main()")
endif
call say ("write_energies")
call write_energies (output_unit, post_scf=.true.)
end if
#ifdef WITH_MOLMECH
!Calculations electrostatic field produced by QM cluster
!at MM atoms
if (operations_qm_mm_new .and. qm_mm) then !!!!!!!!!!!!AS
call QMfield_at_mm_points() !!!!!!!!!!!!AS
end if !!!!!!!!!!!!AS
#endif
! Do post-SCF calculation of matrix elements needed for response
! calculations with tdfrt response program:
ASSERT(comm_same(operations_response))
if (operations_response) then
#ifdef WITH_RESPONSE
call say ("call response_main()")
call response_main ()
call say ("done response_main()")
#else
ABORT ('recompile -DWITH_RESPONSE')
#endif
endif
MEMSET (0)
#ifndef WITH_EPE
if (operations_gradients) then
call say ("call main_gradient()")
call main_gradient (loop) ! (1)
call say ("done main_gradient()")
endif
#else
! do EPE calculations, FIXME: why inlining so much staff into this high
! level sub?
if (operations_qm_epe .and. epe_relaxation) then
call say ("main_epe_block")
call main_epe_block()
call get_energy (tot=energy)
energy2 = energy
call get_epe_energies (lattice_energy=epe_latt_energy, &
epg_cluster_reg_I=cluster_regI, eshort_coupling_au=eshort)
energy = energy + epe_latt_energy
print*,'energy, energy2, eshort', energy, energy2, eshort
print*,'epe_side_optimized_energy', energy+eshort
epe_side_optimized_energy = energy + eshort
print*,'cluster_regI', cluster_regI
print*,'epe_latt_energy', epe_latt_energy
call write_to_trace_unit ('epe_convergence_check')
call epe_convergence_check (epe_side_energy_converged, loop)
endif
! Calculate gradients
if (operations_gradients .and. .not. epe_relaxation .or. &
epe_relaxation .and. epe_side_energy_converged) then
!
! Regular branch here!
!
call say ("Starting main_gradient() ...")
if (epe_relaxation .and. epe_side_energy_converged) then
call write_to_trace_unit ('epe_relaxation .and. epe_side_energy_converged')
endif
call main_gradient (loop) ! (2)
call say ("Done with the integral part for gradients routine.")
elseif (operations_gradients) then
!
! FIXME: clean up is the task of finalize_geometry()
! that is called anyway. Why doing it here?
!
ABORT ('please adapt')
endif
#endif
#ifdef WITH_MOLMECH
if (operations_qm_mm_new) then !!!!!!!!!!!!!!AS
if (imomm) then
call say ("calling molecular mechanics module to perform IMOMM job")
call main_molmech (imomm_mm_small)
call main_molmech (imomm_mm_large)
call say ("QMMM job: summing up different gradients and writing gxfile")
call sum_up_grads_and_write_gx()
elseif (qm_mm .or. (qm_mm_1 .and. qm_mm_1_task ==0)) then
call say ("calling molecular mechanics module to perform QM_MM (1) job")
call main_molmech (qm_mm_run)
call write_gx_qmmm()
end if
endif !!!!!!!!!!!!!!AS
#endif
MEMSET (0)
!
! Various shutdown and deallocation work:
!
call finalize_geometry()
1112 continue
! Otherwise there will be multiple copies printed:
if (output_unit > 0 .and. stdout_unit > 0) then ! yes, AND!
call write_to_output_units (" ------------------------------------")
call write_to_output_units (" - -")
call write_to_output_units (" - main_master: End of Run No. ", inte=loop)
call write_to_output_units (" - -")
call write_to_output_units (" ------------------------------------")
endif
!
! If max_geo_iteration was set to zero in the input,
! dont even try to run optimizer. This will not overwrite
! gxfile with the new geometry updated by optimizer algorithm.
! May be usefull for work with alternative external optimizers:
!
if (operations_geo_opt .and. loop > max_geo_loop) then
WARN ('loop leaped beyond max_geo_loop')
tasks = tasks - 1
exit geometry_loop
! there is not point to cycle geometry_loop so far as the first
! condition it checks after entry is again "loop > max_geo_loop"
endif
if (operations_geo_opt .or. operations_gx_test) then
#ifdef WITH_EFP
if (efp .and. qm_fixed) then
if (geom_converged (loop)) then
tasks = tasks - 1
end if
cycle geometry_loop
end if
#endif
call say ("call optimizer_step()")
geometry_converged = optimizer_step (unique_atom_iwork)
if (geometry_converged) then
call say ("Geometry converged")
! DONT exit geometry_loop
tasks = tasks - 1
! nothing below seems to depend on it:
! operations_geo_opt = .false.
! operations_gx_test = .false.
! But so far it doesnt work as the read_input()
! resets them to true again!
endif
#ifdef NEW_EPE
if (operations_qm_epe .and. geometry_converged) then
if (get_qm_references .and. .not. qm_ref_run) then
tasks = tasks + 1
qm_ref_run=.true.
!one more cycle to calculate and save
!epe_reference and pg_epe_reference
end if
endif
#endif
else
! a default calculation (energy/gradients) requires one run:
tasks = tasks - 1
endif
enddo geometry_loop
#ifdef WITH_MOLMECH
if (operations_qm_mm_new .and. qm_mm_1) then
qm_mm_1_task=qm_mm_1_task+1
! FIXME: I need to repeat all once again because ...
if (qm_mm_1_task == 1) goto 2001 ! enter the task loop again
! QUESTION: is this logic only to enter geometry_loop twice?
end if
#endif
!]]=== eof MAIN LOOP OVER TASKS/GEOMETRIES ================================
! print timing
if (output_timing_summary .or. output_timing_slaves .or. output_timing_detailedsummary) then
call say ("printing timing")
if (output_timing_summary .or. output_timing_detailedsummary) &
call timer_print_summary (integralpar_int_part_name)
if (output_timing_slaves) &
call timer_print_slavetiming (integralpar_int_part_name)
endif
call say ("done")
contains
#ifdef WITH_EPE
subroutine epe_convergence_check (epe_side_energy_converged, i_iter)
use epecom_module, only: epe_rel_converged, &
epe_side_optimized_energy, epeside_energy_limit, &
epe_side_optimized_energy_prev, epe_basic_action=>basic_action
use filename_module, only: data_dir
implicit none
logical, intent (out) :: epe_side_energy_converged
integer (i4_kind), intent (in) :: i_iter
! *** end of interface ***
if (epe_relaxation .and. .not. epe_basic_action .eq. 0) then
inquire (file=trim (data_dir) // "/epe_rel_unconverged", &
exist=epe_rel_converged)
epe_rel_converged = .not. epe_rel_converged
print *, 'epe_convergence_check: epe_relaxation, epe_rel_converged', &
epe_relaxation, epe_rel_converged
epe_side_energy_converged = abs (epe_side_optimized_energy_prev - &
epe_side_optimized_energy) .lt. 0.00002 !? embed_convergence_limit
if (.not. epe_side_energy_converged) &
epe_side_energy_converged = abs (epe_side_optimized_energy_prev - &
epe_side_optimized_energy) .lt. 0.0004 .and. i_iter .gt. 10
print *, 'epe_side_energy_jump' , &
epe_side_optimized_energy_prev - epe_side_optimized_energy, &
epeside_energy_limit
epe_side_optimized_energy_prev = epe_side_optimized_energy
if (.not. epe_rel_converged .or. .not. epe_side_energy_converged) then
print *, 'optimizer: epe relaxation is still not converged', &
epe_rel_converged, epe_side_energy_converged, &
epe_side_optimized_energy_prev - &
epe_side_optimized_energy
call write_to_trace_unit ('.not. epe_side_energy_converged')
else
call write_to_trace_unit ('epe_side_energy_converged')
endif
else
epe_side_energy_converged = epe_basic_action .eq. 0
endif
end subroutine epe_convergence_check
#endif
function optimizer_step (geo_loop) result (converged)
!
! Executed by all workers, though most of the work is done
! serially. The input "geo_loop" schould be the same everywhere.
! The output will be the same on all workers.
!
use comm, only: comm_rank, comm_bcast, comm_barrier
implicit none
integer (i4_kind), intent (in) :: geo_loop
logical :: converged
! *** end of interface ***
call comm_barrier() ! paranoya
! One of us does the work that includes some file-system mangling:
if (comm_rank() == 0) then
call do_optimizer_step (geo_loop, converged)
endif
! Broadcast the result:
call comm_bcast (converged)
call comm_barrier() ! paranoya
end function optimizer_step
subroutine do_optimizer_step (geo_loop, geo_conv)
!
! FIXME: down the call chain there is some file-system mangling
! such as (re)writing gx- and hessian files. This code may not
! work as intended if executed by more than one worker.
!
#ifdef WITH_EPE
use epecom_module, only: cross_boundary_3b, epe_rel_converged, &
epe_basic_action => basic_action
#endif
#ifdef WITH_OPTIMIZER
use optimizer, only: main_opt
use operations_module, only: operations_task, namelist_tasks_used
#endif
implicit none
integer (i4_kind), intent (in) :: geo_loop
logical, intent (out) :: geo_conv
! *** end of interface ***
logical :: conv, stop_after_eperelaxation, convert_internal
character (len=32) :: optimizer_task='GeoOpt'
geo_conv = .false.
#ifdef WITH_EPE
if (operations_qm_epe) then
stop_after_eperelaxation=.true.
if (epe_relaxation .and. .not. epe_basic_action .eq. 0) then
if (.not. epe_rel_converged .or. .not. epe_side_energy_converged) then
print*,'optimizer: epe relaxation is still not converged'
call write_to_trace_unit ('optimizer epe relaxation is still not converged')
return
endif
endif
end if
#endif
#ifdef WITH_OPTIMIZER
if (namelist_tasks_used) then
optimizer_task=operations_task
else
optimizer_task='GeoOpt'
WARN ('assuming GeoOpt in optimizer')
endif
print*,'optimizer: call main_opt(', optimizer_task,')'
convert_internal=.false.
#ifdef WITH_EPE
if (epe_relaxation) then
call main_opt (task=optimizer_task , converged=conv &
, stop_after_eperelaxation=stop_after_eperelaxation &
, convert_internal=convert_internal &
, cross_boundary_3b=cross_boundary_3b)
else
#endif
call main_opt (task=optimizer_task , converged=conv &
, stop_after_eperelaxation=stop_after_eperelaxation &
, convert_internal=convert_internal)
#ifdef WITH_EPE
endif
#endif
if (convert_internal) &
call main_opt (task=optimizer_task , converged=conv &
, stop_after_eperelaxation=stop_after_eperelaxation &
, convert_internal=convert_internal)
! returns converged=true if converged
#else
DPRINT 'optimizer: DONT call main_opt()'
ABORT ('recompile with -DWITH_OPTIMIZER')
#endif
#ifdef WITH_EPE
if (operations_qm_epe) then
if (epe_relaxation .and. stop_after_eperelaxation) then
conv=.true.
endif
end if
#endif
! set output flag:
geo_conv = conv
if (.not. conv) then
call write_to_output_units&
("optimizer: geometry not yet converged in loop", inte =geo_loop)
call write_to_trace_unit&
("optimizer: geometry not yet converged in loop", inte =geo_loop)
endif
end subroutine do_optimizer_step
subroutine legal (version)
use comm_module, only: comm_print_conf
implicit none