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
// *****************************************************************************
/*!
  \file      src/Inciter/DiagCG.cpp
  \copyright 2012-2015 J. Bakosi,
             2016-2018 Los Alamos National Security, LLC.,
             2019-2021 Triad National Security, LLC.
             All rights reserved. See the LICENSE file for details.
  \brief     DiagCG for a PDE system with continuous Galerkin without a matrix
  \details   DiagCG advances a system of partial differential equations (PDEs)
    using continuous Galerkin (CG) finite element (FE) spatial discretization
    (using linear shapefunctions on tetrahedron elements) combined with a time
    stepping scheme that is equivalent to the Lax-Wendroff (LW) scheme within
    the unstructured-mesh FE context and treats discontinuities with
    flux-corrected transport (FCT). Only the diagonal entries of the left-hand
    side matrix are non-zero thus it does not need a matrix-based linear solver.
  \see The documentation in DiagCG.h.
*/
// *****************************************************************************

#include "DiagCG.hpp"
#include "Vector.hpp"
#include "Reader.hpp"
#include "ContainerUtil.hpp"
#include "UnsMesh.hpp"
#include "Inciter/InputDeck/InputDeck.hpp"
#include "DerivedData.hpp"
#include "CGPDE.hpp"
#include "Discretization.hpp"
#include "DistFCT.hpp"
#include "DiagReducer.hpp"
#include "NodeBC.hpp"
#include "Refiner.hpp"
#include "Reorder.hpp"
#include "Integrate/Mass.hpp"
#include "FieldOutput.hpp"

namespace inciter {

extern ctr::InputDeck g_inputdeck;
extern ctr::InputDeck g_inputdeck_defaults;
extern std::vector< CGPDE > g_cgpde;

} // inciter::

using inciter::DiagCG;

DiagCG::DiagCG( const CProxy_Discretization& disc,
                const CProxy_Ghosts&,
                const std::map< int, std::vector< std::size_t > >& bface,
                const std::map< int, std::vector< std::size_t > >& bnode,
                const std::vector< std::size_t >& triinpoel ) :
  m_disc( disc ),
  m_initial( 1 ),
  m_nlhs( 0 ),
  m_nrhs( 0 ),
  m_nnorm( 0 ),
  m_bnode( bnode ),
  m_bface( bface ),
  m_triinpoel( tk::remap(triinpoel,Disc()->Lid()) ),
  m_u( Disc()->Gid().size(),
       g_inputdeck.get< tag::component >().nprop( Disc()->MeshId() ) ),
  m_ul( m_u.nunk(), m_u.nprop() ),
  m_du( m_u.nunk(), m_u.nprop() ),
  m_ue( Disc()->Inpoel().size()/4, m_u.nprop() ),
  m_lhs( m_u.nunk(), m_u.nprop() ),
  m_rhs( m_u.nunk(), m_u.nprop() ),
  m_bcdir(),
  m_lhsc(),
  m_rhsc(),
  m_difc(),
  m_vol(),
  m_bnorm(),
  m_bnormc(),
  m_symbcnodemap(),
  m_symbcnodes(),
  m_farfieldbcnodes(),
  m_diag(),
  m_boxnodes(),
  m_dtp( m_u.nunk(), 0.0 ),
  m_tp( m_u.nunk(), g_inputdeck.get< tag::discr, tag::t0 >() ),
  m_finished( 0 ),
  m_nusermeshblk( 0 ),
  m_nodeblockid()
// *****************************************************************************
//  Constructor
//! \param[in] disc Discretization proxy
//! \param[in] bface Boundary-faces mapped to side set ids
//! \param[in] bnode Boundary-node lists mapped to side set ids
//! \param[in] triinpoel Boundary-face connectivity
// *****************************************************************************
{
  usesAtSync = true;    // enable migration at AtSync

  auto d = Disc();

  // Perform optional operator-access-pattern mesh node reordering
  if (g_inputdeck.get< tag::discr, tag::operator_reorder >()) {

    const auto& inpoel = d->Inpoel();

    // Create new local ids based on access pattern of PDE operators
    std::unordered_map< std::size_t, std::size_t > map;
    std::size_t n = 0;

    for (std::size_t e=0; e<inpoel.size()/4; ++e)
      for (std::size_t i=0; i<4; ++i) {
        std::size_t o = inpoel[e*4+i];
        if (map.find(o) == end(map)) map[o] = n++;<--- Searching before insertion is not necessary.
      }

    Assert( map.size() == d->Gid().size(), "Map size mismatch" );

    // Remap data in bound Discretization object
    d->remap( map );
    // Remap local ids in DistFCT
    d->FCT()->remap( *d );
    // Remap boundary triangle face connectivity
    tk::remap( m_triinpoel, map );

  }

  // Activate SDAG wait
  thisProxy[ thisIndex ].wait4norm();
  thisProxy[ thisIndex ].wait4lhs();

  // Query nodes at which symmetry BCs are specified
  auto bn = d->bcnodes< tag::bc, tag::bcsym >( m_bface, m_triinpoel );
  // Query nodes at which farfield BCs are specified
  auto far = d->bcnodes< tag::bc, tag::bcfarfield >( m_bface, m_triinpoel );

  // Merge BC data where boundary-point normals are required
  for (const auto& [s,n] : far) bn[s].insert( begin(n), end(n) );

  // Compute boundary point normals
  bnorm( bn );
}

void
DiagCG::bnorm( const std::unordered_map< int,
                 std::unordered_set< std::size_t > >& bcnodes )
// *****************************************************************************
//  Compute boundary point normals
//! \param[in] bcnodes Local node ids associated to side set ids at which BCs
//!    are set that require normals
// *****************************************************************************
{
  auto d = Disc();

  m_bnorm = cg::bnorm( m_bface, m_triinpoel, d->Coord(), d->Gid(), bcnodes );

  // Send our nodal normal contributions to neighbor chares
  if (d->NodeCommMap().empty())
    comnorm_complete();
  else
    for (const auto& [ neighborchare, sharednodes ] : d->NodeCommMap()) {
      std::unordered_map< int,
        std::unordered_map< std::size_t, std::array< tk::real, 4 > > > exp;
      for (auto i : sharednodes) {
        for (const auto& [s,norms] : m_bnorm) {
          auto j = norms.find(i);
          if (j != end(norms)) exp[s][i] = j->second;
        }
      }
      thisProxy[ neighborchare ].comnorm( exp );
    }

  ownnorm_complete();
}

void
DiagCG::comnorm( const std::unordered_map< int,
      std::unordered_map< std::size_t, std::array< tk::real, 4 > > >& innorm )
// *****************************************************************************
// Receive boundary point normals on chare-boundaries
//! \param[in] innorm Incoming partial sums of boundary point normal
//!   contributions to normals (first 3 components), inverse distance squared
//!   (4th component)
// *****************************************************************************
{
  // Buffer up inccoming boundary-point normal vector contributions
  for (const auto& [s,norms] : innorm) {
    auto& bnorms = m_bnormc[s];
    for (const auto& [p,n] : norms) {
      auto& bnorm = bnorms[p];
      bnorm[0] += n[0];
      bnorm[1] += n[1];
      bnorm[2] += n[2];
      bnorm[3] += n[3];
    }
  }

  if (++m_nnorm == Disc()->NodeCommMap().size()) {
    m_nnorm = 0;
    comnorm_complete();
  }
}

void
DiagCG::normfinal()
// *****************************************************************************
//  Finish computing boundary point normals
// *****************************************************************************
{
  auto d = Disc();
  const auto& lid = d->Lid();

  // Combine own and communicated contributions to boundary point normals
  for (const auto& [s,norms] : m_bnormc)
    for (const auto& [p,n] : norms) {
      auto k = m_bnorm.find(s);
      if (k != end(m_bnorm)) {
        auto j = k->second.find(p);
        if (j != end(k->second)) {
          auto& norm = j->second;
          norm[0] += n[0];
          norm[1] += n[1];
          norm[2] += n[2];
          norm[3] += n[3];
        }
      }
    }
  tk::destroy( m_bnormc );

  // Divie summed point normals by the sum of inverse distance squared
  for (auto& [s,norms] : m_bnorm)
    for (auto& [p,n] : norms) {
      n[0] /= n[3];
      n[1] /= n[3];
      n[2] /= n[3];
      Assert( (n[0]*n[0] + n[1]*n[1] + n[2]*n[2] - 1.0) <
              std::numeric_limits< tk::real >::epsilon(), "Non-unit normal" );
    }

  // Replace global->local ids associated to boundary point normals
  decltype(m_bnorm) bnorm;
  for (auto& [s,norms] : m_bnorm) {
    auto& bnorms = bnorm[s];
    for (auto&& [g,n] : norms)
      bnorms[ tk::cref_find(lid,g) ] = std::move(n);
  }
  m_bnorm = std::move(bnorm);

  // Prepare unique set of symmetry BC nodes
  m_symbcnodemap = d->bcnodes< tag::bc, tag::bcsym >( m_bface, m_triinpoel );
  for (const auto& [s,nodes] : m_symbcnodemap)
    m_symbcnodes.insert( begin(nodes), end(nodes) );

  // Prepare unique set of farfield BC nodes
  auto far = d->bcnodes< tag::bc, tag::bcfarfield >( m_bface, m_triinpoel );
  for (const auto& [s,nodes] : far)
    m_farfieldbcnodes.insert( begin(nodes), end(nodes) );

  // If farfield BC is set on a node, will not also set symmetry BC
  for (auto fn : m_farfieldbcnodes) {
    m_symbcnodes.erase(fn);
    for (auto& [s,nodes] : m_symbcnodemap) nodes.erase(fn);
  }

  // Signal the runtime system that the workers have been created
  std::vector< std::size_t > meshdata{ m_initial, d->MeshId() };
  contribute( meshdata, CkReduction::sum_ulong,
    CkCallback(CkReductionTarget(Transporter,comfinal), Disc()->Tr()) );
}

void
DiagCG::registerReducers()
// *****************************************************************************
//  Configure Charm++ reduction types initiated from this chare array
//! \details Since this is a [initnode] routine, the runtime system executes the
//!   routine exactly once on every logical node early on in the Charm++ init
//!   sequence. Must be static as it is called without an object. See also:
//!   Section "Initializations at Program Startup" at in the Charm++ manual
//!   http://charm.cs.illinois.edu/manuals/html/charm++/manual.html.
// *****************************************************************************
{
  NodeDiagnostics::registerReducers();
}

void
DiagCG::ResumeFromSync()
// *****************************************************************************
//  Return from migration
//! \details This is called when load balancing (LB) completes. The presence of
//!   this function does not affect whether or not we block on LB.
// *****************************************************************************
{
  if (Disc()->It() == 0) Throw( "it = 0 in ResumeFromSync()" );

  if (!g_inputdeck.get< tag::cmd, tag::nonblocking >()) next();
}

void
DiagCG::setup()
// *****************************************************************************
// Set and output initial conditions and mesh to file
// *****************************************************************************
{
  auto d = Disc();

  // Determine nodes inside user-defined IC box
  g_cgpde[d->MeshId()].IcBoxNodes( d->Coord(), d->Inpoel(),
    d->ElemBlockId(), m_boxnodes, m_nodeblockid, m_nusermeshblk );

  // Compute volume of user-defined box IC
  d->boxvol( m_boxnodes, m_nodeblockid, m_nusermeshblk );

  // Query time history field output labels from all PDEs integrated
  const auto& hist_points = g_inputdeck.get< tag::history, tag::point >();
  if (!hist_points.empty()) {
    std::vector< std::string > histnames;
    auto n = g_cgpde[d->MeshId()].histNames();
    histnames.insert( end(histnames), begin(n), end(n) );
    d->histheader( std::move(histnames) );
  }
}

void
DiagCG::box( tk::real v, const std::vector< tk::real >& blkvols )
// *****************************************************************************
// Receive total box IC volume and set conditions in box
//! \param[in] v Total volume within user-specified box
//! \param[in] blkvols Vector of mesh block discrete volumes with user ICs
// *****************************************************************************
{
  Assert(blkvols.size() == m_nusermeshblk,
    "Incorrect size of block volume vector");
  auto d = Disc();
  const auto& coord = d->Coord();

  // Store user-defined box IC volume
  d->Boxvol() = v;
  d->MeshBlkVol() = blkvols;

  // Set initial conditions for all PDEs
  g_cgpde[d->MeshId()].initialize( coord, m_u, d->T(), d->Boxvol(), m_boxnodes,
    d->MeshBlkVol(), m_nodeblockid );

  // Apply symmetry BCs on initial conditions
  g_cgpde[d->MeshId()].symbc( m_u, coord, m_bnorm, m_symbcnodes );
  // Apply farfield BCs on initial conditions
  g_cgpde[d->MeshId()].farfieldbc( m_u, coord, m_bnorm, m_farfieldbcnodes );

  // Output initial conditions to file (regardless of whether it was requested)
  writeFields( CkCallback(CkIndex_DiagCG::init(), thisProxy[thisIndex]) );
}

void
DiagCG::init()
// *****************************************************************************
// Initially compute left hand side diagonal matrix
// *****************************************************************************
{
  lhs();
}

void
DiagCG::next()
// *****************************************************************************
// Continue to next time step
// *****************************************************************************
{
  dt();
}

void
DiagCG::lhs()
// *****************************************************************************
// Compute the left-hand side of transport equations
// *****************************************************************************
{
  auto d = Disc();

  // Compute lumped mass lhs required for both high and low order solutions
  m_lhs = tk::lump( m_u.nprop(), d->Coord(), d->Inpoel() );

  if (d->NodeCommMap().empty())
    comlhs_complete();
  else // send contributions of lhs to chare-boundary nodes to fellow chares
    for (const auto& [c,n] : d->NodeCommMap()) {
      std::vector< std::vector< tk::real > > l( n.size() );
      std::size_t j = 0;
      for (auto i : n) l[ j++ ] = m_lhs[ tk::cref_find(d->Lid(),i) ];
      thisProxy[c].comlhs( std::vector<std::size_t>(begin(n),end(n)), l );
    }

  ownlhs_complete();
}

void
DiagCG::comlhs( const std::vector< std::size_t >& gid,
                const std::vector< std::vector< tk::real > >& L )
// *****************************************************************************
//  Receive contributions to left-hand side diagonal matrix on chare-boundaries
//! \param[in] gid Global mesh node IDs at which we receive LHS contributions
//! \param[in] L Partial contributions of LHS to chare-boundary nodes
//! \details This function receives contributions to m_lhs, which stores the
//!   diagonal (lumped) mass matrix at mesh nodes. While m_lhs stores
//!   own contributions, m_lhsc collects the neighbor chare contributions during
//!   communication. This way work on m_lhs and m_lhsc is overlapped. The two
//!   are combined in lhsmerge().
// *****************************************************************************
{
  Assert( L.size() == gid.size(), "Size mismatch" );

  using tk::operator+=;

  for (std::size_t i=0; i<gid.size(); ++i)
    m_lhsc[ gid[i] ] += L[i];

  if (++m_nlhs == Disc()->NodeCommMap().size()) {
    m_nlhs = 0;
    comlhs_complete();
  }
}

void
DiagCG::lhsmerge()
// *****************************************************************************
// The own and communication portion of the left-hand side is complete
// *****************************************************************************
{
  // Combine own and communicated contributions to left hand side
  for (const auto& b : m_lhsc) {
    auto lid = tk::cref_find( Disc()->Lid(), b.first );
    for (ncomp_t c=0; c<m_lhs.nprop(); ++c)
      m_lhs(lid,c) += b.second[c];
  }

  // Clear receive buffer
  tk::destroy(m_lhsc);

  // Continue after lhs is complete
  if (m_initial) {
    // Start timer measuring time stepping wall clock time
    Disc()->Timer().zero();
    // Zero grind-timer
    Disc()->grindZero();
    // Continue to next time step
    next();
  } else {
    lhs_complete();
  }
}

void
DiagCG::dt()
// *****************************************************************************
// Compute time step size
// *****************************************************************************
{
  tk::real mindt = std::numeric_limits< tk::real >::max();

  auto const_dt = g_inputdeck.get< tag::discr, tag::dt >();
  auto def_const_dt = g_inputdeck_defaults.get< tag::discr, tag::dt >();
  auto eps = std::numeric_limits< tk::real >::epsilon();

  auto d = Disc();

  // use constant dt if configured
  if (std::abs(const_dt - def_const_dt) > eps) {

    mindt = const_dt;

  } else {      // compute dt based on CFL

    // find the minimum dt across all PDEs integrated
    auto eqdt = g_cgpde[d->MeshId()].dt( d->Coord(), d->Inpoel(), d->T(),
      d->Dtn(), m_u, d->Vol(), d->Vol() );
    if (eqdt < mindt) mindt = eqdt;

  }

  // Actiavate SDAG waits for time step
  thisProxy[ thisIndex ].wait4rhs();
  thisProxy[ thisIndex ].wait4out();

  // Activate SDAG-waits for FCT
  d->FCT()->next();

  // Contribute to minimum dt across all chares the advance to next step
  contribute( sizeof(tk::real), &mindt, CkReduction::min_double,
              CkCallback(CkReductionTarget(DiagCG,advance), thisProxy) );
}

void
DiagCG::advance( tk::real newdt, tk::real )
// *****************************************************************************
// Advance equations to next time step
//! \param[in] newdt Size of this new time step
// *****************************************************************************
{
  auto d = Disc();

  // Set new time step size
  d->setdt( newdt );

  // Compute rhs for next time step
  rhs();
}

void
DiagCG::rhs()
// *****************************************************************************
// Compute right-hand side of transport equations
// *****************************************************************************
{
  auto d = Disc();
  const auto& lid = d->Lid();
  const auto& inpoel = d->Inpoel();

  // Sum nodal averages to elements (1st term of gather)
  m_ue.fill( 0.0 );
  for (std::size_t e=0; e<inpoel.size()/4; ++e)
    for (ncomp_t c=0; c<m_u.nprop(); ++c)
      for (std::size_t a=0; a<4; ++a)
        m_ue(e,c) += m_u(inpoel[e*4+a],c)/4.0;

  // Scatter the right-hand side for chare-boundary cells only
  m_rhs.fill( 0.0 );
  g_cgpde[d->MeshId()].rhs( d->T(), d->Dt(), d->Coord(), d->Inpoel(), m_u, m_ue,
    m_rhs );

  // Compute mass diffusion
  auto dif = d->FCT()->diff( *d, m_u );

  // Query and match user-specified boundary conditions to side sets
  m_bcdir = match( d->MeshId(), m_u.nprop(), d->T(), d->Dt(), m_tp, m_dtp,
                   d->Coord(), lid, m_bnode, /* increment = */ true );

  // Send rhs data on chare-boundary nodes to fellow chares
  if (d->NodeCommMap().empty())
    comrhs_complete();
  else  // send contributions of rhs to chare-boundary nodes to fellow chares
    for (const auto& [c,n] : d->NodeCommMap()) {
      std::vector< std::vector< tk::real > > r( n.size() );
      std::vector< std::vector< tk::real > > D( n.size() );
      std::size_t j = 0;
      for (auto i : n) {
        auto k = tk::cref_find( lid, i );
        r[j] = m_rhs[k];
        D[j] = dif[k];
        ++j;
      }
      thisProxy[c].comrhs( std::vector<std::size_t>(begin(n),end(n)), r, D );
    }

  ownrhs_complete( dif );
}

void
DiagCG::comrhs( const std::vector< std::size_t >& gid,
                const std::vector< std::vector< tk::real > >& R,
                const std::vector< std::vector< tk::real > >& D )
// *****************************************************************************
//  Receive contributions to right-hand side vector on chare-boundaries
//! \param[in] gid Global mesh node IDs at which we receive RHS contributions
//! \param[in] R Partial contributions of RHS to chare-boundary nodes
//! \param[in] D Partial contributions to chare-boundary nodes
//! \details This function receives contributions to m_rhs, which stores the
//!   right hand side vector at mesh nodes. While m_rhs stores own
//!   contributions, m_rhsc collects the neighbor chare contributions during
//!   communication. This way work on m_rhs and m_rhsc is overlapped. The two
//!   are combined in solve(). This function also receives contributions to
//!   mass diffusion term of the right hand side vector at mesh nodes.
// *****************************************************************************
{
  Assert( R.size() == gid.size(), "Size mismatch" );
  Assert( D.size() == gid.size(), "Size mismatch" );

  using tk::operator+=;

  for (std::size_t i=0; i<gid.size(); ++i) {
    m_rhsc[ gid[i] ] += R[i];
    m_difc[ gid[i] ] += D[i];
  }

  if (++m_nrhs == Disc()->NodeCommMap().size()) {
    m_nrhs = 0;
    comrhs_complete();
  }
}

void
DiagCG::solve( tk::Fields& dif )<--- Parameter 'dif' can be declared with const
// *****************************************************************************
//  Solve low and high order diagonal systems
//! \param[in,out] dif Mass diffusion own contribution
// *****************************************************************************
{
  const auto ncomp = m_rhs.nprop();

  auto d = Disc();

  // Combine own and communicated contributions to rhs
  for (const auto& b : m_rhsc) {
    auto lid = tk::cref_find( d->Lid(), b.first );
    for (ncomp_t c=0; c<ncomp; ++c) m_rhs(lid,c) += b.second[c];
  }

  // Combine own and communicated contributions to mass diffusion
  for (const auto& b : m_difc) {
    auto lid = tk::cref_find( d->Lid(), b.first );
    for (ncomp_t c=0; c<ncomp; ++c) dif(lid,c) += b.second[c];
  }

  // Clear receive buffers
  tk::destroy(m_rhsc);
  tk::destroy(m_difc);

  // Set Dirichlet BCs for lhs and both low and high order rhs vectors. Note
  // that the low order rhs (more precisely the mass-diffusion term) is set to
  // zero instead of the solution increment at Dirichlet BCs, because for the
  // low order solution the right hand side is the sum of the high order right
  // hand side and mass diffusion so the low order system is L = R + D, where L
  // is the lumped mass matrix, R is the high order RHS, and D is
  // mass diffusion, and R already will have the Dirichlet BC set.
  for (const auto& [b,bc] : m_bcdir) {
    for (ncomp_t c=0; c<ncomp; ++c) {
      if (bc[c].first) {
        m_lhs( b, c ) = 1.0;
        m_rhs( b, c ) = bc[c].second;
        dif( b, c ) = 0.0;
      }
    }
  }

  // Solve low and high order diagonal systems and update low order solution
  auto dul = (m_rhs + dif) / m_lhs;

  m_ul = m_u + dul;
  m_du = m_rhs / m_lhs;

  const auto& coord = d->Coord();
  // Apply symmetry BCs
  g_cgpde[d->MeshId()].symbc( dul, coord, m_bnorm, m_symbcnodes );
  g_cgpde[d->MeshId()].symbc( m_ul, coord, m_bnorm, m_symbcnodes );
  g_cgpde[d->MeshId()].symbc( m_du, coord, m_bnorm, m_symbcnodes );
  // Apply farfield BCs
  g_cgpde[d->MeshId()].farfieldbc( m_ul, coord, m_bnorm, m_farfieldbcnodes );
  g_cgpde[d->MeshId()].farfieldbc( m_du, coord, m_bnorm, m_farfieldbcnodes );

  // Continue with FCT
  d->FCT()->aec( *d, m_du, m_u, m_bcdir, m_symbcnodemap, m_bnorm );
  d->FCT()->alw( m_u, m_ul, std::move(dul), thisProxy );
}

void
DiagCG::writeFields( CkCallback c ) const
// *****************************************************************************
// Output mesh-based fields to file
//! \param[in] c Function to continue with after the write
// *****************************************************************************
{
  if (g_inputdeck.get< tag::cmd, tag::benchmark >()) {

    c.send();

  } else {

    auto d = Disc();
    const auto& coord = d->Coord();

    // Query fields names requested by user
    auto nodefieldnames = numericFieldNames( tk::Centering::NODE );
    // Collect field output from numerical solution requested by user
    auto nodefields = numericFieldOutput( m_u, tk::Centering::NODE );
    // Collect field output names for analytical solutions
    analyticFieldNames( g_cgpde[d->MeshId()], tk::Centering::NODE,
      nodefieldnames );

    // Collect field output from analytical solutions (if exist)
    auto t = d->T();
    analyticFieldOutput( g_cgpde[d->MeshId()], tk::Centering::NODE, coord[0],
      coord[1], coord[2], t, nodefields );

    // Query and collect block and surface field names from PDEs integrated
    std::vector< std::string > nodesurfnames;
    auto sn = g_cgpde[d->MeshId()].surfNames();
    nodesurfnames.insert( end(nodesurfnames), begin(sn), end(sn) );

    // Collect node field solution
    auto u = m_u;
    std::vector< std::vector< tk::real > > nodesurfs;
    auto so = g_cgpde[d->MeshId()].surfOutput( tk::bfacenodes(m_bface,
      m_triinpoel), u );
    nodesurfs.insert( end(nodesurfs), begin(so), end(so) );

    // Query refinement data
    //auto dtref = g_inputdeck.get< tag::amr, tag::dtref >();

    std::tuple< std::vector< std::string >,
                std::vector< std::vector< tk::real > >,
                std::vector< std::string >,
                std::vector< std::vector< tk::real > > > r;
    /*if (dtref)*/ r = d->Ref()->refinementFields();

    auto& refinement_elemfieldnames = std::get< 0 >( r );
    auto& refinement_elemfields = std::get< 1 >( r );
    auto& refinement_nodefieldnames = std::get< 2 >( r );
    auto& refinement_nodefields = std::get< 3 >( r );

    nodefieldnames.insert( end(nodefieldnames),
      begin(refinement_nodefieldnames), end(refinement_nodefieldnames) );
    nodefields.insert( end(nodefields),
      begin(refinement_nodefields), end(refinement_nodefields) );

    auto elemfieldnames = std::move(refinement_elemfieldnames);
    auto elemfields = std::move(refinement_elemfields);

    // Collect FCT field data (for debugging)
    auto f = d->FCT()->fields();

    const auto& fct_elemfieldnames = std::get< 0 >( f );
    const auto& fct_elemfields = std::get< 1 >( f );
    const auto& fct_nodefieldnames = std::get< 2 >( f );
    const auto& fct_nodefields = std::get< 3 >( f );

    nodefieldnames.insert( end(nodefieldnames),
      begin(fct_nodefieldnames), end(fct_nodefieldnames) );
    nodefields.insert( end(nodefields),
      begin(fct_nodefields), end(fct_nodefields) );

    elemfieldnames.insert( end(elemfieldnames),
      begin(fct_elemfieldnames), end(fct_elemfieldnames) );
    elemfields.insert( end(elemfields),
      begin(fct_elemfields), end(fct_elemfields) );

    Assert( elemfieldnames.size() == elemfields.size(), "Size mismatch" );
    Assert( nodefieldnames.size() == nodefields.size(), "Size mismatch" );

    // Send mesh and fields data (solution dump) for output to file
    d->write( d->Inpoel(), coord, m_bface, tk::remap( m_bnode,d->Lid() ),
              m_triinpoel, elemfieldnames, nodefieldnames, {}, nodesurfnames,
              elemfields, nodefields, {}, nodesurfs, c );

  }
}

void
DiagCG::update( const tk::Fields& a, [[maybe_unused]] tk::Fields&& dul )
// *****************************************************************************
// Prepare for next step
//! \param[in] a Limited antidiffusive element contributions
//! \param[in] dul Low order solution increment
// *****************************************************************************
{
  auto d = Disc();<--- Variable 'd' is assigned a value that is never used.

  // Verify that the change in the solution at those nodes where Dirichlet
  // boundary conditions are set is exactly the amount the BCs prescribe
  Assert( correctBC(a,dul,m_bcdir), "Dirichlet boundary condition incorrect" );

  // Apply limited antidiffusive element contributions to low order solution
  auto un = m_u;
  if (g_inputdeck.get< tag::discr, tag::fct >())
    m_u = m_ul + a;
  else
    m_u = m_u + m_du;

  // Compute diagnostics, e.g., residuals
  auto diag_computed = m_diag.compute( *d, m_u, un, m_bnorm,
                                        m_symbcnodes, m_farfieldbcnodes );
  // Increase number of iterations and physical time
  d->next();
  // Continue to mesh refinement (if configured)
  if (!diag_computed) refine( std::vector< tk::real >( m_u.nprop(), 0.0 ) );
}

void
DiagCG::refine( [[maybe_unused]] const std::vector< tk::real >& l2res )
// *****************************************************************************
// Optionally refine/derefine mesh
//! \param[in] l2res L2-norms of the residual for each scalar component
//!   computed across the whole problem
// *****************************************************************************
{
  auto d = Disc();

  auto dtref = g_inputdeck.get< tag::amr, tag::dtref >();
  auto dtfreq = g_inputdeck.get< tag::amr, tag::dtfreq >();

  // if t>0 refinement enabled and we hit the dtref frequency
  if (dtref && !(d->It() % dtfreq)) {   // h-refine

    // Activate SDAG waits for re-computing the left-hand side
    thisProxy[ thisIndex ].wait4lhs();

    d->startvol();
    d->Ref()->dtref( m_bface, m_bnode, m_triinpoel );
    d->refined() = 1;

  } else {      // do not h-refine

    d->refined() = 0;
    lhs_complete();
    resized();

  }
}

void
DiagCG::resizePostAMR(
  const std::vector< std::size_t >& /*ginpoel*/,
  const tk::UnsMesh::Chunk& chunk,
  const tk::UnsMesh::Coords& coord,
  const std::unordered_map< std::size_t, tk::UnsMesh::Edge >& addedNodes,
  const std::unordered_map< std::size_t, std::size_t >& /*addedTets*/,
  const std::set< std::size_t >& removedNodes,
  const std::unordered_map< std::size_t, std::size_t >& amrNodeMap,
  const tk::NodeCommMap& nodeCommMap,
  const std::map< int, std::vector< std::size_t > >& /*bface*/,
  const std::map< int, std::vector< std::size_t > >& bnode,
  const std::vector< std::size_t >& /*triinpoel*/,
  const std::unordered_map< std::size_t, std::set< std::size_t > >& elemblockid )
// *****************************************************************************
//  Receive new mesh from Refiner
//! \param[in] ginpoel Mesh connectivity with global node ids
//! \param[in] chunk New mesh chunk (connectivity and global<->local id maps)
//! \param[in] coord New mesh node coordinates
//! \param[in] addedNodes Newly added mesh nodes and their parents (local ids)
//! \param[in] addedTets Newly added mesh cells and their parents (local ids)
//! \param[in] removedNodes Newly removed mesh node local ids
//! \param[in] amrNodeMap Node id map after amr (local ids)
//! \param[in] nodeCommMap New node communication map
//! \param[in] bnode Boundary-node lists mapped to side set ids
//! \param[in] elemblockid Local tet ids associated with mesh block ids
// *****************************************************************************
{
  auto d = Disc();

  // Set flag that indicates that we are during time stepping
  m_initial = 0;

  // Zero field output iteration count between two mesh refinement steps
  d->Itf() = 0;

  // Increase number of iterations with mesh refinement
  ++d->Itr();

  // Resize mesh data structures
  d->resizePostAMR( chunk, coord, amrNodeMap, nodeCommMap, removedNodes,
    elemblockid );

  Assert(coord[0].size() == m_u.nunk()-removedNodes.size()+addedNodes.size(),
    "Incorrect vector length post-AMR: expected length after resizing = " +
    std::to_string(coord[0].size()) + ", actual unknown vector length = " +
    std::to_string(m_u.nunk()-removedNodes.size()+addedNodes.size()));

  // Remove newly removed nodes from solution vectors
  m_u.rm(removedNodes);
  m_ul.rm(removedNodes);
  m_du.rm(removedNodes);
  m_lhs.rm(removedNodes);
  m_rhs.rm(removedNodes);

  // Resize auxiliary solution vectors
  auto nelem = d->Inpoel().size()/4;
  auto npoin = coord[0].size();
  auto nprop = m_u.nprop();
  m_u.resize( npoin );
  m_ul.resize( npoin );
  m_du.resize( npoin );
  m_ue.resize( nelem );
  m_lhs.resize( npoin );
  m_rhs.resize( npoin );

  // Update solution on new mesh
  for (const auto& n : addedNodes)
    for (std::size_t c=0; c<nprop; ++c) {
      Assert(n.first < m_u.nunk(), "Added node index out of bounds post-AMR");
      Assert(n.second[0] < m_u.nunk() && n.second[1] < m_u.nunk(),
        "Indices of parent-edge nodes out of bounds post-AMR");
      m_u(n.first,c) = (m_u(n.second[0],c) + m_u(n.second[1],c))/2.0;
    }

  // Update physical-boundary node lists
  m_bnode = bnode;

  // Resize FCT data structures
  d->FCT()->resize( npoin, nodeCommMap, d->Bid(), d->Lid(), d->Inpoel() );

  auto meshid = d->MeshId();
  contribute( sizeof(std::size_t), &meshid, CkReduction::nop,
              CkCallback(CkReductionTarget(Transporter,resized), d->Tr()) );
}

void
DiagCG::resized()
// *****************************************************************************
// Resizing data sutrctures after mesh refinement has been completed
// *****************************************************************************
{
  resize_complete();
}

void
DiagCG::out()
// *****************************************************************************
// Output mesh field data
// *****************************************************************************
{
  auto d = Disc();

  // Output time history
  if (d->histiter() or d->histtime() or d->histrange()) {
    std::vector< std::vector< tk::real > > hist;
    auto h = g_cgpde[d->MeshId()].histOutput( d->Hist(), d->Inpoel(), m_u );
    hist.insert( end(hist), begin(h), end(h) );
    d->history( std::move(hist) );
  }

  // Output field data
  if (d->fielditer() or d->fieldtime() or d->fieldrange() or d->finished())
    writeFields( CkCallback(CkIndex_DiagCG::step(), thisProxy[thisIndex]) );
  else
    step();
}

void
DiagCG::evalLB( int nrestart )
// *****************************************************************************
// Evaluate whether to do load balancing
//! \param[in] nrestart Number of times restarted
// *****************************************************************************
{
  auto d = Disc();

  // Detect if just returned from a checkpoint and if so, zero timers
  d->restarted( nrestart );

  const auto lbfreq = g_inputdeck.get< tag::cmd, tag::lbfreq >();
  const auto nonblocking = g_inputdeck.get< tag::cmd, tag::nonblocking >();

  // Load balancing if user frequency is reached or after the second time-step
  if ( (d->It()) % lbfreq == 0 || d->It() == 2 ) {

    AtSync();
    if (nonblocking) next();

  } else {

    next();

  }
}

void
DiagCG::evalRestart()
// *****************************************************************************
// Evaluate whether to save checkpoint/restart
// *****************************************************************************
{
  auto d = Disc();

  const auto rsfreq = g_inputdeck.get< tag::cmd, tag::rsfreq >();
  const auto benchmark = g_inputdeck.get< tag::cmd, tag::benchmark >();

  if (not benchmark and not (d->It() % rsfreq)) {

    std::vector< std::size_t > meshdata{ /* finished = */ 0, d->MeshId() };
    contribute( meshdata, CkReduction::nop,
      CkCallback(CkReductionTarget(Transporter,checkpoint), d->Tr()) );

  } else {

    evalLB( /* nrestart = */ -1 );

  }
}

void
DiagCG::step()
// *****************************************************************************
// Evaluate whether to continue with next time step
// *****************************************************************************
{
  auto d = Disc();

  // Output one-liner status report to screen
  d->status();

  const auto term = g_inputdeck.get< tag::discr, tag::term >();
  const auto nstep = g_inputdeck.get< tag::discr, tag::nstep >();
  const auto eps = std::numeric_limits< tk::real >::epsilon();

  // If neither max iterations nor max time reached, continue, otherwise finish
  if (std::fabs(d->T()-term) > eps && d->It() < nstep) {

    evalRestart();

  } else {

    auto meshid = d->MeshId();
    d->contribute( sizeof(std::size_t), &meshid, CkReduction::nop,
                   CkCallback(CkReductionTarget(Transporter,finish), d->Tr()) );

  }
}

#include "NoWarning/diagcg.def.h"