File: sourcestats.c

package info (click to toggle)
chrony 1.21z-5%2Betch1
  • links: PTS
  • area: main
  • in suites: etch
  • size: 2,080 kB
  • ctags: 2,385
  • sloc: ansic: 14,901; yacc: 858; sh: 588; perl: 426; makefile: 148
file content (931 lines) | stat: -rw-r--r-- 28,972 bytes parent folder | download | duplicates (3)
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
/*
  $Header: /cvs/src/chrony/sourcestats.c,v 1.40 2003/09/22 21:22:30 richard Exp $

  =======================================================================

  chronyd/chronyc - Programs for keeping computer clocks accurate.

 **********************************************************************
 * Copyright (C) Richard P. Curnow  1997-2003
 * 
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of version 2 of the GNU General Public License as
 * published by the Free Software Foundation.
 * 
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 * 
 * You should have received a copy of the GNU General Public License along
 * with this program; if not, write to the Free Software Foundation, Inc.,
 * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA
 * 
 **********************************************************************

  =======================================================================

  This file contains the routines that do the statistical
  analysis on the samples obtained from the sources,
  to determined frequencies and error bounds. */

#include "sysincl.h"

#include "sourcestats.h"
#include "memory.h"
#include "regress.h"
#include "util.h"
#include "conf.h"
#include "logging.h"
#include "local.h"
#include "mkdirpp.h"

/* ================================================== */
/* Define the maxumum number of samples that we want
   to store per source */
#define MAX_SAMPLES 64

/* This is the assumed worst case bound on an unknown frequency,
   2000ppm, which would be pretty bad */
#define WORST_CASE_FREQ_BOUND (2000.0/1.0e6)

/* Day number of 1 Jan 1970 */
#define MJD_1970 40587

/* ================================================== */
/* File to which statistics are logged, NULL if none */
static FILE *logfile = NULL;
static char *logfilename = NULL;
static unsigned long logwrites = 0;

#define STATISTICS_LOG "statistics.log"

/* ================================================== */
/* This data structure is used to hold the history of data from the
   source */

struct SST_Stats_Record {

  /* Reference ID of source, used for logging to statistics log */
  unsigned long refid;

  /* Number of samples currently stored.  sample[n_samples-1] is the
     newest.  The samples are expected to be sorted in order, but that
     probably doesn't matter.  */
  int n_samples;

  /* The index in the registers of the best individual sample that we
     are holding, in terms of the minimum root distance at the present
     time */
  int best_single_sample;

  /* This is the estimated offset (+ve => local fast) at a particular time */
  double estimated_offset;
  double estimated_offset_sd;
  struct timeval offset_time;

  /* Number of runs of the same sign amongst the residuals */
  int nruns;

  /* This value contains the estimated frequency.  This is the number
     of seconds that the local clock gains relative to the reference
     source per unit local time.  (Positive => local clock fast,
     negative => local clock slow) */
  double estimated_frequency;

  /* This is the assumed worst case bounds on the estimated frequency.
     We assume that the true frequency lies within +/- half this much
     about estimated_frequency */
  double skew;

  /* This is the direction the skew went in at the last sample */
  SST_Skew_Direction skew_dirn;

  /* This is the estimated residual variance of the data points */
  double variance;

  /* This array contains the sample epochs, in terms of the local
     clock. */
  struct timeval sample_times[MAX_SAMPLES];

  /* This is an array of offsets, in seconds, corresponding to the
     sample times.  In this module, we use the convention that
     positive means the local clock is FAST of the source and negative
     means it is SLOW.  This is contrary to the convention in the NTP
     stuff; that part of the code is written to correspond with
     RFC1305 conventions. */
  double offsets[MAX_SAMPLES];

  /* This is an array of the offsets as originally measured.  Local
     clock fast of real time is indicated by positive values.  This
     array is not slewed to adjust the readings when we apply
     adjustments to the local clock, as is done for the array
     'offset'. */
  double orig_offsets[MAX_SAMPLES];

  /* This is an array of peer delays, in seconds, being the roundtrip
     measurement delay to the peer */
  double peer_delays[MAX_SAMPLES];

  /* This is an array of peer dispersions, being the skew and local
     precision dispersion terms from sampling the peer */
  double peer_dispersions[MAX_SAMPLES];

  /* This array contains the root delays of each sample, in seconds */
  double root_delays[MAX_SAMPLES];

  /* This array contains the root dispersions of each sample at the
     time of the measurements */
  double root_dispersions[MAX_SAMPLES];

  /* This array contains the weights to be used in the regression
     analysis for each of the samples. */
  double weights[MAX_SAMPLES];

  /* This array contains the strata that were associated with the sources
     at the times the samples were generated */
  int strata[MAX_SAMPLES];

};

/* ================================================== */

void
SST_Initialise(void)
{
  char *direc;

  if (CNF_GetLogStatistics()) {
    direc = CNF_GetLogDir();
    if (!mkdir_and_parents(direc)) {
      LOG(LOGS_ERR, LOGF_SourceStats, "Could not create directory %s", direc);
      logfile = NULL;
    } else {
      logfilename = MallocArray(char, 2 + strlen(direc) + strlen(STATISTICS_LOG));
      strcpy(logfilename, direc);
      strcat(logfilename, "/");
      strcat(logfilename, STATISTICS_LOG);
      logfile = fopen(logfilename, "a");
      if (!logfile) {
        LOG(LOGS_WARN, LOGF_SourceStats, "Couldn't open logfile %s for update", logfilename);
      }
    }
  }
}

/* ================================================== */

void
SST_Finalise(void)
{
  if (logfile) {
    fclose(logfile);
  }
}

/* ================================================== */
/* This function creates a new instance of the statistics handler */

SST_Stats
SST_CreateInstance(unsigned long refid)
{
  SST_Stats inst;
  inst = MallocNew(struct SST_Stats_Record);
  inst->refid = refid;
  inst->n_samples = 0;
  inst->estimated_frequency = 0;
  inst->skew = 2000.0e-6;
  inst->skew_dirn = SST_Skew_Nochange;
  inst->estimated_offset = 0.0;
  inst->estimated_offset_sd = 86400.0; /* Assume it's at least within a day! */
  inst->variance = 16.0;
  inst->nruns = 0;
  return inst;
}

/* ================================================== */
/* This function deletes an instance of the statistics handler. */

void
SST_DeleteInstance(SST_Stats inst)
{
  Free(inst);
  return;
}

/* ================================================== */

static void
move_stats_entry(SST_Stats inst, int src, int dest)
{
  inst->sample_times[dest] = inst->sample_times[src];
  inst->offsets[dest] = inst->offsets[src];
  inst->orig_offsets[dest] = inst->orig_offsets[src];
  inst->peer_delays[dest] = inst->peer_delays[src];
  inst->peer_dispersions[dest] = inst->peer_dispersions[src];
  inst->root_delays[dest] = inst->root_delays[src];
  inst->root_dispersions[dest] = inst->root_dispersions[src];
  inst->weights[dest] = inst->weights[src];
  inst->strata[dest] = inst->strata[src];
}

/* ================================================== */
/* This function is called to prune the register down when it is full.
   For now, just discard the oldest sample.  */

static void
prune_register(SST_Stats inst, int new_oldest, int *bad_points)
{
  int i, j;

  if (!(new_oldest < inst->n_samples)) {
    CROAK("new_oldest should be < n_samples");
  }
   
  for (i=0, j=new_oldest; j<inst->n_samples; j++) {
    if (!bad_points || !bad_points[j]) {
      if (j != i) {
        move_stats_entry(inst, j, i);
      }
      i++;
    }
  }
  inst->n_samples = i;

}

/* ================================================== */

void
SST_AccumulateSample(SST_Stats inst, struct timeval *sample_time,
                     double offset,
                     double peer_delay, double peer_dispersion,
                     double root_delay, double root_dispersion,
                     int stratum)
{
  int n;
#if 0
  double root_distance;
#endif

  if (inst->n_samples == MAX_SAMPLES) {
    prune_register(inst, 1, NULL);
  }

  n = inst->n_samples;

  inst->sample_times[n] = *sample_time;
  inst->offsets[n] = offset;
  inst->orig_offsets[n] = offset;
  inst->peer_delays[n] = peer_delay;
  inst->peer_dispersions[n] = peer_dispersion;
  inst->root_delays[n] = root_delay;
  inst->root_dispersions[n] = root_dispersion;

#if 0
  /* The weight is worked out when we run the regression algorithm */
  root_distance = root_dispersion + 0.5 * fabs(root_delay);
  
  /* For now, this is the formula for the weight functions */
  inst->weights[n] = root_distance * root_distance;
#endif

  inst->strata[n] = stratum;
 
  ++inst->n_samples;

}

/* ================================================== */
/* This function is used by both the regression routines to find the
   time interval between each historical sample and the most recent
   one */

static void
convert_to_intervals(SST_Stats inst, double *times_back)
{
  struct timeval *newest_tv;
  int i;

  newest_tv = &(inst->sample_times[inst->n_samples - 1]);
  for (i=0; i<inst->n_samples; i++) {
    /* The entries in times_back[] should end up negative */
    UTI_DiffTimevalsToDouble(&(times_back[i]), &(inst->sample_times[i]), newest_tv);
  }
}

/* ================================================== */

static void
find_best_sample_index(SST_Stats inst, double *times_back)
{
  /* With the value of skew that has been computed, see which of the
     samples offers the tightest bound on root distance */

  double root_distance, best_root_distance;
  double elapsed;
  int i, n, best_index;

  n = inst->n_samples - 1;
  best_root_distance = inst->root_dispersions[n] + 0.5 * fabs(inst->root_delays[n]);
  best_index = n;
#if 0
  LOG(LOGS_INFO, LOGF_SourceStats, "n=%d brd=%f", n, best_root_distance);
#endif
  for (i=0; i<n; i++) {
    elapsed = -times_back[i];
#if 0
    LOG(LOGS_INFO, LOGF_SourceStats, "n=%d i=%d latest=[%s] doing=[%s] elapsed=%f", n, i, 
        UTI_TimevalToString(&(inst->sample_times[n])),
        UTI_TimevalToString(&(inst->sample_times[i])),
        elapsed);
#endif

    /* Because the loop does not consider the most recent sample, this assertion must hold */
    if (elapsed <= 0.0) {
      LOG(LOGS_ERR, LOGF_SourceStats, "Elapsed<0! n=%d i=%d latest=[%s] doing=[%s] elapsed=%f",
          n, i, 
          UTI_TimevalToString(&(inst->sample_times[n])),
          UTI_TimevalToString(&(inst->sample_times[i])),
          elapsed);

      elapsed = fabs(elapsed);
    }

    root_distance = inst->root_dispersions[i] + elapsed * inst->skew + 0.5 * fabs(inst->root_delays[i]);
    if (root_distance < best_root_distance) {
      best_root_distance = root_distance;
      best_index = i;
    }
  }
  
  inst->best_single_sample = best_index;

#if 0
  LOG(LOGS_INFO, LOGF_SourceStats, "n=%d best_index=%d", n, best_index);
#endif

  return;
}

/* ================================================== */

/* This defines the assumed ratio between the standard deviation of
   the samples and the peer distance as measured from the round trip
   time.  E.g. a value of 4 means that we think the standard deviation
   is a quarter of the peer distance */

#define SD_TO_DIST_RATIO 8.0

/* ================================================== */
/* This function runs the linear regression operation on the data.  It
   finds the set of most recent samples that give the tightest
   confidence interval for the frequency, and truncates the register
   down to that number of samples */

void
SST_DoNewRegression(SST_Stats inst)
{
  double times_back[MAX_SAMPLES];
  double peer_distances[MAX_SAMPLES];

  int bad_points[MAX_SAMPLES];
  int degrees_of_freedom;
  int best_start;
  double est_intercept, est_slope, est_var, est_intercept_sd, est_slope_sd;
  int i, nruns;
  double min_distance;
  double sd_weight;
  double old_skew, old_freq, stress;

  int regression_ok;

  convert_to_intervals(inst, times_back);

  if (inst->n_samples > 0) {
    for (i=0; i<inst->n_samples; i++) {
      peer_distances[i] = 0.5 * fabs(inst->peer_delays[i]) + inst->peer_dispersions[i];
    }
  
    min_distance = peer_distances[0];
    for (i=1; i<inst->n_samples; i++) {
      if (peer_distances[i] < min_distance) {
        min_distance = peer_distances[i];
      }
    }

    /* And now, work out the weight vector */

    for (i=0; i<inst->n_samples; i++) {
      sd_weight = 1.0 + SD_TO_DIST_RATIO * (peer_distances[i] - min_distance) / min_distance;
      inst->weights[i] = sd_weight * sd_weight;
    } 
  }         

  regression_ok = RGR_FindBestRegression(times_back, inst->offsets, inst->weights,
                                         inst->n_samples,
                                         &est_intercept, &est_slope, &est_var,
                                         &est_intercept_sd, &est_slope_sd,
                                         &best_start, &nruns, &degrees_of_freedom);

  /* This is a legacy of when the regression routine found outliers
     for us.  We don't use it anymore. */
  memset((void *) bad_points, 0, MAX_SAMPLES * sizeof(int));

  if (regression_ok) {

    old_skew = inst->skew;
    old_freq = inst->estimated_frequency;
  
    inst->estimated_frequency = est_slope;
    inst->skew = est_slope_sd * RGR_GetTCoef(degrees_of_freedom);
    inst->estimated_offset = est_intercept;
    inst->offset_time = inst->sample_times[inst->n_samples - 1];
    inst->estimated_offset_sd = est_intercept_sd;
    inst->variance = est_var;
    inst->nruns = nruns;

    stress = fabs(old_freq - inst->estimated_frequency) / old_skew;

    if (best_start > 0) {
      /* If we are throwing old data away, retain the current
         assumptions about the skew */
      inst->skew_dirn = SST_Skew_Nochange;
    } else {
      if (inst->skew < old_skew) {
        inst->skew_dirn = SST_Skew_Decrease;
      } else {
        inst->skew_dirn = SST_Skew_Increase;
      }
    }

    if (logfile) {

      if (((logwrites++) % 32) == 0) {
        fprintf(logfile,
                "==============================================================================================================\n"
                "   Date (UTC) Time     IP Address    Std dev'n Est offset  Offset sd  Diff freq   Est skew  Stress  Ns  Bs  Nr\n"
                "==============================================================================================================\n");
      }
      
            
      fprintf(logfile, "%s %-15s %10.3e %10.3e %10.3e %10.3e %10.3e %7.1e %3d %3d %3d\n",
              UTI_TimeToLogForm(inst->offset_time.tv_sec),
              UTI_IPToDottedQuad(inst->refid),
              sqrt(inst->variance),
              inst->estimated_offset,
              inst->estimated_offset_sd,
              inst->estimated_frequency,
              inst->skew,
              stress,
              inst->n_samples,
              best_start, nruns);

      fflush(logfile);
    }

    prune_register(inst, best_start, bad_points);

  } else {
#if 0
    LOG(LOGS_INFO, LOGF_SourceStats, "too few points (%d) for regression", inst->n_samples);
#endif
    inst->estimated_frequency = 0.0;
    inst->skew = WORST_CASE_FREQ_BOUND;
  }

  find_best_sample_index(inst, times_back);

}

/* ================================================== */
/* This function does a simple regression on what is in the register,
   without trying to optimise the error bounds on the frequency by
   deleting old samples */

void
SST_DoUpdateRegression(SST_Stats inst)
{
  double times_back[MAX_SAMPLES];
  double freq_error_bound;
  double est_intercept, est_slope, est_var_base, est_intercept_sd, est_slope_sd;

  convert_to_intervals(inst, times_back);

  if (inst->n_samples >= 3) { /* Otherwise, we're wasting our time - we
                                 can't do a useful linear regression
                                 with less than 3 points */
    
    RGR_WeightedRegression(times_back, inst->offsets, inst->weights,
                           inst->n_samples,
                           &est_intercept, &est_slope, &est_var_base,
                           &est_intercept_sd, &est_slope_sd);

    freq_error_bound = est_slope_sd * RGR_GetTCoef(inst->n_samples - 2);

    inst->estimated_frequency = est_slope;
    inst->skew = freq_error_bound;

  } else {
    inst->estimated_frequency = 0.0;
    inst->skew = WORST_CASE_FREQ_BOUND;
  }

  find_best_sample_index(inst, times_back);

}

/* ================================================== */

void
SST_GetReferenceData(SST_Stats inst, struct timeval *now, 
                     int *stratum, double *offset,
                     double *root_delay, double *root_dispersion,
                     double *frequency, double *skew)
{

  double elapsed;
  int n;

  *frequency = inst->estimated_frequency;
  *skew = inst->skew;

  n = inst->best_single_sample;

  UTI_DiffTimevalsToDouble(&elapsed, now, &(inst->sample_times[n]));
  *root_delay = inst->root_delays[n];
  *root_dispersion = inst->root_dispersions[n] + elapsed * inst->skew;
  *offset = inst->offsets[n] + elapsed * inst->estimated_frequency;
  *stratum = inst->strata[n];

#ifdef TRACEON
  LOG(LOGS_INFO, LOGF_SourceStats, "n=%d freq=%f skew=%f del=%f disp=%f ofs=%f str=%d",
      n, *frequency, *skew, *root_delay, *root_dispersion, *offset, *stratum);
#endif

  return;
}

/* ================================================== */
/* Return the assumed worst case range of values that this source's
   frequency lies within.  Frequency is defined as the amount of time
   the local clock gains relative to the source per unit local clock
   time. */
void
SST_GetFrequencyRange(SST_Stats inst,
                      double *lo, double *hi)
{
  double freq, skew;
  freq = inst->estimated_frequency;
  skew = inst->skew;
  *lo = freq - skew;
  *hi = freq + skew;
}

/* ================================================== */

/* ================================================== */

void
SST_GetSelectionData(SST_Stats inst, struct timeval *now,
                     int *stratum,
                     double *best_offset, double *best_root_delay,
                     double *best_root_dispersion,
                     double *variance, int *average_ok)
{
  double average_offset;
  double sample_elapsed;
  double elapsed;
  int n;
  double peer_distance;
  
  n = inst->best_single_sample;
  *stratum = inst->strata[n];
  *variance = inst->variance;

  peer_distance = inst->peer_dispersions[n] + 0.5 * fabs(inst->peer_delays[n]);
  UTI_DiffTimevalsToDouble(&elapsed, now, &(inst->offset_time));

  UTI_DiffTimevalsToDouble(&sample_elapsed, now, &(inst->sample_times[n]));
  *best_offset = inst->offsets[n] + sample_elapsed * inst->estimated_frequency;
  *best_root_delay = inst->root_delays[n];
  *best_root_dispersion = inst->root_dispersions[n] + sample_elapsed * inst->skew;

  average_offset = inst->estimated_offset + inst->estimated_frequency * elapsed;
  if (fabs(average_offset - *best_offset) <= peer_distance) {
    *average_ok = 1;
  } else {
    *average_ok = 0;
  }

#ifdef TRACEON
  LOG(LOGS_INFO, LOGF_SourceStats, "n=%d off=%f del=%f dis=%f var=%f pdist=%f avoff=%f avok=%d",
      n, *best_offset, *best_root_delay, *best_root_dispersion, *variance,
      peer_distance, average_offset, *average_ok);
#endif

  return;
}

/* ================================================== */

void
SST_GetTrackingData(SST_Stats inst, struct timeval *now,
                    double *average_offset, double *offset_sd,
                    double *accrued_dispersion,
                    double *frequency, double *skew)
{
  int n;
  double peer_distance;
  double elapsed_offset, elapsed_sample;

  n = inst->best_single_sample;

  *frequency = inst->estimated_frequency;
  *skew = inst->skew;

  peer_distance = inst->peer_dispersions[n] + 0.5 * fabs(inst->peer_delays[n]);
  UTI_DiffTimevalsToDouble(&elapsed_offset, now, &(inst->offset_time));
  *average_offset = inst->estimated_offset + inst->estimated_frequency * elapsed_offset;
  *offset_sd = inst->estimated_offset_sd + elapsed_offset * inst->skew;

  UTI_DiffTimevalsToDouble(&elapsed_sample, now, &(inst->sample_times[n]));
  *accrued_dispersion = inst->skew * elapsed_sample;

#ifdef TRACEON
  LOG(LOGS_INFO, LOGF_SourceStats, "n=%d freq=%f (%.3fppm) skew=%f (%.3fppm) pdist=%f avoff=%f offsd=%f accrdis=%f",
      n, *frequency, 1.0e6* *frequency, *skew, 1.0e6* *skew, peer_distance, *average_offset, *offset_sd, *accrued_dispersion);
#endif

}

/* ================================================== */

void
SST_SlewSamples(SST_Stats inst, struct timeval *when, double dfreq, double doffset)
{
  int n, i;
  double elapsed;
  double delta_time;
  struct timeval *sample, prev;
  double prev_offset, prev_freq;

  n = inst->n_samples;

  for (i=0; i<n; i++) {
    sample = &(inst->sample_times[i]);
    prev = *sample;
#if 0
    UTI_AdjustTimeval(sample, when, sample, dfreq, doffset);
    /* Can't easily use this because we need to slew offset */
#endif
    UTI_DiffTimevalsToDouble(&elapsed, when, sample);
    delta_time = elapsed * dfreq - doffset;
    UTI_AddDoubleToTimeval(sample, delta_time, sample);
    prev_offset = inst->offsets[i];
    inst->offsets[i] += delta_time;
#ifdef TRACEON
    LOG(LOGS_INFO, LOGF_SourceStats, "i=%d old_st=[%s] new_st=[%s] old_off=%f new_off=%f",
        i, UTI_TimevalToString(&prev), UTI_TimevalToString(sample),
        prev_offset, inst->offsets[i]);
#endif
  }

  /* Do a half-baked update to the regression estimates */
  UTI_DiffTimevalsToDouble(&elapsed, when, &(inst->offset_time));
  prev = inst->offset_time;
  delta_time = elapsed * dfreq - doffset;
  UTI_AddDoubleToTimeval(&(inst->offset_time), delta_time, &(inst->offset_time));
  prev_offset = inst->estimated_offset;
  prev_freq = inst->estimated_frequency;
  inst->estimated_offset += delta_time;
  inst->estimated_frequency -= dfreq;

#ifdef TRACEON
  LOG(LOGS_INFO, LOGF_SourceStats, "old_off_time=[%s] new=[%s] old_off=%f new_off=%f old_freq=%.3fppm new_freq=%.3fppm",
      UTI_TimevalToString(&prev), UTI_TimevalToString(&(inst->offset_time)),
      prev_offset, inst->estimated_offset,
      1.0e6*prev_freq, 1.0e6*inst->estimated_frequency);
#endif

  return;
}

/* ================================================== */

double
SST_PredictOffset(SST_Stats inst, struct timeval *when)
{
  double elapsed;
  
  if (inst->n_samples < 3) {
    /* We don't have any useful statistics, and presumably the poll
       interval is minimal.  We can't do any useful prediction other
       than use the latest sample */
    return inst->offsets[inst->n_samples - 1];
  } else {
    UTI_DiffTimevalsToDouble(&elapsed, when, &inst->offset_time);
    return inst->estimated_offset + elapsed * inst->estimated_frequency;
  }

}

/* ================================================== */

double
SST_MinRoundTripDelay(SST_Stats inst)
{
  double min_delay, delay;
  int i;

  if (inst->n_samples == 0) {
    return DBL_MAX;
  } else {
    min_delay = fabs(inst->peer_delays[0]);
    for (i=1; i<inst->n_samples; i++) {
      delay = fabs(inst->peer_delays[i]);
      if (delay < min_delay) {
        min_delay = delay;
      }
    }
    return min_delay;
  }
}

/* ================================================== */
/* This is used to save the register to a file, so that we can reload
   it after restarting the daemon */

void
SST_SaveToFile(SST_Stats inst, FILE *out)
{
  int i;

  fprintf(out, "%d\n", inst->n_samples);

  for(i=0; i<inst->n_samples; i++) {

    fprintf(out, "%08lx %08lx %.6e %.6e %.6e %.6e %.6e %.6e %.6e %d\n",
            (unsigned long) inst->sample_times[i].tv_sec,
            (unsigned long) inst->sample_times[i].tv_usec,
            inst->offsets[i],
            inst->orig_offsets[i],
            inst->peer_delays[i],
            inst->peer_dispersions[i],
            inst->root_delays[i],
            inst->root_dispersions[i],
            inst->weights[i],
            inst->strata[i]);

  }
}

/* ================================================== */
/* This is used to reload samples from a file */

int
SST_LoadFromFile(SST_Stats inst, FILE *in)
{
  int i, line_number;
  char line[1024];
  unsigned long sec, usec;

  if (fgets(line, sizeof(line), in) &&
      (sscanf(line, "%d", &inst->n_samples) == 1)) {

    line_number = 2;

    for (i=0; i<inst->n_samples; i++) {
      if (!fgets(line, sizeof(line), in) ||
          (sscanf(line, "%lx%lx%lf%lf%lf%lf%lf%lf%lf%d\n",
                  &(sec), &(usec),
                  &(inst->offsets[i]),
                  &(inst->orig_offsets[i]),
                  &(inst->peer_delays[i]),
                  &(inst->peer_dispersions[i]),
                  &(inst->root_delays[i]),
                  &(inst->root_dispersions[i]),
                  &(inst->weights[i]),
                  &(inst->strata[i])) != 10)) {

        /* This is the branch taken if the read FAILED */

        LOG(LOGS_WARN, LOGF_SourceStats, "Failed to read data from line %d of dump file", line_number);
        inst->n_samples = 0; /* Load abandoned if any sign of corruption */
        return 0;
      } else {

        /* This is the branch taken if the read is SUCCESSFUL */
        inst->sample_times[i].tv_sec = sec;
        inst->sample_times[i].tv_usec = usec;

        line_number++;
      }
    }
  } else {
    LOG(LOGS_WARN, LOGF_SourceStats, "Could not read number of samples from dump file");
    inst->n_samples = 0; /* Load abandoned if any sign of corruption */
    return 0;
  }

  return 1;

}

/* ================================================== */

void
SST_DoSourceReport(SST_Stats inst, RPT_SourceReport *report, struct timeval *now)
{
  int n, nb;
  double est_offset, est_err, elapsed, sample_elapsed;
  struct timeval ago;

  if (inst->n_samples > 0) {
    n = inst->n_samples - 1;
    report->orig_latest_meas = (long)(0.5 + 1.0e6 * inst->orig_offsets[n]);
    report->latest_meas = (long)(0.5 + 1.0e6 * inst->offsets[n]);
    report->latest_meas_err = (unsigned long)(0.5 + 1.0e6 * (0.5*inst->root_delays[n] + inst->root_dispersions[n]));
    report->stratum = inst->strata[n];

    UTI_DiffTimevals(&ago, now, &inst->sample_times[n]);
    report->latest_meas_ago = ago.tv_sec;

    if (inst->n_samples > 3) {
      UTI_DiffTimevalsToDouble(&elapsed, now, &inst->offset_time);
      nb = inst->best_single_sample;
      UTI_DiffTimevalsToDouble(&sample_elapsed, now, &(inst->sample_times[nb]));
      est_offset = inst->estimated_offset + elapsed * inst->estimated_frequency;
      est_err = (inst->estimated_offset_sd +
                 sample_elapsed * inst->skew +
                 (0.5*inst->root_delays[nb] + inst->root_dispersions[nb]));
      report->est_offset = (long)(0.5 + 1.0e6 * est_offset);
      report->est_offset_err = (unsigned long) (0.5 + 1.0e6 * est_err);
      report->resid_freq = (long) (0.5 * 1.0e9 * inst->estimated_frequency);
      report->resid_skew = (unsigned long) (0.5 + 1.0e9 * inst->skew);
    } else {
      report->est_offset = report->latest_meas;
      report->est_offset_err = report->latest_meas_err;
      report->resid_freq = 0;
      report->resid_skew = 0;
    }
  } else {
    report->orig_latest_meas = 0;
    report->latest_meas = 0;
    report->latest_meas_err = 0;
    report->stratum = 0;
    report->est_offset = 0;
    report->est_offset_err = 0;
    report->resid_freq = 0;
    report->resid_skew = 0;
  }
}


/* ================================================== */

SST_Skew_Direction SST_LastSkewChange(SST_Stats inst)
{
  return inst->skew_dirn;
}

/* ================================================== */

void
SST_DoSourcestatsReport(SST_Stats inst, RPT_SourcestatsReport *report)
{
  double dspan;
  int n;

  report->n_samples = inst->n_samples;
  report->n_runs = inst->nruns;

  if (inst->n_samples > 1) {
    n = inst->n_samples - 1;
    UTI_DiffTimevalsToDouble(&dspan, &inst->sample_times[n], &inst->sample_times[0]);
    report->span_seconds = (unsigned long) (dspan + 0.5);
  } else {
    report->span_seconds = 0;
  }

  report->resid_freq_ppm = 1.0e6 * inst->estimated_frequency;
  report->skew_ppm = 1.0e6 * inst->skew;
  report->sd_us = 1.0e6 * sqrt(inst->variance);
}

/* ================================================== */

void
SST_CycleLogFile(void)
{
  if (logfile && logfilename) {
    fclose(logfile);
    logfile = fopen(logfilename, "a");
    if (!logfile) {
      LOG(LOGS_WARN, LOGF_SourceStats, "Could not reopen logfile %s", logfilename);
    }
    logwrites = 0;
  }
}

/* ================================================== */