summaryrefslogtreecommitdiffstats
path: root/sys/vm/vm_phys.c
blob: d26b8b58095297ce4b23f0e713dac992f247637d (plain)
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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
/*-
 * Copyright (c) 2002-2006 Rice University
 * Copyright (c) 2007 Alan L. Cox <alc@cs.rice.edu>
 * All rights reserved.
 *
 * This software was developed for the FreeBSD Project by Alan L. Cox,
 * Olivier Crameri, Peter Druschel, Sitaram Iyer, and Juan Navarro.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
 * A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT
 * HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
 * OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
 * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY
 * WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

/*
 *	Physical memory system implementation
 *
 * Any external functions defined by this module are only to be used by the
 * virtual memory system.
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include "opt_ddb.h"
#include "opt_vm.h"

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/lock.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mutex.h>
#if MAXMEMDOM > 1
#include <sys/proc.h>
#endif
#include <sys/queue.h>
#include <sys/rwlock.h>
#include <sys/sbuf.h>
#include <sys/sysctl.h>
#include <sys/tree.h>
#include <sys/vmmeter.h>
#include <sys/seq.h>

#include <ddb/ddb.h>

#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/vm_kern.h>
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/vm_phys.h>

#include <vm/vm_domain.h>

_Static_assert(sizeof(long) * NBBY >= VM_PHYSSEG_MAX,
    "Too many physsegs.");

struct mem_affinity *mem_affinity;
int *mem_locality;

int vm_ndomains = 1;

struct vm_phys_seg vm_phys_segs[VM_PHYSSEG_MAX];
int vm_phys_nsegs;

struct vm_phys_fictitious_seg;
static int vm_phys_fictitious_cmp(struct vm_phys_fictitious_seg *,
    struct vm_phys_fictitious_seg *);

RB_HEAD(fict_tree, vm_phys_fictitious_seg) vm_phys_fictitious_tree =
    RB_INITIALIZER(_vm_phys_fictitious_tree);

struct vm_phys_fictitious_seg {
	RB_ENTRY(vm_phys_fictitious_seg) node;
	/* Memory region data */
	vm_paddr_t	start;
	vm_paddr_t	end;
	vm_page_t	first_page;
};

RB_GENERATE_STATIC(fict_tree, vm_phys_fictitious_seg, node,
    vm_phys_fictitious_cmp);

static struct rwlock vm_phys_fictitious_reg_lock;
MALLOC_DEFINE(M_FICT_PAGES, "vm_fictitious", "Fictitious VM pages");

static struct vm_freelist
    vm_phys_free_queues[MAXMEMDOM][VM_NFREELIST][VM_NFREEPOOL][VM_NFREEORDER];

static int vm_nfreelists;

/*
 * Provides the mapping from VM_FREELIST_* to free list indices (flind).
 */
static int vm_freelist_to_flind[VM_NFREELIST];

CTASSERT(VM_FREELIST_DEFAULT == 0);

#ifdef VM_FREELIST_ISADMA
#define	VM_ISADMA_BOUNDARY	16777216
#endif
#ifdef VM_FREELIST_DMA32
#define	VM_DMA32_BOUNDARY	((vm_paddr_t)1 << 32)
#endif

/*
 * Enforce the assumptions made by vm_phys_add_seg() and vm_phys_init() about
 * the ordering of the free list boundaries.
 */
#if defined(VM_ISADMA_BOUNDARY) && defined(VM_LOWMEM_BOUNDARY)
CTASSERT(VM_ISADMA_BOUNDARY < VM_LOWMEM_BOUNDARY);
#endif
#if defined(VM_LOWMEM_BOUNDARY) && defined(VM_DMA32_BOUNDARY)
CTASSERT(VM_LOWMEM_BOUNDARY < VM_DMA32_BOUNDARY);
#endif

static int cnt_prezero;
SYSCTL_INT(_vm_stats_misc, OID_AUTO, cnt_prezero, CTLFLAG_RD,
    &cnt_prezero, 0, "The number of physical pages prezeroed at idle time");

static int sysctl_vm_phys_free(SYSCTL_HANDLER_ARGS);
SYSCTL_OID(_vm, OID_AUTO, phys_free, CTLTYPE_STRING | CTLFLAG_RD,
    NULL, 0, sysctl_vm_phys_free, "A", "Phys Free Info");

static int sysctl_vm_phys_segs(SYSCTL_HANDLER_ARGS);
SYSCTL_OID(_vm, OID_AUTO, phys_segs, CTLTYPE_STRING | CTLFLAG_RD,
    NULL, 0, sysctl_vm_phys_segs, "A", "Phys Seg Info");

#if MAXMEMDOM > 1
static int sysctl_vm_phys_locality(SYSCTL_HANDLER_ARGS);
SYSCTL_OID(_vm, OID_AUTO, phys_locality, CTLTYPE_STRING | CTLFLAG_RD,
    NULL, 0, sysctl_vm_phys_locality, "A", "Phys Locality Info");
#endif

SYSCTL_INT(_vm, OID_AUTO, ndomains, CTLFLAG_RD,
    &vm_ndomains, 0, "Number of physical memory domains available.");

/*
 * Default to first-touch + round-robin.
 */
static struct mtx vm_default_policy_mtx;
MTX_SYSINIT(vm_default_policy, &vm_default_policy_mtx, "default policy mutex",
    MTX_DEF);
#if MAXMEMDOM > 1
static struct vm_domain_policy vm_default_policy =
    VM_DOMAIN_POLICY_STATIC_INITIALISER(VM_POLICY_FIRST_TOUCH_ROUND_ROBIN, 0);
#else
/* Use round-robin so the domain policy code will only try once per allocation */
static struct vm_domain_policy vm_default_policy =
    VM_DOMAIN_POLICY_STATIC_INITIALISER(VM_POLICY_ROUND_ROBIN, 0);
#endif

static vm_page_t vm_phys_alloc_domain_pages(int domain, int flind, int pool,
    int order);
static void _vm_phys_create_seg(vm_paddr_t start, vm_paddr_t end, int domain);
static void vm_phys_create_seg(vm_paddr_t start, vm_paddr_t end);
static int vm_phys_paddr_to_segind(vm_paddr_t pa);
static void vm_phys_split_pages(vm_page_t m, int oind, struct vm_freelist *fl,
    int order);

static int
sysctl_vm_default_policy(SYSCTL_HANDLER_ARGS)
{
	char policy_name[32];
	int error;

	mtx_lock(&vm_default_policy_mtx);

	/* Map policy to output string */
	switch (vm_default_policy.p.policy) {
	case VM_POLICY_FIRST_TOUCH:
		strcpy(policy_name, "first-touch");
		break;
	case VM_POLICY_FIRST_TOUCH_ROUND_ROBIN:
		strcpy(policy_name, "first-touch-rr");
		break;
	case VM_POLICY_ROUND_ROBIN:
	default:
		strcpy(policy_name, "rr");
		break;
	}
	mtx_unlock(&vm_default_policy_mtx);

	error = sysctl_handle_string(oidp, &policy_name[0],
	    sizeof(policy_name), req);
	if (error != 0 || req->newptr == NULL)
		return (error);

	mtx_lock(&vm_default_policy_mtx);
	/* Set: match on the subset of policies that make sense as a default */
	if (strcmp("first-touch-rr", policy_name) == 0) {
		vm_domain_policy_set(&vm_default_policy,
		    VM_POLICY_FIRST_TOUCH_ROUND_ROBIN, 0);
	} else if (strcmp("first-touch", policy_name) == 0) {
		vm_domain_policy_set(&vm_default_policy,
		    VM_POLICY_FIRST_TOUCH, 0);
	} else if (strcmp("rr", policy_name) == 0) {
		vm_domain_policy_set(&vm_default_policy,
		    VM_POLICY_ROUND_ROBIN, 0);
	} else {
		error = EINVAL;
		goto finish;
	}

	error = 0;
finish:
	mtx_unlock(&vm_default_policy_mtx);
	return (error);
}

SYSCTL_PROC(_vm, OID_AUTO, default_policy, CTLTYPE_STRING | CTLFLAG_RW,
    0, 0, sysctl_vm_default_policy, "A",
    "Default policy (rr, first-touch, first-touch-rr");

/*
 * Red-black tree helpers for vm fictitious range management.
 */
static inline int
vm_phys_fictitious_in_range(struct vm_phys_fictitious_seg *p,
    struct vm_phys_fictitious_seg *range)
{

	KASSERT(range->start != 0 && range->end != 0,
	    ("Invalid range passed on search for vm_fictitious page"));
	if (p->start >= range->end)
		return (1);
	if (p->start < range->start)
		return (-1);

	return (0);
}

static int
vm_phys_fictitious_cmp(struct vm_phys_fictitious_seg *p1,
    struct vm_phys_fictitious_seg *p2)
{

	/* Check if this is a search for a page */
	if (p1->end == 0)
		return (vm_phys_fictitious_in_range(p1, p2));

	KASSERT(p2->end != 0,
    ("Invalid range passed as second parameter to vm fictitious comparison"));

	/* Searching to add a new range */
	if (p1->end <= p2->start)
		return (-1);
	if (p1->start >= p2->end)
		return (1);

	panic("Trying to add overlapping vm fictitious ranges:\n"
	    "[%#jx:%#jx] and [%#jx:%#jx]", (uintmax_t)p1->start,
	    (uintmax_t)p1->end, (uintmax_t)p2->start, (uintmax_t)p2->end);
}

static __inline int
vm_rr_selectdomain(void)
{
#if MAXMEMDOM > 1
	struct thread *td;

	td = curthread;

	td->td_dom_rr_idx++;
	td->td_dom_rr_idx %= vm_ndomains;
	return (td->td_dom_rr_idx);
#else
	return (0);
#endif
}

/*
 * Initialise a VM domain iterator.
 *
 * Check the thread policy, then the proc policy,
 * then default to the system policy.
 *
 * Later on the various layers will have this logic
 * plumbed into them and the phys code will be explicitly
 * handed a VM domain policy to use.
 */
static void
vm_policy_iterator_init(struct vm_domain_iterator *vi)
{
#if MAXMEMDOM > 1
	struct vm_domain_policy lcl;
#endif

	vm_domain_iterator_init(vi);

#if MAXMEMDOM > 1
	/* Copy out the thread policy */
	vm_domain_policy_localcopy(&lcl, &curthread->td_vm_dom_policy);
	if (lcl.p.policy != VM_POLICY_NONE) {
		/* Thread policy is present; use it */
		vm_domain_iterator_set_policy(vi, &lcl);
		return;
	}

	vm_domain_policy_localcopy(&lcl,
	    &curthread->td_proc->p_vm_dom_policy);
	if (lcl.p.policy != VM_POLICY_NONE) {
		/* Process policy is present; use it */
		vm_domain_iterator_set_policy(vi, &lcl);
		return;
	}
#endif
	/* Use system default policy */
	vm_domain_iterator_set_policy(vi, &vm_default_policy);
}

static void
vm_policy_iterator_finish(struct vm_domain_iterator *vi)
{

	vm_domain_iterator_cleanup(vi);
}

boolean_t
vm_phys_domain_intersects(long mask, vm_paddr_t low, vm_paddr_t high)
{
	struct vm_phys_seg *s;
	int idx;

	while ((idx = ffsl(mask)) != 0) {
		idx--;	/* ffsl counts from 1 */
		mask &= ~(1UL << idx);
		s = &vm_phys_segs[idx];
		if (low < s->end && high > s->start)
			return (TRUE);
	}
	return (FALSE);
}

/*
 * Outputs the state of the physical memory allocator, specifically,
 * the amount of physical memory in each free list.
 */
static int
sysctl_vm_phys_free(SYSCTL_HANDLER_ARGS)
{
	struct sbuf sbuf;
	struct vm_freelist *fl;
	int dom, error, flind, oind, pind;

	error = sysctl_wire_old_buffer(req, 0);
	if (error != 0)
		return (error);
	sbuf_new_for_sysctl(&sbuf, NULL, 128 * vm_ndomains, req);
	for (dom = 0; dom < vm_ndomains; dom++) {
		sbuf_printf(&sbuf,"\nDOMAIN %d:\n", dom);
		for (flind = 0; flind < vm_nfreelists; flind++) {
			sbuf_printf(&sbuf, "\nFREE LIST %d:\n"
			    "\n  ORDER (SIZE)  |  NUMBER"
			    "\n              ", flind);
			for (pind = 0; pind < VM_NFREEPOOL; pind++)
				sbuf_printf(&sbuf, "  |  POOL %d", pind);
			sbuf_printf(&sbuf, "\n--            ");
			for (pind = 0; pind < VM_NFREEPOOL; pind++)
				sbuf_printf(&sbuf, "-- --      ");
			sbuf_printf(&sbuf, "--\n");
			for (oind = VM_NFREEORDER - 1; oind >= 0; oind--) {
				sbuf_printf(&sbuf, "  %2d (%6dK)", oind,
				    1 << (PAGE_SHIFT - 10 + oind));
				for (pind = 0; pind < VM_NFREEPOOL; pind++) {
				fl = vm_phys_free_queues[dom][flind][pind];
					sbuf_printf(&sbuf, "  |  %6d",
					    fl[oind].lcnt);
				}
				sbuf_printf(&sbuf, "\n");
			}
		}
	}
	error = sbuf_finish(&sbuf);
	sbuf_delete(&sbuf);
	return (error);
}

/*
 * Outputs the set of physical memory segments.
 */
static int
sysctl_vm_phys_segs(SYSCTL_HANDLER_ARGS)
{
	struct sbuf sbuf;
	struct vm_phys_seg *seg;
	int error, segind;

	error = sysctl_wire_old_buffer(req, 0);
	if (error != 0)
		return (error);
	sbuf_new_for_sysctl(&sbuf, NULL, 128, req);
	for (segind = 0; segind < vm_phys_nsegs; segind++) {
		sbuf_printf(&sbuf, "\nSEGMENT %d:\n\n", segind);
		seg = &vm_phys_segs[segind];
		sbuf_printf(&sbuf, "start:     %#jx\n",
		    (uintmax_t)seg->start);
		sbuf_printf(&sbuf, "end:       %#jx\n",
		    (uintmax_t)seg->end);
		sbuf_printf(&sbuf, "domain:    %d\n", seg->domain);
		sbuf_printf(&sbuf, "free list: %p\n", seg->free_queues);
	}
	error = sbuf_finish(&sbuf);
	sbuf_delete(&sbuf);
	return (error);
}

/*
 * Return affinity, or -1 if there's no affinity information.
 */
int
vm_phys_mem_affinity(int f, int t)
{

#if MAXMEMDOM > 1
	if (mem_locality == NULL)
		return (-1);
	if (f >= vm_ndomains || t >= vm_ndomains)
		return (-1);
	return (mem_locality[f * vm_ndomains + t]);
#else
	return (-1);
#endif
}

#if MAXMEMDOM > 1
/*
 * Outputs the VM locality table.
 */
static int
sysctl_vm_phys_locality(SYSCTL_HANDLER_ARGS)
{
	struct sbuf sbuf;
	int error, i, j;

	error = sysctl_wire_old_buffer(req, 0);
	if (error != 0)
		return (error);
	sbuf_new_for_sysctl(&sbuf, NULL, 128, req);

	sbuf_printf(&sbuf, "\n");

	for (i = 0; i < vm_ndomains; i++) {
		sbuf_printf(&sbuf, "%d: ", i);
		for (j = 0; j < vm_ndomains; j++) {
			sbuf_printf(&sbuf, "%d ", vm_phys_mem_affinity(i, j));
		}
		sbuf_printf(&sbuf, "\n");
	}
	error = sbuf_finish(&sbuf);
	sbuf_delete(&sbuf);
	return (error);
}
#endif

static void
vm_freelist_add(struct vm_freelist *fl, vm_page_t m, int order, int tail)
{

	m->order = order;
	if (tail)
		TAILQ_INSERT_TAIL(&fl[order].pl, m, plinks.q);
	else
		TAILQ_INSERT_HEAD(&fl[order].pl, m, plinks.q);
	fl[order].lcnt++;
}

static void
vm_freelist_rem(struct vm_freelist *fl, vm_page_t m, int order)
{

	TAILQ_REMOVE(&fl[order].pl, m, plinks.q);
	fl[order].lcnt--;
	m->order = VM_NFREEORDER;
}

/*
 * Create a physical memory segment.
 */
static void
_vm_phys_create_seg(vm_paddr_t start, vm_paddr_t end, int domain)
{
	struct vm_phys_seg *seg;

	KASSERT(vm_phys_nsegs < VM_PHYSSEG_MAX,
	    ("vm_phys_create_seg: increase VM_PHYSSEG_MAX"));
	KASSERT(domain < vm_ndomains,
	    ("vm_phys_create_seg: invalid domain provided"));
	seg = &vm_phys_segs[vm_phys_nsegs++];
	while (seg > vm_phys_segs && (seg - 1)->start >= end) {
		*seg = *(seg - 1);
		seg--;
	}
	seg->start = start;
	seg->end = end;
	seg->domain = domain;
}

static void
vm_phys_create_seg(vm_paddr_t start, vm_paddr_t end)
{
	int i;

	if (mem_affinity == NULL) {
		_vm_phys_create_seg(start, end, 0);
		return;
	}

	for (i = 0;; i++) {
		if (mem_affinity[i].end == 0)
			panic("Reached end of affinity info");
		if (mem_affinity[i].end <= start)
			continue;
		if (mem_affinity[i].start > start)
			panic("No affinity info for start %jx",
			    (uintmax_t)start);
		if (mem_affinity[i].end >= end) {
			_vm_phys_create_seg(start, end,
			    mem_affinity[i].domain);
			break;
		}
		_vm_phys_create_seg(start, mem_affinity[i].end,
		    mem_affinity[i].domain);
		start = mem_affinity[i].end;
	}
}

/*
 * Add a physical memory segment.
 */
void
vm_phys_add_seg(vm_paddr_t start, vm_paddr_t end)
{
	vm_paddr_t paddr;

	KASSERT((start & PAGE_MASK) == 0,
	    ("vm_phys_define_seg: start is not page aligned"));
	KASSERT((end & PAGE_MASK) == 0,
	    ("vm_phys_define_seg: end is not page aligned"));

	/*
	 * Split the physical memory segment if it spans two or more free
	 * list boundaries.
	 */
	paddr = start;
#ifdef	VM_FREELIST_ISADMA
	if (paddr < VM_ISADMA_BOUNDARY && end > VM_ISADMA_BOUNDARY) {
		vm_phys_create_seg(paddr, VM_ISADMA_BOUNDARY);
		paddr = VM_ISADMA_BOUNDARY;
	}
#endif
#ifdef	VM_FREELIST_LOWMEM
	if (paddr < VM_LOWMEM_BOUNDARY && end > VM_LOWMEM_BOUNDARY) {
		vm_phys_create_seg(paddr, VM_LOWMEM_BOUNDARY);
		paddr = VM_LOWMEM_BOUNDARY;
	}
#endif
#ifdef	VM_FREELIST_DMA32
	if (paddr < VM_DMA32_BOUNDARY && end > VM_DMA32_BOUNDARY) {
		vm_phys_create_seg(paddr, VM_DMA32_BOUNDARY);
		paddr = VM_DMA32_BOUNDARY;
	}
#endif
	vm_phys_create_seg(paddr, end);
}

/*
 * Initialize the physical memory allocator.
 *
 * Requires that vm_page_array is initialized!
 */
void
vm_phys_init(void)
{
	struct vm_freelist *fl;
	struct vm_phys_seg *seg;
	u_long npages;
	int dom, flind, freelist, oind, pind, segind;

	/*
	 * Compute the number of free lists, and generate the mapping from the
	 * manifest constants VM_FREELIST_* to the free list indices.
	 *
	 * Initially, the entries of vm_freelist_to_flind[] are set to either
	 * 0 or 1 to indicate which free lists should be created.
	 */
	npages = 0;
	for (segind = vm_phys_nsegs - 1; segind >= 0; segind--) {
		seg = &vm_phys_segs[segind];
#ifdef	VM_FREELIST_ISADMA
		if (seg->end <= VM_ISADMA_BOUNDARY)
			vm_freelist_to_flind[VM_FREELIST_ISADMA] = 1;
		else
#endif
#ifdef	VM_FREELIST_LOWMEM
		if (seg->end <= VM_LOWMEM_BOUNDARY)
			vm_freelist_to_flind[VM_FREELIST_LOWMEM] = 1;
		else
#endif
#ifdef	VM_FREELIST_DMA32
		if (
#ifdef	VM_DMA32_NPAGES_THRESHOLD
		    /*
		     * Create the DMA32 free list only if the amount of
		     * physical memory above physical address 4G exceeds the
		     * given threshold.
		     */
		    npages > VM_DMA32_NPAGES_THRESHOLD &&
#endif
		    seg->end <= VM_DMA32_BOUNDARY)
			vm_freelist_to_flind[VM_FREELIST_DMA32] = 1;
		else
#endif
		{
			npages += atop(seg->end - seg->start);
			vm_freelist_to_flind[VM_FREELIST_DEFAULT] = 1;
		}
	}
	/* Change each entry into a running total of the free lists. */
	for (freelist = 1; freelist < VM_NFREELIST; freelist++) {
		vm_freelist_to_flind[freelist] +=
		    vm_freelist_to_flind[freelist - 1];
	}
	vm_nfreelists = vm_freelist_to_flind[VM_NFREELIST - 1];
	KASSERT(vm_nfreelists > 0, ("vm_phys_init: no free lists"));
	/* Change each entry into a free list index. */
	for (freelist = 0; freelist < VM_NFREELIST; freelist++)
		vm_freelist_to_flind[freelist]--;

	/*
	 * Initialize the first_page and free_queues fields of each physical
	 * memory segment.
	 */
#ifdef VM_PHYSSEG_SPARSE
	npages = 0;
#endif
	for (segind = 0; segind < vm_phys_nsegs; segind++) {
		seg = &vm_phys_segs[segind];
#ifdef VM_PHYSSEG_SPARSE
		seg->first_page = &vm_page_array[npages];
		npages += atop(seg->end - seg->start);
#else
		seg->first_page = PHYS_TO_VM_PAGE(seg->start);
#endif
#ifdef	VM_FREELIST_ISADMA
		if (seg->end <= VM_ISADMA_BOUNDARY) {
			flind = vm_freelist_to_flind[VM_FREELIST_ISADMA];
			KASSERT(flind >= 0,
			    ("vm_phys_init: ISADMA flind < 0"));
		} else
#endif
#ifdef	VM_FREELIST_LOWMEM
		if (seg->end <= VM_LOWMEM_BOUNDARY) {
			flind = vm_freelist_to_flind[VM_FREELIST_LOWMEM];
			KASSERT(flind >= 0,
			    ("vm_phys_init: LOWMEM flind < 0"));
		} else
#endif
#ifdef	VM_FREELIST_DMA32
		if (seg->end <= VM_DMA32_BOUNDARY) {
			flind = vm_freelist_to_flind[VM_FREELIST_DMA32];
			KASSERT(flind >= 0,
			    ("vm_phys_init: DMA32 flind < 0"));
		} else
#endif
		{
			flind = vm_freelist_to_flind[VM_FREELIST_DEFAULT];
			KASSERT(flind >= 0,
			    ("vm_phys_init: DEFAULT flind < 0"));
		}
		seg->free_queues = &vm_phys_free_queues[seg->domain][flind];
	}

	/*
	 * Initialize the free queues.
	 */
	for (dom = 0; dom < vm_ndomains; dom++) {
		for (flind = 0; flind < vm_nfreelists; flind++) {
			for (pind = 0; pind < VM_NFREEPOOL; pind++) {
				fl = vm_phys_free_queues[dom][flind][pind];
				for (oind = 0; oind < VM_NFREEORDER; oind++)
					TAILQ_INIT(&fl[oind].pl);
			}
		}
	}

	rw_init(&vm_phys_fictitious_reg_lock, "vmfctr");
}

/*
 * Split a contiguous, power of two-sized set of physical pages.
 */
static __inline void
vm_phys_split_pages(vm_page_t m, int oind, struct vm_freelist *fl, int order)
{
	vm_page_t m_buddy;

	while (oind > order) {
		oind--;
		m_buddy = &m[1 << oind];
		KASSERT(m_buddy->order == VM_NFREEORDER,
		    ("vm_phys_split_pages: page %p has unexpected order %d",
		    m_buddy, m_buddy->order));
		vm_freelist_add(fl, m_buddy, oind, 0);
        }
}

/*
 * Initialize a physical page and add it to the free lists.
 */
void
vm_phys_add_page(vm_paddr_t pa)
{
	vm_page_t m;
	struct vm_domain *vmd;

	vm_cnt.v_page_count++;
	m = vm_phys_paddr_to_vm_page(pa);
	m->phys_addr = pa;
	m->queue = PQ_NONE;
	m->segind = vm_phys_paddr_to_segind(pa);
	vmd = vm_phys_domain(m);
	vmd->vmd_page_count++;
	vmd->vmd_segs |= 1UL << m->segind;
	KASSERT(m->order == VM_NFREEORDER,
	    ("vm_phys_add_page: page %p has unexpected order %d",
	    m, m->order));
	m->pool = VM_FREEPOOL_DEFAULT;
	pmap_page_init(m);
	mtx_lock(&vm_page_queue_free_mtx);
	vm_phys_freecnt_adj(m, 1);
	vm_phys_free_pages(m, 0);
	mtx_unlock(&vm_page_queue_free_mtx);
}

/*
 * Allocate a contiguous, power of two-sized set of physical pages
 * from the free lists.
 *
 * The free page queues must be locked.
 */
vm_page_t
vm_phys_alloc_pages(int pool, int order)
{
	vm_page_t m;
	int domain, flind;
	struct vm_domain_iterator vi;

	KASSERT(pool < VM_NFREEPOOL,
	    ("vm_phys_alloc_pages: pool %d is out of range", pool));
	KASSERT(order < VM_NFREEORDER,
	    ("vm_phys_alloc_pages: order %d is out of range", order));

	vm_policy_iterator_init(&vi);

	while ((vm_domain_iterator_run(&vi, &domain)) == 0) {
		for (flind = 0; flind < vm_nfreelists; flind++) {
			m = vm_phys_alloc_domain_pages(domain, flind, pool,
			    order);
			if (m != NULL)
				return (m);
		}
	}

	vm_policy_iterator_finish(&vi);
	return (NULL);
}

/*
 * Allocate a contiguous, power of two-sized set of physical pages from the
 * specified free list.  The free list must be specified using one of the
 * manifest constants VM_FREELIST_*.
 *
 * The free page queues must be locked.
 */
vm_page_t
vm_phys_alloc_freelist_pages(int freelist, int pool, int order)
{
	vm_page_t m;
	struct vm_domain_iterator vi;
	int domain;

	KASSERT(freelist < VM_NFREELIST,
	    ("vm_phys_alloc_freelist_pages: freelist %d is out of range",
	    freelist));
	KASSERT(pool < VM_NFREEPOOL,
	    ("vm_phys_alloc_freelist_pages: pool %d is out of range", pool));
	KASSERT(order < VM_NFREEORDER,
	    ("vm_phys_alloc_freelist_pages: order %d is out of range", order));

	vm_policy_iterator_init(&vi);

	while ((vm_domain_iterator_run(&vi, &domain)) == 0) {
		m = vm_phys_alloc_domain_pages(domain,
		    vm_freelist_to_flind[freelist], pool, order);
		if (m != NULL)
			return (m);
	}

	vm_policy_iterator_finish(&vi);
	return (NULL);
}

static vm_page_t
vm_phys_alloc_domain_pages(int domain, int flind, int pool, int order)
{	
	struct vm_freelist *fl;
	struct vm_freelist *alt;
	int oind, pind;
	vm_page_t m;

	mtx_assert(&vm_page_queue_free_mtx, MA_OWNED);
	fl = &vm_phys_free_queues[domain][flind][pool][0];
	for (oind = order; oind < VM_NFREEORDER; oind++) {
		m = TAILQ_FIRST(&fl[oind].pl);
		if (m != NULL) {
			vm_freelist_rem(fl, m, oind);
			vm_phys_split_pages(m, oind, fl, order);
			return (m);
		}
	}

	/*
	 * The given pool was empty.  Find the largest
	 * contiguous, power-of-two-sized set of pages in any
	 * pool.  Transfer these pages to the given pool, and
	 * use them to satisfy the allocation.
	 */
	for (oind = VM_NFREEORDER - 1; oind >= order; oind--) {
		for (pind = 0; pind < VM_NFREEPOOL; pind++) {
			alt = &vm_phys_free_queues[domain][flind][pind][0];
			m = TAILQ_FIRST(&alt[oind].pl);
			if (m != NULL) {
				vm_freelist_rem(alt, m, oind);
				vm_phys_set_pool(pool, m, oind);
				vm_phys_split_pages(m, oind, fl, order);
				return (m);
			}
		}
	}
	return (NULL);
}

/*
 * Find the vm_page corresponding to the given physical address.
 */
vm_page_t
vm_phys_paddr_to_vm_page(vm_paddr_t pa)
{
	struct vm_phys_seg *seg;
	int segind;

	for (segind = 0; segind < vm_phys_nsegs; segind++) {
		seg = &vm_phys_segs[segind];
		if (pa >= seg->start && pa < seg->end)
			return (&seg->first_page[atop(pa - seg->start)]);
	}
	return (NULL);
}

vm_page_t
vm_phys_fictitious_to_vm_page(vm_paddr_t pa)
{
	struct vm_phys_fictitious_seg tmp, *seg;
	vm_page_t m;

	m = NULL;
	tmp.start = pa;
	tmp.end = 0;

	rw_rlock(&vm_phys_fictitious_reg_lock);
	seg = RB_FIND(fict_tree, &vm_phys_fictitious_tree, &tmp);
	rw_runlock(&vm_phys_fictitious_reg_lock);
	if (seg == NULL)
		return (NULL);

	m = &seg->first_page[atop(pa - seg->start)];
	KASSERT((m->flags & PG_FICTITIOUS) != 0, ("%p not fictitious", m));

	return (m);
}

static inline void
vm_phys_fictitious_init_range(vm_page_t range, vm_paddr_t start,
    long page_count, vm_memattr_t memattr)
{
	long i;

	for (i = 0; i < page_count; i++) {
		vm_page_initfake(&range[i], start + PAGE_SIZE * i, memattr);
		range[i].oflags &= ~VPO_UNMANAGED;
		range[i].busy_lock = VPB_UNBUSIED;
	}
}

int
vm_phys_fictitious_reg_range(vm_paddr_t start, vm_paddr_t end,
    vm_memattr_t memattr)
{
	struct vm_phys_fictitious_seg *seg;
	vm_page_t fp;
	long page_count;
#ifdef VM_PHYSSEG_DENSE
	long pi, pe;
	long dpage_count;
#endif

	KASSERT(start < end,
	    ("Start of segment isn't less than end (start: %jx end: %jx)",
	    (uintmax_t)start, (uintmax_t)end));

	page_count = (end - start) / PAGE_SIZE;

#ifdef VM_PHYSSEG_DENSE
	pi = atop(start);
	pe = atop(end);
	if (pi >= first_page && (pi - first_page) < vm_page_array_size) {
		fp = &vm_page_array[pi - first_page];
		if ((pe - first_page) > vm_page_array_size) {
			/*
			 * We have a segment that starts inside
			 * of vm_page_array, but ends outside of it.
			 *
			 * Use vm_page_array pages for those that are
			 * inside of the vm_page_array range, and
			 * allocate the remaining ones.
			 */
			dpage_count = vm_page_array_size - (pi - first_page);
			vm_phys_fictitious_init_range(fp, start, dpage_count,
			    memattr);
			page_count -= dpage_count;
			start += ptoa(dpage_count);
			goto alloc;
		}
		/*
		 * We can allocate the full range from vm_page_array,
		 * so there's no need to register the range in the tree.
		 */
		vm_phys_fictitious_init_range(fp, start, page_count, memattr);
		return (0);
	} else if (pe > first_page && (pe - first_page) < vm_page_array_size) {
		/*
		 * We have a segment that ends inside of vm_page_array,
		 * but starts outside of it.
		 */
		fp = &vm_page_array[0];
		dpage_count = pe - first_page;
		vm_phys_fictitious_init_range(fp, ptoa(first_page), dpage_count,
		    memattr);
		end -= ptoa(dpage_count);
		page_count -= dpage_count;
		goto alloc;
	} else if (pi < first_page && pe > (first_page + vm_page_array_size)) {
		/*
		 * Trying to register a fictitious range that expands before
		 * and after vm_page_array.
		 */
		return (EINVAL);
	} else {
alloc:
#endif
		fp = malloc(page_count * sizeof(struct vm_page), M_FICT_PAGES,
		    M_WAITOK | M_ZERO);
#ifdef VM_PHYSSEG_DENSE
	}
#endif
	vm_phys_fictitious_init_range(fp, start, page_count, memattr);

	seg = malloc(sizeof(*seg), M_FICT_PAGES, M_WAITOK | M_ZERO);
	seg->start = start;
	seg->end = end;
	seg->first_page = fp;

	rw_wlock(&vm_phys_fictitious_reg_lock);
	RB_INSERT(fict_tree, &vm_phys_fictitious_tree, seg);
	rw_wunlock(&vm_phys_fictitious_reg_lock);

	return (0);
}

void
vm_phys_fictitious_unreg_range(vm_paddr_t start, vm_paddr_t end)
{
	struct vm_phys_fictitious_seg *seg, tmp;
#ifdef VM_PHYSSEG_DENSE
	long pi, pe;
#endif

	KASSERT(start < end,
	    ("Start of segment isn't less than end (start: %jx end: %jx)",
	    (uintmax_t)start, (uintmax_t)end));

#ifdef VM_PHYSSEG_DENSE
	pi = atop(start);
	pe = atop(end);
	if (pi >= first_page && (pi - first_page) < vm_page_array_size) {
		if ((pe - first_page) <= vm_page_array_size) {
			/*
			 * This segment was allocated using vm_page_array
			 * only, there's nothing to do since those pages
			 * were never added to the tree.
			 */
			return;
		}
		/*
		 * We have a segment that starts inside
		 * of vm_page_array, but ends outside of it.
		 *
		 * Calculate how many pages were added to the
		 * tree and free them.
		 */
		start = ptoa(first_page + vm_page_array_size);
	} else if (pe > first_page && (pe - first_page) < vm_page_array_size) {
		/*
		 * We have a segment that ends inside of vm_page_array,
		 * but starts outside of it.
		 */
		end = ptoa(first_page);
	} else if (pi < first_page && pe > (first_page + vm_page_array_size)) {
		/* Since it's not possible to register such a range, panic. */
		panic(
		    "Unregistering not registered fictitious range [%#jx:%#jx]",
		    (uintmax_t)start, (uintmax_t)end);
	}
#endif
	tmp.start = start;
	tmp.end = 0;

	rw_wlock(&vm_phys_fictitious_reg_lock);
	seg = RB_FIND(fict_tree, &vm_phys_fictitious_tree, &tmp);
	if (seg->start != start || seg->end != end) {
		rw_wunlock(&vm_phys_fictitious_reg_lock);
		panic(
		    "Unregistering not registered fictitious range [%#jx:%#jx]",
		    (uintmax_t)start, (uintmax_t)end);
	}
	RB_REMOVE(fict_tree, &vm_phys_fictitious_tree, seg);
	rw_wunlock(&vm_phys_fictitious_reg_lock);
	free(seg->first_page, M_FICT_PAGES);
	free(seg, M_FICT_PAGES);
}

/*
 * Find the segment containing the given physical address.
 */
static int
vm_phys_paddr_to_segind(vm_paddr_t pa)
{
	struct vm_phys_seg *seg;
	int segind;

	for (segind = 0; segind < vm_phys_nsegs; segind++) {
		seg = &vm_phys_segs[segind];
		if (pa >= seg->start && pa < seg->end)
			return (segind);
	}
	panic("vm_phys_paddr_to_segind: paddr %#jx is not in any segment" ,
	    (uintmax_t)pa);
}

/*
 * Free a contiguous, power of two-sized set of physical pages.
 *
 * The free page queues must be locked.
 */
void
vm_phys_free_pages(vm_page_t m, int order)
{
	struct vm_freelist *fl;
	struct vm_phys_seg *seg;
	vm_paddr_t pa;
	vm_page_t m_buddy;

	KASSERT(m->order == VM_NFREEORDER,
	    ("vm_phys_free_pages: page %p has unexpected order %d",
	    m, m->order));
	KASSERT(m->pool < VM_NFREEPOOL,
	    ("vm_phys_free_pages: page %p has unexpected pool %d",
	    m, m->pool));
	KASSERT(order < VM_NFREEORDER,
	    ("vm_phys_free_pages: order %d is out of range", order));
	mtx_assert(&vm_page_queue_free_mtx, MA_OWNED);
	seg = &vm_phys_segs[m->segind];
	if (order < VM_NFREEORDER - 1) {
		pa = VM_PAGE_TO_PHYS(m);
		do {
			pa ^= ((vm_paddr_t)1 << (PAGE_SHIFT + order));
			if (pa < seg->start || pa >= seg->end)
				break;
			m_buddy = &seg->first_page[atop(pa - seg->start)];
			if (m_buddy->order != order)
				break;
			fl = (*seg->free_queues)[m_buddy->pool];
			vm_freelist_rem(fl, m_buddy, order);
			if (m_buddy->pool != m->pool)
				vm_phys_set_pool(m->pool, m_buddy, order);
			order++;
			pa &= ~(((vm_paddr_t)1 << (PAGE_SHIFT + order)) - 1);
			m = &seg->first_page[atop(pa - seg->start)];
		} while (order < VM_NFREEORDER - 1);
	}
	fl = (*seg->free_queues)[m->pool];
	vm_freelist_add(fl, m, order, 1);
}

/*
 * Free a contiguous, arbitrarily sized set of physical pages.
 *
 * The free page queues must be locked.
 */
void
vm_phys_free_contig(vm_page_t m, u_long npages)
{
	u_int n;
	int order;

	/*
	 * Avoid unnecessary coalescing by freeing the pages in the largest
	 * possible power-of-two-sized subsets.
	 */
	mtx_assert(&vm_page_queue_free_mtx, MA_OWNED);
	for (;; npages -= n) {
		/*
		 * Unsigned "min" is used here so that "order" is assigned
		 * "VM_NFREEORDER - 1" when "m"'s physical address is zero
		 * or the low-order bits of its physical address are zero
		 * because the size of a physical address exceeds the size of
		 * a long.
		 */
		order = min(ffsl(VM_PAGE_TO_PHYS(m) >> PAGE_SHIFT) - 1,
		    VM_NFREEORDER - 1);
		n = 1 << order;
		if (npages < n)
			break;
		vm_phys_free_pages(m, order);
		m += n;
	}
	/* The residual "npages" is less than "1 << (VM_NFREEORDER - 1)". */
	for (; npages > 0; npages -= n) {
		order = flsl(npages) - 1;
		n = 1 << order;
		vm_phys_free_pages(m, order);
		m += n;
	}
}

/*
 * Set the pool for a contiguous, power of two-sized set of physical pages. 
 */
void
vm_phys_set_pool(int pool, vm_page_t m, int order)
{
	vm_page_t m_tmp;

	for (m_tmp = m; m_tmp < &m[1 << order]; m_tmp++)
		m_tmp->pool = pool;
}

/*
 * Search for the given physical page "m" in the free lists.  If the search
 * succeeds, remove "m" from the free lists and return TRUE.  Otherwise, return
 * FALSE, indicating that "m" is not in the free lists.
 *
 * The free page queues must be locked.
 */
boolean_t
vm_phys_unfree_page(vm_page_t m)
{
	struct vm_freelist *fl;
	struct vm_phys_seg *seg;
	vm_paddr_t pa, pa_half;
	vm_page_t m_set, m_tmp;
	int order;

	mtx_assert(&vm_page_queue_free_mtx, MA_OWNED);

	/*
	 * First, find the contiguous, power of two-sized set of free
	 * physical pages containing the given physical page "m" and
	 * assign it to "m_set".
	 */
	seg = &vm_phys_segs[m->segind];
	for (m_set = m, order = 0; m_set->order == VM_NFREEORDER &&
	    order < VM_NFREEORDER - 1; ) {
		order++;
		pa = m->phys_addr & (~(vm_paddr_t)0 << (PAGE_SHIFT + order));
		if (pa >= seg->start)
			m_set = &seg->first_page[atop(pa - seg->start)];
		else
			return (FALSE);
	}
	if (m_set->order < order)
		return (FALSE);
	if (m_set->order == VM_NFREEORDER)
		return (FALSE);
	KASSERT(m_set->order < VM_NFREEORDER,
	    ("vm_phys_unfree_page: page %p has unexpected order %d",
	    m_set, m_set->order));

	/*
	 * Next, remove "m_set" from the free lists.  Finally, extract
	 * "m" from "m_set" using an iterative algorithm: While "m_set"
	 * is larger than a page, shrink "m_set" by returning the half
	 * of "m_set" that does not contain "m" to the free lists.
	 */
	fl = (*seg->free_queues)[m_set->pool];
	order = m_set->order;
	vm_freelist_rem(fl, m_set, order);
	while (order > 0) {
		order--;
		pa_half = m_set->phys_addr ^ (1 << (PAGE_SHIFT + order));
		if (m->phys_addr < pa_half)
			m_tmp = &seg->first_page[atop(pa_half - seg->start)];
		else {
			m_tmp = m_set;
			m_set = &seg->first_page[atop(pa_half - seg->start)];
		}
		vm_freelist_add(fl, m_tmp, order, 0);
	}
	KASSERT(m_set == m, ("vm_phys_unfree_page: fatal inconsistency"));
	return (TRUE);
}

/*
 * Try to zero one physical page.  Used by an idle priority thread.
 */
boolean_t
vm_phys_zero_pages_idle(void)
{
	static struct vm_freelist *fl;
	static int flind, oind, pind;
	vm_page_t m, m_tmp;
	int domain;

	domain = vm_rr_selectdomain();
	fl = vm_phys_free_queues[domain][0][0];
	mtx_assert(&vm_page_queue_free_mtx, MA_OWNED);
	for (;;) {
		TAILQ_FOREACH_REVERSE(m, &fl[oind].pl, pglist, plinks.q) {
			for (m_tmp = m; m_tmp < &m[1 << oind]; m_tmp++) {
				if ((m_tmp->flags & (PG_CACHED | PG_ZERO)) == 0) {
					vm_phys_unfree_page(m_tmp);
					vm_phys_freecnt_adj(m, -1);
					mtx_unlock(&vm_page_queue_free_mtx);
					pmap_zero_page_idle(m_tmp);
					m_tmp->flags |= PG_ZERO;
					mtx_lock(&vm_page_queue_free_mtx);
					vm_phys_freecnt_adj(m, 1);
					vm_phys_free_pages(m_tmp, 0);
					vm_page_zero_count++;
					cnt_prezero++;
					return (TRUE);
				}
			}
		}
		oind++;
		if (oind == VM_NFREEORDER) {
			oind = 0;
			pind++;
			if (pind == VM_NFREEPOOL) {
				pind = 0;
				flind++;
				if (flind == vm_nfreelists)
					flind = 0;
			}
			fl = vm_phys_free_queues[domain][flind][pind];
		}
	}
}

/*
 * Allocate a contiguous set of physical pages of the given size
 * "npages" from the free lists.  All of the physical pages must be at
 * or above the given physical address "low" and below the given
 * physical address "high".  The given value "alignment" determines the
 * alignment of the first physical page in the set.  If the given value
 * "boundary" is non-zero, then the set of physical pages cannot cross
 * any physical address boundary that is a multiple of that value.  Both
 * "alignment" and "boundary" must be a power of two.
 */
vm_page_t
vm_phys_alloc_contig(u_long npages, vm_paddr_t low, vm_paddr_t high,
    u_long alignment, vm_paddr_t boundary)
{
	struct vm_freelist *fl;
	struct vm_phys_seg *seg;
	vm_paddr_t pa, pa_last, size;
	vm_page_t m, m_ret;
	u_long npages_end;
	int domain, flind, oind, order, pind;
	struct vm_domain_iterator vi;

	mtx_assert(&vm_page_queue_free_mtx, MA_OWNED);
	size = npages << PAGE_SHIFT;
	KASSERT(size != 0,
	    ("vm_phys_alloc_contig: size must not be 0"));
	KASSERT((alignment & (alignment - 1)) == 0,
	    ("vm_phys_alloc_contig: alignment must be a power of 2"));
	KASSERT((boundary & (boundary - 1)) == 0,
	    ("vm_phys_alloc_contig: boundary must be a power of 2"));
	/* Compute the queue that is the best fit for npages. */
	for (order = 0; (1 << order) < npages; order++);

	vm_policy_iterator_init(&vi);

restartdom:
	if (vm_domain_iterator_run(&vi, &domain) != 0) {
		vm_policy_iterator_finish(&vi);
		return (NULL);
	}

	for (flind = 0; flind < vm_nfreelists; flind++) {
		for (oind = min(order, VM_NFREEORDER - 1); oind < VM_NFREEORDER; oind++) {
			for (pind = 0; pind < VM_NFREEPOOL; pind++) {
				fl = &vm_phys_free_queues[domain][flind][pind][0];
				TAILQ_FOREACH(m_ret, &fl[oind].pl, plinks.q) {
					/*
					 * A free list may contain physical pages
					 * from one or more segments.
					 */
					seg = &vm_phys_segs[m_ret->segind];
					if (seg->start > high ||
					    low >= seg->end)
						continue;

					/*
					 * Is the size of this allocation request
					 * larger than the largest block size?
					 */
					if (order >= VM_NFREEORDER) {
						/*
						 * Determine if a sufficient number
						 * of subsequent blocks to satisfy
						 * the allocation request are free.
						 */
						pa = VM_PAGE_TO_PHYS(m_ret);
						pa_last = pa + size;
						for (;;) {
							pa += 1 << (PAGE_SHIFT + VM_NFREEORDER - 1);
							if (pa >= pa_last)
								break;
							if (pa < seg->start ||
							    pa >= seg->end)
								break;
							m = &seg->first_page[atop(pa - seg->start)];
							if (m->order != VM_NFREEORDER - 1)
								break;
						}
						/* If not, continue to the next block. */
						if (pa < pa_last)
							continue;
					}

					/*
					 * Determine if the blocks are within the given range,
					 * satisfy the given alignment, and do not cross the
					 * given boundary.
					 */
					pa = VM_PAGE_TO_PHYS(m_ret);
					if (pa >= low &&
					    pa + size <= high &&
					    (pa & (alignment - 1)) == 0 &&
					    ((pa ^ (pa + size - 1)) & ~(boundary - 1)) == 0)
						goto done;
				}
			}
		}
	}
	if (!vm_domain_iterator_isdone(&vi))
		goto restartdom;
	vm_policy_iterator_finish(&vi);
	return (NULL);
done:
	for (m = m_ret; m < &m_ret[npages]; m = &m[1 << oind]) {
		fl = (*seg->free_queues)[m->pool];
		vm_freelist_rem(fl, m, m->order);
	}
	if (m_ret->pool != VM_FREEPOOL_DEFAULT)
		vm_phys_set_pool(VM_FREEPOOL_DEFAULT, m_ret, oind);
	fl = (*seg->free_queues)[m_ret->pool];
	vm_phys_split_pages(m_ret, oind, fl, order);
	/* Return excess pages to the free lists. */
	npages_end = roundup2(npages, 1 << imin(oind, order));
	if (npages < npages_end)
		vm_phys_free_contig(&m_ret[npages], npages_end - npages);
	return (m_ret);
}

#ifdef DDB
/*
 * Show the number of physical pages in each of the free lists.
 */
DB_SHOW_COMMAND(freepages, db_show_freepages)
{
	struct vm_freelist *fl;
	int flind, oind, pind, dom;

	for (dom = 0; dom < vm_ndomains; dom++) {
		db_printf("DOMAIN: %d\n", dom);
		for (flind = 0; flind < vm_nfreelists; flind++) {
			db_printf("FREE LIST %d:\n"
			    "\n  ORDER (SIZE)  |  NUMBER"
			    "\n              ", flind);
			for (pind = 0; pind < VM_NFREEPOOL; pind++)
				db_printf("  |  POOL %d", pind);
			db_printf("\n--            ");
			for (pind = 0; pind < VM_NFREEPOOL; pind++)
				db_printf("-- --      ");
			db_printf("--\n");
			for (oind = VM_NFREEORDER - 1; oind >= 0; oind--) {
				db_printf("  %2.2d (%6.6dK)", oind,
				    1 << (PAGE_SHIFT - 10 + oind));
				for (pind = 0; pind < VM_NFREEPOOL; pind++) {
				fl = vm_phys_free_queues[dom][flind][pind];
					db_printf("  |  %6.6d", fl[oind].lcnt);
				}
				db_printf("\n");
			}
			db_printf("\n");
		}
		db_printf("\n");
	}
}
#endif
OpenPOWER on IntegriCloud