summaryrefslogtreecommitdiff
path: root/sys/arch/i386/i386/pmap.c
blob: dace4b52bafccdafd9022094589f2fab9171145a (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
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
/*	$OpenBSD: pmap.c,v 1.18 1997/08/17 17:31:37 grr Exp $	*/
/*	$NetBSD: pmap.c,v 1.36 1996/05/03 19:42:22 christos Exp $	*/

/*
 * Copyright (c) 1993, 1994, 1995 Charles M. Hannum.  All rights reserved.
 * Copyright (c) 1991 Regents of the University of California.
 * All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * the Systems Programming Group of the University of Utah Computer
 * Science Department and William Jolitz of UUNET Technologies Inc.
 *
 * 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.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
 *
 *	@(#)pmap.c	7.7 (Berkeley)	5/12/91
 */

/*
 * Derived originally from an old hp300 version by Mike Hibler.  The version
 * by William Jolitz has been heavily modified to allow non-contiguous
 * mapping of physical memory by Wolfgang Solfrank, and to fix several bugs
 * and greatly speedup it up by Charles Hannum.
 * 
 * A recursive map [a pde which points to the page directory] is used to map
 * the page tables using the pagetables themselves. This is done to reduce
 * the impact on kernel virtual memory for lots of sparse address space, and
 * to reduce the cost of memory to each process.
 */

/*
 *	Manages physical address maps.
 *
 *	In addition to hardware address maps, this
 *	module is called upon to provide software-use-only
 *	maps which may or may not be stored in the same
 *	form as hardware maps.  These pseudo-maps are
 *	used to store intermediate results from copy
 *	operations to and from address spaces.
 *
 *	Since the information managed by this module is
 *	also stored by the logical address mapping module,
 *	this module may throw away valid virtual-to-physical
 *	mappings at almost any time.  However, invalidations
 *	of virtual-to-physical mappings must be done as
 *	requested.
 *
 *	In order to cope with hardware architectures which
 *	make virtual-to-physical map invalidates expensive,
 *	this module may delay invalidate or reduced protection
 *	operations until such time as they are actually
 *	necessary.  This module is given full information as
 *	to which processors are currently using which maps,
 *	and to when physical maps must be made correct.
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/malloc.h>
#include <sys/user.h>

#include <vm/vm.h>
#include <vm/vm_kern.h>
#include <vm/vm_page.h>

#include <machine/cpu.h>

#include <dev/isa/isareg.h>
#include <i386/isa/isa_machdep.h>

#include "isa.h"
#include "isadma.h"

/*
 * Allocate various and sundry SYSMAPs used in the days of old VM
 * and not yet converted.  XXX.
 */
#define	BSDVM_COMPAT	1

#ifdef DEBUG
struct {
	int kernel;	/* entering kernel mapping */
	int user;	/* entering user mapping */
	int ptpneeded;	/* needed to allocate a PT page */
	int pwchange;	/* no mapping change, just wiring or protection */
	int wchange;	/* no mapping change, just wiring */
	int mchange;	/* was mapped but mapping to different page */
	int managed;	/* a managed page */
	int firstpv;	/* first mapping for this PA */
	int secondpv;	/* second mapping for this PA */
	int ci;		/* cache inhibited */
	int unmanaged;	/* not a managed page */
	int flushes;	/* cache flushes */
} enter_stats;
struct {
	int calls;
	int removes;
	int pvfirst;
	int pvsearch;
	int ptinvalid;
	int uflushes;
	int sflushes;
} remove_stats;

int pmapdebug = 0 /* 0xffff */;
#define	PDB_FOLLOW	0x0001
#define	PDB_INIT	0x0002
#define	PDB_ENTER	0x0004
#define	PDB_REMOVE	0x0008
#define	PDB_CREATE	0x0010
#define	PDB_PTPAGE	0x0020
#define	PDB_CACHE	0x0040
#define	PDB_BITS	0x0080
#define	PDB_COLLECT	0x0100
#define	PDB_PROTECT	0x0200
#define	PDB_PDRTAB	0x0400
#define	PDB_PARANOIA	0x2000
#define	PDB_WIRING	0x4000
#define	PDB_PVDUMP	0x8000
#endif

/*
 * Get PDEs and PTEs for user/kernel address space
 */
#define	pmap_pde(m, v)	(&((m)->pm_pdir[((vm_offset_t)(v) >> PDSHIFT)&1023]))

/*
 * Empty PTEs and PDEs are always 0, but checking only the valid bit allows
 * the compiler to generate `testb' rather than `testl'.
 */
#define	pmap_pde_v(pde)			(*(pde) & PG_V)
#define	pmap_pte_pa(pte)		(*(pte) & PG_FRAME)
#define	pmap_pte_w(pte)			(*(pte) & PG_W)
#define	pmap_pte_m(pte)			(*(pte) & PG_M)
#define	pmap_pte_u(pte)			(*(pte) & PG_U)
#define	pmap_pte_v(pte)			(*(pte) & PG_V)
#define	pmap_pte_set_w(pte, v)		((v) ? (*(pte) |= PG_W) : (*(pte) &= ~PG_W))
#define	pmap_pte_set_prot(pte, v)	((*(pte) &= ~PG_PROT), (*(pte) |= (v)))

/*
 * Given a map and a machine independent protection code,
 * convert to a vax protection code.
 */
pt_entry_t	protection_codes[8];

struct pmap	kernel_pmap_store;

vm_offset_t    	avail_start;	/* PA of first available physical page */
vm_offset_t	avail_end;	/* PA of last available physical page */
vm_offset_t	virtual_avail;  /* VA of first avail page (after kernel bss)*/
vm_offset_t	virtual_end;	/* VA of last avail page (end of kernel AS) */
int		npages;

boolean_t	pmap_initialized = FALSE;	/* Has pmap_init completed? */
char		*pmap_attributes;	/* reference and modify bits */
TAILQ_HEAD(pv_page_list, pv_page) pv_page_freelist;
int		pv_nfree;

pt_entry_t *pmap_pte __P((pmap_t, vm_offset_t));
struct pv_entry * pmap_alloc_pv __P((void));
void pmap_free_pv __P((struct pv_entry *));
void i386_protection_init __P((void));
void pmap_collect_pv __P((void));
__inline void pmap_remove_pv __P((pmap_t, vm_offset_t, u_int));
__inline void pmap_enter_pv __P((pmap_t, vm_offset_t, u_int));
void pmap_deactivate __P((pmap_t, struct pcb *));
void pmap_remove_all __P((vm_offset_t));
void pads __P((pmap_t pm));
void pmap_dump_pvlist __P((vm_offset_t phys, char *m));
void pmap_pvdump __P((vm_offset_t pa));

#if BSDVM_COMPAT
#include <sys/msgbuf.h>

/*
 * All those kernel PT submaps that BSD is so fond of
 */
pt_entry_t	*CMAP1, *CMAP2, *XXX_mmap;
caddr_t		CADDR1, CADDR2, vmmap;
pt_entry_t	*msgbufmap;
#endif	/* BSDVM_COMPAT */

/*
 *	Bootstrap the system enough to run with virtual memory.
 *	Map the kernel's code and data, and allocate the system page table.
 *
 *	On the I386 this is called after mapping has already been enabled
 *	and just syncs the pmap module with what has already been done.
 *	[We can't call it easily with mapping off since the kernel is not
 *	mapped with PA == VA, hence we would have to relocate every address
 *	from the linked base (virtual) address to the actual (physical)
 *	address starting relative to 0]
 */

void
pmap_bootstrap(virtual_start)
	vm_offset_t virtual_start;
{
#if BSDVM_COMPAT
	vm_offset_t va;
	pt_entry_t *pte;
#endif

	/* XXX: allow for msgbuf */
	avail_end -= i386_round_page(sizeof(struct msgbuf));

	virtual_avail = virtual_start;
	virtual_end = VM_MAX_KERNEL_ADDRESS;

	/*
	 * Initialize protection array.
	 */
	i386_protection_init();

#ifdef notdef
	/*
	 * Create Kernel page directory table and page maps.
	 * [ currently done in locore. i have wild and crazy ideas -wfj ]
	 */
	bzero(firstaddr, (1+NKPDE)*NBPG);
	pmap_kernel()->pm_pdir = firstaddr + VM_MIN_KERNEL_ADDRESS;
	pmap_kernel()->pm_ptab = firstaddr + VM_MIN_KERNEL_ADDRESS + NBPG;

	firstaddr += NBPG;
	for (x = i386_btod(VM_MIN_KERNEL_ADDRESS);
	     x < i386_btod(VM_MIN_KERNEL_ADDRESS) + NKPDE; x++) {
		pd_entry_t *pde;
		pde = pmap_kernel()->pm_pdir + x;
		*pde = (firstaddr + x*NBPG) | PG_V | PG_KW;
	}
#else
	pmap_kernel()->pm_pdir =
	    (pd_entry_t *)(proc0.p_addr->u_pcb.pcb_cr3 + KERNBASE);
#endif

	simple_lock_init(&pmap_kernel()->pm_lock);
	pmap_kernel()->pm_count = 1;

#if BSDVM_COMPAT
	/*
	 * Allocate all the submaps we need
	 */
#define	SYSMAP(c, p, v, n)	\
	v = (c)va; va += ((n)*NBPG); p = pte; pte += (n);

	va = virtual_avail;
	pte = pmap_pte(pmap_kernel(), va);

	SYSMAP(caddr_t		,CMAP1		,CADDR1	   ,1		)
	SYSMAP(caddr_t		,CMAP2		,CADDR2	   ,1		)
	SYSMAP(caddr_t		,XXX_mmap	,vmmap	   ,1		)
	SYSMAP(struct msgbuf *	,msgbufmap	,msgbufp   ,1		)
	virtual_avail = va;
#endif

	/*
	 * Reserve pmap space for mapping physical pages during dump.
	 */
	virtual_avail = reserve_dumppages(virtual_avail);

	/*
	 * reserve special hunk of memory for use by bus dma as a bounce
	 * buffer (contiguous virtual *and* physical memory).  XXX
	 */
#if NISA > 0 && NISADMA > 0
	if (ctob(physmem) >= 0x1000000) {
		isaphysmem = pmap_steal_memory(DMA_BOUNCE * NBPG);
		isaphysmempgs = DMA_BOUNCE;
	} else {
		isaphysmem = pmap_steal_memory(DMA_BOUNCE_LOW * NBPG);
		isaphysmempgs = DMA_BOUNCE_LOW;
	}
#endif

	/* flawed, no mappings?? */
	if (ctob(physmem) > 31*1024*1024 && MAXKPDE != NKPDE) {
		vm_offset_t p;
		int i;

		p = pmap_steal_memory((MAXKPDE-NKPDE+1) * NBPG);
		bzero((void *)p, (MAXKPDE-NKPDE+1) * NBPG);
		p = round_page(p);
		for (i = NKPDE; i < MAXKPDE; i++, p += NBPG)
			PTD[KPTDI+i] = (pd_entry_t)p |
			    PG_V | PG_KW;
	}

	pmap_update();
}

void
pmap_virtual_space(startp, endp)
	vm_offset_t *startp;
	vm_offset_t *endp;
{

	*startp = virtual_avail;
	*endp = virtual_end;
}

/*
 *	Initialize the pmap module.
 *	Called by vm_init, to initialize any structures that the pmap
 *	system needs to map virtual memory.
 */
void
pmap_init()
{
	vm_offset_t addr;
	vm_size_t s;

	if (PAGE_SIZE != NBPG)
		panic("pmap_init: CLSIZE != 1");

	npages = pmap_page_index(avail_end - 1) + 1;
	s = (vm_size_t) (sizeof(struct pv_entry) * npages + npages);
	s = round_page(s);
	addr = (vm_offset_t) kmem_alloc(kernel_map, s);
	pv_table = (struct pv_entry *) addr;
	addr += sizeof(struct pv_entry) * npages;
	pmap_attributes = (char *) addr;
	TAILQ_INIT(&pv_page_freelist);

#ifdef DEBUG
	if (pmapdebug & PDB_INIT)
		printf("pmap_init: %x bytes (%x pgs): tbl %x attr %x\n",
		       s, npages, pv_table, pmap_attributes);
#endif

	/*
	 * Now it is safe to enable pv_table recording.
	 */
	pmap_initialized = TRUE;
}

struct pv_entry *
pmap_alloc_pv()
{
	struct pv_page *pvp;
	struct pv_entry *pv;
	int i;

	if (pv_nfree == 0) {
		pvp = (struct pv_page *)kmem_alloc(kernel_map, NBPG);
		if (pvp == 0)
			panic("pmap_alloc_pv: kmem_alloc() failed");
		pvp->pvp_pgi.pgi_freelist = pv = &pvp->pvp_pv[1];
		for (i = NPVPPG - 2; i; i--, pv++)
			pv->pv_next = pv + 1;
		pv->pv_next = 0;
		pv_nfree += pvp->pvp_pgi.pgi_nfree = NPVPPG - 1;
		TAILQ_INSERT_HEAD(&pv_page_freelist, pvp, pvp_pgi.pgi_list);
		pv = &pvp->pvp_pv[0];
	} else {
		--pv_nfree;
		pvp = pv_page_freelist.tqh_first;
		if (--pvp->pvp_pgi.pgi_nfree == 0) {
			TAILQ_REMOVE(&pv_page_freelist, pvp, pvp_pgi.pgi_list);
		}
		pv = pvp->pvp_pgi.pgi_freelist;
#ifdef DIAGNOSTIC
		if (pv == 0)
			panic("pmap_alloc_pv: pgi_nfree inconsistent");
#endif
		pvp->pvp_pgi.pgi_freelist = pv->pv_next;
	}
	return pv;
}

void
pmap_free_pv(pv)
	struct pv_entry *pv;
{
	register struct pv_page *pvp;

	pvp = (struct pv_page *) trunc_page(pv);
	switch (++pvp->pvp_pgi.pgi_nfree) {
	case 1:
		TAILQ_INSERT_TAIL(&pv_page_freelist, pvp, pvp_pgi.pgi_list);
	default:
		pv->pv_next = pvp->pvp_pgi.pgi_freelist;
		pvp->pvp_pgi.pgi_freelist = pv;
		++pv_nfree;
		break;
	case NPVPPG:
		pv_nfree -= NPVPPG - 1;
		TAILQ_REMOVE(&pv_page_freelist, pvp, pvp_pgi.pgi_list);
		kmem_free(kernel_map, (vm_offset_t)pvp, NBPG);
		break;
	}
}

void
pmap_collect_pv()
{
	struct pv_page_list pv_page_collectlist;
	struct pv_page *pvp, *npvp;
	struct pv_entry *ph, *ppv, *pv, *npv;
	int s;

	TAILQ_INIT(&pv_page_collectlist);

	for (pvp = pv_page_freelist.tqh_first; pvp; pvp = npvp) {
		if (pv_nfree < NPVPPG)
			break;
		npvp = pvp->pvp_pgi.pgi_list.tqe_next;
		if (pvp->pvp_pgi.pgi_nfree > NPVPPG / 3) {
			TAILQ_REMOVE(&pv_page_freelist, pvp, pvp_pgi.pgi_list);
			TAILQ_INSERT_TAIL(&pv_page_collectlist, pvp, pvp_pgi.pgi_list);
			pv_nfree -= pvp->pvp_pgi.pgi_nfree;
			pvp->pvp_pgi.pgi_nfree = -1;
		}
	}

	if (pv_page_collectlist.tqh_first == 0)
		return;

	for (ph = &pv_table[npages - 1]; ph >= &pv_table[0]; ph--) {
		if (ph->pv_pmap == 0)
			continue;
		s = splimp();
		for (ppv = ph; (pv = ppv->pv_next) != 0; ) {
			pvp = (struct pv_page *) trunc_page(pv);
			if (pvp->pvp_pgi.pgi_nfree == -1) {
				pvp = pv_page_freelist.tqh_first;
				if (--pvp->pvp_pgi.pgi_nfree == 0) {
					TAILQ_REMOVE(&pv_page_freelist, pvp, pvp_pgi.pgi_list);
				}
				npv = pvp->pvp_pgi.pgi_freelist;
#ifdef DIAGNOSTIC
				if (npv == 0)
					panic("pmap_collect_pv: pgi_nfree inconsistent");
#endif
				pvp->pvp_pgi.pgi_freelist = npv->pv_next;
				*npv = *pv;
				ppv->pv_next = npv;
				ppv = npv;
			} else
				ppv = pv;
		}
		splx(s);
	}

	for (pvp = pv_page_collectlist.tqh_first; pvp; pvp = npvp) {
		npvp = pvp->pvp_pgi.pgi_list.tqe_next;
		kmem_free(kernel_map, (vm_offset_t)pvp, NBPG);
	}
}

__inline void
pmap_enter_pv(pmap, va, pind)
	register pmap_t pmap;
	vm_offset_t va;
	u_int pind;
{	
	register struct pv_entry *pv, *npv;
	int s;

	if (!pmap_initialized)
		return;

#ifdef DEBUG
	if (pmapdebug & PDB_ENTER)
		printf("pmap_enter_pv: pv %x: %x/%x/%x\n",
		       pv, pv->pv_va, pv->pv_pmap, pv->pv_next);
#endif

	pv = &pv_table[pind];
	s = splimp();

	if (pv->pv_pmap == NULL) {
		/*
		 * No entries yet, use header as the first entry
		 */
#ifdef DEBUG
		enter_stats.firstpv++;
#endif
		pv->pv_va = va;
		pv->pv_pmap = pmap;
		pv->pv_next = NULL;
	} else {
		/*
		 * There is at least one other VA mapping this page.
		 * Place this entry after the header.
		 */
#ifdef DEBUG
		for (npv = pv; npv; npv = npv->pv_next)
			if (pmap == npv->pv_pmap && va == npv->pv_va)
				panic("pmap_enter_pv: already in pv_tab");
#endif
		npv = pmap_alloc_pv();
		npv->pv_va = va;
		npv->pv_pmap = pmap;
		npv->pv_next = pv->pv_next;
		pv->pv_next = npv;
#ifdef DEBUG
		if (!npv->pv_next)
			enter_stats.secondpv++;
#endif
	}
	splx(s);
}

__inline void
pmap_remove_pv(pmap, va, pind)
	register pmap_t pmap;
	vm_offset_t va;
	u_int pind;
{	
	register struct pv_entry *pv, *npv;
	int s;

	/*
	 * Remove from the PV table (raise IPL since we
	 * may be called at interrupt time).
	 */
	pv = &pv_table[pind];
	s = splimp();

	/*
	 * If it is the first entry on the list, it is actually
	 * in the header and we must copy the following entry up
	 * to the header.  Otherwise we must search the list for
	 * the entry.  In either case we free the now unused entry.
	 */
	if (pmap == pv->pv_pmap && va == pv->pv_va) {
		npv = pv->pv_next;
		if (npv) {
			*pv = *npv;
			pmap_free_pv(npv);
		} else
			pv->pv_pmap = NULL;
	} else {
		for (npv = pv->pv_next; npv; pv = npv, npv = npv->pv_next) {
			if (pmap == npv->pv_pmap && va == npv->pv_va)
				break;
		}
		if (npv) {
			pv->pv_next = npv->pv_next;
			pmap_free_pv(npv);
		}
	}
	splx(s);
}

/*
 *	Used to map a range of physical addresses into kernel
 *	virtual address space.
 *
 *	For now, VM is already on, we only need to map the
 *	specified memory.
 */
vm_offset_t
pmap_map(va, spa, epa, prot)
	vm_offset_t va, spa, epa;
	int prot;
{

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_map(%x, %x, %x, %x)\n", va, spa, epa, prot);
#endif

	while (spa < epa) {
		pmap_enter(pmap_kernel(), va, spa, prot, FALSE);
		va += NBPG;
		spa += NBPG;
	}
	return va;
}

/*
 *	Create and return a physical map.
 *
 *	If the size specified for the map
 *	is zero, the map is an actual physical
 *	map, and may be referenced by the
 *	hardware.
 *
 *	If the size specified is non-zero,
 *	the map will be used in software only, and
 *	is bounded by that size.
 *
 * [ just allocate a ptd and mark it uninitialize -- should we track
 *   with a table which process has which ptd? -wfj ]
 */
pmap_t
pmap_create(size)
	vm_size_t size;
{
	register pmap_t pmap;

#ifdef DEBUG
	if (pmapdebug & (PDB_FOLLOW|PDB_CREATE))
		printf("pmap_create(%x)\n", size);
#endif

	/*
	 * Software use map does not need a pmap
	 */
	if (size)
		return NULL;

	pmap = (pmap_t) malloc(sizeof *pmap, M_VMPMAP, M_WAITOK);
	bzero(pmap, sizeof(*pmap));
	pmap_pinit(pmap);
	return pmap;
}

/*
 * Initialize a preallocated and zeroed pmap structure,
 * such as one in a vmspace structure.
 */
void
pmap_pinit(pmap)
	register struct pmap *pmap;
{

#ifdef DEBUG
	if (pmapdebug & (PDB_FOLLOW|PDB_CREATE))
		printf("pmap_pinit(%x)\n", pmap);
#endif

	/*
	 * No need to allocate page table space yet but we do need a
	 * valid page directory table.
	 */
	pmap->pm_pdir = (pd_entry_t *) kmem_alloc(kernel_map, NBPG);

	/* wire in kernel global address entries */
	bcopy(&PTD[KPTDI], &pmap->pm_pdir[KPTDI], MAXKPDE *
	    sizeof(pd_entry_t));

	/* install self-referential address mapping entry */
	pmap->pm_pdir[PTDPTDI] = pmap_extract(pmap_kernel(),
	    (vm_offset_t)pmap->pm_pdir) | PG_V | PG_KW;

	pmap->pm_count = 1;
	simple_lock_init(&pmap->pm_lock);
}

/*
 *	Retire the given physical map from service.
 *	Should only be called if the map contains
 *	no valid mappings.
 */
void
pmap_destroy(pmap)
	register pmap_t pmap;
{
	int count;

	if (pmap == NULL)
		return;

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_destroy(%x)\n", pmap);
#endif

	simple_lock(&pmap->pm_lock);
	count = --pmap->pm_count;
	simple_unlock(&pmap->pm_lock);
	if (count == 0) {
		pmap_release(pmap);
		free((caddr_t)pmap, M_VMPMAP);
	}
}

/*
 * Release any resources held by the given physical map.
 * Called when a pmap initialized by pmap_pinit is being released.
 * Should only be called if the map contains no valid mappings.
 */
void
pmap_release(pmap)
	register struct pmap *pmap;
{

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_release(%x)\n", pmap);
#endif

#ifdef DIAGNOSTICx
	/* sometimes 1, sometimes 0; could rearrange pmap_destroy */
	if (pmap->pm_count != 1)
		panic("pmap_release count");
#endif

	kmem_free(kernel_map, (vm_offset_t)pmap->pm_pdir, NBPG);
}

/*
 *	Add a reference to the specified pmap.
 */
void
pmap_reference(pmap)
	pmap_t pmap;
{

	if (pmap == NULL)
		return;

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_reference(%x)", pmap);
#endif

	simple_lock(&pmap->pm_lock);
	pmap->pm_count++;
	simple_unlock(&pmap->pm_lock);
}

void
pmap_activate(pmap, pcb)
	pmap_t pmap;
	struct pcb *pcb;
{

	if (pmap /*&& pmap->pm_pdchanged */) {
		pcb->pcb_cr3 =
		    pmap_extract(pmap_kernel(), (vm_offset_t)pmap->pm_pdir);
		if (pmap == &curproc->p_vmspace->vm_pmap)
			lcr3(pcb->pcb_cr3);
		pmap->pm_pdchanged = FALSE;
	}
}

void
pmap_deactivate(pmap, pcb)
	pmap_t pmap;
	struct pcb *pcb;
{
}

/*
 *	Remove the given range of addresses from the specified map.
 *
 *	It is assumed that the start and end are properly
 *	rounded to the page size.
 */
void
pmap_remove(pmap, sva, eva)
	struct pmap *pmap;
	register vm_offset_t sva, eva;
{
	register pt_entry_t *pte;
	vm_offset_t pa;
	u_int pind;
	int flush = 0;

	sva &= PG_FRAME;
	eva &= PG_FRAME;

	/*
	 * We need to acquire a pointer to a page table page before entering
	 * the following loop.
	 */
	while (sva < eva) {
		pte = pmap_pte(pmap, sva);
		if (pte)
			break;
		sva = (sva & PD_MASK) + NBPD;
	}

	while (sva < eva) {
		/* only check once in a while */
		if ((sva & PT_MASK) == 0) {
			if (!pmap_pde_v(pmap_pde(pmap, sva))) {
				/* We can race ahead here, to the next pde. */
				sva += NBPD;
				pte += i386_btop(NBPD);
				continue;
			}
		}

		pte = pmap_pte(pmap, sva);
		if (pte == NULL) {
			/* We can race ahead here, to the next pde. */
			sva = (sva & PD_MASK) + NBPD;
			continue;
		}

		if (!pmap_pte_v(pte)) {
#ifdef __GNUC__
			/*
			 * Scan ahead in a tight loop for the next used PTE in
			 * this page.  We don't scan the whole region here
			 * because we don't want to zero-fill unused page table
			 * pages.
			 */
			int n, m;

			n = min(eva - sva, NBPD - (sva & PT_MASK)) >> PGSHIFT;
			__asm __volatile(
			    "cld\n\trepe\n\tscasl\n\tje 1f\n\tincl %1\n\t1:"
			    : "=D" (pte), "=c" (m)
			    : "0" (pte), "1" (n), "a" (0));
			sva += (n - m) << PGSHIFT;
			if (!m)
				continue;
			/* Overshot. */
			--pte;
#else
			goto next;
#endif
		}

		flush = 1;

		/*
		 * Update statistics
		 */
		if (pmap_pte_w(pte))
			pmap->pm_stats.wired_count--;
		pmap->pm_stats.resident_count--;

		pa = pmap_pte_pa(pte);

		/*
		 * Invalidate the PTEs.
		 * XXX: should cluster them up and invalidate as many
		 * as possible at once.
		 */
#ifdef DEBUG
		if (pmapdebug & PDB_REMOVE)
			printf("remove: inv pte at %x(%x) ", pte, *pte);
#endif

#ifdef needednotdone
reduce wiring count on page table pages as references drop
#endif

		if ((pind = pmap_page_index(pa)) != -1) {
			pmap_attributes[pind] |= *pte & (PG_M | PG_U);
			pmap_remove_pv(pmap, sva, pind);
		}

		*pte = 0;

#ifndef __GNUC__
	next:
#endif
		sva += NBPG;
		pte++;
	}

	if (flush)
		pmap_update();
}

/*
 *	Routine:	pmap_remove_all
 *	Function:
 *		Removes this physical page from
 *		all physical maps in which it resides.
 *		Reflects back modify bits to the pager.
 */
void
pmap_remove_all(pa)
	vm_offset_t pa;
{
	struct pv_entry *ph, *pv, *npv;
	register pmap_t pmap;
	register pt_entry_t *pte;
	u_int pind;
	int s;

#ifdef DEBUG
	if (pmapdebug & (PDB_FOLLOW|PDB_REMOVE|PDB_PROTECT))
		printf("pmap_remove_all(%x)", pa);
	/*pmap_pvdump(pa);*/
#endif

	if ((pind = pmap_page_index(pa)) == -1)
		return;

	pv = ph = &pv_table[pind];
	s = splimp();

	if (ph->pv_pmap == NULL) {
		splx(s);
		return;
	}

	while (pv) {
		pmap = pv->pv_pmap;
		pte = pmap_pte(pmap, pv->pv_va);

#ifdef DEBUG
		if (!pte || !pmap_pte_v(pte) || pmap_pte_pa(pte) != pa)
			panic("pmap_remove_all: bad mapping");
#endif

		/*
		 * Update statistics
		 */
		if (pmap_pte_w(pte))
			pmap->pm_stats.wired_count--;
		pmap->pm_stats.resident_count--;

		/*
		 * Invalidate the PTEs.
		 * XXX: should cluster them up and invalidate as many
		 * as possible at once.
		 */
#ifdef DEBUG
		if (pmapdebug & PDB_REMOVE)
			printf("remove: inv pte at %x(%x) ", pte, *pte);
#endif

#ifdef needednotdone
reduce wiring count on page table pages as references drop
#endif

		/*
		 * Update saved attributes for managed page
		 */
		pmap_attributes[pind] |= *pte & (PG_M | PG_U);

		*pte = 0;

		npv = pv->pv_next;
		if (pv == ph)
			ph->pv_pmap = NULL;
		else
			pmap_free_pv(pv);
		pv = npv;
	}
	splx(s);

	pmap_update();
}

/*
 *	Set the physical protection on the
 *	specified range of this map as requested.
 */
void
pmap_protect(pmap, sva, eva, prot)
	register pmap_t pmap;
	vm_offset_t sva, eva;
	vm_prot_t prot;
{
	register pt_entry_t *pte;
	register int i386prot;
	int flush = 0;

#ifdef DEBUG
	if (pmapdebug & (PDB_FOLLOW|PDB_PROTECT))
		printf("pmap_protect(%x, %x, %x, %x)", pmap, sva, eva, prot);
#endif

	if ((prot & VM_PROT_READ) == VM_PROT_NONE) {
		pmap_remove(pmap, sva, eva);
		return;
	}

	if (prot & VM_PROT_WRITE)
		return;

	sva &= PG_FRAME;
	eva &= PG_FRAME;

	/*
	 * We need to acquire a pointer to a page table page before entering
	 * the following loop.
	 */
	while (sva < eva) {
		pte = pmap_pte(pmap, sva);
		if (pte)
			break;
		sva = (sva & PD_MASK) + NBPD;
	}

	while (sva < eva) {
		/* only check once in a while */
		if ((sva & PT_MASK) == 0) {
			if (!pmap_pde_v(pmap_pde(pmap, sva))) {
				/* We can race ahead here, to the next pde. */
				sva += NBPD;
				pte += i386_btop(NBPD);
				continue;
			}
		}

		if (!pmap_pte_v(pte)) {
#ifdef __GNUC__
			/*
			 * Scan ahead in a tight loop for the next used PTE in
			 * this page.  We don't scan the whole region here
			 * because we don't want to zero-fill unused page table
			 * pages.
			 */
			int n, m;

			n = min(eva - sva, NBPD - (sva & PT_MASK)) >> PGSHIFT;
			__asm __volatile(
			    "cld\n\trepe\n\tscasl\n\tje 1f\n\tincl %1\n\t1:"
			    : "=D" (pte), "=c" (m)
			    : "0" (pte), "1" (n), "a" (0));
			sva += (n - m) << PGSHIFT;
			if (!m)
				continue;
			/* Overshot. */
			--pte;
#else
			goto next;
#endif
		}

		flush = 1;

		i386prot = protection_codes[prot];
		if (sva < VM_MAXUSER_ADDRESS)	/* see also pmap_enter() */
			i386prot |= PG_u;
		else if (sva < VM_MAX_ADDRESS)
			i386prot |= PG_u | PG_RW;
		pmap_pte_set_prot(pte, i386prot);

#ifndef __GNUC__
	next:
#endif
		sva += NBPG;
		pte++;
	}

	if (flush)
		pmap_update();
}

/*
 *	Insert the given physical page (p) at
 *	the specified virtual address (v) in the
 *	target physical map with the protection requested.
 *
 *	If specified, the page will be wired down, meaning
 *	that the related pte can not be reclaimed.
 *
 *	NB:  This is the only routine which MAY NOT lazy-evaluate
 *	or lose information.  That is, this routine must actually
 *	insert this page into the given map NOW.
 */
void
pmap_enter(pmap, va, pa, prot, wired)
	register pmap_t pmap;
	vm_offset_t va;
	register vm_offset_t pa;
	vm_prot_t prot;
	boolean_t wired;
{
	register pt_entry_t *pte;
	register pt_entry_t npte;
	u_int pind;
	int flush = 0;
	boolean_t cacheable;

#ifdef DEBUG
	if (pmapdebug & (PDB_FOLLOW|PDB_ENTER))
		printf("pmap_enter(%x, %x, %x, %x, %x)",
		       pmap, va, pa, prot, wired);
#endif

	if (pmap == NULL)
		return;

	if (va >= VM_MAX_KERNEL_ADDRESS)
		panic("pmap_enter: too big");
	/* also, should not muck with PTD va! */

#ifdef DEBUG
	if (pmap == pmap_kernel())
		enter_stats.kernel++;
	else
		enter_stats.user++;
#endif

	pte = pmap_pte(pmap, va);
	if (!pte) {
		/*
		 * Page Directory table entry not valid, we need a new PT page
		 *
		 * we want to vm_fault in a new zero-filled PT page for our
		 * use.   in order to do this, we want to call vm_fault()
		 * with the VA of where we want to put the PTE.   but in
		 * order to call vm_fault() we need to know which vm_map
		 * we are faulting in.    in the m68k pmap's this is easy
		 * since all PT pages live in one global vm_map ("pt_map")
		 * and we have a lot of virtual space we can use for the
		 * pt_map (since the kernel doesn't have to share its 4GB
		 * address space with processes).    but in the i386 port
		 * the kernel must live in the top part of the virtual 
		 * address space and PT pages live in their process' vm_map
		 * rather than a global one.    the problem is that we have
		 * no way of knowing which vm_map is the correct one to 
		 * fault on.
		 * 
		 * XXX: see NetBSD PR#1834 and Mycroft's posting to 
		 *	tech-kern on 7 Jan 1996.
		 *
		 * rather than always calling panic, we try and make an 
		 * educated guess as to which vm_map to use by using curproc.
		 * this is a workaround and may not fully solve the problem?
	 	 */
		struct vm_map *vmap;
		int rv;
		vm_offset_t v;

		if (curproc == NULL || curproc->p_vmspace == NULL ||
		    pmap != &curproc->p_vmspace->vm_pmap)
			panic("ptdi %x", pmap->pm_pdir[PTDPTDI]);

		/* our guess about the vm_map was good!  fault it in.  */

		vmap = &curproc->p_vmspace->vm_map;
		v = trunc_page(vtopte(va));
		printf("faulting in a pt page map %x va %x\n", vmap, v);
		rv = vm_fault(vmap, v, VM_PROT_READ|VM_PROT_WRITE, FALSE);
		if (rv != KERN_SUCCESS)
			panic("ptdi2 %x", pmap->pm_pdir[PTDPTDI]);
		vm_map_pageable(vmap, v, round_page(v+1), FALSE);
		pte = pmap_pte(pmap, va);
		if (!pte) 
			panic("ptdi3 %x", pmap->pm_pdir[PTDPTDI]);
	}
#ifdef DEBUG
	if (pmapdebug & PDB_ENTER)
		printf("enter: pte %x, *pte %x ", pte, *pte);
#endif

	if (pmap_pte_v(pte)) {
		register vm_offset_t opa;

		/*
		 * Check for wiring change and adjust statistics.
		 */
		if ((wired && !pmap_pte_w(pte)) ||
		    (!wired && pmap_pte_w(pte))) {
			/*
			 * We don't worry about wiring PT pages as they remain
			 * resident as long as there are valid mappings in them.
			 * Hence, if a user page is wired, the PT page will be also.
			 */
#ifdef DEBUG
			if (pmapdebug & PDB_ENTER)
				printf("enter: wiring change -> %x ", wired);
#endif
			if (wired)
				pmap->pm_stats.wired_count++;
			else
				pmap->pm_stats.wired_count--;
#ifdef DEBUG
			enter_stats.wchange++;
#endif
		}

		flush = 1;
		opa = pmap_pte_pa(pte);

		/*
		 * Mapping has not changed, must be protection or wiring change.
		 */
		if (opa == pa) {
#ifdef DEBUG
			enter_stats.pwchange++;
#endif
			goto validate;
		}
		
		/*
		 * Mapping has changed, invalidate old range and fall through to
		 * handle validating new mapping.
		 */
#ifdef DEBUG
		if (pmapdebug & PDB_ENTER)
			printf("enter: removing old mapping %x pa %x ", va, opa);
#endif
		if ((pind = pmap_page_index(opa)) != -1) {
			pmap_attributes[pind] |= *pte & (PG_M | PG_U);
			pmap_remove_pv(pmap, va, pind);
		}
#ifdef DEBUG
		enter_stats.mchange++;
#endif
	} else {
		/*
		 * Increment counters
		 */
		pmap->pm_stats.resident_count++;
		if (wired)
			pmap->pm_stats.wired_count++;
	}

	/*
	 * Enter on the PV list if part of our managed memory
	 */
	if ((pind = pmap_page_index(pa)) != -1) {
#ifdef DEBUG
		enter_stats.managed++;
#endif
		pmap_enter_pv(pmap, va, pind);
		cacheable = TRUE;
	} else if (pmap_initialized) {
#ifdef DEBUG
		enter_stats.unmanaged++;
#endif
		/*
		 * Assumption: if it is not part of our managed memory
		 * then it must be device memory which may be volatile.
		 */
		cacheable = FALSE;
	}

validate:
	/*
	 * Now validate mapping with desired protection/wiring.
	 * Assume uniform modified and referenced status for all
	 * I386 pages in a MACH page.
	 */
	npte = (pa & PG_FRAME) | protection_codes[prot] | PG_V;
	if (wired)
		npte |= PG_W;

	if (va < VM_MAXUSER_ADDRESS)	/* i.e. below USRSTACK */
		npte |= PG_u;
	else if (va < VM_MAX_ADDRESS)
		/*
		 * Page tables need to be user RW, for some reason, and the
		 * user area must be writable too.  Anything above
		 * VM_MAXUSER_ADDRESS is protected from user access by
		 * the user data and code segment descriptors, so this is OK.
		 */
		npte |= PG_u | PG_RW;

#ifdef DEBUG
	if (pmapdebug & PDB_ENTER)
		printf("enter: new pte value %x ", npte);
#endif

	*pte = npte;
	if (flush)
		pmap_update();
}

/*
 *      pmap_page_protect:
 *
 *      Lower the permission for all mappings to a given page.
 */
void
pmap_page_protect(phys, prot)
	vm_offset_t     phys;
	vm_prot_t       prot;
{

	switch (prot) {
	case VM_PROT_READ:
	case VM_PROT_READ|VM_PROT_EXECUTE:
		pmap_copy_on_write(phys);
		break;
	case VM_PROT_ALL:
		break;
	default:
		pmap_remove_all(phys);
		break;
	}
}

/*
 *	Routine:	pmap_change_wiring
 *	Function:	Change the wiring attribute for a map/virtual-address
 *			pair.
 *	In/out conditions:
 *			The mapping must already exist in the pmap.
 */
void
pmap_change_wiring(pmap, va, wired)
	register pmap_t pmap;
	vm_offset_t va;
	boolean_t wired;
{
	register pt_entry_t *pte;

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_change_wiring(%x, %x, %x)", pmap, va, wired);
#endif

	pte = pmap_pte(pmap, va);
	if (!pte)
		return;

#ifdef DEBUG
	/*
	 * Page not valid.  Should this ever happen?
	 * Just continue and change wiring anyway.
	 */
	if (!pmap_pte_v(pte)) {
		if (pmapdebug & PDB_PARANOIA)
			printf("pmap_change_wiring: invalid PTE for %x ", va);
	}
#endif

	if ((wired && !pmap_pte_w(pte)) || (!wired && pmap_pte_w(pte))) {
		if (wired)
			pmap->pm_stats.wired_count++;
		else
			pmap->pm_stats.wired_count--;
		pmap_pte_set_w(pte, wired);
	}
}

/*
 *	Routine:	pmap_pte
 *	Function:
 *		Extract the page table entry associated
 *		with the given map/virtual_address pair.
 */
pt_entry_t *
pmap_pte(pmap, va)
	register pmap_t pmap;
	vm_offset_t va;
{
	pt_entry_t *ptp;

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_pte(%x, %x) ->\n", pmap, va);
#endif

	if (!pmap || !pmap_pde_v(pmap_pde(pmap, va)))
		return NULL;

	if ((pmap->pm_pdir[PTDPTDI] & PG_FRAME) == (PTDpde & PG_FRAME) ||
	    pmap == pmap_kernel())
		/* current address space or kernel */
		ptp = PTmap;
	else {
		/* alternate address space */
		if ((pmap->pm_pdir[PTDPTDI] & PG_FRAME) != (APTDpde & PG_FRAME)) {
			APTDpde = pmap->pm_pdir[PTDPTDI];
			pmap_update();
		}
		ptp = APTmap;
	}

	return ptp + i386_btop(va);
}

/*
 *	Routine:	pmap_extract
 *	Function:
 *		Extract the physical page address associated
 *		with the given map/virtual_address pair.
 */
vm_offset_t
pmap_extract(pmap, va)
	register pmap_t pmap;
	vm_offset_t va;
{
	register pt_entry_t *pte;
	register vm_offset_t pa;

#ifdef DEBUGx
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_extract(%x, %x) -> ", pmap, va);
#endif

	pte = pmap_pte(pmap, va);
	if (!pte)
		return NULL;
	if (!pmap_pte_v(pte))
		return NULL;

	pa = pmap_pte_pa(pte);
#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("%x\n", pa);
#endif
	return pa | (va & ~PG_FRAME);
}

/*
 *	Copy the range specified by src_addr/len
 *	from the source map to the range dst_addr/len
 *	in the destination map.
 *
 *	This routine is only advisory and need not do anything.
 */
void
pmap_copy(dst_pmap, src_pmap, dst_addr, len, src_addr)
	pmap_t dst_pmap, src_pmap;
	vm_offset_t dst_addr, src_addr;
	vm_size_t len;
{

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_copy(%x, %x, %x, %x, %x)",
		       dst_pmap, src_pmap, dst_addr, len, src_addr);
#endif
}

/*
 *	Routine:	pmap_collect
 *	Function:
 *		Garbage collects the physical map system for
 *		pages which are no longer used.
 *		Success need not be guaranteed -- that is, there
 *		may well be pages which are not referenced, but
 *		others may be collected.
 *	Usage:
 *		Called by the pageout daemon when pages are scarce.
 * [ needs to be written -wfj ]  XXXX
 */
void
pmap_collect(pmap)
	pmap_t pmap;
{
#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_collect(%x) ", pmap);
#endif

	if (pmap != pmap_kernel())
		return;

}

#if DEBUG
void
pmap_dump_pvlist(phys, m)
	vm_offset_t phys;
	char *m;
{
	register struct pv_entry *pv;

	if (!(pmapdebug & PDB_PARANOIA))
		return;

	if (!pmap_initialized)
		return;
	printf("%s %08x:", m, phys);
	pv = &pv_table[pmap_page_index(phys)];
	if (pv->pv_pmap == NULL) {
		printf(" no mappings\n");
		return;
	}
	for (; pv; pv = pv->pv_next)
		printf(" pmap %08x va %08x", pv->pv_pmap, pv->pv_va);
	printf("\n");
}
#else
#define	pmap_dump_pvlist(a,b)
#endif

/*
 *	pmap_zero_page zeros the specified by mapping it into
 *	virtual memory and using bzero to clear its contents.
 */
void
pmap_zero_page(phys)
	register vm_offset_t phys;
{

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_zero_page(%x)", phys);
#endif

	pmap_dump_pvlist(phys, "pmap_zero_page: phys");
	*CMAP2 = (phys & PG_FRAME) | PG_V | PG_KW /*| PG_N*/;
	pmap_update();
	bzero(CADDR2, NBPG);
}

/*
 *	pmap_copy_page copies the specified page by mapping
 *	it into virtual memory and using bcopy to copy its
 *	contents.
 */
void
pmap_copy_page(src, dst)
	register vm_offset_t src, dst;
{

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_copy_page(%x, %x)", src, dst);
#endif

	pmap_dump_pvlist(src, "pmap_copy_page: src");
	pmap_dump_pvlist(dst, "pmap_copy_page: dst");
	*CMAP1 = (src & PG_FRAME) | PG_V | PG_KR;
	*CMAP2 = (dst & PG_FRAME) | PG_V | PG_KW /*| PG_N*/;
	pmap_update();
	bcopy(CADDR1, CADDR2, NBPG);
}

/*
 *	Routine:	pmap_pageable
 *	Function:
 *		Make the specified pages (by pmap, offset)
 *		pageable (or not) as requested.
 *
 *		A page which is not pageable may not take
 *		a fault; therefore, its page table entry
 *		must remain valid for the duration.
 *
 *		This routine is merely advisory; pmap_enter
 *		will specify that these pages are to be wired
 *		down (or not) as appropriate.
 */

void
pmap_pageable(pmap, sva, eva, pageable)
	pmap_t pmap;
	vm_offset_t sva, eva;
	boolean_t pageable;
{

#ifdef DEBUG
	if (pmapdebug & PDB_FOLLOW)
		printf("pmap_pageable(%x, %x, %x, %x)",
		       pmap, sva, eva, pageable);
#endif

	/*
	 * If we are making a PT page pageable then all valid
	 * mappings must be gone from that page.  Hence it should
	 * be all zeros and there is no need to clean it.
	 * Assumption:
	 *	- PT pages have only one pv_table entry
	 *	- PT pages are the only single-page allocations
	 *	  between the user stack and kernel va's 
	 * See also pmap_enter & pmap_protect for rehashes of this...
	 */

	if (pageable &&
	    pmap == pmap_kernel() &&
	    sva >= VM_MAXUSER_ADDRESS && eva <= VM_MAX_ADDRESS &&
	    eva - sva == NBPG) {
		register vm_offset_t pa;
		register pt_entry_t *pte;
#ifdef DIAGNOSTIC
		u_int pind;
		register struct pv_entry *pv;
#endif

#ifdef DEBUG
		if ((pmapdebug & (PDB_FOLLOW|PDB_PTPAGE)) == PDB_PTPAGE)
			printf("pmap_pageable(%x, %x, %x, %x)",
			       pmap, sva, eva, pageable);
#endif

		pte = pmap_pte(pmap, sva);
		if (!pte)
			return;
		if (!pmap_pte_v(pte))
			return;

		pa = pmap_pte_pa(pte);

#ifdef DIAGNOSTIC
		if ((*pte & (PG_u | PG_RW)) != (PG_u | PG_RW))
			printf("pmap_pageable: unexpected pte=%x va %x\n",
				*pte, sva);

		if ((pind = pmap_page_index(pa)) == -1) {
			printf("pmap_pageable: invalid pa %x va %x\n",
				pa, sva);
			return;
		}

		pv = &pv_table[pind];
		if (pv->pv_va != sva || pv->pv_next) {
			printf("pmap_pageable: bad PT page va %x next %x\n",
			       pv->pv_va, pv->pv_next);
			return;
		}
#endif

		/*
		 * Mark it unmodified to avoid pageout
		 */
		pmap_clear_modify(pa);

#ifdef needsomethinglikethis
		if (pmapdebug & PDB_PTPAGE)
			printf("pmap_pageable: PT page %x(%x) unmodified\n",
			       sva, *pmap_pte(pmap, sva));
		if (pmapdebug & PDB_WIRING)
			pmap_check_wiring("pageable", sva);
#endif
	}
}

/*
 * Miscellaneous support routines follow
 */
void
i386_protection_init()
{

	protection_codes[VM_PROT_NONE | VM_PROT_NONE | VM_PROT_NONE] = 0;
	protection_codes[VM_PROT_NONE | VM_PROT_NONE | VM_PROT_EXECUTE] =
	protection_codes[VM_PROT_NONE | VM_PROT_READ | VM_PROT_NONE] =
	protection_codes[VM_PROT_NONE | VM_PROT_READ | VM_PROT_EXECUTE] = PG_RO;
	protection_codes[VM_PROT_WRITE | VM_PROT_NONE | VM_PROT_NONE] =
	protection_codes[VM_PROT_WRITE | VM_PROT_NONE | VM_PROT_EXECUTE] =
	protection_codes[VM_PROT_WRITE | VM_PROT_READ | VM_PROT_NONE] =
	protection_codes[VM_PROT_WRITE | VM_PROT_READ | VM_PROT_EXECUTE] = PG_RW;
}

boolean_t
pmap_testbit(pa, setbits)
	register vm_offset_t pa;
	int setbits;
{
	register struct pv_entry *pv;
	register pt_entry_t *pte;
	u_int pind;
	int s;

	if ((pind = pmap_page_index(pa)) == -1)
		return FALSE;

	pv = &pv_table[pind];
	s = splimp();

	/*
	 * Check saved info first
	 */
	if (pmap_attributes[pind] & setbits) {
		splx(s);
		return TRUE;
	}

	/*
	 * Not found, check current mappings returning
	 * immediately if found.
	 */
	if (pv->pv_pmap != NULL) {
		for (; pv; pv = pv->pv_next) {
			pte = pmap_pte(pv->pv_pmap, pv->pv_va);
			if (*pte & setbits) {
				splx(s);
				return TRUE;
			}
		}
	}
	splx(s);
	return FALSE;
}

/*
 * Modify pte bits for all ptes corresponding to the given physical address.
 * We use `maskbits' rather than `clearbits' because we're always passing
 * constants and the latter would require an extra inversion at run-time.
 */
void
pmap_changebit(pa, setbits, maskbits)
	register vm_offset_t pa;
	int setbits, maskbits;
{
	register struct pv_entry *pv;
	register pt_entry_t *pte;
	vm_offset_t va;
	u_int pind;
	int s;

#ifdef DEBUG
	if (pmapdebug & PDB_BITS)
		printf("pmap_changebit(%x, %x, %x)",
		       pa, setbits, ~maskbits);
#endif

	if ((pind = pmap_page_index(pa)) == -1)
		return;

	pv = &pv_table[pind];
	s = splimp();

	/*
	 * Clear saved attributes (modify, reference)
	 */
	if (~maskbits)
		pmap_attributes[pind] &= maskbits;

	/*
	 * Loop over all current mappings setting/clearing as appropos
	 * If setting RO do we need to clear the VAC?
	 */
	if (pv->pv_pmap != NULL) {
		for (; pv; pv = pv->pv_next) {
			va = pv->pv_va;

			/*
			 * XXX don't write protect pager mappings
			 */
			if ((PG_RO && setbits == PG_RO) ||
			    (PG_RW && maskbits == ~PG_RW)) {
				extern vm_offset_t pager_sva, pager_eva;

				if (va >= pager_sva && va < pager_eva)
					continue;
			}

			pte = pmap_pte(pv->pv_pmap, va);
			*pte = (*pte & maskbits) | setbits;
		}
		pmap_update();
	}
	splx(s);
}

void
pmap_prefault(map, v, l)
	vm_map_t map;
	vm_offset_t v;
	vm_size_t l;
{
	vm_offset_t pv, pv2;

	for (pv = v; pv < v + l ; pv += ~PD_MASK + 1) {
		if (!pmap_pde_v(pmap_pde(map->pmap, pv))) {
			pv2 = trunc_page(vtopte(pv));
			vm_fault(map, pv2, VM_PROT_READ, FALSE);
		}
		pv &= PD_MASK;
	}
}

#ifdef DEBUG
void
pmap_pvdump(pa)
	vm_offset_t pa;
{
	register struct pv_entry *pv;

	printf("pa %x", pa);
	for (pv = &pv_table[pmap_page_index(pa)]; pv; pv = pv->pv_next) {
		printf(" -> pmap %x, va %x", pv->pv_pmap, pv->pv_va);
		pads(pv->pv_pmap);
	}
	printf(" ");
}

#ifdef notyet
void
pmap_check_wiring(str, va)
	char *str;
	vm_offset_t va;
{
	vm_map_entry_t entry;
	register int count, *pte;

	va = trunc_page(va);
	if (!pmap_pde_v(pmap_pde(pmap_kernel(), va)) ||
	    !pmap_pte_v(pmap_pte(pmap_kernel(), va)))
		return;

	if (!vm_map_lookup_entry(pt_map, va, &entry)) {
		printf("wired_check: entry for %x not found\n", va);
		return;
	}
	count = 0;
	for (pte = (int *)va; pte < (int *)(va + NBPG); pte++)
		if (*pte)
			count++;
	if (entry->wired_count != count)
		printf("*%s*: %x: w%d/a%d\n",
		       str, va, entry->wired_count, count);
}
#endif

/* print address space of pmap*/
void
pads(pm)
	pmap_t pm;
{
	unsigned va, i, j;
	register pt_entry_t *pte;

	if (pm == pmap_kernel())
		return;
	for (i = 0; i < 1024; i++) 
		if (pmap_pde_v(&pm->pm_pdir[i]))
			for (j = 0; j < 1024 ; j++) {
				va = (i << PDSHIFT) | (j << PGSHIFT);
				if (pm == pmap_kernel() &&
				    va < VM_MIN_KERNEL_ADDRESS)
					continue;
				if (pm != pmap_kernel() &&
				    va > VM_MAX_ADDRESS)
					continue;
				pte = pmap_pte(pm, va);
				if (pmap_pte_v(pte)) 
					printf("%x:%x ", va, *pte); 
			}
}
#endif