aboutsummaryrefslogtreecommitdiff
path: root/sys/kern/kern_malloc.c
blob: c39f14359bf9988d95a39544479594ffc023e3f4 (plain) (blame)
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
/*-
 * Copyright (c) 1987, 1991, 1993
 *	The Regents of the University of California.
 * Copyright (c) 2005-2009 Robert N. M. Watson
 * All rights reserved.
 *
 * 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.
 * 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.
 *
 *	@(#)kern_malloc.c	8.3 (Berkeley) 1/4/94
 */

/*
 * Kernel malloc(9) implementation -- general purpose kernel memory allocator
 * based on memory types.  Back end is implemented using the UMA(9) zone
 * allocator.  A set of fixed-size buckets are used for smaller allocations,
 * and a special UMA allocation interface is used for larger allocations.
 * Callers declare memory types, and statistics are maintained independently
 * for each memory type.  Statistics are maintained per-CPU for performance
 * reasons.  See malloc(9) and comments in malloc.h for a detailed
 * description.
 */

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

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

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kdb.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mutex.h>
#include <sys/vmmeter.h>
#include <sys/proc.h>
#include <sys/sbuf.h>
#include <sys/sysctl.h>
#include <sys/time.h>
#include <sys/vmem.h>

#include <vm/vm.h>
#include <vm/pmap.h>
#include <vm/vm_pageout.h>
#include <vm/vm_param.h>
#include <vm/vm_kern.h>
#include <vm/vm_extern.h>
#include <vm/vm_map.h>
#include <vm/vm_page.h>
#include <vm/uma.h>
#include <vm/uma_int.h>
#include <vm/uma_dbg.h>

#ifdef DEBUG_MEMGUARD
#include <vm/memguard.h>
#endif
#ifdef DEBUG_REDZONE
#include <vm/redzone.h>
#endif

#if defined(INVARIANTS) && defined(__i386__)
#include <machine/cpu.h>
#endif

#include <ddb/ddb.h>

#ifdef KDTRACE_HOOKS
#include <sys/dtrace_bsd.h>

dtrace_malloc_probe_func_t	dtrace_malloc_probe;
#endif

/*
 * When realloc() is called, if the new size is sufficiently smaller than
 * the old size, realloc() will allocate a new, smaller block to avoid
 * wasting memory. 'Sufficiently smaller' is defined as: newsize <=
 * oldsize / 2^n, where REALLOC_FRACTION defines the value of 'n'.
 */
#ifndef REALLOC_FRACTION
#define	REALLOC_FRACTION	1	/* new block if <= half the size */
#endif

/*
 * Centrally define some common malloc types.
 */
MALLOC_DEFINE(M_CACHE, "cache", "Various Dynamically allocated caches");
MALLOC_DEFINE(M_DEVBUF, "devbuf", "device driver memory");
MALLOC_DEFINE(M_TEMP, "temp", "misc temporary data buffers");

MALLOC_DEFINE(M_IP6OPT, "ip6opt", "IPv6 options");
MALLOC_DEFINE(M_IP6NDP, "ip6ndp", "IPv6 Neighbor Discovery");

static struct malloc_type *kmemstatistics;
static int kmemcount;

#define KMEM_ZSHIFT	4
#define KMEM_ZBASE	16
#define KMEM_ZMASK	(KMEM_ZBASE - 1)

#define KMEM_ZMAX	65536
#define KMEM_ZSIZE	(KMEM_ZMAX >> KMEM_ZSHIFT)
static uint8_t kmemsize[KMEM_ZSIZE + 1];

#ifndef MALLOC_DEBUG_MAXZONES
#define	MALLOC_DEBUG_MAXZONES	1
#endif
static int numzones = MALLOC_DEBUG_MAXZONES;

/*
 * Small malloc(9) memory allocations are allocated from a set of UMA buckets
 * of various sizes.
 *
 * XXX: The comment here used to read "These won't be powers of two for
 * long."  It's possible that a significant amount of wasted memory could be
 * recovered by tuning the sizes of these buckets.
 */
struct {
	int kz_size;
	char *kz_name;
	uma_zone_t kz_zone[MALLOC_DEBUG_MAXZONES];
} kmemzones[] = {
	{16, "16", },
	{32, "32", },
	{64, "64", },
	{128, "128", },
	{256, "256", },
	{512, "512", },
	{1024, "1024", },
	{2048, "2048", },
	{4096, "4096", },
	{8192, "8192", },
	{16384, "16384", },
	{32768, "32768", },
	{65536, "65536", },
	{0, NULL},
};

/*
 * Zone to allocate malloc type descriptions from.  For ABI reasons, memory
 * types are described by a data structure passed by the declaring code, but
 * the malloc(9) implementation has its own data structure describing the
 * type and statistics.  This permits the malloc(9)-internal data structures
 * to be modified without breaking binary-compiled kernel modules that
 * declare malloc types.
 */
static uma_zone_t mt_zone;

u_long vm_kmem_size;
SYSCTL_ULONG(_vm, OID_AUTO, kmem_size, CTLFLAG_RDTUN, &vm_kmem_size, 0,
    "Size of kernel memory");

static u_long kmem_zmax = KMEM_ZMAX;
SYSCTL_ULONG(_vm, OID_AUTO, kmem_zmax, CTLFLAG_RDTUN, &kmem_zmax, 0,
    "Maximum allocation size that malloc(9) would use UMA as backend");

static u_long vm_kmem_size_min;
SYSCTL_ULONG(_vm, OID_AUTO, kmem_size_min, CTLFLAG_RDTUN, &vm_kmem_size_min, 0,
    "Minimum size of kernel memory");

static u_long vm_kmem_size_max;
SYSCTL_ULONG(_vm, OID_AUTO, kmem_size_max, CTLFLAG_RDTUN, &vm_kmem_size_max, 0,
    "Maximum size of kernel memory");

static u_int vm_kmem_size_scale;
SYSCTL_UINT(_vm, OID_AUTO, kmem_size_scale, CTLFLAG_RDTUN, &vm_kmem_size_scale, 0,
    "Scale factor for kernel memory size");

static int sysctl_kmem_map_size(SYSCTL_HANDLER_ARGS);
SYSCTL_PROC(_vm, OID_AUTO, kmem_map_size,
    CTLFLAG_RD | CTLTYPE_ULONG | CTLFLAG_MPSAFE, NULL, 0,
    sysctl_kmem_map_size, "LU", "Current kmem allocation size");

static int sysctl_kmem_map_free(SYSCTL_HANDLER_ARGS);
SYSCTL_PROC(_vm, OID_AUTO, kmem_map_free,
    CTLFLAG_RD | CTLTYPE_ULONG | CTLFLAG_MPSAFE, NULL, 0,
    sysctl_kmem_map_free, "LU", "Free space in kmem");

/*
 * The malloc_mtx protects the kmemstatistics linked list.
 */
struct mtx malloc_mtx;

#ifdef MALLOC_PROFILE
uint64_t krequests[KMEM_ZSIZE + 1];

static int sysctl_kern_mprof(SYSCTL_HANDLER_ARGS);
#endif

static int sysctl_kern_malloc_stats(SYSCTL_HANDLER_ARGS);

/*
 * time_uptime of the last malloc(9) failure (induced or real).
 */
static time_t t_malloc_fail;

#if defined(MALLOC_MAKE_FAILURES) || (MALLOC_DEBUG_MAXZONES > 1)
static SYSCTL_NODE(_debug, OID_AUTO, malloc, CTLFLAG_RD, 0,
    "Kernel malloc debugging options");
#endif

/*
 * malloc(9) fault injection -- cause malloc failures every (n) mallocs when
 * the caller specifies M_NOWAIT.  If set to 0, no failures are caused.
 */
#ifdef MALLOC_MAKE_FAILURES
static int malloc_failure_rate;
static int malloc_nowait_count;
static int malloc_failure_count;
SYSCTL_INT(_debug_malloc, OID_AUTO, failure_rate, CTLFLAG_RWTUN,
    &malloc_failure_rate, 0, "Every (n) mallocs with M_NOWAIT will fail");
SYSCTL_INT(_debug_malloc, OID_AUTO, failure_count, CTLFLAG_RD,
    &malloc_failure_count, 0, "Number of imposed M_NOWAIT malloc failures");
#endif

static int
sysctl_kmem_map_size(SYSCTL_HANDLER_ARGS)
{
	u_long size;

	size = vmem_size(kmem_arena, VMEM_ALLOC);
	return (sysctl_handle_long(oidp, &size, 0, req));
}

static int
sysctl_kmem_map_free(SYSCTL_HANDLER_ARGS)
{
	u_long size;

	size = vmem_size(kmem_arena, VMEM_FREE);
	return (sysctl_handle_long(oidp, &size, 0, req));
}

/*
 * malloc(9) uma zone separation -- sub-page buffer overruns in one
 * malloc type will affect only a subset of other malloc types.
 */
#if MALLOC_DEBUG_MAXZONES > 1
static void
tunable_set_numzones(void)
{

	TUNABLE_INT_FETCH("debug.malloc.numzones",
	    &numzones);

	/* Sanity check the number of malloc uma zones. */
	if (numzones <= 0)
		numzones = 1;
	if (numzones > MALLOC_DEBUG_MAXZONES)
		numzones = MALLOC_DEBUG_MAXZONES;
}
SYSINIT(numzones, SI_SUB_TUNABLES, SI_ORDER_ANY, tunable_set_numzones, NULL);
SYSCTL_INT(_debug_malloc, OID_AUTO, numzones, CTLFLAG_RDTUN | CTLFLAG_NOFETCH,
    &numzones, 0, "Number of malloc uma subzones");

/*
 * Any number that changes regularly is an okay choice for the
 * offset.  Build numbers are pretty good of you have them.
 */
static u_int zone_offset = __FreeBSD_version;
TUNABLE_INT("debug.malloc.zone_offset", &zone_offset);
SYSCTL_UINT(_debug_malloc, OID_AUTO, zone_offset, CTLFLAG_RDTUN,
    &zone_offset, 0, "Separate malloc types by examining the "
    "Nth character in the malloc type short description.");

static u_int
mtp_get_subzone(const char *desc)
{
	size_t len;
	u_int val;

	if (desc == NULL || (len = strlen(desc)) == 0)
		return (0);
	val = desc[zone_offset % len];
	return (val % numzones);
}
#elif MALLOC_DEBUG_MAXZONES == 0
#error "MALLOC_DEBUG_MAXZONES must be positive."
#else
static inline u_int
mtp_get_subzone(const char *desc)
{

	return (0);
}
#endif /* MALLOC_DEBUG_MAXZONES > 1 */

int
malloc_last_fail(void)
{

	return (time_uptime - t_malloc_fail);
}

/*
 * An allocation has succeeded -- update malloc type statistics for the
 * amount of bucket size.  Occurs within a critical section so that the
 * thread isn't preempted and doesn't migrate while updating per-PCU
 * statistics.
 */
static void
malloc_type_zone_allocated(struct malloc_type *mtp, unsigned long size,
    int zindx)
{
	struct malloc_type_internal *mtip;
	struct malloc_type_stats *mtsp;

	critical_enter();
	mtip = mtp->ks_handle;
	mtsp = &mtip->mti_stats[curcpu];
	if (size > 0) {
		mtsp->mts_memalloced += size;
		mtsp->mts_numallocs++;
	}
	if (zindx != -1)
		mtsp->mts_size |= 1 << zindx;

#ifdef KDTRACE_HOOKS
	if (dtrace_malloc_probe != NULL) {
		uint32_t probe_id = mtip->mti_probes[DTMALLOC_PROBE_MALLOC];
		if (probe_id != 0)
			(dtrace_malloc_probe)(probe_id,
			    (uintptr_t) mtp, (uintptr_t) mtip,
			    (uintptr_t) mtsp, size, zindx);
	}
#endif

	critical_exit();
}

void
malloc_type_allocated(struct malloc_type *mtp, unsigned long size)
{

	if (size > 0)
		malloc_type_zone_allocated(mtp, size, -1);
}

/*
 * A free operation has occurred -- update malloc type statistics for the
 * amount of the bucket size.  Occurs within a critical section so that the
 * thread isn't preempted and doesn't migrate while updating per-CPU
 * statistics.
 */
void
malloc_type_freed(struct malloc_type *mtp, unsigned long size)
{
	struct malloc_type_internal *mtip;
	struct malloc_type_stats *mtsp;

	critical_enter();
	mtip = mtp->ks_handle;
	mtsp = &mtip->mti_stats[curcpu];
	mtsp->mts_memfreed += size;
	mtsp->mts_numfrees++;

#ifdef KDTRACE_HOOKS
	if (dtrace_malloc_probe != NULL) {
		uint32_t probe_id = mtip->mti_probes[DTMALLOC_PROBE_FREE];
		if (probe_id != 0)
			(dtrace_malloc_probe)(probe_id,
			    (uintptr_t) mtp, (uintptr_t) mtip,
			    (uintptr_t) mtsp, size, 0);
	}
#endif

	critical_exit();
}

/*
 *	contigmalloc:
 *
 *	Allocate a block of physically contiguous memory.
 *
 *	If M_NOWAIT is set, this routine will not block and return NULL if
 *	the allocation fails.
 */
void *
contigmalloc(unsigned long size, struct malloc_type *type, int flags,
    vm_paddr_t low, vm_paddr_t high, unsigned long alignment,
    vm_paddr_t boundary)
{
	void *ret;

	ret = (void *)kmem_alloc_contig(kernel_arena, size, flags, low, high,
	    alignment, boundary, VM_MEMATTR_DEFAULT);
	if (ret != NULL)
		malloc_type_allocated(type, round_page(size));
	return (ret);
}

/*
 *	contigfree:
 *
 *	Free a block of memory allocated by contigmalloc.
 *
 *	This routine may not block.
 */
void
contigfree(void *addr, unsigned long size, struct malloc_type *type)
{

	kmem_free(kernel_arena, (vm_offset_t)addr, size);
	malloc_type_freed(type, round_page(size));
}

/*
 *	malloc:
 *
 *	Allocate a block of memory.
 *
 *	If M_NOWAIT is set, this routine will not block and return NULL if
 *	the allocation fails.
 */
void *
malloc(unsigned long size, struct malloc_type *mtp, int flags)
{
	int indx;
	struct malloc_type_internal *mtip;
	caddr_t va;
	uma_zone_t zone;
#if defined(DIAGNOSTIC) || defined(DEBUG_REDZONE)
	unsigned long osize = size;
#endif

#ifdef INVARIANTS
	KASSERT(mtp->ks_magic == M_MAGIC, ("malloc: bad malloc type magic"));
	/*
	 * Check that exactly one of M_WAITOK or M_NOWAIT is specified.
	 */
	indx = flags & (M_WAITOK | M_NOWAIT);
	if (indx != M_NOWAIT && indx != M_WAITOK) {
		static	struct timeval lasterr;
		static	int curerr, once;
		if (once == 0 && ppsratecheck(&lasterr, &curerr, 1)) {
			printf("Bad malloc flags: %x\n", indx);
			kdb_backtrace();
			flags |= M_WAITOK;
			once++;
		}
	}
#endif
#ifdef MALLOC_MAKE_FAILURES
	if ((flags & M_NOWAIT) && (malloc_failure_rate != 0)) {
		atomic_add_int(&malloc_nowait_count, 1);
		if ((malloc_nowait_count % malloc_failure_rate) == 0) {
			atomic_add_int(&malloc_failure_count, 1);
			t_malloc_fail = time_uptime;
			return (NULL);
		}
	}
#endif
	if (flags & M_WAITOK)
		KASSERT(curthread->td_intr_nesting_level == 0,
		   ("malloc(M_WAITOK) in interrupt context"));

#ifdef DEBUG_MEMGUARD
	if (memguard_cmp_mtp(mtp, size)) {
		va = memguard_alloc(size, flags);
		if (va != NULL)
			return (va);
		/* This is unfortunate but should not be fatal. */
	}
#endif

#ifdef DEBUG_REDZONE
	size = redzone_size_ntor(size);
#endif

	if (size <= kmem_zmax) {
		mtip = mtp->ks_handle;
		if (size & KMEM_ZMASK)
			size = (size & ~KMEM_ZMASK) + KMEM_ZBASE;
		indx = kmemsize[size >> KMEM_ZSHIFT];
		KASSERT(mtip->mti_zone < numzones,
		    ("mti_zone %u out of range %d",
		    mtip->mti_zone, numzones));
		zone = kmemzones[indx].kz_zone[mtip->mti_zone];
#ifdef MALLOC_PROFILE
		krequests[size >> KMEM_ZSHIFT]++;
#endif
		va = uma_zalloc(zone, flags);
		if (va != NULL)
			size = zone->uz_size;
		malloc_type_zone_allocated(mtp, va == NULL ? 0 : size, indx);
	} else {
		size = roundup(size, PAGE_SIZE);
		zone = NULL;
		va = uma_large_malloc(size, flags);
		malloc_type_allocated(mtp, va == NULL ? 0 : size);
	}
	if (flags & M_WAITOK)
		KASSERT(va != NULL, ("malloc(M_WAITOK) returned NULL"));
	else if (va == NULL)
		t_malloc_fail = time_uptime;
#ifdef DIAGNOSTIC
	if (va != NULL && !(flags & M_ZERO)) {
		memset(va, 0x70, osize);
	}
#endif
#ifdef DEBUG_REDZONE
	if (va != NULL)
		va = redzone_setup(va, osize);
#endif
	return ((void *) va);
}

/*
 *	free:
 *
 *	Free a block of memory allocated by malloc.
 *
 *	This routine may not block.
 */
void
free(void *addr, struct malloc_type *mtp)
{
	uma_slab_t slab;
	u_long size;

	KASSERT(mtp->ks_magic == M_MAGIC, ("free: bad malloc type magic"));

	/* free(NULL, ...) does nothing */
	if (addr == NULL)
		return;

#ifdef DEBUG_MEMGUARD
	if (is_memguard_addr(addr)) {
		memguard_free(addr);
		return;
	}
#endif

#ifdef DEBUG_REDZONE
	redzone_check(addr);
	addr = redzone_addr_ntor(addr);
#endif

	slab = vtoslab((vm_offset_t)addr & (~UMA_SLAB_MASK));

	if (slab == NULL)
		panic("free: address %p(%p) has not been allocated.\n",
		    addr, (void *)((u_long)addr & (~UMA_SLAB_MASK)));

	if (!(slab->us_flags & UMA_SLAB_MALLOC)) {
#ifdef INVARIANTS
		struct malloc_type **mtpp = addr;
#endif
		size = slab->us_keg->uk_size;
#ifdef INVARIANTS
		/*
		 * Cache a pointer to the malloc_type that most recently freed
		 * this memory here.  This way we know who is most likely to
		 * have stepped on it later.
		 *
		 * This code assumes that size is a multiple of 8 bytes for
		 * 64 bit machines
		 */
		mtpp = (struct malloc_type **)
		    ((unsigned long)mtpp & ~UMA_ALIGN_PTR);
		mtpp += (size - sizeof(struct malloc_type *)) /
		    sizeof(struct malloc_type *);
		*mtpp = mtp;
#endif
		uma_zfree_arg(LIST_FIRST(&slab->us_keg->uk_zones), addr, slab);
	} else {
		size = slab->us_size;
		uma_large_free(slab);
	}
	malloc_type_freed(mtp, size);
}

/*
 *	realloc: change the size of a memory block
 */
void *
realloc(void *addr, unsigned long size, struct malloc_type *mtp, int flags)
{
	uma_slab_t slab;
	unsigned long alloc;
	void *newaddr;

	KASSERT(mtp->ks_magic == M_MAGIC,
	    ("realloc: bad malloc type magic"));

	/* realloc(NULL, ...) is equivalent to malloc(...) */
	if (addr == NULL)
		return (malloc(size, mtp, flags));

	/*
	 * XXX: Should report free of old memory and alloc of new memory to
	 * per-CPU stats.
	 */

#ifdef DEBUG_MEMGUARD
	if (is_memguard_addr(addr))
		return (memguard_realloc(addr, size, mtp, flags));
#endif

#ifdef DEBUG_REDZONE
	slab = NULL;
	alloc = redzone_get_size(addr);
#else
	slab = vtoslab((vm_offset_t)addr & ~(UMA_SLAB_MASK));

	/* Sanity check */
	KASSERT(slab != NULL,
	    ("realloc: address %p out of range", (void *)addr));

	/* Get the size of the original block */
	if (!(slab->us_flags & UMA_SLAB_MALLOC))
		alloc = slab->us_keg->uk_size;
	else
		alloc = slab->us_size;

	/* Reuse the original block if appropriate */
	if (size <= alloc
	    && (size > (alloc >> REALLOC_FRACTION) || alloc == MINALLOCSIZE))
		return (addr);
#endif /* !DEBUG_REDZONE */

	/* Allocate a new, bigger (or smaller) block */
	if ((newaddr = malloc(size, mtp, flags)) == NULL)
		return (NULL);

	/* Copy over original contents */
	bcopy(addr, newaddr, min(size, alloc));
	free(addr, mtp);
	return (newaddr);
}

/*
 *	reallocf: same as realloc() but free memory on failure.
 */
void *
reallocf(void *addr, unsigned long size, struct malloc_type *mtp, int flags)
{
	void *mem;

	if ((mem = realloc(addr, size, mtp, flags)) == NULL)
		free(addr, mtp);
	return (mem);
}

/*
 * Wake the page daemon when we exhaust KVA.  It will call the lowmem handler
 * and uma_reclaim() callbacks in a context that is safe.
 */
static void
kmem_reclaim(vmem_t *vm, int flags)
{

	pagedaemon_wakeup();
}

#ifndef __sparc64__
CTASSERT(VM_KMEM_SIZE_SCALE >= 1);
#endif

/*
 * Initialize the kernel memory (kmem) arena.
 */
void
kmeminit(void)
{
	u_long mem_size;
	u_long tmp;

#ifdef VM_KMEM_SIZE
	if (vm_kmem_size == 0)
		vm_kmem_size = VM_KMEM_SIZE;
#endif
#ifdef VM_KMEM_SIZE_MIN
	if (vm_kmem_size_min == 0)
		vm_kmem_size_min = VM_KMEM_SIZE_MIN;
#endif
#ifdef VM_KMEM_SIZE_MAX
	if (vm_kmem_size_max == 0)
		vm_kmem_size_max = VM_KMEM_SIZE_MAX;
#endif
	/*
	 * Calculate the amount of kernel virtual address (KVA) space that is
	 * preallocated to the kmem arena.  In order to support a wide range
	 * of machines, it is a function of the physical memory size,
	 * specifically,
	 *
	 *	min(max(physical memory size / VM_KMEM_SIZE_SCALE,
	 *	    VM_KMEM_SIZE_MIN), VM_KMEM_SIZE_MAX)
	 *
	 * Every architecture must define an integral value for
	 * VM_KMEM_SIZE_SCALE.  However, the definitions of VM_KMEM_SIZE_MIN
	 * and VM_KMEM_SIZE_MAX, which represent respectively the floor and
	 * ceiling on this preallocation, are optional.  Typically,
	 * VM_KMEM_SIZE_MAX is itself a function of the available KVA space on
	 * a given architecture.
	 */
	mem_size = vm_cnt.v_page_count;

	if (vm_kmem_size_scale < 1)
		vm_kmem_size_scale = VM_KMEM_SIZE_SCALE;

	/*
	 * Check if we should use defaults for the "vm_kmem_size"
	 * variable:
	 */
	if (vm_kmem_size == 0) {
		vm_kmem_size = (mem_size / vm_kmem_size_scale) * PAGE_SIZE;

		if (vm_kmem_size_min > 0 && vm_kmem_size < vm_kmem_size_min)
			vm_kmem_size = vm_kmem_size_min;
		if (vm_kmem_size_max > 0 && vm_kmem_size >= vm_kmem_size_max)
			vm_kmem_size = vm_kmem_size_max;
	}

	/*
	 * The amount of KVA space that is preallocated to the
	 * kmem arena can be set statically at compile-time or manually
	 * through the kernel environment.  However, it is still limited to
	 * twice the physical memory size, which has been sufficient to handle
	 * the most severe cases of external fragmentation in the kmem arena. 
	 */
	if (vm_kmem_size / 2 / PAGE_SIZE > mem_size)
		vm_kmem_size = 2 * mem_size * PAGE_SIZE;

	vm_kmem_size = round_page(vm_kmem_size);
#ifdef DEBUG_MEMGUARD
	tmp = memguard_fudge(vm_kmem_size, kernel_map);
#else
	tmp = vm_kmem_size;
#endif
	vmem_init(kmem_arena, "kmem arena", kva_alloc(tmp), tmp, PAGE_SIZE,
	    0, 0);
	vmem_set_reclaim(kmem_arena, kmem_reclaim);

#ifdef DEBUG_MEMGUARD
	/*
	 * Initialize MemGuard if support compiled in.  MemGuard is a
	 * replacement allocator used for detecting tamper-after-free
	 * scenarios as they occur.  It is only used for debugging.
	 */
	memguard_init(kmem_arena);
#endif
}

/*
 * Initialize the kernel memory allocator
 */
/* ARGSUSED*/
static void
mallocinit(void *dummy)
{
	int i;
	uint8_t indx;

	mtx_init(&malloc_mtx, "malloc", NULL, MTX_DEF);

	kmeminit();

	uma_startup2();

	if (kmem_zmax < PAGE_SIZE || kmem_zmax > KMEM_ZMAX)
		kmem_zmax = KMEM_ZMAX;

	mt_zone = uma_zcreate("mt_zone", sizeof(struct malloc_type_internal),
#ifdef INVARIANTS
	    mtrash_ctor, mtrash_dtor, mtrash_init, mtrash_fini,
#else
	    NULL, NULL, NULL, NULL,
#endif
	    UMA_ALIGN_PTR, UMA_ZONE_MALLOC);
	for (i = 0, indx = 0; kmemzones[indx].kz_size != 0; indx++) {
		int size = kmemzones[indx].kz_size;
		char *name = kmemzones[indx].kz_name;
		int subzone;

		for (subzone = 0; subzone < numzones; subzone++) {
			kmemzones[indx].kz_zone[subzone] =
			    uma_zcreate(name, size,
#ifdef INVARIANTS
			    mtrash_ctor, mtrash_dtor, mtrash_init, mtrash_fini,
#else
			    NULL, NULL, NULL, NULL,
#endif
			    UMA_ALIGN_PTR, UMA_ZONE_MALLOC);
		}		    
		for (;i <= size; i+= KMEM_ZBASE)
			kmemsize[i >> KMEM_ZSHIFT] = indx;

	}
}
SYSINIT(kmem, SI_SUB_KMEM, SI_ORDER_SECOND, mallocinit, NULL);

void
malloc_init(void *data)
{
	struct malloc_type_internal *mtip;
	struct malloc_type *mtp;

	KASSERT(vm_cnt.v_page_count != 0, ("malloc_register before vm_init"));

	mtp = data;
	if (mtp->ks_magic != M_MAGIC)
		panic("malloc_init: bad malloc type magic");

	mtip = uma_zalloc(mt_zone, M_WAITOK | M_ZERO);
	mtp->ks_handle = mtip;
	mtip->mti_zone = mtp_get_subzone(mtp->ks_shortdesc);

	mtx_lock(&malloc_mtx);
	mtp->ks_next = kmemstatistics;
	kmemstatistics = mtp;
	kmemcount++;
	mtx_unlock(&malloc_mtx);
}

void
malloc_uninit(void *data)
{
	struct malloc_type_internal *mtip;
	struct malloc_type_stats *mtsp;
	struct malloc_type *mtp, *temp;
	uma_slab_t slab;
	long temp_allocs, temp_bytes;
	int i;

	mtp = data;
	KASSERT(mtp->ks_magic == M_MAGIC,
	    ("malloc_uninit: bad malloc type magic"));
	KASSERT(mtp->ks_handle != NULL, ("malloc_deregister: cookie NULL"));

	mtx_lock(&malloc_mtx);
	mtip = mtp->ks_handle;
	mtp->ks_handle = NULL;
	if (mtp != kmemstatistics) {
		for (temp = kmemstatistics; temp != NULL;
		    temp = temp->ks_next) {
			if (temp->ks_next == mtp) {
				temp->ks_next = mtp->ks_next;
				break;
			}
		}
		KASSERT(temp,
		    ("malloc_uninit: type '%s' not found", mtp->ks_shortdesc));
	} else
		kmemstatistics = mtp->ks_next;
	kmemcount--;
	mtx_unlock(&malloc_mtx);

	/*
	 * Look for memory leaks.
	 */
	temp_allocs = temp_bytes = 0;
	for (i = 0; i < MAXCPU; i++) {
		mtsp = &mtip->mti_stats[i];
		temp_allocs += mtsp->mts_numallocs;
		temp_allocs -= mtsp->mts_numfrees;
		temp_bytes += mtsp->mts_memalloced;
		temp_bytes -= mtsp->mts_memfreed;
	}
	if (temp_allocs > 0 || temp_bytes > 0) {
		printf("Warning: memory type %s leaked memory on destroy "
		    "(%ld allocations, %ld bytes leaked).\n", mtp->ks_shortdesc,
		    temp_allocs, temp_bytes);
	}

	slab = vtoslab((vm_offset_t) mtip & (~UMA_SLAB_MASK));
	uma_zfree_arg(mt_zone, mtip, slab);
}

struct malloc_type *
malloc_desc2type(const char *desc)
{
	struct malloc_type *mtp;

	mtx_assert(&malloc_mtx, MA_OWNED);
	for (mtp = kmemstatistics; mtp != NULL; mtp = mtp->ks_next) {
		if (strcmp(mtp->ks_shortdesc, desc) == 0)
			return (mtp);
	}
	return (NULL);
}

static int
sysctl_kern_malloc_stats(SYSCTL_HANDLER_ARGS)
{
	struct malloc_type_stream_header mtsh;
	struct malloc_type_internal *mtip;
	struct malloc_type_header mth;
	struct malloc_type *mtp;
	int error, i;
	struct sbuf sbuf;

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

	/*
	 * Insert stream header.
	 */
	bzero(&mtsh, sizeof(mtsh));
	mtsh.mtsh_version = MALLOC_TYPE_STREAM_VERSION;
	mtsh.mtsh_maxcpus = MAXCPU;
	mtsh.mtsh_count = kmemcount;
	(void)sbuf_bcat(&sbuf, &mtsh, sizeof(mtsh));

	/*
	 * Insert alternating sequence of type headers and type statistics.
	 */
	for (mtp = kmemstatistics; mtp != NULL; mtp = mtp->ks_next) {
		mtip = (struct malloc_type_internal *)mtp->ks_handle;

		/*
		 * Insert type header.
		 */
		bzero(&mth, sizeof(mth));
		strlcpy(mth.mth_name, mtp->ks_shortdesc, MALLOC_MAX_NAME);
		(void)sbuf_bcat(&sbuf, &mth, sizeof(mth));

		/*
		 * Insert type statistics for each CPU.
		 */
		for (i = 0; i < MAXCPU; i++) {
			(void)sbuf_bcat(&sbuf, &mtip->mti_stats[i],
			    sizeof(mtip->mti_stats[i]));
		}
	}
	mtx_unlock(&malloc_mtx);
	error = sbuf_finish(&sbuf);
	sbuf_delete(&sbuf);
	return (error);
}

SYSCTL_PROC(_kern, OID_AUTO, malloc_stats, CTLFLAG_RD|CTLTYPE_STRUCT,
    0, 0, sysctl_kern_malloc_stats, "s,malloc_type_ustats",
    "Return malloc types");

SYSCTL_INT(_kern, OID_AUTO, malloc_count, CTLFLAG_RD, &kmemcount, 0,
    "Count of kernel malloc types");

void
malloc_type_list(malloc_type_list_func_t *func, void *arg)
{
	struct malloc_type *mtp, **bufmtp;
	int count, i;
	size_t buflen;

	mtx_lock(&malloc_mtx);
restart:
	mtx_assert(&malloc_mtx, MA_OWNED);
	count = kmemcount;
	mtx_unlock(&malloc_mtx);

	buflen = sizeof(struct malloc_type *) * count;
	bufmtp = malloc(buflen, M_TEMP, M_WAITOK);

	mtx_lock(&malloc_mtx);

	if (count < kmemcount) {
		free(bufmtp, M_TEMP);
		goto restart;
	}

	for (mtp = kmemstatistics, i = 0; mtp != NULL; mtp = mtp->ks_next, i++)
		bufmtp[i] = mtp;

	mtx_unlock(&malloc_mtx);

	for (i = 0; i < count; i++)
		(func)(bufmtp[i], arg);

	free(bufmtp, M_TEMP);
}

#ifdef DDB
DB_SHOW_COMMAND(malloc, db_show_malloc)
{
	struct malloc_type_internal *mtip;
	struct malloc_type *mtp;
	uint64_t allocs, frees;
	uint64_t alloced, freed;
	int i;

	db_printf("%18s %12s  %12s %12s\n", "Type", "InUse", "MemUse",
	    "Requests");
	for (mtp = kmemstatistics; mtp != NULL; mtp = mtp->ks_next) {
		mtip = (struct malloc_type_internal *)mtp->ks_handle;
		allocs = 0;
		frees = 0;
		alloced = 0;
		freed = 0;
		for (i = 0; i < MAXCPU; i++) {
			allocs += mtip->mti_stats[i].mts_numallocs;
			frees += mtip->mti_stats[i].mts_numfrees;
			alloced += mtip->mti_stats[i].mts_memalloced;
			freed += mtip->mti_stats[i].mts_memfreed;
		}
		db_printf("%18s %12ju %12juK %12ju\n",
		    mtp->ks_shortdesc, allocs - frees,
		    (alloced - freed + 1023) / 1024, allocs);
		if (db_pager_quit)
			break;
	}
}

#if MALLOC_DEBUG_MAXZONES > 1
DB_SHOW_COMMAND(multizone_matches, db_show_multizone_matches)
{
	struct malloc_type_internal *mtip;
	struct malloc_type *mtp;
	u_int subzone;

	if (!have_addr) {
		db_printf("Usage: show multizone_matches <malloc type/addr>\n");
		return;
	}
	mtp = (void *)addr;
	if (mtp->ks_magic != M_MAGIC) {
		db_printf("Magic %lx does not match expected %x\n",
		    mtp->ks_magic, M_MAGIC);
		return;
	}

	mtip = mtp->ks_handle;
	subzone = mtip->mti_zone;

	for (mtp = kmemstatistics; mtp != NULL; mtp = mtp->ks_next) {
		mtip = mtp->ks_handle;
		if (mtip->mti_zone != subzone)
			continue;
		db_printf("%s\n", mtp->ks_shortdesc);
		if (db_pager_quit)
			break;
	}
}
#endif /* MALLOC_DEBUG_MAXZONES > 1 */
#endif /* DDB */

#ifdef MALLOC_PROFILE

static int
sysctl_kern_mprof(SYSCTL_HANDLER_ARGS)
{
	struct sbuf sbuf;
	uint64_t count;
	uint64_t waste;
	uint64_t mem;
	int error;
	int rsize;
	int size;
	int i;

	waste = 0;
	mem = 0;

	error = sysctl_wire_old_buffer(req, 0);
	if (error != 0)
		return (error);
	sbuf_new_for_sysctl(&sbuf, NULL, 128, req);
	sbuf_printf(&sbuf, 
	    "\n  Size                    Requests  Real Size\n");
	for (i = 0; i < KMEM_ZSIZE; i++) {
		size = i << KMEM_ZSHIFT;
		rsize = kmemzones[kmemsize[i]].kz_size;
		count = (long long unsigned)krequests[i];

		sbuf_printf(&sbuf, "%6d%28llu%11d\n", size,
		    (unsigned long long)count, rsize);

		if ((rsize * count) > (size * count))
			waste += (rsize * count) - (size * count);
		mem += (rsize * count);
	}
	sbuf_printf(&sbuf,
	    "\nTotal memory used:\t%30llu\nTotal Memory wasted:\t%30llu\n",
	    (unsigned long long)mem, (unsigned long long)waste);
	error = sbuf_finish(&sbuf);
	sbuf_delete(&sbuf);
	return (error);
}

SYSCTL_OID(_kern, OID_AUTO, mprof, CTLTYPE_STRING|CTLFLAG_RD,
    NULL, 0, sysctl_kern_mprof, "A", "Malloc Profiling");
#endif /* MALLOC_PROFILE */