summaryrefslogtreecommitdiffstats
path: root/sys/kern/uipc_sem.c
blob: c9903aabc049883a5d518ce0977c1797c2c412af (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
/*-
 * Copyright (c) 2002 Alfred Perlstein <alfred@FreeBSD.org>
 * Copyright (c) 2003-2005 SPARTA, Inc.
 * Copyright (c) 2005 Robert N. M. Watson
 * All rights reserved.
 *
 * This software was developed for the FreeBSD Project in part by Network
 * Associates Laboratories, the Security Research Division of Network
 * Associates, Inc. under DARPA/SPAWAR contract N66001-01-C-8035 ("CBOSS"),
 * as part of the DARPA CHATS research program.
 *
 * 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 AUTHOR 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 AUTHOR 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.
 */

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

#include "opt_mac.h"
#include "opt_posix.h"

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/sysproto.h>
#include <sys/eventhandler.h>
#include <sys/kernel.h>
#include <sys/ksem.h>
#include <sys/priv.h>
#include <sys/proc.h>
#include <sys/posix4.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/module.h>
#include <sys/condvar.h>
#include <sys/sem.h>
#include <sys/uio.h>
#include <sys/semaphore.h>
#include <sys/syscall.h>
#include <sys/stat.h>
#include <sys/sysent.h>
#include <sys/sysctl.h>
#include <sys/time.h>
#include <sys/malloc.h>
#include <sys/fcntl.h>
#include <sys/_semaphore.h>

#include <security/mac/mac_framework.h>

static int sem_count_proc(struct proc *p);
static struct ksem *sem_lookup_byname(const char *name);
static int sem_create(struct thread *td, const char *name,
    struct ksem **ksret, mode_t mode, unsigned int value);
static void sem_free(struct ksem *ksnew);
static int sem_perm(struct thread *td, struct ksem *ks);
static void sem_enter(struct proc *p, struct ksem *ks);
static int sem_leave(struct proc *p, struct ksem *ks);
static void sem_exechook(void *arg, struct proc *p, struct image_params *imgp);
static void sem_exithook(void *arg, struct proc *p);
static void sem_forkhook(void *arg, struct proc *p1, struct proc *p2,
    int flags);
static int sem_hasopen(struct thread *td, struct ksem *ks);

static int kern_sem_close(struct thread *td, semid_t id);
static int kern_sem_post(struct thread *td, semid_t id);
static int kern_sem_wait(struct thread *td, semid_t id, int tryflag,
    struct timespec *abstime);
static int kern_sem_init(struct thread *td, int dir, unsigned int value,
    semid_t *idp);
static int kern_sem_open(struct thread *td, int dir, const char *name,
    int oflag, mode_t mode, unsigned int value, semid_t *idp);
static int kern_sem_unlink(struct thread *td, const char *name);

#ifndef SEM_MAX
#define SEM_MAX	30
#endif

#define SEM_MAX_NAMELEN	14

#define SEM_TO_ID(x)	((intptr_t)(x))
#define ID_TO_SEM(x)	id_to_sem(x)

/*
 * available semaphores go here, this includes sem_init and any semaphores
 * created via sem_open that have not yet been unlinked.
 */
LIST_HEAD(, ksem) ksem_head = LIST_HEAD_INITIALIZER(&ksem_head);
/*
 * semaphores still in use but have been sem_unlink()'d go here.
 */
LIST_HEAD(, ksem) ksem_deadhead = LIST_HEAD_INITIALIZER(&ksem_deadhead);

static struct mtx sem_lock;
static MALLOC_DEFINE(M_SEM, "sems", "semaphore data");

static int nsems = 0;
SYSCTL_DECL(_p1003_1b);
SYSCTL_INT(_p1003_1b, OID_AUTO, nsems, CTLFLAG_RD, &nsems, 0, "");

static eventhandler_tag sem_exit_tag, sem_exec_tag, sem_fork_tag;

#ifdef SEM_DEBUG
#define DP(x)	printf x
#else
#define DP(x)
#endif

static __inline
void
sem_ref(struct ksem *ks)
{

	mtx_assert(&sem_lock, MA_OWNED);
	ks->ks_ref++;
	DP(("sem_ref: ks = %p, ref = %d\n", ks, ks->ks_ref));
}

static __inline
void
sem_rel(struct ksem *ks)
{

	mtx_assert(&sem_lock, MA_OWNED);
	DP(("sem_rel: ks = %p, ref = %d\n", ks, ks->ks_ref - 1));
	if (--ks->ks_ref == 0)
		sem_free(ks);
}

static __inline struct ksem *id_to_sem(semid_t id);

static __inline
struct ksem *
id_to_sem(semid_t id)
{
	struct ksem *ks;

	mtx_assert(&sem_lock, MA_OWNED);
	DP(("id_to_sem: id = %0x,%p\n", id, (struct ksem *)id));
	LIST_FOREACH(ks, &ksem_head, ks_entry) {
		DP(("id_to_sem: ks = %p\n", ks));
		if (ks == (struct ksem *)id)
			return (ks);
	}
	return (NULL);
}

static struct ksem *
sem_lookup_byname(const char *name)
{
	struct ksem *ks;

	mtx_assert(&sem_lock, MA_OWNED);
	LIST_FOREACH(ks, &ksem_head, ks_entry)
		if (ks->ks_name != NULL && strcmp(ks->ks_name, name) == 0)
			return (ks);
	return (NULL);
}

static int
sem_create(struct thread *td, const char *name, struct ksem **ksret,
    mode_t mode, unsigned int value)
{
	struct ksem *ret;
	struct proc *p;
	struct ucred *uc;
	size_t len;
	int error;

	DP(("sem_create\n"));
	p = td->td_proc;
	uc = td->td_ucred;
	if (value > SEM_VALUE_MAX)
		return (EINVAL);
	ret = malloc(sizeof(*ret), M_SEM, M_WAITOK | M_ZERO);
	if (name != NULL) {
		len = strlen(name);
		if (len > SEM_MAX_NAMELEN) {
			free(ret, M_SEM);
			return (ENAMETOOLONG);
		}
		/* name must start with a '/' but not contain one. */
		if (*name != '/' || len < 2 || index(name + 1, '/') != NULL) {
			free(ret, M_SEM);
			return (EINVAL);
		}
		ret->ks_name = malloc(len + 1, M_SEM, M_WAITOK);
		strcpy(ret->ks_name, name);
	} else {
		ret->ks_name = NULL;
	}
	ret->ks_mode = mode;
	ret->ks_value = value;
	ret->ks_ref = 1;
	ret->ks_waiters = 0;
	ret->ks_uid = uc->cr_uid;
	ret->ks_gid = uc->cr_gid;
	ret->ks_onlist = 0;
	cv_init(&ret->ks_cv, "sem");
	LIST_INIT(&ret->ks_users);
#ifdef MAC
	mac_init_posix_sem(ret);
	mac_create_posix_sem(uc, ret);
#endif
	if (name != NULL)
		sem_enter(td->td_proc, ret);
	*ksret = ret;
	mtx_lock(&sem_lock);
	if (nsems >= p31b_getcfg(CTL_P1003_1B_SEM_NSEMS_MAX)) {
		sem_leave(td->td_proc, ret);
		sem_free(ret);
		error = ENFILE;
	} else {
		nsems++;
		error = 0;
	}
	mtx_unlock(&sem_lock);
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_init_args {
	unsigned int value;
	semid_t *idp;
};
int ksem_init(struct thread *td, struct ksem_init_args *uap);
#endif
int
ksem_init(struct thread *td, struct ksem_init_args *uap)
{
	int error;

	error = kern_sem_init(td, UIO_USERSPACE, uap->value, uap->idp);
	return (error);
}

static int
kern_sem_init(struct thread *td, int dir, unsigned int value, semid_t *idp)
{
	struct ksem *ks;
	semid_t id;
	int error;

	error = sem_create(td, NULL, &ks, S_IRWXU | S_IRWXG, value);
	if (error)
		return (error);
	id = SEM_TO_ID(ks);
	if (dir == UIO_USERSPACE) {
		error = copyout(&id, idp, sizeof(id));
		if (error) {
			mtx_lock(&sem_lock);
			sem_rel(ks);
			mtx_unlock(&sem_lock);
			return (error);
		}
	} else {
		*idp = id;
	}
	mtx_lock(&sem_lock);
	LIST_INSERT_HEAD(&ksem_head, ks, ks_entry);
	ks->ks_onlist = 1;
	mtx_unlock(&sem_lock);
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_open_args {
	char *name;
	int oflag;
	mode_t mode;
	unsigned int value;
	semid_t *idp;	
};
int ksem_open(struct thread *td, struct ksem_open_args *uap);
#endif
int
ksem_open(struct thread *td, struct ksem_open_args *uap)
{
	char name[SEM_MAX_NAMELEN + 1];
	size_t done;
	int error;

	error = copyinstr(uap->name, name, SEM_MAX_NAMELEN + 1, &done);
	if (error)
		return (error);
	DP((">>> sem_open start\n"));
	error = kern_sem_open(td, UIO_USERSPACE,
	    name, uap->oflag, uap->mode, uap->value, uap->idp);
	DP(("<<< sem_open end\n"));
	return (error);
}

static int
kern_sem_open(struct thread *td, int dir, const char *name, int oflag,
    mode_t mode, unsigned int value, semid_t *idp)
{
	struct ksem *ksnew, *ks;
	int error;
	semid_t id;

	ksnew = NULL;
	mtx_lock(&sem_lock);
	ks = sem_lookup_byname(name);
	/*
	 * If we found it but O_EXCL is set, error.
	 */
	if (ks != NULL && (oflag & O_EXCL) != 0) {
		mtx_unlock(&sem_lock);
		return (EEXIST);
	}
	/*
	 * If we didn't find it...
	 */
	if (ks == NULL) {
		/*
		 * didn't ask for creation? error.
		 */
		if ((oflag & O_CREAT) == 0) {
			mtx_unlock(&sem_lock);
			return (ENOENT);
		}
		/*
		 * We may block during creation, so drop the lock.
		 */
		mtx_unlock(&sem_lock);
		error = sem_create(td, name, &ksnew, mode, value);
		if (error != 0)
			return (error);
		id = SEM_TO_ID(ksnew);
		if (dir == UIO_USERSPACE) {
			DP(("about to copyout! %d to %p\n", id, idp));
			error = copyout(&id, idp, sizeof(id));
			if (error) {
				mtx_lock(&sem_lock);
				sem_leave(td->td_proc, ksnew);
				sem_rel(ksnew);
				mtx_unlock(&sem_lock);
				return (error);
			}
		} else {
			DP(("about to set! %d to %p\n", id, idp));
			*idp = id;
		}
		/*
		 * We need to make sure we haven't lost a race while
		 * allocating during creation.
		 */
		mtx_lock(&sem_lock);
		ks = sem_lookup_byname(name);
		if (ks != NULL) {
			/* we lost... */
			sem_leave(td->td_proc, ksnew);
			sem_rel(ksnew);
			/* we lost and we can't loose... */
			if ((oflag & O_EXCL) != 0) {
				mtx_unlock(&sem_lock);
				return (EEXIST);
			}
		} else {
			DP(("sem_create: about to add to list...\n"));
			LIST_INSERT_HEAD(&ksem_head, ksnew, ks_entry); 
			DP(("sem_create: setting list bit...\n"));
			ksnew->ks_onlist = 1;
			DP(("sem_create: done, about to unlock...\n"));
		}
	} else {
#ifdef MAC
		error = mac_check_posix_sem_open(td->td_ucred, ks);
		if (error)
			goto err_open;
#endif
		/*
		 * if we aren't the creator, then enforce permissions.
		 */
		error = sem_perm(td, ks);
		if (error)
			goto err_open;
		sem_ref(ks);
		mtx_unlock(&sem_lock);
		id = SEM_TO_ID(ks);
		if (dir == UIO_USERSPACE) {
			error = copyout(&id, idp, sizeof(id));
			if (error) {
				mtx_lock(&sem_lock);
				sem_rel(ks);
				mtx_unlock(&sem_lock);
				return (error);
			}
		} else {
			*idp = id;
		}
		sem_enter(td->td_proc, ks);
		mtx_lock(&sem_lock);
		sem_rel(ks);
	}
err_open:
	mtx_unlock(&sem_lock);
	return (error);
}

static int
sem_perm(struct thread *td, struct ksem *ks)
{
	struct ucred *uc;

	/*
	 * XXXRW: This permission routine appears to be incorrect.  If the
	 * user matches, we shouldn't go on to the group if the user
	 * permissions don't allow the action?  Not changed for now.  To fix,
	 * change from a series of if (); if (); to if () else if () else...
	 */
	uc = td->td_ucred;
	DP(("sem_perm: uc(%d,%d) ks(%d,%d,%o)\n",
	    uc->cr_uid, uc->cr_gid,
	     ks->ks_uid, ks->ks_gid, ks->ks_mode));
	if ((uc->cr_uid == ks->ks_uid) && (ks->ks_mode & S_IWUSR) != 0)
		return (0);
	if ((uc->cr_gid == ks->ks_gid) && (ks->ks_mode & S_IWGRP) != 0)
		return (0);
	if ((ks->ks_mode & S_IWOTH) != 0)
		return (0);
	return (priv_check(td, PRIV_SEM_WRITE));
}

static void
sem_free(struct ksem *ks)
{

	nsems--;
	if (ks->ks_onlist)
		LIST_REMOVE(ks, ks_entry);
	if (ks->ks_name != NULL)
		free(ks->ks_name, M_SEM);
	cv_destroy(&ks->ks_cv);
	free(ks, M_SEM);
}

static __inline struct kuser *sem_getuser(struct proc *p, struct ksem *ks);

static __inline struct kuser *
sem_getuser(struct proc *p, struct ksem *ks)
{
	struct kuser *k;

	LIST_FOREACH(k, &ks->ks_users, ku_next)
		if (k->ku_pid == p->p_pid)
			return (k);
	return (NULL);
}

static int
sem_hasopen(struct thread *td, struct ksem *ks)
{
	
	return ((ks->ks_name == NULL && sem_perm(td, ks) == 0)
	    || sem_getuser(td->td_proc, ks) != NULL);
}

static int
sem_leave(struct proc *p, struct ksem *ks)
{
	struct kuser *k;

	DP(("sem_leave: ks = %p\n", ks));
	k = sem_getuser(p, ks);
	DP(("sem_leave: ks = %p, k = %p\n", ks, k));
	if (k != NULL) {
		LIST_REMOVE(k, ku_next);
		sem_rel(ks);
		DP(("sem_leave: about to free k\n"));
		free(k, M_SEM);
		DP(("sem_leave: returning\n"));
		return (0);
	}
	return (EINVAL);
}

static void
sem_enter(p, ks)
	struct proc *p;
	struct ksem *ks;
{
	struct kuser *ku, *k;

	ku = malloc(sizeof(*ku), M_SEM, M_WAITOK);
	ku->ku_pid = p->p_pid;
	mtx_lock(&sem_lock);
	k = sem_getuser(p, ks);
	if (k != NULL) {
		mtx_unlock(&sem_lock);
		free(ku, M_TEMP);
		return;
	}
	LIST_INSERT_HEAD(&ks->ks_users, ku, ku_next);
	sem_ref(ks);
	mtx_unlock(&sem_lock);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_unlink_args {
	char *name;
};
int ksem_unlink(struct thread *td, struct ksem_unlink_args *uap);
#endif
int
ksem_unlink(struct thread *td, struct ksem_unlink_args *uap)
{
	char name[SEM_MAX_NAMELEN + 1];
	size_t done;
	int error;

	error = copyinstr(uap->name, name, SEM_MAX_NAMELEN + 1, &done);
	return (error ? error :
	    kern_sem_unlink(td, name));
}

static int
kern_sem_unlink(struct thread *td, const char *name)
{
	struct ksem *ks;
	int error;

	mtx_lock(&sem_lock);
	ks = sem_lookup_byname(name);
	if (ks != NULL) {
#ifdef MAC
		error = mac_check_posix_sem_unlink(td->td_ucred, ks);
		if (error) {
			mtx_unlock(&sem_lock);
			return (error);
		}
#endif
		error = sem_perm(td, ks);
	} else
		error = ENOENT;
	DP(("sem_unlink: '%s' ks = %p, error = %d\n", name, ks, error));
	if (error == 0) {
		LIST_REMOVE(ks, ks_entry);
		LIST_INSERT_HEAD(&ksem_deadhead, ks, ks_entry); 
		sem_rel(ks);
	}
	mtx_unlock(&sem_lock);
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_close_args {
	semid_t id;
};
int ksem_close(struct thread *td, struct ksem_close_args *uap);
#endif
int
ksem_close(struct thread *td, struct ksem_close_args *uap)
{

	return (kern_sem_close(td, uap->id));
}

static int
kern_sem_close(struct thread *td, semid_t id)
{
	struct ksem *ks;
	int error;

	error = EINVAL;
	mtx_lock(&sem_lock);
	ks = ID_TO_SEM(id);
	/* this is not a valid operation for unnamed sems */
	if (ks != NULL && ks->ks_name != NULL)
		error = sem_leave(td->td_proc, ks);
	mtx_unlock(&sem_lock);
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_post_args {
	semid_t id;
};
int ksem_post(struct thread *td, struct ksem_post_args *uap);
#endif
int
ksem_post(struct thread *td, struct ksem_post_args *uap)
{

	return (kern_sem_post(td, uap->id));
}

static int
kern_sem_post(struct thread *td, semid_t id)
{
	struct ksem *ks;
	int error;

	mtx_lock(&sem_lock);
	ks = ID_TO_SEM(id);
	if (ks == NULL || !sem_hasopen(td, ks)) {
		error = EINVAL;
		goto err;
	}
#ifdef MAC
	error = mac_check_posix_sem_post(td->td_ucred, ks);
	if (error)
		goto err;
#endif
	if (ks->ks_value == SEM_VALUE_MAX) {
		error = EOVERFLOW;
		goto err;
	}
	++ks->ks_value;
	if (ks->ks_waiters > 0)
		cv_signal(&ks->ks_cv);
	error = 0;
err:
	mtx_unlock(&sem_lock);
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_wait_args {
	semid_t id;
};
int ksem_wait(struct thread *td, struct ksem_wait_args *uap);
#endif
int
ksem_wait(struct thread *td, struct ksem_wait_args *uap)
{

	return (kern_sem_wait(td, uap->id, 0, NULL));
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_timedwait_args {
	semid_t id;
	const struct timespec *abstime;
};
int ksem_timedwait(struct thread *td, struct ksem_timedwait_args *uap);
#endif
int
ksem_timedwait(struct thread *td, struct ksem_timedwait_args *uap)
{
	struct timespec abstime;
	struct timespec *ts;
	int error;

	/* We allow a null timespec (wait forever). */
	if (uap->abstime == NULL)
		ts = NULL;
	else {
		error = copyin(uap->abstime, &abstime, sizeof(abstime));
		if (error != 0)
			return (error);
		if (abstime.tv_nsec >= 1000000000 || abstime.tv_nsec < 0)
			return (EINVAL);
		ts = &abstime;
	}
	return (kern_sem_wait(td, uap->id, 0, ts));
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_trywait_args {
	semid_t id;
};
int ksem_trywait(struct thread *td, struct ksem_trywait_args *uap);
#endif
int
ksem_trywait(struct thread *td, struct ksem_trywait_args *uap)
{

	return (kern_sem_wait(td, uap->id, 1, NULL));
}

static int
kern_sem_wait(struct thread *td, semid_t id, int tryflag,
    struct timespec *abstime)
{
	struct timespec ts1, ts2;
	struct timeval tv;
	struct ksem *ks;
	int error;

	DP((">>> kern_sem_wait entered!\n"));
	mtx_lock(&sem_lock);
	ks = ID_TO_SEM(id);
	if (ks == NULL) {
		DP(("kern_sem_wait ks == NULL\n"));
		error = EINVAL;
		goto err;
	}
	sem_ref(ks);
	if (!sem_hasopen(td, ks)) {
		DP(("kern_sem_wait hasopen failed\n"));
		error = EINVAL;
		goto err;
	}
#ifdef MAC
	error = mac_check_posix_sem_wait(td->td_ucred, ks);
	if (error) {
		DP(("kern_sem_wait mac failed\n"));
		goto err;
	}
#endif
	DP(("kern_sem_wait value = %d, tryflag %d\n", ks->ks_value, tryflag));
	if (ks->ks_value == 0) {
		ks->ks_waiters++;
		if (tryflag != 0)
			error = EAGAIN;
		else if (abstime == NULL)
			error = cv_wait_sig(&ks->ks_cv, &sem_lock);
		else {
			for (;;) {
				ts1 = *abstime;
				getnanotime(&ts2);
				timespecsub(&ts1, &ts2);
				TIMESPEC_TO_TIMEVAL(&tv, &ts1);
				if (tv.tv_sec < 0) {
					error = ETIMEDOUT;
					break;
				}
				error = cv_timedwait_sig(&ks->ks_cv,
				    &sem_lock, tvtohz(&tv));
				if (error != EWOULDBLOCK)
					break;
			}
		}
		ks->ks_waiters--;
		if (error)
			goto err;
	}
	ks->ks_value--;
	error = 0;
err:
	if (ks != NULL)
		sem_rel(ks);
	mtx_unlock(&sem_lock);
	DP(("<<< kern_sem_wait leaving, error = %d\n", error));
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_getvalue_args {
	semid_t id;
	int *val;
};
int ksem_getvalue(struct thread *td, struct ksem_getvalue_args *uap);
#endif
int
ksem_getvalue(struct thread *td, struct ksem_getvalue_args *uap)
{
	struct ksem *ks;
	int error, val;

	mtx_lock(&sem_lock);
	ks = ID_TO_SEM(uap->id);
	if (ks == NULL || !sem_hasopen(td, ks)) {
		mtx_unlock(&sem_lock);
		return (EINVAL);
	}
#ifdef MAC
	error = mac_check_posix_sem_getvalue(td->td_ucred, ks);
	if (error) {
		mtx_unlock(&sem_lock);
		return (error);
	}
#endif
	val = ks->ks_value;
	mtx_unlock(&sem_lock);
	error = copyout(&val, uap->val, sizeof(val));
	return (error);
}

#ifndef _SYS_SYSPROTO_H_
struct ksem_destroy_args {
	semid_t id;
};
int ksem_destroy(struct thread *td, struct ksem_destroy_args *uap);
#endif
int
ksem_destroy(struct thread *td, struct ksem_destroy_args *uap)
{
	struct ksem *ks;
	int error;

	mtx_lock(&sem_lock);
	ks = ID_TO_SEM(uap->id);
	if (ks == NULL || !sem_hasopen(td, ks) ||
	    ks->ks_name != NULL) {
		error = EINVAL;
		goto err;
	}
#ifdef MAC
	error = mac_check_posix_sem_destroy(td->td_ucred, ks);
	if (error)
		goto err;
#endif
	if (ks->ks_waiters != 0) {
		error = EBUSY;
		goto err;
	}
	sem_rel(ks);
	error = 0;
err:
	mtx_unlock(&sem_lock);
	return (error);
}

/*
 * Count the number of kusers associated with a proc, so as to guess at how
 * many to allocate when forking.
 */
static int
sem_count_proc(struct proc *p)
{
	struct ksem *ks;
	struct kuser *ku;
	int count;

	mtx_assert(&sem_lock, MA_OWNED);

	count = 0;
	LIST_FOREACH(ks, &ksem_head, ks_entry) {
		LIST_FOREACH(ku, &ks->ks_users, ku_next) {
			if (ku->ku_pid == p->p_pid)
				count++;
		}
	}
	LIST_FOREACH(ks, &ksem_deadhead, ks_entry) {
		LIST_FOREACH(ku, &ks->ks_users, ku_next) {
			if (ku->ku_pid == p->p_pid)
				count++;
		}
	}
	return (count);
}

/*
 * When a process forks, the child process must gain a reference to each open
 * semaphore in the parent process, whether it is unlinked or not.  This
 * requires allocating a kuser structure for each semaphore reference in the
 * new process.  Because the set of semaphores in the parent can change while
 * the fork is in progress, we have to handle races -- first we attempt to
 * allocate enough storage to acquire references to each of the semaphores,
 * then we enter the semaphores and release the temporary references.
 */
static void
sem_forkhook(void *arg, struct proc *p1, struct proc *p2, int flags)
{
	struct ksem *ks, **sem_array;
	int count, i, new_count;
	struct kuser *ku;

	mtx_lock(&sem_lock);
	count = sem_count_proc(p1);
	if (count == 0) {
		mtx_unlock(&sem_lock);
		return;
	}
race_lost:
	mtx_assert(&sem_lock, MA_OWNED);
	mtx_unlock(&sem_lock);
	sem_array = malloc(sizeof(struct ksem *) * count, M_TEMP, M_WAITOK);
	mtx_lock(&sem_lock);
	new_count = sem_count_proc(p1);
	if (count < new_count) {
		/* Lost race, repeat and allocate more storage. */
		free(sem_array, M_TEMP);
		count = new_count;
		goto race_lost;
	}
	/*
	 * Given an array capable of storing an adequate number of semaphore
	 * references, now walk the list of semaphores and acquire a new
	 * reference for any semaphore opened by p1.
	 */
	count = new_count;
	i = 0;
	LIST_FOREACH(ks, &ksem_head, ks_entry) {
		LIST_FOREACH(ku, &ks->ks_users, ku_next) {
			if (ku->ku_pid == p1->p_pid) {
				sem_ref(ks);
				sem_array[i] = ks;
				i++;
				break;
			}
		}
	}
	LIST_FOREACH(ks, &ksem_deadhead, ks_entry) {
		LIST_FOREACH(ku, &ks->ks_users, ku_next) {
			if (ku->ku_pid == p1->p_pid) {
				sem_ref(ks);
				sem_array[i] = ks;
				i++;
				break;
			}
		}
	}
	mtx_unlock(&sem_lock);
	KASSERT(i == count, ("sem_forkhook: i != count (%d, %d)", i, count));
	/*
	 * Now cause p2 to enter each of the referenced semaphores, then
	 * release our temporary reference.  This is pretty inefficient.
	 * Finally, free our temporary array.
	 */
	for (i = 0; i < count; i++) {
		sem_enter(p2, sem_array[i]);
		mtx_lock(&sem_lock);
		sem_rel(sem_array[i]);
		mtx_unlock(&sem_lock);
	}
	free(sem_array, M_TEMP);
}

static void
sem_exechook(void *arg, struct proc *p, struct image_params *imgp __unused)
{
   	sem_exithook(arg, p);   	
}

static void
sem_exithook(void *arg, struct proc *p)
{
	struct ksem *ks, *ksnext;

	mtx_lock(&sem_lock);
	ks = LIST_FIRST(&ksem_head);
	while (ks != NULL) {
		ksnext = LIST_NEXT(ks, ks_entry);
		sem_leave(p, ks);
		ks = ksnext;
	}
	ks = LIST_FIRST(&ksem_deadhead);
	while (ks != NULL) {
		ksnext = LIST_NEXT(ks, ks_entry);
		sem_leave(p, ks);
		ks = ksnext;
	}
	mtx_unlock(&sem_lock);
}

static int
sem_modload(struct module *module, int cmd, void *arg)
{
        int error = 0;

        switch (cmd) {
        case MOD_LOAD:
		mtx_init(&sem_lock, "sem", "semaphore", MTX_DEF);
		p31b_setcfg(CTL_P1003_1B_SEM_NSEMS_MAX, SEM_MAX);
		p31b_setcfg(CTL_P1003_1B_SEM_VALUE_MAX, SEM_VALUE_MAX);
		sem_exit_tag = EVENTHANDLER_REGISTER(process_exit, sem_exithook,
		    NULL, EVENTHANDLER_PRI_ANY);
		sem_exec_tag = EVENTHANDLER_REGISTER(process_exec, sem_exechook,
		    NULL, EVENTHANDLER_PRI_ANY);
		sem_fork_tag = EVENTHANDLER_REGISTER(process_fork, sem_forkhook, NULL, EVENTHANDLER_PRI_ANY);
                break;
        case MOD_UNLOAD:
		if (nsems != 0) {
			error = EOPNOTSUPP;
			break;
		}
		EVENTHANDLER_DEREGISTER(process_exit, sem_exit_tag);
		EVENTHANDLER_DEREGISTER(process_exec, sem_exec_tag);
		EVENTHANDLER_DEREGISTER(process_fork, sem_fork_tag);
		mtx_destroy(&sem_lock);
                break;
        case MOD_SHUTDOWN:
                break;
        default:
                error = EINVAL;
                break;
        }
        return (error);
}

static moduledata_t sem_mod = {
        "sem",
        &sem_modload,
        NULL
};

SYSCALL_MODULE_HELPER(ksem_init);
SYSCALL_MODULE_HELPER(ksem_open);
SYSCALL_MODULE_HELPER(ksem_unlink);
SYSCALL_MODULE_HELPER(ksem_close);
SYSCALL_MODULE_HELPER(ksem_post);
SYSCALL_MODULE_HELPER(ksem_wait);
SYSCALL_MODULE_HELPER(ksem_timedwait);
SYSCALL_MODULE_HELPER(ksem_trywait);
SYSCALL_MODULE_HELPER(ksem_getvalue);
SYSCALL_MODULE_HELPER(ksem_destroy);

DECLARE_MODULE(sem, sem_mod, SI_SUB_SYSV_SEM, SI_ORDER_FIRST);
MODULE_VERSION(sem, 1);
OpenPOWER on IntegriCloud