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
|
/* $FreeBSD$ */
/*
* Implementation of SVID semaphores
*
* Author: Daniel Boulet
*
* This software is provided ``AS IS'' without any warranties of any kind.
*/
#include "opt_sysvipc.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/sysproto.h>
#include <sys/kernel.h>
#include <sys/proc.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/sem.h>
#include <sys/syscall.h>
#include <sys/sysent.h>
#include <sys/sysctl.h>
#include <sys/malloc.h>
#include <sys/jail.h>
static MALLOC_DEFINE(M_SEM, "sem", "SVID compatible semaphores");
#ifdef SEM_DEBUG
#define DPRINTF(a) printf a
#else
#define DPRINTF(a)
#endif
static void seminit(void);
static int sysvsem_modload(struct module *, int, void *);
static int semunload(void);
static void semexit_myhook(struct proc *p);
static int sysctl_sema(SYSCTL_HANDLER_ARGS);
static int semvalid(int semid, struct semid_ds *semaptr);
#ifndef _SYS_SYSPROTO_H_
struct __semctl_args;
int __semctl(struct thread *td, struct __semctl_args *uap);
struct semget_args;
int semget(struct thread *td, struct semget_args *uap);
struct semop_args;
int semop(struct thread *td, struct semop_args *uap);
#endif
static struct sem_undo *semu_alloc(struct thread *td);
static int semundo_adjust(struct thread *td, struct sem_undo **supptr,
int semid, int semnum, int adjval);
static void semundo_clear(int semid, int semnum);
/* XXX casting to (sy_call_t *) is bogus, as usual. */
static sy_call_t *semcalls[] = {
(sy_call_t *)__semctl, (sy_call_t *)semget,
(sy_call_t *)semop
};
static struct mtx sem_mtx; /* semaphore global lock */
static int semtot = 0;
static struct semid_ds *sema; /* semaphore id pool */
static struct mtx *sema_mtx; /* semaphore id pool mutexes*/
static struct sem *sem; /* semaphore pool */
SLIST_HEAD(, sem_undo) semu_list; /* list of active undo structures */
static int *semu; /* undo structure pool */
#define SEMUNDO_MTX sem_mtx
#define SEMUNDO_LOCK() mtx_lock(&SEMUNDO_MTX);
#define SEMUNDO_UNLOCK() mtx_unlock(&SEMUNDO_MTX);
#define SEMUNDO_LOCKASSERT(how) mtx_assert(&SEMUNDO_MTX, (how));
struct sem {
u_short semval; /* semaphore value */
pid_t sempid; /* pid of last operation */
u_short semncnt; /* # awaiting semval > cval */
u_short semzcnt; /* # awaiting semval = 0 */
};
/*
* Undo structure (one per process)
*/
struct sem_undo {
SLIST_ENTRY(sem_undo) un_next; /* ptr to next active undo structure */
struct proc *un_proc; /* owner of this structure */
short un_cnt; /* # of active entries */
struct undo {
short un_adjval; /* adjust on exit values */
short un_num; /* semaphore # */
int un_id; /* semid */
} un_ent[1]; /* undo entries */
};
/*
* Configuration parameters
*/
#ifndef SEMMNI
#define SEMMNI 10 /* # of semaphore identifiers */
#endif
#ifndef SEMMNS
#define SEMMNS 60 /* # of semaphores in system */
#endif
#ifndef SEMUME
#define SEMUME 10 /* max # of undo entries per process */
#endif
#ifndef SEMMNU
#define SEMMNU 30 /* # of undo structures in system */
#endif
/* shouldn't need tuning */
#ifndef SEMMAP
#define SEMMAP 30 /* # of entries in semaphore map */
#endif
#ifndef SEMMSL
#define SEMMSL SEMMNS /* max # of semaphores per id */
#endif
#ifndef SEMOPM
#define SEMOPM 100 /* max # of operations per semop call */
#endif
#define SEMVMX 32767 /* semaphore maximum value */
#define SEMAEM 16384 /* adjust on exit max value */
/*
* Due to the way semaphore memory is allocated, we have to ensure that
* SEMUSZ is properly aligned.
*/
#define SEM_ALIGN(bytes) (((bytes) + (sizeof(long) - 1)) & ~(sizeof(long) - 1))
/* actual size of an undo structure */
#define SEMUSZ SEM_ALIGN(offsetof(struct sem_undo, un_ent[SEMUME]))
/*
* Macro to find a particular sem_undo vector
*/
#define SEMU(ix) \
((struct sem_undo *)(((intptr_t)semu)+ix * seminfo.semusz))
/*
* semaphore info struct
*/
struct seminfo seminfo = {
SEMMAP, /* # of entries in semaphore map */
SEMMNI, /* # of semaphore identifiers */
SEMMNS, /* # of semaphores in system */
SEMMNU, /* # of undo structures in system */
SEMMSL, /* max # of semaphores per id */
SEMOPM, /* max # of operations per semop call */
SEMUME, /* max # of undo entries per process */
SEMUSZ, /* size in bytes of undo structure */
SEMVMX, /* semaphore maximum value */
SEMAEM /* adjust on exit max value */
};
SYSCTL_DECL(_kern_ipc);
SYSCTL_INT(_kern_ipc, OID_AUTO, semmap, CTLFLAG_RW, &seminfo.semmap, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semmni, CTLFLAG_RD, &seminfo.semmni, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semmns, CTLFLAG_RD, &seminfo.semmns, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semmnu, CTLFLAG_RD, &seminfo.semmnu, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semmsl, CTLFLAG_RW, &seminfo.semmsl, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semopm, CTLFLAG_RD, &seminfo.semopm, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semume, CTLFLAG_RD, &seminfo.semume, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semusz, CTLFLAG_RD, &seminfo.semusz, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semvmx, CTLFLAG_RW, &seminfo.semvmx, 0, "");
SYSCTL_INT(_kern_ipc, OID_AUTO, semaem, CTLFLAG_RW, &seminfo.semaem, 0, "");
SYSCTL_PROC(_kern_ipc, OID_AUTO, sema, CTLFLAG_RD,
NULL, 0, sysctl_sema, "", "");
static void
seminit(void)
{
int i;
TUNABLE_INT_FETCH("kern.ipc.semmap", &seminfo.semmap);
TUNABLE_INT_FETCH("kern.ipc.semmni", &seminfo.semmni);
TUNABLE_INT_FETCH("kern.ipc.semmns", &seminfo.semmns);
TUNABLE_INT_FETCH("kern.ipc.semmnu", &seminfo.semmnu);
TUNABLE_INT_FETCH("kern.ipc.semmsl", &seminfo.semmsl);
TUNABLE_INT_FETCH("kern.ipc.semopm", &seminfo.semopm);
TUNABLE_INT_FETCH("kern.ipc.semume", &seminfo.semume);
TUNABLE_INT_FETCH("kern.ipc.semusz", &seminfo.semusz);
TUNABLE_INT_FETCH("kern.ipc.semvmx", &seminfo.semvmx);
TUNABLE_INT_FETCH("kern.ipc.semaem", &seminfo.semaem);
sem = malloc(sizeof(struct sem) * seminfo.semmns, M_SEM, 0);
sema = malloc(sizeof(struct semid_ds) * seminfo.semmni, M_SEM,
0);
sema_mtx = malloc(sizeof(struct mtx) * seminfo.semmni, M_SEM,
M_ZERO);
semu = malloc(seminfo.semmnu * seminfo.semusz, M_SEM, 0);
for (i = 0; i < seminfo.semmni; i++) {
sema[i].sem_base = 0;
sema[i].sem_perm.mode = 0;
}
for (i = 0; i < seminfo.semmni; i++)
mtx_init(&sema_mtx[i], "semid", NULL, MTX_DEF);
for (i = 0; i < seminfo.semmnu; i++) {
struct sem_undo *suptr = SEMU(i);
suptr->un_proc = NULL;
}
SLIST_INIT(&semu_list);
at_exit(semexit_myhook);
mtx_init(&sem_mtx, "sem", NULL, MTX_DEF);
}
static int
semunload(void)
{
int i;
if (semtot != 0)
return (EBUSY);
free(sem, M_SEM);
free(sema, M_SEM);
free(semu, M_SEM);
rm_at_exit(semexit_myhook);
for (i = 0; i < seminfo.semmni; i++)
mtx_destroy(&sema_mtx[i]);
mtx_destroy(&sem_mtx);
return (0);
}
static int
sysvsem_modload(struct module *module, int cmd, void *arg)
{
int error = 0;
switch (cmd) {
case MOD_LOAD:
seminit();
break;
case MOD_UNLOAD:
error = semunload();
break;
case MOD_SHUTDOWN:
break;
default:
error = EINVAL;
break;
}
return (error);
}
static moduledata_t sysvsem_mod = {
"sysvsem",
&sysvsem_modload,
NULL
};
SYSCALL_MODULE_HELPER(semsys);
SYSCALL_MODULE_HELPER(__semctl);
SYSCALL_MODULE_HELPER(semget);
SYSCALL_MODULE_HELPER(semop);
DECLARE_MODULE(sysvsem, sysvsem_mod,
SI_SUB_SYSV_SEM, SI_ORDER_FIRST);
MODULE_VERSION(sysvsem, 1);
/*
* Entry point for all SEM calls
*
* MPSAFE
*/
int
semsys(td, uap)
struct thread *td;
/* XXX actually varargs. */
struct semsys_args /* {
u_int which;
int a2;
int a3;
int a4;
int a5;
} */ *uap;
{
int error;
if (!jail_sysvipc_allowed && jailed(td->td_ucred))
return (ENOSYS);
if (uap->which >= sizeof(semcalls)/sizeof(semcalls[0]))
return (EINVAL);
error = (*semcalls[uap->which])(td, &uap->a2);
return (error);
}
/*
* Allocate a new sem_undo structure for a process
* (returns ptr to structure or NULL if no more room)
*/
static struct sem_undo *
semu_alloc(td)
struct thread *td;
{
int i;
struct sem_undo *suptr;
struct sem_undo **supptr;
int attempt;
SEMUNDO_LOCKASSERT(MA_OWNED);
/*
* Try twice to allocate something.
* (we'll purge any empty structures after the first pass so
* two passes are always enough)
*/
for (attempt = 0; attempt < 2; attempt++) {
/*
* Look for a free structure.
* Fill it in and return it if we find one.
*/
for (i = 0; i < seminfo.semmnu; i++) {
suptr = SEMU(i);
if (suptr->un_proc == NULL) {
SLIST_INSERT_HEAD(&semu_list, suptr, un_next);
suptr->un_cnt = 0;
suptr->un_proc = td->td_proc;
return(suptr);
}
}
/*
* We didn't find a free one, if this is the first attempt
* then try to free some structures.
*/
if (attempt == 0) {
/* All the structures are in use - try to free some */
int did_something = 0;
SLIST_FOREACH_PREVPTR(suptr, supptr, &semu_list,
un_next) {
if (suptr->un_cnt == 0) {
suptr->un_proc = NULL;
did_something = 1;
*supptr = SLIST_NEXT(suptr, un_next);
}
}
/* If we didn't free anything then just give-up */
if (!did_something)
return(NULL);
} else {
/*
* The second pass failed even though we freed
* something after the first pass!
* This is IMPOSSIBLE!
*/
panic("semu_alloc - second attempt failed");
}
}
return (NULL);
}
/*
* Adjust a particular entry for a particular proc
*/
static int
semundo_adjust(td, supptr, semid, semnum, adjval)
struct thread *td;
struct sem_undo **supptr;
int semid, semnum;
int adjval;
{
struct proc *p = td->td_proc;
struct sem_undo *suptr;
struct undo *sunptr;
int i;
SEMUNDO_LOCKASSERT(MA_OWNED);
/* Look for and remember the sem_undo if the caller doesn't provide
it */
suptr = *supptr;
if (suptr == NULL) {
SLIST_FOREACH(suptr, &semu_list, un_next) {
if (suptr->un_proc == p) {
*supptr = suptr;
break;
}
}
if (suptr == NULL) {
if (adjval == 0)
return(0);
suptr = semu_alloc(td);
if (suptr == NULL)
return(ENOSPC);
*supptr = suptr;
}
}
/*
* Look for the requested entry and adjust it (delete if adjval becomes
* 0).
*/
sunptr = &suptr->un_ent[0];
for (i = 0; i < suptr->un_cnt; i++, sunptr++) {
if (sunptr->un_id != semid || sunptr->un_num != semnum)
continue;
if (adjval != 0) {
adjval += sunptr->un_adjval;
if (adjval > seminfo.semaem || adjval < -seminfo.semaem)
return (ERANGE);
}
sunptr->un_adjval = adjval;
if (sunptr->un_adjval == 0) {
suptr->un_cnt--;
if (i < suptr->un_cnt)
suptr->un_ent[i] =
suptr->un_ent[suptr->un_cnt];
}
return(0);
}
/* Didn't find the right entry - create it */
if (adjval == 0)
return(0);
if (adjval > seminfo.semaem || adjval < -seminfo.semaem)
return (ERANGE);
if (suptr->un_cnt != seminfo.semume) {
sunptr = &suptr->un_ent[suptr->un_cnt];
suptr->un_cnt++;
sunptr->un_adjval = adjval;
sunptr->un_id = semid; sunptr->un_num = semnum;
} else
return(EINVAL);
return(0);
}
static void
semundo_clear(semid, semnum)
int semid, semnum;
{
struct sem_undo *suptr;
SEMUNDO_LOCKASSERT(MA_OWNED);
SLIST_FOREACH(suptr, &semu_list, un_next) {
struct undo *sunptr = &suptr->un_ent[0];
int i = 0;
while (i < suptr->un_cnt) {
if (sunptr->un_id == semid) {
if (semnum == -1 || sunptr->un_num == semnum) {
suptr->un_cnt--;
if (i < suptr->un_cnt) {
suptr->un_ent[i] =
suptr->un_ent[suptr->un_cnt];
continue;
}
}
if (semnum != -1)
break;
}
i++, sunptr++;
}
}
}
static int
semvalid(semid, semaptr)
int semid;
struct semid_ds *semaptr;
{
return ((semaptr->sem_perm.mode & SEM_ALLOC) == 0 ||
semaptr->sem_perm.seq != IPCID_TO_SEQ(semid) ? EINVAL : 0);
}
/*
* Note that the user-mode half of this passes a union, not a pointer
*/
#ifndef _SYS_SYSPROTO_H_
struct __semctl_args {
int semid;
int semnum;
int cmd;
union semun *arg;
};
#endif
/*
* MPSAFE
*/
int
__semctl(td, uap)
struct thread *td;
struct __semctl_args *uap;
{
int semid = uap->semid;
int semnum = uap->semnum;
int cmd = uap->cmd;
u_short *array;
union semun *arg = uap->arg;
union semun real_arg;
struct ucred *cred = td->td_ucred;
int i, rval, error;
struct semid_ds sbuf;
struct semid_ds *semaptr;
struct mtx *sema_mtxp;
u_short usval, count;
DPRINTF(("call to semctl(%d, %d, %d, 0x%x)\n",
semid, semnum, cmd, arg));
if (!jail_sysvipc_allowed && jailed(td->td_ucred))
return (ENOSYS);
array = NULL;
switch(cmd) {
case SEM_STAT:
if (semid < 0 || semid >= seminfo.semmni)
return (EINVAL);
if ((error = copyin(arg, &real_arg, sizeof(real_arg))) != 0)
return (error);
semaptr = &sema[semid];
sema_mtxp = &sema_mtx[semid];
mtx_lock(sema_mtxp);
if ((semaptr->sem_perm.mode & SEM_ALLOC) == 0 ) {
error = EINVAL;
goto done2;
}
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
mtx_unlock(sema_mtxp);
error = copyout(semaptr, real_arg.buf, sizeof(struct semid_ds));
rval = IXSEQ_TO_IPCID(semid,semaptr->sem_perm);
if (error == 0)
td->td_retval[0] = rval;
return (error);
}
semid = IPCID_TO_IX(semid);
if (semid < 0 || semid >= seminfo.semmni)
return (EINVAL);
semaptr = &sema[semid];
sema_mtxp = &sema_mtx[semid];
error = 0;
rval = 0;
switch (cmd) {
case IPC_RMID:
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_M)))
goto done2;
semaptr->sem_perm.cuid = cred->cr_uid;
semaptr->sem_perm.uid = cred->cr_uid;
semtot -= semaptr->sem_nsems;
for (i = semaptr->sem_base - sem; i < semtot; i++)
sem[i] = sem[i + semaptr->sem_nsems];
for (i = 0; i < seminfo.semmni; i++) {
if ((sema[i].sem_perm.mode & SEM_ALLOC) &&
sema[i].sem_base > semaptr->sem_base)
sema[i].sem_base -= semaptr->sem_nsems;
}
semaptr->sem_perm.mode = 0;
SEMUNDO_LOCK();
semundo_clear(semid, -1);
SEMUNDO_UNLOCK();
wakeup(semaptr);
break;
case IPC_SET:
if ((error = copyin(arg, &real_arg, sizeof(real_arg))) != 0)
goto done2;
if ((error = copyin(real_arg.buf, &sbuf, sizeof(sbuf))) != 0)
goto done2;
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_M)))
goto done2;
semaptr->sem_perm.uid = sbuf.sem_perm.uid;
semaptr->sem_perm.gid = sbuf.sem_perm.gid;
semaptr->sem_perm.mode = (semaptr->sem_perm.mode & ~0777) |
(sbuf.sem_perm.mode & 0777);
semaptr->sem_ctime = time_second;
break;
case IPC_STAT:
if ((error = copyin(arg, &real_arg, sizeof(real_arg))) != 0)
goto done2;
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
sbuf = *semaptr;
mtx_unlock(sema_mtxp);
error = copyout(semaptr, real_arg.buf,
sizeof(struct semid_ds));
break;
case GETNCNT:
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
if (semnum < 0 || semnum >= semaptr->sem_nsems) {
error = EINVAL;
goto done2;
}
rval = semaptr->sem_base[semnum].semncnt;
break;
case GETPID:
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
if (semnum < 0 || semnum >= semaptr->sem_nsems) {
error = EINVAL;
goto done2;
}
rval = semaptr->sem_base[semnum].sempid;
break;
case GETVAL:
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
if (semnum < 0 || semnum >= semaptr->sem_nsems) {
error = EINVAL;
goto done2;
}
rval = semaptr->sem_base[semnum].semval;
break;
case GETALL:
if ((error = copyin(arg, &real_arg, sizeof(real_arg))) != 0)
goto done2;
array = malloc(sizeof(*array) * semaptr->sem_nsems, M_TEMP,
0);
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
for (i = 0; i < semaptr->sem_nsems; i++)
array[i] = semaptr->sem_base[i].semval;
mtx_unlock(sema_mtxp);
error = copyout(array, real_arg.array,
i * sizeof(real_arg.array[0]));
break;
case GETZCNT:
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_R)))
goto done2;
if (semnum < 0 || semnum >= semaptr->sem_nsems) {
error = EINVAL;
goto done2;
}
rval = semaptr->sem_base[semnum].semzcnt;
break;
case SETVAL:
if ((error = copyin(arg, &real_arg, sizeof(real_arg))) != 0)
goto done2;
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_W)))
goto done2;
if (semnum < 0 || semnum >= semaptr->sem_nsems) {
error = EINVAL;
goto done2;
}
if (real_arg.val < 0 || real_arg.val > seminfo.semvmx) {
error = ERANGE;
goto done2;
}
semaptr->sem_base[semnum].semval = real_arg.val;
SEMUNDO_LOCK();
semundo_clear(semid, semnum);
SEMUNDO_UNLOCK();
wakeup(semaptr);
break;
case SETALL:
mtx_lock(sema_mtxp);
raced:
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
count = semaptr->sem_nsems;
mtx_unlock(sema_mtxp);
if ((error = copyin(arg, &real_arg, sizeof(real_arg))) != 0)
goto done2;
array = malloc(sizeof(*array) * count, M_TEMP, 0);
copyin(real_arg.array, array, count * sizeof(*array));
if (error)
break;
mtx_lock(sema_mtxp);
if ((error = semvalid(uap->semid, semaptr)) != 0)
goto done2;
/* we could have raced? */
if (count != semaptr->sem_nsems) {
free(array, M_TEMP);
array = NULL;
goto raced;
}
if ((error = ipcperm(td, &semaptr->sem_perm, IPC_W)))
goto done2;
for (i = 0; i < semaptr->sem_nsems; i++) {
usval = array[i];
if (usval > seminfo.semvmx) {
error = ERANGE;
break;
}
semaptr->sem_base[i].semval = usval;
}
SEMUNDO_LOCK();
semundo_clear(semid, -1);
SEMUNDO_UNLOCK();
wakeup(semaptr);
break;
default:
error = EINVAL;
break;
}
if (error == 0)
td->td_retval[0] = rval;
done2:
if (mtx_owned(sema_mtxp))
mtx_unlock(sema_mtxp);
if (array != NULL)
free(array, M_TEMP);
return(error);
}
#ifndef _SYS_SYSPROTO_H_
struct semget_args {
key_t key;
int nsems;
int semflg;
};
#endif
/*
* MPSAFE
*/
int
semget(td, uap)
struct thread *td;
struct semget_args *uap;
{
int semid, error = 0;
int key = uap->key;
int nsems = uap->nsems;
int semflg = uap->semflg;
struct ucred *cred = td->td_ucred;
DPRINTF(("semget(0x%x, %d, 0%o)\n", key, nsems, semflg));
if (!jail_sysvipc_allowed && jailed(td->td_ucred))
return (ENOSYS);
mtx_lock(&Giant);
if (key != IPC_PRIVATE) {
for (semid = 0; semid < seminfo.semmni; semid++) {
if ((sema[semid].sem_perm.mode & SEM_ALLOC) &&
sema[semid].sem_perm.key == key)
break;
}
if (semid < seminfo.semmni) {
DPRINTF(("found public key\n"));
if ((error = ipcperm(td, &sema[semid].sem_perm,
semflg & 0700))) {
goto done2;
}
if (nsems > 0 && sema[semid].sem_nsems < nsems) {
DPRINTF(("too small\n"));
error = EINVAL;
goto done2;
}
if ((semflg & IPC_CREAT) && (semflg & IPC_EXCL)) {
DPRINTF(("not exclusive\n"));
error = EEXIST;
goto done2;
}
goto found;
}
}
DPRINTF(("need to allocate the semid_ds\n"));
if (key == IPC_PRIVATE || (semflg & IPC_CREAT)) {
if (nsems <= 0 || nsems > seminfo.semmsl) {
DPRINTF(("nsems out of range (0<%d<=%d)\n", nsems,
seminfo.semmsl));
error = EINVAL;
goto done2;
}
if (nsems > seminfo.semmns - semtot) {
DPRINTF((
"not enough semaphores left (need %d, got %d)\n",
nsems, seminfo.semmns - semtot));
error = ENOSPC;
goto done2;
}
for (semid = 0; semid < seminfo.semmni; semid++) {
if ((sema[semid].sem_perm.mode & SEM_ALLOC) == 0)
break;
}
if (semid == seminfo.semmni) {
DPRINTF(("no more semid_ds's available\n"));
error = ENOSPC;
goto done2;
}
DPRINTF(("semid %d is available\n", semid));
sema[semid].sem_perm.key = key;
sema[semid].sem_perm.cuid = cred->cr_uid;
sema[semid].sem_perm.uid = cred->cr_uid;
sema[semid].sem_perm.cgid = cred->cr_gid;
sema[semid].sem_perm.gid = cred->cr_gid;
sema[semid].sem_perm.mode = (semflg & 0777) | SEM_ALLOC;
sema[semid].sem_perm.seq =
(sema[semid].sem_perm.seq + 1) & 0x7fff;
sema[semid].sem_nsems = nsems;
sema[semid].sem_otime = 0;
sema[semid].sem_ctime = time_second;
sema[semid].sem_base = &sem[semtot];
semtot += nsems;
bzero(sema[semid].sem_base,
sizeof(sema[semid].sem_base[0])*nsems);
DPRINTF(("sembase = 0x%x, next = 0x%x\n", sema[semid].sem_base,
&sem[semtot]));
} else {
DPRINTF(("didn't find it and wasn't asked to create it\n"));
error = ENOENT;
goto done2;
}
found:
td->td_retval[0] = IXSEQ_TO_IPCID(semid, sema[semid].sem_perm);
done2:
mtx_unlock(&Giant);
return (error);
}
#ifndef _SYS_SYSPROTO_H_
struct semop_args {
int semid;
struct sembuf *sops;
size_t nsops;
};
#endif
/*
* MPSAFE
*/
int
semop(td, uap)
struct thread *td;
struct semop_args *uap;
{
int semid = uap->semid;
size_t nsops = uap->nsops;
struct sembuf *sops;
struct semid_ds *semaptr;
struct sembuf *sopptr = 0;
struct sem *semptr = 0;
struct sem_undo *suptr;
struct mtx *sema_mtxp;
size_t i, j;
int error;
int do_wakeup, do_undos;
DPRINTF(("call to semop(%d, 0x%x, %u)\n", semid, sops, nsops));
if (!jail_sysvipc_allowed && jailed(td->td_ucred))
return (ENOSYS);
semid = IPCID_TO_IX(semid); /* Convert back to zero origin */
if (semid < 0 || semid >= seminfo.semmni)
return (EINVAL);
/* Allocate memory for sem_ops */
if (nsops > seminfo.semopm) {
DPRINTF(("too many sops (max=%d, nsops=%d)\n", seminfo.semopm,
nsops));
return (E2BIG);
}
sops = malloc(nsops * sizeof(sops[0]), M_SEM, 0);
if ((error = copyin(uap->sops, sops, nsops * sizeof(sops[0]))) != 0) {
DPRINTF(("error = %d from copyin(%08x, %08x, %d)\n", error,
uap->sops, sops, nsops * sizeof(sops[0])));
free(sops, M_SEM);
return (error);
}
semaptr = &sema[semid];
sema_mtxp = &sema_mtx[semid];
mtx_lock(sema_mtxp);
if ((semaptr->sem_perm.mode & SEM_ALLOC) == 0) {
error = EINVAL;
goto done2;
}
if (semaptr->sem_perm.seq != IPCID_TO_SEQ(uap->semid)) {
error = EINVAL;
goto done2;
}
/*
* Initial pass thru sops to see what permissions are needed.
* Also perform any checks that don't need repeating on each
* attempt to satisfy the request vector.
*/
j = 0; /* permission needed */
do_undos = 0;
for (i = 0; i < nsops; i++) {
sopptr = &sops[i];
if (sopptr->sem_num >= semaptr->sem_nsems) {
error = EFBIG;
goto done2;
}
if (sopptr->sem_flg & SEM_UNDO && sopptr->sem_op != 0)
do_undos = 1;
j |= (sopptr->sem_op == 0) ? SEM_R : SEM_A;
}
if ((error = ipcperm(td, &semaptr->sem_perm, j))) {
DPRINTF(("error = %d from ipaccess\n", error));
goto done2;
}
/*
* Loop trying to satisfy the vector of requests.
* If we reach a point where we must wait, any requests already
* performed are rolled back and we go to sleep until some other
* process wakes us up. At this point, we start all over again.
*
* This ensures that from the perspective of other tasks, a set
* of requests is atomic (never partially satisfied).
*/
for (;;) {
do_wakeup = 0;
error = 0; /* error return if necessary */
for (i = 0; i < nsops; i++) {
sopptr = &sops[i];
semptr = &semaptr->sem_base[sopptr->sem_num];
DPRINTF((
"semop: semaptr=%x, sem_base=%x, "
"semptr=%x, sem[%d]=%d : op=%d, flag=%s\n",
semaptr, semaptr->sem_base, semptr,
sopptr->sem_num, semptr->semval, sopptr->sem_op,
(sopptr->sem_flg & IPC_NOWAIT) ?
"nowait" : "wait"));
if (sopptr->sem_op < 0) {
if (semptr->semval + sopptr->sem_op < 0) {
DPRINTF(("semop: can't do it now\n"));
break;
} else {
semptr->semval += sopptr->sem_op;
if (semptr->semval == 0 &&
semptr->semzcnt > 0)
do_wakeup = 1;
}
} else if (sopptr->sem_op == 0) {
if (semptr->semval != 0) {
DPRINTF(("semop: not zero now\n"));
break;
}
} else if (semptr->semval + sopptr->sem_op >
seminfo.semvmx) {
error = ERANGE;
break;
} else {
if (semptr->semncnt > 0)
do_wakeup = 1;
semptr->semval += sopptr->sem_op;
}
}
/*
* Did we get through the entire vector?
*/
if (i >= nsops)
goto done;
/*
* No ... rollback anything that we've already done
*/
DPRINTF(("semop: rollback 0 through %d\n", i-1));
for (j = 0; j < i; j++)
semaptr->sem_base[sops[j].sem_num].semval -=
sops[j].sem_op;
/* If we detected an error, return it */
if (error != 0)
goto done2;
/*
* If the request that we couldn't satisfy has the
* NOWAIT flag set then return with EAGAIN.
*/
if (sopptr->sem_flg & IPC_NOWAIT) {
error = EAGAIN;
goto done2;
}
if (sopptr->sem_op == 0)
semptr->semzcnt++;
else
semptr->semncnt++;
DPRINTF(("semop: good night!\n"));
error = msleep(semaptr, sema_mtxp, (PZERO - 4) | PCATCH,
"semwait", 0);
DPRINTF(("semop: good morning (error=%d)!\n", error));
if (error != 0) {
error = EINTR;
goto done2;
}
DPRINTF(("semop: good morning!\n"));
/*
* Make sure that the semaphore still exists
*/
if ((semaptr->sem_perm.mode & SEM_ALLOC) == 0 ||
semaptr->sem_perm.seq != IPCID_TO_SEQ(uap->semid)) {
error = EIDRM;
goto done2;
}
/*
* The semaphore is still alive. Readjust the count of
* waiting processes.
*/
if (sopptr->sem_op == 0)
semptr->semzcnt--;
else
semptr->semncnt--;
}
done:
/*
* Process any SEM_UNDO requests.
*/
if (do_undos) {
SEMUNDO_LOCK();
suptr = NULL;
for (i = 0; i < nsops; i++) {
/*
* We only need to deal with SEM_UNDO's for non-zero
* op's.
*/
int adjval;
if ((sops[i].sem_flg & SEM_UNDO) == 0)
continue;
adjval = sops[i].sem_op;
if (adjval == 0)
continue;
error = semundo_adjust(td, &suptr, semid,
sops[i].sem_num, -adjval);
if (error == 0)
continue;
/*
* Oh-Oh! We ran out of either sem_undo's or undo's.
* Rollback the adjustments to this point and then
* rollback the semaphore ups and down so we can return
* with an error with all structures restored. We
* rollback the undo's in the exact reverse order that
* we applied them. This guarantees that we won't run
* out of space as we roll things back out.
*/
for (j = i - 1; j >= 0; j--) {
if ((sops[j].sem_flg & SEM_UNDO) == 0)
continue;
adjval = sops[j].sem_op;
if (adjval == 0)
continue;
if (semundo_adjust(td, &suptr, semid,
sops[j].sem_num, adjval) != 0)
panic("semop - can't undo undos");
}
for (j = 0; j < nsops; j++)
semaptr->sem_base[sops[j].sem_num].semval -=
sops[j].sem_op;
DPRINTF(("error = %d from semundo_adjust\n", error));
SEMUNDO_UNLOCK();
goto done2;
} /* loop through the sops */
SEMUNDO_UNLOCK();
} /* if (do_undos) */
/* We're definitely done - set the sempid's and time */
for (i = 0; i < nsops; i++) {
sopptr = &sops[i];
semptr = &semaptr->sem_base[sopptr->sem_num];
semptr->sempid = td->td_proc->p_pid;
}
semaptr->sem_otime = time_second;
/*
* Do a wakeup if any semaphore was up'd whilst something was
* sleeping on it.
*/
if (do_wakeup) {
DPRINTF(("semop: doing wakeup\n"));
wakeup(semaptr);
DPRINTF(("semop: back from wakeup\n"));
}
DPRINTF(("semop: done\n"));
td->td_retval[0] = 0;
done2:
mtx_unlock(sema_mtxp);
free(sops, M_SEM);
return (error);
}
/*
* Go through the undo structures for this process and apply the adjustments to
* semaphores.
*/
static void
semexit_myhook(p)
struct proc *p;
{
struct sem_undo *suptr;
struct sem_undo **supptr;
/*
* Go through the chain of undo vectors looking for one
* associated with this process.
*/
SEMUNDO_LOCK();
SLIST_FOREACH_PREVPTR(suptr, supptr, &semu_list, un_next) {
if (suptr->un_proc == p)
break;
}
SEMUNDO_UNLOCK();
if (suptr == NULL)
return;
DPRINTF(("proc @%08x has undo structure with %d entries\n", p,
suptr->un_cnt));
/*
* If there are any active undo elements then process them.
*/
if (suptr->un_cnt > 0) {
int ix;
for (ix = 0; ix < suptr->un_cnt; ix++) {
int semid = suptr->un_ent[ix].un_id;
int semnum = suptr->un_ent[ix].un_num;
int adjval = suptr->un_ent[ix].un_adjval;
struct semid_ds *semaptr;
struct mtx *sema_mtxp;
semaptr = &sema[semid];
sema_mtxp = &sema_mtx[semid];
mtx_lock(sema_mtxp);
SEMUNDO_LOCK();
if ((semaptr->sem_perm.mode & SEM_ALLOC) == 0)
panic("semexit - semid not allocated");
if (semnum >= semaptr->sem_nsems)
panic("semexit - semnum out of range");
DPRINTF((
"semexit: %08x id=%d num=%d(adj=%d) ; sem=%d\n",
suptr->un_proc, suptr->un_ent[ix].un_id,
suptr->un_ent[ix].un_num,
suptr->un_ent[ix].un_adjval,
semaptr->sem_base[semnum].semval));
if (adjval < 0) {
if (semaptr->sem_base[semnum].semval < -adjval)
semaptr->sem_base[semnum].semval = 0;
else
semaptr->sem_base[semnum].semval +=
adjval;
} else
semaptr->sem_base[semnum].semval += adjval;
wakeup(semaptr);
DPRINTF(("semexit: back from wakeup\n"));
mtx_unlock(sema_mtxp);
SEMUNDO_UNLOCK();
}
}
/*
* Deallocate the undo vector.
*/
DPRINTF(("removing vector\n"));
suptr->un_proc = NULL;
*supptr = SLIST_NEXT(suptr, un_next);
}
static int
sysctl_sema(SYSCTL_HANDLER_ARGS)
{
return (SYSCTL_OUT(req, sema,
sizeof(struct semid_ds) * seminfo.semmni));
}
|