forked from apple/darwin-xnu
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsys_pipe.c
1817 lines (1542 loc) · 46 KB
/
sys_pipe.c
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
/*
* Copyright (c) 1996 John S. Dyson
* 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 immediately at the beginning of the file, without modification,
* this list of conditions, and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Absolutely no warranty of function or purpose is made by the author
* John S. Dyson.
* 4. Modifications may be freely made to this file if the above conditions
* are met.
*/
/*
* Copyright (c) 2003-2014 Apple Inc. All rights reserved.
*
* @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
* This file contains Original Code and/or Modifications of Original Code
* as defined in and that are subject to the Apple Public Source License
* Version 2.0 (the 'License'). You may not use this file except in
* compliance with the License. The rights granted to you under the License
* may not be used to create, or enable the creation or redistribution of,
* unlawful or unlicensed copies of an Apple operating system, or to
* circumvent, violate, or enable the circumvention or violation of, any
* terms of an Apple operating system software license agreement.
*
* Please obtain a copy of the License at
* http://www.opensource.apple.com/apsl/ and read it before using this file.
*
* The Original Code and all software distributed under the License are
* distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
* Please see the License for the specific language governing rights and
* limitations under the License.
*
* @APPLE_OSREFERENCE_LICENSE_HEADER_END@
*/
/*
* NOTICE: This file was modified by SPARTA, Inc. in 2005 to introduce
* support for mandatory and extensible security protections. This notice
* is included in support of clause 2.2 (b) of the Apple Public License,
* Version 2.0.
*/
/*
* This file contains a high-performance replacement for the socket-based
* pipes scheme originally used in FreeBSD/4.4Lite. It does not support
* all features of sockets, but does do everything that pipes normally
* do.
*
* Pipes are implemented as circular buffers. Following are the valid states in pipes operations
*
* _________________________________
* 1. |_________________________________| r=w, c=0
*
* _________________________________
* 2. |__r:::::wc_______________________| r <= w , c > 0
*
* _________________________________
* 3. |::::wc_____r:::::::::::::::::::::| r>w , c > 0
*
* _________________________________
* 4. |:::::::wrc:::::::::::::::::::::::| w=r, c = Max size
*
*
* Nomenclature:-
* a-z define the steps in a program flow
* 1-4 are the states as defined aboe
* Action: is what file operation is done on the pipe
*
* Current:None Action: initialize with size M=200
* a. State 1 ( r=0, w=0, c=0)
*
* Current: a Action: write(100) (w < M)
* b. State 2 (r=0, w=100, c=100)
*
* Current: b Action: write(100) (w = M-w)
* c. State 4 (r=0,w=0,c=200)
*
* Current: b Action: read(70) ( r < c )
* d. State 2(r=70,w=100,c=30)
*
* Current: d Action: write(75) ( w < (m-w))
* e. State 2 (r=70,w=175,c=105)
*
* Current: d Action: write(110) ( w > (m-w))
* f. State 3 (r=70,w=10,c=140)
*
* Current: d Action: read(30) (r >= c )
* g. State 1 (r=100,w=100,c=0)
*
*/
/*
* This code create half duplex pipe buffers for facilitating file like
* operations on pipes. The initial buffer is very small, but this can
* dynamically change to larger sizes based on usage. The buffer size is never
* reduced. The total amount of kernel memory used is governed by maxpipekva.
* In case of dynamic expansion limit is reached, the output thread is blocked
* until the pipe buffer empties enough to continue.
*
* In order to limit the resource use of pipes, two sysctls exist:
*
* kern.ipc.maxpipekva - This is a hard limit on the amount of pageable
* address space available to us in pipe_map.
*
* Memory usage may be monitored through the sysctls
* kern.ipc.pipes, kern.ipc.pipekva.
*
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/filedesc.h>
#include <sys/kernel.h>
#include <sys/vnode.h>
#include <sys/proc_internal.h>
#include <sys/kauth.h>
#include <sys/file_internal.h>
#include <sys/stat.h>
#include <sys/ioctl.h>
#include <sys/fcntl.h>
#include <sys/malloc.h>
#include <sys/syslog.h>
#include <sys/unistd.h>
#include <sys/resourcevar.h>
#include <sys/aio_kern.h>
#include <sys/signalvar.h>
#include <sys/pipe.h>
#include <sys/sysproto.h>
#include <sys/proc_info.h>
#include <security/audit/audit.h>
#include <sys/kdebug.h>
#include <kern/zalloc.h>
#include <kern/kalloc.h>
#include <vm/vm_kern.h>
#include <libkern/OSAtomic.h>
#include <libkern/section_keywords.h>
#if CONFIG_MACF
#include <security/mac_framework.h>
#endif
#define f_flag f_fglob->fg_flag
#define f_msgcount f_fglob->fg_msgcount
#define f_cred f_fglob->fg_cred
#define f_ops f_fglob->fg_ops
#define f_offset f_fglob->fg_offset
#define f_data f_fglob->fg_data
/*
* interfaces to the outside world exported through file operations
*/
static int pipe_read(struct fileproc *fp, struct uio *uio,
int flags, vfs_context_t ctx);
static int pipe_write(struct fileproc *fp, struct uio *uio,
int flags, vfs_context_t ctx);
static int pipe_close(struct fileglob *fg, vfs_context_t ctx);
static int pipe_select(struct fileproc *fp, int which, void * wql,
vfs_context_t ctx);
static int pipe_kqfilter(struct fileproc *fp, struct knote *kn,
struct kevent_internal_s *kev, vfs_context_t ctx);
static int pipe_ioctl(struct fileproc *fp, u_long cmd, caddr_t data,
vfs_context_t ctx);
static int pipe_drain(struct fileproc *fp,vfs_context_t ctx);
static const struct fileops pipeops = {
.fo_type = DTYPE_PIPE,
.fo_read = pipe_read,
.fo_write = pipe_write,
.fo_ioctl = pipe_ioctl,
.fo_select = pipe_select,
.fo_close = pipe_close,
.fo_kqfilter = pipe_kqfilter,
.fo_drain = pipe_drain,
};
static void filt_pipedetach(struct knote *kn);
static int filt_piperead(struct knote *kn, long hint);
static int filt_pipereadtouch(struct knote *kn, struct kevent_internal_s *kev);
static int filt_pipereadprocess(struct knote *kn, struct filt_process_s *data, struct kevent_internal_s *kev);
static int filt_pipewrite(struct knote *kn, long hint);
static int filt_pipewritetouch(struct knote *kn, struct kevent_internal_s *kev);
static int filt_pipewriteprocess(struct knote *kn, struct filt_process_s *data, struct kevent_internal_s *kev);
SECURITY_READ_ONLY_EARLY(struct filterops) pipe_rfiltops = {
.f_isfd = 1,
.f_detach = filt_pipedetach,
.f_event = filt_piperead,
.f_touch = filt_pipereadtouch,
.f_process = filt_pipereadprocess,
};
SECURITY_READ_ONLY_EARLY(struct filterops) pipe_wfiltops = {
.f_isfd = 1,
.f_detach = filt_pipedetach,
.f_event = filt_pipewrite,
.f_touch = filt_pipewritetouch,
.f_process = filt_pipewriteprocess,
};
static int nbigpipe; /* for compatibility sake. no longer used */
static int amountpipes; /* total number of pipes in system */
static int amountpipekva; /* total memory used by pipes */
int maxpipekva __attribute__((used)) = PIPE_KVAMAX; /* allowing 16MB max. */
#if PIPE_SYSCTLS
SYSCTL_DECL(_kern_ipc);
SYSCTL_INT(_kern_ipc, OID_AUTO, maxpipekva, CTLFLAG_RD|CTLFLAG_LOCKED,
&maxpipekva, 0, "Pipe KVA limit");
SYSCTL_INT(_kern_ipc, OID_AUTO, maxpipekvawired, CTLFLAG_RW|CTLFLAG_LOCKED,
&maxpipekvawired, 0, "Pipe KVA wired limit");
SYSCTL_INT(_kern_ipc, OID_AUTO, pipes, CTLFLAG_RD|CTLFLAG_LOCKED,
&amountpipes, 0, "Current # of pipes");
SYSCTL_INT(_kern_ipc, OID_AUTO, bigpipes, CTLFLAG_RD|CTLFLAG_LOCKED,
&nbigpipe, 0, "Current # of big pipes");
SYSCTL_INT(_kern_ipc, OID_AUTO, pipekva, CTLFLAG_RD|CTLFLAG_LOCKED,
&amountpipekva, 0, "Pipe KVA usage");
SYSCTL_INT(_kern_ipc, OID_AUTO, pipekvawired, CTLFLAG_RD|CTLFLAG_LOCKED,
&amountpipekvawired, 0, "Pipe wired KVA usage");
#endif
static void pipeclose(struct pipe *cpipe);
static void pipe_free_kmem(struct pipe *cpipe);
static int pipe_create(struct pipe **cpipep);
static int pipespace(struct pipe *cpipe, int size);
static int choose_pipespace(unsigned long current, unsigned long expected);
static int expand_pipespace(struct pipe *p, int target_size);
static void pipeselwakeup(struct pipe *cpipe, struct pipe *spipe);
static __inline int pipeio_lock(struct pipe *cpipe, int catch);
static __inline void pipeio_unlock(struct pipe *cpipe);
extern int postpipeevent(struct pipe *, int);
extern void evpipefree(struct pipe *cpipe);
static lck_grp_t *pipe_mtx_grp;
static lck_attr_t *pipe_mtx_attr;
static lck_grp_attr_t *pipe_mtx_grp_attr;
static zone_t pipe_zone;
#define MAX_PIPESIZE(pipe) ( MAX(PIPE_SIZE, (pipe)->pipe_buffer.size) )
#define PIPE_GARBAGE_AGE_LIMIT 5000 /* In milliseconds */
#define PIPE_GARBAGE_QUEUE_LIMIT 32000
struct pipe_garbage {
struct pipe *pg_pipe;
struct pipe_garbage *pg_next;
uint64_t pg_timestamp;
};
static zone_t pipe_garbage_zone;
static struct pipe_garbage *pipe_garbage_head = NULL;
static struct pipe_garbage *pipe_garbage_tail = NULL;
static uint64_t pipe_garbage_age_limit = PIPE_GARBAGE_AGE_LIMIT;
static int pipe_garbage_count = 0;
static lck_mtx_t *pipe_garbage_lock;
static void pipe_garbage_collect(struct pipe *cpipe);
SYSINIT(vfs, SI_SUB_VFS, SI_ORDER_ANY, pipeinit, NULL);
/* initial setup done at time of sysinit */
void
pipeinit(void)
{
nbigpipe=0;
vm_size_t zone_size;
zone_size = 8192 * sizeof(struct pipe);
pipe_zone = zinit(sizeof(struct pipe), zone_size, 4096, "pipe zone");
/* allocate lock group attribute and group for pipe mutexes */
pipe_mtx_grp_attr = lck_grp_attr_alloc_init();
pipe_mtx_grp = lck_grp_alloc_init("pipe", pipe_mtx_grp_attr);
/* allocate the lock attribute for pipe mutexes */
pipe_mtx_attr = lck_attr_alloc_init();
/*
* Set up garbage collection for dead pipes
*/
zone_size = (PIPE_GARBAGE_QUEUE_LIMIT + 20) *
sizeof(struct pipe_garbage);
pipe_garbage_zone = (zone_t)zinit(sizeof(struct pipe_garbage),
zone_size, 4096, "pipe garbage zone");
pipe_garbage_lock = lck_mtx_alloc_init(pipe_mtx_grp, pipe_mtx_attr);
}
#ifndef CONFIG_EMBEDDED
/* Bitmap for things to touch in pipe_touch() */
#define PIPE_ATIME 0x00000001 /* time of last access */
#define PIPE_MTIME 0x00000002 /* time of last modification */
#define PIPE_CTIME 0x00000004 /* time of last status change */
static void
pipe_touch(struct pipe *tpipe, int touch)
{
struct timespec now;
nanotime(&now);
if (touch & PIPE_ATIME) {
tpipe->st_atimespec.tv_sec = now.tv_sec;
tpipe->st_atimespec.tv_nsec = now.tv_nsec;
}
if (touch & PIPE_MTIME) {
tpipe->st_mtimespec.tv_sec = now.tv_sec;
tpipe->st_mtimespec.tv_nsec = now.tv_nsec;
}
if (touch & PIPE_CTIME) {
tpipe->st_ctimespec.tv_sec = now.tv_sec;
tpipe->st_ctimespec.tv_nsec = now.tv_nsec;
}
}
#endif
static const unsigned int pipesize_blocks[] = {512,1024,2048,4096, 4096 * 2, PIPE_SIZE , PIPE_SIZE * 4 };
/*
* finds the right size from possible sizes in pipesize_blocks
* returns the size which matches max(current,expected)
*/
static int
choose_pipespace(unsigned long current, unsigned long expected)
{
int i = sizeof(pipesize_blocks)/sizeof(unsigned int) -1;
unsigned long target;
/*
* assert that we always get an atomic transaction sized pipe buffer,
* even if the system pipe buffer high-water mark has been crossed.
*/
assert(PIPE_BUF == pipesize_blocks[0]);
if (expected > current)
target = expected;
else
target = current;
while ( i >0 && pipesize_blocks[i-1] > target) {
i=i-1;
}
return pipesize_blocks[i];
}
/*
* expand the size of pipe while there is data to be read,
* and then free the old buffer once the current buffered
* data has been transferred to new storage.
* Required: PIPE_LOCK and io lock to be held by caller.
* returns 0 on success or no expansion possible
*/
static int
expand_pipespace(struct pipe *p, int target_size)
{
struct pipe tmp, oldpipe;
int error;
tmp.pipe_buffer.buffer = 0;
if (p->pipe_buffer.size >= (unsigned) target_size) {
return 0; /* the existing buffer is max size possible */
}
/* create enough space in the target */
error = pipespace(&tmp, target_size);
if (error != 0)
return (error);
oldpipe.pipe_buffer.buffer = p->pipe_buffer.buffer;
oldpipe.pipe_buffer.size = p->pipe_buffer.size;
memcpy(tmp.pipe_buffer.buffer, p->pipe_buffer.buffer, p->pipe_buffer.size);
if (p->pipe_buffer.cnt > 0 && p->pipe_buffer.in <= p->pipe_buffer.out ){
/* we are in State 3 and need extra copying for read to be consistent */
memcpy(&tmp.pipe_buffer.buffer[p->pipe_buffer.size], p->pipe_buffer.buffer, p->pipe_buffer.size);
p->pipe_buffer.in += p->pipe_buffer.size;
}
p->pipe_buffer.buffer = tmp.pipe_buffer.buffer;
p->pipe_buffer.size = tmp.pipe_buffer.size;
pipe_free_kmem(&oldpipe);
return 0;
}
/*
* The pipe system call for the DTYPE_PIPE type of pipes
*
* returns:
* FREAD | fd0 | -->[struct rpipe] --> |~~buffer~~| \
* (pipe_mutex)
* FWRITE | fd1 | -->[struct wpipe] --X /
*/
/* ARGSUSED */
int
pipe(proc_t p, __unused struct pipe_args *uap, int32_t *retval)
{
struct fileproc *rf, *wf;
struct pipe *rpipe, *wpipe;
lck_mtx_t *pmtx;
int fd, error;
if ((pmtx = lck_mtx_alloc_init(pipe_mtx_grp, pipe_mtx_attr)) == NULL)
return (ENOMEM);
rpipe = wpipe = NULL;
if (pipe_create(&rpipe) || pipe_create(&wpipe)) {
error = ENFILE;
goto freepipes;
}
/*
* allocate the space for the normal I/O direction up
* front... we'll delay the allocation for the other
* direction until a write actually occurs (most likely it won't)...
*/
error = pipespace(rpipe, choose_pipespace(rpipe->pipe_buffer.size, 0));
if (error)
goto freepipes;
TAILQ_INIT(&rpipe->pipe_evlist);
TAILQ_INIT(&wpipe->pipe_evlist);
error = falloc(p, &rf, &fd, vfs_context_current());
if (error) {
goto freepipes;
}
retval[0] = fd;
/*
* for now we'll create half-duplex pipes(refer returns section above).
* this is what we've always supported..
*/
rf->f_flag = FREAD;
rf->f_data = (caddr_t)rpipe;
rf->f_ops = &pipeops;
error = falloc(p, &wf, &fd, vfs_context_current());
if (error) {
fp_free(p, retval[0], rf);
goto freepipes;
}
wf->f_flag = FWRITE;
wf->f_data = (caddr_t)wpipe;
wf->f_ops = &pipeops;
rpipe->pipe_peer = wpipe;
wpipe->pipe_peer = rpipe;
/* both structures share the same mutex */
rpipe->pipe_mtxp = wpipe->pipe_mtxp = pmtx;
retval[1] = fd;
#if CONFIG_MACF
/*
* XXXXXXXX SHOULD NOT HOLD FILE_LOCK() XXXXXXXXXXXX
*
* struct pipe represents a pipe endpoint. The MAC label is shared
* between the connected endpoints. As a result mac_pipe_label_init() and
* mac_pipe_label_associate() should only be called on one of the endpoints
* after they have been connected.
*/
mac_pipe_label_init(rpipe);
mac_pipe_label_associate(kauth_cred_get(), rpipe);
wpipe->pipe_label = rpipe->pipe_label;
#endif
proc_fdlock_spin(p);
procfdtbl_releasefd(p, retval[0], NULL);
procfdtbl_releasefd(p, retval[1], NULL);
fp_drop(p, retval[0], rf, 1);
fp_drop(p, retval[1], wf, 1);
proc_fdunlock(p);
return (0);
freepipes:
pipeclose(rpipe);
pipeclose(wpipe);
lck_mtx_free(pmtx, pipe_mtx_grp);
return (error);
}
int
pipe_stat(struct pipe *cpipe, void *ub, int isstat64)
{
#if CONFIG_MACF
int error;
#endif
int pipe_size = 0;
int pipe_count;
struct stat *sb = (struct stat *)0; /* warning avoidance ; protected by isstat64 */
struct stat64 * sb64 = (struct stat64 *)0; /* warning avoidance ; protected by isstat64 */
if (cpipe == NULL)
return (EBADF);
PIPE_LOCK(cpipe);
#if CONFIG_MACF
error = mac_pipe_check_stat(kauth_cred_get(), cpipe);
if (error) {
PIPE_UNLOCK(cpipe);
return (error);
}
#endif
if (cpipe->pipe_buffer.buffer == 0) {
/* must be stat'ing the write fd */
if (cpipe->pipe_peer) {
/* the peer still exists, use it's info */
pipe_size = MAX_PIPESIZE(cpipe->pipe_peer);
pipe_count = cpipe->pipe_peer->pipe_buffer.cnt;
} else {
pipe_count = 0;
}
} else {
pipe_size = MAX_PIPESIZE(cpipe);
pipe_count = cpipe->pipe_buffer.cnt;
}
/*
* since peer's buffer is setup ouside of lock
* we might catch it in transient state
*/
if (pipe_size == 0)
pipe_size = MAX(PIPE_SIZE, pipesize_blocks[0]);
if (isstat64 != 0) {
sb64 = (struct stat64 *)ub;
bzero(sb64, sizeof(*sb64));
sb64->st_mode = S_IFIFO | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP;
sb64->st_blksize = pipe_size;
sb64->st_size = pipe_count;
sb64->st_blocks = (sb64->st_size + sb64->st_blksize - 1) / sb64->st_blksize;
sb64->st_uid = kauth_getuid();
sb64->st_gid = kauth_getgid();
sb64->st_atimespec.tv_sec = cpipe->st_atimespec.tv_sec;
sb64->st_atimespec.tv_nsec = cpipe->st_atimespec.tv_nsec;
sb64->st_mtimespec.tv_sec = cpipe->st_mtimespec.tv_sec;
sb64->st_mtimespec.tv_nsec = cpipe->st_mtimespec.tv_nsec;
sb64->st_ctimespec.tv_sec = cpipe->st_ctimespec.tv_sec;
sb64->st_ctimespec.tv_nsec = cpipe->st_ctimespec.tv_nsec;
/*
* Return a relatively unique inode number based on the current
* address of this pipe's struct pipe. This number may be recycled
* relatively quickly.
*/
sb64->st_ino = (ino64_t)VM_KERNEL_ADDRPERM((uintptr_t)cpipe);
} else {
sb = (struct stat *)ub;
bzero(sb, sizeof(*sb));
sb->st_mode = S_IFIFO | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP;
sb->st_blksize = pipe_size;
sb->st_size = pipe_count;
sb->st_blocks = (sb->st_size + sb->st_blksize - 1) / sb->st_blksize;
sb->st_uid = kauth_getuid();
sb->st_gid = kauth_getgid();
sb->st_atimespec.tv_sec = cpipe->st_atimespec.tv_sec;
sb->st_atimespec.tv_nsec = cpipe->st_atimespec.tv_nsec;
sb->st_mtimespec.tv_sec = cpipe->st_mtimespec.tv_sec;
sb->st_mtimespec.tv_nsec = cpipe->st_mtimespec.tv_nsec;
sb->st_ctimespec.tv_sec = cpipe->st_ctimespec.tv_sec;
sb->st_ctimespec.tv_nsec = cpipe->st_ctimespec.tv_nsec;
/*
* Return a relatively unique inode number based on the current
* address of this pipe's struct pipe. This number may be recycled
* relatively quickly.
*/
sb->st_ino = (ino_t)VM_KERNEL_ADDRPERM((uintptr_t)cpipe);
}
PIPE_UNLOCK(cpipe);
/*
* POSIX: Left as 0: st_dev, st_nlink, st_rdev, st_flags, st_gen,
* st_uid, st_gid.
*
* XXX (st_dev) should be unique, but there is no device driver that
* XXX is associated with pipes, since they are implemented via a
* XXX struct fileops indirection rather than as FS objects.
*/
return (0);
}
/*
* Allocate kva for pipe circular buffer, the space is pageable
* This routine will 'realloc' the size of a pipe safely, if it fails
* it will retain the old buffer.
* If it fails it will return ENOMEM.
*/
static int
pipespace(struct pipe *cpipe, int size)
{
vm_offset_t buffer;
if (size <= 0)
return(EINVAL);
if ((buffer = (vm_offset_t)kalloc(size)) == 0 )
return(ENOMEM);
/* free old resources if we're resizing */
pipe_free_kmem(cpipe);
cpipe->pipe_buffer.buffer = (caddr_t)buffer;
cpipe->pipe_buffer.size = size;
cpipe->pipe_buffer.in = 0;
cpipe->pipe_buffer.out = 0;
cpipe->pipe_buffer.cnt = 0;
OSAddAtomic(1, &amountpipes);
OSAddAtomic(cpipe->pipe_buffer.size, &amountpipekva);
return (0);
}
/*
* initialize and allocate VM and memory for pipe
*/
static int
pipe_create(struct pipe **cpipep)
{
struct pipe *cpipe;
cpipe = (struct pipe *)zalloc(pipe_zone);
if ((*cpipep = cpipe) == NULL)
return (ENOMEM);
/*
* protect so pipespace or pipeclose don't follow a junk pointer
* if pipespace() fails.
*/
bzero(cpipe, sizeof *cpipe);
#ifndef CONFIG_EMBEDDED
/* Initial times are all the time of creation of the pipe */
pipe_touch(cpipe, PIPE_ATIME | PIPE_MTIME | PIPE_CTIME);
#endif
return (0);
}
/*
* lock a pipe for I/O, blocking other access
*/
static inline int
pipeio_lock(struct pipe *cpipe, int catch)
{
int error;
while (cpipe->pipe_state & PIPE_LOCKFL) {
cpipe->pipe_state |= PIPE_LWANT;
error = msleep(cpipe, PIPE_MTX(cpipe), catch ? (PRIBIO | PCATCH) : PRIBIO,
"pipelk", 0);
if (error != 0)
return (error);
}
cpipe->pipe_state |= PIPE_LOCKFL;
return (0);
}
/*
* unlock a pipe I/O lock
*/
static inline void
pipeio_unlock(struct pipe *cpipe)
{
cpipe->pipe_state &= ~PIPE_LOCKFL;
if (cpipe->pipe_state & PIPE_LWANT) {
cpipe->pipe_state &= ~PIPE_LWANT;
wakeup(cpipe);
}
}
/*
* wakeup anyone whos blocked in select
*/
static void
pipeselwakeup(struct pipe *cpipe, struct pipe *spipe)
{
if (cpipe->pipe_state & PIPE_SEL) {
cpipe->pipe_state &= ~PIPE_SEL;
selwakeup(&cpipe->pipe_sel);
}
if (cpipe->pipe_state & PIPE_KNOTE)
KNOTE(&cpipe->pipe_sel.si_note, 1);
postpipeevent(cpipe, EV_RWBYTES);
if (spipe && (spipe->pipe_state & PIPE_ASYNC) && spipe->pipe_pgid) {
if (spipe->pipe_pgid < 0)
gsignal(-spipe->pipe_pgid, SIGIO);
else
proc_signal(spipe->pipe_pgid, SIGIO);
}
}
/*
* Read n bytes from the buffer. Semantics are similar to file read.
* returns: number of bytes read from the buffer
*/
/* ARGSUSED */
static int
pipe_read(struct fileproc *fp, struct uio *uio, __unused int flags,
__unused vfs_context_t ctx)
{
struct pipe *rpipe = (struct pipe *)fp->f_data;
int error;
int nread = 0;
u_int size;
PIPE_LOCK(rpipe);
++rpipe->pipe_busy;
error = pipeio_lock(rpipe, 1);
if (error)
goto unlocked_error;
#if CONFIG_MACF
error = mac_pipe_check_read(kauth_cred_get(), rpipe);
if (error)
goto locked_error;
#endif
while (uio_resid(uio)) {
/*
* normal pipe buffer receive
*/
if (rpipe->pipe_buffer.cnt > 0) {
/*
* # bytes to read is min( bytes from read pointer until end of buffer,
* total unread bytes,
* user requested byte count)
*/
size = rpipe->pipe_buffer.size - rpipe->pipe_buffer.out;
if (size > rpipe->pipe_buffer.cnt)
size = rpipe->pipe_buffer.cnt;
// LP64todo - fix this!
if (size > (u_int) uio_resid(uio))
size = (u_int) uio_resid(uio);
PIPE_UNLOCK(rpipe); /* we still hold io lock.*/
error = uiomove(
&rpipe->pipe_buffer.buffer[rpipe->pipe_buffer.out],
size, uio);
PIPE_LOCK(rpipe);
if (error)
break;
rpipe->pipe_buffer.out += size;
if (rpipe->pipe_buffer.out >= rpipe->pipe_buffer.size)
rpipe->pipe_buffer.out = 0;
rpipe->pipe_buffer.cnt -= size;
/*
* If there is no more to read in the pipe, reset
* its pointers to the beginning. This improves
* cache hit stats.
*/
if (rpipe->pipe_buffer.cnt == 0) {
rpipe->pipe_buffer.in = 0;
rpipe->pipe_buffer.out = 0;
}
nread += size;
} else {
/*
* detect EOF condition
* read returns 0 on EOF, no need to set error
*/
if (rpipe->pipe_state & (PIPE_DRAIN | PIPE_EOF)) {
break;
}
/*
* If the "write-side" has been blocked, wake it up now.
*/
if (rpipe->pipe_state & PIPE_WANTW) {
rpipe->pipe_state &= ~PIPE_WANTW;
wakeup(rpipe);
}
/*
* Break if some data was read in previous iteration.
*/
if (nread > 0)
break;
/*
* Unlock the pipe buffer for our remaining processing.
* We will either break out with an error or we will
* sleep and relock to loop.
*/
pipeio_unlock(rpipe);
/*
* Handle non-blocking mode operation or
* wait for more data.
*/
if (fp->f_flag & FNONBLOCK) {
error = EAGAIN;
} else {
rpipe->pipe_state |= PIPE_WANTR;
error = msleep(rpipe, PIPE_MTX(rpipe), PRIBIO | PCATCH, "piperd", 0);
if (error == 0)
error = pipeio_lock(rpipe, 1);
}
if (error)
goto unlocked_error;
}
}
#if CONFIG_MACF
locked_error:
#endif
pipeio_unlock(rpipe);
unlocked_error:
--rpipe->pipe_busy;
/*
* PIPE_WANT processing only makes sense if pipe_busy is 0.
*/
if ((rpipe->pipe_busy == 0) && (rpipe->pipe_state & PIPE_WANT)) {
rpipe->pipe_state &= ~(PIPE_WANT|PIPE_WANTW);
wakeup(rpipe);
} else if (rpipe->pipe_buffer.cnt < rpipe->pipe_buffer.size) {
/*
* Handle write blocking hysteresis.
*/
if (rpipe->pipe_state & PIPE_WANTW) {
rpipe->pipe_state &= ~PIPE_WANTW;
wakeup(rpipe);
}
}
if ((rpipe->pipe_buffer.size - rpipe->pipe_buffer.cnt) > 0)
pipeselwakeup(rpipe, rpipe->pipe_peer);
#ifndef CONFIG_EMBEDDED
/* update last read time */
pipe_touch(rpipe, PIPE_ATIME);
#endif
PIPE_UNLOCK(rpipe);
return (error);
}
/*
* perform a write of n bytes into the read side of buffer. Since
* pipes are unidirectional a write is meant to be read by the otherside only.
*/
static int
pipe_write(struct fileproc *fp, struct uio *uio, __unused int flags,
__unused vfs_context_t ctx)
{
int error = 0;
int orig_resid;
int pipe_size;
struct pipe *wpipe, *rpipe;
// LP64todo - fix this!
orig_resid = uio_resid(uio);
int space;
rpipe = (struct pipe *)fp->f_data;
PIPE_LOCK(rpipe);
wpipe = rpipe->pipe_peer;
/*
* detect loss of pipe read side, issue SIGPIPE if lost.
*/
if (wpipe == NULL || (wpipe->pipe_state & (PIPE_DRAIN | PIPE_EOF))) {
PIPE_UNLOCK(rpipe);
return (EPIPE);
}
#if CONFIG_MACF
error = mac_pipe_check_write(kauth_cred_get(), wpipe);
if (error) {
PIPE_UNLOCK(rpipe);
return (error);
}
#endif
++wpipe->pipe_busy;
pipe_size = 0;
/*
* need to allocate some storage... we delay the allocation
* until the first write on fd[0] to avoid allocating storage for both
* 'pipe ends'... most pipes are half-duplex with the writes targeting
* fd[1], so allocating space for both ends is a waste...
*/
if ( wpipe->pipe_buffer.buffer == 0 || (
(unsigned)orig_resid > wpipe->pipe_buffer.size - wpipe->pipe_buffer.cnt &&
amountpipekva < maxpipekva ) ) {
pipe_size = choose_pipespace(wpipe->pipe_buffer.size, wpipe->pipe_buffer.cnt + orig_resid);
}
if (pipe_size) {
/*
* need to do initial allocation or resizing of pipe
* holding both structure and io locks.
*/
if ((error = pipeio_lock(wpipe, 1)) == 0) {
if (wpipe->pipe_buffer.cnt == 0)
error = pipespace(wpipe, pipe_size);
else
error = expand_pipespace(wpipe, pipe_size);
pipeio_unlock(wpipe);
/* allocation failed */
if (wpipe->pipe_buffer.buffer == 0)
error = ENOMEM;
}
if (error) {
/*
* If an error occurred unbusy and return, waking up any pending
* readers.
*/
--wpipe->pipe_busy;
if ((wpipe->pipe_busy == 0) &&
(wpipe->pipe_state & PIPE_WANT)) {
wpipe->pipe_state &= ~(PIPE_WANT | PIPE_WANTR);
wakeup(wpipe);
}
PIPE_UNLOCK(rpipe);
return(error);
}
}
while (uio_resid(uio)) {
retrywrite:
space = wpipe->pipe_buffer.size - wpipe->pipe_buffer.cnt;
/* Writes of size <= PIPE_BUF must be atomic. */
if ((space < uio_resid(uio)) && (orig_resid <= PIPE_BUF))
space = 0;
if (space > 0) {
if ((error = pipeio_lock(wpipe,1)) == 0) {
int size; /* Transfer size */
int segsize; /* first segment to transfer */
if (wpipe->pipe_state & (PIPE_DRAIN | PIPE_EOF)) {
pipeio_unlock(wpipe);
error = EPIPE;
break;
}
/*
* If a process blocked in pipeio_lock, our
* value for space might be bad... the mutex
* is dropped while we're blocked
*/
if (space > (int)(wpipe->pipe_buffer.size -
wpipe->pipe_buffer.cnt)) {
pipeio_unlock(wpipe);
goto retrywrite;
}
/*
* Transfer size is minimum of uio transfer
* and free space in pipe buffer.