summaryrefslogtreecommitdiffstats
path: root/api/src/glfs.c
blob: 897d3eab80916a1ee30a54b82e1b0405d079113b (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
/*
  Copyright (c) 2012 Red Hat, Inc. <http://www.redhat.com>
  This file is part of GlusterFS.

  This file is licensed to you under your choice of the GNU Lesser
  General Public License, version 3 or any later version (LGPLv3 or
  later), or the GNU General Public License, version 2 (GPLv2), in all
  cases as published by the Free Software Foundation.
*/


/*
  TODO:
  - merge locks in glfs_posix_lock for lock self-healing
  - set proper pid/lk_owner to call frames (currently buried in syncop)
  - fix logging.c/h to store logfp and loglevel in glusterfs_ctx_t and
    reach it via THIS.
  - update syncop functions to accept/return xdata. ???
  - protocol/client to reconnect immediately after portmap disconnect.
  - handle SEEK_END failure in _lseek()
  - handle umask (per filesystem?)
  - make itables LRU based
  - 0-copy for readv/writev
  - reconcile the open/creat mess
*/

#include <unistd.h>
#include <string.h>
#include <stdlib.h>
#include <stdio.h>
#include <inttypes.h>
#include <sys/types.h>
#include <unistd.h>
#include <limits.h>

#ifndef _CONFIG_H
#define _CONFIG_H
#include "config.h"
#endif

#include "glusterfs.h"
#include "logging.h"
#include "stack.h"
#include "event.h"
#include "glfs-mem-types.h"
#include "common-utils.h"
#include "syncop.h"
#include "call-stub.h"
#include "gfapi-messages.h"

#include "glfs.h"
#include "glfs-internal.h"
#include "hashfn.h"
#include "rpc-clnt.h"


static gf_boolean_t
vol_assigned (cmd_args_t *args)
{
	return args->volfile || args->volfile_server;
}


static int
glusterfs_ctx_defaults_init (glusterfs_ctx_t *ctx)
{
	call_pool_t   *pool = NULL;
	int	       ret = -1;

	if (!ctx) {
		goto err;
        }

        ret = xlator_mem_acct_init (THIS, glfs_mt_end + 1);
        if (ret != 0) {
                gf_msg (THIS->name, GF_LOG_ERROR, ENOMEM,
                        API_MSG_MEM_ACCT_INIT_FAILED,
                        "Memory accounting init failed");
                return ret;
        }

        /* reset ret to -1 so that we don't need to explicitly
         * set it in all error paths before "goto err"
         */

        ret = -1;

	ctx->process_uuid = generate_glusterfs_ctx_id ();
	if (!ctx->process_uuid) {
		goto err;
	}

	ctx->page_size	= 128 * GF_UNIT_KB;

	ctx->iobuf_pool = iobuf_pool_new ();
	if (!ctx->iobuf_pool) {
		goto err;
	}

	ctx->event_pool = event_pool_new (DEFAULT_EVENT_POOL_SIZE,
                                          STARTING_EVENT_THREADS);
	if (!ctx->event_pool) {
		goto err;
	}

	ctx->env = syncenv_new (0, 0, 0);
	if (!ctx->env) {
		goto err;
	}

	pool = GF_CALLOC (1, sizeof (call_pool_t),
			  glfs_mt_call_pool_t);
	if (!pool) {
		goto err;
	}

	/* frame_mem_pool size 112 * 4k */
	pool->frame_mem_pool = mem_pool_new (call_frame_t, 4096);
	if (!pool->frame_mem_pool) {
		goto err;
	}
	/* stack_mem_pool size 256 * 1024 */
	pool->stack_mem_pool = mem_pool_new (call_stack_t, 1024);
	if (!pool->stack_mem_pool) {
		goto err;
	}

	ctx->stub_mem_pool = mem_pool_new (call_stub_t, 1024);
	if (!ctx->stub_mem_pool) {
		goto err;
	}

	ctx->dict_pool = mem_pool_new (dict_t, GF_MEMPOOL_COUNT_OF_DICT_T);
	if (!ctx->dict_pool)
		goto err;

	ctx->dict_pair_pool = mem_pool_new (data_pair_t,
					    GF_MEMPOOL_COUNT_OF_DATA_PAIR_T);
	if (!ctx->dict_pair_pool)
		goto err;

	ctx->dict_data_pool = mem_pool_new (data_t, GF_MEMPOOL_COUNT_OF_DATA_T);
	if (!ctx->dict_data_pool)
		goto err;

        ctx->logbuf_pool = mem_pool_new (log_buf_t,
                                         GF_MEMPOOL_COUNT_OF_LRU_BUF_T);
        if (!ctx->logbuf_pool)
                goto err;

	INIT_LIST_HEAD (&pool->all_frames);
	INIT_LIST_HEAD (&ctx->cmd_args.xlator_options);
        INIT_LIST_HEAD (&ctx->cmd_args.volfile_servers);

	LOCK_INIT (&pool->lock);
	ctx->pool = pool;

	pthread_mutex_init (&(ctx->lock), NULL);

	ret = 0;
err:
	if (ret && pool) {
		if (pool->frame_mem_pool)
			mem_pool_destroy (pool->frame_mem_pool);
		if (pool->stack_mem_pool)
			mem_pool_destroy (pool->stack_mem_pool);
		GF_FREE (pool);
	}

	if (ret && ctx) {
		if (ctx->stub_mem_pool)
			mem_pool_destroy (ctx->stub_mem_pool);
		if (ctx->dict_pool)
			mem_pool_destroy (ctx->dict_pool);
		if (ctx->dict_data_pool)
			mem_pool_destroy (ctx->dict_data_pool);
		if (ctx->dict_pair_pool)
			mem_pool_destroy (ctx->dict_pair_pool);
                if (ctx->logbuf_pool)
                        mem_pool_destroy (ctx->logbuf_pool);
	}

	return ret;
}


static int
create_master (struct glfs *fs)
{
	int		 ret = 0;
	xlator_t	*master = NULL;

	master = GF_CALLOC (1, sizeof (*master),
			    glfs_mt_xlator_t);
	if (!master)
		goto err;

	master->name = gf_strdup ("gfapi");
	if (!master->name)
		goto err;

	if (xlator_set_type (master, "mount/api") == -1) {
		gf_msg ("glfs", GF_LOG_ERROR, 0,
                        API_MSG_MASTER_XLATOR_INIT_FAILED, "master xlator "
                        "for %s initialization failed", fs->volname);
		goto err;
	}

	master->ctx	 = fs->ctx;
	master->private	 = fs;
	master->options	 = get_new_dict ();
	if (!master->options)
		goto err;


	ret = xlator_init (master);
	if (ret) {
		gf_msg ("glfs", GF_LOG_ERROR, 0,
                        API_MSG_GFAPI_XLATOR_INIT_FAILED,
			"failed to initialize gfapi translator");
		goto err;
	}

	fs->ctx->master = master;
	THIS = master;

	return 0;

err:
	if (master) {
		xlator_destroy (master);
	}

	return -1;
}


static FILE *
get_volfp (struct glfs *fs)
{
	cmd_args_t  *cmd_args = NULL;
	FILE	    *specfp = NULL;

	cmd_args = &fs->ctx->cmd_args;

	if ((specfp = fopen (cmd_args->volfile, "r")) == NULL) {
		gf_msg ("glfs", GF_LOG_ERROR, errno,
                        API_MSG_VOLFILE_OPEN_FAILED,
			"volume file %s open failed: %s",
			cmd_args->volfile,
			strerror (errno));
		return NULL;
	}

	gf_msg_debug ("glfs", 0, "loading volume file %s", cmd_args->volfile);

	return specfp;
}


int
glfs_volumes_init (struct glfs *fs)
{
	FILE		   *fp = NULL;
	cmd_args_t	   *cmd_args = NULL;
	int		    ret = 0;

	cmd_args = &fs->ctx->cmd_args;

	if (!vol_assigned (cmd_args))
		return -1;

	if (cmd_args->volfile_server) {
		ret = glfs_mgmt_init (fs);
		goto out;
	}

	fp = get_volfp (fs);

	if (!fp) {
		gf_msg ("glfs", GF_LOG_ERROR, ENOENT,
                        API_MSG_VOL_SPEC_FILE_ERROR,
			"Cannot reach volume specification file");
		ret = -1;
		goto out;
	}

	ret = glfs_process_volfp (fs, fp);
	if (ret)
		goto out;

out:
	return ret;
}


///////////////////////////////////////////////////////////////////////////////


int
pub_glfs_set_xlator_option (struct glfs *fs, const char *xlator,
                            const char *key, const char *value)
{
	xlator_cmdline_option_t *option = NULL;

        DECLARE_OLD_THIS;
        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

	option = GF_CALLOC (1, sizeof (*option),
			    glfs_mt_xlator_cmdline_option_t);
	if (!option)
		goto enomem;

	INIT_LIST_HEAD (&option->cmd_args);

	option->volume = gf_strdup (xlator);
	if (!option->volume)
		goto enomem;
	option->key = gf_strdup (key);
	if (!option->key)
		goto enomem;
	option->value = gf_strdup (value);
	if (!option->value)
		goto enomem;

	list_add (&option->cmd_args, &fs->ctx->cmd_args.xlator_options);

        __GLFS_EXIT_FS;

	return 0;
enomem:
	errno = ENOMEM;

	if (!option) {
                __GLFS_EXIT_FS;
		return -1;
        }

	GF_FREE (option->volume);
	GF_FREE (option->key);
	GF_FREE (option->value);
	GF_FREE (option);

        __GLFS_EXIT_FS;

invalid_fs:
	return -1;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_set_xlator_option, 3.4.0);


int
pub_glfs_unset_volfile_server (struct glfs *fs, const char *transport,
                               const char *host, const int port)
{
        cmd_args_t       *cmd_args = NULL;
        server_cmdline_t *server = NULL;
        int               ret = -1;

        if (!transport || !host || !port) {
                errno = EINVAL;
                return ret;
        }

        DECLARE_OLD_THIS;
        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

        cmd_args = &fs->ctx->cmd_args;
        list_for_each_entry(server, &cmd_args->curr_server->list, list) {
                if ((!strcmp(server->volfile_server, host) &&
                     !strcmp(server->transport, transport) &&
                     (server->port == port))) {
                        list_del (&server->list);
                        ret = 0;
                        goto out;
                }
        }

out:
        __GLFS_EXIT_FS;

invalid_fs:
        return ret;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_unset_volfile_server, 3.5.1);


int
pub_glfs_set_volfile_server (struct glfs *fs, const char *transport,
                             const char *host, int port)
{
        cmd_args_t            *cmd_args = NULL;
        server_cmdline_t      *server = NULL;
        server_cmdline_t      *tmp = NULL;
        int                    ret = -1;

        if (!fs || !host) {
                errno = EINVAL;
                return ret;
        }

        DECLARE_OLD_THIS;
        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

        cmd_args = &fs->ctx->cmd_args;

        cmd_args->max_connect_attempts = 1;

        server = GF_CALLOC (1, sizeof (server_cmdline_t),
                            glfs_mt_server_cmdline_t);

        if (!server) {
                errno = ENOMEM;
                goto out;
        }

        INIT_LIST_HEAD (&server->list);

        server->volfile_server = gf_strdup (host);
        if (!server->volfile_server) {
                errno = ENOMEM;
                goto out;
        }

        if (transport) {
                server->transport = gf_strdup (transport);
                if (!server->transport) {
                        errno = ENOMEM;
                        goto out;
                }
        }

        server->port = port;

        if (!cmd_args->volfile_server) {
                cmd_args->volfile_server = server->volfile_server;
                cmd_args->volfile_server_transport = server->transport;
                cmd_args->volfile_server_port = server->port;
                cmd_args->curr_server = server;
        }

        list_for_each_entry(tmp, &cmd_args->volfile_servers, list) {
                if ((!strcmp(tmp->volfile_server, host) &&
                     !strcmp(tmp->transport, transport) &&
                     (tmp->port == port))) {
                        errno = EEXIST;
                        ret = -1;
                        goto out;
                }
        }

        list_add_tail (&server->list, &cmd_args->volfile_servers);

        ret = 0;
out:
        if (ret == -1) {
                if (server) {
                        GF_FREE (server->volfile_server);
                        GF_FREE (server->transport);
                        GF_FREE (server);
                }
        }

        __GLFS_EXIT_FS;

invalid_fs:
        return ret;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_set_volfile_server, 3.4.0);


int
pub_glfs_setfsuid (uid_t fsuid)
{
         /* TODO:
         * - Set the THIS and restore it appropriately
         */
	return syncopctx_setfsuid (&fsuid);
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_setfsuid, 3.4.2);


int
pub_glfs_setfsgid (gid_t fsgid)
{
         /* TODO:
         * - Set the THIS and restore it appropriately
         */
	return syncopctx_setfsgid (&fsgid);
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_setfsgid, 3.4.2);


int
pub_glfs_setfsgroups (size_t size, const gid_t *list)
{
         /* TODO:
         * - Set the THIS and restore it appropriately
         */
	return syncopctx_setfsgroups(size, list);
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_setfsgroups, 3.4.2);


struct glfs *
pub_glfs_from_glfd (struct glfs_fd *glfd)
{
	return glfd->fs;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_from_glfd, 3.4.0);


struct glfs_fd *
glfs_fd_new (struct glfs *fs)
{
	struct glfs_fd  *glfd = NULL;

	glfd = GF_CALLOC (1, sizeof (*glfd), glfs_mt_glfs_fd_t);
	if (!glfd)
		return NULL;

	glfd->fs = fs;

	INIT_LIST_HEAD (&glfd->openfds);

	return glfd;
}


void
glfs_fd_bind (struct glfs_fd *glfd)
{
	struct glfs *fs = NULL;

	fs = glfd->fs;

	glfs_lock (fs);
	{
		list_add_tail (&glfd->openfds, &fs->openfds);
	}
	glfs_unlock (fs);
}

void
glfs_fd_destroy (struct glfs_fd *glfd)
{
	if (!glfd)
		return;

	glfs_lock (glfd->fs);
	{
		list_del_init (&glfd->openfds);
	}
	glfs_unlock (glfd->fs);

	if (glfd->fd)
		fd_unref (glfd->fd);

	GF_FREE (glfd->readdirbuf);

	GF_FREE (glfd);
}


static void *
glfs_poller (void *data)
{
	struct glfs  *fs = NULL;

	fs = data;

	event_dispatch (fs->ctx->event_pool);

	return NULL;
}

/*
 * please note all the variable initializations done here probably
 * need to be added in 'glfs_new_from_ctx()' as well, so that,
 * we do not miss out while adding new members in 'fs'.
*/


struct glfs *
pub_glfs_new (const char *volname)
{
	struct glfs     *fs             = NULL;
	int              ret            = -1;
	glusterfs_ctx_t *ctx            = NULL;
        xlator_t        *old_THIS       = NULL;

        if (!volname) {
                errno = EINVAL;
                return NULL;
        }

        fs = CALLOC (1, sizeof (*fs));
        if (!fs)
                return NULL;

        ret = pthread_mutex_init (&fs->mutex, NULL);
        if (ret != 0)
                goto freefs;

        ret = pthread_cond_init (&fs->cond, NULL);
        if (ret != 0)
                goto mutex_destroy;

        ret = pthread_cond_init (&fs->child_down_cond, NULL);
        if (ret != 0)
                goto cond_destroy;

        ret = pthread_mutex_init (&fs->upcall_list_mutex, NULL);
        if (ret != 0)
                goto cond_child_destroy;

        ctx = glusterfs_ctx_new ();
        if (!ctx)
                goto fini;

        /* first globals init, for gf_mem_acct_enable_set () */

        ret = glusterfs_globals_init (ctx);
        if (ret)
                goto fini;

        old_THIS = THIS;
        /* THIS is set to NULL so that we do not modify the caller xlators'
         * ctx, instead we set the global_xlator->ctx
         */
        THIS = NULL;
        THIS->ctx = ctx;

        /* then ctx_defaults_init, for xlator_mem_acct_init(THIS) */

        ret = glusterfs_ctx_defaults_init (ctx);
        if (ret)
                goto fini;

        fs->ctx = ctx;

        ret = glfs_set_logging (fs, "/dev/null", 0);
        if (ret)
                goto fini;

        fs->ctx->cmd_args.volfile_id = gf_strdup (volname);
        if (!(fs->ctx->cmd_args.volfile_id))
                goto fini;

        fs->volname = strdup (volname);
        if (!fs->volname)
                goto fini;

        INIT_LIST_HEAD (&fs->openfds);
        INIT_LIST_HEAD (&fs->upcall_list);

        fs->pin_refcnt = 0;

        goto out;

cond_child_destroy:
        pthread_cond_destroy (&fs->child_down_cond);

cond_destroy:
        pthread_cond_destroy (&fs->cond);

mutex_destroy:
        pthread_mutex_destroy (&fs->mutex);

freefs:
        FREE (fs);
        fs = NULL;
        goto out;
fini:
        /*
         * When pthread_*init() fails there is no way for other cleanup
         * funtions (glfs_fini/glfs_free_from_ctx) to know which of them succeded
         * and which did not(unless there is a flag). Hence pthread cleanup is done
         * in this funtion.Anything that fails after pthread_*_init() succeeds, should
         * directly call glfs_fini() to cleanup the resources.
         */

         glfs_fini (fs);
         fs = NULL;
out:
        if (old_THIS)
                THIS = old_THIS;

        return fs;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_new, 3.4.0);


struct glfs *
priv_glfs_new_from_ctx (glusterfs_ctx_t *ctx)
{
        struct glfs    *fs = NULL;

        if (!ctx)
                return NULL;

        fs = CALLOC (1, sizeof (*fs));
        if (!fs)
                return NULL;
        fs->ctx = ctx;

        (void) pthread_cond_init (&fs->cond, NULL);
        (void) pthread_cond_init (&fs->child_down_cond, NULL);

        (void) pthread_mutex_init (&fs->mutex, NULL);

        INIT_LIST_HEAD (&fs->openfds);

        INIT_LIST_HEAD (&fs->upcall_list);
        pthread_mutex_init (&fs->upcall_list_mutex, NULL);

        fs->pin_refcnt = 0;

        return fs;
}

GFAPI_SYMVER_PRIVATE_DEFAULT(glfs_new_from_ctx, 3.7.0);


void
priv_glfs_free_from_ctx (struct glfs *fs)
{
        upcall_entry       *u_list   = NULL;
        upcall_entry       *tmp      = NULL;

        if (!fs)
                return;

        /* cleanup upcall structures */
        list_for_each_entry_safe (u_list, tmp,
                                  &fs->upcall_list,
                                  upcall_list) {
                list_del_init (&u_list->upcall_list);
                GF_FREE (u_list->upcall_data.data);
                GF_FREE (u_list);
        }
        (void) pthread_mutex_destroy (&fs->upcall_list_mutex);

        (void) pthread_cond_destroy (&fs->cond);
        (void) pthread_cond_destroy (&fs->child_down_cond);

        (void) pthread_mutex_destroy (&fs->mutex);

        if (fs->volname)
                FREE (fs->volname);

        FREE (fs);
}

GFAPI_SYMVER_PRIVATE_DEFAULT(glfs_free_from_ctx, 3.7.0);


int
pub_glfs_set_volfile (struct glfs *fs, const char *volfile)
{
	cmd_args_t  *cmd_args = NULL;

	cmd_args = &fs->ctx->cmd_args;

	if (vol_assigned (cmd_args))
		return -1;

	cmd_args->volfile = gf_strdup (volfile);
        if (!cmd_args->volfile)
                return -1;
	return 0;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_set_volfile, 3.4.0);


int
pub_glfs_set_logging (struct glfs *fs, const char *logfile, int loglevel)
{
        int  ret = -1;
        char *tmplog = NULL;

        DECLARE_OLD_THIS;
        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

        if (!logfile) {
                ret = gf_set_log_file_path (&fs->ctx->cmd_args);
                if (ret)
                        goto out;
                tmplog = fs->ctx->cmd_args.log_file;
        } else {
                tmplog = (char *)logfile;
        }

        /* finish log set parameters before init */
        if (loglevel >= 0)
                gf_log_set_loglevel (loglevel);

        ret = gf_log_init (fs->ctx, tmplog, NULL);
        if (ret)
                goto out;

        ret = gf_log_inject_timer_event (fs->ctx);
        if (ret)
                goto out;

out:
        __GLFS_EXIT_FS;

invalid_fs:
        return ret;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_set_logging, 3.4.0);


int
glfs_init_wait (struct glfs *fs)
{
	int   ret = -1;

	/* Always a top-down call, use glfs_lock() */
	glfs_lock (fs);
	{
		while (!fs->init)
			pthread_cond_wait (&fs->cond,
					   &fs->mutex);
		ret = fs->ret;
		errno = fs->err;
	}
	glfs_unlock (fs);

	return ret;
}


void
priv_glfs_init_done (struct glfs *fs, int ret)
{
	glfs_init_cbk init_cbk;

	if (!fs) {
		gf_msg ("glfs", GF_LOG_ERROR, EINVAL, API_MSG_GLFS_FSOBJ_NULL,
			"fs is NULL");
		goto out;
	}

	init_cbk = fs->init_cbk;

	/* Always a bottom-up call, use mutex_lock() */
	pthread_mutex_lock (&fs->mutex);
	{
		fs->init = 1;
		fs->ret = ret;
		fs->err = errno;

		if (!init_cbk)
			pthread_cond_broadcast (&fs->cond);
	}
	pthread_mutex_unlock (&fs->mutex);

	if (init_cbk)
		init_cbk (fs, ret);
out:
	return;
}

GFAPI_SYMVER_PRIVATE_DEFAULT(glfs_init_done, 3.4.0);


int
glfs_init_common (struct glfs *fs)
{
	int  ret = -1;

	ret = create_master (fs);
	if (ret)
		return ret;

	ret = gf_thread_create (&fs->poller, NULL, glfs_poller, fs);
	if (ret)
		return ret;

	ret = glfs_volumes_init (fs);
	if (ret)
		return ret;

	fs->dev_id = gf_dm_hashfn (fs->volname, strlen (fs->volname));
	return ret;
}


int
glfs_init_async (struct glfs *fs, glfs_init_cbk cbk)
{
	int  ret = -1;

	if (!fs || !fs->ctx) {
		gf_msg ("glfs", GF_LOG_ERROR, EINVAL, API_MSG_INVALID_ENTRY,
			"fs is not properly initialized.");
		errno = EINVAL;
		return ret;
	}

	fs->init_cbk = cbk;

	ret = glfs_init_common (fs);

	return ret;
}


int
pub_glfs_init (struct glfs *fs)
{
	int  ret = -1;

        DECLARE_OLD_THIS;

	if (!fs || !fs->ctx) {
		gf_msg ("glfs", GF_LOG_ERROR, EINVAL, API_MSG_INVALID_ENTRY,
			"fs is not properly initialized.");
		errno = EINVAL;
		return ret;
	}

        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

	ret = glfs_init_common (fs);
	if (ret)
		goto out;

	ret = glfs_init_wait (fs);
out:
        __GLFS_EXIT_FS;

        /* Set the initial current working directory to "/" */
        if (ret >= 0) {
                ret = glfs_chdir (fs, "/");
        }

invalid_fs:
	return ret;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_init, 3.4.0);

static int
glusterfs_ctx_destroy (glusterfs_ctx_t *ctx)
{
        call_pool_t       *pool            = NULL;
        int               ret              = 0;
        glusterfs_graph_t *trav_graph      = NULL;
        glusterfs_graph_t *tmp             = NULL;

        if (ctx == NULL)
                return 0;

        /* For all the graphs, crawl through the xlator_t structs and free
         * all its members except for the mem_acct member,
         * as GF_FREE will be referencing it.
         */
        list_for_each_entry_safe (trav_graph, tmp, &ctx->graphs, list) {
                xlator_tree_free_members (trav_graph->first);
        }

        /* Free the memory pool */
        if (ctx->stub_mem_pool)
                mem_pool_destroy (ctx->stub_mem_pool);
        if (ctx->dict_pool)
                mem_pool_destroy (ctx->dict_pool);
        if (ctx->dict_data_pool)
                mem_pool_destroy (ctx->dict_data_pool);
        if (ctx->dict_pair_pool)
                mem_pool_destroy (ctx->dict_pair_pool);
        if (ctx->logbuf_pool)
                mem_pool_destroy (ctx->logbuf_pool);

        pool = ctx->pool;
        if (pool) {
                if (pool->frame_mem_pool)
                        mem_pool_destroy (pool->frame_mem_pool);
                if (pool->stack_mem_pool)
                        mem_pool_destroy (pool->stack_mem_pool);
                LOCK_DESTROY (&pool->lock);
                GF_FREE (pool);
        }

        /* Free the event pool */
        ret = event_pool_destroy (ctx->event_pool);

        /* Free the iobuf pool */
        iobuf_pool_destroy (ctx->iobuf_pool);

        GF_FREE (ctx->process_uuid);
        GF_FREE (ctx->cmd_args.volfile_id);

        pthread_mutex_destroy (&(ctx->lock));
        pthread_mutex_destroy (&(ctx->notify_lock));
        pthread_cond_destroy (&(ctx->notify_cond));

        /* Free all the graph structs and its containing xlator_t structs
         * from this point there should be no reference to GF_FREE/GF_CALLOC
         * as it will try to access mem_acct and the below funtion would
         * have freed the same.
         */
        list_for_each_entry_safe (trav_graph, tmp, &ctx->graphs, list) {
                glusterfs_graph_destroy_residual (trav_graph);
        }

        FREE (ctx);

        return ret;
}

int
pub_glfs_fini (struct glfs *fs)
{
        int                ret = -1;
        int                countdown = 100;
        xlator_t           *subvol = NULL;
        glusterfs_ctx_t    *ctx = NULL;
        glusterfs_graph_t  *graph = NULL;
        call_pool_t        *call_pool = NULL;
        int                fs_init = 0;
        int                err = -1;

        DECLARE_OLD_THIS;

        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

        ctx = fs->ctx;
        if (!ctx) {
                goto free_fs;
        }

        if (ctx->mgmt) {
                rpc_clnt_disable (ctx->mgmt);
                ctx->mgmt = NULL;
        }

        call_pool = fs->ctx->pool;

        while (countdown--) {
                /* give some time for background frames to finish */
                pthread_mutex_lock (&fs->mutex);
                {
                        /* Do we need to increase countdown? */
                        if ((!call_pool->cnt) && (!fs->pin_refcnt)) {
                                gf_msg_trace ("glfs", 0,
                                        "call_pool_cnt - %ld,"
                                        "pin_refcnt - %d",
                                        call_pool->cnt, fs->pin_refcnt);

                                ctx->cleanup_started = 1;
                                pthread_mutex_unlock (&fs->mutex);
                                break;
                        }
                }
                pthread_mutex_unlock (&fs->mutex);
                usleep (100000);
        }

        /* leaked frames may exist, we ignore */

        /*We deem glfs_fini as successful if there are no pending frames in the call
         *pool*/
        ret = (call_pool->cnt == 0)? 0: -1;

        pthread_mutex_lock (&fs->mutex);
        {
                fs_init = fs->init;
        }
        pthread_mutex_unlock (&fs->mutex);

        if (fs_init != 0) {
                subvol = glfs_active_subvol (fs);
                if (subvol) {
                        /* PARENT_DOWN within glfs_subvol_done() is issued
                           only on graph switch (new graph should activiate
                           and decrement the extra @winds count taken in
                           glfs_graph_setup()

                           Since we are explicitly destroying,
                           PARENT_DOWN is necessary
                        */
                        xlator_notify (subvol, GF_EVENT_PARENT_DOWN, subvol, 0);
                        /* Here we wait for GF_EVENT_CHILD_DOWN before exiting,
                           in case of asynchrnous cleanup
                        */
                        graph = subvol->graph;
                        err = pthread_mutex_lock (&fs->mutex);
                        if (err != 0) {
                                gf_msg ("glfs", GF_LOG_ERROR, err,
                                        API_MSG_FSMUTEX_LOCK_FAILED,
                                        "pthread lock on glfs mutex, "
                                        "returned error: (%s)", strerror (err));
                                goto fail;
                        }
                        /* check and wait for CHILD_DOWN for active subvol*/
                        {
                                while (graph->used) {
                                        err = pthread_cond_wait (&fs->child_down_cond,
                                                                 &fs->mutex);
                                        if (err != 0)
                                                gf_msg ("glfs", GF_LOG_INFO, err,
                                                        API_MSG_COND_WAIT_FAILED,
                                                        "%s cond wait failed %s",
                                                        subvol->name,
                                                        strerror (err));
                                }
                        }

                        err = pthread_mutex_unlock (&fs->mutex);
                        if (err != 0) {
                                gf_msg ("glfs", GF_LOG_ERROR, err,
                                        API_MSG_FSMUTEX_UNLOCK_FAILED,
                                        "pthread unlock on glfs mutex, "
                                        "returned error: (%s)", strerror (err));
                                goto fail;
                        }
                }
                glfs_subvol_done (fs, subvol);
        }

        ctx->cleanup_started = 1;

        if (fs_init != 0) {
                /* Destroy all the inode tables of all the graphs.
                 * NOTE:
                 * - inode objects should be destroyed before calling fini()
                 *   of each xlator, as fini() and forget() of the xlators
                 *   can share few common locks or data structures, calling
                 *   fini first might destroy those required by forget
                 *   ( eg: in quick-read)
                 * - The call to inode_table_destroy_all is not required when
                 *   the cleanup during graph switch is implemented to perform
                 *   inode table destroy.
                 */
                inode_table_destroy_all (ctx);

                /* Call fini() of all the xlators in the active graph
                 * NOTE:
                 * - xlator fini() should be called before destroying any of
                 *   the threads. (eg: fini() in protocol-client uses timer
                 *   thread) */
                glusterfs_graph_deactivate (ctx->active);

                /* Join the syncenv_processor threads and cleanup
                 * syncenv resources*/
                syncenv_destroy (ctx->env);

                /* Join the poller thread */
                if (event_dispatch_destroy (ctx->event_pool) != 0)
                        ret = -1;
        }

        /* log infra has to be brought down before destroying
         * timer registry, as logging uses timer infra
         */
        if (gf_log_fini (ctx) != 0)
                ret = -1;

        /* Join the timer thread */
        if (fs_init != 0) {
                gf_timer_registry_destroy (ctx);
        }

        /* Destroy the context and the global pools */
        if (glusterfs_ctx_destroy (ctx) != 0)
                ret = -1;

free_fs:
        glfs_free_from_ctx (fs);

fail:
        if (!ret)
                ret = err;

        __GLFS_EXIT_FS;

invalid_fs:
        return ret;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_fini, 3.4.0);


ssize_t
pub_glfs_get_volfile (struct glfs *fs, void *buf, size_t len)
{
        ssize_t         res = -1;

        DECLARE_OLD_THIS;
        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

        glfs_lock(fs);
        if (len >= fs->oldvollen) {
                gf_msg_trace ("glfs", 0, "copying %zu to %p", len, buf);
                memcpy(buf,fs->oldvolfile,len);
                res = len;
        }
        else {
                res = len - fs->oldvollen;
                gf_msg_trace ("glfs", 0, "buffer is %zd too short", -res);
        }
        glfs_unlock(fs);

        __GLFS_EXIT_FS;

invalid_fs:
        return res;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_get_volfile, 3.6.0);

int
pub_glfs_ipc (struct glfs *fs, int opcode)
{
	xlator_t        *subvol = NULL;
        int             ret = -1;

	DECLARE_OLD_THIS;
        __GLFS_ENTRY_VALIDATE_FS (fs, invalid_fs);

	subvol = glfs_active_subvol (fs);
	if (!subvol) {
		ret = -1;
		errno = EIO;
		goto out;
	}

	ret = syncop_ipc (subvol, opcode, NULL, NULL);
        DECODE_SYNCOP_ERR (ret);

out:
        glfs_subvol_done (fs, subvol);
        __GLFS_EXIT_FS;

invalid_fs:
        return ret;
}

GFAPI_SYMVER_PUBLIC_DEFAULT(glfs_ipc, 3.7.0);