xref: /freebsd/sbin/hastd/primary.c (revision 3e0efd2ec4fcb4cd68fb8ccf8aea6fc6151c454b)
1 /*-
2  * Copyright (c) 2009 The FreeBSD Foundation
3  * Copyright (c) 2010-2011 Pawel Jakub Dawidek <pawel@dawidek.net>
4  * All rights reserved.
5  *
6  * This software was developed by Pawel Jakub Dawidek under sponsorship from
7  * the FreeBSD Foundation.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  *
18  * THIS SOFTWARE IS PROVIDED BY THE AUTHORS AND CONTRIBUTORS ``AS IS'' AND
19  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE LIABLE
22  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
23  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
24  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
25  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
27  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28  * SUCH DAMAGE.
29  */
30 
31 #include <sys/cdefs.h>
32 __FBSDID("$FreeBSD$");
33 
34 #include <sys/types.h>
35 #include <sys/time.h>
36 #include <sys/bio.h>
37 #include <sys/disk.h>
38 #include <sys/refcount.h>
39 #include <sys/stat.h>
40 
41 #include <geom/gate/g_gate.h>
42 
43 #include <err.h>
44 #include <errno.h>
45 #include <fcntl.h>
46 #include <libgeom.h>
47 #include <pthread.h>
48 #include <signal.h>
49 #include <stdint.h>
50 #include <stdio.h>
51 #include <string.h>
52 #include <sysexits.h>
53 #include <unistd.h>
54 
55 #include <activemap.h>
56 #include <nv.h>
57 #include <rangelock.h>
58 
59 #include "control.h"
60 #include "event.h"
61 #include "hast.h"
62 #include "hast_proto.h"
63 #include "hastd.h"
64 #include "hooks.h"
65 #include "metadata.h"
66 #include "proto.h"
67 #include "pjdlog.h"
68 #include "subr.h"
69 #include "synch.h"
70 
71 /* The is only one remote component for now. */
72 #define	ISREMOTE(no)	((no) == 1)
73 
74 struct hio {
75 	/*
76 	 * Number of components we are still waiting for.
77 	 * When this field goes to 0, we can send the request back to the
78 	 * kernel. Each component has to decrease this counter by one
79 	 * even on failure.
80 	 */
81 	unsigned int		 hio_countdown;
82 	/*
83 	 * Each component has a place to store its own error.
84 	 * Once the request is handled by all components we can decide if the
85 	 * request overall is successful or not.
86 	 */
87 	int			*hio_errors;
88 	/*
89 	 * Structure used to communicate with GEOM Gate class.
90 	 */
91 	struct g_gate_ctl_io	 hio_ggio;
92 	/*
93 	 * Request was already confirmed to GEOM Gate.
94 	 */
95 	bool			 hio_done;
96 	/*
97 	 * Remember replication from the time the request was initiated,
98 	 * so we won't get confused when replication changes on reload.
99 	 */
100 	int			 hio_replication;
101 	TAILQ_ENTRY(hio)	*hio_next;
102 };
103 #define	hio_free_next	hio_next[0]
104 #define	hio_done_next	hio_next[0]
105 
106 /*
107  * Free list holds unused structures. When free list is empty, we have to wait
108  * until some in-progress requests are freed.
109  */
110 static TAILQ_HEAD(, hio) hio_free_list;
111 static pthread_mutex_t hio_free_list_lock;
112 static pthread_cond_t hio_free_list_cond;
113 /*
114  * There is one send list for every component. One requests is placed on all
115  * send lists - each component gets the same request, but each component is
116  * responsible for managing his own send list.
117  */
118 static TAILQ_HEAD(, hio) *hio_send_list;
119 static pthread_mutex_t *hio_send_list_lock;
120 static pthread_cond_t *hio_send_list_cond;
121 /*
122  * There is one recv list for every component, although local components don't
123  * use recv lists as local requests are done synchronously.
124  */
125 static TAILQ_HEAD(, hio) *hio_recv_list;
126 static pthread_mutex_t *hio_recv_list_lock;
127 static pthread_cond_t *hio_recv_list_cond;
128 /*
129  * Request is placed on done list by the slowest component (the one that
130  * decreased hio_countdown from 1 to 0).
131  */
132 static TAILQ_HEAD(, hio) hio_done_list;
133 static pthread_mutex_t hio_done_list_lock;
134 static pthread_cond_t hio_done_list_cond;
135 /*
136  * Structure below are for interaction with sync thread.
137  */
138 static bool sync_inprogress;
139 static pthread_mutex_t sync_lock;
140 static pthread_cond_t sync_cond;
141 /*
142  * The lock below allows to synchornize access to remote connections.
143  */
144 static pthread_rwlock_t *hio_remote_lock;
145 
146 /*
147  * Lock to synchronize metadata updates. Also synchronize access to
148  * hr_primary_localcnt and hr_primary_remotecnt fields.
149  */
150 static pthread_mutex_t metadata_lock;
151 
152 /*
153  * Maximum number of outstanding I/O requests.
154  */
155 #define	HAST_HIO_MAX	256
156 /*
157  * Number of components. At this point there are only two components: local
158  * and remote, but in the future it might be possible to use multiple local
159  * and remote components.
160  */
161 #define	HAST_NCOMPONENTS	2
162 
163 #define	ISCONNECTED(res, no)	\
164 	((res)->hr_remotein != NULL && (res)->hr_remoteout != NULL)
165 
166 #define	QUEUE_INSERT1(hio, name, ncomp)	do {				\
167 	bool _wakeup;							\
168 									\
169 	mtx_lock(&hio_##name##_list_lock[(ncomp)]);			\
170 	_wakeup = TAILQ_EMPTY(&hio_##name##_list[(ncomp)]);		\
171 	TAILQ_INSERT_TAIL(&hio_##name##_list[(ncomp)], (hio),		\
172 	    hio_next[(ncomp)]);						\
173 	mtx_unlock(&hio_##name##_list_lock[ncomp]);			\
174 	if (_wakeup)							\
175 		cv_signal(&hio_##name##_list_cond[(ncomp)]);		\
176 } while (0)
177 #define	QUEUE_INSERT2(hio, name)	do {				\
178 	bool _wakeup;							\
179 									\
180 	mtx_lock(&hio_##name##_list_lock);				\
181 	_wakeup = TAILQ_EMPTY(&hio_##name##_list);			\
182 	TAILQ_INSERT_TAIL(&hio_##name##_list, (hio), hio_##name##_next);\
183 	mtx_unlock(&hio_##name##_list_lock);				\
184 	if (_wakeup)							\
185 		cv_signal(&hio_##name##_list_cond);			\
186 } while (0)
187 #define	QUEUE_TAKE1(hio, name, ncomp, timeout)	do {			\
188 	bool _last;							\
189 									\
190 	mtx_lock(&hio_##name##_list_lock[(ncomp)]);			\
191 	_last = false;							\
192 	while (((hio) = TAILQ_FIRST(&hio_##name##_list[(ncomp)])) == NULL && !_last) { \
193 		cv_timedwait(&hio_##name##_list_cond[(ncomp)],		\
194 		    &hio_##name##_list_lock[(ncomp)], (timeout));	\
195 		if ((timeout) != 0)					\
196 			_last = true;					\
197 	}								\
198 	if (hio != NULL) {						\
199 		TAILQ_REMOVE(&hio_##name##_list[(ncomp)], (hio),	\
200 		    hio_next[(ncomp)]);					\
201 	}								\
202 	mtx_unlock(&hio_##name##_list_lock[(ncomp)]);			\
203 } while (0)
204 #define	QUEUE_TAKE2(hio, name)	do {					\
205 	mtx_lock(&hio_##name##_list_lock);				\
206 	while (((hio) = TAILQ_FIRST(&hio_##name##_list)) == NULL) {	\
207 		cv_wait(&hio_##name##_list_cond,			\
208 		    &hio_##name##_list_lock);				\
209 	}								\
210 	TAILQ_REMOVE(&hio_##name##_list, (hio), hio_##name##_next);	\
211 	mtx_unlock(&hio_##name##_list_lock);				\
212 } while (0)
213 
214 #define	SYNCREQ(hio)		do {					\
215 	(hio)->hio_ggio.gctl_unit = -1;					\
216 	(hio)->hio_ggio.gctl_seq = 1;					\
217 } while (0)
218 #define	ISSYNCREQ(hio)		((hio)->hio_ggio.gctl_unit == -1)
219 #define	SYNCREQDONE(hio)	do { (hio)->hio_ggio.gctl_unit = -2; } while (0)
220 #define	ISSYNCREQDONE(hio)	((hio)->hio_ggio.gctl_unit == -2)
221 
222 static struct hast_resource *gres;
223 
224 static pthread_mutex_t range_lock;
225 static struct rangelocks *range_regular;
226 static bool range_regular_wait;
227 static pthread_cond_t range_regular_cond;
228 static struct rangelocks *range_sync;
229 static bool range_sync_wait;
230 static pthread_cond_t range_sync_cond;
231 static bool fullystarted;
232 
233 static void *ggate_recv_thread(void *arg);
234 static void *local_send_thread(void *arg);
235 static void *remote_send_thread(void *arg);
236 static void *remote_recv_thread(void *arg);
237 static void *ggate_send_thread(void *arg);
238 static void *sync_thread(void *arg);
239 static void *guard_thread(void *arg);
240 
241 static void
242 cleanup(struct hast_resource *res)
243 {
244 	int rerrno;
245 
246 	/* Remember errno. */
247 	rerrno = errno;
248 
249 	/* Destroy ggate provider if we created one. */
250 	if (res->hr_ggateunit >= 0) {
251 		struct g_gate_ctl_destroy ggiod;
252 
253 		bzero(&ggiod, sizeof(ggiod));
254 		ggiod.gctl_version = G_GATE_VERSION;
255 		ggiod.gctl_unit = res->hr_ggateunit;
256 		ggiod.gctl_force = 1;
257 		if (ioctl(res->hr_ggatefd, G_GATE_CMD_DESTROY, &ggiod) == -1) {
258 			pjdlog_errno(LOG_WARNING,
259 			    "Unable to destroy hast/%s device",
260 			    res->hr_provname);
261 		}
262 		res->hr_ggateunit = -1;
263 	}
264 
265 	/* Restore errno. */
266 	errno = rerrno;
267 }
268 
269 static __dead2 void
270 primary_exit(int exitcode, const char *fmt, ...)
271 {
272 	va_list ap;
273 
274 	PJDLOG_ASSERT(exitcode != EX_OK);
275 	va_start(ap, fmt);
276 	pjdlogv_errno(LOG_ERR, fmt, ap);
277 	va_end(ap);
278 	cleanup(gres);
279 	exit(exitcode);
280 }
281 
282 static __dead2 void
283 primary_exitx(int exitcode, const char *fmt, ...)
284 {
285 	va_list ap;
286 
287 	va_start(ap, fmt);
288 	pjdlogv(exitcode == EX_OK ? LOG_INFO : LOG_ERR, fmt, ap);
289 	va_end(ap);
290 	cleanup(gres);
291 	exit(exitcode);
292 }
293 
294 static int
295 hast_activemap_flush(struct hast_resource *res)
296 {
297 	const unsigned char *buf;
298 	size_t size;
299 
300 	buf = activemap_bitmap(res->hr_amp, &size);
301 	PJDLOG_ASSERT(buf != NULL);
302 	PJDLOG_ASSERT((size % res->hr_local_sectorsize) == 0);
303 	if (pwrite(res->hr_localfd, buf, size, METADATA_SIZE) !=
304 	    (ssize_t)size) {
305 		pjdlog_errno(LOG_ERR, "Unable to flush activemap to disk");
306 		return (-1);
307 	}
308 	if (res->hr_metaflush == 1 && g_flush(res->hr_localfd) == -1) {
309 		if (errno == EOPNOTSUPP) {
310 			pjdlog_warning("The %s provider doesn't support flushing write cache. Disabling it.",
311 			    res->hr_localpath);
312 			res->hr_metaflush = 0;
313 		} else {
314 			pjdlog_errno(LOG_ERR,
315 			    "Unable to flush disk cache on activemap update");
316 			return (-1);
317 		}
318 	}
319 	return (0);
320 }
321 
322 static bool
323 real_remote(const struct hast_resource *res)
324 {
325 
326 	return (strcmp(res->hr_remoteaddr, "none") != 0);
327 }
328 
329 static void
330 init_environment(struct hast_resource *res __unused)
331 {
332 	struct hio *hio;
333 	unsigned int ii, ncomps;
334 
335 	/*
336 	 * In the future it might be per-resource value.
337 	 */
338 	ncomps = HAST_NCOMPONENTS;
339 
340 	/*
341 	 * Allocate memory needed by lists.
342 	 */
343 	hio_send_list = malloc(sizeof(hio_send_list[0]) * ncomps);
344 	if (hio_send_list == NULL) {
345 		primary_exitx(EX_TEMPFAIL,
346 		    "Unable to allocate %zu bytes of memory for send lists.",
347 		    sizeof(hio_send_list[0]) * ncomps);
348 	}
349 	hio_send_list_lock = malloc(sizeof(hio_send_list_lock[0]) * ncomps);
350 	if (hio_send_list_lock == NULL) {
351 		primary_exitx(EX_TEMPFAIL,
352 		    "Unable to allocate %zu bytes of memory for send list locks.",
353 		    sizeof(hio_send_list_lock[0]) * ncomps);
354 	}
355 	hio_send_list_cond = malloc(sizeof(hio_send_list_cond[0]) * ncomps);
356 	if (hio_send_list_cond == NULL) {
357 		primary_exitx(EX_TEMPFAIL,
358 		    "Unable to allocate %zu bytes of memory for send list condition variables.",
359 		    sizeof(hio_send_list_cond[0]) * ncomps);
360 	}
361 	hio_recv_list = malloc(sizeof(hio_recv_list[0]) * ncomps);
362 	if (hio_recv_list == NULL) {
363 		primary_exitx(EX_TEMPFAIL,
364 		    "Unable to allocate %zu bytes of memory for recv lists.",
365 		    sizeof(hio_recv_list[0]) * ncomps);
366 	}
367 	hio_recv_list_lock = malloc(sizeof(hio_recv_list_lock[0]) * ncomps);
368 	if (hio_recv_list_lock == NULL) {
369 		primary_exitx(EX_TEMPFAIL,
370 		    "Unable to allocate %zu bytes of memory for recv list locks.",
371 		    sizeof(hio_recv_list_lock[0]) * ncomps);
372 	}
373 	hio_recv_list_cond = malloc(sizeof(hio_recv_list_cond[0]) * ncomps);
374 	if (hio_recv_list_cond == NULL) {
375 		primary_exitx(EX_TEMPFAIL,
376 		    "Unable to allocate %zu bytes of memory for recv list condition variables.",
377 		    sizeof(hio_recv_list_cond[0]) * ncomps);
378 	}
379 	hio_remote_lock = malloc(sizeof(hio_remote_lock[0]) * ncomps);
380 	if (hio_remote_lock == NULL) {
381 		primary_exitx(EX_TEMPFAIL,
382 		    "Unable to allocate %zu bytes of memory for remote connections locks.",
383 		    sizeof(hio_remote_lock[0]) * ncomps);
384 	}
385 
386 	/*
387 	 * Initialize lists, their locks and theirs condition variables.
388 	 */
389 	TAILQ_INIT(&hio_free_list);
390 	mtx_init(&hio_free_list_lock);
391 	cv_init(&hio_free_list_cond);
392 	for (ii = 0; ii < HAST_NCOMPONENTS; ii++) {
393 		TAILQ_INIT(&hio_send_list[ii]);
394 		mtx_init(&hio_send_list_lock[ii]);
395 		cv_init(&hio_send_list_cond[ii]);
396 		TAILQ_INIT(&hio_recv_list[ii]);
397 		mtx_init(&hio_recv_list_lock[ii]);
398 		cv_init(&hio_recv_list_cond[ii]);
399 		rw_init(&hio_remote_lock[ii]);
400 	}
401 	TAILQ_INIT(&hio_done_list);
402 	mtx_init(&hio_done_list_lock);
403 	cv_init(&hio_done_list_cond);
404 	mtx_init(&metadata_lock);
405 
406 	/*
407 	 * Allocate requests pool and initialize requests.
408 	 */
409 	for (ii = 0; ii < HAST_HIO_MAX; ii++) {
410 		hio = malloc(sizeof(*hio));
411 		if (hio == NULL) {
412 			primary_exitx(EX_TEMPFAIL,
413 			    "Unable to allocate %zu bytes of memory for hio request.",
414 			    sizeof(*hio));
415 		}
416 		hio->hio_countdown = 0;
417 		hio->hio_errors = malloc(sizeof(hio->hio_errors[0]) * ncomps);
418 		if (hio->hio_errors == NULL) {
419 			primary_exitx(EX_TEMPFAIL,
420 			    "Unable allocate %zu bytes of memory for hio errors.",
421 			    sizeof(hio->hio_errors[0]) * ncomps);
422 		}
423 		hio->hio_next = malloc(sizeof(hio->hio_next[0]) * ncomps);
424 		if (hio->hio_next == NULL) {
425 			primary_exitx(EX_TEMPFAIL,
426 			    "Unable allocate %zu bytes of memory for hio_next field.",
427 			    sizeof(hio->hio_next[0]) * ncomps);
428 		}
429 		hio->hio_ggio.gctl_version = G_GATE_VERSION;
430 		hio->hio_ggio.gctl_data = malloc(MAXPHYS);
431 		if (hio->hio_ggio.gctl_data == NULL) {
432 			primary_exitx(EX_TEMPFAIL,
433 			    "Unable to allocate %zu bytes of memory for gctl_data.",
434 			    MAXPHYS);
435 		}
436 		hio->hio_ggio.gctl_length = MAXPHYS;
437 		hio->hio_ggio.gctl_error = 0;
438 		TAILQ_INSERT_HEAD(&hio_free_list, hio, hio_free_next);
439 	}
440 }
441 
442 static bool
443 init_resuid(struct hast_resource *res)
444 {
445 
446 	mtx_lock(&metadata_lock);
447 	if (res->hr_resuid != 0) {
448 		mtx_unlock(&metadata_lock);
449 		return (false);
450 	} else {
451 		/* Initialize unique resource identifier. */
452 		arc4random_buf(&res->hr_resuid, sizeof(res->hr_resuid));
453 		mtx_unlock(&metadata_lock);
454 		if (metadata_write(res) == -1)
455 			exit(EX_NOINPUT);
456 		return (true);
457 	}
458 }
459 
460 static void
461 init_local(struct hast_resource *res)
462 {
463 	unsigned char *buf;
464 	size_t mapsize;
465 
466 	if (metadata_read(res, true) == -1)
467 		exit(EX_NOINPUT);
468 	mtx_init(&res->hr_amp_lock);
469 	if (activemap_init(&res->hr_amp, res->hr_datasize, res->hr_extentsize,
470 	    res->hr_local_sectorsize, res->hr_keepdirty) == -1) {
471 		primary_exit(EX_TEMPFAIL, "Unable to create activemap");
472 	}
473 	mtx_init(&range_lock);
474 	cv_init(&range_regular_cond);
475 	if (rangelock_init(&range_regular) == -1)
476 		primary_exit(EX_TEMPFAIL, "Unable to create regular range lock");
477 	cv_init(&range_sync_cond);
478 	if (rangelock_init(&range_sync) == -1)
479 		primary_exit(EX_TEMPFAIL, "Unable to create sync range lock");
480 	mapsize = activemap_ondisk_size(res->hr_amp);
481 	buf = calloc(1, mapsize);
482 	if (buf == NULL) {
483 		primary_exitx(EX_TEMPFAIL,
484 		    "Unable to allocate buffer for activemap.");
485 	}
486 	if (pread(res->hr_localfd, buf, mapsize, METADATA_SIZE) !=
487 	    (ssize_t)mapsize) {
488 		primary_exit(EX_NOINPUT, "Unable to read activemap");
489 	}
490 	activemap_copyin(res->hr_amp, buf, mapsize);
491 	free(buf);
492 	if (res->hr_resuid != 0)
493 		return;
494 	/*
495 	 * We're using provider for the first time. Initialize local and remote
496 	 * counters. We don't initialize resuid here, as we want to do it just
497 	 * in time. The reason for this is that we want to inform secondary
498 	 * that there were no writes yet, so there is no need to synchronize
499 	 * anything.
500 	 */
501 	res->hr_primary_localcnt = 0;
502 	res->hr_primary_remotecnt = 0;
503 	if (metadata_write(res) == -1)
504 		exit(EX_NOINPUT);
505 }
506 
507 static int
508 primary_connect(struct hast_resource *res, struct proto_conn **connp)
509 {
510 	struct proto_conn *conn;
511 	int16_t val;
512 
513 	val = 1;
514 	if (proto_send(res->hr_conn, &val, sizeof(val)) == -1) {
515 		primary_exit(EX_TEMPFAIL,
516 		    "Unable to send connection request to parent");
517 	}
518 	if (proto_recv(res->hr_conn, &val, sizeof(val)) == -1) {
519 		primary_exit(EX_TEMPFAIL,
520 		    "Unable to receive reply to connection request from parent");
521 	}
522 	if (val != 0) {
523 		errno = val;
524 		pjdlog_errno(LOG_WARNING, "Unable to connect to %s",
525 		    res->hr_remoteaddr);
526 		return (-1);
527 	}
528 	if (proto_connection_recv(res->hr_conn, true, &conn) == -1) {
529 		primary_exit(EX_TEMPFAIL,
530 		    "Unable to receive connection from parent");
531 	}
532 	if (proto_connect_wait(conn, res->hr_timeout) == -1) {
533 		pjdlog_errno(LOG_WARNING, "Unable to connect to %s",
534 		    res->hr_remoteaddr);
535 		proto_close(conn);
536 		return (-1);
537 	}
538 	/* Error in setting timeout is not critical, but why should it fail? */
539 	if (proto_timeout(conn, res->hr_timeout) == -1)
540 		pjdlog_errno(LOG_WARNING, "Unable to set connection timeout");
541 
542 	*connp = conn;
543 
544 	return (0);
545 }
546 
547 static int
548 init_remote(struct hast_resource *res, struct proto_conn **inp,
549     struct proto_conn **outp)
550 {
551 	struct proto_conn *in, *out;
552 	struct nv *nvout, *nvin;
553 	const unsigned char *token;
554 	unsigned char *map;
555 	const char *errmsg;
556 	int32_t extentsize;
557 	int64_t datasize;
558 	uint32_t mapsize;
559 	size_t size;
560 	int error;
561 
562 	PJDLOG_ASSERT((inp == NULL && outp == NULL) || (inp != NULL && outp != NULL));
563 	PJDLOG_ASSERT(real_remote(res));
564 
565 	in = out = NULL;
566 	errmsg = NULL;
567 
568 	if (primary_connect(res, &out) == -1)
569 		return (ECONNREFUSED);
570 
571 	error = ECONNABORTED;
572 
573 	/*
574 	 * First handshake step.
575 	 * Setup outgoing connection with remote node.
576 	 */
577 	nvout = nv_alloc();
578 	nv_add_string(nvout, res->hr_name, "resource");
579 	if (nv_error(nvout) != 0) {
580 		pjdlog_common(LOG_WARNING, 0, nv_error(nvout),
581 		    "Unable to allocate header for connection with %s",
582 		    res->hr_remoteaddr);
583 		nv_free(nvout);
584 		goto close;
585 	}
586 	if (hast_proto_send(res, out, nvout, NULL, 0) == -1) {
587 		pjdlog_errno(LOG_WARNING,
588 		    "Unable to send handshake header to %s",
589 		    res->hr_remoteaddr);
590 		nv_free(nvout);
591 		goto close;
592 	}
593 	nv_free(nvout);
594 	if (hast_proto_recv_hdr(out, &nvin) == -1) {
595 		pjdlog_errno(LOG_WARNING,
596 		    "Unable to receive handshake header from %s",
597 		    res->hr_remoteaddr);
598 		goto close;
599 	}
600 	errmsg = nv_get_string(nvin, "errmsg");
601 	if (errmsg != NULL) {
602 		pjdlog_warning("%s", errmsg);
603 		if (nv_exists(nvin, "wait"))
604 			error = EBUSY;
605 		nv_free(nvin);
606 		goto close;
607 	}
608 	token = nv_get_uint8_array(nvin, &size, "token");
609 	if (token == NULL) {
610 		pjdlog_warning("Handshake header from %s has no 'token' field.",
611 		    res->hr_remoteaddr);
612 		nv_free(nvin);
613 		goto close;
614 	}
615 	if (size != sizeof(res->hr_token)) {
616 		pjdlog_warning("Handshake header from %s contains 'token' of wrong size (got %zu, expected %zu).",
617 		    res->hr_remoteaddr, size, sizeof(res->hr_token));
618 		nv_free(nvin);
619 		goto close;
620 	}
621 	bcopy(token, res->hr_token, sizeof(res->hr_token));
622 	nv_free(nvin);
623 
624 	/*
625 	 * Second handshake step.
626 	 * Setup incoming connection with remote node.
627 	 */
628 	if (primary_connect(res, &in) == -1)
629 		goto close;
630 
631 	nvout = nv_alloc();
632 	nv_add_string(nvout, res->hr_name, "resource");
633 	nv_add_uint8_array(nvout, res->hr_token, sizeof(res->hr_token),
634 	    "token");
635 	if (res->hr_resuid == 0) {
636 		/*
637 		 * The resuid field was not yet initialized.
638 		 * Because we do synchronization inside init_resuid(), it is
639 		 * possible that someone already initialized it, the function
640 		 * will return false then, but if we successfully initialized
641 		 * it, we will get true. True means that there were no writes
642 		 * to this resource yet and we want to inform secondary that
643 		 * synchronization is not needed by sending "virgin" argument.
644 		 */
645 		if (init_resuid(res))
646 			nv_add_int8(nvout, 1, "virgin");
647 	}
648 	nv_add_uint64(nvout, res->hr_resuid, "resuid");
649 	nv_add_uint64(nvout, res->hr_primary_localcnt, "localcnt");
650 	nv_add_uint64(nvout, res->hr_primary_remotecnt, "remotecnt");
651 	if (nv_error(nvout) != 0) {
652 		pjdlog_common(LOG_WARNING, 0, nv_error(nvout),
653 		    "Unable to allocate header for connection with %s",
654 		    res->hr_remoteaddr);
655 		nv_free(nvout);
656 		goto close;
657 	}
658 	if (hast_proto_send(res, in, nvout, NULL, 0) == -1) {
659 		pjdlog_errno(LOG_WARNING,
660 		    "Unable to send handshake header to %s",
661 		    res->hr_remoteaddr);
662 		nv_free(nvout);
663 		goto close;
664 	}
665 	nv_free(nvout);
666 	if (hast_proto_recv_hdr(out, &nvin) == -1) {
667 		pjdlog_errno(LOG_WARNING,
668 		    "Unable to receive handshake header from %s",
669 		    res->hr_remoteaddr);
670 		goto close;
671 	}
672 	errmsg = nv_get_string(nvin, "errmsg");
673 	if (errmsg != NULL) {
674 		pjdlog_warning("%s", errmsg);
675 		nv_free(nvin);
676 		goto close;
677 	}
678 	datasize = nv_get_int64(nvin, "datasize");
679 	if (datasize != res->hr_datasize) {
680 		pjdlog_warning("Data size differs between nodes (local=%jd, remote=%jd).",
681 		    (intmax_t)res->hr_datasize, (intmax_t)datasize);
682 		nv_free(nvin);
683 		goto close;
684 	}
685 	extentsize = nv_get_int32(nvin, "extentsize");
686 	if (extentsize != res->hr_extentsize) {
687 		pjdlog_warning("Extent size differs between nodes (local=%zd, remote=%zd).",
688 		    (ssize_t)res->hr_extentsize, (ssize_t)extentsize);
689 		nv_free(nvin);
690 		goto close;
691 	}
692 	res->hr_secondary_localcnt = nv_get_uint64(nvin, "localcnt");
693 	res->hr_secondary_remotecnt = nv_get_uint64(nvin, "remotecnt");
694 	res->hr_syncsrc = nv_get_uint8(nvin, "syncsrc");
695 	if (nv_exists(nvin, "virgin")) {
696 		/*
697 		 * Secondary was reinitialized, bump localcnt if it is 0 as
698 		 * only we have the data.
699 		 */
700 		PJDLOG_ASSERT(res->hr_syncsrc == HAST_SYNCSRC_PRIMARY);
701 		PJDLOG_ASSERT(res->hr_secondary_localcnt == 0);
702 
703 		if (res->hr_primary_localcnt == 0) {
704 			PJDLOG_ASSERT(res->hr_secondary_remotecnt == 0);
705 
706 			mtx_lock(&metadata_lock);
707 			res->hr_primary_localcnt++;
708 			pjdlog_debug(1, "Increasing localcnt to %ju.",
709 			    (uintmax_t)res->hr_primary_localcnt);
710 			(void)metadata_write(res);
711 			mtx_unlock(&metadata_lock);
712 		}
713 	}
714 	map = NULL;
715 	mapsize = nv_get_uint32(nvin, "mapsize");
716 	if (mapsize > 0) {
717 		map = malloc(mapsize);
718 		if (map == NULL) {
719 			pjdlog_error("Unable to allocate memory for remote activemap (mapsize=%ju).",
720 			    (uintmax_t)mapsize);
721 			nv_free(nvin);
722 			goto close;
723 		}
724 		/*
725 		 * Remote node have some dirty extents on its own, lets
726 		 * download its activemap.
727 		 */
728 		if (hast_proto_recv_data(res, out, nvin, map,
729 		    mapsize) == -1) {
730 			pjdlog_errno(LOG_ERR,
731 			    "Unable to receive remote activemap");
732 			nv_free(nvin);
733 			free(map);
734 			goto close;
735 		}
736 		/*
737 		 * Merge local and remote bitmaps.
738 		 */
739 		activemap_merge(res->hr_amp, map, mapsize);
740 		free(map);
741 		/*
742 		 * Now that we merged bitmaps from both nodes, flush it to the
743 		 * disk before we start to synchronize.
744 		 */
745 		(void)hast_activemap_flush(res);
746 	}
747 	nv_free(nvin);
748 #ifdef notyet
749 	/* Setup directions. */
750 	if (proto_send(out, NULL, 0) == -1)
751 		pjdlog_errno(LOG_WARNING, "Unable to set connection direction");
752 	if (proto_recv(in, NULL, 0) == -1)
753 		pjdlog_errno(LOG_WARNING, "Unable to set connection direction");
754 #endif
755 	pjdlog_info("Connected to %s.", res->hr_remoteaddr);
756 	if (inp != NULL && outp != NULL) {
757 		*inp = in;
758 		*outp = out;
759 	} else {
760 		res->hr_remotein = in;
761 		res->hr_remoteout = out;
762 	}
763 	event_send(res, EVENT_CONNECT);
764 	return (0);
765 close:
766 	if (errmsg != NULL && strcmp(errmsg, "Split-brain condition!") == 0)
767 		event_send(res, EVENT_SPLITBRAIN);
768 	proto_close(out);
769 	if (in != NULL)
770 		proto_close(in);
771 	return (error);
772 }
773 
774 static void
775 sync_start(void)
776 {
777 
778 	mtx_lock(&sync_lock);
779 	sync_inprogress = true;
780 	mtx_unlock(&sync_lock);
781 	cv_signal(&sync_cond);
782 }
783 
784 static void
785 sync_stop(void)
786 {
787 
788 	mtx_lock(&sync_lock);
789 	if (sync_inprogress)
790 		sync_inprogress = false;
791 	mtx_unlock(&sync_lock);
792 }
793 
794 static void
795 init_ggate(struct hast_resource *res)
796 {
797 	struct g_gate_ctl_create ggiocreate;
798 	struct g_gate_ctl_cancel ggiocancel;
799 
800 	/*
801 	 * We communicate with ggate via /dev/ggctl. Open it.
802 	 */
803 	res->hr_ggatefd = open("/dev/" G_GATE_CTL_NAME, O_RDWR);
804 	if (res->hr_ggatefd == -1)
805 		primary_exit(EX_OSFILE, "Unable to open /dev/" G_GATE_CTL_NAME);
806 	/*
807 	 * Create provider before trying to connect, as connection failure
808 	 * is not critical, but may take some time.
809 	 */
810 	bzero(&ggiocreate, sizeof(ggiocreate));
811 	ggiocreate.gctl_version = G_GATE_VERSION;
812 	ggiocreate.gctl_mediasize = res->hr_datasize;
813 	ggiocreate.gctl_sectorsize = res->hr_local_sectorsize;
814 	ggiocreate.gctl_flags = 0;
815 	ggiocreate.gctl_maxcount = 0;
816 	ggiocreate.gctl_timeout = 0;
817 	ggiocreate.gctl_unit = G_GATE_NAME_GIVEN;
818 	snprintf(ggiocreate.gctl_name, sizeof(ggiocreate.gctl_name), "hast/%s",
819 	    res->hr_provname);
820 	if (ioctl(res->hr_ggatefd, G_GATE_CMD_CREATE, &ggiocreate) == 0) {
821 		pjdlog_info("Device hast/%s created.", res->hr_provname);
822 		res->hr_ggateunit = ggiocreate.gctl_unit;
823 		return;
824 	}
825 	if (errno != EEXIST) {
826 		primary_exit(EX_OSERR, "Unable to create hast/%s device",
827 		    res->hr_provname);
828 	}
829 	pjdlog_debug(1,
830 	    "Device hast/%s already exists, we will try to take it over.",
831 	    res->hr_provname);
832 	/*
833 	 * If we received EEXIST, we assume that the process who created the
834 	 * provider died and didn't clean up. In that case we will start from
835 	 * where he left of.
836 	 */
837 	bzero(&ggiocancel, sizeof(ggiocancel));
838 	ggiocancel.gctl_version = G_GATE_VERSION;
839 	ggiocancel.gctl_unit = G_GATE_NAME_GIVEN;
840 	snprintf(ggiocancel.gctl_name, sizeof(ggiocancel.gctl_name), "hast/%s",
841 	    res->hr_provname);
842 	if (ioctl(res->hr_ggatefd, G_GATE_CMD_CANCEL, &ggiocancel) == 0) {
843 		pjdlog_info("Device hast/%s recovered.", res->hr_provname);
844 		res->hr_ggateunit = ggiocancel.gctl_unit;
845 		return;
846 	}
847 	primary_exit(EX_OSERR, "Unable to take over hast/%s device",
848 	    res->hr_provname);
849 }
850 
851 void
852 hastd_primary(struct hast_resource *res)
853 {
854 	pthread_t td;
855 	pid_t pid;
856 	int error, mode, debuglevel;
857 
858 	/*
859 	 * Create communication channel for sending control commands from
860 	 * parent to child.
861 	 */
862 	if (proto_client(NULL, "socketpair://", &res->hr_ctrl) == -1) {
863 		/* TODO: There's no need for this to be fatal error. */
864 		KEEP_ERRNO((void)pidfile_remove(pfh));
865 		pjdlog_exit(EX_OSERR,
866 		    "Unable to create control sockets between parent and child");
867 	}
868 	/*
869 	 * Create communication channel for sending events from child to parent.
870 	 */
871 	if (proto_client(NULL, "socketpair://", &res->hr_event) == -1) {
872 		/* TODO: There's no need for this to be fatal error. */
873 		KEEP_ERRNO((void)pidfile_remove(pfh));
874 		pjdlog_exit(EX_OSERR,
875 		    "Unable to create event sockets between child and parent");
876 	}
877 	/*
878 	 * Create communication channel for sending connection requests from
879 	 * child to parent.
880 	 */
881 	if (proto_client(NULL, "socketpair://", &res->hr_conn) == -1) {
882 		/* TODO: There's no need for this to be fatal error. */
883 		KEEP_ERRNO((void)pidfile_remove(pfh));
884 		pjdlog_exit(EX_OSERR,
885 		    "Unable to create connection sockets between child and parent");
886 	}
887 
888 	pid = fork();
889 	if (pid == -1) {
890 		/* TODO: There's no need for this to be fatal error. */
891 		KEEP_ERRNO((void)pidfile_remove(pfh));
892 		pjdlog_exit(EX_TEMPFAIL, "Unable to fork");
893 	}
894 
895 	if (pid > 0) {
896 		/* This is parent. */
897 		/* Declare that we are receiver. */
898 		proto_recv(res->hr_event, NULL, 0);
899 		proto_recv(res->hr_conn, NULL, 0);
900 		/* Declare that we are sender. */
901 		proto_send(res->hr_ctrl, NULL, 0);
902 		res->hr_workerpid = pid;
903 		return;
904 	}
905 
906 	gres = res;
907 	mode = pjdlog_mode_get();
908 	debuglevel = pjdlog_debug_get();
909 
910 	/* Declare that we are sender. */
911 	proto_send(res->hr_event, NULL, 0);
912 	proto_send(res->hr_conn, NULL, 0);
913 	/* Declare that we are receiver. */
914 	proto_recv(res->hr_ctrl, NULL, 0);
915 	descriptors_cleanup(res);
916 
917 	descriptors_assert(res, mode);
918 
919 	pjdlog_init(mode);
920 	pjdlog_debug_set(debuglevel);
921 	pjdlog_prefix_set("[%s] (%s) ", res->hr_name, role2str(res->hr_role));
922 	setproctitle("%s (%s)", res->hr_name, role2str(res->hr_role));
923 
924 	init_local(res);
925 	init_ggate(res);
926 	init_environment(res);
927 
928 	if (drop_privs(res) != 0) {
929 		cleanup(res);
930 		exit(EX_CONFIG);
931 	}
932 	pjdlog_info("Privileges successfully dropped.");
933 
934 	/*
935 	 * Create the guard thread first, so we can handle signals from the
936 	 * very beginning.
937 	 */
938 	error = pthread_create(&td, NULL, guard_thread, res);
939 	PJDLOG_ASSERT(error == 0);
940 	/*
941 	 * Create the control thread before sending any event to the parent,
942 	 * as we can deadlock when parent sends control request to worker,
943 	 * but worker has no control thread started yet, so parent waits.
944 	 * In the meantime worker sends an event to the parent, but parent
945 	 * is unable to handle the event, because it waits for control
946 	 * request response.
947 	 */
948 	error = pthread_create(&td, NULL, ctrl_thread, res);
949 	PJDLOG_ASSERT(error == 0);
950 	if (real_remote(res)) {
951 		error = init_remote(res, NULL, NULL);
952 		if (error == 0) {
953 			sync_start();
954 		} else if (error == EBUSY) {
955 			time_t start = time(NULL);
956 
957 			pjdlog_warning("Waiting for remote node to become %s for %ds.",
958 			    role2str(HAST_ROLE_SECONDARY),
959 			    res->hr_timeout);
960 			for (;;) {
961 				sleep(1);
962 				error = init_remote(res, NULL, NULL);
963 				if (error != EBUSY)
964 					break;
965 				if (time(NULL) > start + res->hr_timeout)
966 					break;
967 			}
968 			if (error == EBUSY) {
969 				pjdlog_warning("Remote node is still %s, starting anyway.",
970 				    role2str(HAST_ROLE_PRIMARY));
971 			}
972 		}
973 	}
974 	error = pthread_create(&td, NULL, ggate_recv_thread, res);
975 	PJDLOG_ASSERT(error == 0);
976 	error = pthread_create(&td, NULL, local_send_thread, res);
977 	PJDLOG_ASSERT(error == 0);
978 	error = pthread_create(&td, NULL, remote_send_thread, res);
979 	PJDLOG_ASSERT(error == 0);
980 	error = pthread_create(&td, NULL, remote_recv_thread, res);
981 	PJDLOG_ASSERT(error == 0);
982 	error = pthread_create(&td, NULL, ggate_send_thread, res);
983 	PJDLOG_ASSERT(error == 0);
984 	fullystarted = true;
985 	(void)sync_thread(res);
986 }
987 
988 static void
989 reqlog(int loglevel, int debuglevel, struct g_gate_ctl_io *ggio, const char *fmt, ...)
990 {
991 	char msg[1024];
992 	va_list ap;
993 
994 	va_start(ap, fmt);
995 	(void)vsnprintf(msg, sizeof(msg), fmt, ap);
996 	va_end(ap);
997 	switch (ggio->gctl_cmd) {
998 	case BIO_READ:
999 		(void)snprlcat(msg, sizeof(msg), "READ(%ju, %ju).",
1000 		    (uintmax_t)ggio->gctl_offset,
1001 		    (uintmax_t)ggio->gctl_length);
1002 		break;
1003 	case BIO_DELETE:
1004 		(void)snprlcat(msg, sizeof(msg), "DELETE(%ju, %ju).",
1005 		    (uintmax_t)ggio->gctl_offset,
1006 		    (uintmax_t)ggio->gctl_length);
1007 		break;
1008 	case BIO_FLUSH:
1009 		(void)snprlcat(msg, sizeof(msg), "FLUSH.");
1010 		break;
1011 	case BIO_WRITE:
1012 		(void)snprlcat(msg, sizeof(msg), "WRITE(%ju, %ju).",
1013 		    (uintmax_t)ggio->gctl_offset,
1014 		    (uintmax_t)ggio->gctl_length);
1015 		break;
1016 	default:
1017 		(void)snprlcat(msg, sizeof(msg), "UNKNOWN(%u).",
1018 		    (unsigned int)ggio->gctl_cmd);
1019 		break;
1020 	}
1021 	pjdlog_common(loglevel, debuglevel, -1, "%s", msg);
1022 }
1023 
1024 static void
1025 remote_close(struct hast_resource *res, int ncomp)
1026 {
1027 
1028 	rw_wlock(&hio_remote_lock[ncomp]);
1029 	/*
1030 	 * Check for a race between dropping rlock and acquiring wlock -
1031 	 * another thread can close connection in-between.
1032 	 */
1033 	if (!ISCONNECTED(res, ncomp)) {
1034 		PJDLOG_ASSERT(res->hr_remotein == NULL);
1035 		PJDLOG_ASSERT(res->hr_remoteout == NULL);
1036 		rw_unlock(&hio_remote_lock[ncomp]);
1037 		return;
1038 	}
1039 
1040 	PJDLOG_ASSERT(res->hr_remotein != NULL);
1041 	PJDLOG_ASSERT(res->hr_remoteout != NULL);
1042 
1043 	pjdlog_debug(2, "Closing incoming connection to %s.",
1044 	    res->hr_remoteaddr);
1045 	proto_close(res->hr_remotein);
1046 	res->hr_remotein = NULL;
1047 	pjdlog_debug(2, "Closing outgoing connection to %s.",
1048 	    res->hr_remoteaddr);
1049 	proto_close(res->hr_remoteout);
1050 	res->hr_remoteout = NULL;
1051 
1052 	rw_unlock(&hio_remote_lock[ncomp]);
1053 
1054 	pjdlog_warning("Disconnected from %s.", res->hr_remoteaddr);
1055 
1056 	/*
1057 	 * Stop synchronization if in-progress.
1058 	 */
1059 	sync_stop();
1060 
1061 	event_send(res, EVENT_DISCONNECT);
1062 }
1063 
1064 /*
1065  * Acknowledge write completion to the kernel, but don't update activemap yet.
1066  */
1067 static void
1068 write_complete(struct hast_resource *res, struct hio *hio)
1069 {
1070 	struct g_gate_ctl_io *ggio;
1071 	unsigned int ncomp;
1072 
1073 	PJDLOG_ASSERT(!hio->hio_done);
1074 
1075 	ggio = &hio->hio_ggio;
1076 	PJDLOG_ASSERT(ggio->gctl_cmd == BIO_WRITE);
1077 
1078 	/*
1079 	 * Bump local count if this is first write after
1080 	 * connection failure with remote node.
1081 	 */
1082 	ncomp = 1;
1083 	rw_rlock(&hio_remote_lock[ncomp]);
1084 	if (!ISCONNECTED(res, ncomp)) {
1085 		mtx_lock(&metadata_lock);
1086 		if (res->hr_primary_localcnt == res->hr_secondary_remotecnt) {
1087 			res->hr_primary_localcnt++;
1088 			pjdlog_debug(1, "Increasing localcnt to %ju.",
1089 			    (uintmax_t)res->hr_primary_localcnt);
1090 			(void)metadata_write(res);
1091 		}
1092 		mtx_unlock(&metadata_lock);
1093 	}
1094 	rw_unlock(&hio_remote_lock[ncomp]);
1095 	if (ioctl(res->hr_ggatefd, G_GATE_CMD_DONE, ggio) == -1)
1096 		primary_exit(EX_OSERR, "G_GATE_CMD_DONE failed");
1097 	hio->hio_done = true;
1098 }
1099 
1100 /*
1101  * Thread receives ggate I/O requests from the kernel and passes them to
1102  * appropriate threads:
1103  * WRITE - always goes to both local_send and remote_send threads
1104  * READ (when the block is up-to-date on local component) -
1105  *	only local_send thread
1106  * READ (when the block isn't up-to-date on local component) -
1107  *	only remote_send thread
1108  * DELETE - always goes to both local_send and remote_send threads
1109  * FLUSH - always goes to both local_send and remote_send threads
1110  */
1111 static void *
1112 ggate_recv_thread(void *arg)
1113 {
1114 	struct hast_resource *res = arg;
1115 	struct g_gate_ctl_io *ggio;
1116 	struct hio *hio;
1117 	unsigned int ii, ncomp, ncomps;
1118 	int error;
1119 
1120 	for (;;) {
1121 		pjdlog_debug(2, "ggate_recv: Taking free request.");
1122 		QUEUE_TAKE2(hio, free);
1123 		pjdlog_debug(2, "ggate_recv: (%p) Got free request.", hio);
1124 		ggio = &hio->hio_ggio;
1125 		ggio->gctl_unit = res->hr_ggateunit;
1126 		ggio->gctl_length = MAXPHYS;
1127 		ggio->gctl_error = 0;
1128 		hio->hio_done = false;
1129 		hio->hio_replication = res->hr_replication;
1130 		pjdlog_debug(2,
1131 		    "ggate_recv: (%p) Waiting for request from the kernel.",
1132 		    hio);
1133 		if (ioctl(res->hr_ggatefd, G_GATE_CMD_START, ggio) == -1) {
1134 			if (sigexit_received)
1135 				pthread_exit(NULL);
1136 			primary_exit(EX_OSERR, "G_GATE_CMD_START failed");
1137 		}
1138 		error = ggio->gctl_error;
1139 		switch (error) {
1140 		case 0:
1141 			break;
1142 		case ECANCELED:
1143 			/* Exit gracefully. */
1144 			if (!sigexit_received) {
1145 				pjdlog_debug(2,
1146 				    "ggate_recv: (%p) Received cancel from the kernel.",
1147 				    hio);
1148 				pjdlog_info("Received cancel from the kernel, exiting.");
1149 			}
1150 			pthread_exit(NULL);
1151 		case ENOMEM:
1152 			/*
1153 			 * Buffer too small? Impossible, we allocate MAXPHYS
1154 			 * bytes - request can't be bigger than that.
1155 			 */
1156 			/* FALLTHROUGH */
1157 		case ENXIO:
1158 		default:
1159 			primary_exitx(EX_OSERR, "G_GATE_CMD_START failed: %s.",
1160 			    strerror(error));
1161 		}
1162 
1163 		ncomp = 0;
1164 		ncomps = HAST_NCOMPONENTS;
1165 
1166 		for (ii = 0; ii < ncomps; ii++)
1167 			hio->hio_errors[ii] = EINVAL;
1168 		reqlog(LOG_DEBUG, 2, ggio,
1169 		    "ggate_recv: (%p) Request received from the kernel: ",
1170 		    hio);
1171 
1172 		/*
1173 		 * Inform all components about new write request.
1174 		 * For read request prefer local component unless the given
1175 		 * range is out-of-date, then use remote component.
1176 		 */
1177 		switch (ggio->gctl_cmd) {
1178 		case BIO_READ:
1179 			res->hr_stat_read++;
1180 			ncomps = 1;
1181 			mtx_lock(&metadata_lock);
1182 			if (res->hr_syncsrc == HAST_SYNCSRC_UNDEF ||
1183 			    res->hr_syncsrc == HAST_SYNCSRC_PRIMARY) {
1184 				/*
1185 				 * This range is up-to-date on local component,
1186 				 * so handle request locally.
1187 				 */
1188 				 /* Local component is 0 for now. */
1189 				ncomp = 0;
1190 			} else /* if (res->hr_syncsrc ==
1191 			    HAST_SYNCSRC_SECONDARY) */ {
1192 				PJDLOG_ASSERT(res->hr_syncsrc ==
1193 				    HAST_SYNCSRC_SECONDARY);
1194 				/*
1195 				 * This range is out-of-date on local component,
1196 				 * so send request to the remote node.
1197 				 */
1198 				 /* Remote component is 1 for now. */
1199 				ncomp = 1;
1200 			}
1201 			mtx_unlock(&metadata_lock);
1202 			break;
1203 		case BIO_WRITE:
1204 			res->hr_stat_write++;
1205 			if (res->hr_resuid == 0 &&
1206 			    res->hr_primary_localcnt == 0) {
1207 				/* This is first write. */
1208 				res->hr_primary_localcnt = 1;
1209 			}
1210 			for (;;) {
1211 				mtx_lock(&range_lock);
1212 				if (rangelock_islocked(range_sync,
1213 				    ggio->gctl_offset, ggio->gctl_length)) {
1214 					pjdlog_debug(2,
1215 					    "regular: Range offset=%jd length=%zu locked.",
1216 					    (intmax_t)ggio->gctl_offset,
1217 					    (size_t)ggio->gctl_length);
1218 					range_regular_wait = true;
1219 					cv_wait(&range_regular_cond, &range_lock);
1220 					range_regular_wait = false;
1221 					mtx_unlock(&range_lock);
1222 					continue;
1223 				}
1224 				if (rangelock_add(range_regular,
1225 				    ggio->gctl_offset, ggio->gctl_length) == -1) {
1226 					mtx_unlock(&range_lock);
1227 					pjdlog_debug(2,
1228 					    "regular: Range offset=%jd length=%zu is already locked, waiting.",
1229 					    (intmax_t)ggio->gctl_offset,
1230 					    (size_t)ggio->gctl_length);
1231 					sleep(1);
1232 					continue;
1233 				}
1234 				mtx_unlock(&range_lock);
1235 				break;
1236 			}
1237 			mtx_lock(&res->hr_amp_lock);
1238 			if (activemap_write_start(res->hr_amp,
1239 			    ggio->gctl_offset, ggio->gctl_length)) {
1240 				res->hr_stat_activemap_update++;
1241 				(void)hast_activemap_flush(res);
1242 			}
1243 			mtx_unlock(&res->hr_amp_lock);
1244 			break;
1245 		case BIO_DELETE:
1246 			res->hr_stat_delete++;
1247 			break;
1248 		case BIO_FLUSH:
1249 			res->hr_stat_flush++;
1250 			break;
1251 		}
1252 		pjdlog_debug(2,
1253 		    "ggate_recv: (%p) Moving request to the send queues.", hio);
1254 		refcount_init(&hio->hio_countdown, ncomps);
1255 		for (ii = ncomp; ii < ncomp + ncomps; ii++)
1256 			QUEUE_INSERT1(hio, send, ii);
1257 	}
1258 	/* NOTREACHED */
1259 	return (NULL);
1260 }
1261 
1262 /*
1263  * Thread reads from or writes to local component.
1264  * If local read fails, it redirects it to remote_send thread.
1265  */
1266 static void *
1267 local_send_thread(void *arg)
1268 {
1269 	struct hast_resource *res = arg;
1270 	struct g_gate_ctl_io *ggio;
1271 	struct hio *hio;
1272 	unsigned int ncomp, rncomp;
1273 	ssize_t ret;
1274 
1275 	/* Local component is 0 for now. */
1276 	ncomp = 0;
1277 	/* Remote component is 1 for now. */
1278 	rncomp = 1;
1279 
1280 	for (;;) {
1281 		pjdlog_debug(2, "local_send: Taking request.");
1282 		QUEUE_TAKE1(hio, send, ncomp, 0);
1283 		pjdlog_debug(2, "local_send: (%p) Got request.", hio);
1284 		ggio = &hio->hio_ggio;
1285 		switch (ggio->gctl_cmd) {
1286 		case BIO_READ:
1287 			ret = pread(res->hr_localfd, ggio->gctl_data,
1288 			    ggio->gctl_length,
1289 			    ggio->gctl_offset + res->hr_localoff);
1290 			if (ret == ggio->gctl_length)
1291 				hio->hio_errors[ncomp] = 0;
1292 			else if (!ISSYNCREQ(hio)) {
1293 				/*
1294 				 * If READ failed, try to read from remote node.
1295 				 */
1296 				if (ret == -1) {
1297 					reqlog(LOG_WARNING, 0, ggio,
1298 					    "Local request failed (%s), trying remote node. ",
1299 					    strerror(errno));
1300 				} else if (ret != ggio->gctl_length) {
1301 					reqlog(LOG_WARNING, 0, ggio,
1302 					    "Local request failed (%zd != %jd), trying remote node. ",
1303 					    ret, (intmax_t)ggio->gctl_length);
1304 				}
1305 				QUEUE_INSERT1(hio, send, rncomp);
1306 				continue;
1307 			}
1308 			break;
1309 		case BIO_WRITE:
1310 			ret = pwrite(res->hr_localfd, ggio->gctl_data,
1311 			    ggio->gctl_length,
1312 			    ggio->gctl_offset + res->hr_localoff);
1313 			if (ret == -1) {
1314 				hio->hio_errors[ncomp] = errno;
1315 				reqlog(LOG_WARNING, 0, ggio,
1316 				    "Local request failed (%s): ",
1317 				    strerror(errno));
1318 			} else if (ret != ggio->gctl_length) {
1319 				hio->hio_errors[ncomp] = EIO;
1320 				reqlog(LOG_WARNING, 0, ggio,
1321 				    "Local request failed (%zd != %jd): ",
1322 				    ret, (intmax_t)ggio->gctl_length);
1323 			} else {
1324 				hio->hio_errors[ncomp] = 0;
1325 				if (hio->hio_replication ==
1326 				    HAST_REPLICATION_ASYNC &&
1327 				    !ISSYNCREQ(hio)) {
1328 					ggio->gctl_error = 0;
1329 					write_complete(res, hio);
1330 				}
1331 			}
1332 			break;
1333 		case BIO_DELETE:
1334 			ret = g_delete(res->hr_localfd,
1335 			    ggio->gctl_offset + res->hr_localoff,
1336 			    ggio->gctl_length);
1337 			if (ret == -1) {
1338 				hio->hio_errors[ncomp] = errno;
1339 				reqlog(LOG_WARNING, 0, ggio,
1340 				    "Local request failed (%s): ",
1341 				    strerror(errno));
1342 			} else {
1343 				hio->hio_errors[ncomp] = 0;
1344 			}
1345 			break;
1346 		case BIO_FLUSH:
1347 			if (!res->hr_localflush) {
1348 				ret = -1;
1349 				errno = EOPNOTSUPP;
1350 				break;
1351 			}
1352 			ret = g_flush(res->hr_localfd);
1353 			if (ret == -1) {
1354 				if (errno == EOPNOTSUPP)
1355 					res->hr_localflush = false;
1356 				hio->hio_errors[ncomp] = errno;
1357 				reqlog(LOG_WARNING, 0, ggio,
1358 				    "Local request failed (%s): ",
1359 				    strerror(errno));
1360 			} else {
1361 				hio->hio_errors[ncomp] = 0;
1362 			}
1363 			break;
1364 		}
1365 		if (!refcount_release(&hio->hio_countdown))
1366 			continue;
1367 		if (ISSYNCREQ(hio)) {
1368 			mtx_lock(&sync_lock);
1369 			SYNCREQDONE(hio);
1370 			mtx_unlock(&sync_lock);
1371 			cv_signal(&sync_cond);
1372 		} else {
1373 			pjdlog_debug(2,
1374 			    "local_send: (%p) Moving request to the done queue.",
1375 			    hio);
1376 			QUEUE_INSERT2(hio, done);
1377 		}
1378 	}
1379 	/* NOTREACHED */
1380 	return (NULL);
1381 }
1382 
1383 static void
1384 keepalive_send(struct hast_resource *res, unsigned int ncomp)
1385 {
1386 	struct nv *nv;
1387 
1388 	rw_rlock(&hio_remote_lock[ncomp]);
1389 
1390 	if (!ISCONNECTED(res, ncomp)) {
1391 		rw_unlock(&hio_remote_lock[ncomp]);
1392 		return;
1393 	}
1394 
1395 	PJDLOG_ASSERT(res->hr_remotein != NULL);
1396 	PJDLOG_ASSERT(res->hr_remoteout != NULL);
1397 
1398 	nv = nv_alloc();
1399 	nv_add_uint8(nv, HIO_KEEPALIVE, "cmd");
1400 	if (nv_error(nv) != 0) {
1401 		rw_unlock(&hio_remote_lock[ncomp]);
1402 		nv_free(nv);
1403 		pjdlog_debug(1,
1404 		    "keepalive_send: Unable to prepare header to send.");
1405 		return;
1406 	}
1407 	if (hast_proto_send(res, res->hr_remoteout, nv, NULL, 0) == -1) {
1408 		rw_unlock(&hio_remote_lock[ncomp]);
1409 		pjdlog_common(LOG_DEBUG, 1, errno,
1410 		    "keepalive_send: Unable to send request");
1411 		nv_free(nv);
1412 		remote_close(res, ncomp);
1413 		return;
1414 	}
1415 
1416 	rw_unlock(&hio_remote_lock[ncomp]);
1417 	nv_free(nv);
1418 	pjdlog_debug(2, "keepalive_send: Request sent.");
1419 }
1420 
1421 /*
1422  * Thread sends request to secondary node.
1423  */
1424 static void *
1425 remote_send_thread(void *arg)
1426 {
1427 	struct hast_resource *res = arg;
1428 	struct g_gate_ctl_io *ggio;
1429 	time_t lastcheck, now;
1430 	struct hio *hio;
1431 	struct nv *nv;
1432 	unsigned int ncomp;
1433 	bool wakeup;
1434 	uint64_t offset, length;
1435 	uint8_t cmd;
1436 	void *data;
1437 
1438 	/* Remote component is 1 for now. */
1439 	ncomp = 1;
1440 	lastcheck = time(NULL);
1441 
1442 	for (;;) {
1443 		pjdlog_debug(2, "remote_send: Taking request.");
1444 		QUEUE_TAKE1(hio, send, ncomp, HAST_KEEPALIVE);
1445 		if (hio == NULL) {
1446 			now = time(NULL);
1447 			if (lastcheck + HAST_KEEPALIVE <= now) {
1448 				keepalive_send(res, ncomp);
1449 				lastcheck = now;
1450 			}
1451 			continue;
1452 		}
1453 		pjdlog_debug(2, "remote_send: (%p) Got request.", hio);
1454 		ggio = &hio->hio_ggio;
1455 		switch (ggio->gctl_cmd) {
1456 		case BIO_READ:
1457 			cmd = HIO_READ;
1458 			data = NULL;
1459 			offset = ggio->gctl_offset;
1460 			length = ggio->gctl_length;
1461 			break;
1462 		case BIO_WRITE:
1463 			cmd = HIO_WRITE;
1464 			data = ggio->gctl_data;
1465 			offset = ggio->gctl_offset;
1466 			length = ggio->gctl_length;
1467 			break;
1468 		case BIO_DELETE:
1469 			cmd = HIO_DELETE;
1470 			data = NULL;
1471 			offset = ggio->gctl_offset;
1472 			length = ggio->gctl_length;
1473 			break;
1474 		case BIO_FLUSH:
1475 			cmd = HIO_FLUSH;
1476 			data = NULL;
1477 			offset = 0;
1478 			length = 0;
1479 			break;
1480 		default:
1481 			PJDLOG_ABORT("invalid condition");
1482 		}
1483 		nv = nv_alloc();
1484 		nv_add_uint8(nv, cmd, "cmd");
1485 		nv_add_uint64(nv, (uint64_t)ggio->gctl_seq, "seq");
1486 		nv_add_uint64(nv, offset, "offset");
1487 		nv_add_uint64(nv, length, "length");
1488 		if (nv_error(nv) != 0) {
1489 			hio->hio_errors[ncomp] = nv_error(nv);
1490 			pjdlog_debug(2,
1491 			    "remote_send: (%p) Unable to prepare header to send.",
1492 			    hio);
1493 			reqlog(LOG_ERR, 0, ggio,
1494 			    "Unable to prepare header to send (%s): ",
1495 			    strerror(nv_error(nv)));
1496 			/* Move failed request immediately to the done queue. */
1497 			goto done_queue;
1498 		}
1499 		/*
1500 		 * Protect connection from disappearing.
1501 		 */
1502 		rw_rlock(&hio_remote_lock[ncomp]);
1503 		if (!ISCONNECTED(res, ncomp)) {
1504 			rw_unlock(&hio_remote_lock[ncomp]);
1505 			hio->hio_errors[ncomp] = ENOTCONN;
1506 			goto done_queue;
1507 		}
1508 		/*
1509 		 * Move the request to recv queue before sending it, because
1510 		 * in different order we can get reply before we move request
1511 		 * to recv queue.
1512 		 */
1513 		pjdlog_debug(2,
1514 		    "remote_send: (%p) Moving request to the recv queue.",
1515 		    hio);
1516 		mtx_lock(&hio_recv_list_lock[ncomp]);
1517 		wakeup = TAILQ_EMPTY(&hio_recv_list[ncomp]);
1518 		TAILQ_INSERT_TAIL(&hio_recv_list[ncomp], hio, hio_next[ncomp]);
1519 		mtx_unlock(&hio_recv_list_lock[ncomp]);
1520 		if (hast_proto_send(res, res->hr_remoteout, nv, data,
1521 		    data != NULL ? length : 0) == -1) {
1522 			hio->hio_errors[ncomp] = errno;
1523 			rw_unlock(&hio_remote_lock[ncomp]);
1524 			pjdlog_debug(2,
1525 			    "remote_send: (%p) Unable to send request.", hio);
1526 			reqlog(LOG_ERR, 0, ggio,
1527 			    "Unable to send request (%s): ",
1528 			    strerror(hio->hio_errors[ncomp]));
1529 			remote_close(res, ncomp);
1530 			/*
1531 			 * Take request back from the receive queue and move
1532 			 * it immediately to the done queue.
1533 			 */
1534 			mtx_lock(&hio_recv_list_lock[ncomp]);
1535 			TAILQ_REMOVE(&hio_recv_list[ncomp], hio,
1536 			    hio_next[ncomp]);
1537 			mtx_unlock(&hio_recv_list_lock[ncomp]);
1538 			goto done_queue;
1539 		}
1540 		rw_unlock(&hio_remote_lock[ncomp]);
1541 		nv_free(nv);
1542 		if (wakeup)
1543 			cv_signal(&hio_recv_list_cond[ncomp]);
1544 		continue;
1545 done_queue:
1546 		nv_free(nv);
1547 		if (ISSYNCREQ(hio)) {
1548 			if (!refcount_release(&hio->hio_countdown))
1549 				continue;
1550 			mtx_lock(&sync_lock);
1551 			SYNCREQDONE(hio);
1552 			mtx_unlock(&sync_lock);
1553 			cv_signal(&sync_cond);
1554 			continue;
1555 		}
1556 		if (ggio->gctl_cmd == BIO_WRITE) {
1557 			mtx_lock(&res->hr_amp_lock);
1558 			if (activemap_need_sync(res->hr_amp, ggio->gctl_offset,
1559 			    ggio->gctl_length)) {
1560 				(void)hast_activemap_flush(res);
1561 			}
1562 			mtx_unlock(&res->hr_amp_lock);
1563 		}
1564 		if (!refcount_release(&hio->hio_countdown))
1565 			continue;
1566 		pjdlog_debug(2,
1567 		    "remote_send: (%p) Moving request to the done queue.",
1568 		    hio);
1569 		QUEUE_INSERT2(hio, done);
1570 	}
1571 	/* NOTREACHED */
1572 	return (NULL);
1573 }
1574 
1575 /*
1576  * Thread receives answer from secondary node and passes it to ggate_send
1577  * thread.
1578  */
1579 static void *
1580 remote_recv_thread(void *arg)
1581 {
1582 	struct hast_resource *res = arg;
1583 	struct g_gate_ctl_io *ggio;
1584 	struct hio *hio;
1585 	struct nv *nv;
1586 	unsigned int ncomp;
1587 	uint64_t seq;
1588 	int error;
1589 
1590 	/* Remote component is 1 for now. */
1591 	ncomp = 1;
1592 
1593 	for (;;) {
1594 		/* Wait until there is anything to receive. */
1595 		mtx_lock(&hio_recv_list_lock[ncomp]);
1596 		while (TAILQ_EMPTY(&hio_recv_list[ncomp])) {
1597 			pjdlog_debug(2, "remote_recv: No requests, waiting.");
1598 			cv_wait(&hio_recv_list_cond[ncomp],
1599 			    &hio_recv_list_lock[ncomp]);
1600 		}
1601 		mtx_unlock(&hio_recv_list_lock[ncomp]);
1602 
1603 		rw_rlock(&hio_remote_lock[ncomp]);
1604 		if (!ISCONNECTED(res, ncomp)) {
1605 			rw_unlock(&hio_remote_lock[ncomp]);
1606 			/*
1607 			 * Connection is dead, so move all pending requests to
1608 			 * the done queue (one-by-one).
1609 			 */
1610 			mtx_lock(&hio_recv_list_lock[ncomp]);
1611 			hio = TAILQ_FIRST(&hio_recv_list[ncomp]);
1612 			PJDLOG_ASSERT(hio != NULL);
1613 			TAILQ_REMOVE(&hio_recv_list[ncomp], hio,
1614 			    hio_next[ncomp]);
1615 			mtx_unlock(&hio_recv_list_lock[ncomp]);
1616 			goto done_queue;
1617 		}
1618 		if (hast_proto_recv_hdr(res->hr_remotein, &nv) == -1) {
1619 			pjdlog_errno(LOG_ERR,
1620 			    "Unable to receive reply header");
1621 			rw_unlock(&hio_remote_lock[ncomp]);
1622 			remote_close(res, ncomp);
1623 			continue;
1624 		}
1625 		rw_unlock(&hio_remote_lock[ncomp]);
1626 		seq = nv_get_uint64(nv, "seq");
1627 		if (seq == 0) {
1628 			pjdlog_error("Header contains no 'seq' field.");
1629 			nv_free(nv);
1630 			continue;
1631 		}
1632 		mtx_lock(&hio_recv_list_lock[ncomp]);
1633 		TAILQ_FOREACH(hio, &hio_recv_list[ncomp], hio_next[ncomp]) {
1634 			if (hio->hio_ggio.gctl_seq == seq) {
1635 				TAILQ_REMOVE(&hio_recv_list[ncomp], hio,
1636 				    hio_next[ncomp]);
1637 				break;
1638 			}
1639 		}
1640 		mtx_unlock(&hio_recv_list_lock[ncomp]);
1641 		if (hio == NULL) {
1642 			pjdlog_error("Found no request matching received 'seq' field (%ju).",
1643 			    (uintmax_t)seq);
1644 			nv_free(nv);
1645 			continue;
1646 		}
1647 		ggio = &hio->hio_ggio;
1648 		error = nv_get_int16(nv, "error");
1649 		if (error != 0) {
1650 			/* Request failed on remote side. */
1651 			hio->hio_errors[ncomp] = error;
1652 			reqlog(LOG_WARNING, 0, ggio,
1653 			    "Remote request failed (%s): ", strerror(error));
1654 			nv_free(nv);
1655 			goto done_queue;
1656 		}
1657 		switch (ggio->gctl_cmd) {
1658 		case BIO_READ:
1659 			rw_rlock(&hio_remote_lock[ncomp]);
1660 			if (!ISCONNECTED(res, ncomp)) {
1661 				rw_unlock(&hio_remote_lock[ncomp]);
1662 				nv_free(nv);
1663 				goto done_queue;
1664 			}
1665 			if (hast_proto_recv_data(res, res->hr_remotein, nv,
1666 			    ggio->gctl_data, ggio->gctl_length) == -1) {
1667 				hio->hio_errors[ncomp] = errno;
1668 				pjdlog_errno(LOG_ERR,
1669 				    "Unable to receive reply data");
1670 				rw_unlock(&hio_remote_lock[ncomp]);
1671 				nv_free(nv);
1672 				remote_close(res, ncomp);
1673 				goto done_queue;
1674 			}
1675 			rw_unlock(&hio_remote_lock[ncomp]);
1676 			break;
1677 		case BIO_WRITE:
1678 		case BIO_DELETE:
1679 		case BIO_FLUSH:
1680 			break;
1681 		default:
1682 			PJDLOG_ABORT("invalid condition");
1683 		}
1684 		hio->hio_errors[ncomp] = 0;
1685 		nv_free(nv);
1686 done_queue:
1687 		if (!refcount_release(&hio->hio_countdown))
1688 			continue;
1689 		if (ISSYNCREQ(hio)) {
1690 			mtx_lock(&sync_lock);
1691 			SYNCREQDONE(hio);
1692 			mtx_unlock(&sync_lock);
1693 			cv_signal(&sync_cond);
1694 		} else {
1695 			pjdlog_debug(2,
1696 			    "remote_recv: (%p) Moving request to the done queue.",
1697 			    hio);
1698 			QUEUE_INSERT2(hio, done);
1699 		}
1700 	}
1701 	/* NOTREACHED */
1702 	return (NULL);
1703 }
1704 
1705 /*
1706  * Thread sends answer to the kernel.
1707  */
1708 static void *
1709 ggate_send_thread(void *arg)
1710 {
1711 	struct hast_resource *res = arg;
1712 	struct g_gate_ctl_io *ggio;
1713 	struct hio *hio;
1714 	unsigned int ii, ncomps;
1715 
1716 	ncomps = HAST_NCOMPONENTS;
1717 
1718 	for (;;) {
1719 		pjdlog_debug(2, "ggate_send: Taking request.");
1720 		QUEUE_TAKE2(hio, done);
1721 		pjdlog_debug(2, "ggate_send: (%p) Got request.", hio);
1722 		ggio = &hio->hio_ggio;
1723 		for (ii = 0; ii < ncomps; ii++) {
1724 			if (hio->hio_errors[ii] == 0) {
1725 				/*
1726 				 * One successful request is enough to declare
1727 				 * success.
1728 				 */
1729 				ggio->gctl_error = 0;
1730 				break;
1731 			}
1732 		}
1733 		if (ii == ncomps) {
1734 			/*
1735 			 * None of the requests were successful.
1736 			 * Use the error from local component except the
1737 			 * case when we did only remote request.
1738 			 */
1739 			if (ggio->gctl_cmd == BIO_READ &&
1740 			    res->hr_syncsrc == HAST_SYNCSRC_SECONDARY)
1741 				ggio->gctl_error = hio->hio_errors[1];
1742 			else
1743 				ggio->gctl_error = hio->hio_errors[0];
1744 		}
1745 		if (ggio->gctl_error == 0 && ggio->gctl_cmd == BIO_WRITE) {
1746 			mtx_lock(&res->hr_amp_lock);
1747 			if (activemap_write_complete(res->hr_amp,
1748 			    ggio->gctl_offset, ggio->gctl_length)) {
1749 				res->hr_stat_activemap_update++;
1750 				(void)hast_activemap_flush(res);
1751 			}
1752 			mtx_unlock(&res->hr_amp_lock);
1753 		}
1754 		if (ggio->gctl_cmd == BIO_WRITE) {
1755 			/*
1756 			 * Unlock range we locked.
1757 			 */
1758 			mtx_lock(&range_lock);
1759 			rangelock_del(range_regular, ggio->gctl_offset,
1760 			    ggio->gctl_length);
1761 			if (range_sync_wait)
1762 				cv_signal(&range_sync_cond);
1763 			mtx_unlock(&range_lock);
1764 			if (!hio->hio_done)
1765 				write_complete(res, hio);
1766 		} else {
1767 			if (ioctl(res->hr_ggatefd, G_GATE_CMD_DONE, ggio) == -1) {
1768 				primary_exit(EX_OSERR,
1769 				    "G_GATE_CMD_DONE failed");
1770 			}
1771 		}
1772 		pjdlog_debug(2,
1773 		    "ggate_send: (%p) Moving request to the free queue.", hio);
1774 		QUEUE_INSERT2(hio, free);
1775 	}
1776 	/* NOTREACHED */
1777 	return (NULL);
1778 }
1779 
1780 /*
1781  * Thread synchronize local and remote components.
1782  */
1783 static void *
1784 sync_thread(void *arg __unused)
1785 {
1786 	struct hast_resource *res = arg;
1787 	struct hio *hio;
1788 	struct g_gate_ctl_io *ggio;
1789 	struct timeval tstart, tend, tdiff;
1790 	unsigned int ii, ncomp, ncomps;
1791 	off_t offset, length, synced;
1792 	bool dorewind;
1793 	int syncext;
1794 
1795 	ncomps = HAST_NCOMPONENTS;
1796 	dorewind = true;
1797 	synced = 0;
1798 	offset = -1;
1799 
1800 	for (;;) {
1801 		mtx_lock(&sync_lock);
1802 		if (offset >= 0 && !sync_inprogress) {
1803 			gettimeofday(&tend, NULL);
1804 			timersub(&tend, &tstart, &tdiff);
1805 			pjdlog_info("Synchronization interrupted after %#.0T. "
1806 			    "%NB synchronized so far.", &tdiff,
1807 			    (intmax_t)synced);
1808 			event_send(res, EVENT_SYNCINTR);
1809 		}
1810 		while (!sync_inprogress) {
1811 			dorewind = true;
1812 			synced = 0;
1813 			cv_wait(&sync_cond, &sync_lock);
1814 		}
1815 		mtx_unlock(&sync_lock);
1816 		/*
1817 		 * Obtain offset at which we should synchronize.
1818 		 * Rewind synchronization if needed.
1819 		 */
1820 		mtx_lock(&res->hr_amp_lock);
1821 		if (dorewind)
1822 			activemap_sync_rewind(res->hr_amp);
1823 		offset = activemap_sync_offset(res->hr_amp, &length, &syncext);
1824 		if (syncext != -1) {
1825 			/*
1826 			 * We synchronized entire syncext extent, we can mark
1827 			 * it as clean now.
1828 			 */
1829 			if (activemap_extent_complete(res->hr_amp, syncext))
1830 				(void)hast_activemap_flush(res);
1831 		}
1832 		mtx_unlock(&res->hr_amp_lock);
1833 		if (dorewind) {
1834 			dorewind = false;
1835 			if (offset == -1)
1836 				pjdlog_info("Nodes are in sync.");
1837 			else {
1838 				pjdlog_info("Synchronization started. %NB to go.",
1839 				    (intmax_t)(res->hr_extentsize *
1840 				    activemap_ndirty(res->hr_amp)));
1841 				event_send(res, EVENT_SYNCSTART);
1842 				gettimeofday(&tstart, NULL);
1843 			}
1844 		}
1845 		if (offset == -1) {
1846 			sync_stop();
1847 			pjdlog_debug(1, "Nothing to synchronize.");
1848 			/*
1849 			 * Synchronization complete, make both localcnt and
1850 			 * remotecnt equal.
1851 			 */
1852 			ncomp = 1;
1853 			rw_rlock(&hio_remote_lock[ncomp]);
1854 			if (ISCONNECTED(res, ncomp)) {
1855 				if (synced > 0) {
1856 					int64_t bps;
1857 
1858 					gettimeofday(&tend, NULL);
1859 					timersub(&tend, &tstart, &tdiff);
1860 					bps = (int64_t)((double)synced /
1861 					    ((double)tdiff.tv_sec +
1862 					    (double)tdiff.tv_usec / 1000000));
1863 					pjdlog_info("Synchronization complete. "
1864 					    "%NB synchronized in %#.0lT (%NB/sec).",
1865 					    (intmax_t)synced, &tdiff,
1866 					    (intmax_t)bps);
1867 					event_send(res, EVENT_SYNCDONE);
1868 				}
1869 				mtx_lock(&metadata_lock);
1870 				res->hr_syncsrc = HAST_SYNCSRC_UNDEF;
1871 				res->hr_primary_localcnt =
1872 				    res->hr_secondary_remotecnt;
1873 				res->hr_primary_remotecnt =
1874 				    res->hr_secondary_localcnt;
1875 				pjdlog_debug(1,
1876 				    "Setting localcnt to %ju and remotecnt to %ju.",
1877 				    (uintmax_t)res->hr_primary_localcnt,
1878 				    (uintmax_t)res->hr_primary_remotecnt);
1879 				(void)metadata_write(res);
1880 				mtx_unlock(&metadata_lock);
1881 			}
1882 			rw_unlock(&hio_remote_lock[ncomp]);
1883 			continue;
1884 		}
1885 		pjdlog_debug(2, "sync: Taking free request.");
1886 		QUEUE_TAKE2(hio, free);
1887 		pjdlog_debug(2, "sync: (%p) Got free request.", hio);
1888 		/*
1889 		 * Lock the range we are going to synchronize. We don't want
1890 		 * race where someone writes between our read and write.
1891 		 */
1892 		for (;;) {
1893 			mtx_lock(&range_lock);
1894 			if (rangelock_islocked(range_regular, offset, length)) {
1895 				pjdlog_debug(2,
1896 				    "sync: Range offset=%jd length=%jd locked.",
1897 				    (intmax_t)offset, (intmax_t)length);
1898 				range_sync_wait = true;
1899 				cv_wait(&range_sync_cond, &range_lock);
1900 				range_sync_wait = false;
1901 				mtx_unlock(&range_lock);
1902 				continue;
1903 			}
1904 			if (rangelock_add(range_sync, offset, length) == -1) {
1905 				mtx_unlock(&range_lock);
1906 				pjdlog_debug(2,
1907 				    "sync: Range offset=%jd length=%jd is already locked, waiting.",
1908 				    (intmax_t)offset, (intmax_t)length);
1909 				sleep(1);
1910 				continue;
1911 			}
1912 			mtx_unlock(&range_lock);
1913 			break;
1914 		}
1915 		/*
1916 		 * First read the data from synchronization source.
1917 		 */
1918 		SYNCREQ(hio);
1919 		ggio = &hio->hio_ggio;
1920 		ggio->gctl_cmd = BIO_READ;
1921 		ggio->gctl_offset = offset;
1922 		ggio->gctl_length = length;
1923 		ggio->gctl_error = 0;
1924 		hio->hio_done = false;
1925 		hio->hio_replication = res->hr_replication;
1926 		for (ii = 0; ii < ncomps; ii++)
1927 			hio->hio_errors[ii] = EINVAL;
1928 		reqlog(LOG_DEBUG, 2, ggio, "sync: (%p) Sending sync request: ",
1929 		    hio);
1930 		pjdlog_debug(2, "sync: (%p) Moving request to the send queue.",
1931 		    hio);
1932 		mtx_lock(&metadata_lock);
1933 		if (res->hr_syncsrc == HAST_SYNCSRC_PRIMARY) {
1934 			/*
1935 			 * This range is up-to-date on local component,
1936 			 * so handle request locally.
1937 			 */
1938 			 /* Local component is 0 for now. */
1939 			ncomp = 0;
1940 		} else /* if (res->hr_syncsrc == HAST_SYNCSRC_SECONDARY) */ {
1941 			PJDLOG_ASSERT(res->hr_syncsrc == HAST_SYNCSRC_SECONDARY);
1942 			/*
1943 			 * This range is out-of-date on local component,
1944 			 * so send request to the remote node.
1945 			 */
1946 			 /* Remote component is 1 for now. */
1947 			ncomp = 1;
1948 		}
1949 		mtx_unlock(&metadata_lock);
1950 		refcount_init(&hio->hio_countdown, 1);
1951 		QUEUE_INSERT1(hio, send, ncomp);
1952 
1953 		/*
1954 		 * Let's wait for READ to finish.
1955 		 */
1956 		mtx_lock(&sync_lock);
1957 		while (!ISSYNCREQDONE(hio))
1958 			cv_wait(&sync_cond, &sync_lock);
1959 		mtx_unlock(&sync_lock);
1960 
1961 		if (hio->hio_errors[ncomp] != 0) {
1962 			pjdlog_error("Unable to read synchronization data: %s.",
1963 			    strerror(hio->hio_errors[ncomp]));
1964 			goto free_queue;
1965 		}
1966 
1967 		/*
1968 		 * We read the data from synchronization source, now write it
1969 		 * to synchronization target.
1970 		 */
1971 		SYNCREQ(hio);
1972 		ggio->gctl_cmd = BIO_WRITE;
1973 		for (ii = 0; ii < ncomps; ii++)
1974 			hio->hio_errors[ii] = EINVAL;
1975 		reqlog(LOG_DEBUG, 2, ggio, "sync: (%p) Sending sync request: ",
1976 		    hio);
1977 		pjdlog_debug(2, "sync: (%p) Moving request to the send queue.",
1978 		    hio);
1979 		mtx_lock(&metadata_lock);
1980 		if (res->hr_syncsrc == HAST_SYNCSRC_PRIMARY) {
1981 			/*
1982 			 * This range is up-to-date on local component,
1983 			 * so we update remote component.
1984 			 */
1985 			 /* Remote component is 1 for now. */
1986 			ncomp = 1;
1987 		} else /* if (res->hr_syncsrc == HAST_SYNCSRC_SECONDARY) */ {
1988 			PJDLOG_ASSERT(res->hr_syncsrc == HAST_SYNCSRC_SECONDARY);
1989 			/*
1990 			 * This range is out-of-date on local component,
1991 			 * so we update it.
1992 			 */
1993 			 /* Local component is 0 for now. */
1994 			ncomp = 0;
1995 		}
1996 		mtx_unlock(&metadata_lock);
1997 
1998 		pjdlog_debug(2, "sync: (%p) Moving request to the send queue.",
1999 		    hio);
2000 		refcount_init(&hio->hio_countdown, 1);
2001 		QUEUE_INSERT1(hio, send, ncomp);
2002 
2003 		/*
2004 		 * Let's wait for WRITE to finish.
2005 		 */
2006 		mtx_lock(&sync_lock);
2007 		while (!ISSYNCREQDONE(hio))
2008 			cv_wait(&sync_cond, &sync_lock);
2009 		mtx_unlock(&sync_lock);
2010 
2011 		if (hio->hio_errors[ncomp] != 0) {
2012 			pjdlog_error("Unable to write synchronization data: %s.",
2013 			    strerror(hio->hio_errors[ncomp]));
2014 			goto free_queue;
2015 		}
2016 
2017 		synced += length;
2018 free_queue:
2019 		mtx_lock(&range_lock);
2020 		rangelock_del(range_sync, offset, length);
2021 		if (range_regular_wait)
2022 			cv_signal(&range_regular_cond);
2023 		mtx_unlock(&range_lock);
2024 		pjdlog_debug(2, "sync: (%p) Moving request to the free queue.",
2025 		    hio);
2026 		QUEUE_INSERT2(hio, free);
2027 	}
2028 	/* NOTREACHED */
2029 	return (NULL);
2030 }
2031 
2032 void
2033 primary_config_reload(struct hast_resource *res, struct nv *nv)
2034 {
2035 	unsigned int ii, ncomps;
2036 	int modified, vint;
2037 	const char *vstr;
2038 
2039 	pjdlog_info("Reloading configuration...");
2040 
2041 	PJDLOG_ASSERT(res->hr_role == HAST_ROLE_PRIMARY);
2042 	PJDLOG_ASSERT(gres == res);
2043 	nv_assert(nv, "remoteaddr");
2044 	nv_assert(nv, "sourceaddr");
2045 	nv_assert(nv, "replication");
2046 	nv_assert(nv, "checksum");
2047 	nv_assert(nv, "compression");
2048 	nv_assert(nv, "timeout");
2049 	nv_assert(nv, "exec");
2050 	nv_assert(nv, "metaflush");
2051 
2052 	ncomps = HAST_NCOMPONENTS;
2053 
2054 #define MODIFIED_REMOTEADDR	0x01
2055 #define MODIFIED_SOURCEADDR	0x02
2056 #define MODIFIED_REPLICATION	0x04
2057 #define MODIFIED_CHECKSUM	0x08
2058 #define MODIFIED_COMPRESSION	0x10
2059 #define MODIFIED_TIMEOUT	0x20
2060 #define MODIFIED_EXEC		0x40
2061 #define MODIFIED_METAFLUSH	0x80
2062 	modified = 0;
2063 
2064 	vstr = nv_get_string(nv, "remoteaddr");
2065 	if (strcmp(gres->hr_remoteaddr, vstr) != 0) {
2066 		/*
2067 		 * Don't copy res->hr_remoteaddr to gres just yet.
2068 		 * We want remote_close() to log disconnect from the old
2069 		 * addresses, not from the new ones.
2070 		 */
2071 		modified |= MODIFIED_REMOTEADDR;
2072 	}
2073 	vstr = nv_get_string(nv, "sourceaddr");
2074 	if (strcmp(gres->hr_sourceaddr, vstr) != 0) {
2075 		strlcpy(gres->hr_sourceaddr, vstr, sizeof(gres->hr_sourceaddr));
2076 		modified |= MODIFIED_SOURCEADDR;
2077 	}
2078 	vint = nv_get_int32(nv, "replication");
2079 	if (gres->hr_replication != vint) {
2080 		gres->hr_replication = vint;
2081 		modified |= MODIFIED_REPLICATION;
2082 	}
2083 	vint = nv_get_int32(nv, "checksum");
2084 	if (gres->hr_checksum != vint) {
2085 		gres->hr_checksum = vint;
2086 		modified |= MODIFIED_CHECKSUM;
2087 	}
2088 	vint = nv_get_int32(nv, "compression");
2089 	if (gres->hr_compression != vint) {
2090 		gres->hr_compression = vint;
2091 		modified |= MODIFIED_COMPRESSION;
2092 	}
2093 	vint = nv_get_int32(nv, "timeout");
2094 	if (gres->hr_timeout != vint) {
2095 		gres->hr_timeout = vint;
2096 		modified |= MODIFIED_TIMEOUT;
2097 	}
2098 	vstr = nv_get_string(nv, "exec");
2099 	if (strcmp(gres->hr_exec, vstr) != 0) {
2100 		strlcpy(gres->hr_exec, vstr, sizeof(gres->hr_exec));
2101 		modified |= MODIFIED_EXEC;
2102 	}
2103 	vint = nv_get_int32(nv, "metaflush");
2104 	if (gres->hr_metaflush != vint) {
2105 		gres->hr_metaflush = vint;
2106 		modified |= MODIFIED_METAFLUSH;
2107 	}
2108 
2109 	/*
2110 	 * Change timeout for connected sockets.
2111 	 * Don't bother if we need to reconnect.
2112 	 */
2113 	if ((modified & MODIFIED_TIMEOUT) != 0 &&
2114 	    (modified & (MODIFIED_REMOTEADDR | MODIFIED_SOURCEADDR)) == 0) {
2115 		for (ii = 0; ii < ncomps; ii++) {
2116 			if (!ISREMOTE(ii))
2117 				continue;
2118 			rw_rlock(&hio_remote_lock[ii]);
2119 			if (!ISCONNECTED(gres, ii)) {
2120 				rw_unlock(&hio_remote_lock[ii]);
2121 				continue;
2122 			}
2123 			rw_unlock(&hio_remote_lock[ii]);
2124 			if (proto_timeout(gres->hr_remotein,
2125 			    gres->hr_timeout) == -1) {
2126 				pjdlog_errno(LOG_WARNING,
2127 				    "Unable to set connection timeout");
2128 			}
2129 			if (proto_timeout(gres->hr_remoteout,
2130 			    gres->hr_timeout) == -1) {
2131 				pjdlog_errno(LOG_WARNING,
2132 				    "Unable to set connection timeout");
2133 			}
2134 		}
2135 	}
2136 	if ((modified & (MODIFIED_REMOTEADDR | MODIFIED_SOURCEADDR)) != 0) {
2137 		for (ii = 0; ii < ncomps; ii++) {
2138 			if (!ISREMOTE(ii))
2139 				continue;
2140 			remote_close(gres, ii);
2141 		}
2142 		if (modified & MODIFIED_REMOTEADDR) {
2143 			vstr = nv_get_string(nv, "remoteaddr");
2144 			strlcpy(gres->hr_remoteaddr, vstr,
2145 			    sizeof(gres->hr_remoteaddr));
2146 		}
2147 	}
2148 #undef	MODIFIED_REMOTEADDR
2149 #undef	MODIFIED_SOURCEADDR
2150 #undef	MODIFIED_REPLICATION
2151 #undef	MODIFIED_CHECKSUM
2152 #undef	MODIFIED_COMPRESSION
2153 #undef	MODIFIED_TIMEOUT
2154 #undef	MODIFIED_EXEC
2155 #undef	MODIFIED_METAFLUSH
2156 
2157 	pjdlog_info("Configuration reloaded successfully.");
2158 }
2159 
2160 static void
2161 guard_one(struct hast_resource *res, unsigned int ncomp)
2162 {
2163 	struct proto_conn *in, *out;
2164 
2165 	if (!ISREMOTE(ncomp))
2166 		return;
2167 
2168 	rw_rlock(&hio_remote_lock[ncomp]);
2169 
2170 	if (!real_remote(res)) {
2171 		rw_unlock(&hio_remote_lock[ncomp]);
2172 		return;
2173 	}
2174 
2175 	if (ISCONNECTED(res, ncomp)) {
2176 		PJDLOG_ASSERT(res->hr_remotein != NULL);
2177 		PJDLOG_ASSERT(res->hr_remoteout != NULL);
2178 		rw_unlock(&hio_remote_lock[ncomp]);
2179 		pjdlog_debug(2, "remote_guard: Connection to %s is ok.",
2180 		    res->hr_remoteaddr);
2181 		return;
2182 	}
2183 
2184 	PJDLOG_ASSERT(res->hr_remotein == NULL);
2185 	PJDLOG_ASSERT(res->hr_remoteout == NULL);
2186 	/*
2187 	 * Upgrade the lock. It doesn't have to be atomic as no other thread
2188 	 * can change connection status from disconnected to connected.
2189 	 */
2190 	rw_unlock(&hio_remote_lock[ncomp]);
2191 	pjdlog_debug(2, "remote_guard: Reconnecting to %s.",
2192 	    res->hr_remoteaddr);
2193 	in = out = NULL;
2194 	if (init_remote(res, &in, &out) == 0) {
2195 		rw_wlock(&hio_remote_lock[ncomp]);
2196 		PJDLOG_ASSERT(res->hr_remotein == NULL);
2197 		PJDLOG_ASSERT(res->hr_remoteout == NULL);
2198 		PJDLOG_ASSERT(in != NULL && out != NULL);
2199 		res->hr_remotein = in;
2200 		res->hr_remoteout = out;
2201 		rw_unlock(&hio_remote_lock[ncomp]);
2202 		pjdlog_info("Successfully reconnected to %s.",
2203 		    res->hr_remoteaddr);
2204 		sync_start();
2205 	} else {
2206 		/* Both connections should be NULL. */
2207 		PJDLOG_ASSERT(res->hr_remotein == NULL);
2208 		PJDLOG_ASSERT(res->hr_remoteout == NULL);
2209 		PJDLOG_ASSERT(in == NULL && out == NULL);
2210 		pjdlog_debug(2, "remote_guard: Reconnect to %s failed.",
2211 		    res->hr_remoteaddr);
2212 	}
2213 }
2214 
2215 /*
2216  * Thread guards remote connections and reconnects when needed, handles
2217  * signals, etc.
2218  */
2219 static void *
2220 guard_thread(void *arg)
2221 {
2222 	struct hast_resource *res = arg;
2223 	unsigned int ii, ncomps;
2224 	struct timespec timeout;
2225 	time_t lastcheck, now;
2226 	sigset_t mask;
2227 	int signo;
2228 
2229 	ncomps = HAST_NCOMPONENTS;
2230 	lastcheck = time(NULL);
2231 
2232 	PJDLOG_VERIFY(sigemptyset(&mask) == 0);
2233 	PJDLOG_VERIFY(sigaddset(&mask, SIGINT) == 0);
2234 	PJDLOG_VERIFY(sigaddset(&mask, SIGTERM) == 0);
2235 
2236 	timeout.tv_sec = HAST_KEEPALIVE;
2237 	timeout.tv_nsec = 0;
2238 	signo = -1;
2239 
2240 	for (;;) {
2241 		switch (signo) {
2242 		case SIGINT:
2243 		case SIGTERM:
2244 			sigexit_received = true;
2245 			primary_exitx(EX_OK,
2246 			    "Termination signal received, exiting.");
2247 			break;
2248 		default:
2249 			break;
2250 		}
2251 
2252 		/*
2253 		 * Don't check connections until we fully started,
2254 		 * as we may still be looping, waiting for remote node
2255 		 * to switch from primary to secondary.
2256 		 */
2257 		if (fullystarted) {
2258 			pjdlog_debug(2, "remote_guard: Checking connections.");
2259 			now = time(NULL);
2260 			if (lastcheck + HAST_KEEPALIVE <= now) {
2261 				for (ii = 0; ii < ncomps; ii++)
2262 					guard_one(res, ii);
2263 				lastcheck = now;
2264 			}
2265 		}
2266 		signo = sigtimedwait(&mask, NULL, &timeout);
2267 	}
2268 	/* NOTREACHED */
2269 	return (NULL);
2270 }
2271