xref: /spdk/module/bdev/rbd/bdev_rbd.c (revision 5a7d428d0fa7ace275e7dc5fe97f6a2ae6ad012d)
1 /*-
2  *   BSD LICENSE
3  *
4  *   Copyright (c) Intel Corporation.
5  *   All rights reserved.
6  *
7  *   Redistribution and use in source and binary forms, with or without
8  *   modification, are permitted provided that the following conditions
9  *   are met:
10  *
11  *     * Redistributions of source code must retain the above copyright
12  *       notice, this list of conditions and the following disclaimer.
13  *     * Redistributions in binary form must reproduce the above copyright
14  *       notice, this list of conditions and the following disclaimer in
15  *       the documentation and/or other materials provided with the
16  *       distribution.
17  *     * Neither the name of Intel Corporation nor the names of its
18  *       contributors may be used to endorse or promote products derived
19  *       from this software without specific prior written permission.
20  *
21  *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22  *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23  *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
24  *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
25  *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
26  *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
27  *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
28  *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
29  *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
30  *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
31  *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
32  */
33 
34 #include "spdk/stdinc.h"
35 
36 #include "bdev_rbd.h"
37 
38 #include <rbd/librbd.h>
39 #include <rados/librados.h>
40 #include <sys/eventfd.h>
41 #include <sys/epoll.h>
42 
43 #include "spdk/env.h"
44 #include "spdk/bdev.h"
45 #include "spdk/thread.h"
46 #include "spdk/json.h"
47 #include "spdk/string.h"
48 #include "spdk/util.h"
49 #include "spdk/likely.h"
50 
51 #include "spdk/bdev_module.h"
52 #include "spdk/log.h"
53 
54 #define SPDK_RBD_QUEUE_DEPTH 128
55 #define MAX_EVENTS_PER_POLL 128
56 
57 static int bdev_rbd_count = 0;
58 
59 struct bdev_rbd {
60 	struct spdk_bdev disk;
61 	char *rbd_name;
62 	char *user_id;
63 	char *pool_name;
64 	char **config;
65 	rbd_image_info_t info;
66 	TAILQ_ENTRY(bdev_rbd) tailq;
67 	struct spdk_poller *reset_timer;
68 	struct spdk_bdev_io *reset_bdev_io;
69 };
70 
71 struct bdev_rbd_group_channel {
72 	struct spdk_poller *poller;
73 	int epoll_fd;
74 };
75 
76 struct bdev_rbd_io_channel {
77 	rados_ioctx_t io_ctx;
78 	rados_t cluster;
79 	int pfd;
80 	rbd_image_t image;
81 	struct bdev_rbd *disk;
82 	struct bdev_rbd_group_channel *group_ch;
83 };
84 
85 struct bdev_rbd_io {
86 	size_t	total_len;
87 };
88 
89 static void
90 bdev_rbd_free(struct bdev_rbd *rbd)
91 {
92 	if (!rbd) {
93 		return;
94 	}
95 
96 	free(rbd->disk.name);
97 	free(rbd->rbd_name);
98 	free(rbd->user_id);
99 	free(rbd->pool_name);
100 	bdev_rbd_free_config(rbd->config);
101 	free(rbd);
102 }
103 
104 void
105 bdev_rbd_free_config(char **config)
106 {
107 	char **entry;
108 
109 	if (config) {
110 		for (entry = config; *entry; entry++) {
111 			free(*entry);
112 		}
113 		free(config);
114 	}
115 }
116 
117 char **
118 bdev_rbd_dup_config(const char *const *config)
119 {
120 	size_t count;
121 	char **copy;
122 
123 	if (!config) {
124 		return NULL;
125 	}
126 	for (count = 0; config[count]; count++) {}
127 	copy = calloc(count + 1, sizeof(*copy));
128 	if (!copy) {
129 		return NULL;
130 	}
131 	for (count = 0; config[count]; count++) {
132 		if (!(copy[count] = strdup(config[count]))) {
133 			bdev_rbd_free_config(copy);
134 			return NULL;
135 		}
136 	}
137 	return copy;
138 }
139 
140 static int
141 bdev_rados_context_init(const char *user_id, const char *rbd_pool_name, const char *const *config,
142 			rados_t *cluster, rados_ioctx_t *io_ctx)
143 {
144 	int ret;
145 
146 	ret = rados_create(cluster, user_id);
147 	if (ret < 0) {
148 		SPDK_ERRLOG("Failed to create rados_t struct\n");
149 		return -1;
150 	}
151 
152 	if (config) {
153 		const char *const *entry = config;
154 		while (*entry) {
155 			ret = rados_conf_set(*cluster, entry[0], entry[1]);
156 			if (ret < 0) {
157 				SPDK_ERRLOG("Failed to set %s = %s\n", entry[0], entry[1]);
158 				rados_shutdown(*cluster);
159 				return -1;
160 			}
161 			entry += 2;
162 		}
163 	} else {
164 		ret = rados_conf_read_file(*cluster, NULL);
165 		if (ret < 0) {
166 			SPDK_ERRLOG("Failed to read conf file\n");
167 			rados_shutdown(*cluster);
168 			return -1;
169 		}
170 	}
171 
172 	ret = rados_connect(*cluster);
173 	if (ret < 0) {
174 		SPDK_ERRLOG("Failed to connect to rbd_pool\n");
175 		rados_shutdown(*cluster);
176 		return -1;
177 	}
178 
179 	ret = rados_ioctx_create(*cluster, rbd_pool_name, io_ctx);
180 
181 	if (ret < 0) {
182 		SPDK_ERRLOG("Failed to create ioctx\n");
183 		rados_shutdown(*cluster);
184 		return -1;
185 	}
186 
187 	return 0;
188 }
189 
190 static int
191 bdev_rbd_init(struct bdev_rbd *rbd)
192 {
193 	int ret = 0;
194 	rados_t cluster = NULL;
195 	rados_ioctx_t io_ctx = NULL;
196 	rbd_image_t image = NULL;
197 
198 	ret = bdev_rados_context_init(rbd->user_id, rbd->pool_name, (const char *const *)rbd->config,
199 				      &cluster, &io_ctx);
200 	if (ret < 0) {
201 		SPDK_ERRLOG("Failed to create rados context for user_id=%s and rbd_pool=%s\n",
202 			    rbd->user_id ? rbd->user_id : "admin (the default)", rbd->pool_name);
203 		return -1;
204 	}
205 
206 	ret = rbd_open(io_ctx, rbd->rbd_name, &image, NULL);
207 	if (ret < 0) {
208 		SPDK_ERRLOG("Failed to open specified rbd device\n");
209 		goto end;
210 	}
211 	ret = rbd_stat(image, &rbd->info, sizeof(rbd->info));
212 	rbd_close(image);
213 	if (ret < 0) {
214 		SPDK_ERRLOG("Failed to stat specified rbd device\n");
215 	}
216 
217 end:
218 	rados_ioctx_destroy(io_ctx);
219 	rados_shutdown(cluster);
220 	return ret;
221 }
222 
223 static void
224 bdev_rbd_exit(rbd_image_t image)
225 {
226 	rbd_flush(image);
227 	rbd_close(image);
228 }
229 
230 static void
231 bdev_rbd_finish_aiocb(rbd_completion_t cb, void *arg)
232 {
233 	/* Doing nothing here */
234 }
235 
236 static void
237 bdev_rbd_start_aio(struct spdk_io_channel *ch, struct spdk_bdev_io *bdev_io,
238 		   struct iovec *iov, int iovcnt, uint64_t offset, size_t len)
239 {
240 	struct bdev_rbd_io_channel *rbdio_ch = spdk_io_channel_get_ctx(ch);
241 	int ret;
242 	rbd_completion_t comp;
243 	struct bdev_rbd_io *rbd_io;
244 	rbd_image_t image = rbdio_ch->image;
245 
246 	ret = rbd_aio_create_completion(bdev_io, bdev_rbd_finish_aiocb,
247 					&comp);
248 	if (ret < 0) {
249 		goto err;
250 	}
251 
252 	if (bdev_io->type == SPDK_BDEV_IO_TYPE_READ) {
253 		rbd_io = (struct bdev_rbd_io *)bdev_io->driver_ctx;
254 		rbd_io->total_len = len;
255 		if (spdk_likely(iovcnt == 1)) {
256 			ret = rbd_aio_read(image, offset, iov[0].iov_len, iov[0].iov_base, comp);
257 		} else {
258 			ret = rbd_aio_readv(image, iov, iovcnt, offset, comp);
259 		}
260 	} else if (bdev_io->type == SPDK_BDEV_IO_TYPE_WRITE) {
261 		if (spdk_likely(iovcnt == 1)) {
262 			ret = rbd_aio_write(image, offset, iov[0].iov_len, iov[0].iov_base, comp);
263 		} else {
264 			ret = rbd_aio_writev(image, iov, iovcnt, offset, comp);
265 		}
266 	} else if (bdev_io->type == SPDK_BDEV_IO_TYPE_FLUSH) {
267 		ret = rbd_aio_flush(image, comp);
268 	}
269 
270 	if (ret < 0) {
271 		rbd_aio_release(comp);
272 		goto err;
273 	}
274 
275 	return;
276 
277 err:
278 	spdk_bdev_io_complete(bdev_io, SPDK_BDEV_IO_STATUS_FAILED);
279 }
280 
281 static int bdev_rbd_library_init(void);
282 
283 static void bdev_rbd_library_fini(void);
284 
285 static int
286 bdev_rbd_get_ctx_size(void)
287 {
288 	return sizeof(struct bdev_rbd_io);
289 }
290 
291 static struct spdk_bdev_module rbd_if = {
292 	.name = "rbd",
293 	.module_init = bdev_rbd_library_init,
294 	.module_fini = bdev_rbd_library_fini,
295 	.get_ctx_size = bdev_rbd_get_ctx_size,
296 
297 };
298 SPDK_BDEV_MODULE_REGISTER(rbd, &rbd_if)
299 
300 static int
301 bdev_rbd_reset_timer(void *arg)
302 {
303 	struct bdev_rbd *disk = arg;
304 
305 	/*
306 	 * TODO: This should check if any I/O is still in flight before completing the reset.
307 	 * For now, just complete after the timer expires.
308 	 */
309 	spdk_bdev_io_complete(disk->reset_bdev_io, SPDK_BDEV_IO_STATUS_SUCCESS);
310 	spdk_poller_unregister(&disk->reset_timer);
311 	disk->reset_bdev_io = NULL;
312 
313 	return SPDK_POLLER_BUSY;
314 }
315 
316 static void
317 bdev_rbd_reset(struct bdev_rbd *disk, struct spdk_bdev_io *bdev_io)
318 {
319 	/*
320 	 * HACK: Since librbd doesn't provide any way to cancel outstanding aio, just kick off a
321 	 * timer to wait for in-flight I/O to complete.
322 	 */
323 	assert(disk->reset_bdev_io == NULL);
324 	disk->reset_bdev_io = bdev_io;
325 	disk->reset_timer = SPDK_POLLER_REGISTER(bdev_rbd_reset_timer, disk, 1 * 1000 * 1000);
326 }
327 
328 static int
329 bdev_rbd_destruct(void *ctx)
330 {
331 	struct bdev_rbd *rbd = ctx;
332 
333 	spdk_io_device_unregister(rbd, NULL);
334 
335 	bdev_rbd_free(rbd);
336 	return 0;
337 }
338 
339 static void
340 bdev_rbd_get_buf_cb(struct spdk_io_channel *ch, struct spdk_bdev_io *bdev_io,
341 		    bool success)
342 {
343 	if (!success) {
344 		spdk_bdev_io_complete(bdev_io, SPDK_BDEV_IO_STATUS_FAILED);
345 		return;
346 	}
347 
348 	bdev_rbd_start_aio(ch,
349 			   bdev_io,
350 			   bdev_io->u.bdev.iovs,
351 			   bdev_io->u.bdev.iovcnt,
352 			   bdev_io->u.bdev.offset_blocks * bdev_io->bdev->blocklen,
353 			   bdev_io->u.bdev.num_blocks * bdev_io->bdev->blocklen);
354 }
355 
356 static void
357 bdev_rbd_submit_request(struct spdk_io_channel *ch, struct spdk_bdev_io *bdev_io)
358 {
359 	switch (bdev_io->type) {
360 	case SPDK_BDEV_IO_TYPE_READ:
361 		spdk_bdev_io_get_buf(bdev_io, bdev_rbd_get_buf_cb,
362 				     bdev_io->u.bdev.num_blocks * bdev_io->bdev->blocklen);
363 		break;
364 
365 	case SPDK_BDEV_IO_TYPE_WRITE:
366 	case SPDK_BDEV_IO_TYPE_FLUSH:
367 		bdev_rbd_start_aio(ch,
368 				   bdev_io,
369 				   bdev_io->u.bdev.iovs,
370 				   bdev_io->u.bdev.iovcnt,
371 				   bdev_io->u.bdev.offset_blocks * bdev_io->bdev->blocklen,
372 				   bdev_io->u.bdev.num_blocks * bdev_io->bdev->blocklen);
373 		break;
374 
375 	case SPDK_BDEV_IO_TYPE_RESET:
376 		bdev_rbd_reset((struct bdev_rbd *)bdev_io->bdev->ctxt,
377 			       bdev_io);
378 		break;
379 
380 	default:
381 		SPDK_ERRLOG("Unsupported IO type =%d\n", bdev_io->type);
382 		spdk_bdev_io_complete(bdev_io, SPDK_BDEV_IO_STATUS_FAILED);
383 		break;
384 	}
385 }
386 
387 static bool
388 bdev_rbd_io_type_supported(void *ctx, enum spdk_bdev_io_type io_type)
389 {
390 	switch (io_type) {
391 	case SPDK_BDEV_IO_TYPE_READ:
392 	case SPDK_BDEV_IO_TYPE_WRITE:
393 	case SPDK_BDEV_IO_TYPE_FLUSH:
394 	case SPDK_BDEV_IO_TYPE_RESET:
395 		return true;
396 
397 	default:
398 		return false;
399 	}
400 }
401 
402 static void
403 bdev_rbd_io_poll(struct bdev_rbd_io_channel *ch)
404 {
405 	int i, io_status, rc;
406 	rbd_completion_t comps[SPDK_RBD_QUEUE_DEPTH];
407 	struct spdk_bdev_io *bdev_io;
408 	struct bdev_rbd_io *rbd_io;
409 	enum spdk_bdev_io_status bio_status;
410 
411 	rc = rbd_poll_io_events(ch->image, comps, SPDK_RBD_QUEUE_DEPTH);
412 	for (i = 0; i < rc; i++) {
413 		bdev_io = rbd_aio_get_arg(comps[i]);
414 		rbd_io = (struct bdev_rbd_io *)bdev_io->driver_ctx;
415 		io_status = rbd_aio_get_return_value(comps[i]);
416 		bio_status = SPDK_BDEV_IO_STATUS_SUCCESS;
417 
418 		if (bdev_io->type == SPDK_BDEV_IO_TYPE_READ) {
419 			if ((int)rbd_io->total_len != io_status) {
420 				bio_status = SPDK_BDEV_IO_STATUS_FAILED;
421 			}
422 		} else {
423 			/* For others, 0 means success */
424 			if (io_status != 0) {
425 				bio_status = SPDK_BDEV_IO_STATUS_FAILED;
426 			}
427 		}
428 
429 		rbd_aio_release(comps[i]);
430 
431 		spdk_bdev_io_complete(bdev_io, bio_status);
432 	}
433 }
434 
435 static void
436 bdev_rbd_free_channel(struct bdev_rbd_io_channel *ch)
437 {
438 	if (!ch) {
439 		return;
440 	}
441 
442 	if (ch->image) {
443 		bdev_rbd_exit(ch->image);
444 	}
445 
446 	if (ch->io_ctx) {
447 		rados_ioctx_destroy(ch->io_ctx);
448 	}
449 
450 	if (ch->cluster) {
451 		rados_shutdown(ch->cluster);
452 	}
453 
454 	if (ch->pfd >= 0) {
455 		close(ch->pfd);
456 	}
457 
458 	if (ch->group_ch) {
459 		spdk_put_io_channel(spdk_io_channel_from_ctx(ch->group_ch));
460 	}
461 }
462 
463 static void *
464 bdev_rbd_handle(void *arg)
465 {
466 	struct bdev_rbd_io_channel *ch = arg;
467 	void *ret = arg;
468 	int rc;
469 
470 	rc = bdev_rados_context_init(ch->disk->user_id, ch->disk->pool_name,
471 				     (const char *const *)ch->disk->config,
472 				     &ch->cluster, &ch->io_ctx);
473 	if (rc < 0) {
474 		SPDK_ERRLOG("Failed to create rados context for user_id %s and rbd_pool=%s\n",
475 			    ch->disk->user_id ? ch->disk->user_id : "admin (the default)", ch->disk->pool_name);
476 		ret = NULL;
477 		goto end;
478 	}
479 
480 	if (rbd_open(ch->io_ctx, ch->disk->rbd_name, &ch->image, NULL) < 0) {
481 		SPDK_ERRLOG("Failed to open specified rbd device\n");
482 		ret = NULL;
483 	}
484 
485 end:
486 	return ret;
487 }
488 
489 static int
490 bdev_rbd_create_cb(void *io_device, void *ctx_buf)
491 {
492 	struct bdev_rbd_io_channel *ch = ctx_buf;
493 	int ret;
494 	struct epoll_event event;
495 
496 	ch->disk = io_device;
497 	ch->image = NULL;
498 	ch->io_ctx = NULL;
499 	ch->pfd = -1;
500 
501 	if (spdk_call_unaffinitized(bdev_rbd_handle, ch) == NULL) {
502 		goto err;
503 	}
504 
505 	ch->pfd = eventfd(0, EFD_NONBLOCK);
506 	if (ch->pfd < 0) {
507 		SPDK_ERRLOG("Failed to get eventfd\n");
508 		goto err;
509 	}
510 
511 	ret = rbd_set_image_notification(ch->image, ch->pfd, EVENT_TYPE_EVENTFD);
512 	if (ret < 0) {
513 		SPDK_ERRLOG("Failed to set rbd image notification\n");
514 		goto err;
515 	}
516 
517 	ch->group_ch = spdk_io_channel_get_ctx(spdk_get_io_channel(&rbd_if));
518 	assert(ch->group_ch != NULL);
519 	memset(&event, 0, sizeof(event));
520 	event.events = EPOLLIN;
521 	event.data.ptr = ch;
522 
523 	ret = epoll_ctl(ch->group_ch->epoll_fd, EPOLL_CTL_ADD, ch->pfd, &event);
524 	if (ret < 0) {
525 		SPDK_ERRLOG("Failed to add the fd of ch(%p) to the epoll group from group_ch=%p\n", ch,
526 			    ch->group_ch);
527 		goto err;
528 	}
529 
530 	return 0;
531 
532 err:
533 	bdev_rbd_free_channel(ch);
534 	return -1;
535 }
536 
537 static void
538 bdev_rbd_destroy_cb(void *io_device, void *ctx_buf)
539 {
540 	struct bdev_rbd_io_channel *io_channel = ctx_buf;
541 	int rc;
542 
543 	rc = epoll_ctl(io_channel->group_ch->epoll_fd, EPOLL_CTL_DEL,
544 		       io_channel->pfd, NULL);
545 	if (rc < 0) {
546 		SPDK_ERRLOG("Failed to remove fd on io_channel=%p from the polling group=%p\n",
547 			    io_channel, io_channel->group_ch);
548 	}
549 
550 	bdev_rbd_free_channel(io_channel);
551 }
552 
553 static struct spdk_io_channel *
554 bdev_rbd_get_io_channel(void *ctx)
555 {
556 	struct bdev_rbd *rbd_bdev = ctx;
557 
558 	return spdk_get_io_channel(rbd_bdev);
559 }
560 
561 static int
562 bdev_rbd_dump_info_json(void *ctx, struct spdk_json_write_ctx *w)
563 {
564 	struct bdev_rbd *rbd_bdev = ctx;
565 
566 	spdk_json_write_named_object_begin(w, "rbd");
567 
568 	spdk_json_write_named_string(w, "pool_name", rbd_bdev->pool_name);
569 
570 	spdk_json_write_named_string(w, "rbd_name", rbd_bdev->rbd_name);
571 
572 	if (rbd_bdev->user_id) {
573 		spdk_json_write_named_string(w, "user_id", rbd_bdev->user_id);
574 	}
575 
576 	if (rbd_bdev->config) {
577 		char **entry = rbd_bdev->config;
578 
579 		spdk_json_write_named_object_begin(w, "config");
580 		while (*entry) {
581 			spdk_json_write_named_string(w, entry[0], entry[1]);
582 			entry += 2;
583 		}
584 		spdk_json_write_object_end(w);
585 	}
586 
587 	spdk_json_write_object_end(w);
588 
589 	return 0;
590 }
591 
592 static void
593 bdev_rbd_write_config_json(struct spdk_bdev *bdev, struct spdk_json_write_ctx *w)
594 {
595 	struct bdev_rbd *rbd = bdev->ctxt;
596 
597 	spdk_json_write_object_begin(w);
598 
599 	spdk_json_write_named_string(w, "method", "bdev_rbd_create");
600 
601 	spdk_json_write_named_object_begin(w, "params");
602 	spdk_json_write_named_string(w, "name", bdev->name);
603 	spdk_json_write_named_string(w, "pool_name", rbd->pool_name);
604 	spdk_json_write_named_string(w, "rbd_name", rbd->rbd_name);
605 	spdk_json_write_named_uint32(w, "block_size", bdev->blocklen);
606 	if (rbd->user_id) {
607 		spdk_json_write_named_string(w, "user_id", rbd->user_id);
608 	}
609 
610 	if (rbd->config) {
611 		char **entry = rbd->config;
612 
613 		spdk_json_write_named_object_begin(w, "config");
614 		while (*entry) {
615 			spdk_json_write_named_string(w, entry[0], entry[1]);
616 			entry += 2;
617 		}
618 		spdk_json_write_object_end(w);
619 	}
620 
621 	spdk_json_write_object_end(w);
622 
623 	spdk_json_write_object_end(w);
624 }
625 
626 static const struct spdk_bdev_fn_table rbd_fn_table = {
627 	.destruct		= bdev_rbd_destruct,
628 	.submit_request		= bdev_rbd_submit_request,
629 	.io_type_supported	= bdev_rbd_io_type_supported,
630 	.get_io_channel		= bdev_rbd_get_io_channel,
631 	.dump_info_json		= bdev_rbd_dump_info_json,
632 	.write_config_json	= bdev_rbd_write_config_json,
633 };
634 
635 int
636 bdev_rbd_create(struct spdk_bdev **bdev, const char *name, const char *user_id,
637 		const char *pool_name,
638 		const char *const *config,
639 		const char *rbd_name,
640 		uint32_t block_size)
641 {
642 	struct bdev_rbd *rbd;
643 	int ret;
644 
645 	if ((pool_name == NULL) || (rbd_name == NULL)) {
646 		return -EINVAL;
647 	}
648 
649 	rbd = calloc(1, sizeof(struct bdev_rbd));
650 	if (rbd == NULL) {
651 		SPDK_ERRLOG("Failed to allocate bdev_rbd struct\n");
652 		return -ENOMEM;
653 	}
654 
655 	rbd->rbd_name = strdup(rbd_name);
656 	if (!rbd->rbd_name) {
657 		bdev_rbd_free(rbd);
658 		return -ENOMEM;
659 	}
660 
661 	if (user_id) {
662 		rbd->user_id = strdup(user_id);
663 		if (!rbd->user_id) {
664 			bdev_rbd_free(rbd);
665 			return -ENOMEM;
666 		}
667 	}
668 
669 	rbd->pool_name = strdup(pool_name);
670 	if (!rbd->pool_name) {
671 		bdev_rbd_free(rbd);
672 		return -ENOMEM;
673 	}
674 
675 	if (config && !(rbd->config = bdev_rbd_dup_config(config))) {
676 		bdev_rbd_free(rbd);
677 		return -ENOMEM;
678 	}
679 
680 	ret = bdev_rbd_init(rbd);
681 	if (ret < 0) {
682 		bdev_rbd_free(rbd);
683 		SPDK_ERRLOG("Failed to init rbd device\n");
684 		return ret;
685 	}
686 
687 	if (name) {
688 		rbd->disk.name = strdup(name);
689 	} else {
690 		rbd->disk.name = spdk_sprintf_alloc("Ceph%d", bdev_rbd_count);
691 	}
692 	if (!rbd->disk.name) {
693 		bdev_rbd_free(rbd);
694 		return -ENOMEM;
695 	}
696 	rbd->disk.product_name = "Ceph Rbd Disk";
697 	bdev_rbd_count++;
698 
699 	rbd->disk.write_cache = 0;
700 	rbd->disk.blocklen = block_size;
701 	rbd->disk.blockcnt = rbd->info.size / rbd->disk.blocklen;
702 	rbd->disk.ctxt = rbd;
703 	rbd->disk.fn_table = &rbd_fn_table;
704 	rbd->disk.module = &rbd_if;
705 
706 	SPDK_NOTICELOG("Add %s rbd disk to lun\n", rbd->disk.name);
707 
708 	spdk_io_device_register(rbd, bdev_rbd_create_cb,
709 				bdev_rbd_destroy_cb,
710 				sizeof(struct bdev_rbd_io_channel),
711 				rbd_name);
712 	ret = spdk_bdev_register(&rbd->disk);
713 	if (ret) {
714 		spdk_io_device_unregister(rbd, NULL);
715 		bdev_rbd_free(rbd);
716 		return ret;
717 	}
718 
719 	*bdev = &(rbd->disk);
720 
721 	return ret;
722 }
723 
724 void
725 bdev_rbd_delete(struct spdk_bdev *bdev, spdk_delete_rbd_complete cb_fn, void *cb_arg)
726 {
727 	if (!bdev || bdev->module != &rbd_if) {
728 		cb_fn(cb_arg, -ENODEV);
729 		return;
730 	}
731 
732 	spdk_bdev_unregister(bdev, cb_fn, cb_arg);
733 }
734 
735 int
736 bdev_rbd_resize(struct spdk_bdev *bdev, const uint64_t new_size_in_mb)
737 {
738 	struct spdk_io_channel *ch;
739 	struct bdev_rbd_io_channel *rbd_io_ch;
740 	int rc;
741 	uint64_t new_size_in_byte;
742 	uint64_t current_size_in_mb;
743 
744 	if (bdev->module != &rbd_if) {
745 		return -EINVAL;
746 	}
747 
748 	current_size_in_mb = bdev->blocklen * bdev->blockcnt / (1024 * 1024);
749 	if (current_size_in_mb > new_size_in_mb) {
750 		SPDK_ERRLOG("The new bdev size must be lager than current bdev size.\n");
751 		return -EINVAL;
752 	}
753 
754 	ch = bdev_rbd_get_io_channel(bdev);
755 	rbd_io_ch = spdk_io_channel_get_ctx(ch);
756 	new_size_in_byte = new_size_in_mb * 1024 * 1024;
757 
758 	rc = rbd_resize(rbd_io_ch->image, new_size_in_byte);
759 	spdk_put_io_channel(ch);
760 	if (rc != 0) {
761 		SPDK_ERRLOG("failed to resize the ceph bdev.\n");
762 		return rc;
763 	}
764 
765 	rc = spdk_bdev_notify_blockcnt_change(bdev, new_size_in_byte / bdev->blocklen);
766 	if (rc != 0) {
767 		SPDK_ERRLOG("failed to notify block cnt change.\n");
768 		return rc;
769 	}
770 
771 	return rc;
772 }
773 
774 static int
775 bdev_rbd_group_poll(void *arg)
776 {
777 	struct bdev_rbd_group_channel *group_ch = arg;
778 	struct epoll_event events[MAX_EVENTS_PER_POLL];
779 	int num_events, i;
780 
781 	num_events = epoll_wait(group_ch->epoll_fd, events, MAX_EVENTS_PER_POLL, 0);
782 
783 	if (num_events <= 0) {
784 		return SPDK_POLLER_IDLE;
785 	}
786 
787 	for (i = 0; i < num_events; i++) {
788 		bdev_rbd_io_poll((struct bdev_rbd_io_channel *)events[i].data.ptr);
789 	}
790 
791 	return SPDK_POLLER_BUSY;
792 }
793 
794 static int
795 bdev_rbd_group_create_cb(void *io_device, void *ctx_buf)
796 {
797 	struct bdev_rbd_group_channel *ch = ctx_buf;
798 
799 	ch->epoll_fd = epoll_create1(0);
800 	if (ch->epoll_fd < 0) {
801 		SPDK_ERRLOG("Could not create epoll fd on io device=%p\n", io_device);
802 		return -1;
803 	}
804 
805 	ch->poller = SPDK_POLLER_REGISTER(bdev_rbd_group_poll, ch, 0);
806 
807 	return 0;
808 }
809 
810 static void
811 bdev_rbd_group_destroy_cb(void *io_device, void *ctx_buf)
812 {
813 	struct bdev_rbd_group_channel *ch = ctx_buf;
814 
815 	if (ch->epoll_fd >= 0) {
816 		close(ch->epoll_fd);
817 	}
818 
819 	spdk_poller_unregister(&ch->poller);
820 }
821 
822 static int
823 bdev_rbd_library_init(void)
824 {
825 	spdk_io_device_register(&rbd_if, bdev_rbd_group_create_cb, bdev_rbd_group_destroy_cb,
826 				sizeof(struct bdev_rbd_group_channel), "bdev_rbd_poll_groups");
827 
828 	return 0;
829 }
830 
831 static void
832 bdev_rbd_library_fini(void)
833 {
834 	spdk_io_device_unregister(&rbd_if, NULL);
835 }
836 
837 SPDK_LOG_REGISTER_COMPONENT(bdev_rbd)
838