projects/linux·blob master · refs

2806 lines · 69.9 KiB

1// SPDX-License-Identifier: GPL-2.0-or-later
2/*
3 * Network block device - make block devices work over TCP
4 *
5 * Note that you can not swap over this thing, yet. Seems to work but
6 * deadlocks sometimes - you can not swap over TCP in general.
7 *
8 * Copyright 1997-2000, 2008 Pavel Machek <[email protected]>
9 * Parts copyright 2001 Steven Whitehouse <[email protected]>
10 *
11 * (part of code stolen from loop.c)
12 */
13
14#define pr_fmt(fmt) "nbd: " fmt
15
16#include <linux/major.h>
17
18#include <linux/blkdev.h>
19#include <linux/module.h>
20#include <linux/init.h>
21#include <linux/sched.h>
22#include <linux/sched/mm.h>
23#include <linux/fs.h>
24#include <linux/bio.h>
25#include <linux/stat.h>
26#include <linux/errno.h>
27#include <linux/file.h>
28#include <linux/ioctl.h>
29#include <linux/mutex.h>
30#include <linux/compiler.h>
31#include <linux/completion.h>
32#include <linux/err.h>
33#include <linux/kernel.h>
34#include <linux/slab.h>
35#include <net/sock.h>
36#include <linux/net.h>
37#include <linux/kthread.h>
38#include <linux/types.h>
39#include <linux/debugfs.h>
40#include <linux/blk-mq.h>
41
42#include <linux/uaccess.h>
43#include <asm/types.h>
44
45#include <linux/nbd.h>
46#include <linux/nbd-netlink.h>
47#include <net/genetlink.h>
48
49#define CREATE_TRACE_POINTS
50#include <trace/events/nbd.h>
51
52static DEFINE_IDR(nbd_index_idr);
53static DEFINE_MUTEX(nbd_index_mutex);
54static struct workqueue_struct *nbd_del_wq;
55static int nbd_total_devices = 0;
56
57struct nbd_sock {
58 struct socket *sock;
59 struct mutex tx_lock;
60 struct request *pending;
61 int sent;
62 bool dead;
63 int fallback_index;
64 int cookie;
65 struct work_struct work;
66};
67
68struct recv_thread_args {
69 struct work_struct work;
70 struct nbd_device *nbd;
71 struct nbd_sock *nsock;
72 int index;
73};
74
75struct link_dead_args {
76 struct work_struct work;
77 int index;
78};
79
80#define NBD_RT_TIMEDOUT 0
81#define NBD_RT_DISCONNECT_REQUESTED 1
82#define NBD_RT_DISCONNECTED 2
83#define NBD_RT_HAS_PID_FILE 3
84#define NBD_RT_HAS_CONFIG_REF 4
85#define NBD_RT_BOUND 5
86#define NBD_RT_DISCONNECT_ON_CLOSE 6
87#define NBD_RT_HAS_BACKEND_FILE 7
88
89#define NBD_DESTROY_ON_DISCONNECT 0
90#define NBD_DISCONNECT_REQUESTED 1
91
92struct nbd_config {
93 u32 flags;
94 unsigned long runtime_flags;
95 u64 dead_conn_timeout;
96
97 struct nbd_sock **socks;
98 int num_connections;
99 atomic_t live_connections;
100 wait_queue_head_t conn_wait;
101
102 atomic_t recv_threads;
103 wait_queue_head_t recv_wq;
104 unsigned int blksize_bits;
105 loff_t bytesize;
106#if IS_ENABLED(CONFIG_DEBUG_FS)
107 struct dentry *dbg_dir;
108#endif
109};
110
111static inline unsigned int nbd_blksize(struct nbd_config *config)
112{
113 return 1u << config->blksize_bits;
114}
115
116struct nbd_device {
117 struct blk_mq_tag_set tag_set;
118
119 int index;
120 refcount_t config_refs;
121 refcount_t refs;
122 struct nbd_config *config;
123 struct mutex config_lock;
124 struct gendisk *disk;
125 struct workqueue_struct *recv_workq;
126 struct work_struct remove_work;
127
128 struct list_head list;
129 struct task_struct *task_setup;
130
131 unsigned long flags;
132 pid_t pid; /* pid of nbd-client, if attached */
133
134 char *backend;
135};
136
137#define NBD_CMD_REQUEUED 1
138/*
139 * This flag will be set if nbd_queue_rq() succeed, and will be checked and
140 * cleared in completion. Both setting and clearing of the flag are protected
141 * by cmd->lock.
142 */
143#define NBD_CMD_INFLIGHT 2
144
145/* Just part of request header or data payload is sent successfully */
146#define NBD_CMD_PARTIAL_SEND 3
147
148struct nbd_cmd {
149 struct nbd_device *nbd;
150 struct mutex lock;
151 int index;
152 int cookie;
153 int retries;
154 blk_status_t status;
155 unsigned long flags;
156 u32 cmd_cookie;
157};
158
159#if IS_ENABLED(CONFIG_DEBUG_FS)
160static struct dentry *nbd_dbg_dir;
161#endif
162
163#define nbd_name(nbd) ((nbd)->disk->disk_name)
164
165#define NBD_DEF_BLKSIZE_BITS 10
166
167static unsigned int nbds_max = 16;
168static int max_part = 16;
169static int pre_defined_connections = 1;
170static int part_shift;
171
172static int nbd_dev_dbg_init(struct nbd_device *nbd);
173static void nbd_dev_dbg_close(struct nbd_device *nbd);
174static void nbd_config_put(struct nbd_device *nbd);
175static void nbd_connect_reply(struct genl_info *info, int index);
176static int nbd_genl_status(struct sk_buff *skb, struct genl_info *info);
177static void nbd_dead_link_work(struct work_struct *work);
178static void nbd_disconnect_and_put(struct nbd_device *nbd);
179
180static inline struct device *nbd_to_dev(struct nbd_device *nbd)
181{
182 return disk_to_dev(nbd->disk);
183}
184
185static void nbd_requeue_cmd(struct nbd_cmd *cmd)
186{
187 struct request *req = blk_mq_rq_from_pdu(cmd);
188
189 lockdep_assert_held(&cmd->lock);
190
191 /*
192 * Clear INFLIGHT flag so that this cmd won't be completed in
193 * normal completion path
194 *
195 * INFLIGHT flag will be set when the cmd is queued to nbd next
196 * time.
197 */
198 __clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
199
200 if (!test_and_set_bit(NBD_CMD_REQUEUED, &cmd->flags))
201 blk_mq_requeue_request(req, true);
202}
203
204#define NBD_COOKIE_BITS 32
205
206static u64 nbd_cmd_handle(struct nbd_cmd *cmd)
207{
208 struct request *req = blk_mq_rq_from_pdu(cmd);
209 u32 tag = blk_mq_unique_tag(req);
210 u64 cookie = cmd->cmd_cookie;
211
212 return (cookie << NBD_COOKIE_BITS) | tag;
213}
214
215static u32 nbd_handle_to_tag(u64 handle)
216{
217 return (u32)handle;
218}
219
220static u32 nbd_handle_to_cookie(u64 handle)
221{
222 return (u32)(handle >> NBD_COOKIE_BITS);
223}
224
225static const char *nbdcmd_to_ascii(int cmd)
226{
227 switch (cmd) {
228 case NBD_CMD_READ: return "read";
229 case NBD_CMD_WRITE: return "write";
230 case NBD_CMD_DISC: return "disconnect";
231 case NBD_CMD_FLUSH: return "flush";
232 case NBD_CMD_TRIM: return "trim/discard";
233 }
234 return "invalid";
235}
236
237static ssize_t pid_show(struct device *dev,
238 struct device_attribute *attr, char *buf)
239{
240 struct gendisk *disk = dev_to_disk(dev);
241 struct nbd_device *nbd = disk->private_data;
242
243 return sprintf(buf, "%d\n", nbd->pid);
244}
245
246static const struct device_attribute pid_attr = {
247 .attr = { .name = "pid", .mode = 0444},
248 .show = pid_show,
249};
250
251static ssize_t backend_show(struct device *dev,
252 struct device_attribute *attr, char *buf)
253{
254 struct gendisk *disk = dev_to_disk(dev);
255 struct nbd_device *nbd = disk->private_data;
256
257 return sprintf(buf, "%s\n", nbd->backend ?: "");
258}
259
260static const struct device_attribute backend_attr = {
261 .attr = { .name = "backend", .mode = 0444},
262 .show = backend_show,
263};
264
265static void nbd_dev_remove(struct nbd_device *nbd)
266{
267 struct gendisk *disk = nbd->disk;
268
269 del_gendisk(disk);
270 blk_mq_free_tag_set(&nbd->tag_set);
271
272 /*
273 * Remove from idr after del_gendisk() completes, so if the same ID is
274 * reused, the following add_disk() will succeed.
275 */
276 mutex_lock(&nbd_index_mutex);
277 idr_remove(&nbd_index_idr, nbd->index);
278 mutex_unlock(&nbd_index_mutex);
279 destroy_workqueue(nbd->recv_workq);
280 put_disk(disk);
281}
282
283static void nbd_dev_remove_work(struct work_struct *work)
284{
285 nbd_dev_remove(container_of(work, struct nbd_device, remove_work));
286}
287
288static void nbd_put(struct nbd_device *nbd)
289{
290 if (!refcount_dec_and_test(&nbd->refs))
291 return;
292
293 /* Call del_gendisk() asynchrounously to prevent deadlock */
294 if (test_bit(NBD_DESTROY_ON_DISCONNECT, &nbd->flags))
295 queue_work(nbd_del_wq, &nbd->remove_work);
296 else
297 nbd_dev_remove(nbd);
298}
299
300static int nbd_disconnected(struct nbd_config *config)
301{
302 return test_bit(NBD_RT_DISCONNECTED, &config->runtime_flags) ||
303 test_bit(NBD_RT_DISCONNECT_REQUESTED, &config->runtime_flags);
304}
305
306static void nbd_mark_nsock_dead(struct nbd_device *nbd, struct nbd_sock *nsock,
307 int notify)
308{
309 if (!nsock->dead && notify && !nbd_disconnected(nbd->config)) {
310 struct link_dead_args *args;
311 args = kmalloc_obj(struct link_dead_args, GFP_NOIO);
312 if (args) {
313 INIT_WORK(&args->work, nbd_dead_link_work);
314 args->index = nbd->index;
315 queue_work(system_percpu_wq, &args->work);
316 }
317 }
318 if (!nsock->dead) {
319 kernel_sock_shutdown(nsock->sock, SHUT_RDWR);
320 if (atomic_dec_return(&nbd->config->live_connections) == 0) {
321 if (test_and_clear_bit(NBD_RT_DISCONNECT_REQUESTED,
322 &nbd->config->runtime_flags)) {
323 set_bit(NBD_RT_DISCONNECTED,
324 &nbd->config->runtime_flags);
325 dev_info(nbd_to_dev(nbd),
326 "Disconnected due to user request.\n");
327 }
328 }
329 }
330 nsock->dead = true;
331 nsock->pending = NULL;
332 nsock->sent = 0;
333}
334
335static void nbd_apply_limits(struct queue_limits *lim, u32 flags)
336{
337 lim->features &= ~(BLK_FEAT_WRITE_CACHE | BLK_FEAT_FUA | BLK_FEAT_ROTATIONAL);
338 lim->max_hw_discard_sectors = 0;
339 lim->max_write_zeroes_sectors = 0;
340
341 if (flags & NBD_FLAG_SEND_TRIM)
342 lim->max_hw_discard_sectors = UINT_MAX >> SECTOR_SHIFT;
343 if (flags & NBD_FLAG_SEND_FLUSH) {
344 lim->features |= BLK_FEAT_WRITE_CACHE;
345 if (flags & NBD_FLAG_SEND_FUA)
346 lim->features |= BLK_FEAT_FUA;
347 }
348
349 if (flags & NBD_FLAG_ROTATIONAL)
350 lim->features |= BLK_FEAT_ROTATIONAL;
351 if (flags & NBD_FLAG_SEND_WRITE_ZEROES)
352 lim->max_write_zeroes_sectors = UINT_MAX >> SECTOR_SHIFT;
353}
354
355static int nbd_set_size(struct nbd_device *nbd, loff_t bytesize, loff_t blksize)
356{
357 struct queue_limits lim;
358 int error;
359
360 if (!blksize)
361 blksize = 1u << NBD_DEF_BLKSIZE_BITS;
362
363 if (blk_validate_block_size(blksize))
364 return -EINVAL;
365
366 if (bytesize < 0)
367 return -EINVAL;
368
369 nbd->config->bytesize = bytesize;
370 nbd->config->blksize_bits = __ffs(blksize);
371
372 if (!nbd->pid)
373 return 0;
374
375 lim = queue_limits_start_update(nbd->disk->queue);
376 nbd_apply_limits(&lim, nbd->config->flags);
377 lim.logical_block_size = blksize;
378 lim.physical_block_size = blksize;
379 /* No need freeze with 0 capacity and write cache disabled */
380 if (!get_capacity(nbd->disk) && !blk_queue_write_cache(nbd->disk->queue))
381 error = queue_limits_commit_update(nbd->disk->queue, &lim);
382 else
383 error = queue_limits_commit_update_frozen(nbd->disk->queue, &lim);
384 if (error)
385 return error;
386
387 if (max_part)
388 set_bit(GD_NEED_PART_SCAN, &nbd->disk->state);
389 if (!set_capacity_and_notify(nbd->disk, bytesize >> 9))
390 kobject_uevent(&nbd_to_dev(nbd)->kobj, KOBJ_CHANGE);
391 return 0;
392}
393
394static void nbd_complete_rq(struct request *req)
395{
396 struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
397
398 dev_dbg(nbd_to_dev(cmd->nbd), "request %p: %s\n", req,
399 cmd->status ? "failed" : "done");
400
401 blk_mq_end_request(req, cmd->status);
402}
403
404/*
405 * Forcibly shutdown the socket causing all listeners to error
406 */
407static void sock_shutdown(struct nbd_device *nbd)
408{
409 struct nbd_config *config = nbd->config;
410 int i;
411
412 if (config->num_connections == 0)
413 return;
414 if (test_and_set_bit(NBD_RT_DISCONNECTED, &config->runtime_flags))
415 return;
416
417 for (i = 0; i < config->num_connections; i++) {
418 struct nbd_sock *nsock = config->socks[i];
419 mutex_lock(&nsock->tx_lock);
420 nbd_mark_nsock_dead(nbd, nsock, 0);
421 mutex_unlock(&nsock->tx_lock);
422 }
423 dev_warn(disk_to_dev(nbd->disk), "shutting down sockets\n");
424}
425
426static u32 req_to_nbd_cmd_type(struct request *req)
427{
428 switch (req_op(req)) {
429 case REQ_OP_DISCARD:
430 return NBD_CMD_TRIM;
431 case REQ_OP_FLUSH:
432 return NBD_CMD_FLUSH;
433 case REQ_OP_WRITE:
434 return NBD_CMD_WRITE;
435 case REQ_OP_READ:
436 return NBD_CMD_READ;
437 case REQ_OP_WRITE_ZEROES:
438 return NBD_CMD_WRITE_ZEROES;
439 default:
440 return U32_MAX;
441 }
442}
443
444static struct nbd_config *nbd_get_config_unlocked(struct nbd_device *nbd)
445{
446 if (refcount_inc_not_zero(&nbd->config_refs)) {
447 /*
448 * Add smp_mb__after_atomic to ensure that reading nbd->config_refs
449 * and reading nbd->config is ordered. The pair is the barrier in
450 * nbd_alloc_and_init_config(), avoid nbd->config_refs is set
451 * before nbd->config.
452 */
453 smp_mb__after_atomic();
454 return nbd->config;
455 }
456
457 return NULL;
458}
459
460static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req)
461{
462 struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
463 struct nbd_device *nbd = cmd->nbd;
464 struct nbd_config *config;
465
466 if (!mutex_trylock(&cmd->lock))
467 return BLK_EH_RESET_TIMER;
468
469 /* partial send is handled in nbd_sock's work function */
470 if (test_bit(NBD_CMD_PARTIAL_SEND, &cmd->flags)) {
471 mutex_unlock(&cmd->lock);
472 return BLK_EH_RESET_TIMER;
473 }
474
475 if (!test_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
476 mutex_unlock(&cmd->lock);
477 return BLK_EH_DONE;
478 }
479
480 config = nbd_get_config_unlocked(nbd);
481 if (!config) {
482 cmd->status = BLK_STS_TIMEOUT;
483 __clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
484 mutex_unlock(&cmd->lock);
485 goto done;
486 }
487
488 if (config->num_connections > 1 ||
489 (config->num_connections == 1 && nbd->tag_set.timeout)) {
490 dev_err_ratelimited(nbd_to_dev(nbd),
491 "Connection timed out, retrying (%d/%d alive)\n",
492 atomic_read(&config->live_connections),
493 config->num_connections);
494 /*
495 * Hooray we have more connections, requeue this IO, the submit
496 * path will put it on a real connection. Or if only one
497 * connection is configured, the submit path will wait util
498 * a new connection is reconfigured or util dead timeout.
499 */
500 if (config->socks) {
501 if (cmd->index < config->num_connections) {
502 struct nbd_sock *nsock =
503 config->socks[cmd->index];
504 mutex_lock(&nsock->tx_lock);
505 /* We can have multiple outstanding requests, so
506 * we don't want to mark the nsock dead if we've
507 * already reconnected with a new socket, so
508 * only mark it dead if its the same socket we
509 * were sent out on.
510 */
511 if (cmd->cookie == nsock->cookie)
512 nbd_mark_nsock_dead(nbd, nsock, 1);
513 mutex_unlock(&nsock->tx_lock);
514 }
515 nbd_requeue_cmd(cmd);
516 mutex_unlock(&cmd->lock);
517 nbd_config_put(nbd);
518 return BLK_EH_DONE;
519 }
520 }
521
522 if (!nbd->tag_set.timeout) {
523 /*
524 * Userspace sets timeout=0 to disable socket disconnection,
525 * so just warn and reset the timer.
526 */
527 struct nbd_sock *nsock = config->socks[cmd->index];
528 cmd->retries++;
529 dev_info(nbd_to_dev(nbd), "Possible stuck request %p: control (%s@%llu,%uB). Runtime %u seconds\n",
530 req, nbdcmd_to_ascii(req_to_nbd_cmd_type(req)),
531 (unsigned long long)blk_rq_pos(req) << 9,
532 blk_rq_bytes(req), (req->timeout / HZ) * cmd->retries);
533
534 mutex_lock(&nsock->tx_lock);
535 if (cmd->cookie != nsock->cookie) {
536 nbd_requeue_cmd(cmd);
537 mutex_unlock(&nsock->tx_lock);
538 mutex_unlock(&cmd->lock);
539 nbd_config_put(nbd);
540 return BLK_EH_DONE;
541 }
542 mutex_unlock(&nsock->tx_lock);
543 mutex_unlock(&cmd->lock);
544 nbd_config_put(nbd);
545 return BLK_EH_RESET_TIMER;
546 }
547
548 dev_err_ratelimited(nbd_to_dev(nbd), "Connection timed out\n");
549 set_bit(NBD_RT_TIMEDOUT, &config->runtime_flags);
550 cmd->status = BLK_STS_IOERR;
551 __clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
552 mutex_unlock(&cmd->lock);
553 sock_shutdown(nbd);
554 nbd_config_put(nbd);
555done:
556 blk_mq_complete_request(req);
557 return BLK_EH_DONE;
558}
559
560static int __sock_xmit(struct nbd_device *nbd, struct socket *sock, int send,
561 struct iov_iter *iter, int msg_flags, int *sent)
562{
563 int result;
564 struct msghdr msg = {} ;
565 unsigned int noreclaim_flag;
566
567 if (unlikely(!sock)) {
568 dev_err_ratelimited(disk_to_dev(nbd->disk),
569 "Attempted %s on closed socket in sock_xmit\n",
570 (send ? "send" : "recv"));
571 return -EINVAL;
572 }
573
574 msg.msg_iter = *iter;
575
576 noreclaim_flag = memalloc_noreclaim_save();
577
578 scoped_with_kernel_creds() {
579 do {
580 sock->sk->sk_allocation = GFP_NOIO | __GFP_MEMALLOC;
581 sock->sk->sk_use_task_frag = false;
582 msg.msg_flags = msg_flags | MSG_NOSIGNAL;
583
584 if (send)
585 result = sock_sendmsg(sock, &msg);
586 else
587 result = sock_recvmsg(sock, &msg, msg.msg_flags);
588
589 if (result <= 0) {
590 if (result == 0)
591 result = -EPIPE; /* short read */
592 break;
593 }
594 if (sent)
595 *sent += result;
596 } while (msg_data_left(&msg));
597 }
598
599 memalloc_noreclaim_restore(noreclaim_flag);
600
601 return result;
602}
603
604/*
605 * Send or receive packet. Return a positive value on success and
606 * negtive value on failure, and never return 0.
607 */
608static int sock_xmit(struct nbd_device *nbd, int index, int send,
609 struct iov_iter *iter, int msg_flags, int *sent)
610{
611 struct nbd_config *config = nbd->config;
612 struct socket *sock = config->socks[index]->sock;
613
614 return __sock_xmit(nbd, sock, send, iter, msg_flags, sent);
615}
616
617/*
618 * Different settings for sk->sk_sndtimeo can result in different return values
619 * if there is a signal pending when we enter sendmsg, because reasons?
620 */
621static inline int was_interrupted(int result)
622{
623 return result == -ERESTARTSYS || result == -EINTR;
624}
625
626/*
627 * We've already sent header or part of data payload, have no choice but
628 * to set pending and schedule it in work.
629 *
630 * And we have to return BLK_STS_OK to block core, otherwise this same
631 * request may be re-dispatched with different tag, but our header has
632 * been sent out with old tag, and this way does confuse reply handling.
633 */
634static void nbd_sched_pending_work(struct nbd_device *nbd,
635 struct nbd_sock *nsock,
636 struct nbd_cmd *cmd, int sent)
637{
638 struct request *req = blk_mq_rq_from_pdu(cmd);
639
640 /* pending work should be scheduled only once */
641 WARN_ON_ONCE(test_bit(NBD_CMD_PARTIAL_SEND, &cmd->flags));
642
643 nsock->pending = req;
644 nsock->sent = sent;
645 set_bit(NBD_CMD_PARTIAL_SEND, &cmd->flags);
646 refcount_inc(&nbd->config_refs);
647 schedule_work(&nsock->work);
648}
649
650/*
651 * Returns BLK_STS_RESOURCE if the caller should retry after a delay.
652 * Returns BLK_STS_IOERR if sending failed.
653 */
654static blk_status_t nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd,
655 int index)
656{
657 struct request *req = blk_mq_rq_from_pdu(cmd);
658 struct nbd_config *config = nbd->config;
659 struct nbd_sock *nsock = config->socks[index];
660 int result;
661 struct nbd_request request = {.magic = htonl(NBD_REQUEST_MAGIC)};
662 struct kvec iov = {.iov_base = &request, .iov_len = sizeof(request)};
663 struct iov_iter from;
664 struct bio *bio;
665 u64 handle;
666 u32 type;
667 u32 nbd_cmd_flags = 0;
668 int sent = nsock->sent, skip = 0;
669
670 lockdep_assert_held(&cmd->lock);
671 lockdep_assert_held(&nsock->tx_lock);
672
673 iov_iter_kvec(&from, ITER_SOURCE, &iov, 1, sizeof(request));
674
675 type = req_to_nbd_cmd_type(req);
676 if (type == U32_MAX)
677 return BLK_STS_IOERR;
678
679 if (rq_data_dir(req) == WRITE &&
680 (config->flags & NBD_FLAG_READ_ONLY)) {
681 dev_err_ratelimited(disk_to_dev(nbd->disk),
682 "Write on read-only\n");
683 return BLK_STS_IOERR;
684 }
685
686 if (req->cmd_flags & REQ_FUA)
687 nbd_cmd_flags |= NBD_CMD_FLAG_FUA;
688 if ((req->cmd_flags & REQ_NOUNMAP) && (type == NBD_CMD_WRITE_ZEROES))
689 nbd_cmd_flags |= NBD_CMD_FLAG_NO_HOLE;
690
691 /* We did a partial send previously, and we at least sent the whole
692 * request struct, so just go and send the rest of the pages in the
693 * request.
694 */
695 if (sent) {
696 if (sent >= sizeof(request)) {
697 skip = sent - sizeof(request);
698
699 /* initialize handle for tracing purposes */
700 handle = nbd_cmd_handle(cmd);
701
702 goto send_pages;
703 }
704 iov_iter_advance(&from, sent);
705 } else {
706 cmd->cmd_cookie++;
707 }
708 cmd->index = index;
709 cmd->cookie = nsock->cookie;
710 cmd->retries = 0;
711 request.type = htonl(type | nbd_cmd_flags);
712 if (type != NBD_CMD_FLUSH) {
713 request.from = cpu_to_be64((u64)blk_rq_pos(req) << 9);
714 request.len = htonl(blk_rq_bytes(req));
715 }
716 handle = nbd_cmd_handle(cmd);
717 request.cookie = cpu_to_be64(handle);
718
719 trace_nbd_send_request(&request, nbd->index, blk_mq_rq_from_pdu(cmd));
720
721 dev_dbg(nbd_to_dev(nbd), "request %p: sending control (%s@%llu,%uB)\n",
722 req, nbdcmd_to_ascii(type),
723 (unsigned long long)blk_rq_pos(req) << 9, blk_rq_bytes(req));
724 result = sock_xmit(nbd, index, 1, &from,
725 (type == NBD_CMD_WRITE) ? MSG_MORE : 0, &sent);
726 trace_nbd_header_sent(req, handle);
727 if (result < 0) {
728 if (was_interrupted(result)) {
729 /* If we haven't sent anything we can just return BUSY,
730 * however if we have sent something we need to make
731 * sure we only allow this req to be sent until we are
732 * completely done.
733 */
734 if (sent) {
735 nbd_sched_pending_work(nbd, nsock, cmd, sent);
736 return BLK_STS_OK;
737 }
738 set_bit(NBD_CMD_REQUEUED, &cmd->flags);
739 return BLK_STS_RESOURCE;
740 }
741 dev_err_ratelimited(disk_to_dev(nbd->disk),
742 "Send control failed (result %d)\n", result);
743 goto requeue;
744 }
745send_pages:
746 if (type != NBD_CMD_WRITE)
747 goto out;
748
749 bio = req->bio;
750 while (bio) {
751 struct bio *next = bio->bi_next;
752 struct bvec_iter iter;
753 struct bio_vec bvec;
754
755 bio_for_each_segment(bvec, bio, iter) {
756 bool is_last = !next && bio_iter_last(bvec, iter);
757 int flags = is_last ? 0 : MSG_MORE;
758
759 dev_dbg(nbd_to_dev(nbd), "request %p: sending %d bytes data\n",
760 req, bvec.bv_len);
761 iov_iter_bvec(&from, ITER_SOURCE, &bvec, 1, bvec.bv_len);
762 if (skip) {
763 if (skip >= iov_iter_count(&from)) {
764 skip -= iov_iter_count(&from);
765 continue;
766 }
767 iov_iter_advance(&from, skip);
768 skip = 0;
769 }
770 result = sock_xmit(nbd, index, 1, &from, flags, &sent);
771 if (result < 0) {
772 if (was_interrupted(result)) {
773 nbd_sched_pending_work(nbd, nsock, cmd, sent);
774 return BLK_STS_OK;
775 }
776 dev_err(disk_to_dev(nbd->disk),
777 "Send data failed (result %d)\n",
778 result);
779 goto requeue;
780 }
781 /*
782 * The completion might already have come in,
783 * so break for the last one instead of letting
784 * the iterator do it. This prevents use-after-free
785 * of the bio.
786 */
787 if (is_last)
788 break;
789 }
790 bio = next;
791 }
792out:
793 trace_nbd_payload_sent(req, handle);
794 nsock->pending = NULL;
795 nsock->sent = 0;
796 __set_bit(NBD_CMD_INFLIGHT, &cmd->flags);
797 return BLK_STS_OK;
798
799requeue:
800 /*
801 * Can't requeue in case we are dealing with partial send
802 *
803 * We must run from pending work function.
804 * */
805 if (test_bit(NBD_CMD_PARTIAL_SEND, &cmd->flags))
806 return BLK_STS_OK;
807
808 /* retry on a different socket */
809 dev_err_ratelimited(disk_to_dev(nbd->disk),
810 "Request send failed, requeueing\n");
811 nbd_mark_nsock_dead(nbd, nsock, 1);
812 nbd_requeue_cmd(cmd);
813 return BLK_STS_OK;
814}
815
816/* handle partial sending */
817static void nbd_pending_cmd_work(struct work_struct *work)
818{
819 struct nbd_sock *nsock = container_of(work, struct nbd_sock, work);
820 struct request *req = nsock->pending;
821 struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
822 struct nbd_device *nbd = cmd->nbd;
823 unsigned long deadline = READ_ONCE(req->deadline);
824 unsigned int wait_ms = 2;
825
826 mutex_lock(&cmd->lock);
827
828 WARN_ON_ONCE(test_bit(NBD_CMD_REQUEUED, &cmd->flags));
829 if (WARN_ON_ONCE(!test_bit(NBD_CMD_PARTIAL_SEND, &cmd->flags)))
830 goto out;
831
832 mutex_lock(&nsock->tx_lock);
833 while (true) {
834 nbd_send_cmd(nbd, cmd, cmd->index);
835 if (!nsock->pending)
836 break;
837
838 /* don't bother timeout handler for partial sending */
839 if (READ_ONCE(jiffies) + msecs_to_jiffies(wait_ms) >= deadline) {
840 cmd->status = BLK_STS_IOERR;
841 blk_mq_complete_request(req);
842 break;
843 }
844 msleep(wait_ms);
845 wait_ms *= 2;
846 }
847 mutex_unlock(&nsock->tx_lock);
848 clear_bit(NBD_CMD_PARTIAL_SEND, &cmd->flags);
849out:
850 mutex_unlock(&cmd->lock);
851 nbd_config_put(nbd);
852}
853
854static int nbd_read_reply(struct nbd_device *nbd, struct socket *sock,
855 struct nbd_reply *reply)
856{
857 struct kvec iov = {.iov_base = reply, .iov_len = sizeof(*reply)};
858 struct iov_iter to;
859 int result;
860
861 reply->magic = 0;
862 iov_iter_kvec(&to, ITER_DEST, &iov, 1, sizeof(*reply));
863 result = __sock_xmit(nbd, sock, 0, &to, MSG_WAITALL, NULL);
864 if (result < 0) {
865 if (!nbd_disconnected(nbd->config))
866 dev_err(disk_to_dev(nbd->disk),
867 "Receive control failed (result %d)\n", result);
868 return result;
869 }
870
871 if (ntohl(reply->magic) != NBD_REPLY_MAGIC) {
872 dev_err(disk_to_dev(nbd->disk), "Wrong magic (0x%lx)\n",
873 (unsigned long)ntohl(reply->magic));
874 return -EPROTO;
875 }
876
877 return 0;
878}
879
880/* NULL returned = something went wrong, inform userspace */
881static struct nbd_cmd *nbd_handle_reply(struct nbd_device *nbd, int index,
882 struct nbd_reply *reply)
883{
884 int result;
885 struct nbd_cmd *cmd;
886 struct request *req = NULL;
887 u64 handle;
888 u16 hwq;
889 u32 tag;
890 int ret = 0;
891
892 handle = be64_to_cpu(reply->cookie);
893 tag = nbd_handle_to_tag(handle);
894 hwq = blk_mq_unique_tag_to_hwq(tag);
895 if (hwq < nbd->tag_set.nr_hw_queues)
896 req = blk_mq_tag_to_rq(nbd->tag_set.tags[hwq],
897 blk_mq_unique_tag_to_tag(tag));
898 if (!req || !blk_mq_request_started(req)) {
899 dev_err(disk_to_dev(nbd->disk), "Unexpected reply (%d) %p\n",
900 tag, req);
901 return ERR_PTR(-ENOENT);
902 }
903 trace_nbd_header_received(req, handle);
904 cmd = blk_mq_rq_to_pdu(req);
905
906 mutex_lock(&cmd->lock);
907 if (!test_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
908 dev_err(disk_to_dev(nbd->disk), "Suspicious reply %d (status %u flags %lu)",
909 tag, cmd->status, cmd->flags);
910 ret = -ENOENT;
911 goto out;
912 }
913 if (cmd->index != index) {
914 dev_err(disk_to_dev(nbd->disk), "Unexpected reply %d from different sock %d (expected %d)",
915 tag, index, cmd->index);
916 ret = -ENOENT;
917 goto out;
918 }
919 if (cmd->cmd_cookie != nbd_handle_to_cookie(handle)) {
920 dev_err(disk_to_dev(nbd->disk), "Double reply on req %p, cmd_cookie %u, handle cookie %u\n",
921 req, cmd->cmd_cookie, nbd_handle_to_cookie(handle));
922 ret = -ENOENT;
923 goto out;
924 }
925 if (cmd->status != BLK_STS_OK) {
926 dev_err(disk_to_dev(nbd->disk), "Command already handled %p\n",
927 req);
928 ret = -ENOENT;
929 goto out;
930 }
931 if (test_bit(NBD_CMD_REQUEUED, &cmd->flags)) {
932 dev_err(disk_to_dev(nbd->disk), "Raced with timeout on req %p\n",
933 req);
934 ret = -ENOENT;
935 goto out;
936 }
937 if (ntohl(reply->error)) {
938 dev_err(disk_to_dev(nbd->disk), "Other side returned error (%d)\n",
939 ntohl(reply->error));
940 cmd->status = BLK_STS_IOERR;
941 goto out;
942 }
943
944 dev_dbg(nbd_to_dev(nbd), "request %p: got reply\n", req);
945 if (rq_data_dir(req) != WRITE) {
946 struct req_iterator iter;
947 struct bio_vec bvec;
948 struct iov_iter to;
949
950 rq_for_each_segment(bvec, req, iter) {
951 iov_iter_bvec(&to, ITER_DEST, &bvec, 1, bvec.bv_len);
952 result = sock_xmit(nbd, index, 0, &to, MSG_WAITALL, NULL);
953 if (result < 0) {
954 dev_err(disk_to_dev(nbd->disk), "Receive data failed (result %d)\n",
955 result);
956 /*
957 * If we've disconnected, we need to make sure we
958 * complete this request, otherwise error out
959 * and let the timeout stuff handle resubmitting
960 * this request onto another connection.
961 */
962 if (nbd_disconnected(nbd->config)) {
963 cmd->status = BLK_STS_IOERR;
964 goto out;
965 }
966 ret = -EIO;
967 goto out;
968 }
969 dev_dbg(nbd_to_dev(nbd), "request %p: got %d bytes data\n",
970 req, bvec.bv_len);
971 }
972 }
973out:
974 trace_nbd_payload_received(req, handle);
975 mutex_unlock(&cmd->lock);
976 return ret ? ERR_PTR(ret) : cmd;
977}
978
979static void recv_work(struct work_struct *work)
980{
981 struct recv_thread_args *args = container_of(work,
982 struct recv_thread_args,
983 work);
984 struct nbd_device *nbd = args->nbd;
985 struct nbd_config *config = nbd->config;
986 struct request_queue *q = nbd->disk->queue;
987 struct nbd_sock *nsock = args->nsock;
988 struct nbd_cmd *cmd;
989 struct request *rq;
990
991 while (1) {
992 struct nbd_reply reply;
993
994 if (nbd_read_reply(nbd, nsock->sock, &reply))
995 break;
996
997 /*
998 * Grab .q_usage_counter so request pool won't go away, then no
999 * request use-after-free is possible during nbd_handle_reply().
1000 * If queue is frozen, there won't be any inflight requests, we
1001 * needn't to handle the incoming garbage message.
1002 */
1003 if (!percpu_ref_tryget(&q->q_usage_counter)) {
1004 dev_err(disk_to_dev(nbd->disk), "%s: no io inflight\n",
1005 __func__);
1006 break;
1007 }
1008
1009 cmd = nbd_handle_reply(nbd, args->index, &reply);
1010 if (IS_ERR(cmd)) {
1011 percpu_ref_put(&q->q_usage_counter);
1012 break;
1013 }
1014
1015 rq = blk_mq_rq_from_pdu(cmd);
1016 if (likely(!blk_should_fake_timeout(rq->q))) {
1017 bool complete;
1018
1019 mutex_lock(&cmd->lock);
1020 complete = __test_and_clear_bit(NBD_CMD_INFLIGHT,
1021 &cmd->flags);
1022 mutex_unlock(&cmd->lock);
1023 if (complete)
1024 blk_mq_complete_request(rq);
1025 }
1026 percpu_ref_put(&q->q_usage_counter);
1027 }
1028
1029 mutex_lock(&nsock->tx_lock);
1030 nbd_mark_nsock_dead(nbd, nsock, 1);
1031 mutex_unlock(&nsock->tx_lock);
1032
1033 atomic_dec(&config->recv_threads);
1034 wake_up(&config->recv_wq);
1035 nbd_config_put(nbd);
1036 kfree(args);
1037}
1038
1039static bool nbd_clear_req(struct request *req, void *data)
1040{
1041 struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
1042
1043 /* don't abort one completed request */
1044 if (blk_mq_request_completed(req))
1045 return true;
1046
1047 mutex_lock(&cmd->lock);
1048 if (!__test_and_clear_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
1049 mutex_unlock(&cmd->lock);
1050 return true;
1051 }
1052 cmd->status = BLK_STS_IOERR;
1053 mutex_unlock(&cmd->lock);
1054
1055 blk_mq_complete_request(req);
1056 return true;
1057}
1058
1059static void nbd_clear_que(struct nbd_device *nbd)
1060{
1061 blk_mq_quiesce_queue(nbd->disk->queue);
1062 blk_mq_tagset_busy_iter(&nbd->tag_set, nbd_clear_req, NULL);
1063 blk_mq_unquiesce_queue(nbd->disk->queue);
1064 dev_dbg(disk_to_dev(nbd->disk), "queue cleared\n");
1065}
1066
1067static int find_fallback(struct nbd_device *nbd, int index)
1068{
1069 struct nbd_config *config = nbd->config;
1070 int new_index = -1;
1071 struct nbd_sock *nsock = config->socks[index];
1072 int fallback = nsock->fallback_index;
1073 int i;
1074
1075 if (test_bit(NBD_RT_DISCONNECTED, &config->runtime_flags))
1076 return new_index;
1077
1078 if (config->num_connections <= 1)
1079 goto no_fallback;
1080
1081 if (fallback >= 0 && fallback < config->num_connections &&
1082 !config->socks[fallback]->dead)
1083 return fallback;
1084
1085 for (i = 0; i < config->num_connections; i++) {
1086 if (i != index && !config->socks[i]->dead) {
1087 new_index = i;
1088 break;
1089 }
1090 }
1091 nsock->fallback_index = new_index;
1092 if (new_index >= 0)
1093 return new_index;
1094
1095no_fallback:
1096 dev_err_ratelimited(disk_to_dev(nbd->disk),
1097 "Dead connection, failed to find a fallback\n");
1098 return new_index;
1099}
1100
1101static int wait_for_reconnect(struct nbd_device *nbd)
1102{
1103 struct nbd_config *config = nbd->config;
1104 if (!config->dead_conn_timeout)
1105 return 0;
1106
1107 if (!wait_event_timeout(config->conn_wait,
1108 test_bit(NBD_RT_DISCONNECTED,
1109 &config->runtime_flags) ||
1110 atomic_read(&config->live_connections) > 0,
1111 config->dead_conn_timeout))
1112 return 0;
1113
1114 return !test_bit(NBD_RT_DISCONNECTED, &config->runtime_flags);
1115}
1116
1117static blk_status_t nbd_handle_cmd(struct nbd_cmd *cmd, int index)
1118{
1119 struct request *req = blk_mq_rq_from_pdu(cmd);
1120 struct nbd_device *nbd = cmd->nbd;
1121 struct nbd_config *config;
1122 struct nbd_sock *nsock;
1123 blk_status_t ret;
1124
1125 lockdep_assert_held(&cmd->lock);
1126
1127 config = nbd_get_config_unlocked(nbd);
1128 if (!config) {
1129 dev_err_ratelimited(disk_to_dev(nbd->disk),
1130 "Socks array is empty\n");
1131 return BLK_STS_IOERR;
1132 }
1133
1134 if (index >= config->num_connections) {
1135 dev_err_ratelimited(disk_to_dev(nbd->disk),
1136 "Attempted send on invalid socket\n");
1137 nbd_config_put(nbd);
1138 return BLK_STS_IOERR;
1139 }
1140 cmd->status = BLK_STS_OK;
1141again:
1142 nsock = config->socks[index];
1143 mutex_lock(&nsock->tx_lock);
1144 if (nsock->dead) {
1145 int old_index = index;
1146 index = find_fallback(nbd, index);
1147 mutex_unlock(&nsock->tx_lock);
1148 if (index < 0) {
1149 if (wait_for_reconnect(nbd)) {
1150 index = old_index;
1151 goto again;
1152 }
1153 /* All the sockets should already be down at this point,
1154 * we just want to make sure that DISCONNECTED is set so
1155 * any requests that come in that were queue'ed waiting
1156 * for the reconnect timer don't trigger the timer again
1157 * and instead just error out.
1158 */
1159 sock_shutdown(nbd);
1160 nbd_config_put(nbd);
1161 return BLK_STS_IOERR;
1162 }
1163 goto again;
1164 }
1165
1166 /* Handle the case that we have a pending request that was partially
1167 * transmitted that _has_ to be serviced first. We need to call requeue
1168 * here so that it gets put _after_ the request that is already on the
1169 * dispatch list.
1170 */
1171 blk_mq_start_request(req);
1172 if (unlikely(nsock->pending && nsock->pending != req)) {
1173 nbd_requeue_cmd(cmd);
1174 ret = BLK_STS_OK;
1175 goto out;
1176 }
1177 ret = nbd_send_cmd(nbd, cmd, index);
1178out:
1179 mutex_unlock(&nsock->tx_lock);
1180 nbd_config_put(nbd);
1181 return ret;
1182}
1183
1184static blk_status_t nbd_queue_rq(struct blk_mq_hw_ctx *hctx,
1185 const struct blk_mq_queue_data *bd)
1186{
1187 struct nbd_cmd *cmd = blk_mq_rq_to_pdu(bd->rq);
1188 blk_status_t ret;
1189
1190 /*
1191 * Since we look at the bio's to send the request over the network we
1192 * need to make sure the completion work doesn't mark this request done
1193 * before we are done doing our send. This keeps us from dereferencing
1194 * freed data if we have particularly fast completions (ie we get the
1195 * completion before we exit sock_xmit on the last bvec) or in the case
1196 * that the server is misbehaving (or there was an error) before we're
1197 * done sending everything over the wire.
1198 */
1199 mutex_lock(&cmd->lock);
1200 clear_bit(NBD_CMD_REQUEUED, &cmd->flags);
1201
1202 /* We can be called directly from the user space process, which means we
1203 * could possibly have signals pending so our sendmsg will fail. In
1204 * this case we need to return that we are busy, otherwise error out as
1205 * appropriate.
1206 */
1207 ret = nbd_handle_cmd(cmd, hctx->queue_num);
1208 mutex_unlock(&cmd->lock);
1209
1210 return ret;
1211}
1212
1213static struct socket *nbd_get_socket(struct nbd_device *nbd, unsigned long fd,
1214 int *err)
1215{
1216 struct socket *sock;
1217
1218 *err = 0;
1219 sock = sockfd_lookup(fd, err);
1220 if (!sock)
1221 return NULL;
1222
1223 if (!sk_is_tcp(sock->sk) &&
1224 !sk_is_stream_unix(sock->sk)) {
1225 dev_err(disk_to_dev(nbd->disk), "Unsupported socket: should be TCP or UNIX.\n");
1226 *err = -EINVAL;
1227 sockfd_put(sock);
1228 return NULL;
1229 }
1230
1231 if (sock->ops->shutdown == sock_no_shutdown) {
1232 dev_err(disk_to_dev(nbd->disk), "Unsupported socket: shutdown callout must be supported.\n");
1233 *err = -EINVAL;
1234 sockfd_put(sock);
1235 return NULL;
1236 }
1237
1238 return sock;
1239}
1240
1241#ifdef CONFIG_DEBUG_LOCK_ALLOC
1242static struct lock_class_key nbd_key[3];
1243static struct lock_class_key nbd_slock_key[3];
1244
1245static void nbd_reclassify_socket(struct socket *sock)
1246{
1247 struct sock *sk = sock->sk;
1248
1249 if (!sock_allow_reclassification(sk))
1250 return;
1251
1252 switch (sk->sk_family) {
1253 case AF_INET:
1254 sock_lock_init_class_and_name(sk, "slock-AF_INET-NBD",
1255 &nbd_slock_key[0],
1256 "sk_lock-AF_INET-NBD",
1257 &nbd_key[0]);
1258 break;
1259 case AF_INET6:
1260 sock_lock_init_class_and_name(sk, "slock-AF_INET6-NBD",
1261 &nbd_slock_key[1],
1262 "sk_lock-AF_INET6-NBD",
1263 &nbd_key[1]);
1264 break;
1265 case AF_UNIX:
1266 sock_lock_init_class_and_name(sk, "slock-AF_UNIX-NBD",
1267 &nbd_slock_key[2],
1268 "sk_lock-AF_UNIX-NBD",
1269 &nbd_key[2]);
1270 break;
1271 }
1272}
1273#else
1274static inline void nbd_reclassify_socket(struct socket *sock) {}
1275#endif
1276
1277static int nbd_add_socket(struct nbd_device *nbd, unsigned long arg,
1278 bool netlink)
1279{
1280 struct nbd_config *config = nbd->config;
1281 struct socket *sock;
1282 struct nbd_sock **socks;
1283 struct nbd_sock *nsock;
1284 int err;
1285
1286 /* Arg will be cast to int, check it to avoid overflow */
1287 if (arg > INT_MAX)
1288 return -EINVAL;
1289
1290 if (nbd->pid) {
1291 dev_err(disk_to_dev(nbd->disk),
1292 "Cannot add socket to a running device\n");
1293 return -EBUSY;
1294 }
1295
1296 sock = nbd_get_socket(nbd, arg, &err);
1297 if (!sock)
1298 return err;
1299 nbd_reclassify_socket(sock);
1300
1301 if (!netlink && !nbd->task_setup &&
1302 !test_bit(NBD_RT_BOUND, &config->runtime_flags))
1303 nbd->task_setup = current;
1304
1305 if (!netlink &&
1306 (nbd->task_setup != current ||
1307 test_bit(NBD_RT_BOUND, &config->runtime_flags))) {
1308 dev_err(disk_to_dev(nbd->disk),
1309 "Device being setup by another task");
1310 err = -EBUSY;
1311 goto put_socket;
1312 }
1313
1314 nsock = kzalloc_obj(*nsock);
1315 if (!nsock) {
1316 err = -ENOMEM;
1317 goto put_socket;
1318 }
1319
1320 socks = krealloc(config->socks, (config->num_connections + 1) *
1321 sizeof(struct nbd_sock *), GFP_KERNEL);
1322 if (!socks) {
1323 kfree(nsock);
1324 err = -ENOMEM;
1325 goto put_socket;
1326 }
1327
1328 config->socks = socks;
1329
1330 nsock->fallback_index = -1;
1331 nsock->dead = false;
1332 mutex_init(&nsock->tx_lock);
1333 nsock->sock = sock;
1334 nsock->pending = NULL;
1335 nsock->sent = 0;
1336 nsock->cookie = 0;
1337 INIT_WORK(&nsock->work, nbd_pending_cmd_work);
1338 socks[config->num_connections++] = nsock;
1339 atomic_inc(&config->live_connections);
1340
1341 return 0;
1342
1343put_socket:
1344 sockfd_put(sock);
1345 return err;
1346}
1347
1348static int nbd_genl_reconnect_sock_cb(struct nbd_device *nbd, unsigned long arg)
1349{
1350 struct nbd_config *config = nbd->config;
1351 struct socket *sock, *old;
1352 struct recv_thread_args *args;
1353 int i;
1354 int err;
1355
1356 sock = nbd_get_socket(nbd, arg, &err);
1357 if (!sock)
1358 return err;
1359
1360 args = kzalloc_obj(*args);
1361 if (!args) {
1362 sockfd_put(sock);
1363 return -ENOMEM;
1364 }
1365
1366 for (i = 0; i < config->num_connections; i++) {
1367 struct nbd_sock *nsock = config->socks[i];
1368
1369 if (!nsock->dead)
1370 continue;
1371
1372 mutex_lock(&nsock->tx_lock);
1373 if (!nsock->dead) {
1374 mutex_unlock(&nsock->tx_lock);
1375 continue;
1376 }
1377 sk_set_memalloc(sock->sk);
1378 if (nbd->tag_set.timeout)
1379 sock->sk->sk_sndtimeo = nbd->tag_set.timeout;
1380 atomic_inc(&config->recv_threads);
1381 refcount_inc(&nbd->config_refs);
1382 old = nsock->sock;
1383 nsock->fallback_index = -1;
1384 nsock->sock = sock;
1385 nsock->dead = false;
1386 INIT_WORK(&args->work, recv_work);
1387 args->index = i;
1388 args->nbd = nbd;
1389 args->nsock = nsock;
1390 nsock->cookie++;
1391 mutex_unlock(&nsock->tx_lock);
1392 sockfd_put(old);
1393
1394 clear_bit(NBD_RT_DISCONNECTED, &config->runtime_flags);
1395
1396 /* We take the tx_mutex in an error path in the recv_work, so we
1397 * need to queue_work outside of the tx_mutex.
1398 */
1399 queue_work(nbd->recv_workq, &args->work);
1400
1401 atomic_inc(&config->live_connections);
1402 wake_up(&config->conn_wait);
1403 dev_info(nbd_to_dev(nbd), "reconnected socket\n");
1404 return 0;
1405 }
1406 sockfd_put(sock);
1407 kfree(args);
1408 return 1;
1409}
1410
1411static void nbd_bdev_reset(struct nbd_device *nbd)
1412{
1413 if (disk_openers(nbd->disk) > 1)
1414 return;
1415 set_capacity(nbd->disk, 0);
1416}
1417
1418static void nbd_parse_flags(struct nbd_device *nbd)
1419{
1420 if (nbd->config->flags & NBD_FLAG_READ_ONLY)
1421 set_disk_ro(nbd->disk, true);
1422 else
1423 set_disk_ro(nbd->disk, false);
1424}
1425
1426static void send_disconnects(struct nbd_device *nbd)
1427{
1428 struct nbd_config *config = nbd->config;
1429 struct nbd_request request = {
1430 .magic = htonl(NBD_REQUEST_MAGIC),
1431 .type = htonl(NBD_CMD_DISC),
1432 };
1433 struct kvec iov = {.iov_base = &request, .iov_len = sizeof(request)};
1434 struct iov_iter from;
1435 int i, ret;
1436
1437 for (i = 0; i < config->num_connections; i++) {
1438 struct nbd_sock *nsock = config->socks[i];
1439
1440 iov_iter_kvec(&from, ITER_SOURCE, &iov, 1, sizeof(request));
1441 mutex_lock(&nsock->tx_lock);
1442 ret = sock_xmit(nbd, i, 1, &from, 0, NULL);
1443 if (ret < 0)
1444 dev_err(disk_to_dev(nbd->disk),
1445 "Send disconnect failed %d\n", ret);
1446 mutex_unlock(&nsock->tx_lock);
1447 }
1448}
1449
1450static int nbd_disconnect(struct nbd_device *nbd)
1451{
1452 struct nbd_config *config = nbd->config;
1453
1454 dev_info(disk_to_dev(nbd->disk), "NBD_DISCONNECT\n");
1455 set_bit(NBD_RT_DISCONNECT_REQUESTED, &config->runtime_flags);
1456 set_bit(NBD_DISCONNECT_REQUESTED, &nbd->flags);
1457 send_disconnects(nbd);
1458 return 0;
1459}
1460
1461static void nbd_clear_sock(struct nbd_device *nbd)
1462{
1463 sock_shutdown(nbd);
1464 nbd_clear_que(nbd);
1465 nbd->task_setup = NULL;
1466}
1467
1468static void nbd_config_put(struct nbd_device *nbd)
1469{
1470 if (refcount_dec_and_mutex_lock(&nbd->config_refs,
1471 &nbd->config_lock)) {
1472 struct nbd_config *config = nbd->config;
1473 struct queue_limits lim;
1474 nbd_dev_dbg_close(nbd);
1475 invalidate_disk(nbd->disk);
1476 /* reset queue limits to default */
1477 lim = queue_limits_start_update(nbd->disk->queue);
1478 nbd_apply_limits(&lim, 0);
1479 queue_limits_commit_update(nbd->disk->queue, &lim);
1480 if (nbd->config->bytesize)
1481 kobject_uevent(&nbd_to_dev(nbd)->kobj, KOBJ_CHANGE);
1482 if (test_and_clear_bit(NBD_RT_HAS_PID_FILE,
1483 &config->runtime_flags))
1484 device_remove_file(disk_to_dev(nbd->disk), &pid_attr);
1485 nbd->pid = 0;
1486 if (test_and_clear_bit(NBD_RT_HAS_BACKEND_FILE,
1487 &config->runtime_flags)) {
1488 device_remove_file(disk_to_dev(nbd->disk), &backend_attr);
1489 kfree(nbd->backend);
1490 nbd->backend = NULL;
1491 }
1492 nbd_clear_sock(nbd);
1493 if (config->num_connections) {
1494 int i;
1495 for (i = 0; i < config->num_connections; i++) {
1496 sockfd_put(config->socks[i]->sock);
1497 kfree(config->socks[i]);
1498 }
1499 kfree(config->socks);
1500 }
1501 kfree(nbd->config);
1502 nbd->config = NULL;
1503
1504 nbd->tag_set.timeout = 0;
1505
1506 mutex_unlock(&nbd->config_lock);
1507 nbd_put(nbd);
1508 module_put(THIS_MODULE);
1509 }
1510}
1511
1512static int nbd_start_device(struct nbd_device *nbd)
1513{
1514 struct nbd_config *config = nbd->config;
1515 int num_connections = config->num_connections;
1516 int error = 0, i;
1517
1518 if (nbd->pid)
1519 return -EBUSY;
1520 if (!config->socks)
1521 return -EINVAL;
1522 if (num_connections > 1 &&
1523 !(config->flags & NBD_FLAG_CAN_MULTI_CONN)) {
1524 dev_err(disk_to_dev(nbd->disk), "server does not support multiple connections per device.\n");
1525 return -EINVAL;
1526 }
1527
1528retry:
1529 mutex_unlock(&nbd->config_lock);
1530 blk_mq_update_nr_hw_queues(&nbd->tag_set, num_connections);
1531 mutex_lock(&nbd->config_lock);
1532
1533 /* if another code path updated nr_hw_queues, retry until succeed */
1534 if (num_connections != config->num_connections) {
1535 num_connections = config->num_connections;
1536 goto retry;
1537 }
1538
1539 nbd->pid = task_pid_nr(current);
1540
1541 nbd_parse_flags(nbd);
1542
1543 error = device_create_file(disk_to_dev(nbd->disk), &pid_attr);
1544 if (error) {
1545 dev_err(disk_to_dev(nbd->disk), "device_create_file failed for pid!\n");
1546 return error;
1547 }
1548 set_bit(NBD_RT_HAS_PID_FILE, &config->runtime_flags);
1549
1550 nbd_dev_dbg_init(nbd);
1551 for (i = 0; i < num_connections; i++) {
1552 struct recv_thread_args *args;
1553
1554 args = kzalloc_obj(*args);
1555 if (!args) {
1556 sock_shutdown(nbd);
1557 /*
1558 * If num_connections is m (2 < m),
1559 * and NO.1 ~ NO.n(1 < n < m) kzallocs are successful.
1560 * But NO.(n + 1) failed. We still have n recv threads.
1561 * So, add flush_workqueue here to prevent recv threads
1562 * dropping the last config_refs and trying to destroy
1563 * the workqueue from inside the workqueue.
1564 */
1565 if (i)
1566 flush_workqueue(nbd->recv_workq);
1567 return -ENOMEM;
1568 }
1569 sk_set_memalloc(config->socks[i]->sock->sk);
1570 if (nbd->tag_set.timeout)
1571 config->socks[i]->sock->sk->sk_sndtimeo =
1572 nbd->tag_set.timeout;
1573 atomic_inc(&config->recv_threads);
1574 refcount_inc(&nbd->config_refs);
1575 INIT_WORK(&args->work, recv_work);
1576 args->nbd = nbd;
1577 args->nsock = config->socks[i];
1578 args->index = i;
1579 queue_work(nbd->recv_workq, &args->work);
1580 }
1581 return nbd_set_size(nbd, config->bytesize, nbd_blksize(config));
1582}
1583
1584static int nbd_start_device_ioctl(struct nbd_device *nbd)
1585{
1586 struct nbd_config *config = nbd->config;
1587 int ret;
1588
1589 ret = nbd_start_device(nbd);
1590 if (ret)
1591 return ret;
1592
1593 if (max_part)
1594 set_bit(GD_NEED_PART_SCAN, &nbd->disk->state);
1595 mutex_unlock(&nbd->config_lock);
1596 ret = wait_event_interruptible(config->recv_wq,
1597 atomic_read(&config->recv_threads) == 0);
1598 if (ret) {
1599 sock_shutdown(nbd);
1600 nbd_clear_que(nbd);
1601 }
1602
1603 flush_workqueue(nbd->recv_workq);
1604 mutex_lock(&nbd->config_lock);
1605 nbd_bdev_reset(nbd);
1606 /* user requested, ignore socket errors */
1607 if (test_bit(NBD_RT_DISCONNECT_REQUESTED, &config->runtime_flags))
1608 ret = 0;
1609 if (test_bit(NBD_RT_TIMEDOUT, &config->runtime_flags))
1610 ret = -ETIMEDOUT;
1611 return ret;
1612}
1613
1614static void nbd_clear_sock_ioctl(struct nbd_device *nbd)
1615{
1616 nbd_clear_sock(nbd);
1617 disk_force_media_change(nbd->disk);
1618 nbd_bdev_reset(nbd);
1619 if (test_and_clear_bit(NBD_RT_HAS_CONFIG_REF,
1620 &nbd->config->runtime_flags))
1621 nbd_config_put(nbd);
1622}
1623
1624static void nbd_set_cmd_timeout(struct nbd_device *nbd, u64 timeout)
1625{
1626 nbd->tag_set.timeout = timeout * HZ;
1627 if (timeout)
1628 blk_queue_rq_timeout(nbd->disk->queue, timeout * HZ);
1629 else
1630 blk_queue_rq_timeout(nbd->disk->queue, 30 * HZ);
1631}
1632
1633/* Must be called with config_lock held */
1634static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
1635 unsigned int cmd, unsigned long arg)
1636{
1637 struct nbd_config *config = nbd->config;
1638 loff_t bytesize;
1639
1640 switch (cmd) {
1641 case NBD_DISCONNECT:
1642 return nbd_disconnect(nbd);
1643 case NBD_CLEAR_SOCK:
1644 nbd_clear_sock_ioctl(nbd);
1645 return 0;
1646 case NBD_SET_SOCK:
1647 return nbd_add_socket(nbd, arg, false);
1648 case NBD_SET_BLKSIZE:
1649 return nbd_set_size(nbd, config->bytesize, arg);
1650 case NBD_SET_SIZE:
1651 return nbd_set_size(nbd, arg, nbd_blksize(config));
1652 case NBD_SET_SIZE_BLOCKS:
1653 if (check_shl_overflow(arg, config->blksize_bits, &bytesize))
1654 return -EINVAL;
1655 return nbd_set_size(nbd, bytesize, nbd_blksize(config));
1656 case NBD_SET_TIMEOUT:
1657 nbd_set_cmd_timeout(nbd, arg);
1658 return 0;
1659
1660 case NBD_SET_FLAGS:
1661 config->flags = arg;
1662 return 0;
1663 case NBD_DO_IT:
1664 return nbd_start_device_ioctl(nbd);
1665 case NBD_CLEAR_QUE:
1666 /*
1667 * This is for compatibility only. The queue is always cleared
1668 * by NBD_DO_IT or NBD_CLEAR_SOCK.
1669 */
1670 return 0;
1671 case NBD_PRINT_DEBUG:
1672 /*
1673 * For compatibility only, we no longer keep a list of
1674 * outstanding requests.
1675 */
1676 return 0;
1677 }
1678 return -ENOTTY;
1679}
1680
1681static int nbd_ioctl(struct block_device *bdev, blk_mode_t mode,
1682 unsigned int cmd, unsigned long arg)
1683{
1684 struct nbd_device *nbd = bdev->bd_disk->private_data;
1685 struct nbd_config *config = nbd->config;
1686 int error = -EINVAL;
1687
1688 if (!capable(CAP_SYS_ADMIN))
1689 return -EPERM;
1690
1691 /* The block layer will pass back some non-nbd ioctls in case we have
1692 * special handling for them, but we don't so just return an error.
1693 */
1694 if (_IOC_TYPE(cmd) != 0xab)
1695 return -EINVAL;
1696
1697 mutex_lock(&nbd->config_lock);
1698
1699 /* Don't allow ioctl operations on a nbd device that was created with
1700 * netlink, unless it's DISCONNECT or CLEAR_SOCK, which are fine.
1701 */
1702 if (!test_bit(NBD_RT_BOUND, &config->runtime_flags) ||
1703 (cmd == NBD_DISCONNECT || cmd == NBD_CLEAR_SOCK))
1704 error = __nbd_ioctl(bdev, nbd, cmd, arg);
1705 else
1706 dev_err(nbd_to_dev(nbd), "Cannot use ioctl interface on a netlink controlled device.\n");
1707 mutex_unlock(&nbd->config_lock);
1708 return error;
1709}
1710
1711static int nbd_alloc_and_init_config(struct nbd_device *nbd)
1712{
1713 struct nbd_config *config;
1714
1715 if (WARN_ON(nbd->config))
1716 return -EINVAL;
1717
1718 if (!try_module_get(THIS_MODULE))
1719 return -ENODEV;
1720
1721 config = kzalloc_obj(struct nbd_config, GFP_NOFS);
1722 if (!config) {
1723 module_put(THIS_MODULE);
1724 return -ENOMEM;
1725 }
1726
1727 atomic_set(&config->recv_threads, 0);
1728 init_waitqueue_head(&config->recv_wq);
1729 init_waitqueue_head(&config->conn_wait);
1730 config->blksize_bits = NBD_DEF_BLKSIZE_BITS;
1731 atomic_set(&config->live_connections, 0);
1732
1733 nbd->config = config;
1734 /*
1735 * Order refcount_set(&nbd->config_refs, 1) and nbd->config assignment,
1736 * its pair is the barrier in nbd_get_config_unlocked().
1737 * So nbd_get_config_unlocked() won't see nbd->config as null after
1738 * refcount_inc_not_zero() succeed.
1739 */
1740 smp_mb__before_atomic();
1741 refcount_set(&nbd->config_refs, 1);
1742
1743 return 0;
1744}
1745
1746static int nbd_open(struct gendisk *disk, blk_mode_t mode)
1747{
1748 struct nbd_device *nbd;
1749 struct nbd_config *config;
1750 int ret = 0;
1751
1752 mutex_lock(&nbd_index_mutex);
1753 nbd = disk->private_data;
1754 if (!nbd) {
1755 ret = -ENXIO;
1756 goto out;
1757 }
1758 if (!refcount_inc_not_zero(&nbd->refs)) {
1759 ret = -ENXIO;
1760 goto out;
1761 }
1762
1763 config = nbd_get_config_unlocked(nbd);
1764 if (!config) {
1765 mutex_lock(&nbd->config_lock);
1766 if (refcount_inc_not_zero(&nbd->config_refs)) {
1767 mutex_unlock(&nbd->config_lock);
1768 goto out;
1769 }
1770 ret = nbd_alloc_and_init_config(nbd);
1771 if (ret) {
1772 mutex_unlock(&nbd->config_lock);
1773 goto out;
1774 }
1775
1776 refcount_inc(&nbd->refs);
1777 mutex_unlock(&nbd->config_lock);
1778 if (max_part)
1779 set_bit(GD_NEED_PART_SCAN, &disk->state);
1780 } else if (nbd_disconnected(config)) {
1781 if (max_part)
1782 set_bit(GD_NEED_PART_SCAN, &disk->state);
1783 }
1784out:
1785 mutex_unlock(&nbd_index_mutex);
1786 return ret;
1787}
1788
1789static void nbd_release(struct gendisk *disk)
1790{
1791 struct nbd_device *nbd = disk->private_data;
1792
1793 if (test_bit(NBD_RT_DISCONNECT_ON_CLOSE, &nbd->config->runtime_flags) &&
1794 disk_openers(disk) == 0)
1795 nbd_disconnect_and_put(nbd);
1796
1797 nbd_config_put(nbd);
1798 nbd_put(nbd);
1799}
1800
1801static void nbd_free_disk(struct gendisk *disk)
1802{
1803 struct nbd_device *nbd = disk->private_data;
1804
1805 kfree(nbd);
1806}
1807
1808static const struct block_device_operations nbd_fops =
1809{
1810 .owner = THIS_MODULE,
1811 .open = nbd_open,
1812 .release = nbd_release,
1813 .ioctl = nbd_ioctl,
1814 .compat_ioctl = nbd_ioctl,
1815 .free_disk = nbd_free_disk,
1816};
1817
1818#if IS_ENABLED(CONFIG_DEBUG_FS)
1819
1820static int nbd_dbg_tasks_show(struct seq_file *s, void *unused)
1821{
1822 struct nbd_device *nbd = s->private;
1823
1824 if (nbd->pid)
1825 seq_printf(s, "recv: %d\n", nbd->pid);
1826
1827 return 0;
1828}
1829
1830DEFINE_SHOW_ATTRIBUTE(nbd_dbg_tasks);
1831
1832static int nbd_dbg_flags_show(struct seq_file *s, void *unused)
1833{
1834 struct nbd_device *nbd = s->private;
1835 u32 flags = nbd->config->flags;
1836
1837 seq_printf(s, "Hex: 0x%08x\n\n", flags);
1838
1839 seq_puts(s, "Known flags:\n");
1840
1841 if (flags & NBD_FLAG_HAS_FLAGS)
1842 seq_puts(s, "NBD_FLAG_HAS_FLAGS\n");
1843 if (flags & NBD_FLAG_READ_ONLY)
1844 seq_puts(s, "NBD_FLAG_READ_ONLY\n");
1845 if (flags & NBD_FLAG_SEND_FLUSH)
1846 seq_puts(s, "NBD_FLAG_SEND_FLUSH\n");
1847 if (flags & NBD_FLAG_SEND_FUA)
1848 seq_puts(s, "NBD_FLAG_SEND_FUA\n");
1849 if (flags & NBD_FLAG_SEND_TRIM)
1850 seq_puts(s, "NBD_FLAG_SEND_TRIM\n");
1851 if (flags & NBD_FLAG_SEND_WRITE_ZEROES)
1852 seq_puts(s, "NBD_FLAG_SEND_WRITE_ZEROES\n");
1853 if (flags & NBD_FLAG_ROTATIONAL)
1854 seq_puts(s, "NBD_FLAG_ROTATIONAL\n");
1855
1856 return 0;
1857}
1858
1859DEFINE_SHOW_ATTRIBUTE(nbd_dbg_flags);
1860
1861static int nbd_dev_dbg_init(struct nbd_device *nbd)
1862{
1863 struct dentry *dir;
1864 struct nbd_config *config = nbd->config;
1865
1866 if (!nbd_dbg_dir)
1867 return -EIO;
1868
1869 dir = debugfs_create_dir(nbd_name(nbd), nbd_dbg_dir);
1870 if (IS_ERR(dir)) {
1871 dev_err(nbd_to_dev(nbd), "Failed to create debugfs dir for '%s'\n",
1872 nbd_name(nbd));
1873 return -EIO;
1874 }
1875 config->dbg_dir = dir;
1876
1877 debugfs_create_file("tasks", 0444, dir, nbd, &nbd_dbg_tasks_fops);
1878 debugfs_create_u64("size_bytes", 0444, dir, &config->bytesize);
1879 debugfs_create_u32("timeout", 0444, dir, &nbd->tag_set.timeout);
1880 debugfs_create_u32("blocksize_bits", 0444, dir, &config->blksize_bits);
1881 debugfs_create_file("flags", 0444, dir, nbd, &nbd_dbg_flags_fops);
1882
1883 return 0;
1884}
1885
1886static void nbd_dev_dbg_close(struct nbd_device *nbd)
1887{
1888 debugfs_remove_recursive(nbd->config->dbg_dir);
1889}
1890
1891static int nbd_dbg_init(void)
1892{
1893 struct dentry *dbg_dir;
1894
1895 dbg_dir = debugfs_create_dir("nbd", NULL);
1896 if (IS_ERR(dbg_dir))
1897 return -EIO;
1898
1899 nbd_dbg_dir = dbg_dir;
1900
1901 return 0;
1902}
1903
1904static void nbd_dbg_close(void)
1905{
1906 debugfs_remove_recursive(nbd_dbg_dir);
1907}
1908
1909#else /* IS_ENABLED(CONFIG_DEBUG_FS) */
1910
1911static int nbd_dev_dbg_init(struct nbd_device *nbd)
1912{
1913 return 0;
1914}
1915
1916static void nbd_dev_dbg_close(struct nbd_device *nbd)
1917{
1918}
1919
1920static int nbd_dbg_init(void)
1921{
1922 return 0;
1923}
1924
1925static void nbd_dbg_close(void)
1926{
1927}
1928
1929#endif
1930
1931static int nbd_init_request(struct blk_mq_tag_set *set, struct request *rq,
1932 unsigned int hctx_idx, int numa_node)
1933{
1934 struct nbd_cmd *cmd = blk_mq_rq_to_pdu(rq);
1935 cmd->nbd = set->driver_data;
1936 cmd->flags = 0;
1937 mutex_init(&cmd->lock);
1938 return 0;
1939}
1940
1941static const struct blk_mq_ops nbd_mq_ops = {
1942 .queue_rq = nbd_queue_rq,
1943 .complete = nbd_complete_rq,
1944 .init_request = nbd_init_request,
1945 .timeout = nbd_xmit_timeout,
1946};
1947
1948static struct nbd_device *nbd_dev_add(int index, unsigned int refs,
1949 int nr_hw_queues)
1950{
1951 struct queue_limits lim = {
1952 .max_hw_sectors = 65536,
1953 .io_opt = 256 << SECTOR_SHIFT,
1954 .max_segments = USHRT_MAX,
1955 .max_segment_size = UINT_MAX,
1956 };
1957 struct nbd_device *nbd;
1958 struct gendisk *disk;
1959 int err = -ENOMEM;
1960
1961 nbd = kzalloc_obj(struct nbd_device);
1962 if (!nbd)
1963 goto out;
1964
1965 nbd->tag_set.ops = &nbd_mq_ops;
1966 nbd->tag_set.nr_hw_queues = nr_hw_queues;
1967 nbd->tag_set.queue_depth = 128;
1968 nbd->tag_set.numa_node = NUMA_NO_NODE;
1969 nbd->tag_set.cmd_size = sizeof(struct nbd_cmd);
1970 nbd->tag_set.flags = BLK_MQ_F_BLOCKING;
1971 nbd->tag_set.driver_data = nbd;
1972 INIT_WORK(&nbd->remove_work, nbd_dev_remove_work);
1973 nbd->backend = NULL;
1974
1975 err = blk_mq_alloc_tag_set(&nbd->tag_set);
1976 if (err)
1977 goto out_free_nbd;
1978
1979 mutex_lock(&nbd_index_mutex);
1980 if (index >= 0) {
1981 err = idr_alloc(&nbd_index_idr, nbd, index, index + 1,
1982 GFP_KERNEL);
1983 if (err == -ENOSPC)
1984 err = -EEXIST;
1985 } else {
1986 err = idr_alloc(&nbd_index_idr, nbd, 0,
1987 (MINORMASK >> part_shift) + 1, GFP_KERNEL);
1988 if (err >= 0)
1989 index = err;
1990 }
1991 nbd->index = index;
1992 mutex_unlock(&nbd_index_mutex);
1993 if (err < 0)
1994 goto out_free_tags;
1995
1996 disk = blk_mq_alloc_disk(&nbd->tag_set, &lim, NULL);
1997 if (IS_ERR(disk)) {
1998 err = PTR_ERR(disk);
1999 goto out_free_idr;
2000 }
2001 nbd->disk = disk;
2002
2003 nbd->recv_workq = alloc_workqueue("nbd%d-recv",
2004 WQ_MEM_RECLAIM | WQ_HIGHPRI |
2005 WQ_UNBOUND, 0, nbd->index);
2006 if (!nbd->recv_workq) {
2007 dev_err(disk_to_dev(nbd->disk), "Could not allocate knbd recv work queue.\n");
2008 err = -ENOMEM;
2009 goto out_err_disk;
2010 }
2011
2012 mutex_init(&nbd->config_lock);
2013 refcount_set(&nbd->config_refs, 0);
2014 /*
2015 * Start out with a zero references to keep other threads from using
2016 * this device until it is fully initialized.
2017 */
2018 refcount_set(&nbd->refs, 0);
2019 INIT_LIST_HEAD(&nbd->list);
2020 disk->major = NBD_MAJOR;
2021 disk->first_minor = index << part_shift;
2022 disk->minors = 1 << part_shift;
2023 disk->fops = &nbd_fops;
2024 disk->private_data = nbd;
2025 sprintf(disk->disk_name, "nbd%d", index);
2026 err = add_disk(disk);
2027 if (err)
2028 goto out_free_work;
2029
2030 /*
2031 * Now publish the device.
2032 */
2033 refcount_set(&nbd->refs, refs);
2034 nbd_total_devices++;
2035 return nbd;
2036
2037out_free_work:
2038 destroy_workqueue(nbd->recv_workq);
2039out_err_disk:
2040 put_disk(disk);
2041out_free_idr:
2042 mutex_lock(&nbd_index_mutex);
2043 idr_remove(&nbd_index_idr, index);
2044 mutex_unlock(&nbd_index_mutex);
2045out_free_tags:
2046 blk_mq_free_tag_set(&nbd->tag_set);
2047out_free_nbd:
2048 kfree(nbd);
2049out:
2050 return ERR_PTR(err);
2051}
2052
2053static struct nbd_device *nbd_find_get_unused(void)
2054{
2055 struct nbd_device *nbd;
2056 int id;
2057
2058 lockdep_assert_held(&nbd_index_mutex);
2059
2060 idr_for_each_entry(&nbd_index_idr, nbd, id) {
2061 if (refcount_read(&nbd->config_refs) ||
2062 test_bit(NBD_DESTROY_ON_DISCONNECT, &nbd->flags))
2063 continue;
2064 if (refcount_inc_not_zero(&nbd->refs))
2065 return nbd;
2066 }
2067
2068 return NULL;
2069}
2070
2071/* Netlink interface. */
2072static const struct nla_policy nbd_attr_policy[NBD_ATTR_MAX + 1] = {
2073 [NBD_ATTR_INDEX] = { .type = NLA_U32 },
2074 [NBD_ATTR_SIZE_BYTES] = { .type = NLA_U64 },
2075 [NBD_ATTR_BLOCK_SIZE_BYTES] = { .type = NLA_U64 },
2076 [NBD_ATTR_TIMEOUT] = { .type = NLA_U64 },
2077 [NBD_ATTR_SERVER_FLAGS] = { .type = NLA_U64 },
2078 [NBD_ATTR_CLIENT_FLAGS] = { .type = NLA_U64 },
2079 [NBD_ATTR_SOCKETS] = { .type = NLA_NESTED},
2080 [NBD_ATTR_DEAD_CONN_TIMEOUT] = { .type = NLA_U64 },
2081 [NBD_ATTR_DEVICE_LIST] = { .type = NLA_NESTED},
2082 [NBD_ATTR_BACKEND_IDENTIFIER] = { .type = NLA_STRING},
2083};
2084
2085static const struct nla_policy nbd_sock_policy[NBD_SOCK_MAX + 1] = {
2086 [NBD_SOCK_FD] = { .type = NLA_U32 },
2087};
2088
2089/* We don't use this right now since we don't parse the incoming list, but we
2090 * still want it here so userspace knows what to expect.
2091 */
2092static const struct nla_policy __attribute__((unused))
2093nbd_device_policy[NBD_DEVICE_ATTR_MAX + 1] = {
2094 [NBD_DEVICE_INDEX] = { .type = NLA_U32 },
2095 [NBD_DEVICE_CONNECTED] = { .type = NLA_U8 },
2096};
2097
2098static int nbd_genl_size_set(struct genl_info *info, struct nbd_device *nbd)
2099{
2100 struct nbd_config *config = nbd->config;
2101 u64 bsize = nbd_blksize(config);
2102 u64 bytes = config->bytesize;
2103
2104 if (info->attrs[NBD_ATTR_SIZE_BYTES])
2105 bytes = nla_get_u64(info->attrs[NBD_ATTR_SIZE_BYTES]);
2106
2107 if (info->attrs[NBD_ATTR_BLOCK_SIZE_BYTES])
2108 bsize = nla_get_u64(info->attrs[NBD_ATTR_BLOCK_SIZE_BYTES]);
2109
2110 if (bytes != config->bytesize || bsize != nbd_blksize(config))
2111 return nbd_set_size(nbd, bytes, bsize);
2112 return 0;
2113}
2114
2115/*
2116 * Walk the NBD_ATTR_SOCKETS nested list can call @cb for each socket fd.
2117 *
2118 * Return the number of fds walked, or a negative errno.
2119 */
2120static int nbd_genl_foreach_sock(struct genl_info *info,
2121 int (*cb)(struct nbd_device *nbd, unsigned long fd),
2122 struct nbd_device *nbd)
2123{
2124 struct nlattr *attr;
2125 int rem, count = 0;
2126
2127 if (!info->attrs[NBD_ATTR_SOCKETS])
2128 return 0;
2129
2130 nla_for_each_nested(attr, info->attrs[NBD_ATTR_SOCKETS], rem) {
2131 struct nlattr *socks[NBD_SOCK_MAX + 1];
2132 int ret;
2133
2134 if (nla_type(attr) != NBD_SOCK_ITEM) {
2135 pr_err("socks must be embedded in a SOCK_ITEM attr\n");
2136 return -EINVAL;
2137 }
2138
2139 if (nla_parse_nested_deprecated(socks, NBD_SOCK_MAX,
2140 attr,
2141 nbd_sock_policy,
2142 info->extack)) {
2143 pr_err("error processing sock list\n");
2144 return -EINVAL;
2145 }
2146
2147 if (!socks[NBD_SOCK_FD])
2148 continue;
2149
2150 count++;
2151 if (cb) {
2152 ret = cb(nbd, (int)nla_get_u32(socks[NBD_SOCK_FD]));
2153 if (ret > 0)
2154 return count;
2155 if (ret < 0)
2156 return ret;
2157 }
2158 }
2159 return count;
2160}
2161
2162static int nbd_genl_connect_sock_cb(struct nbd_device *nbd, unsigned long fd)
2163{
2164 return nbd_add_socket(nbd, fd, true);
2165}
2166
2167static int nbd_genl_connect(struct sk_buff *skb, struct genl_info *info)
2168{
2169 struct nbd_device *nbd;
2170 struct nbd_config *config;
2171 int index = -1;
2172 int ret;
2173 bool put_dev = false;
2174
2175 if (!netlink_capable(skb, CAP_SYS_ADMIN))
2176 return -EPERM;
2177
2178 if (info->attrs[NBD_ATTR_INDEX]) {
2179 index = nla_get_u32(info->attrs[NBD_ATTR_INDEX]);
2180
2181 /*
2182 * Too big first_minor can cause duplicate creation of
2183 * sysfs files/links, since index << part_shift might overflow, or
2184 * MKDEV() expect that the max bits of first_minor is 20.
2185 */
2186 if (index < 0 || index > MINORMASK >> part_shift) {
2187 pr_err("illegal input index %d\n", index);
2188 return -EINVAL;
2189 }
2190 }
2191 if (GENL_REQ_ATTR_CHECK(info, NBD_ATTR_SOCKETS)) {
2192 pr_err("must specify at least one socket\n");
2193 return -EINVAL;
2194 }
2195 if (GENL_REQ_ATTR_CHECK(info, NBD_ATTR_SIZE_BYTES)) {
2196 pr_err("must specify a size in bytes for the device\n");
2197 return -EINVAL;
2198 }
2199again:
2200 mutex_lock(&nbd_index_mutex);
2201 if (index == -1) {
2202 nbd = nbd_find_get_unused();
2203 } else {
2204 nbd = idr_find(&nbd_index_idr, index);
2205 if (nbd) {
2206 if ((test_bit(NBD_DESTROY_ON_DISCONNECT, &nbd->flags) &&
2207 test_bit(NBD_DISCONNECT_REQUESTED, &nbd->flags)) ||
2208 !refcount_inc_not_zero(&nbd->refs)) {
2209 mutex_unlock(&nbd_index_mutex);
2210 pr_err("device at index %d is going down\n",
2211 index);
2212 return -EINVAL;
2213 }
2214 }
2215 }
2216 mutex_unlock(&nbd_index_mutex);
2217
2218 if (!nbd) {
2219 ret = nbd_genl_foreach_sock(info, NULL, NULL);
2220 if (ret < 0)
2221 return ret;
2222
2223 nbd = nbd_dev_add(index, 2, ret > 0 ? ret : 1);
2224 if (IS_ERR(nbd)) {
2225 pr_err("failed to add new device\n");
2226 return PTR_ERR(nbd);
2227 }
2228 }
2229
2230 mutex_lock(&nbd->config_lock);
2231 if (refcount_read(&nbd->config_refs)) {
2232 mutex_unlock(&nbd->config_lock);
2233 nbd_put(nbd);
2234 if (index == -1)
2235 goto again;
2236 pr_err("nbd%d already in use\n", index);
2237 return -EBUSY;
2238 }
2239
2240 ret = nbd_alloc_and_init_config(nbd);
2241 if (ret) {
2242 mutex_unlock(&nbd->config_lock);
2243 nbd_put(nbd);
2244 pr_err("couldn't allocate config\n");
2245 return ret;
2246 }
2247
2248 config = nbd->config;
2249 set_bit(NBD_RT_BOUND, &config->runtime_flags);
2250 ret = nbd_genl_size_set(info, nbd);
2251 if (ret)
2252 goto out;
2253
2254 if (info->attrs[NBD_ATTR_TIMEOUT])
2255 nbd_set_cmd_timeout(nbd,
2256 nla_get_u64(info->attrs[NBD_ATTR_TIMEOUT]));
2257 if (info->attrs[NBD_ATTR_DEAD_CONN_TIMEOUT]) {
2258 config->dead_conn_timeout =
2259 nla_get_u64(info->attrs[NBD_ATTR_DEAD_CONN_TIMEOUT]);
2260 config->dead_conn_timeout *= HZ;
2261 }
2262 if (info->attrs[NBD_ATTR_SERVER_FLAGS])
2263 config->flags =
2264 nla_get_u64(info->attrs[NBD_ATTR_SERVER_FLAGS]);
2265 if (info->attrs[NBD_ATTR_CLIENT_FLAGS]) {
2266 u64 flags = nla_get_u64(info->attrs[NBD_ATTR_CLIENT_FLAGS]);
2267 if (flags & NBD_CFLAG_DESTROY_ON_DISCONNECT) {
2268 /*
2269 * We have 1 ref to keep the device around, and then 1
2270 * ref for our current operation here, which will be
2271 * inherited by the config. If we already have
2272 * DESTROY_ON_DISCONNECT set then we know we don't have
2273 * that extra ref already held so we don't need the
2274 * put_dev.
2275 */
2276 if (!test_and_set_bit(NBD_DESTROY_ON_DISCONNECT,
2277 &nbd->flags))
2278 put_dev = true;
2279 } else {
2280 if (test_and_clear_bit(NBD_DESTROY_ON_DISCONNECT,
2281 &nbd->flags))
2282 refcount_inc(&nbd->refs);
2283 }
2284 if (flags & NBD_CFLAG_DISCONNECT_ON_CLOSE) {
2285 set_bit(NBD_RT_DISCONNECT_ON_CLOSE,
2286 &config->runtime_flags);
2287 }
2288 }
2289
2290 ret = nbd_genl_foreach_sock(info, nbd_genl_connect_sock_cb, nbd);
2291 if (ret < 0)
2292 goto out;
2293
2294 if (info->attrs[NBD_ATTR_BACKEND_IDENTIFIER]) {
2295 nbd->backend = nla_strdup(info->attrs[NBD_ATTR_BACKEND_IDENTIFIER],
2296 GFP_KERNEL);
2297 if (!nbd->backend) {
2298 ret = -ENOMEM;
2299 goto out;
2300 }
2301 }
2302 ret = device_create_file(disk_to_dev(nbd->disk), &backend_attr);
2303 if (ret) {
2304 dev_err(disk_to_dev(nbd->disk),
2305 "device_create_file failed for backend!\n");
2306 goto out;
2307 }
2308 set_bit(NBD_RT_HAS_BACKEND_FILE, &config->runtime_flags);
2309
2310 ret = nbd_start_device(nbd);
2311out:
2312 if (!ret) {
2313 set_bit(NBD_RT_HAS_CONFIG_REF, &config->runtime_flags);
2314 refcount_inc(&nbd->config_refs);
2315 nbd_connect_reply(info, nbd->index);
2316 }
2317 mutex_unlock(&nbd->config_lock);
2318
2319 nbd_config_put(nbd);
2320 if (put_dev)
2321 nbd_put(nbd);
2322 return ret;
2323}
2324
2325static void nbd_disconnect_and_put(struct nbd_device *nbd)
2326{
2327 mutex_lock(&nbd->config_lock);
2328 nbd_disconnect(nbd);
2329 sock_shutdown(nbd);
2330 wake_up(&nbd->config->conn_wait);
2331 /*
2332 * Make sure recv thread has finished, we can safely call nbd_clear_que()
2333 * to cancel the inflight I/Os.
2334 */
2335 flush_workqueue(nbd->recv_workq);
2336 nbd_clear_que(nbd);
2337 nbd->task_setup = NULL;
2338 clear_bit(NBD_RT_BOUND, &nbd->config->runtime_flags);
2339 mutex_unlock(&nbd->config_lock);
2340
2341 if (test_and_clear_bit(NBD_RT_HAS_CONFIG_REF,
2342 &nbd->config->runtime_flags))
2343 nbd_config_put(nbd);
2344}
2345
2346static int nbd_genl_disconnect(struct sk_buff *skb, struct genl_info *info)
2347{
2348 struct nbd_device *nbd;
2349 int index;
2350
2351 if (!netlink_capable(skb, CAP_SYS_ADMIN))
2352 return -EPERM;
2353
2354 if (GENL_REQ_ATTR_CHECK(info, NBD_ATTR_INDEX)) {
2355 pr_err("must specify an index to disconnect\n");
2356 return -EINVAL;
2357 }
2358 index = nla_get_u32(info->attrs[NBD_ATTR_INDEX]);
2359 mutex_lock(&nbd_index_mutex);
2360 nbd = idr_find(&nbd_index_idr, index);
2361 if (!nbd) {
2362 mutex_unlock(&nbd_index_mutex);
2363 pr_err("couldn't find device at index %d\n", index);
2364 return -EINVAL;
2365 }
2366 if (!refcount_inc_not_zero(&nbd->refs)) {
2367 mutex_unlock(&nbd_index_mutex);
2368 pr_err("device at index %d is going down\n", index);
2369 return -EINVAL;
2370 }
2371 mutex_unlock(&nbd_index_mutex);
2372 if (!refcount_inc_not_zero(&nbd->config_refs))
2373 goto put_nbd;
2374 nbd_disconnect_and_put(nbd);
2375 nbd_config_put(nbd);
2376put_nbd:
2377 nbd_put(nbd);
2378 return 0;
2379}
2380
2381static int nbd_genl_reconfigure(struct sk_buff *skb, struct genl_info *info)
2382{
2383 struct nbd_device *nbd = NULL;
2384 struct nbd_config *config;
2385 int index;
2386 int ret = 0;
2387 bool put_dev = false;
2388
2389 if (!netlink_capable(skb, CAP_SYS_ADMIN))
2390 return -EPERM;
2391
2392 if (GENL_REQ_ATTR_CHECK(info, NBD_ATTR_INDEX)) {
2393 pr_err("must specify a device to reconfigure\n");
2394 return -EINVAL;
2395 }
2396 index = nla_get_u32(info->attrs[NBD_ATTR_INDEX]);
2397 mutex_lock(&nbd_index_mutex);
2398 nbd = idr_find(&nbd_index_idr, index);
2399 if (!nbd) {
2400 mutex_unlock(&nbd_index_mutex);
2401 pr_err("couldn't find a device at index %d\n", index);
2402 return -EINVAL;
2403 }
2404 if (nbd->backend) {
2405 if (info->attrs[NBD_ATTR_BACKEND_IDENTIFIER]) {
2406 if (nla_strcmp(info->attrs[NBD_ATTR_BACKEND_IDENTIFIER],
2407 nbd->backend)) {
2408 mutex_unlock(&nbd_index_mutex);
2409 dev_err(nbd_to_dev(nbd),
2410 "backend image doesn't match with %s\n",
2411 nbd->backend);
2412 return -EINVAL;
2413 }
2414 } else {
2415 mutex_unlock(&nbd_index_mutex);
2416 dev_err(nbd_to_dev(nbd), "must specify backend\n");
2417 return -EINVAL;
2418 }
2419 }
2420 if (!refcount_inc_not_zero(&nbd->refs)) {
2421 mutex_unlock(&nbd_index_mutex);
2422 pr_err("device at index %d is going down\n", index);
2423 return -EINVAL;
2424 }
2425 mutex_unlock(&nbd_index_mutex);
2426
2427 config = nbd_get_config_unlocked(nbd);
2428 if (!config) {
2429 dev_err(nbd_to_dev(nbd),
2430 "not configured, cannot reconfigure\n");
2431 nbd_put(nbd);
2432 return -EINVAL;
2433 }
2434
2435 mutex_lock(&nbd->config_lock);
2436 if (!test_bit(NBD_RT_BOUND, &config->runtime_flags) ||
2437 !nbd->pid) {
2438 dev_err(nbd_to_dev(nbd),
2439 "not configured, cannot reconfigure\n");
2440 ret = -EINVAL;
2441 goto out;
2442 }
2443
2444 ret = nbd_genl_size_set(info, nbd);
2445 if (ret)
2446 goto out;
2447
2448 if (info->attrs[NBD_ATTR_TIMEOUT])
2449 nbd_set_cmd_timeout(nbd,
2450 nla_get_u64(info->attrs[NBD_ATTR_TIMEOUT]));
2451 if (info->attrs[NBD_ATTR_DEAD_CONN_TIMEOUT]) {
2452 config->dead_conn_timeout =
2453 nla_get_u64(info->attrs[NBD_ATTR_DEAD_CONN_TIMEOUT]);
2454 config->dead_conn_timeout *= HZ;
2455 }
2456 if (info->attrs[NBD_ATTR_CLIENT_FLAGS]) {
2457 u64 flags = nla_get_u64(info->attrs[NBD_ATTR_CLIENT_FLAGS]);
2458 if (flags & NBD_CFLAG_DESTROY_ON_DISCONNECT) {
2459 if (!test_and_set_bit(NBD_DESTROY_ON_DISCONNECT,
2460 &nbd->flags))
2461 put_dev = true;
2462 } else {
2463 if (test_and_clear_bit(NBD_DESTROY_ON_DISCONNECT,
2464 &nbd->flags))
2465 refcount_inc(&nbd->refs);
2466 }
2467
2468 if (flags & NBD_CFLAG_DISCONNECT_ON_CLOSE) {
2469 set_bit(NBD_RT_DISCONNECT_ON_CLOSE,
2470 &config->runtime_flags);
2471 } else {
2472 clear_bit(NBD_RT_DISCONNECT_ON_CLOSE,
2473 &config->runtime_flags);
2474 }
2475 }
2476
2477 ret = nbd_genl_foreach_sock(info, nbd_genl_reconnect_sock_cb, nbd);
2478 /* foreach_sock returns a positive count on success; doit must return 0 */
2479 if (ret >= 0)
2480 ret = 0;
2481out:
2482 mutex_unlock(&nbd->config_lock);
2483 nbd_config_put(nbd);
2484 nbd_put(nbd);
2485 if (put_dev)
2486 nbd_put(nbd);
2487 return ret;
2488}
2489
2490static const struct genl_small_ops nbd_connect_genl_ops[] = {
2491 {
2492 .cmd = NBD_CMD_CONNECT,
2493 .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2494 .doit = nbd_genl_connect,
2495 },
2496 {
2497 .cmd = NBD_CMD_DISCONNECT,
2498 .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2499 .doit = nbd_genl_disconnect,
2500 },
2501 {
2502 .cmd = NBD_CMD_RECONFIGURE,
2503 .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2504 .doit = nbd_genl_reconfigure,
2505 },
2506 {
2507 .cmd = NBD_CMD_STATUS,
2508 .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2509 .doit = nbd_genl_status,
2510 },
2511};
2512
2513static const struct genl_multicast_group nbd_mcast_grps[] = {
2514 { .name = NBD_GENL_MCAST_GROUP_NAME, },
2515};
2516
2517static struct genl_family nbd_genl_family __ro_after_init = {
2518 .hdrsize = 0,
2519 .name = NBD_GENL_FAMILY_NAME,
2520 .version = NBD_GENL_VERSION,
2521 .module = THIS_MODULE,
2522 .small_ops = nbd_connect_genl_ops,
2523 .n_small_ops = ARRAY_SIZE(nbd_connect_genl_ops),
2524 .resv_start_op = NBD_CMD_STATUS + 1,
2525 .maxattr = NBD_ATTR_MAX,
2526 .netnsok = 1,
2527 .policy = nbd_attr_policy,
2528 .mcgrps = nbd_mcast_grps,
2529 .n_mcgrps = ARRAY_SIZE(nbd_mcast_grps),
2530};
2531MODULE_ALIAS_GENL_FAMILY(NBD_GENL_FAMILY_NAME);
2532
2533static int populate_nbd_status(struct nbd_device *nbd, struct sk_buff *reply)
2534{
2535 struct nlattr *dev_opt;
2536 u8 connected = 0;
2537 int ret;
2538
2539 /* This is a little racey, but for status it's ok. The
2540 * reason we don't take a ref here is because we can't
2541 * take a ref in the index == -1 case as we would need
2542 * to put under the nbd_index_mutex, which could
2543 * deadlock if we are configured to remove ourselves
2544 * once we're disconnected.
2545 */
2546 if (refcount_read(&nbd->config_refs))
2547 connected = 1;
2548 dev_opt = nla_nest_start_noflag(reply, NBD_DEVICE_ITEM);
2549 if (!dev_opt)
2550 return -EMSGSIZE;
2551 ret = nla_put_u32(reply, NBD_DEVICE_INDEX, nbd->index);
2552 if (ret)
2553 return -EMSGSIZE;
2554 ret = nla_put_u8(reply, NBD_DEVICE_CONNECTED,
2555 connected);
2556 if (ret)
2557 return -EMSGSIZE;
2558 nla_nest_end(reply, dev_opt);
2559 return 0;
2560}
2561
2562static int status_cb(int id, void *ptr, void *data)
2563{
2564 struct nbd_device *nbd = ptr;
2565 return populate_nbd_status(nbd, (struct sk_buff *)data);
2566}
2567
2568static int nbd_genl_status(struct sk_buff *skb, struct genl_info *info)
2569{
2570 struct nlattr *dev_list;
2571 struct sk_buff *reply;
2572 void *reply_head;
2573 size_t msg_size;
2574 int index = -1;
2575 int ret = -ENOMEM;
2576
2577 if (info->attrs[NBD_ATTR_INDEX])
2578 index = nla_get_u32(info->attrs[NBD_ATTR_INDEX]);
2579
2580 mutex_lock(&nbd_index_mutex);
2581
2582 msg_size = nla_total_size(nla_attr_size(sizeof(u32)) +
2583 nla_attr_size(sizeof(u8)));
2584 msg_size *= (index == -1) ? nbd_total_devices : 1;
2585
2586 reply = genlmsg_new(msg_size, GFP_KERNEL);
2587 if (!reply)
2588 goto out;
2589 reply_head = genlmsg_put_reply(reply, info, &nbd_genl_family, 0,
2590 NBD_CMD_STATUS);
2591 if (!reply_head) {
2592 nlmsg_free(reply);
2593 goto out;
2594 }
2595
2596 dev_list = nla_nest_start_noflag(reply, NBD_ATTR_DEVICE_LIST);
2597 if (!dev_list) {
2598 nlmsg_free(reply);
2599 ret = -EMSGSIZE;
2600 goto out;
2601 }
2602
2603 if (index == -1) {
2604 ret = idr_for_each(&nbd_index_idr, &status_cb, reply);
2605 if (ret) {
2606 nlmsg_free(reply);
2607 goto out;
2608 }
2609 } else {
2610 struct nbd_device *nbd;
2611 nbd = idr_find(&nbd_index_idr, index);
2612 if (nbd) {
2613 ret = populate_nbd_status(nbd, reply);
2614 if (ret) {
2615 nlmsg_free(reply);
2616 goto out;
2617 }
2618 }
2619 }
2620 nla_nest_end(reply, dev_list);
2621 genlmsg_end(reply, reply_head);
2622 ret = genlmsg_reply(reply, info);
2623out:
2624 mutex_unlock(&nbd_index_mutex);
2625 return ret;
2626}
2627
2628static void nbd_connect_reply(struct genl_info *info, int index)
2629{
2630 struct sk_buff *skb;
2631 void *msg_head;
2632 int ret;
2633
2634 skb = genlmsg_new(nla_total_size(sizeof(u32)), GFP_KERNEL);
2635 if (!skb)
2636 return;
2637 msg_head = genlmsg_put_reply(skb, info, &nbd_genl_family, 0,
2638 NBD_CMD_CONNECT);
2639 if (!msg_head) {
2640 nlmsg_free(skb);
2641 return;
2642 }
2643 ret = nla_put_u32(skb, NBD_ATTR_INDEX, index);
2644 if (ret) {
2645 nlmsg_free(skb);
2646 return;
2647 }
2648 genlmsg_end(skb, msg_head);
2649 genlmsg_reply(skb, info);
2650}
2651
2652static void nbd_mcast_index(int index)
2653{
2654 struct sk_buff *skb;
2655 void *msg_head;
2656 int ret;
2657
2658 skb = genlmsg_new(nla_total_size(sizeof(u32)), GFP_KERNEL);
2659 if (!skb)
2660 return;
2661 msg_head = genlmsg_put(skb, 0, 0, &nbd_genl_family, 0,
2662 NBD_CMD_LINK_DEAD);
2663 if (!msg_head) {
2664 nlmsg_free(skb);
2665 return;
2666 }
2667 ret = nla_put_u32(skb, NBD_ATTR_INDEX, index);
2668 if (ret) {
2669 nlmsg_free(skb);
2670 return;
2671 }
2672 genlmsg_end(skb, msg_head);
2673 genlmsg_multicast(&nbd_genl_family, skb, 0, 0, GFP_KERNEL);
2674}
2675
2676static void nbd_dead_link_work(struct work_struct *work)
2677{
2678 struct link_dead_args *args = container_of(work, struct link_dead_args,
2679 work);
2680 nbd_mcast_index(args->index);
2681 kfree(args);
2682}
2683
2684static int __init nbd_init(void)
2685{
2686 int i;
2687
2688 BUILD_BUG_ON(sizeof(struct nbd_request) != 28);
2689
2690 if (max_part < 0) {
2691 pr_err("max_part must be >= 0\n");
2692 return -EINVAL;
2693 }
2694
2695 part_shift = 0;
2696 if (max_part > 0) {
2697 part_shift = fls(max_part);
2698
2699 /*
2700 * Adjust max_part according to part_shift as it is exported
2701 * to user space so that user can know the max number of
2702 * partition kernel should be able to manage.
2703 *
2704 * Note that -1 is required because partition 0 is reserved
2705 * for the whole disk.
2706 */
2707 max_part = (1UL << part_shift) - 1;
2708 }
2709
2710 if ((1UL << part_shift) > DISK_MAX_PARTS)
2711 return -EINVAL;
2712
2713 if (nbds_max > 1UL << (MINORBITS - part_shift))
2714 return -EINVAL;
2715
2716 /* An excessively large value will be adjusted in blk_mq_alloc_tag_set */
2717 if (pre_defined_connections < 1) {
2718 pr_err("pre_defined_connections must be >= 1\n");
2719 return -EINVAL;
2720 }
2721
2722 if (register_blkdev(NBD_MAJOR, "nbd"))
2723 return -EIO;
2724
2725 nbd_del_wq = alloc_workqueue("nbd-del", WQ_UNBOUND, 0);
2726 if (!nbd_del_wq) {
2727 unregister_blkdev(NBD_MAJOR, "nbd");
2728 return -ENOMEM;
2729 }
2730
2731 if (genl_register_family(&nbd_genl_family)) {
2732 destroy_workqueue(nbd_del_wq);
2733 unregister_blkdev(NBD_MAJOR, "nbd");
2734 return -EINVAL;
2735 }
2736 nbd_dbg_init();
2737
2738 /*
2739 * Set to the intended connection count so nbd_start_device() can skip
2740 * the queue-freezing blk_mq_update_nr_hw_queues() call.
2741 */
2742 for (i = 0; i < nbds_max; i++)
2743 nbd_dev_add(i, 1, pre_defined_connections);
2744 return 0;
2745}
2746
2747static int nbd_exit_cb(int id, void *ptr, void *data)
2748{
2749 struct list_head *list = (struct list_head *)data;
2750 struct nbd_device *nbd = ptr;
2751
2752 /* Skip nbd that is being removed asynchronously */
2753 if (refcount_read(&nbd->refs))
2754 list_add_tail(&nbd->list, list);
2755
2756 return 0;
2757}
2758
2759static void __exit nbd_cleanup(void)
2760{
2761 struct nbd_device *nbd;
2762 LIST_HEAD(del_list);
2763
2764 /*
2765 * Unregister netlink interface prior to waiting
2766 * for the completion of netlink commands.
2767 */
2768 genl_unregister_family(&nbd_genl_family);
2769
2770 nbd_dbg_close();
2771
2772 mutex_lock(&nbd_index_mutex);
2773 idr_for_each(&nbd_index_idr, &nbd_exit_cb, &del_list);
2774 mutex_unlock(&nbd_index_mutex);
2775
2776 while (!list_empty(&del_list)) {
2777 nbd = list_first_entry(&del_list, struct nbd_device, list);
2778 list_del_init(&nbd->list);
2779 if (refcount_read(&nbd->config_refs))
2780 pr_err("possibly leaking nbd_config (ref %d)\n",
2781 refcount_read(&nbd->config_refs));
2782 if (refcount_read(&nbd->refs) != 1)
2783 pr_err("possibly leaking a device\n");
2784 nbd_put(nbd);
2785 }
2786
2787 /* Also wait for nbd_dev_remove_work() completes */
2788 destroy_workqueue(nbd_del_wq);
2789
2790 idr_destroy(&nbd_index_idr);
2791 unregister_blkdev(NBD_MAJOR, "nbd");
2792}
2793
2794module_init(nbd_init);
2795module_exit(nbd_cleanup);
2796
2797MODULE_DESCRIPTION("Network Block Device");
2798MODULE_LICENSE("GPL");
2799
2800module_param(nbds_max, int, 0444);
2801MODULE_PARM_DESC(nbds_max, "number of network block devices to initialize (default: 16)");
2802module_param(max_part, int, 0444);
2803MODULE_PARM_DESC(max_part, "number of partitions per device (default: 16)");
2804module_param(pre_defined_connections, int, 0444);
2805MODULE_PARM_DESC(pre_defined_connections,
2806"number of connections for devices pre-created at module load (default: 1)");