4 This file is part of DRBD by Philipp Reisner and Lars Ellenberg.
6 Copyright (C) 2001-2008, LINBIT Information Technologies GmbH.
7 Copyright (C) 1999-2008, Philipp Reisner <philipp.reisner@linbit.com>.
8 Copyright (C) 2002-2008, Lars Ellenberg <lars.ellenberg@linbit.com>.
10 Thanks to Carter Burden, Bart Grantham and Gennadiy Nerubayev
11 from Logicworks, Inc. for making SDP replication support possible.
13 drbd is free software; you can redistribute it and/or modify
14 it under the terms of the GNU General Public License as published by
15 the Free Software Foundation; either version 2, or (at your option)
18 drbd is distributed in the hope that it will be useful,
19 but WITHOUT ANY WARRANTY; without even the implied warranty of
20 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
21 GNU General Public License for more details.
23 You should have received a copy of the GNU General Public License
24 along with drbd; see the file COPYING. If not, write to
25 the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
29 #include <linux/module.h>
30 #include <linux/drbd.h>
31 #include <asm/uaccess.h>
32 #include <asm/types.h>
34 #include <linux/ctype.h>
35 #include <linux/mutex.h>
37 #include <linux/file.h>
38 #include <linux/proc_fs.h>
39 #include <linux/init.h>
41 #include <linux/memcontrol.h>
42 #include <linux/mm_inline.h>
43 #include <linux/slab.h>
44 #include <linux/random.h>
45 #include <linux/reboot.h>
46 #include <linux/notifier.h>
47 #include <linux/kthread.h>
49 #define __KERNEL_SYSCALLS__
50 #include <linux/unistd.h>
51 #include <linux/vmalloc.h>
53 #include <linux/drbd_limits.h>
55 #include "drbd_req.h" /* only for _req_mod in tl_release and tl_clear */
59 static DEFINE_MUTEX(drbd_main_mutex);
60 int drbdd_init(struct drbd_thread *);
61 int drbd_worker(struct drbd_thread *);
62 int drbd_asender(struct drbd_thread *);
65 static int drbd_open(struct block_device *bdev, fmode_t mode);
66 static int drbd_release(struct gendisk *gd, fmode_t mode);
67 static int w_md_sync(struct drbd_work *w, int unused);
68 static void md_sync_timer_fn(unsigned long data);
69 static int w_bitmap_io(struct drbd_work *w, int unused);
70 static int w_go_diskless(struct drbd_work *w, int unused);
72 MODULE_AUTHOR("Philipp Reisner <phil@linbit.com>, "
73 "Lars Ellenberg <lars@linbit.com>");
74 MODULE_DESCRIPTION("drbd - Distributed Replicated Block Device v" REL_VERSION);
75 MODULE_VERSION(REL_VERSION);
76 MODULE_LICENSE("GPL");
77 MODULE_PARM_DESC(minor_count, "Approximate number of drbd devices ("
78 __stringify(DRBD_MINOR_COUNT_MIN) "-" __stringify(DRBD_MINOR_COUNT_MAX) ")");
79 MODULE_ALIAS_BLOCKDEV_MAJOR(DRBD_MAJOR);
81 #include <linux/moduleparam.h>
82 /* allow_open_on_secondary */
83 MODULE_PARM_DESC(allow_oos, "DONT USE!");
84 /* thanks to these macros, if compiled into the kernel (not-module),
85 * this becomes the boot parameter drbd.minor_count */
86 module_param(minor_count, uint, 0444);
87 module_param(disable_sendpage, bool, 0644);
88 module_param(allow_oos, bool, 0);
89 module_param(proc_details, int, 0644);
91 #ifdef CONFIG_DRBD_FAULT_INJECTION
94 static int fault_count;
96 /* bitmap of enabled faults */
97 module_param(enable_faults, int, 0664);
98 /* fault rate % value - applies to all enabled faults */
99 module_param(fault_rate, int, 0664);
100 /* count of faults inserted */
101 module_param(fault_count, int, 0664);
102 /* bitmap of devices to insert faults on */
103 module_param(fault_devs, int, 0644);
106 /* module parameter, defined */
107 unsigned int minor_count = DRBD_MINOR_COUNT_DEF;
108 int disable_sendpage;
110 int proc_details; /* Detail level in proc drbd*/
112 /* Module parameter for setting the user mode helper program
113 * to run. Default is /sbin/drbdadm */
114 char usermode_helper[80] = "/sbin/drbdadm";
116 module_param_string(usermode_helper, usermode_helper, sizeof(usermode_helper), 0644);
118 /* in 2.6.x, our device mapping and config info contains our virtual gendisks
119 * as member "struct gendisk *vdisk;"
122 struct list_head drbd_tconns; /* list of struct drbd_tconn */
123 DEFINE_MUTEX(drbd_cfg_mutex);
125 struct kmem_cache *drbd_request_cache;
126 struct kmem_cache *drbd_ee_cache; /* peer requests */
127 struct kmem_cache *drbd_bm_ext_cache; /* bitmap extents */
128 struct kmem_cache *drbd_al_ext_cache; /* activity log extents */
129 mempool_t *drbd_request_mempool;
130 mempool_t *drbd_ee_mempool;
131 mempool_t *drbd_md_io_page_pool;
132 struct bio_set *drbd_md_io_bio_set;
134 /* I do not use a standard mempool, because:
135 1) I want to hand out the pre-allocated objects first.
136 2) I want to be able to interrupt sleeping allocation with a signal.
137 Note: This is a single linked list, the next pointer is the private
138 member of struct page.
140 struct page *drbd_pp_pool;
141 spinlock_t drbd_pp_lock;
143 wait_queue_head_t drbd_pp_wait;
145 DEFINE_RATELIMIT_STATE(drbd_ratelimit_state, 5 * HZ, 5);
147 static const struct block_device_operations drbd_ops = {
148 .owner = THIS_MODULE,
150 .release = drbd_release,
153 static void bio_destructor_drbd(struct bio *bio)
155 bio_free(bio, drbd_md_io_bio_set);
158 struct bio *bio_alloc_drbd(gfp_t gfp_mask)
162 if (!drbd_md_io_bio_set)
163 return bio_alloc(gfp_mask, 1);
165 bio = bio_alloc_bioset(gfp_mask, 1, drbd_md_io_bio_set);
168 bio->bi_destructor = bio_destructor_drbd;
173 /* When checking with sparse, and this is an inline function, sparse will
174 give tons of false positives. When this is a real functions sparse works.
176 int _get_ldev_if_state(struct drbd_conf *mdev, enum drbd_disk_state mins)
180 atomic_inc(&mdev->local_cnt);
181 io_allowed = (mdev->state.disk >= mins);
183 if (atomic_dec_and_test(&mdev->local_cnt))
184 wake_up(&mdev->misc_wait);
192 * DOC: The transfer log
194 * The transfer log is a single linked list of &struct drbd_tl_epoch objects.
195 * mdev->tconn->newest_tle points to the head, mdev->tconn->oldest_tle points to the tail
196 * of the list. There is always at least one &struct drbd_tl_epoch object.
198 * Each &struct drbd_tl_epoch has a circular double linked list of requests
201 static int tl_init(struct drbd_tconn *tconn)
203 struct drbd_tl_epoch *b;
205 /* during device minor initialization, we may well use GFP_KERNEL */
206 b = kmalloc(sizeof(struct drbd_tl_epoch), GFP_KERNEL);
209 INIT_LIST_HEAD(&b->requests);
210 INIT_LIST_HEAD(&b->w.list);
214 b->w.cb = NULL; /* if this is != NULL, we need to dec_ap_pending in tl_clear */
216 tconn->oldest_tle = b;
217 tconn->newest_tle = b;
218 INIT_LIST_HEAD(&tconn->out_of_sequence_requests);
223 static void tl_cleanup(struct drbd_tconn *tconn)
225 if (tconn->oldest_tle != tconn->newest_tle)
226 conn_err(tconn, "ASSERT FAILED: oldest_tle == newest_tle\n");
227 if (!list_empty(&tconn->out_of_sequence_requests))
228 conn_err(tconn, "ASSERT FAILED: list_empty(out_of_sequence_requests)\n");
229 kfree(tconn->oldest_tle);
230 tconn->oldest_tle = NULL;
231 kfree(tconn->unused_spare_tle);
232 tconn->unused_spare_tle = NULL;
236 * _tl_add_barrier() - Adds a barrier to the transfer log
237 * @mdev: DRBD device.
238 * @new: Barrier to be added before the current head of the TL.
240 * The caller must hold the req_lock.
242 void _tl_add_barrier(struct drbd_tconn *tconn, struct drbd_tl_epoch *new)
244 struct drbd_tl_epoch *newest_before;
246 INIT_LIST_HEAD(&new->requests);
247 INIT_LIST_HEAD(&new->w.list);
248 new->w.cb = NULL; /* if this is != NULL, we need to dec_ap_pending in tl_clear */
252 newest_before = tconn->newest_tle;
253 /* never send a barrier number == 0, because that is special-cased
254 * when using TCQ for our write ordering code */
255 new->br_number = (newest_before->br_number+1) ?: 1;
256 if (tconn->newest_tle != new) {
257 tconn->newest_tle->next = new;
258 tconn->newest_tle = new;
263 * tl_release() - Free or recycle the oldest &struct drbd_tl_epoch object of the TL
264 * @mdev: DRBD device.
265 * @barrier_nr: Expected identifier of the DRBD write barrier packet.
266 * @set_size: Expected number of requests before that barrier.
268 * In case the passed barrier_nr or set_size does not match the oldest
269 * &struct drbd_tl_epoch objects this function will cause a termination
272 void tl_release(struct drbd_tconn *tconn, unsigned int barrier_nr,
273 unsigned int set_size)
275 struct drbd_conf *mdev;
276 struct drbd_tl_epoch *b, *nob; /* next old barrier */
277 struct list_head *le, *tle;
278 struct drbd_request *r;
280 spin_lock_irq(&tconn->req_lock);
282 b = tconn->oldest_tle;
284 /* first some paranoia code */
286 conn_err(tconn, "BAD! BarrierAck #%u received, but no epoch in tl!?\n",
290 if (b->br_number != barrier_nr) {
291 conn_err(tconn, "BAD! BarrierAck #%u received, expected #%u!\n",
292 barrier_nr, b->br_number);
295 if (b->n_writes != set_size) {
296 conn_err(tconn, "BAD! BarrierAck #%u received with n_writes=%u, expected n_writes=%u!\n",
297 barrier_nr, set_size, b->n_writes);
301 /* Clean up list of requests processed during current epoch */
302 list_for_each_safe(le, tle, &b->requests) {
303 r = list_entry(le, struct drbd_request, tl_requests);
304 _req_mod(r, BARRIER_ACKED);
306 /* There could be requests on the list waiting for completion
307 of the write to the local disk. To avoid corruptions of
308 slab's data structures we have to remove the lists head.
310 Also there could have been a barrier ack out of sequence, overtaking
311 the write acks - which would be a bug and violating write ordering.
312 To not deadlock in case we lose connection while such requests are
313 still pending, we need some way to find them for the
314 _req_mode(CONNECTION_LOST_WHILE_PENDING).
316 These have been list_move'd to the out_of_sequence_requests list in
317 _req_mod(, BARRIER_ACKED) above.
319 list_del_init(&b->requests);
323 if (test_and_clear_bit(CREATE_BARRIER, &mdev->flags)) {
324 _tl_add_barrier(tconn, b);
326 tconn->oldest_tle = nob;
327 /* if nob == NULL b was the only barrier, and becomes the new
328 barrier. Therefore tconn->oldest_tle points already to b */
330 D_ASSERT(nob != NULL);
331 tconn->oldest_tle = nob;
335 spin_unlock_irq(&tconn->req_lock);
336 dec_ap_pending(mdev);
341 spin_unlock_irq(&tconn->req_lock);
342 conn_request_state(tconn, NS(conn, C_PROTOCOL_ERROR), CS_HARD);
347 * _tl_restart() - Walks the transfer log, and applies an action to all requests
348 * @mdev: DRBD device.
349 * @what: The action/event to perform with all request objects
351 * @what might be one of CONNECTION_LOST_WHILE_PENDING, RESEND, FAIL_FROZEN_DISK_IO,
352 * RESTART_FROZEN_DISK_IO.
354 void _tl_restart(struct drbd_tconn *tconn, enum drbd_req_event what)
356 struct drbd_tl_epoch *b, *tmp, **pn;
357 struct list_head *le, *tle, carry_reads;
358 struct drbd_request *req;
359 int rv, n_writes, n_reads;
361 b = tconn->oldest_tle;
362 pn = &tconn->oldest_tle;
366 INIT_LIST_HEAD(&carry_reads);
367 list_for_each_safe(le, tle, &b->requests) {
368 req = list_entry(le, struct drbd_request, tl_requests);
369 rv = _req_mod(req, what);
371 n_writes += (rv & MR_WRITE) >> MR_WRITE_SHIFT;
372 n_reads += (rv & MR_READ) >> MR_READ_SHIFT;
377 if (what == RESEND) {
378 b->n_writes = n_writes;
379 if (b->w.cb == NULL) {
380 b->w.cb = w_send_barrier;
381 inc_ap_pending(b->w.mdev);
382 set_bit(CREATE_BARRIER, &b->w.mdev->flags);
385 drbd_queue_work(&tconn->data.work, &b->w);
390 list_add(&carry_reads, &b->requests);
391 /* there could still be requests on that ring list,
392 * in case local io is still pending */
393 list_del(&b->requests);
395 /* dec_ap_pending corresponding to queue_barrier.
396 * the newest barrier may not have been queued yet,
397 * in which case w.cb is still NULL. */
399 dec_ap_pending(b->w.mdev);
401 if (b == tconn->newest_tle) {
402 /* recycle, but reinit! */
404 conn_err(tconn, "ASSERT FAILED tmp == NULL");
405 INIT_LIST_HEAD(&b->requests);
406 list_splice(&carry_reads, &b->requests);
407 INIT_LIST_HEAD(&b->w.list);
409 b->br_number = net_random();
419 list_splice(&carry_reads, &b->requests);
425 * tl_clear() - Clears all requests and &struct drbd_tl_epoch objects out of the TL
426 * @mdev: DRBD device.
428 * This is called after the connection to the peer was lost. The storage covered
429 * by the requests on the transfer gets marked as our of sync. Called from the
430 * receiver thread and the worker thread.
432 void tl_clear(struct drbd_tconn *tconn)
434 struct drbd_conf *mdev;
435 struct list_head *le, *tle;
436 struct drbd_request *r;
439 spin_lock_irq(&tconn->req_lock);
441 _tl_restart(tconn, CONNECTION_LOST_WHILE_PENDING);
443 /* we expect this list to be empty. */
444 if (!list_empty(&tconn->out_of_sequence_requests))
445 conn_err(tconn, "ASSERT FAILED list_empty(&out_of_sequence_requests)\n");
447 /* but just in case, clean it up anyways! */
448 list_for_each_safe(le, tle, &tconn->out_of_sequence_requests) {
449 r = list_entry(le, struct drbd_request, tl_requests);
450 /* It would be nice to complete outside of spinlock.
451 * But this is easier for now. */
452 _req_mod(r, CONNECTION_LOST_WHILE_PENDING);
455 /* ensure bit indicating barrier is required is clear */
456 idr_for_each_entry(&tconn->volumes, mdev, vnr)
457 clear_bit(CREATE_BARRIER, &mdev->flags);
459 spin_unlock_irq(&tconn->req_lock);
462 void tl_restart(struct drbd_tconn *tconn, enum drbd_req_event what)
464 spin_lock_irq(&tconn->req_lock);
465 _tl_restart(tconn, what);
466 spin_unlock_irq(&tconn->req_lock);
469 static int drbd_thread_setup(void *arg)
471 struct drbd_thread *thi = (struct drbd_thread *) arg;
472 struct drbd_tconn *tconn = thi->tconn;
476 snprintf(current->comm, sizeof(current->comm), "drbd_%c_%s",
477 thi->name[0], thi->tconn->name);
480 retval = thi->function(thi);
482 spin_lock_irqsave(&thi->t_lock, flags);
484 /* if the receiver has been "EXITING", the last thing it did
485 * was set the conn state to "StandAlone",
486 * if now a re-connect request comes in, conn state goes C_UNCONNECTED,
487 * and receiver thread will be "started".
488 * drbd_thread_start needs to set "RESTARTING" in that case.
489 * t_state check and assignment needs to be within the same spinlock,
490 * so either thread_start sees EXITING, and can remap to RESTARTING,
491 * or thread_start see NONE, and can proceed as normal.
494 if (thi->t_state == RESTARTING) {
495 conn_info(tconn, "Restarting %s thread\n", thi->name);
496 thi->t_state = RUNNING;
497 spin_unlock_irqrestore(&thi->t_lock, flags);
504 complete(&thi->stop);
505 spin_unlock_irqrestore(&thi->t_lock, flags);
507 conn_info(tconn, "Terminating %s\n", current->comm);
509 /* Release mod reference taken when thread was started */
510 module_put(THIS_MODULE);
514 static void drbd_thread_init(struct drbd_tconn *tconn, struct drbd_thread *thi,
515 int (*func) (struct drbd_thread *), char *name)
517 spin_lock_init(&thi->t_lock);
520 thi->function = func;
522 strncpy(thi->name, name, ARRAY_SIZE(thi->name));
525 int drbd_thread_start(struct drbd_thread *thi)
527 struct drbd_tconn *tconn = thi->tconn;
528 struct task_struct *nt;
531 /* is used from state engine doing drbd_thread_stop_nowait,
532 * while holding the req lock irqsave */
533 spin_lock_irqsave(&thi->t_lock, flags);
535 switch (thi->t_state) {
537 conn_info(tconn, "Starting %s thread (from %s [%d])\n",
538 thi->name, current->comm, current->pid);
540 /* Get ref on module for thread - this is released when thread exits */
541 if (!try_module_get(THIS_MODULE)) {
542 conn_err(tconn, "Failed to get module reference in drbd_thread_start\n");
543 spin_unlock_irqrestore(&thi->t_lock, flags);
547 init_completion(&thi->stop);
548 thi->reset_cpu_mask = 1;
549 thi->t_state = RUNNING;
550 spin_unlock_irqrestore(&thi->t_lock, flags);
551 flush_signals(current); /* otherw. may get -ERESTARTNOINTR */
553 nt = kthread_create(drbd_thread_setup, (void *) thi,
554 "drbd_%c_%s", thi->name[0], thi->tconn->name);
557 conn_err(tconn, "Couldn't start thread\n");
559 module_put(THIS_MODULE);
562 spin_lock_irqsave(&thi->t_lock, flags);
564 thi->t_state = RUNNING;
565 spin_unlock_irqrestore(&thi->t_lock, flags);
569 thi->t_state = RESTARTING;
570 conn_info(tconn, "Restarting %s thread (from %s [%d])\n",
571 thi->name, current->comm, current->pid);
576 spin_unlock_irqrestore(&thi->t_lock, flags);
584 void _drbd_thread_stop(struct drbd_thread *thi, int restart, int wait)
588 enum drbd_thread_state ns = restart ? RESTARTING : EXITING;
590 /* may be called from state engine, holding the req lock irqsave */
591 spin_lock_irqsave(&thi->t_lock, flags);
593 if (thi->t_state == NONE) {
594 spin_unlock_irqrestore(&thi->t_lock, flags);
596 drbd_thread_start(thi);
600 if (thi->t_state != ns) {
601 if (thi->task == NULL) {
602 spin_unlock_irqrestore(&thi->t_lock, flags);
608 init_completion(&thi->stop);
609 if (thi->task != current)
610 force_sig(DRBD_SIGKILL, thi->task);
613 spin_unlock_irqrestore(&thi->t_lock, flags);
616 wait_for_completion(&thi->stop);
619 static struct drbd_thread *drbd_task_to_thread(struct drbd_tconn *tconn, struct task_struct *task)
621 struct drbd_thread *thi =
622 task == tconn->receiver.task ? &tconn->receiver :
623 task == tconn->asender.task ? &tconn->asender :
624 task == tconn->worker.task ? &tconn->worker : NULL;
629 char *drbd_task_to_thread_name(struct drbd_tconn *tconn, struct task_struct *task)
631 struct drbd_thread *thi = drbd_task_to_thread(tconn, task);
632 return thi ? thi->name : task->comm;
635 int conn_lowest_minor(struct drbd_tconn *tconn)
638 struct drbd_conf *mdev;
640 mdev = idr_get_next(&tconn->volumes, &vnr);
643 return mdev_to_minor(mdev);
648 * drbd_calc_cpu_mask() - Generate CPU masks, spread over all CPUs
649 * @mdev: DRBD device.
651 * Forces all threads of a device onto the same CPU. This is beneficial for
652 * DRBD's performance. May be overwritten by user's configuration.
654 void drbd_calc_cpu_mask(struct drbd_tconn *tconn)
659 if (cpumask_weight(tconn->cpu_mask))
662 ord = conn_lowest_minor(tconn) % cpumask_weight(cpu_online_mask);
663 for_each_online_cpu(cpu) {
665 cpumask_set_cpu(cpu, tconn->cpu_mask);
669 /* should not be reached */
670 cpumask_setall(tconn->cpu_mask);
674 * drbd_thread_current_set_cpu() - modifies the cpu mask of the _current_ thread
675 * @mdev: DRBD device.
676 * @thi: drbd_thread object
678 * call in the "main loop" of _all_ threads, no need for any mutex, current won't die
681 void drbd_thread_current_set_cpu(struct drbd_thread *thi)
683 struct task_struct *p = current;
685 if (!thi->reset_cpu_mask)
687 thi->reset_cpu_mask = 0;
688 set_cpus_allowed_ptr(p, thi->tconn->cpu_mask);
692 static void prepare_header80(struct p_header80 *h, enum drbd_packet cmd, int size)
694 h->magic = cpu_to_be32(DRBD_MAGIC);
695 h->command = cpu_to_be16(cmd);
696 h->length = cpu_to_be16(size);
699 static void prepare_header95(struct p_header95 *h, enum drbd_packet cmd, int size)
701 h->magic = cpu_to_be16(DRBD_MAGIC_BIG);
702 h->command = cpu_to_be16(cmd);
703 h->length = cpu_to_be32(size);
706 static void _prepare_header(struct drbd_tconn *tconn, int vnr, struct p_header *h,
707 enum drbd_packet cmd, int size)
709 if (tconn->agreed_pro_version >= 95)
710 prepare_header95(&h->h95, cmd, size);
712 prepare_header80(&h->h80, cmd, size);
715 static void prepare_header(struct drbd_conf *mdev, struct p_header *h,
716 enum drbd_packet cmd, int size)
718 _prepare_header(mdev->tconn, mdev->vnr, h, cmd, size);
721 /* the appropriate socket mutex must be held already */
722 int _conn_send_cmd(struct drbd_tconn *tconn, int vnr, struct socket *sock,
723 enum drbd_packet cmd, struct p_header *h, size_t size,
728 _prepare_header(tconn, vnr, h, cmd, size - sizeof(struct p_header));
729 err = drbd_send_all(tconn, sock, h, size, msg_flags);
730 if (err && !signal_pending(current))
731 conn_warn(tconn, "short send %s size=%d\n",
732 cmdname(cmd), (int)size);
736 /* don't pass the socket. we may only look at it
737 * when we hold the appropriate socket mutex.
739 int conn_send_cmd(struct drbd_tconn *tconn, int vnr, struct drbd_socket *sock,
740 enum drbd_packet cmd, struct p_header *h, size_t size)
744 mutex_lock(&sock->mutex);
746 err = _conn_send_cmd(tconn, vnr, sock->socket, cmd, h, size, 0);
747 mutex_unlock(&sock->mutex);
751 int conn_send_cmd2(struct drbd_tconn *tconn, enum drbd_packet cmd, char *data,
757 prepare_header80(&h, cmd, size);
758 err = drbd_get_data_sock(tconn);
760 err = drbd_send_all(tconn, tconn->data.socket, &h, sizeof(h), 0);
762 err = drbd_send_all(tconn, tconn->data.socket, data, size, 0);
763 drbd_put_data_sock(tconn);
768 int drbd_send_ping(struct drbd_tconn *tconn)
771 return !conn_send_cmd(tconn, 0, &tconn->meta, P_PING, &h, sizeof(h));
774 int drbd_send_ping_ack(struct drbd_tconn *tconn)
777 return !conn_send_cmd(tconn, 0, &tconn->meta, P_PING_ACK, &h, sizeof(h));
780 int drbd_send_sync_param(struct drbd_conf *mdev)
782 struct p_rs_param_95 *p;
785 const int apv = mdev->tconn->agreed_pro_version;
787 size = apv <= 87 ? sizeof(struct p_rs_param)
788 : apv == 88 ? sizeof(struct p_rs_param)
789 + strlen(mdev->tconn->net_conf->verify_alg) + 1
790 : apv <= 94 ? sizeof(struct p_rs_param_89)
791 : /* apv >= 95 */ sizeof(struct p_rs_param_95);
793 /* used from admin command context and receiver/worker context.
794 * to avoid kmalloc, grab the socket right here,
795 * then use the pre-allocated sbuf there */
796 mutex_lock(&mdev->tconn->data.mutex);
797 sock = mdev->tconn->data.socket;
799 if (likely(sock != NULL)) {
800 enum drbd_packet cmd =
801 apv >= 89 ? P_SYNC_PARAM89 : P_SYNC_PARAM;
803 p = mdev->tconn->data.sbuf;
805 /* initialize verify_alg and csums_alg */
806 memset(p->verify_alg, 0, 2 * SHARED_SECRET_MAX);
808 if (get_ldev(mdev)) {
809 p->rate = cpu_to_be32(mdev->ldev->dc.resync_rate);
810 p->c_plan_ahead = cpu_to_be32(mdev->ldev->dc.c_plan_ahead);
811 p->c_delay_target = cpu_to_be32(mdev->ldev->dc.c_delay_target);
812 p->c_fill_target = cpu_to_be32(mdev->ldev->dc.c_fill_target);
813 p->c_max_rate = cpu_to_be32(mdev->ldev->dc.c_max_rate);
816 p->rate = cpu_to_be32(DRBD_RATE_DEF);
817 p->c_plan_ahead = cpu_to_be32(DRBD_C_PLAN_AHEAD_DEF);
818 p->c_delay_target = cpu_to_be32(DRBD_C_DELAY_TARGET_DEF);
819 p->c_fill_target = cpu_to_be32(DRBD_C_FILL_TARGET_DEF);
820 p->c_max_rate = cpu_to_be32(DRBD_C_MAX_RATE_DEF);
824 strcpy(p->verify_alg, mdev->tconn->net_conf->verify_alg);
826 strcpy(p->csums_alg, mdev->tconn->net_conf->csums_alg);
828 err = _drbd_send_cmd(mdev, sock, cmd, &p->head, size, 0);
832 mutex_unlock(&mdev->tconn->data.mutex);
837 int drbd_send_protocol(struct drbd_tconn *tconn)
839 struct p_protocol *p;
842 size = sizeof(struct p_protocol);
844 if (tconn->agreed_pro_version >= 87)
845 size += strlen(tconn->net_conf->integrity_alg) + 1;
847 /* we must not recurse into our own queue,
848 * as that is blocked during handshake */
849 p = kmalloc(size, GFP_NOIO);
853 p->protocol = cpu_to_be32(tconn->net_conf->wire_protocol);
854 p->after_sb_0p = cpu_to_be32(tconn->net_conf->after_sb_0p);
855 p->after_sb_1p = cpu_to_be32(tconn->net_conf->after_sb_1p);
856 p->after_sb_2p = cpu_to_be32(tconn->net_conf->after_sb_2p);
857 p->two_primaries = cpu_to_be32(tconn->net_conf->two_primaries);
860 if (tconn->net_conf->want_lose)
862 if (tconn->net_conf->dry_run) {
863 if (tconn->agreed_pro_version >= 92)
866 conn_err(tconn, "--dry-run is not supported by peer");
871 p->conn_flags = cpu_to_be32(cf);
873 if (tconn->agreed_pro_version >= 87)
874 strcpy(p->integrity_alg, tconn->net_conf->integrity_alg);
876 err = conn_send_cmd2(tconn, P_PROTOCOL, p->head.payload, size - sizeof(struct p_header));
881 int _drbd_send_uuids(struct drbd_conf *mdev, u64 uuid_flags)
886 if (!get_ldev_if_state(mdev, D_NEGOTIATING))
889 for (i = UI_CURRENT; i < UI_SIZE; i++)
890 p.uuid[i] = mdev->ldev ? cpu_to_be64(mdev->ldev->md.uuid[i]) : 0;
892 mdev->comm_bm_set = drbd_bm_total_weight(mdev);
893 p.uuid[UI_SIZE] = cpu_to_be64(mdev->comm_bm_set);
894 uuid_flags |= mdev->tconn->net_conf->want_lose ? 1 : 0;
895 uuid_flags |= test_bit(CRASHED_PRIMARY, &mdev->flags) ? 2 : 0;
896 uuid_flags |= mdev->new_state_tmp.disk == D_INCONSISTENT ? 4 : 0;
897 p.uuid[UI_FLAGS] = cpu_to_be64(uuid_flags);
901 return drbd_send_cmd(mdev, &mdev->tconn->data, P_UUIDS, &p.head, sizeof(p));
904 int drbd_send_uuids(struct drbd_conf *mdev)
906 return _drbd_send_uuids(mdev, 0);
909 int drbd_send_uuids_skip_initial_sync(struct drbd_conf *mdev)
911 return _drbd_send_uuids(mdev, 8);
914 void drbd_print_uuids(struct drbd_conf *mdev, const char *text)
916 if (get_ldev_if_state(mdev, D_NEGOTIATING)) {
917 u64 *uuid = mdev->ldev->md.uuid;
918 dev_info(DEV, "%s %016llX:%016llX:%016llX:%016llX\n",
920 (unsigned long long)uuid[UI_CURRENT],
921 (unsigned long long)uuid[UI_BITMAP],
922 (unsigned long long)uuid[UI_HISTORY_START],
923 (unsigned long long)uuid[UI_HISTORY_END]);
926 dev_info(DEV, "%s effective data uuid: %016llX\n",
928 (unsigned long long)mdev->ed_uuid);
932 void drbd_gen_and_send_sync_uuid(struct drbd_conf *mdev)
937 D_ASSERT(mdev->state.disk == D_UP_TO_DATE);
939 uuid = mdev->ldev->md.uuid[UI_BITMAP] + UUID_NEW_BM_OFFSET;
940 drbd_uuid_set(mdev, UI_BITMAP, uuid);
941 drbd_print_uuids(mdev, "updated sync UUID");
943 p.uuid = cpu_to_be64(uuid);
945 drbd_send_cmd(mdev, &mdev->tconn->data, P_SYNC_UUID, &p.head, sizeof(p));
948 int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply, enum dds_flags flags)
951 sector_t d_size, u_size;
952 int q_order_type, max_bio_size;
954 if (get_ldev_if_state(mdev, D_NEGOTIATING)) {
955 D_ASSERT(mdev->ldev->backing_bdev);
956 d_size = drbd_get_max_capacity(mdev->ldev);
957 u_size = mdev->ldev->dc.disk_size;
958 q_order_type = drbd_queue_order_type(mdev);
959 max_bio_size = queue_max_hw_sectors(mdev->ldev->backing_bdev->bd_disk->queue) << 9;
960 max_bio_size = min_t(int, max_bio_size, DRBD_MAX_BIO_SIZE);
965 q_order_type = QUEUE_ORDERED_NONE;
966 max_bio_size = DRBD_MAX_BIO_SIZE; /* ... multiple BIOs per peer_request */
969 p.d_size = cpu_to_be64(d_size);
970 p.u_size = cpu_to_be64(u_size);
971 p.c_size = cpu_to_be64(trigger_reply ? 0 : drbd_get_capacity(mdev->this_bdev));
972 p.max_bio_size = cpu_to_be32(max_bio_size);
973 p.queue_order_type = cpu_to_be16(q_order_type);
974 p.dds_flags = cpu_to_be16(flags);
976 return drbd_send_cmd(mdev, &mdev->tconn->data, P_SIZES, &p.head, sizeof(p));
980 * drbd_send_state() - Sends the drbd state to the peer
981 * @mdev: DRBD device.
983 int drbd_send_state(struct drbd_conf *mdev)
989 mutex_lock(&mdev->tconn->data.mutex);
991 p.state = cpu_to_be32(mdev->state.i); /* Within the send mutex */
992 sock = mdev->tconn->data.socket;
994 if (likely(sock != NULL))
995 err = _drbd_send_cmd(mdev, sock, P_STATE, &p.head, sizeof(p), 0);
997 mutex_unlock(&mdev->tconn->data.mutex);
1002 int _conn_send_state_req(struct drbd_tconn *tconn, int vnr, enum drbd_packet cmd,
1003 union drbd_state mask, union drbd_state val)
1005 struct p_req_state p;
1007 p.mask = cpu_to_be32(mask.i);
1008 p.val = cpu_to_be32(val.i);
1010 return conn_send_cmd(tconn, vnr, &tconn->data, cmd, &p.head, sizeof(p));
1013 void drbd_send_sr_reply(struct drbd_conf *mdev, enum drbd_state_rv retcode)
1015 struct p_req_state_reply p;
1017 p.retcode = cpu_to_be32(retcode);
1019 drbd_send_cmd(mdev, &mdev->tconn->meta, P_STATE_CHG_REPLY, &p.head, sizeof(p));
1022 int conn_send_sr_reply(struct drbd_tconn *tconn, enum drbd_state_rv retcode)
1024 struct p_req_state_reply p;
1025 enum drbd_packet cmd = tconn->agreed_pro_version < 100 ? P_STATE_CHG_REPLY : P_CONN_ST_CHG_REPLY;
1027 p.retcode = cpu_to_be32(retcode);
1029 return !conn_send_cmd(tconn, 0, &tconn->meta, cmd, &p.head, sizeof(p));
1032 static void dcbp_set_code(struct p_compressed_bm *p, enum drbd_bitmap_code code)
1034 BUG_ON(code & ~0xf);
1035 p->encoding = (p->encoding & ~0xf) | code;
1038 static void dcbp_set_start(struct p_compressed_bm *p, int set)
1040 p->encoding = (p->encoding & ~0x80) | (set ? 0x80 : 0);
1043 static void dcbp_set_pad_bits(struct p_compressed_bm *p, int n)
1046 p->encoding = (p->encoding & (~0x7 << 4)) | (n << 4);
1049 int fill_bitmap_rle_bits(struct drbd_conf *mdev,
1050 struct p_compressed_bm *p,
1051 struct bm_xfer_ctx *c)
1053 struct bitstream bs;
1054 unsigned long plain_bits;
1061 /* may we use this feature? */
1062 if ((mdev->tconn->net_conf->use_rle == 0) ||
1063 (mdev->tconn->agreed_pro_version < 90))
1066 if (c->bit_offset >= c->bm_bits)
1067 return 0; /* nothing to do. */
1069 /* use at most thus many bytes */
1070 bitstream_init(&bs, p->code, BM_PACKET_VLI_BYTES_MAX, 0);
1071 memset(p->code, 0, BM_PACKET_VLI_BYTES_MAX);
1072 /* plain bits covered in this code string */
1075 /* p->encoding & 0x80 stores whether the first run length is set.
1076 * bit offset is implicit.
1077 * start with toggle == 2 to be able to tell the first iteration */
1080 /* see how much plain bits we can stuff into one packet
1081 * using RLE and VLI. */
1083 tmp = (toggle == 0) ? _drbd_bm_find_next_zero(mdev, c->bit_offset)
1084 : _drbd_bm_find_next(mdev, c->bit_offset);
1087 rl = tmp - c->bit_offset;
1089 if (toggle == 2) { /* first iteration */
1091 /* the first checked bit was set,
1092 * store start value, */
1093 dcbp_set_start(p, 1);
1094 /* but skip encoding of zero run length */
1098 dcbp_set_start(p, 0);
1101 /* paranoia: catch zero runlength.
1102 * can only happen if bitmap is modified while we scan it. */
1104 dev_err(DEV, "unexpected zero runlength while encoding bitmap "
1105 "t:%u bo:%lu\n", toggle, c->bit_offset);
1109 bits = vli_encode_bits(&bs, rl);
1110 if (bits == -ENOBUFS) /* buffer full */
1113 dev_err(DEV, "error while encoding bitmap: %d\n", bits);
1119 c->bit_offset = tmp;
1120 } while (c->bit_offset < c->bm_bits);
1122 len = bs.cur.b - p->code + !!bs.cur.bit;
1124 if (plain_bits < (len << 3)) {
1125 /* incompressible with this method.
1126 * we need to rewind both word and bit position. */
1127 c->bit_offset -= plain_bits;
1128 bm_xfer_ctx_bit_to_word_offset(c);
1129 c->bit_offset = c->word_offset * BITS_PER_LONG;
1133 /* RLE + VLI was able to compress it just fine.
1134 * update c->word_offset. */
1135 bm_xfer_ctx_bit_to_word_offset(c);
1137 /* store pad_bits */
1138 dcbp_set_pad_bits(p, (8 - bs.cur.bit) & 0x7);
1144 * send_bitmap_rle_or_plain
1146 * Return 0 when done, 1 when another iteration is needed, and a negative error
1147 * code upon failure.
1150 send_bitmap_rle_or_plain(struct drbd_conf *mdev,
1151 struct p_header *h, struct bm_xfer_ctx *c)
1153 struct p_compressed_bm *p = (void*)h;
1154 unsigned long num_words;
1157 len = fill_bitmap_rle_bits(mdev, p, c);
1163 dcbp_set_code(p, RLE_VLI_Bits);
1164 err = _drbd_send_cmd(mdev, mdev->tconn->data.socket,
1165 P_COMPRESSED_BITMAP, h,
1166 sizeof(*p) + len, 0);
1169 c->bytes[0] += sizeof(*p) + len;
1171 if (c->bit_offset >= c->bm_bits)
1174 /* was not compressible.
1175 * send a buffer full of plain text bits instead. */
1176 num_words = min_t(size_t, BM_PACKET_WORDS, c->bm_words - c->word_offset);
1177 len = num_words * sizeof(long);
1179 drbd_bm_get_lel(mdev, c->word_offset, num_words, (unsigned long*)h->payload);
1180 err = _drbd_send_cmd(mdev, mdev->tconn->data.socket, P_BITMAP,
1181 h, sizeof(struct p_header80) + len, 0);
1182 c->word_offset += num_words;
1183 c->bit_offset = c->word_offset * BITS_PER_LONG;
1186 c->bytes[1] += sizeof(struct p_header80) + len;
1188 if (c->bit_offset > c->bm_bits)
1189 c->bit_offset = c->bm_bits;
1193 INFO_bm_xfer_stats(mdev, "send", c);
1201 /* See the comment at receive_bitmap() */
1202 static int _drbd_send_bitmap(struct drbd_conf *mdev)
1204 struct bm_xfer_ctx c;
1208 if (!expect(mdev->bitmap))
1211 /* maybe we should use some per thread scratch page,
1212 * and allocate that during initial device creation? */
1213 p = (struct p_header *) __get_free_page(GFP_NOIO);
1215 dev_err(DEV, "failed to allocate one page buffer in %s\n", __func__);
1219 if (get_ldev(mdev)) {
1220 if (drbd_md_test_flag(mdev->ldev, MDF_FULL_SYNC)) {
1221 dev_info(DEV, "Writing the whole bitmap, MDF_FullSync was set.\n");
1222 drbd_bm_set_all(mdev);
1223 if (drbd_bm_write(mdev)) {
1224 /* write_bm did fail! Leave full sync flag set in Meta P_DATA
1225 * but otherwise process as per normal - need to tell other
1226 * side that a full resync is required! */
1227 dev_err(DEV, "Failed to write bitmap to disk!\n");
1229 drbd_md_clear_flag(mdev, MDF_FULL_SYNC);
1236 c = (struct bm_xfer_ctx) {
1237 .bm_bits = drbd_bm_bits(mdev),
1238 .bm_words = drbd_bm_words(mdev),
1242 err = send_bitmap_rle_or_plain(mdev, p, &c);
1245 free_page((unsigned long) p);
1249 int drbd_send_bitmap(struct drbd_conf *mdev)
1253 if (drbd_get_data_sock(mdev->tconn))
1255 err = !_drbd_send_bitmap(mdev);
1256 drbd_put_data_sock(mdev->tconn);
1259 void drbd_send_b_ack(struct drbd_conf *mdev, u32 barrier_nr, u32 set_size)
1261 struct p_barrier_ack p;
1263 p.barrier = barrier_nr;
1264 p.set_size = cpu_to_be32(set_size);
1266 if (mdev->state.conn >= C_CONNECTED)
1267 drbd_send_cmd(mdev, &mdev->tconn->meta, P_BARRIER_ACK, &p.head, sizeof(p));
1271 * _drbd_send_ack() - Sends an ack packet
1272 * @mdev: DRBD device.
1273 * @cmd: Packet command code.
1274 * @sector: sector, needs to be in big endian byte order
1275 * @blksize: size in byte, needs to be in big endian byte order
1276 * @block_id: Id, big endian byte order
1278 static int _drbd_send_ack(struct drbd_conf *mdev, enum drbd_packet cmd,
1279 u64 sector, u32 blksize, u64 block_id)
1281 struct p_block_ack p;
1284 p.block_id = block_id;
1285 p.blksize = blksize;
1286 p.seq_num = cpu_to_be32(atomic_inc_return(&mdev->packet_seq));
1288 if (!mdev->tconn->meta.socket || mdev->state.conn < C_CONNECTED)
1290 return drbd_send_cmd(mdev, &mdev->tconn->meta, cmd, &p.head, sizeof(p));
1293 /* dp->sector and dp->block_id already/still in network byte order,
1294 * data_size is payload size according to dp->head,
1295 * and may need to be corrected for digest size. */
1296 void drbd_send_ack_dp(struct drbd_conf *mdev, enum drbd_packet cmd,
1297 struct p_data *dp, int data_size)
1299 data_size -= (mdev->tconn->agreed_pro_version >= 87 && mdev->tconn->integrity_r_tfm) ?
1300 crypto_hash_digestsize(mdev->tconn->integrity_r_tfm) : 0;
1301 _drbd_send_ack(mdev, cmd, dp->sector, cpu_to_be32(data_size),
1305 void drbd_send_ack_rp(struct drbd_conf *mdev, enum drbd_packet cmd,
1306 struct p_block_req *rp)
1308 _drbd_send_ack(mdev, cmd, rp->sector, rp->blksize, rp->block_id);
1312 * drbd_send_ack() - Sends an ack packet
1313 * @mdev: DRBD device
1314 * @cmd: packet command code
1315 * @peer_req: peer request
1317 int drbd_send_ack(struct drbd_conf *mdev, enum drbd_packet cmd,
1318 struct drbd_peer_request *peer_req)
1320 return _drbd_send_ack(mdev, cmd,
1321 cpu_to_be64(peer_req->i.sector),
1322 cpu_to_be32(peer_req->i.size),
1323 peer_req->block_id);
1326 /* This function misuses the block_id field to signal if the blocks
1327 * are is sync or not. */
1328 int drbd_send_ack_ex(struct drbd_conf *mdev, enum drbd_packet cmd,
1329 sector_t sector, int blksize, u64 block_id)
1331 return _drbd_send_ack(mdev, cmd,
1332 cpu_to_be64(sector),
1333 cpu_to_be32(blksize),
1334 cpu_to_be64(block_id));
1337 int drbd_send_drequest(struct drbd_conf *mdev, int cmd,
1338 sector_t sector, int size, u64 block_id)
1340 struct p_block_req p;
1342 p.sector = cpu_to_be64(sector);
1343 p.block_id = block_id;
1344 p.blksize = cpu_to_be32(size);
1346 return drbd_send_cmd(mdev, &mdev->tconn->data, cmd, &p.head, sizeof(p));
1349 int drbd_send_drequest_csum(struct drbd_conf *mdev, sector_t sector, int size,
1350 void *digest, int digest_size, enum drbd_packet cmd)
1353 struct p_block_req p;
1355 prepare_header(mdev, &p.head, cmd, sizeof(p) - sizeof(struct p_header) + digest_size);
1356 p.sector = cpu_to_be64(sector);
1357 p.block_id = ID_SYNCER /* unused */;
1358 p.blksize = cpu_to_be32(size);
1360 mutex_lock(&mdev->tconn->data.mutex);
1361 err = drbd_send_all(mdev->tconn, mdev->tconn->data.socket, &p, sizeof(p), 0);
1363 err = drbd_send_all(mdev->tconn, mdev->tconn->data.socket, digest, digest_size, 0);
1364 mutex_unlock(&mdev->tconn->data.mutex);
1368 int drbd_send_ov_request(struct drbd_conf *mdev, sector_t sector, int size)
1370 struct p_block_req p;
1372 p.sector = cpu_to_be64(sector);
1373 p.block_id = ID_SYNCER /* unused */;
1374 p.blksize = cpu_to_be32(size);
1376 return drbd_send_cmd(mdev, &mdev->tconn->data, P_OV_REQUEST, &p.head, sizeof(p));
1379 /* called on sndtimeo
1380 * returns false if we should retry,
1381 * true if we think connection is dead
1383 static int we_should_drop_the_connection(struct drbd_tconn *tconn, struct socket *sock)
1386 /* long elapsed = (long)(jiffies - mdev->last_received); */
1388 drop_it = tconn->meta.socket == sock
1389 || !tconn->asender.task
1390 || get_t_state(&tconn->asender) != RUNNING
1391 || tconn->cstate < C_WF_REPORT_PARAMS;
1396 drop_it = !--tconn->ko_count;
1398 conn_err(tconn, "[%s/%d] sock_sendmsg time expired, ko = %u\n",
1399 current->comm, current->pid, tconn->ko_count);
1400 request_ping(tconn);
1403 return drop_it; /* && (mdev->state == R_PRIMARY) */;
1406 static void drbd_update_congested(struct drbd_tconn *tconn)
1408 struct sock *sk = tconn->data.socket->sk;
1409 if (sk->sk_wmem_queued > sk->sk_sndbuf * 4 / 5)
1410 set_bit(NET_CONGESTED, &tconn->flags);
1413 /* The idea of sendpage seems to be to put some kind of reference
1414 * to the page into the skb, and to hand it over to the NIC. In
1415 * this process get_page() gets called.
1417 * As soon as the page was really sent over the network put_page()
1418 * gets called by some part of the network layer. [ NIC driver? ]
1420 * [ get_page() / put_page() increment/decrement the count. If count
1421 * reaches 0 the page will be freed. ]
1423 * This works nicely with pages from FSs.
1424 * But this means that in protocol A we might signal IO completion too early!
1426 * In order not to corrupt data during a resync we must make sure
1427 * that we do not reuse our own buffer pages (EEs) to early, therefore
1428 * we have the net_ee list.
1430 * XFS seems to have problems, still, it submits pages with page_count == 0!
1431 * As a workaround, we disable sendpage on pages
1432 * with page_count == 0 or PageSlab.
1434 static int _drbd_no_send_page(struct drbd_conf *mdev, struct page *page,
1435 int offset, size_t size, unsigned msg_flags)
1437 struct socket *socket;
1441 socket = mdev->tconn->data.socket;
1442 addr = kmap(page) + offset;
1443 err = drbd_send_all(mdev->tconn, socket, addr, size, msg_flags);
1446 mdev->send_cnt += size >> 9;
1450 static int _drbd_send_page(struct drbd_conf *mdev, struct page *page,
1451 int offset, size_t size, unsigned msg_flags)
1453 struct socket *socket = mdev->tconn->data.socket;
1454 mm_segment_t oldfs = get_fs();
1458 /* e.g. XFS meta- & log-data is in slab pages, which have a
1459 * page_count of 0 and/or have PageSlab() set.
1460 * we cannot use send_page for those, as that does get_page();
1461 * put_page(); and would cause either a VM_BUG directly, or
1462 * __page_cache_release a page that would actually still be referenced
1463 * by someone, leading to some obscure delayed Oops somewhere else. */
1464 if (disable_sendpage || (page_count(page) < 1) || PageSlab(page))
1465 return _drbd_no_send_page(mdev, page, offset, size, msg_flags);
1467 msg_flags |= MSG_NOSIGNAL;
1468 drbd_update_congested(mdev->tconn);
1473 sent = socket->ops->sendpage(socket, page, offset, len, msg_flags);
1475 if (sent == -EAGAIN) {
1476 if (we_should_drop_the_connection(mdev->tconn, socket))
1480 dev_warn(DEV, "%s: size=%d len=%d sent=%d\n",
1481 __func__, (int)size, len, sent);
1488 } while (len > 0 /* THINK && mdev->cstate >= C_CONNECTED*/);
1490 clear_bit(NET_CONGESTED, &mdev->tconn->flags);
1494 mdev->send_cnt += size >> 9;
1499 static int _drbd_send_bio(struct drbd_conf *mdev, struct bio *bio)
1501 struct bio_vec *bvec;
1503 /* hint all but last page with MSG_MORE */
1504 __bio_for_each_segment(bvec, bio, i, 0) {
1507 err = _drbd_no_send_page(mdev, bvec->bv_page,
1508 bvec->bv_offset, bvec->bv_len,
1509 i == bio->bi_vcnt - 1 ? 0 : MSG_MORE);
1516 static int _drbd_send_zc_bio(struct drbd_conf *mdev, struct bio *bio)
1518 struct bio_vec *bvec;
1520 /* hint all but last page with MSG_MORE */
1521 __bio_for_each_segment(bvec, bio, i, 0) {
1524 err = _drbd_send_page(mdev, bvec->bv_page,
1525 bvec->bv_offset, bvec->bv_len,
1526 i == bio->bi_vcnt - 1 ? 0 : MSG_MORE);
1533 static int _drbd_send_zc_ee(struct drbd_conf *mdev,
1534 struct drbd_peer_request *peer_req)
1536 struct page *page = peer_req->pages;
1537 unsigned len = peer_req->i.size;
1540 /* hint all but last page with MSG_MORE */
1541 page_chain_for_each(page) {
1542 unsigned l = min_t(unsigned, len, PAGE_SIZE);
1544 err = _drbd_send_page(mdev, page, 0, l,
1545 page_chain_next(page) ? MSG_MORE : 0);
1553 static u32 bio_flags_to_wire(struct drbd_conf *mdev, unsigned long bi_rw)
1555 if (mdev->tconn->agreed_pro_version >= 95)
1556 return (bi_rw & REQ_SYNC ? DP_RW_SYNC : 0) |
1557 (bi_rw & REQ_FUA ? DP_FUA : 0) |
1558 (bi_rw & REQ_FLUSH ? DP_FLUSH : 0) |
1559 (bi_rw & REQ_DISCARD ? DP_DISCARD : 0);
1561 return bi_rw & REQ_SYNC ? DP_RW_SYNC : 0;
1564 /* Used to send write requests
1565 * R_PRIMARY -> Peer (P_DATA)
1567 int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req)
1571 unsigned int dp_flags = 0;
1575 err = drbd_get_data_sock(mdev->tconn);
1579 dgs = (mdev->tconn->agreed_pro_version >= 87 && mdev->tconn->integrity_w_tfm) ?
1580 crypto_hash_digestsize(mdev->tconn->integrity_w_tfm) : 0;
1582 prepare_header(mdev, &p.head, P_DATA, sizeof(p) - sizeof(struct p_header) + dgs + req->i.size);
1583 p.sector = cpu_to_be64(req->i.sector);
1584 p.block_id = (unsigned long)req;
1585 p.seq_num = cpu_to_be32(req->seq_num = atomic_inc_return(&mdev->packet_seq));
1587 dp_flags = bio_flags_to_wire(mdev, req->master_bio->bi_rw);
1589 if (mdev->state.conn >= C_SYNC_SOURCE &&
1590 mdev->state.conn <= C_PAUSED_SYNC_T)
1591 dp_flags |= DP_MAY_SET_IN_SYNC;
1593 p.dp_flags = cpu_to_be32(dp_flags);
1594 set_bit(UNPLUG_REMOTE, &mdev->flags);
1595 err = drbd_send_all(mdev->tconn, mdev->tconn->data.socket, &p,
1596 sizeof(p), dgs ? MSG_MORE : 0);
1598 dgb = mdev->tconn->int_dig_out;
1599 drbd_csum_bio(mdev, mdev->tconn->integrity_w_tfm, req->master_bio, dgb);
1600 err = drbd_send_all(mdev->tconn, mdev->tconn->data.socket, dgb, dgs, 0);
1603 /* For protocol A, we have to memcpy the payload into
1604 * socket buffers, as we may complete right away
1605 * as soon as we handed it over to tcp, at which point the data
1606 * pages may become invalid.
1608 * For data-integrity enabled, we copy it as well, so we can be
1609 * sure that even if the bio pages may still be modified, it
1610 * won't change the data on the wire, thus if the digest checks
1611 * out ok after sending on this side, but does not fit on the
1612 * receiving side, we sure have detected corruption elsewhere.
1614 if (mdev->tconn->net_conf->wire_protocol == DRBD_PROT_A || dgs)
1615 err = _drbd_send_bio(mdev, req->master_bio);
1617 err = _drbd_send_zc_bio(mdev, req->master_bio);
1619 /* double check digest, sometimes buffers have been modified in flight. */
1620 if (dgs > 0 && dgs <= 64) {
1621 /* 64 byte, 512 bit, is the largest digest size
1622 * currently supported in kernel crypto. */
1623 unsigned char digest[64];
1624 drbd_csum_bio(mdev, mdev->tconn->integrity_w_tfm, req->master_bio, digest);
1625 if (memcmp(mdev->tconn->int_dig_out, digest, dgs)) {
1627 "Digest mismatch, buffer modified by upper layers during write: %llus +%u\n",
1628 (unsigned long long)req->i.sector, req->i.size);
1630 } /* else if (dgs > 64) {
1631 ... Be noisy about digest too large ...
1635 drbd_put_data_sock(mdev->tconn);
1640 /* answer packet, used to send data back for read requests:
1641 * Peer -> (diskless) R_PRIMARY (P_DATA_REPLY)
1642 * C_SYNC_SOURCE -> C_SYNC_TARGET (P_RS_DATA_REPLY)
1644 int drbd_send_block(struct drbd_conf *mdev, enum drbd_packet cmd,
1645 struct drbd_peer_request *peer_req)
1652 dgs = (mdev->tconn->agreed_pro_version >= 87 && mdev->tconn->integrity_w_tfm) ?
1653 crypto_hash_digestsize(mdev->tconn->integrity_w_tfm) : 0;
1655 prepare_header(mdev, &p.head, cmd, sizeof(p) -
1656 sizeof(struct p_header80) +
1657 dgs + peer_req->i.size);
1658 p.sector = cpu_to_be64(peer_req->i.sector);
1659 p.block_id = peer_req->block_id;
1660 p.seq_num = 0; /* unused */
1662 /* Only called by our kernel thread.
1663 * This one may be interrupted by DRBD_SIG and/or DRBD_SIGKILL
1664 * in response to admin command or module unload.
1666 err = drbd_get_data_sock(mdev->tconn);
1669 err = drbd_send_all(mdev->tconn, mdev->tconn->data.socket, &p,
1670 sizeof(p), dgs ? MSG_MORE : 0);
1672 dgb = mdev->tconn->int_dig_out;
1673 drbd_csum_ee(mdev, mdev->tconn->integrity_w_tfm, peer_req, dgb);
1674 err = drbd_send_all(mdev->tconn, mdev->tconn->data.socket, dgb,
1678 err = _drbd_send_zc_ee(mdev, peer_req);
1679 drbd_put_data_sock(mdev->tconn);
1684 int drbd_send_out_of_sync(struct drbd_conf *mdev, struct drbd_request *req)
1686 struct p_block_desc p;
1688 p.sector = cpu_to_be64(req->i.sector);
1689 p.blksize = cpu_to_be32(req->i.size);
1691 return drbd_send_cmd(mdev, &mdev->tconn->data, P_OUT_OF_SYNC, &p.head, sizeof(p));
1695 drbd_send distinguishes two cases:
1697 Packets sent via the data socket "sock"
1698 and packets sent via the meta data socket "msock"
1701 -----------------+-------------------------+------------------------------
1702 timeout conf.timeout / 2 conf.timeout / 2
1703 timeout action send a ping via msock Abort communication
1704 and close all sockets
1708 * you must have down()ed the appropriate [m]sock_mutex elsewhere!
1710 int drbd_send(struct drbd_tconn *tconn, struct socket *sock,
1711 void *buf, size_t size, unsigned msg_flags)
1720 /* THINK if (signal_pending) return ... ? */
1725 msg.msg_name = NULL;
1726 msg.msg_namelen = 0;
1727 msg.msg_control = NULL;
1728 msg.msg_controllen = 0;
1729 msg.msg_flags = msg_flags | MSG_NOSIGNAL;
1731 if (sock == tconn->data.socket) {
1732 tconn->ko_count = tconn->net_conf->ko_count;
1733 drbd_update_congested(tconn);
1737 * tcp_sendmsg does _not_ use its size parameter at all ?
1739 * -EAGAIN on timeout, -EINTR on signal.
1742 * do we need to block DRBD_SIG if sock == &meta.socket ??
1743 * otherwise wake_asender() might interrupt some send_*Ack !
1745 rv = kernel_sendmsg(sock, &msg, &iov, 1, size);
1746 if (rv == -EAGAIN) {
1747 if (we_should_drop_the_connection(tconn, sock))
1753 flush_signals(current);
1761 } while (sent < size);
1763 if (sock == tconn->data.socket)
1764 clear_bit(NET_CONGESTED, &tconn->flags);
1767 if (rv != -EAGAIN) {
1768 conn_err(tconn, "%s_sendmsg returned %d\n",
1769 sock == tconn->meta.socket ? "msock" : "sock",
1771 conn_request_state(tconn, NS(conn, C_BROKEN_PIPE), CS_HARD);
1773 conn_request_state(tconn, NS(conn, C_TIMEOUT), CS_HARD);
1780 * drbd_send_all - Send an entire buffer
1782 * Returns 0 upon success and a negative error value otherwise.
1784 int drbd_send_all(struct drbd_tconn *tconn, struct socket *sock, void *buffer,
1785 size_t size, unsigned msg_flags)
1789 err = drbd_send(tconn, sock, buffer, size, msg_flags);
1797 static int drbd_open(struct block_device *bdev, fmode_t mode)
1799 struct drbd_conf *mdev = bdev->bd_disk->private_data;
1800 unsigned long flags;
1803 mutex_lock(&drbd_main_mutex);
1804 spin_lock_irqsave(&mdev->tconn->req_lock, flags);
1805 /* to have a stable mdev->state.role
1806 * and no race with updating open_cnt */
1808 if (mdev->state.role != R_PRIMARY) {
1809 if (mode & FMODE_WRITE)
1811 else if (!allow_oos)
1817 spin_unlock_irqrestore(&mdev->tconn->req_lock, flags);
1818 mutex_unlock(&drbd_main_mutex);
1823 static int drbd_release(struct gendisk *gd, fmode_t mode)
1825 struct drbd_conf *mdev = gd->private_data;
1826 mutex_lock(&drbd_main_mutex);
1828 mutex_unlock(&drbd_main_mutex);
1832 static void drbd_set_defaults(struct drbd_conf *mdev)
1834 /* Beware! The actual layout differs
1835 * between big endian and little endian */
1836 mdev->state = (union drbd_state) {
1837 { .role = R_SECONDARY,
1839 .conn = C_STANDALONE,
1848 void drbd_init_set_defaults(struct drbd_conf *mdev)
1850 /* the memset(,0,) did most of this.
1851 * note: only assignments, no allocation in here */
1853 drbd_set_defaults(mdev);
1855 atomic_set(&mdev->ap_bio_cnt, 0);
1856 atomic_set(&mdev->ap_pending_cnt, 0);
1857 atomic_set(&mdev->rs_pending_cnt, 0);
1858 atomic_set(&mdev->unacked_cnt, 0);
1859 atomic_set(&mdev->local_cnt, 0);
1860 atomic_set(&mdev->pp_in_use_by_net, 0);
1861 atomic_set(&mdev->rs_sect_in, 0);
1862 atomic_set(&mdev->rs_sect_ev, 0);
1863 atomic_set(&mdev->ap_in_flight, 0);
1865 mutex_init(&mdev->md_io_mutex);
1866 mutex_init(&mdev->own_state_mutex);
1867 mdev->state_mutex = &mdev->own_state_mutex;
1869 spin_lock_init(&mdev->al_lock);
1870 spin_lock_init(&mdev->peer_seq_lock);
1871 spin_lock_init(&mdev->epoch_lock);
1873 INIT_LIST_HEAD(&mdev->active_ee);
1874 INIT_LIST_HEAD(&mdev->sync_ee);
1875 INIT_LIST_HEAD(&mdev->done_ee);
1876 INIT_LIST_HEAD(&mdev->read_ee);
1877 INIT_LIST_HEAD(&mdev->net_ee);
1878 INIT_LIST_HEAD(&mdev->resync_reads);
1879 INIT_LIST_HEAD(&mdev->resync_work.list);
1880 INIT_LIST_HEAD(&mdev->unplug_work.list);
1881 INIT_LIST_HEAD(&mdev->go_diskless.list);
1882 INIT_LIST_HEAD(&mdev->md_sync_work.list);
1883 INIT_LIST_HEAD(&mdev->start_resync_work.list);
1884 INIT_LIST_HEAD(&mdev->bm_io_work.w.list);
1886 mdev->resync_work.cb = w_resync_timer;
1887 mdev->unplug_work.cb = w_send_write_hint;
1888 mdev->go_diskless.cb = w_go_diskless;
1889 mdev->md_sync_work.cb = w_md_sync;
1890 mdev->bm_io_work.w.cb = w_bitmap_io;
1891 mdev->start_resync_work.cb = w_start_resync;
1893 mdev->resync_work.mdev = mdev;
1894 mdev->unplug_work.mdev = mdev;
1895 mdev->go_diskless.mdev = mdev;
1896 mdev->md_sync_work.mdev = mdev;
1897 mdev->bm_io_work.w.mdev = mdev;
1898 mdev->start_resync_work.mdev = mdev;
1900 init_timer(&mdev->resync_timer);
1901 init_timer(&mdev->md_sync_timer);
1902 init_timer(&mdev->start_resync_timer);
1903 init_timer(&mdev->request_timer);
1904 mdev->resync_timer.function = resync_timer_fn;
1905 mdev->resync_timer.data = (unsigned long) mdev;
1906 mdev->md_sync_timer.function = md_sync_timer_fn;
1907 mdev->md_sync_timer.data = (unsigned long) mdev;
1908 mdev->start_resync_timer.function = start_resync_timer_fn;
1909 mdev->start_resync_timer.data = (unsigned long) mdev;
1910 mdev->request_timer.function = request_timer_fn;
1911 mdev->request_timer.data = (unsigned long) mdev;
1913 init_waitqueue_head(&mdev->misc_wait);
1914 init_waitqueue_head(&mdev->state_wait);
1915 init_waitqueue_head(&mdev->ee_wait);
1916 init_waitqueue_head(&mdev->al_wait);
1917 init_waitqueue_head(&mdev->seq_wait);
1919 /* mdev->tconn->agreed_pro_version gets initialized in drbd_connect() */
1920 mdev->write_ordering = WO_bdev_flush;
1921 mdev->resync_wenr = LC_FREE;
1922 mdev->peer_max_bio_size = DRBD_MAX_BIO_SIZE_SAFE;
1923 mdev->local_max_bio_size = DRBD_MAX_BIO_SIZE_SAFE;
1926 void drbd_mdev_cleanup(struct drbd_conf *mdev)
1929 if (mdev->tconn->receiver.t_state != NONE)
1930 dev_err(DEV, "ASSERT FAILED: receiver t_state == %d expected 0.\n",
1931 mdev->tconn->receiver.t_state);
1933 /* no need to lock it, I'm the only thread alive */
1934 if (atomic_read(&mdev->current_epoch->epoch_size) != 0)
1935 dev_err(DEV, "epoch_size:%d\n", atomic_read(&mdev->current_epoch->epoch_size));
1945 mdev->rs_failed = 0;
1946 mdev->rs_last_events = 0;
1947 mdev->rs_last_sect_ev = 0;
1948 for (i = 0; i < DRBD_SYNC_MARKS; i++) {
1949 mdev->rs_mark_left[i] = 0;
1950 mdev->rs_mark_time[i] = 0;
1952 D_ASSERT(mdev->tconn->net_conf == NULL);
1954 drbd_set_my_capacity(mdev, 0);
1956 /* maybe never allocated. */
1957 drbd_bm_resize(mdev, 0, 1);
1958 drbd_bm_cleanup(mdev);
1961 drbd_free_resources(mdev);
1962 clear_bit(AL_SUSPENDED, &mdev->flags);
1965 * currently we drbd_init_ee only on module load, so
1966 * we may do drbd_release_ee only on module unload!
1968 D_ASSERT(list_empty(&mdev->active_ee));
1969 D_ASSERT(list_empty(&mdev->sync_ee));
1970 D_ASSERT(list_empty(&mdev->done_ee));
1971 D_ASSERT(list_empty(&mdev->read_ee));
1972 D_ASSERT(list_empty(&mdev->net_ee));
1973 D_ASSERT(list_empty(&mdev->resync_reads));
1974 D_ASSERT(list_empty(&mdev->tconn->data.work.q));
1975 D_ASSERT(list_empty(&mdev->tconn->meta.work.q));
1976 D_ASSERT(list_empty(&mdev->resync_work.list));
1977 D_ASSERT(list_empty(&mdev->unplug_work.list));
1978 D_ASSERT(list_empty(&mdev->go_diskless.list));
1980 drbd_set_defaults(mdev);
1984 static void drbd_destroy_mempools(void)
1988 while (drbd_pp_pool) {
1989 page = drbd_pp_pool;
1990 drbd_pp_pool = (struct page *)page_private(page);
1995 /* D_ASSERT(atomic_read(&drbd_pp_vacant)==0); */
1997 if (drbd_md_io_bio_set)
1998 bioset_free(drbd_md_io_bio_set);
1999 if (drbd_md_io_page_pool)
2000 mempool_destroy(drbd_md_io_page_pool);
2001 if (drbd_ee_mempool)
2002 mempool_destroy(drbd_ee_mempool);
2003 if (drbd_request_mempool)
2004 mempool_destroy(drbd_request_mempool);
2006 kmem_cache_destroy(drbd_ee_cache);
2007 if (drbd_request_cache)
2008 kmem_cache_destroy(drbd_request_cache);
2009 if (drbd_bm_ext_cache)
2010 kmem_cache_destroy(drbd_bm_ext_cache);
2011 if (drbd_al_ext_cache)
2012 kmem_cache_destroy(drbd_al_ext_cache);
2014 drbd_md_io_bio_set = NULL;
2015 drbd_md_io_page_pool = NULL;
2016 drbd_ee_mempool = NULL;
2017 drbd_request_mempool = NULL;
2018 drbd_ee_cache = NULL;
2019 drbd_request_cache = NULL;
2020 drbd_bm_ext_cache = NULL;
2021 drbd_al_ext_cache = NULL;
2026 static int drbd_create_mempools(void)
2029 const int number = (DRBD_MAX_BIO_SIZE/PAGE_SIZE) * minor_count;
2032 /* prepare our caches and mempools */
2033 drbd_request_mempool = NULL;
2034 drbd_ee_cache = NULL;
2035 drbd_request_cache = NULL;
2036 drbd_bm_ext_cache = NULL;
2037 drbd_al_ext_cache = NULL;
2038 drbd_pp_pool = NULL;
2039 drbd_md_io_page_pool = NULL;
2040 drbd_md_io_bio_set = NULL;
2043 drbd_request_cache = kmem_cache_create(
2044 "drbd_req", sizeof(struct drbd_request), 0, 0, NULL);
2045 if (drbd_request_cache == NULL)
2048 drbd_ee_cache = kmem_cache_create(
2049 "drbd_ee", sizeof(struct drbd_peer_request), 0, 0, NULL);
2050 if (drbd_ee_cache == NULL)
2053 drbd_bm_ext_cache = kmem_cache_create(
2054 "drbd_bm", sizeof(struct bm_extent), 0, 0, NULL);
2055 if (drbd_bm_ext_cache == NULL)
2058 drbd_al_ext_cache = kmem_cache_create(
2059 "drbd_al", sizeof(struct lc_element), 0, 0, NULL);
2060 if (drbd_al_ext_cache == NULL)
2064 drbd_md_io_bio_set = bioset_create(DRBD_MIN_POOL_PAGES, 0);
2065 if (drbd_md_io_bio_set == NULL)
2068 drbd_md_io_page_pool = mempool_create_page_pool(DRBD_MIN_POOL_PAGES, 0);
2069 if (drbd_md_io_page_pool == NULL)
2072 drbd_request_mempool = mempool_create(number,
2073 mempool_alloc_slab, mempool_free_slab, drbd_request_cache);
2074 if (drbd_request_mempool == NULL)
2077 drbd_ee_mempool = mempool_create(number,
2078 mempool_alloc_slab, mempool_free_slab, drbd_ee_cache);
2079 if (drbd_ee_mempool == NULL)
2082 /* drbd's page pool */
2083 spin_lock_init(&drbd_pp_lock);
2085 for (i = 0; i < number; i++) {
2086 page = alloc_page(GFP_HIGHUSER);
2089 set_page_private(page, (unsigned long)drbd_pp_pool);
2090 drbd_pp_pool = page;
2092 drbd_pp_vacant = number;
2097 drbd_destroy_mempools(); /* in case we allocated some */
2101 static int drbd_notify_sys(struct notifier_block *this, unsigned long code,
2104 /* just so we have it. you never know what interesting things we
2105 * might want to do here some day...
2111 static struct notifier_block drbd_notifier = {
2112 .notifier_call = drbd_notify_sys,
2115 static void drbd_release_ee_lists(struct drbd_conf *mdev)
2119 rr = drbd_release_ee(mdev, &mdev->active_ee);
2121 dev_err(DEV, "%d EEs in active list found!\n", rr);
2123 rr = drbd_release_ee(mdev, &mdev->sync_ee);
2125 dev_err(DEV, "%d EEs in sync list found!\n", rr);
2127 rr = drbd_release_ee(mdev, &mdev->read_ee);
2129 dev_err(DEV, "%d EEs in read list found!\n", rr);
2131 rr = drbd_release_ee(mdev, &mdev->done_ee);
2133 dev_err(DEV, "%d EEs in done list found!\n", rr);
2135 rr = drbd_release_ee(mdev, &mdev->net_ee);
2137 dev_err(DEV, "%d EEs in net list found!\n", rr);
2140 /* caution. no locking. */
2141 void drbd_delete_device(unsigned int minor)
2143 struct drbd_conf *mdev = minor_to_mdev(minor);
2148 idr_remove(&mdev->tconn->volumes, mdev->vnr);
2149 idr_remove(&minors, minor);
2152 /* paranoia asserts */
2153 D_ASSERT(mdev->open_cnt == 0);
2154 D_ASSERT(list_empty(&mdev->tconn->data.work.q));
2155 /* end paranoia asserts */
2157 del_gendisk(mdev->vdisk);
2159 /* cleanup stuff that may have been allocated during
2160 * device (re-)configuration or state changes */
2162 if (mdev->this_bdev)
2163 bdput(mdev->this_bdev);
2165 drbd_free_resources(mdev);
2167 drbd_release_ee_lists(mdev);
2169 lc_destroy(mdev->act_log);
2170 lc_destroy(mdev->resync);
2172 kfree(mdev->p_uuid);
2173 /* mdev->p_uuid = NULL; */
2175 /* cleanup the rest that has been
2176 * allocated from drbd_new_device
2177 * and actually free the mdev itself */
2178 drbd_free_mdev(mdev);
2181 static void drbd_cleanup(void)
2184 struct drbd_conf *mdev;
2186 unregister_reboot_notifier(&drbd_notifier);
2188 /* first remove proc,
2189 * drbdsetup uses it's presence to detect
2190 * whether DRBD is loaded.
2191 * If we would get stuck in proc removal,
2192 * but have netlink already deregistered,
2193 * some drbdsetup commands may wait forever
2197 remove_proc_entry("drbd", NULL);
2199 drbd_genl_unregister();
2201 idr_for_each_entry(&minors, mdev, i)
2202 drbd_delete_device(i);
2203 drbd_destroy_mempools();
2204 unregister_blkdev(DRBD_MAJOR, "drbd");
2206 idr_destroy(&minors);
2208 printk(KERN_INFO "drbd: module cleanup done.\n");
2212 * drbd_congested() - Callback for pdflush
2213 * @congested_data: User data
2214 * @bdi_bits: Bits pdflush is currently interested in
2216 * Returns 1<<BDI_async_congested and/or 1<<BDI_sync_congested if we are congested.
2218 static int drbd_congested(void *congested_data, int bdi_bits)
2220 struct drbd_conf *mdev = congested_data;
2221 struct request_queue *q;
2225 if (!may_inc_ap_bio(mdev)) {
2226 /* DRBD has frozen IO */
2232 if (get_ldev(mdev)) {
2233 q = bdev_get_queue(mdev->ldev->backing_bdev);
2234 r = bdi_congested(&q->backing_dev_info, bdi_bits);
2240 if (bdi_bits & (1 << BDI_async_congested) && test_bit(NET_CONGESTED, &mdev->tconn->flags)) {
2241 r |= (1 << BDI_async_congested);
2242 reason = reason == 'b' ? 'a' : 'n';
2246 mdev->congestion_reason = reason;
2250 static void drbd_init_workqueue(struct drbd_work_queue* wq)
2252 sema_init(&wq->s, 0);
2253 spin_lock_init(&wq->q_lock);
2254 INIT_LIST_HEAD(&wq->q);
2257 struct drbd_tconn *conn_by_name(const char *name)
2259 struct drbd_tconn *tconn;
2261 if (!name || !name[0])
2264 mutex_lock(&drbd_cfg_mutex);
2265 list_for_each_entry(tconn, &drbd_tconns, all_tconn) {
2266 if (!strcmp(tconn->name, name))
2271 mutex_unlock(&drbd_cfg_mutex);
2275 static int drbd_alloc_socket(struct drbd_socket *socket)
2277 socket->rbuf = (void *) __get_free_page(GFP_KERNEL);
2280 socket->sbuf = (void *) __get_free_page(GFP_KERNEL);
2286 static void drbd_free_socket(struct drbd_socket *socket)
2288 free_page((unsigned long) socket->sbuf);
2289 free_page((unsigned long) socket->rbuf);
2292 struct drbd_tconn *drbd_new_tconn(const char *name)
2294 struct drbd_tconn *tconn;
2296 tconn = kzalloc(sizeof(struct drbd_tconn), GFP_KERNEL);
2300 tconn->name = kstrdup(name, GFP_KERNEL);
2304 if (drbd_alloc_socket(&tconn->data))
2306 if (drbd_alloc_socket(&tconn->meta))
2309 if (!zalloc_cpumask_var(&tconn->cpu_mask, GFP_KERNEL))
2312 if (!tl_init(tconn))
2315 tconn->cstate = C_STANDALONE;
2316 mutex_init(&tconn->cstate_mutex);
2317 spin_lock_init(&tconn->req_lock);
2318 atomic_set(&tconn->net_cnt, 0);
2319 init_waitqueue_head(&tconn->net_cnt_wait);
2320 init_waitqueue_head(&tconn->ping_wait);
2321 idr_init(&tconn->volumes);
2323 drbd_init_workqueue(&tconn->data.work);
2324 mutex_init(&tconn->data.mutex);
2326 drbd_init_workqueue(&tconn->meta.work);
2327 mutex_init(&tconn->meta.mutex);
2329 drbd_thread_init(tconn, &tconn->receiver, drbdd_init, "receiver");
2330 drbd_thread_init(tconn, &tconn->worker, drbd_worker, "worker");
2331 drbd_thread_init(tconn, &tconn->asender, drbd_asender, "asender");
2333 tconn->res_opts = (struct res_opts) {
2334 {}, 0, /* cpu_mask */
2335 DRBD_ON_NO_DATA_DEF, /* on_no_data */
2338 mutex_lock(&drbd_cfg_mutex);
2339 list_add_tail(&tconn->all_tconn, &drbd_tconns);
2340 mutex_unlock(&drbd_cfg_mutex);
2346 free_cpumask_var(tconn->cpu_mask);
2347 drbd_free_socket(&tconn->meta);
2348 drbd_free_socket(&tconn->data);
2355 void drbd_free_tconn(struct drbd_tconn *tconn)
2357 list_del(&tconn->all_tconn);
2358 idr_destroy(&tconn->volumes);
2360 free_cpumask_var(tconn->cpu_mask);
2361 drbd_free_socket(&tconn->meta);
2362 drbd_free_socket(&tconn->data);
2364 kfree(tconn->int_dig_out);
2365 kfree(tconn->int_dig_in);
2366 kfree(tconn->int_dig_vv);
2370 enum drbd_ret_code conn_new_minor(struct drbd_tconn *tconn, unsigned int minor, int vnr)
2372 struct drbd_conf *mdev;
2373 struct gendisk *disk;
2374 struct request_queue *q;
2376 int minor_got = minor;
2377 enum drbd_ret_code err = ERR_NOMEM;
2379 mdev = minor_to_mdev(minor);
2381 return ERR_MINOR_EXISTS;
2383 /* GFP_KERNEL, we are outside of all write-out paths */
2384 mdev = kzalloc(sizeof(struct drbd_conf), GFP_KERNEL);
2388 mdev->tconn = tconn;
2389 mdev->minor = minor;
2392 drbd_init_set_defaults(mdev);
2394 q = blk_alloc_queue(GFP_KERNEL);
2398 q->queuedata = mdev;
2400 disk = alloc_disk(1);
2405 set_disk_ro(disk, true);
2408 disk->major = DRBD_MAJOR;
2409 disk->first_minor = minor;
2410 disk->fops = &drbd_ops;
2411 sprintf(disk->disk_name, "drbd%d", minor);
2412 disk->private_data = mdev;
2414 mdev->this_bdev = bdget(MKDEV(DRBD_MAJOR, minor));
2415 /* we have no partitions. we contain only ourselves. */
2416 mdev->this_bdev->bd_contains = mdev->this_bdev;
2418 q->backing_dev_info.congested_fn = drbd_congested;
2419 q->backing_dev_info.congested_data = mdev;
2421 blk_queue_make_request(q, drbd_make_request);
2422 /* Setting the max_hw_sectors to an odd value of 8kibyte here
2423 This triggers a max_bio_size message upon first attach or connect */
2424 blk_queue_max_hw_sectors(q, DRBD_MAX_BIO_SIZE_SAFE >> 8);
2425 blk_queue_bounce_limit(q, BLK_BOUNCE_ANY);
2426 blk_queue_merge_bvec(q, drbd_merge_bvec);
2427 q->queue_lock = &mdev->tconn->req_lock; /* needed since we use */
2429 mdev->md_io_page = alloc_page(GFP_KERNEL);
2430 if (!mdev->md_io_page)
2431 goto out_no_io_page;
2433 if (drbd_bm_init(mdev))
2435 mdev->read_requests = RB_ROOT;
2436 mdev->write_requests = RB_ROOT;
2438 mdev->current_epoch = kzalloc(sizeof(struct drbd_epoch), GFP_KERNEL);
2439 if (!mdev->current_epoch)
2442 INIT_LIST_HEAD(&mdev->current_epoch->list);
2445 if (!idr_pre_get(&minors, GFP_KERNEL))
2446 goto out_no_minor_idr;
2447 if (idr_get_new_above(&minors, mdev, minor, &minor_got))
2448 goto out_no_minor_idr;
2449 if (minor_got != minor) {
2450 err = ERR_MINOR_EXISTS;
2451 drbd_msg_put_info("requested minor exists already");
2452 goto out_idr_remove_minor;
2455 if (!idr_pre_get(&tconn->volumes, GFP_KERNEL))
2456 goto out_idr_remove_minor;
2457 if (idr_get_new_above(&tconn->volumes, mdev, vnr, &vnr_got))
2458 goto out_idr_remove_minor;
2459 if (vnr_got != vnr) {
2460 err = ERR_INVALID_REQUEST;
2461 drbd_msg_put_info("requested volume exists already");
2462 goto out_idr_remove_vol;
2466 /* inherit the connection state */
2467 mdev->state.conn = tconn->cstate;
2468 if (mdev->state.conn == C_WF_REPORT_PARAMS)
2469 drbd_connected(vnr, mdev, tconn);
2474 idr_remove(&tconn->volumes, vnr_got);
2475 out_idr_remove_minor:
2476 idr_remove(&minors, minor_got);
2479 kfree(mdev->current_epoch);
2481 drbd_bm_cleanup(mdev);
2483 __free_page(mdev->md_io_page);
2487 blk_cleanup_queue(q);
2493 /* counterpart of drbd_new_device.
2494 * last part of drbd_delete_device. */
2495 void drbd_free_mdev(struct drbd_conf *mdev)
2497 kfree(mdev->current_epoch);
2498 if (mdev->bitmap) /* should no longer be there. */
2499 drbd_bm_cleanup(mdev);
2500 __free_page(mdev->md_io_page);
2501 put_disk(mdev->vdisk);
2502 blk_cleanup_queue(mdev->rq_queue);
2507 int __init drbd_init(void)
2511 BUILD_BUG_ON(sizeof(struct p_header80) != sizeof(struct p_header95));
2512 BUILD_BUG_ON(sizeof(struct p_handshake) != 80);
2514 if (minor_count < DRBD_MINOR_COUNT_MIN || minor_count > DRBD_MINOR_COUNT_MAX) {
2516 "drbd: invalid minor_count (%d)\n", minor_count);
2524 err = register_blkdev(DRBD_MAJOR, "drbd");
2527 "drbd: unable to register block device major %d\n",
2532 err = drbd_genl_register();
2534 printk(KERN_ERR "drbd: unable to register generic netlink family\n");
2539 register_reboot_notifier(&drbd_notifier);
2542 * allocate all necessary structs
2546 init_waitqueue_head(&drbd_pp_wait);
2548 drbd_proc = NULL; /* play safe for drbd_cleanup */
2551 err = drbd_create_mempools();
2555 drbd_proc = proc_create_data("drbd", S_IFREG | S_IRUGO , NULL, &drbd_proc_fops, NULL);
2557 printk(KERN_ERR "drbd: unable to register proc file\n");
2561 rwlock_init(&global_state_lock);
2562 INIT_LIST_HEAD(&drbd_tconns);
2564 printk(KERN_INFO "drbd: initialized. "
2565 "Version: " REL_VERSION " (api:%d/proto:%d-%d)\n",
2566 API_VERSION, PRO_VERSION_MIN, PRO_VERSION_MAX);
2567 printk(KERN_INFO "drbd: %s\n", drbd_buildtag());
2568 printk(KERN_INFO "drbd: registered as block device major %d\n",
2571 return 0; /* Success! */
2576 /* currently always the case */
2577 printk(KERN_ERR "drbd: ran out of memory\n");
2579 printk(KERN_ERR "drbd: initialization failure\n");
2583 void drbd_free_bc(struct drbd_backing_dev *ldev)
2588 blkdev_put(ldev->backing_bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
2589 blkdev_put(ldev->md_bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
2594 void drbd_free_sock(struct drbd_tconn *tconn)
2596 if (tconn->data.socket) {
2597 mutex_lock(&tconn->data.mutex);
2598 kernel_sock_shutdown(tconn->data.socket, SHUT_RDWR);
2599 sock_release(tconn->data.socket);
2600 tconn->data.socket = NULL;
2601 mutex_unlock(&tconn->data.mutex);
2603 if (tconn->meta.socket) {
2604 mutex_lock(&tconn->meta.mutex);
2605 kernel_sock_shutdown(tconn->meta.socket, SHUT_RDWR);
2606 sock_release(tconn->meta.socket);
2607 tconn->meta.socket = NULL;
2608 mutex_unlock(&tconn->meta.mutex);
2613 void drbd_free_resources(struct drbd_conf *mdev)
2615 crypto_free_hash(mdev->tconn->csums_tfm);
2616 mdev->tconn->csums_tfm = NULL;
2617 crypto_free_hash(mdev->tconn->verify_tfm);
2618 mdev->tconn->verify_tfm = NULL;
2619 crypto_free_hash(mdev->tconn->cram_hmac_tfm);
2620 mdev->tconn->cram_hmac_tfm = NULL;
2621 crypto_free_hash(mdev->tconn->integrity_w_tfm);
2622 mdev->tconn->integrity_w_tfm = NULL;
2623 crypto_free_hash(mdev->tconn->integrity_r_tfm);
2624 mdev->tconn->integrity_r_tfm = NULL;
2626 drbd_free_sock(mdev->tconn);
2629 drbd_free_bc(mdev->ldev);
2630 mdev->ldev = NULL;);
2633 /* meta data management */
2635 struct meta_data_on_disk {
2636 u64 la_size; /* last agreed size. */
2637 u64 uuid[UI_SIZE]; /* UUIDs. */
2640 u32 flags; /* MDF */
2643 u32 al_offset; /* offset to this block */
2644 u32 al_nr_extents; /* important for restoring the AL */
2645 /* `-- act_log->nr_elements <-- ldev->dc.al_extents */
2646 u32 bm_offset; /* offset to the bitmap, from here */
2647 u32 bm_bytes_per_bit; /* BM_BLOCK_SIZE */
2648 u32 la_peer_max_bio_size; /* last peer max_bio_size */
2649 u32 reserved_u32[3];
2654 * drbd_md_sync() - Writes the meta data super block if the MD_DIRTY flag bit is set
2655 * @mdev: DRBD device.
2657 void drbd_md_sync(struct drbd_conf *mdev)
2659 struct meta_data_on_disk *buffer;
2663 del_timer(&mdev->md_sync_timer);
2664 /* timer may be rearmed by drbd_md_mark_dirty() now. */
2665 if (!test_and_clear_bit(MD_DIRTY, &mdev->flags))
2668 /* We use here D_FAILED and not D_ATTACHING because we try to write
2669 * metadata even if we detach due to a disk failure! */
2670 if (!get_ldev_if_state(mdev, D_FAILED))
2673 mutex_lock(&mdev->md_io_mutex);
2674 buffer = (struct meta_data_on_disk *)page_address(mdev->md_io_page);
2675 memset(buffer, 0, 512);
2677 buffer->la_size = cpu_to_be64(drbd_get_capacity(mdev->this_bdev));
2678 for (i = UI_CURRENT; i < UI_SIZE; i++)
2679 buffer->uuid[i] = cpu_to_be64(mdev->ldev->md.uuid[i]);
2680 buffer->flags = cpu_to_be32(mdev->ldev->md.flags);
2681 buffer->magic = cpu_to_be32(DRBD_MD_MAGIC);
2683 buffer->md_size_sect = cpu_to_be32(mdev->ldev->md.md_size_sect);
2684 buffer->al_offset = cpu_to_be32(mdev->ldev->md.al_offset);
2685 buffer->al_nr_extents = cpu_to_be32(mdev->act_log->nr_elements);
2686 buffer->bm_bytes_per_bit = cpu_to_be32(BM_BLOCK_SIZE);
2687 buffer->device_uuid = cpu_to_be64(mdev->ldev->md.device_uuid);
2689 buffer->bm_offset = cpu_to_be32(mdev->ldev->md.bm_offset);
2690 buffer->la_peer_max_bio_size = cpu_to_be32(mdev->peer_max_bio_size);
2692 D_ASSERT(drbd_md_ss__(mdev, mdev->ldev) == mdev->ldev->md.md_offset);
2693 sector = mdev->ldev->md.md_offset;
2695 if (drbd_md_sync_page_io(mdev, mdev->ldev, sector, WRITE)) {
2696 /* this was a try anyways ... */
2697 dev_err(DEV, "meta data update failed!\n");
2698 drbd_chk_io_error(mdev, 1, true);
2701 /* Update mdev->ldev->md.la_size_sect,
2702 * since we updated it on metadata. */
2703 mdev->ldev->md.la_size_sect = drbd_get_capacity(mdev->this_bdev);
2705 mutex_unlock(&mdev->md_io_mutex);
2710 * drbd_md_read() - Reads in the meta data super block
2711 * @mdev: DRBD device.
2712 * @bdev: Device from which the meta data should be read in.
2714 * Return 0 (NO_ERROR) on success, and an enum drbd_ret_code in case
2715 * something goes wrong. Currently only: ERR_IO_MD_DISK, ERR_MD_INVALID.
2717 int drbd_md_read(struct drbd_conf *mdev, struct drbd_backing_dev *bdev)
2719 struct meta_data_on_disk *buffer;
2720 int i, rv = NO_ERROR;
2722 if (!get_ldev_if_state(mdev, D_ATTACHING))
2723 return ERR_IO_MD_DISK;
2725 mutex_lock(&mdev->md_io_mutex);
2726 buffer = (struct meta_data_on_disk *)page_address(mdev->md_io_page);
2728 if (drbd_md_sync_page_io(mdev, bdev, bdev->md.md_offset, READ)) {
2729 /* NOTE: can't do normal error processing here as this is
2730 called BEFORE disk is attached */
2731 dev_err(DEV, "Error while reading metadata.\n");
2732 rv = ERR_IO_MD_DISK;
2736 if (buffer->magic != cpu_to_be32(DRBD_MD_MAGIC)) {
2737 dev_err(DEV, "Error while reading metadata, magic not found.\n");
2738 rv = ERR_MD_INVALID;
2741 if (be32_to_cpu(buffer->al_offset) != bdev->md.al_offset) {
2742 dev_err(DEV, "unexpected al_offset: %d (expected %d)\n",
2743 be32_to_cpu(buffer->al_offset), bdev->md.al_offset);
2744 rv = ERR_MD_INVALID;
2747 if (be32_to_cpu(buffer->bm_offset) != bdev->md.bm_offset) {
2748 dev_err(DEV, "unexpected bm_offset: %d (expected %d)\n",
2749 be32_to_cpu(buffer->bm_offset), bdev->md.bm_offset);
2750 rv = ERR_MD_INVALID;
2753 if (be32_to_cpu(buffer->md_size_sect) != bdev->md.md_size_sect) {
2754 dev_err(DEV, "unexpected md_size: %u (expected %u)\n",
2755 be32_to_cpu(buffer->md_size_sect), bdev->md.md_size_sect);
2756 rv = ERR_MD_INVALID;
2760 if (be32_to_cpu(buffer->bm_bytes_per_bit) != BM_BLOCK_SIZE) {
2761 dev_err(DEV, "unexpected bm_bytes_per_bit: %u (expected %u)\n",
2762 be32_to_cpu(buffer->bm_bytes_per_bit), BM_BLOCK_SIZE);
2763 rv = ERR_MD_INVALID;
2767 bdev->md.la_size_sect = be64_to_cpu(buffer->la_size);
2768 for (i = UI_CURRENT; i < UI_SIZE; i++)
2769 bdev->md.uuid[i] = be64_to_cpu(buffer->uuid[i]);
2770 bdev->md.flags = be32_to_cpu(buffer->flags);
2771 bdev->dc.al_extents = be32_to_cpu(buffer->al_nr_extents);
2772 bdev->md.device_uuid = be64_to_cpu(buffer->device_uuid);
2774 spin_lock_irq(&mdev->tconn->req_lock);
2775 if (mdev->state.conn < C_CONNECTED) {
2777 peer = be32_to_cpu(buffer->la_peer_max_bio_size);
2778 peer = max_t(int, peer, DRBD_MAX_BIO_SIZE_SAFE);
2779 mdev->peer_max_bio_size = peer;
2781 spin_unlock_irq(&mdev->tconn->req_lock);
2783 if (bdev->dc.al_extents < 7)
2784 bdev->dc.al_extents = 127;
2787 mutex_unlock(&mdev->md_io_mutex);
2794 * drbd_md_mark_dirty() - Mark meta data super block as dirty
2795 * @mdev: DRBD device.
2797 * Call this function if you change anything that should be written to
2798 * the meta-data super block. This function sets MD_DIRTY, and starts a
2799 * timer that ensures that within five seconds you have to call drbd_md_sync().
2802 void drbd_md_mark_dirty_(struct drbd_conf *mdev, unsigned int line, const char *func)
2804 if (!test_and_set_bit(MD_DIRTY, &mdev->flags)) {
2805 mod_timer(&mdev->md_sync_timer, jiffies + HZ);
2806 mdev->last_md_mark_dirty.line = line;
2807 mdev->last_md_mark_dirty.func = func;
2811 void drbd_md_mark_dirty(struct drbd_conf *mdev)
2813 if (!test_and_set_bit(MD_DIRTY, &mdev->flags))
2814 mod_timer(&mdev->md_sync_timer, jiffies + 5*HZ);
2818 static void drbd_uuid_move_history(struct drbd_conf *mdev) __must_hold(local)
2822 for (i = UI_HISTORY_START; i < UI_HISTORY_END; i++)
2823 mdev->ldev->md.uuid[i+1] = mdev->ldev->md.uuid[i];
2826 void _drbd_uuid_set(struct drbd_conf *mdev, int idx, u64 val) __must_hold(local)
2828 if (idx == UI_CURRENT) {
2829 if (mdev->state.role == R_PRIMARY)
2834 drbd_set_ed_uuid(mdev, val);
2837 mdev->ldev->md.uuid[idx] = val;
2838 drbd_md_mark_dirty(mdev);
2842 void drbd_uuid_set(struct drbd_conf *mdev, int idx, u64 val) __must_hold(local)
2844 if (mdev->ldev->md.uuid[idx]) {
2845 drbd_uuid_move_history(mdev);
2846 mdev->ldev->md.uuid[UI_HISTORY_START] = mdev->ldev->md.uuid[idx];
2848 _drbd_uuid_set(mdev, idx, val);
2852 * drbd_uuid_new_current() - Creates a new current UUID
2853 * @mdev: DRBD device.
2855 * Creates a new current UUID, and rotates the old current UUID into
2856 * the bitmap slot. Causes an incremental resync upon next connect.
2858 void drbd_uuid_new_current(struct drbd_conf *mdev) __must_hold(local)
2861 unsigned long long bm_uuid = mdev->ldev->md.uuid[UI_BITMAP];
2864 dev_warn(DEV, "bm UUID was already set: %llX\n", bm_uuid);
2866 mdev->ldev->md.uuid[UI_BITMAP] = mdev->ldev->md.uuid[UI_CURRENT];
2868 get_random_bytes(&val, sizeof(u64));
2869 _drbd_uuid_set(mdev, UI_CURRENT, val);
2870 drbd_print_uuids(mdev, "new current UUID");
2871 /* get it to stable storage _now_ */
2875 void drbd_uuid_set_bm(struct drbd_conf *mdev, u64 val) __must_hold(local)
2877 if (mdev->ldev->md.uuid[UI_BITMAP] == 0 && val == 0)
2881 drbd_uuid_move_history(mdev);
2882 mdev->ldev->md.uuid[UI_HISTORY_START] = mdev->ldev->md.uuid[UI_BITMAP];
2883 mdev->ldev->md.uuid[UI_BITMAP] = 0;
2885 unsigned long long bm_uuid = mdev->ldev->md.uuid[UI_BITMAP];
2887 dev_warn(DEV, "bm UUID was already set: %llX\n", bm_uuid);
2889 mdev->ldev->md.uuid[UI_BITMAP] = val & ~((u64)1);
2891 drbd_md_mark_dirty(mdev);
2895 * drbd_bmio_set_n_write() - io_fn for drbd_queue_bitmap_io() or drbd_bitmap_io()
2896 * @mdev: DRBD device.
2898 * Sets all bits in the bitmap and writes the whole bitmap to stable storage.
2900 int drbd_bmio_set_n_write(struct drbd_conf *mdev)
2904 if (get_ldev_if_state(mdev, D_ATTACHING)) {
2905 drbd_md_set_flag(mdev, MDF_FULL_SYNC);
2907 drbd_bm_set_all(mdev);
2909 rv = drbd_bm_write(mdev);
2912 drbd_md_clear_flag(mdev, MDF_FULL_SYNC);
2923 * drbd_bmio_clear_n_write() - io_fn for drbd_queue_bitmap_io() or drbd_bitmap_io()
2924 * @mdev: DRBD device.
2926 * Clears all bits in the bitmap and writes the whole bitmap to stable storage.
2928 int drbd_bmio_clear_n_write(struct drbd_conf *mdev)
2932 drbd_resume_al(mdev);
2933 if (get_ldev_if_state(mdev, D_ATTACHING)) {
2934 drbd_bm_clear_all(mdev);
2935 rv = drbd_bm_write(mdev);
2942 static int w_bitmap_io(struct drbd_work *w, int unused)
2944 struct bm_io_work *work = container_of(w, struct bm_io_work, w);
2945 struct drbd_conf *mdev = w->mdev;
2948 D_ASSERT(atomic_read(&mdev->ap_bio_cnt) == 0);
2950 if (get_ldev(mdev)) {
2951 drbd_bm_lock(mdev, work->why, work->flags);
2952 rv = work->io_fn(mdev);
2953 drbd_bm_unlock(mdev);
2957 clear_bit_unlock(BITMAP_IO, &mdev->flags);
2958 wake_up(&mdev->misc_wait);
2961 work->done(mdev, rv);
2963 clear_bit(BITMAP_IO_QUEUED, &mdev->flags);
2970 void drbd_ldev_destroy(struct drbd_conf *mdev)
2972 lc_destroy(mdev->resync);
2973 mdev->resync = NULL;
2974 lc_destroy(mdev->act_log);
2975 mdev->act_log = NULL;
2977 drbd_free_bc(mdev->ldev);
2978 mdev->ldev = NULL;);
2980 clear_bit(GO_DISKLESS, &mdev->flags);
2983 static int w_go_diskless(struct drbd_work *w, int unused)
2985 struct drbd_conf *mdev = w->mdev;
2987 D_ASSERT(mdev->state.disk == D_FAILED);
2988 /* we cannot assert local_cnt == 0 here, as get_ldev_if_state will
2989 * inc/dec it frequently. Once we are D_DISKLESS, no one will touch
2990 * the protected members anymore, though, so once put_ldev reaches zero
2991 * again, it will be safe to free them. */
2992 drbd_force_state(mdev, NS(disk, D_DISKLESS));
2996 void drbd_go_diskless(struct drbd_conf *mdev)
2998 D_ASSERT(mdev->state.disk == D_FAILED);
2999 if (!test_and_set_bit(GO_DISKLESS, &mdev->flags))
3000 drbd_queue_work(&mdev->tconn->data.work, &mdev->go_diskless);
3004 * drbd_queue_bitmap_io() - Queues an IO operation on the whole bitmap
3005 * @mdev: DRBD device.
3006 * @io_fn: IO callback to be called when bitmap IO is possible
3007 * @done: callback to be called after the bitmap IO was performed
3008 * @why: Descriptive text of the reason for doing the IO
3010 * While IO on the bitmap happens we freeze application IO thus we ensure
3011 * that drbd_set_out_of_sync() can not be called. This function MAY ONLY be
3012 * called from worker context. It MUST NOT be used while a previous such
3013 * work is still pending!
3015 void drbd_queue_bitmap_io(struct drbd_conf *mdev,
3016 int (*io_fn)(struct drbd_conf *),
3017 void (*done)(struct drbd_conf *, int),
3018 char *why, enum bm_flag flags)
3020 D_ASSERT(current == mdev->tconn->worker.task);
3022 D_ASSERT(!test_bit(BITMAP_IO_QUEUED, &mdev->flags));
3023 D_ASSERT(!test_bit(BITMAP_IO, &mdev->flags));
3024 D_ASSERT(list_empty(&mdev->bm_io_work.w.list));
3025 if (mdev->bm_io_work.why)
3026 dev_err(DEV, "FIXME going to queue '%s' but '%s' still pending?\n",
3027 why, mdev->bm_io_work.why);
3029 mdev->bm_io_work.io_fn = io_fn;
3030 mdev->bm_io_work.done = done;
3031 mdev->bm_io_work.why = why;
3032 mdev->bm_io_work.flags = flags;
3034 spin_lock_irq(&mdev->tconn->req_lock);
3035 set_bit(BITMAP_IO, &mdev->flags);
3036 if (atomic_read(&mdev->ap_bio_cnt) == 0) {
3037 if (!test_and_set_bit(BITMAP_IO_QUEUED, &mdev->flags))
3038 drbd_queue_work(&mdev->tconn->data.work, &mdev->bm_io_work.w);
3040 spin_unlock_irq(&mdev->tconn->req_lock);
3044 * drbd_bitmap_io() - Does an IO operation on the whole bitmap
3045 * @mdev: DRBD device.
3046 * @io_fn: IO callback to be called when bitmap IO is possible
3047 * @why: Descriptive text of the reason for doing the IO
3049 * freezes application IO while that the actual IO operations runs. This
3050 * functions MAY NOT be called from worker context.
3052 int drbd_bitmap_io(struct drbd_conf *mdev, int (*io_fn)(struct drbd_conf *),
3053 char *why, enum bm_flag flags)
3057 D_ASSERT(current != mdev->tconn->worker.task);
3059 if ((flags & BM_LOCKED_SET_ALLOWED) == 0)
3060 drbd_suspend_io(mdev);
3062 drbd_bm_lock(mdev, why, flags);
3064 drbd_bm_unlock(mdev);
3066 if ((flags & BM_LOCKED_SET_ALLOWED) == 0)
3067 drbd_resume_io(mdev);
3072 void drbd_md_set_flag(struct drbd_conf *mdev, int flag) __must_hold(local)
3074 if ((mdev->ldev->md.flags & flag) != flag) {
3075 drbd_md_mark_dirty(mdev);
3076 mdev->ldev->md.flags |= flag;
3080 void drbd_md_clear_flag(struct drbd_conf *mdev, int flag) __must_hold(local)
3082 if ((mdev->ldev->md.flags & flag) != 0) {
3083 drbd_md_mark_dirty(mdev);
3084 mdev->ldev->md.flags &= ~flag;
3087 int drbd_md_test_flag(struct drbd_backing_dev *bdev, int flag)
3089 return (bdev->md.flags & flag) != 0;
3092 static void md_sync_timer_fn(unsigned long data)
3094 struct drbd_conf *mdev = (struct drbd_conf *) data;
3096 drbd_queue_work_front(&mdev->tconn->data.work, &mdev->md_sync_work);
3099 static int w_md_sync(struct drbd_work *w, int unused)
3101 struct drbd_conf *mdev = w->mdev;
3103 dev_warn(DEV, "md_sync_timer expired! Worker calls drbd_md_sync().\n");
3105 dev_warn(DEV, "last md_mark_dirty: %s:%u\n",
3106 mdev->last_md_mark_dirty.func, mdev->last_md_mark_dirty.line);
3112 const char *cmdname(enum drbd_packet cmd)
3114 /* THINK may need to become several global tables
3115 * when we want to support more than
3116 * one PRO_VERSION */
3117 static const char *cmdnames[] = {
3119 [P_DATA_REPLY] = "DataReply",
3120 [P_RS_DATA_REPLY] = "RSDataReply",
3121 [P_BARRIER] = "Barrier",
3122 [P_BITMAP] = "ReportBitMap",
3123 [P_BECOME_SYNC_TARGET] = "BecomeSyncTarget",
3124 [P_BECOME_SYNC_SOURCE] = "BecomeSyncSource",
3125 [P_UNPLUG_REMOTE] = "UnplugRemote",
3126 [P_DATA_REQUEST] = "DataRequest",
3127 [P_RS_DATA_REQUEST] = "RSDataRequest",
3128 [P_SYNC_PARAM] = "SyncParam",
3129 [P_SYNC_PARAM89] = "SyncParam89",
3130 [P_PROTOCOL] = "ReportProtocol",
3131 [P_UUIDS] = "ReportUUIDs",
3132 [P_SIZES] = "ReportSizes",
3133 [P_STATE] = "ReportState",
3134 [P_SYNC_UUID] = "ReportSyncUUID",
3135 [P_AUTH_CHALLENGE] = "AuthChallenge",
3136 [P_AUTH_RESPONSE] = "AuthResponse",
3138 [P_PING_ACK] = "PingAck",
3139 [P_RECV_ACK] = "RecvAck",
3140 [P_WRITE_ACK] = "WriteAck",
3141 [P_RS_WRITE_ACK] = "RSWriteAck",
3142 [P_DISCARD_WRITE] = "DiscardWrite",
3143 [P_NEG_ACK] = "NegAck",
3144 [P_NEG_DREPLY] = "NegDReply",
3145 [P_NEG_RS_DREPLY] = "NegRSDReply",
3146 [P_BARRIER_ACK] = "BarrierAck",
3147 [P_STATE_CHG_REQ] = "StateChgRequest",
3148 [P_STATE_CHG_REPLY] = "StateChgReply",
3149 [P_OV_REQUEST] = "OVRequest",
3150 [P_OV_REPLY] = "OVReply",
3151 [P_OV_RESULT] = "OVResult",
3152 [P_CSUM_RS_REQUEST] = "CsumRSRequest",
3153 [P_RS_IS_IN_SYNC] = "CsumRSIsInSync",
3154 [P_COMPRESSED_BITMAP] = "CBitmap",
3155 [P_DELAY_PROBE] = "DelayProbe",
3156 [P_OUT_OF_SYNC] = "OutOfSync",
3157 [P_RETRY_WRITE] = "RetryWrite",
3160 if (cmd == P_HAND_SHAKE_M)
3161 return "HandShakeM";
3162 if (cmd == P_HAND_SHAKE_S)
3163 return "HandShakeS";
3164 if (cmd == P_HAND_SHAKE)
3166 if (cmd >= ARRAY_SIZE(cmdnames))
3168 return cmdnames[cmd];
3172 * drbd_wait_misc - wait for a request to make progress
3173 * @mdev: device associated with the request
3174 * @i: the struct drbd_interval embedded in struct drbd_request or
3175 * struct drbd_peer_request
3177 int drbd_wait_misc(struct drbd_conf *mdev, struct drbd_interval *i)
3179 struct net_conf *net_conf = mdev->tconn->net_conf;
3185 timeout = MAX_SCHEDULE_TIMEOUT;
3186 if (net_conf->ko_count)
3187 timeout = net_conf->timeout * HZ / 10 * net_conf->ko_count;
3189 /* Indicate to wake up mdev->misc_wait on progress. */
3191 prepare_to_wait(&mdev->misc_wait, &wait, TASK_INTERRUPTIBLE);
3192 spin_unlock_irq(&mdev->tconn->req_lock);
3193 timeout = schedule_timeout(timeout);
3194 finish_wait(&mdev->misc_wait, &wait);
3195 spin_lock_irq(&mdev->tconn->req_lock);
3196 if (!timeout || mdev->state.conn < C_CONNECTED)
3198 if (signal_pending(current))
3199 return -ERESTARTSYS;
3203 #ifdef CONFIG_DRBD_FAULT_INJECTION
3204 /* Fault insertion support including random number generator shamelessly
3205 * stolen from kernel/rcutorture.c */
3206 struct fault_random_state {
3207 unsigned long state;
3208 unsigned long count;
3211 #define FAULT_RANDOM_MULT 39916801 /* prime */
3212 #define FAULT_RANDOM_ADD 479001701 /* prime */
3213 #define FAULT_RANDOM_REFRESH 10000
3216 * Crude but fast random-number generator. Uses a linear congruential
3217 * generator, with occasional help from get_random_bytes().
3219 static unsigned long
3220 _drbd_fault_random(struct fault_random_state *rsp)
3224 if (!rsp->count--) {
3225 get_random_bytes(&refresh, sizeof(refresh));
3226 rsp->state += refresh;
3227 rsp->count = FAULT_RANDOM_REFRESH;
3229 rsp->state = rsp->state * FAULT_RANDOM_MULT + FAULT_RANDOM_ADD;
3230 return swahw32(rsp->state);
3234 _drbd_fault_str(unsigned int type) {
3235 static char *_faults[] = {
3236 [DRBD_FAULT_MD_WR] = "Meta-data write",
3237 [DRBD_FAULT_MD_RD] = "Meta-data read",
3238 [DRBD_FAULT_RS_WR] = "Resync write",
3239 [DRBD_FAULT_RS_RD] = "Resync read",
3240 [DRBD_FAULT_DT_WR] = "Data write",
3241 [DRBD_FAULT_DT_RD] = "Data read",
3242 [DRBD_FAULT_DT_RA] = "Data read ahead",
3243 [DRBD_FAULT_BM_ALLOC] = "BM allocation",
3244 [DRBD_FAULT_AL_EE] = "EE allocation",
3245 [DRBD_FAULT_RECEIVE] = "receive data corruption",
3248 return (type < DRBD_FAULT_MAX) ? _faults[type] : "**Unknown**";
3252 _drbd_insert_fault(struct drbd_conf *mdev, unsigned int type)
3254 static struct fault_random_state rrs = {0, 0};
3256 unsigned int ret = (
3258 ((1 << mdev_to_minor(mdev)) & fault_devs) != 0) &&
3259 (((_drbd_fault_random(&rrs) % 100) + 1) <= fault_rate));
3264 if (__ratelimit(&drbd_ratelimit_state))
3265 dev_warn(DEV, "***Simulating %s failure\n",
3266 _drbd_fault_str(type));
3273 const char *drbd_buildtag(void)
3275 /* DRBD built from external sources has here a reference to the
3276 git hash of the source code. */
3278 static char buildtag[38] = "\0uilt-in";
3280 if (buildtag[0] == 0) {
3281 #ifdef CONFIG_MODULES
3282 if (THIS_MODULE != NULL)
3283 sprintf(buildtag, "srcversion: %-24s", THIS_MODULE->srcversion);
3292 module_init(drbd_init)
3293 module_exit(drbd_cleanup)
3295 EXPORT_SYMBOL(drbd_conn_str);
3296 EXPORT_SYMBOL(drbd_role_str);
3297 EXPORT_SYMBOL(drbd_disk_str);
3298 EXPORT_SYMBOL(drbd_set_st_err_str);