blob: 1cfcc44fd4841d60578e80750e1ecfd74920b538 [file] [log] [blame]
Philipp Reisnerb411b362009-09-25 16:07:19 -07001/*
2 drbd_receiver.c
3
4 This file is part of DRBD by Philipp Reisner and Lars Ellenberg.
5
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>.
9
10 drbd is free software; you can redistribute it and/or modify
11 it under the terms of the GNU General Public License as published by
12 the Free Software Foundation; either version 2, or (at your option)
13 any later version.
14
15 drbd is distributed in the hope that it will be useful,
16 but WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 GNU General Public License for more details.
19
20 You should have received a copy of the GNU General Public License
21 along with drbd; see the file COPYING. If not, write to
22 the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
23 */
24
25
Philipp Reisnerb411b362009-09-25 16:07:19 -070026#include <linux/module.h>
27
28#include <asm/uaccess.h>
29#include <net/sock.h>
30
Philipp Reisnerb411b362009-09-25 16:07:19 -070031#include <linux/drbd.h>
32#include <linux/fs.h>
33#include <linux/file.h>
34#include <linux/in.h>
35#include <linux/mm.h>
36#include <linux/memcontrol.h>
37#include <linux/mm_inline.h>
38#include <linux/slab.h>
Philipp Reisnerb411b362009-09-25 16:07:19 -070039#include <linux/pkt_sched.h>
40#define __KERNEL_SYSCALLS__
41#include <linux/unistd.h>
42#include <linux/vmalloc.h>
43#include <linux/random.h>
Philipp Reisnerb411b362009-09-25 16:07:19 -070044#include <linux/string.h>
45#include <linux/scatterlist.h>
46#include "drbd_int.h"
Philipp Reisnerb411b362009-09-25 16:07:19 -070047#include "drbd_req.h"
48
49#include "drbd_vli.h"
50
Philipp Reisnerb411b362009-09-25 16:07:19 -070051enum finish_epoch {
52 FE_STILL_LIVE,
53 FE_DESTROYED,
54 FE_RECYCLED,
55};
56
57static int drbd_do_handshake(struct drbd_conf *mdev);
58static int drbd_do_auth(struct drbd_conf *mdev);
59
60static enum finish_epoch drbd_may_finish_epoch(struct drbd_conf *, struct drbd_epoch *, enum epoch_event);
61static int e_end_block(struct drbd_conf *, struct drbd_work *, int);
62
Philipp Reisnerb411b362009-09-25 16:07:19 -070063
64#define GFP_TRY (__GFP_HIGHMEM | __GFP_NOWARN)
65
Lars Ellenberg45bb9122010-05-14 17:10:48 +020066/*
67 * some helper functions to deal with single linked page lists,
68 * page->private being our "next" pointer.
69 */
70
71/* If at least n pages are linked at head, get n pages off.
72 * Otherwise, don't modify head, and return NULL.
73 * Locking is the responsibility of the caller.
74 */
75static struct page *page_chain_del(struct page **head, int n)
76{
77 struct page *page;
78 struct page *tmp;
79
80 BUG_ON(!n);
81 BUG_ON(!head);
82
83 page = *head;
Philipp Reisner23ce4222010-05-20 13:35:31 +020084
85 if (!page)
86 return NULL;
87
Lars Ellenberg45bb9122010-05-14 17:10:48 +020088 while (page) {
89 tmp = page_chain_next(page);
90 if (--n == 0)
91 break; /* found sufficient pages */
92 if (tmp == NULL)
93 /* insufficient pages, don't use any of them. */
94 return NULL;
95 page = tmp;
96 }
97
98 /* add end of list marker for the returned list */
99 set_page_private(page, 0);
100 /* actual return value, and adjustment of head */
101 page = *head;
102 *head = tmp;
103 return page;
104}
105
106/* may be used outside of locks to find the tail of a (usually short)
107 * "private" page chain, before adding it back to a global chain head
108 * with page_chain_add() under a spinlock. */
109static struct page *page_chain_tail(struct page *page, int *len)
110{
111 struct page *tmp;
112 int i = 1;
113 while ((tmp = page_chain_next(page)))
114 ++i, page = tmp;
115 if (len)
116 *len = i;
117 return page;
118}
119
120static int page_chain_free(struct page *page)
121{
122 struct page *tmp;
123 int i = 0;
124 page_chain_for_each_safe(page, tmp) {
125 put_page(page);
126 ++i;
127 }
128 return i;
129}
130
131static void page_chain_add(struct page **head,
132 struct page *chain_first, struct page *chain_last)
133{
134#if 1
135 struct page *tmp;
136 tmp = page_chain_tail(chain_first, NULL);
137 BUG_ON(tmp != chain_last);
138#endif
139
140 /* add chain to head */
141 set_page_private(chain_last, (unsigned long)*head);
142 *head = chain_first;
143}
144
145static struct page *drbd_pp_first_pages_or_try_alloc(struct drbd_conf *mdev, int number)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700146{
147 struct page *page = NULL;
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200148 struct page *tmp = NULL;
149 int i = 0;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700150
151 /* Yes, testing drbd_pp_vacant outside the lock is racy.
152 * So what. It saves a spin_lock. */
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200153 if (drbd_pp_vacant >= number) {
Philipp Reisnerb411b362009-09-25 16:07:19 -0700154 spin_lock(&drbd_pp_lock);
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200155 page = page_chain_del(&drbd_pp_pool, number);
156 if (page)
157 drbd_pp_vacant -= number;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700158 spin_unlock(&drbd_pp_lock);
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200159 if (page)
160 return page;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700161 }
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200162
Philipp Reisnerb411b362009-09-25 16:07:19 -0700163 /* GFP_TRY, because we must not cause arbitrary write-out: in a DRBD
164 * "criss-cross" setup, that might cause write-out on some other DRBD,
165 * which in turn might block on the other node at this very place. */
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200166 for (i = 0; i < number; i++) {
167 tmp = alloc_page(GFP_TRY);
168 if (!tmp)
169 break;
170 set_page_private(tmp, (unsigned long)page);
171 page = tmp;
172 }
173
174 if (i == number)
175 return page;
176
177 /* Not enough pages immediately available this time.
178 * No need to jump around here, drbd_pp_alloc will retry this
179 * function "soon". */
180 if (page) {
181 tmp = page_chain_tail(page, NULL);
182 spin_lock(&drbd_pp_lock);
183 page_chain_add(&drbd_pp_pool, page, tmp);
184 drbd_pp_vacant += i;
185 spin_unlock(&drbd_pp_lock);
186 }
187 return NULL;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700188}
189
Philipp Reisnerb411b362009-09-25 16:07:19 -0700190static void reclaim_net_ee(struct drbd_conf *mdev, struct list_head *to_be_freed)
191{
192 struct drbd_epoch_entry *e;
193 struct list_head *le, *tle;
194
195 /* The EEs are always appended to the end of the list. Since
196 they are sent in order over the wire, they have to finish
197 in order. As soon as we see the first not finished we can
198 stop to examine the list... */
199
200 list_for_each_safe(le, tle, &mdev->net_ee) {
201 e = list_entry(le, struct drbd_epoch_entry, w.list);
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200202 if (drbd_ee_has_active_page(e))
Philipp Reisnerb411b362009-09-25 16:07:19 -0700203 break;
204 list_move(le, to_be_freed);
205 }
206}
207
208static void drbd_kick_lo_and_reclaim_net(struct drbd_conf *mdev)
209{
210 LIST_HEAD(reclaimed);
211 struct drbd_epoch_entry *e, *t;
212
Philipp Reisnerb411b362009-09-25 16:07:19 -0700213 spin_lock_irq(&mdev->req_lock);
214 reclaim_net_ee(mdev, &reclaimed);
215 spin_unlock_irq(&mdev->req_lock);
216
217 list_for_each_entry_safe(e, t, &reclaimed, w.list)
Lars Ellenberg435f0742010-09-06 12:30:25 +0200218 drbd_free_net_ee(mdev, e);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700219}
220
221/**
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200222 * drbd_pp_alloc() - Returns @number pages, retries forever (or until signalled)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700223 * @mdev: DRBD device.
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200224 * @number: number of pages requested
225 * @retry: whether to retry, if not enough pages are available right now
Philipp Reisnerb411b362009-09-25 16:07:19 -0700226 *
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200227 * Tries to allocate number pages, first from our own page pool, then from
228 * the kernel, unless this allocation would exceed the max_buffers setting.
229 * Possibly retry until DRBD frees sufficient pages somewhere else.
230 *
231 * Returns a page chain linked via page->private.
Philipp Reisnerb411b362009-09-25 16:07:19 -0700232 */
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200233static struct page *drbd_pp_alloc(struct drbd_conf *mdev, unsigned number, bool retry)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700234{
235 struct page *page = NULL;
236 DEFINE_WAIT(wait);
237
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200238 /* Yes, we may run up to @number over max_buffers. If we
239 * follow it strictly, the admin will get it wrong anyways. */
240 if (atomic_read(&mdev->pp_in_use) < mdev->net_conf->max_buffers)
241 page = drbd_pp_first_pages_or_try_alloc(mdev, number);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700242
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200243 while (page == NULL) {
Philipp Reisnerb411b362009-09-25 16:07:19 -0700244 prepare_to_wait(&drbd_pp_wait, &wait, TASK_INTERRUPTIBLE);
245
246 drbd_kick_lo_and_reclaim_net(mdev);
247
248 if (atomic_read(&mdev->pp_in_use) < mdev->net_conf->max_buffers) {
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200249 page = drbd_pp_first_pages_or_try_alloc(mdev, number);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700250 if (page)
251 break;
252 }
253
254 if (!retry)
255 break;
256
257 if (signal_pending(current)) {
258 dev_warn(DEV, "drbd_pp_alloc interrupted!\n");
259 break;
260 }
261
262 schedule();
263 }
264 finish_wait(&drbd_pp_wait, &wait);
265
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200266 if (page)
267 atomic_add(number, &mdev->pp_in_use);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700268 return page;
269}
270
271/* Must not be used from irq, as that may deadlock: see drbd_pp_alloc.
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200272 * Is also used from inside an other spin_lock_irq(&mdev->req_lock);
273 * Either links the page chain back to the global pool,
274 * or returns all pages to the system. */
Lars Ellenberg435f0742010-09-06 12:30:25 +0200275static void drbd_pp_free(struct drbd_conf *mdev, struct page *page, int is_net)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700276{
Lars Ellenberg435f0742010-09-06 12:30:25 +0200277 atomic_t *a = is_net ? &mdev->pp_in_use_by_net : &mdev->pp_in_use;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700278 int i;
Lars Ellenberg435f0742010-09-06 12:30:25 +0200279
Lars Ellenberg1816a2b2010-11-11 15:19:07 +0100280 if (drbd_pp_vacant > (DRBD_MAX_BIO_SIZE/PAGE_SIZE)*minor_count)
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200281 i = page_chain_free(page);
282 else {
283 struct page *tmp;
284 tmp = page_chain_tail(page, &i);
285 spin_lock(&drbd_pp_lock);
286 page_chain_add(&drbd_pp_pool, page, tmp);
287 drbd_pp_vacant += i;
288 spin_unlock(&drbd_pp_lock);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700289 }
Lars Ellenberg435f0742010-09-06 12:30:25 +0200290 i = atomic_sub_return(i, a);
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200291 if (i < 0)
Lars Ellenberg435f0742010-09-06 12:30:25 +0200292 dev_warn(DEV, "ASSERTION FAILED: %s: %d < 0\n",
293 is_net ? "pp_in_use_by_net" : "pp_in_use", i);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700294 wake_up(&drbd_pp_wait);
295}
296
297/*
298You need to hold the req_lock:
299 _drbd_wait_ee_list_empty()
300
301You must not have the req_lock:
302 drbd_free_ee()
303 drbd_alloc_ee()
304 drbd_init_ee()
305 drbd_release_ee()
306 drbd_ee_fix_bhs()
307 drbd_process_done_ee()
308 drbd_clear_done_ee()
309 drbd_wait_ee_list_empty()
310*/
311
312struct drbd_epoch_entry *drbd_alloc_ee(struct drbd_conf *mdev,
313 u64 id,
314 sector_t sector,
315 unsigned int data_size,
316 gfp_t gfp_mask) __must_hold(local)
317{
Philipp Reisnerb411b362009-09-25 16:07:19 -0700318 struct drbd_epoch_entry *e;
319 struct page *page;
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200320 unsigned nr_pages = (data_size + PAGE_SIZE -1) >> PAGE_SHIFT;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700321
Andreas Gruenbacher0cf9d272010-12-07 10:43:29 +0100322 if (drbd_insert_fault(mdev, DRBD_FAULT_AL_EE))
Philipp Reisnerb411b362009-09-25 16:07:19 -0700323 return NULL;
324
325 e = mempool_alloc(drbd_ee_mempool, gfp_mask & ~__GFP_HIGHMEM);
326 if (!e) {
327 if (!(gfp_mask & __GFP_NOWARN))
328 dev_err(DEV, "alloc_ee: Allocation of an EE failed\n");
329 return NULL;
330 }
331
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200332 page = drbd_pp_alloc(mdev, nr_pages, (gfp_mask & __GFP_WAIT));
333 if (!page)
334 goto fail;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700335
Andreas Gruenbacher8b946252011-01-20 15:23:07 +0100336 drbd_clear_interval(&e->i);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700337 e->epoch = NULL;
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200338 e->mdev = mdev;
339 e->pages = page;
340 atomic_set(&e->pending_bios, 0);
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +0100341 e->i.size = data_size;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700342 e->flags = 0;
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +0100343 e->i.sector = sector;
Andreas Gruenbacher9a8e7752011-01-11 14:04:09 +0100344 /*
345 * The block_id is opaque to the receiver. It is not endianness
346 * converted, and sent back to the sender unchanged.
347 */
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200348 e->block_id = id;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700349
Philipp Reisnerb411b362009-09-25 16:07:19 -0700350 return e;
351
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200352 fail:
Philipp Reisnerb411b362009-09-25 16:07:19 -0700353 mempool_free(e, drbd_ee_mempool);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700354 return NULL;
355}
356
Lars Ellenberg435f0742010-09-06 12:30:25 +0200357void drbd_free_some_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e, int is_net)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700358{
Lars Ellenbergc36c3ce2010-08-11 20:42:55 +0200359 if (e->flags & EE_HAS_DIGEST)
360 kfree(e->digest);
Lars Ellenberg435f0742010-09-06 12:30:25 +0200361 drbd_pp_free(mdev, e->pages, is_net);
Lars Ellenberg45bb9122010-05-14 17:10:48 +0200362 D_ASSERT(atomic_read(&e->pending_bios) == 0);
Andreas Gruenbacher8b946252011-01-20 15:23:07 +0100363 D_ASSERT(drbd_interval_empty(&e->i));
Philipp Reisnerb411b362009-09-25 16:07:19 -0700364 mempool_free(e, drbd_ee_mempool);
365}
366
367int drbd_release_ee(struct drbd_conf *mdev, struct list_head *list)
368{
369 LIST_HEAD(work_list);
370 struct drbd_epoch_entry *e, *t;
371 int count = 0;
Lars Ellenberg435f0742010-09-06 12:30:25 +0200372 int is_net = list == &mdev->net_ee;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700373
374 spin_lock_irq(&mdev->req_lock);
375 list_splice_init(list, &work_list);
376 spin_unlock_irq(&mdev->req_lock);
377
378 list_for_each_entry_safe(e, t, &work_list, w.list) {
Lars Ellenberg435f0742010-09-06 12:30:25 +0200379 drbd_free_some_ee(mdev, e, is_net);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700380 count++;
381 }
382 return count;
383}
384
385
386/*
387 * This function is called from _asender only_
Andreas Gruenbacher8554df12011-01-25 15:37:43 +0100388 * but see also comments in _req_mod(,BARRIER_ACKED)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700389 * and receive_Barrier.
390 *
391 * Move entries from net_ee to done_ee, if ready.
392 * Grab done_ee, call all callbacks, free the entries.
393 * The callbacks typically send out ACKs.
394 */
395static int drbd_process_done_ee(struct drbd_conf *mdev)
396{
397 LIST_HEAD(work_list);
398 LIST_HEAD(reclaimed);
399 struct drbd_epoch_entry *e, *t;
400 int ok = (mdev->state.conn >= C_WF_REPORT_PARAMS);
401
402 spin_lock_irq(&mdev->req_lock);
403 reclaim_net_ee(mdev, &reclaimed);
404 list_splice_init(&mdev->done_ee, &work_list);
405 spin_unlock_irq(&mdev->req_lock);
406
407 list_for_each_entry_safe(e, t, &reclaimed, w.list)
Lars Ellenberg435f0742010-09-06 12:30:25 +0200408 drbd_free_net_ee(mdev, e);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700409
410 /* possible callbacks here:
411 * e_end_block, and e_end_resync_block, e_send_discard_ack.
412 * all ignore the last argument.
413 */
414 list_for_each_entry_safe(e, t, &work_list, w.list) {
Philipp Reisnerb411b362009-09-25 16:07:19 -0700415 /* list_del not necessary, next/prev members not touched */
416 ok = e->w.cb(mdev, &e->w, !ok) && ok;
417 drbd_free_ee(mdev, e);
418 }
419 wake_up(&mdev->ee_wait);
420
421 return ok;
422}
423
424void _drbd_wait_ee_list_empty(struct drbd_conf *mdev, struct list_head *head)
425{
426 DEFINE_WAIT(wait);
427
428 /* avoids spin_lock/unlock
429 * and calling prepare_to_wait in the fast path */
430 while (!list_empty(head)) {
431 prepare_to_wait(&mdev->ee_wait, &wait, TASK_UNINTERRUPTIBLE);
432 spin_unlock_irq(&mdev->req_lock);
Jens Axboe7eaceac2011-03-10 08:52:07 +0100433 io_schedule();
Philipp Reisnerb411b362009-09-25 16:07:19 -0700434 finish_wait(&mdev->ee_wait, &wait);
435 spin_lock_irq(&mdev->req_lock);
436 }
437}
438
439void drbd_wait_ee_list_empty(struct drbd_conf *mdev, struct list_head *head)
440{
441 spin_lock_irq(&mdev->req_lock);
442 _drbd_wait_ee_list_empty(mdev, head);
443 spin_unlock_irq(&mdev->req_lock);
444}
445
446/* see also kernel_accept; which is only present since 2.6.18.
447 * also we want to log which part of it failed, exactly */
448static int drbd_accept(struct drbd_conf *mdev, const char **what,
449 struct socket *sock, struct socket **newsock)
450{
451 struct sock *sk = sock->sk;
452 int err = 0;
453
454 *what = "listen";
455 err = sock->ops->listen(sock, 5);
456 if (err < 0)
457 goto out;
458
459 *what = "sock_create_lite";
460 err = sock_create_lite(sk->sk_family, sk->sk_type, sk->sk_protocol,
461 newsock);
462 if (err < 0)
463 goto out;
464
465 *what = "accept";
466 err = sock->ops->accept(sock, *newsock, 0);
467 if (err < 0) {
468 sock_release(*newsock);
469 *newsock = NULL;
470 goto out;
471 }
472 (*newsock)->ops = sock->ops;
473
474out:
475 return err;
476}
477
478static int drbd_recv_short(struct drbd_conf *mdev, struct socket *sock,
479 void *buf, size_t size, int flags)
480{
481 mm_segment_t oldfs;
482 struct kvec iov = {
483 .iov_base = buf,
484 .iov_len = size,
485 };
486 struct msghdr msg = {
487 .msg_iovlen = 1,
488 .msg_iov = (struct iovec *)&iov,
489 .msg_flags = (flags ? flags : MSG_WAITALL | MSG_NOSIGNAL)
490 };
491 int rv;
492
493 oldfs = get_fs();
494 set_fs(KERNEL_DS);
495 rv = sock_recvmsg(sock, &msg, size, msg.msg_flags);
496 set_fs(oldfs);
497
498 return rv;
499}
500
501static int drbd_recv(struct drbd_conf *mdev, void *buf, size_t size)
502{
503 mm_segment_t oldfs;
504 struct kvec iov = {
505 .iov_base = buf,
506 .iov_len = size,
507 };
508 struct msghdr msg = {
509 .msg_iovlen = 1,
510 .msg_iov = (struct iovec *)&iov,
511 .msg_flags = MSG_WAITALL | MSG_NOSIGNAL
512 };
513 int rv;
514
515 oldfs = get_fs();
516 set_fs(KERNEL_DS);
517
518 for (;;) {
519 rv = sock_recvmsg(mdev->data.socket, &msg, size, msg.msg_flags);
520 if (rv == size)
521 break;
522
523 /* Note:
524 * ECONNRESET other side closed the connection
525 * ERESTARTSYS (on sock) we got a signal
526 */
527
528 if (rv < 0) {
529 if (rv == -ECONNRESET)
530 dev_info(DEV, "sock was reset by peer\n");
531 else if (rv != -ERESTARTSYS)
532 dev_err(DEV, "sock_recvmsg returned %d\n", rv);
533 break;
534 } else if (rv == 0) {
535 dev_info(DEV, "sock was shut down by peer\n");
536 break;
537 } else {
538 /* signal came in, or peer/link went down,
539 * after we read a partial message
540 */
541 /* D_ASSERT(signal_pending(current)); */
542 break;
543 }
544 };
545
546 set_fs(oldfs);
547
548 if (rv != size)
549 drbd_force_state(mdev, NS(conn, C_BROKEN_PIPE));
550
551 return rv;
552}
553
Lars Ellenberg5dbf1672010-05-25 16:18:01 +0200554/* quoting tcp(7):
555 * On individual connections, the socket buffer size must be set prior to the
556 * listen(2) or connect(2) calls in order to have it take effect.
557 * This is our wrapper to do so.
558 */
559static void drbd_setbufsize(struct socket *sock, unsigned int snd,
560 unsigned int rcv)
561{
562 /* open coded SO_SNDBUF, SO_RCVBUF */
563 if (snd) {
564 sock->sk->sk_sndbuf = snd;
565 sock->sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
566 }
567 if (rcv) {
568 sock->sk->sk_rcvbuf = rcv;
569 sock->sk->sk_userlocks |= SOCK_RCVBUF_LOCK;
570 }
571}
572
Philipp Reisnerb411b362009-09-25 16:07:19 -0700573static struct socket *drbd_try_connect(struct drbd_conf *mdev)
574{
575 const char *what;
576 struct socket *sock;
577 struct sockaddr_in6 src_in6;
578 int err;
579 int disconnect_on_error = 1;
580
581 if (!get_net_conf(mdev))
582 return NULL;
583
584 what = "sock_create_kern";
585 err = sock_create_kern(((struct sockaddr *)mdev->net_conf->my_addr)->sa_family,
586 SOCK_STREAM, IPPROTO_TCP, &sock);
587 if (err < 0) {
588 sock = NULL;
589 goto out;
590 }
591
592 sock->sk->sk_rcvtimeo =
593 sock->sk->sk_sndtimeo = mdev->net_conf->try_connect_int*HZ;
Lars Ellenberg5dbf1672010-05-25 16:18:01 +0200594 drbd_setbufsize(sock, mdev->net_conf->sndbuf_size,
595 mdev->net_conf->rcvbuf_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700596
597 /* explicitly bind to the configured IP as source IP
598 * for the outgoing connections.
599 * This is needed for multihomed hosts and to be
600 * able to use lo: interfaces for drbd.
601 * Make sure to use 0 as port number, so linux selects
602 * a free one dynamically.
603 */
604 memcpy(&src_in6, mdev->net_conf->my_addr,
605 min_t(int, mdev->net_conf->my_addr_len, sizeof(src_in6)));
606 if (((struct sockaddr *)mdev->net_conf->my_addr)->sa_family == AF_INET6)
607 src_in6.sin6_port = 0;
608 else
609 ((struct sockaddr_in *)&src_in6)->sin_port = 0; /* AF_INET & AF_SCI */
610
611 what = "bind before connect";
612 err = sock->ops->bind(sock,
613 (struct sockaddr *) &src_in6,
614 mdev->net_conf->my_addr_len);
615 if (err < 0)
616 goto out;
617
618 /* connect may fail, peer not yet available.
619 * stay C_WF_CONNECTION, don't go Disconnecting! */
620 disconnect_on_error = 0;
621 what = "connect";
622 err = sock->ops->connect(sock,
623 (struct sockaddr *)mdev->net_conf->peer_addr,
624 mdev->net_conf->peer_addr_len, 0);
625
626out:
627 if (err < 0) {
628 if (sock) {
629 sock_release(sock);
630 sock = NULL;
631 }
632 switch (-err) {
633 /* timeout, busy, signal pending */
634 case ETIMEDOUT: case EAGAIN: case EINPROGRESS:
635 case EINTR: case ERESTARTSYS:
636 /* peer not (yet) available, network problem */
637 case ECONNREFUSED: case ENETUNREACH:
638 case EHOSTDOWN: case EHOSTUNREACH:
639 disconnect_on_error = 0;
640 break;
641 default:
642 dev_err(DEV, "%s failed, err = %d\n", what, err);
643 }
644 if (disconnect_on_error)
645 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
646 }
647 put_net_conf(mdev);
648 return sock;
649}
650
651static struct socket *drbd_wait_for_connect(struct drbd_conf *mdev)
652{
653 int timeo, err;
654 struct socket *s_estab = NULL, *s_listen;
655 const char *what;
656
657 if (!get_net_conf(mdev))
658 return NULL;
659
660 what = "sock_create_kern";
661 err = sock_create_kern(((struct sockaddr *)mdev->net_conf->my_addr)->sa_family,
662 SOCK_STREAM, IPPROTO_TCP, &s_listen);
663 if (err) {
664 s_listen = NULL;
665 goto out;
666 }
667
668 timeo = mdev->net_conf->try_connect_int * HZ;
669 timeo += (random32() & 1) ? timeo / 7 : -timeo / 7; /* 28.5% random jitter */
670
671 s_listen->sk->sk_reuse = 1; /* SO_REUSEADDR */
672 s_listen->sk->sk_rcvtimeo = timeo;
673 s_listen->sk->sk_sndtimeo = timeo;
Lars Ellenberg5dbf1672010-05-25 16:18:01 +0200674 drbd_setbufsize(s_listen, mdev->net_conf->sndbuf_size,
675 mdev->net_conf->rcvbuf_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700676
677 what = "bind before listen";
678 err = s_listen->ops->bind(s_listen,
679 (struct sockaddr *) mdev->net_conf->my_addr,
680 mdev->net_conf->my_addr_len);
681 if (err < 0)
682 goto out;
683
684 err = drbd_accept(mdev, &what, s_listen, &s_estab);
685
686out:
687 if (s_listen)
688 sock_release(s_listen);
689 if (err < 0) {
690 if (err != -EAGAIN && err != -EINTR && err != -ERESTARTSYS) {
691 dev_err(DEV, "%s failed, err = %d\n", what, err);
692 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
693 }
694 }
695 put_net_conf(mdev);
696
697 return s_estab;
698}
699
700static int drbd_send_fp(struct drbd_conf *mdev,
701 struct socket *sock, enum drbd_packets cmd)
702{
Philipp Reisner02918be2010-08-20 14:35:10 +0200703 struct p_header80 *h = &mdev->data.sbuf.header.h80;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700704
705 return _drbd_send_cmd(mdev, sock, cmd, h, sizeof(*h), 0);
706}
707
708static enum drbd_packets drbd_recv_fp(struct drbd_conf *mdev, struct socket *sock)
709{
Philipp Reisner02918be2010-08-20 14:35:10 +0200710 struct p_header80 *h = &mdev->data.rbuf.header.h80;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700711 int rr;
712
713 rr = drbd_recv_short(mdev, sock, h, sizeof(*h), 0);
714
Andreas Gruenbacherca9bc122011-01-11 13:47:24 +0100715 if (rr == sizeof(*h) && h->magic == cpu_to_be32(DRBD_MAGIC))
Philipp Reisnerb411b362009-09-25 16:07:19 -0700716 return be16_to_cpu(h->command);
717
718 return 0xffff;
719}
720
721/**
722 * drbd_socket_okay() - Free the socket if its connection is not okay
723 * @mdev: DRBD device.
724 * @sock: pointer to the pointer to the socket.
725 */
726static int drbd_socket_okay(struct drbd_conf *mdev, struct socket **sock)
727{
728 int rr;
729 char tb[4];
730
731 if (!*sock)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +0100732 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700733
734 rr = drbd_recv_short(mdev, *sock, tb, 4, MSG_DONTWAIT | MSG_PEEK);
735
736 if (rr > 0 || rr == -EAGAIN) {
Andreas Gruenbacher81e84652010-12-09 15:03:57 +0100737 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700738 } else {
739 sock_release(*sock);
740 *sock = NULL;
Andreas Gruenbacher81e84652010-12-09 15:03:57 +0100741 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700742 }
743}
744
745/*
746 * return values:
747 * 1 yes, we have a valid connection
748 * 0 oops, did not work out, please try again
749 * -1 peer talks different language,
750 * no point in trying again, please go standalone.
751 * -2 We do not have a network config...
752 */
753static int drbd_connect(struct drbd_conf *mdev)
754{
755 struct socket *s, *sock, *msock;
756 int try, h, ok;
757
758 D_ASSERT(!mdev->data.socket);
759
Philipp Reisnerb411b362009-09-25 16:07:19 -0700760 if (drbd_request_state(mdev, NS(conn, C_WF_CONNECTION)) < SS_SUCCESS)
761 return -2;
762
763 clear_bit(DISCARD_CONCURRENT, &mdev->flags);
764
765 sock = NULL;
766 msock = NULL;
767
768 do {
769 for (try = 0;;) {
770 /* 3 tries, this should take less than a second! */
771 s = drbd_try_connect(mdev);
772 if (s || ++try >= 3)
773 break;
774 /* give the other side time to call bind() & listen() */
Philipp Reisner20ee6392011-01-18 15:28:59 +0100775 schedule_timeout_interruptible(HZ / 10);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700776 }
777
778 if (s) {
779 if (!sock) {
780 drbd_send_fp(mdev, s, P_HAND_SHAKE_S);
781 sock = s;
782 s = NULL;
783 } else if (!msock) {
784 drbd_send_fp(mdev, s, P_HAND_SHAKE_M);
785 msock = s;
786 s = NULL;
787 } else {
788 dev_err(DEV, "Logic error in drbd_connect()\n");
789 goto out_release_sockets;
790 }
791 }
792
793 if (sock && msock) {
Philipp Reisnera8e40792011-05-13 12:03:55 +0200794 schedule_timeout_interruptible(mdev->net_conf->ping_timeo*HZ/10);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700795 ok = drbd_socket_okay(mdev, &sock);
796 ok = drbd_socket_okay(mdev, &msock) && ok;
797 if (ok)
798 break;
799 }
800
801retry:
802 s = drbd_wait_for_connect(mdev);
803 if (s) {
804 try = drbd_recv_fp(mdev, s);
805 drbd_socket_okay(mdev, &sock);
806 drbd_socket_okay(mdev, &msock);
807 switch (try) {
808 case P_HAND_SHAKE_S:
809 if (sock) {
810 dev_warn(DEV, "initial packet S crossed\n");
811 sock_release(sock);
812 }
813 sock = s;
814 break;
815 case P_HAND_SHAKE_M:
816 if (msock) {
817 dev_warn(DEV, "initial packet M crossed\n");
818 sock_release(msock);
819 }
820 msock = s;
821 set_bit(DISCARD_CONCURRENT, &mdev->flags);
822 break;
823 default:
824 dev_warn(DEV, "Error receiving initial packet\n");
825 sock_release(s);
826 if (random32() & 1)
827 goto retry;
828 }
829 }
830
831 if (mdev->state.conn <= C_DISCONNECTING)
832 goto out_release_sockets;
833 if (signal_pending(current)) {
834 flush_signals(current);
835 smp_rmb();
Andreas Gruenbachere77a0a52011-01-25 15:43:39 +0100836 if (get_t_state(&mdev->receiver) == EXITING)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700837 goto out_release_sockets;
838 }
839
840 if (sock && msock) {
841 ok = drbd_socket_okay(mdev, &sock);
842 ok = drbd_socket_okay(mdev, &msock) && ok;
843 if (ok)
844 break;
845 }
846 } while (1);
847
848 msock->sk->sk_reuse = 1; /* SO_REUSEADDR */
849 sock->sk->sk_reuse = 1; /* SO_REUSEADDR */
850
851 sock->sk->sk_allocation = GFP_NOIO;
852 msock->sk->sk_allocation = GFP_NOIO;
853
854 sock->sk->sk_priority = TC_PRIO_INTERACTIVE_BULK;
855 msock->sk->sk_priority = TC_PRIO_INTERACTIVE;
856
Philipp Reisnerb411b362009-09-25 16:07:19 -0700857 /* NOT YET ...
858 * sock->sk->sk_sndtimeo = mdev->net_conf->timeout*HZ/10;
859 * sock->sk->sk_rcvtimeo = MAX_SCHEDULE_TIMEOUT;
860 * first set it to the P_HAND_SHAKE timeout,
861 * which we set to 4x the configured ping_timeout. */
862 sock->sk->sk_sndtimeo =
863 sock->sk->sk_rcvtimeo = mdev->net_conf->ping_timeo*4*HZ/10;
864
865 msock->sk->sk_sndtimeo = mdev->net_conf->timeout*HZ/10;
866 msock->sk->sk_rcvtimeo = mdev->net_conf->ping_int*HZ;
867
868 /* we don't want delays.
Lucas De Marchi25985ed2011-03-30 22:57:33 -0300869 * we use TCP_CORK where appropriate, though */
Philipp Reisnerb411b362009-09-25 16:07:19 -0700870 drbd_tcp_nodelay(sock);
871 drbd_tcp_nodelay(msock);
872
873 mdev->data.socket = sock;
874 mdev->meta.socket = msock;
875 mdev->last_received = jiffies;
876
877 D_ASSERT(mdev->asender.task == NULL);
878
879 h = drbd_do_handshake(mdev);
880 if (h <= 0)
881 return h;
882
883 if (mdev->cram_hmac_tfm) {
884 /* drbd_request_state(mdev, NS(conn, WFAuth)); */
Johannes Thomab10d96c2010-01-07 16:02:50 +0100885 switch (drbd_do_auth(mdev)) {
886 case -1:
Philipp Reisnerb411b362009-09-25 16:07:19 -0700887 dev_err(DEV, "Authentication of peer failed\n");
888 return -1;
Johannes Thomab10d96c2010-01-07 16:02:50 +0100889 case 0:
890 dev_err(DEV, "Authentication of peer failed, trying again.\n");
891 return 0;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700892 }
893 }
894
895 if (drbd_request_state(mdev, NS(conn, C_WF_REPORT_PARAMS)) < SS_SUCCESS)
896 return 0;
897
898 sock->sk->sk_sndtimeo = mdev->net_conf->timeout*HZ/10;
899 sock->sk->sk_rcvtimeo = MAX_SCHEDULE_TIMEOUT;
900
901 atomic_set(&mdev->packet_seq, 0);
902 mdev->peer_seq = 0;
903
904 drbd_thread_start(&mdev->asender);
905
Philipp Reisner148efa12011-01-15 00:21:15 +0100906 if (drbd_send_protocol(mdev) == -1)
Philipp Reisner7e2455c2010-04-22 14:50:23 +0200907 return -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700908 drbd_send_sync_param(mdev, &mdev->sync_conf);
Philipp Reisnere89b5912010-03-24 17:11:33 +0100909 drbd_send_sizes(mdev, 0, 0);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700910 drbd_send_uuids(mdev);
911 drbd_send_state(mdev);
912 clear_bit(USE_DEGR_WFC_T, &mdev->flags);
913 clear_bit(RESIZE_PENDING, &mdev->flags);
Philipp Reisner7fde2be2011-03-01 11:08:28 +0100914 mod_timer(&mdev->request_timer, jiffies + HZ); /* just start it here. */
Philipp Reisnerb411b362009-09-25 16:07:19 -0700915
916 return 1;
917
918out_release_sockets:
919 if (sock)
920 sock_release(sock);
921 if (msock)
922 sock_release(msock);
923 return -1;
924}
925
Philipp Reisner02918be2010-08-20 14:35:10 +0200926static int drbd_recv_header(struct drbd_conf *mdev, enum drbd_packets *cmd, unsigned int *packet_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700927{
Philipp Reisner02918be2010-08-20 14:35:10 +0200928 union p_header *h = &mdev->data.rbuf.header;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700929 int r;
930
931 r = drbd_recv(mdev, h, sizeof(*h));
Philipp Reisnerb411b362009-09-25 16:07:19 -0700932 if (unlikely(r != sizeof(*h))) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +0100933 if (!signal_pending(current))
934 dev_warn(DEV, "short read expecting header on sock: r=%d\n", r);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +0100935 return false;
Philipp Reisner02918be2010-08-20 14:35:10 +0200936 }
937
Andreas Gruenbacherca9bc122011-01-11 13:47:24 +0100938 if (likely(h->h80.magic == cpu_to_be32(DRBD_MAGIC))) {
Philipp Reisner02918be2010-08-20 14:35:10 +0200939 *cmd = be16_to_cpu(h->h80.command);
940 *packet_size = be16_to_cpu(h->h80.length);
Andreas Gruenbacherca9bc122011-01-11 13:47:24 +0100941 } else if (h->h95.magic == cpu_to_be16(DRBD_MAGIC_BIG)) {
Philipp Reisner02918be2010-08-20 14:35:10 +0200942 *cmd = be16_to_cpu(h->h95.command);
943 *packet_size = be32_to_cpu(h->h95.length);
944 } else {
Lars Ellenberg004352f2010-10-05 20:13:58 +0200945 dev_err(DEV, "magic?? on data m: 0x%08x c: %d l: %d\n",
946 be32_to_cpu(h->h80.magic),
947 be16_to_cpu(h->h80.command),
948 be16_to_cpu(h->h80.length));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +0100949 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700950 }
951 mdev->last_received = jiffies;
952
Andreas Gruenbacher81e84652010-12-09 15:03:57 +0100953 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700954}
955
Philipp Reisner2451fc32010-08-24 13:43:11 +0200956static void drbd_flush(struct drbd_conf *mdev)
Philipp Reisnerb411b362009-09-25 16:07:19 -0700957{
958 int rv;
959
960 if (mdev->write_ordering >= WO_bdev_flush && get_ldev(mdev)) {
Dmitry Monakhovfbd9b092010-04-28 17:55:06 +0400961 rv = blkdev_issue_flush(mdev->ldev->backing_bdev, GFP_KERNEL,
Christoph Hellwigdd3932e2010-09-16 20:51:46 +0200962 NULL);
Philipp Reisnerb411b362009-09-25 16:07:19 -0700963 if (rv) {
964 dev_err(DEV, "local disk flush failed with status %d\n", rv);
965 /* would rather check on EOPNOTSUPP, but that is not reliable.
966 * don't try again for ANY return value != 0
967 * if (rv == -EOPNOTSUPP) */
968 drbd_bump_write_ordering(mdev, WO_drain_io);
969 }
970 put_ldev(mdev);
971 }
Philipp Reisnerb411b362009-09-25 16:07:19 -0700972}
973
974/**
975 * drbd_may_finish_epoch() - Applies an epoch_event to the epoch's state, eventually finishes it.
976 * @mdev: DRBD device.
977 * @epoch: Epoch object.
978 * @ev: Epoch event.
979 */
980static enum finish_epoch drbd_may_finish_epoch(struct drbd_conf *mdev,
981 struct drbd_epoch *epoch,
982 enum epoch_event ev)
983{
Philipp Reisner2451fc32010-08-24 13:43:11 +0200984 int epoch_size;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700985 struct drbd_epoch *next_epoch;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700986 enum finish_epoch rv = FE_STILL_LIVE;
987
988 spin_lock(&mdev->epoch_lock);
989 do {
990 next_epoch = NULL;
Philipp Reisnerb411b362009-09-25 16:07:19 -0700991
992 epoch_size = atomic_read(&epoch->epoch_size);
993
994 switch (ev & ~EV_CLEANUP) {
995 case EV_PUT:
996 atomic_dec(&epoch->active);
997 break;
998 case EV_GOT_BARRIER_NR:
999 set_bit(DE_HAVE_BARRIER_NUMBER, &epoch->flags);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001000 break;
1001 case EV_BECAME_LAST:
1002 /* nothing to do*/
1003 break;
1004 }
1005
Philipp Reisnerb411b362009-09-25 16:07:19 -07001006 if (epoch_size != 0 &&
1007 atomic_read(&epoch->active) == 0 &&
Philipp Reisner2451fc32010-08-24 13:43:11 +02001008 test_bit(DE_HAVE_BARRIER_NUMBER, &epoch->flags)) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001009 if (!(ev & EV_CLEANUP)) {
1010 spin_unlock(&mdev->epoch_lock);
1011 drbd_send_b_ack(mdev, epoch->barrier_nr, epoch_size);
1012 spin_lock(&mdev->epoch_lock);
1013 }
1014 dec_unacked(mdev);
1015
1016 if (mdev->current_epoch != epoch) {
1017 next_epoch = list_entry(epoch->list.next, struct drbd_epoch, list);
1018 list_del(&epoch->list);
1019 ev = EV_BECAME_LAST | (ev & EV_CLEANUP);
1020 mdev->epochs--;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001021 kfree(epoch);
1022
1023 if (rv == FE_STILL_LIVE)
1024 rv = FE_DESTROYED;
1025 } else {
1026 epoch->flags = 0;
1027 atomic_set(&epoch->epoch_size, 0);
Uwe Kleine-König698f9312010-07-02 20:41:51 +02001028 /* atomic_set(&epoch->active, 0); is already zero */
Philipp Reisnerb411b362009-09-25 16:07:19 -07001029 if (rv == FE_STILL_LIVE)
1030 rv = FE_RECYCLED;
Philipp Reisner2451fc32010-08-24 13:43:11 +02001031 wake_up(&mdev->ee_wait);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001032 }
1033 }
1034
1035 if (!next_epoch)
1036 break;
1037
1038 epoch = next_epoch;
1039 } while (1);
1040
1041 spin_unlock(&mdev->epoch_lock);
1042
Philipp Reisnerb411b362009-09-25 16:07:19 -07001043 return rv;
1044}
1045
1046/**
1047 * drbd_bump_write_ordering() - Fall back to an other write ordering method
1048 * @mdev: DRBD device.
1049 * @wo: Write ordering method to try.
1050 */
1051void drbd_bump_write_ordering(struct drbd_conf *mdev, enum write_ordering_e wo) __must_hold(local)
1052{
1053 enum write_ordering_e pwo;
1054 static char *write_ordering_str[] = {
1055 [WO_none] = "none",
1056 [WO_drain_io] = "drain",
1057 [WO_bdev_flush] = "flush",
Philipp Reisnerb411b362009-09-25 16:07:19 -07001058 };
1059
1060 pwo = mdev->write_ordering;
1061 wo = min(pwo, wo);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001062 if (wo == WO_bdev_flush && mdev->ldev->dc.no_disk_flush)
1063 wo = WO_drain_io;
1064 if (wo == WO_drain_io && mdev->ldev->dc.no_disk_drain)
1065 wo = WO_none;
1066 mdev->write_ordering = wo;
Philipp Reisner2451fc32010-08-24 13:43:11 +02001067 if (pwo != mdev->write_ordering || wo == WO_bdev_flush)
Philipp Reisnerb411b362009-09-25 16:07:19 -07001068 dev_info(DEV, "Method to ensure write ordering: %s\n", write_ordering_str[mdev->write_ordering]);
1069}
1070
1071/**
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001072 * drbd_submit_ee()
1073 * @mdev: DRBD device.
1074 * @e: epoch entry
1075 * @rw: flag field, see bio->bi_rw
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001076 *
1077 * May spread the pages to multiple bios,
1078 * depending on bio_add_page restrictions.
1079 *
1080 * Returns 0 if all bios have been submitted,
1081 * -ENOMEM if we could not allocate enough bios,
1082 * -ENOSPC (any better suggestion?) if we have not been able to bio_add_page a
1083 * single page to an empty bio (which should never happen and likely indicates
1084 * that the lower level IO stack is in some way broken). This has been observed
1085 * on certain Xen deployments.
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001086 */
1087/* TODO allocate from our own bio_set. */
1088int drbd_submit_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e,
1089 const unsigned rw, const int fault_type)
1090{
1091 struct bio *bios = NULL;
1092 struct bio *bio;
1093 struct page *page = e->pages;
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001094 sector_t sector = e->i.sector;
1095 unsigned ds = e->i.size;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001096 unsigned n_bios = 0;
1097 unsigned nr_pages = (ds + PAGE_SIZE -1) >> PAGE_SHIFT;
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001098 int err = -ENOMEM;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001099
1100 /* In most cases, we will only need one bio. But in case the lower
1101 * level restrictions happen to be different at this offset on this
1102 * side than those of the sending peer, we may need to submit the
1103 * request in more than one bio. */
1104next_bio:
1105 bio = bio_alloc(GFP_NOIO, nr_pages);
1106 if (!bio) {
1107 dev_err(DEV, "submit_ee: Allocation of a bio failed\n");
1108 goto fail;
1109 }
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001110 /* > e->i.sector, unless this is the first bio */
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001111 bio->bi_sector = sector;
1112 bio->bi_bdev = mdev->ldev->backing_bdev;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001113 bio->bi_rw = rw;
1114 bio->bi_private = e;
1115 bio->bi_end_io = drbd_endio_sec;
1116
1117 bio->bi_next = bios;
1118 bios = bio;
1119 ++n_bios;
1120
1121 page_chain_for_each(page) {
1122 unsigned len = min_t(unsigned, ds, PAGE_SIZE);
1123 if (!bio_add_page(bio, page, len, 0)) {
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001124 /* A single page must always be possible!
1125 * But in case it fails anyways,
1126 * we deal with it, and complain (below). */
1127 if (bio->bi_vcnt == 0) {
1128 dev_err(DEV,
1129 "bio_add_page failed for len=%u, "
1130 "bi_vcnt=0 (bi_sector=%llu)\n",
1131 len, (unsigned long long)bio->bi_sector);
1132 err = -ENOSPC;
1133 goto fail;
1134 }
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001135 goto next_bio;
1136 }
1137 ds -= len;
1138 sector += len >> 9;
1139 --nr_pages;
1140 }
1141 D_ASSERT(page == NULL);
1142 D_ASSERT(ds == 0);
1143
1144 atomic_set(&e->pending_bios, n_bios);
1145 do {
1146 bio = bios;
1147 bios = bios->bi_next;
1148 bio->bi_next = NULL;
1149
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001150 drbd_generic_make_request(mdev, fault_type, bio);
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001151 } while (bios);
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001152 return 0;
1153
1154fail:
1155 while (bios) {
1156 bio = bios;
1157 bios = bios->bi_next;
1158 bio_put(bio);
1159 }
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001160 return err;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001161}
1162
Philipp Reisner02918be2010-08-20 14:35:10 +02001163static int receive_Barrier(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07001164{
Philipp Reisner2451fc32010-08-24 13:43:11 +02001165 int rv;
Philipp Reisner02918be2010-08-20 14:35:10 +02001166 struct p_barrier *p = &mdev->data.rbuf.barrier;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001167 struct drbd_epoch *epoch;
1168
Philipp Reisnerb411b362009-09-25 16:07:19 -07001169 inc_unacked(mdev);
1170
Philipp Reisnerb411b362009-09-25 16:07:19 -07001171 mdev->current_epoch->barrier_nr = p->barrier;
1172 rv = drbd_may_finish_epoch(mdev, mdev->current_epoch, EV_GOT_BARRIER_NR);
1173
1174 /* P_BARRIER_ACK may imply that the corresponding extent is dropped from
1175 * the activity log, which means it would not be resynced in case the
1176 * R_PRIMARY crashes now.
1177 * Therefore we must send the barrier_ack after the barrier request was
1178 * completed. */
1179 switch (mdev->write_ordering) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001180 case WO_none:
1181 if (rv == FE_RECYCLED)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001182 return true;
Philipp Reisner2451fc32010-08-24 13:43:11 +02001183
1184 /* receiver context, in the writeout path of the other node.
1185 * avoid potential distributed deadlock */
1186 epoch = kmalloc(sizeof(struct drbd_epoch), GFP_NOIO);
1187 if (epoch)
1188 break;
1189 else
1190 dev_warn(DEV, "Allocation of an epoch failed, slowing down\n");
1191 /* Fall through */
Philipp Reisnerb411b362009-09-25 16:07:19 -07001192
1193 case WO_bdev_flush:
1194 case WO_drain_io:
Philipp Reisnerb411b362009-09-25 16:07:19 -07001195 drbd_wait_ee_list_empty(mdev, &mdev->active_ee);
Philipp Reisner2451fc32010-08-24 13:43:11 +02001196 drbd_flush(mdev);
1197
1198 if (atomic_read(&mdev->current_epoch->epoch_size)) {
1199 epoch = kmalloc(sizeof(struct drbd_epoch), GFP_NOIO);
1200 if (epoch)
1201 break;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001202 }
1203
Philipp Reisner2451fc32010-08-24 13:43:11 +02001204 epoch = mdev->current_epoch;
1205 wait_event(mdev->ee_wait, atomic_read(&epoch->epoch_size) == 0);
1206
1207 D_ASSERT(atomic_read(&epoch->active) == 0);
1208 D_ASSERT(epoch->flags == 0);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001209
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001210 return true;
Philipp Reisner2451fc32010-08-24 13:43:11 +02001211 default:
1212 dev_err(DEV, "Strangeness in mdev->write_ordering %d\n", mdev->write_ordering);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001213 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001214 }
1215
1216 epoch->flags = 0;
1217 atomic_set(&epoch->epoch_size, 0);
1218 atomic_set(&epoch->active, 0);
1219
1220 spin_lock(&mdev->epoch_lock);
1221 if (atomic_read(&mdev->current_epoch->epoch_size)) {
1222 list_add(&epoch->list, &mdev->current_epoch->list);
1223 mdev->current_epoch = epoch;
1224 mdev->epochs++;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001225 } else {
1226 /* The current_epoch got recycled while we allocated this one... */
1227 kfree(epoch);
1228 }
1229 spin_unlock(&mdev->epoch_lock);
1230
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001231 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001232}
1233
1234/* used from receive_RSDataReply (recv_resync_read)
1235 * and from receive_Data */
1236static struct drbd_epoch_entry *
1237read_in_block(struct drbd_conf *mdev, u64 id, sector_t sector, int data_size) __must_hold(local)
1238{
Lars Ellenberg66660322010-04-06 12:15:04 +02001239 const sector_t capacity = drbd_get_capacity(mdev->this_bdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001240 struct drbd_epoch_entry *e;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001241 struct page *page;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001242 int dgs, ds, rr;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001243 void *dig_in = mdev->int_dig_in;
1244 void *dig_vv = mdev->int_dig_vv;
Philipp Reisner6b4388a2010-04-26 14:11:45 +02001245 unsigned long *data;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001246
1247 dgs = (mdev->agreed_pro_version >= 87 && mdev->integrity_r_tfm) ?
1248 crypto_hash_digestsize(mdev->integrity_r_tfm) : 0;
1249
1250 if (dgs) {
1251 rr = drbd_recv(mdev, dig_in, dgs);
1252 if (rr != dgs) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01001253 if (!signal_pending(current))
1254 dev_warn(DEV,
1255 "short read receiving data digest: read %d expected %d\n",
1256 rr, dgs);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001257 return NULL;
1258 }
1259 }
1260
1261 data_size -= dgs;
1262
Philipp Reisnerd07c9c12011-01-20 16:49:33 +01001263 ERR_IF(data_size == 0) return NULL;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001264 ERR_IF(data_size & 0x1ff) return NULL;
Lars Ellenberg1816a2b2010-11-11 15:19:07 +01001265 ERR_IF(data_size > DRBD_MAX_BIO_SIZE) return NULL;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001266
Lars Ellenberg66660322010-04-06 12:15:04 +02001267 /* even though we trust out peer,
1268 * we sometimes have to double check. */
1269 if (sector + (data_size>>9) > capacity) {
Lars Ellenbergfdda6542011-01-24 15:11:01 +01001270 dev_err(DEV, "request from peer beyond end of local disk: "
1271 "capacity: %llus < sector: %llus + size: %u\n",
Lars Ellenberg66660322010-04-06 12:15:04 +02001272 (unsigned long long)capacity,
1273 (unsigned long long)sector, data_size);
1274 return NULL;
1275 }
1276
Philipp Reisnerb411b362009-09-25 16:07:19 -07001277 /* GFP_NOIO, because we must not cause arbitrary write-out: in a DRBD
1278 * "criss-cross" setup, that might cause write-out on some other DRBD,
1279 * which in turn might block on the other node at this very place. */
1280 e = drbd_alloc_ee(mdev, id, sector, data_size, GFP_NOIO);
1281 if (!e)
1282 return NULL;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001283
Philipp Reisnerb411b362009-09-25 16:07:19 -07001284 ds = data_size;
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001285 page = e->pages;
1286 page_chain_for_each(page) {
1287 unsigned len = min_t(int, ds, PAGE_SIZE);
Philipp Reisner6b4388a2010-04-26 14:11:45 +02001288 data = kmap(page);
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001289 rr = drbd_recv(mdev, data, len);
Andreas Gruenbacher0cf9d272010-12-07 10:43:29 +01001290 if (drbd_insert_fault(mdev, DRBD_FAULT_RECEIVE)) {
Philipp Reisner6b4388a2010-04-26 14:11:45 +02001291 dev_err(DEV, "Fault injection: Corrupting data on receive\n");
1292 data[0] = data[0] ^ (unsigned long)-1;
1293 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07001294 kunmap(page);
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001295 if (rr != len) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001296 drbd_free_ee(mdev, e);
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01001297 if (!signal_pending(current))
1298 dev_warn(DEV, "short read receiving data: read %d expected %d\n",
1299 rr, len);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001300 return NULL;
1301 }
1302 ds -= rr;
1303 }
1304
1305 if (dgs) {
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001306 drbd_csum_ee(mdev, mdev->integrity_r_tfm, e, dig_vv);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001307 if (memcmp(dig_in, dig_vv, dgs)) {
Lars Ellenberg470be442010-11-10 10:36:52 +01001308 dev_err(DEV, "Digest integrity check FAILED: %llus +%u\n",
1309 (unsigned long long)sector, data_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001310 drbd_bcast_ee(mdev, "digest failed",
1311 dgs, dig_in, dig_vv, e);
1312 drbd_free_ee(mdev, e);
1313 return NULL;
1314 }
1315 }
1316 mdev->recv_cnt += data_size>>9;
1317 return e;
1318}
1319
1320/* drbd_drain_block() just takes a data block
1321 * out of the socket input buffer, and discards it.
1322 */
1323static int drbd_drain_block(struct drbd_conf *mdev, int data_size)
1324{
1325 struct page *page;
1326 int rr, rv = 1;
1327 void *data;
1328
Lars Ellenbergc3470cd2010-04-01 16:57:19 +02001329 if (!data_size)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001330 return true;
Lars Ellenbergc3470cd2010-04-01 16:57:19 +02001331
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001332 page = drbd_pp_alloc(mdev, 1, 1);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001333
1334 data = kmap(page);
1335 while (data_size) {
1336 rr = drbd_recv(mdev, data, min_t(int, data_size, PAGE_SIZE));
1337 if (rr != min_t(int, data_size, PAGE_SIZE)) {
1338 rv = 0;
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01001339 if (!signal_pending(current))
1340 dev_warn(DEV,
1341 "short read receiving data: read %d expected %d\n",
1342 rr, min_t(int, data_size, PAGE_SIZE));
Philipp Reisnerb411b362009-09-25 16:07:19 -07001343 break;
1344 }
1345 data_size -= rr;
1346 }
1347 kunmap(page);
Lars Ellenberg435f0742010-09-06 12:30:25 +02001348 drbd_pp_free(mdev, page, 0);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001349 return rv;
1350}
1351
1352static int recv_dless_read(struct drbd_conf *mdev, struct drbd_request *req,
1353 sector_t sector, int data_size)
1354{
1355 struct bio_vec *bvec;
1356 struct bio *bio;
1357 int dgs, rr, i, expect;
1358 void *dig_in = mdev->int_dig_in;
1359 void *dig_vv = mdev->int_dig_vv;
1360
1361 dgs = (mdev->agreed_pro_version >= 87 && mdev->integrity_r_tfm) ?
1362 crypto_hash_digestsize(mdev->integrity_r_tfm) : 0;
1363
1364 if (dgs) {
1365 rr = drbd_recv(mdev, dig_in, dgs);
1366 if (rr != dgs) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01001367 if (!signal_pending(current))
1368 dev_warn(DEV,
1369 "short read receiving data reply digest: read %d expected %d\n",
1370 rr, dgs);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001371 return 0;
1372 }
1373 }
1374
1375 data_size -= dgs;
1376
1377 /* optimistically update recv_cnt. if receiving fails below,
1378 * we disconnect anyways, and counters will be reset. */
1379 mdev->recv_cnt += data_size>>9;
1380
1381 bio = req->master_bio;
1382 D_ASSERT(sector == bio->bi_sector);
1383
1384 bio_for_each_segment(bvec, bio, i) {
1385 expect = min_t(int, data_size, bvec->bv_len);
1386 rr = drbd_recv(mdev,
1387 kmap(bvec->bv_page)+bvec->bv_offset,
1388 expect);
1389 kunmap(bvec->bv_page);
1390 if (rr != expect) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01001391 if (!signal_pending(current))
1392 dev_warn(DEV, "short read receiving data reply: "
1393 "read %d expected %d\n",
1394 rr, expect);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001395 return 0;
1396 }
1397 data_size -= rr;
1398 }
1399
1400 if (dgs) {
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001401 drbd_csum_bio(mdev, mdev->integrity_r_tfm, bio, dig_vv);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001402 if (memcmp(dig_in, dig_vv, dgs)) {
1403 dev_err(DEV, "Digest integrity check FAILED. Broken NICs?\n");
1404 return 0;
1405 }
1406 }
1407
1408 D_ASSERT(data_size == 0);
1409 return 1;
1410}
1411
1412/* e_end_resync_block() is called via
1413 * drbd_process_done_ee() by asender only */
1414static int e_end_resync_block(struct drbd_conf *mdev, struct drbd_work *w, int unused)
1415{
1416 struct drbd_epoch_entry *e = (struct drbd_epoch_entry *)w;
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001417 sector_t sector = e->i.sector;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001418 int ok;
1419
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001420 D_ASSERT(drbd_interval_empty(&e->i));
Philipp Reisnerb411b362009-09-25 16:07:19 -07001421
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001422 if (likely((e->flags & EE_WAS_ERROR) == 0)) {
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001423 drbd_set_in_sync(mdev, sector, e->i.size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001424 ok = drbd_send_ack(mdev, P_RS_WRITE_ACK, e);
1425 } else {
1426 /* Record failure to sync */
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001427 drbd_rs_failed_io(mdev, sector, e->i.size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001428
1429 ok = drbd_send_ack(mdev, P_NEG_ACK, e);
1430 }
1431 dec_unacked(mdev);
1432
1433 return ok;
1434}
1435
1436static int recv_resync_read(struct drbd_conf *mdev, sector_t sector, int data_size) __releases(local)
1437{
1438 struct drbd_epoch_entry *e;
1439
1440 e = read_in_block(mdev, ID_SYNCER, sector, data_size);
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001441 if (!e)
1442 goto fail;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001443
1444 dec_rs_pending(mdev);
1445
Philipp Reisnerb411b362009-09-25 16:07:19 -07001446 inc_unacked(mdev);
1447 /* corresponding dec_unacked() in e_end_resync_block()
1448 * respective _drbd_clear_done_ee */
1449
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001450 e->w.cb = e_end_resync_block;
1451
Philipp Reisnerb411b362009-09-25 16:07:19 -07001452 spin_lock_irq(&mdev->req_lock);
1453 list_add(&e->w.list, &mdev->sync_ee);
1454 spin_unlock_irq(&mdev->req_lock);
1455
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001456 atomic_add(data_size >> 9, &mdev->rs_sect_ev);
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001457 if (drbd_submit_ee(mdev, e, WRITE, DRBD_FAULT_RS_WR) == 0)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001458 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001459
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001460 /* don't care for the reason here */
1461 dev_err(DEV, "submit failed, triggering re-connect\n");
Lars Ellenberg22cc37a2010-09-14 20:40:41 +02001462 spin_lock_irq(&mdev->req_lock);
1463 list_del(&e->w.list);
1464 spin_unlock_irq(&mdev->req_lock);
1465
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001466 drbd_free_ee(mdev, e);
1467fail:
1468 put_ldev(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001469 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001470}
1471
Andreas Gruenbacher668eebc2011-01-20 17:14:26 +01001472static struct drbd_request *
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01001473find_request(struct drbd_conf *mdev, struct rb_root *root, u64 id,
1474 sector_t sector, bool missing_ok, const char *func)
Andreas Gruenbacher668eebc2011-01-20 17:14:26 +01001475{
Andreas Gruenbacher668eebc2011-01-20 17:14:26 +01001476 struct drbd_request *req;
1477
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01001478 /* Request object according to our peer */
1479 req = (struct drbd_request *)(unsigned long)id;
1480 if (drbd_contains_interval(root, sector, &req->i))
Andreas Gruenbacher668eebc2011-01-20 17:14:26 +01001481 return req;
Andreas Gruenbacherc3afd8f2011-01-20 22:25:40 +01001482 if (!missing_ok) {
1483 dev_err(DEV, "%s: failed to find request %lu, sector %llus\n", func,
1484 (unsigned long)id, (unsigned long long)sector);
1485 }
Andreas Gruenbacher668eebc2011-01-20 17:14:26 +01001486 return NULL;
1487}
1488
Philipp Reisner02918be2010-08-20 14:35:10 +02001489static int receive_DataReply(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07001490{
1491 struct drbd_request *req;
1492 sector_t sector;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001493 int ok;
Philipp Reisner02918be2010-08-20 14:35:10 +02001494 struct p_data *p = &mdev->data.rbuf.data;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001495
1496 sector = be64_to_cpu(p->sector);
1497
1498 spin_lock_irq(&mdev->req_lock);
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01001499 req = find_request(mdev, &mdev->read_requests, p->block_id, sector, false, __func__);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001500 spin_unlock_irq(&mdev->req_lock);
Andreas Gruenbacherc3afd8f2011-01-20 22:25:40 +01001501 if (unlikely(!req))
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001502 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001503
Bart Van Assche24c48302011-05-21 18:32:29 +02001504 /* hlist_del(&req->collision) is done in _req_may_be_done, to avoid
Philipp Reisnerb411b362009-09-25 16:07:19 -07001505 * special casing it there for the various failure cases.
1506 * still no race with drbd_fail_pending_reads */
1507 ok = recv_dless_read(mdev, req, sector, data_size);
1508
1509 if (ok)
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01001510 req_mod(req, DATA_RECEIVED);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001511 /* else: nothing. handled from drbd_disconnect...
1512 * I don't think we may complete this just yet
1513 * in case we are "on-disconnect: freeze" */
1514
1515 return ok;
1516}
1517
Philipp Reisner02918be2010-08-20 14:35:10 +02001518static int receive_RSDataReply(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07001519{
1520 sector_t sector;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001521 int ok;
Philipp Reisner02918be2010-08-20 14:35:10 +02001522 struct p_data *p = &mdev->data.rbuf.data;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001523
1524 sector = be64_to_cpu(p->sector);
1525 D_ASSERT(p->block_id == ID_SYNCER);
1526
1527 if (get_ldev(mdev)) {
1528 /* data is submitted to disk within recv_resync_read.
1529 * corresponding put_ldev done below on error,
Andreas Gruenbacher9c508422011-01-14 21:19:36 +01001530 * or in drbd_endio_sec. */
Philipp Reisnerb411b362009-09-25 16:07:19 -07001531 ok = recv_resync_read(mdev, sector, data_size);
1532 } else {
1533 if (__ratelimit(&drbd_ratelimit_state))
1534 dev_err(DEV, "Can not write resync data to local disk.\n");
1535
1536 ok = drbd_drain_block(mdev, data_size);
1537
Lars Ellenberg2b2bf212010-10-06 11:46:55 +02001538 drbd_send_ack_dp(mdev, P_NEG_ACK, p, data_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001539 }
1540
Philipp Reisner778f2712010-07-06 11:14:00 +02001541 atomic_add(data_size >> 9, &mdev->rs_sect_in);
1542
Philipp Reisnerb411b362009-09-25 16:07:19 -07001543 return ok;
1544}
1545
1546/* e_end_block() is called via drbd_process_done_ee().
1547 * this means this function only runs in the asender thread
1548 */
1549static int e_end_block(struct drbd_conf *mdev, struct drbd_work *w, int cancel)
1550{
1551 struct drbd_epoch_entry *e = (struct drbd_epoch_entry *)w;
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001552 sector_t sector = e->i.sector;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001553 int ok = 1, pcmd;
1554
Philipp Reisnerb411b362009-09-25 16:07:19 -07001555 if (mdev->net_conf->wire_protocol == DRBD_PROT_C) {
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001556 if (likely((e->flags & EE_WAS_ERROR) == 0)) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001557 pcmd = (mdev->state.conn >= C_SYNC_SOURCE &&
1558 mdev->state.conn <= C_PAUSED_SYNC_T &&
1559 e->flags & EE_MAY_SET_IN_SYNC) ?
1560 P_RS_WRITE_ACK : P_WRITE_ACK;
1561 ok &= drbd_send_ack(mdev, pcmd, e);
1562 if (pcmd == P_RS_WRITE_ACK)
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001563 drbd_set_in_sync(mdev, sector, e->i.size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001564 } else {
1565 ok = drbd_send_ack(mdev, P_NEG_ACK, e);
1566 /* we expect it to be marked out of sync anyways...
1567 * maybe assert this? */
1568 }
1569 dec_unacked(mdev);
1570 }
1571 /* we delete from the conflict detection hash _after_ we sent out the
1572 * P_WRITE_ACK / P_NEG_ACK, to get the sequence number right. */
1573 if (mdev->net_conf->two_primaries) {
1574 spin_lock_irq(&mdev->req_lock);
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001575 D_ASSERT(!drbd_interval_empty(&e->i));
1576 drbd_remove_interval(&mdev->epoch_entries, &e->i);
1577 drbd_clear_interval(&e->i);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001578 spin_unlock_irq(&mdev->req_lock);
Andreas Gruenbacherbb3bfe92011-01-21 15:59:23 +01001579 } else
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001580 D_ASSERT(drbd_interval_empty(&e->i));
Philipp Reisnerb411b362009-09-25 16:07:19 -07001581
1582 drbd_may_finish_epoch(mdev, e->epoch, EV_PUT + (cancel ? EV_CLEANUP : 0));
1583
1584 return ok;
1585}
1586
1587static int e_send_discard_ack(struct drbd_conf *mdev, struct drbd_work *w, int unused)
1588{
1589 struct drbd_epoch_entry *e = (struct drbd_epoch_entry *)w;
1590 int ok = 1;
1591
1592 D_ASSERT(mdev->net_conf->wire_protocol == DRBD_PROT_C);
1593 ok = drbd_send_ack(mdev, P_DISCARD_ACK, e);
1594
1595 spin_lock_irq(&mdev->req_lock);
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001596 D_ASSERT(!drbd_interval_empty(&e->i));
1597 drbd_remove_interval(&mdev->epoch_entries, &e->i);
1598 drbd_clear_interval(&e->i);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001599 spin_unlock_irq(&mdev->req_lock);
1600
1601 dec_unacked(mdev);
1602
1603 return ok;
1604}
1605
1606/* Called from receive_Data.
1607 * Synchronize packets on sock with packets on msock.
1608 *
1609 * This is here so even when a P_DATA packet traveling via sock overtook an Ack
1610 * packet traveling on msock, they are still processed in the order they have
1611 * been sent.
1612 *
1613 * Note: we don't care for Ack packets overtaking P_DATA packets.
1614 *
1615 * In case packet_seq is larger than mdev->peer_seq number, there are
1616 * outstanding packets on the msock. We wait for them to arrive.
1617 * In case we are the logically next packet, we update mdev->peer_seq
1618 * ourselves. Correctly handles 32bit wrap around.
1619 *
1620 * Assume we have a 10 GBit connection, that is about 1<<30 byte per second,
1621 * about 1<<21 sectors per second. So "worst" case, we have 1<<3 == 8 seconds
1622 * for the 24bit wrap (historical atomic_t guarantee on some archs), and we have
1623 * 1<<9 == 512 seconds aka ages for the 32bit wrap around...
1624 *
1625 * returns 0 if we may process the packet,
1626 * -ERESTARTSYS if we were interrupted (by disconnect signal). */
1627static int drbd_wait_peer_seq(struct drbd_conf *mdev, const u32 packet_seq)
1628{
1629 DEFINE_WAIT(wait);
1630 unsigned int p_seq;
1631 long timeout;
1632 int ret = 0;
1633 spin_lock(&mdev->peer_seq_lock);
1634 for (;;) {
1635 prepare_to_wait(&mdev->seq_wait, &wait, TASK_INTERRUPTIBLE);
1636 if (seq_le(packet_seq, mdev->peer_seq+1))
1637 break;
1638 if (signal_pending(current)) {
1639 ret = -ERESTARTSYS;
1640 break;
1641 }
1642 p_seq = mdev->peer_seq;
1643 spin_unlock(&mdev->peer_seq_lock);
1644 timeout = schedule_timeout(30*HZ);
1645 spin_lock(&mdev->peer_seq_lock);
1646 if (timeout == 0 && p_seq == mdev->peer_seq) {
1647 ret = -ETIMEDOUT;
1648 dev_err(DEV, "ASSERT FAILED waited 30 seconds for sequence update, forcing reconnect\n");
1649 break;
1650 }
1651 }
1652 finish_wait(&mdev->seq_wait, &wait);
1653 if (mdev->peer_seq+1 == packet_seq)
1654 mdev->peer_seq++;
1655 spin_unlock(&mdev->peer_seq_lock);
1656 return ret;
1657}
1658
Lars Ellenberg688593c2010-11-17 22:25:03 +01001659/* see also bio_flags_to_wire()
1660 * DRBD_REQ_*, because we need to semantically map the flags to data packet
1661 * flags and back. We may replicate to other kernel versions. */
1662static unsigned long wire_flags_to_bio(struct drbd_conf *mdev, u32 dpf)
Philipp Reisner76d2e7e2010-08-25 11:58:05 +02001663{
Lars Ellenberg688593c2010-11-17 22:25:03 +01001664 return (dpf & DP_RW_SYNC ? REQ_SYNC : 0) |
1665 (dpf & DP_FUA ? REQ_FUA : 0) |
1666 (dpf & DP_FLUSH ? REQ_FLUSH : 0) |
1667 (dpf & DP_DISCARD ? REQ_DISCARD : 0);
Philipp Reisner76d2e7e2010-08-25 11:58:05 +02001668}
1669
Philipp Reisnerb411b362009-09-25 16:07:19 -07001670/* mirrored write */
Philipp Reisner02918be2010-08-20 14:35:10 +02001671static int receive_Data(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07001672{
1673 sector_t sector;
1674 struct drbd_epoch_entry *e;
Philipp Reisner02918be2010-08-20 14:35:10 +02001675 struct p_data *p = &mdev->data.rbuf.data;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001676 int rw = WRITE;
1677 u32 dp_flags;
1678
Philipp Reisnerb411b362009-09-25 16:07:19 -07001679 if (!get_ldev(mdev)) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001680 spin_lock(&mdev->peer_seq_lock);
1681 if (mdev->peer_seq+1 == be32_to_cpu(p->seq_num))
1682 mdev->peer_seq++;
1683 spin_unlock(&mdev->peer_seq_lock);
1684
Lars Ellenberg2b2bf212010-10-06 11:46:55 +02001685 drbd_send_ack_dp(mdev, P_NEG_ACK, p, data_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001686 atomic_inc(&mdev->current_epoch->epoch_size);
1687 return drbd_drain_block(mdev, data_size);
1688 }
1689
1690 /* get_ldev(mdev) successful.
1691 * Corresponding put_ldev done either below (on various errors),
Andreas Gruenbacher9c508422011-01-14 21:19:36 +01001692 * or in drbd_endio_sec, if we successfully submit the data at
Philipp Reisnerb411b362009-09-25 16:07:19 -07001693 * the end of this function. */
1694
1695 sector = be64_to_cpu(p->sector);
1696 e = read_in_block(mdev, p->block_id, sector, data_size);
1697 if (!e) {
1698 put_ldev(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001699 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001700 }
1701
Philipp Reisnerb411b362009-09-25 16:07:19 -07001702 e->w.cb = e_end_block;
1703
Lars Ellenberg688593c2010-11-17 22:25:03 +01001704 dp_flags = be32_to_cpu(p->dp_flags);
1705 rw |= wire_flags_to_bio(mdev, dp_flags);
1706
1707 if (dp_flags & DP_MAY_SET_IN_SYNC)
1708 e->flags |= EE_MAY_SET_IN_SYNC;
1709
Philipp Reisnerb411b362009-09-25 16:07:19 -07001710 spin_lock(&mdev->epoch_lock);
1711 e->epoch = mdev->current_epoch;
1712 atomic_inc(&e->epoch->epoch_size);
1713 atomic_inc(&e->epoch->active);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001714 spin_unlock(&mdev->epoch_lock);
1715
Philipp Reisnerb411b362009-09-25 16:07:19 -07001716 /* I'm the receiver, I do hold a net_cnt reference. */
1717 if (!mdev->net_conf->two_primaries) {
1718 spin_lock_irq(&mdev->req_lock);
1719 } else {
1720 /* don't get the req_lock yet,
1721 * we may sleep in drbd_wait_peer_seq */
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001722 const int size = e->i.size;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001723 const int discard = test_bit(DISCARD_CONCURRENT, &mdev->flags);
1724 DEFINE_WAIT(wait);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001725 int first;
1726
1727 D_ASSERT(mdev->net_conf->wire_protocol == DRBD_PROT_C);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001728
1729 /* conflict detection and handling:
1730 * 1. wait on the sequence number,
1731 * in case this data packet overtook ACK packets.
Andreas Gruenbacherbb3bfe92011-01-21 15:59:23 +01001732 * 2. check our interval trees for conflicting requests:
1733 * we only need to check the write_requests tree; the
1734 * epoch_entries tree cannot contain any overlaps because
1735 * they were already eliminated on the submitting node.
Philipp Reisnerb411b362009-09-25 16:07:19 -07001736 *
1737 * Note: for two_primaries, we are protocol C,
1738 * so there cannot be any request that is DONE
1739 * but still on the transfer log.
1740 *
Andreas Gruenbacherbb3bfe92011-01-21 15:59:23 +01001741 * unconditionally add to the epoch_entries tree.
Philipp Reisnerb411b362009-09-25 16:07:19 -07001742 *
1743 * if no conflicting request is found:
1744 * submit.
1745 *
1746 * if any conflicting request is found
1747 * that has not yet been acked,
1748 * AND I have the "discard concurrent writes" flag:
1749 * queue (via done_ee) the P_DISCARD_ACK; OUT.
1750 *
1751 * if any conflicting request is found:
1752 * block the receiver, waiting on misc_wait
1753 * until no more conflicting requests are there,
1754 * or we get interrupted (disconnect).
1755 *
1756 * we do not just write after local io completion of those
1757 * requests, but only after req is done completely, i.e.
1758 * we wait for the P_DISCARD_ACK to arrive!
1759 *
1760 * then proceed normally, i.e. submit.
1761 */
1762 if (drbd_wait_peer_seq(mdev, be32_to_cpu(p->seq_num)))
1763 goto out_interrupted;
1764
1765 spin_lock_irq(&mdev->req_lock);
1766
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001767 drbd_insert_interval(&mdev->epoch_entries, &e->i);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001768
Philipp Reisnerb411b362009-09-25 16:07:19 -07001769 first = 1;
1770 for (;;) {
Andreas Gruenbacherde696712011-01-20 15:00:24 +01001771 struct drbd_interval *i;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001772 int have_unacked = 0;
1773 int have_conflict = 0;
1774 prepare_to_wait(&mdev->misc_wait, &wait,
1775 TASK_INTERRUPTIBLE);
Andreas Gruenbacherde696712011-01-20 15:00:24 +01001776
1777 i = drbd_find_overlap(&mdev->write_requests, sector, size);
1778 if (i) {
1779 struct drbd_request *req2 =
1780 container_of(i, struct drbd_request, i);
1781
1782 /* only ALERT on first iteration,
1783 * we may be woken up early... */
1784 if (first)
1785 dev_alert(DEV, "%s[%u] Concurrent local write detected!"
1786 " new: %llus +%u; pending: %llus +%u\n",
1787 current->comm, current->pid,
1788 (unsigned long long)sector, size,
1789 (unsigned long long)req2->i.sector, req2->i.size);
1790 if (req2->rq_state & RQ_NET_PENDING)
1791 ++have_unacked;
1792 ++have_conflict;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001793 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07001794 if (!have_conflict)
1795 break;
1796
1797 /* Discard Ack only for the _first_ iteration */
1798 if (first && discard && have_unacked) {
1799 dev_alert(DEV, "Concurrent write! [DISCARD BY FLAG] sec=%llus\n",
1800 (unsigned long long)sector);
1801 inc_unacked(mdev);
1802 e->w.cb = e_send_discard_ack;
1803 list_add_tail(&e->w.list, &mdev->done_ee);
1804
1805 spin_unlock_irq(&mdev->req_lock);
1806
1807 /* we could probably send that P_DISCARD_ACK ourselves,
1808 * but I don't like the receiver using the msock */
1809
1810 put_ldev(mdev);
1811 wake_asender(mdev);
1812 finish_wait(&mdev->misc_wait, &wait);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001813 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001814 }
1815
1816 if (signal_pending(current)) {
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001817 drbd_remove_interval(&mdev->epoch_entries, &e->i);
1818 drbd_clear_interval(&e->i);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001819
1820 spin_unlock_irq(&mdev->req_lock);
1821
1822 finish_wait(&mdev->misc_wait, &wait);
1823 goto out_interrupted;
1824 }
1825
1826 spin_unlock_irq(&mdev->req_lock);
1827 if (first) {
1828 first = 0;
1829 dev_alert(DEV, "Concurrent write! [W AFTERWARDS] "
1830 "sec=%llus\n", (unsigned long long)sector);
1831 } else if (discard) {
1832 /* we had none on the first iteration.
1833 * there must be none now. */
1834 D_ASSERT(have_unacked == 0);
1835 }
1836 schedule();
1837 spin_lock_irq(&mdev->req_lock);
1838 }
1839 finish_wait(&mdev->misc_wait, &wait);
1840 }
1841
1842 list_add(&e->w.list, &mdev->active_ee);
1843 spin_unlock_irq(&mdev->req_lock);
1844
1845 switch (mdev->net_conf->wire_protocol) {
1846 case DRBD_PROT_C:
1847 inc_unacked(mdev);
1848 /* corresponding dec_unacked() in e_end_block()
1849 * respective _drbd_clear_done_ee */
1850 break;
1851 case DRBD_PROT_B:
1852 /* I really don't like it that the receiver thread
1853 * sends on the msock, but anyways */
1854 drbd_send_ack(mdev, P_RECV_ACK, e);
1855 break;
1856 case DRBD_PROT_A:
1857 /* nothing to do */
1858 break;
1859 }
1860
Lars Ellenberg6719fb02010-10-18 23:04:07 +02001861 if (mdev->state.pdsk < D_INCONSISTENT) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001862 /* In case we have the only disk of the cluster, */
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001863 drbd_set_out_of_sync(mdev, e->i.sector, e->i.size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001864 e->flags |= EE_CALL_AL_COMPLETE_IO;
Lars Ellenberg6719fb02010-10-18 23:04:07 +02001865 e->flags &= ~EE_MAY_SET_IN_SYNC;
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001866 drbd_al_begin_io(mdev, e->i.sector);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001867 }
1868
Lars Ellenberg45bb9122010-05-14 17:10:48 +02001869 if (drbd_submit_ee(mdev, e, rw, DRBD_FAULT_DT_WR) == 0)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001870 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001871
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001872 /* don't care for the reason here */
1873 dev_err(DEV, "submit failed, triggering re-connect\n");
Lars Ellenberg22cc37a2010-09-14 20:40:41 +02001874 spin_lock_irq(&mdev->req_lock);
1875 list_del(&e->w.list);
Andreas Gruenbacher8b946252011-01-20 15:23:07 +01001876 drbd_remove_interval(&mdev->epoch_entries, &e->i);
1877 drbd_clear_interval(&e->i);
Lars Ellenberg22cc37a2010-09-14 20:40:41 +02001878 spin_unlock_irq(&mdev->req_lock);
1879 if (e->flags & EE_CALL_AL_COMPLETE_IO)
Andreas Gruenbacher010f6e62011-01-14 20:59:35 +01001880 drbd_al_complete_io(mdev, e->i.sector);
Lars Ellenberg22cc37a2010-09-14 20:40:41 +02001881
Philipp Reisnerb411b362009-09-25 16:07:19 -07001882out_interrupted:
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01001883 drbd_may_finish_epoch(mdev, e->epoch, EV_PUT + EV_CLEANUP);
Philipp Reisnerb411b362009-09-25 16:07:19 -07001884 put_ldev(mdev);
1885 drbd_free_ee(mdev, e);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001886 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001887}
1888
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001889/* We may throttle resync, if the lower device seems to be busy,
1890 * and current sync rate is above c_min_rate.
1891 *
1892 * To decide whether or not the lower device is busy, we use a scheme similar
1893 * to MD RAID is_mddev_idle(): if the partition stats reveal "significant"
1894 * (more than 64 sectors) of activity we cannot account for with our own resync
1895 * activity, it obviously is "busy".
1896 *
1897 * The current sync rate used here uses only the most recent two step marks,
1898 * to have a short time average so we can react faster.
1899 */
Philipp Reisnere3555d82010-11-07 15:56:29 +01001900int drbd_rs_should_slow_down(struct drbd_conf *mdev, sector_t sector)
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001901{
1902 struct gendisk *disk = mdev->ldev->backing_bdev->bd_contains->bd_disk;
1903 unsigned long db, dt, dbdt;
Philipp Reisnere3555d82010-11-07 15:56:29 +01001904 struct lc_element *tmp;
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001905 int curr_events;
1906 int throttle = 0;
1907
1908 /* feature disabled? */
1909 if (mdev->sync_conf.c_min_rate == 0)
1910 return 0;
1911
Philipp Reisnere3555d82010-11-07 15:56:29 +01001912 spin_lock_irq(&mdev->al_lock);
1913 tmp = lc_find(mdev->resync, BM_SECT_TO_EXT(sector));
1914 if (tmp) {
1915 struct bm_extent *bm_ext = lc_entry(tmp, struct bm_extent, lce);
1916 if (test_bit(BME_PRIORITY, &bm_ext->flags)) {
1917 spin_unlock_irq(&mdev->al_lock);
1918 return 0;
1919 }
1920 /* Do not slow down if app IO is already waiting for this extent */
1921 }
1922 spin_unlock_irq(&mdev->al_lock);
1923
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001924 curr_events = (int)part_stat_read(&disk->part0, sectors[0]) +
1925 (int)part_stat_read(&disk->part0, sectors[1]) -
1926 atomic_read(&mdev->rs_sect_ev);
Philipp Reisnere3555d82010-11-07 15:56:29 +01001927
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001928 if (!mdev->rs_last_events || curr_events - mdev->rs_last_events > 64) {
1929 unsigned long rs_left;
1930 int i;
1931
1932 mdev->rs_last_events = curr_events;
1933
1934 /* sync speed average over the last 2*DRBD_SYNC_MARK_STEP,
1935 * approx. */
Lars Ellenberg2649f082010-11-05 10:05:47 +01001936 i = (mdev->rs_last_mark + DRBD_SYNC_MARKS-1) % DRBD_SYNC_MARKS;
1937
1938 if (mdev->state.conn == C_VERIFY_S || mdev->state.conn == C_VERIFY_T)
1939 rs_left = mdev->ov_left;
1940 else
1941 rs_left = drbd_bm_total_weight(mdev) - mdev->rs_failed;
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02001942
1943 dt = ((long)jiffies - (long)mdev->rs_mark_time[i]) / HZ;
1944 if (!dt)
1945 dt++;
1946 db = mdev->rs_mark_left[i] - rs_left;
1947 dbdt = Bit2KB(db/dt);
1948
1949 if (dbdt > mdev->sync_conf.c_min_rate)
1950 throttle = 1;
1951 }
1952 return throttle;
1953}
1954
1955
Philipp Reisner02918be2010-08-20 14:35:10 +02001956static int receive_DataRequest(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int digest_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07001957{
1958 sector_t sector;
1959 const sector_t capacity = drbd_get_capacity(mdev->this_bdev);
1960 struct drbd_epoch_entry *e;
1961 struct digest_info *di = NULL;
Philipp Reisnerb18b37b2010-10-13 15:32:44 +02001962 int size, verb;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001963 unsigned int fault_type;
Philipp Reisner02918be2010-08-20 14:35:10 +02001964 struct p_block_req *p = &mdev->data.rbuf.block_req;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001965
1966 sector = be64_to_cpu(p->sector);
1967 size = be32_to_cpu(p->blksize);
1968
Lars Ellenberg1816a2b2010-11-11 15:19:07 +01001969 if (size <= 0 || (size & 0x1ff) != 0 || size > DRBD_MAX_BIO_SIZE) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07001970 dev_err(DEV, "%s:%d: sector: %llus, size: %u\n", __FILE__, __LINE__,
1971 (unsigned long long)sector, size);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001972 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001973 }
1974 if (sector + (size>>9) > capacity) {
1975 dev_err(DEV, "%s:%d: sector: %llus, size: %u\n", __FILE__, __LINE__,
1976 (unsigned long long)sector, size);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01001977 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07001978 }
1979
1980 if (!get_ldev_if_state(mdev, D_UP_TO_DATE)) {
Philipp Reisnerb18b37b2010-10-13 15:32:44 +02001981 verb = 1;
1982 switch (cmd) {
1983 case P_DATA_REQUEST:
1984 drbd_send_ack_rp(mdev, P_NEG_DREPLY, p);
1985 break;
1986 case P_RS_DATA_REQUEST:
1987 case P_CSUM_RS_REQUEST:
1988 case P_OV_REQUEST:
1989 drbd_send_ack_rp(mdev, P_NEG_RS_DREPLY , p);
1990 break;
1991 case P_OV_REPLY:
1992 verb = 0;
1993 dec_rs_pending(mdev);
1994 drbd_send_ack_ex(mdev, P_OV_RESULT, sector, size, ID_IN_SYNC);
1995 break;
1996 default:
1997 dev_err(DEV, "unexpected command (%s) in receive_DataRequest\n",
1998 cmdname(cmd));
1999 }
2000 if (verb && __ratelimit(&drbd_ratelimit_state))
Philipp Reisnerb411b362009-09-25 16:07:19 -07002001 dev_err(DEV, "Can not satisfy peer's read request, "
2002 "no local data.\n");
Philipp Reisnerb18b37b2010-10-13 15:32:44 +02002003
Lars Ellenberga821cc42010-09-06 12:31:37 +02002004 /* drain possibly payload */
2005 return drbd_drain_block(mdev, digest_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002006 }
2007
2008 /* GFP_NOIO, because we must not cause arbitrary write-out: in a DRBD
2009 * "criss-cross" setup, that might cause write-out on some other DRBD,
2010 * which in turn might block on the other node at this very place. */
2011 e = drbd_alloc_ee(mdev, p->block_id, sector, size, GFP_NOIO);
2012 if (!e) {
2013 put_ldev(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002014 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002015 }
2016
Philipp Reisner02918be2010-08-20 14:35:10 +02002017 switch (cmd) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002018 case P_DATA_REQUEST:
2019 e->w.cb = w_e_end_data_req;
2020 fault_type = DRBD_FAULT_DT_RD;
Lars Ellenberg80a40e42010-08-11 23:28:00 +02002021 /* application IO, don't drbd_rs_begin_io */
2022 goto submit;
2023
Philipp Reisnerb411b362009-09-25 16:07:19 -07002024 case P_RS_DATA_REQUEST:
2025 e->w.cb = w_e_end_rsdata_req;
2026 fault_type = DRBD_FAULT_RS_RD;
Lars Ellenberg5f9915b2010-11-09 14:15:24 +01002027 /* used in the sector offset progress display */
2028 mdev->bm_resync_fo = BM_SECT_TO_BIT(sector);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002029 break;
2030
2031 case P_OV_REPLY:
2032 case P_CSUM_RS_REQUEST:
2033 fault_type = DRBD_FAULT_RS_RD;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002034 di = kmalloc(sizeof(*di) + digest_size, GFP_NOIO);
2035 if (!di)
2036 goto out_free_e;
2037
2038 di->digest_size = digest_size;
2039 di->digest = (((char *)di)+sizeof(struct digest_info));
2040
Lars Ellenbergc36c3ce2010-08-11 20:42:55 +02002041 e->digest = di;
2042 e->flags |= EE_HAS_DIGEST;
2043
Philipp Reisnerb411b362009-09-25 16:07:19 -07002044 if (drbd_recv(mdev, di->digest, digest_size) != digest_size)
2045 goto out_free_e;
2046
Philipp Reisner02918be2010-08-20 14:35:10 +02002047 if (cmd == P_CSUM_RS_REQUEST) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002048 D_ASSERT(mdev->agreed_pro_version >= 89);
2049 e->w.cb = w_e_end_csum_rs_req;
Lars Ellenberg5f9915b2010-11-09 14:15:24 +01002050 /* used in the sector offset progress display */
2051 mdev->bm_resync_fo = BM_SECT_TO_BIT(sector);
Philipp Reisner02918be2010-08-20 14:35:10 +02002052 } else if (cmd == P_OV_REPLY) {
Lars Ellenberg2649f082010-11-05 10:05:47 +01002053 /* track progress, we may need to throttle */
2054 atomic_add(size >> 9, &mdev->rs_sect_in);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002055 e->w.cb = w_e_end_ov_reply;
2056 dec_rs_pending(mdev);
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02002057 /* drbd_rs_begin_io done when we sent this request,
2058 * but accounting still needs to be done. */
2059 goto submit_for_resync;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002060 }
2061 break;
2062
2063 case P_OV_REQUEST:
Philipp Reisnerb411b362009-09-25 16:07:19 -07002064 if (mdev->ov_start_sector == ~(sector_t)0 &&
2065 mdev->agreed_pro_version >= 90) {
Lars Ellenbergde228bb2010-11-05 09:43:15 +01002066 unsigned long now = jiffies;
2067 int i;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002068 mdev->ov_start_sector = sector;
2069 mdev->ov_position = sector;
Lars Ellenberg30b743a2010-11-05 09:39:06 +01002070 mdev->ov_left = drbd_bm_bits(mdev) - BM_SECT_TO_BIT(sector);
2071 mdev->rs_total = mdev->ov_left;
Lars Ellenbergde228bb2010-11-05 09:43:15 +01002072 for (i = 0; i < DRBD_SYNC_MARKS; i++) {
2073 mdev->rs_mark_left[i] = mdev->ov_left;
2074 mdev->rs_mark_time[i] = now;
2075 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07002076 dev_info(DEV, "Online Verify start sector: %llu\n",
2077 (unsigned long long)sector);
2078 }
2079 e->w.cb = w_e_end_ov_req;
2080 fault_type = DRBD_FAULT_RS_RD;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002081 break;
2082
Philipp Reisnerb411b362009-09-25 16:07:19 -07002083 default:
2084 dev_err(DEV, "unexpected command (%s) in receive_DataRequest\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02002085 cmdname(cmd));
Philipp Reisnerb411b362009-09-25 16:07:19 -07002086 fault_type = DRBD_FAULT_MAX;
Lars Ellenberg80a40e42010-08-11 23:28:00 +02002087 goto out_free_e;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002088 }
2089
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02002090 /* Throttle, drbd_rs_begin_io and submit should become asynchronous
2091 * wrt the receiver, but it is not as straightforward as it may seem.
2092 * Various places in the resync start and stop logic assume resync
2093 * requests are processed in order, requeuing this on the worker thread
2094 * introduces a bunch of new code for synchronization between threads.
2095 *
2096 * Unlimited throttling before drbd_rs_begin_io may stall the resync
2097 * "forever", throttling after drbd_rs_begin_io will lock that extent
2098 * for application writes for the same time. For now, just throttle
2099 * here, where the rest of the code expects the receiver to sleep for
2100 * a while, anyways.
2101 */
Philipp Reisnerb411b362009-09-25 16:07:19 -07002102
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02002103 /* Throttle before drbd_rs_begin_io, as that locks out application IO;
2104 * this defers syncer requests for some time, before letting at least
2105 * on request through. The resync controller on the receiving side
2106 * will adapt to the incoming rate accordingly.
2107 *
2108 * We cannot throttle here if remote is Primary/SyncTarget:
2109 * we would also throttle its application reads.
2110 * In that case, throttling is done on the SyncTarget only.
2111 */
Philipp Reisnere3555d82010-11-07 15:56:29 +01002112 if (mdev->state.peer != R_PRIMARY && drbd_rs_should_slow_down(mdev, sector))
2113 schedule_timeout_uninterruptible(HZ/10);
2114 if (drbd_rs_begin_io(mdev, sector))
Lars Ellenberg80a40e42010-08-11 23:28:00 +02002115 goto out_free_e;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002116
Lars Ellenberg0f0601f2010-08-11 23:40:24 +02002117submit_for_resync:
2118 atomic_add(size >> 9, &mdev->rs_sect_ev);
2119
Lars Ellenberg80a40e42010-08-11 23:28:00 +02002120submit:
Philipp Reisnerb411b362009-09-25 16:07:19 -07002121 inc_unacked(mdev);
Lars Ellenberg80a40e42010-08-11 23:28:00 +02002122 spin_lock_irq(&mdev->req_lock);
2123 list_add_tail(&e->w.list, &mdev->read_ee);
2124 spin_unlock_irq(&mdev->req_lock);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002125
Lars Ellenberg45bb9122010-05-14 17:10:48 +02002126 if (drbd_submit_ee(mdev, e, READ, fault_type) == 0)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002127 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002128
Lars Ellenberg10f6d9922011-01-24 14:47:09 +01002129 /* don't care for the reason here */
2130 dev_err(DEV, "submit failed, triggering re-connect\n");
Lars Ellenberg22cc37a2010-09-14 20:40:41 +02002131 spin_lock_irq(&mdev->req_lock);
2132 list_del(&e->w.list);
2133 spin_unlock_irq(&mdev->req_lock);
2134 /* no drbd_rs_complete_io(), we are dropping the connection anyways */
2135
Philipp Reisnerb411b362009-09-25 16:07:19 -07002136out_free_e:
Philipp Reisnerb411b362009-09-25 16:07:19 -07002137 put_ldev(mdev);
2138 drbd_free_ee(mdev, e);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002139 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002140}
2141
2142static int drbd_asb_recover_0p(struct drbd_conf *mdev) __must_hold(local)
2143{
2144 int self, peer, rv = -100;
2145 unsigned long ch_self, ch_peer;
2146
2147 self = mdev->ldev->md.uuid[UI_BITMAP] & 1;
2148 peer = mdev->p_uuid[UI_BITMAP] & 1;
2149
2150 ch_peer = mdev->p_uuid[UI_SIZE];
2151 ch_self = mdev->comm_bm_set;
2152
2153 switch (mdev->net_conf->after_sb_0p) {
2154 case ASB_CONSENSUS:
2155 case ASB_DISCARD_SECONDARY:
2156 case ASB_CALL_HELPER:
2157 dev_err(DEV, "Configuration error.\n");
2158 break;
2159 case ASB_DISCONNECT:
2160 break;
2161 case ASB_DISCARD_YOUNGER_PRI:
2162 if (self == 0 && peer == 1) {
2163 rv = -1;
2164 break;
2165 }
2166 if (self == 1 && peer == 0) {
2167 rv = 1;
2168 break;
2169 }
2170 /* Else fall through to one of the other strategies... */
2171 case ASB_DISCARD_OLDER_PRI:
2172 if (self == 0 && peer == 1) {
2173 rv = 1;
2174 break;
2175 }
2176 if (self == 1 && peer == 0) {
2177 rv = -1;
2178 break;
2179 }
2180 /* Else fall through to one of the other strategies... */
Lars Ellenbergad19bf62009-10-14 09:36:49 +02002181 dev_warn(DEV, "Discard younger/older primary did not find a decision\n"
Philipp Reisnerb411b362009-09-25 16:07:19 -07002182 "Using discard-least-changes instead\n");
2183 case ASB_DISCARD_ZERO_CHG:
2184 if (ch_peer == 0 && ch_self == 0) {
2185 rv = test_bit(DISCARD_CONCURRENT, &mdev->flags)
2186 ? -1 : 1;
2187 break;
2188 } else {
2189 if (ch_peer == 0) { rv = 1; break; }
2190 if (ch_self == 0) { rv = -1; break; }
2191 }
2192 if (mdev->net_conf->after_sb_0p == ASB_DISCARD_ZERO_CHG)
2193 break;
2194 case ASB_DISCARD_LEAST_CHG:
2195 if (ch_self < ch_peer)
2196 rv = -1;
2197 else if (ch_self > ch_peer)
2198 rv = 1;
2199 else /* ( ch_self == ch_peer ) */
2200 /* Well, then use something else. */
2201 rv = test_bit(DISCARD_CONCURRENT, &mdev->flags)
2202 ? -1 : 1;
2203 break;
2204 case ASB_DISCARD_LOCAL:
2205 rv = -1;
2206 break;
2207 case ASB_DISCARD_REMOTE:
2208 rv = 1;
2209 }
2210
2211 return rv;
2212}
2213
2214static int drbd_asb_recover_1p(struct drbd_conf *mdev) __must_hold(local)
2215{
Andreas Gruenbacher6184ea22010-12-09 14:23:27 +01002216 int hg, rv = -100;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002217
2218 switch (mdev->net_conf->after_sb_1p) {
2219 case ASB_DISCARD_YOUNGER_PRI:
2220 case ASB_DISCARD_OLDER_PRI:
2221 case ASB_DISCARD_LEAST_CHG:
2222 case ASB_DISCARD_LOCAL:
2223 case ASB_DISCARD_REMOTE:
2224 dev_err(DEV, "Configuration error.\n");
2225 break;
2226 case ASB_DISCONNECT:
2227 break;
2228 case ASB_CONSENSUS:
2229 hg = drbd_asb_recover_0p(mdev);
2230 if (hg == -1 && mdev->state.role == R_SECONDARY)
2231 rv = hg;
2232 if (hg == 1 && mdev->state.role == R_PRIMARY)
2233 rv = hg;
2234 break;
2235 case ASB_VIOLENTLY:
2236 rv = drbd_asb_recover_0p(mdev);
2237 break;
2238 case ASB_DISCARD_SECONDARY:
2239 return mdev->state.role == R_PRIMARY ? 1 : -1;
2240 case ASB_CALL_HELPER:
2241 hg = drbd_asb_recover_0p(mdev);
2242 if (hg == -1 && mdev->state.role == R_PRIMARY) {
Andreas Gruenbacherbb437942010-12-09 14:02:35 +01002243 enum drbd_state_rv rv2;
2244
2245 drbd_set_role(mdev, R_SECONDARY, 0);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002246 /* drbd_change_state() does not sleep while in SS_IN_TRANSIENT_STATE,
2247 * we might be here in C_WF_REPORT_PARAMS which is transient.
2248 * we do not need to wait for the after state change work either. */
Andreas Gruenbacherbb437942010-12-09 14:02:35 +01002249 rv2 = drbd_change_state(mdev, CS_VERBOSE, NS(role, R_SECONDARY));
2250 if (rv2 != SS_SUCCESS) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002251 drbd_khelper(mdev, "pri-lost-after-sb");
2252 } else {
2253 dev_warn(DEV, "Successfully gave up primary role.\n");
2254 rv = hg;
2255 }
2256 } else
2257 rv = hg;
2258 }
2259
2260 return rv;
2261}
2262
2263static int drbd_asb_recover_2p(struct drbd_conf *mdev) __must_hold(local)
2264{
Andreas Gruenbacher6184ea22010-12-09 14:23:27 +01002265 int hg, rv = -100;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002266
2267 switch (mdev->net_conf->after_sb_2p) {
2268 case ASB_DISCARD_YOUNGER_PRI:
2269 case ASB_DISCARD_OLDER_PRI:
2270 case ASB_DISCARD_LEAST_CHG:
2271 case ASB_DISCARD_LOCAL:
2272 case ASB_DISCARD_REMOTE:
2273 case ASB_CONSENSUS:
2274 case ASB_DISCARD_SECONDARY:
2275 dev_err(DEV, "Configuration error.\n");
2276 break;
2277 case ASB_VIOLENTLY:
2278 rv = drbd_asb_recover_0p(mdev);
2279 break;
2280 case ASB_DISCONNECT:
2281 break;
2282 case ASB_CALL_HELPER:
2283 hg = drbd_asb_recover_0p(mdev);
2284 if (hg == -1) {
Andreas Gruenbacherbb437942010-12-09 14:02:35 +01002285 enum drbd_state_rv rv2;
2286
Philipp Reisnerb411b362009-09-25 16:07:19 -07002287 /* drbd_change_state() does not sleep while in SS_IN_TRANSIENT_STATE,
2288 * we might be here in C_WF_REPORT_PARAMS which is transient.
2289 * we do not need to wait for the after state change work either. */
Andreas Gruenbacherbb437942010-12-09 14:02:35 +01002290 rv2 = drbd_change_state(mdev, CS_VERBOSE, NS(role, R_SECONDARY));
2291 if (rv2 != SS_SUCCESS) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002292 drbd_khelper(mdev, "pri-lost-after-sb");
2293 } else {
2294 dev_warn(DEV, "Successfully gave up primary role.\n");
2295 rv = hg;
2296 }
2297 } else
2298 rv = hg;
2299 }
2300
2301 return rv;
2302}
2303
2304static void drbd_uuid_dump(struct drbd_conf *mdev, char *text, u64 *uuid,
2305 u64 bits, u64 flags)
2306{
2307 if (!uuid) {
2308 dev_info(DEV, "%s uuid info vanished while I was looking!\n", text);
2309 return;
2310 }
2311 dev_info(DEV, "%s %016llX:%016llX:%016llX:%016llX bits:%llu flags:%llX\n",
2312 text,
2313 (unsigned long long)uuid[UI_CURRENT],
2314 (unsigned long long)uuid[UI_BITMAP],
2315 (unsigned long long)uuid[UI_HISTORY_START],
2316 (unsigned long long)uuid[UI_HISTORY_END],
2317 (unsigned long long)bits,
2318 (unsigned long long)flags);
2319}
2320
2321/*
2322 100 after split brain try auto recover
2323 2 C_SYNC_SOURCE set BitMap
2324 1 C_SYNC_SOURCE use BitMap
2325 0 no Sync
2326 -1 C_SYNC_TARGET use BitMap
2327 -2 C_SYNC_TARGET set BitMap
2328 -100 after split brain, disconnect
2329-1000 unrelated data
Philipp Reisner4a23f262011-01-11 17:42:17 +01002330-1091 requires proto 91
2331-1096 requires proto 96
Philipp Reisnerb411b362009-09-25 16:07:19 -07002332 */
2333static int drbd_uuid_compare(struct drbd_conf *mdev, int *rule_nr) __must_hold(local)
2334{
2335 u64 self, peer;
2336 int i, j;
2337
2338 self = mdev->ldev->md.uuid[UI_CURRENT] & ~((u64)1);
2339 peer = mdev->p_uuid[UI_CURRENT] & ~((u64)1);
2340
2341 *rule_nr = 10;
2342 if (self == UUID_JUST_CREATED && peer == UUID_JUST_CREATED)
2343 return 0;
2344
2345 *rule_nr = 20;
2346 if ((self == UUID_JUST_CREATED || self == (u64)0) &&
2347 peer != UUID_JUST_CREATED)
2348 return -2;
2349
2350 *rule_nr = 30;
2351 if (self != UUID_JUST_CREATED &&
2352 (peer == UUID_JUST_CREATED || peer == (u64)0))
2353 return 2;
2354
2355 if (self == peer) {
2356 int rct, dc; /* roles at crash time */
2357
2358 if (mdev->p_uuid[UI_BITMAP] == (u64)0 && mdev->ldev->md.uuid[UI_BITMAP] != (u64)0) {
2359
2360 if (mdev->agreed_pro_version < 91)
Philipp Reisner4a23f262011-01-11 17:42:17 +01002361 return -1091;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002362
2363 if ((mdev->ldev->md.uuid[UI_BITMAP] & ~((u64)1)) == (mdev->p_uuid[UI_HISTORY_START] & ~((u64)1)) &&
2364 (mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1)) == (mdev->p_uuid[UI_HISTORY_START + 1] & ~((u64)1))) {
2365 dev_info(DEV, "was SyncSource, missed the resync finished event, corrected myself:\n");
2366 drbd_uuid_set_bm(mdev, 0UL);
2367
2368 drbd_uuid_dump(mdev, "self", mdev->ldev->md.uuid,
2369 mdev->state.disk >= D_NEGOTIATING ? drbd_bm_total_weight(mdev) : 0, 0);
2370 *rule_nr = 34;
2371 } else {
2372 dev_info(DEV, "was SyncSource (peer failed to write sync_uuid)\n");
2373 *rule_nr = 36;
2374 }
2375
2376 return 1;
2377 }
2378
2379 if (mdev->ldev->md.uuid[UI_BITMAP] == (u64)0 && mdev->p_uuid[UI_BITMAP] != (u64)0) {
2380
2381 if (mdev->agreed_pro_version < 91)
Philipp Reisner4a23f262011-01-11 17:42:17 +01002382 return -1091;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002383
2384 if ((mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1)) == (mdev->p_uuid[UI_BITMAP] & ~((u64)1)) &&
2385 (mdev->ldev->md.uuid[UI_HISTORY_START + 1] & ~((u64)1)) == (mdev->p_uuid[UI_HISTORY_START] & ~((u64)1))) {
2386 dev_info(DEV, "was SyncTarget, peer missed the resync finished event, corrected peer:\n");
2387
2388 mdev->p_uuid[UI_HISTORY_START + 1] = mdev->p_uuid[UI_HISTORY_START];
2389 mdev->p_uuid[UI_HISTORY_START] = mdev->p_uuid[UI_BITMAP];
2390 mdev->p_uuid[UI_BITMAP] = 0UL;
2391
2392 drbd_uuid_dump(mdev, "peer", mdev->p_uuid, mdev->p_uuid[UI_SIZE], mdev->p_uuid[UI_FLAGS]);
2393 *rule_nr = 35;
2394 } else {
2395 dev_info(DEV, "was SyncTarget (failed to write sync_uuid)\n");
2396 *rule_nr = 37;
2397 }
2398
2399 return -1;
2400 }
2401
2402 /* Common power [off|failure] */
2403 rct = (test_bit(CRASHED_PRIMARY, &mdev->flags) ? 1 : 0) +
2404 (mdev->p_uuid[UI_FLAGS] & 2);
2405 /* lowest bit is set when we were primary,
2406 * next bit (weight 2) is set when peer was primary */
2407 *rule_nr = 40;
2408
2409 switch (rct) {
2410 case 0: /* !self_pri && !peer_pri */ return 0;
2411 case 1: /* self_pri && !peer_pri */ return 1;
2412 case 2: /* !self_pri && peer_pri */ return -1;
2413 case 3: /* self_pri && peer_pri */
2414 dc = test_bit(DISCARD_CONCURRENT, &mdev->flags);
2415 return dc ? -1 : 1;
2416 }
2417 }
2418
2419 *rule_nr = 50;
2420 peer = mdev->p_uuid[UI_BITMAP] & ~((u64)1);
2421 if (self == peer)
2422 return -1;
2423
2424 *rule_nr = 51;
2425 peer = mdev->p_uuid[UI_HISTORY_START] & ~((u64)1);
2426 if (self == peer) {
Philipp Reisner4a23f262011-01-11 17:42:17 +01002427 if (mdev->agreed_pro_version < 96 ?
2428 (mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1)) ==
2429 (mdev->p_uuid[UI_HISTORY_START + 1] & ~((u64)1)) :
2430 peer + UUID_NEW_BM_OFFSET == (mdev->p_uuid[UI_BITMAP] & ~((u64)1))) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002431 /* The last P_SYNC_UUID did not get though. Undo the last start of
2432 resync as sync source modifications of the peer's UUIDs. */
2433
2434 if (mdev->agreed_pro_version < 91)
Philipp Reisner4a23f262011-01-11 17:42:17 +01002435 return -1091;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002436
2437 mdev->p_uuid[UI_BITMAP] = mdev->p_uuid[UI_HISTORY_START];
2438 mdev->p_uuid[UI_HISTORY_START] = mdev->p_uuid[UI_HISTORY_START + 1];
Philipp Reisner4a23f262011-01-11 17:42:17 +01002439
2440 dev_info(DEV, "Did not got last syncUUID packet, corrected:\n");
2441 drbd_uuid_dump(mdev, "peer", mdev->p_uuid, mdev->p_uuid[UI_SIZE], mdev->p_uuid[UI_FLAGS]);
2442
Philipp Reisnerb411b362009-09-25 16:07:19 -07002443 return -1;
2444 }
2445 }
2446
2447 *rule_nr = 60;
2448 self = mdev->ldev->md.uuid[UI_CURRENT] & ~((u64)1);
2449 for (i = UI_HISTORY_START; i <= UI_HISTORY_END; i++) {
2450 peer = mdev->p_uuid[i] & ~((u64)1);
2451 if (self == peer)
2452 return -2;
2453 }
2454
2455 *rule_nr = 70;
2456 self = mdev->ldev->md.uuid[UI_BITMAP] & ~((u64)1);
2457 peer = mdev->p_uuid[UI_CURRENT] & ~((u64)1);
2458 if (self == peer)
2459 return 1;
2460
2461 *rule_nr = 71;
2462 self = mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1);
2463 if (self == peer) {
Philipp Reisner4a23f262011-01-11 17:42:17 +01002464 if (mdev->agreed_pro_version < 96 ?
2465 (mdev->ldev->md.uuid[UI_HISTORY_START + 1] & ~((u64)1)) ==
2466 (mdev->p_uuid[UI_HISTORY_START] & ~((u64)1)) :
2467 self + UUID_NEW_BM_OFFSET == (mdev->ldev->md.uuid[UI_BITMAP] & ~((u64)1))) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002468 /* The last P_SYNC_UUID did not get though. Undo the last start of
2469 resync as sync source modifications of our UUIDs. */
2470
2471 if (mdev->agreed_pro_version < 91)
Philipp Reisner4a23f262011-01-11 17:42:17 +01002472 return -1091;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002473
2474 _drbd_uuid_set(mdev, UI_BITMAP, mdev->ldev->md.uuid[UI_HISTORY_START]);
2475 _drbd_uuid_set(mdev, UI_HISTORY_START, mdev->ldev->md.uuid[UI_HISTORY_START + 1]);
2476
Philipp Reisner4a23f262011-01-11 17:42:17 +01002477 dev_info(DEV, "Last syncUUID did not get through, corrected:\n");
Philipp Reisnerb411b362009-09-25 16:07:19 -07002478 drbd_uuid_dump(mdev, "self", mdev->ldev->md.uuid,
2479 mdev->state.disk >= D_NEGOTIATING ? drbd_bm_total_weight(mdev) : 0, 0);
2480
2481 return 1;
2482 }
2483 }
2484
2485
2486 *rule_nr = 80;
Philipp Reisnerd8c2a362009-11-18 15:52:51 +01002487 peer = mdev->p_uuid[UI_CURRENT] & ~((u64)1);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002488 for (i = UI_HISTORY_START; i <= UI_HISTORY_END; i++) {
2489 self = mdev->ldev->md.uuid[i] & ~((u64)1);
2490 if (self == peer)
2491 return 2;
2492 }
2493
2494 *rule_nr = 90;
2495 self = mdev->ldev->md.uuid[UI_BITMAP] & ~((u64)1);
2496 peer = mdev->p_uuid[UI_BITMAP] & ~((u64)1);
2497 if (self == peer && self != ((u64)0))
2498 return 100;
2499
2500 *rule_nr = 100;
2501 for (i = UI_HISTORY_START; i <= UI_HISTORY_END; i++) {
2502 self = mdev->ldev->md.uuid[i] & ~((u64)1);
2503 for (j = UI_HISTORY_START; j <= UI_HISTORY_END; j++) {
2504 peer = mdev->p_uuid[j] & ~((u64)1);
2505 if (self == peer)
2506 return -100;
2507 }
2508 }
2509
2510 return -1000;
2511}
2512
2513/* drbd_sync_handshake() returns the new conn state on success, or
2514 CONN_MASK (-1) on failure.
2515 */
2516static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_role peer_role,
2517 enum drbd_disk_state peer_disk) __must_hold(local)
2518{
2519 int hg, rule_nr;
2520 enum drbd_conns rv = C_MASK;
2521 enum drbd_disk_state mydisk;
2522
2523 mydisk = mdev->state.disk;
2524 if (mydisk == D_NEGOTIATING)
2525 mydisk = mdev->new_state_tmp.disk;
2526
2527 dev_info(DEV, "drbd_sync_handshake:\n");
2528 drbd_uuid_dump(mdev, "self", mdev->ldev->md.uuid, mdev->comm_bm_set, 0);
2529 drbd_uuid_dump(mdev, "peer", mdev->p_uuid,
2530 mdev->p_uuid[UI_SIZE], mdev->p_uuid[UI_FLAGS]);
2531
2532 hg = drbd_uuid_compare(mdev, &rule_nr);
2533
2534 dev_info(DEV, "uuid_compare()=%d by rule %d\n", hg, rule_nr);
2535
2536 if (hg == -1000) {
2537 dev_alert(DEV, "Unrelated data, aborting!\n");
2538 return C_MASK;
2539 }
Philipp Reisner4a23f262011-01-11 17:42:17 +01002540 if (hg < -1000) {
2541 dev_alert(DEV, "To resolve this both sides have to support at least protocol %d\n", -hg - 1000);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002542 return C_MASK;
2543 }
2544
2545 if ((mydisk == D_INCONSISTENT && peer_disk > D_INCONSISTENT) ||
2546 (peer_disk == D_INCONSISTENT && mydisk > D_INCONSISTENT)) {
2547 int f = (hg == -100) || abs(hg) == 2;
2548 hg = mydisk > D_INCONSISTENT ? 1 : -1;
2549 if (f)
2550 hg = hg*2;
2551 dev_info(DEV, "Becoming sync %s due to disk states.\n",
2552 hg > 0 ? "source" : "target");
2553 }
2554
Adam Gandelman3a11a482010-04-08 16:48:23 -07002555 if (abs(hg) == 100)
2556 drbd_khelper(mdev, "initial-split-brain");
2557
Philipp Reisnerb411b362009-09-25 16:07:19 -07002558 if (hg == 100 || (hg == -100 && mdev->net_conf->always_asbp)) {
2559 int pcount = (mdev->state.role == R_PRIMARY)
2560 + (peer_role == R_PRIMARY);
2561 int forced = (hg == -100);
2562
2563 switch (pcount) {
2564 case 0:
2565 hg = drbd_asb_recover_0p(mdev);
2566 break;
2567 case 1:
2568 hg = drbd_asb_recover_1p(mdev);
2569 break;
2570 case 2:
2571 hg = drbd_asb_recover_2p(mdev);
2572 break;
2573 }
2574 if (abs(hg) < 100) {
2575 dev_warn(DEV, "Split-Brain detected, %d primaries, "
2576 "automatically solved. Sync from %s node\n",
2577 pcount, (hg < 0) ? "peer" : "this");
2578 if (forced) {
2579 dev_warn(DEV, "Doing a full sync, since"
2580 " UUIDs where ambiguous.\n");
2581 hg = hg*2;
2582 }
2583 }
2584 }
2585
2586 if (hg == -100) {
2587 if (mdev->net_conf->want_lose && !(mdev->p_uuid[UI_FLAGS]&1))
2588 hg = -1;
2589 if (!mdev->net_conf->want_lose && (mdev->p_uuid[UI_FLAGS]&1))
2590 hg = 1;
2591
2592 if (abs(hg) < 100)
2593 dev_warn(DEV, "Split-Brain detected, manually solved. "
2594 "Sync from %s node\n",
2595 (hg < 0) ? "peer" : "this");
2596 }
2597
2598 if (hg == -100) {
Lars Ellenberg580b9762010-02-26 23:15:23 +01002599 /* FIXME this log message is not correct if we end up here
2600 * after an attempted attach on a diskless node.
2601 * We just refuse to attach -- well, we drop the "connection"
2602 * to that disk, in a way... */
Adam Gandelman3a11a482010-04-08 16:48:23 -07002603 dev_alert(DEV, "Split-Brain detected but unresolved, dropping connection!\n");
Philipp Reisnerb411b362009-09-25 16:07:19 -07002604 drbd_khelper(mdev, "split-brain");
2605 return C_MASK;
2606 }
2607
2608 if (hg > 0 && mydisk <= D_INCONSISTENT) {
2609 dev_err(DEV, "I shall become SyncSource, but I am inconsistent!\n");
2610 return C_MASK;
2611 }
2612
2613 if (hg < 0 && /* by intention we do not use mydisk here. */
2614 mdev->state.role == R_PRIMARY && mdev->state.disk >= D_CONSISTENT) {
2615 switch (mdev->net_conf->rr_conflict) {
2616 case ASB_CALL_HELPER:
2617 drbd_khelper(mdev, "pri-lost");
2618 /* fall through */
2619 case ASB_DISCONNECT:
2620 dev_err(DEV, "I shall become SyncTarget, but I am primary!\n");
2621 return C_MASK;
2622 case ASB_VIOLENTLY:
2623 dev_warn(DEV, "Becoming SyncTarget, violating the stable-data"
2624 "assumption\n");
2625 }
2626 }
2627
Philipp Reisnercf14c2e2010-02-02 21:03:50 +01002628 if (mdev->net_conf->dry_run || test_bit(CONN_DRY_RUN, &mdev->flags)) {
2629 if (hg == 0)
2630 dev_info(DEV, "dry-run connect: No resync, would become Connected immediately.\n");
2631 else
2632 dev_info(DEV, "dry-run connect: Would become %s, doing a %s resync.",
2633 drbd_conn_str(hg > 0 ? C_SYNC_SOURCE : C_SYNC_TARGET),
2634 abs(hg) >= 2 ? "full" : "bit-map based");
2635 return C_MASK;
2636 }
2637
Philipp Reisnerb411b362009-09-25 16:07:19 -07002638 if (abs(hg) >= 2) {
2639 dev_info(DEV, "Writing the whole bitmap, full sync required after drbd_sync_handshake.\n");
Lars Ellenberg20ceb2b2011-01-21 10:56:44 +01002640 if (drbd_bitmap_io(mdev, &drbd_bmio_set_n_write, "set_n_write from sync_handshake",
2641 BM_LOCKED_SET_ALLOWED))
Philipp Reisnerb411b362009-09-25 16:07:19 -07002642 return C_MASK;
2643 }
2644
2645 if (hg > 0) { /* become sync source. */
2646 rv = C_WF_BITMAP_S;
2647 } else if (hg < 0) { /* become sync target */
2648 rv = C_WF_BITMAP_T;
2649 } else {
2650 rv = C_CONNECTED;
2651 if (drbd_bm_total_weight(mdev)) {
2652 dev_info(DEV, "No resync, but %lu bits in bitmap!\n",
2653 drbd_bm_total_weight(mdev));
2654 }
2655 }
2656
2657 return rv;
2658}
2659
2660/* returns 1 if invalid */
2661static int cmp_after_sb(enum drbd_after_sb_p peer, enum drbd_after_sb_p self)
2662{
2663 /* ASB_DISCARD_REMOTE - ASB_DISCARD_LOCAL is valid */
2664 if ((peer == ASB_DISCARD_REMOTE && self == ASB_DISCARD_LOCAL) ||
2665 (self == ASB_DISCARD_REMOTE && peer == ASB_DISCARD_LOCAL))
2666 return 0;
2667
2668 /* any other things with ASB_DISCARD_REMOTE or ASB_DISCARD_LOCAL are invalid */
2669 if (peer == ASB_DISCARD_REMOTE || peer == ASB_DISCARD_LOCAL ||
2670 self == ASB_DISCARD_REMOTE || self == ASB_DISCARD_LOCAL)
2671 return 1;
2672
2673 /* everything else is valid if they are equal on both sides. */
2674 if (peer == self)
2675 return 0;
2676
2677 /* everything es is invalid. */
2678 return 1;
2679}
2680
Philipp Reisner02918be2010-08-20 14:35:10 +02002681static int receive_protocol(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07002682{
Philipp Reisner02918be2010-08-20 14:35:10 +02002683 struct p_protocol *p = &mdev->data.rbuf.protocol;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002684 int p_proto, p_after_sb_0p, p_after_sb_1p, p_after_sb_2p;
Philipp Reisnercf14c2e2010-02-02 21:03:50 +01002685 int p_want_lose, p_two_primaries, cf;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002686 char p_integrity_alg[SHARED_SECRET_MAX] = "";
2687
Philipp Reisnerb411b362009-09-25 16:07:19 -07002688 p_proto = be32_to_cpu(p->protocol);
2689 p_after_sb_0p = be32_to_cpu(p->after_sb_0p);
2690 p_after_sb_1p = be32_to_cpu(p->after_sb_1p);
2691 p_after_sb_2p = be32_to_cpu(p->after_sb_2p);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002692 p_two_primaries = be32_to_cpu(p->two_primaries);
Philipp Reisnercf14c2e2010-02-02 21:03:50 +01002693 cf = be32_to_cpu(p->conn_flags);
2694 p_want_lose = cf & CF_WANT_LOSE;
2695
2696 clear_bit(CONN_DRY_RUN, &mdev->flags);
2697
2698 if (cf & CF_DRY_RUN)
2699 set_bit(CONN_DRY_RUN, &mdev->flags);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002700
2701 if (p_proto != mdev->net_conf->wire_protocol) {
2702 dev_err(DEV, "incompatible communication protocols\n");
2703 goto disconnect;
2704 }
2705
2706 if (cmp_after_sb(p_after_sb_0p, mdev->net_conf->after_sb_0p)) {
2707 dev_err(DEV, "incompatible after-sb-0pri settings\n");
2708 goto disconnect;
2709 }
2710
2711 if (cmp_after_sb(p_after_sb_1p, mdev->net_conf->after_sb_1p)) {
2712 dev_err(DEV, "incompatible after-sb-1pri settings\n");
2713 goto disconnect;
2714 }
2715
2716 if (cmp_after_sb(p_after_sb_2p, mdev->net_conf->after_sb_2p)) {
2717 dev_err(DEV, "incompatible after-sb-2pri settings\n");
2718 goto disconnect;
2719 }
2720
2721 if (p_want_lose && mdev->net_conf->want_lose) {
2722 dev_err(DEV, "both sides have the 'want_lose' flag set\n");
2723 goto disconnect;
2724 }
2725
2726 if (p_two_primaries != mdev->net_conf->two_primaries) {
2727 dev_err(DEV, "incompatible setting of the two-primaries options\n");
2728 goto disconnect;
2729 }
2730
2731 if (mdev->agreed_pro_version >= 87) {
2732 unsigned char *my_alg = mdev->net_conf->integrity_alg;
2733
2734 if (drbd_recv(mdev, p_integrity_alg, data_size) != data_size)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002735 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002736
2737 p_integrity_alg[SHARED_SECRET_MAX-1] = 0;
2738 if (strcmp(p_integrity_alg, my_alg)) {
2739 dev_err(DEV, "incompatible setting of the data-integrity-alg\n");
2740 goto disconnect;
2741 }
2742 dev_info(DEV, "data-integrity-alg: %s\n",
2743 my_alg[0] ? my_alg : (unsigned char *)"<not-used>");
2744 }
2745
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002746 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002747
2748disconnect:
2749 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002750 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002751}
2752
2753/* helper function
2754 * input: alg name, feature name
2755 * return: NULL (alg name was "")
2756 * ERR_PTR(error) if something goes wrong
2757 * or the crypto hash ptr, if it worked out ok. */
2758struct crypto_hash *drbd_crypto_alloc_digest_safe(const struct drbd_conf *mdev,
2759 const char *alg, const char *name)
2760{
2761 struct crypto_hash *tfm;
2762
2763 if (!alg[0])
2764 return NULL;
2765
2766 tfm = crypto_alloc_hash(alg, 0, CRYPTO_ALG_ASYNC);
2767 if (IS_ERR(tfm)) {
2768 dev_err(DEV, "Can not allocate \"%s\" as %s (reason: %ld)\n",
2769 alg, name, PTR_ERR(tfm));
2770 return tfm;
2771 }
2772 if (!drbd_crypto_is_hash(crypto_hash_tfm(tfm))) {
2773 crypto_free_hash(tfm);
2774 dev_err(DEV, "\"%s\" is not a digest (%s)\n", alg, name);
2775 return ERR_PTR(-EINVAL);
2776 }
2777 return tfm;
2778}
2779
Philipp Reisner02918be2010-08-20 14:35:10 +02002780static int receive_SyncParam(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int packet_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07002781{
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002782 int ok = true;
Philipp Reisner02918be2010-08-20 14:35:10 +02002783 struct p_rs_param_95 *p = &mdev->data.rbuf.rs_param_95;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002784 unsigned int header_size, data_size, exp_max_sz;
2785 struct crypto_hash *verify_tfm = NULL;
2786 struct crypto_hash *csums_tfm = NULL;
2787 const int apv = mdev->agreed_pro_version;
Philipp Reisner778f2712010-07-06 11:14:00 +02002788 int *rs_plan_s = NULL;
2789 int fifo_size = 0;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002790
2791 exp_max_sz = apv <= 87 ? sizeof(struct p_rs_param)
2792 : apv == 88 ? sizeof(struct p_rs_param)
2793 + SHARED_SECRET_MAX
Philipp Reisner8e26f9c2010-07-06 17:25:54 +02002794 : apv <= 94 ? sizeof(struct p_rs_param_89)
2795 : /* apv >= 95 */ sizeof(struct p_rs_param_95);
Philipp Reisnerb411b362009-09-25 16:07:19 -07002796
Philipp Reisner02918be2010-08-20 14:35:10 +02002797 if (packet_size > exp_max_sz) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07002798 dev_err(DEV, "SyncParam packet too long: received %u, expected <= %u bytes\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02002799 packet_size, exp_max_sz);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002800 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002801 }
2802
2803 if (apv <= 88) {
Philipp Reisner02918be2010-08-20 14:35:10 +02002804 header_size = sizeof(struct p_rs_param) - sizeof(struct p_header80);
2805 data_size = packet_size - header_size;
Philipp Reisner8e26f9c2010-07-06 17:25:54 +02002806 } else if (apv <= 94) {
Philipp Reisner02918be2010-08-20 14:35:10 +02002807 header_size = sizeof(struct p_rs_param_89) - sizeof(struct p_header80);
2808 data_size = packet_size - header_size;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002809 D_ASSERT(data_size == 0);
Philipp Reisner8e26f9c2010-07-06 17:25:54 +02002810 } else {
Philipp Reisner02918be2010-08-20 14:35:10 +02002811 header_size = sizeof(struct p_rs_param_95) - sizeof(struct p_header80);
2812 data_size = packet_size - header_size;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002813 D_ASSERT(data_size == 0);
2814 }
2815
2816 /* initialize verify_alg and csums_alg */
2817 memset(p->verify_alg, 0, 2 * SHARED_SECRET_MAX);
2818
Philipp Reisner02918be2010-08-20 14:35:10 +02002819 if (drbd_recv(mdev, &p->head.payload, header_size) != header_size)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002820 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002821
2822 mdev->sync_conf.rate = be32_to_cpu(p->rate);
2823
2824 if (apv >= 88) {
2825 if (apv == 88) {
2826 if (data_size > SHARED_SECRET_MAX) {
2827 dev_err(DEV, "verify-alg too long, "
2828 "peer wants %u, accepting only %u byte\n",
2829 data_size, SHARED_SECRET_MAX);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002830 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002831 }
2832
2833 if (drbd_recv(mdev, p->verify_alg, data_size) != data_size)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002834 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002835
2836 /* we expect NUL terminated string */
2837 /* but just in case someone tries to be evil */
2838 D_ASSERT(p->verify_alg[data_size-1] == 0);
2839 p->verify_alg[data_size-1] = 0;
2840
2841 } else /* apv >= 89 */ {
2842 /* we still expect NUL terminated strings */
2843 /* but just in case someone tries to be evil */
2844 D_ASSERT(p->verify_alg[SHARED_SECRET_MAX-1] == 0);
2845 D_ASSERT(p->csums_alg[SHARED_SECRET_MAX-1] == 0);
2846 p->verify_alg[SHARED_SECRET_MAX-1] = 0;
2847 p->csums_alg[SHARED_SECRET_MAX-1] = 0;
2848 }
2849
2850 if (strcmp(mdev->sync_conf.verify_alg, p->verify_alg)) {
2851 if (mdev->state.conn == C_WF_REPORT_PARAMS) {
2852 dev_err(DEV, "Different verify-alg settings. me=\"%s\" peer=\"%s\"\n",
2853 mdev->sync_conf.verify_alg, p->verify_alg);
2854 goto disconnect;
2855 }
2856 verify_tfm = drbd_crypto_alloc_digest_safe(mdev,
2857 p->verify_alg, "verify-alg");
2858 if (IS_ERR(verify_tfm)) {
2859 verify_tfm = NULL;
2860 goto disconnect;
2861 }
2862 }
2863
2864 if (apv >= 89 && strcmp(mdev->sync_conf.csums_alg, p->csums_alg)) {
2865 if (mdev->state.conn == C_WF_REPORT_PARAMS) {
2866 dev_err(DEV, "Different csums-alg settings. me=\"%s\" peer=\"%s\"\n",
2867 mdev->sync_conf.csums_alg, p->csums_alg);
2868 goto disconnect;
2869 }
2870 csums_tfm = drbd_crypto_alloc_digest_safe(mdev,
2871 p->csums_alg, "csums-alg");
2872 if (IS_ERR(csums_tfm)) {
2873 csums_tfm = NULL;
2874 goto disconnect;
2875 }
2876 }
2877
Philipp Reisner8e26f9c2010-07-06 17:25:54 +02002878 if (apv > 94) {
2879 mdev->sync_conf.rate = be32_to_cpu(p->rate);
2880 mdev->sync_conf.c_plan_ahead = be32_to_cpu(p->c_plan_ahead);
2881 mdev->sync_conf.c_delay_target = be32_to_cpu(p->c_delay_target);
2882 mdev->sync_conf.c_fill_target = be32_to_cpu(p->c_fill_target);
2883 mdev->sync_conf.c_max_rate = be32_to_cpu(p->c_max_rate);
Philipp Reisner778f2712010-07-06 11:14:00 +02002884
2885 fifo_size = (mdev->sync_conf.c_plan_ahead * 10 * SLEEP_TIME) / HZ;
2886 if (fifo_size != mdev->rs_plan_s.size && fifo_size > 0) {
2887 rs_plan_s = kzalloc(sizeof(int) * fifo_size, GFP_KERNEL);
2888 if (!rs_plan_s) {
2889 dev_err(DEV, "kmalloc of fifo_buffer failed");
2890 goto disconnect;
2891 }
2892 }
Philipp Reisner8e26f9c2010-07-06 17:25:54 +02002893 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07002894
2895 spin_lock(&mdev->peer_seq_lock);
2896 /* lock against drbd_nl_syncer_conf() */
2897 if (verify_tfm) {
2898 strcpy(mdev->sync_conf.verify_alg, p->verify_alg);
2899 mdev->sync_conf.verify_alg_len = strlen(p->verify_alg) + 1;
2900 crypto_free_hash(mdev->verify_tfm);
2901 mdev->verify_tfm = verify_tfm;
2902 dev_info(DEV, "using verify-alg: \"%s\"\n", p->verify_alg);
2903 }
2904 if (csums_tfm) {
2905 strcpy(mdev->sync_conf.csums_alg, p->csums_alg);
2906 mdev->sync_conf.csums_alg_len = strlen(p->csums_alg) + 1;
2907 crypto_free_hash(mdev->csums_tfm);
2908 mdev->csums_tfm = csums_tfm;
2909 dev_info(DEV, "using csums-alg: \"%s\"\n", p->csums_alg);
2910 }
Philipp Reisner778f2712010-07-06 11:14:00 +02002911 if (fifo_size != mdev->rs_plan_s.size) {
2912 kfree(mdev->rs_plan_s.values);
2913 mdev->rs_plan_s.values = rs_plan_s;
2914 mdev->rs_plan_s.size = fifo_size;
2915 mdev->rs_planed = 0;
2916 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07002917 spin_unlock(&mdev->peer_seq_lock);
2918 }
2919
2920 return ok;
2921disconnect:
2922 /* just for completeness: actually not needed,
2923 * as this is not reached if csums_tfm was ok. */
2924 crypto_free_hash(csums_tfm);
2925 /* but free the verify_tfm again, if csums_tfm did not work out */
2926 crypto_free_hash(verify_tfm);
2927 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002928 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002929}
2930
Philipp Reisnerb411b362009-09-25 16:07:19 -07002931/* warn if the arguments differ by more than 12.5% */
2932static void warn_if_differ_considerably(struct drbd_conf *mdev,
2933 const char *s, sector_t a, sector_t b)
2934{
2935 sector_t d;
2936 if (a == 0 || b == 0)
2937 return;
2938 d = (a > b) ? (a - b) : (b - a);
2939 if (d > (a>>3) || d > (b>>3))
2940 dev_warn(DEV, "Considerable difference in %s: %llus vs. %llus\n", s,
2941 (unsigned long long)a, (unsigned long long)b);
2942}
2943
Philipp Reisner02918be2010-08-20 14:35:10 +02002944static int receive_sizes(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07002945{
Philipp Reisner02918be2010-08-20 14:35:10 +02002946 struct p_sizes *p = &mdev->data.rbuf.sizes;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002947 enum determine_dev_size dd = unchanged;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002948 sector_t p_size, p_usize, my_usize;
2949 int ldsc = 0; /* local disk size changed */
Philipp Reisnere89b5912010-03-24 17:11:33 +01002950 enum dds_flags ddsf;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002951
Philipp Reisnerb411b362009-09-25 16:07:19 -07002952 p_size = be64_to_cpu(p->d_size);
2953 p_usize = be64_to_cpu(p->u_size);
2954
2955 if (p_size == 0 && mdev->state.disk == D_DISKLESS) {
2956 dev_err(DEV, "some backing storage is needed\n");
2957 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002958 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002959 }
2960
2961 /* just store the peer's disk size for now.
2962 * we still need to figure out whether we accept that. */
2963 mdev->p_size = p_size;
2964
Philipp Reisnerb411b362009-09-25 16:07:19 -07002965 if (get_ldev(mdev)) {
2966 warn_if_differ_considerably(mdev, "lower level device sizes",
2967 p_size, drbd_get_max_capacity(mdev->ldev));
2968 warn_if_differ_considerably(mdev, "user requested size",
2969 p_usize, mdev->ldev->dc.disk_size);
2970
2971 /* if this is the first connect, or an otherwise expected
2972 * param exchange, choose the minimum */
2973 if (mdev->state.conn == C_WF_REPORT_PARAMS)
2974 p_usize = min_not_zero((sector_t)mdev->ldev->dc.disk_size,
2975 p_usize);
2976
2977 my_usize = mdev->ldev->dc.disk_size;
2978
2979 if (mdev->ldev->dc.disk_size != p_usize) {
2980 mdev->ldev->dc.disk_size = p_usize;
2981 dev_info(DEV, "Peer sets u_size to %lu sectors\n",
2982 (unsigned long)mdev->ldev->dc.disk_size);
2983 }
2984
2985 /* Never shrink a device with usable data during connect.
2986 But allow online shrinking if we are connected. */
Philipp Reisnera393db62009-12-22 13:35:52 +01002987 if (drbd_new_dev_size(mdev, mdev->ldev, 0) <
Philipp Reisnerb411b362009-09-25 16:07:19 -07002988 drbd_get_capacity(mdev->this_bdev) &&
2989 mdev->state.disk >= D_OUTDATED &&
2990 mdev->state.conn < C_CONNECTED) {
2991 dev_err(DEV, "The peer's disk size is too small!\n");
2992 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
2993 mdev->ldev->dc.disk_size = my_usize;
2994 put_ldev(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01002995 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07002996 }
2997 put_ldev(mdev);
2998 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07002999
Philipp Reisnere89b5912010-03-24 17:11:33 +01003000 ddsf = be16_to_cpu(p->dds_flags);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003001 if (get_ldev(mdev)) {
Bart Van Assche24c48302011-05-21 18:32:29 +02003002 dd = drbd_determine_dev_size(mdev, ddsf);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003003 put_ldev(mdev);
3004 if (dd == dev_size_error)
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003005 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003006 drbd_md_sync(mdev);
3007 } else {
3008 /* I am diskless, need to accept the peer's size. */
3009 drbd_set_my_capacity(mdev, p_size);
3010 }
3011
Philipp Reisner99432fc2011-05-20 16:39:13 +02003012 mdev->peer_max_bio_size = be32_to_cpu(p->max_bio_size);
3013 drbd_reconsider_max_bio_size(mdev);
3014
Philipp Reisnerb411b362009-09-25 16:07:19 -07003015 if (get_ldev(mdev)) {
3016 if (mdev->ldev->known_size != drbd_get_capacity(mdev->ldev->backing_bdev)) {
3017 mdev->ldev->known_size = drbd_get_capacity(mdev->ldev->backing_bdev);
3018 ldsc = 1;
3019 }
3020
Philipp Reisnerb411b362009-09-25 16:07:19 -07003021 put_ldev(mdev);
3022 }
3023
3024 if (mdev->state.conn > C_WF_REPORT_PARAMS) {
3025 if (be64_to_cpu(p->c_size) !=
3026 drbd_get_capacity(mdev->this_bdev) || ldsc) {
3027 /* we have different sizes, probably peer
3028 * needs to know my new size... */
Philipp Reisnere89b5912010-03-24 17:11:33 +01003029 drbd_send_sizes(mdev, 0, ddsf);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003030 }
3031 if (test_and_clear_bit(RESIZE_PENDING, &mdev->flags) ||
3032 (dd == grew && mdev->state.conn == C_CONNECTED)) {
3033 if (mdev->state.pdsk >= D_INCONSISTENT &&
Philipp Reisnere89b5912010-03-24 17:11:33 +01003034 mdev->state.disk >= D_INCONSISTENT) {
3035 if (ddsf & DDSF_NO_RESYNC)
3036 dev_info(DEV, "Resync of new storage suppressed with --assume-clean\n");
3037 else
3038 resync_after_online_grow(mdev);
3039 } else
Philipp Reisnerb411b362009-09-25 16:07:19 -07003040 set_bit(RESYNC_AFTER_NEG, &mdev->flags);
3041 }
3042 }
3043
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003044 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003045}
3046
Philipp Reisner02918be2010-08-20 14:35:10 +02003047static int receive_uuids(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003048{
Philipp Reisner02918be2010-08-20 14:35:10 +02003049 struct p_uuids *p = &mdev->data.rbuf.uuids;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003050 u64 *p_uuid;
Lars Ellenberg62b0da32011-01-20 13:25:21 +01003051 int i, updated_uuids = 0;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003052
Philipp Reisnerb411b362009-09-25 16:07:19 -07003053 p_uuid = kmalloc(sizeof(u64)*UI_EXTENDED_SIZE, GFP_NOIO);
3054
3055 for (i = UI_CURRENT; i < UI_EXTENDED_SIZE; i++)
3056 p_uuid[i] = be64_to_cpu(p->uuid[i]);
3057
3058 kfree(mdev->p_uuid);
3059 mdev->p_uuid = p_uuid;
3060
3061 if (mdev->state.conn < C_CONNECTED &&
3062 mdev->state.disk < D_INCONSISTENT &&
3063 mdev->state.role == R_PRIMARY &&
3064 (mdev->ed_uuid & ~((u64)1)) != (p_uuid[UI_CURRENT] & ~((u64)1))) {
3065 dev_err(DEV, "Can only connect to data with current UUID=%016llX\n",
3066 (unsigned long long)mdev->ed_uuid);
3067 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003068 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003069 }
3070
3071 if (get_ldev(mdev)) {
3072 int skip_initial_sync =
3073 mdev->state.conn == C_CONNECTED &&
3074 mdev->agreed_pro_version >= 90 &&
3075 mdev->ldev->md.uuid[UI_CURRENT] == UUID_JUST_CREATED &&
3076 (p_uuid[UI_FLAGS] & 8);
3077 if (skip_initial_sync) {
3078 dev_info(DEV, "Accepted new current UUID, preparing to skip initial sync\n");
3079 drbd_bitmap_io(mdev, &drbd_bmio_clear_n_write,
Lars Ellenberg20ceb2b2011-01-21 10:56:44 +01003080 "clear_n_write from receive_uuids",
3081 BM_LOCKED_TEST_ALLOWED);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003082 _drbd_uuid_set(mdev, UI_CURRENT, p_uuid[UI_CURRENT]);
3083 _drbd_uuid_set(mdev, UI_BITMAP, 0);
3084 _drbd_set_state(_NS2(mdev, disk, D_UP_TO_DATE, pdsk, D_UP_TO_DATE),
3085 CS_VERBOSE, NULL);
3086 drbd_md_sync(mdev);
Lars Ellenberg62b0da32011-01-20 13:25:21 +01003087 updated_uuids = 1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003088 }
3089 put_ldev(mdev);
Philipp Reisner18a50fa2010-06-21 14:14:15 +02003090 } else if (mdev->state.disk < D_INCONSISTENT &&
3091 mdev->state.role == R_PRIMARY) {
3092 /* I am a diskless primary, the peer just created a new current UUID
3093 for me. */
Lars Ellenberg62b0da32011-01-20 13:25:21 +01003094 updated_uuids = drbd_set_ed_uuid(mdev, p_uuid[UI_CURRENT]);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003095 }
3096
3097 /* Before we test for the disk state, we should wait until an eventually
3098 ongoing cluster wide state change is finished. That is important if
3099 we are primary and are detaching from our disk. We need to see the
3100 new disk state... */
3101 wait_event(mdev->misc_wait, !test_bit(CLUSTER_ST_CHANGE, &mdev->flags));
3102 if (mdev->state.conn >= C_CONNECTED && mdev->state.disk < D_INCONSISTENT)
Lars Ellenberg62b0da32011-01-20 13:25:21 +01003103 updated_uuids |= drbd_set_ed_uuid(mdev, p_uuid[UI_CURRENT]);
3104
3105 if (updated_uuids)
3106 drbd_print_uuids(mdev, "receiver updated UUIDs to");
Philipp Reisnerb411b362009-09-25 16:07:19 -07003107
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003108 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003109}
3110
3111/**
3112 * convert_state() - Converts the peer's view of the cluster state to our point of view
3113 * @ps: The state as seen by the peer.
3114 */
3115static union drbd_state convert_state(union drbd_state ps)
3116{
3117 union drbd_state ms;
3118
3119 static enum drbd_conns c_tab[] = {
3120 [C_CONNECTED] = C_CONNECTED,
3121
3122 [C_STARTING_SYNC_S] = C_STARTING_SYNC_T,
3123 [C_STARTING_SYNC_T] = C_STARTING_SYNC_S,
3124 [C_DISCONNECTING] = C_TEAR_DOWN, /* C_NETWORK_FAILURE, */
3125 [C_VERIFY_S] = C_VERIFY_T,
3126 [C_MASK] = C_MASK,
3127 };
3128
3129 ms.i = ps.i;
3130
3131 ms.conn = c_tab[ps.conn];
3132 ms.peer = ps.role;
3133 ms.role = ps.peer;
3134 ms.pdsk = ps.disk;
3135 ms.disk = ps.pdsk;
3136 ms.peer_isp = (ps.aftr_isp | ps.user_isp);
3137
3138 return ms;
3139}
3140
Philipp Reisner02918be2010-08-20 14:35:10 +02003141static int receive_req_state(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003142{
Philipp Reisner02918be2010-08-20 14:35:10 +02003143 struct p_req_state *p = &mdev->data.rbuf.req_state;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003144 union drbd_state mask, val;
Andreas Gruenbacherbf885f82010-12-08 00:39:32 +01003145 enum drbd_state_rv rv;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003146
Philipp Reisnerb411b362009-09-25 16:07:19 -07003147 mask.i = be32_to_cpu(p->mask);
3148 val.i = be32_to_cpu(p->val);
3149
3150 if (test_bit(DISCARD_CONCURRENT, &mdev->flags) &&
3151 test_bit(CLUSTER_ST_CHANGE, &mdev->flags)) {
3152 drbd_send_sr_reply(mdev, SS_CONCURRENT_ST_CHG);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003153 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003154 }
3155
3156 mask = convert_state(mask);
3157 val = convert_state(val);
3158
3159 rv = drbd_change_state(mdev, CS_VERBOSE, mask, val);
3160
3161 drbd_send_sr_reply(mdev, rv);
3162 drbd_md_sync(mdev);
3163
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003164 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003165}
3166
Philipp Reisner02918be2010-08-20 14:35:10 +02003167static int receive_state(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003168{
Philipp Reisner02918be2010-08-20 14:35:10 +02003169 struct p_state *p = &mdev->data.rbuf.state;
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003170 union drbd_state os, ns, peer_state;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003171 enum drbd_disk_state real_peer_disk;
Philipp Reisner65d922c2010-06-16 16:18:09 +02003172 enum chg_state_flags cs_flags;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003173 int rv;
3174
Philipp Reisnerb411b362009-09-25 16:07:19 -07003175 peer_state.i = be32_to_cpu(p->state);
3176
3177 real_peer_disk = peer_state.disk;
3178 if (peer_state.disk == D_NEGOTIATING) {
3179 real_peer_disk = mdev->p_uuid[UI_FLAGS] & 4 ? D_INCONSISTENT : D_CONSISTENT;
3180 dev_info(DEV, "real peer disk state = %s\n", drbd_disk_str(real_peer_disk));
3181 }
3182
3183 spin_lock_irq(&mdev->req_lock);
3184 retry:
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003185 os = ns = mdev->state;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003186 spin_unlock_irq(&mdev->req_lock);
3187
Lars Ellenberge9ef7bb2010-10-07 15:55:39 +02003188 /* peer says his disk is uptodate, while we think it is inconsistent,
3189 * and this happens while we think we have a sync going on. */
3190 if (os.pdsk == D_INCONSISTENT && real_peer_disk == D_UP_TO_DATE &&
3191 os.conn > C_CONNECTED && os.disk == D_UP_TO_DATE) {
3192 /* If we are (becoming) SyncSource, but peer is still in sync
3193 * preparation, ignore its uptodate-ness to avoid flapping, it
3194 * will change to inconsistent once the peer reaches active
3195 * syncing states.
3196 * It may have changed syncer-paused flags, however, so we
3197 * cannot ignore this completely. */
3198 if (peer_state.conn > C_CONNECTED &&
3199 peer_state.conn < C_SYNC_SOURCE)
3200 real_peer_disk = D_INCONSISTENT;
3201
3202 /* if peer_state changes to connected at the same time,
3203 * it explicitly notifies us that it finished resync.
3204 * Maybe we should finish it up, too? */
3205 else if (os.conn >= C_SYNC_SOURCE &&
3206 peer_state.conn == C_CONNECTED) {
3207 if (drbd_bm_total_weight(mdev) <= mdev->rs_failed)
3208 drbd_resync_finished(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003209 return true;
Lars Ellenberge9ef7bb2010-10-07 15:55:39 +02003210 }
3211 }
3212
3213 /* peer says his disk is inconsistent, while we think it is uptodate,
3214 * and this happens while the peer still thinks we have a sync going on,
3215 * but we think we are already done with the sync.
3216 * We ignore this to avoid flapping pdsk.
3217 * This should not happen, if the peer is a recent version of drbd. */
3218 if (os.pdsk == D_UP_TO_DATE && real_peer_disk == D_INCONSISTENT &&
3219 os.conn == C_CONNECTED && peer_state.conn > C_SYNC_SOURCE)
3220 real_peer_disk = D_UP_TO_DATE;
3221
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003222 if (ns.conn == C_WF_REPORT_PARAMS)
3223 ns.conn = C_CONNECTED;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003224
Philipp Reisner67531712010-10-27 12:21:30 +02003225 if (peer_state.conn == C_AHEAD)
3226 ns.conn = C_BEHIND;
3227
Philipp Reisnerb411b362009-09-25 16:07:19 -07003228 if (mdev->p_uuid && peer_state.disk >= D_NEGOTIATING &&
3229 get_ldev_if_state(mdev, D_NEGOTIATING)) {
3230 int cr; /* consider resync */
3231
3232 /* if we established a new connection */
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003233 cr = (os.conn < C_CONNECTED);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003234 /* if we had an established connection
3235 * and one of the nodes newly attaches a disk */
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003236 cr |= (os.conn == C_CONNECTED &&
Philipp Reisnerb411b362009-09-25 16:07:19 -07003237 (peer_state.disk == D_NEGOTIATING ||
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003238 os.disk == D_NEGOTIATING));
Philipp Reisnerb411b362009-09-25 16:07:19 -07003239 /* if we have both been inconsistent, and the peer has been
3240 * forced to be UpToDate with --overwrite-data */
3241 cr |= test_bit(CONSIDER_RESYNC, &mdev->flags);
3242 /* if we had been plain connected, and the admin requested to
3243 * start a sync by "invalidate" or "invalidate-remote" */
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003244 cr |= (os.conn == C_CONNECTED &&
Philipp Reisnerb411b362009-09-25 16:07:19 -07003245 (peer_state.conn >= C_STARTING_SYNC_S &&
3246 peer_state.conn <= C_WF_BITMAP_T));
3247
3248 if (cr)
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003249 ns.conn = drbd_sync_handshake(mdev, peer_state.role, real_peer_disk);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003250
3251 put_ldev(mdev);
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003252 if (ns.conn == C_MASK) {
3253 ns.conn = C_CONNECTED;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003254 if (mdev->state.disk == D_NEGOTIATING) {
Lars Ellenberg82f59cc2010-10-16 12:13:47 +02003255 drbd_force_state(mdev, NS(disk, D_FAILED));
Philipp Reisnerb411b362009-09-25 16:07:19 -07003256 } else if (peer_state.disk == D_NEGOTIATING) {
3257 dev_err(DEV, "Disk attach process on the peer node was aborted.\n");
3258 peer_state.disk = D_DISKLESS;
Lars Ellenberg580b9762010-02-26 23:15:23 +01003259 real_peer_disk = D_DISKLESS;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003260 } else {
Philipp Reisnercf14c2e2010-02-02 21:03:50 +01003261 if (test_and_clear_bit(CONN_DRY_RUN, &mdev->flags))
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003262 return false;
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003263 D_ASSERT(os.conn == C_WF_REPORT_PARAMS);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003264 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003265 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003266 }
3267 }
3268 }
3269
3270 spin_lock_irq(&mdev->req_lock);
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003271 if (mdev->state.i != os.i)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003272 goto retry;
3273 clear_bit(CONSIDER_RESYNC, &mdev->flags);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003274 ns.peer = peer_state.role;
3275 ns.pdsk = real_peer_disk;
3276 ns.peer_isp = (peer_state.aftr_isp | peer_state.user_isp);
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003277 if ((ns.conn == C_CONNECTED || ns.conn == C_WF_BITMAP_S) && ns.disk == D_NEGOTIATING)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003278 ns.disk = mdev->new_state_tmp.disk;
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003279 cs_flags = CS_VERBOSE + (os.conn < C_CONNECTED && ns.conn >= C_CONNECTED ? 0 : CS_HARD);
3280 if (ns.pdsk == D_CONSISTENT && is_susp(ns) && ns.conn == C_CONNECTED && os.conn < C_CONNECTED &&
Philipp Reisner481c6f52010-06-22 14:03:27 +02003281 test_bit(NEW_CUR_UUID, &mdev->flags)) {
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01003282 /* Do not allow tl_restart(RESEND) for a rebooted peer. We can only allow this
Philipp Reisner481c6f52010-06-22 14:03:27 +02003283 for temporal network outages! */
3284 spin_unlock_irq(&mdev->req_lock);
3285 dev_err(DEV, "Aborting Connect, can not thaw IO with an only Consistent peer\n");
3286 tl_clear(mdev);
3287 drbd_uuid_new_current(mdev);
3288 clear_bit(NEW_CUR_UUID, &mdev->flags);
3289 drbd_force_state(mdev, NS2(conn, C_PROTOCOL_ERROR, susp, 0));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003290 return false;
Philipp Reisner481c6f52010-06-22 14:03:27 +02003291 }
Philipp Reisner65d922c2010-06-16 16:18:09 +02003292 rv = _drbd_set_state(mdev, ns, cs_flags, NULL);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003293 ns = mdev->state;
3294 spin_unlock_irq(&mdev->req_lock);
3295
3296 if (rv < SS_SUCCESS) {
3297 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003298 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003299 }
3300
Lars Ellenberg4ac4aad2010-07-22 17:39:26 +02003301 if (os.conn > C_WF_REPORT_PARAMS) {
3302 if (ns.conn > C_CONNECTED && peer_state.conn <= C_CONNECTED &&
Philipp Reisnerb411b362009-09-25 16:07:19 -07003303 peer_state.disk != D_NEGOTIATING ) {
3304 /* we want resync, peer has not yet decided to sync... */
3305 /* Nowadays only used when forcing a node into primary role and
3306 setting its disk to UpToDate with that */
3307 drbd_send_uuids(mdev);
3308 drbd_send_state(mdev);
3309 }
3310 }
3311
3312 mdev->net_conf->want_lose = 0;
3313
3314 drbd_md_sync(mdev); /* update connected indicator, la_size, ... */
3315
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003316 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003317}
3318
Philipp Reisner02918be2010-08-20 14:35:10 +02003319static int receive_sync_uuid(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003320{
Philipp Reisner02918be2010-08-20 14:35:10 +02003321 struct p_rs_uuid *p = &mdev->data.rbuf.rs_uuid;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003322
3323 wait_event(mdev->misc_wait,
3324 mdev->state.conn == C_WF_SYNC_UUID ||
Philipp Reisnerc4752ef2010-10-27 17:32:36 +02003325 mdev->state.conn == C_BEHIND ||
Philipp Reisnerb411b362009-09-25 16:07:19 -07003326 mdev->state.conn < C_CONNECTED ||
3327 mdev->state.disk < D_NEGOTIATING);
3328
3329 /* D_ASSERT( mdev->state.conn == C_WF_SYNC_UUID ); */
3330
Philipp Reisnerb411b362009-09-25 16:07:19 -07003331 /* Here the _drbd_uuid_ functions are right, current should
3332 _not_ be rotated into the history */
3333 if (get_ldev_if_state(mdev, D_NEGOTIATING)) {
3334 _drbd_uuid_set(mdev, UI_CURRENT, be64_to_cpu(p->uuid));
3335 _drbd_uuid_set(mdev, UI_BITMAP, 0UL);
3336
Lars Ellenberg62b0da32011-01-20 13:25:21 +01003337 drbd_print_uuids(mdev, "updated sync uuid");
Philipp Reisnerb411b362009-09-25 16:07:19 -07003338 drbd_start_resync(mdev, C_SYNC_TARGET);
3339
3340 put_ldev(mdev);
3341 } else
3342 dev_err(DEV, "Ignoring SyncUUID packet!\n");
3343
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003344 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003345}
3346
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003347/**
3348 * receive_bitmap_plain
3349 *
3350 * Return 0 when done, 1 when another iteration is needed, and a negative error
3351 * code upon failure.
3352 */
3353static int
Philipp Reisner02918be2010-08-20 14:35:10 +02003354receive_bitmap_plain(struct drbd_conf *mdev, unsigned int data_size,
3355 unsigned long *buffer, struct bm_xfer_ctx *c)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003356{
3357 unsigned num_words = min_t(size_t, BM_PACKET_WORDS, c->bm_words - c->word_offset);
3358 unsigned want = num_words * sizeof(long);
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003359 int err;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003360
Philipp Reisner02918be2010-08-20 14:35:10 +02003361 if (want != data_size) {
3362 dev_err(DEV, "%s:want (%u) != data_size (%u)\n", __func__, want, data_size);
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003363 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003364 }
3365 if (want == 0)
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003366 return 0;
3367 err = drbd_recv(mdev, buffer, want);
3368 if (err != want) {
3369 if (err >= 0)
3370 err = -EIO;
3371 return err;
3372 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07003373
3374 drbd_bm_merge_lel(mdev, c->word_offset, num_words, buffer);
3375
3376 c->word_offset += num_words;
3377 c->bit_offset = c->word_offset * BITS_PER_LONG;
3378 if (c->bit_offset > c->bm_bits)
3379 c->bit_offset = c->bm_bits;
3380
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003381 return 1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003382}
3383
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003384/**
3385 * recv_bm_rle_bits
3386 *
3387 * Return 0 when done, 1 when another iteration is needed, and a negative error
3388 * code upon failure.
3389 */
3390static int
Philipp Reisnerb411b362009-09-25 16:07:19 -07003391recv_bm_rle_bits(struct drbd_conf *mdev,
3392 struct p_compressed_bm *p,
3393 struct bm_xfer_ctx *c)
3394{
3395 struct bitstream bs;
3396 u64 look_ahead;
3397 u64 rl;
3398 u64 tmp;
3399 unsigned long s = c->bit_offset;
3400 unsigned long e;
Lars Ellenberg004352f2010-10-05 20:13:58 +02003401 int len = be16_to_cpu(p->head.length) - (sizeof(*p) - sizeof(p->head));
Philipp Reisnerb411b362009-09-25 16:07:19 -07003402 int toggle = DCBP_get_start(p);
3403 int have;
3404 int bits;
3405
3406 bitstream_init(&bs, p->code, len, DCBP_get_pad_bits(p));
3407
3408 bits = bitstream_get_bits(&bs, &look_ahead, 64);
3409 if (bits < 0)
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003410 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003411
3412 for (have = bits; have > 0; s += rl, toggle = !toggle) {
3413 bits = vli_decode_bits(&rl, look_ahead);
3414 if (bits <= 0)
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003415 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003416
3417 if (toggle) {
3418 e = s + rl -1;
3419 if (e >= c->bm_bits) {
3420 dev_err(DEV, "bitmap overflow (e:%lu) while decoding bm RLE packet\n", e);
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003421 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003422 }
3423 _drbd_bm_set_bits(mdev, s, e);
3424 }
3425
3426 if (have < bits) {
3427 dev_err(DEV, "bitmap decoding error: h:%d b:%d la:0x%08llx l:%u/%u\n",
3428 have, bits, look_ahead,
3429 (unsigned int)(bs.cur.b - p->code),
3430 (unsigned int)bs.buf_len);
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003431 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003432 }
3433 look_ahead >>= bits;
3434 have -= bits;
3435
3436 bits = bitstream_get_bits(&bs, &tmp, 64 - have);
3437 if (bits < 0)
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003438 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003439 look_ahead |= tmp << have;
3440 have += bits;
3441 }
3442
3443 c->bit_offset = s;
3444 bm_xfer_ctx_bit_to_word_offset(c);
3445
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003446 return (s != c->bm_bits);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003447}
3448
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003449/**
3450 * decode_bitmap_c
3451 *
3452 * Return 0 when done, 1 when another iteration is needed, and a negative error
3453 * code upon failure.
3454 */
3455static int
Philipp Reisnerb411b362009-09-25 16:07:19 -07003456decode_bitmap_c(struct drbd_conf *mdev,
3457 struct p_compressed_bm *p,
3458 struct bm_xfer_ctx *c)
3459{
3460 if (DCBP_get_code(p) == RLE_VLI_Bits)
3461 return recv_bm_rle_bits(mdev, p, c);
3462
3463 /* other variants had been implemented for evaluation,
3464 * but have been dropped as this one turned out to be "best"
3465 * during all our tests. */
3466
3467 dev_err(DEV, "receive_bitmap_c: unknown encoding %u\n", p->encoding);
3468 drbd_force_state(mdev, NS(conn, C_PROTOCOL_ERROR));
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003469 return -EIO;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003470}
3471
3472void INFO_bm_xfer_stats(struct drbd_conf *mdev,
3473 const char *direction, struct bm_xfer_ctx *c)
3474{
3475 /* what would it take to transfer it "plaintext" */
Philipp Reisner0b70a132010-08-20 13:36:10 +02003476 unsigned plain = sizeof(struct p_header80) *
Philipp Reisnerb411b362009-09-25 16:07:19 -07003477 ((c->bm_words+BM_PACKET_WORDS-1)/BM_PACKET_WORDS+1)
3478 + c->bm_words * sizeof(long);
3479 unsigned total = c->bytes[0] + c->bytes[1];
3480 unsigned r;
3481
3482 /* total can not be zero. but just in case: */
3483 if (total == 0)
3484 return;
3485
3486 /* don't report if not compressed */
3487 if (total >= plain)
3488 return;
3489
3490 /* total < plain. check for overflow, still */
3491 r = (total > UINT_MAX/1000) ? (total / (plain/1000))
3492 : (1000 * total / plain);
3493
3494 if (r > 1000)
3495 r = 1000;
3496
3497 r = 1000 - r;
3498 dev_info(DEV, "%s bitmap stats [Bytes(packets)]: plain %u(%u), RLE %u(%u), "
3499 "total %u; compression: %u.%u%%\n",
3500 direction,
3501 c->bytes[1], c->packets[1],
3502 c->bytes[0], c->packets[0],
3503 total, r/10, r % 10);
3504}
3505
3506/* Since we are processing the bitfield from lower addresses to higher,
3507 it does not matter if the process it in 32 bit chunks or 64 bit
3508 chunks as long as it is little endian. (Understand it as byte stream,
3509 beginning with the lowest byte...) If we would use big endian
3510 we would need to process it from the highest address to the lowest,
3511 in order to be agnostic to the 32 vs 64 bits issue.
3512
3513 returns 0 on failure, 1 if we successfully received it. */
Philipp Reisner02918be2010-08-20 14:35:10 +02003514static int receive_bitmap(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003515{
3516 struct bm_xfer_ctx c;
3517 void *buffer;
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003518 int err;
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003519 int ok = false;
Philipp Reisner02918be2010-08-20 14:35:10 +02003520 struct p_header80 *h = &mdev->data.rbuf.header.h80;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003521
Lars Ellenberg20ceb2b2011-01-21 10:56:44 +01003522 drbd_bm_lock(mdev, "receive bitmap", BM_LOCKED_SET_ALLOWED);
3523 /* you are supposed to send additional out-of-sync information
3524 * if you actually set bits during this phase */
Philipp Reisnerb411b362009-09-25 16:07:19 -07003525
3526 /* maybe we should use some per thread scratch page,
3527 * and allocate that during initial device creation? */
3528 buffer = (unsigned long *) __get_free_page(GFP_NOIO);
3529 if (!buffer) {
3530 dev_err(DEV, "failed to allocate one page buffer in %s\n", __func__);
3531 goto out;
3532 }
3533
3534 c = (struct bm_xfer_ctx) {
3535 .bm_bits = drbd_bm_bits(mdev),
3536 .bm_words = drbd_bm_words(mdev),
3537 };
3538
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003539 for(;;) {
Philipp Reisner02918be2010-08-20 14:35:10 +02003540 if (cmd == P_BITMAP) {
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003541 err = receive_bitmap_plain(mdev, data_size, buffer, &c);
Philipp Reisner02918be2010-08-20 14:35:10 +02003542 } else if (cmd == P_COMPRESSED_BITMAP) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07003543 /* MAYBE: sanity check that we speak proto >= 90,
3544 * and the feature is enabled! */
3545 struct p_compressed_bm *p;
3546
Philipp Reisner02918be2010-08-20 14:35:10 +02003547 if (data_size > BM_PACKET_PAYLOAD_BYTES) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07003548 dev_err(DEV, "ReportCBitmap packet too large\n");
3549 goto out;
3550 }
3551 /* use the page buff */
3552 p = buffer;
3553 memcpy(p, h, sizeof(*h));
Philipp Reisner02918be2010-08-20 14:35:10 +02003554 if (drbd_recv(mdev, p->head.payload, data_size) != data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003555 goto out;
Lars Ellenberg004352f2010-10-05 20:13:58 +02003556 if (data_size <= (sizeof(*p) - sizeof(p->head))) {
3557 dev_err(DEV, "ReportCBitmap packet too small (l:%u)\n", data_size);
Andreas Gruenbacher78fcbda2010-12-10 22:18:27 +01003558 goto out;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003559 }
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003560 err = decode_bitmap_c(mdev, p, &c);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003561 } else {
Philipp Reisner02918be2010-08-20 14:35:10 +02003562 dev_warn(DEV, "receive_bitmap: cmd neither ReportBitMap nor ReportCBitMap (is 0x%x)", cmd);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003563 goto out;
3564 }
3565
Philipp Reisner02918be2010-08-20 14:35:10 +02003566 c.packets[cmd == P_BITMAP]++;
3567 c.bytes[cmd == P_BITMAP] += sizeof(struct p_header80) + data_size;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003568
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003569 if (err <= 0) {
3570 if (err < 0)
3571 goto out;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003572 break;
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003573 }
Philipp Reisner02918be2010-08-20 14:35:10 +02003574 if (!drbd_recv_header(mdev, &cmd, &data_size))
Philipp Reisnerb411b362009-09-25 16:07:19 -07003575 goto out;
Andreas Gruenbacher2c464072010-12-11 21:53:12 +01003576 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07003577
3578 INFO_bm_xfer_stats(mdev, "receive", &c);
3579
3580 if (mdev->state.conn == C_WF_BITMAP_T) {
Andreas Gruenbacherde1f8e42010-12-10 21:04:00 +01003581 enum drbd_state_rv rv;
3582
Philipp Reisnerb411b362009-09-25 16:07:19 -07003583 ok = !drbd_send_bitmap(mdev);
3584 if (!ok)
3585 goto out;
3586 /* Omit CS_ORDERED with this state transition to avoid deadlocks. */
Andreas Gruenbacherde1f8e42010-12-10 21:04:00 +01003587 rv = _drbd_request_state(mdev, NS(conn, C_WF_SYNC_UUID), CS_VERBOSE);
3588 D_ASSERT(rv == SS_SUCCESS);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003589 } else if (mdev->state.conn != C_WF_BITMAP_S) {
3590 /* admin may have requested C_DISCONNECTING,
3591 * other threads may have noticed network errors */
3592 dev_info(DEV, "unexpected cstate (%s) in receive_bitmap\n",
3593 drbd_conn_str(mdev->state.conn));
3594 }
3595
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003596 ok = true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003597 out:
Lars Ellenberg20ceb2b2011-01-21 10:56:44 +01003598 drbd_bm_unlock(mdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003599 if (ok && mdev->state.conn == C_WF_BITMAP_S)
3600 drbd_start_resync(mdev, C_SYNC_SOURCE);
3601 free_page((unsigned long) buffer);
3602 return ok;
3603}
3604
Philipp Reisner02918be2010-08-20 14:35:10 +02003605static int receive_skip(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003606{
3607 /* TODO zero copy sink :) */
3608 static char sink[128];
3609 int size, want, r;
3610
Philipp Reisner02918be2010-08-20 14:35:10 +02003611 dev_warn(DEV, "skipping unknown optional packet type %d, l: %d!\n",
3612 cmd, data_size);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003613
Philipp Reisner02918be2010-08-20 14:35:10 +02003614 size = data_size;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003615 while (size > 0) {
3616 want = min_t(int, size, sizeof(sink));
3617 r = drbd_recv(mdev, sink, want);
3618 ERR_IF(r <= 0) break;
3619 size -= r;
3620 }
3621 return size == 0;
3622}
3623
Philipp Reisner02918be2010-08-20 14:35:10 +02003624static int receive_UnplugRemote(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
Philipp Reisnerb411b362009-09-25 16:07:19 -07003625{
Philipp Reisnerb411b362009-09-25 16:07:19 -07003626 /* Make sure we've acked all the TCP data associated
3627 * with the data requests being unplugged */
3628 drbd_tcp_quickack(mdev->data.socket);
3629
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003630 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003631}
3632
Philipp Reisner73a01a12010-10-27 14:33:00 +02003633static int receive_out_of_sync(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned int data_size)
3634{
3635 struct p_block_desc *p = &mdev->data.rbuf.block_desc;
3636
Lars Ellenbergf735e3632010-12-17 21:06:18 +01003637 switch (mdev->state.conn) {
3638 case C_WF_SYNC_UUID:
3639 case C_WF_BITMAP_T:
3640 case C_BEHIND:
3641 break;
3642 default:
3643 dev_err(DEV, "ASSERT FAILED cstate = %s, expected: WFSyncUUID|WFBitMapT|Behind\n",
3644 drbd_conn_str(mdev->state.conn));
3645 }
3646
Philipp Reisner73a01a12010-10-27 14:33:00 +02003647 drbd_set_out_of_sync(mdev, be64_to_cpu(p->sector), be32_to_cpu(p->blksize));
3648
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01003649 return true;
Philipp Reisner73a01a12010-10-27 14:33:00 +02003650}
3651
Philipp Reisner02918be2010-08-20 14:35:10 +02003652typedef int (*drbd_cmd_handler_f)(struct drbd_conf *, enum drbd_packets cmd, unsigned int to_receive);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003653
Philipp Reisner02918be2010-08-20 14:35:10 +02003654struct data_cmd {
3655 int expect_payload;
3656 size_t pkt_size;
3657 drbd_cmd_handler_f function;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003658};
3659
Philipp Reisner02918be2010-08-20 14:35:10 +02003660static struct data_cmd drbd_cmd_handler[] = {
3661 [P_DATA] = { 1, sizeof(struct p_data), receive_Data },
3662 [P_DATA_REPLY] = { 1, sizeof(struct p_data), receive_DataReply },
3663 [P_RS_DATA_REPLY] = { 1, sizeof(struct p_data), receive_RSDataReply } ,
3664 [P_BARRIER] = { 0, sizeof(struct p_barrier), receive_Barrier } ,
3665 [P_BITMAP] = { 1, sizeof(struct p_header80), receive_bitmap } ,
3666 [P_COMPRESSED_BITMAP] = { 1, sizeof(struct p_header80), receive_bitmap } ,
3667 [P_UNPLUG_REMOTE] = { 0, sizeof(struct p_header80), receive_UnplugRemote },
3668 [P_DATA_REQUEST] = { 0, sizeof(struct p_block_req), receive_DataRequest },
3669 [P_RS_DATA_REQUEST] = { 0, sizeof(struct p_block_req), receive_DataRequest },
3670 [P_SYNC_PARAM] = { 1, sizeof(struct p_header80), receive_SyncParam },
3671 [P_SYNC_PARAM89] = { 1, sizeof(struct p_header80), receive_SyncParam },
3672 [P_PROTOCOL] = { 1, sizeof(struct p_protocol), receive_protocol },
3673 [P_UUIDS] = { 0, sizeof(struct p_uuids), receive_uuids },
3674 [P_SIZES] = { 0, sizeof(struct p_sizes), receive_sizes },
3675 [P_STATE] = { 0, sizeof(struct p_state), receive_state },
3676 [P_STATE_CHG_REQ] = { 0, sizeof(struct p_req_state), receive_req_state },
3677 [P_SYNC_UUID] = { 0, sizeof(struct p_rs_uuid), receive_sync_uuid },
3678 [P_OV_REQUEST] = { 0, sizeof(struct p_block_req), receive_DataRequest },
3679 [P_OV_REPLY] = { 1, sizeof(struct p_block_req), receive_DataRequest },
3680 [P_CSUM_RS_REQUEST] = { 1, sizeof(struct p_block_req), receive_DataRequest },
3681 [P_DELAY_PROBE] = { 0, sizeof(struct p_delay_probe93), receive_skip },
Philipp Reisner73a01a12010-10-27 14:33:00 +02003682 [P_OUT_OF_SYNC] = { 0, sizeof(struct p_block_desc), receive_out_of_sync },
Philipp Reisner02918be2010-08-20 14:35:10 +02003683 /* anything missing from this table is in
3684 * the asender_tbl, see get_asender_cmd */
3685 [P_MAX_CMD] = { 0, 0, NULL },
3686};
3687
3688/* All handler functions that expect a sub-header get that sub-heder in
3689 mdev->data.rbuf.header.head.payload.
3690
3691 Usually in mdev->data.rbuf.header.head the callback can find the usual
3692 p_header, but they may not rely on that. Since there is also p_header95 !
3693 */
Philipp Reisnerb411b362009-09-25 16:07:19 -07003694
3695static void drbdd(struct drbd_conf *mdev)
3696{
Philipp Reisner02918be2010-08-20 14:35:10 +02003697 union p_header *header = &mdev->data.rbuf.header;
3698 unsigned int packet_size;
3699 enum drbd_packets cmd;
3700 size_t shs; /* sub header size */
3701 int rv;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003702
Andreas Gruenbachere77a0a52011-01-25 15:43:39 +01003703 while (get_t_state(&mdev->receiver) == RUNNING) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07003704 drbd_thread_current_set_cpu(mdev);
Philipp Reisner02918be2010-08-20 14:35:10 +02003705 if (!drbd_recv_header(mdev, &cmd, &packet_size))
3706 goto err_out;
3707
3708 if (unlikely(cmd >= P_MAX_CMD || !drbd_cmd_handler[cmd].function)) {
3709 dev_err(DEV, "unknown packet type %d, l: %d!\n", cmd, packet_size);
3710 goto err_out;
Lars Ellenberg0b33a912009-11-16 15:58:04 +01003711 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07003712
Philipp Reisner02918be2010-08-20 14:35:10 +02003713 shs = drbd_cmd_handler[cmd].pkt_size - sizeof(union p_header);
Philipp Reisner02918be2010-08-20 14:35:10 +02003714 if (packet_size - shs > 0 && !drbd_cmd_handler[cmd].expect_payload) {
3715 dev_err(DEV, "No payload expected %s l:%d\n", cmdname(cmd), packet_size);
3716 goto err_out;
3717 }
3718
Lars Ellenbergc13f7e12010-10-29 23:32:01 +02003719 if (shs) {
3720 rv = drbd_recv(mdev, &header->h80.payload, shs);
3721 if (unlikely(rv != shs)) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01003722 if (!signal_pending(current))
3723 dev_warn(DEV, "short read while reading sub header: rv=%d\n", rv);
Lars Ellenbergc13f7e12010-10-29 23:32:01 +02003724 goto err_out;
3725 }
3726 }
3727
Philipp Reisner02918be2010-08-20 14:35:10 +02003728 rv = drbd_cmd_handler[cmd].function(mdev, cmd, packet_size - shs);
3729
3730 if (unlikely(!rv)) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07003731 dev_err(DEV, "error receiving %s, l: %d!\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02003732 cmdname(cmd), packet_size);
3733 goto err_out;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003734 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07003735 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07003736
Philipp Reisner02918be2010-08-20 14:35:10 +02003737 if (0) {
3738 err_out:
3739 drbd_force_state(mdev, NS(conn, C_PROTOCOL_ERROR));
Philipp Reisnerb411b362009-09-25 16:07:19 -07003740 }
Lars Ellenberg856c50c2010-10-14 13:37:40 +02003741 /* If we leave here, we probably want to update at least the
3742 * "Connected" indicator on stable storage. Do so explicitly here. */
3743 drbd_md_sync(mdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003744}
3745
3746void drbd_flush_workqueue(struct drbd_conf *mdev)
3747{
3748 struct drbd_wq_barrier barr;
3749
3750 barr.w.cb = w_prev_work_done;
3751 init_completion(&barr.done);
3752 drbd_queue_work(&mdev->data.work, &barr.w);
3753 wait_for_completion(&barr.done);
3754}
3755
3756static void drbd_disconnect(struct drbd_conf *mdev)
3757{
3758 enum drbd_fencing_p fp;
3759 union drbd_state os, ns;
3760 int rv = SS_UNKNOWN_ERROR;
3761 unsigned int i;
3762
3763 if (mdev->state.conn == C_STANDALONE)
3764 return;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003765
3766 /* asender does not clean up anything. it must not interfere, either */
3767 drbd_thread_stop(&mdev->asender);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003768 drbd_free_sock(mdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003769
Philipp Reisner85719572010-07-21 10:20:17 +02003770 /* wait for current activity to cease. */
Philipp Reisnerb411b362009-09-25 16:07:19 -07003771 spin_lock_irq(&mdev->req_lock);
3772 _drbd_wait_ee_list_empty(mdev, &mdev->active_ee);
3773 _drbd_wait_ee_list_empty(mdev, &mdev->sync_ee);
3774 _drbd_wait_ee_list_empty(mdev, &mdev->read_ee);
3775 spin_unlock_irq(&mdev->req_lock);
3776
3777 /* We do not have data structures that would allow us to
3778 * get the rs_pending_cnt down to 0 again.
3779 * * On C_SYNC_TARGET we do not have any data structures describing
3780 * the pending RSDataRequest's we have sent.
3781 * * On C_SYNC_SOURCE there is no data structure that tracks
3782 * the P_RS_DATA_REPLY blocks that we sent to the SyncTarget.
3783 * And no, it is not the sum of the reference counts in the
3784 * resync_LRU. The resync_LRU tracks the whole operation including
3785 * the disk-IO, while the rs_pending_cnt only tracks the blocks
3786 * on the fly. */
3787 drbd_rs_cancel_all(mdev);
3788 mdev->rs_total = 0;
3789 mdev->rs_failed = 0;
3790 atomic_set(&mdev->rs_pending_cnt, 0);
3791 wake_up(&mdev->misc_wait);
3792
Philipp Reisner7fde2be2011-03-01 11:08:28 +01003793 del_timer(&mdev->request_timer);
3794
Philipp Reisnerb411b362009-09-25 16:07:19 -07003795 /* make sure syncer is stopped and w_resume_next_sg queued */
3796 del_timer_sync(&mdev->resync_timer);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003797 resync_timer_fn((unsigned long)mdev);
3798
Philipp Reisnerb411b362009-09-25 16:07:19 -07003799 /* wait for all w_e_end_data_req, w_e_end_rsdata_req, w_send_barrier,
3800 * w_make_resync_request etc. which may still be on the worker queue
3801 * to be "canceled" */
3802 drbd_flush_workqueue(mdev);
3803
3804 /* This also does reclaim_net_ee(). If we do this too early, we might
3805 * miss some resync ee and pages.*/
3806 drbd_process_done_ee(mdev);
3807
3808 kfree(mdev->p_uuid);
3809 mdev->p_uuid = NULL;
3810
Philipp Reisnerfb22c402010-09-08 23:20:21 +02003811 if (!is_susp(mdev->state))
Philipp Reisnerb411b362009-09-25 16:07:19 -07003812 tl_clear(mdev);
3813
Philipp Reisnerb411b362009-09-25 16:07:19 -07003814 dev_info(DEV, "Connection closed\n");
3815
3816 drbd_md_sync(mdev);
3817
3818 fp = FP_DONT_CARE;
3819 if (get_ldev(mdev)) {
3820 fp = mdev->ldev->dc.fencing;
3821 put_ldev(mdev);
3822 }
3823
Philipp Reisner87f7be42010-06-11 13:56:33 +02003824 if (mdev->state.role == R_PRIMARY && fp >= FP_RESOURCE && mdev->state.pdsk >= D_UNKNOWN)
3825 drbd_try_outdate_peer_async(mdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003826
3827 spin_lock_irq(&mdev->req_lock);
3828 os = mdev->state;
3829 if (os.conn >= C_UNCONNECTED) {
3830 /* Do not restart in case we are C_DISCONNECTING */
3831 ns = os;
3832 ns.conn = C_UNCONNECTED;
3833 rv = _drbd_set_state(mdev, ns, CS_VERBOSE, NULL);
3834 }
3835 spin_unlock_irq(&mdev->req_lock);
3836
3837 if (os.conn == C_DISCONNECTING) {
Philipp Reisner84dfb9f2010-06-23 11:20:05 +02003838 wait_event(mdev->net_cnt_wait, atomic_read(&mdev->net_cnt) == 0);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003839
Philipp Reisnerb411b362009-09-25 16:07:19 -07003840 crypto_free_hash(mdev->cram_hmac_tfm);
3841 mdev->cram_hmac_tfm = NULL;
3842
3843 kfree(mdev->net_conf);
3844 mdev->net_conf = NULL;
3845 drbd_request_state(mdev, NS(conn, C_STANDALONE));
3846 }
3847
Lars Ellenberg20ceb2b2011-01-21 10:56:44 +01003848 /* serialize with bitmap writeout triggered by the state change,
3849 * if any. */
3850 wait_event(mdev->misc_wait, !test_bit(BITMAP_IO, &mdev->flags));
3851
Philipp Reisnerb411b362009-09-25 16:07:19 -07003852 /* tcp_close and release of sendpage pages can be deferred. I don't
3853 * want to use SO_LINGER, because apparently it can be deferred for
3854 * more than 20 seconds (longest time I checked).
3855 *
3856 * Actually we don't care for exactly when the network stack does its
3857 * put_page(), but release our reference on these pages right here.
3858 */
3859 i = drbd_release_ee(mdev, &mdev->net_ee);
3860 if (i)
3861 dev_info(DEV, "net_ee not empty, killed %u entries\n", i);
Lars Ellenberg435f0742010-09-06 12:30:25 +02003862 i = atomic_read(&mdev->pp_in_use_by_net);
3863 if (i)
3864 dev_info(DEV, "pp_in_use_by_net = %d, expected 0\n", i);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003865 i = atomic_read(&mdev->pp_in_use);
3866 if (i)
Lars Ellenberg45bb9122010-05-14 17:10:48 +02003867 dev_info(DEV, "pp_in_use = %d, expected 0\n", i);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003868
3869 D_ASSERT(list_empty(&mdev->read_ee));
3870 D_ASSERT(list_empty(&mdev->active_ee));
3871 D_ASSERT(list_empty(&mdev->sync_ee));
3872 D_ASSERT(list_empty(&mdev->done_ee));
3873
3874 /* ok, no more ee's on the fly, it is safe to reset the epoch_size */
3875 atomic_set(&mdev->current_epoch->epoch_size, 0);
3876 D_ASSERT(list_empty(&mdev->current_epoch->list));
3877}
3878
3879/*
3880 * We support PRO_VERSION_MIN to PRO_VERSION_MAX. The protocol version
3881 * we can agree on is stored in agreed_pro_version.
3882 *
3883 * feature flags and the reserved array should be enough room for future
3884 * enhancements of the handshake protocol, and possible plugins...
3885 *
3886 * for now, they are expected to be zero, but ignored.
3887 */
3888static int drbd_send_handshake(struct drbd_conf *mdev)
3889{
3890 /* ASSERT current == mdev->receiver ... */
3891 struct p_handshake *p = &mdev->data.sbuf.handshake;
3892 int ok;
3893
3894 if (mutex_lock_interruptible(&mdev->data.mutex)) {
3895 dev_err(DEV, "interrupted during initial handshake\n");
3896 return 0; /* interrupted. not ok. */
3897 }
3898
3899 if (mdev->data.socket == NULL) {
3900 mutex_unlock(&mdev->data.mutex);
3901 return 0;
3902 }
3903
3904 memset(p, 0, sizeof(*p));
3905 p->protocol_min = cpu_to_be32(PRO_VERSION_MIN);
3906 p->protocol_max = cpu_to_be32(PRO_VERSION_MAX);
3907 ok = _drbd_send_cmd( mdev, mdev->data.socket, P_HAND_SHAKE,
Philipp Reisner0b70a132010-08-20 13:36:10 +02003908 (struct p_header80 *)p, sizeof(*p), 0 );
Philipp Reisnerb411b362009-09-25 16:07:19 -07003909 mutex_unlock(&mdev->data.mutex);
3910 return ok;
3911}
3912
3913/*
3914 * return values:
3915 * 1 yes, we have a valid connection
3916 * 0 oops, did not work out, please try again
3917 * -1 peer talks different language,
3918 * no point in trying again, please go standalone.
3919 */
3920static int drbd_do_handshake(struct drbd_conf *mdev)
3921{
3922 /* ASSERT current == mdev->receiver ... */
3923 struct p_handshake *p = &mdev->data.rbuf.handshake;
Philipp Reisner02918be2010-08-20 14:35:10 +02003924 const int expect = sizeof(struct p_handshake) - sizeof(struct p_header80);
3925 unsigned int length;
3926 enum drbd_packets cmd;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003927 int rv;
3928
3929 rv = drbd_send_handshake(mdev);
3930 if (!rv)
3931 return 0;
3932
Philipp Reisner02918be2010-08-20 14:35:10 +02003933 rv = drbd_recv_header(mdev, &cmd, &length);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003934 if (!rv)
3935 return 0;
3936
Philipp Reisner02918be2010-08-20 14:35:10 +02003937 if (cmd != P_HAND_SHAKE) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07003938 dev_err(DEV, "expected HandShake packet, received: %s (0x%04x)\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02003939 cmdname(cmd), cmd);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003940 return -1;
3941 }
3942
Philipp Reisner02918be2010-08-20 14:35:10 +02003943 if (length != expect) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07003944 dev_err(DEV, "expected HandShake length: %u, received: %u\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02003945 expect, length);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003946 return -1;
3947 }
3948
3949 rv = drbd_recv(mdev, &p->head.payload, expect);
3950
3951 if (rv != expect) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01003952 if (!signal_pending(current))
3953 dev_warn(DEV, "short read receiving handshake packet: l=%u\n", rv);
Philipp Reisnerb411b362009-09-25 16:07:19 -07003954 return 0;
3955 }
3956
Philipp Reisnerb411b362009-09-25 16:07:19 -07003957 p->protocol_min = be32_to_cpu(p->protocol_min);
3958 p->protocol_max = be32_to_cpu(p->protocol_max);
3959 if (p->protocol_max == 0)
3960 p->protocol_max = p->protocol_min;
3961
3962 if (PRO_VERSION_MAX < p->protocol_min ||
3963 PRO_VERSION_MIN > p->protocol_max)
3964 goto incompat;
3965
3966 mdev->agreed_pro_version = min_t(int, PRO_VERSION_MAX, p->protocol_max);
3967
3968 dev_info(DEV, "Handshake successful: "
3969 "Agreed network protocol version %d\n", mdev->agreed_pro_version);
3970
3971 return 1;
3972
3973 incompat:
3974 dev_err(DEV, "incompatible DRBD dialects: "
3975 "I support %d-%d, peer supports %d-%d\n",
3976 PRO_VERSION_MIN, PRO_VERSION_MAX,
3977 p->protocol_min, p->protocol_max);
3978 return -1;
3979}
3980
3981#if !defined(CONFIG_CRYPTO_HMAC) && !defined(CONFIG_CRYPTO_HMAC_MODULE)
3982static int drbd_do_auth(struct drbd_conf *mdev)
3983{
3984 dev_err(DEV, "This kernel was build without CONFIG_CRYPTO_HMAC.\n");
3985 dev_err(DEV, "You need to disable 'cram-hmac-alg' in drbd.conf.\n");
Johannes Thomab10d96c2010-01-07 16:02:50 +01003986 return -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07003987}
3988#else
3989#define CHALLENGE_LEN 64
Johannes Thomab10d96c2010-01-07 16:02:50 +01003990
3991/* Return value:
3992 1 - auth succeeded,
3993 0 - failed, try again (network error),
3994 -1 - auth failed, don't try again.
3995*/
3996
Philipp Reisnerb411b362009-09-25 16:07:19 -07003997static int drbd_do_auth(struct drbd_conf *mdev)
3998{
3999 char my_challenge[CHALLENGE_LEN]; /* 64 Bytes... */
4000 struct scatterlist sg;
4001 char *response = NULL;
4002 char *right_response = NULL;
4003 char *peers_ch = NULL;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004004 unsigned int key_len = strlen(mdev->net_conf->shared_secret);
4005 unsigned int resp_size;
4006 struct hash_desc desc;
Philipp Reisner02918be2010-08-20 14:35:10 +02004007 enum drbd_packets cmd;
4008 unsigned int length;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004009 int rv;
4010
4011 desc.tfm = mdev->cram_hmac_tfm;
4012 desc.flags = 0;
4013
4014 rv = crypto_hash_setkey(mdev->cram_hmac_tfm,
4015 (u8 *)mdev->net_conf->shared_secret, key_len);
4016 if (rv) {
4017 dev_err(DEV, "crypto_hash_setkey() failed with %d\n", rv);
Johannes Thomab10d96c2010-01-07 16:02:50 +01004018 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004019 goto fail;
4020 }
4021
4022 get_random_bytes(my_challenge, CHALLENGE_LEN);
4023
4024 rv = drbd_send_cmd2(mdev, P_AUTH_CHALLENGE, my_challenge, CHALLENGE_LEN);
4025 if (!rv)
4026 goto fail;
4027
Philipp Reisner02918be2010-08-20 14:35:10 +02004028 rv = drbd_recv_header(mdev, &cmd, &length);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004029 if (!rv)
4030 goto fail;
4031
Philipp Reisner02918be2010-08-20 14:35:10 +02004032 if (cmd != P_AUTH_CHALLENGE) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004033 dev_err(DEV, "expected AuthChallenge packet, received: %s (0x%04x)\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02004034 cmdname(cmd), cmd);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004035 rv = 0;
4036 goto fail;
4037 }
4038
Philipp Reisner02918be2010-08-20 14:35:10 +02004039 if (length > CHALLENGE_LEN * 2) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004040 dev_err(DEV, "expected AuthChallenge payload too big.\n");
Johannes Thomab10d96c2010-01-07 16:02:50 +01004041 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004042 goto fail;
4043 }
4044
Philipp Reisner02918be2010-08-20 14:35:10 +02004045 peers_ch = kmalloc(length, GFP_NOIO);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004046 if (peers_ch == NULL) {
4047 dev_err(DEV, "kmalloc of peers_ch failed\n");
Johannes Thomab10d96c2010-01-07 16:02:50 +01004048 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004049 goto fail;
4050 }
4051
Philipp Reisner02918be2010-08-20 14:35:10 +02004052 rv = drbd_recv(mdev, peers_ch, length);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004053
Philipp Reisner02918be2010-08-20 14:35:10 +02004054 if (rv != length) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01004055 if (!signal_pending(current))
4056 dev_warn(DEV, "short read AuthChallenge: l=%u\n", rv);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004057 rv = 0;
4058 goto fail;
4059 }
4060
4061 resp_size = crypto_hash_digestsize(mdev->cram_hmac_tfm);
4062 response = kmalloc(resp_size, GFP_NOIO);
4063 if (response == NULL) {
4064 dev_err(DEV, "kmalloc of response failed\n");
Johannes Thomab10d96c2010-01-07 16:02:50 +01004065 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004066 goto fail;
4067 }
4068
4069 sg_init_table(&sg, 1);
Philipp Reisner02918be2010-08-20 14:35:10 +02004070 sg_set_buf(&sg, peers_ch, length);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004071
4072 rv = crypto_hash_digest(&desc, &sg, sg.length, response);
4073 if (rv) {
4074 dev_err(DEV, "crypto_hash_digest() failed with %d\n", rv);
Johannes Thomab10d96c2010-01-07 16:02:50 +01004075 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004076 goto fail;
4077 }
4078
4079 rv = drbd_send_cmd2(mdev, P_AUTH_RESPONSE, response, resp_size);
4080 if (!rv)
4081 goto fail;
4082
Philipp Reisner02918be2010-08-20 14:35:10 +02004083 rv = drbd_recv_header(mdev, &cmd, &length);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004084 if (!rv)
4085 goto fail;
4086
Philipp Reisner02918be2010-08-20 14:35:10 +02004087 if (cmd != P_AUTH_RESPONSE) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004088 dev_err(DEV, "expected AuthResponse packet, received: %s (0x%04x)\n",
Philipp Reisner02918be2010-08-20 14:35:10 +02004089 cmdname(cmd), cmd);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004090 rv = 0;
4091 goto fail;
4092 }
4093
Philipp Reisner02918be2010-08-20 14:35:10 +02004094 if (length != resp_size) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004095 dev_err(DEV, "expected AuthResponse payload of wrong size\n");
4096 rv = 0;
4097 goto fail;
4098 }
4099
4100 rv = drbd_recv(mdev, response , resp_size);
4101
4102 if (rv != resp_size) {
Lars Ellenberg0ddc5542011-01-21 12:35:15 +01004103 if (!signal_pending(current))
4104 dev_warn(DEV, "short read receiving AuthResponse: l=%u\n", rv);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004105 rv = 0;
4106 goto fail;
4107 }
4108
4109 right_response = kmalloc(resp_size, GFP_NOIO);
Julia Lawall2d1ee872009-12-27 22:27:11 +01004110 if (right_response == NULL) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004111 dev_err(DEV, "kmalloc of right_response failed\n");
Johannes Thomab10d96c2010-01-07 16:02:50 +01004112 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004113 goto fail;
4114 }
4115
4116 sg_set_buf(&sg, my_challenge, CHALLENGE_LEN);
4117
4118 rv = crypto_hash_digest(&desc, &sg, sg.length, right_response);
4119 if (rv) {
4120 dev_err(DEV, "crypto_hash_digest() failed with %d\n", rv);
Johannes Thomab10d96c2010-01-07 16:02:50 +01004121 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004122 goto fail;
4123 }
4124
4125 rv = !memcmp(response, right_response, resp_size);
4126
4127 if (rv)
4128 dev_info(DEV, "Peer authenticated using %d bytes of '%s' HMAC\n",
4129 resp_size, mdev->net_conf->cram_hmac_alg);
Johannes Thomab10d96c2010-01-07 16:02:50 +01004130 else
4131 rv = -1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004132
4133 fail:
4134 kfree(peers_ch);
4135 kfree(response);
4136 kfree(right_response);
4137
4138 return rv;
4139}
4140#endif
4141
4142int drbdd_init(struct drbd_thread *thi)
4143{
4144 struct drbd_conf *mdev = thi->mdev;
4145 unsigned int minor = mdev_to_minor(mdev);
4146 int h;
4147
4148 sprintf(current->comm, "drbd%d_receiver", minor);
4149
4150 dev_info(DEV, "receiver (re)started\n");
4151
4152 do {
4153 h = drbd_connect(mdev);
4154 if (h == 0) {
4155 drbd_disconnect(mdev);
Philipp Reisner20ee6392011-01-18 15:28:59 +01004156 schedule_timeout_interruptible(HZ);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004157 }
4158 if (h == -1) {
4159 dev_warn(DEV, "Discarding network configuration.\n");
4160 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
4161 }
4162 } while (h == 0);
4163
4164 if (h > 0) {
4165 if (get_net_conf(mdev)) {
4166 drbdd(mdev);
4167 put_net_conf(mdev);
4168 }
4169 }
4170
4171 drbd_disconnect(mdev);
4172
4173 dev_info(DEV, "receiver terminated\n");
4174 return 0;
4175}
4176
4177/* ********* acknowledge sender ******** */
4178
Philipp Reisner0b70a132010-08-20 13:36:10 +02004179static int got_RqSReply(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004180{
4181 struct p_req_state_reply *p = (struct p_req_state_reply *)h;
4182
4183 int retcode = be32_to_cpu(p->retcode);
4184
4185 if (retcode >= SS_SUCCESS) {
4186 set_bit(CL_ST_CHG_SUCCESS, &mdev->flags);
4187 } else {
4188 set_bit(CL_ST_CHG_FAIL, &mdev->flags);
4189 dev_err(DEV, "Requested state change failed by peer: %s (%d)\n",
4190 drbd_set_st_err_str(retcode), retcode);
4191 }
4192 wake_up(&mdev->state_wait);
4193
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004194 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004195}
4196
Philipp Reisner0b70a132010-08-20 13:36:10 +02004197static int got_Ping(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004198{
4199 return drbd_send_ping_ack(mdev);
4200
4201}
4202
Philipp Reisner0b70a132010-08-20 13:36:10 +02004203static int got_PingAck(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004204{
4205 /* restore idle timeout */
4206 mdev->meta.socket->sk->sk_rcvtimeo = mdev->net_conf->ping_int*HZ;
Philipp Reisner309d1602010-03-02 15:03:44 +01004207 if (!test_and_set_bit(GOT_PING_ACK, &mdev->flags))
4208 wake_up(&mdev->misc_wait);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004209
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004210 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004211}
4212
Philipp Reisner0b70a132010-08-20 13:36:10 +02004213static int got_IsInSync(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004214{
4215 struct p_block_ack *p = (struct p_block_ack *)h;
4216 sector_t sector = be64_to_cpu(p->sector);
4217 int blksize = be32_to_cpu(p->blksize);
4218
4219 D_ASSERT(mdev->agreed_pro_version >= 89);
4220
4221 update_peer_seq(mdev, be32_to_cpu(p->seq_num));
4222
Lars Ellenberg1d53f092010-09-05 01:13:24 +02004223 if (get_ldev(mdev)) {
4224 drbd_rs_complete_io(mdev, sector);
4225 drbd_set_in_sync(mdev, sector, blksize);
4226 /* rs_same_csums is supposed to count in units of BM_BLOCK_SIZE */
4227 mdev->rs_same_csum += (blksize >> BM_BLOCK_SHIFT);
4228 put_ldev(mdev);
4229 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07004230 dec_rs_pending(mdev);
Philipp Reisner778f2712010-07-06 11:14:00 +02004231 atomic_add(blksize >> 9, &mdev->rs_sect_in);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004232
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004233 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004234}
4235
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01004236static int
4237validate_req_change_req_state(struct drbd_conf *mdev, u64 id, sector_t sector,
4238 struct rb_root *root, const char *func,
4239 enum drbd_req_event what, bool missing_ok)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004240{
4241 struct drbd_request *req;
4242 struct bio_and_error m;
4243
4244 spin_lock_irq(&mdev->req_lock);
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01004245 req = find_request(mdev, root, id, sector, missing_ok, func);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004246 if (unlikely(!req)) {
4247 spin_unlock_irq(&mdev->req_lock);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004248 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004249 }
4250 __req_mod(req, what, &m);
4251 spin_unlock_irq(&mdev->req_lock);
4252
4253 if (m.bio)
4254 complete_master_bio(mdev, &m);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004255 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004256}
4257
Philipp Reisner0b70a132010-08-20 13:36:10 +02004258static int got_BlockAck(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004259{
4260 struct p_block_ack *p = (struct p_block_ack *)h;
4261 sector_t sector = be64_to_cpu(p->sector);
4262 int blksize = be32_to_cpu(p->blksize);
4263 enum drbd_req_event what;
4264
4265 update_peer_seq(mdev, be32_to_cpu(p->seq_num));
4266
Andreas Gruenbacher579b57e2011-01-13 18:40:57 +01004267 if (p->block_id == ID_SYNCER) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004268 drbd_set_in_sync(mdev, sector, blksize);
4269 dec_rs_pending(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004270 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004271 }
4272 switch (be16_to_cpu(h->command)) {
4273 case P_RS_WRITE_ACK:
4274 D_ASSERT(mdev->net_conf->wire_protocol == DRBD_PROT_C);
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01004275 what = WRITE_ACKED_BY_PEER_AND_SIS;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004276 break;
4277 case P_WRITE_ACK:
4278 D_ASSERT(mdev->net_conf->wire_protocol == DRBD_PROT_C);
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01004279 what = WRITE_ACKED_BY_PEER;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004280 break;
4281 case P_RECV_ACK:
4282 D_ASSERT(mdev->net_conf->wire_protocol == DRBD_PROT_B);
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01004283 what = RECV_ACKED_BY_PEER;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004284 break;
4285 case P_DISCARD_ACK:
4286 D_ASSERT(mdev->net_conf->wire_protocol == DRBD_PROT_C);
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01004287 what = CONFLICT_DISCARDED_BY_PEER;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004288 break;
4289 default:
4290 D_ASSERT(0);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004291 return false;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004292 }
4293
4294 return validate_req_change_req_state(mdev, p->block_id, sector,
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01004295 &mdev->write_requests, __func__,
4296 what, false);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004297}
4298
Philipp Reisner0b70a132010-08-20 13:36:10 +02004299static int got_NegAck(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004300{
4301 struct p_block_ack *p = (struct p_block_ack *)h;
4302 sector_t sector = be64_to_cpu(p->sector);
Philipp Reisner2deb8332011-01-17 18:39:18 +01004303 int size = be32_to_cpu(p->blksize);
Andreas Gruenbacherc3afd8f2011-01-20 22:25:40 +01004304 bool missing_ok = mdev->net_conf->wire_protocol == DRBD_PROT_A ||
4305 mdev->net_conf->wire_protocol == DRBD_PROT_B;
4306 bool found;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004307
4308 update_peer_seq(mdev, be32_to_cpu(p->seq_num));
4309
Andreas Gruenbacher579b57e2011-01-13 18:40:57 +01004310 if (p->block_id == ID_SYNCER) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004311 dec_rs_pending(mdev);
4312 drbd_rs_failed_io(mdev, sector, size);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004313 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004314 }
Philipp Reisner2deb8332011-01-17 18:39:18 +01004315
Andreas Gruenbacherc3afd8f2011-01-20 22:25:40 +01004316 found = validate_req_change_req_state(mdev, p->block_id, sector,
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01004317 &mdev->write_requests, __func__,
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01004318 NEG_ACKED, missing_ok);
Andreas Gruenbacherc3afd8f2011-01-20 22:25:40 +01004319 if (!found) {
4320 /* Protocol A has no P_WRITE_ACKs, but has P_NEG_ACKs.
4321 The master bio might already be completed, therefore the
4322 request is no longer in the collision hash. */
4323 /* In Protocol B we might already have got a P_RECV_ACK
4324 but then get a P_NEG_ACK afterwards. */
4325 if (!missing_ok)
Philipp Reisner2deb8332011-01-17 18:39:18 +01004326 return false;
Andreas Gruenbacherc3afd8f2011-01-20 22:25:40 +01004327 drbd_set_out_of_sync(mdev, sector, size);
Philipp Reisner2deb8332011-01-17 18:39:18 +01004328 }
Philipp Reisner2deb8332011-01-17 18:39:18 +01004329 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004330}
4331
Philipp Reisner0b70a132010-08-20 13:36:10 +02004332static int got_NegDReply(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004333{
4334 struct p_block_ack *p = (struct p_block_ack *)h;
4335 sector_t sector = be64_to_cpu(p->sector);
4336
4337 update_peer_seq(mdev, be32_to_cpu(p->seq_num));
4338 dev_err(DEV, "Got NegDReply; Sector %llus, len %u; Fail original request.\n",
4339 (unsigned long long)sector, be32_to_cpu(p->blksize));
4340
4341 return validate_req_change_req_state(mdev, p->block_id, sector,
Andreas Gruenbacherbc9c5c42011-01-21 18:00:55 +01004342 &mdev->read_requests, __func__,
Andreas Gruenbacher8554df12011-01-25 15:37:43 +01004343 NEG_ACKED, false);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004344}
4345
Philipp Reisner0b70a132010-08-20 13:36:10 +02004346static int got_NegRSDReply(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004347{
4348 sector_t sector;
4349 int size;
4350 struct p_block_ack *p = (struct p_block_ack *)h;
4351
4352 sector = be64_to_cpu(p->sector);
4353 size = be32_to_cpu(p->blksize);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004354
4355 update_peer_seq(mdev, be32_to_cpu(p->seq_num));
4356
4357 dec_rs_pending(mdev);
4358
4359 if (get_ldev_if_state(mdev, D_FAILED)) {
4360 drbd_rs_complete_io(mdev, sector);
Philipp Reisnerd612d302010-12-27 10:53:28 +01004361 switch (be16_to_cpu(h->command)) {
4362 case P_NEG_RS_DREPLY:
4363 drbd_rs_failed_io(mdev, sector, size);
4364 case P_RS_CANCEL:
4365 break;
4366 default:
4367 D_ASSERT(0);
4368 put_ldev(mdev);
4369 return false;
4370 }
Philipp Reisnerb411b362009-09-25 16:07:19 -07004371 put_ldev(mdev);
4372 }
4373
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004374 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004375}
4376
Philipp Reisner0b70a132010-08-20 13:36:10 +02004377static int got_BarrierAck(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004378{
4379 struct p_barrier_ack *p = (struct p_barrier_ack *)h;
4380
4381 tl_release(mdev, p->barrier, be32_to_cpu(p->set_size));
4382
Philipp Reisnerc4752ef2010-10-27 17:32:36 +02004383 if (mdev->state.conn == C_AHEAD &&
4384 atomic_read(&mdev->ap_in_flight) == 0 &&
Philipp Reisner370a43e2011-01-14 16:03:11 +01004385 !test_and_set_bit(AHEAD_TO_SYNC_SOURCE, &mdev->current_epoch->flags)) {
4386 mdev->start_resync_timer.expires = jiffies + HZ;
4387 add_timer(&mdev->start_resync_timer);
Philipp Reisnerc4752ef2010-10-27 17:32:36 +02004388 }
4389
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004390 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004391}
4392
Philipp Reisner0b70a132010-08-20 13:36:10 +02004393static int got_OVResult(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisnerb411b362009-09-25 16:07:19 -07004394{
4395 struct p_block_ack *p = (struct p_block_ack *)h;
4396 struct drbd_work *w;
4397 sector_t sector;
4398 int size;
4399
4400 sector = be64_to_cpu(p->sector);
4401 size = be32_to_cpu(p->blksize);
4402
4403 update_peer_seq(mdev, be32_to_cpu(p->seq_num));
4404
4405 if (be64_to_cpu(p->block_id) == ID_OUT_OF_SYNC)
4406 drbd_ov_oos_found(mdev, sector, size);
4407 else
4408 ov_oos_print(mdev);
4409
Lars Ellenberg1d53f092010-09-05 01:13:24 +02004410 if (!get_ldev(mdev))
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004411 return true;
Lars Ellenberg1d53f092010-09-05 01:13:24 +02004412
Philipp Reisnerb411b362009-09-25 16:07:19 -07004413 drbd_rs_complete_io(mdev, sector);
4414 dec_rs_pending(mdev);
4415
Lars Ellenbergea5442a2010-11-05 09:48:01 +01004416 --mdev->ov_left;
4417
4418 /* let's advance progress step marks only for every other megabyte */
4419 if ((mdev->ov_left & 0x200) == 0x200)
4420 drbd_advance_rs_marks(mdev, mdev->ov_left);
4421
4422 if (mdev->ov_left == 0) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004423 w = kmalloc(sizeof(*w), GFP_NOIO);
4424 if (w) {
4425 w->cb = w_ov_finished;
4426 drbd_queue_work_front(&mdev->data.work, w);
4427 } else {
4428 dev_err(DEV, "kmalloc(w) failed.");
4429 ov_oos_print(mdev);
4430 drbd_resync_finished(mdev);
4431 }
4432 }
Lars Ellenberg1d53f092010-09-05 01:13:24 +02004433 put_ldev(mdev);
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004434 return true;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004435}
4436
Philipp Reisner02918be2010-08-20 14:35:10 +02004437static int got_skip(struct drbd_conf *mdev, struct p_header80 *h)
Philipp Reisner0ced55a2010-04-30 15:26:20 +02004438{
Andreas Gruenbacher81e84652010-12-09 15:03:57 +01004439 return true;
Philipp Reisner0ced55a2010-04-30 15:26:20 +02004440}
4441
Philipp Reisnerb411b362009-09-25 16:07:19 -07004442struct asender_cmd {
4443 size_t pkt_size;
Philipp Reisner0b70a132010-08-20 13:36:10 +02004444 int (*process)(struct drbd_conf *mdev, struct p_header80 *h);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004445};
4446
4447static struct asender_cmd *get_asender_cmd(int cmd)
4448{
4449 static struct asender_cmd asender_tbl[] = {
4450 /* anything missing from this table is in
4451 * the drbd_cmd_handler (drbd_default_handler) table,
4452 * see the beginning of drbdd() */
Philipp Reisner0b70a132010-08-20 13:36:10 +02004453 [P_PING] = { sizeof(struct p_header80), got_Ping },
4454 [P_PING_ACK] = { sizeof(struct p_header80), got_PingAck },
Philipp Reisnerb411b362009-09-25 16:07:19 -07004455 [P_RECV_ACK] = { sizeof(struct p_block_ack), got_BlockAck },
4456 [P_WRITE_ACK] = { sizeof(struct p_block_ack), got_BlockAck },
4457 [P_RS_WRITE_ACK] = { sizeof(struct p_block_ack), got_BlockAck },
4458 [P_DISCARD_ACK] = { sizeof(struct p_block_ack), got_BlockAck },
4459 [P_NEG_ACK] = { sizeof(struct p_block_ack), got_NegAck },
4460 [P_NEG_DREPLY] = { sizeof(struct p_block_ack), got_NegDReply },
4461 [P_NEG_RS_DREPLY] = { sizeof(struct p_block_ack), got_NegRSDReply},
4462 [P_OV_RESULT] = { sizeof(struct p_block_ack), got_OVResult },
4463 [P_BARRIER_ACK] = { sizeof(struct p_barrier_ack), got_BarrierAck },
4464 [P_STATE_CHG_REPLY] = { sizeof(struct p_req_state_reply), got_RqSReply },
4465 [P_RS_IS_IN_SYNC] = { sizeof(struct p_block_ack), got_IsInSync },
Philipp Reisner02918be2010-08-20 14:35:10 +02004466 [P_DELAY_PROBE] = { sizeof(struct p_delay_probe93), got_skip },
Philipp Reisnerd612d302010-12-27 10:53:28 +01004467 [P_RS_CANCEL] = { sizeof(struct p_block_ack), got_NegRSDReply},
Philipp Reisnerb411b362009-09-25 16:07:19 -07004468 [P_MAX_CMD] = { 0, NULL },
4469 };
4470 if (cmd > P_MAX_CMD || asender_tbl[cmd].process == NULL)
4471 return NULL;
4472 return &asender_tbl[cmd];
4473}
4474
4475int drbd_asender(struct drbd_thread *thi)
4476{
4477 struct drbd_conf *mdev = thi->mdev;
Philipp Reisner02918be2010-08-20 14:35:10 +02004478 struct p_header80 *h = &mdev->meta.rbuf.header.h80;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004479 struct asender_cmd *cmd = NULL;
4480
4481 int rv, len;
4482 void *buf = h;
4483 int received = 0;
Philipp Reisner0b70a132010-08-20 13:36:10 +02004484 int expect = sizeof(struct p_header80);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004485 int empty;
Lars Ellenbergf36af182011-03-09 22:44:55 +01004486 int ping_timeout_active = 0;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004487
4488 sprintf(current->comm, "drbd%d_asender", mdev_to_minor(mdev));
4489
4490 current->policy = SCHED_RR; /* Make this a realtime task! */
4491 current->rt_priority = 2; /* more important than all other tasks */
4492
Andreas Gruenbachere77a0a52011-01-25 15:43:39 +01004493 while (get_t_state(thi) == RUNNING) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004494 drbd_thread_current_set_cpu(mdev);
4495 if (test_and_clear_bit(SEND_PING, &mdev->flags)) {
4496 ERR_IF(!drbd_send_ping(mdev)) goto reconnect;
4497 mdev->meta.socket->sk->sk_rcvtimeo =
4498 mdev->net_conf->ping_timeo*HZ/10;
Lars Ellenbergf36af182011-03-09 22:44:55 +01004499 ping_timeout_active = 1;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004500 }
4501
4502 /* conditionally cork;
4503 * it may hurt latency if we cork without much to send */
4504 if (!mdev->net_conf->no_cork &&
4505 3 < atomic_read(&mdev->unacked_cnt))
4506 drbd_tcp_cork(mdev->meta.socket);
4507 while (1) {
4508 clear_bit(SIGNAL_ASENDER, &mdev->flags);
4509 flush_signals(current);
Lars Ellenberg0f8488e2010-10-13 18:19:23 +02004510 if (!drbd_process_done_ee(mdev))
Philipp Reisnerb411b362009-09-25 16:07:19 -07004511 goto reconnect;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004512 /* to avoid race with newly queued ACKs */
4513 set_bit(SIGNAL_ASENDER, &mdev->flags);
4514 spin_lock_irq(&mdev->req_lock);
4515 empty = list_empty(&mdev->done_ee);
4516 spin_unlock_irq(&mdev->req_lock);
4517 /* new ack may have been queued right here,
4518 * but then there is also a signal pending,
4519 * and we start over... */
4520 if (empty)
4521 break;
4522 }
4523 /* but unconditionally uncork unless disabled */
4524 if (!mdev->net_conf->no_cork)
4525 drbd_tcp_uncork(mdev->meta.socket);
4526
4527 /* short circuit, recv_msg would return EINTR anyways. */
4528 if (signal_pending(current))
4529 continue;
4530
4531 rv = drbd_recv_short(mdev, mdev->meta.socket,
4532 buf, expect-received, 0);
4533 clear_bit(SIGNAL_ASENDER, &mdev->flags);
4534
4535 flush_signals(current);
4536
4537 /* Note:
4538 * -EINTR (on meta) we got a signal
4539 * -EAGAIN (on meta) rcvtimeo expired
4540 * -ECONNRESET other side closed the connection
4541 * -ERESTARTSYS (on data) we got a signal
4542 * rv < 0 other than above: unexpected error!
4543 * rv == expected: full header or command
4544 * rv < expected: "woken" by signal during receive
4545 * rv == 0 : "connection shut down by peer"
4546 */
4547 if (likely(rv > 0)) {
4548 received += rv;
4549 buf += rv;
4550 } else if (rv == 0) {
4551 dev_err(DEV, "meta connection shut down by peer.\n");
4552 goto reconnect;
4553 } else if (rv == -EAGAIN) {
Lars Ellenbergcb6518c2011-06-20 14:44:45 +02004554 /* If the data socket received something meanwhile,
4555 * that is good enough: peer is still alive. */
4556 if (time_after(mdev->last_received,
4557 jiffies - mdev->meta.socket->sk->sk_rcvtimeo))
4558 continue;
Lars Ellenbergf36af182011-03-09 22:44:55 +01004559 if (ping_timeout_active) {
Philipp Reisnerb411b362009-09-25 16:07:19 -07004560 dev_err(DEV, "PingAck did not arrive in time.\n");
4561 goto reconnect;
4562 }
4563 set_bit(SEND_PING, &mdev->flags);
4564 continue;
4565 } else if (rv == -EINTR) {
4566 continue;
4567 } else {
4568 dev_err(DEV, "sock_recvmsg returned %d\n", rv);
4569 goto reconnect;
4570 }
4571
4572 if (received == expect && cmd == NULL) {
Andreas Gruenbacherca9bc122011-01-11 13:47:24 +01004573 if (unlikely(h->magic != cpu_to_be32(DRBD_MAGIC))) {
Lars Ellenberg004352f2010-10-05 20:13:58 +02004574 dev_err(DEV, "magic?? on meta m: 0x%08x c: %d l: %d\n",
4575 be32_to_cpu(h->magic),
4576 be16_to_cpu(h->command),
4577 be16_to_cpu(h->length));
Philipp Reisnerb411b362009-09-25 16:07:19 -07004578 goto reconnect;
4579 }
4580 cmd = get_asender_cmd(be16_to_cpu(h->command));
4581 len = be16_to_cpu(h->length);
4582 if (unlikely(cmd == NULL)) {
Lars Ellenberg004352f2010-10-05 20:13:58 +02004583 dev_err(DEV, "unknown command?? on meta m: 0x%08x c: %d l: %d\n",
4584 be32_to_cpu(h->magic),
4585 be16_to_cpu(h->command),
4586 be16_to_cpu(h->length));
Philipp Reisnerb411b362009-09-25 16:07:19 -07004587 goto disconnect;
4588 }
4589 expect = cmd->pkt_size;
Philipp Reisner0b70a132010-08-20 13:36:10 +02004590 ERR_IF(len != expect-sizeof(struct p_header80))
Philipp Reisnerb411b362009-09-25 16:07:19 -07004591 goto reconnect;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004592 }
4593 if (received == expect) {
Lars Ellenbergcb6518c2011-06-20 14:44:45 +02004594 mdev->last_received = jiffies;
Philipp Reisnerb411b362009-09-25 16:07:19 -07004595 D_ASSERT(cmd != NULL);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004596 if (!cmd->process(mdev, h))
4597 goto reconnect;
4598
Lars Ellenbergf36af182011-03-09 22:44:55 +01004599 /* the idle_timeout (ping-int)
4600 * has been restored in got_PingAck() */
4601 if (cmd == get_asender_cmd(P_PING_ACK))
4602 ping_timeout_active = 0;
4603
Philipp Reisnerb411b362009-09-25 16:07:19 -07004604 buf = h;
4605 received = 0;
Philipp Reisner0b70a132010-08-20 13:36:10 +02004606 expect = sizeof(struct p_header80);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004607 cmd = NULL;
4608 }
4609 }
4610
4611 if (0) {
4612reconnect:
4613 drbd_force_state(mdev, NS(conn, C_NETWORK_FAILURE));
Lars Ellenberg856c50c2010-10-14 13:37:40 +02004614 drbd_md_sync(mdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004615 }
4616 if (0) {
4617disconnect:
4618 drbd_force_state(mdev, NS(conn, C_DISCONNECTING));
Lars Ellenberg856c50c2010-10-14 13:37:40 +02004619 drbd_md_sync(mdev);
Philipp Reisnerb411b362009-09-25 16:07:19 -07004620 }
4621 clear_bit(SIGNAL_ASENDER, &mdev->flags);
4622
4623 D_ASSERT(mdev->state.conn < C_CONNECTED);
4624 dev_info(DEV, "asender terminated\n");
4625
4626 return 0;
4627}