drbd_main.c 83.4 KB
Newer Older
P
Philipp Reisner 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
/*
   drbd.c

   This file is part of DRBD by Philipp Reisner and Lars Ellenberg.

   Copyright (C) 2001-2008, LINBIT Information Technologies GmbH.
   Copyright (C) 1999-2008, Philipp Reisner <philipp.reisner@linbit.com>.
   Copyright (C) 2002-2008, Lars Ellenberg <lars.ellenberg@linbit.com>.

   Thanks to Carter Burden, Bart Grantham and Gennadiy Nerubayev
   from Logicworks, Inc. for making SDP replication support possible.

   drbd is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2, or (at your option)
   any later version.

   drbd is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with drbd; see the file COPYING.  If not, write to
   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.

 */

#include <linux/module.h>
#include <linux/drbd.h>
#include <asm/uaccess.h>
#include <asm/types.h>
#include <net/sock.h>
#include <linux/ctype.h>
35
#include <linux/mutex.h>
P
Philipp Reisner 已提交
36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/proc_fs.h>
#include <linux/init.h>
#include <linux/mm.h>
#include <linux/memcontrol.h>
#include <linux/mm_inline.h>
#include <linux/slab.h>
#include <linux/random.h>
#include <linux/reboot.h>
#include <linux/notifier.h>
#include <linux/kthread.h>

#define __KERNEL_SYSCALLS__
#include <linux/unistd.h>
#include <linux/vmalloc.h>

#include <linux/drbd_limits.h>
#include "drbd_int.h"
#include "drbd_req.h" /* only for _req_mod in tl_release and tl_clear */

#include "drbd_vli.h"

59
static DEFINE_MUTEX(drbd_main_mutex);
P
Philipp Reisner 已提交
60 61 62 63 64 65 66 67 68 69
int drbdd_init(struct drbd_thread *);
int drbd_worker(struct drbd_thread *);
int drbd_asender(struct drbd_thread *);

int drbd_init(void);
static int drbd_open(struct block_device *bdev, fmode_t mode);
static int drbd_release(struct gendisk *gd, fmode_t mode);
static int w_md_sync(struct drbd_conf *mdev, struct drbd_work *w, int unused);
static void md_sync_timer_fn(unsigned long data);
static int w_bitmap_io(struct drbd_conf *mdev, struct drbd_work *w, int unused);
70
static int w_go_diskless(struct drbd_conf *mdev, struct drbd_work *w, int unused);
P
Philipp Reisner 已提交
71 72 73 74 75 76

MODULE_AUTHOR("Philipp Reisner <phil@linbit.com>, "
	      "Lars Ellenberg <lars@linbit.com>");
MODULE_DESCRIPTION("drbd - Distributed Replicated Block Device v" REL_VERSION);
MODULE_VERSION(REL_VERSION);
MODULE_LICENSE("GPL");
77 78
MODULE_PARM_DESC(minor_count, "Maximum number of drbd devices ("
		 __stringify(DRBD_MINOR_COUNT_MIN) "-" __stringify(DRBD_MINOR_COUNT_MAX) ")");
P
Philipp Reisner 已提交
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107
MODULE_ALIAS_BLOCKDEV_MAJOR(DRBD_MAJOR);

#include <linux/moduleparam.h>
/* allow_open_on_secondary */
MODULE_PARM_DESC(allow_oos, "DONT USE!");
/* thanks to these macros, if compiled into the kernel (not-module),
 * this becomes the boot parameter drbd.minor_count */
module_param(minor_count, uint, 0444);
module_param(disable_sendpage, bool, 0644);
module_param(allow_oos, bool, 0);
module_param(cn_idx, uint, 0444);
module_param(proc_details, int, 0644);

#ifdef CONFIG_DRBD_FAULT_INJECTION
int enable_faults;
int fault_rate;
static int fault_count;
int fault_devs;
/* bitmap of enabled faults */
module_param(enable_faults, int, 0664);
/* fault rate % value - applies to all enabled faults */
module_param(fault_rate, int, 0664);
/* count of faults inserted */
module_param(fault_count, int, 0664);
/* bitmap of devices to insert faults on */
module_param(fault_devs, int, 0644);
#endif

/* module parameter, defined */
108
unsigned int minor_count = DRBD_MINOR_COUNT_DEF;
P
Philipp Reisner 已提交
109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
int disable_sendpage;
int allow_oos;
unsigned int cn_idx = CN_IDX_DRBD;
int proc_details;       /* Detail level in proc drbd*/

/* Module parameter for setting the user mode helper program
 * to run. Default is /sbin/drbdadm */
char usermode_helper[80] = "/sbin/drbdadm";

module_param_string(usermode_helper, usermode_helper, sizeof(usermode_helper), 0644);

/* in 2.6.x, our device mapping and config info contains our virtual gendisks
 * as member "struct gendisk *vdisk;"
 */
struct drbd_conf **minor_table;
124
struct list_head drbd_tconns;  /* list of struct drbd_tconn */
P
Philipp Reisner 已提交
125 126

struct kmem_cache *drbd_request_cache;
A
Andreas Gruenbacher 已提交
127
struct kmem_cache *drbd_ee_cache;	/* peer requests */
P
Philipp Reisner 已提交
128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
struct kmem_cache *drbd_bm_ext_cache;	/* bitmap extents */
struct kmem_cache *drbd_al_ext_cache;	/* activity log extents */
mempool_t *drbd_request_mempool;
mempool_t *drbd_ee_mempool;

/* I do not use a standard mempool, because:
   1) I want to hand out the pre-allocated objects first.
   2) I want to be able to interrupt sleeping allocation with a signal.
   Note: This is a single linked list, the next pointer is the private
	 member of struct page.
 */
struct page *drbd_pp_pool;
spinlock_t   drbd_pp_lock;
int          drbd_pp_vacant;
wait_queue_head_t drbd_pp_wait;

DEFINE_RATELIMIT_STATE(drbd_ratelimit_state, 5 * HZ, 5);

146
static const struct block_device_operations drbd_ops = {
P
Philipp Reisner 已提交
147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
	.owner =   THIS_MODULE,
	.open =    drbd_open,
	.release = drbd_release,
};

#define ARRY_SIZE(A) (sizeof(A)/sizeof(A[0]))

#ifdef __CHECKER__
/* When checking with sparse, and this is an inline function, sparse will
   give tons of false positives. When this is a real functions sparse works.
 */
int _get_ldev_if_state(struct drbd_conf *mdev, enum drbd_disk_state mins)
{
	int io_allowed;

	atomic_inc(&mdev->local_cnt);
	io_allowed = (mdev->state.disk >= mins);
	if (!io_allowed) {
		if (atomic_dec_and_test(&mdev->local_cnt))
			wake_up(&mdev->misc_wait);
	}
	return io_allowed;
}

#endif

/**
 * DOC: The transfer log
 *
 * The transfer log is a single linked list of &struct drbd_tl_epoch objects.
177
 * mdev->tconn->newest_tle points to the head, mdev->tconn->oldest_tle points to the tail
P
Philipp Reisner 已提交
178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194
 * of the list. There is always at least one &struct drbd_tl_epoch object.
 *
 * Each &struct drbd_tl_epoch has a circular double linked list of requests
 * attached.
 */
static int tl_init(struct drbd_conf *mdev)
{
	struct drbd_tl_epoch *b;

	/* during device minor initialization, we may well use GFP_KERNEL */
	b = kmalloc(sizeof(struct drbd_tl_epoch), GFP_KERNEL);
	if (!b)
		return 0;
	INIT_LIST_HEAD(&b->requests);
	INIT_LIST_HEAD(&b->w.list);
	b->next = NULL;
	b->br_number = 4711;
195
	b->n_writes = 0;
P
Philipp Reisner 已提交
196 197
	b->w.cb = NULL; /* if this is != NULL, we need to dec_ap_pending in tl_clear */

198 199 200
	mdev->tconn->oldest_tle = b;
	mdev->tconn->newest_tle = b;
	INIT_LIST_HEAD(&mdev->tconn->out_of_sequence_requests);
P
Philipp Reisner 已提交
201 202 203 204 205 206

	return 1;
}

static void tl_cleanup(struct drbd_conf *mdev)
{
207 208 209 210 211 212
	D_ASSERT(mdev->tconn->oldest_tle == mdev->tconn->newest_tle);
	D_ASSERT(list_empty(&mdev->tconn->out_of_sequence_requests));
	kfree(mdev->tconn->oldest_tle);
	mdev->tconn->oldest_tle = NULL;
	kfree(mdev->tconn->unused_spare_tle);
	mdev->tconn->unused_spare_tle = NULL;
213 214
}

P
Philipp Reisner 已提交
215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
/**
 * _tl_add_barrier() - Adds a barrier to the transfer log
 * @mdev:	DRBD device.
 * @new:	Barrier to be added before the current head of the TL.
 *
 * The caller must hold the req_lock.
 */
void _tl_add_barrier(struct drbd_conf *mdev, struct drbd_tl_epoch *new)
{
	struct drbd_tl_epoch *newest_before;

	INIT_LIST_HEAD(&new->requests);
	INIT_LIST_HEAD(&new->w.list);
	new->w.cb = NULL; /* if this is != NULL, we need to dec_ap_pending in tl_clear */
	new->next = NULL;
230
	new->n_writes = 0;
P
Philipp Reisner 已提交
231

232
	newest_before = mdev->tconn->newest_tle;
P
Philipp Reisner 已提交
233 234 235
	/* never send a barrier number == 0, because that is special-cased
	 * when using TCQ for our write ordering code */
	new->br_number = (newest_before->br_number+1) ?: 1;
236 237 238
	if (mdev->tconn->newest_tle != new) {
		mdev->tconn->newest_tle->next = new;
		mdev->tconn->newest_tle = new;
P
Philipp Reisner 已提交
239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258
	}
}

/**
 * tl_release() - Free or recycle the oldest &struct drbd_tl_epoch object of the TL
 * @mdev:	DRBD device.
 * @barrier_nr:	Expected identifier of the DRBD write barrier packet.
 * @set_size:	Expected number of requests before that barrier.
 *
 * In case the passed barrier_nr or set_size does not match the oldest
 * &struct drbd_tl_epoch objects this function will cause a termination
 * of the connection.
 */
void tl_release(struct drbd_conf *mdev, unsigned int barrier_nr,
		       unsigned int set_size)
{
	struct drbd_tl_epoch *b, *nob; /* next old barrier */
	struct list_head *le, *tle;
	struct drbd_request *r;

259
	spin_lock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
260

261
	b = mdev->tconn->oldest_tle;
P
Philipp Reisner 已提交
262 263 264 265 266 267 268 269 270 271 272 273

	/* first some paranoia code */
	if (b == NULL) {
		dev_err(DEV, "BAD! BarrierAck #%u received, but no epoch in tl!?\n",
			barrier_nr);
		goto bail;
	}
	if (b->br_number != barrier_nr) {
		dev_err(DEV, "BAD! BarrierAck #%u received, expected #%u!\n",
			barrier_nr, b->br_number);
		goto bail;
	}
274 275 276
	if (b->n_writes != set_size) {
		dev_err(DEV, "BAD! BarrierAck #%u received with n_writes=%u, expected n_writes=%u!\n",
			barrier_nr, set_size, b->n_writes);
P
Philipp Reisner 已提交
277 278 279 280 281 282
		goto bail;
	}

	/* Clean up list of requests processed during current epoch */
	list_for_each_safe(le, tle, &b->requests) {
		r = list_entry(le, struct drbd_request, tl_requests);
283
		_req_mod(r, BARRIER_ACKED);
P
Philipp Reisner 已提交
284 285 286 287 288 289 290 291 292
	}
	/* There could be requests on the list waiting for completion
	   of the write to the local disk. To avoid corruptions of
	   slab's data structures we have to remove the lists head.

	   Also there could have been a barrier ack out of sequence, overtaking
	   the write acks - which would be a bug and violating write ordering.
	   To not deadlock in case we lose connection while such requests are
	   still pending, we need some way to find them for the
293
	   _req_mode(CONNECTION_LOST_WHILE_PENDING).
P
Philipp Reisner 已提交
294 295

	   These have been list_move'd to the out_of_sequence_requests list in
296
	   _req_mod(, BARRIER_ACKED) above.
P
Philipp Reisner 已提交
297 298 299 300 301 302 303
	   */
	list_del_init(&b->requests);

	nob = b->next;
	if (test_and_clear_bit(CREATE_BARRIER, &mdev->flags)) {
		_tl_add_barrier(mdev, b);
		if (nob)
304
			mdev->tconn->oldest_tle = nob;
P
Philipp Reisner 已提交
305
		/* if nob == NULL b was the only barrier, and becomes the new
306
		   barrier. Therefore mdev->tconn->oldest_tle points already to b */
P
Philipp Reisner 已提交
307 308
	} else {
		D_ASSERT(nob != NULL);
309
		mdev->tconn->oldest_tle = nob;
P
Philipp Reisner 已提交
310 311 312
		kfree(b);
	}

313
	spin_unlock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
314 315 316 317 318
	dec_ap_pending(mdev);

	return;

bail:
319
	spin_unlock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
320 321 322
	drbd_force_state(mdev, NS(conn, C_PROTOCOL_ERROR));
}

323

P
Philipp Reisner 已提交
324
/**
325
 * _tl_restart() - Walks the transfer log, and applies an action to all requests
P
Philipp Reisner 已提交
326
 * @mdev:	DRBD device.
327
 * @what:       The action/event to perform with all request objects
P
Philipp Reisner 已提交
328
 *
329 330
 * @what might be one of CONNECTION_LOST_WHILE_PENDING, RESEND, FAIL_FROZEN_DISK_IO,
 * RESTART_FROZEN_DISK_IO.
P
Philipp Reisner 已提交
331
 */
332
void _tl_restart(struct drbd_conf *mdev, enum drbd_req_event what)
P
Philipp Reisner 已提交
333
{
334
	struct drbd_tl_epoch *b, *tmp, **pn;
335
	struct list_head *le, *tle, carry_reads;
336 337
	struct drbd_request *req;
	int rv, n_writes, n_reads;
P
Philipp Reisner 已提交
338

339 340
	b = mdev->tconn->oldest_tle;
	pn = &mdev->tconn->oldest_tle;
P
Philipp Reisner 已提交
341
	while (b) {
342 343
		n_writes = 0;
		n_reads = 0;
344
		INIT_LIST_HEAD(&carry_reads);
P
Philipp Reisner 已提交
345
		list_for_each_safe(le, tle, &b->requests) {
346 347 348 349 350
			req = list_entry(le, struct drbd_request, tl_requests);
			rv = _req_mod(req, what);

			n_writes += (rv & MR_WRITE) >> MR_WRITE_SHIFT;
			n_reads  += (rv & MR_READ) >> MR_READ_SHIFT;
P
Philipp Reisner 已提交
351 352 353
		}
		tmp = b->next;

354
		if (n_writes) {
355
			if (what == RESEND) {
356 357 358 359 360 361 362
				b->n_writes = n_writes;
				if (b->w.cb == NULL) {
					b->w.cb = w_send_barrier;
					inc_ap_pending(mdev);
					set_bit(CREATE_BARRIER, &mdev->flags);
				}

363
				drbd_queue_work(&mdev->tconn->data.work, &b->w);
364 365 366
			}
			pn = &b->next;
		} else {
367 368
			if (n_reads)
				list_add(&carry_reads, &b->requests);
369 370 371 372 373 374 375 376 377 378
			/* there could still be requests on that ring list,
			 * in case local io is still pending */
			list_del(&b->requests);

			/* dec_ap_pending corresponding to queue_barrier.
			 * the newest barrier may not have been queued yet,
			 * in which case w.cb is still NULL. */
			if (b->w.cb != NULL)
				dec_ap_pending(mdev);

379
			if (b == mdev->tconn->newest_tle) {
380 381 382
				/* recycle, but reinit! */
				D_ASSERT(tmp == NULL);
				INIT_LIST_HEAD(&b->requests);
383
				list_splice(&carry_reads, &b->requests);
384 385 386 387 388 389 390 391 392 393
				INIT_LIST_HEAD(&b->w.list);
				b->w.cb = NULL;
				b->br_number = net_random();
				b->n_writes = 0;

				*pn = b;
				break;
			}
			*pn = tmp;
			kfree(b);
P
Philipp Reisner 已提交
394 395
		}
		b = tmp;
396
		list_splice(&carry_reads, &b->requests);
P
Philipp Reisner 已提交
397
	}
398 399
}

P
Philipp Reisner 已提交
400 401 402 403 404 405 406 407 408 409 410 411 412 413

/**
 * tl_clear() - Clears all requests and &struct drbd_tl_epoch objects out of the TL
 * @mdev:	DRBD device.
 *
 * This is called after the connection to the peer was lost. The storage covered
 * by the requests on the transfer gets marked as our of sync. Called from the
 * receiver thread and the worker thread.
 */
void tl_clear(struct drbd_conf *mdev)
{
	struct list_head *le, *tle;
	struct drbd_request *r;

414
	spin_lock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
415

416
	_tl_restart(mdev, CONNECTION_LOST_WHILE_PENDING);
P
Philipp Reisner 已提交
417 418

	/* we expect this list to be empty. */
419
	D_ASSERT(list_empty(&mdev->tconn->out_of_sequence_requests));
P
Philipp Reisner 已提交
420 421

	/* but just in case, clean it up anyways! */
422
	list_for_each_safe(le, tle, &mdev->tconn->out_of_sequence_requests) {
P
Philipp Reisner 已提交
423 424 425
		r = list_entry(le, struct drbd_request, tl_requests);
		/* It would be nice to complete outside of spinlock.
		 * But this is easier for now. */
426
		_req_mod(r, CONNECTION_LOST_WHILE_PENDING);
P
Philipp Reisner 已提交
427 428 429 430 431
	}

	/* ensure bit indicating barrier is required is clear */
	clear_bit(CREATE_BARRIER, &mdev->flags);

432
	spin_unlock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
433 434
}

435 436
void tl_restart(struct drbd_conf *mdev, enum drbd_req_event what)
{
437
	spin_lock_irq(&mdev->tconn->req_lock);
438
	_tl_restart(mdev, what);
439
	spin_unlock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
440 441 442 443 444 445 446 447 448 449 450 451 452 453
}

static int drbd_thread_setup(void *arg)
{
	struct drbd_thread *thi = (struct drbd_thread *) arg;
	struct drbd_conf *mdev = thi->mdev;
	unsigned long flags;
	int retval;

restart:
	retval = thi->function(thi);

	spin_lock_irqsave(&thi->t_lock, flags);

454
	/* if the receiver has been "EXITING", the last thing it did
P
Philipp Reisner 已提交
455 456 457
	 * was set the conn state to "StandAlone",
	 * if now a re-connect request comes in, conn state goes C_UNCONNECTED,
	 * and receiver thread will be "started".
458
	 * drbd_thread_start needs to set "RESTARTING" in that case.
P
Philipp Reisner 已提交
459
	 * t_state check and assignment needs to be within the same spinlock,
460 461
	 * so either thread_start sees EXITING, and can remap to RESTARTING,
	 * or thread_start see NONE, and can proceed as normal.
P
Philipp Reisner 已提交
462 463
	 */

464
	if (thi->t_state == RESTARTING) {
465
		dev_info(DEV, "Restarting %s thread\n", thi->name);
466
		thi->t_state = RUNNING;
P
Philipp Reisner 已提交
467 468 469 470 471
		spin_unlock_irqrestore(&thi->t_lock, flags);
		goto restart;
	}

	thi->task = NULL;
472
	thi->t_state = NONE;
P
Philipp Reisner 已提交
473 474 475 476 477 478 479 480 481 482 483 484
	smp_mb();
	complete(&thi->stop);
	spin_unlock_irqrestore(&thi->t_lock, flags);

	dev_info(DEV, "Terminating %s\n", current->comm);

	/* Release mod reference taken when thread was started */
	module_put(THIS_MODULE);
	return retval;
}

static void drbd_thread_init(struct drbd_conf *mdev, struct drbd_thread *thi,
485
			     int (*func) (struct drbd_thread *), char *name)
P
Philipp Reisner 已提交
486 487 488
{
	spin_lock_init(&thi->t_lock);
	thi->task    = NULL;
489
	thi->t_state = NONE;
P
Philipp Reisner 已提交
490 491
	thi->function = func;
	thi->mdev = mdev;
492
	strncpy(thi->name, name, ARRAY_SIZE(thi->name));
P
Philipp Reisner 已提交
493 494 495 496 497 498 499 500 501 502 503 504 505
}

int drbd_thread_start(struct drbd_thread *thi)
{
	struct drbd_conf *mdev = thi->mdev;
	struct task_struct *nt;
	unsigned long flags;

	/* is used from state engine doing drbd_thread_stop_nowait,
	 * while holding the req lock irqsave */
	spin_lock_irqsave(&thi->t_lock, flags);

	switch (thi->t_state) {
506
	case NONE:
P
Philipp Reisner 已提交
507
		dev_info(DEV, "Starting %s thread (from %s [%d])\n",
508
			 thi->name, current->comm, current->pid);
P
Philipp Reisner 已提交
509 510 511 512 513

		/* Get ref on module for thread - this is released when thread exits */
		if (!try_module_get(THIS_MODULE)) {
			dev_err(DEV, "Failed to get module reference in drbd_thread_start\n");
			spin_unlock_irqrestore(&thi->t_lock, flags);
514
			return false;
P
Philipp Reisner 已提交
515 516 517 518 519
		}

		init_completion(&thi->stop);
		D_ASSERT(thi->task == NULL);
		thi->reset_cpu_mask = 1;
520
		thi->t_state = RUNNING;
P
Philipp Reisner 已提交
521 522 523 524
		spin_unlock_irqrestore(&thi->t_lock, flags);
		flush_signals(current); /* otherw. may get -ERESTARTNOINTR */

		nt = kthread_create(drbd_thread_setup, (void *) thi,
525
				    "drbd%d_%s", mdev_to_minor(mdev), thi->name);
P
Philipp Reisner 已提交
526 527 528 529 530

		if (IS_ERR(nt)) {
			dev_err(DEV, "Couldn't start thread\n");

			module_put(THIS_MODULE);
531
			return false;
P
Philipp Reisner 已提交
532 533 534
		}
		spin_lock_irqsave(&thi->t_lock, flags);
		thi->task = nt;
535
		thi->t_state = RUNNING;
P
Philipp Reisner 已提交
536 537 538
		spin_unlock_irqrestore(&thi->t_lock, flags);
		wake_up_process(nt);
		break;
539 540
	case EXITING:
		thi->t_state = RESTARTING;
P
Philipp Reisner 已提交
541
		dev_info(DEV, "Restarting %s thread (from %s [%d])\n",
542
				thi->name, current->comm, current->pid);
P
Philipp Reisner 已提交
543
		/* fall through */
544 545
	case RUNNING:
	case RESTARTING:
P
Philipp Reisner 已提交
546 547 548 549 550
	default:
		spin_unlock_irqrestore(&thi->t_lock, flags);
		break;
	}

551
	return true;
P
Philipp Reisner 已提交
552 553 554 555 556 557 558
}


void _drbd_thread_stop(struct drbd_thread *thi, int restart, int wait)
{
	unsigned long flags;

559
	enum drbd_thread_state ns = restart ? RESTARTING : EXITING;
P
Philipp Reisner 已提交
560 561 562 563

	/* may be called from state engine, holding the req lock irqsave */
	spin_lock_irqsave(&thi->t_lock, flags);

564
	if (thi->t_state == NONE) {
P
Philipp Reisner 已提交
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
		spin_unlock_irqrestore(&thi->t_lock, flags);
		if (restart)
			drbd_thread_start(thi);
		return;
	}

	if (thi->t_state != ns) {
		if (thi->task == NULL) {
			spin_unlock_irqrestore(&thi->t_lock, flags);
			return;
		}

		thi->t_state = ns;
		smp_mb();
		init_completion(&thi->stop);
		if (thi->task != current)
			force_sig(DRBD_SIGKILL, thi->task);

	}

	spin_unlock_irqrestore(&thi->t_lock, flags);

	if (wait)
		wait_for_completion(&thi->stop);
}

591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607
static struct drbd_thread *drbd_task_to_thread(struct drbd_conf *mdev, struct task_struct *task)
{
	struct drbd_tconn *tconn = mdev->tconn;
	struct drbd_thread *thi =
		task == tconn->receiver.task ? &tconn->receiver :
		task == tconn->asender.task  ? &tconn->asender :
		task == tconn->worker.task   ? &tconn->worker : NULL;

	return thi;
}

char *drbd_task_to_thread_name(struct drbd_conf *mdev, struct task_struct *task)
{
	struct drbd_thread *thi = drbd_task_to_thread(mdev, task);
	return thi ? thi->name : task->comm;
}

P
Philipp Reisner 已提交
608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637
#ifdef CONFIG_SMP
/**
 * drbd_calc_cpu_mask() - Generate CPU masks, spread over all CPUs
 * @mdev:	DRBD device.
 *
 * Forces all threads of a device onto the same CPU. This is beneficial for
 * DRBD's performance. May be overwritten by user's configuration.
 */
void drbd_calc_cpu_mask(struct drbd_conf *mdev)
{
	int ord, cpu;

	/* user override. */
	if (cpumask_weight(mdev->cpu_mask))
		return;

	ord = mdev_to_minor(mdev) % cpumask_weight(cpu_online_mask);
	for_each_online_cpu(cpu) {
		if (ord-- == 0) {
			cpumask_set_cpu(cpu, mdev->cpu_mask);
			return;
		}
	}
	/* should not be reached */
	cpumask_setall(mdev->cpu_mask);
}

/**
 * drbd_thread_current_set_cpu() - modifies the cpu mask of the _current_ thread
 * @mdev:	DRBD device.
638
 * @thi:	drbd_thread object
P
Philipp Reisner 已提交
639 640 641 642
 *
 * call in the "main loop" of _all_ threads, no need for any mutex, current won't die
 * prematurely.
 */
643
void drbd_thread_current_set_cpu(struct drbd_conf *mdev, struct drbd_thread *thi)
P
Philipp Reisner 已提交
644 645
{
	struct task_struct *p = current;
646

P
Philipp Reisner 已提交
647 648 649 650 651 652 653
	if (!thi->reset_cpu_mask)
		return;
	thi->reset_cpu_mask = 0;
	set_cpus_allowed_ptr(p, mdev->cpu_mask);
}
#endif

P
Philipp Reisner 已提交
654
static void prepare_header80(struct drbd_conf *mdev, struct p_header80 *h,
655
			     enum drbd_packet cmd, int size)
P
Philipp Reisner 已提交
656 657 658 659 660 661 662
{
	h->magic   = cpu_to_be32(DRBD_MAGIC);
	h->command = cpu_to_be16(cmd);
	h->length  = cpu_to_be16(size);
}

static void prepare_header95(struct drbd_conf *mdev, struct p_header95 *h,
663
			     enum drbd_packet cmd, int size)
P
Philipp Reisner 已提交
664 665 666 667 668 669 670
{
	h->magic   = cpu_to_be16(DRBD_MAGIC_BIG);
	h->command = cpu_to_be16(cmd);
	h->length  = cpu_to_be32(size);
}

static void prepare_header(struct drbd_conf *mdev, struct p_header *h,
671
			   enum drbd_packet cmd, int size)
P
Philipp Reisner 已提交
672 673 674 675 676 677 678
{
	if (mdev->tconn->agreed_pro_version >= 100 || size > DRBD_MAX_SIZE_H80_PACKET)
		prepare_header95(mdev, &h->h95, cmd, size);
	else
		prepare_header80(mdev, &h->h80, cmd, size);
}

P
Philipp Reisner 已提交
679 680
/* the appropriate socket mutex must be held already */
int _drbd_send_cmd(struct drbd_conf *mdev, struct socket *sock,
681 682
		   enum drbd_packet cmd, struct p_header *h, size_t size,
		   unsigned msg_flags)
P
Philipp Reisner 已提交
683 684 685
{
	int sent, ok;

686 687 688 689
	if (!expect(h))
		return false;
	if (!expect(size))
		return false;
P
Philipp Reisner 已提交
690

P
Philipp Reisner 已提交
691
	prepare_header(mdev, h, cmd, size - sizeof(struct p_header));
P
Philipp Reisner 已提交
692

693
	sent = drbd_send(mdev->tconn, sock, h, size, msg_flags);
P
Philipp Reisner 已提交
694 695

	ok = (sent == size);
696 697
	if (!ok && !signal_pending(current))
		dev_warn(DEV, "short sent %s size=%d sent=%d\n",
P
Philipp Reisner 已提交
698 699 700 701 702 703 704 705
		    cmdname(cmd), (int)size, sent);
	return ok;
}

/* don't pass the socket. we may only look at it
 * when we hold the appropriate socket mutex.
 */
int drbd_send_cmd(struct drbd_conf *mdev, int use_data_socket,
706
		  enum drbd_packet cmd, struct p_header *h, size_t size)
P
Philipp Reisner 已提交
707 708 709 710 711
{
	int ok = 0;
	struct socket *sock;

	if (use_data_socket) {
712 713
		mutex_lock(&mdev->tconn->data.mutex);
		sock = mdev->tconn->data.socket;
P
Philipp Reisner 已提交
714
	} else {
715 716
		mutex_lock(&mdev->tconn->meta.mutex);
		sock = mdev->tconn->meta.socket;
P
Philipp Reisner 已提交
717 718 719 720 721 722 723 724
	}

	/* drbd_disconnect() could have called drbd_free_sock()
	 * while we were waiting in down()... */
	if (likely(sock != NULL))
		ok = _drbd_send_cmd(mdev, sock, cmd, h, size, 0);

	if (use_data_socket)
725
		mutex_unlock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
726
	else
727
		mutex_unlock(&mdev->tconn->meta.mutex);
P
Philipp Reisner 已提交
728 729 730
	return ok;
}

731
int drbd_send_cmd2(struct drbd_conf *mdev, enum drbd_packet cmd, char *data,
P
Philipp Reisner 已提交
732 733
		   size_t size)
{
P
Philipp Reisner 已提交
734
	struct p_header h;
P
Philipp Reisner 已提交
735 736
	int ok;

P
Philipp Reisner 已提交
737
	prepare_header(mdev, &h, cmd, size);
P
Philipp Reisner 已提交
738 739 740 741 742

	if (!drbd_get_data_sock(mdev))
		return 0;

	ok = (sizeof(h) ==
743
		drbd_send(mdev->tconn, mdev->tconn->data.socket, &h, sizeof(h), 0));
P
Philipp Reisner 已提交
744
	ok = ok && (size ==
745
		drbd_send(mdev->tconn, mdev->tconn->data.socket, data, size, 0));
P
Philipp Reisner 已提交
746 747 748 749 750 751 752 753

	drbd_put_data_sock(mdev);

	return ok;
}

int drbd_send_sync_param(struct drbd_conf *mdev, struct syncer_conf *sc)
{
754
	struct p_rs_param_95 *p;
P
Philipp Reisner 已提交
755 756
	struct socket *sock;
	int size, rv;
757
	const int apv = mdev->tconn->agreed_pro_version;
P
Philipp Reisner 已提交
758 759 760 761

	size = apv <= 87 ? sizeof(struct p_rs_param)
		: apv == 88 ? sizeof(struct p_rs_param)
			+ strlen(mdev->sync_conf.verify_alg) + 1
762 763
		: apv <= 94 ? sizeof(struct p_rs_param_89)
		: /* apv >= 95 */ sizeof(struct p_rs_param_95);
P
Philipp Reisner 已提交
764 765 766 767

	/* used from admin command context and receiver/worker context.
	 * to avoid kmalloc, grab the socket right here,
	 * then use the pre-allocated sbuf there */
768 769
	mutex_lock(&mdev->tconn->data.mutex);
	sock = mdev->tconn->data.socket;
P
Philipp Reisner 已提交
770 771

	if (likely(sock != NULL)) {
772 773
		enum drbd_packet cmd =
			apv >= 89 ? P_SYNC_PARAM89 : P_SYNC_PARAM;
P
Philipp Reisner 已提交
774

775
		p = &mdev->tconn->data.sbuf.rs_param_95;
P
Philipp Reisner 已提交
776 777 778 779 780

		/* initialize verify_alg and csums_alg */
		memset(p->verify_alg, 0, 2 * SHARED_SECRET_MAX);

		p->rate = cpu_to_be32(sc->rate);
781 782 783 784
		p->c_plan_ahead = cpu_to_be32(sc->c_plan_ahead);
		p->c_delay_target = cpu_to_be32(sc->c_delay_target);
		p->c_fill_target = cpu_to_be32(sc->c_fill_target);
		p->c_max_rate = cpu_to_be32(sc->c_max_rate);
P
Philipp Reisner 已提交
785 786 787 788 789 790 791 792 793 794

		if (apv >= 88)
			strcpy(p->verify_alg, mdev->sync_conf.verify_alg);
		if (apv >= 89)
			strcpy(p->csums_alg, mdev->sync_conf.csums_alg);

		rv = _drbd_send_cmd(mdev, sock, cmd, &p->head, size, 0);
	} else
		rv = 0; /* not ok */

795
	mutex_unlock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
796 797 798 799 800 801 802

	return rv;
}

int drbd_send_protocol(struct drbd_conf *mdev)
{
	struct p_protocol *p;
803
	int size, cf, rv;
P
Philipp Reisner 已提交
804 805 806

	size = sizeof(struct p_protocol);

807
	if (mdev->tconn->agreed_pro_version >= 87)
808
		size += strlen(mdev->tconn->net_conf->integrity_alg) + 1;
P
Philipp Reisner 已提交
809 810 811 812 813 814 815

	/* we must not recurse into our own queue,
	 * as that is blocked during handshake */
	p = kmalloc(size, GFP_NOIO);
	if (p == NULL)
		return 0;

816 817 818 819 820
	p->protocol      = cpu_to_be32(mdev->tconn->net_conf->wire_protocol);
	p->after_sb_0p   = cpu_to_be32(mdev->tconn->net_conf->after_sb_0p);
	p->after_sb_1p   = cpu_to_be32(mdev->tconn->net_conf->after_sb_1p);
	p->after_sb_2p   = cpu_to_be32(mdev->tconn->net_conf->after_sb_2p);
	p->two_primaries = cpu_to_be32(mdev->tconn->net_conf->two_primaries);
P
Philipp Reisner 已提交
821

822
	cf = 0;
823
	if (mdev->tconn->net_conf->want_lose)
824
		cf |= CF_WANT_LOSE;
825
	if (mdev->tconn->net_conf->dry_run) {
826
		if (mdev->tconn->agreed_pro_version >= 92)
827 828 829
			cf |= CF_DRY_RUN;
		else {
			dev_err(DEV, "--dry-run is not supported by peer");
D
Dan Carpenter 已提交
830
			kfree(p);
831
			return -1;
832 833 834 835
		}
	}
	p->conn_flags    = cpu_to_be32(cf);

836
	if (mdev->tconn->agreed_pro_version >= 87)
837
		strcpy(p->integrity_alg, mdev->tconn->net_conf->integrity_alg);
P
Philipp Reisner 已提交
838

839
	rv = drbd_send_cmd(mdev, USE_DATA_SOCKET, P_PROTOCOL, &p->head, size);
P
Philipp Reisner 已提交
840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856
	kfree(p);
	return rv;
}

int _drbd_send_uuids(struct drbd_conf *mdev, u64 uuid_flags)
{
	struct p_uuids p;
	int i;

	if (!get_ldev_if_state(mdev, D_NEGOTIATING))
		return 1;

	for (i = UI_CURRENT; i < UI_SIZE; i++)
		p.uuid[i] = mdev->ldev ? cpu_to_be64(mdev->ldev->md.uuid[i]) : 0;

	mdev->comm_bm_set = drbd_bm_total_weight(mdev);
	p.uuid[UI_SIZE] = cpu_to_be64(mdev->comm_bm_set);
857
	uuid_flags |= mdev->tconn->net_conf->want_lose ? 1 : 0;
P
Philipp Reisner 已提交
858 859 860 861 862 863
	uuid_flags |= test_bit(CRASHED_PRIMARY, &mdev->flags) ? 2 : 0;
	uuid_flags |= mdev->new_state_tmp.disk == D_INCONSISTENT ? 4 : 0;
	p.uuid[UI_FLAGS] = cpu_to_be64(uuid_flags);

	put_ldev(mdev);

864
	return drbd_send_cmd(mdev, USE_DATA_SOCKET, P_UUIDS, &p.head, sizeof(p));
P
Philipp Reisner 已提交
865 866 867 868 869 870 871 872 873 874 875 876
}

int drbd_send_uuids(struct drbd_conf *mdev)
{
	return _drbd_send_uuids(mdev, 0);
}

int drbd_send_uuids_skip_initial_sync(struct drbd_conf *mdev)
{
	return _drbd_send_uuids(mdev, 8);
}

877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
void drbd_print_uuids(struct drbd_conf *mdev, const char *text)
{
	if (get_ldev_if_state(mdev, D_NEGOTIATING)) {
		u64 *uuid = mdev->ldev->md.uuid;
		dev_info(DEV, "%s %016llX:%016llX:%016llX:%016llX\n",
		     text,
		     (unsigned long long)uuid[UI_CURRENT],
		     (unsigned long long)uuid[UI_BITMAP],
		     (unsigned long long)uuid[UI_HISTORY_START],
		     (unsigned long long)uuid[UI_HISTORY_END]);
		put_ldev(mdev);
	} else {
		dev_info(DEV, "%s effective data uuid: %016llX\n",
				text,
				(unsigned long long)mdev->ed_uuid);
	}
}

895
int drbd_gen_and_send_sync_uuid(struct drbd_conf *mdev)
P
Philipp Reisner 已提交
896 897
{
	struct p_rs_uuid p;
898 899 900
	u64 uuid;

	D_ASSERT(mdev->state.disk == D_UP_TO_DATE);
P
Philipp Reisner 已提交
901

902
	uuid = mdev->ldev->md.uuid[UI_BITMAP] + UUID_NEW_BM_OFFSET;
903
	drbd_uuid_set(mdev, UI_BITMAP, uuid);
904
	drbd_print_uuids(mdev, "updated sync UUID");
905 906
	drbd_md_sync(mdev);
	p.uuid = cpu_to_be64(uuid);
P
Philipp Reisner 已提交
907

908
	return drbd_send_cmd(mdev, USE_DATA_SOCKET, P_SYNC_UUID, &p.head, sizeof(p));
P
Philipp Reisner 已提交
909 910
}

911
int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply, enum dds_flags flags)
P
Philipp Reisner 已提交
912 913 914
{
	struct p_sizes p;
	sector_t d_size, u_size;
915
	int q_order_type, max_bio_size;
P
Philipp Reisner 已提交
916 917 918 919 920 921 922
	int ok;

	if (get_ldev_if_state(mdev, D_NEGOTIATING)) {
		D_ASSERT(mdev->ldev->backing_bdev);
		d_size = drbd_get_max_capacity(mdev->ldev);
		u_size = mdev->ldev->dc.disk_size;
		q_order_type = drbd_queue_order_type(mdev);
923 924
		max_bio_size = queue_max_hw_sectors(mdev->ldev->backing_bdev->bd_disk->queue) << 9;
		max_bio_size = min_t(int, max_bio_size, DRBD_MAX_BIO_SIZE);
P
Philipp Reisner 已提交
925 926 927 928 929
		put_ldev(mdev);
	} else {
		d_size = 0;
		u_size = 0;
		q_order_type = QUEUE_ORDERED_NONE;
930
		max_bio_size = DRBD_MAX_BIO_SIZE; /* ... multiple BIOs per peer_request */
P
Philipp Reisner 已提交
931 932 933 934 935
	}

	p.d_size = cpu_to_be64(d_size);
	p.u_size = cpu_to_be64(u_size);
	p.c_size = cpu_to_be64(trigger_reply ? 0 : drbd_get_capacity(mdev->this_bdev));
936
	p.max_bio_size = cpu_to_be32(max_bio_size);
937 938
	p.queue_order_type = cpu_to_be16(q_order_type);
	p.dds_flags = cpu_to_be16(flags);
P
Philipp Reisner 已提交
939

940
	ok = drbd_send_cmd(mdev, USE_DATA_SOCKET, P_SIZES, &p.head, sizeof(p));
P
Philipp Reisner 已提交
941 942 943 944 945 946 947 948 949 950 951 952 953
	return ok;
}

/**
 * drbd_send_state() - Sends the drbd state to the peer
 * @mdev:	DRBD device.
 */
int drbd_send_state(struct drbd_conf *mdev)
{
	struct socket *sock;
	struct p_state p;
	int ok = 0;

954
	mutex_lock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
955 956

	p.state = cpu_to_be32(mdev->state.i); /* Within the send mutex */
957
	sock = mdev->tconn->data.socket;
P
Philipp Reisner 已提交
958 959

	if (likely(sock != NULL)) {
960
		ok = _drbd_send_cmd(mdev, sock, P_STATE, &p.head, sizeof(p), 0);
P
Philipp Reisner 已提交
961 962
	}

963
	mutex_unlock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
964 965 966 967 968 969 970 971 972 973 974 975

	return ok;
}

int drbd_send_state_req(struct drbd_conf *mdev,
	union drbd_state mask, union drbd_state val)
{
	struct p_req_state p;

	p.mask    = cpu_to_be32(mask.i);
	p.val     = cpu_to_be32(val.i);

976
	return drbd_send_cmd(mdev, USE_DATA_SOCKET, P_STATE_CHG_REQ, &p.head, sizeof(p));
P
Philipp Reisner 已提交
977 978
}

979
int drbd_send_sr_reply(struct drbd_conf *mdev, enum drbd_state_rv retcode)
P
Philipp Reisner 已提交
980 981 982 983 984
{
	struct p_req_state_reply p;

	p.retcode    = cpu_to_be32(retcode);

985
	return drbd_send_cmd(mdev, USE_META_SOCKET, P_STATE_CHG_REPLY, &p.head, sizeof(p));
P
Philipp Reisner 已提交
986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
}

int fill_bitmap_rle_bits(struct drbd_conf *mdev,
	struct p_compressed_bm *p,
	struct bm_xfer_ctx *c)
{
	struct bitstream bs;
	unsigned long plain_bits;
	unsigned long tmp;
	unsigned long rl;
	unsigned len;
	unsigned toggle;
	int bits;

	/* may we use this feature? */
	if ((mdev->sync_conf.use_rle == 0) ||
1002
		(mdev->tconn->agreed_pro_version < 90))
P
Philipp Reisner 已提交
1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081
			return 0;

	if (c->bit_offset >= c->bm_bits)
		return 0; /* nothing to do. */

	/* use at most thus many bytes */
	bitstream_init(&bs, p->code, BM_PACKET_VLI_BYTES_MAX, 0);
	memset(p->code, 0, BM_PACKET_VLI_BYTES_MAX);
	/* plain bits covered in this code string */
	plain_bits = 0;

	/* p->encoding & 0x80 stores whether the first run length is set.
	 * bit offset is implicit.
	 * start with toggle == 2 to be able to tell the first iteration */
	toggle = 2;

	/* see how much plain bits we can stuff into one packet
	 * using RLE and VLI. */
	do {
		tmp = (toggle == 0) ? _drbd_bm_find_next_zero(mdev, c->bit_offset)
				    : _drbd_bm_find_next(mdev, c->bit_offset);
		if (tmp == -1UL)
			tmp = c->bm_bits;
		rl = tmp - c->bit_offset;

		if (toggle == 2) { /* first iteration */
			if (rl == 0) {
				/* the first checked bit was set,
				 * store start value, */
				DCBP_set_start(p, 1);
				/* but skip encoding of zero run length */
				toggle = !toggle;
				continue;
			}
			DCBP_set_start(p, 0);
		}

		/* paranoia: catch zero runlength.
		 * can only happen if bitmap is modified while we scan it. */
		if (rl == 0) {
			dev_err(DEV, "unexpected zero runlength while encoding bitmap "
			    "t:%u bo:%lu\n", toggle, c->bit_offset);
			return -1;
		}

		bits = vli_encode_bits(&bs, rl);
		if (bits == -ENOBUFS) /* buffer full */
			break;
		if (bits <= 0) {
			dev_err(DEV, "error while encoding bitmap: %d\n", bits);
			return 0;
		}

		toggle = !toggle;
		plain_bits += rl;
		c->bit_offset = tmp;
	} while (c->bit_offset < c->bm_bits);

	len = bs.cur.b - p->code + !!bs.cur.bit;

	if (plain_bits < (len << 3)) {
		/* incompressible with this method.
		 * we need to rewind both word and bit position. */
		c->bit_offset -= plain_bits;
		bm_xfer_ctx_bit_to_word_offset(c);
		c->bit_offset = c->word_offset * BITS_PER_LONG;
		return 0;
	}

	/* RLE + VLI was able to compress it just fine.
	 * update c->word_offset. */
	bm_xfer_ctx_bit_to_word_offset(c);

	/* store pad_bits */
	DCBP_set_pad_bits(p, (8 - bs.cur.bit) & 0x7);

	return len;
}

1082 1083 1084 1085 1086 1087 1088
/**
 * send_bitmap_rle_or_plain
 *
 * Return 0 when done, 1 when another iteration is needed, and a negative error
 * code upon failure.
 */
static int
P
Philipp Reisner 已提交
1089
send_bitmap_rle_or_plain(struct drbd_conf *mdev,
1090
			 struct p_header *h, struct bm_xfer_ctx *c)
P
Philipp Reisner 已提交
1091 1092 1093 1094 1095 1096 1097 1098 1099
{
	struct p_compressed_bm *p = (void*)h;
	unsigned long num_words;
	int len;
	int ok;

	len = fill_bitmap_rle_bits(mdev, p, c);

	if (len < 0)
1100
		return -EIO;
P
Philipp Reisner 已提交
1101 1102 1103

	if (len) {
		DCBP_set_code(p, RLE_VLI_Bits);
1104
		ok = _drbd_send_cmd(mdev, mdev->tconn->data.socket, P_COMPRESSED_BITMAP, h,
P
Philipp Reisner 已提交
1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118
			sizeof(*p) + len, 0);

		c->packets[0]++;
		c->bytes[0] += sizeof(*p) + len;

		if (c->bit_offset >= c->bm_bits)
			len = 0; /* DONE */
	} else {
		/* was not compressible.
		 * send a buffer full of plain text bits instead. */
		num_words = min_t(size_t, BM_PACKET_WORDS, c->bm_words - c->word_offset);
		len = num_words * sizeof(long);
		if (len)
			drbd_bm_get_lel(mdev, c->word_offset, num_words, (unsigned long*)h->payload);
1119
		ok = _drbd_send_cmd(mdev, mdev->tconn->data.socket, P_BITMAP,
1120
				   h, sizeof(struct p_header80) + len, 0);
P
Philipp Reisner 已提交
1121 1122 1123 1124
		c->word_offset += num_words;
		c->bit_offset = c->word_offset * BITS_PER_LONG;

		c->packets[1]++;
1125
		c->bytes[1] += sizeof(struct p_header80) + len;
P
Philipp Reisner 已提交
1126 1127 1128 1129

		if (c->bit_offset > c->bm_bits)
			c->bit_offset = c->bm_bits;
	}
1130 1131 1132 1133 1134 1135 1136 1137
	if (ok) {
		if (len == 0) {
			INFO_bm_xfer_stats(mdev, "send", c);
			return 0;
		} else
			return 1;
	}
	return -EIO;
P
Philipp Reisner 已提交
1138 1139 1140 1141 1142 1143
}

/* See the comment at receive_bitmap() */
int _drbd_send_bitmap(struct drbd_conf *mdev)
{
	struct bm_xfer_ctx c;
1144
	struct p_header *p;
1145
	int err;
P
Philipp Reisner 已提交
1146

1147 1148
	if (!expect(mdev->bitmap))
		return false;
P
Philipp Reisner 已提交
1149 1150 1151

	/* maybe we should use some per thread scratch page,
	 * and allocate that during initial device creation? */
1152
	p = (struct p_header *) __get_free_page(GFP_NOIO);
P
Philipp Reisner 已提交
1153 1154
	if (!p) {
		dev_err(DEV, "failed to allocate one page buffer in %s\n", __func__);
1155
		return false;
P
Philipp Reisner 已提交
1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180
	}

	if (get_ldev(mdev)) {
		if (drbd_md_test_flag(mdev->ldev, MDF_FULL_SYNC)) {
			dev_info(DEV, "Writing the whole bitmap, MDF_FullSync was set.\n");
			drbd_bm_set_all(mdev);
			if (drbd_bm_write(mdev)) {
				/* write_bm did fail! Leave full sync flag set in Meta P_DATA
				 * but otherwise process as per normal - need to tell other
				 * side that a full resync is required! */
				dev_err(DEV, "Failed to write bitmap to disk!\n");
			} else {
				drbd_md_clear_flag(mdev, MDF_FULL_SYNC);
				drbd_md_sync(mdev);
			}
		}
		put_ldev(mdev);
	}

	c = (struct bm_xfer_ctx) {
		.bm_bits = drbd_bm_bits(mdev),
		.bm_words = drbd_bm_words(mdev),
	};

	do {
1181 1182
		err = send_bitmap_rle_or_plain(mdev, p, &c);
	} while (err > 0);
P
Philipp Reisner 已提交
1183 1184

	free_page((unsigned long) p);
1185
	return err == 0;
P
Philipp Reisner 已提交
1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207
}

int drbd_send_bitmap(struct drbd_conf *mdev)
{
	int err;

	if (!drbd_get_data_sock(mdev))
		return -1;
	err = !_drbd_send_bitmap(mdev);
	drbd_put_data_sock(mdev);
	return err;
}

int drbd_send_b_ack(struct drbd_conf *mdev, u32 barrier_nr, u32 set_size)
{
	int ok;
	struct p_barrier_ack p;

	p.barrier  = barrier_nr;
	p.set_size = cpu_to_be32(set_size);

	if (mdev->state.conn < C_CONNECTED)
1208
		return false;
1209
	ok = drbd_send_cmd(mdev, USE_META_SOCKET, P_BARRIER_ACK, &p.head, sizeof(p));
P
Philipp Reisner 已提交
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220
	return ok;
}

/**
 * _drbd_send_ack() - Sends an ack packet
 * @mdev:	DRBD device.
 * @cmd:	Packet command code.
 * @sector:	sector, needs to be in big endian byte order
 * @blksize:	size in byte, needs to be in big endian byte order
 * @block_id:	Id, big endian byte order
 */
1221 1222
static int _drbd_send_ack(struct drbd_conf *mdev, enum drbd_packet cmd,
			  u64 sector, u32 blksize, u64 block_id)
P
Philipp Reisner 已提交
1223 1224 1225 1226 1227 1228 1229 1230 1231
{
	int ok;
	struct p_block_ack p;

	p.sector   = sector;
	p.block_id = block_id;
	p.blksize  = blksize;
	p.seq_num  = cpu_to_be32(atomic_add_return(1, &mdev->packet_seq));

1232
	if (!mdev->tconn->meta.socket || mdev->state.conn < C_CONNECTED)
1233
		return false;
1234
	ok = drbd_send_cmd(mdev, USE_META_SOCKET, cmd, &p.head, sizeof(p));
P
Philipp Reisner 已提交
1235 1236 1237
	return ok;
}

1238 1239 1240
/* dp->sector and dp->block_id already/still in network byte order,
 * data_size is payload size according to dp->head,
 * and may need to be corrected for digest size. */
1241
int drbd_send_ack_dp(struct drbd_conf *mdev, enum drbd_packet cmd,
1242
		     struct p_data *dp, int data_size)
P
Philipp Reisner 已提交
1243
{
1244 1245
	data_size -= (mdev->tconn->agreed_pro_version >= 87 && mdev->tconn->integrity_r_tfm) ?
		crypto_hash_digestsize(mdev->tconn->integrity_r_tfm) : 0;
P
Philipp Reisner 已提交
1246 1247 1248 1249
	return _drbd_send_ack(mdev, cmd, dp->sector, cpu_to_be32(data_size),
			      dp->block_id);
}

1250
int drbd_send_ack_rp(struct drbd_conf *mdev, enum drbd_packet cmd,
P
Philipp Reisner 已提交
1251 1252 1253 1254 1255 1256 1257
		     struct p_block_req *rp)
{
	return _drbd_send_ack(mdev, cmd, rp->sector, rp->blksize, rp->block_id);
}

/**
 * drbd_send_ack() - Sends an ack packet
1258 1259 1260
 * @mdev:	DRBD device
 * @cmd:	packet command code
 * @peer_req:	peer request
P
Philipp Reisner 已提交
1261
 */
1262
int drbd_send_ack(struct drbd_conf *mdev, enum drbd_packet cmd,
1263
		  struct drbd_peer_request *peer_req)
P
Philipp Reisner 已提交
1264 1265
{
	return _drbd_send_ack(mdev, cmd,
1266 1267 1268
			      cpu_to_be64(peer_req->i.sector),
			      cpu_to_be32(peer_req->i.size),
			      peer_req->block_id);
P
Philipp Reisner 已提交
1269 1270 1271 1272
}

/* This function misuses the block_id field to signal if the blocks
 * are is sync or not. */
1273
int drbd_send_ack_ex(struct drbd_conf *mdev, enum drbd_packet cmd,
P
Philipp Reisner 已提交
1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
		     sector_t sector, int blksize, u64 block_id)
{
	return _drbd_send_ack(mdev, cmd,
			      cpu_to_be64(sector),
			      cpu_to_be32(blksize),
			      cpu_to_be64(block_id));
}

int drbd_send_drequest(struct drbd_conf *mdev, int cmd,
		       sector_t sector, int size, u64 block_id)
{
	int ok;
	struct p_block_req p;

	p.sector   = cpu_to_be64(sector);
	p.block_id = block_id;
	p.blksize  = cpu_to_be32(size);

1292
	ok = drbd_send_cmd(mdev, USE_DATA_SOCKET, cmd, &p.head, sizeof(p));
P
Philipp Reisner 已提交
1293 1294 1295
	return ok;
}

1296 1297
int drbd_send_drequest_csum(struct drbd_conf *mdev, sector_t sector, int size,
			    void *digest, int digest_size, enum drbd_packet cmd)
P
Philipp Reisner 已提交
1298 1299 1300 1301
{
	int ok;
	struct p_block_req p;

P
Philipp Reisner 已提交
1302
	prepare_header(mdev, &p.head, cmd, sizeof(p) - sizeof(struct p_header) + digest_size);
P
Philipp Reisner 已提交
1303
	p.sector   = cpu_to_be64(sector);
1304
	p.block_id = ID_SYNCER /* unused */;
P
Philipp Reisner 已提交
1305 1306
	p.blksize  = cpu_to_be32(size);

1307
	mutex_lock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
1308

1309 1310
	ok = (sizeof(p) == drbd_send(mdev->tconn, mdev->tconn->data.socket, &p, sizeof(p), 0));
	ok = ok && (digest_size == drbd_send(mdev->tconn, mdev->tconn->data.socket, digest, digest_size, 0));
P
Philipp Reisner 已提交
1311

1312
	mutex_unlock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322

	return ok;
}

int drbd_send_ov_request(struct drbd_conf *mdev, sector_t sector, int size)
{
	int ok;
	struct p_block_req p;

	p.sector   = cpu_to_be64(sector);
1323
	p.block_id = ID_SYNCER /* unused */;
P
Philipp Reisner 已提交
1324 1325
	p.blksize  = cpu_to_be32(size);

1326
	ok = drbd_send_cmd(mdev, USE_DATA_SOCKET, P_OV_REQUEST, &p.head, sizeof(p));
P
Philipp Reisner 已提交
1327 1328 1329 1330
	return ok;
}

/* called on sndtimeo
1331 1332
 * returns false if we should retry,
 * true if we think connection is dead
P
Philipp Reisner 已提交
1333
 */
1334
static int we_should_drop_the_connection(struct drbd_tconn *tconn, struct socket *sock)
P
Philipp Reisner 已提交
1335 1336 1337 1338
{
	int drop_it;
	/* long elapsed = (long)(jiffies - mdev->last_received); */

1339 1340 1341 1342
	drop_it =   tconn->meta.socket == sock
		|| !tconn->asender.task
		|| get_t_state(&tconn->asender) != RUNNING
		|| tconn->volume0->state.conn < C_CONNECTED;
P
Philipp Reisner 已提交
1343 1344

	if (drop_it)
1345
		return true;
P
Philipp Reisner 已提交
1346

1347
	drop_it = !--tconn->ko_count;
P
Philipp Reisner 已提交
1348
	if (!drop_it) {
1349 1350 1351
		conn_err(tconn, "[%s/%d] sock_sendmsg time expired, ko = %u\n",
			 current->comm, current->pid, tconn->ko_count);
		request_ping(tconn);
P
Philipp Reisner 已提交
1352 1353 1354 1355 1356
	}

	return drop_it; /* && (mdev->state == R_PRIMARY) */;
}

1357
static void drbd_update_congested(struct drbd_tconn *tconn)
1358
{
1359
	struct sock *sk = tconn->data.socket->sk;
1360
	if (sk->sk_wmem_queued > sk->sk_sndbuf * 4 / 5)
1361
		set_bit(NET_CONGESTED, &tconn->flags);
1362 1363
}

P
Philipp Reisner 已提交
1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385
/* The idea of sendpage seems to be to put some kind of reference
 * to the page into the skb, and to hand it over to the NIC. In
 * this process get_page() gets called.
 *
 * As soon as the page was really sent over the network put_page()
 * gets called by some part of the network layer. [ NIC driver? ]
 *
 * [ get_page() / put_page() increment/decrement the count. If count
 *   reaches 0 the page will be freed. ]
 *
 * This works nicely with pages from FSs.
 * But this means that in protocol A we might signal IO completion too early!
 *
 * In order not to corrupt data during a resync we must make sure
 * that we do not reuse our own buffer pages (EEs) to early, therefore
 * we have the net_ee list.
 *
 * XFS seems to have problems, still, it submits pages with page_count == 0!
 * As a workaround, we disable sendpage on pages
 * with page_count == 0 or PageSlab.
 */
static int _drbd_no_send_page(struct drbd_conf *mdev, struct page *page,
L
Lars Ellenberg 已提交
1386
		   int offset, size_t size, unsigned msg_flags)
P
Philipp Reisner 已提交
1387
{
1388
	int sent = drbd_send(mdev->tconn, mdev->tconn->data.socket, kmap(page) + offset, size, msg_flags);
P
Philipp Reisner 已提交
1389 1390 1391 1392 1393 1394 1395
	kunmap(page);
	if (sent == size)
		mdev->send_cnt += size>>9;
	return sent == size;
}

static int _drbd_send_page(struct drbd_conf *mdev, struct page *page,
L
Lars Ellenberg 已提交
1396
		    int offset, size_t size, unsigned msg_flags)
P
Philipp Reisner 已提交
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408
{
	mm_segment_t oldfs = get_fs();
	int sent, ok;
	int len = size;

	/* e.g. XFS meta- & log-data is in slab pages, which have a
	 * page_count of 0 and/or have PageSlab() set.
	 * we cannot use send_page for those, as that does get_page();
	 * put_page(); and would cause either a VM_BUG directly, or
	 * __page_cache_release a page that would actually still be referenced
	 * by someone, leading to some obscure delayed Oops somewhere else. */
	if (disable_sendpage || (page_count(page) < 1) || PageSlab(page))
L
Lars Ellenberg 已提交
1409
		return _drbd_no_send_page(mdev, page, offset, size, msg_flags);
P
Philipp Reisner 已提交
1410

L
Lars Ellenberg 已提交
1411
	msg_flags |= MSG_NOSIGNAL;
1412
	drbd_update_congested(mdev->tconn);
P
Philipp Reisner 已提交
1413 1414
	set_fs(KERNEL_DS);
	do {
1415
		sent = mdev->tconn->data.socket->ops->sendpage(mdev->tconn->data.socket, page,
P
Philipp Reisner 已提交
1416
							offset, len,
L
Lars Ellenberg 已提交
1417
							msg_flags);
P
Philipp Reisner 已提交
1418
		if (sent == -EAGAIN) {
1419
			if (we_should_drop_the_connection(mdev->tconn,
1420
							  mdev->tconn->data.socket))
P
Philipp Reisner 已提交
1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433
				break;
			else
				continue;
		}
		if (sent <= 0) {
			dev_warn(DEV, "%s: size=%d len=%d sent=%d\n",
			     __func__, (int)size, len, sent);
			break;
		}
		len    -= sent;
		offset += sent;
	} while (len > 0 /* THINK && mdev->cstate >= C_CONNECTED*/);
	set_fs(oldfs);
1434
	clear_bit(NET_CONGESTED, &mdev->tconn->flags);
P
Philipp Reisner 已提交
1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445

	ok = (len == 0);
	if (likely(ok))
		mdev->send_cnt += size>>9;
	return ok;
}

static int _drbd_send_bio(struct drbd_conf *mdev, struct bio *bio)
{
	struct bio_vec *bvec;
	int i;
L
Lars Ellenberg 已提交
1446
	/* hint all but last page with MSG_MORE */
P
Philipp Reisner 已提交
1447 1448
	__bio_for_each_segment(bvec, bio, i, 0) {
		if (!_drbd_no_send_page(mdev, bvec->bv_page,
L
Lars Ellenberg 已提交
1449 1450
				     bvec->bv_offset, bvec->bv_len,
				     i == bio->bi_vcnt -1 ? 0 : MSG_MORE))
P
Philipp Reisner 已提交
1451 1452 1453 1454 1455 1456 1457 1458 1459
			return 0;
	}
	return 1;
}

static int _drbd_send_zc_bio(struct drbd_conf *mdev, struct bio *bio)
{
	struct bio_vec *bvec;
	int i;
L
Lars Ellenberg 已提交
1460
	/* hint all but last page with MSG_MORE */
P
Philipp Reisner 已提交
1461 1462
	__bio_for_each_segment(bvec, bio, i, 0) {
		if (!_drbd_send_page(mdev, bvec->bv_page,
L
Lars Ellenberg 已提交
1463 1464
				     bvec->bv_offset, bvec->bv_len,
				     i == bio->bi_vcnt -1 ? 0 : MSG_MORE))
P
Philipp Reisner 已提交
1465 1466 1467 1468 1469
			return 0;
	}
	return 1;
}

1470 1471
static int _drbd_send_zc_ee(struct drbd_conf *mdev,
			    struct drbd_peer_request *peer_req)
1472
{
1473 1474 1475
	struct page *page = peer_req->pages;
	unsigned len = peer_req->i.size;

L
Lars Ellenberg 已提交
1476
	/* hint all but last page with MSG_MORE */
1477 1478
	page_chain_for_each(page) {
		unsigned l = min_t(unsigned, len, PAGE_SIZE);
L
Lars Ellenberg 已提交
1479 1480
		if (!_drbd_send_page(mdev, page, 0, l,
				page_chain_next(page) ? MSG_MORE : 0))
1481 1482 1483 1484 1485 1486
			return 0;
		len -= l;
	}
	return 1;
}

1487 1488
static u32 bio_flags_to_wire(struct drbd_conf *mdev, unsigned long bi_rw)
{
1489
	if (mdev->tconn->agreed_pro_version >= 95)
1490 1491 1492 1493 1494
		return  (bi_rw & REQ_SYNC ? DP_RW_SYNC : 0) |
			(bi_rw & REQ_FUA ? DP_FUA : 0) |
			(bi_rw & REQ_FLUSH ? DP_FLUSH : 0) |
			(bi_rw & REQ_DISCARD ? DP_DISCARD : 0);
	else
J
Jens Axboe 已提交
1495
		return bi_rw & REQ_SYNC ? DP_RW_SYNC : 0;
1496 1497
}

P
Philipp Reisner 已提交
1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511
/* Used to send write requests
 * R_PRIMARY -> Peer	(P_DATA)
 */
int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req)
{
	int ok = 1;
	struct p_data p;
	unsigned int dp_flags = 0;
	void *dgb;
	int dgs;

	if (!drbd_get_data_sock(mdev))
		return 0;

1512 1513
	dgs = (mdev->tconn->agreed_pro_version >= 87 && mdev->tconn->integrity_w_tfm) ?
		crypto_hash_digestsize(mdev->tconn->integrity_w_tfm) : 0;
P
Philipp Reisner 已提交
1514

P
Philipp Reisner 已提交
1515
	prepare_header(mdev, &p.head, P_DATA, sizeof(p) - sizeof(struct p_header) + dgs + req->i.size);
1516
	p.sector   = cpu_to_be64(req->i.sector);
P
Philipp Reisner 已提交
1517
	p.block_id = (unsigned long)req;
P
Philipp Reisner 已提交
1518
	p.seq_num  = cpu_to_be32(req->seq_num = atomic_add_return(1, &mdev->packet_seq));
P
Philipp Reisner 已提交
1519

1520 1521
	dp_flags = bio_flags_to_wire(mdev, req->master_bio->bi_rw);

P
Philipp Reisner 已提交
1522 1523 1524 1525 1526 1527 1528
	if (mdev->state.conn >= C_SYNC_SOURCE &&
	    mdev->state.conn <= C_PAUSED_SYNC_T)
		dp_flags |= DP_MAY_SET_IN_SYNC;

	p.dp_flags = cpu_to_be32(dp_flags);
	set_bit(UNPLUG_REMOTE, &mdev->flags);
	ok = (sizeof(p) ==
1529
		drbd_send(mdev->tconn, mdev->tconn->data.socket, &p, sizeof(p), dgs ? MSG_MORE : 0));
P
Philipp Reisner 已提交
1530
	if (ok && dgs) {
1531 1532
		dgb = mdev->tconn->int_dig_out;
		drbd_csum_bio(mdev, mdev->tconn->integrity_w_tfm, req->master_bio, dgb);
1533
		ok = dgs == drbd_send(mdev->tconn, mdev->tconn->data.socket, dgb, dgs, 0);
P
Philipp Reisner 已提交
1534 1535
	}
	if (ok) {
1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546
		/* For protocol A, we have to memcpy the payload into
		 * socket buffers, as we may complete right away
		 * as soon as we handed it over to tcp, at which point the data
		 * pages may become invalid.
		 *
		 * For data-integrity enabled, we copy it as well, so we can be
		 * sure that even if the bio pages may still be modified, it
		 * won't change the data on the wire, thus if the digest checks
		 * out ok after sending on this side, but does not fit on the
		 * receiving side, we sure have detected corruption elsewhere.
		 */
1547
		if (mdev->tconn->net_conf->wire_protocol == DRBD_PROT_A || dgs)
P
Philipp Reisner 已提交
1548 1549 1550
			ok = _drbd_send_bio(mdev, req->master_bio);
		else
			ok = _drbd_send_zc_bio(mdev, req->master_bio);
1551 1552 1553

		/* double check digest, sometimes buffers have been modified in flight. */
		if (dgs > 0 && dgs <= 64) {
B
Bart Van Assche 已提交
1554
			/* 64 byte, 512 bit, is the largest digest size
1555 1556
			 * currently supported in kernel crypto. */
			unsigned char digest[64];
1557 1558
			drbd_csum_bio(mdev, mdev->tconn->integrity_w_tfm, req->master_bio, digest);
			if (memcmp(mdev->tconn->int_dig_out, digest, dgs)) {
1559 1560
				dev_warn(DEV,
					"Digest mismatch, buffer modified by upper layers during write: %llus +%u\n",
1561
					(unsigned long long)req->i.sector, req->i.size);
1562 1563 1564 1565
			}
		} /* else if (dgs > 64) {
		     ... Be noisy about digest too large ...
		} */
P
Philipp Reisner 已提交
1566 1567 1568
	}

	drbd_put_data_sock(mdev);
1569

P
Philipp Reisner 已提交
1570 1571 1572 1573 1574 1575 1576
	return ok;
}

/* answer packet, used to send data back for read requests:
 *  Peer       -> (diskless) R_PRIMARY   (P_DATA_REPLY)
 *  C_SYNC_SOURCE -> C_SYNC_TARGET         (P_RS_DATA_REPLY)
 */
1577
int drbd_send_block(struct drbd_conf *mdev, enum drbd_packet cmd,
1578
		    struct drbd_peer_request *peer_req)
P
Philipp Reisner 已提交
1579 1580 1581 1582 1583 1584
{
	int ok;
	struct p_data p;
	void *dgb;
	int dgs;

1585 1586
	dgs = (mdev->tconn->agreed_pro_version >= 87 && mdev->tconn->integrity_w_tfm) ?
		crypto_hash_digestsize(mdev->tconn->integrity_w_tfm) : 0;
P
Philipp Reisner 已提交
1587

1588 1589 1590 1591 1592
	prepare_header(mdev, &p.head, cmd, sizeof(p) -
					   sizeof(struct p_header80) +
					   dgs + peer_req->i.size);
	p.sector   = cpu_to_be64(peer_req->i.sector);
	p.block_id = peer_req->block_id;
1593
	p.seq_num = 0;  /* unused */
P
Philipp Reisner 已提交
1594 1595 1596 1597 1598 1599 1600 1601

	/* Only called by our kernel thread.
	 * This one may be interrupted by DRBD_SIG and/or DRBD_SIGKILL
	 * in response to admin command or module unload.
	 */
	if (!drbd_get_data_sock(mdev))
		return 0;

1602
	ok = sizeof(p) == drbd_send(mdev->tconn, mdev->tconn->data.socket, &p, sizeof(p), dgs ? MSG_MORE : 0);
P
Philipp Reisner 已提交
1603
	if (ok && dgs) {
1604
		dgb = mdev->tconn->int_dig_out;
1605
		drbd_csum_ee(mdev, mdev->tconn->integrity_w_tfm, peer_req, dgb);
1606
		ok = dgs == drbd_send(mdev->tconn, mdev->tconn->data.socket, dgb, dgs, 0);
P
Philipp Reisner 已提交
1607 1608
	}
	if (ok)
1609
		ok = _drbd_send_zc_ee(mdev, peer_req);
P
Philipp Reisner 已提交
1610 1611

	drbd_put_data_sock(mdev);
1612

P
Philipp Reisner 已提交
1613 1614 1615
	return ok;
}

1616 1617 1618 1619
int drbd_send_oos(struct drbd_conf *mdev, struct drbd_request *req)
{
	struct p_block_desc p;

1620 1621
	p.sector  = cpu_to_be64(req->i.sector);
	p.blksize = cpu_to_be32(req->i.size);
1622 1623 1624 1625

	return drbd_send_cmd(mdev, USE_DATA_SOCKET, P_OUT_OF_SYNC, &p.head, sizeof(p));
}

P
Philipp Reisner 已提交
1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641
/*
  drbd_send distinguishes two cases:

  Packets sent via the data socket "sock"
  and packets sent via the meta data socket "msock"

		    sock                      msock
  -----------------+-------------------------+------------------------------
  timeout           conf.timeout / 2          conf.timeout / 2
  timeout action    send a ping via msock     Abort communication
					      and close all sockets
*/

/*
 * you must have down()ed the appropriate [m]sock_mutex elsewhere!
 */
1642
int drbd_send(struct drbd_tconn *tconn, struct socket *sock,
P
Philipp Reisner 已提交
1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662
	      void *buf, size_t size, unsigned msg_flags)
{
	struct kvec iov;
	struct msghdr msg;
	int rv, sent = 0;

	if (!sock)
		return -1000;

	/* THINK  if (signal_pending) return ... ? */

	iov.iov_base = buf;
	iov.iov_len  = size;

	msg.msg_name       = NULL;
	msg.msg_namelen    = 0;
	msg.msg_control    = NULL;
	msg.msg_controllen = 0;
	msg.msg_flags      = msg_flags | MSG_NOSIGNAL;

1663 1664 1665
	if (sock == tconn->data.socket) {
		tconn->ko_count = tconn->net_conf->ko_count;
		drbd_update_congested(tconn);
P
Philipp Reisner 已提交
1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678
	}
	do {
		/* STRANGE
		 * tcp_sendmsg does _not_ use its size parameter at all ?
		 *
		 * -EAGAIN on timeout, -EINTR on signal.
		 */
/* THINK
 * do we need to block DRBD_SIG if sock == &meta.socket ??
 * otherwise wake_asender() might interrupt some send_*Ack !
 */
		rv = kernel_sendmsg(sock, &msg, &iov, 1, size);
		if (rv == -EAGAIN) {
1679
			if (we_should_drop_the_connection(tconn, sock))
P
Philipp Reisner 已提交
1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694
				break;
			else
				continue;
		}
		if (rv == -EINTR) {
			flush_signals(current);
			rv = 0;
		}
		if (rv < 0)
			break;
		sent += rv;
		iov.iov_base += rv;
		iov.iov_len  -= rv;
	} while (sent < size);

1695 1696
	if (sock == tconn->data.socket)
		clear_bit(NET_CONGESTED, &tconn->flags);
P
Philipp Reisner 已提交
1697 1698 1699

	if (rv <= 0) {
		if (rv != -EAGAIN) {
1700 1701 1702 1703
			conn_err(tconn, "%s_sendmsg returned %d\n",
				 sock == tconn->meta.socket ? "msock" : "sock",
				 rv);
			drbd_force_state(tconn->volume0, NS(conn, C_BROKEN_PIPE));
P
Philipp Reisner 已提交
1704
		} else
1705
			drbd_force_state(tconn->volume0, NS(conn, C_TIMEOUT));
P
Philipp Reisner 已提交
1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716
	}

	return sent;
}

static int drbd_open(struct block_device *bdev, fmode_t mode)
{
	struct drbd_conf *mdev = bdev->bd_disk->private_data;
	unsigned long flags;
	int rv = 0;

1717
	mutex_lock(&drbd_main_mutex);
1718
	spin_lock_irqsave(&mdev->tconn->req_lock, flags);
P
Philipp Reisner 已提交
1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730
	/* to have a stable mdev->state.role
	 * and no race with updating open_cnt */

	if (mdev->state.role != R_PRIMARY) {
		if (mode & FMODE_WRITE)
			rv = -EROFS;
		else if (!allow_oos)
			rv = -EMEDIUMTYPE;
	}

	if (!rv)
		mdev->open_cnt++;
1731
	spin_unlock_irqrestore(&mdev->tconn->req_lock, flags);
1732
	mutex_unlock(&drbd_main_mutex);
P
Philipp Reisner 已提交
1733 1734 1735 1736 1737 1738 1739

	return rv;
}

static int drbd_release(struct gendisk *gd, fmode_t mode)
{
	struct drbd_conf *mdev = gd->private_data;
1740
	mutex_lock(&drbd_main_mutex);
P
Philipp Reisner 已提交
1741
	mdev->open_cnt--;
1742
	mutex_unlock(&drbd_main_mutex);
P
Philipp Reisner 已提交
1743 1744 1745 1746 1747
	return 0;
}

static void drbd_set_defaults(struct drbd_conf *mdev)
{
1748 1749 1750 1751 1752 1753 1754 1755 1756
	/* This way we get a compile error when sync_conf grows,
	   and we forgot to initialize it here */
	mdev->sync_conf = (struct syncer_conf) {
		/* .rate = */		DRBD_RATE_DEF,
		/* .after = */		DRBD_AFTER_DEF,
		/* .al_extents = */	DRBD_AL_EXTENTS_DEF,
		/* .verify_alg = */	{}, 0,
		/* .cpu_mask = */	{}, 0,
		/* .csums_alg = */	{}, 0,
1757
		/* .use_rle = */	0,
1758 1759 1760 1761
		/* .on_no_data = */	DRBD_ON_NO_DATA_DEF,
		/* .c_plan_ahead = */	DRBD_C_PLAN_AHEAD_DEF,
		/* .c_delay_target = */	DRBD_C_DELAY_TARGET_DEF,
		/* .c_fill_target = */	DRBD_C_FILL_TARGET_DEF,
1762 1763
		/* .c_max_rate = */	DRBD_C_MAX_RATE_DEF,
		/* .c_min_rate = */	DRBD_C_MIN_RATE_DEF
1764 1765 1766 1767
	};

	/* Have to use that way, because the layout differs between
	   big endian and little endian */
P
Philipp Reisner 已提交
1768 1769 1770 1771 1772 1773
	mdev->state = (union drbd_state) {
		{ .role = R_SECONDARY,
		  .peer = R_UNKNOWN,
		  .conn = C_STANDALONE,
		  .disk = D_DISKLESS,
		  .pdsk = D_UNKNOWN,
1774 1775 1776
		  .susp = 0,
		  .susp_nod = 0,
		  .susp_fen = 0
P
Philipp Reisner 已提交
1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792
		} };
}

void drbd_init_set_defaults(struct drbd_conf *mdev)
{
	/* the memset(,0,) did most of this.
	 * note: only assignments, no allocation in here */

	drbd_set_defaults(mdev);

	atomic_set(&mdev->ap_bio_cnt, 0);
	atomic_set(&mdev->ap_pending_cnt, 0);
	atomic_set(&mdev->rs_pending_cnt, 0);
	atomic_set(&mdev->unacked_cnt, 0);
	atomic_set(&mdev->local_cnt, 0);
	atomic_set(&mdev->pp_in_use, 0);
1793
	atomic_set(&mdev->pp_in_use_by_net, 0);
1794
	atomic_set(&mdev->rs_sect_in, 0);
1795
	atomic_set(&mdev->rs_sect_ev, 0);
1796
	atomic_set(&mdev->ap_in_flight, 0);
P
Philipp Reisner 已提交
1797 1798

	mutex_init(&mdev->md_io_mutex);
1799 1800 1801 1802
	mutex_init(&mdev->tconn->data.mutex);
	mutex_init(&mdev->tconn->meta.mutex);
	sema_init(&mdev->tconn->data.work.s, 0);
	sema_init(&mdev->tconn->meta.work.s, 0);
P
Philipp Reisner 已提交
1803 1804
	mutex_init(&mdev->state_mutex);

1805 1806
	spin_lock_init(&mdev->tconn->data.work.q_lock);
	spin_lock_init(&mdev->tconn->meta.work.q_lock);
P
Philipp Reisner 已提交
1807 1808

	spin_lock_init(&mdev->al_lock);
1809
	spin_lock_init(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
1810 1811 1812 1813 1814 1815 1816 1817 1818
	spin_lock_init(&mdev->peer_seq_lock);
	spin_lock_init(&mdev->epoch_lock);

	INIT_LIST_HEAD(&mdev->active_ee);
	INIT_LIST_HEAD(&mdev->sync_ee);
	INIT_LIST_HEAD(&mdev->done_ee);
	INIT_LIST_HEAD(&mdev->read_ee);
	INIT_LIST_HEAD(&mdev->net_ee);
	INIT_LIST_HEAD(&mdev->resync_reads);
1819 1820
	INIT_LIST_HEAD(&mdev->tconn->data.work.q);
	INIT_LIST_HEAD(&mdev->tconn->meta.work.q);
P
Philipp Reisner 已提交
1821 1822
	INIT_LIST_HEAD(&mdev->resync_work.list);
	INIT_LIST_HEAD(&mdev->unplug_work.list);
1823
	INIT_LIST_HEAD(&mdev->go_diskless.list);
P
Philipp Reisner 已提交
1824
	INIT_LIST_HEAD(&mdev->md_sync_work.list);
1825
	INIT_LIST_HEAD(&mdev->start_resync_work.list);
P
Philipp Reisner 已提交
1826
	INIT_LIST_HEAD(&mdev->bm_io_work.w.list);
1827

1828
	mdev->resync_work.cb  = w_resync_timer;
P
Philipp Reisner 已提交
1829
	mdev->unplug_work.cb  = w_send_write_hint;
1830
	mdev->go_diskless.cb  = w_go_diskless;
P
Philipp Reisner 已提交
1831 1832
	mdev->md_sync_work.cb = w_md_sync;
	mdev->bm_io_work.w.cb = w_bitmap_io;
1833
	mdev->start_resync_work.cb = w_start_resync;
P
Philipp Reisner 已提交
1834 1835
	init_timer(&mdev->resync_timer);
	init_timer(&mdev->md_sync_timer);
1836
	init_timer(&mdev->start_resync_timer);
1837
	init_timer(&mdev->request_timer);
P
Philipp Reisner 已提交
1838 1839 1840 1841
	mdev->resync_timer.function = resync_timer_fn;
	mdev->resync_timer.data = (unsigned long) mdev;
	mdev->md_sync_timer.function = md_sync_timer_fn;
	mdev->md_sync_timer.data = (unsigned long) mdev;
1842 1843
	mdev->start_resync_timer.function = start_resync_timer_fn;
	mdev->start_resync_timer.data = (unsigned long) mdev;
1844 1845
	mdev->request_timer.function = request_timer_fn;
	mdev->request_timer.data = (unsigned long) mdev;
P
Philipp Reisner 已提交
1846 1847 1848 1849 1850 1851 1852

	init_waitqueue_head(&mdev->misc_wait);
	init_waitqueue_head(&mdev->state_wait);
	init_waitqueue_head(&mdev->ee_wait);
	init_waitqueue_head(&mdev->al_wait);
	init_waitqueue_head(&mdev->seq_wait);

1853 1854 1855
	drbd_thread_init(mdev, &mdev->tconn->receiver, drbdd_init, "receiver");
	drbd_thread_init(mdev, &mdev->tconn->worker, drbd_worker, "worker");
	drbd_thread_init(mdev, &mdev->tconn->asender, drbd_asender, "asender");
P
Philipp Reisner 已提交
1856

P
Philipp Reisner 已提交
1857
	/* mdev->tconn->agreed_pro_version gets initialized in drbd_connect() */
1858
	mdev->write_ordering = WO_bdev_flush;
P
Philipp Reisner 已提交
1859
	mdev->resync_wenr = LC_FREE;
1860 1861
	mdev->peer_max_bio_size = DRBD_MAX_BIO_SIZE_SAFE;
	mdev->local_max_bio_size = DRBD_MAX_BIO_SIZE_SAFE;
P
Philipp Reisner 已提交
1862 1863 1864 1865
}

void drbd_mdev_cleanup(struct drbd_conf *mdev)
{
1866
	int i;
1867
	if (mdev->tconn->receiver.t_state != NONE)
P
Philipp Reisner 已提交
1868
		dev_err(DEV, "ASSERT FAILED: receiver t_state == %d expected 0.\n",
1869
				mdev->tconn->receiver.t_state);
P
Philipp Reisner 已提交
1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882

	/* no need to lock it, I'm the only thread alive */
	if (atomic_read(&mdev->current_epoch->epoch_size) !=  0)
		dev_err(DEV, "epoch_size:%d\n", atomic_read(&mdev->current_epoch->epoch_size));
	mdev->al_writ_cnt  =
	mdev->bm_writ_cnt  =
	mdev->read_cnt     =
	mdev->recv_cnt     =
	mdev->send_cnt     =
	mdev->writ_cnt     =
	mdev->p_size       =
	mdev->rs_start     =
	mdev->rs_total     =
1883 1884
	mdev->rs_failed    = 0;
	mdev->rs_last_events = 0;
1885
	mdev->rs_last_sect_ev = 0;
1886 1887 1888 1889
	for (i = 0; i < DRBD_SYNC_MARKS; i++) {
		mdev->rs_mark_left[i] = 0;
		mdev->rs_mark_time[i] = 0;
	}
1890
	D_ASSERT(mdev->tconn->net_conf == NULL);
P
Philipp Reisner 已提交
1891 1892 1893 1894

	drbd_set_my_capacity(mdev, 0);
	if (mdev->bitmap) {
		/* maybe never allocated. */
1895
		drbd_bm_resize(mdev, 0, 1);
P
Philipp Reisner 已提交
1896 1897 1898 1899
		drbd_bm_cleanup(mdev);
	}

	drbd_free_resources(mdev);
1900
	clear_bit(AL_SUSPENDED, &mdev->flags);
P
Philipp Reisner 已提交
1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911

	/*
	 * currently we drbd_init_ee only on module load, so
	 * we may do drbd_release_ee only on module unload!
	 */
	D_ASSERT(list_empty(&mdev->active_ee));
	D_ASSERT(list_empty(&mdev->sync_ee));
	D_ASSERT(list_empty(&mdev->done_ee));
	D_ASSERT(list_empty(&mdev->read_ee));
	D_ASSERT(list_empty(&mdev->net_ee));
	D_ASSERT(list_empty(&mdev->resync_reads));
1912 1913
	D_ASSERT(list_empty(&mdev->tconn->data.work.q));
	D_ASSERT(list_empty(&mdev->tconn->meta.work.q));
P
Philipp Reisner 已提交
1914 1915
	D_ASSERT(list_empty(&mdev->resync_work.list));
	D_ASSERT(list_empty(&mdev->unplug_work.list));
1916
	D_ASSERT(list_empty(&mdev->go_diskless.list));
1917 1918

	drbd_set_defaults(mdev);
P
Philipp Reisner 已提交
1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
}


static void drbd_destroy_mempools(void)
{
	struct page *page;

	while (drbd_pp_pool) {
		page = drbd_pp_pool;
		drbd_pp_pool = (struct page *)page_private(page);
		__free_page(page);
		drbd_pp_vacant--;
	}

	/* D_ASSERT(atomic_read(&drbd_pp_vacant)==0); */

	if (drbd_ee_mempool)
		mempool_destroy(drbd_ee_mempool);
	if (drbd_request_mempool)
		mempool_destroy(drbd_request_mempool);
	if (drbd_ee_cache)
		kmem_cache_destroy(drbd_ee_cache);
	if (drbd_request_cache)
		kmem_cache_destroy(drbd_request_cache);
	if (drbd_bm_ext_cache)
		kmem_cache_destroy(drbd_bm_ext_cache);
	if (drbd_al_ext_cache)
		kmem_cache_destroy(drbd_al_ext_cache);

	drbd_ee_mempool      = NULL;
	drbd_request_mempool = NULL;
	drbd_ee_cache        = NULL;
	drbd_request_cache   = NULL;
	drbd_bm_ext_cache    = NULL;
	drbd_al_ext_cache    = NULL;

	return;
}

static int drbd_create_mempools(void)
{
	struct page *page;
1961
	const int number = (DRBD_MAX_BIO_SIZE/PAGE_SIZE) * minor_count;
P
Philipp Reisner 已提交
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978
	int i;

	/* prepare our caches and mempools */
	drbd_request_mempool = NULL;
	drbd_ee_cache        = NULL;
	drbd_request_cache   = NULL;
	drbd_bm_ext_cache    = NULL;
	drbd_al_ext_cache    = NULL;
	drbd_pp_pool         = NULL;

	/* caches */
	drbd_request_cache = kmem_cache_create(
		"drbd_req", sizeof(struct drbd_request), 0, 0, NULL);
	if (drbd_request_cache == NULL)
		goto Enomem;

	drbd_ee_cache = kmem_cache_create(
1979
		"drbd_ee", sizeof(struct drbd_peer_request), 0, 0, NULL);
P
Philipp Reisner 已提交
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
	if (drbd_ee_cache == NULL)
		goto Enomem;

	drbd_bm_ext_cache = kmem_cache_create(
		"drbd_bm", sizeof(struct bm_extent), 0, 0, NULL);
	if (drbd_bm_ext_cache == NULL)
		goto Enomem;

	drbd_al_ext_cache = kmem_cache_create(
		"drbd_al", sizeof(struct lc_element), 0, 0, NULL);
	if (drbd_al_ext_cache == NULL)
		goto Enomem;

	/* mempools */
	drbd_request_mempool = mempool_create(number,
		mempool_alloc_slab, mempool_free_slab, drbd_request_cache);
	if (drbd_request_mempool == NULL)
		goto Enomem;

	drbd_ee_mempool = mempool_create(number,
		mempool_alloc_slab, mempool_free_slab, drbd_ee_cache);
2001
	if (drbd_ee_mempool == NULL)
P
Philipp Reisner 已提交
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071
		goto Enomem;

	/* drbd's page pool */
	spin_lock_init(&drbd_pp_lock);

	for (i = 0; i < number; i++) {
		page = alloc_page(GFP_HIGHUSER);
		if (!page)
			goto Enomem;
		set_page_private(page, (unsigned long)drbd_pp_pool);
		drbd_pp_pool = page;
	}
	drbd_pp_vacant = number;

	return 0;

Enomem:
	drbd_destroy_mempools(); /* in case we allocated some */
	return -ENOMEM;
}

static int drbd_notify_sys(struct notifier_block *this, unsigned long code,
	void *unused)
{
	/* just so we have it.  you never know what interesting things we
	 * might want to do here some day...
	 */

	return NOTIFY_DONE;
}

static struct notifier_block drbd_notifier = {
	.notifier_call = drbd_notify_sys,
};

static void drbd_release_ee_lists(struct drbd_conf *mdev)
{
	int rr;

	rr = drbd_release_ee(mdev, &mdev->active_ee);
	if (rr)
		dev_err(DEV, "%d EEs in active list found!\n", rr);

	rr = drbd_release_ee(mdev, &mdev->sync_ee);
	if (rr)
		dev_err(DEV, "%d EEs in sync list found!\n", rr);

	rr = drbd_release_ee(mdev, &mdev->read_ee);
	if (rr)
		dev_err(DEV, "%d EEs in read list found!\n", rr);

	rr = drbd_release_ee(mdev, &mdev->done_ee);
	if (rr)
		dev_err(DEV, "%d EEs in done list found!\n", rr);

	rr = drbd_release_ee(mdev, &mdev->net_ee);
	if (rr)
		dev_err(DEV, "%d EEs in net list found!\n", rr);
}

/* caution. no locking.
 * currently only used from module cleanup code. */
static void drbd_delete_device(unsigned int minor)
{
	struct drbd_conf *mdev = minor_to_mdev(minor);

	if (!mdev)
		return;

	/* paranoia asserts */
2072
	D_ASSERT(mdev->open_cnt == 0);
2073
	D_ASSERT(list_empty(&mdev->tconn->data.work.q));
P
Philipp Reisner 已提交
2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084
	/* end paranoia asserts */

	del_gendisk(mdev->vdisk);

	/* cleanup stuff that may have been allocated during
	 * device (re-)configuration or state changes */

	if (mdev->this_bdev)
		bdput(mdev->this_bdev);

	drbd_free_resources(mdev);
2085
	drbd_free_tconn(mdev->tconn);
P
Philipp Reisner 已提交
2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106

	drbd_release_ee_lists(mdev);

	lc_destroy(mdev->act_log);
	lc_destroy(mdev->resync);

	kfree(mdev->p_uuid);
	/* mdev->p_uuid = NULL; */

	/* cleanup the rest that has been
	 * allocated from drbd_new_device
	 * and actually free the mdev itself */
	drbd_free_mdev(mdev);
}

static void drbd_cleanup(void)
{
	unsigned int i;

	unregister_reboot_notifier(&drbd_notifier);

2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117
	/* first remove proc,
	 * drbdsetup uses it's presence to detect
	 * whether DRBD is loaded.
	 * If we would get stuck in proc removal,
	 * but have netlink already deregistered,
	 * some drbdsetup commands may wait forever
	 * for an answer.
	 */
	if (drbd_proc)
		remove_proc_entry("drbd", NULL);

P
Philipp Reisner 已提交
2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147
	drbd_nl_cleanup();

	if (minor_table) {
		i = minor_count;
		while (i--)
			drbd_delete_device(i);
		drbd_destroy_mempools();
	}

	kfree(minor_table);

	unregister_blkdev(DRBD_MAJOR, "drbd");

	printk(KERN_INFO "drbd: module cleanup done.\n");
}

/**
 * drbd_congested() - Callback for pdflush
 * @congested_data:	User data
 * @bdi_bits:		Bits pdflush is currently interested in
 *
 * Returns 1<<BDI_async_congested and/or 1<<BDI_sync_congested if we are congested.
 */
static int drbd_congested(void *congested_data, int bdi_bits)
{
	struct drbd_conf *mdev = congested_data;
	struct request_queue *q;
	char reason = '-';
	int r = 0;

2148
	if (!may_inc_ap_bio(mdev)) {
P
Philipp Reisner 已提交
2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162
		/* DRBD has frozen IO */
		r = bdi_bits;
		reason = 'd';
		goto out;
	}

	if (get_ldev(mdev)) {
		q = bdev_get_queue(mdev->ldev->backing_bdev);
		r = bdi_congested(&q->backing_dev_info, bdi_bits);
		put_ldev(mdev);
		if (r)
			reason = 'b';
	}

2163
	if (bdi_bits & (1 << BDI_async_congested) && test_bit(NET_CONGESTED, &mdev->tconn->flags)) {
P
Philipp Reisner 已提交
2164 2165 2166 2167 2168 2169 2170 2171 2172
		r |= (1 << BDI_async_congested);
		reason = reason == 'b' ? 'a' : 'n';
	}

out:
	mdev->congestion_reason = reason;
	return r;
}

2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184
struct drbd_tconn *drbd_new_tconn(char *name)
{
	struct drbd_tconn *tconn;

	tconn = kzalloc(sizeof(struct drbd_tconn), GFP_KERNEL);
	if (!tconn)
		return NULL;

	tconn->name = kstrdup(name, GFP_KERNEL);
	if (!tconn->name)
		goto fail;

2185 2186 2187
	atomic_set(&tconn->net_cnt, 0);
	init_waitqueue_head(&tconn->net_cnt_wait);

2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207
	write_lock_irq(&global_state_lock);
	list_add(&tconn->all_tconn, &drbd_tconns);
	write_unlock_irq(&global_state_lock);

	return tconn;

fail:
	kfree(tconn->name);
	kfree(tconn);

	return NULL;
}

void drbd_free_tconn(struct drbd_tconn *tconn)
{
	write_lock_irq(&global_state_lock);
	list_del(&tconn->all_tconn);
	write_unlock_irq(&global_state_lock);

	kfree(tconn->name);
2208 2209 2210
	kfree(tconn->int_dig_out);
	kfree(tconn->int_dig_in);
	kfree(tconn->int_dig_vv);
2211 2212 2213
	kfree(tconn);
}

P
Philipp Reisner 已提交
2214 2215 2216 2217 2218
struct drbd_conf *drbd_new_device(unsigned int minor)
{
	struct drbd_conf *mdev;
	struct gendisk *disk;
	struct request_queue *q;
2219
	char conn_name[9]; /* drbd1234N */
P
Philipp Reisner 已提交
2220 2221 2222 2223 2224

	/* GFP_KERNEL, we are outside of all write-out paths */
	mdev = kzalloc(sizeof(struct drbd_conf), GFP_KERNEL);
	if (!mdev)
		return NULL;
2225 2226
	sprintf(conn_name, "drbd%d", minor);
	mdev->tconn = drbd_new_tconn(conn_name);
2227 2228 2229
	if (!mdev->tconn)
		goto out_no_tconn;

P
Philipp Reisner 已提交
2230 2231 2232
	if (!zalloc_cpumask_var(&mdev->cpu_mask, GFP_KERNEL))
		goto out_no_cpumask;

2233
	mdev->tconn->volume0 = mdev;
P
Philipp Reisner 已提交
2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248
	mdev->minor = minor;

	drbd_init_set_defaults(mdev);

	q = blk_alloc_queue(GFP_KERNEL);
	if (!q)
		goto out_no_q;
	mdev->rq_queue = q;
	q->queuedata   = mdev;

	disk = alloc_disk(1);
	if (!disk)
		goto out_no_disk;
	mdev->vdisk = disk;

2249
	set_disk_ro(disk, true);
P
Philipp Reisner 已提交
2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264

	disk->queue = q;
	disk->major = DRBD_MAJOR;
	disk->first_minor = minor;
	disk->fops = &drbd_ops;
	sprintf(disk->disk_name, "drbd%d", minor);
	disk->private_data = mdev;

	mdev->this_bdev = bdget(MKDEV(DRBD_MAJOR, minor));
	/* we have no partitions. we contain only ourselves. */
	mdev->this_bdev->bd_contains = mdev->this_bdev;

	q->backing_dev_info.congested_fn = drbd_congested;
	q->backing_dev_info.congested_data = mdev;

2265
	blk_queue_make_request(q, drbd_make_request);
2266 2267 2268
	/* Setting the max_hw_sectors to an odd value of 8kibyte here
	   This triggers a max_bio_size message upon first attach or connect */
	blk_queue_max_hw_sectors(q, DRBD_MAX_BIO_SIZE_SAFE >> 8);
P
Philipp Reisner 已提交
2269 2270
	blk_queue_bounce_limit(q, BLK_BOUNCE_ANY);
	blk_queue_merge_bvec(q, drbd_merge_bvec);
2271
	q->queue_lock = &mdev->tconn->req_lock; /* needed since we use */
P
Philipp Reisner 已提交
2272 2273 2274 2275 2276 2277 2278 2279 2280 2281

	mdev->md_io_page = alloc_page(GFP_KERNEL);
	if (!mdev->md_io_page)
		goto out_no_io_page;

	if (drbd_bm_init(mdev))
		goto out_no_bitmap;
	/* no need to lock access, we are still initializing this minor device. */
	if (!tl_init(mdev))
		goto out_no_tl;
2282
	mdev->read_requests = RB_ROOT;
2283
	mdev->write_requests = RB_ROOT;
P
Philipp Reisner 已提交
2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308

	mdev->current_epoch = kzalloc(sizeof(struct drbd_epoch), GFP_KERNEL);
	if (!mdev->current_epoch)
		goto out_no_epoch;

	INIT_LIST_HEAD(&mdev->current_epoch->list);
	mdev->epochs = 1;

	return mdev;

/* out_whatever_else:
	kfree(mdev->current_epoch); */
out_no_epoch:
	tl_cleanup(mdev);
out_no_tl:
	drbd_bm_cleanup(mdev);
out_no_bitmap:
	__free_page(mdev->md_io_page);
out_no_io_page:
	put_disk(disk);
out_no_disk:
	blk_cleanup_queue(q);
out_no_q:
	free_cpumask_var(mdev->cpu_mask);
out_no_cpumask:
2309 2310
	drbd_free_tconn(mdev->tconn);
out_no_tconn:
P
Philipp Reisner 已提交
2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334
	kfree(mdev);
	return NULL;
}

/* counterpart of drbd_new_device.
 * last part of drbd_delete_device. */
void drbd_free_mdev(struct drbd_conf *mdev)
{
	kfree(mdev->current_epoch);
	tl_cleanup(mdev);
	if (mdev->bitmap) /* should no longer be there. */
		drbd_bm_cleanup(mdev);
	__free_page(mdev->md_io_page);
	put_disk(mdev->vdisk);
	blk_cleanup_queue(mdev->rq_queue);
	free_cpumask_var(mdev->cpu_mask);
	kfree(mdev);
}


int __init drbd_init(void)
{
	int err;

P
Philipp Reisner 已提交
2335 2336
	BUILD_BUG_ON(sizeof(struct p_header80) != sizeof(struct p_header95));
	BUILD_BUG_ON(sizeof(struct p_handshake) != 80);
P
Philipp Reisner 已提交
2337

2338
	if (minor_count < DRBD_MINOR_COUNT_MIN || minor_count > DRBD_MINOR_COUNT_MAX) {
P
Philipp Reisner 已提交
2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378
		printk(KERN_ERR
			"drbd: invalid minor_count (%d)\n", minor_count);
#ifdef MODULE
		return -EINVAL;
#else
		minor_count = 8;
#endif
	}

	err = drbd_nl_init();
	if (err)
		return err;

	err = register_blkdev(DRBD_MAJOR, "drbd");
	if (err) {
		printk(KERN_ERR
		       "drbd: unable to register block device major %d\n",
		       DRBD_MAJOR);
		return err;
	}

	register_reboot_notifier(&drbd_notifier);

	/*
	 * allocate all necessary structs
	 */
	err = -ENOMEM;

	init_waitqueue_head(&drbd_pp_wait);

	drbd_proc = NULL; /* play safe for drbd_cleanup */
	minor_table = kzalloc(sizeof(struct drbd_conf *)*minor_count,
				GFP_KERNEL);
	if (!minor_table)
		goto Enomem;

	err = drbd_create_mempools();
	if (err)
		goto Enomem;

2379
	drbd_proc = proc_create_data("drbd", S_IFREG | S_IRUGO , NULL, &drbd_proc_fops, NULL);
P
Philipp Reisner 已提交
2380 2381 2382 2383 2384 2385
	if (!drbd_proc)	{
		printk(KERN_ERR "drbd: unable to register proc file\n");
		goto Enomem;
	}

	rwlock_init(&global_state_lock);
2386
	INIT_LIST_HEAD(&drbd_tconns);
P
Philipp Reisner 已提交
2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412

	printk(KERN_INFO "drbd: initialized. "
	       "Version: " REL_VERSION " (api:%d/proto:%d-%d)\n",
	       API_VERSION, PRO_VERSION_MIN, PRO_VERSION_MAX);
	printk(KERN_INFO "drbd: %s\n", drbd_buildtag());
	printk(KERN_INFO "drbd: registered as block device major %d\n",
		DRBD_MAJOR);
	printk(KERN_INFO "drbd: minor_table @ 0x%p\n", minor_table);

	return 0; /* Success! */

Enomem:
	drbd_cleanup();
	if (err == -ENOMEM)
		/* currently always the case */
		printk(KERN_ERR "drbd: ran out of memory\n");
	else
		printk(KERN_ERR "drbd: initialization failure\n");
	return err;
}

void drbd_free_bc(struct drbd_backing_dev *ldev)
{
	if (ldev == NULL)
		return;

2413 2414
	blkdev_put(ldev->backing_bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
	blkdev_put(ldev->md_bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL);
P
Philipp Reisner 已提交
2415 2416 2417 2418 2419 2420

	kfree(ldev);
}

void drbd_free_sock(struct drbd_conf *mdev)
{
2421 2422 2423 2424 2425 2426
	if (mdev->tconn->data.socket) {
		mutex_lock(&mdev->tconn->data.mutex);
		kernel_sock_shutdown(mdev->tconn->data.socket, SHUT_RDWR);
		sock_release(mdev->tconn->data.socket);
		mdev->tconn->data.socket = NULL;
		mutex_unlock(&mdev->tconn->data.mutex);
P
Philipp Reisner 已提交
2427
	}
2428 2429 2430 2431 2432 2433
	if (mdev->tconn->meta.socket) {
		mutex_lock(&mdev->tconn->meta.mutex);
		kernel_sock_shutdown(mdev->tconn->meta.socket, SHUT_RDWR);
		sock_release(mdev->tconn->meta.socket);
		mdev->tconn->meta.socket = NULL;
		mutex_unlock(&mdev->tconn->meta.mutex);
P
Philipp Reisner 已提交
2434 2435 2436 2437 2438 2439 2440 2441 2442 2443
	}
}


void drbd_free_resources(struct drbd_conf *mdev)
{
	crypto_free_hash(mdev->csums_tfm);
	mdev->csums_tfm = NULL;
	crypto_free_hash(mdev->verify_tfm);
	mdev->verify_tfm = NULL;
2444 2445 2446 2447 2448 2449
	crypto_free_hash(mdev->tconn->cram_hmac_tfm);
	mdev->tconn->cram_hmac_tfm = NULL;
	crypto_free_hash(mdev->tconn->integrity_w_tfm);
	mdev->tconn->integrity_w_tfm = NULL;
	crypto_free_hash(mdev->tconn->integrity_r_tfm);
	mdev->tconn->integrity_r_tfm = NULL;
P
Philipp Reisner 已提交
2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472

	drbd_free_sock(mdev);

	__no_warn(local,
		  drbd_free_bc(mdev->ldev);
		  mdev->ldev = NULL;);
}

/* meta data management */

struct meta_data_on_disk {
	u64 la_size;           /* last agreed size. */
	u64 uuid[UI_SIZE];   /* UUIDs. */
	u64 device_uuid;
	u64 reserved_u64_1;
	u32 flags;             /* MDF */
	u32 magic;
	u32 md_size_sect;
	u32 al_offset;         /* offset to this block */
	u32 al_nr_extents;     /* important for restoring the AL */
	      /* `-- act_log->nr_elements <-- sync_conf.al_extents */
	u32 bm_offset;         /* offset to the bitmap, from here */
	u32 bm_bytes_per_bit;  /* BM_BLOCK_SIZE */
2473 2474
	u32 la_peer_max_bio_size;   /* last peer max_bio_size */
	u32 reserved_u32[3];
P
Philipp Reisner 已提交
2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487

} __packed;

/**
 * drbd_md_sync() - Writes the meta data super block if the MD_DIRTY flag bit is set
 * @mdev:	DRBD device.
 */
void drbd_md_sync(struct drbd_conf *mdev)
{
	struct meta_data_on_disk *buffer;
	sector_t sector;
	int i;

2488 2489
	del_timer(&mdev->md_sync_timer);
	/* timer may be rearmed by drbd_md_mark_dirty() now. */
P
Philipp Reisner 已提交
2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514
	if (!test_and_clear_bit(MD_DIRTY, &mdev->flags))
		return;

	/* We use here D_FAILED and not D_ATTACHING because we try to write
	 * metadata even if we detach due to a disk failure! */
	if (!get_ldev_if_state(mdev, D_FAILED))
		return;

	mutex_lock(&mdev->md_io_mutex);
	buffer = (struct meta_data_on_disk *)page_address(mdev->md_io_page);
	memset(buffer, 0, 512);

	buffer->la_size = cpu_to_be64(drbd_get_capacity(mdev->this_bdev));
	for (i = UI_CURRENT; i < UI_SIZE; i++)
		buffer->uuid[i] = cpu_to_be64(mdev->ldev->md.uuid[i]);
	buffer->flags = cpu_to_be32(mdev->ldev->md.flags);
	buffer->magic = cpu_to_be32(DRBD_MD_MAGIC);

	buffer->md_size_sect  = cpu_to_be32(mdev->ldev->md.md_size_sect);
	buffer->al_offset     = cpu_to_be32(mdev->ldev->md.al_offset);
	buffer->al_nr_extents = cpu_to_be32(mdev->act_log->nr_elements);
	buffer->bm_bytes_per_bit = cpu_to_be32(BM_BLOCK_SIZE);
	buffer->device_uuid = cpu_to_be64(mdev->ldev->md.device_uuid);

	buffer->bm_offset = cpu_to_be32(mdev->ldev->md.bm_offset);
2515
	buffer->la_peer_max_bio_size = cpu_to_be32(mdev->peer_max_bio_size);
P
Philipp Reisner 已提交
2516 2517 2518 2519

	D_ASSERT(drbd_md_ss__(mdev, mdev->ldev) == mdev->ldev->md.md_offset);
	sector = mdev->ldev->md.md_offset;

2520
	if (!drbd_md_sync_page_io(mdev, mdev->ldev, sector, WRITE)) {
P
Philipp Reisner 已提交
2521 2522
		/* this was a try anyways ... */
		dev_err(DEV, "meta data update failed!\n");
2523
		drbd_chk_io_error(mdev, 1, true);
P
Philipp Reisner 已提交
2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538
	}

	/* Update mdev->ldev->md.la_size_sect,
	 * since we updated it on metadata. */
	mdev->ldev->md.la_size_sect = drbd_get_capacity(mdev->this_bdev);

	mutex_unlock(&mdev->md_io_mutex);
	put_ldev(mdev);
}

/**
 * drbd_md_read() - Reads in the meta data super block
 * @mdev:	DRBD device.
 * @bdev:	Device from which the meta data should be read in.
 *
2539
 * Return 0 (NO_ERROR) on success, and an enum drbd_ret_code in case
P
Philipp Reisner 已提交
2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553
 * something goes wrong.  Currently only: ERR_IO_MD_DISK, ERR_MD_INVALID.
 */
int drbd_md_read(struct drbd_conf *mdev, struct drbd_backing_dev *bdev)
{
	struct meta_data_on_disk *buffer;
	int i, rv = NO_ERROR;

	if (!get_ldev_if_state(mdev, D_ATTACHING))
		return ERR_IO_MD_DISK;

	mutex_lock(&mdev->md_io_mutex);
	buffer = (struct meta_data_on_disk *)page_address(mdev->md_io_page);

	if (!drbd_md_sync_page_io(mdev, bdev, bdev->md.md_offset, READ)) {
L
Lucas De Marchi 已提交
2554
		/* NOTE: can't do normal error processing here as this is
P
Philipp Reisner 已提交
2555 2556 2557 2558 2559 2560
		   called BEFORE disk is attached */
		dev_err(DEV, "Error while reading metadata.\n");
		rv = ERR_IO_MD_DISK;
		goto err;
	}

2561
	if (buffer->magic != cpu_to_be32(DRBD_MD_MAGIC)) {
P
Philipp Reisner 已提交
2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598
		dev_err(DEV, "Error while reading metadata, magic not found.\n");
		rv = ERR_MD_INVALID;
		goto err;
	}
	if (be32_to_cpu(buffer->al_offset) != bdev->md.al_offset) {
		dev_err(DEV, "unexpected al_offset: %d (expected %d)\n",
		    be32_to_cpu(buffer->al_offset), bdev->md.al_offset);
		rv = ERR_MD_INVALID;
		goto err;
	}
	if (be32_to_cpu(buffer->bm_offset) != bdev->md.bm_offset) {
		dev_err(DEV, "unexpected bm_offset: %d (expected %d)\n",
		    be32_to_cpu(buffer->bm_offset), bdev->md.bm_offset);
		rv = ERR_MD_INVALID;
		goto err;
	}
	if (be32_to_cpu(buffer->md_size_sect) != bdev->md.md_size_sect) {
		dev_err(DEV, "unexpected md_size: %u (expected %u)\n",
		    be32_to_cpu(buffer->md_size_sect), bdev->md.md_size_sect);
		rv = ERR_MD_INVALID;
		goto err;
	}

	if (be32_to_cpu(buffer->bm_bytes_per_bit) != BM_BLOCK_SIZE) {
		dev_err(DEV, "unexpected bm_bytes_per_bit: %u (expected %u)\n",
		    be32_to_cpu(buffer->bm_bytes_per_bit), BM_BLOCK_SIZE);
		rv = ERR_MD_INVALID;
		goto err;
	}

	bdev->md.la_size_sect = be64_to_cpu(buffer->la_size);
	for (i = UI_CURRENT; i < UI_SIZE; i++)
		bdev->md.uuid[i] = be64_to_cpu(buffer->uuid[i]);
	bdev->md.flags = be32_to_cpu(buffer->flags);
	mdev->sync_conf.al_extents = be32_to_cpu(buffer->al_nr_extents);
	bdev->md.device_uuid = be64_to_cpu(buffer->device_uuid);

2599
	spin_lock_irq(&mdev->tconn->req_lock);
2600 2601 2602 2603 2604 2605
	if (mdev->state.conn < C_CONNECTED) {
		int peer;
		peer = be32_to_cpu(buffer->la_peer_max_bio_size);
		peer = max_t(int, peer, DRBD_MAX_BIO_SIZE_SAFE);
		mdev->peer_max_bio_size = peer;
	}
2606
	spin_unlock_irq(&mdev->tconn->req_lock);
2607

P
Philipp Reisner 已提交
2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
	if (mdev->sync_conf.al_extents < 7)
		mdev->sync_conf.al_extents = 127;

 err:
	mutex_unlock(&mdev->md_io_mutex);
	put_ldev(mdev);

	return rv;
}

/**
 * drbd_md_mark_dirty() - Mark meta data super block as dirty
 * @mdev:	DRBD device.
 *
 * Call this function if you change anything that should be written to
 * the meta-data super block. This function sets MD_DIRTY, and starts a
 * timer that ensures that within five seconds you have to call drbd_md_sync().
 */
2626
#ifdef DEBUG
2627 2628 2629 2630 2631 2632 2633 2634 2635
void drbd_md_mark_dirty_(struct drbd_conf *mdev, unsigned int line, const char *func)
{
	if (!test_and_set_bit(MD_DIRTY, &mdev->flags)) {
		mod_timer(&mdev->md_sync_timer, jiffies + HZ);
		mdev->last_md_mark_dirty.line = line;
		mdev->last_md_mark_dirty.func = func;
	}
}
#else
P
Philipp Reisner 已提交
2636 2637
void drbd_md_mark_dirty(struct drbd_conf *mdev)
{
2638
	if (!test_and_set_bit(MD_DIRTY, &mdev->flags))
2639
		mod_timer(&mdev->md_sync_timer, jiffies + 5*HZ);
P
Philipp Reisner 已提交
2640
}
2641
#endif
P
Philipp Reisner 已提交
2642 2643 2644 2645 2646

static void drbd_uuid_move_history(struct drbd_conf *mdev) __must_hold(local)
{
	int i;

2647
	for (i = UI_HISTORY_START; i < UI_HISTORY_END; i++)
P
Philipp Reisner 已提交
2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685
		mdev->ldev->md.uuid[i+1] = mdev->ldev->md.uuid[i];
}

void _drbd_uuid_set(struct drbd_conf *mdev, int idx, u64 val) __must_hold(local)
{
	if (idx == UI_CURRENT) {
		if (mdev->state.role == R_PRIMARY)
			val |= 1;
		else
			val &= ~((u64)1);

		drbd_set_ed_uuid(mdev, val);
	}

	mdev->ldev->md.uuid[idx] = val;
	drbd_md_mark_dirty(mdev);
}


void drbd_uuid_set(struct drbd_conf *mdev, int idx, u64 val) __must_hold(local)
{
	if (mdev->ldev->md.uuid[idx]) {
		drbd_uuid_move_history(mdev);
		mdev->ldev->md.uuid[UI_HISTORY_START] = mdev->ldev->md.uuid[idx];
	}
	_drbd_uuid_set(mdev, idx, val);
}

/**
 * drbd_uuid_new_current() - Creates a new current UUID
 * @mdev:	DRBD device.
 *
 * Creates a new current UUID, and rotates the old current UUID into
 * the bitmap slot. Causes an incremental resync upon next connect.
 */
void drbd_uuid_new_current(struct drbd_conf *mdev) __must_hold(local)
{
	u64 val;
2686 2687 2688 2689
	unsigned long long bm_uuid = mdev->ldev->md.uuid[UI_BITMAP];

	if (bm_uuid)
		dev_warn(DEV, "bm UUID was already set: %llX\n", bm_uuid);
P
Philipp Reisner 已提交
2690 2691 2692 2693 2694

	mdev->ldev->md.uuid[UI_BITMAP] = mdev->ldev->md.uuid[UI_CURRENT];

	get_random_bytes(&val, sizeof(u64));
	_drbd_uuid_set(mdev, UI_CURRENT, val);
2695
	drbd_print_uuids(mdev, "new current UUID");
2696 2697
	/* get it to stable storage _now_ */
	drbd_md_sync(mdev);
P
Philipp Reisner 已提交
2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709
}

void drbd_uuid_set_bm(struct drbd_conf *mdev, u64 val) __must_hold(local)
{
	if (mdev->ldev->md.uuid[UI_BITMAP] == 0 && val == 0)
		return;

	if (val == 0) {
		drbd_uuid_move_history(mdev);
		mdev->ldev->md.uuid[UI_HISTORY_START] = mdev->ldev->md.uuid[UI_BITMAP];
		mdev->ldev->md.uuid[UI_BITMAP] = 0;
	} else {
2710 2711 2712
		unsigned long long bm_uuid = mdev->ldev->md.uuid[UI_BITMAP];
		if (bm_uuid)
			dev_warn(DEV, "bm UUID was already set: %llX\n", bm_uuid);
P
Philipp Reisner 已提交
2713

2714
		mdev->ldev->md.uuid[UI_BITMAP] = val & ~((u64)1);
P
Philipp Reisner 已提交
2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756
	}
	drbd_md_mark_dirty(mdev);
}

/**
 * drbd_bmio_set_n_write() - io_fn for drbd_queue_bitmap_io() or drbd_bitmap_io()
 * @mdev:	DRBD device.
 *
 * Sets all bits in the bitmap and writes the whole bitmap to stable storage.
 */
int drbd_bmio_set_n_write(struct drbd_conf *mdev)
{
	int rv = -EIO;

	if (get_ldev_if_state(mdev, D_ATTACHING)) {
		drbd_md_set_flag(mdev, MDF_FULL_SYNC);
		drbd_md_sync(mdev);
		drbd_bm_set_all(mdev);

		rv = drbd_bm_write(mdev);

		if (!rv) {
			drbd_md_clear_flag(mdev, MDF_FULL_SYNC);
			drbd_md_sync(mdev);
		}

		put_ldev(mdev);
	}

	return rv;
}

/**
 * drbd_bmio_clear_n_write() - io_fn for drbd_queue_bitmap_io() or drbd_bitmap_io()
 * @mdev:	DRBD device.
 *
 * Clears all bits in the bitmap and writes the whole bitmap to stable storage.
 */
int drbd_bmio_clear_n_write(struct drbd_conf *mdev)
{
	int rv = -EIO;

2757
	drbd_resume_al(mdev);
P
Philipp Reisner 已提交
2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769
	if (get_ldev_if_state(mdev, D_ATTACHING)) {
		drbd_bm_clear_all(mdev);
		rv = drbd_bm_write(mdev);
		put_ldev(mdev);
	}

	return rv;
}

static int w_bitmap_io(struct drbd_conf *mdev, struct drbd_work *w, int unused)
{
	struct bm_io_work *work = container_of(w, struct bm_io_work, w);
2770
	int rv = -EIO;
P
Philipp Reisner 已提交
2771 2772 2773

	D_ASSERT(atomic_read(&mdev->ap_bio_cnt) == 0);

2774
	if (get_ldev(mdev)) {
2775
		drbd_bm_lock(mdev, work->why, work->flags);
2776 2777 2778 2779
		rv = work->io_fn(mdev);
		drbd_bm_unlock(mdev);
		put_ldev(mdev);
	}
P
Philipp Reisner 已提交
2780 2781

	clear_bit(BITMAP_IO, &mdev->flags);
2782
	smp_mb__after_clear_bit();
P
Philipp Reisner 已提交
2783 2784 2785 2786 2787 2788 2789
	wake_up(&mdev->misc_wait);

	if (work->done)
		work->done(mdev, rv);

	clear_bit(BITMAP_IO_QUEUED, &mdev->flags);
	work->why = NULL;
2790
	work->flags = 0;
P
Philipp Reisner 已提交
2791 2792 2793 2794

	return 1;
}

2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811
void drbd_ldev_destroy(struct drbd_conf *mdev)
{
	lc_destroy(mdev->resync);
	mdev->resync = NULL;
	lc_destroy(mdev->act_log);
	mdev->act_log = NULL;
	__no_warn(local,
		drbd_free_bc(mdev->ldev);
		mdev->ldev = NULL;);

	if (mdev->md_io_tmpp) {
		__free_page(mdev->md_io_tmpp);
		mdev->md_io_tmpp = NULL;
	}
	clear_bit(GO_DISKLESS, &mdev->flags);
}

2812 2813 2814
static int w_go_diskless(struct drbd_conf *mdev, struct drbd_work *w, int unused)
{
	D_ASSERT(mdev->state.disk == D_FAILED);
2815 2816
	/* we cannot assert local_cnt == 0 here, as get_ldev_if_state will
	 * inc/dec it frequently. Once we are D_DISKLESS, no one will touch
2817 2818
	 * the protected members anymore, though, so once put_ldev reaches zero
	 * again, it will be safe to free them. */
2819 2820 2821 2822 2823 2824 2825 2826
	drbd_force_state(mdev, NS(disk, D_DISKLESS));
	return 1;
}

void drbd_go_diskless(struct drbd_conf *mdev)
{
	D_ASSERT(mdev->state.disk == D_FAILED);
	if (!test_and_set_bit(GO_DISKLESS, &mdev->flags))
2827
		drbd_queue_work(&mdev->tconn->data.work, &mdev->go_diskless);
2828 2829
}

P
Philipp Reisner 已提交
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844
/**
 * drbd_queue_bitmap_io() - Queues an IO operation on the whole bitmap
 * @mdev:	DRBD device.
 * @io_fn:	IO callback to be called when bitmap IO is possible
 * @done:	callback to be called after the bitmap IO was performed
 * @why:	Descriptive text of the reason for doing the IO
 *
 * While IO on the bitmap happens we freeze application IO thus we ensure
 * that drbd_set_out_of_sync() can not be called. This function MAY ONLY be
 * called from worker context. It MUST NOT be used while a previous such
 * work is still pending!
 */
void drbd_queue_bitmap_io(struct drbd_conf *mdev,
			  int (*io_fn)(struct drbd_conf *),
			  void (*done)(struct drbd_conf *, int),
2845
			  char *why, enum bm_flag flags)
P
Philipp Reisner 已提交
2846
{
2847
	D_ASSERT(current == mdev->tconn->worker.task);
P
Philipp Reisner 已提交
2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858

	D_ASSERT(!test_bit(BITMAP_IO_QUEUED, &mdev->flags));
	D_ASSERT(!test_bit(BITMAP_IO, &mdev->flags));
	D_ASSERT(list_empty(&mdev->bm_io_work.w.list));
	if (mdev->bm_io_work.why)
		dev_err(DEV, "FIXME going to queue '%s' but '%s' still pending?\n",
			why, mdev->bm_io_work.why);

	mdev->bm_io_work.io_fn = io_fn;
	mdev->bm_io_work.done = done;
	mdev->bm_io_work.why = why;
2859
	mdev->bm_io_work.flags = flags;
P
Philipp Reisner 已提交
2860

2861
	spin_lock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
2862 2863
	set_bit(BITMAP_IO, &mdev->flags);
	if (atomic_read(&mdev->ap_bio_cnt) == 0) {
2864
		if (!test_and_set_bit(BITMAP_IO_QUEUED, &mdev->flags))
2865
			drbd_queue_work(&mdev->tconn->data.work, &mdev->bm_io_work.w);
P
Philipp Reisner 已提交
2866
	}
2867
	spin_unlock_irq(&mdev->tconn->req_lock);
P
Philipp Reisner 已提交
2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878
}

/**
 * drbd_bitmap_io() -  Does an IO operation on the whole bitmap
 * @mdev:	DRBD device.
 * @io_fn:	IO callback to be called when bitmap IO is possible
 * @why:	Descriptive text of the reason for doing the IO
 *
 * freezes application IO while that the actual IO operations runs. This
 * functions MAY NOT be called from worker context.
 */
2879 2880
int drbd_bitmap_io(struct drbd_conf *mdev, int (*io_fn)(struct drbd_conf *),
		char *why, enum bm_flag flags)
P
Philipp Reisner 已提交
2881 2882 2883
{
	int rv;

2884
	D_ASSERT(current != mdev->tconn->worker.task);
P
Philipp Reisner 已提交
2885

2886 2887
	if ((flags & BM_LOCKED_SET_ALLOWED) == 0)
		drbd_suspend_io(mdev);
P
Philipp Reisner 已提交
2888

2889
	drbd_bm_lock(mdev, why, flags);
P
Philipp Reisner 已提交
2890 2891 2892
	rv = io_fn(mdev);
	drbd_bm_unlock(mdev);

2893 2894
	if ((flags & BM_LOCKED_SET_ALLOWED) == 0)
		drbd_resume_io(mdev);
P
Philipp Reisner 已提交
2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922

	return rv;
}

void drbd_md_set_flag(struct drbd_conf *mdev, int flag) __must_hold(local)
{
	if ((mdev->ldev->md.flags & flag) != flag) {
		drbd_md_mark_dirty(mdev);
		mdev->ldev->md.flags |= flag;
	}
}

void drbd_md_clear_flag(struct drbd_conf *mdev, int flag) __must_hold(local)
{
	if ((mdev->ldev->md.flags & flag) != 0) {
		drbd_md_mark_dirty(mdev);
		mdev->ldev->md.flags &= ~flag;
	}
}
int drbd_md_test_flag(struct drbd_backing_dev *bdev, int flag)
{
	return (bdev->md.flags & flag) != 0;
}

static void md_sync_timer_fn(unsigned long data)
{
	struct drbd_conf *mdev = (struct drbd_conf *) data;

2923
	drbd_queue_work_front(&mdev->tconn->data.work, &mdev->md_sync_work);
P
Philipp Reisner 已提交
2924 2925 2926 2927 2928
}

static int w_md_sync(struct drbd_conf *mdev, struct drbd_work *w, int unused)
{
	dev_warn(DEV, "md_sync_timer expired! Worker calls drbd_md_sync().\n");
2929 2930 2931 2932
#ifdef DEBUG
	dev_warn(DEV, "last md_mark_dirty: %s:%u\n",
		mdev->last_md_mark_dirty.func, mdev->last_md_mark_dirty.line);
#endif
P
Philipp Reisner 已提交
2933 2934 2935 2936
	drbd_md_sync(mdev);
	return 1;
}

2937
const char *cmdname(enum drbd_packet cmd)
2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995
{
	/* THINK may need to become several global tables
	 * when we want to support more than
	 * one PRO_VERSION */
	static const char *cmdnames[] = {
		[P_DATA]	        = "Data",
		[P_DATA_REPLY]	        = "DataReply",
		[P_RS_DATA_REPLY]	= "RSDataReply",
		[P_BARRIER]	        = "Barrier",
		[P_BITMAP]	        = "ReportBitMap",
		[P_BECOME_SYNC_TARGET]  = "BecomeSyncTarget",
		[P_BECOME_SYNC_SOURCE]  = "BecomeSyncSource",
		[P_UNPLUG_REMOTE]	= "UnplugRemote",
		[P_DATA_REQUEST]	= "DataRequest",
		[P_RS_DATA_REQUEST]     = "RSDataRequest",
		[P_SYNC_PARAM]	        = "SyncParam",
		[P_SYNC_PARAM89]	= "SyncParam89",
		[P_PROTOCOL]            = "ReportProtocol",
		[P_UUIDS]	        = "ReportUUIDs",
		[P_SIZES]	        = "ReportSizes",
		[P_STATE]	        = "ReportState",
		[P_SYNC_UUID]           = "ReportSyncUUID",
		[P_AUTH_CHALLENGE]      = "AuthChallenge",
		[P_AUTH_RESPONSE]	= "AuthResponse",
		[P_PING]		= "Ping",
		[P_PING_ACK]	        = "PingAck",
		[P_RECV_ACK]	        = "RecvAck",
		[P_WRITE_ACK]	        = "WriteAck",
		[P_RS_WRITE_ACK]	= "RSWriteAck",
		[P_DISCARD_ACK]	        = "DiscardAck",
		[P_NEG_ACK]	        = "NegAck",
		[P_NEG_DREPLY]	        = "NegDReply",
		[P_NEG_RS_DREPLY]	= "NegRSDReply",
		[P_BARRIER_ACK]	        = "BarrierAck",
		[P_STATE_CHG_REQ]       = "StateChgRequest",
		[P_STATE_CHG_REPLY]     = "StateChgReply",
		[P_OV_REQUEST]          = "OVRequest",
		[P_OV_REPLY]            = "OVReply",
		[P_OV_RESULT]           = "OVResult",
		[P_CSUM_RS_REQUEST]     = "CsumRSRequest",
		[P_RS_IS_IN_SYNC]	= "CsumRSIsInSync",
		[P_COMPRESSED_BITMAP]   = "CBitmap",
		[P_DELAY_PROBE]         = "DelayProbe",
		[P_OUT_OF_SYNC]		= "OutOfSync",
		[P_MAX_CMD]	        = NULL,
	};

	if (cmd == P_HAND_SHAKE_M)
		return "HandShakeM";
	if (cmd == P_HAND_SHAKE_S)
		return "HandShakeS";
	if (cmd == P_HAND_SHAKE)
		return "HandShake";
	if (cmd >= P_MAX_CMD)
		return "Unknown";
	return cmdnames[cmd];
}

P
Philipp Reisner 已提交
2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016
#ifdef CONFIG_DRBD_FAULT_INJECTION
/* Fault insertion support including random number generator shamelessly
 * stolen from kernel/rcutorture.c */
struct fault_random_state {
	unsigned long state;
	unsigned long count;
};

#define FAULT_RANDOM_MULT 39916801  /* prime */
#define FAULT_RANDOM_ADD	479001701 /* prime */
#define FAULT_RANDOM_REFRESH 10000

/*
 * Crude but fast random-number generator.  Uses a linear congruential
 * generator, with occasional help from get_random_bytes().
 */
static unsigned long
_drbd_fault_random(struct fault_random_state *rsp)
{
	long refresh;

3017
	if (!rsp->count--) {
P
Philipp Reisner 已提交
3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036
		get_random_bytes(&refresh, sizeof(refresh));
		rsp->state += refresh;
		rsp->count = FAULT_RANDOM_REFRESH;
	}
	rsp->state = rsp->state * FAULT_RANDOM_MULT + FAULT_RANDOM_ADD;
	return swahw32(rsp->state);
}

static char *
_drbd_fault_str(unsigned int type) {
	static char *_faults[] = {
		[DRBD_FAULT_MD_WR] = "Meta-data write",
		[DRBD_FAULT_MD_RD] = "Meta-data read",
		[DRBD_FAULT_RS_WR] = "Resync write",
		[DRBD_FAULT_RS_RD] = "Resync read",
		[DRBD_FAULT_DT_WR] = "Data write",
		[DRBD_FAULT_DT_RD] = "Data read",
		[DRBD_FAULT_DT_RA] = "Data read ahead",
		[DRBD_FAULT_BM_ALLOC] = "BM allocation",
3037 3038
		[DRBD_FAULT_AL_EE] = "EE allocation",
		[DRBD_FAULT_RECEIVE] = "receive data corruption",
P
Philipp Reisner 已提交
3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056
	};

	return (type < DRBD_FAULT_MAX) ? _faults[type] : "**Unknown**";
}

unsigned int
_drbd_insert_fault(struct drbd_conf *mdev, unsigned int type)
{
	static struct fault_random_state rrs = {0, 0};

	unsigned int ret = (
		(fault_devs == 0 ||
			((1 << mdev_to_minor(mdev)) & fault_devs) != 0) &&
		(((_drbd_fault_random(&rrs) % 100) + 1) <= fault_rate));

	if (ret) {
		fault_count++;

3057
		if (__ratelimit(&drbd_ratelimit_state))
P
Philipp Reisner 已提交
3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091
			dev_warn(DEV, "***Simulating %s failure\n",
				_drbd_fault_str(type));
	}

	return ret;
}
#endif

const char *drbd_buildtag(void)
{
	/* DRBD built from external sources has here a reference to the
	   git hash of the source code. */

	static char buildtag[38] = "\0uilt-in";

	if (buildtag[0] == 0) {
#ifdef CONFIG_MODULES
		if (THIS_MODULE != NULL)
			sprintf(buildtag, "srcversion: %-24s", THIS_MODULE->srcversion);
		else
#endif
			buildtag[0] = 'b';
	}

	return buildtag;
}

module_init(drbd_init)
module_exit(drbd_cleanup)

EXPORT_SYMBOL(drbd_conn_str);
EXPORT_SYMBOL(drbd_role_str);
EXPORT_SYMBOL(drbd_disk_str);
EXPORT_SYMBOL(drbd_set_st_err_str);