fs-writeback.c 70.5 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8 9 10
/*
 * fs/fs-writeback.c
 *
 * Copyright (C) 2002, Linus Torvalds.
 *
 * Contains all the functions related to writing back and waiting
 * upon dirty inodes against superblocks, and writing back dirty
 * pages against inodes.  ie: data writeback.  Writeout of the
 * inode itself is not handled here.
 *
11
 * 10Apr2002	Andrew Morton
Linus Torvalds's avatar
Linus Torvalds committed
12 13 14 15 16
 *		Split out of fs/inode.c
 *		Additions for address_space-based writeback
 */

#include <linux/kernel.h>
17
#include <linux/export.h>
Linus Torvalds's avatar
Linus Torvalds committed
18
#include <linux/spinlock.h>
19
#include <linux/slab.h>
Linus Torvalds's avatar
Linus Torvalds committed
20 21 22
#include <linux/sched.h>
#include <linux/fs.h>
#include <linux/mm.h>
23
#include <linux/pagemap.h>
24
#include <linux/kthread.h>
Linus Torvalds's avatar
Linus Torvalds committed
25 26 27
#include <linux/writeback.h>
#include <linux/blkdev.h>
#include <linux/backing-dev.h>
28
#include <linux/tracepoint.h>
29
#include <linux/device.h>
30
#include <linux/memcontrol.h>
31
#include "internal.h"
Linus Torvalds's avatar
Linus Torvalds committed
32

33 34 35
/*
 * 4MB minimal write chunk size
 */
36
#define MIN_WRITEBACK_PAGES	(4096UL >> (PAGE_SHIFT - 10))
37

38 39 40 41
struct wb_completion {
	atomic_t		cnt;
};

42 43 44
/*
 * Passed into wb_writeback(), essentially a subset of writeback_control
 */
45
struct wb_writeback_work {
46 47
	long nr_pages;
	struct super_block *sb;
48
	unsigned long *older_than_this;
49
	enum writeback_sync_modes sync_mode;
50
	unsigned int tagged_writepages:1;
51 52 53
	unsigned int for_kupdate:1;
	unsigned int range_cyclic:1;
	unsigned int for_background:1;
54
	unsigned int for_sync:1;	/* sync(2) WB_SYNC_ALL writeback */
55
	unsigned int auto_free:1;	/* free on completion */
56
	enum wb_reason reason;		/* why was writeback initiated? */
57

58
	struct list_head list;		/* pending work list */
59
	struct wb_completion *done;	/* set if the caller waits */
60 61
};

62 63 64 65 66 67 68 69 70 71 72 73 74
/*
 * If one wants to wait for one or more wb_writeback_works, each work's
 * ->done should be set to a wb_completion defined using the following
 * macro.  Once all work items are issued with wb_queue_work(), the caller
 * can wait for the completion of all using wb_wait_for_completion().  Work
 * items which are waited upon aren't freed automatically on completion.
 */
#define DEFINE_WB_COMPLETION_ONSTACK(cmpl)				\
	struct wb_completion cmpl = {					\
		.cnt		= ATOMIC_INIT(1),			\
	}


75 76 77 78 79 80 81 82 83 84 85 86
/*
 * If an inode is constantly having its pages dirtied, but then the
 * updates stop dirtytime_expire_interval seconds in the past, it's
 * possible for the worst case time between when an inode has its
 * timestamps updated and when they finally get written out to be two
 * dirtytime_expire_intervals.  We set the default to 12 hours (in
 * seconds), which means most of the time inodes will have their
 * timestamps written to disk after 12 hours, but in the worst case a
 * few inodes might not their timestamps updated for 24 hours.
 */
unsigned int dirtytime_expire_interval = 12 * 60 * 60;

87 88
static inline struct inode *wb_inode(struct list_head *head)
{
89
	return list_entry(head, struct inode, i_io_list);
90 91
}

92 93 94 95 96 97 98 99
/*
 * Include the creation of the trace points after defining the
 * wb_writeback_work structure and inline functions so that the definition
 * remains local to this file.
 */
#define CREATE_TRACE_POINTS
#include <trace/events/writeback.h>

100 101
EXPORT_TRACEPOINT_SYMBOL_GPL(wbc_writepage);

102 103 104 105 106 107
static bool wb_io_lists_populated(struct bdi_writeback *wb)
{
	if (wb_has_dirty_io(wb)) {
		return false;
	} else {
		set_bit(WB_has_dirty_io, &wb->state);
108
		WARN_ON_ONCE(!wb->avg_write_bandwidth);
109 110
		atomic_long_add(wb->avg_write_bandwidth,
				&wb->bdi->tot_write_bandwidth);
111 112 113 114 115 116 117
		return true;
	}
}

static void wb_io_lists_depopulated(struct bdi_writeback *wb)
{
	if (wb_has_dirty_io(wb) && list_empty(&wb->b_dirty) &&
118
	    list_empty(&wb->b_io) && list_empty(&wb->b_more_io)) {
119
		clear_bit(WB_has_dirty_io, &wb->state);
120 121
		WARN_ON_ONCE(atomic_long_sub_return(wb->avg_write_bandwidth,
					&wb->bdi->tot_write_bandwidth) < 0);
122
	}
123 124 125
}

/**
126
 * inode_io_list_move_locked - move an inode onto a bdi_writeback IO list
127 128 129 130
 * @inode: inode to be moved
 * @wb: target bdi_writeback
 * @head: one of @wb->b_{dirty|io|more_io}
 *
131
 * Move @inode->i_io_list to @list of @wb and set %WB_has_dirty_io.
132 133 134
 * Returns %true if @inode is the first occupant of the !dirty_time IO
 * lists; otherwise, %false.
 */
135
static bool inode_io_list_move_locked(struct inode *inode,
136 137 138 139 140
				      struct bdi_writeback *wb,
				      struct list_head *head)
{
	assert_spin_locked(&wb->list_lock);

141
	list_move(&inode->i_io_list, head);
142 143 144 145 146 147 148 149 150 151

	/* dirty_time doesn't count as dirty_io until expiration */
	if (head != &wb->b_dirty_time)
		return wb_io_lists_populated(wb);

	wb_io_lists_depopulated(wb);
	return false;
}

/**
152
 * inode_io_list_del_locked - remove an inode from its bdi_writeback IO list
153 154 155 156 157 158
 * @inode: inode to be removed
 * @wb: bdi_writeback @inode is being removed from
 *
 * Remove @inode which may be on one of @wb->b_{dirty|io|more_io} lists and
 * clear %WB_has_dirty_io if all are empty afterwards.
 */
159
static void inode_io_list_del_locked(struct inode *inode,
160 161 162 163
				     struct bdi_writeback *wb)
{
	assert_spin_locked(&wb->list_lock);

164
	list_del_init(&inode->i_io_list);
165 166 167
	wb_io_lists_depopulated(wb);
}

168
static void wb_wakeup(struct bdi_writeback *wb)
169
{
170 171 172 173
	spin_lock_bh(&wb->work_lock);
	if (test_bit(WB_registered, &wb->state))
		mod_delayed_work(bdi_wq, &wb->dwork, 0);
	spin_unlock_bh(&wb->work_lock);
174 175
}

176 177 178 179 180 181 182 183 184 185 186
static void finish_writeback_work(struct bdi_writeback *wb,
				  struct wb_writeback_work *work)
{
	struct wb_completion *done = work->done;

	if (work->auto_free)
		kfree(work);
	if (done && atomic_dec_and_test(&done->cnt))
		wake_up_all(&wb->bdi->wb_waitq);
}

187 188
static void wb_queue_work(struct bdi_writeback *wb,
			  struct wb_writeback_work *work)
189
{
190
	trace_writeback_queue(wb, work);
191

192 193
	if (work->done)
		atomic_inc(&work->done->cnt);
194 195 196 197 198 199 200 201 202

	spin_lock_bh(&wb->work_lock);

	if (test_bit(WB_registered, &wb->state)) {
		list_add_tail(&work->list, &wb->work_list);
		mod_delayed_work(bdi_wq, &wb->dwork, 0);
	} else
		finish_writeback_work(wb, work);

203
	spin_unlock_bh(&wb->work_lock);
Linus Torvalds's avatar
Linus Torvalds committed
204 205
}

206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223
/**
 * wb_wait_for_completion - wait for completion of bdi_writeback_works
 * @bdi: bdi work items were issued to
 * @done: target wb_completion
 *
 * Wait for one or more work items issued to @bdi with their ->done field
 * set to @done, which should have been defined with
 * DEFINE_WB_COMPLETION_ONSTACK().  This function returns after all such
 * work items are completed.  Work items which are waited upon aren't freed
 * automatically on completion.
 */
static void wb_wait_for_completion(struct backing_dev_info *bdi,
				   struct wb_completion *done)
{
	atomic_dec(&done->cnt);		/* put down the initial count */
	wait_event(bdi->wb_waitq, !atomic_read(&done->cnt));
}

224 225
#ifdef CONFIG_CGROUP_WRITEBACK

226 227 228 229 230 231 232 233 234 235 236 237 238 239
/* parameters for foreign inode detection, see wb_detach_inode() */
#define WB_FRN_TIME_SHIFT	13	/* 1s = 2^13, upto 8 secs w/ 16bit */
#define WB_FRN_TIME_AVG_SHIFT	3	/* avg = avg * 7/8 + new * 1/8 */
#define WB_FRN_TIME_CUT_DIV	2	/* ignore rounds < avg / 2 */
#define WB_FRN_TIME_PERIOD	(2 * (1 << WB_FRN_TIME_SHIFT))	/* 2s */

#define WB_FRN_HIST_SLOTS	16	/* inode->i_wb_frn_history is 16bit */
#define WB_FRN_HIST_UNIT	(WB_FRN_TIME_PERIOD / WB_FRN_HIST_SLOTS)
					/* each slot's duration is 2s / 16 */
#define WB_FRN_HIST_THR_SLOTS	(WB_FRN_HIST_SLOTS / 2)
					/* if foreign slots >= 8, switch */
#define WB_FRN_HIST_MAX_SLOTS	(WB_FRN_HIST_THR_SLOTS / 2 + 1)
					/* one round can affect upto 5 slots */

240 241 242
static atomic_t isw_nr_in_flight = ATOMIC_INIT(0);
static struct workqueue_struct *isw_wq;

243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
void __inode_attach_wb(struct inode *inode, struct page *page)
{
	struct backing_dev_info *bdi = inode_to_bdi(inode);
	struct bdi_writeback *wb = NULL;

	if (inode_cgwb_enabled(inode)) {
		struct cgroup_subsys_state *memcg_css;

		if (page) {
			memcg_css = mem_cgroup_css_from_page(page);
			wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC);
		} else {
			/* must pin memcg_css, see wb_get_create() */
			memcg_css = task_get_css(current, memory_cgrp_id);
			wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC);
			css_put(memcg_css);
		}
	}

	if (!wb)
		wb = &bdi->wb;

	/*
	 * There may be multiple instances of this function racing to
	 * update the same inode.  Use cmpxchg() to tell the winner.
	 */
	if (unlikely(cmpxchg(&inode->i_wb, NULL, wb)))
		wb_put(wb);
}

273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298
/**
 * locked_inode_to_wb_and_lock_list - determine a locked inode's wb and lock it
 * @inode: inode of interest with i_lock held
 *
 * Returns @inode's wb with its list_lock held.  @inode->i_lock must be
 * held on entry and is released on return.  The returned wb is guaranteed
 * to stay @inode's associated wb until its list_lock is released.
 */
static struct bdi_writeback *
locked_inode_to_wb_and_lock_list(struct inode *inode)
	__releases(&inode->i_lock)
	__acquires(&wb->list_lock)
{
	while (true) {
		struct bdi_writeback *wb = inode_to_wb(inode);

		/*
		 * inode_to_wb() association is protected by both
		 * @inode->i_lock and @wb->list_lock but list_lock nests
		 * outside i_lock.  Drop i_lock and verify that the
		 * association hasn't changed after acquiring list_lock.
		 */
		wb_get(wb);
		spin_unlock(&inode->i_lock);
		spin_lock(&wb->list_lock);

299
		/* i_wb may have changed inbetween, can't use inode_to_wb() */
300 301 302 303
		if (likely(wb == inode->i_wb)) {
			wb_put(wb);	/* @inode already has ref */
			return wb;
		}
304 305

		spin_unlock(&wb->list_lock);
306
		wb_put(wb);
307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
		cpu_relax();
		spin_lock(&inode->i_lock);
	}
}

/**
 * inode_to_wb_and_lock_list - determine an inode's wb and lock it
 * @inode: inode of interest
 *
 * Same as locked_inode_to_wb_and_lock_list() but @inode->i_lock isn't held
 * on entry.
 */
static struct bdi_writeback *inode_to_wb_and_lock_list(struct inode *inode)
	__acquires(&wb->list_lock)
{
	spin_lock(&inode->i_lock);
	return locked_inode_to_wb_and_lock_list(inode);
}

326 327 328 329 330 331 332 333 334 335 336 337 338
struct inode_switch_wbs_context {
	struct inode		*inode;
	struct bdi_writeback	*new_wb;

	struct rcu_head		rcu_head;
	struct work_struct	work;
};

static void inode_switch_wbs_work_fn(struct work_struct *work)
{
	struct inode_switch_wbs_context *isw =
		container_of(work, struct inode_switch_wbs_context, work);
	struct inode *inode = isw->inode;
339 340
	struct address_space *mapping = inode->i_mapping;
	struct bdi_writeback *old_wb = inode->i_wb;
341
	struct bdi_writeback *new_wb = isw->new_wb;
342 343 344
	struct radix_tree_iter iter;
	bool switched = false;
	void **slot;
345 346 347 348 349 350

	/*
	 * By the time control reaches here, RCU grace period has passed
	 * since I_WB_SWITCH assertion and all wb stat update transactions
	 * between unlocked_inode_to_wb_begin/end() are guaranteed to be
	 * synchronizing against mapping->tree_lock.
351 352 353 354
	 *
	 * Grabbing old_wb->list_lock, inode->i_lock and mapping->tree_lock
	 * gives us exclusion against all wb related operations on @inode
	 * including IO list manipulations and stat updates.
355
	 */
356 357 358 359 360 361 362
	if (old_wb < new_wb) {
		spin_lock(&old_wb->list_lock);
		spin_lock_nested(&new_wb->list_lock, SINGLE_DEPTH_NESTING);
	} else {
		spin_lock(&new_wb->list_lock);
		spin_lock_nested(&old_wb->list_lock, SINGLE_DEPTH_NESTING);
	}
363
	spin_lock(&inode->i_lock);
364 365 366 367
	spin_lock_irq(&mapping->tree_lock);

	/*
	 * Once I_FREEING is visible under i_lock, the eviction path owns
368
	 * the inode and we shouldn't modify ->i_io_list.
369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
	 */
	if (unlikely(inode->i_state & I_FREEING))
		goto skip_switch;

	/*
	 * Count and transfer stats.  Note that PAGECACHE_TAG_DIRTY points
	 * to possibly dirty pages while PAGECACHE_TAG_WRITEBACK points to
	 * pages actually under underwriteback.
	 */
	radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, 0,
				   PAGECACHE_TAG_DIRTY) {
		struct page *page = radix_tree_deref_slot_protected(slot,
							&mapping->tree_lock);
		if (likely(page) && PageDirty(page)) {
			__dec_wb_stat(old_wb, WB_RECLAIMABLE);
			__inc_wb_stat(new_wb, WB_RECLAIMABLE);
		}
	}

	radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, 0,
				   PAGECACHE_TAG_WRITEBACK) {
		struct page *page = radix_tree_deref_slot_protected(slot,
							&mapping->tree_lock);
		if (likely(page)) {
			WARN_ON_ONCE(!PageWriteback(page));
			__dec_wb_stat(old_wb, WB_WRITEBACK);
			__inc_wb_stat(new_wb, WB_WRITEBACK);
		}
	}

	wb_get(new_wb);

	/*
	 * Transfer to @new_wb's IO list if necessary.  The specific list
	 * @inode was on is ignored and the inode is put on ->b_dirty which
	 * is always correct including from ->b_dirty_time.  The transfer
	 * preserves @inode->dirtied_when ordering.
	 */
407
	if (!list_empty(&inode->i_io_list)) {
408 409
		struct inode *pos;

410
		inode_io_list_del_locked(inode, old_wb);
411
		inode->i_wb = new_wb;
412
		list_for_each_entry(pos, &new_wb->b_dirty, i_io_list)
413 414 415
			if (time_after_eq(inode->dirtied_when,
					  pos->dirtied_when))
				break;
416
		inode_io_list_move_locked(inode, new_wb, pos->i_io_list.prev);
417 418 419
	} else {
		inode->i_wb = new_wb;
	}
420

421
	/* ->i_wb_frn updates may race wbc_detach_inode() but doesn't matter */
422 423 424
	inode->i_wb_frn_winner = 0;
	inode->i_wb_frn_avg_time = 0;
	inode->i_wb_frn_history = 0;
425 426
	switched = true;
skip_switch:
427 428 429 430 431 432
	/*
	 * Paired with load_acquire in unlocked_inode_to_wb_begin() and
	 * ensures that the new wb is visible if they see !I_WB_SWITCH.
	 */
	smp_store_release(&inode->i_state, inode->i_state & ~I_WB_SWITCH);

433
	spin_unlock_irq(&mapping->tree_lock);
434
	spin_unlock(&inode->i_lock);
435 436
	spin_unlock(&new_wb->list_lock);
	spin_unlock(&old_wb->list_lock);
437

438 439 440 441
	if (switched) {
		wb_wakeup(new_wb);
		wb_put(old_wb);
	}
442
	wb_put(new_wb);
443 444

	iput(inode);
445
	kfree(isw);
446 447

	atomic_dec(&isw_nr_in_flight);
448 449 450 451 452 453 454 455 456
}

static void inode_switch_wbs_rcu_fn(struct rcu_head *rcu_head)
{
	struct inode_switch_wbs_context *isw = container_of(rcu_head,
				struct inode_switch_wbs_context, rcu_head);

	/* needs to grab bh-unsafe locks, bounce to work item */
	INIT_WORK(&isw->work, inode_switch_wbs_work_fn);
457
	queue_work(isw_wq, &isw->work);
458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492
}

/**
 * inode_switch_wbs - change the wb association of an inode
 * @inode: target inode
 * @new_wb_id: ID of the new wb
 *
 * Switch @inode's wb association to the wb identified by @new_wb_id.  The
 * switching is performed asynchronously and may fail silently.
 */
static void inode_switch_wbs(struct inode *inode, int new_wb_id)
{
	struct backing_dev_info *bdi = inode_to_bdi(inode);
	struct cgroup_subsys_state *memcg_css;
	struct inode_switch_wbs_context *isw;

	/* noop if seems to be already in progress */
	if (inode->i_state & I_WB_SWITCH)
		return;

	isw = kzalloc(sizeof(*isw), GFP_ATOMIC);
	if (!isw)
		return;

	/* find and pin the new wb */
	rcu_read_lock();
	memcg_css = css_from_id(new_wb_id, &memory_cgrp_subsys);
	if (memcg_css)
		isw->new_wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC);
	rcu_read_unlock();
	if (!isw->new_wb)
		goto out_free;

	/* while holding I_WB_SWITCH, no one else can update the association */
	spin_lock(&inode->i_lock);
493 494 495 496 497 498
	if (!(inode->i_sb->s_flags & MS_ACTIVE) ||
	    inode->i_state & (I_WB_SWITCH | I_FREEING) ||
	    inode_to_wb(inode) == isw->new_wb) {
		spin_unlock(&inode->i_lock);
		goto out_free;
	}
499
	inode->i_state |= I_WB_SWITCH;
500
	__iget(inode);
501 502 503 504
	spin_unlock(&inode->i_lock);

	isw->inode = inode;

505 506
	atomic_inc(&isw_nr_in_flight);

507 508 509 510 511 512 513 514 515 516 517 518 519 520 521
	/*
	 * In addition to synchronizing among switchers, I_WB_SWITCH tells
	 * the RCU protected stat update paths to grab the mapping's
	 * tree_lock so that stat transfer can synchronize against them.
	 * Let's continue after I_WB_SWITCH is guaranteed to be visible.
	 */
	call_rcu(&isw->rcu_head, inode_switch_wbs_rcu_fn);
	return;

out_free:
	if (isw->new_wb)
		wb_put(isw->new_wb);
	kfree(isw);
}

522 523 524 525 526 527 528 529 530 531 532 533 534
/**
 * wbc_attach_and_unlock_inode - associate wbc with target inode and unlock it
 * @wbc: writeback_control of interest
 * @inode: target inode
 *
 * @inode is locked and about to be written back under the control of @wbc.
 * Record @inode's writeback context into @wbc and unlock the i_lock.  On
 * writeback completion, wbc_detach_inode() should be called.  This is used
 * to track the cgroup writeback context.
 */
void wbc_attach_and_unlock_inode(struct writeback_control *wbc,
				 struct inode *inode)
{
535 536 537 538 539
	if (!inode_cgwb_enabled(inode)) {
		spin_unlock(&inode->i_lock);
		return;
	}

540
	wbc->wb = inode_to_wb(inode);
541 542 543 544 545 546 547 548 549
	wbc->inode = inode;

	wbc->wb_id = wbc->wb->memcg_css->id;
	wbc->wb_lcand_id = inode->i_wb_frn_winner;
	wbc->wb_tcand_id = 0;
	wbc->wb_bytes = 0;
	wbc->wb_lcand_bytes = 0;
	wbc->wb_tcand_bytes = 0;

550 551
	wb_get(wbc->wb);
	spin_unlock(&inode->i_lock);
552 553 554 555 556 557 558

	/*
	 * A dying wb indicates that the memcg-blkcg mapping has changed
	 * and a new wb is already serving the memcg.  Switch immediately.
	 */
	if (unlikely(wb_dying(wbc->wb)))
		inode_switch_wbs(inode, wbc->wb_id);
559 560 561
}

/**
562 563
 * wbc_detach_inode - disassociate wbc from inode and perform foreign detection
 * @wbc: writeback_control of the just finished writeback
564 565 566
 *
 * To be called after a writeback attempt of an inode finishes and undoes
 * wbc_attach_and_unlock_inode().  Can be called under any context.
567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
 *
 * As concurrent write sharing of an inode is expected to be very rare and
 * memcg only tracks page ownership on first-use basis severely confining
 * the usefulness of such sharing, cgroup writeback tracks ownership
 * per-inode.  While the support for concurrent write sharing of an inode
 * is deemed unnecessary, an inode being written to by different cgroups at
 * different points in time is a lot more common, and, more importantly,
 * charging only by first-use can too readily lead to grossly incorrect
 * behaviors (single foreign page can lead to gigabytes of writeback to be
 * incorrectly attributed).
 *
 * To resolve this issue, cgroup writeback detects the majority dirtier of
 * an inode and transfers the ownership to it.  To avoid unnnecessary
 * oscillation, the detection mechanism keeps track of history and gives
 * out the switch verdict only if the foreign usage pattern is stable over
 * a certain amount of time and/or writeback attempts.
 *
 * On each writeback attempt, @wbc tries to detect the majority writer
 * using Boyer-Moore majority vote algorithm.  In addition to the byte
 * count from the majority voting, it also counts the bytes written for the
 * current wb and the last round's winner wb (max of last round's current
 * wb, the winner from two rounds ago, and the last round's majority
 * candidate).  Keeping track of the historical winner helps the algorithm
 * to semi-reliably detect the most active writer even when it's not the
 * absolute majority.
 *
 * Once the winner of the round is determined, whether the winner is
 * foreign or not and how much IO time the round consumed is recorded in
 * inode->i_wb_frn_history.  If the amount of recorded foreign IO time is
 * over a certain threshold, the switch verdict is given.
597 598 599
 */
void wbc_detach_inode(struct writeback_control *wbc)
{
600 601
	struct bdi_writeback *wb = wbc->wb;
	struct inode *inode = wbc->inode;
602 603
	unsigned long avg_time, max_bytes, max_time;
	u16 history;
604 605
	int max_id;

606 607 608 609 610 611
	if (!wb)
		return;

	history = inode->i_wb_frn_history;
	avg_time = inode->i_wb_frn_avg_time;

612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663
	/* pick the winner of this round */
	if (wbc->wb_bytes >= wbc->wb_lcand_bytes &&
	    wbc->wb_bytes >= wbc->wb_tcand_bytes) {
		max_id = wbc->wb_id;
		max_bytes = wbc->wb_bytes;
	} else if (wbc->wb_lcand_bytes >= wbc->wb_tcand_bytes) {
		max_id = wbc->wb_lcand_id;
		max_bytes = wbc->wb_lcand_bytes;
	} else {
		max_id = wbc->wb_tcand_id;
		max_bytes = wbc->wb_tcand_bytes;
	}

	/*
	 * Calculate the amount of IO time the winner consumed and fold it
	 * into the running average kept per inode.  If the consumed IO
	 * time is lower than avag / WB_FRN_TIME_CUT_DIV, ignore it for
	 * deciding whether to switch or not.  This is to prevent one-off
	 * small dirtiers from skewing the verdict.
	 */
	max_time = DIV_ROUND_UP((max_bytes >> PAGE_SHIFT) << WB_FRN_TIME_SHIFT,
				wb->avg_write_bandwidth);
	if (avg_time)
		avg_time += (max_time >> WB_FRN_TIME_AVG_SHIFT) -
			    (avg_time >> WB_FRN_TIME_AVG_SHIFT);
	else
		avg_time = max_time;	/* immediate catch up on first run */

	if (max_time >= avg_time / WB_FRN_TIME_CUT_DIV) {
		int slots;

		/*
		 * The switch verdict is reached if foreign wb's consume
		 * more than a certain proportion of IO time in a
		 * WB_FRN_TIME_PERIOD.  This is loosely tracked by 16 slot
		 * history mask where each bit represents one sixteenth of
		 * the period.  Determine the number of slots to shift into
		 * history from @max_time.
		 */
		slots = min(DIV_ROUND_UP(max_time, WB_FRN_HIST_UNIT),
			    (unsigned long)WB_FRN_HIST_MAX_SLOTS);
		history <<= slots;
		if (wbc->wb_id != max_id)
			history |= (1U << slots) - 1;

		/*
		 * Switch if the current wb isn't the consistent winner.
		 * If there are multiple closely competing dirtiers, the
		 * inode may switch across them repeatedly over time, which
		 * is okay.  The main goal is avoiding keeping an inode on
		 * the wrong wb for an extended period of time.
		 */
664 665
		if (hweight32(history) > WB_FRN_HIST_THR_SLOTS)
			inode_switch_wbs(inode, max_id);
666 667 668 669 670 671 672 673 674 675
	}

	/*
	 * Multiple instances of this function may race to update the
	 * following fields but we don't mind occassional inaccuracies.
	 */
	inode->i_wb_frn_winner = max_id;
	inode->i_wb_frn_avg_time = min(avg_time, (unsigned long)U16_MAX);
	inode->i_wb_frn_history = history;

676 677 678 679
	wb_put(wbc->wb);
	wbc->wb = NULL;
}

680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721
/**
 * wbc_account_io - account IO issued during writeback
 * @wbc: writeback_control of the writeback in progress
 * @page: page being written out
 * @bytes: number of bytes being written out
 *
 * @bytes from @page are about to written out during the writeback
 * controlled by @wbc.  Keep the book for foreign inode detection.  See
 * wbc_detach_inode().
 */
void wbc_account_io(struct writeback_control *wbc, struct page *page,
		    size_t bytes)
{
	int id;

	/*
	 * pageout() path doesn't attach @wbc to the inode being written
	 * out.  This is intentional as we don't want the function to block
	 * behind a slow cgroup.  Ultimately, we want pageout() to kick off
	 * regular writeback instead of writing things out itself.
	 */
	if (!wbc->wb)
		return;

	id = mem_cgroup_css_from_page(page)->id;

	if (id == wbc->wb_id) {
		wbc->wb_bytes += bytes;
		return;
	}

	if (id == wbc->wb_lcand_id)
		wbc->wb_lcand_bytes += bytes;

	/* Boyer-Moore majority vote algorithm */
	if (!wbc->wb_tcand_bytes)
		wbc->wb_tcand_id = id;
	if (id == wbc->wb_tcand_id)
		wbc->wb_tcand_bytes += bytes;
	else
		wbc->wb_tcand_bytes -= min(bytes, wbc->wb_tcand_bytes);
}
722
EXPORT_SYMBOL_GPL(wbc_account_io);
723

724 725
/**
 * inode_congested - test whether an inode is congested
726
 * @inode: inode to test for congestion (may be NULL)
727 728 729 730 731 732 733 734 735
 * @cong_bits: mask of WB_[a]sync_congested bits to test
 *
 * Tests whether @inode is congested.  @cong_bits is the mask of congestion
 * bits to test and the return value is the mask of set bits.
 *
 * If cgroup writeback is enabled for @inode, the congestion state is
 * determined by whether the cgwb (cgroup bdi_writeback) for the blkcg
 * associated with @inode is congested; otherwise, the root wb's congestion
 * state is used.
736 737 738
 *
 * @inode is allowed to be NULL as this function is often called on
 * mapping->host which is NULL for the swapper space.
739 740 741
 */
int inode_congested(struct inode *inode, int cong_bits)
{
742 743 744 745
	/*
	 * Once set, ->i_wb never becomes NULL while the inode is alive.
	 * Start transaction iff ->i_wb is visible.
	 */
746
	if (inode && inode_to_wb_is_valid(inode)) {
747 748 749 750 751 752 753
		struct bdi_writeback *wb;
		bool locked, congested;

		wb = unlocked_inode_to_wb_begin(inode, &locked);
		congested = wb_congested(wb, cong_bits);
		unlocked_inode_to_wb_end(inode, locked);
		return congested;
754 755 756 757 758 759
	}

	return wb_congested(&inode_to_bdi(inode)->wb, cong_bits);
}
EXPORT_SYMBOL_GPL(inode_congested);

760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
/**
 * wb_split_bdi_pages - split nr_pages to write according to bandwidth
 * @wb: target bdi_writeback to split @nr_pages to
 * @nr_pages: number of pages to write for the whole bdi
 *
 * Split @wb's portion of @nr_pages according to @wb's write bandwidth in
 * relation to the total write bandwidth of all wb's w/ dirty inodes on
 * @wb->bdi.
 */
static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages)
{
	unsigned long this_bw = wb->avg_write_bandwidth;
	unsigned long tot_bw = atomic_long_read(&wb->bdi->tot_write_bandwidth);

	if (nr_pages == LONG_MAX)
		return LONG_MAX;

	/*
	 * This may be called on clean wb's and proportional distribution
	 * may not make sense, just use the original @nr_pages in those
	 * cases.  In general, we wanna err on the side of writing more.
	 */
	if (!tot_bw || this_bw >= tot_bw)
		return nr_pages;
	else
		return DIV_ROUND_UP_ULL((u64)nr_pages * this_bw, tot_bw);
}

788 789 790 791 792 793 794 795 796 797 798 799 800 801 802
/**
 * bdi_split_work_to_wbs - split a wb_writeback_work to all wb's of a bdi
 * @bdi: target backing_dev_info
 * @base_work: wb_writeback_work to issue
 * @skip_if_busy: skip wb's which already have writeback in progress
 *
 * Split and issue @base_work to all wb's (bdi_writeback's) of @bdi which
 * have dirty inodes.  If @base_work->nr_page isn't %LONG_MAX, it's
 * distributed to the busy wbs according to each wb's proportion in the
 * total active write bandwidth of @bdi.
 */
static void bdi_split_work_to_wbs(struct backing_dev_info *bdi,
				  struct wb_writeback_work *base_work,
				  bool skip_if_busy)
{
803
	struct bdi_writeback *last_wb = NULL;
804 805
	struct bdi_writeback *wb = list_entry(&bdi->wb_list,
					      struct bdi_writeback, bdi_node);
806 807 808 809

	might_sleep();
restart:
	rcu_read_lock();
810
	list_for_each_entry_continue_rcu(wb, &bdi->wb_list, bdi_node) {
811 812 813 814 815
		DEFINE_WB_COMPLETION_ONSTACK(fallback_work_done);
		struct wb_writeback_work fallback_work;
		struct wb_writeback_work *work;
		long nr_pages;

816 817 818 819 820
		if (last_wb) {
			wb_put(last_wb);
			last_wb = NULL;
		}

821 822 823 824 825 826
		/* SYNC_ALL writes out I_DIRTY_TIME too */
		if (!wb_has_dirty_io(wb) &&
		    (base_work->sync_mode == WB_SYNC_NONE ||
		     list_empty(&wb->b_dirty_time)))
			continue;
		if (skip_if_busy && writeback_in_progress(wb))
827 828
			continue;

829 830 831 832 833 834 835 836 837
		nr_pages = wb_split_bdi_pages(wb, base_work->nr_pages);

		work = kmalloc(sizeof(*work), GFP_ATOMIC);
		if (work) {
			*work = *base_work;
			work->nr_pages = nr_pages;
			work->auto_free = 1;
			wb_queue_work(wb, work);
			continue;
838
		}
839 840 841 842 843 844 845 846 847 848

		/* alloc failed, execute synchronously using on-stack fallback */
		work = &fallback_work;
		*work = *base_work;
		work->nr_pages = nr_pages;
		work->auto_free = 0;
		work->done = &fallback_work_done;

		wb_queue_work(wb, work);

849 850 851 852 853 854 855 856
		/*
		 * Pin @wb so that it stays on @bdi->wb_list.  This allows
		 * continuing iteration from @wb after dropping and
		 * regrabbing rcu read lock.
		 */
		wb_get(wb);
		last_wb = wb;

857 858 859
		rcu_read_unlock();
		wb_wait_for_completion(bdi, &fallback_work_done);
		goto restart;
860 861
	}
	rcu_read_unlock();
862 863 864

	if (last_wb)
		wb_put(last_wb);
865 866
}

867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893
/**
 * cgroup_writeback_umount - flush inode wb switches for umount
 *
 * This function is called when a super_block is about to be destroyed and
 * flushes in-flight inode wb switches.  An inode wb switch goes through
 * RCU and then workqueue, so the two need to be flushed in order to ensure
 * that all previously scheduled switches are finished.  As wb switches are
 * rare occurrences and synchronize_rcu() can take a while, perform
 * flushing iff wb switches are in flight.
 */
void cgroup_writeback_umount(void)
{
	if (atomic_read(&isw_nr_in_flight)) {
		synchronize_rcu();
		flush_workqueue(isw_wq);
	}
}

static int __init cgroup_writeback_init(void)
{
	isw_wq = alloc_workqueue("inode_switch_wbs", 0, 0);
	if (!isw_wq)
		return -ENOMEM;
	return 0;
}
fs_initcall(cgroup_writeback_init);

894 895
#else	/* CONFIG_CGROUP_WRITEBACK */

896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916
static struct bdi_writeback *
locked_inode_to_wb_and_lock_list(struct inode *inode)
	__releases(&inode->i_lock)
	__acquires(&wb->list_lock)
{
	struct bdi_writeback *wb = inode_to_wb(inode);

	spin_unlock(&inode->i_lock);
	spin_lock(&wb->list_lock);
	return wb;
}

static struct bdi_writeback *inode_to_wb_and_lock_list(struct inode *inode)
	__acquires(&wb->list_lock)
{
	struct bdi_writeback *wb = inode_to_wb(inode);

	spin_lock(&wb->list_lock);
	return wb;
}

917 918 919 920 921
static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages)
{
	return nr_pages;
}

922 923 924 925 926 927
static void bdi_split_work_to_wbs(struct backing_dev_info *bdi,
				  struct wb_writeback_work *base_work,
				  bool skip_if_busy)
{
	might_sleep();

928
	if (!skip_if_busy || !writeback_in_progress(&bdi->wb)) {
929 930 931 932 933
		base_work->auto_free = 0;
		wb_queue_work(&bdi->wb, base_work);
	}
}

934 935
#endif	/* CONFIG_CGROUP_WRITEBACK */

936 937
void wb_start_writeback(struct bdi_writeback *wb, long nr_pages,
			bool range_cyclic, enum wb_reason reason)
938
{
939 940 941 942 943 944 945 946 947
	struct wb_writeback_work *work;

	if (!wb_has_dirty_io(wb))
		return;

	/*
	 * This is WB_SYNC_NONE writeback, so if allocation fails just
	 * wakeup the thread for old dirty data writeback
	 */
948 949
	work = kzalloc(sizeof(*work),
		       GFP_NOWAIT | __GFP_NOMEMALLOC | __GFP_NOWARN);
950
	if (!work) {
951
		trace_writeback_nowork(wb);
952 953 954 955 956 957 958 959
		wb_wakeup(wb);
		return;
	}

	work->sync_mode	= WB_SYNC_NONE;
	work->nr_pages	= nr_pages;
	work->range_cyclic = range_cyclic;
	work->reason	= reason;
960
	work->auto_free	= 1;
961 962

	wb_queue_work(wb, work);
963
}
964

965
/**
966 967
 * wb_start_background_writeback - start background writeback
 * @wb: bdi_writback to write from
968 969
 *
 * Description:
970
 *   This makes sure WB_SYNC_NONE background writeback happens. When
971
 *   this function returns, it is only guaranteed that for given wb
972 973
 *   some IO is happening if we are over background dirty threshold.
 *   Caller need not hold sb s_umount semaphore.
974
 */
975
void wb_start_background_writeback(struct bdi_writeback *wb)
976
{
977 978 979 980
	/*
	 * We just wake up the flusher thread. It will perform background
	 * writeback as soon as there is no other work to do.
	 */
981
	trace_writeback_wake_background(wb);
982
	wb_wakeup(wb);
Linus Torvalds's avatar
Linus Torvalds committed
983 984
}

985 986 987
/*
 * Remove the inode from the writeback list it is on.
 */
988
void inode_io_list_del(struct inode *inode)
989
{
990
	struct bdi_writeback *wb;
991

992
	wb = inode_to_wb_and_lock_list(inode);
993
	inode_io_list_del_locked(inode, wb);
994
	spin_unlock(&wb->list_lock);
995 996
}

997 998 999 1000 1001 1002 1003 1004 1005 1006
/*
 * mark an inode as under writeback on the sb
 */
void sb_mark_inode_writeback(struct inode *inode)
{
	struct super_block *sb = inode->i_sb;
	unsigned long flags;

	if (list_empty(&inode->i_wb_list)) {
		spin_lock_irqsave(&sb->s_inode_wblist_lock, flags);
1007
		if (list_empty(&inode->i_wb_list)) {
1008
			list_add_tail(&inode->i_wb_list, &sb->s_inodes_wb);
1009 1010
			trace_sb_mark_inode_writeback(inode);
		}
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
		spin_unlock_irqrestore(&sb->s_inode_wblist_lock, flags);
	}
}

/*
 * clear an inode as under writeback on the sb
 */
void sb_clear_inode_writeback(struct inode *inode)
{
	struct super_block *sb = inode->i_sb;
	unsigned long flags;

	if (!list_empty(&inode->i_wb_list)) {
		spin_lock_irqsave(&sb->s_inode_wblist_lock, flags);
1025 1026 1027 1028
		if (!list_empty(&inode->i_wb_list)) {
			list_del_init(&inode->i_wb_list);
			trace_sb_clear_inode_writeback(inode);
		}
1029 1030 1031 1032
		spin_unlock_irqrestore(&sb->s_inode_wblist_lock, flags);
	}
}

1033 1034 1035 1036 1037
/*
 * Redirty an inode: set its when-it-was dirtied timestamp and move it to the
 * furthest end of its superblock's dirty-inode list.
 *
 * Before stamping the inode's ->dirtied_when, we check to see whether it is
1038
 * already the most-recently-dirtied inode on the b_dirty list.  If that is
1039 1040 1041
 * the case then the inode must have been redirtied while it was being written
 * out and we don't reset its dirtied_when.
 */
1042
static void redirty_tail(struct inode *inode, struct bdi_writeback *wb)
1043
{
1044
	if (!list_empty(&wb->b_dirty)) {
1045
		struct inode *tail;
1046

1047
		tail = wb_inode(wb->b_dirty.next);
1048
		if (time_before(inode->dirtied_when, tail->dirtied_when))
1049 1050
			inode->dirtied_when = jiffies;
	}
1051
	inode_io_list_move_locked(inode, wb, &wb->b_dirty);
1052 1053
}

1054
/*
1055
 * requeue inode for re-scanning after bdi->b_io list is exhausted.
1056
 */
1057
static void requeue_io(struct inode *inode, struct bdi_writeback *wb)
1058
{
1059
	inode_io_list_move_locked(inode, wb, &wb->b_more_io);
1060 1061
}

Joern Engel's avatar
Joern Engel committed
1062 1063
static void inode_sync_complete(struct inode *inode)
{
1064
	inode->i_state &= ~I_SYNC;
1065 1066
	/* If inode is clean an unused, put it into LRU now... */
	inode_add_lru(inode);
1067
	/* Waiters must see I_SYNC cleared before being woken up */
Joern Engel's avatar
Joern Engel committed
1068 1069 1070 1071
	smp_mb();
	wake_up_bit(&inode->i_state, __I_SYNC);
}

1072 1073 1074 1075 1076 1077 1078 1079
static bool inode_dirtied_after(struct inode *inode, unsigned long t)
{
	bool ret = time_after(inode->dirtied_when, t);
#ifndef CONFIG_64BIT
	/*
	 * For inodes being constantly redirtied, dirtied_when can get stuck.
	 * It _appears_ to be in the future, but is actually in distant past.
	 * This test is necessary to prevent such wrapped-around relative times
1080
	 * from permanently stopping the whole bdi writeback.
1081 1082 1083 1084 1085 1086
	 */
	ret = ret && time_before_eq(inode->dirtied_when, jiffies);
#endif
	return ret;
}

1087 1088
#define EXPIRE_DIRTY_ATIME 0x0001

1089
/*
1090
 * Move expired (dirtied before work->older_than_this) dirty inodes from
1091
 * @delaying_queue to @dispatch_queue.
1092
 */
1093
static int move_expired_inodes(struct list_head *delaying_queue,
1094
			       struct list_head *dispatch_queue,
1095
			       int flags,
1096
			       struct wb_writeback_work *work)
1097
{
1098 1099
	unsigned long *older_than_this = NULL;
	unsigned long expire_time;
1100 1101
	LIST_HEAD(tmp);
	struct list_head *pos, *node;
1102
	struct super_block *sb = NULL;
1103
	struct inode *inode;
1104
	int do_sb_sort = 0;
1105
	int moved = 0;
1106

1107 1108
	if ((flags & EXPIRE_DIRTY_ATIME) == 0)
		older_than_this = work->older_than_this;
1109 1110
	else if (!work->for_sync) {
		expire_time = jiffies - (dirtytime_expire_interval * HZ);
1111 1112
		older_than_this = &expire_time;
	}
1113
	while (!list_empty(delaying_queue)) {
1114
		inode = wb_inode(delaying_queue->prev);
1115 1116
		if (older_than_this &&
		    inode_dirtied_after(inode, *older_than_this))
1117
			break;
1118
		list_move(&inode->i_io_list, &tmp);
1119
		moved++;
1120 1121
		if (flags & EXPIRE_DIRTY_ATIME)
			set_bit(__I_DIRTY_TIME_EXPIRED, &inode->i_state);
1122 1123
		if (sb_is_blkdev_sb(inode->i_sb))
			continue;
1124 1125 1126
		if (sb && sb != inode->i_sb)
			do_sb_sort = 1;
		sb = inode->i_sb;
1127 1128
	}

1129 1130 1131
	/* just one sb in list, splice to dispatch_queue and we're done */
	if (!do_sb_sort) {
		list_splice(&tmp, dispatch_queue);
1132
		goto out;
1133 1134
	}

1135 1136
	/* Move inodes from one superblock together */
	while (!list_empty(&tmp)) {
1137
		sb = wb_inode(tmp.prev)->i_sb;
1138
		list_for_each_prev_safe(pos, node, &tmp) {
1139
			inode = wb_inode(pos);
1140
			if (inode->i_sb == sb)
1141
				list_move(&inode->i_io_list, dispatch_queue);
1142
		}
1143
	}
1144 1145
out:
	return moved;
1146 1147 1148 1149
}

/*
 * Queue all expired dirty inodes for io, eldest first.
1150 1151 1152 1153 1154 1155 1156 1157
 * Before
 *         newly dirtied     b_dirty    b_io    b_more_io
 *         =============>    gf         edc     BA
 * After
 *         newly dirtied     b_dirty    b_io    b_more_io
 *         =============>    g          fBAedc
 *                                           |
 *                                           +--> dequeue for IO
1158
 */
1159
static void queue_io(struct bdi_writeback *wb, struct wb_writeback_work *work)
1160
{
1161
	int moved;
1162

1163
	assert_spin_locked(&wb->list_lock);
1164
	list_splice_init(&wb->b_more_io, &wb->b_io);
1165 1166 1167
	moved = move_expired_inodes(&wb->b_dirty, &wb->b_io, 0, work);
	moved += move_expired_inodes(&wb->b_dirty_time, &wb->b_io,
				     EXPIRE_DIRTY_ATIME, work);
1168 1169
	if (moved)
		wb_io_lists_populated(wb);
1170
	trace_writeback_queue_io(wb, work, moved);
1171 1172
}

1173
static int write_inode(struct inode *inode, struct writeback_control *wbc)
1174
{
1175 1176 1177 1178 1179 1180 1181 1182
	int ret;

	if (inode->i_sb->s_op->write_inode && !is_bad_inode(inode)) {
		trace_writeback_write_inode_start(inode, wbc);
		ret = inode->i_sb->s_op->write_inode(inode, wbc);
		trace_writeback_write_inode(inode, wbc);
		return ret;
	}
1183
	return 0;
1184 1185
}

Linus Torvalds's avatar
Linus Torvalds committed
1186
/*
1187 1188
 * Wait for writeback on an inode to complete. Called with i_lock held.
 * Caller must make sure inode cannot go away when we drop i_lock.
1189
 */
1190 1191 1192
static void __inode_wait_for_writeback(struct inode *inode)
	__releases(inode->i_lock)
	__acquires(inode->i_lock)
1193 1194 1195 1196 1197
{
	DEFINE_WAIT_BIT(wq, &inode->i_state, __I_SYNC);
	wait_queue_head_t *wqh;

	wqh = bit_waitqueue(&inode->i_state, __I_SYNC);
1198 1199
	while (inode->i_state & I_SYNC) {
		spin_unlock(&inode->i_lock);
1200 1201
		__wait_on_bit(wqh, &wq, bit_wait,
			      TASK_UNINTERRUPTIBLE);
1202
		spin_lock(&inode->i_lock);
1203
	}
1204 1205
}

1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235
/*
 * Wait for writeback on an inode to complete. Caller must have inode pinned.
 */
void inode_wait_for_writeback(struct inode *inode)
{
	spin_lock(&inode->i_lock);
	__inode_wait_for_writeback(inode);
	spin_unlock(&inode->i_lock);
}

/*
 * Sleep until I_SYNC is cleared. This function must be called with i_lock
 * held and drops it. It is aimed for callers not holding any inode reference
 * so once i_lock is dropped, inode can go away.
 */
static void inode_sleep_on_writeback(struct inode *inode)
	__releases(inode->i_lock)
{
	DEFINE_WAIT(wait);
	wait_queue_head_t *wqh = bit_waitqueue(&inode->i_state, __I_SYNC);
	int sleep;

	prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE);
	sleep = inode->i_state & I_SYNC;
	spin_unlock(&inode->i_lock);
	if (sleep)
		schedule();
	finish_wait(wqh, &wait);
}

1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258
/*
 * Find proper writeback list for the inode depending on its current state and
 * possibly also change of its state while we were doing writeback.  Here we
 * handle things such as livelock prevention or fairness of writeback among
 * inodes. This function can be called only by flusher thread - noone else
 * processes all inodes in writeback lists and requeueing inodes behind flusher
 * thread's back can have unexpected consequences.
 */
static void requeue_inode(struct inode *inode, struct bdi_writeback *wb,
			  struct writeback_control *wbc)
{
	if (inode->i_state & I_FREEING)
		return;

	/*
	 * Sync livelock prevention. Each inode is tagged and synced in one
	 * shot. If still dirty, it will be redirty_tail()'ed below.  Update
	 * the dirty time to prevent enqueue and sync it again.
	 */
	if ((inode->i_state & I_DIRTY) &&
	    (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_writepages))
		inode->dirtied_when = jiffies;

1259 1260 1261 1262 1263 1264 1265 1266 1267
	if (wbc->pages_skipped) {
		/*
		 * writeback is not making progress due to locked
		 * buffers. Skip this inode for now.
		 */
		redirty_tail(inode, wb);
		return;
	}

1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292
	if (mapping_tagged(inode->i_mapping, PAGECACHE_TAG_DIRTY)) {
		/*
		 * We didn't write back all the pages.  nfs_writepages()
		 * sometimes bales out without doing anything.
		 */
		if (wbc->nr_to_write <= 0) {
			/* Slice used up. Queue for next turn. */
			requeue_io(inode, wb);
		} else {
			/*
			 * Writeback blocked by something other than
			 * congestion. Delay the inode for some time to
			 * avoid spinning on the CPU (100% iowait)
			 * retrying writeback of the dirty page/inode
			 * that cannot be performed immediately.
			 */
			redirty_tail(inode, wb);
		}
	} else if (inode->i_state & I_DIRTY) {
		/*
		 * Filesystems can dirty the inode during writeback operations,
		 * such as delayed allocation during submission or metadata
		 * updates after data IO completion.
		 */
		redirty_tail(inode, wb);
1293
	} else if (inode->i_state & I_DIRTY_TIME) {
1294
		inode->dirtied_when = jiffies;
1295
		inode_io_list_move_locked(inode, wb, &wb->b_dirty_time);
1296 1297
	} else {
		/* The inode is clean. Remove from writeback lists. */
1298
		inode_io_list_del_locked(inode, wb);
1299 1300 1301
	}
}

1302
/*
1303 1304 1305
 * Write out an inode and its dirty pages. Do not update the writeback list
 * linkage. That is left to the caller. The caller is also responsible for
 * setting I_SYNC flag and calling inode_sync_complete() to clear it.
Linus Torvalds's avatar
Linus Torvalds committed
1306 1307
 */
static int
1308
__writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
Linus Torvalds's avatar
Linus Torvalds committed
1309 1310
{
	struct address_space *mapping = inode->i_mapping;
1311
	long nr_to_write = wbc->nr_to_write;
1312
	unsigned dirty;
Linus Torvalds's avatar
Linus Torvalds committed
1313 1314
	int ret;

1315
	WARN_ON(!(inode->i_state & I_SYNC));
Linus Torvalds's avatar
Linus Torvalds committed
1316

1317 1318
	trace_writeback_single_inode_start(inode, wbc, nr_to_write);

Linus Torvalds's avatar
Linus Torvalds committed
1319 1320
	ret = do_writepages(mapping, wbc);

1321 1322 1323
	/*
	 * Make sure to wait on the data before writing out the metadata.
	 * This is important for filesystems that modify metadata on data
1324 1325 1326
	 * I/O completion. We don't do it for sync(2) writeback because it has a
	 * separate, external IO completion path and ->sync_fs for guaranteeing
	 * inode metadata is written back correctly.
1327
	 */
1328
	if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync) {
1329
		int err = filemap_fdatawait(mapping);
Linus Torvalds's avatar
Linus Torvalds committed
1330 1331 1332 1333
		if (ret == 0)
			ret = err;
	}

1334 1335 1336 1337 1338
	/*
	 * Some filesystems may redirty the inode during the writeback
	 * due to delalloc, clear dirty metadata flags right before
	 * write_inode()
	 */
1339
	spin_lock(&inode->i_lock);
1340

1341
	dirty = inode->i_state & I_DIRTY;
1342 1343
	if (inode->i_state & I_DIRTY_TIME) {
		if ((dirty & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) ||
1344
		    wbc->sync_mode == WB_SYNC_ALL ||
1345 1346 1347 1348 1349 1350 1351 1352 1353
		    unlikely(inode->i_state & I_DIRTY_TIME_EXPIRED) ||
		    unlikely(time_after(jiffies,
					(inode->dirtied_time_when +
					 dirtytime_expire_interval * HZ)))) {
			dirty |= I_DIRTY_TIME | I_DIRTY_TIME_EXPIRED;
			trace_writeback_lazytime(inode);
		}
	} else
		inode->i_state &= ~I_DIRTY_TIME_EXPIRED;
1354
	inode->i_state &= ~dirty;
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371

	/*
	 * Paired with smp_mb() in __mark_inode_dirty().  This allows
	 * __mark_inode_dirty() to test i_state without grabbing i_lock -
	 * either they see the I_DIRTY bits cleared or we see the dirtied
	 * inode.
	 *
	 * I_DIRTY_PAGES is always cleared together above even if @mapping
	 * still has dirty pages.  The flag is reinstated after smp_mb() if
	 * necessary.  This guarantees that either __mark_inode_dirty()
	 * sees clear I_DIRTY_PAGES or we see PAGECACHE_TAG_DIRTY.
	 */
	smp_mb();

	if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY))
		inode->i_state |= I_DIRTY_PAGES;

1372
	spin_unlock(&inode->i_lock);
1373

1374 1375
	if (dirty & I_DIRTY_TIME)
		mark_inode_dirty_sync(inode);
1376
	/* Don't write the inode if only I_DIRTY_PAGES was set */
1377
	if (dirty & ~I_DIRTY_PAGES) {
1378
		int err = write_inode(inode, wbc);
Linus Torvalds's avatar
Linus Torvalds committed
1379 1380 1381
		if (ret == 0)
			ret = err;
	}
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393
	trace_writeback_single_inode(inode, wbc, nr_to_write);
	return ret;
}

/*
 * Write out an inode's dirty pages. Either the caller has an active reference
 * on the inode or the inode has I_WILL_FREE set.
 *
 * This function is designed to be called for writing back one inode which
 * we go e.g. from filesystem. Flusher thread uses __writeback_single_inode()
 * and does more profound writeback list handling in writeback_sb_inodes().
 */
1394 1395
static int writeback_single_inode(struct inode *inode,
				  struct writeback_control *wbc)
1396
{
1397
	struct bdi_writeback *wb;
1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409
	int ret = 0;

	spin_lock(&inode->i_lock);
	if (!atomic_read(&inode->i_count))
		WARN_ON(!(inode->i_state & (I_WILL_FREE|I_FREEING)));
	else
		WARN_ON(inode->i_state & I_WILL_FREE);

	if (inode->i_state & I_SYNC) {
		if (wbc->sync_mode != WB_SYNC_ALL)
			goto out;
		/*
1410 1411 1412
		 * It's a data-integrity sync. We must wait. Since callers hold
		 * inode reference or inode has I_WILL_FREE set, it cannot go
		 * away under us.
1413
		 */
1414
		__inode_wait_for_writeback(inode);
1415 1416 1417
	}
	WARN_ON(inode->i_state & I_SYNC);
	/*
1418 1419 1420 1421 1422 1423
	 * Skip inode if it is clean and we have no outstanding writeback in
	 * WB_SYNC_ALL mode. We don't want to mess with writeback lists in this
	 * function since flusher thread may be doing for example sync in
	 * parallel and if we move the inode, it could get skipped. So here we
	 * make sure inode is on some writeback list and leave it there unless
	 * we have completely cleaned the inode.
1424
	 */
1425
	if (!(inode->i_state & I_DIRTY_ALL) &&
1426 1427
	    (wbc->sync_mode != WB_SYNC_ALL ||
	     !mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK)))
1428 1429
		goto out;
	inode->i_state |= I_SYNC;
1430
	wbc_attach_and_unlock_inode(wbc, inode);
1431

1432
	ret = __writeback_single_inode(inode, wbc);
Linus Torvalds's avatar
Linus Torvalds committed
1433

1434
	wbc_detach_inode(wbc);
1435 1436

	wb = inode_to_wb_and_lock_list(inode);
1437
	spin_lock(&inode->i_lock);
1438 1439 1440 1441
	/*
	 * If inode is clean, remove it from writeback lists. Otherwise don't
	 * touch it. See comment above for explanation.
	 */
1442
	if (!(inode->i_state & I_DIRTY_ALL))
1443
		inode_io_list_del_locked(inode, wb);
1444
	spin_unlock(&wb->list_lock);
Joern Engel's avatar
Joern Engel committed
1445
	inode_sync_complete(inode);
1446 1447
out:
	spin_unlock(&inode->i_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1448 1449 1450
	return ret;
}

1451
static long writeback_chunk_size(struct bdi_writeback *wb,
1452
				 struct wb_writeback_work *work)
1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470
{
	long pages;

	/*
	 * WB_SYNC_ALL mode does livelock avoidance by syncing dirty
	 * inodes/pages in one big loop. Setting wbc.nr_to_write=LONG_MAX
	 * here avoids calling into writeback_inodes_wb() more than once.
	 *
	 * The intended call sequence for WB_SYNC_ALL writeback is:
	 *
	 *      wb_writeback()
	 *          writeback_sb_inodes()       <== called only once
	 *              write_cache_pages()     <== called once for each inode
	 *                   (quickly) tag currently dirty pages
	 *                   (maybe slowly) sync all tagged pages
	 */
	if (work->sync_mode == WB_SYNC_ALL || work->tagged_writepages)
		pages = LONG_MAX;
1471
	else {
1472
		pages = min(wb->avg_write_bandwidth / 2,
1473
			    global_wb_domain.dirty_limit / DIRTY_SCOPE);
1474 1475 1476 1477
		pages = min(pages, work->nr_pages);
		pages = round_down(pages + MIN_WRITEBACK_PAGES,
				   MIN_WRITEBACK_PAGES);
	}
1478 1479 1480 1481

	return pages;
}

1482 1483
/*
 * Write a portion of b_io inodes which belong to @sb.
1484
 *
1485
 * Return the number of pages and/or inodes written.
1486 1487 1488 1489
 *
 * NOTE! This is called with wb->list_lock held, and will
 * unlock and relock that for each inode it ends up doing
 * IO for.
1490
 */
1491 1492 1493
static long writeback_sb_inodes(struct super_block *sb,
				struct bdi_writeback *wb,
				struct wb_writeback_work *work)
Linus Torvalds's avatar
Linus Torvalds committed
1494
{
1495 1496 1497 1498 1499
	struct writeback_control wbc = {
		.sync_mode		= work->sync_mode,
		.tagged_writepages	= work->tagged_writepages,
		.for_kupdate		= work->for_kupdate,
		.for_background		= work->for_background,
1500
		.for_sync		= work->for_sync,
1501 1502 1503 1504 1505 1506 1507 1508
		.range_cyclic		= work->range_cyclic,
		.range_start		= 0,
		.range_end		= LLONG_MAX,
	};
	unsigned long start_time = jiffies;
	long write_chunk;
	long wrote = 0;  /* count both pages and inodes */

1509
	while (!list_empty(&wb->b_io)) {
1510
		struct inode *inode = wb_inode(wb->b_io.prev);
1511
		struct bdi_writeback *tmp_wb;
1512 1513

		if (inode->i_sb != sb) {
1514
			if (work->sb) {
1515 1516 1517 1518 1519
				/*
				 * We only want to write back data for this
				 * superblock, move all inodes not belonging
				 * to it back onto the dirty list.
				 */
1520
				redirty_tail(inode, wb);
1521 1522 1523 1524 1525 1526 1527 1528
				continue;
			}

			/*
			 * The inode belongs to a different superblock.
			 * Bounce back to the caller to unpin this and
			 * pin the next superblock.
			 */
1529
			break;
1530 1531
		}

1532
		/*
1533 1534
		 * Don't bother with new inodes or inodes being freed, first
		 * kind does not need periodic writeout yet, and for the latter
1535 1536
		 * kind writeout is handled by the freer.
		 */
1537
		spin_lock(&inode->i_lock);
1538
		if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) {
1539
			spin_unlock(&inode->i_lock);
1540
			redirty_tail(inode, wb);
1541 1542
			continue;
		}
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557
		if ((inode->i_state & I_SYNC) && wbc.sync_mode != WB_SYNC_ALL) {
			/*
			 * If this inode is locked for writeback and we are not
			 * doing writeback-for-data-integrity, move it to
			 * b_more_io so that writeback can proceed with the
			 * other inodes on s_io.
			 *
			 * We'll have another go at writing back this inode
			 * when we completed a full scan of b_io.
			 */
			spin_unlock(&inode->i_lock);
			requeue_io(inode, wb);
			trace_writeback_sb_inodes_requeue(inode);
			continue;
		}
1558 1559
		spin_unlock(&wb->list_lock);

1560 1561 1562 1563 1564
		/*
		 * We already requeued the inode if it had I_SYNC set and we
		 * are doing WB_SYNC_NONE writeback. So this catches only the
		 * WB_SYNC_ALL case.
		 */
1565 1566 1567 1568
		if (inode->i_state & I_SYNC) {
			/* Wait for I_SYNC. This function drops i_lock... */
			inode_sleep_on_writeback(inode);
			/* Inode may be gone, start again */
1569
			spin_lock(&wb->list_lock);
1570 1571
			continue;
		}
1572
		inode->i_state |= I_SYNC;
1573
		wbc_attach_and_unlock_inode(&wbc, inode);
1574

1575
		write_chunk = writeback_chunk_size(wb, work);
1576 1577
		wbc.nr_to_write = write_chunk;
		wbc.pages_skipped = 0;
1578

1579 1580 1581 1582
		/*
		 * We use I_SYNC to pin the inode in memory. While it is set
		 * evict_inode() will wait so the inode cannot be freed.
		 */
1583
		__writeback_single_inode(inode, &wbc);
1584

1585
		wbc_detach_inode(&wbc);
1586 1587
		work->nr_pages -= write_chunk - wbc.nr_to_write;
		wrote += write_chunk - wbc.nr_to_write;
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601

		if (need_resched()) {
			/*
			 * We're trying to balance between building up a nice
			 * long list of IOs to improve our merge rate, and
			 * getting those IOs out quickly for anyone throttling
			 * in balance_dirty_pages().  cond_resched() doesn't
			 * unplug, so get our IOs out the door before we
			 * give up the CPU.
			 */
			blk_flush_plug(current);
			cond_resched();
		}

1602 1603 1604 1605 1606
		/*
		 * Requeue @inode if still dirty.  Be careful as @inode may
		 * have been switched to another wb in the meantime.
		 */
		tmp_wb = inode_to_wb_and_lock_list(inode);
1607
		spin_lock(&inode->i_lock);