5 * SOME HIGH LEVEL CODE DOCUMENTATION:
7 * Bcache mostly works with cache sets, cache devices, and backing devices.
9 * Support for multiple cache devices hasn't quite been finished off yet, but
10 * it's about 95% plumbed through. A cache set and its cache devices is sort of
11 * like a md raid array and its component devices. Most of the code doesn't care
12 * about individual cache devices, the main abstraction is the cache set.
14 * Multiple cache devices is intended to give us the ability to mirror dirty
15 * cached data and metadata, without mirroring clean cached data.
17 * Backing devices are different, in that they have a lifetime independent of a
18 * cache set. When you register a newly formatted backing device it'll come up
19 * in passthrough mode, and then you can attach and detach a backing device from
20 * a cache set at runtime - while it's mounted and in use. Detaching implicitly
21 * invalidates any cached data for that backing device.
23 * A cache set can have multiple (many) backing devices attached to it.
25 * There's also flash only volumes - this is the reason for the distinction
26 * between struct cached_dev and struct bcache_device. A flash only volume
27 * works much like a bcache device that has a backing device, except the
28 * "cached" data is always dirty. The end result is that we get thin
29 * provisioning with very little additional code.
31 * Flash only volumes work but they're not production ready because the moving
32 * garbage collector needs more work. More on that later.
36 * Bcache is primarily designed for caching, which means that in normal
37 * operation all of our available space will be allocated. Thus, we need an
38 * efficient way of deleting things from the cache so we can write new things to
41 * To do this, we first divide the cache device up into buckets. A bucket is the
42 * unit of allocation; they're typically around 1 mb - anywhere from 128k to 2M+
45 * Each bucket has a 16 bit priority, and an 8 bit generation associated with
46 * it. The gens and priorities for all the buckets are stored contiguously and
47 * packed on disk (in a linked list of buckets - aside from the superblock, all
48 * of bcache's metadata is stored in buckets).
50 * The priority is used to implement an LRU. We reset a bucket's priority when
51 * we allocate it or on cache it, and every so often we decrement the priority
52 * of each bucket. It could be used to implement something more sophisticated,
53 * if anyone ever gets around to it.
55 * The generation is used for invalidating buckets. Each pointer also has an 8
56 * bit generation embedded in it; for a pointer to be considered valid, its gen
57 * must match the gen of the bucket it points into. Thus, to reuse a bucket all
58 * we have to do is increment its gen (and write its new gen to disk; we batch
61 * Bcache is entirely COW - we never write twice to a bucket, even buckets that
62 * contain metadata (including btree nodes).
66 * Bcache is in large part design around the btree.
68 * At a high level, the btree is just an index of key -> ptr tuples.
70 * Keys represent extents, and thus have a size field. Keys also have a variable
71 * number of pointers attached to them (potentially zero, which is handy for
72 * invalidating the cache).
74 * The key itself is an inode:offset pair. The inode number corresponds to a
75 * backing device or a flash only volume. The offset is the ending offset of the
76 * extent within the inode - not the starting offset; this makes lookups
77 * slightly more convenient.
79 * Pointers contain the cache device id, the offset on that device, and an 8 bit
80 * generation number. More on the gen later.
82 * Index lookups are not fully abstracted - cache lookups in particular are
83 * still somewhat mixed in with the btree code, but things are headed in that
86 * Updates are fairly well abstracted, though. There are two different ways of
87 * updating the btree; insert and replace.
89 * BTREE_INSERT will just take a list of keys and insert them into the btree -
90 * overwriting (possibly only partially) any extents they overlap with. This is
91 * used to update the index after a write.
93 * BTREE_REPLACE is really cmpxchg(); it inserts a key into the btree iff it is
94 * overwriting a key that matches another given key. This is used for inserting
95 * data into the cache after a cache miss, and for background writeback, and for
96 * the moving garbage collector.
98 * There is no "delete" operation; deleting things from the index is
99 * accomplished by either by invalidating pointers (by incrementing a bucket's
100 * gen) or by inserting a key with 0 pointers - which will overwrite anything
101 * previously present at that location in the index.
103 * This means that there are always stale/invalid keys in the btree. They're
104 * filtered out by the code that iterates through a btree node, and removed when
105 * a btree node is rewritten.
109 * Our unit of allocation is a bucket, and we we can't arbitrarily allocate and
110 * free smaller than a bucket - so, that's how big our btree nodes are.
112 * (If buckets are really big we'll only use part of the bucket for a btree node
113 * - no less than 1/4th - but a bucket still contains no more than a single
114 * btree node. I'd actually like to change this, but for now we rely on the
115 * bucket's gen for deleting btree nodes when we rewrite/split a node.)
117 * Anyways, btree nodes are big - big enough to be inefficient with a textbook
118 * btree implementation.
120 * The way this is solved is that btree nodes are internally log structured; we
121 * can append new keys to an existing btree node without rewriting it. This
122 * means each set of keys we write is sorted, but the node is not.
124 * We maintain this log structure in memory - keeping 1Mb of keys sorted would
125 * be expensive, and we have to distinguish between the keys we have written and
126 * the keys we haven't. So to do a lookup in a btree node, we have to search
127 * each sorted set. But we do merge written sets together lazily, so the cost of
128 * these extra searches is quite low (normally most of the keys in a btree node
129 * will be in one big set, and then there'll be one or two sets that are much
132 * This log structure makes bcache's btree more of a hybrid between a
133 * conventional btree and a compacting data structure, with some of the
134 * advantages of both.
136 * GARBAGE COLLECTION:
138 * We can't just invalidate any bucket - it might contain dirty data or
139 * metadata. If it once contained dirty data, other writes might overwrite it
140 * later, leaving no valid pointers into that bucket in the index.
142 * Thus, the primary purpose of garbage collection is to find buckets to reuse.
143 * It also counts how much valid data it each bucket currently contains, so that
144 * allocation can reuse buckets sooner when they've been mostly overwritten.
146 * It also does some things that are really internal to the btree
147 * implementation. If a btree node contains pointers that are stale by more than
148 * some threshold, it rewrites the btree node to avoid the bucket's generation
149 * wrapping around. It also merges adjacent btree nodes if they're empty enough.
153 * Bcache's journal is not necessary for consistency; we always strictly
154 * order metadata writes so that the btree and everything else is consistent on
155 * disk in the event of an unclean shutdown, and in fact bcache had writeback
156 * caching (with recovery from unclean shutdown) before journalling was
159 * Rather, the journal is purely a performance optimization; we can't complete a
160 * write until we've updated the index on disk, otherwise the cache would be
161 * inconsistent in the event of an unclean shutdown. This means that without the
162 * journal, on random write workloads we constantly have to update all the leaf
163 * nodes in the btree, and those writes will be mostly empty (appending at most
164 * a few keys each) - highly inefficient in terms of amount of metadata writes,
165 * and it puts more strain on the various btree resorting/compacting code.
167 * The journal is just a log of keys we've inserted; on startup we just reinsert
168 * all the keys in the open journal entries. That means that when we're updating
169 * a node in the btree, we can wait until a 4k block of keys fills up before
172 * For simplicity, we only journal updates to leaf nodes; updates to parent
173 * nodes are rare enough (since our leaf nodes are huge) that it wasn't worth
174 * the complexity to deal with journalling them (in particular, journal replay)
175 * - updates to non leaf nodes just happen synchronously (see btree_split()).
179 #define pr_fmt(fmt) "bcachefs: %s() " fmt "\n", __func__
181 #include <linux/bug.h>
182 #include <linux/bio.h>
183 #include <linux/closure.h>
184 #include <linux/kobject.h>
185 #include <linux/lglock.h>
186 #include <linux/list.h>
187 #include <linux/mutex.h>
188 #include <linux/percpu-refcount.h>
189 #include <linux/radix-tree.h>
190 #include <linux/rbtree.h>
191 #include <linux/rhashtable.h>
192 #include <linux/rwsem.h>
193 #include <linux/seqlock.h>
194 #include <linux/shrinker.h>
195 #include <linux/types.h>
196 #include <linux/workqueue.h>
197 #include <linux/zstd.h>
199 #include "bcachefs_format.h"
205 #include <linux/dynamic_fault.h>
207 #define bch2_fs_init_fault(name) \
208 dynamic_fault("bcachefs:bch_fs_init:" name)
209 #define bch2_meta_read_fault(name) \
210 dynamic_fault("bcachefs:meta:read:" name)
211 #define bch2_meta_write_fault(name) \
212 dynamic_fault("bcachefs:meta:write:" name)
215 #define bch2_fmt(_c, fmt) "bcachefs (%s): " fmt "\n", ((_c)->name)
217 #define bch2_fmt(_c, fmt) fmt "\n"
220 #define bch_info(c, fmt, ...) \
221 printk(KERN_INFO bch2_fmt(c, fmt), ##__VA_ARGS__)
222 #define bch_notice(c, fmt, ...) \
223 printk(KERN_NOTICE bch2_fmt(c, fmt), ##__VA_ARGS__)
224 #define bch_warn(c, fmt, ...) \
225 printk(KERN_WARNING bch2_fmt(c, fmt), ##__VA_ARGS__)
226 #define bch_err(c, fmt, ...) \
227 printk(KERN_ERR bch2_fmt(c, fmt), ##__VA_ARGS__)
229 #define bch_verbose(c, fmt, ...) \
231 if ((c)->opts.verbose_recovery) \
232 bch_info(c, fmt, ##__VA_ARGS__); \
235 #define pr_verbose_init(opts, fmt, ...) \
237 if (opt_get(opts, verbose_init)) \
238 pr_info(fmt, ##__VA_ARGS__); \
241 /* Parameters that are useful for debugging, but should always be compiled in: */
242 #define BCH_DEBUG_PARAMS_ALWAYS() \
243 BCH_DEBUG_PARAM(key_merging_disabled, \
244 "Disables merging of extents") \
245 BCH_DEBUG_PARAM(btree_gc_always_rewrite, \
246 "Causes mark and sweep to compact and rewrite every " \
247 "btree node it traverses") \
248 BCH_DEBUG_PARAM(btree_gc_rewrite_disabled, \
249 "Disables rewriting of btree nodes during mark and sweep")\
250 BCH_DEBUG_PARAM(btree_shrinker_disabled, \
251 "Disables the shrinker callback for the btree node cache")
253 /* Parameters that should only be compiled in in debug mode: */
254 #define BCH_DEBUG_PARAMS_DEBUG() \
255 BCH_DEBUG_PARAM(expensive_debug_checks, \
256 "Enables various runtime debugging checks that " \
257 "significantly affect performance") \
258 BCH_DEBUG_PARAM(debug_check_bkeys, \
259 "Run bkey_debugcheck (primarily checking GC/allocation "\
260 "information) when iterating over keys") \
261 BCH_DEBUG_PARAM(verify_btree_ondisk, \
262 "Reread btree nodes at various points to verify the " \
263 "mergesort in the read path against modifications " \
266 #define BCH_DEBUG_PARAMS_ALL() BCH_DEBUG_PARAMS_ALWAYS() BCH_DEBUG_PARAMS_DEBUG()
268 #ifdef CONFIG_BCACHEFS_DEBUG
269 #define BCH_DEBUG_PARAMS() BCH_DEBUG_PARAMS_ALL()
271 #define BCH_DEBUG_PARAMS() BCH_DEBUG_PARAMS_ALWAYS()
274 /* name, frequency_units, duration_units */
275 #define BCH_TIME_STATS() \
276 BCH_TIME_STAT(btree_node_mem_alloc, sec, us) \
277 BCH_TIME_STAT(btree_gc, sec, ms) \
278 BCH_TIME_STAT(btree_split, sec, us) \
279 BCH_TIME_STAT(btree_sort, ms, us) \
280 BCH_TIME_STAT(btree_read, ms, us) \
281 BCH_TIME_STAT(journal_write, us, us) \
282 BCH_TIME_STAT(journal_delay, ms, us) \
283 BCH_TIME_STAT(journal_blocked, sec, ms) \
284 BCH_TIME_STAT(journal_flush_seq, us, us)
286 #include "alloc_types.h"
287 #include "buckets_types.h"
288 #include "clock_types.h"
289 #include "journal_types.h"
290 #include "keylist_types.h"
291 #include "quota_types.h"
292 #include "super_types.h"
295 * Number of nodes we might have to allocate in a worst case btree split
296 * operation - we split all the way up to the root, then allocate a new root.
298 #define btree_reserve_required_nodes(depth) (((depth) + 1) * 2 + 1)
300 /* Number of nodes btree coalesce will try to coalesce at once */
301 #define GC_MERGE_NODES 4U
303 /* Maximum number of nodes we might need to allocate atomically: */
304 #define BTREE_RESERVE_MAX \
305 (btree_reserve_required_nodes(BTREE_MAX_DEPTH) + GC_MERGE_NODES)
307 /* Size of the freelist we allocate btree nodes from: */
308 #define BTREE_NODE_RESERVE (BTREE_RESERVE_MAX * 4)
311 struct crypto_blkcipher;
315 GC_PHASE_SB = BTREE_ID_NR + 1,
316 GC_PHASE_PENDING_DELETE,
328 u64 sectors[2][BCH_DATA_NR];
333 struct percpu_ref ref;
334 struct completion ref_completion;
335 struct percpu_ref io_ref;
336 struct completion io_ref_completion;
342 * Cached version of this device's member info from superblock
343 * Committed by bch2_write_super() -> bch_fs_mi_update()
345 struct bch_member_cpu mi;
347 char name[BDEVNAME_SIZE];
349 struct bch_sb_handle disk_sb;
352 struct bch_devs_mask self;
354 /* biosets used in cloned bios for writing multiple replicas */
355 struct bio_set replica_set;
359 * Per-bucket arrays are protected by c->usage_lock, bucket_lock and
360 * gc_lock, for device resize - holding any is sufficient for access:
361 * Or rcu_read_lock(), but only for ptr_stale():
363 struct bucket_array __rcu *buckets;
364 unsigned long *buckets_dirty;
365 /* most out of date gen in the btree */
367 struct rw_semaphore bucket_lock;
369 struct bch_dev_usage __percpu *usage_percpu;
370 struct bch_dev_usage usage_cached;
373 struct task_struct *alloc_thread;
376 * free: Buckets that are ready to be used
378 * free_inc: Incoming buckets - these are buckets that currently have
379 * cached data in them, and we can't reuse them until after we write
380 * their new gen to disk. After prio_write() finishes writing the new
381 * gens/prios, they'll be moved to the free list (and possibly discarded
384 alloc_fifo free[RESERVE_NR];
386 spinlock_t freelist_lock;
387 unsigned nr_invalidated;
389 u8 open_buckets_partial[OPEN_BUCKETS_COUNT];
390 unsigned open_buckets_partial_nr;
392 size_t fifo_last_bucket;
394 /* last calculated minimum prio */
397 atomic_long_t saturated_count;
398 size_t inc_gen_needs_gc;
399 size_t inc_gen_really_needs_gc;
400 u64 allocator_journal_seq_flush;
401 bool allocator_invalidating_data;
403 alloc_heap alloc_heap;
406 struct task_struct *copygc_thread;
407 copygc_heap copygc_heap;
408 struct bch_pd_controller copygc_pd;
409 struct write_point copygc_write_point;
411 struct journal_device journal;
413 struct work_struct io_error_work;
415 /* The rest of this all shows up in sysfs */
418 struct io_count __percpu *io_done;
422 * Flag bits for what phase of startup/shutdown the cache set is at, how we're
423 * shutting down, etc.:
425 * BCH_FS_UNREGISTERING means we're not just shutting down, we're detaching
426 * all the backing devices first (their cached data gets invalidated, and they
427 * won't automatically reattach).
432 BCH_FS_ALLOC_READ_DONE,
433 BCH_FS_ALLOCATOR_STARTED,
434 BCH_FS_INITIAL_GC_DONE,
439 BCH_FS_WRITE_DISABLE_COMPLETE,
448 BCH_FS_FSCK_FIXED_ERRORS,
450 BCH_FS_REBUILD_REPLICAS,
451 BCH_FS_HOLD_BTREE_WRITES,
456 struct dentry *btree;
457 struct dentry *btree_format;
458 struct dentry *failed;
463 struct task_struct *migrate;
464 struct bch_pd_controller pd;
466 struct bch_devs_mask devs;
467 struct write_point wp;
480 struct list_head list;
482 struct kobject internal;
483 struct kobject opts_dir;
484 struct kobject time_stats;
488 struct device *chardev;
489 struct super_block *vfs_sb;
492 /* ro/rw, add/remove devices: */
493 struct mutex state_lock;
494 enum bch_fs_state state;
496 /* Counts outstanding writes, for clean transition to read-only */
497 struct percpu_ref writes;
498 struct work_struct read_only_work;
500 struct bch_dev __rcu *devs[BCH_SB_MEMBERS_MAX];
502 struct bch_replicas_cpu __rcu *replicas;
503 struct bch_replicas_cpu __rcu *replicas_gc;
504 struct mutex replicas_gc_lock;
506 struct bch_disk_groups_cpu __rcu *disk_groups;
508 struct bch_opts opts;
510 /* Updated by bch2_sb_update():*/
515 u16 encoded_extent_max;
527 struct bch_sb *disk_sb;
528 unsigned disk_sb_order;
530 unsigned short block_bits; /* ilog2(block_size) */
532 u16 btree_foreground_merge_threshold;
534 struct closure sb_write;
535 struct mutex sb_lock;
538 struct bio_set btree_bio;
540 struct btree_root btree_roots[BTREE_ID_NR];
541 bool btree_roots_dirty;
542 struct mutex btree_root_lock;
544 struct btree_cache btree_cache;
546 mempool_t btree_reserve_pool;
549 * Cache of allocated btree nodes - if we allocate a btree node and
550 * don't use it, if we free it that space can't be reused until going
551 * _all_ the way through the allocator (which exposes us to a livelock
552 * when allocating btree reserves fail halfway through) - instead, we
553 * can stick them here:
555 struct btree_alloc btree_reserve_cache[BTREE_NODE_RESERVE * 2];
556 unsigned btree_reserve_cache_nr;
557 struct mutex btree_reserve_cache_lock;
559 mempool_t btree_interior_update_pool;
560 struct list_head btree_interior_update_list;
561 struct mutex btree_interior_update_lock;
562 struct closure_waitlist btree_interior_update_wait;
564 struct workqueue_struct *wq;
565 /* copygc needs its own workqueue for index updates.. */
566 struct workqueue_struct *copygc_wq;
569 struct delayed_work pd_controllers_update;
570 unsigned pd_controllers_update_seconds;
574 * These contain all r/w devices - i.e. devices we can currently
577 struct bch_devs_mask rw_devs[BCH_DATA_NR];
578 struct bch_tier tiers[BCH_TIER_MAX];
579 /* NULL if we only have devices in one tier: */
580 struct bch_devs_mask *fastest_devs;
581 struct bch_tier *fastest_tier;
583 u64 capacity; /* sectors */
586 * When capacity _decreases_ (due to a disk being removed), we
587 * increment capacity_gen - this invalidates outstanding reservations
588 * and forces them to be revalidated
592 atomic64_t sectors_available;
594 struct bch_fs_usage __percpu *usage_percpu;
595 struct bch_fs_usage usage_cached;
596 struct lglock usage_lock;
598 struct closure_waitlist freelist_wait;
601 * When we invalidate buckets, we use both the priority and the amount
602 * of good data to determine which buckets to reuse first - to weight
603 * those together consistently we keep track of the smallest nonzero
604 * priority of any bucket.
606 struct prio_clock prio_clock[2];
608 struct io_clock io_clock[2];
611 spinlock_t freelist_lock;
612 u8 open_buckets_freelist;
613 u8 open_buckets_nr_free;
614 struct closure_waitlist open_buckets_wait;
615 struct open_bucket open_buckets[OPEN_BUCKETS_COUNT];
617 struct write_point btree_write_point;
619 struct write_point write_points[WRITE_POINT_COUNT];
620 struct hlist_head write_points_hash[WRITE_POINT_COUNT];
621 struct mutex write_points_hash_lock;
623 /* GARBAGE COLLECTION */
624 struct task_struct *gc_thread;
626 unsigned long gc_count;
629 * Tracks GC's progress - everything in the range [ZERO_KEY..gc_cur_pos]
630 * has been marked by GC.
632 * gc_cur_phase is a superset of btree_ids (BTREE_ID_EXTENTS etc.)
634 * gc_cur_phase == GC_PHASE_DONE indicates that gc is finished/not
635 * currently running, and gc marks are currently valid
637 * Protected by gc_pos_lock. Only written to by GC thread, so GC thread
638 * can read without a lock.
640 seqcount_t gc_pos_lock;
641 struct gc_pos gc_pos;
644 * The allocation code needs gc_mark in struct bucket to be correct, but
645 * it's not while a gc is in progress.
647 struct rw_semaphore gc_lock;
650 struct bio_set bio_read;
651 struct bio_set bio_read_split;
652 struct bio_set bio_write;
653 struct mutex bio_bounce_pages_lock;
654 mempool_t bio_bounce_pages;
656 mempool_t compression_bounce[2];
657 mempool_t compress_workspace[BCH_COMPRESSION_NR];
658 mempool_t decompress_workspace;
659 ZSTD_parameters zstd_params;
661 struct crypto_shash *sha256;
662 struct crypto_skcipher *chacha20;
663 struct crypto_shash *poly1305;
665 atomic64_t key_version;
667 /* VFS IO PATH - fs-io.c */
668 struct bio_set writepage_bioset;
669 struct bio_set dio_write_bioset;
670 struct bio_set dio_read_bioset;
672 struct bio_list btree_write_error_list;
673 struct work_struct btree_write_error_work;
674 spinlock_t btree_write_error_lock;
677 struct list_head fsck_errors;
678 struct mutex fsck_error_lock;
682 wait_queue_head_t writeback_wait;
683 atomic_t writeback_pages;
684 unsigned writeback_pages_max;
685 atomic_long_t nr_inodes;
688 struct bch_memquota_type quotas[QTYP_NR];
691 struct dentry *debug;
692 struct btree_debug btree_debug[BTREE_ID_NR];
693 #ifdef CONFIG_BCACHEFS_DEBUG
694 struct btree *verify_data;
695 struct btree_node *verify_ondisk;
696 struct mutex verify_lock;
699 u64 unused_inode_hint;
702 * A btree node on disk could have too many bsets for an iterator to fit
703 * on the stack - have to dynamically allocate them
707 mempool_t btree_bounce_pool;
709 struct journal journal;
711 unsigned bucket_journal_seq;
713 /* The rest of this all shows up in sysfs */
714 atomic_long_t read_realloc_races;
715 atomic_long_t extent_migrate_done;
716 atomic_long_t extent_migrate_raced;
718 unsigned btree_gc_periodic:1;
719 unsigned copy_gc_enabled:1;
720 unsigned tiering_enabled:1;
721 unsigned tiering_percent;
723 #define BCH_DEBUG_PARAM(name, description) bool name;
724 BCH_DEBUG_PARAMS_ALL()
725 #undef BCH_DEBUG_PARAM
727 #define BCH_TIME_STAT(name, frequency_units, duration_units) \
728 struct time_stats name##_time;
733 static inline void bch2_set_ra_pages(struct bch_fs *c, unsigned ra_pages)
735 #ifndef NO_BCACHEFS_FS
737 c->vfs_sb->s_bdi->ra_pages = ra_pages;
741 static inline bool bch2_fs_running(struct bch_fs *c)
743 return c->state == BCH_FS_RO || c->state == BCH_FS_RW;
746 static inline unsigned bucket_bytes(const struct bch_dev *ca)
748 return ca->mi.bucket_size << 9;
751 static inline unsigned block_bytes(const struct bch_fs *c)
753 return c->opts.block_size << 9;
756 #endif /* _BCACHEFS_H */