summaryrefslogtreecommitdiff
path: root/fs/bcachefs/bcachefs.h
diff options
context:
space:
mode:
Diffstat (limited to 'fs/bcachefs/bcachefs.h')
-rw-r--r--fs/bcachefs/bcachefs.h221
1 files changed, 102 insertions, 119 deletions
diff --git a/fs/bcachefs/bcachefs.h b/fs/bcachefs/bcachefs.h
index 69d0d60d50e3..161cf2f05d2a 100644
--- a/fs/bcachefs/bcachefs.h
+++ b/fs/bcachefs/bcachefs.h
@@ -205,13 +205,16 @@
#include <linux/zstd.h>
#include "bcachefs_format.h"
+#include "btree_journal_iter_types.h"
+#include "disk_accounting_types.h"
#include "errcode.h"
#include "fifo.h"
#include "nocow_locking_types.h"
#include "opts.h"
-#include "recovery_types.h"
+#include "recovery_passes_types.h"
#include "sb-errors_types.h"
#include "seqmutex.h"
+#include "time_stats.h"
#include "util.h"
#ifdef CONFIG_BCACHEFS_DEBUG
@@ -265,6 +268,11 @@ do { \
#define bch2_fmt(_c, fmt) bch2_log_msg(_c, fmt "\n")
+void bch2_print_str(struct bch_fs *, const char *);
+
+__printf(2, 3)
+void bch2_print_opts(struct bch_opts *, const char *, ...);
+
__printf(2, 3)
void __bch2_print(struct bch_fs *c, const char *fmt, ...);
@@ -286,6 +294,8 @@ do { \
#define bch_info(c, fmt, ...) \
bch2_print(c, KERN_INFO bch2_fmt(c, fmt), ##__VA_ARGS__)
+#define bch_info_ratelimited(c, fmt, ...) \
+ bch2_print_ratelimited(c, KERN_INFO bch2_fmt(c, fmt), ##__VA_ARGS__)
#define bch_notice(c, fmt, ...) \
bch2_print(c, KERN_NOTICE bch2_fmt(c, fmt), ##__VA_ARGS__)
#define bch_warn(c, fmt, ...) \
@@ -345,6 +355,12 @@ do { \
bch_info(c, fmt, ##__VA_ARGS__); \
} while (0)
+#define bch_verbose_ratelimited(c, fmt, ...) \
+do { \
+ if ((c)->opts.verbose) \
+ bch_info_ratelimited(c, fmt, ##__VA_ARGS__); \
+} while (0)
+
#define pr_verbose_init(opts, fmt, ...) \
do { \
if (opt_get(opts, verbose)) \
@@ -355,6 +371,8 @@ do { \
#define BCH_DEBUG_PARAMS_ALWAYS() \
BCH_DEBUG_PARAM(key_merging_disabled, \
"Disables merging of extents") \
+ BCH_DEBUG_PARAM(btree_node_merging_disabled, \
+ "Disables merging of btree nodes") \
BCH_DEBUG_PARAM(btree_gc_always_rewrite, \
"Causes mark and sweep to compact and rewrite every " \
"btree node it traverses") \
@@ -438,6 +456,7 @@ BCH_DEBUG_PARAMS_DEBUG()
x(blocked_journal_low_on_space) \
x(blocked_journal_low_on_pin) \
x(blocked_journal_max_in_flight) \
+ x(blocked_key_cache_flush) \
x(blocked_allocate) \
x(blocked_allocate_open_bucket) \
x(blocked_write_buffer_full) \
@@ -451,7 +470,9 @@ enum bch_time_stats {
};
#include "alloc_types.h"
+#include "btree_gc_types.h"
#include "btree_types.h"
+#include "btree_node_scan_types.h"
#include "btree_write_buffer_types.h"
#include "buckets_types.h"
#include "buckets_waiting_for_journal_types.h"
@@ -463,6 +484,7 @@ enum bch_time_stats {
#include "quota_types.h"
#include "rebalance_types.h"
#include "replicas_types.h"
+#include "sb-members_types.h"
#include "subvolume_types.h"
#include "super_types.h"
#include "thread_with_file_types.h"
@@ -480,55 +502,24 @@ enum bch_time_stats {
struct btree;
-enum gc_phase {
- GC_PHASE_NOT_RUNNING,
- GC_PHASE_START,
- GC_PHASE_SB,
-
- GC_PHASE_BTREE_stripes,
- GC_PHASE_BTREE_extents,
- GC_PHASE_BTREE_inodes,
- GC_PHASE_BTREE_dirents,
- GC_PHASE_BTREE_xattrs,
- GC_PHASE_BTREE_alloc,
- GC_PHASE_BTREE_quotas,
- GC_PHASE_BTREE_reflink,
- GC_PHASE_BTREE_subvolumes,
- GC_PHASE_BTREE_snapshots,
- GC_PHASE_BTREE_lru,
- GC_PHASE_BTREE_freespace,
- GC_PHASE_BTREE_need_discard,
- GC_PHASE_BTREE_backpointers,
- GC_PHASE_BTREE_bucket_gens,
- GC_PHASE_BTREE_snapshot_trees,
- GC_PHASE_BTREE_deleted_inodes,
- GC_PHASE_BTREE_logged_ops,
- GC_PHASE_BTREE_rebalance_work,
-
- GC_PHASE_PENDING_DELETE,
-};
-
-struct gc_pos {
- enum gc_phase phase;
- struct bpos pos;
- unsigned level;
-};
-
-struct reflink_gc {
- u64 offset;
- u32 size;
- u32 refcount;
-};
-
-typedef GENRADIX(struct reflink_gc) reflink_gc_table;
-
struct io_count {
u64 sectors[2][BCH_DATA_NR];
};
+struct discard_in_flight {
+ bool in_progress:1;
+ u64 bucket:63;
+};
+
struct bch_dev {
struct kobject kobj;
+#ifdef CONFIG_BCACHEFS_DEBUG
+ atomic_long_t ref;
+ bool dying;
+ unsigned long last_put;
+#else
struct percpu_ref ref;
+#endif
struct completion ref_completion;
struct percpu_ref io_ref;
struct completion io_ref_completion;
@@ -554,36 +545,38 @@ struct bch_dev {
struct bch_devs_mask self;
- /* biosets used in cloned bios for writing multiple replicas */
- struct bio_set replica_set;
-
/*
* Buckets:
- * Per-bucket arrays are protected by c->mark_lock, bucket_lock and
- * gc_lock, for device resize - holding any is sufficient for access:
- * Or rcu_read_lock(), but only for ptr_stale():
+ * Per-bucket arrays are protected by either rcu_read_lock or
+ * state_lock, for device resize.
*/
- struct bucket_array __rcu *buckets_gc;
+ GENRADIX(struct bucket) buckets_gc;
struct bucket_gens __rcu *bucket_gens;
u8 *oldest_gen;
unsigned long *buckets_nouse;
- struct rw_semaphore bucket_lock;
- struct bch_dev_usage *usage_base;
- struct bch_dev_usage __percpu *usage[JOURNAL_BUF_NR];
- struct bch_dev_usage __percpu *usage_gc;
+ unsigned long *bucket_backpointer_mismatches;
+ unsigned long *bucket_backpointer_empty;
+
+ struct bch_dev_usage __percpu *usage;
/* Allocator: */
- u64 new_fs_bucket_idx;
- u64 alloc_cursor;
+ u64 alloc_cursor[3];
unsigned nr_open_buckets;
+ unsigned nr_partial_buckets;
unsigned nr_btree_reserve;
size_t inc_gen_needs_gc;
size_t inc_gen_really_needs_gc;
size_t buckets_waiting_on_journal;
+ struct work_struct invalidate_work;
+ struct work_struct discard_work;
+ struct mutex discard_buckets_in_flight_lock;
+ DARRAY(struct discard_in_flight) discard_buckets_in_flight;
+ struct work_struct discard_fast_work;
+
atomic64_t rebalance_work;
struct journal_device journal;
@@ -593,7 +586,7 @@ struct bch_dev {
/* The rest of this all shows up in sysfs */
atomic64_t cur_latency[2];
- struct bch2_time_stats io_latency[2];
+ struct bch2_time_stats_quantiles io_latency[2];
#define CONGESTED_MAX 1024
atomic_t congested;
@@ -609,7 +602,11 @@ struct bch_dev {
*/
#define BCH_FS_FLAGS() \
+ x(new_fs) \
x(started) \
+ x(clean_recovery) \
+ x(btree_running) \
+ x(accounting_replay_done) \
x(may_go_rw) \
x(rw) \
x(was_rw) \
@@ -618,14 +615,15 @@ struct bch_dev {
x(going_ro) \
x(write_disable_complete) \
x(clean_shutdown) \
+ x(recovery_running) \
x(fsck_running) \
x(initial_gc_unfixed) \
- x(need_another_gc) \
x(need_delete_dead_snapshots) \
x(error) \
x(topology_error) \
x(errors_fixed) \
- x(errors_not_fixed)
+ x(errors_not_fixed) \
+ x(no_invalid_checks)
enum bch_fs_flags {
#define x(n) BCH_FS_##n,
@@ -662,38 +660,15 @@ struct journal_seq_blacklist_table {
} entries[];
};
-struct journal_keys {
- struct journal_key {
- u64 journal_seq;
- u32 journal_offset;
- enum btree_id btree_id:8;
- unsigned level:8;
- bool allocated;
- bool overwritten;
- struct bkey_i *k;
- } *d;
- /*
- * Gap buffer: instead of all the empty space in the array being at the
- * end of the buffer - from @nr to @size - the empty space is at @gap.
- * This means that sequential insertions are O(n) instead of O(n^2).
- */
- size_t gap;
- size_t nr;
- size_t size;
- atomic_t ref;
- bool initial_ref_held;
-};
-
struct btree_trans_buf {
struct btree_trans *trans;
};
-#define REPLICAS_DELTA_LIST_MAX (1U << 16)
-
#define BCACHEFS_ROOT_SUBVOL_INUM \
((subvol_inum) { BCACHEFS_ROOT_SUBVOL, BCACHEFS_ROOT_INO })
#define BCH_WRITE_REFS() \
+ x(journal) \
x(trans) \
x(write) \
x(promote) \
@@ -702,9 +677,14 @@ struct btree_trans_buf {
x(stripe_delete) \
x(reflink) \
x(fallocate) \
+ x(fsync) \
+ x(dio_write) \
x(discard) \
+ x(discard_fast) \
+ x(check_discard_freespace_key) \
x(invalidate) \
x(delete_dead_snapshots) \
+ x(gc_gens) \
x(snapshot_delete_pagecache) \
x(sysfs) \
x(btree_write_buffer)
@@ -745,6 +725,12 @@ struct bch_fs {
struct percpu_ref writes;
#endif
/*
+ * Certain operations are only allowed in single threaded mode, during
+ * recovery, and we want to assert that this is the case:
+ */
+ struct task_struct *recovery_task;
+
+ /*
* Analagous to c->writes, for asynchronous ops that don't necessarily
* need fs to be read-write
*/
@@ -755,15 +741,14 @@ struct bch_fs {
struct bch_dev __rcu *devs[BCH_SB_MEMBERS_MAX];
+ struct bch_accounting_mem accounting;
+
struct bch_replicas_cpu replicas;
struct bch_replicas_cpu replicas_gc;
struct mutex replicas_gc_lock;
- mempool_t replicas_delta_pool;
struct journal_entry_res btree_root_journal_res;
- struct journal_entry_res replicas_journal_res;
struct journal_entry_res clock_journal_res;
- struct journal_entry_res dev_usage_journal_res;
struct bch_disk_groups_cpu __rcu *disk_groups;
@@ -775,6 +760,8 @@ struct bch_fs {
__uuid_t user_uuid;
u16 version;
+ u16 version_incompat;
+ u16 version_incompat_allowed;
u16 version_min;
u16 version_upgrade_complete;
@@ -789,7 +776,8 @@ struct bch_fs {
unsigned nsec_per_time_unit;
u64 features;
u64 compat;
- unsigned long errors_silent[BITS_TO_LONGS(BCH_SB_ERR_MAX)];
+ unsigned long errors_silent[BITS_TO_LONGS(BCH_FSCK_ERR_MAX)];
+ u64 btrees_lost_data;
} sb;
@@ -804,7 +792,6 @@ struct bch_fs {
/* snapshot.c: */
struct snapshot_table __rcu *snapshots;
- size_t snapshot_table_size;
struct mutex snapshot_table_lock;
struct rw_semaphore snapshot_create_lock;
@@ -815,7 +802,8 @@ struct bch_fs {
/* BTREE CACHE */
struct bio_set btree_bio;
- struct workqueue_struct *io_complete_wq;
+ struct workqueue_struct *btree_read_complete_wq;
+ struct workqueue_struct *btree_write_submit_wq;
struct btree_root btree_roots_known[BTREE_ID_NR];
DARRAY(struct btree_root) btree_roots_extra;
@@ -843,8 +831,11 @@ struct bch_fs {
struct workqueue_struct *btree_interior_update_worker;
struct work_struct btree_interior_update_work;
- struct list_head pending_node_rewrites;
- struct mutex pending_node_rewrites_lock;
+ struct workqueue_struct *btree_node_rewrite_worker;
+ struct list_head btree_node_rewrites;
+ struct list_head btree_node_rewrites_pending;
+ spinlock_t btree_node_rewrites_lock;
+ struct closure_waitlist btree_node_rewrites_wait;
/* btree_io.c: */
spinlock_t btree_write_error_lock;
@@ -881,8 +872,10 @@ struct bch_fs {
/* ALLOCATION */
struct bch_devs_mask rw_devs[BCH_DATA_NR];
+ unsigned long rw_devs_change_count;
u64 capacity; /* sectors */
+ u64 reserved; /* sectors */
/*
* When capacity _decreases_ (due to a disk being removed), we
@@ -900,27 +893,20 @@ struct bch_fs {
struct percpu_rw_semaphore mark_lock;
seqcount_t usage_lock;
- struct bch_fs_usage *usage_base;
- struct bch_fs_usage __percpu *usage[JOURNAL_BUF_NR];
- struct bch_fs_usage __percpu *usage_gc;
+ struct bch_fs_usage_base __percpu *usage;
u64 __percpu *online_reserved;
- /* single element mempool: */
- struct mutex usage_scratch_lock;
- struct bch_fs_usage_online *usage_scratch;
+ unsigned long allocator_last_stuck;
struct io_clock io_clock[2];
/* JOURNAL SEQ BLACKLIST */
struct journal_seq_blacklist_table *
journal_seq_blacklist_table;
- struct work_struct journal_seq_blacklist_gc_work;
/* ALLOCATOR */
spinlock_t freelist_lock;
struct closure_waitlist freelist_wait;
- u64 blocked_allocate;
- u64 blocked_allocate_open_bucket;
open_bucket_idx_t open_buckets_freelist;
open_bucket_idx_t open_buckets_nr_free;
@@ -940,12 +926,9 @@ struct bch_fs {
unsigned write_points_nr;
struct buckets_waiting_for_journal buckets_waiting_for_journal;
- struct work_struct discard_work;
- struct work_struct invalidate_work;
/* GARBAGE COLLECTION */
- struct task_struct *gc_thread;
- atomic_t kick_gc;
+ struct work_struct gc_gens_work;
unsigned long gc_count;
enum btree_id gc_gens_btree;
@@ -975,6 +958,7 @@ struct bch_fs {
struct bio_set bio_read;
struct bio_set bio_read_split;
struct bio_set bio_write;
+ struct bio_set replica_set;
struct mutex bio_bounce_pages_lock;
mempool_t bio_bounce_pages;
struct bucket_nocow_lock_table
@@ -982,8 +966,7 @@ struct bch_fs {
struct rhashtable promote_table;
mempool_t compression_bounce[2];
- mempool_t compress_workspace[BCH_COMPRESSION_TYPE_NR];
- mempool_t decompress_workspace;
+ mempool_t compress_workspace[BCH_COMPRESSION_OPT_NR];
size_t zstd_workspace_size;
struct crypto_shash *sha256;
@@ -1041,6 +1024,8 @@ struct bch_fs {
/* fs.c */
struct list_head vfs_inodes_list;
struct mutex vfs_inodes_lock;
+ struct rhashtable vfs_inodes_table;
+ struct rhltable vfs_inodes_by_inum_table;
/* VFS IO PATH - fs-io.c */
struct bio_set writepage_bioset;
@@ -1062,12 +1047,12 @@ struct bch_fs {
* for signaling to the toplevel code which pass we want to run now.
*/
enum bch_recovery_pass curr_recovery_pass;
- /* bitmap of explicitly enabled recovery passes: */
- u64 recovery_passes_explicit;
+ enum bch_recovery_pass next_recovery_pass;
/* bitmask of recovery passes that we actually ran */
u64 recovery_passes_complete;
/* never rewinds version of curr_recovery_pass */
enum bch_recovery_pass recovery_pass_done;
+ spinlock_t recovery_pass_lock;
struct semaphore online_fsck_mutex;
/* DEBUG JUNK */
@@ -1078,9 +1063,6 @@ struct bch_fs {
struct btree_node *verify_ondisk;
struct mutex verify_lock;
- u64 *unused_inode_hints;
- unsigned inode_shard_bits;
-
/*
* A btree node on disk could have too many bsets for an iterator to fit
* on the stack - have to dynamically allocate them
@@ -1095,15 +1077,13 @@ struct bch_fs {
struct journal_keys journal_keys;
struct list_head journal_iters;
+ struct find_btree_nodes found_btree_nodes;
+
u64 last_bucket_seq_cleanup;
u64 counters_on_mount[BCH_COUNTER_NR];
u64 __percpu *counters;
- unsigned btree_gc_periodic:1;
- unsigned copy_gc_enabled:1;
- bool promote_whole_extents;
-
struct bch2_time_stats times[BCH_TIME_STAT_NR];
struct btree_transaction_stats btree_transaction_stats[BCH_TRANSACTIONS_NR];
@@ -1212,12 +1192,15 @@ static inline bool btree_id_cached(const struct bch_fs *c, enum btree_id btree)
static inline struct timespec64 bch2_time_to_timespec(const struct bch_fs *c, s64 time)
{
struct timespec64 t;
+ s64 sec;
s32 rem;
time += c->sb.time_base_lo;
- t.tv_sec = div_s64_rem(time, c->sb.time_units_per_sec, &rem);
- t.tv_nsec = rem * c->sb.nsec_per_time_unit;
+ sec = div_s64_rem(time, c->sb.time_units_per_sec, &rem);
+
+ set_normalized_timespec64(&t, sec, rem * (s64)c->sb.nsec_per_time_unit);
+
return t;
}
@@ -1235,9 +1218,9 @@ static inline s64 bch2_current_time(const struct bch_fs *c)
return timespec_to_bch2_time(c, now);
}
-static inline bool bch2_dev_exists2(const struct bch_fs *c, unsigned dev)
+static inline u64 bch2_current_io_time(const struct bch_fs *c, int rw)
{
- return dev < c->sb.nr_devices && c->devs[dev];
+ return max(1ULL, (u64) atomic64_read(&c->io_clock[rw].now) & LRU_TIME_MAX);
}
static inline struct stdio_redirect *bch2_fs_stdio_redirect(struct bch_fs *c)
generated by cgit (git 2.34.1) at 2025-07-25 20:12:08 +0000