summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--Documentation/filesystems/bcachefs/future/idle_work.rst6
-rw-r--r--fs/bcachefs/backpointers.c49
-rw-r--r--fs/bcachefs/btree_io.c7
-rw-r--r--fs/bcachefs/extent_update.c8
-rw-r--r--fs/bcachefs/journal.c2
-rw-r--r--fs/bcachefs/opts.h7
-rw-r--r--fs/bcachefs/recovery.c1
-rw-r--r--fs/bcachefs/recovery_passes_format.h2
-rw-r--r--fs/bcachefs/sb-counters_format.h10
-rw-r--r--fs/bcachefs/super-io.c51
-rw-r--r--fs/bcachefs/super-io.h3
-rw-r--r--fs/bcachefs/super.c6
-rw-r--r--mm/shrinker.c4
13 files changed, 102 insertions, 54 deletions
diff --git a/Documentation/filesystems/bcachefs/future/idle_work.rst b/Documentation/filesystems/bcachefs/future/idle_work.rst
index 59a332509dcd..f1202113dde0 100644
--- a/Documentation/filesystems/bcachefs/future/idle_work.rst
+++ b/Documentation/filesystems/bcachefs/future/idle_work.rst
@@ -11,10 +11,10 @@ idle" so the system can go to sleep. We don't want to be dribbling out
background work while the system should be idle.
The complicating factor is that there are a number of background tasks, which
-form a heirarchy (or a digraph, depending on how you divide it up) - one
+form a hierarchy (or a digraph, depending on how you divide it up) - one
background task may generate work for another.
-Thus proper idle detection needs to model this heirarchy.
+Thus proper idle detection needs to model this hierarchy.
- Foreground writes
- Page cache writeback
@@ -51,7 +51,7 @@ IDLE REGIME
When the system becomes idle, we should start flushing our pending work
quicker so the system can go to sleep.
-Note that the definition of "idle" depends on where in the heirarchy a task
+Note that the definition of "idle" depends on where in the hierarchy a task
is - a task should start flushing work more quickly when the task above it has
stopped generating new work.
diff --git a/fs/bcachefs/backpointers.c b/fs/bcachefs/backpointers.c
index c43aaab4c108..cb25cddb759b 100644
--- a/fs/bcachefs/backpointers.c
+++ b/fs/bcachefs/backpointers.c
@@ -532,10 +532,6 @@ static int check_bp_exists(struct btree_trans *trans,
struct btree_iter other_extent_iter = {};
CLASS(printbuf, buf)();
- if (bpos_lt(bp->k.p, s->bp_start) ||
- bpos_gt(bp->k.p, s->bp_end))
- return 0;
-
CLASS(btree_iter, bp_iter)(trans, BTREE_ID_backpointers, bp->k.p, 0);
struct bkey_s_c bp_k = bch2_btree_iter_peek_slot(&bp_iter);
int ret = bkey_err(bp_k);
@@ -690,6 +686,10 @@ static int check_extent_to_backpointers(struct btree_trans *trans,
struct bkey_i_backpointer bp;
bch2_extent_ptr_to_bp(c, btree, level, k, p, entry, &bp);
+ if (bpos_lt(bp.k.p, s->bp_start) ||
+ bpos_gt(bp.k.p, s->bp_end))
+ continue;
+
int ret = !empty
? check_bp_exists(trans, s, &bp, k)
: bch2_bucket_backpointer_mod(trans, k, &bp, true);
@@ -809,8 +809,6 @@ static int bch2_check_extents_to_backpointers_pass(struct btree_trans *trans,
for (enum btree_id btree_id = 0;
btree_id < btree_id_nr_alive(c);
btree_id++) {
- /* btree_type_has_ptrs should probably include BTREE_ID_stripes,
- * definitely her... */
int level, depth = btree_type_has_ptrs(btree_id) ? 0 : 1;
ret = commit_do(trans, NULL, NULL,
@@ -899,7 +897,7 @@ static int check_bucket_backpointer_mismatch(struct btree_trans *trans, struct b
struct bkey_s_c_backpointer bp = bkey_s_c_to_backpointer(bp_k);
- if (c->sb.version_upgrade_complete >= bcachefs_metadata_version_backpointer_bucket_gen &&
+ if (c->sb.version_upgrade_complete < bcachefs_metadata_version_backpointer_bucket_gen &&
(bp.v->bucket_gen != a->gen ||
bp.v->pad)) {
ret = bch2_backpointer_del(trans, bp_k.k->p);
@@ -931,6 +929,14 @@ static int check_bucket_backpointer_mismatch(struct btree_trans *trans, struct b
if (sectors[ALLOC_dirty] != a->dirty_sectors ||
sectors[ALLOC_cached] != a->cached_sectors ||
sectors[ALLOC_stripe] != a->stripe_sectors) {
+ /*
+ * Post 1.14 upgrade, we assume that backpointers are mostly
+ * correct and a sector count mismatch is probably due to a
+ * write buffer race
+ *
+ * Pre upgrade, we expect all the buckets to be wrong, a write
+ * buffer flush is pointless:
+ */
if (c->sb.version_upgrade_complete >= bcachefs_metadata_version_backpointer_bucket_gen) {
ret = bch2_backpointers_maybe_flush(trans, alloc_k, last_flushed);
if (ret)
@@ -978,12 +984,22 @@ static bool backpointer_node_has_missing(struct bch_fs *c, struct bkey_s_c k)
goto next;
struct bpos bucket = bp_pos_to_bucket(ca, pos);
- u64 next = ca->mi.nbuckets;
-
- unsigned long *bitmap = READ_ONCE(ca->bucket_backpointer_mismatch.buckets);
- if (bitmap)
- next = min_t(u64, next,
- find_next_bit(bitmap, ca->mi.nbuckets, bucket.offset));
+ u64 next = min(bucket.offset, ca->mi.nbuckets);
+
+ unsigned long *mismatch = READ_ONCE(ca->bucket_backpointer_mismatch.buckets);
+ unsigned long *empty = READ_ONCE(ca->bucket_backpointer_empty.buckets);
+ /*
+ * Find the first bucket with mismatches - but
+ * not empty buckets; we don't need to pin those
+ * because we just recreate all backpointers in
+ * those buckets
+ */
+ if (mismatch && empty)
+ next = find_next_andnot_bit(mismatch, empty, ca->mi.nbuckets, next);
+ else if (mismatch)
+ next = find_next_bit(mismatch, ca->mi.nbuckets, next);
+ else
+ next = ca->mi.nbuckets;
bucket.offset = next;
if (bucket.offset == ca->mi.nbuckets)
@@ -1110,17 +1126,18 @@ int bch2_check_extents_to_backpointers(struct bch_fs *c)
if (ret)
goto err;
- u64 nr_buckets = 0, nr_mismatches = 0;
+ u64 nr_buckets = 0, nr_mismatches = 0, nr_empty = 0;
for_each_member_device(c, ca) {
nr_buckets += ca->mi.nbuckets;
nr_mismatches += ca->bucket_backpointer_mismatch.nr;
+ nr_empty += ca->bucket_backpointer_empty.nr;
}
if (!nr_mismatches)
goto err;
- bch_info(c, "scanning for missing backpointers in %llu/%llu buckets",
- nr_mismatches, nr_buckets);
+ bch_info(c, "scanning for missing backpointers in %llu/%llu buckets, %llu buckets with no backpointers",
+ nr_mismatches - nr_empty, nr_buckets, nr_empty);
while (1) {
ret = bch2_pin_backpointer_nodes_with_missing(trans, s.bp_start, &s.bp_end);
diff --git a/fs/bcachefs/btree_io.c b/fs/bcachefs/btree_io.c
index 276cf088539e..2e3dd9bacac5 100644
--- a/fs/bcachefs/btree_io.c
+++ b/fs/bcachefs/btree_io.c
@@ -131,10 +131,10 @@ static void *btree_bounce_alloc(struct bch_fs *c, size_t size,
BUG_ON(size > c->opts.btree_node_size);
*used_mempool = false;
- p = kvmalloc(size, GFP_NOWAIT);
+ p = kvmalloc(size, GFP_NOWAIT|__GFP_ACCOUNT|__GFP_RECLAIMABLE);
if (!p) {
*used_mempool = true;
- p = mempool_alloc(&c->btree_bounce_pool, GFP_NOFS);
+ p = mempool_alloc(&c->btree_bounce_pool, GFP_NOFS|__GFP_ACCOUNT|__GFP_RECLAIMABLE);
}
memalloc_nofs_restore(flags);
return p;
@@ -1014,6 +1014,7 @@ static int validate_bset_keys(struct bch_fs *c, struct btree *b,
k = bkey_p_next(k);
continue;
drop_this_key:
+ ret = 0;
next_good_key = k->u64s;
if (!next_good_key ||
@@ -1470,7 +1471,7 @@ start:
}
prt_newline(&buf);
- if (failed.nr)
+ if (ret || failed.nr)
bch2_print_str_ratelimited(c, KERN_ERR, buf.buf);
async_object_list_del(c, btree_read_bio, rb->list_idx);
diff --git a/fs/bcachefs/extent_update.c b/fs/bcachefs/extent_update.c
index 7ddb156c765c..73eb28090bc7 100644
--- a/fs/bcachefs/extent_update.c
+++ b/fs/bcachefs/extent_update.c
@@ -115,9 +115,15 @@ int bch2_extent_trim_atomic(struct btree_trans *trans,
copy.flags |= BTREE_ITER_nofilter_whiteouts;
+ /*
+ * We're doing our own whiteout filtering, but we still need to pass a
+ * max key to avoid popping an assert in bch2_snapshot_is_ancestor():
+ */
struct bkey_s_c k;
unsigned nr_iters = 0;
- for_each_btree_key_continue_norestart(copy, 0, k, ret) {
+ for_each_btree_key_max_continue_norestart(copy,
+ POS(insert->k.p.inode, U64_MAX),
+ 0, k, ret) {
unsigned offset = 0;
if (bkey_gt(iter->pos, bkey_start_pos(k.k)))
diff --git a/fs/bcachefs/journal.c b/fs/bcachefs/journal.c
index 07869436a964..d56959f12210 100644
--- a/fs/bcachefs/journal.c
+++ b/fs/bcachefs/journal.c
@@ -737,9 +737,9 @@ int bch2_journal_res_get_slowpath(struct journal *j, struct journal_res *res,
return ret;
CLASS(printbuf, buf)();
+ prt_printf(&buf, bch2_fmt(c, "Journal stuck? Waited for 10 seconds, err %s"), bch2_err_str(ret));
bch2_journal_debug_to_text(&buf, j);
bch2_print_str(c, KERN_ERR, buf.buf);
- prt_printf(&buf, bch2_fmt(c, "Journal stuck? Waited for 10 seconds, err %s"), bch2_err_str(ret));
closure_wait_event(&j->async_wait,
!bch2_err_matches(ret = __journal_res_get(j, res, flags), BCH_ERR_operation_blocked) ||
diff --git a/fs/bcachefs/opts.h b/fs/bcachefs/opts.h
index 84ce69a7f131..31a3abcbd83e 100644
--- a/fs/bcachefs/opts.h
+++ b/fs/bcachefs/opts.h
@@ -242,7 +242,7 @@ enum fsck_err_opts {
x(inodes_32bit, u8, \
OPT_FS|OPT_INODE|OPT_FORMAT|OPT_MOUNT|OPT_RUNTIME, \
OPT_BOOL(), \
- BCH_SB_INODE_32BIT, true, \
+ BCH_SB_INODE_32BIT, false, \
NULL, "Constrain inode numbers to 32 bits") \
x(shard_inode_numbers_bits, u8, \
OPT_FS|OPT_FORMAT, \
@@ -321,6 +321,11 @@ enum fsck_err_opts {
OPT_BOOL(), \
BCH2_NO_SB_OPT, false, \
NULL, "Don't kick drives out when splitbrain detected")\
+ x(no_version_check, u8, \
+ OPT_HIDDEN, \
+ OPT_BOOL(), \
+ BCH2_NO_SB_OPT, false, \
+ NULL, "Don't fail reading the superblock due to incompatible version")\
x(verbose, u8, \
OPT_FS|OPT_MOUNT|OPT_RUNTIME, \
OPT_BOOL(), \
diff --git a/fs/bcachefs/recovery.c b/fs/bcachefs/recovery.c
index 29e81f96db0f..8280ca333f5b 100644
--- a/fs/bcachefs/recovery.c
+++ b/fs/bcachefs/recovery.c
@@ -64,7 +64,6 @@ int bch2_btree_lost_data(struct bch_fs *c,
* but in debug mode we want the next fsck run to be clean:
*/
ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_check_lrus, 0, &write_sb) ?: ret;
- ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_check_backpointers_to_extents, 0, &write_sb) ?: ret;
#endif
write_sb |= !__test_and_set_bit_le64(BCH_FSCK_ERR_lru_entry_bad, ext->errors_silent);
diff --git a/fs/bcachefs/recovery_passes_format.h b/fs/bcachefs/recovery_passes_format.h
index b63c20558d3d..2696eee00345 100644
--- a/fs/bcachefs/recovery_passes_format.h
+++ b/fs/bcachefs/recovery_passes_format.h
@@ -37,7 +37,7 @@
x(check_alloc_info, 10, PASS_ONLINE|PASS_FSCK_ALLOC) \
x(check_lrus, 11, PASS_ONLINE|PASS_FSCK_ALLOC) \
x(check_btree_backpointers, 12, PASS_ONLINE|PASS_FSCK_ALLOC) \
- x(check_backpointers_to_extents, 13, PASS_ONLINE|PASS_FSCK_DEBUG) \
+ x(check_backpointers_to_extents, 13, PASS_ONLINE) \
x(check_extents_to_backpointers, 14, PASS_ONLINE|PASS_FSCK_ALLOC) \
x(check_alloc_to_lru_refs, 15, PASS_ONLINE|PASS_FSCK_ALLOC) \
x(fs_freespace_init, 16, PASS_ALWAYS|PASS_SILENT) \
diff --git a/fs/bcachefs/sb-counters_format.h b/fs/bcachefs/sb-counters_format.h
index 44bc12573a0c..bfeb713dd210 100644
--- a/fs/bcachefs/sb-counters_format.h
+++ b/fs/bcachefs/sb-counters_format.h
@@ -22,7 +22,7 @@ enum counters_flags {
x(io_read_split, 33, TYPE_COUNTER) \
x(io_read_reuse_race, 34, TYPE_COUNTER) \
x(io_read_retry, 32, TYPE_COUNTER) \
- x(io_read_fail_and_poison, 82, TYPE_COUNTER) \
+ x(io_read_fail_and_poison, 95, TYPE_COUNTER) \
x(io_write, 1, TYPE_SECTORS) \
x(io_move, 2, TYPE_SECTORS) \
x(io_move_read, 35, TYPE_SECTORS) \
@@ -124,4 +124,12 @@ struct bch_sb_field_counters {
__le64 d[];
};
+static inline void __maybe_unused check_bch_counter_ids_unique(void) {
+ switch(0){
+#define x(t, n, ...) case (n):
+ BCH_PERSISTENT_COUNTERS()
+#undef x
+ }
+}
+
#endif /* _BCACHEFS_SB_COUNTERS_FORMAT_H */
diff --git a/fs/bcachefs/super-io.c b/fs/bcachefs/super-io.c
index 369465a4de77..5897380c4c08 100644
--- a/fs/bcachefs/super-io.c
+++ b/fs/bcachefs/super-io.c
@@ -379,7 +379,7 @@ static int bch2_sb_compatible(struct bch_sb *sb, struct printbuf *out)
return 0;
}
-int bch2_sb_validate(struct bch_sb *sb, u64 read_offset,
+int bch2_sb_validate(struct bch_sb *sb, struct bch_opts *opts, u64 read_offset,
enum bch_validate_flags flags, struct printbuf *out)
{
enum bch_opt_id opt_id;
@@ -389,28 +389,30 @@ int bch2_sb_validate(struct bch_sb *sb, u64 read_offset,
if (ret)
return ret;
- u64 incompat = le64_to_cpu(sb->features[0]) & (~0ULL << BCH_FEATURE_NR);
- unsigned incompat_bit = 0;
- if (incompat)
- incompat_bit = __ffs64(incompat);
- else if (sb->features[1])
- incompat_bit = 64 + __ffs64(le64_to_cpu(sb->features[1]));
-
- if (incompat_bit) {
- prt_printf(out, "Filesystem has incompatible feature bit %u, highest supported %s (%u)",
- incompat_bit,
- bch2_sb_features[BCH_FEATURE_NR - 1],
- BCH_FEATURE_NR - 1);
- return -BCH_ERR_invalid_sb_features;
- }
+ if (!opts->no_version_check) {
+ u64 incompat = le64_to_cpu(sb->features[0]) & (~0ULL << BCH_FEATURE_NR);
+ unsigned incompat_bit = 0;
+ if (incompat)
+ incompat_bit = __ffs64(incompat);
+ else if (sb->features[1])
+ incompat_bit = 64 + __ffs64(le64_to_cpu(sb->features[1]));
+
+ if (incompat_bit) {
+ prt_printf(out, "Filesystem has incompatible feature bit %u, highest supported %s (%u)",
+ incompat_bit,
+ bch2_sb_features[BCH_FEATURE_NR - 1],
+ BCH_FEATURE_NR - 1);
+ return -BCH_ERR_invalid_sb_features;
+ }
- if (BCH_VERSION_MAJOR(le16_to_cpu(sb->version)) > BCH_VERSION_MAJOR(bcachefs_metadata_version_current) ||
- BCH_SB_VERSION_INCOMPAT(sb) > bcachefs_metadata_version_current) {
- prt_str(out, "Filesystem has incompatible version ");
- bch2_version_to_text(out, le16_to_cpu(sb->version));
- prt_str(out, ", current version ");
- bch2_version_to_text(out, bcachefs_metadata_version_current);
- return -BCH_ERR_invalid_sb_features;
+ if (BCH_VERSION_MAJOR(le16_to_cpu(sb->version)) > BCH_VERSION_MAJOR(bcachefs_metadata_version_current) ||
+ BCH_SB_VERSION_INCOMPAT(sb) > bcachefs_metadata_version_current) {
+ prt_str(out, "Filesystem has incompatible version ");
+ bch2_version_to_text(out, le16_to_cpu(sb->version));
+ prt_str(out, ", current version ");
+ bch2_version_to_text(out, bcachefs_metadata_version_current);
+ return -BCH_ERR_invalid_sb_features;
+ }
}
if (bch2_is_zero(sb->user_uuid.b, sizeof(sb->user_uuid))) {
@@ -915,7 +917,7 @@ got_super:
sb->have_layout = true;
- ret = bch2_sb_validate(sb->sb, offset, 0, &err);
+ ret = bch2_sb_validate(sb->sb, opts, offset, 0, &err);
if (ret) {
bch2_print_opts(opts, KERN_ERR "bcachefs (%s): error validating superblock: %s\n",
path, err.buf);
@@ -1081,9 +1083,10 @@ int bch2_write_super(struct bch_fs *c)
bch2_sb_from_fs(c, (*ca));
darray_for_each(online_devices, ca) {
+ struct bch_opts opts = bch2_opts_empty();
printbuf_reset(&err);
- ret = bch2_sb_validate((*ca)->disk_sb.sb, 0, BCH_VALIDATE_write, &err);
+ ret = bch2_sb_validate((*ca)->disk_sb.sb, &opts, 0, BCH_VALIDATE_write, &err);
if (ret) {
bch2_fs_inconsistent(c, "sb invalid before write: %s", err.buf);
goto out;
diff --git a/fs/bcachefs/super-io.h b/fs/bcachefs/super-io.h
index a3b7a90f2533..82cb3a3ceeae 100644
--- a/fs/bcachefs/super-io.h
+++ b/fs/bcachefs/super-io.h
@@ -92,7 +92,8 @@ int bch2_sb_from_fs(struct bch_fs *, struct bch_dev *);
void bch2_free_super(struct bch_sb_handle *);
int bch2_sb_realloc(struct bch_sb_handle *, unsigned);
-int bch2_sb_validate(struct bch_sb *, u64, enum bch_validate_flags, struct printbuf *);
+int bch2_sb_validate(struct bch_sb *, struct bch_opts *, u64,
+ enum bch_validate_flags, struct printbuf *);
int bch2_read_super(const char *, struct bch_opts *, struct bch_sb_handle *);
int bch2_read_super_silent(const char *, struct bch_opts *, struct bch_sb_handle *);
diff --git a/fs/bcachefs/super.c b/fs/bcachefs/super.c
index 09e7f8ae9922..ee3b30b1c2b5 100644
--- a/fs/bcachefs/super.c
+++ b/fs/bcachefs/super.c
@@ -1021,6 +1021,12 @@ static int bch2_fs_opt_version_init(struct bch_fs *c)
prt_bitflags(&p, bch2_recovery_passes, sb_passes);
}
+ u64 btrees_lost_data = le64_to_cpu(ext->btrees_lost_data);
+ if (btrees_lost_data) {
+ prt_str(&p, "\nsuperblock indicates damage to following btrees:\n ");
+ prt_bitflags(&p, __bch2_btree_ids, btrees_lost_data);
+ }
+
if (bch2_check_version_downgrade(c)) {
prt_str(&p, "\nVersion downgrade required:");
diff --git a/mm/shrinker.c b/mm/shrinker.c
index c94eedf2cfd8..4a76364d2b7e 100644
--- a/mm/shrinker.c
+++ b/mm/shrinker.c
@@ -833,7 +833,9 @@ void shrinker_to_text(struct seq_buf *out, struct shrinker *shrinker)
};
unsigned long nr_freed = atomic_long_read(&shrinker->objects_freed);
- seq_buf_puts(out, shrinker->name);
+ seq_buf_printf(out, "%ps", shrinker->scan_objects);
+ if (shrinker->name)
+ seq_buf_printf(out, ": %s", shrinker->name);
seq_buf_putc(out, '\n');
seq_buf_printf(out, "objects: %lu\n", shrinker->count_objects(shrinker, &sc));
generated by cgit v1.2.3 (git 2.46.0) at 2025-08-11 14:17:46 +0000