diff options
Diffstat (limited to 'fs')
59 files changed, 570 insertions, 604 deletions
diff --git a/fs/autofs/inode.c b/fs/autofs/inode.c index a5083d447a62..1f5db6863663 100644 --- a/fs/autofs/inode.c +++ b/fs/autofs/inode.c @@ -309,9 +309,7 @@ static int autofs_fill_super(struct super_block *s, struct fs_context *fc) struct autofs_fs_context *ctx = fc->fs_private; struct autofs_sb_info *sbi = s->s_fs_info; struct inode *root_inode; - struct dentry *root; struct autofs_info *ino; - int ret = -ENOMEM; pr_debug("starting up, sbi = %p\n", sbi); @@ -328,56 +326,44 @@ static int autofs_fill_super(struct super_block *s, struct fs_context *fc) */ ino = autofs_new_ino(sbi); if (!ino) - goto fail; + return -ENOMEM; root_inode = autofs_get_inode(s, S_IFDIR | 0755); + if (!root_inode) + return -ENOMEM; + root_inode->i_uid = ctx->uid; root_inode->i_gid = ctx->gid; + root_inode->i_fop = &autofs_root_operations; + root_inode->i_op = &autofs_dir_inode_operations; - root = d_make_root(root_inode); - if (!root) - goto fail_ino; - - root->d_fsdata = ino; + s->s_root = d_make_root(root_inode); + if (unlikely(!s->s_root)) { + autofs_free_ino(ino); + return -ENOMEM; + } + s->s_root->d_fsdata = ino; if (ctx->pgrp_set) { sbi->oz_pgrp = find_get_pid(ctx->pgrp); - if (!sbi->oz_pgrp) { - ret = invalf(fc, "Could not find process group %d", - ctx->pgrp); - goto fail_dput; - } - } else { + if (!sbi->oz_pgrp) + return invalf(fc, "Could not find process group %d", + ctx->pgrp); + } else sbi->oz_pgrp = get_task_pid(current, PIDTYPE_PGID); - } if (autofs_type_trigger(sbi->type)) - __managed_dentry_set_managed(root); - - root_inode->i_fop = &autofs_root_operations; - root_inode->i_op = &autofs_dir_inode_operations; + /* s->s_root won't be contended so there's little to + * be gained by not taking the d_lock when setting + * d_flags, even when a lot mounts are being done. + */ + managed_dentry_set_managed(s->s_root); pr_debug("pipe fd = %d, pgrp = %u\n", sbi->pipefd, pid_nr(sbi->oz_pgrp)); sbi->flags &= ~AUTOFS_SBI_CATATONIC; - - /* - * Success! Install the root dentry now to indicate completion. - */ - s->s_root = root; return 0; - - /* - * Failure ... clean up. - */ -fail_dput: - dput(root); - goto fail; -fail_ino: - autofs_free_ino(ino); -fail: - return ret; } /* diff --git a/fs/bcachefs/backpointers.c b/fs/bcachefs/backpointers.c index ef02c9bb0354..23c0834a97a4 100644 --- a/fs/bcachefs/backpointers.c +++ b/fs/bcachefs/backpointers.c @@ -313,17 +313,17 @@ struct btree *bch2_backpointer_get_node(struct btree_trans *trans, bp.level - 1, 0); b = bch2_btree_iter_peek_node(iter); - if (IS_ERR(b)) + if (IS_ERR_OR_NULL(b)) goto err; BUG_ON(b->c.level != bp.level - 1); - if (b && extent_matches_bp(c, bp.btree_id, bp.level, - bkey_i_to_s_c(&b->key), - bucket, bp)) + if (extent_matches_bp(c, bp.btree_id, bp.level, + bkey_i_to_s_c(&b->key), + bucket, bp)) return b; - if (b && btree_node_will_make_reachable(b)) { + if (btree_node_will_make_reachable(b)) { b = ERR_PTR(-BCH_ERR_backpointer_to_overwritten_btree_node); } else { backpointer_not_found(trans, bp_pos, bp, bkey_i_to_s_c(&b->key)); diff --git a/fs/bcachefs/bcachefs.h b/fs/bcachefs/bcachefs.h index 9cb8684959ee..403aa3389fcc 100644 --- a/fs/bcachefs/bcachefs.h +++ b/fs/bcachefs/bcachefs.h @@ -617,7 +617,7 @@ struct journal_seq_blacklist_table { u64 start; u64 end; bool dirty; - } entries[0]; + } entries[]; }; struct journal_keys { diff --git a/fs/bcachefs/btree_iter.c b/fs/bcachefs/btree_iter.c index c2adf3fbb0b3..6fa90bcd7016 100644 --- a/fs/bcachefs/btree_iter.c +++ b/fs/bcachefs/btree_iter.c @@ -3087,8 +3087,6 @@ void bch2_trans_put(struct btree_trans *trans) srcu_read_unlock(&c->btree_trans_barrier, trans->srcu_idx); } - bch2_journal_preres_put(&c->journal, &trans->journal_preres); - kfree(trans->extra_journal_entries.data); if (trans->fs_usage_deltas) { diff --git a/fs/bcachefs/btree_key_cache.c b/fs/bcachefs/btree_key_cache.c index 9b78f78a75b5..37fbf22de8fc 100644 --- a/fs/bcachefs/btree_key_cache.c +++ b/fs/bcachefs/btree_key_cache.c @@ -89,10 +89,13 @@ static void bkey_cached_free(struct btree_key_cache *bc, ck->btree_trans_barrier_seq = start_poll_synchronize_srcu(&c->btree_trans_barrier); - if (ck->c.lock.readers) + if (ck->c.lock.readers) { list_move_tail(&ck->list, &bc->freed_pcpu); - else + bc->nr_freed_pcpu++; + } else { list_move_tail(&ck->list, &bc->freed_nonpcpu); + bc->nr_freed_nonpcpu++; + } atomic_long_inc(&bc->nr_freed); kfree(ck->k); @@ -109,6 +112,8 @@ static void __bkey_cached_move_to_freelist_ordered(struct btree_key_cache *bc, { struct bkey_cached *pos; + bc->nr_freed_nonpcpu++; + list_for_each_entry_reverse(pos, &bc->freed_nonpcpu, list) { if (ULONG_CMP_GE(ck->btree_trans_barrier_seq, pos->btree_trans_barrier_seq)) { @@ -158,6 +163,7 @@ static void bkey_cached_move_to_freelist(struct btree_key_cache *bc, #else mutex_lock(&bc->lock); list_move_tail(&ck->list, &bc->freed_nonpcpu); + bc->nr_freed_nonpcpu++; mutex_unlock(&bc->lock); #endif } else { @@ -217,6 +223,7 @@ bkey_cached_alloc(struct btree_trans *trans, struct btree_path *path, f->nr < ARRAY_SIZE(f->objs) / 2) { ck = list_last_entry(&bc->freed_nonpcpu, struct bkey_cached, list); list_del_init(&ck->list); + bc->nr_freed_nonpcpu--; f->objs[f->nr++] = ck; } @@ -229,6 +236,7 @@ bkey_cached_alloc(struct btree_trans *trans, struct btree_path *path, if (!list_empty(&bc->freed_nonpcpu)) { ck = list_last_entry(&bc->freed_nonpcpu, struct bkey_cached, list); list_del_init(&ck->list); + bc->nr_freed_nonpcpu--; } mutex_unlock(&bc->lock); #endif @@ -664,7 +672,6 @@ static int btree_key_cache_flush_pos(struct btree_trans *trans, goto out; bch2_journal_pin_drop(j, &ck->journal); - bch2_journal_preres_put(j, &ck->res); BUG_ON(!btree_node_locked(c_iter.path, 0)); @@ -762,18 +769,6 @@ bool bch2_btree_insert_key_cached(struct btree_trans *trans, BUG_ON(insert->k.u64s > ck->u64s); - if (likely(!(flags & BTREE_INSERT_JOURNAL_REPLAY))) { - int difference; - - BUG_ON(jset_u64s(insert->k.u64s) > trans->journal_preres.u64s); - - difference = jset_u64s(insert->k.u64s) - ck->res.u64s; - if (difference > 0) { - trans->journal_preres.u64s -= difference; - ck->res.u64s += difference; - } - } - bkey_copy(ck->k, insert); ck->valid = true; @@ -850,6 +845,8 @@ static unsigned long bch2_btree_key_cache_scan(struct shrinker *shrink, * Newest freed entries are at the end of the list - once we hit one * that's too new to be freed, we can bail out: */ + scanned += bc->nr_freed_nonpcpu; + list_for_each_entry_safe(ck, t, &bc->freed_nonpcpu, list) { if (!poll_state_synchronize_srcu(&c->btree_trans_barrier, ck->btree_trans_barrier_seq)) @@ -859,13 +856,15 @@ static unsigned long bch2_btree_key_cache_scan(struct shrinker *shrink, six_lock_exit(&ck->c.lock); kmem_cache_free(bch2_key_cache, ck); atomic_long_dec(&bc->nr_freed); - scanned++; freed++; + bc->nr_freed_nonpcpu--; } if (scanned >= nr) goto out; + scanned += bc->nr_freed_pcpu; + list_for_each_entry_safe(ck, t, &bc->freed_pcpu, list) { if (!poll_state_synchronize_srcu(&c->btree_trans_barrier, ck->btree_trans_barrier_seq)) @@ -875,8 +874,8 @@ static unsigned long bch2_btree_key_cache_scan(struct shrinker *shrink, six_lock_exit(&ck->c.lock); kmem_cache_free(bch2_key_cache, ck); atomic_long_dec(&bc->nr_freed); - scanned++; freed++; + bc->nr_freed_pcpu--; } if (scanned >= nr) @@ -982,6 +981,9 @@ void bch2_fs_btree_key_cache_exit(struct btree_key_cache *bc) } #endif + BUG_ON(list_count_nodes(&bc->freed_pcpu) != bc->nr_freed_pcpu); + BUG_ON(list_count_nodes(&bc->freed_nonpcpu) != bc->nr_freed_nonpcpu); + list_splice(&bc->freed_pcpu, &items); list_splice(&bc->freed_nonpcpu, &items); @@ -991,7 +993,6 @@ void bch2_fs_btree_key_cache_exit(struct btree_key_cache *bc) cond_resched(); bch2_journal_pin_drop(&c->journal, &ck->journal); - bch2_journal_preres_put(&c->journal, &ck->res); list_del(&ck->list); kfree(ck->k); diff --git a/fs/bcachefs/btree_key_cache_types.h b/fs/bcachefs/btree_key_cache_types.h new file mode 100644 index 000000000000..290e4e57df5b --- /dev/null +++ b/fs/bcachefs/btree_key_cache_types.h @@ -0,0 +1,34 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _BCACHEFS_BTREE_KEY_CACHE_TYPES_H +#define _BCACHEFS_BTREE_KEY_CACHE_TYPES_H + +struct btree_key_cache_freelist { + struct bkey_cached *objs[16]; + unsigned nr; +}; + +struct btree_key_cache { + struct mutex lock; + struct rhashtable table; + bool table_init_done; + + struct list_head freed_pcpu; + size_t nr_freed_pcpu; + struct list_head freed_nonpcpu; + size_t nr_freed_nonpcpu; + + struct shrinker *shrink; + unsigned shrink_iter; + struct btree_key_cache_freelist __percpu *pcpu_freed; + + atomic_long_t nr_freed; + atomic_long_t nr_keys; + atomic_long_t nr_dirty; +}; + +struct bkey_cached_key { + u32 btree_id; + struct bpos pos; +} __packed __aligned(4); + +#endif /* _BCACHEFS_BTREE_KEY_CACHE_TYPES_H */ diff --git a/fs/bcachefs/btree_trans_commit.c b/fs/bcachefs/btree_trans_commit.c index decad7b66c59..12907beda98c 100644 --- a/fs/bcachefs/btree_trans_commit.c +++ b/fs/bcachefs/btree_trans_commit.c @@ -78,6 +78,53 @@ inline void bch2_btree_node_prep_for_write(struct btree_trans *trans, bch2_btree_init_next(trans, b); } +static noinline int trans_lock_write_fail(struct btree_trans *trans, struct btree_insert_entry *i) +{ + while (--i >= trans->updates) { + if (same_leaf_as_prev(trans, i)) + continue; + + bch2_btree_node_unlock_write(trans, i->path, insert_l(i)->b); + } + + trace_and_count(trans->c, trans_restart_would_deadlock_write, trans); + return btree_trans_restart(trans, BCH_ERR_transaction_restart_would_deadlock_write); +} + +static inline int bch2_trans_lock_write(struct btree_trans *trans) +{ + struct btree_insert_entry *i; + + EBUG_ON(trans->write_locked); + + trans_for_each_update(trans, i) { + if (same_leaf_as_prev(trans, i)) + continue; + + if (bch2_btree_node_lock_write(trans, i->path, &insert_l(i)->b->c)) + return trans_lock_write_fail(trans, i); + + if (!i->cached) + bch2_btree_node_prep_for_write(trans, i->path, insert_l(i)->b); + } + + trans->write_locked = true; + return 0; +} + +static inline void bch2_trans_unlock_write(struct btree_trans *trans) +{ + if (likely(trans->write_locked)) { + struct btree_insert_entry *i; + + trans_for_each_update(trans, i) + if (!same_leaf_as_prev(trans, i)) + bch2_btree_node_unlock_write_inlined(trans, i->path, + insert_l(i)->b); + trans->write_locked = false; + } +} + /* Inserting into a given leaf node (last stage of insert): */ /* Handle overwrites and do insert, for non extents: */ @@ -276,17 +323,6 @@ static inline void btree_insert_entry_checks(struct btree_trans *trans, bch2_snapshot_is_internal_node(trans->c, i->k->k.p.snapshot)); } -static noinline int -bch2_trans_journal_preres_get_cold(struct btree_trans *trans, unsigned flags, - unsigned long trace_ip) -{ - return drop_locks_do(trans, - bch2_journal_preres_get(&trans->c->journal, - &trans->journal_preres, - trans->journal_preres_u64s, - (flags & BCH_WATERMARK_MASK))); -} - static __always_inline int bch2_trans_journal_res_get(struct btree_trans *trans, unsigned flags) { @@ -321,6 +357,45 @@ static inline int btree_key_can_insert(struct btree_trans *trans, return 0; } +noinline static int +btree_key_can_insert_cached_slowpath(struct btree_trans *trans, unsigned flags, + struct btree_path *path, unsigned new_u64s) +{ + struct bch_fs *c = trans->c; + struct btree_insert_entry *i; + struct bkey_cached *ck = (void *) path->l[0].b; + struct bkey_i *new_k; + int ret; + + bch2_trans_unlock_write(trans); + bch2_trans_unlock(trans); + + new_k = kmalloc(new_u64s * sizeof(u64), GFP_KERNEL); + if (!new_k) { + bch_err(c, "error allocating memory for key cache key, btree %s u64s %u", + bch2_btree_id_str(path->btree_id), new_u64s); + return -BCH_ERR_ENOMEM_btree_key_cache_insert; + } + + ret = bch2_trans_relock(trans) ?: + bch2_trans_lock_write(trans); + if (unlikely(ret)) { + kfree(new_k); + return ret; + } + + memcpy(new_k, ck->k, ck->u64s * sizeof(u64)); + + trans_for_each_update(trans, i) + if (i->old_v == &ck->k->v) + i->old_v = &new_k->v; + + kfree(ck->k); + ck->u64s = new_u64s; + ck->k = new_k; + return 0; +} + static int btree_key_can_insert_cached(struct btree_trans *trans, unsigned flags, struct btree_path *path, unsigned u64s) { @@ -347,12 +422,9 @@ static int btree_key_can_insert_cached(struct btree_trans *trans, unsigned flags return 0; new_u64s = roundup_pow_of_two(u64s); - new_k = krealloc(ck->k, new_u64s * sizeof(u64), GFP_NOFS); - if (!new_k) { - bch_err(c, "error allocating memory for key cache key, btree %s u64s %u", - bch2_btree_id_str(path->btree_id), new_u64s); - return -BCH_ERR_ENOMEM_btree_key_cache_insert; - } + new_k = krealloc(ck->k, new_u64s * sizeof(u64), GFP_NOWAIT); + if (unlikely(!new_k)) + return btree_key_can_insert_cached_slowpath(trans, flags, path, new_u64s); trans_for_each_update(trans, i) if (i->old_v == &ck->k->v) @@ -732,37 +804,6 @@ revert_fs_usage: return ret; } -static noinline int trans_lock_write_fail(struct btree_trans *trans, struct btree_insert_entry *i) -{ - while (--i >= trans->updates) { - if (same_leaf_as_prev(trans, i)) - continue; - - bch2_btree_node_unlock_write(trans, i->path, insert_l(i)->b); - } - - trace_and_count(trans->c, trans_restart_would_deadlock_write, trans); - return btree_trans_restart(trans, BCH_ERR_transaction_restart_would_deadlock_write); -} - -static inline int trans_lock_write(struct btree_trans *trans) -{ - struct btree_insert_entry *i; - - trans_for_each_update(trans, i) { - if (same_leaf_as_prev(trans, i)) - continue; - - if (bch2_btree_node_lock_write(trans, i->path, &insert_l(i)->b->c)) - return trans_lock_write_fail(trans, i); - - if (!i->cached) - bch2_btree_node_prep_for_write(trans, i->path, insert_l(i)->b); - } - - return 0; -} - static noinline void bch2_drop_overwrites_from_journal(struct btree_trans *trans) { struct btree_insert_entry *i; @@ -830,15 +871,7 @@ static inline int do_bch2_trans_commit(struct btree_trans *trans, unsigned flags } } - ret = bch2_journal_preres_get(&c->journal, - &trans->journal_preres, trans->journal_preres_u64s, - (flags & BCH_WATERMARK_MASK)|JOURNAL_RES_GET_NONBLOCK); - if (unlikely(ret == -BCH_ERR_journal_preres_get_blocked)) - ret = bch2_trans_journal_preres_get_cold(trans, flags, trace_ip); - if (unlikely(ret)) - return ret; - - ret = trans_lock_write(trans); + ret = bch2_trans_lock_write(trans); if (unlikely(ret)) return ret; @@ -847,10 +880,7 @@ static inline int do_bch2_trans_commit(struct btree_trans *trans, unsigned flags if (!ret && unlikely(trans->journal_replay_not_finished)) bch2_drop_overwrites_from_journal(trans); - trans_for_each_update(trans, i) - if (!same_leaf_as_prev(trans, i)) - bch2_btree_node_unlock_write_inlined(trans, i->path, - insert_l(i)->b); + bch2_trans_unlock_write(trans); if (!ret && trans->journal_pin) bch2_journal_pin_add(&c->journal, trans->journal_res.seq, @@ -1003,7 +1033,6 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags) struct bch_fs *c = trans->c; struct btree_insert_entry *i = NULL; struct btree_write_buffered_key *wb; - unsigned u64s; int ret = 0; if (!trans->nr_updates && @@ -1063,13 +1092,8 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags) EBUG_ON(test_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags)); - memset(&trans->journal_preres, 0, sizeof(trans->journal_preres)); - trans->journal_u64s = trans->extra_journal_entries.nr; - trans->journal_preres_u64s = 0; - trans->journal_transaction_names = READ_ONCE(c->opts.journal_transaction_names); - if (trans->journal_transaction_names) trans->journal_u64s += jset_u64s(JSET_ENTRY_LOG_U64s); @@ -1085,16 +1109,11 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags) if (i->key_cache_already_flushed) continue; - /* we're going to journal the key being updated: */ - u64s = jset_u64s(i->k->k.u64s); - if (i->cached && - likely(!(flags & BTREE_INSERT_JOURNAL_REPLAY))) - trans->journal_preres_u64s += u64s; - if (i->flags & BTREE_UPDATE_NOJOURNAL) continue; - trans->journal_u64s += u64s; + /* we're going to journal the key being updated: */ + trans->journal_u64s += jset_u64s(i->k->k.u64s); /* and we're also going to log the overwrite: */ if (trans->journal_transaction_names) @@ -1126,8 +1145,6 @@ retry: trace_and_count(c, transaction_commit, trans, _RET_IP_); out: - bch2_journal_preres_put(&c->journal, &trans->journal_preres); - if (likely(!(flags & BTREE_INSERT_NOCHECK_RW))) bch2_write_ref_put(c, BCH_WRITE_REF_trans); out_reset: diff --git a/fs/bcachefs/btree_types.h b/fs/bcachefs/btree_types.h index 941841a0c5bf..60453ba86c4b 100644 --- a/fs/bcachefs/btree_types.h +++ b/fs/bcachefs/btree_types.h @@ -5,7 +5,7 @@ #include <linux/list.h> #include <linux/rhashtable.h> -//#include "bkey_methods.h" +#include "btree_key_cache_types.h" #include "buckets_types.h" #include "darray.h" #include "errcode.h" @@ -312,31 +312,6 @@ struct btree_iter { #endif }; -struct btree_key_cache_freelist { - struct bkey_cached *objs[16]; - unsigned nr; -}; - -struct btree_key_cache { - struct mutex lock; - struct rhashtable table; - bool table_init_done; - struct list_head freed_pcpu; - struct list_head freed_nonpcpu; - struct shrinker *shrink; - unsigned shrink_iter; - struct btree_key_cache_freelist __percpu *pcpu_freed; - - atomic_long_t nr_freed; - atomic_long_t nr_keys; - atomic_long_t nr_dirty; -}; - -struct bkey_cached_key { - u32 btree_id; - struct bpos pos; -} __packed __aligned(4); - #define BKEY_CACHED_ACCESSED 0 #define BKEY_CACHED_DIRTY 1 @@ -352,7 +327,6 @@ struct bkey_cached { struct rhash_head hash; struct list_head list; - struct journal_preres res; struct journal_entry_pin journal; u64 seq; @@ -389,11 +363,7 @@ struct btree_insert_entry { unsigned long ip_allocated; }; -#ifndef CONFIG_LOCKDEP #define BTREE_ITER_MAX 64 -#else -#define BTREE_ITER_MAX 32 -#endif struct btree_trans_commit_hook; typedef int (btree_trans_commit_hook_fn)(struct btree_trans *, struct btree_trans_commit_hook *); @@ -434,6 +404,7 @@ struct btree_trans { bool journal_transaction_names:1; bool journal_replay_not_finished:1; bool notrace_relock_fail:1; + bool write_locked:1; enum bch_errcode restarted:16; u32 restart_count; unsigned long last_begin_ip; @@ -465,11 +436,9 @@ struct btree_trans { struct journal_entry_pin *journal_pin; struct journal_res journal_res; - struct journal_preres journal_preres; u64 *journal_seq; struct disk_reservation *disk_res; unsigned journal_u64s; - unsigned journal_preres_u64s; struct replicas_delta_list *fs_usage_deltas; }; diff --git a/fs/bcachefs/btree_update_interior.c b/fs/bcachefs/btree_update_interior.c index 39c2db68123b..76f27bc9fa24 100644 --- a/fs/bcachefs/btree_update_interior.c +++ b/fs/bcachefs/btree_update_interior.c @@ -513,8 +513,6 @@ static void bch2_btree_update_free(struct btree_update *as, struct btree_trans * up_read(&c->gc_lock); as->took_gc_lock = false; - bch2_journal_preres_put(&c->journal, &as->journal_preres); - bch2_journal_pin_drop(&c->journal, &as->journal); bch2_journal_pin_flush(&c->journal, &as->journal); bch2_disk_reservation_put(c, &as->disk_res); @@ -734,8 +732,6 @@ err: bch2_journal_pin_drop(&c->journal, &as->journal); - bch2_journal_preres_put(&c->journal, &as->journal_preres); - mutex_lock(&c->btree_interior_update_lock); for (i = 0; i < as->nr_new_nodes; i++) { b = as->new_nodes[i]; @@ -1047,7 +1043,6 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, unsigned nr_nodes[2] = { 0, 0 }; unsigned update_level = level; enum bch_watermark watermark = flags & BCH_WATERMARK_MASK; - unsigned journal_flags = 0; int ret = 0; u32 restart_count = trans->restart_count; @@ -1061,10 +1056,6 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, flags &= ~BCH_WATERMARK_MASK; flags |= watermark; - if (flags & BTREE_INSERT_JOURNAL_RECLAIM) - journal_flags |= JOURNAL_RES_GET_NONBLOCK; - journal_flags |= watermark; - while (1) { nr_nodes[!!update_level] += 1 + split; update_level++; @@ -1129,27 +1120,6 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, if (ret) goto err; - ret = bch2_journal_preres_get(&c->journal, &as->journal_preres, - BTREE_UPDATE_JOURNAL_RES, - journal_flags|JOURNAL_RES_GET_NONBLOCK); - if (ret) { - if (flags & BTREE_INSERT_JOURNAL_RECLAIM) { - ret = -BCH_ERR_journal_reclaim_would_deadlock; - goto err; - } - - ret = drop_locks_do(trans, - bch2_journal_preres_get(&c->journal, &as->journal_preres, - BTREE_UPDATE_JOURNAL_RES, - journal_flags)); - if (ret == -BCH_ERR_journal_preres_get_blocked) { - trace_and_count(c, trans_restart_journal_preres_get, trans, _RET_IP_, journal_flags); - ret = btree_trans_restart(trans, BCH_ERR_transaction_restart_journal_preres_get); - } - if (ret) - goto err; - } - ret = bch2_disk_reservation_get(c, &as->disk_res, (nr_nodes[0] + nr_nodes[1]) * btree_sectors(c), c->opts.metadata_replicas, diff --git a/fs/bcachefs/btree_update_interior.h b/fs/bcachefs/btree_update_interior.h index 4df21512d640..031076e75fa1 100644 --- a/fs/bcachefs/btree_update_interior.h +++ b/fs/bcachefs/btree_update_interior.h @@ -55,7 +55,6 @@ struct btree_update { unsigned update_level; struct disk_reservation disk_res; - struct journal_preres journal_preres; /* * BTREE_INTERIOR_UPDATING_NODE: diff --git a/fs/bcachefs/data_update.c b/fs/bcachefs/data_update.c index 0771a6d880bf..5ed66202c226 100644 --- a/fs/bcachefs/data_update.c +++ b/fs/bcachefs/data_update.c @@ -239,6 +239,34 @@ restart_drop_extra_replicas: next_pos = insert->k.p; + /* + * Check for nonce offset inconsistency: + * This is debug code - we've been seeing this bug rarely, and + * it's been hard to reproduce, so this should give us some more + * information when it does occur: + */ + struct printbuf err = PRINTBUF; + int invalid = bch2_bkey_invalid(c, bkey_i_to_s_c(insert), __btree_node_type(0, m->btree_id), 0, &err); + printbuf_exit(&err); + + if (invalid) { + struct printbuf buf = PRINTBUF; + + prt_str(&buf, "about to insert invalid key in data update path"); + prt_str(&buf, "\nold: "); + bch2_bkey_val_to_text(&buf, c, old); + prt_str(&buf, "\nk: "); + bch2_bkey_val_to_text(&buf, c, k); + prt_str(&buf, "\nnew: "); + bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(insert)); + + bch2_print_string_as_lines(KERN_ERR, buf.buf); + printbuf_exit(&buf); + + bch2_fatal_error(c); + goto out; + } + ret = bch2_insert_snapshot_whiteouts(trans, m->btree_id, k.k->p, bkey_start_pos(&insert->k)) ?: bch2_insert_snapshot_whiteouts(trans, m->btree_id, diff --git a/fs/bcachefs/disk_groups.c b/fs/bcachefs/disk_groups.c index d613695abf9f..4d0cb0ccff32 100644 --- a/fs/bcachefs/disk_groups.c +++ b/fs/bcachefs/disk_groups.c @@ -555,6 +555,7 @@ void bch2_target_to_text(struct printbuf *out, struct bch_fs *c, unsigned v) case TARGET_DEV: { struct bch_dev *ca; + out->atomic++; rcu_read_lock(); ca = t.dev < c->sb.nr_devices ? rcu_dereference(c->devs[t.dev]) @@ -570,6 +571,7 @@ void bch2_target_to_text(struct printbuf *out, struct bch_fs *c, unsigned v) } rcu_read_unlock(); + out->atomic--; break; } case TARGET_GROUP: @@ -580,7 +582,7 @@ void bch2_target_to_text(struct printbuf *out, struct bch_fs *c, unsigned v) } } -void bch2_target_to_text_sb(struct printbuf *out, struct bch_sb *sb, unsigned v) +static void bch2_target_to_text_sb(struct printbuf *out, struct bch_sb *sb, unsigned v) { struct target t = target_decode(v); diff --git a/fs/bcachefs/ec.c b/fs/bcachefs/ec.c index 875f7c5a6fca..2a77de18c004 100644 --- a/fs/bcachefs/ec.c +++ b/fs/bcachefs/ec.c @@ -1373,6 +1373,15 @@ ec_new_stripe_head_alloc(struct bch_fs *c, unsigned target, h->nr_active_devs++; rcu_read_unlock(); + + /* + * If we only have redundancy + 1 devices, we're better off with just + * replication: + */ + if (h->nr_active_devs < h->redundancy + 2) + bch_err(c, "insufficient devices available to create stripe (have %u, need %u) - mismatched bucket sizes?", + h->nr_active_devs, h->redundancy + 2); + list_add(&h->list, &c->ec_stripe_head_list); return h; } @@ -1424,6 +1433,11 @@ __bch2_ec_stripe_head_get(struct btree_trans *trans, h = ec_new_stripe_head_alloc(c, target, algo, redundancy, watermark); found: + if (!IS_ERR_OR_NULL(h) && + h->nr_active_devs < h->redundancy + 2) { + mutex_unlock(&h->lock); + h = NULL; + } mutex_unlock(&c->ec_stripe_head_lock); return h; } @@ -1681,8 +1695,6 @@ struct ec_stripe_head *bch2_ec_stripe_head_get(struct btree_trans *trans, int ret; h = __bch2_ec_stripe_head_get(trans, target, algo, redundancy, watermark); - if (!h) - bch_err(c, "no stripe head"); if (IS_ERR_OR_NULL(h)) return h; diff --git a/fs/bcachefs/fs-io-pagecache.c b/fs/bcachefs/fs-io-pagecache.c index 8bd9bcdd27f7..ff664fd0d8ef 100644 --- a/fs/bcachefs/fs-io-pagecache.c +++ b/fs/bcachefs/fs-io-pagecache.c @@ -13,7 +13,7 @@ int bch2_filemap_get_contig_folios_d(struct address_space *mapping, loff_t start, u64 end, - int fgp_flags, gfp_t gfp, + fgf_t fgp_flags, gfp_t gfp, folios *fs) { struct folio *f; diff --git a/fs/bcachefs/fs-io-pagecache.h b/fs/bcachefs/fs-io-pagecache.h index a2222ad586e9..27f712ae37a6 100644 --- a/fs/bcachefs/fs-io-pagecache.h +++ b/fs/bcachefs/fs-io-pagecache.h @@ -7,7 +7,7 @@ typedef DARRAY(struct folio *) folios; int bch2_filemap_get_contig_folios_d(struct address_space *, loff_t, - u64, int, gfp_t, folios *); + u64, fgf_t, gfp_t, folios *); int bch2_write_invalidate_inode_pages_range(struct address_space *, loff_t, loff_t); /* diff --git a/fs/bcachefs/fs.c b/fs/bcachefs/fs.c index 166d8d8abe68..8ef817304e4a 100644 --- a/fs/bcachefs/fs.c +++ b/fs/bcachefs/fs.c @@ -1922,10 +1922,7 @@ out: return dget(sb->s_root); err_put_super: - sb->s_fs_info = NULL; - c->vfs_sb = NULL; deactivate_locked_super(sb); - bch2_fs_stop(c); return ERR_PTR(bch2_err_class(ret)); } @@ -1933,11 +1930,8 @@ static void bch2_kill_sb(struct super_block *sb) { struct bch_fs *c = sb->s_fs_info; - if (c) - c->vfs_sb = NULL; generic_shutdown_super(sb); - if (c) - bch2_fs_free(c); + bch2_fs_free(c); } static struct file_system_type bcache_fs_type = { diff --git a/fs/bcachefs/fsck.c b/fs/bcachefs/fsck.c index 9f3e9bd3d767..e0c5cd119acc 100644 --- a/fs/bcachefs/fsck.c +++ b/fs/bcachefs/fsck.c @@ -2220,7 +2220,7 @@ static int nlink_cmp(const void *_l, const void *_r) const struct nlink *l = _l; const struct nlink *r = _r; - return cmp_int(l->inum, r->inum) ?: cmp_int(l->snapshot, r->snapshot); + return cmp_int(l->inum, r->inum); } static void inc_link(struct bch_fs *c, struct snapshots_seen *s, diff --git a/fs/bcachefs/inode.c b/fs/bcachefs/inode.c index def77f2d8802..c7849b0753e7 100644 --- a/fs/bcachefs/inode.c +++ b/fs/bcachefs/inode.c @@ -1134,7 +1134,7 @@ static int may_delete_deleted_inode(struct btree_trans *trans, * unlinked inodes in the snapshot leaves: */ *need_another_pass = true; - return 0; + goto out; } ret = 1; @@ -1169,8 +1169,10 @@ again: */ for_each_btree_key(trans, iter, BTREE_ID_deleted_inodes, POS_MIN, BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS, k, ret) { - ret = lockrestart_do(trans, may_delete_deleted_inode(trans, &iter, k.k->p, - &need_another_pass)); + ret = commit_do(trans, NULL, NULL, + BTREE_INSERT_NOFAIL| + BTREE_INSERT_LAZY_RW, + may_delete_deleted_inode(trans, &iter, k.k->p, &need_another_pass)); if (ret < 0) break; diff --git a/fs/bcachefs/io_write.c b/fs/bcachefs/io_write.c index f02b3f7d26a0..d704a8f829c8 100644 --- a/fs/bcachefs/io_write.c +++ b/fs/bcachefs/io_write.c @@ -795,7 +795,7 @@ static int bch2_write_decrypt(struct bch_write_op *op) * checksum: */ csum = bch2_checksum_bio(c, op->crc.csum_type, nonce, &op->wbio.bio); - if (bch2_crc_cmp(op->crc.csum, csum)) + if (bch2_crc_cmp(op->crc.csum, csum) && !c->opts.no_data_io) return -EIO; ret = bch2_encrypt_bio(c, op->crc.csum_type, nonce, &op->wbio.bio); diff --git a/fs/bcachefs/journal.c b/fs/bcachefs/journal.c index 5b5d69f2316b..23a9b7845d11 100644 --- a/fs/bcachefs/journal.c +++ b/fs/bcachefs/journal.c @@ -526,36 +526,6 @@ int bch2_journal_res_get_slowpath(struct journal *j, struct journal_res *res, return ret; } -/* journal_preres: */ - -static bool journal_preres_available(struct journal *j, - struct journal_preres *res, - unsigned new_u64s, - unsigned flags) -{ - bool ret = bch2_journal_preres_get_fast(j, res, new_u64s, flags, true); - - if (!ret && mutex_trylock(&j->reclaim_lock)) { - bch2_journal_reclaim(j); - mutex_unlock(&j->reclaim_lock); - } - - return ret; -} - -int __bch2_journal_preres_get(struct journal *j, - struct journal_preres *res, - unsigned new_u64s, - unsigned flags) -{ - int ret; - - closure_wait_event(&j->preres_wait, - (ret = bch2_journal_error(j)) || - journal_preres_available(j, res, new_u64s, flags)); - return ret; -} - /* journal_entry_res: */ void bch2_journal_entry_res_resize(struct journal *j, @@ -1306,7 +1276,6 @@ void __bch2_journal_debug_to_text(struct printbuf *out, struct journal *j) prt_printf(out, "last_seq:\t\t%llu\n", journal_last_seq(j)); prt_printf(out, "last_seq_ondisk:\t%llu\n", j->last_seq_ondisk); prt_printf(out, "flushed_seq_ondisk:\t%llu\n", j->flushed_seq_ondisk); - prt_printf(out, "prereserved:\t\t%u/%u\n", j->prereserved.reserved, j->prereserved.remaining); prt_printf(out, "watermark:\t\t%s\n", bch2_watermarks[j->watermark]); prt_printf(out, "each entry reserved:\t%u\n", j->entry_u64s_reserved); prt_printf(out, "nr flush writes:\t%llu\n", j->nr_flush_writes); diff --git a/fs/bcachefs/journal.h b/fs/bcachefs/journal.h index 011711e99c8d..c85d01cf4948 100644 --- a/fs/bcachefs/journal.h +++ b/fs/bcachefs/journal.h @@ -395,104 +395,6 @@ out: return 0; } -/* journal_preres: */ - -static inline void journal_set_watermark(struct journal *j) -{ - union journal_preres_state s = READ_ONCE(j->prereserved); - unsigned watermark = BCH_WATERMARK_stripe; - - if (fifo_free(&j->pin) < j->pin.size / 4) - watermark = max_t(unsigned, watermark, BCH_WATERMARK_copygc); - if (fifo_free(&j->pin) < j->pin.size / 8) - watermark = max_t(unsigned, watermark, BCH_WATERMARK_reclaim); - - if (s.reserved > s.remaining) - watermark = max_t(unsigned, watermark, BCH_WATERMARK_copygc); - if (!s.remaining) - watermark = max_t(unsigned, watermark, BCH_WATERMARK_reclaim); - - if (watermark == j->watermark) - return; - - swap(watermark, j->watermark); - if (watermark > j->watermark) - journal_wake(j); -} - -static inline void bch2_journal_preres_put(struct journal *j, - struct journal_preres *res) -{ - union journal_preres_state s = { .reserved = res->u64s }; - - if (!res->u64s) - return; - - s.v = atomic64_sub_return(s.v, &j->prereserved.counter); - res->u64s = 0; - - if (unlikely(s.waiting)) { - clear_bit(ilog2((((union journal_preres_state) { .waiting = 1 }).v)), - (unsigned long *) &j->prereserved.v); - closure_wake_up(&j->preres_wait); - } - - if (s.reserved <= s.remaining && j->watermark) - journal_set_watermark(j); -} - -int __bch2_journal_preres_get(struct journal *, - struct journal_preres *, unsigned, unsigned); - -static inline int bch2_journal_preres_get_fast(struct journal *j, - struct journal_preres *res, - unsigned new_u64s, - unsigned flags, - bool set_waiting) -{ - int d = new_u64s - res->u64s; - union journal_preres_state old, new; - u64 v = atomic64_read(&j->prereserved.counter); - enum bch_watermark watermark = flags & BCH_WATERMARK_MASK; - int ret; - - do { - old.v = new.v = v; - ret = 0; - - if (watermark == BCH_WATERMARK_reclaim || - new.reserved + d < new.remaining) { - new.reserved += d; - ret = 1; - } else if (set_waiting && !new.waiting) - new.waiting = true; - else - return 0; - } while ((v = atomic64_cmpxchg(&j->prereserved.counter, - old.v, new.v)) != old.v); - - if (ret) - res->u64s += d; - return ret; -} - -static inline int bch2_journal_preres_get(struct journal *j, - struct journal_preres *res, - unsigned new_u64s, - unsigned flags) -{ - if (new_u64s <= res->u64s) - return 0; - - if (bch2_journal_preres_get_fast(j, res, new_u64s, flags, false)) - return 0; - - if (flags & JOURNAL_RES_GET_NONBLOCK) - return -BCH_ERR_journal_preres_get_blocked; - - return __bch2_journal_preres_get(j, res, new_u64s, flags); -} - /* journal_entry_res: */ void bch2_journal_entry_res_resize(struct journal *, diff --git a/fs/bcachefs/journal_io.c b/fs/bcachefs/journal_io.c index f4bc2cdbfdd7..786a09285509 100644 --- a/fs/bcachefs/journal_io.c +++ b/fs/bcachefs/journal_io.c @@ -1079,6 +1079,12 @@ found: if (ja->bucket_seq[ja->cur_idx] && ja->sectors_free == ca->mi.bucket_size) { +#if 0 + /* + * Debug code for ZNS support, where we (probably) want to be + * correlated where we stopped in the journal to the zone write + * points: + */ bch_err(c, "ja->sectors_free == ca->mi.bucket_size"); bch_err(c, "cur_idx %u/%u", ja->cur_idx, ja->nr); for (i = 0; i < 3; i++) { @@ -1086,6 +1092,7 @@ found: bch_err(c, "bucket_seq[%u] = %llu", idx, ja->bucket_seq[idx]); } +#endif ja->sectors_free = 0; } diff --git a/fs/bcachefs/journal_reclaim.c b/fs/bcachefs/journal_reclaim.c index 9a584aaaa2eb..e63c6eda86af 100644 --- a/fs/bcachefs/journal_reclaim.c +++ b/fs/bcachefs/journal_reclaim.c @@ -50,16 +50,21 @@ unsigned bch2_journal_dev_buckets_available(struct journal *j, return available; } -static void journal_set_remaining(struct journal *j, unsigned u64s_remaining) +static inline void journal_set_watermark(struct journal *j, bool low_on_space) { - union journal_preres_state old, new; - u64 v = atomic64_read(&j->prereserved.counter); + unsigned watermark = BCH_WATERMARK_stripe; - do { - old.v = new.v = v; - new.remaining = u64s_remaining; - } while ((v = atomic64_cmpxchg(&j->prereserved.counter, - old.v, new.v)) != old.v); + if (low_on_space) + watermark = max_t(unsigned, watermark, BCH_WATERMARK_reclaim); + if (fifo_free(&j->pin) < j->pin.size / 4) + watermark = max_t(unsigned, watermark, BCH_WATERMARK_reclaim); + + if (watermark == j->watermark) + return; + + swap(watermark, j->watermark); + if (watermark > j->watermark) + journal_wake(j); } static struct journal_space @@ -162,7 +167,6 @@ void bch2_journal_space_available(struct journal *j) struct bch_fs *c = container_of(j, struct bch_fs, journal); struct bch_dev *ca; unsigned clean, clean_ondisk, total; - s64 u64s_remaining = 0; unsigned max_entry_size = min(j->buf[0].buf_size >> 9, j->buf[1].buf_size >> 9); unsigned i, nr_online = 0, nr_devs_want; @@ -222,16 +226,10 @@ void bch2_journal_space_available(struct journal *j) else clear_bit(JOURNAL_MAY_SKIP_FLUSH, &j->flags); - u64s_remaining = (u64) clean << 6; - u64s_remaining -= (u64) total << 3; - u64s_remaining = max(0LL, u64s_remaining); - u64s_remaining /= 4; - u64s_remaining = min_t(u64, u64s_remaining, U32_MAX); + journal_set_watermark(j, clean * 4 <= total); out: j->cur_entry_sectors = !ret ? j->space[journal_space_discarded].next_entry : 0; j->cur_entry_error = ret; - journal_set_remaining(j, u64s_remaining); - journal_set_watermark(j); if (!ret) journal_wake(j); @@ -555,11 +553,6 @@ static u64 journal_seq_to_flush(struct journal *j) /* Try to keep the journal at most half full: */ nr_buckets = ja->nr / 2; - /* And include pre-reservations: */ - nr_buckets += DIV_ROUND_UP(j->prereserved.reserved, - (ca->mi.bucket_size << 6) - - journal_entry_overhead(j)); - nr_buckets = min(nr_buckets, ja->nr); bucket_to_flush = (ja->cur_idx + nr_buckets) % ja->nr; @@ -638,10 +631,7 @@ static int __bch2_journal_reclaim(struct journal *j, bool direct, bool kicked) msecs_to_jiffies(c->opts.journal_reclaim_delay))) min_nr = 1; - if (j->prereserved.reserved * 4 > j->prereserved.remaining) - min_nr = 1; - - if (fifo_free(&j->pin) <= 32) + if (j->watermark != BCH_WATERMARK_stripe) min_nr = 1; if (atomic_read(&c->btree_cache.dirty) * 2 > c->btree_cache.used) @@ -652,8 +642,6 @@ static int __bch2_journal_reclaim(struct journal *j, bool direct, bool kicked) trace_and_count(c, journal_reclaim_start, c, direct, kicked, min_nr, min_key_cache, - j->prereserved.reserved, - j->prereserved.remaining, atomic_read(&c->btree_cache.dirty), c->btree_cache.used, atomic_long_read(&c->btree_key_cache.nr_dirty), diff --git a/fs/bcachefs/journal_types.h b/fs/bcachefs/journal_types.h index 42504e16acb6..a756b69582e3 100644 --- a/fs/bcachefs/journal_types.h +++ b/fs/bcachefs/journal_types.h @@ -76,14 +76,6 @@ struct journal_res { u64 seq; }; -/* - * For reserving space in the journal prior to getting a reservation on a - * particular journal entry: - */ -struct journal_preres { - unsigned u64s; -}; - union journal_res_state { struct { atomic64_t counter; @@ -104,22 +96,6 @@ union journal_res_state { }; }; -union journal_preres_state { - struct { - atomic64_t counter; - }; - - struct { - u64 v; - }; - - struct { - u64 waiting:1, - reserved:31, - remaining:32; - }; -}; - /* bytes: */ #define JOURNAL_ENTRY_SIZE_MIN (64U << 10) /* 64k */ #define JOURNAL_ENTRY_SIZE_MAX (4U << 20) /* 4M */ @@ -180,8 +156,6 @@ struct journal { union journal_res_state reservations; enum bch_watermark watermark; - union journal_preres_state prereserved; - } __aligned(SMP_CACHE_BYTES); unsigned long flags; diff --git a/fs/bcachefs/six.c b/fs/bcachefs/six.c index b775cf0fb7cb..97790445e67a 100644 --- a/fs/bcachefs/six.c +++ b/fs/bcachefs/six.c @@ -163,8 +163,11 @@ static int __do_six_trylock(struct six_lock *lock, enum six_lock_type type, this_cpu_sub(*lock->readers, !ret); preempt_enable(); - if (!ret && (old & SIX_LOCK_WAITING_write)) - ret = -1 - SIX_LOCK_write; + if (!ret) { + smp_mb(); + if (atomic_read(&lock->state) & SIX_LOCK_WAITING_write) + ret = -1 - SIX_LOCK_write; + } } else if (type == SIX_LOCK_write && lock->readers) { if (try) { atomic_add(SIX_LOCK_HELD_write, &lock->state); diff --git a/fs/bcachefs/subvolume_types.h b/fs/bcachefs/subvolume_types.h index 86833445af20..2d2e66a4e468 100644 --- a/fs/bcachefs/subvolume_types.h +++ b/fs/bcachefs/subvolume_types.h @@ -20,7 +20,7 @@ struct snapshot_t { }; struct snapshot_table { - struct snapshot_t s[0]; + DECLARE_FLEX_ARRAY(struct snapshot_t, s); }; typedef struct { diff --git a/fs/bcachefs/trace.h b/fs/bcachefs/trace.h index 893304a1f06e..7857671159b4 100644 --- a/fs/bcachefs/trace.h +++ b/fs/bcachefs/trace.h @@ -196,10 +196,9 @@ DEFINE_EVENT(bio, journal_write, TRACE_EVENT(journal_reclaim_start, TP_PROTO(struct bch_fs *c, bool direct, bool kicked, u64 min_nr, u64 min_key_cache, - u64 prereserved, u64 prereserved_total, u64 btree_cache_dirty, u64 btree_cache_total, u64 btree_key_cache_dirty, u64 btree_key_cache_total), - TP_ARGS(c, direct, kicked, min_nr, min_key_cache, prereserved, prereserved_total, + TP_ARGS(c, direct, kicked, min_nr, min_key_cache, btree_cache_dirty, btree_cache_total, btree_key_cache_dirty, btree_key_cache_total), @@ -209,8 +208,6 @@ TRACE_EVENT(journal_reclaim_start, __field(bool, kicked ) __field(u64, min_nr ) __field(u64, min_key_cache ) - __field(u64, prereserved ) - __field(u64, prereserved_total ) __field(u64, btree_cache_dirty ) __field(u64, btree_cache_total ) __field(u64, btree_key_cache_dirty ) @@ -223,22 +220,18 @@ TRACE_EVENT(journal_reclaim_start, __entry->kicked = kicked; __entry->min_nr = min_nr; __entry->min_key_cache = min_key_cache; - __entry->prereserved = prereserved; - __entry->prereserved_total = prereserved_total; __entry->btree_cache_dirty = btree_cache_dirty; __entry->btree_cache_total = btree_cache_total; __entry->btree_key_cache_dirty = btree_key_cache_dirty; __entry->btree_key_cache_total = btree_key_cache_total; ), - TP_printk("%d,%d direct %u kicked %u min %llu key cache %llu prereserved %llu/%llu btree cache %llu/%llu key cache %llu/%llu", + TP_printk("%d,%d direct %u kicked %u min %llu key cache %llu btree cache %llu/%llu key cache %llu/%llu", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->direct, __entry->kicked, __entry->min_nr, __entry->min_key_cache, - __entry->prereserved, - __entry->prereserved_total, __entry->btree_cache_dirty, __entry->btree_cache_total, __entry->btree_key_cache_dirty, diff --git a/fs/bcachefs/xattr.c b/fs/bcachefs/xattr.c index a39ff0c296ec..79d982674c18 100644 --- a/fs/bcachefs/xattr.c +++ b/fs/bcachefs/xattr.c @@ -552,6 +552,14 @@ static int bch2_xattr_bcachefs_set(const struct xattr_handler *handler, s.v = v + 1; s.defined = true; } else { + /* + * Check if this option was set on the parent - if so, switched + * back to inheriting from the parent: + * + * rename() also has to deal with keeping inherited options up + * to date - see bch2_reinherit_attrs() + */ + spin_lock(&dentry->d_lock); if (!IS_ROOT(dentry)) { struct bch_inode_info *dir = to_bch_ei(d_inode(dentry->d_parent)); @@ -560,6 +568,7 @@ static int bch2_xattr_bcachefs_set(const struct xattr_handler *handler, } else { s.v = 0; } + spin_unlock(&dentry->d_lock); s.defined = false; } diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index a25dd3d20008..b0e8774c435a 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -998,6 +998,14 @@ static int ecryptfs_getattr_link(struct mnt_idmap *idmap, return rc; } +static int ecryptfs_do_getattr(const struct path *path, struct kstat *stat, + u32 request_mask, unsigned int flags) +{ + if (flags & AT_GETATTR_NOSEC) + return vfs_getattr_nosec(path, stat, request_mask, flags); + return vfs_getattr(path, stat, request_mask, flags); +} + static int ecryptfs_getattr(struct mnt_idmap *idmap, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int flags) @@ -1006,8 +1014,8 @@ static int ecryptfs_getattr(struct mnt_idmap *idmap, struct kstat lower_stat; int rc; - rc = vfs_getattr(ecryptfs_dentry_to_lower_path(dentry), &lower_stat, - request_mask, flags); + rc = ecryptfs_do_getattr(ecryptfs_dentry_to_lower_path(dentry), + &lower_stat, request_mask, flags); if (!rc) { fsstack_copy_attr_all(d_inode(dentry), ecryptfs_inode_to_lower(d_inode(dentry))); diff --git a/fs/erofs/Kconfig b/fs/erofs/Kconfig index e540648dedc2..1d318f85232d 100644 --- a/fs/erofs/Kconfig +++ b/fs/erofs/Kconfig @@ -21,7 +21,7 @@ config EROFS_FS performance under extremely memory pressure without extra cost. See the documentation at <file:Documentation/filesystems/erofs.rst> - for more details. + and the web pages at <https://erofs.docs.kernel.org> for more details. If unsure, say N. diff --git a/fs/erofs/data.c b/fs/erofs/data.c index 029c761670bf..c98aeda8abb2 100644 --- a/fs/erofs/data.c +++ b/fs/erofs/data.c @@ -220,7 +220,7 @@ int erofs_map_dev(struct super_block *sb, struct erofs_map_dev *map) up_read(&devs->rwsem); return 0; } - map->m_bdev = dif->bdev_handle->bdev; + map->m_bdev = dif->bdev_handle ? dif->bdev_handle->bdev : NULL; map->m_daxdev = dif->dax_dev; map->m_dax_part_off = dif->dax_part_off; map->m_fscache = dif->fscache; @@ -238,7 +238,8 @@ int erofs_map_dev(struct super_block *sb, struct erofs_map_dev *map) if (map->m_pa >= startoff && map->m_pa < startoff + length) { map->m_pa -= startoff; - map->m_bdev = dif->bdev_handle->bdev; + map->m_bdev = dif->bdev_handle ? + dif->bdev_handle->bdev : NULL; map->m_daxdev = dif->dax_dev; map->m_dax_part_off = dif->dax_part_off; map->m_fscache = dif->fscache; diff --git a/fs/erofs/inode.c b/fs/erofs/inode.c index b8ad05b4509d..14a79d3226ab 100644 --- a/fs/erofs/inode.c +++ b/fs/erofs/inode.c @@ -15,11 +15,11 @@ static void *erofs_read_inode(struct erofs_buf *buf, struct erofs_sb_info *sbi = EROFS_SB(sb); struct erofs_inode *vi = EROFS_I(inode); const erofs_off_t inode_loc = erofs_iloc(inode); - erofs_blk_t blkaddr, nblks = 0; void *kaddr; struct erofs_inode_compact *dic; struct erofs_inode_extended *die, *copied = NULL; + union erofs_inode_i_u iu; unsigned int ifmt; int err; @@ -35,9 +35,8 @@ static void *erofs_read_inode(struct erofs_buf *buf, dic = kaddr + *ofs; ifmt = le16_to_cpu(dic->i_format); - if (ifmt & ~EROFS_I_ALL) { - erofs_err(inode->i_sb, "unsupported i_format %u of nid %llu", + erofs_err(sb, "unsupported i_format %u of nid %llu", ifmt, vi->nid); err = -EOPNOTSUPP; goto err_out; @@ -45,7 +44,7 @@ static void *erofs_read_inode(struct erofs_buf *buf, vi->datalayout = erofs_inode_datalayout(ifmt); if (vi->datalayout >= EROFS_INODE_DATALAYOUT_MAX) { - erofs_err(inode->i_sb, "unsupported datalayout %u of nid %llu", + erofs_err(sb, "unsupported datalayout %u of nid %llu", vi->datalayout, vi->nid); err = -EOPNOTSUPP; goto err_out; @@ -82,40 +81,15 @@ static void *erofs_read_inode(struct erofs_buf *buf, vi->xattr_isize = erofs_xattr_ibody_size(die->i_xattr_icount); inode->i_mode = le16_to_cpu(die->i_mode); - switch (inode->i_mode & S_IFMT) { - case S_IFREG: - case S_IFDIR: - case S_IFLNK: - vi->raw_blkaddr = le32_to_cpu(die->i_u.raw_blkaddr); - break; - case S_IFCHR: - case S_IFBLK: - inode->i_rdev = - new_decode_dev(le32_to_cpu(die->i_u.rdev)); - break; - case S_IFIFO: - case S_IFSOCK: - inode->i_rdev = 0; - break; - default: - goto bogusimode; - } + iu = die->i_u; i_uid_write(inode, le32_to_cpu(die->i_uid)); i_gid_write(inode, le32_to_cpu(die->i_gid)); set_nlink(inode, le32_to_cpu(die->i_nlink)); - - /* extended inode has its own timestamp */ + /* each extended inode has its own timestamp */ inode_set_ctime(inode, le64_to_cpu(die->i_mtime), le32_to_cpu(die->i_mtime_nsec)); inode->i_size = le64_to_cpu(die->i_size); - - /* total blocks for compressed files */ - if (erofs_inode_is_data_compressed(vi->datalayout)) - nblks = le32_to_cpu(die->i_u.compressed_blocks); - else if (vi->datalayout == EROFS_INODE_CHUNK_BASED) - /* fill chunked inode summary info */ - vi->chunkformat = le16_to_cpu(die->i_u.c.format); kfree(copied); copied = NULL; break; @@ -125,49 +99,51 @@ static void *erofs_read_inode(struct erofs_buf *buf, vi->xattr_isize = erofs_xattr_ibody_size(dic->i_xattr_icount); inode->i_mode = le16_to_cpu(dic->i_mode); - switch (inode->i_mode & S_IFMT) { - case S_IFREG: - case S_IFDIR: - case S_IFLNK: - vi->raw_blkaddr = le32_to_cpu(dic->i_u.raw_blkaddr); - break; - case S_IFCHR: - case S_IFBLK: - inode->i_rdev = - new_decode_dev(le32_to_cpu(dic->i_u.rdev)); - break; - case S_IFIFO: - case S_IFSOCK: - inode->i_rdev = 0; - break; - default: - goto bogusimode; - } + iu = dic->i_u; i_uid_write(inode, le16_to_cpu(dic->i_uid)); i_gid_write(inode, le16_to_cpu(dic->i_gid)); set_nlink(inode, le16_to_cpu(dic->i_nlink)); - /* use build time for compact inodes */ inode_set_ctime(inode, sbi->build_time, sbi->build_time_nsec); inode->i_size = le32_to_cpu(dic->i_size); - if (erofs_inode_is_data_compressed(vi->datalayout)) - nblks = le32_to_cpu(dic->i_u.compressed_blocks); - else if (vi->datalayout == EROFS_INODE_CHUNK_BASED) - vi->chunkformat = le16_to_cpu(dic->i_u.c.format); break; default: - erofs_err(inode->i_sb, - "unsupported on-disk inode version %u of nid %llu", + erofs_err(sb, "unsupported on-disk inode version %u of nid %llu", erofs_inode_version(ifmt), vi->nid); err = -EOPNOTSUPP; goto err_out; } - if (vi->datalayout == EROFS_INODE_CHUNK_BASED) { + switch (inode->i_mode & S_IFMT) { + case S_IFREG: + case S_IFDIR: + case S_IFLNK: + vi->raw_blkaddr = le32_to_cpu(iu.raw_blkaddr); + break; + case S_IFCHR: + case S_IFBLK: + inode->i_rdev = new_decode_dev(le32_to_cpu(iu.rdev)); + break; + case S_IFIFO: + case S_IFSOCK: + inode->i_rdev = 0; + break; + default: + erofs_err(sb, "bogus i_mode (%o) @ nid %llu", inode->i_mode, + vi->nid); + err = -EFSCORRUPTED; + goto err_out; + } + + /* total blocks for compressed files */ + if (erofs_inode_is_data_compressed(vi->datalayout)) { + nblks = le32_to_cpu(iu.compressed_blocks); + } else if (vi->datalayout == EROFS_INODE_CHUNK_BASED) { + /* fill chunked inode summary info */ + vi->chunkformat = le16_to_cpu(iu.c.format); if (vi->chunkformat & ~EROFS_CHUNK_FORMAT_ALL) { - erofs_err(inode->i_sb, - "unsupported chunk format %x of nid %llu", + erofs_err(sb, "unsupported chunk format %x of nid %llu", vi->chunkformat, vi->nid); err = -EOPNOTSUPP; goto err_out; @@ -191,10 +167,6 @@ static void *erofs_read_inode(struct erofs_buf *buf, inode->i_blocks = nblks << (sb->s_blocksize_bits - 9); return kaddr; -bogusimode: - erofs_err(inode->i_sb, "bogus i_mode (%o) @ nid %llu", - inode->i_mode, vi->nid); - err = -EFSCORRUPTED; err_out: DBG_BUGON(1); kfree(copied); diff --git a/fs/inode.c b/fs/inode.c index edcd8a61975f..f238d987dec9 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -215,6 +215,8 @@ int inode_init_always(struct super_block *sb, struct inode *inode) lockdep_set_class_and_name(&mapping->invalidate_lock, &sb->s_type->invalidate_lock_key, "mapping.invalidate_lock"); + if (sb->s_iflags & SB_I_STABLE_WRITES) + mapping_set_stable_writes(mapping); inode->i_private = NULL; inode->i_mapping = mapping; INIT_HLIST_HEAD(&inode->i_dentry); /* buggered by rcu freeing */ diff --git a/fs/libfs.c b/fs/libfs.c index e9440d55073c..c2aa6fd4795c 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -399,6 +399,8 @@ static loff_t offset_dir_llseek(struct file *file, loff_t offset, int whence) return -EINVAL; } + /* In this case, ->private_data is protected by f_pos_lock */ + file->private_data = NULL; return vfs_setpos(file, offset, U32_MAX); } @@ -428,7 +430,7 @@ static bool offset_dir_emit(struct dir_context *ctx, struct dentry *dentry) inode->i_ino, fs_umode_to_dtype(inode->i_mode)); } -static void offset_iterate_dir(struct inode *inode, struct dir_context *ctx) +static void *offset_iterate_dir(struct inode *inode, struct dir_context *ctx) { struct offset_ctx *so_ctx = inode->i_op->get_offset_ctx(inode); XA_STATE(xas, &so_ctx->xa, ctx->pos); @@ -437,7 +439,7 @@ static void offset_iterate_dir(struct inode *inode, struct dir_context *ctx) while (true) { dentry = offset_find_next(&xas); if (!dentry) - break; + return ERR_PTR(-ENOENT); if (!offset_dir_emit(ctx, dentry)) { dput(dentry); @@ -447,6 +449,7 @@ static void offset_iterate_dir(struct inode *inode, struct dir_context *ctx) dput(dentry); ctx->pos = xas.xa_index + 1; } + return NULL; } /** @@ -479,7 +482,12 @@ static int offset_readdir(struct file *file, struct dir_context *ctx) if (!dir_emit_dots(file, ctx)) return 0; - offset_iterate_dir(d_inode(dir), ctx); + /* In this case, ->private_data is protected by f_pos_lock */ + if (ctx->pos == 2) + file->private_data = NULL; + else if (file->private_data == ERR_PTR(-ENOENT)) + return 0; + file->private_data = offset_iterate_dir(d_inode(dir), ctx); return 0; } diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index 929248c6ca84..4cbe0434cbb8 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -84,8 +84,8 @@ int nfsd_net_reply_cache_init(struct nfsd_net *nn); void nfsd_net_reply_cache_destroy(struct nfsd_net *nn); int nfsd_reply_cache_init(struct nfsd_net *); void nfsd_reply_cache_shutdown(struct nfsd_net *); -int nfsd_cache_lookup(struct svc_rqst *rqstp, - struct nfsd_cacherep **cacherep); +int nfsd_cache_lookup(struct svc_rqst *rqstp, unsigned int start, + unsigned int len, struct nfsd_cacherep **cacherep); void nfsd_cache_update(struct svc_rqst *rqstp, struct nfsd_cacherep *rp, int cachetype, __be32 *statp); int nfsd_reply_cache_stats_show(struct seq_file *m, void *v); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 4045c852a450..40415929e2ae 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2804,7 +2804,7 @@ static int client_opens_release(struct inode *inode, struct file *file) /* XXX: alternatively, we could get/drop in seq start/stop */ drop_client(clp); - return 0; + return seq_release(inode, file); } static const struct file_operations client_states_fops = { diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index fd56a52aa5fb..d3273a396659 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -369,33 +369,52 @@ nfsd_reply_cache_scan(struct shrinker *shrink, struct shrink_control *sc) return freed; } -/* - * Walk an xdr_buf and get a CRC for at most the first RC_CSUMLEN bytes +/** + * nfsd_cache_csum - Checksum incoming NFS Call arguments + * @buf: buffer containing a whole RPC Call message + * @start: starting byte of the NFS Call header + * @remaining: size of the NFS Call header, in bytes + * + * Compute a weak checksum of the leading bytes of an NFS procedure + * call header to help verify that a retransmitted Call matches an + * entry in the duplicate reply cache. + * + * To avoid assumptions about how the RPC message is laid out in + * @buf and what else it might contain (eg, a GSS MIC suffix), the + * caller passes us the exact location and length of the NFS Call + * header. + * + * Returns a 32-bit checksum value, as defined in RFC 793. */ -static __wsum -nfsd_cache_csum(struct svc_rqst *rqstp) +static __wsum nfsd_cache_csum(struct xdr_buf *buf, unsigned int start, + unsigned int remaining) { + unsigned int base, len; + struct xdr_buf subbuf; + __wsum csum = 0; + void *p; int idx; - unsigned int base; - __wsum csum; - struct xdr_buf *buf = &rqstp->rq_arg; - const unsigned char *p = buf->head[0].iov_base; - size_t csum_len = min_t(size_t, buf->head[0].iov_len + buf->page_len, - RC_CSUMLEN); - size_t len = min(buf->head[0].iov_len, csum_len); + + if (remaining > RC_CSUMLEN) + remaining = RC_CSUMLEN; + if (xdr_buf_subsegment(buf, &subbuf, start, remaining)) + return csum; /* rq_arg.head first */ - csum = csum_partial(p, len, 0); - csum_len -= len; + if (subbuf.head[0].iov_len) { + len = min_t(unsigned int, subbuf.head[0].iov_len, remaining); + csum = csum_partial(subbuf.head[0].iov_base, len, csum); + remaining -= len; + } /* Continue into page array */ - idx = buf->page_base / PAGE_SIZE; - base = buf->page_base & ~PAGE_MASK; - while (csum_len) { - p = page_address(buf->pages[idx]) + base; - len = min_t(size_t, PAGE_SIZE - base, csum_len); + idx = subbuf.page_base / PAGE_SIZE; + base = subbuf.page_base & ~PAGE_MASK; + while (remaining) { + p = page_address(subbuf.pages[idx]) + base; + len = min_t(unsigned int, PAGE_SIZE - base, remaining); csum = csum_partial(p, len, csum); - csum_len -= len; + remaining -= len; base = 0; ++idx; } @@ -466,6 +485,8 @@ out: /** * nfsd_cache_lookup - Find an entry in the duplicate reply cache * @rqstp: Incoming Call to find + * @start: starting byte in @rqstp->rq_arg of the NFS Call header + * @len: size of the NFS Call header, in bytes * @cacherep: OUT: DRC entry for this request * * Try to find an entry matching the current call in the cache. When none @@ -479,7 +500,8 @@ out: * %RC_REPLY: Reply from cache * %RC_DROPIT: Do not process the request further */ -int nfsd_cache_lookup(struct svc_rqst *rqstp, struct nfsd_cacherep **cacherep) +int nfsd_cache_lookup(struct svc_rqst *rqstp, unsigned int start, + unsigned int len, struct nfsd_cacherep **cacherep) { struct nfsd_net *nn; struct nfsd_cacherep *rp, *found; @@ -495,7 +517,7 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp, struct nfsd_cacherep **cacherep) goto out; } - csum = nfsd_cache_csum(rqstp); + csum = nfsd_cache_csum(&rqstp->rq_arg, start, len); /* * Since the common case is a cache miss followed by an insert, @@ -641,24 +663,17 @@ void nfsd_cache_update(struct svc_rqst *rqstp, struct nfsd_cacherep *rp, return; } -/* - * Copy cached reply to current reply buffer. Should always fit. - * FIXME as reply is in a page, we should just attach the page, and - * keep a refcount.... - */ static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *data) { - struct kvec *vec = &rqstp->rq_res.head[0]; - - if (vec->iov_len + data->iov_len > PAGE_SIZE) { - printk(KERN_WARNING "nfsd: cached reply too large (%zd).\n", - data->iov_len); - return 0; - } - memcpy((char*)vec->iov_base + vec->iov_len, data->iov_base, data->iov_len); - vec->iov_len += data->iov_len; - return 1; + __be32 *p; + + p = xdr_reserve_space(&rqstp->rq_res_stream, data->iov_len); + if (unlikely(!p)) + return false; + memcpy(p, data->iov_base, data->iov_len); + xdr_commit_encode(&rqstp->rq_res_stream); + return true; } /* diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index d6122bb2d167..fe61d9bbcc1f 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -981,6 +981,8 @@ int nfsd_dispatch(struct svc_rqst *rqstp) const struct svc_procedure *proc = rqstp->rq_procinfo; __be32 *statp = rqstp->rq_accept_statp; struct nfsd_cacherep *rp; + unsigned int start, len; + __be32 *nfs_reply; /* * Give the xdr decoder a chance to change this if it wants @@ -988,6 +990,13 @@ int nfsd_dispatch(struct svc_rqst *rqstp) */ rqstp->rq_cachetype = proc->pc_cachetype; + /* + * ->pc_decode advances the argument stream past the NFS + * Call header, so grab the header's starting location and + * size now for the call to nfsd_cache_lookup(). + */ + start = xdr_stream_pos(&rqstp->rq_arg_stream); + len = xdr_stream_remaining(&rqstp->rq_arg_stream); if (!proc->pc_decode(rqstp, &rqstp->rq_arg_stream)) goto out_decode_err; @@ -1001,7 +1010,7 @@ int nfsd_dispatch(struct svc_rqst *rqstp) smp_store_release(&rqstp->rq_status_counter, rqstp->rq_status_counter | 1); rp = NULL; - switch (nfsd_cache_lookup(rqstp, &rp)) { + switch (nfsd_cache_lookup(rqstp, start, len, &rp)) { case RC_DOIT: break; case RC_REPLY: @@ -1010,6 +1019,7 @@ int nfsd_dispatch(struct svc_rqst *rqstp) goto out_dropit; } + nfs_reply = xdr_inline_decode(&rqstp->rq_res_stream, 0); *statp = proc->pc_func(rqstp); if (test_bit(RQ_DROPME, &rqstp->rq_flags)) goto out_update_drop; @@ -1023,7 +1033,7 @@ int nfsd_dispatch(struct svc_rqst *rqstp) */ smp_store_release(&rqstp->rq_status_counter, rqstp->rq_status_counter + 1); - nfsd_cache_update(rqstp, rp, rqstp->rq_cachetype, statp + 1); + nfsd_cache_update(rqstp, rp, rqstp->rq_cachetype, nfs_reply); out_cached_reply: return 1; diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c index 345b8f161ca4..c63b31a460be 100644 --- a/fs/overlayfs/inode.c +++ b/fs/overlayfs/inode.c @@ -171,7 +171,7 @@ int ovl_getattr(struct mnt_idmap *idmap, const struct path *path, type = ovl_path_real(dentry, &realpath); old_cred = ovl_override_creds(dentry->d_sb); - err = vfs_getattr(&realpath, stat, request_mask, flags); + err = ovl_do_getattr(&realpath, stat, request_mask, flags); if (err) goto out; @@ -196,8 +196,8 @@ int ovl_getattr(struct mnt_idmap *idmap, const struct path *path, (!is_dir ? STATX_NLINK : 0); ovl_path_lower(dentry, &realpath); - err = vfs_getattr(&realpath, &lowerstat, - lowermask, flags); + err = ovl_do_getattr(&realpath, &lowerstat, lowermask, + flags); if (err) goto out; @@ -249,8 +249,8 @@ int ovl_getattr(struct mnt_idmap *idmap, const struct path *path, ovl_path_lowerdata(dentry, &realpath); if (realpath.dentry) { - err = vfs_getattr(&realpath, &lowerdatastat, - lowermask, flags); + err = ovl_do_getattr(&realpath, &lowerdatastat, + lowermask, flags); if (err) goto out; } else { diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h index ca88b2636a57..05c3dd597fa8 100644 --- a/fs/overlayfs/overlayfs.h +++ b/fs/overlayfs/overlayfs.h @@ -408,6 +408,14 @@ static inline bool ovl_open_flags_need_copy_up(int flags) return ((OPEN_FMODE(flags) & FMODE_WRITE) || (flags & O_TRUNC)); } +static inline int ovl_do_getattr(const struct path *path, struct kstat *stat, + u32 request_mask, unsigned int flags) +{ + if (flags & AT_GETATTR_NOSEC) + return vfs_getattr_nosec(path, stat, request_mask, flags); + return vfs_getattr(path, stat, request_mask, flags); +} + /* util.c */ int ovl_get_write_access(struct dentry *dentry); void ovl_put_write_access(struct dentry *dentry); diff --git a/fs/overlayfs/params.c b/fs/overlayfs/params.c index ddab9ea267d1..3fe2dde1598f 100644 --- a/fs/overlayfs/params.c +++ b/fs/overlayfs/params.c @@ -430,7 +430,7 @@ static int ovl_parse_param_lowerdir(const char *name, struct fs_context *fc) struct ovl_fs_context *ctx = fc->fs_private; struct ovl_fs_context_layer *l; char *dup = NULL, *iter; - ssize_t nr_lower = 0, nr = 0, nr_data = 0; + ssize_t nr_lower, nr; bool data_layer = false; /* @@ -482,6 +482,7 @@ static int ovl_parse_param_lowerdir(const char *name, struct fs_context *fc) iter = dup; l = ctx->lower; for (nr = 0; nr < nr_lower; nr++, l++) { + ctx->nr++; memset(l, 0, sizeof(*l)); err = ovl_mount_dir(iter, &l->path); @@ -498,10 +499,10 @@ static int ovl_parse_param_lowerdir(const char *name, struct fs_context *fc) goto out_put; if (data_layer) - nr_data++; + ctx->nr_data++; /* Calling strchr() again would overrun. */ - if ((nr + 1) == nr_lower) + if (ctx->nr == nr_lower) break; err = -EINVAL; @@ -511,7 +512,7 @@ static int ovl_parse_param_lowerdir(const char *name, struct fs_context *fc) * This is a regular layer so we require that * there are no data layers. */ - if ((ctx->nr_data + nr_data) > 0) { + if (ctx->nr_data > 0) { pr_err("regular lower layers cannot follow data lower layers"); goto out_put; } @@ -524,8 +525,6 @@ static int ovl_parse_param_lowerdir(const char *name, struct fs_context *fc) data_layer = true; iter++; } - ctx->nr = nr_lower; - ctx->nr_data += nr_data; kfree(dup); return 0; diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c index 50a201e9cd39..c3f020ca13a8 100644 --- a/fs/overlayfs/util.c +++ b/fs/overlayfs/util.c @@ -978,7 +978,7 @@ int ovl_set_protattr(struct inode *inode, struct dentry *upper, return 0; } -/** +/* * Caller must hold a reference to inode to prevent it from being freed while * it is marked inuse. */ diff --git a/fs/smb/client/cifs_spnego.c b/fs/smb/client/cifs_spnego.c index 6f3285f1dfee..af7849e5974f 100644 --- a/fs/smb/client/cifs_spnego.c +++ b/fs/smb/client/cifs_spnego.c @@ -64,8 +64,8 @@ struct key_type cifs_spnego_key_type = { * strlen(";sec=ntlmsspi") */ #define MAX_MECH_STR_LEN 13 -/* strlen of "host=" */ -#define HOST_KEY_LEN 5 +/* strlen of ";host=" */ +#define HOST_KEY_LEN 6 /* strlen of ";ip4=" or ";ip6=" */ #define IP_KEY_LEN 5 diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index 57c2a7df3457..f896f60c924b 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -2065,6 +2065,12 @@ void __cifs_put_smb_ses(struct cifs_ses *ses) ses->chans[i].server = NULL; } + /* we now account for primary channel in iface->refcount */ + if (ses->chans[0].iface) { + kref_put(&ses->chans[0].iface->refcount, release_iface); + ses->chans[0].server = NULL; + } + sesInfoFree(ses); cifs_put_tcp_session(server, 0); } diff --git a/fs/smb/client/sess.c b/fs/smb/client/sess.c index 0bb2ac929061..8b2d7c1ca428 100644 --- a/fs/smb/client/sess.c +++ b/fs/smb/client/sess.c @@ -322,28 +322,32 @@ cifs_disable_secondary_channels(struct cifs_ses *ses) iface = ses->chans[i].iface; server = ses->chans[i].server; + /* + * remove these references first, since we need to unlock + * the chan_lock here, since iface_lock is a higher lock + */ + ses->chans[i].iface = NULL; + ses->chans[i].server = NULL; + spin_unlock(&ses->chan_lock); + if (iface) { spin_lock(&ses->iface_lock); kref_put(&iface->refcount, release_iface); - ses->chans[i].iface = NULL; iface->num_channels--; if (iface->weight_fulfilled) iface->weight_fulfilled--; spin_unlock(&ses->iface_lock); } - spin_unlock(&ses->chan_lock); - if (server && !server->terminate) { - server->terminate = true; - cifs_signal_cifsd_for_reconnect(server, false); - } - spin_lock(&ses->chan_lock); - if (server) { - ses->chans[i].server = NULL; + if (!server->terminate) { + server->terminate = true; + cifs_signal_cifsd_for_reconnect(server, false); + } cifs_put_tcp_session(server, false); } + spin_lock(&ses->chan_lock); } done: diff --git a/fs/smb/client/smb2transport.c b/fs/smb/client/smb2transport.c index 84ea67301303..5a3ca62d2f07 100644 --- a/fs/smb/client/smb2transport.c +++ b/fs/smb/client/smb2transport.c @@ -458,6 +458,8 @@ generate_smb3signingkey(struct cifs_ses *ses, ptriplet->encryption.context, ses->smb3encryptionkey, SMB3_ENC_DEC_KEY_SIZE); + if (rc) + return rc; rc = generate_key(ses, ptriplet->decryption.label, ptriplet->decryption.context, ses->smb3decryptionkey, @@ -466,9 +468,6 @@ generate_smb3signingkey(struct cifs_ses *ses, return rc; } - if (rc) - return rc; - #ifdef CONFIG_CIFS_DEBUG_DUMP_KEYS cifs_dbg(VFS, "%s: dumping generated AES session keys\n", __func__); /* diff --git a/fs/stat.c b/fs/stat.c index 24bb0209e459..f721d26ec3f7 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -133,7 +133,8 @@ int vfs_getattr_nosec(const struct path *path, struct kstat *stat, idmap = mnt_idmap(path->mnt); if (inode->i_op->getattr) return inode->i_op->getattr(idmap, path, stat, - request_mask, query_flags); + request_mask, + query_flags | AT_GETATTR_NOSEC); generic_fillattr(idmap, request_mask, inode, stat); return 0; @@ -166,6 +167,9 @@ int vfs_getattr(const struct path *path, struct kstat *stat, { int retval; + if (WARN_ON_ONCE(query_flags & AT_GETATTR_NOSEC)) + return -EPERM; + retval = security_inode_getattr(path); if (retval) return retval; diff --git a/fs/xfs/Kconfig b/fs/xfs/Kconfig index ed0bc8cbc703..567fb37274d3 100644 --- a/fs/xfs/Kconfig +++ b/fs/xfs/Kconfig @@ -147,7 +147,7 @@ config XFS_ONLINE_SCRUB_STATS bool "XFS online metadata check usage data collection" default y depends on XFS_ONLINE_SCRUB - select XFS_DEBUG + select DEBUG_FS help If you say Y here, the kernel will gather usage data about the online metadata check subsystem. This includes the number diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index 3069194527dd..100ab5931b31 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -2275,16 +2275,37 @@ xfs_alloc_min_freelist( ASSERT(mp->m_alloc_maxlevels > 0); + /* + * For a btree shorter than the maximum height, the worst case is that + * every level gets split and a new level is added, then while inserting + * another entry to refill the AGFL, every level under the old root gets + * split again. This is: + * + * (full height split reservation) + (AGFL refill split height) + * = (current height + 1) + (current height - 1) + * = (new height) + (new height - 2) + * = 2 * new height - 2 + * + * For a btree of maximum height, the worst case is that every level + * under the root gets split, then while inserting another entry to + * refill the AGFL, every level under the root gets split again. This is + * also: + * + * 2 * (current height - 1) + * = 2 * (new height - 1) + * = 2 * new height - 2 + */ + /* space needed by-bno freespace btree */ min_free = min_t(unsigned int, levels[XFS_BTNUM_BNOi] + 1, - mp->m_alloc_maxlevels); + mp->m_alloc_maxlevels) * 2 - 2; /* space needed by-size freespace btree */ min_free += min_t(unsigned int, levels[XFS_BTNUM_CNTi] + 1, - mp->m_alloc_maxlevels); + mp->m_alloc_maxlevels) * 2 - 2; /* space needed reverse mapping used space btree */ if (xfs_has_rmapbt(mp)) min_free += min_t(unsigned int, levels[XFS_BTNUM_RMAPi] + 1, - mp->m_rmap_maxlevels); + mp->m_rmap_maxlevels) * 2 - 2; return min_free; } diff --git a/fs/xfs/libxfs/xfs_defer.c b/fs/xfs/libxfs/xfs_defer.c index bcfb6a4203cd..f71679ce23b9 100644 --- a/fs/xfs/libxfs/xfs_defer.c +++ b/fs/xfs/libxfs/xfs_defer.c @@ -245,21 +245,18 @@ xfs_defer_create_intents( return ret; } -/* Abort all the intents that were committed. */ STATIC void -xfs_defer_trans_abort( - struct xfs_trans *tp, - struct list_head *dop_pending) +xfs_defer_pending_abort( + struct xfs_mount *mp, + struct list_head *dop_list) { struct xfs_defer_pending *dfp; const struct xfs_defer_op_type *ops; - trace_xfs_defer_trans_abort(tp, _RET_IP_); - /* Abort intent items that don't have a done item. */ - list_for_each_entry(dfp, dop_pending, dfp_list) { + list_for_each_entry(dfp, dop_list, dfp_list) { ops = defer_op_types[dfp->dfp_type]; - trace_xfs_defer_pending_abort(tp->t_mountp, dfp); + trace_xfs_defer_pending_abort(mp, dfp); if (dfp->dfp_intent && !dfp->dfp_done) { ops->abort_intent(dfp->dfp_intent); dfp->dfp_intent = NULL; @@ -267,6 +264,16 @@ xfs_defer_trans_abort( } } +/* Abort all the intents that were committed. */ +STATIC void +xfs_defer_trans_abort( + struct xfs_trans *tp, + struct list_head *dop_pending) +{ + trace_xfs_defer_trans_abort(tp, _RET_IP_); + xfs_defer_pending_abort(tp->t_mountp, dop_pending); +} + /* * Capture resources that the caller said not to release ("held") when the * transaction commits. Caller is responsible for zero-initializing @dres. @@ -756,12 +763,13 @@ xfs_defer_ops_capture( /* Release all resources that we used to capture deferred ops. */ void -xfs_defer_ops_capture_free( +xfs_defer_ops_capture_abort( struct xfs_mount *mp, struct xfs_defer_capture *dfc) { unsigned short i; + xfs_defer_pending_abort(mp, &dfc->dfc_dfops); xfs_defer_cancel_list(mp, &dfc->dfc_dfops); for (i = 0; i < dfc->dfc_held.dr_bufs; i++) @@ -802,7 +810,7 @@ xfs_defer_ops_capture_and_commit( /* Commit the transaction and add the capture structure to the list. */ error = xfs_trans_commit(tp); if (error) { - xfs_defer_ops_capture_free(mp, dfc); + xfs_defer_ops_capture_abort(mp, dfc); return error; } diff --git a/fs/xfs/libxfs/xfs_defer.h b/fs/xfs/libxfs/xfs_defer.h index 114a3a4930a3..8788ad5f6a73 100644 --- a/fs/xfs/libxfs/xfs_defer.h +++ b/fs/xfs/libxfs/xfs_defer.h @@ -121,7 +121,7 @@ int xfs_defer_ops_capture_and_commit(struct xfs_trans *tp, struct list_head *capture_list); void xfs_defer_ops_continue(struct xfs_defer_capture *d, struct xfs_trans *tp, struct xfs_defer_resources *dres); -void xfs_defer_ops_capture_free(struct xfs_mount *mp, +void xfs_defer_ops_capture_abort(struct xfs_mount *mp, struct xfs_defer_capture *d); void xfs_defer_resources_rele(struct xfs_defer_resources *dres); diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c index 543f3748c2a3..137a65bda95d 100644 --- a/fs/xfs/libxfs/xfs_inode_buf.c +++ b/fs/xfs/libxfs/xfs_inode_buf.c @@ -510,6 +510,9 @@ xfs_dinode_verify( if (mode && nextents + naextents > nblocks) return __this_address; + if (nextents + naextents == 0 && nblocks != 0) + return __this_address; + if (S_ISDIR(mode) && nextents > mp->m_dir_geo->max_extents) return __this_address; diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 3dc47937da5d..3beb470f1892 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -569,6 +569,14 @@ extern void xfs_setup_inode(struct xfs_inode *ip); extern void xfs_setup_iops(struct xfs_inode *ip); extern void xfs_diflags_to_iflags(struct xfs_inode *ip, bool init); +static inline void xfs_update_stable_writes(struct xfs_inode *ip) +{ + if (bdev_stable_writes(xfs_inode_buftarg(ip)->bt_bdev)) + mapping_set_stable_writes(VFS_I(ip)->i_mapping); + else + mapping_clear_stable_writes(VFS_I(ip)->i_mapping); +} + /* * When setting up a newly allocated inode, we need to call * xfs_finish_inode_setup() once the inode is fully instantiated at diff --git a/fs/xfs/xfs_inode_item_recover.c b/fs/xfs/xfs_inode_item_recover.c index 0e5dba2343ea..144198a6b270 100644 --- a/fs/xfs/xfs_inode_item_recover.c +++ b/fs/xfs/xfs_inode_item_recover.c @@ -286,6 +286,7 @@ xlog_recover_inode_commit_pass2( struct xfs_log_dinode *ldip; uint isize; int need_free = 0; + xfs_failaddr_t fa; if (item->ri_buf[0].i_len == sizeof(struct xfs_inode_log_format)) { in_f = item->ri_buf[0].i_addr; @@ -369,24 +370,26 @@ xlog_recover_inode_commit_pass2( * superblock flag to determine whether we need to look at di_flushiter * to skip replay when the on disk inode is newer than the log one */ - if (!xfs_has_v3inodes(mp) && - ldip->di_flushiter < be16_to_cpu(dip->di_flushiter)) { - /* - * Deal with the wrap case, DI_MAX_FLUSH is less - * than smaller numbers - */ - if (be16_to_cpu(dip->di_flushiter) == DI_MAX_FLUSH && - ldip->di_flushiter < (DI_MAX_FLUSH >> 1)) { - /* do nothing */ - } else { - trace_xfs_log_recover_inode_skip(log, in_f); - error = 0; - goto out_release; + if (!xfs_has_v3inodes(mp)) { + if (ldip->di_flushiter < be16_to_cpu(dip->di_flushiter)) { + /* + * Deal with the wrap case, DI_MAX_FLUSH is less + * than smaller numbers + */ + if (be16_to_cpu(dip->di_flushiter) == DI_MAX_FLUSH && + ldip->di_flushiter < (DI_MAX_FLUSH >> 1)) { + /* do nothing */ + } else { + trace_xfs_log_recover_inode_skip(log, in_f); + error = 0; + goto out_release; + } } + + /* Take the opportunity to reset the flush iteration count */ + ldip->di_flushiter = 0; } - /* Take the opportunity to reset the flush iteration count */ - ldip->di_flushiter = 0; if (unlikely(S_ISREG(ldip->di_mode))) { if ((ldip->di_format != XFS_DINODE_FMT_EXTENTS) && @@ -528,8 +531,19 @@ out_owner_change: (dip->di_mode != 0)) error = xfs_recover_inode_owner_change(mp, dip, in_f, buffer_list); - /* re-generate the checksum. */ + /* re-generate the checksum and validate the recovered inode. */ xfs_dinode_calc_crc(log->l_mp, dip); + fa = xfs_dinode_verify(log->l_mp, in_f->ilf_ino, dip); + if (fa) { + XFS_CORRUPTION_ERROR( + "Bad dinode after recovery", + XFS_ERRLEVEL_LOW, mp, dip, sizeof(*dip)); + xfs_alert(mp, + "Metadata corruption detected at %pS, inode 0x%llx", + fa, in_f->ilf_ino); + error = -EFSCORRUPTED; + goto out_release; + } ASSERT(bp->b_mount == mp); bp->b_flags |= _XBF_LOGRECOVERY; diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c index a82470e027f7..6c3919687ea6 100644 --- a/fs/xfs/xfs_ioctl.c +++ b/fs/xfs/xfs_ioctl.c @@ -1121,23 +1121,25 @@ xfs_ioctl_setattr_xflags( struct fileattr *fa) { struct xfs_mount *mp = ip->i_mount; + bool rtflag = (fa->fsx_xflags & FS_XFLAG_REALTIME); uint64_t i_flags2; - /* Can't change realtime flag if any extents are allocated. */ - if ((ip->i_df.if_nextents || ip->i_delayed_blks) && - XFS_IS_REALTIME_INODE(ip) != (fa->fsx_xflags & FS_XFLAG_REALTIME)) - return -EINVAL; + if (rtflag != XFS_IS_REALTIME_INODE(ip)) { + /* Can't change realtime flag if any extents are allocated. */ + if (ip->i_df.if_nextents || ip->i_delayed_blks) + return -EINVAL; + } - /* If realtime flag is set then must have realtime device */ - if (fa->fsx_xflags & FS_XFLAG_REALTIME) { + if (rtflag) { + /* If realtime flag is set then must have realtime device */ if (mp->m_sb.sb_rblocks == 0 || mp->m_sb.sb_rextsize == 0 || xfs_extlen_to_rtxmod(mp, ip->i_extsize)) return -EINVAL; - } - /* Clear reflink if we are actually able to set the rt flag. */ - if ((fa->fsx_xflags & FS_XFLAG_REALTIME) && xfs_is_reflink_inode(ip)) - ip->i_diflags2 &= ~XFS_DIFLAG2_REFLINK; + /* Clear reflink if we are actually able to set the rt flag. */ + if (xfs_is_reflink_inode(ip)) + ip->i_diflags2 &= ~XFS_DIFLAG2_REFLINK; + } /* diflags2 only valid for v3 inodes. */ i_flags2 = xfs_flags2diflags2(ip, fa->fsx_xflags); @@ -1148,6 +1150,14 @@ xfs_ioctl_setattr_xflags( ip->i_diflags2 = i_flags2; xfs_diflags_to_iflags(ip, false); + + /* + * Make the stable writes flag match that of the device the inode + * resides on when flipping the RT flag. + */ + if (rtflag != XFS_IS_REALTIME_INODE(ip) && S_ISREG(VFS_I(ip)->i_mode)) + xfs_update_stable_writes(ip); + xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); XFS_STATS_INC(mp, xs_ig_attrchg); diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index fdfda4fba12b..a0d77f5f512e 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -1299,6 +1299,13 @@ xfs_setup_inode( mapping_set_gfp_mask(inode->i_mapping, (gfp_mask & ~(__GFP_FS))); /* + * For real-time inodes update the stable write flags to that of the RT + * device instead of the data device. + */ + if (S_ISREG(inode->i_mode) && XFS_IS_REALTIME_INODE(ip)) + xfs_update_stable_writes(ip); + + /* * If there is no attribute fork no ACL can exist on this inode, * and it can't have any file capabilities attached to it either. */ diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 51c100c86177..ee206facf0dc 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -1893,9 +1893,7 @@ xlog_write_iclog( * the buffer manually, the code needs to be kept in sync * with the I/O completion path. */ - xlog_state_done_syncing(iclog); - up(&iclog->ic_sema); - return; + goto sync; } /* @@ -1925,20 +1923,17 @@ xlog_write_iclog( * avoid shutdown re-entering this path and erroring out again. */ if (log->l_targ != log->l_mp->m_ddev_targp && - blkdev_issue_flush(log->l_mp->m_ddev_targp->bt_bdev)) { - xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR); - return; - } + blkdev_issue_flush(log->l_mp->m_ddev_targp->bt_bdev)) + goto shutdown; } if (iclog->ic_flags & XLOG_ICL_NEED_FUA) iclog->ic_bio.bi_opf |= REQ_FUA; iclog->ic_flags &= ~(XLOG_ICL_NEED_FLUSH | XLOG_ICL_NEED_FUA); - if (xlog_map_iclog_data(&iclog->ic_bio, iclog->ic_data, count)) { - xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR); - return; - } + if (xlog_map_iclog_data(&iclog->ic_bio, iclog->ic_data, count)) + goto shutdown; + if (is_vmalloc_addr(iclog->ic_data)) flush_kernel_vmap_range(iclog->ic_data, count); @@ -1959,6 +1954,12 @@ xlog_write_iclog( } submit_bio(&iclog->ic_bio); + return; +shutdown: + xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR); +sync: + xlog_state_done_syncing(iclog); + up(&iclog->ic_sema); } /* diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 13b94d2e605b..a1e18b24971a 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -2511,7 +2511,7 @@ xlog_abort_defer_ops( list_for_each_entry_safe(dfc, next, capture_list, dfc_list) { list_del_init(&dfc->dfc_list); - xfs_defer_ops_capture_free(mp, dfc); + xfs_defer_ops_capture_abort(mp, dfc); } } diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c index 658edee8381d..e5b62dc28466 100644 --- a/fs/xfs/xfs_reflink.c +++ b/fs/xfs/xfs_reflink.c @@ -784,6 +784,7 @@ xfs_reflink_end_cow_extent( } } del = got; + xfs_trim_extent(&del, *offset_fsb, end_fsb - *offset_fsb); /* Grab the corresponding mapping in the data fork. */ nmaps = 1; |