diff options
Diffstat (limited to 'fs')
288 files changed, 6453 insertions, 4322 deletions
diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index e51e7d88980a..1d847a939f29 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -98,14 +98,25 @@ static struct file_system_type anon_inode_fs_type = { .kill_sb = kill_anon_super, }; -static struct inode *anon_inode_make_secure_inode( - const char *name, - const struct inode *context_inode) +/** + * anon_inode_make_secure_inode - allocate an anonymous inode with security context + * @sb: [in] Superblock to allocate from + * @name: [in] Name of the class of the newfile (e.g., "secretmem") + * @context_inode: + * [in] Optional parent inode for security inheritance + * + * The function ensures proper security initialization through the LSM hook + * security_inode_init_security_anon(). + * + * Return: Pointer to new inode on success, ERR_PTR on failure. + */ +struct inode *anon_inode_make_secure_inode(struct super_block *sb, const char *name, + const struct inode *context_inode) { struct inode *inode; int error; - inode = alloc_anon_inode(anon_inode_mnt->mnt_sb); + inode = alloc_anon_inode(sb); if (IS_ERR(inode)) return inode; inode->i_flags &= ~S_PRIVATE; @@ -118,6 +129,7 @@ static struct inode *anon_inode_make_secure_inode( } return inode; } +EXPORT_SYMBOL_GPL_FOR_MODULES(anon_inode_make_secure_inode, "kvm"); static struct file *__anon_inode_getfile(const char *name, const struct file_operations *fops, @@ -132,7 +144,8 @@ static struct file *__anon_inode_getfile(const char *name, return ERR_PTR(-ENOENT); if (make_inode) { - inode = anon_inode_make_secure_inode(name, context_inode); + inode = anon_inode_make_secure_inode(anon_inode_mnt->mnt_sb, + name, context_inode); if (IS_ERR(inode)) { file = ERR_CAST(inode); goto err; diff --git a/fs/bcachefs/alloc_background.c b/fs/bcachefs/alloc_background.c index 173e81c2bbcb..66de46318620 100644 --- a/fs/bcachefs/alloc_background.c +++ b/fs/bcachefs/alloc_background.c @@ -21,7 +21,6 @@ #include "error.h" #include "lru.h" #include "recovery.h" -#include "trace.h" #include "varint.h" #include <linux/kthread.h> @@ -337,11 +336,10 @@ void bch2_alloc_v4_swab(struct bkey_s k) a->stripe_sectors = swab32(a->stripe_sectors); } -void bch2_alloc_to_text(struct printbuf *out, struct bch_fs *c, struct bkey_s_c k) +static inline void __bch2_alloc_v4_to_text(struct printbuf *out, struct bch_fs *c, + unsigned dev, const struct bch_alloc_v4 *a) { - struct bch_alloc_v4 _a; - const struct bch_alloc_v4 *a = bch2_alloc_to_v4(k, &_a); - struct bch_dev *ca = c ? bch2_dev_bucket_tryget_noerror(c, k.k->p) : NULL; + struct bch_dev *ca = c ? bch2_dev_tryget_noerror(c, dev) : NULL; prt_newline(out); printbuf_indent_add(out, 2); @@ -369,6 +367,19 @@ void bch2_alloc_to_text(struct printbuf *out, struct bch_fs *c, struct bkey_s_c bch2_dev_put(ca); } +void bch2_alloc_to_text(struct printbuf *out, struct bch_fs *c, struct bkey_s_c k) +{ + struct bch_alloc_v4 _a; + const struct bch_alloc_v4 *a = bch2_alloc_to_v4(k, &_a); + + __bch2_alloc_v4_to_text(out, c, k.k->p.inode, a); +} + +void bch2_alloc_v4_to_text(struct printbuf *out, struct bch_fs *c, struct bkey_s_c k) +{ + __bch2_alloc_v4_to_text(out, c, k.k->p.inode, bkey_s_c_to_alloc_v4(k).v); +} + void __bch2_alloc_to_v4(struct bkey_s_c k, struct bch_alloc_v4 *out) { if (k.k->type == KEY_TYPE_alloc_v4) { @@ -697,8 +708,8 @@ static int __need_discard_or_freespace_err(struct btree_trans *trans, set ? "" : "un", bch2_btree_id_str(btree), buf.buf); - if (ret == -BCH_ERR_fsck_ignore || - ret == -BCH_ERR_fsck_errors_not_fixed) + if (bch2_err_matches(ret, BCH_ERR_fsck_ignore) || + bch2_err_matches(ret, BCH_ERR_fsck_errors_not_fixed)) ret = 0; printbuf_exit(&buf); @@ -854,7 +865,7 @@ int bch2_trigger_alloc(struct btree_trans *trans, struct bch_dev *ca = bch2_dev_bucket_tryget(c, new.k->p); if (!ca) - return -BCH_ERR_trigger_alloc; + return bch_err_throw(c, trigger_alloc); struct bch_alloc_v4 old_a_convert; const struct bch_alloc_v4 *old_a = bch2_alloc_to_v4(old, &old_a_convert); @@ -988,14 +999,11 @@ int bch2_trigger_alloc(struct btree_trans *trans, } if (new_a->gen != old_a->gen) { - rcu_read_lock(); + guard(rcu)(); u8 *gen = bucket_gen(ca, new.k->p.offset); - if (unlikely(!gen)) { - rcu_read_unlock(); + if (unlikely(!gen)) goto invalid_bucket; - } *gen = new_a->gen; - rcu_read_unlock(); } #define eval_state(_a, expr) ({ const struct bch_alloc_v4 *a = _a; expr; }) @@ -1021,15 +1029,12 @@ int bch2_trigger_alloc(struct btree_trans *trans, } if ((flags & BTREE_TRIGGER_gc) && (flags & BTREE_TRIGGER_insert)) { - rcu_read_lock(); + guard(rcu)(); struct bucket *g = gc_bucket(ca, new.k->p.offset); - if (unlikely(!g)) { - rcu_read_unlock(); + if (unlikely(!g)) goto invalid_bucket; - } g->gen_valid = 1; g->gen = new_a->gen; - rcu_read_unlock(); } err: fsck_err: @@ -1039,7 +1044,7 @@ fsck_err: invalid_bucket: bch2_fs_inconsistent(c, "reference to invalid bucket\n%s", (bch2_bkey_val_to_text(&buf, c, new.s_c), buf.buf)); - ret = -BCH_ERR_trigger_alloc; + ret = bch_err_throw(c, trigger_alloc); goto err; } @@ -1105,13 +1110,12 @@ static bool next_bucket(struct bch_fs *c, struct bch_dev **ca, struct bpos *buck bucket->offset = 0; } - rcu_read_lock(); + guard(rcu)(); *ca = __bch2_next_dev_idx(c, bucket->inode, NULL); if (*ca) { *bucket = POS((*ca)->dev_idx, (*ca)->mi.first_bucket); bch2_dev_get(*ca); } - rcu_read_unlock(); return *ca != NULL; } @@ -1402,6 +1406,9 @@ int bch2_check_discard_freespace_key(struct btree_trans *trans, struct btree_ite : BCH_DATA_free; struct printbuf buf = PRINTBUF; + unsigned fsck_flags = (async_repair ? FSCK_ERR_NO_LOG : 0)| + FSCK_CAN_FIX|FSCK_CAN_IGNORE; + struct bpos bucket = iter->pos; bucket.offset &= ~(~0ULL << 56); u64 genbits = iter->pos.offset & (~0ULL << 56); @@ -1415,9 +1422,10 @@ int bch2_check_discard_freespace_key(struct btree_trans *trans, struct btree_ite return ret; if (!bch2_dev_bucket_exists(c, bucket)) { - if (fsck_err(trans, need_discard_freespace_key_to_invalid_dev_bucket, - "entry in %s btree for nonexistant dev:bucket %llu:%llu", - bch2_btree_id_str(iter->btree_id), bucket.inode, bucket.offset)) + if (__fsck_err(trans, fsck_flags, + need_discard_freespace_key_to_invalid_dev_bucket, + "entry in %s btree for nonexistant dev:bucket %llu:%llu", + bch2_btree_id_str(iter->btree_id), bucket.inode, bucket.offset)) goto delete; ret = 1; goto out; @@ -1429,7 +1437,8 @@ int bch2_check_discard_freespace_key(struct btree_trans *trans, struct btree_ite if (a->data_type != state || (state == BCH_DATA_free && genbits != alloc_freespace_genbits(*a))) { - if (fsck_err(trans, need_discard_freespace_key_bad, + if (__fsck_err(trans, fsck_flags, + need_discard_freespace_key_bad, "%s\nincorrectly set at %s:%llu:%llu:0 (free %u, genbits %llu should be %llu)", (bch2_bkey_val_to_text(&buf, c, alloc_k), buf.buf), bch2_btree_id_str(iter->btree_id), @@ -1454,7 +1463,7 @@ delete: ret = bch2_btree_bit_mod_iter(trans, iter, false) ?: bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc) ?: - -BCH_ERR_transaction_restart_commit; + bch_err_throw(c, transaction_restart_commit); goto out; } else { /* @@ -1777,14 +1786,16 @@ int bch2_check_alloc_to_lru_refs(struct bch_fs *c) static int discard_in_flight_add(struct bch_dev *ca, u64 bucket, bool in_progress) { + struct bch_fs *c = ca->fs; int ret; mutex_lock(&ca->discard_buckets_in_flight_lock); - darray_for_each(ca->discard_buckets_in_flight, i) - if (i->bucket == bucket) { - ret = -BCH_ERR_EEXIST_discard_in_flight_add; - goto out; - } + struct discard_in_flight *i = + darray_find_p(ca->discard_buckets_in_flight, i, i->bucket == bucket); + if (i) { + ret = bch_err_throw(c, EEXIST_discard_in_flight_add); + goto out; + } ret = darray_push(&ca->discard_buckets_in_flight, ((struct discard_in_flight) { .in_progress = in_progress, @@ -1798,14 +1809,11 @@ out: static void discard_in_flight_remove(struct bch_dev *ca, u64 bucket) { mutex_lock(&ca->discard_buckets_in_flight_lock); - darray_for_each(ca->discard_buckets_in_flight, i) - if (i->bucket == bucket) { - BUG_ON(!i->in_progress); - darray_remove_item(&ca->discard_buckets_in_flight, i); - goto found; - } - BUG(); -found: + struct discard_in_flight *i = + darray_find_p(ca->discard_buckets_in_flight, i, i->bucket == bucket); + BUG_ON(!i || !i->in_progress); + + darray_remove_item(&ca->discard_buckets_in_flight, i); mutex_unlock(&ca->discard_buckets_in_flight_lock); } @@ -2504,7 +2512,7 @@ void bch2_recalc_capacity(struct bch_fs *c) lockdep_assert_held(&c->state_lock); - rcu_read_lock(); + guard(rcu)(); for_each_member_device_rcu(c, ca, NULL) { struct block_device *bdev = READ_ONCE(ca->disk_sb.bdev); if (bdev) @@ -2549,7 +2557,6 @@ void bch2_recalc_capacity(struct bch_fs *c) bucket_size_max = max_t(unsigned, bucket_size_max, ca->mi.bucket_size); } - rcu_read_unlock(); bch2_set_ra_pages(c, ra_pages); @@ -2574,10 +2581,9 @@ u64 bch2_min_rw_member_capacity(struct bch_fs *c) { u64 ret = U64_MAX; - rcu_read_lock(); + guard(rcu)(); for_each_rw_member_rcu(c, ca) ret = min(ret, ca->mi.nbuckets * ca->mi.bucket_size); - rcu_read_unlock(); return ret; } diff --git a/fs/bcachefs/alloc_background.h b/fs/bcachefs/alloc_background.h index 4f94c6a661bf..0cc5adc55b6f 100644 --- a/fs/bcachefs/alloc_background.h +++ b/fs/bcachefs/alloc_background.h @@ -13,11 +13,9 @@ static inline bool bch2_dev_bucket_exists(struct bch_fs *c, struct bpos pos) { - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, pos.inode); - bool ret = ca && bucket_valid(ca, pos.offset); - rcu_read_unlock(); - return ret; + return ca && bucket_valid(ca, pos.offset); } static inline u64 bucket_to_u64(struct bpos bucket) @@ -253,6 +251,7 @@ int bch2_alloc_v4_validate(struct bch_fs *, struct bkey_s_c, struct bkey_validate_context); void bch2_alloc_v4_swab(struct bkey_s); void bch2_alloc_to_text(struct printbuf *, struct bch_fs *, struct bkey_s_c); +void bch2_alloc_v4_to_text(struct printbuf *, struct bch_fs *, struct bkey_s_c); #define bch2_bkey_ops_alloc ((struct bkey_ops) { \ .key_validate = bch2_alloc_v1_validate, \ @@ -277,7 +276,7 @@ void bch2_alloc_to_text(struct printbuf *, struct bch_fs *, struct bkey_s_c); #define bch2_bkey_ops_alloc_v4 ((struct bkey_ops) { \ .key_validate = bch2_alloc_v4_validate, \ - .val_to_text = bch2_alloc_to_text, \ + .val_to_text = bch2_alloc_v4_to_text, \ .swab = bch2_alloc_v4_swab, \ .trigger = bch2_trigger_alloc, \ .min_val_size = 48, \ diff --git a/fs/bcachefs/alloc_foreground.c b/fs/bcachefs/alloc_foreground.c index 1a52c12c51ae..b58525ec7b4d 100644 --- a/fs/bcachefs/alloc_foreground.c +++ b/fs/bcachefs/alloc_foreground.c @@ -69,10 +69,9 @@ const char * const bch2_watermarks[] = { void bch2_reset_alloc_cursors(struct bch_fs *c) { - rcu_read_lock(); + guard(rcu)(); for_each_member_device_rcu(c, ca, NULL) memset(ca->alloc_cursor, 0, sizeof(ca->alloc_cursor)); - rcu_read_unlock(); } static void bch2_open_bucket_hash_add(struct bch_fs *c, struct open_bucket *ob) @@ -166,9 +165,8 @@ static void open_bucket_free_unused(struct bch_fs *c, struct open_bucket *ob) ARRAY_SIZE(c->open_buckets_partial)); spin_lock(&c->freelist_lock); - rcu_read_lock(); - bch2_dev_rcu(c, ob->dev)->nr_partial_buckets++; - rcu_read_unlock(); + scoped_guard(rcu) + bch2_dev_rcu(c, ob->dev)->nr_partial_buckets++; ob->on_partial_list = true; c->open_buckets_partial[c->open_buckets_partial_nr++] = @@ -229,7 +227,7 @@ static struct open_bucket *__try_alloc_bucket(struct bch_fs *c, track_event_change(&c->times[BCH_TIME_blocked_allocate_open_bucket], true); spin_unlock(&c->freelist_lock); - return ERR_PTR(-BCH_ERR_open_buckets_empty); + return ERR_PTR(bch_err_throw(c, open_buckets_empty)); } /* Recheck under lock: */ @@ -513,7 +511,8 @@ again: bch2_dev_usage_read_fast(ca, &req->usage); avail = dev_buckets_free(ca, req->usage, req->watermark); - if (req->usage.buckets[BCH_DATA_need_discard] > avail) + if (req->usage.buckets[BCH_DATA_need_discard] > + min(avail, ca->mi.nbuckets >> 7)) bch2_dev_do_discards(ca); if (req->usage.buckets[BCH_DATA_need_gc_gens] > avail) @@ -535,7 +534,7 @@ again: track_event_change(&c->times[BCH_TIME_blocked_allocate], true); - ob = ERR_PTR(-BCH_ERR_freelist_empty); + ob = ERR_PTR(bch_err_throw(c, freelist_empty)); goto err; } @@ -560,7 +559,7 @@ alloc: } err: if (!ob) - ob = ERR_PTR(-BCH_ERR_no_buckets_found); + ob = ERR_PTR(bch_err_throw(c, no_buckets_found)); if (!IS_ERR(ob)) ob->data_type = req->data_type; @@ -603,18 +602,18 @@ static int __dev_stripe_cmp(struct dev_stripe_state *stripe, #define dev_stripe_cmp(l, r) __dev_stripe_cmp(stripe, l, r) -struct dev_alloc_list bch2_dev_alloc_list(struct bch_fs *c, - struct dev_stripe_state *stripe, - struct bch_devs_mask *devs) +void bch2_dev_alloc_list(struct bch_fs *c, + struct dev_stripe_state *stripe, + struct bch_devs_mask *devs, + struct dev_alloc_list *ret) { - struct dev_alloc_list ret = { .nr = 0 }; - unsigned i; + ret->nr = 0; + unsigned i; for_each_set_bit(i, devs->d, BCH_SB_MEMBERS_MAX) - ret.data[ret.nr++] = i; + ret->data[ret->nr++] = i; - bubble_sort(ret.data, ret.nr, dev_stripe_cmp); - return ret; + bubble_sort(ret->data, ret->nr, dev_stripe_cmp); } static const u64 stripe_clock_hand_rescale = 1ULL << 62; /* trigger rescale at */ @@ -705,18 +704,19 @@ static int add_new_bucket(struct bch_fs *c, return 0; } -int bch2_bucket_alloc_set_trans(struct btree_trans *trans, - struct alloc_request *req, - struct dev_stripe_state *stripe, - struct closure *cl) +inline int bch2_bucket_alloc_set_trans(struct btree_trans *trans, + struct alloc_request *req, + struct dev_stripe_state *stripe, + struct closure *cl) { struct bch_fs *c = trans->c; - int ret = -BCH_ERR_insufficient_devices; + int ret = 0; BUG_ON(req->nr_effective >= req->nr_replicas); - struct dev_alloc_list devs_sorted = bch2_dev_alloc_list(c, stripe, &req->devs_may_alloc); - darray_for_each(devs_sorted, i) { + bch2_dev_alloc_list(c, stripe, &req->devs_may_alloc, &req->devs_sorted); + + darray_for_each(req->devs_sorted, i) { req->ca = bch2_dev_tryget_noerror(c, *i); if (!req->ca) continue; @@ -739,13 +739,16 @@ int bch2_bucket_alloc_set_trans(struct btree_trans *trans, continue; } - if (add_new_bucket(c, req, ob)) { - ret = 0; + ret = add_new_bucket(c, req, ob); + if (ret) break; - } } - return ret; + if (ret == 1) + return 0; + if (ret) + return ret; + return bch_err_throw(c, insufficient_devices); } /* Allocate from stripes: */ @@ -776,9 +779,9 @@ static int bucket_alloc_from_stripe(struct btree_trans *trans, if (!h) return 0; - struct dev_alloc_list devs_sorted = - bch2_dev_alloc_list(c, &req->wp->stripe, &req->devs_may_alloc); - darray_for_each(devs_sorted, i) + bch2_dev_alloc_list(c, &req->wp->stripe, &req->devs_may_alloc, &req->devs_sorted); + + darray_for_each(req->devs_sorted, i) for (unsigned ec_idx = 0; ec_idx < h->s->nr_data; ec_idx++) { if (!h->s->blocks[ec_idx]) continue; @@ -872,9 +875,8 @@ static int bucket_alloc_set_partial(struct bch_fs *c, i); ob->on_partial_list = false; - rcu_read_lock(); - bch2_dev_rcu(c, ob->dev)->nr_partial_buckets--; - rcu_read_unlock(); + scoped_guard(rcu) + bch2_dev_rcu(c, ob->dev)->nr_partial_buckets--; ret = add_new_bucket(c, req, ob); if (ret) @@ -1056,9 +1058,8 @@ void bch2_open_buckets_stop(struct bch_fs *c, struct bch_dev *ca, ob->on_partial_list = false; - rcu_read_lock(); - bch2_dev_rcu(c, ob->dev)->nr_partial_buckets--; - rcu_read_unlock(); + scoped_guard(rcu) + bch2_dev_rcu(c, ob->dev)->nr_partial_buckets--; spin_unlock(&c->freelist_lock); bch2_open_bucket_put(c, ob); @@ -1086,14 +1087,11 @@ static struct write_point *__writepoint_find(struct hlist_head *head, { struct write_point *wp; - rcu_read_lock(); + guard(rcu)(); hlist_for_each_entry_rcu(wp, head, node) if (wp->write_point == write_point) - goto out; - wp = NULL; -out: - rcu_read_unlock(); - return wp; + return wp; + return NULL; } static inline bool too_many_writepoints(struct bch_fs *c, unsigned factor) @@ -1104,7 +1102,7 @@ static inline bool too_many_writepoints(struct bch_fs *c, unsigned factor) return stranded * factor > free; } -static bool try_increase_writepoints(struct bch_fs *c) +static noinline bool try_increase_writepoints(struct bch_fs *c) { struct write_point *wp; @@ -1117,7 +1115,7 @@ static bool try_increase_writepoints(struct bch_fs *c) return true; } -static bool try_decrease_writepoints(struct btree_trans *trans, unsigned old_nr) +static noinline bool try_decrease_writepoints(struct btree_trans *trans, unsigned old_nr) { struct bch_fs *c = trans->c; struct write_point *wp; @@ -1379,11 +1377,11 @@ err: goto retry; if (cl && bch2_err_matches(ret, BCH_ERR_open_buckets_empty)) - ret = -BCH_ERR_bucket_alloc_blocked; + ret = bch_err_throw(c, bucket_alloc_blocked); if (cl && !(flags & BCH_WRITE_alloc_nowait) && bch2_err_matches(ret, BCH_ERR_freelist_empty)) - ret = -BCH_ERR_bucket_alloc_blocked; + ret = bch_err_throw(c, bucket_alloc_blocked); return ret; } @@ -1637,19 +1635,16 @@ static noinline void bch2_print_allocator_stuck(struct bch_fs *c) bch2_printbuf_make_room(&buf, 4096); - rcu_read_lock(); buf.atomic++; - - for_each_online_member_rcu(c, ca) { - prt_printf(&buf, "Dev %u:\n", ca->dev_idx); - printbuf_indent_add(&buf, 2); - bch2_dev_alloc_debug_to_text(&buf, ca); - printbuf_indent_sub(&buf, 2); - prt_newline(&buf); - } - + scoped_guard(rcu) + for_each_online_member_rcu(c, ca) { + prt_printf(&buf, "Dev %u:\n", ca->dev_idx); + printbuf_indent_add(&buf, 2); + bch2_dev_alloc_debug_to_text(&buf, ca); + printbuf_indent_sub(&buf, 2); + prt_newline(&buf); + } --buf.atomic; - rcu_read_unlock(); prt_printf(&buf, "Copygc debug:\n"); printbuf_indent_add(&buf, 2); diff --git a/fs/bcachefs/alloc_foreground.h b/fs/bcachefs/alloc_foreground.h index 2e01c7b61ed1..1b3fc8460096 100644 --- a/fs/bcachefs/alloc_foreground.h +++ b/fs/bcachefs/alloc_foreground.h @@ -42,6 +42,7 @@ struct alloc_request { struct bch_devs_mask devs_may_alloc; /* bch2_bucket_alloc_set_trans(): */ + struct dev_alloc_list devs_sorted; struct bch_dev_usage usage; /* bch2_bucket_alloc_trans(): */ @@ -71,9 +72,10 @@ struct alloc_request { struct bch_devs_mask scratch_devs_may_alloc; }; -struct dev_alloc_list bch2_dev_alloc_list(struct bch_fs *, - struct dev_stripe_state *, - struct bch_devs_mask *); +void bch2_dev_alloc_list(struct bch_fs *, + struct dev_stripe_state *, + struct bch_devs_mask *, + struct dev_alloc_list *); void bch2_dev_stripe_increment(struct bch_dev *, struct dev_stripe_state *); static inline struct bch_dev *ob_dev(struct bch_fs *c, struct open_bucket *ob) diff --git a/fs/bcachefs/backpointers.c b/fs/bcachefs/backpointers.c index cde7dd115267..77d93beb3c8f 100644 --- a/fs/bcachefs/backpointers.c +++ b/fs/bcachefs/backpointers.c @@ -48,17 +48,19 @@ void bch2_backpointer_to_text(struct printbuf *out, struct bch_fs *c, struct bke { struct bkey_s_c_backpointer bp = bkey_s_c_to_backpointer(k); - rcu_read_lock(); - struct bch_dev *ca = bch2_dev_rcu_noerror(c, bp.k->p.inode); - if (ca) { - u32 bucket_offset; - struct bpos bucket = bp_pos_to_bucket_and_offset(ca, bp.k->p, &bucket_offset); - rcu_read_unlock(); + struct bch_dev *ca; + u32 bucket_offset; + struct bpos bucket; + scoped_guard(rcu) { + ca = bch2_dev_rcu_noerror(c, bp.k->p.inode); + if (ca) + bucket = bp_pos_to_bucket_and_offset(ca, bp.k->p, &bucket_offset); + } + + if (ca) prt_printf(out, "bucket=%llu:%llu:%u ", bucket.inode, bucket.offset, bucket_offset); - } else { - rcu_read_unlock(); + else prt_printf(out, "sector=%llu:%llu ", bp.k->p.inode, bp.k->p.offset >> MAX_EXTENT_COMPRESS_RATIO_SHIFT); - } bch2_btree_id_level_to_text(out, bp.v->btree_id, bp.v->level); prt_str(out, " data_type="); @@ -140,7 +142,7 @@ static noinline int backpointer_mod_err(struct btree_trans *trans, } if (!will_check && __bch2_inconsistent_error(c, &buf)) - ret = -BCH_ERR_erofs_unfixed_errors; + ret = bch_err_throw(c, erofs_unfixed_errors); bch_err(c, "%s", buf.buf); printbuf_exit(&buf); @@ -293,7 +295,7 @@ static struct btree *__bch2_backpointer_get_node(struct btree_trans *trans, return b; if (btree_node_will_make_reachable(b)) { - b = ERR_PTR(-BCH_ERR_backpointer_to_overwritten_btree_node); + b = ERR_PTR(bch_err_throw(c, backpointer_to_overwritten_btree_node)); } else { int ret = backpointer_target_not_found(trans, bp, bkey_i_to_s_c(&b->key), last_flushed, commit); @@ -591,6 +593,7 @@ check_existing_bp: bkey_for_each_ptr(other_extent_ptrs, ptr) if (ptr->dev == bp->k.p.inode && dev_ptr_stale_rcu(ca, ptr)) { + rcu_read_unlock(); ret = drop_dev_and_update(trans, other_bp.v->btree_id, other_extent, bp->k.p.inode); if (ret) @@ -648,7 +651,7 @@ check_existing_bp: prt_newline(&buf); bch2_bkey_val_to_text(&buf, c, other_extent); bch_err(c, "%s", buf.buf); - ret = -BCH_ERR_fsck_repair_unimplemented; + ret = bch_err_throw(c, fsck_repair_unimplemented); goto err; missing: printbuf_reset(&buf); @@ -679,26 +682,23 @@ static int check_extent_to_backpointers(struct btree_trans *trans, if (p.ptr.dev == BCH_SB_MEMBER_INVALID) continue; - rcu_read_lock(); - struct bch_dev *ca = bch2_dev_rcu_noerror(c, p.ptr.dev); - if (!ca) { - rcu_read_unlock(); - continue; - } + bool empty; + { + /* scoped_guard() is a loop, so it breaks continue */ + guard(rcu)(); + struct bch_dev *ca = bch2_dev_rcu_noerror(c, p.ptr.dev); + if (!ca) + continue; - if (p.ptr.cached && dev_ptr_stale_rcu(ca, &p.ptr)) { - rcu_read_unlock(); - continue; - } + if (p.ptr.cached && dev_ptr_stale_rcu(ca, &p.ptr)) + continue; - u64 b = PTR_BUCKET_NR(ca, &p.ptr); - if (!bch2_bucket_bitmap_test(&ca->bucket_backpointer_mismatch, b)) { - rcu_read_unlock(); - continue; - } + u64 b = PTR_BUCKET_NR(ca, &p.ptr); + if (!bch2_bucket_bitmap_test(&ca->bucket_backpointer_mismatch, b)) + continue; - bool empty = bch2_bucket_bitmap_test(&ca->bucket_backpointer_empty, b); - rcu_read_unlock(); + empty = bch2_bucket_bitmap_test(&ca->bucket_backpointer_empty, b); + } struct bkey_i_backpointer bp; bch2_extent_ptr_to_bp(c, btree, level, k, p, entry, &bp); @@ -953,7 +953,7 @@ static int check_bucket_backpointer_mismatch(struct btree_trans *trans, struct b sectors[ALLOC_cached] > a->cached_sectors || sectors[ALLOC_stripe] > a->stripe_sectors) { ret = check_bucket_backpointers_to_extents(trans, ca, alloc_k.k->p) ?: - -BCH_ERR_transaction_restart_nested; + bch_err_throw(c, transaction_restart_nested); goto err; } @@ -981,7 +981,7 @@ static bool backpointer_node_has_missing(struct bch_fs *c, struct bkey_s_c k) case KEY_TYPE_btree_ptr_v2: { bool ret = false; - rcu_read_lock(); + guard(rcu)(); struct bpos pos = bkey_s_c_to_btree_ptr_v2(k).v->min_key; while (pos.inode <= k.k->p.inode) { if (pos.inode >= c->sb.nr_devices) @@ -1009,7 +1009,6 @@ static bool backpointer_node_has_missing(struct bch_fs *c, struct bkey_s_c k) next: pos = SPOS(pos.inode + 1, 0, 0); } - rcu_read_unlock(); return ret; } @@ -1352,7 +1351,7 @@ static int bch2_bucket_bitmap_set(struct bch_dev *ca, struct bucket_bitmap *b, u b->buckets = kvcalloc(BITS_TO_LONGS(ca->mi.nbuckets), sizeof(unsigned long), GFP_KERNEL); if (!b->buckets) - return -BCH_ERR_ENOMEM_backpointer_mismatches_bitmap; + return bch_err_throw(ca->fs, ENOMEM_backpointer_mismatches_bitmap); } b->nr += !__test_and_set_bit(bit, b->buckets); @@ -1361,7 +1360,8 @@ static int bch2_bucket_bitmap_set(struct bch_dev *ca, struct bucket_bitmap *b, u return 0; } -int bch2_bucket_bitmap_resize(struct bucket_bitmap *b, u64 old_size, u64 new_size) +int bch2_bucket_bitmap_resize(struct bch_dev *ca, struct bucket_bitmap *b, + u64 old_size, u64 new_size) { scoped_guard(mutex, &b->lock) { if (!b->buckets) @@ -1370,7 +1370,7 @@ int bch2_bucket_bitmap_resize(struct bucket_bitmap *b, u64 old_size, u64 new_siz unsigned long *n = kvcalloc(BITS_TO_LONGS(new_size), sizeof(unsigned long), GFP_KERNEL); if (!n) - return -BCH_ERR_ENOMEM_backpointer_mismatches_bitmap; + return bch_err_throw(ca->fs, ENOMEM_backpointer_mismatches_bitmap); memcpy(n, b->buckets, BITS_TO_LONGS(min(old_size, new_size)) * sizeof(unsigned long)); diff --git a/fs/bcachefs/backpointers.h b/fs/bcachefs/backpointers.h index 6840561084ce..7e71afee1ac0 100644 --- a/fs/bcachefs/backpointers.h +++ b/fs/bcachefs/backpointers.h @@ -53,11 +53,10 @@ static inline struct bpos bp_pos_to_bucket_and_offset(const struct bch_dev *ca, static inline bool bp_pos_to_bucket_nodev_noerror(struct bch_fs *c, struct bpos bp_pos, struct bpos *bucket) { - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, bp_pos.inode); if (ca) *bucket = bp_pos_to_bucket(ca, bp_pos); - rcu_read_unlock(); return ca != NULL; } @@ -195,7 +194,7 @@ static inline bool bch2_bucket_bitmap_test(struct bucket_bitmap *b, u64 i) return bitmap && test_bit(i, bitmap); } -int bch2_bucket_bitmap_resize(struct bucket_bitmap *, u64, u64); +int bch2_bucket_bitmap_resize(struct bch_dev *, struct bucket_bitmap *, u64, u64); void bch2_bucket_bitmap_free(struct bucket_bitmap *); #endif /* _BCACHEFS_BACKPOINTERS_BACKGROUND_H */ diff --git a/fs/bcachefs/bcachefs.h b/fs/bcachefs/bcachefs.h index 7824da2af9d0..ddfacad0f70c 100644 --- a/fs/bcachefs/bcachefs.h +++ b/fs/bcachefs/bcachefs.h @@ -183,6 +183,16 @@ #define pr_fmt(fmt) "%s() " fmt "\n", __func__ #endif +#ifdef CONFIG_BCACHEFS_DEBUG +#define ENUMERATED_REF_DEBUG +#endif + +#ifndef dynamic_fault +#define dynamic_fault(...) 0 +#endif + +#define race_fault(...) dynamic_fault("bcachefs:race") + #include <linux/backing-dev-defs.h> #include <linux/bug.h> #include <linux/bio.h> @@ -219,15 +229,30 @@ #include "time_stats.h" #include "util.h" -#ifdef CONFIG_BCACHEFS_DEBUG -#define ENUMERATED_REF_DEBUG -#endif - -#ifndef dynamic_fault -#define dynamic_fault(...) 0 -#endif +#include "alloc_types.h" +#include "async_objs_types.h" +#include "btree_gc_types.h" +#include "btree_types.h" +#include "btree_node_scan_types.h" +#include "btree_write_buffer_types.h" +#include "buckets_types.h" +#include "buckets_waiting_for_journal_types.h" +#include "clock_types.h" +#include "disk_groups_types.h" +#include "ec_types.h" +#include "enumerated_ref_types.h" +#include "journal_types.h" +#include "keylist_types.h" +#include "quota_types.h" +#include "rebalance_types.h" +#include "recovery_passes_types.h" +#include "replicas_types.h" +#include "sb-members_types.h" +#include "subvolume_types.h" +#include "super_types.h" +#include "thread_with_file_types.h" -#define race_fault(...) dynamic_fault("bcachefs:race") +#include "trace.h" #define count_event(_c, _name) this_cpu_inc((_c)->counters[BCH_COUNTER_##_name]) @@ -271,7 +296,6 @@ do { \ #define bch2_fmt(_c, fmt) bch2_log_msg(_c, fmt "\n") void bch2_print_str(struct bch_fs *, const char *, const char *); -void bch2_print_str_nonblocking(struct bch_fs *, const char *, const char *); __printf(2, 3) void bch2_print_opts(struct bch_opts *, const char *, ...); @@ -380,6 +404,14 @@ do { \ pr_info(fmt, ##__VA_ARGS__); \ } while (0) +static inline int __bch2_err_trace(struct bch_fs *c, int err) +{ + trace_error_throw(c, err, _THIS_IP_); + return err; +} + +#define bch_err_throw(_c, _err) __bch2_err_trace(_c, -BCH_ERR_##_err) + /* Parameters that are useful for debugging, but should always be compiled in: */ #define BCH_DEBUG_PARAMS_ALWAYS() \ BCH_DEBUG_PARAM(key_merging_disabled, \ @@ -486,29 +518,6 @@ enum bch_time_stats { BCH_TIME_STAT_NR }; -#include "alloc_types.h" -#include "async_objs_types.h" -#include "btree_gc_types.h" -#include "btree_types.h" -#include "btree_node_scan_types.h" -#include "btree_write_buffer_types.h" -#include "buckets_types.h" -#include "buckets_waiting_for_journal_types.h" -#include "clock_types.h" -#include "disk_groups_types.h" -#include "ec_types.h" -#include "enumerated_ref_types.h" -#include "journal_types.h" -#include "keylist_types.h" -#include "quota_types.h" -#include "rebalance_types.h" -#include "recovery_passes_types.h" -#include "replicas_types.h" -#include "sb-members_types.h" -#include "subvolume_types.h" -#include "super_types.h" -#include "thread_with_file_types.h" - /* Number of nodes btree coalesce will try to coalesce at once */ #define GC_MERGE_NODES 4U @@ -758,7 +767,8 @@ struct btree_trans_buf { x(sysfs) \ x(btree_write_buffer) \ x(btree_node_scrub) \ - x(async_recovery_passes) + x(async_recovery_passes) \ + x(ioctl_data) enum bch_write_ref { #define x(n) BCH_WRITE_REF_##n, @@ -853,9 +863,7 @@ struct bch_fs { DARRAY(enum bcachefs_metadata_version) incompat_versions_requested; -#ifdef CONFIG_UNICODE struct unicode_map *cf_encoding; -#endif struct bch_sb_handle disk_sb; @@ -1275,4 +1283,13 @@ static inline bool bch2_discard_opt_enabled(struct bch_fs *c, struct bch_dev *ca : ca->mi.discard; } +static inline bool bch2_fs_casefold_enabled(struct bch_fs *c) +{ +#ifdef CONFIG_UNICODE + return !c->opts.casefold_disabled; +#else + return false; +#endif +} + #endif /* _BCACHEFS_H */ diff --git a/fs/bcachefs/btree_cache.c b/fs/bcachefs/btree_cache.c index 8557cbd3d818..83c9860e6b82 100644 --- a/fs/bcachefs/btree_cache.c +++ b/fs/bcachefs/btree_cache.c @@ -85,7 +85,7 @@ void bch2_btree_node_to_freelist(struct bch_fs *c, struct btree *b) six_unlock_intent(&b->c.lock); } -static void __btree_node_data_free(struct btree_cache *bc, struct btree *b) +void __btree_node_data_free(struct btree *b) { BUG_ON(!list_empty(&b->list)); BUG_ON(btree_node_hashed(b)); @@ -112,16 +112,17 @@ static void __btree_node_data_free(struct btree_cache *bc, struct btree *b) munmap(b->aux_data, btree_aux_data_bytes(b)); #endif b->aux_data = NULL; - - btree_node_to_freedlist(bc, b); } static void btree_node_data_free(struct btree_cache *bc, struct btree *b) { BUG_ON(list_empty(&b->list)); list_del_init(&b->list); + + __btree_node_data_free(b); + --bc->nr_freeable; - __btree_node_data_free(bc, b); + btree_node_to_freedlist(bc, b); } static int bch2_btree_cache_cmp_fn(struct rhashtable_compare_arg *arg, @@ -149,7 +150,7 @@ static int btree_node_data_alloc(struct bch_fs *c, struct btree *b, gfp_t gfp) b->data = kvmalloc(btree_buf_bytes(b), gfp); if (!b->data) - return -BCH_ERR_ENOMEM_btree_node_mem_alloc; + return bch_err_throw(c, ENOMEM_btree_node_mem_alloc); #ifdef __KERNEL__ b->aux_data = kvmalloc(btree_aux_data_bytes(b), gfp); #else @@ -162,7 +163,7 @@ static int btree_node_data_alloc(struct bch_fs *c, struct btree *b, gfp_t gfp) if (!b->aux_data) { kvfree(b->data); b->data = NULL; - return -BCH_ERR_ENOMEM_btree_node_mem_alloc; + return bch_err_throw(c, ENOMEM_btree_node_mem_alloc); } return 0; @@ -185,10 +186,7 @@ static struct btree *__btree_node_mem_alloc(struct bch_fs *c, gfp_t gfp) struct btree *__bch2_btree_node_mem_alloc(struct bch_fs *c) { - struct btree_cache *bc = &c->btree_cache; - struct btree *b; - - b = __btree_node_mem_alloc(c, GFP_KERNEL); + struct btree *b = __btree_node_mem_alloc(c, GFP_KERNEL); if (!b) return NULL; @@ -198,8 +196,6 @@ struct btree *__bch2_btree_node_mem_alloc(struct bch_fs *c) } bch2_btree_lock_init(&b->c, 0, GFP_KERNEL); - - __bch2_btree_node_to_freelist(bc, b); return b; } @@ -353,21 +349,21 @@ static int __btree_node_reclaim_checks(struct bch_fs *c, struct btree *b, if (btree_node_noevict(b)) { bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_noevict]++; - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } if (btree_node_write_blocked(b)) { bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_write_blocked]++; - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } if (btree_node_will_make_reachable(b)) { bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_will_make_reachable]++; - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } if (btree_node_dirty(b)) { if (!flush) { bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_dirty]++; - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } if (locked) { @@ -393,7 +389,7 @@ static int __btree_node_reclaim_checks(struct bch_fs *c, struct btree *b, bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_read_in_flight]++; else if (btree_node_write_in_flight(b)) bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_write_in_flight]++; - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } if (locked) @@ -424,13 +420,13 @@ retry_unlocked: if (!six_trylock_intent(&b->c.lock)) { bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_lock_intent]++; - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } if (!six_trylock_write(&b->c.lock)) { bc->not_freed[BCH_BTREE_CACHE_NOT_FREED_lock_write]++; six_unlock_intent(&b->c.lock); - return -BCH_ERR_ENOMEM_btree_node_reclaim; + return bch_err_throw(c, ENOMEM_btree_node_reclaim); } /* recheck under lock */ @@ -524,7 +520,8 @@ restart: --touched;; } else if (!btree_node_reclaim(c, b)) { __bch2_btree_node_hash_remove(bc, b); - __btree_node_data_free(bc, b); + __btree_node_data_free(b); + btree_node_to_freedlist(bc, b); freed++; bc->nr_freed++; @@ -652,9 +649,12 @@ int bch2_fs_btree_cache_init(struct bch_fs *c) bch2_recalc_btree_reserve(c); - for (i = 0; i < bc->nr_reserve; i++) - if (!__bch2_btree_node_mem_alloc(c)) + for (i = 0; i < bc->nr_reserve; i++) { + struct btree *b = __bch2_btree_node_mem_alloc(c); + if (!b) goto err; + __bch2_btree_node_to_freelist(bc, b); + } list_splice_init(&bc->live[0].list, &bc->freeable); @@ -682,7 +682,7 @@ int bch2_fs_btree_cache_init(struct bch_fs *c) return 0; err: - return -BCH_ERR_ENOMEM_fs_btree_cache_init; + return bch_err_throw(c, ENOMEM_fs_btree_cache_init); } void bch2_fs_btree_cache_init_early(struct btree_cache *bc) @@ -727,7 +727,7 @@ int bch2_btree_cache_cannibalize_lock(struct btree_trans *trans, struct closure if (!cl) { trace_and_count(c, btree_cache_cannibalize_lock_fail, trans); - return -BCH_ERR_ENOMEM_btree_cache_cannibalize_lock; + return bch_err_throw(c, ENOMEM_btree_cache_cannibalize_lock); } closure_wait(&bc->alloc_wait, cl); @@ -741,7 +741,7 @@ int bch2_btree_cache_cannibalize_lock(struct btree_trans *trans, struct closure } trace_and_count(c, btree_cache_cannibalize_lock_fail, trans); - return -BCH_ERR_btree_cache_cannibalize_lock_blocked; + return bch_err_throw(c, btree_cache_cannibalize_lock_blocked); success: trace_and_count(c, btree_cache_cannibalize_lock, trans); diff --git a/fs/bcachefs/btree_cache.h b/fs/bcachefs/btree_cache.h index ca3c1b145330..be275f87a60e 100644 --- a/fs/bcachefs/btree_cache.h +++ b/fs/bcachefs/btree_cache.h @@ -30,6 +30,7 @@ void bch2_btree_node_update_key_early(struct btree_trans *, enum btree_id, unsig void bch2_btree_cache_cannibalize_unlock(struct btree_trans *); int bch2_btree_cache_cannibalize_lock(struct btree_trans *, struct closure *); +void __btree_node_data_free(struct btree *); struct btree *__bch2_btree_node_mem_alloc(struct bch_fs *); struct btree *bch2_btree_node_mem_alloc(struct btree_trans *, bool); diff --git a/fs/bcachefs/btree_gc.c b/fs/bcachefs/btree_gc.c index 91b6395421df..bac108e93823 100644 --- a/fs/bcachefs/btree_gc.c +++ b/fs/bcachefs/btree_gc.c @@ -150,7 +150,7 @@ static int set_node_min(struct bch_fs *c, struct btree *b, struct bpos new_min) new = kmalloc_array(BKEY_BTREE_PTR_U64s_MAX, sizeof(u64), GFP_KERNEL); if (!new) - return -BCH_ERR_ENOMEM_gc_repair_key; + return bch_err_throw(c, ENOMEM_gc_repair_key); btree_ptr_to_v2(b, new); b->data->min_key = new_min; @@ -190,7 +190,7 @@ static int set_node_max(struct bch_fs *c, struct btree *b, struct bpos new_max) new = kmalloc_array(BKEY_BTREE_PTR_U64s_MAX, sizeof(u64), GFP_KERNEL); if (!new) - return -BCH_ERR_ENOMEM_gc_repair_key; + return bch_err_throw(c, ENOMEM_gc_repair_key); btree_ptr_to_v2(b, new); b->data->max_key = new_max; @@ -397,7 +397,11 @@ again: continue; } - ret = btree_check_node_boundaries(trans, b, prev, cur, pulled_from_scan); + ret = lockrestart_do(trans, + btree_check_node_boundaries(trans, b, prev, cur, pulled_from_scan)); + if (ret < 0) + goto err; + if (ret == DID_FILL_FROM_SCAN) { new_pass = true; ret = 0; @@ -438,7 +442,8 @@ again: if (!ret && !IS_ERR_OR_NULL(prev)) { BUG_ON(cur); - ret = btree_repair_node_end(trans, b, prev, pulled_from_scan); + ret = lockrestart_do(trans, + btree_repair_node_end(trans, b, prev, pulled_from_scan)); if (ret == DID_FILL_FROM_SCAN) { new_pass = true; ret = 0; @@ -498,8 +503,14 @@ again: prt_newline(&buf); bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key)); + /* + * XXX: we're not passing the trans object here because we're not set up + * to handle a transaction restart - this code needs to be rewritten + * when we start doing online topology repair + */ + bch2_trans_unlock_long(trans); if (mustfix_fsck_err_on(!have_child, - trans, btree_node_topology_interior_node_empty, + c, btree_node_topology_interior_node_empty, "empty interior btree node at %s", buf.buf)) ret = DROP_THIS_NODE; err: @@ -519,49 +530,72 @@ fsck_err: bch2_bkey_buf_exit(&prev_k, c); bch2_bkey_buf_exit(&cur_k, c); printbuf_exit(&buf); + bch_err_fn(c, ret); return ret; } -int bch2_check_topology(struct bch_fs *c) +static int bch2_check_root(struct btree_trans *trans, enum btree_id btree, + bool *reconstructed_root) { - struct btree_trans *trans = bch2_trans_get(c); - struct bpos pulled_from_scan = POS_MIN; + struct bch_fs *c = trans->c; + struct btree_root *r = bch2_btree_id_root(c, btree); struct printbuf buf = PRINTBUF; int ret = 0; - bch2_trans_srcu_unlock(trans); + bch2_btree_id_to_text(&buf, btree); - for (unsigned i = 0; i < btree_id_nr_alive(c) && !ret; i++) { - struct btree_root *r = bch2_btree_id_root(c, i); - bool reconstructed_root = false; + if (r->error) { + bch_info(c, "btree root %s unreadable, must recover from scan", buf.buf); - printbuf_reset(&buf); - bch2_btree_id_to_text(&buf, i); + ret = bch2_btree_has_scanned_nodes(c, btree); + if (ret < 0) + goto err; - if (r->error) { -reconstruct_root: - bch_info(c, "btree root %s unreadable, must recover from scan", buf.buf); + if (!ret) { + __fsck_err(trans, + FSCK_CAN_FIX|(!btree_id_important(btree) ? FSCK_AUTOFIX : 0), + btree_root_unreadable_and_scan_found_nothing, + "no nodes found for btree %s, continue?", buf.buf); r->alive = false; r->error = 0; + bch2_btree_root_alloc_fake_trans(trans, btree, 0); + } else { + r->alive = false; + r->error = 0; + bch2_btree_root_alloc_fake_trans(trans, btree, 1); - if (!bch2_btree_has_scanned_nodes(c, i)) { - __fsck_err(trans, - FSCK_CAN_FIX|(!btree_id_important(i) ? FSCK_AUTOFIX : 0), - btree_root_unreadable_and_scan_found_nothing, - "no nodes found for btree %s, continue?", buf.buf); - bch2_btree_root_alloc_fake_trans(trans, i, 0); - } else { - bch2_btree_root_alloc_fake_trans(trans, i, 1); - bch2_shoot_down_journal_keys(c, i, 1, BTREE_MAX_DEPTH, POS_MIN, SPOS_MAX); - ret = bch2_get_scanned_nodes(c, i, 0, POS_MIN, SPOS_MAX); - if (ret) - break; - } - - reconstructed_root = true; + bch2_shoot_down_journal_keys(c, btree, 1, BTREE_MAX_DEPTH, POS_MIN, SPOS_MAX); + ret = bch2_get_scanned_nodes(c, btree, 0, POS_MIN, SPOS_MAX); + if (ret) + goto err; } + *reconstructed_root = true; + } +err: +fsck_err: + printbuf_exit(&buf); + bch_err_fn(c, ret); + return ret; +} + +int bch2_check_topology(struct bch_fs *c) +{ + struct btree_trans *trans = bch2_trans_get(c); + struct bpos pulled_from_scan = POS_MIN; + int ret = 0; + + bch2_trans_srcu_unlock(trans); + + for (unsigned i = 0; i < btree_id_nr_alive(c) && !ret; i++) { + bool reconstructed_root = false; +recover: + ret = lockrestart_do(trans, bch2_check_root(trans, i, &reconstructed_root)); + if (ret) + break; + + struct btree_root *r = bch2_btree_id_root(c, i); struct btree *b = r->b; btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_read); @@ -575,17 +609,21 @@ reconstruct_root: r->b = NULL; - if (!reconstructed_root) - goto reconstruct_root; + if (!reconstructed_root) { + r->error = -EIO; + goto recover; + } + struct printbuf buf = PRINTBUF; + bch2_btree_id_to_text(&buf, i); bch_err(c, "empty btree root %s", buf.buf); + printbuf_exit(&buf); bch2_btree_root_alloc_fake_trans(trans, i, 0); r->alive = false; ret = 0; } } -fsck_err: - printbuf_exit(&buf); + bch2_trans_put(trans); return ret; } @@ -935,7 +973,7 @@ static int bch2_gc_alloc_start(struct bch_fs *c) ret = genradix_prealloc(&ca->buckets_gc, ca->mi.nbuckets, GFP_KERNEL); if (ret) { bch2_dev_put(ca); - ret = -BCH_ERR_ENOMEM_gc_alloc_start; + ret = bch_err_throw(c, ENOMEM_gc_alloc_start); break; } } @@ -1093,42 +1131,41 @@ static int gc_btree_gens_key(struct btree_trans *trans, { struct bch_fs *c = trans->c; struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); - struct bkey_i *u; - int ret; if (unlikely(test_bit(BCH_FS_going_ro, &c->flags))) return -EROFS; - rcu_read_lock(); - bkey_for_each_ptr(ptrs, ptr) { - struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); - if (!ca) - continue; + bool too_stale = false; + scoped_guard(rcu) { + bkey_for_each_ptr(ptrs, ptr) { + struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); + if (!ca) + continue; - if (dev_ptr_stale(ca, ptr) > 16) { - rcu_read_unlock(); - goto update; + too_stale |= dev_ptr_stale(ca, ptr) > 16; } + + if (!too_stale) + bkey_for_each_ptr(ptrs, ptr) { + struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); + if (!ca) + continue; + + u8 *gen = &ca->oldest_gen[PTR_BUCKET_NR(ca, ptr)]; + if (gen_after(*gen, ptr->gen)) + *gen = ptr->gen; + } } - bkey_for_each_ptr(ptrs, ptr) { - struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); - if (!ca) - continue; + if (too_stale) { + struct bkey_i *u = bch2_bkey_make_mut(trans, iter, &k, 0); + int ret = PTR_ERR_OR_ZERO(u); + if (ret) + return ret; - u8 *gen = &ca->oldest_gen[PTR_BUCKET_NR(ca, ptr)]; - if (gen_after(*gen, ptr->gen)) - *gen = ptr->gen; + bch2_extent_normalize(c, bkey_i_to_s(u)); } - rcu_read_unlock(); - return 0; -update: - u = bch2_bkey_make_mut(trans, iter, &k, 0); - ret = PTR_ERR_OR_ZERO(u); - if (ret) - return ret; - bch2_extent_normalize(c, bkey_i_to_s(u)); return 0; } @@ -1181,7 +1218,7 @@ int bch2_gc_gens(struct bch_fs *c) ca->oldest_gen = kvmalloc(gens->nbuckets, GFP_KERNEL); if (!ca->oldest_gen) { bch2_dev_put(ca); - ret = -BCH_ERR_ENOMEM_gc_gens; + ret = bch_err_throw(c, ENOMEM_gc_gens); goto err; } diff --git a/fs/bcachefs/btree_io.c b/fs/bcachefs/btree_io.c index 34018296053a..590cd29f3e86 100644 --- a/fs/bcachefs/btree_io.c +++ b/fs/bcachefs/btree_io.c @@ -557,7 +557,9 @@ static int __btree_err(int ret, const char *fmt, ...) { if (c->recovery.curr_pass == BCH_RECOVERY_PASS_scan_for_btree_nodes) - return -BCH_ERR_fsck_fix; + return ret == -BCH_ERR_btree_node_read_err_fixable + ? bch_err_throw(c, fsck_fix) + : ret; bool have_retry = false; int ret2; @@ -566,15 +568,15 @@ static int __btree_err(int ret, bch2_mark_btree_validate_failure(failed, ca->dev_idx); struct extent_ptr_decoded pick; - have_retry = !bch2_bkey_pick_read_device(c, + have_retry = bch2_bkey_pick_read_device(c, bkey_i_to_s_c(&b->key), - failed, &pick, -1); + failed, &pick, -1) == 1; } if (!have_retry && ret == -BCH_ERR_btree_node_read_err_want_retry) - ret = -BCH_ERR_btree_node_read_err_fixable; + ret = bch_err_throw(c, btree_node_read_err_fixable); if (!have_retry && ret == -BCH_ERR_btree_node_read_err_must_retry) - ret = -BCH_ERR_btree_node_read_err_bad_node; + ret = bch_err_throw(c, btree_node_read_err_bad_node); bch2_sb_error_count(c, err_type); @@ -602,18 +604,17 @@ static int __btree_err(int ret, switch (ret) { case -BCH_ERR_btree_node_read_err_fixable: ret2 = bch2_fsck_err_opt(c, FSCK_CAN_FIX, err_type); - if (ret2 != -BCH_ERR_fsck_fix && - ret2 != -BCH_ERR_fsck_ignore) { + if (!bch2_err_matches(ret2, BCH_ERR_fsck_fix) && + !bch2_err_matches(ret2, BCH_ERR_fsck_ignore)) { ret = ret2; goto fsck_err; } if (!have_retry) - ret = -BCH_ERR_fsck_fix; + ret = bch_err_throw(c, fsck_fix); goto out; case -BCH_ERR_btree_node_read_err_bad_node: prt_str(&out, ", "); - ret = __bch2_topology_error(c, &out); break; } @@ -631,18 +632,17 @@ static int __btree_err(int ret, switch (ret) { case -BCH_ERR_btree_node_read_err_fixable: ret2 = __bch2_fsck_err(c, NULL, FSCK_CAN_FIX, err_type, "%s", out.buf); - if (ret2 != -BCH_ERR_fsck_fix && - ret2 != -BCH_ERR_fsck_ignore) { + if (!bch2_err_matches(ret2, BCH_ERR_fsck_fix) && + !bch2_err_matches(ret2, BCH_ERR_fsck_ignore)) { ret = ret2; goto fsck_err; } if (!have_retry) - ret = -BCH_ERR_fsck_fix; + ret = bch_err_throw(c, fsck_fix); goto out; case -BCH_ERR_btree_node_read_err_bad_node: prt_str(&out, ", "); - ret = __bch2_topology_error(c, &out); break; } print: @@ -660,7 +660,7 @@ fsck_err: failed, err_msg, \ msg, ##__VA_ARGS__); \ \ - if (_ret != -BCH_ERR_fsck_fix) { \ + if (!bch2_err_matches(_ret, BCH_ERR_fsck_fix)) { \ ret = _ret; \ goto fsck_err; \ } \ @@ -723,12 +723,11 @@ void bch2_btree_node_drop_keys_outside_node(struct btree *b) static int validate_bset(struct bch_fs *c, struct bch_dev *ca, struct btree *b, struct bset *i, - unsigned offset, unsigned sectors, int write, + unsigned offset, int write, struct bch_io_failures *failed, struct printbuf *err_msg) { unsigned version = le16_to_cpu(i->version); - unsigned ptr_written = btree_ptr_sectors_written(bkey_i_to_s_c(&b->key)); struct printbuf buf1 = PRINTBUF; struct printbuf buf2 = PRINTBUF; int ret = 0; @@ -741,16 +740,22 @@ static int validate_bset(struct bch_fs *c, struct bch_dev *ca, BCH_VERSION_MAJOR(version), BCH_VERSION_MINOR(version)); - if (btree_err_on(version < c->sb.version_min, + if (c->recovery.curr_pass != BCH_RECOVERY_PASS_scan_for_btree_nodes && + btree_err_on(version < c->sb.version_min, -BCH_ERR_btree_node_read_err_fixable, c, NULL, b, i, NULL, btree_node_bset_older_than_sb_min, "bset version %u older than superblock version_min %u", version, c->sb.version_min)) { - mutex_lock(&c->sb_lock); - c->disk_sb.sb->version_min = cpu_to_le16(version); - bch2_write_super(c); - mutex_unlock(&c->sb_lock); + if (bch2_version_compatible(version)) { + mutex_lock(&c->sb_lock); + c->disk_sb.sb->version_min = cpu_to_le16(version); + bch2_write_super(c); + mutex_unlock(&c->sb_lock); + } else { + /* We have no idea what's going on: */ + i->version = cpu_to_le16(c->sb.version); + } } if (btree_err_on(BCH_VERSION_MAJOR(version) > @@ -772,15 +777,6 @@ static int validate_bset(struct bch_fs *c, struct bch_dev *ca, btree_node_unsupported_version, "BSET_SEPARATE_WHITEOUTS no longer supported"); - if (!write && - btree_err_on(offset + sectors > (ptr_written ?: btree_sectors(c)), - -BCH_ERR_btree_node_read_err_fixable, - c, ca, b, i, NULL, - bset_past_end_of_btree_node, - "bset past end of btree node (offset %u len %u but written %zu)", - offset, sectors, ptr_written ?: btree_sectors(c))) - i->u64s = 0; - btree_err_on(offset && !i->u64s, -BCH_ERR_btree_node_read_err_fixable, c, ca, b, i, NULL, @@ -1045,6 +1041,7 @@ got_good_key: le16_add_cpu(&i->u64s, -next_good_key); memmove_u64s_down(k, (u64 *) k + next_good_key, (u64 *) vstruct_end(i) - (u64 *) k); set_btree_node_need_rewrite(b); + set_btree_node_need_rewrite_error(b); } fsck_err: printbuf_exit(&buf); @@ -1144,6 +1141,14 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca, "unknown checksum type %llu", BSET_CSUM_TYPE(i)); if (first) { + sectors = vstruct_sectors(b->data, c->block_bits); + if (btree_err_on(b->written + sectors > (ptr_written ?: btree_sectors(c)), + -BCH_ERR_btree_node_read_err_fixable, + c, ca, b, i, NULL, + bset_past_end_of_btree_node, + "bset past end of btree node (offset %u len %u but written %zu)", + b->written, sectors, ptr_written ?: btree_sectors(c))) + i->u64s = 0; if (good_csum_type) { struct bch_csum csum = csum_vstruct(c, BSET_CSUM_TYPE(i), nonce, b->data); bool csum_bad = bch2_crc_cmp(b->data->csum, csum); @@ -1171,9 +1176,15 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca, c, NULL, b, NULL, NULL, btree_node_unsupported_version, "btree node does not have NEW_EXTENT_OVERWRITE set"); - - sectors = vstruct_sectors(b->data, c->block_bits); } else { + sectors = vstruct_sectors(bne, c->block_bits); + if (btree_err_on(b->written + sectors > (ptr_written ?: btree_sectors(c)), + -BCH_ERR_btree_node_read_err_fixable, + c, ca, b, i, NULL, + bset_past_end_of_btree_node, + "bset past end of btree node (offset %u len %u but written %zu)", + b->written, sectors, ptr_written ?: btree_sectors(c))) + i->u64s = 0; if (good_csum_type) { struct bch_csum csum = csum_vstruct(c, BSET_CSUM_TYPE(i), nonce, bne); bool csum_bad = bch2_crc_cmp(bne->csum, csum); @@ -1194,14 +1205,12 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca, "decrypting btree node: %s", bch2_err_str(ret))) goto fsck_err; } - - sectors = vstruct_sectors(bne, c->block_bits); } b->version_ondisk = min(b->version_ondisk, le16_to_cpu(i->version)); - ret = validate_bset(c, ca, b, i, b->written, sectors, READ, failed, err_msg); + ret = validate_bset(c, ca, b, i, b->written, READ, failed, err_msg); if (ret) goto fsck_err; @@ -1286,9 +1295,6 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca, btree_bounce_free(c, btree_buf_bytes(b), used_mempool, sorted); - if (updated_range) - bch2_btree_node_drop_keys_outside_node(b); - i = &b->data->keys; for (k = i->start; k != vstruct_last(i);) { struct bkey tmp; @@ -1305,6 +1311,7 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca, (u64 *) vstruct_end(i) - (u64 *) k); set_btree_bset_end(b, b->set); set_btree_node_need_rewrite(b); + set_btree_node_need_rewrite_error(b); continue; } if (ret) @@ -1325,17 +1332,50 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca, btree_node_reset_sib_u64s(b); - rcu_read_lock(); - bkey_for_each_ptr(bch2_bkey_ptrs(bkey_i_to_s(&b->key)), ptr) { - struct bch_dev *ca2 = bch2_dev_rcu(c, ptr->dev); + if (updated_range) + bch2_btree_node_drop_keys_outside_node(b); - if (!ca2 || ca2->mi.state != BCH_MEMBER_STATE_rw) - set_btree_node_need_rewrite(b); + /* + * XXX: + * + * We deadlock if too many btree updates require node rewrites while + * we're still in journal replay. + * + * This is because btree node rewrites generate more updates for the + * interior updates (alloc, backpointers), and if those updates touch + * new nodes and generate more rewrites - well, you see the problem. + * + * The biggest cause is that we don't use the btree write buffer (for + * the backpointer updates - this needs some real thought on locking in + * order to fix. + * + * The problem with this workaround (not doing the rewrite for degraded + * nodes in journal replay) is that those degraded nodes persist, and we + * don't want that (this is a real bug when a btree node write completes + * with fewer replicas than we wanted and leaves a degraded node due to + * device _removal_, i.e. the device went away mid write). + * + * It's less of a bug here, but still a problem because we don't yet + * have a way of tracking degraded data - we another index (all + * extents/btree nodes, by replicas entry) in order to fix properly + * (re-replicate degraded data at the earliest possible time). + */ + if (c->recovery.passes_complete & BIT_ULL(BCH_RECOVERY_PASS_journal_replay)) { + scoped_guard(rcu) + bkey_for_each_ptr(bch2_bkey_ptrs(bkey_i_to_s(&b->key)), ptr) { + struct bch_dev *ca2 = bch2_dev_rcu(c, ptr->dev); + + if (!ca2 || ca2->mi.state != BCH_MEMBER_STATE_rw) { + set_btree_node_need_rewrite(b); + set_btree_node_need_rewrite_degraded(b); + } + } } - rcu_read_unlock(); - if (!ptr_written) + if (!ptr_written) { set_btree_node_need_rewrite(b); + set_btree_node_need_rewrite_ptr_written_zero(b); + } fsck_err: mempool_free(iter, &c->fill_iter); printbuf_exit(&buf); @@ -1366,7 +1406,7 @@ static void btree_node_read_work(struct work_struct *work) ret = bch2_bkey_pick_read_device(c, bkey_i_to_s_c(&b->key), &failed, &rb->pick, -1); - if (ret) { + if (ret <= 0) { set_btree_node_read_error(b); break; } @@ -1688,7 +1728,7 @@ static int btree_node_read_all_replicas(struct bch_fs *c, struct btree *b, bool ra = kzalloc(sizeof(*ra), GFP_NOFS); if (!ra) - return -BCH_ERR_ENOMEM_btree_node_read_all_replicas; + return bch_err_throw(c, ENOMEM_btree_node_read_all_replicas); closure_init(&ra->cl, NULL); ra->c = c; @@ -1870,7 +1910,7 @@ static int __bch2_btree_root_read(struct btree_trans *trans, enum btree_id id, bch2_btree_node_hash_remove(&c->btree_cache, b); mutex_unlock(&c->btree_cache.lock); - ret = -BCH_ERR_btree_node_read_error; + ret = bch_err_throw(c, btree_node_read_error); goto err; } @@ -1971,28 +2011,12 @@ static void btree_node_scrub_work(struct work_struct *work) prt_newline(&err); if (!btree_node_scrub_check(c, scrub->buf, scrub->written, &err)) { - struct btree_trans *trans = bch2_trans_get(c); - - struct btree_iter iter; - bch2_trans_node_iter_init(trans, &iter, scrub->btree, - scrub->key.k->k.p, 0, scrub->level - 1, 0); - - struct btree *b; - int ret = lockrestart_do(trans, - PTR_ERR_OR_ZERO(b = bch2_btree_iter_peek_node(trans, &iter))); - if (ret) - goto err; - - if (bkey_i_to_btree_ptr_v2(&b->key)->v.seq == scrub->seq) { - bch_err(c, "error validating btree node during scrub on %s at btree %s", - scrub->ca->name, err.buf); - - ret = bch2_btree_node_rewrite(trans, &iter, b, 0, 0); - } -err: - bch2_trans_iter_exit(trans, &iter); - bch2_trans_begin(trans); - bch2_trans_put(trans); + int ret = bch2_trans_do(c, + bch2_btree_node_rewrite_key(trans, scrub->btree, scrub->level - 1, + scrub->key.k, 0)); + if (!bch2_err_matches(ret, ENOENT) && + !bch2_err_matches(ret, EROFS)) + bch_err_fn_ratelimited(c, ret); } printbuf_exit(&err); @@ -2020,7 +2044,7 @@ int bch2_btree_node_scrub(struct btree_trans *trans, struct bch_fs *c = trans->c; if (!enumerated_ref_tryget(&c->writes, BCH_WRITE_REF_btree_node_scrub)) - return -BCH_ERR_erofs_no_writes; + return bch_err_throw(c, erofs_no_writes); struct extent_ptr_decoded pick; int ret = bch2_bkey_pick_read_device(c, k, NULL, &pick, dev); @@ -2030,7 +2054,7 @@ int bch2_btree_node_scrub(struct btree_trans *trans, struct bch_dev *ca = bch2_dev_get_ioref(c, pick.ptr.dev, READ, BCH_DEV_READ_REF_btree_node_scrub); if (!ca) { - ret = -BCH_ERR_device_offline; + ret = bch_err_throw(c, device_offline); goto err; } @@ -2167,7 +2191,7 @@ static void btree_node_write_work(struct work_struct *work) bch2_dev_list_has_dev(wbio->wbio.failed, ptr->dev)); if (!bch2_bkey_nr_ptrs(bkey_i_to_s_c(&wbio->key))) { - ret = -BCH_ERR_btree_node_write_all_failed; + ret = bch_err_throw(c, btree_node_write_all_failed); goto err; } @@ -2256,7 +2280,7 @@ static void btree_node_write_endio(struct bio *bio) } static int validate_bset_for_write(struct bch_fs *c, struct btree *b, - struct bset *i, unsigned sectors) + struct bset *i) { int ret = bch2_bkey_validate(c, bkey_i_to_s_c(&b->key), (struct bkey_validate_context) { @@ -2271,7 +2295,7 @@ static int validate_bset_for_write(struct bch_fs *c, struct btree *b, } ret = validate_bset_keys(c, b, i, WRITE, NULL, NULL) ?: - validate_bset(c, NULL, b, i, b->written, sectors, WRITE, NULL, NULL); + validate_bset(c, NULL, b, i, b->written, WRITE, NULL, NULL); if (ret) { bch2_inconsistent_error(c); dump_stack(); @@ -2464,7 +2488,7 @@ do_write: /* if we're going to be encrypting, check metadata validity first: */ if (validate_before_checksum && - validate_bset_for_write(c, b, i, sectors_to_write)) + validate_bset_for_write(c, b, i)) goto err; ret = bset_encrypt(c, i, b->written << 9); @@ -2481,7 +2505,7 @@ do_write: /* if we're not encrypting, check metadata after checksumming: */ if (!validate_before_checksum && - validate_bset_for_write(c, b, i, sectors_to_write)) + validate_bset_for_write(c, b, i)) goto err; /* diff --git a/fs/bcachefs/btree_iter.c b/fs/bcachefs/btree_iter.c index b4bf4217a3fa..f8829b667ad3 100644 --- a/fs/bcachefs/btree_iter.c +++ b/fs/bcachefs/btree_iter.c @@ -890,8 +890,7 @@ static noinline void btree_node_mem_ptr_set(struct btree_trans *trans, static noinline int btree_node_iter_and_journal_peek(struct btree_trans *trans, struct btree_path *path, - unsigned flags, - struct bkey_buf *out) + unsigned flags) { struct bch_fs *c = trans->c; struct btree_path_level *l = path_l(path); @@ -915,7 +914,7 @@ static noinline int btree_node_iter_and_journal_peek(struct btree_trans *trans, goto err; } - bch2_bkey_buf_reassemble(out, c, k); + bkey_reassemble(&trans->btree_path_down, k); if ((flags & BTREE_ITER_prefetch) && c->opts.btree_node_prefetch) @@ -926,6 +925,22 @@ err: return ret; } +static noinline_for_stack int btree_node_missing_err(struct btree_trans *trans, + struct btree_path *path) +{ + struct bch_fs *c = trans->c; + struct printbuf buf = PRINTBUF; + + prt_str(&buf, "node not found at pos "); + bch2_bpos_to_text(&buf, path->pos); + prt_str(&buf, " within parent node "); + bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&path_l(path)->b->key)); + + bch2_fs_fatal_error(c, "%s", buf.buf); + printbuf_exit(&buf); + return bch_err_throw(c, btree_need_topology_repair); +} + static __always_inline int btree_path_down(struct btree_trans *trans, struct btree_path *path, unsigned flags, @@ -936,51 +951,38 @@ static __always_inline int btree_path_down(struct btree_trans *trans, struct btree *b; unsigned level = path->level - 1; enum six_lock_type lock_type = __btree_lock_want(path, level); - struct bkey_buf tmp; int ret; EBUG_ON(!btree_node_locked(path, path->level)); - bch2_bkey_buf_init(&tmp); - if (unlikely(trans->journal_replay_not_finished)) { - ret = btree_node_iter_and_journal_peek(trans, path, flags, &tmp); + ret = btree_node_iter_and_journal_peek(trans, path, flags); if (ret) - goto err; + return ret; } else { struct bkey_packed *k = bch2_btree_node_iter_peek(&l->iter, l->b); - if (!k) { - struct printbuf buf = PRINTBUF; - - prt_str(&buf, "node not found at pos "); - bch2_bpos_to_text(&buf, path->pos); - prt_str(&buf, " within parent node "); - bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&l->b->key)); - - bch2_fs_fatal_error(c, "%s", buf.buf); - printbuf_exit(&buf); - ret = -BCH_ERR_btree_need_topology_repair; - goto err; - } + if (unlikely(!k)) + return btree_node_missing_err(trans, path); - bch2_bkey_buf_unpack(&tmp, c, l->b, k); + bch2_bkey_unpack(l->b, &trans->btree_path_down, k); - if ((flags & BTREE_ITER_prefetch) && + if (unlikely((flags & BTREE_ITER_prefetch)) && c->opts.btree_node_prefetch) { ret = btree_path_prefetch(trans, path); if (ret) - goto err; + return ret; } } - b = bch2_btree_node_get(trans, path, tmp.k, level, lock_type, trace_ip); + b = bch2_btree_node_get(trans, path, &trans->btree_path_down, + level, lock_type, trace_ip); ret = PTR_ERR_OR_ZERO(b); if (unlikely(ret)) - goto err; + return ret; - if (likely(!trans->journal_replay_not_finished && - tmp.k->k.type == KEY_TYPE_btree_ptr_v2) && - unlikely(b != btree_node_mem_ptr(tmp.k))) + if (unlikely(b != btree_node_mem_ptr(&trans->btree_path_down)) && + likely(!trans->journal_replay_not_finished && + trans->btree_path_down.k.type == KEY_TYPE_btree_ptr_v2)) btree_node_mem_ptr_set(trans, path, level + 1, b); if (btree_node_read_locked(path, level + 1)) @@ -992,9 +994,7 @@ static __always_inline int btree_path_down(struct btree_trans *trans, bch2_btree_path_level_init(trans, path, b); bch2_btree_path_verify_locks(trans, path); -err: - bch2_bkey_buf_exit(&tmp, c); - return ret; + return 0; } static int bch2_btree_path_traverse_all(struct btree_trans *trans) @@ -1006,7 +1006,7 @@ static int bch2_btree_path_traverse_all(struct btree_trans *trans) int ret = 0; if (trans->in_traverse_all) - return -BCH_ERR_transaction_restart_in_traverse_all; + return bch_err_throw(c, transaction_restart_in_traverse_all); trans->in_traverse_all = true; retry_all: @@ -2076,14 +2076,14 @@ inline bool bch2_btree_iter_rewind(struct btree_trans *trans, struct btree_iter static noinline void bch2_btree_trans_peek_prev_updates(struct btree_trans *trans, struct btree_iter *iter, - struct bkey_s_c *k) + struct bpos search_key, struct bkey_s_c *k) { struct bpos end = path_l(btree_iter_path(trans, iter))->b->data->min_key; trans_for_each_update(trans, i) if (!i->key_cache_already_flushed && i->btree_id == iter->btree_id && - bpos_le(i->k->k.p, iter->pos) && + bpos_le(i->k->k.p, search_key) && bpos_ge(i->k->k.p, k->k ? k->k->p : end)) { iter->k = i->k->k; *k = bkey_i_to_s_c(i->k); @@ -2092,6 +2092,7 @@ void bch2_btree_trans_peek_prev_updates(struct btree_trans *trans, struct btree_ static noinline void bch2_btree_trans_peek_updates(struct btree_trans *trans, struct btree_iter *iter, + struct bpos search_key, struct bkey_s_c *k) { struct btree_path *path = btree_iter_path(trans, iter); @@ -2100,7 +2101,7 @@ void bch2_btree_trans_peek_updates(struct btree_trans *trans, struct btree_iter trans_for_each_update(trans, i) if (!i->key_cache_already_flushed && i->btree_id == iter->btree_id && - bpos_ge(i->k->k.p, path->pos) && + bpos_ge(i->k->k.p, search_key) && bpos_le(i->k->k.p, k->k ? k->k->p : end)) { iter->k = i->k->k; *k = bkey_i_to_s_c(i->k); @@ -2122,13 +2123,14 @@ void bch2_btree_trans_peek_slot_updates(struct btree_trans *trans, struct btree_ static struct bkey_i *bch2_btree_journal_peek(struct btree_trans *trans, struct btree_iter *iter, + struct bpos search_pos, struct bpos end_pos) { struct btree_path *path = btree_iter_path(trans, iter); return bch2_journal_keys_peek_max(trans->c, iter->btree_id, path->level, - path->pos, + search_pos, end_pos, &iter->journal_idx); } @@ -2138,7 +2140,7 @@ struct bkey_s_c btree_trans_peek_slot_journal(struct btree_trans *trans, struct btree_iter *iter) { struct btree_path *path = btree_iter_path(trans, iter); - struct bkey_i *k = bch2_btree_journal_peek(trans, iter, path->pos); + struct bkey_i *k = bch2_btree_journal_peek(trans, iter, path->pos, path->pos); if (k) { iter->k = k->k; @@ -2151,11 +2153,12 @@ struct bkey_s_c btree_trans_peek_slot_journal(struct btree_trans *trans, static noinline void btree_trans_peek_journal(struct btree_trans *trans, struct btree_iter *iter, + struct bpos search_key, struct bkey_s_c *k) { struct btree_path *path = btree_iter_path(trans, iter); struct bkey_i *next_journal = - bch2_btree_journal_peek(trans, iter, + bch2_btree_journal_peek(trans, iter, search_key, k->k ? k->k->p : path_l(path)->b->key.k.p); if (next_journal) { iter->k = next_journal->k; @@ -2165,13 +2168,14 @@ void btree_trans_peek_journal(struct btree_trans *trans, static struct bkey_i *bch2_btree_journal_peek_prev(struct btree_trans *trans, struct btree_iter *iter, + struct bpos search_key, struct bpos end_pos) { struct btree_path *path = btree_iter_path(trans, iter); return bch2_journal_keys_peek_prev_min(trans->c, iter->btree_id, path->level, - path->pos, + search_key, end_pos, &iter->journal_idx); } @@ -2179,12 +2183,13 @@ static struct bkey_i *bch2_btree_journal_peek_prev(struct btree_trans *trans, static noinline void btree_trans_peek_prev_journal(struct btree_trans *trans, struct btree_iter *iter, + struct bpos search_key, struct bkey_s_c *k) { struct btree_path *path = btree_iter_path(trans, iter); struct bkey_i *next_journal = - bch2_btree_journal_peek_prev(trans, iter, - k->k ? k->k->p : path_l(path)->b->key.k.p); + bch2_btree_journal_peek_prev(trans, iter, search_key, + k->k ? k->k->p : path_l(path)->b->data->min_key); if (next_journal) { iter->k = next_journal->k; @@ -2292,11 +2297,11 @@ static struct bkey_s_c __bch2_btree_iter_peek(struct btree_trans *trans, struct } if (unlikely(iter->flags & BTREE_ITER_with_journal)) - btree_trans_peek_journal(trans, iter, &k); + btree_trans_peek_journal(trans, iter, search_key, &k); if (unlikely((iter->flags & BTREE_ITER_with_updates) && trans->nr_updates)) - bch2_btree_trans_peek_updates(trans, iter, &k); + bch2_btree_trans_peek_updates(trans, iter, search_key, &k); if (k.k && bkey_deleted(k.k)) { /* @@ -2326,6 +2331,20 @@ static struct bkey_s_c __bch2_btree_iter_peek(struct btree_trans *trans, struct } bch2_btree_iter_verify(trans, iter); + + if (trace___btree_iter_peek_enabled()) { + CLASS(printbuf, buf)(); + + int ret = bkey_err(k); + if (ret) + prt_str(&buf, bch2_err_str(ret)); + else if (k.k) + bch2_bkey_val_to_text(&buf, trans->c, k); + else + prt_str(&buf, "(null)"); + trace___btree_iter_peek(trans->c, buf.buf); + } + return k; } @@ -2484,6 +2503,19 @@ out_no_locked: bch2_btree_iter_verify_entry_exit(iter); + if (trace_btree_iter_peek_max_enabled()) { + CLASS(printbuf, buf)(); + + int ret = bkey_err(k); + if (ret) + prt_str(&buf, bch2_err_str(ret)); + else if (k.k) + bch2_bkey_val_to_text(&buf, trans->c, k); + else + prt_str(&buf, "(null)"); + trace_btree_iter_peek_max(trans->c, buf.buf); + } + return k; end: bch2_btree_iter_set_pos(trans, iter, end); @@ -2557,11 +2589,11 @@ static struct bkey_s_c __bch2_btree_iter_peek_prev(struct btree_trans *trans, st } if (unlikely(iter->flags & BTREE_ITER_with_journal)) - btree_trans_peek_prev_journal(trans, iter, &k); + btree_trans_peek_prev_journal(trans, iter, search_key, &k); if (unlikely((iter->flags & BTREE_ITER_with_updates) && trans->nr_updates)) - bch2_btree_trans_peek_prev_updates(trans, iter, &k); + bch2_btree_trans_peek_prev_updates(trans, iter, search_key, &k); if (likely(k.k && !bkey_deleted(k.k))) { break; @@ -2724,6 +2756,19 @@ out_no_locked: bch2_btree_iter_verify_entry_exit(iter); bch2_btree_iter_verify(trans, iter); + + if (trace_btree_iter_peek_prev_min_enabled()) { + CLASS(printbuf, buf)(); + + int ret = bkey_err(k); + if (ret) + prt_str(&buf, bch2_err_str(ret)); + else if (k.k) + bch2_bkey_val_to_text(&buf, trans->c, k); + else + prt_str(&buf, "(null)"); + trace_btree_iter_peek_prev_min(trans->c, buf.buf); + } return k; end: bch2_btree_iter_set_pos(trans, iter, end); @@ -2767,8 +2812,10 @@ struct bkey_s_c bch2_btree_iter_peek_slot(struct btree_trans *trans, struct btre /* extents can't span inode numbers: */ if ((iter->flags & BTREE_ITER_is_extents) && unlikely(iter->pos.offset == KEY_OFFSET_MAX)) { - if (iter->pos.inode == KEY_INODE_MAX) - return bkey_s_c_null; + if (iter->pos.inode == KEY_INODE_MAX) { + k = bkey_s_c_null; + goto out2; + } bch2_btree_iter_set_pos(trans, iter, bpos_nosnap_successor(iter->pos)); } @@ -2785,8 +2832,10 @@ struct bkey_s_c bch2_btree_iter_peek_slot(struct btree_trans *trans, struct btre } struct btree_path *path = btree_iter_path(trans, iter); - if (unlikely(!btree_path_node(path, path->level))) - return bkey_s_c_null; + if (unlikely(!btree_path_node(path, path->level))) { + k = bkey_s_c_null; + goto out2; + } btree_path_set_should_be_locked(trans, path); @@ -2879,7 +2928,20 @@ out: bch2_btree_iter_verify(trans, iter); ret = bch2_btree_iter_verify_ret(trans, iter, k); if (unlikely(ret)) - return bkey_s_c_err(ret); + k = bkey_s_c_err(ret); +out2: + if (trace_btree_iter_peek_slot_enabled()) { + CLASS(printbuf, buf)(); + + int ret = bkey_err(k); + if (ret) + prt_str(&buf, bch2_err_str(ret)); + else if (k.k) + bch2_bkey_val_to_text(&buf, trans->c, k); + else + prt_str(&buf, "(null)"); + trace_btree_iter_peek_slot(trans->c, buf.buf); + } return k; } @@ -3132,6 +3194,10 @@ void *__bch2_trans_kmalloc(struct btree_trans *trans, size_t size, unsigned long if (WARN_ON_ONCE(new_bytes > BTREE_TRANS_MEM_MAX)) { #ifdef CONFIG_BCACHEFS_TRANS_KMALLOC_TRACE struct printbuf buf = PRINTBUF; + bch2_log_msg_start(c, &buf); + prt_printf(&buf, "bump allocator exceeded BTREE_TRANS_MEM_MAX (%u)\n", + BTREE_TRANS_MEM_MAX); + bch2_trans_kmalloc_trace_to_text(&buf, &trans->trans_kmalloc_trace); bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); @@ -3159,46 +3225,32 @@ void *__bch2_trans_kmalloc(struct btree_trans *trans, size_t size, unsigned long mutex_unlock(&s->lock); } - if (trans->used_mempool) { - if (trans->mem_bytes >= new_bytes) - goto out_change_top; - - /* No more space from mempool item, need malloc new one */ - new_mem = kmalloc(new_bytes, GFP_NOWAIT|__GFP_NOWARN); - if (unlikely(!new_mem)) { - bch2_trans_unlock(trans); - - new_mem = kmalloc(new_bytes, GFP_KERNEL); - if (!new_mem) - return ERR_PTR(-BCH_ERR_ENOMEM_trans_kmalloc); + if (trans->used_mempool || new_bytes > BTREE_TRANS_MEM_MAX) { + EBUG_ON(trans->mem_bytes >= new_bytes); + return ERR_PTR(-BCH_ERR_ENOMEM_trans_kmalloc); + } - ret = bch2_trans_relock(trans); - if (ret) { - kfree(new_mem); - return ERR_PTR(ret); - } - } - memcpy(new_mem, trans->mem, trans->mem_top); - trans->used_mempool = false; - mempool_free(trans->mem, &c->btree_trans_mem_pool); - goto out_new_mem; + if (old_bytes) { + trans->realloc_bytes_required = new_bytes; + trace_and_count(c, trans_restart_mem_realloced, trans, _RET_IP_, new_bytes); + return ERR_PTR(btree_trans_restart_ip(trans, + BCH_ERR_transaction_restart_mem_realloced, _RET_IP_)); } - new_mem = krealloc(trans->mem, new_bytes, GFP_NOWAIT|__GFP_NOWARN); + EBUG_ON(trans->mem); + + new_mem = kmalloc(new_bytes, GFP_NOWAIT|__GFP_NOWARN); if (unlikely(!new_mem)) { bch2_trans_unlock(trans); - new_mem = krealloc(trans->mem, new_bytes, GFP_KERNEL); + new_mem = kmalloc(new_bytes, GFP_KERNEL); if (!new_mem && new_bytes <= BTREE_TRANS_MEM_MAX) { new_mem = mempool_alloc(&c->btree_trans_mem_pool, GFP_KERNEL); new_bytes = BTREE_TRANS_MEM_MAX; - memcpy(new_mem, trans->mem, trans->mem_top); trans->used_mempool = true; - kfree(trans->mem); } - if (!new_mem) - return ERR_PTR(-BCH_ERR_ENOMEM_trans_kmalloc); + EBUG_ON(!new_mem); trans->mem = new_mem; trans->mem_bytes = new_bytes; @@ -3207,18 +3259,10 @@ void *__bch2_trans_kmalloc(struct btree_trans *trans, size_t size, unsigned long if (ret) return ERR_PTR(ret); } -out_new_mem: + trans->mem = new_mem; trans->mem_bytes = new_bytes; - if (old_bytes) { - trace_and_count(c, trans_restart_mem_realloced, trans, _RET_IP_, new_bytes); - return ERR_PTR(btree_trans_restart_ip(trans, - BCH_ERR_transaction_restart_mem_realloced, _RET_IP_)); - } -out_change_top: - bch2_trans_kmalloc_trace(trans, size, ip); - p = trans->mem + trans->mem_top; trans->mem_top += size; memset(p, 0, size); @@ -3279,6 +3323,27 @@ u32 bch2_trans_begin(struct btree_trans *trans) trans->restart_count++; trans->mem_top = 0; + if (trans->restarted == BCH_ERR_transaction_restart_mem_realloced) { + EBUG_ON(!trans->mem || !trans->mem_bytes); + unsigned new_bytes = trans->realloc_bytes_required; + void *new_mem = krealloc(trans->mem, new_bytes, GFP_NOWAIT|__GFP_NOWARN); + if (unlikely(!new_mem)) { + bch2_trans_unlock(trans); + new_mem = krealloc(trans->mem, new_bytes, GFP_KERNEL); + + EBUG_ON(new_bytes > BTREE_TRANS_MEM_MAX); + + if (!new_mem) { + new_mem = mempool_alloc(&trans->c->btree_trans_mem_pool, GFP_KERNEL); + new_bytes = BTREE_TRANS_MEM_MAX; + trans->used_mempool = true; + kfree(trans->mem); + } + } + trans->mem = new_mem; + trans->mem_bytes = new_bytes; + } + trans_for_each_path(trans, path, i) { path->should_be_locked = false; @@ -3568,13 +3633,12 @@ bch2_btree_bkey_cached_common_to_text(struct printbuf *out, struct btree_bkey_cached_common *b) { struct six_lock_count c = six_lock_counts(&b->lock); - struct task_struct *owner; pid_t pid; - rcu_read_lock(); - owner = READ_ONCE(b->lock.owner); - pid = owner ? owner->pid : 0; - rcu_read_unlock(); + scoped_guard(rcu) { + struct task_struct *owner = READ_ONCE(b->lock.owner); + pid = owner ? owner->pid : 0; + } prt_printf(out, "\t%px %c ", b, b->cached ? 'c' : 'b'); bch2_btree_id_to_text(out, b->btree_id); @@ -3603,7 +3667,7 @@ void bch2_btree_trans_to_text(struct printbuf *out, struct btree_trans *trans) prt_printf(out, "%i %s\n", task ? task->pid : 0, trans->fn); /* trans->paths is rcu protected vs. freeing */ - rcu_read_lock(); + guard(rcu)(); out->atomic++; struct btree_path *paths = rcu_dereference(trans->paths); @@ -3646,7 +3710,6 @@ void bch2_btree_trans_to_text(struct printbuf *out, struct btree_trans *trans) } out: --out->atomic; - rcu_read_unlock(); } void bch2_fs_btree_iter_exit(struct bch_fs *c) diff --git a/fs/bcachefs/btree_iter.h b/fs/bcachefs/btree_iter.h index 2cabb5f0f484..09dd3e52622e 100644 --- a/fs/bcachefs/btree_iter.h +++ b/fs/bcachefs/btree_iter.h @@ -963,16 +963,6 @@ struct bkey_s_c bch2_btree_iter_peek_and_restart_outlined(struct btree_trans *, _p; \ }) -#define bch2_trans_run(_c, _do) \ -({ \ - struct btree_trans *trans = bch2_trans_get(_c); \ - int _ret = (_do); \ - bch2_trans_put(trans); \ - _ret; \ -}) - -#define bch2_trans_do(_c, _do) bch2_trans_run(_c, lockrestart_do(trans, _do)) - struct btree_trans *__bch2_trans_get(struct bch_fs *, unsigned); void bch2_trans_put(struct btree_trans *); @@ -990,6 +980,27 @@ unsigned bch2_trans_get_fn_idx(const char *); __bch2_trans_get(_c, trans_fn_idx); \ }) +/* + * We don't use DEFINE_CLASS() because using a function for the constructor + * breaks bch2_trans_get()'s use of __func__ + */ +typedef struct btree_trans * class_btree_trans_t; +static inline void class_btree_trans_destructor(struct btree_trans **p) +{ + struct btree_trans *trans = *p; + bch2_trans_put(trans); +} + +#define class_btree_trans_constructor(_c) bch2_trans_get(_c) + +#define bch2_trans_run(_c, _do) \ +({ \ + CLASS(btree_trans, trans)(_c); \ + (_do); \ +}) + +#define bch2_trans_do(_c, _do) bch2_trans_run(_c, lockrestart_do(trans, _do)) + void bch2_btree_trans_to_text(struct printbuf *, struct btree_trans *); void bch2_fs_btree_iter_exit(struct bch_fs *); diff --git a/fs/bcachefs/btree_journal_iter.c b/fs/bcachefs/btree_journal_iter.c index ade3b5addd75..ea839560a136 100644 --- a/fs/bcachefs/btree_journal_iter.c +++ b/fs/bcachefs/btree_journal_iter.c @@ -137,12 +137,15 @@ struct bkey_i *bch2_journal_keys_peek_prev_min(struct bch_fs *c, enum btree_id b struct journal_key *k; BUG_ON(*idx > keys->nr); + + if (!keys->nr) + return NULL; search: if (!*idx) *idx = __bch2_journal_key_search(keys, btree_id, level, pos); - while (*idx && - __journal_key_cmp(btree_id, level, end_pos, idx_to_key(keys, *idx - 1)) <= 0) { + while (*idx < keys->nr && + __journal_key_cmp(btree_id, level, end_pos, idx_to_key(keys, *idx)) >= 0) { (*idx)++; iters++; if (iters == 10) { @@ -151,18 +154,23 @@ search: } } + if (*idx == keys->nr) + --(*idx); + struct bkey_i *ret = NULL; rcu_read_lock(); /* for overwritten_ranges */ - while ((k = *idx < keys->nr ? idx_to_key(keys, *idx) : NULL)) { + while (true) { + k = idx_to_key(keys, *idx); if (__journal_key_cmp(btree_id, level, end_pos, k) > 0) break; if (k->overwritten) { if (k->overwritten_range) - *idx = rcu_dereference(k->overwritten_range)->start - 1; - else - *idx -= 1; + *idx = rcu_dereference(k->overwritten_range)->start; + if (!*idx) + break; + --(*idx); continue; } @@ -171,6 +179,8 @@ search: break; } + if (!*idx) + break; --(*idx); iters++; if (iters == 10) { @@ -292,7 +302,7 @@ int bch2_journal_key_insert_take(struct bch_fs *c, enum btree_id id, if (!new_keys.data) { bch_err(c, "%s: error allocating new key array (size %zu)", __func__, new_keys.size); - return -BCH_ERR_ENOMEM_journal_key_insert; + return bch_err_throw(c, ENOMEM_journal_key_insert); } /* Since @keys was full, there was no gap: */ @@ -331,7 +341,7 @@ int bch2_journal_key_insert(struct bch_fs *c, enum btree_id id, n = kmalloc(bkey_bytes(&k->k), GFP_KERNEL); if (!n) - return -BCH_ERR_ENOMEM_journal_key_insert; + return bch_err_throw(c, ENOMEM_journal_key_insert); bkey_copy(n, k); ret = bch2_journal_key_insert_take(c, id, level, n); @@ -457,11 +467,9 @@ static void bch2_journal_iter_advance(struct journal_iter *iter) static struct bkey_s_c bch2_journal_iter_peek(struct journal_iter *iter) { - struct bkey_s_c ret = bkey_s_c_null; - journal_iter_verify(iter); - rcu_read_lock(); + guard(rcu)(); while (iter->idx < iter->keys->size) { struct journal_key *k = iter->keys->data + iter->idx; @@ -470,19 +478,16 @@ static struct bkey_s_c bch2_journal_iter_peek(struct journal_iter *iter) break; BUG_ON(cmp); - if (!k->overwritten) { - ret = bkey_i_to_s_c(k->k); - break; - } + if (!k->overwritten) + return bkey_i_to_s_c(k->k); if (k->overwritten_range) iter->idx = idx_to_pos(iter->keys, rcu_dereference(k->overwritten_range)->end); else bch2_journal_iter_advance(iter); } - rcu_read_unlock(); - return ret; + return bkey_s_c_null; } static void bch2_journal_iter_exit(struct journal_iter *iter) @@ -646,10 +651,11 @@ static int journal_sort_key_cmp(const void *_l, const void *_r) { const struct journal_key *l = _l; const struct journal_key *r = _r; + int rewind = l->rewind && r->rewind ? -1 : 1; return journal_key_cmp(l, r) ?: - cmp_int(l->journal_seq, r->journal_seq) ?: - cmp_int(l->journal_offset, r->journal_offset); + ((cmp_int(l->journal_seq, r->journal_seq) ?: + cmp_int(l->journal_offset, r->journal_offset)) * rewind); } void bch2_journal_keys_put(struct bch_fs *c) @@ -718,6 +724,8 @@ int bch2_journal_keys_sort(struct bch_fs *c) struct journal_keys *keys = &c->journal_keys; size_t nr_read = 0; + u64 rewind_seq = c->opts.journal_rewind ?: U64_MAX; + genradix_for_each(&c->journal_entries, iter, _i) { i = *_i; @@ -726,28 +734,43 @@ int bch2_journal_keys_sort(struct bch_fs *c) cond_resched(); - for_each_jset_key(k, entry, &i->j) { - struct journal_key n = (struct journal_key) { - .btree_id = entry->btree_id, - .level = entry->level, - .k = k, - .journal_seq = le64_to_cpu(i->j.seq), - .journal_offset = k->_data - i->j._data, - }; - - if (darray_push(keys, n)) { - __journal_keys_sort(keys); - - if (keys->nr * 8 > keys->size * 7) { - bch_err(c, "Too many journal keys for slowpath; have %zu compacted, buf size %zu, processed %zu keys at seq %llu", - keys->nr, keys->size, nr_read, le64_to_cpu(i->j.seq)); - return -BCH_ERR_ENOMEM_journal_keys_sort; + vstruct_for_each(&i->j, entry) { + bool rewind = !entry->level && + !btree_id_is_alloc(entry->btree_id) && + le64_to_cpu(i->j.seq) >= rewind_seq; + + if (entry->type != (rewind + ? BCH_JSET_ENTRY_overwrite + : BCH_JSET_ENTRY_btree_keys)) + continue; + + if (!rewind && le64_to_cpu(i->j.seq) < c->journal_replay_seq_start) + continue; + + jset_entry_for_each_key(entry, k) { + struct journal_key n = (struct journal_key) { + .btree_id = entry->btree_id, + .level = entry->level, + .rewind = rewind, + .k = k, + .journal_seq = le64_to_cpu(i->j.seq), + .journal_offset = k->_data - i->j._data, + }; + + if (darray_push(keys, n)) { + __journal_keys_sort(keys); + + if (keys->nr * 8 > keys->size * 7) { + bch_err(c, "Too many journal keys for slowpath; have %zu compacted, buf size %zu, processed %zu keys at seq %llu", + keys->nr, keys->size, nr_read, le64_to_cpu(i->j.seq)); + return bch_err_throw(c, ENOMEM_journal_keys_sort); + } + + BUG_ON(darray_push(keys, n)); } - BUG_ON(darray_push(keys, n)); + nr_read++; } - - nr_read++; } } diff --git a/fs/bcachefs/btree_journal_iter_types.h b/fs/bcachefs/btree_journal_iter_types.h index 8b773823704f..86aacb254fb2 100644 --- a/fs/bcachefs/btree_journal_iter_types.h +++ b/fs/bcachefs/btree_journal_iter_types.h @@ -11,8 +11,9 @@ struct journal_key { u32 journal_offset; enum btree_id btree_id:8; unsigned level:8; - bool allocated; - bool overwritten; + bool allocated:1; + bool overwritten:1; + bool rewind:1; struct journal_key_range_overwritten __rcu * overwritten_range; struct bkey_i *k; diff --git a/fs/bcachefs/btree_key_cache.c b/fs/bcachefs/btree_key_cache.c index 9da950e7eb7d..d96188b92db2 100644 --- a/fs/bcachefs/btree_key_cache.c +++ b/fs/bcachefs/btree_key_cache.c @@ -187,27 +187,23 @@ lock: static struct bkey_cached * bkey_cached_reuse(struct btree_key_cache *c) { - struct bucket_table *tbl; + + guard(rcu)(); + struct bucket_table *tbl = rht_dereference_rcu(c->table.tbl, &c->table); struct rhash_head *pos; struct bkey_cached *ck; - unsigned i; - rcu_read_lock(); - tbl = rht_dereference_rcu(c->table.tbl, &c->table); - for (i = 0; i < tbl->size; i++) + for (unsigned i = 0; i < tbl->size; i++) rht_for_each_entry_rcu(ck, pos, tbl, i, hash) { if (!test_bit(BKEY_CACHED_DIRTY, &ck->flags) && bkey_cached_lock_for_evict(ck)) { if (bkey_cached_evict(c, ck)) - goto out; + return ck; six_unlock_write(&ck->c.lock); six_unlock_intent(&ck->c.lock); } } - ck = NULL; -out: - rcu_read_unlock(); - return ck; + return NULL; } static int btree_key_cache_create(struct btree_trans *trans, @@ -242,7 +238,7 @@ static int btree_key_cache_create(struct btree_trans *trans, if (unlikely(!ck)) { bch_err(c, "error allocating memory for key cache item, btree %s", bch2_btree_id_str(ck_path->btree_id)); - return -BCH_ERR_ENOMEM_btree_key_cache_create; + return bch_err_throw(c, ENOMEM_btree_key_cache_create); } } @@ -260,7 +256,7 @@ static int btree_key_cache_create(struct btree_trans *trans, if (unlikely(!new_k)) { bch_err(trans->c, "error allocating memory for key cache key, btree %s u64s %u", bch2_btree_id_str(ck->key.btree_id), key_u64s); - ret = -BCH_ERR_ENOMEM_btree_key_cache_fill; + ret = bch_err_throw(c, ENOMEM_btree_key_cache_fill); } else if (ret) { kfree(new_k); goto err; @@ -826,20 +822,20 @@ int bch2_fs_btree_key_cache_init(struct btree_key_cache *bc) bc->nr_pending = alloc_percpu(size_t); if (!bc->nr_pending) - return -BCH_ERR_ENOMEM_fs_btree_cache_init; + return bch_err_throw(c, ENOMEM_fs_btree_cache_init); if (rcu_pending_init(&bc->pending[0], &c->btree_trans_barrier, __bkey_cached_free) || rcu_pending_init(&bc->pending[1], &c->btree_trans_barrier, __bkey_cached_free)) - return -BCH_ERR_ENOMEM_fs_btree_cache_init; + return bch_err_throw(c, ENOMEM_fs_btree_cache_init); if (rhashtable_init(&bc->table, &bch2_btree_key_cache_params)) - return -BCH_ERR_ENOMEM_fs_btree_cache_init; + return bch_err_throw(c, ENOMEM_fs_btree_cache_init); bc->table_init_done = true; shrink = shrinker_alloc(0, "%s-btree_key_cache", c->name); if (!shrink) - return -BCH_ERR_ENOMEM_fs_btree_cache_init; + return bch_err_throw(c, ENOMEM_fs_btree_cache_init); bc->shrink = shrink; shrink->count_objects = bch2_btree_key_cache_count; shrink->scan_objects = bch2_btree_key_cache_scan; diff --git a/fs/bcachefs/btree_locking.c b/fs/bcachefs/btree_locking.c index 2f2aed0c9916..bed2b4b6ffb9 100644 --- a/fs/bcachefs/btree_locking.c +++ b/fs/bcachefs/btree_locking.c @@ -194,6 +194,30 @@ static int btree_trans_abort_preference(struct btree_trans *trans) return 3; } +static noinline __noreturn void break_cycle_fail(struct lock_graph *g) +{ + struct printbuf buf = PRINTBUF; + buf.atomic++; + + prt_printf(&buf, bch2_fmt(g->g->trans->c, "cycle of nofail locks")); + + for (struct trans_waiting_for_lock *i = g->g; i < g->g + g->nr; i++) { + struct btree_trans *trans = i->trans; + + bch2_btree_trans_to_text(&buf, trans); + + prt_printf(&buf, "backtrace:\n"); + printbuf_indent_add(&buf, 2); + bch2_prt_task_backtrace(&buf, trans->locking_wait.task, 2, GFP_NOWAIT); + printbuf_indent_sub(&buf, 2); + prt_newline(&buf); + } + + bch2_print_str(g->g->trans->c, KERN_ERR, buf.buf); + printbuf_exit(&buf); + BUG(); +} + static noinline int break_cycle(struct lock_graph *g, struct printbuf *cycle, struct trans_waiting_for_lock *from) { @@ -219,28 +243,8 @@ static noinline int break_cycle(struct lock_graph *g, struct printbuf *cycle, } } - if (unlikely(!best)) { - struct printbuf buf = PRINTBUF; - buf.atomic++; - - prt_printf(&buf, bch2_fmt(g->g->trans->c, "cycle of nofail locks")); - - for (i = g->g; i < g->g + g->nr; i++) { - struct btree_trans *trans = i->trans; - - bch2_btree_trans_to_text(&buf, trans); - - prt_printf(&buf, "backtrace:\n"); - printbuf_indent_add(&buf, 2); - bch2_prt_task_backtrace(&buf, trans->locking_wait.task, 2, GFP_NOWAIT); - printbuf_indent_sub(&buf, 2); - prt_newline(&buf); - } - - bch2_print_str_nonblocking(g->g->trans->c, KERN_ERR, buf.buf); - printbuf_exit(&buf); - BUG(); - } + if (unlikely(!best)) + break_cycle_fail(g); ret = abort_lock(g, abort); out: @@ -255,15 +259,14 @@ static int lock_graph_descend(struct lock_graph *g, struct btree_trans *trans, struct printbuf *cycle) { struct btree_trans *orig_trans = g->g->trans; - struct trans_waiting_for_lock *i; - for (i = g->g; i < g->g + g->nr; i++) + for (struct trans_waiting_for_lock *i = g->g; i < g->g + g->nr; i++) if (i->trans == trans) { closure_put(&trans->ref); return break_cycle(g, cycle, i); } - if (g->nr == ARRAY_SIZE(g->g)) { + if (unlikely(g->nr == ARRAY_SIZE(g->g))) { closure_put(&trans->ref); if (orig_trans->lock_may_not_fail) @@ -308,7 +311,7 @@ int bch2_check_for_deadlock(struct btree_trans *trans, struct printbuf *cycle) lock_graph_down(&g, trans); /* trans->paths is rcu protected vs. freeing */ - rcu_read_lock(); + guard(rcu)(); if (cycle) cycle->atomic++; next: @@ -406,7 +409,6 @@ up: out: if (cycle) --cycle->atomic; - rcu_read_unlock(); return ret; } @@ -769,7 +771,7 @@ static inline void __bch2_trans_unlock(struct btree_trans *trans) } static noinline __cold void bch2_trans_relock_fail(struct btree_trans *trans, struct btree_path *path, - struct get_locks_fail *f, bool trace) + struct get_locks_fail *f, bool trace, ulong ip) { if (!trace) goto out; @@ -794,7 +796,7 @@ static noinline __cold void bch2_trans_relock_fail(struct btree_trans *trans, st prt_printf(&buf, " total locked %u.%u.%u", c.n[0], c.n[1], c.n[2]); } - trace_trans_restart_relock(trans, _RET_IP_, buf.buf); + trace_trans_restart_relock(trans, ip, buf.buf); printbuf_exit(&buf); } @@ -804,7 +806,7 @@ out: bch2_trans_verify_locks(trans); } -static inline int __bch2_trans_relock(struct btree_trans *trans, bool trace) +static inline int __bch2_trans_relock(struct btree_trans *trans, bool trace, ulong ip) { bch2_trans_verify_locks(trans); @@ -823,7 +825,7 @@ static inline int __bch2_trans_relock(struct btree_trans *trans, bool trace) if (path->should_be_locked && (ret = btree_path_get_locks(trans, path, false, &f, BCH_ERR_transaction_restart_relock))) { - bch2_trans_relock_fail(trans, path, &f, trace); + bch2_trans_relock_fail(trans, path, &f, trace, ip); return ret; } } @@ -836,12 +838,12 @@ out: int bch2_trans_relock(struct btree_trans *trans) { - return __bch2_trans_relock(trans, true); + return __bch2_trans_relock(trans, true, _RET_IP_); } int bch2_trans_relock_notrace(struct btree_trans *trans) { - return __bch2_trans_relock(trans, false); + return __bch2_trans_relock(trans, false, _RET_IP_); } void bch2_trans_unlock(struct btree_trans *trans) diff --git a/fs/bcachefs/btree_locking.h b/fs/bcachefs/btree_locking.h index 9adca77e2580..f2173a3316f4 100644 --- a/fs/bcachefs/btree_locking.h +++ b/fs/bcachefs/btree_locking.h @@ -417,8 +417,10 @@ static inline void btree_path_set_should_be_locked(struct btree_trans *trans, st EBUG_ON(!btree_node_locked(path, path->level)); EBUG_ON(path->uptodate); - path->should_be_locked = true; - trace_btree_path_should_be_locked(trans, path); + if (!path->should_be_locked) { + path->should_be_locked = true; + trace_btree_path_should_be_locked(trans, path); + } } static inline void __btree_path_set_level_up(struct btree_trans *trans, diff --git a/fs/bcachefs/btree_node_scan.c b/fs/bcachefs/btree_node_scan.c index 5a97a6b8a757..42c9eb2c786e 100644 --- a/fs/bcachefs/btree_node_scan.c +++ b/fs/bcachefs/btree_node_scan.c @@ -75,39 +75,6 @@ static inline u64 bkey_journal_seq(struct bkey_s_c k) } } -static bool found_btree_node_is_readable(struct btree_trans *trans, - struct found_btree_node *f) -{ - struct { __BKEY_PADDED(k, BKEY_BTREE_PTR_VAL_U64s_MAX); } tmp; - - found_btree_node_to_key(&tmp.k, f); - - struct btree *b = bch2_btree_node_get_noiter(trans, &tmp.k, f->btree_id, f->level, false); - bool ret = !IS_ERR_OR_NULL(b); - if (!ret) - return ret; - - f->sectors_written = b->written; - f->journal_seq = le64_to_cpu(b->data->keys.journal_seq); - - struct bkey_s_c k; - struct bkey unpacked; - struct btree_node_iter iter; - for_each_btree_node_key_unpack(b, k, &iter, &unpacked) - f->journal_seq = max(f->journal_seq, bkey_journal_seq(k)); - - six_unlock_read(&b->c.lock); - - /* - * We might update this node's range; if that happens, we need the node - * to be re-read so the read path can trim keys that are no longer in - * this node - */ - if (b != btree_node_root(trans->c, b)) - bch2_btree_node_evict(trans, &tmp.k); - return ret; -} - static int found_btree_node_cmp_cookie(const void *_l, const void *_r) { const struct found_btree_node *l = _l; @@ -159,17 +126,17 @@ static const struct min_heap_callbacks found_btree_node_heap_cbs = { }; static void try_read_btree_node(struct find_btree_nodes *f, struct bch_dev *ca, - struct bio *bio, struct btree_node *bn, u64 offset) + struct btree *b, struct bio *bio, u64 offset) { struct bch_fs *c = container_of(f, struct bch_fs, found_btree_nodes); + struct btree_node *bn = b->data; bio_reset(bio, ca->disk_sb.bdev, REQ_OP_READ); bio->bi_iter.bi_sector = offset; - bch2_bio_map(bio, bn, PAGE_SIZE); + bch2_bio_map(bio, b->data, c->opts.block_size); u64 submit_time = local_clock(); submit_bio_wait(bio); - bch2_account_io_completion(ca, BCH_MEMBER_ERROR_read, submit_time, !bio->bi_status); if (bio->bi_status) { @@ -201,6 +168,14 @@ static void try_read_btree_node(struct find_btree_nodes *f, struct bch_dev *ca, if (BTREE_NODE_ID(bn) >= BTREE_ID_NR_MAX) return; + bio_reset(bio, ca->disk_sb.bdev, REQ_OP_READ); + bio->bi_iter.bi_sector = offset; + bch2_bio_map(bio, b->data, c->opts.btree_node_size); + + submit_time = local_clock(); + submit_bio_wait(bio); + bch2_account_io_completion(ca, BCH_MEMBER_ERROR_read, submit_time, !bio->bi_status); + rcu_read_lock(); struct found_btree_node n = { .btree_id = BTREE_NODE_ID(bn), @@ -217,7 +192,20 @@ static void try_read_btree_node(struct find_btree_nodes *f, struct bch_dev *ca, }; rcu_read_unlock(); - if (bch2_trans_run(c, found_btree_node_is_readable(trans, &n))) { + found_btree_node_to_key(&b->key, &n); + + CLASS(printbuf, buf)(); + if (!bch2_btree_node_read_done(c, ca, b, NULL, &buf)) { + /* read_done will swap out b->data for another buffer */ + bn = b->data; + /* + * Grab journal_seq here because we want the max journal_seq of + * any bset; read_done sorts down to a single set and picks the + * max journal_seq + */ + n.journal_seq = le64_to_cpu(bn->keys.journal_seq), + n.sectors_written = b->written; + mutex_lock(&f->lock); if (BSET_BIG_ENDIAN(&bn->keys) != CPU_BIG_ENDIAN) { bch_err(c, "try_read_btree_node() can't handle endian conversion"); @@ -237,12 +225,20 @@ static int read_btree_nodes_worker(void *p) struct find_btree_nodes_worker *w = p; struct bch_fs *c = container_of(w->f, struct bch_fs, found_btree_nodes); struct bch_dev *ca = w->ca; - void *buf = (void *) __get_free_page(GFP_KERNEL); - struct bio *bio = bio_alloc(NULL, 1, 0, GFP_KERNEL); unsigned long last_print = jiffies; + struct btree *b = NULL; + struct bio *bio = NULL; + + b = __bch2_btree_node_mem_alloc(c); + if (!b) { + bch_err(c, "read_btree_nodes_worker: error allocating buf"); + w->f->ret = -ENOMEM; + goto err; + } - if (!buf || !bio) { - bch_err(c, "read_btree_nodes_worker: error allocating bio/buf"); + bio = bio_alloc(NULL, buf_pages(b->data, c->opts.btree_node_size), 0, GFP_KERNEL); + if (!bio) { + bch_err(c, "read_btree_nodes_worker: error allocating bio"); w->f->ret = -ENOMEM; goto err; } @@ -266,11 +262,13 @@ static int read_btree_nodes_worker(void *p) !bch2_dev_btree_bitmap_marked_sectors(ca, sector, btree_sectors(c))) continue; - try_read_btree_node(w->f, ca, bio, buf, sector); + try_read_btree_node(w->f, ca, b, bio, sector); } err: + if (b) + __btree_node_data_free(b); + kfree(b); bio_put(bio); - free_page((unsigned long) buf); enumerated_ref_put(&ca->io_ref[READ], BCH_DEV_READ_REF_btree_node_scan); closure_put(w->cl); kfree(w); @@ -363,6 +361,8 @@ static int handle_overwrites(struct bch_fs *c, min_heap_sift_down(nodes_heap, 0, &found_btree_node_heap_cbs, NULL); } } + + cond_resched(); } return 0; @@ -519,8 +519,12 @@ bool bch2_btree_node_is_stale(struct bch_fs *c, struct btree *b) return false; } -bool bch2_btree_has_scanned_nodes(struct bch_fs *c, enum btree_id btree) +int bch2_btree_has_scanned_nodes(struct bch_fs *c, enum btree_id btree) { + int ret = bch2_run_print_explicit_recovery_pass(c, BCH_RECOVERY_PASS_scan_for_btree_nodes); + if (ret) + return ret; + struct found_btree_node search = { .btree_id = btree, .level = 0, diff --git a/fs/bcachefs/btree_node_scan.h b/fs/bcachefs/btree_node_scan.h index 08687b209787..66e6f9ed19d0 100644 --- a/fs/bcachefs/btree_node_scan.h +++ b/fs/bcachefs/btree_node_scan.h @@ -4,7 +4,7 @@ int bch2_scan_for_btree_nodes(struct bch_fs *); bool bch2_btree_node_is_stale(struct bch_fs *, struct btree *); -bool bch2_btree_has_scanned_nodes(struct bch_fs *, enum btree_id); +int bch2_btree_has_scanned_nodes(struct bch_fs *, enum btree_id); int bch2_get_scanned_nodes(struct bch_fs *, enum btree_id, unsigned, struct bpos, struct bpos); void bch2_find_btree_nodes_exit(struct find_btree_nodes *); diff --git a/fs/bcachefs/btree_trans_commit.c b/fs/bcachefs/btree_trans_commit.c index 1c03c965d836..639ef75b3dbd 100644 --- a/fs/bcachefs/btree_trans_commit.c +++ b/fs/bcachefs/btree_trans_commit.c @@ -376,7 +376,7 @@ static inline int btree_key_can_insert(struct btree_trans *trans, struct btree *b, unsigned u64s) { if (!bch2_btree_node_insert_fits(b, u64s)) - return -BCH_ERR_btree_insert_btree_node_full; + return bch_err_throw(trans->c, btree_insert_btree_node_full); return 0; } @@ -394,9 +394,10 @@ btree_key_can_insert_cached_slowpath(struct btree_trans *trans, unsigned flags, new_k = kmalloc(new_u64s * sizeof(u64), GFP_KERNEL); if (!new_k) { - bch_err(trans->c, "error allocating memory for key cache key, btree %s u64s %u", + struct bch_fs *c = trans->c; + 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; + return bch_err_throw(c, ENOMEM_btree_key_cache_insert); } ret = bch2_trans_relock(trans) ?: @@ -432,7 +433,7 @@ static int btree_key_can_insert_cached(struct btree_trans *trans, unsigned flags if (watermark < BCH_WATERMARK_reclaim && !test_bit(BKEY_CACHED_DIRTY, &ck->flags) && bch2_btree_key_cache_must_wait(c)) - return -BCH_ERR_btree_insert_need_journal_reclaim; + return bch_err_throw(c, btree_insert_need_journal_reclaim); /* * bch2_varint_decode can read past the end of the buffer by at most 7 @@ -594,12 +595,13 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags, int ret = 0; bch2_trans_verify_not_unlocked_or_in_restart(trans); - +#if 0 + /* todo: bring back dynamic fault injection */ if (race_fault()) { trace_and_count(c, trans_restart_fault_inject, trans, trace_ip); return btree_trans_restart(trans, BCH_ERR_transaction_restart_fault_inject); } - +#endif /* * Check if the insert will fit in the leaf node with the write lock * held, otherwise another thread could write the node changing the @@ -756,6 +758,8 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags, btree_trans_journal_entries_start(trans), trans->journal_entries.u64s); + EBUG_ON(trans->journal_res.u64s < trans->journal_entries.u64s); + trans->journal_res.offset += trans->journal_entries.u64s; trans->journal_res.u64s -= trans->journal_entries.u64s; @@ -894,7 +898,7 @@ int bch2_trans_commit_error(struct btree_trans *trans, unsigned flags, */ if ((flags & BCH_TRANS_COMMIT_journal_reclaim) && watermark < BCH_WATERMARK_reclaim) { - ret = -BCH_ERR_journal_reclaim_would_deadlock; + ret = bch_err_throw(c, journal_reclaim_would_deadlock); goto out; } @@ -966,13 +970,26 @@ do_bch2_trans_commit_to_journal_replay(struct btree_trans *trans) for (struct jset_entry *i = btree_trans_journal_entries_start(trans); i != btree_trans_journal_entries_top(trans); - i = vstruct_next(i)) + i = vstruct_next(i)) { if (i->type == BCH_JSET_ENTRY_btree_keys || i->type == BCH_JSET_ENTRY_write_buffer_keys) { - int ret = bch2_journal_key_insert(c, i->btree_id, i->level, i->start); - if (ret) - return ret; + jset_entry_for_each_key(i, k) { + int ret = bch2_journal_key_insert(c, i->btree_id, i->level, k); + if (ret) + return ret; + } + } + + if (i->type == BCH_JSET_ENTRY_btree_root) { + guard(mutex)(&c->btree_root_lock); + + struct btree_root *r = bch2_btree_id_root(c, i->btree_id); + + bkey_copy(&r->key, i->start); + r->level = i->level; + r->alive = true; } + } for (struct bkey_i *i = btree_trans_subbuf_base(trans, &trans->accounting); i != btree_trans_subbuf_top(trans, &trans->accounting); @@ -989,6 +1006,7 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags) { struct btree_insert_entry *errored_at = NULL; struct bch_fs *c = trans->c; + unsigned journal_u64s = 0; int ret = 0; bch2_trans_verify_not_unlocked_or_in_restart(trans); @@ -1011,16 +1029,16 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags) if (unlikely(!test_bit(BCH_FS_may_go_rw, &c->flags))) ret = do_bch2_trans_commit_to_journal_replay(trans); else - ret = -BCH_ERR_erofs_trans_commit; + ret = bch_err_throw(c, erofs_trans_commit); goto out_reset; } EBUG_ON(test_bit(BCH_FS_clean_shutdown, &c->flags)); - trans->journal_u64s = trans->journal_entries.u64s + jset_u64s(trans->accounting.u64s); + journal_u64s = jset_u64s(trans->accounting.u64s); 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); + journal_u64s += jset_u64s(JSET_ENTRY_LOG_U64s); trans_for_each_update(trans, i) { struct btree_path *path = trans->paths + i->path; @@ -1040,11 +1058,11 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags) continue; /* we're going to journal the key being updated: */ - trans->journal_u64s += jset_u64s(i->k->k.u64s); + journal_u64s += jset_u64s(i->k->k.u64s); /* and we're also going to log the overwrite: */ if (trans->journal_transaction_names) - trans->journal_u64s += jset_u64s(i->old_k.u64s); + journal_u64s += jset_u64s(i->old_k.u64s); } if (trans->extra_disk_res) { @@ -1062,6 +1080,8 @@ retry: memset(&trans->journal_res, 0, sizeof(trans->journal_res)); memset(&trans->fs_usage_delta, 0, sizeof(trans->fs_usage_delta)); + trans->journal_u64s = journal_u64s + trans->journal_entries.u64s; + ret = do_bch2_trans_commit(trans, flags, &errored_at, _RET_IP_); /* make sure we didn't drop or screw up locks: */ @@ -1093,7 +1113,7 @@ err: * restart: */ if (flags & BCH_TRANS_COMMIT_no_journal_res) { - ret = -BCH_ERR_transaction_restart_nested; + ret = bch_err_throw(c, transaction_restart_nested); goto out; } diff --git a/fs/bcachefs/btree_types.h b/fs/bcachefs/btree_types.h index 9d641bf9d2a2..112170fd9c8f 100644 --- a/fs/bcachefs/btree_types.h +++ b/fs/bcachefs/btree_types.h @@ -497,6 +497,7 @@ struct btree_trans { void *mem; unsigned mem_top; unsigned mem_bytes; + unsigned realloc_bytes_required; #ifdef CONFIG_BCACHEFS_TRANS_KMALLOC_TRACE darray_trans_kmalloc_trace trans_kmalloc_trace; #endif @@ -555,6 +556,8 @@ struct btree_trans { unsigned journal_u64s; unsigned extra_disk_res; /* XXX kill */ + __BKEY_PADDED(btree_path_down, BKEY_BTREE_PTR_VAL_U64s_MAX); + #ifdef CONFIG_DEBUG_LOCK_ALLOC struct lockdep_map dep_map; #endif @@ -615,6 +618,9 @@ enum btree_write_type { x(dying) \ x(fake) \ x(need_rewrite) \ + x(need_rewrite_error) \ + x(need_rewrite_degraded) \ + x(need_rewrite_ptr_written_zero) \ x(never_write) \ x(pinned) @@ -639,6 +645,32 @@ static inline void clear_btree_node_ ## flag(struct btree *b) \ BTREE_FLAGS() #undef x +#define BTREE_NODE_REWRITE_REASON() \ + x(none) \ + x(unknown) \ + x(error) \ + x(degraded) \ + x(ptr_written_zero) + +enum btree_node_rewrite_reason { +#define x(n) BTREE_NODE_REWRITE_##n, + BTREE_NODE_REWRITE_REASON() +#undef x +}; + +static inline enum btree_node_rewrite_reason btree_node_rewrite_reason(struct btree *b) +{ + if (btree_node_need_rewrite_ptr_written_zero(b)) + return BTREE_NODE_REWRITE_ptr_written_zero; + if (btree_node_need_rewrite_degraded(b)) + return BTREE_NODE_REWRITE_degraded; + if (btree_node_need_rewrite_error(b)) + return BTREE_NODE_REWRITE_error; + if (btree_node_need_rewrite(b)) + return BTREE_NODE_REWRITE_unknown; + return BTREE_NODE_REWRITE_none; +} + static inline struct btree_write *btree_current_write(struct btree *b) { return b->writes + btree_node_write_idx(b); diff --git a/fs/bcachefs/btree_update.c b/fs/bcachefs/btree_update.c index 5dac09c98026..ee657b9f4b96 100644 --- a/fs/bcachefs/btree_update.c +++ b/fs/bcachefs/btree_update.c @@ -123,65 +123,44 @@ static int need_whiteout_for_snapshot(struct btree_trans *trans, } int __bch2_insert_snapshot_whiteouts(struct btree_trans *trans, - enum btree_id id, - struct bpos old_pos, - struct bpos new_pos) + enum btree_id btree, struct bpos pos, + snapshot_id_list *s) { - struct bch_fs *c = trans->c; - struct btree_iter old_iter, new_iter = {}; - struct bkey_s_c old_k, new_k; - snapshot_id_list s; - struct bkey_i *update; int ret = 0; - if (!bch2_snapshot_has_children(c, old_pos.snapshot)) - return 0; + darray_for_each(*s, id) { + pos.snapshot = *id; - darray_init(&s); - - bch2_trans_iter_init(trans, &old_iter, id, old_pos, - BTREE_ITER_not_extents| - BTREE_ITER_all_snapshots); - while ((old_k = bch2_btree_iter_prev(trans, &old_iter)).k && - !(ret = bkey_err(old_k)) && - bkey_eq(old_pos, old_k.k->p)) { - struct bpos whiteout_pos = - SPOS(new_pos.inode, new_pos.offset, old_k.k->p.snapshot); - - if (!bch2_snapshot_is_ancestor(c, old_k.k->p.snapshot, old_pos.snapshot) || - snapshot_list_has_ancestor(c, &s, old_k.k->p.snapshot)) - continue; - - new_k = bch2_bkey_get_iter(trans, &new_iter, id, whiteout_pos, - BTREE_ITER_not_extents| - BTREE_ITER_intent); - ret = bkey_err(new_k); + struct btree_iter iter; + struct bkey_s_c k = bch2_bkey_get_iter(trans, &iter, btree, pos, + BTREE_ITER_not_extents| + BTREE_ITER_intent); + ret = bkey_err(k); if (ret) break; - if (new_k.k->type == KEY_TYPE_deleted) { - update = bch2_trans_kmalloc(trans, sizeof(struct bkey_i)); + if (k.k->type == KEY_TYPE_deleted) { + struct bkey_i *update = bch2_trans_kmalloc(trans, sizeof(struct bkey_i)); ret = PTR_ERR_OR_ZERO(update); - if (ret) + if (ret) { + bch2_trans_iter_exit(trans, &iter); break; + } bkey_init(&update->k); - update->k.p = whiteout_pos; + update->k.p = pos; update->k.type = KEY_TYPE_whiteout; - ret = bch2_trans_update(trans, &new_iter, update, + ret = bch2_trans_update(trans, &iter, update, BTREE_UPDATE_internal_snapshot_node); } - bch2_trans_iter_exit(trans, &new_iter); + bch2_trans_iter_exit(trans, &iter); - ret = snapshot_list_add(c, &s, old_k.k->p.snapshot); if (ret) break; } - bch2_trans_iter_exit(trans, &new_iter); - bch2_trans_iter_exit(trans, &old_iter); - darray_exit(&s); + darray_exit(s); return ret; } @@ -570,20 +549,26 @@ void *__bch2_trans_subbuf_alloc(struct btree_trans *trans, unsigned u64s) { unsigned new_top = buf->u64s + u64s; - unsigned old_size = buf->size; + unsigned new_size = buf->size; - if (new_top > buf->size) - buf->size = roundup_pow_of_two(new_top); + BUG_ON(roundup_pow_of_two(new_top) > U16_MAX); - void *n = bch2_trans_kmalloc_nomemzero(trans, buf->size * sizeof(u64)); + if (new_top > new_size) + new_size = roundup_pow_of_two(new_top); + + void *n = bch2_trans_kmalloc_nomemzero(trans, new_size * sizeof(u64)); if (IS_ERR(n)) return n; + unsigned offset = (u64 *) n - (u64 *) trans->mem; + BUG_ON(offset > U16_MAX); + if (buf->u64s) memcpy(n, btree_trans_subbuf_base(trans, buf), - old_size * sizeof(u64)); + buf->size * sizeof(u64)); buf->base = (u64 *) n - (u64 *) trans->mem; + buf->size = new_size; void *p = btree_trans_subbuf_top(trans, buf); buf->u64s = new_top; @@ -608,7 +593,7 @@ int bch2_bkey_get_empty_slot(struct btree_trans *trans, struct btree_iter *iter, BUG_ON(k.k->type != KEY_TYPE_deleted); if (bkey_gt(k.k->p, end)) { - ret = -BCH_ERR_ENOSPC_btree_slot; + ret = bch_err_throw(trans->c, ENOSPC_btree_slot); goto err; } diff --git a/fs/bcachefs/btree_update.h b/fs/bcachefs/btree_update.h index f907eaa8b185..0b98ab959719 100644 --- a/fs/bcachefs/btree_update.h +++ b/fs/bcachefs/btree_update.h @@ -4,6 +4,7 @@ #include "btree_iter.h" #include "journal.h" +#include "snapshot.h" struct bch_fs; struct btree; @@ -74,7 +75,7 @@ static inline int bch2_btree_delete_at_buffered(struct btree_trans *trans, } int __bch2_insert_snapshot_whiteouts(struct btree_trans *, enum btree_id, - struct bpos, struct bpos); + struct bpos, snapshot_id_list *); /* * For use when splitting extents in existing snapshots: @@ -88,11 +89,20 @@ static inline int bch2_insert_snapshot_whiteouts(struct btree_trans *trans, struct bpos old_pos, struct bpos new_pos) { + BUG_ON(old_pos.snapshot != new_pos.snapshot); + if (!btree_type_has_snapshots(btree) || bkey_eq(old_pos, new_pos)) return 0; - return __bch2_insert_snapshot_whiteouts(trans, btree, old_pos, new_pos); + snapshot_id_list s; + int ret = bch2_get_snapshot_overwrites(trans, btree, old_pos, &s); + if (ret) + return ret; + + return s.nr + ? __bch2_insert_snapshot_whiteouts(trans, btree, new_pos, &s) + : 0; } int bch2_trans_update_extent_overwrite(struct btree_trans *, struct btree_iter *, @@ -160,8 +170,7 @@ bch2_trans_jset_entry_alloc(struct btree_trans *trans, unsigned u64s) int bch2_btree_insert_clone_trans(struct btree_trans *, enum btree_id, struct bkey_i *); -int bch2_btree_write_buffer_insert_err(struct btree_trans *, - enum btree_id, struct bkey_i *); +int bch2_btree_write_buffer_insert_err(struct bch_fs *, enum btree_id, struct bkey_i *); static inline int __must_check bch2_trans_update_buffered(struct btree_trans *trans, enum btree_id btree, @@ -172,7 +181,7 @@ static inline int __must_check bch2_trans_update_buffered(struct btree_trans *tr EBUG_ON(k->k.u64s > BTREE_WRITE_BUFERED_U64s_MAX); if (unlikely(!btree_type_uses_write_buffer(btree))) { - int ret = bch2_btree_write_buffer_insert_err(trans, btree, k); + int ret = bch2_btree_write_buffer_insert_err(trans->c, btree, k); dump_stack(); return ret; } diff --git a/fs/bcachefs/btree_update_interior.c b/fs/bcachefs/btree_update_interior.c index 74e65714fecd..553059b33bfd 100644 --- a/fs/bcachefs/btree_update_interior.c +++ b/fs/bcachefs/btree_update_interior.c @@ -57,8 +57,6 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b) struct bkey_buf prev; int ret = 0; - printbuf_indent_add_nextline(&buf, 2); - BUG_ON(b->key.k.type == KEY_TYPE_btree_ptr_v2 && !bpos_eq(bkey_i_to_btree_ptr_v2(&b->key)->v.min_key, b->data->min_key)); @@ -69,20 +67,23 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b) if (b == btree_node_root(c, b)) { if (!bpos_eq(b->data->min_key, POS_MIN)) { - ret = __bch2_topology_error(c, &buf); - + bch2_log_msg_start(c, &buf); + prt_printf(&buf, "btree root with incorrect min_key: "); bch2_bpos_to_text(&buf, b->data->min_key); - log_fsck_err(trans, btree_root_bad_min_key, - "btree root with incorrect min_key: %s", buf.buf); - goto out; + prt_newline(&buf); + + bch2_count_fsck_err(c, btree_root_bad_min_key, &buf); + goto err; } if (!bpos_eq(b->data->max_key, SPOS_MAX)) { - ret = __bch2_topology_error(c, &buf); + bch2_log_msg_start(c, &buf); + prt_printf(&buf, "btree root with incorrect max_key: "); bch2_bpos_to_text(&buf, b->data->max_key); - log_fsck_err(trans, btree_root_bad_max_key, - "btree root with incorrect max_key: %s", buf.buf); - goto out; + prt_newline(&buf); + + bch2_count_fsck_err(c, btree_root_bad_max_key, &buf); + goto err; } } @@ -100,19 +101,15 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b) : bpos_successor(prev.k->k.p); if (!bpos_eq(expected_min, bp.v->min_key)) { - ret = __bch2_topology_error(c, &buf); - - prt_str(&buf, "end of prev node doesn't match start of next node\nin "); - bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level); - prt_str(&buf, " node "); - bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key)); + prt_str(&buf, "end of prev node doesn't match start of next node"); prt_str(&buf, "\nprev "); bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(prev.k)); prt_str(&buf, "\nnext "); bch2_bkey_val_to_text(&buf, c, k); + prt_newline(&buf); - log_fsck_err(trans, btree_node_topology_bad_min_key, "%s", buf.buf); - goto out; + bch2_count_fsck_err(c, btree_node_topology_bad_min_key, &buf); + goto err; } bch2_bkey_buf_reassemble(&prev, c, k); @@ -120,32 +117,34 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b) } if (bkey_deleted(&prev.k->k)) { - ret = __bch2_topology_error(c, &buf); - - prt_str(&buf, "empty interior node\nin "); - bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level); - prt_str(&buf, " node "); - bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key)); - - log_fsck_err(trans, btree_node_topology_empty_interior_node, "%s", buf.buf); - } else if (!bpos_eq(prev.k->k.p, b->key.k.p)) { - ret = __bch2_topology_error(c, &buf); + prt_printf(&buf, "empty interior node\n"); + bch2_count_fsck_err(c, btree_node_topology_empty_interior_node, &buf); + goto err; + } - prt_str(&buf, "last child node doesn't end at end of parent node\nin "); - bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level); - prt_str(&buf, " node "); - bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key)); - prt_str(&buf, "\nlast key "); + if (!bpos_eq(prev.k->k.p, b->key.k.p)) { + prt_str(&buf, "last child node doesn't end at end of parent node\nchild: "); bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(prev.k)); + prt_newline(&buf); - log_fsck_err(trans, btree_node_topology_bad_max_key, "%s", buf.buf); + bch2_count_fsck_err(c, btree_node_topology_bad_max_key, &buf); + goto err; } out: -fsck_err: bch2_btree_and_journal_iter_exit(&iter); bch2_bkey_buf_exit(&prev, c); printbuf_exit(&buf); return ret; +err: + bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level); + prt_char(&buf, ' '); + bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key)); + prt_newline(&buf); + + ret = __bch2_topology_error(c, &buf); + bch2_print_str(c, KERN_ERR, buf.buf); + BUG_ON(!ret); + goto out; } /* Calculate ideal packed bkey format for new btree nodes: */ @@ -685,12 +684,31 @@ static void btree_update_nodes_written(struct btree_update *as) /* * Wait for any in flight writes to finish before we free the old nodes - * on disk: + * on disk. But we haven't pinned those old nodes in the btree cache, + * they might have already been evicted. + * + * The update we're completing deleted references to those nodes from the + * btree, so we know if they've been evicted they can't be pulled back in. + * We just have to check if the nodes we have pointers to are still those + * old nodes, and haven't been reused. + * + * This can't be done locklessly because the data buffer might have been + * vmalloc allocated, and they're not RCU freed. We also need the + * __no_kmsan_checks annotation because even with the btree node read + * lock, nothing tells us that the data buffer has been initialized (if + * the btree node has been reused for a different node, and the data + * buffer swapped for a new data buffer). */ for (i = 0; i < as->nr_old_nodes; i++) { b = as->old_nodes[i]; - if (btree_node_seq_matches(b, as->old_nodes_seq[i])) + bch2_trans_begin(trans); + btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_read); + bool seq_matches = btree_node_seq_matches(b, as->old_nodes_seq[i]); + six_unlock_read(&b->c.lock); + bch2_trans_unlock_long(trans); + + if (seq_matches) wait_on_bit_io(&b->flags, BTREE_NODE_write_in_flight_inner, TASK_UNINTERRUPTIBLE); } @@ -1120,6 +1138,13 @@ static void bch2_btree_update_done(struct btree_update *as, struct btree_trans * start_time); } +static const char * const btree_node_reawrite_reason_strs[] = { +#define x(n) #n, + BTREE_NODE_REWRITE_REASON() +#undef x + NULL, +}; + static struct btree_update * bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, unsigned level_start, bool split, @@ -1214,6 +1239,15 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, list_add_tail(&as->list, &c->btree_interior_update_list); mutex_unlock(&c->btree_interior_update_lock); + struct btree *b = btree_path_node(path, path->level); + as->node_start = b->data->min_key; + as->node_end = b->data->max_key; + as->node_needed_rewrite = btree_node_rewrite_reason(b); + as->node_written = b->written; + as->node_sectors = btree_buf_bytes(b) >> 9; + as->node_remaining = __bch2_btree_u64s_remaining(b, + btree_bkey_last(b, bset_tree_last(b))); + /* * We don't want to allocate if we're in an error state, that can cause * deadlock on emergency shutdown due to open buckets getting stuck in @@ -1245,7 +1279,7 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, if (bch2_err_matches(ret, ENOSPC) && (flags & BCH_TRANS_COMMIT_journal_reclaim) && watermark < BCH_WATERMARK_reclaim) { - ret = -BCH_ERR_journal_reclaim_would_deadlock; + ret = bch_err_throw(c, journal_reclaim_would_deadlock); goto err; } @@ -1253,10 +1287,11 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path, do { ret = bch2_btree_reserve_get(trans, as, nr_nodes, target, flags, &cl); - + if (!bch2_err_matches(ret, BCH_ERR_operation_blocked)) + break; bch2_trans_unlock(trans); bch2_wait_on_allocator(c, &cl); - } while (bch2_err_matches(ret, BCH_ERR_operation_blocked)); + } while (1); } if (ret) { @@ -2090,6 +2125,9 @@ int __bch2_foreground_maybe_merge(struct btree_trans *trans, if (ret) goto err; + as->node_start = prev->data->min_key; + as->node_end = next->data->max_key; + trace_and_count(c, btree_node_merge, trans, b); n = bch2_btree_node_alloc(as, trans, b->c.level); @@ -2178,7 +2216,7 @@ static int get_iter_to_node(struct btree_trans *trans, struct btree_iter *iter, if (btree_iter_path(trans, iter)->l[b->c.level].b != b) { /* node has been freed: */ BUG_ON(!btree_node_dying(b)); - ret = -BCH_ERR_btree_node_dying; + ret = bch_err_throw(trans->c, btree_node_dying); goto err; } @@ -2256,9 +2294,9 @@ err: goto out; } -static int bch2_btree_node_rewrite_key(struct btree_trans *trans, - enum btree_id btree, unsigned level, - struct bkey_i *k, unsigned flags) +int bch2_btree_node_rewrite_key(struct btree_trans *trans, + enum btree_id btree, unsigned level, + struct bkey_i *k, unsigned flags) { struct btree_iter iter; bch2_trans_node_iter_init(trans, &iter, @@ -2330,9 +2368,8 @@ static void async_btree_node_rewrite_work(struct work_struct *work) int ret = bch2_trans_do(c, bch2_btree_node_rewrite_key(trans, a->btree_id, a->level, a->key.k, 0)); - if (ret != -ENOENT && - !bch2_err_matches(ret, EROFS) && - ret != -BCH_ERR_journal_shutdown) + if (!bch2_err_matches(ret, ENOENT) && + !bch2_err_matches(ret, EROFS)) bch_err_fn_ratelimited(c, ret); spin_lock(&c->btree_node_rewrites_lock); @@ -2663,9 +2700,19 @@ static void bch2_btree_update_to_text(struct printbuf *out, struct btree_update prt_str(out, " "); bch2_btree_id_to_text(out, as->btree_id); - prt_printf(out, " l=%u-%u mode=%s nodes_written=%u cl.remaining=%u journal_seq=%llu\n", + prt_printf(out, " l=%u-%u ", as->update_level_start, - as->update_level_end, + as->update_level_end); + bch2_bpos_to_text(out, as->node_start); + prt_char(out, ' '); + bch2_bpos_to_text(out, as->node_end); + prt_printf(out, "\nwritten %u/%u u64s_remaining %u need_rewrite %s", + as->node_written, + as->node_sectors, + as->node_remaining, + btree_node_reawrite_reason_strs[as->node_needed_rewrite]); + + prt_printf(out, "\nmode=%s nodes_written=%u cl.remaining=%u journal_seq=%llu\n", bch2_btree_update_modes[as->mode], as->nodes_written, closure_nr_remaining(&as->cl), @@ -2792,16 +2839,16 @@ int bch2_fs_btree_interior_update_init(struct bch_fs *c) c->btree_interior_update_worker = alloc_workqueue("btree_update", WQ_UNBOUND|WQ_MEM_RECLAIM, 8); if (!c->btree_interior_update_worker) - return -BCH_ERR_ENOMEM_btree_interior_update_worker_init; + return bch_err_throw(c, ENOMEM_btree_interior_update_worker_init); c->btree_node_rewrite_worker = alloc_ordered_workqueue("btree_node_rewrite", WQ_UNBOUND); if (!c->btree_node_rewrite_worker) - return -BCH_ERR_ENOMEM_btree_interior_update_worker_init; + return bch_err_throw(c, ENOMEM_btree_interior_update_worker_init); if (mempool_init_kmalloc_pool(&c->btree_interior_update_pool, 1, sizeof(struct btree_update))) - return -BCH_ERR_ENOMEM_btree_interior_update_pool_init; + return bch_err_throw(c, ENOMEM_btree_interior_update_pool_init); return 0; } diff --git a/fs/bcachefs/btree_update_interior.h b/fs/bcachefs/btree_update_interior.h index 7fe793788a79..ac04e45a8515 100644 --- a/fs/bcachefs/btree_update_interior.h +++ b/fs/bcachefs/btree_update_interior.h @@ -57,6 +57,13 @@ struct btree_update { unsigned took_gc_lock:1; enum btree_id btree_id; + struct bpos node_start; + struct bpos node_end; + enum btree_node_rewrite_reason node_needed_rewrite; + u16 node_written; + u16 node_sectors; + u16 node_remaining; + unsigned update_level_start; unsigned update_level_end; @@ -169,6 +176,9 @@ static inline int bch2_foreground_maybe_merge(struct btree_trans *trans, int bch2_btree_node_rewrite(struct btree_trans *, struct btree_iter *, struct btree *, unsigned, unsigned); +int bch2_btree_node_rewrite_key(struct btree_trans *, + enum btree_id, unsigned, + struct bkey_i *, unsigned); int bch2_btree_node_rewrite_pos(struct btree_trans *, enum btree_id, unsigned, struct bpos, unsigned, unsigned); diff --git a/fs/bcachefs/btree_write_buffer.c b/fs/bcachefs/btree_write_buffer.c index efb0c64d0aac..4b095235a0d2 100644 --- a/fs/bcachefs/btree_write_buffer.c +++ b/fs/bcachefs/btree_write_buffer.c @@ -267,10 +267,9 @@ out: BUG_ON(wb->sorted.size < wb->flushing.keys.nr); } -int bch2_btree_write_buffer_insert_err(struct btree_trans *trans, +int bch2_btree_write_buffer_insert_err(struct bch_fs *c, enum btree_id btree, struct bkey_i *k) { - struct bch_fs *c = trans->c; struct printbuf buf = PRINTBUF; prt_printf(&buf, "attempting to do write buffer update on non wb btree="); @@ -332,7 +331,7 @@ static int bch2_btree_write_buffer_flush_locked(struct btree_trans *trans) struct btree_write_buffered_key *k = &wb->flushing.keys.data[i->idx]; if (unlikely(!btree_type_uses_write_buffer(k->btree))) { - ret = bch2_btree_write_buffer_insert_err(trans, k->btree, &k->k); + ret = bch2_btree_write_buffer_insert_err(trans->c, k->btree, &k->k); goto err; } @@ -394,7 +393,7 @@ static int bch2_btree_write_buffer_flush_locked(struct btree_trans *trans) bool accounting_accumulated = false; do { if (race_fault()) { - ret = -BCH_ERR_journal_reclaim_would_deadlock; + ret = bch_err_throw(c, journal_reclaim_would_deadlock); break; } @@ -633,7 +632,7 @@ int bch2_btree_write_buffer_tryflush(struct btree_trans *trans) struct bch_fs *c = trans->c; if (!enumerated_ref_tryget(&c->writes, BCH_WRITE_REF_btree_write_buffer)) - return -BCH_ERR_erofs_no_writes; + return bch_err_throw(c, erofs_no_writes); int ret = bch2_btree_write_buffer_flush_nocheck_rw(trans); enumerated_ref_put(&c->writes, BCH_WRITE_REF_btree_write_buffer); @@ -676,7 +675,10 @@ int bch2_btree_write_buffer_maybe_flush(struct btree_trans *trans, goto err; bch2_bkey_buf_copy(last_flushed, c, tmp.k); - ret = -BCH_ERR_transaction_restart_write_buffer_flush; + + /* can we avoid the unconditional restart? */ + trace_and_count(c, trans_restart_write_buffer_flush, trans, _RET_IP_); + ret = bch_err_throw(c, transaction_restart_write_buffer_flush); } err: bch2_bkey_buf_exit(&tmp, c); diff --git a/fs/bcachefs/btree_write_buffer.h b/fs/bcachefs/btree_write_buffer.h index 05f56fd1eed0..c351d21aca0b 100644 --- a/fs/bcachefs/btree_write_buffer.h +++ b/fs/bcachefs/btree_write_buffer.h @@ -89,6 +89,12 @@ static inline int bch2_journal_key_to_wb(struct bch_fs *c, struct journal_keys_to_wb *dst, enum btree_id btree, struct bkey_i *k) { + if (unlikely(!btree_type_uses_write_buffer(btree))) { + int ret = bch2_btree_write_buffer_insert_err(c, btree, k); + dump_stack(); + return ret; + } + EBUG_ON(!dst->seq); return k->k.type == KEY_TYPE_accounting diff --git a/fs/bcachefs/buckets.c b/fs/bcachefs/buckets.c index 09eb5a543ae4..f25903c10e8a 100644 --- a/fs/bcachefs/buckets.c +++ b/fs/bcachefs/buckets.c @@ -221,6 +221,20 @@ static int bch2_check_fix_ptr(struct btree_trans *trans, bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { if (!p.ptr.cached && data_type == BCH_DATA_btree) { + switch (g->data_type) { + case BCH_DATA_sb: + bch_err(c, "btree and superblock in the same bucket - cannot repair"); + ret = bch_err_throw(c, fsck_repair_unimplemented); + goto out; + case BCH_DATA_journal: + ret = bch2_dev_journal_bucket_delete(ca, PTR_BUCKET_NR(ca, &p.ptr)); + bch_err_msg(c, ret, "error deleting journal bucket %zu", + PTR_BUCKET_NR(ca, &p.ptr)); + if (ret) + goto out; + break; + } + g->data_type = data_type; g->stripe_sectors = 0; g->dirty_sectors = 0; @@ -270,6 +284,9 @@ int bch2_check_fix_ptrs(struct btree_trans *trans, struct printbuf buf = PRINTBUF; int ret = 0; + /* We don't yet do btree key updates correctly for when we're RW */ + BUG_ON(test_bit(BCH_FS_rw, &c->flags)); + bkey_for_each_ptr_decode(k.k, ptrs_c, p, entry_c) { ret = bch2_check_fix_ptr(trans, k, p, entry_c, &do_update); if (ret) @@ -277,20 +294,13 @@ int bch2_check_fix_ptrs(struct btree_trans *trans, } if (do_update) { - if (flags & BTREE_TRIGGER_is_root) { - bch_err(c, "cannot update btree roots yet"); - ret = -EINVAL; - goto err; - } - struct bkey_i *new = bch2_bkey_make_mut_noupdate(trans, k); ret = PTR_ERR_OR_ZERO(new); if (ret) goto err; - rcu_read_lock(); - bch2_bkey_drop_ptrs(bkey_i_to_s(new), ptr, !bch2_dev_exists(c, ptr->dev)); - rcu_read_unlock(); + scoped_guard(rcu) + bch2_bkey_drop_ptrs(bkey_i_to_s(new), ptr, !bch2_dev_exists(c, ptr->dev)); if (level) { /* @@ -299,14 +309,11 @@ int bch2_check_fix_ptrs(struct btree_trans *trans, * sort it out: */ struct bkey_ptrs ptrs = bch2_bkey_ptrs(bkey_i_to_s(new)); - rcu_read_lock(); - bkey_for_each_ptr(ptrs, ptr) { - struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); - struct bucket *g = PTR_GC_BUCKET(ca, ptr); - - ptr->gen = g->gen; - } - rcu_read_unlock(); + scoped_guard(rcu) + bkey_for_each_ptr(ptrs, ptr) { + struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); + ptr->gen = PTR_GC_BUCKET(ca, ptr)->gen; + } } else { struct bkey_ptrs ptrs; union bch_extent_entry *entry; @@ -370,19 +377,41 @@ found: bch_info(c, "new key %s", buf.buf); } - struct btree_iter iter; - bch2_trans_node_iter_init(trans, &iter, btree, new->k.p, 0, level, - BTREE_ITER_intent|BTREE_ITER_all_snapshots); - ret = bch2_btree_iter_traverse(trans, &iter) ?: - bch2_trans_update(trans, &iter, new, - BTREE_UPDATE_internal_snapshot_node| - BTREE_TRIGGER_norun); - bch2_trans_iter_exit(trans, &iter); - if (ret) - goto err; + if (!(flags & BTREE_TRIGGER_is_root)) { + struct btree_iter iter; + bch2_trans_node_iter_init(trans, &iter, btree, new->k.p, 0, level, + BTREE_ITER_intent|BTREE_ITER_all_snapshots); + ret = bch2_btree_iter_traverse(trans, &iter) ?: + bch2_trans_update(trans, &iter, new, + BTREE_UPDATE_internal_snapshot_node| + BTREE_TRIGGER_norun); + bch2_trans_iter_exit(trans, &iter); + if (ret) + goto err; + + if (level) + bch2_btree_node_update_key_early(trans, btree, level - 1, k, new); + } else { + struct jset_entry *e = bch2_trans_jset_entry_alloc(trans, + jset_u64s(new->k.u64s)); + ret = PTR_ERR_OR_ZERO(e); + if (ret) + goto err; + + journal_entry_set(e, + BCH_JSET_ENTRY_btree_root, + btree, level - 1, + new, new->k.u64s); - if (level) - bch2_btree_node_update_key_early(trans, btree, level - 1, k, new); + /* + * no locking, we're single threaded and not rw yet, see + * the big assertino above that we repeat here: + */ + BUG_ON(test_bit(BCH_FS_rw, &c->flags)); + + struct btree *b = bch2_btree_id_root(c, btree)->b; + bkey_copy(&b->key, new); + } } err: printbuf_exit(&buf); @@ -406,7 +435,15 @@ static int bucket_ref_update_err(struct btree_trans *trans, struct printbuf *buf if (insert) { bch2_trans_updates_to_text(buf, trans); __bch2_inconsistent_error(c, buf); - ret = -BCH_ERR_bucket_ref_update; + /* + * If we're in recovery, run_explicit_recovery_pass might give + * us an error code for rewinding recovery + */ + if (!ret) + ret = bch_err_throw(c, bucket_ref_update); + } else { + /* Always ignore overwrite errors, so that deletion works */ + ret = 0; } if (print || insert) @@ -595,7 +632,7 @@ static int bch2_trigger_pointer(struct btree_trans *trans, struct bch_dev *ca = bch2_dev_tryget(c, p.ptr.dev); if (unlikely(!ca)) { if (insert && p.ptr.dev != BCH_SB_MEMBER_INVALID) - ret = -BCH_ERR_trigger_pointer; + ret = bch_err_throw(c, trigger_pointer); goto err; } @@ -603,7 +640,7 @@ static int bch2_trigger_pointer(struct btree_trans *trans, if (!bucket_valid(ca, bucket.offset)) { if (insert) { bch2_dev_bucket_missing(ca, bucket.offset); - ret = -BCH_ERR_trigger_pointer; + ret = bch_err_throw(c, trigger_pointer); } goto err; } @@ -625,7 +662,7 @@ static int bch2_trigger_pointer(struct btree_trans *trans, if (bch2_fs_inconsistent_on(!g, c, "reference to invalid bucket on device %u\n %s", p.ptr.dev, (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { - ret = -BCH_ERR_trigger_pointer; + ret = bch_err_throw(c, trigger_pointer); goto err; } @@ -651,6 +688,8 @@ static int bch2_trigger_stripe_ptr(struct btree_trans *trans, s64 sectors, enum btree_iter_update_trigger_flags flags) { + struct bch_fs *c = trans->c; + if (flags & BTREE_TRIGGER_transactional) { struct btree_iter iter; struct bkey_i_stripe *s = bch2_bkey_get_mut_typed(trans, &iter, @@ -668,7 +707,7 @@ static int bch2_trigger_stripe_ptr(struct btree_trans *trans, bch2_trans_inconsistent(trans, "stripe pointer doesn't match stripe %llu", (u64) p.ec.idx); - ret = -BCH_ERR_trigger_stripe_pointer; + ret = bch_err_throw(c, trigger_stripe_pointer); goto err; } @@ -688,13 +727,11 @@ err: } if (flags & BTREE_TRIGGER_gc) { - struct bch_fs *c = trans->c; - struct gc_stripe *m = genradix_ptr_alloc(&c->gc_stripes, p.ec.idx, GFP_KERNEL); if (!m) { bch_err(c, "error allocating memory for gc_stripes, idx %llu", (u64) p.ec.idx); - return -BCH_ERR_ENOMEM_mark_stripe_ptr; + return bch_err_throw(c, ENOMEM_mark_stripe_ptr); } gc_stripe_lock(m); @@ -709,7 +746,7 @@ err: __bch2_inconsistent_error(c, &buf); bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); - return -BCH_ERR_trigger_stripe_pointer; + return bch_err_throw(c, trigger_stripe_pointer); } m->block_sectors[p.ec.block] += sectors; @@ -732,8 +769,7 @@ err: static int __trigger_extent(struct btree_trans *trans, enum btree_id btree_id, unsigned level, struct bkey_s_c k, - enum btree_iter_update_trigger_flags flags, - s64 *replicas_sectors) + enum btree_iter_update_trigger_flags flags) { bool gc = flags & BTREE_TRIGGER_gc; struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); @@ -744,6 +780,8 @@ static int __trigger_extent(struct btree_trans *trans, : BCH_DATA_user; int ret = 0; + s64 replicas_sectors = 0; + struct disk_accounting_pos acc_replicas_key; memset(&acc_replicas_key, 0, sizeof(acc_replicas_key)); acc_replicas_key.type = BCH_DISK_ACCOUNTING_replicas; @@ -770,7 +808,7 @@ static int __trigger_extent(struct btree_trans *trans, if (ret) return ret; } else if (!p.has_ec) { - *replicas_sectors += disk_sectors; + replicas_sectors += disk_sectors; replicas_entry_add_dev(&acc_replicas_key.replicas, p.ptr.dev); } else { ret = bch2_trigger_stripe_ptr(trans, k, p, data_type, disk_sectors, flags); @@ -808,13 +846,13 @@ static int __trigger_extent(struct btree_trans *trans, } if (acc_replicas_key.replicas.nr_devs) { - ret = bch2_disk_accounting_mod(trans, &acc_replicas_key, replicas_sectors, 1, gc); + ret = bch2_disk_accounting_mod(trans, &acc_replicas_key, &replicas_sectors, 1, gc); if (ret) return ret; } if (acc_replicas_key.replicas.nr_devs && !level && k.k->p.snapshot) { - ret = bch2_disk_accounting_mod2_nr(trans, gc, replicas_sectors, 1, snapshot, k.k->p.snapshot); + ret = bch2_disk_accounting_mod2_nr(trans, gc, &replicas_sectors, 1, snapshot, k.k->p.snapshot); if (ret) return ret; } @@ -830,7 +868,7 @@ static int __trigger_extent(struct btree_trans *trans, } if (level) { - ret = bch2_disk_accounting_mod2_nr(trans, gc, replicas_sectors, 1, btree, btree_id); + ret = bch2_disk_accounting_mod2_nr(trans, gc, &replicas_sectors, 1, btree, btree_id); if (ret) return ret; } else { @@ -839,7 +877,7 @@ static int __trigger_extent(struct btree_trans *trans, s64 v[3] = { insert ? 1 : -1, insert ? k.k->size : -((s64) k.k->size), - *replicas_sectors, + replicas_sectors, }; ret = bch2_disk_accounting_mod2(trans, gc, v, inum, k.k->p.inode); if (ret) @@ -871,20 +909,16 @@ int bch2_trigger_extent(struct btree_trans *trans, return 0; if (flags & (BTREE_TRIGGER_transactional|BTREE_TRIGGER_gc)) { - s64 old_replicas_sectors = 0, new_replicas_sectors = 0; - if (old.k->type) { int ret = __trigger_extent(trans, btree, level, old, - flags & ~BTREE_TRIGGER_insert, - &old_replicas_sectors); + flags & ~BTREE_TRIGGER_insert); if (ret) return ret; } if (new.k->type) { int ret = __trigger_extent(trans, btree, level, new.s_c, - flags & ~BTREE_TRIGGER_overwrite, - &new_replicas_sectors); + flags & ~BTREE_TRIGGER_overwrite); if (ret) return ret; } @@ -971,15 +1005,16 @@ static int __bch2_trans_mark_metadata_bucket(struct btree_trans *trans, bch2_data_type_str(type), bch2_data_type_str(type)); - bool print = bch2_count_fsck_err(c, bucket_metadata_type_mismatch, &buf); + bch2_count_fsck_err(c, bucket_metadata_type_mismatch, &buf); - bch2_run_explicit_recovery_pass(c, &buf, + ret = bch2_run_explicit_recovery_pass(c, &buf, BCH_RECOVERY_PASS_check_allocations, 0); - if (print) - bch2_print_str(c, KERN_ERR, buf.buf); + /* Always print, this is always fatal */ + bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); - ret = -BCH_ERR_metadata_bucket_inconsistency; + if (!ret) + ret = bch_err_throw(c, metadata_bucket_inconsistency); goto err; } @@ -1032,7 +1067,7 @@ static int bch2_mark_metadata_bucket(struct btree_trans *trans, struct bch_dev * err_unlock: bucket_unlock(g); err: - return -BCH_ERR_metadata_bucket_inconsistency; + return bch_err_throw(c, metadata_bucket_inconsistency); } int bch2_trans_mark_metadata_bucket(struct btree_trans *trans, @@ -1247,7 +1282,7 @@ recalculate: ret = 0; } else { atomic64_set(&c->sectors_available, sectors_available); - ret = -BCH_ERR_ENOSPC_disk_reservation; + ret = bch_err_throw(c, ENOSPC_disk_reservation); } mutex_unlock(&c->sectors_available_lock); @@ -1276,7 +1311,7 @@ int bch2_buckets_nouse_alloc(struct bch_fs *c) GFP_KERNEL|__GFP_ZERO); if (!ca->buckets_nouse) { bch2_dev_put(ca); - return -BCH_ERR_ENOMEM_buckets_nouse; + return bch_err_throw(c, ENOMEM_buckets_nouse); } } @@ -1301,12 +1336,12 @@ int bch2_dev_buckets_resize(struct bch_fs *c, struct bch_dev *ca, u64 nbuckets) lockdep_assert_held(&c->state_lock); if (resize && ca->buckets_nouse) - return -BCH_ERR_no_resize_with_buckets_nouse; + return bch_err_throw(c, no_resize_with_buckets_nouse); bucket_gens = bch2_kvmalloc(struct_size(bucket_gens, b, nbuckets), GFP_KERNEL|__GFP_ZERO); if (!bucket_gens) { - ret = -BCH_ERR_ENOMEM_bucket_gens; + ret = bch_err_throw(c, ENOMEM_bucket_gens); goto err; } @@ -1325,9 +1360,9 @@ int bch2_dev_buckets_resize(struct bch_fs *c, struct bch_dev *ca, u64 nbuckets) sizeof(bucket_gens->b[0]) * copy); } - ret = bch2_bucket_bitmap_resize(&ca->bucket_backpointer_mismatch, + ret = bch2_bucket_bitmap_resize(ca, &ca->bucket_backpointer_mismatch, ca->mi.nbuckets, nbuckets) ?: - bch2_bucket_bitmap_resize(&ca->bucket_backpointer_empty, + bch2_bucket_bitmap_resize(ca, &ca->bucket_backpointer_empty, ca->mi.nbuckets, nbuckets); rcu_assign_pointer(ca->bucket_gens, bucket_gens); @@ -1354,7 +1389,7 @@ int bch2_dev_buckets_alloc(struct bch_fs *c, struct bch_dev *ca) { ca->usage = alloc_percpu(struct bch_dev_usage_full); if (!ca->usage) - return -BCH_ERR_ENOMEM_usage_init; + return bch_err_throw(c, ENOMEM_usage_init); return bch2_dev_buckets_resize(c, ca, ca->mi.nbuckets); } diff --git a/fs/bcachefs/buckets.h b/fs/bcachefs/buckets.h index af1532de4a37..49a3807a5eab 100644 --- a/fs/bcachefs/buckets.h +++ b/fs/bcachefs/buckets.h @@ -84,10 +84,8 @@ static inline int bucket_gen_get_rcu(struct bch_dev *ca, size_t b) static inline int bucket_gen_get(struct bch_dev *ca, size_t b) { - rcu_read_lock(); - int ret = bucket_gen_get_rcu(ca, b); - rcu_read_unlock(); - return ret; + guard(rcu)(); + return bucket_gen_get_rcu(ca, b); } static inline size_t PTR_BUCKET_NR(const struct bch_dev *ca, @@ -156,10 +154,8 @@ static inline int dev_ptr_stale_rcu(struct bch_dev *ca, const struct bch_extent_ */ static inline int dev_ptr_stale(struct bch_dev *ca, const struct bch_extent_ptr *ptr) { - rcu_read_lock(); - int ret = dev_ptr_stale_rcu(ca, ptr); - rcu_read_unlock(); - return ret; + guard(rcu)(); + return dev_ptr_stale_rcu(ca, ptr); } /* Device usage: */ diff --git a/fs/bcachefs/buckets_waiting_for_journal.c b/fs/bcachefs/buckets_waiting_for_journal.c index c8a488e6b7b8..832eff93acb6 100644 --- a/fs/bcachefs/buckets_waiting_for_journal.c +++ b/fs/bcachefs/buckets_waiting_for_journal.c @@ -108,7 +108,8 @@ int bch2_set_bucket_needs_journal_commit(struct buckets_waiting_for_journal *b, realloc: n = kvmalloc(sizeof(*n) + (sizeof(n->d[0]) << new_bits), GFP_KERNEL); if (!n) { - ret = -BCH_ERR_ENOMEM_buckets_waiting_for_journal_set; + struct bch_fs *c = container_of(b, struct bch_fs, buckets_waiting_for_journal); + ret = bch_err_throw(c, ENOMEM_buckets_waiting_for_journal_set); goto out; } diff --git a/fs/bcachefs/chardev.c b/fs/bcachefs/chardev.c index 4066946b26bc..5ea89aa2b0c4 100644 --- a/fs/bcachefs/chardev.c +++ b/fs/bcachefs/chardev.c @@ -319,6 +319,7 @@ static int bch2_data_thread(void *arg) ctx->stats.ret = BCH_IOCTL_DATA_EVENT_RET_done; ctx->stats.data_type = (int) DATA_PROGRESS_DATA_TYPE_done; } + enumerated_ref_put(&ctx->c->writes, BCH_WRITE_REF_ioctl_data); return 0; } @@ -378,15 +379,24 @@ static long bch2_ioctl_data(struct bch_fs *c, struct bch_data_ctx *ctx; int ret; - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; + if (!enumerated_ref_tryget(&c->writes, BCH_WRITE_REF_ioctl_data)) + return -EROFS; - if (arg.op >= BCH_DATA_OP_NR || arg.flags) - return -EINVAL; + if (!capable(CAP_SYS_ADMIN)) { + ret = -EPERM; + goto put_ref; + } + + if (arg.op >= BCH_DATA_OP_NR || arg.flags) { + ret = -EINVAL; + goto put_ref; + } ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); - if (!ctx) - return -ENOMEM; + if (!ctx) { + ret = -ENOMEM; + goto put_ref; + } ctx->c = c; ctx->arg = arg; @@ -395,11 +405,16 @@ static long bch2_ioctl_data(struct bch_fs *c, &bcachefs_data_ops, bch2_data_thread); if (ret < 0) - kfree(ctx); + goto cleanup; + return ret; +cleanup: + kfree(ctx); +put_ref: + enumerated_ref_put(&c->writes, BCH_WRITE_REF_ioctl_data); return ret; } -static long bch2_ioctl_fs_usage(struct bch_fs *c, +static noinline_for_stack long bch2_ioctl_fs_usage(struct bch_fs *c, struct bch_ioctl_fs_usage __user *user_arg) { struct bch_ioctl_fs_usage arg = {}; @@ -469,7 +484,7 @@ err: } /* obsolete, didn't allow for new data types: */ -static long bch2_ioctl_dev_usage(struct bch_fs *c, +static noinline_for_stack long bch2_ioctl_dev_usage(struct bch_fs *c, struct bch_ioctl_dev_usage __user *user_arg) { struct bch_ioctl_dev_usage arg; @@ -613,15 +628,12 @@ static long bch2_ioctl_disk_get_idx(struct bch_fs *c, if (!dev) return -EINVAL; - rcu_read_lock(); + guard(rcu)(); for_each_online_member_rcu(c, ca) - if (ca->dev == dev) { - rcu_read_unlock(); + if (ca->dev == dev) return ca->dev_idx; - } - rcu_read_unlock(); - return -BCH_ERR_ENOENT_dev_idx_not_found; + return bch_err_throw(c, ENOENT_dev_idx_not_found); } static long bch2_ioctl_disk_resize(struct bch_fs *c, diff --git a/fs/bcachefs/checksum.c b/fs/bcachefs/checksum.c index d3e2e4f776c6..a6795e73f0b9 100644 --- a/fs/bcachefs/checksum.c +++ b/fs/bcachefs/checksum.c @@ -173,7 +173,7 @@ int bch2_encrypt(struct bch_fs *c, unsigned type, if (bch2_fs_inconsistent_on(!c->chacha20_key_set, c, "attempting to encrypt without encryption key")) - return -BCH_ERR_no_encryption_key; + return bch_err_throw(c, no_encryption_key); bch2_chacha20(&c->chacha20_key, nonce, data, len); return 0; @@ -262,7 +262,7 @@ int __bch2_encrypt_bio(struct bch_fs *c, unsigned type, if (bch2_fs_inconsistent_on(!c->chacha20_key_set, c, "attempting to encrypt without encryption key")) - return -BCH_ERR_no_encryption_key; + return bch_err_throw(c, no_encryption_key); bch2_chacha20_init(&chacha_state, &c->chacha20_key, nonce); @@ -375,7 +375,7 @@ int bch2_rechecksum_bio(struct bch_fs *c, struct bio *bio, prt_str(&buf, ")"); WARN_RATELIMIT(1, "%s", buf.buf); printbuf_exit(&buf); - return -BCH_ERR_recompute_checksum; + return bch_err_throw(c, recompute_checksum); } for (i = splits; i < splits + ARRAY_SIZE(splits); i++) { @@ -659,7 +659,7 @@ int bch2_enable_encryption(struct bch_fs *c, bool keyed) crypt = bch2_sb_field_resize(&c->disk_sb, crypt, sizeof(*crypt) / sizeof(u64)); if (!crypt) { - ret = -BCH_ERR_ENOSPC_sb_crypt; + ret = bch_err_throw(c, ENOSPC_sb_crypt); goto err; } diff --git a/fs/bcachefs/clock.c b/fs/bcachefs/clock.c index f57f9f4774e6..8e9264b5a84e 100644 --- a/fs/bcachefs/clock.c +++ b/fs/bcachefs/clock.c @@ -53,7 +53,6 @@ void bch2_io_timer_del(struct io_clock *clock, struct io_timer *timer) struct io_clock_wait { struct io_timer io_timer; - struct timer_list cpu_timer; struct task_struct *task; int expired; }; @@ -67,15 +66,6 @@ static void io_clock_wait_fn(struct io_timer *timer) wake_up_process(wait->task); } -static void io_clock_cpu_timeout(struct timer_list *timer) -{ - struct io_clock_wait *wait = container_of(timer, - struct io_clock_wait, cpu_timer); - - wait->expired = 1; - wake_up_process(wait->task); -} - void bch2_io_clock_schedule_timeout(struct io_clock *clock, u64 until) { struct io_clock_wait wait = { @@ -90,8 +80,8 @@ void bch2_io_clock_schedule_timeout(struct io_clock *clock, u64 until) bch2_io_timer_del(clock, &wait.io_timer); } -void bch2_kthread_io_clock_wait(struct io_clock *clock, - u64 io_until, unsigned long cpu_timeout) +unsigned long bch2_kthread_io_clock_wait_once(struct io_clock *clock, + u64 io_until, unsigned long cpu_timeout) { bool kthread = (current->flags & PF_KTHREAD) != 0; struct io_clock_wait wait = { @@ -103,27 +93,26 @@ void bch2_kthread_io_clock_wait(struct io_clock *clock, bch2_io_timer_add(clock, &wait.io_timer); - timer_setup_on_stack(&wait.cpu_timer, io_clock_cpu_timeout, 0); - - if (cpu_timeout != MAX_SCHEDULE_TIMEOUT) - mod_timer(&wait.cpu_timer, cpu_timeout + jiffies); - - do { - set_current_state(TASK_INTERRUPTIBLE); - if (kthread && kthread_should_stop()) - break; - - if (wait.expired) - break; - - schedule(); + set_current_state(TASK_INTERRUPTIBLE); + if (!(kthread && kthread_should_stop())) { + cpu_timeout = schedule_timeout(cpu_timeout); try_to_freeze(); - } while (0); + } __set_current_state(TASK_RUNNING); - timer_delete_sync(&wait.cpu_timer); - timer_destroy_on_stack(&wait.cpu_timer); bch2_io_timer_del(clock, &wait.io_timer); + return cpu_timeout; +} + +void bch2_kthread_io_clock_wait(struct io_clock *clock, + u64 io_until, unsigned long cpu_timeout) +{ + bool kthread = (current->flags & PF_KTHREAD) != 0; + + while (!(kthread && kthread_should_stop()) && + cpu_timeout && + atomic64_read(&clock->now) < io_until) + cpu_timeout = bch2_kthread_io_clock_wait_once(clock, io_until, cpu_timeout); } static struct io_timer *get_expired_timer(struct io_clock *clock, u64 now) diff --git a/fs/bcachefs/clock.h b/fs/bcachefs/clock.h index 82c79c8baf92..8769be2aa21e 100644 --- a/fs/bcachefs/clock.h +++ b/fs/bcachefs/clock.h @@ -4,6 +4,7 @@ void bch2_io_timer_add(struct io_clock *, struct io_timer *); void bch2_io_timer_del(struct io_clock *, struct io_timer *); +unsigned long bch2_kthread_io_clock_wait_once(struct io_clock *, u64, unsigned long); void bch2_kthread_io_clock_wait(struct io_clock *, u64, unsigned long); void __bch2_increment_clock(struct io_clock *, u64); diff --git a/fs/bcachefs/compress.c b/fs/bcachefs/compress.c index 1bca61d17092..b37b1f325f0a 100644 --- a/fs/bcachefs/compress.c +++ b/fs/bcachefs/compress.c @@ -187,7 +187,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src, __bch2_compression_types[crc.compression_type])) ret = bch2_check_set_has_compressed_data(c, opt); else - ret = -BCH_ERR_compression_workspace_not_initialized; + ret = bch_err_throw(c, compression_workspace_not_initialized); if (ret) goto err; } @@ -200,7 +200,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src, ret2 = LZ4_decompress_safe_partial(src_data.b, dst_data, src_len, dst_len, dst_len); if (ret2 != dst_len) - ret = -BCH_ERR_decompress_lz4; + ret = bch_err_throw(c, decompress_lz4); break; case BCH_COMPRESSION_TYPE_gzip: { z_stream strm = { @@ -219,7 +219,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src, mempool_free(workspace, workspace_pool); if (ret2 != Z_STREAM_END) - ret = -BCH_ERR_decompress_gzip; + ret = bch_err_throw(c, decompress_gzip); break; } case BCH_COMPRESSION_TYPE_zstd: { @@ -227,7 +227,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src, size_t real_src_len = le32_to_cpup(src_data.b); if (real_src_len > src_len - 4) { - ret = -BCH_ERR_decompress_zstd_src_len_bad; + ret = bch_err_throw(c, decompress_zstd_src_len_bad); goto err; } @@ -241,7 +241,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src, mempool_free(workspace, workspace_pool); if (ret2 != dst_len) - ret = -BCH_ERR_decompress_zstd; + ret = bch_err_throw(c, decompress_zstd); break; } default: @@ -270,7 +270,7 @@ int bch2_bio_uncompress_inplace(struct bch_write_op *op, bch2_write_op_error(op, op->pos.offset, "extent too big to decompress (%u > %u)", crc->uncompressed_size << 9, c->opts.encoded_extent_max); - return -BCH_ERR_decompress_exceeded_max_encoded_extent; + return bch_err_throw(c, decompress_exceeded_max_encoded_extent); } data = __bounce_alloc(c, dst_len, WRITE); @@ -314,7 +314,7 @@ int bch2_bio_uncompress(struct bch_fs *c, struct bio *src, if (crc.uncompressed_size << 9 > c->opts.encoded_extent_max || crc.compressed_size << 9 > c->opts.encoded_extent_max) - return -BCH_ERR_decompress_exceeded_max_encoded_extent; + return bch_err_throw(c, decompress_exceeded_max_encoded_extent); dst_data = dst_len == dst_iter.bi_size ? __bio_map_or_bounce(c, dst, dst_iter, WRITE) @@ -656,12 +656,12 @@ static int __bch2_fs_compress_init(struct bch_fs *c, u64 features) if (!mempool_initialized(&c->compression_bounce[READ]) && mempool_init_kvmalloc_pool(&c->compression_bounce[READ], 1, c->opts.encoded_extent_max)) - return -BCH_ERR_ENOMEM_compression_bounce_read_init; + return bch_err_throw(c, ENOMEM_compression_bounce_read_init); if (!mempool_initialized(&c->compression_bounce[WRITE]) && mempool_init_kvmalloc_pool(&c->compression_bounce[WRITE], 1, c->opts.encoded_extent_max)) - return -BCH_ERR_ENOMEM_compression_bounce_write_init; + return bch_err_throw(c, ENOMEM_compression_bounce_write_init); for (i = compression_types; i < compression_types + ARRAY_SIZE(compression_types); @@ -675,7 +675,7 @@ static int __bch2_fs_compress_init(struct bch_fs *c, u64 features) if (mempool_init_kvmalloc_pool( &c->compress_workspace[i->type], 1, i->compress_workspace)) - return -BCH_ERR_ENOMEM_compression_workspace_init; + return bch_err_throw(c, ENOMEM_compression_workspace_init); } return 0; diff --git a/fs/bcachefs/darray.h b/fs/bcachefs/darray.h index 50ec3decfe8c..4080ee99aadd 100644 --- a/fs/bcachefs/darray.h +++ b/fs/bcachefs/darray.h @@ -8,6 +8,7 @@ * Inspired by CCAN's darray */ +#include <linux/cleanup.h> #include <linux/slab.h> #define DARRAY_PREALLOCATED(_type, _nr) \ @@ -87,7 +88,23 @@ int __bch2_darray_resize_noprof(darray_char *, size_t, size_t, gfp_t); #define darray_remove_item(_d, _pos) \ array_remove_item((_d)->data, (_d)->nr, (_pos) - (_d)->data) -#define __darray_for_each(_d, _i) \ +#define darray_find_p(_d, _i, cond) \ +({ \ + typeof((_d).data) _ret = NULL; \ + \ + darray_for_each(_d, _i) \ + if (cond) { \ + _ret = _i; \ + break; \ + } \ + _ret; \ +}) + +#define darray_find(_d, _item) darray_find_p(_d, _i, *_i == _item) + +/* Iteration: */ + +#define __darray_for_each(_d, _i) \ for ((_i) = (_d).data; _i < (_d).data + (_d).nr; _i++) #define darray_for_each(_d, _i) \ @@ -96,6 +113,8 @@ int __bch2_darray_resize_noprof(darray_char *, size_t, size_t, gfp_t); #define darray_for_each_reverse(_d, _i) \ for (typeof(&(_d).data[0]) _i = (_d).data + (_d).nr - 1; _i >= (_d).data && (_d).nr; --_i) +/* Init/exit */ + #define darray_init(_d) \ do { \ (_d)->nr = 0; \ @@ -111,4 +130,29 @@ do { \ darray_init(_d); \ } while (0) +#define DEFINE_DARRAY_CLASS(_type) \ +DEFINE_CLASS(_type, _type, darray_exit(&(_T)), (_type) {}, void) + +#define DEFINE_DARRAY(_type) \ +typedef DARRAY(_type) darray_##_type; \ +DEFINE_DARRAY_CLASS(darray_##_type) + +#define DEFINE_DARRAY_NAMED(_name, _type) \ +typedef DARRAY(_type) _name; \ +DEFINE_DARRAY_CLASS(_name) + +DEFINE_DARRAY_CLASS(darray_char); +DEFINE_DARRAY_CLASS(darray_str) +DEFINE_DARRAY_CLASS(darray_const_str) + +DEFINE_DARRAY_CLASS(darray_u8) +DEFINE_DARRAY_CLASS(darray_u16) +DEFINE_DARRAY_CLASS(darray_u32) +DEFINE_DARRAY_CLASS(darray_u64) + +DEFINE_DARRAY_CLASS(darray_s8) +DEFINE_DARRAY_CLASS(darray_s16) +DEFINE_DARRAY_CLASS(darray_s32) +DEFINE_DARRAY_CLASS(darray_s64) + #endif /* _BCACHEFS_DARRAY_H */ diff --git a/fs/bcachefs/data_update.c b/fs/bcachefs/data_update.c index c34e5b88ba9d..e848e210a9bf 100644 --- a/fs/bcachefs/data_update.c +++ b/fs/bcachefs/data_update.c @@ -66,37 +66,46 @@ static void bkey_nocow_unlock(struct bch_fs *c, struct bkey_s_c k) } } -static bool bkey_nocow_lock(struct bch_fs *c, struct moving_context *ctxt, struct bkey_s_c k) +static noinline_for_stack +bool __bkey_nocow_lock(struct bch_fs *c, struct moving_context *ctxt, struct bkey_ptrs_c ptrs, + const struct bch_extent_ptr *start) { - struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); + if (!ctxt) { + bkey_for_each_ptr(ptrs, ptr) { + if (ptr == start) + break; + + struct bch_dev *ca = bch2_dev_have_ref(c, ptr->dev); + struct bpos bucket = PTR_BUCKET_POS(ca, ptr); + bch2_bucket_nocow_unlock(&c->nocow_locks, bucket, 0); + } + return false; + } - bkey_for_each_ptr(ptrs, ptr) { + __bkey_for_each_ptr(start, ptrs.end, ptr) { struct bch_dev *ca = bch2_dev_have_ref(c, ptr->dev); struct bpos bucket = PTR_BUCKET_POS(ca, ptr); - if (ctxt) { - bool locked; - - move_ctxt_wait_event(ctxt, - (locked = bch2_bucket_nocow_trylock(&c->nocow_locks, bucket, 0)) || - list_empty(&ctxt->ios)); + bool locked; + move_ctxt_wait_event(ctxt, + (locked = bch2_bucket_nocow_trylock(&c->nocow_locks, bucket, 0)) || + list_empty(&ctxt->ios)); + if (!locked) + bch2_bucket_nocow_lock(&c->nocow_locks, bucket, 0); + } + return true; +} - if (!locked) - bch2_bucket_nocow_lock(&c->nocow_locks, bucket, 0); - } else { - if (!bch2_bucket_nocow_trylock(&c->nocow_locks, bucket, 0)) { - bkey_for_each_ptr(ptrs, ptr2) { - if (ptr2 == ptr) - break; +static bool bkey_nocow_lock(struct bch_fs *c, struct moving_context *ctxt, struct bkey_ptrs_c ptrs) +{ + bkey_for_each_ptr(ptrs, ptr) { + struct bch_dev *ca = bch2_dev_have_ref(c, ptr->dev); + struct bpos bucket = PTR_BUCKET_POS(ca, ptr); - ca = bch2_dev_have_ref(c, ptr2->dev); - bucket = PTR_BUCKET_POS(ca, ptr2); - bch2_bucket_nocow_unlock(&c->nocow_locks, bucket, 0); - } - return false; - } - } + if (!bch2_bucket_nocow_trylock(&c->nocow_locks, bucket, 0)) + return __bkey_nocow_lock(c, ctxt, ptrs, ptr); } + return true; } @@ -240,13 +249,14 @@ static int data_update_invalid_bkey(struct data_update *m, 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)); + prt_newline(&buf); bch2_fs_emergency_read_only2(c, &buf); bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); - return -BCH_ERR_invalid_bkey; + return bch_err_throw(c, invalid_bkey); } static int __bch2_data_update_index_update(struct btree_trans *trans, @@ -367,21 +377,21 @@ restart_drop_conflicting_replicas: bch2_bkey_durability(c, bkey_i_to_s_c(&new->k_i)); /* Now, drop excess replicas: */ - rcu_read_lock(); + scoped_guard(rcu) { restart_drop_extra_replicas: - bkey_for_each_ptr_decode(old.k, bch2_bkey_ptrs(bkey_i_to_s(insert)), p, entry) { - unsigned ptr_durability = bch2_extent_ptr_durability(c, &p); + bkey_for_each_ptr_decode(old.k, bch2_bkey_ptrs(bkey_i_to_s(insert)), p, entry) { + unsigned ptr_durability = bch2_extent_ptr_durability(c, &p); - if (!p.ptr.cached && - durability - ptr_durability >= m->op.opts.data_replicas) { - durability -= ptr_durability; + if (!p.ptr.cached && + durability - ptr_durability >= m->op.opts.data_replicas) { + durability -= ptr_durability; - bch2_extent_ptr_set_cached(c, &m->op.opts, - bkey_i_to_s(insert), &entry->ptr); - goto restart_drop_extra_replicas; + bch2_extent_ptr_set_cached(c, &m->op.opts, + bkey_i_to_s(insert), &entry->ptr); + goto restart_drop_extra_replicas; + } } } - rcu_read_unlock(); /* Finally, add the pointers we just wrote: */ extent_for_each_ptr_decode(extent_i_to_s(new), p, entry) @@ -523,8 +533,9 @@ void bch2_data_update_exit(struct data_update *update) bch2_bkey_buf_exit(&update->k, c); } -static int bch2_update_unwritten_extent(struct btree_trans *trans, - struct data_update *update) +static noinline_for_stack +int bch2_update_unwritten_extent(struct btree_trans *trans, + struct data_update *update) { struct bch_fs *c = update->op.c; struct bkey_i_extent *e; @@ -716,18 +727,10 @@ int bch2_extent_drop_ptrs(struct btree_trans *trans, bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc); } -int bch2_data_update_bios_init(struct data_update *m, struct bch_fs *c, - struct bch_io_opts *io_opts) +static int __bch2_data_update_bios_init(struct data_update *m, struct bch_fs *c, + struct bch_io_opts *io_opts, + unsigned buf_bytes) { - struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(bkey_i_to_s_c(m->k.k)); - const union bch_extent_entry *entry; - struct extent_ptr_decoded p; - - /* write path might have to decompress data: */ - unsigned buf_bytes = 0; - bkey_for_each_ptr_decode(&m->k.k->k, ptrs, p, entry) - buf_bytes = max_t(unsigned, buf_bytes, p.crc.uncompressed_size << 9); - unsigned nr_vecs = DIV_ROUND_UP(buf_bytes, PAGE_SIZE); m->bvecs = kmalloc_array(nr_vecs, sizeof*(m->bvecs), GFP_KERNEL); @@ -751,11 +754,26 @@ int bch2_data_update_bios_init(struct data_update *m, struct bch_fs *c, return 0; } +int bch2_data_update_bios_init(struct data_update *m, struct bch_fs *c, + struct bch_io_opts *io_opts) +{ + struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(bkey_i_to_s_c(m->k.k)); + const union bch_extent_entry *entry; + struct extent_ptr_decoded p; + + /* write path might have to decompress data: */ + unsigned buf_bytes = 0; + bkey_for_each_ptr_decode(&m->k.k->k, ptrs, p, entry) + buf_bytes = max_t(unsigned, buf_bytes, p.crc.uncompressed_size << 9); + + return __bch2_data_update_bios_init(m, c, io_opts, buf_bytes); +} + static int can_write_extent(struct bch_fs *c, struct data_update *m) { if ((m->op.flags & BCH_WRITE_alloc_nowait) && unlikely(c->open_buckets_nr_free <= bch2_open_buckets_reserved(m->op.watermark))) - return -BCH_ERR_data_update_done_would_block; + return bch_err_throw(c, data_update_done_would_block); unsigned target = m->op.flags & BCH_WRITE_only_specified_devs ? m->op.target @@ -765,7 +783,8 @@ static int can_write_extent(struct bch_fs *c, struct data_update *m) darray_for_each(m->op.devs_have, i) __clear_bit(*i, devs.d); - rcu_read_lock(); + guard(rcu)(); + unsigned nr_replicas = 0, i; for_each_set_bit(i, devs.d, BCH_SB_MEMBERS_MAX) { struct bch_dev *ca = bch2_dev_rcu_noerror(c, i); @@ -782,12 +801,11 @@ static int can_write_extent(struct bch_fs *c, struct data_update *m) if (nr_replicas >= m->op.nr_replicas) break; } - rcu_read_unlock(); if (!nr_replicas) - return -BCH_ERR_data_update_done_no_rw_devs; + return bch_err_throw(c, data_update_done_no_rw_devs); if (nr_replicas < m->op.nr_replicas) - return -BCH_ERR_insufficient_devices; + return bch_err_throw(c, insufficient_devices); return 0; } @@ -802,19 +820,21 @@ int bch2_data_update_init(struct btree_trans *trans, struct bkey_s_c k) { struct bch_fs *c = trans->c; - struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); - const union bch_extent_entry *entry; - struct extent_ptr_decoded p; - unsigned reserve_sectors = k.k->size * data_opts.extra_replicas; int ret = 0; - /* - * fs is corrupt we have a key for a snapshot node that doesn't exist, - * and we have to check for this because we go rw before repairing the - * snapshots table - just skip it, we can move it later. - */ - if (unlikely(k.k->p.snapshot && !bch2_snapshot_exists(c, k.k->p.snapshot))) - return -BCH_ERR_data_update_done_no_snapshot; + if (k.k->p.snapshot) { + ret = bch2_check_key_has_snapshot(trans, iter, k); + if (bch2_err_matches(ret, BCH_ERR_recovery_will_run)) { + /* Can't repair yet, waiting on other recovery passes */ + return bch_err_throw(c, data_update_done_no_snapshot); + } + if (ret < 0) + return ret; + if (ret) /* key was deleted */ + return bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc) ?: + bch_err_throw(c, data_update_done_no_snapshot); + ret = 0; + } bch2_bkey_buf_init(&m->k); bch2_bkey_buf_reassemble(&m->k, c, k); @@ -842,10 +862,17 @@ int bch2_data_update_init(struct btree_trans *trans, unsigned durability_have = 0, durability_removing = 0; + struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(bkey_i_to_s_c(m->k.k)); + const union bch_extent_entry *entry; + struct extent_ptr_decoded p; + unsigned reserve_sectors = k.k->size * data_opts.extra_replicas; + unsigned buf_bytes = 0; + bool unwritten = false; + unsigned ptr_bit = 1; bkey_for_each_ptr_decode(k.k, ptrs, p, entry) { if (!p.ptr.cached) { - rcu_read_lock(); + guard(rcu)(); if (ptr_bit & m->data_opts.rewrite_ptrs) { if (crc_is_compressed(p.crc)) reserve_sectors += k.k->size; @@ -856,7 +883,6 @@ int bch2_data_update_init(struct btree_trans *trans, bch2_dev_list_add_dev(&m->op.devs_have, p.ptr.dev); durability_have += bch2_extent_ptr_durability(c, &p); } - rcu_read_unlock(); } /* @@ -872,6 +898,9 @@ int bch2_data_update_init(struct btree_trans *trans, if (p.crc.compression_type == BCH_COMPRESSION_TYPE_incompressible) m->op.incompressible = true; + buf_bytes = max_t(unsigned, buf_bytes, p.crc.uncompressed_size << 9); + unwritten |= p.ptr.unwritten; + ptr_bit <<= 1; } @@ -910,7 +939,7 @@ int bch2_data_update_init(struct btree_trans *trans, if (iter) ret = bch2_extent_drop_ptrs(trans, iter, k, io_opts, &m->data_opts); if (!ret) - ret = -BCH_ERR_data_update_done_no_writes_needed; + ret = bch_err_throw(c, data_update_done_no_writes_needed); goto out_bkey_buf_exit; } @@ -941,23 +970,25 @@ int bch2_data_update_init(struct btree_trans *trans, } if (!bkey_get_dev_refs(c, k)) { - ret = -BCH_ERR_data_update_done_no_dev_refs; + ret = bch_err_throw(c, data_update_done_no_dev_refs); goto out_put_disk_res; } if (c->opts.nocow_enabled && - !bkey_nocow_lock(c, ctxt, k)) { - ret = -BCH_ERR_nocow_lock_blocked; + !bkey_nocow_lock(c, ctxt, ptrs)) { + ret = bch_err_throw(c, nocow_lock_blocked); goto out_put_dev_refs; } - if (bkey_extent_is_unwritten(k)) { + if (unwritten) { ret = bch2_update_unwritten_extent(trans, m) ?: - -BCH_ERR_data_update_done_unwritten; + bch_err_throw(c, data_update_done_unwritten); goto out_nocow_unlock; } - ret = bch2_data_update_bios_init(m, c, io_opts); + bch2_trans_unlock(trans); + + ret = __bch2_data_update_bios_init(m, c, io_opts, buf_bytes); if (ret) goto out_nocow_unlock; diff --git a/fs/bcachefs/debug.c b/fs/bcachefs/debug.c index 4fa70634c90e..07c2a0f73cc2 100644 --- a/fs/bcachefs/debug.c +++ b/fs/bcachefs/debug.c @@ -153,8 +153,6 @@ void __bch2_btree_verify(struct bch_fs *c, struct btree *b) c->verify_data = __bch2_btree_node_mem_alloc(c); if (!c->verify_data) goto out; - - list_del_init(&c->verify_data->list); } BUG_ON(b->nsets != 1); @@ -492,6 +490,8 @@ static void bch2_cached_btree_node_to_text(struct printbuf *out, struct bch_fs * prt_printf(out, "journal pin %px:\t%llu\n", &b->writes[1].journal, b->writes[1].journal.seq); + prt_printf(out, "ob:\t%u\n", b->ob.nr); + printbuf_indent_sub(out, 2); } @@ -508,27 +508,27 @@ static ssize_t bch2_cached_btree_nodes_read(struct file *file, char __user *buf, i->ret = 0; do { - struct bucket_table *tbl; - struct rhash_head *pos; - struct btree *b; - ret = bch2_debugfs_flush_buf(i); if (ret) return ret; - rcu_read_lock(); i->buf.atomic++; - tbl = rht_dereference_rcu(c->btree_cache.table.tbl, - &c->btree_cache.table); - if (i->iter < tbl->size) { - rht_for_each_entry_rcu(b, pos, tbl, i->iter, hash) - bch2_cached_btree_node_to_text(&i->buf, c, b); - i->iter++; - } else { - done = true; + scoped_guard(rcu) { + struct bucket_table *tbl = + rht_dereference_rcu(c->btree_cache.table.tbl, + &c->btree_cache.table); + if (i->iter < tbl->size) { + struct rhash_head *pos; + struct btree *b; + + rht_for_each_entry_rcu(b, pos, tbl, i->iter, hash) + bch2_cached_btree_node_to_text(&i->buf, c, b); + i->iter++; + } else { + done = true; + } } --i->buf.atomic; - rcu_read_unlock(); } while (!done); if (i->buf.allocation_failure) @@ -584,6 +584,8 @@ static ssize_t bch2_btree_transactions_read(struct file *file, char __user *buf, i->ubuf = buf; i->size = size; i->ret = 0; + + int srcu_idx = srcu_read_lock(&c->btree_trans_barrier); restart: seqmutex_lock(&c->btree_trans_lock); list_sort(&c->btree_trans_list, list_ptr_order_cmp); @@ -597,6 +599,11 @@ restart: if (!closure_get_not_zero(&trans->ref)) continue; + if (!trans->srcu_held) { + closure_put(&trans->ref); + continue; + } + u32 seq = seqmutex_unlock(&c->btree_trans_lock); bch2_btree_trans_to_text(&i->buf, trans); @@ -618,6 +625,8 @@ restart: } seqmutex_unlock(&c->btree_trans_lock); unlocked: + srcu_read_unlock(&c->btree_trans_barrier, srcu_idx); + if (i->buf.allocation_failure) ret = -ENOMEM; diff --git a/fs/bcachefs/dirent.c b/fs/bcachefs/dirent.c index d198001838f3..28875c5c86ad 100644 --- a/fs/bcachefs/dirent.c +++ b/fs/bcachefs/dirent.c @@ -13,12 +13,15 @@ #include <linux/dcache.h> +#ifdef CONFIG_UNICODE int bch2_casefold(struct btree_trans *trans, const struct bch_hash_info *info, const struct qstr *str, struct qstr *out_cf) { *out_cf = (struct qstr) QSTR_INIT(NULL, 0); -#ifdef CONFIG_UNICODE + if (!bch2_fs_casefold_enabled(trans->c)) + return -EOPNOTSUPP; + unsigned char *buf = bch2_trans_kmalloc(trans, BCH_NAME_MAX + 1); int ret = PTR_ERR_OR_ZERO(buf); if (ret) @@ -30,10 +33,8 @@ int bch2_casefold(struct btree_trans *trans, const struct bch_hash_info *info, *out_cf = (struct qstr) QSTR_INIT(buf, ret); return 0; -#else - return -EOPNOTSUPP; -#endif } +#endif static unsigned bch2_dirent_name_bytes(struct bkey_s_c_dirent d) { @@ -231,70 +232,66 @@ void bch2_dirent_to_text(struct printbuf *out, struct bch_fs *c, struct bkey_s_c prt_printf(out, " type %s", bch2_d_type_str(d.v->d_type)); } -static struct bkey_i_dirent *dirent_alloc_key(struct btree_trans *trans, - subvol_inum dir, - u8 type, - int name_len, int cf_name_len, - u64 dst) +int bch2_dirent_init_name(struct bch_fs *c, + struct bkey_i_dirent *dirent, + const struct bch_hash_info *hash_info, + const struct qstr *name, + const struct qstr *cf_name) { - struct bkey_i_dirent *dirent; - unsigned u64s = BKEY_U64s + dirent_val_u64s(name_len, cf_name_len); + EBUG_ON(hash_info->cf_encoding == NULL && cf_name); + int cf_len = 0; - BUG_ON(u64s > U8_MAX); - - dirent = bch2_trans_kmalloc(trans, u64s * sizeof(u64)); - if (IS_ERR(dirent)) - return dirent; + if (name->len > BCH_NAME_MAX) + return -ENAMETOOLONG; - bkey_dirent_init(&dirent->k_i); - dirent->k.u64s = u64s; + dirent->v.d_casefold = hash_info->cf_encoding != NULL; - if (type != DT_SUBVOL) { - dirent->v.d_inum = cpu_to_le64(dst); + if (!dirent->v.d_casefold) { + memcpy(&dirent->v.d_name[0], name->name, name->len); + memset(&dirent->v.d_name[name->len], 0, + bkey_val_bytes(&dirent->k) - + offsetof(struct bch_dirent, d_name) - + name->len); } else { - dirent->v.d_parent_subvol = cpu_to_le32(dir.subvol); - dirent->v.d_child_subvol = cpu_to_le32(dst); - } + if (!bch2_fs_casefold_enabled(c)) + return -EOPNOTSUPP; - dirent->v.d_type = type; - dirent->v.d_unused = 0; - dirent->v.d_casefold = cf_name_len ? 1 : 0; +#ifdef CONFIG_UNICODE + memcpy(&dirent->v.d_cf_name_block.d_names[0], name->name, name->len); - return dirent; -} + char *cf_out = &dirent->v.d_cf_name_block.d_names[name->len]; -static void dirent_init_regular_name(struct bkey_i_dirent *dirent, - const struct qstr *name) -{ - EBUG_ON(dirent->v.d_casefold); + if (cf_name) { + cf_len = cf_name->len; - memcpy(&dirent->v.d_name[0], name->name, name->len); - memset(&dirent->v.d_name[name->len], 0, - bkey_val_bytes(&dirent->k) - - offsetof(struct bch_dirent, d_name) - - name->len); -} + memcpy(cf_out, cf_name->name, cf_name->len); + } else { + cf_len = utf8_casefold(hash_info->cf_encoding, name, + cf_out, + bkey_val_end(bkey_i_to_s(&dirent->k_i)) - (void *) cf_out); + if (cf_len <= 0) + return cf_len; + } -static void dirent_init_casefolded_name(struct bkey_i_dirent *dirent, - const struct qstr *name, - const struct qstr *cf_name) -{ - EBUG_ON(!dirent->v.d_casefold); - EBUG_ON(!cf_name->len); - - dirent->v.d_cf_name_block.d_name_len = cpu_to_le16(name->len); - dirent->v.d_cf_name_block.d_cf_name_len = cpu_to_le16(cf_name->len); - memcpy(&dirent->v.d_cf_name_block.d_names[0], name->name, name->len); - memcpy(&dirent->v.d_cf_name_block.d_names[name->len], cf_name->name, cf_name->len); - memset(&dirent->v.d_cf_name_block.d_names[name->len + cf_name->len], 0, - bkey_val_bytes(&dirent->k) - - offsetof(struct bch_dirent, d_cf_name_block.d_names) - - name->len + cf_name->len); - - EBUG_ON(bch2_dirent_get_casefold_name(dirent_i_to_s_c(dirent)).len != cf_name->len); + memset(&dirent->v.d_cf_name_block.d_names[name->len + cf_len], 0, + bkey_val_bytes(&dirent->k) - + offsetof(struct bch_dirent, d_cf_name_block.d_names) - + name->len + cf_len); + + dirent->v.d_cf_name_block.d_name_len = cpu_to_le16(name->len); + dirent->v.d_cf_name_block.d_cf_name_len = cpu_to_le16(cf_len); + + EBUG_ON(bch2_dirent_get_casefold_name(dirent_i_to_s_c(dirent)).len != cf_len); +#endif + } + + unsigned u64s = dirent_val_u64s(name->len, cf_len); + BUG_ON(u64s > bkey_val_u64s(&dirent->k)); + set_bkey_val_u64s(&dirent->k, u64s); + return 0; } -static struct bkey_i_dirent *dirent_create_key(struct btree_trans *trans, +struct bkey_i_dirent *bch2_dirent_create_key(struct btree_trans *trans, const struct bch_hash_info *hash_info, subvol_inum dir, u8 type, @@ -302,31 +299,28 @@ static struct bkey_i_dirent *dirent_create_key(struct btree_trans *trans, const struct qstr *cf_name, u64 dst) { - struct bkey_i_dirent *dirent; - struct qstr _cf_name; - - if (name->len > BCH_NAME_MAX) - return ERR_PTR(-ENAMETOOLONG); + struct bkey_i_dirent *dirent = bch2_trans_kmalloc(trans, BKEY_U64s_MAX * sizeof(u64)); + if (IS_ERR(dirent)) + return dirent; - if (hash_info->cf_encoding && !cf_name) { - int ret = bch2_casefold(trans, hash_info, name, &_cf_name); - if (ret) - return ERR_PTR(ret); + bkey_dirent_init(&dirent->k_i); + dirent->k.u64s = BKEY_U64s_MAX; - cf_name = &_cf_name; + if (type != DT_SUBVOL) { + dirent->v.d_inum = cpu_to_le64(dst); + } else { + dirent->v.d_parent_subvol = cpu_to_le32(dir.subvol); + dirent->v.d_child_subvol = cpu_to_le32(dst); } - dirent = dirent_alloc_key(trans, dir, type, name->len, cf_name ? cf_name->len : 0, dst); - if (IS_ERR(dirent)) - return dirent; + dirent->v.d_type = type; + dirent->v.d_unused = 0; - if (cf_name) - dirent_init_casefolded_name(dirent, name, cf_name); - else - dirent_init_regular_name(dirent, name); + int ret = bch2_dirent_init_name(trans->c, dirent, hash_info, name, cf_name); + if (ret) + return ERR_PTR(ret); EBUG_ON(bch2_dirent_get_name(dirent_i_to_s_c(dirent)).len != name->len); - return dirent; } @@ -341,7 +335,7 @@ int bch2_dirent_create_snapshot(struct btree_trans *trans, struct bkey_i_dirent *dirent; int ret; - dirent = dirent_create_key(trans, hash_info, dir_inum, type, name, NULL, dst_inum); + dirent = bch2_dirent_create_key(trans, hash_info, dir_inum, type, name, NULL, dst_inum); ret = PTR_ERR_OR_ZERO(dirent); if (ret) return ret; @@ -365,7 +359,7 @@ int bch2_dirent_create(struct btree_trans *trans, subvol_inum dir, struct bkey_i_dirent *dirent; int ret; - dirent = dirent_create_key(trans, hash_info, dir, type, name, NULL, dst_inum); + dirent = bch2_dirent_create_key(trans, hash_info, dir, type, name, NULL, dst_inum); ret = PTR_ERR_OR_ZERO(dirent); if (ret) return ret; @@ -402,8 +396,8 @@ int bch2_dirent_read_target(struct btree_trans *trans, subvol_inum dir, } int bch2_dirent_rename(struct btree_trans *trans, - subvol_inum src_dir, struct bch_hash_info *src_hash, u64 *src_dir_i_size, - subvol_inum dst_dir, struct bch_hash_info *dst_hash, u64 *dst_dir_i_size, + subvol_inum src_dir, struct bch_hash_info *src_hash, + subvol_inum dst_dir, struct bch_hash_info *dst_hash, const struct qstr *src_name, subvol_inum *src_inum, u64 *src_offset, const struct qstr *dst_name, subvol_inum *dst_inum, u64 *dst_offset, enum bch_rename_mode mode) @@ -470,8 +464,8 @@ int bch2_dirent_rename(struct btree_trans *trans, *src_offset = dst_iter.pos.offset; /* Create new dst key: */ - new_dst = dirent_create_key(trans, dst_hash, dst_dir, 0, dst_name, - dst_hash->cf_encoding ? &dst_name_lookup : NULL, 0); + new_dst = bch2_dirent_create_key(trans, dst_hash, dst_dir, 0, dst_name, + dst_hash->cf_encoding ? &dst_name_lookup : NULL, 0); ret = PTR_ERR_OR_ZERO(new_dst); if (ret) goto out; @@ -481,8 +475,8 @@ int bch2_dirent_rename(struct btree_trans *trans, /* Create new src key: */ if (mode == BCH_RENAME_EXCHANGE) { - new_src = dirent_create_key(trans, src_hash, src_dir, 0, src_name, - src_hash->cf_encoding ? &src_name_lookup : NULL, 0); + new_src = bch2_dirent_create_key(trans, src_hash, src_dir, 0, src_name, + src_hash->cf_encoding ? &src_name_lookup : NULL, 0); ret = PTR_ERR_OR_ZERO(new_src); if (ret) goto out; @@ -542,14 +536,6 @@ int bch2_dirent_rename(struct btree_trans *trans, new_src->v.d_type == DT_SUBVOL) new_src->v.d_parent_subvol = cpu_to_le32(src_dir.subvol); - if (old_dst.k) - *dst_dir_i_size -= bkey_bytes(old_dst.k); - *src_dir_i_size -= bkey_bytes(old_src.k); - - if (mode == BCH_RENAME_EXCHANGE) - *src_dir_i_size += bkey_bytes(&new_src->k); - *dst_dir_i_size += bkey_bytes(&new_dst->k); - ret = bch2_trans_update(trans, &dst_iter, &new_dst->k_i, 0); if (ret) goto out; @@ -656,7 +642,7 @@ int bch2_empty_dir_snapshot(struct btree_trans *trans, u64 dir, u32 subvol, u32 struct bkey_s_c_dirent d = bkey_s_c_to_dirent(k); if (d.v->d_type == DT_SUBVOL && le32_to_cpu(d.v->d_parent_subvol) != subvol) continue; - ret = -BCH_ERR_ENOTEMPTY_dir_not_empty; + ret = bch_err_throw(trans->c, ENOTEMPTY_dir_not_empty); break; } bch2_trans_iter_exit(trans, &iter); @@ -692,7 +678,9 @@ static int bch2_dir_emit(struct dir_context *ctx, struct bkey_s_c_dirent d, subv return !ret; } -int bch2_readdir(struct bch_fs *c, subvol_inum inum, struct dir_context *ctx) +int bch2_readdir(struct bch_fs *c, subvol_inum inum, + struct bch_hash_info *hash_info, + struct dir_context *ctx) { struct bkey_buf sk; bch2_bkey_buf_init(&sk); @@ -710,7 +698,11 @@ int bch2_readdir(struct bch_fs *c, subvol_inum inum, struct dir_context *ctx) struct bkey_s_c_dirent dirent = bkey_i_to_s_c_dirent(sk.k); subvol_inum target; - int ret2 = bch2_dirent_read_target(trans, inum, dirent, &target); + + bool need_second_pass = false; + int ret2 = bch2_str_hash_check_key(trans, NULL, &bch2_dirent_hash_desc, + hash_info, &iter, k, &need_second_pass) ?: + bch2_dirent_read_target(trans, inum, dirent, &target); if (ret2 > 0) continue; @@ -740,7 +732,7 @@ static int lookup_first_inode(struct btree_trans *trans, u64 inode_nr, ret = bch2_inode_unpack(k, inode); goto found; } - ret = -BCH_ERR_ENOENT_inode; + ret = bch_err_throw(trans->c, ENOENT_inode); found: bch_err_msg(trans->c, ret, "fetching inode %llu", inode_nr); bch2_trans_iter_exit(trans, &iter); diff --git a/fs/bcachefs/dirent.h b/fs/bcachefs/dirent.h index d3e7ae669575..0417608c18d5 100644 --- a/fs/bcachefs/dirent.h +++ b/fs/bcachefs/dirent.h @@ -23,8 +23,16 @@ struct bch_fs; struct bch_hash_info; struct bch_inode_info; +#ifdef CONFIG_UNICODE int bch2_casefold(struct btree_trans *, const struct bch_hash_info *, const struct qstr *, struct qstr *); +#else +static inline int bch2_casefold(struct btree_trans *trans, const struct bch_hash_info *info, + const struct qstr *str, struct qstr *out_cf) +{ + return -EOPNOTSUPP; +} +#endif static inline int bch2_maybe_casefold(struct btree_trans *trans, const struct bch_hash_info *info, @@ -38,7 +46,7 @@ static inline int bch2_maybe_casefold(struct btree_trans *trans, } } -struct qstr bch2_dirent_get_name(struct bkey_s_c_dirent d); +struct qstr bch2_dirent_get_name(struct bkey_s_c_dirent); static inline unsigned dirent_val_u64s(unsigned len, unsigned cf_len) { @@ -59,6 +67,15 @@ static inline void dirent_copy_target(struct bkey_i_dirent *dst, dst->v.d_type = src.v->d_type; } +int bch2_dirent_init_name(struct bch_fs *, + struct bkey_i_dirent *, + const struct bch_hash_info *, + const struct qstr *, + const struct qstr *); +struct bkey_i_dirent *bch2_dirent_create_key(struct btree_trans *, + const struct bch_hash_info *, subvol_inum, u8, + const struct qstr *, const struct qstr *, u64); + int bch2_dirent_create_snapshot(struct btree_trans *, u32, u64, u32, const struct bch_hash_info *, u8, const struct qstr *, u64, u64 *, @@ -80,8 +97,8 @@ enum bch_rename_mode { }; int bch2_dirent_rename(struct btree_trans *, - subvol_inum, struct bch_hash_info *, u64 *, - subvol_inum, struct bch_hash_info *, u64 *, + subvol_inum, struct bch_hash_info *, + subvol_inum, struct bch_hash_info *, const struct qstr *, subvol_inum *, u64 *, const struct qstr *, subvol_inum *, u64 *, enum bch_rename_mode); @@ -95,7 +112,7 @@ u64 bch2_dirent_lookup(struct bch_fs *, subvol_inum, int bch2_empty_dir_snapshot(struct btree_trans *, u64, u32, u32); int bch2_empty_dir_trans(struct btree_trans *, subvol_inum); -int bch2_readdir(struct bch_fs *, subvol_inum, struct dir_context *); +int bch2_readdir(struct bch_fs *, subvol_inum, struct bch_hash_info *, struct dir_context *); int bch2_fsck_remove_dirent(struct btree_trans *, struct bpos); diff --git a/fs/bcachefs/disk_accounting.c b/fs/bcachefs/disk_accounting.c index b3840ff7c407..f7528cd69c73 100644 --- a/fs/bcachefs/disk_accounting.c +++ b/fs/bcachefs/disk_accounting.c @@ -390,7 +390,7 @@ static int __bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accoun err: free_percpu(n.v[1]); free_percpu(n.v[0]); - return -BCH_ERR_ENOMEM_disk_accounting; + return bch_err_throw(c, ENOMEM_disk_accounting); } int bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accounting a, @@ -401,7 +401,7 @@ int bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accounting a, if (mode != BCH_ACCOUNTING_read && accounting_to_replicas(&r.e, a.k->p) && !bch2_replicas_marked_locked(c, &r.e)) - return -BCH_ERR_btree_insert_need_mark_replicas; + return bch_err_throw(c, btree_insert_need_mark_replicas); percpu_up_read(&c->mark_lock); percpu_down_write(&c->mark_lock); @@ -419,7 +419,7 @@ int bch2_accounting_mem_insert_locked(struct bch_fs *c, struct bkey_s_c_accounti if (mode != BCH_ACCOUNTING_read && accounting_to_replicas(&r.e, a.k->p) && !bch2_replicas_marked_locked(c, &r.e)) - return -BCH_ERR_btree_insert_need_mark_replicas; + return bch_err_throw(c, btree_insert_need_mark_replicas); return __bch2_accounting_mem_insert(c, a); } @@ -559,7 +559,7 @@ int bch2_gc_accounting_start(struct bch_fs *c) sizeof(u64), GFP_KERNEL); if (!e->v[1]) { bch2_accounting_free_counters(acc, true); - ret = -BCH_ERR_ENOMEM_disk_accounting; + ret = bch_err_throw(c, ENOMEM_disk_accounting); break; } } @@ -618,7 +618,9 @@ int bch2_gc_accounting_done(struct bch_fs *c) for (unsigned j = 0; j < nr; j++) src_v[j] -= dst_v[j]; - if (fsck_err(trans, accounting_mismatch, "%s", buf.buf)) { + bch2_trans_unlock_long(trans); + + if (fsck_err(c, accounting_mismatch, "%s", buf.buf)) { percpu_up_write(&c->mark_lock); ret = commit_do(trans, NULL, NULL, 0, bch2_disk_accounting_mod(trans, &acc_k, src_v, nr, false)); @@ -737,7 +739,7 @@ invalid_device: bch2_disk_accounting_mod(trans, acc, v, nr, false)) ?: -BCH_ERR_remove_disk_accounting_entry; } else { - ret = -BCH_ERR_remove_disk_accounting_entry; + ret = bch_err_throw(c, remove_disk_accounting_entry); } goto fsck_err; } @@ -897,8 +899,8 @@ int bch2_accounting_read(struct bch_fs *c) case BCH_DISK_ACCOUNTING_replicas: fs_usage_data_type_to_base(usage, k.replicas.data_type, v[0]); break; - case BCH_DISK_ACCOUNTING_dev_data_type: - rcu_read_lock(); + case BCH_DISK_ACCOUNTING_dev_data_type: { + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, k.dev_data_type.dev); if (ca) { struct bch_dev_usage_type __percpu *d = &ca->usage->d[k.dev_data_type.data_type]; @@ -910,9 +912,9 @@ int bch2_accounting_read(struct bch_fs *c) k.dev_data_type.data_type == BCH_DATA_journal) usage->hidden += v[0] * ca->mi.bucket_size; } - rcu_read_unlock(); break; } + } } preempt_enable(); fsck_err: @@ -1006,19 +1008,18 @@ void bch2_verify_accounting_clean(struct bch_fs *c) case BCH_DISK_ACCOUNTING_replicas: fs_usage_data_type_to_base(&base, acc_k.replicas.data_type, a.v->d[0]); break; - case BCH_DISK_ACCOUNTING_dev_data_type: { - rcu_read_lock(); - struct bch_dev *ca = bch2_dev_rcu_noerror(c, acc_k.dev_data_type.dev); - if (!ca) { - rcu_read_unlock(); - continue; + case BCH_DISK_ACCOUNTING_dev_data_type: + { + guard(rcu)(); /* scoped guard is a loop, and doesn't play nicely with continue */ + struct bch_dev *ca = bch2_dev_rcu_noerror(c, acc_k.dev_data_type.dev); + if (!ca) + continue; + + v[0] = percpu_u64_get(&ca->usage->d[acc_k.dev_data_type.data_type].buckets); + v[1] = percpu_u64_get(&ca->usage->d[acc_k.dev_data_type.data_type].sectors); + v[2] = percpu_u64_get(&ca->usage->d[acc_k.dev_data_type.data_type].fragmented); } - v[0] = percpu_u64_get(&ca->usage->d[acc_k.dev_data_type.data_type].buckets); - v[1] = percpu_u64_get(&ca->usage->d[acc_k.dev_data_type.data_type].sectors); - v[2] = percpu_u64_get(&ca->usage->d[acc_k.dev_data_type.data_type].fragmented); - rcu_read_unlock(); - if (memcmp(a.v->d, v, 3 * sizeof(u64))) { struct printbuf buf = PRINTBUF; @@ -1032,7 +1033,6 @@ void bch2_verify_accounting_clean(struct bch_fs *c) mismatch = true; } } - } 0; }))); diff --git a/fs/bcachefs/disk_accounting.h b/fs/bcachefs/disk_accounting.h index f6098e33ab30..d61abebf3e0b 100644 --- a/fs/bcachefs/disk_accounting.h +++ b/fs/bcachefs/disk_accounting.h @@ -174,17 +174,17 @@ static inline int bch2_accounting_mem_mod_locked(struct btree_trans *trans, case BCH_DISK_ACCOUNTING_replicas: fs_usage_data_type_to_base(&trans->fs_usage_delta, acc_k.replicas.data_type, a.v->d[0]); break; - case BCH_DISK_ACCOUNTING_dev_data_type: - rcu_read_lock(); + case BCH_DISK_ACCOUNTING_dev_data_type: { + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, acc_k.dev_data_type.dev); if (ca) { this_cpu_add(ca->usage->d[acc_k.dev_data_type.data_type].buckets, a.v->d[0]); this_cpu_add(ca->usage->d[acc_k.dev_data_type.data_type].sectors, a.v->d[1]); this_cpu_add(ca->usage->d[acc_k.dev_data_type.data_type].fragmented, a.v->d[2]); } - rcu_read_unlock(); break; } + } } unsigned idx; diff --git a/fs/bcachefs/disk_groups.c b/fs/bcachefs/disk_groups.c index c20ecf5e5381..cde842ac1886 100644 --- a/fs/bcachefs/disk_groups.c +++ b/fs/bcachefs/disk_groups.c @@ -130,7 +130,7 @@ int bch2_sb_disk_groups_to_cpu(struct bch_fs *c) cpu_g = kzalloc(struct_size(cpu_g, entries, nr_groups), GFP_KERNEL); if (!cpu_g) - return -BCH_ERR_ENOMEM_disk_groups_to_cpu; + return bch_err_throw(c, ENOMEM_disk_groups_to_cpu); cpu_g->nr = nr_groups; @@ -170,36 +170,28 @@ int bch2_sb_disk_groups_to_cpu(struct bch_fs *c) const struct bch_devs_mask *bch2_target_to_mask(struct bch_fs *c, unsigned target) { struct target t = target_decode(target); - struct bch_devs_mask *devs; - rcu_read_lock(); + guard(rcu)(); switch (t.type) { case TARGET_NULL: - devs = NULL; - break; + return NULL; case TARGET_DEV: { struct bch_dev *ca = t.dev < c->sb.nr_devices ? rcu_dereference(c->devs[t.dev]) : NULL; - devs = ca ? &ca->self : NULL; - break; + return ca ? &ca->self : NULL; } case TARGET_GROUP: { struct bch_disk_groups_cpu *g = rcu_dereference(c->disk_groups); - devs = g && t.group < g->nr && !g->entries[t.group].deleted + return g && t.group < g->nr && !g->entries[t.group].deleted ? &g->entries[t.group].devs : NULL; - break; } default: BUG(); } - - rcu_read_unlock(); - - return devs; } bool bch2_dev_in_target(struct bch_fs *c, unsigned dev, unsigned target) @@ -384,7 +376,7 @@ void bch2_disk_groups_to_text(struct printbuf *out, struct bch_fs *c) bch2_printbuf_make_room(out, 4096); out->atomic++; - rcu_read_lock(); + guard(rcu)(); struct bch_disk_groups_cpu *g = rcu_dereference(c->disk_groups); for (unsigned i = 0; i < (g ? g->nr : 0); i++) { @@ -405,16 +397,14 @@ next: prt_newline(out); } - rcu_read_unlock(); out->atomic--; } void bch2_disk_path_to_text(struct printbuf *out, struct bch_fs *c, unsigned v) { out->atomic++; - rcu_read_lock(); + guard(rcu)(); __bch2_disk_path_to_text(out, rcu_dereference(c->disk_groups), v), - rcu_read_unlock(); --out->atomic; } @@ -535,13 +525,11 @@ void bch2_target_to_text(struct printbuf *out, struct bch_fs *c, unsigned v) switch (t.type) { case TARGET_NULL: prt_printf(out, "none"); - break; + return; case TARGET_DEV: { - struct bch_dev *ca; - out->atomic++; - rcu_read_lock(); - ca = t.dev < c->sb.nr_devices + guard(rcu)(); + struct bch_dev *ca = t.dev < c->sb.nr_devices ? rcu_dereference(c->devs[t.dev]) : NULL; @@ -552,13 +540,12 @@ void bch2_target_to_text(struct printbuf *out, struct bch_fs *c, unsigned v) else prt_printf(out, "invalid device %u", t.dev); - rcu_read_unlock(); out->atomic--; - break; + return; } case TARGET_GROUP: bch2_disk_path_to_text(out, c, t.group); - break; + return; default: BUG(); } diff --git a/fs/bcachefs/ec.c b/fs/bcachefs/ec.c index c581426e3894..543dbba9b14f 100644 --- a/fs/bcachefs/ec.c +++ b/fs/bcachefs/ec.c @@ -213,7 +213,7 @@ static int __mark_stripe_bucket(struct btree_trans *trans, a->dirty_sectors, a->stripe, s.k->p.offset, (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) { - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } @@ -224,7 +224,7 @@ static int __mark_stripe_bucket(struct btree_trans *trans, a->dirty_sectors, a->cached_sectors, (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) { - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } } else { @@ -234,7 +234,7 @@ static int __mark_stripe_bucket(struct btree_trans *trans, bucket.inode, bucket.offset, a->gen, a->stripe, (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) { - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } @@ -244,7 +244,7 @@ static int __mark_stripe_bucket(struct btree_trans *trans, bch2_data_type_str(a->data_type), bch2_data_type_str(data_type), (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) { - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } @@ -256,7 +256,7 @@ static int __mark_stripe_bucket(struct btree_trans *trans, a->dirty_sectors, a->cached_sectors, (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) { - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } } @@ -295,7 +295,7 @@ static int mark_stripe_bucket(struct btree_trans *trans, struct bch_dev *ca = bch2_dev_tryget(c, ptr->dev); if (unlikely(!ca)) { if (ptr->dev != BCH_SB_MEMBER_INVALID && !(flags & BTREE_TRIGGER_overwrite)) - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } @@ -325,7 +325,7 @@ static int mark_stripe_bucket(struct btree_trans *trans, if (bch2_fs_inconsistent_on(!g, c, "reference to invalid bucket on device %u\n%s", ptr->dev, (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) { - ret = -BCH_ERR_mark_stripe; + ret = bch_err_throw(c, mark_stripe); goto err; } @@ -428,7 +428,7 @@ int bch2_trigger_stripe(struct btree_trans *trans, gc = genradix_ptr_alloc(&c->gc_stripes, idx, GFP_KERNEL); if (!gc) { bch_err(c, "error allocating memory for gc_stripes, idx %llu", idx); - return -BCH_ERR_ENOMEM_mark_stripe; + return bch_err_throw(c, ENOMEM_mark_stripe); } /* @@ -536,7 +536,8 @@ static void ec_stripe_buf_exit(struct ec_stripe_buf *buf) } /* XXX: this is a non-mempoolified memory allocation: */ -static int ec_stripe_buf_init(struct ec_stripe_buf *buf, +static int ec_stripe_buf_init(struct bch_fs *c, + struct ec_stripe_buf *buf, unsigned offset, unsigned size) { struct bch_stripe *v = &bkey_i_to_stripe(&buf->key)->v; @@ -564,7 +565,7 @@ static int ec_stripe_buf_init(struct ec_stripe_buf *buf, return 0; err: ec_stripe_buf_exit(buf); - return -BCH_ERR_ENOMEM_stripe_buf; + return bch_err_throw(c, ENOMEM_stripe_buf); } /* Checksumming: */ @@ -840,7 +841,7 @@ int bch2_ec_read_extent(struct btree_trans *trans, struct bch_read_bio *rbio, buf = kzalloc(sizeof(*buf), GFP_NOFS); if (!buf) - return -BCH_ERR_ENOMEM_ec_read_extent; + return bch_err_throw(c, ENOMEM_ec_read_extent); ret = lockrestart_do(trans, get_stripe_key_trans(trans, rbio->pick.ec.idx, buf)); if (ret) { @@ -861,7 +862,7 @@ int bch2_ec_read_extent(struct btree_trans *trans, struct bch_read_bio *rbio, goto err; } - ret = ec_stripe_buf_init(buf, offset, bio_sectors(&rbio->bio)); + ret = ec_stripe_buf_init(c, buf, offset, bio_sectors(&rbio->bio)); if (ret) { msg = "-ENOMEM"; goto err; @@ -894,7 +895,7 @@ err: bch_err_ratelimited(c, "error doing reconstruct read: %s\n %s", msg, msgbuf.buf); printbuf_exit(&msgbuf); - ret = -BCH_ERR_stripe_reconstruct; + ret = bch_err_throw(c, stripe_reconstruct); goto out; } @@ -904,7 +905,7 @@ static int __ec_stripe_mem_alloc(struct bch_fs *c, size_t idx, gfp_t gfp) { if (c->gc_pos.phase != GC_PHASE_not_running && !genradix_ptr_alloc(&c->gc_stripes, idx, gfp)) - return -BCH_ERR_ENOMEM_ec_stripe_mem_alloc; + return bch_err_throw(c, ENOMEM_ec_stripe_mem_alloc); return 0; } @@ -1129,7 +1130,7 @@ static int ec_stripe_update_extent(struct btree_trans *trans, bch2_fs_inconsistent(c, "%s", buf.buf); printbuf_exit(&buf); - return -BCH_ERR_erasure_coding_found_btree_node; + return bch_err_throw(c, erasure_coding_found_btree_node); } k = bch2_backpointer_get_key(trans, bp, &iter, BTREE_ITER_intent, last_flushed); @@ -1195,7 +1196,7 @@ static int ec_stripe_update_bucket(struct btree_trans *trans, struct ec_stripe_b struct bch_dev *ca = bch2_dev_tryget(c, ptr.dev); if (!ca) - return -BCH_ERR_ENOENT_dev_not_found; + return bch_err_throw(c, ENOENT_dev_not_found); struct bpos bucket_pos = PTR_BUCKET_POS(ca, &ptr); @@ -1256,7 +1257,7 @@ static void zero_out_rest_of_ec_bucket(struct bch_fs *c, struct bch_dev *ca = bch2_dev_get_ioref(c, ob->dev, WRITE, BCH_DEV_WRITE_REF_ec_bucket_zero); if (!ca) { - s->err = -BCH_ERR_erofs_no_writes; + s->err = bch_err_throw(c, erofs_no_writes); return; } @@ -1320,7 +1321,7 @@ static void ec_stripe_create(struct ec_stripe_new *s) if (ec_do_recov(c, &s->existing_stripe)) { bch_err(c, "error creating stripe: error reading existing stripe"); - ret = -BCH_ERR_ec_block_read; + ret = bch_err_throw(c, ec_block_read); goto err; } @@ -1346,7 +1347,7 @@ static void ec_stripe_create(struct ec_stripe_new *s) if (ec_nr_failed(&s->new_stripe)) { bch_err(c, "error creating stripe: error writing redundancy buckets"); - ret = -BCH_ERR_ec_block_write; + ret = bch_err_throw(c, ec_block_write); goto err; } @@ -1578,26 +1579,26 @@ static struct ec_stripe_new *ec_new_stripe_alloc(struct bch_fs *c, struct ec_str static void ec_stripe_head_devs_update(struct bch_fs *c, struct ec_stripe_head *h) { struct bch_devs_mask devs = h->devs; + unsigned nr_devs, nr_devs_with_durability; - rcu_read_lock(); - h->devs = target_rw_devs(c, BCH_DATA_user, h->disk_label - ? group_to_target(h->disk_label - 1) - : 0); - unsigned nr_devs = dev_mask_nr(&h->devs); + scoped_guard(rcu) { + h->devs = target_rw_devs(c, BCH_DATA_user, h->disk_label + ? group_to_target(h->disk_label - 1) + : 0); + nr_devs = dev_mask_nr(&h->devs); - for_each_member_device_rcu(c, ca, &h->devs) - if (!ca->mi.durability) - __clear_bit(ca->dev_idx, h->devs.d); - unsigned nr_devs_with_durability = dev_mask_nr(&h->devs); + for_each_member_device_rcu(c, ca, &h->devs) + if (!ca->mi.durability) + __clear_bit(ca->dev_idx, h->devs.d); + nr_devs_with_durability = dev_mask_nr(&h->devs); - h->blocksize = pick_blocksize(c, &h->devs); + h->blocksize = pick_blocksize(c, &h->devs); - h->nr_active_devs = 0; - for_each_member_device_rcu(c, ca, &h->devs) - if (ca->mi.bucket_size == h->blocksize) - h->nr_active_devs++; - - rcu_read_unlock(); + h->nr_active_devs = 0; + for_each_member_device_rcu(c, ca, &h->devs) + if (ca->mi.bucket_size == h->blocksize) + h->nr_active_devs++; + } /* * If we only have redundancy + 1 devices, we're better off with just @@ -1865,7 +1866,7 @@ static int init_new_stripe_from_existing(struct bch_fs *c, struct ec_stripe_new s->nr_data = existing_v->nr_blocks - existing_v->nr_redundant; - int ret = ec_stripe_buf_init(&s->existing_stripe, 0, le16_to_cpu(existing_v->sectors)); + int ret = ec_stripe_buf_init(c, &s->existing_stripe, 0, le16_to_cpu(existing_v->sectors)); if (ret) { bch2_stripe_close(c, s); return ret; @@ -1925,7 +1926,7 @@ static int __bch2_ec_stripe_head_reuse(struct btree_trans *trans, struct ec_stri } bch2_trans_iter_exit(trans, &lru_iter); if (!ret) - ret = -BCH_ERR_stripe_alloc_blocked; + ret = bch_err_throw(c, stripe_alloc_blocked); if (ret == 1) ret = 0; if (ret) @@ -1966,7 +1967,7 @@ static int __bch2_ec_stripe_head_reserve(struct btree_trans *trans, struct ec_st continue; } - ret = -BCH_ERR_ENOSPC_stripe_create; + ret = bch_err_throw(c, ENOSPC_stripe_create); break; } @@ -2024,7 +2025,7 @@ struct ec_stripe_head *bch2_ec_stripe_head_get(struct btree_trans *trans, if (!h->s) { h->s = ec_new_stripe_alloc(c, h); if (!h->s) { - ret = -BCH_ERR_ENOMEM_ec_new_stripe_alloc; + ret = bch_err_throw(c, ENOMEM_ec_new_stripe_alloc); bch_err(c, "failed to allocate new stripe"); goto err; } @@ -2089,7 +2090,7 @@ alloc_existing: goto err; allocate_buf: - ret = ec_stripe_buf_init(&s->new_stripe, 0, h->blocksize); + ret = ec_stripe_buf_init(c, &s->new_stripe, 0, h->blocksize); if (ret) goto err; @@ -2115,6 +2116,7 @@ int bch2_invalidate_stripe_to_dev(struct btree_trans *trans, if (k.k->type != KEY_TYPE_stripe) return 0; + struct bch_fs *c = trans->c; struct bkey_i_stripe *s = bch2_bkey_make_mut_typed(trans, iter, &k, 0, stripe); int ret = PTR_ERR_OR_ZERO(s); @@ -2141,23 +2143,22 @@ int bch2_invalidate_stripe_to_dev(struct btree_trans *trans, unsigned nr_good = 0; - rcu_read_lock(); - bkey_for_each_ptr(ptrs, ptr) { - if (ptr->dev == dev_idx) - ptr->dev = BCH_SB_MEMBER_INVALID; + scoped_guard(rcu) + bkey_for_each_ptr(ptrs, ptr) { + if (ptr->dev == dev_idx) + ptr->dev = BCH_SB_MEMBER_INVALID; - struct bch_dev *ca = bch2_dev_rcu(trans->c, ptr->dev); - nr_good += ca && ca->mi.state != BCH_MEMBER_STATE_failed; - } - rcu_read_unlock(); + struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); + nr_good += ca && ca->mi.state != BCH_MEMBER_STATE_failed; + } if (nr_good < s->v.nr_blocks && !(flags & BCH_FORCE_IF_DATA_DEGRADED)) - return -BCH_ERR_remove_would_lose_data; + return bch_err_throw(c, remove_would_lose_data); unsigned nr_data = s->v.nr_blocks - s->v.nr_redundant; if (nr_good < nr_data && !(flags & BCH_FORCE_IF_DATA_LOST)) - return -BCH_ERR_remove_would_lose_data; + return bch_err_throw(c, remove_would_lose_data); sectors = -sectors; @@ -2178,14 +2179,15 @@ static int bch2_invalidate_stripe_to_dev_from_alloc(struct btree_trans *trans, s return 0; if (a->stripe_sectors) { - bch_err(trans->c, "trying to invalidate device in stripe when bucket has stripe data"); - return -BCH_ERR_invalidate_stripe_to_dev; + struct bch_fs *c = trans->c; + bch_err(c, "trying to invalidate device in stripe when bucket has stripe data"); + return bch_err_throw(c, invalidate_stripe_to_dev); } struct btree_iter iter; struct bkey_s_c_stripe s = bch2_bkey_get_iter_typed(trans, &iter, BTREE_ID_stripes, POS(0, a->stripe), - BTREE_ITER_slots, stripe); + BTREE_ITER_slots, stripe); int ret = bkey_err(s); if (ret) return ret; diff --git a/fs/bcachefs/errcode.c b/fs/bcachefs/errcode.c index 43557bebd0f8..c39cf304c681 100644 --- a/fs/bcachefs/errcode.c +++ b/fs/bcachefs/errcode.c @@ -13,12 +13,13 @@ static const char * const bch2_errcode_strs[] = { NULL }; -static unsigned bch2_errcode_parents[] = { +static const unsigned bch2_errcode_parents[] = { #define x(class, err) [BCH_ERR_##err - BCH_ERR_START] = class, BCH_ERRCODES() #undef x }; +__attribute__((const)) const char *bch2_err_str(int err) { const char *errstr; @@ -36,6 +37,7 @@ const char *bch2_err_str(int err) return errstr ?: "(Invalid error)"; } +__attribute__((const)) bool __bch2_err_matches(int err, int class) { err = abs(err); diff --git a/fs/bcachefs/errcode.h b/fs/bcachefs/errcode.h index 62843e772b2c..acc3b7b67704 100644 --- a/fs/bcachefs/errcode.h +++ b/fs/bcachefs/errcode.h @@ -137,7 +137,6 @@ x(BCH_ERR_transaction_restart, transaction_restart_relock) \ x(BCH_ERR_transaction_restart, transaction_restart_relock_path) \ x(BCH_ERR_transaction_restart, transaction_restart_relock_path_intent) \ - x(BCH_ERR_transaction_restart, transaction_restart_relock_after_fill) \ x(BCH_ERR_transaction_restart, transaction_restart_too_many_iters) \ x(BCH_ERR_transaction_restart, transaction_restart_lock_node_reused) \ x(BCH_ERR_transaction_restart, transaction_restart_fill_relock) \ @@ -148,11 +147,8 @@ x(BCH_ERR_transaction_restart, transaction_restart_would_deadlock_write)\ x(BCH_ERR_transaction_restart, transaction_restart_deadlock_recursion_limit)\ x(BCH_ERR_transaction_restart, transaction_restart_upgrade) \ - x(BCH_ERR_transaction_restart, transaction_restart_key_cache_upgrade) \ x(BCH_ERR_transaction_restart, transaction_restart_key_cache_fill) \ x(BCH_ERR_transaction_restart, transaction_restart_key_cache_raced) \ - x(BCH_ERR_transaction_restart, transaction_restart_key_cache_realloced)\ - x(BCH_ERR_transaction_restart, transaction_restart_journal_preres_get) \ x(BCH_ERR_transaction_restart, transaction_restart_split_race) \ x(BCH_ERR_transaction_restart, transaction_restart_write_buffer_flush) \ x(BCH_ERR_transaction_restart, transaction_restart_nested) \ @@ -182,9 +178,12 @@ x(BCH_ERR_fsck, fsck_errors_not_fixed) \ x(BCH_ERR_fsck, fsck_repair_unimplemented) \ x(BCH_ERR_fsck, fsck_repair_impossible) \ - x(EINVAL, restart_recovery) \ - x(EINVAL, cannot_rewind_recovery) \ + x(EINVAL, recovery_will_run) \ + x(BCH_ERR_recovery_will_run, restart_recovery) \ + x(BCH_ERR_recovery_will_run, cannot_rewind_recovery) \ + x(BCH_ERR_recovery_will_run, recovery_pass_will_run) \ x(0, data_update_done) \ + x(0, bkey_was_deleted) \ x(BCH_ERR_data_update_done, data_update_done_would_block) \ x(BCH_ERR_data_update_done, data_update_done_unwritten) \ x(BCH_ERR_data_update_done, data_update_done_no_writes_needed) \ @@ -211,6 +210,8 @@ x(EINVAL, remove_would_lose_data) \ x(EINVAL, no_resize_with_buckets_nouse) \ x(EINVAL, inode_unpack_error) \ + x(EINVAL, inode_not_unlinked) \ + x(EINVAL, inode_has_child_snapshot) \ x(EINVAL, varint_decode_error) \ x(EINVAL, erasure_coding_found_btree_node) \ x(EINVAL, option_negative) \ @@ -236,7 +237,6 @@ x(BCH_ERR_journal_res_blocked, journal_buf_enomem) \ x(BCH_ERR_journal_res_blocked, journal_stuck) \ x(BCH_ERR_journal_res_blocked, journal_retry_open) \ - x(BCH_ERR_journal_res_blocked, journal_preres_get_blocked) \ x(BCH_ERR_journal_res_blocked, bucket_alloc_blocked) \ x(BCH_ERR_journal_res_blocked, stripe_alloc_blocked) \ x(BCH_ERR_invalid, invalid_sb) \ @@ -282,7 +282,6 @@ x(EIO, sb_not_downgraded) \ x(EIO, btree_node_write_all_failed) \ x(EIO, btree_node_read_error) \ - x(EIO, btree_node_read_validate_error) \ x(EIO, btree_need_topology_repair) \ x(EIO, bucket_ref_update) \ x(EIO, trigger_alloc) \ @@ -357,9 +356,11 @@ enum bch_errcode { BCH_ERR_MAX }; -const char *bch2_err_str(int); -bool __bch2_err_matches(int, int); +__attribute__((const)) const char *bch2_err_str(int); +__attribute__((const)) bool __bch2_err_matches(int, int); + +__attribute__((const)) static inline bool _bch2_err_matches(int err, int class) { return err < 0 && __bch2_err_matches(err, class); diff --git a/fs/bcachefs/error.c b/fs/bcachefs/error.c index c2cad28635bf..267e73d9d7e6 100644 --- a/fs/bcachefs/error.c +++ b/fs/bcachefs/error.c @@ -69,7 +69,7 @@ static bool bch2_fs_trans_inconsistent(struct bch_fs *c, struct btree_trans *tra if (trans) bch2_trans_updates_to_text(&buf, trans); bool ret = __bch2_inconsistent_error(c, &buf); - bch2_print_str_nonblocking(c, KERN_ERR, buf.buf); + bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); return ret; @@ -100,10 +100,10 @@ int __bch2_topology_error(struct bch_fs *c, struct printbuf *out) set_bit(BCH_FS_topology_error, &c->flags); if (!test_bit(BCH_FS_in_recovery, &c->flags)) { __bch2_inconsistent_error(c, out); - return -BCH_ERR_btree_need_topology_repair; + return bch_err_throw(c, btree_need_topology_repair); } else { return bch2_run_explicit_recovery_pass(c, out, BCH_RECOVERY_PASS_check_topology, 0) ?: - -BCH_ERR_btree_node_read_validate_error; + bch_err_throw(c, btree_need_topology_repair); } } @@ -403,23 +403,23 @@ int bch2_fsck_err_opt(struct bch_fs *c, if (test_bit(BCH_FS_in_fsck, &c->flags)) { if (!(flags & (FSCK_CAN_FIX|FSCK_CAN_IGNORE))) - return -BCH_ERR_fsck_repair_unimplemented; + return bch_err_throw(c, fsck_repair_unimplemented); switch (c->opts.fix_errors) { case FSCK_FIX_exit: - return -BCH_ERR_fsck_errors_not_fixed; + return bch_err_throw(c, fsck_errors_not_fixed); case FSCK_FIX_yes: if (flags & FSCK_CAN_FIX) - return -BCH_ERR_fsck_fix; + return bch_err_throw(c, fsck_fix); fallthrough; case FSCK_FIX_no: if (flags & FSCK_CAN_IGNORE) - return -BCH_ERR_fsck_ignore; - return -BCH_ERR_fsck_errors_not_fixed; + return bch_err_throw(c, fsck_ignore); + return bch_err_throw(c, fsck_errors_not_fixed); case FSCK_FIX_ask: if (flags & FSCK_AUTOFIX) - return -BCH_ERR_fsck_fix; - return -BCH_ERR_fsck_ask; + return bch_err_throw(c, fsck_fix); + return bch_err_throw(c, fsck_ask); default: BUG(); } @@ -427,12 +427,12 @@ int bch2_fsck_err_opt(struct bch_fs *c, if ((flags & FSCK_AUTOFIX) && (c->opts.errors == BCH_ON_ERROR_continue || c->opts.errors == BCH_ON_ERROR_fix_safe)) - return -BCH_ERR_fsck_fix; + return bch_err_throw(c, fsck_fix); if (c->opts.errors == BCH_ON_ERROR_continue && (flags & FSCK_CAN_IGNORE)) - return -BCH_ERR_fsck_ignore; - return -BCH_ERR_fsck_errors_not_fixed; + return bch_err_throw(c, fsck_ignore); + return bch_err_throw(c, fsck_errors_not_fixed); } } @@ -444,7 +444,7 @@ int __bch2_fsck_err(struct bch_fs *c, { va_list args; struct printbuf buf = PRINTBUF, *out = &buf; - int ret = -BCH_ERR_fsck_ignore; + int ret = 0; const char *action_orig = "fix?", *action = action_orig; might_sleep(); @@ -474,8 +474,8 @@ int __bch2_fsck_err(struct bch_fs *c, if (test_bit(err, c->sb.errors_silent)) return flags & FSCK_CAN_FIX - ? -BCH_ERR_fsck_fix - : -BCH_ERR_fsck_ignore; + ? bch_err_throw(c, fsck_fix) + : bch_err_throw(c, fsck_ignore); printbuf_indent_add_nextline(out, 2); @@ -517,10 +517,10 @@ int __bch2_fsck_err(struct bch_fs *c, prt_str(out, ", "); if (flags & FSCK_CAN_FIX) { prt_actioning(out, action); - ret = -BCH_ERR_fsck_fix; + ret = bch_err_throw(c, fsck_fix); } else { prt_str(out, ", continuing"); - ret = -BCH_ERR_fsck_ignore; + ret = bch_err_throw(c, fsck_ignore); } goto print; @@ -532,18 +532,18 @@ int __bch2_fsck_err(struct bch_fs *c, "run fsck, and forward to devs so error can be marked for self-healing"); inconsistent = true; print = true; - ret = -BCH_ERR_fsck_errors_not_fixed; + ret = bch_err_throw(c, fsck_errors_not_fixed); } else if (flags & FSCK_CAN_FIX) { prt_str(out, ", "); prt_actioning(out, action); - ret = -BCH_ERR_fsck_fix; + ret = bch_err_throw(c, fsck_fix); } else { prt_str(out, ", continuing"); - ret = -BCH_ERR_fsck_ignore; + ret = bch_err_throw(c, fsck_ignore); } } else if (c->opts.fix_errors == FSCK_FIX_exit) { prt_str(out, ", exiting"); - ret = -BCH_ERR_fsck_errors_not_fixed; + ret = bch_err_throw(c, fsck_errors_not_fixed); } else if (flags & FSCK_CAN_FIX) { int fix = s && s->fix ? s->fix @@ -562,30 +562,37 @@ int __bch2_fsck_err(struct bch_fs *c, : FSCK_FIX_yes; ret = ret & 1 - ? -BCH_ERR_fsck_fix - : -BCH_ERR_fsck_ignore; + ? bch_err_throw(c, fsck_fix) + : bch_err_throw(c, fsck_ignore); } else if (fix == FSCK_FIX_yes || (c->opts.nochanges && !(flags & FSCK_CAN_IGNORE))) { prt_str(out, ", "); prt_actioning(out, action); - ret = -BCH_ERR_fsck_fix; + ret = bch_err_throw(c, fsck_fix); } else { prt_str(out, ", not "); prt_actioning(out, action); + ret = bch_err_throw(c, fsck_ignore); + } + } else { + if (flags & FSCK_CAN_IGNORE) { + prt_str(out, ", continuing"); + ret = bch_err_throw(c, fsck_ignore); + } else { + prt_str(out, " (repair unimplemented)"); + ret = bch_err_throw(c, fsck_repair_unimplemented); } - } else if (!(flags & FSCK_CAN_IGNORE)) { - prt_str(out, " (repair unimplemented)"); } - if (ret == -BCH_ERR_fsck_ignore && + if (bch2_err_matches(ret, BCH_ERR_fsck_ignore) && (c->opts.fix_errors == FSCK_FIX_exit || !(flags & FSCK_CAN_IGNORE))) - ret = -BCH_ERR_fsck_errors_not_fixed; + ret = bch_err_throw(c, fsck_errors_not_fixed); if (test_bit(BCH_FS_in_fsck, &c->flags) && - (ret != -BCH_ERR_fsck_fix && - ret != -BCH_ERR_fsck_ignore)) { + (!bch2_err_matches(ret, BCH_ERR_fsck_fix) && + !bch2_err_matches(ret, BCH_ERR_fsck_ignore))) { exiting = true; print = true; } @@ -614,25 +621,32 @@ print: if (s) s->ret = ret; + if (trans && + !(flags & FSCK_ERR_NO_LOG) && + ret == -BCH_ERR_fsck_fix) + ret = bch2_trans_log_str(trans, bch2_sb_error_strs[err]) ?: ret; +err_unlock: + mutex_unlock(&c->fsck_error_msgs_lock); +err: /* * We don't yet track whether the filesystem currently has errors, for * log_fsck_err()s: that would require us to track for every error type * which recovery pass corrects it, to get the fsck exit status correct: */ - if (flags & FSCK_CAN_FIX) { - if (ret == -BCH_ERR_fsck_fix) { - set_bit(BCH_FS_errors_fixed, &c->flags); - } else { - set_bit(BCH_FS_errors_not_fixed, &c->flags); - set_bit(BCH_FS_error, &c->flags); - } + if (bch2_err_matches(ret, BCH_ERR_transaction_restart)) { + /* nothing */ + } else if (bch2_err_matches(ret, BCH_ERR_fsck_fix)) { + set_bit(BCH_FS_errors_fixed, &c->flags); + } else { + set_bit(BCH_FS_errors_not_fixed, &c->flags); + set_bit(BCH_FS_error, &c->flags); } -err_unlock: - mutex_unlock(&c->fsck_error_msgs_lock); -err: + if (action != action_orig) kfree(action); printbuf_exit(&buf); + + BUG_ON(!ret); return ret; } @@ -650,12 +664,12 @@ int __bch2_bkey_fsck_err(struct bch_fs *c, const char *fmt, ...) { if (from.flags & BCH_VALIDATE_silent) - return -BCH_ERR_fsck_delete_bkey; + return bch_err_throw(c, fsck_delete_bkey); unsigned fsck_flags = 0; if (!(from.flags & (BCH_VALIDATE_write|BCH_VALIDATE_commit))) { if (test_bit(err, c->sb.errors_silent)) - return -BCH_ERR_fsck_delete_bkey; + return bch_err_throw(c, fsck_delete_bkey); fsck_flags |= FSCK_AUTOFIX|FSCK_CAN_FIX; } diff --git a/fs/bcachefs/error.h b/fs/bcachefs/error.h index 5123d4c86770..0c3c3a24fc6f 100644 --- a/fs/bcachefs/error.h +++ b/fs/bcachefs/error.h @@ -105,13 +105,13 @@ void bch2_free_fsck_errs(struct bch_fs *); #define fsck_err_wrap(_do) \ ({ \ int _ret = _do; \ - if (_ret != -BCH_ERR_fsck_fix && \ - _ret != -BCH_ERR_fsck_ignore) { \ + if (!bch2_err_matches(_ret, BCH_ERR_fsck_fix) && \ + !bch2_err_matches(_ret, BCH_ERR_fsck_ignore)) { \ ret = _ret; \ goto fsck_err; \ } \ \ - _ret == -BCH_ERR_fsck_fix; \ + bch2_err_matches(_ret, BCH_ERR_fsck_fix); \ }) #define __fsck_err(...) fsck_err_wrap(bch2_fsck_err(__VA_ARGS__)) @@ -170,10 +170,10 @@ do { \ int _ret = __bch2_bkey_fsck_err(c, k, from, \ BCH_FSCK_ERR_##_err_type, \ _err_msg, ##__VA_ARGS__); \ - if (_ret != -BCH_ERR_fsck_fix && \ - _ret != -BCH_ERR_fsck_ignore) \ + if (!bch2_err_matches(_ret, BCH_ERR_fsck_fix) && \ + !bch2_err_matches(_ret, BCH_ERR_fsck_ignore)) \ ret = _ret; \ - ret = -BCH_ERR_fsck_delete_bkey; \ + ret = bch_err_throw(c, fsck_delete_bkey); \ goto fsck_err; \ } while (0) diff --git a/fs/bcachefs/extent_update.c b/fs/bcachefs/extent_update.c index b899ee75f5b9..e76e58a568bf 100644 --- a/fs/bcachefs/extent_update.c +++ b/fs/bcachefs/extent_update.c @@ -139,6 +139,17 @@ int bch2_extent_trim_atomic(struct btree_trans *trans, if (ret) return ret; - bch2_cut_back(end, k); + /* tracepoint */ + + if (bpos_lt(end, k->k.p)) { + if (trace_extent_trim_atomic_enabled()) { + CLASS(printbuf, buf)(); + bch2_bpos_to_text(&buf, end); + prt_newline(&buf); + bch2_bkey_val_to_text(&buf, trans->c, bkey_i_to_s_c(k)); + trace_extent_trim_atomic(trans->c, buf.buf); + } + bch2_cut_back(end, k); + } return 0; } diff --git a/fs/bcachefs/extents.c b/fs/bcachefs/extents.c index 1ac9897f189d..83cbd77dcb9c 100644 --- a/fs/bcachefs/extents.c +++ b/fs/bcachefs/extents.c @@ -50,34 +50,34 @@ void bch2_io_failures_to_text(struct printbuf *out, struct bch_io_failures *failed) { static const char * const error_types[] = { - "io", "checksum", "ec reconstruct", NULL + "btree validate", "io", "checksum", "ec reconstruct", NULL }; for (struct bch_dev_io_failures *f = failed->devs; f < failed->devs + failed->nr; f++) { unsigned errflags = - ((!!f->failed_io) << 0) | - ((!!f->failed_csum_nr) << 1) | - ((!!f->failed_ec) << 2); - - if (!errflags) - continue; + ((!!f->failed_btree_validate) << 0) | + ((!!f->failed_io) << 1) | + ((!!f->failed_csum_nr) << 2) | + ((!!f->failed_ec) << 3); bch2_printbuf_make_room(out, 1024); - rcu_read_lock(); out->atomic++; - struct bch_dev *ca = bch2_dev_rcu_noerror(c, f->dev); - if (ca) - prt_str(out, ca->name); - else - prt_printf(out, "(invalid device %u)", f->dev); + scoped_guard(rcu) { + struct bch_dev *ca = bch2_dev_rcu_noerror(c, f->dev); + if (ca) + prt_str(out, ca->name); + else + prt_printf(out, "(invalid device %u)", f->dev); + } --out->atomic; - rcu_read_unlock(); prt_char(out, ' '); - if (is_power_of_2(errflags)) { + if (!errflags) { + prt_str(out, "no error - confused"); + } else if (is_power_of_2(errflags)) { prt_bitflags(out, error_types, errflags); prt_str(out, " error"); } else { @@ -193,7 +193,7 @@ int bch2_bkey_pick_read_device(struct bch_fs *c, struct bkey_s_c k, bool have_dirty_ptrs = false, have_pick = false; if (k.k->type == KEY_TYPE_error) - return -BCH_ERR_key_type_error; + return bch_err_throw(c, key_type_error); rcu_read_lock(); struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); @@ -286,17 +286,17 @@ int bch2_bkey_pick_read_device(struct bch_fs *c, struct bkey_s_c k, if (!have_dirty_ptrs) return 0; if (have_missing_devs) - return -BCH_ERR_no_device_to_read_from; + return bch_err_throw(c, no_device_to_read_from); if (have_csum_errors) - return -BCH_ERR_data_read_csum_err; + return bch_err_throw(c, data_read_csum_err); if (have_io_errors) - return -BCH_ERR_data_read_io_err; + return bch_err_throw(c, data_read_io_err); /* * If we get here, we have pointers (bkey_ptrs_validate() ensures that), * but they don't point to valid devices: */ - return -BCH_ERR_no_devices_valid; + return bch_err_throw(c, no_devices_valid); } /* KEY_TYPE_btree_ptr: */ @@ -407,6 +407,8 @@ bool bch2_extent_merge(struct bch_fs *c, struct bkey_s l, struct bkey_s_c r) lp.crc = bch2_extent_crc_unpack(l.k, NULL); rp.crc = bch2_extent_crc_unpack(r.k, NULL); + guard(rcu)(); + while (__bkey_ptr_next_decode(l.k, l_ptrs.end, lp, en_l) && __bkey_ptr_next_decode(r.k, r_ptrs.end, rp, en_r)) { if (lp.ptr.offset + lp.crc.offset + lp.crc.live_size != @@ -418,10 +420,8 @@ bool bch2_extent_merge(struct bch_fs *c, struct bkey_s l, struct bkey_s_c r) return false; /* Extents may not straddle buckets: */ - rcu_read_lock(); struct bch_dev *ca = bch2_dev_rcu(c, lp.ptr.dev); bool same_bucket = ca && PTR_BUCKET_NR(ca, &lp.ptr) == PTR_BUCKET_NR(ca, &rp.ptr); - rcu_read_unlock(); if (!same_bucket) return false; @@ -838,11 +838,9 @@ unsigned bch2_bkey_durability(struct bch_fs *c, struct bkey_s_c k) struct extent_ptr_decoded p; unsigned durability = 0; - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr_decode(k.k, ptrs, p, entry) durability += bch2_extent_ptr_durability(c, &p); - rcu_read_unlock(); - return durability; } @@ -853,12 +851,10 @@ static unsigned bch2_bkey_durability_safe(struct bch_fs *c, struct bkey_s_c k) struct extent_ptr_decoded p; unsigned durability = 0; - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr_decode(k.k, ptrs, p, entry) if (p.ptr.dev < c->sb.nr_devices && c->devs[p.ptr.dev]) durability += bch2_extent_ptr_durability(c, &p); - rcu_read_unlock(); - return durability; } @@ -1015,20 +1011,16 @@ bool bch2_bkey_has_target(struct bch_fs *c, struct bkey_s_c k, unsigned target) { struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); struct bch_dev *ca; - bool ret = false; - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr(ptrs, ptr) if (bch2_dev_in_target(c, ptr->dev, target) && (ca = bch2_dev_rcu(c, ptr->dev)) && (!ptr->cached || - !dev_ptr_stale_rcu(ca, ptr))) { - ret = true; - break; - } - rcu_read_unlock(); + !dev_ptr_stale_rcu(ca, ptr))) + return true; - return ret; + return false; } bool bch2_bkey_matches_ptr(struct bch_fs *c, struct bkey_s_c k, @@ -1142,7 +1134,7 @@ void bch2_extent_ptr_set_cached(struct bch_fs *c, bool have_cached_ptr; unsigned drop_dev = ptr->dev; - rcu_read_lock(); + guard(rcu)(); restart_drop_ptrs: ptrs = bch2_bkey_ptrs(k); have_cached_ptr = false; @@ -1175,10 +1167,8 @@ restart_drop_ptrs: goto drop; ptr->cached = true; - rcu_read_unlock(); return; drop: - rcu_read_unlock(); bch2_bkey_drop_ptr_noerror(k, ptr); } @@ -1194,12 +1184,11 @@ bool bch2_extent_normalize(struct bch_fs *c, struct bkey_s k) { struct bch_dev *ca; - rcu_read_lock(); + guard(rcu)(); bch2_bkey_drop_ptrs(k, ptr, ptr->cached && (!(ca = bch2_dev_rcu(c, ptr->dev)) || dev_ptr_stale_rcu(ca, ptr) > 0)); - rcu_read_unlock(); return bkey_deleted(k.k); } @@ -1217,7 +1206,7 @@ bool bch2_extent_normalize_by_opts(struct bch_fs *c, struct bkey_ptrs ptrs; bool have_cached_ptr; - rcu_read_lock(); + guard(rcu)(); restart_drop_ptrs: ptrs = bch2_bkey_ptrs(k); have_cached_ptr = false; @@ -1230,7 +1219,6 @@ restart_drop_ptrs: } have_cached_ptr = true; } - rcu_read_unlock(); return bkey_deleted(k.k); } @@ -1238,7 +1226,7 @@ restart_drop_ptrs: void bch2_extent_ptr_to_text(struct printbuf *out, struct bch_fs *c, const struct bch_extent_ptr *ptr) { out->atomic++; - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, ptr->dev); if (!ca) { prt_printf(out, "ptr: %u:%llu gen %u%s", ptr->dev, @@ -1262,7 +1250,6 @@ void bch2_extent_ptr_to_text(struct printbuf *out, struct bch_fs *c, const struc else if (stale) prt_printf(out, " invalid"); } - rcu_read_unlock(); --out->atomic; } @@ -1528,7 +1515,7 @@ int bch2_bkey_ptrs_validate(struct bch_fs *c, struct bkey_s_c k, struct bch_compression_opt opt = __bch2_compression_decode(r->compression); prt_printf(err, "invalid compression opt %u:%u", opt.type, opt.level); - return -BCH_ERR_invalid_bkey; + return bch_err_throw(c, invalid_bkey); } #endif break; diff --git a/fs/bcachefs/fs-io-buffered.c b/fs/bcachefs/fs-io-buffered.c index e3a75dcca60c..66bacdd49f78 100644 --- a/fs/bcachefs/fs-io-buffered.c +++ b/fs/bcachefs/fs-io-buffered.c @@ -394,17 +394,9 @@ struct bch_writepage_state { struct bch_io_opts opts; struct bch_folio_sector *tmp; unsigned tmp_sectors; + struct blk_plug plug; }; -static inline struct bch_writepage_state bch_writepage_state_init(struct bch_fs *c, - struct bch_inode_info *inode) -{ - struct bch_writepage_state ret = { 0 }; - - bch2_inode_opts_get(&ret.opts, c, &inode->ei_inode); - return ret; -} - /* * Determine when a writepage io is full. We have to limit writepage bios to a * single page per bvec (i.e. 1MB with 4k pages) because that is the limit to @@ -666,17 +658,17 @@ do_io: int bch2_writepages(struct address_space *mapping, struct writeback_control *wbc) { struct bch_fs *c = mapping->host->i_sb->s_fs_info; - struct bch_writepage_state w = - bch_writepage_state_init(c, to_bch_ei(mapping->host)); - struct blk_plug plug; - int ret; + struct bch_writepage_state *w = kzalloc(sizeof(*w), GFP_NOFS|__GFP_NOFAIL); - blk_start_plug(&plug); - ret = write_cache_pages(mapping, wbc, __bch2_writepage, &w); - if (w.io) - bch2_writepage_do_io(&w); - blk_finish_plug(&plug); - kfree(w.tmp); + bch2_inode_opts_get(&w->opts, c, &to_bch_ei(mapping->host)->ei_inode); + + blk_start_plug(&w->plug); + int ret = write_cache_pages(mapping, wbc, __bch2_writepage, w); + if (w->io) + bch2_writepage_do_io(w); + blk_finish_plug(&w->plug); + kfree(w->tmp); + kfree(w); return bch2_err_class(ret); } diff --git a/fs/bcachefs/fs-io-pagecache.c b/fs/bcachefs/fs-io-pagecache.c index fbae9c1de746..c2cc405822f2 100644 --- a/fs/bcachefs/fs-io-pagecache.c +++ b/fs/bcachefs/fs-io-pagecache.c @@ -447,7 +447,7 @@ static int __bch2_folio_reservation_get(struct bch_fs *c, if (!reserved) { bch2_disk_reservation_put(c, &disk_res); - return -BCH_ERR_ENOSPC_disk_reservation; + return bch_err_throw(c, ENOSPC_disk_reservation); } break; } diff --git a/fs/bcachefs/fs-io.c b/fs/bcachefs/fs-io.c index b1e9ee28fc0f..a233f45875e9 100644 --- a/fs/bcachefs/fs-io.c +++ b/fs/bcachefs/fs-io.c @@ -71,12 +71,12 @@ void bch2_inode_flush_nocow_writes_async(struct bch_fs *c, memset(&inode->ei_devs_need_flush, 0, sizeof(inode->ei_devs_need_flush)); for_each_set_bit(dev, devs.d, BCH_SB_MEMBERS_MAX) { - rcu_read_lock(); - ca = rcu_dereference(c->devs[dev]); - if (ca && !enumerated_ref_tryget(&ca->io_ref[WRITE], - BCH_DEV_WRITE_REF_nocow_flush)) - ca = NULL; - rcu_read_unlock(); + scoped_guard(rcu) { + ca = rcu_dereference(c->devs[dev]); + if (ca && !enumerated_ref_tryget(&ca->io_ref[WRITE], + BCH_DEV_WRITE_REF_nocow_flush)) + ca = NULL; + } if (!ca) continue; diff --git a/fs/bcachefs/fs-ioctl.c b/fs/bcachefs/fs-ioctl.c index 05361a793206..4e72e654da96 100644 --- a/fs/bcachefs/fs-ioctl.c +++ b/fs/bcachefs/fs-ioctl.c @@ -268,13 +268,13 @@ static long bch2_ioctl_subvolume_create(struct bch_fs *c, struct file *filp, } if (dst_dentry->d_inode) { - error = -BCH_ERR_EEXIST_subvolume_create; + error = bch_err_throw(c, EEXIST_subvolume_create); goto err3; } dir = dst_path.dentry->d_inode; if (IS_DEADDIR(dir)) { - error = -BCH_ERR_ENOENT_directory_dead; + error = bch_err_throw(c, ENOENT_directory_dead); goto err3; } diff --git a/fs/bcachefs/fs.c b/fs/bcachefs/fs.c index ddfe89d84966..e54e4f255b22 100644 --- a/fs/bcachefs/fs.c +++ b/fs/bcachefs/fs.c @@ -124,8 +124,9 @@ retry: goto err; struct bch_extent_rebalance new_r = bch2_inode_rebalance_opts_get(c, &inode_u); + bool rebalance_changed = memcmp(&old_r, &new_r, sizeof(new_r)); - if (memcmp(&old_r, &new_r, sizeof(new_r))) { + if (rebalance_changed) { ret = bch2_set_rebalance_needs_scan_trans(trans, inode_u.bi_inum); if (ret) goto err; @@ -146,6 +147,9 @@ err: if (bch2_err_matches(ret, BCH_ERR_transaction_restart)) goto retry; + if (rebalance_changed) + bch2_rebalance_wakeup(c); + bch2_fs_fatal_err_on(bch2_err_matches(ret, ENOENT), c, "%s: inode %llu:%llu not found when updating", bch2_err_str(ret), @@ -718,7 +722,6 @@ static struct dentry *bch2_lookup(struct inode *vdir, struct dentry *dentry, if (IS_ERR(inode)) inode = NULL; -#ifdef CONFIG_UNICODE if (!inode && IS_CASEFOLDED(vdir)) { /* * Do not cache a negative dentry in casefolded directories @@ -733,7 +736,6 @@ static struct dentry *bch2_lookup(struct inode *vdir, struct dentry *dentry, */ return NULL; } -#endif return d_splice_alias(&inode->v, dentry); } @@ -1569,11 +1571,12 @@ static int bch2_vfs_readdir(struct file *file, struct dir_context *ctx) { struct bch_inode_info *inode = file_bch_inode(file); struct bch_fs *c = inode->v.i_sb->s_fs_info; + struct bch_hash_info hash = bch2_hash_info_init(c, &inode->ei_inode); if (!dir_emit_dots(file, ctx)) return 0; - int ret = bch2_readdir(c, inode_inum(inode), ctx); + int ret = bch2_readdir(c, inode_inum(inode), &hash, ctx); bch_err_fn(c, ret); return bch2_err_class(ret); @@ -1727,7 +1730,8 @@ static int bch2_fileattr_set(struct mnt_idmap *idmap, bch2_write_inode(c, inode, fssetxattr_inode_update_fn, &s, ATTR_CTIME); mutex_unlock(&inode->ei_update_lock); - return ret; + + return bch2_err_class(ret); } static const struct file_operations bch_file_operations = { @@ -2002,14 +2006,14 @@ retry: goto err; if (k.k->type != KEY_TYPE_dirent) { - ret = -BCH_ERR_ENOENT_dirent_doesnt_match_inode; + ret = bch_err_throw(c, ENOENT_dirent_doesnt_match_inode); goto err; } d = bkey_s_c_to_dirent(k); ret = bch2_dirent_read_target(trans, inode_inum(dir), d, &target); if (ret > 0) - ret = -BCH_ERR_ENOENT_dirent_doesnt_match_inode; + ret = bch_err_throw(c, ENOENT_dirent_doesnt_match_inode); if (ret) goto err; @@ -2175,7 +2179,13 @@ static void bch2_evict_inode(struct inode *vinode) KEY_TYPE_QUOTA_WARN); bch2_quota_acct(c, inode->ei_qid, Q_INO, -1, KEY_TYPE_QUOTA_WARN); - bch2_inode_rm(c, inode_inum(inode)); + int ret = bch2_inode_rm(c, inode_inum(inode)); + if (ret && !bch2_err_matches(ret, EROFS)) { + bch_err_msg(c, ret, "VFS incorrectly tried to delete inode %llu:%llu", + inode->ei_inum.subvol, + inode->ei_inum.inum); + bch2_sb_error_count(c, BCH_FSCK_ERR_vfs_bad_inode_rm); + } /* * If we are deleting, we need it present in the vfs hash table @@ -2322,14 +2332,13 @@ static int bch2_show_devname(struct seq_file *seq, struct dentry *root) struct bch_fs *c = root->d_sb->s_fs_info; bool first = true; - rcu_read_lock(); + guard(rcu)(); for_each_online_member_rcu(c, ca) { if (!first) seq_putc(seq, ':'); first = false; seq_puts(seq, ca->disk_sb.sb_name); } - rcu_read_unlock(); return 0; } @@ -2480,6 +2489,14 @@ static int bch2_fs_get_tree(struct fs_context *fc) if (ret) goto err_stop_fs; + /* + * We might be doing a RO mount because other options required it, or we + * have no alloc info and it's a small image with no room to regenerate + * it + */ + if (c->opts.read_only) + fc->sb_flags |= SB_RDONLY; + sb = sget(fc->fs_type, NULL, bch2_set_super, fc->sb_flags|SB_NOSEC, c); ret = PTR_ERR_OR_ZERO(sb); if (ret) @@ -2526,16 +2543,16 @@ got_sb: sb->s_bdi->ra_pages = VM_READAHEAD_PAGES; - rcu_read_lock(); - for_each_online_member_rcu(c, ca) { - struct block_device *bdev = ca->disk_sb.bdev; + scoped_guard(rcu) { + for_each_online_member_rcu(c, ca) { + struct block_device *bdev = ca->disk_sb.bdev; - /* XXX: create an anonymous device for multi device filesystems */ - sb->s_bdev = bdev; - sb->s_dev = bdev->bd_dev; - break; + /* XXX: create an anonymous device for multi device filesystems */ + sb->s_bdev = bdev; + sb->s_dev = bdev->bd_dev; + break; + } } - rcu_read_unlock(); c->dev = sb->s_dev; @@ -2547,9 +2564,10 @@ got_sb: sb->s_shrink->seeks = 0; #ifdef CONFIG_UNICODE - sb->s_encoding = c->cf_encoding; -#endif + if (bch2_fs_casefold_enabled(c)) + sb->s_encoding = c->cf_encoding; generic_set_sb_d_ops(sb); +#endif vinode = bch2_vfs_inode_get(c, BCACHEFS_ROOT_SUBVOL_INUM); ret = PTR_ERR_OR_ZERO(vinode); diff --git a/fs/bcachefs/fsck.c b/fs/bcachefs/fsck.c index 49f46df8340e..856eb2b41896 100644 --- a/fs/bcachefs/fsck.c +++ b/fs/bcachefs/fsck.c @@ -12,6 +12,7 @@ #include "fs.h" #include "fsck.h" #include "inode.h" +#include "io_misc.h" #include "keylist.h" #include "namei.h" #include "recovery_passes.h" @@ -23,14 +24,15 @@ #include <linux/bsearch.h> #include <linux/dcache.h> /* struct qstr */ -static int dirent_points_to_inode_nowarn(struct bkey_s_c_dirent d, +static int dirent_points_to_inode_nowarn(struct bch_fs *c, + struct bkey_s_c_dirent d, struct bch_inode_unpacked *inode) { if (d.v->d_type == DT_SUBVOL ? le32_to_cpu(d.v->d_child_subvol) == inode->bi_subvol : le64_to_cpu(d.v->d_inum) == inode->bi_inum) return 0; - return -BCH_ERR_ENOENT_dirent_doesnt_match_inode; + return bch_err_throw(c, ENOENT_dirent_doesnt_match_inode); } static void dirent_inode_mismatch_msg(struct printbuf *out, @@ -49,7 +51,7 @@ static int dirent_points_to_inode(struct bch_fs *c, struct bkey_s_c_dirent dirent, struct bch_inode_unpacked *inode) { - int ret = dirent_points_to_inode_nowarn(dirent, inode); + int ret = dirent_points_to_inode_nowarn(c, dirent, inode); if (ret) { struct printbuf buf = PRINTBUF; dirent_inode_mismatch_msg(&buf, c, dirent, inode); @@ -152,7 +154,7 @@ static int find_snapshot_tree_subvol(struct btree_trans *trans, goto found; } } - ret = -BCH_ERR_ENOENT_no_snapshot_tree_subvol; + ret = bch_err_throw(trans->c, ENOENT_no_snapshot_tree_subvol); found: bch2_trans_iter_exit(trans, &iter); return ret; @@ -229,7 +231,7 @@ static int lookup_lostfound(struct btree_trans *trans, u32 snapshot, if (d_type != DT_DIR) { bch_err(c, "error looking up lost+found: not a directory"); - return -BCH_ERR_ENOENT_not_directory; + return bch_err_throw(c, ENOENT_not_directory); } /* @@ -326,7 +328,8 @@ static inline bool inode_should_reattach(struct bch_inode_unpacked *inode) (inode->bi_flags & BCH_INODE_has_child_snapshot)) return false; - return !inode->bi_dir && !(inode->bi_flags & BCH_INODE_unlinked); + return !bch2_inode_has_backpointer(inode) && + !(inode->bi_flags & BCH_INODE_unlinked); } static int maybe_delete_dirent(struct btree_trans *trans, struct bpos d_pos, u32 snapshot) @@ -371,6 +374,18 @@ static int reattach_inode(struct btree_trans *trans, struct bch_inode_unpacked * if (inode->bi_subvol) { inode->bi_parent_subvol = BCACHEFS_ROOT_SUBVOL; + struct btree_iter subvol_iter; + struct bkey_i_subvolume *subvol = + bch2_bkey_get_mut_typed(trans, &subvol_iter, + BTREE_ID_subvolumes, POS(0, inode->bi_subvol), + 0, subvolume); + ret = PTR_ERR_OR_ZERO(subvol); + if (ret) + return ret; + + subvol->v.fs_path_parent = BCACHEFS_ROOT_SUBVOL; + bch2_trans_iter_exit(trans, &subvol_iter); + u64 root_inum; ret = subvol_lookup(trans, inode->bi_parent_subvol, &dirent_snapshot, &root_inum); @@ -386,6 +401,8 @@ static int reattach_inode(struct btree_trans *trans, struct bch_inode_unpacked * if (ret) return ret; + bch_verbose(c, "got lostfound inum %llu", lostfound.bi_inum); + lostfound.bi_nlink += S_ISDIR(inode->bi_mode); /* ensure lost+found inode is also present in inode snapshot */ @@ -422,6 +439,16 @@ static int reattach_inode(struct btree_trans *trans, struct bch_inode_unpacked * if (ret) return ret; + { + CLASS(printbuf, buf)(); + ret = bch2_inum_snapshot_to_path(trans, inode->bi_inum, + inode->bi_snapshot, NULL, &buf); + if (ret) + return ret; + + bch_info(c, "reattached at %s", buf.buf); + } + /* * Fix up inodes in child snapshots: if they should also be reattached * update the backpointer field, if they should not be we need to emit @@ -489,13 +516,21 @@ static struct bkey_s_c_dirent dirent_get_by_pos(struct btree_trans *trans, static int remove_backpointer(struct btree_trans *trans, struct bch_inode_unpacked *inode) { - if (!inode->bi_dir) + if (!bch2_inode_has_backpointer(inode)) return 0; + u32 snapshot = inode->bi_snapshot; + + if (inode->bi_parent_subvol) { + int ret = bch2_subvolume_get_snapshot(trans, inode->bi_parent_subvol, &snapshot); + if (ret) + return ret; + } + struct bch_fs *c = trans->c; struct btree_iter iter; struct bkey_s_c_dirent d = dirent_get_by_pos(trans, &iter, - SPOS(inode->bi_dir, inode->bi_dir_offset, inode->bi_snapshot)); + SPOS(inode->bi_dir, inode->bi_dir_offset, snapshot)); int ret = bkey_err(d) ?: dirent_points_to_inode(c, d, inode) ?: bch2_fsck_remove_dirent(trans, d.k->p); @@ -531,7 +566,7 @@ static int reconstruct_subvol(struct btree_trans *trans, u32 snapshotid, u32 sub if (!bch2_snapshot_is_leaf(c, snapshotid)) { bch_err(c, "need to reconstruct subvol, but have interior node snapshot"); - return -BCH_ERR_fsck_repair_unimplemented; + return bch_err_throw(c, fsck_repair_unimplemented); } /* @@ -643,11 +678,6 @@ static int reconstruct_inode(struct btree_trans *trans, enum btree_id btree, u32 return __bch2_fsck_write_inode(trans, &new_inode); } -struct snapshots_seen { - struct bpos pos; - snapshot_id_list ids; -}; - static inline void snapshots_seen_exit(struct snapshots_seen *s) { darray_exit(&s->ids); @@ -699,14 +729,8 @@ static int snapshots_seen_update(struct bch_fs *c, struct snapshots_seen *s, static bool key_visible_in_snapshot(struct bch_fs *c, struct snapshots_seen *seen, u32 id, u32 ancestor) { - ssize_t i; - EBUG_ON(id > ancestor); - /* @ancestor should be the snapshot most recently added to @seen */ - EBUG_ON(ancestor != seen->pos.snapshot); - EBUG_ON(ancestor != darray_last(seen->ids)); - if (id == ancestor) return true; @@ -722,11 +746,8 @@ static bool key_visible_in_snapshot(struct bch_fs *c, struct snapshots_seen *see * numerically, since snapshot ID lists are kept sorted, so if we find * an id that's an ancestor of @id we're done: */ - - for (i = seen->ids.nr - 2; - i >= 0 && seen->ids.data[i] >= id; - --i) - if (bch2_snapshot_is_ancestor(c, id, seen->ids.data[i])) + darray_for_each_reverse(seen->ids, i) + if (*i != ancestor && bch2_snapshot_is_ancestor(c, id, *i)) return false; return true; @@ -810,7 +831,7 @@ static int add_inode(struct bch_fs *c, struct inode_walker *w, if (!n->whiteout) { return bch2_inode_unpack(inode, &n->inode); } else { - n->inode.bi_inum = inode.k->p.inode; + n->inode.bi_inum = inode.k->p.offset; n->inode.bi_snapshot = inode.k->p.snapshot; return 0; } @@ -890,14 +911,11 @@ lookup_inode_for_snapshot(struct btree_trans *trans, struct inode_walker *w, str { struct bch_fs *c = trans->c; - struct inode_walker_entry *i; - __darray_for_each(w->inodes, i) - if (bch2_snapshot_is_ancestor(c, k.k->p.snapshot, i->inode.bi_snapshot)) - goto found; + struct inode_walker_entry *i = darray_find_p(w->inodes, i, + bch2_snapshot_is_ancestor(c, k.k->p.snapshot, i->inode.bi_snapshot)); - return NULL; -found: - BUG_ON(k.k->p.snapshot > i->inode.bi_snapshot); + if (!i) + return NULL; struct printbuf buf = PRINTBUF; int ret = 0; @@ -910,17 +928,15 @@ found: w->last_pos.inode, k.k->p.snapshot, i->inode.bi_snapshot, (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { - struct bch_inode_unpacked new = i->inode; - struct bkey_i whiteout; - - new.bi_snapshot = k.k->p.snapshot; - if (!i->whiteout) { + struct bch_inode_unpacked new = i->inode; + new.bi_snapshot = k.k->p.snapshot; ret = __bch2_fsck_write_inode(trans, &new); } else { + struct bkey_i whiteout; bkey_init(&whiteout.k); whiteout.k.type = KEY_TYPE_whiteout; - whiteout.k.p = SPOS(0, i->inode.bi_inum, i->inode.bi_snapshot); + whiteout.k.p = SPOS(0, i->inode.bi_inum, k.k->p.snapshot); ret = bch2_btree_insert_nonextent(trans, BTREE_ID_inodes, &whiteout, BTREE_UPDATE_internal_snapshot_node); @@ -947,7 +963,7 @@ found: if (ret) goto fsck_err; - ret = -BCH_ERR_transaction_restart_nested; + ret = bch_err_throw(c, transaction_restart_nested); goto fsck_err; } @@ -992,7 +1008,8 @@ int bch2_fsck_update_backpointers(struct btree_trans *trans, int ret = 0; if (d->v.d_type == DT_SUBVOL) { - BUG(); + bch_err(trans->c, "%s does not support DT_SUBVOL", __func__); + ret = -BCH_ERR_fsck_repair_unimplemented; } else { ret = get_visible_inodes(trans, &target, s, le64_to_cpu(d->v.d_inum)); if (ret) @@ -1048,7 +1065,7 @@ static int check_inode_dirent_inode(struct btree_trans *trans, if (ret && !bch2_err_matches(ret, ENOENT)) return ret; - if ((ret || dirent_points_to_inode_nowarn(d, inode)) && + if ((ret || dirent_points_to_inode_nowarn(c, d, inode)) && inode->bi_subvol && (inode->bi_flags & BCH_INODE_has_child_snapshot)) { /* Older version of a renamed subvolume root: we won't have a @@ -1069,7 +1086,7 @@ static int check_inode_dirent_inode(struct btree_trans *trans, trans, inode_points_to_missing_dirent, "inode points to missing dirent\n%s", (bch2_inode_unpacked_to_text(&buf, inode), buf.buf)) || - fsck_err_on(!ret && dirent_points_to_inode_nowarn(d, inode), + fsck_err_on(!ret && dirent_points_to_inode_nowarn(c, d, inode), trans, inode_points_to_wrong_dirent, "%s", (printbuf_reset(&buf), @@ -1141,13 +1158,14 @@ static int check_inode(struct btree_trans *trans, if (ret) goto err; - if (u.bi_dir || u.bi_dir_offset) { + if (bch2_inode_has_backpointer(&u)) { ret = check_inode_dirent_inode(trans, &u, &do_update); if (ret) goto err; } - if (fsck_err_on(u.bi_dir && (u.bi_flags & BCH_INODE_unlinked), + if (fsck_err_on(bch2_inode_has_backpointer(&u) && + (u.bi_flags & BCH_INODE_unlinked), trans, inode_unlinked_but_has_dirent, "inode unlinked but has dirent\n%s", (printbuf_reset(&buf), @@ -1174,6 +1192,14 @@ static int check_inode(struct btree_trans *trans, ret = 0; } + if (fsck_err_on(S_ISDIR(u.bi_mode) && u.bi_size, + trans, inode_dir_has_nonzero_i_size, + "directory %llu:%u with nonzero i_size %lli", + u.bi_inum, u.bi_snapshot, u.bi_size)) { + u.bi_size = 0; + do_update = true; + } + ret = bch2_inode_has_child_snapshots(trans, k.k->p); if (ret < 0) goto err; @@ -1436,6 +1462,7 @@ static int check_key_has_inode(struct btree_trans *trans, { struct bch_fs *c = trans->c; struct printbuf buf = PRINTBUF; + struct btree_iter iter2 = {}; int ret = PTR_ERR_OR_ZERO(i); if (ret) return ret; @@ -1445,40 +1472,105 @@ static int check_key_has_inode(struct btree_trans *trans, bool have_inode = i && !i->whiteout; - if (!have_inode && (c->sb.btrees_lost_data & BIT_ULL(BTREE_ID_inodes))) { - ret = reconstruct_inode(trans, iter->btree_id, k.k->p.snapshot, k.k->p.inode) ?: - bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc); - if (ret) - goto err; + if (!have_inode && (c->sb.btrees_lost_data & BIT_ULL(BTREE_ID_inodes))) + goto reconstruct; - inode->last_pos.inode--; - ret = -BCH_ERR_transaction_restart_nested; - goto err; + if (have_inode && btree_matches_i_mode(iter->btree_id, i->inode.bi_mode)) + goto out; + + prt_printf(&buf, ", "); + + bool have_old_inode = false; + darray_for_each(inode->inodes, i2) + if (!i2->whiteout && + bch2_snapshot_is_ancestor(c, k.k->p.snapshot, i2->inode.bi_snapshot) && + btree_matches_i_mode(iter->btree_id, i2->inode.bi_mode)) { + prt_printf(&buf, "but found good inode in older snapshot\n"); + bch2_inode_unpacked_to_text(&buf, &i2->inode); + prt_newline(&buf); + have_old_inode = true; + break; + } + + struct bkey_s_c k2; + unsigned nr_keys = 0; + + prt_printf(&buf, "found keys:\n"); + + for_each_btree_key_max_norestart(trans, iter2, iter->btree_id, + SPOS(k.k->p.inode, 0, k.k->p.snapshot), + POS(k.k->p.inode, U64_MAX), + 0, k2, ret) { + nr_keys++; + if (nr_keys <= 10) { + bch2_bkey_val_to_text(&buf, c, k2); + prt_newline(&buf); + } + if (nr_keys >= 100) + break; } - if (fsck_err_on(!have_inode, - trans, key_in_missing_inode, - "key in missing inode:\n%s", - (printbuf_reset(&buf), - bch2_bkey_val_to_text(&buf, c, k), buf.buf))) - goto delete; + if (ret) + goto err; - if (fsck_err_on(have_inode && !btree_matches_i_mode(iter->btree_id, i->inode.bi_mode), - trans, key_in_wrong_inode_type, - "key for wrong inode mode %o:\n%s", - i->inode.bi_mode, - (printbuf_reset(&buf), - bch2_bkey_val_to_text(&buf, c, k), buf.buf))) - goto delete; + if (nr_keys > 100) + prt_printf(&buf, "found > %u keys for this missing inode\n", nr_keys); + else if (nr_keys > 10) + prt_printf(&buf, "found %u keys for this missing inode\n", nr_keys); + + if (!have_inode) { + if (fsck_err_on(!have_inode, + trans, key_in_missing_inode, + "key in missing inode%s", buf.buf)) { + /* + * Maybe a deletion that raced with data move, or something + * weird like that? But if we know the inode was deleted, or + * it's just a few keys, we can safely delete them. + * + * If it's many keys, we should probably recreate the inode + */ + if (have_old_inode || nr_keys <= 2) + goto delete; + else + goto reconstruct; + } + } else { + /* + * not autofix, this one would be a giant wtf - bit error in the + * inode corrupting i_mode? + * + * may want to try repairing inode instead of deleting + */ + if (fsck_err_on(!btree_matches_i_mode(iter->btree_id, i->inode.bi_mode), + trans, key_in_wrong_inode_type, + "key for wrong inode mode %o%s", + i->inode.bi_mode, buf.buf)) + goto delete; + } out: err: fsck_err: + bch2_trans_iter_exit(trans, &iter2); printbuf_exit(&buf); bch_err_fn(c, ret); return ret; delete: + /* + * XXX: print out more info + * count up extents for this inode, check if we have different inode in + * an older snapshot version, perhaps decide if we want to reconstitute + */ ret = bch2_btree_delete_at(trans, iter, BTREE_UPDATE_internal_snapshot_node); goto out; +reconstruct: + ret = reconstruct_inode(trans, iter->btree_id, k.k->p.snapshot, k.k->p.inode) ?: + bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc); + if (ret) + goto err; + + inode->last_pos.inode--; + ret = bch_err_throw(c, transaction_restart_nested); + goto out; } static int check_i_sectors_notnested(struct btree_trans *trans, struct inode_walker *w) @@ -1569,7 +1661,7 @@ static int extent_ends_at(struct bch_fs *c, sizeof(seen->ids.data[0]) * seen->ids.size, GFP_KERNEL); if (!n.seen.ids.data) - return -BCH_ERR_ENOMEM_fsck_extent_ends_at; + return bch_err_throw(c, ENOMEM_fsck_extent_ends_at); __darray_for_each(extent_ends->e, i) { if (i->snapshot == k.k->p.snapshot) { @@ -1619,7 +1711,7 @@ static int overlapping_extents_found(struct btree_trans *trans, bch_err(c, "%s: error finding first overlapping extent when repairing, got%s", __func__, buf.buf); - ret = -BCH_ERR_internal_fsck_err; + ret = bch_err_throw(c, internal_fsck_err); goto err; } @@ -1644,7 +1736,7 @@ static int overlapping_extents_found(struct btree_trans *trans, pos2.size != k2.k->size) { bch_err(c, "%s: error finding seconding overlapping extent when repairing%s", __func__, buf.buf); - ret = -BCH_ERR_internal_fsck_err; + ret = bch_err_throw(c, internal_fsck_err); goto err; } @@ -1692,7 +1784,7 @@ static int overlapping_extents_found(struct btree_trans *trans, * We overwrote the second extent - restart * check_extent() from the top: */ - ret = -BCH_ERR_transaction_restart_nested; + ret = bch_err_throw(c, transaction_restart_nested); } } fsck_err: @@ -1820,20 +1912,19 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter, !key_visible_in_snapshot(c, s, i->inode.bi_snapshot, k.k->p.snapshot)) continue; - if (fsck_err_on(k.k->p.offset > round_up(i->inode.bi_size, block_bytes(c)) >> 9 && + u64 last_block = round_up(i->inode.bi_size, block_bytes(c)) >> 9; + + if (fsck_err_on(k.k->p.offset > last_block && !bkey_extent_is_reservation(k), trans, extent_past_end_of_inode, "extent type past end of inode %llu:%u, i_size %llu\n%s", i->inode.bi_inum, i->inode.bi_snapshot, i->inode.bi_size, (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { - struct btree_iter iter2; - - bch2_trans_copy_iter(trans, &iter2, iter); - bch2_btree_iter_set_snapshot(trans, &iter2, i->inode.bi_snapshot); - ret = bch2_btree_iter_traverse(trans, &iter2) ?: - bch2_btree_delete_at(trans, &iter2, - BTREE_UPDATE_internal_snapshot_node); - bch2_trans_iter_exit(trans, &iter2); + ret = bch2_fpunch_snapshot(trans, + SPOS(i->inode.bi_inum, + last_block, + i->inode.bi_snapshot), + POS(i->inode.bi_inum, U64_MAX)); if (ret) goto err; @@ -1947,14 +2038,22 @@ static int check_subdir_count_notnested(struct btree_trans *trans, struct inode_ continue; } - if (fsck_err_on(i->inode.bi_nlink != i->count, - trans, inode_dir_wrong_nlink, - "directory %llu:%u with wrong i_nlink: got %u, should be %llu", - w->last_pos.inode, i->inode.bi_snapshot, i->inode.bi_nlink, i->count)) { - i->inode.bi_nlink = i->count; - ret = bch2_fsck_write_inode(trans, &i->inode); - if (ret) - break; + if (i->inode.bi_nlink != i->count) { + CLASS(printbuf, buf)(); + + lockrestart_do(trans, + bch2_inum_snapshot_to_path(trans, w->last_pos.inode, + i->inode.bi_snapshot, NULL, &buf)); + + if (fsck_err_on(i->inode.bi_nlink != i->count, + trans, inode_dir_wrong_nlink, + "directory with wrong i_nlink: got %u, should be %llu\n%s", + i->inode.bi_nlink, i->count, buf.buf)) { + i->inode.bi_nlink = i->count; + ret = bch2_fsck_write_inode(trans, &i->inode); + if (ret) + break; + } } } fsck_err: @@ -2045,7 +2144,7 @@ static int check_dirent_to_subvol(struct btree_trans *trans, struct btree_iter * (bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf))) { if (!new_parent_subvol) { bch_err(c, "could not find a subvol for snapshot %u", d.k->p.snapshot); - return -BCH_ERR_fsck_repair_unimplemented; + return bch_err_throw(c, fsck_repair_unimplemented); } struct bkey_i_dirent *new_dirent = bch2_bkey_make_mut_typed(trans, iter, &d.s_c, 0, dirent); @@ -2107,7 +2206,7 @@ static int check_dirent_to_subvol(struct btree_trans *trans, struct btree_iter * if (ret) { bch_err(c, "subvol %u points to missing inode root %llu", target_subvol, target_inum); - ret = -BCH_ERR_fsck_repair_unimplemented; + ret = bch_err_throw(c, fsck_repair_unimplemented); goto err; } @@ -2139,7 +2238,8 @@ static int check_dirent(struct btree_trans *trans, struct btree_iter *iter, struct bch_hash_info *hash_info, struct inode_walker *dir, struct inode_walker *target, - struct snapshots_seen *s) + struct snapshots_seen *s, + bool *need_second_pass) { struct bch_fs *c = trans->c; struct inode_walker_entry *i; @@ -2181,7 +2281,10 @@ static int check_dirent(struct btree_trans *trans, struct btree_iter *iter, *hash_info = bch2_hash_info_init(c, &i->inode); dir->first_this_inode = false; - ret = bch2_str_hash_check_key(trans, s, &bch2_dirent_hash_desc, hash_info, iter, k); + hash_info->cf_encoding = bch2_inode_casefold(c, &i->inode) ? c->cf_encoding : NULL; + + ret = bch2_str_hash_check_key(trans, s, &bch2_dirent_hash_desc, hash_info, + iter, k, need_second_pass); if (ret < 0) goto err; if (ret) { @@ -2202,31 +2305,34 @@ static int check_dirent(struct btree_trans *trans, struct btree_iter *iter, (printbuf_reset(&buf), bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { - struct qstr name = bch2_dirent_get_name(d); - u32 subvol = d.v->d_type == DT_SUBVOL - ? le32_to_cpu(d.v->d_parent_subvol) - : 0; + subvol_inum dir_inum = { .subvol = d.v->d_type == DT_SUBVOL + ? le32_to_cpu(d.v->d_parent_subvol) + : 0, + }; u64 target = d.v->d_type == DT_SUBVOL ? le32_to_cpu(d.v->d_child_subvol) : le64_to_cpu(d.v->d_inum); - u64 dir_offset; + struct qstr name = bch2_dirent_get_name(d); + + struct bkey_i_dirent *new_d = + bch2_dirent_create_key(trans, hash_info, dir_inum, + d.v->d_type, &name, NULL, target); + ret = PTR_ERR_OR_ZERO(new_d); + if (ret) + goto out; + + new_d->k.p.inode = d.k->p.inode; + new_d->k.p.snapshot = d.k->p.snapshot; - ret = bch2_hash_delete_at(trans, + struct btree_iter dup_iter = {}; + ret = bch2_hash_delete_at(trans, bch2_dirent_hash_desc, hash_info, iter, BTREE_UPDATE_internal_snapshot_node) ?: - bch2_dirent_create_snapshot(trans, subvol, - d.k->p.inode, d.k->p.snapshot, - hash_info, - d.v->d_type, - &name, - target, - &dir_offset, - BTREE_ITER_with_updates| - BTREE_UPDATE_internal_snapshot_node| - STR_HASH_must_create) ?: - bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc); - - /* might need another check_dirents pass */ + bch2_str_hash_repair_key(trans, s, + &bch2_dirent_hash_desc, hash_info, + iter, bkey_i_to_s_c(&new_d->k_i), + &dup_iter, bkey_s_c_null, + need_second_pass); goto out; } @@ -2294,7 +2400,6 @@ out: err: fsck_err: printbuf_exit(&buf); - bch_err_fn(c, ret); return ret; } @@ -2308,16 +2413,31 @@ int bch2_check_dirents(struct bch_fs *c) struct inode_walker target = inode_walker_init(); struct snapshots_seen s; struct bch_hash_info hash_info; + bool need_second_pass = false, did_second_pass = false; + int ret; snapshots_seen_init(&s); - - int ret = bch2_trans_run(c, - for_each_btree_key(trans, iter, BTREE_ID_dirents, +again: + ret = bch2_trans_run(c, + for_each_btree_key_commit(trans, iter, BTREE_ID_dirents, POS(BCACHEFS_ROOT_INO, 0), BTREE_ITER_prefetch|BTREE_ITER_all_snapshots, k, - check_dirent(trans, &iter, k, &hash_info, &dir, &target, &s)) ?: + NULL, NULL, BCH_TRANS_COMMIT_no_enospc, + check_dirent(trans, &iter, k, &hash_info, &dir, &target, &s, + &need_second_pass)) ?: check_subdir_count_notnested(trans, &dir)); + if (!ret && need_second_pass && !did_second_pass) { + bch_info(c, "check_dirents requires second pass"); + swap(did_second_pass, need_second_pass); + goto again; + } + + if (!ret && need_second_pass) { + bch_err(c, "dirents not repairing"); + ret = -EINVAL; + } + snapshots_seen_exit(&s); inode_walker_exit(&dir); inode_walker_exit(&target); @@ -2331,16 +2451,14 @@ static int check_xattr(struct btree_trans *trans, struct btree_iter *iter, struct inode_walker *inode) { struct bch_fs *c = trans->c; - struct inode_walker_entry *i; - int ret; - ret = bch2_check_key_has_snapshot(trans, iter, k); + int ret = bch2_check_key_has_snapshot(trans, iter, k); if (ret < 0) return ret; if (ret) return 0; - i = walk_inode(trans, inode, k); + struct inode_walker_entry *i = walk_inode(trans, inode, k); ret = PTR_ERR_OR_ZERO(i); if (ret) return ret; @@ -2356,9 +2474,9 @@ static int check_xattr(struct btree_trans *trans, struct btree_iter *iter, *hash_info = bch2_hash_info_init(c, &i->inode); inode->first_this_inode = false; - ret = bch2_str_hash_check_key(trans, NULL, &bch2_xattr_hash_desc, hash_info, iter, k); - bch_err_fn(c, ret); - return ret; + bool need_second_pass = false; + return bch2_str_hash_check_key(trans, NULL, &bch2_xattr_hash_desc, hash_info, + iter, k, &need_second_pass); } /* @@ -2470,6 +2588,11 @@ static int check_subvol_path(struct btree_trans *trans, struct btree_iter *iter, if (k.k->type != KEY_TYPE_subvolume) return 0; + subvol_inum start = { + .subvol = k.k->p.offset, + .inum = le64_to_cpu(bkey_s_c_to_subvolume(k).v->inode), + }; + while (k.k->p.offset != BCACHEFS_ROOT_SUBVOL) { ret = darray_push(&subvol_path, k.k->p.offset); if (ret) @@ -2488,11 +2611,11 @@ static int check_subvol_path(struct btree_trans *trans, struct btree_iter *iter, if (darray_u32_has(&subvol_path, parent)) { printbuf_reset(&buf); - prt_printf(&buf, "subvolume loop:\n"); + prt_printf(&buf, "subvolume loop: "); - darray_for_each_reverse(subvol_path, i) - prt_printf(&buf, "%u ", *i); - prt_printf(&buf, "%u", parent); + ret = bch2_inum_to_path(trans, start, &buf); + if (ret) + goto err; if (fsck_err(trans, subvol_loop, "%s", buf.buf)) ret = reattach_subvol(trans, s); @@ -2536,19 +2659,13 @@ int bch2_check_subvolume_structure(struct bch_fs *c) return ret; } -struct pathbuf_entry { - u64 inum; - u32 snapshot; -}; - -typedef DARRAY(struct pathbuf_entry) pathbuf; - -static int bch2_bi_depth_renumber_one(struct btree_trans *trans, struct pathbuf_entry *p, +static int bch2_bi_depth_renumber_one(struct btree_trans *trans, + u64 inum, u32 snapshot, u32 new_depth) { struct btree_iter iter; struct bkey_s_c k = bch2_bkey_get_iter(trans, &iter, BTREE_ID_inodes, - SPOS(0, p->inum, p->snapshot), 0); + SPOS(0, inum, snapshot), 0); struct bch_inode_unpacked inode; int ret = bkey_err(k) ?: @@ -2567,14 +2684,15 @@ err: return ret; } -static int bch2_bi_depth_renumber(struct btree_trans *trans, pathbuf *path, u32 new_bi_depth) +static int bch2_bi_depth_renumber(struct btree_trans *trans, darray_u64 *path, + u32 snapshot, u32 new_bi_depth) { u32 restart_count = trans->restart_count; int ret = 0; darray_for_each_reverse(*path, i) { ret = nested_lockrestart_do(trans, - bch2_bi_depth_renumber_one(trans, i, new_bi_depth)); + bch2_bi_depth_renumber_one(trans, *i, snapshot, new_bi_depth)); bch_err_fn(trans->c, ret); if (ret) break; @@ -2585,37 +2703,36 @@ static int bch2_bi_depth_renumber(struct btree_trans *trans, pathbuf *path, u32 return ret ?: trans_was_restarted(trans, restart_count); } -static bool path_is_dup(pathbuf *p, u64 inum, u32 snapshot) -{ - darray_for_each(*p, i) - if (i->inum == inum && - i->snapshot == snapshot) - return true; - return false; -} - static int check_path_loop(struct btree_trans *trans, struct bkey_s_c inode_k) { struct bch_fs *c = trans->c; struct btree_iter inode_iter = {}; - pathbuf path = {}; + darray_u64 path = {}; struct printbuf buf = PRINTBUF; u32 snapshot = inode_k.k->p.snapshot; bool redo_bi_depth = false; u32 min_bi_depth = U32_MAX; int ret = 0; + struct bpos start = inode_k.k->p; + struct bch_inode_unpacked inode; ret = bch2_inode_unpack(inode_k, &inode); if (ret) return ret; - while (!inode.bi_subvol) { + /* + * If we're running full fsck, check_dirents() will have already ran, + * and we shouldn't see any missing backpointers here - otherwise that's + * handled separately, by check_unreachable_inodes + */ + while (!inode.bi_subvol && + bch2_inode_has_backpointer(&inode)) { struct btree_iter dirent_iter; struct bkey_s_c_dirent d; - u32 parent_snapshot = snapshot; - d = inode_get_dirent(trans, &dirent_iter, &inode, &parent_snapshot); + d = dirent_get_by_pos(trans, &dirent_iter, + SPOS(inode.bi_dir, inode.bi_dir_offset, snapshot)); ret = bkey_err(d.s_c); if (ret && !bch2_err_matches(ret, ENOENT)) goto out; @@ -2633,15 +2750,10 @@ static int check_path_loop(struct btree_trans *trans, struct bkey_s_c inode_k) bch2_trans_iter_exit(trans, &dirent_iter); - ret = darray_push(&path, ((struct pathbuf_entry) { - .inum = inode.bi_inum, - .snapshot = snapshot, - })); + ret = darray_push(&path, inode.bi_inum); if (ret) return ret; - snapshot = parent_snapshot; - bch2_trans_iter_exit(trans, &inode_iter); inode_k = bch2_bkey_get_iter(trans, &inode_iter, BTREE_ID_inodes, SPOS(0, inode.bi_dir, snapshot), 0); @@ -2663,21 +2775,28 @@ static int check_path_loop(struct btree_trans *trans, struct bkey_s_c inode_k) break; inode = parent_inode; - snapshot = inode_k.k->p.snapshot; redo_bi_depth = true; - if (path_is_dup(&path, inode.bi_inum, snapshot)) { + if (darray_find(path, inode.bi_inum)) { printbuf_reset(&buf); - prt_printf(&buf, "directory structure loop:\n"); - darray_for_each_reverse(path, i) - prt_printf(&buf, "%llu:%u ", i->inum, i->snapshot); - prt_printf(&buf, "%llu:%u", inode.bi_inum, snapshot); + prt_printf(&buf, "directory structure loop in snapshot %u: ", + snapshot); + + ret = bch2_inum_snapshot_to_path(trans, start.offset, start.snapshot, NULL, &buf); + if (ret) + goto out; + + if (c->opts.verbose) { + prt_newline(&buf); + darray_for_each(path, i) + prt_printf(&buf, "%llu ", *i); + } if (fsck_err(trans, dir_loop, "%s", buf.buf)) { ret = remove_backpointer(trans, &inode); bch_err_msg(c, ret, "removing dirent"); if (ret) - break; + goto out; ret = reattach_inode(trans, &inode); bch_err_msg(c, ret, "reattaching inode %llu", inode.bi_inum); @@ -2691,7 +2810,7 @@ static int check_path_loop(struct btree_trans *trans, struct bkey_s_c inode_k) min_bi_depth = 0; if (redo_bi_depth) - ret = bch2_bi_depth_renumber(trans, &path, min_bi_depth); + ret = bch2_bi_depth_renumber(trans, &path, snapshot, min_bi_depth); out: fsck_err: bch2_trans_iter_exit(trans, &inode_iter); @@ -2708,7 +2827,7 @@ fsck_err: int bch2_check_directory_structure(struct bch_fs *c) { int ret = bch2_trans_run(c, - for_each_btree_key_commit(trans, iter, BTREE_ID_inodes, POS_MIN, + for_each_btree_key_reverse_commit(trans, iter, BTREE_ID_inodes, POS_MIN, BTREE_ITER_intent| BTREE_ITER_prefetch| BTREE_ITER_all_snapshots, k, @@ -2747,7 +2866,7 @@ static int add_nlink(struct bch_fs *c, struct nlink_table *t, if (!d) { bch_err(c, "fsck: error allocating memory for nlink_table, size %zu", new_size); - return -BCH_ERR_ENOMEM_fsck_add_nlink; + return bch_err_throw(c, ENOMEM_fsck_add_nlink); } if (t->d) diff --git a/fs/bcachefs/fsck.h b/fs/bcachefs/fsck.h index 574948278cd4..e5fe7cf7b251 100644 --- a/fs/bcachefs/fsck.h +++ b/fs/bcachefs/fsck.h @@ -4,6 +4,12 @@ #include "str_hash.h" +/* recoverds snapshot IDs of overwrites at @pos */ +struct snapshots_seen { + struct bpos pos; + snapshot_id_list ids; +}; + int bch2_fsck_update_backpointers(struct btree_trans *, struct snapshots_seen *, const struct bch_hash_desc, diff --git a/fs/bcachefs/inode.c b/fs/bcachefs/inode.c index 5cf70108ae2f..ef4cc7395b86 100644 --- a/fs/bcachefs/inode.c +++ b/fs/bcachefs/inode.c @@ -38,6 +38,7 @@ static const char * const bch2_inode_flag_strs[] = { #undef x static int delete_ancestor_snapshot_inodes(struct btree_trans *, struct bpos); +static int may_delete_deleted_inum(struct btree_trans *, subvol_inum); static const u8 byte_table[8] = { 1, 2, 3, 4, 6, 8, 10, 13 }; @@ -1041,7 +1042,7 @@ again: goto found_slot; if (!ret && start == min) - ret = -BCH_ERR_ENOSPC_inode_create; + ret = bch_err_throw(trans->c, ENOSPC_inode_create); if (ret) { bch2_trans_iter_exit(trans, iter); @@ -1130,19 +1131,23 @@ int bch2_inode_rm(struct bch_fs *c, subvol_inum inum) u32 snapshot; int ret; + ret = lockrestart_do(trans, may_delete_deleted_inum(trans, inum)); + if (ret) + goto err2; + /* * If this was a directory, there shouldn't be any real dirents left - * but there could be whiteouts (from hash collisions) that we should * delete: * - * XXX: the dirent could ideally would delete whiteouts when they're no + * XXX: the dirent code ideally would delete whiteouts when they're no * longer needed */ ret = bch2_inode_delete_keys(trans, inum, BTREE_ID_extents) ?: bch2_inode_delete_keys(trans, inum, BTREE_ID_xattrs) ?: bch2_inode_delete_keys(trans, inum, BTREE_ID_dirents); if (ret) - goto err; + goto err2; retry: bch2_trans_begin(trans); @@ -1161,7 +1166,7 @@ retry: bch2_fs_inconsistent(c, "inode %llu:%u not found when deleting", inum.inum, snapshot); - ret = -BCH_ERR_ENOENT_inode; + ret = bch_err_throw(c, ENOENT_inode); goto err; } @@ -1260,7 +1265,14 @@ int bch2_inode_set_casefold(struct btree_trans *trans, subvol_inum inum, { struct bch_fs *c = trans->c; -#ifdef CONFIG_UNICODE +#ifndef CONFIG_UNICODE + bch_err(c, "Cannot use casefolding on a kernel without CONFIG_UNICODE"); + return -EOPNOTSUPP; +#endif + + if (c->opts.casefold_disabled) + return -EOPNOTSUPP; + int ret = 0; /* Not supported on individual files. */ if (!S_ISDIR(bi->bi_mode)) @@ -1284,10 +1296,6 @@ int bch2_inode_set_casefold(struct btree_trans *trans, subvol_inum inum, bi->bi_fields_set |= BIT(Inode_opt_casefold); return bch2_maybe_propagate_has_case_insensitive(trans, inum, bi); -#else - bch_err(c, "Cannot use casefolding on a kernel without CONFIG_UNICODE"); - return -EOPNOTSUPP; -#endif } static noinline int __bch2_inode_rm_snapshot(struct btree_trans *trans, u64 inum, u32 snapshot) @@ -1328,7 +1336,7 @@ retry: bch2_fs_inconsistent(c, "inode %llu:%u not found when deleting", inum, snapshot); - ret = -BCH_ERR_ENOENT_inode; + ret = bch_err_throw(c, ENOENT_inode); goto err; } @@ -1392,10 +1400,8 @@ int bch2_inode_rm_snapshot(struct btree_trans *trans, u64 inum, u32 snapshot) delete_ancestor_snapshot_inodes(trans, SPOS(0, inum, snapshot)); } -static int may_delete_deleted_inode(struct btree_trans *trans, - struct btree_iter *iter, - struct bpos pos, - bool *need_another_pass) +static int may_delete_deleted_inode(struct btree_trans *trans, struct bpos pos, + bool from_deleted_inodes) { struct bch_fs *c = trans->c; struct btree_iter inode_iter; @@ -1409,12 +1415,14 @@ static int may_delete_deleted_inode(struct btree_trans *trans, if (ret) return ret; - ret = bkey_is_inode(k.k) ? 0 : -BCH_ERR_ENOENT_inode; - if (fsck_err_on(!bkey_is_inode(k.k), + ret = bkey_is_inode(k.k) ? 0 : bch_err_throw(c, ENOENT_inode); + if (fsck_err_on(from_deleted_inodes && ret, trans, deleted_inode_missing, "nonexistent inode %llu:%u in deleted_inodes btree", pos.offset, pos.snapshot)) goto delete; + if (ret) + goto out; ret = bch2_inode_unpack(k, &inode); if (ret) @@ -1422,7 +1430,8 @@ static int may_delete_deleted_inode(struct btree_trans *trans, if (S_ISDIR(inode.bi_mode)) { ret = bch2_empty_dir_snapshot(trans, pos.offset, 0, pos.snapshot); - if (fsck_err_on(bch2_err_matches(ret, ENOTEMPTY), + if (fsck_err_on(from_deleted_inodes && + bch2_err_matches(ret, ENOTEMPTY), trans, deleted_inode_is_dir, "non empty directory %llu:%u in deleted_inodes btree", pos.offset, pos.snapshot)) @@ -1431,17 +1440,25 @@ static int may_delete_deleted_inode(struct btree_trans *trans, goto out; } - if (fsck_err_on(!(inode.bi_flags & BCH_INODE_unlinked), + ret = inode.bi_flags & BCH_INODE_unlinked ? 0 : bch_err_throw(c, inode_not_unlinked); + if (fsck_err_on(from_deleted_inodes && ret, trans, deleted_inode_not_unlinked, "non-deleted inode %llu:%u in deleted_inodes btree", pos.offset, pos.snapshot)) goto delete; + if (ret) + goto out; + + ret = !(inode.bi_flags & BCH_INODE_has_child_snapshot) + ? 0 : bch_err_throw(c, inode_has_child_snapshot); - if (fsck_err_on(inode.bi_flags & BCH_INODE_has_child_snapshot, + if (fsck_err_on(from_deleted_inodes && ret, trans, deleted_inode_has_child_snapshots, "inode with child snapshots %llu:%u in deleted_inodes btree", pos.offset, pos.snapshot)) goto delete; + if (ret) + goto out; ret = bch2_inode_has_child_snapshots(trans, k.k->p); if (ret < 0) @@ -1458,19 +1475,28 @@ static int may_delete_deleted_inode(struct btree_trans *trans, if (ret) goto out; } + + if (!from_deleted_inodes) { + ret = bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc) ?: + bch_err_throw(c, inode_has_child_snapshot); + goto out; + } + goto delete; } - if (test_bit(BCH_FS_clean_recovery, &c->flags) && - !fsck_err(trans, deleted_inode_but_clean, - "filesystem marked as clean but have deleted inode %llu:%u", - pos.offset, pos.snapshot)) { - ret = 0; - goto out; - } + if (from_deleted_inodes) { + if (test_bit(BCH_FS_clean_recovery, &c->flags) && + !fsck_err(trans, deleted_inode_but_clean, + "filesystem marked as clean but have deleted inode %llu:%u", + pos.offset, pos.snapshot)) { + ret = 0; + goto out; + } - ret = 1; + ret = 1; + } out: fsck_err: bch2_trans_iter_exit(trans, &inode_iter); @@ -1481,12 +1507,19 @@ delete: goto out; } +static int may_delete_deleted_inum(struct btree_trans *trans, subvol_inum inum) +{ + u32 snapshot; + + return bch2_subvolume_get_snapshot(trans, inum.subvol, &snapshot) ?: + may_delete_deleted_inode(trans, SPOS(0, inum.inum, snapshot), false); +} + int bch2_delete_dead_inodes(struct bch_fs *c) { struct btree_trans *trans = bch2_trans_get(c); - bool need_another_pass; int ret; -again: + /* * if we ran check_inodes() unlinked inodes will have already been * cleaned up but the write buffer will be out of sync; therefore we @@ -1496,8 +1529,6 @@ again: if (ret) goto err; - need_another_pass = false; - /* * Weird transaction restart handling here because on successful delete, * bch2_inode_rm_snapshot() will return a nested transaction restart, @@ -1507,7 +1538,7 @@ again: ret = for_each_btree_key_commit(trans, iter, BTREE_ID_deleted_inodes, POS_MIN, BTREE_ITER_prefetch|BTREE_ITER_all_snapshots, k, NULL, NULL, BCH_TRANS_COMMIT_no_enospc, ({ - ret = may_delete_deleted_inode(trans, &iter, k.k->p, &need_another_pass); + ret = may_delete_deleted_inode(trans, k.k->p, true); if (ret > 0) { bch_verbose_ratelimited(c, "deleting unlinked inode %llu:%u", k.k->p.offset, k.k->p.snapshot); @@ -1528,10 +1559,8 @@ again: ret; })); - - if (!ret && need_another_pass) - goto again; err: bch2_trans_put(trans); + bch_err_fn(c, ret); return ret; } diff --git a/fs/bcachefs/inode.h b/fs/bcachefs/inode.h index 77ad2d549541..b8ec3e628d90 100644 --- a/fs/bcachefs/inode.h +++ b/fs/bcachefs/inode.h @@ -254,6 +254,11 @@ static inline bool bch2_inode_casefold(struct bch_fs *c, const struct bch_inode_ : c->opts.casefold; } +static inline bool bch2_inode_has_backpointer(const struct bch_inode_unpacked *bi) +{ + return bi->bi_dir || bi->bi_dir_offset; +} + /* i_nlink: */ static inline unsigned nlink_bias(umode_t mode) @@ -283,15 +288,6 @@ static inline void bch2_inode_nlink_set(struct bch_inode_unpacked *bi, int bch2_inode_nlink_inc(struct bch_inode_unpacked *); void bch2_inode_nlink_dec(struct btree_trans *, struct bch_inode_unpacked *); -static inline bool bch2_inode_should_have_single_bp(struct bch_inode_unpacked *inode) -{ - bool inode_has_bp = inode->bi_dir || inode->bi_dir_offset; - - return S_ISDIR(inode->bi_mode) || - inode->bi_subvol || - (!inode->bi_nlink && inode_has_bp); -} - struct bch_opts bch2_inode_opts_to_opts(struct bch_inode_unpacked *); void bch2_inode_opts_get(struct bch_io_opts *, struct bch_fs *, struct bch_inode_unpacked *); diff --git a/fs/bcachefs/io_misc.c b/fs/bcachefs/io_misc.c index cc07729a4b62..07023667a475 100644 --- a/fs/bcachefs/io_misc.c +++ b/fs/bcachefs/io_misc.c @@ -91,7 +91,7 @@ int bch2_extent_fallocate(struct btree_trans *trans, opts.data_replicas, BCH_WATERMARK_normal, 0, &cl, &wp); if (bch2_err_matches(ret, BCH_ERR_operation_blocked)) - ret = -BCH_ERR_transaction_restart_nested; + ret = bch_err_throw(c, transaction_restart_nested); if (ret) goto err; @@ -135,6 +135,33 @@ err_noprint: return ret; } +/* For fsck */ +int bch2_fpunch_snapshot(struct btree_trans *trans, struct bpos start, struct bpos end) +{ + u32 restart_count = trans->restart_count; + struct bch_fs *c = trans->c; + struct disk_reservation disk_res = bch2_disk_reservation_init(c, 0); + unsigned max_sectors = KEY_SIZE_MAX & (~0 << c->block_bits); + struct bkey_i delete; + + int ret = for_each_btree_key_max_commit(trans, iter, BTREE_ID_extents, + start, end, 0, k, + &disk_res, NULL, BCH_TRANS_COMMIT_no_enospc, ({ + bkey_init(&delete.k); + delete.k.p = iter.pos; + + /* create the biggest key we can */ + bch2_key_resize(&delete.k, max_sectors); + bch2_cut_back(end, &delete); + + bch2_extent_trim_atomic(trans, &iter, &delete) ?: + bch2_trans_update(trans, &iter, &delete, 0); + })); + + bch2_disk_reservation_put(c, &disk_res); + return ret ?: trans_was_restarted(trans, restart_count); +} + /* * Returns -BCH_ERR_transacton_restart if we had to drop locks: */ diff --git a/fs/bcachefs/io_misc.h b/fs/bcachefs/io_misc.h index 9cb44a7c43c1..b93e4d4b3c0c 100644 --- a/fs/bcachefs/io_misc.h +++ b/fs/bcachefs/io_misc.h @@ -5,6 +5,8 @@ int bch2_extent_fallocate(struct btree_trans *, subvol_inum, struct btree_iter *, u64, struct bch_io_opts, s64 *, struct write_point_specifier); + +int bch2_fpunch_snapshot(struct btree_trans *, struct bpos, struct bpos); int bch2_fpunch_at(struct btree_trans *, struct btree_iter *, subvol_inum, u64, s64 *); int bch2_fpunch(struct bch_fs *c, subvol_inum, u64, u64, s64 *); diff --git a/fs/bcachefs/io_read.c b/fs/bcachefs/io_read.c index cc708d46557e..e0874ad9a6cf 100644 --- a/fs/bcachefs/io_read.c +++ b/fs/bcachefs/io_read.c @@ -56,7 +56,7 @@ static bool bch2_target_congested(struct bch_fs *c, u16 target) if (!target) return false; - rcu_read_lock(); + guard(rcu)(); devs = bch2_target_to_mask(c, target) ?: &c->rw_devs[BCH_DATA_user]; @@ -73,7 +73,6 @@ static bool bch2_target_congested(struct bch_fs *c, u16 target) total += max(congested, 0LL); nr++; } - rcu_read_unlock(); return get_random_u32_below(nr * CONGESTED_MAX) < total; } @@ -138,21 +137,21 @@ static inline int should_promote(struct bch_fs *c, struct bkey_s_c k, BUG_ON(!opts.promote_target); if (!(flags & BCH_READ_may_promote)) - return -BCH_ERR_nopromote_may_not; + return bch_err_throw(c, nopromote_may_not); if (bch2_bkey_has_target(c, k, opts.promote_target)) - return -BCH_ERR_nopromote_already_promoted; + return bch_err_throw(c, nopromote_already_promoted); if (bkey_extent_is_unwritten(k)) - return -BCH_ERR_nopromote_unwritten; + return bch_err_throw(c, nopromote_unwritten); if (bch2_target_congested(c, opts.promote_target)) - return -BCH_ERR_nopromote_congested; + return bch_err_throw(c, nopromote_congested); } if (rhashtable_lookup_fast(&c->promote_table, &pos, bch_promote_params)) - return -BCH_ERR_nopromote_in_flight; + return bch_err_throw(c, nopromote_in_flight); return 0; } @@ -167,6 +166,7 @@ static noinline void promote_free(struct bch_read_bio *rbio) BUG_ON(ret); async_object_list_del(c, promote, op->list_idx); + async_object_list_del(c, rbio, rbio->list_idx); bch2_data_update_exit(&op->write); @@ -240,7 +240,7 @@ static struct bch_read_bio *__promote_alloc(struct btree_trans *trans, struct promote_op *op = kzalloc(sizeof(*op), GFP_KERNEL); if (!op) { - ret = -BCH_ERR_nopromote_enomem; + ret = bch_err_throw(c, nopromote_enomem); goto err_put; } @@ -249,7 +249,7 @@ static struct bch_read_bio *__promote_alloc(struct btree_trans *trans, if (rhashtable_lookup_insert_fast(&c->promote_table, &op->hash, bch_promote_params)) { - ret = -BCH_ERR_nopromote_in_flight; + ret = bch_err_throw(c, nopromote_in_flight); goto err; } @@ -344,6 +344,10 @@ static struct bch_read_bio *promote_alloc(struct btree_trans *trans, *bounce = true; *read_full = promote_full; + + if (have_io_error(failed)) + orig->self_healing = true; + return promote; nopromote: trace_io_read_nopromote(c, ret); @@ -453,6 +457,10 @@ static void bch2_rbio_done(struct bch_read_bio *rbio) if (rbio->start_time) bch2_time_stats_update(&rbio->c->times[BCH_TIME_data_read], rbio->start_time); +#ifdef CONFIG_BCACHEFS_ASYNC_OBJECT_LISTS + if (rbio->list_idx) + async_object_list_del(rbio->c, rbio, rbio->list_idx); +#endif bio_endio(&rbio->bio); } @@ -545,7 +553,7 @@ retry: if (!bkey_and_val_eq(k, bkey_i_to_s_c(u->k.k))) { /* extent we wanted to read no longer exists: */ - rbio->ret = -BCH_ERR_data_read_key_overwritten; + rbio->ret = bch_err_throw(trans->c, data_read_key_overwritten); goto err; } @@ -636,12 +644,15 @@ static void bch2_rbio_retry(struct work_struct *work) prt_str(&buf, "(internal move) "); prt_str(&buf, "data read error, "); - if (!ret) + if (!ret) { prt_str(&buf, "successful retry"); - else + if (rbio->self_healing) + prt_str(&buf, ", self healing"); + } else prt_str(&buf, bch2_err_str(ret)); prt_newline(&buf); + if (!bkey_deleted(&sk.k->k)) { bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(sk.k)); prt_newline(&buf); @@ -1036,7 +1047,7 @@ int __bch2_read_extent(struct btree_trans *trans, struct bch_read_bio *orig, if ((bch2_bkey_extent_flags(k) & BIT_ULL(BCH_EXTENT_FLAG_poisoned)) && !orig->data_update) - return -BCH_ERR_extent_poisoned; + return bch_err_throw(c, extent_poisoned); retry_pick: ret = bch2_bkey_pick_read_device(c, k, failed, &pick, dev); @@ -1074,7 +1085,7 @@ retry_pick: bch_err_ratelimited(c, "%s", buf.buf); printbuf_exit(&buf); - ret = -BCH_ERR_data_read_no_encryption_key; + ret = bch_err_throw(c, data_read_no_encryption_key); goto err; } @@ -1128,7 +1139,7 @@ retry_pick: if (ca) enumerated_ref_put(&ca->io_ref[READ], BCH_DEV_READ_REF_io_read); - rbio->ret = -BCH_ERR_data_read_buffer_too_small; + rbio->ret = bch_err_throw(c, data_read_buffer_too_small); goto out_read_done; } @@ -1333,7 +1344,7 @@ hole: * have to signal that: */ if (u) - orig->ret = -BCH_ERR_data_read_key_overwritten; + orig->ret = bch_err_throw(c, data_read_key_overwritten); zero_fill_bio_iter(&orig->bio, iter); out_read_done: @@ -1485,7 +1496,12 @@ void bch2_read_bio_to_text(struct printbuf *out, struct bch_read_bio *rbio) prt_printf(out, "have_ioref:\t%u\n", rbio->have_ioref); prt_printf(out, "narrow_crcs:\t%u\n", rbio->narrow_crcs); prt_printf(out, "context:\t%u\n", rbio->context); - prt_printf(out, "ret:\t%s\n", bch2_err_str(rbio->ret)); + + int ret = READ_ONCE(rbio->ret); + if (ret < 0) + prt_printf(out, "ret:\t%s\n", bch2_err_str(ret)); + else + prt_printf(out, "ret:\t%i\n", ret); prt_printf(out, "flags:\t"); bch2_prt_bitflags(out, bch2_read_bio_flags, rbio->flags); @@ -1510,18 +1526,18 @@ int bch2_fs_io_read_init(struct bch_fs *c) c->opts.btree_node_size, c->opts.encoded_extent_max) / PAGE_SIZE, 0)) - return -BCH_ERR_ENOMEM_bio_bounce_pages_init; + return bch_err_throw(c, ENOMEM_bio_bounce_pages_init); if (bioset_init(&c->bio_read, 1, offsetof(struct bch_read_bio, bio), BIOSET_NEED_BVECS)) - return -BCH_ERR_ENOMEM_bio_read_init; + return bch_err_throw(c, ENOMEM_bio_read_init); if (bioset_init(&c->bio_read_split, 1, offsetof(struct bch_read_bio, bio), BIOSET_NEED_BVECS)) - return -BCH_ERR_ENOMEM_bio_read_split_init; + return bch_err_throw(c, ENOMEM_bio_read_split_init); if (rhashtable_init(&c->promote_table, &bch_promote_params)) - return -BCH_ERR_ENOMEM_promote_table_init; + return bch_err_throw(c, ENOMEM_promote_table_init); return 0; } diff --git a/fs/bcachefs/io_read.h b/fs/bcachefs/io_read.h index c08b9c047b3e..9c5ddbf861b3 100644 --- a/fs/bcachefs/io_read.h +++ b/fs/bcachefs/io_read.h @@ -44,6 +44,7 @@ struct bch_read_bio { have_ioref:1, narrow_crcs:1, saw_error:1, + self_healing:1, context:2; }; u16 _state; @@ -91,6 +92,8 @@ static inline int bch2_read_indirect_extent(struct btree_trans *trans, return 0; *data_btree = BTREE_ID_reflink; + + struct bch_fs *c = trans->c; struct btree_iter iter; struct bkey_s_c k = bch2_lookup_indirect_extent(trans, &iter, offset_into_extent, @@ -102,10 +105,10 @@ static inline int bch2_read_indirect_extent(struct btree_trans *trans, if (bkey_deleted(k.k)) { bch2_trans_iter_exit(trans, &iter); - return -BCH_ERR_missing_indirect_extent; + return bch_err_throw(c, missing_indirect_extent); } - bch2_bkey_buf_reassemble(extent, trans->c, k); + bch2_bkey_buf_reassemble(extent, c, k); bch2_trans_iter_exit(trans, &iter); return 0; } diff --git a/fs/bcachefs/io_write.c b/fs/bcachefs/io_write.c index 52a60982a66b..88b1eec8eff3 100644 --- a/fs/bcachefs/io_write.c +++ b/fs/bcachefs/io_write.c @@ -558,6 +558,7 @@ static void bch2_write_done(struct closure *cl) static noinline int bch2_write_drop_io_error_ptrs(struct bch_write_op *op) { + struct bch_fs *c = op->c; struct keylist *keys = &op->insert_keys; struct bkey_i *src, *dst = keys->keys, *n; @@ -569,7 +570,7 @@ static noinline int bch2_write_drop_io_error_ptrs(struct bch_write_op *op) test_bit(ptr->dev, op->failed.d)); if (!bch2_bkey_nr_ptrs(bkey_i_to_s_c(src))) - return -BCH_ERR_data_write_io; + return bch_err_throw(c, data_write_io); } if (dst != src) @@ -976,7 +977,7 @@ csum_err: op->crc.csum_type < BCH_CSUM_NR ? __bch2_csum_types[op->crc.csum_type] : "(unknown)"); - return -BCH_ERR_data_write_csum; + return bch_err_throw(c, data_write_csum); } static int bch2_write_extent(struct bch_write_op *op, struct write_point *wp, @@ -1208,16 +1209,13 @@ static bool bch2_extent_is_writeable(struct bch_write_op *op, e = bkey_s_c_to_extent(k); - rcu_read_lock(); + guard(rcu)(); extent_for_each_ptr_decode(e, p, entry) { - if (crc_is_encoded(p.crc) || p.has_ec) { - rcu_read_unlock(); + if (crc_is_encoded(p.crc) || p.has_ec) return false; - } replicas += bch2_extent_ptr_durability(c, &p); } - rcu_read_unlock(); return replicas >= op->opts.data_replicas; } @@ -1290,7 +1288,7 @@ static void bch2_nocow_write_convert_unwritten(struct bch_write_op *op) static void __bch2_nocow_write_done(struct bch_write_op *op) { if (unlikely(op->flags & BCH_WRITE_io_error)) { - op->error = -BCH_ERR_data_write_io; + op->error = bch_err_throw(op->c, data_write_io); } else if (unlikely(op->flags & BCH_WRITE_convert_unwritten)) bch2_nocow_write_convert_unwritten(op); } @@ -1483,10 +1481,10 @@ err_bucket_stale: "pointer to invalid bucket in nocow path on device %llu\n %s", stale_at->b.inode, (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { - ret = -BCH_ERR_data_write_invalid_ptr; + ret = bch_err_throw(c, data_write_invalid_ptr); } else { /* We can retry this: */ - ret = -BCH_ERR_transaction_restart; + ret = bch_err_throw(c, transaction_restart); } printbuf_exit(&buf); @@ -1693,18 +1691,18 @@ CLOSURE_CALLBACK(bch2_write) if (unlikely(bio->bi_iter.bi_size & (c->opts.block_size - 1))) { bch2_write_op_error(op, op->pos.offset, "misaligned write"); - op->error = -BCH_ERR_data_write_misaligned; + op->error = bch_err_throw(c, data_write_misaligned); goto err; } if (c->opts.nochanges) { - op->error = -BCH_ERR_erofs_no_writes; + op->error = bch_err_throw(c, erofs_no_writes); goto err; } if (!(op->flags & BCH_WRITE_move) && !enumerated_ref_tryget(&c->writes, BCH_WRITE_REF_write)) { - op->error = -BCH_ERR_erofs_no_writes; + op->error = bch_err_throw(c, erofs_no_writes); goto err; } @@ -1776,7 +1774,7 @@ int bch2_fs_io_write_init(struct bch_fs *c) { if (bioset_init(&c->bio_write, 1, offsetof(struct bch_write_bio, bio), BIOSET_NEED_BVECS) || bioset_init(&c->replica_set, 4, offsetof(struct bch_write_bio, bio), 0)) - return -BCH_ERR_ENOMEM_bio_write_init; + return bch_err_throw(c, ENOMEM_bio_write_init); return 0; } diff --git a/fs/bcachefs/journal.c b/fs/bcachefs/journal.c index 09b70fd140a1..f22b05e02c1e 100644 --- a/fs/bcachefs/journal.c +++ b/fs/bcachefs/journal.c @@ -397,7 +397,7 @@ static int journal_entry_open(struct journal *j) BUG_ON(BCH_SB_CLEAN(c->disk_sb.sb)); if (j->blocked) - return -BCH_ERR_journal_blocked; + return bch_err_throw(c, journal_blocked); if (j->cur_entry_error) return j->cur_entry_error; @@ -407,23 +407,23 @@ static int journal_entry_open(struct journal *j) return ret; if (!fifo_free(&j->pin)) - return -BCH_ERR_journal_pin_full; + return bch_err_throw(c, journal_pin_full); if (nr_unwritten_journal_entries(j) == ARRAY_SIZE(j->buf)) - return -BCH_ERR_journal_max_in_flight; + return bch_err_throw(c, journal_max_in_flight); if (atomic64_read(&j->seq) - j->seq_write_started == JOURNAL_STATE_BUF_NR) - return -BCH_ERR_journal_max_open; + return bch_err_throw(c, journal_max_open); if (unlikely(journal_cur_seq(j) >= JOURNAL_SEQ_MAX)) { bch_err(c, "cannot start: journal seq overflow"); if (bch2_fs_emergency_read_only_locked(c)) bch_err(c, "fatal error - emergency read only"); - return -BCH_ERR_journal_shutdown; + return bch_err_throw(c, journal_shutdown); } if (!j->free_buf && !buf->data) - return -BCH_ERR_journal_buf_enomem; /* will retry after write completion frees up a buf */ + return bch_err_throw(c, journal_buf_enomem); /* will retry after write completion frees up a buf */ BUG_ON(!j->cur_entry_sectors); @@ -447,7 +447,7 @@ static int journal_entry_open(struct journal *j) u64s = clamp_t(int, u64s, 0, JOURNAL_ENTRY_CLOSED_VAL - 1); if (u64s <= (ssize_t) j->early_journal_entries.nr) - return -BCH_ERR_journal_full; + return bch_err_throw(c, journal_full); if (fifo_empty(&j->pin) && j->reclaim_thread) wake_up_process(j->reclaim_thread); @@ -464,7 +464,7 @@ static int journal_entry_open(struct journal *j) journal_cur_seq(j)); if (bch2_fs_emergency_read_only_locked(c)) bch_err(c, "fatal error - emergency read only"); - return -BCH_ERR_journal_shutdown; + return bch_err_throw(c, journal_shutdown); } BUG_ON(j->pin.back - 1 != atomic64_read(&j->seq)); @@ -597,16 +597,16 @@ retry: return ret; if (j->blocked) - return -BCH_ERR_journal_blocked; + return bch_err_throw(c, journal_blocked); if ((flags & BCH_WATERMARK_MASK) < j->watermark) { - ret = -BCH_ERR_journal_full; + ret = bch_err_throw(c, journal_full); can_discard = j->can_discard; goto out; } if (nr_unwritten_journal_entries(j) == ARRAY_SIZE(j->buf) && !journal_entry_is_open(j)) { - ret = -BCH_ERR_journal_max_in_flight; + ret = bch_err_throw(c, journal_max_in_flight); goto out; } @@ -647,7 +647,7 @@ out: goto retry; if (journal_error_check_stuck(j, ret, flags)) - ret = -BCH_ERR_journal_stuck; + ret = bch_err_throw(c, journal_stuck); if (ret == -BCH_ERR_journal_max_in_flight && track_event_change(&c->times[BCH_TIME_blocked_journal_max_in_flight], true) && @@ -708,10 +708,9 @@ static unsigned max_dev_latency(struct bch_fs *c) { u64 nsecs = 0; - rcu_read_lock(); + guard(rcu)(); for_each_rw_member_rcu(c, ca) nsecs = max(nsecs, ca->io_latency[WRITE].stats.max_duration); - rcu_read_unlock(); return nsecs_to_jiffies(nsecs); } @@ -813,6 +812,7 @@ out: int bch2_journal_flush_seq_async(struct journal *j, u64 seq, struct closure *parent) { + struct bch_fs *c = container_of(j, struct bch_fs, journal); struct journal_buf *buf; int ret = 0; @@ -828,7 +828,7 @@ int bch2_journal_flush_seq_async(struct journal *j, u64 seq, /* Recheck under lock: */ if (j->err_seq && seq >= j->err_seq) { - ret = -BCH_ERR_journal_flush_err; + ret = bch_err_throw(c, journal_flush_err); goto out; } @@ -999,7 +999,7 @@ int bch2_journal_meta(struct journal *j) struct bch_fs *c = container_of(j, struct bch_fs, journal); if (!enumerated_ref_tryget(&c->writes, BCH_WRITE_REF_journal)) - return -BCH_ERR_erofs_no_writes; + return bch_err_throw(c, erofs_no_writes); int ret = __bch2_journal_meta(j); enumerated_ref_put(&c->writes, BCH_WRITE_REF_journal); @@ -1132,7 +1132,7 @@ static int bch2_set_nr_journal_buckets_iter(struct bch_dev *ca, unsigned nr, new_buckets = kcalloc(nr, sizeof(u64), GFP_KERNEL); new_bucket_seq = kcalloc(nr, sizeof(u64), GFP_KERNEL); if (!bu || !ob || !new_buckets || !new_bucket_seq) { - ret = -BCH_ERR_ENOMEM_set_nr_journal_buckets; + ret = bch_err_throw(c, ENOMEM_set_nr_journal_buckets); goto err_free; } @@ -1283,7 +1283,7 @@ static int bch2_set_nr_journal_buckets_loop(struct bch_fs *c, struct bch_dev *ca ret = 0; /* wait and retry */ bch2_disk_reservation_put(c, &disk_res); - closure_sync(&cl); + bch2_wait_on_allocator(c, &cl); } return ret; @@ -1304,6 +1304,66 @@ int bch2_set_nr_journal_buckets(struct bch_fs *c, struct bch_dev *ca, return ret; } +int bch2_dev_journal_bucket_delete(struct bch_dev *ca, u64 b) +{ + struct bch_fs *c = ca->fs; + struct journal *j = &c->journal; + struct journal_device *ja = &ca->journal; + + guard(mutex)(&c->sb_lock); + unsigned pos; + for (pos = 0; pos < ja->nr; pos++) + if (ja->buckets[pos] == b) + break; + + if (pos == ja->nr) { + bch_err(ca, "journal bucket %llu not found when deleting", b); + return -EINVAL; + } + + u64 *new_buckets = kcalloc(ja->nr, sizeof(u64), GFP_KERNEL);; + if (!new_buckets) + return bch_err_throw(c, ENOMEM_set_nr_journal_buckets); + + memcpy(new_buckets, ja->buckets, ja->nr * sizeof(u64)); + memmove(&new_buckets[pos], + &new_buckets[pos + 1], + (ja->nr - 1 - pos) * sizeof(new_buckets[0])); + + int ret = bch2_journal_buckets_to_sb(c, ca, ja->buckets, ja->nr - 1) ?: + bch2_write_super(c); + if (ret) { + kfree(new_buckets); + return ret; + } + + scoped_guard(spinlock, &j->lock) { + if (pos < ja->discard_idx) + --ja->discard_idx; + if (pos < ja->dirty_idx_ondisk) + --ja->dirty_idx_ondisk; + if (pos < ja->dirty_idx) + --ja->dirty_idx; + if (pos < ja->cur_idx) + --ja->cur_idx; + + ja->nr--; + + memmove(&ja->buckets[pos], + &ja->buckets[pos + 1], + (ja->nr - pos) * sizeof(ja->buckets[0])); + + memmove(&ja->bucket_seq[pos], + &ja->bucket_seq[pos + 1], + (ja->nr - pos) * sizeof(ja->bucket_seq[0])); + + bch2_journal_space_available(j); + } + + kfree(new_buckets); + return 0; +} + int bch2_dev_journal_alloc(struct bch_dev *ca, bool new_fs) { struct bch_fs *c = ca->fs; @@ -1313,14 +1373,14 @@ int bch2_dev_journal_alloc(struct bch_dev *ca, bool new_fs) if (c->sb.features & BIT_ULL(BCH_FEATURE_small_image)) { bch_err(c, "cannot allocate journal, filesystem is an unresized image file"); - return -BCH_ERR_erofs_filesystem_full; + return bch_err_throw(c, erofs_filesystem_full); } unsigned nr; int ret; if (dynamic_fault("bcachefs:add:journal_alloc")) { - ret = -BCH_ERR_ENOMEM_set_nr_journal_buckets; + ret = bch_err_throw(c, ENOMEM_set_nr_journal_buckets); goto err; } @@ -1414,14 +1474,13 @@ void bch2_fs_journal_stop(struct journal *j) clear_bit(JOURNAL_running, &j->flags); } -int bch2_fs_journal_start(struct journal *j, u64 cur_seq) +int bch2_fs_journal_start(struct journal *j, u64 last_seq, u64 cur_seq) { struct bch_fs *c = container_of(j, struct bch_fs, journal); struct journal_entry_pin_list *p; struct journal_replay *i, **_i; struct genradix_iter iter; bool had_entries = false; - u64 last_seq = cur_seq, nr, seq; /* * @@ -1435,17 +1494,11 @@ int bch2_fs_journal_start(struct journal *j, u64 cur_seq) return -EINVAL; } - genradix_for_each_reverse(&c->journal_entries, iter, _i) { - i = *_i; - - if (journal_replay_ignore(i)) - continue; - - last_seq = le64_to_cpu(i->j.last_seq); - break; - } + /* Clean filesystem? */ + if (!last_seq) + last_seq = cur_seq; - nr = cur_seq - last_seq; + u64 nr = cur_seq - last_seq; /* * Extra fudge factor, in case we crashed when the journal pin fifo was @@ -1459,7 +1512,7 @@ int bch2_fs_journal_start(struct journal *j, u64 cur_seq) init_fifo(&j->pin, roundup_pow_of_two(nr), GFP_KERNEL); if (!j->pin.data) { bch_err(c, "error reallocating journal fifo (%llu open entries)", nr); - return -BCH_ERR_ENOMEM_journal_pin_fifo; + return bch_err_throw(c, ENOMEM_journal_pin_fifo); } j->replay_journal_seq = last_seq; @@ -1472,6 +1525,7 @@ int bch2_fs_journal_start(struct journal *j, u64 cur_seq) j->pin.back = cur_seq; atomic64_set(&j->seq, cur_seq - 1); + u64 seq; fifo_for_each_entry_ptr(p, &j->pin, seq) journal_pin_list_init(p, 1); @@ -1547,6 +1601,7 @@ void bch2_dev_journal_exit(struct bch_dev *ca) int bch2_dev_journal_init(struct bch_dev *ca, struct bch_sb *sb) { + struct bch_fs *c = ca->fs; struct journal_device *ja = &ca->journal; struct bch_sb_field_journal *journal_buckets = bch2_sb_field_get(sb, journal); @@ -1566,7 +1621,7 @@ int bch2_dev_journal_init(struct bch_dev *ca, struct bch_sb *sb) ja->bucket_seq = kcalloc(ja->nr, sizeof(u64), GFP_KERNEL); if (!ja->bucket_seq) - return -BCH_ERR_ENOMEM_dev_journal_init; + return bch_err_throw(c, ENOMEM_dev_journal_init); unsigned nr_bvecs = DIV_ROUND_UP(JOURNAL_ENTRY_SIZE_MAX, PAGE_SIZE); @@ -1574,7 +1629,7 @@ int bch2_dev_journal_init(struct bch_dev *ca, struct bch_sb *sb) ja->bio[i] = kzalloc(struct_size(ja->bio[i], bio.bi_inline_vecs, nr_bvecs), GFP_KERNEL); if (!ja->bio[i]) - return -BCH_ERR_ENOMEM_dev_journal_init; + return bch_err_throw(c, ENOMEM_dev_journal_init); ja->bio[i]->ca = ca; ja->bio[i]->buf_idx = i; @@ -1583,7 +1638,7 @@ int bch2_dev_journal_init(struct bch_dev *ca, struct bch_sb *sb) ja->buckets = kcalloc(ja->nr, sizeof(u64), GFP_KERNEL); if (!ja->buckets) - return -BCH_ERR_ENOMEM_dev_journal_init; + return bch_err_throw(c, ENOMEM_dev_journal_init); if (journal_buckets_v2) { unsigned nr = bch2_sb_field_journal_v2_nr_entries(journal_buckets_v2); @@ -1637,10 +1692,12 @@ void bch2_fs_journal_init_early(struct journal *j) int bch2_fs_journal_init(struct journal *j) { + struct bch_fs *c = container_of(j, struct bch_fs, journal); + j->free_buf_size = j->buf_size_want = JOURNAL_ENTRY_SIZE_MIN; j->free_buf = kvmalloc(j->free_buf_size, GFP_KERNEL); if (!j->free_buf) - return -BCH_ERR_ENOMEM_journal_buf; + return bch_err_throw(c, ENOMEM_journal_buf); for (unsigned i = 0; i < ARRAY_SIZE(j->buf); i++) j->buf[i].idx = i; @@ -1648,7 +1705,7 @@ int bch2_fs_journal_init(struct journal *j) j->wq = alloc_workqueue("bcachefs_journal", WQ_HIGHPRI|WQ_FREEZABLE|WQ_UNBOUND|WQ_MEM_RECLAIM, 512); if (!j->wq) - return -BCH_ERR_ENOMEM_fs_other_alloc; + return bch_err_throw(c, ENOMEM_fs_other_alloc); return 0; } @@ -1672,7 +1729,7 @@ void __bch2_journal_debug_to_text(struct printbuf *out, struct journal *j) printbuf_tabstop_push(out, 28); out->atomic++; - rcu_read_lock(); + guard(rcu)(); s = READ_ONCE(j->reservations); prt_printf(out, "flags:\t"); @@ -1763,8 +1820,6 @@ void __bch2_journal_debug_to_text(struct printbuf *out, struct journal *j) prt_printf(out, "replicas want %u need %u\n", c->opts.metadata_replicas, c->opts.metadata_replicas_required); - rcu_read_unlock(); - --out->atomic; } diff --git a/fs/bcachefs/journal.h b/fs/bcachefs/journal.h index 8ff00a0ec778..977907038d98 100644 --- a/fs/bcachefs/journal.h +++ b/fs/bcachefs/journal.h @@ -444,15 +444,16 @@ struct journal_buf *bch2_next_write_buffer_flush_journal_buf(struct journal *, u void __bch2_journal_debug_to_text(struct printbuf *, struct journal *); void bch2_journal_debug_to_text(struct printbuf *, struct journal *); -int bch2_set_nr_journal_buckets(struct bch_fs *, struct bch_dev *, - unsigned nr); +int bch2_set_nr_journal_buckets(struct bch_fs *, struct bch_dev *, unsigned); +int bch2_dev_journal_bucket_delete(struct bch_dev *, u64); + int bch2_dev_journal_alloc(struct bch_dev *, bool); int bch2_fs_journal_alloc(struct bch_fs *); void bch2_dev_journal_stop(struct journal *, struct bch_dev *); void bch2_fs_journal_stop(struct journal *); -int bch2_fs_journal_start(struct journal *, u64); +int bch2_fs_journal_start(struct journal *, u64, u64); void bch2_journal_set_replay_done(struct journal *); void bch2_dev_journal_exit(struct bch_dev *); diff --git a/fs/bcachefs/journal_io.c b/fs/bcachefs/journal_io.c index 63bb207208b2..9e028dbcc3d0 100644 --- a/fs/bcachefs/journal_io.c +++ b/fs/bcachefs/journal_io.c @@ -49,25 +49,27 @@ void bch2_journal_pos_from_member_info_resume(struct bch_fs *c) mutex_unlock(&c->sb_lock); } -void bch2_journal_ptrs_to_text(struct printbuf *out, struct bch_fs *c, - struct journal_replay *j) +static void bch2_journal_ptr_to_text(struct printbuf *out, struct bch_fs *c, struct journal_ptr *p) +{ + struct bch_dev *ca = bch2_dev_tryget_noerror(c, p->dev); + prt_printf(out, "%s %u:%u:%u (sector %llu)", + ca ? ca->name : "(invalid dev)", + p->dev, p->bucket, p->bucket_offset, p->sector); + bch2_dev_put(ca); +} + +void bch2_journal_ptrs_to_text(struct printbuf *out, struct bch_fs *c, struct journal_replay *j) { darray_for_each(j->ptrs, i) { if (i != j->ptrs.data) prt_printf(out, " "); - prt_printf(out, "%u:%u:%u (sector %llu)", - i->dev, i->bucket, i->bucket_offset, i->sector); + bch2_journal_ptr_to_text(out, c, i); } } -static void bch2_journal_replay_to_text(struct printbuf *out, struct bch_fs *c, - struct journal_replay *j) +static void bch2_journal_datetime_to_text(struct printbuf *out, struct jset *j) { - prt_printf(out, "seq %llu ", le64_to_cpu(j->j.seq)); - - bch2_journal_ptrs_to_text(out, c, j); - - for_each_jset_entry_type(entry, &j->j, BCH_JSET_ENTRY_datetime) { + for_each_jset_entry_type(entry, j, BCH_JSET_ENTRY_datetime) { struct jset_entry_datetime *datetime = container_of(entry, struct jset_entry_datetime, entry); bch2_prt_datetime(out, le64_to_cpu(datetime->seconds)); @@ -75,6 +77,15 @@ static void bch2_journal_replay_to_text(struct printbuf *out, struct bch_fs *c, } } +static void bch2_journal_replay_to_text(struct printbuf *out, struct bch_fs *c, + struct journal_replay *j) +{ + prt_printf(out, "seq %llu ", le64_to_cpu(j->j.seq)); + bch2_journal_datetime_to_text(out, &j->j); + prt_char(out, ' '); + bch2_journal_ptrs_to_text(out, c, j); +} + static struct nonce journal_nonce(const struct jset *jset) { return (struct nonce) {{ @@ -149,6 +160,9 @@ static int journal_entry_add(struct bch_fs *c, struct bch_dev *ca, struct printbuf buf = PRINTBUF; int ret = JOURNAL_ENTRY_ADD_OK; + if (last_seq && c->opts.journal_rewind) + last_seq = min(last_seq, c->opts.journal_rewind); + if (!c->journal.oldest_seq_found_ondisk || le64_to_cpu(j->seq) < c->journal.oldest_seq_found_ondisk) c->journal.oldest_seq_found_ondisk = le64_to_cpu(j->seq); @@ -188,7 +202,7 @@ static int journal_entry_add(struct bch_fs *c, struct bch_dev *ca, journal_entry_radix_idx(c, le64_to_cpu(j->seq)), GFP_KERNEL); if (!_i) - return -BCH_ERR_ENOMEM_journal_entry_add; + return bch_err_throw(c, ENOMEM_journal_entry_add); /* * Duplicate journal entries? If so we want the one that didn't have a @@ -231,7 +245,7 @@ static int journal_entry_add(struct bch_fs *c, struct bch_dev *ca, replace: i = kvmalloc(offsetof(struct journal_replay, j) + bytes, GFP_KERNEL); if (!i) - return -BCH_ERR_ENOMEM_journal_entry_add; + return bch_err_throw(c, ENOMEM_journal_entry_add); darray_init(&i->ptrs); i->csum_good = entry_ptr.csum_good; @@ -311,7 +325,7 @@ static void journal_entry_err_msg(struct printbuf *out, bch2_sb_error_count(c, BCH_FSCK_ERR_##_err); \ if (bch2_fs_inconsistent(c, \ "corrupt metadata before write: %s\n", _buf.buf)) {\ - ret = -BCH_ERR_fsck_errors_not_fixed; \ + ret = bch_err_throw(c, fsck_errors_not_fixed); \ goto fsck_err; \ } \ break; \ @@ -418,6 +432,10 @@ static void journal_entry_btree_keys_to_text(struct printbuf *out, struct bch_fs bool first = true; jset_entry_for_each_key(entry, k) { + /* We may be called on entries that haven't been validated: */ + if (!k->k.u64s) + break; + if (!first) { prt_newline(out); bch2_prt_jset_entry_type(out, entry->type); @@ -1005,19 +1023,19 @@ struct journal_read_buf { size_t size; }; -static int journal_read_buf_realloc(struct journal_read_buf *b, +static int journal_read_buf_realloc(struct bch_fs *c, struct journal_read_buf *b, size_t new_size) { void *n; /* the bios are sized for this many pages, max: */ if (new_size > JOURNAL_ENTRY_SIZE_MAX) - return -BCH_ERR_ENOMEM_journal_read_buf_realloc; + return bch_err_throw(c, ENOMEM_journal_read_buf_realloc); new_size = roundup_pow_of_two(new_size); n = kvmalloc(new_size, GFP_KERNEL); if (!n) - return -BCH_ERR_ENOMEM_journal_read_buf_realloc; + return bch_err_throw(c, ENOMEM_journal_read_buf_realloc); kvfree(b->data); b->data = n; @@ -1037,7 +1055,6 @@ static int journal_read_bucket(struct bch_dev *ca, u64 offset = bucket_to_sector(ca, ja->buckets[bucket]), end = offset + ca->mi.bucket_size; bool saw_bad = false, csum_good; - struct printbuf err = PRINTBUF; int ret = 0; pr_debug("reading %u", bucket); @@ -1053,7 +1070,7 @@ reread: bio = bio_kmalloc(nr_bvecs, GFP_KERNEL); if (!bio) - return -BCH_ERR_ENOMEM_journal_read_bucket; + return bch_err_throw(c, ENOMEM_journal_read_bucket); bio_init(bio, ca->disk_sb.bdev, bio->bi_inline_vecs, nr_bvecs, REQ_OP_READ); bio->bi_iter.bi_sector = offset; @@ -1064,7 +1081,7 @@ reread: kfree(bio); if (!ret && bch2_meta_read_fault("journal")) - ret = -BCH_ERR_EIO_fault_injected; + ret = bch_err_throw(c, EIO_fault_injected); bch2_account_io_completion(ca, BCH_MEMBER_ERROR_read, submit_time, !ret); @@ -1078,7 +1095,7 @@ reread: * found on a different device, and missing or * no journal entries will be handled later */ - goto out; + return 0; } j = buf->data; @@ -1092,15 +1109,15 @@ reread: break; case JOURNAL_ENTRY_REREAD: if (vstruct_bytes(j) > buf->size) { - ret = journal_read_buf_realloc(buf, + ret = journal_read_buf_realloc(c, buf, vstruct_bytes(j)); if (ret) - goto err; + return ret; } goto reread; case JOURNAL_ENTRY_NONE: if (!saw_bad) - goto out; + return 0; /* * On checksum error we don't really trust the size * field of the journal entry we read, so try reading @@ -1109,7 +1126,7 @@ reread: sectors = block_sectors(c); goto next_block; default: - goto err; + return ret; } if (le64_to_cpu(j->seq) > ja->highest_seq_found) { @@ -1126,22 +1143,20 @@ reread: * bucket: */ if (le64_to_cpu(j->seq) < ja->bucket_seq[bucket]) - goto out; + return 0; ja->bucket_seq[bucket] = le64_to_cpu(j->seq); - enum bch_csum_type csum_type = JSET_CSUM_TYPE(j); struct bch_csum csum; csum_good = jset_csum_good(c, j, &csum); bch2_account_io_completion(ca, BCH_MEMBER_ERROR_checksum, 0, csum_good); if (!csum_good) { - bch_err_dev_ratelimited(ca, "%s", - (printbuf_reset(&err), - prt_str(&err, "journal "), - bch2_csum_err_msg(&err, csum_type, j->csum, csum), - err.buf)); + /* + * Don't print an error here, we'll print the error + * later if we need this journal entry + */ saw_bad = true; } @@ -1153,6 +1168,7 @@ reread: mutex_lock(&jlist->lock); ret = journal_entry_add(c, ca, (struct journal_ptr) { .csum_good = csum_good, + .csum = csum, .dev = ca->dev_idx, .bucket = bucket, .bucket_offset = offset - @@ -1167,7 +1183,7 @@ reread: case JOURNAL_ENTRY_ADD_OUT_OF_RANGE: break; default: - goto err; + return ret; } next_block: pr_debug("next"); @@ -1176,11 +1192,7 @@ next_block: j = ((void *) j) + (sectors << 9); } -out: - ret = 0; -err: - printbuf_exit(&err); - return ret; + return 0; } static CLOSURE_CALLBACK(bch2_journal_read_device) @@ -1197,7 +1209,7 @@ static CLOSURE_CALLBACK(bch2_journal_read_device) if (!ja->nr) goto out; - ret = journal_read_buf_realloc(&buf, PAGE_SIZE); + ret = journal_read_buf_realloc(c, &buf, PAGE_SIZE); if (ret) goto err; @@ -1229,13 +1241,105 @@ err: goto out; } +noinline_for_stack +static void bch2_journal_print_checksum_error(struct bch_fs *c, struct journal_replay *j) +{ + struct printbuf buf = PRINTBUF; + enum bch_csum_type csum_type = JSET_CSUM_TYPE(&j->j); + bool have_good = false; + + prt_printf(&buf, "invalid journal checksum(s) at seq %llu ", le64_to_cpu(j->j.seq)); + bch2_journal_datetime_to_text(&buf, &j->j); + prt_newline(&buf); + + darray_for_each(j->ptrs, ptr) + if (!ptr->csum_good) { + bch2_journal_ptr_to_text(&buf, c, ptr); + prt_char(&buf, ' '); + bch2_csum_to_text(&buf, csum_type, ptr->csum); + prt_newline(&buf); + } else { + have_good = true; + } + + prt_printf(&buf, "should be "); + bch2_csum_to_text(&buf, csum_type, j->j.csum); + + if (have_good) + prt_printf(&buf, "\n(had good copy on another device)"); + + bch2_print_str(c, KERN_ERR, buf.buf); + printbuf_exit(&buf); +} + +noinline_for_stack +static int bch2_journal_check_for_missing(struct bch_fs *c, u64 start_seq, u64 end_seq) +{ + struct printbuf buf = PRINTBUF; + int ret = 0; + + struct genradix_iter radix_iter; + struct journal_replay *i, **_i, *prev = NULL; + u64 seq = start_seq; + + genradix_for_each(&c->journal_entries, radix_iter, _i) { + i = *_i; + + if (journal_replay_ignore(i)) + continue; + + BUG_ON(seq > le64_to_cpu(i->j.seq)); + + while (seq < le64_to_cpu(i->j.seq)) { + while (seq < le64_to_cpu(i->j.seq) && + bch2_journal_seq_is_blacklisted(c, seq, false)) + seq++; + + if (seq == le64_to_cpu(i->j.seq)) + break; + + u64 missing_start = seq; + + while (seq < le64_to_cpu(i->j.seq) && + !bch2_journal_seq_is_blacklisted(c, seq, false)) + seq++; + + u64 missing_end = seq - 1; + + printbuf_reset(&buf); + prt_printf(&buf, "journal entries %llu-%llu missing! (replaying %llu-%llu)", + missing_start, missing_end, + start_seq, end_seq); + + prt_printf(&buf, "\nprev at "); + if (prev) { + bch2_journal_ptrs_to_text(&buf, c, prev); + prt_printf(&buf, " size %zu", vstruct_sectors(&prev->j, c->block_bits)); + } else + prt_printf(&buf, "(none)"); + + prt_printf(&buf, "\nnext at "); + bch2_journal_ptrs_to_text(&buf, c, i); + prt_printf(&buf, ", continue?"); + + fsck_err(c, journal_entries_missing, "%s", buf.buf); + } + + prev = i; + seq++; + } +fsck_err: + printbuf_exit(&buf); + return ret; +} + int bch2_journal_read(struct bch_fs *c, u64 *last_seq, u64 *blacklist_seq, u64 *start_seq) { struct journal_list jlist; - struct journal_replay *i, **_i, *prev = NULL; + struct journal_replay *i, **_i; struct genradix_iter radix_iter; struct printbuf buf = PRINTBUF; bool degraded = false, last_write_torn = false; @@ -1326,14 +1430,24 @@ int bch2_journal_read(struct bch_fs *c, return 0; } - bch_info(c, "journal read done, replaying entries %llu-%llu", - *last_seq, *blacklist_seq - 1); + printbuf_reset(&buf); + prt_printf(&buf, "journal read done, replaying entries %llu-%llu", + *last_seq, *blacklist_seq - 1); - if (*start_seq != *blacklist_seq) - bch_info(c, "dropped unflushed entries %llu-%llu", - *blacklist_seq, *start_seq - 1); + /* + * Drop blacklisted entries and entries older than last_seq (or start of + * journal rewind: + */ + u64 drop_before = *last_seq; + if (c->opts.journal_rewind) { + drop_before = min(drop_before, c->opts.journal_rewind); + prt_printf(&buf, " (rewinding from %llu)", c->opts.journal_rewind); + } - /* Drop blacklisted entries and entries older than last_seq: */ + *last_seq = drop_before; + if (*start_seq != *blacklist_seq) + prt_printf(&buf, " (unflushed %llu-%llu)", *blacklist_seq, *start_seq - 1); + bch_info(c, "%s", buf.buf); genradix_for_each(&c->journal_entries, radix_iter, _i) { i = *_i; @@ -1341,7 +1455,7 @@ int bch2_journal_read(struct bch_fs *c, continue; seq = le64_to_cpu(i->j.seq); - if (seq < *last_seq) { + if (seq < drop_before) { journal_replay_free(c, i, false); continue; } @@ -1354,56 +1468,9 @@ int bch2_journal_read(struct bch_fs *c, } } - /* Check for missing entries: */ - seq = *last_seq; - genradix_for_each(&c->journal_entries, radix_iter, _i) { - i = *_i; - - if (journal_replay_ignore(i)) - continue; - - BUG_ON(seq > le64_to_cpu(i->j.seq)); - - while (seq < le64_to_cpu(i->j.seq)) { - u64 missing_start, missing_end; - struct printbuf buf1 = PRINTBUF, buf2 = PRINTBUF; - - while (seq < le64_to_cpu(i->j.seq) && - bch2_journal_seq_is_blacklisted(c, seq, false)) - seq++; - - if (seq == le64_to_cpu(i->j.seq)) - break; - - missing_start = seq; - - while (seq < le64_to_cpu(i->j.seq) && - !bch2_journal_seq_is_blacklisted(c, seq, false)) - seq++; - - if (prev) { - bch2_journal_ptrs_to_text(&buf1, c, prev); - prt_printf(&buf1, " size %zu", vstruct_sectors(&prev->j, c->block_bits)); - } else - prt_printf(&buf1, "(none)"); - bch2_journal_ptrs_to_text(&buf2, c, i); - - missing_end = seq - 1; - fsck_err(c, journal_entries_missing, - "journal entries %llu-%llu missing! (replaying %llu-%llu)\n" - "prev at %s\n" - "next at %s, continue?", - missing_start, missing_end, - *last_seq, *blacklist_seq - 1, - buf1.buf, buf2.buf); - - printbuf_exit(&buf1); - printbuf_exit(&buf2); - } - - prev = i; - seq++; - } + ret = bch2_journal_check_for_missing(c, drop_before, *blacklist_seq - 1); + if (ret) + goto err; genradix_for_each(&c->journal_entries, radix_iter, _i) { union bch_replicas_padded replicas = { @@ -1416,15 +1483,15 @@ int bch2_journal_read(struct bch_fs *c, if (journal_replay_ignore(i)) continue; - darray_for_each(i->ptrs, ptr) { - struct bch_dev *ca = bch2_dev_have_ref(c, ptr->dev); - - if (!ptr->csum_good) - bch_err_dev_offset(ca, ptr->sector, - "invalid journal checksum, seq %llu%s", - le64_to_cpu(i->j.seq), - i->csum_good ? " (had good copy on another device)" : ""); - } + /* + * Don't print checksum errors until we know we're going to use + * a given journal entry: + */ + darray_for_each(i->ptrs, ptr) + if (!ptr->csum_good) { + bch2_journal_print_checksum_error(c, i); + break; + } ret = jset_validate(c, bch2_dev_have_ref(c, i->ptrs.data[0].dev), @@ -1467,7 +1534,7 @@ static void journal_advance_devs_to_next_bucket(struct journal *j, { struct bch_fs *c = container_of(j, struct bch_fs, journal); - rcu_read_lock(); + guard(rcu)(); darray_for_each(*devs, i) { struct bch_dev *ca = rcu_dereference(c->devs[*i]); if (!ca) @@ -1489,7 +1556,6 @@ static void journal_advance_devs_to_next_bucket(struct journal *j, ja->bucket_seq[ja->cur_idx] = le64_to_cpu(seq); } } - rcu_read_unlock(); } static void __journal_write_alloc(struct journal *j, @@ -1559,7 +1625,7 @@ static int journal_write_alloc(struct journal *j, struct journal_buf *w, retry_target: devs = target_rw_devs(c, BCH_DATA_journal, target); - devs_sorted = bch2_dev_alloc_list(c, &j->wp.stripe, &devs); + bch2_dev_alloc_list(c, &j->wp.stripe, &devs, &devs_sorted); retry_alloc: __journal_write_alloc(j, w, &devs_sorted, sectors, replicas, replicas_want); @@ -1581,6 +1647,16 @@ retry_alloc: done: BUG_ON(bkey_val_u64s(&w->key.k) > BCH_REPLICAS_MAX); +#if 0 + /* + * XXX: we need a way to alert the user when we go degraded for any + * reason + */ + if (*replicas < min(replicas_want, + dev_mask_nr(&c->rw_devs[BCH_DATA_free]))) { + } +#endif + return *replicas >= replicas_need ? 0 : -BCH_ERR_insufficient_journal_devices; } @@ -1628,7 +1704,7 @@ static CLOSURE_CALLBACK(journal_write_done) : j->noflush_write_time, j->write_start_time); if (!w->devs_written.nr) { - err = -BCH_ERR_journal_write_err; + err = bch_err_throw(c, journal_write_err); } else { bch2_devlist_to_replicas(&replicas.e, BCH_DATA_journal, w->devs_written); @@ -1640,9 +1716,10 @@ static CLOSURE_CALLBACK(journal_write_done) bch2_log_msg_start(c, &buf); if (err == -BCH_ERR_journal_write_err) - prt_printf(&buf, "unable to write journal to sufficient devices"); + prt_printf(&buf, "unable to write journal to sufficient devices\n"); else - prt_printf(&buf, "journal write error marking replicas: %s", bch2_err_str(err)); + prt_printf(&buf, "journal write error marking replicas: %s\n", + bch2_err_str(err)); bch2_fs_emergency_read_only2(c, &buf); @@ -1690,6 +1767,7 @@ static CLOSURE_CALLBACK(journal_write_done) closure_wake_up(&c->freelist_wait); bch2_reset_alloc_cursors(c); + do_discards = true; } j->seq_ondisk = seq; @@ -2058,7 +2136,7 @@ CLOSURE_CALLBACK(bch2_journal_write) struct journal *j = container_of(w, struct journal, buf[w->idx]); struct bch_fs *c = container_of(j, struct bch_fs, journal); union bch_replicas_padded replicas; - unsigned nr_rw_members = dev_mask_nr(&c->rw_devs[BCH_DATA_journal]); + unsigned nr_rw_members = dev_mask_nr(&c->rw_devs[BCH_DATA_free]); int ret; BUG_ON(BCH_SB_CLEAN(c->disk_sb.sb)); diff --git a/fs/bcachefs/journal_io.h b/fs/bcachefs/journal_io.h index 12b39fcb4424..6fa82c4050fe 100644 --- a/fs/bcachefs/journal_io.h +++ b/fs/bcachefs/journal_io.h @@ -9,6 +9,7 @@ void bch2_journal_pos_from_member_info_resume(struct bch_fs *); struct journal_ptr { bool csum_good; + struct bch_csum csum; u8 dev; u32 bucket; u32 bucket_offset; diff --git a/fs/bcachefs/journal_reclaim.c b/fs/bcachefs/journal_reclaim.c index 70f36f6bc482..0042d43b8e57 100644 --- a/fs/bcachefs/journal_reclaim.c +++ b/fs/bcachefs/journal_reclaim.c @@ -83,18 +83,20 @@ static struct journal_space journal_dev_space_available(struct journal *j, struct bch_dev *ca, enum journal_space_from from) { + struct bch_fs *c = container_of(j, struct bch_fs, journal); struct journal_device *ja = &ca->journal; unsigned sectors, buckets, unwritten; + unsigned bucket_size_aligned = round_down(ca->mi.bucket_size, block_sectors(c)); u64 seq; if (from == journal_space_total) return (struct journal_space) { - .next_entry = ca->mi.bucket_size, - .total = ca->mi.bucket_size * ja->nr, + .next_entry = bucket_size_aligned, + .total = bucket_size_aligned * ja->nr, }; buckets = bch2_journal_dev_buckets_available(j, ja, from); - sectors = ja->sectors_free; + sectors = round_down(ja->sectors_free, block_sectors(c)); /* * We that we don't allocate the space for a journal entry @@ -109,7 +111,7 @@ journal_dev_space_available(struct journal *j, struct bch_dev *ca, continue; /* entry won't fit on this device, skip: */ - if (unwritten > ca->mi.bucket_size) + if (unwritten > bucket_size_aligned) continue; if (unwritten >= sectors) { @@ -119,7 +121,7 @@ journal_dev_space_available(struct journal *j, struct bch_dev *ca, } buckets--; - sectors = ca->mi.bucket_size; + sectors = bucket_size_aligned; } sectors -= unwritten; @@ -127,12 +129,12 @@ journal_dev_space_available(struct journal *j, struct bch_dev *ca, if (sectors < ca->mi.bucket_size && buckets) { buckets--; - sectors = ca->mi.bucket_size; + sectors = bucket_size_aligned; } return (struct journal_space) { .next_entry = sectors, - .total = sectors + buckets * ca->mi.bucket_size, + .total = sectors + buckets * bucket_size_aligned, }; } @@ -146,7 +148,6 @@ static struct journal_space __journal_space_available(struct journal *j, unsigne BUG_ON(nr_devs_want > ARRAY_SIZE(dev_space)); - rcu_read_lock(); for_each_member_device_rcu(c, ca, &c->rw_devs[BCH_DATA_journal]) { if (!ca->journal.nr || !ca->mi.durability) @@ -164,12 +165,17 @@ static struct journal_space __journal_space_available(struct journal *j, unsigne array_insert_item(dev_space, nr_devs, pos, space); } - rcu_read_unlock(); if (nr_devs < nr_devs_want) return (struct journal_space) { 0, 0 }; /* + * It's possible for bucket size to be misaligned w.r.t. the filesystem + * block size: + */ + min_bucket_size = round_down(min_bucket_size, block_sectors(c)); + + /* * We sorted largest to smallest, and we want the smallest out of the * @nr_devs_want largest devices: */ @@ -189,8 +195,8 @@ void bch2_journal_space_available(struct journal *j) int ret = 0; lockdep_assert_held(&j->lock); + guard(rcu)(); - rcu_read_lock(); for_each_member_device_rcu(c, ca, &c->rw_devs[BCH_DATA_journal]) { struct journal_device *ja = &ca->journal; @@ -210,7 +216,6 @@ void bch2_journal_space_available(struct journal *j) max_entry_size = min_t(unsigned, max_entry_size, ca->mi.bucket_size); nr_online++; } - rcu_read_unlock(); j->can_discard = can_discard; @@ -221,15 +226,13 @@ void bch2_journal_space_available(struct journal *j) prt_printf(&buf, "insufficient writeable journal devices available: have %u, need %u\n" "rw journal devs:", nr_online, metadata_replicas_required(c)); - rcu_read_lock(); for_each_member_device_rcu(c, ca, &c->rw_devs[BCH_DATA_journal]) prt_printf(&buf, " %s", ca->name); - rcu_read_unlock(); bch_err(c, "%s", buf.buf); printbuf_exit(&buf); } - ret = -BCH_ERR_insufficient_journal_devices; + ret = bch_err_throw(c, insufficient_journal_devices); goto out; } @@ -243,7 +246,7 @@ void bch2_journal_space_available(struct journal *j) total = j->space[journal_space_total].total; if (!j->space[journal_space_discarded].next_entry) - ret = -BCH_ERR_journal_full; + ret = bch_err_throw(c, journal_full); if ((j->space[journal_space_clean_ondisk].next_entry < j->space[journal_space_clean_ondisk].total) && @@ -256,8 +259,7 @@ void bch2_journal_space_available(struct journal *j) bch2_journal_set_watermark(j); out: j->cur_entry_sectors = !ret - ? round_down(j->space[journal_space_discarded].next_entry, - block_sectors(c)) + ? j->space[journal_space_discarded].next_entry : 0; j->cur_entry_error = ret; @@ -625,9 +627,9 @@ static u64 journal_seq_to_flush(struct journal *j) struct bch_fs *c = container_of(j, struct bch_fs, journal); u64 seq_to_flush = 0; - spin_lock(&j->lock); + guard(spinlock)(&j->lock); + guard(rcu)(); - rcu_read_lock(); for_each_rw_member_rcu(c, ca) { struct journal_device *ja = &ca->journal; unsigned nr_buckets, bucket_to_flush; @@ -642,15 +644,11 @@ static u64 journal_seq_to_flush(struct journal *j) seq_to_flush = max(seq_to_flush, ja->bucket_seq[bucket_to_flush]); } - rcu_read_unlock(); /* Also flush if the pin fifo is more than half full */ - seq_to_flush = max_t(s64, seq_to_flush, - (s64) journal_cur_seq(j) - - (j->pin.size >> 1)); - spin_unlock(&j->lock); - - return seq_to_flush; + return max_t(s64, seq_to_flush, + (s64) journal_cur_seq(j) - + (j->pin.size >> 1)); } /** diff --git a/fs/bcachefs/journal_sb.c b/fs/bcachefs/journal_sb.c index 62b910f2fb27..0cb9b93f13e7 100644 --- a/fs/bcachefs/journal_sb.c +++ b/fs/bcachefs/journal_sb.c @@ -210,7 +210,7 @@ int bch2_journal_buckets_to_sb(struct bch_fs *c, struct bch_dev *ca, j = bch2_sb_field_resize(&ca->disk_sb, journal_v2, (sizeof(*j) + sizeof(j->d[0]) * nr_compacted) / sizeof(u64)); if (!j) - return -BCH_ERR_ENOSPC_sb_journal; + return bch_err_throw(c, ENOSPC_sb_journal); bch2_sb_field_delete(&ca->disk_sb, BCH_SB_FIELD_journal); diff --git a/fs/bcachefs/journal_seq_blacklist.c b/fs/bcachefs/journal_seq_blacklist.c index c5a7d800a0f5..af4fe416d9ec 100644 --- a/fs/bcachefs/journal_seq_blacklist.c +++ b/fs/bcachefs/journal_seq_blacklist.c @@ -78,7 +78,7 @@ int bch2_journal_seq_blacklist_add(struct bch_fs *c, u64 start, u64 end) bl = bch2_sb_field_resize(&c->disk_sb, journal_seq_blacklist, sb_blacklist_u64s(nr + 1)); if (!bl) { - ret = -BCH_ERR_ENOSPC_sb_journal_seq_blacklist; + ret = bch_err_throw(c, ENOSPC_sb_journal_seq_blacklist); goto out; } @@ -152,7 +152,7 @@ int bch2_blacklist_table_initialize(struct bch_fs *c) t = kzalloc(struct_size(t, entries, nr), GFP_KERNEL); if (!t) - return -BCH_ERR_ENOMEM_blacklist_table_init; + return bch_err_throw(c, ENOMEM_blacklist_table_init); t->nr = nr; diff --git a/fs/bcachefs/lru.c b/fs/bcachefs/lru.c index 2f63fc6d456f..57b5b3263b08 100644 --- a/fs/bcachefs/lru.c +++ b/fs/bcachefs/lru.c @@ -145,13 +145,11 @@ static u64 bkey_lru_type_idx(struct bch_fs *c, case BCH_LRU_fragmentation: { a = bch2_alloc_to_v4(k, &a_convert); - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, k.k->p.inode); - u64 idx = ca + return ca ? alloc_lru_idx_fragmentation(*a, ca) : 0; - rcu_read_unlock(); - return idx; } case BCH_LRU_stripes: return k.k->type == KEY_TYPE_stripe diff --git a/fs/bcachefs/migrate.c b/fs/bcachefs/migrate.c index bb7a92270c09..f296cce95338 100644 --- a/fs/bcachefs/migrate.c +++ b/fs/bcachefs/migrate.c @@ -35,7 +35,7 @@ static int drop_dev_ptrs(struct bch_fs *c, struct bkey_s k, nr_good = bch2_bkey_durability(c, k.s_c); if ((!nr_good && !(flags & lost)) || (nr_good < replicas && !(flags & degraded))) - return -BCH_ERR_remove_would_lose_data; + return bch_err_throw(c, remove_would_lose_data); return 0; } @@ -156,7 +156,7 @@ static int bch2_dev_metadata_drop(struct bch_fs *c, /* don't handle this yet: */ if (flags & BCH_FORCE_IF_METADATA_LOST) - return -BCH_ERR_remove_with_metadata_missing_unimplemented; + return bch_err_throw(c, remove_with_metadata_missing_unimplemented); trans = bch2_trans_get(c); bch2_bkey_buf_init(&k); diff --git a/fs/bcachefs/move.c b/fs/bcachefs/move.c index 79f4722621d5..eec591e947bd 100644 --- a/fs/bcachefs/move.c +++ b/fs/bcachefs/move.c @@ -38,30 +38,74 @@ const char * const bch2_data_ops_strs[] = { NULL }; -static void trace_io_move2(struct bch_fs *c, struct bkey_s_c k, - struct bch_io_opts *io_opts, - struct data_update_opts *data_opts) +struct evacuate_bucket_arg { + struct bpos bucket; + int gen; + struct data_update_opts data_opts; +}; + +static bool evacuate_bucket_pred(struct bch_fs *, void *, + enum btree_id, struct bkey_s_c, + struct bch_io_opts *, + struct data_update_opts *); + +static noinline void +trace_io_move2(struct bch_fs *c, struct bkey_s_c k, + struct bch_io_opts *io_opts, + struct data_update_opts *data_opts) { - if (trace_io_move_enabled()) { - struct printbuf buf = PRINTBUF; + struct printbuf buf = PRINTBUF; - bch2_bkey_val_to_text(&buf, c, k); - prt_newline(&buf); - bch2_data_update_opts_to_text(&buf, c, io_opts, data_opts); - trace_io_move(c, buf.buf); - printbuf_exit(&buf); - } + bch2_bkey_val_to_text(&buf, c, k); + prt_newline(&buf); + bch2_data_update_opts_to_text(&buf, c, io_opts, data_opts); + trace_io_move(c, buf.buf); + printbuf_exit(&buf); } -static void trace_io_move_read2(struct bch_fs *c, struct bkey_s_c k) +static noinline void trace_io_move_read2(struct bch_fs *c, struct bkey_s_c k) { - if (trace_io_move_read_enabled()) { - struct printbuf buf = PRINTBUF; + struct printbuf buf = PRINTBUF; - bch2_bkey_val_to_text(&buf, c, k); - trace_io_move_read(c, buf.buf); - printbuf_exit(&buf); + bch2_bkey_val_to_text(&buf, c, k); + trace_io_move_read(c, buf.buf); + printbuf_exit(&buf); +} + +static noinline void +trace_io_move_pred2(struct bch_fs *c, struct bkey_s_c k, + struct bch_io_opts *io_opts, + struct data_update_opts *data_opts, + move_pred_fn pred, void *_arg, bool p) +{ + struct printbuf buf = PRINTBUF; + + prt_printf(&buf, "%ps: %u", pred, p); + + if (pred == evacuate_bucket_pred) { + struct evacuate_bucket_arg *arg = _arg; + prt_printf(&buf, " gen=%u", arg->gen); } + + prt_newline(&buf); + bch2_bkey_val_to_text(&buf, c, k); + prt_newline(&buf); + bch2_data_update_opts_to_text(&buf, c, io_opts, data_opts); + trace_io_move_pred(c, buf.buf); + printbuf_exit(&buf); +} + +static noinline void +trace_io_move_evacuate_bucket2(struct bch_fs *c, struct bpos bucket, int gen) +{ + struct printbuf buf = PRINTBUF; + + prt_printf(&buf, "bucket: "); + bch2_bpos_to_text(&buf, bucket); + prt_printf(&buf, " gen: %i\n", gen); + + trace_io_move_evacuate_bucket(c, buf.buf); + printbuf_exit(&buf); } struct moving_io { @@ -298,7 +342,8 @@ int bch2_move_extent(struct moving_context *ctxt, struct bch_fs *c = trans->c; int ret = -ENOMEM; - trace_io_move2(c, k, &io_opts, &data_opts); + if (trace_io_move_enabled()) + trace_io_move2(c, k, &io_opts, &data_opts); this_cpu_add(c->counters[BCH_COUNTER_io_move], k.k->size); if (ctxt->stats) @@ -314,16 +359,14 @@ int bch2_move_extent(struct moving_context *ctxt, return 0; } - /* - * Before memory allocations & taking nocow locks in - * bch2_data_update_init(): - */ - bch2_trans_unlock(trans); - - struct moving_io *io = kzalloc(sizeof(struct moving_io), GFP_KERNEL); + struct moving_io *io = allocate_dropping_locks(trans, ret, + kzalloc(sizeof(struct moving_io), _gfp)); if (!io) goto err; + if (ret) + goto err_free; + INIT_LIST_HEAD(&io->io_list); io->write.ctxt = ctxt; io->read_sectors = k.k->size; @@ -343,6 +386,8 @@ int bch2_move_extent(struct moving_context *ctxt, io->write.op.c = c; io->write.data_opts = data_opts; + bch2_trans_unlock(trans); + ret = bch2_data_update_bios_init(&io->write, c, &io_opts); if (ret) goto err_free; @@ -364,7 +409,8 @@ int bch2_move_extent(struct moving_context *ctxt, atomic_inc(&io->b->count); } - trace_io_move_read2(c, k); + if (trace_io_move_read_enabled()) + trace_io_move_read2(c, k); mutex_lock(&ctxt->lock); atomic_add(io->read_sectors, &ctxt->read_sectors); @@ -390,9 +436,6 @@ int bch2_move_extent(struct moving_context *ctxt, err_free: kfree(io); err: - if (bch2_err_matches(ret, BCH_ERR_data_update_done)) - return 0; - if (bch2_err_matches(ret, EROFS) || bch2_err_matches(ret, BCH_ERR_transaction_restart)) return ret; @@ -408,6 +451,9 @@ err: trace_io_move_start_fail(c, buf.buf); printbuf_exit(&buf); } + + if (bch2_err_matches(ret, BCH_ERR_data_update_done)) + return 0; return ret; } @@ -496,6 +542,7 @@ int bch2_move_get_io_opts_one(struct btree_trans *trans, bch2_inode_opts_get(io_opts, c, &inode); } bch2_trans_iter_exit(trans, &inode_iter); + /* seem to be spinning here? */ out: return bch2_get_update_rebalance_opts(trans, io_opts, extent_iter, extent_k); } @@ -910,7 +957,13 @@ static int __bch2_move_data_phys(struct moving_context *ctxt, } struct data_update_opts data_opts = {}; - if (!pred(c, arg, bp.v->btree_id, k, &io_opts, &data_opts)) { + bool p = pred(c, arg, bp.v->btree_id, k, &io_opts, &data_opts); + + if (trace_io_move_pred_enabled()) + trace_io_move_pred2(c, k, &io_opts, &data_opts, + pred, arg, p); + + if (!p) { bch2_trans_iter_exit(trans, &iter); goto next; } @@ -918,7 +971,7 @@ static int __bch2_move_data_phys(struct moving_context *ctxt, if (data_opts.scrub && !bch2_dev_idx_is_online(c, data_opts.read_dev)) { bch2_trans_iter_exit(trans, &iter); - ret = -BCH_ERR_device_offline; + ret = bch_err_throw(c, device_offline); break; } @@ -993,12 +1046,6 @@ int bch2_move_data_phys(struct bch_fs *c, return ret; } -struct evacuate_bucket_arg { - struct bpos bucket; - int gen; - struct data_update_opts data_opts; -}; - static bool evacuate_bucket_pred(struct bch_fs *c, void *_arg, enum btree_id btree, struct bkey_s_c k, struct bch_io_opts *io_opts, @@ -1025,8 +1072,13 @@ int bch2_evacuate_bucket(struct moving_context *ctxt, struct bpos bucket, int gen, struct data_update_opts data_opts) { + struct bch_fs *c = ctxt->trans->c; struct evacuate_bucket_arg arg = { bucket, gen, data_opts, }; + count_event(c, io_move_evacuate_bucket); + if (trace_io_move_evacuate_bucket_enabled()) + trace_io_move_evacuate_bucket2(c, bucket, gen); + return __bch2_move_data_phys(ctxt, bucket_in_flight, bucket.inode, bucket.offset, @@ -1124,7 +1176,7 @@ static bool rereplicate_pred(struct bch_fs *c, void *arg, ? c->opts.metadata_replicas : io_opts->data_replicas; - rcu_read_lock(); + guard(rcu)(); struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k); unsigned i = 0; bkey_for_each_ptr(ptrs, ptr) { @@ -1134,7 +1186,6 @@ static bool rereplicate_pred(struct bch_fs *c, void *arg, data_opts->kill_ptrs |= BIT(i); i++; } - rcu_read_unlock(); if (!data_opts->kill_ptrs && (!nr_good || nr_good >= replicas)) @@ -1242,7 +1293,7 @@ static bool drop_extra_replicas_pred(struct bch_fs *c, void *arg, struct extent_ptr_decoded p; unsigned i = 0; - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr_decode(k.k, bch2_bkey_ptrs_c(k), p, entry) { unsigned d = bch2_extent_ptr_durability(c, &p); @@ -1253,7 +1304,6 @@ static bool drop_extra_replicas_pred(struct bch_fs *c, void *arg, i++; } - rcu_read_unlock(); return data_opts->kill_ptrs != 0; } diff --git a/fs/bcachefs/movinggc.c b/fs/bcachefs/movinggc.c index e7a2a13554d7..5e6de91a8763 100644 --- a/fs/bcachefs/movinggc.c +++ b/fs/bcachefs/movinggc.c @@ -28,7 +28,7 @@ #include <linux/wait.h> struct buckets_in_flight { - struct rhashtable table; + struct rhashtable *table; struct move_bucket *first; struct move_bucket *last; size_t nr; @@ -71,7 +71,7 @@ static int bch2_bucket_is_movable(struct btree_trans *trans, if (ret) return ret; - struct bch_dev *ca = bch2_dev_tryget(c, k.k->p.inode); + struct bch_dev *ca = bch2_dev_bucket_tryget(c, k.k->p); if (!ca) goto out; @@ -98,7 +98,7 @@ out: static void move_bucket_free(struct buckets_in_flight *list, struct move_bucket *b) { - int ret = rhashtable_remove_fast(&list->table, &b->hash, + int ret = rhashtable_remove_fast(list->table, &b->hash, bch_move_bucket_params); BUG_ON(ret); kfree(b); @@ -133,7 +133,7 @@ static void move_buckets_wait(struct moving_context *ctxt, static bool bucket_in_flight(struct buckets_in_flight *list, struct move_bucket_key k) { - return rhashtable_lookup_fast(&list->table, &k, bch_move_bucket_params); + return rhashtable_lookup_fast(list->table, &k, bch_move_bucket_params); } static int bch2_copygc_get_buckets(struct moving_context *ctxt, @@ -185,7 +185,7 @@ static int bch2_copygc_get_buckets(struct moving_context *ctxt, goto err; } - ret2 = rhashtable_lookup_insert_fast(&buckets_in_flight->table, &b_i->hash, + ret2 = rhashtable_lookup_insert_fast(buckets_in_flight->table, &b_i->hash, bch_move_bucket_params); BUG_ON(ret2); @@ -293,11 +293,9 @@ u64 bch2_copygc_wait_amount(struct bch_fs *c) { u64 wait = U64_MAX; - rcu_read_lock(); + guard(rcu)(); for_each_rw_member_rcu(c, ca) wait = min(wait, bch2_copygc_dev_wait_amount(ca)); - rcu_read_unlock(); - return wait; } @@ -321,21 +319,21 @@ void bch2_copygc_wait_to_text(struct printbuf *out, struct bch_fs *c) bch2_printbuf_make_room(out, 4096); - rcu_read_lock(); + struct task_struct *t; out->atomic++; + scoped_guard(rcu) { + prt_printf(out, "Currently calculated wait:\n"); + for_each_rw_member_rcu(c, ca) { + prt_printf(out, " %s:\t", ca->name); + prt_human_readable_u64(out, bch2_copygc_dev_wait_amount(ca)); + prt_newline(out); + } - prt_printf(out, "Currently calculated wait:\n"); - for_each_rw_member_rcu(c, ca) { - prt_printf(out, " %s:\t", ca->name); - prt_human_readable_u64(out, bch2_copygc_dev_wait_amount(ca)); - prt_newline(out); + t = rcu_dereference(c->copygc_thread); + if (t) + get_task_struct(t); } - - struct task_struct *t = rcu_dereference(c->copygc_thread); - if (t) - get_task_struct(t); --out->atomic; - rcu_read_unlock(); if (t) { bch2_prt_task_backtrace(out, t, 0, GFP_KERNEL); @@ -352,10 +350,13 @@ static int bch2_copygc_thread(void *arg) struct buckets_in_flight buckets = {}; u64 last, wait; - int ret = rhashtable_init(&buckets.table, &bch_move_bucket_params); + buckets.table = kzalloc(sizeof(*buckets.table), GFP_KERNEL); + int ret = !buckets.table + ? -ENOMEM + : rhashtable_init(buckets.table, &bch_move_bucket_params); bch_err_msg(c, ret, "allocating copygc buckets in flight"); if (ret) - return ret; + goto err; set_freezable(); @@ -423,11 +424,12 @@ static int bch2_copygc_thread(void *arg) } move_buckets_wait(&ctxt, &buckets, true); - rhashtable_destroy(&buckets.table); + rhashtable_destroy(buckets.table); bch2_moving_ctxt_exit(&ctxt); bch2_move_stats_exit(&move_stats, c); - - return 0; +err: + kfree(buckets.table); + return ret; } void bch2_copygc_stop(struct bch_fs *c) diff --git a/fs/bcachefs/movinggc.h b/fs/bcachefs/movinggc.h index b9683d22bab0..f615910d6f98 100644 --- a/fs/bcachefs/movinggc.h +++ b/fs/bcachefs/movinggc.h @@ -7,11 +7,10 @@ void bch2_copygc_wait_to_text(struct printbuf *, struct bch_fs *); static inline void bch2_copygc_wakeup(struct bch_fs *c) { - rcu_read_lock(); + guard(rcu)(); struct task_struct *p = rcu_dereference(c->copygc_thread); if (p) wake_up_process(p); - rcu_read_unlock(); } void bch2_copygc_stop(struct bch_fs *); diff --git a/fs/bcachefs/namei.c b/fs/bcachefs/namei.c index a84b69d6caef..c3f87c59922d 100644 --- a/fs/bcachefs/namei.c +++ b/fs/bcachefs/namei.c @@ -175,6 +175,16 @@ int bch2_create_trans(struct btree_trans *trans, new_inode->bi_dir_offset = dir_offset; } + if (S_ISDIR(mode)) { + ret = bch2_maybe_propagate_has_case_insensitive(trans, + (subvol_inum) { + new_inode->bi_subvol ?: dir.subvol, + new_inode->bi_inum }, + new_inode); + if (ret) + goto err; + } + if (S_ISDIR(mode) && !new_inode->bi_subvol) new_inode->bi_depth = dir_u->bi_depth + 1; @@ -287,7 +297,7 @@ int bch2_unlink_trans(struct btree_trans *trans, } if (deleting_subvol && !inode_u->bi_subvol) { - ret = -BCH_ERR_ENOENT_not_subvol; + ret = bch_err_throw(c, ENOENT_not_subvol); goto err; } @@ -425,8 +435,8 @@ int bch2_rename_trans(struct btree_trans *trans, } ret = bch2_dirent_rename(trans, - src_dir, &src_hash, &src_dir_u->bi_size, - dst_dir, &dst_hash, &dst_dir_u->bi_size, + src_dir, &src_hash, + dst_dir, &dst_hash, src_name, &src_inum, &src_offset, dst_name, &dst_inum, &dst_offset, mode); @@ -615,14 +625,26 @@ static int __bch2_inum_to_path(struct btree_trans *trans, { unsigned orig_pos = path->pos; int ret = 0; + DARRAY(subvol_inum) inums = {}; + + if (!snapshot) { + ret = bch2_subvolume_get_snapshot(trans, subvol, &snapshot); + if (ret) + goto disconnected; + } while (true) { - if (!snapshot) { - ret = bch2_subvolume_get_snapshot(trans, subvol, &snapshot); - if (ret) - goto disconnected; + subvol_inum n = (subvol_inum) { subvol ?: snapshot, inum }; + + if (darray_find_p(inums, i, i->subvol == n.subvol && i->inum == n.inum)) { + prt_str_reversed(path, "(loop)"); + break; } + ret = darray_push(&inums, n); + if (ret) + goto err; + struct bch_inode_unpacked inode; ret = bch2_inode_find_by_inum_snapshot(trans, inum, snapshot, &inode, 0); if (ret) @@ -633,14 +655,16 @@ static int __bch2_inum_to_path(struct btree_trans *trans, break; if (!inode.bi_dir && !inode.bi_dir_offset) { - ret = -BCH_ERR_ENOENT_inode_no_backpointer; + ret = bch_err_throw(trans->c, ENOENT_inode_no_backpointer); goto disconnected; } inum = inode.bi_dir; if (inode.bi_parent_subvol) { subvol = inode.bi_parent_subvol; - snapshot = 0; + ret = bch2_subvolume_get_snapshot(trans, inode.bi_parent_subvol, &snapshot); + if (ret) + goto disconnected; } struct btree_iter d_iter; @@ -652,6 +676,7 @@ static int __bch2_inum_to_path(struct btree_trans *trans, goto disconnected; struct qstr dirent_name = bch2_dirent_get_name(d); + prt_bytes_reversed(path, dirent_name.name, dirent_name.len); prt_char(path, '/'); @@ -667,8 +692,10 @@ out: goto err; reverse_bytes(path->buf + orig_pos, path->pos - orig_pos); + darray_exit(&inums); return 0; err: + darray_exit(&inums); return ret; disconnected: if (bch2_err_matches(ret, BCH_ERR_transaction_restart)) @@ -707,8 +734,7 @@ static int bch2_check_dirent_inode_dirent(struct btree_trans *trans, if (inode_points_to_dirent(target, d)) return 0; - if (!target->bi_dir && - !target->bi_dir_offset) { + if (!bch2_inode_has_backpointer(target)) { fsck_err_on(S_ISDIR(target->bi_mode), trans, inode_dir_missing_backpointer, "directory with missing backpointer\n%s", @@ -733,15 +759,6 @@ static int bch2_check_dirent_inode_dirent(struct btree_trans *trans, return __bch2_fsck_write_inode(trans, target); } - if (bch2_inode_should_have_single_bp(target) && - !fsck_err(trans, inode_wrong_backpointer, - "dirent points to inode that does not point back:\n%s", - (bch2_bkey_val_to_text(&buf, c, d.s_c), - prt_newline(&buf), - bch2_inode_unpacked_to_text(&buf, target), - buf.buf))) - goto err; - struct bkey_s_c_dirent bp_dirent = bch2_bkey_get_iter_typed(trans, &bp_iter, BTREE_ID_dirents, SPOS(target->bi_dir, target->bi_dir_offset, target->bi_snapshot), @@ -768,6 +785,7 @@ static int bch2_check_dirent_inode_dirent(struct btree_trans *trans, ret = __bch2_fsck_write_inode(trans, target); } } else { + printbuf_reset(&buf); bch2_bkey_val_to_text(&buf, c, d.s_c); prt_newline(&buf); bch2_bkey_val_to_text(&buf, c, bp_dirent.s_c); @@ -857,7 +875,8 @@ int __bch2_check_dirent_target(struct btree_trans *trans, n->v.d_inum = cpu_to_le64(target->bi_inum); } - ret = bch2_trans_update(trans, dirent_iter, &n->k_i, 0); + ret = bch2_trans_update(trans, dirent_iter, &n->k_i, + BTREE_UPDATE_internal_snapshot_node); if (ret) goto err; } diff --git a/fs/bcachefs/opts.h b/fs/bcachefs/opts.h index 2a02606254b3..63f8e254495c 100644 --- a/fs/bcachefs/opts.h +++ b/fs/bcachefs/opts.h @@ -234,6 +234,11 @@ enum fsck_err_opts { OPT_BOOL(), \ BCH_SB_CASEFOLD, false, \ NULL, "Dirent lookups are casefolded") \ + x(casefold_disabled, u8, \ + OPT_FS|OPT_MOUNT, \ + OPT_BOOL(), \ + BCH2_NO_SB_OPT, false, \ + NULL, "Disable casefolding filesystem wide") \ x(inodes_32bit, u8, \ OPT_FS|OPT_INODE|OPT_FORMAT|OPT_MOUNT|OPT_RUNTIME, \ OPT_BOOL(), \ @@ -379,6 +384,11 @@ enum fsck_err_opts { OPT_BOOL(), \ BCH2_NO_SB_OPT, false, \ NULL, "Exit recovery immediately prior to journal replay")\ + x(journal_rewind, u64, \ + OPT_FS|OPT_MOUNT, \ + OPT_UINT(0, U64_MAX), \ + BCH2_NO_SB_OPT, 0, \ + NULL, "Rewind journal") \ x(recovery_passes, u64, \ OPT_FS|OPT_MOUNT, \ OPT_BITFIELD(bch2_recovery_passes), \ diff --git a/fs/bcachefs/printbuf.h b/fs/bcachefs/printbuf.h index 1ca476adbf6f..8f4e28d440ac 100644 --- a/fs/bcachefs/printbuf.h +++ b/fs/bcachefs/printbuf.h @@ -140,6 +140,14 @@ void bch2_prt_bitflags_vector(struct printbuf *, const char * const[], .size = _size, \ }) +static inline struct printbuf bch2_printbuf_init(void) +{ + return PRINTBUF; +} + +DEFINE_CLASS(printbuf, struct printbuf, + bch2_printbuf_exit(&_T), bch2_printbuf_init(), void) + /* * Returns size remaining of output buffer: */ diff --git a/fs/bcachefs/quota.c b/fs/bcachefs/quota.c index 3d4755d73af7..f241efb1fb50 100644 --- a/fs/bcachefs/quota.c +++ b/fs/bcachefs/quota.c @@ -527,7 +527,7 @@ int bch2_fs_quota_read(struct bch_fs *c) struct bch_sb_field_quota *sb_quota = bch2_sb_get_or_create_quota(&c->disk_sb); if (!sb_quota) { mutex_unlock(&c->sb_lock); - return -BCH_ERR_ENOSPC_sb_quota; + return bch_err_throw(c, ENOSPC_sb_quota); } bch2_sb_quota_read(c); @@ -572,7 +572,7 @@ static int bch2_quota_enable(struct super_block *sb, unsigned uflags) mutex_lock(&c->sb_lock); sb_quota = bch2_sb_get_or_create_quota(&c->disk_sb); if (!sb_quota) { - ret = -BCH_ERR_ENOSPC_sb_quota; + ret = bch_err_throw(c, ENOSPC_sb_quota); goto unlock; } @@ -726,7 +726,7 @@ static int bch2_quota_set_info(struct super_block *sb, int type, mutex_lock(&c->sb_lock); sb_quota = bch2_sb_get_or_create_quota(&c->disk_sb); if (!sb_quota) { - ret = -BCH_ERR_ENOSPC_sb_quota; + ret = bch_err_throw(c, ENOSPC_sb_quota); goto unlock; } diff --git a/fs/bcachefs/rcu_pending.c b/fs/bcachefs/rcu_pending.c index bef2aa1b8bcd..b1438be9d690 100644 --- a/fs/bcachefs/rcu_pending.c +++ b/fs/bcachefs/rcu_pending.c @@ -182,11 +182,6 @@ static inline void kfree_bulk(size_t nr, void ** p) while (nr--) kfree(*p); } - -#define local_irq_save(flags) \ -do { \ - flags = 0; \ -} while (0) #endif static noinline void __process_finished_items(struct rcu_pending *pending, @@ -429,9 +424,15 @@ __rcu_pending_enqueue(struct rcu_pending *pending, struct rcu_head *head, BUG_ON((ptr != NULL) != (pending->process == RCU_PENDING_KVFREE_FN)); - local_irq_save(flags); - p = this_cpu_ptr(pending->p); - spin_lock(&p->lock); + /* We could technically be scheduled before taking the lock and end up + * using a different cpu's rcu_pending_pcpu: that's ok, it needs a lock + * anyways + * + * And we have to do it this way to avoid breaking PREEMPT_RT, which + * redefines how spinlocks work: + */ + p = raw_cpu_ptr(pending->p); + spin_lock_irqsave(&p->lock, flags); rcu_gp_poll_state_t seq = __get_state_synchronize_rcu(pending->srcu); restart: if (may_sleep && @@ -520,9 +521,8 @@ check_expired: goto free_node; } - local_irq_save(flags); - p = this_cpu_ptr(pending->p); - spin_lock(&p->lock); + p = raw_cpu_ptr(pending->p); + spin_lock_irqsave(&p->lock, flags); goto restart; } diff --git a/fs/bcachefs/rebalance.c b/fs/bcachefs/rebalance.c index de1ec9e0caa0..1c345b86b1c0 100644 --- a/fs/bcachefs/rebalance.c +++ b/fs/bcachefs/rebalance.c @@ -80,13 +80,12 @@ static inline unsigned bch2_bkey_ptrs_need_move(struct bch_fs *c, unsigned ptr_bit = 1; unsigned rewrite_ptrs = 0; - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr(ptrs, ptr) { if (!ptr->cached && !bch2_dev_in_target(c, ptr->dev, opts->background_target)) rewrite_ptrs |= ptr_bit; ptr_bit <<= 1; } - rcu_read_unlock(); return rewrite_ptrs; } @@ -135,12 +134,11 @@ u64 bch2_bkey_sectors_need_rebalance(struct bch_fs *c, struct bkey_s_c k) } incompressible: if (opts->background_target) { - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr_decode(k.k, ptrs, p, entry) if (!p.ptr.cached && !bch2_dev_in_target(c, p.ptr.dev, opts->background_target)) sectors += p.crc.compressed_size; - rcu_read_unlock(); } return sectors; @@ -445,7 +443,7 @@ static int do_rebalance_extent(struct moving_context *ctxt, if (bch2_err_matches(ret, ENOMEM)) { /* memory allocation failure, wait for some IO to finish */ bch2_move_ctxt_wait_for_io(ctxt); - ret = -BCH_ERR_transaction_restart_nested; + ret = bch_err_throw(c, transaction_restart_nested); } if (bch2_err_matches(ret, BCH_ERR_transaction_restart)) @@ -527,7 +525,7 @@ static void rebalance_wait(struct bch_fs *c) r->state = BCH_REBALANCE_waiting; } - bch2_kthread_io_clock_wait(clock, r->wait_iotime_end, MAX_SCHEDULE_TIMEOUT); + bch2_kthread_io_clock_wait_once(clock, r->wait_iotime_end, MAX_SCHEDULE_TIMEOUT); } static bool bch2_rebalance_enabled(struct bch_fs *c) @@ -544,6 +542,7 @@ static int do_rebalance(struct moving_context *ctxt) struct bch_fs_rebalance *r = &c->rebalance; struct btree_iter rebalance_work_iter, extent_iter = {}; struct bkey_s_c k; + u32 kick = r->kick; int ret = 0; bch2_trans_begin(trans); @@ -593,7 +592,8 @@ static int do_rebalance(struct moving_context *ctxt) if (!ret && !kthread_should_stop() && !atomic64_read(&r->work_stats.sectors_seen) && - !atomic64_read(&r->scan_stats.sectors_seen)) { + !atomic64_read(&r->scan_stats.sectors_seen) && + kick == r->kick) { bch2_moving_ctxt_flush_all(ctxt); bch2_trans_unlock_long(trans); rebalance_wait(c); @@ -677,11 +677,12 @@ void bch2_rebalance_status_to_text(struct printbuf *out, struct bch_fs *c) } prt_newline(out); - rcu_read_lock(); - struct task_struct *t = rcu_dereference(c->rebalance.thread); - if (t) - get_task_struct(t); - rcu_read_unlock(); + struct task_struct *t; + scoped_guard(rcu) { + t = rcu_dereference(c->rebalance.thread); + if (t) + get_task_struct(t); + } if (t) { bch2_prt_task_backtrace(out, t, 0, GFP_KERNEL); @@ -794,7 +795,7 @@ static int check_rebalance_work_one(struct btree_trans *trans, BTREE_ID_extents, POS_MIN, BTREE_ITER_prefetch| BTREE_ITER_all_snapshots); - return -BCH_ERR_transaction_restart_nested; + return bch_err_throw(c, transaction_restart_nested); } if (!extent_k.k && !rebalance_k.k) diff --git a/fs/bcachefs/rebalance.h b/fs/bcachefs/rebalance.h index 5d9214fe1a22..7a565ea7dbfc 100644 --- a/fs/bcachefs/rebalance.h +++ b/fs/bcachefs/rebalance.h @@ -39,13 +39,11 @@ int bch2_set_fs_needs_rebalance(struct bch_fs *); static inline void bch2_rebalance_wakeup(struct bch_fs *c) { - struct task_struct *p; - - rcu_read_lock(); - p = rcu_dereference(c->rebalance.thread); + c->rebalance.kick++; + guard(rcu)(); + struct task_struct *p = rcu_dereference(c->rebalance.thread); if (p) wake_up_process(p); - rcu_read_unlock(); } void bch2_rebalance_status_to_text(struct printbuf *, struct bch_fs *); diff --git a/fs/bcachefs/rebalance_types.h b/fs/bcachefs/rebalance_types.h index 33d77286f1d5..c659da149fa3 100644 --- a/fs/bcachefs/rebalance_types.h +++ b/fs/bcachefs/rebalance_types.h @@ -18,6 +18,7 @@ enum bch_rebalance_states { struct bch_fs_rebalance { struct task_struct __rcu *thread; + u32 kick; struct bch_pd_controller pd; enum bch_rebalance_states state; diff --git a/fs/bcachefs/recovery.c b/fs/bcachefs/recovery.c index 4fca57575565..c94debb12d2f 100644 --- a/fs/bcachefs/recovery.c +++ b/fs/bcachefs/recovery.c @@ -99,9 +99,11 @@ int bch2_btree_lost_data(struct bch_fs *c, goto out; case BTREE_ID_snapshots: ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_reconstruct_snapshots, 0) ?: ret; + ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_check_topology, 0) ?: ret; ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_scan_for_btree_nodes, 0) ?: ret; goto out; default: + ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_check_topology, 0) ?: ret; ret = __bch2_run_explicit_recovery_pass(c, msg, BCH_RECOVERY_PASS_scan_for_btree_nodes, 0) ?: ret; goto out; } @@ -272,6 +274,28 @@ static int bch2_journal_replay_key(struct btree_trans *trans, struct btree_path *path = btree_iter_path(trans, &iter); if (unlikely(!btree_path_node(path, k->level))) { + struct bch_fs *c = trans->c; + + CLASS(printbuf, buf)(); + prt_str(&buf, "btree="); + bch2_btree_id_to_text(&buf, k->btree_id); + prt_printf(&buf, " level=%u ", k->level); + bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(k->k)); + + if (!(c->recovery.passes_complete & (BIT_ULL(BCH_RECOVERY_PASS_scan_for_btree_nodes)| + BIT_ULL(BCH_RECOVERY_PASS_check_topology)))) { + bch_err(c, "have key in journal replay for btree depth that does not exist, confused\n%s", + buf.buf); + ret = -EINVAL; + } + + if (!k->allocated) { + bch_notice(c, "dropping key in journal replay for depth that does not exist because we're recovering from scan\n%s", + buf.buf); + k->overwritten = true; + goto out; + } + bch2_trans_iter_exit(trans, &iter); bch2_trans_node_iter_init(trans, &iter, k->btree_id, k->k->k.p, BTREE_MAX_DEPTH, 0, iter_flags); @@ -594,6 +618,7 @@ static int read_btree_roots(struct bch_fs *c) buf.buf, bch2_err_str(ret))) { if (btree_id_is_alloc(i)) r->error = 0; + ret = 0; } } @@ -679,7 +704,7 @@ static bool check_version_upgrade(struct bch_fs *c) ret = true; } - if (new_version > c->sb.version_incompat && + if (new_version > c->sb.version_incompat_allowed && c->opts.version_upgrade == BCH_VERSION_UPGRADE_incompatible) { struct printbuf buf = PRINTBUF; @@ -739,7 +764,24 @@ int bch2_fs_recovery(struct bch_fs *c) ? min(c->opts.recovery_pass_last, BCH_RECOVERY_PASS_snapshots_read) : BCH_RECOVERY_PASS_snapshots_read; c->opts.nochanges = true; + } + + if (c->opts.nochanges) c->opts.read_only = true; + + if (c->opts.journal_rewind) { + bch_info(c, "rewinding journal, fsck required"); + c->opts.fsck = true; + } + + if (go_rw_in_recovery(c)) { + /* + * start workqueues/kworkers early - kthread creation checks for + * pending signals, which is _very_ annoying + */ + ret = bch2_fs_init_rw(c); + if (ret) + goto err; } mutex_lock(&c->sb_lock); @@ -879,7 +921,7 @@ int bch2_fs_recovery(struct bch_fs *c) use_clean: if (!clean) { bch_err(c, "no superblock clean section found"); - ret = -BCH_ERR_fsck_repair_impossible; + ret = bch_err_throw(c, fsck_repair_impossible); goto err; } @@ -950,7 +992,7 @@ use_clean: ret = bch2_journal_log_msg(c, "starting journal at entry %llu, replaying %llu-%llu", journal_seq, last_seq, blacklist_seq - 1) ?: - bch2_fs_journal_start(&c->journal, journal_seq); + bch2_fs_journal_start(&c->journal, last_seq, journal_seq); if (ret) goto err; @@ -1093,13 +1135,6 @@ use_clean: out: bch2_flush_fsck_errs(c); - if (!c->opts.retain_recovery_info) { - bch2_journal_keys_put_initial(c); - bch2_find_btree_nodes_exit(&c->found_btree_nodes); - } - if (!IS_ERR(clean)) - kfree(clean); - if (!ret && test_bit(BCH_FS_need_delete_dead_snapshots, &c->flags) && !c->opts.nochanges) { @@ -1108,6 +1143,9 @@ out: } bch_err_fn(c, ret); +final_out: + if (!IS_ERR(clean)) + kfree(clean); return ret; err: fsck_err: @@ -1115,13 +1153,13 @@ fsck_err: struct printbuf buf = PRINTBUF; bch2_log_msg_start(c, &buf); - prt_printf(&buf, "error in recovery: %s", bch2_err_str(ret)); + prt_printf(&buf, "error in recovery: %s\n", bch2_err_str(ret)); bch2_fs_emergency_read_only2(c, &buf); bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); } - return ret; + goto final_out; } int bch2_fs_initialize(struct bch_fs *c) @@ -1170,7 +1208,7 @@ int bch2_fs_initialize(struct bch_fs *c) * journal_res_get() will crash if called before this has * set up the journal.pin FIFO and journal.cur pointer: */ - ret = bch2_fs_journal_start(&c->journal, 1); + ret = bch2_fs_journal_start(&c->journal, 1, 1); if (ret) goto err; diff --git a/fs/bcachefs/recovery_passes.c b/fs/bcachefs/recovery_passes.c index dabb29b08ad0..6a039e011064 100644 --- a/fs/bcachefs/recovery_passes.c +++ b/fs/bcachefs/recovery_passes.c @@ -103,20 +103,20 @@ static void bch2_sb_recovery_passes_to_text(struct printbuf *out, prt_tab(out); bch2_pr_time_units(out, le32_to_cpu(i->last_runtime) * NSEC_PER_SEC); + + if (BCH_RECOVERY_PASS_NO_RATELIMIT(i)) + prt_str(out, " (no ratelimit)"); + prt_newline(out); } } -static void bch2_sb_recovery_pass_complete(struct bch_fs *c, - enum bch_recovery_pass pass, - s64 start_time) +static struct recovery_pass_entry *bch2_sb_recovery_pass_entry(struct bch_fs *c, + enum bch_recovery_pass pass) { enum bch_recovery_pass_stable stable = bch2_recovery_pass_to_stable(pass); - s64 end_time = ktime_get_real_seconds(); - mutex_lock(&c->sb_lock); - struct bch_sb_field_ext *ext = bch2_sb_field_get(c->disk_sb.sb, ext); - __clear_bit_le64(stable, ext->recovery_passes_required); + lockdep_assert_held(&c->sb_lock); struct bch_sb_field_recovery_passes *r = bch2_sb_field_get(c->disk_sb.sb, recovery_passes); @@ -127,15 +127,43 @@ static void bch2_sb_recovery_pass_complete(struct bch_fs *c, r = bch2_sb_field_resize(&c->disk_sb, recovery_passes, u64s); if (!r) { bch_err(c, "error creating recovery_passes sb section"); - goto out; + return NULL; } } - r->start[stable].last_run = cpu_to_le64(end_time); - r->start[stable].last_runtime = cpu_to_le32(max(0, end_time - start_time)); -out: + return r->start + stable; +} + +static void bch2_sb_recovery_pass_complete(struct bch_fs *c, + enum bch_recovery_pass pass, + s64 start_time) +{ + guard(mutex)(&c->sb_lock); + struct bch_sb_field_ext *ext = bch2_sb_field_get(c->disk_sb.sb, ext); + __clear_bit_le64(bch2_recovery_pass_to_stable(pass), + ext->recovery_passes_required); + + struct recovery_pass_entry *e = bch2_sb_recovery_pass_entry(c, pass); + if (e) { + s64 end_time = ktime_get_real_seconds(); + e->last_run = cpu_to_le64(end_time); + e->last_runtime = cpu_to_le32(max(0, end_time - start_time)); + SET_BCH_RECOVERY_PASS_NO_RATELIMIT(e, false); + } + bch2_write_super(c); - mutex_unlock(&c->sb_lock); +} + +void bch2_recovery_pass_set_no_ratelimit(struct bch_fs *c, + enum bch_recovery_pass pass) +{ + guard(mutex)(&c->sb_lock); + + struct recovery_pass_entry *e = bch2_sb_recovery_pass_entry(c, pass); + if (e && !BCH_RECOVERY_PASS_NO_RATELIMIT(e)) { + SET_BCH_RECOVERY_PASS_NO_RATELIMIT(e, false); + bch2_write_super(c); + } } static bool bch2_recovery_pass_want_ratelimit(struct bch_fs *c, enum bch_recovery_pass pass) @@ -157,6 +185,9 @@ static bool bch2_recovery_pass_want_ratelimit(struct bch_fs *c, enum bch_recover */ ret = (u64) le32_to_cpu(i->last_runtime) * 100 > ktime_get_real_seconds() - le64_to_cpu(i->last_run); + + if (BCH_RECOVERY_PASS_NO_RATELIMIT(i)) + ret = false; } return ret; @@ -186,11 +217,7 @@ static int bch2_set_may_go_rw(struct bch_fs *c) set_bit(BCH_FS_may_go_rw, &c->flags); - if (keys->nr || - !c->opts.read_only || - !c->sb.clean || - c->opts.recovery_passes || - (c->opts.fsck && !(c->sb.features & BIT_ULL(BCH_FEATURE_no_alloc_info)))) { + if (go_rw_in_recovery(c)) { if (c->sb.features & BIT_ULL(BCH_FEATURE_no_alloc_info)) { bch_info(c, "mounting a filesystem with no alloc info read-write; will recreate"); bch2_reconstruct_alloc(c); @@ -263,8 +290,13 @@ static bool recovery_pass_needs_set(struct bch_fs *c, enum bch_run_recovery_pass_flags *flags) { struct bch_fs_recovery *r = &c->recovery; - bool in_recovery = test_bit(BCH_FS_in_recovery, &c->flags); - bool persistent = !in_recovery || !(*flags & RUN_RECOVERY_PASS_nopersistent); + + /* + * Never run scan_for_btree_nodes persistently: check_topology will run + * it if required + */ + if (pass == BCH_RECOVERY_PASS_scan_for_btree_nodes) + *flags |= RUN_RECOVERY_PASS_nopersistent; if ((*flags & RUN_RECOVERY_PASS_ratelimit) && !bch2_recovery_pass_want_ratelimit(c, pass)) @@ -279,6 +311,11 @@ static bool recovery_pass_needs_set(struct bch_fs *c, * Otherwise, we run run_explicit_recovery_pass when we find damage, so * it should run again even if it's already run: */ + bool in_recovery = test_bit(BCH_FS_in_recovery, &c->flags); + bool persistent = !in_recovery || !(*flags & RUN_RECOVERY_PASS_nopersistent); + bool rewind = in_recovery && + r->curr_pass > pass && + !(r->passes_complete & BIT_ULL(pass)); if (persistent ? !(c->sb.recovery_passes_required & BIT_ULL(pass)) @@ -289,6 +326,9 @@ static bool recovery_pass_needs_set(struct bch_fs *c, (r->passes_ratelimiting & BIT_ULL(pass))) return true; + if (rewind) + return true; + return false; } @@ -315,10 +355,12 @@ int __bch2_run_explicit_recovery_pass(struct bch_fs *c, goto out; bool in_recovery = test_bit(BCH_FS_in_recovery, &c->flags); - bool rewind = in_recovery && r->curr_pass > pass; + bool rewind = in_recovery && + r->curr_pass > pass && + !(r->passes_complete & BIT_ULL(pass)); bool ratelimit = flags & RUN_RECOVERY_PASS_ratelimit; - if (!(in_recovery && (flags & RUN_RECOVERY_PASS_nopersistent))) { + if (!(flags & RUN_RECOVERY_PASS_nopersistent)) { struct bch_sb_field_ext *ext = bch2_sb_field_get(c->disk_sb.sb, ext); __set_bit_le64(bch2_recovery_pass_to_stable(pass), ext->recovery_passes_required); } @@ -327,7 +369,7 @@ int __bch2_run_explicit_recovery_pass(struct bch_fs *c, (!in_recovery || r->curr_pass >= BCH_RECOVERY_PASS_set_may_go_rw)) { prt_printf(out, "need recovery pass %s (%u), but already rw\n", bch2_recovery_passes[pass], pass); - ret = -BCH_ERR_cannot_rewind_recovery; + ret = bch_err_throw(c, cannot_rewind_recovery); goto out; } @@ -347,7 +389,7 @@ int __bch2_run_explicit_recovery_pass(struct bch_fs *c, if (rewind) { r->next_pass = pass; r->passes_complete &= (1ULL << pass) >> 1; - ret = -BCH_ERR_restart_recovery; + ret = bch_err_throw(c, restart_recovery); } } else { prt_printf(out, "scheduling recovery pass %s (%u)%s\n", @@ -371,10 +413,8 @@ int bch2_run_explicit_recovery_pass(struct bch_fs *c, { int ret = 0; - scoped_guard(mutex, &c->sb_lock) { - if (!recovery_pass_needs_set(c, pass, &flags)) - return 0; - + if (recovery_pass_needs_set(c, pass, &flags)) { + guard(mutex)(&c->sb_lock); ret = __bch2_run_explicit_recovery_pass(c, out, pass, flags); bch2_write_super(c); } @@ -382,9 +422,38 @@ int bch2_run_explicit_recovery_pass(struct bch_fs *c, return ret; } +/* + * Returns 0 if @pass has run recently, otherwise one of + * -BCH_ERR_restart_recovery + * -BCH_ERR_recovery_pass_will_run + */ +int bch2_require_recovery_pass(struct bch_fs *c, + struct printbuf *out, + enum bch_recovery_pass pass) +{ + if (test_bit(BCH_FS_in_recovery, &c->flags) && + c->recovery.passes_complete & BIT_ULL(pass)) + return 0; + + guard(mutex)(&c->sb_lock); + + if (bch2_recovery_pass_want_ratelimit(c, pass)) + return 0; + + enum bch_run_recovery_pass_flags flags = 0; + int ret = 0; + + if (recovery_pass_needs_set(c, pass, &flags)) { + ret = __bch2_run_explicit_recovery_pass(c, out, pass, flags); + bch2_write_super(c); + } + + return ret ?: bch_err_throw(c, recovery_pass_will_run); +} + int bch2_run_print_explicit_recovery_pass(struct bch_fs *c, enum bch_recovery_pass pass) { - enum bch_run_recovery_pass_flags flags = RUN_RECOVERY_PASS_nopersistent; + enum bch_run_recovery_pass_flags flags = 0; if (!recovery_pass_needs_set(c, pass, &flags)) return 0; diff --git a/fs/bcachefs/recovery_passes.h b/fs/bcachefs/recovery_passes.h index dc0d2014ff9b..2117f0ce1922 100644 --- a/fs/bcachefs/recovery_passes.h +++ b/fs/bcachefs/recovery_passes.h @@ -10,11 +10,22 @@ u64 bch2_recovery_passes_from_stable(u64 v); u64 bch2_fsck_recovery_passes(void); +void bch2_recovery_pass_set_no_ratelimit(struct bch_fs *, enum bch_recovery_pass); + enum bch_run_recovery_pass_flags { RUN_RECOVERY_PASS_nopersistent = BIT(0), RUN_RECOVERY_PASS_ratelimit = BIT(1), }; +static inline bool go_rw_in_recovery(struct bch_fs *c) +{ + return (c->journal_keys.nr || + !c->opts.read_only || + !c->sb.clean || + c->opts.recovery_passes || + (c->opts.fsck && !(c->sb.features & BIT_ULL(BCH_FEATURE_no_alloc_info)))); +} + int bch2_run_print_explicit_recovery_pass(struct bch_fs *, enum bch_recovery_pass); int __bch2_run_explicit_recovery_pass(struct bch_fs *, struct printbuf *, @@ -24,6 +35,9 @@ int bch2_run_explicit_recovery_pass(struct bch_fs *, struct printbuf *, enum bch_recovery_pass, enum bch_run_recovery_pass_flags); +int bch2_require_recovery_pass(struct bch_fs *, struct printbuf *, + enum bch_recovery_pass); + int bch2_run_online_recovery_passes(struct bch_fs *, u64); int bch2_run_recovery_passes(struct bch_fs *, enum bch_recovery_pass); diff --git a/fs/bcachefs/recovery_passes_format.h b/fs/bcachefs/recovery_passes_format.h index c434eafbca19..b63c20558d3d 100644 --- a/fs/bcachefs/recovery_passes_format.h +++ b/fs/bcachefs/recovery_passes_format.h @@ -87,6 +87,8 @@ struct recovery_pass_entry { __le32 flags; }; +LE32_BITMASK(BCH_RECOVERY_PASS_NO_RATELIMIT, struct recovery_pass_entry, flags, 0, 1) + struct bch_sb_field_recovery_passes { struct bch_sb_field field; struct recovery_pass_entry start[]; diff --git a/fs/bcachefs/reflink.c b/fs/bcachefs/reflink.c index 3a13dbcab6ba..92b90cfe622b 100644 --- a/fs/bcachefs/reflink.c +++ b/fs/bcachefs/reflink.c @@ -64,6 +64,9 @@ void bch2_reflink_p_to_text(struct printbuf *out, struct bch_fs *c, REFLINK_P_IDX(p.v), le32_to_cpu(p.v->front_pad), le32_to_cpu(p.v->back_pad)); + + if (REFLINK_P_ERROR(p.v)) + prt_str(out, " error"); } bool bch2_reflink_p_merge(struct bch_fs *c, struct bkey_s _l, struct bkey_s_c _r) @@ -269,13 +272,12 @@ struct bkey_s_c bch2_lookup_indirect_extent(struct btree_trans *trans, return k; if (unlikely(!bkey_extent_is_reflink_data(k.k))) { - unsigned size = min((u64) k.k->size, - REFLINK_P_IDX(p.v) + p.k->size + le32_to_cpu(p.v->back_pad) - - reflink_offset); - bch2_key_resize(&iter->k, size); + u64 missing_end = min(k.k->p.offset, + REFLINK_P_IDX(p.v) + p.k->size + le32_to_cpu(p.v->back_pad)); + BUG_ON(reflink_offset == missing_end); int ret = bch2_indirect_extent_missing_error(trans, p, reflink_offset, - k.k->p.offset, should_commit); + missing_end, should_commit); if (ret) { bch2_trans_iter_exit(trans, iter); return bkey_s_c_err(ret); @@ -312,7 +314,7 @@ static int trans_trigger_reflink_p_segment(struct btree_trans *trans, if (!bkey_refcount_c(k)) { if (!(flags & BTREE_TRIGGER_overwrite)) - ret = -BCH_ERR_missing_indirect_extent; + ret = bch_err_throw(c, missing_indirect_extent); goto next; } @@ -612,7 +614,7 @@ s64 bch2_remap_range(struct bch_fs *c, int ret = 0, ret2 = 0; if (!enumerated_ref_tryget(&c->writes, BCH_WRITE_REF_reflink)) - return -BCH_ERR_erofs_no_writes; + return bch_err_throw(c, erofs_no_writes); bch2_check_set_feature(c, BCH_FEATURE_reflink); @@ -711,7 +713,8 @@ s64 bch2_remap_range(struct bch_fs *c, SET_REFLINK_P_IDX(&dst_p->v, offset); if (reflink_p_may_update_opts_field && - may_change_src_io_path_opts) + may_change_src_io_path_opts && + REFLINK_P_MAY_UPDATE_OPTIONS(src_p.v)) SET_REFLINK_P_MAY_UPDATE_OPTIONS(&dst_p->v, true); } else { BUG(); @@ -847,7 +850,7 @@ int bch2_gc_reflink_start(struct bch_fs *c) struct reflink_gc *r = genradix_ptr_alloc(&c->reflink_gc_table, c->reflink_gc_nr++, GFP_KERNEL); if (!r) { - ret = -BCH_ERR_ENOMEM_gc_reflink_start; + ret = bch_err_throw(c, ENOMEM_gc_reflink_start); break; } diff --git a/fs/bcachefs/replicas.c b/fs/bcachefs/replicas.c index 477ef0997949..8383bd7fdb3f 100644 --- a/fs/bcachefs/replicas.c +++ b/fs/bcachefs/replicas.c @@ -119,7 +119,7 @@ int bch2_replicas_entry_validate(struct bch_replicas_entry_v1 *r, return 0; bad: bch2_replicas_entry_to_text(err, r); - return -BCH_ERR_invalid_replicas_entry; + return bch_err_throw(c, invalid_replicas_entry); } void bch2_cpu_replicas_to_text(struct printbuf *out, @@ -311,7 +311,7 @@ static int bch2_mark_replicas_slowpath(struct bch_fs *c, !__replicas_has_entry(&c->replicas_gc, new_entry)) { new_gc = cpu_replicas_add_entry(c, &c->replicas_gc, new_entry); if (!new_gc.entries) { - ret = -BCH_ERR_ENOMEM_cpu_replicas; + ret = bch_err_throw(c, ENOMEM_cpu_replicas); goto err; } } @@ -319,7 +319,7 @@ static int bch2_mark_replicas_slowpath(struct bch_fs *c, if (!__replicas_has_entry(&c->replicas, new_entry)) { new_r = cpu_replicas_add_entry(c, &c->replicas, new_entry); if (!new_r.entries) { - ret = -BCH_ERR_ENOMEM_cpu_replicas; + ret = bch_err_throw(c, ENOMEM_cpu_replicas); goto err; } @@ -422,7 +422,7 @@ int bch2_replicas_gc_start(struct bch_fs *c, unsigned typemask) if (!c->replicas_gc.entries) { mutex_unlock(&c->sb_lock); bch_err(c, "error allocating c->replicas_gc"); - return -BCH_ERR_ENOMEM_replicas_gc; + return bch_err_throw(c, ENOMEM_replicas_gc); } for_each_cpu_replicas_entry(&c->replicas, e) @@ -458,7 +458,7 @@ retry: new.entries = kcalloc(nr, new.entry_size, GFP_KERNEL); if (!new.entries) { bch_err(c, "error allocating c->replicas_gc"); - return -BCH_ERR_ENOMEM_replicas_gc; + return bch_err_throw(c, ENOMEM_replicas_gc); } mutex_lock(&c->sb_lock); @@ -622,7 +622,7 @@ static int bch2_cpu_replicas_to_sb_replicas_v0(struct bch_fs *c, sb_r = bch2_sb_field_resize(&c->disk_sb, replicas_v0, DIV_ROUND_UP(bytes, sizeof(u64))); if (!sb_r) - return -BCH_ERR_ENOSPC_sb_replicas; + return bch_err_throw(c, ENOSPC_sb_replicas); bch2_sb_field_delete(&c->disk_sb, BCH_SB_FIELD_replicas); sb_r = bch2_sb_field_get(c->disk_sb.sb, replicas_v0); @@ -667,7 +667,7 @@ static int bch2_cpu_replicas_to_sb_replicas(struct bch_fs *c, sb_r = bch2_sb_field_resize(&c->disk_sb, replicas, DIV_ROUND_UP(bytes, sizeof(u64))); if (!sb_r) - return -BCH_ERR_ENOSPC_sb_replicas; + return bch_err_throw(c, ENOSPC_sb_replicas); bch2_sb_field_delete(&c->disk_sb, BCH_SB_FIELD_replicas_v0); sb_r = bch2_sb_field_get(c->disk_sb.sb, replicas); @@ -819,19 +819,18 @@ bool bch2_have_enough_devs(struct bch_fs *c, struct bch_devs_mask devs, if (e->data_type == BCH_DATA_cached) continue; - rcu_read_lock(); - for (unsigned i = 0; i < e->nr_devs; i++) { - if (e->devs[i] == BCH_SB_MEMBER_INVALID) { - nr_failed++; - continue; - } + scoped_guard(rcu) + for (unsigned i = 0; i < e->nr_devs; i++) { + if (e->devs[i] == BCH_SB_MEMBER_INVALID) { + nr_failed++; + continue; + } - nr_online += test_bit(e->devs[i], devs.d); + nr_online += test_bit(e->devs[i], devs.d); - struct bch_dev *ca = bch2_dev_rcu_noerror(c, e->devs[i]); - nr_failed += !ca || ca->mi.state == BCH_MEMBER_STATE_failed; - } - rcu_read_unlock(); + struct bch_dev *ca = bch2_dev_rcu_noerror(c, e->devs[i]); + nr_failed += !ca || ca->mi.state == BCH_MEMBER_STATE_failed; + } if (nr_online + nr_failed == e->nr_devs) continue; diff --git a/fs/bcachefs/sb-counters_format.h b/fs/bcachefs/sb-counters_format.h index 7c0c9c842b4e..b868702a431a 100644 --- a/fs/bcachefs/sb-counters_format.h +++ b/fs/bcachefs/sb-counters_format.h @@ -26,6 +26,7 @@ enum counters_flags { x(io_move_write_fail, 82, TYPE_COUNTER) \ x(io_move_start_fail, 39, TYPE_COUNTER) \ x(io_move_created_rebalance, 83, TYPE_COUNTER) \ + x(io_move_evacuate_bucket, 84, TYPE_COUNTER) \ x(bucket_invalidate, 3, TYPE_COUNTER) \ x(bucket_discard, 4, TYPE_COUNTER) \ x(bucket_discard_fast, 79, TYPE_COUNTER) \ diff --git a/fs/bcachefs/sb-downgrade.c b/fs/bcachefs/sb-downgrade.c index 861fce1630f0..1506d05e0665 100644 --- a/fs/bcachefs/sb-downgrade.c +++ b/fs/bcachefs/sb-downgrade.c @@ -253,6 +253,7 @@ DOWNGRADE_TABLE() static int downgrade_table_extra(struct bch_fs *c, darray_char *table) { + unsigned dst_offset = table->nr; struct bch_sb_field_downgrade_entry *dst = (void *) &darray_top(*table); unsigned bytes = sizeof(*dst) + sizeof(dst->errors[0]) * le16_to_cpu(dst->nr_errors); int ret = 0; @@ -268,6 +269,9 @@ static int downgrade_table_extra(struct bch_fs *c, darray_char *table) if (ret) return ret; + dst = (void *) &table->data[dst_offset]; + dst->nr_errors = cpu_to_le16(nr_errors + 1); + /* open coded __set_bit_le64, as dst is packed and * dst->recovery_passes is misaligned */ unsigned b = BCH_RECOVERY_PASS_STABLE_check_allocations; @@ -278,7 +282,6 @@ static int downgrade_table_extra(struct bch_fs *c, darray_char *table) break; } - dst->nr_errors = cpu_to_le16(nr_errors); return ret; } @@ -417,7 +420,7 @@ int bch2_sb_downgrade_update(struct bch_fs *c) d = bch2_sb_field_resize(&c->disk_sb, downgrade, sb_u64s); if (!d) { - ret = -BCH_ERR_ENOSPC_sb_downgrade; + ret = bch_err_throw(c, ENOSPC_sb_downgrade); goto out; } diff --git a/fs/bcachefs/sb-errors.c b/fs/bcachefs/sb-errors.c index 013a96883b4e..48853efdc105 100644 --- a/fs/bcachefs/sb-errors.c +++ b/fs/bcachefs/sb-errors.c @@ -78,6 +78,28 @@ const struct bch_sb_field_ops bch_sb_field_ops_errors = { .to_text = bch2_sb_errors_to_text, }; +void bch2_fs_errors_to_text(struct printbuf *out, struct bch_fs *c) +{ + if (out->nr_tabstops < 1) + printbuf_tabstop_push(out, 48); + if (out->nr_tabstops < 2) + printbuf_tabstop_push(out, 8); + if (out->nr_tabstops < 3) + printbuf_tabstop_push(out, 16); + + guard(mutex)(&c->fsck_error_counts_lock); + + bch_sb_errors_cpu *e = &c->fsck_error_counts; + darray_for_each(*e, i) { + bch2_sb_error_id_to_text(out, i->id); + prt_tab(out); + prt_u64(out, i->nr); + prt_tab(out); + bch2_prt_datetime(out, i->last_error_time); + prt_newline(out); + } +} + void bch2_sb_error_count(struct bch_fs *c, enum bch_sb_error_id err) { bch_sb_errors_cpu *e = &c->fsck_error_counts; diff --git a/fs/bcachefs/sb-errors.h b/fs/bcachefs/sb-errors.h index b2357b8e6107..e86267264692 100644 --- a/fs/bcachefs/sb-errors.h +++ b/fs/bcachefs/sb-errors.h @@ -7,6 +7,7 @@ extern const char * const bch2_sb_error_strs[]; void bch2_sb_error_id_to_text(struct printbuf *, enum bch_sb_error_id); +void bch2_fs_errors_to_text(struct printbuf *, struct bch_fs *); extern const struct bch_sb_field_ops bch_sb_field_ops_errors; diff --git a/fs/bcachefs/sb-errors_format.h b/fs/bcachefs/sb-errors_format.h index 0bfb151da9cf..d154b7651d28 100644 --- a/fs/bcachefs/sb-errors_format.h +++ b/fs/bcachefs/sb-errors_format.h @@ -3,9 +3,10 @@ #define _BCACHEFS_SB_ERRORS_FORMAT_H enum bch_fsck_flags { - FSCK_CAN_FIX = 1 << 0, - FSCK_CAN_IGNORE = 1 << 1, - FSCK_AUTOFIX = 1 << 2, + FSCK_CAN_FIX = BIT(0), + FSCK_CAN_IGNORE = BIT(1), + FSCK_AUTOFIX = BIT(2), + FSCK_ERR_NO_LOG = BIT(3), }; #define BCH_SB_ERRS() \ @@ -134,7 +135,7 @@ enum bch_fsck_flags { x(bucket_gens_to_invalid_buckets, 121, FSCK_AUTOFIX) \ x(bucket_gens_nonzero_for_invalid_buckets, 122, FSCK_AUTOFIX) \ x(need_discard_freespace_key_to_invalid_dev_bucket, 123, 0) \ - x(need_discard_freespace_key_bad, 124, 0) \ + x(need_discard_freespace_key_bad, 124, FSCK_AUTOFIX) \ x(discarding_bucket_not_in_need_discard_btree, 291, 0) \ x(backpointer_bucket_offset_wrong, 125, 0) \ x(backpointer_level_bad, 294, 0) \ @@ -165,7 +166,7 @@ enum bch_fsck_flags { x(ptr_to_missing_replicas_entry, 149, FSCK_AUTOFIX) \ x(ptr_to_missing_stripe, 150, 0) \ x(ptr_to_incorrect_stripe, 151, 0) \ - x(ptr_gen_newer_than_bucket_gen, 152, 0) \ + x(ptr_gen_newer_than_bucket_gen, 152, FSCK_AUTOFIX) \ x(ptr_too_stale, 153, 0) \ x(stale_dirty_ptr, 154, FSCK_AUTOFIX) \ x(ptr_bucket_data_type_mismatch, 155, 0) \ @@ -217,7 +218,7 @@ enum bch_fsck_flags { x(inode_str_hash_invalid, 194, 0) \ x(inode_v3_fields_start_bad, 195, 0) \ x(inode_snapshot_mismatch, 196, 0) \ - x(snapshot_key_missing_inode_snapshot, 314, 0) \ + x(snapshot_key_missing_inode_snapshot, 314, FSCK_AUTOFIX) \ x(inode_unlinked_but_clean, 197, 0) \ x(inode_unlinked_but_nlink_nonzero, 198, 0) \ x(inode_unlinked_and_not_open, 281, 0) \ @@ -232,10 +233,11 @@ enum bch_fsck_flags { x(inode_dir_multiple_links, 206, FSCK_AUTOFIX) \ x(inode_dir_missing_backpointer, 284, FSCK_AUTOFIX) \ x(inode_dir_unlinked_but_not_empty, 286, FSCK_AUTOFIX) \ + x(inode_dir_has_nonzero_i_size, 319, FSCK_AUTOFIX) \ x(inode_multiple_links_but_nlink_0, 207, FSCK_AUTOFIX) \ x(inode_wrong_backpointer, 208, FSCK_AUTOFIX) \ x(inode_wrong_nlink, 209, FSCK_AUTOFIX) \ - x(inode_has_child_snapshots_wrong, 287, 0) \ + x(inode_has_child_snapshots_wrong, 287, FSCK_AUTOFIX) \ x(inode_unreachable, 210, FSCK_AUTOFIX) \ x(inode_journal_seq_in_future, 299, FSCK_AUTOFIX) \ x(inode_i_sectors_underflow, 312, FSCK_AUTOFIX) \ @@ -243,26 +245,27 @@ enum bch_fsck_flags { x(inode_parent_has_case_insensitive_not_set, 317, FSCK_AUTOFIX) \ x(vfs_inode_i_blocks_underflow, 311, FSCK_AUTOFIX) \ x(vfs_inode_i_blocks_not_zero_at_truncate, 313, FSCK_AUTOFIX) \ + x(vfs_bad_inode_rm, 320, 0) \ x(deleted_inode_but_clean, 211, FSCK_AUTOFIX) \ x(deleted_inode_missing, 212, FSCK_AUTOFIX) \ x(deleted_inode_is_dir, 213, FSCK_AUTOFIX) \ x(deleted_inode_not_unlinked, 214, FSCK_AUTOFIX) \ x(deleted_inode_has_child_snapshots, 288, FSCK_AUTOFIX) \ x(extent_overlapping, 215, 0) \ - x(key_in_missing_inode, 216, 0) \ + x(key_in_missing_inode, 216, FSCK_AUTOFIX) \ x(key_in_wrong_inode_type, 217, 0) \ - x(extent_past_end_of_inode, 218, 0) \ + x(extent_past_end_of_inode, 218, FSCK_AUTOFIX) \ x(dirent_empty_name, 219, 0) \ x(dirent_val_too_big, 220, 0) \ x(dirent_name_too_long, 221, 0) \ x(dirent_name_embedded_nul, 222, 0) \ x(dirent_name_dot_or_dotdot, 223, 0) \ x(dirent_name_has_slash, 224, 0) \ - x(dirent_d_type_wrong, 225, 0) \ + x(dirent_d_type_wrong, 225, FSCK_AUTOFIX) \ x(inode_bi_parent_wrong, 226, 0) \ x(dirent_in_missing_dir_inode, 227, 0) \ x(dirent_in_non_dir_inode, 228, 0) \ - x(dirent_to_missing_inode, 229, 0) \ + x(dirent_to_missing_inode, 229, FSCK_AUTOFIX) \ x(dirent_to_overwritten_inode, 302, 0) \ x(dirent_to_missing_subvol, 230, 0) \ x(dirent_to_itself, 231, 0) \ @@ -277,8 +280,8 @@ enum bch_fsck_flags { x(root_dir_missing, 239, 0) \ x(root_inode_not_dir, 240, 0) \ x(dir_loop, 241, 0) \ - x(hash_table_key_duplicate, 242, 0) \ - x(hash_table_key_wrong_offset, 243, 0) \ + x(hash_table_key_duplicate, 242, FSCK_AUTOFIX) \ + x(hash_table_key_wrong_offset, 243, FSCK_AUTOFIX) \ x(unlinked_inode_not_on_deleted_list, 244, FSCK_AUTOFIX) \ x(reflink_p_front_pad_bad, 245, 0) \ x(journal_entry_dup_same_device, 246, 0) \ @@ -298,7 +301,7 @@ enum bch_fsck_flags { x(btree_node_bkey_bad_u64s, 260, 0) \ x(btree_node_topology_empty_interior_node, 261, 0) \ x(btree_ptr_v2_min_key_bad, 262, 0) \ - x(btree_root_unreadable_and_scan_found_nothing, 263, FSCK_AUTOFIX) \ + x(btree_root_unreadable_and_scan_found_nothing, 263, 0) \ x(snapshot_node_missing, 264, FSCK_AUTOFIX) \ x(dup_backpointer_to_bad_csum_extent, 265, 0) \ x(btree_bitmap_not_marked, 266, FSCK_AUTOFIX) \ @@ -311,7 +314,7 @@ enum bch_fsck_flags { x(accounting_mismatch, 272, FSCK_AUTOFIX) \ x(accounting_replicas_not_marked, 273, 0) \ x(accounting_to_invalid_device, 289, 0) \ - x(invalid_btree_id, 274, 0) \ + x(invalid_btree_id, 274, FSCK_AUTOFIX) \ x(alloc_key_io_time_bad, 275, 0) \ x(alloc_key_fragmentation_lru_wrong, 276, FSCK_AUTOFIX) \ x(accounting_key_junk_at_end, 277, FSCK_AUTOFIX) \ @@ -328,7 +331,7 @@ enum bch_fsck_flags { x(dirent_stray_data_after_cf_name, 305, 0) \ x(rebalance_work_incorrectly_set, 309, FSCK_AUTOFIX) \ x(rebalance_work_incorrectly_unset, 310, FSCK_AUTOFIX) \ - x(MAX, 319, 0) + x(MAX, 321, 0) enum bch_sb_error_id { #define x(t, n, ...) BCH_FSCK_ERR_##t = n, diff --git a/fs/bcachefs/sb-members.c b/fs/bcachefs/sb-members.c index 3398906660a5..6245e342a8a8 100644 --- a/fs/bcachefs/sb-members.c +++ b/fs/bcachefs/sb-members.c @@ -101,7 +101,7 @@ static int sb_members_v2_resize_entries(struct bch_fs *c) mi = bch2_sb_field_resize(&c->disk_sb, members_v2, u64s); if (!mi) - return -BCH_ERR_ENOSPC_sb_members_v2; + return bch_err_throw(c, ENOSPC_sb_members_v2); for (int i = c->disk_sb.sb->nr_devices - 1; i >= 0; --i) { void *dst = (void *) mi->_members + (i * sizeof(struct bch_member)); @@ -325,9 +325,17 @@ static void bch2_sb_members_v1_to_text(struct printbuf *out, struct bch_sb *sb, { struct bch_sb_field_members_v1 *mi = field_to_type(f, members_v1); struct bch_sb_field_disk_groups *gi = bch2_sb_field_get(sb, disk_groups); - unsigned i; - for (i = 0; i < sb->nr_devices; i++) + if (vstruct_end(&mi->field) <= (void *) &mi->_members[0]) { + prt_printf(out, "field ends before start of entries"); + return; + } + + unsigned nr = (vstruct_end(&mi->field) - (void *) &mi->_members[0]) / sizeof(mi->_members[0]); + if (nr != sb->nr_devices) + prt_printf(out, "nr_devices mismatch: have %i entries, should be %u", nr, sb->nr_devices); + + for (unsigned i = 0; i < min(sb->nr_devices, nr); i++) member_to_text(out, members_v1_get(mi, i), gi, sb, i); } @@ -341,9 +349,27 @@ static void bch2_sb_members_v2_to_text(struct printbuf *out, struct bch_sb *sb, { struct bch_sb_field_members_v2 *mi = field_to_type(f, members_v2); struct bch_sb_field_disk_groups *gi = bch2_sb_field_get(sb, disk_groups); - unsigned i; - for (i = 0; i < sb->nr_devices; i++) + if (vstruct_end(&mi->field) <= (void *) &mi->_members[0]) { + prt_printf(out, "field ends before start of entries"); + return; + } + + if (!le16_to_cpu(mi->member_bytes)) { + prt_printf(out, "member_bytes 0"); + return; + } + + unsigned nr = (vstruct_end(&mi->field) - (void *) &mi->_members[0]) / le16_to_cpu(mi->member_bytes); + if (nr != sb->nr_devices) + prt_printf(out, "nr_devices mismatch: have %i entries, should be %u", nr, sb->nr_devices); + + /* + * We call to_text() on superblock sections that haven't passed + * validate, so we can't trust sb->nr_devices. + */ + + for (unsigned i = 0; i < min(sb->nr_devices, nr); i++) member_to_text(out, members_v2_get(mi, i), gi, sb, i); } @@ -378,14 +404,13 @@ void bch2_sb_members_from_cpu(struct bch_fs *c) { struct bch_sb_field_members_v2 *mi = bch2_sb_field_get(c->disk_sb.sb, members_v2); - rcu_read_lock(); + guard(rcu)(); for_each_member_device_rcu(c, ca, NULL) { struct bch_member *m = __bch2_members_v2_get_mut(mi, ca->dev_idx); for (unsigned e = 0; e < BCH_MEMBER_ERROR_NR; e++) m->errors[e] = cpu_to_le64(atomic64_read(&ca->errors[e])); } - rcu_read_unlock(); } void bch2_dev_io_errors_to_text(struct printbuf *out, struct bch_dev *ca) @@ -443,20 +468,14 @@ void bch2_dev_errors_reset(struct bch_dev *ca) bool bch2_dev_btree_bitmap_marked(struct bch_fs *c, struct bkey_s_c k) { - bool ret = true; - rcu_read_lock(); + guard(rcu)(); bkey_for_each_ptr(bch2_bkey_ptrs_c(k), ptr) { struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev); - if (!ca) - continue; - - if (!bch2_dev_btree_bitmap_marked_sectors(ca, ptr->offset, btree_sectors(c))) { - ret = false; - break; - } + if (ca && + !bch2_dev_btree_bitmap_marked_sectors(ca, ptr->offset, btree_sectors(c))) + return false; } - rcu_read_unlock(); - return ret; + return true; } static void __bch2_dev_btree_bitmap_mark(struct bch_sb_field_members_v2 *mi, unsigned dev, diff --git a/fs/bcachefs/sb-members.h b/fs/bcachefs/sb-members.h index 6bd9b86aee5b..8d8a8a857648 100644 --- a/fs/bcachefs/sb-members.h +++ b/fs/bcachefs/sb-members.h @@ -28,12 +28,9 @@ static inline struct bch_dev *bch2_dev_rcu(struct bch_fs *, unsigned); static inline bool bch2_dev_idx_is_online(struct bch_fs *c, unsigned dev) { - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu(c, dev); - bool ret = ca && bch2_dev_is_online(ca); - rcu_read_unlock(); - - return ret; + return ca && bch2_dev_is_online(ca); } static inline bool bch2_dev_is_healthy(struct bch_dev *ca) @@ -142,12 +139,10 @@ static inline void bch2_dev_put(struct bch_dev *ca) static inline struct bch_dev *bch2_get_next_dev(struct bch_fs *c, struct bch_dev *ca) { - rcu_read_lock(); + guard(rcu)(); bch2_dev_put(ca); if ((ca = __bch2_next_dev(c, ca, NULL))) bch2_dev_get(ca); - rcu_read_unlock(); - return ca; } @@ -166,7 +161,7 @@ static inline struct bch_dev *bch2_get_next_online_dev(struct bch_fs *c, unsigned state_mask, int rw, unsigned ref_idx) { - rcu_read_lock(); + guard(rcu)(); if (ca) enumerated_ref_put(&ca->io_ref[rw], ref_idx); @@ -174,7 +169,6 @@ static inline struct bch_dev *bch2_get_next_online_dev(struct bch_fs *c, (!((1 << ca->mi.state) & state_mask) || !enumerated_ref_tryget(&ca->io_ref[rw], ref_idx))) ; - rcu_read_unlock(); return ca; } @@ -239,11 +233,10 @@ static inline struct bch_dev *bch2_dev_rcu(struct bch_fs *c, unsigned dev) static inline struct bch_dev *bch2_dev_tryget_noerror(struct bch_fs *c, unsigned dev) { - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu_noerror(c, dev); if (ca) bch2_dev_get(ca); - rcu_read_unlock(); return ca; } @@ -299,19 +292,16 @@ static inline struct bch_dev *bch2_dev_get_ioref(struct bch_fs *c, unsigned dev, { might_sleep(); - rcu_read_lock(); + guard(rcu)(); struct bch_dev *ca = bch2_dev_rcu(c, dev); - if (ca && !enumerated_ref_tryget(&ca->io_ref[rw], ref_idx)) - ca = NULL; - rcu_read_unlock(); + if (!ca || !enumerated_ref_tryget(&ca->io_ref[rw], ref_idx)) + return NULL; - if (ca && - (ca->mi.state == BCH_MEMBER_STATE_rw || - (ca->mi.state == BCH_MEMBER_STATE_ro && rw == READ))) + if (ca->mi.state == BCH_MEMBER_STATE_rw || + (ca->mi.state == BCH_MEMBER_STATE_ro && rw == READ)) return ca; - if (ca) - enumerated_ref_put(&ca->io_ref[rw], ref_idx); + enumerated_ref_put(&ca->io_ref[rw], ref_idx); return NULL; } diff --git a/fs/bcachefs/six.c b/fs/bcachefs/six.c index 7c403427fbdb..538c324f4765 100644 --- a/fs/bcachefs/six.c +++ b/fs/bcachefs/six.c @@ -339,12 +339,9 @@ static inline bool six_owner_running(struct six_lock *lock) * acquiring the lock and setting the owner field. If we're an RT task * that will live-lock because we won't let the owner complete. */ - rcu_read_lock(); + guard(rcu)(); struct task_struct *owner = READ_ONCE(lock->owner); - bool ret = owner ? owner_on_cpu(owner) : !rt_or_dl_task(current); - rcu_read_unlock(); - - return ret; + return owner ? owner_on_cpu(owner) : !rt_or_dl_task(current); } static inline bool six_optimistic_spin(struct six_lock *lock, diff --git a/fs/bcachefs/snapshot.c b/fs/bcachefs/snapshot.c index 00d62d1190ef..4c43d2a2c1f5 100644 --- a/fs/bcachefs/snapshot.c +++ b/fs/bcachefs/snapshot.c @@ -54,7 +54,7 @@ int bch2_snapshot_tree_lookup(struct btree_trans *trans, u32 id, BTREE_ITER_with_updates, snapshot_tree, s); if (bch2_err_matches(ret, ENOENT)) - ret = -BCH_ERR_ENOENT_snapshot_tree; + ret = bch_err_throw(trans->c, ENOENT_snapshot_tree); return ret; } @@ -67,7 +67,7 @@ __bch2_snapshot_tree_create(struct btree_trans *trans) struct bkey_i_snapshot_tree *s_t; if (ret == -BCH_ERR_ENOSPC_btree_slot) - ret = -BCH_ERR_ENOSPC_snapshot_tree; + ret = bch_err_throw(trans->c, ENOSPC_snapshot_tree); if (ret) return ERR_PTR(ret); @@ -105,11 +105,8 @@ static bool __bch2_snapshot_is_ancestor_early(struct snapshot_table *t, u32 id, static bool bch2_snapshot_is_ancestor_early(struct bch_fs *c, u32 id, u32 ancestor) { - rcu_read_lock(); - bool ret = __bch2_snapshot_is_ancestor_early(rcu_dereference(c->snapshots), id, ancestor); - rcu_read_unlock(); - - return ret; + guard(rcu)(); + return __bch2_snapshot_is_ancestor_early(rcu_dereference(c->snapshots), id, ancestor); } static inline u32 get_ancestor_below(struct snapshot_table *t, u32 id, u32 ancestor) @@ -138,28 +135,25 @@ static bool test_ancestor_bitmap(struct snapshot_table *t, u32 id, u32 ancestor) bool __bch2_snapshot_is_ancestor(struct bch_fs *c, u32 id, u32 ancestor) { - bool ret; +#ifdef CONFIG_BCACHEFS_DEBUG + u32 orig_id = id; +#endif - rcu_read_lock(); + guard(rcu)(); struct snapshot_table *t = rcu_dereference(c->snapshots); - if (unlikely(c->recovery.pass_done < BCH_RECOVERY_PASS_check_snapshots)) { - ret = __bch2_snapshot_is_ancestor_early(t, id, ancestor); - goto out; - } + if (unlikely(c->recovery.pass_done < BCH_RECOVERY_PASS_check_snapshots)) + return __bch2_snapshot_is_ancestor_early(t, id, ancestor); if (likely(ancestor >= IS_ANCESTOR_BITMAP)) while (id && id < ancestor - IS_ANCESTOR_BITMAP) id = get_ancestor_below(t, id, ancestor); - ret = id && id < ancestor + bool ret = id && id < ancestor ? test_ancestor_bitmap(t, id, ancestor) : id == ancestor; - EBUG_ON(ret != __bch2_snapshot_is_ancestor_early(t, id, ancestor)); -out: - rcu_read_unlock(); - + EBUG_ON(ret != __bch2_snapshot_is_ancestor_early(t, orig_id, ancestor)); return ret; } @@ -293,7 +287,7 @@ static int bch2_snapshot_table_make_room(struct bch_fs *c, u32 id) mutex_lock(&c->snapshot_table_lock); int ret = snapshot_t_mut(c, id) ? 0 - : -BCH_ERR_ENOMEM_mark_snapshot; + : bch_err_throw(c, ENOMEM_mark_snapshot); mutex_unlock(&c->snapshot_table_lock); return ret; } @@ -312,7 +306,7 @@ static int __bch2_mark_snapshot(struct btree_trans *trans, t = snapshot_t_mut(c, id); if (!t) { - ret = -BCH_ERR_ENOMEM_mark_snapshot; + ret = bch_err_throw(c, ENOMEM_mark_snapshot); goto err; } @@ -412,10 +406,10 @@ static u32 bch2_snapshot_tree_next(struct bch_fs *c, u32 id) u32 bch2_snapshot_oldest_subvol(struct bch_fs *c, u32 snapshot_root, snapshot_id_list *skip) { + guard(rcu)(); u32 id, subvol = 0, s; retry: id = snapshot_root; - rcu_read_lock(); while (id && bch2_snapshot_exists(c, id)) { if (!(skip && snapshot_list_has_id(skip, id))) { s = snapshot_t(c, id)->subvol; @@ -427,7 +421,6 @@ retry: if (id == snapshot_root) break; } - rcu_read_unlock(); if (!subvol && skip) { skip = NULL; @@ -617,18 +610,14 @@ static int snapshot_tree_ptr_good(struct btree_trans *trans, u32 bch2_snapshot_skiplist_get(struct bch_fs *c, u32 id) { - const struct snapshot_t *s; - if (!id) return 0; - rcu_read_lock(); - s = snapshot_t(c, id); - if (s->parent) - id = bch2_snapshot_nth_parent(c, id, get_random_u32_below(s->depth)); - rcu_read_unlock(); - - return id; + guard(rcu)(); + const struct snapshot_t *s = snapshot_t(c, id); + return s->parent + ? bch2_snapshot_nth_parent(c, id, get_random_u32_below(s->depth)) + : id; } static int snapshot_skiplist_good(struct btree_trans *trans, u32 id, struct bch_snapshot s) @@ -882,7 +871,8 @@ static int check_snapshot_exists(struct btree_trans *trans, u32 id) for_each_btree_key_norestart(trans, iter, BTREE_ID_snapshot_trees, POS_MIN, 0, k, ret) { - if (le32_to_cpu(bkey_s_c_to_snapshot_tree(k).v->root_snapshot) == id) { + if (k.k->type == KEY_TYPE_snapshot_tree && + le32_to_cpu(bkey_s_c_to_snapshot_tree(k).v->root_snapshot) == id) { tree_id = k.k->p.offset; break; } @@ -910,7 +900,8 @@ static int check_snapshot_exists(struct btree_trans *trans, u32 id) for_each_btree_key_norestart(trans, iter, BTREE_ID_subvolumes, POS_MIN, 0, k, ret) { - if (le32_to_cpu(bkey_s_c_to_subvolume(k).v->snapshot) == id) { + if (k.k->type == KEY_TYPE_subvolume && + le32_to_cpu(bkey_s_c_to_subvolume(k).v->snapshot) == id) { snapshot->v.subvol = cpu_to_le32(k.k->p.offset); SET_BCH_SNAPSHOT_SUBVOL(&snapshot->v, true); break; @@ -947,10 +938,7 @@ static inline bool same_snapshot(struct snapshot_tree_reconstruct *r, struct bpo static inline bool snapshot_id_lists_have_common(snapshot_id_list *l, snapshot_id_list *r) { - darray_for_each(*l, i) - if (snapshot_list_has_id(r, *i)) - return true; - return false; + return darray_find_p(*l, i, snapshot_list_has_id(r, *i)) != NULL; } static void snapshot_id_list_to_text(struct printbuf *out, snapshot_id_list *s) @@ -1022,7 +1010,7 @@ int bch2_reconstruct_snapshots(struct bch_fs *c) "snapshot node %u from tree %s missing, recreate?", *id, buf.buf)) { if (t->nr > 1) { bch_err(c, "cannot reconstruct snapshot trees with multiple nodes"); - ret = -BCH_ERR_fsck_repair_unimplemented; + ret = bch_err_throw(c, fsck_repair_unimplemented); goto err; } @@ -1061,24 +1049,73 @@ int __bch2_check_key_has_snapshot(struct btree_trans *trans, ret = bch2_btree_delete_at(trans, iter, BTREE_UPDATE_internal_snapshot_node) ?: 1; - /* - * Snapshot missing: we should have caught this with btree_lost_data and - * kicked off reconstruct_snapshots, so if we end up here we have no - * idea what happened: - */ - if (fsck_err_on(state == SNAPSHOT_ID_empty, - trans, bkey_in_missing_snapshot, - "key in missing snapshot %s, delete?", - (bch2_btree_id_to_text(&buf, iter->btree_id), - prt_char(&buf, ' '), - bch2_bkey_val_to_text(&buf, c, k), buf.buf))) - ret = bch2_btree_delete_at(trans, iter, - BTREE_UPDATE_internal_snapshot_node) ?: 1; + if (state == SNAPSHOT_ID_empty) { + /* + * Snapshot missing: we should have caught this with btree_lost_data and + * kicked off reconstruct_snapshots, so if we end up here we have no + * idea what happened. + * + * Do not delete unless we know that subvolumes and snapshots + * are consistent: + * + * XXX: + * + * We could be smarter here, and instead of using the generic + * recovery pass ratelimiting, track if there have been any + * changes to the snapshots or inodes btrees since those passes + * last ran. + */ + ret = bch2_require_recovery_pass(c, &buf, BCH_RECOVERY_PASS_check_snapshots) ?: ret; + ret = bch2_require_recovery_pass(c, &buf, BCH_RECOVERY_PASS_check_subvols) ?: ret; + + if (c->sb.btrees_lost_data & BIT_ULL(BTREE_ID_snapshots)) + ret = bch2_require_recovery_pass(c, &buf, BCH_RECOVERY_PASS_reconstruct_snapshots) ?: ret; + + unsigned repair_flags = FSCK_CAN_IGNORE | (!ret ? FSCK_CAN_FIX : 0); + + if (__fsck_err(trans, repair_flags, bkey_in_missing_snapshot, + "key in missing snapshot %s, delete?", + (bch2_btree_id_to_text(&buf, iter->btree_id), + prt_char(&buf, ' '), + bch2_bkey_val_to_text(&buf, c, k), buf.buf))) { + ret = bch2_btree_delete_at(trans, iter, + BTREE_UPDATE_internal_snapshot_node) ?: 1; + } + } fsck_err: printbuf_exit(&buf); return ret; } +int __bch2_get_snapshot_overwrites(struct btree_trans *trans, + enum btree_id btree, struct bpos pos, + snapshot_id_list *s) +{ + struct bch_fs *c = trans->c; + struct btree_iter iter; + struct bkey_s_c k; + int ret = 0; + + for_each_btree_key_reverse_norestart(trans, iter, btree, bpos_predecessor(pos), + BTREE_ITER_all_snapshots, k, ret) { + if (!bkey_eq(k.k->p, pos)) + break; + + if (!bch2_snapshot_is_ancestor(c, k.k->p.snapshot, pos.snapshot) || + snapshot_list_has_ancestor(c, s, k.k->p.snapshot)) + continue; + + ret = snapshot_list_add(c, s, k.k->p.snapshot); + if (ret) + break; + } + bch2_trans_iter_exit(trans, &iter); + if (ret) + darray_exit(s); + + return ret; +} + /* * Mark a snapshot as deleted, for future cleanup: */ @@ -1263,7 +1300,7 @@ static int create_snapids(struct btree_trans *trans, u32 parent, u32 tree, goto err; if (!k.k || !k.k->p.offset) { - ret = -BCH_ERR_ENOSPC_snapshot_create; + ret = bch_err_throw(c, ENOSPC_snapshot_create); goto err; } @@ -1399,10 +1436,8 @@ int bch2_snapshot_node_create(struct btree_trans *trans, u32 parent, static inline u32 interior_delete_has_id(interior_delete_list *l, u32 id) { - darray_for_each(*l, i) - if (i->id == id) - return i->live_child; - return 0; + struct snapshot_interior_delete *i = darray_find_p(*l, i, i->id == id); + return i ? i->live_child : 0; } static unsigned __live_child(struct snapshot_table *t, u32 id, @@ -1434,11 +1469,9 @@ static unsigned live_child(struct bch_fs *c, u32 id) { struct snapshot_delete *d = &c->snapshot_delete; - rcu_read_lock(); - u32 ret = __live_child(rcu_dereference(c->snapshots), id, - &d->delete_leaves, &d->delete_interior); - rcu_read_unlock(); - return ret; + guard(rcu)(); + return __live_child(rcu_dereference(c->snapshots), id, + &d->delete_leaves, &d->delete_interior); } static bool snapshot_id_dying(struct snapshot_delete *d, unsigned id) @@ -1695,7 +1728,7 @@ static int check_should_delete_snapshot(struct btree_trans *trans, struct bkey_s static inline u32 bch2_snapshot_nth_parent_skip(struct bch_fs *c, u32 id, u32 n, interior_delete_list *skip) { - rcu_read_lock(); + guard(rcu)(); while (interior_delete_has_id(skip, id)) id = __bch2_snapshot_parent(c, id); @@ -1704,7 +1737,6 @@ static inline u32 bch2_snapshot_nth_parent_skip(struct bch_fs *c, u32 id, u32 n, id = __bch2_snapshot_parent(c, id); } while (interior_delete_has_id(skip, id)); } - rcu_read_unlock(); return id; } @@ -1870,6 +1902,8 @@ err: d->running = false; mutex_unlock(&d->progress_lock); bch2_trans_put(trans); + + bch2_recovery_pass_set_no_ratelimit(c, BCH_RECOVERY_PASS_check_snapshots); out_unlock: mutex_unlock(&d->lock); if (!bch2_err_matches(ret, EROFS)) @@ -1905,7 +1939,7 @@ void bch2_delete_dead_snapshots_async(struct bch_fs *c) BUG_ON(!test_bit(BCH_FS_may_go_rw, &c->flags)); - if (!queue_work(c->write_ref_wq, &c->snapshot_delete.work)) + if (!queue_work(system_long_wq, &c->snapshot_delete.work)) enumerated_ref_put(&c->writes, BCH_WRITE_REF_delete_dead_snapshots); } diff --git a/fs/bcachefs/snapshot.h b/fs/bcachefs/snapshot.h index 382a171f5413..6766bf673ed9 100644 --- a/fs/bcachefs/snapshot.h +++ b/fs/bcachefs/snapshot.h @@ -46,12 +46,9 @@ static inline const struct snapshot_t *snapshot_t(struct bch_fs *c, u32 id) static inline u32 bch2_snapshot_tree(struct bch_fs *c, u32 id) { - rcu_read_lock(); + guard(rcu)(); const struct snapshot_t *s = snapshot_t(c, id); - id = s ? s->tree : 0; - rcu_read_unlock(); - - return id; + return s ? s->tree : 0; } static inline u32 __bch2_snapshot_parent_early(struct bch_fs *c, u32 id) @@ -62,11 +59,8 @@ static inline u32 __bch2_snapshot_parent_early(struct bch_fs *c, u32 id) static inline u32 bch2_snapshot_parent_early(struct bch_fs *c, u32 id) { - rcu_read_lock(); - id = __bch2_snapshot_parent_early(c, id); - rcu_read_unlock(); - - return id; + guard(rcu)(); + return __bch2_snapshot_parent_early(c, id); } static inline u32 __bch2_snapshot_parent(struct bch_fs *c, u32 id) @@ -88,20 +82,15 @@ static inline u32 __bch2_snapshot_parent(struct bch_fs *c, u32 id) static inline u32 bch2_snapshot_parent(struct bch_fs *c, u32 id) { - rcu_read_lock(); - id = __bch2_snapshot_parent(c, id); - rcu_read_unlock(); - - return id; + guard(rcu)(); + return __bch2_snapshot_parent(c, id); } static inline u32 bch2_snapshot_nth_parent(struct bch_fs *c, u32 id, u32 n) { - rcu_read_lock(); + guard(rcu)(); while (n--) id = __bch2_snapshot_parent(c, id); - rcu_read_unlock(); - return id; } @@ -110,13 +99,11 @@ u32 bch2_snapshot_skiplist_get(struct bch_fs *, u32); static inline u32 bch2_snapshot_root(struct bch_fs *c, u32 id) { - u32 parent; + guard(rcu)(); - rcu_read_lock(); + u32 parent; while ((parent = __bch2_snapshot_parent(c, id))) id = parent; - rcu_read_unlock(); - return id; } @@ -128,11 +115,8 @@ static inline enum snapshot_id_state __bch2_snapshot_id_state(struct bch_fs *c, static inline enum snapshot_id_state bch2_snapshot_id_state(struct bch_fs *c, u32 id) { - rcu_read_lock(); - enum snapshot_id_state ret = __bch2_snapshot_id_state(c, id); - rcu_read_unlock(); - - return ret; + guard(rcu)(); + return __bch2_snapshot_id_state(c, id); } static inline bool bch2_snapshot_exists(struct bch_fs *c, u32 id) @@ -142,12 +126,9 @@ static inline bool bch2_snapshot_exists(struct bch_fs *c, u32 id) static inline int bch2_snapshot_is_internal_node(struct bch_fs *c, u32 id) { - rcu_read_lock(); + guard(rcu)(); const struct snapshot_t *s = snapshot_t(c, id); - int ret = s ? s->children[0] : -BCH_ERR_invalid_snapshot_node; - rcu_read_unlock(); - - return ret; + return s ? s->children[0] : -BCH_ERR_invalid_snapshot_node; } static inline int bch2_snapshot_is_leaf(struct bch_fs *c, u32 id) @@ -160,13 +141,8 @@ static inline int bch2_snapshot_is_leaf(struct bch_fs *c, u32 id) static inline u32 bch2_snapshot_depth(struct bch_fs *c, u32 parent) { - u32 depth; - - rcu_read_lock(); - depth = parent ? snapshot_t(c, parent)->depth + 1 : 0; - rcu_read_unlock(); - - return depth; + guard(rcu)(); + return parent ? snapshot_t(c, parent)->depth + 1 : 0; } bool __bch2_snapshot_is_ancestor(struct bch_fs *, u32, u32); @@ -180,20 +156,14 @@ static inline bool bch2_snapshot_is_ancestor(struct bch_fs *c, u32 id, u32 ances static inline bool bch2_snapshot_has_children(struct bch_fs *c, u32 id) { - rcu_read_lock(); + guard(rcu)(); const struct snapshot_t *t = snapshot_t(c, id); - bool ret = t && (t->children[0]|t->children[1]) != 0; - rcu_read_unlock(); - - return ret; + return t && (t->children[0]|t->children[1]) != 0; } static inline bool snapshot_list_has_id(snapshot_id_list *s, u32 id) { - darray_for_each(*s, i) - if (*i == id) - return true; - return false; + return darray_find(*s, id) != NULL; } static inline bool snapshot_list_has_ancestor(struct bch_fs *c, snapshot_id_list *s, u32 id) @@ -258,6 +228,25 @@ static inline int bch2_check_key_has_snapshot(struct btree_trans *trans, : __bch2_check_key_has_snapshot(trans, iter, k); } +int __bch2_get_snapshot_overwrites(struct btree_trans *, + enum btree_id, struct bpos, + snapshot_id_list *); + +/* + * Get a list of snapshot IDs that have overwritten a given key: + */ +static inline int bch2_get_snapshot_overwrites(struct btree_trans *trans, + enum btree_id btree, struct bpos pos, + snapshot_id_list *s) +{ + darray_init(s); + + return bch2_snapshot_has_children(trans->c, pos.snapshot) + ? __bch2_get_snapshot_overwrites(trans, btree, pos, s) + : 0; + +} + int bch2_snapshot_node_set_deleted(struct btree_trans *, u32); int __bch2_key_has_snapshot_overwrites(struct btree_trans *, enum btree_id, struct bpos); diff --git a/fs/bcachefs/str_hash.c b/fs/bcachefs/str_hash.c index 0cbf5508a32c..3e9f59226bdf 100644 --- a/fs/bcachefs/str_hash.c +++ b/fs/bcachefs/str_hash.c @@ -31,14 +31,16 @@ static int bch2_dirent_has_target(struct btree_trans *trans, struct bkey_s_c_dir } } -static noinline int fsck_rename_dirent(struct btree_trans *trans, - struct snapshots_seen *s, - const struct bch_hash_desc desc, - struct bch_hash_info *hash_info, - struct bkey_s_c_dirent old) +static int bch2_fsck_rename_dirent(struct btree_trans *trans, + struct snapshots_seen *s, + const struct bch_hash_desc desc, + struct bch_hash_info *hash_info, + struct bkey_s_c_dirent old, + bool *updated_before_k_pos) { + struct bch_fs *c = trans->c; struct qstr old_name = bch2_dirent_get_name(old); - struct bkey_i_dirent *new = bch2_trans_kmalloc(trans, bkey_bytes(old.k) + 32); + struct bkey_i_dirent *new = bch2_trans_kmalloc(trans, BKEY_U64s_MAX * sizeof(u64)); int ret = PTR_ERR_OR_ZERO(new); if (ret) return ret; @@ -47,28 +49,39 @@ static noinline int fsck_rename_dirent(struct btree_trans *trans, dirent_copy_target(new, old); new->k.p = old.k->p; + char *renamed_buf = bch2_trans_kmalloc(trans, old_name.len + 20); + ret = PTR_ERR_OR_ZERO(renamed_buf); + if (ret) + return ret; + for (unsigned i = 0; i < 1000; i++) { - unsigned len = sprintf(new->v.d_name, "%.*s.fsck_renamed-%u", - old_name.len, old_name.name, i); - unsigned u64s = BKEY_U64s + dirent_val_u64s(len, 0); + new->k.u64s = BKEY_U64s_MAX; - if (u64s > U8_MAX) - return -EINVAL; + struct qstr renamed_name = (struct qstr) QSTR_INIT(renamed_buf, + sprintf(renamed_buf, "%.*s.fsck_renamed-%u", + old_name.len, old_name.name, i)); - new->k.u64s = u64s; + ret = bch2_dirent_init_name(c, new, hash_info, &renamed_name, NULL); + if (ret) + return ret; ret = bch2_hash_set_in_snapshot(trans, bch2_dirent_hash_desc, hash_info, (subvol_inum) { 0, old.k->p.inode }, old.k->p.snapshot, &new->k_i, - BTREE_UPDATE_internal_snapshot_node); - if (!bch2_err_matches(ret, EEXIST)) + BTREE_UPDATE_internal_snapshot_node| + STR_HASH_must_create); + if (ret && !bch2_err_matches(ret, EEXIST)) break; + if (!ret) { + if (bpos_lt(new->k.p, old.k->p)) + *updated_before_k_pos = true; + break; + } } - if (ret) - return ret; - - return bch2_fsck_update_backpointers(trans, s, desc, hash_info, &new->k_i); + ret = ret ?: bch2_fsck_update_backpointers(trans, s, desc, hash_info, &new->k_i); + bch_err_fn(c, ret); + return ret; } static noinline int hash_pick_winner(struct btree_trans *trans, @@ -186,7 +199,7 @@ int bch2_repair_inode_hash_info(struct btree_trans *trans, #endif bch2_print_str(c, KERN_ERR, buf.buf); printbuf_exit(&buf); - ret = -BCH_ERR_fsck_repair_unimplemented; + ret = bch_err_throw(c, fsck_repair_unimplemented); goto err; } @@ -221,11 +234,115 @@ static noinline int check_inode_hash_info_matches_root(struct btree_trans *trans return ret; } +/* Put a str_hash key in its proper location, checking for duplicates */ +int bch2_str_hash_repair_key(struct btree_trans *trans, + struct snapshots_seen *s, + const struct bch_hash_desc *desc, + struct bch_hash_info *hash_info, + struct btree_iter *k_iter, struct bkey_s_c k, + struct btree_iter *dup_iter, struct bkey_s_c dup_k, + bool *updated_before_k_pos) +{ + struct bch_fs *c = trans->c; + struct printbuf buf = PRINTBUF; + bool free_snapshots_seen = false; + int ret = 0; + + if (!s) { + s = bch2_trans_kmalloc(trans, sizeof(*s)); + ret = PTR_ERR_OR_ZERO(s); + if (ret) + goto out; + + s->pos = k_iter->pos; + darray_init(&s->ids); + + ret = bch2_get_snapshot_overwrites(trans, desc->btree_id, k_iter->pos, &s->ids); + if (ret) + goto out; + + free_snapshots_seen = true; + } + + if (!dup_k.k) { + struct bkey_i *new = bch2_bkey_make_mut_noupdate(trans, k); + ret = PTR_ERR_OR_ZERO(new); + if (ret) + goto out; + + dup_k = bch2_hash_set_or_get_in_snapshot(trans, dup_iter, *desc, hash_info, + (subvol_inum) { 0, new->k.p.inode }, + new->k.p.snapshot, new, + STR_HASH_must_create| + BTREE_ITER_with_updates| + BTREE_UPDATE_internal_snapshot_node); + ret = bkey_err(dup_k); + if (ret) + goto out; + if (dup_k.k) + goto duplicate_entries; + + if (bpos_lt(new->k.p, k.k->p)) + *updated_before_k_pos = true; + + ret = bch2_insert_snapshot_whiteouts(trans, desc->btree_id, + k_iter->pos, new->k.p) ?: + bch2_hash_delete_at(trans, *desc, hash_info, k_iter, + BTREE_ITER_with_updates| + BTREE_UPDATE_internal_snapshot_node) ?: + bch2_fsck_update_backpointers(trans, s, *desc, hash_info, new) ?: + bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc) ?: + -BCH_ERR_transaction_restart_commit; + } else { +duplicate_entries: + ret = hash_pick_winner(trans, *desc, hash_info, k, dup_k); + if (ret < 0) + goto out; + + if (!fsck_err(trans, hash_table_key_duplicate, + "duplicate hash table keys%s:\n%s", + ret != 2 ? "" : ", both point to valid inodes", + (printbuf_reset(&buf), + bch2_bkey_val_to_text(&buf, c, k), + prt_newline(&buf), + bch2_bkey_val_to_text(&buf, c, dup_k), + buf.buf))) + goto out; + + switch (ret) { + case 0: + ret = bch2_hash_delete_at(trans, *desc, hash_info, k_iter, 0); + break; + case 1: + ret = bch2_hash_delete_at(trans, *desc, hash_info, dup_iter, 0); + break; + case 2: + ret = bch2_fsck_rename_dirent(trans, s, *desc, hash_info, + bkey_s_c_to_dirent(k), + updated_before_k_pos) ?: + bch2_hash_delete_at(trans, *desc, hash_info, k_iter, + BTREE_ITER_with_updates); + goto out; + } + + ret = bch2_trans_commit(trans, NULL, NULL, 0) ?: + -BCH_ERR_transaction_restart_commit; + } +out: +fsck_err: + bch2_trans_iter_exit(trans, dup_iter); + printbuf_exit(&buf); + if (free_snapshots_seen) + darray_exit(&s->ids); + return ret; +} + int __bch2_str_hash_check_key(struct btree_trans *trans, struct snapshots_seen *s, const struct bch_hash_desc *desc, struct bch_hash_info *hash_info, - struct btree_iter *k_iter, struct bkey_s_c hash_k) + struct btree_iter *k_iter, struct bkey_s_c hash_k, + bool *updated_before_k_pos) { struct bch_fs *c = trans->c; struct btree_iter iter = {}; @@ -239,24 +356,31 @@ int __bch2_str_hash_check_key(struct btree_trans *trans, for_each_btree_key_norestart(trans, iter, desc->btree_id, SPOS(hash_k.k->p.inode, hash, hash_k.k->p.snapshot), - BTREE_ITER_slots, k, ret) { + BTREE_ITER_slots| + BTREE_ITER_with_updates, k, ret) { if (bkey_eq(k.k->p, hash_k.k->p)) break; if (k.k->type == desc->key_type && - !desc->cmp_bkey(k, hash_k)) - goto duplicate_entries; + !desc->cmp_bkey(k, hash_k)) { + ret = check_inode_hash_info_matches_root(trans, hash_k.k->p.inode, + hash_info) ?: + bch2_str_hash_repair_key(trans, s, desc, hash_info, + k_iter, hash_k, + &iter, k, updated_before_k_pos); + break; + } - if (bkey_deleted(k.k)) { - bch2_trans_iter_exit(trans, &iter); + if (bkey_deleted(k.k)) goto bad_hash; - } } -out: bch2_trans_iter_exit(trans, &iter); +out: +fsck_err: printbuf_exit(&buf); return ret; bad_hash: + bch2_trans_iter_exit(trans, &iter); /* * Before doing any repair, check hash_info itself: */ @@ -265,64 +389,12 @@ bad_hash: goto out; if (fsck_err(trans, hash_table_key_wrong_offset, - "hash table key at wrong offset: btree %s inode %llu offset %llu, hashed to %llu\n%s", - bch2_btree_id_str(desc->btree_id), hash_k.k->p.inode, hash_k.k->p.offset, hash, - (printbuf_reset(&buf), - bch2_bkey_val_to_text(&buf, c, hash_k), buf.buf))) { - struct bkey_i *new = bch2_bkey_make_mut_noupdate(trans, hash_k); - if (IS_ERR(new)) - return PTR_ERR(new); - - k = bch2_hash_set_or_get_in_snapshot(trans, &iter, *desc, hash_info, - (subvol_inum) { 0, hash_k.k->p.inode }, - hash_k.k->p.snapshot, new, - STR_HASH_must_create| - BTREE_ITER_with_updates| - BTREE_UPDATE_internal_snapshot_node); - ret = bkey_err(k); - if (ret) - goto out; - if (k.k) - goto duplicate_entries; - - ret = bch2_hash_delete_at(trans, *desc, hash_info, k_iter, - BTREE_UPDATE_internal_snapshot_node) ?: - bch2_fsck_update_backpointers(trans, s, *desc, hash_info, new) ?: - bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc) ?: - -BCH_ERR_transaction_restart_nested; - goto out; - } -fsck_err: - goto out; -duplicate_entries: - ret = hash_pick_winner(trans, *desc, hash_info, hash_k, k); - if (ret < 0) - goto out; - - if (!fsck_err(trans, hash_table_key_duplicate, - "duplicate hash table keys%s:\n%s", - ret != 2 ? "" : ", both point to valid inodes", - (printbuf_reset(&buf), - bch2_bkey_val_to_text(&buf, c, hash_k), - prt_newline(&buf), - bch2_bkey_val_to_text(&buf, c, k), - buf.buf))) - goto out; - - switch (ret) { - case 0: - ret = bch2_hash_delete_at(trans, *desc, hash_info, k_iter, 0); - break; - case 1: - ret = bch2_hash_delete_at(trans, *desc, hash_info, &iter, 0); - break; - case 2: - ret = fsck_rename_dirent(trans, s, *desc, hash_info, bkey_s_c_to_dirent(hash_k)) ?: - bch2_hash_delete_at(trans, *desc, hash_info, k_iter, 0); - goto out; - } - - ret = bch2_trans_commit(trans, NULL, NULL, 0) ?: - -BCH_ERR_transaction_restart_nested; + "hash table key at wrong offset: should be at %llu\n%s", + hash, + (bch2_bkey_val_to_text(&buf, c, hash_k), buf.buf))) + ret = bch2_str_hash_repair_key(trans, s, desc, hash_info, + k_iter, hash_k, + &iter, bkey_s_c_null, + updated_before_k_pos); goto out; } diff --git a/fs/bcachefs/str_hash.h b/fs/bcachefs/str_hash.h index 6762b3627e1b..8979ac2d7a3b 100644 --- a/fs/bcachefs/str_hash.h +++ b/fs/bcachefs/str_hash.h @@ -48,9 +48,7 @@ bch2_hash_info_init(struct bch_fs *c, const struct bch_inode_unpacked *bi) struct bch_hash_info info = { .inum_snapshot = bi->bi_snapshot, .type = INODE_STR_HASH(bi), -#ifdef CONFIG_UNICODE .cf_encoding = bch2_inode_casefold(c, bi) ? c->cf_encoding : NULL, -#endif .siphash_key = { .k0 = bi->bi_hash_seed } }; @@ -261,6 +259,7 @@ struct bkey_s_c bch2_hash_set_or_get_in_snapshot(struct btree_trans *trans, struct bkey_i *insert, enum btree_iter_update_trigger_flags flags) { + struct bch_fs *c = trans->c; struct btree_iter slot = {}; struct bkey_s_c k; bool found = false; @@ -288,7 +287,7 @@ struct bkey_s_c bch2_hash_set_or_get_in_snapshot(struct btree_trans *trans, } if (!ret) - ret = -BCH_ERR_ENOSPC_str_hash_create; + ret = bch_err_throw(c, ENOSPC_str_hash_create); out: bch2_trans_iter_exit(trans, &slot); bch2_trans_iter_exit(trans, iter); @@ -300,7 +299,7 @@ not_found: bch2_trans_iter_exit(trans, &slot); return k; } else if (!found && (flags & STR_HASH_must_replace)) { - ret = -BCH_ERR_ENOENT_str_hash_set_must_replace; + ret = bch_err_throw(c, ENOENT_str_hash_set_must_replace); } else { if (!found && slot.path) swap(*iter, slot); @@ -328,7 +327,7 @@ int bch2_hash_set_in_snapshot(struct btree_trans *trans, return ret; if (k.k) { bch2_trans_iter_exit(trans, &iter); - return -BCH_ERR_EEXIST_str_hash_set; + return bch_err_throw(trans->c, EEXIST_str_hash_set); } return 0; @@ -397,17 +396,27 @@ int bch2_hash_delete(struct btree_trans *trans, int bch2_repair_inode_hash_info(struct btree_trans *, struct bch_inode_unpacked *); struct snapshots_seen; +int bch2_str_hash_repair_key(struct btree_trans *, + struct snapshots_seen *, + const struct bch_hash_desc *, + struct bch_hash_info *, + struct btree_iter *, struct bkey_s_c, + struct btree_iter *, struct bkey_s_c, + bool *); + int __bch2_str_hash_check_key(struct btree_trans *, struct snapshots_seen *, const struct bch_hash_desc *, struct bch_hash_info *, - struct btree_iter *, struct bkey_s_c); + struct btree_iter *, struct bkey_s_c, + bool *); static inline int bch2_str_hash_check_key(struct btree_trans *trans, struct snapshots_seen *s, const struct bch_hash_desc *desc, struct bch_hash_info *hash_info, - struct btree_iter *k_iter, struct bkey_s_c hash_k) + struct btree_iter *k_iter, struct bkey_s_c hash_k, + bool *updated_before_k_pos) { if (hash_k.k->type != desc->key_type) return 0; @@ -415,7 +424,8 @@ static inline int bch2_str_hash_check_key(struct btree_trans *trans, if (likely(desc->hash_bkey(hash_info, hash_k) == hash_k.k->p.offset)) return 0; - return __bch2_str_hash_check_key(trans, s, desc, hash_info, k_iter, hash_k); + return __bch2_str_hash_check_key(trans, s, desc, hash_info, k_iter, hash_k, + updated_before_k_pos); } #endif /* _BCACHEFS_STR_HASH_H */ diff --git a/fs/bcachefs/subvolume.c b/fs/bcachefs/subvolume.c index 35c9f86a73c1..020587449123 100644 --- a/fs/bcachefs/subvolume.c +++ b/fs/bcachefs/subvolume.c @@ -130,10 +130,20 @@ static int check_subvol(struct btree_trans *trans, "subvolume %llu points to missing subvolume root %llu:%u", k.k->p.offset, le64_to_cpu(subvol.v->inode), le32_to_cpu(subvol.v->snapshot))) { - ret = bch2_subvolume_delete(trans, iter->pos.offset); - bch_err_msg(c, ret, "deleting subvolume %llu", iter->pos.offset); - ret = ret ?: -BCH_ERR_transaction_restart_nested; - goto err; + /* + * Recreate - any contents that are still disconnected + * will then get reattached under lost+found + */ + bch2_inode_init_early(c, &inode); + bch2_inode_init_late(c, &inode, bch2_current_time(c), + 0, 0, S_IFDIR|0700, 0, NULL); + inode.bi_inum = le64_to_cpu(subvol.v->inode); + inode.bi_snapshot = le32_to_cpu(subvol.v->snapshot); + inode.bi_subvol = k.k->p.offset; + inode.bi_parent_subvol = le32_to_cpu(subvol.v->fs_path_parent); + ret = __bch2_fsck_write_inode(trans, &inode); + if (ret) + goto err; } } else { goto err; @@ -141,13 +151,9 @@ static int check_subvol(struct btree_trans *trans, if (!BCH_SUBVOLUME_SNAP(subvol.v)) { u32 snapshot_root = bch2_snapshot_root(c, le32_to_cpu(subvol.v->snapshot)); - u32 snapshot_tree; - struct bch_snapshot_tree st; - - rcu_read_lock(); - snapshot_tree = snapshot_t(c, snapshot_root)->tree; - rcu_read_unlock(); + u32 snapshot_tree = bch2_snapshot_tree(c, snapshot_root); + struct bch_snapshot_tree st; ret = bch2_snapshot_tree_lookup(trans, snapshot_tree, &st); bch2_fs_inconsistent_on(bch2_err_matches(ret, ENOENT), c, @@ -259,6 +265,13 @@ void bch2_subvolume_to_text(struct printbuf *out, struct bch_fs *c, prt_printf(out, " creation_parent %u", le32_to_cpu(s.v->creation_parent)); prt_printf(out, " fs_parent %u", le32_to_cpu(s.v->fs_path_parent)); } + + if (BCH_SUBVOLUME_RO(s.v)) + prt_printf(out, " ro"); + if (BCH_SUBVOLUME_SNAP(s.v)) + prt_printf(out, " snapshot"); + if (BCH_SUBVOLUME_UNLINKED(s.v)) + prt_printf(out, " unlinked"); } static int subvolume_children_mod(struct btree_trans *trans, struct bpos pos, bool set) @@ -486,9 +499,12 @@ err: static int bch2_subvolume_delete(struct btree_trans *trans, u32 subvolid) { - return bch2_subvolumes_reparent(trans, subvolid) ?: + int ret = bch2_subvolumes_reparent(trans, subvolid) ?: commit_do(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc, __bch2_subvolume_delete(trans, subvolid)); + + bch2_recovery_pass_set_no_ratelimit(trans->c, BCH_RECOVERY_PASS_check_subvols); + return ret; } static void bch2_subvolume_wait_for_pagecache_and_delete(struct work_struct *work) @@ -597,7 +613,7 @@ int bch2_subvolume_create(struct btree_trans *trans, u64 inode, ret = bch2_bkey_get_empty_slot(trans, &dst_iter, BTREE_ID_subvolumes, POS(0, U32_MAX)); if (ret == -BCH_ERR_ENOSPC_btree_slot) - ret = -BCH_ERR_ENOSPC_subvolume_create; + ret = bch_err_throw(c, ENOSPC_subvolume_create); if (ret) return ret; @@ -703,8 +719,9 @@ static int __bch2_fs_upgrade_for_subvolumes(struct btree_trans *trans) return ret; if (!bkey_is_inode(k.k)) { - bch_err(trans->c, "root inode not found"); - ret = -BCH_ERR_ENOENT_inode; + struct bch_fs *c = trans->c; + bch_err(c, "root inode not found"); + ret = bch_err_throw(c, ENOENT_inode); goto err; } diff --git a/fs/bcachefs/super-io.c b/fs/bcachefs/super-io.c index 6687b9235d3c..6c2e1d647403 100644 --- a/fs/bcachefs/super-io.c +++ b/fs/bcachefs/super-io.c @@ -1112,7 +1112,7 @@ int bch2_write_super(struct bch_fs *c) prt_str(&buf, ")"); bch2_fs_fatal_error(c, ": %s", buf.buf); printbuf_exit(&buf); - ret = -BCH_ERR_sb_not_downgraded; + ret = bch_err_throw(c, sb_not_downgraded); goto out; } @@ -1142,7 +1142,7 @@ int bch2_write_super(struct bch_fs *c) if (c->opts.errors != BCH_ON_ERROR_continue && c->opts.errors != BCH_ON_ERROR_fix_safe) { - ret = -BCH_ERR_erofs_sb_err; + ret = bch_err_throw(c, erofs_sb_err); bch2_fs_fatal_error(c, "%s", buf.buf); } else { bch_err(c, "%s", buf.buf); @@ -1161,7 +1161,7 @@ int bch2_write_super(struct bch_fs *c) ca->disk_sb.seq); bch2_fs_fatal_error(c, "%s", buf.buf); printbuf_exit(&buf); - ret = -BCH_ERR_erofs_sb_err; + ret = bch_err_throw(c, erofs_sb_err); } } @@ -1215,7 +1215,7 @@ int bch2_write_super(struct bch_fs *c) !can_mount_with_written), c, ": Unable to write superblock to sufficient devices (from %ps)", (void *) _RET_IP_)) - ret = -BCH_ERR_erofs_sb_err; + ret = bch_err_throw(c, erofs_sb_err); out: /* Make new options visible after they're persistent: */ bch2_sb_update(c); diff --git a/fs/bcachefs/super.c b/fs/bcachefs/super.c index 11579b74c640..c46b1053a02c 100644 --- a/fs/bcachefs/super.c +++ b/fs/bcachefs/super.c @@ -104,7 +104,7 @@ const char * const bch2_dev_write_refs[] = { #undef x static void __bch2_print_str(struct bch_fs *c, const char *prefix, - const char *str, bool nonblocking) + const char *str) { #ifdef __KERNEL__ struct stdio_redirect *stdio = bch2_fs_stdio_redirect(c); @@ -114,17 +114,12 @@ static void __bch2_print_str(struct bch_fs *c, const char *prefix, return; } #endif - bch2_print_string_as_lines(KERN_ERR, str, nonblocking); + bch2_print_string_as_lines(KERN_ERR, str); } void bch2_print_str(struct bch_fs *c, const char *prefix, const char *str) { - __bch2_print_str(c, prefix, str, false); -} - -void bch2_print_str_nonblocking(struct bch_fs *c, const char *prefix, const char *str) -{ - __bch2_print_str(c, prefix, str, true); + __bch2_print_str(c, prefix, str); } __printf(2, 0) @@ -215,27 +210,20 @@ static int bch2_dev_alloc(struct bch_fs *, unsigned); static int bch2_dev_sysfs_online(struct bch_fs *, struct bch_dev *); static void bch2_dev_io_ref_stop(struct bch_dev *, int); static void __bch2_dev_read_only(struct bch_fs *, struct bch_dev *); -static int bch2_fs_init_rw(struct bch_fs *); struct bch_fs *bch2_dev_to_fs(dev_t dev) { - struct bch_fs *c; - - mutex_lock(&bch_fs_list_lock); - rcu_read_lock(); + guard(mutex)(&bch_fs_list_lock); + guard(rcu)(); + struct bch_fs *c; list_for_each_entry(c, &bch_fs_list, list) for_each_member_device_rcu(c, ca, NULL) if (ca->disk_sb.bdev && ca->disk_sb.bdev->bd_dev == dev) { closure_get(&c->cl); - goto found; + return c; } - c = NULL; -found: - rcu_read_unlock(); - mutex_unlock(&bch_fs_list_lock); - - return c; + return NULL; } static struct bch_fs *__bch2_uuid_to_fs(__uuid_t uuid) @@ -480,16 +468,16 @@ static int __bch2_fs_read_write(struct bch_fs *c, bool early) BUG_ON(!test_bit(BCH_FS_may_go_rw, &c->flags)); if (WARN_ON(c->sb.features & BIT_ULL(BCH_FEATURE_no_alloc_info))) - return -BCH_ERR_erofs_no_alloc_info; + return bch_err_throw(c, erofs_no_alloc_info); if (test_bit(BCH_FS_initial_gc_unfixed, &c->flags)) { bch_err(c, "cannot go rw, unfixed btree errors"); - return -BCH_ERR_erofs_unfixed_errors; + return bch_err_throw(c, erofs_unfixed_errors); } if (c->sb.features & BIT_ULL(BCH_FEATURE_small_image)) { bch_err(c, "cannot go rw, filesystem is an unresized image file"); - return -BCH_ERR_erofs_filesystem_full; + return bch_err_throw(c, erofs_filesystem_full); } if (test_bit(BCH_FS_rw, &c->flags)) @@ -507,13 +495,12 @@ static int __bch2_fs_read_write(struct bch_fs *c, bool early) clear_bit(BCH_FS_clean_shutdown, &c->flags); - rcu_read_lock(); - for_each_online_member_rcu(c, ca) - if (ca->mi.state == BCH_MEMBER_STATE_rw) { - bch2_dev_allocator_add(c, ca); - enumerated_ref_start(&ca->io_ref[WRITE]); - } - rcu_read_unlock(); + scoped_guard(rcu) + for_each_online_member_rcu(c, ca) + if (ca->mi.state == BCH_MEMBER_STATE_rw) { + bch2_dev_allocator_add(c, ca); + enumerated_ref_start(&ca->io_ref[WRITE]); + } bch2_recalc_capacity(c); @@ -571,13 +558,13 @@ int bch2_fs_read_write(struct bch_fs *c) { if (c->opts.recovery_pass_last && c->opts.recovery_pass_last < BCH_RECOVERY_PASS_journal_replay) - return -BCH_ERR_erofs_norecovery; + return bch_err_throw(c, erofs_norecovery); if (c->opts.nochanges) - return -BCH_ERR_erofs_nochanges; + return bch_err_throw(c, erofs_nochanges); if (c->sb.features & BIT_ULL(BCH_FEATURE_no_alloc_info)) - return -BCH_ERR_erofs_no_alloc_info; + return bch_err_throw(c, erofs_no_alloc_info); return __bch2_fs_read_write(c, false); } @@ -762,7 +749,7 @@ static int bch2_fs_online(struct bch_fs *c) if (c->sb.multi_device && __bch2_uuid_to_fs(c->sb.uuid)) { bch_err(c, "filesystem UUID already open"); - return -BCH_ERR_filesystem_uuid_already_open; + return bch_err_throw(c, filesystem_uuid_already_open); } ret = bch2_fs_chardev_init(c); @@ -806,7 +793,7 @@ err: return ret; } -static int bch2_fs_init_rw(struct bch_fs *c) +int bch2_fs_init_rw(struct bch_fs *c) { if (test_bit(BCH_FS_rw_init_done, &c->flags)) return 0; @@ -821,7 +808,7 @@ static int bch2_fs_init_rw(struct bch_fs *c) WQ_HIGHPRI|WQ_FREEZABLE|WQ_MEM_RECLAIM, 1)) || !(c->write_ref_wq = alloc_workqueue("bcachefs_write_ref", WQ_FREEZABLE, 0))) - return -BCH_ERR_ENOMEM_fs_other_alloc; + return bch_err_throw(c, ENOMEM_fs_other_alloc); int ret = bch2_fs_btree_interior_update_init(c) ?: bch2_fs_btree_write_buffer_init(c) ?: @@ -1002,7 +989,7 @@ static struct bch_fs *bch2_fs_alloc(struct bch_sb *sb, struct bch_opts *opts, mempool_init_kvmalloc_pool(&c->btree_bounce_pool, 1, c->opts.btree_node_size) || mempool_init_kmalloc_pool(&c->large_bkey_pool, 1, 2048)) { - ret = -BCH_ERR_ENOMEM_fs_other_alloc; + ret = bch_err_throw(c, ENOMEM_fs_other_alloc); goto err; } @@ -1027,21 +1014,29 @@ static struct bch_fs *bch2_fs_alloc(struct bch_sb *sb, struct bch_opts *opts, if (ret) goto err; + if (go_rw_in_recovery(c)) { + /* + * start workqueues/kworkers early - kthread creation checks for + * pending signals, which is _very_ annoying + */ + ret = bch2_fs_init_rw(c); + if (ret) + goto err; + } + #ifdef CONFIG_UNICODE - /* Default encoding until we can potentially have more as an option. */ - c->cf_encoding = utf8_load(BCH_FS_DEFAULT_UTF8_ENCODING); - if (IS_ERR(c->cf_encoding)) { - printk(KERN_ERR "Cannot load UTF-8 encoding for filesystem. Version: %u.%u.%u", - unicode_major(BCH_FS_DEFAULT_UTF8_ENCODING), - unicode_minor(BCH_FS_DEFAULT_UTF8_ENCODING), - unicode_rev(BCH_FS_DEFAULT_UTF8_ENCODING)); - ret = -EINVAL; - goto err; + if (bch2_fs_casefold_enabled(c)) { + /* Default encoding until we can potentially have more as an option. */ + c->cf_encoding = utf8_load(BCH_FS_DEFAULT_UTF8_ENCODING); + if (IS_ERR(c->cf_encoding)) { + printk(KERN_ERR "Cannot load UTF-8 encoding for filesystem. Version: %u.%u.%u", + unicode_major(BCH_FS_DEFAULT_UTF8_ENCODING), + unicode_minor(BCH_FS_DEFAULT_UTF8_ENCODING), + unicode_rev(BCH_FS_DEFAULT_UTF8_ENCODING)); + ret = -EINVAL; + goto err; + } } - bch_info(c, "Using encoding defined by superblock: utf8-%u.%u.%u", - unicode_major(BCH_FS_DEFAULT_UTF8_ENCODING), - unicode_minor(BCH_FS_DEFAULT_UTF8_ENCODING), - unicode_rev(BCH_FS_DEFAULT_UTF8_ENCODING)); #else if (c->sb.features & BIT_ULL(BCH_FEATURE_casefolding)) { printk(KERN_ERR "Cannot mount a filesystem with casefolding on a kernel without CONFIG_UNICODE\n"); @@ -1083,12 +1078,13 @@ noinline_for_stack static void print_mount_opts(struct bch_fs *c) { enum bch_opt_id i; - struct printbuf p = PRINTBUF; - bool first = true; + CLASS(printbuf, p)(); + bch2_log_msg_start(c, &p); prt_str(&p, "starting version "); bch2_version_to_text(&p, c->sb.version); + bool first = true; for (i = 0; i < bch2_opts_nr; i++) { const struct bch_option *opt = &bch2_opt_table[i]; u64 v = bch2_opt_get_by_id(&c->opts, i); @@ -1105,17 +1101,24 @@ static void print_mount_opts(struct bch_fs *c) } if (c->sb.version_incompat_allowed != c->sb.version) { - prt_printf(&p, "\n allowing incompatible features above "); + prt_printf(&p, "\nallowing incompatible features above "); bch2_version_to_text(&p, c->sb.version_incompat_allowed); } if (c->opts.verbose) { - prt_printf(&p, "\n features: "); + prt_printf(&p, "\nfeatures: "); prt_bitflags(&p, bch2_sb_features, c->sb.features); } - bch_info(c, "%s", p.buf); - printbuf_exit(&p); + if (c->sb.multi_device) { + prt_printf(&p, "\nwith devices"); + for_each_online_member(c, ca, BCH_DEV_READ_REF_bch2_online_devs) { + prt_char(&p, ' '); + prt_str(&p, ca->name); + } + } + + bch2_print_str(c, KERN_INFO, p.buf); } static bool bch2_fs_may_start(struct bch_fs *c) @@ -1159,8 +1162,14 @@ int bch2_fs_start(struct bch_fs *c) print_mount_opts(c); + if (c->cf_encoding) + bch_info(c, "Using encoding defined by superblock: utf8-%u.%u.%u", + unicode_major(BCH_FS_DEFAULT_UTF8_ENCODING), + unicode_minor(BCH_FS_DEFAULT_UTF8_ENCODING), + unicode_rev(BCH_FS_DEFAULT_UTF8_ENCODING)); + if (!bch2_fs_may_start(c)) - return -BCH_ERR_insufficient_devices_to_start; + return bch_err_throw(c, insufficient_devices_to_start); down_write(&c->state_lock); mutex_lock(&c->sb_lock); @@ -1171,7 +1180,7 @@ int bch2_fs_start(struct bch_fs *c) sizeof(struct bch_sb_field_ext) / sizeof(u64))) { mutex_unlock(&c->sb_lock); up_write(&c->state_lock); - ret = -BCH_ERR_ENOSPC_sb; + ret = bch_err_throw(c, ENOSPC_sb); goto err; } @@ -1182,22 +1191,20 @@ int bch2_fs_start(struct bch_fs *c) goto err; } - rcu_read_lock(); - for_each_online_member_rcu(c, ca) - bch2_members_v2_get_mut(c->disk_sb.sb, ca->dev_idx)->last_mount = - cpu_to_le64(now); - rcu_read_unlock(); + scoped_guard(rcu) + for_each_online_member_rcu(c, ca) + bch2_members_v2_get_mut(c->disk_sb.sb, ca->dev_idx)->last_mount = + cpu_to_le64(now); /* * Dno't write superblock yet: recovery might have to downgrade */ mutex_unlock(&c->sb_lock); - rcu_read_lock(); - for_each_online_member_rcu(c, ca) - if (ca->mi.state == BCH_MEMBER_STATE_rw) - bch2_dev_allocator_add(c, ca); - rcu_read_unlock(); + scoped_guard(rcu) + for_each_online_member_rcu(c, ca) + if (ca->mi.state == BCH_MEMBER_STATE_rw) + bch2_dev_allocator_add(c, ca); bch2_recalc_capacity(c); up_write(&c->state_lock); @@ -1215,7 +1222,7 @@ int bch2_fs_start(struct bch_fs *c) goto err; if (bch2_fs_init_fault("fs_start")) { - ret = -BCH_ERR_injected_fs_start; + ret = bch_err_throw(c, injected_fs_start); goto err; } @@ -1242,11 +1249,11 @@ static int bch2_dev_may_add(struct bch_sb *sb, struct bch_fs *c) struct bch_member m = bch2_sb_member_get(sb, sb->dev_idx); if (le16_to_cpu(sb->block_size) != block_sectors(c)) - return -BCH_ERR_mismatched_block_size; + return bch_err_throw(c, mismatched_block_size); if (le16_to_cpu(m.bucket_size) < BCH_SB_BTREE_NODE_SIZE(c->disk_sb.sb)) - return -BCH_ERR_bucket_size_too_small; + return bch_err_throw(c, bucket_size_too_small); return 0; } @@ -1557,7 +1564,7 @@ static int bch2_dev_alloc(struct bch_fs *c, unsigned dev_idx) bch2_dev_attach(c, ca, dev_idx); return 0; err: - return -BCH_ERR_ENOMEM_dev_alloc; + return bch_err_throw(c, ENOMEM_dev_alloc); } static int __bch2_dev_attach_bdev(struct bch_dev *ca, struct bch_sb_handle *sb) @@ -1567,13 +1574,13 @@ static int __bch2_dev_attach_bdev(struct bch_dev *ca, struct bch_sb_handle *sb) if (bch2_dev_is_online(ca)) { bch_err(ca, "already have device online in slot %u", sb->sb->dev_idx); - return -BCH_ERR_device_already_online; + return bch_err_throw(ca->fs, device_already_online); } if (get_capacity(sb->bdev->bd_disk) < ca->mi.bucket_size * ca->mi.nbuckets) { bch_err(ca, "cannot online: device too small"); - return -BCH_ERR_device_size_too_small; + return bch_err_throw(ca->fs, device_size_too_small); } BUG_ON(!enumerated_ref_is_zero(&ca->io_ref[READ])); @@ -1725,7 +1732,7 @@ int __bch2_dev_set_state(struct bch_fs *c, struct bch_dev *ca, return 0; if (!bch2_dev_state_allowed(c, ca, new_state, flags)) - return -BCH_ERR_device_state_not_allowed; + return bch_err_throw(c, device_state_not_allowed); if (new_state != BCH_MEMBER_STATE_rw) __bch2_dev_read_only(c, ca); @@ -1778,7 +1785,7 @@ int bch2_dev_remove(struct bch_fs *c, struct bch_dev *ca, int flags) if (!bch2_dev_state_allowed(c, ca, BCH_MEMBER_STATE_failed, flags)) { bch_err(ca, "Cannot remove without losing data"); - ret = -BCH_ERR_device_state_not_allowed; + ret = bch_err_throw(c, device_state_not_allowed); goto err; } @@ -1914,7 +1921,7 @@ int bch2_dev_add(struct bch_fs *c, const char *path) if (list_empty(&c->list)) { mutex_lock(&bch_fs_list_lock); if (__bch2_uuid_to_fs(c->sb.uuid)) - ret = -BCH_ERR_filesystem_uuid_already_open; + ret = bch_err_throw(c, filesystem_uuid_already_open); else list_add(&c->list, &bch_fs_list); mutex_unlock(&bch_fs_list_lock); @@ -2001,6 +2008,22 @@ int bch2_dev_add(struct bch_fs *c, const char *path) goto err_late; } + /* + * We just changed the superblock UUID, invalidate cache and send a + * uevent to update /dev/disk/by-uuid + */ + invalidate_bdev(ca->disk_sb.bdev); + + char uuid_str[37]; + snprintf(uuid_str, sizeof(uuid_str), "UUID=%pUb", &c->sb.uuid); + + char *envp[] = { + "CHANGE=uuid", + uuid_str, + NULL, + }; + kobject_uevent_env(&ca->disk_sb.bdev->bd_device.kobj, KOBJ_CHANGE, envp); + up_write(&c->state_lock); out: printbuf_exit(&label); @@ -2101,7 +2124,7 @@ int bch2_dev_offline(struct bch_fs *c, struct bch_dev *ca, int flags) if (!bch2_dev_state_allowed(c, ca, BCH_MEMBER_STATE_failed, flags)) { bch_err(ca, "Cannot offline required disk"); up_write(&c->state_lock); - return -BCH_ERR_device_state_not_allowed; + return bch_err_throw(c, device_state_not_allowed); } __bch2_dev_offline(c, ca); @@ -2140,7 +2163,7 @@ int bch2_dev_resize(struct bch_fs *c, struct bch_dev *ca, u64 nbuckets) if (nbuckets > BCH_MEMBER_NBUCKETS_MAX) { bch_err(ca, "New device size too big (%llu greater than max %u)", nbuckets, BCH_MEMBER_NBUCKETS_MAX); - ret = -BCH_ERR_device_size_too_big; + ret = bch_err_throw(c, device_size_too_big); goto err; } @@ -2148,7 +2171,7 @@ int bch2_dev_resize(struct bch_fs *c, struct bch_dev *ca, u64 nbuckets) get_capacity(ca->disk_sb.bdev->bd_disk) < ca->mi.bucket_size * nbuckets) { bch_err(ca, "New size larger than device"); - ret = -BCH_ERR_device_size_too_small; + ret = bch_err_throw(c, device_size_too_small); goto err; } @@ -2383,7 +2406,7 @@ struct bch_fs *bch2_fs_open(darray_const_str *devices, } if (opts->nochanges && !opts->read_only) { - ret = -BCH_ERR_erofs_nochanges; + ret = bch_err_throw(c, erofs_nochanges); goto err_print; } diff --git a/fs/bcachefs/super.h b/fs/bcachefs/super.h index dc52f06cb2b9..e90bab9afe78 100644 --- a/fs/bcachefs/super.h +++ b/fs/bcachefs/super.h @@ -46,6 +46,7 @@ void __bch2_fs_stop(struct bch_fs *); void bch2_fs_free(struct bch_fs *); void bch2_fs_stop(struct bch_fs *); +int bch2_fs_init_rw(struct bch_fs *); int bch2_fs_start(struct bch_fs *); struct bch_fs *bch2_fs_open(darray_const_str *, struct bch_opts *); diff --git a/fs/bcachefs/sysfs.c b/fs/bcachefs/sysfs.c index 1a55196d69f1..05848375cea2 100644 --- a/fs/bcachefs/sysfs.c +++ b/fs/bcachefs/sysfs.c @@ -26,6 +26,7 @@ #include "disk_groups.h" #include "ec.h" #include "enumerated_ref.h" +#include "error.h" #include "inode.h" #include "journal.h" #include "journal_reclaim.h" @@ -37,6 +38,7 @@ #include "rebalance.h" #include "recovery_passes.h" #include "replicas.h" +#include "sb-errors.h" #include "super-io.h" #include "tests.h" @@ -143,6 +145,7 @@ do { \ write_attribute(trigger_gc); write_attribute(trigger_discards); write_attribute(trigger_invalidates); +write_attribute(trigger_journal_commit); write_attribute(trigger_journal_flush); write_attribute(trigger_journal_writes); write_attribute(trigger_btree_cache_shrink); @@ -151,6 +154,7 @@ write_attribute(trigger_btree_updates); write_attribute(trigger_freelist_wakeup); write_attribute(trigger_recalc_capacity); write_attribute(trigger_delete_dead_snapshots); +write_attribute(trigger_emergency_read_only); read_attribute(gc_gens_pos); read_attribute(uuid); @@ -172,6 +176,7 @@ read_attribute(btree_write_stats); read_attribute(btree_cache_size); read_attribute(compression_stats); +read_attribute(errors); read_attribute(journal_debug); read_attribute(btree_cache); read_attribute(btree_key_cache); @@ -353,6 +358,9 @@ SHOW(bch2_fs) if (attr == &sysfs_compression_stats) bch2_compression_stats_to_text(out, c); + if (attr == &sysfs_errors) + bch2_fs_errors_to_text(out, c); + if (attr == &sysfs_new_stripes) bch2_new_stripes_to_text(out, c); @@ -428,6 +436,9 @@ STORE(bch2_fs) if (attr == &sysfs_trigger_invalidates) bch2_do_invalidates(c); + if (attr == &sysfs_trigger_journal_commit) + bch2_journal_flush(&c->journal); + if (attr == &sysfs_trigger_journal_flush) { bch2_journal_flush_all_pins(&c->journal); bch2_journal_meta(&c->journal); @@ -448,6 +459,16 @@ STORE(bch2_fs) if (attr == &sysfs_trigger_delete_dead_snapshots) __bch2_delete_dead_snapshots(c); + if (attr == &sysfs_trigger_emergency_read_only) { + struct printbuf buf = PRINTBUF; + bch2_log_msg_start(c, &buf); + + prt_printf(&buf, "shutdown by sysfs\n"); + bch2_fs_emergency_read_only2(c, &buf); + bch2_print_str(c, KERN_ERR, buf.buf); + printbuf_exit(&buf); + } + #ifdef CONFIG_BCACHEFS_TESTS if (attr == &sysfs_perf_test) { char *tmp = kstrdup(buf, GFP_KERNEL), *p = tmp; @@ -483,6 +504,7 @@ struct attribute *bch2_fs_files[] = { &sysfs_recovery_status, &sysfs_compression_stats, + &sysfs_errors, #ifdef CONFIG_BCACHEFS_TESTS &sysfs_perf_test, @@ -571,6 +593,7 @@ struct attribute *bch2_fs_internal_files[] = { &sysfs_trigger_gc, &sysfs_trigger_discards, &sysfs_trigger_invalidates, + &sysfs_trigger_journal_commit, &sysfs_trigger_journal_flush, &sysfs_trigger_journal_writes, &sysfs_trigger_btree_cache_shrink, @@ -579,6 +602,7 @@ struct attribute *bch2_fs_internal_files[] = { &sysfs_trigger_freelist_wakeup, &sysfs_trigger_recalc_capacity, &sysfs_trigger_delete_dead_snapshots, + &sysfs_trigger_emergency_read_only, &sysfs_gc_gens_pos, diff --git a/fs/bcachefs/trace.h b/fs/bcachefs/trace.h index 8cb5b40704fd..9c5a9c551f03 100644 --- a/fs/bcachefs/trace.h +++ b/fs/bcachefs/trace.h @@ -199,6 +199,50 @@ DECLARE_EVENT_CLASS(bio, (unsigned long long)__entry->sector, __entry->nr_sector) ); +/* errors */ + +TRACE_EVENT(error_throw, + TP_PROTO(struct bch_fs *c, int bch_err, unsigned long ip), + TP_ARGS(c, bch_err, ip), + + TP_STRUCT__entry( + __field(dev_t, dev ) + __field(int, err ) + __array(char, err_str, 32 ) + __array(char, ip, 32 ) + ), + + TP_fast_assign( + __entry->dev = c->dev; + __entry->err = bch_err; + strscpy(__entry->err_str, bch2_err_str(bch_err), sizeof(__entry->err_str)); + snprintf(__entry->ip, sizeof(__entry->ip), "%ps", (void *) ip); + ), + + TP_printk("%d,%d %s ret %s", MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->ip, __entry->err_str) +); + +TRACE_EVENT(error_downcast, + TP_PROTO(int bch_err, int std_err, unsigned long ip), + TP_ARGS(bch_err, std_err, ip), + + TP_STRUCT__entry( + __array(char, bch_err, 32 ) + __array(char, std_err, 32 ) + __array(char, ip, 32 ) + ), + + TP_fast_assign( + strscpy(__entry->bch_err, bch2_err_str(bch_err), sizeof(__entry->bch_err)); + strscpy(__entry->std_err, bch2_err_str(std_err), sizeof(__entry->std_err)); + snprintf(__entry->ip, sizeof(__entry->ip), "%ps", (void *) ip); + ), + + TP_printk("%s ret %s -> %s %s", __entry->ip, + __entry->bch_err, __entry->std_err, __entry->ip) +); + /* disk_accounting.c */ TRACE_EVENT(accounting_mem_insert, @@ -1036,34 +1080,14 @@ TRACE_EVENT(trans_blocked_journal_reclaim, __entry->must_wait) ); -TRACE_EVENT(trans_restart_journal_preres_get, - TP_PROTO(struct btree_trans *trans, - unsigned long caller_ip, - unsigned flags), - TP_ARGS(trans, caller_ip, flags), - - TP_STRUCT__entry( - __array(char, trans_fn, 32 ) - __field(unsigned long, caller_ip ) - __field(unsigned, flags ) - ), - - TP_fast_assign( - strscpy(__entry->trans_fn, trans->fn, sizeof(__entry->trans_fn)); - __entry->caller_ip = caller_ip; - __entry->flags = flags; - ), - - TP_printk("%s %pS %x", __entry->trans_fn, - (void *) __entry->caller_ip, - __entry->flags) -); - +#if 0 +/* todo: bring back dynamic fault injection */ DEFINE_EVENT(transaction_event, trans_restart_fault_inject, TP_PROTO(struct btree_trans *trans, unsigned long caller_ip), TP_ARGS(trans, caller_ip) ); +#endif DEFINE_EVENT(transaction_event, trans_traverse_all, TP_PROTO(struct btree_trans *trans, @@ -1151,19 +1175,6 @@ DEFINE_EVENT(transaction_restart_iter, trans_restart_relock_parent_for_fill, TP_ARGS(trans, caller_ip, path) ); -DEFINE_EVENT(transaction_restart_iter, trans_restart_relock_after_fill, - TP_PROTO(struct btree_trans *trans, - unsigned long caller_ip, - struct btree_path *path), - TP_ARGS(trans, caller_ip, path) -); - -DEFINE_EVENT(transaction_event, trans_restart_key_cache_upgrade, - TP_PROTO(struct btree_trans *trans, - unsigned long caller_ip), - TP_ARGS(trans, caller_ip) -); - DEFINE_EVENT(transaction_restart_iter, trans_restart_relock_key_cache_fill, TP_PROTO(struct btree_trans *trans, unsigned long caller_ip, @@ -1185,13 +1196,6 @@ DEFINE_EVENT(transaction_restart_iter, trans_restart_relock_path_intent, TP_ARGS(trans, caller_ip, path) ); -DEFINE_EVENT(transaction_restart_iter, trans_restart_traverse, - TP_PROTO(struct btree_trans *trans, - unsigned long caller_ip, - struct btree_path *path), - TP_ARGS(trans, caller_ip, path) -); - DEFINE_EVENT(transaction_restart_iter, trans_restart_memory_allocation_failure, TP_PROTO(struct btree_trans *trans, unsigned long caller_ip, @@ -1250,44 +1254,6 @@ TRACE_EVENT(trans_restart_mem_realloced, __entry->bytes) ); -TRACE_EVENT(trans_restart_key_cache_key_realloced, - TP_PROTO(struct btree_trans *trans, - unsigned long caller_ip, - struct btree_path *path, - unsigned old_u64s, - unsigned new_u64s), - TP_ARGS(trans, caller_ip, path, old_u64s, new_u64s), - - TP_STRUCT__entry( - __array(char, trans_fn, 32 ) - __field(unsigned long, caller_ip ) - __field(enum btree_id, btree_id ) - TRACE_BPOS_entries(pos) - __field(u32, old_u64s ) - __field(u32, new_u64s ) - ), - - TP_fast_assign( - strscpy(__entry->trans_fn, trans->fn, sizeof(__entry->trans_fn)); - __entry->caller_ip = caller_ip; - - __entry->btree_id = path->btree_id; - TRACE_BPOS_assign(pos, path->pos); - __entry->old_u64s = old_u64s; - __entry->new_u64s = new_u64s; - ), - - TP_printk("%s %pS btree %s pos %llu:%llu:%u old_u64s %u new_u64s %u", - __entry->trans_fn, - (void *) __entry->caller_ip, - bch2_btree_id_str(__entry->btree_id), - __entry->pos_inode, - __entry->pos_offset, - __entry->pos_snapshot, - __entry->old_u64s, - __entry->new_u64s) -); - DEFINE_EVENT(transaction_event, trans_restart_write_buffer_flush, TP_PROTO(struct btree_trans *trans, unsigned long caller_ip), @@ -1431,28 +1397,44 @@ DEFINE_EVENT(fs_str, data_update, TP_ARGS(c, str) ); +DEFINE_EVENT(fs_str, io_move_pred, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) +); + DEFINE_EVENT(fs_str, io_move_created_rebalance, TP_PROTO(struct bch_fs *c, const char *str), TP_ARGS(c, str) ); -TRACE_EVENT(error_downcast, - TP_PROTO(int bch_err, int std_err, unsigned long ip), - TP_ARGS(bch_err, std_err, ip), +DEFINE_EVENT(fs_str, io_move_evacuate_bucket, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) +); - TP_STRUCT__entry( - __array(char, bch_err, 32 ) - __array(char, std_err, 32 ) - __array(char, ip, 32 ) - ), +DEFINE_EVENT(fs_str, extent_trim_atomic, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) +); - TP_fast_assign( - strscpy(__entry->bch_err, bch2_err_str(bch_err), sizeof(__entry->bch_err)); - strscpy(__entry->std_err, bch2_err_str(std_err), sizeof(__entry->std_err)); - snprintf(__entry->ip, sizeof(__entry->ip), "%ps", (void *) ip); - ), +DEFINE_EVENT(fs_str, btree_iter_peek_slot, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) +); + +DEFINE_EVENT(fs_str, __btree_iter_peek, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) +); - TP_printk("%s -> %s %s", __entry->bch_err, __entry->std_err, __entry->ip) +DEFINE_EVENT(fs_str, btree_iter_peek_max, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) +); + +DEFINE_EVENT(fs_str, btree_iter_peek_prev_min, + TP_PROTO(struct bch_fs *c, const char *str), + TP_ARGS(c, str) ); #ifdef CONFIG_BCACHEFS_PATH_TRACEPOINTS @@ -1867,21 +1849,6 @@ TRACE_EVENT(btree_path_free, __entry->dup_locked) ); -TRACE_EVENT(btree_path_free_trans_begin, - TP_PROTO(btree_path_idx_t path), - TP_ARGS(path), - - TP_STRUCT__entry( - __field(btree_path_idx_t, idx ) - ), - - TP_fast_assign( - __entry->idx = path; - ), - - TP_printk(" path %3u", __entry->idx) -); - #else /* CONFIG_BCACHEFS_PATH_TRACEPOINTS */ #ifndef _TRACE_BCACHEFS_H @@ -1899,7 +1866,6 @@ static inline void trace_btree_path_traverse_start(struct btree_trans *trans, st static inline void trace_btree_path_traverse_end(struct btree_trans *trans, struct btree_path *path) {} static inline void trace_btree_path_set_pos(struct btree_trans *trans, struct btree_path *path, struct bpos *new_pos) {} static inline void trace_btree_path_free(struct btree_trans *trans, btree_path_idx_t path, struct btree_path *dup) {} -static inline void trace_btree_path_free_trans_begin(btree_path_idx_t path) {} #endif #endif /* CONFIG_BCACHEFS_PATH_TRACEPOINTS */ diff --git a/fs/bcachefs/util.c b/fs/bcachefs/util.c index dc3817f545fa..df9a6071fe18 100644 --- a/fs/bcachefs/util.c +++ b/fs/bcachefs/util.c @@ -262,8 +262,7 @@ static bool string_is_spaces(const char *str) return true; } -void bch2_print_string_as_lines(const char *prefix, const char *lines, - bool nonblocking) +void bch2_print_string_as_lines(const char *prefix, const char *lines) { bool locked = false; const char *p; @@ -273,12 +272,7 @@ void bch2_print_string_as_lines(const char *prefix, const char *lines, return; } - if (!nonblocking) { - console_lock(); - locked = true; - } else { - locked = console_trylock(); - } + locked = console_trylock(); while (*lines) { p = strchrnul(lines, '\n'); diff --git a/fs/bcachefs/util.h b/fs/bcachefs/util.h index 0a4b1d433621..6488f098d140 100644 --- a/fs/bcachefs/util.h +++ b/fs/bcachefs/util.h @@ -214,7 +214,7 @@ u64 bch2_read_flag_list(const char *, const char * const[]); void bch2_prt_u64_base2_nbits(struct printbuf *, u64, unsigned); void bch2_prt_u64_base2(struct printbuf *, u64); -void bch2_print_string_as_lines(const char *, const char *, bool); +void bch2_print_string_as_lines(const char *, const char *); typedef DARRAY(unsigned long) bch_stacktrace; int bch2_save_backtrace(bch_stacktrace *stack, struct task_struct *, unsigned, gfp_t); diff --git a/fs/btrfs/block-group.h b/fs/btrfs/block-group.h index 9de356bcb411..aa176cc9a324 100644 --- a/fs/btrfs/block-group.h +++ b/fs/btrfs/block-group.h @@ -83,6 +83,8 @@ enum btrfs_block_group_flags { BLOCK_GROUP_FLAG_ZONED_DATA_RELOC, /* Does the block group need to be added to the free space tree? */ BLOCK_GROUP_FLAG_NEEDS_FREE_SPACE, + /* Set after we add a new block group to the free space tree. */ + BLOCK_GROUP_FLAG_FREE_SPACE_ADDED, /* Indicate that the block group is placed on a sequential zone */ BLOCK_GROUP_FLAG_SEQUENTIAL_ZONE, /* diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index c7cc24a5dd5e..8c597fa60523 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -1377,7 +1377,10 @@ static int btrfs_wq_run_delayed_node(struct btrfs_delayed_root *delayed_root, void btrfs_assert_delayed_root_empty(struct btrfs_fs_info *fs_info) { - WARN_ON(btrfs_first_delayed_node(fs_info->delayed_root)); + struct btrfs_delayed_node *node = btrfs_first_delayed_node(fs_info->delayed_root); + + if (WARN_ON(node)) + refcount_dec(&node->refs); } static bool could_end_wait(struct btrfs_delayed_root *delayed_root, int seq) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 1beb9458f622..0d6ad7512f21 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1835,6 +1835,8 @@ void btrfs_put_root(struct btrfs_root *root) if (refcount_dec_and_test(&root->refs)) { if (WARN_ON(!xa_empty(&root->inodes))) xa_destroy(&root->inodes); + if (WARN_ON(!xa_empty(&root->delayed_nodes))) + xa_destroy(&root->delayed_nodes); WARN_ON(test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state)); if (root->anon_dev) free_anon_bdev(root->anon_dev); @@ -2156,8 +2158,7 @@ static int load_global_roots_objectid(struct btrfs_root *tree_root, found = true; root = read_tree_root_path(tree_root, path, &key); if (IS_ERR(root)) { - if (!btrfs_test_opt(fs_info, IGNOREBADROOTS)) - ret = PTR_ERR(root); + ret = PTR_ERR(root); break; } set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state); @@ -4310,8 +4311,8 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info) * * So wait for all ongoing ordered extents to complete and then run * delayed iputs. This works because once we reach this point no one - * can either create new ordered extents nor create delayed iputs - * through some other means. + * can create new ordered extents, but delayed iputs can still be added + * by a reclaim worker (see comments further below). * * Also note that btrfs_wait_ordered_roots() is not safe here, because * it waits for BTRFS_ORDERED_COMPLETE to be set on an ordered extent, @@ -4322,15 +4323,29 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info) btrfs_flush_workqueue(fs_info->endio_write_workers); /* Ordered extents for free space inodes. */ btrfs_flush_workqueue(fs_info->endio_freespace_worker); + /* + * Run delayed iputs in case an async reclaim worker is waiting for them + * to be run as mentioned above. + */ btrfs_run_delayed_iputs(fs_info); - /* There should be no more workload to generate new delayed iputs. */ - set_bit(BTRFS_FS_STATE_NO_DELAYED_IPUT, &fs_info->fs_state); cancel_work_sync(&fs_info->async_reclaim_work); cancel_work_sync(&fs_info->async_data_reclaim_work); cancel_work_sync(&fs_info->preempt_reclaim_work); cancel_work_sync(&fs_info->em_shrinker_work); + /* + * Run delayed iputs again because an async reclaim worker may have + * added new ones if it was flushing delalloc: + * + * shrink_delalloc() -> btrfs_start_delalloc_roots() -> + * start_delalloc_inodes() -> btrfs_add_delayed_iput() + */ + btrfs_run_delayed_iputs(fs_info); + + /* There should be no more workload to generate new delayed iputs. */ + set_bit(BTRFS_FS_STATE_NO_DELAYED_IPUT, &fs_info->fs_state); + /* Cancel or finish ongoing discard work */ btrfs_discard_cleanup(fs_info); diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 849199768664..1dc931c4937f 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -4312,7 +4312,6 @@ static int try_release_subpage_extent_buffer(struct folio *folio) spin_unlock(&eb->refs_lock); continue; } - xa_unlock_irq(&fs_info->buffer_tree); /* * If tree ref isn't set then we know the ref on this eb is a @@ -4329,6 +4328,7 @@ static int try_release_subpage_extent_buffer(struct folio *folio) * check the folio private at the end. And * release_extent_buffer() will release the refs_lock. */ + xa_unlock_irq(&fs_info->buffer_tree); release_extent_buffer(eb); xa_lock_irq(&fs_info->buffer_tree); } diff --git a/fs/btrfs/free-space-tree.c b/fs/btrfs/free-space-tree.c index 0c573d46639a..a83c268f7f87 100644 --- a/fs/btrfs/free-space-tree.c +++ b/fs/btrfs/free-space-tree.c @@ -1115,11 +1115,21 @@ static int populate_free_space_tree(struct btrfs_trans_handle *trans, ret = btrfs_search_slot_for_read(extent_root, &key, path, 1, 0); if (ret < 0) goto out_locked; - ASSERT(ret == 0); + /* + * If ret is 1 (no key found), it means this is an empty block group, + * without any extents allocated from it and there's no block group + * item (key BTRFS_BLOCK_GROUP_ITEM_KEY) located in the extent tree + * because we are using the block group tree feature, so block group + * items are stored in the block group tree. It also means there are no + * extents allocated for block groups with a start offset beyond this + * block group's end offset (this is the last, highest, block group). + */ + if (!btrfs_fs_compat_ro(trans->fs_info, BLOCK_GROUP_TREE)) + ASSERT(ret == 0); start = block_group->start; end = block_group->start + block_group->length; - while (1) { + while (ret == 0) { btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); if (key.type == BTRFS_EXTENT_ITEM_KEY || @@ -1149,8 +1159,6 @@ static int populate_free_space_tree(struct btrfs_trans_handle *trans, ret = btrfs_next_item(extent_root, path); if (ret < 0) goto out_locked; - if (ret) - break; } if (start < end) { ret = __add_to_free_space_tree(trans, block_group, path2, @@ -1233,6 +1241,7 @@ static int clear_free_space_tree(struct btrfs_trans_handle *trans, { BTRFS_PATH_AUTO_FREE(path); struct btrfs_key key; + struct rb_node *node; int nr; int ret; @@ -1261,6 +1270,16 @@ static int clear_free_space_tree(struct btrfs_trans_handle *trans, btrfs_release_path(path); } + node = rb_first_cached(&trans->fs_info->block_group_cache_tree); + while (node) { + struct btrfs_block_group *bg; + + bg = rb_entry(node, struct btrfs_block_group, cache_node); + clear_bit(BLOCK_GROUP_FLAG_FREE_SPACE_ADDED, &bg->runtime_flags); + node = rb_next(node); + cond_resched(); + } + return 0; } @@ -1350,12 +1369,18 @@ int btrfs_rebuild_free_space_tree(struct btrfs_fs_info *fs_info) block_group = rb_entry(node, struct btrfs_block_group, cache_node); + + if (test_bit(BLOCK_GROUP_FLAG_FREE_SPACE_ADDED, + &block_group->runtime_flags)) + goto next; + ret = populate_free_space_tree(trans, block_group); if (ret) { btrfs_abort_transaction(trans, ret); btrfs_end_transaction(trans); return ret; } +next: if (btrfs_should_end_transaction(trans)) { btrfs_end_transaction(trans); trans = btrfs_start_transaction(free_space_root, 1); @@ -1382,6 +1407,29 @@ static int __add_block_group_free_space(struct btrfs_trans_handle *trans, clear_bit(BLOCK_GROUP_FLAG_NEEDS_FREE_SPACE, &block_group->runtime_flags); + /* + * While rebuilding the free space tree we may allocate new metadata + * block groups while modifying the free space tree. + * + * Because during the rebuild (at btrfs_rebuild_free_space_tree()) we + * can use multiple transactions, every time btrfs_end_transaction() is + * called at btrfs_rebuild_free_space_tree() we finish the creation of + * new block groups by calling btrfs_create_pending_block_groups(), and + * that in turn calls us, through add_block_group_free_space(), to add + * a free space info item and a free space extent item for the block + * group. + * + * Then later btrfs_rebuild_free_space_tree() may find such new block + * groups and processes them with populate_free_space_tree(), which can + * fail with EEXIST since there are already items for the block group in + * the free space tree. Notice that we say "may find" because a new + * block group may be added to the block groups rbtree in a node before + * or after the block group currently being processed by the rebuild + * process. So signal the rebuild process to skip such new block groups + * if it finds them. + */ + set_bit(BLOCK_GROUP_FLAG_FREE_SPACE_ADDED, &block_group->runtime_flags); + ret = add_new_free_space_info(trans, block_group, path); if (ret) return ret; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index c0c778243bf1..fc66872b4c74 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4250,9 +4250,9 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans, ret = btrfs_del_inode_ref(trans, root, name, ino, dir_ino, &index); if (ret) { - btrfs_info(fs_info, - "failed to delete reference to %.*s, inode %llu parent %llu", - name->len, name->name, ino, dir_ino); + btrfs_crit(fs_info, + "failed to delete reference to %.*s, root %llu inode %llu parent %llu", + name->len, name->name, btrfs_root_id(root), ino, dir_ino); btrfs_abort_transaction(trans, ret); goto err; } @@ -4710,7 +4710,6 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info; int ret = 0; struct btrfs_trans_handle *trans; - u64 last_unlink_trans; struct fscrypt_name fname; if (inode->i_size > BTRFS_EMPTY_DIR_SIZE) @@ -4736,6 +4735,23 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) goto out_notrans; } + /* + * Propagate the last_unlink_trans value of the deleted dir to its + * parent directory. This is to prevent an unrecoverable log tree in the + * case we do something like this: + * 1) create dir foo + * 2) create snapshot under dir foo + * 3) delete the snapshot + * 4) rmdir foo + * 5) mkdir foo + * 6) fsync foo or some file inside foo + * + * This is because we can't unlink other roots when replaying the dir + * deletes for directory foo. + */ + if (BTRFS_I(inode)->last_unlink_trans >= trans->transid) + btrfs_record_snapshot_destroy(trans, BTRFS_I(dir)); + if (unlikely(btrfs_ino(BTRFS_I(inode)) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) { ret = btrfs_unlink_subvol(trans, BTRFS_I(dir), dentry); goto out; @@ -4745,27 +4761,11 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) if (ret) goto out; - last_unlink_trans = BTRFS_I(inode)->last_unlink_trans; - /* now the directory is empty */ ret = btrfs_unlink_inode(trans, BTRFS_I(dir), BTRFS_I(d_inode(dentry)), &fname.disk_name); - if (!ret) { + if (!ret) btrfs_i_size_write(BTRFS_I(inode), 0); - /* - * Propagate the last_unlink_trans value of the deleted dir to - * its parent directory. This is to prevent an unrecoverable - * log tree in the case we do something like this: - * 1) create dir foo - * 2) create snapshot under dir foo - * 3) delete the snapshot - * 4) rmdir foo - * 5) mkdir foo - * 6) fsync foo or some file inside foo - */ - if (last_unlink_trans >= trans->transid) - BTRFS_I(dir)->last_unlink_trans = last_unlink_trans; - } out: btrfs_end_transaction(trans); out_notrans: @@ -8059,6 +8059,7 @@ static int btrfs_rename_exchange(struct inode *old_dir, int ret; int ret2; bool need_abort = false; + bool logs_pinned = false; struct fscrypt_name old_fname, new_fname; struct fscrypt_str *old_name, *new_name; @@ -8182,6 +8183,31 @@ static int btrfs_rename_exchange(struct inode *old_dir, inode_inc_iversion(new_inode); simple_rename_timestamp(old_dir, old_dentry, new_dir, new_dentry); + if (old_ino != BTRFS_FIRST_FREE_OBJECTID && + new_ino != BTRFS_FIRST_FREE_OBJECTID) { + /* + * If we are renaming in the same directory (and it's not for + * root entries) pin the log early to prevent any concurrent + * task from logging the directory after we removed the old + * entries and before we add the new entries, otherwise that + * task can sync a log without any entry for the inodes we are + * renaming and therefore replaying that log, if a power failure + * happens after syncing the log, would result in deleting the + * inodes. + * + * If the rename affects two different directories, we want to + * make sure the that there's no log commit that contains + * updates for only one of the directories but not for the + * other. + * + * If we are renaming an entry for a root, we don't care about + * log updates since we called btrfs_set_log_full_commit(). + */ + btrfs_pin_log_trans(root); + btrfs_pin_log_trans(dest); + logs_pinned = true; + } + if (old_dentry->d_parent != new_dentry->d_parent) { btrfs_record_unlink_dir(trans, BTRFS_I(old_dir), BTRFS_I(old_inode), true); @@ -8253,30 +8279,23 @@ static int btrfs_rename_exchange(struct inode *old_dir, BTRFS_I(new_inode)->dir_index = new_idx; /* - * Now pin the logs of the roots. We do it to ensure that no other task - * can sync the logs while we are in progress with the rename, because - * that could result in an inconsistency in case any of the inodes that - * are part of this rename operation were logged before. + * Do the log updates for all inodes. + * + * If either entry is for a root we don't need to update the logs since + * we've called btrfs_set_log_full_commit() before. */ - if (old_ino != BTRFS_FIRST_FREE_OBJECTID) - btrfs_pin_log_trans(root); - if (new_ino != BTRFS_FIRST_FREE_OBJECTID) - btrfs_pin_log_trans(dest); - - /* Do the log updates for all inodes. */ - if (old_ino != BTRFS_FIRST_FREE_OBJECTID) + if (logs_pinned) { btrfs_log_new_name(trans, old_dentry, BTRFS_I(old_dir), old_rename_ctx.index, new_dentry->d_parent); - if (new_ino != BTRFS_FIRST_FREE_OBJECTID) btrfs_log_new_name(trans, new_dentry, BTRFS_I(new_dir), new_rename_ctx.index, old_dentry->d_parent); + } - /* Now unpin the logs. */ - if (old_ino != BTRFS_FIRST_FREE_OBJECTID) +out_fail: + if (logs_pinned) { btrfs_end_log_trans(root); - if (new_ino != BTRFS_FIRST_FREE_OBJECTID) btrfs_end_log_trans(dest); -out_fail: + } ret2 = btrfs_end_transaction(trans); ret = ret ? ret : ret2; out_notrans: @@ -8326,6 +8345,7 @@ static int btrfs_rename(struct mnt_idmap *idmap, int ret2; u64 old_ino = btrfs_ino(BTRFS_I(old_inode)); struct fscrypt_name old_fname, new_fname; + bool logs_pinned = false; if (btrfs_ino(BTRFS_I(new_dir)) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) return -EPERM; @@ -8460,6 +8480,29 @@ static int btrfs_rename(struct mnt_idmap *idmap, inode_inc_iversion(old_inode); simple_rename_timestamp(old_dir, old_dentry, new_dir, new_dentry); + if (old_ino != BTRFS_FIRST_FREE_OBJECTID) { + /* + * If we are renaming in the same directory (and it's not a + * root entry) pin the log to prevent any concurrent task from + * logging the directory after we removed the old entry and + * before we add the new entry, otherwise that task can sync + * a log without any entry for the inode we are renaming and + * therefore replaying that log, if a power failure happens + * after syncing the log, would result in deleting the inode. + * + * If the rename affects two different directories, we want to + * make sure the that there's no log commit that contains + * updates for only one of the directories but not for the + * other. + * + * If we are renaming an entry for a root, we don't care about + * log updates since we called btrfs_set_log_full_commit(). + */ + btrfs_pin_log_trans(root); + btrfs_pin_log_trans(dest); + logs_pinned = true; + } + if (old_dentry->d_parent != new_dentry->d_parent) btrfs_record_unlink_dir(trans, BTRFS_I(old_dir), BTRFS_I(old_inode), true); @@ -8524,7 +8567,7 @@ static int btrfs_rename(struct mnt_idmap *idmap, if (old_inode->i_nlink == 1) BTRFS_I(old_inode)->dir_index = index; - if (old_ino != BTRFS_FIRST_FREE_OBJECTID) + if (logs_pinned) btrfs_log_new_name(trans, old_dentry, BTRFS_I(old_dir), rename_ctx.index, new_dentry->d_parent); @@ -8540,6 +8583,10 @@ static int btrfs_rename(struct mnt_idmap *idmap, } } out_fail: + if (logs_pinned) { + btrfs_end_log_trans(root); + btrfs_end_log_trans(dest); + } ret2 = btrfs_end_transaction(trans); ret = ret ? ret : ret2; out_notrans: diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 913acef3f0a9..8a60983a697c 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -666,14 +666,14 @@ static noinline int create_subvol(struct mnt_idmap *idmap, goto out; } + btrfs_record_new_subvolume(trans, BTRFS_I(dir)); + ret = btrfs_create_new_inode(trans, &new_inode_args); if (ret) { btrfs_abort_transaction(trans, ret); goto out; } - btrfs_record_new_subvolume(trans, BTRFS_I(dir)); - d_instantiate_new(dentry, new_inode_args.inode); new_inode_args.inode = NULL; @@ -3139,7 +3139,7 @@ static long btrfs_ioctl_scrub(struct file *file, void __user *arg) return -EPERM; if (btrfs_fs_incompat(fs_info, EXTENT_TREE_V2)) { - btrfs_err(fs_info, "scrub is not supported on extent tree v2 yet"); + btrfs_err(fs_info, "scrub: extent tree v2 not yet supported"); return -EINVAL; } diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index ce36fafc771e..7cd5e76a783c 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -557,7 +557,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 num_bytes, */ for (i = 0; i < ipath->fspath->elem_cnt; ++i) btrfs_warn_in_rcu(fs_info, -"%s at logical %llu on dev %s, physical %llu, root %llu, inode %llu, offset %llu, length %u, links %u (path: %s)", +"scrub: %s at logical %llu on dev %s, physical %llu root %llu inode %llu offset %llu length %u links %u (path: %s)", swarn->errstr, swarn->logical, btrfs_dev_name(swarn->dev), swarn->physical, @@ -571,7 +571,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 num_bytes, err: btrfs_warn_in_rcu(fs_info, - "%s at logical %llu on dev %s, physical %llu, root %llu, inode %llu, offset %llu: path resolving failed with ret=%d", + "scrub: %s at logical %llu on dev %s, physical %llu root %llu inode %llu offset %llu: path resolving failed with ret=%d", swarn->errstr, swarn->logical, btrfs_dev_name(swarn->dev), swarn->physical, @@ -596,7 +596,7 @@ static void scrub_print_common_warning(const char *errstr, struct btrfs_device * /* Super block error, no need to search extent tree. */ if (is_super) { - btrfs_warn_in_rcu(fs_info, "%s on device %s, physical %llu", + btrfs_warn_in_rcu(fs_info, "scrub: %s on device %s, physical %llu", errstr, btrfs_dev_name(dev), physical); return; } @@ -631,14 +631,14 @@ static void scrub_print_common_warning(const char *errstr, struct btrfs_device * &ref_level); if (ret < 0) { btrfs_warn(fs_info, - "failed to resolve tree backref for logical %llu: %d", - swarn.logical, ret); + "scrub: failed to resolve tree backref for logical %llu: %d", + swarn.logical, ret); break; } if (ret > 0) break; btrfs_warn_in_rcu(fs_info, -"%s at logical %llu on dev %s, physical %llu: metadata %s (level %d) in tree %llu", +"scrub: %s at logical %llu on dev %s, physical %llu: metadata %s (level %d) in tree %llu", errstr, swarn.logical, btrfs_dev_name(dev), swarn.physical, (ref_level ? "node" : "leaf"), ref_level, ref_root); @@ -718,7 +718,7 @@ static void scrub_verify_one_metadata(struct scrub_stripe *stripe, int sector_nr scrub_bitmap_set_meta_error(stripe, sector_nr, sectors_per_tree); scrub_bitmap_set_error(stripe, sector_nr, sectors_per_tree); btrfs_warn_rl(fs_info, - "tree block %llu mirror %u has bad bytenr, has %llu want %llu", + "scrub: tree block %llu mirror %u has bad bytenr, has %llu want %llu", logical, stripe->mirror_num, btrfs_stack_header_bytenr(header), logical); return; @@ -728,7 +728,7 @@ static void scrub_verify_one_metadata(struct scrub_stripe *stripe, int sector_nr scrub_bitmap_set_meta_error(stripe, sector_nr, sectors_per_tree); scrub_bitmap_set_error(stripe, sector_nr, sectors_per_tree); btrfs_warn_rl(fs_info, - "tree block %llu mirror %u has bad fsid, has %pU want %pU", + "scrub: tree block %llu mirror %u has bad fsid, has %pU want %pU", logical, stripe->mirror_num, header->fsid, fs_info->fs_devices->fsid); return; @@ -738,7 +738,7 @@ static void scrub_verify_one_metadata(struct scrub_stripe *stripe, int sector_nr scrub_bitmap_set_meta_error(stripe, sector_nr, sectors_per_tree); scrub_bitmap_set_error(stripe, sector_nr, sectors_per_tree); btrfs_warn_rl(fs_info, - "tree block %llu mirror %u has bad chunk tree uuid, has %pU want %pU", + "scrub: tree block %llu mirror %u has bad chunk tree uuid, has %pU want %pU", logical, stripe->mirror_num, header->chunk_tree_uuid, fs_info->chunk_tree_uuid); return; @@ -760,7 +760,7 @@ static void scrub_verify_one_metadata(struct scrub_stripe *stripe, int sector_nr scrub_bitmap_set_meta_error(stripe, sector_nr, sectors_per_tree); scrub_bitmap_set_error(stripe, sector_nr, sectors_per_tree); btrfs_warn_rl(fs_info, - "tree block %llu mirror %u has bad csum, has " CSUM_FMT " want " CSUM_FMT, +"scrub: tree block %llu mirror %u has bad csum, has " CSUM_FMT " want " CSUM_FMT, logical, stripe->mirror_num, CSUM_FMT_VALUE(fs_info->csum_size, on_disk_csum), CSUM_FMT_VALUE(fs_info->csum_size, calculated_csum)); @@ -771,7 +771,7 @@ static void scrub_verify_one_metadata(struct scrub_stripe *stripe, int sector_nr scrub_bitmap_set_meta_gen_error(stripe, sector_nr, sectors_per_tree); scrub_bitmap_set_error(stripe, sector_nr, sectors_per_tree); btrfs_warn_rl(fs_info, - "tree block %llu mirror %u has bad generation, has %llu want %llu", + "scrub: tree block %llu mirror %u has bad generation, has %llu want %llu", logical, stripe->mirror_num, btrfs_stack_header_generation(header), stripe->sectors[sector_nr].generation); @@ -814,7 +814,7 @@ static void scrub_verify_one_sector(struct scrub_stripe *stripe, int sector_nr) */ if (unlikely(sector_nr + sectors_per_tree > stripe->nr_sectors)) { btrfs_warn_rl(fs_info, - "tree block at %llu crosses stripe boundary %llu", + "scrub: tree block at %llu crosses stripe boundary %llu", stripe->logical + (sector_nr << fs_info->sectorsize_bits), stripe->logical); @@ -1046,12 +1046,12 @@ skip: if (repaired) { if (dev) { btrfs_err_rl_in_rcu(fs_info, - "fixed up error at logical %llu on dev %s physical %llu", + "scrub: fixed up error at logical %llu on dev %s physical %llu", stripe->logical, btrfs_dev_name(dev), physical); } else { btrfs_err_rl_in_rcu(fs_info, - "fixed up error at logical %llu on mirror %u", + "scrub: fixed up error at logical %llu on mirror %u", stripe->logical, stripe->mirror_num); } continue; @@ -1060,12 +1060,12 @@ skip: /* The remaining are all for unrepaired. */ if (dev) { btrfs_err_rl_in_rcu(fs_info, - "unable to fixup (regular) error at logical %llu on dev %s physical %llu", +"scrub: unable to fixup (regular) error at logical %llu on dev %s physical %llu", stripe->logical, btrfs_dev_name(dev), physical); } else { btrfs_err_rl_in_rcu(fs_info, - "unable to fixup (regular) error at logical %llu on mirror %u", + "scrub: unable to fixup (regular) error at logical %llu on mirror %u", stripe->logical, stripe->mirror_num); } @@ -1593,8 +1593,7 @@ static int sync_write_pointer_for_zoned(struct scrub_ctx *sctx, u64 logical, physical, sctx->write_pointer); if (ret) - btrfs_err(fs_info, - "zoned: failed to recover write pointer"); + btrfs_err(fs_info, "scrub: zoned: failed to recover write pointer"); } mutex_unlock(&sctx->wr_lock); btrfs_dev_clear_zone_empty(sctx->wr_tgtdev, physical); @@ -1658,7 +1657,7 @@ static int scrub_find_fill_first_stripe(struct btrfs_block_group *bg, int ret; if (unlikely(!extent_root || !csum_root)) { - btrfs_err(fs_info, "no valid extent or csum root for scrub"); + btrfs_err(fs_info, "scrub: no valid extent or csum root found"); return -EUCLEAN; } memset(stripe->sectors, 0, sizeof(struct scrub_sector_verification) * @@ -1907,7 +1906,7 @@ static bool stripe_has_metadata_error(struct scrub_stripe *stripe) struct btrfs_fs_info *fs_info = stripe->bg->fs_info; btrfs_err(fs_info, - "stripe %llu has unrepaired metadata sector at %llu", + "scrub: stripe %llu has unrepaired metadata sector at logical %llu", stripe->logical, stripe->logical + (i << fs_info->sectorsize_bits)); return true; @@ -2167,7 +2166,7 @@ static int scrub_raid56_parity_stripe(struct scrub_ctx *sctx, bitmap_and(&error, &error, &has_extent, stripe->nr_sectors); if (!bitmap_empty(&error, stripe->nr_sectors)) { btrfs_err(fs_info, -"unrepaired sectors detected, full stripe %llu data stripe %u errors %*pbl", +"scrub: unrepaired sectors detected, full stripe %llu data stripe %u errors %*pbl", full_stripe_start, i, stripe->nr_sectors, &error); ret = -EIO; @@ -2789,14 +2788,14 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, ro_set = 0; } else if (ret == -ETXTBSY) { btrfs_warn(fs_info, - "skipping scrub of block group %llu due to active swapfile", + "scrub: skipping scrub of block group %llu due to active swapfile", cache->start); scrub_pause_off(fs_info); ret = 0; goto skip_unfreeze; } else { - btrfs_warn(fs_info, - "failed setting block group ro: %d", ret); + btrfs_warn(fs_info, "scrub: failed setting block group ro: %d", + ret); btrfs_unfreeze_block_group(cache); btrfs_put_block_group(cache); scrub_pause_off(fs_info); @@ -2892,13 +2891,13 @@ static int scrub_one_super(struct scrub_ctx *sctx, struct btrfs_device *dev, ret = btrfs_check_super_csum(fs_info, sb); if (ret != 0) { btrfs_err_rl(fs_info, - "super block at physical %llu devid %llu has bad csum", + "scrub: super block at physical %llu devid %llu has bad csum", physical, dev->devid); return -EIO; } if (btrfs_super_generation(sb) != generation) { btrfs_err_rl(fs_info, -"super block at physical %llu devid %llu has bad generation %llu expect %llu", +"scrub: super block at physical %llu devid %llu has bad generation %llu expect %llu", physical, dev->devid, btrfs_super_generation(sb), generation); return -EUCLEAN; @@ -3059,7 +3058,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start, !test_bit(BTRFS_DEV_STATE_WRITEABLE, &dev->dev_state)) { mutex_unlock(&fs_info->fs_devices->device_list_mutex); btrfs_err_in_rcu(fs_info, - "scrub on devid %llu: filesystem on %s is not writable", + "scrub: devid %llu: filesystem on %s is not writable", devid, btrfs_dev_name(dev)); ret = -EROFS; goto out; diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 97e933113b82..cea8a7e9d6d3 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -143,6 +143,9 @@ static struct btrfs_inode *btrfs_iget_logging(u64 objectid, struct btrfs_root *r unsigned int nofs_flag; struct btrfs_inode *inode; + /* Only meant to be called for subvolume roots and not for log roots. */ + ASSERT(is_fstree(btrfs_root_id(root))); + /* * We're holding a transaction handle whether we are logging or * replaying a log tree, so we must make sure NOFS semantics apply @@ -604,21 +607,6 @@ static int read_alloc_one_name(struct extent_buffer *eb, void *start, int len, return 0; } -/* - * simple helper to read an inode off the disk from a given root - * This can only be called for subvolume roots and not for the log - */ -static noinline struct btrfs_inode *read_one_inode(struct btrfs_root *root, - u64 objectid) -{ - struct btrfs_inode *inode; - - inode = btrfs_iget_logging(objectid, root); - if (IS_ERR(inode)) - return NULL; - return inode; -} - /* replays a single extent in 'eb' at 'slot' with 'key' into the * subvolume 'root'. path is released on entry and should be released * on exit. @@ -668,15 +656,15 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, extent_end = ALIGN(start + size, fs_info->sectorsize); } else { - ret = 0; - goto out; + btrfs_err(fs_info, + "unexpected extent type=%d root=%llu inode=%llu offset=%llu", + found_type, btrfs_root_id(root), key->objectid, key->offset); + return -EUCLEAN; } - inode = read_one_inode(root, key->objectid); - if (!inode) { - ret = -EIO; - goto out; - } + inode = btrfs_iget_logging(key->objectid, root); + if (IS_ERR(inode)) + return PTR_ERR(inode); /* * first check to see if we already have this extent in the @@ -948,9 +936,10 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, btrfs_release_path(path); - inode = read_one_inode(root, location.objectid); - if (!inode) { - ret = -EIO; + inode = btrfs_iget_logging(location.objectid, root); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); + inode = NULL; goto out; } @@ -961,7 +950,8 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, ret = unlink_inode_for_log_replay(trans, dir, inode, &name); out: kfree(name.name); - iput(&inode->vfs_inode); + if (inode) + iput(&inode->vfs_inode); return ret; } @@ -1072,7 +1062,9 @@ again: search_key.type = BTRFS_INODE_REF_KEY; search_key.offset = parent_objectid; ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0); - if (ret == 0) { + if (ret < 0) { + return ret; + } else if (ret == 0) { struct btrfs_inode_ref *victim_ref; unsigned long ptr; unsigned long ptr_end; @@ -1145,13 +1137,13 @@ again: struct fscrypt_str victim_name; extref = (struct btrfs_inode_extref *)(base + cur_offset); + victim_name.len = btrfs_inode_extref_name_len(leaf, extref); if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid) goto next; ret = read_alloc_one_name(leaf, &extref->name, - btrfs_inode_extref_name_len(leaf, extref), - &victim_name); + victim_name.len, &victim_name); if (ret) return ret; @@ -1166,18 +1158,18 @@ again: kfree(victim_name.name); return ret; } else if (!ret) { - ret = -ENOENT; - victim_parent = read_one_inode(root, - parent_objectid); - if (victim_parent) { + victim_parent = btrfs_iget_logging(parent_objectid, root); + if (IS_ERR(victim_parent)) { + ret = PTR_ERR(victim_parent); + } else { inc_nlink(&inode->vfs_inode); btrfs_release_path(path); ret = unlink_inode_for_log_replay(trans, victim_parent, inode, &victim_name); + iput(&victim_parent->vfs_inode); } - iput(&victim_parent->vfs_inode); kfree(victim_name.name); if (ret) return ret; @@ -1314,9 +1306,9 @@ again: struct btrfs_inode *dir; btrfs_release_path(path); - dir = read_one_inode(root, parent_id); - if (!dir) { - ret = -ENOENT; + dir = btrfs_iget_logging(parent_id, root); + if (IS_ERR(dir)) { + ret = PTR_ERR(dir); kfree(name.name); goto out; } @@ -1388,15 +1380,17 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, * copy the back ref in. The link count fixup code will take * care of the rest */ - dir = read_one_inode(root, parent_objectid); - if (!dir) { - ret = -ENOENT; + dir = btrfs_iget_logging(parent_objectid, root); + if (IS_ERR(dir)) { + ret = PTR_ERR(dir); + dir = NULL; goto out; } - inode = read_one_inode(root, inode_objectid); - if (!inode) { - ret = -EIO; + inode = btrfs_iget_logging(inode_objectid, root); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); + inode = NULL; goto out; } @@ -1408,11 +1402,13 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, * parent object can change from one array * item to another. */ - if (!dir) - dir = read_one_inode(root, parent_objectid); if (!dir) { - ret = -ENOENT; - goto out; + dir = btrfs_iget_logging(parent_objectid, root); + if (IS_ERR(dir)) { + ret = PTR_ERR(dir); + dir = NULL; + goto out; + } } } else { ret = ref_get_fields(eb, ref_ptr, &name, &ref_index); @@ -1681,9 +1677,9 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans, break; btrfs_release_path(path); - inode = read_one_inode(root, key.offset); - if (!inode) { - ret = -EIO; + inode = btrfs_iget_logging(key.offset, root); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); break; } @@ -1719,9 +1715,9 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans, struct btrfs_inode *inode; struct inode *vfs_inode; - inode = read_one_inode(root, objectid); - if (!inode) - return -EIO; + inode = btrfs_iget_logging(objectid, root); + if (IS_ERR(inode)) + return PTR_ERR(inode); vfs_inode = &inode->vfs_inode; key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID; @@ -1760,14 +1756,14 @@ static noinline int insert_one_name(struct btrfs_trans_handle *trans, struct btrfs_inode *dir; int ret; - inode = read_one_inode(root, location->objectid); - if (!inode) - return -ENOENT; + inode = btrfs_iget_logging(location->objectid, root); + if (IS_ERR(inode)) + return PTR_ERR(inode); - dir = read_one_inode(root, dirid); - if (!dir) { + dir = btrfs_iget_logging(dirid, root); + if (IS_ERR(dir)) { iput(&inode->vfs_inode); - return -EIO; + return PTR_ERR(dir); } ret = btrfs_add_link(trans, dir, inode, name, 1, index); @@ -1844,9 +1840,9 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, bool update_size = true; bool name_added = false; - dir = read_one_inode(root, key->objectid); - if (!dir) - return -EIO; + dir = btrfs_iget_logging(key->objectid, root); + if (IS_ERR(dir)) + return PTR_ERR(dir); ret = read_alloc_one_name(eb, di + 1, btrfs_dir_name_len(eb, di), &name); if (ret) @@ -2146,9 +2142,10 @@ static noinline int check_item_in_log(struct btrfs_trans_handle *trans, btrfs_dir_item_key_to_cpu(eb, di, &location); btrfs_release_path(path); btrfs_release_path(log_path); - inode = read_one_inode(root, location.objectid); - if (!inode) { - ret = -EIO; + inode = btrfs_iget_logging(location.objectid, root); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); + inode = NULL; goto out; } @@ -2300,14 +2297,17 @@ static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans, if (!log_path) return -ENOMEM; - dir = read_one_inode(root, dirid); - /* it isn't an error if the inode isn't there, that can happen - * because we replay the deletes before we copy in the inode item - * from the log + dir = btrfs_iget_logging(dirid, root); + /* + * It isn't an error if the inode isn't there, that can happen because + * we replay the deletes before we copy in the inode item from the log. */ - if (!dir) { + if (IS_ERR(dir)) { btrfs_free_path(log_path); - return 0; + ret = PTR_ERR(dir); + if (ret == -ENOENT) + ret = 0; + return ret; } range_start = 0; @@ -2466,9 +2466,9 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb, struct btrfs_inode *inode; u64 from; - inode = read_one_inode(root, key.objectid); - if (!inode) { - ret = -EIO; + inode = btrfs_iget_logging(key.objectid, root); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); break; } from = ALIGN(i_size_read(&inode->vfs_inode), @@ -7447,6 +7447,8 @@ void btrfs_record_snapshot_destroy(struct btrfs_trans_handle *trans, * full log sync. * Also we don't need to worry with renames, since btrfs_rename() marks the log * for full commit when renaming a subvolume. + * + * Must be called before creating the subvolume entry in its parent directory. */ void btrfs_record_new_subvolume(const struct btrfs_trans_handle *trans, struct btrfs_inode *dir) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 89835071cfea..f475b4b7c457 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3282,6 +3282,12 @@ int btrfs_remove_chunk(struct btrfs_trans_handle *trans, u64 chunk_offset) device->bytes_used - dev_extent_len); atomic64_add(dev_extent_len, &fs_info->free_chunk_space); btrfs_clear_space_info_full(fs_info); + + if (list_empty(&device->post_commit_list)) { + list_add_tail(&device->post_commit_list, + &trans->transaction->dev_update_list); + } + mutex_unlock(&fs_info->chunk_mutex); } } diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index b5b0156d5b95..9430b34d3cbb 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -1403,7 +1403,8 @@ static int btrfs_load_block_group_single(struct btrfs_block_group *bg, static int btrfs_load_block_group_dup(struct btrfs_block_group *bg, struct btrfs_chunk_map *map, struct zone_info *zone_info, - unsigned long *active) + unsigned long *active, + u64 last_alloc) { struct btrfs_fs_info *fs_info = bg->fs_info; @@ -1426,6 +1427,13 @@ static int btrfs_load_block_group_dup(struct btrfs_block_group *bg, zone_info[1].physical); return -EIO; } + + if (zone_info[0].alloc_offset == WP_CONVENTIONAL) + zone_info[0].alloc_offset = last_alloc; + + if (zone_info[1].alloc_offset == WP_CONVENTIONAL) + zone_info[1].alloc_offset = last_alloc; + if (zone_info[0].alloc_offset != zone_info[1].alloc_offset) { btrfs_err(bg->fs_info, "zoned: write pointer offset mismatch of zones in DUP profile"); @@ -1446,7 +1454,8 @@ static int btrfs_load_block_group_dup(struct btrfs_block_group *bg, static int btrfs_load_block_group_raid1(struct btrfs_block_group *bg, struct btrfs_chunk_map *map, struct zone_info *zone_info, - unsigned long *active) + unsigned long *active, + u64 last_alloc) { struct btrfs_fs_info *fs_info = bg->fs_info; int i; @@ -1461,10 +1470,12 @@ static int btrfs_load_block_group_raid1(struct btrfs_block_group *bg, bg->zone_capacity = min_not_zero(zone_info[0].capacity, zone_info[1].capacity); for (i = 0; i < map->num_stripes; i++) { - if (zone_info[i].alloc_offset == WP_MISSING_DEV || - zone_info[i].alloc_offset == WP_CONVENTIONAL) + if (zone_info[i].alloc_offset == WP_MISSING_DEV) continue; + if (zone_info[i].alloc_offset == WP_CONVENTIONAL) + zone_info[i].alloc_offset = last_alloc; + if ((zone_info[0].alloc_offset != zone_info[i].alloc_offset) && !btrfs_test_opt(fs_info, DEGRADED)) { btrfs_err(fs_info, @@ -1494,7 +1505,8 @@ static int btrfs_load_block_group_raid1(struct btrfs_block_group *bg, static int btrfs_load_block_group_raid0(struct btrfs_block_group *bg, struct btrfs_chunk_map *map, struct zone_info *zone_info, - unsigned long *active) + unsigned long *active, + u64 last_alloc) { struct btrfs_fs_info *fs_info = bg->fs_info; @@ -1505,10 +1517,29 @@ static int btrfs_load_block_group_raid0(struct btrfs_block_group *bg, } for (int i = 0; i < map->num_stripes; i++) { - if (zone_info[i].alloc_offset == WP_MISSING_DEV || - zone_info[i].alloc_offset == WP_CONVENTIONAL) + if (zone_info[i].alloc_offset == WP_MISSING_DEV) continue; + if (zone_info[i].alloc_offset == WP_CONVENTIONAL) { + u64 stripe_nr, full_stripe_nr; + u64 stripe_offset; + int stripe_index; + + stripe_nr = div64_u64(last_alloc, map->stripe_size); + stripe_offset = stripe_nr * map->stripe_size; + full_stripe_nr = div_u64(stripe_nr, map->num_stripes); + div_u64_rem(stripe_nr, map->num_stripes, &stripe_index); + + zone_info[i].alloc_offset = + full_stripe_nr * map->stripe_size; + + if (stripe_index > i) + zone_info[i].alloc_offset += map->stripe_size; + else if (stripe_index == i) + zone_info[i].alloc_offset += + (last_alloc - stripe_offset); + } + if (test_bit(0, active) != test_bit(i, active)) { if (!btrfs_zone_activate(bg)) return -EIO; @@ -1526,7 +1557,8 @@ static int btrfs_load_block_group_raid0(struct btrfs_block_group *bg, static int btrfs_load_block_group_raid10(struct btrfs_block_group *bg, struct btrfs_chunk_map *map, struct zone_info *zone_info, - unsigned long *active) + unsigned long *active, + u64 last_alloc) { struct btrfs_fs_info *fs_info = bg->fs_info; @@ -1537,8 +1569,7 @@ static int btrfs_load_block_group_raid10(struct btrfs_block_group *bg, } for (int i = 0; i < map->num_stripes; i++) { - if (zone_info[i].alloc_offset == WP_MISSING_DEV || - zone_info[i].alloc_offset == WP_CONVENTIONAL) + if (zone_info[i].alloc_offset == WP_MISSING_DEV) continue; if (test_bit(0, active) != test_bit(i, active)) { @@ -1549,6 +1580,29 @@ static int btrfs_load_block_group_raid10(struct btrfs_block_group *bg, set_bit(BLOCK_GROUP_FLAG_ZONE_IS_ACTIVE, &bg->runtime_flags); } + if (zone_info[i].alloc_offset == WP_CONVENTIONAL) { + u64 stripe_nr, full_stripe_nr; + u64 stripe_offset; + int stripe_index; + + stripe_nr = div64_u64(last_alloc, map->stripe_size); + stripe_offset = stripe_nr * map->stripe_size; + full_stripe_nr = div_u64(stripe_nr, + map->num_stripes / map->sub_stripes); + div_u64_rem(stripe_nr, + (map->num_stripes / map->sub_stripes), + &stripe_index); + + zone_info[i].alloc_offset = + full_stripe_nr * map->stripe_size; + + if (stripe_index > (i / map->sub_stripes)) + zone_info[i].alloc_offset += map->stripe_size; + else if (stripe_index == (i / map->sub_stripes)) + zone_info[i].alloc_offset += + (last_alloc - stripe_offset); + } + if ((i % map->sub_stripes) == 0) { bg->zone_capacity += zone_info[i].capacity; bg->alloc_offset += zone_info[i].alloc_offset; @@ -1637,18 +1691,22 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new) ret = btrfs_load_block_group_single(cache, &zone_info[0], active); break; case BTRFS_BLOCK_GROUP_DUP: - ret = btrfs_load_block_group_dup(cache, map, zone_info, active); + ret = btrfs_load_block_group_dup(cache, map, zone_info, active, + last_alloc); break; case BTRFS_BLOCK_GROUP_RAID1: case BTRFS_BLOCK_GROUP_RAID1C3: case BTRFS_BLOCK_GROUP_RAID1C4: - ret = btrfs_load_block_group_raid1(cache, map, zone_info, active); + ret = btrfs_load_block_group_raid1(cache, map, zone_info, + active, last_alloc); break; case BTRFS_BLOCK_GROUP_RAID0: - ret = btrfs_load_block_group_raid0(cache, map, zone_info, active); + ret = btrfs_load_block_group_raid0(cache, map, zone_info, + active, last_alloc); break; case BTRFS_BLOCK_GROUP_RAID10: - ret = btrfs_load_block_group_raid10(cache, map, zone_info, active); + ret = btrfs_load_block_group_raid10(cache, map, zone_info, + active, last_alloc); break; case BTRFS_BLOCK_GROUP_RAID5: case BTRFS_BLOCK_GROUP_RAID6: diff --git a/fs/cachefiles/io.c b/fs/cachefiles/io.c index c08e4a66ac07..3e0576d9db1d 100644 --- a/fs/cachefiles/io.c +++ b/fs/cachefiles/io.c @@ -347,8 +347,6 @@ int __cachefiles_write(struct cachefiles_object *object, default: ki->was_async = false; cachefiles_write_complete(&ki->iocb, ret); - if (ret > 0) - ret = 0; break; } diff --git a/fs/cachefiles/ondemand.c b/fs/cachefiles/ondemand.c index d9bc67176128..a7ed86fa98bb 100644 --- a/fs/cachefiles/ondemand.c +++ b/fs/cachefiles/ondemand.c @@ -83,10 +83,8 @@ static ssize_t cachefiles_ondemand_fd_write_iter(struct kiocb *kiocb, trace_cachefiles_ondemand_fd_write(object, file_inode(file), pos, len); ret = __cachefiles_write(object, file, pos, iter, NULL, NULL); - if (!ret) { - ret = len; + if (ret > 0) kiocb->ki_pos += ret; - } out: fput(file); diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index b95c4cb21c13..60a621b00c65 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -409,6 +409,15 @@ static void ceph_netfs_issue_read(struct netfs_io_subrequest *subreq) struct page **pages; size_t page_off; + /* + * FIXME: io_iter.count needs to be corrected to aligned + * length. Otherwise, iov_iter_get_pages_alloc2() operates + * with the initial unaligned length value. As a result, + * ceph_msg_data_cursor_init() triggers BUG_ON() in the case + * if msg->sparse_read_total > msg->data_length. + */ + subreq->io_iter.count = len; + err = iov_iter_get_pages_alloc2(&subreq->io_iter, &pages, len, &page_off); if (err < 0) { doutc(cl, "%llx.%llx failed to allocate pages, %d\n", diff --git a/fs/ceph/export.c b/fs/ceph/export.c index 150076ced937..b2f2af104679 100644 --- a/fs/ceph/export.c +++ b/fs/ceph/export.c @@ -33,12 +33,19 @@ struct ceph_nfs_snapfh { u32 hash; } __attribute__ ((packed)); +#define BYTES_PER_U32 (sizeof(u32)) +#define CEPH_FH_BASIC_SIZE \ + (sizeof(struct ceph_nfs_fh) / BYTES_PER_U32) +#define CEPH_FH_WITH_PARENT_SIZE \ + (sizeof(struct ceph_nfs_confh) / BYTES_PER_U32) +#define CEPH_FH_SNAPPED_INODE_SIZE \ + (sizeof(struct ceph_nfs_snapfh) / BYTES_PER_U32) + static int ceph_encode_snapfh(struct inode *inode, u32 *rawfh, int *max_len, struct inode *parent_inode) { struct ceph_client *cl = ceph_inode_to_client(inode); - static const int snap_handle_length = - sizeof(struct ceph_nfs_snapfh) >> 2; + static const int snap_handle_length = CEPH_FH_SNAPPED_INODE_SIZE; struct ceph_nfs_snapfh *sfh = (void *)rawfh; u64 snapid = ceph_snap(inode); int ret; @@ -88,10 +95,8 @@ static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, struct inode *parent_inode) { struct ceph_client *cl = ceph_inode_to_client(inode); - static const int handle_length = - sizeof(struct ceph_nfs_fh) >> 2; - static const int connected_handle_length = - sizeof(struct ceph_nfs_confh) >> 2; + static const int handle_length = CEPH_FH_BASIC_SIZE; + static const int connected_handle_length = CEPH_FH_WITH_PARENT_SIZE; int type; if (ceph_snap(inode) != CEPH_NOSNAP) @@ -308,7 +313,7 @@ static struct dentry *ceph_fh_to_dentry(struct super_block *sb, if (fh_type != FILEID_INO32_GEN && fh_type != FILEID_INO32_GEN_PARENT) return NULL; - if (fh_len < sizeof(*fh) / 4) + if (fh_len < sizeof(*fh) / BYTES_PER_U32) return NULL; doutc(fsc->client, "%llx\n", fh->ino); @@ -427,7 +432,7 @@ static struct dentry *ceph_fh_to_parent(struct super_block *sb, if (fh_type != FILEID_INO32_GEN_PARENT) return NULL; - if (fh_len < sizeof(*cfh) / 4) + if (fh_len < sizeof(*cfh) / BYTES_PER_U32) return NULL; doutc(fsc->client, "%llx\n", cfh->parent_ino); diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 851d70200c6b..a7254cab44cc 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -2616,7 +2616,7 @@ static int ceph_zero_objects(struct inode *inode, loff_t offset, loff_t length) s32 stripe_unit = ci->i_layout.stripe_unit; s32 stripe_count = ci->i_layout.stripe_count; s32 object_size = ci->i_layout.object_size; - u64 object_set_size = object_size * stripe_count; + u64 object_set_size = (u64) object_size * stripe_count; u64 nearly, t; /* round offset up to next period boundary */ diff --git a/fs/ceph/super.c b/fs/ceph/super.c index f3951253e393..2b8438d8a324 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -1033,8 +1033,7 @@ void ceph_umount_begin(struct super_block *sb) struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb); doutc(fsc->client, "starting forced umount\n"); - if (!fsc) - return; + fsc->mount_state = CEPH_MOUNT_SHUTDOWN; __ceph_umount_begin(fsc); } @@ -1227,6 +1226,7 @@ static int ceph_set_super(struct super_block *s, struct fs_context *fc) s->s_time_min = 0; s->s_time_max = U32_MAX; s->s_flags |= SB_NODIRATIME | SB_NOATIME; + s->s_magic = CEPH_SUPER_MAGIC; ceph_fscrypt_set_ops(s); diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index e01d5f29f4d2..6dd3a524cd35 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -509,7 +509,7 @@ static void add_scan(struct dlm_ls *ls, struct dlm_rsb *r) void dlm_rsb_scan(struct timer_list *timer) { - struct dlm_ls *ls = from_timer(ls, timer, ls_scan_timer); + struct dlm_ls *ls = timer_container_of(ls, timer, ls_scan_timer); int our_nodeid = dlm_our_nodeid(); struct dlm_rsb *r; int rv; diff --git a/fs/efivarfs/super.c b/fs/efivarfs/super.c index c900d98bf494..284d6dbba2ec 100644 --- a/fs/efivarfs/super.c +++ b/fs/efivarfs/super.c @@ -390,10 +390,16 @@ static int efivarfs_reconfigure(struct fs_context *fc) return 0; } +static void efivarfs_free(struct fs_context *fc) +{ + kfree(fc->s_fs_info); +} + static const struct fs_context_operations efivarfs_context_ops = { .get_tree = efivarfs_get_tree, .parse_param = efivarfs_parse_param, .reconfigure = efivarfs_reconfigure, + .free = efivarfs_free, }; static int efivarfs_check_missing(efi_char16_t *name16, efi_guid_t vendor, diff --git a/fs/erofs/data.c b/fs/erofs/data.c index 6a329c329f43..16e4a6bd9b97 100644 --- a/fs/erofs/data.c +++ b/fs/erofs/data.c @@ -214,9 +214,11 @@ int erofs_map_dev(struct super_block *sb, struct erofs_map_dev *map) /* * bit 30: I/O error occurred on this folio + * bit 29: CPU has dirty data in D-cache (needs aliasing handling); * bit 0 - 29: remaining parts to complete this folio */ -#define EROFS_ONLINEFOLIO_EIO (1 << 30) +#define EROFS_ONLINEFOLIO_EIO 30 +#define EROFS_ONLINEFOLIO_DIRTY 29 void erofs_onlinefolio_init(struct folio *folio) { @@ -233,19 +235,23 @@ void erofs_onlinefolio_split(struct folio *folio) atomic_inc((atomic_t *)&folio->private); } -void erofs_onlinefolio_end(struct folio *folio, int err) +void erofs_onlinefolio_end(struct folio *folio, int err, bool dirty) { int orig, v; do { orig = atomic_read((atomic_t *)&folio->private); - v = (orig - 1) | (err ? EROFS_ONLINEFOLIO_EIO : 0); + DBG_BUGON(orig <= 0); + v = dirty << EROFS_ONLINEFOLIO_DIRTY; + v |= (orig - 1) | (!!err << EROFS_ONLINEFOLIO_EIO); } while (atomic_cmpxchg((atomic_t *)&folio->private, orig, v) != orig); - if (v & ~EROFS_ONLINEFOLIO_EIO) + if (v & (BIT(EROFS_ONLINEFOLIO_DIRTY) - 1)) return; folio->private = 0; - folio_end_read(folio, !(v & EROFS_ONLINEFOLIO_EIO)); + if (v & BIT(EROFS_ONLINEFOLIO_DIRTY)) + flush_dcache_folio(folio); + folio_end_read(folio, !(v & BIT(EROFS_ONLINEFOLIO_EIO))); } static int erofs_iomap_begin(struct inode *inode, loff_t offset, loff_t length, @@ -351,11 +357,16 @@ int erofs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, */ static int erofs_read_folio(struct file *file, struct folio *folio) { + trace_erofs_read_folio(folio, true); + return iomap_read_folio(folio, &erofs_iomap_ops); } static void erofs_readahead(struct readahead_control *rac) { + trace_erofs_readahead(rac->mapping->host, readahead_index(rac), + readahead_count(rac), true); + return iomap_readahead(rac, &erofs_iomap_ops); } diff --git a/fs/erofs/decompressor.c b/fs/erofs/decompressor.c index bf62e2836b60..358061d7b660 100644 --- a/fs/erofs/decompressor.c +++ b/fs/erofs/decompressor.c @@ -301,13 +301,11 @@ static int z_erofs_transform_plain(struct z_erofs_decompress_req *rq, cur = min(cur, rq->outputsize); if (cur && rq->out[0]) { kin = kmap_local_page(rq->in[nrpages_in - 1]); - if (rq->out[0] == rq->in[nrpages_in - 1]) { + if (rq->out[0] == rq->in[nrpages_in - 1]) memmove(kin + rq->pageofs_out, kin + pi, cur); - flush_dcache_page(rq->out[0]); - } else { + else memcpy_to_page(rq->out[0], rq->pageofs_out, kin + pi, cur); - } kunmap_local(kin); } rq->outputsize -= cur; @@ -325,14 +323,12 @@ static int z_erofs_transform_plain(struct z_erofs_decompress_req *rq, po = (rq->pageofs_out + cur + pi) & ~PAGE_MASK; DBG_BUGON(no >= nrpages_out); cnt = min(insz - pi, PAGE_SIZE - po); - if (rq->out[no] == rq->in[ni]) { + if (rq->out[no] == rq->in[ni]) memmove(kin + po, kin + rq->pageofs_in + pi, cnt); - flush_dcache_page(rq->out[no]); - } else if (rq->out[no]) { + else if (rq->out[no]) memcpy_to_page(rq->out[no], po, kin + rq->pageofs_in + pi, cnt); - } pi += cnt; } while (pi < insz); kunmap_local(kin); diff --git a/fs/erofs/dir.c b/fs/erofs/dir.c index 2fae209d0274..3e4b38bec0aa 100644 --- a/fs/erofs/dir.c +++ b/fs/erofs/dir.c @@ -58,6 +58,11 @@ static int erofs_readdir(struct file *f, struct dir_context *ctx) struct erofs_dirent *de; unsigned int nameoff, maxsize; + if (fatal_signal_pending(current)) { + err = -ERESTARTSYS; + break; + } + de = erofs_bread(&buf, dbstart, true); if (IS_ERR(de)) { erofs_err(sb, "failed to readdir of logical block %llu of nid %llu", @@ -88,6 +93,7 @@ static int erofs_readdir(struct file *f, struct dir_context *ctx) break; ctx->pos = dbstart + maxsize; ofs = 0; + cond_resched(); } erofs_put_metabuf(&buf); if (EROFS_I(dir)->dot_omitted && ctx->pos == dir->i_size) { diff --git a/fs/erofs/fileio.c b/fs/erofs/fileio.c index 7d81f504bff0..91781718199e 100644 --- a/fs/erofs/fileio.c +++ b/fs/erofs/fileio.c @@ -38,7 +38,7 @@ static void erofs_fileio_ki_complete(struct kiocb *iocb, long ret) } else { bio_for_each_folio_all(fi, &rq->bio) { DBG_BUGON(folio_test_uptodate(fi.folio)); - erofs_onlinefolio_end(fi.folio, ret); + erofs_onlinefolio_end(fi.folio, ret, false); } } bio_uninit(&rq->bio); @@ -47,6 +47,7 @@ static void erofs_fileio_ki_complete(struct kiocb *iocb, long ret) static void erofs_fileio_rq_submit(struct erofs_fileio_rq *rq) { + const struct cred *old_cred; struct iov_iter iter; int ret; @@ -60,7 +61,9 @@ static void erofs_fileio_rq_submit(struct erofs_fileio_rq *rq) rq->iocb.ki_flags = IOCB_DIRECT; iov_iter_bvec(&iter, ITER_DEST, rq->bvecs, rq->bio.bi_vcnt, rq->bio.bi_iter.bi_size); + old_cred = override_creds(rq->iocb.ki_filp->f_cred); ret = vfs_iocb_iter_read(rq->iocb.ki_filp, &rq->iocb, &iter); + revert_creds(old_cred); if (ret != -EIOCBQUEUED) erofs_fileio_ki_complete(&rq->iocb, ret); } @@ -93,8 +96,6 @@ static int erofs_fileio_scan_folio(struct erofs_fileio *io, struct folio *folio) struct erofs_map_blocks *map = &io->map; unsigned int cur = 0, end = folio_size(folio), len, attached = 0; loff_t pos = folio_pos(folio), ofs; - struct iov_iter iter; - struct bio_vec bv; int err = 0; erofs_onlinefolio_init(folio); @@ -119,13 +120,7 @@ static int erofs_fileio_scan_folio(struct erofs_fileio *io, struct folio *folio) err = PTR_ERR(src); break; } - bvec_set_folio(&bv, folio, len, cur); - iov_iter_bvec(&iter, ITER_DEST, &bv, 1, len); - if (copy_to_iter(src, len, &iter) != len) { - erofs_put_metabuf(&buf); - err = -EIO; - break; - } + memcpy_to_folio(folio, cur, src, len); erofs_put_metabuf(&buf); } else if (!(map->m_flags & EROFS_MAP_MAPPED)) { folio_zero_segment(folio, cur, cur + len); @@ -159,7 +154,7 @@ io_retry: } cur += len; } - erofs_onlinefolio_end(folio, err); + erofs_onlinefolio_end(folio, err, false); return err; } diff --git a/fs/erofs/internal.h b/fs/erofs/internal.h index a32c03a80c70..06b867d2fc3b 100644 --- a/fs/erofs/internal.h +++ b/fs/erofs/internal.h @@ -315,10 +315,12 @@ static inline struct folio *erofs_grab_folio_nowait(struct address_space *as, /* The length of extent is full */ #define EROFS_MAP_FULL_MAPPED 0x0008 /* Located in the special packed inode */ -#define EROFS_MAP_FRAGMENT 0x0010 +#define __EROFS_MAP_FRAGMENT 0x0010 /* The extent refers to partial decompressed data */ #define EROFS_MAP_PARTIAL_REF 0x0020 +#define EROFS_MAP_FRAGMENT (EROFS_MAP_MAPPED | __EROFS_MAP_FRAGMENT) + struct erofs_map_blocks { struct erofs_buf buf; @@ -390,7 +392,7 @@ int erofs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, int erofs_map_blocks(struct inode *inode, struct erofs_map_blocks *map); void erofs_onlinefolio_init(struct folio *folio); void erofs_onlinefolio_split(struct folio *folio); -void erofs_onlinefolio_end(struct folio *folio, int err); +void erofs_onlinefolio_end(struct folio *folio, int err, bool dirty); struct inode *erofs_iget(struct super_block *sb, erofs_nid_t nid); int erofs_getattr(struct mnt_idmap *idmap, const struct path *path, struct kstat *stat, u32 request_mask, diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c index fe8071844724..e3f28a1bb945 100644 --- a/fs/erofs/zdata.c +++ b/fs/erofs/zdata.c @@ -1034,7 +1034,7 @@ static int z_erofs_scan_folio(struct z_erofs_frontend *f, if (!(map->m_flags & EROFS_MAP_MAPPED)) { folio_zero_segment(folio, cur, end); tight = false; - } else if (map->m_flags & EROFS_MAP_FRAGMENT) { + } else if (map->m_flags & __EROFS_MAP_FRAGMENT) { erofs_off_t fpos = offset + cur - map->m_la; err = z_erofs_read_fragment(inode->i_sb, folio, cur, @@ -1091,7 +1091,7 @@ static int z_erofs_scan_folio(struct z_erofs_frontend *f, tight = (bs == PAGE_SIZE); } } while ((end = cur) > 0); - erofs_onlinefolio_end(folio, err); + erofs_onlinefolio_end(folio, err, false); return err; } @@ -1196,7 +1196,7 @@ static void z_erofs_fill_other_copies(struct z_erofs_backend *be, int err) cur += len; } kunmap_local(dst); - erofs_onlinefolio_end(page_folio(bvi->bvec.page), err); + erofs_onlinefolio_end(page_folio(bvi->bvec.page), err, true); list_del(p); kfree(bvi); } @@ -1355,7 +1355,7 @@ static int z_erofs_decompress_pcluster(struct z_erofs_backend *be, int err) DBG_BUGON(z_erofs_page_is_invalidated(page)); if (!z_erofs_is_shortlived_page(page)) { - erofs_onlinefolio_end(page_folio(page), err); + erofs_onlinefolio_end(page_folio(page), err, true); continue; } if (pcl->algorithmformat != Z_EROFS_COMPRESSION_LZ4) { diff --git a/fs/erofs/zmap.c b/fs/erofs/zmap.c index 14ea47f954f5..f1a15ff22147 100644 --- a/fs/erofs/zmap.c +++ b/fs/erofs/zmap.c @@ -413,8 +413,7 @@ static int z_erofs_map_blocks_fo(struct inode *inode, !vi->z_tailextent_headlcn) { map->m_la = 0; map->m_llen = inode->i_size; - map->m_flags = EROFS_MAP_MAPPED | - EROFS_MAP_FULL_MAPPED | EROFS_MAP_FRAGMENT; + map->m_flags = EROFS_MAP_FRAGMENT; return 0; } initial_lcn = ofs >> lclusterbits; @@ -489,7 +488,7 @@ static int z_erofs_map_blocks_fo(struct inode *inode, goto unmap_out; } } else if (fragment && m.lcn == vi->z_tailextent_headlcn) { - map->m_flags |= EROFS_MAP_FRAGMENT; + map->m_flags = EROFS_MAP_FRAGMENT; } else { map->m_pa = erofs_pos(sb, m.pblk); err = z_erofs_get_extent_compressedlen(&m, initial_lcn); @@ -597,6 +596,10 @@ static int z_erofs_map_blocks_ext(struct inode *inode, if (la > map->m_la) { r = mid; + if (la > lend) { + DBG_BUGON(1); + return -EFSCORRUPTED; + } lend = la; } else { l = mid + 1; @@ -613,7 +616,7 @@ static int z_erofs_map_blocks_ext(struct inode *inode, if (lstart < lend) { map->m_la = lstart; if (last && (vi->z_advise & Z_EROFS_ADVISE_FRAGMENT_PCLUSTER)) { - map->m_flags |= EROFS_MAP_MAPPED | EROFS_MAP_FRAGMENT; + map->m_flags = EROFS_MAP_FRAGMENT; vi->z_fragmentoff = map->m_plen; if (recsz > offsetof(struct z_erofs_extent, pstart_lo)) vi->z_fragmentoff |= map->m_pa << 32; @@ -635,12 +638,6 @@ static int z_erofs_map_blocks_ext(struct inode *inode, } } map->m_llen = lend - map->m_la; - if (!last && map->m_llen < sb->s_blocksize) { - erofs_err(sb, "extent too small %llu @ offset %llu of nid %llu", - map->m_llen, map->m_la, vi->nid); - DBG_BUGON(1); - return -EFSCORRUPTED; - } return 0; } @@ -799,7 +796,7 @@ static int z_erofs_iomap_begin_report(struct inode *inode, loff_t offset, iomap->length = map.m_llen; if (map.m_flags & EROFS_MAP_MAPPED) { iomap->type = IOMAP_MAPPED; - iomap->addr = map.m_flags & EROFS_MAP_FRAGMENT ? + iomap->addr = map.m_flags & __EROFS_MAP_FRAGMENT ? IOMAP_NULL_ADDR : map.m_pa; } else { iomap->type = IOMAP_HOLE; diff --git a/fs/eventpoll.c b/fs/eventpoll.c index d4dbffdedd08..0fbf5dfedb24 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -883,7 +883,7 @@ static bool __ep_remove(struct eventpoll *ep, struct epitem *epi, bool force) kfree_rcu(epi, rcu); percpu_counter_dec(&ep->user->epoll_watches); - return ep_refcount_dec_and_test(ep); + return true; } /* @@ -891,14 +891,14 @@ static bool __ep_remove(struct eventpoll *ep, struct epitem *epi, bool force) */ static void ep_remove_safe(struct eventpoll *ep, struct epitem *epi) { - WARN_ON_ONCE(__ep_remove(ep, epi, false)); + if (__ep_remove(ep, epi, false)) + WARN_ON_ONCE(ep_refcount_dec_and_test(ep)); } static void ep_clear_and_put(struct eventpoll *ep) { struct rb_node *rbp, *next; struct epitem *epi; - bool dispose; /* We need to release all tasks waiting for these file */ if (waitqueue_active(&ep->poll_wait)) @@ -931,10 +931,8 @@ static void ep_clear_and_put(struct eventpoll *ep) cond_resched(); } - dispose = ep_refcount_dec_and_test(ep); mutex_unlock(&ep->mtx); - - if (dispose) + if (ep_refcount_dec_and_test(ep)) ep_free(ep); } @@ -1137,7 +1135,7 @@ again: dispose = __ep_remove(ep, epi, true); mutex_unlock(&ep->mtx); - if (dispose) + if (dispose && ep_refcount_dec_and_test(ep)) ep_free(ep); goto again; } diff --git a/fs/exec.c b/fs/exec.c index 1f5fdd2e096e..ba400aafd640 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -114,6 +114,9 @@ static inline void put_binfmt(struct linux_binfmt * fmt) bool path_noexec(const struct path *path) { + /* If it's an anonymous inode make sure that we catch any shenanigans. */ + VFS_WARN_ON_ONCE(IS_ANON_FILE(d_inode(path->dentry)) && + !(path->mnt->mnt_sb->s_iflags & SB_I_NOEXEC)); return (path->mnt->mnt_flags & MNT_NOEXEC) || (path->mnt->mnt_sb->s_iflags & SB_I_NOEXEC); } @@ -781,13 +784,15 @@ static struct file *do_open_execat(int fd, struct filename *name, int flags) if (IS_ERR(file)) return file; + if (path_noexec(&file->f_path)) + return ERR_PTR(-EACCES); + /* * In the past the regular type check was here. It moved to may_open() in * 633fb6ac3980 ("exec: move S_ISREG() check earlier"). Since then it is * an invariant that all non-regular files error out before we get here. */ - if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode)) || - path_noexec(&file->f_path)) + if (WARN_ON_ONCE(!S_ISREG(file_inode(file)->i_mode))) return ERR_PTR(-EACCES); err = exe_file_deny_write_access(file); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index a7f80ca01174..c7d39da7e733 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3627,7 +3627,7 @@ int ext4_feature_set_ok(struct super_block *sb, int readonly) */ static void print_daily_error_info(struct timer_list *t) { - struct ext4_sb_info *sbi = from_timer(sbi, t, s_err_report); + struct ext4_sb_info *sbi = timer_container_of(sbi, t, s_err_report); struct super_block *sb = sbi->s_sb; struct ext4_super_block *es = sbi->s_es; diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 6bd3de64f2a8..696131e655ed 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -35,6 +35,17 @@ #include <trace/events/f2fs.h> #include <uapi/linux/f2fs.h> +static void f2fs_zero_post_eof_page(struct inode *inode, loff_t new_size) +{ + loff_t old_size = i_size_read(inode); + + if (old_size >= new_size) + return; + + /* zero or drop pages only in range of [old_size, new_size] */ + truncate_pagecache(inode, old_size); +} + static vm_fault_t f2fs_filemap_fault(struct vm_fault *vmf) { struct inode *inode = file_inode(vmf->vma->vm_file); @@ -103,8 +114,13 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault *vmf) f2fs_bug_on(sbi, f2fs_has_inline_data(inode)); + filemap_invalidate_lock(inode->i_mapping); + f2fs_zero_post_eof_page(inode, (folio->index + 1) << PAGE_SHIFT); + filemap_invalidate_unlock(inode->i_mapping); + file_update_time(vmf->vma->vm_file); filemap_invalidate_lock_shared(inode->i_mapping); + folio_lock(folio); if (unlikely(folio->mapping != inode->i_mapping || folio_pos(folio) > i_size_read(inode) || @@ -1109,6 +1125,8 @@ int f2fs_setattr(struct mnt_idmap *idmap, struct dentry *dentry, f2fs_down_write(&fi->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); + if (attr->ia_size > old_size) + f2fs_zero_post_eof_page(inode, attr->ia_size); truncate_setsize(inode, attr->ia_size); if (attr->ia_size <= old_size) @@ -1227,6 +1245,10 @@ static int f2fs_punch_hole(struct inode *inode, loff_t offset, loff_t len) if (ret) return ret; + filemap_invalidate_lock(inode->i_mapping); + f2fs_zero_post_eof_page(inode, offset + len); + filemap_invalidate_unlock(inode->i_mapping); + pg_start = ((unsigned long long) offset) >> PAGE_SHIFT; pg_end = ((unsigned long long) offset + len) >> PAGE_SHIFT; @@ -1510,6 +1532,8 @@ static int f2fs_do_collapse(struct inode *inode, loff_t offset, loff_t len) f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); + f2fs_zero_post_eof_page(inode, offset + len); + f2fs_lock_op(sbi); f2fs_drop_extent_tree(inode); truncate_pagecache(inode, offset); @@ -1631,6 +1655,10 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len, if (ret) return ret; + filemap_invalidate_lock(mapping); + f2fs_zero_post_eof_page(inode, offset + len); + filemap_invalidate_unlock(mapping); + pg_start = ((unsigned long long) offset) >> PAGE_SHIFT; pg_end = ((unsigned long long) offset + len) >> PAGE_SHIFT; @@ -1762,6 +1790,8 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len) /* avoid gc operation during block exchange */ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(mapping); + + f2fs_zero_post_eof_page(inode, offset + len); truncate_pagecache(inode, offset); while (!ret && idx > pg_start) { @@ -1819,6 +1849,10 @@ static int f2fs_expand_inode_data(struct inode *inode, loff_t offset, if (err) return err; + filemap_invalidate_lock(inode->i_mapping); + f2fs_zero_post_eof_page(inode, offset + len); + filemap_invalidate_unlock(inode->i_mapping); + f2fs_balance_fs(sbi, true); pg_start = ((unsigned long long)offset) >> PAGE_SHIFT; @@ -4860,6 +4894,10 @@ static ssize_t f2fs_write_checks(struct kiocb *iocb, struct iov_iter *from) err = file_modified(file); if (err) return err; + + filemap_invalidate_lock(inode->i_mapping); + f2fs_zero_post_eof_page(inode, iocb->ki_pos + iov_iter_count(from)); + filemap_invalidate_unlock(inode->i_mapping); return count; } diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 1cb4cba7f961..bfe104db284e 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -2078,7 +2078,6 @@ write_node: if (!__write_node_folio(folio, false, &submitted, wbc, do_balance, io_type, NULL)) { - folio_unlock(folio); folio_batch_release(&fbatch); ret = -EIO; goto out; diff --git a/fs/file.c b/fs/file.c index 3a3146664cf3..b6db031545e6 100644 --- a/fs/file.c +++ b/fs/file.c @@ -1198,8 +1198,12 @@ bool file_seek_cur_needs_f_lock(struct file *file) if (!(file->f_mode & FMODE_ATOMIC_POS) && !file->f_op->iterate_shared) return false; - VFS_WARN_ON_ONCE((file_count(file) > 1) && - !mutex_is_locked(&file->f_pos_lock)); + /* + * Note that we are not guaranteed to be called after fdget_pos() on + * this file obj, in which case the caller is expected to provide the + * appropriate locking. + */ + return true; } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index f102afc03359..47006d0753f1 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1147,7 +1147,7 @@ static ssize_t fuse_send_write_pages(struct fuse_io_args *ia, static ssize_t fuse_fill_write_pages(struct fuse_io_args *ia, struct address_space *mapping, struct iov_iter *ii, loff_t pos, - unsigned int max_pages) + unsigned int max_folios) { struct fuse_args_pages *ap = &ia->ap; struct fuse_conn *fc = get_fuse_conn(mapping->host); @@ -1157,12 +1157,11 @@ static ssize_t fuse_fill_write_pages(struct fuse_io_args *ia, int err = 0; num = min(iov_iter_count(ii), fc->max_write); - num = min(num, max_pages << PAGE_SHIFT); ap->args.in_pages = true; ap->descs[0].offset = offset; - while (num) { + while (num && ap->num_folios < max_folios) { size_t tmp; struct folio *folio; pgoff_t index = pos >> PAGE_SHIFT; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index bfe8d8af46f3..9572bdef49ee 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -9,6 +9,7 @@ #include "fuse_i.h" #include "dev_uring_i.h" +#include <linux/dax.h> #include <linux/pagemap.h> #include <linux/slab.h> #include <linux/file.h> @@ -162,6 +163,9 @@ static void fuse_evict_inode(struct inode *inode) /* Will write inode on close/munmap and in all other dirtiers */ WARN_ON(inode->i_state & I_DIRTY_INODE); + if (FUSE_IS_DAX(inode)) + dax_break_layout_final(inode); + truncate_inode_pages_final(&inode->i_data); clear_inode(inode); if (inode->i_sb->s_flags & SB_ACTIVE) { diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c index 3729391a18f3..fb4519158f3a 100644 --- a/fs/iomap/buffered-io.c +++ b/fs/iomap/buffered-io.c @@ -71,6 +71,9 @@ static void iomap_set_range_uptodate(struct folio *folio, size_t off, unsigned long flags; bool uptodate = true; + if (folio_test_uptodate(folio)) + return; + if (ifs) { spin_lock_irqsave(&ifs->state_lock, flags); uptodate = ifs_set_range_uptodate(folio, ifs, off, len); diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index d5da9817df9b..33e6a620c103 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -1440,9 +1440,16 @@ static int isofs_read_inode(struct inode *inode, int relocated) inode->i_op = &page_symlink_inode_operations; inode_nohighmem(inode); inode->i_data.a_ops = &isofs_symlink_aops; - } else + } else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode) || + S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) { /* XXX - parse_rock_ridge_inode() had already set i_rdev. */ init_special_inode(inode, inode->i_mode, inode->i_rdev); + } else { + printk(KERN_DEBUG "ISOFS: Invalid file type 0%04o for inode %lu.\n", + inode->i_mode, inode->i_ino); + ret = -EIO; + goto fail; + } ret = 0; out: diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 6d5e76848733..d480b94117cd 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -134,7 +134,7 @@ static __be32 jbd2_superblock_csum(journal_superblock_t *sb) static void commit_timeout(struct timer_list *t) { - journal_t *journal = from_timer(journal, t, j_commit_timer); + journal_t *journal = timer_container_of(journal, t, j_commit_timer); wake_up_process(journal->j_task); } diff --git a/fs/jffs2/erase.c b/fs/jffs2/erase.c index ef3a1e1b6cb0..fda9f4d6093f 100644 --- a/fs/jffs2/erase.c +++ b/fs/jffs2/erase.c @@ -425,7 +425,9 @@ static void jffs2_mark_erased_block(struct jffs2_sb_info *c, struct jffs2_eraseb .totlen = cpu_to_je32(c->cleanmarker_size) }; - jffs2_prealloc_raw_node_refs(c, jeb, 1); + ret = jffs2_prealloc_raw_node_refs(c, jeb, 1); + if (ret) + goto filebad; marker.hdr_crc = cpu_to_je32(crc32(0, &marker, sizeof(struct jffs2_unknown_node)-4)); diff --git a/fs/jffs2/scan.c b/fs/jffs2/scan.c index 29671e33a171..62879c218d4b 100644 --- a/fs/jffs2/scan.c +++ b/fs/jffs2/scan.c @@ -256,7 +256,9 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) jffs2_dbg(1, "%s(): Skipping %d bytes in nextblock to ensure page alignment\n", __func__, skip); - jffs2_prealloc_raw_node_refs(c, c->nextblock, 1); + ret = jffs2_prealloc_raw_node_refs(c, c->nextblock, 1); + if (ret) + goto out; jffs2_scan_dirty_space(c, c->nextblock, skip); } #endif diff --git a/fs/jffs2/summary.c b/fs/jffs2/summary.c index 4fe64519870f..d83372d3e1a0 100644 --- a/fs/jffs2/summary.c +++ b/fs/jffs2/summary.c @@ -858,7 +858,10 @@ int jffs2_sum_write_sumnode(struct jffs2_sb_info *c) spin_unlock(&c->erase_completion_lock); jeb = c->nextblock; - jffs2_prealloc_raw_node_refs(c, jeb, 1); + ret = jffs2_prealloc_raw_node_refs(c, jeb, 1); + + if (ret) + goto out; if (!c->summary->sum_num || !c->summary->sum_list_head) { JFFS2_WARNING("Empty summary info!!!\n"); @@ -872,6 +875,8 @@ int jffs2_sum_write_sumnode(struct jffs2_sb_info *c) datasize += padsize; ret = jffs2_sum_write_data(c, jeb, infosize, datasize, padsize); + +out: spin_lock(&c->erase_completion_lock); return ret; } diff --git a/fs/libfs.c b/fs/libfs.c index 9ea0ecc325a8..6f487fc6be34 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -1649,12 +1649,10 @@ struct inode *alloc_anon_inode(struct super_block *s) */ inode->i_state = I_DIRTY; /* - * Historically anonymous inodes didn't have a type at all and - * userspace has come to rely on this. Internally they're just - * regular files but S_IFREG is masked off when reporting - * information to userspace. + * Historically anonymous inodes don't have a type at all and + * userspace has come to rely on this. */ - inode->i_mode = S_IFREG | S_IRUSR | S_IWUSR; + inode->i_mode = S_IRUSR | S_IWUSR; inode->i_uid = current_fsuid(); inode->i_gid = current_fsgid(); inode->i_flags |= S_PRIVATE | S_ANON_INODE; diff --git a/fs/namei.c b/fs/namei.c index 4bb889fc980b..c26a7ee42184 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2917,7 +2917,8 @@ static int lookup_one_common(struct mnt_idmap *idmap, * @base: base directory to lookup from * * Look up a dentry by name in the dcache, returning NULL if it does not - * currently exist. The function does not try to create a dentry. + * currently exist. The function does not try to create a dentry and if one + * is found it doesn't try to revalidate it. * * Note that this routine is purely a helper for filesystem usage and should * not be called by generic code. It does no permission checking. @@ -2933,7 +2934,7 @@ struct dentry *try_lookup_noperm(struct qstr *name, struct dentry *base) if (err) return ERR_PTR(err); - return lookup_dcache(name, base, 0); + return d_lookup(base, name); } EXPORT_SYMBOL(try_lookup_noperm); @@ -3057,14 +3058,22 @@ EXPORT_SYMBOL(lookup_one_positive_unlocked); * Note that this routine is purely a helper for filesystem usage and should * not be called by generic code. It does no permission checking. * - * Unlike lookup_noperm, it should be called without the parent + * Unlike lookup_noperm(), it should be called without the parent * i_rwsem held, and will take the i_rwsem itself if necessary. + * + * Unlike try_lookup_noperm() it *does* revalidate the dentry if it already + * existed. */ struct dentry *lookup_noperm_unlocked(struct qstr *name, struct dentry *base) { struct dentry *ret; + int err; - ret = try_lookup_noperm(name, base); + err = lookup_noperm_common(name, base); + if (err) + return ERR_PTR(err); + + ret = lookup_dcache(name, base, 0); if (!ret) ret = lookup_slow(name, base, 0); return ret; @@ -3471,7 +3480,7 @@ static int may_open(struct mnt_idmap *idmap, const struct path *path, return -EACCES; break; default: - VFS_BUG_ON_INODE(1, inode); + VFS_BUG_ON_INODE(!IS_ANON_FILE(inode), inode); } error = inode_permission(idmap, inode, MAY_OPEN | acc_mode); diff --git a/fs/namespace.c b/fs/namespace.c index 2f2e93927f46..54c59e091919 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2310,21 +2310,62 @@ out: return dst_mnt; } -/* Caller should check returned pointer for errors */ +static inline bool extend_array(struct path **res, struct path **to_free, + unsigned n, unsigned *count, unsigned new_count) +{ + struct path *p; -struct vfsmount *collect_mounts(const struct path *path) + if (likely(n < *count)) + return true; + p = kmalloc_array(new_count, sizeof(struct path), GFP_KERNEL); + if (p && *count) + memcpy(p, *res, *count * sizeof(struct path)); + *count = new_count; + kfree(*to_free); + *to_free = *res = p; + return p; +} + +struct path *collect_paths(const struct path *path, + struct path *prealloc, unsigned count) { - struct mount *tree; - namespace_lock(); - if (!check_mnt(real_mount(path->mnt))) - tree = ERR_PTR(-EINVAL); - else - tree = copy_tree(real_mount(path->mnt), path->dentry, - CL_COPY_ALL | CL_PRIVATE); - namespace_unlock(); - if (IS_ERR(tree)) - return ERR_CAST(tree); - return &tree->mnt; + struct mount *root = real_mount(path->mnt); + struct mount *child; + struct path *res = prealloc, *to_free = NULL; + unsigned n = 0; + + guard(rwsem_read)(&namespace_sem); + + if (!check_mnt(root)) + return ERR_PTR(-EINVAL); + if (!extend_array(&res, &to_free, 0, &count, 32)) + return ERR_PTR(-ENOMEM); + res[n++] = *path; + list_for_each_entry(child, &root->mnt_mounts, mnt_child) { + if (!is_subdir(child->mnt_mountpoint, path->dentry)) + continue; + for (struct mount *m = child; m; m = next_mnt(m, child)) { + if (!extend_array(&res, &to_free, n, &count, 2 * count)) + return ERR_PTR(-ENOMEM); + res[n].mnt = &m->mnt; + res[n].dentry = m->mnt.mnt_root; + n++; + } + } + if (!extend_array(&res, &to_free, n, &count, count + 1)) + return ERR_PTR(-ENOMEM); + memset(res + n, 0, (count - n) * sizeof(struct path)); + for (struct path *p = res; p->mnt; p++) + path_get(p); + return res; +} + +void drop_collected_paths(struct path *paths, struct path *prealloc) +{ + for (struct path *p = paths; p->mnt; p++) + path_put(p); + if (paths != prealloc) + kfree(paths); } static void free_mnt_ns(struct mnt_namespace *); @@ -2401,16 +2442,7 @@ void dissolve_on_fput(struct vfsmount *mnt) free_mnt_ns(ns); } -void drop_collected_mounts(struct vfsmount *mnt) -{ - namespace_lock(); - lock_mount_hash(); - umount_tree(real_mount(mnt), 0); - unlock_mount_hash(); - namespace_unlock(); -} - -bool has_locked_children(struct mount *mnt, struct dentry *dentry) +static bool __has_locked_children(struct mount *mnt, struct dentry *dentry) { struct mount *child; @@ -2424,6 +2456,16 @@ bool has_locked_children(struct mount *mnt, struct dentry *dentry) return false; } +bool has_locked_children(struct mount *mnt, struct dentry *dentry) +{ + bool res; + + read_seqlock_excl(&mount_lock); + res = __has_locked_children(mnt, dentry); + read_sequnlock_excl(&mount_lock); + return res; +} + /* * Check that there aren't references to earlier/same mount namespaces in the * specified subtree. Such references can act as pins for mount namespaces @@ -2468,23 +2510,27 @@ struct vfsmount *clone_private_mount(const struct path *path) if (IS_MNT_UNBINDABLE(old_mnt)) return ERR_PTR(-EINVAL); - if (mnt_has_parent(old_mnt)) { - if (!check_mnt(old_mnt)) - return ERR_PTR(-EINVAL); - } else { - if (!is_mounted(&old_mnt->mnt)) - return ERR_PTR(-EINVAL); - - /* Make sure this isn't something purely kernel internal. */ - if (!is_anon_ns(old_mnt->mnt_ns)) + /* + * Make sure the source mount is acceptable. + * Anything mounted in our mount namespace is allowed. + * Otherwise, it must be the root of an anonymous mount + * namespace, and we need to make sure no namespace + * loops get created. + */ + if (!check_mnt(old_mnt)) { + if (!is_mounted(&old_mnt->mnt) || + !is_anon_ns(old_mnt->mnt_ns) || + mnt_has_parent(old_mnt)) return ERR_PTR(-EINVAL); - /* Make sure we don't create mount namespace loops. */ if (!check_for_nsfs_mounts(old_mnt)) return ERR_PTR(-EINVAL); } - if (has_locked_children(old_mnt, path->dentry)) + if (!ns_capable(old_mnt->mnt_ns->user_ns, CAP_SYS_ADMIN)) + return ERR_PTR(-EPERM); + + if (__has_locked_children(old_mnt, path->dentry)) return ERR_PTR(-EINVAL); new_mnt = clone_mnt(old_mnt, path->dentry, CL_PRIVATE); @@ -2497,21 +2543,6 @@ struct vfsmount *clone_private_mount(const struct path *path) } EXPORT_SYMBOL_GPL(clone_private_mount); -int iterate_mounts(int (*f)(struct vfsmount *, void *), void *arg, - struct vfsmount *root) -{ - struct mount *mnt; - int res = f(root, arg); - if (res) - return res; - list_for_each_entry(mnt, &real_mount(root)->mnt_list, mnt_list) { - res = f(&mnt->mnt, arg); - if (res) - return res; - } - return 0; -} - static void lock_mnt_tree(struct mount *mnt) { struct mount *p; @@ -2737,14 +2768,14 @@ static int attach_recursive_mnt(struct mount *source_mnt, hlist_for_each_entry_safe(child, n, &tree_list, mnt_hash) { struct mount *q; hlist_del_init(&child->mnt_hash); - q = __lookup_mnt(&child->mnt_parent->mnt, - child->mnt_mountpoint); - if (q) - mnt_change_mountpoint(child, smp, q); /* Notice when we are propagating across user namespaces */ if (child->mnt_parent->mnt_ns->user_ns != user_ns) lock_mnt_tree(child); child->mnt.mnt_flags &= ~MNT_LOCKED; + q = __lookup_mnt(&child->mnt_parent->mnt, + child->mnt_mountpoint); + if (q) + mnt_change_mountpoint(child, smp, q); commit_tree(child); } put_mountpoint(smp); @@ -2930,6 +2961,10 @@ static int do_change_type(struct path *path, int ms_flags) return -EINVAL; namespace_lock(); + if (!check_mnt(mnt)) { + err = -EINVAL; + goto out_unlock; + } if (type == MS_SHARED) { err = invent_group_ids(mnt, recurse); if (err) @@ -3021,7 +3056,7 @@ static struct mount *__do_loopback(struct path *old_path, int recurse) if (!may_copy_tree(old_path)) return mnt; - if (!recurse && has_locked_children(old, old_path->dentry)) + if (!recurse && __has_locked_children(old, old_path->dentry)) return mnt; if (recurse) @@ -3414,7 +3449,7 @@ static int do_set_group(struct path *from_path, struct path *to_path) goto out; /* From mount should not have locked children in place of To's root */ - if (has_locked_children(from, to->mnt.mnt_root)) + if (__has_locked_children(from, to->mnt.mnt_root)) goto out; /* Setting sharing groups is only allowed on private mounts */ @@ -3428,7 +3463,7 @@ static int do_set_group(struct path *from_path, struct path *to_path) if (IS_MNT_SLAVE(from)) { struct mount *m = from->mnt_master; - list_add(&to->mnt_slave, &m->mnt_slave_list); + list_add(&to->mnt_slave, &from->mnt_slave); to->mnt_master = m; } @@ -3453,18 +3488,25 @@ out: * Check if path is overmounted, i.e., if there's a mount on top of * @path->mnt with @path->dentry as mountpoint. * - * Context: This function expects namespace_lock() to be held. + * Context: namespace_sem must be held at least shared. + * MUST NOT be called under lock_mount_hash() (there one should just + * call __lookup_mnt() and check if it returns NULL). * Return: If path is overmounted true is returned, false if not. */ static inline bool path_overmounted(const struct path *path) { + unsigned seq = read_seqbegin(&mount_lock); + bool no_child; + rcu_read_lock(); - if (unlikely(__lookup_mnt(path->mnt, path->dentry))) { - rcu_read_unlock(); - return true; - } + no_child = !__lookup_mnt(path->mnt, path->dentry); rcu_read_unlock(); - return false; + if (need_seqretry(&mount_lock, seq)) { + read_seqlock_excl(&mount_lock); + no_child = !__lookup_mnt(path->mnt, path->dentry); + read_sequnlock_excl(&mount_lock); + } + return unlikely(!no_child); } /** @@ -3623,37 +3665,41 @@ static int do_move_mount(struct path *old_path, ns = old->mnt_ns; err = -EINVAL; - if (!may_use_mount(p)) - goto out; - /* The thing moved must be mounted... */ if (!is_mounted(&old->mnt)) goto out; - /* ... and either ours or the root of anon namespace */ - if (!(attached ? check_mnt(old) : is_anon_ns(ns))) - goto out; - - if (is_anon_ns(ns) && ns == p->mnt_ns) { + if (check_mnt(old)) { + /* if the source is in our namespace... */ + /* ... it should be detachable from parent */ + if (!mnt_has_parent(old) || IS_MNT_LOCKED(old)) + goto out; + /* ... and the target should be in our namespace */ + if (!check_mnt(p)) + goto out; + } else { /* - * Ending up with two files referring to the root of the - * same anonymous mount namespace would cause an error - * as this would mean trying to move the same mount - * twice into the mount tree which would be rejected - * later. But be explicit about it right here. + * otherwise the source must be the root of some anon namespace. + * AV: check for mount being root of an anon namespace is worth + * an inlined predicate... */ - goto out; - } else if (is_anon_ns(p->mnt_ns)) { + if (!is_anon_ns(ns) || mnt_has_parent(old)) + goto out; /* - * Don't allow moving an attached mount tree to an - * anonymous mount tree. + * Bail out early if the target is within the same namespace - + * subsequent checks would've rejected that, but they lose + * some corner cases if we check it early. */ - goto out; + if (ns == p->mnt_ns) + goto out; + /* + * Target should be either in our namespace or in an acceptable + * anon namespace, sensu check_anonymous_mnt(). + */ + if (!may_use_mount(p)) + goto out; } - if (old->mnt.mnt_flags & MNT_LOCKED) - goto out; - if (!path_mounted(old_path)) goto out; @@ -5261,16 +5307,12 @@ SYSCALL_DEFINE5(open_tree_attr, int, dfd, const char __user *, filename, kattr.kflags |= MOUNT_KATTR_RECURSE; ret = wants_mount_setattr(uattr, usize, &kattr); - if (ret < 0) - return ret; - - if (ret) { + if (ret > 0) { ret = do_mount_setattr(&file->f_path, &kattr); - if (ret) - return ret; - finish_mount_kattr(&kattr); } + if (ret) + return ret; } fd = get_unused_fd_flags(flags & O_CLOEXEC); @@ -6233,7 +6275,11 @@ void put_mnt_ns(struct mnt_namespace *ns) { if (!refcount_dec_and_test(&ns->ns.count)) return; - drop_collected_mounts(&ns->root->mnt); + namespace_lock(); + lock_mount_hash(); + umount_tree(ns->root, 0); + unlock_mount_hash(); + namespace_unlock(); free_mnt_ns(ns); } diff --git a/fs/netfs/buffered_write.c b/fs/netfs/buffered_write.c index 72a3e6db2524..f27ea5099a68 100644 --- a/fs/netfs/buffered_write.c +++ b/fs/netfs/buffered_write.c @@ -53,30 +53,40 @@ static struct folio *netfs_grab_folio_for_write(struct address_space *mapping, * data written into the pagecache until we can find out from the server what * the values actually are. */ -static void netfs_update_i_size(struct netfs_inode *ctx, struct inode *inode, - loff_t i_size, loff_t pos, size_t copied) +void netfs_update_i_size(struct netfs_inode *ctx, struct inode *inode, + loff_t pos, size_t copied) { + loff_t i_size, end = pos + copied; blkcnt_t add; size_t gap; + if (end <= i_size_read(inode)) + return; + if (ctx->ops->update_i_size) { - ctx->ops->update_i_size(inode, pos); + ctx->ops->update_i_size(inode, end); return; } - i_size_write(inode, pos); + spin_lock(&inode->i_lock); + + i_size = i_size_read(inode); + if (end > i_size) { + i_size_write(inode, end); #if IS_ENABLED(CONFIG_FSCACHE) - fscache_update_cookie(ctx->cache, NULL, &pos); + fscache_update_cookie(ctx->cache, NULL, &end); #endif - gap = SECTOR_SIZE - (i_size & (SECTOR_SIZE - 1)); - if (copied > gap) { - add = DIV_ROUND_UP(copied - gap, SECTOR_SIZE); + gap = SECTOR_SIZE - (i_size & (SECTOR_SIZE - 1)); + if (copied > gap) { + add = DIV_ROUND_UP(copied - gap, SECTOR_SIZE); - inode->i_blocks = min_t(blkcnt_t, - DIV_ROUND_UP(pos, SECTOR_SIZE), - inode->i_blocks + add); + inode->i_blocks = min_t(blkcnt_t, + DIV_ROUND_UP(end, SECTOR_SIZE), + inode->i_blocks + add); + } } + spin_unlock(&inode->i_lock); } /** @@ -111,7 +121,7 @@ ssize_t netfs_perform_write(struct kiocb *iocb, struct iov_iter *iter, struct folio *folio = NULL, *writethrough = NULL; unsigned int bdp_flags = (iocb->ki_flags & IOCB_NOWAIT) ? BDP_ASYNC : 0; ssize_t written = 0, ret, ret2; - loff_t i_size, pos = iocb->ki_pos; + loff_t pos = iocb->ki_pos; size_t max_chunk = mapping_max_folio_size(mapping); bool maybe_trouble = false; @@ -344,10 +354,8 @@ ssize_t netfs_perform_write(struct kiocb *iocb, struct iov_iter *iter, flush_dcache_folio(folio); /* Update the inode size if we moved the EOF marker */ + netfs_update_i_size(ctx, inode, pos, copied); pos += copied; - i_size = i_size_read(inode); - if (pos > i_size) - netfs_update_i_size(ctx, inode, i_size, pos, copied); written += copied; if (likely(!wreq)) { diff --git a/fs/netfs/direct_write.c b/fs/netfs/direct_write.c index fa9a5bf3c6d5..a16660ab7f83 100644 --- a/fs/netfs/direct_write.c +++ b/fs/netfs/direct_write.c @@ -9,20 +9,6 @@ #include <linux/uio.h> #include "internal.h" -static void netfs_cleanup_dio_write(struct netfs_io_request *wreq) -{ - struct inode *inode = wreq->inode; - unsigned long long end = wreq->start + wreq->transferred; - - if (!wreq->error && - i_size_read(inode) < end) { - if (wreq->netfs_ops->update_i_size) - wreq->netfs_ops->update_i_size(inode, end); - else - i_size_write(inode, end); - } -} - /* * Perform an unbuffered write where we may have to do an RMW operation on an * encrypted file. This can also be used for direct I/O writes. @@ -98,7 +84,6 @@ ssize_t netfs_unbuffered_write_iter_locked(struct kiocb *iocb, struct iov_iter * if (async) wreq->iocb = iocb; wreq->len = iov_iter_count(&wreq->buffer.iter); - wreq->cleanup = netfs_cleanup_dio_write; ret = netfs_unbuffered_write(wreq, is_sync_kiocb(iocb), wreq->len); if (ret < 0) { _debug("begin = %zd", ret); @@ -106,7 +91,6 @@ ssize_t netfs_unbuffered_write_iter_locked(struct kiocb *iocb, struct iov_iter * } if (!async) { - trace_netfs_rreq(wreq, netfs_rreq_trace_wait_ip); ret = netfs_wait_for_write(wreq); if (ret > 0) iocb->ki_pos += ret; diff --git a/fs/netfs/internal.h b/fs/netfs/internal.h index e2ee9183392b..d4f16fefd965 100644 --- a/fs/netfs/internal.h +++ b/fs/netfs/internal.h @@ -28,6 +28,12 @@ int netfs_prefetch_for_write(struct file *file, struct folio *folio, size_t offset, size_t len); /* + * buffered_write.c + */ +void netfs_update_i_size(struct netfs_inode *ctx, struct inode *inode, + loff_t pos, size_t copied); + +/* * main.c */ extern unsigned int netfs_debug; @@ -267,14 +273,32 @@ static inline void netfs_wake_rreq_flag(struct netfs_io_request *rreq, enum netfs_rreq_trace trace) { if (test_bit(rreq_flag, &rreq->flags)) { - trace_netfs_rreq(rreq, trace); clear_bit_unlock(rreq_flag, &rreq->flags); smp_mb__after_atomic(); /* Set flag before task state */ + trace_netfs_rreq(rreq, trace); wake_up(&rreq->waitq); } } /* + * Test the NETFS_RREQ_IN_PROGRESS flag, inserting an appropriate barrier. + */ +static inline bool netfs_check_rreq_in_progress(const struct netfs_io_request *rreq) +{ + /* Order read of flags before read of anything else, such as error. */ + return test_bit_acquire(NETFS_RREQ_IN_PROGRESS, &rreq->flags); +} + +/* + * Test the NETFS_SREQ_IN_PROGRESS flag, inserting an appropriate barrier. + */ +static inline bool netfs_check_subreq_in_progress(const struct netfs_io_subrequest *subreq) +{ + /* Order read of flags before read of anything else, such as error. */ + return test_bit_acquire(NETFS_SREQ_IN_PROGRESS, &subreq->flags); +} + +/* * fscache-cache.c */ #ifdef CONFIG_PROC_FS diff --git a/fs/netfs/main.c b/fs/netfs/main.c index 3db401d269e7..73da6c9f5777 100644 --- a/fs/netfs/main.c +++ b/fs/netfs/main.c @@ -58,15 +58,15 @@ static int netfs_requests_seq_show(struct seq_file *m, void *v) if (v == &netfs_io_requests) { seq_puts(m, - "REQUEST OR REF FL ERR OPS COVERAGE\n" - "======== == === == ==== === =========\n" + "REQUEST OR REF FLAG ERR OPS COVERAGE\n" + "======== == === ==== ==== === =========\n" ); return 0; } rreq = list_entry(v, struct netfs_io_request, proc_link); seq_printf(m, - "%08x %s %3d %2lx %4ld %3d @%04llx %llx/%llx", + "%08x %s %3d %4lx %4ld %3d @%04llx %llx/%llx", rreq->debug_id, netfs_origins[rreq->origin], refcount_read(&rreq->ref), diff --git a/fs/netfs/misc.c b/fs/netfs/misc.c index 43b67a28a8fa..20748bcfbf59 100644 --- a/fs/netfs/misc.c +++ b/fs/netfs/misc.c @@ -356,22 +356,22 @@ void netfs_wait_for_in_progress_stream(struct netfs_io_request *rreq, DEFINE_WAIT(myself); list_for_each_entry(subreq, &stream->subrequests, rreq_link) { - if (!test_bit(NETFS_SREQ_IN_PROGRESS, &subreq->flags)) + if (!netfs_check_subreq_in_progress(subreq)) continue; - trace_netfs_rreq(rreq, netfs_rreq_trace_wait_queue); + trace_netfs_rreq(rreq, netfs_rreq_trace_wait_quiesce); for (;;) { prepare_to_wait(&rreq->waitq, &myself, TASK_UNINTERRUPTIBLE); - if (!test_bit(NETFS_SREQ_IN_PROGRESS, &subreq->flags)) + if (!netfs_check_subreq_in_progress(subreq)) break; trace_netfs_sreq(subreq, netfs_sreq_trace_wait_for); schedule(); - trace_netfs_rreq(rreq, netfs_rreq_trace_woke_queue); } } + trace_netfs_rreq(rreq, netfs_rreq_trace_waited_quiesce); finish_wait(&rreq->waitq, &myself); } @@ -381,7 +381,12 @@ void netfs_wait_for_in_progress_stream(struct netfs_io_request *rreq, static int netfs_collect_in_app(struct netfs_io_request *rreq, bool (*collector)(struct netfs_io_request *rreq)) { - bool need_collect = false, inactive = true; + bool need_collect = false, inactive = true, done = true; + + if (!netfs_check_rreq_in_progress(rreq)) { + trace_netfs_rreq(rreq, netfs_rreq_trace_recollect); + return 1; /* Done */ + } for (int i = 0; i < NR_IO_STREAMS; i++) { struct netfs_io_subrequest *subreq; @@ -395,14 +400,16 @@ static int netfs_collect_in_app(struct netfs_io_request *rreq, struct netfs_io_subrequest, rreq_link); if (subreq && - (!test_bit(NETFS_SREQ_IN_PROGRESS, &subreq->flags) || + (!netfs_check_subreq_in_progress(subreq) || test_bit(NETFS_SREQ_MADE_PROGRESS, &subreq->flags))) { need_collect = true; break; } + if (subreq || !test_bit(NETFS_RREQ_ALL_QUEUED, &rreq->flags)) + done = false; } - if (!need_collect && !inactive) + if (!need_collect && !inactive && !done) return 0; /* Sleep */ __set_current_state(TASK_RUNNING); @@ -423,14 +430,13 @@ static int netfs_collect_in_app(struct netfs_io_request *rreq, /* * Wait for a request to complete, successfully or otherwise. */ -static ssize_t netfs_wait_for_request(struct netfs_io_request *rreq, - bool (*collector)(struct netfs_io_request *rreq)) +static ssize_t netfs_wait_for_in_progress(struct netfs_io_request *rreq, + bool (*collector)(struct netfs_io_request *rreq)) { DEFINE_WAIT(myself); ssize_t ret; for (;;) { - trace_netfs_rreq(rreq, netfs_rreq_trace_wait_queue); prepare_to_wait(&rreq->waitq, &myself, TASK_UNINTERRUPTIBLE); if (!test_bit(NETFS_RREQ_OFFLOAD_COLLECTION, &rreq->flags)) { @@ -440,18 +446,22 @@ static ssize_t netfs_wait_for_request(struct netfs_io_request *rreq, case 1: goto all_collected; case 2: + if (!netfs_check_rreq_in_progress(rreq)) + break; + cond_resched(); continue; } } - if (!test_bit(NETFS_RREQ_IN_PROGRESS, &rreq->flags)) + if (!netfs_check_rreq_in_progress(rreq)) break; + trace_netfs_rreq(rreq, netfs_rreq_trace_wait_ip); schedule(); - trace_netfs_rreq(rreq, netfs_rreq_trace_woke_queue); } all_collected: + trace_netfs_rreq(rreq, netfs_rreq_trace_waited_ip); finish_wait(&rreq->waitq, &myself); ret = rreq->error; @@ -478,12 +488,12 @@ all_collected: ssize_t netfs_wait_for_read(struct netfs_io_request *rreq) { - return netfs_wait_for_request(rreq, netfs_read_collection); + return netfs_wait_for_in_progress(rreq, netfs_read_collection); } ssize_t netfs_wait_for_write(struct netfs_io_request *rreq) { - return netfs_wait_for_request(rreq, netfs_write_collection); + return netfs_wait_for_in_progress(rreq, netfs_write_collection); } /* @@ -494,10 +504,8 @@ static void netfs_wait_for_pause(struct netfs_io_request *rreq, { DEFINE_WAIT(myself); - trace_netfs_rreq(rreq, netfs_rreq_trace_wait_pause); - for (;;) { - trace_netfs_rreq(rreq, netfs_rreq_trace_wait_queue); + trace_netfs_rreq(rreq, netfs_rreq_trace_wait_pause); prepare_to_wait(&rreq->waitq, &myself, TASK_UNINTERRUPTIBLE); if (!test_bit(NETFS_RREQ_OFFLOAD_COLLECTION, &rreq->flags)) { @@ -507,19 +515,23 @@ static void netfs_wait_for_pause(struct netfs_io_request *rreq, case 1: goto all_collected; case 2: + if (!netfs_check_rreq_in_progress(rreq) || + !test_bit(NETFS_RREQ_PAUSE, &rreq->flags)) + break; + cond_resched(); continue; } } - if (!test_bit(NETFS_RREQ_IN_PROGRESS, &rreq->flags) || + if (!netfs_check_rreq_in_progress(rreq) || !test_bit(NETFS_RREQ_PAUSE, &rreq->flags)) break; schedule(); - trace_netfs_rreq(rreq, netfs_rreq_trace_woke_queue); } all_collected: + trace_netfs_rreq(rreq, netfs_rreq_trace_waited_pause); finish_wait(&rreq->waitq, &myself); } diff --git a/fs/netfs/read_collect.c b/fs/netfs/read_collect.c index 96ee18af28ef..3e804da1e1eb 100644 --- a/fs/netfs/read_collect.c +++ b/fs/netfs/read_collect.c @@ -218,7 +218,7 @@ reassess: stream->collected_to = front->start; } - if (test_bit(NETFS_SREQ_IN_PROGRESS, &front->flags)) + if (netfs_check_subreq_in_progress(front)) notes |= HIT_PENDING; smp_rmb(); /* Read counters after IN_PROGRESS flag. */ transferred = READ_ONCE(front->transferred); @@ -293,7 +293,9 @@ reassess: spin_lock(&rreq->lock); remove = front; - trace_netfs_sreq(front, netfs_sreq_trace_discard); + trace_netfs_sreq(front, + notes & ABANDON_SREQ ? + netfs_sreq_trace_abandoned : netfs_sreq_trace_consumed); list_del_init(&front->rreq_link); front = list_first_entry_or_null(&stream->subrequests, struct netfs_io_subrequest, rreq_link); @@ -353,9 +355,11 @@ static void netfs_rreq_assess_dio(struct netfs_io_request *rreq) if (rreq->iocb) { rreq->iocb->ki_pos += rreq->transferred; - if (rreq->iocb->ki_complete) + if (rreq->iocb->ki_complete) { + trace_netfs_rreq(rreq, netfs_rreq_trace_ki_complete); rreq->iocb->ki_complete( rreq->iocb, rreq->error ? rreq->error : rreq->transferred); + } } if (rreq->netfs_ops->done) rreq->netfs_ops->done(rreq); @@ -379,9 +383,11 @@ static void netfs_rreq_assess_single(struct netfs_io_request *rreq) if (rreq->iocb) { rreq->iocb->ki_pos += rreq->transferred; - if (rreq->iocb->ki_complete) + if (rreq->iocb->ki_complete) { + trace_netfs_rreq(rreq, netfs_rreq_trace_ki_complete); rreq->iocb->ki_complete( rreq->iocb, rreq->error ? rreq->error : rreq->transferred); + } } if (rreq->netfs_ops->done) rreq->netfs_ops->done(rreq); @@ -445,7 +451,7 @@ void netfs_read_collection_worker(struct work_struct *work) struct netfs_io_request *rreq = container_of(work, struct netfs_io_request, work); netfs_see_request(rreq, netfs_rreq_trace_see_work); - if (test_bit(NETFS_RREQ_IN_PROGRESS, &rreq->flags)) { + if (netfs_check_rreq_in_progress(rreq)) { if (netfs_read_collection(rreq)) /* Drop the ref from the IN_PROGRESS flag. */ netfs_put_request(rreq, netfs_rreq_trace_put_work_ip); diff --git a/fs/netfs/read_pgpriv2.c b/fs/netfs/read_pgpriv2.c index 5bbe906a551d..8097bc069c1d 100644 --- a/fs/netfs/read_pgpriv2.c +++ b/fs/netfs/read_pgpriv2.c @@ -110,6 +110,8 @@ static struct netfs_io_request *netfs_pgpriv2_begin_copy_to_cache( if (!creq->io_streams[1].avail) goto cancel_put; + __set_bit(NETFS_RREQ_OFFLOAD_COLLECTION, &creq->flags); + trace_netfs_copy2cache(rreq, creq); trace_netfs_write(creq, netfs_write_trace_copy_to_cache); netfs_stat(&netfs_n_wh_copy_to_cache); rreq->copy_to_cache = creq; @@ -154,6 +156,9 @@ void netfs_pgpriv2_end_copy_to_cache(struct netfs_io_request *rreq) netfs_issue_write(creq, &creq->io_streams[1]); smp_wmb(); /* Write lists before ALL_QUEUED. */ set_bit(NETFS_RREQ_ALL_QUEUED, &creq->flags); + trace_netfs_rreq(rreq, netfs_rreq_trace_end_copy_to_cache); + if (list_empty_careful(&creq->io_streams[1].subrequests)) + netfs_wake_collector(creq); netfs_put_request(creq, netfs_rreq_trace_put_return); creq->copy_to_cache = NULL; diff --git a/fs/netfs/write_collect.c b/fs/netfs/write_collect.c index e2b102ffb768..0f3a36852a4d 100644 --- a/fs/netfs/write_collect.c +++ b/fs/netfs/write_collect.c @@ -240,7 +240,7 @@ reassess_streams: } /* Stall if the front is still undergoing I/O. */ - if (test_bit(NETFS_SREQ_IN_PROGRESS, &front->flags)) { + if (netfs_check_subreq_in_progress(front)) { notes |= HIT_PENDING; break; } @@ -393,8 +393,10 @@ bool netfs_write_collection(struct netfs_io_request *wreq) ictx->ops->invalidate_cache(wreq); } - if (wreq->cleanup) - wreq->cleanup(wreq); + if ((wreq->origin == NETFS_UNBUFFERED_WRITE || + wreq->origin == NETFS_DIO_WRITE) && + !wreq->error) + netfs_update_i_size(ictx, &ictx->inode, wreq->start, wreq->transferred); if (wreq->origin == NETFS_DIO_WRITE && wreq->mapping->nrpages) { @@ -419,9 +421,11 @@ bool netfs_write_collection(struct netfs_io_request *wreq) if (wreq->iocb) { size_t written = min(wreq->transferred, wreq->len); wreq->iocb->ki_pos += written; - if (wreq->iocb->ki_complete) + if (wreq->iocb->ki_complete) { + trace_netfs_rreq(wreq, netfs_rreq_trace_ki_complete); wreq->iocb->ki_complete( wreq->iocb, wreq->error ? wreq->error : written); + } wreq->iocb = VFS_PTR_POISON; } @@ -434,7 +438,7 @@ void netfs_write_collection_worker(struct work_struct *work) struct netfs_io_request *rreq = container_of(work, struct netfs_io_request, work); netfs_see_request(rreq, netfs_rreq_trace_see_work); - if (test_bit(NETFS_RREQ_IN_PROGRESS, &rreq->flags)) { + if (netfs_check_rreq_in_progress(rreq)) { if (netfs_write_collection(rreq)) /* Drop the ref from the IN_PROGRESS flag. */ netfs_put_request(rreq, netfs_rreq_trace_put_work_ip); diff --git a/fs/netfs/write_retry.c b/fs/netfs/write_retry.c index 9d1d8a8bab72..fc9c3e0d34d8 100644 --- a/fs/netfs/write_retry.c +++ b/fs/netfs/write_retry.c @@ -146,14 +146,13 @@ static void netfs_retry_write_stream(struct netfs_io_request *wreq, subreq = netfs_alloc_subrequest(wreq); subreq->source = to->source; subreq->start = start; - subreq->debug_index = atomic_inc_return(&wreq->subreq_counter); subreq->stream_nr = to->stream_nr; subreq->retry_count = 1; trace_netfs_sreq_ref(wreq->debug_id, subreq->debug_index, refcount_read(&subreq->ref), netfs_sreq_trace_new); - netfs_get_subrequest(subreq, netfs_sreq_trace_get_resubmit); + trace_netfs_sreq(subreq, netfs_sreq_trace_split); list_add(&subreq->rreq_link, &to->rreq_link); to = list_next_entry(to, rreq_link); diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 6d63b958c4bb..cf35ad3f818a 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -180,7 +180,7 @@ struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init) clp->cl_proto = cl_init->proto; clp->cl_nconnect = cl_init->nconnect; clp->cl_max_connect = cl_init->max_connect ? cl_init->max_connect : 1; - clp->cl_net = get_net(cl_init->net); + clp->cl_net = get_net_track(cl_init->net, &clp->cl_ns_tracker, GFP_KERNEL); #if IS_ENABLED(CONFIG_NFS_LOCALIO) seqlock_init(&clp->cl_boot_lock); @@ -250,7 +250,7 @@ void nfs_free_client(struct nfs_client *clp) if (!IS_ERR(clp->cl_rpcclient)) rpc_shutdown_client(clp->cl_rpcclient); - put_net(clp->cl_net); + put_net_track(clp->cl_net, &clp->cl_ns_tracker); put_nfs_version(clp->cl_nfs_mod); kfree(clp->cl_hostname); kfree(clp->cl_acceptor); @@ -439,7 +439,7 @@ struct nfs_client *nfs_get_client(const struct nfs_client_initdata *cl_init) spin_unlock(&nn->nfs_client_lock); new = rpc_ops->init_client(new, cl_init); if (!IS_ERR(new)) - nfs_local_probe(new); + nfs_local_probe_async(new); return new; } diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 8bdbc4dca89c..10ef46e29b25 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -1021,13 +1021,6 @@ out: nfs_inode_find_state_and_recover(inode, stateid); } -void nfs_remove_bad_delegation(struct inode *inode, - const nfs4_stateid *stateid) -{ - nfs_revoke_delegation(inode, stateid); -} -EXPORT_SYMBOL_GPL(nfs_remove_bad_delegation); - void nfs_delegation_mark_returned(struct inode *inode, const nfs4_stateid *stateid) { @@ -1070,6 +1063,24 @@ out_rcu_unlock: } /** + * nfs_remove_bad_delegation - handle delegations that are unusable + * @inode: inode to process + * @stateid: the delegation's stateid + * + * If the server ACK-ed our FREE_STATEID then clean + * up the delegation, else mark and keep the revoked state. + */ +void nfs_remove_bad_delegation(struct inode *inode, + const nfs4_stateid *stateid) +{ + if (stateid && stateid->type == NFS4_FREED_STATEID_TYPE) + nfs_delegation_mark_returned(inode, stateid); + else + nfs_revoke_delegation(inode, stateid); +} +EXPORT_SYMBOL_GPL(nfs_remove_bad_delegation); + +/** * nfs_expire_unused_delegation_types * @clp: client to process * @flags: delegation types to expire diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index e6909cafab68..4bea008dbebd 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1105,6 +1105,7 @@ static void ff_layout_reset_read(struct nfs_pgio_header *hdr) } static int ff_layout_async_handle_error_v4(struct rpc_task *task, + u32 op_status, struct nfs4_state *state, struct nfs_client *clp, struct pnfs_layout_segment *lseg, @@ -1115,32 +1116,42 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task, struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx); struct nfs4_slot_table *tbl = &clp->cl_session->fc_slot_table; - switch (task->tk_status) { - case -NFS4ERR_BADSESSION: - case -NFS4ERR_BADSLOT: - case -NFS4ERR_BAD_HIGH_SLOT: - case -NFS4ERR_DEADSESSION: - case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: - case -NFS4ERR_SEQ_FALSE_RETRY: - case -NFS4ERR_SEQ_MISORDERED: + switch (op_status) { + case NFS4_OK: + case NFS4ERR_NXIO: + break; + case NFSERR_PERM: + if (!task->tk_xprt) + break; + xprt_force_disconnect(task->tk_xprt); + goto out_retry; + case NFS4ERR_BADSESSION: + case NFS4ERR_BADSLOT: + case NFS4ERR_BAD_HIGH_SLOT: + case NFS4ERR_DEADSESSION: + case NFS4ERR_CONN_NOT_BOUND_TO_SESSION: + case NFS4ERR_SEQ_FALSE_RETRY: + case NFS4ERR_SEQ_MISORDERED: dprintk("%s ERROR %d, Reset session. Exchangeid " "flags 0x%x\n", __func__, task->tk_status, clp->cl_exchange_flags); nfs4_schedule_session_recovery(clp->cl_session, task->tk_status); - break; - case -NFS4ERR_DELAY: - case -NFS4ERR_GRACE: + goto out_retry; + case NFS4ERR_DELAY: + nfs_inc_stats(lseg->pls_layout->plh_inode, NFSIOS_DELAY); + fallthrough; + case NFS4ERR_GRACE: rpc_delay(task, FF_LAYOUT_POLL_RETRY_MAX); - break; - case -NFS4ERR_RETRY_UNCACHED_REP: - break; + goto out_retry; + case NFS4ERR_RETRY_UNCACHED_REP: + goto out_retry; /* Invalidate Layout errors */ - case -NFS4ERR_PNFS_NO_LAYOUT: - case -ESTALE: /* mapped NFS4ERR_STALE */ - case -EBADHANDLE: /* mapped NFS4ERR_BADHANDLE */ - case -EISDIR: /* mapped NFS4ERR_ISDIR */ - case -NFS4ERR_FHEXPIRED: - case -NFS4ERR_WRONG_TYPE: + case NFS4ERR_PNFS_NO_LAYOUT: + case NFS4ERR_STALE: + case NFS4ERR_BADHANDLE: + case NFS4ERR_ISDIR: + case NFS4ERR_FHEXPIRED: + case NFS4ERR_WRONG_TYPE: dprintk("%s Invalid layout error %d\n", __func__, task->tk_status); /* @@ -1153,6 +1164,11 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task, pnfs_destroy_layout(NFS_I(inode)); rpc_wake_up(&tbl->slot_tbl_waitq); goto reset; + default: + break; + } + + switch (task->tk_status) { /* RPC connection errors */ case -ENETDOWN: case -ENETUNREACH: @@ -1172,27 +1188,56 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task, nfs4_delete_deviceid(devid->ld, devid->nfs_client, &devid->deviceid); rpc_wake_up(&tbl->slot_tbl_waitq); - fallthrough; + break; default: - if (ff_layout_avoid_mds_available_ds(lseg)) - return -NFS4ERR_RESET_TO_PNFS; -reset: - dprintk("%s Retry through MDS. Error %d\n", __func__, - task->tk_status); - return -NFS4ERR_RESET_TO_MDS; + break; } + + if (ff_layout_avoid_mds_available_ds(lseg)) + return -NFS4ERR_RESET_TO_PNFS; +reset: + dprintk("%s Retry through MDS. Error %d\n", __func__, + task->tk_status); + return -NFS4ERR_RESET_TO_MDS; + +out_retry: task->tk_status = 0; return -EAGAIN; } /* Retry all errors through either pNFS or MDS except for -EJUKEBOX */ static int ff_layout_async_handle_error_v3(struct rpc_task *task, + u32 op_status, struct nfs_client *clp, struct pnfs_layout_segment *lseg, u32 idx) { struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx); + switch (op_status) { + case NFS_OK: + case NFSERR_NXIO: + break; + case NFSERR_PERM: + if (!task->tk_xprt) + break; + xprt_force_disconnect(task->tk_xprt); + goto out_retry; + case NFSERR_ACCES: + case NFSERR_BADHANDLE: + case NFSERR_FBIG: + case NFSERR_IO: + case NFSERR_NOSPC: + case NFSERR_ROFS: + case NFSERR_STALE: + goto out_reset_to_pnfs; + case NFSERR_JUKEBOX: + nfs_inc_stats(lseg->pls_layout->plh_inode, NFSIOS_DELAY); + goto out_retry; + default: + break; + } + switch (task->tk_status) { /* File access problems. Don't mark the device as unavailable */ case -EACCES: @@ -1216,6 +1261,7 @@ static int ff_layout_async_handle_error_v3(struct rpc_task *task, nfs4_delete_deviceid(devid->ld, devid->nfs_client, &devid->deviceid); } +out_reset_to_pnfs: /* FIXME: Need to prevent infinite looping here. */ return -NFS4ERR_RESET_TO_PNFS; out_retry: @@ -1226,6 +1272,7 @@ out_retry: } static int ff_layout_async_handle_error(struct rpc_task *task, + u32 op_status, struct nfs4_state *state, struct nfs_client *clp, struct pnfs_layout_segment *lseg, @@ -1244,10 +1291,11 @@ static int ff_layout_async_handle_error(struct rpc_task *task, switch (vers) { case 3: - return ff_layout_async_handle_error_v3(task, clp, lseg, idx); - case 4: - return ff_layout_async_handle_error_v4(task, state, clp, + return ff_layout_async_handle_error_v3(task, op_status, clp, lseg, idx); + case 4: + return ff_layout_async_handle_error_v4(task, op_status, state, + clp, lseg, idx); default: /* should never happen */ WARN_ON_ONCE(1); @@ -1300,6 +1348,7 @@ static void ff_layout_io_track_ds_error(struct pnfs_layout_segment *lseg, switch (status) { case NFS4ERR_DELAY: case NFS4ERR_GRACE: + case NFS4ERR_PERM: break; case NFS4ERR_NXIO: ff_layout_mark_ds_unreachable(lseg, idx); @@ -1332,7 +1381,8 @@ static int ff_layout_read_done_cb(struct rpc_task *task, trace_ff_layout_read_error(hdr, task->tk_status); } - err = ff_layout_async_handle_error(task, hdr->args.context->state, + err = ff_layout_async_handle_error(task, hdr->res.op_status, + hdr->args.context->state, hdr->ds_clp, hdr->lseg, hdr->pgio_mirror_idx); @@ -1505,7 +1555,8 @@ static int ff_layout_write_done_cb(struct rpc_task *task, trace_ff_layout_write_error(hdr, task->tk_status); } - err = ff_layout_async_handle_error(task, hdr->args.context->state, + err = ff_layout_async_handle_error(task, hdr->res.op_status, + hdr->args.context->state, hdr->ds_clp, hdr->lseg, hdr->pgio_mirror_idx); @@ -1554,8 +1605,9 @@ static int ff_layout_commit_done_cb(struct rpc_task *task, trace_ff_layout_commit_error(data, task->tk_status); } - err = ff_layout_async_handle_error(task, NULL, data->ds_clp, - data->lseg, data->ds_commit_index); + err = ff_layout_async_handle_error(task, data->res.op_status, + NULL, data->ds_clp, data->lseg, + data->ds_commit_index); trace_nfs4_pnfs_commit_ds(data, err); switch (err) { diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 4a304cf17c4b..656d5c50bbce 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -400,7 +400,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, * keep ds_clp even if DS is local, so that if local IO cannot * proceed somehow, we can fall back to NFS whenever we want. */ - nfs_local_probe(ds->ds_clp); + nfs_local_probe_async(ds->ds_clp); max_payload = nfs_block_size(rpc_max_payload(ds->ds_clp->cl_rpcclient), NULL); diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 119e447758b9..a2fa6bc4d74e 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -557,6 +557,8 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) set_nlink(inode, fattr->nlink); else if (fattr_supported & NFS_ATTR_FATTR_NLINK) nfs_set_cache_invalid(inode, NFS_INO_INVALID_NLINK); + else + set_nlink(inode, 1); if (fattr->valid & NFS_ATTR_FATTR_OWNER) inode->i_uid = fattr->uid; else if (fattr_supported & NFS_ATTR_FATTR_OWNER) @@ -633,6 +635,34 @@ nfs_fattr_fixup_delegated(struct inode *inode, struct nfs_fattr *fattr) } } +static void nfs_set_timestamps_to_ts(struct inode *inode, struct iattr *attr) +{ + unsigned int cache_flags = 0; + + if (attr->ia_valid & ATTR_MTIME_SET) { + struct timespec64 ctime = inode_get_ctime(inode); + struct timespec64 mtime = inode_get_mtime(inode); + struct timespec64 now; + int updated = 0; + + now = inode_set_ctime_current(inode); + if (!timespec64_equal(&now, &ctime)) + updated |= S_CTIME; + + inode_set_mtime_to_ts(inode, attr->ia_mtime); + if (!timespec64_equal(&now, &mtime)) + updated |= S_MTIME; + + inode_maybe_inc_iversion(inode, updated); + cache_flags |= NFS_INO_INVALID_CTIME | NFS_INO_INVALID_MTIME; + } + if (attr->ia_valid & ATTR_ATIME_SET) { + inode_set_atime_to_ts(inode, attr->ia_atime); + cache_flags |= NFS_INO_INVALID_ATIME; + } + NFS_I(inode)->cache_validity &= ~cache_flags; +} + static void nfs_update_timestamps(struct inode *inode, unsigned int ia_valid) { enum file_time_flags time_flags = 0; @@ -701,14 +731,27 @@ nfs_setattr(struct mnt_idmap *idmap, struct dentry *dentry, if (nfs_have_delegated_mtime(inode) && attr->ia_valid & ATTR_MTIME) { spin_lock(&inode->i_lock); - nfs_update_timestamps(inode, attr->ia_valid); + if (attr->ia_valid & ATTR_MTIME_SET) { + nfs_set_timestamps_to_ts(inode, attr); + attr->ia_valid &= ~(ATTR_MTIME|ATTR_MTIME_SET| + ATTR_ATIME|ATTR_ATIME_SET); + } else { + nfs_update_timestamps(inode, attr->ia_valid); + attr->ia_valid &= ~(ATTR_MTIME|ATTR_ATIME); + } spin_unlock(&inode->i_lock); - attr->ia_valid &= ~(ATTR_MTIME | ATTR_ATIME); } else if (nfs_have_delegated_atime(inode) && attr->ia_valid & ATTR_ATIME && !(attr->ia_valid & ATTR_MTIME)) { - nfs_update_delegated_atime(inode); - attr->ia_valid &= ~ATTR_ATIME; + if (attr->ia_valid & ATTR_ATIME_SET) { + spin_lock(&inode->i_lock); + nfs_set_timestamps_to_ts(inode, attr); + spin_unlock(&inode->i_lock); + attr->ia_valid &= ~(ATTR_ATIME|ATTR_ATIME_SET); + } else { + nfs_update_delegated_atime(inode); + attr->ia_valid &= ~ATTR_ATIME; + } } /* Optimization: if the end result is no change, don't RPC */ @@ -2546,15 +2589,26 @@ EXPORT_SYMBOL_GPL(nfs_net_id); static int nfs_net_init(struct net *net) { struct nfs_net *nn = net_generic(net, nfs_net_id); + int err; nfs_clients_init(net); if (!rpc_proc_register(net, &nn->rpcstats)) { - nfs_clients_exit(net); - return -ENOMEM; + err = -ENOMEM; + goto err_proc_rpc; } - return nfs_fs_proc_net_init(net); + err = nfs_fs_proc_net_init(net); + if (err) + goto err_proc_nfs; + + return 0; + +err_proc_nfs: + rpc_proc_unregister(net, "nfs"); +err_proc_rpc: + nfs_clients_exit(net); + return err; } static void nfs_net_exit(struct net *net) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 6655e5f32ec6..69c2c10ee658 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -455,7 +455,6 @@ extern int nfs_wait_bit_killable(struct wait_bit_key *key, int mode); #if IS_ENABLED(CONFIG_NFS_LOCALIO) /* localio.c */ -extern void nfs_local_probe(struct nfs_client *); extern void nfs_local_probe_async(struct nfs_client *); extern void nfs_local_probe_async_work(struct work_struct *); extern struct nfsd_file *nfs_local_open_fh(struct nfs_client *, diff --git a/fs/nfs/localio.c b/fs/nfs/localio.c index 4ec952f9f47d..510d0a16cfe9 100644 --- a/fs/nfs/localio.c +++ b/fs/nfs/localio.c @@ -171,7 +171,7 @@ static bool nfs_server_uuid_is_local(struct nfs_client *clp) * - called after alloc_client and init_client (so cl_rpcclient exists) * - this function is idempotent, it can be called for old or new clients */ -void nfs_local_probe(struct nfs_client *clp) +static void nfs_local_probe(struct nfs_client *clp) { /* Disallow localio if disabled via sysfs or AUTH_SYS isn't used */ if (!localio_enabled || @@ -191,14 +191,16 @@ void nfs_local_probe(struct nfs_client *clp) nfs_localio_enable_client(clp); nfs_uuid_end(&clp->cl_uuid); } -EXPORT_SYMBOL_GPL(nfs_local_probe); void nfs_local_probe_async_work(struct work_struct *work) { struct nfs_client *clp = container_of(work, struct nfs_client, cl_local_probe_work); + if (!refcount_inc_not_zero(&clp->cl_count)) + return; nfs_local_probe(clp); + nfs_put_client(clp); } void nfs_local_probe_async(struct nfs_client *clp) @@ -207,14 +209,16 @@ void nfs_local_probe_async(struct nfs_client *clp) } EXPORT_SYMBOL_GPL(nfs_local_probe_async); -static inline struct nfsd_file *nfs_local_file_get(struct nfsd_file *nf) +static inline void nfs_local_file_put(struct nfsd_file *localio) { - return nfs_to->nfsd_file_get(nf); -} + /* nfs_to_nfsd_file_put_local() expects an __rcu pointer + * but we have a __kernel pointer. It is always safe + * to cast a __kernel pointer to an __rcu pointer + * because the cast only weakens what is known about the pointer. + */ + struct nfsd_file __rcu *nf = (struct nfsd_file __rcu*) localio; -static inline void nfs_local_file_put(struct nfsd_file *nf) -{ - nfs_to->nfsd_file_put(nf); + nfs_to_nfsd_file_put_local(&nf); } /* @@ -226,12 +230,13 @@ static inline void nfs_local_file_put(struct nfsd_file *nf) static struct nfsd_file * __nfs_local_open_fh(struct nfs_client *clp, const struct cred *cred, struct nfs_fh *fh, struct nfs_file_localio *nfl, + struct nfsd_file __rcu **pnf, const fmode_t mode) { struct nfsd_file *localio; localio = nfs_open_local_fh(&clp->cl_uuid, clp->cl_rpcclient, - cred, fh, nfl, mode); + cred, fh, nfl, pnf, mode); if (IS_ERR(localio)) { int status = PTR_ERR(localio); trace_nfs_local_open_fh(fh, mode, status); @@ -258,7 +263,7 @@ nfs_local_open_fh(struct nfs_client *clp, const struct cred *cred, struct nfs_fh *fh, struct nfs_file_localio *nfl, const fmode_t mode) { - struct nfsd_file *nf, *new, __rcu **pnf; + struct nfsd_file *nf, __rcu **pnf; if (!nfs_server_is_local(clp)) return NULL; @@ -270,29 +275,9 @@ nfs_local_open_fh(struct nfs_client *clp, const struct cred *cred, else pnf = &nfl->ro_file; - new = NULL; - rcu_read_lock(); - nf = rcu_dereference(*pnf); - if (!nf) { - rcu_read_unlock(); - new = __nfs_local_open_fh(clp, cred, fh, nfl, mode); - if (IS_ERR(new)) - return NULL; - rcu_read_lock(); - /* try to swap in the pointer */ - spin_lock(&clp->cl_uuid.lock); - nf = rcu_dereference_protected(*pnf, 1); - if (!nf) { - nf = new; - new = NULL; - rcu_assign_pointer(*pnf, nf); - } - spin_unlock(&clp->cl_uuid.lock); - } - nf = nfs_local_file_get(nf); - rcu_read_unlock(); - if (new) - nfs_to_nfsd_file_put_local(new); + nf = __nfs_local_open_fh(clp, cred, fh, nfl, pnf, mode); + if (IS_ERR(nf)) + return NULL; return nf; } EXPORT_SYMBOL_GPL(nfs_local_open_fh); diff --git a/fs/nfs/nfs42.h b/fs/nfs/nfs42.h index 0282d93c8bcc..aafd15a4afce 100644 --- a/fs/nfs/nfs42.h +++ b/fs/nfs/nfs42.h @@ -21,6 +21,7 @@ int nfs42_proc_allocate(struct file *, loff_t, loff_t); ssize_t nfs42_proc_copy(struct file *, loff_t, struct file *, loff_t, size_t, struct nl4_server *, nfs4_stateid *, bool); int nfs42_proc_deallocate(struct file *, loff_t, loff_t); +int nfs42_proc_zero_range(struct file *, loff_t, loff_t); loff_t nfs42_proc_llseek(struct file *, loff_t, int); int nfs42_proc_layoutstats_generic(struct nfs_server *, struct nfs42_layoutstat_data *); diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c index 5cf52ece96ac..01c01f45358b 100644 --- a/fs/nfs/nfs42proc.c +++ b/fs/nfs/nfs42proc.c @@ -146,7 +146,8 @@ int nfs42_proc_allocate(struct file *filep, loff_t offset, loff_t len) err = nfs42_proc_fallocate(&msg, filep, offset, len); if (err == -EOPNOTSUPP) - NFS_SERVER(inode)->caps &= ~NFS_CAP_ALLOCATE; + NFS_SERVER(inode)->caps &= ~(NFS_CAP_ALLOCATE | + NFS_CAP_ZERO_RANGE); inode_unlock(inode); return err; @@ -169,7 +170,31 @@ int nfs42_proc_deallocate(struct file *filep, loff_t offset, loff_t len) if (err == 0) truncate_pagecache_range(inode, offset, (offset + len) -1); if (err == -EOPNOTSUPP) - NFS_SERVER(inode)->caps &= ~NFS_CAP_DEALLOCATE; + NFS_SERVER(inode)->caps &= ~(NFS_CAP_DEALLOCATE | + NFS_CAP_ZERO_RANGE); + + inode_unlock(inode); + return err; +} + +int nfs42_proc_zero_range(struct file *filep, loff_t offset, loff_t len) +{ + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_ZERO_RANGE], + }; + struct inode *inode = file_inode(filep); + int err; + + if (!nfs_server_capable(inode, NFS_CAP_ZERO_RANGE)) + return -EOPNOTSUPP; + + inode_lock(inode); + + err = nfs42_proc_fallocate(&msg, filep, offset, len); + if (err == 0) + truncate_pagecache_range(inode, offset, (offset + len) -1); + if (err == -EOPNOTSUPP) + NFS_SERVER(inode)->caps &= ~NFS_CAP_ZERO_RANGE; inode_unlock(inode); return err; diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index b1b663468249..4cc915d5741d 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -174,6 +174,18 @@ decode_putfh_maxsz + \ decode_deallocate_maxsz + \ decode_getattr_maxsz) +#define NFS4_enc_zero_range_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ + encode_putfh_maxsz + \ + encode_deallocate_maxsz + \ + encode_allocate_maxsz + \ + encode_getattr_maxsz) +#define NFS4_dec_zero_range_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ + decode_putfh_maxsz + \ + decode_deallocate_maxsz + \ + decode_allocate_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_read_plus_sz (compound_encode_hdr_maxsz + \ encode_sequence_maxsz + \ encode_putfh_maxsz + \ @@ -649,6 +661,27 @@ static void nfs4_xdr_enc_deallocate(struct rpc_rqst *req, } /* + * Encode ZERO_RANGE request + */ +static void nfs4_xdr_enc_zero_range(struct rpc_rqst *req, + struct xdr_stream *xdr, + const void *data) +{ + const struct nfs42_falloc_args *args = data; + struct compound_hdr hdr = { + .minorversion = nfs4_xdr_minorversion(&args->seq_args), + }; + + encode_compound_hdr(xdr, req, &hdr); + encode_sequence(xdr, &args->seq_args, &hdr); + encode_putfh(xdr, args->falloc_fh, &hdr); + encode_deallocate(xdr, args, &hdr); + encode_allocate(xdr, args, &hdr); + encode_getfattr(xdr, args->falloc_bitmask, &hdr); + encode_nops(&hdr); +} + +/* * Encode READ_PLUS request */ static void nfs4_xdr_enc_read_plus(struct rpc_rqst *req, @@ -1511,6 +1544,37 @@ out: } /* + * Decode ZERO_RANGE request + */ +static int nfs4_xdr_dec_zero_range(struct rpc_rqst *rqstp, + struct xdr_stream *xdr, + void *data) +{ + struct nfs42_falloc_res *res = data; + struct compound_hdr hdr; + int status; + + status = decode_compound_hdr(xdr, &hdr); + if (status) + goto out; + status = decode_sequence(xdr, &res->seq_res, rqstp); + if (status) + goto out; + status = decode_putfh(xdr); + if (status) + goto out; + status = decode_deallocate(xdr, res); + if (status) + goto out; + status = decode_allocate(xdr, res); + if (status) + goto out; + decode_getfattr(xdr, res->falloc_fattr, res->falloc_server); +out: + return status; +} + +/* * Decode READ_PLUS request */ static int nfs4_xdr_dec_read_plus(struct rpc_rqst *rqstp, diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 7d383d29a995..d3ca91f60fc1 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -67,8 +67,7 @@ struct nfs4_minor_version_ops { void (*free_lock_state)(struct nfs_server *, struct nfs4_lock_state *); int (*test_and_free_expired)(struct nfs_server *, - const nfs4_stateid *, - const struct cred *); + nfs4_stateid *, const struct cred *); struct nfs_seqid * (*alloc_seqid)(struct nfs_seqid_counter *, gfp_t); void (*session_trunk)(struct rpc_clnt *clnt, diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c index 1cd9652f3c28..5e9d66f3466c 100644 --- a/fs/nfs/nfs4file.c +++ b/fs/nfs/nfs4file.c @@ -225,8 +225,14 @@ static long nfs42_fallocate(struct file *filep, int mode, loff_t offset, loff_t if (!S_ISREG(inode->i_mode)) return -EOPNOTSUPP; - if ((mode != 0) && (mode != (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE))) + switch (mode) { + case 0: + case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE: + case FALLOC_FL_ZERO_RANGE: + break; + default: return -EOPNOTSUPP; + } ret = inode_newsize_ok(inode, offset + len); if (ret < 0) @@ -234,6 +240,8 @@ static long nfs42_fallocate(struct file *filep, int mode, loff_t offset, loff_t if (mode & FALLOC_FL_PUNCH_HOLE) return nfs42_proc_deallocate(filep, offset, len); + else if (mode & FALLOC_FL_ZERO_RANGE) + return nfs42_proc_zero_range(filep, offset ,len); return nfs42_proc_allocate(filep, offset, len); } diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b1d2122bd5a7..341740fa293d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -105,7 +105,7 @@ static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, bool is_privileged); static int nfs41_test_stateid(struct nfs_server *, const nfs4_stateid *, const struct cred *); -static int nfs41_free_stateid(struct nfs_server *, const nfs4_stateid *, +static int nfs41_free_stateid(struct nfs_server *, nfs4_stateid *, const struct cred *, bool); #endif @@ -325,14 +325,14 @@ static void nfs4_bitmap_copy_adjust(__u32 *dst, const __u32 *src, if (nfs_have_delegated_mtime(inode)) { if (!(cache_validity & NFS_INO_INVALID_ATIME)) - dst[1] &= ~FATTR4_WORD1_TIME_ACCESS; + dst[1] &= ~(FATTR4_WORD1_TIME_ACCESS|FATTR4_WORD1_TIME_ACCESS_SET); if (!(cache_validity & NFS_INO_INVALID_MTIME)) - dst[1] &= ~FATTR4_WORD1_TIME_MODIFY; + dst[1] &= ~(FATTR4_WORD1_TIME_MODIFY|FATTR4_WORD1_TIME_MODIFY_SET); if (!(cache_validity & NFS_INO_INVALID_CTIME)) - dst[1] &= ~FATTR4_WORD1_TIME_METADATA; + dst[1] &= ~(FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY_SET); } else if (nfs_have_delegated_atime(inode)) { if (!(cache_validity & NFS_INO_INVALID_ATIME)) - dst[1] &= ~FATTR4_WORD1_TIME_ACCESS; + dst[1] &= ~(FATTR4_WORD1_TIME_ACCESS|FATTR4_WORD1_TIME_ACCESS_SET); } } @@ -2903,16 +2903,14 @@ static int nfs40_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st } static int nfs40_test_and_free_expired_stateid(struct nfs_server *server, - const nfs4_stateid *stateid, - const struct cred *cred) + nfs4_stateid *stateid, const struct cred *cred) { return -NFS4ERR_BAD_STATEID; } #if defined(CONFIG_NFS_V4_1) static int nfs41_test_and_free_expired_stateid(struct nfs_server *server, - const nfs4_stateid *stateid, - const struct cred *cred) + nfs4_stateid *stateid, const struct cred *cred) { int status; @@ -2921,6 +2919,7 @@ static int nfs41_test_and_free_expired_stateid(struct nfs_server *server, break; case NFS4_INVALID_STATEID_TYPE: case NFS4_SPECIAL_STATEID_TYPE: + case NFS4_FREED_STATEID_TYPE: return -NFS4ERR_BAD_STATEID; case NFS4_REVOKED_STATEID_TYPE: goto out_free; @@ -3976,8 +3975,9 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f FATTR4_WORD0_CASE_INSENSITIVE | FATTR4_WORD0_CASE_PRESERVING; if (minorversion) - bitmask[2] = FATTR4_WORD2_SUPPATTR_EXCLCREAT | - FATTR4_WORD2_OPEN_ARGUMENTS; + bitmask[2] = FATTR4_WORD2_SUPPATTR_EXCLCREAT; + if (minorversion > 1) + bitmask[2] |= FATTR4_WORD2_OPEN_ARGUMENTS; status = nfs4_call_sync(server->client, server, &msg, &args.seq_args, &res.seq_res, 0); if (status == 0) { @@ -5164,13 +5164,15 @@ static int nfs4_do_create(struct inode *dir, struct dentry *dentry, struct nfs4_ } static struct dentry *nfs4_do_mkdir(struct inode *dir, struct dentry *dentry, - struct nfs4_createdata *data) + struct nfs4_createdata *data, int *statusp) { - int status = nfs4_call_sync(NFS_SERVER(dir)->client, NFS_SERVER(dir), &data->msg, + struct dentry *ret; + + *statusp = nfs4_call_sync(NFS_SERVER(dir)->client, NFS_SERVER(dir), &data->msg, &data->arg.seq_args, &data->res.seq_res, 1); - if (status) - return ERR_PTR(status); + if (*statusp) + return NULL; spin_lock(&dir->i_lock); /* Creating a directory bumps nlink in the parent */ @@ -5179,7 +5181,11 @@ static struct dentry *nfs4_do_mkdir(struct inode *dir, struct dentry *dentry, data->res.fattr->time_start, NFS_INO_INVALID_DATA); spin_unlock(&dir->i_lock); - return nfs_add_or_obtain(dentry, data->res.fh, data->res.fattr); + ret = nfs_add_or_obtain(dentry, data->res.fh, data->res.fattr); + if (!IS_ERR(ret)) + return ret; + *statusp = PTR_ERR(ret); + return NULL; } static void nfs4_free_createdata(struct nfs4_createdata *data) @@ -5240,17 +5246,18 @@ static int nfs4_proc_symlink(struct inode *dir, struct dentry *dentry, static struct dentry *_nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr, - struct nfs4_label *label) + struct nfs4_label *label, int *statusp) { struct nfs4_createdata *data; - struct dentry *ret = ERR_PTR(-ENOMEM); + struct dentry *ret = NULL; + *statusp = -ENOMEM; data = nfs4_alloc_createdata(dir, &dentry->d_name, sattr, NF4DIR); if (data == NULL) goto out; data->arg.label = label; - ret = nfs4_do_mkdir(dir, dentry, data); + ret = nfs4_do_mkdir(dir, dentry, data, statusp); nfs4_free_createdata(data); out: @@ -5273,11 +5280,12 @@ static struct dentry *nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry, if (!(server->attr_bitmask[2] & FATTR4_WORD2_MODE_UMASK)) sattr->ia_mode &= ~current_umask(); do { - alias = _nfs4_proc_mkdir(dir, dentry, sattr, label); - err = PTR_ERR_OR_ZERO(alias); + alias = _nfs4_proc_mkdir(dir, dentry, sattr, label, &err); trace_nfs4_mkdir(dir, &dentry->d_name, err); - err = nfs4_handle_exception(NFS_SERVER(dir), err, - &exception); + if (err) + alias = ERR_PTR(nfs4_handle_exception(NFS_SERVER(dir), + err, + &exception)); } while (exception.retry); nfs4_label_release_security(label); @@ -6211,6 +6219,8 @@ static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen, struct nfs_server *server = NFS_SERVER(inode); int ret; + if (unlikely(NFS_FH(inode)->size == 0)) + return -ENODATA; if (!nfs4_server_supports_acls(server, type)) return -EOPNOTSUPP; ret = nfs_revalidate_inode(inode, NFS_INO_INVALID_CHANGE); @@ -6285,6 +6295,9 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, { struct nfs4_exception exception = { }; int err; + + if (unlikely(NFS_FH(inode)->size == 0)) + return -ENODATA; do { err = __nfs4_proc_set_acl(inode, buf, buflen, type); trace_nfs4_set_acl(inode, err); @@ -10611,7 +10624,7 @@ static const struct rpc_call_ops nfs41_free_stateid_ops = { * Note: this function is always asynchronous. */ static int nfs41_free_stateid(struct nfs_server *server, - const nfs4_stateid *stateid, + nfs4_stateid *stateid, const struct cred *cred, bool privileged) { @@ -10651,6 +10664,7 @@ static int nfs41_free_stateid(struct nfs_server *server, if (IS_ERR(task)) return PTR_ERR(task); rpc_put_task(task); + stateid->type = NFS4_FREED_STATEID_TYPE; return 0; } @@ -10817,6 +10831,7 @@ static const struct nfs4_minor_version_ops nfs_v4_2_minor_ops = { | NFS_CAP_OFFLOAD_CANCEL | NFS_CAP_COPY_NOTIFY | NFS_CAP_DEALLOCATE + | NFS_CAP_ZERO_RANGE | NFS_CAP_SEEK | NFS_CAP_LAYOUTSTATS | NFS_CAP_CLONE @@ -10852,7 +10867,7 @@ const struct nfs4_minor_version_ops *nfs_v4_minor_ops[] = { static ssize_t nfs4_listxattr(struct dentry *dentry, char *list, size_t size) { - ssize_t error, error2, error3; + ssize_t error, error2, error3, error4; size_t left = size; error = generic_listxattr(dentry, list, left); @@ -10875,8 +10890,16 @@ static ssize_t nfs4_listxattr(struct dentry *dentry, char *list, size_t size) error3 = nfs4_listxattr_nfs4_user(d_inode(dentry), list, left); if (error3 < 0) return error3; + if (list) { + list += error3; + left -= error3; + } + + error4 = security_inode_listsecurity(d_inode(dentry), list, left); + if (error4 < 0) + return error4; - error += error2 + error3; + error += error2 + error3 + error4; if (size && error > size) return -ERANGE; return error; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 55bef5fbfa47..318afde38057 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -7711,6 +7711,7 @@ const struct rpc_procinfo nfs4_procedures[] = { PROC42(LISTXATTRS, enc_listxattrs, dec_listxattrs), PROC42(REMOVEXATTR, enc_removexattr, dec_removexattr), PROC42(READ_PLUS, enc_read_plus, dec_read_plus), + PROC42(ZERO_RANGE, enc_zero_range, dec_zero_range), }; static unsigned int nfs_version4_counts[ARRAY_SIZE(nfs4_procedures)]; diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 3adb7d0dbec7..1a7ec68bde15 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -2059,8 +2059,10 @@ static void nfs_layoutget_begin(struct pnfs_layout_hdr *lo) static void nfs_layoutget_end(struct pnfs_layout_hdr *lo) { if (atomic_dec_and_test(&lo->plh_outstanding) && - test_and_clear_bit(NFS_LAYOUT_DRAIN, &lo->plh_flags)) + test_and_clear_bit(NFS_LAYOUT_DRAIN, &lo->plh_flags)) { + smp_mb__after_atomic(); wake_up_bit(&lo->plh_flags, NFS_LAYOUT_DRAIN); + } } static bool pnfs_is_first_layoutget(struct pnfs_layout_hdr *lo) diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c index 91ef486f40b9..b4ccdf78d4dd 100644 --- a/fs/nfs/pnfs_nfs.c +++ b/fs/nfs/pnfs_nfs.c @@ -830,10 +830,16 @@ static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv, .servername = clp->cl_hostname, .connect_timeout = connect_timeout, .reconnect_timeout = connect_timeout, + .xprtsec = clp->cl_xprtsec, }; - if (da->da_transport != clp->cl_proto) + if (da->da_transport != clp->cl_proto && + clp->cl_proto != XPRT_TRANSPORT_TCP_TLS) continue; + if (da->da_transport == XPRT_TRANSPORT_TCP && + mds_srv->nfs_client->cl_proto == XPRT_TRANSPORT_TCP_TLS) + xprt_args.ident = XPRT_TRANSPORT_TCP_TLS; + if (da->da_addr.ss_family != clp->cl_addr.ss_family) continue; /* Add this address as an alias */ @@ -841,6 +847,9 @@ static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv, rpc_clnt_test_and_add_xprt, NULL); continue; } + if (da->da_transport == XPRT_TRANSPORT_TCP && + mds_srv->nfs_client->cl_proto == XPRT_TRANSPORT_TCP_TLS) + da->da_transport = XPRT_TRANSPORT_TCP_TLS; clp = get_v3_ds_connect(mds_srv, &da->da_addr, da->da_addrlen, da->da_transport, diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 81bd1b9aba17..3c1fa320b3f1 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -56,7 +56,8 @@ static int nfs_return_empty_folio(struct folio *folio) { folio_zero_segment(folio, 0, folio_size(folio)); folio_mark_uptodate(folio); - folio_unlock(folio); + if (nfs_netfs_folio_unlock(folio)) + folio_unlock(folio); return 0; } diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 9eea9e62afc9..91b5503b6f74 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -1052,6 +1052,16 @@ int nfs_reconfigure(struct fs_context *fc) sync_filesystem(sb); /* + * The SB_RDONLY flag has been removed from the superblock during + * mounts to prevent interference between different filesystems. + * Similarly, it is also necessary to ignore the SB_RDONLY flag + * during reconfiguration; otherwise, it may also result in the + * creation of redundant superblocks when mounting a directory with + * different rw and ro flags multiple times. + */ + fc->sb_flags_mask &= ~SB_RDONLY; + + /* * Userspace mount programs that send binary options generally send * them populated with default values. We have no way to know which * ones were explicitly specified. Fall back to legacy behavior and @@ -1308,8 +1318,17 @@ int nfs_get_tree_common(struct fs_context *fc) if (IS_ERR(server)) return PTR_ERR(server); + /* + * When NFS_MOUNT_UNSHARED is not set, NFS forces the sharing of a + * superblock among each filesystem that mounts sub-directories + * belonging to a single exported root path. + * To prevent interference between different filesystems, the + * SB_RDONLY flag should be removed from the superblock. + */ if (server->flags & NFS_MOUNT_UNSHARED) compare_super = NULL; + else + fc->sb_flags &= ~SB_RDONLY; /* -o noac implies -o sync */ if (server->flags & NFS_MOUNT_NOAC) diff --git a/fs/nfs/sysfs.c b/fs/nfs/sysfs.c index 37cb2b776435..545148d42dcc 100644 --- a/fs/nfs/sysfs.c +++ b/fs/nfs/sysfs.c @@ -387,6 +387,33 @@ static inline void nfs_sysfs_add_nfsv41_server(struct nfs_server *server) } #endif /* CONFIG_NFS_V4_1 */ +#if IS_ENABLED(CONFIG_NFS_LOCALIO) + +static ssize_t +localio_show(struct kobject *kobj, struct kobj_attribute *attr, + char *buf) +{ + struct nfs_server *server = container_of(kobj, struct nfs_server, kobj); + bool localio = nfs_server_is_local(server->nfs_client); + return sysfs_emit(buf, "%d\n", localio); +} + +static struct kobj_attribute nfs_sysfs_attr_localio = __ATTR_RO(localio); + +static void nfs_sysfs_add_nfs_localio_server(struct nfs_server *server) +{ + int ret = sysfs_create_file_ns(&server->kobj, &nfs_sysfs_attr_localio.attr, + nfs_netns_server_namespace(&server->kobj)); + if (ret < 0) + pr_warn("NFS: sysfs_create_file_ns for server-%d failed (%d)\n", + server->s_sysfs_id, ret); +} +#else +static inline void nfs_sysfs_add_nfs_localio_server(struct nfs_server *server) +{ +} +#endif /* IS_ENABLED(CONFIG_NFS_LOCALIO) */ + void nfs_sysfs_add_server(struct nfs_server *server) { int ret; @@ -405,6 +432,7 @@ void nfs_sysfs_add_server(struct nfs_server *server) server->s_sysfs_id, ret); nfs_sysfs_add_nfsv41_server(server); + nfs_sysfs_add_nfs_localio_server(server); } EXPORT_SYMBOL_GPL(nfs_sysfs_add_server); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 23df8b214474..374fc6b34c79 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -632,19 +632,19 @@ static void nfs_write_error(struct nfs_page *req, int error) * Find an associated nfs write request, and prepare to flush it out * May return an error if the user signalled nfs_wait_on_request(). */ -static int nfs_page_async_flush(struct folio *folio, - struct writeback_control *wbc, - struct nfs_pageio_descriptor *pgio) +static int nfs_do_writepage(struct folio *folio, struct writeback_control *wbc, + struct nfs_pageio_descriptor *pgio) { struct nfs_page *req; - int ret = 0; + int ret; + + nfs_pageio_cond_complete(pgio, folio->index); req = nfs_lock_and_join_requests(folio); if (!req) - goto out; - ret = PTR_ERR(req); + return 0; if (IS_ERR(req)) - goto out; + return PTR_ERR(req); nfs_folio_set_writeback(folio); WARN_ON_ONCE(test_bit(PG_CLEAN, &req->wb_flags)); @@ -654,7 +654,6 @@ static int nfs_page_async_flush(struct folio *folio, if (nfs_error_is_fatal_on_server(ret)) goto out_launder; - ret = 0; if (!nfs_pageio_add_request(pgio, req)) { ret = pgio->pg_error; /* @@ -662,28 +661,20 @@ static int nfs_page_async_flush(struct folio *folio, */ if (nfs_error_is_fatal_on_server(ret)) goto out_launder; - if (wbc->sync_mode == WB_SYNC_NONE) - ret = AOP_WRITEPAGE_ACTIVATE; folio_redirty_for_writepage(wbc, folio); nfs_redirty_request(req); pgio->pg_error = 0; - } else - nfs_add_stats(folio->mapping->host, - NFSIOS_WRITEPAGES, 1); -out: - return ret; + return ret; + } + + nfs_add_stats(folio->mapping->host, NFSIOS_WRITEPAGES, 1); + return 0; + out_launder: nfs_write_error(req, ret); return 0; } -static int nfs_do_writepage(struct folio *folio, struct writeback_control *wbc, - struct nfs_pageio_descriptor *pgio) -{ - nfs_pageio_cond_complete(pgio, folio->index); - return nfs_page_async_flush(folio, wbc, pgio); -} - /* * Write an mmapped page to the server. */ @@ -703,17 +694,6 @@ static int nfs_writepage_locked(struct folio *folio, return err; } -static int nfs_writepages_callback(struct folio *folio, - struct writeback_control *wbc, void *data) -{ - int ret; - - ret = nfs_do_writepage(folio, wbc, data); - if (ret != AOP_WRITEPAGE_ACTIVATE) - folio_unlock(folio); - return ret; -} - static void nfs_io_completion_commit(void *inode) { nfs_commit_inode(inode, 0); @@ -749,11 +729,15 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) } do { + struct folio *folio = NULL; + nfs_pageio_init_write(&pgio, inode, priority, false, &nfs_async_write_completion_ops); pgio.pg_io_completion = ioc; - err = write_cache_pages(mapping, wbc, nfs_writepages_callback, - &pgio); + while ((folio = writeback_iter(mapping, wbc, folio, &err))) { + err = nfs_do_writepage(folio, wbc, &pgio); + folio_unlock(folio); + } pgio.pg_error = 0; nfs_pageio_complete(&pgio); if (err == -EAGAIN && mntflags & NFS_MOUNT_SOFTERR) diff --git a/fs/nfs_common/nfslocalio.c b/fs/nfs_common/nfslocalio.c index 6a0bdea6d644..05c7c16e37ab 100644 --- a/fs/nfs_common/nfslocalio.c +++ b/fs/nfs_common/nfslocalio.c @@ -151,8 +151,7 @@ EXPORT_SYMBOL_GPL(nfs_localio_enable_client); */ static bool nfs_uuid_put(nfs_uuid_t *nfs_uuid) { - LIST_HEAD(local_files); - struct nfs_file_localio *nfl, *tmp; + struct nfs_file_localio *nfl; spin_lock(&nfs_uuid->lock); if (unlikely(!rcu_access_pointer(nfs_uuid->net))) { @@ -166,17 +165,42 @@ static bool nfs_uuid_put(nfs_uuid_t *nfs_uuid) nfs_uuid->dom = NULL; } - list_splice_init(&nfs_uuid->files, &local_files); - spin_unlock(&nfs_uuid->lock); - /* Walk list of files and ensure their last references dropped */ - list_for_each_entry_safe(nfl, tmp, &local_files, list) { - nfs_close_local_fh(nfl); + + while ((nfl = list_first_entry_or_null(&nfs_uuid->files, + struct nfs_file_localio, + list)) != NULL) { + /* If nfs_uuid is already NULL, nfs_close_local_fh is + * closing and we must wait, else we unlink and close. + */ + if (rcu_access_pointer(nfl->nfs_uuid) == NULL) { + /* nfs_close_local_fh() is doing the + * close and we must wait. until it unlinks + */ + wait_var_event_spinlock(nfl, + list_first_entry_or_null( + &nfs_uuid->files, + struct nfs_file_localio, + list) != nfl, + &nfs_uuid->lock); + continue; + } + + /* Remove nfl from nfs_uuid->files list */ + list_del_init(&nfl->list); + spin_unlock(&nfs_uuid->lock); + + nfs_to_nfsd_file_put_local(&nfl->ro_file); + nfs_to_nfsd_file_put_local(&nfl->rw_file); cond_resched(); - } - spin_lock(&nfs_uuid->lock); - BUG_ON(!list_empty(&nfs_uuid->files)); + spin_lock(&nfs_uuid->lock); + /* Now we can allow racing nfs_close_local_fh() to + * skip the locking. + */ + RCU_INIT_POINTER(nfl->nfs_uuid, NULL); + wake_up_var_locked(&nfl->nfs_uuid, &nfs_uuid->lock); + } /* Remove client from nn->local_clients */ if (nfs_uuid->list_lock) { @@ -237,6 +261,7 @@ static void nfs_uuid_add_file(nfs_uuid_t *nfs_uuid, struct nfs_file_localio *nfl struct nfsd_file *nfs_open_local_fh(nfs_uuid_t *uuid, struct rpc_clnt *rpc_clnt, const struct cred *cred, const struct nfs_fh *nfs_fh, struct nfs_file_localio *nfl, + struct nfsd_file __rcu **pnf, const fmode_t fmode) { struct net *net; @@ -261,10 +286,9 @@ struct nfsd_file *nfs_open_local_fh(nfs_uuid_t *uuid, rcu_read_unlock(); /* We have an implied reference to net thanks to nfsd_net_try_get */ localio = nfs_to->nfsd_open_local_fh(net, uuid->dom, rpc_clnt, - cred, nfs_fh, fmode); - if (IS_ERR(localio)) - nfs_to_nfsd_net_put(net); - else + cred, nfs_fh, pnf, fmode); + nfs_to_nfsd_net_put(net); + if (!IS_ERR(localio)) nfs_uuid_add_file(uuid, nfl); return localio; @@ -273,8 +297,6 @@ EXPORT_SYMBOL_GPL(nfs_open_local_fh); void nfs_close_local_fh(struct nfs_file_localio *nfl) { - struct nfsd_file *ro_nf = NULL; - struct nfsd_file *rw_nf = NULL; nfs_uuid_t *nfs_uuid; rcu_read_lock(); @@ -285,28 +307,39 @@ void nfs_close_local_fh(struct nfs_file_localio *nfl) return; } - ro_nf = rcu_access_pointer(nfl->ro_file); - rw_nf = rcu_access_pointer(nfl->rw_file); - if (ro_nf || rw_nf) { - spin_lock(&nfs_uuid->lock); - if (ro_nf) - ro_nf = rcu_dereference_protected(xchg(&nfl->ro_file, NULL), 1); - if (rw_nf) - rw_nf = rcu_dereference_protected(xchg(&nfl->rw_file, NULL), 1); - - /* Remove nfl from nfs_uuid->files list */ - RCU_INIT_POINTER(nfl->nfs_uuid, NULL); - list_del_init(&nfl->list); + spin_lock(&nfs_uuid->lock); + if (!rcu_access_pointer(nfl->nfs_uuid)) { + /* nfs_uuid_put has finished here */ spin_unlock(&nfs_uuid->lock); rcu_read_unlock(); - - if (ro_nf) - nfs_to_nfsd_file_put_local(ro_nf); - if (rw_nf) - nfs_to_nfsd_file_put_local(rw_nf); return; } + if (list_empty(&nfs_uuid->files)) { + /* nfs_uuid_put() has started closing files, wait for it + * to finished + */ + spin_unlock(&nfs_uuid->lock); + rcu_read_unlock(); + wait_var_event(&nfl->nfs_uuid, + rcu_access_pointer(nfl->nfs_uuid) == NULL); + return; + } + /* tell nfs_uuid_put() to wait for us */ + RCU_INIT_POINTER(nfl->nfs_uuid, NULL); + spin_unlock(&nfs_uuid->lock); rcu_read_unlock(); + + nfs_to_nfsd_file_put_local(&nfl->ro_file); + nfs_to_nfsd_file_put_local(&nfl->rw_file); + + /* Remove nfl from nfs_uuid->files list and signal nfs_uuid_put() + * that we are done. The moment we drop the spinlock the + * nfs_uuid could be freed. + */ + spin_lock(&nfs_uuid->lock); + list_del_init(&nfl->list); + wake_up_var_locked(&nfl->nfs_uuid, &nfs_uuid->lock); + spin_unlock(&nfs_uuid->lock); } EXPORT_SYMBOL_GPL(nfs_close_local_fh); diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c index ab85e6a2454f..e108b6c705b4 100644 --- a/fs/nfsd/filecache.c +++ b/fs/nfsd/filecache.c @@ -378,15 +378,41 @@ nfsd_file_put(struct nfsd_file *nf) * the reference of the nfsd_file. */ struct net * -nfsd_file_put_local(struct nfsd_file *nf) +nfsd_file_put_local(struct nfsd_file __rcu **pnf) { - struct net *net = nf->nf_net; + struct nfsd_file *nf; + struct net *net = NULL; - nfsd_file_put(nf); + nf = unrcu_pointer(xchg(pnf, NULL)); + if (nf) { + net = nf->nf_net; + nfsd_file_put(nf); + } return net; } /** + * nfsd_file_get_local - get nfsd_file reference and reference to net + * @nf: nfsd_file of which to put the reference + * + * Get reference to both the nfsd_file and nf->nf_net. + */ +struct nfsd_file * +nfsd_file_get_local(struct nfsd_file *nf) +{ + struct net *net = nf->nf_net; + + if (nfsd_net_try_get(net)) { + nf = nfsd_file_get(nf); + if (!nf) + nfsd_net_put(net); + } else { + nf = NULL; + } + return nf; +} + +/** * nfsd_file_file - get the backing file of an nfsd_file * @nf: nfsd_file of which to access the backing file. * diff --git a/fs/nfsd/filecache.h b/fs/nfsd/filecache.h index 5865f9c72712..722b26c71e45 100644 --- a/fs/nfsd/filecache.h +++ b/fs/nfsd/filecache.h @@ -62,7 +62,8 @@ void nfsd_file_cache_shutdown(void); int nfsd_file_cache_start_net(struct net *net); void nfsd_file_cache_shutdown_net(struct net *net); void nfsd_file_put(struct nfsd_file *nf); -struct net *nfsd_file_put_local(struct nfsd_file *nf); +struct net *nfsd_file_put_local(struct nfsd_file __rcu **nf); +struct nfsd_file *nfsd_file_get_local(struct nfsd_file *nf); struct nfsd_file *nfsd_file_get(struct nfsd_file *nf); struct file *nfsd_file_file(struct nfsd_file *nf); void nfsd_file_close_inode_sync(struct inode *inode); diff --git a/fs/nfsd/localio.c b/fs/nfsd/localio.c index 238647fa379e..80d9ff6608a7 100644 --- a/fs/nfsd/localio.c +++ b/fs/nfsd/localio.c @@ -24,21 +24,6 @@ #include "filecache.h" #include "cache.h" -static const struct nfsd_localio_operations nfsd_localio_ops = { - .nfsd_net_try_get = nfsd_net_try_get, - .nfsd_net_put = nfsd_net_put, - .nfsd_open_local_fh = nfsd_open_local_fh, - .nfsd_file_put_local = nfsd_file_put_local, - .nfsd_file_get = nfsd_file_get, - .nfsd_file_put = nfsd_file_put, - .nfsd_file_file = nfsd_file_file, -}; - -void nfsd_localio_ops_init(void) -{ - nfs_to = &nfsd_localio_ops; -} - /** * nfsd_open_local_fh - lookup a local filehandle @nfs_fh and map to nfsd_file * @@ -47,6 +32,7 @@ void nfsd_localio_ops_init(void) * @rpc_clnt: rpc_clnt that the client established * @cred: cred that the client established * @nfs_fh: filehandle to lookup + * @nfp: place to find the nfsd_file, or store it if it was non-NULL * @fmode: fmode_t to use for open * * This function maps a local fh to a path on a local filesystem. @@ -57,10 +43,11 @@ void nfsd_localio_ops_init(void) * set. Caller (NFS client) is responsible for calling nfsd_net_put and * nfsd_file_put (via nfs_to_nfsd_file_put_local). */ -struct nfsd_file * +static struct nfsd_file * nfsd_open_local_fh(struct net *net, struct auth_domain *dom, struct rpc_clnt *rpc_clnt, const struct cred *cred, - const struct nfs_fh *nfs_fh, const fmode_t fmode) + const struct nfs_fh *nfs_fh, struct nfsd_file __rcu **pnf, + const fmode_t fmode) { int mayflags = NFSD_MAY_LOCALIO; struct svc_cred rq_cred; @@ -71,6 +58,15 @@ nfsd_open_local_fh(struct net *net, struct auth_domain *dom, if (nfs_fh->size > NFS4_FHSIZE) return ERR_PTR(-EINVAL); + if (!nfsd_net_try_get(net)) + return ERR_PTR(-ENXIO); + + rcu_read_lock(); + localio = nfsd_file_get(rcu_dereference(*pnf)); + rcu_read_unlock(); + if (localio) + return localio; + /* nfs_fh -> svc_fh */ fh_init(&fh, NFS4_FHSIZE); fh.fh_handle.fh_size = nfs_fh->size; @@ -92,9 +88,47 @@ nfsd_open_local_fh(struct net *net, struct auth_domain *dom, if (rq_cred.cr_group_info) put_group_info(rq_cred.cr_group_info); + if (!IS_ERR(localio)) { + struct nfsd_file *new; + if (!nfsd_net_try_get(net)) { + nfsd_file_put(localio); + nfsd_net_put(net); + return ERR_PTR(-ENXIO); + } + nfsd_file_get(localio); + again: + new = unrcu_pointer(cmpxchg(pnf, NULL, RCU_INITIALIZER(localio))); + if (new) { + /* Some other thread installed an nfsd_file */ + if (nfsd_file_get(new) == NULL) + goto again; + /* + * Drop the ref we were going to install and the + * one we were going to return. + */ + nfsd_file_put(localio); + nfsd_file_put(localio); + localio = new; + } + } else + nfsd_net_put(net); + return localio; } -EXPORT_SYMBOL_GPL(nfsd_open_local_fh); + +static const struct nfsd_localio_operations nfsd_localio_ops = { + .nfsd_net_try_get = nfsd_net_try_get, + .nfsd_net_put = nfsd_net_put, + .nfsd_open_local_fh = nfsd_open_local_fh, + .nfsd_file_put_local = nfsd_file_put_local, + .nfsd_file_get_local = nfsd_file_get_local, + .nfsd_file_file = nfsd_file_file, +}; + +void nfsd_localio_ops_init(void) +{ + nfs_to = &nfsd_localio_ops; +} /* * UUID_IS_LOCAL XDR functions diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index ccb00aa93be0..e00b2aea8da2 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -1409,6 +1409,7 @@ void nfsd41_cb_referring_call(struct nfsd4_callback *cb, out: if (!rcl->__nr_referring_calls) { cb->cb_nr_referring_call_list--; + list_del(&rcl->__list); kfree(rcl); } } diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 3f3e9f6c4250..6a42cc7a845a 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1611,7 +1611,7 @@ out_unlock: */ int nfsd_nl_threads_set_doit(struct sk_buff *skb, struct genl_info *info) { - int *nthreads, count = 0, nrpools, i, ret = -EOPNOTSUPP, rem; + int *nthreads, nrpools = 0, i, ret = -EOPNOTSUPP, rem; struct net *net = genl_info_net(info); struct nfsd_net *nn = net_generic(net, nfsd_net_id); const struct nlattr *attr; @@ -1623,12 +1623,11 @@ int nfsd_nl_threads_set_doit(struct sk_buff *skb, struct genl_info *info) /* count number of SERVER_THREADS values */ nlmsg_for_each_attr(attr, info->nlhdr, GENL_HDRLEN, rem) { if (nla_type(attr) == NFSD_A_SERVER_THREADS) - count++; + nrpools++; } mutex_lock(&nfsd_mutex); - nrpools = max(count, nfsd_nrpools(net)); nthreads = kcalloc(nrpools, sizeof(int), GFP_KERNEL); if (!nthreads) { ret = -ENOMEM; diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index 6613b8fcceb0..5cf7328d5360 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -472,11 +472,18 @@ static int __nilfs_read_inode(struct super_block *sb, inode->i_op = &nilfs_symlink_inode_operations; inode_nohighmem(inode); inode->i_mapping->a_ops = &nilfs_aops; - } else { + } else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode) || + S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) { inode->i_op = &nilfs_special_inode_operations; init_special_inode( inode, inode->i_mode, huge_decode_dev(le64_to_cpu(raw_inode->i_device_code))); + } else { + nilfs_error(sb, + "invalid file type bits in mode 0%o for inode %lu", + inode->i_mode, ino); + err = -EIO; + goto failed_unmap; } nilfs_ifile_unmap_inode(raw_inode); brelse(bh); diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 61a4141f8d6b..f15ca6fc400d 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -2485,7 +2485,7 @@ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode) static void nilfs_construction_timeout(struct timer_list *t) { - struct nilfs_sc_info *sci = from_timer(sci, t, sc_timer); + struct nilfs_sc_info *sci = timer_container_of(sci, t, sc_timer); wake_up_process(sci->sc_task); } diff --git a/fs/notify/dnotify/dnotify.c b/fs/notify/dnotify/dnotify.c index c4cdaf5fa7ed..9fb73bafd41d 100644 --- a/fs/notify/dnotify/dnotify.c +++ b/fs/notify/dnotify/dnotify.c @@ -308,6 +308,10 @@ int fcntl_dirnotify(int fd, struct file *filp, unsigned int arg) goto out_err; } + error = file_f_owner_allocate(filp); + if (error) + goto out_err; + /* new fsnotify mark, we expect most fcntl calls to add a new mark */ new_dn_mark = kmem_cache_alloc(dnotify_mark_cache, GFP_KERNEL); if (!new_dn_mark) { @@ -315,10 +319,6 @@ int fcntl_dirnotify(int fd, struct file *filp, unsigned int arg) goto out_err; } - error = file_f_owner_allocate(filp); - if (error) - goto out_err; - /* set up the new_fsn_mark and new_dn_mark */ new_fsn_mark = &new_dn_mark->fsn_mark; fsnotify_init_mark(new_fsn_mark, dnotify_group); diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 43e652a2adaf..b05d4e9d13b2 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -1488,7 +1488,8 @@ static void o2net_sc_send_keep_req(struct work_struct *work) * where shutdown is going to be involved */ static void o2net_idle_timer(struct timer_list *t) { - struct o2net_sock_container *sc = from_timer(sc, t, sc_idle_timeout); + struct o2net_sock_container *sc = timer_container_of(sc, t, + sc_idle_timeout); struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num); #ifdef CONFIG_DEBUG_FS unsigned long msecs = ktime_to_ms(ktime_get()) - diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c index 969b458100fe..dfea7bd800cb 100644 --- a/fs/overlayfs/file.c +++ b/fs/overlayfs/file.c @@ -48,8 +48,8 @@ static struct file *ovl_open_realfile(const struct file *file, if (!inode_owner_or_capable(real_idmap, realinode)) flags &= ~O_NOATIME; - realfile = backing_file_open(&file->f_path, flags, realpath, - current_cred()); + realfile = backing_file_open(file_user_path((struct file *) file), + flags, realpath, current_cred()); } ovl_revert_creds(old_cred); diff --git a/fs/overlayfs/namei.c b/fs/overlayfs/namei.c index bf722daf19a9..2043f0369059 100644 --- a/fs/overlayfs/namei.c +++ b/fs/overlayfs/namei.c @@ -16,6 +16,7 @@ struct ovl_lookup_data { struct super_block *sb; + struct dentry *dentry; const struct ovl_layer *layer; struct qstr name; bool is_dir; @@ -24,6 +25,7 @@ struct ovl_lookup_data { bool stop; bool last; char *redirect; + char *upperredirect; int metacopy; /* Referring to last redirect xattr */ bool absolute_redirect; @@ -1024,6 +1026,31 @@ int ovl_verify_lowerdata(struct dentry *dentry) return ovl_maybe_validate_verity(dentry); } +/* + * Following redirects/metacopy can have security consequences: it's like a + * symlink into the lower layer without the permission checks. + * + * This is only a problem if the upper layer is untrusted (e.g comes from an USB + * drive). This can allow a non-readable file or directory to become readable. + * + * Only following redirects when redirects are enabled disables this attack + * vector when not necessary. + */ +static bool ovl_check_follow_redirect(struct ovl_lookup_data *d) +{ + struct ovl_fs *ofs = OVL_FS(d->sb); + + if (d->metacopy && !ofs->config.metacopy) { + pr_warn_ratelimited("refusing to follow metacopy origin for (%pd2)\n", d->dentry); + return false; + } + if ((d->redirect || d->upperredirect) && !ovl_redirect_follow(ofs)) { + pr_warn_ratelimited("refusing to follow redirect for (%pd2)\n", d->dentry); + return false; + } + return true; +} + struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) { @@ -1039,7 +1066,7 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, unsigned int ctr = 0; struct inode *inode = NULL; bool upperopaque = false; - char *upperredirect = NULL; + bool check_redirect = (ovl_redirect_follow(ofs) || ofs->numdatalayer); struct dentry *this; unsigned int i; int err; @@ -1047,12 +1074,14 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, int metacopy_size = 0; struct ovl_lookup_data d = { .sb = dentry->d_sb, + .dentry = dentry, .name = dentry->d_name, .is_dir = false, .opaque = false, .stop = false, - .last = ovl_redirect_follow(ofs) ? false : !ovl_numlower(poe), + .last = check_redirect ? false : !ovl_numlower(poe), .redirect = NULL, + .upperredirect = NULL, .metacopy = 0, }; @@ -1094,8 +1123,8 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, if (d.redirect) { err = -ENOMEM; - upperredirect = kstrdup(d.redirect, GFP_KERNEL); - if (!upperredirect) + d.upperredirect = kstrdup(d.redirect, GFP_KERNEL); + if (!d.upperredirect) goto out_put_upper; if (d.redirect[0] == '/') poe = roe; @@ -1113,7 +1142,12 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, for (i = 0; !d.stop && i < ovl_numlower(poe); i++) { struct ovl_path lower = ovl_lowerstack(poe)[i]; - if (!ovl_redirect_follow(ofs)) + if (!ovl_check_follow_redirect(&d)) { + err = -EPERM; + goto out_put; + } + + if (!check_redirect) d.last = i == ovl_numlower(poe) - 1; else if (d.is_dir || !ofs->numdatalayer) d.last = lower.layer->idx == ovl_numlower(roe); @@ -1126,13 +1160,6 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, if (!this) continue; - if ((uppermetacopy || d.metacopy) && !ofs->config.metacopy) { - dput(this); - err = -EPERM; - pr_warn_ratelimited("refusing to follow metacopy origin for (%pd2)\n", dentry); - goto out_put; - } - /* * If no origin fh is stored in upper of a merge dir, store fh * of lower dir and set upper parent "impure". @@ -1185,23 +1212,6 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, ctr++; } - /* - * Following redirects can have security consequences: it's like - * a symlink into the lower layer without the permission checks. - * This is only a problem if the upper layer is untrusted (e.g - * comes from an USB drive). This can allow a non-readable file - * or directory to become readable. - * - * Only following redirects when redirects are enabled disables - * this attack vector when not necessary. - */ - err = -EPERM; - if (d.redirect && !ovl_redirect_follow(ofs)) { - pr_warn_ratelimited("refusing to follow redirect for (%pd2)\n", - dentry); - goto out_put; - } - if (d.stop) break; @@ -1212,10 +1222,16 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, } } - /* Defer lookup of lowerdata in data-only layers to first access */ + /* + * Defer lookup of lowerdata in data-only layers to first access. + * Don't require redirect=follow and metacopy=on in this case. + */ if (d.metacopy && ctr && ofs->numdatalayer && d.absolute_redirect) { d.metacopy = 0; ctr++; + } else if (!ovl_check_follow_redirect(&d)) { + err = -EPERM; + goto out_put; } /* @@ -1298,20 +1314,26 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, /* * It's safe to assign upperredirect here: the previous - * assignment of happens only if upperdentry is non-NULL, and + * assignment happens only if upperdentry is non-NULL, and * this one only if upperdentry is NULL. */ - upperredirect = ovl_get_redirect_xattr(ofs, &upperpath, 0); - if (IS_ERR(upperredirect)) { - err = PTR_ERR(upperredirect); - upperredirect = NULL; + d.upperredirect = ovl_get_redirect_xattr(ofs, &upperpath, 0); + if (IS_ERR(d.upperredirect)) { + err = PTR_ERR(d.upperredirect); + d.upperredirect = NULL; goto out_free_oe; } + err = ovl_check_metacopy_xattr(ofs, &upperpath, NULL); if (err < 0) goto out_free_oe; - uppermetacopy = err; + d.metacopy = uppermetacopy = err; metacopy_size = err; + + if (!ovl_check_follow_redirect(&d)) { + err = -EPERM; + goto out_free_oe; + } } if (upperdentry || ctr) { @@ -1319,7 +1341,7 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, .upperdentry = upperdentry, .oe = oe, .index = index, - .redirect = upperredirect, + .redirect = d.upperredirect, }; /* Store lowerdata redirect for lazy lookup */ @@ -1361,7 +1383,7 @@ out_put_upper: kfree(origin_path); } dput(upperdentry); - kfree(upperredirect); + kfree(d.upperredirect); out: kfree(d.redirect); ovl_revert_creds(old_cred); @@ -1371,7 +1393,7 @@ out: bool ovl_lower_positive(struct dentry *dentry) { struct ovl_entry *poe = OVL_E(dentry->d_parent); - struct qstr *name = &dentry->d_name; + const struct qstr *name = &dentry->d_name; const struct cred *old_cred; unsigned int i; bool positive = false; @@ -1394,9 +1416,15 @@ bool ovl_lower_positive(struct dentry *dentry) struct dentry *this; struct ovl_path *parentpath = &ovl_lowerstack(poe)[i]; + /* + * We need to make a non-const copy of dentry->d_name, + * because lookup_one_positive_unlocked() will hash name + * with parentpath base, which is on another (lower fs). + */ this = lookup_one_positive_unlocked( mnt_idmap(parentpath->layer->mnt), - name, parentpath->dentry); + &QSTR_LEN(name->name, name->len), + parentpath->dentry); if (IS_ERR(this)) { switch (PTR_ERR(this)) { case -ENOENT: diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h index 8baaba0a3fe5..497323128e5f 100644 --- a/fs/overlayfs/overlayfs.h +++ b/fs/overlayfs/overlayfs.h @@ -246,9 +246,11 @@ static inline struct dentry *ovl_do_mkdir(struct ovl_fs *ofs, struct dentry *dentry, umode_t mode) { - dentry = vfs_mkdir(ovl_upper_mnt_idmap(ofs), dir, dentry, mode); - pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, PTR_ERR_OR_ZERO(dentry)); - return dentry; + struct dentry *ret; + + ret = vfs_mkdir(ovl_upper_mnt_idmap(ofs), dir, dentry, mode); + pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, PTR_ERR_OR_ZERO(ret)); + return ret; } static inline int ovl_do_mknod(struct ovl_fs *ofs, diff --git a/fs/overlayfs/ovl_entry.h b/fs/overlayfs/ovl_entry.h index cb449ab310a7..afb7762f873f 100644 --- a/fs/overlayfs/ovl_entry.h +++ b/fs/overlayfs/ovl_entry.h @@ -51,7 +51,7 @@ struct ovl_path { struct ovl_entry { unsigned int __numlower; - struct ovl_path __lowerstack[]; + struct ovl_path __lowerstack[] __counted_by(__numlower); }; /* private information held for overlayfs's superblock */ diff --git a/fs/overlayfs/params.c b/fs/overlayfs/params.c index 6759f7d040c8..f42488c01957 100644 --- a/fs/overlayfs/params.c +++ b/fs/overlayfs/params.c @@ -871,18 +871,6 @@ int ovl_fs_params_verify(const struct ovl_fs_context *ctx, config->uuid = OVL_UUID_NULL; } - /* Resolve verity -> metacopy dependency */ - if (config->verity_mode && !config->metacopy) { - /* Don't allow explicit specified conflicting combinations */ - if (set.metacopy) { - pr_err("conflicting options: metacopy=off,verity=%s\n", - ovl_verity_mode(config)); - return -EINVAL; - } - /* Otherwise automatically enable metacopy. */ - config->metacopy = true; - } - /* * This is to make the logic below simpler. It doesn't make any other * difference, since redirect_dir=on is only used for upper. @@ -890,18 +878,13 @@ int ovl_fs_params_verify(const struct ovl_fs_context *ctx, if (!config->upperdir && config->redirect_mode == OVL_REDIRECT_FOLLOW) config->redirect_mode = OVL_REDIRECT_ON; - /* Resolve verity -> metacopy -> redirect_dir dependency */ + /* metacopy -> redirect_dir dependency */ if (config->metacopy && config->redirect_mode != OVL_REDIRECT_ON) { if (set.metacopy && set.redirect) { pr_err("conflicting options: metacopy=on,redirect_dir=%s\n", ovl_redirect_mode(config)); return -EINVAL; } - if (config->verity_mode && set.redirect) { - pr_err("conflicting options: verity=%s,redirect_dir=%s\n", - ovl_verity_mode(config), ovl_redirect_mode(config)); - return -EINVAL; - } if (set.redirect) { /* * There was an explicit redirect_dir=... that resulted @@ -970,7 +953,7 @@ int ovl_fs_params_verify(const struct ovl_fs_context *ctx, } - /* Resolve userxattr -> !redirect && !metacopy && !verity dependency */ + /* Resolve userxattr -> !redirect && !metacopy dependency */ if (config->userxattr) { if (set.redirect && config->redirect_mode != OVL_REDIRECT_NOFOLLOW) { @@ -982,11 +965,6 @@ int ovl_fs_params_verify(const struct ovl_fs_context *ctx, pr_err("conflicting options: userxattr,metacopy=on\n"); return -EINVAL; } - if (config->verity_mode) { - pr_err("conflicting options: userxattr,verity=%s\n", - ovl_verity_mode(config)); - return -EINVAL; - } /* * Silently disable default setting of redirect and metacopy. * This shall be the default in the future as well: these @@ -1025,11 +1003,6 @@ int ovl_fs_params_verify(const struct ovl_fs_context *ctx, */ } - if (ctx->nr_data > 0 && !config->metacopy) { - pr_err("lower data-only dirs require metacopy support.\n"); - return -EINVAL; - } - return 0; } @@ -1078,17 +1051,16 @@ int ovl_show_options(struct seq_file *m, struct dentry *dentry) seq_printf(m, ",redirect_dir=%s", ovl_redirect_mode(&ofs->config)); if (ofs->config.index != ovl_index_def) - seq_printf(m, ",index=%s", ofs->config.index ? "on" : "off"); + seq_printf(m, ",index=%s", str_on_off(ofs->config.index)); if (ofs->config.uuid != ovl_uuid_def()) seq_printf(m, ",uuid=%s", ovl_uuid_mode(&ofs->config)); if (ofs->config.nfs_export != ovl_nfs_export_def) - seq_printf(m, ",nfs_export=%s", ofs->config.nfs_export ? - "on" : "off"); + seq_printf(m, ",nfs_export=%s", + str_on_off(ofs->config.nfs_export)); if (ofs->config.xino != ovl_xino_def() && !ovl_same_fs(ofs)) seq_printf(m, ",xino=%s", ovl_xino_mode(&ofs->config)); if (ofs->config.metacopy != ovl_metacopy_def) - seq_printf(m, ",metacopy=%s", - ofs->config.metacopy ? "on" : "off"); + seq_printf(m, ",metacopy=%s", str_on_off(ofs->config.metacopy)); if (ofs->config.ovl_volatile) seq_puts(m, ",volatile"); if (ofs->config.userxattr) diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c index 44e208da417c..474c80d210d1 100644 --- a/fs/overlayfs/readdir.c +++ b/fs/overlayfs/readdir.c @@ -13,6 +13,7 @@ #include <linux/security.h> #include <linux/cred.h> #include <linux/ratelimit.h> +#include <linux/overflow.h> #include "overlayfs.h" struct ovl_cache_entry { @@ -147,9 +148,8 @@ static struct ovl_cache_entry *ovl_cache_entry_new(struct ovl_readdir_data *rdd, u64 ino, unsigned int d_type) { struct ovl_cache_entry *p; - size_t size = offsetof(struct ovl_cache_entry, name[len + 1]); - p = kmalloc(size, GFP_KERNEL); + p = kmalloc(struct_size(p, name, len + 1), GFP_KERNEL); if (!p) return NULL; diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c index 0819c739cc2f..dcccb4b4a66c 100644 --- a/fs/overlayfs/util.c +++ b/fs/overlayfs/util.c @@ -15,6 +15,7 @@ #include <linux/uuid.h> #include <linux/namei.h> #include <linux/ratelimit.h> +#include <linux/overflow.h> #include "overlayfs.h" /* Get write access to upper mnt - may fail if upper sb was remounted ro */ @@ -145,9 +146,9 @@ void ovl_stack_free(struct ovl_path *stack, unsigned int n) struct ovl_entry *ovl_alloc_entry(unsigned int numlower) { - size_t size = offsetof(struct ovl_entry, __lowerstack[numlower]); - struct ovl_entry *oe = kzalloc(size, GFP_KERNEL); + struct ovl_entry *oe; + oe = kzalloc(struct_size(oe, __lowerstack, numlower), GFP_KERNEL); if (oe) oe->__numlower = numlower; @@ -305,7 +306,9 @@ enum ovl_path_type ovl_path_realdata(struct dentry *dentry, struct path *path) struct dentry *ovl_dentry_upper(struct dentry *dentry) { - return ovl_upperdentry_dereference(OVL_I(d_inode(dentry))); + struct inode *inode = d_inode(dentry); + + return inode ? ovl_upperdentry_dereference(OVL_I(inode)) : NULL; } struct dentry *ovl_dentry_lower(struct dentry *dentry) diff --git a/fs/pidfs.c b/fs/pidfs.c index c1f0a067be40..4625e097e3a0 100644 --- a/fs/pidfs.c +++ b/fs/pidfs.c @@ -319,7 +319,7 @@ static long pidfd_info(struct file *file, unsigned int cmd, unsigned long arg) if (!c) return -ESRCH; - if (!(kinfo.mask & PIDFD_INFO_COREDUMP)) { + if ((kinfo.mask & PIDFD_INFO_COREDUMP) && !(kinfo.coredump_mask)) { task_lock(task); if (task->mm) kinfo.coredump_mask = pidfs_coredump_mask(task->mm->flags); @@ -366,7 +366,7 @@ static long pidfd_info(struct file *file, unsigned int cmd, unsigned long arg) kinfo.pid = task_pid_vnr(task); kinfo.mask |= PIDFD_INFO_PID; - if (kinfo.pid == 0 || kinfo.tgid == 0 || (kinfo.ppid == 0 && kinfo.pid != 1)) + if (kinfo.pid == 0 || kinfo.tgid == 0) return -ESRCH; copy_out: diff --git a/fs/pnode.h b/fs/pnode.h index 34b6247af01d..2d026fb98b18 100644 --- a/fs/pnode.h +++ b/fs/pnode.h @@ -28,8 +28,6 @@ #define CL_SHARED_TO_SLAVE 0x20 #define CL_COPY_MNT_NS_FILE 0x40 -#define CL_COPY_ALL (CL_COPY_UNBINDABLE | CL_COPY_MNT_NS_FILE) - static inline void set_mnt_shared(struct mount *mnt) { mnt->mnt.mnt_flags &= ~MNT_SHARED_MASK; diff --git a/fs/proc/inode.c b/fs/proc/inode.c index a3eb3b740f76..3604b616311c 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -42,7 +42,7 @@ static void proc_evict_inode(struct inode *inode) head = ei->sysctl; if (head) { - RCU_INIT_POINTER(ei->sysctl, NULL); + WRITE_ONCE(ei->sysctl, NULL); proc_sys_evict_inode(inode, head); } } diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index cc9d74a06ff0..08b78150cdde 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -918,17 +918,21 @@ static int proc_sys_compare(const struct dentry *dentry, struct ctl_table_header *head; struct inode *inode; - /* Although proc doesn't have negative dentries, rcu-walk means - * that inode here can be NULL */ - /* AV: can it, indeed? */ - inode = d_inode_rcu(dentry); - if (!inode) - return 1; if (name->len != len) return 1; if (memcmp(name->name, str, len)) return 1; - head = rcu_dereference(PROC_I(inode)->sysctl); + + // false positive is fine here - we'll recheck anyway + if (d_in_lookup(dentry)) + return 0; + + inode = d_inode_rcu(dentry); + // we just might have run into dentry in the middle of __dentry_kill() + if (!inode) + return 1; + + head = READ_ONCE(PROC_I(inode)->sysctl); return !head || !sysctl_is_seen(head); } diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 27972c0749e7..751479eb128f 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -36,9 +36,9 @@ void task_mem(struct seq_file *m, struct mm_struct *mm) unsigned long text, lib, swap, anon, file, shmem; unsigned long hiwater_vm, total_vm, hiwater_rss, total_rss; - anon = get_mm_counter(mm, MM_ANONPAGES); - file = get_mm_counter(mm, MM_FILEPAGES); - shmem = get_mm_counter(mm, MM_SHMEMPAGES); + anon = get_mm_counter_sum(mm, MM_ANONPAGES); + file = get_mm_counter_sum(mm, MM_FILEPAGES); + shmem = get_mm_counter_sum(mm, MM_SHMEMPAGES); /* * Note: to minimize their overhead, mm maintains hiwater_vm and @@ -59,7 +59,7 @@ void task_mem(struct seq_file *m, struct mm_struct *mm) text = min(text, mm->exec_vm << PAGE_SHIFT); lib = (mm->exec_vm << PAGE_SHIFT) - text; - swap = get_mm_counter(mm, MM_SWAPENTS); + swap = get_mm_counter_sum(mm, MM_SWAPENTS); SEQ_PUT_DEC("VmPeak:\t", hiwater_vm); SEQ_PUT_DEC(" kB\nVmSize:\t", total_vm); SEQ_PUT_DEC(" kB\nVmLck:\t", mm->locked_vm); @@ -92,12 +92,12 @@ unsigned long task_statm(struct mm_struct *mm, unsigned long *shared, unsigned long *text, unsigned long *data, unsigned long *resident) { - *shared = get_mm_counter(mm, MM_FILEPAGES) + - get_mm_counter(mm, MM_SHMEMPAGES); + *shared = get_mm_counter_sum(mm, MM_FILEPAGES) + + get_mm_counter_sum(mm, MM_SHMEMPAGES); *text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK)) >> PAGE_SHIFT; *data = mm->data_vm + mm->stack_vm; - *resident = *shared + get_mm_counter(mm, MM_ANONPAGES); + *resident = *shared + get_mm_counter_sum(mm, MM_ANONPAGES); return mm->total_vm; } @@ -2182,7 +2182,7 @@ static unsigned long pagemap_thp_category(struct pagemap_scan_private *p, categories |= PAGE_IS_FILE; } - if (is_zero_pfn(pmd_pfn(pmd))) + if (is_huge_zero_pmd(pmd)) categories |= PAGE_IS_PFNZERO; if (pmd_soft_dirty(pmd)) categories |= PAGE_IS_SOFT_DIRTY; diff --git a/fs/resctrl/ctrlmondata.c b/fs/resctrl/ctrlmondata.c index 6ed2dfd4dbbd..d98e0d2de09f 100644 --- a/fs/resctrl/ctrlmondata.c +++ b/fs/resctrl/ctrlmondata.c @@ -594,9 +594,10 @@ int rdtgroup_mondata_show(struct seq_file *m, void *arg) struct rmid_read rr = {0}; struct rdt_mon_domain *d; struct rdtgroup *rdtgrp; + int domid, cpu, ret = 0; struct rdt_resource *r; + struct cacheinfo *ci; struct mon_data *md; - int domid, ret = 0; rdtgrp = rdtgroup_kn_lock_live(of->kn); if (!rdtgrp) { @@ -623,10 +624,14 @@ int rdtgroup_mondata_show(struct seq_file *m, void *arg) * one that matches this cache id. */ list_for_each_entry(d, &r->mon_domains, hdr.list) { - if (d->ci->id == domid) { - rr.ci = d->ci; + if (d->ci_id == domid) { + rr.ci_id = d->ci_id; + cpu = cpumask_any(&d->hdr.cpu_mask); + ci = get_cpu_cacheinfo_level(cpu, RESCTRL_L3_CACHE); + if (!ci) + continue; mon_event_read(&rr, r, NULL, rdtgrp, - &d->ci->shared_cpu_map, evtid, false); + &ci->shared_cpu_map, evtid, false); goto checkresult; } } diff --git a/fs/resctrl/internal.h b/fs/resctrl/internal.h index 9a8cf6f11151..0a1eedba2b03 100644 --- a/fs/resctrl/internal.h +++ b/fs/resctrl/internal.h @@ -98,7 +98,7 @@ struct mon_data { * domains in @r sharing L3 @ci.id * @evtid: Which monitor event to read. * @first: Initialize MBM counter when true. - * @ci: Cacheinfo for L3. Only set when @d is NULL. Used when summing domains. + * @ci_id: Cacheinfo id for L3. Only set when @d is NULL. Used when summing domains. * @err: Error encountered when reading counter. * @val: Returned value of event counter. If @rgrp is a parent resource group, * @val includes the sum of event counts from its child resource groups. @@ -112,7 +112,7 @@ struct rmid_read { struct rdt_mon_domain *d; enum resctrl_event_id evtid; bool first; - struct cacheinfo *ci; + unsigned int ci_id; int err; u64 val; void *arch_mon_ctx; diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c index bde2801289d3..f5637855c3ac 100644 --- a/fs/resctrl/monitor.c +++ b/fs/resctrl/monitor.c @@ -361,6 +361,7 @@ static int __mon_event_count(u32 closid, u32 rmid, struct rmid_read *rr) { int cpu = smp_processor_id(); struct rdt_mon_domain *d; + struct cacheinfo *ci; struct mbm_state *m; int err, ret; u64 tval = 0; @@ -388,7 +389,8 @@ static int __mon_event_count(u32 closid, u32 rmid, struct rmid_read *rr) } /* Summing domains that share a cache, must be on a CPU for that cache. */ - if (!cpumask_test_cpu(cpu, &rr->ci->shared_cpu_map)) + ci = get_cpu_cacheinfo_level(cpu, RESCTRL_L3_CACHE); + if (!ci || ci->id != rr->ci_id) return -EINVAL; /* @@ -400,7 +402,7 @@ static int __mon_event_count(u32 closid, u32 rmid, struct rmid_read *rr) */ ret = -EINVAL; list_for_each_entry(d, &rr->r->mon_domains, hdr.list) { - if (d->ci->id != rr->ci->id) + if (d->ci_id != rr->ci_id) continue; err = resctrl_arch_rmid_read(rr->r, d, closid, rmid, rr->evtid, &tval, rr->arch_mon_ctx); diff --git a/fs/resctrl/rdtgroup.c b/fs/resctrl/rdtgroup.c index cc37f58b47dd..77d08229d855 100644 --- a/fs/resctrl/rdtgroup.c +++ b/fs/resctrl/rdtgroup.c @@ -536,6 +536,8 @@ static ssize_t rdtgroup_cpus_write(struct kernfs_open_file *of, goto unlock; } + rdt_last_cmd_clear(); + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED || rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { ret = -EINVAL; @@ -3034,7 +3036,7 @@ static void rmdir_mondata_subdir_allrdtgrp(struct rdt_resource *r, char name[32]; snc_mode = r->mon_scope == RESCTRL_L3_NODE; - sprintf(name, "mon_%s_%02d", r->name, snc_mode ? d->ci->id : d->hdr.id); + sprintf(name, "mon_%s_%02d", r->name, snc_mode ? d->ci_id : d->hdr.id); if (snc_mode) sprintf(subname, "mon_sub_%s_%02d", r->name, d->hdr.id); @@ -3059,7 +3061,7 @@ static int mon_add_all_files(struct kernfs_node *kn, struct rdt_mon_domain *d, return -EPERM; list_for_each_entry(mevt, &r->evt_list, list) { - domid = do_sum ? d->ci->id : d->hdr.id; + domid = do_sum ? d->ci_id : d->hdr.id; priv = mon_get_kn_priv(r->rid, domid, mevt, do_sum); if (WARN_ON_ONCE(!priv)) return -EINVAL; @@ -3087,7 +3089,7 @@ static int mkdir_mondata_subdir(struct kernfs_node *parent_kn, lockdep_assert_held(&rdtgroup_mutex); snc_mode = r->mon_scope == RESCTRL_L3_NODE; - sprintf(name, "mon_%s_%02d", r->name, snc_mode ? d->ci->id : d->hdr.id); + sprintf(name, "mon_%s_%02d", r->name, snc_mode ? d->ci_id : d->hdr.id); kn = kernfs_find_and_get(parent_kn, name); if (kn) { /* @@ -3472,6 +3474,8 @@ static int mkdir_rdt_prepare(struct kernfs_node *parent_kn, goto out_unlock; } + rdt_last_cmd_clear(); + /* * Check that the parent directory for a monitor group is a "mon_groups" * directory. diff --git a/fs/smb/client/cached_dir.c b/fs/smb/client/cached_dir.c index 89d2dbbb742c..368e870624da 100644 --- a/fs/smb/client/cached_dir.c +++ b/fs/smb/client/cached_dir.c @@ -155,6 +155,7 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, struct cached_fids *cfids; const char *npath; int retries = 0, cur_sleep = 1; + __le32 lease_flags = 0; if (cifs_sb->root == NULL) return -ENOENT; @@ -201,6 +202,8 @@ replay_again: } spin_unlock(&cfids->cfid_list_lock); + pfid = &cfid->fid; + /* * Skip any prefix paths in @path as lookup_noperm_positive_unlocked() ends up * calling ->lookup() which already adds those through @@ -222,6 +225,25 @@ replay_again: rc = -ENOENT; goto out; } + if (dentry->d_parent && server->dialect >= SMB30_PROT_ID) { + struct cached_fid *parent_cfid; + + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry(parent_cfid, &cfids->entries, entry) { + if (parent_cfid->dentry == dentry->d_parent) { + cifs_dbg(FYI, "found a parent cached file handle\n"); + if (parent_cfid->has_lease && parent_cfid->time) { + lease_flags + |= SMB2_LEASE_FLAG_PARENT_LEASE_KEY_SET_LE; + memcpy(pfid->parent_lease_key, + parent_cfid->fid.lease_key, + SMB2_LEASE_KEY_SIZE); + } + break; + } + } + spin_unlock(&cfids->cfid_list_lock); + } } cfid->dentry = dentry; cfid->tcon = tcon; @@ -236,7 +258,6 @@ replay_again: if (smb3_encryption_required(tcon)) flags |= CIFS_TRANSFORM_REQ; - pfid = &cfid->fid; server->ops->new_lease_key(pfid); memset(rqst, 0, sizeof(rqst)); @@ -256,6 +277,7 @@ replay_again: FILE_READ_EA, .disposition = FILE_OPEN, .fid = pfid, + .lease_flags = lease_flags, .replay = !!(retries), }; @@ -487,8 +509,17 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) spin_lock(&cfids->cfid_list_lock); list_for_each_entry(cfid, &cfids->entries, entry) { tmp_list = kmalloc(sizeof(*tmp_list), GFP_ATOMIC); - if (tmp_list == NULL) - break; + if (tmp_list == NULL) { + /* + * If the malloc() fails, we won't drop all + * dentries, and unmounting is likely to trigger + * a 'Dentry still in use' error. + */ + cifs_tcon_dbg(VFS, "Out of memory while dropping dentries\n"); + spin_unlock(&cfids->cfid_list_lock); + spin_unlock(&cifs_sb->tlink_tree_lock); + goto done; + } spin_lock(&cfid->fid_lock); tmp_list->dentry = cfid->dentry; cfid->dentry = NULL; @@ -500,6 +531,7 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) } spin_unlock(&cifs_sb->tlink_tree_lock); +done: list_for_each_entry_safe(tmp_list, q, &entry, entry) { list_del(&tmp_list->entry); dput(tmp_list->dentry); diff --git a/fs/smb/client/cached_dir.h b/fs/smb/client/cached_dir.h index 1dfe79d947a6..a28f7cae3caa 100644 --- a/fs/smb/client/cached_dir.h +++ b/fs/smb/client/cached_dir.h @@ -21,12 +21,12 @@ struct cached_dirent { struct cached_dirents { bool is_valid:1; bool is_failed:1; - struct dir_context *ctx; /* - * Only used to make sure we only take entries - * from a single context. Never dereferenced. - */ + struct file *file; /* + * Used to associate the cache with a single + * open file instance. + */ struct mutex de_mutex; - int pos; /* Expected ctx->pos */ + loff_t pos; /* Expected ctx->pos */ struct list_head entries; }; diff --git a/fs/smb/client/cifs_debug.c b/fs/smb/client/cifs_debug.c index e03c890de0a0..3fdf75737d43 100644 --- a/fs/smb/client/cifs_debug.c +++ b/fs/smb/client/cifs_debug.c @@ -362,6 +362,10 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) c = 0; spin_lock(&cifs_tcp_ses_lock); list_for_each_entry(server, &cifs_tcp_ses_list, tcp_ses_list) { +#ifdef CONFIG_CIFS_SMB_DIRECT + struct smbdirect_socket_parameters *sp; +#endif + /* channel info will be printed as a part of sessions below */ if (SERVER_IS_CHAN(server)) continue; @@ -383,25 +387,26 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) seq_printf(m, "\nSMBDirect transport not available"); goto skip_rdma; } + sp = &server->smbd_conn->socket.parameters; seq_printf(m, "\nSMBDirect (in hex) protocol version: %x " "transport status: %x", server->smbd_conn->protocol, - server->smbd_conn->transport_status); + server->smbd_conn->socket.status); seq_printf(m, "\nConn receive_credit_max: %x " "send_credit_target: %x max_send_size: %x", - server->smbd_conn->receive_credit_max, - server->smbd_conn->send_credit_target, - server->smbd_conn->max_send_size); + sp->recv_credit_max, + sp->send_credit_target, + sp->max_send_size); seq_printf(m, "\nConn max_fragmented_recv_size: %x " "max_fragmented_send_size: %x max_receive_size:%x", - server->smbd_conn->max_fragmented_recv_size, - server->smbd_conn->max_fragmented_send_size, - server->smbd_conn->max_receive_size); + sp->max_fragmented_recv_size, + sp->max_fragmented_send_size, + sp->max_recv_size); seq_printf(m, "\nConn keep_alive_interval: %x " "max_readwrite_size: %x rdma_readwrite_threshold: %x", - server->smbd_conn->keep_alive_interval, - server->smbd_conn->max_readwrite_size, + sp->keepalive_interval_msec * 1000, + sp->max_read_write_size, server->smbd_conn->rdma_readwrite_threshold); seq_printf(m, "\nDebug count_get_receive_buffer: %x " "count_put_receive_buffer: %x count_send_empty: %x", @@ -1100,7 +1105,7 @@ static ssize_t cifs_security_flags_proc_write(struct file *file, if ((count < 1) || (count > 11)) return -EINVAL; - memset(flags_string, 0, 12); + memset(flags_string, 0, sizeof(flags_string)); if (copy_from_user(flags_string, buffer, count)) return -EFAULT; diff --git a/fs/smb/client/cifs_ioctl.h b/fs/smb/client/cifs_ioctl.h index 26327442e383..b51ce64fcccf 100644 --- a/fs/smb/client/cifs_ioctl.h +++ b/fs/smb/client/cifs_ioctl.h @@ -61,7 +61,7 @@ struct smb_query_info { struct smb3_key_debug_info { __u64 Suid; __u16 cipher_type; - __u8 auth_key[16]; /* SMB2_NTLMV2_SESSKEY_SIZE */ + __u8 auth_key[SMB2_NTLMV2_SESSKEY_SIZE]; __u8 smb3encryptionkey[SMB3_SIGN_KEY_SIZE]; __u8 smb3decryptionkey[SMB3_SIGN_KEY_SIZE]; } __packed; diff --git a/fs/smb/client/cifsfs.c b/fs/smb/client/cifsfs.c index fb04e263611c..0a5266ecfd15 100644 --- a/fs/smb/client/cifsfs.c +++ b/fs/smb/client/cifsfs.c @@ -70,7 +70,6 @@ bool require_gcm_256; /* false by default */ bool enable_negotiate_signing; /* false by default */ unsigned int global_secflags = CIFSSEC_DEF; /* unsigned int ntlmv2_support = 0; */ -unsigned int sign_CIFS_PDUs = 1; /* * Global transaction id (XID) information diff --git a/fs/smb/client/cifsfs.h b/fs/smb/client/cifsfs.h index ca435a3841b8..b9ec9fe16a98 100644 --- a/fs/smb/client/cifsfs.h +++ b/fs/smb/client/cifsfs.h @@ -145,6 +145,6 @@ extern const struct export_operations cifs_export_ops; #endif /* CONFIG_CIFS_NFSD_EXPORT */ /* when changing internal version - update following two lines at same time */ -#define SMB3_PRODUCT_BUILD 54 -#define CIFS_VERSION "2.54" +#define SMB3_PRODUCT_BUILD 55 +#define CIFS_VERSION "2.55" #endif /* _CIFSFS_H */ diff --git a/fs/smb/client/cifsglob.h b/fs/smb/client/cifsglob.h index 3b32116b0b49..89160bc34d35 100644 --- a/fs/smb/client/cifsglob.h +++ b/fs/smb/client/cifsglob.h @@ -556,7 +556,7 @@ struct smb_version_operations { void (*set_oplock_level)(struct cifsInodeInfo *cinode, __u32 oplock, __u16 epoch, bool *purge_cache); /* create lease context buffer for CREATE request */ - char * (*create_lease_buf)(u8 *lease_key, u8 oplock); + char * (*create_lease_buf)(u8 *lease_key, u8 oplock, u8 *parent_lease_key, __le32 le_flags); /* parse lease context buffer and return oplock/epoch info */ __u8 (*parse_lease_buf)(void *buf, __u16 *epoch, char *lkey); ssize_t (*copychunk_range)(const unsigned int, @@ -709,6 +709,7 @@ inc_rfc1001_len(void *buf, int count) struct TCP_Server_Info { struct list_head tcp_ses_list; struct list_head smb_ses_list; + struct list_head rlist; /* reconnect list */ spinlock_t srv_lock; /* protect anything here that is not protected */ __u64 conn_id; /* connection identifier (useful for debugging) */ int srv_count; /* reference counter */ @@ -773,8 +774,10 @@ struct TCP_Server_Info { char workstation_RFC1001_name[RFC1001_NAME_LEN_WITH_NULL]; __u32 sequence_number; /* for signing, protected by srv_mutex */ __u32 reconnect_instance; /* incremented on each reconnect */ + __le32 session_key_id; /* retrieved from negotiate response and send in session setup request */ struct session_key session_key; unsigned long lstrp; /* when we got last response from this server */ + unsigned long neg_start; /* when negotiate started (jiffies) */ struct cifs_secmech secmech; /* crypto sec mech functs, descriptors */ #define CIFS_NEGFLAVOR_UNENCAP 1 /* wct == 17, but no ext_sec */ #define CIFS_NEGFLAVOR_EXTENDED 2 /* wct == 17, ext_sec bit set */ @@ -1084,6 +1087,7 @@ struct cifs_chan { }; #define CIFS_SES_FLAG_SCALE_CHANNELS (0x1) +#define CIFS_SES_FLAGS_PENDING_QUERY_INTERFACES (0x2) /* * Session structure. One of these for each uid session with a particular host @@ -1300,6 +1304,7 @@ struct cifs_tcon { bool use_persistent:1; /* use persistent instead of durable handles */ bool no_lease:1; /* Do not request leases on files or directories */ bool use_witness:1; /* use witness protocol */ + bool dummy:1; /* dummy tcon used for reconnecting channels */ __le32 capabilities; __u32 share_flags; __u32 maximal_access; @@ -1441,6 +1446,7 @@ struct cifs_open_parms { bool reconnect:1; bool replay:1; /* indicates that this open is for a replay */ struct kvec *ea_cctx; + __le32 lease_flags; }; struct cifs_fid { @@ -1448,6 +1454,7 @@ struct cifs_fid { __u64 persistent_fid; /* persist file id for smb2 */ __u64 volatile_fid; /* volatile file id for smb2 */ __u8 lease_key[SMB2_LEASE_KEY_SIZE]; /* lease key for smb2 */ + __u8 parent_lease_key[SMB2_LEASE_KEY_SIZE]; __u8 create_guid[16]; __u32 access; struct cifs_pending_open *pending_open; @@ -1988,8 +1995,7 @@ require use of the stronger protocol */ * TCP_Server_Info-> TCP_Server_Info cifs_get_tcp_session * reconnect_mutex * TCP_Server_Info->srv_mutex TCP_Server_Info cifs_get_tcp_session - * cifs_ses->session_mutex cifs_ses sesInfoAlloc - * cifs_tcon + * cifs_ses->session_mutex cifs_ses sesInfoAlloc * cifs_tcon->open_file_lock cifs_tcon->openFileList tconInfoAlloc * cifs_tcon->pending_opens * cifs_tcon->stat_lock cifs_tcon->bytes_read tconInfoAlloc @@ -2008,21 +2014,25 @@ require use of the stronger protocol */ * ->oplock_credits * ->reconnect_instance * cifs_ses->ses_lock (anything that is not protected by another lock and can change) + * sesInfoAlloc * cifs_ses->iface_lock cifs_ses->iface_list sesInfoAlloc * ->iface_count * ->iface_last_update - * cifs_ses->chan_lock cifs_ses->chans + * cifs_ses->chan_lock cifs_ses->chans sesInfoAlloc * ->chans_need_reconnect * ->chans_in_reconnect * cifs_tcon->tc_lock (anything that is not protected by another lock and can change) + * tcon_info_alloc * inode->i_rwsem, taken by fs/netfs/locking.c e.g. should be taken before cifsInodeInfo locks * cifsInodeInfo->open_file_lock cifsInodeInfo->openFileList cifs_alloc_inode * cifsInodeInfo->writers_lock cifsInodeInfo->writers cifsInodeInfo_alloc * cifsInodeInfo->lock_sem cifsInodeInfo->llist cifs_init_once * ->can_cache_brlcks * cifsInodeInfo->deferred_lock cifsInodeInfo->deferred_closes cifsInodeInfo_alloc - * cached_fids->cfid_list_lock cifs_tcon->cfids->entries init_cached_dirs - * cifsFileInfo->fh_mutex cifsFileInfo cifs_new_fileinfo + * cached_fids->cfid_list_lock cifs_tcon->cfids->entries init_cached_dirs + * cached_fid->fid_lock (anything that is not protected by another lock and can change) + * init_cached_dir + * cifsFileInfo->fh_mutex cifsFileInfo cifs_new_fileinfo * cifsFileInfo->file_info_lock cifsFileInfo->count cifs_new_fileinfo * ->invalidHandle initiate_cifs_search * ->oplock_break_cancelled diff --git a/fs/smb/client/cifspdu.h b/fs/smb/client/cifspdu.h index 1b79fe07476f..d9cf7db0ac35 100644 --- a/fs/smb/client/cifspdu.h +++ b/fs/smb/client/cifspdu.h @@ -597,7 +597,7 @@ typedef union smb_com_session_setup_andx { __le16 MaxBufferSize; __le16 MaxMpxCount; __le16 VcNumber; - __u32 SessionKey; + __le32 SessionKey; __le16 SecurityBlobLength; __u32 Reserved; __le32 Capabilities; /* see below */ @@ -616,7 +616,7 @@ typedef union smb_com_session_setup_andx { __le16 MaxBufferSize; __le16 MaxMpxCount; __le16 VcNumber; - __u32 SessionKey; + __le32 SessionKey; __le16 CaseInsensitivePasswordLength; /* ASCII password len */ __le16 CaseSensitivePasswordLength; /* Unicode password length*/ __u32 Reserved; /* see below */ @@ -654,7 +654,7 @@ typedef union smb_com_session_setup_andx { __le16 MaxBufferSize; __le16 MaxMpxCount; __le16 VcNumber; - __u32 SessionKey; + __le32 SessionKey; __le16 PasswordLength; __u32 Reserved; /* encrypt key len and offset */ __le16 ByteCount; diff --git a/fs/smb/client/cifsproto.h b/fs/smb/client/cifsproto.h index 66093fa78aed..045227ed4efc 100644 --- a/fs/smb/client/cifsproto.h +++ b/fs/smb/client/cifsproto.h @@ -136,6 +136,7 @@ extern int SendReceiveBlockingLock(const unsigned int xid, struct smb_hdr *out_buf, int *bytes_returned); +void smb2_query_server_interfaces(struct work_struct *work); void cifs_signal_cifsd_for_reconnect(struct TCP_Server_Info *server, bool all_channels); diff --git a/fs/smb/client/cifssmb.c b/fs/smb/client/cifssmb.c index 477792c07d45..75142f49d65d 100644 --- a/fs/smb/client/cifssmb.c +++ b/fs/smb/client/cifssmb.c @@ -498,6 +498,7 @@ CIFSSMBNegotiate(const unsigned int xid, server->max_rw = le32_to_cpu(pSMBr->MaxRawSize); cifs_dbg(NOISY, "Max buf = %d\n", ses->server->maxBuf); server->capabilities = le32_to_cpu(pSMBr->Capabilities); + server->session_key_id = pSMBr->SessionKey; server->timeAdj = (int)(__s16)le16_to_cpu(pSMBr->ServerTimeZone); server->timeAdj *= 60; @@ -1333,7 +1334,12 @@ cifs_readv_callback(struct mid_q_entry *mid) cifs_stats_bytes_read(tcon, rdata->got_bytes); break; case MID_REQUEST_SUBMITTED: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_req_submitted); + goto do_retry; case MID_RETRY_NEEDED: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_retry_needed); +do_retry: + __set_bit(NETFS_SREQ_NEED_RETRY, &rdata->subreq.flags); rdata->result = -EAGAIN; if (server->sign && rdata->got_bytes) /* reset bytes number since we can not check a sign */ @@ -1342,8 +1348,14 @@ cifs_readv_callback(struct mid_q_entry *mid) task_io_account_read(rdata->got_bytes); cifs_stats_bytes_read(tcon, rdata->got_bytes); break; + case MID_RESPONSE_MALFORMED: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_malformed); + rdata->result = -EIO; + break; default: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_unknown); rdata->result = -EIO; + break; } if (rdata->result == -ENODATA) { @@ -1712,10 +1724,21 @@ cifs_writev_callback(struct mid_q_entry *mid) } break; case MID_REQUEST_SUBMITTED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_req_submitted); + __set_bit(NETFS_SREQ_NEED_RETRY, &wdata->subreq.flags); + result = -EAGAIN; + break; case MID_RETRY_NEEDED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_retry_needed); + __set_bit(NETFS_SREQ_NEED_RETRY, &wdata->subreq.flags); result = -EAGAIN; break; + case MID_RESPONSE_MALFORMED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_malformed); + result = -EIO; + break; default: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_unknown); result = -EIO; break; } @@ -2753,10 +2776,10 @@ int cifs_query_reparse_point(const unsigned int xid, io_req->TotalParameterCount = 0; io_req->TotalDataCount = 0; - io_req->MaxParameterCount = cpu_to_le32(2); + io_req->MaxParameterCount = cpu_to_le32(0); /* BB find exact data count max from sess structure BB */ io_req->MaxDataCount = cpu_to_le32(CIFSMaxBufSize & 0xFFFFFF00); - io_req->MaxSetupCount = 4; + io_req->MaxSetupCount = 1; io_req->Reserved = 0; io_req->ParameterOffset = 0; io_req->DataCount = 0; @@ -2783,6 +2806,22 @@ int cifs_query_reparse_point(const unsigned int xid, goto error; } + /* SetupCount must be 1, otherwise offset to ByteCount is incorrect. */ + if (io_rsp->SetupCount != 1) { + rc = -EIO; + goto error; + } + + /* + * ReturnedDataLen is output length of executed IOCTL. + * DataCount is output length transferred over network. + * Check that we have full FSCTL_GET_REPARSE_POINT buffer. + */ + if (data_count != le16_to_cpu(io_rsp->ReturnedDataLen)) { + rc = -EIO; + goto error; + } + end = 2 + get_bcc(&io_rsp->hdr) + (__u8 *)&io_rsp->ByteCount; start = (__u8 *)&io_rsp->hdr.Protocol + data_offset; if (start >= end) { diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index 6bf04d9a5491..205f547ca49e 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -97,7 +97,7 @@ static int reconn_set_ipaddr_from_hostname(struct TCP_Server_Info *server) return rc; } -static void smb2_query_server_interfaces(struct work_struct *work) +void smb2_query_server_interfaces(struct work_struct *work) { int rc; int xid; @@ -116,18 +116,22 @@ static void smb2_query_server_interfaces(struct work_struct *work) rc = server->ops->query_server_interfaces(xid, tcon, false); free_xid(xid); - if (rc) { - if (rc == -EOPNOTSUPP) - return; - + if (rc) cifs_dbg(FYI, "%s: failed to query server interfaces: %d\n", __func__, rc); - } queue_delayed_work(cifsiod_wq, &tcon->query_interfaces, (SMB_INTERFACE_POLL_INTERVAL * HZ)); } +#define set_need_reco(server) \ +do { \ + spin_lock(&server->srv_lock); \ + if (server->tcpStatus != CifsExiting) \ + server->tcpStatus = CifsNeedReconnect; \ + spin_unlock(&server->srv_lock); \ +} while (0) + /* * Update the tcpStatus for the server. * This is used to signal the cifsd thread to call cifs_reconnect @@ -141,39 +145,45 @@ void cifs_signal_cifsd_for_reconnect(struct TCP_Server_Info *server, bool all_channels) { - struct TCP_Server_Info *pserver; + struct TCP_Server_Info *nserver; struct cifs_ses *ses; + LIST_HEAD(reco); int i; - /* If server is a channel, select the primary channel */ - pserver = SERVER_IS_CHAN(server) ? server->primary_server : server; - /* if we need to signal just this channel */ if (!all_channels) { - spin_lock(&server->srv_lock); - if (server->tcpStatus != CifsExiting) - server->tcpStatus = CifsNeedReconnect; - spin_unlock(&server->srv_lock); + set_need_reco(server); return; } - spin_lock(&cifs_tcp_ses_lock); - list_for_each_entry(ses, &pserver->smb_ses_list, smb_ses_list) { - if (cifs_ses_exiting(ses)) - continue; - spin_lock(&ses->chan_lock); - for (i = 0; i < ses->chan_count; i++) { - if (!ses->chans[i].server) + if (SERVER_IS_CHAN(server)) + server = server->primary_server; + scoped_guard(spinlock, &cifs_tcp_ses_lock) { + set_need_reco(server); + list_for_each_entry(ses, &server->smb_ses_list, smb_ses_list) { + spin_lock(&ses->ses_lock); + if (ses->ses_status == SES_EXITING) { + spin_unlock(&ses->ses_lock); continue; - - spin_lock(&ses->chans[i].server->srv_lock); - if (ses->chans[i].server->tcpStatus != CifsExiting) - ses->chans[i].server->tcpStatus = CifsNeedReconnect; - spin_unlock(&ses->chans[i].server->srv_lock); + } + spin_lock(&ses->chan_lock); + for (i = 1; i < ses->chan_count; i++) { + nserver = ses->chans[i].server; + if (!nserver) + continue; + nserver->srv_count++; + list_add(&nserver->rlist, &reco); + } + spin_unlock(&ses->chan_lock); + spin_unlock(&ses->ses_lock); } - spin_unlock(&ses->chan_lock); } - spin_unlock(&cifs_tcp_ses_lock); + + list_for_each_entry_safe(server, nserver, &reco, rlist) { + list_del_init(&server->rlist); + set_need_reco(server); + cifs_put_tcp_session(server, 0); + } } /* @@ -377,6 +387,13 @@ static int __cifs_reconnect(struct TCP_Server_Info *server, if (!cifs_tcp_ses_needs_reconnect(server, 1)) return 0; + /* + * if smb session has been marked for reconnect, also reconnect all + * connections. This way, the other connections do not end up bad. + */ + if (mark_smb_session) + cifs_signal_cifsd_for_reconnect(server, mark_smb_session); + cifs_mark_tcp_ses_conns_for_reconnect(server, mark_smb_session); cifs_abort_connection(server); @@ -385,7 +402,8 @@ static int __cifs_reconnect(struct TCP_Server_Info *server, try_to_freeze(); cifs_server_lock(server); - if (!cifs_swn_set_server_dstaddr(server)) { + if (!cifs_swn_set_server_dstaddr(server) && + !SERVER_IS_CHAN(server)) { /* resolve the hostname again to make sure that IP address is up-to-date */ rc = reconn_set_ipaddr_from_hostname(server); cifs_dbg(FYI, "%s: reconn_set_ipaddr_from_hostname: rc=%d\n", __func__, rc); @@ -661,12 +679,12 @@ server_unresponsive(struct TCP_Server_Info *server) /* * If we're in the process of mounting a share or reconnecting a session * and the server abruptly shut down (e.g. socket wasn't closed, packet - * had been ACK'ed but no SMB response), don't wait longer than 20s to - * negotiate protocol. + * had been ACK'ed but no SMB response), don't wait longer than 20s from + * when negotiate actually started. */ spin_lock(&server->srv_lock); if (server->tcpStatus == CifsInNegotiate && - time_after(jiffies, server->lstrp + 20 * HZ)) { + time_after(jiffies, server->neg_start + 20 * HZ)) { spin_unlock(&server->srv_lock); cifs_reconnect(server, false); return true; @@ -2862,20 +2880,14 @@ cifs_get_tcon(struct cifs_ses *ses, struct smb3_fs_context *ctx) tcon->max_cached_dirs = ctx->max_cached_dirs; tcon->nodelete = ctx->nodelete; tcon->local_lease = ctx->local_lease; - INIT_LIST_HEAD(&tcon->pending_opens); tcon->status = TID_GOOD; - INIT_DELAYED_WORK(&tcon->query_interfaces, - smb2_query_server_interfaces); if (ses->server->dialect >= SMB30_PROT_ID && (ses->server->capabilities & SMB2_GLOBAL_CAP_MULTI_CHANNEL)) { /* schedule query interfaces poll */ queue_delayed_work(cifsiod_wq, &tcon->query_interfaces, (SMB_INTERFACE_POLL_INTERVAL * HZ)); } -#ifdef CONFIG_CIFS_DFS_UPCALL - INIT_DELAYED_WORK(&tcon->dfs_cache_work, dfs_cache_refresh); -#endif spin_lock(&cifs_tcp_ses_lock); list_add(&tcon->tcon_list, &ses->tcon_list); spin_unlock(&cifs_tcp_ses_lock); @@ -3714,9 +3726,15 @@ int cifs_mount_get_tcon(struct cifs_mount_ctx *mnt_ctx) goto out; } - /* if new SMB3.11 POSIX extensions are supported do not remap / and \ */ - if (tcon->posix_extensions) + /* + * if new SMB3.11 POSIX extensions are supported, do not change anything in the + * path (i.e., do not remap / and \ and do not map any special characters) + */ + if (tcon->posix_extensions) { cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_POSIX_PATHS; + cifs_sb->mnt_cifs_flags &= ~(CIFS_MOUNT_MAP_SFM_CHR | + CIFS_MOUNT_MAP_SPECIAL_CHR); + } #ifdef CONFIG_CIFS_ALLOW_INSECURE_LEGACY /* tell server which Unix caps we support */ @@ -4189,7 +4207,9 @@ retry: return 0; } + server->lstrp = jiffies; server->tcpStatus = CifsInNegotiate; + server->neg_start = jiffies; spin_unlock(&server->srv_lock); rc = server->ops->negotiate(xid, ses, server); diff --git a/fs/smb/client/dir.c b/fs/smb/client/dir.c index d1e95632ac54..5223edf6d11a 100644 --- a/fs/smb/client/dir.c +++ b/fs/smb/client/dir.c @@ -23,6 +23,7 @@ #include "fs_context.h" #include "cifs_ioctl.h" #include "fscache.h" +#include "cached_dir.h" static void renew_parental_timestamps(struct dentry *direntry) @@ -189,7 +190,9 @@ static int cifs_do_create(struct inode *inode, struct dentry *direntry, unsigned int disposition; struct TCP_Server_Info *server = tcon->ses->server; struct cifs_open_parms oparms; + struct cached_fid *parent_cfid = NULL; int rdwr_for_fscache = 0; + __le32 lease_flags = 0; *oplock = 0; if (tcon->ses->server->oplocks) @@ -311,7 +314,28 @@ static int cifs_do_create(struct inode *inode, struct dentry *direntry, unsigned if (!tcon->unix_ext && (mode & S_IWUGO) == 0) create_options |= CREATE_OPTION_READONLY; + retry_open: + if (tcon->cfids && direntry->d_parent && server->dialect >= SMB30_PROT_ID) { + parent_cfid = NULL; + spin_lock(&tcon->cfids->cfid_list_lock); + list_for_each_entry(parent_cfid, &tcon->cfids->entries, entry) { + if (parent_cfid->dentry == direntry->d_parent) { + cifs_dbg(FYI, "found a parent cached file handle\n"); + if (parent_cfid->has_lease && parent_cfid->time) { + lease_flags + |= SMB2_LEASE_FLAG_PARENT_LEASE_KEY_SET_LE; + memcpy(fid->parent_lease_key, + parent_cfid->fid.lease_key, + SMB2_LEASE_KEY_SIZE); + parent_cfid->dirents.is_valid = false; + } + break; + } + } + spin_unlock(&tcon->cfids->cfid_list_lock); + } + oparms = (struct cifs_open_parms) { .tcon = tcon, .cifs_sb = cifs_sb, @@ -320,6 +344,7 @@ retry_open: .disposition = disposition, .path = full_path, .fid = fid, + .lease_flags = lease_flags, .mode = mode, }; rc = server->ops->open(xid, &oparms, oplock, buf); diff --git a/fs/smb/client/file.c b/fs/smb/client/file.c index d2df10b8e6fd..1421bde045c2 100644 --- a/fs/smb/client/file.c +++ b/fs/smb/client/file.c @@ -52,6 +52,7 @@ static void cifs_prepare_write(struct netfs_io_subrequest *subreq) struct netfs_io_stream *stream = &req->rreq.io_streams[subreq->stream_nr]; struct TCP_Server_Info *server; struct cifsFileInfo *open_file = req->cfile; + struct cifs_sb_info *cifs_sb = CIFS_SB(wdata->rreq->inode->i_sb); size_t wsize = req->rreq.wsize; int rc; @@ -63,6 +64,10 @@ static void cifs_prepare_write(struct netfs_io_subrequest *subreq) server = cifs_pick_channel(tlink_tcon(open_file->tlink)->ses); wdata->server = server; + if (cifs_sb->ctx->wsize == 0) + cifs_negotiate_wsize(server, cifs_sb->ctx, + tlink_tcon(req->cfile->tlink)); + retry: if (open_file->invalidHandle) { rc = cifs_reopen_file(open_file, false); @@ -160,10 +165,9 @@ static int cifs_prepare_read(struct netfs_io_subrequest *subreq) server = cifs_pick_channel(tlink_tcon(req->cfile->tlink)->ses); rdata->server = server; - if (cifs_sb->ctx->rsize == 0) { + if (cifs_sb->ctx->rsize == 0) cifs_negotiate_rsize(server, cifs_sb->ctx, tlink_tcon(req->cfile->tlink)); - } rc = server->ops->wait_mtu_credits(server, cifs_sb->ctx->rsize, &size, &rdata->credits); @@ -999,15 +1003,18 @@ int cifs_open(struct inode *inode, struct file *file) rc = cifs_get_readable_path(tcon, full_path, &cfile); } if (rc == 0) { - if (file->f_flags == cfile->f_flags) { + unsigned int oflags = file->f_flags & ~(O_CREAT|O_EXCL|O_TRUNC); + unsigned int cflags = cfile->f_flags & ~(O_CREAT|O_EXCL|O_TRUNC); + + if (cifs_convert_flags(oflags, 0) == cifs_convert_flags(cflags, 0) && + (oflags & (O_SYNC|O_DIRECT)) == (cflags & (O_SYNC|O_DIRECT))) { file->private_data = cfile; spin_lock(&CIFS_I(inode)->deferred_lock); cifs_del_deferred_close(cfile); spin_unlock(&CIFS_I(inode)->deferred_lock); goto use_cache; - } else { - _cifsFileInfo_put(cfile, true, false); } + _cifsFileInfo_put(cfile, true, false); } else { /* hard link on the defeered close file */ rc = cifs_get_hardlink_path(tcon, inode, file); @@ -3081,7 +3088,8 @@ void cifs_oplock_break(struct work_struct *work) struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo, oplock_break); struct inode *inode = d_inode(cfile->dentry); - struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + struct super_block *sb = inode->i_sb; + struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct cifsInodeInfo *cinode = CIFS_I(inode); struct cifs_tcon *tcon; struct TCP_Server_Info *server; @@ -3091,6 +3099,12 @@ void cifs_oplock_break(struct work_struct *work) __u64 persistent_fid, volatile_fid; __u16 net_fid; + /* + * Hold a reference to the superblock to prevent it and its inodes from + * being freed while we are accessing cinode. Otherwise, _cifsFileInfo_put() + * may release the last reference to the sb and trigger inode eviction. + */ + cifs_sb_active(sb); wait_on_bit(&cinode->flags, CIFS_INODE_PENDING_WRITERS, TASK_UNINTERRUPTIBLE); @@ -3163,6 +3177,7 @@ oplock_break_ack: cifs_put_tlink(tlink); out: cifs_done_oplock_break(cinode); + cifs_sb_deactive(sb); } static int cifs_swap_activate(struct swap_info_struct *sis, diff --git a/fs/smb/client/fs_context.c b/fs/smb/client/fs_context.c index a634a34d4086..59ccc2229ab3 100644 --- a/fs/smb/client/fs_context.c +++ b/fs/smb/client/fs_context.c @@ -1824,10 +1824,14 @@ static int smb3_fs_context_parse_param(struct fs_context *fc, cifs_errorf(fc, "symlinkroot mount options must be absolute path\n"); goto cifs_parse_mount_err; } - kfree(ctx->symlinkroot); - ctx->symlinkroot = kstrdup(param->string, GFP_KERNEL); - if (!ctx->symlinkroot) + if (strnlen(param->string, PATH_MAX) == PATH_MAX) { + cifs_errorf(fc, "symlinkroot path too long (max path length: %u)\n", + PATH_MAX - 1); goto cifs_parse_mount_err; + } + kfree(ctx->symlinkroot); + ctx->symlinkroot = param->string; + param->string = NULL; break; } /* case Opt_ignore: - is ignored as expected ... */ @@ -1837,13 +1841,6 @@ static int smb3_fs_context_parse_param(struct fs_context *fc, goto cifs_parse_mount_err; } - /* - * By default resolve all native absolute symlinks relative to "/mnt/". - * Same default has drvfs driver running in WSL for resolving SMB shares. - */ - if (!ctx->symlinkroot) - ctx->symlinkroot = kstrdup("/mnt/", GFP_KERNEL); - return 0; cifs_parse_mount_err: diff --git a/fs/smb/client/ioctl.c b/fs/smb/client/ioctl.c index 56439da4f119..0a9935ce05a5 100644 --- a/fs/smb/client/ioctl.c +++ b/fs/smb/client/ioctl.c @@ -506,7 +506,7 @@ long cifs_ioctl(struct file *filep, unsigned int command, unsigned long arg) le16_to_cpu(tcon->ses->server->cipher_type); pkey_inf.Suid = tcon->ses->Suid; memcpy(pkey_inf.auth_key, tcon->ses->auth_key.response, - 16 /* SMB2_NTLMV2_SESSKEY_SIZE */); + SMB2_NTLMV2_SESSKEY_SIZE); memcpy(pkey_inf.smb3decryptionkey, tcon->ses->smb3decryptionkey, SMB3_SIGN_KEY_SIZE); memcpy(pkey_inf.smb3encryptionkey, diff --git a/fs/smb/client/misc.c b/fs/smb/client/misc.c index 7b6ed9b23e71..da23cc12a52c 100644 --- a/fs/smb/client/misc.c +++ b/fs/smb/client/misc.c @@ -151,6 +151,12 @@ tcon_info_alloc(bool dir_leases_enabled, enum smb3_tcon_ref_trace trace) #ifdef CONFIG_CIFS_DFS_UPCALL INIT_LIST_HEAD(&ret_buf->dfs_ses_list); #endif + INIT_LIST_HEAD(&ret_buf->pending_opens); + INIT_DELAYED_WORK(&ret_buf->query_interfaces, + smb2_query_server_interfaces); +#ifdef CONFIG_CIFS_DFS_UPCALL + INIT_DELAYED_WORK(&ret_buf->dfs_cache_work, dfs_cache_refresh); +#endif return ret_buf; } @@ -326,6 +332,14 @@ check_smb_hdr(struct smb_hdr *smb) if (smb->Command == SMB_COM_LOCKING_ANDX) return 0; + /* + * Windows NT server returns error resposne (e.g. STATUS_DELETE_PENDING + * or STATUS_OBJECT_NAME_NOT_FOUND or ERRDOS/ERRbadfile or any other) + * for some TRANS2 requests without the RESPONSE flag set in header. + */ + if (smb->Command == SMB_COM_TRANSACTION2 && smb->Status.CifsError != 0) + return 0; + cifs_dbg(VFS, "Server sent request, not response. mid=%u\n", get_mid(smb)); return 1; diff --git a/fs/smb/client/namespace.c b/fs/smb/client/namespace.c index 778daf11f1db..52a520349cb7 100644 --- a/fs/smb/client/namespace.c +++ b/fs/smb/client/namespace.c @@ -146,6 +146,9 @@ static char *automount_fullpath(struct dentry *dentry, void *page) } spin_unlock(&tcon->tc_lock); + if (unlikely(!page)) + return ERR_PTR(-ENOMEM); + s = dentry_path_raw(dentry, page, PATH_MAX); if (IS_ERR(s)) return s; diff --git a/fs/smb/client/readdir.c b/fs/smb/client/readdir.c index f9f11cbf89be..4e5460206397 100644 --- a/fs/smb/client/readdir.c +++ b/fs/smb/client/readdir.c @@ -264,7 +264,7 @@ cifs_posix_to_fattr(struct cifs_fattr *fattr, struct smb2_posix_info *info, /* The Mode field in the response can now include the file type as well */ fattr->cf_mode = wire_mode_to_posix(le32_to_cpu(info->Mode), fattr->cf_cifsattrs & ATTR_DIRECTORY); - fattr->cf_dtype = S_DT(le32_to_cpu(info->Mode)); + fattr->cf_dtype = S_DT(fattr->cf_mode); switch (fattr->cf_mode & S_IFMT) { case S_IFLNK: @@ -851,9 +851,9 @@ static bool emit_cached_dirents(struct cached_dirents *cde, } static void update_cached_dirents_count(struct cached_dirents *cde, - struct dir_context *ctx) + struct file *file) { - if (cde->ctx != ctx) + if (cde->file != file) return; if (cde->is_valid || cde->is_failed) return; @@ -862,9 +862,9 @@ static void update_cached_dirents_count(struct cached_dirents *cde, } static void finished_cached_dirents_count(struct cached_dirents *cde, - struct dir_context *ctx) + struct dir_context *ctx, struct file *file) { - if (cde->ctx != ctx) + if (cde->file != file) return; if (cde->is_valid || cde->is_failed) return; @@ -877,11 +877,12 @@ static void finished_cached_dirents_count(struct cached_dirents *cde, static void add_cached_dirent(struct cached_dirents *cde, struct dir_context *ctx, const char *name, int namelen, - struct cifs_fattr *fattr) + struct cifs_fattr *fattr, + struct file *file) { struct cached_dirent *de; - if (cde->ctx != ctx) + if (cde->file != file) return; if (cde->is_valid || cde->is_failed) return; @@ -911,7 +912,8 @@ static void add_cached_dirent(struct cached_dirents *cde, static bool cifs_dir_emit(struct dir_context *ctx, const char *name, int namelen, struct cifs_fattr *fattr, - struct cached_fid *cfid) + struct cached_fid *cfid, + struct file *file) { bool rc; ino_t ino = cifs_uniqueid_to_ino_t(fattr->cf_uniqueid); @@ -923,7 +925,7 @@ static bool cifs_dir_emit(struct dir_context *ctx, if (cfid) { mutex_lock(&cfid->dirents.de_mutex); add_cached_dirent(&cfid->dirents, ctx, name, namelen, - fattr); + fattr, file); mutex_unlock(&cfid->dirents.de_mutex); } @@ -1023,7 +1025,7 @@ static int cifs_filldir(char *find_entry, struct file *file, cifs_prime_dcache(file_dentry(file), &name, &fattr); return !cifs_dir_emit(ctx, name.name, name.len, - &fattr, cfid); + &fattr, cfid, file); } @@ -1074,8 +1076,8 @@ int cifs_readdir(struct file *file, struct dir_context *ctx) * we need to initialize scanning and storing the * directory content. */ - if (ctx->pos == 0 && cfid->dirents.ctx == NULL) { - cfid->dirents.ctx = ctx; + if (ctx->pos == 0 && cfid->dirents.file == NULL) { + cfid->dirents.file = file; cfid->dirents.pos = 2; } /* @@ -1143,7 +1145,7 @@ int cifs_readdir(struct file *file, struct dir_context *ctx) } else { if (cfid) { mutex_lock(&cfid->dirents.de_mutex); - finished_cached_dirents_count(&cfid->dirents, ctx); + finished_cached_dirents_count(&cfid->dirents, ctx, file); mutex_unlock(&cfid->dirents.de_mutex); } cifs_dbg(FYI, "Could not find entry\n"); @@ -1184,7 +1186,7 @@ int cifs_readdir(struct file *file, struct dir_context *ctx) ctx->pos++; if (cfid) { mutex_lock(&cfid->dirents.de_mutex); - update_cached_dirents_count(&cfid->dirents, ctx); + update_cached_dirents_count(&cfid->dirents, file); mutex_unlock(&cfid->dirents.de_mutex); } diff --git a/fs/smb/client/reparse.c b/fs/smb/client/reparse.c index bb25e77c5540..5fa29a97ac15 100644 --- a/fs/smb/client/reparse.c +++ b/fs/smb/client/reparse.c @@ -57,6 +57,7 @@ static int create_native_symlink(const unsigned int xid, struct inode *inode, struct reparse_symlink_data_buffer *buf = NULL; struct cifs_open_info_data data = {}; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + const char *symroot = cifs_sb->ctx->symlinkroot; struct inode *new; struct kvec iov; __le16 *path = NULL; @@ -82,7 +83,8 @@ static int create_native_symlink(const unsigned int xid, struct inode *inode, .symlink_target = symlink_target, }; - if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) && symname[0] == '/') { + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) && + symroot && symname[0] == '/') { /* * This is a request to create an absolute symlink on the server * which does not support POSIX paths, and expects symlink in @@ -92,7 +94,7 @@ static int create_native_symlink(const unsigned int xid, struct inode *inode, * ensure compatibility of this symlink stored in absolute form * on the SMB server. */ - if (!strstarts(symname, cifs_sb->ctx->symlinkroot)) { + if (!strstarts(symname, symroot)) { /* * If the absolute Linux symlink target path is not * inside "symlinkroot" location then there is no way @@ -101,12 +103,12 @@ static int create_native_symlink(const unsigned int xid, struct inode *inode, cifs_dbg(VFS, "absolute symlink '%s' cannot be converted to NT format " "because it is outside of symlinkroot='%s'\n", - symname, cifs_sb->ctx->symlinkroot); + symname, symroot); rc = -EINVAL; goto out; } - len = strlen(cifs_sb->ctx->symlinkroot); - if (cifs_sb->ctx->symlinkroot[len-1] != '/') + len = strlen(symroot); + if (symroot[len - 1] != '/') len++; if (symname[len] >= 'a' && symname[len] <= 'z' && (symname[len+1] == '/' || symname[len+1] == '\0')) { @@ -782,6 +784,7 @@ int smb2_parse_native_symlink(char **target, const char *buf, unsigned int len, const char *full_path, struct cifs_sb_info *cifs_sb) { + const char *symroot = cifs_sb->ctx->symlinkroot; char sep = CIFS_DIR_SEP(cifs_sb); char *linux_target = NULL; char *smb_target = NULL; @@ -815,7 +818,8 @@ int smb2_parse_native_symlink(char **target, const char *buf, unsigned int len, goto out; } - if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) && !relative) { + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) && + symroot && !relative) { /* * This is an absolute symlink from the server which does not * support POSIX paths, so the symlink is in NT-style path. @@ -875,15 +879,8 @@ globalroot: abs_path += sizeof("\\DosDevices\\")-1; else if (strstarts(abs_path, "\\GLOBAL??\\")) abs_path += sizeof("\\GLOBAL??\\")-1; - else { - /* Unhandled absolute symlink, points outside of DOS/Win32 */ - cifs_dbg(VFS, - "absolute symlink '%s' cannot be converted from NT format " - "because points to unknown target\n", - smb_target); - rc = -EIO; - goto out; - } + else + goto out_unhandled_target; /* Sometimes path separator after \?? is double backslash */ if (abs_path[0] == '\\') @@ -910,25 +907,19 @@ globalroot: abs_path++; abs_path[0] = drive_letter; } else { - /* Unhandled absolute symlink. Report an error. */ - cifs_dbg(VFS, - "absolute symlink '%s' cannot be converted from NT format " - "because points to unknown target\n", - smb_target); - rc = -EIO; - goto out; + goto out_unhandled_target; } abs_path_len = strlen(abs_path)+1; - symlinkroot_len = strlen(cifs_sb->ctx->symlinkroot); - if (cifs_sb->ctx->symlinkroot[symlinkroot_len-1] == '/') + symlinkroot_len = strlen(symroot); + if (symroot[symlinkroot_len - 1] == '/') symlinkroot_len--; linux_target = kmalloc(symlinkroot_len + 1 + abs_path_len, GFP_KERNEL); if (!linux_target) { rc = -ENOMEM; goto out; } - memcpy(linux_target, cifs_sb->ctx->symlinkroot, symlinkroot_len); + memcpy(linux_target, symroot, symlinkroot_len); linux_target[symlinkroot_len] = '/'; memcpy(linux_target + symlinkroot_len + 1, abs_path, abs_path_len); } else if (smb_target[0] == sep && relative) { @@ -966,6 +957,7 @@ globalroot: * These paths have same format as Linux symlinks, so no * conversion is needed. */ +out_unhandled_target: linux_target = smb_target; smb_target = NULL; } @@ -1172,7 +1164,6 @@ out: if (!have_xattr_dev && (tag == IO_REPARSE_TAG_LX_CHR || tag == IO_REPARSE_TAG_LX_BLK)) return false; - fattr->cf_dtype = S_DT(fattr->cf_mode); return true; } diff --git a/fs/smb/client/sess.c b/fs/smb/client/sess.c index b3fa9ee26912..330bc3d25bad 100644 --- a/fs/smb/client/sess.c +++ b/fs/smb/client/sess.c @@ -445,6 +445,10 @@ cifs_chan_update_iface(struct cifs_ses *ses, struct TCP_Server_Info *server) ses->chans[chan_index].iface = iface; spin_unlock(&ses->chan_lock); + + spin_lock(&server->srv_lock); + memcpy(&server->dstaddr, &iface->sockaddr, sizeof(server->dstaddr)); + spin_unlock(&server->srv_lock); } static int @@ -494,8 +498,7 @@ cifs_ses_add_channel(struct cifs_ses *ses, ctx->domainauto = ses->domainAuto; ctx->domainname = ses->domainName; - /* no hostname for extra channels */ - ctx->server_hostname = ""; + ctx->server_hostname = ses->server->hostname; ctx->username = ses->user_name; ctx->password = ses->password; @@ -628,6 +631,7 @@ static __u32 cifs_ssetup_hdr(struct cifs_ses *ses, USHRT_MAX)); pSMB->req.MaxMpxCount = cpu_to_le16(server->maxReq); pSMB->req.VcNumber = cpu_to_le16(1); + pSMB->req.SessionKey = server->session_key_id; /* Now no need to set SMBFLG_CASELESS or obsolete CANONICAL PATH */ @@ -1684,22 +1688,22 @@ _sess_auth_rawntlmssp_assemble_req(struct sess_data *sess_data) pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base; capabilities = cifs_ssetup_hdr(ses, server, pSMB); - if ((pSMB->req.hdr.Flags2 & SMBFLG2_UNICODE) == 0) { - cifs_dbg(VFS, "NTLMSSP requires Unicode support\n"); - return -ENOSYS; - } - pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; capabilities |= CAP_EXTENDED_SECURITY; pSMB->req.Capabilities |= cpu_to_le32(capabilities); bcc_ptr = sess_data->iov[2].iov_base; - /* unicode strings must be word aligned */ - if (!IS_ALIGNED(sess_data->iov[0].iov_len + sess_data->iov[1].iov_len, 2)) { - *bcc_ptr = 0; - bcc_ptr++; + + if (pSMB->req.hdr.Flags2 & SMBFLG2_UNICODE) { + /* unicode strings must be word aligned */ + if (!IS_ALIGNED(sess_data->iov[0].iov_len + sess_data->iov[1].iov_len, 2)) { + *bcc_ptr = 0; + bcc_ptr++; + } + unicode_oslm_strings(&bcc_ptr, sess_data->nls_cp); + } else { + ascii_oslm_strings(&bcc_ptr, sess_data->nls_cp); } - unicode_oslm_strings(&bcc_ptr, sess_data->nls_cp); sess_data->iov[2].iov_len = (long) bcc_ptr - (long) sess_data->iov[2].iov_base; diff --git a/fs/smb/client/smb2inode.c b/fs/smb/client/smb2inode.c index 2a3e46b8e15a..a11a2a693c51 100644 --- a/fs/smb/client/smb2inode.c +++ b/fs/smb/client/smb2inode.c @@ -1346,7 +1346,8 @@ struct inode *smb2_get_reparse_inode(struct cifs_open_info_data *data, * empty object on the server. */ if (!(le32_to_cpu(tcon->fsAttrInfo.Attributes) & FILE_SUPPORTS_REPARSE_POINTS)) - return ERR_PTR(-EOPNOTSUPP); + if (!tcon->posix_extensions) + return ERR_PTR(-EOPNOTSUPP); oparms = CIFS_OPARMS(cifs_sb, tcon, full_path, SYNCHRONIZE | DELETE | diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index 2fe8eeb98535..938a8a7c5d21 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -504,6 +504,9 @@ smb3_negotiate_wsize(struct cifs_tcon *tcon, struct smb3_fs_context *ctx) wsize = min_t(unsigned int, wsize, server->max_write); #ifdef CONFIG_CIFS_SMB_DIRECT if (server->rdma) { + struct smbdirect_socket_parameters *sp = + &server->smbd_conn->socket.parameters; + if (server->sign) /* * Account for SMB2 data transfer packet header and @@ -511,12 +514,12 @@ smb3_negotiate_wsize(struct cifs_tcon *tcon, struct smb3_fs_context *ctx) */ wsize = min_t(unsigned int, wsize, - server->smbd_conn->max_fragmented_send_size - + sp->max_fragmented_send_size - SMB2_READWRITE_PDU_HEADER_SIZE - sizeof(struct smb2_transform_hdr)); else wsize = min_t(unsigned int, - wsize, server->smbd_conn->max_readwrite_size); + wsize, sp->max_read_write_size); } #endif if (!(server->capabilities & SMB2_GLOBAL_CAP_LARGE_MTU)) @@ -552,6 +555,9 @@ smb3_negotiate_rsize(struct cifs_tcon *tcon, struct smb3_fs_context *ctx) rsize = min_t(unsigned int, rsize, server->max_read); #ifdef CONFIG_CIFS_SMB_DIRECT if (server->rdma) { + struct smbdirect_socket_parameters *sp = + &server->smbd_conn->socket.parameters; + if (server->sign) /* * Account for SMB2 data transfer packet header and @@ -559,12 +565,12 @@ smb3_negotiate_rsize(struct cifs_tcon *tcon, struct smb3_fs_context *ctx) */ rsize = min_t(unsigned int, rsize, - server->smbd_conn->max_fragmented_recv_size - + sp->max_fragmented_recv_size - SMB2_READWRITE_PDU_HEADER_SIZE - sizeof(struct smb2_transform_hdr)); else rsize = min_t(unsigned int, - rsize, server->smbd_conn->max_readwrite_size); + rsize, sp->max_read_write_size); } #endif @@ -4069,7 +4075,7 @@ map_oplock_to_lease(u8 oplock) } static char * -smb2_create_lease_buf(u8 *lease_key, u8 oplock) +smb2_create_lease_buf(u8 *lease_key, u8 oplock, u8 *parent_lease_key, __le32 flags) { struct create_lease *buf; @@ -4095,7 +4101,7 @@ smb2_create_lease_buf(u8 *lease_key, u8 oplock) } static char * -smb3_create_lease_buf(u8 *lease_key, u8 oplock) +smb3_create_lease_buf(u8 *lease_key, u8 oplock, u8 *parent_lease_key, __le32 flags) { struct create_lease_v2 *buf; @@ -4105,6 +4111,9 @@ smb3_create_lease_buf(u8 *lease_key, u8 oplock) memcpy(&buf->lcontext.LeaseKey, lease_key, SMB2_LEASE_KEY_SIZE); buf->lcontext.LeaseState = map_oplock_to_lease(oplock); + buf->lcontext.LeaseFlags = flags; + if (flags & SMB2_LEASE_FLAG_PARENT_LEASE_KEY_SET_LE) + memcpy(&buf->lcontext.ParentLeaseKey, parent_lease_key, SMB2_LEASE_KEY_SIZE); buf->ccontext.DataOffset = cpu_to_le16(offsetof (struct create_lease_v2, lcontext)); @@ -4307,6 +4316,7 @@ crypt_message(struct TCP_Server_Info *server, int num_rqst, u8 key[SMB3_ENC_DEC_KEY_SIZE]; struct aead_request *req; u8 *iv; + DECLARE_CRYPTO_WAIT(wait); unsigned int crypt_len = le32_to_cpu(tr_hdr->OriginalMessageSize); void *creq; size_t sensitive_size; @@ -4357,7 +4367,11 @@ crypt_message(struct TCP_Server_Info *server, int num_rqst, aead_request_set_crypt(req, sg, sg, crypt_len, iv); aead_request_set_ad(req, assoc_data_len); - rc = enc ? crypto_aead_encrypt(req) : crypto_aead_decrypt(req); + aead_request_set_callback(req, CRYPTO_TFM_REQ_MAY_BACKLOG, + crypto_req_done, &wait); + + rc = crypto_wait_req(enc ? crypto_aead_encrypt(req) + : crypto_aead_decrypt(req), &wait); if (!rc && enc) memcpy(&tr_hdr->Signature, sign, SMB2_SIGNATURE_SIZE); @@ -5246,7 +5260,8 @@ static int smb2_make_node(unsigned int xid, struct inode *inode, if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) { rc = cifs_sfu_make_node(xid, inode, dentry, tcon, full_path, mode, dev); - } else if (le32_to_cpu(tcon->fsAttrInfo.Attributes) & FILE_SUPPORTS_REPARSE_POINTS) { + } else if ((le32_to_cpu(tcon->fsAttrInfo.Attributes) & FILE_SUPPORTS_REPARSE_POINTS) + || (tcon->posix_extensions)) { rc = smb2_mknod_reparse(xid, inode, dentry, tcon, full_path, mode, dev); } diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c index 399185ca7cac..2df93a75e3b8 100644 --- a/fs/smb/client/smb2pdu.c +++ b/fs/smb/client/smb2pdu.c @@ -36,6 +36,7 @@ #include "smb2glob.h" #include "cifspdu.h" #include "cifs_spnego.h" +#include "../common/smbdirect/smbdirect.h" #include "smbdirect.h" #include "trace.h" #ifdef CONFIG_CIFS_DFS_UPCALL @@ -411,14 +412,23 @@ skip_sess_setup: if (!rc && (server->capabilities & SMB2_GLOBAL_CAP_MULTI_CHANNEL) && server->ops->query_server_interfaces) { - mutex_unlock(&ses->session_mutex); - /* - * query server network interfaces, in case they change + * query server network interfaces, in case they change. + * Also mark the session as pending this update while the query + * is in progress. This will be used to avoid calling + * smb2_reconnect recursively. */ + ses->flags |= CIFS_SES_FLAGS_PENDING_QUERY_INTERFACES; xid = get_xid(); rc = server->ops->query_server_interfaces(xid, tcon, false); free_xid(xid); + ses->flags &= ~CIFS_SES_FLAGS_PENDING_QUERY_INTERFACES; + + if (!tcon->ipc && !tcon->dummy) + queue_delayed_work(cifsiod_wq, &tcon->query_interfaces, + (SMB_INTERFACE_POLL_INTERVAL * HZ)); + + mutex_unlock(&ses->session_mutex); if (rc == -EOPNOTSUPP && ses->chan_count > 1) { /* @@ -438,11 +448,8 @@ skip_sess_setup: if (ses->chan_max > ses->chan_count && ses->iface_count && !SERVER_IS_CHAN(server)) { - if (ses->chan_count == 1) { + if (ses->chan_count == 1) cifs_server_dbg(VFS, "supports multichannel now\n"); - queue_delayed_work(cifsiod_wq, &tcon->query_interfaces, - (SMB_INTERFACE_POLL_INTERVAL * HZ)); - } cifs_try_adding_channels(ses); } @@ -560,11 +567,18 @@ static int smb2_ioctl_req_init(u32 opcode, struct cifs_tcon *tcon, struct TCP_Server_Info *server, void **request_buf, unsigned int *total_len) { - /* Skip reconnect only for FSCTL_VALIDATE_NEGOTIATE_INFO IOCTLs */ - if (opcode == FSCTL_VALIDATE_NEGOTIATE_INFO) { + /* + * Skip reconnect in one of the following cases: + * 1. For FSCTL_VALIDATE_NEGOTIATE_INFO IOCTLs + * 2. For FSCTL_QUERY_NETWORK_INTERFACE_INFO IOCTL when called from + * smb2_reconnect (indicated by CIFS_SES_FLAG_SCALE_CHANNELS ses flag) + */ + if (opcode == FSCTL_VALIDATE_NEGOTIATE_INFO || + (opcode == FSCTL_QUERY_NETWORK_INTERFACE_INFO && + (tcon->ses->flags & CIFS_SES_FLAGS_PENDING_QUERY_INTERFACES))) return __smb2_plain_req_init(SMB2_IOCTL, tcon, server, request_buf, total_len); - } + return smb2_plain_req_init(SMB2_IOCTL, tcon, server, request_buf, total_len); } @@ -2392,11 +2406,16 @@ static int add_lease_context(struct TCP_Server_Info *server, struct smb2_create_req *req, struct kvec *iov, - unsigned int *num_iovec, u8 *lease_key, __u8 *oplock) + unsigned int *num_iovec, + u8 *lease_key, + __u8 *oplock, + u8 *parent_lease_key, + __le32 flags) { unsigned int num = *num_iovec; - iov[num].iov_base = server->ops->create_lease_buf(lease_key, *oplock); + iov[num].iov_base = server->ops->create_lease_buf(lease_key, *oplock, + parent_lease_key, flags); if (iov[num].iov_base == NULL) return -ENOMEM; iov[num].iov_len = server->vals->create_lease_size; @@ -3069,7 +3088,9 @@ SMB2_open_init(struct cifs_tcon *tcon, struct TCP_Server_Info *server, req->RequestedOplockLevel = *oplock; /* no srv lease support */ else { rc = add_lease_context(server, req, iov, &n_iov, - oparms->fid->lease_key, oplock); + oparms->fid->lease_key, oplock, + oparms->fid->parent_lease_key, + oparms->lease_flags); if (rc) return rc; } @@ -4208,10 +4229,8 @@ void smb2_reconnect_server(struct work_struct *work) } goto done; } - tcon->status = TID_GOOD; - tcon->retry = false; - tcon->need_reconnect = false; + tcon->dummy = true; /* now reconnect sessions for necessary channels */ list_for_each_entry_safe(ses, ses2, &tmp_ses_list, rlist) { @@ -4442,10 +4461,10 @@ smb2_new_read_req(void **buf, unsigned int *total_len, #ifdef CONFIG_CIFS_SMB_DIRECT /* * If we want to do a RDMA write, fill in and append - * smbd_buffer_descriptor_v1 to the end of read request + * smbdirect_buffer_descriptor_v1 to the end of read request */ if (rdata && smb3_use_rdma_offload(io_parms)) { - struct smbd_buffer_descriptor_v1 *v1; + struct smbdirect_buffer_descriptor_v1 *v1; bool need_invalidate = server->dialect == SMB30_PROT_ID; rdata->mr = smbd_register_mr(server->smbd_conn, &rdata->subreq.io_iter, @@ -4459,8 +4478,8 @@ smb2_new_read_req(void **buf, unsigned int *total_len, req->ReadChannelInfoOffset = cpu_to_le16(offsetof(struct smb2_read_req, Buffer)); req->ReadChannelInfoLength = - cpu_to_le16(sizeof(struct smbd_buffer_descriptor_v1)); - v1 = (struct smbd_buffer_descriptor_v1 *) &req->Buffer[0]; + cpu_to_le16(sizeof(struct smbdirect_buffer_descriptor_v1)); + v1 = (struct smbdirect_buffer_descriptor_v1 *) &req->Buffer[0]; v1->offset = cpu_to_le64(rdata->mr->mr->iova); v1->token = cpu_to_le32(rdata->mr->mr->rkey); v1->length = cpu_to_le32(rdata->mr->mr->length); @@ -4546,7 +4565,11 @@ smb2_readv_callback(struct mid_q_entry *mid) cifs_stats_bytes_read(tcon, rdata->got_bytes); break; case MID_REQUEST_SUBMITTED: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_req_submitted); + goto do_retry; case MID_RETRY_NEEDED: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_retry_needed); +do_retry: __set_bit(NETFS_SREQ_NEED_RETRY, &rdata->subreq.flags); rdata->result = -EAGAIN; if (server->sign && rdata->got_bytes) @@ -4557,11 +4580,15 @@ smb2_readv_callback(struct mid_q_entry *mid) cifs_stats_bytes_read(tcon, rdata->got_bytes); break; case MID_RESPONSE_MALFORMED: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_malformed); credits.value = le16_to_cpu(shdr->CreditRequest); credits.instance = server->reconnect_instance; - fallthrough; + rdata->result = -EIO; + break; default: + trace_netfs_sreq(&rdata->subreq, netfs_sreq_trace_io_unknown); rdata->result = -EIO; + break; } #ifdef CONFIG_CIFS_SMB_DIRECT /* @@ -4814,11 +4841,14 @@ smb2_writev_callback(struct mid_q_entry *mid) switch (mid->mid_state) { case MID_RESPONSE_RECEIVED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_progress); credits.value = le16_to_cpu(rsp->hdr.CreditRequest); credits.instance = server->reconnect_instance; result = smb2_check_receive(mid, server, 0); - if (result != 0) + if (result != 0) { + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_bad); break; + } written = le32_to_cpu(rsp->DataLength); /* @@ -4840,14 +4870,23 @@ smb2_writev_callback(struct mid_q_entry *mid) } break; case MID_REQUEST_SUBMITTED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_req_submitted); + __set_bit(NETFS_SREQ_NEED_RETRY, &wdata->subreq.flags); + result = -EAGAIN; + break; case MID_RETRY_NEEDED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_retry_needed); + __set_bit(NETFS_SREQ_NEED_RETRY, &wdata->subreq.flags); result = -EAGAIN; break; case MID_RESPONSE_MALFORMED: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_malformed); credits.value = le16_to_cpu(rsp->hdr.CreditRequest); credits.instance = server->reconnect_instance; - fallthrough; + result = -EIO; + break; default: + trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_unknown); result = -EIO; break; } @@ -4887,7 +4926,6 @@ smb2_writev_callback(struct mid_q_entry *mid) server->credits, server->in_flight, 0, cifs_trace_rw_credits_write_response_clear); wdata->credits.value = 0; - trace_netfs_sreq(&wdata->subreq, netfs_sreq_trace_io_progress); cifs_write_subrequest_terminated(wdata, result ?: written); release_mid(mid); trace_smb3_rw_credits(rreq_debug_id, subreq_debug_index, 0, @@ -4968,10 +5006,10 @@ smb2_async_writev(struct cifs_io_subrequest *wdata) #ifdef CONFIG_CIFS_SMB_DIRECT /* * If we want to do a server RDMA read, fill in and append - * smbd_buffer_descriptor_v1 to the end of write request + * smbdirect_buffer_descriptor_v1 to the end of write request */ if (smb3_use_rdma_offload(io_parms)) { - struct smbd_buffer_descriptor_v1 *v1; + struct smbdirect_buffer_descriptor_v1 *v1; bool need_invalidate = server->dialect == SMB30_PROT_ID; wdata->mr = smbd_register_mr(server->smbd_conn, &wdata->subreq.io_iter, @@ -4990,8 +5028,8 @@ smb2_async_writev(struct cifs_io_subrequest *wdata) req->WriteChannelInfoOffset = cpu_to_le16(offsetof(struct smb2_write_req, Buffer)); req->WriteChannelInfoLength = - cpu_to_le16(sizeof(struct smbd_buffer_descriptor_v1)); - v1 = (struct smbd_buffer_descriptor_v1 *) &req->Buffer[0]; + cpu_to_le16(sizeof(struct smbdirect_buffer_descriptor_v1)); + v1 = (struct smbdirect_buffer_descriptor_v1 *) &req->Buffer[0]; v1->offset = cpu_to_le64(wdata->mr->mr->iova); v1->token = cpu_to_le32(wdata->mr->mr->rkey); v1->length = cpu_to_le32(wdata->mr->mr->length); @@ -5917,71 +5955,6 @@ posix_qfsinf_exit: } int -SMB2_QFS_info(const unsigned int xid, struct cifs_tcon *tcon, - u64 persistent_fid, u64 volatile_fid, struct kstatfs *fsdata) -{ - struct smb_rqst rqst; - struct smb2_query_info_rsp *rsp = NULL; - struct kvec iov; - struct kvec rsp_iov; - int rc = 0; - int resp_buftype; - struct cifs_ses *ses = tcon->ses; - struct TCP_Server_Info *server; - struct smb2_fs_full_size_info *info = NULL; - int flags = 0; - int retries = 0, cur_sleep = 1; - -replay_again: - /* reinitialize for possible replay */ - flags = 0; - server = cifs_pick_channel(ses); - - rc = build_qfs_info_req(&iov, tcon, server, - FS_FULL_SIZE_INFORMATION, - sizeof(struct smb2_fs_full_size_info), - persistent_fid, volatile_fid); - if (rc) - return rc; - - if (smb3_encryption_required(tcon)) - flags |= CIFS_TRANSFORM_REQ; - - memset(&rqst, 0, sizeof(struct smb_rqst)); - rqst.rq_iov = &iov; - rqst.rq_nvec = 1; - - if (retries) - smb2_set_replay(server, &rqst); - - rc = cifs_send_recv(xid, ses, server, - &rqst, &resp_buftype, flags, &rsp_iov); - free_qfs_info_req(&iov); - if (rc) { - cifs_stats_fail_inc(tcon, SMB2_QUERY_INFO_HE); - goto qfsinf_exit; - } - rsp = (struct smb2_query_info_rsp *)rsp_iov.iov_base; - - info = (struct smb2_fs_full_size_info *)( - le16_to_cpu(rsp->OutputBufferOffset) + (char *)rsp); - rc = smb2_validate_iov(le16_to_cpu(rsp->OutputBufferOffset), - le32_to_cpu(rsp->OutputBufferLength), &rsp_iov, - sizeof(struct smb2_fs_full_size_info)); - if (!rc) - smb2_copy_fs_info_to_kstatfs(info, fsdata); - -qfsinf_exit: - free_rsp_buf(resp_buftype, rsp_iov.iov_base); - - if (is_replayable_error(rc) && - smb2_should_replay(tcon, &retries, &cur_sleep)) - goto replay_again; - - return rc; -} - -int SMB2_QFS_attr(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_fid, u64 volatile_fid, int level) { diff --git a/fs/smb/client/smb2proto.h b/fs/smb/client/smb2proto.h index 4662c7e2d259..035aa1624053 100644 --- a/fs/smb/client/smb2proto.h +++ b/fs/smb/client/smb2proto.h @@ -259,9 +259,6 @@ extern int smb2_handle_cancelled_close(struct cifs_tcon *tcon, __u64 volatile_fid); extern int smb2_handle_cancelled_mid(struct mid_q_entry *mid, struct TCP_Server_Info *server); void smb2_cancelled_close_fid(struct work_struct *work); -extern int SMB2_QFS_info(const unsigned int xid, struct cifs_tcon *tcon, - u64 persistent_file_id, u64 volatile_file_id, - struct kstatfs *FSData); extern int SMB311_posix_qfs_info(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_file_id, u64 volatile_file_id, struct kstatfs *FSData); diff --git a/fs/smb/client/smbdirect.c b/fs/smb/client/smbdirect.c index b0b7254661e9..754e94a0e07f 100644 --- a/fs/smb/client/smbdirect.c +++ b/fs/smb/client/smbdirect.c @@ -7,6 +7,7 @@ #include <linux/module.h> #include <linux/highmem.h> #include <linux/folio_queue.h> +#include "../common/smbdirect/smbdirect_pdu.h" #include "smbdirect.h" #include "cifs_debug.h" #include "cifsproto.h" @@ -50,9 +51,6 @@ struct smb_extract_to_rdma { static ssize_t smb_extract_iter_to_rdma(struct iov_iter *iter, size_t len, struct smb_extract_to_rdma *rdma); -/* SMBD version number */ -#define SMBD_V1 0x0100 - /* Port numbers for SMBD transport */ #define SMB_PORT 445 #define SMBD_PORT 5445 @@ -165,10 +163,11 @@ static void smbd_disconnect_rdma_work(struct work_struct *work) { struct smbd_connection *info = container_of(work, struct smbd_connection, disconnect_work); + struct smbdirect_socket *sc = &info->socket; - if (info->transport_status == SMBD_CONNECTED) { - info->transport_status = SMBD_DISCONNECTING; - rdma_disconnect(info->id); + if (sc->status == SMBDIRECT_SOCKET_CONNECTED) { + sc->status = SMBDIRECT_SOCKET_DISCONNECTING; + rdma_disconnect(sc->rdma.cm_id); } } @@ -182,6 +181,7 @@ static int smbd_conn_upcall( struct rdma_cm_id *id, struct rdma_cm_event *event) { struct smbd_connection *info = id->context; + struct smbdirect_socket *sc = &info->socket; log_rdma_event(INFO, "event=%d status=%d\n", event->event, event->status); @@ -205,7 +205,7 @@ static int smbd_conn_upcall( case RDMA_CM_EVENT_ESTABLISHED: log_rdma_event(INFO, "connected event=%d\n", event->event); - info->transport_status = SMBD_CONNECTED; + sc->status = SMBDIRECT_SOCKET_CONNECTED; wake_up_interruptible(&info->conn_wait); break; @@ -213,20 +213,20 @@ static int smbd_conn_upcall( case RDMA_CM_EVENT_UNREACHABLE: case RDMA_CM_EVENT_REJECTED: log_rdma_event(INFO, "connecting failed event=%d\n", event->event); - info->transport_status = SMBD_DISCONNECTED; + sc->status = SMBDIRECT_SOCKET_DISCONNECTED; wake_up_interruptible(&info->conn_wait); break; case RDMA_CM_EVENT_DEVICE_REMOVAL: case RDMA_CM_EVENT_DISCONNECTED: /* This happens when we fail the negotiation */ - if (info->transport_status == SMBD_NEGOTIATE_FAILED) { - info->transport_status = SMBD_DISCONNECTED; + if (sc->status == SMBDIRECT_SOCKET_NEGOTIATE_FAILED) { + sc->status = SMBDIRECT_SOCKET_DISCONNECTED; wake_up(&info->conn_wait); break; } - info->transport_status = SMBD_DISCONNECTED; + sc->status = SMBDIRECT_SOCKET_DISCONNECTED; wake_up_interruptible(&info->disconn_wait); wake_up_interruptible(&info->wait_reassembly_queue); wake_up_interruptible_all(&info->wait_send_queue); @@ -275,6 +275,8 @@ static void send_done(struct ib_cq *cq, struct ib_wc *wc) int i; struct smbd_request *request = container_of(wc->wr_cqe, struct smbd_request, cqe); + struct smbd_connection *info = request->info; + struct smbdirect_socket *sc = &info->socket; log_rdma_send(INFO, "smbd_request 0x%p completed wc->status=%d\n", request, wc->status); @@ -286,7 +288,7 @@ static void send_done(struct ib_cq *cq, struct ib_wc *wc) } for (i = 0; i < request->num_sge; i++) - ib_dma_unmap_single(request->info->id->device, + ib_dma_unmap_single(sc->ib.dev, request->sge[i].addr, request->sge[i].length, DMA_TO_DEVICE); @@ -299,7 +301,7 @@ static void send_done(struct ib_cq *cq, struct ib_wc *wc) mempool_free(request, request->info->request_mempool); } -static void dump_smbd_negotiate_resp(struct smbd_negotiate_resp *resp) +static void dump_smbdirect_negotiate_resp(struct smbdirect_negotiate_resp *resp) { log_rdma_event(INFO, "resp message min_version %u max_version %u negotiated_version %u credits_requested %u credits_granted %u status %u max_readwrite_size %u preferred_send_size %u max_receive_size %u max_fragmented_size %u\n", resp->min_version, resp->max_version, @@ -318,15 +320,17 @@ static bool process_negotiation_response( struct smbd_response *response, int packet_length) { struct smbd_connection *info = response->info; - struct smbd_negotiate_resp *packet = smbd_response_payload(response); + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; + struct smbdirect_negotiate_resp *packet = smbd_response_payload(response); - if (packet_length < sizeof(struct smbd_negotiate_resp)) { + if (packet_length < sizeof(struct smbdirect_negotiate_resp)) { log_rdma_event(ERR, "error: packet_length=%d\n", packet_length); return false; } - if (le16_to_cpu(packet->negotiated_version) != SMBD_V1) { + if (le16_to_cpu(packet->negotiated_version) != SMBDIRECT_V1) { log_rdma_event(ERR, "error: negotiated_version=%x\n", le16_to_cpu(packet->negotiated_version)); return false; @@ -347,20 +351,20 @@ static bool process_negotiation_response( atomic_set(&info->receive_credits, 0); - if (le32_to_cpu(packet->preferred_send_size) > info->max_receive_size) { + if (le32_to_cpu(packet->preferred_send_size) > sp->max_recv_size) { log_rdma_event(ERR, "error: preferred_send_size=%d\n", le32_to_cpu(packet->preferred_send_size)); return false; } - info->max_receive_size = le32_to_cpu(packet->preferred_send_size); + sp->max_recv_size = le32_to_cpu(packet->preferred_send_size); if (le32_to_cpu(packet->max_receive_size) < SMBD_MIN_RECEIVE_SIZE) { log_rdma_event(ERR, "error: max_receive_size=%d\n", le32_to_cpu(packet->max_receive_size)); return false; } - info->max_send_size = min_t(int, info->max_send_size, - le32_to_cpu(packet->max_receive_size)); + sp->max_send_size = min_t(u32, sp->max_send_size, + le32_to_cpu(packet->max_receive_size)); if (le32_to_cpu(packet->max_fragmented_size) < SMBD_MIN_FRAGMENTED_SIZE) { @@ -368,18 +372,18 @@ static bool process_negotiation_response( le32_to_cpu(packet->max_fragmented_size)); return false; } - info->max_fragmented_send_size = + sp->max_fragmented_send_size = le32_to_cpu(packet->max_fragmented_size); info->rdma_readwrite_threshold = - rdma_readwrite_threshold > info->max_fragmented_send_size ? - info->max_fragmented_send_size : + rdma_readwrite_threshold > sp->max_fragmented_send_size ? + sp->max_fragmented_send_size : rdma_readwrite_threshold; - info->max_readwrite_size = min_t(u32, + sp->max_read_write_size = min_t(u32, le32_to_cpu(packet->max_readwrite_size), info->max_frmr_depth * PAGE_SIZE); - info->max_frmr_depth = info->max_readwrite_size / PAGE_SIZE; + info->max_frmr_depth = sp->max_read_write_size / PAGE_SIZE; return true; } @@ -393,8 +397,9 @@ static void smbd_post_send_credits(struct work_struct *work) struct smbd_connection *info = container_of(work, struct smbd_connection, post_send_credits_work); + struct smbdirect_socket *sc = &info->socket; - if (info->transport_status != SMBD_CONNECTED) { + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) { wake_up(&info->wait_receive_queues); return; } @@ -448,7 +453,7 @@ static void smbd_post_send_credits(struct work_struct *work) /* Called from softirq, when recv is done */ static void recv_done(struct ib_cq *cq, struct ib_wc *wc) { - struct smbd_data_transfer *data_transfer; + struct smbdirect_data_transfer *data_transfer; struct smbd_response *response = container_of(wc->wr_cqe, struct smbd_response, cqe); struct smbd_connection *info = response->info; @@ -474,7 +479,7 @@ static void recv_done(struct ib_cq *cq, struct ib_wc *wc) switch (response->type) { /* SMBD negotiation response */ case SMBD_NEGOTIATE_RESP: - dump_smbd_negotiate_resp(smbd_response_payload(response)); + dump_smbdirect_negotiate_resp(smbd_response_payload(response)); info->full_packet_received = true; info->negotiate_done = process_negotiation_response(response, wc->byte_len); @@ -531,7 +536,7 @@ static void recv_done(struct ib_cq *cq, struct ib_wc *wc) /* Send a KEEP_ALIVE response right away if requested */ info->keep_alive_requested = KEEP_ALIVE_NONE; if (le16_to_cpu(data_transfer->flags) & - SMB_DIRECT_RESPONSE_REQUESTED) { + SMBDIRECT_FLAG_RESPONSE_REQUESTED) { info->keep_alive_requested = KEEP_ALIVE_PENDING; } @@ -635,32 +640,34 @@ static int smbd_ia_open( struct smbd_connection *info, struct sockaddr *dstaddr, int port) { + struct smbdirect_socket *sc = &info->socket; int rc; - info->id = smbd_create_id(info, dstaddr, port); - if (IS_ERR(info->id)) { - rc = PTR_ERR(info->id); + sc->rdma.cm_id = smbd_create_id(info, dstaddr, port); + if (IS_ERR(sc->rdma.cm_id)) { + rc = PTR_ERR(sc->rdma.cm_id); goto out1; } + sc->ib.dev = sc->rdma.cm_id->device; - if (!frwr_is_supported(&info->id->device->attrs)) { + if (!frwr_is_supported(&sc->ib.dev->attrs)) { log_rdma_event(ERR, "Fast Registration Work Requests (FRWR) is not supported\n"); log_rdma_event(ERR, "Device capability flags = %llx max_fast_reg_page_list_len = %u\n", - info->id->device->attrs.device_cap_flags, - info->id->device->attrs.max_fast_reg_page_list_len); + sc->ib.dev->attrs.device_cap_flags, + sc->ib.dev->attrs.max_fast_reg_page_list_len); rc = -EPROTONOSUPPORT; goto out2; } info->max_frmr_depth = min_t(int, smbd_max_frmr_depth, - info->id->device->attrs.max_fast_reg_page_list_len); + sc->ib.dev->attrs.max_fast_reg_page_list_len); info->mr_type = IB_MR_TYPE_MEM_REG; - if (info->id->device->attrs.kernel_cap_flags & IBK_SG_GAPS_REG) + if (sc->ib.dev->attrs.kernel_cap_flags & IBK_SG_GAPS_REG) info->mr_type = IB_MR_TYPE_SG_GAPS; - info->pd = ib_alloc_pd(info->id->device, 0); - if (IS_ERR(info->pd)) { - rc = PTR_ERR(info->pd); + sc->ib.pd = ib_alloc_pd(sc->ib.dev, 0); + if (IS_ERR(sc->ib.pd)) { + rc = PTR_ERR(sc->ib.pd); log_rdma_event(ERR, "ib_alloc_pd() returned %d\n", rc); goto out2; } @@ -668,8 +675,8 @@ static int smbd_ia_open( return 0; out2: - rdma_destroy_id(info->id); - info->id = NULL; + rdma_destroy_id(sc->rdma.cm_id); + sc->rdma.cm_id = NULL; out1: return rc; @@ -683,10 +690,12 @@ out1: */ static int smbd_post_send_negotiate_req(struct smbd_connection *info) { + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; struct ib_send_wr send_wr; int rc = -ENOMEM; struct smbd_request *request; - struct smbd_negotiate_req *packet; + struct smbdirect_negotiate_req *packet; request = mempool_alloc(info->request_mempool, GFP_KERNEL); if (!request) @@ -695,29 +704,29 @@ static int smbd_post_send_negotiate_req(struct smbd_connection *info) request->info = info; packet = smbd_request_payload(request); - packet->min_version = cpu_to_le16(SMBD_V1); - packet->max_version = cpu_to_le16(SMBD_V1); + packet->min_version = cpu_to_le16(SMBDIRECT_V1); + packet->max_version = cpu_to_le16(SMBDIRECT_V1); packet->reserved = 0; - packet->credits_requested = cpu_to_le16(info->send_credit_target); - packet->preferred_send_size = cpu_to_le32(info->max_send_size); - packet->max_receive_size = cpu_to_le32(info->max_receive_size); + packet->credits_requested = cpu_to_le16(sp->send_credit_target); + packet->preferred_send_size = cpu_to_le32(sp->max_send_size); + packet->max_receive_size = cpu_to_le32(sp->max_recv_size); packet->max_fragmented_size = - cpu_to_le32(info->max_fragmented_recv_size); + cpu_to_le32(sp->max_fragmented_recv_size); request->num_sge = 1; request->sge[0].addr = ib_dma_map_single( - info->id->device, (void *)packet, + sc->ib.dev, (void *)packet, sizeof(*packet), DMA_TO_DEVICE); - if (ib_dma_mapping_error(info->id->device, request->sge[0].addr)) { + if (ib_dma_mapping_error(sc->ib.dev, request->sge[0].addr)) { rc = -EIO; goto dma_mapping_failed; } request->sge[0].length = sizeof(*packet); - request->sge[0].lkey = info->pd->local_dma_lkey; + request->sge[0].lkey = sc->ib.pd->local_dma_lkey; ib_dma_sync_single_for_device( - info->id->device, request->sge[0].addr, + sc->ib.dev, request->sge[0].addr, request->sge[0].length, DMA_TO_DEVICE); request->cqe.done = send_done; @@ -734,14 +743,14 @@ static int smbd_post_send_negotiate_req(struct smbd_connection *info) request->sge[0].length, request->sge[0].lkey); atomic_inc(&info->send_pending); - rc = ib_post_send(info->id->qp, &send_wr, NULL); + rc = ib_post_send(sc->ib.qp, &send_wr, NULL); if (!rc) return 0; /* if we reach here, post send failed */ log_rdma_send(ERR, "ib_post_send failed rc=%d\n", rc); atomic_dec(&info->send_pending); - ib_dma_unmap_single(info->id->device, request->sge[0].addr, + ib_dma_unmap_single(sc->ib.dev, request->sge[0].addr, request->sge[0].length, DMA_TO_DEVICE); smbd_disconnect_rdma_connection(info); @@ -774,10 +783,10 @@ static int manage_credits_prior_sending(struct smbd_connection *info) /* * Check if we need to send a KEEP_ALIVE message * The idle connection timer triggers a KEEP_ALIVE message when expires - * SMB_DIRECT_RESPONSE_REQUESTED is set in the message flag to have peer send + * SMBDIRECT_FLAG_RESPONSE_REQUESTED is set in the message flag to have peer send * back a response. * return value: - * 1 if SMB_DIRECT_RESPONSE_REQUESTED needs to be set + * 1 if SMBDIRECT_FLAG_RESPONSE_REQUESTED needs to be set * 0: otherwise */ static int manage_keep_alive_before_sending(struct smbd_connection *info) @@ -793,6 +802,8 @@ static int manage_keep_alive_before_sending(struct smbd_connection *info) static int smbd_post_send(struct smbd_connection *info, struct smbd_request *request) { + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; struct ib_send_wr send_wr; int rc, i; @@ -801,7 +812,7 @@ static int smbd_post_send(struct smbd_connection *info, "rdma_request sge[%d] addr=0x%llx length=%u\n", i, request->sge[i].addr, request->sge[i].length); ib_dma_sync_single_for_device( - info->id->device, + sc->ib.dev, request->sge[i].addr, request->sge[i].length, DMA_TO_DEVICE); @@ -816,7 +827,7 @@ static int smbd_post_send(struct smbd_connection *info, send_wr.opcode = IB_WR_SEND; send_wr.send_flags = IB_SEND_SIGNALED; - rc = ib_post_send(info->id->qp, &send_wr, NULL); + rc = ib_post_send(sc->ib.qp, &send_wr, NULL); if (rc) { log_rdma_send(ERR, "ib_post_send failed rc=%d\n", rc); smbd_disconnect_rdma_connection(info); @@ -824,7 +835,7 @@ static int smbd_post_send(struct smbd_connection *info, } else /* Reset timer for idle connection after packet is sent */ mod_delayed_work(info->workqueue, &info->idle_timer_work, - info->keep_alive_interval*HZ); + msecs_to_jiffies(sp->keepalive_interval_msec)); return rc; } @@ -833,22 +844,24 @@ static int smbd_post_send_iter(struct smbd_connection *info, struct iov_iter *iter, int *_remaining_data_length) { + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; int i, rc; int header_length; int data_length; struct smbd_request *request; - struct smbd_data_transfer *packet; + struct smbdirect_data_transfer *packet; int new_credits = 0; wait_credit: /* Wait for send credits. A SMBD packet needs one credit */ rc = wait_event_interruptible(info->wait_send_queue, atomic_read(&info->send_credits) > 0 || - info->transport_status != SMBD_CONNECTED); + sc->status != SMBDIRECT_SOCKET_CONNECTED); if (rc) goto err_wait_credit; - if (info->transport_status != SMBD_CONNECTED) { + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) { log_outgoing(ERR, "disconnected not sending on wait_credit\n"); rc = -EAGAIN; goto err_wait_credit; @@ -860,17 +873,17 @@ wait_credit: wait_send_queue: wait_event(info->wait_post_send, - atomic_read(&info->send_pending) < info->send_credit_target || - info->transport_status != SMBD_CONNECTED); + atomic_read(&info->send_pending) < sp->send_credit_target || + sc->status != SMBDIRECT_SOCKET_CONNECTED); - if (info->transport_status != SMBD_CONNECTED) { + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) { log_outgoing(ERR, "disconnected not sending on wait_send_queue\n"); rc = -EAGAIN; goto err_wait_send_queue; } if (unlikely(atomic_inc_return(&info->send_pending) > - info->send_credit_target)) { + sp->send_credit_target)) { atomic_dec(&info->send_pending); goto wait_send_queue; } @@ -890,12 +903,14 @@ wait_send_queue: .nr_sge = 1, .max_sge = SMBDIRECT_MAX_SEND_SGE, .sge = request->sge, - .device = info->id->device, - .local_dma_lkey = info->pd->local_dma_lkey, + .device = sc->ib.dev, + .local_dma_lkey = sc->ib.pd->local_dma_lkey, .direction = DMA_TO_DEVICE, }; + size_t payload_len = umin(*_remaining_data_length, + sp->max_send_size - sizeof(*packet)); - rc = smb_extract_iter_to_rdma(iter, *_remaining_data_length, + rc = smb_extract_iter_to_rdma(iter, payload_len, &extract); if (rc < 0) goto err_dma; @@ -909,7 +924,7 @@ wait_send_queue: /* Fill in the packet header */ packet = smbd_request_payload(request); - packet->credits_requested = cpu_to_le16(info->send_credit_target); + packet->credits_requested = cpu_to_le16(sp->send_credit_target); new_credits = manage_credits_prior_sending(info); atomic_add(new_credits, &info->receive_credits); @@ -919,7 +934,7 @@ wait_send_queue: packet->flags = 0; if (manage_keep_alive_before_sending(info)) - packet->flags |= cpu_to_le16(SMB_DIRECT_RESPONSE_REQUESTED); + packet->flags |= cpu_to_le16(SMBDIRECT_FLAG_RESPONSE_REQUESTED); packet->reserved = 0; if (!data_length) @@ -938,23 +953,23 @@ wait_send_queue: le32_to_cpu(packet->remaining_data_length)); /* Map the packet to DMA */ - header_length = sizeof(struct smbd_data_transfer); + header_length = sizeof(struct smbdirect_data_transfer); /* If this is a packet without payload, don't send padding */ if (!data_length) - header_length = offsetof(struct smbd_data_transfer, padding); + header_length = offsetof(struct smbdirect_data_transfer, padding); - request->sge[0].addr = ib_dma_map_single(info->id->device, + request->sge[0].addr = ib_dma_map_single(sc->ib.dev, (void *)packet, header_length, DMA_TO_DEVICE); - if (ib_dma_mapping_error(info->id->device, request->sge[0].addr)) { + if (ib_dma_mapping_error(sc->ib.dev, request->sge[0].addr)) { rc = -EIO; request->sge[0].addr = 0; goto err_dma; } request->sge[0].length = header_length; - request->sge[0].lkey = info->pd->local_dma_lkey; + request->sge[0].lkey = sc->ib.pd->local_dma_lkey; rc = smbd_post_send(info, request); if (!rc) @@ -963,7 +978,7 @@ wait_send_queue: err_dma: for (i = 0; i < request->num_sge; i++) if (request->sge[i].addr) - ib_dma_unmap_single(info->id->device, + ib_dma_unmap_single(sc->ib.dev, request->sge[i].addr, request->sge[i].length, DMA_TO_DEVICE); @@ -1000,6 +1015,27 @@ static int smbd_post_send_empty(struct smbd_connection *info) return smbd_post_send_iter(info, NULL, &remaining_data_length); } +static int smbd_post_send_full_iter(struct smbd_connection *info, + struct iov_iter *iter, + int *_remaining_data_length) +{ + int rc = 0; + + /* + * smbd_post_send_iter() respects the + * negotiated max_send_size, so we need to + * loop until the full iter is posted + */ + + while (iov_iter_count(iter) > 0) { + rc = smbd_post_send_iter(info, iter, _remaining_data_length); + if (rc < 0) + break; + } + + return rc; +} + /* * Post a receive request to the transport * The remote peer can only send data when a receive request is posted @@ -1008,17 +1044,19 @@ static int smbd_post_send_empty(struct smbd_connection *info) static int smbd_post_recv( struct smbd_connection *info, struct smbd_response *response) { + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; struct ib_recv_wr recv_wr; int rc = -EIO; response->sge.addr = ib_dma_map_single( - info->id->device, response->packet, - info->max_receive_size, DMA_FROM_DEVICE); - if (ib_dma_mapping_error(info->id->device, response->sge.addr)) + sc->ib.dev, response->packet, + sp->max_recv_size, DMA_FROM_DEVICE); + if (ib_dma_mapping_error(sc->ib.dev, response->sge.addr)) return rc; - response->sge.length = info->max_receive_size; - response->sge.lkey = info->pd->local_dma_lkey; + response->sge.length = sp->max_recv_size; + response->sge.lkey = sc->ib.pd->local_dma_lkey; response->cqe.done = recv_done; @@ -1027,9 +1065,9 @@ static int smbd_post_recv( recv_wr.sg_list = &response->sge; recv_wr.num_sge = 1; - rc = ib_post_recv(info->id->qp, &recv_wr, NULL); + rc = ib_post_recv(sc->ib.qp, &recv_wr, NULL); if (rc) { - ib_dma_unmap_single(info->id->device, response->sge.addr, + ib_dma_unmap_single(sc->ib.dev, response->sge.addr, response->sge.length, DMA_FROM_DEVICE); smbd_disconnect_rdma_connection(info); log_rdma_recv(ERR, "ib_post_recv failed rc=%d\n", rc); @@ -1187,9 +1225,10 @@ static struct smbd_response *get_receive_buffer(struct smbd_connection *info) static void put_receive_buffer( struct smbd_connection *info, struct smbd_response *response) { + struct smbdirect_socket *sc = &info->socket; unsigned long flags; - ib_dma_unmap_single(info->id->device, response->sge.addr, + ib_dma_unmap_single(sc->ib.dev, response->sge.addr, response->sge.length, DMA_FROM_DEVICE); spin_lock_irqsave(&info->receive_queue_lock, flags); @@ -1264,6 +1303,8 @@ static void idle_connection_timer(struct work_struct *work) struct smbd_connection *info = container_of( work, struct smbd_connection, idle_timer_work.work); + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; if (info->keep_alive_requested != KEEP_ALIVE_NONE) { log_keep_alive(ERR, @@ -1278,7 +1319,7 @@ static void idle_connection_timer(struct work_struct *work) /* Setup the next idle timeout work */ queue_delayed_work(info->workqueue, &info->idle_timer_work, - info->keep_alive_interval*HZ); + msecs_to_jiffies(sp->keepalive_interval_msec)); } /* @@ -1289,6 +1330,8 @@ static void idle_connection_timer(struct work_struct *work) void smbd_destroy(struct TCP_Server_Info *server) { struct smbd_connection *info = server->smbd_conn; + struct smbdirect_socket *sc; + struct smbdirect_socket_parameters *sp; struct smbd_response *response; unsigned long flags; @@ -1296,19 +1339,22 @@ void smbd_destroy(struct TCP_Server_Info *server) log_rdma_event(INFO, "rdma session already destroyed\n"); return; } + sc = &info->socket; + sp = &sc->parameters; log_rdma_event(INFO, "destroying rdma session\n"); - if (info->transport_status != SMBD_DISCONNECTED) { - rdma_disconnect(server->smbd_conn->id); + if (sc->status != SMBDIRECT_SOCKET_DISCONNECTED) { + rdma_disconnect(sc->rdma.cm_id); log_rdma_event(INFO, "wait for transport being disconnected\n"); wait_event_interruptible( info->disconn_wait, - info->transport_status == SMBD_DISCONNECTED); + sc->status == SMBDIRECT_SOCKET_DISCONNECTED); } log_rdma_event(INFO, "destroying qp\n"); - ib_drain_qp(info->id->qp); - rdma_destroy_qp(info->id); + ib_drain_qp(sc->ib.qp); + rdma_destroy_qp(sc->rdma.cm_id); + sc->ib.qp = NULL; log_rdma_event(INFO, "cancelling idle timer\n"); cancel_delayed_work_sync(&info->idle_timer_work); @@ -1336,7 +1382,7 @@ void smbd_destroy(struct TCP_Server_Info *server) log_rdma_event(INFO, "free receive buffers\n"); wait_event(info->wait_receive_queues, info->count_receive_queue + info->count_empty_packet_queue - == info->receive_credit_max); + == sp->recv_credit_max); destroy_receive_buffers(info); /* @@ -1355,10 +1401,10 @@ void smbd_destroy(struct TCP_Server_Info *server) } destroy_mr_list(info); - ib_free_cq(info->send_cq); - ib_free_cq(info->recv_cq); - ib_dealloc_pd(info->pd); - rdma_destroy_id(info->id); + ib_free_cq(sc->ib.send_cq); + ib_free_cq(sc->ib.recv_cq); + ib_dealloc_pd(sc->ib.pd); + rdma_destroy_id(sc->rdma.cm_id); /* free mempools */ mempool_destroy(info->request_mempool); @@ -1367,7 +1413,7 @@ void smbd_destroy(struct TCP_Server_Info *server) mempool_destroy(info->response_mempool); kmem_cache_destroy(info->response_cache); - info->transport_status = SMBD_DESTROYED; + sc->status = SMBDIRECT_SOCKET_DESTROYED; destroy_workqueue(info->workqueue); log_rdma_event(INFO, "rdma session destroyed\n"); @@ -1392,7 +1438,7 @@ int smbd_reconnect(struct TCP_Server_Info *server) * This is possible if transport is disconnected and we haven't received * notification from RDMA, but upper layer has detected timeout */ - if (server->smbd_conn->transport_status == SMBD_CONNECTED) { + if (server->smbd_conn->socket.status == SMBDIRECT_SOCKET_CONNECTED) { log_rdma_event(INFO, "disconnecting transport\n"); smbd_destroy(server); } @@ -1424,37 +1470,47 @@ static void destroy_caches_and_workqueue(struct smbd_connection *info) #define MAX_NAME_LEN 80 static int allocate_caches_and_workqueue(struct smbd_connection *info) { + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; char name[MAX_NAME_LEN]; int rc; + if (WARN_ON_ONCE(sp->max_recv_size < sizeof(struct smbdirect_data_transfer))) + return -ENOMEM; + scnprintf(name, MAX_NAME_LEN, "smbd_request_%p", info); info->request_cache = kmem_cache_create( name, sizeof(struct smbd_request) + - sizeof(struct smbd_data_transfer), + sizeof(struct smbdirect_data_transfer), 0, SLAB_HWCACHE_ALIGN, NULL); if (!info->request_cache) return -ENOMEM; info->request_mempool = - mempool_create(info->send_credit_target, mempool_alloc_slab, + mempool_create(sp->send_credit_target, mempool_alloc_slab, mempool_free_slab, info->request_cache); if (!info->request_mempool) goto out1; scnprintf(name, MAX_NAME_LEN, "smbd_response_%p", info); + + struct kmem_cache_args response_args = { + .align = __alignof__(struct smbd_response), + .useroffset = (offsetof(struct smbd_response, packet) + + sizeof(struct smbdirect_data_transfer)), + .usersize = sp->max_recv_size - sizeof(struct smbdirect_data_transfer), + }; info->response_cache = - kmem_cache_create( - name, - sizeof(struct smbd_response) + - info->max_receive_size, - 0, SLAB_HWCACHE_ALIGN, NULL); + kmem_cache_create(name, + sizeof(struct smbd_response) + sp->max_recv_size, + &response_args, SLAB_HWCACHE_ALIGN); if (!info->response_cache) goto out2; info->response_mempool = - mempool_create(info->receive_credit_max, mempool_alloc_slab, + mempool_create(sp->recv_credit_max, mempool_alloc_slab, mempool_free_slab, info->response_cache); if (!info->response_mempool) goto out3; @@ -1464,7 +1520,7 @@ static int allocate_caches_and_workqueue(struct smbd_connection *info) if (!info->workqueue) goto out4; - rc = allocate_receive_buffers(info, info->receive_credit_max); + rc = allocate_receive_buffers(info, sp->recv_credit_max); if (rc) { log_rdma_event(ERR, "failed to allocate receive buffers\n"); goto out5; @@ -1491,6 +1547,8 @@ static struct smbd_connection *_smbd_get_connection( { int rc; struct smbd_connection *info; + struct smbdirect_socket *sc; + struct smbdirect_socket_parameters *sp; struct rdma_conn_param conn_param; struct ib_qp_init_attr qp_attr; struct sockaddr_in *addr_in = (struct sockaddr_in *) dstaddr; @@ -1500,101 +1558,102 @@ static struct smbd_connection *_smbd_get_connection( info = kzalloc(sizeof(struct smbd_connection), GFP_KERNEL); if (!info) return NULL; + sc = &info->socket; + sp = &sc->parameters; - info->transport_status = SMBD_CONNECTING; + sc->status = SMBDIRECT_SOCKET_CONNECTING; rc = smbd_ia_open(info, dstaddr, port); if (rc) { log_rdma_event(INFO, "smbd_ia_open rc=%d\n", rc); goto create_id_failed; } - if (smbd_send_credit_target > info->id->device->attrs.max_cqe || - smbd_send_credit_target > info->id->device->attrs.max_qp_wr) { + if (smbd_send_credit_target > sc->ib.dev->attrs.max_cqe || + smbd_send_credit_target > sc->ib.dev->attrs.max_qp_wr) { log_rdma_event(ERR, "consider lowering send_credit_target = %d. Possible CQE overrun, device reporting max_cqe %d max_qp_wr %d\n", smbd_send_credit_target, - info->id->device->attrs.max_cqe, - info->id->device->attrs.max_qp_wr); + sc->ib.dev->attrs.max_cqe, + sc->ib.dev->attrs.max_qp_wr); goto config_failed; } - if (smbd_receive_credit_max > info->id->device->attrs.max_cqe || - smbd_receive_credit_max > info->id->device->attrs.max_qp_wr) { + if (smbd_receive_credit_max > sc->ib.dev->attrs.max_cqe || + smbd_receive_credit_max > sc->ib.dev->attrs.max_qp_wr) { log_rdma_event(ERR, "consider lowering receive_credit_max = %d. Possible CQE overrun, device reporting max_cqe %d max_qp_wr %d\n", smbd_receive_credit_max, - info->id->device->attrs.max_cqe, - info->id->device->attrs.max_qp_wr); + sc->ib.dev->attrs.max_cqe, + sc->ib.dev->attrs.max_qp_wr); goto config_failed; } - info->receive_credit_max = smbd_receive_credit_max; - info->send_credit_target = smbd_send_credit_target; - info->max_send_size = smbd_max_send_size; - info->max_fragmented_recv_size = smbd_max_fragmented_recv_size; - info->max_receive_size = smbd_max_receive_size; - info->keep_alive_interval = smbd_keep_alive_interval; + sp->recv_credit_max = smbd_receive_credit_max; + sp->send_credit_target = smbd_send_credit_target; + sp->max_send_size = smbd_max_send_size; + sp->max_fragmented_recv_size = smbd_max_fragmented_recv_size; + sp->max_recv_size = smbd_max_receive_size; + sp->keepalive_interval_msec = smbd_keep_alive_interval * 1000; - if (info->id->device->attrs.max_send_sge < SMBDIRECT_MAX_SEND_SGE || - info->id->device->attrs.max_recv_sge < SMBDIRECT_MAX_RECV_SGE) { + if (sc->ib.dev->attrs.max_send_sge < SMBDIRECT_MAX_SEND_SGE || + sc->ib.dev->attrs.max_recv_sge < SMBDIRECT_MAX_RECV_SGE) { log_rdma_event(ERR, "device %.*s max_send_sge/max_recv_sge = %d/%d too small\n", IB_DEVICE_NAME_MAX, - info->id->device->name, - info->id->device->attrs.max_send_sge, - info->id->device->attrs.max_recv_sge); + sc->ib.dev->name, + sc->ib.dev->attrs.max_send_sge, + sc->ib.dev->attrs.max_recv_sge); goto config_failed; } - info->send_cq = NULL; - info->recv_cq = NULL; - info->send_cq = - ib_alloc_cq_any(info->id->device, info, - info->send_credit_target, IB_POLL_SOFTIRQ); - if (IS_ERR(info->send_cq)) { - info->send_cq = NULL; + sc->ib.send_cq = + ib_alloc_cq_any(sc->ib.dev, info, + sp->send_credit_target, IB_POLL_SOFTIRQ); + if (IS_ERR(sc->ib.send_cq)) { + sc->ib.send_cq = NULL; goto alloc_cq_failed; } - info->recv_cq = - ib_alloc_cq_any(info->id->device, info, - info->receive_credit_max, IB_POLL_SOFTIRQ); - if (IS_ERR(info->recv_cq)) { - info->recv_cq = NULL; + sc->ib.recv_cq = + ib_alloc_cq_any(sc->ib.dev, info, + sp->recv_credit_max, IB_POLL_SOFTIRQ); + if (IS_ERR(sc->ib.recv_cq)) { + sc->ib.recv_cq = NULL; goto alloc_cq_failed; } memset(&qp_attr, 0, sizeof(qp_attr)); qp_attr.event_handler = smbd_qp_async_error_upcall; qp_attr.qp_context = info; - qp_attr.cap.max_send_wr = info->send_credit_target; - qp_attr.cap.max_recv_wr = info->receive_credit_max; + qp_attr.cap.max_send_wr = sp->send_credit_target; + qp_attr.cap.max_recv_wr = sp->recv_credit_max; qp_attr.cap.max_send_sge = SMBDIRECT_MAX_SEND_SGE; qp_attr.cap.max_recv_sge = SMBDIRECT_MAX_RECV_SGE; qp_attr.cap.max_inline_data = 0; qp_attr.sq_sig_type = IB_SIGNAL_REQ_WR; qp_attr.qp_type = IB_QPT_RC; - qp_attr.send_cq = info->send_cq; - qp_attr.recv_cq = info->recv_cq; + qp_attr.send_cq = sc->ib.send_cq; + qp_attr.recv_cq = sc->ib.recv_cq; qp_attr.port_num = ~0; - rc = rdma_create_qp(info->id, info->pd, &qp_attr); + rc = rdma_create_qp(sc->rdma.cm_id, sc->ib.pd, &qp_attr); if (rc) { log_rdma_event(ERR, "rdma_create_qp failed %i\n", rc); goto create_qp_failed; } + sc->ib.qp = sc->rdma.cm_id->qp; memset(&conn_param, 0, sizeof(conn_param)); conn_param.initiator_depth = 0; conn_param.responder_resources = - min(info->id->device->attrs.max_qp_rd_atom, + min(sc->ib.dev->attrs.max_qp_rd_atom, SMBD_CM_RESPONDER_RESOURCES); info->responder_resources = conn_param.responder_resources; log_rdma_mr(INFO, "responder_resources=%d\n", info->responder_resources); /* Need to send IRD/ORD in private data for iWARP */ - info->id->device->ops.get_port_immutable( - info->id->device, info->id->port_num, &port_immutable); + sc->ib.dev->ops.get_port_immutable( + sc->ib.dev, sc->rdma.cm_id->port_num, &port_immutable); if (port_immutable.core_cap_flags & RDMA_CORE_PORT_IWARP) { ird_ord_hdr[0] = info->responder_resources; ird_ord_hdr[1] = 1; @@ -1615,16 +1674,16 @@ static struct smbd_connection *_smbd_get_connection( init_waitqueue_head(&info->conn_wait); init_waitqueue_head(&info->disconn_wait); init_waitqueue_head(&info->wait_reassembly_queue); - rc = rdma_connect(info->id, &conn_param); + rc = rdma_connect(sc->rdma.cm_id, &conn_param); if (rc) { log_rdma_event(ERR, "rdma_connect() failed with %i\n", rc); goto rdma_connect_failed; } wait_event_interruptible( - info->conn_wait, info->transport_status != SMBD_CONNECTING); + info->conn_wait, sc->status != SMBDIRECT_SOCKET_CONNECTING); - if (info->transport_status != SMBD_CONNECTED) { + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) { log_rdma_event(ERR, "rdma_connect failed port=%d\n", port); goto rdma_connect_failed; } @@ -1640,7 +1699,7 @@ static struct smbd_connection *_smbd_get_connection( init_waitqueue_head(&info->wait_send_queue); INIT_DELAYED_WORK(&info->idle_timer_work, idle_connection_timer); queue_delayed_work(info->workqueue, &info->idle_timer_work, - info->keep_alive_interval*HZ); + msecs_to_jiffies(sp->keepalive_interval_msec)); init_waitqueue_head(&info->wait_send_pending); atomic_set(&info->send_pending, 0); @@ -1675,26 +1734,26 @@ allocate_mr_failed: negotiation_failed: cancel_delayed_work_sync(&info->idle_timer_work); destroy_caches_and_workqueue(info); - info->transport_status = SMBD_NEGOTIATE_FAILED; + sc->status = SMBDIRECT_SOCKET_NEGOTIATE_FAILED; init_waitqueue_head(&info->conn_wait); - rdma_disconnect(info->id); + rdma_disconnect(sc->rdma.cm_id); wait_event(info->conn_wait, - info->transport_status == SMBD_DISCONNECTED); + sc->status == SMBDIRECT_SOCKET_DISCONNECTED); allocate_cache_failed: rdma_connect_failed: - rdma_destroy_qp(info->id); + rdma_destroy_qp(sc->rdma.cm_id); create_qp_failed: alloc_cq_failed: - if (info->send_cq) - ib_free_cq(info->send_cq); - if (info->recv_cq) - ib_free_cq(info->recv_cq); + if (sc->ib.send_cq) + ib_free_cq(sc->ib.send_cq); + if (sc->ib.recv_cq) + ib_free_cq(sc->ib.recv_cq); config_failed: - ib_dealloc_pd(info->pd); - rdma_destroy_id(info->id); + ib_dealloc_pd(sc->ib.pd); + rdma_destroy_id(sc->rdma.cm_id); create_id_failed: kfree(info); @@ -1719,34 +1778,39 @@ try_again: } /* - * Receive data from receive reassembly queue + * Receive data from the transport's receive reassembly queue * All the incoming data packets are placed in reassembly queue - * buf: the buffer to read data into + * iter: the buffer to read data into * size: the length of data to read * return value: actual data read - * Note: this implementation copies the data from reassebmly queue to receive + * + * Note: this implementation copies the data from reassembly queue to receive * buffers used by upper layer. This is not the optimal code path. A better way * to do it is to not have upper layer allocate its receive buffers but rather * borrow the buffer from reassembly queue, and return it after data is * consumed. But this will require more changes to upper layer code, and also * need to consider packet boundaries while they still being reassembled. */ -static int smbd_recv_buf(struct smbd_connection *info, char *buf, - unsigned int size) +int smbd_recv(struct smbd_connection *info, struct msghdr *msg) { + struct smbdirect_socket *sc = &info->socket; struct smbd_response *response; - struct smbd_data_transfer *data_transfer; + struct smbdirect_data_transfer *data_transfer; + size_t size = iov_iter_count(&msg->msg_iter); int to_copy, to_read, data_read, offset; u32 data_length, remaining_data_length, data_offset; int rc; + if (WARN_ON_ONCE(iov_iter_rw(&msg->msg_iter) == WRITE)) + return -EINVAL; /* It's a bug in upper layer to get there */ + again: /* * No need to hold the reassembly queue lock all the time as we are * the only one reading from the front of the queue. The transport * may add more entries to the back of the queue at the same time */ - log_read(INFO, "size=%d info->reassembly_data_length=%d\n", size, + log_read(INFO, "size=%zd info->reassembly_data_length=%d\n", size, info->reassembly_data_length); if (info->reassembly_data_length >= size) { int queue_length; @@ -1784,7 +1848,10 @@ again: if (response->first_segment && size == 4) { unsigned int rfc1002_len = data_length + remaining_data_length; - *((__be32 *)buf) = cpu_to_be32(rfc1002_len); + __be32 rfc1002_hdr = cpu_to_be32(rfc1002_len); + if (copy_to_iter(&rfc1002_hdr, sizeof(rfc1002_hdr), + &msg->msg_iter) != sizeof(rfc1002_hdr)) + return -EFAULT; data_read = 4; response->first_segment = false; log_read(INFO, "returning rfc1002 length %d\n", @@ -1793,10 +1860,9 @@ again: } to_copy = min_t(int, data_length - offset, to_read); - memcpy( - buf + data_read, - (char *)data_transfer + data_offset + offset, - to_copy); + if (copy_to_iter((char *)data_transfer + data_offset + offset, + to_copy, &msg->msg_iter) != to_copy) + return -EFAULT; /* move on to the next buffer? */ if (to_copy == data_length - offset) { @@ -1848,12 +1914,12 @@ read_rfc1002_done: rc = wait_event_interruptible( info->wait_reassembly_queue, info->reassembly_data_length >= size || - info->transport_status != SMBD_CONNECTED); + sc->status != SMBDIRECT_SOCKET_CONNECTED); /* Don't return any data if interrupted */ if (rc) return rc; - if (info->transport_status != SMBD_CONNECTED) { + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) { log_read(ERR, "disconnected\n"); return -ECONNABORTED; } @@ -1862,89 +1928,6 @@ read_rfc1002_done: } /* - * Receive a page from receive reassembly queue - * page: the page to read data into - * to_read: the length of data to read - * return value: actual data read - */ -static int smbd_recv_page(struct smbd_connection *info, - struct page *page, unsigned int page_offset, - unsigned int to_read) -{ - int ret; - char *to_address; - void *page_address; - - /* make sure we have the page ready for read */ - ret = wait_event_interruptible( - info->wait_reassembly_queue, - info->reassembly_data_length >= to_read || - info->transport_status != SMBD_CONNECTED); - if (ret) - return ret; - - /* now we can read from reassembly queue and not sleep */ - page_address = kmap_atomic(page); - to_address = (char *) page_address + page_offset; - - log_read(INFO, "reading from page=%p address=%p to_read=%d\n", - page, to_address, to_read); - - ret = smbd_recv_buf(info, to_address, to_read); - kunmap_atomic(page_address); - - return ret; -} - -/* - * Receive data from transport - * msg: a msghdr point to the buffer, can be ITER_KVEC or ITER_BVEC - * return: total bytes read, or 0. SMB Direct will not do partial read. - */ -int smbd_recv(struct smbd_connection *info, struct msghdr *msg) -{ - char *buf; - struct page *page; - unsigned int to_read, page_offset; - int rc; - - if (iov_iter_rw(&msg->msg_iter) == WRITE) { - /* It's a bug in upper layer to get there */ - cifs_dbg(VFS, "Invalid msg iter dir %u\n", - iov_iter_rw(&msg->msg_iter)); - rc = -EINVAL; - goto out; - } - - switch (iov_iter_type(&msg->msg_iter)) { - case ITER_KVEC: - buf = msg->msg_iter.kvec->iov_base; - to_read = msg->msg_iter.kvec->iov_len; - rc = smbd_recv_buf(info, buf, to_read); - break; - - case ITER_BVEC: - page = msg->msg_iter.bvec->bv_page; - page_offset = msg->msg_iter.bvec->bv_offset; - to_read = msg->msg_iter.bvec->bv_len; - rc = smbd_recv_page(info, page, page_offset, to_read); - break; - - default: - /* It's a bug in upper layer to get there */ - cifs_dbg(VFS, "Invalid msg type %d\n", - iov_iter_type(&msg->msg_iter)); - rc = -EINVAL; - } - -out: - /* SMBDirect will read it all or nothing */ - if (rc > 0) - msg->msg_iter.count = 0; - return rc; -} - -/* * Send data to transport * Each rqst is transported as a SMBDirect payload * rqst: the data to write @@ -1954,12 +1937,14 @@ int smbd_send(struct TCP_Server_Info *server, int num_rqst, struct smb_rqst *rqst_array) { struct smbd_connection *info = server->smbd_conn; + struct smbdirect_socket *sc = &info->socket; + struct smbdirect_socket_parameters *sp = &sc->parameters; struct smb_rqst *rqst; struct iov_iter iter; unsigned int remaining_data_length, klen; int rc, i, rqst_idx; - if (info->transport_status != SMBD_CONNECTED) + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) return -EAGAIN; /* @@ -1971,10 +1956,10 @@ int smbd_send(struct TCP_Server_Info *server, for (i = 0; i < num_rqst; i++) remaining_data_length += smb_rqst_len(server, &rqst_array[i]); - if (unlikely(remaining_data_length > info->max_fragmented_send_size)) { + if (unlikely(remaining_data_length > sp->max_fragmented_send_size)) { /* assertion: payload never exceeds negotiated maximum */ log_write(ERR, "payload size %d > max size %d\n", - remaining_data_length, info->max_fragmented_send_size); + remaining_data_length, sp->max_fragmented_send_size); return -EINVAL; } @@ -2000,14 +1985,14 @@ int smbd_send(struct TCP_Server_Info *server, klen += rqst->rq_iov[i].iov_len; iov_iter_kvec(&iter, ITER_SOURCE, rqst->rq_iov, rqst->rq_nvec, klen); - rc = smbd_post_send_iter(info, &iter, &remaining_data_length); + rc = smbd_post_send_full_iter(info, &iter, &remaining_data_length); if (rc < 0) break; if (iov_iter_count(&rqst->rq_iter) > 0) { /* And then the data pages if there are any */ - rc = smbd_post_send_iter(info, &rqst->rq_iter, - &remaining_data_length); + rc = smbd_post_send_full_iter(info, &rqst->rq_iter, + &remaining_data_length); if (rc < 0) break; } @@ -2053,6 +2038,7 @@ static void smbd_mr_recovery_work(struct work_struct *work) { struct smbd_connection *info = container_of(work, struct smbd_connection, mr_recovery_work); + struct smbdirect_socket *sc = &info->socket; struct smbd_mr *smbdirect_mr; int rc; @@ -2070,7 +2056,7 @@ static void smbd_mr_recovery_work(struct work_struct *work) } smbdirect_mr->mr = ib_alloc_mr( - info->pd, info->mr_type, + sc->ib.pd, info->mr_type, info->max_frmr_depth); if (IS_ERR(smbdirect_mr->mr)) { log_rdma_mr(ERR, "ib_alloc_mr failed mr_type=%x max_frmr_depth=%x\n", @@ -2099,12 +2085,13 @@ static void smbd_mr_recovery_work(struct work_struct *work) static void destroy_mr_list(struct smbd_connection *info) { + struct smbdirect_socket *sc = &info->socket; struct smbd_mr *mr, *tmp; cancel_work_sync(&info->mr_recovery_work); list_for_each_entry_safe(mr, tmp, &info->mr_list, list) { if (mr->state == MR_INVALIDATED) - ib_dma_unmap_sg(info->id->device, mr->sgt.sgl, + ib_dma_unmap_sg(sc->ib.dev, mr->sgt.sgl, mr->sgt.nents, mr->dir); ib_dereg_mr(mr->mr); kfree(mr->sgt.sgl); @@ -2121,6 +2108,7 @@ static void destroy_mr_list(struct smbd_connection *info) */ static int allocate_mr_list(struct smbd_connection *info) { + struct smbdirect_socket *sc = &info->socket; int i; struct smbd_mr *smbdirect_mr, *tmp; @@ -2136,7 +2124,7 @@ static int allocate_mr_list(struct smbd_connection *info) smbdirect_mr = kzalloc(sizeof(*smbdirect_mr), GFP_KERNEL); if (!smbdirect_mr) goto cleanup_entries; - smbdirect_mr->mr = ib_alloc_mr(info->pd, info->mr_type, + smbdirect_mr->mr = ib_alloc_mr(sc->ib.pd, info->mr_type, info->max_frmr_depth); if (IS_ERR(smbdirect_mr->mr)) { log_rdma_mr(ERR, "ib_alloc_mr failed mr_type=%x max_frmr_depth=%x\n", @@ -2181,20 +2169,20 @@ cleanup_entries: */ static struct smbd_mr *get_mr(struct smbd_connection *info) { + struct smbdirect_socket *sc = &info->socket; struct smbd_mr *ret; int rc; again: rc = wait_event_interruptible(info->wait_mr, atomic_read(&info->mr_ready_count) || - info->transport_status != SMBD_CONNECTED); + sc->status != SMBDIRECT_SOCKET_CONNECTED); if (rc) { log_rdma_mr(ERR, "wait_event_interruptible rc=%x\n", rc); return NULL; } - if (info->transport_status != SMBD_CONNECTED) { - log_rdma_mr(ERR, "info->transport_status=%x\n", - info->transport_status); + if (sc->status != SMBDIRECT_SOCKET_CONNECTED) { + log_rdma_mr(ERR, "sc->status=%x\n", sc->status); return NULL; } @@ -2247,6 +2235,7 @@ struct smbd_mr *smbd_register_mr(struct smbd_connection *info, struct iov_iter *iter, bool writing, bool need_invalidate) { + struct smbdirect_socket *sc = &info->socket; struct smbd_mr *smbdirect_mr; int rc, num_pages; enum dma_data_direction dir; @@ -2276,7 +2265,7 @@ struct smbd_mr *smbd_register_mr(struct smbd_connection *info, num_pages, iov_iter_count(iter), info->max_frmr_depth); smbd_iter_to_mr(info, iter, &smbdirect_mr->sgt, info->max_frmr_depth); - rc = ib_dma_map_sg(info->id->device, smbdirect_mr->sgt.sgl, + rc = ib_dma_map_sg(sc->ib.dev, smbdirect_mr->sgt.sgl, smbdirect_mr->sgt.nents, dir); if (!rc) { log_rdma_mr(ERR, "ib_dma_map_sg num_pages=%x dir=%x rc=%x\n", @@ -2312,7 +2301,7 @@ struct smbd_mr *smbd_register_mr(struct smbd_connection *info, * on IB_WR_REG_MR. Hardware enforces a barrier and order of execution * on the next ib_post_send when we actually send I/O to remote peer */ - rc = ib_post_send(info->id->qp, ®_wr->wr, NULL); + rc = ib_post_send(sc->ib.qp, ®_wr->wr, NULL); if (!rc) return smbdirect_mr; @@ -2321,7 +2310,7 @@ struct smbd_mr *smbd_register_mr(struct smbd_connection *info, /* If all failed, attempt to recover this MR by setting it MR_ERROR*/ map_mr_error: - ib_dma_unmap_sg(info->id->device, smbdirect_mr->sgt.sgl, + ib_dma_unmap_sg(sc->ib.dev, smbdirect_mr->sgt.sgl, smbdirect_mr->sgt.nents, smbdirect_mr->dir); dma_map_error: @@ -2359,6 +2348,7 @@ int smbd_deregister_mr(struct smbd_mr *smbdirect_mr) { struct ib_send_wr *wr; struct smbd_connection *info = smbdirect_mr->conn; + struct smbdirect_socket *sc = &info->socket; int rc = 0; if (smbdirect_mr->need_invalidate) { @@ -2372,7 +2362,7 @@ int smbd_deregister_mr(struct smbd_mr *smbdirect_mr) wr->send_flags = IB_SEND_SIGNALED; init_completion(&smbdirect_mr->invalidate_done); - rc = ib_post_send(info->id->qp, wr, NULL); + rc = ib_post_send(sc->ib.qp, wr, NULL); if (rc) { log_rdma_mr(ERR, "ib_post_send failed rc=%x\n", rc); smbd_disconnect_rdma_connection(info); @@ -2389,7 +2379,7 @@ int smbd_deregister_mr(struct smbd_mr *smbdirect_mr) if (smbdirect_mr->state == MR_INVALIDATED) { ib_dma_unmap_sg( - info->id->device, smbdirect_mr->sgt.sgl, + sc->ib.dev, smbdirect_mr->sgt.sgl, smbdirect_mr->sgt.nents, smbdirect_mr->dir); smbdirect_mr->state = MR_READY; @@ -2552,13 +2542,14 @@ static ssize_t smb_extract_folioq_to_rdma(struct iov_iter *iter, size_t fsize = folioq_folio_size(folioq, slot); if (offset < fsize) { - size_t part = umin(maxsize - ret, fsize - offset); + size_t part = umin(maxsize, fsize - offset); if (!smb_set_sge(rdma, folio_page(folio, 0), offset, part)) return -EIO; offset += part; ret += part; + maxsize -= part; } if (offset >= fsize) { @@ -2573,7 +2564,7 @@ static ssize_t smb_extract_folioq_to_rdma(struct iov_iter *iter, slot = 0; } } - } while (rdma->nr_sge < rdma->max_sge || maxsize > 0); + } while (rdma->nr_sge < rdma->max_sge && maxsize > 0); iter->folioq = folioq; iter->folioq_slot = slot; diff --git a/fs/smb/client/smbdirect.h b/fs/smb/client/smbdirect.h index c08e3665150d..75b3f491c3ad 100644 --- a/fs/smb/client/smbdirect.h +++ b/fs/smb/client/smbdirect.h @@ -15,6 +15,9 @@ #include <rdma/rdma_cm.h> #include <linux/mempool.h> +#include "../common/smbdirect/smbdirect.h" +#include "../common/smbdirect/smbdirect_socket.h" + extern int rdma_readwrite_threshold; extern int smbd_max_frmr_depth; extern int smbd_keep_alive_interval; @@ -50,14 +53,8 @@ enum smbd_connection_status { * 5. mempools for allocating packets */ struct smbd_connection { - enum smbd_connection_status transport_status; - - /* RDMA related */ - struct rdma_cm_id *id; - struct ib_qp_init_attr qp_attr; - struct ib_pd *pd; - struct ib_cq *send_cq, *recv_cq; - struct ib_device_attr dev_attr; + struct smbdirect_socket socket; + int ri_rc; struct completion ri_done; wait_queue_head_t conn_wait; @@ -72,15 +69,7 @@ struct smbd_connection { spinlock_t lock_new_credits_offered; int new_credits_offered; - /* Connection parameters defined in [MS-SMBD] 3.1.1.1 */ - int receive_credit_max; - int send_credit_target; - int max_send_size; - int max_fragmented_recv_size; - int max_fragmented_send_size; - int max_receive_size; - int keep_alive_interval; - int max_readwrite_size; + /* dynamic connection parameters defined in [MS-SMBD] 3.1.1.1 */ enum keep_alive_status keep_alive_requested; int protocol; atomic_t send_credits; @@ -177,54 +166,6 @@ enum smbd_message_type { SMBD_TRANSFER_DATA, }; -#define SMB_DIRECT_RESPONSE_REQUESTED 0x0001 - -/* SMBD negotiation request packet [MS-SMBD] 2.2.1 */ -struct smbd_negotiate_req { - __le16 min_version; - __le16 max_version; - __le16 reserved; - __le16 credits_requested; - __le32 preferred_send_size; - __le32 max_receive_size; - __le32 max_fragmented_size; -} __packed; - -/* SMBD negotiation response packet [MS-SMBD] 2.2.2 */ -struct smbd_negotiate_resp { - __le16 min_version; - __le16 max_version; - __le16 negotiated_version; - __le16 reserved; - __le16 credits_requested; - __le16 credits_granted; - __le32 status; - __le32 max_readwrite_size; - __le32 preferred_send_size; - __le32 max_receive_size; - __le32 max_fragmented_size; -} __packed; - -/* SMBD data transfer packet with payload [MS-SMBD] 2.2.3 */ -struct smbd_data_transfer { - __le16 credits_requested; - __le16 credits_granted; - __le16 flags; - __le16 reserved; - __le32 remaining_data_length; - __le32 data_offset; - __le32 data_length; - __le32 padding; - __u8 buffer[]; -} __packed; - -/* The packet fields for a registered RDMA buffer */ -struct smbd_buffer_descriptor_v1 { - __le64 offset; - __le32 token; - __le32 length; -} __packed; - /* Maximum number of SGEs used by smbdirect.c in any send work request */ #define SMBDIRECT_MAX_SEND_SGE 6 diff --git a/fs/smb/client/trace.h b/fs/smb/client/trace.h index 52bcb55d9952..93e5b2bb9f28 100644 --- a/fs/smb/client/trace.h +++ b/fs/smb/client/trace.h @@ -140,7 +140,7 @@ DECLARE_EVENT_CLASS(smb3_rw_err_class, __entry->len = len; __entry->rc = rc; ), - TP_printk("\tR=%08x[%x] xid=%u sid=0x%llx tid=0x%x fid=0x%llx offset=0x%llx len=0x%x rc=%d", + TP_printk("R=%08x[%x] xid=%u sid=0x%llx tid=0x%x fid=0x%llx offset=0x%llx len=0x%x rc=%d", __entry->rreq_debug_id, __entry->rreq_debug_index, __entry->xid, __entry->sesid, __entry->tid, __entry->fid, __entry->offset, __entry->len, __entry->rc) @@ -190,7 +190,7 @@ DECLARE_EVENT_CLASS(smb3_other_err_class, __entry->len = len; __entry->rc = rc; ), - TP_printk("\txid=%u sid=0x%llx tid=0x%x fid=0x%llx offset=0x%llx len=0x%x rc=%d", + TP_printk("xid=%u sid=0x%llx tid=0x%x fid=0x%llx offset=0x%llx len=0x%x rc=%d", __entry->xid, __entry->sesid, __entry->tid, __entry->fid, __entry->offset, __entry->len, __entry->rc) ) @@ -247,7 +247,7 @@ DECLARE_EVENT_CLASS(smb3_copy_range_err_class, __entry->len = len; __entry->rc = rc; ), - TP_printk("\txid=%u sid=0x%llx tid=0x%x source fid=0x%llx source offset=0x%llx target fid=0x%llx target offset=0x%llx len=0x%x rc=%d", + TP_printk("xid=%u sid=0x%llx tid=0x%x source fid=0x%llx source offset=0x%llx target fid=0x%llx target offset=0x%llx len=0x%x rc=%d", __entry->xid, __entry->sesid, __entry->tid, __entry->target_fid, __entry->src_offset, __entry->target_fid, __entry->target_offset, __entry->len, __entry->rc) ) @@ -298,7 +298,7 @@ DECLARE_EVENT_CLASS(smb3_copy_range_done_class, __entry->target_offset = target_offset; __entry->len = len; ), - TP_printk("\txid=%u sid=0x%llx tid=0x%x source fid=0x%llx source offset=0x%llx target fid=0x%llx target offset=0x%llx len=0x%x", + TP_printk("xid=%u sid=0x%llx tid=0x%x source fid=0x%llx source offset=0x%llx target fid=0x%llx target offset=0x%llx len=0x%x", __entry->xid, __entry->sesid, __entry->tid, __entry->target_fid, __entry->src_offset, __entry->target_fid, __entry->target_offset, __entry->len) ) @@ -482,7 +482,7 @@ DECLARE_EVENT_CLASS(smb3_fd_class, __entry->tid = tid; __entry->sesid = sesid; ), - TP_printk("\txid=%u sid=0x%llx tid=0x%x fid=0x%llx", + TP_printk("xid=%u sid=0x%llx tid=0x%x fid=0x%llx", __entry->xid, __entry->sesid, __entry->tid, __entry->fid) ) @@ -521,7 +521,7 @@ DECLARE_EVENT_CLASS(smb3_fd_err_class, __entry->sesid = sesid; __entry->rc = rc; ), - TP_printk("\txid=%u sid=0x%llx tid=0x%x fid=0x%llx rc=%d", + TP_printk("xid=%u sid=0x%llx tid=0x%x fid=0x%llx rc=%d", __entry->xid, __entry->sesid, __entry->tid, __entry->fid, __entry->rc) ) @@ -794,7 +794,7 @@ DECLARE_EVENT_CLASS(smb3_cmd_err_class, __entry->status = status; __entry->rc = rc; ), - TP_printk("\tsid=0x%llx tid=0x%x cmd=%u mid=%llu status=0x%x rc=%d", + TP_printk("sid=0x%llx tid=0x%x cmd=%u mid=%llu status=0x%x rc=%d", __entry->sesid, __entry->tid, __entry->cmd, __entry->mid, __entry->status, __entry->rc) ) @@ -829,7 +829,7 @@ DECLARE_EVENT_CLASS(smb3_cmd_done_class, __entry->cmd = cmd; __entry->mid = mid; ), - TP_printk("\tsid=0x%llx tid=0x%x cmd=%u mid=%llu", + TP_printk("sid=0x%llx tid=0x%x cmd=%u mid=%llu", __entry->sesid, __entry->tid, __entry->cmd, __entry->mid) ) @@ -867,7 +867,7 @@ DECLARE_EVENT_CLASS(smb3_mid_class, __entry->when_sent = when_sent; __entry->when_received = when_received; ), - TP_printk("\tcmd=%u mid=%llu pid=%u, when_sent=%lu when_rcv=%lu", + TP_printk("cmd=%u mid=%llu pid=%u, when_sent=%lu when_rcv=%lu", __entry->cmd, __entry->mid, __entry->pid, __entry->when_sent, __entry->when_received) ) @@ -898,7 +898,7 @@ DECLARE_EVENT_CLASS(smb3_exit_err_class, __assign_str(func_name); __entry->rc = rc; ), - TP_printk("\t%s: xid=%u rc=%d", + TP_printk("%s: xid=%u rc=%d", __get_str(func_name), __entry->xid, __entry->rc) ) @@ -924,7 +924,7 @@ DECLARE_EVENT_CLASS(smb3_sync_err_class, __entry->ino = ino; __entry->rc = rc; ), - TP_printk("\tino=%lu rc=%d", + TP_printk("ino=%lu rc=%d", __entry->ino, __entry->rc) ) @@ -950,7 +950,7 @@ DECLARE_EVENT_CLASS(smb3_enter_exit_class, __entry->xid = xid; __assign_str(func_name); ), - TP_printk("\t%s: xid=%u", + TP_printk("%s: xid=%u", __get_str(func_name), __entry->xid) ) diff --git a/fs/smb/client/transport.c b/fs/smb/client/transport.c index 266af17aa7d9..191783f553ce 100644 --- a/fs/smb/client/transport.c +++ b/fs/smb/client/transport.c @@ -1018,14 +1018,16 @@ struct TCP_Server_Info *cifs_pick_channel(struct cifs_ses *ses) uint index = 0; unsigned int min_in_flight = UINT_MAX, max_in_flight = 0; struct TCP_Server_Info *server = NULL; - int i; + int i, start, cur; if (!ses) return NULL; spin_lock(&ses->chan_lock); + start = atomic_inc_return(&ses->chan_seq); for (i = 0; i < ses->chan_count; i++) { - server = ses->chans[i].server; + cur = (start + i) % ses->chan_count; + server = ses->chans[cur].server; if (!server || server->terminate) continue; @@ -1042,17 +1044,15 @@ struct TCP_Server_Info *cifs_pick_channel(struct cifs_ses *ses) */ if (server->in_flight < min_in_flight) { min_in_flight = server->in_flight; - index = i; + index = cur; } if (server->in_flight > max_in_flight) max_in_flight = server->in_flight; } /* if all channels are equally loaded, fall back to round-robin */ - if (min_in_flight == max_in_flight) { - index = (uint)atomic_inc_return(&ses->chan_seq); - index %= ses->chan_count; - } + if (min_in_flight == max_in_flight) + index = (uint)start % ses->chan_count; server = ses->chans[index].server; spin_unlock(&ses->chan_lock); diff --git a/fs/smb/common/smbdirect/smbdirect.h b/fs/smb/common/smbdirect/smbdirect.h new file mode 100644 index 000000000000..b9a385344ff3 --- /dev/null +++ b/fs/smb/common/smbdirect/smbdirect.h @@ -0,0 +1,37 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * Copyright (C) 2017, Microsoft Corporation. + * Copyright (C) 2018, LG Electronics. + */ + +#ifndef __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_H__ +#define __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_H__ + +/* SMB-DIRECT buffer descriptor V1 structure [MS-SMBD] 2.2.3.1 */ +struct smbdirect_buffer_descriptor_v1 { + __le64 offset; + __le32 token; + __le32 length; +} __packed; + +/* + * Connection parameters mostly from [MS-SMBD] 3.1.1.1 + * + * These are setup and negotiated at the beginning of a + * connection and remain constant unless explicitly changed. + * + * Some values are important for the upper layer. + */ +struct smbdirect_socket_parameters { + __u16 recv_credit_max; + __u16 send_credit_target; + __u32 max_send_size; + __u32 max_fragmented_send_size; + __u32 max_recv_size; + __u32 max_fragmented_recv_size; + __u32 max_read_write_size; + __u32 keepalive_interval_msec; + __u32 keepalive_timeout_msec; +} __packed; + +#endif /* __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_H__ */ diff --git a/fs/smb/common/smbdirect/smbdirect_pdu.h b/fs/smb/common/smbdirect/smbdirect_pdu.h new file mode 100644 index 000000000000..ae9fdb05ce23 --- /dev/null +++ b/fs/smb/common/smbdirect/smbdirect_pdu.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * Copyright (c) 2017 Stefan Metzmacher + */ + +#ifndef __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_PDU_H__ +#define __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_PDU_H__ + +#define SMBDIRECT_V1 0x0100 + +/* SMBD negotiation request packet [MS-SMBD] 2.2.1 */ +struct smbdirect_negotiate_req { + __le16 min_version; + __le16 max_version; + __le16 reserved; + __le16 credits_requested; + __le32 preferred_send_size; + __le32 max_receive_size; + __le32 max_fragmented_size; +} __packed; + +/* SMBD negotiation response packet [MS-SMBD] 2.2.2 */ +struct smbdirect_negotiate_resp { + __le16 min_version; + __le16 max_version; + __le16 negotiated_version; + __le16 reserved; + __le16 credits_requested; + __le16 credits_granted; + __le32 status; + __le32 max_readwrite_size; + __le32 preferred_send_size; + __le32 max_receive_size; + __le32 max_fragmented_size; +} __packed; + +#define SMBDIRECT_DATA_MIN_HDR_SIZE 0x14 +#define SMBDIRECT_DATA_OFFSET 0x18 + +#define SMBDIRECT_FLAG_RESPONSE_REQUESTED 0x0001 + +/* SMBD data transfer packet with payload [MS-SMBD] 2.2.3 */ +struct smbdirect_data_transfer { + __le16 credits_requested; + __le16 credits_granted; + __le16 flags; + __le16 reserved; + __le32 remaining_data_length; + __le32 data_offset; + __le32 data_length; + __le32 padding; + __u8 buffer[]; +} __packed; + +#endif /* __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_PDU_H__ */ diff --git a/fs/smb/common/smbdirect/smbdirect_socket.h b/fs/smb/common/smbdirect/smbdirect_socket.h new file mode 100644 index 000000000000..e5b15cc44a7b --- /dev/null +++ b/fs/smb/common/smbdirect/smbdirect_socket.h @@ -0,0 +1,43 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * Copyright (c) 2025 Stefan Metzmacher + */ + +#ifndef __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_SOCKET_H__ +#define __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_SOCKET_H__ + +enum smbdirect_socket_status { + SMBDIRECT_SOCKET_CREATED, + SMBDIRECT_SOCKET_CONNECTING, + SMBDIRECT_SOCKET_CONNECTED, + SMBDIRECT_SOCKET_NEGOTIATE_FAILED, + SMBDIRECT_SOCKET_DISCONNECTING, + SMBDIRECT_SOCKET_DISCONNECTED, + SMBDIRECT_SOCKET_DESTROYED +}; + +struct smbdirect_socket { + enum smbdirect_socket_status status; + + /* RDMA related */ + struct { + struct rdma_cm_id *cm_id; + } rdma; + + /* IB verbs related */ + struct { + struct ib_pd *pd; + struct ib_cq *send_cq; + struct ib_cq *recv_cq; + + /* + * shortcuts for rdma.cm_id->{qp,device}; + */ + struct ib_qp *qp; + struct ib_device *dev; + } ib; + + struct smbdirect_socket_parameters parameters; +}; + +#endif /* __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_SOCKET_H__ */ diff --git a/fs/smb/server/Kconfig b/fs/smb/server/Kconfig index cf70e96ad4de..4a23a5e7e8fe 100644 --- a/fs/smb/server/Kconfig +++ b/fs/smb/server/Kconfig @@ -11,6 +11,7 @@ config SMB_SERVER select CRYPTO_HMAC select CRYPTO_ECB select CRYPTO_LIB_DES + select CRYPTO_LIB_SHA256 select CRYPTO_SHA256 select CRYPTO_CMAC select CRYPTO_SHA512 diff --git a/fs/smb/server/auth.c b/fs/smb/server/auth.c index b3d121052408..d99871c21451 100644 --- a/fs/smb/server/auth.c +++ b/fs/smb/server/auth.c @@ -979,40 +979,6 @@ out: return rc; } -int ksmbd_gen_sd_hash(struct ksmbd_conn *conn, char *sd_buf, int len, - __u8 *pi_hash) -{ - int rc; - struct ksmbd_crypto_ctx *ctx = NULL; - - ctx = ksmbd_crypto_ctx_find_sha256(); - if (!ctx) { - ksmbd_debug(AUTH, "could not alloc sha256\n"); - return -ENOMEM; - } - - rc = crypto_shash_init(CRYPTO_SHA256(ctx)); - if (rc) { - ksmbd_debug(AUTH, "could not init shashn"); - goto out; - } - - rc = crypto_shash_update(CRYPTO_SHA256(ctx), sd_buf, len); - if (rc) { - ksmbd_debug(AUTH, "could not update with n\n"); - goto out; - } - - rc = crypto_shash_final(CRYPTO_SHA256(ctx), pi_hash); - if (rc) { - ksmbd_debug(AUTH, "Could not generate hash err : %d\n", rc); - goto out; - } -out: - ksmbd_release_crypto_ctx(ctx); - return rc; -} - static int ksmbd_get_encryption_key(struct ksmbd_work *work, __u64 ses_id, int enc, u8 *key) { diff --git a/fs/smb/server/auth.h b/fs/smb/server/auth.h index 362b6159a6cf..6879a1bd1b91 100644 --- a/fs/smb/server/auth.h +++ b/fs/smb/server/auth.h @@ -66,6 +66,4 @@ int ksmbd_gen_smb311_encryptionkey(struct ksmbd_conn *conn, struct ksmbd_session *sess); int ksmbd_gen_preauth_integrity_hash(struct ksmbd_conn *conn, char *buf, __u8 *pi_hash); -int ksmbd_gen_sd_hash(struct ksmbd_conn *conn, char *sd_buf, int len, - __u8 *pi_hash); #endif diff --git a/fs/smb/server/connection.c b/fs/smb/server/connection.c index 83764c230e9d..3f04a2977ba8 100644 --- a/fs/smb/server/connection.c +++ b/fs/smb/server/connection.c @@ -40,7 +40,7 @@ void ksmbd_conn_free(struct ksmbd_conn *conn) kvfree(conn->request_buf); kfree(conn->preauth_info); if (atomic_dec_and_test(&conn->refcnt)) { - ksmbd_free_transport(conn->transport); + conn->transport->ops->free_transport(conn->transport); kfree(conn); } } diff --git a/fs/smb/server/connection.h b/fs/smb/server/connection.h index 14620e147dda..dd3e0e3f7bf0 100644 --- a/fs/smb/server/connection.h +++ b/fs/smb/server/connection.h @@ -108,6 +108,7 @@ struct ksmbd_conn { __le16 signing_algorithm; bool binding; atomic_t refcnt; + bool is_aapl; }; struct ksmbd_conn_ops { @@ -132,6 +133,7 @@ struct ksmbd_transport_ops { void *buf, unsigned int len, struct smb2_buffer_desc_v1 *desc, unsigned int desc_len); + void (*free_transport)(struct ksmbd_transport *kt); }; struct ksmbd_transport { diff --git a/fs/smb/server/crypto_ctx.c b/fs/smb/server/crypto_ctx.c index ce733dc9a4a3..80bd68c8635e 100644 --- a/fs/smb/server/crypto_ctx.c +++ b/fs/smb/server/crypto_ctx.c @@ -75,9 +75,6 @@ static struct shash_desc *alloc_shash_desc(int id) case CRYPTO_SHASH_CMACAES: tfm = crypto_alloc_shash("cmac(aes)", 0, 0); break; - case CRYPTO_SHASH_SHA256: - tfm = crypto_alloc_shash("sha256", 0, 0); - break; case CRYPTO_SHASH_SHA512: tfm = crypto_alloc_shash("sha512", 0, 0); break; @@ -198,11 +195,6 @@ struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_cmacaes(void) return ____crypto_shash_ctx_find(CRYPTO_SHASH_CMACAES); } -struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_sha256(void) -{ - return ____crypto_shash_ctx_find(CRYPTO_SHASH_SHA256); -} - struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_sha512(void) { return ____crypto_shash_ctx_find(CRYPTO_SHASH_SHA512); diff --git a/fs/smb/server/crypto_ctx.h b/fs/smb/server/crypto_ctx.h index 4a367c62f653..ac64801d52d3 100644 --- a/fs/smb/server/crypto_ctx.h +++ b/fs/smb/server/crypto_ctx.h @@ -13,7 +13,6 @@ enum { CRYPTO_SHASH_HMACMD5 = 0, CRYPTO_SHASH_HMACSHA256, CRYPTO_SHASH_CMACAES, - CRYPTO_SHASH_SHA256, CRYPTO_SHASH_SHA512, CRYPTO_SHASH_MAX, }; @@ -39,14 +38,12 @@ struct ksmbd_crypto_ctx { #define CRYPTO_HMACMD5(c) ((c)->desc[CRYPTO_SHASH_HMACMD5]) #define CRYPTO_HMACSHA256(c) ((c)->desc[CRYPTO_SHASH_HMACSHA256]) #define CRYPTO_CMACAES(c) ((c)->desc[CRYPTO_SHASH_CMACAES]) -#define CRYPTO_SHA256(c) ((c)->desc[CRYPTO_SHASH_SHA256]) #define CRYPTO_SHA512(c) ((c)->desc[CRYPTO_SHASH_SHA512]) #define CRYPTO_HMACMD5_TFM(c) ((c)->desc[CRYPTO_SHASH_HMACMD5]->tfm) #define CRYPTO_HMACSHA256_TFM(c)\ ((c)->desc[CRYPTO_SHASH_HMACSHA256]->tfm) #define CRYPTO_CMACAES_TFM(c) ((c)->desc[CRYPTO_SHASH_CMACAES]->tfm) -#define CRYPTO_SHA256_TFM(c) ((c)->desc[CRYPTO_SHASH_SHA256]->tfm) #define CRYPTO_SHA512_TFM(c) ((c)->desc[CRYPTO_SHASH_SHA512]->tfm) #define CRYPTO_GCM(c) ((c)->ccmaes[CRYPTO_AEAD_AES_GCM]) @@ -57,7 +54,6 @@ struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_hmacmd5(void); struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_hmacsha256(void); struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_cmacaes(void); struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_sha512(void); -struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_sha256(void); struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_gcm(void); struct ksmbd_crypto_ctx *ksmbd_crypto_ctx_find_ccm(void); void ksmbd_crypto_destroy(void); diff --git a/fs/smb/server/server.c b/fs/smb/server/server.c index ab533c602987..8c9c49c3a0a4 100644 --- a/fs/smb/server/server.c +++ b/fs/smb/server/server.c @@ -631,6 +631,5 @@ MODULE_SOFTDEP("pre: sha512"); MODULE_SOFTDEP("pre: aead2"); MODULE_SOFTDEP("pre: ccm"); MODULE_SOFTDEP("pre: gcm"); -MODULE_SOFTDEP("pre: crc32"); module_init(ksmbd_server_init) module_exit(ksmbd_server_exit) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 8d414239b3fe..63d17cea2e95 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -1607,17 +1607,18 @@ static int krb5_authenticate(struct ksmbd_work *work, out_len = work->response_sz - (le16_to_cpu(rsp->SecurityBufferOffset) + 4); - /* Check previous session */ - prev_sess_id = le64_to_cpu(req->PreviousSessionId); - if (prev_sess_id && prev_sess_id != sess->id) - destroy_previous_session(conn, sess->user, prev_sess_id); - retval = ksmbd_krb5_authenticate(sess, in_blob, in_len, out_blob, &out_len); if (retval) { ksmbd_debug(SMB, "krb5 authentication failed\n"); return -EINVAL; } + + /* Check previous session */ + prev_sess_id = le64_to_cpu(req->PreviousSessionId); + if (prev_sess_id && prev_sess_id != sess->id) + destroy_previous_session(conn, sess->user, prev_sess_id); + rsp->SecurityBufferLength = cpu_to_le16(out_len); if ((conn->sign || server_conf.enforced_signing) || @@ -2874,7 +2875,7 @@ int smb2_open(struct ksmbd_work *work) int req_op_level = 0, open_flags = 0, may_flags = 0, file_info = 0; int rc = 0; int contxt_cnt = 0, query_disk_id = 0; - int maximal_access_ctxt = 0, posix_ctxt = 0; + bool maximal_access_ctxt = false, posix_ctxt = false; int s_type = 0; int next_off = 0; char *name = NULL; @@ -2903,6 +2904,27 @@ int smb2_open(struct ksmbd_work *work) return create_smb2_pipe(work); } + if (req->CreateContextsOffset && tcon->posix_extensions) { + context = smb2_find_context_vals(req, SMB2_CREATE_TAG_POSIX, 16); + if (IS_ERR(context)) { + rc = PTR_ERR(context); + goto err_out2; + } else if (context) { + struct create_posix *posix = (struct create_posix *)context; + + if (le16_to_cpu(context->DataOffset) + + le32_to_cpu(context->DataLength) < + sizeof(struct create_posix) - 4) { + rc = -EINVAL; + goto err_out2; + } + ksmbd_debug(SMB, "get posix context\n"); + + posix_mode = le32_to_cpu(posix->Mode); + posix_ctxt = true; + } + } + if (req->NameLength) { name = smb2_get_name((char *)req + le16_to_cpu(req->NameOffset), le16_to_cpu(req->NameLength), @@ -2925,9 +2947,11 @@ int smb2_open(struct ksmbd_work *work) goto err_out2; } - rc = ksmbd_validate_filename(name); - if (rc < 0) - goto err_out2; + if (posix_ctxt == false) { + rc = ksmbd_validate_filename(name); + if (rc < 0) + goto err_out2; + } if (ksmbd_share_veto_filename(share, name)) { rc = -ENOENT; @@ -3085,28 +3109,6 @@ int smb2_open(struct ksmbd_work *work) rc = -EBADF; goto err_out2; } - - if (tcon->posix_extensions) { - context = smb2_find_context_vals(req, - SMB2_CREATE_TAG_POSIX, 16); - if (IS_ERR(context)) { - rc = PTR_ERR(context); - goto err_out2; - } else if (context) { - struct create_posix *posix = - (struct create_posix *)context; - if (le16_to_cpu(context->DataOffset) + - le32_to_cpu(context->DataLength) < - sizeof(struct create_posix) - 4) { - rc = -EINVAL; - goto err_out2; - } - ksmbd_debug(SMB, "get posix context\n"); - - posix_mode = le32_to_cpu(posix->Mode); - posix_ctxt = 1; - } - } } if (ksmbd_override_fsids(work)) { @@ -3539,6 +3541,15 @@ int smb2_open(struct ksmbd_work *work) ksmbd_debug(SMB, "get query on disk id context\n"); query_disk_id = 1; } + + if (conn->is_aapl == false) { + context = smb2_find_context_vals(req, SMB2_CREATE_AAPL, 4); + if (IS_ERR(context)) { + rc = PTR_ERR(context); + goto err_out1; + } else if (context) + conn->is_aapl = true; + } } rc = ksmbd_vfs_getattr(&path, &stat); @@ -3978,7 +3989,10 @@ static int smb2_populate_readdir_entry(struct ksmbd_conn *conn, int info_level, if (dinfo->EaSize) dinfo->ExtFileAttributes = FILE_ATTRIBUTE_REPARSE_POINT_LE; dinfo->Reserved = 0; - dinfo->UniqueId = cpu_to_le64(ksmbd_kstat->kstat->ino); + if (conn->is_aapl) + dinfo->UniqueId = 0; + else + dinfo->UniqueId = cpu_to_le64(ksmbd_kstat->kstat->ino); if (d_info->hide_dot_file && d_info->name[0] == '.') dinfo->ExtFileAttributes |= FILE_ATTRIBUTE_HIDDEN_LE; memcpy(dinfo->FileName, conv_name, conv_len); @@ -3995,7 +4009,10 @@ static int smb2_populate_readdir_entry(struct ksmbd_conn *conn, int info_level, smb2_get_reparse_tag_special_file(ksmbd_kstat->kstat->mode); if (fibdinfo->EaSize) fibdinfo->ExtFileAttributes = FILE_ATTRIBUTE_REPARSE_POINT_LE; - fibdinfo->UniqueId = cpu_to_le64(ksmbd_kstat->kstat->ino); + if (conn->is_aapl) + fibdinfo->UniqueId = 0; + else + fibdinfo->UniqueId = cpu_to_le64(ksmbd_kstat->kstat->ino); fibdinfo->ShortNameLength = 0; fibdinfo->Reserved = 0; fibdinfo->Reserved2 = cpu_to_le16(0); @@ -4855,8 +4872,13 @@ static int get_file_standard_info(struct smb2_query_info_rsp *rsp, sinfo = (struct smb2_file_standard_info *)rsp->Buffer; delete_pending = ksmbd_inode_pending_delete(fp); - sinfo->AllocationSize = cpu_to_le64(stat.blocks << 9); - sinfo->EndOfFile = S_ISDIR(stat.mode) ? 0 : cpu_to_le64(stat.size); + if (ksmbd_stream_fd(fp) == false) { + sinfo->AllocationSize = cpu_to_le64(stat.blocks << 9); + sinfo->EndOfFile = S_ISDIR(stat.mode) ? 0 : cpu_to_le64(stat.size); + } else { + sinfo->AllocationSize = cpu_to_le64(fp->stream.size); + sinfo->EndOfFile = cpu_to_le64(fp->stream.size); + } sinfo->NumberOfLinks = cpu_to_le32(get_nlink(&stat) - delete_pending); sinfo->DeletePending = delete_pending; sinfo->Directory = S_ISDIR(stat.mode) ? 1 : 0; @@ -4919,9 +4941,14 @@ static int get_file_all_info(struct ksmbd_work *work, file_info->ChangeTime = cpu_to_le64(time); file_info->Attributes = fp->f_ci->m_fattr; file_info->Pad1 = 0; - file_info->AllocationSize = - cpu_to_le64(stat.blocks << 9); - file_info->EndOfFile = S_ISDIR(stat.mode) ? 0 : cpu_to_le64(stat.size); + if (ksmbd_stream_fd(fp) == false) { + file_info->AllocationSize = + cpu_to_le64(stat.blocks << 9); + file_info->EndOfFile = S_ISDIR(stat.mode) ? 0 : cpu_to_le64(stat.size); + } else { + file_info->AllocationSize = cpu_to_le64(fp->stream.size); + file_info->EndOfFile = cpu_to_le64(fp->stream.size); + } file_info->NumberOfLinks = cpu_to_le32(get_nlink(&stat) - delete_pending); file_info->DeletePending = delete_pending; @@ -4930,7 +4957,10 @@ static int get_file_all_info(struct ksmbd_work *work, file_info->IndexNumber = cpu_to_le64(stat.ino); file_info->EASize = 0; file_info->AccessFlags = fp->daccess; - file_info->CurrentByteOffset = cpu_to_le64(fp->filp->f_pos); + if (ksmbd_stream_fd(fp) == false) + file_info->CurrentByteOffset = cpu_to_le64(fp->filp->f_pos); + else + file_info->CurrentByteOffset = cpu_to_le64(fp->stream.pos); file_info->Mode = fp->coption; file_info->AlignmentRequirement = 0; conv_len = smbConvertToUTF16((__le16 *)file_info->FileName, filename, @@ -5118,8 +5148,13 @@ static int get_file_network_open_info(struct smb2_query_info_rsp *rsp, time = ksmbd_UnixTimeToNT(stat.ctime); file_info->ChangeTime = cpu_to_le64(time); file_info->Attributes = fp->f_ci->m_fattr; - file_info->AllocationSize = cpu_to_le64(stat.blocks << 9); - file_info->EndOfFile = S_ISDIR(stat.mode) ? 0 : cpu_to_le64(stat.size); + if (ksmbd_stream_fd(fp) == false) { + file_info->AllocationSize = cpu_to_le64(stat.blocks << 9); + file_info->EndOfFile = S_ISDIR(stat.mode) ? 0 : cpu_to_le64(stat.size); + } else { + file_info->AllocationSize = cpu_to_le64(fp->stream.size); + file_info->EndOfFile = cpu_to_le64(fp->stream.size); + } file_info->Reserved = cpu_to_le32(0); rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_ntwrk_info)); @@ -5142,7 +5177,11 @@ static void get_file_position_info(struct smb2_query_info_rsp *rsp, struct smb2_file_pos_info *file_info; file_info = (struct smb2_file_pos_info *)rsp->Buffer; - file_info->CurrentByteOffset = cpu_to_le64(fp->filp->f_pos); + if (ksmbd_stream_fd(fp) == false) + file_info->CurrentByteOffset = cpu_to_le64(fp->filp->f_pos); + else + file_info->CurrentByteOffset = cpu_to_le64(fp->stream.pos); + rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_pos_info)); } @@ -5231,8 +5270,13 @@ static int find_file_posix_info(struct smb2_query_info_rsp *rsp, file_info->ChangeTime = cpu_to_le64(time); file_info->DosAttributes = fp->f_ci->m_fattr; file_info->Inode = cpu_to_le64(stat.ino); - file_info->EndOfFile = cpu_to_le64(stat.size); - file_info->AllocationSize = cpu_to_le64(stat.blocks << 9); + if (ksmbd_stream_fd(fp) == false) { + file_info->EndOfFile = cpu_to_le64(stat.size); + file_info->AllocationSize = cpu_to_le64(stat.blocks << 9); + } else { + file_info->EndOfFile = cpu_to_le64(fp->stream.size); + file_info->AllocationSize = cpu_to_le64(fp->stream.size); + } file_info->HardLinks = cpu_to_le32(stat.nlink); file_info->Mode = cpu_to_le32(stat.mode & 0777); switch (stat.mode & S_IFMT) { @@ -6174,6 +6218,9 @@ static int set_file_allocation_info(struct ksmbd_work *work, if (!(fp->daccess & FILE_WRITE_DATA_LE)) return -EACCES; + if (ksmbd_stream_fd(fp) == true) + return 0; + rc = vfs_getattr(&fp->filp->f_path, &stat, STATX_BASIC_STATS, AT_STATX_SYNC_AS_STAT); if (rc) @@ -6232,7 +6279,8 @@ static int set_end_of_file_info(struct ksmbd_work *work, struct ksmbd_file *fp, * truncate of some filesystem like FAT32 fill zero data in * truncated range. */ - if (inode->i_sb->s_magic != MSDOS_SUPER_MAGIC) { + if (inode->i_sb->s_magic != MSDOS_SUPER_MAGIC && + ksmbd_stream_fd(fp) == false) { ksmbd_debug(SMB, "truncated to newsize %lld\n", newsize); rc = ksmbd_vfs_truncate(work, fp, newsize); if (rc) { @@ -6305,7 +6353,13 @@ static int set_file_position_info(struct ksmbd_file *fp, return -EINVAL; } - fp->filp->f_pos = current_byte_offset; + if (ksmbd_stream_fd(fp) == false) + fp->filp->f_pos = current_byte_offset; + else { + if (current_byte_offset > XATTR_SIZE_MAX) + current_byte_offset = XATTR_SIZE_MAX; + fp->stream.pos = current_byte_offset; + } return 0; } @@ -8519,11 +8573,6 @@ static void smb20_oplock_break_ack(struct ksmbd_work *work) goto err_out; } - opinfo->op_state = OPLOCK_STATE_NONE; - wake_up_interruptible_all(&opinfo->oplock_q); - opinfo_put(opinfo); - ksmbd_fd_put(work, fp); - rsp->StructureSize = cpu_to_le16(24); rsp->OplockLevel = rsp_oplevel; rsp->Reserved = 0; @@ -8531,16 +8580,15 @@ static void smb20_oplock_break_ack(struct ksmbd_work *work) rsp->VolatileFid = volatile_id; rsp->PersistentFid = persistent_id; ret = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_oplock_break)); - if (!ret) - return; - + if (ret) { err_out: + smb2_set_err_rsp(work); + } + opinfo->op_state = OPLOCK_STATE_NONE; wake_up_interruptible_all(&opinfo->oplock_q); - opinfo_put(opinfo); ksmbd_fd_put(work, fp); - smb2_set_err_rsp(work); } static int check_lease_state(struct lease *lease, __le32 req_state) @@ -8670,11 +8718,6 @@ static void smb21_lease_break_ack(struct ksmbd_work *work) } lease_state = lease->state; - opinfo->op_state = OPLOCK_STATE_NONE; - wake_up_interruptible_all(&opinfo->oplock_q); - atomic_dec(&opinfo->breaking_cnt); - wake_up_interruptible_all(&opinfo->oplock_brk); - opinfo_put(opinfo); rsp->StructureSize = cpu_to_le16(36); rsp->Reserved = 0; @@ -8683,16 +8726,16 @@ static void smb21_lease_break_ack(struct ksmbd_work *work) rsp->LeaseState = lease_state; rsp->LeaseDuration = 0; ret = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_lease_ack)); - if (!ret) - return; - + if (ret) { err_out: + smb2_set_err_rsp(work); + } + + opinfo->op_state = OPLOCK_STATE_NONE; wake_up_interruptible_all(&opinfo->oplock_q); atomic_dec(&opinfo->breaking_cnt); wake_up_interruptible_all(&opinfo->oplock_brk); - opinfo_put(opinfo); - smb2_set_err_rsp(work); } /** diff --git a/fs/smb/server/smb2pdu.h b/fs/smb/server/smb2pdu.h index 17a0b18a8406..16ae8a10490b 100644 --- a/fs/smb/server/smb2pdu.h +++ b/fs/smb/server/smb2pdu.h @@ -63,6 +63,9 @@ struct preauth_integrity_info { #define SMB2_SESSION_TIMEOUT (10 * HZ) +/* Apple Defined Contexts */ +#define SMB2_CREATE_AAPL "AAPL" + struct create_durable_req_v2 { struct create_context_hdr ccontext; __u8 Name[8]; diff --git a/fs/smb/server/transport_rdma.c b/fs/smb/server/transport_rdma.c index 4998df04ab95..c6cbe0d56e32 100644 --- a/fs/smb/server/transport_rdma.c +++ b/fs/smb/server/transport_rdma.c @@ -159,7 +159,8 @@ struct smb_direct_transport { }; #define KSMBD_TRANS(t) ((struct ksmbd_transport *)&((t)->transport)) - +#define SMBD_TRANS(t) ((struct smb_direct_transport *)container_of(t, \ + struct smb_direct_transport, transport)) enum { SMB_DIRECT_MSG_NEGOTIATE_REQ = 0, SMB_DIRECT_MSG_DATA_TRANSFER @@ -410,6 +411,11 @@ err: return NULL; } +static void smb_direct_free_transport(struct ksmbd_transport *kt) +{ + kfree(SMBD_TRANS(kt)); +} + static void free_transport(struct smb_direct_transport *t) { struct smb_direct_recvmsg *recvmsg; @@ -427,7 +433,8 @@ static void free_transport(struct smb_direct_transport *t) if (t->qp) { ib_drain_qp(t->qp); ib_mr_pool_destroy(t->qp, &t->qp->rdma_mrs); - ib_destroy_qp(t->qp); + t->qp = NULL; + rdma_destroy_qp(t->cm_id); } ksmbd_debug(RDMA, "drain the reassembly queue\n"); @@ -455,7 +462,6 @@ static void free_transport(struct smb_direct_transport *t) smb_direct_destroy_pools(t); ksmbd_conn_free(KSMBD_TRANS(t)->conn); - kfree(t); } static struct smb_direct_sendmsg @@ -1935,8 +1941,8 @@ static int smb_direct_create_qpair(struct smb_direct_transport *t, return 0; err: if (t->qp) { - ib_destroy_qp(t->qp); t->qp = NULL; + rdma_destroy_qp(t->cm_id); } if (t->recv_cq) { ib_destroy_cq(t->recv_cq); @@ -2281,4 +2287,5 @@ static const struct ksmbd_transport_ops ksmbd_smb_direct_transport_ops = { .read = smb_direct_read, .rdma_read = smb_direct_rdma_read, .rdma_write = smb_direct_rdma_write, + .free_transport = smb_direct_free_transport, }; diff --git a/fs/smb/server/transport_tcp.c b/fs/smb/server/transport_tcp.c index abedf510899a..4e9f98db9ff4 100644 --- a/fs/smb/server/transport_tcp.c +++ b/fs/smb/server/transport_tcp.c @@ -93,7 +93,7 @@ static struct tcp_transport *alloc_transport(struct socket *client_sk) return t; } -void ksmbd_free_transport(struct ksmbd_transport *kt) +static void ksmbd_tcp_free_transport(struct ksmbd_transport *kt) { struct tcp_transport *t = TCP_TRANS(kt); @@ -656,4 +656,5 @@ static const struct ksmbd_transport_ops ksmbd_tcp_transport_ops = { .read = ksmbd_tcp_read, .writev = ksmbd_tcp_writev, .disconnect = ksmbd_tcp_disconnect, + .free_transport = ksmbd_tcp_free_transport, }; diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c index baf0d3031a44..d3437f6644e3 100644 --- a/fs/smb/server/vfs.c +++ b/fs/smb/server/vfs.c @@ -4,6 +4,7 @@ * Copyright (C) 2018 Samsung Electronics Co., Ltd. */ +#include <crypto/sha2.h> #include <linux/kernel.h> #include <linux/fs.h> #include <linux/filelock.h> @@ -292,6 +293,7 @@ static int ksmbd_vfs_stream_read(struct ksmbd_file *fp, char *buf, loff_t *pos, if (v_len - *pos < count) count = v_len - *pos; + fp->stream.pos = v_len; memcpy(buf, &stream_buf[*pos], count); @@ -455,8 +457,8 @@ static int ksmbd_vfs_stream_write(struct ksmbd_file *fp, char *buf, loff_t *pos, true); if (err < 0) goto out; - - fp->filp->f_pos = *pos; + else + fp->stream.pos = size; err = 0; out: kvfree(stream_buf); @@ -1280,6 +1282,7 @@ out1: err = ksmbd_vfs_lock_parent(parent_path->dentry, path->dentry); if (err) { + mnt_drop_write(parent_path->mnt); path_put(path); path_put(parent_path); } @@ -1476,11 +1479,7 @@ int ksmbd_vfs_set_sd_xattr(struct ksmbd_conn *conn, acl.sd_buf = (char *)pntsd; acl.sd_size = len; - rc = ksmbd_gen_sd_hash(conn, acl.sd_buf, acl.sd_size, acl.hash); - if (rc) { - pr_err("failed to generate hash for ndr acl\n"); - return rc; - } + sha256(acl.sd_buf, acl.sd_size, acl.hash); smb_acl = ksmbd_vfs_make_xattr_posix_acl(idmap, inode, ACL_TYPE_ACCESS); @@ -1495,12 +1494,7 @@ int ksmbd_vfs_set_sd_xattr(struct ksmbd_conn *conn, goto out; } - rc = ksmbd_gen_sd_hash(conn, acl_ndr.data, acl_ndr.offset, - acl.posix_acl_hash); - if (rc) { - pr_err("failed to generate hash for ndr acl\n"); - goto out; - } + sha256(acl_ndr.data, acl_ndr.offset, acl.posix_acl_hash); rc = ndr_encode_v4_ntacl(&sd_ndr, &acl); if (rc) { @@ -1557,11 +1551,7 @@ int ksmbd_vfs_get_sd_xattr(struct ksmbd_conn *conn, goto out_free; } - rc = ksmbd_gen_sd_hash(conn, acl_ndr.data, acl_ndr.offset, cmp_hash); - if (rc) { - pr_err("failed to generate hash for ndr acl\n"); - goto out_free; - } + sha256(acl_ndr.data, acl_ndr.offset, cmp_hash); if (memcmp(cmp_hash, acl.posix_acl_hash, XATTR_SD_HASH_SIZE)) { pr_err("hash value diff\n"); diff --git a/fs/smb/server/vfs_cache.h b/fs/smb/server/vfs_cache.h index 5bbb179736c2..0708155b5caf 100644 --- a/fs/smb/server/vfs_cache.h +++ b/fs/smb/server/vfs_cache.h @@ -44,6 +44,7 @@ struct ksmbd_lock { struct stream { char *name; ssize_t size; + loff_t pos; }; struct ksmbd_inode { diff --git a/fs/super.c b/fs/super.c index 21799e213fd7..80418ca8e215 100644 --- a/fs/super.c +++ b/fs/super.c @@ -964,8 +964,10 @@ void iterate_supers_type(struct file_system_type *type, spin_unlock(&sb_lock); locked = super_lock_shared(sb); - if (locked) + if (locked) { f(sb, arg); + super_unlock_shared(sb); + } spin_lock(&sb_lock); if (p) diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index ee954e64ce7f..e28ab4395e5c 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -985,7 +985,7 @@ int ubifs_jnl_write_inode(struct ubifs_info *c, const struct inode *inode) dbg_jnl("ino %lu, nlink %u", inode->i_ino, inode->i_nlink); if (kill_xattrs && ui->xattr_cnt > ubifs_xattr_max_cnt(c)) { - ubifs_err(c, "Cannot delete inode, it has too much xattrs!"); + ubifs_err(c, "Cannot delete inode, it has too many xattrs!"); err = -EPERM; ubifs_ro_mode(c, err); return err; diff --git a/fs/ufs/super.c b/fs/ufs/super.c index eea718ac66b4..6e4585169f94 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -397,7 +397,7 @@ static int ufs_parse_param(struct fs_context *fc, struct fs_parameter *param) pr_err("ufstype can't be changed during remount\n"); return -EINVAL; } - if (!ctx->flavour) { + if (ctx->flavour) { pr_err("conflicting ufstype options\n"); return -EINVAL; } diff --git a/fs/xattr.c b/fs/xattr.c index 8ec5b0204bfd..600ae97969cf 100644 --- a/fs/xattr.c +++ b/fs/xattr.c @@ -1479,6 +1479,7 @@ ssize_t simple_xattr_list(struct inode *inode, struct simple_xattrs *xattrs, buffer += err; } remaining_size -= err; + err = 0; read_lock(&xattrs->lock); for (rbp = rb_first(&xattrs->rb_root); rbp; rbp = rb_next(rbp)) { diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index 7839efe050bf..000cc7f4a3ce 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -3444,16 +3444,41 @@ xfs_alloc_read_agf( set_bit(XFS_AGSTATE_AGF_INIT, &pag->pag_opstate); } + #ifdef DEBUG - else if (!xfs_is_shutdown(mp)) { - ASSERT(pag->pagf_freeblks == be32_to_cpu(agf->agf_freeblks)); - ASSERT(pag->pagf_btreeblks == be32_to_cpu(agf->agf_btreeblks)); - ASSERT(pag->pagf_flcount == be32_to_cpu(agf->agf_flcount)); - ASSERT(pag->pagf_longest == be32_to_cpu(agf->agf_longest)); - ASSERT(pag->pagf_bno_level == be32_to_cpu(agf->agf_bno_level)); - ASSERT(pag->pagf_cnt_level == be32_to_cpu(agf->agf_cnt_level)); + /* + * It's possible for the AGF to be out of sync if the block device is + * silently dropping writes. This can happen in fstests with dmflakey + * enabled, which allows the buffer to be cleaned and reclaimed by + * memory pressure and then re-read from disk here. We will get a + * stale version of the AGF from disk, and nothing good can happen from + * here. Hence if we detect this situation, immediately shut down the + * filesystem. + * + * This can also happen if we are already in the middle of a forced + * shutdown, so don't bother checking if we are already shut down. + */ + if (!xfs_is_shutdown(pag_mount(pag))) { + bool ok = true; + + ok &= pag->pagf_freeblks == be32_to_cpu(agf->agf_freeblks); + ok &= pag->pagf_freeblks == be32_to_cpu(agf->agf_freeblks); + ok &= pag->pagf_btreeblks == be32_to_cpu(agf->agf_btreeblks); + ok &= pag->pagf_flcount == be32_to_cpu(agf->agf_flcount); + ok &= pag->pagf_longest == be32_to_cpu(agf->agf_longest); + ok &= pag->pagf_bno_level == be32_to_cpu(agf->agf_bno_level); + ok &= pag->pagf_cnt_level == be32_to_cpu(agf->agf_cnt_level); + + if (XFS_IS_CORRUPT(pag_mount(pag), !ok)) { + xfs_ag_mark_sick(pag, XFS_SICK_AG_AGF); + xfs_trans_brelse(tp, agfbp); + xfs_force_shutdown(pag_mount(pag), + SHUTDOWN_CORRUPT_ONDISK); + return -EFSCORRUPTED; + } } -#endif +#endif /* DEBUG */ + if (agfbpp) *agfbpp = agfbp; else diff --git a/fs/xfs/libxfs/xfs_group.c b/fs/xfs/libxfs/xfs_group.c index e9d76bcdc820..20ad7c309489 100644 --- a/fs/xfs/libxfs/xfs_group.c +++ b/fs/xfs/libxfs/xfs_group.c @@ -163,7 +163,8 @@ xfs_group_free( xfs_defer_drain_free(&xg->xg_intents_drain); #ifdef __KERNEL__ - kfree(xg->xg_busy_extents); + if (xfs_group_has_extent_busy(xg->xg_mount, xg->xg_type)) + kfree(xg->xg_busy_extents); #endif if (uninit) @@ -189,9 +190,11 @@ xfs_group_insert( xg->xg_type = type; #ifdef __KERNEL__ - xg->xg_busy_extents = xfs_extent_busy_alloc(); - if (!xg->xg_busy_extents) - return -ENOMEM; + if (xfs_group_has_extent_busy(mp, type)) { + xg->xg_busy_extents = xfs_extent_busy_alloc(); + if (!xg->xg_busy_extents) + return -ENOMEM; + } spin_lock_init(&xg->xg_state_lock); xfs_hooks_init(&xg->xg_rmap_update_hooks); #endif @@ -210,7 +213,8 @@ xfs_group_insert( out_drain: xfs_defer_drain_free(&xg->xg_intents_drain); #ifdef __KERNEL__ - kfree(xg->xg_busy_extents); + if (xfs_group_has_extent_busy(xg->xg_mount, xg->xg_type)) + kfree(xg->xg_busy_extents); #endif return error; } diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c index 0c47b5c6ca7d..750111634d9f 100644 --- a/fs/xfs/libxfs/xfs_ialloc.c +++ b/fs/xfs/libxfs/xfs_ialloc.c @@ -2801,12 +2801,35 @@ xfs_ialloc_read_agi( set_bit(XFS_AGSTATE_AGI_INIT, &pag->pag_opstate); } +#ifdef DEBUG /* - * It's possible for these to be out of sync if - * we are in the middle of a forced shutdown. + * It's possible for the AGF to be out of sync if the block device is + * silently dropping writes. This can happen in fstests with dmflakey + * enabled, which allows the buffer to be cleaned and reclaimed by + * memory pressure and then re-read from disk here. We will get a + * stale version of the AGF from disk, and nothing good can happen from + * here. Hence if we detect this situation, immediately shut down the + * filesystem. + * + * This can also happen if we are already in the middle of a forced + * shutdown, so don't bother checking if we are already shut down. */ - ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) || - xfs_is_shutdown(pag_mount(pag))); + if (!xfs_is_shutdown(pag_mount(pag))) { + bool ok = true; + + ok &= pag->pagi_freecount == be32_to_cpu(agi->agi_freecount); + ok &= pag->pagi_count == be32_to_cpu(agi->agi_count); + + if (XFS_IS_CORRUPT(pag_mount(pag), !ok)) { + xfs_ag_mark_sick(pag, XFS_SICK_AG_AGI); + xfs_trans_brelse(tp, agibp); + xfs_force_shutdown(pag_mount(pag), + SHUTDOWN_CORRUPT_ONDISK); + return -EFSCORRUPTED; + } + } +#endif /* DEBUG */ + if (agibpp) *agibpp = agibp; else diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 8af83bd161f9..f9ef3b2a332a 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -1683,7 +1683,7 @@ xfs_free_buftarg( fs_put_dax(btp->bt_daxdev, btp->bt_mount); /* the main block device is closed by kill_block_super */ if (btp->bt_bdev != btp->bt_mount->m_super->s_bdev) - bdev_fput(btp->bt_bdev_file); + bdev_fput(btp->bt_file); kfree(btp); } @@ -1712,8 +1712,8 @@ xfs_configure_buftarg_atomic_writes( max_bytes = 0; } - btp->bt_bdev_awu_min = min_bytes; - btp->bt_bdev_awu_max = max_bytes; + btp->bt_awu_min = min_bytes; + btp->bt_awu_max = max_bytes; } /* Configure a buffer target that abstracts a block device. */ @@ -1738,14 +1738,9 @@ xfs_configure_buftarg( return -EINVAL; } - /* - * Flush the block device pagecache so our bios see anything dirtied - * before mount. - */ if (bdev_can_atomic_write(btp->bt_bdev)) xfs_configure_buftarg_atomic_writes(btp); - - return sync_blockdev(btp->bt_bdev); + return 0; } int @@ -1803,7 +1798,7 @@ xfs_alloc_buftarg( btp = kzalloc(sizeof(*btp), GFP_KERNEL | __GFP_NOFAIL); btp->bt_mount = mp; - btp->bt_bdev_file = bdev_file; + btp->bt_file = bdev_file; btp->bt_bdev = file_bdev(bdev_file); btp->bt_dev = btp->bt_bdev->bd_dev; btp->bt_daxdev = fs_dax_get_by_bdev(btp->bt_bdev, &btp->bt_dax_part_off, @@ -2082,44 +2077,6 @@ xfs_buf_delwri_submit( return error; } -/* - * Push a single buffer on a delwri queue. - * - * The purpose of this function is to submit a single buffer of a delwri queue - * and return with the buffer still on the original queue. - * - * The buffer locking and queue management logic between _delwri_pushbuf() and - * _delwri_queue() guarantee that the buffer cannot be queued to another list - * before returning. - */ -int -xfs_buf_delwri_pushbuf( - struct xfs_buf *bp, - struct list_head *buffer_list) -{ - int error; - - ASSERT(bp->b_flags & _XBF_DELWRI_Q); - - trace_xfs_buf_delwri_pushbuf(bp, _RET_IP_); - - xfs_buf_lock(bp); - bp->b_flags &= ~(_XBF_DELWRI_Q | XBF_ASYNC); - bp->b_flags |= XBF_WRITE; - xfs_buf_submit(bp); - - /* - * The buffer is now locked, under I/O but still on the original delwri - * queue. Wait for I/O completion, restore the DELWRI_Q flag and - * return with the buffer unlocked and still on the original queue. - */ - error = xfs_buf_iowait(bp); - bp->b_flags |= _XBF_DELWRI_Q; - xfs_buf_unlock(bp); - - return error; -} - void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref) { /* diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h index 9d2ab567cf81..b269e115d9ac 100644 --- a/fs/xfs/xfs_buf.h +++ b/fs/xfs/xfs_buf.h @@ -94,7 +94,6 @@ void xfs_buf_cache_destroy(struct xfs_buf_cache *bch); */ struct xfs_buftarg { dev_t bt_dev; - struct file *bt_bdev_file; struct block_device *bt_bdev; struct dax_device *bt_daxdev; struct file *bt_file; @@ -112,9 +111,9 @@ struct xfs_buftarg { struct percpu_counter bt_readahead_count; struct ratelimit_state bt_ioerror_rl; - /* Atomic write unit values, bytes */ - unsigned int bt_bdev_awu_min; - unsigned int bt_bdev_awu_max; + /* Hardware atomic write unit values, bytes */ + unsigned int bt_awu_min; + unsigned int bt_awu_max; /* built-in cache, if we're not using the perag one */ struct xfs_buf_cache bt_cache[]; @@ -326,7 +325,6 @@ extern bool xfs_buf_delwri_queue(struct xfs_buf *, struct list_head *); void xfs_buf_delwri_queue_here(struct xfs_buf *bp, struct list_head *bl); extern int xfs_buf_delwri_submit(struct list_head *); extern int xfs_buf_delwri_submit_nowait(struct list_head *); -extern int xfs_buf_delwri_pushbuf(struct xfs_buf *, struct list_head *); static inline xfs_daddr_t xfs_buf_daddr(struct xfs_buf *bp) { @@ -376,7 +374,6 @@ extern void xfs_buftarg_wait(struct xfs_buftarg *); extern void xfs_buftarg_drain(struct xfs_buftarg *); int xfs_configure_buftarg(struct xfs_buftarg *btp, unsigned int sectorsize); -#define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev) #define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev) int xfs_buf_reverify(struct xfs_buf *bp, const struct xfs_buf_ops *ops); diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 90139e0f3271..7fc54725c5f6 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -32,6 +32,61 @@ static inline struct xfs_buf_log_item *BUF_ITEM(struct xfs_log_item *lip) return container_of(lip, struct xfs_buf_log_item, bli_item); } +static void +xfs_buf_item_get_format( + struct xfs_buf_log_item *bip, + int count) +{ + ASSERT(bip->bli_formats == NULL); + bip->bli_format_count = count; + + if (count == 1) { + bip->bli_formats = &bip->__bli_format; + return; + } + + bip->bli_formats = kzalloc(count * sizeof(struct xfs_buf_log_format), + GFP_KERNEL | __GFP_NOFAIL); +} + +static void +xfs_buf_item_free_format( + struct xfs_buf_log_item *bip) +{ + if (bip->bli_formats != &bip->__bli_format) { + kfree(bip->bli_formats); + bip->bli_formats = NULL; + } +} + +static void +xfs_buf_item_free( + struct xfs_buf_log_item *bip) +{ + xfs_buf_item_free_format(bip); + kvfree(bip->bli_item.li_lv_shadow); + kmem_cache_free(xfs_buf_item_cache, bip); +} + +/* + * xfs_buf_item_relse() is called when the buf log item is no longer needed. + */ +static void +xfs_buf_item_relse( + struct xfs_buf_log_item *bip) +{ + struct xfs_buf *bp = bip->bli_buf; + + trace_xfs_buf_item_relse(bp, _RET_IP_); + + ASSERT(!test_bit(XFS_LI_IN_AIL, &bip->bli_item.li_flags)); + ASSERT(atomic_read(&bip->bli_refcount) == 0); + + bp->b_log_item = NULL; + xfs_buf_rele(bp); + xfs_buf_item_free(bip); +} + /* Is this log iovec plausibly large enough to contain the buffer log format? */ bool xfs_buf_log_check_iovec( @@ -390,6 +445,42 @@ xfs_buf_item_pin( } /* + * For a stale BLI, process all the necessary completions that must be + * performed when the final BLI reference goes away. The buffer will be + * referenced and locked here - we return to the caller with the buffer still + * referenced and locked for them to finalise processing of the buffer. + */ +static void +xfs_buf_item_finish_stale( + struct xfs_buf_log_item *bip) +{ + struct xfs_buf *bp = bip->bli_buf; + struct xfs_log_item *lip = &bip->bli_item; + + ASSERT(bip->bli_flags & XFS_BLI_STALE); + ASSERT(xfs_buf_islocked(bp)); + ASSERT(bp->b_flags & XBF_STALE); + ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL); + ASSERT(list_empty(&lip->li_trans)); + ASSERT(!bp->b_transp); + + if (bip->bli_flags & XFS_BLI_STALE_INODE) { + xfs_buf_item_done(bp); + xfs_buf_inode_iodone(bp); + ASSERT(list_empty(&bp->b_li_list)); + return; + } + + /* + * We may or may not be on the AIL here, xfs_trans_ail_delete() will do + * the right thing regardless of the situation in which we are called. + */ + xfs_trans_ail_delete(lip, SHUTDOWN_LOG_IO_ERROR); + xfs_buf_item_relse(bip); + ASSERT(bp->b_log_item == NULL); +} + +/* * This is called to unpin the buffer associated with the buf log item which was * previously pinned with a call to xfs_buf_item_pin(). We enter this function * with a buffer pin count, a buffer reference and a BLI reference. @@ -438,13 +529,6 @@ xfs_buf_item_unpin( } if (stale) { - ASSERT(bip->bli_flags & XFS_BLI_STALE); - ASSERT(xfs_buf_islocked(bp)); - ASSERT(bp->b_flags & XBF_STALE); - ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL); - ASSERT(list_empty(&lip->li_trans)); - ASSERT(!bp->b_transp); - trace_xfs_buf_item_unpin_stale(bip); /* @@ -455,22 +539,7 @@ xfs_buf_item_unpin( * processing is complete. */ xfs_buf_rele(bp); - - /* - * If we get called here because of an IO error, we may or may - * not have the item on the AIL. xfs_trans_ail_delete() will - * take care of that situation. xfs_trans_ail_delete() drops - * the AIL lock. - */ - if (bip->bli_flags & XFS_BLI_STALE_INODE) { - xfs_buf_item_done(bp); - xfs_buf_inode_iodone(bp); - ASSERT(list_empty(&bp->b_li_list)); - } else { - xfs_trans_ail_delete(lip, SHUTDOWN_LOG_IO_ERROR); - xfs_buf_item_relse(bp); - ASSERT(bp->b_log_item == NULL); - } + xfs_buf_item_finish_stale(bip); xfs_buf_relse(bp); return; } @@ -543,43 +612,42 @@ xfs_buf_item_push( * Drop the buffer log item refcount and take appropriate action. This helper * determines whether the bli must be freed or not, since a decrement to zero * does not necessarily mean the bli is unused. - * - * Return true if the bli is freed, false otherwise. */ -bool +void xfs_buf_item_put( struct xfs_buf_log_item *bip) { - struct xfs_log_item *lip = &bip->bli_item; - bool aborted; - bool dirty; + + ASSERT(xfs_buf_islocked(bip->bli_buf)); /* drop the bli ref and return if it wasn't the last one */ if (!atomic_dec_and_test(&bip->bli_refcount)) - return false; + return; - /* - * We dropped the last ref and must free the item if clean or aborted. - * If the bli is dirty and non-aborted, the buffer was clean in the - * transaction but still awaiting writeback from previous changes. In - * that case, the bli is freed on buffer writeback completion. - */ - aborted = test_bit(XFS_LI_ABORTED, &lip->li_flags) || - xlog_is_shutdown(lip->li_log); - dirty = bip->bli_flags & XFS_BLI_DIRTY; - if (dirty && !aborted) - return false; + /* If the BLI is in the AIL, then it is still dirty and in use */ + if (test_bit(XFS_LI_IN_AIL, &bip->bli_item.li_flags)) { + ASSERT(bip->bli_flags & XFS_BLI_DIRTY); + return; + } /* - * The bli is aborted or clean. An aborted item may be in the AIL - * regardless of dirty state. For example, consider an aborted - * transaction that invalidated a dirty bli and cleared the dirty - * state. + * In shutdown conditions, we can be asked to free a dirty BLI that + * isn't in the AIL. This can occur due to a checkpoint aborting a BLI + * instead of inserting it into the AIL at checkpoint IO completion. If + * there's another bli reference (e.g. a btree cursor holds a clean + * reference) and it is released via xfs_trans_brelse(), we can get here + * with that aborted, dirty BLI. In this case, it is safe to free the + * dirty BLI immediately, as it is not in the AIL and there are no + * other references to it. + * + * We should never get here with a stale BLI via that path as + * xfs_trans_brelse() specifically holds onto stale buffers rather than + * releasing them. */ - if (aborted) - xfs_trans_ail_delete(lip, 0); - xfs_buf_item_relse(bip->bli_buf); - return true; + ASSERT(!(bip->bli_flags & XFS_BLI_DIRTY) || + test_bit(XFS_LI_ABORTED, &bip->bli_item.li_flags)); + ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); + xfs_buf_item_relse(bip); } /* @@ -600,6 +668,15 @@ xfs_buf_item_put( * if necessary but do not unlock the buffer. This is for support of * xfs_trans_bhold(). Make sure the XFS_BLI_HOLD field is cleared if we don't * free the item. + * + * If the XFS_BLI_STALE flag is set, the last reference to the BLI *must* + * perform a completion abort of any objects attached to the buffer for IO + * tracking purposes. This generally only happens in shutdown situations, + * normally xfs_buf_item_unpin() will drop the last BLI reference and perform + * completion processing. However, because transaction completion can race with + * checkpoint completion during a shutdown, this release context may end up + * being the last active reference to the BLI and so needs to perform this + * cleanup. */ STATIC void xfs_buf_item_release( @@ -607,18 +684,19 @@ xfs_buf_item_release( { struct xfs_buf_log_item *bip = BUF_ITEM(lip); struct xfs_buf *bp = bip->bli_buf; - bool released; bool hold = bip->bli_flags & XFS_BLI_HOLD; bool stale = bip->bli_flags & XFS_BLI_STALE; -#if defined(DEBUG) || defined(XFS_WARN) - bool ordered = bip->bli_flags & XFS_BLI_ORDERED; - bool dirty = bip->bli_flags & XFS_BLI_DIRTY; bool aborted = test_bit(XFS_LI_ABORTED, &lip->li_flags); + bool dirty = bip->bli_flags & XFS_BLI_DIRTY; +#if defined(DEBUG) || defined(XFS_WARN) + bool ordered = bip->bli_flags & XFS_BLI_ORDERED; #endif trace_xfs_buf_item_release(bip); + ASSERT(xfs_buf_islocked(bp)); + /* * The bli dirty state should match whether the blf has logged segments * except for ordered buffers, where only the bli should be dirty. @@ -634,16 +712,56 @@ xfs_buf_item_release( bp->b_transp = NULL; bip->bli_flags &= ~(XFS_BLI_LOGGED | XFS_BLI_HOLD | XFS_BLI_ORDERED); + /* If there are other references, then we have nothing to do. */ + if (!atomic_dec_and_test(&bip->bli_refcount)) + goto out_release; + + /* + * Stale buffer completion frees the BLI, unlocks and releases the + * buffer. Neither the BLI or buffer are safe to reference after this + * call, so there's nothing more we need to do here. + * + * If we get here with a stale buffer and references to the BLI remain, + * we must not unlock the buffer as the last BLI reference owns lock + * context, not us. + */ + if (stale) { + xfs_buf_item_finish_stale(bip); + xfs_buf_relse(bp); + ASSERT(!hold); + return; + } + /* - * Unref the item and unlock the buffer unless held or stale. Stale - * buffers remain locked until final unpin unless the bli is freed by - * the unref call. The latter implies shutdown because buffer - * invalidation dirties the bli and transaction. + * Dirty or clean, aborted items are done and need to be removed from + * the AIL and released. This frees the BLI, but leaves the buffer + * locked and referenced. */ - released = xfs_buf_item_put(bip); - if (hold || (stale && !released)) + if (aborted || xlog_is_shutdown(lip->li_log)) { + ASSERT(list_empty(&bip->bli_buf->b_li_list)); + xfs_buf_item_done(bp); + goto out_release; + } + + /* + * Clean, unreferenced BLIs can be immediately freed, leaving the buffer + * locked and referenced. + * + * Dirty, unreferenced BLIs *must* be in the AIL awaiting writeback. + */ + if (!dirty) + xfs_buf_item_relse(bip); + else + ASSERT(test_bit(XFS_LI_IN_AIL, &lip->li_flags)); + + /* Not safe to reference the BLI from here */ +out_release: + /* + * If we get here with a stale buffer, we must not unlock the + * buffer as the last BLI reference owns lock context, not us. + */ + if (stale || hold) return; - ASSERT(!stale || aborted); xfs_buf_relse(bp); } @@ -729,33 +847,6 @@ static const struct xfs_item_ops xfs_buf_item_ops = { .iop_push = xfs_buf_item_push, }; -STATIC void -xfs_buf_item_get_format( - struct xfs_buf_log_item *bip, - int count) -{ - ASSERT(bip->bli_formats == NULL); - bip->bli_format_count = count; - - if (count == 1) { - bip->bli_formats = &bip->__bli_format; - return; - } - - bip->bli_formats = kzalloc(count * sizeof(struct xfs_buf_log_format), - GFP_KERNEL | __GFP_NOFAIL); -} - -STATIC void -xfs_buf_item_free_format( - struct xfs_buf_log_item *bip) -{ - if (bip->bli_formats != &bip->__bli_format) { - kfree(bip->bli_formats); - bip->bli_formats = NULL; - } -} - /* * Allocate a new buf log item to go with the given buffer. * Set the buffer's b_log_item field to point to the new @@ -976,34 +1067,6 @@ xfs_buf_item_dirty_format( return false; } -STATIC void -xfs_buf_item_free( - struct xfs_buf_log_item *bip) -{ - xfs_buf_item_free_format(bip); - kvfree(bip->bli_item.li_lv_shadow); - kmem_cache_free(xfs_buf_item_cache, bip); -} - -/* - * xfs_buf_item_relse() is called when the buf log item is no longer needed. - */ -void -xfs_buf_item_relse( - struct xfs_buf *bp) -{ - struct xfs_buf_log_item *bip = bp->b_log_item; - - trace_xfs_buf_item_relse(bp, _RET_IP_); - ASSERT(!test_bit(XFS_LI_IN_AIL, &bip->bli_item.li_flags)); - - if (atomic_read(&bip->bli_refcount)) - return; - bp->b_log_item = NULL; - xfs_buf_rele(bp); - xfs_buf_item_free(bip); -} - void xfs_buf_item_done( struct xfs_buf *bp) @@ -1023,5 +1086,5 @@ xfs_buf_item_done( xfs_trans_ail_delete(&bp->b_log_item->bli_item, (bp->b_flags & _XBF_LOGRECOVERY) ? 0 : SHUTDOWN_CORRUPT_INCORE); - xfs_buf_item_relse(bp); + xfs_buf_item_relse(bp->b_log_item); } diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h index e10e324cd245..416890b84f8c 100644 --- a/fs/xfs/xfs_buf_item.h +++ b/fs/xfs/xfs_buf_item.h @@ -49,8 +49,7 @@ struct xfs_buf_log_item { int xfs_buf_item_init(struct xfs_buf *, struct xfs_mount *); void xfs_buf_item_done(struct xfs_buf *bp); -void xfs_buf_item_relse(struct xfs_buf *); -bool xfs_buf_item_put(struct xfs_buf_log_item *); +void xfs_buf_item_put(struct xfs_buf_log_item *bip); void xfs_buf_item_log(struct xfs_buf_log_item *, uint, uint); bool xfs_buf_item_dirty_format(struct xfs_buf_log_item *); void xfs_buf_inode_iodone(struct xfs_buf *); diff --git a/fs/xfs/xfs_discard.c b/fs/xfs/xfs_discard.c index 94d0873bcd62..603d51365645 100644 --- a/fs/xfs/xfs_discard.c +++ b/fs/xfs/xfs_discard.c @@ -103,24 +103,6 @@ xfs_discard_endio( bio_put(bio); } -static inline struct block_device * -xfs_group_bdev( - const struct xfs_group *xg) -{ - struct xfs_mount *mp = xg->xg_mount; - - switch (xg->xg_type) { - case XG_TYPE_AG: - return mp->m_ddev_targp->bt_bdev; - case XG_TYPE_RTG: - return mp->m_rtdev_targp->bt_bdev; - default: - ASSERT(0); - break; - } - return NULL; -} - /* * Walk the discard list and issue discards on all the busy extents in the * list. We plug and chain the bios so that we only need a single completion @@ -138,11 +120,14 @@ xfs_discard_extents( blk_start_plug(&plug); list_for_each_entry(busyp, &extents->extent_list, list) { - trace_xfs_discard_extent(busyp->group, busyp->bno, - busyp->length); + struct xfs_group *xg = busyp->group; + struct xfs_buftarg *btp = + xfs_group_type_buftarg(xg->xg_mount, xg->xg_type); + + trace_xfs_discard_extent(xg, busyp->bno, busyp->length); - error = __blkdev_issue_discard(xfs_group_bdev(busyp->group), - xfs_gbno_to_daddr(busyp->group, busyp->bno), + error = __blkdev_issue_discard(btp->bt_bdev, + xfs_gbno_to_daddr(xg, busyp->bno), XFS_FSB_TO_BB(mp, busyp->length), GFP_KERNEL, &bio); if (error && error != -EOPNOTSUPP) { diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index b4e32f0860b7..0bd8022e47b4 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -1398,11 +1398,9 @@ xfs_qm_dqflush( ASSERT(XFS_DQ_IS_LOCKED(dqp)); ASSERT(!completion_done(&dqp->q_flush)); + ASSERT(atomic_read(&dqp->q_pincount) == 0); trace_xfs_dqflush(dqp); - - xfs_qm_dqunpin_wait(dqp); - fa = xfs_qm_dqflush_check(dqp); if (fa) { xfs_alert(mp, "corrupt dquot ID 0x%x in memory at %pS", diff --git a/fs/xfs/xfs_extent_busy.h b/fs/xfs/xfs_extent_busy.h index f069b04e8ea1..3e6e019b6146 100644 --- a/fs/xfs/xfs_extent_busy.h +++ b/fs/xfs/xfs_extent_busy.h @@ -68,4 +68,12 @@ static inline void xfs_extent_busy_sort(struct list_head *list) list_sort(NULL, list, xfs_extent_busy_ag_cmp); } +/* + * Zoned RTGs don't need to track busy extents, as the actual block freeing only + * happens by a zone reset, which forces out all transactions that touched the + * to be reset zone first. + */ +#define xfs_group_has_extent_busy(mp, type) \ + ((type) == XG_TYPE_AG || !xfs_has_zoned((mp))) + #endif /* __XFS_EXTENT_BUSY_H__ */ diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 48254a72071b..38e365b16348 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -752,7 +752,7 @@ xfs_file_dio_write_atomic( * HW offload should be faster, so try that first if it is already * known that the write length is not too large. */ - if (ocount > xfs_inode_buftarg(ip)->bt_bdev_awu_max) + if (ocount > xfs_inode_buftarg(ip)->bt_awu_max) dops = &xfs_atomic_write_cow_iomap_ops; else dops = &xfs_direct_write_iomap_ops; @@ -1335,9 +1335,10 @@ xfs_falloc_allocate_range( } #define XFS_FALLOC_FL_SUPPORTED \ - (FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | \ - FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE | \ - FALLOC_FL_INSERT_RANGE | FALLOC_FL_UNSHARE_RANGE) + (FALLOC_FL_ALLOCATE_RANGE | FALLOC_FL_KEEP_SIZE | \ + FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE | \ + FALLOC_FL_ZERO_RANGE | FALLOC_FL_INSERT_RANGE | \ + FALLOC_FL_UNSHARE_RANGE) STATIC long __xfs_file_fallocate( diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c index 726e29b837e6..bbc2f2973dcc 100644 --- a/fs/xfs/xfs_icache.c +++ b/fs/xfs/xfs_icache.c @@ -979,7 +979,15 @@ xfs_reclaim_inode( */ if (xlog_is_shutdown(ip->i_mount->m_log)) { xfs_iunpin_wait(ip); + /* + * Avoid a ABBA deadlock on the inode cluster buffer vs + * concurrent xfs_ifree_cluster() trying to mark the inode + * stale. We don't need the inode locked to run the flush abort + * code, but the flush abort needs to lock the cluster buffer. + */ + xfs_iunlock(ip, XFS_ILOCK_EXCL); xfs_iflush_shutdown_abort(ip); + xfs_ilock(ip, XFS_ILOCK_EXCL); goto reclaim; } if (xfs_ipincount(ip)) diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index ee3e0f284287..761a996a857c 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -1635,7 +1635,7 @@ retry: iip = ip->i_itemp; if (__xfs_iflags_test(ip, XFS_IFLUSHING)) { ASSERT(!list_empty(&iip->ili_item.li_bio_list)); - ASSERT(iip->ili_last_fields); + ASSERT(iip->ili_last_fields || xlog_is_shutdown(mp->m_log)); goto out_iunlock; } diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index d7e2b902ef5c..07fbdcc4cbf5 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -358,7 +358,7 @@ static inline bool xfs_inode_has_bigrtalloc(const struct xfs_inode *ip) static inline bool xfs_inode_can_hw_atomic_write(const struct xfs_inode *ip) { - return xfs_inode_buftarg(ip)->bt_bdev_awu_max > 0; + return xfs_inode_buftarg(ip)->bt_awu_max > 0; } /* diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index c6cb0b6b9e46..285e27ff89e2 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -758,11 +758,14 @@ xfs_inode_item_push( * completed and items removed from the AIL before the next push * attempt. */ + trace_xfs_inode_push_stale(ip, _RET_IP_); return XFS_ITEM_PINNED; } - if (xfs_ipincount(ip) > 0 || xfs_buf_ispinned(bp)) + if (xfs_ipincount(ip) > 0 || xfs_buf_ispinned(bp)) { + trace_xfs_inode_push_pinned(ip, _RET_IP_); return XFS_ITEM_PINNED; + } if (xfs_iflags_test(ip, XFS_IFLUSHING)) return XFS_ITEM_FLUSHING; diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index ff05e6b1b0bb..ec30b78bf5c4 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -827,7 +827,7 @@ xfs_bmap_hw_atomic_write_possible( /* * The ->iomap_begin caller should ensure this, but check anyway. */ - return len <= xfs_inode_buftarg(ip)->bt_bdev_awu_max; + return len <= xfs_inode_buftarg(ip)->bt_awu_max; } static int diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 8cddbb7c149b..01e597290eb5 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -665,7 +665,7 @@ xfs_get_atomic_write_max_opt( * less than our out of place write limit, but we don't want to exceed * the awu_max. */ - return min(awu_max, xfs_inode_buftarg(ip)->bt_bdev_awu_max); + return min(awu_max, xfs_inode_buftarg(ip)->bt_awu_max); } static void diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index f66d2d430e4f..a80cb6b9969a 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -793,8 +793,10 @@ xlog_cil_ail_insert( struct xfs_log_item *lip = lv->lv_item; xfs_lsn_t item_lsn; - if (aborted) + if (aborted) { + trace_xlog_ail_insert_abort(lip); set_bit(XFS_LI_ABORTED, &lip->li_flags); + } if (lip->li_ops->flags & XFS_ITEM_RELEASE_WHEN_COMMITTED) { lip->li_ops->iop_release(lip); diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 29276fe60df9..0b690bc119d7 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -171,19 +171,16 @@ xfs_readsb( ASSERT(mp->m_ddev_targp != NULL); /* - * For the initial read, we must guess at the sector - * size based on the block device. It's enough to - * get the sb_sectsize out of the superblock and - * then reread with the proper length. - * We don't verify it yet, because it may not be complete. + * In the first pass, use the device sector size to just read enough + * of the superblock to extract the XFS sector size. + * + * The device sector size must be smaller than or equal to the XFS + * sector size and thus we can always read the superblock. Once we know + * the XFS sector size, re-read it and run the buffer verifier. */ - sector_size = xfs_getsize_buftarg(mp->m_ddev_targp); + sector_size = mp->m_ddev_targp->bt_logical_sectorsize; buf_ops = NULL; - /* - * Allocate a (locked) buffer to hold the superblock. This will be kept - * around at all times to optimize access to the superblock. - */ reread: error = xfs_buf_read_uncached(mp->m_ddev_targp, XFS_SB_DADDR, BTOBB(sector_size), &bp, buf_ops); @@ -247,6 +244,10 @@ reread: /* no need to be quiet anymore, so reset the buf ops */ bp->b_ops = &xfs_sb_buf_ops; + /* + * Keep a pointer of the sb buffer around instead of caching it in the + * buffer cache because we access it frequently. + */ mp->m_sb_bp = bp; xfs_buf_unlock(bp); return 0; @@ -678,68 +679,46 @@ static inline unsigned int max_pow_of_two_factor(const unsigned int nr) } /* - * If the data device advertises atomic write support, limit the size of data - * device atomic writes to the greatest power-of-two factor of the AG size so - * that every atomic write unit aligns with the start of every AG. This is - * required so that the per-AG allocations for an atomic write will always be + * If the underlying device advertises atomic write support, limit the size of + * atomic writes to the greatest power-of-two factor of the group size so + * that every atomic write unit aligns with the start of every group. This is + * required so that the allocations for an atomic write will always be * aligned compatibly with the alignment requirements of the storage. * - * If the data device doesn't advertise atomic writes, then there are no - * alignment restrictions and the largest out-of-place write we can do - * ourselves is the number of blocks that user files can allocate from any AG. - */ -static inline xfs_extlen_t xfs_calc_perag_awu_max(struct xfs_mount *mp) -{ - if (mp->m_ddev_targp->bt_bdev_awu_min > 0) - return max_pow_of_two_factor(mp->m_sb.sb_agblocks); - return rounddown_pow_of_two(mp->m_ag_max_usable); -} - -/* - * Reflink on the realtime device requires rtgroups, and atomic writes require - * reflink. - * - * If the realtime device advertises atomic write support, limit the size of - * data device atomic writes to the greatest power-of-two factor of the rtgroup - * size so that every atomic write unit aligns with the start of every rtgroup. - * This is required so that the per-rtgroup allocations for an atomic write - * will always be aligned compatibly with the alignment requirements of the - * storage. - * - * If the rt device doesn't advertise atomic writes, then there are no - * alignment restrictions and the largest out-of-place write we can do - * ourselves is the number of blocks that user files can allocate from any - * rtgroup. + * If the device doesn't advertise atomic writes, then there are no alignment + * restrictions and the largest out-of-place write we can do ourselves is the + * number of blocks that user files can allocate from any group. */ -static inline xfs_extlen_t xfs_calc_rtgroup_awu_max(struct xfs_mount *mp) +static xfs_extlen_t +xfs_calc_group_awu_max( + struct xfs_mount *mp, + enum xfs_group_type type) { - struct xfs_groups *rgs = &mp->m_groups[XG_TYPE_RTG]; + struct xfs_groups *g = &mp->m_groups[type]; + struct xfs_buftarg *btp = xfs_group_type_buftarg(mp, type); - if (rgs->blocks == 0) + if (g->blocks == 0) return 0; - if (mp->m_rtdev_targp && mp->m_rtdev_targp->bt_bdev_awu_min > 0) - return max_pow_of_two_factor(rgs->blocks); - return rounddown_pow_of_two(rgs->blocks); + if (btp && btp->bt_awu_min > 0) + return max_pow_of_two_factor(g->blocks); + return rounddown_pow_of_two(g->blocks); } /* Compute the maximum atomic write unit size for each section. */ static inline void xfs_calc_atomic_write_unit_max( - struct xfs_mount *mp) + struct xfs_mount *mp, + enum xfs_group_type type) { - struct xfs_groups *ags = &mp->m_groups[XG_TYPE_AG]; - struct xfs_groups *rgs = &mp->m_groups[XG_TYPE_RTG]; + struct xfs_groups *g = &mp->m_groups[type]; const xfs_extlen_t max_write = xfs_calc_atomic_write_max(mp); const xfs_extlen_t max_ioend = xfs_reflink_max_atomic_cow(mp); - const xfs_extlen_t max_agsize = xfs_calc_perag_awu_max(mp); - const xfs_extlen_t max_rgsize = xfs_calc_rtgroup_awu_max(mp); - - ags->awu_max = min3(max_write, max_ioend, max_agsize); - rgs->awu_max = min3(max_write, max_ioend, max_rgsize); + const xfs_extlen_t max_gsize = xfs_calc_group_awu_max(mp, type); - trace_xfs_calc_atomic_write_unit_max(mp, max_write, max_ioend, - max_agsize, max_rgsize); + g->awu_max = min3(max_write, max_ioend, max_gsize); + trace_xfs_calc_atomic_write_unit_max(mp, type, max_write, max_ioend, + max_gsize, g->awu_max); } /* @@ -757,7 +736,8 @@ xfs_set_max_atomic_write_opt( max(mp->m_groups[XG_TYPE_AG].blocks, mp->m_groups[XG_TYPE_RTG].blocks); const xfs_extlen_t max_group_write = - max(xfs_calc_perag_awu_max(mp), xfs_calc_rtgroup_awu_max(mp)); + max(xfs_calc_group_awu_max(mp, XG_TYPE_AG), + xfs_calc_group_awu_max(mp, XG_TYPE_RTG)); int error; if (new_max_bytes == 0) @@ -813,7 +793,8 @@ set_limit: return error; } - xfs_calc_atomic_write_unit_max(mp); + xfs_calc_atomic_write_unit_max(mp, XG_TYPE_AG); + xfs_calc_atomic_write_unit_max(mp, XG_TYPE_RTG); mp->m_awu_max_bytes = new_max_bytes; return 0; } diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index d85084f9f317..97de44c32272 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -802,4 +802,21 @@ static inline void xfs_mod_sb_delalloc(struct xfs_mount *mp, int64_t delta) int xfs_set_max_atomic_write_opt(struct xfs_mount *mp, unsigned long long new_max_bytes); +static inline struct xfs_buftarg * +xfs_group_type_buftarg( + struct xfs_mount *mp, + enum xfs_group_type type) +{ + switch (type) { + case XG_TYPE_AG: + return mp->m_ddev_targp; + case XG_TYPE_RTG: + return mp->m_rtdev_targp; + default: + ASSERT(0); + break; + } + return NULL; +} + #endif /* __XFS_MOUNT_H__ */ diff --git a/fs/xfs/xfs_mru_cache.c b/fs/xfs/xfs_mru_cache.c index 08443ceec329..866c71d9fbae 100644 --- a/fs/xfs/xfs_mru_cache.c +++ b/fs/xfs/xfs_mru_cache.c @@ -320,7 +320,7 @@ xfs_mru_cache_create( xfs_mru_cache_free_func_t free_func) { struct xfs_mru_cache *mru = NULL; - int err = 0, grp; + int grp; unsigned int grp_time; if (mrup) @@ -341,8 +341,8 @@ xfs_mru_cache_create( mru->lists = kzalloc(mru->grp_count * sizeof(*mru->lists), GFP_KERNEL | __GFP_NOFAIL); if (!mru->lists) { - err = -ENOMEM; - goto exit; + kfree(mru); + return -ENOMEM; } for (grp = 0; grp < mru->grp_count; grp++) @@ -361,14 +361,7 @@ xfs_mru_cache_create( mru->free_func = free_func; mru->data = data; *mrup = mru; - -exit: - if (err && mru && mru->lists) - kfree(mru->lists); - if (err && mru) - kfree(mru); - - return err; + return 0; } /* @@ -425,10 +418,6 @@ xfs_mru_cache_insert( { int error = -EINVAL; - ASSERT(mru && mru->lists); - if (!mru || !mru->lists) - goto out_free; - error = -ENOMEM; if (radix_tree_preload(GFP_KERNEL)) goto out_free; diff --git a/fs/xfs/xfs_notify_failure.c b/fs/xfs/xfs_notify_failure.c index 3545dc1d953c..42e9c72b85c0 100644 --- a/fs/xfs/xfs_notify_failure.c +++ b/fs/xfs/xfs_notify_failure.c @@ -253,8 +253,7 @@ xfs_dax_notify_dev_failure( return -EOPNOTSUPP; } - error = xfs_dax_translate_range(type == XG_TYPE_RTG ? - mp->m_rtdev_targp : mp->m_ddev_targp, + error = xfs_dax_translate_range(xfs_group_type_buftarg(mp, type), offset, len, &daddr, &bblen); if (error) return error; diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 417439b58785..fa135ac26471 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -134,6 +134,7 @@ xfs_qm_dqpurge( dqp->q_flags |= XFS_DQFLAG_FREEING; + xfs_qm_dqunpin_wait(dqp); xfs_dqflock(dqp); /* @@ -465,6 +466,7 @@ xfs_qm_dquot_isolate( struct xfs_dquot *dqp = container_of(item, struct xfs_dquot, q_lru); struct xfs_qm_isolate *isol = arg; + enum lru_status ret = LRU_SKIP; if (!xfs_dqlock_nowait(dqp)) goto out_miss_busy; @@ -478,6 +480,16 @@ xfs_qm_dquot_isolate( goto out_miss_unlock; /* + * If the dquot is pinned or dirty, rotate it to the end of the LRU to + * give some time for it to be cleaned before we try to isolate it + * again. + */ + ret = LRU_ROTATE; + if (XFS_DQ_IS_DIRTY(dqp) || atomic_read(&dqp->q_pincount) > 0) { + goto out_miss_unlock; + } + + /* * This dquot has acquired a reference in the meantime remove it from * the freelist and try again. */ @@ -492,41 +504,14 @@ xfs_qm_dquot_isolate( } /* - * If the dquot is dirty, flush it. If it's already being flushed, just - * skip it so there is time for the IO to complete before we try to - * reclaim it again on the next LRU pass. + * The dquot may still be under IO, in which case the flush lock will be + * held. If we can't get the flush lock now, just skip over the dquot as + * if it was dirty. */ if (!xfs_dqflock_nowait(dqp)) goto out_miss_unlock; - if (XFS_DQ_IS_DIRTY(dqp)) { - struct xfs_buf *bp = NULL; - int error; - - trace_xfs_dqreclaim_dirty(dqp); - - /* we have to drop the LRU lock to flush the dquot */ - spin_unlock(&lru->lock); - - error = xfs_dquot_use_attached_buf(dqp, &bp); - if (!bp || error == -EAGAIN) { - xfs_dqfunlock(dqp); - goto out_unlock_dirty; - } - - /* - * dqflush completes dqflock on error, and the delwri ioend - * does it on success. - */ - error = xfs_qm_dqflush(dqp, bp); - if (error) - goto out_unlock_dirty; - - xfs_buf_delwri_queue(bp, &isol->buffers); - xfs_buf_relse(bp); - goto out_unlock_dirty; - } - + ASSERT(!XFS_DQ_IS_DIRTY(dqp)); xfs_dquot_detach_buf(dqp); xfs_dqfunlock(dqp); @@ -548,13 +533,7 @@ out_miss_unlock: out_miss_busy: trace_xfs_dqreclaim_busy(dqp); XFS_STATS_INC(dqp->q_mount, xs_qm_dqreclaim_misses); - return LRU_SKIP; - -out_unlock_dirty: - trace_xfs_dqreclaim_busy(dqp); - XFS_STATS_INC(dqp->q_mount, xs_qm_dqreclaim_misses); - xfs_dqunlock(dqp); - return LRU_RETRY; + return ret; } static unsigned long @@ -1486,7 +1465,6 @@ xfs_qm_flush_one( struct xfs_dquot *dqp, void *data) { - struct xfs_mount *mp = dqp->q_mount; struct list_head *buffer_list = data; struct xfs_buf *bp = NULL; int error = 0; @@ -1497,34 +1475,8 @@ xfs_qm_flush_one( if (!XFS_DQ_IS_DIRTY(dqp)) goto out_unlock; - /* - * The only way the dquot is already flush locked by the time quotacheck - * gets here is if reclaim flushed it before the dqadjust walk dirtied - * it for the final time. Quotacheck collects all dquot bufs in the - * local delwri queue before dquots are dirtied, so reclaim can't have - * possibly queued it for I/O. The only way out is to push the buffer to - * cycle the flush lock. - */ - if (!xfs_dqflock_nowait(dqp)) { - /* buf is pinned in-core by delwri list */ - error = xfs_buf_incore(mp->m_ddev_targp, dqp->q_blkno, - mp->m_quotainfo->qi_dqchunklen, 0, &bp); - if (error) - goto out_unlock; - - if (!(bp->b_flags & _XBF_DELWRI_Q)) { - error = -EAGAIN; - xfs_buf_relse(bp); - goto out_unlock; - } - xfs_buf_unlock(bp); - - xfs_buf_delwri_pushbuf(bp, buffer_list); - xfs_buf_rele(bp); - - error = -EAGAIN; - goto out_unlock; - } + xfs_qm_dqunpin_wait(dqp); + xfs_dqflock(dqp); error = xfs_dquot_use_attached_buf(dqp, &bp); if (error) diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index 6484c596ecea..736eb0924573 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -1259,6 +1259,8 @@ xfs_growfs_check_rtgeom( kfree(nmp); + trace_xfs_growfs_check_rtgeom(mp, min_logfsbs); + if (min_logfsbs > mp->m_sb.sb_logblocks) return -EINVAL; diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 0bc4b5489078..bb0a82635a77 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -2020,14 +2020,13 @@ xfs_remount_rw( int error; if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp && - bdev_read_only(mp->m_logdev_targp->bt_bdev)) { + xfs_readonly_buftarg(mp->m_logdev_targp)) { xfs_warn(mp, "ro->rw transition prohibited by read-only logdev"); return -EACCES; } - if (mp->m_rtdev_targp && - bdev_read_only(mp->m_rtdev_targp->bt_bdev)) { + if (mp->m_rtdev_targp && xfs_readonly_buftarg(mp->m_rtdev_targp)) { xfs_warn(mp, "ro->rw transition prohibited by read-only rtdev"); return -EACCES; diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 01d284a1c759..78be223b13b2 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -171,36 +171,33 @@ DEFINE_ATTR_LIST_EVENT(xfs_attr_leaf_list); DEFINE_ATTR_LIST_EVENT(xfs_attr_node_list); TRACE_EVENT(xfs_calc_atomic_write_unit_max, - TP_PROTO(struct xfs_mount *mp, unsigned int max_write, - unsigned int max_ioend, unsigned int max_agsize, - unsigned int max_rgsize), - TP_ARGS(mp, max_write, max_ioend, max_agsize, max_rgsize), + TP_PROTO(struct xfs_mount *mp, enum xfs_group_type type, + unsigned int max_write, unsigned int max_ioend, + unsigned int max_gsize, unsigned int awu_max), + TP_ARGS(mp, type, max_write, max_ioend, max_gsize, awu_max), TP_STRUCT__entry( __field(dev_t, dev) + __field(enum xfs_group_type, type) __field(unsigned int, max_write) __field(unsigned int, max_ioend) - __field(unsigned int, max_agsize) - __field(unsigned int, max_rgsize) - __field(unsigned int, data_awu_max) - __field(unsigned int, rt_awu_max) + __field(unsigned int, max_gsize) + __field(unsigned int, awu_max) ), TP_fast_assign( __entry->dev = mp->m_super->s_dev; + __entry->type = type; __entry->max_write = max_write; __entry->max_ioend = max_ioend; - __entry->max_agsize = max_agsize; - __entry->max_rgsize = max_rgsize; - __entry->data_awu_max = mp->m_groups[XG_TYPE_AG].awu_max; - __entry->rt_awu_max = mp->m_groups[XG_TYPE_RTG].awu_max; + __entry->max_gsize = max_gsize; + __entry->awu_max = awu_max; ), - TP_printk("dev %d:%d max_write %u max_ioend %u max_agsize %u max_rgsize %u data_awu_max %u rt_awu_max %u", + TP_printk("dev %d:%d %s max_write %u max_ioend %u max_gsize %u awu_max %u", MAJOR(__entry->dev), MINOR(__entry->dev), + __print_symbolic(__entry->type, XG_TYPE_STRINGS), __entry->max_write, __entry->max_ioend, - __entry->max_agsize, - __entry->max_rgsize, - __entry->data_awu_max, - __entry->rt_awu_max) + __entry->max_gsize, + __entry->awu_max) ); TRACE_EVENT(xfs_calc_max_atomic_write_fsblocks, @@ -778,7 +775,6 @@ DEFINE_BUF_EVENT(xfs_buf_iowait_done); DEFINE_BUF_EVENT(xfs_buf_delwri_queue); DEFINE_BUF_EVENT(xfs_buf_delwri_queued); DEFINE_BUF_EVENT(xfs_buf_delwri_split); -DEFINE_BUF_EVENT(xfs_buf_delwri_pushbuf); DEFINE_BUF_EVENT(xfs_buf_get_uncached); DEFINE_BUF_EVENT(xfs_buf_item_relse); DEFINE_BUF_EVENT(xfs_buf_iodone_async); @@ -1147,6 +1143,7 @@ DECLARE_EVENT_CLASS(xfs_iref_class, __field(xfs_ino_t, ino) __field(int, count) __field(int, pincount) + __field(unsigned long, iflags) __field(unsigned long, caller_ip) ), TP_fast_assign( @@ -1154,13 +1151,15 @@ DECLARE_EVENT_CLASS(xfs_iref_class, __entry->ino = ip->i_ino; __entry->count = atomic_read(&VFS_I(ip)->i_count); __entry->pincount = atomic_read(&ip->i_pincount); + __entry->iflags = ip->i_flags; __entry->caller_ip = caller_ip; ), - TP_printk("dev %d:%d ino 0x%llx count %d pincount %d caller %pS", + TP_printk("dev %d:%d ino 0x%llx count %d pincount %d iflags 0x%lx caller %pS", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->count, __entry->pincount, + __entry->iflags, (char *)__entry->caller_ip) ) @@ -1250,6 +1249,8 @@ DEFINE_IREF_EVENT(xfs_irele); DEFINE_IREF_EVENT(xfs_inode_pin); DEFINE_IREF_EVENT(xfs_inode_unpin); DEFINE_IREF_EVENT(xfs_inode_unpin_nowait); +DEFINE_IREF_EVENT(xfs_inode_push_pinned); +DEFINE_IREF_EVENT(xfs_inode_push_stale); DECLARE_EVENT_CLASS(xfs_namespace_class, TP_PROTO(struct xfs_inode *dp, const struct xfs_name *name), @@ -1654,6 +1655,8 @@ DEFINE_LOG_ITEM_EVENT(xfs_ail_flushing); DEFINE_LOG_ITEM_EVENT(xfs_cil_whiteout_mark); DEFINE_LOG_ITEM_EVENT(xfs_cil_whiteout_skip); DEFINE_LOG_ITEM_EVENT(xfs_cil_whiteout_unpin); +DEFINE_LOG_ITEM_EVENT(xlog_ail_insert_abort); +DEFINE_LOG_ITEM_EVENT(xfs_trans_free_abort); DECLARE_EVENT_CLASS(xfs_ail_class, TP_PROTO(struct xfs_log_item *lip, xfs_lsn_t old_lsn, xfs_lsn_t new_lsn), diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index c6657072361a..b4a07af513ba 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -742,8 +742,10 @@ xfs_trans_free_items( list_for_each_entry_safe(lip, next, &tp->t_items, li_trans) { xfs_trans_del_item(lip); - if (abort) + if (abort) { + trace_xfs_trans_free_abort(lip); set_bit(XFS_LI_ABORTED, &lip->li_flags); + } if (lip->li_ops->iop_release) lip->li_ops->iop_release(lip); } diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c index 0f641a9091ec..ac5cecec9aa1 100644 --- a/fs/xfs/xfs_xattr.c +++ b/fs/xfs/xfs_xattr.c @@ -243,7 +243,7 @@ __xfs_xattr_put_listent( offset = context->buffer + context->count; memcpy(offset, prefix, prefix_len); offset += prefix_len; - strncpy(offset, (char *)name, namelen); /* real name */ + memcpy(offset, (char *)name, namelen); /* real name */ offset += namelen; *offset = '\0'; diff --git a/fs/xfs/xfs_zone_alloc.c b/fs/xfs/xfs_zone_alloc.c index 80add26c0111..01315ed75502 100644 --- a/fs/xfs/xfs_zone_alloc.c +++ b/fs/xfs/xfs_zone_alloc.c @@ -727,7 +727,7 @@ xfs_select_zone( for (;;) { prepare_to_wait(&zi->zi_zone_wait, &wait, TASK_UNINTERRUPTIBLE); oz = xfs_select_zone_nowait(mp, write_hint, pack_tight); - if (oz) + if (oz || xfs_is_shutdown(mp)) break; schedule(); } @@ -777,26 +777,6 @@ xfs_mark_rtg_boundary( ioend->io_flags |= IOMAP_IOEND_BOUNDARY; } -static void -xfs_submit_zoned_bio( - struct iomap_ioend *ioend, - struct xfs_open_zone *oz, - bool is_seq) -{ - ioend->io_bio.bi_iter.bi_sector = ioend->io_sector; - ioend->io_private = oz; - atomic_inc(&oz->oz_ref); /* for xfs_zoned_end_io */ - - if (is_seq) { - ioend->io_bio.bi_opf &= ~REQ_OP_WRITE; - ioend->io_bio.bi_opf |= REQ_OP_ZONE_APPEND; - } else { - xfs_mark_rtg_boundary(ioend); - } - - submit_bio(&ioend->io_bio); -} - /* * Cache the last zone written to for an inode so that it is considered first * for subsequent writes. @@ -891,6 +871,26 @@ xfs_zone_cache_create_association( xfs_mru_cache_insert(mp->m_zone_cache, ip->i_ino, &item->mru); } +static void +xfs_submit_zoned_bio( + struct iomap_ioend *ioend, + struct xfs_open_zone *oz, + bool is_seq) +{ + ioend->io_bio.bi_iter.bi_sector = ioend->io_sector; + ioend->io_private = oz; + atomic_inc(&oz->oz_ref); /* for xfs_zoned_end_io */ + + if (is_seq) { + ioend->io_bio.bi_opf &= ~REQ_OP_WRITE; + ioend->io_bio.bi_opf |= REQ_OP_ZONE_APPEND; + } else { + xfs_mark_rtg_boundary(ioend); + } + + submit_bio(&ioend->io_bio); +} + void xfs_zone_alloc_and_submit( struct iomap_ioend *ioend, |