diff options
Diffstat (limited to 'fs')
38 files changed, 957 insertions, 437 deletions
@@ -192,7 +192,6 @@ static struct file *aio_private_file(struct kioctx *ctx, loff_t nr_pages) } file->f_flags = O_RDWR; - file->private_data = ctx; return file; } @@ -202,7 +201,7 @@ static struct dentry *aio_mount(struct file_system_type *fs_type, static const struct dentry_operations ops = { .d_dname = simple_dname, }; - return mount_pseudo(fs_type, "aio:", NULL, &ops, 0xa10a10a1); + return mount_pseudo(fs_type, "aio:", NULL, &ops, AIO_RING_MAGIC); } /* aio_setup @@ -556,8 +555,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) struct aio_ring *ring; spin_lock(&mm->ioctx_lock); - rcu_read_lock(); - table = rcu_dereference(mm->ioctx_table); + table = rcu_dereference_raw(mm->ioctx_table); while (1) { if (table) @@ -565,7 +563,6 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) if (!table->table[i]) { ctx->id = i; table->table[i] = ctx; - rcu_read_unlock(); spin_unlock(&mm->ioctx_lock); /* While kioctx setup is in progress, @@ -579,8 +576,6 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) } new_nr = (table ? table->nr : 1) * 4; - - rcu_read_unlock(); spin_unlock(&mm->ioctx_lock); table = kzalloc(sizeof(*table) + sizeof(struct kioctx *) * @@ -591,8 +586,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) table->nr = new_nr; spin_lock(&mm->ioctx_lock); - rcu_read_lock(); - old = rcu_dereference(mm->ioctx_table); + old = rcu_dereference_raw(mm->ioctx_table); if (!old) { rcu_assign_pointer(mm->ioctx_table, table); @@ -739,12 +733,9 @@ static int kill_ioctx(struct mm_struct *mm, struct kioctx *ctx, spin_lock(&mm->ioctx_lock); - rcu_read_lock(); - table = rcu_dereference(mm->ioctx_table); - + table = rcu_dereference_raw(mm->ioctx_table); WARN_ON(ctx != table->table[ctx->id]); table->table[ctx->id] = NULL; - rcu_read_unlock(); spin_unlock(&mm->ioctx_lock); /* percpu_ref_kill() will do the necessary call_rcu() */ @@ -793,40 +784,30 @@ EXPORT_SYMBOL(wait_on_sync_kiocb); */ void exit_aio(struct mm_struct *mm) { - struct kioctx_table *table; - struct kioctx *ctx; - unsigned i = 0; - - while (1) { - rcu_read_lock(); - table = rcu_dereference(mm->ioctx_table); - - do { - if (!table || i >= table->nr) { - rcu_read_unlock(); - rcu_assign_pointer(mm->ioctx_table, NULL); - if (table) - kfree(table); - return; - } + struct kioctx_table *table = rcu_dereference_raw(mm->ioctx_table); + int i; - ctx = table->table[i++]; - } while (!ctx); + if (!table) + return; - rcu_read_unlock(); + for (i = 0; i < table->nr; ++i) { + struct kioctx *ctx = table->table[i]; + if (!ctx) + continue; /* - * We don't need to bother with munmap() here - - * exit_mmap(mm) is coming and it'll unmap everything. - * Since aio_free_ring() uses non-zero ->mmap_size - * as indicator that it needs to unmap the area, - * just set it to 0; aio_free_ring() is the only - * place that uses ->mmap_size, so it's safe. + * We don't need to bother with munmap() here - exit_mmap(mm) + * is coming and it'll unmap everything. And we simply can't, + * this is not necessarily our ->mm. + * Since kill_ioctx() uses non-zero ->mmap_size as indicator + * that it needs to unmap the area, just set it to 0. */ ctx->mmap_size = 0; - kill_ioctx(mm, ctx, NULL); } + + RCU_INIT_POINTER(mm->ioctx_table, NULL); + kfree(table); } static void put_reqs_available(struct kioctx *ctx, unsigned nr) @@ -834,10 +815,8 @@ static void put_reqs_available(struct kioctx *ctx, unsigned nr) struct kioctx_cpu *kcpu; unsigned long flags; - preempt_disable(); - kcpu = this_cpu_ptr(ctx->cpu); - local_irq_save(flags); + kcpu = this_cpu_ptr(ctx->cpu); kcpu->reqs_available += nr; while (kcpu->reqs_available >= ctx->req_batch * 2) { @@ -846,7 +825,6 @@ static void put_reqs_available(struct kioctx *ctx, unsigned nr) } local_irq_restore(flags); - preempt_enable(); } static bool get_reqs_available(struct kioctx *ctx) @@ -855,10 +833,8 @@ static bool get_reqs_available(struct kioctx *ctx) bool ret = false; unsigned long flags; - preempt_disable(); - kcpu = this_cpu_ptr(ctx->cpu); - local_irq_save(flags); + kcpu = this_cpu_ptr(ctx->cpu); if (!kcpu->reqs_available) { int old, avail = atomic_read(&ctx->reqs_available); @@ -878,7 +854,6 @@ static bool get_reqs_available(struct kioctx *ctx) kcpu->reqs_available--; out: local_irq_restore(flags); - preempt_enable(); return ret; } @@ -1047,7 +1022,7 @@ void aio_complete(struct kiocb *iocb, long res, long res2) } EXPORT_SYMBOL(aio_complete); -/* aio_read_events +/* aio_read_events_ring * Pull an event off of the ioctx's event ring. Returns the number of * events fetched */ @@ -1270,12 +1245,12 @@ static ssize_t aio_setup_vectored_rw(struct kiocb *kiocb, if (compat) ret = compat_rw_copy_check_uvector(rw, (struct compat_iovec __user *)buf, - *nr_segs, 1, *iovec, iovec); + *nr_segs, UIO_FASTIOV, *iovec, iovec); else #endif ret = rw_copy_check_uvector(rw, (struct iovec __user *)buf, - *nr_segs, 1, *iovec, iovec); + *nr_segs, UIO_FASTIOV, *iovec, iovec); if (ret < 0) return ret; @@ -1299,9 +1274,8 @@ static ssize_t aio_setup_single_vector(struct kiocb *kiocb, } /* - * aio_setup_iocb: - * Performs the initial checks and aio retry method - * setup for the kiocb at the time of io submission. + * aio_run_iocb: + * Performs the initial checks and io submission. */ static ssize_t aio_run_iocb(struct kiocb *req, unsigned opcode, char __user *buf, bool compat) @@ -1313,7 +1287,7 @@ static ssize_t aio_run_iocb(struct kiocb *req, unsigned opcode, fmode_t mode; aio_rw_op *rw_op; rw_iter_op *iter_op; - struct iovec inline_vec, *iovec = &inline_vec; + struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; struct iov_iter iter; switch (opcode) { @@ -1348,7 +1322,7 @@ rw_common: if (!ret) ret = rw_verify_area(rw, file, &req->ki_pos, req->ki_nbytes); if (ret < 0) { - if (iovec != &inline_vec) + if (iovec != inline_vecs) kfree(iovec); return ret; } @@ -1395,7 +1369,7 @@ rw_common: return -EINVAL; } - if (iovec != &inline_vec) + if (iovec != inline_vecs) kfree(iovec); if (ret != -EIOCBQUEUED) { diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index e25564bfcb46..54a201dac7f9 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -276,9 +276,8 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path, } if (ret > 0) goto next; - ret = ulist_add_merge(parents, eb->start, - (uintptr_t)eie, - (u64 *)&old, GFP_NOFS); + ret = ulist_add_merge_ptr(parents, eb->start, + eie, (void **)&old, GFP_NOFS); if (ret < 0) break; if (!ret && extent_item_pos) { @@ -1001,16 +1000,19 @@ again: ret = -EIO; goto out; } + btrfs_tree_read_lock(eb); + btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK); ret = find_extent_in_eb(eb, bytenr, *extent_item_pos, &eie); + btrfs_tree_read_unlock_blocking(eb); free_extent_buffer(eb); if (ret < 0) goto out; ref->inode_list = eie; } - ret = ulist_add_merge(refs, ref->parent, - (uintptr_t)ref->inode_list, - (u64 *)&eie, GFP_NOFS); + ret = ulist_add_merge_ptr(refs, ref->parent, + ref->inode_list, + (void **)&eie, GFP_NOFS); if (ret < 0) goto out; if (!ret && extent_item_pos) { diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index 4794923c410c..43527fd78825 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -84,12 +84,6 @@ struct btrfs_inode { */ struct list_head delalloc_inodes; - /* - * list for tracking inodes that must be sent to disk before a - * rename or truncate commit - */ - struct list_head ordered_operations; - /* node for the red-black tree that links inodes in subvolume root */ struct rb_node rb_node; diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index aeab453b8e24..44ee5d2e52a4 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -280,9 +280,9 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans, WARN_ON(btrfs_header_generation(buf) > trans->transid); if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID) - ret = btrfs_inc_ref(trans, root, cow, 1, 1); + ret = btrfs_inc_ref(trans, root, cow, 1); else - ret = btrfs_inc_ref(trans, root, cow, 0, 1); + ret = btrfs_inc_ref(trans, root, cow, 0); if (ret) return ret; @@ -1035,14 +1035,14 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, if ((owner == root->root_key.objectid || root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) && !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)) { - ret = btrfs_inc_ref(trans, root, buf, 1, 1); + ret = btrfs_inc_ref(trans, root, buf, 1); BUG_ON(ret); /* -ENOMEM */ if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) { - ret = btrfs_dec_ref(trans, root, buf, 0, 1); + ret = btrfs_dec_ref(trans, root, buf, 0); BUG_ON(ret); /* -ENOMEM */ - ret = btrfs_inc_ref(trans, root, cow, 1, 1); + ret = btrfs_inc_ref(trans, root, cow, 1); BUG_ON(ret); /* -ENOMEM */ } new_flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF; @@ -1050,9 +1050,9 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) - ret = btrfs_inc_ref(trans, root, cow, 1, 1); + ret = btrfs_inc_ref(trans, root, cow, 1); else - ret = btrfs_inc_ref(trans, root, cow, 0, 1); + ret = btrfs_inc_ref(trans, root, cow, 0); BUG_ON(ret); /* -ENOMEM */ } if (new_flags != 0) { @@ -1069,11 +1069,11 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) { if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) - ret = btrfs_inc_ref(trans, root, cow, 1, 1); + ret = btrfs_inc_ref(trans, root, cow, 1); else - ret = btrfs_inc_ref(trans, root, cow, 0, 1); + ret = btrfs_inc_ref(trans, root, cow, 0); BUG_ON(ret); /* -ENOMEM */ - ret = btrfs_dec_ref(trans, root, buf, 1, 1); + ret = btrfs_dec_ref(trans, root, buf, 1); BUG_ON(ret); /* -ENOMEM */ } clean_tree_block(trans, root, buf); diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index be91397f4e92..8e29b614fe93 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3326,9 +3326,9 @@ int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes, u64 min_alloc_size, u64 empty_size, u64 hint_byte, struct btrfs_key *ins, int is_data, int delalloc); int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref, int no_quota); + struct extent_buffer *buf, int full_backref); int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref, int no_quota); + struct extent_buffer *buf, int full_backref); int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, u64 flags, diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 08e65e9cf2aa..d0ed9e664f7d 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -60,8 +60,6 @@ static void end_workqueue_fn(struct btrfs_work *work); static void free_fs_root(struct btrfs_root *root); static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, int read_only); -static void btrfs_destroy_ordered_operations(struct btrfs_transaction *t, - struct btrfs_root *root); static void btrfs_destroy_ordered_extents(struct btrfs_root *root); static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, struct btrfs_root *root); @@ -3829,34 +3827,6 @@ static void btrfs_error_commit_super(struct btrfs_root *root) btrfs_cleanup_transaction(root); } -static void btrfs_destroy_ordered_operations(struct btrfs_transaction *t, - struct btrfs_root *root) -{ - struct btrfs_inode *btrfs_inode; - struct list_head splice; - - INIT_LIST_HEAD(&splice); - - mutex_lock(&root->fs_info->ordered_operations_mutex); - spin_lock(&root->fs_info->ordered_root_lock); - - list_splice_init(&t->ordered_operations, &splice); - while (!list_empty(&splice)) { - btrfs_inode = list_entry(splice.next, struct btrfs_inode, - ordered_operations); - - list_del_init(&btrfs_inode->ordered_operations); - spin_unlock(&root->fs_info->ordered_root_lock); - - btrfs_invalidate_inodes(btrfs_inode->root); - - spin_lock(&root->fs_info->ordered_root_lock); - } - - spin_unlock(&root->fs_info->ordered_root_lock); - mutex_unlock(&root->fs_info->ordered_operations_mutex); -} - static void btrfs_destroy_ordered_extents(struct btrfs_root *root) { struct btrfs_ordered_extent *ordered; @@ -4093,8 +4063,6 @@ again: void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans, struct btrfs_root *root) { - btrfs_destroy_ordered_operations(cur_trans, root); - btrfs_destroy_delayed_refs(cur_trans, root); cur_trans->state = TRANS_STATE_COMMIT_START; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 813537f362f9..102ed3143976 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3057,7 +3057,7 @@ out: static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, - int full_backref, int inc, int no_quota) + int full_backref, int inc) { u64 bytenr; u64 num_bytes; @@ -3111,7 +3111,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, key.offset -= btrfs_file_extent_offset(buf, fi); ret = process_func(trans, root, bytenr, num_bytes, parent, ref_root, key.objectid, - key.offset, no_quota); + key.offset, 1); if (ret) goto fail; } else { @@ -3119,7 +3119,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, num_bytes = btrfs_level_size(root, level - 1); ret = process_func(trans, root, bytenr, num_bytes, parent, ref_root, level - 1, 0, - no_quota); + 1); if (ret) goto fail; } @@ -3130,15 +3130,15 @@ fail: } int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref, int no_quota) + struct extent_buffer *buf, int full_backref) { - return __btrfs_mod_ref(trans, root, buf, full_backref, 1, no_quota); + return __btrfs_mod_ref(trans, root, buf, full_backref, 1); } int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref, int no_quota) + struct extent_buffer *buf, int full_backref) { - return __btrfs_mod_ref(trans, root, buf, full_backref, 0, no_quota); + return __btrfs_mod_ref(trans, root, buf, full_backref, 0); } static int write_one_cache_group(struct btrfs_trans_handle *trans, @@ -7478,6 +7478,220 @@ reada: wc->reada_slot = slot; } +static int account_leaf_items(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct extent_buffer *eb) +{ + int nr = btrfs_header_nritems(eb); + int i, extent_type, ret; + struct btrfs_key key; + struct btrfs_file_extent_item *fi; + u64 bytenr, num_bytes; + + for (i = 0; i < nr; i++) { + btrfs_item_key_to_cpu(eb, &key, i); + + if (key.type != BTRFS_EXTENT_DATA_KEY) + continue; + + fi = btrfs_item_ptr(eb, i, struct btrfs_file_extent_item); + /* filter out non qgroup-accountable extents */ + extent_type = btrfs_file_extent_type(eb, fi); + + if (extent_type == BTRFS_FILE_EXTENT_INLINE) + continue; + + bytenr = btrfs_file_extent_disk_bytenr(eb, fi); + if (!bytenr) + continue; + + num_bytes = btrfs_file_extent_disk_num_bytes(eb, fi); + + ret = btrfs_qgroup_record_ref(trans, root->fs_info, + root->objectid, + bytenr, num_bytes, + BTRFS_QGROUP_OPER_SUB_SUBTREE, 0); + if (ret) + return ret; + } + return 0; +} + +/* + * Walk up the tree from the bottom, freeing leaves and any interior + * nodes which have had all slots visited. If a node (leaf or + * interior) is freed, the node above it will have it's slot + * incremented. The root node will never be freed. + * + * At the end of this function, we should have a path which has all + * slots incremented to the next position for a search. If we need to + * read a new node it will be NULL and the node above it will have the + * correct slot selected for a later read. + * + * If we increment the root nodes slot counter past the number of + * elements, 1 is returned to signal completion of the search. + */ +static int adjust_slots_upwards(struct btrfs_root *root, + struct btrfs_path *path, int root_level) +{ + int level = 0; + int nr, slot; + struct extent_buffer *eb; + + if (root_level == 0) + return 1; + + while (level <= root_level) { + eb = path->nodes[level]; + nr = btrfs_header_nritems(eb); + path->slots[level]++; + slot = path->slots[level]; + if (slot >= nr || level == 0) { + /* + * Don't free the root - we will detect this + * condition after our loop and return a + * positive value for caller to stop walking the tree. + */ + if (level != root_level) { + btrfs_tree_unlock_rw(eb, path->locks[level]); + path->locks[level] = 0; + + free_extent_buffer(eb); + path->nodes[level] = NULL; + path->slots[level] = 0; + } + } else { + /* + * We have a valid slot to walk back down + * from. Stop here so caller can process these + * new nodes. + */ + break; + } + + level++; + } + + eb = path->nodes[root_level]; + if (path->slots[root_level] >= btrfs_header_nritems(eb)) + return 1; + + return 0; +} + +/* + * root_eb is the subtree root and is locked before this function is called. + */ +static int account_shared_subtree(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct extent_buffer *root_eb, + u64 root_gen, + int root_level) +{ + int ret = 0; + int level; + struct extent_buffer *eb = root_eb; + struct btrfs_path *path = NULL; + + BUG_ON(root_level < 0 || root_level > BTRFS_MAX_LEVEL); + BUG_ON(root_eb == NULL); + + if (!root->fs_info->quota_enabled) + return 0; + + if (!extent_buffer_uptodate(root_eb)) { + ret = btrfs_read_buffer(root_eb, root_gen); + if (ret) + goto out; + } + + if (root_level == 0) { + ret = account_leaf_items(trans, root, root_eb); + goto out; + } + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + /* + * Walk down the tree. Missing extent blocks are filled in as + * we go. Metadata is accounted every time we read a new + * extent block. + * + * When we reach a leaf, we account for file extent items in it, + * walk back up the tree (adjusting slot pointers as we go) + * and restart the search process. + */ + extent_buffer_get(root_eb); /* For path */ + path->nodes[root_level] = root_eb; + path->slots[root_level] = 0; + path->locks[root_level] = 0; /* so release_path doesn't try to unlock */ +walk_down: + level = root_level; + while (level >= 0) { + if (path->nodes[level] == NULL) { + int child_bsize = root->nodesize; + int parent_slot; + u64 child_gen; + u64 child_bytenr; + + /* We need to get child blockptr/gen from + * parent before we can read it. */ + eb = path->nodes[level + 1]; + parent_slot = path->slots[level + 1]; + child_bytenr = btrfs_node_blockptr(eb, parent_slot); + child_gen = btrfs_node_ptr_generation(eb, parent_slot); + + eb = read_tree_block(root, child_bytenr, child_bsize, + child_gen); + if (!eb || !extent_buffer_uptodate(eb)) { + ret = -EIO; + goto out; + } + + path->nodes[level] = eb; + path->slots[level] = 0; + + btrfs_tree_read_lock(eb); + btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK); + path->locks[level] = BTRFS_READ_LOCK_BLOCKING; + + ret = btrfs_qgroup_record_ref(trans, root->fs_info, + root->objectid, + child_bytenr, + child_bsize, + BTRFS_QGROUP_OPER_SUB_SUBTREE, + 0); + if (ret) + goto out; + + } + + if (level == 0) { + ret = account_leaf_items(trans, root, path->nodes[level]); + if (ret) + goto out; + + /* Nonzero return here means we completed our search */ + ret = adjust_slots_upwards(root, path, root_level); + if (ret) + break; + + /* Restart search with new slots */ + goto walk_down; + } + + level--; + } + + ret = 0; +out: + btrfs_free_path(path); + + return ret; +} + /* * helper to process tree block while walking down the tree. * @@ -7532,9 +7746,9 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans, /* wc->stage == UPDATE_BACKREF */ if (!(wc->flags[level] & flag)) { BUG_ON(!path->locks[level]); - ret = btrfs_inc_ref(trans, root, eb, 1, wc->for_reloc); + ret = btrfs_inc_ref(trans, root, eb, 1); BUG_ON(ret); /* -ENOMEM */ - ret = btrfs_dec_ref(trans, root, eb, 0, wc->for_reloc); + ret = btrfs_dec_ref(trans, root, eb, 0); BUG_ON(ret); /* -ENOMEM */ ret = btrfs_set_disk_extent_flags(trans, root, eb->start, eb->len, flag, @@ -7581,6 +7795,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans, int level = wc->level; int reada = 0; int ret = 0; + bool need_account = false; generation = btrfs_node_ptr_generation(path->nodes[level], path->slots[level]); @@ -7626,6 +7841,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans, if (wc->stage == DROP_REFERENCE) { if (wc->refs[level - 1] > 1) { + need_account = true; if (level == 1 && (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF)) goto skip; @@ -7689,6 +7905,16 @@ skip: parent = 0; } + if (need_account) { + ret = account_shared_subtree(trans, root, next, + generation, level - 1); + if (ret) { + printk_ratelimited(KERN_ERR "BTRFS: %s Error " + "%d accounting shared subtree. Quota " + "is out of sync, rescan required.\n", + root->fs_info->sb->s_id, ret); + } + } ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent, root->root_key.objectid, level - 1, 0, 0); BUG_ON(ret); /* -ENOMEM */ @@ -7769,12 +7995,17 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans, if (wc->refs[level] == 1) { if (level == 0) { if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) - ret = btrfs_dec_ref(trans, root, eb, 1, - wc->for_reloc); + ret = btrfs_dec_ref(trans, root, eb, 1); else - ret = btrfs_dec_ref(trans, root, eb, 0, - wc->for_reloc); + ret = btrfs_dec_ref(trans, root, eb, 0); BUG_ON(ret); /* -ENOMEM */ + ret = account_leaf_items(trans, root, eb); + if (ret) { + printk_ratelimited(KERN_ERR "BTRFS: %s Error " + "%d accounting leaf items. Quota " + "is out of sync, rescan required.\n", + root->fs_info->sb->s_id, ret); + } } /* make block locked assertion in clean_tree_block happy */ if (!path->locks[level] && @@ -7900,6 +8131,8 @@ int btrfs_drop_snapshot(struct btrfs_root *root, int level; bool root_dropped = false; + btrfs_debug(root->fs_info, "Drop subvolume %llu", root->objectid); + path = btrfs_alloc_path(); if (!path) { err = -ENOMEM; @@ -8025,6 +8258,24 @@ int btrfs_drop_snapshot(struct btrfs_root *root, goto out_end_trans; } + /* + * Qgroup update accounting is run from + * delayed ref handling. This usually works + * out because delayed refs are normally the + * only way qgroup updates are added. However, + * we may have added updates during our tree + * walk so run qgroups here to make sure we + * don't lose any updates. + */ + ret = btrfs_delayed_qgroup_accounting(trans, + root->fs_info); + if (ret) + printk_ratelimited(KERN_ERR "BTRFS: Failure %d " + "running qgroup updates " + "during snapshot delete. " + "Quota is out of sync, " + "rescan required.\n", ret); + btrfs_end_transaction_throttle(trans, tree_root); if (!for_reloc && btrfs_need_cleaner_sleep(root)) { pr_debug("BTRFS: drop snapshot early exit\n"); @@ -8078,6 +8329,14 @@ int btrfs_drop_snapshot(struct btrfs_root *root, } root_dropped = true; out_end_trans: + ret = btrfs_delayed_qgroup_accounting(trans, tree_root->fs_info); + if (ret) + printk_ratelimited(KERN_ERR "BTRFS: Failure %d " + "running qgroup updates " + "during snapshot delete. " + "Quota is out of sync, " + "rescan required.\n", ret); + btrfs_end_transaction_throttle(trans, tree_root); out_free: kfree(wc); diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index f46cfe45d686..54c84daec9b5 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -756,7 +756,7 @@ again: found_next = 1; if (ret != 0) goto insert; - slot = 0; + slot = path->slots[0]; } btrfs_item_key_to_cpu(path->nodes[0], &found_key, slot); if (found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID || diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 1f2b99cb55ea..d3afac292d67 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1838,33 +1838,9 @@ out: int btrfs_release_file(struct inode *inode, struct file *filp) { - /* - * ordered_data_close is set by settattr when we are about to truncate - * a file from a non-zero size to a zero size. This tries to - * flush down new bytes that may have been written if the - * application were using truncate to replace a file in place. - */ - if (test_and_clear_bit(BTRFS_INODE_ORDERED_DATA_CLOSE, - &BTRFS_I(inode)->runtime_flags)) { - struct btrfs_trans_handle *trans; - struct btrfs_root *root = BTRFS_I(inode)->root; - - /* - * We need to block on a committing transaction to keep us from - * throwing a ordered operation on to the list and causing - * something like sync to deadlock trying to flush out this - * inode. - */ - trans = btrfs_start_transaction(root, 0); - if (IS_ERR(trans)) - return PTR_ERR(trans); - btrfs_add_ordered_operation(trans, BTRFS_I(inode)->root, inode); - btrfs_end_transaction(trans, root); - if (inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT) - filemap_flush(inode->i_mapping); - } if (filp->private_data) btrfs_ioctl_trans_end(filp); + filemap_flush(inode->i_mapping); return 0; } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 3183742d6f0d..03708ef3deef 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -709,6 +709,18 @@ retry: unlock_extent(io_tree, async_extent->start, async_extent->start + async_extent->ram_size - 1); + + /* + * we need to redirty the pages if we decide to + * fallback to uncompressed IO, otherwise we + * will not submit these pages down to lower + * layers. + */ + extent_range_redirty_for_io(inode, + async_extent->start, + async_extent->start + + async_extent->ram_size - 1); + goto retry; } goto out_free; @@ -7939,27 +7951,6 @@ static int btrfs_truncate(struct inode *inode) BUG_ON(ret); /* - * setattr is responsible for setting the ordered_data_close flag, - * but that is only tested during the last file release. That - * could happen well after the next commit, leaving a great big - * window where new writes may get lost if someone chooses to write - * to this file after truncating to zero - * - * The inode doesn't have any dirty data here, and so if we commit - * this is a noop. If someone immediately starts writing to the inode - * it is very likely we'll catch some of their writes in this - * transaction, and the commit will find this file on the ordered - * data list with good things to send down. - * - * This is a best effort solution, there is still a window where - * using truncate to replace the contents of the file will - * end up with a zero length file after a crash. - */ - if (inode->i_size == 0 && test_bit(BTRFS_INODE_ORDERED_DATA_CLOSE, - &BTRFS_I(inode)->runtime_flags)) - btrfs_add_ordered_operation(trans, root, inode); - - /* * So if we truncate and then write and fsync we normally would just * write the extents that changed, which is a problem if we need to * first truncate that entire inode. So set this flag so we write out @@ -8106,7 +8097,6 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) mutex_init(&ei->delalloc_mutex); btrfs_ordered_inode_tree_init(&ei->ordered_tree); INIT_LIST_HEAD(&ei->delalloc_inodes); - INIT_LIST_HEAD(&ei->ordered_operations); RB_CLEAR_NODE(&ei->rb_node); return inode; @@ -8146,17 +8136,6 @@ void btrfs_destroy_inode(struct inode *inode) if (!root) goto free; - /* - * Make sure we're properly removed from the ordered operation - * lists. - */ - smp_mb(); - if (!list_empty(&BTRFS_I(inode)->ordered_operations)) { - spin_lock(&root->fs_info->ordered_root_lock); - list_del_init(&BTRFS_I(inode)->ordered_operations); - spin_unlock(&root->fs_info->ordered_root_lock); - } - if (test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM, &BTRFS_I(inode)->runtime_flags)) { btrfs_info(root->fs_info, "inode %llu still on the orphan list", @@ -8338,12 +8317,10 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, ret = 0; /* - * we're using rename to replace one file with another. - * and the replacement file is large. Start IO on it now so - * we don't add too much work to the end of the transaction + * we're using rename to replace one file with another. Start IO on it + * now so we don't add too much work to the end of the transaction */ - if (new_inode && S_ISREG(old_inode->i_mode) && new_inode->i_size && - old_inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT) + if (new_inode && S_ISREG(old_inode->i_mode) && new_inode->i_size) filemap_flush(old_inode->i_mapping); /* close the racy window with snapshot create/destroy ioctl */ @@ -8391,12 +8368,6 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, */ btrfs_pin_log_trans(root); } - /* - * make sure the inode gets flushed if it is replacing - * something. - */ - if (new_inode && new_inode->i_size && S_ISREG(old_inode->i_mode)) - btrfs_add_ordered_operation(trans, root, old_inode); inode_inc_iversion(old_dir); inode_inc_iversion(new_dir); diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index 7187b14faa6c..963895c1f801 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -571,18 +571,6 @@ void btrfs_remove_ordered_extent(struct inode *inode, trace_btrfs_ordered_extent_remove(inode, entry); - /* - * we have no more ordered extents for this inode and - * no dirty pages. We can safely remove it from the - * list of ordered extents - */ - if (RB_EMPTY_ROOT(&tree->tree) && - !mapping_tagged(inode->i_mapping, PAGECACHE_TAG_DIRTY)) { - spin_lock(&root->fs_info->ordered_root_lock); - list_del_init(&BTRFS_I(inode)->ordered_operations); - spin_unlock(&root->fs_info->ordered_root_lock); - } - if (!root->nr_ordered_extents) { spin_lock(&root->fs_info->ordered_root_lock); BUG_ON(list_empty(&root->ordered_root)); @@ -687,81 +675,6 @@ void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr) } /* - * this is used during transaction commit to write all the inodes - * added to the ordered operation list. These files must be fully on - * disk before the transaction commits. - * - * we have two modes here, one is to just start the IO via filemap_flush - * and the other is to wait for all the io. When we wait, we have an - * extra check to make sure the ordered operation list really is empty - * before we return - */ -int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans, - struct btrfs_root *root, int wait) -{ - struct btrfs_inode *btrfs_inode; - struct inode *inode; - struct btrfs_transaction *cur_trans = trans->transaction; - struct list_head splice; - struct list_head works; - struct btrfs_delalloc_work *work, *next; - int ret = 0; - - INIT_LIST_HEAD(&splice); - INIT_LIST_HEAD(&works); - - mutex_lock(&root->fs_info->ordered_extent_flush_mutex); - spin_lock(&root->fs_info->ordered_root_lock); - list_splice_init(&cur_trans->ordered_operations, &splice); - while (!list_empty(&splice)) { - btrfs_inode = list_entry(splice.next, struct btrfs_inode, - ordered_operations); - inode = &btrfs_inode->vfs_inode; - - list_del_init(&btrfs_inode->ordered_operations); - - /* - * the inode may be getting freed (in sys_unlink path). - */ - inode = igrab(inode); - if (!inode) - continue; - - if (!wait) - list_add_tail(&BTRFS_I(inode)->ordered_operations, - &cur_trans->ordered_operations); - spin_unlock(&root->fs_info->ordered_root_lock); - - work = btrfs_alloc_delalloc_work(inode, wait, 1); - if (!work) { - spin_lock(&root->fs_info->ordered_root_lock); - if (list_empty(&BTRFS_I(inode)->ordered_operations)) - list_add_tail(&btrfs_inode->ordered_operations, - &splice); - list_splice_tail(&splice, - &cur_trans->ordered_operations); - spin_unlock(&root->fs_info->ordered_root_lock); - ret = -ENOMEM; - goto out; - } - list_add_tail(&work->list, &works); - btrfs_queue_work(root->fs_info->flush_workers, - &work->work); - - cond_resched(); - spin_lock(&root->fs_info->ordered_root_lock); - } - spin_unlock(&root->fs_info->ordered_root_lock); -out: - list_for_each_entry_safe(work, next, &works, list) { - list_del_init(&work->list); - btrfs_wait_and_free_delalloc_work(work); - } - mutex_unlock(&root->fs_info->ordered_extent_flush_mutex); - return ret; -} - -/* * Used to start IO or wait for a given ordered extent to finish. * * If wait is one, this effectively waits on page writeback for all the pages @@ -1120,42 +1033,6 @@ out: return index; } - -/* - * add a given inode to the list of inodes that must be fully on - * disk before a transaction commit finishes. - * - * This basically gives us the ext3 style data=ordered mode, and it is mostly - * used to make sure renamed files are fully on disk. - * - * It is a noop if the inode is already fully on disk. - * - * If trans is not null, we'll do a friendly check for a transaction that - * is already flushing things and force the IO down ourselves. - */ -void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct inode *inode) -{ - struct btrfs_transaction *cur_trans = trans->transaction; - u64 last_mod; - - last_mod = max(BTRFS_I(inode)->generation, BTRFS_I(inode)->last_trans); - - /* - * if this file hasn't been changed since the last transaction - * commit, we can safely return without doing anything - */ - if (last_mod <= root->fs_info->last_trans_committed) - return; - - spin_lock(&root->fs_info->ordered_root_lock); - if (list_empty(&BTRFS_I(inode)->ordered_operations)) { - list_add_tail(&BTRFS_I(inode)->ordered_operations, - &cur_trans->ordered_operations); - } - spin_unlock(&root->fs_info->ordered_root_lock); -} - int __init ordered_data_init(void) { btrfs_ordered_extent_cache = kmem_cache_create("btrfs_ordered_extent", diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index 246897058efb..d81a274d621e 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -190,11 +190,6 @@ int btrfs_ordered_update_i_size(struct inode *inode, u64 offset, struct btrfs_ordered_extent *ordered); int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr, u32 *sum, int len); -int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans, - struct btrfs_root *root, int wait); -void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct inode *inode); int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr); void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr); void btrfs_get_logged_extents(struct inode *inode, diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 98cb6b2630f9..b497498484be 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -1201,6 +1201,50 @@ out: mutex_unlock(&fs_info->qgroup_ioctl_lock); return ret; } + +static int comp_oper_exist(struct btrfs_qgroup_operation *oper1, + struct btrfs_qgroup_operation *oper2) +{ + /* + * Ignore seq and type here, we're looking for any operation + * at all related to this extent on that root. + */ + if (oper1->bytenr < oper2->bytenr) + return -1; + if (oper1->bytenr > oper2->bytenr) + return 1; + if (oper1->ref_root < oper2->ref_root) + return -1; + if (oper1->ref_root > oper2->ref_root) + return 1; + return 0; +} + +static int qgroup_oper_exists(struct btrfs_fs_info *fs_info, + struct btrfs_qgroup_operation *oper) +{ + struct rb_node *n; + struct btrfs_qgroup_operation *cur; + int cmp; + + spin_lock(&fs_info->qgroup_op_lock); + n = fs_info->qgroup_op_tree.rb_node; + while (n) { + cur = rb_entry(n, struct btrfs_qgroup_operation, n); + cmp = comp_oper_exist(cur, oper); + if (cmp < 0) { + n = n->rb_right; + } else if (cmp) { + n = n->rb_left; + } else { + spin_unlock(&fs_info->qgroup_op_lock); + return -EEXIST; + } + } + spin_unlock(&fs_info->qgroup_op_lock); + return 0; +} + static int comp_oper(struct btrfs_qgroup_operation *oper1, struct btrfs_qgroup_operation *oper2) { @@ -1290,6 +1334,23 @@ int btrfs_qgroup_record_ref(struct btrfs_trans_handle *trans, oper->seq = atomic_inc_return(&fs_info->qgroup_op_seq); INIT_LIST_HEAD(&oper->elem.list); oper->elem.seq = 0; + + if (type == BTRFS_QGROUP_OPER_SUB_SUBTREE) { + /* + * If any operation for this bytenr/ref_root combo + * exists, then we know it's not exclusively owned and + * shouldn't be queued up. + * + * This also catches the case where we have a cloned + * extent that gets queued up multiple times during + * drop snapshot. + */ + if (qgroup_oper_exists(fs_info, oper)) { + kfree(oper); + return 0; + } + } + ret = insert_qgroup_oper(fs_info, oper); if (ret) { /* Shouldn't happen so have an assert for developers */ @@ -1884,6 +1945,111 @@ out: } /* + * Process a reference to a shared subtree. This type of operation is + * queued during snapshot removal when we encounter extents which are + * shared between more than one root. + */ +static int qgroup_subtree_accounting(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, + struct btrfs_qgroup_operation *oper) +{ + struct ulist *roots = NULL; + struct ulist_node *unode; + struct ulist_iterator uiter; + struct btrfs_qgroup_list *glist; + struct ulist *parents; + int ret = 0; + int err; + struct btrfs_qgroup *qg; + u64 root_obj = 0; + struct seq_list elem = {}; + + parents = ulist_alloc(GFP_NOFS); + if (!parents) + return -ENOMEM; + + btrfs_get_tree_mod_seq(fs_info, &elem); + ret = btrfs_find_all_roots(trans, fs_info, oper->bytenr, + elem.seq, &roots); + btrfs_put_tree_mod_seq(fs_info, &elem); + if (ret < 0) + return ret; + + if (roots->nnodes != 1) + goto out; + + ULIST_ITER_INIT(&uiter); + unode = ulist_next(roots, &uiter); /* Only want 1 so no need to loop */ + /* + * If we find our ref root then that means all refs + * this extent has to the root have not yet been + * deleted. In that case, we do nothing and let the + * last ref for this bytenr drive our update. + * + * This can happen for example if an extent is + * referenced multiple times in a snapshot (clone, + * etc). If we are in the middle of snapshot removal, + * queued updates for such an extent will find the + * root if we have not yet finished removing the + * snapshot. + */ + if (unode->val == oper->ref_root) + goto out; + + root_obj = unode->val; + BUG_ON(!root_obj); + + spin_lock(&fs_info->qgroup_lock); + qg = find_qgroup_rb(fs_info, root_obj); + if (!qg) + goto out_unlock; + + qg->excl += oper->num_bytes; + qg->excl_cmpr += oper->num_bytes; + qgroup_dirty(fs_info, qg); + + /* + * Adjust counts for parent groups. First we find all + * parents, then in the 2nd loop we do the adjustment + * while adding parents of the parents to our ulist. + */ + list_for_each_entry(glist, &qg->groups, next_group) { + err = ulist_add(parents, glist->group->qgroupid, + ptr_to_u64(glist->group), GFP_ATOMIC); + if (err < 0) { + ret = err; + goto out_unlock; + } + } + + ULIST_ITER_INIT(&uiter); + while ((unode = ulist_next(parents, &uiter))) { + qg = u64_to_ptr(unode->aux); + qg->excl += oper->num_bytes; + qg->excl_cmpr += oper->num_bytes; + qgroup_dirty(fs_info, qg); + + /* Add any parents of the parents */ + list_for_each_entry(glist, &qg->groups, next_group) { + err = ulist_add(parents, glist->group->qgroupid, + ptr_to_u64(glist->group), GFP_ATOMIC); + if (err < 0) { + ret = err; + goto out_unlock; + } + } + } + +out_unlock: + spin_unlock(&fs_info->qgroup_lock); + +out: + ulist_free(roots); + ulist_free(parents); + return ret; +} + +/* * btrfs_qgroup_account_ref is called for every ref that is added to or deleted * from the fs. First, all roots referencing the extent are searched, and * then the space is accounted accordingly to the different roots. The @@ -1920,6 +2086,9 @@ static int btrfs_qgroup_account(struct btrfs_trans_handle *trans, case BTRFS_QGROUP_OPER_SUB_SHARED: ret = qgroup_shared_accounting(trans, fs_info, oper); break; + case BTRFS_QGROUP_OPER_SUB_SUBTREE: + ret = qgroup_subtree_accounting(trans, fs_info, oper); + break; default: ASSERT(0); } diff --git a/fs/btrfs/qgroup.h b/fs/btrfs/qgroup.h index 5952ff1fbd7a..18cc68ca3090 100644 --- a/fs/btrfs/qgroup.h +++ b/fs/btrfs/qgroup.h @@ -44,6 +44,7 @@ enum btrfs_qgroup_operation_type { BTRFS_QGROUP_OPER_ADD_SHARED, BTRFS_QGROUP_OPER_SUB_EXCL, BTRFS_QGROUP_OPER_SUB_SHARED, + BTRFS_QGROUP_OPER_SUB_SUBTREE, }; struct btrfs_qgroup_operation { diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 67b48b9a03e0..c4124de4435b 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1665,6 +1665,21 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes) return 0; } +/* + * Calculate numbers for 'df', pessimistic in case of mixed raid profiles. + * + * If there's a redundant raid level at DATA block groups, use the respective + * multiplier to scale the sizes. + * + * Unused device space usage is based on simulating the chunk allocator + * algorithm that respects the device sizes, order of allocations and the + * 'alloc_start' value, this is a close approximation of the actual use but + * there are other factors that may change the result (like a new metadata + * chunk). + * + * FIXME: not accurate for mixed block groups, total and free/used are ok, + * available appears slightly larger. + */ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) { struct btrfs_fs_info *fs_info = btrfs_sb(dentry->d_sb); @@ -1675,6 +1690,8 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) u64 total_free_data = 0; int bits = dentry->d_sb->s_blocksize_bits; __be32 *fsid = (__be32 *)fs_info->fsid; + unsigned factor = 1; + struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv; int ret; /* holding chunk_muext to avoid allocating new chunks */ @@ -1682,30 +1699,52 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) rcu_read_lock(); list_for_each_entry_rcu(found, head, list) { if (found->flags & BTRFS_BLOCK_GROUP_DATA) { + int i; + total_free_data += found->disk_total - found->disk_used; total_free_data -= btrfs_account_ro_block_groups_free_space(found); + + for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) { + if (!list_empty(&found->block_groups[i])) { + switch (i) { + case BTRFS_RAID_DUP: + case BTRFS_RAID_RAID1: + case BTRFS_RAID_RAID10: + factor = 2; + } + } + } } total_used += found->disk_used; } + rcu_read_unlock(); - buf->f_namelen = BTRFS_NAME_LEN; - buf->f_blocks = btrfs_super_total_bytes(disk_super) >> bits; - buf->f_bfree = buf->f_blocks - (total_used >> bits); - buf->f_bsize = dentry->d_sb->s_blocksize; - buf->f_type = BTRFS_SUPER_MAGIC; + buf->f_blocks = div_u64(btrfs_super_total_bytes(disk_super), factor); + buf->f_blocks >>= bits; + buf->f_bfree = buf->f_blocks - (div_u64(total_used, factor) >> bits); + + /* Account global block reserve as used, it's in logical size already */ + spin_lock(&block_rsv->lock); + buf->f_bfree -= block_rsv->size >> bits; + spin_unlock(&block_rsv->lock); + buf->f_bavail = total_free_data; ret = btrfs_calc_avail_data_space(fs_info->tree_root, &total_free_data); if (ret) { mutex_unlock(&fs_info->chunk_mutex); return ret; } - buf->f_bavail += total_free_data; + buf->f_bavail += div_u64(total_free_data, factor); buf->f_bavail = buf->f_bavail >> bits; mutex_unlock(&fs_info->chunk_mutex); + buf->f_type = BTRFS_SUPER_MAGIC; + buf->f_bsize = dentry->d_sb->s_blocksize; + buf->f_namelen = BTRFS_NAME_LEN; + /* We treat it as constant endianness (it doesn't matter _which_) because we want the fsid to come out the same whether mounted on a big-endian or little-endian host */ diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 5f379affdf23..d89c6d3542ca 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -218,7 +218,6 @@ loop: spin_lock_init(&cur_trans->delayed_refs.lock); INIT_LIST_HEAD(&cur_trans->pending_snapshots); - INIT_LIST_HEAD(&cur_trans->ordered_operations); INIT_LIST_HEAD(&cur_trans->pending_chunks); INIT_LIST_HEAD(&cur_trans->switch_commits); list_add_tail(&cur_trans->list, &fs_info->trans_list); @@ -1612,27 +1611,6 @@ static void cleanup_transaction(struct btrfs_trans_handle *trans, kmem_cache_free(btrfs_trans_handle_cachep, trans); } -static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans, - struct btrfs_root *root) -{ - int ret; - - ret = btrfs_run_delayed_items(trans, root); - if (ret) - return ret; - - /* - * rename don't use btrfs_join_transaction, so, once we - * set the transaction to blocked above, we aren't going - * to get any new ordered operations. We can safely run - * it here and no for sure that nothing new will be added - * to the list - */ - ret = btrfs_run_ordered_operations(trans, root, 1); - - return ret; -} - static inline int btrfs_start_delalloc_flush(struct btrfs_fs_info *fs_info) { if (btrfs_test_opt(fs_info->tree_root, FLUSHONCOMMIT)) @@ -1653,13 +1631,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, struct btrfs_transaction *prev_trans = NULL; int ret; - ret = btrfs_run_ordered_operations(trans, root, 0); - if (ret) { - btrfs_abort_transaction(trans, root, ret); - btrfs_end_transaction(trans, root); - return ret; - } - /* Stop the commit early if ->aborted is set */ if (unlikely(ACCESS_ONCE(cur_trans->aborted))) { ret = cur_trans->aborted; @@ -1740,7 +1711,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, if (ret) goto cleanup_transaction; - ret = btrfs_flush_all_pending_stuffs(trans, root); + ret = btrfs_run_delayed_items(trans, root); if (ret) goto cleanup_transaction; @@ -1748,7 +1719,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, extwriter_counter_read(cur_trans) == 0); /* some pending stuffs might be added after the previous flush. */ - ret = btrfs_flush_all_pending_stuffs(trans, root); + ret = btrfs_run_delayed_items(trans, root); if (ret) goto cleanup_transaction; diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 7dd558ed0716..579be51b27e5 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -55,7 +55,6 @@ struct btrfs_transaction { wait_queue_head_t writer_wait; wait_queue_head_t commit_wait; struct list_head pending_snapshots; - struct list_head ordered_operations; struct list_head pending_chunks; struct list_head switch_commits; struct btrfs_delayed_ref_root delayed_refs; diff --git a/fs/btrfs/ulist.h b/fs/btrfs/ulist.h index 7f78cbf5cf41..4c29db604bbe 100644 --- a/fs/btrfs/ulist.h +++ b/fs/btrfs/ulist.h @@ -57,6 +57,21 @@ void ulist_free(struct ulist *ulist); int ulist_add(struct ulist *ulist, u64 val, u64 aux, gfp_t gfp_mask); int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux, u64 *old_aux, gfp_t gfp_mask); + +/* just like ulist_add_merge() but take a pointer for the aux data */ +static inline int ulist_add_merge_ptr(struct ulist *ulist, u64 val, void *aux, + void **old_aux, gfp_t gfp_mask) +{ +#if BITS_PER_LONG == 32 + u64 old64 = (uintptr_t)*old_aux; + int ret = ulist_add_merge(ulist, val, (uintptr_t)aux, &old64, gfp_mask); + *old_aux = (void *)((uintptr_t)old64); + return ret; +#else + return ulist_add_merge(ulist, val, (u64)aux, (u64 *)old_aux, gfp_mask); +#endif +} + struct ulist_node *ulist_next(struct ulist *ulist, struct ulist_iterator *uiter); diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index ac4f260155c8..889b98455750 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -207,6 +207,19 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf) return 0; } +static long cifs_fallocate(struct file *file, int mode, loff_t off, loff_t len) +{ + struct super_block *sb = file->f_path.dentry->d_sb; + struct cifs_sb_info *cifs_sb = CIFS_SB(sb); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); + struct TCP_Server_Info *server = tcon->ses->server; + + if (server->ops->fallocate) + return server->ops->fallocate(file, tcon, mode, off, len); + + return -EOPNOTSUPP; +} + static int cifs_permission(struct inode *inode, int mask) { struct cifs_sb_info *cifs_sb; @@ -812,8 +825,9 @@ static int cifs_setlease(struct file *file, long arg, struct file_lock **lease) if (!(S_ISREG(inode->i_mode))) return -EINVAL; - /* check if file is oplocked */ - if (((arg == F_RDLCK) && CIFS_CACHE_READ(CIFS_I(inode))) || + /* Check if file is oplocked if this is request for new lease */ + if (arg == F_UNLCK || + ((arg == F_RDLCK) && CIFS_CACHE_READ(CIFS_I(inode))) || ((arg == F_WRLCK) && CIFS_CACHE_WRITE(CIFS_I(inode)))) return generic_setlease(file, arg, lease); else if (tlink_tcon(cfile->tlink)->local_lease && @@ -908,6 +922,7 @@ const struct file_operations cifs_file_ops = { .unlocked_ioctl = cifs_ioctl, #endif /* CONFIG_CIFS_POSIX */ .setlease = cifs_setlease, + .fallocate = cifs_fallocate, }; const struct file_operations cifs_file_strict_ops = { @@ -927,6 +942,7 @@ const struct file_operations cifs_file_strict_ops = { .unlocked_ioctl = cifs_ioctl, #endif /* CONFIG_CIFS_POSIX */ .setlease = cifs_setlease, + .fallocate = cifs_fallocate, }; const struct file_operations cifs_file_direct_ops = { @@ -947,6 +963,7 @@ const struct file_operations cifs_file_direct_ops = { #endif /* CONFIG_CIFS_POSIX */ .llseek = cifs_llseek, .setlease = cifs_setlease, + .fallocate = cifs_fallocate, }; const struct file_operations cifs_file_nobrl_ops = { @@ -965,6 +982,7 @@ const struct file_operations cifs_file_nobrl_ops = { .unlocked_ioctl = cifs_ioctl, #endif /* CONFIG_CIFS_POSIX */ .setlease = cifs_setlease, + .fallocate = cifs_fallocate, }; const struct file_operations cifs_file_strict_nobrl_ops = { @@ -983,6 +1001,7 @@ const struct file_operations cifs_file_strict_nobrl_ops = { .unlocked_ioctl = cifs_ioctl, #endif /* CONFIG_CIFS_POSIX */ .setlease = cifs_setlease, + .fallocate = cifs_fallocate, }; const struct file_operations cifs_file_direct_nobrl_ops = { @@ -1002,6 +1021,7 @@ const struct file_operations cifs_file_direct_nobrl_ops = { #endif /* CONFIG_CIFS_POSIX */ .llseek = cifs_llseek, .setlease = cifs_setlease, + .fallocate = cifs_fallocate, }; const struct file_operations cifs_dir_ops = { diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 0012e1e291d4..dfc731b02aa9 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -409,6 +409,10 @@ struct smb_version_operations { /* get mtu credits */ int (*wait_mtu_credits)(struct TCP_Server_Info *, unsigned int, unsigned int *, unsigned int *); + /* check if we need to issue closedir */ + bool (*dir_needs_close)(struct cifsFileInfo *); + long (*fallocate)(struct file *, struct cifs_tcon *, int, loff_t, + loff_t); }; struct smb_version_values { @@ -883,6 +887,7 @@ struct cifs_tcon { for this mount even if server would support */ bool local_lease:1; /* check leases (only) on local system not remote */ bool broken_posix_open; /* e.g. Samba server versions < 3.3.2, 3.2.9 */ + bool broken_sparse_sup; /* if server or share does not support sparse */ bool need_reconnect:1; /* connection reset, tid now invalid */ #ifdef CONFIG_CIFS_SMB2 bool print:1; /* set if connection to printer share */ diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index 33df36ef9d52..5f9822ac0245 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -2253,6 +2253,29 @@ typedef struct { /* minimum includes first three fields, and empty FS Name */ #define MIN_FS_ATTR_INFO_SIZE 12 + +/* List of FileSystemAttributes - see 2.5.1 of MS-FSCC */ +#define FILE_SUPPORT_INTEGRITY_STREAMS 0x04000000 +#define FILE_SUPPORTS_USN_JOURNAL 0x02000000 +#define FILE_SUPPORTS_OPEN_BY_FILE_ID 0x01000000 +#define FILE_SUPPORTS_EXTENDED_ATTRIBUTES 0x00800000 +#define FILE_SUPPORTS_HARD_LINKS 0x00400000 +#define FILE_SUPPORTS_TRANSACTIONS 0x00200000 +#define FILE_SEQUENTIAL_WRITE_ONCE 0x00100000 +#define FILE_READ_ONLY_VOLUME 0x00080000 +#define FILE_NAMED_STREAMS 0x00040000 +#define FILE_SUPPORTS_ENCRYPTION 0x00020000 +#define FILE_SUPPORTS_OBJECT_IDS 0x00010000 +#define FILE_VOLUME_IS_COMPRESSED 0x00008000 +#define FILE_SUPPORTS_REMOTE_STORAGE 0x00000100 +#define FILE_SUPPORTS_REPARSE_POINTS 0x00000080 +#define FILE_SUPPORTS_SPARSE_FILES 0x00000040 +#define FILE_VOLUME_QUOTAS 0x00000020 +#define FILE_FILE_COMPRESSION 0x00000010 +#define FILE_PERSISTENT_ACLS 0x00000008 +#define FILE_UNICODE_ON_DISK 0x00000004 +#define FILE_CASE_PRESERVED_NAMES 0x00000002 +#define FILE_CASE_SENSITIVE_SEARCH 0x00000001 typedef struct { __le32 Attributes; __le32 MaxPathNameComponentLength; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 4ab2f79ffa7a..d5fec92e0360 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -762,7 +762,7 @@ int cifs_closedir(struct inode *inode, struct file *file) cifs_dbg(FYI, "Freeing private data in close dir\n"); spin_lock(&cifs_file_list_lock); - if (!cfile->srch_inf.endOfSearch && !cfile->invalidHandle) { + if (server->ops->dir_needs_close(cfile)) { cfile->invalidHandle = true; spin_unlock(&cifs_file_list_lock); if (server->ops->close_dir) diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 426d6c6ad8bf..949ec909ec9a 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1727,6 +1727,12 @@ unlink_target: target_dentry, to_name); } + /* force revalidate to go get info when needed */ + CIFS_I(source_dir)->time = CIFS_I(target_dir)->time = 0; + + source_dir->i_ctime = source_dir->i_mtime = target_dir->i_ctime = + target_dir->i_mtime = current_fs_time(source_dir->i_sb); + cifs_rename_exit: kfree(info_buf_source); kfree(from_name); diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 81340c6253eb..b7415d596dbd 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -574,13 +574,6 @@ void cifs_set_oplock_level(struct cifsInodeInfo *cinode, __u32 oplock) cinode->oplock = 0; } -static int -cifs_oplock_break_wait(void *unused) -{ - schedule(); - return signal_pending(current) ? -ERESTARTSYS : 0; -} - /* * We wait for oplock breaks to be processed before we attempt to perform * writes. diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index b15862e0f68c..798c80a41c88 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -593,7 +593,7 @@ find_cifs_entry(const unsigned int xid, struct cifs_tcon *tcon, loff_t pos, /* close and restart search */ cifs_dbg(FYI, "search backing up - close and restart search\n"); spin_lock(&cifs_file_list_lock); - if (!cfile->srch_inf.endOfSearch && !cfile->invalidHandle) { + if (server->ops->dir_needs_close(cfile)) { cfile->invalidHandle = true; spin_unlock(&cifs_file_list_lock); if (server->ops->close) diff --git a/fs/cifs/smb1ops.c b/fs/cifs/smb1ops.c index 5e8c22d6c7b9..1a6df4b03f67 100644 --- a/fs/cifs/smb1ops.c +++ b/fs/cifs/smb1ops.c @@ -1015,6 +1015,12 @@ cifs_wp_retry_size(struct inode *inode) return CIFS_SB(inode->i_sb)->wsize; } +static bool +cifs_dir_needs_close(struct cifsFileInfo *cfile) +{ + return !cfile->srch_inf.endOfSearch && !cfile->invalidHandle; +} + struct smb_version_operations smb1_operations = { .send_cancel = send_nt_cancel, .compare_fids = cifs_compare_fids, @@ -1086,6 +1092,7 @@ struct smb_version_operations smb1_operations = { .create_mf_symlink = cifs_create_mf_symlink, .is_read_op = cifs_is_read_op, .wp_retry_size = cifs_wp_retry_size, + .dir_needs_close = cifs_dir_needs_close, #ifdef CONFIG_CIFS_XATTR .query_all_EAs = CIFSSMBQAllEAs, .set_EA = CIFSSMBSetEA, diff --git a/fs/cifs/smb2maperror.c b/fs/cifs/smb2maperror.c index e31a9dfdcd39..af59d03db492 100644 --- a/fs/cifs/smb2maperror.c +++ b/fs/cifs/smb2maperror.c @@ -214,7 +214,7 @@ static const struct status_to_posix_error smb2_error_map_table[] = { {STATUS_BREAKPOINT, -EIO, "STATUS_BREAKPOINT"}, {STATUS_SINGLE_STEP, -EIO, "STATUS_SINGLE_STEP"}, {STATUS_BUFFER_OVERFLOW, -EIO, "STATUS_BUFFER_OVERFLOW"}, - {STATUS_NO_MORE_FILES, -EIO, "STATUS_NO_MORE_FILES"}, + {STATUS_NO_MORE_FILES, -ENODATA, "STATUS_NO_MORE_FILES"}, {STATUS_WAKE_SYSTEM_DEBUGGER, -EIO, "STATUS_WAKE_SYSTEM_DEBUGGER"}, {STATUS_HANDLES_CLOSED, -EIO, "STATUS_HANDLES_CLOSED"}, {STATUS_NO_INHERITANCE, -EIO, "STATUS_NO_INHERITANCE"}, @@ -298,7 +298,7 @@ static const struct status_to_posix_error smb2_error_map_table[] = { {STATUS_INVALID_PARAMETER, -EINVAL, "STATUS_INVALID_PARAMETER"}, {STATUS_NO_SUCH_DEVICE, -ENODEV, "STATUS_NO_SUCH_DEVICE"}, {STATUS_NO_SUCH_FILE, -ENOENT, "STATUS_NO_SUCH_FILE"}, - {STATUS_INVALID_DEVICE_REQUEST, -EIO, "STATUS_INVALID_DEVICE_REQUEST"}, + {STATUS_INVALID_DEVICE_REQUEST, -EOPNOTSUPP, "STATUS_INVALID_DEVICE_REQUEST"}, {STATUS_END_OF_FILE, -ENODATA, "STATUS_END_OF_FILE"}, {STATUS_WRONG_VOLUME, -EIO, "STATUS_WRONG_VOLUME"}, {STATUS_NO_MEDIA_IN_DEVICE, -EIO, "STATUS_NO_MEDIA_IN_DEVICE"}, diff --git a/fs/cifs/smb2misc.c b/fs/cifs/smb2misc.c index f2e6ac29a8d6..4aa7a0f07d6e 100644 --- a/fs/cifs/smb2misc.c +++ b/fs/cifs/smb2misc.c @@ -178,9 +178,24 @@ smb2_check_message(char *buf, unsigned int length) /* Windows 7 server returns 24 bytes more */ if (clc_len + 20 == len && command == SMB2_OPLOCK_BREAK_HE) return 0; - /* server can return one byte more */ + /* server can return one byte more due to implied bcc[0] */ if (clc_len == 4 + len + 1) return 0; + + /* + * MacOS server pads after SMB2.1 write response with 3 bytes + * of junk. Other servers match RFC1001 len to actual + * SMB2/SMB3 frame length (header + smb2 response specific data) + * Log the server error (once), but allow it and continue + * since the frame is parseable. + */ + if (clc_len < 4 /* RFC1001 header size */ + len) { + printk_once(KERN_WARNING + "SMB2 server sent bad RFC1001 len %d not %d\n", + len, clc_len - 4); + return 0; + } + return 1; } return 0; diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c index 77f8aeb9c2fc..5a48aa290dfe 100644 --- a/fs/cifs/smb2ops.c +++ b/fs/cifs/smb2ops.c @@ -731,11 +731,72 @@ smb2_sync_write(const unsigned int xid, struct cifsFileInfo *cfile, return SMB2_write(xid, parms, written, iov, nr_segs); } +/* Set or clear the SPARSE_FILE attribute based on value passed in setsparse */ +static bool smb2_set_sparse(const unsigned int xid, struct cifs_tcon *tcon, + struct cifsFileInfo *cfile, struct inode *inode, __u8 setsparse) +{ + struct cifsInodeInfo *cifsi; + int rc; + + cifsi = CIFS_I(inode); + + /* if file already sparse don't bother setting sparse again */ + if ((cifsi->cifsAttrs & FILE_ATTRIBUTE_SPARSE_FILE) && setsparse) + return true; /* already sparse */ + + if (!(cifsi->cifsAttrs & FILE_ATTRIBUTE_SPARSE_FILE) && !setsparse) + return true; /* already not sparse */ + + /* + * Can't check for sparse support on share the usual way via the + * FS attribute info (FILE_SUPPORTS_SPARSE_FILES) on the share + * since Samba server doesn't set the flag on the share, yet + * supports the set sparse FSCTL and returns sparse correctly + * in the file attributes. If we fail setting sparse though we + * mark that server does not support sparse files for this share + * to avoid repeatedly sending the unsupported fsctl to server + * if the file is repeatedly extended. + */ + if (tcon->broken_sparse_sup) + return false; + + rc = SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid, + cfile->fid.volatile_fid, FSCTL_SET_SPARSE, + true /* is_fctl */, &setsparse, 1, NULL, NULL); + if (rc) { + tcon->broken_sparse_sup = true; + cifs_dbg(FYI, "set sparse rc = %d\n", rc); + return false; + } + + if (setsparse) + cifsi->cifsAttrs |= FILE_ATTRIBUTE_SPARSE_FILE; + else + cifsi->cifsAttrs &= (~FILE_ATTRIBUTE_SPARSE_FILE); + + return true; +} + static int smb2_set_file_size(const unsigned int xid, struct cifs_tcon *tcon, struct cifsFileInfo *cfile, __u64 size, bool set_alloc) { __le64 eof = cpu_to_le64(size); + struct inode *inode; + + /* + * If extending file more than one page make sparse. Many Linux fs + * make files sparse by default when extending via ftruncate + */ + inode = cfile->dentry->d_inode; + + if (!set_alloc && (size > inode->i_size + 8192)) { + __u8 set_sparse = 1; + + /* whether set sparse succeeds or not, extend the file */ + smb2_set_sparse(xid, tcon, cfile, inode, set_sparse); + } + return SMB2_set_eof(xid, tcon, cfile->fid.persistent_fid, cfile->fid.volatile_fid, cfile->pid, &eof, false); } @@ -954,6 +1015,105 @@ smb2_query_symlink(const unsigned int xid, struct cifs_tcon *tcon, return rc; } +static long smb3_zero_range(struct file *file, struct cifs_tcon *tcon, + loff_t offset, loff_t len, bool keep_size) +{ + struct inode *inode; + struct cifsInodeInfo *cifsi; + struct cifsFileInfo *cfile = file->private_data; + struct file_zero_data_information fsctl_buf; + long rc; + unsigned int xid; + + xid = get_xid(); + + inode = cfile->dentry->d_inode; + cifsi = CIFS_I(inode); + + /* if file not oplocked can't be sure whether asking to extend size */ + if (!CIFS_CACHE_READ(cifsi)) + if (keep_size == false) + return -EOPNOTSUPP; + + /* + * Must check if file sparse since fallocate -z (zero range) assumes + * non-sparse allocation + */ + if (!(cifsi->cifsAttrs & FILE_ATTRIBUTE_SPARSE_FILE)) + return -EOPNOTSUPP; + + /* + * need to make sure we are not asked to extend the file since the SMB3 + * fsctl does not change the file size. In the future we could change + * this to zero the first part of the range then set the file size + * which for a non sparse file would zero the newly extended range + */ + if (keep_size == false) + if (i_size_read(inode) < offset + len) + return -EOPNOTSUPP; + + cifs_dbg(FYI, "offset %lld len %lld", offset, len); + + fsctl_buf.FileOffset = cpu_to_le64(offset); + fsctl_buf.BeyondFinalZero = cpu_to_le64(offset + len); + + rc = SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid, + cfile->fid.volatile_fid, FSCTL_SET_ZERO_DATA, + true /* is_fctl */, (char *)&fsctl_buf, + sizeof(struct file_zero_data_information), NULL, NULL); + free_xid(xid); + return rc; +} + +static long smb3_punch_hole(struct file *file, struct cifs_tcon *tcon, + loff_t offset, loff_t len) +{ + struct inode *inode; + struct cifsInodeInfo *cifsi; + struct cifsFileInfo *cfile = file->private_data; + struct file_zero_data_information fsctl_buf; + long rc; + unsigned int xid; + __u8 set_sparse = 1; + + xid = get_xid(); + + inode = cfile->dentry->d_inode; + cifsi = CIFS_I(inode); + + /* Need to make file sparse, if not already, before freeing range. */ + /* Consider adding equivalent for compressed since it could also work */ + if (!smb2_set_sparse(xid, tcon, cfile, inode, set_sparse)) + return -EOPNOTSUPP; + + cifs_dbg(FYI, "offset %lld len %lld", offset, len); + + fsctl_buf.FileOffset = cpu_to_le64(offset); + fsctl_buf.BeyondFinalZero = cpu_to_le64(offset + len); + + rc = SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid, + cfile->fid.volatile_fid, FSCTL_SET_ZERO_DATA, + true /* is_fctl */, (char *)&fsctl_buf, + sizeof(struct file_zero_data_information), NULL, NULL); + free_xid(xid); + return rc; +} + +static long smb3_fallocate(struct file *file, struct cifs_tcon *tcon, int mode, + loff_t off, loff_t len) +{ + /* KEEP_SIZE already checked for by do_fallocate */ + if (mode & FALLOC_FL_PUNCH_HOLE) + return smb3_punch_hole(file, tcon, off, len); + else if (mode & FALLOC_FL_ZERO_RANGE) { + if (mode & FALLOC_FL_KEEP_SIZE) + return smb3_zero_range(file, tcon, off, len, true); + return smb3_zero_range(file, tcon, off, len, false); + } + + return -EOPNOTSUPP; +} + static void smb2_downgrade_oplock(struct TCP_Server_Info *server, struct cifsInodeInfo *cinode, bool set_level2) @@ -1161,6 +1321,12 @@ smb2_wp_retry_size(struct inode *inode) SMB2_MAX_BUFFER_SIZE); } +static bool +smb2_dir_needs_close(struct cifsFileInfo *cfile) +{ + return !cfile->invalidHandle; +} + struct smb_version_operations smb20_operations = { .compare_fids = smb2_compare_fids, .setup_request = smb2_setup_request, @@ -1236,6 +1402,7 @@ struct smb_version_operations smb20_operations = { .parse_lease_buf = smb2_parse_lease_buf, .clone_range = smb2_clone_range, .wp_retry_size = smb2_wp_retry_size, + .dir_needs_close = smb2_dir_needs_close, }; struct smb_version_operations smb21_operations = { @@ -1313,6 +1480,7 @@ struct smb_version_operations smb21_operations = { .parse_lease_buf = smb2_parse_lease_buf, .clone_range = smb2_clone_range, .wp_retry_size = smb2_wp_retry_size, + .dir_needs_close = smb2_dir_needs_close, }; struct smb_version_operations smb30_operations = { @@ -1393,6 +1561,8 @@ struct smb_version_operations smb30_operations = { .clone_range = smb2_clone_range, .validate_negotiate = smb3_validate_negotiate, .wp_retry_size = smb2_wp_retry_size, + .dir_needs_close = smb2_dir_needs_close, + .fallocate = smb3_fallocate, }; struct smb_version_values smb20_values = { diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c index 42ebc1a8be6c..fa0dd044213b 100644 --- a/fs/cifs/smb2pdu.c +++ b/fs/cifs/smb2pdu.c @@ -907,7 +907,8 @@ tcon_exit: tcon_error_exit: if (rsp->hdr.Status == STATUS_BAD_NETWORK_NAME) { cifs_dbg(VFS, "BAD_NETWORK_NAME: %s\n", tree); - tcon->bad_network_name = true; + if (tcon) + tcon->bad_network_name = true; } goto tcon_exit; } @@ -1224,7 +1225,9 @@ SMB2_ioctl(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_fid, cifs_dbg(FYI, "SMB2 IOCTL\n"); - *out_data = NULL; + if (out_data != NULL) + *out_data = NULL; + /* zero out returned data len, in case of error */ if (plen) *plen = 0; @@ -2177,6 +2180,10 @@ SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, rsp = (struct smb2_query_directory_rsp *)iov[0].iov_base; if (rc) { + if (rc == -ENODATA && rsp->hdr.Status == STATUS_NO_MORE_FILES) { + srch_inf->endOfSearch = true; + rc = 0; + } cifs_stats_fail_inc(tcon, SMB2_QUERY_DIRECTORY_HE); goto qdir_exit; } @@ -2214,11 +2221,6 @@ SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, else cifs_dbg(VFS, "illegal search buffer type\n"); - if (rsp->hdr.Status == STATUS_NO_MORE_FILES) - srch_inf->endOfSearch = 1; - else - srch_inf->endOfSearch = 0; - return rc; qdir_exit: diff --git a/fs/cifs/smb2pdu.h b/fs/cifs/smb2pdu.h index 69f3595d3952..fbe486c285a9 100644 --- a/fs/cifs/smb2pdu.h +++ b/fs/cifs/smb2pdu.h @@ -573,6 +573,12 @@ struct copychunk_ioctl { __u32 Reserved2; } __packed; +/* this goes in the ioctl buffer when doing FSCTL_SET_ZERO_DATA */ +struct file_zero_data_information { + __le64 FileOffset; + __le64 BeyondFinalZero; +} __packed; + struct copychunk_ioctl_rsp { __le32 ChunksWritten; __le32 ChunkBytesWritten; diff --git a/fs/cifs/smbfsctl.h b/fs/cifs/smbfsctl.h index 0e538b5c9622..83efa59535be 100644 --- a/fs/cifs/smbfsctl.h +++ b/fs/cifs/smbfsctl.h @@ -63,7 +63,7 @@ #define FSCTL_SET_OBJECT_ID_EXTENDED 0x000900BC /* BB add struct */ #define FSCTL_CREATE_OR_GET_OBJECT_ID 0x000900C0 /* BB add struct */ #define FSCTL_SET_SPARSE 0x000900C4 /* BB add struct */ -#define FSCTL_SET_ZERO_DATA 0x000900C8 /* BB add struct */ +#define FSCTL_SET_ZERO_DATA 0x000980C8 #define FSCTL_SET_ENCRYPTION 0x000900D7 /* BB add struct */ #define FSCTL_ENCRYPTION_FSCTL_IO 0x000900DB /* BB add struct */ #define FSCTL_WRITE_RAW_ENCRYPTED 0x000900DF /* BB add struct */ diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 08cdfe5461e3..622e88249024 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -2828,8 +2828,9 @@ static int ext3_statfs (struct dentry * dentry, struct kstatfs * buf) */ overhead += ngroups * (2 + sbi->s_itb_per_group); - /* Add the journal blocks as well */ - overhead += sbi->s_journal->j_maxlen; + /* Add the internal journal blocks as well */ + if (sbi->s_journal && !sbi->journal_bdev) + overhead += sbi->s_journal->j_maxlen; sbi->s_overhead_last = overhead; smp_wmb(); diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 4556ce1af5b0..5ddaf8625d3b 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -61,7 +61,7 @@ static void isofs_put_super(struct super_block *sb) return; } -static int isofs_read_inode(struct inode *); +static int isofs_read_inode(struct inode *, int relocated); static int isofs_statfs (struct dentry *, struct kstatfs *); static struct kmem_cache *isofs_inode_cachep; @@ -1259,7 +1259,7 @@ out_toomany: goto out; } -static int isofs_read_inode(struct inode *inode) +static int isofs_read_inode(struct inode *inode, int relocated) { struct super_block *sb = inode->i_sb; struct isofs_sb_info *sbi = ISOFS_SB(sb); @@ -1404,7 +1404,7 @@ static int isofs_read_inode(struct inode *inode) */ if (!high_sierra) { - parse_rock_ridge_inode(de, inode); + parse_rock_ridge_inode(de, inode, relocated); /* if we want uid/gid set, override the rock ridge setting */ if (sbi->s_uid_set) inode->i_uid = sbi->s_uid; @@ -1483,9 +1483,10 @@ static int isofs_iget5_set(struct inode *ino, void *data) * offset that point to the underlying meta-data for the inode. The * code below is otherwise similar to the iget() code in * include/linux/fs.h */ -struct inode *isofs_iget(struct super_block *sb, - unsigned long block, - unsigned long offset) +struct inode *__isofs_iget(struct super_block *sb, + unsigned long block, + unsigned long offset, + int relocated) { unsigned long hashval; struct inode *inode; @@ -1507,7 +1508,7 @@ struct inode *isofs_iget(struct super_block *sb, return ERR_PTR(-ENOMEM); if (inode->i_state & I_NEW) { - ret = isofs_read_inode(inode); + ret = isofs_read_inode(inode, relocated); if (ret < 0) { iget_failed(inode); inode = ERR_PTR(ret); diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h index 99167238518d..0ac4c1f73fbd 100644 --- a/fs/isofs/isofs.h +++ b/fs/isofs/isofs.h @@ -107,7 +107,7 @@ extern int iso_date(char *, int); struct inode; /* To make gcc happy */ -extern int parse_rock_ridge_inode(struct iso_directory_record *, struct inode *); +extern int parse_rock_ridge_inode(struct iso_directory_record *, struct inode *, int relocated); extern int get_rock_ridge_filename(struct iso_directory_record *, char *, struct inode *); extern int isofs_name_translate(struct iso_directory_record *, char *, struct inode *); @@ -118,9 +118,24 @@ extern struct dentry *isofs_lookup(struct inode *, struct dentry *, unsigned int extern struct buffer_head *isofs_bread(struct inode *, sector_t); extern int isofs_get_blocks(struct inode *, sector_t, struct buffer_head **, unsigned long); -extern struct inode *isofs_iget(struct super_block *sb, - unsigned long block, - unsigned long offset); +struct inode *__isofs_iget(struct super_block *sb, + unsigned long block, + unsigned long offset, + int relocated); + +static inline struct inode *isofs_iget(struct super_block *sb, + unsigned long block, + unsigned long offset) +{ + return __isofs_iget(sb, block, offset, 0); +} + +static inline struct inode *isofs_iget_reloc(struct super_block *sb, + unsigned long block, + unsigned long offset) +{ + return __isofs_iget(sb, block, offset, 1); +} /* Because the inode number is no longer relevant to finding the * underlying meta-data for an inode, we are free to choose a more diff --git a/fs/isofs/rock.c b/fs/isofs/rock.c index c0bf42472e40..f488bbae541a 100644 --- a/fs/isofs/rock.c +++ b/fs/isofs/rock.c @@ -288,12 +288,16 @@ eio: goto out; } +#define RR_REGARD_XA 1 +#define RR_RELOC_DE 2 + static int parse_rock_ridge_inode_internal(struct iso_directory_record *de, - struct inode *inode, int regard_xa) + struct inode *inode, int flags) { int symlink_len = 0; int cnt, sig; + unsigned int reloc_block; struct inode *reloc; struct rock_ridge *rr; int rootflag; @@ -305,7 +309,7 @@ parse_rock_ridge_inode_internal(struct iso_directory_record *de, init_rock_state(&rs, inode); setup_rock_ridge(de, inode, &rs); - if (regard_xa) { + if (flags & RR_REGARD_XA) { rs.chr += 14; rs.len -= 14; if (rs.len < 0) @@ -485,12 +489,22 @@ repeat: "relocated directory\n"); goto out; case SIG('C', 'L'): - ISOFS_I(inode)->i_first_extent = - isonum_733(rr->u.CL.location); - reloc = - isofs_iget(inode->i_sb, - ISOFS_I(inode)->i_first_extent, - 0); + if (flags & RR_RELOC_DE) { + printk(KERN_ERR + "ISOFS: Recursive directory relocation " + "is not supported\n"); + goto eio; + } + reloc_block = isonum_733(rr->u.CL.location); + if (reloc_block == ISOFS_I(inode)->i_iget5_block && + ISOFS_I(inode)->i_iget5_offset == 0) { + printk(KERN_ERR + "ISOFS: Directory relocation points to " + "itself\n"); + goto eio; + } + ISOFS_I(inode)->i_first_extent = reloc_block; + reloc = isofs_iget_reloc(inode->i_sb, reloc_block, 0); if (IS_ERR(reloc)) { ret = PTR_ERR(reloc); goto out; @@ -637,9 +651,11 @@ static char *get_symlink_chunk(char *rpnt, struct rock_ridge *rr, char *plimit) return rpnt; } -int parse_rock_ridge_inode(struct iso_directory_record *de, struct inode *inode) +int parse_rock_ridge_inode(struct iso_directory_record *de, struct inode *inode, + int relocated) { - int result = parse_rock_ridge_inode_internal(de, inode, 0); + int flags = relocated ? RR_RELOC_DE : 0; + int result = parse_rock_ridge_inode_internal(de, inode, flags); /* * if rockridge flag was reset and we didn't look for attributes @@ -647,7 +663,8 @@ int parse_rock_ridge_inode(struct iso_directory_record *de, struct inode *inode) */ if ((ISOFS_SB(inode->i_sb)->s_rock_offset == -1) && (ISOFS_SB(inode->i_sb)->s_rock == 2)) { - result = parse_rock_ridge_inode_internal(de, inode, 14); + result = parse_rock_ridge_inode_internal(de, inode, + flags | RR_REGARD_XA); } return result; } diff --git a/fs/locks.c b/fs/locks.c index a6f54802d277..cb66fb05ad4a 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -247,6 +247,18 @@ void locks_free_lock(struct file_lock *fl) } EXPORT_SYMBOL(locks_free_lock); +static void +locks_dispose_list(struct list_head *dispose) +{ + struct file_lock *fl; + + while (!list_empty(dispose)) { + fl = list_first_entry(dispose, struct file_lock, fl_block); + list_del_init(&fl->fl_block); + locks_free_lock(fl); + } +} + void locks_init_lock(struct file_lock *fl) { memset(fl, 0, sizeof(struct file_lock)); @@ -285,7 +297,8 @@ EXPORT_SYMBOL(__locks_copy_lock); void locks_copy_lock(struct file_lock *new, struct file_lock *fl) { - locks_release_private(new); + /* "new" must be a freshly-initialized lock */ + WARN_ON_ONCE(new->fl_ops); __locks_copy_lock(new, fl); new->fl_file = fl->fl_file; @@ -650,12 +663,16 @@ static void locks_unlink_lock(struct file_lock **thisfl_p) * * Must be called with i_lock held! */ -static void locks_delete_lock(struct file_lock **thisfl_p) +static void locks_delete_lock(struct file_lock **thisfl_p, + struct list_head *dispose) { struct file_lock *fl = *thisfl_p; locks_unlink_lock(thisfl_p); - locks_free_lock(fl); + if (dispose) + list_add(&fl->fl_block, dispose); + else + locks_free_lock(fl); } /* Determine if lock sys_fl blocks lock caller_fl. Common functionality @@ -811,6 +828,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) struct inode * inode = file_inode(filp); int error = 0; int found = 0; + LIST_HEAD(dispose); if (!(request->fl_flags & FL_ACCESS) && (request->fl_type != F_UNLCK)) { new_fl = locks_alloc_lock(); @@ -833,7 +851,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) if (request->fl_type == fl->fl_type) goto out; found = 1; - locks_delete_lock(before); + locks_delete_lock(before, &dispose); break; } @@ -880,6 +898,7 @@ out: spin_unlock(&inode->i_lock); if (new_fl) locks_free_lock(new_fl); + locks_dispose_list(&dispose); return error; } @@ -893,6 +912,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str struct file_lock **before; int error; bool added = false; + LIST_HEAD(dispose); /* * We may need two file_lock structures for this operation, @@ -988,7 +1008,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str else request->fl_end = fl->fl_end; if (added) { - locks_delete_lock(before); + locks_delete_lock(before, &dispose); continue; } request = fl; @@ -1018,21 +1038,24 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str * one (This may happen several times). */ if (added) { - locks_delete_lock(before); + locks_delete_lock(before, &dispose); continue; } - /* Replace the old lock with the new one. - * Wake up anybody waiting for the old one, - * as the change in lock type might satisfy - * their needs. + /* + * Replace the old lock with new_fl, and + * remove the old one. It's safe to do the + * insert here since we know that we won't be + * using new_fl later, and that the lock is + * just replacing an existing lock. */ - locks_wake_up_blocks(fl); - fl->fl_start = request->fl_start; - fl->fl_end = request->fl_end; - fl->fl_type = request->fl_type; - locks_release_private(fl); - locks_copy_private(fl, request); - request = fl; + error = -ENOLCK; + if (!new_fl) + goto out; + locks_copy_lock(new_fl, request); + request = new_fl; + new_fl = NULL; + locks_delete_lock(before, &dispose); + locks_insert_lock(before, request); added = true; } } @@ -1093,6 +1116,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str locks_free_lock(new_fl); if (new_fl2) locks_free_lock(new_fl2); + locks_dispose_list(&dispose); return error; } @@ -1268,7 +1292,7 @@ int lease_modify(struct file_lock **before, int arg) printk(KERN_ERR "locks_delete_lock: fasync == %p\n", fl->fl_fasync); fl->fl_fasync = NULL; } - locks_delete_lock(before); + locks_delete_lock(before, NULL); } return 0; } @@ -1737,13 +1761,10 @@ static int do_fcntl_add_lease(unsigned int fd, struct file *filp, long arg) ret = fl; spin_lock(&inode->i_lock); error = __vfs_setlease(filp, arg, &ret); - if (error) { - spin_unlock(&inode->i_lock); - locks_free_lock(fl); - goto out_free_fasync; - } - if (ret != fl) - locks_free_lock(fl); + if (error) + goto out_unlock; + if (ret == fl) + fl = NULL; /* * fasync_insert_entry() returns the old entry if any. @@ -1755,9 +1776,10 @@ static int do_fcntl_add_lease(unsigned int fd, struct file *filp, long arg) new = NULL; error = __f_setown(filp, task_pid(current), PIDTYPE_PID, 0); +out_unlock: spin_unlock(&inode->i_lock); - -out_free_fasync: + if (fl) + locks_free_lock(fl); if (new) fasync_free(new); return error; @@ -2320,6 +2342,7 @@ void locks_remove_file(struct file *filp) struct inode * inode = file_inode(filp); struct file_lock *fl; struct file_lock **before; + LIST_HEAD(dispose); if (!inode->i_flock) return; @@ -2365,12 +2388,13 @@ void locks_remove_file(struct file *filp) fl->fl_type, fl->fl_flags, fl->fl_start, fl->fl_end); - locks_delete_lock(before); + locks_delete_lock(before, &dispose); continue; } before = &fl->fl_next; } spin_unlock(&inode->i_lock); + locks_dispose_list(&dispose); } /** @@ -2452,7 +2476,11 @@ static void lock_get_status(struct seq_file *f, struct file_lock *fl, seq_puts(f, "FLOCK ADVISORY "); } } else if (IS_LEASE(fl)) { - seq_puts(f, "LEASE "); + if (fl->fl_flags & FL_DELEG) + seq_puts(f, "DELEG "); + else + seq_puts(f, "LEASE "); + if (lease_breaking(fl)) seq_puts(f, "BREAKING "); else if (fl->fl_file) diff --git a/fs/udf/namei.c b/fs/udf/namei.c index 9737cba1357d..83a06001742b 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -1014,7 +1014,7 @@ static int udf_symlink(struct inode *dir, struct dentry *dentry, fi = udf_add_entry(dir, dentry, &fibh, &cfi, &err); if (!fi) - goto out_no_entry; + goto out_fail; cfi.icb.extLength = cpu_to_le32(sb->s_blocksize); cfi.icb.extLocation = cpu_to_lelb(iinfo->i_location); if (UDF_SB(inode->i_sb)->s_lvid_bh) { @@ -1036,6 +1036,7 @@ out: out_no_entry: up_write(&iinfo->i_data_sem); +out_fail: inode_dec_link_count(inode); iput(inode); goto out; |