diff options
Diffstat (limited to 'fs')
59 files changed, 2597 insertions, 1441 deletions
diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index eb159aaa5a11..89b156d85d63 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -59,22 +59,19 @@ struct posix_acl *btrfs_get_acl(struct inode *inode, int type) if (!value) return ERR_PTR(-ENOMEM); size = __btrfs_getxattr(inode, name, value, size); - if (size > 0) { - acl = posix_acl_from_xattr(value, size); - if (IS_ERR(acl)) { - kfree(value); - return acl; - } - set_cached_acl(inode, type, acl); - } - kfree(value); + } + if (size > 0) { + acl = posix_acl_from_xattr(value, size); } else if (size == -ENOENT || size == -ENODATA || size == 0) { /* FIXME, who returns -ENOENT? I think nobody */ acl = NULL; - set_cached_acl(inode, type, acl); } else { acl = ERR_PTR(-EIO); } + kfree(value); + + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); return acl; } diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index d9f99a16edd6..5a5d325a3935 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -103,11 +103,6 @@ struct btrfs_inode { */ u64 delalloc_bytes; - /* total number of bytes that may be used for this inode for - * delalloc - */ - u64 reserved_bytes; - /* * the size of the file stored in the metadata on disk. data=ordered * means the in-memory i_size might be larger than the size on disk @@ -115,9 +110,6 @@ struct btrfs_inode { */ u64 disk_i_size; - /* flags field from the on disk inode */ - u32 flags; - /* * if this is a directory then index_cnt is the counter for the index * number for new files that are created @@ -132,6 +124,15 @@ struct btrfs_inode { u64 last_unlink_trans; /* + * Number of bytes outstanding that are going to need csums. This is + * used in ENOSPC accounting. + */ + u64 csum_bytes; + + /* flags field from the on disk inode */ + u32 flags; + + /* * Counters to keep track of the number of extent item's we may use due * to delalloc and such. outstanding_extents is the number of extent * items we think we'll end up using, and reserved_extents is the number diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index 8ec5d86f1734..14f1c5a0b2d2 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -85,7 +85,8 @@ struct compressed_bio { static inline int compressed_bio_size(struct btrfs_root *root, unsigned long disk_size) { - u16 csum_size = btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); + return sizeof(struct compressed_bio) + ((disk_size + root->sectorsize - 1) / root->sectorsize) * csum_size; diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 011cab3aca8d..0fe615e4ea38 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -902,9 +902,10 @@ static noinline int balance_level(struct btrfs_trans_handle *trans, orig_ptr = btrfs_node_blockptr(mid, orig_slot); - if (level < BTRFS_MAX_LEVEL - 1) + if (level < BTRFS_MAX_LEVEL - 1) { parent = path->nodes[level + 1]; - pslot = path->slots[level + 1]; + pslot = path->slots[level + 1]; + } /* * deal with the case where there is only one pointer in the root @@ -1107,9 +1108,10 @@ static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans, mid = path->nodes[level]; WARN_ON(btrfs_header_generation(mid) != trans->transid); - if (level < BTRFS_MAX_LEVEL - 1) + if (level < BTRFS_MAX_LEVEL - 1) { parent = path->nodes[level + 1]; - pslot = path->slots[level + 1]; + pslot = path->slots[level + 1]; + } if (!parent) return 1; diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 370af767440d..b9ba59ff9292 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -30,6 +30,7 @@ #include <linux/kobject.h> #include <trace/events/btrfs.h> #include <asm/kmap_types.h> +#include <linux/pagemap.h> #include "extent_io.h" #include "extent_map.h" #include "async-thread.h" @@ -360,6 +361,47 @@ struct btrfs_header { #define BTRFS_LABEL_SIZE 256 /* + * just in case we somehow lose the roots and are not able to mount, + * we store an array of the roots from previous transactions + * in the super. + */ +#define BTRFS_NUM_BACKUP_ROOTS 4 +struct btrfs_root_backup { + __le64 tree_root; + __le64 tree_root_gen; + + __le64 chunk_root; + __le64 chunk_root_gen; + + __le64 extent_root; + __le64 extent_root_gen; + + __le64 fs_root; + __le64 fs_root_gen; + + __le64 dev_root; + __le64 dev_root_gen; + + __le64 csum_root; + __le64 csum_root_gen; + + __le64 total_bytes; + __le64 bytes_used; + __le64 num_devices; + /* future */ + __le64 unsed_64[4]; + + u8 tree_root_level; + u8 chunk_root_level; + u8 extent_root_level; + u8 fs_root_level; + u8 dev_root_level; + u8 csum_root_level; + /* future and to align */ + u8 unused_8[10]; +} __attribute__ ((__packed__)); + +/* * the super block basically lists the main trees of the FS * it currently lacks any block count etc etc */ @@ -405,6 +447,7 @@ struct btrfs_super_block { /* future expansion */ __le64 reserved[31]; u8 sys_chunk_array[BTRFS_SYSTEM_CHUNK_ARRAY_SIZE]; + struct btrfs_root_backup super_roots[BTRFS_NUM_BACKUP_ROOTS]; } __attribute__ ((__packed__)); /* @@ -772,14 +815,8 @@ struct btrfs_space_info { struct btrfs_block_rsv { u64 size; u64 reserved; - u64 freed[2]; struct btrfs_space_info *space_info; - struct list_head list; spinlock_t lock; - atomic_t usage; - unsigned int priority:8; - unsigned int durable:1; - unsigned int refill_used:1; unsigned int full:1; }; @@ -840,10 +877,10 @@ struct btrfs_block_group_cache { spinlock_t lock; u64 pinned; u64 reserved; - u64 reserved_pinned; u64 bytes_super; u64 flags; u64 sectorsize; + u64 cache_generation; unsigned int ro:1; unsigned int dirty:1; unsigned int iref:1; @@ -899,6 +936,10 @@ struct btrfs_fs_info { spinlock_t block_group_cache_lock; struct rb_root block_group_cache_tree; + /* keep track of unallocated space */ + spinlock_t free_chunk_lock; + u64 free_chunk_space; + struct extent_io_tree freed_extents[2]; struct extent_io_tree *pinned_extents; @@ -916,14 +957,11 @@ struct btrfs_fs_info { struct btrfs_block_rsv trans_block_rsv; /* block reservation for chunk tree */ struct btrfs_block_rsv chunk_block_rsv; + /* block reservation for delayed operations */ + struct btrfs_block_rsv delayed_block_rsv; struct btrfs_block_rsv empty_block_rsv; - /* list of block reservations that cross multiple transactions */ - struct list_head durable_block_rsv_list; - - struct mutex durable_block_rsv_mutex; - u64 generation; u64 last_trans_committed; @@ -942,8 +980,8 @@ struct btrfs_fs_info { wait_queue_head_t transaction_blocked_wait; wait_queue_head_t async_submit_wait; - struct btrfs_super_block super_copy; - struct btrfs_super_block super_for_commit; + struct btrfs_super_block *super_copy; + struct btrfs_super_block *super_for_commit; struct block_device *__bdev; struct super_block *sb; struct inode *btree_inode; @@ -1124,6 +1162,9 @@ struct btrfs_fs_info { /* readahead tree */ spinlock_t reada_lock; struct radix_tree_root reada_tree; + + /* next backup root to be overwritten */ + int backup_root_index; }; /* @@ -1368,6 +1409,7 @@ struct btrfs_ioctl_defrag_range_args { #define BTRFS_MOUNT_ENOSPC_DEBUG (1 << 15) #define BTRFS_MOUNT_AUTO_DEFRAG (1 << 16) #define BTRFS_MOUNT_INODE_MAP_CACHE (1 << 17) +#define BTRFS_MOUNT_RECOVERY (1 << 18) #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) @@ -1983,6 +2025,55 @@ static inline bool btrfs_root_readonly(struct btrfs_root *root) return root->root_item.flags & BTRFS_ROOT_SUBVOL_RDONLY; } +/* struct btrfs_root_backup */ +BTRFS_SETGET_STACK_FUNCS(backup_tree_root, struct btrfs_root_backup, + tree_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_tree_root_gen, struct btrfs_root_backup, + tree_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_tree_root_level, struct btrfs_root_backup, + tree_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_chunk_root, struct btrfs_root_backup, + chunk_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_chunk_root_gen, struct btrfs_root_backup, + chunk_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_chunk_root_level, struct btrfs_root_backup, + chunk_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_extent_root, struct btrfs_root_backup, + extent_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_extent_root_gen, struct btrfs_root_backup, + extent_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_extent_root_level, struct btrfs_root_backup, + extent_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_fs_root, struct btrfs_root_backup, + fs_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_fs_root_gen, struct btrfs_root_backup, + fs_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_fs_root_level, struct btrfs_root_backup, + fs_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_dev_root, struct btrfs_root_backup, + dev_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_dev_root_gen, struct btrfs_root_backup, + dev_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_dev_root_level, struct btrfs_root_backup, + dev_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_csum_root, struct btrfs_root_backup, + csum_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_csum_root_gen, struct btrfs_root_backup, + csum_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_csum_root_level, struct btrfs_root_backup, + csum_root_level, 8); +BTRFS_SETGET_STACK_FUNCS(backup_total_bytes, struct btrfs_root_backup, + total_bytes, 64); +BTRFS_SETGET_STACK_FUNCS(backup_bytes_used, struct btrfs_root_backup, + bytes_used, 64); +BTRFS_SETGET_STACK_FUNCS(backup_num_devices, struct btrfs_root_backup, + num_devices, 64); + /* struct btrfs_super_block */ BTRFS_SETGET_STACK_FUNCS(super_bytenr, struct btrfs_super_block, bytenr, 64); @@ -2134,6 +2225,11 @@ static inline bool btrfs_mixed_space_info(struct btrfs_space_info *space_info) (space_info->flags & BTRFS_BLOCK_GROUP_DATA)); } +static inline gfp_t btrfs_alloc_write_mask(struct address_space *mapping) +{ + return mapping_gfp_mask(mapping) & ~__GFP_FS; +} + /* extent-tree.c */ static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root, unsigned num_items) @@ -2142,6 +2238,17 @@ static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root, 3 * num_items; } +/* + * Doing a truncate won't result in new nodes or leaves, just what we need for + * COW. + */ +static inline u64 btrfs_calc_trunc_metadata_size(struct btrfs_root *root, + unsigned num_items) +{ + return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) * + num_items; +} + void btrfs_put_block_group(struct btrfs_block_group_cache *cache); int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_root *root, unsigned long count); @@ -2151,6 +2258,9 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans, u64 num_bytes, u64 *refs, u64 *flags); int btrfs_pin_extent(struct btrfs_root *root, u64 bytenr, u64 num, int reserved); +int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + u64 bytenr, u64 num_bytes); int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 objectid, u64 offset, u64 bytenr); @@ -2201,8 +2311,8 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, u64 root_objectid, u64 owner, u64 offset); int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len); -int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, - u64 num_bytes, int reserve, int sinfo); +int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, + u64 start, u64 len); int btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, @@ -2245,25 +2355,23 @@ void btrfs_init_block_rsv(struct btrfs_block_rsv *rsv); struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_root *root); void btrfs_free_block_rsv(struct btrfs_root *root, struct btrfs_block_rsv *rsv); -void btrfs_add_durable_block_rsv(struct btrfs_fs_info *fs_info, - struct btrfs_block_rsv *rsv); -int btrfs_block_rsv_add(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +int btrfs_block_rsv_add(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 num_bytes); -int btrfs_block_rsv_check(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +int btrfs_block_rsv_add_noflush(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, + u64 num_bytes); +int btrfs_block_rsv_check(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, int min_factor); +int btrfs_block_rsv_refill(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, - u64 min_reserved, int min_factor); + u64 min_reserved); int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv, struct btrfs_block_rsv *dst_rsv, u64 num_bytes); void btrfs_block_rsv_release(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 num_bytes); -int btrfs_truncate_reserve_metadata(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_block_rsv *rsv); int btrfs_set_block_group_ro(struct btrfs_root *root, struct btrfs_block_group_cache *cache); int btrfs_set_block_group_rw(struct btrfs_root *root, @@ -2384,6 +2492,18 @@ static inline int btrfs_fs_closing(struct btrfs_fs_info *fs_info) smp_mb(); return fs_info->closing; } +static inline void free_fs_info(struct btrfs_fs_info *fs_info) +{ + kfree(fs_info->delayed_root); + kfree(fs_info->extent_root); + kfree(fs_info->tree_root); + kfree(fs_info->chunk_root); + kfree(fs_info->dev_root); + kfree(fs_info->csum_root); + kfree(fs_info->super_copy); + kfree(fs_info->super_for_commit); + kfree(fs_info); +} /* root-item.c */ int btrfs_find_root_ref(struct btrfs_root *tree_root, @@ -2584,11 +2704,6 @@ int btrfs_update_inode(struct btrfs_trans_handle *trans, int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode); int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode); int btrfs_orphan_cleanup(struct btrfs_root *root); -void btrfs_orphan_pre_snapshot(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending, - u64 *bytes_to_reserve); -void btrfs_orphan_post_snapshot(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending); void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size); diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index b52c672f4c18..bbe8496d5339 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -591,7 +591,7 @@ static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans, return 0; src_rsv = trans->block_rsv; - dst_rsv = &root->fs_info->global_block_rsv; + dst_rsv = &root->fs_info->delayed_block_rsv; num_bytes = btrfs_calc_trans_metadata_size(root, 1); ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes); @@ -609,7 +609,7 @@ static void btrfs_delayed_item_release_metadata(struct btrfs_root *root, if (!item->bytes_reserved) return; - rsv = &root->fs_info->global_block_rsv; + rsv = &root->fs_info->delayed_block_rsv; btrfs_block_rsv_release(root, rsv, item->bytes_reserved); } @@ -624,13 +624,36 @@ static int btrfs_delayed_inode_reserve_metadata( u64 num_bytes; int ret; - if (!trans->bytes_reserved) - return 0; - src_rsv = trans->block_rsv; - dst_rsv = &root->fs_info->global_block_rsv; + dst_rsv = &root->fs_info->delayed_block_rsv; num_bytes = btrfs_calc_trans_metadata_size(root, 1); + + /* + * btrfs_dirty_inode will update the inode under btrfs_join_transaction + * which doesn't reserve space for speed. This is a problem since we + * still need to reserve space for this update, so try to reserve the + * space. + * + * Now if src_rsv == delalloc_block_rsv we'll let it just steal since + * we're accounted for. + */ + if (!trans->bytes_reserved && + src_rsv != &root->fs_info->delalloc_block_rsv) { + ret = btrfs_block_rsv_add_noflush(root, dst_rsv, num_bytes); + /* + * Since we're under a transaction reserve_metadata_bytes could + * try to commit the transaction which will make it return + * EAGAIN to make us stop the transaction we have, so return + * ENOSPC instead so that btrfs_dirty_inode knows what to do. + */ + if (ret == -EAGAIN) + ret = -ENOSPC; + if (!ret) + node->bytes_reserved = num_bytes; + return ret; + } + ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes); if (!ret) node->bytes_reserved = num_bytes; @@ -646,7 +669,7 @@ static void btrfs_delayed_inode_release_metadata(struct btrfs_root *root, if (!node->bytes_reserved) return; - rsv = &root->fs_info->global_block_rsv; + rsv = &root->fs_info->delayed_block_rsv; btrfs_block_rsv_release(root, rsv, node->bytes_reserved); node->bytes_reserved = 0; @@ -1026,7 +1049,7 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans, path->leave_spinning = 1; block_rsv = trans->block_rsv; - trans->block_rsv = &root->fs_info->global_block_rsv; + trans->block_rsv = &root->fs_info->delayed_block_rsv; delayed_root = btrfs_get_delayed_root(root); @@ -1069,7 +1092,7 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, path->leave_spinning = 1; block_rsv = trans->block_rsv; - trans->block_rsv = &node->root->fs_info->global_block_rsv; + trans->block_rsv = &node->root->fs_info->delayed_block_rsv; ret = btrfs_insert_delayed_items(trans, path, node->root, node); if (!ret) @@ -1149,7 +1172,7 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work) goto free_path; block_rsv = trans->block_rsv; - trans->block_rsv = &root->fs_info->global_block_rsv; + trans->block_rsv = &root->fs_info->delayed_block_rsv; ret = btrfs_insert_delayed_items(trans, path, root, delayed_node); if (!ret) @@ -1686,11 +1709,8 @@ int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans, } ret = btrfs_delayed_inode_reserve_metadata(trans, root, delayed_node); - /* - * we must reserve enough space when we start a new transaction, - * so reserving metadata failure is impossible - */ - BUG_ON(ret); + if (ret) + goto release_node; fill_stack_inode_item(trans, &delayed_node->inode_item, inode); delayed_node->inode_dirty = 1; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 2151828aa142..cedfbfb278eb 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -256,8 +256,7 @@ void btrfs_csum_final(u32 crc, char *result) static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf, int verify) { - u16 csum_size = - btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); char *result = NULL; unsigned long len; unsigned long cur_len; @@ -1204,10 +1203,12 @@ static int find_and_setup_root(struct btrfs_root *tree_root, generation = btrfs_root_generation(&root->root_item); blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item)); + root->commit_root = NULL; root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item), blocksize, generation); if (!root->node || !btrfs_buffer_uptodate(root->node, generation)) { free_extent_buffer(root->node); + root->node = NULL; return -EIO; } root->commit_root = btrfs_root_node(root); @@ -1646,6 +1647,228 @@ sleep: return 0; } +/* + * this will find the highest generation in the array of + * root backups. The index of the highest array is returned, + * or -1 if we can't find anything. + * + * We check to make sure the array is valid by comparing the + * generation of the latest root in the array with the generation + * in the super block. If they don't match we pitch it. + */ +static int find_newest_super_backup(struct btrfs_fs_info *info, u64 newest_gen) +{ + u64 cur; + int newest_index = -1; + struct btrfs_root_backup *root_backup; + int i; + + for (i = 0; i < BTRFS_NUM_BACKUP_ROOTS; i++) { + root_backup = info->super_copy->super_roots + i; + cur = btrfs_backup_tree_root_gen(root_backup); + if (cur == newest_gen) + newest_index = i; + } + + /* check to see if we actually wrapped around */ + if (newest_index == BTRFS_NUM_BACKUP_ROOTS - 1) { + root_backup = info->super_copy->super_roots; + cur = btrfs_backup_tree_root_gen(root_backup); + if (cur == newest_gen) + newest_index = 0; + } + return newest_index; +} + + +/* + * find the oldest backup so we know where to store new entries + * in the backup array. This will set the backup_root_index + * field in the fs_info struct + */ +static void find_oldest_super_backup(struct btrfs_fs_info *info, + u64 newest_gen) +{ + int newest_index = -1; + + newest_index = find_newest_super_backup(info, newest_gen); + /* if there was garbage in there, just move along */ + if (newest_index == -1) { + info->backup_root_index = 0; + } else { + info->backup_root_index = (newest_index + 1) % BTRFS_NUM_BACKUP_ROOTS; + } +} + +/* + * copy all the root pointers into the super backup array. + * this will bump the backup pointer by one when it is + * done + */ +static void backup_super_roots(struct btrfs_fs_info *info) +{ + int next_backup; + struct btrfs_root_backup *root_backup; + int last_backup; + + next_backup = info->backup_root_index; + last_backup = (next_backup + BTRFS_NUM_BACKUP_ROOTS - 1) % + BTRFS_NUM_BACKUP_ROOTS; + + /* + * just overwrite the last backup if we're at the same generation + * this happens only at umount + */ + root_backup = info->super_for_commit->super_roots + last_backup; + if (btrfs_backup_tree_root_gen(root_backup) == + btrfs_header_generation(info->tree_root->node)) + next_backup = last_backup; + + root_backup = info->super_for_commit->super_roots + next_backup; + + /* + * make sure all of our padding and empty slots get zero filled + * regardless of which ones we use today + */ + memset(root_backup, 0, sizeof(*root_backup)); + + info->backup_root_index = (next_backup + 1) % BTRFS_NUM_BACKUP_ROOTS; + + btrfs_set_backup_tree_root(root_backup, info->tree_root->node->start); + btrfs_set_backup_tree_root_gen(root_backup, + btrfs_header_generation(info->tree_root->node)); + + btrfs_set_backup_tree_root_level(root_backup, + btrfs_header_level(info->tree_root->node)); + + btrfs_set_backup_chunk_root(root_backup, info->chunk_root->node->start); + btrfs_set_backup_chunk_root_gen(root_backup, + btrfs_header_generation(info->chunk_root->node)); + btrfs_set_backup_chunk_root_level(root_backup, + btrfs_header_level(info->chunk_root->node)); + + btrfs_set_backup_extent_root(root_backup, info->extent_root->node->start); + btrfs_set_backup_extent_root_gen(root_backup, + btrfs_header_generation(info->extent_root->node)); + btrfs_set_backup_extent_root_level(root_backup, + btrfs_header_level(info->extent_root->node)); + + btrfs_set_backup_fs_root(root_backup, info->fs_root->node->start); + btrfs_set_backup_fs_root_gen(root_backup, + btrfs_header_generation(info->fs_root->node)); + btrfs_set_backup_fs_root_level(root_backup, + btrfs_header_level(info->fs_root->node)); + + btrfs_set_backup_dev_root(root_backup, info->dev_root->node->start); + btrfs_set_backup_dev_root_gen(root_backup, + btrfs_header_generation(info->dev_root->node)); + btrfs_set_backup_dev_root_level(root_backup, + btrfs_header_level(info->dev_root->node)); + + btrfs_set_backup_csum_root(root_backup, info->csum_root->node->start); + btrfs_set_backup_csum_root_gen(root_backup, + btrfs_header_generation(info->csum_root->node)); + btrfs_set_backup_csum_root_level(root_backup, + btrfs_header_level(info->csum_root->node)); + + btrfs_set_backup_total_bytes(root_backup, + btrfs_super_total_bytes(info->super_copy)); + btrfs_set_backup_bytes_used(root_backup, + btrfs_super_bytes_used(info->super_copy)); + btrfs_set_backup_num_devices(root_backup, + btrfs_super_num_devices(info->super_copy)); + + /* + * if we don't copy this out to the super_copy, it won't get remembered + * for the next commit + */ + memcpy(&info->super_copy->super_roots, + &info->super_for_commit->super_roots, + sizeof(*root_backup) * BTRFS_NUM_BACKUP_ROOTS); +} + +/* + * this copies info out of the root backup array and back into + * the in-memory super block. It is meant to help iterate through + * the array, so you send it the number of backups you've already + * tried and the last backup index you used. + * + * this returns -1 when it has tried all the backups + */ +static noinline int next_root_backup(struct btrfs_fs_info *info, + struct btrfs_super_block *super, + int *num_backups_tried, int *backup_index) +{ + struct btrfs_root_backup *root_backup; + int newest = *backup_index; + + if (*num_backups_tried == 0) { + u64 gen = btrfs_super_generation(super); + + newest = find_newest_super_backup(info, gen); + if (newest == -1) + return -1; + + *backup_index = newest; + *num_backups_tried = 1; + } else if (*num_backups_tried == BTRFS_NUM_BACKUP_ROOTS) { + /* we've tried all the backups, all done */ + return -1; + } else { + /* jump to the next oldest backup */ + newest = (*backup_index + BTRFS_NUM_BACKUP_ROOTS - 1) % + BTRFS_NUM_BACKUP_ROOTS; + *backup_index = newest; + *num_backups_tried += 1; + } + root_backup = super->super_roots + newest; + + btrfs_set_super_generation(super, + btrfs_backup_tree_root_gen(root_backup)); + btrfs_set_super_root(super, btrfs_backup_tree_root(root_backup)); + btrfs_set_super_root_level(super, + btrfs_backup_tree_root_level(root_backup)); + btrfs_set_super_bytes_used(super, btrfs_backup_bytes_used(root_backup)); + + /* + * fixme: the total bytes and num_devices need to match or we should + * need a fsck + */ + btrfs_set_super_total_bytes(super, btrfs_backup_total_bytes(root_backup)); + btrfs_set_super_num_devices(super, btrfs_backup_num_devices(root_backup)); + return 0; +} + +/* helper to cleanup tree roots */ +static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root) +{ + free_extent_buffer(info->tree_root->node); + free_extent_buffer(info->tree_root->commit_root); + free_extent_buffer(info->dev_root->node); + free_extent_buffer(info->dev_root->commit_root); + free_extent_buffer(info->extent_root->node); + free_extent_buffer(info->extent_root->commit_root); + free_extent_buffer(info->csum_root->node); + free_extent_buffer(info->csum_root->commit_root); + + info->tree_root->node = NULL; + info->tree_root->commit_root = NULL; + info->dev_root->node = NULL; + info->dev_root->commit_root = NULL; + info->extent_root->node = NULL; + info->extent_root->commit_root = NULL; + info->csum_root->node = NULL; + info->csum_root->commit_root = NULL; + + if (chunk_root) { + free_extent_buffer(info->chunk_root->node); + free_extent_buffer(info->chunk_root->commit_root); + info->chunk_root->node = NULL; + info->chunk_root->commit_root = NULL; + } +} + + struct btrfs_root *open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_devices, char *options) @@ -1673,6 +1896,8 @@ struct btrfs_root *open_ctree(struct super_block *sb, int ret; int err = -EINVAL; + int num_backups_tried = 0; + int backup_index = 0; struct btrfs_super_block *disk_super; @@ -1717,6 +1942,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, spin_lock_init(&fs_info->fs_roots_radix_lock); spin_lock_init(&fs_info->delayed_iput_lock); spin_lock_init(&fs_info->defrag_inodes_lock); + spin_lock_init(&fs_info->free_chunk_lock); mutex_init(&fs_info->reloc_mutex); init_completion(&fs_info->kobj_unregister); @@ -1734,8 +1960,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_init_block_rsv(&fs_info->trans_block_rsv); btrfs_init_block_rsv(&fs_info->chunk_block_rsv); btrfs_init_block_rsv(&fs_info->empty_block_rsv); - INIT_LIST_HEAD(&fs_info->durable_block_rsv_list); - mutex_init(&fs_info->durable_block_rsv_mutex); + btrfs_init_block_rsv(&fs_info->delayed_block_rsv); atomic_set(&fs_info->nr_async_submits, 0); atomic_set(&fs_info->async_delalloc_pages, 0); atomic_set(&fs_info->async_submit_draining, 0); @@ -1746,6 +1971,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, fs_info->metadata_ratio = 0; fs_info->defrag_inodes = RB_ROOT; fs_info->trans_no_join = 0; + fs_info->free_chunk_space = 0; /* readahead state */ INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT); @@ -1839,14 +2065,14 @@ struct btrfs_root *open_ctree(struct super_block *sb, goto fail_alloc; } - memcpy(&fs_info->super_copy, bh->b_data, sizeof(fs_info->super_copy)); - memcpy(&fs_info->super_for_commit, &fs_info->super_copy, - sizeof(fs_info->super_for_commit)); + memcpy(fs_info->super_copy, bh->b_data, sizeof(*fs_info->super_copy)); + memcpy(fs_info->super_for_commit, fs_info->super_copy, + sizeof(*fs_info->super_for_commit)); brelse(bh); - memcpy(fs_info->fsid, fs_info->super_copy.fsid, BTRFS_FSID_SIZE); + memcpy(fs_info->fsid, fs_info->super_copy->fsid, BTRFS_FSID_SIZE); - disk_super = &fs_info->super_copy; + disk_super = fs_info->super_copy; if (!btrfs_super_root(disk_super)) goto fail_alloc; @@ -1856,6 +2082,13 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY); /* + * run through our array of backup supers and setup + * our ring pointer to the oldest one + */ + generation = btrfs_super_generation(disk_super); + find_oldest_super_backup(fs_info, generation); + + /* * In the long term, we'll store the compression type in the super * block, and it'll be used for per file compression control. */ @@ -2017,7 +2250,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, if (!test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) { printk(KERN_WARNING "btrfs: failed to read chunk root on %s\n", sb->s_id); - goto fail_chunk_root; + goto fail_tree_roots; } btrfs_set_root_node(&chunk_root->root_item, chunk_root->node); chunk_root->commit_root = btrfs_root_node(chunk_root); @@ -2032,11 +2265,12 @@ struct btrfs_root *open_ctree(struct super_block *sb, if (ret) { printk(KERN_WARNING "btrfs: failed to read chunk tree on %s\n", sb->s_id); - goto fail_chunk_root; + goto fail_tree_roots; } btrfs_close_extra_devices(fs_devices); +retry_root_backup: blocksize = btrfs_level_size(tree_root, btrfs_super_root_level(disk_super)); generation = btrfs_super_generation(disk_super); @@ -2044,32 +2278,33 @@ struct btrfs_root *open_ctree(struct super_block *sb, tree_root->node = read_tree_block(tree_root, btrfs_super_root(disk_super), blocksize, generation); - if (!tree_root->node) - goto fail_chunk_root; - if (!test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) { + if (!tree_root->node || + !test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) { printk(KERN_WARNING "btrfs: failed to read tree root on %s\n", sb->s_id); - goto fail_tree_root; + + goto recovery_tree_root; } + btrfs_set_root_node(&tree_root->root_item, tree_root->node); tree_root->commit_root = btrfs_root_node(tree_root); ret = find_and_setup_root(tree_root, fs_info, BTRFS_EXTENT_TREE_OBJECTID, extent_root); if (ret) - goto fail_tree_root; + goto recovery_tree_root; extent_root->track_dirty = 1; ret = find_and_setup_root(tree_root, fs_info, BTRFS_DEV_TREE_OBJECTID, dev_root); if (ret) - goto fail_extent_root; + goto recovery_tree_root; dev_root->track_dirty = 1; ret = find_and_setup_root(tree_root, fs_info, BTRFS_CSUM_TREE_OBJECTID, csum_root); if (ret) - goto fail_dev_root; + goto recovery_tree_root; csum_root->track_dirty = 1; @@ -2202,20 +2437,10 @@ fail_cleaner: fail_block_groups: btrfs_free_block_groups(fs_info); - free_extent_buffer(csum_root->node); - free_extent_buffer(csum_root->commit_root); -fail_dev_root: - free_extent_buffer(dev_root->node); - free_extent_buffer(dev_root->commit_root); -fail_extent_root: - free_extent_buffer(extent_root->node); - free_extent_buffer(extent_root->commit_root); -fail_tree_root: - free_extent_buffer(tree_root->node); - free_extent_buffer(tree_root->commit_root); -fail_chunk_root: - free_extent_buffer(chunk_root->node); - free_extent_buffer(chunk_root->commit_root); + +fail_tree_roots: + free_root_pointers(fs_info, 1); + fail_sb_buffer: btrfs_stop_workers(&fs_info->generic_worker); btrfs_stop_workers(&fs_info->fixup_workers); @@ -2230,7 +2455,6 @@ fail_sb_buffer: btrfs_stop_workers(&fs_info->delayed_workers); btrfs_stop_workers(&fs_info->caching_workers); fail_alloc: - kfree(fs_info->delayed_root); fail_iput: invalidate_inode_pages2(fs_info->btree_inode->i_mapping); iput(fs_info->btree_inode); @@ -2242,13 +2466,27 @@ fail_bdi: fail_srcu: cleanup_srcu_struct(&fs_info->subvol_srcu); fail: - kfree(extent_root); - kfree(tree_root); - kfree(fs_info); - kfree(chunk_root); - kfree(dev_root); - kfree(csum_root); + free_fs_info(fs_info); return ERR_PTR(err); + +recovery_tree_root: + + if (!btrfs_test_opt(tree_root, RECOVERY)) + goto fail_tree_roots; + + free_root_pointers(fs_info, 0); + + /* don't use the log in recovery mode, it won't be valid */ + btrfs_set_super_log_root(disk_super, 0); + + /* we can't trust the free space cache either */ + btrfs_set_opt(fs_info->mount_opt, CLEAR_CACHE); + + ret = next_root_backup(fs_info, fs_info->super_copy, + &num_backups_tried, &backup_index); + if (ret == -1) + goto fail_block_groups; + goto retry_root_backup; } static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate) @@ -2416,10 +2654,11 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors) int total_errors = 0; u64 flags; - max_errors = btrfs_super_num_devices(&root->fs_info->super_copy) - 1; + max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1; do_barriers = !btrfs_test_opt(root, NOBARRIER); + backup_super_roots(root->fs_info); - sb = &root->fs_info->super_for_commit; + sb = root->fs_info->super_for_commit; dev_item = &sb->dev_item; mutex_lock(&root->fs_info->fs_devices->device_list_mutex); @@ -2623,8 +2862,6 @@ int close_ctree(struct btrfs_root *root) /* clear out the rbtree of defraggable inodes */ btrfs_run_defrag_inodes(root->fs_info); - btrfs_put_block_group_cache(fs_info); - /* * Here come 2 situations when btrfs is broken to flip readonly: * @@ -2650,6 +2887,8 @@ int close_ctree(struct btrfs_root *root) printk(KERN_ERR "btrfs: commit super ret %d\n", ret); } + btrfs_put_block_group_cache(fs_info); + kthread_stop(root->fs_info->transaction_kthread); kthread_stop(root->fs_info->cleaner_kthread); @@ -2681,7 +2920,6 @@ int close_ctree(struct btrfs_root *root) del_fs_roots(fs_info); iput(fs_info->btree_inode); - kfree(fs_info->delayed_root); btrfs_stop_workers(&fs_info->generic_worker); btrfs_stop_workers(&fs_info->fixup_workers); @@ -2703,12 +2941,7 @@ int close_ctree(struct btrfs_root *root) bdi_destroy(&fs_info->bdi); cleanup_srcu_struct(&fs_info->subvol_srcu); - kfree(fs_info->extent_root); - kfree(fs_info->tree_root); - kfree(fs_info->chunk_root); - kfree(fs_info->dev_root); - kfree(fs_info->csum_root); - kfree(fs_info); + free_fs_info(fs_info); return 0; } @@ -2814,7 +3047,8 @@ int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid) return ret; } -int btree_lock_page_hook(struct page *page) +static int btree_lock_page_hook(struct page *page, void *data, + void (*flush_fn)(void *)) { struct inode *inode = page->mapping->host; struct btrfs_root *root = BTRFS_I(inode)->root; @@ -2831,7 +3065,10 @@ int btree_lock_page_hook(struct page *page) if (!eb) goto out; - btrfs_tree_lock(eb); + if (!btrfs_try_tree_write_lock(eb)) { + flush_fn(data); + btrfs_tree_lock(eb); + } btrfs_set_header_flag(eb, BTRFS_HEADER_FLAG_WRITTEN); if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &eb->bflags)) { @@ -2846,7 +3083,10 @@ int btree_lock_page_hook(struct page *page) btrfs_tree_unlock(eb); free_extent_buffer(eb); out: - lock_page(page); + if (!trylock_page(page)) { + flush_fn(data); + lock_page(page); + } return 0; } diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index b3bdb5c1390f..c99d0a8f13fa 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -85,8 +85,6 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info); int btrfs_add_log_tree(struct btrfs_trans_handle *trans, struct btrfs_root *root); -int btree_lock_page_hook(struct page *page); - #ifdef CONFIG_DEBUG_LOCK_ALLOC void btrfs_init_lockdep(void); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index f5be06a2462f..23e936c3de76 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -23,6 +23,7 @@ #include <linux/rcupdate.h> #include <linux/kthread.h> #include <linux/slab.h> +#include <linux/ratelimit.h> #include "compat.h" #include "hash.h" #include "ctree.h" @@ -52,6 +53,21 @@ enum { CHUNK_ALLOC_LIMITED = 2, }; +/* + * Control how reservations are dealt with. + * + * RESERVE_FREE - freeing a reservation. + * RESERVE_ALLOC - allocating space and we need to update bytes_may_use for + * ENOSPC accounting + * RESERVE_ALLOC_NO_ACCOUNT - allocating space and we should not update + * bytes_may_use as the ENOSPC accounting is done elsewhere + */ +enum { + RESERVE_FREE = 0, + RESERVE_ALLOC = 1, + RESERVE_ALLOC_NO_ACCOUNT = 2, +}; + static int update_block_group(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, int alloc); @@ -81,6 +97,8 @@ static int find_next_key(struct btrfs_path *path, int level, struct btrfs_key *key); static void dump_space_info(struct btrfs_space_info *info, u64 bytes, int dump_block_groups); +static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, + u64 num_bytes, int reserve); static noinline int block_group_cache_done(struct btrfs_block_group_cache *cache) @@ -104,7 +122,6 @@ void btrfs_put_block_group(struct btrfs_block_group_cache *cache) if (atomic_dec_and_test(&cache->count)) { WARN_ON(cache->pinned > 0); WARN_ON(cache->reserved > 0); - WARN_ON(cache->reserved_pinned > 0); kfree(cache->free_space_ctl); kfree(cache); } @@ -465,7 +482,8 @@ static int cache_block_group(struct btrfs_block_group_cache *cache, * we likely hold important locks. */ if (trans && (!trans->transaction->in_commit) && - (root && root != root->fs_info->tree_root)) { + (root && root != root->fs_info->tree_root) && + btrfs_test_opt(root, SPACE_CACHE)) { spin_lock(&cache->lock); if (cache->cached != BTRFS_CACHE_NO) { spin_unlock(&cache->lock); @@ -2700,6 +2718,13 @@ again: goto again; } + /* We've already setup this transaction, go ahead and exit */ + if (block_group->cache_generation == trans->transid && + i_size_read(inode)) { + dcs = BTRFS_DC_SETUP; + goto out_put; + } + /* * We want to set the generation to 0, that way if anything goes wrong * from here on out we know not to trust this cache when we load up next @@ -2749,12 +2774,15 @@ again: if (!ret) dcs = BTRFS_DC_SETUP; btrfs_free_reserved_data_space(inode, num_pages); + out_put: iput(inode); out_free: btrfs_release_path(path); out: spin_lock(&block_group->lock); + if (!ret) + block_group->cache_generation = trans->transid; block_group->disk_cache_state = dcs; spin_unlock(&block_group->lock); @@ -3122,16 +3150,13 @@ commit_trans: return -ENOSPC; } data_sinfo->bytes_may_use += bytes; - BTRFS_I(inode)->reserved_bytes += bytes; spin_unlock(&data_sinfo->lock); return 0; } /* - * called when we are clearing an delalloc extent from the - * inode's io_tree or there was an error for whatever reason - * after calling btrfs_check_data_free_space + * Called if we need to clear a data reservation for this inode. */ void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes) { @@ -3144,7 +3169,6 @@ void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes) data_sinfo = BTRFS_I(inode)->space_info; spin_lock(&data_sinfo->lock); data_sinfo->bytes_may_use -= bytes; - BTRFS_I(inode)->reserved_bytes -= bytes; spin_unlock(&data_sinfo->lock); } @@ -3165,6 +3189,7 @@ static int should_alloc_chunk(struct btrfs_root *root, struct btrfs_space_info *sinfo, u64 alloc_bytes, int force) { + struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv; u64 num_bytes = sinfo->total_bytes - sinfo->bytes_readonly; u64 num_allocated = sinfo->bytes_used + sinfo->bytes_reserved; u64 thresh; @@ -3173,11 +3198,18 @@ static int should_alloc_chunk(struct btrfs_root *root, return 1; /* + * We need to take into account the global rsv because for all intents + * and purposes it's used space. Don't worry about locking the + * global_rsv, it doesn't change except when the transaction commits. + */ + num_allocated += global_rsv->size; + + /* * in limited mode, we want to have some free space up to * about 1% of the FS size. */ if (force == CHUNK_ALLOC_LIMITED) { - thresh = btrfs_super_total_bytes(&root->fs_info->super_copy); + thresh = btrfs_super_total_bytes(root->fs_info->super_copy); thresh = max_t(u64, 64 * 1024 * 1024, div_factor_fine(thresh, 1)); @@ -3199,7 +3231,7 @@ static int should_alloc_chunk(struct btrfs_root *root, if (num_allocated + alloc_bytes < div_factor(num_bytes, 8)) return 0; - thresh = btrfs_super_total_bytes(&root->fs_info->super_copy); + thresh = btrfs_super_total_bytes(root->fs_info->super_copy); /* 256MB or 5% of the FS */ thresh = max_t(u64, 256 * 1024 * 1024, div_factor_fine(thresh, 5)); @@ -3302,24 +3334,26 @@ out: /* * shrink metadata reservation for delalloc */ -static int shrink_delalloc(struct btrfs_trans_handle *trans, - struct btrfs_root *root, u64 to_reclaim, int sync) +static int shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, + bool wait_ordered) { struct btrfs_block_rsv *block_rsv; struct btrfs_space_info *space_info; + struct btrfs_trans_handle *trans; u64 reserved; u64 max_reclaim; u64 reclaimed = 0; long time_left; - int nr_pages = (2 * 1024 * 1024) >> PAGE_CACHE_SHIFT; + unsigned long nr_pages = (2 * 1024 * 1024) >> PAGE_CACHE_SHIFT; int loops = 0; unsigned long progress; + trans = (struct btrfs_trans_handle *)current->journal_info; block_rsv = &root->fs_info->delalloc_block_rsv; space_info = block_rsv->space_info; smp_mb(); - reserved = space_info->bytes_reserved; + reserved = space_info->bytes_may_use; progress = space_info->reservation_progress; if (reserved == 0) @@ -3334,7 +3368,8 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, } max_reclaim = min(reserved, to_reclaim); - + nr_pages = max_t(unsigned long, nr_pages, + max_reclaim >> PAGE_CACHE_SHIFT); while (loops < 1024) { /* have the flusher threads jump in and do some IO */ smp_mb(); @@ -3343,9 +3378,9 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, writeback_inodes_sb_nr_if_idle(root->fs_info->sb, nr_pages); spin_lock(&space_info->lock); - if (reserved > space_info->bytes_reserved) - reclaimed += reserved - space_info->bytes_reserved; - reserved = space_info->bytes_reserved; + if (reserved > space_info->bytes_may_use) + reclaimed += reserved - space_info->bytes_may_use; + reserved = space_info->bytes_may_use; spin_unlock(&space_info->lock); loops++; @@ -3356,11 +3391,15 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, if (trans && trans->transaction->blocked) return -EAGAIN; - time_left = schedule_timeout_interruptible(1); + if (wait_ordered && !trans) { + btrfs_wait_ordered_extents(root, 0, 0); + } else { + time_left = schedule_timeout_interruptible(1); - /* We were interrupted, exit */ - if (time_left) - break; + /* We were interrupted, exit */ + if (time_left) + break; + } /* we've kicked the IO a few times, if anything has been freed, * exit. There is no sense in looping here for a long time @@ -3375,34 +3414,90 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, } } - if (reclaimed >= to_reclaim && !trans) - btrfs_wait_ordered_extents(root, 0, 0); + return reclaimed >= to_reclaim; } -/* - * Retries tells us how many times we've called reserve_metadata_bytes. The - * idea is if this is the first call (retries == 0) then we will add to our - * reserved count if we can't make the allocation in order to hold our place - * while we go and try and free up space. That way for retries > 1 we don't try - * and add space, we just check to see if the amount of unused space is >= the - * total space, meaning that our reservation is valid. +/** + * maybe_commit_transaction - possibly commit the transaction if its ok to + * @root - the root we're allocating for + * @bytes - the number of bytes we want to reserve + * @force - force the commit * - * However if we don't intend to retry this reservation, pass -1 as retries so - * that it short circuits this logic. + * This will check to make sure that committing the transaction will actually + * get us somewhere and then commit the transaction if it does. Otherwise it + * will return -ENOSPC. */ -static int reserve_metadata_bytes(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +static int may_commit_transaction(struct btrfs_root *root, + struct btrfs_space_info *space_info, + u64 bytes, int force) +{ + struct btrfs_block_rsv *delayed_rsv = &root->fs_info->delayed_block_rsv; + struct btrfs_trans_handle *trans; + + trans = (struct btrfs_trans_handle *)current->journal_info; + if (trans) + return -EAGAIN; + + if (force) + goto commit; + + /* See if there is enough pinned space to make this reservation */ + spin_lock(&space_info->lock); + if (space_info->bytes_pinned >= bytes) { + spin_unlock(&space_info->lock); + goto commit; + } + spin_unlock(&space_info->lock); + + /* + * See if there is some space in the delayed insertion reservation for + * this reservation. + */ + if (space_info != delayed_rsv->space_info) + return -ENOSPC; + + spin_lock(&delayed_rsv->lock); + if (delayed_rsv->size < bytes) { + spin_unlock(&delayed_rsv->lock); + return -ENOSPC; + } + spin_unlock(&delayed_rsv->lock); + +commit: + trans = btrfs_join_transaction(root); + if (IS_ERR(trans)) + return -ENOSPC; + + return btrfs_commit_transaction(trans, root); +} + +/** + * reserve_metadata_bytes - try to reserve bytes from the block_rsv's space + * @root - the root we're allocating for + * @block_rsv - the block_rsv we're allocating for + * @orig_bytes - the number of bytes we want + * @flush - wether or not we can flush to make our reservation + * + * This will reserve orgi_bytes number of bytes from the space info associated + * with the block_rsv. If there is not enough space it will make an attempt to + * flush out space to make room. It will do this by flushing delalloc if + * possible or committing the transaction. If flush is 0 then no attempts to + * regain reservations will be made and this will fail if there is not enough + * space already. + */ +static int reserve_metadata_bytes(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 orig_bytes, int flush) { struct btrfs_space_info *space_info = block_rsv->space_info; - u64 unused; + u64 used; u64 num_bytes = orig_bytes; int retries = 0; int ret = 0; bool committed = false; bool flushing = false; + bool wait_ordered = false; again: ret = 0; @@ -3419,7 +3514,7 @@ again: * deadlock since we are waiting for the flusher to finish, but * hold the current transaction open. */ - if (trans) + if (current->journal_info) return -EAGAIN; ret = wait_event_interruptible(space_info->wait, !space_info->flush); @@ -3431,9 +3526,9 @@ again: } ret = -ENOSPC; - unused = space_info->bytes_used + space_info->bytes_reserved + - space_info->bytes_pinned + space_info->bytes_readonly + - space_info->bytes_may_use; + used = space_info->bytes_used + space_info->bytes_reserved + + space_info->bytes_pinned + space_info->bytes_readonly + + space_info->bytes_may_use; /* * The idea here is that we've not already over-reserved the block group @@ -3442,10 +3537,9 @@ again: * lets start flushing stuff first and then come back and try to make * our reservation. */ - if (unused <= space_info->total_bytes) { - unused = space_info->total_bytes - unused; - if (unused >= num_bytes) { - space_info->bytes_reserved += orig_bytes; + if (used <= space_info->total_bytes) { + if (used + orig_bytes <= space_info->total_bytes) { + space_info->bytes_may_use += orig_bytes; ret = 0; } else { /* @@ -3461,10 +3555,64 @@ again: * amount plus the amount of bytes that we need for this * reservation. */ - num_bytes = unused - space_info->total_bytes + + wait_ordered = true; + num_bytes = used - space_info->total_bytes + (orig_bytes * (retries + 1)); } + if (ret) { + u64 profile = btrfs_get_alloc_profile(root, 0); + u64 avail; + + /* + * If we have a lot of space that's pinned, don't bother doing + * the overcommit dance yet and just commit the transaction. + */ + avail = (space_info->total_bytes - space_info->bytes_used) * 8; + do_div(avail, 10); + if (space_info->bytes_pinned >= avail && flush && !committed) { + space_info->flush = 1; + flushing = true; + spin_unlock(&space_info->lock); + ret = may_commit_transaction(root, space_info, + orig_bytes, 1); + if (ret) + goto out; + committed = true; + goto again; + } + + spin_lock(&root->fs_info->free_chunk_lock); + avail = root->fs_info->free_chunk_space; + + /* + * If we have dup, raid1 or raid10 then only half of the free + * space is actually useable. + */ + if (profile & (BTRFS_BLOCK_GROUP_DUP | + BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID10)) + avail >>= 1; + + /* + * If we aren't flushing don't let us overcommit too much, say + * 1/8th of the space. If we can flush, let it overcommit up to + * 1/2 of the space. + */ + if (flush) + avail >>= 3; + else + avail >>= 1; + spin_unlock(&root->fs_info->free_chunk_lock); + + if (used + num_bytes < space_info->total_bytes + avail) { + space_info->bytes_may_use += orig_bytes; + ret = 0; + } else { + wait_ordered = true; + } + } + /* * Couldn't make our reservation, save our place so while we're trying * to reclaim space we can actually use it instead of somebody else @@ -3484,7 +3632,7 @@ again: * We do synchronous shrinking since we don't actually unreserve * metadata until after the IO is completed. */ - ret = shrink_delalloc(trans, root, num_bytes, 1); + ret = shrink_delalloc(root, num_bytes, wait_ordered); if (ret < 0) goto out; @@ -3496,35 +3644,17 @@ again: * so go back around and try again. */ if (retries < 2) { + wait_ordered = true; retries++; goto again; } - /* - * Not enough space to be reclaimed, don't bother committing the - * transaction. - */ - spin_lock(&space_info->lock); - if (space_info->bytes_pinned < orig_bytes) - ret = -ENOSPC; - spin_unlock(&space_info->lock); - if (ret) - goto out; - - ret = -EAGAIN; - if (trans) - goto out; - ret = -ENOSPC; if (committed) goto out; - trans = btrfs_join_transaction(root); - if (IS_ERR(trans)) - goto out; - ret = btrfs_commit_transaction(trans, root); + ret = may_commit_transaction(root, space_info, orig_bytes, 0); if (!ret) { - trans = NULL; committed = true; goto again; } @@ -3542,10 +3672,12 @@ out: static struct btrfs_block_rsv *get_block_rsv(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - struct btrfs_block_rsv *block_rsv; - if (root->ref_cows) + struct btrfs_block_rsv *block_rsv = NULL; + + if (root->ref_cows || root == root->fs_info->csum_root) block_rsv = trans->block_rsv; - else + + if (!block_rsv) block_rsv = root->block_rsv; if (!block_rsv) @@ -3616,7 +3748,7 @@ static void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv, } if (num_bytes) { spin_lock(&space_info->lock); - space_info->bytes_reserved -= num_bytes; + space_info->bytes_may_use -= num_bytes; space_info->reservation_progress++; spin_unlock(&space_info->lock); } @@ -3640,9 +3772,6 @@ void btrfs_init_block_rsv(struct btrfs_block_rsv *rsv) { memset(rsv, 0, sizeof(*rsv)); spin_lock_init(&rsv->lock); - atomic_set(&rsv->usage, 1); - rsv->priority = 6; - INIT_LIST_HEAD(&rsv->list); } struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_root *root) @@ -3663,38 +3792,38 @@ struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_root *root) void btrfs_free_block_rsv(struct btrfs_root *root, struct btrfs_block_rsv *rsv) { - if (rsv && atomic_dec_and_test(&rsv->usage)) { - btrfs_block_rsv_release(root, rsv, (u64)-1); - if (!rsv->durable) - kfree(rsv); - } + btrfs_block_rsv_release(root, rsv, (u64)-1); + kfree(rsv); } -/* - * make the block_rsv struct be able to capture freed space. - * the captured space will re-add to the the block_rsv struct - * after transaction commit - */ -void btrfs_add_durable_block_rsv(struct btrfs_fs_info *fs_info, - struct btrfs_block_rsv *block_rsv) +int btrfs_block_rsv_add(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, + u64 num_bytes) { - block_rsv->durable = 1; - mutex_lock(&fs_info->durable_block_rsv_mutex); - list_add_tail(&block_rsv->list, &fs_info->durable_block_rsv_list); - mutex_unlock(&fs_info->durable_block_rsv_mutex); + int ret; + + if (num_bytes == 0) + return 0; + + ret = reserve_metadata_bytes(root, block_rsv, num_bytes, 1); + if (!ret) { + block_rsv_add_bytes(block_rsv, num_bytes, 1); + return 0; + } + + return ret; } -int btrfs_block_rsv_add(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_block_rsv *block_rsv, - u64 num_bytes) +int btrfs_block_rsv_add_noflush(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, + u64 num_bytes) { int ret; if (num_bytes == 0) return 0; - ret = reserve_metadata_bytes(trans, root, block_rsv, num_bytes, 1); + ret = reserve_metadata_bytes(root, block_rsv, num_bytes, 0); if (!ret) { block_rsv_add_bytes(block_rsv, num_bytes, 1); return 0; @@ -3703,55 +3832,52 @@ int btrfs_block_rsv_add(struct btrfs_trans_handle *trans, return ret; } -int btrfs_block_rsv_check(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_block_rsv *block_rsv, - u64 min_reserved, int min_factor) +int btrfs_block_rsv_check(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, int min_factor) { u64 num_bytes = 0; - int commit_trans = 0; int ret = -ENOSPC; if (!block_rsv) return 0; spin_lock(&block_rsv->lock); - if (min_factor > 0) - num_bytes = div_factor(block_rsv->size, min_factor); - if (min_reserved > num_bytes) - num_bytes = min_reserved; + num_bytes = div_factor(block_rsv->size, min_factor); + if (block_rsv->reserved >= num_bytes) + ret = 0; + spin_unlock(&block_rsv->lock); - if (block_rsv->reserved >= num_bytes) { + return ret; +} + +int btrfs_block_rsv_refill(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, + u64 min_reserved) +{ + u64 num_bytes = 0; + int ret = -ENOSPC; + + if (!block_rsv) + return 0; + + spin_lock(&block_rsv->lock); + num_bytes = min_reserved; + if (block_rsv->reserved >= num_bytes) ret = 0; - } else { + else num_bytes -= block_rsv->reserved; - if (block_rsv->durable && - block_rsv->freed[0] + block_rsv->freed[1] >= num_bytes) - commit_trans = 1; - } spin_unlock(&block_rsv->lock); + if (!ret) return 0; - if (block_rsv->refill_used) { - ret = reserve_metadata_bytes(trans, root, block_rsv, - num_bytes, 0); - if (!ret) { - block_rsv_add_bytes(block_rsv, num_bytes, 0); - return 0; - } - } - - if (commit_trans) { - if (trans) - return -EAGAIN; - trans = btrfs_join_transaction(root); - BUG_ON(IS_ERR(trans)); - ret = btrfs_commit_transaction(trans, root); + ret = reserve_metadata_bytes(root, block_rsv, num_bytes, 1); + if (!ret) { + block_rsv_add_bytes(block_rsv, num_bytes, 0); return 0; } - return -ENOSPC; + return ret; } int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv, @@ -3783,7 +3909,7 @@ static u64 calc_global_metadata_size(struct btrfs_fs_info *fs_info) u64 num_bytes; u64 meta_used; u64 data_used; - int csum_size = btrfs_super_csum_size(&fs_info->super_copy); + int csum_size = btrfs_super_csum_size(fs_info->super_copy); sinfo = __find_space_info(fs_info, BTRFS_BLOCK_GROUP_DATA); spin_lock(&sinfo->lock); @@ -3827,12 +3953,12 @@ static void update_global_block_rsv(struct btrfs_fs_info *fs_info) if (sinfo->total_bytes > num_bytes) { num_bytes = sinfo->total_bytes - num_bytes; block_rsv->reserved += num_bytes; - sinfo->bytes_reserved += num_bytes; + sinfo->bytes_may_use += num_bytes; } if (block_rsv->reserved >= block_rsv->size) { num_bytes = block_rsv->reserved - block_rsv->size; - sinfo->bytes_reserved -= num_bytes; + sinfo->bytes_may_use -= num_bytes; sinfo->reservation_progress++; block_rsv->reserved = block_rsv->size; block_rsv->full = 1; @@ -3848,16 +3974,13 @@ static void init_global_block_rsv(struct btrfs_fs_info *fs_info) space_info = __find_space_info(fs_info, BTRFS_BLOCK_GROUP_SYSTEM); fs_info->chunk_block_rsv.space_info = space_info; - fs_info->chunk_block_rsv.priority = 10; space_info = __find_space_info(fs_info, BTRFS_BLOCK_GROUP_METADATA); fs_info->global_block_rsv.space_info = space_info; - fs_info->global_block_rsv.priority = 10; - fs_info->global_block_rsv.refill_used = 1; fs_info->delalloc_block_rsv.space_info = space_info; fs_info->trans_block_rsv.space_info = space_info; fs_info->empty_block_rsv.space_info = space_info; - fs_info->empty_block_rsv.priority = 10; + fs_info->delayed_block_rsv.space_info = space_info; fs_info->extent_root->block_rsv = &fs_info->global_block_rsv; fs_info->csum_root->block_rsv = &fs_info->global_block_rsv; @@ -3865,10 +3988,6 @@ static void init_global_block_rsv(struct btrfs_fs_info *fs_info) fs_info->tree_root->block_rsv = &fs_info->global_block_rsv; fs_info->chunk_root->block_rsv = &fs_info->chunk_block_rsv; - btrfs_add_durable_block_rsv(fs_info, &fs_info->global_block_rsv); - - btrfs_add_durable_block_rsv(fs_info, &fs_info->delalloc_block_rsv); - update_global_block_rsv(fs_info); } @@ -3881,37 +4000,8 @@ static void release_global_block_rsv(struct btrfs_fs_info *fs_info) WARN_ON(fs_info->trans_block_rsv.reserved > 0); WARN_ON(fs_info->chunk_block_rsv.size > 0); WARN_ON(fs_info->chunk_block_rsv.reserved > 0); -} - -int btrfs_truncate_reserve_metadata(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_block_rsv *rsv) -{ - struct btrfs_block_rsv *trans_rsv = &root->fs_info->trans_block_rsv; - u64 num_bytes; - int ret; - - /* - * Truncate should be freeing data, but give us 2 items just in case it - * needs to use some space. We may want to be smarter about this in the - * future. - */ - num_bytes = btrfs_calc_trans_metadata_size(root, 2); - - /* We already have enough bytes, just return */ - if (rsv->reserved >= num_bytes) - return 0; - - num_bytes -= rsv->reserved; - - /* - * You should have reserved enough space before hand to do this, so this - * should not fail. - */ - ret = block_rsv_migrate_bytes(trans_rsv, rsv, num_bytes); - BUG_ON(ret); - - return 0; + WARN_ON(fs_info->delayed_block_rsv.size > 0); + WARN_ON(fs_info->delayed_block_rsv.reserved > 0); } void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans, @@ -3920,9 +4010,7 @@ void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans, if (!trans->bytes_reserved) return; - BUG_ON(trans->block_rsv != &root->fs_info->trans_block_rsv); - btrfs_block_rsv_release(root, trans->block_rsv, - trans->bytes_reserved); + btrfs_block_rsv_release(root, trans->block_rsv, trans->bytes_reserved); trans->bytes_reserved = 0; } @@ -3964,11 +4052,19 @@ int btrfs_snap_reserve_metadata(struct btrfs_trans_handle *trans, return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes); } +/** + * drop_outstanding_extent - drop an outstanding extent + * @inode: the inode we're dropping the extent for + * + * This is called when we are freeing up an outstanding extent, either called + * after an error or after an extent is written. This will return the number of + * reserved extents that need to be freed. This must be called with + * BTRFS_I(inode)->lock held. + */ static unsigned drop_outstanding_extent(struct inode *inode) { unsigned dropped_extents = 0; - spin_lock(&BTRFS_I(inode)->lock); BUG_ON(!BTRFS_I(inode)->outstanding_extents); BTRFS_I(inode)->outstanding_extents--; @@ -3978,19 +4074,70 @@ static unsigned drop_outstanding_extent(struct inode *inode) */ if (BTRFS_I(inode)->outstanding_extents >= BTRFS_I(inode)->reserved_extents) - goto out; + return 0; dropped_extents = BTRFS_I(inode)->reserved_extents - BTRFS_I(inode)->outstanding_extents; BTRFS_I(inode)->reserved_extents -= dropped_extents; -out: - spin_unlock(&BTRFS_I(inode)->lock); return dropped_extents; } -static u64 calc_csum_metadata_size(struct inode *inode, u64 num_bytes) +/** + * calc_csum_metadata_size - return the amount of metada space that must be + * reserved/free'd for the given bytes. + * @inode: the inode we're manipulating + * @num_bytes: the number of bytes in question + * @reserve: 1 if we are reserving space, 0 if we are freeing space + * + * This adjusts the number of csum_bytes in the inode and then returns the + * correct amount of metadata that must either be reserved or freed. We + * calculate how many checksums we can fit into one leaf and then divide the + * number of bytes that will need to be checksumed by this value to figure out + * how many checksums will be required. If we are adding bytes then the number + * may go up and we will return the number of additional bytes that must be + * reserved. If it is going down we will return the number of bytes that must + * be freed. + * + * This must be called with BTRFS_I(inode)->lock held. + */ +static u64 calc_csum_metadata_size(struct inode *inode, u64 num_bytes, + int reserve) { - return num_bytes >>= 3; + struct btrfs_root *root = BTRFS_I(inode)->root; + u64 csum_size; + int num_csums_per_leaf; + int num_csums; + int old_csums; + + if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM && + BTRFS_I(inode)->csum_bytes == 0) + return 0; + + old_csums = (int)div64_u64(BTRFS_I(inode)->csum_bytes, root->sectorsize); + if (reserve) + BTRFS_I(inode)->csum_bytes += num_bytes; + else + BTRFS_I(inode)->csum_bytes -= num_bytes; + csum_size = BTRFS_LEAF_DATA_SIZE(root) - sizeof(struct btrfs_item); + num_csums_per_leaf = (int)div64_u64(csum_size, + sizeof(struct btrfs_csum_item) + + sizeof(struct btrfs_disk_key)); + num_csums = (int)div64_u64(BTRFS_I(inode)->csum_bytes, root->sectorsize); + num_csums = num_csums + num_csums_per_leaf - 1; + num_csums = num_csums / num_csums_per_leaf; + + old_csums = old_csums + num_csums_per_leaf - 1; + old_csums = old_csums / num_csums_per_leaf; + + /* No change, no need to reserve more */ + if (old_csums == num_csums) + return 0; + + if (reserve) + return btrfs_calc_trans_metadata_size(root, + num_csums - old_csums); + + return btrfs_calc_trans_metadata_size(root, old_csums - num_csums); } int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) @@ -3999,9 +4146,13 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) struct btrfs_block_rsv *block_rsv = &root->fs_info->delalloc_block_rsv; u64 to_reserve = 0; unsigned nr_extents = 0; + int flush = 1; int ret; - if (btrfs_transaction_in_commit(root->fs_info)) + if (btrfs_is_free_space_inode(root, inode)) + flush = 0; + + if (flush && btrfs_transaction_in_commit(root->fs_info)) schedule_timeout(1); num_bytes = ALIGN(num_bytes, root->sectorsize); @@ -4017,18 +4168,29 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) to_reserve = btrfs_calc_trans_metadata_size(root, nr_extents); } + to_reserve += calc_csum_metadata_size(inode, num_bytes, 1); spin_unlock(&BTRFS_I(inode)->lock); - to_reserve += calc_csum_metadata_size(inode, num_bytes); - ret = reserve_metadata_bytes(NULL, root, block_rsv, to_reserve, 1); + ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush); if (ret) { + u64 to_free = 0; unsigned dropped; + + spin_lock(&BTRFS_I(inode)->lock); + dropped = drop_outstanding_extent(inode); + to_free = calc_csum_metadata_size(inode, num_bytes, 0); + spin_unlock(&BTRFS_I(inode)->lock); + to_free += btrfs_calc_trans_metadata_size(root, dropped); + /* - * We don't need the return value since our reservation failed, - * we just need to clean up our counter. + * Somebody could have come in and twiddled with the + * reservation, so if we have to free more than we would have + * reserved from this reservation go ahead and release those + * bytes. */ - dropped = drop_outstanding_extent(inode); - WARN_ON(dropped > 1); + to_free -= to_reserve; + if (to_free) + btrfs_block_rsv_release(root, block_rsv, to_free); return ret; } @@ -4037,6 +4199,15 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) return 0; } +/** + * btrfs_delalloc_release_metadata - release a metadata reservation for an inode + * @inode: the inode to release the reservation for + * @num_bytes: the number of bytes we're releasing + * + * This will release the metadata reservation for an inode. This can be called + * once we complete IO for a given set of bytes to release their metadata + * reservations. + */ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) { struct btrfs_root *root = BTRFS_I(inode)->root; @@ -4044,9 +4215,11 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) unsigned dropped; num_bytes = ALIGN(num_bytes, root->sectorsize); + spin_lock(&BTRFS_I(inode)->lock); dropped = drop_outstanding_extent(inode); - to_free = calc_csum_metadata_size(inode, num_bytes); + to_free = calc_csum_metadata_size(inode, num_bytes, 0); + spin_unlock(&BTRFS_I(inode)->lock); if (dropped > 0) to_free += btrfs_calc_trans_metadata_size(root, dropped); @@ -4054,6 +4227,21 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) to_free); } +/** + * btrfs_delalloc_reserve_space - reserve data and metadata space for delalloc + * @inode: inode we're writing to + * @num_bytes: the number of bytes we want to allocate + * + * This will do the following things + * + * o reserve space in the data space info for num_bytes + * o reserve space in the metadata space info based on number of outstanding + * extents and how much csums will be needed + * o add to the inodes ->delalloc_bytes + * o add it to the fs_info's delalloc inodes list. + * + * This will return 0 for success and -ENOSPC if there is no space left. + */ int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes) { int ret; @@ -4071,6 +4259,19 @@ int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes) return 0; } +/** + * btrfs_delalloc_release_space - release data and metadata space for delalloc + * @inode: inode we're releasing space for + * @num_bytes: the number of bytes we want to free up + * + * This must be matched with a call to btrfs_delalloc_reserve_space. This is + * called in the case that we don't need the metadata AND data reservations + * anymore. So if there is an error or we insert an inline extent. + * + * This function will release the metadata space that was not used and will + * decrement ->delalloc_bytes and remove it from the fs_info delalloc_inodes + * list if there are no delalloc bytes left. + */ void btrfs_delalloc_release_space(struct inode *inode, u64 num_bytes) { btrfs_delalloc_release_metadata(inode, num_bytes); @@ -4090,12 +4291,12 @@ static int update_block_group(struct btrfs_trans_handle *trans, /* block accounting for super block */ spin_lock(&info->delalloc_lock); - old_val = btrfs_super_bytes_used(&info->super_copy); + old_val = btrfs_super_bytes_used(info->super_copy); if (alloc) old_val += num_bytes; else old_val -= num_bytes; - btrfs_set_super_bytes_used(&info->super_copy, old_val); + btrfs_set_super_bytes_used(info->super_copy, old_val); spin_unlock(&info->delalloc_lock); while (total) { @@ -4123,7 +4324,7 @@ static int update_block_group(struct btrfs_trans_handle *trans, spin_lock(&cache->space_info->lock); spin_lock(&cache->lock); - if (btrfs_super_cache_generation(&info->super_copy) != 0 && + if (btrfs_test_opt(root, SPACE_CACHE) && cache->disk_cache_state < BTRFS_DC_CLEAR) cache->disk_cache_state = BTRFS_DC_CLEAR; @@ -4135,7 +4336,6 @@ static int update_block_group(struct btrfs_trans_handle *trans, btrfs_set_block_group_used(&cache->item, old_val); cache->reserved -= num_bytes; cache->space_info->bytes_reserved -= num_bytes; - cache->space_info->reservation_progress++; cache->space_info->bytes_used += num_bytes; cache->space_info->disk_used += num_bytes * factor; spin_unlock(&cache->lock); @@ -4187,7 +4387,6 @@ static int pin_down_extent(struct btrfs_root *root, if (reserved) { cache->reserved -= num_bytes; cache->space_info->bytes_reserved -= num_bytes; - cache->space_info->reservation_progress++; } spin_unlock(&cache->lock); spin_unlock(&cache->space_info->lock); @@ -4215,45 +4414,82 @@ int btrfs_pin_extent(struct btrfs_root *root, } /* - * update size of reserved extents. this function may return -EAGAIN - * if 'reserve' is true or 'sinfo' is false. + * this function must be called within transaction + */ +int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + u64 bytenr, u64 num_bytes) +{ + struct btrfs_block_group_cache *cache; + + cache = btrfs_lookup_block_group(root->fs_info, bytenr); + BUG_ON(!cache); + + /* + * pull in the free space cache (if any) so that our pin + * removes the free space from the cache. We have load_only set + * to one because the slow code to read in the free extents does check + * the pinned extents. + */ + cache_block_group(cache, trans, root, 1); + + pin_down_extent(root, cache, bytenr, num_bytes, 0); + + /* remove us from the free space cache (if we're there at all) */ + btrfs_remove_free_space(cache, bytenr, num_bytes); + btrfs_put_block_group(cache); + return 0; +} + +/** + * btrfs_update_reserved_bytes - update the block_group and space info counters + * @cache: The cache we are manipulating + * @num_bytes: The number of bytes in question + * @reserve: One of the reservation enums + * + * This is called by the allocator when it reserves space, or by somebody who is + * freeing space that was never actually used on disk. For example if you + * reserve some space for a new leaf in transaction A and before transaction A + * commits you free that leaf, you call this with reserve set to 0 in order to + * clear the reservation. + * + * Metadata reservations should be called with RESERVE_ALLOC so we do the proper + * ENOSPC accounting. For data we handle the reservation through clearing the + * delalloc bits in the io_tree. We have to do this since we could end up + * allocating less disk space for the amount of data we have reserved in the + * case of compression. + * + * If this is a reservation and the block group has become read only we cannot + * make the reservation and return -EAGAIN, otherwise this function always + * succeeds. */ -int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, - u64 num_bytes, int reserve, int sinfo) +static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, + u64 num_bytes, int reserve) { + struct btrfs_space_info *space_info = cache->space_info; int ret = 0; - if (sinfo) { - struct btrfs_space_info *space_info = cache->space_info; - spin_lock(&space_info->lock); - spin_lock(&cache->lock); - if (reserve) { - if (cache->ro) { - ret = -EAGAIN; - } else { - cache->reserved += num_bytes; - space_info->bytes_reserved += num_bytes; - } - } else { - if (cache->ro) - space_info->bytes_readonly += num_bytes; - cache->reserved -= num_bytes; - space_info->bytes_reserved -= num_bytes; - space_info->reservation_progress++; - } - spin_unlock(&cache->lock); - spin_unlock(&space_info->lock); - } else { - spin_lock(&cache->lock); + spin_lock(&space_info->lock); + spin_lock(&cache->lock); + if (reserve != RESERVE_FREE) { if (cache->ro) { ret = -EAGAIN; } else { - if (reserve) - cache->reserved += num_bytes; - else - cache->reserved -= num_bytes; + cache->reserved += num_bytes; + space_info->bytes_reserved += num_bytes; + if (reserve == RESERVE_ALLOC) { + BUG_ON(space_info->bytes_may_use < num_bytes); + space_info->bytes_may_use -= num_bytes; + } } - spin_unlock(&cache->lock); + } else { + if (cache->ro) + space_info->bytes_readonly += num_bytes; + cache->reserved -= num_bytes; + space_info->bytes_reserved -= num_bytes; + space_info->reservation_progress++; } + spin_unlock(&cache->lock); + spin_unlock(&space_info->lock); return ret; } @@ -4319,13 +4555,8 @@ static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end) spin_lock(&cache->lock); cache->pinned -= len; cache->space_info->bytes_pinned -= len; - if (cache->ro) { + if (cache->ro) cache->space_info->bytes_readonly += len; - } else if (cache->reserved_pinned > 0) { - len = min(len, cache->reserved_pinned); - cache->reserved_pinned -= len; - cache->space_info->bytes_reserved += len; - } spin_unlock(&cache->lock); spin_unlock(&cache->space_info->lock); } @@ -4340,11 +4571,8 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, { struct btrfs_fs_info *fs_info = root->fs_info; struct extent_io_tree *unpin; - struct btrfs_block_rsv *block_rsv; - struct btrfs_block_rsv *next_rsv; u64 start; u64 end; - int idx; int ret; if (fs_info->pinned_extents == &fs_info->freed_extents[0]) @@ -4367,30 +4595,6 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, cond_resched(); } - mutex_lock(&fs_info->durable_block_rsv_mutex); - list_for_each_entry_safe(block_rsv, next_rsv, - &fs_info->durable_block_rsv_list, list) { - - idx = trans->transid & 0x1; - if (block_rsv->freed[idx] > 0) { - block_rsv_add_bytes(block_rsv, - block_rsv->freed[idx], 0); - block_rsv->freed[idx] = 0; - } - if (atomic_read(&block_rsv->usage) == 0) { - btrfs_block_rsv_release(root, block_rsv, (u64)-1); - - if (block_rsv->freed[0] == 0 && - block_rsv->freed[1] == 0) { - list_del_init(&block_rsv->list); - kfree(block_rsv); - } - } else { - btrfs_block_rsv_release(root, block_rsv, 0); - } - } - mutex_unlock(&fs_info->durable_block_rsv_mutex); - return 0; } @@ -4668,7 +4872,6 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, struct extent_buffer *buf, u64 parent, int last_ref) { - struct btrfs_block_rsv *block_rsv; struct btrfs_block_group_cache *cache = NULL; int ret; @@ -4683,64 +4886,24 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, if (!last_ref) return; - block_rsv = get_block_rsv(trans, root); cache = btrfs_lookup_block_group(root->fs_info, buf->start); - if (block_rsv->space_info != cache->space_info) - goto out; if (btrfs_header_generation(buf) == trans->transid) { if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) { ret = check_ref_cleanup(trans, root, buf->start); if (!ret) - goto pin; + goto out; } if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) { pin_down_extent(root, cache, buf->start, buf->len, 1); - goto pin; + goto out; } WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags)); btrfs_add_free_space(cache, buf->start, buf->len); - ret = btrfs_update_reserved_bytes(cache, buf->len, 0, 0); - if (ret == -EAGAIN) { - /* block group became read-only */ - btrfs_update_reserved_bytes(cache, buf->len, 0, 1); - goto out; - } - - ret = 1; - spin_lock(&block_rsv->lock); - if (block_rsv->reserved < block_rsv->size) { - block_rsv->reserved += buf->len; - ret = 0; - } - spin_unlock(&block_rsv->lock); - - if (ret) { - spin_lock(&cache->space_info->lock); - cache->space_info->bytes_reserved -= buf->len; - cache->space_info->reservation_progress++; - spin_unlock(&cache->space_info->lock); - } - goto out; - } -pin: - if (block_rsv->durable && !cache->ro) { - ret = 0; - spin_lock(&cache->lock); - if (!cache->ro) { - cache->reserved_pinned += buf->len; - ret = 1; - } - spin_unlock(&cache->lock); - - if (ret) { - spin_lock(&block_rsv->lock); - block_rsv->freed[trans->transid & 0x1] += buf->len; - spin_unlock(&block_rsv->lock); - } + btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE); } out: /* @@ -4883,10 +5046,13 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans, int last_ptr_loop = 0; int loop = 0; int index = 0; + int alloc_type = (data & BTRFS_BLOCK_GROUP_DATA) ? + RESERVE_ALLOC_NO_ACCOUNT : RESERVE_ALLOC; bool found_uncached_bg = false; bool failed_cluster_refill = false; bool failed_alloc = false; bool use_cluster = true; + bool have_caching_bg = false; u64 ideal_cache_percent = 0; u64 ideal_cache_offset = 0; @@ -4969,6 +5135,7 @@ ideal_cache: } } search: + have_caching_bg = false; down_read(&space_info->groups_sem); list_for_each_entry(block_group, &space_info->block_groups[index], list) { @@ -5177,6 +5344,8 @@ refill_cluster: failed_alloc = true; goto have_block_group; } else if (!offset) { + if (!cached) + have_caching_bg = true; goto loop; } checks: @@ -5202,8 +5371,8 @@ checks: search_start - offset); BUG_ON(offset > search_start); - ret = btrfs_update_reserved_bytes(block_group, num_bytes, 1, - (data & BTRFS_BLOCK_GROUP_DATA)); + ret = btrfs_update_reserved_bytes(block_group, num_bytes, + alloc_type); if (ret == -EAGAIN) { btrfs_add_free_space(block_group, offset, num_bytes); goto loop; @@ -5227,6 +5396,9 @@ loop: } up_read(&space_info->groups_sem); + if (!ins->objectid && loop >= LOOP_CACHING_WAIT && have_caching_bg) + goto search; + if (!ins->objectid && ++index < BTRFS_NR_RAID_TYPES) goto search; @@ -5325,7 +5497,8 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes, int index = 0; spin_lock(&info->lock); - printk(KERN_INFO "space_info has %llu free, is %sfull\n", + printk(KERN_INFO "space_info %llu has %llu free, is %sfull\n", + (unsigned long long)info->flags, (unsigned long long)(info->total_bytes - info->bytes_used - info->bytes_pinned - info->bytes_reserved - info->bytes_readonly), @@ -5411,7 +5584,8 @@ again: return ret; } -int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len) +static int __btrfs_free_reserved_extent(struct btrfs_root *root, + u64 start, u64 len, int pin) { struct btrfs_block_group_cache *cache; int ret = 0; @@ -5426,8 +5600,12 @@ int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len) if (btrfs_test_opt(root, DISCARD)) ret = btrfs_discard_extent(root, start, len, NULL); - btrfs_add_free_space(cache, start, len); - btrfs_update_reserved_bytes(cache, len, 0, 1); + if (pin) + pin_down_extent(root, cache, start, len, 1); + else { + btrfs_add_free_space(cache, start, len); + btrfs_update_reserved_bytes(cache, len, RESERVE_FREE); + } btrfs_put_block_group(cache); trace_btrfs_reserved_extent_free(root, start, len); @@ -5435,6 +5613,18 @@ int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len) return ret; } +int btrfs_free_reserved_extent(struct btrfs_root *root, + u64 start, u64 len) +{ + return __btrfs_free_reserved_extent(root, start, len, 0); +} + +int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, + u64 start, u64 len) +{ + return __btrfs_free_reserved_extent(root, start, len, 1); +} + static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 parent, u64 root_objectid, @@ -5630,7 +5820,8 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, put_caching_control(caching_ctl); } - ret = btrfs_update_reserved_bytes(block_group, ins->offset, 1, 1); + ret = btrfs_update_reserved_bytes(block_group, ins->offset, + RESERVE_ALLOC_NO_ACCOUNT); BUG_ON(ret); btrfs_put_block_group(block_group); ret = alloc_reserved_file_extent(trans, root, 0, root_objectid, @@ -5687,8 +5878,7 @@ use_block_rsv(struct btrfs_trans_handle *trans, block_rsv = get_block_rsv(trans, root); if (block_rsv->size == 0) { - ret = reserve_metadata_bytes(trans, root, block_rsv, - blocksize, 0); + ret = reserve_metadata_bytes(root, block_rsv, blocksize, 0); /* * If we couldn't reserve metadata bytes try and use some from * the global reserve. @@ -5708,13 +5898,15 @@ use_block_rsv(struct btrfs_trans_handle *trans, if (!ret) return block_rsv; if (ret) { - WARN_ON(1); - ret = reserve_metadata_bytes(trans, root, block_rsv, blocksize, - 0); + static DEFINE_RATELIMIT_STATE(_rs, + DEFAULT_RATELIMIT_INTERVAL, + /*DEFAULT_RATELIMIT_BURST*/ 2); + if (__ratelimit(&_rs)) { + printk(KERN_DEBUG "btrfs: block rsv returned %d\n", ret); + WARN_ON(1); + } + ret = reserve_metadata_bytes(root, block_rsv, blocksize, 0); if (!ret) { - spin_lock(&block_rsv->lock); - block_rsv->size += blocksize; - spin_unlock(&block_rsv->lock); return block_rsv; } else if (ret && block_rsv != global_rsv) { ret = block_rsv_use_bytes(global_rsv, blocksize); @@ -6592,12 +6784,9 @@ static int set_block_group_ro(struct btrfs_block_group_cache *cache, int force) cache->bytes_super - btrfs_block_group_used(&cache->item); if (sinfo->bytes_used + sinfo->bytes_reserved + sinfo->bytes_pinned + - sinfo->bytes_may_use + sinfo->bytes_readonly + - cache->reserved_pinned + num_bytes + min_allocable_bytes <= - sinfo->total_bytes) { + sinfo->bytes_may_use + sinfo->bytes_readonly + num_bytes + + min_allocable_bytes <= sinfo->total_bytes) { sinfo->bytes_readonly += num_bytes; - sinfo->bytes_reserved += cache->reserved_pinned; - cache->reserved_pinned = 0; cache->ro = 1; ret = 0; } @@ -6964,7 +7153,8 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info) struct btrfs_space_info, list); if (space_info->bytes_pinned > 0 || - space_info->bytes_reserved > 0) { + space_info->bytes_reserved > 0 || + space_info->bytes_may_use > 0) { WARN_ON(1); dump_space_info(space_info, 0, 0); } @@ -7006,14 +7196,12 @@ int btrfs_read_block_groups(struct btrfs_root *root) return -ENOMEM; path->reada = 1; - cache_gen = btrfs_super_cache_generation(&root->fs_info->super_copy); - if (cache_gen != 0 && - btrfs_super_generation(&root->fs_info->super_copy) != cache_gen) + cache_gen = btrfs_super_cache_generation(root->fs_info->super_copy); + if (btrfs_test_opt(root, SPACE_CACHE) && + btrfs_super_generation(root->fs_info->super_copy) != cache_gen) need_clear = 1; if (btrfs_test_opt(root, CLEAR_CACHE)) need_clear = 1; - if (!btrfs_test_opt(root, SPACE_CACHE) && cache_gen) - printk(KERN_INFO "btrfs: disk space caching is enabled\n"); while (1) { ret = find_first_block_group(root, path, &key); @@ -7252,7 +7440,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, goto out; } - inode = lookup_free_space_inode(root, block_group, path); + inode = lookup_free_space_inode(tree_root, block_group, path); if (!IS_ERR(inode)) { ret = btrfs_orphan_add(trans, inode); BUG_ON(ret); @@ -7268,7 +7456,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, spin_unlock(&block_group->lock); } /* One for our lookup ref */ - iput(inode); + btrfs_add_delayed_iput(inode); } key.objectid = BTRFS_FREE_SPACE_OBJECTID; @@ -7339,7 +7527,7 @@ int btrfs_init_space_info(struct btrfs_fs_info *fs_info) int mixed = 0; int ret; - disk_super = &fs_info->super_copy; + disk_super = fs_info->super_copy; if (!btrfs_super_root(disk_super)) return 1; diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index deba714236d7..c12705682c65 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -894,6 +894,194 @@ search_again: goto again; } +/** + * convert_extent - convert all bits in a given range from one bit to another + * @tree: the io tree to search + * @start: the start offset in bytes + * @end: the end offset in bytes (inclusive) + * @bits: the bits to set in this range + * @clear_bits: the bits to clear in this range + * @mask: the allocation mask + * + * This will go through and set bits for the given range. If any states exist + * already in this range they are set with the given bit and cleared of the + * clear_bits. This is only meant to be used by things that are mergeable, ie + * converting from say DELALLOC to DIRTY. This is not meant to be used with + * boundary bits like LOCK. + */ +int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, + int bits, int clear_bits, gfp_t mask) +{ + struct extent_state *state; + struct extent_state *prealloc = NULL; + struct rb_node *node; + int err = 0; + u64 last_start; + u64 last_end; + +again: + if (!prealloc && (mask & __GFP_WAIT)) { + prealloc = alloc_extent_state(mask); + if (!prealloc) + return -ENOMEM; + } + + spin_lock(&tree->lock); + /* + * this search will find all the extents that end after + * our range starts. + */ + node = tree_search(tree, start); + if (!node) { + prealloc = alloc_extent_state_atomic(prealloc); + if (!prealloc) + return -ENOMEM; + err = insert_state(tree, prealloc, start, end, &bits); + prealloc = NULL; + BUG_ON(err == -EEXIST); + goto out; + } + state = rb_entry(node, struct extent_state, rb_node); +hit_next: + last_start = state->start; + last_end = state->end; + + /* + * | ---- desired range ---- | + * | state | + * + * Just lock what we found and keep going + */ + if (state->start == start && state->end <= end) { + struct rb_node *next_node; + + set_state_bits(tree, state, &bits); + clear_state_bit(tree, state, &clear_bits, 0); + + merge_state(tree, state); + if (last_end == (u64)-1) + goto out; + + start = last_end + 1; + next_node = rb_next(&state->rb_node); + if (next_node && start < end && prealloc && !need_resched()) { + state = rb_entry(next_node, struct extent_state, + rb_node); + if (state->start == start) + goto hit_next; + } + goto search_again; + } + + /* + * | ---- desired range ---- | + * | state | + * or + * | ------------- state -------------- | + * + * We need to split the extent we found, and may flip bits on + * second half. + * + * If the extent we found extends past our + * range, we just split and search again. It'll get split + * again the next time though. + * + * If the extent we found is inside our range, we set the + * desired bit on it. + */ + if (state->start < start) { + prealloc = alloc_extent_state_atomic(prealloc); + if (!prealloc) + return -ENOMEM; + err = split_state(tree, state, prealloc, start); + BUG_ON(err == -EEXIST); + prealloc = NULL; + if (err) + goto out; + if (state->end <= end) { + set_state_bits(tree, state, &bits); + clear_state_bit(tree, state, &clear_bits, 0); + merge_state(tree, state); + if (last_end == (u64)-1) + goto out; + start = last_end + 1; + } + goto search_again; + } + /* + * | ---- desired range ---- | + * | state | or | state | + * + * There's a hole, we need to insert something in it and + * ignore the extent we found. + */ + if (state->start > start) { + u64 this_end; + if (end < last_start) + this_end = end; + else + this_end = last_start - 1; + + prealloc = alloc_extent_state_atomic(prealloc); + if (!prealloc) + return -ENOMEM; + + /* + * Avoid to free 'prealloc' if it can be merged with + * the later extent. + */ + err = insert_state(tree, prealloc, start, this_end, + &bits); + BUG_ON(err == -EEXIST); + if (err) { + free_extent_state(prealloc); + prealloc = NULL; + goto out; + } + prealloc = NULL; + start = this_end + 1; + goto search_again; + } + /* + * | ---- desired range ---- | + * | state | + * We need to split the extent, and set the bit + * on the first half + */ + if (state->start <= end && state->end > end) { + prealloc = alloc_extent_state_atomic(prealloc); + if (!prealloc) + return -ENOMEM; + + err = split_state(tree, state, prealloc, end + 1); + BUG_ON(err == -EEXIST); + + set_state_bits(tree, prealloc, &bits); + clear_state_bit(tree, prealloc, &clear_bits, 0); + + merge_state(tree, prealloc); + prealloc = NULL; + goto out; + } + + goto search_again; + +out: + spin_unlock(&tree->lock); + if (prealloc) + free_extent_state(prealloc); + + return err; + +search_again: + if (start > end) + goto out; + spin_unlock(&tree->lock); + if (mask & __GFP_WAIT) + cond_resched(); + goto again; +} + /* wrappers around set/clear extent bit */ int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask) @@ -919,7 +1107,7 @@ int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached_state, gfp_t mask) { return set_extent_bit(tree, start, end, - EXTENT_DELALLOC | EXTENT_DIRTY | EXTENT_UPTODATE, + EXTENT_DELALLOC | EXTENT_UPTODATE, 0, NULL, cached_state, mask); } @@ -2136,6 +2324,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, int compressed; int write_flags; unsigned long nr_written = 0; + bool fill_delalloc = true; if (wbc->sync_mode == WB_SYNC_ALL) write_flags = WRITE_SYNC; @@ -2145,6 +2334,9 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, trace___extent_writepage(page, inode, wbc); WARN_ON(!PageLocked(page)); + + ClearPageError(page); + pg_offset = i_size & (PAGE_CACHE_SIZE - 1); if (page->index > end_index || (page->index == end_index && !pg_offset)) { @@ -2166,10 +2358,13 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, set_page_extent_mapped(page); + if (!tree->ops || !tree->ops->fill_delalloc) + fill_delalloc = false; + delalloc_start = start; delalloc_end = 0; page_started = 0; - if (!epd->extent_locked) { + if (!epd->extent_locked && fill_delalloc) { u64 delalloc_to_write = 0; /* * make sure the wbc mapping index is at least updated @@ -2421,10 +2616,16 @@ retry: * swizzled back from swapper_space to tmpfs file * mapping */ - if (tree->ops && tree->ops->write_cache_pages_lock_hook) - tree->ops->write_cache_pages_lock_hook(page); - else - lock_page(page); + if (tree->ops && + tree->ops->write_cache_pages_lock_hook) { + tree->ops->write_cache_pages_lock_hook(page, + data, flush_fn); + } else { + if (!trylock_page(page)) { + flush_fn(data); + lock_page(page); + } + } if (unlikely(page->mapping != mapping)) { unlock_page(page); @@ -3204,6 +3405,7 @@ int clear_extent_buffer_dirty(struct extent_io_tree *tree, PAGECACHE_TAG_DIRTY); } spin_unlock_irq(&page->mapping->tree_lock); + ClearPageError(page); unlock_page(page); } return 0; diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index fcaf49bcb880..697570eed9e8 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -17,6 +17,7 @@ #define EXTENT_NODATASUM (1 << 10) #define EXTENT_DO_ACCOUNTING (1 << 11) #define EXTENT_FIRST_DELALLOC (1 << 12) +#define EXTENT_NEED_WAIT (1 << 13) #define EXTENT_IOBITS (EXTENT_LOCKED | EXTENT_WRITEBACK) #define EXTENT_CTLBITS (EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC) @@ -86,7 +87,8 @@ struct extent_io_ops { struct extent_state *other); void (*split_extent_hook)(struct inode *inode, struct extent_state *orig, u64 split); - int (*write_cache_pages_lock_hook)(struct page *page); + int (*write_cache_pages_lock_hook)(struct page *page, void *data, + void (*flush_fn)(void *)); }; struct extent_io_tree { @@ -215,6 +217,8 @@ int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); +int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, + int bits, int clear_bits, gfp_t mask); int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached_state, gfp_t mask); int find_first_extent_bit(struct extent_io_tree *tree, u64 start, diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index a1cb7821becd..c7fb3a4247d3 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -91,8 +91,7 @@ struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans, struct btrfs_csum_item *item; struct extent_buffer *leaf; u64 csum_offset = 0; - u16 csum_size = - btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); int csums_in_item; file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID; @@ -162,8 +161,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root, u64 item_last_offset = 0; u64 disk_bytenr; u32 diff; - u16 csum_size = - btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); int ret; struct btrfs_path *path; struct btrfs_csum_item *item = NULL; @@ -290,7 +288,7 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, int ret; size_t size; u64 csum_end; - u16 csum_size = btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); path = btrfs_alloc_path(); if (!path) @@ -492,8 +490,7 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans, u64 bytenr, u64 len) { struct extent_buffer *leaf; - u16 csum_size = - btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); u64 csum_end; u64 end_byte = bytenr + len; u32 blocksize_bits = root->fs_info->sb->s_blocksize_bits; @@ -549,8 +546,7 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, u64 csum_end; struct extent_buffer *leaf; int ret; - u16 csum_size = - btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); int blocksize_bits = root->fs_info->sb->s_blocksize_bits; root = root->fs_info->csum_root; @@ -676,8 +672,7 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans, struct btrfs_sector_sum *sector_sum; u32 nritems; u32 ins_size; - u16 csum_size = - btrfs_super_csum_size(&root->fs_info->super_copy); + u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy); path = btrfs_alloc_path(); if (!path) diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index e4e57d59edb7..f2e928289600 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1069,6 +1069,7 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file, int i; unsigned long index = pos >> PAGE_CACHE_SHIFT; struct inode *inode = fdentry(file)->d_inode; + gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping); int err = 0; int faili = 0; u64 start_pos; @@ -1080,7 +1081,7 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file, again: for (i = 0; i < num_pages; i++) { pages[i] = find_or_create_page(inode->i_mapping, index + i, - GFP_NOFS); + mask); if (!pages[i]) { faili = i - 1; err = -ENOMEM; @@ -1615,10 +1616,6 @@ static long btrfs_fallocate(struct file *file, int mode, goto out; } - ret = btrfs_check_data_free_space(inode, alloc_end - alloc_start); - if (ret) - goto out; - locked_end = alloc_end - 1; while (1) { struct btrfs_ordered_extent *ordered; @@ -1664,11 +1661,27 @@ static long btrfs_fallocate(struct file *file, int mode, if (em->block_start == EXTENT_MAP_HOLE || (cur_offset >= inode->i_size && !test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) { + + /* + * Make sure we have enough space before we do the + * allocation. + */ + ret = btrfs_check_data_free_space(inode, last_byte - + cur_offset); + if (ret) { + free_extent_map(em); + break; + } + ret = btrfs_prealloc_file_range(inode, mode, cur_offset, last_byte - cur_offset, 1 << inode->i_blkbits, offset + len, &alloc_hint); + + /* Let go of our reservation. */ + btrfs_free_reserved_data_space(inode, last_byte - + cur_offset); if (ret < 0) { free_extent_map(em); break; @@ -1694,8 +1707,6 @@ static long btrfs_fallocate(struct file *file, int mode, } unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end, &cached_state, GFP_NOFS); - - btrfs_free_reserved_data_space(inode, alloc_end - alloc_start); out: mutex_unlock(&inode->i_mutex); return ret; diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 41ac927401d0..7a15fcfb3e1f 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -20,6 +20,7 @@ #include <linux/sched.h> #include <linux/slab.h> #include <linux/math64.h> +#include <linux/ratelimit.h> #include "ctree.h" #include "free-space-cache.h" #include "transaction.h" @@ -84,6 +85,7 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root, *block_group, struct btrfs_path *path) { struct inode *inode = NULL; + u32 flags = BTRFS_INODE_NODATASUM | BTRFS_INODE_NODATACOW; spin_lock(&block_group->lock); if (block_group->inode) @@ -98,13 +100,14 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root, return inode; spin_lock(&block_group->lock); - if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM) { + if (!((BTRFS_I(inode)->flags & flags) == flags)) { printk(KERN_INFO "Old style space inode found, converting.\n"); - BTRFS_I(inode)->flags &= ~BTRFS_INODE_NODATASUM; + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM | + BTRFS_INODE_NODATACOW; block_group->disk_cache_state = BTRFS_DC_CLEAR; } - if (!btrfs_fs_closing(root->fs_info)) { + if (!block_group->iref) { block_group->inode = igrab(inode); block_group->iref = 1; } @@ -122,12 +125,17 @@ int __create_free_space_inode(struct btrfs_root *root, struct btrfs_free_space_header *header; struct btrfs_inode_item *inode_item; struct extent_buffer *leaf; + u64 flags = BTRFS_INODE_NOCOMPRESS | BTRFS_INODE_PREALLOC; int ret; ret = btrfs_insert_empty_inode(trans, root, path, ino); if (ret) return ret; + /* We inline crc's for the free disk space cache */ + if (ino != BTRFS_FREE_INO_OBJECTID) + flags |= BTRFS_INODE_NODATASUM | BTRFS_INODE_NODATACOW; + leaf = path->nodes[0]; inode_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_inode_item); @@ -140,8 +148,7 @@ int __create_free_space_inode(struct btrfs_root *root, btrfs_set_inode_uid(leaf, inode_item, 0); btrfs_set_inode_gid(leaf, inode_item, 0); btrfs_set_inode_mode(leaf, inode_item, S_IFREG | 0600); - btrfs_set_inode_flags(leaf, inode_item, BTRFS_INODE_NOCOMPRESS | - BTRFS_INODE_PREALLOC); + btrfs_set_inode_flags(leaf, inode_item, flags); btrfs_set_inode_nlink(leaf, inode_item, 1); btrfs_set_inode_transid(leaf, inode_item, trans->transid); btrfs_set_inode_block_group(leaf, inode_item, offset); @@ -191,16 +198,24 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root, struct inode *inode) { struct btrfs_block_rsv *rsv; + u64 needed_bytes; loff_t oldsize; int ret = 0; rsv = trans->block_rsv; - trans->block_rsv = root->orphan_block_rsv; - ret = btrfs_block_rsv_check(trans, root, - root->orphan_block_rsv, - 0, 5); - if (ret) - return ret; + trans->block_rsv = &root->fs_info->global_block_rsv; + + /* 1 for slack space, 1 for updating the inode */ + needed_bytes = btrfs_calc_trunc_metadata_size(root, 1) + + btrfs_calc_trans_metadata_size(root, 1); + + spin_lock(&trans->block_rsv->lock); + if (trans->block_rsv->reserved < needed_bytes) { + spin_unlock(&trans->block_rsv->lock); + trans->block_rsv = rsv; + return -ENOSPC; + } + spin_unlock(&trans->block_rsv->lock); oldsize = i_size_read(inode); btrfs_i_size_write(inode, 0); @@ -213,13 +228,15 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root, ret = btrfs_truncate_inode_items(trans, root, inode, 0, BTRFS_EXTENT_DATA_KEY); - trans->block_rsv = rsv; if (ret) { + trans->block_rsv = rsv; WARN_ON(1); return ret; } ret = btrfs_update_inode(trans, root, inode); + trans->block_rsv = rsv; + return ret; } @@ -242,26 +259,342 @@ static int readahead_cache(struct inode *inode) return 0; } +struct io_ctl { + void *cur, *orig; + struct page *page; + struct page **pages; + struct btrfs_root *root; + unsigned long size; + int index; + int num_pages; + unsigned check_crcs:1; +}; + +static int io_ctl_init(struct io_ctl *io_ctl, struct inode *inode, + struct btrfs_root *root) +{ + memset(io_ctl, 0, sizeof(struct io_ctl)); + io_ctl->num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> + PAGE_CACHE_SHIFT; + io_ctl->pages = kzalloc(sizeof(struct page *) * io_ctl->num_pages, + GFP_NOFS); + if (!io_ctl->pages) + return -ENOMEM; + io_ctl->root = root; + if (btrfs_ino(inode) != BTRFS_FREE_INO_OBJECTID) + io_ctl->check_crcs = 1; + return 0; +} + +static void io_ctl_free(struct io_ctl *io_ctl) +{ + kfree(io_ctl->pages); +} + +static void io_ctl_unmap_page(struct io_ctl *io_ctl) +{ + if (io_ctl->cur) { + kunmap(io_ctl->page); + io_ctl->cur = NULL; + io_ctl->orig = NULL; + } +} + +static void io_ctl_map_page(struct io_ctl *io_ctl, int clear) +{ + WARN_ON(io_ctl->cur); + BUG_ON(io_ctl->index >= io_ctl->num_pages); + io_ctl->page = io_ctl->pages[io_ctl->index++]; + io_ctl->cur = kmap(io_ctl->page); + io_ctl->orig = io_ctl->cur; + io_ctl->size = PAGE_CACHE_SIZE; + if (clear) + memset(io_ctl->cur, 0, PAGE_CACHE_SIZE); +} + +static void io_ctl_drop_pages(struct io_ctl *io_ctl) +{ + int i; + + io_ctl_unmap_page(io_ctl); + + for (i = 0; i < io_ctl->num_pages; i++) { + ClearPageChecked(io_ctl->pages[i]); + unlock_page(io_ctl->pages[i]); + page_cache_release(io_ctl->pages[i]); + } +} + +static int io_ctl_prepare_pages(struct io_ctl *io_ctl, struct inode *inode, + int uptodate) +{ + struct page *page; + gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping); + int i; + + for (i = 0; i < io_ctl->num_pages; i++) { + page = find_or_create_page(inode->i_mapping, i, mask); + if (!page) { + io_ctl_drop_pages(io_ctl); + return -ENOMEM; + } + io_ctl->pages[i] = page; + if (uptodate && !PageUptodate(page)) { + btrfs_readpage(NULL, page); + lock_page(page); + if (!PageUptodate(page)) { + printk(KERN_ERR "btrfs: error reading free " + "space cache\n"); + io_ctl_drop_pages(io_ctl); + return -EIO; + } + } + } + + return 0; +} + +static void io_ctl_set_generation(struct io_ctl *io_ctl, u64 generation) +{ + u64 *val; + + io_ctl_map_page(io_ctl, 1); + + /* + * Skip the csum areas. If we don't check crcs then we just have a + * 64bit chunk at the front of the first page. + */ + if (io_ctl->check_crcs) { + io_ctl->cur += (sizeof(u32) * io_ctl->num_pages); + io_ctl->size -= sizeof(u64) + (sizeof(u32) * io_ctl->num_pages); + } else { + io_ctl->cur += sizeof(u64); + io_ctl->size -= sizeof(u64) * 2; + } + + val = io_ctl->cur; + *val = cpu_to_le64(generation); + io_ctl->cur += sizeof(u64); +} + +static int io_ctl_check_generation(struct io_ctl *io_ctl, u64 generation) +{ + u64 *gen; + + /* + * Skip the crc area. If we don't check crcs then we just have a 64bit + * chunk at the front of the first page. + */ + if (io_ctl->check_crcs) { + io_ctl->cur += sizeof(u32) * io_ctl->num_pages; + io_ctl->size -= sizeof(u64) + + (sizeof(u32) * io_ctl->num_pages); + } else { + io_ctl->cur += sizeof(u64); + io_ctl->size -= sizeof(u64) * 2; + } + + gen = io_ctl->cur; + if (le64_to_cpu(*gen) != generation) { + printk_ratelimited(KERN_ERR "btrfs: space cache generation " + "(%Lu) does not match inode (%Lu)\n", *gen, + generation); + io_ctl_unmap_page(io_ctl); + return -EIO; + } + io_ctl->cur += sizeof(u64); + return 0; +} + +static void io_ctl_set_crc(struct io_ctl *io_ctl, int index) +{ + u32 *tmp; + u32 crc = ~(u32)0; + unsigned offset = 0; + + if (!io_ctl->check_crcs) { + io_ctl_unmap_page(io_ctl); + return; + } + + if (index == 0) + offset = sizeof(u32) * io_ctl->num_pages;; + + crc = btrfs_csum_data(io_ctl->root, io_ctl->orig + offset, crc, + PAGE_CACHE_SIZE - offset); + btrfs_csum_final(crc, (char *)&crc); + io_ctl_unmap_page(io_ctl); + tmp = kmap(io_ctl->pages[0]); + tmp += index; + *tmp = crc; + kunmap(io_ctl->pages[0]); +} + +static int io_ctl_check_crc(struct io_ctl *io_ctl, int index) +{ + u32 *tmp, val; + u32 crc = ~(u32)0; + unsigned offset = 0; + + if (!io_ctl->check_crcs) { + io_ctl_map_page(io_ctl, 0); + return 0; + } + + if (index == 0) + offset = sizeof(u32) * io_ctl->num_pages; + + tmp = kmap(io_ctl->pages[0]); + tmp += index; + val = *tmp; + kunmap(io_ctl->pages[0]); + + io_ctl_map_page(io_ctl, 0); + crc = btrfs_csum_data(io_ctl->root, io_ctl->orig + offset, crc, + PAGE_CACHE_SIZE - offset); + btrfs_csum_final(crc, (char *)&crc); + if (val != crc) { + printk_ratelimited(KERN_ERR "btrfs: csum mismatch on free " + "space cache\n"); + io_ctl_unmap_page(io_ctl); + return -EIO; + } + + return 0; +} + +static int io_ctl_add_entry(struct io_ctl *io_ctl, u64 offset, u64 bytes, + void *bitmap) +{ + struct btrfs_free_space_entry *entry; + + if (!io_ctl->cur) + return -ENOSPC; + + entry = io_ctl->cur; + entry->offset = cpu_to_le64(offset); + entry->bytes = cpu_to_le64(bytes); + entry->type = (bitmap) ? BTRFS_FREE_SPACE_BITMAP : + BTRFS_FREE_SPACE_EXTENT; + io_ctl->cur += sizeof(struct btrfs_free_space_entry); + io_ctl->size -= sizeof(struct btrfs_free_space_entry); + + if (io_ctl->size >= sizeof(struct btrfs_free_space_entry)) + return 0; + + io_ctl_set_crc(io_ctl, io_ctl->index - 1); + + /* No more pages to map */ + if (io_ctl->index >= io_ctl->num_pages) + return 0; + + /* map the next page */ + io_ctl_map_page(io_ctl, 1); + return 0; +} + +static int io_ctl_add_bitmap(struct io_ctl *io_ctl, void *bitmap) +{ + if (!io_ctl->cur) + return -ENOSPC; + + /* + * If we aren't at the start of the current page, unmap this one and + * map the next one if there is any left. + */ + if (io_ctl->cur != io_ctl->orig) { + io_ctl_set_crc(io_ctl, io_ctl->index - 1); + if (io_ctl->index >= io_ctl->num_pages) + return -ENOSPC; + io_ctl_map_page(io_ctl, 0); + } + + memcpy(io_ctl->cur, bitmap, PAGE_CACHE_SIZE); + io_ctl_set_crc(io_ctl, io_ctl->index - 1); + if (io_ctl->index < io_ctl->num_pages) + io_ctl_map_page(io_ctl, 0); + return 0; +} + +static void io_ctl_zero_remaining_pages(struct io_ctl *io_ctl) +{ + /* + * If we're not on the boundary we know we've modified the page and we + * need to crc the page. + */ + if (io_ctl->cur != io_ctl->orig) + io_ctl_set_crc(io_ctl, io_ctl->index - 1); + else + io_ctl_unmap_page(io_ctl); + + while (io_ctl->index < io_ctl->num_pages) { + io_ctl_map_page(io_ctl, 1); + io_ctl_set_crc(io_ctl, io_ctl->index - 1); + } +} + +static int io_ctl_read_entry(struct io_ctl *io_ctl, + struct btrfs_free_space *entry, u8 *type) +{ + struct btrfs_free_space_entry *e; + + e = io_ctl->cur; + entry->offset = le64_to_cpu(e->offset); + entry->bytes = le64_to_cpu(e->bytes); + *type = e->type; + io_ctl->cur += sizeof(struct btrfs_free_space_entry); + io_ctl->size -= sizeof(struct btrfs_free_space_entry); + + if (io_ctl->size >= sizeof(struct btrfs_free_space_entry)) + return 0; + + io_ctl_unmap_page(io_ctl); + + if (io_ctl->index >= io_ctl->num_pages) + return 0; + + return io_ctl_check_crc(io_ctl, io_ctl->index); +} + +static int io_ctl_read_bitmap(struct io_ctl *io_ctl, + struct btrfs_free_space *entry) +{ + int ret; + + if (io_ctl->cur && io_ctl->cur != io_ctl->orig) + io_ctl_unmap_page(io_ctl); + + ret = io_ctl_check_crc(io_ctl, io_ctl->index); + if (ret) + return ret; + + memcpy(entry->bitmap, io_ctl->cur, PAGE_CACHE_SIZE); + io_ctl_unmap_page(io_ctl); + + return 0; +} + int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, struct btrfs_free_space_ctl *ctl, struct btrfs_path *path, u64 offset) { struct btrfs_free_space_header *header; struct extent_buffer *leaf; - struct page *page; + struct io_ctl io_ctl; struct btrfs_key key; + struct btrfs_free_space *e, *n; struct list_head bitmaps; u64 num_entries; u64 num_bitmaps; u64 generation; - pgoff_t index = 0; + u8 type; int ret = 0; INIT_LIST_HEAD(&bitmaps); /* Nothing in the space cache, goodbye */ if (!i_size_read(inode)) - goto out; + return 0; key.objectid = BTRFS_FREE_SPACE_OBJECTID; key.offset = offset; @@ -269,11 +602,10 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) - goto out; + return 0; else if (ret > 0) { btrfs_release_path(path); - ret = 0; - goto out; + return 0; } ret = -1; @@ -291,169 +623,100 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, " not match free space cache generation (%llu)\n", (unsigned long long)BTRFS_I(inode)->generation, (unsigned long long)generation); - goto out; + return 0; } if (!num_entries) - goto out; + return 0; + io_ctl_init(&io_ctl, inode, root); ret = readahead_cache(inode); if (ret) goto out; - while (1) { - struct btrfs_free_space_entry *entry; - struct btrfs_free_space *e; - void *addr; - unsigned long offset = 0; - int need_loop = 0; + ret = io_ctl_prepare_pages(&io_ctl, inode, 1); + if (ret) + goto out; - if (!num_entries && !num_bitmaps) - break; + ret = io_ctl_check_crc(&io_ctl, 0); + if (ret) + goto free_cache; - page = find_or_create_page(inode->i_mapping, index, GFP_NOFS); - if (!page) + ret = io_ctl_check_generation(&io_ctl, generation); + if (ret) + goto free_cache; + + while (num_entries) { + e = kmem_cache_zalloc(btrfs_free_space_cachep, + GFP_NOFS); + if (!e) goto free_cache; - if (!PageUptodate(page)) { - btrfs_readpage(NULL, page); - lock_page(page); - if (!PageUptodate(page)) { - unlock_page(page); - page_cache_release(page); - printk(KERN_ERR "btrfs: error reading free " - "space cache\n"); - goto free_cache; - } + ret = io_ctl_read_entry(&io_ctl, e, &type); + if (ret) { + kmem_cache_free(btrfs_free_space_cachep, e); + goto free_cache; } - addr = kmap(page); - if (index == 0) { - u64 *gen; + if (!e->bytes) { + kmem_cache_free(btrfs_free_space_cachep, e); + goto free_cache; + } - /* - * We put a bogus crc in the front of the first page in - * case old kernels try to mount a fs with the new - * format to make sure they discard the cache. - */ - addr += sizeof(u64); - offset += sizeof(u64); - - gen = addr; - if (*gen != BTRFS_I(inode)->generation) { - printk(KERN_ERR "btrfs: space cache generation" - " (%llu) does not match inode (%llu)\n", - (unsigned long long)*gen, - (unsigned long long) - BTRFS_I(inode)->generation); - kunmap(page); - unlock_page(page); - page_cache_release(page); + if (type == BTRFS_FREE_SPACE_EXTENT) { + spin_lock(&ctl->tree_lock); + ret = link_free_space(ctl, e); + spin_unlock(&ctl->tree_lock); + if (ret) { + printk(KERN_ERR "Duplicate entries in " + "free space cache, dumping\n"); + kmem_cache_free(btrfs_free_space_cachep, e); goto free_cache; } - addr += sizeof(u64); - offset += sizeof(u64); - } - entry = addr; - - while (1) { - if (!num_entries) - break; - - need_loop = 1; - e = kmem_cache_zalloc(btrfs_free_space_cachep, - GFP_NOFS); - if (!e) { - kunmap(page); - unlock_page(page); - page_cache_release(page); + } else { + BUG_ON(!num_bitmaps); + num_bitmaps--; + e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); + if (!e->bitmap) { + kmem_cache_free( + btrfs_free_space_cachep, e); goto free_cache; } - - e->offset = le64_to_cpu(entry->offset); - e->bytes = le64_to_cpu(entry->bytes); - if (!e->bytes) { - kunmap(page); + spin_lock(&ctl->tree_lock); + ret = link_free_space(ctl, e); + ctl->total_bitmaps++; + ctl->op->recalc_thresholds(ctl); + spin_unlock(&ctl->tree_lock); + if (ret) { + printk(KERN_ERR "Duplicate entries in " + "free space cache, dumping\n"); kmem_cache_free(btrfs_free_space_cachep, e); - unlock_page(page); - page_cache_release(page); goto free_cache; } - - if (entry->type == BTRFS_FREE_SPACE_EXTENT) { - spin_lock(&ctl->tree_lock); - ret = link_free_space(ctl, e); - spin_unlock(&ctl->tree_lock); - if (ret) { - printk(KERN_ERR "Duplicate entries in " - "free space cache, dumping\n"); - kunmap(page); - unlock_page(page); - page_cache_release(page); - goto free_cache; - } - } else { - e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); - if (!e->bitmap) { - kunmap(page); - kmem_cache_free( - btrfs_free_space_cachep, e); - unlock_page(page); - page_cache_release(page); - goto free_cache; - } - spin_lock(&ctl->tree_lock); - ret = link_free_space(ctl, e); - ctl->total_bitmaps++; - ctl->op->recalc_thresholds(ctl); - spin_unlock(&ctl->tree_lock); - if (ret) { - printk(KERN_ERR "Duplicate entries in " - "free space cache, dumping\n"); - kunmap(page); - unlock_page(page); - page_cache_release(page); - goto free_cache; - } - list_add_tail(&e->list, &bitmaps); - } - - num_entries--; - offset += sizeof(struct btrfs_free_space_entry); - if (offset + sizeof(struct btrfs_free_space_entry) >= - PAGE_CACHE_SIZE) - break; - entry++; + list_add_tail(&e->list, &bitmaps); } - /* - * We read an entry out of this page, we need to move on to the - * next page. - */ - if (need_loop) { - kunmap(page); - goto next; - } + num_entries--; + } - /* - * We add the bitmaps at the end of the entries in order that - * the bitmap entries are added to the cache. - */ - e = list_entry(bitmaps.next, struct btrfs_free_space, list); + /* + * We add the bitmaps at the end of the entries in order that + * the bitmap entries are added to the cache. + */ + list_for_each_entry_safe(e, n, &bitmaps, list) { list_del_init(&e->list); - memcpy(e->bitmap, addr, PAGE_CACHE_SIZE); - kunmap(page); - num_bitmaps--; -next: - unlock_page(page); - page_cache_release(page); - index++; + ret = io_ctl_read_bitmap(&io_ctl, e); + if (ret) + goto free_cache; } + io_ctl_drop_pages(&io_ctl); ret = 1; out: + io_ctl_free(&io_ctl); return ret; free_cache: + io_ctl_drop_pages(&io_ctl); __btrfs_remove_free_space_cache(ctl); goto out; } @@ -465,7 +728,7 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, struct btrfs_root *root = fs_info->tree_root; struct inode *inode; struct btrfs_path *path; - int ret; + int ret = 0; bool matched; u64 used = btrfs_block_group_used(&block_group->item); @@ -497,6 +760,14 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, return 0; } + /* We may have converted the inode and made the cache invalid. */ + spin_lock(&block_group->lock); + if (block_group->disk_cache_state != BTRFS_DC_WRITTEN) { + spin_unlock(&block_group->lock); + goto out; + } + spin_unlock(&block_group->lock); + ret = __load_free_space_cache(fs_info->tree_root, inode, ctl, path, block_group->key.objectid); btrfs_free_path(path); @@ -530,6 +801,19 @@ out: return ret; } +/** + * __btrfs_write_out_cache - write out cached info to an inode + * @root - the root the inode belongs to + * @ctl - the free space cache we are going to write out + * @block_group - the block_group for this cache if it belongs to a block_group + * @trans - the trans handle + * @path - the path to use + * @offset - the offset for the key we'll insert + * + * This function writes out a free space cache struct to disk for quick recovery + * on mount. This will return 0 if it was successfull in writing the cache out, + * and -1 if it was not. + */ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, struct btrfs_free_space_ctl *ctl, struct btrfs_block_group_cache *block_group, @@ -540,42 +824,24 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, struct extent_buffer *leaf; struct rb_node *node; struct list_head *pos, *n; - struct page **pages; - struct page *page; struct extent_state *cached_state = NULL; struct btrfs_free_cluster *cluster = NULL; struct extent_io_tree *unpin = NULL; + struct io_ctl io_ctl; struct list_head bitmap_list; struct btrfs_key key; u64 start, end, len; - u64 bytes = 0; - u32 crc = ~(u32)0; - int index = 0, num_pages = 0; int entries = 0; int bitmaps = 0; - int ret = -1; - bool next_page = false; - bool out_of_space = false; + int ret; + int err = -1; INIT_LIST_HEAD(&bitmap_list); - node = rb_first(&ctl->free_space_offset); - if (!node) - return 0; - if (!i_size_read(inode)) return -1; - num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> - PAGE_CACHE_SHIFT; - - filemap_write_and_wait(inode->i_mapping); - btrfs_wait_ordered_range(inode, inode->i_size & - ~(root->sectorsize - 1), (u64)-1); - - pages = kzalloc(sizeof(struct page *) * num_pages, GFP_NOFS); - if (!pages) - return -1; + io_ctl_init(&io_ctl, inode, root); /* Get the cluster for this block_group if it exists */ if (block_group && !list_empty(&block_group->cluster_list)) @@ -589,30 +855,9 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, */ unpin = root->fs_info->pinned_extents; - /* - * Lock all pages first so we can lock the extent safely. - * - * NOTE: Because we hold the ref the entire time we're going to write to - * the page find_get_page should never fail, so we don't do a check - * after find_get_page at this point. Just putting this here so people - * know and don't freak out. - */ - while (index < num_pages) { - page = find_or_create_page(inode->i_mapping, index, GFP_NOFS); - if (!page) { - int i; - - for (i = 0; i < num_pages; i++) { - unlock_page(pages[i]); - page_cache_release(pages[i]); - } - goto out; - } - pages[index] = page; - index++; - } + /* Lock all pages first so we can lock the extent safely. */ + io_ctl_prepare_pages(&io_ctl, inode, 0); - index = 0; lock_extent_bits(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, 0, &cached_state, GFP_NOFS); @@ -623,189 +868,111 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, if (block_group) start = block_group->key.objectid; - /* Write out the extent entries */ - do { - struct btrfs_free_space_entry *entry; - void *addr, *orig; - unsigned long offset = 0; + node = rb_first(&ctl->free_space_offset); + if (!node && cluster) { + node = rb_first(&cluster->root); + cluster = NULL; + } - next_page = false; + /* Make sure we can fit our crcs into the first page */ + if (io_ctl.check_crcs && + (io_ctl.num_pages * sizeof(u32)) >= PAGE_CACHE_SIZE) { + WARN_ON(1); + goto out_nospc; + } - if (index >= num_pages) { - out_of_space = true; - break; - } + io_ctl_set_generation(&io_ctl, trans->transid); - page = pages[index]; + /* Write out the extent entries */ + while (node) { + struct btrfs_free_space *e; - orig = addr = kmap(page); - if (index == 0) { - u64 *gen; + e = rb_entry(node, struct btrfs_free_space, offset_index); + entries++; - /* - * We're going to put in a bogus crc for this page to - * make sure that old kernels who aren't aware of this - * format will be sure to discard the cache. - */ - addr += sizeof(u64); - offset += sizeof(u64); + ret = io_ctl_add_entry(&io_ctl, e->offset, e->bytes, + e->bitmap); + if (ret) + goto out_nospc; - gen = addr; - *gen = trans->transid; - addr += sizeof(u64); - offset += sizeof(u64); + if (e->bitmap) { + list_add_tail(&e->list, &bitmap_list); + bitmaps++; } - entry = addr; - - memset(addr, 0, PAGE_CACHE_SIZE - offset); - while (node && !next_page) { - struct btrfs_free_space *e; - - e = rb_entry(node, struct btrfs_free_space, offset_index); - entries++; - - entry->offset = cpu_to_le64(e->offset); - entry->bytes = cpu_to_le64(e->bytes); - if (e->bitmap) { - entry->type = BTRFS_FREE_SPACE_BITMAP; - list_add_tail(&e->list, &bitmap_list); - bitmaps++; - } else { - entry->type = BTRFS_FREE_SPACE_EXTENT; - } - node = rb_next(node); - if (!node && cluster) { - node = rb_first(&cluster->root); - cluster = NULL; - } - offset += sizeof(struct btrfs_free_space_entry); - if (offset + sizeof(struct btrfs_free_space_entry) >= - PAGE_CACHE_SIZE) - next_page = true; - entry++; + node = rb_next(node); + if (!node && cluster) { + node = rb_first(&cluster->root); + cluster = NULL; } + } - /* - * We want to add any pinned extents to our free space cache - * so we don't leak the space - */ - while (block_group && !next_page && - (start < block_group->key.objectid + - block_group->key.offset)) { - ret = find_first_extent_bit(unpin, start, &start, &end, - EXTENT_DIRTY); - if (ret) { - ret = 0; - break; - } - - /* This pinned extent is out of our range */ - if (start >= block_group->key.objectid + - block_group->key.offset) - break; - - len = block_group->key.objectid + - block_group->key.offset - start; - len = min(len, end + 1 - start); - - entries++; - entry->offset = cpu_to_le64(start); - entry->bytes = cpu_to_le64(len); - entry->type = BTRFS_FREE_SPACE_EXTENT; - - start = end + 1; - offset += sizeof(struct btrfs_free_space_entry); - if (offset + sizeof(struct btrfs_free_space_entry) >= - PAGE_CACHE_SIZE) - next_page = true; - entry++; + /* + * We want to add any pinned extents to our free space cache + * so we don't leak the space + */ + while (block_group && (start < block_group->key.objectid + + block_group->key.offset)) { + ret = find_first_extent_bit(unpin, start, &start, &end, + EXTENT_DIRTY); + if (ret) { + ret = 0; + break; } - /* Generate bogus crc value */ - if (index == 0) { - u32 *tmp; - crc = btrfs_csum_data(root, orig + sizeof(u64), crc, - PAGE_CACHE_SIZE - sizeof(u64)); - btrfs_csum_final(crc, (char *)&crc); - crc++; - tmp = orig; - *tmp = crc; - } + /* This pinned extent is out of our range */ + if (start >= block_group->key.objectid + + block_group->key.offset) + break; - kunmap(page); + len = block_group->key.objectid + + block_group->key.offset - start; + len = min(len, end + 1 - start); - bytes += PAGE_CACHE_SIZE; + entries++; + ret = io_ctl_add_entry(&io_ctl, start, len, NULL); + if (ret) + goto out_nospc; - index++; - } while (node || next_page); + start = end + 1; + } /* Write out the bitmaps */ list_for_each_safe(pos, n, &bitmap_list) { - void *addr; struct btrfs_free_space *entry = list_entry(pos, struct btrfs_free_space, list); - if (index >= num_pages) { - out_of_space = true; - break; - } - page = pages[index]; - - addr = kmap(page); - memcpy(addr, entry->bitmap, PAGE_CACHE_SIZE); - kunmap(page); - bytes += PAGE_CACHE_SIZE; - + ret = io_ctl_add_bitmap(&io_ctl, entry->bitmap); + if (ret) + goto out_nospc; list_del_init(&entry->list); - index++; - } - - if (out_of_space) { - btrfs_drop_pages(pages, num_pages); - unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, - i_size_read(inode) - 1, &cached_state, - GFP_NOFS); - ret = 0; - goto out; } /* Zero out the rest of the pages just to make sure */ - while (index < num_pages) { - void *addr; - - page = pages[index]; - addr = kmap(page); - memset(addr, 0, PAGE_CACHE_SIZE); - kunmap(page); - bytes += PAGE_CACHE_SIZE; - index++; - } + io_ctl_zero_remaining_pages(&io_ctl); - ret = btrfs_dirty_pages(root, inode, pages, num_pages, 0, - bytes, &cached_state); - btrfs_drop_pages(pages, num_pages); + ret = btrfs_dirty_pages(root, inode, io_ctl.pages, io_ctl.num_pages, + 0, i_size_read(inode), &cached_state); + io_ctl_drop_pages(&io_ctl); unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, &cached_state, GFP_NOFS); - if (ret) { - ret = 0; + if (ret) goto out; - } - BTRFS_I(inode)->generation = trans->transid; - filemap_write_and_wait(inode->i_mapping); + ret = filemap_write_and_wait(inode->i_mapping); + if (ret) + goto out; key.objectid = BTRFS_FREE_SPACE_OBJECTID; key.offset = offset; key.type = 0; - ret = btrfs_search_slot(trans, root, &key, path, 1, 1); + ret = btrfs_search_slot(trans, root, &key, path, 0, 1); if (ret < 0) { - ret = -1; - clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, bytes - 1, - EXTENT_DIRTY | EXTENT_DELALLOC | - EXTENT_DO_ACCOUNTING, 0, 0, NULL, GFP_NOFS); + clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, inode->i_size - 1, + EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, NULL, + GFP_NOFS); goto out; } leaf = path->nodes[0]; @@ -816,15 +983,16 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); if (found_key.objectid != BTRFS_FREE_SPACE_OBJECTID || found_key.offset != offset) { - ret = -1; - clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, bytes - 1, - EXTENT_DIRTY | EXTENT_DELALLOC | - EXTENT_DO_ACCOUNTING, 0, 0, NULL, - GFP_NOFS); + clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, + inode->i_size - 1, + EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, + NULL, GFP_NOFS); btrfs_release_path(path); goto out; } } + + BTRFS_I(inode)->generation = trans->transid; header = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_free_space_header); btrfs_set_free_space_entries(leaf, header, entries); @@ -833,16 +1001,26 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, btrfs_mark_buffer_dirty(leaf); btrfs_release_path(path); - ret = 1; - + err = 0; out: - kfree(pages); - if (ret != 1) { - invalidate_inode_pages2_range(inode->i_mapping, 0, index); + io_ctl_free(&io_ctl); + if (err) { + invalidate_inode_pages2(inode->i_mapping); BTRFS_I(inode)->generation = 0; } btrfs_update_inode(trans, root, inode); - return ret; + return err; + +out_nospc: + list_for_each_safe(pos, n, &bitmap_list) { + struct btrfs_free_space *entry = + list_entry(pos, struct btrfs_free_space, list); + list_del_init(&entry->list); + } + io_ctl_drop_pages(&io_ctl); + unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, + i_size_read(inode) - 1, &cached_state, GFP_NOFS); + goto out; } int btrfs_write_out_cache(struct btrfs_root *root, @@ -869,14 +1047,15 @@ int btrfs_write_out_cache(struct btrfs_root *root, ret = __btrfs_write_out_cache(root, inode, ctl, block_group, trans, path, block_group->key.objectid); - if (ret < 0) { + if (ret) { spin_lock(&block_group->lock); block_group->disk_cache_state = BTRFS_DC_ERROR; spin_unlock(&block_group->lock); ret = 0; - +#ifdef DEBUG printk(KERN_ERR "btrfs: failed to write free space cace " "for block group %llu\n", block_group->key.objectid); +#endif } iput(inode); @@ -1701,6 +1880,7 @@ again: ctl->total_bitmaps--; } kmem_cache_free(btrfs_free_space_cachep, info); + ret = 0; goto out_lock; } @@ -1708,7 +1888,8 @@ again: unlink_free_space(ctl, info); info->offset += bytes; info->bytes -= bytes; - link_free_space(ctl, info); + ret = link_free_space(ctl, info); + WARN_ON(ret); goto out_lock; } @@ -2472,9 +2653,19 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, spin_unlock(&ctl->tree_lock); if (bytes >= minlen) { - int update_ret; - update_ret = btrfs_update_reserved_bytes(block_group, - bytes, 1, 1); + struct btrfs_space_info *space_info; + int update = 0; + + space_info = block_group->space_info; + spin_lock(&space_info->lock); + spin_lock(&block_group->lock); + if (!block_group->ro) { + block_group->reserved += bytes; + space_info->bytes_reserved += bytes; + update = 1; + } + spin_unlock(&block_group->lock); + spin_unlock(&space_info->lock); ret = btrfs_error_discard_extent(fs_info->extent_root, start, @@ -2482,9 +2673,16 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, &actually_trimmed); btrfs_add_free_space(block_group, start, bytes); - if (!update_ret) - btrfs_update_reserved_bytes(block_group, - bytes, 0, 1); + if (update) { + spin_lock(&space_info->lock); + spin_lock(&block_group->lock); + if (block_group->ro) + space_info->bytes_readonly += bytes; + block_group->reserved -= bytes; + space_info->bytes_reserved -= bytes; + spin_unlock(&space_info->lock); + spin_unlock(&block_group->lock); + } if (ret) break; @@ -2643,9 +2841,13 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root, return 0; ret = __btrfs_write_out_cache(root, inode, ctl, NULL, trans, path, 0); - if (ret < 0) + if (ret) { + btrfs_delalloc_release_metadata(inode, inode->i_size); +#ifdef DEBUG printk(KERN_ERR "btrfs: failed to write free ino cache " "for root %llu\n", root->root_key.objectid); +#endif + } iput(inode); return ret; diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c index b4087e0fa871..53dcbdf446cd 100644 --- a/fs/btrfs/inode-map.c +++ b/fs/btrfs/inode-map.c @@ -465,14 +465,16 @@ again: /* Just to make sure we have enough space */ prealloc += 8 * PAGE_CACHE_SIZE; - ret = btrfs_check_data_free_space(inode, prealloc); + ret = btrfs_delalloc_reserve_space(inode, prealloc); if (ret) goto out_put; ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, prealloc, prealloc, prealloc, &alloc_hint); - if (ret) + if (ret) { + btrfs_delalloc_release_space(inode, prealloc); goto out_put; + } btrfs_free_reserved_data_space(inode, prealloc); out_put: diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index b2d004ad66a0..b6b612e14ed7 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -393,7 +393,10 @@ again: (BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS))) { WARN_ON(pages); pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS); - BUG_ON(!pages); + if (!pages) { + /* just bail out to the uncompressed code */ + goto cont; + } if (BTRFS_I(inode)->force_compress) compress_type = BTRFS_I(inode)->force_compress; @@ -424,6 +427,7 @@ again: will_compress = 1; } } +cont: if (start == 0) { trans = btrfs_join_transaction(root); BUG_ON(IS_ERR(trans)); @@ -820,7 +824,7 @@ static noinline int cow_file_range(struct inode *inode, } BUG_ON(disk_num_bytes > - btrfs_super_total_bytes(&root->fs_info->super_copy)); + btrfs_super_total_bytes(root->fs_info->super_copy)); alloc_hint = get_extent_allocation_hint(inode, start, num_bytes); btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0); @@ -1792,12 +1796,12 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) } ret = 0; out: - if (nolock) { - if (trans) - btrfs_end_transaction_nolock(trans, root); - } else { + if (root != root->fs_info->tree_root) btrfs_delalloc_release_metadata(inode, ordered_extent->len); - if (trans) + if (trans) { + if (nolock) + btrfs_end_transaction_nolock(trans, root); + else btrfs_end_transaction(trans, root); } @@ -2079,89 +2083,6 @@ void btrfs_run_delayed_iputs(struct btrfs_root *root) up_read(&root->fs_info->cleanup_work_sem); } -/* - * calculate extra metadata reservation when snapshotting a subvolume - * contains orphan files. - */ -void btrfs_orphan_pre_snapshot(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending, - u64 *bytes_to_reserve) -{ - struct btrfs_root *root; - struct btrfs_block_rsv *block_rsv; - u64 num_bytes; - int index; - - root = pending->root; - if (!root->orphan_block_rsv || list_empty(&root->orphan_list)) - return; - - block_rsv = root->orphan_block_rsv; - - /* orphan block reservation for the snapshot */ - num_bytes = block_rsv->size; - - /* - * after the snapshot is created, COWing tree blocks may use more - * space than it frees. So we should make sure there is enough - * reserved space. - */ - index = trans->transid & 0x1; - if (block_rsv->reserved + block_rsv->freed[index] < block_rsv->size) { - num_bytes += block_rsv->size - - (block_rsv->reserved + block_rsv->freed[index]); - } - - *bytes_to_reserve += num_bytes; -} - -void btrfs_orphan_post_snapshot(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending) -{ - struct btrfs_root *root = pending->root; - struct btrfs_root *snap = pending->snap; - struct btrfs_block_rsv *block_rsv; - u64 num_bytes; - int index; - int ret; - - if (!root->orphan_block_rsv || list_empty(&root->orphan_list)) - return; - - /* refill source subvolume's orphan block reservation */ - block_rsv = root->orphan_block_rsv; - index = trans->transid & 0x1; - if (block_rsv->reserved + block_rsv->freed[index] < block_rsv->size) { - num_bytes = block_rsv->size - - (block_rsv->reserved + block_rsv->freed[index]); - ret = btrfs_block_rsv_migrate(&pending->block_rsv, - root->orphan_block_rsv, - num_bytes); - BUG_ON(ret); - } - - /* setup orphan block reservation for the snapshot */ - block_rsv = btrfs_alloc_block_rsv(snap); - BUG_ON(!block_rsv); - - btrfs_add_durable_block_rsv(root->fs_info, block_rsv); - snap->orphan_block_rsv = block_rsv; - - num_bytes = root->orphan_block_rsv->size; - ret = btrfs_block_rsv_migrate(&pending->block_rsv, - block_rsv, num_bytes); - BUG_ON(ret); - -#if 0 - /* insert orphan item for the snapshot */ - WARN_ON(!root->orphan_item_inserted); - ret = btrfs_insert_orphan_item(trans, root->fs_info->tree_root, - snap->root_key.objectid); - BUG_ON(ret); - snap->orphan_item_inserted = 1; -#endif -} - enum btrfs_orphan_cleanup_state { ORPHAN_CLEANUP_STARTED = 1, ORPHAN_CLEANUP_DONE = 2, @@ -2247,9 +2168,6 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode) } spin_unlock(&root->orphan_lock); - if (block_rsv) - btrfs_add_durable_block_rsv(root->fs_info, block_rsv); - /* grab metadata reservation from transaction handle */ if (reserve) { ret = btrfs_orphan_reserve_metadata(trans, inode); @@ -2316,6 +2234,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) struct btrfs_key key, found_key; struct btrfs_trans_handle *trans; struct inode *inode; + u64 last_objectid = 0; int ret = 0, nr_unlink = 0, nr_truncate = 0; if (cmpxchg(&root->orphan_cleanup_state, 0, ORPHAN_CLEANUP_STARTED)) @@ -2367,41 +2286,49 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) * crossing root thing. we store the inode number in the * offset of the orphan item. */ + + if (found_key.offset == last_objectid) { + printk(KERN_ERR "btrfs: Error removing orphan entry, " + "stopping orphan cleanup\n"); + ret = -EINVAL; + goto out; + } + + last_objectid = found_key.offset; + found_key.objectid = found_key.offset; found_key.type = BTRFS_INODE_ITEM_KEY; found_key.offset = 0; inode = btrfs_iget(root->fs_info->sb, &found_key, root, NULL); - if (IS_ERR(inode)) { - ret = PTR_ERR(inode); + ret = PTR_RET(inode); + if (ret && ret != -ESTALE) goto out; - } /* - * add this inode to the orphan list so btrfs_orphan_del does - * the proper thing when we hit it - */ - spin_lock(&root->orphan_lock); - list_add(&BTRFS_I(inode)->i_orphan, &root->orphan_list); - spin_unlock(&root->orphan_lock); - - /* - * if this is a bad inode, means we actually succeeded in - * removing the inode, but not the orphan record, which means - * we need to manually delete the orphan since iput will just - * do a destroy_inode + * Inode is already gone but the orphan item is still there, + * kill the orphan item. */ - if (is_bad_inode(inode)) { - trans = btrfs_start_transaction(root, 0); + if (ret == -ESTALE) { + trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto out; } - btrfs_orphan_del(trans, inode); + ret = btrfs_del_orphan_item(trans, root, + found_key.objectid); + BUG_ON(ret); btrfs_end_transaction(trans, root); - iput(inode); continue; } + /* + * add this inode to the orphan list so btrfs_orphan_del does + * the proper thing when we hit it + */ + spin_lock(&root->orphan_lock); + list_add(&BTRFS_I(inode)->i_orphan, &root->orphan_list); + spin_unlock(&root->orphan_lock); + /* if we have links, this was a truncate, lets do that */ if (inode->i_nlink) { if (!S_ISREG(inode->i_mode)) { @@ -2835,7 +2762,16 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, u64 ino = btrfs_ino(inode); u64 dir_ino = btrfs_ino(dir); - trans = btrfs_start_transaction(root, 10); + /* + * 1 for the possible orphan item + * 1 for the dir item + * 1 for the dir index + * 1 for the inode ref + * 1 for the inode ref in the tree log + * 2 for the dir entries in the log + * 1 for the inode + */ + trans = btrfs_start_transaction(root, 8); if (!IS_ERR(trans) || PTR_ERR(trans) != -ENOSPC) return trans; @@ -2858,7 +2794,8 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, return ERR_PTR(-ENOMEM); } - trans = btrfs_start_transaction(root, 0); + /* 1 for the orphan item */ + trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { btrfs_free_path(path); root->fs_info->enospc_unlink = 0; @@ -2963,6 +2900,12 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, err = 0; out: btrfs_free_path(path); + /* Migrate the orphan reservation over */ + if (!err) + err = btrfs_block_rsv_migrate(trans->block_rsv, + &root->fs_info->global_block_rsv, + trans->bytes_reserved); + if (err) { btrfs_end_transaction(trans, root); root->fs_info->enospc_unlink = 0; @@ -2977,6 +2920,9 @@ static void __unlink_end_trans(struct btrfs_trans_handle *trans, struct btrfs_root *root) { if (trans->block_rsv == &root->fs_info->global_block_rsv) { + btrfs_block_rsv_release(root, trans->block_rsv, + trans->bytes_reserved); + trans->block_rsv = &root->fs_info->trans_block_rsv; BUG_ON(!root->fs_info->enospc_unlink); root->fs_info->enospc_unlink = 0; } @@ -3368,6 +3314,7 @@ static int btrfs_truncate_page(struct address_space *mapping, loff_t from) pgoff_t index = from >> PAGE_CACHE_SHIFT; unsigned offset = from & (PAGE_CACHE_SIZE-1); struct page *page; + gfp_t mask = btrfs_alloc_write_mask(mapping); int ret = 0; u64 page_start; u64 page_end; @@ -3380,7 +3327,7 @@ static int btrfs_truncate_page(struct address_space *mapping, loff_t from) ret = -ENOMEM; again: - page = find_or_create_page(mapping, index, GFP_NOFS); + page = find_or_create_page(mapping, index, mask); if (!page) { btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE); goto out; @@ -3613,6 +3560,8 @@ void btrfs_evict_inode(struct inode *inode) { struct btrfs_trans_handle *trans; struct btrfs_root *root = BTRFS_I(inode)->root; + struct btrfs_block_rsv *rsv, *global_rsv; + u64 min_size = btrfs_calc_trunc_metadata_size(root, 1); unsigned long nr; int ret; @@ -3640,22 +3589,55 @@ void btrfs_evict_inode(struct inode *inode) goto no_delete; } + rsv = btrfs_alloc_block_rsv(root); + if (!rsv) { + btrfs_orphan_del(NULL, inode); + goto no_delete; + } + rsv->size = min_size; + global_rsv = &root->fs_info->global_block_rsv; + btrfs_i_size_write(inode, 0); + /* + * This is a bit simpler than btrfs_truncate since + * + * 1) We've already reserved our space for our orphan item in the + * unlink. + * 2) We're going to delete the inode item, so we don't need to update + * it at all. + * + * So we just need to reserve some slack space in case we add bytes when + * doing the truncate. + */ while (1) { - trans = btrfs_join_transaction(root); - BUG_ON(IS_ERR(trans)); - trans->block_rsv = root->orphan_block_rsv; + ret = btrfs_block_rsv_refill(root, rsv, min_size); + + /* + * Try and steal from the global reserve since we will + * likely not use this space anyway, we want to try as + * hard as possible to get this to work. + */ + if (ret) + ret = btrfs_block_rsv_migrate(global_rsv, rsv, min_size); - ret = btrfs_block_rsv_check(trans, root, - root->orphan_block_rsv, 0, 5); if (ret) { - BUG_ON(ret != -EAGAIN); - ret = btrfs_commit_transaction(trans, root); - BUG_ON(ret); - continue; + printk(KERN_WARNING "Could not get space for a " + "delete, will truncate on mount %d\n", ret); + btrfs_orphan_del(NULL, inode); + btrfs_free_block_rsv(root, rsv); + goto no_delete; + } + + trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + btrfs_orphan_del(NULL, inode); + btrfs_free_block_rsv(root, rsv); + goto no_delete; } + trans->block_rsv = rsv; + ret = btrfs_truncate_inode_items(trans, root, inode, 0, 0); if (ret != -EAGAIN) break; @@ -3664,14 +3646,17 @@ void btrfs_evict_inode(struct inode *inode) btrfs_end_transaction(trans, root); trans = NULL; btrfs_btree_balance_dirty(root, nr); - } + btrfs_free_block_rsv(root, rsv); + if (ret == 0) { + trans->block_rsv = root->orphan_block_rsv; ret = btrfs_orphan_del(trans, inode); BUG_ON(ret); } + trans->block_rsv = &root->fs_info->trans_block_rsv; if (!(root == root->fs_info->tree_root || root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)) btrfs_return_ino(root, btrfs_ino(inode)); @@ -5795,8 +5780,7 @@ again: if (test_bit(BTRFS_ORDERED_NOCOW, &ordered->flags)) { ret = btrfs_ordered_update_i_size(inode, 0, ordered); if (!ret) - ret = btrfs_update_inode(trans, root, inode); - err = ret; + err = btrfs_update_inode(trans, root, inode); goto out; } @@ -6541,6 +6525,7 @@ static int btrfs_truncate(struct inode *inode) struct btrfs_trans_handle *trans; unsigned long nr; u64 mask = root->sectorsize - 1; + u64 min_size = btrfs_calc_trunc_metadata_size(root, 1); ret = btrfs_truncate_page(inode->i_mapping, inode->i_size); if (ret) @@ -6588,19 +6573,23 @@ static int btrfs_truncate(struct inode *inode) rsv = btrfs_alloc_block_rsv(root); if (!rsv) return -ENOMEM; - btrfs_add_durable_block_rsv(root->fs_info, rsv); + rsv->size = min_size; + /* + * 1 for the truncate slack space + * 1 for the orphan item we're going to add + * 1 for the orphan item deletion + * 1 for updating the inode. + */ trans = btrfs_start_transaction(root, 4); if (IS_ERR(trans)) { err = PTR_ERR(trans); goto out; } - /* - * Reserve space for the truncate process. Truncate should be adding - * space, but if there are snapshots it may end up using space. - */ - ret = btrfs_truncate_reserve_metadata(trans, root, rsv); + /* Migrate the slack space for the truncate to our reserve */ + ret = btrfs_block_rsv_migrate(&root->fs_info->trans_block_rsv, rsv, + min_size); BUG_ON(ret); ret = btrfs_orphan_add(trans, inode); @@ -6609,21 +6598,6 @@ static int btrfs_truncate(struct inode *inode) goto out; } - nr = trans->blocks_used; - btrfs_end_transaction(trans, root); - btrfs_btree_balance_dirty(root, nr); - - /* - * Ok so we've already migrated our bytes over for the truncate, so here - * just reserve the one slot we need for updating the inode. - */ - trans = btrfs_start_transaction(root, 1); - if (IS_ERR(trans)) { - err = PTR_ERR(trans); - goto out; - } - trans->block_rsv = rsv; - /* * setattr is responsible for setting the ordered_data_close flag, * but that is only tested during the last file release. That @@ -6645,20 +6619,30 @@ static int btrfs_truncate(struct inode *inode) btrfs_add_ordered_operation(trans, root, inode); while (1) { + ret = btrfs_block_rsv_refill(root, rsv, min_size); + if (ret) { + /* + * This can only happen with the original transaction we + * started above, every other time we shouldn't have a + * transaction started yet. + */ + if (ret == -EAGAIN) + goto end_trans; + err = ret; + break; + } + if (!trans) { - trans = btrfs_start_transaction(root, 3); + /* Just need the 1 for updating the inode */ + trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { err = PTR_ERR(trans); goto out; } - - ret = btrfs_truncate_reserve_metadata(trans, root, - rsv); - BUG_ON(ret); - - trans->block_rsv = rsv; } + trans->block_rsv = rsv; + ret = btrfs_truncate_inode_items(trans, root, inode, inode->i_size, BTRFS_EXTENT_DATA_KEY); @@ -6673,7 +6657,7 @@ static int btrfs_truncate(struct inode *inode) err = ret; break; } - +end_trans: nr = trans->blocks_used; btrfs_end_transaction(trans, root); trans = NULL; @@ -6755,9 +6739,9 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) ei->last_sub_trans = 0; ei->logged_trans = 0; ei->delalloc_bytes = 0; - ei->reserved_bytes = 0; ei->disk_i_size = 0; ei->flags = 0; + ei->csum_bytes = 0; ei->index_cnt = (u64)-1; ei->last_unlink_trans = 0; @@ -6803,6 +6787,8 @@ void btrfs_destroy_inode(struct inode *inode) WARN_ON(inode->i_data.nrpages); WARN_ON(BTRFS_I(inode)->outstanding_extents); WARN_ON(BTRFS_I(inode)->reserved_extents); + WARN_ON(BTRFS_I(inode)->delalloc_bytes); + WARN_ON(BTRFS_I(inode)->csum_bytes); /* * This can happen where we create an inode, but somebody else also diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 538f65a79ec5..8f6e14279409 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -117,7 +117,7 @@ void btrfs_update_iflags(struct inode *inode) /* * Inherit flags from the parent inode. * - * Unlike extN we don't have any flags we don't want to inherit currently. + * Currently only the compression flags and the cow flags are inherited. */ void btrfs_inherit_iflags(struct inode *inode, struct inode *dir) { @@ -128,12 +128,17 @@ void btrfs_inherit_iflags(struct inode *inode, struct inode *dir) flags = BTRFS_I(dir)->flags; - if (S_ISREG(inode->i_mode)) - flags &= ~BTRFS_INODE_DIRSYNC; - else if (!S_ISDIR(inode->i_mode)) - flags &= (BTRFS_INODE_NODUMP | BTRFS_INODE_NOATIME); + if (flags & BTRFS_INODE_NOCOMPRESS) { + BTRFS_I(inode)->flags &= ~BTRFS_INODE_COMPRESS; + BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS; + } else if (flags & BTRFS_INODE_COMPRESS) { + BTRFS_I(inode)->flags &= ~BTRFS_INODE_NOCOMPRESS; + BTRFS_I(inode)->flags |= BTRFS_INODE_COMPRESS; + } + + if (flags & BTRFS_INODE_NODATACOW) + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; - BTRFS_I(inode)->flags = flags; btrfs_update_iflags(inode); } @@ -277,6 +282,7 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) struct fstrim_range range; u64 minlen = ULLONG_MAX; u64 num_devices = 0; + u64 total_bytes = btrfs_super_total_bytes(root->fs_info->super_copy); int ret; if (!capable(CAP_SYS_ADMIN)) @@ -295,12 +301,15 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) } } rcu_read_unlock(); + if (!num_devices) return -EOPNOTSUPP; - if (copy_from_user(&range, arg, sizeof(range))) return -EFAULT; + if (range.start > total_bytes) + return -EINVAL; + range.len = min(range.len, total_bytes - range.start); range.minlen = max(range.minlen, minlen); ret = btrfs_trim_fs(root, &range); if (ret < 0) @@ -760,7 +769,7 @@ static int should_defrag_range(struct inode *inode, u64 start, u64 len, int ret = 1; /* - * make sure that once we start defragging and extent, we keep on + * make sure that once we start defragging an extent, we keep on * defragging it */ if (start < *defrag_end) @@ -805,7 +814,6 @@ static int should_defrag_range(struct inode *inode, u64 start, u64 len, * extent will force at least part of that big extent to be defragged. */ if (ret) { - *last_len += len; *defrag_end = extent_map_end(em); } else { *last_len = 0; @@ -843,6 +851,7 @@ static int cluster_pages_for_defrag(struct inode *inode, int i_done; struct btrfs_ordered_extent *ordered; struct extent_state *cached_state = NULL; + gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping); if (isize == 0) return 0; @@ -860,7 +869,7 @@ again: for (i = 0; i < num_pages; i++) { struct page *page; page = find_or_create_page(inode->i_mapping, - start_index + i, GFP_NOFS); + start_index + i, mask); if (!page) break; @@ -972,18 +981,20 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, struct btrfs_super_block *disk_super; struct file_ra_state *ra = NULL; unsigned long last_index; + u64 isize = i_size_read(inode); u64 features; u64 last_len = 0; u64 skip = 0; u64 defrag_end = 0; u64 newer_off = range->start; - int newer_left = 0; unsigned long i; + unsigned long ra_index = 0; int ret; int defrag_count = 0; int compress_type = BTRFS_COMPRESS_ZLIB; int extent_thresh = range->extent_thresh; - int newer_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT; + int max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT; + int cluster = max_cluster; u64 new_align = ~((u64)128 * 1024 - 1); struct page **pages = NULL; @@ -997,7 +1008,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, compress_type = range->compress_type; } - if (inode->i_size == 0) + if (isize == 0) return 0; /* @@ -1013,7 +1024,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, ra = &file->f_ra; } - pages = kmalloc(sizeof(struct page *) * newer_cluster, + pages = kmalloc(sizeof(struct page *) * max_cluster, GFP_NOFS); if (!pages) { ret = -ENOMEM; @@ -1022,10 +1033,10 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, /* find the last page to defrag */ if (range->start + range->len > range->start) { - last_index = min_t(u64, inode->i_size - 1, + last_index = min_t(u64, isize - 1, range->start + range->len - 1) >> PAGE_CACHE_SHIFT; } else { - last_index = (inode->i_size - 1) >> PAGE_CACHE_SHIFT; + last_index = (isize - 1) >> PAGE_CACHE_SHIFT; } if (newer_than) { @@ -1038,16 +1049,24 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, * the extents in the file evenly spaced */ i = (newer_off & new_align) >> PAGE_CACHE_SHIFT; - newer_left = newer_cluster; } else goto out_ra; } else { i = range->start >> PAGE_CACHE_SHIFT; } if (!max_to_defrag) - max_to_defrag = last_index - 1; + max_to_defrag = last_index; + + /* + * make writeback starts from i, so the defrag range can be + * written sequentially. + */ + if (i < inode->i_mapping->writeback_index) + inode->i_mapping->writeback_index = i; - while (i <= last_index && defrag_count < max_to_defrag) { + while (i <= last_index && defrag_count < max_to_defrag && + (i < (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> + PAGE_CACHE_SHIFT)) { /* * make sure we stop running if someone unmounts * the FS @@ -1070,18 +1089,31 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, i = max(i + 1, next); continue; } + + if (!newer_than) { + cluster = (PAGE_CACHE_ALIGN(defrag_end) >> + PAGE_CACHE_SHIFT) - i; + cluster = min(cluster, max_cluster); + } else { + cluster = max_cluster; + } + if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) BTRFS_I(inode)->force_compress = compress_type; - btrfs_force_ra(inode->i_mapping, ra, file, i, newer_cluster); + if (i + cluster > ra_index) { + ra_index = max(i, ra_index); + btrfs_force_ra(inode->i_mapping, ra, file, ra_index, + cluster); + ra_index += max_cluster; + } - ret = cluster_pages_for_defrag(inode, pages, i, newer_cluster); + ret = cluster_pages_for_defrag(inode, pages, i, cluster); if (ret < 0) goto out_ra; defrag_count += ret; balance_dirty_pages_ratelimited_nr(inode->i_mapping, ret); - i += ret; if (newer_than) { if (newer_off == (u64)-1) @@ -1096,12 +1128,17 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, if (!ret) { range->start = newer_off; i = (newer_off & new_align) >> PAGE_CACHE_SHIFT; - newer_left = newer_cluster; } else { break; } } else { - i++; + if (ret > 0) { + i += ret; + last_len += ret << PAGE_CACHE_SHIFT; + } else { + i++; + last_len = 0; + } } } @@ -1127,16 +1164,14 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, mutex_unlock(&inode->i_mutex); } - disk_super = &root->fs_info->super_copy; + disk_super = root->fs_info->super_copy; features = btrfs_super_incompat_flags(disk_super); if (range->compress_type == BTRFS_COMPRESS_LZO) { features |= BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO; btrfs_set_super_incompat_flags(disk_super, features); } - if (!file) - kfree(ra); - return defrag_count; + ret = defrag_count; out_ra: if (!file) @@ -2578,7 +2613,7 @@ static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp) return PTR_ERR(trans); } - dir_id = btrfs_super_root_dir(&root->fs_info->super_copy); + dir_id = btrfs_super_root_dir(root->fs_info->super_copy); di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path, dir_id, "default", 7, 1); if (IS_ERR_OR_NULL(di)) { @@ -2594,7 +2629,7 @@ static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp) btrfs_mark_buffer_dirty(path->nodes[0]); btrfs_free_path(path); - disk_super = &root->fs_info->super_copy; + disk_super = root->fs_info->super_copy; features = btrfs_super_incompat_flags(disk_super); if (!(features & BTRFS_FEATURE_INCOMPAT_DEFAULT_SUBVOL)) { features |= BTRFS_FEATURE_INCOMPAT_DEFAULT_SUBVOL; diff --git a/fs/btrfs/print-tree.c b/fs/btrfs/print-tree.c index fb2605d998e9..f38e452486b8 100644 --- a/fs/btrfs/print-tree.c +++ b/fs/btrfs/print-tree.c @@ -158,8 +158,7 @@ static void print_extent_ref_v0(struct extent_buffer *eb, int slot) void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l) { int i; - u32 type; - u32 nr = btrfs_header_nritems(l); + u32 type, nr; struct btrfs_item *item; struct btrfs_root_item *ri; struct btrfs_dir_item *di; @@ -172,6 +171,11 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l) struct btrfs_key key; struct btrfs_key found_key; + if (!l) + return; + + nr = btrfs_header_nritems(l); + printk(KERN_INFO "leaf %llu total ptrs %d free space %d\n", (unsigned long long)btrfs_header_bytenr(l), nr, btrfs_leaf_free_space(root, l)); diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 59bb1764273d..24d654ce7a06 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -2041,8 +2041,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc, BUG_ON(IS_ERR(trans)); trans->block_rsv = rc->block_rsv; - ret = btrfs_block_rsv_check(trans, root, rc->block_rsv, - min_reserved, 0); + ret = btrfs_block_rsv_refill(root, rc->block_rsv, min_reserved); if (ret) { BUG_ON(ret != -EAGAIN); ret = btrfs_commit_transaction(trans, root); @@ -2152,8 +2151,7 @@ int prepare_to_merge(struct reloc_control *rc, int err) again: if (!err) { num_bytes = rc->merging_rsv_size; - ret = btrfs_block_rsv_add(NULL, root, rc->block_rsv, - num_bytes); + ret = btrfs_block_rsv_add(root, rc->block_rsv, num_bytes); if (ret) err = ret; } @@ -2427,7 +2425,7 @@ static int reserve_metadata_space(struct btrfs_trans_handle *trans, num_bytes = calcu_metadata_size(rc, node, 1) * 2; trans->block_rsv = rc->block_rsv; - ret = btrfs_block_rsv_add(trans, root, rc->block_rsv, num_bytes); + ret = btrfs_block_rsv_add(root, rc->block_rsv, num_bytes); if (ret) { if (ret == -EAGAIN) rc->commit_transaction = 1; @@ -2922,6 +2920,7 @@ static int relocate_file_extent_cluster(struct inode *inode, unsigned long last_index; struct page *page; struct file_ra_state *ra; + gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping); int nr = 0; int ret = 0; @@ -2956,7 +2955,7 @@ static int relocate_file_extent_cluster(struct inode *inode, ra, NULL, index, last_index + 1 - index); page = find_or_create_page(inode->i_mapping, index, - GFP_NOFS); + mask); if (!page) { btrfs_delalloc_release_metadata(inode, PAGE_CACHE_SIZE); @@ -3323,8 +3322,11 @@ static int find_data_references(struct reloc_control *rc, } key.objectid = ref_objectid; - key.offset = ref_offset; key.type = BTRFS_EXTENT_DATA_KEY; + if (ref_offset > ((u64)-1 << 32)) + key.offset = 0; + else + key.offset = ref_offset; path->search_commit_root = 1; path->skip_locking = 1; @@ -3645,14 +3647,11 @@ int prepare_to_relocate(struct reloc_control *rc) * btrfs_init_reloc_root will use them when there * is no reservation in transaction handle. */ - ret = btrfs_block_rsv_add(NULL, rc->extent_root, rc->block_rsv, + ret = btrfs_block_rsv_add(rc->extent_root, rc->block_rsv, rc->extent_root->nodesize * 256); if (ret) return ret; - rc->block_rsv->refill_used = 1; - btrfs_add_durable_block_rsv(rc->extent_root->fs_info, rc->block_rsv); - memset(&rc->cluster, 0, sizeof(rc->cluster)); rc->search_start = rc->block_group->key.objectid; rc->extents_found = 0; @@ -3777,8 +3776,7 @@ restart: } } - ret = btrfs_block_rsv_check(trans, rc->extent_root, - rc->block_rsv, 0, 5); + ret = btrfs_block_rsv_check(rc->extent_root, rc->block_rsv, 5); if (ret < 0) { if (ret != -EAGAIN) { err = ret; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index f930f2776589..5bc4ec827b3d 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -179,7 +179,7 @@ struct scrub_dev *scrub_setup_dev(struct btrfs_device *dev) sdev->curr = -1; atomic_set(&sdev->in_flight, 0); atomic_set(&sdev->cancel_req, 0); - sdev->csum_size = btrfs_super_csum_size(&fs_info->super_copy); + sdev->csum_size = btrfs_super_csum_size(fs_info->super_copy); INIT_LIST_HEAD(&sdev->csum_list); spin_lock_init(&sdev->list_lock); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 15634d4648d7..57080dffdfc6 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -40,6 +40,7 @@ #include <linux/magic.h> #include <linux/slab.h> #include <linux/cleancache.h> +#include <linux/mnt_namespace.h> #include "compat.h" #include "delayed-inode.h" #include "ctree.h" @@ -58,6 +59,7 @@ #include <trace/events/btrfs.h> static const struct super_operations btrfs_super_ops; +static struct file_system_type btrfs_fs_type; static const char *btrfs_decode_error(struct btrfs_fs_info *fs_info, int errno, char nbuf[16]) @@ -162,7 +164,7 @@ enum { Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, - Opt_inode_cache, Opt_err, + Opt_inode_cache, Opt_no_space_cache, Opt_recovery, Opt_err, }; static match_table_t tokens = { @@ -195,6 +197,8 @@ static match_table_t tokens = { {Opt_subvolrootid, "subvolrootid=%d"}, {Opt_defrag, "autodefrag"}, {Opt_inode_cache, "inode_cache"}, + {Opt_no_space_cache, "no_space_cache"}, + {Opt_recovery, "recovery"}, {Opt_err, NULL}, }; @@ -206,14 +210,19 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) { struct btrfs_fs_info *info = root->fs_info; substring_t args[MAX_OPT_ARGS]; - char *p, *num, *orig; + char *p, *num, *orig = NULL; + u64 cache_gen; int intarg; int ret = 0; char *compress_type; bool compress_force = false; + cache_gen = btrfs_super_cache_generation(root->fs_info->super_copy); + if (cache_gen) + btrfs_set_opt(info->mount_opt, SPACE_CACHE); + if (!options) - return 0; + goto out; /* * strsep changes the string, duplicate it because parse_options @@ -360,9 +369,12 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) btrfs_set_opt(info->mount_opt, DISCARD); break; case Opt_space_cache: - printk(KERN_INFO "btrfs: enabling disk space caching\n"); btrfs_set_opt(info->mount_opt, SPACE_CACHE); break; + case Opt_no_space_cache: + printk(KERN_INFO "btrfs: disabling disk space caching\n"); + btrfs_clear_opt(info->mount_opt, SPACE_CACHE); + break; case Opt_inode_cache: printk(KERN_INFO "btrfs: enabling inode map caching\n"); btrfs_set_opt(info->mount_opt, INODE_MAP_CACHE); @@ -381,6 +393,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) printk(KERN_INFO "btrfs: enabling auto defrag"); btrfs_set_opt(info->mount_opt, AUTO_DEFRAG); break; + case Opt_recovery: + printk(KERN_INFO "btrfs: enabling auto recovery"); + btrfs_set_opt(info->mount_opt, RECOVERY); + break; case Opt_err: printk(KERN_INFO "btrfs: unrecognized mount option " "'%s'\n", p); @@ -391,6 +407,8 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) } } out: + if (!ret && btrfs_test_opt(root, SPACE_CACHE)) + printk(KERN_INFO "btrfs: disk space caching is enabled\n"); kfree(orig); return ret; } @@ -406,12 +424,12 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, u64 *subvol_rootid, struct btrfs_fs_devices **fs_devices) { substring_t args[MAX_OPT_ARGS]; - char *opts, *orig, *p; + char *device_name, *opts, *orig, *p; int error = 0; int intarg; if (!options) - goto out; + return 0; /* * strsep changes the string, duplicate it because parse_options @@ -457,29 +475,24 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, } break; case Opt_device: - error = btrfs_scan_one_device(match_strdup(&args[0]), + device_name = match_strdup(&args[0]); + if (!device_name) { + error = -ENOMEM; + goto out; + } + error = btrfs_scan_one_device(device_name, flags, holder, fs_devices); + kfree(device_name); if (error) - goto out_free_opts; + goto out; break; default: break; } } - out_free_opts: +out: kfree(orig); - out: - /* - * If no subvolume name is specified we use the default one. Allocate - * a copy of the string "." here so that code later in the - * mount path doesn't care if it's the default volume or another one. - */ - if (!*subvol_name) { - *subvol_name = kstrdup(".", GFP_KERNEL); - if (!*subvol_name) - return -ENOMEM; - } return error; } @@ -492,7 +505,6 @@ static struct dentry *get_default_root(struct super_block *sb, struct btrfs_path *path; struct btrfs_key location; struct inode *inode; - struct dentry *dentry; u64 dir_id; int new = 0; @@ -517,7 +529,7 @@ static struct dentry *get_default_root(struct super_block *sb, * will mount by default if we haven't been given a specific subvolume * to mount. */ - dir_id = btrfs_super_root_dir(&root->fs_info->super_copy); + dir_id = btrfs_super_root_dir(root->fs_info->super_copy); di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0); if (IS_ERR(di)) { btrfs_free_path(path); @@ -566,29 +578,7 @@ setup_root: return dget(sb->s_root); } - if (new) { - const struct qstr name = { .name = "/", .len = 1 }; - - /* - * New inode, we need to make the dentry a sibling of s_root so - * everything gets cleaned up properly on unmount. - */ - dentry = d_alloc(sb->s_root, &name); - if (!dentry) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - d_splice_alias(inode, dentry); - } else { - /* - * We found the inode in cache, just find a dentry for it and - * put the reference to the inode we just got. - */ - dentry = d_find_alias(inode); - iput(inode); - } - - return dentry; + return d_obtain_alias(inode); } static int btrfs_fill_super(struct super_block *sb, @@ -719,6 +709,8 @@ static int btrfs_show_options(struct seq_file *seq, struct vfsmount *vfs) seq_puts(seq, ",noacl"); if (btrfs_test_opt(root, SPACE_CACHE)) seq_puts(seq, ",space_cache"); + else + seq_puts(seq, ",no_space_cache"); if (btrfs_test_opt(root, CLEAR_CACHE)) seq_puts(seq, ",clear_cache"); if (btrfs_test_opt(root, USER_SUBVOL_RM_ALLOWED)) @@ -753,6 +745,137 @@ static int btrfs_set_super(struct super_block *s, void *data) return set_anon_super(s, data); } +/* + * subvolumes are identified by ino 256 + */ +static inline int is_subvolume_inode(struct inode *inode) +{ + if (inode && inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) + return 1; + return 0; +} + +/* + * This will strip out the subvol=%s argument for an argument string and add + * subvolid=0 to make sure we get the actual tree root for path walking to the + * subvol we want. + */ +static char *setup_root_args(char *args) +{ + unsigned copied = 0; + unsigned len = strlen(args) + 2; + char *pos; + char *ret; + + /* + * We need the same args as before, but minus + * + * subvol=a + * + * and add + * + * subvolid=0 + * + * which is a difference of 2 characters, so we allocate strlen(args) + + * 2 characters. + */ + ret = kzalloc(len * sizeof(char), GFP_NOFS); + if (!ret) + return NULL; + pos = strstr(args, "subvol="); + + /* This shouldn't happen, but just in case.. */ + if (!pos) { + kfree(ret); + return NULL; + } + + /* + * The subvol=<> arg is not at the front of the string, copy everybody + * up to that into ret. + */ + if (pos != args) { + *pos = '\0'; + strcpy(ret, args); + copied += strlen(args); + pos++; + } + + strncpy(ret + copied, "subvolid=0", len - copied); + + /* Length of subvolid=0 */ + copied += 10; + + /* + * If there is no , after the subvol= option then we know there's no + * other options and we can just return. + */ + pos = strchr(pos, ','); + if (!pos) + return ret; + + /* Copy the rest of the arguments into our buffer */ + strncpy(ret + copied, pos, len - copied); + copied += strlen(pos); + + return ret; +} + +static struct dentry *mount_subvol(const char *subvol_name, int flags, + const char *device_name, char *data) +{ + struct super_block *s; + struct dentry *root; + struct vfsmount *mnt; + struct mnt_namespace *ns_private; + char *newargs; + struct path path; + int error; + + newargs = setup_root_args(data); + if (!newargs) + return ERR_PTR(-ENOMEM); + mnt = vfs_kern_mount(&btrfs_fs_type, flags, device_name, + newargs); + kfree(newargs); + if (IS_ERR(mnt)) + return ERR_CAST(mnt); + + ns_private = create_mnt_ns(mnt); + if (IS_ERR(ns_private)) { + mntput(mnt); + return ERR_CAST(ns_private); + } + + /* + * This will trigger the automount of the subvol so we can just + * drop the mnt we have here and return the dentry that we + * found. + */ + error = vfs_path_lookup(mnt->mnt_root, mnt, subvol_name, + LOOKUP_FOLLOW, &path); + put_mnt_ns(ns_private); + if (error) + return ERR_PTR(error); + + if (!is_subvolume_inode(path.dentry->d_inode)) { + path_put(&path); + mntput(mnt); + error = -EINVAL; + printk(KERN_ERR "btrfs: '%s' is not a valid subvolume\n", + subvol_name); + return ERR_PTR(-EINVAL); + } + + /* Get a ref to the sb and the dentry we found and return it */ + s = path.mnt->mnt_sb; + atomic_inc(&s->s_active); + root = dget(path.dentry); + path_put(&path); + down_write(&s->s_umount); + + return root; +} /* * Find a superblock for the given device / mount point. @@ -784,13 +907,19 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, if (error) return ERR_PTR(error); + if (subvol_name) { + root = mount_subvol(subvol_name, flags, device_name, data); + kfree(subvol_name); + return root; + } + error = btrfs_scan_one_device(device_name, mode, fs_type, &fs_devices); if (error) - goto error_free_subvol_name; + return ERR_PTR(error); error = btrfs_open_devices(fs_devices, mode, fs_type); if (error) - goto error_free_subvol_name; + return ERR_PTR(error); if (!(flags & MS_RDONLY) && fs_devices->rw_devices == 0) { error = -EACCES; @@ -813,88 +942,57 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, fs_info->fs_devices = fs_devices; tree_root->fs_info = fs_info; + fs_info->super_copy = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS); + fs_info->super_for_commit = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS); + if (!fs_info->super_copy || !fs_info->super_for_commit) { + error = -ENOMEM; + goto error_close_devices; + } + bdev = fs_devices->latest_bdev; s = sget(fs_type, btrfs_test_super, btrfs_set_super, tree_root); - if (IS_ERR(s)) - goto error_s; + if (IS_ERR(s)) { + error = PTR_ERR(s); + goto error_close_devices; + } if (s->s_root) { if ((flags ^ s->s_flags) & MS_RDONLY) { deactivate_locked_super(s); - error = -EBUSY; - goto error_close_devices; + return ERR_PTR(-EBUSY); } btrfs_close_devices(fs_devices); - kfree(fs_info); + free_fs_info(fs_info); kfree(tree_root); } else { char b[BDEVNAME_SIZE]; s->s_flags = flags | MS_NOSEC; strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id)); + btrfs_sb(s)->fs_info->bdev_holder = fs_type; error = btrfs_fill_super(s, fs_devices, data, flags & MS_SILENT ? 1 : 0); if (error) { deactivate_locked_super(s); - goto error_free_subvol_name; + return ERR_PTR(error); } - btrfs_sb(s)->fs_info->bdev_holder = fs_type; s->s_flags |= MS_ACTIVE; } - /* if they gave us a subvolume name bind mount into that */ - if (strcmp(subvol_name, ".")) { - struct dentry *new_root; - - root = get_default_root(s, subvol_rootid); - if (IS_ERR(root)) { - error = PTR_ERR(root); - deactivate_locked_super(s); - goto error_free_subvol_name; - } - - mutex_lock(&root->d_inode->i_mutex); - new_root = lookup_one_len(subvol_name, root, - strlen(subvol_name)); - mutex_unlock(&root->d_inode->i_mutex); - - if (IS_ERR(new_root)) { - dput(root); - deactivate_locked_super(s); - error = PTR_ERR(new_root); - goto error_free_subvol_name; - } - if (!new_root->d_inode) { - dput(root); - dput(new_root); - deactivate_locked_super(s); - error = -ENXIO; - goto error_free_subvol_name; - } - dput(root); - root = new_root; - } else { - root = get_default_root(s, subvol_objectid); - if (IS_ERR(root)) { - error = PTR_ERR(root); - deactivate_locked_super(s); - goto error_free_subvol_name; - } + root = get_default_root(s, subvol_objectid); + if (IS_ERR(root)) { + deactivate_locked_super(s); + return root; } - kfree(subvol_name); return root; -error_s: - error = PTR_ERR(s); error_close_devices: btrfs_close_devices(fs_devices); - kfree(fs_info); + free_fs_info(fs_info); kfree(tree_root); -error_free_subvol_name: - kfree(subvol_name); return ERR_PTR(error); } @@ -919,7 +1017,7 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) if (root->fs_info->fs_devices->rw_devices == 0) return -EACCES; - if (btrfs_super_log_root(&root->fs_info->super_copy) != 0) + if (btrfs_super_log_root(root->fs_info->super_copy) != 0) return -EINVAL; ret = btrfs_cleanup_fs_roots(root->fs_info); @@ -1085,7 +1183,7 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes) static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) { struct btrfs_root *root = btrfs_sb(dentry->d_sb); - struct btrfs_super_block *disk_super = &root->fs_info->super_copy; + struct btrfs_super_block *disk_super = root->fs_info->super_copy; struct list_head *head = &root->fs_info->space_info; struct btrfs_space_info *found; u64 total_used = 0; diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index e24b7964a155..29f782cc2cc9 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -275,7 +275,7 @@ static struct btrfs_trans_handle *start_transaction(struct btrfs_root *root, */ if (num_items > 0 && root != root->fs_info->chunk_root) { num_bytes = btrfs_calc_trans_metadata_size(root, num_items); - ret = btrfs_block_rsv_add(NULL, root, + ret = btrfs_block_rsv_add(root, &root->fs_info->trans_block_rsv, num_bytes); if (ret) @@ -418,8 +418,8 @@ static int should_end_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root) { int ret; - ret = btrfs_block_rsv_check(trans, root, - &root->fs_info->global_block_rsv, 0, 5); + + ret = btrfs_block_rsv_check(root, &root->fs_info->global_block_rsv, 5); return ret ? 1 : 0; } @@ -427,17 +427,26 @@ int btrfs_should_end_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root) { struct btrfs_transaction *cur_trans = trans->transaction; + struct btrfs_block_rsv *rsv = trans->block_rsv; int updates; smp_mb(); if (cur_trans->blocked || cur_trans->delayed_refs.flushing) return 1; + /* + * We need to do this in case we're deleting csums so the global block + * rsv get's used instead of the csum block rsv. + */ + trans->block_rsv = NULL; + updates = trans->delayed_ref_updates; trans->delayed_ref_updates = 0; if (updates) btrfs_run_delayed_refs(trans, root, updates); + trans->block_rsv = rsv; + return should_end_transaction(trans, root); } @@ -453,6 +462,8 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, return 0; } + btrfs_trans_release_metadata(trans, root); + trans->block_rsv = NULL; while (count < 4) { unsigned long cur = trans->delayed_ref_updates; trans->delayed_ref_updates = 0; @@ -473,8 +484,6 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, count++; } - btrfs_trans_release_metadata(trans, root); - if (lock && !atomic_read(&root->fs_info->open_ioctl_trans) && should_end_transaction(trans, root)) { trans->transaction->blocked = 1; @@ -562,50 +571,21 @@ int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans, int btrfs_write_marked_extents(struct btrfs_root *root, struct extent_io_tree *dirty_pages, int mark) { - int ret; int err = 0; int werr = 0; - struct page *page; - struct inode *btree_inode = root->fs_info->btree_inode; + struct address_space *mapping = root->fs_info->btree_inode->i_mapping; u64 start = 0; u64 end; - unsigned long index; - while (1) { - ret = find_first_extent_bit(dirty_pages, start, &start, &end, - mark); - if (ret) - break; - while (start <= end) { - cond_resched(); - - index = start >> PAGE_CACHE_SHIFT; - start = (u64)(index + 1) << PAGE_CACHE_SHIFT; - page = find_get_page(btree_inode->i_mapping, index); - if (!page) - continue; - - btree_lock_page_hook(page); - if (!page->mapping) { - unlock_page(page); - page_cache_release(page); - continue; - } - - if (PageWriteback(page)) { - if (PageDirty(page)) - wait_on_page_writeback(page); - else { - unlock_page(page); - page_cache_release(page); - continue; - } - } - err = write_one_page(page, 0); - if (err) - werr = err; - page_cache_release(page); - } + while (!find_first_extent_bit(dirty_pages, start, &start, &end, + mark)) { + convert_extent_bit(dirty_pages, start, end, EXTENT_NEED_WAIT, mark, + GFP_NOFS); + err = filemap_fdatawrite_range(mapping, start, end); + if (err) + werr = err; + cond_resched(); + start = end + 1; } if (err) werr = err; @@ -621,39 +601,20 @@ int btrfs_write_marked_extents(struct btrfs_root *root, int btrfs_wait_marked_extents(struct btrfs_root *root, struct extent_io_tree *dirty_pages, int mark) { - int ret; int err = 0; int werr = 0; - struct page *page; - struct inode *btree_inode = root->fs_info->btree_inode; + struct address_space *mapping = root->fs_info->btree_inode->i_mapping; u64 start = 0; u64 end; - unsigned long index; - while (1) { - ret = find_first_extent_bit(dirty_pages, start, &start, &end, - mark); - if (ret) - break; - - clear_extent_bits(dirty_pages, start, end, mark, GFP_NOFS); - while (start <= end) { - index = start >> PAGE_CACHE_SHIFT; - start = (u64)(index + 1) << PAGE_CACHE_SHIFT; - page = find_get_page(btree_inode->i_mapping, index); - if (!page) - continue; - if (PageDirty(page)) { - btree_lock_page_hook(page); - wait_on_page_writeback(page); - err = write_one_page(page, 0); - if (err) - werr = err; - } - wait_on_page_writeback(page); - page_cache_release(page); - cond_resched(); - } + while (!find_first_extent_bit(dirty_pages, start, &start, &end, + EXTENT_NEED_WAIT)) { + clear_extent_bits(dirty_pages, start, end, EXTENT_NEED_WAIT, GFP_NOFS); + err = filemap_fdatawait_range(mapping, start, end); + if (err) + werr = err; + cond_resched(); + start = end + 1; } if (err) werr = err; @@ -673,7 +634,12 @@ int btrfs_write_and_wait_marked_extents(struct btrfs_root *root, ret = btrfs_write_marked_extents(root, dirty_pages, mark); ret2 = btrfs_wait_marked_extents(root, dirty_pages, mark); - return ret || ret2; + + if (ret) + return ret; + if (ret2) + return ret2; + return 0; } int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans, @@ -911,10 +877,9 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, } btrfs_reloc_pre_snapshot(trans, pending, &to_reserve); - btrfs_orphan_pre_snapshot(trans, pending, &to_reserve); if (to_reserve > 0) { - ret = btrfs_block_rsv_add(trans, root, &pending->block_rsv, + ret = btrfs_block_rsv_add(root, &pending->block_rsv, to_reserve); if (ret) { pending->error = ret; @@ -1002,7 +967,6 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, BUG_ON(IS_ERR(pending->snap)); btrfs_reloc_post_snapshot(trans, pending); - btrfs_orphan_post_snapshot(trans, pending); fail: kfree(new_root_item); trans->block_rsv = rsv; @@ -1032,7 +996,7 @@ static void update_super_roots(struct btrfs_root *root) struct btrfs_root_item *root_item; struct btrfs_super_block *super; - super = &root->fs_info->super_copy; + super = root->fs_info->super_copy; root_item = &root->fs_info->chunk_root->root_item; super->chunk_root = root_item->bytenr; @@ -1043,7 +1007,7 @@ static void update_super_roots(struct btrfs_root *root) super->root = root_item->bytenr; super->generation = root_item->generation; super->root_level = root_item->level; - if (super->cache_generation != 0 || btrfs_test_opt(root, SPACE_CACHE)) + if (btrfs_test_opt(root, SPACE_CACHE)) super->cache_generation = root_item->generation; } @@ -1168,14 +1132,15 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, btrfs_run_ordered_operations(root, 0); + btrfs_trans_release_metadata(trans, root); + trans->block_rsv = NULL; + /* make a pass through all the delayed refs we have so far * any runnings procs may add more while we are here */ ret = btrfs_run_delayed_refs(trans, root, 0); BUG_ON(ret); - btrfs_trans_release_metadata(trans, root); - cur_trans = trans->transaction; /* * set the flushing flag so procs in this transaction have to @@ -1341,12 +1306,12 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, update_super_roots(root); if (!root->fs_info->log_root_recovering) { - btrfs_set_super_log_root(&root->fs_info->super_copy, 0); - btrfs_set_super_log_root_level(&root->fs_info->super_copy, 0); + btrfs_set_super_log_root(root->fs_info->super_copy, 0); + btrfs_set_super_log_root_level(root->fs_info->super_copy, 0); } - memcpy(&root->fs_info->super_for_commit, &root->fs_info->super_copy, - sizeof(root->fs_info->super_copy)); + memcpy(root->fs_info->super_for_commit, root->fs_info->super_copy, + sizeof(*root->fs_info->super_copy)); trans->transaction->blocked = 0; spin_lock(&root->fs_info->trans_lock); diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 786639fca067..f4d81c06d48f 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -276,8 +276,9 @@ static int process_one_buffer(struct btrfs_root *log, struct walk_control *wc, u64 gen) { if (wc->pin) - btrfs_pin_extent(log->fs_info->extent_root, - eb->start, eb->len, 0); + btrfs_pin_extent_for_log_replay(wc->trans, + log->fs_info->extent_root, + eb->start, eb->len); if (btrfs_buffer_uptodate(eb, gen)) { if (wc->write) @@ -1760,7 +1761,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_free_reserved_extent(root, + ret = btrfs_free_and_pin_reserved_extent(root, bytenr, blocksize); BUG_ON(ret); } @@ -1828,7 +1829,7 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans, btrfs_tree_unlock(next); WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_free_reserved_extent(root, + ret = btrfs_free_and_pin_reserved_extent(root, path->nodes[*level]->start, path->nodes[*level]->len); BUG_ON(ret); @@ -1897,7 +1898,7 @@ static int walk_log_tree(struct btrfs_trans_handle *trans, WARN_ON(log->root_key.objectid != BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_free_reserved_extent(log, next->start, + ret = btrfs_free_and_pin_reserved_extent(log, next->start, next->len); BUG_ON(ret); } @@ -2013,10 +2014,10 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, /* wait for previous tree log sync to complete */ if (atomic_read(&root->log_commit[(index1 + 1) % 2])) wait_log_commit(trans, root, root->log_transid - 1); - while (1) { unsigned long batch = root->log_batch; - if (root->log_multiple_pids) { + /* when we're on an ssd, just kick the log commit out */ + if (!btrfs_test_opt(root, SSD) && root->log_multiple_pids) { mutex_unlock(&root->log_mutex); schedule_timeout_uninterruptible(1); mutex_lock(&root->log_mutex); @@ -2117,9 +2118,9 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, BUG_ON(ret); btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); - btrfs_set_super_log_root(&root->fs_info->super_for_commit, + btrfs_set_super_log_root(root->fs_info->super_for_commit, log_root_tree->node->start); - btrfs_set_super_log_root_level(&root->fs_info->super_for_commit, + btrfs_set_super_log_root_level(root->fs_info->super_for_commit, btrfs_header_level(log_root_tree->node)); log_root_tree->log_batch = 0; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 1dccce5bc93d..f1685a2b45c8 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -605,10 +605,8 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices, set_blocksize(bdev, 4096); bh = btrfs_read_dev_super(bdev); - if (!bh) { - ret = -EINVAL; + if (!bh) goto error_close; - } disk_super = (struct btrfs_super_block *)bh->b_data; devid = btrfs_stack_device_id(&disk_super->dev_item); @@ -663,7 +661,7 @@ error: continue; } if (fs_devices->open_devices == 0) { - ret = -EIO; + ret = -EINVAL; goto out; } fs_devices->seeding = seeding; @@ -1021,8 +1019,13 @@ static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans, } BUG_ON(ret); - if (device->bytes_used > 0) - device->bytes_used -= btrfs_dev_extent_length(leaf, extent); + if (device->bytes_used > 0) { + u64 len = btrfs_dev_extent_length(leaf, extent); + device->bytes_used -= len; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space += len; + spin_unlock(&root->fs_info->free_chunk_lock); + } ret = btrfs_del_item(trans, root, path); out: @@ -1364,6 +1367,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) if (ret) goto error_undo; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space = device->total_bytes - + device->bytes_used; + spin_unlock(&root->fs_info->free_chunk_lock); + device->in_fs_metadata = 0; btrfs_scrub_cancel_dev(root, device); @@ -1395,8 +1403,8 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) call_rcu(&device->rcu, free_device); mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); - num_devices = btrfs_super_num_devices(&root->fs_info->super_copy) - 1; - btrfs_set_super_num_devices(&root->fs_info->super_copy, num_devices); + num_devices = btrfs_super_num_devices(root->fs_info->super_copy) - 1; + btrfs_set_super_num_devices(root->fs_info->super_copy, num_devices); if (cur_devices->open_devices == 0) { struct btrfs_fs_devices *fs_devices; @@ -1458,7 +1466,7 @@ static int btrfs_prepare_sprout(struct btrfs_trans_handle *trans, struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices; struct btrfs_fs_devices *old_devices; struct btrfs_fs_devices *seed_devices; - struct btrfs_super_block *disk_super = &root->fs_info->super_copy; + struct btrfs_super_block *disk_super = root->fs_info->super_copy; struct btrfs_device *device; u64 super_flags; @@ -1699,15 +1707,19 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) root->fs_info->fs_devices->num_can_discard++; root->fs_info->fs_devices->total_rw_bytes += device->total_bytes; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space += device->total_bytes; + spin_unlock(&root->fs_info->free_chunk_lock); + if (!blk_queue_nonrot(bdev_get_queue(bdev))) root->fs_info->fs_devices->rotating = 1; - total_bytes = btrfs_super_total_bytes(&root->fs_info->super_copy); - btrfs_set_super_total_bytes(&root->fs_info->super_copy, + total_bytes = btrfs_super_total_bytes(root->fs_info->super_copy); + btrfs_set_super_total_bytes(root->fs_info->super_copy, total_bytes + device->total_bytes); - total_bytes = btrfs_super_num_devices(&root->fs_info->super_copy); - btrfs_set_super_num_devices(&root->fs_info->super_copy, + total_bytes = btrfs_super_num_devices(root->fs_info->super_copy); + btrfs_set_super_num_devices(root->fs_info->super_copy, total_bytes + 1); mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); @@ -1798,7 +1810,7 @@ static int __btrfs_grow_device(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 new_size) { struct btrfs_super_block *super_copy = - &device->dev_root->fs_info->super_copy; + device->dev_root->fs_info->super_copy; u64 old_total = btrfs_super_total_bytes(super_copy); u64 diff = new_size - device->total_bytes; @@ -1857,7 +1869,7 @@ static int btrfs_free_chunk(struct btrfs_trans_handle *trans, static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64 chunk_offset) { - struct btrfs_super_block *super_copy = &root->fs_info->super_copy; + struct btrfs_super_block *super_copy = root->fs_info->super_copy; struct btrfs_disk_key *disk_key; struct btrfs_chunk *chunk; u8 *ptr; @@ -2183,7 +2195,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) bool retried = false; struct extent_buffer *l; struct btrfs_key key; - struct btrfs_super_block *super_copy = &root->fs_info->super_copy; + struct btrfs_super_block *super_copy = root->fs_info->super_copy; u64 old_total = btrfs_super_total_bytes(super_copy); u64 old_size = device->total_bytes; u64 diff = device->total_bytes - new_size; @@ -2200,8 +2212,12 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) lock_chunks(root); device->total_bytes = new_size; - if (device->writeable) + if (device->writeable) { device->fs_devices->total_rw_bytes -= diff; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space -= diff; + spin_unlock(&root->fs_info->free_chunk_lock); + } unlock_chunks(root); again: @@ -2265,6 +2281,9 @@ again: device->total_bytes = old_size; if (device->writeable) device->fs_devices->total_rw_bytes += diff; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space += diff; + spin_unlock(&root->fs_info->free_chunk_lock); unlock_chunks(root); goto done; } @@ -2300,7 +2319,7 @@ static int btrfs_add_system_chunk(struct btrfs_trans_handle *trans, struct btrfs_key *key, struct btrfs_chunk *chunk, int item_size) { - struct btrfs_super_block *super_copy = &root->fs_info->super_copy; + struct btrfs_super_block *super_copy = root->fs_info->super_copy; struct btrfs_disk_key disk_key; u32 array_size; u8 *ptr; @@ -2623,6 +2642,11 @@ static int __finish_chunk_alloc(struct btrfs_trans_handle *trans, index++; } + spin_lock(&extent_root->fs_info->free_chunk_lock); + extent_root->fs_info->free_chunk_space -= (stripe_size * + map->num_stripes); + spin_unlock(&extent_root->fs_info->free_chunk_lock); + index = 0; stripe = &chunk->stripe; while (index < map->num_stripes) { @@ -3624,15 +3648,20 @@ static int read_one_dev(struct btrfs_root *root, fill_device_from_item(leaf, dev_item, device); device->dev_root = root->fs_info->dev_root; device->in_fs_metadata = 1; - if (device->writeable) + if (device->writeable) { device->fs_devices->total_rw_bytes += device->total_bytes; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space += device->total_bytes - + device->bytes_used; + spin_unlock(&root->fs_info->free_chunk_lock); + } ret = 0; return ret; } int btrfs_read_sys_array(struct btrfs_root *root) { - struct btrfs_super_block *super_copy = &root->fs_info->super_copy; + struct btrfs_super_block *super_copy = root->fs_info->super_copy; struct extent_buffer *sb; struct btrfs_disk_key *disk_key; struct btrfs_chunk *chunk; diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index 69565e5fc6a0..a76e41c04b71 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c @@ -127,6 +127,17 @@ static int do_setxattr(struct btrfs_trans_handle *trans, again: ret = btrfs_insert_xattr_item(trans, root, path, btrfs_ino(inode), name, name_len, value, size); + /* + * If we're setting an xattr to a new value but the new value is say + * exactly BTRFS_MAX_XATTR_SIZE, we could end up with EOVERFLOW getting + * back from split_leaf. This is because it thinks we'll be extending + * the existing item size, but we're asking for enough space to add the + * item itself. So if we get EOVERFLOW just set ret to EEXIST and let + * the rest of the function figure it out. + */ + if (ret == -EOVERFLOW) + ret = -EEXIST; + if (ret == -EEXIST) { if (flags & XATTR_CREATE) goto out; diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index e76bfeb68267..30acd22147e1 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -351,9 +351,7 @@ static int build_avpair_blob(struct cifs_ses *ses, const struct nls_table *nls_cp) { unsigned int dlen; - unsigned int wlen; - unsigned int size = 6 * sizeof(struct ntlmssp2_name); - __le64 curtime; + unsigned int size = 2 * sizeof(struct ntlmssp2_name); char *defdmname = "WORKGROUP"; unsigned char *blobptr; struct ntlmssp2_name *attrptr; @@ -365,15 +363,14 @@ build_avpair_blob(struct cifs_ses *ses, const struct nls_table *nls_cp) } dlen = strlen(ses->domainName); - wlen = strlen(ses->server->hostname); - /* The length of this blob is a size which is - * six times the size of a structure which holds name/size + - * two times the unicode length of a domain name + - * two times the unicode length of a server name + - * size of a timestamp (which is 8 bytes). + /* + * The length of this blob is two times the size of a + * structure (av pair) which holds name/size + * ( for NTLMSSP_AV_NB_DOMAIN_NAME followed by NTLMSSP_AV_EOL ) + + * unicode length of a netbios domain name */ - ses->auth_key.len = size + 2 * (2 * dlen) + 2 * (2 * wlen) + 8; + ses->auth_key.len = size + 2 * dlen; ses->auth_key.response = kzalloc(ses->auth_key.len, GFP_KERNEL); if (!ses->auth_key.response) { ses->auth_key.len = 0; @@ -384,44 +381,15 @@ build_avpair_blob(struct cifs_ses *ses, const struct nls_table *nls_cp) blobptr = ses->auth_key.response; attrptr = (struct ntlmssp2_name *) blobptr; + /* + * As defined in MS-NTLM 3.3.2, just this av pair field + * is sufficient as part of the temp + */ attrptr->type = cpu_to_le16(NTLMSSP_AV_NB_DOMAIN_NAME); attrptr->length = cpu_to_le16(2 * dlen); blobptr = (unsigned char *)attrptr + sizeof(struct ntlmssp2_name); cifs_strtoUCS((__le16 *)blobptr, ses->domainName, dlen, nls_cp); - blobptr += 2 * dlen; - attrptr = (struct ntlmssp2_name *) blobptr; - - attrptr->type = cpu_to_le16(NTLMSSP_AV_NB_COMPUTER_NAME); - attrptr->length = cpu_to_le16(2 * wlen); - blobptr = (unsigned char *)attrptr + sizeof(struct ntlmssp2_name); - cifs_strtoUCS((__le16 *)blobptr, ses->server->hostname, wlen, nls_cp); - - blobptr += 2 * wlen; - attrptr = (struct ntlmssp2_name *) blobptr; - - attrptr->type = cpu_to_le16(NTLMSSP_AV_DNS_DOMAIN_NAME); - attrptr->length = cpu_to_le16(2 * dlen); - blobptr = (unsigned char *)attrptr + sizeof(struct ntlmssp2_name); - cifs_strtoUCS((__le16 *)blobptr, ses->domainName, dlen, nls_cp); - - blobptr += 2 * dlen; - attrptr = (struct ntlmssp2_name *) blobptr; - - attrptr->type = cpu_to_le16(NTLMSSP_AV_DNS_COMPUTER_NAME); - attrptr->length = cpu_to_le16(2 * wlen); - blobptr = (unsigned char *)attrptr + sizeof(struct ntlmssp2_name); - cifs_strtoUCS((__le16 *)blobptr, ses->server->hostname, wlen, nls_cp); - - blobptr += 2 * wlen; - attrptr = (struct ntlmssp2_name *) blobptr; - - attrptr->type = cpu_to_le16(NTLMSSP_AV_TIMESTAMP); - attrptr->length = cpu_to_le16(sizeof(__le64)); - blobptr = (unsigned char *)attrptr + sizeof(struct ntlmssp2_name); - curtime = cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - memcpy(blobptr, &curtime, sizeof(__le64)); - return 0; } diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index f93eb948d071..54b8f1e7da94 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -548,6 +548,12 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb) struct inode *dir = dentry->d_inode; struct dentry *child; + if (!dir) { + dput(dentry); + dentry = ERR_PTR(-ENOENT); + break; + } + /* skip separators */ while (*s == sep) s++; @@ -563,10 +569,6 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb) mutex_unlock(&dir->i_mutex); dput(dentry); dentry = child; - if (!dentry->d_inode) { - dput(dentry); - dentry = ERR_PTR(-ENOENT); - } } while (!IS_ERR(dentry)); _FreeXid(xid); kfree(full_path); diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index aac37d99a487..a80f7bd97b90 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -4079,7 +4079,8 @@ int CIFSFindNext(const int xid, struct cifs_tcon *tcon, T2_FNEXT_RSP_PARMS *parms; char *response_data; int rc = 0; - int bytes_returned, name_len; + int bytes_returned; + unsigned int name_len; __u16 params, byte_count; cFYI(1, "In FindNext"); diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 633c246b6775..71beb0201970 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1298,7 +1298,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, /* ignore */ } else if (strnicmp(data, "guest", 5) == 0) { /* ignore */ - } else if (strnicmp(data, "rw", 2) == 0) { + } else if (strnicmp(data, "rw", 2) == 0 && strlen(data) == 2) { /* ignore */ } else if (strnicmp(data, "ro", 2) == 0) { /* ignore */ @@ -1401,7 +1401,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, vol->server_ino = 1; } else if (strnicmp(data, "noserverino", 9) == 0) { vol->server_ino = 0; - } else if (strnicmp(data, "rwpidforward", 4) == 0) { + } else if (strnicmp(data, "rwpidforward", 12) == 0) { vol->rwpidforward = 1; } else if (strnicmp(data, "cifsacl", 7) == 0) { vol->cifs_acl = 1; @@ -2018,7 +2018,7 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) warned_on_ntlm = true; cERROR(1, "default security mechanism requested. The default " "security mechanism will be upgraded from ntlm to " - "ntlmv2 in kernel release 3.1"); + "ntlmv2 in kernel release 3.2"); } ses->overrideSecFlg = volume_info->secFlg; diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 04da6acde85d..12661e1deedd 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1134,7 +1134,7 @@ struct buffer_head *ext3_bread(handle_t *handle, struct inode *inode, return bh; if (buffer_uptodate(bh)) return bh; - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META | REQ_PRIO, 1, &bh); wait_on_buffer(bh); if (buffer_uptodate(bh)) return bh; @@ -2807,7 +2807,7 @@ make_io: trace_ext3_load_inode(inode); get_bh(bh); bh->b_end_io = end_buffer_read_sync; - submit_bh(READ_META, bh); + submit_bh(READ | REQ_META | REQ_PRIO, bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) { ext3_error(inode->i_sb, "ext3_get_inode_loc", diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 5571708b6a58..0629e09f6511 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -922,7 +922,8 @@ restart: bh = ext3_getblk(NULL, dir, b++, 0, &err); bh_use[ra_max] = bh; if (bh) - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META | REQ_PRIO, + 1, &bh); } } if ((bh = bh_use[ra_ptr++]) == NULL) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 18d2558b7624..986e2388f031 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -647,7 +647,7 @@ struct buffer_head *ext4_bread(handle_t *handle, struct inode *inode, return bh; if (buffer_uptodate(bh)) return bh; - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META | REQ_PRIO, 1, &bh); wait_on_buffer(bh); if (buffer_uptodate(bh)) return bh; @@ -3298,7 +3298,7 @@ make_io: trace_ext4_load_inode(inode); get_bh(bh); bh->b_end_io = end_buffer_read_sync; - submit_bh(READ_META, bh); + submit_bh(READ | REQ_META | REQ_PRIO, bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) { EXT4_ERROR_INODE_BLOCK(inode, block, diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index f8068c7bae9f..1c924faeb6c8 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -922,7 +922,8 @@ restart: bh = ext4_getblk(NULL, dir, b++, 0, &err); bh_use[ra_max] = bh; if (bh) - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META | REQ_PRIO, + 1, &bh); } } if ((bh = bh_use[ra_ptr++]) == NULL) diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index 85c62923ee29..598646434362 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -624,9 +624,9 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull) bh->b_end_io = end_buffer_write_sync; get_bh(bh); if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags)) - submit_bh(WRITE_SYNC | REQ_META, bh); + submit_bh(WRITE_SYNC | REQ_META | REQ_PRIO, bh); else - submit_bh(WRITE_FLUSH_FUA | REQ_META, bh); + submit_bh(WRITE_FLUSH_FUA | REQ_META | REQ_PRIO, bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c index 747238cd9f96..be29858900f6 100644 --- a/fs/gfs2/meta_io.c +++ b/fs/gfs2/meta_io.c @@ -37,7 +37,7 @@ static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wb { struct buffer_head *bh, *head; int nr_underway = 0; - int write_op = REQ_META | + int write_op = REQ_META | REQ_PRIO | (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); BUG_ON(!PageLocked(page)); @@ -225,7 +225,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags, } bh->b_end_io = end_buffer_read_sync; get_bh(bh); - submit_bh(READ_SYNC | REQ_META, bh); + submit_bh(READ_SYNC | REQ_META | REQ_PRIO, bh); if (!(flags & DIO_WAIT)) return 0; @@ -435,7 +435,7 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen) if (buffer_uptodate(first_bh)) goto out; if (!buffer_locked(first_bh)) - ll_rw_block(READ_SYNC | REQ_META, 1, &first_bh); + ll_rw_block(READ_SYNC | REQ_META | REQ_PRIO, 1, &first_bh); dblock++; extlen--; diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 3bc073a4cf82..079587e53849 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -224,7 +224,7 @@ static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector, int silent) bio->bi_end_io = end_bio_io_page; bio->bi_private = page; - submit_bio(READ_SYNC | REQ_META, bio); + submit_bio(READ_SYNC | REQ_META | REQ_PRIO, bio); wait_on_page_locked(page); bio_put(bio); if (!PageUptodate(page)) { diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 42e8d23bc047..0e8bb13381e4 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -709,7 +709,7 @@ get_a_page: set_buffer_uptodate(bh); if (!buffer_uptodate(bh)) { - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META | REQ_PRIO, 1, &bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) goto unlock_out; diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index c106ca22e812..d24a9b666a23 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -344,6 +344,7 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) struct inode *root, *inode; struct qstr str; struct nls_table *nls = NULL; + u64 last_fs_block, last_fs_page; int err; err = -EINVAL; @@ -399,9 +400,13 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) if (!sbi->rsrc_clump_blocks) sbi->rsrc_clump_blocks = 1; - err = generic_check_addressable(sbi->alloc_blksz_shift, - sbi->total_blocks); - if (err) { + err = -EFBIG; + last_fs_block = sbi->total_blocks - 1; + last_fs_page = (last_fs_block << sbi->alloc_blksz_shift) >> + PAGE_CACHE_SHIFT; + + if ((last_fs_block > (sector_t)(~0ULL) >> (sbi->alloc_blksz_shift - 9)) || + (last_fs_page > (pgoff_t)(~0ULL))) { printk(KERN_ERR "hfs: filesystem size too large.\n"); goto out_free_vhdr; } @@ -525,8 +530,8 @@ out_close_cat_tree: out_close_ext_tree: hfs_btree_close(sbi->ext_tree); out_free_vhdr: - kfree(sbi->s_vhdr); - kfree(sbi->s_backup_vhdr); + kfree(sbi->s_vhdr_buf); + kfree(sbi->s_backup_vhdr_buf); out_unload_nls: unload_nls(sbi->nls); unload_nls(nls); diff --git a/fs/hfsplus/wrapper.c b/fs/hfsplus/wrapper.c index 10e515a0d452..7daf4b852d1c 100644 --- a/fs/hfsplus/wrapper.c +++ b/fs/hfsplus/wrapper.c @@ -272,9 +272,9 @@ reread: return 0; out_free_backup_vhdr: - kfree(sbi->s_backup_vhdr); + kfree(sbi->s_backup_vhdr_buf); out_free_vhdr: - kfree(sbi->s_vhdr); + kfree(sbi->s_vhdr_buf); out: return error; } diff --git a/fs/namei.c b/fs/namei.c index b52bc685465f..0b3138de2a3b 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -721,12 +721,6 @@ static int follow_automount(struct path *path, unsigned flags, if (!path->dentry->d_op || !path->dentry->d_op->d_automount) return -EREMOTE; - /* We don't want to mount if someone supplied AT_NO_AUTOMOUNT - * and this is the terminal part of the path. - */ - if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_PARENT)) - return -EISDIR; /* we actually want to stop here */ - /* We don't want to mount if someone's just doing a stat - * unless they're stat'ing a directory and appended a '/' to * the name. @@ -739,7 +733,7 @@ static int follow_automount(struct path *path, unsigned flags, * of the daemon to instantiate them before they can be used. */ if (!(flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY | - LOOKUP_OPEN | LOOKUP_CREATE)) && + LOOKUP_OPEN | LOOKUP_CREATE | LOOKUP_AUTOMOUNT)) && path->dentry->d_inode) return -EISDIR; @@ -2616,6 +2610,7 @@ int vfs_rmdir(struct inode *dir, struct dentry *dentry) if (!dir->i_op->rmdir) return -EPERM; + dget(dentry); mutex_lock(&dentry->d_inode->i_mutex); error = -EBUSY; @@ -2636,6 +2631,7 @@ int vfs_rmdir(struct inode *dir, struct dentry *dentry) out: mutex_unlock(&dentry->d_inode->i_mutex); + dput(dentry); if (!error) d_delete(dentry); return error; @@ -3025,6 +3021,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, if (error) return error; + dget(new_dentry); if (target) mutex_lock(&target->i_mutex); @@ -3045,6 +3042,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, out: if (target) mutex_unlock(&target->i_mutex); + dput(new_dentry); if (!error) if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) d_move(old_dentry,new_dentry); diff --git a/fs/namespace.c b/fs/namespace.c index 22bfe8273c68..b4febb29d3bb 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1757,7 +1757,7 @@ static int do_loopback(struct path *path, char *old_name, return err; if (!old_name || !*old_name) return -EINVAL; - err = kern_path(old_name, LOOKUP_FOLLOW, &old_path); + err = kern_path(old_name, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &old_path); if (err) return err; diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 1ec1a85fa71c..3e93e9a1bee1 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -56,6 +56,9 @@ enum nfs4_session_state { NFS4_SESSION_DRAINING, }; +#define NFS4_RENEW_TIMEOUT 0x01 +#define NFS4_RENEW_DELEGATION_CB 0x02 + struct nfs4_minor_version_ops { u32 minor_version; @@ -225,7 +228,7 @@ struct nfs4_state_recovery_ops { }; struct nfs4_state_maintenance_ops { - int (*sched_state_renewal)(struct nfs_client *, struct rpc_cred *); + int (*sched_state_renewal)(struct nfs_client *, struct rpc_cred *, unsigned); struct rpc_cred * (*get_state_renewal_cred_locked)(struct nfs_client *); int (*renew_lease)(struct nfs_client *, struct rpc_cred *); }; @@ -237,8 +240,6 @@ extern const struct inode_operations nfs4_dir_inode_operations; extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *, struct nfs4_setclientid_res *); extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct nfs4_setclientid_res *arg, struct rpc_cred *); extern int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred); -extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *); -extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *); extern int nfs4_init_clientid(struct nfs_client *, struct rpc_cred *); extern int nfs41_init_clientid(struct nfs_client *, struct rpc_cred *); extern int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc); @@ -349,6 +350,7 @@ extern void nfs4_close_sync(struct nfs4_state *, fmode_t); extern void nfs4_state_set_mode_locked(struct nfs4_state *, fmode_t); extern void nfs4_schedule_lease_recovery(struct nfs_client *); extern void nfs4_schedule_state_manager(struct nfs_client *); +extern void nfs4_schedule_path_down_recovery(struct nfs_client *clp); extern void nfs4_schedule_stateid_recovery(const struct nfs_server *, struct nfs4_state *); extern void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags); extern void nfs41_handle_recall_slot(struct nfs_client *clp); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 8c77039e7a81..4700fae1ada0 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3374,9 +3374,13 @@ static void nfs4_renew_done(struct rpc_task *task, void *calldata) if (task->tk_status < 0) { /* Unless we're shutting down, schedule state recovery! */ - if (test_bit(NFS_CS_RENEWD, &clp->cl_res_state) != 0) + if (test_bit(NFS_CS_RENEWD, &clp->cl_res_state) == 0) + return; + if (task->tk_status != NFS4ERR_CB_PATH_DOWN) { nfs4_schedule_lease_recovery(clp); - return; + return; + } + nfs4_schedule_path_down_recovery(clp); } do_renew_lease(clp, timestamp); } @@ -3386,7 +3390,7 @@ static const struct rpc_call_ops nfs4_renew_ops = { .rpc_release = nfs4_renew_release, }; -int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred) +static int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred, unsigned renew_flags) { struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENEW], @@ -3395,9 +3399,11 @@ int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred) }; struct nfs4_renewdata *data; + if (renew_flags == 0) + return 0; if (!atomic_inc_not_zero(&clp->cl_count)) return -EIO; - data = kmalloc(sizeof(*data), GFP_KERNEL); + data = kmalloc(sizeof(*data), GFP_NOFS); if (data == NULL) return -ENOMEM; data->client = clp; @@ -3406,7 +3412,7 @@ int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred) &nfs4_renew_ops, data); } -int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred) +static int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred) { struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENEW], @@ -5504,11 +5510,13 @@ static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, struct rpc_ return rpc_run_task(&task_setup_data); } -static int nfs41_proc_async_sequence(struct nfs_client *clp, struct rpc_cred *cred) +static int nfs41_proc_async_sequence(struct nfs_client *clp, struct rpc_cred *cred, unsigned renew_flags) { struct rpc_task *task; int ret = 0; + if ((renew_flags & NFS4_RENEW_TIMEOUT) == 0) + return 0; task = _nfs41_proc_sequence(clp, cred); if (IS_ERR(task)) ret = PTR_ERR(task); diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index df8e7f3ca56d..dc484c0eae7f 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -60,6 +60,7 @@ nfs4_renew_state(struct work_struct *work) struct rpc_cred *cred; long lease; unsigned long last, now; + unsigned renew_flags = 0; ops = clp->cl_mvops->state_renewal_ops; dprintk("%s: start\n", __func__); @@ -72,18 +73,23 @@ nfs4_renew_state(struct work_struct *work) last = clp->cl_last_renewal; now = jiffies; /* Are we close to a lease timeout? */ - if (time_after(now, last + lease/3)) { + if (time_after(now, last + lease/3)) + renew_flags |= NFS4_RENEW_TIMEOUT; + if (nfs_delegations_present(clp)) + renew_flags |= NFS4_RENEW_DELEGATION_CB; + + if (renew_flags != 0) { cred = ops->get_state_renewal_cred_locked(clp); spin_unlock(&clp->cl_lock); if (cred == NULL) { - if (!nfs_delegations_present(clp)) { + if (!(renew_flags & NFS4_RENEW_DELEGATION_CB)) { set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); goto out; } nfs_expire_all_delegations(clp); } else { /* Queue an asynchronous RENEW. */ - ops->sched_state_renewal(clp, cred); + ops->sched_state_renewal(clp, cred, renew_flags); put_rpccred(cred); goto out_exp; } diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 72ab97ef3d61..39914be40b03 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1038,6 +1038,12 @@ void nfs4_schedule_lease_recovery(struct nfs_client *clp) nfs4_schedule_state_manager(clp); } +void nfs4_schedule_path_down_recovery(struct nfs_client *clp) +{ + nfs_handle_cb_pathdown(clp); + nfs4_schedule_state_manager(clp); +} + static int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state) { diff --git a/fs/nfs/super.c b/fs/nfs/super.c index b961ceac66b4..5b19b6aabe18 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2035,9 +2035,6 @@ static inline void nfs_initialise_sb(struct super_block *sb) sb->s_blocksize = nfs_block_bits(server->wsize, &sb->s_blocksize_bits); - if (server->flags & NFS_MOUNT_NOAC) - sb->s_flags |= MS_SYNCHRONOUS; - sb->s_bdi = &server->backing_dev_info; nfs_super_set_maxbytes(sb, server->maxfilesize); @@ -2249,6 +2246,10 @@ static struct dentry *nfs_fs_mount(struct file_system_type *fs_type, if (server->flags & NFS_MOUNT_UNSHARED) compare_super = NULL; + /* -o noac implies -o sync */ + if (server->flags & NFS_MOUNT_NOAC) + sb_mntdata.mntflags |= MS_SYNCHRONOUS; + /* Get a superblock - note that we may end up sharing one that already exists */ s = sget(fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { @@ -2361,6 +2362,10 @@ nfs_xdev_mount(struct file_system_type *fs_type, int flags, if (server->flags & NFS_MOUNT_UNSHARED) compare_super = NULL; + /* -o noac implies -o sync */ + if (server->flags & NFS_MOUNT_NOAC) + sb_mntdata.mntflags |= MS_SYNCHRONOUS; + /* Get a superblock - note that we may end up sharing one that already exists */ s = sget(&nfs_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { @@ -2628,6 +2633,10 @@ nfs4_remote_mount(struct file_system_type *fs_type, int flags, if (server->flags & NFS4_MOUNT_UNSHARED) compare_super = NULL; + /* -o noac implies -o sync */ + if (server->flags & NFS_MOUNT_NOAC) + sb_mntdata.mntflags |= MS_SYNCHRONOUS; + /* Get a superblock - note that we may end up sharing one that already exists */ s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { @@ -2789,7 +2798,7 @@ static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt, goto out_put_mnt_ns; ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt, - export_path, LOOKUP_FOLLOW, &path); + export_path, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &path); nfs_referral_loop_unprotect(); put_mnt_ns(ns_private); @@ -2916,6 +2925,10 @@ nfs4_xdev_mount(struct file_system_type *fs_type, int flags, if (server->flags & NFS4_MOUNT_UNSHARED) compare_super = NULL; + /* -o noac implies -o sync */ + if (server->flags & NFS_MOUNT_NOAC) + sb_mntdata.mntflags |= MS_SYNCHRONOUS; + /* Get a superblock - note that we may end up sharing one that already exists */ s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { @@ -3003,6 +3016,10 @@ nfs4_remote_referral_mount(struct file_system_type *fs_type, int flags, if (server->flags & NFS4_MOUNT_UNSHARED) compare_super = NULL; + /* -o noac implies -o sync */ + if (server->flags & NFS_MOUNT_NOAC) + sb_mntdata.mntflags |= MS_SYNCHRONOUS; + /* Get a superblock - note that we may end up sharing one that already exists */ s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { diff --git a/fs/nfs/write.c b/fs/nfs/write.c index b39b37f80913..c9bd2a6b7d4b 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -958,7 +958,7 @@ static int nfs_flush_multi(struct nfs_pageio_descriptor *desc, struct list_head if (!data) goto out_bad; data->pagevec[0] = page; - nfs_write_rpcsetup(req, data, wsize, offset, desc->pg_ioflags); + nfs_write_rpcsetup(req, data, len, offset, desc->pg_ioflags); list_add(&data->list, res); requests++; nbytes -= len; diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 25b6a887adb9..5afaa58a8630 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -877,30 +877,54 @@ struct numa_maps_private { struct numa_maps md; }; -static void gather_stats(struct page *page, struct numa_maps *md, int pte_dirty) +static void gather_stats(struct page *page, struct numa_maps *md, int pte_dirty, + unsigned long nr_pages) { int count = page_mapcount(page); - md->pages++; + md->pages += nr_pages; if (pte_dirty || PageDirty(page)) - md->dirty++; + md->dirty += nr_pages; if (PageSwapCache(page)) - md->swapcache++; + md->swapcache += nr_pages; if (PageActive(page) || PageUnevictable(page)) - md->active++; + md->active += nr_pages; if (PageWriteback(page)) - md->writeback++; + md->writeback += nr_pages; if (PageAnon(page)) - md->anon++; + md->anon += nr_pages; if (count > md->mapcount_max) md->mapcount_max = count; - md->node[page_to_nid(page)]++; + md->node[page_to_nid(page)] += nr_pages; +} + +static struct page *can_gather_numa_stats(pte_t pte, struct vm_area_struct *vma, + unsigned long addr) +{ + struct page *page; + int nid; + + if (!pte_present(pte)) + return NULL; + + page = vm_normal_page(vma, addr, pte); + if (!page) + return NULL; + + if (PageReserved(page)) + return NULL; + + nid = page_to_nid(page); + if (!node_isset(nid, node_states[N_HIGH_MEMORY])) + return NULL; + + return page; } static int gather_pte_stats(pmd_t *pmd, unsigned long addr, @@ -912,26 +936,32 @@ static int gather_pte_stats(pmd_t *pmd, unsigned long addr, pte_t *pte; md = walk->private; - orig_pte = pte = pte_offset_map_lock(walk->mm, pmd, addr, &ptl); - do { - struct page *page; - int nid; + spin_lock(&walk->mm->page_table_lock); + if (pmd_trans_huge(*pmd)) { + if (pmd_trans_splitting(*pmd)) { + spin_unlock(&walk->mm->page_table_lock); + wait_split_huge_page(md->vma->anon_vma, pmd); + } else { + pte_t huge_pte = *(pte_t *)pmd; + struct page *page; - if (!pte_present(*pte)) - continue; + page = can_gather_numa_stats(huge_pte, md->vma, addr); + if (page) + gather_stats(page, md, pte_dirty(huge_pte), + HPAGE_PMD_SIZE/PAGE_SIZE); + spin_unlock(&walk->mm->page_table_lock); + return 0; + } + } else { + spin_unlock(&walk->mm->page_table_lock); + } - page = vm_normal_page(md->vma, addr, *pte); + orig_pte = pte = pte_offset_map_lock(walk->mm, pmd, addr, &ptl); + do { + struct page *page = can_gather_numa_stats(*pte, md->vma, addr); if (!page) continue; - - if (PageReserved(page)) - continue; - - nid = page_to_nid(page); - if (!node_isset(nid, node_states[N_HIGH_MEMORY])) - continue; - - gather_stats(page, md, pte_dirty(*pte)); + gather_stats(page, md, pte_dirty(*pte), 1); } while (pte++, addr += PAGE_SIZE, addr != end); pte_unmap_unlock(orig_pte, ptl); @@ -952,7 +982,7 @@ static int gather_hugetbl_stats(pte_t *pte, unsigned long hmask, return 0; md = walk->private; - gather_stats(page, md, pte_dirty(*pte)); + gather_stats(page, md, pte_dirty(*pte), 1); return 0; } diff --git a/fs/quota/quota.c b/fs/quota/quota.c index b34bdb25490c..10b6be3ca280 100644 --- a/fs/quota/quota.c +++ b/fs/quota/quota.c @@ -355,7 +355,7 @@ SYSCALL_DEFINE4(quotactl, unsigned int, cmd, const char __user *, special, * resolution (think about autofs) and thus deadlocks could arise. */ if (cmds == Q_QUOTAON) { - ret = user_path_at(AT_FDCWD, addr, LOOKUP_FOLLOW, &path); + ret = user_path_at(AT_FDCWD, addr, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &path); if (ret) pathp = ERR_PTR(ret); else diff --git a/fs/stat.c b/fs/stat.c index ba5316ffac61..78a3aa83c7ea 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -81,8 +81,6 @@ int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat, if (!(flag & AT_SYMLINK_NOFOLLOW)) lookup_flags |= LOOKUP_FOLLOW; - if (flag & AT_NO_AUTOMOUNT) - lookup_flags |= LOOKUP_NO_AUTOMOUNT; if (flag & AT_EMPTY_PATH) lookup_flags |= LOOKUP_EMPTY; diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 63e971e2b837..8c37dde4c521 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -1300,6 +1300,7 @@ xfs_end_io_direct_write( bool is_async) { struct xfs_ioend *ioend = iocb->private; + struct inode *inode = ioend->io_inode; /* * blockdev_direct_IO can return an error even after the I/O @@ -1331,7 +1332,7 @@ xfs_end_io_direct_write( } /* XXX: probably should move into the real I/O completion handler */ - inode_dio_done(ioend->io_inode); + inode_dio_done(inode); } STATIC ssize_t diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index cac2ecfa6746..ef43fce519a1 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -629,7 +629,7 @@ xfs_buf_item_push( * the xfsbufd to get this buffer written. We have to unlock the buffer * to allow the xfsbufd to write it, too. */ -STATIC void +STATIC bool xfs_buf_item_pushbuf( struct xfs_log_item *lip) { @@ -643,6 +643,7 @@ xfs_buf_item_pushbuf( xfs_buf_delwri_promote(bp); xfs_buf_relse(bp); + return true; } STATIC void diff --git a/fs/xfs/xfs_dquot_item.c b/fs/xfs/xfs_dquot_item.c index 9e0e2fa3f2c8..bb3f71d236d2 100644 --- a/fs/xfs/xfs_dquot_item.c +++ b/fs/xfs/xfs_dquot_item.c @@ -183,13 +183,14 @@ xfs_qm_dqunpin_wait( * search the buffer cache can be a time consuming thing, and AIL lock is a * spinlock. */ -STATIC void +STATIC bool xfs_qm_dquot_logitem_pushbuf( struct xfs_log_item *lip) { struct xfs_dq_logitem *qlip = DQUOT_ITEM(lip); struct xfs_dquot *dqp = qlip->qli_dquot; struct xfs_buf *bp; + bool ret = true; ASSERT(XFS_DQ_IS_LOCKED(dqp)); @@ -201,17 +202,20 @@ xfs_qm_dquot_logitem_pushbuf( if (completion_done(&dqp->q_flush) || !(lip->li_flags & XFS_LI_IN_AIL)) { xfs_dqunlock(dqp); - return; + return true; } bp = xfs_incore(dqp->q_mount->m_ddev_targp, qlip->qli_format.qlf_blkno, dqp->q_mount->m_quotainfo->qi_dqchunklen, XBF_TRYLOCK); xfs_dqunlock(dqp); if (!bp) - return; + return true; if (XFS_BUF_ISDELAYWRITE(bp)) xfs_buf_delwri_promote(bp); + if (xfs_buf_ispinned(bp)) + ret = false; xfs_buf_relse(bp); + return ret; } /* diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index 588406dc6a35..836ad80d4f2b 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -708,13 +708,14 @@ xfs_inode_item_committed( * marked delayed write. If that's the case, we'll promote it and that will * allow the caller to write the buffer by triggering the xfsbufd to run. */ -STATIC void +STATIC bool xfs_inode_item_pushbuf( struct xfs_log_item *lip) { struct xfs_inode_log_item *iip = INODE_ITEM(lip); struct xfs_inode *ip = iip->ili_inode; struct xfs_buf *bp; + bool ret = true; ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED)); @@ -725,7 +726,7 @@ xfs_inode_item_pushbuf( if (completion_done(&ip->i_flush) || !(lip->li_flags & XFS_LI_IN_AIL)) { xfs_iunlock(ip, XFS_ILOCK_SHARED); - return; + return true; } bp = xfs_incore(ip->i_mount->m_ddev_targp, iip->ili_format.ilf_blkno, @@ -733,10 +734,13 @@ xfs_inode_item_pushbuf( xfs_iunlock(ip, XFS_ILOCK_SHARED); if (!bp) - return; + return true; if (XFS_BUF_ISDELAYWRITE(bp)) xfs_buf_delwri_promote(bp); + if (xfs_buf_ispinned(bp)) + ret = false; xfs_buf_relse(bp); + return ret; } /* diff --git a/fs/xfs/xfs_linux.h b/fs/xfs/xfs_linux.h index 1e8a45e74c3e..828662f70d64 100644 --- a/fs/xfs/xfs_linux.h +++ b/fs/xfs/xfs_linux.h @@ -68,6 +68,8 @@ #include <linux/ctype.h> #include <linux/writeback.h> #include <linux/capability.h> +#include <linux/kthread.h> +#include <linux/freezer.h> #include <linux/list_sort.h> #include <asm/page.h> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 2366c54cc4fa..5cf06b85fd9d 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -1652,24 +1652,13 @@ xfs_init_workqueues(void) */ xfs_syncd_wq = alloc_workqueue("xfssyncd", WQ_CPU_INTENSIVE, 8); if (!xfs_syncd_wq) - goto out; - - xfs_ail_wq = alloc_workqueue("xfsail", WQ_CPU_INTENSIVE, 8); - if (!xfs_ail_wq) - goto out_destroy_syncd; - + return -ENOMEM; return 0; - -out_destroy_syncd: - destroy_workqueue(xfs_syncd_wq); -out: - return -ENOMEM; } STATIC void xfs_destroy_workqueues(void) { - destroy_workqueue(xfs_ail_wq); destroy_workqueue(xfs_syncd_wq); } diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 06a9759b6352..53597f4db9b5 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -350,7 +350,7 @@ typedef struct xfs_item_ops { void (*iop_unlock)(xfs_log_item_t *); xfs_lsn_t (*iop_committed)(xfs_log_item_t *, xfs_lsn_t); void (*iop_push)(xfs_log_item_t *); - void (*iop_pushbuf)(xfs_log_item_t *); + bool (*iop_pushbuf)(xfs_log_item_t *); void (*iop_committing)(xfs_log_item_t *, xfs_lsn_t); } xfs_item_ops_t; diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index c15aa29fa169..3a1e7ca54c2d 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -28,8 +28,6 @@ #include "xfs_trans_priv.h" #include "xfs_error.h" -struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ - #ifdef DEBUG /* * Check that the list is sorted as it should be. @@ -356,16 +354,10 @@ xfs_ail_delete( xfs_trans_ail_cursor_clear(ailp, lip); } -/* - * xfs_ail_worker does the work of pushing on the AIL. It will requeue itself - * to run at a later time if there is more work to do to complete the push. - */ -STATIC void -xfs_ail_worker( - struct work_struct *work) +static long +xfsaild_push( + struct xfs_ail *ailp) { - struct xfs_ail *ailp = container_of(to_delayed_work(work), - struct xfs_ail, xa_work); xfs_mount_t *mp = ailp->xa_mount; struct xfs_ail_cursor cur; xfs_log_item_t *lip; @@ -427,8 +419,13 @@ xfs_ail_worker( case XFS_ITEM_PUSHBUF: XFS_STATS_INC(xs_push_ail_pushbuf); - IOP_PUSHBUF(lip); - ailp->xa_last_pushed_lsn = lsn; + + if (!IOP_PUSHBUF(lip)) { + stuck++; + flush_log = 1; + } else { + ailp->xa_last_pushed_lsn = lsn; + } push_xfsbufd = 1; break; @@ -440,7 +437,6 @@ xfs_ail_worker( case XFS_ITEM_LOCKED: XFS_STATS_INC(xs_push_ail_locked); - ailp->xa_last_pushed_lsn = lsn; stuck++; break; @@ -501,20 +497,6 @@ out_done: /* We're past our target or empty, so idle */ ailp->xa_last_pushed_lsn = 0; - /* - * We clear the XFS_AIL_PUSHING_BIT first before checking - * whether the target has changed. If the target has changed, - * this pushes the requeue race directly onto the result of the - * atomic test/set bit, so we are guaranteed that either the - * the pusher that changed the target or ourselves will requeue - * the work (but not both). - */ - clear_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags); - smp_rmb(); - if (XFS_LSN_CMP(ailp->xa_target, target) == 0 || - test_and_set_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags)) - return; - tout = 50; } else if (XFS_LSN_CMP(lsn, target) >= 0) { /* @@ -537,9 +519,30 @@ out_done: tout = 20; } - /* There is more to do, requeue us. */ - queue_delayed_work(xfs_syncd_wq, &ailp->xa_work, - msecs_to_jiffies(tout)); + return tout; +} + +static int +xfsaild( + void *data) +{ + struct xfs_ail *ailp = data; + long tout = 0; /* milliseconds */ + + while (!kthread_should_stop()) { + if (tout && tout <= 20) + __set_current_state(TASK_KILLABLE); + else + __set_current_state(TASK_INTERRUPTIBLE); + schedule_timeout(tout ? + msecs_to_jiffies(tout) : MAX_SCHEDULE_TIMEOUT); + + try_to_freeze(); + + tout = xfsaild_push(ailp); + } + + return 0; } /* @@ -574,8 +577,9 @@ xfs_ail_push( */ smp_wmb(); xfs_trans_ail_copy_lsn(ailp, &ailp->xa_target, &threshold_lsn); - if (!test_and_set_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags)) - queue_delayed_work(xfs_syncd_wq, &ailp->xa_work, 0); + smp_wmb(); + + wake_up_process(ailp->xa_task); } /* @@ -813,9 +817,18 @@ xfs_trans_ail_init( INIT_LIST_HEAD(&ailp->xa_ail); INIT_LIST_HEAD(&ailp->xa_cursors); spin_lock_init(&ailp->xa_lock); - INIT_DELAYED_WORK(&ailp->xa_work, xfs_ail_worker); + + ailp->xa_task = kthread_run(xfsaild, ailp, "xfsaild/%s", + ailp->xa_mount->m_fsname); + if (IS_ERR(ailp->xa_task)) + goto out_free_ailp; + mp->m_ail = ailp; return 0; + +out_free_ailp: + kmem_free(ailp); + return ENOMEM; } void @@ -824,6 +837,6 @@ xfs_trans_ail_destroy( { struct xfs_ail *ailp = mp->m_ail; - cancel_delayed_work_sync(&ailp->xa_work); + kthread_stop(ailp->xa_task); kmem_free(ailp); } diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index 212946b97239..22750b5e4a8f 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -64,23 +64,17 @@ struct xfs_ail_cursor { */ struct xfs_ail { struct xfs_mount *xa_mount; + struct task_struct *xa_task; struct list_head xa_ail; xfs_lsn_t xa_target; struct list_head xa_cursors; spinlock_t xa_lock; - struct delayed_work xa_work; xfs_lsn_t xa_last_pushed_lsn; - unsigned long xa_flags; }; -#define XFS_AIL_PUSHING_BIT 0 - /* * From xfs_trans_ail.c */ - -extern struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ - void xfs_trans_ail_update_bulk(struct xfs_ail *ailp, struct xfs_ail_cursor *cur, struct xfs_log_item **log_items, int nr_items, |
