diff options
Diffstat (limited to 'fs/nilfs2')
33 files changed, 2136 insertions, 768 deletions
diff --git a/fs/nilfs2/Kconfig b/fs/nilfs2/Kconfig index 251da07b2a1..80da8eb2739 100644 --- a/fs/nilfs2/Kconfig +++ b/fs/nilfs2/Kconfig @@ -1,6 +1,5 @@ config NILFS2_FS - tristate "NILFS2 file system support (EXPERIMENTAL)" - depends on EXPERIMENTAL + tristate "NILFS2 file system support" select CRC32 help NILFS2 is a log-structured file system (LFS) supporting continuous diff --git a/fs/nilfs2/alloc.c b/fs/nilfs2/alloc.c index 0a0a66d98cc..741fd02e044 100644 --- a/fs/nilfs2/alloc.c +++ b/fs/nilfs2/alloc.c @@ -398,6 +398,69 @@ nilfs_palloc_rest_groups_in_desc_block(const struct inode *inode, } /** + * nilfs_palloc_count_desc_blocks - count descriptor blocks number + * @inode: inode of metadata file using this allocator + * @desc_blocks: descriptor blocks number [out] + */ +static int nilfs_palloc_count_desc_blocks(struct inode *inode, + unsigned long *desc_blocks) +{ + unsigned long blknum; + int ret; + + ret = nilfs_bmap_last_key(NILFS_I(inode)->i_bmap, &blknum); + if (likely(!ret)) + *desc_blocks = DIV_ROUND_UP( + blknum, NILFS_MDT(inode)->mi_blocks_per_desc_block); + return ret; +} + +/** + * nilfs_palloc_mdt_file_can_grow - check potential opportunity for + * MDT file growing + * @inode: inode of metadata file using this allocator + * @desc_blocks: known current descriptor blocks count + */ +static inline bool nilfs_palloc_mdt_file_can_grow(struct inode *inode, + unsigned long desc_blocks) +{ + return (nilfs_palloc_groups_per_desc_block(inode) * desc_blocks) < + nilfs_palloc_groups_count(inode); +} + +/** + * nilfs_palloc_count_max_entries - count max number of entries that can be + * described by descriptor blocks count + * @inode: inode of metadata file using this allocator + * @nused: current number of used entries + * @nmaxp: max number of entries [out] + */ +int nilfs_palloc_count_max_entries(struct inode *inode, u64 nused, u64 *nmaxp) +{ + unsigned long desc_blocks = 0; + u64 entries_per_desc_block, nmax; + int err; + + err = nilfs_palloc_count_desc_blocks(inode, &desc_blocks); + if (unlikely(err)) + return err; + + entries_per_desc_block = (u64)nilfs_palloc_entries_per_group(inode) * + nilfs_palloc_groups_per_desc_block(inode); + nmax = entries_per_desc_block * desc_blocks; + + if (nused == nmax && + nilfs_palloc_mdt_file_can_grow(inode, desc_blocks)) + nmax += entries_per_desc_block; + + if (nused > nmax) + return -ERANGE; + + *nmaxp = nmax; + return 0; +} + +/** * nilfs_palloc_prepare_alloc_entry - prepare to allocate a persistent object * @inode: inode of metadata file using this allocator * @req: nilfs_palloc_req structure exchanged for the allocation @@ -489,8 +552,8 @@ int nilfs_palloc_prepare_alloc_entry(struct inode *inode, void nilfs_palloc_commit_alloc_entry(struct inode *inode, struct nilfs_palloc_req *req) { - nilfs_mdt_mark_buffer_dirty(req->pr_bitmap_bh); - nilfs_mdt_mark_buffer_dirty(req->pr_desc_bh); + mark_buffer_dirty(req->pr_bitmap_bh); + mark_buffer_dirty(req->pr_desc_bh); nilfs_mdt_mark_dirty(inode); brelse(req->pr_bitmap_bh); @@ -527,8 +590,8 @@ void nilfs_palloc_commit_free_entry(struct inode *inode, kunmap(req->pr_bitmap_bh->b_page); kunmap(req->pr_desc_bh->b_page); - nilfs_mdt_mark_buffer_dirty(req->pr_desc_bh); - nilfs_mdt_mark_buffer_dirty(req->pr_bitmap_bh); + mark_buffer_dirty(req->pr_desc_bh); + mark_buffer_dirty(req->pr_bitmap_bh); nilfs_mdt_mark_dirty(inode); brelse(req->pr_bitmap_bh); @@ -646,7 +709,7 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) unsigned long group, group_offset; int i, j, n, ret; - for (i = 0; i < nitems; i += n) { + for (i = 0; i < nitems; i = j) { group = nilfs_palloc_group(inode, entry_nrs[i], &group_offset); ret = nilfs_palloc_get_desc_block(inode, group, 0, &desc_bh); if (ret < 0) @@ -683,8 +746,8 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) kunmap(bitmap_bh->b_page); kunmap(desc_bh->b_page); - nilfs_mdt_mark_buffer_dirty(desc_bh); - nilfs_mdt_mark_buffer_dirty(bitmap_bh); + mark_buffer_dirty(desc_bh); + mark_buffer_dirty(bitmap_bh); nilfs_mdt_mark_dirty(inode); brelse(bitmap_bh); diff --git a/fs/nilfs2/alloc.h b/fs/nilfs2/alloc.h index f5fde36b9e2..4bd6451b570 100644 --- a/fs/nilfs2/alloc.h +++ b/fs/nilfs2/alloc.h @@ -48,6 +48,8 @@ int nilfs_palloc_get_entry_block(struct inode *, __u64, int, void *nilfs_palloc_block_get_entry(const struct inode *, __u64, const struct buffer_head *, void *); +int nilfs_palloc_count_max_entries(struct inode *, u64, u64 *); + /** * nilfs_palloc_req - persistent allocator request and reply * @pr_entry_nr: entry number (vblocknr or inode number) @@ -76,15 +78,23 @@ int nilfs_palloc_freev(struct inode *, __u64 *, size_t); #define nilfs_clear_bit_atomic ext2_clear_bit_atomic #define nilfs_find_next_zero_bit find_next_zero_bit_le -/* - * persistent object allocator cache +/** + * struct nilfs_bh_assoc - block offset and buffer head association + * @blkoff: block offset + * @bh: buffer head */ - struct nilfs_bh_assoc { unsigned long blkoff; struct buffer_head *bh; }; +/** + * struct nilfs_palloc_cache - persistent object allocator cache + * @lock: cache protecting lock + * @prev_desc: blockgroup descriptors cache + * @prev_bitmap: blockgroup bitmap cache + * @prev_entry: translation entries cache + */ struct nilfs_palloc_cache { spinlock_t lock; struct nilfs_bh_assoc prev_desc; diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index 4723f04e9b1..aadbd0b5e3e 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -34,7 +34,9 @@ struct inode *nilfs_bmap_get_dat(const struct nilfs_bmap *bmap) { - return NILFS_I_NILFS(bmap->b_inode)->ns_dat; + struct the_nilfs *nilfs = bmap->b_inode->i_sb->s_fs_info; + + return nilfs->ns_dat; } static int nilfs_bmap_convert_error(struct nilfs_bmap *bmap, diff --git a/fs/nilfs2/bmap.h b/fs/nilfs2/bmap.h index 40d9f453d31..b89e68076ad 100644 --- a/fs/nilfs2/bmap.h +++ b/fs/nilfs2/bmap.h @@ -135,6 +135,13 @@ struct nilfs_bmap { /* state */ #define NILFS_BMAP_DIRTY 0x00000001 +/** + * struct nilfs_bmap_store - shadow copy of bmap state + * @data: cached raw block mapping of on-disk inode + * @last_allocated_key: cached value of last allocated key for data block + * @last_allocated_ptr: cached value of last allocated ptr for data block + * @state: cached value of state field of bmap structure + */ struct nilfs_bmap_store { __le64 data[NILFS_BMAP_SIZE / sizeof(__le64)]; __u64 last_allocated_key; diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 609cd223eea..a35ae35e693 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -34,12 +34,6 @@ #include "page.h" #include "btnode.h" -void nilfs_btnode_cache_init(struct address_space *btnc, - struct backing_dev_info *bdi) -{ - nilfs_mapping_init(btnc, bdi); -} - void nilfs_btnode_cache_clear(struct address_space *btnc) { invalidate_mapping_pages(btnc, 0, -1); @@ -62,7 +56,7 @@ nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr) BUG(); } memset(bh->b_data, 0, 1 << inode->i_blkbits); - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_bdev = inode->i_sb->s_bdev; bh->b_blocknr = blocknr; set_buffer_mapped(bh); set_buffer_uptodate(bh); @@ -94,10 +88,11 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, if (pblocknr == 0) { pblocknr = blocknr; if (inode->i_ino != NILFS_DAT_INO) { - struct inode *dat = NILFS_I_NILFS(inode)->ns_dat; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; /* blocknr is a virtual block number */ - err = nilfs_dat_translate(dat, blocknr, &pblocknr); + err = nilfs_dat_translate(nilfs->ns_dat, blocknr, + &pblocknr); if (unlikely(err)) { brelse(bh); goto out_locked; @@ -120,7 +115,7 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, goto found; } set_buffer_mapped(bh); - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_bdev = inode->i_sb->s_bdev; bh->b_blocknr = pblocknr; /* set block address for read */ bh->b_end_io = end_buffer_read_sync; get_bh(bh); @@ -259,7 +254,7 @@ void nilfs_btnode_commit_change_key(struct address_space *btnc, "invalid oldkey %lld (newkey=%lld)", (unsigned long long)oldkey, (unsigned long long)newkey); - nilfs_btnode_mark_dirty(obh); + mark_buffer_dirty(obh); spin_lock_irq(&btnc->tree_lock); radix_tree_delete(&btnc->page_tree, oldkey); @@ -271,7 +266,7 @@ void nilfs_btnode_commit_change_key(struct address_space *btnc, unlock_page(opage); } else { nilfs_copy_buffer(nbh, obh); - nilfs_btnode_mark_dirty(nbh); + mark_buffer_dirty(nbh); nbh->b_blocknr = newkey; ctxt->bh = nbh; diff --git a/fs/nilfs2/btnode.h b/fs/nilfs2/btnode.h index 1b8ebd888c2..d876b565ce6 100644 --- a/fs/nilfs2/btnode.h +++ b/fs/nilfs2/btnode.h @@ -29,7 +29,13 @@ #include <linux/fs.h> #include <linux/backing-dev.h> - +/** + * struct nilfs_btnode_chkey_ctxt - change key context + * @oldkey: old key of block's moving content + * @newkey: new key for block's content + * @bh: buffer head of old buffer + * @newbh: buffer head of new buffer + */ struct nilfs_btnode_chkey_ctxt { __u64 oldkey; __u64 newkey; @@ -37,7 +43,6 @@ struct nilfs_btnode_chkey_ctxt { struct buffer_head *newbh; }; -void nilfs_btnode_cache_init(struct address_space *, struct backing_dev_info *); void nilfs_btnode_cache_clear(struct address_space *); struct buffer_head *nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr); @@ -51,7 +56,4 @@ void nilfs_btnode_commit_change_key(struct address_space *, void nilfs_btnode_abort_change_key(struct address_space *, struct nilfs_btnode_chkey_ctxt *); -#define nilfs_btnode_mark_dirty(bh) nilfs_mark_buffer_dirty(bh) - - #endif /* _NILFS_BTNODE_H */ diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index d451ae0e0bf..b2e3ff34762 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -714,7 +714,7 @@ static void nilfs_btree_promote_key(struct nilfs_bmap *btree, nilfs_btree_get_nonroot_node(path, level), path[level].bp_index, key); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); } while ((path[level].bp_index == 0) && (++level < nilfs_btree_height(btree) - 1)); } @@ -739,7 +739,7 @@ static void nilfs_btree_do_insert(struct nilfs_bmap *btree, nilfs_btree_node_insert(node, path[level].bp_index, *keyp, *ptrp, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (path[level].bp_index == 0) nilfs_btree_promote_key(btree, path, level + 1, @@ -777,9 +777,9 @@ static void nilfs_btree_carry_left(struct nilfs_bmap *btree, nilfs_btree_node_move_left(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -823,9 +823,9 @@ static void nilfs_btree_carry_right(struct nilfs_bmap *btree, nilfs_btree_node_move_right(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); path[level + 1].bp_index++; nilfs_btree_promote_key(btree, path, level + 1, @@ -870,9 +870,9 @@ static void nilfs_btree_split(struct nilfs_bmap *btree, nilfs_btree_node_move_right(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); newkey = nilfs_btree_node_get_key(right, 0); newptr = path[level].bp_newreq.bpr_ptr; @@ -919,7 +919,7 @@ static void nilfs_btree_grow(struct nilfs_bmap *btree, nilfs_btree_node_set_level(root, level + 1); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); path[level].bp_bh = path[level].bp_sib_bh; path[level].bp_sib_bh = NULL; @@ -1194,7 +1194,7 @@ static void nilfs_btree_do_delete(struct nilfs_bmap *btree, nilfs_btree_node_delete(node, path[level].bp_index, keyp, ptrp, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (path[level].bp_index == 0) nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -1226,9 +1226,9 @@ static void nilfs_btree_borrow_left(struct nilfs_bmap *btree, nilfs_btree_node_move_right(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -1258,9 +1258,9 @@ static void nilfs_btree_borrow_right(struct nilfs_bmap *btree, nilfs_btree_node_move_left(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); path[level + 1].bp_index++; nilfs_btree_promote_key(btree, path, level + 1, @@ -1289,7 +1289,7 @@ static void nilfs_btree_concat_left(struct nilfs_bmap *btree, nilfs_btree_node_move_left(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); nilfs_btnode_delete(path[level].bp_bh); path[level].bp_bh = path[level].bp_sib_bh; @@ -1315,7 +1315,7 @@ static void nilfs_btree_concat_right(struct nilfs_bmap *btree, nilfs_btree_node_move_left(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); nilfs_btnode_delete(path[level].bp_sib_bh); path[level].bp_sib_bh = NULL; @@ -1346,6 +1346,11 @@ static void nilfs_btree_shrink(struct nilfs_bmap *btree, path[level].bp_bh = NULL; } +static void nilfs_btree_nop(struct nilfs_bmap *btree, + struct nilfs_btree_path *path, + int level, __u64 *keyp, __u64 *ptrp) +{ +} static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree, struct nilfs_btree_path *path, @@ -1356,20 +1361,19 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree, struct buffer_head *bh; struct nilfs_btree_node *node, *parent, *sib; __u64 sibptr; - int pindex, level, ncmin, ncmax, ncblk, ret; + int pindex, dindex, level, ncmin, ncmax, ncblk, ret; ret = 0; stats->bs_nblocks = 0; ncmin = NILFS_BTREE_NODE_NCHILDREN_MIN(nilfs_btree_node_size(btree)); ncblk = nilfs_btree_nchildren_per_block(btree); - for (level = NILFS_BTREE_LEVEL_NODE_MIN; + for (level = NILFS_BTREE_LEVEL_NODE_MIN, dindex = path[level].bp_index; level < nilfs_btree_height(btree) - 1; level++) { node = nilfs_btree_get_nonroot_node(path, level); path[level].bp_oldreq.bpr_ptr = - nilfs_btree_node_get_ptr(node, path[level].bp_index, - ncblk); + nilfs_btree_node_get_ptr(node, dindex, ncblk); ret = nilfs_bmap_prepare_end_ptr(btree, &path[level].bp_oldreq, dat); if (ret < 0) @@ -1383,6 +1387,7 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree, parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); pindex = path[level + 1].bp_index; + dindex = pindex; if (pindex > 0) { /* left sibling */ @@ -1421,6 +1426,14 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree, path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_concat_right; stats->bs_nblocks++; + /* + * When merging right sibling node + * into the current node, pointer to + * the right sibling node must be + * terminated instead. The adjustment + * below is required for that. + */ + dindex = pindex + 1; /* continue; */ } } else { @@ -1431,29 +1444,31 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree, NILFS_BTREE_ROOT_NCHILDREN_MAX) { path[level].bp_op = nilfs_btree_shrink; stats->bs_nblocks += 2; + level++; + path[level].bp_op = nilfs_btree_nop; + goto shrink_root_child; } else { path[level].bp_op = nilfs_btree_do_delete; stats->bs_nblocks++; + goto out; } - - goto out; - } } + /* child of the root node is deleted */ + path[level].bp_op = nilfs_btree_do_delete; + stats->bs_nblocks++; + +shrink_root_child: node = nilfs_btree_get_root(btree); path[level].bp_oldreq.bpr_ptr = - nilfs_btree_node_get_ptr(node, path[level].bp_index, + nilfs_btree_node_get_ptr(node, dindex, NILFS_BTREE_ROOT_NCHILDREN_MAX); ret = nilfs_bmap_prepare_end_ptr(btree, &path[level].bp_oldreq, dat); if (ret < 0) goto err_out_child_node; - /* child of the root node is deleted */ - path[level].bp_op = nilfs_btree_do_delete; - stats->bs_nblocks++; - /* success */ out: *levelp = level; @@ -1709,7 +1724,7 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *btree, nilfs_btree_node_init(node, 0, 1, n, ncblk, keys, ptrs); nilfs_btree_node_insert(node, n, key, dreq->bpr_ptr, ncblk); if (!buffer_dirty(bh)) - nilfs_btnode_mark_dirty(bh); + mark_buffer_dirty(bh); if (!nilfs_bmap_dirty(btree)) nilfs_bmap_set_dirty(btree); @@ -1787,7 +1802,7 @@ static int nilfs_btree_propagate_p(struct nilfs_bmap *btree, { while ((++level < nilfs_btree_height(btree) - 1) && !buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); return 0; } @@ -2229,7 +2244,7 @@ static int nilfs_btree_mark(struct nilfs_bmap *btree, __u64 key, int level) } if (!buffer_dirty(bh)) - nilfs_btnode_mark_dirty(bh); + mark_buffer_dirty(bh); brelse(bh); if (!nilfs_bmap_dirty(btree)) nilfs_bmap_set_dirty(btree); diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c index 5ff15a8a102..0d58075f34e 100644 --- a/fs/nilfs2/cpfile.c +++ b/fs/nilfs2/cpfile.c @@ -216,14 +216,14 @@ int nilfs_cpfile_get_checkpoint(struct inode *cpfile, if (!nilfs_cpfile_is_in_first(cpfile, cno)) nilfs_cpfile_block_add_valid_checkpoints(cpfile, cp_bh, kaddr, 1); - nilfs_mdt_mark_buffer_dirty(cp_bh); + mark_buffer_dirty(cp_bh); - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); le64_add_cpu(&header->ch_ncheckpoints, 1); - kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(header_bh); + kunmap_atomic(kaddr); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); } @@ -286,7 +286,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, __u64 cno; void *kaddr; unsigned long tnicps; - int ret, ncps, nicps, count, i; + int ret, ncps, nicps, nss, count, i; if (unlikely(start == 0 || start > end)) { printk(KERN_ERR "%s: invalid range of checkpoint numbers: " @@ -301,6 +301,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, if (ret < 0) goto out_sem; tnicps = 0; + nss = 0; for (cno = start; cno < end; cno += ncps) { ncps = nilfs_cpfile_checkpoints_in_block(cpfile, cno, end); @@ -313,20 +314,21 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, continue; } - kaddr = kmap_atomic(cp_bh->b_page, KM_USER0); + kaddr = kmap_atomic(cp_bh->b_page); cp = nilfs_cpfile_block_get_checkpoint( cpfile, cno, cp_bh, kaddr); nicps = 0; for (i = 0; i < ncps; i++, cp = (void *)cp + cpsz) { - WARN_ON(nilfs_checkpoint_snapshot(cp)); - if (!nilfs_checkpoint_invalid(cp)) { + if (nilfs_checkpoint_snapshot(cp)) { + nss++; + } else if (!nilfs_checkpoint_invalid(cp)) { nilfs_checkpoint_set_invalid(cp); nicps++; } } if (nicps > 0) { tnicps += nicps; - nilfs_mdt_mark_buffer_dirty(cp_bh); + mark_buffer_dirty(cp_bh); nilfs_mdt_mark_dirty(cpfile); if (!nilfs_cpfile_is_in_first(cpfile, cno)) { count = @@ -334,7 +336,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, cpfile, cp_bh, kaddr, nicps); if (count == 0) { /* make hole */ - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(cp_bh); ret = nilfs_cpfile_delete_checkpoint_block( @@ -349,21 +351,23 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, } } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(cp_bh); } if (tnicps > 0) { - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); le64_add_cpu(&header->ch_ncheckpoints, -(u64)tnicps); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); } brelse(header_bh); + if (nss > 0) + ret = -EBUSY; out_sem: up_write(&NILFS_MDT(cpfile)->mi_sem); @@ -408,7 +412,7 @@ static ssize_t nilfs_cpfile_do_get_cpinfo(struct inode *cpfile, __u64 *cnop, continue; /* skip hole */ } - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); cp = nilfs_cpfile_block_get_checkpoint(cpfile, cno, bh, kaddr); for (i = 0; i < ncps && n < nci; i++, cp = (void *)cp + cpsz) { if (!nilfs_checkpoint_invalid(cp)) { @@ -418,7 +422,7 @@ static ssize_t nilfs_cpfile_do_get_cpinfo(struct inode *cpfile, __u64 *cnop, n++; } } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh); } @@ -451,10 +455,10 @@ static ssize_t nilfs_cpfile_do_get_ssinfo(struct inode *cpfile, __u64 *cnop, ret = nilfs_cpfile_get_header_block(cpfile, &bh); if (ret < 0) goto out; - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, bh, kaddr); curr = le64_to_cpu(header->ch_snapshot_list.ssl_next); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh); if (curr == 0) { ret = 0; @@ -472,7 +476,7 @@ static ssize_t nilfs_cpfile_do_get_ssinfo(struct inode *cpfile, __u64 *cnop, ret = 0; /* No snapshots (started from a hole block) */ goto out; } - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); while (n < nci) { cp = nilfs_cpfile_block_get_checkpoint(cpfile, curr, bh, kaddr); curr = ~(__u64)0; /* Terminator */ @@ -488,7 +492,7 @@ static ssize_t nilfs_cpfile_do_get_ssinfo(struct inode *cpfile, __u64 *cnop, next_blkoff = nilfs_cpfile_get_blkoff(cpfile, next); if (curr_blkoff != next_blkoff) { - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh); ret = nilfs_cpfile_get_checkpoint_block(cpfile, next, 0, &bh); @@ -496,12 +500,12 @@ static ssize_t nilfs_cpfile_do_get_ssinfo(struct inode *cpfile, __u64 *cnop, WARN_ON(ret == -ENOENT); goto out; } - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); } curr = next; curr_blkoff = next_blkoff; } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh); *cnop = curr; ret = n; @@ -592,24 +596,24 @@ static int nilfs_cpfile_set_snapshot(struct inode *cpfile, __u64 cno) ret = nilfs_cpfile_get_checkpoint_block(cpfile, cno, 0, &cp_bh); if (ret < 0) goto out_sem; - kaddr = kmap_atomic(cp_bh->b_page, KM_USER0); + kaddr = kmap_atomic(cp_bh->b_page); cp = nilfs_cpfile_block_get_checkpoint(cpfile, cno, cp_bh, kaddr); if (nilfs_checkpoint_invalid(cp)) { ret = -ENOENT; - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); goto out_cp; } if (nilfs_checkpoint_snapshot(cp)) { ret = 0; - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); goto out_cp; } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); ret = nilfs_cpfile_get_header_block(cpfile, &header_bh); if (ret < 0) goto out_cp; - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); list = &header->ch_snapshot_list; curr_bh = header_bh; @@ -621,13 +625,13 @@ static int nilfs_cpfile_set_snapshot(struct inode *cpfile, __u64 cno) prev_blkoff = nilfs_cpfile_get_blkoff(cpfile, prev); curr = prev; if (curr_blkoff != prev_blkoff) { - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(curr_bh); ret = nilfs_cpfile_get_checkpoint_block(cpfile, curr, 0, &curr_bh); if (ret < 0) goto out_header; - kaddr = kmap_atomic(curr_bh->b_page, KM_USER0); + kaddr = kmap_atomic(curr_bh->b_page); } curr_blkoff = prev_blkoff; cp = nilfs_cpfile_block_get_checkpoint( @@ -635,7 +639,7 @@ static int nilfs_cpfile_set_snapshot(struct inode *cpfile, __u64 cno) list = &cp->cp_snapshot_list; prev = le64_to_cpu(list->ssl_prev); } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); if (prev != 0) { ret = nilfs_cpfile_get_checkpoint_block(cpfile, prev, 0, @@ -647,34 +651,34 @@ static int nilfs_cpfile_set_snapshot(struct inode *cpfile, __u64 cno) get_bh(prev_bh); } - kaddr = kmap_atomic(curr_bh->b_page, KM_USER0); + kaddr = kmap_atomic(curr_bh->b_page); list = nilfs_cpfile_block_get_snapshot_list( cpfile, curr, curr_bh, kaddr); list->ssl_prev = cpu_to_le64(cno); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(cp_bh->b_page, KM_USER0); + kaddr = kmap_atomic(cp_bh->b_page); cp = nilfs_cpfile_block_get_checkpoint(cpfile, cno, cp_bh, kaddr); cp->cp_snapshot_list.ssl_next = cpu_to_le64(curr); cp->cp_snapshot_list.ssl_prev = cpu_to_le64(prev); nilfs_checkpoint_set_snapshot(cp); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(prev_bh->b_page, KM_USER0); + kaddr = kmap_atomic(prev_bh->b_page); list = nilfs_cpfile_block_get_snapshot_list( cpfile, prev, prev_bh, kaddr); list->ssl_next = cpu_to_le64(cno); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); le64_add_cpu(&header->ch_nsnapshots, 1); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - nilfs_mdt_mark_buffer_dirty(prev_bh); - nilfs_mdt_mark_buffer_dirty(curr_bh); - nilfs_mdt_mark_buffer_dirty(cp_bh); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(prev_bh); + mark_buffer_dirty(curr_bh); + mark_buffer_dirty(cp_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); brelse(prev_bh); @@ -710,23 +714,23 @@ static int nilfs_cpfile_clear_snapshot(struct inode *cpfile, __u64 cno) ret = nilfs_cpfile_get_checkpoint_block(cpfile, cno, 0, &cp_bh); if (ret < 0) goto out_sem; - kaddr = kmap_atomic(cp_bh->b_page, KM_USER0); + kaddr = kmap_atomic(cp_bh->b_page); cp = nilfs_cpfile_block_get_checkpoint(cpfile, cno, cp_bh, kaddr); if (nilfs_checkpoint_invalid(cp)) { ret = -ENOENT; - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); goto out_cp; } if (!nilfs_checkpoint_snapshot(cp)) { ret = 0; - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); goto out_cp; } list = &cp->cp_snapshot_list; next = le64_to_cpu(list->ssl_next); prev = le64_to_cpu(list->ssl_prev); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); ret = nilfs_cpfile_get_header_block(cpfile, &header_bh); if (ret < 0) @@ -750,34 +754,34 @@ static int nilfs_cpfile_clear_snapshot(struct inode *cpfile, __u64 cno) get_bh(prev_bh); } - kaddr = kmap_atomic(next_bh->b_page, KM_USER0); + kaddr = kmap_atomic(next_bh->b_page); list = nilfs_cpfile_block_get_snapshot_list( cpfile, next, next_bh, kaddr); list->ssl_prev = cpu_to_le64(prev); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(prev_bh->b_page, KM_USER0); + kaddr = kmap_atomic(prev_bh->b_page); list = nilfs_cpfile_block_get_snapshot_list( cpfile, prev, prev_bh, kaddr); list->ssl_next = cpu_to_le64(next); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(cp_bh->b_page, KM_USER0); + kaddr = kmap_atomic(cp_bh->b_page); cp = nilfs_cpfile_block_get_checkpoint(cpfile, cno, cp_bh, kaddr); cp->cp_snapshot_list.ssl_next = cpu_to_le64(0); cp->cp_snapshot_list.ssl_prev = cpu_to_le64(0); nilfs_checkpoint_clear_snapshot(cp); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); le64_add_cpu(&header->ch_nsnapshots, -1); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - nilfs_mdt_mark_buffer_dirty(next_bh); - nilfs_mdt_mark_buffer_dirty(prev_bh); - nilfs_mdt_mark_buffer_dirty(cp_bh); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(next_bh); + mark_buffer_dirty(prev_bh); + mark_buffer_dirty(cp_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); brelse(prev_bh); @@ -829,13 +833,13 @@ int nilfs_cpfile_is_snapshot(struct inode *cpfile, __u64 cno) ret = nilfs_cpfile_get_checkpoint_block(cpfile, cno, 0, &bh); if (ret < 0) goto out; - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); cp = nilfs_cpfile_block_get_checkpoint(cpfile, cno, bh, kaddr); if (nilfs_checkpoint_invalid(cp)) ret = -ENOENT; else ret = nilfs_checkpoint_snapshot(cp); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh); out: @@ -912,12 +916,12 @@ int nilfs_cpfile_get_stat(struct inode *cpfile, struct nilfs_cpstat *cpstat) ret = nilfs_cpfile_get_header_block(cpfile, &bh); if (ret < 0) goto out_sem; - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); header = nilfs_cpfile_block_get_header(cpfile, bh, kaddr); cpstat->cs_cno = nilfs_mdt_cno(cpfile); cpstat->cs_ncps = le64_to_cpu(header->ch_ncheckpoints); cpstat->cs_nsss = le64_to_cpu(header->ch_nsnapshots); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh); out_sem: @@ -938,6 +942,18 @@ int nilfs_cpfile_read(struct super_block *sb, size_t cpsize, struct inode *cpfile; int err; + if (cpsize > sb->s_blocksize) { + printk(KERN_ERR + "NILFS: too large checkpoint size: %zu bytes.\n", + cpsize); + return -EINVAL; + } else if (cpsize < NILFS_MIN_CHECKPOINT_SIZE) { + printk(KERN_ERR + "NILFS: too small checkpoint size: %zu bytes.\n", + cpsize); + return -EINVAL; + } + cpfile = nilfs_iget_locked(sb, NULL, NILFS_CPFILE_INO); if (unlikely(!cpfile)) return -ENOMEM; diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c index 59e5fe742f7..0d5fada9119 100644 --- a/fs/nilfs2/dat.c +++ b/fs/nilfs2/dat.c @@ -33,6 +33,12 @@ #define NILFS_CNO_MIN ((__u64)1) #define NILFS_CNO_MAX (~(__u64)0) +/** + * struct nilfs_dat_info - on-memory private data of DAT file + * @mi: on-memory private data of metadata file + * @palloc_cache: persistent object allocator cache of DAT file + * @shadow: shadow map of DAT file + */ struct nilfs_dat_info { struct nilfs_mdt_info mi; struct nilfs_palloc_cache palloc_cache; @@ -54,7 +60,7 @@ static int nilfs_dat_prepare_entry(struct inode *dat, static void nilfs_dat_commit_entry(struct inode *dat, struct nilfs_palloc_req *req) { - nilfs_mdt_mark_buffer_dirty(req->pr_entry_bh); + mark_buffer_dirty(req->pr_entry_bh); nilfs_mdt_mark_dirty(dat); brelse(req->pr_entry_bh); } @@ -85,13 +91,13 @@ void nilfs_dat_commit_alloc(struct inode *dat, struct nilfs_palloc_req *req) struct nilfs_dat_entry *entry; void *kaddr; - kaddr = kmap_atomic(req->pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req->pr_entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); entry->de_start = cpu_to_le64(NILFS_CNO_MIN); entry->de_end = cpu_to_le64(NILFS_CNO_MAX); entry->de_blocknr = cpu_to_le64(0); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nilfs_palloc_commit_alloc_entry(dat, req); nilfs_dat_commit_entry(dat, req); @@ -109,13 +115,13 @@ static void nilfs_dat_commit_free(struct inode *dat, struct nilfs_dat_entry *entry; void *kaddr; - kaddr = kmap_atomic(req->pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req->pr_entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); entry->de_start = cpu_to_le64(NILFS_CNO_MIN); entry->de_end = cpu_to_le64(NILFS_CNO_MIN); entry->de_blocknr = cpu_to_le64(0); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nilfs_dat_commit_entry(dat, req); nilfs_palloc_commit_free_entry(dat, req); @@ -136,12 +142,12 @@ void nilfs_dat_commit_start(struct inode *dat, struct nilfs_palloc_req *req, struct nilfs_dat_entry *entry; void *kaddr; - kaddr = kmap_atomic(req->pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req->pr_entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); entry->de_start = cpu_to_le64(nilfs_mdt_cno(dat)); entry->de_blocknr = cpu_to_le64(blocknr); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nilfs_dat_commit_entry(dat, req); } @@ -160,12 +166,12 @@ int nilfs_dat_prepare_end(struct inode *dat, struct nilfs_palloc_req *req) return ret; } - kaddr = kmap_atomic(req->pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req->pr_entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); start = le64_to_cpu(entry->de_start); blocknr = le64_to_cpu(entry->de_blocknr); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); if (blocknr == 0) { ret = nilfs_palloc_prepare_free_entry(dat, req); @@ -186,7 +192,7 @@ void nilfs_dat_commit_end(struct inode *dat, struct nilfs_palloc_req *req, sector_t blocknr; void *kaddr; - kaddr = kmap_atomic(req->pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req->pr_entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); end = start = le64_to_cpu(entry->de_start); @@ -196,7 +202,7 @@ void nilfs_dat_commit_end(struct inode *dat, struct nilfs_palloc_req *req, } entry->de_end = cpu_to_le64(end); blocknr = le64_to_cpu(entry->de_blocknr); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); if (blocknr == 0) nilfs_dat_commit_free(dat, req); @@ -211,12 +217,12 @@ void nilfs_dat_abort_end(struct inode *dat, struct nilfs_palloc_req *req) sector_t blocknr; void *kaddr; - kaddr = kmap_atomic(req->pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req->pr_entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); start = le64_to_cpu(entry->de_start); blocknr = le64_to_cpu(entry->de_blocknr); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); if (start == nilfs_mdt_cno(dat) && blocknr == 0) nilfs_palloc_abort_free_entry(dat, req); @@ -346,22 +352,22 @@ int nilfs_dat_move(struct inode *dat, __u64 vblocknr, sector_t blocknr) } } - kaddr = kmap_atomic(entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, vblocknr, entry_bh, kaddr); if (unlikely(entry->de_blocknr == cpu_to_le64(0))) { printk(KERN_CRIT "%s: vbn = %llu, [%llu, %llu)\n", __func__, (unsigned long long)vblocknr, (unsigned long long)le64_to_cpu(entry->de_start), (unsigned long long)le64_to_cpu(entry->de_end)); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(entry_bh); return -EINVAL; } WARN_ON(blocknr == 0); entry->de_blocknr = cpu_to_le64(blocknr); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - nilfs_mdt_mark_buffer_dirty(entry_bh); + mark_buffer_dirty(entry_bh); nilfs_mdt_mark_dirty(dat); brelse(entry_bh); @@ -409,7 +415,7 @@ int nilfs_dat_translate(struct inode *dat, __u64 vblocknr, sector_t *blocknrp) } } - kaddr = kmap_atomic(entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, vblocknr, entry_bh, kaddr); blocknr = le64_to_cpu(entry->de_blocknr); if (blocknr == 0) { @@ -419,7 +425,7 @@ int nilfs_dat_translate(struct inode *dat, __u64 vblocknr, sector_t *blocknrp) *blocknrp = blocknr; out: - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(entry_bh); return ret; } @@ -440,7 +446,7 @@ ssize_t nilfs_dat_get_vinfo(struct inode *dat, void *buf, unsigned visz, 0, &entry_bh); if (ret < 0) return ret; - kaddr = kmap_atomic(entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(entry_bh->b_page); /* last virtual block number in this block */ first = vinfo->vi_vblocknr; do_div(first, entries_per_block); @@ -456,7 +462,7 @@ ssize_t nilfs_dat_get_vinfo(struct inode *dat, void *buf, unsigned visz, vinfo->vi_end = le64_to_cpu(entry->de_end); vinfo->vi_blocknr = le64_to_cpu(entry->de_blocknr); } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(entry_bh); } @@ -478,6 +484,18 @@ int nilfs_dat_read(struct super_block *sb, size_t entry_size, struct nilfs_dat_info *di; int err; + if (entry_size > sb->s_blocksize) { + printk(KERN_ERR + "NILFS: too large DAT entry size: %zu bytes.\n", + entry_size); + return -EINVAL; + } else if (entry_size < NILFS_MIN_DAT_ENTRY_SIZE) { + printk(KERN_ERR + "NILFS: too small DAT entry size: %zu bytes.\n", + entry_size); + return -EINVAL; + } + dat = nilfs_iget_locked(sb, NULL, NILFS_DAT_INO); if (unlikely(!dat)) return -ENOMEM; diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index 3a1923943b1..197a63e9d10 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -251,27 +251,23 @@ nilfs_type_by_mode[S_IFMT >> S_SHIFT] = { static void nilfs_set_de_type(struct nilfs_dir_entry *de, struct inode *inode) { - mode_t mode = inode->i_mode; + umode_t mode = inode->i_mode; de->file_type = nilfs_type_by_mode[(mode & S_IFMT)>>S_SHIFT]; } -static int nilfs_readdir(struct file *filp, void *dirent, filldir_t filldir) +static int nilfs_readdir(struct file *file, struct dir_context *ctx) { - loff_t pos = filp->f_pos; - struct inode *inode = filp->f_dentry->d_inode; + loff_t pos = ctx->pos; + struct inode *inode = file_inode(file); struct super_block *sb = inode->i_sb; unsigned int offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; unsigned long npages = dir_pages(inode); /* unsigned chunk_mask = ~(nilfs_chunk_size(inode)-1); */ - unsigned char *types = NULL; - int ret; if (pos > inode->i_size - NILFS_DIR_REC_LEN(1)) - goto success; - - types = nilfs_filetype_table; + return 0; for ( ; n < npages; n++, offset = 0) { char *kaddr, *limit; @@ -281,9 +277,8 @@ static int nilfs_readdir(struct file *filp, void *dirent, filldir_t filldir) if (IS_ERR(page)) { nilfs_error(sb, __func__, "bad page in #%lu", inode->i_ino); - filp->f_pos += PAGE_CACHE_SIZE - offset; - ret = -EIO; - goto done; + ctx->pos += PAGE_CACHE_SIZE - offset; + return -EIO; } kaddr = page_address(page); de = (struct nilfs_dir_entry *)(kaddr + offset); @@ -293,35 +288,28 @@ static int nilfs_readdir(struct file *filp, void *dirent, filldir_t filldir) if (de->rec_len == 0) { nilfs_error(sb, __func__, "zero-length directory entry"); - ret = -EIO; nilfs_put_page(page); - goto done; + return -EIO; } if (de->inode) { - int over; - unsigned char d_type = DT_UNKNOWN; + unsigned char t; - if (types && de->file_type < NILFS_FT_MAX) - d_type = types[de->file_type]; + if (de->file_type < NILFS_FT_MAX) + t = nilfs_filetype_table[de->file_type]; + else + t = DT_UNKNOWN; - offset = (char *)de - kaddr; - over = filldir(dirent, de->name, de->name_len, - (n<<PAGE_CACHE_SHIFT) | offset, - le64_to_cpu(de->inode), d_type); - if (over) { + if (!dir_emit(ctx, de->name, de->name_len, + le64_to_cpu(de->inode), t)) { nilfs_put_page(page); - goto success; + return 0; } } - filp->f_pos += nilfs_rec_len_from_disk(de->rec_len); + ctx->pos += nilfs_rec_len_from_disk(de->rec_len); } nilfs_put_page(page); } - -success: - ret = 0; -done: - return ret; + return 0; } /* @@ -602,7 +590,7 @@ int nilfs_make_empty(struct inode *inode, struct inode *parent) unlock_page(page); goto fail; } - kaddr = kmap_atomic(page, KM_USER0); + kaddr = kmap_atomic(page); memset(kaddr, 0, chunk_size); de = (struct nilfs_dir_entry *)kaddr; de->name_len = 1; @@ -617,7 +605,7 @@ int nilfs_make_empty(struct inode *inode, struct inode *parent) de->inode = cpu_to_le64(parent->i_ino); memcpy(de->name, "..\0", 4); nilfs_set_de_type(de, inode); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nilfs_commit_chunk(page, mapping, 0, chunk_size); fail: page_cache_release(page); @@ -678,7 +666,7 @@ not_empty: const struct file_operations nilfs_dir_operations = { .llseek = generic_file_llseek, .read = generic_read_dir, - .readdir = nilfs_readdir, + .iterate = nilfs_readdir, .unlocked_ioctl = nilfs_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = nilfs_compat_ioctl, diff --git a/fs/nilfs2/export.h b/fs/nilfs2/export.h index a71cc412b65..19ccbf9522a 100644 --- a/fs/nilfs2/export.h +++ b/fs/nilfs2/export.h @@ -5,6 +5,14 @@ extern const struct export_operations nilfs_export_ops; +/** + * struct nilfs_fid - NILFS file id type + * @cno: checkpoint number + * @ino: inode number + * @gen: file generation (version) for NFS + * @parent_gen: parent generation (version) for NFS + * @parent_ino: parent inode number + */ struct nilfs_fid { u64 cno; u64 ino; diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c index 397e7325863..24978153c0c 100644 --- a/fs/nilfs2/file.c +++ b/fs/nilfs2/file.c @@ -27,7 +27,7 @@ #include "nilfs.h" #include "segment.h" -int nilfs_sync_file(struct file *file, int datasync) +int nilfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) { /* * Called from fsync() system call @@ -37,36 +37,50 @@ int nilfs_sync_file(struct file *file, int datasync) * This function should be implemented when the writeback function * will be implemented. */ + struct the_nilfs *nilfs; struct inode *inode = file->f_mapping->host; int err; - if (!nilfs_inode_dirty(inode)) - return 0; - - if (datasync) - err = nilfs_construct_dsync_segment(inode->i_sb, inode, 0, - LLONG_MAX); - else - err = nilfs_construct_segment(inode->i_sb); + err = filemap_write_and_wait_range(inode->i_mapping, start, end); + if (err) + return err; + mutex_lock(&inode->i_mutex); + + if (nilfs_inode_dirty(inode)) { + if (datasync) + err = nilfs_construct_dsync_segment(inode->i_sb, inode, + 0, LLONG_MAX); + else + err = nilfs_construct_segment(inode->i_sb); + } + mutex_unlock(&inode->i_mutex); + nilfs = inode->i_sb->s_fs_info; + if (!err && nilfs_test_opt(nilfs, BARRIER)) { + err = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); + if (err != -EIO) + err = 0; + } return err; } static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct nilfs_transaction_info ti; - int ret; + int ret = 0; if (unlikely(nilfs_near_disk_full(inode->i_sb->s_fs_info))) return VM_FAULT_SIGBUS; /* -ENOSPC */ + sb_start_pagefault(inode->i_sb); lock_page(page); if (page->mapping != inode->i_mapping || page_offset(page) >= i_size_read(inode) || !PageUptodate(page)) { unlock_page(page); - return VM_FAULT_NOPAGE; /* make the VM retry the fault */ + ret = -EFAULT; /* make the VM retry the fault */ + goto out; } /* @@ -100,32 +114,35 @@ static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) ret = nilfs_transaction_begin(inode->i_sb, &ti, 1); /* never returns -ENOMEM, but may return -ENOSPC */ if (unlikely(ret)) - return VM_FAULT_SIGBUS; + goto out; - ret = block_page_mkwrite(vma, vmf, nilfs_get_block); - if (ret != VM_FAULT_LOCKED) { + file_update_time(vma->vm_file); + ret = __block_page_mkwrite(vma, vmf, nilfs_get_block); + if (ret) { nilfs_transaction_abort(inode->i_sb); - return ret; + goto out; } nilfs_set_file_dirty(inode, 1 << (PAGE_SHIFT - inode->i_blkbits)); nilfs_transaction_commit(inode->i_sb); mapped: - SetPageChecked(page); - wait_on_page_writeback(page); - return VM_FAULT_LOCKED; + wait_for_stable_page(page); + out: + sb_end_pagefault(inode->i_sb); + return block_page_mkwrite_return(ret); } static const struct vm_operations_struct nilfs_file_vm_ops = { .fault = filemap_fault, + .map_pages = filemap_map_pages, .page_mkwrite = nilfs_page_mkwrite, + .remap_pages = generic_file_remap_pages, }; static int nilfs_file_mmap(struct file *file, struct vm_area_struct *vma) { file_accessed(file); vma->vm_ops = &nilfs_file_vm_ops; - vma->vm_flags |= VM_CAN_NONLINEAR; return 0; } @@ -135,10 +152,10 @@ static int nilfs_file_mmap(struct file *file, struct vm_area_struct *vma) */ const struct file_operations nilfs_file_operations = { .llseek = generic_file_llseek, - .read = do_sync_read, - .write = do_sync_write, - .aio_read = generic_file_aio_read, - .aio_write = generic_file_aio_write, + .read = new_sync_read, + .write = new_sync_write, + .read_iter = generic_file_read_iter, + .write_iter = generic_file_write_iter, .unlocked_ioctl = nilfs_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = nilfs_compat_ioctl, @@ -151,7 +168,6 @@ const struct file_operations nilfs_file_operations = { }; const struct inode_operations nilfs_file_inode_operations = { - .truncate = nilfs_truncate, .setattr = nilfs_setattr, .permission = nilfs_permission, .fiemap = nilfs_fiemap, diff --git a/fs/nilfs2/gcinode.c b/fs/nilfs2/gcinode.c index 1c2a3e23f8b..57ceaf33d17 100644 --- a/fs/nilfs2/gcinode.c +++ b/fs/nilfs2/gcinode.c @@ -48,9 +48,6 @@ #include "dat.h" #include "ifile.h" -static const struct address_space_operations def_gcinode_aops = { -}; - /* * nilfs_gccache_submit_read_data() - add data buffer and submit read request * @inode - gc inode @@ -87,9 +84,9 @@ int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff, goto out; if (pbn == 0) { - struct inode *dat_inode = NILFS_I_NILFS(inode)->ns_dat; - /* use original dat, not gc dat. */ - err = nilfs_dat_translate(dat_inode, vbn, &pbn); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + + err = nilfs_dat_translate(nilfs->ns_dat, vbn, &pbn); if (unlikely(err)) { /* -EIO, -ENOMEM, -ENOENT */ brelse(bh); goto failed; @@ -103,7 +100,7 @@ int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff, } if (!buffer_mapped(bh)) { - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_bdev = inode->i_sb->s_bdev; set_buffer_mapped(bh); } bh->b_blocknr = pbn; @@ -160,15 +157,11 @@ int nilfs_gccache_wait_and_mark_dirty(struct buffer_head *bh) if (buffer_dirty(bh)) return -EEXIST; - if (buffer_nilfs_node(bh)) { - if (nilfs_btree_broken_node_block(bh)) { - clear_buffer_uptodate(bh); - return -EIO; - } - nilfs_btnode_mark_dirty(bh); - } else { - nilfs_mark_buffer_dirty(bh); + if (buffer_nilfs_node(bh) && nilfs_btree_broken_node_block(bh)) { + clear_buffer_uptodate(bh); + return -EIO; } + mark_buffer_dirty(bh); return 0; } @@ -178,7 +171,7 @@ int nilfs_init_gcinode(struct inode *inode) inode->i_mode = S_IFREG; mapping_set_gfp_mask(inode->i_mapping, GFP_NOFS); - inode->i_mapping->a_ops = &def_gcinode_aops; + inode->i_mapping->a_ops = &empty_aops; inode->i_mapping->backing_dev_info = inode->i_sb->s_bdi; ii->i_flags = 0; @@ -198,6 +191,8 @@ void nilfs_remove_all_gcinodes(struct the_nilfs *nilfs) while (!list_empty(head)) { ii = list_first_entry(head, struct nilfs_inode_info, i_dirty); list_del_init(&ii->i_dirty); + truncate_inode_pages(&ii->vfs_inode.i_data, 0); + nilfs_btnode_cache_clear(&ii->i_btnode_cache); iput(&ii->vfs_inode); } } diff --git a/fs/nilfs2/ifile.c b/fs/nilfs2/ifile.c index bfc73d3a30e..6548c7851b4 100644 --- a/fs/nilfs2/ifile.c +++ b/fs/nilfs2/ifile.c @@ -29,7 +29,11 @@ #include "alloc.h" #include "ifile.h" - +/** + * struct nilfs_ifile_info - on-memory private data of ifile + * @mi: on-memory private data of metadata file + * @palloc_cache: persistent object allocator cache of ifile + */ struct nilfs_ifile_info { struct nilfs_mdt_info mi; struct nilfs_palloc_cache palloc_cache; @@ -80,7 +84,7 @@ int nilfs_ifile_create_inode(struct inode *ifile, ino_t *out_ino, return ret; } nilfs_palloc_commit_alloc_entry(ifile, &req); - nilfs_mdt_mark_buffer_dirty(req.pr_entry_bh); + mark_buffer_dirty(req.pr_entry_bh); nilfs_mdt_mark_dirty(ifile); *out_ino = (ino_t)req.pr_entry_nr; *out_bh = req.pr_entry_bh; @@ -122,13 +126,13 @@ int nilfs_ifile_delete_inode(struct inode *ifile, ino_t ino) return ret; } - kaddr = kmap_atomic(req.pr_entry_bh->b_page, KM_USER0); + kaddr = kmap_atomic(req.pr_entry_bh->b_page); raw_inode = nilfs_palloc_block_get_entry(ifile, req.pr_entry_nr, req.pr_entry_bh, kaddr); raw_inode->i_flags = 0; - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - nilfs_mdt_mark_buffer_dirty(req.pr_entry_bh); + mark_buffer_dirty(req.pr_entry_bh); brelse(req.pr_entry_bh); nilfs_palloc_commit_free_entry(ifile, &req); @@ -156,6 +160,28 @@ int nilfs_ifile_get_inode_block(struct inode *ifile, ino_t ino, } /** + * nilfs_ifile_count_free_inodes - calculate free inodes count + * @ifile: ifile inode + * @nmaxinodes: current maximum of available inodes count [out] + * @nfreeinodes: free inodes count [out] + */ +int nilfs_ifile_count_free_inodes(struct inode *ifile, + u64 *nmaxinodes, u64 *nfreeinodes) +{ + u64 nused; + int err; + + *nmaxinodes = 0; + *nfreeinodes = 0; + + nused = atomic64_read(&NILFS_I(ifile)->i_root->inodes_count); + err = nilfs_palloc_count_max_entries(ifile, nused, nmaxinodes); + if (likely(!err)) + *nfreeinodes = *nmaxinodes - nused; + return err; +} + +/** * nilfs_ifile_read - read or get ifile inode * @sb: super block instance * @root: root object diff --git a/fs/nilfs2/ifile.h b/fs/nilfs2/ifile.h index 59b6f2b51df..679674d1337 100644 --- a/fs/nilfs2/ifile.h +++ b/fs/nilfs2/ifile.h @@ -49,6 +49,8 @@ int nilfs_ifile_create_inode(struct inode *, ino_t *, struct buffer_head **); int nilfs_ifile_delete_inode(struct inode *, ino_t); int nilfs_ifile_get_inode_block(struct inode *, ino_t, struct buffer_head **); +int nilfs_ifile_count_free_inodes(struct inode *, u64 *, u64 *); + int nilfs_ifile_read(struct super_block *sb, struct nilfs_root *root, size_t inode_size, struct nilfs_inode *raw_inode, struct inode **inodep); diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index c0aa27490c0..6252b173a46 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -25,7 +25,7 @@ #include <linux/gfp.h> #include <linux/mpage.h> #include <linux/writeback.h> -#include <linux/uio.h> +#include <linux/aio.h> #include "nilfs.h" #include "btnode.h" #include "segment.h" @@ -34,6 +34,13 @@ #include "cpfile.h" #include "ifile.h" +/** + * struct nilfs_iget_args - arguments used during comparison between inodes + * @ino: inode number + * @cno: checkpoint number + * @root: pointer on NILFS root object (mounted checkpoint) + * @for_gc: inode for GC flag + */ struct nilfs_iget_args { u64 ino; __u64 cno; @@ -47,7 +54,7 @@ void nilfs_inode_add_blocks(struct inode *inode, int n) inode_add_bytes(inode, (1 << inode->i_blkbits) * n); if (root) - atomic_add(n, &root->blocks_count); + atomic64_add(n, &root->blocks_count); } void nilfs_inode_sub_blocks(struct inode *inode, int n) @@ -56,7 +63,7 @@ void nilfs_inode_sub_blocks(struct inode *inode, int n) inode_sub_bytes(inode, (1 << inode->i_blkbits) * n); if (root) - atomic_sub(n, &root->blocks_count); + atomic64_sub(n, &root->blocks_count); } /** @@ -74,14 +81,14 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff, struct buffer_head *bh_result, int create) { struct nilfs_inode_info *ii = NILFS_I(inode); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; __u64 blknum = 0; int err = 0, ret; - struct inode *dat = NILFS_I_NILFS(inode)->ns_dat; unsigned maxblocks = bh_result->b_size >> inode->i_blkbits; - down_read(&NILFS_MDT(dat)->mi_sem); + down_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem); ret = nilfs_bmap_lookup_contig(ii->i_bmap, blkoff, &blknum, maxblocks); - up_read(&NILFS_MDT(dat)->mi_sem); + up_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem); if (ret >= 0) { /* found */ map_bh(bh_result, inode->i_sb, blknum); if (ret > 0) @@ -168,6 +175,11 @@ static int nilfs_writepages(struct address_space *mapping, struct inode *inode = mapping->host; int err = 0; + if (inode->i_sb->s_flags & MS_RDONLY) { + nilfs_clear_dirty_pages(mapping, false); + return -EROFS; + } + if (wbc->sync_mode == WB_SYNC_ALL) err = nilfs_construct_dsync_segment(inode->i_sb, inode, wbc->range_start, @@ -180,6 +192,18 @@ static int nilfs_writepage(struct page *page, struct writeback_control *wbc) struct inode *inode = page->mapping->host; int err; + if (inode->i_sb->s_flags & MS_RDONLY) { + /* + * It means that filesystem was remounted in read-only + * mode because of error or metadata corruption. But we + * have dirty pages that try to be flushed in background. + * So, here we simply discard this dirty page. + */ + nilfs_clear_dirty_page(page, false); + unlock_page(page); + return -EROFS; + } + redirty_page_for_writepage(wbc, page); unlock_page(page); @@ -195,17 +219,46 @@ static int nilfs_writepage(struct page *page, struct writeback_control *wbc) static int nilfs_set_page_dirty(struct page *page) { - int ret = __set_page_dirty_buffers(page); + int ret = __set_page_dirty_nobuffers(page); - if (ret) { + if (page_has_buffers(page)) { struct inode *inode = page->mapping->host; - unsigned nr_dirty = 1 << (PAGE_SHIFT - inode->i_blkbits); + unsigned nr_dirty = 0; + struct buffer_head *bh, *head; - nilfs_set_file_dirty(inode, nr_dirty); + /* + * This page is locked by callers, and no other thread + * concurrently marks its buffers dirty since they are + * only dirtied through routines in fs/buffer.c in + * which call sites of mark_buffer_dirty are protected + * by page lock. + */ + bh = head = page_buffers(page); + do { + /* Do not mark hole blocks dirty */ + if (buffer_dirty(bh) || !buffer_mapped(bh)) + continue; + + set_buffer_dirty(bh); + nr_dirty++; + } while (bh = bh->b_this_page, bh != head); + + if (nr_dirty) + nilfs_set_file_dirty(inode, nr_dirty); } return ret; } +void nilfs_write_failed(struct address_space *mapping, loff_t to) +{ + struct inode *inode = mapping->host; + + if (to > inode->i_size) { + truncate_pagecache(inode, inode->i_size); + nilfs_truncate(inode); + } +} + static int nilfs_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) @@ -220,10 +273,7 @@ static int nilfs_write_begin(struct file *file, struct address_space *mapping, err = block_write_begin(mapping, pos, len, flags, pagep, nilfs_get_block); if (unlikely(err)) { - loff_t isize = mapping->host->i_size; - if (pos + len > isize) - vmtruncate(mapping->host, isize); - + nilfs_write_failed(mapping, pos + len); nilfs_transaction_abort(inode->i_sb); } return err; @@ -248,19 +298,21 @@ static int nilfs_write_end(struct file *file, struct address_space *mapping, } static ssize_t -nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t offset, unsigned long nr_segs) +nilfs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; + struct address_space *mapping = file->f_mapping; struct inode *inode = file->f_mapping->host; + size_t count = iov_iter_count(iter); ssize_t size; if (rw == WRITE) return 0; /* Needs synchronization with the cleaner */ - size = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, nilfs_get_block, NULL); + size = blockdev_direct_IO(rw, iocb, inode, iter, offset, + nilfs_get_block); /* * In case of error extending write may have instantiated a few @@ -268,10 +320,10 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, */ if (unlikely((rw & WRITE) && size < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + count; if (end > isize) - vmtruncate(inode, isize); + nilfs_write_failed(mapping, end); } return size; @@ -291,7 +343,7 @@ const struct address_space_operations nilfs_aops = { .is_partially_uptodate = block_is_partially_uptodate, }; -struct inode *nilfs_new_inode(struct inode *dir, int mode) +struct inode *nilfs_new_inode(struct inode *dir, umode_t mode) { struct super_block *sb = dir->i_sb; struct the_nilfs *nilfs = sb->s_fs_info; @@ -318,7 +370,7 @@ struct inode *nilfs_new_inode(struct inode *dir, int mode) goto failed_ifile_create_inode; /* reference count of i_bh inherits from nilfs_mdt_read_block() */ - atomic_inc(&root->inodes_count); + atomic64_inc(&root->inodes_count); inode_init_owner(inode, dir, mode); inode->i_ino = ino; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; @@ -354,7 +406,7 @@ struct inode *nilfs_new_inode(struct inode *dir, int mode) failed_acl: failed_bmap: - inode->i_nlink = 0; + clear_nlink(inode); iput(inode); /* raw_inode will be deleted through generic_delete_inode() */ goto failed; @@ -394,9 +446,9 @@ int nilfs_read_inode_common(struct inode *inode, int err; inode->i_mode = le16_to_cpu(raw_inode->i_mode); - inode->i_uid = (uid_t)le32_to_cpu(raw_inode->i_uid); - inode->i_gid = (gid_t)le32_to_cpu(raw_inode->i_gid); - inode->i_nlink = le16_to_cpu(raw_inode->i_links_count); + i_uid_write(inode, le32_to_cpu(raw_inode->i_uid)); + i_gid_write(inode, le32_to_cpu(raw_inode->i_gid)); + set_nlink(inode, le16_to_cpu(raw_inode->i_links_count)); inode->i_size = le64_to_cpu(raw_inode->i_size); inode->i_atime.tv_sec = le64_to_cpu(raw_inode->i_mtime); inode->i_ctime.tv_sec = le64_to_cpu(raw_inode->i_ctime); @@ -583,8 +635,8 @@ void nilfs_write_inode_common(struct inode *inode, struct nilfs_inode_info *ii = NILFS_I(inode); raw_inode->i_mode = cpu_to_le16(inode->i_mode); - raw_inode->i_uid = cpu_to_le32(inode->i_uid); - raw_inode->i_gid = cpu_to_le32(inode->i_gid); + raw_inode->i_uid = cpu_to_le32(i_uid_read(inode)); + raw_inode->i_gid = cpu_to_le32(i_gid_read(inode)); raw_inode->i_links_count = cpu_to_le16(inode->i_nlink); raw_inode->i_size = cpu_to_le64(inode->i_size); raw_inode->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec); @@ -596,6 +648,16 @@ void nilfs_write_inode_common(struct inode *inode, raw_inode->i_flags = cpu_to_le32(ii->i_flags); raw_inode->i_generation = cpu_to_le32(inode->i_generation); + if (NILFS_ROOT_METADATA_FILE(inode->i_ino)) { + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + + /* zero-fill unused portion in the case of super root block */ + raw_inode->i_xattr = 0; + raw_inode->i_pad = 0; + memset((void *)raw_inode + sizeof(*raw_inode), 0, + nilfs->ns_inode_size - sizeof(*raw_inode)); + } + if (has_bmap) nilfs_bmap_write(ii->i_bmap, raw_inode); else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) @@ -722,25 +784,23 @@ void nilfs_evict_inode(struct inode *inode) int ret; if (inode->i_nlink || !ii->i_root || unlikely(is_bad_inode(inode))) { - if (inode->i_data.nrpages) - truncate_inode_pages(&inode->i_data, 0); - end_writeback(inode); + truncate_inode_pages_final(&inode->i_data); + clear_inode(inode); nilfs_clear_inode(inode); return; } nilfs_transaction_begin(sb, &ti, 0); /* never fails */ - if (inode->i_data.nrpages) - truncate_inode_pages(&inode->i_data, 0); + truncate_inode_pages_final(&inode->i_data); /* TODO: some of the following operations may fail. */ nilfs_truncate_bmap(ii, 0); nilfs_mark_inode_dirty(inode); - end_writeback(inode); + clear_inode(inode); ret = nilfs_ifile_delete_inode(ii->i_root->ifile, inode->i_ino); if (!ret) - atomic_dec(&ii->i_root->inodes_count); + atomic64_dec(&ii->i_root->inodes_count); nilfs_clear_inode(inode); @@ -768,9 +828,9 @@ int nilfs_setattr(struct dentry *dentry, struct iattr *iattr) if ((iattr->ia_valid & ATTR_SIZE) && iattr->ia_size != i_size_read(inode)) { - err = vmtruncate(inode, iattr->ia_size); - if (unlikely(err)) - goto out_err; + inode_dio_wait(inode); + truncate_setsize(inode, iattr->ia_size); + nilfs_truncate(inode); } setattr_copy(inode, iattr); @@ -789,19 +849,14 @@ out_err: return err; } -int nilfs_permission(struct inode *inode, int mask, unsigned int flags) +int nilfs_permission(struct inode *inode, int mask) { - struct nilfs_root *root; - - if (flags & IPERM_FLAG_RCU) - return -ECHILD; - - root = NILFS_I(inode)->i_root; + struct nilfs_root *root = NILFS_I(inode)->i_root; if ((mask & MAY_WRITE) && root && root->cno != NILFS_CPTREE_CURRENT_CNO) return -EROFS; /* snapshot is not writable */ - return generic_permission(inode, mask, flags, NULL); + return generic_permission(inode, mask); } int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh) @@ -872,8 +927,7 @@ int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty) return -EINVAL; /* NILFS_I_DIRTY may remain for freeing inode */ } - list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &nilfs->ns_dirty_files); + list_move_tail(&ii->i_dirty, &nilfs->ns_dirty_files); set_bit(NILFS_I_QUEUED, &ii->i_state); } spin_unlock(&nilfs->ns_inode_lock); @@ -892,7 +946,7 @@ int nilfs_mark_inode_dirty(struct inode *inode) return err; } nilfs_update_inode(inode, ibh); - nilfs_mdt_mark_buffer_dirty(ibh); + mark_buffer_dirty(ibh); nilfs_mdt_mark_dirty(NILFS_I(inode)->i_root->ifile); brelse(ibh); return 0; @@ -908,7 +962,7 @@ int nilfs_mark_inode_dirty(struct inode *inode) * construction. This function can be called both as a single operation * and as a part of indivisible file operations. */ -void nilfs_dirty_inode(struct inode *inode) +void nilfs_dirty_inode(struct inode *inode, int flags) { struct nilfs_transaction_info ti; struct nilfs_mdt_info *mdi = NILFS_MDT(inode); @@ -931,7 +985,7 @@ void nilfs_dirty_inode(struct inode *inode) int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len) { - struct the_nilfs *nilfs = NILFS_I_NILFS(inode); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; __u64 logical = 0, phys = 0, size = 0; __u32 flags = 0; loff_t isize; diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index f2469ba6246..422fb54b737 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -27,7 +27,7 @@ #include <linux/uaccess.h> /* copy_from_user(), copy_to_user() */ #include <linux/vmalloc.h> #include <linux/compat.h> /* compat_ptr() */ -#include <linux/mount.h> /* mnt_want_write(), mnt_drop_write() */ +#include <linux/mount.h> /* mnt_want_write_file(), mnt_drop_write_file() */ #include <linux/buffer_head.h> #include <linux/nilfs2_fs.h> #include "nilfs.h" @@ -37,7 +37,26 @@ #include "sufile.h" #include "dat.h" - +/** + * nilfs_ioctl_wrap_copy - wrapping function of get/set metadata info + * @nilfs: nilfs object + * @argv: vector of arguments from userspace + * @dir: set of direction flags + * @dofunc: concrete function of get/set metadata info + * + * Description: nilfs_ioctl_wrap_copy() gets/sets metadata info by means of + * calling dofunc() function on the basis of @argv argument. + * + * Return Value: On success, 0 is returned and requested metadata info + * is copied into userspace. On error, one of the following + * negative error codes is returned. + * + * %-EINVAL - Invalid arguments from userspace. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EFAULT - Failure during execution of requested operation. + */ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs, struct nilfs_argv *argv, int dir, ssize_t (*dofunc)(struct the_nilfs *, @@ -57,6 +76,14 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs, if (argv->v_size > PAGE_SIZE) return -EINVAL; + /* + * Reject pairs of a start item position (argv->v_index) and a + * total count (argv->v_nmembs) which leads position 'pos' to + * overflow by the increment at the end of the loop. + */ + if (argv->v_index > ~(__u64)0 - argv->v_nmembs) + return -EINVAL; + buf = (void *)__get_free_pages(GFP_NOFS, 0); if (unlikely(!buf)) return -ENOMEM; @@ -99,6 +126,9 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs, return ret; } +/** + * nilfs_ioctl_getflags - ioctl to support lsattr + */ static int nilfs_ioctl_getflags(struct inode *inode, void __user *argp) { unsigned int flags = NILFS_I(inode)->i_flags & FS_FL_USER_VISIBLE; @@ -106,6 +136,9 @@ static int nilfs_ioctl_getflags(struct inode *inode, void __user *argp) return put_user(flags, (int __user *)argp); } +/** + * nilfs_ioctl_setflags - ioctl to support chattr + */ static int nilfs_ioctl_setflags(struct inode *inode, struct file *filp, void __user *argp) { @@ -119,7 +152,7 @@ static int nilfs_ioctl_setflags(struct inode *inode, struct file *filp, if (get_user(flags, (int __user *)argp)) return -EFAULT; - ret = mnt_want_write(filp->f_path.mnt); + ret = mnt_want_write_file(filp); if (ret) return ret; @@ -154,15 +187,37 @@ static int nilfs_ioctl_setflags(struct inode *inode, struct file *filp, ret = nilfs_transaction_commit(inode->i_sb); out: mutex_unlock(&inode->i_mutex); - mnt_drop_write(filp->f_path.mnt); + mnt_drop_write_file(filp); return ret; } +/** + * nilfs_ioctl_getversion - get info about a file's version (generation number) + */ static int nilfs_ioctl_getversion(struct inode *inode, void __user *argp) { return put_user(inode->i_generation, (int __user *)argp); } +/** + * nilfs_ioctl_change_cpmode - change checkpoint mode (checkpoint/snapshot) + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_change_cpmode() function changes mode of + * given checkpoint between checkpoint and snapshot state. This ioctl + * is used in chcp and mkcp utilities. + * + * Return Value: On success, 0 is returned and mode of a checkpoint is + * changed. On error, one of the following negative error codes + * is returned. + * + * %-EPERM - Operation not permitted. + * + * %-EFAULT - Failure during checkpoint mode changing. + */ static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { @@ -174,7 +229,7 @@ static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp, if (!capable(CAP_SYS_ADMIN)) return -EPERM; - ret = mnt_want_write(filp->f_path.mnt); + ret = mnt_want_write_file(filp); if (ret) return ret; @@ -182,7 +237,7 @@ static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp, if (copy_from_user(&cpmode, argp, sizeof(cpmode))) goto out; - down_read(&inode->i_sb->s_umount); + mutex_lock(&nilfs->ns_snapshot_mount_mutex); nilfs_transaction_begin(inode->i_sb, &ti, 0); ret = nilfs_cpfile_change_cpmode( @@ -192,12 +247,31 @@ static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp, else nilfs_transaction_commit(inode->i_sb); /* never fails */ - up_read(&inode->i_sb->s_umount); + mutex_unlock(&nilfs->ns_snapshot_mount_mutex); out: - mnt_drop_write(filp->f_path.mnt); + mnt_drop_write_file(filp); return ret; } +/** + * nilfs_ioctl_delete_checkpoint - remove checkpoint + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_delete_checkpoint() function removes + * checkpoint from NILFS2 file system. This ioctl is used in rmcp + * utility. + * + * Return Value: On success, 0 is returned and a checkpoint is + * removed. On error, one of the following negative error codes + * is returned. + * + * %-EPERM - Operation not permitted. + * + * %-EFAULT - Failure during checkpoint removing. + */ static int nilfs_ioctl_delete_checkpoint(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) @@ -210,7 +284,7 @@ nilfs_ioctl_delete_checkpoint(struct inode *inode, struct file *filp, if (!capable(CAP_SYS_ADMIN)) return -EPERM; - ret = mnt_want_write(filp->f_path.mnt); + ret = mnt_want_write_file(filp); if (ret) return ret; @@ -225,10 +299,25 @@ nilfs_ioctl_delete_checkpoint(struct inode *inode, struct file *filp, else nilfs_transaction_commit(inode->i_sb); /* never fails */ out: - mnt_drop_write(filp->f_path.mnt); + mnt_drop_write_file(filp); return ret; } +/** + * nilfs_ioctl_do_get_cpinfo - callback method getting info about checkpoints + * @nilfs: nilfs object + * @posp: pointer on array of checkpoint's numbers + * @flags: checkpoint mode (checkpoint or snapshot) + * @buf: buffer for storing checkponts' info + * @size: size in bytes of one checkpoint info item in array + * @nmembs: number of checkpoints in array (numbers and infos) + * + * Description: nilfs_ioctl_do_get_cpinfo() function returns info about + * requested checkpoints. The NILFS_IOCTL_GET_CPINFO ioctl is used in + * lscp utility and by nilfs_cleanerd daemon. + * + * Return value: count of nilfs_cpinfo structures in output buffer. + */ static ssize_t nilfs_ioctl_do_get_cpinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, void *buf, size_t size, size_t nmembs) @@ -242,6 +331,27 @@ nilfs_ioctl_do_get_cpinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, return ret; } +/** + * nilfs_ioctl_get_cpstat - get checkpoints statistics + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_get_cpstat() returns information about checkpoints. + * The NILFS_IOCTL_GET_CPSTAT ioctl is used by lscp, rmcp utilities + * and by nilfs_cleanerd daemon. + * + * Return Value: On success, 0 is returned, and checkpoints information is + * copied into userspace pointer @argp. On error, one of the following + * negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EFAULT - Failure during getting checkpoints statistics. + */ static int nilfs_ioctl_get_cpstat(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { @@ -260,6 +370,21 @@ static int nilfs_ioctl_get_cpstat(struct inode *inode, struct file *filp, return ret; } +/** + * nilfs_ioctl_do_get_suinfo - callback method getting segment usage info + * @nilfs: nilfs object + * @posp: pointer on array of segment numbers + * @flags: *not used* + * @buf: buffer for storing suinfo array + * @size: size in bytes of one suinfo item in array + * @nmembs: count of segment numbers and suinfos in array + * + * Description: nilfs_ioctl_do_get_suinfo() function returns segment usage + * info about requested segments. The NILFS_IOCTL_GET_SUINFO ioctl is used + * in lssu, nilfs_resize utilities and by nilfs_cleanerd daemon. + * + * Return value: count of nilfs_suinfo structures in output buffer. + */ static ssize_t nilfs_ioctl_do_get_suinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, void *buf, size_t size, size_t nmembs) @@ -273,6 +398,27 @@ nilfs_ioctl_do_get_suinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, return ret; } +/** + * nilfs_ioctl_get_sustat - get segment usage statistics + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_get_sustat() returns segment usage statistics. + * The NILFS_IOCTL_GET_SUSTAT ioctl is used in lssu, nilfs_resize utilities + * and by nilfs_cleanerd daemon. + * + * Return Value: On success, 0 is returned, and segment usage information is + * copied into userspace pointer @argp. On error, one of the following + * negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EFAULT - Failure during getting segment usage statistics. + */ static int nilfs_ioctl_get_sustat(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { @@ -291,6 +437,21 @@ static int nilfs_ioctl_get_sustat(struct inode *inode, struct file *filp, return ret; } +/** + * nilfs_ioctl_do_get_vinfo - callback method getting virtual blocks info + * @nilfs: nilfs object + * @posp: *not used* + * @flags: *not used* + * @buf: buffer for storing array of nilfs_vinfo structures + * @size: size in bytes of one vinfo item in array + * @nmembs: count of vinfos in array + * + * Description: nilfs_ioctl_do_get_vinfo() function returns information + * on virtual block addresses. The NILFS_IOCTL_GET_VINFO ioctl is used + * by nilfs_cleanerd daemon. + * + * Return value: count of nilfs_vinfo structures in output buffer. + */ static ssize_t nilfs_ioctl_do_get_vinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, void *buf, size_t size, size_t nmembs) @@ -303,6 +464,21 @@ nilfs_ioctl_do_get_vinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, return ret; } +/** + * nilfs_ioctl_do_get_bdescs - callback method getting disk block descriptors + * @nilfs: nilfs object + * @posp: *not used* + * @flags: *not used* + * @buf: buffer for storing array of nilfs_bdesc structures + * @size: size in bytes of one bdesc item in array + * @nmembs: count of bdescs in array + * + * Description: nilfs_ioctl_do_get_bdescs() function returns information + * about descriptors of disk block numbers. The NILFS_IOCTL_GET_BDESCS ioctl + * is used by nilfs_cleanerd daemon. + * + * Return value: count of nilfs_bdescs structures in output buffer. + */ static ssize_t nilfs_ioctl_do_get_bdescs(struct the_nilfs *nilfs, __u64 *posp, int flags, void *buf, size_t size, size_t nmembs) @@ -329,6 +505,29 @@ nilfs_ioctl_do_get_bdescs(struct the_nilfs *nilfs, __u64 *posp, int flags, return nmembs; } +/** + * nilfs_ioctl_get_bdescs - get disk block descriptors + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_do_get_bdescs() function returns information + * about descriptors of disk block numbers. The NILFS_IOCTL_GET_BDESCS ioctl + * is used by nilfs_cleanerd daemon. + * + * Return Value: On success, 0 is returned, and disk block descriptors are + * copied into userspace pointer @argp. On error, one of the following + * negative error codes is returned. + * + * %-EINVAL - Invalid arguments from userspace. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EFAULT - Failure during getting disk block descriptors. + */ static int nilfs_ioctl_get_bdescs(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { @@ -352,6 +551,26 @@ static int nilfs_ioctl_get_bdescs(struct inode *inode, struct file *filp, return ret; } +/** + * nilfs_ioctl_move_inode_block - prepare data/node block for moving by GC + * @inode: inode object + * @vdesc: descriptor of virtual block number + * @buffers: list of moving buffers + * + * Description: nilfs_ioctl_move_inode_block() function registers data/node + * buffer in the GC pagecache and submit read request. + * + * Return Value: On success, 0 is returned. On error, one of the following + * negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-ENOENT - Requested block doesn't exist. + * + * %-EEXIST - Blocks conflict is detected. + */ static int nilfs_ioctl_move_inode_block(struct inode *inode, struct nilfs_vdesc *vdesc, struct list_head *buffers) @@ -397,6 +616,19 @@ static int nilfs_ioctl_move_inode_block(struct inode *inode, return 0; } +/** + * nilfs_ioctl_move_blocks - move valid inode's blocks during garbage collection + * @sb: superblock object + * @argv: vector of arguments from userspace + * @buf: array of nilfs_vdesc structures + * + * Description: nilfs_ioctl_move_blocks() function reads valid data/node + * blocks that garbage collector specified with the array of nilfs_vdesc + * structures and stores them into page caches of GC inodes. + * + * Return Value: Number of processed nilfs_vdesc structures or + * error code, otherwise. + */ static int nilfs_ioctl_move_blocks(struct super_block *sb, struct nilfs_argv *argv, void *buf) { @@ -462,6 +694,25 @@ static int nilfs_ioctl_move_blocks(struct super_block *sb, return ret; } +/** + * nilfs_ioctl_delete_checkpoints - delete checkpoints + * @nilfs: nilfs object + * @argv: vector of arguments from userspace + * @buf: array of periods of checkpoints numbers + * + * Description: nilfs_ioctl_delete_checkpoints() function deletes checkpoints + * in the period from p_start to p_end, excluding p_end itself. The checkpoints + * which have been already deleted are ignored. + * + * Return Value: Number of processed nilfs_period structures or + * error code, otherwise. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EINVAL - invalid checkpoints. + */ static int nilfs_ioctl_delete_checkpoints(struct the_nilfs *nilfs, struct nilfs_argv *argv, void *buf) { @@ -479,6 +730,24 @@ static int nilfs_ioctl_delete_checkpoints(struct the_nilfs *nilfs, return nmembs; } +/** + * nilfs_ioctl_free_vblocknrs - free virtual block numbers + * @nilfs: nilfs object + * @argv: vector of arguments from userspace + * @buf: array of virtual block numbers + * + * Description: nilfs_ioctl_free_vblocknrs() function frees + * the virtual block numbers specified by @buf and @argv->v_nmembs. + * + * Return Value: Number of processed virtual block numbers or + * error code, otherwise. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-ENOENT - The virtual block number have not been allocated. + */ static int nilfs_ioctl_free_vblocknrs(struct the_nilfs *nilfs, struct nilfs_argv *argv, void *buf) { @@ -490,6 +759,24 @@ static int nilfs_ioctl_free_vblocknrs(struct the_nilfs *nilfs, return (ret < 0) ? ret : nmembs; } +/** + * nilfs_ioctl_mark_blocks_dirty - mark blocks dirty + * @nilfs: nilfs object + * @argv: vector of arguments from userspace + * @buf: array of block descriptors + * + * Description: nilfs_ioctl_mark_blocks_dirty() function marks + * metadata file or data blocks as dirty. + * + * Return Value: Number of processed block descriptors or + * error code, otherwise. + * + * %-ENOMEM - Insufficient memory available. + * + * %-EIO - I/O error + * + * %-ENOENT - the specified block does not exist (hole block) + */ static int nilfs_ioctl_mark_blocks_dirty(struct the_nilfs *nilfs, struct nilfs_argv *argv, void *buf) { @@ -571,6 +858,20 @@ int nilfs_ioctl_prepare_clean_segments(struct the_nilfs *nilfs, return ret; } +/** + * nilfs_ioctl_clean_segments - clean segments + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_clean_segments() function makes garbage + * collection operation in the environment of requested parameters + * from userspace. The NILFS_IOCTL_CLEAN_SEGMENTS ioctl is used by + * nilfs_cleanerd daemon. + * + * Return Value: On success, 0 is returned or error code, otherwise. + */ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { @@ -591,7 +892,7 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, if (!capable(CAP_SYS_ADMIN)) return -EPERM; - ret = mnt_want_write(filp->f_path.mnt); + ret = mnt_want_write_file(filp); if (ret) return ret; @@ -603,6 +904,8 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, nsegs = argv[4].v_nmembs; if (argv[4].v_size != argsz[4]) goto out; + if (nsegs > UINT_MAX / sizeof(__u64)) + goto out; /* * argv[4] points to segment numbers this ioctl cleans. We @@ -625,6 +928,9 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, if (argv[n].v_nmembs > nsegs * nilfs->ns_blocks_per_segment) goto out_free; + if (argv[n].v_nmembs >= UINT_MAX / argv[n].v_size) + goto out_free; + len = argv[n].v_size * argv[n].v_nmembs; base = (void __user *)(unsigned long)argv[n].v_base; if (len == 0) { @@ -655,14 +961,15 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, goto out_free; } - vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE); - ret = nilfs_ioctl_move_blocks(inode->i_sb, &argv[0], kbufs[0]); if (ret < 0) printk(KERN_ERR "NILFS: GC failed during preparation: " "cannot read source blocks: err=%d\n", ret); - else + else { + if (nilfs_sb_need_update(nilfs)) + set_nilfs_discontinued(nilfs); ret = nilfs_clean_segments(inode->i_sb, argv, kbufs); + } nilfs_remove_all_gcinodes(nilfs); clear_nilfs_gc_running(nilfs); @@ -672,10 +979,37 @@ out_free: vfree(kbufs[n]); kfree(kbufs[4]); out: - mnt_drop_write(filp->f_path.mnt); + mnt_drop_write_file(filp); return ret; } +/** + * nilfs_ioctl_sync - make a checkpoint + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: nilfs_ioctl_sync() function constructs a logical segment + * for checkpointing. This function guarantees that all modified data + * and metadata are written out to the device when it successfully + * returned. + * + * Return Value: On success, 0 is retured. On errors, one of the following + * negative error code is returned. + * + * %-EROFS - Read only filesystem. + * + * %-EIO - I/O error + * + * %-ENOSPC - No space left on device (only in a panic state). + * + * %-ERESTARTSYS - Interrupted. + * + * %-ENOMEM - Insufficient memory available. + * + * %-EFAULT - Failure during execution of requested operation. + */ static int nilfs_ioctl_sync(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { @@ -687,8 +1021,14 @@ static int nilfs_ioctl_sync(struct inode *inode, struct file *filp, if (ret < 0) return ret; + nilfs = inode->i_sb->s_fs_info; + if (nilfs_test_opt(nilfs, BARRIER)) { + ret = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); + if (ret == -EIO) + return ret; + } + if (argp != NULL) { - nilfs = inode->i_sb->s_fs_info; down_read(&nilfs->ns_segctor_sem); cno = nilfs->ns_cno - 1; up_read(&nilfs->ns_segctor_sem); @@ -698,6 +1038,146 @@ static int nilfs_ioctl_sync(struct inode *inode, struct file *filp, return 0; } +/** + * nilfs_ioctl_resize - resize NILFS2 volume + * @inode: inode object + * @filp: file object + * @argp: pointer on argument from userspace + * + * Return Value: On success, 0 is returned or error code, otherwise. + */ +static int nilfs_ioctl_resize(struct inode *inode, struct file *filp, + void __user *argp) +{ + __u64 newsize; + int ret = -EPERM; + + if (!capable(CAP_SYS_ADMIN)) + goto out; + + ret = mnt_want_write_file(filp); + if (ret) + goto out; + + ret = -EFAULT; + if (copy_from_user(&newsize, argp, sizeof(newsize))) + goto out_drop_write; + + ret = nilfs_resize_fs(inode->i_sb, newsize); + +out_drop_write: + mnt_drop_write_file(filp); +out: + return ret; +} + +/** + * nilfs_ioctl_trim_fs() - trim ioctl handle function + * @inode: inode object + * @argp: pointer on argument from userspace + * + * Decription: nilfs_ioctl_trim_fs is the FITRIM ioctl handle function. It + * checks the arguments from userspace and calls nilfs_sufile_trim_fs, which + * performs the actual trim operation. + * + * Return Value: On success, 0 is returned or negative error code, otherwise. + */ +static int nilfs_ioctl_trim_fs(struct inode *inode, void __user *argp) +{ + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + struct request_queue *q = bdev_get_queue(nilfs->ns_bdev); + struct fstrim_range range; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (!blk_queue_discard(q)) + return -EOPNOTSUPP; + + if (copy_from_user(&range, argp, sizeof(range))) + return -EFAULT; + + range.minlen = max_t(u64, range.minlen, q->limits.discard_granularity); + + down_read(&nilfs->ns_segctor_sem); + ret = nilfs_sufile_trim_fs(nilfs->ns_sufile, &range); + up_read(&nilfs->ns_segctor_sem); + + if (ret < 0) + return ret; + + if (copy_to_user(argp, &range, sizeof(range))) + return -EFAULT; + + return 0; +} + +/** + * nilfs_ioctl_set_alloc_range - limit range of segments to be allocated + * @inode: inode object + * @argp: pointer on argument from userspace + * + * Decription: nilfs_ioctl_set_alloc_range() function defines lower limit + * of segments in bytes and upper limit of segments in bytes. + * The NILFS_IOCTL_SET_ALLOC_RANGE is used by nilfs_resize utility. + * + * Return Value: On success, 0 is returned or error code, otherwise. + */ +static int nilfs_ioctl_set_alloc_range(struct inode *inode, void __user *argp) +{ + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + __u64 range[2]; + __u64 minseg, maxseg; + unsigned long segbytes; + int ret = -EPERM; + + if (!capable(CAP_SYS_ADMIN)) + goto out; + + ret = -EFAULT; + if (copy_from_user(range, argp, sizeof(__u64[2]))) + goto out; + + ret = -ERANGE; + if (range[1] > i_size_read(inode->i_sb->s_bdev->bd_inode)) + goto out; + + segbytes = nilfs->ns_blocks_per_segment * nilfs->ns_blocksize; + + minseg = range[0] + segbytes - 1; + do_div(minseg, segbytes); + maxseg = NILFS_SB2_OFFSET_BYTES(range[1]); + do_div(maxseg, segbytes); + maxseg--; + + ret = nilfs_sufile_set_alloc_range(nilfs->ns_sufile, minseg, maxseg); +out: + return ret; +} + +/** + * nilfs_ioctl_get_info - wrapping function of get metadata info + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * @membsz: size of an item in bytes + * @dofunc: concrete function of getting metadata info + * + * Description: nilfs_ioctl_get_info() gets metadata info by means of + * calling dofunc() function. + * + * Return Value: On success, 0 is returned and requested metadata info + * is copied into userspace. On error, one of the following + * negative error codes is returned. + * + * %-EINVAL - Invalid arguments from userspace. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EFAULT - Failure during execution of requested operation. + */ static int nilfs_ioctl_get_info(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp, size_t membsz, @@ -725,9 +1205,98 @@ static int nilfs_ioctl_get_info(struct inode *inode, struct file *filp, return ret; } +/** + * nilfs_ioctl_set_suinfo - set segment usage info + * @inode: inode object + * @filp: file object + * @cmd: ioctl's request code + * @argp: pointer on argument from userspace + * + * Description: Expects an array of nilfs_suinfo_update structures + * encapsulated in nilfs_argv and updates the segment usage info + * according to the flags in nilfs_suinfo_update. + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-EPERM - Not enough permissions + * + * %-EFAULT - Error copying input data + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EINVAL - Invalid values in input (segment number, flags or nblocks) + */ +static int nilfs_ioctl_set_suinfo(struct inode *inode, struct file *filp, + unsigned int cmd, void __user *argp) +{ + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + struct nilfs_transaction_info ti; + struct nilfs_argv argv; + size_t len; + void __user *base; + void *kbuf; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + ret = mnt_want_write_file(filp); + if (ret) + return ret; + + ret = -EFAULT; + if (copy_from_user(&argv, argp, sizeof(argv))) + goto out; + + ret = -EINVAL; + if (argv.v_size < sizeof(struct nilfs_suinfo_update)) + goto out; + + if (argv.v_nmembs > nilfs->ns_nsegments) + goto out; + + if (argv.v_nmembs >= UINT_MAX / argv.v_size) + goto out; + + len = argv.v_size * argv.v_nmembs; + if (!len) { + ret = 0; + goto out; + } + + base = (void __user *)(unsigned long)argv.v_base; + kbuf = vmalloc(len); + if (!kbuf) { + ret = -ENOMEM; + goto out; + } + + if (copy_from_user(kbuf, base, len)) { + ret = -EFAULT; + goto out_free; + } + + nilfs_transaction_begin(inode->i_sb, &ti, 0); + ret = nilfs_sufile_set_suinfo(nilfs->ns_sufile, kbuf, argv.v_size, + argv.v_nmembs); + if (unlikely(ret < 0)) + nilfs_transaction_abort(inode->i_sb); + else + nilfs_transaction_commit(inode->i_sb); /* never fails */ + +out_free: + vfree(kbuf); +out: + mnt_drop_write_file(filp); + return ret; +} + long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); void __user *argp = (void __user *)arg; switch (cmd) { @@ -751,6 +1320,8 @@ long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return nilfs_ioctl_get_info(inode, filp, cmd, argp, sizeof(struct nilfs_suinfo), nilfs_ioctl_do_get_suinfo); + case NILFS_IOCTL_SET_SUINFO: + return nilfs_ioctl_set_suinfo(inode, filp, cmd, argp); case NILFS_IOCTL_GET_SUSTAT: return nilfs_ioctl_get_sustat(inode, filp, cmd, argp); case NILFS_IOCTL_GET_VINFO: @@ -763,6 +1334,12 @@ long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return nilfs_ioctl_clean_segments(inode, filp, cmd, argp); case NILFS_IOCTL_SYNC: return nilfs_ioctl_sync(inode, filp, cmd, argp); + case NILFS_IOCTL_RESIZE: + return nilfs_ioctl_resize(inode, filp, argp); + case NILFS_IOCTL_SET_ALLOC_RANGE: + return nilfs_ioctl_set_alloc_range(inode, argp); + case FITRIM: + return nilfs_ioctl_trim_fs(inode, argp); default: return -ENOTTY; } @@ -781,6 +1358,21 @@ long nilfs_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) case FS_IOC32_GETVERSION: cmd = FS_IOC_GETVERSION; break; + case NILFS_IOCTL_CHANGE_CPMODE: + case NILFS_IOCTL_DELETE_CHECKPOINT: + case NILFS_IOCTL_GET_CPINFO: + case NILFS_IOCTL_GET_CPSTAT: + case NILFS_IOCTL_GET_SUINFO: + case NILFS_IOCTL_SET_SUINFO: + case NILFS_IOCTL_GET_SUSTAT: + case NILFS_IOCTL_GET_VINFO: + case NILFS_IOCTL_GET_BDESCS: + case NILFS_IOCTL_CLEAN_SEGMENTS: + case NILFS_IOCTL_SYNC: + case NILFS_IOCTL_RESIZE: + case NILFS_IOCTL_SET_ALLOC_RANGE: + case FITRIM: + break; default: return -ENOIOCTLCMD; } diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index a649b05f706..c4dcd1db57e 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -58,15 +58,15 @@ nilfs_mdt_insert_new_block(struct inode *inode, unsigned long block, set_buffer_mapped(bh); - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); memset(kaddr + bh_offset(bh), 0, 1 << inode->i_blkbits); if (init_block) init_block(inode, bh, kaddr); flush_dcache_page(bh->b_page); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); set_buffer_uptodate(bh); - nilfs_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(inode); return 0; } @@ -355,7 +355,7 @@ int nilfs_mdt_mark_block_dirty(struct inode *inode, unsigned long block) err = nilfs_mdt_read_block(inode, block, 0, &bh); if (unlikely(err)) return err; - nilfs_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(inode); brelse(bh); return 0; @@ -375,14 +375,25 @@ int nilfs_mdt_fetch_dirty(struct inode *inode) static int nilfs_mdt_write_page(struct page *page, struct writeback_control *wbc) { - struct inode *inode; + struct inode *inode = page->mapping->host; struct super_block *sb; int err = 0; + if (inode && (inode->i_sb->s_flags & MS_RDONLY)) { + /* + * It means that filesystem was remounted in read-only + * mode because of error or metadata corruption. But we + * have dirty pages that try to be flushed in background. + * So, here we simply discard this dirty page. + */ + nilfs_clear_dirty_page(page, false); + unlock_page(page); + return -EROFS; + } + redirty_page_for_writepage(wbc, page); unlock_page(page); - inode = page->mapping->host; if (!inode) return 0; @@ -450,9 +461,9 @@ int nilfs_mdt_setup_shadow_map(struct inode *inode, INIT_LIST_HEAD(&shadow->frozen_buffers); address_space_init_once(&shadow->frozen_data); - nilfs_mapping_init(&shadow->frozen_data, bdi); + nilfs_mapping_init(&shadow->frozen_data, inode, bdi); address_space_init_once(&shadow->frozen_btnodes); - nilfs_mapping_init(&shadow->frozen_btnodes, bdi); + nilfs_mapping_init(&shadow->frozen_btnodes, inode, bdi); mi->mi_shadow = shadow; return 0; } @@ -561,10 +572,10 @@ void nilfs_mdt_restore_from_shadow_map(struct inode *inode) if (mi->mi_palloc_cache) nilfs_palloc_clear_cache(inode); - nilfs_clear_dirty_pages(inode->i_mapping); + nilfs_clear_dirty_pages(inode->i_mapping, true); nilfs_copy_back_pages(inode->i_mapping, &shadow->frozen_data); - nilfs_clear_dirty_pages(&ii->i_btnode_cache); + nilfs_clear_dirty_pages(&ii->i_btnode_cache, true); nilfs_copy_back_pages(&ii->i_btnode_cache, &shadow->frozen_btnodes); nilfs_bmap_restore(ii->i_bmap, &shadow->bmap_store); diff --git a/fs/nilfs2/mdt.h b/fs/nilfs2/mdt.h index ed68563ec70..ab172e8549c 100644 --- a/fs/nilfs2/mdt.h +++ b/fs/nilfs2/mdt.h @@ -28,6 +28,13 @@ #include "nilfs.h" #include "page.h" +/** + * struct nilfs_shadow_map - shadow mapping of meta data file + * @bmap_store: shadow copy of bmap state + * @frozen_data: shadowed dirty data pages + * @frozen_btnodes: shadowed dirty b-tree nodes' pages + * @frozen_buffers: list of frozen buffers + */ struct nilfs_shadow_map { struct nilfs_bmap_store bmap_store; struct address_space frozen_data; @@ -64,11 +71,6 @@ static inline struct nilfs_mdt_info *NILFS_MDT(const struct inode *inode) return inode->i_private; } -static inline struct the_nilfs *NILFS_I_NILFS(struct inode *inode) -{ - return inode->i_sb->s_fs_info; -} - /* Default GFP flags using highmem */ #define NILFS_MDT_GFP (__GFP_WAIT | __GFP_IO | __GFP_HIGHMEM) @@ -93,8 +95,6 @@ int nilfs_mdt_freeze_buffer(struct inode *inode, struct buffer_head *bh); struct buffer_head *nilfs_mdt_get_frozen_buffer(struct inode *inode, struct buffer_head *bh); -#define nilfs_mdt_mark_buffer_dirty(bh) nilfs_mark_buffer_dirty(bh) - static inline void nilfs_mdt_mark_dirty(struct inode *inode) { if (!test_bit(NILFS_I_DIRTY, &NILFS_I(inode)->i_state)) @@ -108,7 +108,7 @@ static inline void nilfs_mdt_clear_dirty(struct inode *inode) static inline __u64 nilfs_mdt_cno(struct inode *inode) { - return NILFS_I_NILFS(inode)->ns_cno; + return ((struct the_nilfs *)inode->i_sb->s_fs_info)->ns_cno; } #define nilfs_mdt_bgl_lock(inode, bg) \ diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index 546849b3e88..9de78f08989 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -63,7 +63,7 @@ static inline int nilfs_add_nondir(struct dentry *dentry, struct inode *inode) */ static struct dentry * -nilfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) +nilfs_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) { struct inode *inode; ino_t ino; @@ -72,12 +72,7 @@ nilfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) return ERR_PTR(-ENAMETOOLONG); ino = nilfs_inode_by_name(dir, &dentry->d_name); - inode = NULL; - if (ino) { - inode = nilfs_iget(dir->i_sb, NILFS_I(dir)->i_root, ino); - if (IS_ERR(inode)) - return ERR_CAST(inode); - } + inode = ino ? nilfs_iget(dir->i_sb, NILFS_I(dir)->i_root, ino) : NULL; return d_splice_alias(inode, dentry); } @@ -89,8 +84,8 @@ nilfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) * If the create succeeds, we fill in the inode information * with d_instantiate(). */ -static int nilfs_create(struct inode *dir, struct dentry *dentry, int mode, - struct nameidata *nd) +static int nilfs_create(struct inode *dir, struct dentry *dentry, umode_t mode, + bool excl) { struct inode *inode; struct nilfs_transaction_info ti; @@ -117,7 +112,7 @@ static int nilfs_create(struct inode *dir, struct dentry *dentry, int mode, } static int -nilfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) +nilfs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, dev_t rdev) { struct inode *inode; struct nilfs_transaction_info ti; @@ -198,9 +193,6 @@ static int nilfs_link(struct dentry *old_dentry, struct inode *dir, struct nilfs_transaction_info ti; int err; - if (inode->i_nlink >= NILFS_LINK_MAX) - return -EMLINK; - err = nilfs_transaction_begin(dir->i_sb, &ti, 1); if (err) return err; @@ -218,15 +210,12 @@ static int nilfs_link(struct dentry *old_dentry, struct inode *dir, return err; } -static int nilfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) +static int nilfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) { struct inode *inode; struct nilfs_transaction_info ti; int err; - if (dir->i_nlink >= NILFS_LINK_MAX) - return -EMLINK; - err = nilfs_transaction_begin(dir->i_sb, &ti, 1); if (err) return err; @@ -294,7 +283,7 @@ static int nilfs_do_unlink(struct inode *dir, struct dentry *dentry) nilfs_warning(inode->i_sb, __func__, "deleting nonexistent file (%lu), %d\n", inode->i_ino, inode->i_nlink); - inode->i_nlink = 1; + set_nlink(inode, 1); } err = nilfs_delete_entry(de, page); if (err) @@ -405,11 +394,6 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry, drop_nlink(new_inode); nilfs_mark_inode_dirty(new_inode); } else { - if (dir_de) { - err = -EMLINK; - if (new_dir->i_nlink >= NILFS_LINK_MAX) - goto out_dir; - } err = nilfs_add_link(new_dentry, old_inode); if (err) goto out_dir; @@ -457,7 +441,7 @@ static struct dentry *nilfs_get_parent(struct dentry *child) { unsigned long ino; struct inode *inode; - struct qstr dotdot = {.name = "..", .len = 2}; + struct qstr dotdot = QSTR_INIT("..", 2); struct nilfs_root *root; ino = nilfs_inode_by_name(child->d_inode, &dotdot); @@ -524,31 +508,29 @@ static struct dentry *nilfs_fh_to_parent(struct super_block *sb, struct fid *fh, return nilfs_get_dentry(sb, fid->cno, fid->parent_ino, fid->parent_gen); } -static int nilfs_encode_fh(struct dentry *dentry, __u32 *fh, int *lenp, - int connectable) +static int nilfs_encode_fh(struct inode *inode, __u32 *fh, int *lenp, + struct inode *parent) { struct nilfs_fid *fid = (struct nilfs_fid *)fh; - struct inode *inode = dentry->d_inode; struct nilfs_root *root = NILFS_I(inode)->i_root; int type; - if (*lenp < NILFS_FID_SIZE_NON_CONNECTABLE || - (connectable && *lenp < NILFS_FID_SIZE_CONNECTABLE)) - return 255; + if (parent && *lenp < NILFS_FID_SIZE_CONNECTABLE) { + *lenp = NILFS_FID_SIZE_CONNECTABLE; + return FILEID_INVALID; + } + if (*lenp < NILFS_FID_SIZE_NON_CONNECTABLE) { + *lenp = NILFS_FID_SIZE_NON_CONNECTABLE; + return FILEID_INVALID; + } fid->cno = root->cno; fid->ino = inode->i_ino; fid->gen = inode->i_generation; - if (connectable && !S_ISDIR(inode->i_mode)) { - struct inode *parent; - - spin_lock(&dentry->d_lock); - parent = dentry->d_parent->d_inode; + if (parent) { fid->parent_ino = parent->i_ino; fid->parent_gen = parent->i_generation; - spin_unlock(&dentry->d_lock); - type = FILEID_NILFS_WITH_PARENT; *lenp = NILFS_FID_SIZE_CONNECTABLE; } else { diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index a8dd344303c..9bc72dec3fa 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -32,8 +32,21 @@ #include "the_nilfs.h" #include "bmap.h" -/* - * nilfs inode data in memory +/** + * struct nilfs_inode_info - nilfs inode data in memory + * @i_flags: inode flags + * @i_state: dynamic state flags + * @i_bmap: pointer on i_bmap_data + * @i_bmap_data: raw block mapping + * @i_xattr: <TODO> + * @i_dir_start_lookup: page index of last successful search + * @i_cno: checkpoint number for GC inode + * @i_btnode_cache: cached pages of b-tree nodes + * @i_dirty: list for connecting dirty files + * @xattr_sem: semaphore for extended attributes processing + * @i_bh: buffer contains disk inode + * @i_root: root object of the current filesystem tree + * @vfs_inode: VFS inode object */ struct nilfs_inode_info { __u32 i_flags; @@ -80,12 +93,6 @@ static inline struct inode *NILFS_BTNC_I(struct address_space *btnc) return &ii->vfs_inode; } -static inline struct inode *NILFS_AS_I(struct address_space *mapping) -{ - return (mapping->host) ? : - container_of(mapping, struct inode, i_data); -} - /* * Dynamic state flags of NILFS on-memory inode (i_state) */ @@ -241,7 +248,7 @@ extern void nilfs_set_link(struct inode *, struct nilfs_dir_entry *, struct page *, struct inode *); /* file.c */ -extern int nilfs_sync_file(struct file *, int); +extern int nilfs_sync_file(struct file *, loff_t, loff_t, int); /* ioctl.c */ long nilfs_ioctl(struct file *, unsigned int, unsigned long); @@ -252,7 +259,7 @@ int nilfs_ioctl_prepare_clean_segments(struct the_nilfs *, struct nilfs_argv *, /* inode.c */ void nilfs_inode_add_blocks(struct inode *inode, int n); void nilfs_inode_sub_blocks(struct inode *inode, int n); -extern struct inode *nilfs_new_inode(struct inode *, int); +extern struct inode *nilfs_new_inode(struct inode *, umode_t); extern void nilfs_free_inode(struct inode *); extern int nilfs_get_block(struct inode *, sector_t, struct buffer_head *, int); extern void nilfs_set_inode_flags(struct inode *); @@ -270,22 +277,23 @@ extern void nilfs_update_inode(struct inode *, struct buffer_head *); extern void nilfs_truncate(struct inode *); extern void nilfs_evict_inode(struct inode *); extern int nilfs_setattr(struct dentry *, struct iattr *); -int nilfs_permission(struct inode *inode, int mask, unsigned int flags); +extern void nilfs_write_failed(struct address_space *mapping, loff_t to); +int nilfs_permission(struct inode *inode, int mask); int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh); extern int nilfs_inode_dirty(struct inode *); int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty); extern int nilfs_mark_inode_dirty(struct inode *); -extern void nilfs_dirty_inode(struct inode *); +extern void nilfs_dirty_inode(struct inode *, int flags); int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); /* super.c */ extern struct inode *nilfs_alloc_inode(struct super_block *); extern void nilfs_destroy_inode(struct inode *); -extern void nilfs_error(struct super_block *, const char *, const char *, ...) - __attribute__ ((format (printf, 3, 4))); -extern void nilfs_warning(struct super_block *, const char *, const char *, ...) - __attribute__ ((format (printf, 3, 4))); +extern __printf(3, 4) +void nilfs_error(struct super_block *, const char *, const char *, ...); +extern __printf(3, 4) +void nilfs_warning(struct super_block *, const char *, const char *, ...); extern struct nilfs_super_block * nilfs_read_super_block(struct super_block *, u64, int, struct buffer_head **); extern int nilfs_store_magic_and_option(struct super_block *, @@ -298,6 +306,7 @@ struct nilfs_super_block **nilfs_prepare_super(struct super_block *sb, int flip); int nilfs_commit_super(struct super_block *sb, int flag); int nilfs_cleanup_super(struct super_block *sb); +int nilfs_resize_fs(struct super_block *sb, __u64 newsize); int nilfs_attach_checkpoint(struct super_block *sb, __u64 cno, int curr_mnt, struct nilfs_root **root); int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno); diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index 1168059c7ef..da276640f77 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -37,8 +37,7 @@ #define NILFS_BUFFER_INHERENT_BITS \ ((1UL << BH_Uptodate) | (1UL << BH_Mapped) | (1UL << BH_NILFS_Node) | \ - (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Allocated) | \ - (1UL << BH_NILFS_Checked)) + (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Checked)) static struct buffer_head * __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index, @@ -59,19 +58,6 @@ __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index, return bh; } -/* - * Since the page cache of B-tree node pages or data page cache of pseudo - * inodes does not have a valid mapping->host pointer, calling - * mark_buffer_dirty() for their buffers causes a NULL pointer dereference; - * it calls __mark_inode_dirty(NULL) through __set_page_dirty(). - * To avoid this problem, the old style mark_buffer_dirty() is used instead. - */ -void nilfs_mark_buffer_dirty(struct buffer_head *bh) -{ - if (!buffer_dirty(bh) && !test_set_buffer_dirty(bh)) - __set_page_dirty_nobuffers(bh->b_page); -} - struct buffer_head *nilfs_grab_buffer(struct inode *inode, struct address_space *mapping, unsigned long blkoff, @@ -108,6 +94,7 @@ void nilfs_forget_buffer(struct buffer_head *bh) clear_buffer_nilfs_volatile(bh); clear_buffer_nilfs_checked(bh); clear_buffer_nilfs_redirected(bh); + clear_buffer_async_write(bh); clear_buffer_dirty(bh); if (nilfs_page_buffers_clean(page)) __nilfs_clear_page_dirty(page); @@ -133,11 +120,11 @@ void nilfs_copy_buffer(struct buffer_head *dbh, struct buffer_head *sbh) struct page *spage = sbh->b_page, *dpage = dbh->b_page; struct buffer_head *bh; - kaddr0 = kmap_atomic(spage, KM_USER0); - kaddr1 = kmap_atomic(dpage, KM_USER1); + kaddr0 = kmap_atomic(spage); + kaddr1 = kmap_atomic(dpage); memcpy(kaddr1 + bh_offset(dbh), kaddr0 + bh_offset(sbh), sbh->b_size); - kunmap_atomic(kaddr1, KM_USER1); - kunmap_atomic(kaddr0, KM_USER0); + kunmap_atomic(kaddr1); + kunmap_atomic(kaddr0); dbh->b_state = sbh->b_state & NILFS_BUFFER_INHERENT_BITS; dbh->b_blocknr = sbh->b_blocknr; @@ -183,7 +170,7 @@ int nilfs_page_buffers_clean(struct page *page) void nilfs_page_bug(struct page *page) { struct address_space *m; - unsigned long ino = 0; + unsigned long ino; if (unlikely(!page)) { printk(KERN_CRIT "NILFS_PAGE_BUG(NULL)\n"); @@ -191,11 +178,8 @@ void nilfs_page_bug(struct page *page) } m = page->mapping; - if (m) { - struct inode *inode = NILFS_AS_I(m); - if (inode != NULL) - ino = inode->i_ino; - } + ino = m ? m->host->i_ino : 0; + printk(KERN_CRIT "NILFS_PAGE_BUG(%p): cnt=%d index#=%llu flags=0x%lx " "mapping=%p ino=%lu\n", page, atomic_read(&page->_count), @@ -217,56 +201,6 @@ void nilfs_page_bug(struct page *page) } /** - * nilfs_alloc_private_page - allocate a private page with buffer heads - * - * Return Value: On success, a pointer to the allocated page is returned. - * On error, NULL is returned. - */ -struct page *nilfs_alloc_private_page(struct block_device *bdev, int size, - unsigned long state) -{ - struct buffer_head *bh, *head, *tail; - struct page *page; - - page = alloc_page(GFP_NOFS); /* page_count of the returned page is 1 */ - if (unlikely(!page)) - return NULL; - - lock_page(page); - head = alloc_page_buffers(page, size, 0); - if (unlikely(!head)) { - unlock_page(page); - __free_page(page); - return NULL; - } - - bh = head; - do { - bh->b_state = (1UL << BH_NILFS_Allocated) | state; - tail = bh; - bh->b_bdev = bdev; - bh = bh->b_this_page; - } while (bh); - - tail->b_this_page = head; - attach_page_buffers(page, head); - - return page; -} - -void nilfs_free_private_page(struct page *page) -{ - BUG_ON(!PageLocked(page)); - BUG_ON(page->mapping); - - if (page_has_buffers(page) && !try_to_free_buffers(page)) - NILFS_PAGE_BUG(page, "failed to free page"); - - unlock_page(page); - __free_page(page); -} - -/** * nilfs_copy_page -- copy the page with buffers * @dst: destination page * @src: source page @@ -437,7 +371,12 @@ repeat: goto repeat; } -void nilfs_clear_dirty_pages(struct address_space *mapping) +/** + * nilfs_clear_dirty_pages - discard dirty pages in address space + * @mapping: address space with dirty pages for discarding + * @silent: suppress [true] or print [false] warning messages + */ +void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent) { struct pagevec pvec; unsigned int i; @@ -449,25 +388,9 @@ void nilfs_clear_dirty_pages(struct address_space *mapping) PAGEVEC_SIZE)) { for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; - struct buffer_head *bh, *head; lock_page(page); - ClearPageUptodate(page); - ClearPageMappedToDisk(page); - bh = head = page_buffers(page); - do { - lock_buffer(bh); - clear_buffer_dirty(bh); - clear_buffer_nilfs_volatile(bh); - clear_buffer_nilfs_checked(bh); - clear_buffer_nilfs_redirected(bh); - clear_buffer_uptodate(bh); - clear_buffer_mapped(bh); - unlock_buffer(bh); - bh = bh->b_this_page; - } while (bh != head); - - __nilfs_clear_page_dirty(page); + nilfs_clear_dirty_page(page, silent); unlock_page(page); } pagevec_release(&pvec); @@ -475,6 +398,52 @@ void nilfs_clear_dirty_pages(struct address_space *mapping) } } +/** + * nilfs_clear_dirty_page - discard dirty page + * @page: dirty page that will be discarded + * @silent: suppress [true] or print [false] warning messages + */ +void nilfs_clear_dirty_page(struct page *page, bool silent) +{ + struct inode *inode = page->mapping->host; + struct super_block *sb = inode->i_sb; + + BUG_ON(!PageLocked(page)); + + if (!silent) { + nilfs_warning(sb, __func__, + "discard page: offset %lld, ino %lu", + page_offset(page), inode->i_ino); + } + + ClearPageUptodate(page); + ClearPageMappedToDisk(page); + + if (page_has_buffers(page)) { + struct buffer_head *bh, *head; + + bh = head = page_buffers(page); + do { + lock_buffer(bh); + if (!silent) { + nilfs_warning(sb, __func__, + "discard block %llu, size %zu", + (u64)bh->b_blocknr, bh->b_size); + } + clear_buffer_async_write(bh); + clear_buffer_dirty(bh); + clear_buffer_nilfs_volatile(bh); + clear_buffer_nilfs_checked(bh); + clear_buffer_nilfs_redirected(bh); + clear_buffer_uptodate(bh); + clear_buffer_mapped(bh); + unlock_buffer(bh); + } while (bh = bh->b_this_page, bh != head); + } + + __nilfs_clear_page_dirty(page); +} + unsigned nilfs_page_count_clean_buffers(struct page *page, unsigned from, unsigned to) { @@ -492,13 +461,13 @@ unsigned nilfs_page_count_clean_buffers(struct page *page, return nc; } -void nilfs_mapping_init(struct address_space *mapping, +void nilfs_mapping_init(struct address_space *mapping, struct inode *inode, struct backing_dev_info *bdi) { - mapping->host = NULL; + mapping->host = inode; mapping->flags = 0; mapping_set_gfp_mask(mapping, GFP_NOFS); - mapping->assoc_mapping = NULL; + mapping->private_data = NULL; mapping->backing_dev_info = bdi; mapping->a_ops = &empty_aops; } diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h index f06b79ad749..ef30c5c2426 100644 --- a/fs/nilfs2/page.h +++ b/fs/nilfs2/page.h @@ -38,14 +38,12 @@ enum { BH_NILFS_Redirected, }; -BUFFER_FNS(NILFS_Allocated, nilfs_allocated) /* nilfs private buffers */ BUFFER_FNS(NILFS_Node, nilfs_node) /* nilfs node buffers */ BUFFER_FNS(NILFS_Volatile, nilfs_volatile) BUFFER_FNS(NILFS_Checked, nilfs_checked) /* buffer is verified */ BUFFER_FNS(NILFS_Redirected, nilfs_redirected) /* redirected to a copy */ -void nilfs_mark_buffer_dirty(struct buffer_head *bh); int __nilfs_clear_page_dirty(struct page *); struct buffer_head *nilfs_grab_buffer(struct inode *, struct address_space *, @@ -54,14 +52,12 @@ void nilfs_forget_buffer(struct buffer_head *); void nilfs_copy_buffer(struct buffer_head *, struct buffer_head *); int nilfs_page_buffers_clean(struct page *); void nilfs_page_bug(struct page *); -struct page *nilfs_alloc_private_page(struct block_device *, int, - unsigned long); -void nilfs_free_private_page(struct page *); int nilfs_copy_dirty_pages(struct address_space *, struct address_space *); void nilfs_copy_back_pages(struct address_space *, struct address_space *); -void nilfs_clear_dirty_pages(struct address_space *); -void nilfs_mapping_init(struct address_space *mapping, +void nilfs_clear_dirty_page(struct page *, bool); +void nilfs_clear_dirty_pages(struct address_space *, bool); +void nilfs_mapping_init(struct address_space *mapping, struct inode *inode, struct backing_dev_info *bdi); unsigned nilfs_page_count_clean_buffers(struct page *, unsigned, unsigned); unsigned long nilfs_find_uncommitted_extent(struct inode *inode, diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c index ba4a64518f3..ff00a0b7acb 100644 --- a/fs/nilfs2/recovery.c +++ b/fs/nilfs2/recovery.c @@ -387,9 +387,9 @@ static int nilfs_scan_dsync_log(struct the_nilfs *nilfs, sector_t start_blocknr, static void dispose_recovery_list(struct list_head *head) { while (!list_empty(head)) { - struct nilfs_recovery_block *rb - = list_entry(head->next, - struct nilfs_recovery_block, list); + struct nilfs_recovery_block *rb; + + rb = list_first_entry(head, struct nilfs_recovery_block, list); list_del(&rb->list); kfree(rb); } @@ -416,9 +416,9 @@ static int nilfs_segment_list_add(struct list_head *head, __u64 segnum) void nilfs_dispose_segment_list(struct list_head *head) { while (!list_empty(head)) { - struct nilfs_segment_entry *ent - = list_entry(head->next, - struct nilfs_segment_entry, list); + struct nilfs_segment_entry *ent; + + ent = list_first_entry(head, struct nilfs_segment_entry, list); list_del(&ent->list); kfree(ent); } @@ -493,9 +493,9 @@ static int nilfs_recovery_copy_block(struct the_nilfs *nilfs, if (unlikely(!bh_org)) return -EIO; - kaddr = kmap_atomic(page, KM_USER0); + kaddr = kmap_atomic(page); memcpy(kaddr + bh_offset(bh_org), bh_org->b_data, bh_org->b_size); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(bh_org); return 0; } @@ -527,7 +527,8 @@ static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs, if (unlikely(err)) { loff_t isize = inode->i_size; if (pos + blocksize > isize) - vmtruncate(inode, isize); + nilfs_write_failed(inode->i_mapping, + pos + blocksize); goto failed_inode; } diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c index 2853ff20f85..dc3a9efdaab 100644 --- a/fs/nilfs2/segbuf.c +++ b/fs/nilfs2/segbuf.c @@ -227,9 +227,9 @@ static void nilfs_segbuf_fill_in_data_crc(struct nilfs_segment_buffer *segbuf, crc = crc32_le(crc, bh->b_data, bh->b_size); } list_for_each_entry(bh, &segbuf->sb_payload_buffers, b_assoc_buffers) { - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); crc = crc32_le(crc, kaddr + bh_offset(bh), bh->b_size); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); } raw_sum->ss_datasum = cpu_to_le32(crc); } @@ -239,12 +239,15 @@ nilfs_segbuf_fill_in_super_root_crc(struct nilfs_segment_buffer *segbuf, u32 seed) { struct nilfs_super_root *raw_sr; + struct the_nilfs *nilfs = segbuf->sb_super->s_fs_info; + unsigned srsize; u32 crc; raw_sr = (struct nilfs_super_root *)segbuf->sb_super_root->b_data; + srsize = NILFS_SR_BYTES(nilfs->ns_inode_size); crc = crc32_le(seed, (unsigned char *)raw_sr + sizeof(raw_sr->sr_sum), - NILFS_SR_BYTES - sizeof(raw_sr->sr_sum)); + srsize - sizeof(raw_sr->sr_sum)); raw_sr->sr_sum = cpu_to_le32(crc); } @@ -254,18 +257,6 @@ static void nilfs_release_buffers(struct list_head *list) list_for_each_entry_safe(bh, n, list, b_assoc_buffers) { list_del_init(&bh->b_assoc_buffers); - if (buffer_nilfs_allocated(bh)) { - struct page *clone_page = bh->b_page; - - /* remove clone page */ - brelse(bh); - page_cache_release(clone_page); /* for each bh */ - if (page_count(clone_page) <= 2) { - lock_page(clone_page); - nilfs_free_private_page(clone_page); - } - continue; - } brelse(bh); } } @@ -354,8 +345,7 @@ static void nilfs_end_bio_write(struct bio *bio, int err) if (err == -EOPNOTSUPP) { set_bit(BIO_EOPNOTSUPP, &bio->bi_flags); - bio_put(bio); - /* to be detected by submit_seg_bio() */ + /* to be detected by nilfs_segbuf_submit_bio() */ } if (!uptodate) @@ -386,12 +376,12 @@ static int nilfs_segbuf_submit_bio(struct nilfs_segment_buffer *segbuf, bio->bi_private = segbuf; bio_get(bio); submit_bio(mode, bio); + segbuf->sb_nbio++; if (bio_flagged(bio, BIO_EOPNOTSUPP)) { bio_put(bio); err = -EOPNOTSUPP; goto failed; } - segbuf->sb_nbio++; bio_put(bio); wi->bio = NULL; @@ -426,7 +416,8 @@ static struct bio *nilfs_alloc_seg_bio(struct the_nilfs *nilfs, sector_t start, } if (likely(bio)) { bio->bi_bdev = nilfs->ns_bdev; - bio->bi_sector = start << (nilfs->ns_blocksize_bits - 9); + bio->bi_iter.bi_sector = + start << (nilfs->ns_blocksize_bits - 9); } return bio; } diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index afe4f218345..a1a191634ab 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -189,7 +189,7 @@ int nilfs_transaction_begin(struct super_block *sb, if (ret > 0) return 0; - vfs_check_frozen(sb, SB_FREEZE_WRITE); + sb_start_intwrite(sb); nilfs = sb->s_fs_info; down_read(&nilfs->ns_segctor_sem); @@ -205,6 +205,7 @@ int nilfs_transaction_begin(struct super_block *sb, current->journal_info = ti->ti_save; if (ti->ti_flags & NILFS_TI_DYNAMIC_ALLOC) kmem_cache_free(nilfs_transaction_cachep, ti); + sb_end_intwrite(sb); return ret; } @@ -246,6 +247,7 @@ int nilfs_transaction_commit(struct super_block *sb) err = nilfs_construct_segment(sb); if (ti->ti_flags & NILFS_TI_DYNAMIC_ALLOC) kmem_cache_free(nilfs_transaction_cachep, ti); + sb_end_intwrite(sb); return err; } @@ -264,6 +266,7 @@ void nilfs_transaction_abort(struct super_block *sb) current->journal_info = ti->ti_save; if (ti->ti_flags & NILFS_TI_DYNAMIC_ALLOC) kmem_cache_free(nilfs_transaction_cachep, ti); + sb_end_intwrite(sb); } void nilfs_relax_pressure_in_lock(struct super_block *sb) @@ -655,17 +658,14 @@ static size_t nilfs_lookup_dirty_data_buffers(struct inode *inode, if (unlikely(page->index > last)) break; - if (mapping->host) { - lock_page(page); - if (!page_has_buffers(page)) - create_empty_buffers(page, - 1 << inode->i_blkbits, 0); - unlock_page(page); - } + lock_page(page); + if (!page_has_buffers(page)) + create_empty_buffers(page, 1 << inode->i_blkbits, 0); + unlock_page(page); bh = head = page_buffers(page); do { - if (!buffer_dirty(bh)) + if (!buffer_dirty(bh) || buffer_async_write(bh)) continue; get_bh(bh); list_add_tail(&bh->b_assoc_buffers, listp); @@ -699,7 +699,8 @@ static void nilfs_lookup_dirty_node_buffers(struct inode *inode, for (i = 0; i < pagevec_count(&pvec); i++) { bh = head = page_buffers(pvec.pages[i]); do { - if (buffer_dirty(bh)) { + if (buffer_dirty(bh) && + !buffer_async_write(bh)) { get_bh(bh); list_add_tail(&bh->b_assoc_buffers, listp); @@ -809,7 +810,7 @@ static int nilfs_segctor_create_checkpoint(struct nilfs_sc_info *sci) /* The following code is duplicated with cpfile. But, it is needed to collect the checkpoint even if it was not newly created */ - nilfs_mdt_mark_buffer_dirty(bh_cp); + mark_buffer_dirty(bh_cp); nilfs_mdt_mark_dirty(nilfs->ns_cpfile); nilfs_cpfile_put_checkpoint( nilfs->ns_cpfile, nilfs->ns_cno, bh_cp); @@ -835,9 +836,9 @@ static int nilfs_segctor_fill_in_checkpoint(struct nilfs_sc_info *sci) raw_cp->cp_snapshot_list.ssl_next = 0; raw_cp->cp_snapshot_list.ssl_prev = 0; raw_cp->cp_inodes_count = - cpu_to_le64(atomic_read(&sci->sc_root->inodes_count)); + cpu_to_le64(atomic64_read(&sci->sc_root->inodes_count)); raw_cp->cp_blocks_count = - cpu_to_le64(atomic_read(&sci->sc_root->blocks_count)); + cpu_to_le64(atomic64_read(&sci->sc_root->blocks_count)); raw_cp->cp_nblk_inc = cpu_to_le64(sci->sc_nblk_inc + sci->sc_nblk_this_inc); raw_cp->cp_create = cpu_to_le64(sci->sc_seg_ctime); @@ -889,12 +890,14 @@ static void nilfs_segctor_fill_in_super_root(struct nilfs_sc_info *sci, { struct buffer_head *bh_sr; struct nilfs_super_root *raw_sr; - unsigned isz = nilfs->ns_inode_size; + unsigned isz, srsz; bh_sr = NILFS_LAST_SEGBUF(&sci->sc_segbufs)->sb_super_root; raw_sr = (struct nilfs_super_root *)bh_sr->b_data; + isz = nilfs->ns_inode_size; + srsz = NILFS_SR_BYTES(isz); - raw_sr->sr_bytes = cpu_to_le16(NILFS_SR_BYTES); + raw_sr->sr_bytes = cpu_to_le16(srsz); raw_sr->sr_nongc_ctime = cpu_to_le64(nilfs_doing_gc() ? nilfs->ns_nongc_ctime : sci->sc_seg_ctime); @@ -906,6 +909,7 @@ static void nilfs_segctor_fill_in_super_root(struct nilfs_sc_info *sci, NILFS_SR_CPFILE_OFFSET(isz), 1); nilfs_write_inode_common(nilfs->ns_sufile, (void *)raw_sr + NILFS_SR_SUFILE_OFFSET(isz), 1); + memset((void *)raw_sr + srsz, 0, nilfs->ns_blocksize - srsz); } static void nilfs_redirty_inodes(struct list_head *head) @@ -954,8 +958,8 @@ static int nilfs_segctor_apply_buffers(struct nilfs_sc_info *sci, dispose_buffers: while (!list_empty(listp)) { - bh = list_entry(listp->next, struct buffer_head, - b_assoc_buffers); + bh = list_first_entry(listp, struct buffer_head, + b_assoc_buffers); list_del_init(&bh->b_assoc_buffers); brelse(bh); } @@ -1436,17 +1440,19 @@ static int nilfs_segctor_collect(struct nilfs_sc_info *sci, nilfs_clear_logs(&sci->sc_segbufs); - err = nilfs_segctor_extend_segments(sci, nilfs, nadd); - if (unlikely(err)) - return err; - if (sci->sc_stage.flags & NILFS_CF_SUFREED) { err = nilfs_sufile_cancel_freev(nilfs->ns_sufile, sci->sc_freesegs, sci->sc_nfreesegs, NULL); WARN_ON(err); /* do not happen */ + sci->sc_stage.flags &= ~NILFS_CF_SUFREED; } + + err = nilfs_segctor_extend_segments(sci, nilfs, nadd); + if (unlikely(err)) + return err; + nadd = min_t(int, nadd << 1, SC_MAX_SEGDELTA); sci->sc_stage = prev_stage; } @@ -1500,10 +1506,7 @@ nilfs_segctor_update_payload_blocknr(struct nilfs_sc_info *sci, nblocks = le32_to_cpu(finfo->fi_nblocks); ndatablk = le32_to_cpu(finfo->fi_ndatablk); - if (buffer_nilfs_node(bh)) - inode = NILFS_BTNC_I(bh->b_page->mapping); - else - inode = NILFS_AS_I(bh->b_page->mapping); + inode = bh->b_page->mapping->host; if (mode == SC_LSEG_DSYNC) sc_op = &nilfs_sc_dsync_ops; @@ -1556,88 +1559,30 @@ static int nilfs_segctor_assign(struct nilfs_sc_info *sci, int mode) return 0; } -static int -nilfs_copy_replace_page_buffers(struct page *page, struct list_head *out) -{ - struct page *clone_page; - struct buffer_head *bh, *head, *bh2; - void *kaddr; - - bh = head = page_buffers(page); - - clone_page = nilfs_alloc_private_page(bh->b_bdev, bh->b_size, 0); - if (unlikely(!clone_page)) - return -ENOMEM; - - bh2 = page_buffers(clone_page); - kaddr = kmap_atomic(page, KM_USER0); - do { - if (list_empty(&bh->b_assoc_buffers)) - continue; - get_bh(bh2); - page_cache_get(clone_page); /* for each bh */ - memcpy(bh2->b_data, kaddr + bh_offset(bh), bh2->b_size); - bh2->b_blocknr = bh->b_blocknr; - list_replace(&bh->b_assoc_buffers, &bh2->b_assoc_buffers); - list_add_tail(&bh->b_assoc_buffers, out); - } while (bh = bh->b_this_page, bh2 = bh2->b_this_page, bh != head); - kunmap_atomic(kaddr, KM_USER0); - - if (!TestSetPageWriteback(clone_page)) - account_page_writeback(clone_page); - unlock_page(clone_page); - - return 0; -} - -static int nilfs_test_page_to_be_frozen(struct page *page) -{ - struct address_space *mapping = page->mapping; - - if (!mapping || !mapping->host || S_ISDIR(mapping->host->i_mode)) - return 0; - - if (page_mapped(page)) { - ClearPageChecked(page); - return 1; - } - return PageChecked(page); -} - -static int nilfs_begin_page_io(struct page *page, struct list_head *out) +static void nilfs_begin_page_io(struct page *page) { if (!page || PageWriteback(page)) /* For split b-tree node pages, this function may be called twice. We ignore the 2nd or later calls by this check. */ - return 0; + return; lock_page(page); clear_page_dirty_for_io(page); set_page_writeback(page); unlock_page(page); - - if (nilfs_test_page_to_be_frozen(page)) { - int err = nilfs_copy_replace_page_buffers(page, out); - if (unlikely(err)) - return err; - } - return 0; } -static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, - struct page **failed_page) +static void nilfs_segctor_prepare_write(struct nilfs_sc_info *sci) { struct nilfs_segment_buffer *segbuf; struct page *bd_page = NULL, *fs_page = NULL; - struct list_head *list = &sci->sc_copied_buffers; - int err; - *failed_page = NULL; list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) { struct buffer_head *bh; list_for_each_entry(bh, &segbuf->sb_segsum_buffers, b_assoc_buffers) { + set_buffer_async_write(bh); if (bh->b_page != bd_page) { if (bd_page) { lock_page(bd_page); @@ -1651,6 +1596,7 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, list_for_each_entry(bh, &segbuf->sb_payload_buffers, b_assoc_buffers) { + set_buffer_async_write(bh); if (bh == segbuf->sb_super_root) { if (bh->b_page != bd_page) { lock_page(bd_page); @@ -1662,11 +1608,7 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, break; } if (bh->b_page != fs_page) { - err = nilfs_begin_page_io(fs_page, list); - if (unlikely(err)) { - *failed_page = fs_page; - goto out; - } + nilfs_begin_page_io(fs_page); fs_page = bh->b_page; } } @@ -1677,11 +1619,7 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, set_page_writeback(bd_page); unlock_page(bd_page); } - err = nilfs_begin_page_io(fs_page, list); - if (unlikely(err)) - *failed_page = fs_page; - out: - return err; + nilfs_begin_page_io(fs_page); } static int nilfs_segctor_write(struct nilfs_sc_info *sci, @@ -1694,24 +1632,6 @@ static int nilfs_segctor_write(struct nilfs_sc_info *sci, return ret; } -static void __nilfs_end_page_io(struct page *page, int err) -{ - if (!err) { - if (!nilfs_page_buffers_clean(page)) - __set_page_dirty_nobuffers(page); - ClearPageError(page); - } else { - __set_page_dirty_nobuffers(page); - SetPageError(page); - } - - if (buffer_nilfs_allocated(page_buffers(page))) { - if (TestClearPageWriteback(page)) - dec_zone_page_state(page, NR_WRITEBACK); - } else - end_page_writeback(page); -} - static void nilfs_end_page_io(struct page *page, int err) { if (!page) @@ -1738,40 +1658,19 @@ static void nilfs_end_page_io(struct page *page, int err) return; } - __nilfs_end_page_io(page, err); -} - -static void nilfs_clear_copied_buffers(struct list_head *list, int err) -{ - struct buffer_head *bh, *head; - struct page *page; - - while (!list_empty(list)) { - bh = list_entry(list->next, struct buffer_head, - b_assoc_buffers); - page = bh->b_page; - page_cache_get(page); - head = bh = page_buffers(page); - do { - if (!list_empty(&bh->b_assoc_buffers)) { - list_del_init(&bh->b_assoc_buffers); - if (!err) { - set_buffer_uptodate(bh); - clear_buffer_dirty(bh); - clear_buffer_delay(bh); - clear_buffer_nilfs_volatile(bh); - } - brelse(bh); /* for b_assoc_buffers */ - } - } while ((bh = bh->b_this_page) != head); - - __nilfs_end_page_io(page, err); - page_cache_release(page); + if (!err) { + if (!nilfs_page_buffers_clean(page)) + __set_page_dirty_nobuffers(page); + ClearPageError(page); + } else { + __set_page_dirty_nobuffers(page); + SetPageError(page); } + + end_page_writeback(page); } -static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, - int err) +static void nilfs_abort_logs(struct list_head *logs, int err) { struct nilfs_segment_buffer *segbuf; struct page *bd_page = NULL, *fs_page = NULL; @@ -1783,6 +1682,7 @@ static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, list_for_each_entry(segbuf, logs, sb_list) { list_for_each_entry(bh, &segbuf->sb_segsum_buffers, b_assoc_buffers) { + clear_buffer_async_write(bh); if (bh->b_page != bd_page) { if (bd_page) end_page_writeback(bd_page); @@ -1792,6 +1692,7 @@ static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, list_for_each_entry(bh, &segbuf->sb_payload_buffers, b_assoc_buffers) { + clear_buffer_async_write(bh); if (bh == segbuf->sb_super_root) { if (bh->b_page != bd_page) { end_page_writeback(bd_page); @@ -1801,8 +1702,6 @@ static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, } if (bh->b_page != fs_page) { nilfs_end_page_io(fs_page, err); - if (fs_page && fs_page == failed_page) - return; fs_page = bh->b_page; } } @@ -1821,12 +1720,11 @@ static void nilfs_segctor_abort_construction(struct nilfs_sc_info *sci, list_splice_tail_init(&sci->sc_write_logs, &logs); ret = nilfs_wait_on_logs(&logs); - nilfs_abort_logs(&logs, NULL, ret ? : err); + nilfs_abort_logs(&logs, ret ? : err); list_splice_tail_init(&sci->sc_segbufs, &logs); nilfs_cancel_segusage(&logs, nilfs->ns_sufile); nilfs_free_incomplete_logs(&logs, nilfs); - nilfs_clear_copied_buffers(&sci->sc_copied_buffers, err); if (sci->sc_stage.flags & NILFS_CF_SUFREED) { ret = nilfs_sufile_cancel_freev(nilfs->ns_sufile, @@ -1864,6 +1762,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) b_assoc_buffers) { set_buffer_uptodate(bh); clear_buffer_dirty(bh); + clear_buffer_async_write(bh); if (bh->b_page != bd_page) { if (bd_page) end_page_writeback(bd_page); @@ -1885,6 +1784,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) b_assoc_buffers) { set_buffer_uptodate(bh); clear_buffer_dirty(bh); + clear_buffer_async_write(bh); clear_buffer_delay(bh); clear_buffer_nilfs_volatile(bh); clear_buffer_nilfs_redirected(bh); @@ -1920,8 +1820,6 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) nilfs_end_page_io(fs_page, 0); - nilfs_clear_copied_buffers(&sci->sc_copied_buffers, 0); - nilfs_drop_collected_inodes(&sci->sc_dirty_files); if (nilfs_doing_gc()) @@ -1979,7 +1877,7 @@ static int nilfs_segctor_collect_dirty_files(struct nilfs_sc_info *sci, "failed to get inode block.\n"); return err; } - nilfs_mdt_mark_buffer_dirty(ibh); + mark_buffer_dirty(ibh); nilfs_mdt_mark_dirty(ifile); spin_lock(&nilfs->ns_inode_lock); if (likely(!ii->i_bh)) @@ -1991,8 +1889,7 @@ static int nilfs_segctor_collect_dirty_files(struct nilfs_sc_info *sci, clear_bit(NILFS_I_QUEUED, &ii->i_state); set_bit(NILFS_I_BUSY, &ii->i_state); - list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &sci->sc_dirty_files); + list_move_tail(&ii->i_dirty, &sci->sc_dirty_files); } spin_unlock(&nilfs->ns_inode_lock); @@ -2014,8 +1911,7 @@ static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, clear_bit(NILFS_I_BUSY, &ii->i_state); brelse(ii->i_bh); ii->i_bh = NULL; - list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &ti->ti_garbage); + list_move_tail(&ii->i_dirty, &ti->ti_garbage); } spin_unlock(&nilfs->ns_inode_lock); } @@ -2026,7 +1922,6 @@ static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) { struct the_nilfs *nilfs = sci->sc_super->s_fs_info; - struct page *failed_page; int err; sci->sc_stage.scnt = NILFS_ST_INIT; @@ -2081,11 +1976,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) nilfs_segctor_update_segusage(sci, nilfs->ns_sufile); /* Write partial segments */ - err = nilfs_segctor_prepare_write(sci, &failed_page); - if (err) { - nilfs_abort_logs(&sci->sc_segbufs, failed_page, err); - goto failed_to_write; - } + nilfs_segctor_prepare_write(sci); nilfs_add_checksums_on_logs(&sci->sc_segbufs, nilfs->ns_crc_seed); @@ -2430,6 +2321,8 @@ nilfs_remove_written_gcinodes(struct the_nilfs *nilfs, struct list_head *head) if (!test_bit(NILFS_I_UPDATED, &ii->i_state)) continue; list_del_init(&ii->i_dirty); + truncate_inode_pages(&ii->vfs_inode.i_data, 0); + nilfs_btnode_cache_clear(&ii->i_btnode_cache); iput(&ii->vfs_inode); } } @@ -2591,7 +2484,7 @@ static int nilfs_segctor_thread(void *arg) if (freezing(current)) { spin_unlock(&sci->sc_state_lock); - refrigerator(); + try_to_freeze(); spin_lock(&sci->sc_state_lock); } else { DEFINE_WAIT(wait); @@ -2687,7 +2580,6 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct super_block *sb, INIT_LIST_HEAD(&sci->sc_segbufs); INIT_LIST_HEAD(&sci->sc_write_logs); INIT_LIST_HEAD(&sci->sc_gc_inodes); - INIT_LIST_HEAD(&sci->sc_copied_buffers); init_timer(&sci->sc_timer); sci->sc_interval = HZ * NILFS_SC_DEFAULT_TIMEOUT; @@ -2695,7 +2587,7 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct super_block *sb, sci->sc_watermark = NILFS_SC_DEFAULT_WATERMARK; if (nilfs->ns_interval) - sci->sc_interval = nilfs->ns_interval; + sci->sc_interval = HZ * nilfs->ns_interval; if (nilfs->ns_watermark) sci->sc_watermark = nilfs->ns_watermark; return sci; @@ -2741,8 +2633,6 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) if (flag || !nilfs_segctor_confirm(sci)) nilfs_segctor_write_out(sci); - WARN_ON(!list_empty(&sci->sc_copied_buffers)); - if (!list_empty(&sci->sc_dirty_files)) { nilfs_warning(sci->sc_super, __func__, "dirty file(s) after the final construction\n"); diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index 6c02a86745f..38a1d001331 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -92,7 +92,6 @@ struct nilfs_segsum_pointer { * @sc_nblk_inc: Block count of current generation * @sc_dirty_files: List of files to be written * @sc_gc_inodes: List of GC inodes having blocks to be written - * @sc_copied_buffers: List of copied buffers (buffer heads) to freeze data * @sc_freesegs: array of segment numbers to be freed * @sc_nfreesegs: number of segments on @sc_freesegs * @sc_dsync_inode: inode whose data pages are written for a sync operation @@ -136,7 +135,6 @@ struct nilfs_sc_info { struct list_head sc_dirty_files; struct list_head sc_gc_inodes; - struct list_head sc_copied_buffers; __u64 *sc_freesegs; size_t sc_nfreesegs; diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c index 1d6f488ccae..2a869c35c36 100644 --- a/fs/nilfs2/sufile.c +++ b/fs/nilfs2/sufile.c @@ -30,10 +30,18 @@ #include "mdt.h" #include "sufile.h" - +/** + * struct nilfs_sufile_info - on-memory private data of sufile + * @mi: on-memory private data of metadata file + * @ncleansegs: number of clean segments + * @allocmin: lower limit of allocatable segment range + * @allocmax: upper limit of allocatable segment range + */ struct nilfs_sufile_info { struct nilfs_mdt_info mi; - unsigned long ncleansegs; + unsigned long ncleansegs;/* number of clean segments */ + __u64 allocmin; /* lower limit of allocatable segment range */ + __u64 allocmax; /* upper limit of allocatable segment range */ }; static inline struct nilfs_sufile_info *NILFS_SUI(struct inode *sufile) @@ -96,19 +104,26 @@ nilfs_sufile_get_segment_usage_block(struct inode *sufile, __u64 segnum, create, NULL, bhp); } +static int nilfs_sufile_delete_segment_usage_block(struct inode *sufile, + __u64 segnum) +{ + return nilfs_mdt_delete_block(sufile, + nilfs_sufile_get_blkoff(sufile, segnum)); +} + static void nilfs_sufile_mod_counter(struct buffer_head *header_bh, u64 ncleanadd, u64 ndirtyadd) { struct nilfs_sufile_header *header; void *kaddr; - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = kaddr + bh_offset(header_bh); le64_add_cpu(&header->sh_ncleansegs, ncleanadd); le64_add_cpu(&header->sh_ndirtysegs, ndirtyadd); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(header_bh); } /** @@ -248,6 +263,35 @@ int nilfs_sufile_update(struct inode *sufile, __u64 segnum, int create, } /** + * nilfs_sufile_set_alloc_range - limit range of segment to be allocated + * @sufile: inode of segment usage file + * @start: minimum segment number of allocatable region (inclusive) + * @end: maximum segment number of allocatable region (inclusive) + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-ERANGE - invalid segment region + */ +int nilfs_sufile_set_alloc_range(struct inode *sufile, __u64 start, __u64 end) +{ + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); + __u64 nsegs; + int ret = -ERANGE; + + down_write(&NILFS_MDT(sufile)->mi_sem); + nsegs = nilfs_sufile_get_nsegments(sufile); + + if (start <= end && end < nsegs) { + sui->allocmin = start; + sui->allocmax = end; + ret = 0; + } + up_write(&NILFS_MDT(sufile)->mi_sem); + return ret; +} + +/** * nilfs_sufile_alloc - allocate a segment * @sufile: inode of segment usage file * @segnump: pointer to segment number @@ -269,37 +313,56 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) struct buffer_head *header_bh, *su_bh; struct nilfs_sufile_header *header; struct nilfs_segment_usage *su; + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); size_t susz = NILFS_MDT(sufile)->mi_entry_size; __u64 segnum, maxsegnum, last_alloc; void *kaddr; - unsigned long nsegments, ncleansegs, nsus; - int ret, i, j; + unsigned long nsegments, ncleansegs, nsus, cnt; + int ret, j; down_write(&NILFS_MDT(sufile)->mi_sem); ret = nilfs_sufile_get_header_block(sufile, &header_bh); if (ret < 0) goto out_sem; - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = kaddr + bh_offset(header_bh); ncleansegs = le64_to_cpu(header->sh_ncleansegs); last_alloc = le64_to_cpu(header->sh_last_alloc); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nsegments = nilfs_sufile_get_nsegments(sufile); + maxsegnum = sui->allocmax; segnum = last_alloc + 1; - maxsegnum = nsegments - 1; - for (i = 0; i < nsegments; i += nsus) { - if (segnum >= nsegments) { - /* wrap around */ - segnum = 0; - maxsegnum = last_alloc; + if (segnum < sui->allocmin || segnum > sui->allocmax) + segnum = sui->allocmin; + + for (cnt = 0; cnt < nsegments; cnt += nsus) { + if (segnum > maxsegnum) { + if (cnt < sui->allocmax - sui->allocmin + 1) { + /* + * wrap around in the limited region. + * if allocation started from + * sui->allocmin, this never happens. + */ + segnum = sui->allocmin; + maxsegnum = last_alloc; + } else if (segnum > sui->allocmin && + sui->allocmax + 1 < nsegments) { + segnum = sui->allocmax + 1; + maxsegnum = nsegments - 1; + } else if (sui->allocmin > 0) { + segnum = 0; + maxsegnum = sui->allocmin - 1; + } else { + break; /* never happens */ + } } ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 1, &su_bh); if (ret < 0) goto out_header; - kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage( sufile, segnum, su_bh, kaddr); @@ -310,25 +373,25 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) continue; /* found a clean segment */ nilfs_segment_usage_set_dirty(su); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = kaddr + bh_offset(header_bh); le64_add_cpu(&header->sh_ncleansegs, -1); le64_add_cpu(&header->sh_ndirtysegs, 1); header->sh_last_alloc = cpu_to_le64(segnum); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - NILFS_SUI(sufile)->ncleansegs--; - nilfs_mdt_mark_buffer_dirty(header_bh); - nilfs_mdt_mark_buffer_dirty(su_bh); + sui->ncleansegs--; + mark_buffer_dirty(header_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); brelse(su_bh); *segnump = segnum; goto out_header; } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(su_bh); } @@ -350,21 +413,21 @@ void nilfs_sufile_do_cancel_free(struct inode *sufile, __u64 segnum, struct nilfs_segment_usage *su; void *kaddr; - kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); if (unlikely(!nilfs_segment_usage_clean(su))) { printk(KERN_WARNING "%s: segment %llu must be clean\n", __func__, (unsigned long long)segnum); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); return; } nilfs_segment_usage_set_dirty(su); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nilfs_sufile_mod_counter(header_bh, -1, 1); NILFS_SUI(sufile)->ncleansegs--; - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -376,11 +439,11 @@ void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum, void *kaddr; int clean, dirty; - kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); if (su->su_flags == cpu_to_le32(1UL << NILFS_SEGMENT_USAGE_DIRTY) && su->su_nblocks == cpu_to_le32(0)) { - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); return; } clean = nilfs_segment_usage_clean(su); @@ -390,12 +453,12 @@ void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum, su->su_lastmod = cpu_to_le64(0); su->su_nblocks = cpu_to_le32(0); su->su_flags = cpu_to_le32(1UL << NILFS_SEGMENT_USAGE_DIRTY); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); nilfs_sufile_mod_counter(header_bh, clean ? (u64)-1 : 0, dirty ? 0 : 1); NILFS_SUI(sufile)->ncleansegs -= clean; - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -407,12 +470,12 @@ void nilfs_sufile_do_free(struct inode *sufile, __u64 segnum, void *kaddr; int sudirty; - kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); if (nilfs_segment_usage_clean(su)) { printk(KERN_WARNING "%s: segment %llu is already clean\n", __func__, (unsigned long long)segnum); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); return; } WARN_ON(nilfs_segment_usage_error(su)); @@ -420,8 +483,8 @@ void nilfs_sufile_do_free(struct inode *sufile, __u64 segnum, sudirty = nilfs_segment_usage_dirty(su); nilfs_segment_usage_set_clean(su); - kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(su_bh); + kunmap_atomic(kaddr); + mark_buffer_dirty(su_bh); nilfs_sufile_mod_counter(header_bh, 1, sudirty ? (u64)-1 : 0); NILFS_SUI(sufile)->ncleansegs++; @@ -441,7 +504,7 @@ int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum) ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &bh); if (!ret) { - nilfs_mdt_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(sufile); brelse(bh); } @@ -468,15 +531,15 @@ int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, if (ret < 0) goto out_sem; - kaddr = kmap_atomic(bh->b_page, KM_USER0); + kaddr = kmap_atomic(bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, bh, kaddr); WARN_ON(nilfs_segment_usage_error(su)); if (modtime) su->su_lastmod = cpu_to_le64(modtime); su->su_nblocks = cpu_to_le32(nblocks); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); - nilfs_mdt_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(sufile); brelse(bh); @@ -505,7 +568,7 @@ int nilfs_sufile_get_stat(struct inode *sufile, struct nilfs_sustat *sustat) { struct buffer_head *header_bh; struct nilfs_sufile_header *header; - struct the_nilfs *nilfs = NILFS_I_NILFS(sufile); + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; void *kaddr; int ret; @@ -515,7 +578,7 @@ int nilfs_sufile_get_stat(struct inode *sufile, struct nilfs_sustat *sustat) if (ret < 0) goto out_sem; - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = kaddr + bh_offset(header_bh); sustat->ss_nsegs = nilfs_sufile_get_nsegments(sufile); sustat->ss_ncleansegs = le64_to_cpu(header->sh_ncleansegs); @@ -525,7 +588,7 @@ int nilfs_sufile_get_stat(struct inode *sufile, struct nilfs_sustat *sustat) spin_lock(&nilfs->ns_last_segment_lock); sustat->ss_prot_seq = nilfs->ns_prot_seq; spin_unlock(&nilfs->ns_last_segment_lock); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(header_bh); out_sem: @@ -541,22 +604,194 @@ void nilfs_sufile_do_set_error(struct inode *sufile, __u64 segnum, void *kaddr; int suclean; - kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); if (nilfs_segment_usage_error(su)) { - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); return; } suclean = nilfs_segment_usage_clean(su); nilfs_segment_usage_set_error(su); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); if (suclean) { nilfs_sufile_mod_counter(header_bh, -1, 0); NILFS_SUI(sufile)->ncleansegs--; } - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); + nilfs_mdt_mark_dirty(sufile); +} + +/** + * nilfs_sufile_truncate_range - truncate range of segment array + * @sufile: inode of segment usage file + * @start: start segment number (inclusive) + * @end: end segment number (inclusive) + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EINVAL - Invalid number of segments specified + * + * %-EBUSY - Dirty or active segments are present in the range + */ +static int nilfs_sufile_truncate_range(struct inode *sufile, + __u64 start, __u64 end) +{ + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; + struct buffer_head *header_bh; + struct buffer_head *su_bh; + struct nilfs_segment_usage *su, *su2; + size_t susz = NILFS_MDT(sufile)->mi_entry_size; + unsigned long segusages_per_block; + unsigned long nsegs, ncleaned; + __u64 segnum; + void *kaddr; + ssize_t n, nc; + int ret; + int j; + + nsegs = nilfs_sufile_get_nsegments(sufile); + + ret = -EINVAL; + if (start > end || start >= nsegs) + goto out; + + ret = nilfs_sufile_get_header_block(sufile, &header_bh); + if (ret < 0) + goto out; + + segusages_per_block = nilfs_sufile_segment_usages_per_block(sufile); + ncleaned = 0; + + for (segnum = start; segnum <= end; segnum += n) { + n = min_t(unsigned long, + segusages_per_block - + nilfs_sufile_get_offset(sufile, segnum), + end - segnum + 1); + ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, + &su_bh); + if (ret < 0) { + if (ret != -ENOENT) + goto out_header; + /* hole */ + continue; + } + kaddr = kmap_atomic(su_bh->b_page); + su = nilfs_sufile_block_get_segment_usage( + sufile, segnum, su_bh, kaddr); + su2 = su; + for (j = 0; j < n; j++, su = (void *)su + susz) { + if ((le32_to_cpu(su->su_flags) & + ~(1UL << NILFS_SEGMENT_USAGE_ERROR)) || + nilfs_segment_is_active(nilfs, segnum + j)) { + ret = -EBUSY; + kunmap_atomic(kaddr); + brelse(su_bh); + goto out_header; + } + } + nc = 0; + for (su = su2, j = 0; j < n; j++, su = (void *)su + susz) { + if (nilfs_segment_usage_error(su)) { + nilfs_segment_usage_set_clean(su); + nc++; + } + } + kunmap_atomic(kaddr); + if (nc > 0) { + mark_buffer_dirty(su_bh); + ncleaned += nc; + } + brelse(su_bh); + + if (n == segusages_per_block) { + /* make hole */ + nilfs_sufile_delete_segment_usage_block(sufile, segnum); + } + } + ret = 0; + +out_header: + if (ncleaned > 0) { + NILFS_SUI(sufile)->ncleansegs += ncleaned; + nilfs_sufile_mod_counter(header_bh, ncleaned, 0); + nilfs_mdt_mark_dirty(sufile); + } + brelse(header_bh); +out: + return ret; +} + +/** + * nilfs_sufile_resize - resize segment array + * @sufile: inode of segment usage file + * @newnsegs: new number of segments + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-ENOSPC - Enough free space is not left for shrinking + * + * %-EBUSY - Dirty or active segments exist in the region to be truncated + */ +int nilfs_sufile_resize(struct inode *sufile, __u64 newnsegs) +{ + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; + struct buffer_head *header_bh; + struct nilfs_sufile_header *header; + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); + void *kaddr; + unsigned long nsegs, nrsvsegs; + int ret = 0; + + down_write(&NILFS_MDT(sufile)->mi_sem); + + nsegs = nilfs_sufile_get_nsegments(sufile); + if (nsegs == newnsegs) + goto out; + + ret = -ENOSPC; + nrsvsegs = nilfs_nrsvsegs(nilfs, newnsegs); + if (newnsegs < nsegs && nsegs - newnsegs + nrsvsegs > sui->ncleansegs) + goto out; + + ret = nilfs_sufile_get_header_block(sufile, &header_bh); + if (ret < 0) + goto out; + + if (newnsegs > nsegs) { + sui->ncleansegs += newnsegs - nsegs; + } else /* newnsegs < nsegs */ { + ret = nilfs_sufile_truncate_range(sufile, newnsegs, nsegs - 1); + if (ret < 0) + goto out_header; + + sui->ncleansegs -= nsegs - newnsegs; + } + + kaddr = kmap_atomic(header_bh->b_page); + header = kaddr + bh_offset(header_bh); + header->sh_ncleansegs = cpu_to_le64(sui->ncleansegs); + kunmap_atomic(kaddr); + + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(sufile); + nilfs_set_nsegments(nilfs, newnsegs); + +out_header: + brelse(header_bh); +out: + up_write(&NILFS_MDT(sufile)->mi_sem); + return ret; } /** @@ -583,7 +818,7 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, struct nilfs_segment_usage *su; struct nilfs_suinfo *si = buf; size_t susz = NILFS_MDT(sufile)->mi_entry_size; - struct the_nilfs *nilfs = NILFS_I_NILFS(sufile); + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; void *kaddr; unsigned long nsegs, segusages_per_block; ssize_t n; @@ -611,7 +846,7 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, continue; } - kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage( sufile, segnum, su_bh, kaddr); for (j = 0; j < n; @@ -624,7 +859,7 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, si->sui_flags |= (1UL << NILFS_SEGMENT_USAGE_ACTIVE); } - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(su_bh); } ret = nsegs; @@ -635,6 +870,289 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, } /** + * nilfs_sufile_set_suinfo - sets segment usage info + * @sufile: inode of segment usage file + * @buf: array of suinfo_update + * @supsz: byte size of suinfo_update + * @nsup: size of suinfo_update array + * + * Description: Takes an array of nilfs_suinfo_update structs and updates + * segment usage accordingly. Only the fields indicated by the sup_flags + * are updated. + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EINVAL - Invalid values in input (segment number, flags or nblocks) + */ +ssize_t nilfs_sufile_set_suinfo(struct inode *sufile, void *buf, + unsigned supsz, size_t nsup) +{ + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; + struct buffer_head *header_bh, *bh; + struct nilfs_suinfo_update *sup, *supend = buf + supsz * nsup; + struct nilfs_segment_usage *su; + void *kaddr; + unsigned long blkoff, prev_blkoff; + int cleansi, cleansu, dirtysi, dirtysu; + long ncleaned = 0, ndirtied = 0; + int ret = 0; + + if (unlikely(nsup == 0)) + return ret; + + for (sup = buf; sup < supend; sup = (void *)sup + supsz) { + if (sup->sup_segnum >= nilfs->ns_nsegments + || (sup->sup_flags & + (~0UL << __NR_NILFS_SUINFO_UPDATE_FIELDS)) + || (nilfs_suinfo_update_nblocks(sup) && + sup->sup_sui.sui_nblocks > + nilfs->ns_blocks_per_segment)) + return -EINVAL; + } + + down_write(&NILFS_MDT(sufile)->mi_sem); + + ret = nilfs_sufile_get_header_block(sufile, &header_bh); + if (ret < 0) + goto out_sem; + + sup = buf; + blkoff = nilfs_sufile_get_blkoff(sufile, sup->sup_segnum); + ret = nilfs_mdt_get_block(sufile, blkoff, 1, NULL, &bh); + if (ret < 0) + goto out_header; + + for (;;) { + kaddr = kmap_atomic(bh->b_page); + su = nilfs_sufile_block_get_segment_usage( + sufile, sup->sup_segnum, bh, kaddr); + + if (nilfs_suinfo_update_lastmod(sup)) + su->su_lastmod = cpu_to_le64(sup->sup_sui.sui_lastmod); + + if (nilfs_suinfo_update_nblocks(sup)) + su->su_nblocks = cpu_to_le32(sup->sup_sui.sui_nblocks); + + if (nilfs_suinfo_update_flags(sup)) { + /* + * Active flag is a virtual flag projected by running + * nilfs kernel code - drop it not to write it to + * disk. + */ + sup->sup_sui.sui_flags &= + ~(1UL << NILFS_SEGMENT_USAGE_ACTIVE); + + cleansi = nilfs_suinfo_clean(&sup->sup_sui); + cleansu = nilfs_segment_usage_clean(su); + dirtysi = nilfs_suinfo_dirty(&sup->sup_sui); + dirtysu = nilfs_segment_usage_dirty(su); + + if (cleansi && !cleansu) + ++ncleaned; + else if (!cleansi && cleansu) + --ncleaned; + + if (dirtysi && !dirtysu) + ++ndirtied; + else if (!dirtysi && dirtysu) + --ndirtied; + + su->su_flags = cpu_to_le32(sup->sup_sui.sui_flags); + } + + kunmap_atomic(kaddr); + + sup = (void *)sup + supsz; + if (sup >= supend) + break; + + prev_blkoff = blkoff; + blkoff = nilfs_sufile_get_blkoff(sufile, sup->sup_segnum); + if (blkoff == prev_blkoff) + continue; + + /* get different block */ + mark_buffer_dirty(bh); + put_bh(bh); + ret = nilfs_mdt_get_block(sufile, blkoff, 1, NULL, &bh); + if (unlikely(ret < 0)) + goto out_mark; + } + mark_buffer_dirty(bh); + put_bh(bh); + + out_mark: + if (ncleaned || ndirtied) { + nilfs_sufile_mod_counter(header_bh, (u64)ncleaned, + (u64)ndirtied); + NILFS_SUI(sufile)->ncleansegs += ncleaned; + } + nilfs_mdt_mark_dirty(sufile); + out_header: + put_bh(header_bh); + out_sem: + up_write(&NILFS_MDT(sufile)->mi_sem); + return ret; +} + +/** + * nilfs_sufile_trim_fs() - trim ioctl handle function + * @sufile: inode of segment usage file + * @range: fstrim_range structure + * + * start: First Byte to trim + * len: number of Bytes to trim from start + * minlen: minimum extent length in Bytes + * + * Decription: nilfs_sufile_trim_fs goes through all segments containing bytes + * from start to start+len. start is rounded up to the next block boundary + * and start+len is rounded down. For each clean segment blkdev_issue_discard + * function is invoked. + * + * Return Value: On success, 0 is returned or negative error code, otherwise. + */ +int nilfs_sufile_trim_fs(struct inode *sufile, struct fstrim_range *range) +{ + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; + struct buffer_head *su_bh; + struct nilfs_segment_usage *su; + void *kaddr; + size_t n, i, susz = NILFS_MDT(sufile)->mi_entry_size; + sector_t seg_start, seg_end, start_block, end_block; + sector_t start = 0, nblocks = 0; + u64 segnum, segnum_end, minlen, len, max_blocks, ndiscarded = 0; + int ret = 0; + unsigned int sects_per_block; + + sects_per_block = (1 << nilfs->ns_blocksize_bits) / + bdev_logical_block_size(nilfs->ns_bdev); + len = range->len >> nilfs->ns_blocksize_bits; + minlen = range->minlen >> nilfs->ns_blocksize_bits; + max_blocks = ((u64)nilfs->ns_nsegments * nilfs->ns_blocks_per_segment); + + if (!len || range->start >= max_blocks << nilfs->ns_blocksize_bits) + return -EINVAL; + + start_block = (range->start + nilfs->ns_blocksize - 1) >> + nilfs->ns_blocksize_bits; + + /* + * range->len can be very large (actually, it is set to + * ULLONG_MAX by default) - truncate upper end of the range + * carefully so as not to overflow. + */ + if (max_blocks - start_block < len) + end_block = max_blocks - 1; + else + end_block = start_block + len - 1; + + segnum = nilfs_get_segnum_of_block(nilfs, start_block); + segnum_end = nilfs_get_segnum_of_block(nilfs, end_block); + + down_read(&NILFS_MDT(sufile)->mi_sem); + + while (segnum <= segnum_end) { + n = nilfs_sufile_segment_usages_in_block(sufile, segnum, + segnum_end); + + ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, + &su_bh); + if (ret < 0) { + if (ret != -ENOENT) + goto out_sem; + /* hole */ + segnum += n; + continue; + } + + kaddr = kmap_atomic(su_bh->b_page); + su = nilfs_sufile_block_get_segment_usage(sufile, segnum, + su_bh, kaddr); + for (i = 0; i < n; ++i, ++segnum, su = (void *)su + susz) { + if (!nilfs_segment_usage_clean(su)) + continue; + + nilfs_get_segment_range(nilfs, segnum, &seg_start, + &seg_end); + + if (!nblocks) { + /* start new extent */ + start = seg_start; + nblocks = seg_end - seg_start + 1; + continue; + } + + if (start + nblocks == seg_start) { + /* add to previous extent */ + nblocks += seg_end - seg_start + 1; + continue; + } + + /* discard previous extent */ + if (start < start_block) { + nblocks -= start_block - start; + start = start_block; + } + + if (nblocks >= minlen) { + kunmap_atomic(kaddr); + + ret = blkdev_issue_discard(nilfs->ns_bdev, + start * sects_per_block, + nblocks * sects_per_block, + GFP_NOFS, 0); + if (ret < 0) { + put_bh(su_bh); + goto out_sem; + } + + ndiscarded += nblocks; + kaddr = kmap_atomic(su_bh->b_page); + su = nilfs_sufile_block_get_segment_usage( + sufile, segnum, su_bh, kaddr); + } + + /* start new extent */ + start = seg_start; + nblocks = seg_end - seg_start + 1; + } + kunmap_atomic(kaddr); + put_bh(su_bh); + } + + + if (nblocks) { + /* discard last extent */ + if (start < start_block) { + nblocks -= start_block - start; + start = start_block; + } + if (start + nblocks > end_block + 1) + nblocks = end_block - start + 1; + + if (nblocks >= minlen) { + ret = blkdev_issue_discard(nilfs->ns_bdev, + start * sects_per_block, + nblocks * sects_per_block, + GFP_NOFS, 0); + if (!ret) + ndiscarded += nblocks; + } + } + +out_sem: + up_read(&NILFS_MDT(sufile)->mi_sem); + + range->len = ndiscarded << nilfs->ns_blocksize_bits; + return ret; +} + +/** * nilfs_sufile_read - read or get sufile inode * @sb: super block instance * @susize: size of a segment usage entry @@ -651,6 +1169,18 @@ int nilfs_sufile_read(struct super_block *sb, size_t susize, void *kaddr; int err; + if (susize > sb->s_blocksize) { + printk(KERN_ERR + "NILFS: too large segment usage size: %zu bytes.\n", + susize); + return -EINVAL; + } else if (susize < NILFS_MIN_SEGMENT_USAGE_SIZE) { + printk(KERN_ERR + "NILFS: too small segment usage size: %zu bytes.\n", + susize); + return -EINVAL; + } + sufile = nilfs_iget_locked(sb, NULL, NILFS_SUFILE_INO); if (unlikely(!sufile)) return -ENOMEM; @@ -673,12 +1203,15 @@ int nilfs_sufile_read(struct super_block *sb, size_t susize, goto failed; sui = NILFS_SUI(sufile); - kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + kaddr = kmap_atomic(header_bh->b_page); header = kaddr + bh_offset(header_bh); sui->ncleansegs = le64_to_cpu(header->sh_ncleansegs); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); brelse(header_bh); + sui->allocmax = nilfs_sufile_get_nsegments(sufile) - 1; + sui->allocmin = 0; + unlock_new_inode(sufile); out: *inodep = sufile; diff --git a/fs/nilfs2/sufile.h b/fs/nilfs2/sufile.h index a943fbacb45..b8afd72f237 100644 --- a/fs/nilfs2/sufile.h +++ b/fs/nilfs2/sufile.h @@ -31,11 +31,12 @@ static inline unsigned long nilfs_sufile_get_nsegments(struct inode *sufile) { - return NILFS_I_NILFS(sufile)->ns_nsegments; + return ((struct the_nilfs *)sufile->i_sb->s_fs_info)->ns_nsegments; } unsigned long nilfs_sufile_get_ncleansegs(struct inode *sufile); +int nilfs_sufile_set_alloc_range(struct inode *sufile, __u64 start, __u64 end); int nilfs_sufile_alloc(struct inode *, __u64 *); int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum); int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, @@ -43,6 +44,7 @@ int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, int nilfs_sufile_get_stat(struct inode *, struct nilfs_sustat *); ssize_t nilfs_sufile_get_suinfo(struct inode *, __u64, void *, unsigned, size_t); +ssize_t nilfs_sufile_set_suinfo(struct inode *, void *, unsigned , size_t); int nilfs_sufile_updatev(struct inode *, __u64 *, size_t, int, size_t *, void (*dofunc)(struct inode *, __u64, @@ -61,8 +63,10 @@ void nilfs_sufile_do_cancel_free(struct inode *, __u64, struct buffer_head *, void nilfs_sufile_do_set_error(struct inode *, __u64, struct buffer_head *, struct buffer_head *); +int nilfs_sufile_resize(struct inode *sufile, __u64 newnsegs); int nilfs_sufile_read(struct super_block *sb, size_t susize, struct nilfs_inode *raw_inode, struct inode **inodep); +int nilfs_sufile_trim_fs(struct inode *sufile, struct fstrim_range *range); /** * nilfs_sufile_scrap - make a segment garbage diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 062cca06519..8c532b2ca3a 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -56,6 +56,7 @@ #include "btnode.h" #include "page.h" #include "cpfile.h" +#include "sufile.h" /* nilfs_sufile_resize(), nilfs_sufile_set_alloc_range() */ #include "ifile.h" #include "dat.h" #include "segment.h" @@ -165,7 +166,7 @@ struct inode *nilfs_alloc_inode(struct super_block *sb) ii->i_state = 0; ii->i_cno = 0; ii->vfs_inode.i_version = 1; - nilfs_btnode_cache_init(&ii->i_btnode_cache, sb->s_bdi); + nilfs_mapping_init(&ii->i_btnode_cache, &ii->vfs_inode, sb->s_bdi); return &ii->vfs_inode; } @@ -174,8 +175,6 @@ static void nilfs_i_callback(struct rcu_head *head) struct inode *inode = container_of(head, struct inode, i_rcu); struct nilfs_mdt_info *mdi = NILFS_MDT(inode); - INIT_LIST_HEAD(&inode->i_dentry); - if (mdi) { kfree(mdi->mi_bgl); /* kfree(NULL) is safe */ kfree(mdi); @@ -347,6 +346,134 @@ int nilfs_cleanup_super(struct super_block *sb) return ret; } +/** + * nilfs_move_2nd_super - relocate secondary super block + * @sb: super block instance + * @sb2off: new offset of the secondary super block (in bytes) + */ +static int nilfs_move_2nd_super(struct super_block *sb, loff_t sb2off) +{ + struct the_nilfs *nilfs = sb->s_fs_info; + struct buffer_head *nsbh; + struct nilfs_super_block *nsbp; + sector_t blocknr, newblocknr; + unsigned long offset; + int sb2i = -1; /* array index of the secondary superblock */ + int ret = 0; + + /* nilfs->ns_sem must be locked by the caller. */ + if (nilfs->ns_sbh[1] && + nilfs->ns_sbh[1]->b_blocknr > nilfs->ns_first_data_block) { + sb2i = 1; + blocknr = nilfs->ns_sbh[1]->b_blocknr; + } else if (nilfs->ns_sbh[0]->b_blocknr > nilfs->ns_first_data_block) { + sb2i = 0; + blocknr = nilfs->ns_sbh[0]->b_blocknr; + } + if (sb2i >= 0 && (u64)blocknr << nilfs->ns_blocksize_bits == sb2off) + goto out; /* super block location is unchanged */ + + /* Get new super block buffer */ + newblocknr = sb2off >> nilfs->ns_blocksize_bits; + offset = sb2off & (nilfs->ns_blocksize - 1); + nsbh = sb_getblk(sb, newblocknr); + if (!nsbh) { + printk(KERN_WARNING + "NILFS warning: unable to move secondary superblock " + "to block %llu\n", (unsigned long long)newblocknr); + ret = -EIO; + goto out; + } + nsbp = (void *)nsbh->b_data + offset; + memset(nsbp, 0, nilfs->ns_blocksize); + + if (sb2i >= 0) { + memcpy(nsbp, nilfs->ns_sbp[sb2i], nilfs->ns_sbsize); + brelse(nilfs->ns_sbh[sb2i]); + nilfs->ns_sbh[sb2i] = nsbh; + nilfs->ns_sbp[sb2i] = nsbp; + } else if (nilfs->ns_sbh[0]->b_blocknr < nilfs->ns_first_data_block) { + /* secondary super block will be restored to index 1 */ + nilfs->ns_sbh[1] = nsbh; + nilfs->ns_sbp[1] = nsbp; + } else { + brelse(nsbh); + } +out: + return ret; +} + +/** + * nilfs_resize_fs - resize the filesystem + * @sb: super block instance + * @newsize: new size of the filesystem (in bytes) + */ +int nilfs_resize_fs(struct super_block *sb, __u64 newsize) +{ + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_super_block **sbp; + __u64 devsize, newnsegs; + loff_t sb2off; + int ret; + + ret = -ERANGE; + devsize = i_size_read(sb->s_bdev->bd_inode); + if (newsize > devsize) + goto out; + + /* + * Write lock is required to protect some functions depending + * on the number of segments, the number of reserved segments, + * and so forth. + */ + down_write(&nilfs->ns_segctor_sem); + + sb2off = NILFS_SB2_OFFSET_BYTES(newsize); + newnsegs = sb2off >> nilfs->ns_blocksize_bits; + do_div(newnsegs, nilfs->ns_blocks_per_segment); + + ret = nilfs_sufile_resize(nilfs->ns_sufile, newnsegs); + up_write(&nilfs->ns_segctor_sem); + if (ret < 0) + goto out; + + ret = nilfs_construct_segment(sb); + if (ret < 0) + goto out; + + down_write(&nilfs->ns_sem); + nilfs_move_2nd_super(sb, sb2off); + ret = -EIO; + sbp = nilfs_prepare_super(sb, 0); + if (likely(sbp)) { + nilfs_set_log_cursor(sbp[0], nilfs); + /* + * Drop NILFS_RESIZE_FS flag for compatibility with + * mount-time resize which may be implemented in a + * future release. + */ + sbp[0]->s_state = cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & + ~NILFS_RESIZE_FS); + sbp[0]->s_dev_size = cpu_to_le64(newsize); + sbp[0]->s_nsegments = cpu_to_le64(nilfs->ns_nsegments); + if (sbp[1]) + memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); + ret = nilfs_commit_super(sb, NILFS_SB_COMMIT_ALL); + } + up_write(&nilfs->ns_sem); + + /* + * Reset the range of allocatable segments last. This order + * is important in the case of expansion because the secondary + * superblock must be protected from log write until migration + * completes. + */ + if (!ret) + nilfs_sufile_set_alloc_range(nilfs->ns_sufile, 0, newnsegs - 1); +out: + return ret; +} + static void nilfs_put_super(struct super_block *sb) { struct the_nilfs *nilfs = sb->s_fs_info; @@ -427,8 +554,10 @@ int nilfs_attach_checkpoint(struct super_block *sb, __u64 cno, int curr_mnt, if (err) goto failed_bh; - atomic_set(&root->inodes_count, le64_to_cpu(raw_cp->cp_inodes_count)); - atomic_set(&root->blocks_count, le64_to_cpu(raw_cp->cp_blocks_count)); + atomic64_set(&root->inodes_count, + le64_to_cpu(raw_cp->cp_inodes_count)); + atomic64_set(&root->blocks_count, + le64_to_cpu(raw_cp->cp_blocks_count)); nilfs_cpfile_put_checkpoint(nilfs->ns_cpfile, cno, bh_cp); @@ -482,6 +611,7 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf) unsigned long overhead; unsigned long nrsvblocks; sector_t nfreeblocks; + u64 nmaxinodes, nfreeinodes; int err; /* @@ -506,14 +636,34 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf) if (unlikely(err)) return err; + err = nilfs_ifile_count_free_inodes(root->ifile, + &nmaxinodes, &nfreeinodes); + if (unlikely(err)) { + printk(KERN_WARNING + "NILFS warning: fail to count free inodes: err %d.\n", + err); + if (err == -ERANGE) { + /* + * If nilfs_palloc_count_max_entries() returns + * -ERANGE error code then we simply treat + * curent inodes count as maximum possible and + * zero as free inodes value. + */ + nmaxinodes = atomic64_read(&root->inodes_count); + nfreeinodes = 0; + err = 0; + } else + return err; + } + buf->f_type = NILFS_SUPER_MAGIC; buf->f_bsize = sb->s_blocksize; buf->f_blocks = blocks - overhead; buf->f_bfree = nfreeblocks; buf->f_bavail = (buf->f_bfree >= nrsvblocks) ? (buf->f_bfree - nrsvblocks) : 0; - buf->f_files = atomic_read(&root->inodes_count); - buf->f_ffree = 0; /* nilfs_count_free_inodes(sb); */ + buf->f_files = nmaxinodes; + buf->f_ffree = nfreeinodes; buf->f_namelen = NILFS_NAME_LEN; buf->f_fsid.val[0] = (u32)id; buf->f_fsid.val[1] = (u32)(id >> 32); @@ -521,11 +671,11 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf) return 0; } -static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs) +static int nilfs_show_options(struct seq_file *seq, struct dentry *dentry) { - struct super_block *sb = vfs->mnt_sb; + struct super_block *sb = dentry->d_sb; struct the_nilfs *nilfs = sb->s_fs_info; - struct nilfs_root *root = NILFS_I(vfs->mnt_root->d_inode)->i_root; + struct nilfs_root *root = NILFS_I(dentry->d_inode)->i_root; if (!nilfs_test_opt(nilfs, BARRIER)) seq_puts(seq, ",nobarrier"); @@ -549,20 +699,13 @@ static const struct super_operations nilfs_sops = { .alloc_inode = nilfs_alloc_inode, .destroy_inode = nilfs_destroy_inode, .dirty_inode = nilfs_dirty_inode, - /* .write_inode = nilfs_write_inode, */ - /* .put_inode = nilfs_put_inode, */ - /* .drop_inode = nilfs_drop_inode, */ .evict_inode = nilfs_evict_inode, .put_super = nilfs_put_super, - /* .write_super = nilfs_write_super, */ .sync_fs = nilfs_sync_fs, .freeze_fs = nilfs_freeze, .unfreeze_fs = nilfs_unfreeze, - /* .write_super_lockfs */ - /* .unlockfs */ .statfs = nilfs_statfs, .remount_fs = nilfs_remount, - /* .umount_begin */ .show_options = nilfs_show_options }; @@ -790,9 +933,8 @@ static int nilfs_get_root_dentry(struct super_block *sb, if (root->cno == NILFS_CPTREE_CURRENT_CNO) { dentry = d_find_alias(inode); if (!dentry) { - dentry = d_alloc_root(inode); + dentry = d_make_root(inode); if (!dentry) { - iput(inode); ret = -ENOMEM; goto failed_dentry; } @@ -822,6 +964,8 @@ static int nilfs_attach_snapshot(struct super_block *s, __u64 cno, struct nilfs_root *root; int ret; + mutex_lock(&nilfs->ns_snapshot_mount_mutex); + down_read(&nilfs->ns_segctor_sem); ret = nilfs_cpfile_is_snapshot(nilfs->ns_cpfile, cno); up_read(&nilfs->ns_segctor_sem); @@ -846,26 +990,20 @@ static int nilfs_attach_snapshot(struct super_block *s, __u64 cno, ret = nilfs_get_root_dentry(s, root, root_dentry); nilfs_put_root(root); out: + mutex_unlock(&nilfs->ns_snapshot_mount_mutex); return ret; } -static int nilfs_tree_was_touched(struct dentry *root_dentry) -{ - return root_dentry->d_count > 1; -} - /** - * nilfs_try_to_shrink_tree() - try to shrink dentries of a checkpoint + * nilfs_tree_is_busy() - try to shrink dentries of a checkpoint * @root_dentry: root dentry of the tree to be shrunk * * This function returns true if the tree was in-use. */ -static int nilfs_try_to_shrink_tree(struct dentry *root_dentry) +static bool nilfs_tree_is_busy(struct dentry *root_dentry) { - if (have_submounts(root_dentry)) - return true; shrink_dcache_parent(root_dentry); - return nilfs_tree_was_touched(root_dentry); + return d_count(root_dentry) > 1; } int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno) @@ -889,8 +1027,7 @@ int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno) if (inode) { dentry = d_find_alias(inode); if (dentry) { - if (nilfs_tree_was_touched(dentry)) - ret = nilfs_try_to_shrink_tree(dentry); + ret = nilfs_tree_is_busy(dentry); dput(dentry); } iput(inode); @@ -932,6 +1069,7 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_export_op = &nilfs_export_ops; sb->s_root = NULL; sb->s_time_gran = 1; + sb->s_max_links = NILFS_LINK_MAX; bdi = sb->s_bdev->bd_inode->i_mapping->backing_dev_info; sb->s_bdi = bdi ? : &default_backing_dev_info; @@ -991,6 +1129,7 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) unsigned long old_mount_opt; int err; + sync_filesystem(sb); old_sb_flags = sb->s_flags; old_mount_opt = nilfs->ns_mount_opt; @@ -1161,7 +1300,8 @@ nilfs_mount(struct file_system_type *fs_type, int flags, err = -EBUSY; goto failed; } - s = sget(fs_type, nilfs_test_bdev_super, nilfs_set_bdev_super, sd.bdev); + s = sget(fs_type, nilfs_test_bdev_super, nilfs_set_bdev_super, flags, + sd.bdev); mutex_unlock(&sd.bdev->bd_fsfreeze_mutex); if (IS_ERR(s)) { err = PTR_ERR(s); @@ -1174,7 +1314,6 @@ nilfs_mount(struct file_system_type *fs_type, int flags, s_new = true; /* New superblock instance created */ - s->s_flags = flags; s->s_mode = mode; strlcpy(s->s_id, bdevname(sd.bdev, b), sizeof(s->s_id)); sb_set_blocksize(s, block_size(sd.bdev)); @@ -1185,11 +1324,8 @@ nilfs_mount(struct file_system_type *fs_type, int flags, s->s_flags |= MS_ACTIVE; } else if (!sd.cno) { - int busy = false; - - if (nilfs_tree_was_touched(s->s_root)) { - busy = nilfs_try_to_shrink_tree(s->s_root); - if (busy && (flags ^ s->s_flags) & MS_RDONLY) { + if (nilfs_tree_is_busy(s->s_root)) { + if ((flags ^ s->s_flags) & MS_RDONLY) { printk(KERN_ERR "NILFS: the device already " "has a %s mount.\n", (s->s_flags & MS_RDONLY) ? @@ -1197,8 +1333,7 @@ nilfs_mount(struct file_system_type *fs_type, int flags, err = -EBUSY; goto failed_super; } - } - if (!busy) { + } else { /* * Try remount to setup mount states if the current * tree is not mounted and only snapshots use this sb. @@ -1238,6 +1373,7 @@ struct file_system_type nilfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("nilfs2"); static void nilfs_inode_init_once(void *obj) { @@ -1259,6 +1395,12 @@ static void nilfs_segbuf_init_once(void *obj) static void nilfs_destroy_cachep(void) { + /* + * Make sure all delayed rcu free inodes are flushed before we + * destroy cache. + */ + rcu_barrier(); + if (nilfs_inode_cachep) kmem_cache_destroy(nilfs_inode_cachep); if (nilfs_transaction_cachep) diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index d2acd1a651f..8ba8229ba07 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -76,6 +76,7 @@ struct the_nilfs *alloc_nilfs(struct block_device *bdev) nilfs->ns_bdev = bdev; atomic_set(&nilfs->ns_ndirtyblks, 0); init_rwsem(&nilfs->ns_sem); + mutex_init(&nilfs->ns_snapshot_mount_mutex); INIT_LIST_HEAD(&nilfs->ns_dirty_files); INIT_LIST_HEAD(&nilfs->ns_gc_inodes); spin_lock_init(&nilfs->ns_inode_lock); @@ -363,6 +364,24 @@ static unsigned long long nilfs_max_size(unsigned int blkbits) return res; } +/** + * nilfs_nrsvsegs - calculate the number of reserved segments + * @nilfs: nilfs object + * @nsegs: total number of segments + */ +unsigned long nilfs_nrsvsegs(struct the_nilfs *nilfs, unsigned long nsegs) +{ + return max_t(unsigned long, NILFS_MIN_NRSVSEGS, + DIV_ROUND_UP(nsegs * nilfs->ns_r_segments_percentage, + 100)); +} + +void nilfs_set_nsegments(struct the_nilfs *nilfs, unsigned long nsegs) +{ + nilfs->ns_nsegments = nsegs; + nilfs->ns_nrsvsegs = nilfs_nrsvsegs(nilfs, nsegs); +} + static int nilfs_store_disk_layout(struct the_nilfs *nilfs, struct nilfs_super_block *sbp) { @@ -380,6 +399,16 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, return -EINVAL; nilfs->ns_inode_size = le16_to_cpu(sbp->s_inode_size); + if (nilfs->ns_inode_size > nilfs->ns_blocksize) { + printk(KERN_ERR "NILFS: too large inode size: %d bytes.\n", + nilfs->ns_inode_size); + return -EINVAL; + } else if (nilfs->ns_inode_size < NILFS_MIN_INODE_SIZE) { + printk(KERN_ERR "NILFS: too small inode size: %d bytes.\n", + nilfs->ns_inode_size); + return -EINVAL; + } + nilfs->ns_first_ino = le32_to_cpu(sbp->s_first_ino); nilfs->ns_blocks_per_segment = le32_to_cpu(sbp->s_blocks_per_segment); @@ -389,13 +418,15 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, } nilfs->ns_first_data_block = le64_to_cpu(sbp->s_first_data_block); - nilfs->ns_nsegments = le64_to_cpu(sbp->s_nsegments); nilfs->ns_r_segments_percentage = le32_to_cpu(sbp->s_r_segments_percentage); - nilfs->ns_nrsvsegs = - max_t(unsigned long, NILFS_MIN_NRSVSEGS, - DIV_ROUND_UP(nilfs->ns_nsegments * - nilfs->ns_r_segments_percentage, 100)); + if (nilfs->ns_r_segments_percentage < 1 || + nilfs->ns_r_segments_percentage > 99) { + printk(KERN_ERR "NILFS: invalid reserved segments percentage.\n"); + return -EINVAL; + } + + nilfs_set_nsegments(nilfs, le64_to_cpu(sbp->s_nsegments)); nilfs->ns_crc_seed = le32_to_cpu(sbp->s_crc_seed); return 0; } @@ -501,6 +532,7 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, brelse(sbh[1]); sbh[1] = NULL; sbp[1] = NULL; + valid[1] = 0; swp = 0; } if (!valid[swp]) { @@ -742,8 +774,8 @@ nilfs_find_or_create_root(struct the_nilfs *nilfs, __u64 cno) new->ifile = NULL; new->nilfs = nilfs; atomic_set(&new->count, 1); - atomic_set(&new->inodes_count, 0); - atomic_set(&new->blocks_count, 0); + atomic64_set(&new->inodes_count, 0); + atomic64_set(&new->blocks_count, 0); rb_link_node(&new->rb_node, parent, p); rb_insert_color(&new->rb_node, &nilfs->ns_cptree); diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index f4968145c2a..de8cc53b4a5 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -47,11 +47,13 @@ enum { * @ns_flags: flags * @ns_bdev: block device * @ns_sem: semaphore for shared states + * @ns_snapshot_mount_mutex: mutex to protect snapshot mounts * @ns_sbh: buffer heads of on-disk super blocks * @ns_sbp: pointers to super block data * @ns_sbwtime: previous write time of super block * @ns_sbwcount: write count of super block * @ns_sbsize: size of valid data in super block + * @ns_mount_state: file system state * @ns_seg_seq: segment sequence counter * @ns_segnum: index number of the latest full segment. * @ns_nextnum: index number of the full segment index to be used next @@ -99,13 +101,12 @@ struct the_nilfs { struct block_device *ns_bdev; struct rw_semaphore ns_sem; + struct mutex ns_snapshot_mount_mutex; /* * used for * - loading the latest checkpoint exclusively. * - allocating a new full segment. - * - protecting s_dirt in the super_block struct - * (see nilfs_write_super) and the following fields. */ struct buffer_head *ns_sbh[2]; struct nilfs_super_block *ns_sbp[2]; @@ -229,9 +230,8 @@ THE_NILFS_FNS(SB_DIRTY, sb_dirty) * @count: refcount of this structure * @nilfs: nilfs object * @ifile: inode file - * @root: root inode * @inodes_count: number of inodes - * @blocks_count: number of blocks (Reserved) + * @blocks_count: number of blocks */ struct nilfs_root { __u64 cno; @@ -241,8 +241,8 @@ struct nilfs_root { struct the_nilfs *nilfs; struct inode *ifile; - atomic_t inodes_count; - atomic_t blocks_count; + atomic64_t inodes_count; + atomic64_t blocks_count; }; /* Special checkpoint number */ @@ -268,6 +268,8 @@ struct the_nilfs *alloc_nilfs(struct block_device *bdev); void destroy_nilfs(struct the_nilfs *nilfs); int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data); int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb); +unsigned long nilfs_nrsvsegs(struct the_nilfs *nilfs, unsigned long nsegs); +void nilfs_set_nsegments(struct the_nilfs *nilfs, unsigned long nsegs); int nilfs_discard_segments(struct the_nilfs *, __u64 *, size_t); int nilfs_count_free_blocks(struct the_nilfs *, sector_t *); struct nilfs_root *nilfs_lookup_root(struct the_nilfs *nilfs, __u64 cno); |