diff --git a/fs/ext4/Makefile b/fs/ext4/Makefile index 04109460ba9e..56fd8f865930 100644 --- a/fs/ext4/Makefile +++ b/fs/ext4/Makefile @@ -7,7 +7,7 @@ obj-$(CONFIG_EXT4_FS) += ext4.o ext4-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o page-io.o \ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o \ ext4_jbd2.o migrate.o mballoc.o block_validity.o move_extent.o \ - mmp.o + mmp.o indirect.o ext4-$(CONFIG_EXT4_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o ext4-$(CONFIG_EXT4_FS_POSIX_ACL) += acl.o diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 264f6949511e..f8224adf496e 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -620,3 +620,51 @@ unsigned long ext4_bg_num_gdb(struct super_block *sb, ext4_group_t group) } +/** + * ext4_inode_to_goal_block - return a hint for block allocation + * @inode: inode for block allocation + * + * Return the ideal location to start allocating blocks for a + * newly created inode. + */ +ext4_fsblk_t ext4_inode_to_goal_block(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + ext4_group_t block_group; + ext4_grpblk_t colour; + int flex_size = ext4_flex_bg_size(EXT4_SB(inode->i_sb)); + ext4_fsblk_t bg_start; + ext4_fsblk_t last_block; + + block_group = ei->i_block_group; + if (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) { + /* + * If there are at least EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME + * block groups per flexgroup, reserve the first block + * group for directories and special files. Regular + * files will start at the second block group. This + * tends to speed up directory access and improves + * fsck times. + */ + block_group &= ~(flex_size-1); + if (S_ISREG(inode->i_mode)) + block_group++; + } + bg_start = ext4_group_first_block_no(inode->i_sb, block_group); + last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1; + + /* + * If we are doing delayed allocation, we don't need take + * colour into account. + */ + if (test_opt(inode->i_sb, DELALLOC)) + return bg_start; + + if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block) + colour = (current->pid % 16) * + (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16); + else + colour = (current->pid % 16) * ((last_block - bg_start) / 16); + return bg_start + colour; +} + diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index fac90f3fba80..8efb2f0a3447 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -246,3 +246,24 @@ int ext4_data_block_valid(struct ext4_sb_info *sbi, ext4_fsblk_t start_blk, return 1; } +int ext4_check_blockref(const char *function, unsigned int line, + struct inode *inode, __le32 *p, unsigned int max) +{ + struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es; + __le32 *bref = p; + unsigned int blk; + + while (bref < p+max) { + blk = le32_to_cpu(*bref++); + if (blk && + unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb), + blk, 1))) { + es->s_last_error_block = cpu_to_le64(blk); + ext4_error_inode(inode, function, line, blk, + "invalid block"); + return -EIO; + } + } + return 0; +} + diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index fa44df879711..e717dfd2f2b4 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -526,6 +526,7 @@ struct ext4_new_group_data { #define EXT4_FREE_BLOCKS_METADATA 0x0001 #define EXT4_FREE_BLOCKS_FORGET 0x0002 #define EXT4_FREE_BLOCKS_VALIDATED 0x0004 +#define EXT4_FREE_BLOCKS_NO_QUOT_UPDATE 0x0008 /* * ioctl commands @@ -939,6 +940,8 @@ struct ext4_inode_info { #define ext4_find_next_zero_bit find_next_zero_bit_le #define ext4_find_next_bit find_next_bit_le +extern void ext4_set_bits(void *bm, int cur, int len); + /* * Maximal mount counts between two filesystem checks */ @@ -1126,7 +1129,8 @@ struct ext4_sb_info { struct journal_s *s_journal; struct list_head s_orphan; struct mutex s_orphan_lock; - struct mutex s_resize_lock; + unsigned long s_resize_flags; /* Flags indicating if there + is a resizer */ unsigned long s_commit_interval; u32 s_max_batch_time; u32 s_min_batch_time; @@ -1214,6 +1218,9 @@ struct ext4_sb_info { /* Kernel thread for multiple mount protection */ struct task_struct *s_mmp_tsk; + + /* record the last minlen when FITRIM is called. */ + atomic_t s_last_trim_minblks; }; static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb) @@ -1743,6 +1750,7 @@ extern unsigned ext4_init_block_bitmap(struct super_block *sb, struct ext4_group_desc *desc); #define ext4_free_blocks_after_init(sb, group, desc) \ ext4_init_block_bitmap(sb, NULL, group, desc) +ext4_fsblk_t ext4_inode_to_goal_block(struct inode *); /* dir.c */ extern int __ext4_check_dir_entry(const char *, unsigned int, struct inode *, @@ -1793,7 +1801,7 @@ extern void ext4_free_blocks(handle_t *handle, struct inode *inode, unsigned long count, int flags); extern int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t i, struct ext4_group_desc *desc); -extern void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, +extern int ext4_group_add_blocks(handle_t *handle, struct super_block *sb, ext4_fsblk_t block, unsigned long count); extern int ext4_trim_fs(struct super_block *, struct fstrim_range *); @@ -1834,6 +1842,17 @@ extern int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf); extern qsize_t *ext4_get_reserved_space(struct inode *inode); extern void ext4_da_update_reserve_space(struct inode *inode, int used, int quota_claim); + +/* indirect.c */ +extern int ext4_ind_map_blocks(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, int flags); +extern ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb, + const struct iovec *iov, loff_t offset, + unsigned long nr_segs); +extern int ext4_ind_calc_metadata_amount(struct inode *inode, sector_t lblock); +extern int ext4_ind_trans_blocks(struct inode *inode, int nrblocks, int chunk); +extern void ext4_ind_truncate(struct inode *inode); + /* ioctl.c */ extern long ext4_ioctl(struct file *, unsigned int, unsigned long); extern long ext4_compat_ioctl(struct file *, unsigned int, unsigned long); @@ -1855,6 +1874,9 @@ extern int ext4_group_extend(struct super_block *sb, ext4_fsblk_t n_blocks_count); /* super.c */ +extern void *ext4_kvmalloc(size_t size, gfp_t flags); +extern void *ext4_kvzalloc(size_t size, gfp_t flags); +extern void ext4_kvfree(void *ptr); extern void __ext4_error(struct super_block *, const char *, unsigned int, const char *, ...) __attribute__ ((format (printf, 4, 5))); @@ -2067,11 +2089,19 @@ struct ext4_group_info { * 5 free 8-block regions. */ }; -#define EXT4_GROUP_INFO_NEED_INIT_BIT 0 +#define EXT4_GROUP_INFO_NEED_INIT_BIT 0 +#define EXT4_GROUP_INFO_WAS_TRIMMED_BIT 1 #define EXT4_MB_GRP_NEED_INIT(grp) \ (test_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &((grp)->bb_state))) +#define EXT4_MB_GRP_WAS_TRIMMED(grp) \ + (test_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state))) +#define EXT4_MB_GRP_SET_TRIMMED(grp) \ + (set_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state))) +#define EXT4_MB_GRP_CLEAR_TRIMMED(grp) \ + (clear_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state))) + #define EXT4_MAX_CONTENTION 8 #define EXT4_CONTENTION_THRESHOLD 2 @@ -2122,6 +2152,19 @@ static inline void ext4_mark_super_dirty(struct super_block *sb) sb->s_dirt =1; } +/* + * Block validity checking + */ +#define ext4_check_indirect_blockref(inode, bh) \ + ext4_check_blockref(__func__, __LINE__, inode, \ + (__le32 *)(bh)->b_data, \ + EXT4_ADDR_PER_BLOCK((inode)->i_sb)) + +#define ext4_ind_check_inode(inode) \ + ext4_check_blockref(__func__, __LINE__, inode, \ + EXT4_I(inode)->i_data, \ + EXT4_NDIR_BLOCKS) + /* * Inodes and files operations */ @@ -2151,6 +2194,8 @@ extern void ext4_exit_system_zone(void); extern int ext4_data_block_valid(struct ext4_sb_info *sbi, ext4_fsblk_t start_blk, unsigned int count); +extern int ext4_check_blockref(const char *, unsigned int, + struct inode *, __le32 *, unsigned int); /* extents.c */ extern int ext4_ext_tree_init(handle_t *handle, struct inode *); @@ -2230,6 +2275,10 @@ static inline void set_bitmap_uptodate(struct buffer_head *bh) extern wait_queue_head_t ext4__ioend_wq[EXT4_WQ_HASH_SZ]; extern struct mutex ext4__aio_mutex[EXT4_WQ_HASH_SZ]; +#define EXT4_RESIZING 0 +extern int ext4_resize_begin(struct super_block *sb); +extern void ext4_resize_end(struct super_block *sb); + #endif /* __KERNEL__ */ #endif /* _EXT4_H */ diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index f815cc81e7a2..57cf568a98ab 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -114,12 +114,6 @@ static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode, struct ext4_ext_path *path, ext4_lblk_t block) { - struct ext4_inode_info *ei = EXT4_I(inode); - ext4_fsblk_t bg_start; - ext4_fsblk_t last_block; - ext4_grpblk_t colour; - ext4_group_t block_group; - int flex_size = ext4_flex_bg_size(EXT4_SB(inode->i_sb)); int depth; if (path) { @@ -161,36 +155,7 @@ static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode, } /* OK. use inode's group */ - block_group = ei->i_block_group; - if (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) { - /* - * If there are at least EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME - * block groups per flexgroup, reserve the first block - * group for directories and special files. Regular - * files will start at the second block group. This - * tends to speed up directory access and improves - * fsck times. - */ - block_group &= ~(flex_size-1); - if (S_ISREG(inode->i_mode)) - block_group++; - } - bg_start = ext4_group_first_block_no(inode->i_sb, block_group); - last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1; - - /* - * If we are doing delayed allocation, we don't need take - * colour into account. - */ - if (test_opt(inode->i_sb, DELALLOC)) - return bg_start; - - if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block) - colour = (current->pid % 16) * - (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16); - else - colour = (current->pid % 16) * ((last_block - bg_start) / 16); - return bg_start + colour + block; + return ext4_inode_to_goal_block(inode); } /* @@ -776,6 +741,16 @@ static int ext4_ext_insert_index(handle_t *handle, struct inode *inode, logical, le32_to_cpu(curp->p_idx->ei_block)); return -EIO; } + + if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries) + >= le16_to_cpu(curp->p_hdr->eh_max))) { + EXT4_ERROR_INODE(inode, + "eh_entries %d >= eh_max %d!", + le16_to_cpu(curp->p_hdr->eh_entries), + le16_to_cpu(curp->p_hdr->eh_max)); + return -EIO; + } + len = EXT_MAX_INDEX(curp->p_hdr) - curp->p_idx; if (logical > le32_to_cpu(curp->p_idx->ei_block)) { /* insert after */ @@ -805,13 +780,6 @@ static int ext4_ext_insert_index(handle_t *handle, struct inode *inode, ext4_idx_store_pblock(ix, ptr); le16_add_cpu(&curp->p_hdr->eh_entries, 1); - if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries) - > le16_to_cpu(curp->p_hdr->eh_max))) { - EXT4_ERROR_INODE(inode, - "logical %d == ei_block %d!", - logical, le32_to_cpu(curp->p_idx->ei_block)); - return -EIO; - } if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) { EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!"); return -EIO; @@ -1446,8 +1414,7 @@ ext4_ext_next_allocated_block(struct ext4_ext_path *path) * ext4_ext_next_leaf_block: * returns first allocated block from next leaf or EXT_MAX_BLOCKS */ -static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode, - struct ext4_ext_path *path) +static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path) { int depth; @@ -1757,7 +1724,6 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode, goto merge; } -repeat: depth = ext_depth(inode); eh = path[depth].p_hdr; if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) @@ -1765,9 +1731,10 @@ repeat: /* probably next leaf has space for us? */ fex = EXT_LAST_EXTENT(eh); - next = ext4_ext_next_leaf_block(inode, path); - if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block) - && next != EXT_MAX_BLOCKS) { + next = EXT_MAX_BLOCKS; + if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)) + next = ext4_ext_next_leaf_block(path); + if (next != EXT_MAX_BLOCKS) { ext_debug("next leaf block - %d\n", next); BUG_ON(npath != NULL); npath = ext4_ext_find_extent(inode, next, NULL); @@ -1779,7 +1746,7 @@ repeat: ext_debug("next leaf isn't full(%d)\n", le16_to_cpu(eh->eh_entries)); path = npath; - goto repeat; + goto has_space; } ext_debug("next leaf has no free space(%d,%d)\n", le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max)); @@ -1839,7 +1806,7 @@ has_space: ext4_ext_pblock(newext), ext4_ext_is_uninitialized(newext), ext4_ext_get_actual_len(newext), - nearex, len, nearex + 1, nearex + 2); + nearex, len, nearex, nearex + 1); memmove(nearex + 1, nearex, len); path[depth].p_ext = nearex; } @@ -2052,7 +2019,7 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, } /* - * ext4_ext_in_cache() + * ext4_ext_check_cache() * Checks to see if the given block is in the cache. * If it is, the cached extent is stored in the given * cache extent pointer. If the cached extent is a hole, @@ -2134,8 +2101,6 @@ ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block, /* * ext4_ext_rm_idx: * removes index from the index block. - * It's used in truncate case only, thus all requests are for - * last index in the block only. */ static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode, struct ext4_ext_path *path) @@ -2153,6 +2118,13 @@ static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode, err = ext4_ext_get_access(handle, inode, path); if (err) return err; + + if (path->p_idx != EXT_LAST_INDEX(path->p_hdr)) { + int len = EXT_LAST_INDEX(path->p_hdr) - path->p_idx; + len *= sizeof(struct ext4_extent_idx); + memmove(path->p_idx, path->p_idx + 1, len); + } + le16_add_cpu(&path->p_hdr->eh_entries, -1); err = ext4_ext_dirty(handle, inode, path); if (err) @@ -2534,8 +2506,7 @@ ext4_ext_more_to_rm(struct ext4_ext_path *path) return 1; } -static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start, - ext4_lblk_t end) +static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start) { struct super_block *sb = inode->i_sb; int depth = ext_depth(inode); @@ -2575,7 +2546,7 @@ again: if (i == depth) { /* this is leaf block */ err = ext4_ext_rm_leaf(handle, inode, path, - start, end); + start, EXT_MAX_BLOCKS - 1); /* root level has p_bh == NULL, brelse() eats this */ brelse(path[i].p_bh); path[i].p_bh = NULL; @@ -3107,12 +3078,10 @@ static int ext4_convert_unwritten_extents_endio(handle_t *handle, struct ext4_ext_path *path) { struct ext4_extent *ex; - struct ext4_extent_header *eh; int depth; int err = 0; depth = ext_depth(inode); - eh = path[depth].p_hdr; ex = path[depth].p_ext; ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical" @@ -3357,8 +3326,8 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); /* check in cache */ - if (ext4_ext_in_cache(inode, map->m_lblk, &newex) && - ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0)) { + if (!(flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) && + ext4_ext_in_cache(inode, map->m_lblk, &newex)) { if (!newex.ee_start_lo && !newex.ee_start_hi) { if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { /* @@ -3497,8 +3466,27 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext4_ext_mark_uninitialized(ex); - err = ext4_ext_remove_space(inode, map->m_lblk, - map->m_lblk + punched_out); + ext4_ext_invalidate_cache(inode); + + err = ext4_ext_rm_leaf(handle, inode, path, + map->m_lblk, map->m_lblk + punched_out); + + if (!err && path->p_hdr->eh_entries == 0) { + /* + * Punch hole freed all of this sub tree, + * so we need to correct eh_depth + */ + err = ext4_ext_get_access(handle, inode, path); + if (err == 0) { + ext_inode_hdr(inode)->eh_depth = 0; + ext_inode_hdr(inode)->eh_max = + cpu_to_le16(ext4_ext_space_root( + inode, 0)); + + err = ext4_ext_dirty( + handle, inode, path); + } + } goto out2; } @@ -3596,17 +3584,18 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, } err = check_eofblocks_fl(handle, inode, map->m_lblk, path, ar.len); - if (err) - goto out2; - - err = ext4_ext_insert_extent(handle, inode, path, &newex, flags); + if (!err) + err = ext4_ext_insert_extent(handle, inode, path, + &newex, flags); if (err) { + int fb_flags = flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE ? + EXT4_FREE_BLOCKS_NO_QUOT_UPDATE : 0; /* free data blocks we just allocated */ /* not a good idea to call discard here directly, * but otherwise we'd need to call it every free() */ ext4_discard_preallocations(inode); ext4_free_blocks(handle, inode, NULL, ext4_ext_pblock(&newex), - ext4_ext_get_actual_len(&newex), 0); + ext4_ext_get_actual_len(&newex), fb_flags); goto out2; } @@ -3699,7 +3688,7 @@ void ext4_ext_truncate(struct inode *inode) last_block = (inode->i_size + sb->s_blocksize - 1) >> EXT4_BLOCK_SIZE_BITS(sb); - err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1); + err = ext4_ext_remove_space(inode, last_block); /* In a multi-transaction truncate, we only make the final * transaction synchronous. @@ -3835,7 +3824,7 @@ retry: blkbits) >> blkbits)) new_size = offset + len; else - new_size = (map.m_lblk + ret) << blkbits; + new_size = ((loff_t) map.m_lblk + ret) << blkbits; ext4_falloc_update_inode(inode, mode, new_size, (map.m_flags & EXT4_MAP_NEW)); diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index da3bed3e0c29..036f78f7a1ef 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -129,15 +129,30 @@ static int ext4_sync_parent(struct inode *inode) { struct writeback_control wbc; struct dentry *dentry = NULL; + struct inode *next; int ret = 0; - while (inode && ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) { + if (!ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) + return 0; + inode = igrab(inode); + while (ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) { ext4_clear_inode_state(inode, EXT4_STATE_NEWENTRY); - dentry = list_entry(inode->i_dentry.next, - struct dentry, d_alias); - if (!dentry || !dentry->d_parent || !dentry->d_parent->d_inode) + dentry = NULL; + spin_lock(&inode->i_lock); + if (!list_empty(&inode->i_dentry)) { + dentry = list_first_entry(&inode->i_dentry, + struct dentry, d_alias); + dget(dentry); + } + spin_unlock(&inode->i_lock); + if (!dentry) break; - inode = dentry->d_parent->d_inode; + next = igrab(dentry->d_parent->d_inode); + dput(dentry); + if (!next) + break; + iput(inode); + inode = next; ret = sync_mapping_buffers(inode->i_mapping); if (ret) break; @@ -148,6 +163,7 @@ static int ext4_sync_parent(struct inode *inode) if (ret) break; } + iput(inode); return ret; } diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 21bb2f61e502..9c63f273b550 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -1287,7 +1287,7 @@ extern int ext4_init_inode_table(struct super_block *sb, ext4_group_t group, group, used_blks, ext4_itable_unused_count(sb, gdp)); ret = 1; - goto out; + goto err_out; } blk = ext4_inode_table(sb, gdp) + used_blks; diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c new file mode 100644 index 000000000000..b8602cde5b5a --- /dev/null +++ b/fs/ext4/indirect.c @@ -0,0 +1,1482 @@ +/* + * linux/fs/ext4/indirect.c + * + * from + * + * linux/fs/ext4/inode.c + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * + * from + * + * linux/fs/minix/inode.c + * + * Copyright (C) 1991, 1992 Linus Torvalds + * + * Goal-directed block allocation by Stephen Tweedie + * (sct@redhat.com), 1993, 1998 + */ + +#include +#include "ext4_jbd2.h" +#include "truncate.h" + +#include + +typedef struct { + __le32 *p; + __le32 key; + struct buffer_head *bh; +} Indirect; + +static inline void add_chain(Indirect *p, struct buffer_head *bh, __le32 *v) +{ + p->key = *(p->p = v); + p->bh = bh; +} + +/** + * ext4_block_to_path - parse the block number into array of offsets + * @inode: inode in question (we are only interested in its superblock) + * @i_block: block number to be parsed + * @offsets: array to store the offsets in + * @boundary: set this non-zero if the referred-to block is likely to be + * followed (on disk) by an indirect block. + * + * To store the locations of file's data ext4 uses a data structure common + * for UNIX filesystems - tree of pointers anchored in the inode, with + * data blocks at leaves and indirect blocks in intermediate nodes. + * This function translates the block number into path in that tree - + * return value is the path length and @offsets[n] is the offset of + * pointer to (n+1)th node in the nth one. If @block is out of range + * (negative or too large) warning is printed and zero returned. + * + * Note: function doesn't find node addresses, so no IO is needed. All + * we need to know is the capacity of indirect blocks (taken from the + * inode->i_sb). + */ + +/* + * Portability note: the last comparison (check that we fit into triple + * indirect block) is spelled differently, because otherwise on an + * architecture with 32-bit longs and 8Kb pages we might get into trouble + * if our filesystem had 8Kb blocks. We might use long long, but that would + * kill us on x86. Oh, well, at least the sign propagation does not matter - + * i_block would have to be negative in the very beginning, so we would not + * get there at all. + */ + +static int ext4_block_to_path(struct inode *inode, + ext4_lblk_t i_block, + ext4_lblk_t offsets[4], int *boundary) +{ + int ptrs = EXT4_ADDR_PER_BLOCK(inode->i_sb); + int ptrs_bits = EXT4_ADDR_PER_BLOCK_BITS(inode->i_sb); + const long direct_blocks = EXT4_NDIR_BLOCKS, + indirect_blocks = ptrs, + double_blocks = (1 << (ptrs_bits * 2)); + int n = 0; + int final = 0; + + if (i_block < direct_blocks) { + offsets[n++] = i_block; + final = direct_blocks; + } else if ((i_block -= direct_blocks) < indirect_blocks) { + offsets[n++] = EXT4_IND_BLOCK; + offsets[n++] = i_block; + final = ptrs; + } else if ((i_block -= indirect_blocks) < double_blocks) { + offsets[n++] = EXT4_DIND_BLOCK; + offsets[n++] = i_block >> ptrs_bits; + offsets[n++] = i_block & (ptrs - 1); + final = ptrs; + } else if (((i_block -= double_blocks) >> (ptrs_bits * 2)) < ptrs) { + offsets[n++] = EXT4_TIND_BLOCK; + offsets[n++] = i_block >> (ptrs_bits * 2); + offsets[n++] = (i_block >> ptrs_bits) & (ptrs - 1); + offsets[n++] = i_block & (ptrs - 1); + final = ptrs; + } else { + ext4_warning(inode->i_sb, "block %lu > max in inode %lu", + i_block + direct_blocks + + indirect_blocks + double_blocks, inode->i_ino); + } + if (boundary) + *boundary = final - 1 - (i_block & (ptrs - 1)); + return n; +} + +/** + * ext4_get_branch - read the chain of indirect blocks leading to data + * @inode: inode in question + * @depth: depth of the chain (1 - direct pointer, etc.) + * @offsets: offsets of pointers in inode/indirect blocks + * @chain: place to store the result + * @err: here we store the error value + * + * Function fills the array of triples and returns %NULL + * if everything went OK or the pointer to the last filled triple + * (incomplete one) otherwise. Upon the return chain[i].key contains + * the number of (i+1)-th block in the chain (as it is stored in memory, + * i.e. little-endian 32-bit), chain[i].p contains the address of that + * number (it points into struct inode for i==0 and into the bh->b_data + * for i>0) and chain[i].bh points to the buffer_head of i-th indirect + * block for i>0 and NULL for i==0. In other words, it holds the block + * numbers of the chain, addresses they were taken from (and where we can + * verify that chain did not change) and buffer_heads hosting these + * numbers. + * + * Function stops when it stumbles upon zero pointer (absent block) + * (pointer to last triple returned, *@err == 0) + * or when it gets an IO error reading an indirect block + * (ditto, *@err == -EIO) + * or when it reads all @depth-1 indirect blocks successfully and finds + * the whole chain, all way to the data (returns %NULL, *err == 0). + * + * Need to be called with + * down_read(&EXT4_I(inode)->i_data_sem) + */ +static Indirect *ext4_get_branch(struct inode *inode, int depth, + ext4_lblk_t *offsets, + Indirect chain[4], int *err) +{ + struct super_block *sb = inode->i_sb; + Indirect *p = chain; + struct buffer_head *bh; + + *err = 0; + /* i_data is not going away, no lock needed */ + add_chain(chain, NULL, EXT4_I(inode)->i_data + *offsets); + if (!p->key) + goto no_block; + while (--depth) { + bh = sb_getblk(sb, le32_to_cpu(p->key)); + if (unlikely(!bh)) + goto failure; + + if (!bh_uptodate_or_lock(bh)) { + if (bh_submit_read(bh) < 0) { + put_bh(bh); + goto failure; + } + /* validate block references */ + if (ext4_check_indirect_blockref(inode, bh)) { + put_bh(bh); + goto failure; + } + } + + add_chain(++p, bh, (__le32 *)bh->b_data + *++offsets); + /* Reader: end */ + if (!p->key) + goto no_block; + } + return NULL; + +failure: + *err = -EIO; +no_block: + return p; +} + +/** + * ext4_find_near - find a place for allocation with sufficient locality + * @inode: owner + * @ind: descriptor of indirect block. + * + * This function returns the preferred place for block allocation. + * It is used when heuristic for sequential allocation fails. + * Rules are: + * + if there is a block to the left of our position - allocate near it. + * + if pointer will live in indirect block - allocate near that block. + * + if pointer will live in inode - allocate in the same + * cylinder group. + * + * In the latter case we colour the starting block by the callers PID to + * prevent it from clashing with concurrent allocations for a different inode + * in the same block group. The PID is used here so that functionally related + * files will be close-by on-disk. + * + * Caller must make sure that @ind is valid and will stay that way. + */ +static ext4_fsblk_t ext4_find_near(struct inode *inode, Indirect *ind) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + __le32 *start = ind->bh ? (__le32 *) ind->bh->b_data : ei->i_data; + __le32 *p; + + /* Try to find previous block */ + for (p = ind->p - 1; p >= start; p--) { + if (*p) + return le32_to_cpu(*p); + } + + /* No such thing, so let's try location of indirect block */ + if (ind->bh) + return ind->bh->b_blocknr; + + /* + * It is going to be referred to from the inode itself? OK, just put it + * into the same cylinder group then. + */ + return ext4_inode_to_goal_block(inode); +} + +/** + * ext4_find_goal - find a preferred place for allocation. + * @inode: owner + * @block: block we want + * @partial: pointer to the last triple within a chain + * + * Normally this function find the preferred place for block allocation, + * returns it. + * Because this is only used for non-extent files, we limit the block nr + * to 32 bits. + */ +static ext4_fsblk_t ext4_find_goal(struct inode *inode, ext4_lblk_t block, + Indirect *partial) +{ + ext4_fsblk_t goal; + + /* + * XXX need to get goal block from mballoc's data structures + */ + + goal = ext4_find_near(inode, partial); + goal = goal & EXT4_MAX_BLOCK_FILE_PHYS; + return goal; +} + +/** + * ext4_blks_to_allocate - Look up the block map and count the number + * of direct blocks need to be allocated for the given branch. + * + * @branch: chain of indirect blocks + * @k: number of blocks need for indirect blocks + * @blks: number of data blocks to be mapped. + * @blocks_to_boundary: the offset in the indirect block + * + * return the total number of blocks to be allocate, including the + * direct and indirect blocks. + */ +static int ext4_blks_to_allocate(Indirect *branch, int k, unsigned int blks, + int blocks_to_boundary) +{ + unsigned int count = 0; + + /* + * Simple case, [t,d]Indirect block(s) has not allocated yet + * then it's clear blocks on that path have not allocated + */ + if (k > 0) { + /* right now we don't handle cross boundary allocation */ + if (blks < blocks_to_boundary + 1) + count += blks; + else + count += blocks_to_boundary + 1; + return count; + } + + count++; + while (count < blks && count <= blocks_to_boundary && + le32_to_cpu(*(branch[0].p + count)) == 0) { + count++; + } + return count; +} + +/** + * ext4_alloc_blocks: multiple allocate blocks needed for a branch + * @handle: handle for this transaction + * @inode: inode which needs allocated blocks + * @iblock: the logical block to start allocated at + * @goal: preferred physical block of allocation + * @indirect_blks: the number of blocks need to allocate for indirect + * blocks + * @blks: number of desired blocks + * @new_blocks: on return it will store the new block numbers for + * the indirect blocks(if needed) and the first direct block, + * @err: on return it will store the error code + * + * This function will return the number of blocks allocated as + * requested by the passed-in parameters. + */ +static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, + ext4_lblk_t iblock, ext4_fsblk_t goal, + int indirect_blks, int blks, + ext4_fsblk_t new_blocks[4], int *err) +{ + struct ext4_allocation_request ar; + int target, i; + unsigned long count = 0, blk_allocated = 0; + int index = 0; + ext4_fsblk_t current_block = 0; + int ret = 0; + + /* + * Here we try to allocate the requested multiple blocks at once, + * on a best-effort basis. + * To build a branch, we should allocate blocks for + * the indirect blocks(if not allocated yet), and at least + * the first direct block of this branch. That's the + * minimum number of blocks need to allocate(required) + */ + /* first we try to allocate the indirect blocks */ + target = indirect_blks; + while (target > 0) { + count = target; + /* allocating blocks for indirect blocks and direct blocks */ + current_block = ext4_new_meta_blocks(handle, inode, goal, + 0, &count, err); + if (*err) + goto failed_out; + + if (unlikely(current_block + count > EXT4_MAX_BLOCK_FILE_PHYS)) { + EXT4_ERROR_INODE(inode, + "current_block %llu + count %lu > %d!", + current_block, count, + EXT4_MAX_BLOCK_FILE_PHYS); + *err = -EIO; + goto failed_out; + } + + target -= count; + /* allocate blocks for indirect blocks */ + while (index < indirect_blks && count) { + new_blocks[index++] = current_block++; + count--; + } + if (count > 0) { + /* + * save the new block number + * for the first direct block + */ + new_blocks[index] = current_block; + printk(KERN_INFO "%s returned more blocks than " + "requested\n", __func__); + WARN_ON(1); + break; + } + } + + target = blks - count ; + blk_allocated = count; + if (!target) + goto allocated; + /* Now allocate data blocks */ + memset(&ar, 0, sizeof(ar)); + ar.inode = inode; + ar.goal = goal; + ar.len = target; + ar.logical = iblock; + if (S_ISREG(inode->i_mode)) + /* enable in-core preallocation only for regular files */ + ar.flags = EXT4_MB_HINT_DATA; + + current_block = ext4_mb_new_blocks(handle, &ar, err); + if (unlikely(current_block + ar.len > EXT4_MAX_BLOCK_FILE_PHYS)) { + EXT4_ERROR_INODE(inode, + "current_block %llu + ar.len %d > %d!", + current_block, ar.len, + EXT4_MAX_BLOCK_FILE_PHYS); + *err = -EIO; + goto failed_out; + } + + if (*err && (target == blks)) { + /* + * if the allocation failed and we didn't allocate + * any blocks before + */ + goto failed_out; + } + if (!*err) { + if (target == blks) { + /* + * save the new block number + * for the first direct block + */ + new_blocks[index] = current_block; + } + blk_allocated += ar.len; + } +allocated: + /* total number of blocks allocated for direct blocks */ + ret = blk_allocated; + *err = 0; + return ret; +failed_out: + for (i = 0; i < index; i++) + ext4_free_blocks(handle, inode, NULL, new_blocks[i], 1, 0); + return ret; +} + +/** + * ext4_alloc_branch - allocate and set up a chain of blocks. + * @handle: handle for this transaction + * @inode: owner + * @indirect_blks: number of allocated indirect blocks + * @blks: number of allocated direct blocks + * @goal: preferred place for allocation + * @offsets: offsets (in the blocks) to store the pointers to next. + * @branch: place to store the chain in. + * + * This function allocates blocks, zeroes out all but the last one, + * links them into chain and (if we are synchronous) writes them to disk. + * In other words, it prepares a branch that can be spliced onto the + * inode. It stores the information about that chain in the branch[], in + * the same format as ext4_get_branch() would do. We are calling it after + * we had read the existing part of chain and partial points to the last + * triple of that (one with zero ->key). Upon the exit we have the same + * picture as after the successful ext4_get_block(), except that in one + * place chain is disconnected - *branch->p is still zero (we did not + * set the last link), but branch->key contains the number that should + * be placed into *branch->p to fill that gap. + * + * If allocation fails we free all blocks we've allocated (and forget + * their buffer_heads) and return the error value the from failed + * ext4_alloc_block() (normally -ENOSPC). Otherwise we set the chain + * as described above and return 0. + */ +static int ext4_alloc_branch(handle_t *handle, struct inode *inode, + ext4_lblk_t iblock, int indirect_blks, + int *blks, ext4_fsblk_t goal, + ext4_lblk_t *offsets, Indirect *branch) +{ + int blocksize = inode->i_sb->s_blocksize; + int i, n = 0; + int err = 0; + struct buffer_head *bh; + int num; + ext4_fsblk_t new_blocks[4]; + ext4_fsblk_t current_block; + + num = ext4_alloc_blocks(handle, inode, iblock, goal, indirect_blks, + *blks, new_blocks, &err); + if (err) + return err; + + branch[0].key = cpu_to_le32(new_blocks[0]); + /* + * metadata blocks and data blocks are allocated. + */ + for (n = 1; n <= indirect_blks; n++) { + /* + * Get buffer_head for parent block, zero it out + * and set the pointer to new one, then send + * parent to disk. + */ + bh = sb_getblk(inode->i_sb, new_blocks[n-1]); + if (unlikely(!bh)) { + err = -EIO; + goto failed; + } + + branch[n].bh = bh; + lock_buffer(bh); + BUFFER_TRACE(bh, "call get_create_access"); + err = ext4_journal_get_create_access(handle, bh); + if (err) { + /* Don't brelse(bh) here; it's done in + * ext4_journal_forget() below */ + unlock_buffer(bh); + goto failed; + } + + memset(bh->b_data, 0, blocksize); + branch[n].p = (__le32 *) bh->b_data + offsets[n]; + branch[n].key = cpu_to_le32(new_blocks[n]); + *branch[n].p = branch[n].key; + if (n == indirect_blks) { + current_block = new_blocks[n]; + /* + * End of chain, update the last new metablock of + * the chain to point to the new allocated + * data blocks numbers + */ + for (i = 1; i < num; i++) + *(branch[n].p + i) = cpu_to_le32(++current_block); + } + BUFFER_TRACE(bh, "marking uptodate"); + set_buffer_uptodate(bh); + unlock_buffer(bh); + + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (err) + goto failed; + } + *blks = num; + return err; +failed: + /* Allocation failed, free what we already allocated */ + ext4_free_blocks(handle, inode, NULL, new_blocks[0], 1, 0); + for (i = 1; i <= n ; i++) { + /* + * branch[i].bh is newly allocated, so there is no + * need to revoke the block, which is why we don't + * need to set EXT4_FREE_BLOCKS_METADATA. + */ + ext4_free_blocks(handle, inode, NULL, new_blocks[i], 1, + EXT4_FREE_BLOCKS_FORGET); + } + for (i = n+1; i < indirect_blks; i++) + ext4_free_blocks(handle, inode, NULL, new_blocks[i], 1, 0); + + ext4_free_blocks(handle, inode, NULL, new_blocks[i], num, 0); + + return err; +} + +/** + * ext4_splice_branch - splice the allocated branch onto inode. + * @handle: handle for this transaction + * @inode: owner + * @block: (logical) number of block we are adding + * @chain: chain of indirect blocks (with a missing link - see + * ext4_alloc_branch) + * @where: location of missing link + * @num: number of indirect blocks we are adding + * @blks: number of direct blocks we are adding + * + * This function fills the missing link and does all housekeeping needed in + * inode (->i_blocks, etc.). In case of success we end up with the full + * chain to new block and return 0. + */ +static int ext4_splice_branch(handle_t *handle, struct inode *inode, + ext4_lblk_t block, Indirect *where, int num, + int blks) +{ + int i; + int err = 0; + ext4_fsblk_t current_block; + + /* + * If we're splicing into a [td]indirect block (as opposed to the + * inode) then we need to get write access to the [td]indirect block + * before the splice. + */ + if (where->bh) { + BUFFER_TRACE(where->bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, where->bh); + if (err) + goto err_out; + } + /* That's it */ + + *where->p = where->key; + + /* + * Update the host buffer_head or inode to point to more just allocated + * direct blocks blocks + */ + if (num == 0 && blks > 1) { + current_block = le32_to_cpu(where->key) + 1; + for (i = 1; i < blks; i++) + *(where->p + i) = cpu_to_le32(current_block++); + } + + /* We are done with atomic stuff, now do the rest of housekeeping */ + /* had we spliced it onto indirect block? */ + if (where->bh) { + /* + * If we spliced it onto an indirect block, we haven't + * altered the inode. Note however that if it is being spliced + * onto an indirect block at the very end of the file (the + * file is growing) then we *will* alter the inode to reflect + * the new i_size. But that is not done here - it is done in + * generic_commit_write->__mark_inode_dirty->ext4_dirty_inode. + */ + jbd_debug(5, "splicing indirect only\n"); + BUFFER_TRACE(where->bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, inode, where->bh); + if (err) + goto err_out; + } else { + /* + * OK, we spliced it into the inode itself on a direct block. + */ + ext4_mark_inode_dirty(handle, inode); + jbd_debug(5, "splicing direct\n"); + } + return err; + +err_out: + for (i = 1; i <= num; i++) { + /* + * branch[i].bh is newly allocated, so there is no + * need to revoke the block, which is why we don't + * need to set EXT4_FREE_BLOCKS_METADATA. + */ + ext4_free_blocks(handle, inode, where[i].bh, 0, 1, + EXT4_FREE_BLOCKS_FORGET); + } + ext4_free_blocks(handle, inode, NULL, le32_to_cpu(where[num].key), + blks, 0); + + return err; +} + +/* + * The ext4_ind_map_blocks() function handles non-extents inodes + * (i.e., using the traditional indirect/double-indirect i_blocks + * scheme) for ext4_map_blocks(). + * + * Allocation strategy is simple: if we have to allocate something, we will + * have to go the whole way to leaf. So let's do it before attaching anything + * to tree, set linkage between the newborn blocks, write them if sync is + * required, recheck the path, free and repeat if check fails, otherwise + * set the last missing link (that will protect us from any truncate-generated + * removals - all blocks on the path are immune now) and possibly force the + * write on the parent block. + * That has a nice additional property: no special recovery from the failed + * allocations is needed - we simply release blocks and do not touch anything + * reachable from inode. + * + * `handle' can be NULL if create == 0. + * + * return > 0, # of blocks mapped or allocated. + * return = 0, if plain lookup failed. + * return < 0, error case. + * + * The ext4_ind_get_blocks() function should be called with + * down_write(&EXT4_I(inode)->i_data_sem) if allocating filesystem + * blocks (i.e., flags has EXT4_GET_BLOCKS_CREATE set) or + * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system + * blocks. + */ +int ext4_ind_map_blocks(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, + int flags) +{ + int err = -EIO; + ext4_lblk_t offsets[4]; + Indirect chain[4]; + Indirect *partial; + ext4_fsblk_t goal; + int indirect_blks; + int blocks_to_boundary = 0; + int depth; + int count = 0; + ext4_fsblk_t first_block = 0; + + trace_ext4_ind_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); + J_ASSERT(!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))); + J_ASSERT(handle != NULL || (flags & EXT4_GET_BLOCKS_CREATE) == 0); + depth = ext4_block_to_path(inode, map->m_lblk, offsets, + &blocks_to_boundary); + + if (depth == 0) + goto out; + + partial = ext4_get_branch(inode, depth, offsets, chain, &err); + + /* Simplest case - block found, no allocation needed */ + if (!partial) { + first_block = le32_to_cpu(chain[depth - 1].key); + count++; + /*map more blocks*/ + while (count < map->m_len && count <= blocks_to_boundary) { + ext4_fsblk_t blk; + + blk = le32_to_cpu(*(chain[depth-1].p + count)); + + if (blk == first_block + count) + count++; + else + break; + } + goto got_it; + } + + /* Next simple case - plain lookup or failed read of indirect block */ + if ((flags & EXT4_GET_BLOCKS_CREATE) == 0 || err == -EIO) + goto cleanup; + + /* + * Okay, we need to do block allocation. + */ + goal = ext4_find_goal(inode, map->m_lblk, partial); + + /* the number of blocks need to allocate for [d,t]indirect blocks */ + indirect_blks = (chain + depth) - partial - 1; + + /* + * Next look up the indirect map to count the totoal number of + * direct blocks to allocate for this branch. + */ + count = ext4_blks_to_allocate(partial, indirect_blks, + map->m_len, blocks_to_boundary); + /* + * Block out ext4_truncate while we alter the tree + */ + err = ext4_alloc_branch(handle, inode, map->m_lblk, indirect_blks, + &count, goal, + offsets + (partial - chain), partial); + + /* + * The ext4_splice_branch call will free and forget any buffers + * on the new chain if there is a failure, but that risks using + * up transaction credits, especially for bitmaps where the + * credits cannot be returned. Can we handle this somehow? We + * may need to return -EAGAIN upwards in the worst case. --sct + */ + if (!err) + err = ext4_splice_branch(handle, inode, map->m_lblk, + partial, indirect_blks, count); + if (err) + goto cleanup; + + map->m_flags |= EXT4_MAP_NEW; + + ext4_update_inode_fsync_trans(handle, inode, 1); +got_it: + map->m_flags |= EXT4_MAP_MAPPED; + map->m_pblk = le32_to_cpu(chain[depth-1].key); + map->m_len = count; + if (count > blocks_to_boundary) + map->m_flags |= EXT4_MAP_BOUNDARY; + err = count; + /* Clean up and exit */ + partial = chain + depth - 1; /* the whole chain */ +cleanup: + while (partial > chain) { + BUFFER_TRACE(partial->bh, "call brelse"); + brelse(partial->bh); + partial--; + } +out: + trace_ext4_ind_map_blocks_exit(inode, map->m_lblk, + map->m_pblk, map->m_len, err); + return err; +} + +/* + * O_DIRECT for ext3 (or indirect map) based files + * + * If the O_DIRECT write will extend the file then add this inode to the + * orphan list. So recovery will truncate it back to the original size + * if the machine crashes during the write. + * + * If the O_DIRECT write is intantiating holes inside i_size and the machine + * crashes then stale disk data _may_ be exposed inside the file. But current + * VFS code falls back into buffered path in that case so we are safe. + */ +ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb, + const struct iovec *iov, loff_t offset, + unsigned long nr_segs) +{ + struct file *file = iocb->ki_filp; + struct inode *inode = file->f_mapping->host; + struct ext4_inode_info *ei = EXT4_I(inode); + handle_t *handle; + ssize_t ret; + int orphan = 0; + size_t count = iov_length(iov, nr_segs); + int retries = 0; + + if (rw == WRITE) { + loff_t final_size = offset + count; + + if (final_size > inode->i_size) { + /* Credits for sb + inode write */ + handle = ext4_journal_start(inode, 2); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out; + } + ret = ext4_orphan_add(handle, inode); + if (ret) { + ext4_journal_stop(handle); + goto out; + } + orphan = 1; + ei->i_disksize = inode->i_size; + ext4_journal_stop(handle); + } + } + +retry: + if (rw == READ && ext4_should_dioread_nolock(inode)) + ret = __blockdev_direct_IO(rw, iocb, inode, + inode->i_sb->s_bdev, iov, + offset, nr_segs, + ext4_get_block, NULL, NULL, 0); + else { + ret = blockdev_direct_IO(rw, iocb, inode, iov, + offset, nr_segs, ext4_get_block); + + if (unlikely((rw & WRITE) && ret < 0)) { + loff_t isize = i_size_read(inode); + loff_t end = offset + iov_length(iov, nr_segs); + + if (end > isize) + ext4_truncate_failed_write(inode); + } + } + if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry; + + if (orphan) { + int err; + + /* Credits for sb + inode write */ + handle = ext4_journal_start(inode, 2); + if (IS_ERR(handle)) { + /* This is really bad luck. We've written the data + * but cannot extend i_size. Bail out and pretend + * the write failed... */ + ret = PTR_ERR(handle); + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + + goto out; + } + if (inode->i_nlink) + ext4_orphan_del(handle, inode); + if (ret > 0) { + loff_t end = offset + ret; + if (end > inode->i_size) { + ei->i_disksize = end; + i_size_write(inode, end); + /* + * We're going to return a positive `ret' + * here due to non-zero-length I/O, so there's + * no way of reporting error returns from + * ext4_mark_inode_dirty() to userspace. So + * ignore it. + */ + ext4_mark_inode_dirty(handle, inode); + } + } + err = ext4_journal_stop(handle); + if (ret == 0) + ret = err; + } +out: + return ret; +} + +/* + * Calculate the number of metadata blocks need to reserve + * to allocate a new block at @lblocks for non extent file based file + */ +int ext4_ind_calc_metadata_amount(struct inode *inode, sector_t lblock) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + sector_t dind_mask = ~((sector_t)EXT4_ADDR_PER_BLOCK(inode->i_sb) - 1); + int blk_bits; + + if (lblock < EXT4_NDIR_BLOCKS) + return 0; + + lblock -= EXT4_NDIR_BLOCKS; + + if (ei->i_da_metadata_calc_len && + (lblock & dind_mask) == ei->i_da_metadata_calc_last_lblock) { + ei->i_da_metadata_calc_len++; + return 0; + } + ei->i_da_metadata_calc_last_lblock = lblock & dind_mask; + ei->i_da_metadata_calc_len = 1; + blk_bits = order_base_2(lblock); + return (blk_bits / EXT4_ADDR_PER_BLOCK_BITS(inode->i_sb)) + 1; +} + +int ext4_ind_trans_blocks(struct inode *inode, int nrblocks, int chunk) +{ + int indirects; + + /* if nrblocks are contiguous */ + if (chunk) { + /* + * With N contiguous data blocks, we need at most + * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) + 1 indirect blocks, + * 2 dindirect blocks, and 1 tindirect block + */ + return DIV_ROUND_UP(nrblocks, + EXT4_ADDR_PER_BLOCK(inode->i_sb)) + 4; + } + /* + * if nrblocks are not contiguous, worse case, each block touch + * a indirect block, and each indirect block touch a double indirect + * block, plus a triple indirect block + */ + indirects = nrblocks * 2 + 1; + return indirects; +} + +/* + * Truncate transactions can be complex and absolutely huge. So we need to + * be able to restart the transaction at a conventient checkpoint to make + * sure we don't overflow the journal. + * + * start_transaction gets us a new handle for a truncate transaction, + * and extend_transaction tries to extend the existing one a bit. If + * extend fails, we need to propagate the failure up and restart the + * transaction in the top-level truncate loop. --sct + */ +static handle_t *start_transaction(struct inode *inode) +{ + handle_t *result; + + result = ext4_journal_start(inode, ext4_blocks_for_truncate(inode)); + if (!IS_ERR(result)) + return result; + + ext4_std_error(inode->i_sb, PTR_ERR(result)); + return result; +} + +/* + * Try to extend this transaction for the purposes of truncation. + * + * Returns 0 if we managed to create more room. If we can't create more + * room, and the transaction must be restarted we return 1. + */ +static int try_to_extend_transaction(handle_t *handle, struct inode *inode) +{ + if (!ext4_handle_valid(handle)) + return 0; + if (ext4_handle_has_enough_credits(handle, EXT4_RESERVE_TRANS_BLOCKS+1)) + return 0; + if (!ext4_journal_extend(handle, ext4_blocks_for_truncate(inode))) + return 0; + return 1; +} + +/* + * Probably it should be a library function... search for first non-zero word + * or memcmp with zero_page, whatever is better for particular architecture. + * Linus? + */ +static inline int all_zeroes(__le32 *p, __le32 *q) +{ + while (p < q) + if (*p++) + return 0; + return 1; +} + +/** + * ext4_find_shared - find the indirect blocks for partial truncation. + * @inode: inode in question + * @depth: depth of the affected branch + * @offsets: offsets of pointers in that branch (see ext4_block_to_path) + * @chain: place to store the pointers to partial indirect blocks + * @top: place to the (detached) top of branch + * + * This is a helper function used by ext4_truncate(). + * + * When we do truncate() we may have to clean the ends of several + * indirect blocks but leave the blocks themselves alive. Block is + * partially truncated if some data below the new i_size is referred + * from it (and it is on the path to the first completely truncated + * data block, indeed). We have to free the top of that path along + * with everything to the right of the path. Since no allocation + * past the truncation point is possible until ext4_truncate() + * finishes, we may safely do the latter, but top of branch may + * require special attention - pageout below the truncation point + * might try to populate it. + * + * We atomically detach the top of branch from the tree, store the + * block number of its root in *@top, pointers to buffer_heads of + * partially truncated blocks - in @chain[].bh and pointers to + * their last elements that should not be removed - in + * @chain[].p. Return value is the pointer to last filled element + * of @chain. + * + * The work left to caller to do the actual freeing of subtrees: + * a) free the subtree starting from *@top + * b) free the subtrees whose roots are stored in + * (@chain[i].p+1 .. end of @chain[i].bh->b_data) + * c) free the subtrees growing from the inode past the @chain[0]. + * (no partially truncated stuff there). */ + +static Indirect *ext4_find_shared(struct inode *inode, int depth, + ext4_lblk_t offsets[4], Indirect chain[4], + __le32 *top) +{ + Indirect *partial, *p; + int k, err; + + *top = 0; + /* Make k index the deepest non-null offset + 1 */ + for (k = depth; k > 1 && !offsets[k-1]; k--) + ; + partial = ext4_get_branch(inode, k, offsets, chain, &err); + /* Writer: pointers */ + if (!partial) + partial = chain + k-1; + /* + * If the branch acquired continuation since we've looked at it - + * fine, it should all survive and (new) top doesn't belong to us. + */ + if (!partial->key && *partial->p) + /* Writer: end */ + goto no_top; + for (p = partial; (p > chain) && all_zeroes((__le32 *) p->bh->b_data, p->p); p--) + ; + /* + * OK, we've found the last block that must survive. The rest of our + * branch should be detached before unlocking. However, if that rest + * of branch is all ours and does not grow immediately from the inode + * it's easier to cheat and just decrement partial->p. + */ + if (p == chain + k - 1 && p > chain) { + p->p--; + } else { + *top = *p->p; + /* Nope, don't do this in ext4. Must leave the tree intact */ +#if 0 + *p->p = 0; +#endif + } + /* Writer: end */ + + while (partial > p) { + brelse(partial->bh); + partial--; + } +no_top: + return partial; +} + +/* + * Zero a number of block pointers in either an inode or an indirect block. + * If we restart the transaction we must again get write access to the + * indirect block for further modification. + * + * We release `count' blocks on disk, but (last - first) may be greater + * than `count' because there can be holes in there. + * + * Return 0 on success, 1 on invalid block range + * and < 0 on fatal error. + */ +static int ext4_clear_blocks(handle_t *handle, struct inode *inode, + struct buffer_head *bh, + ext4_fsblk_t block_to_free, + unsigned long count, __le32 *first, + __le32 *last) +{ + __le32 *p; + int flags = EXT4_FREE_BLOCKS_FORGET | EXT4_FREE_BLOCKS_VALIDATED; + int err; + + if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) + flags |= EXT4_FREE_BLOCKS_METADATA; + + if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), block_to_free, + count)) { + EXT4_ERROR_INODE(inode, "attempt to clear invalid " + "blocks %llu len %lu", + (unsigned long long) block_to_free, count); + return 1; + } + + if (try_to_extend_transaction(handle, inode)) { + if (bh) { + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (unlikely(err)) + goto out_err; + } + err = ext4_mark_inode_dirty(handle, inode); + if (unlikely(err)) + goto out_err; + err = ext4_truncate_restart_trans(handle, inode, + ext4_blocks_for_truncate(inode)); + if (unlikely(err)) + goto out_err; + if (bh) { + BUFFER_TRACE(bh, "retaking write access"); + err = ext4_journal_get_write_access(handle, bh); + if (unlikely(err)) + goto out_err; + } + } + + for (p = first; p < last; p++) + *p = 0; + + ext4_free_blocks(handle, inode, NULL, block_to_free, count, flags); + return 0; +out_err: + ext4_std_error(inode->i_sb, err); + return err; +} + +/** + * ext4_free_data - free a list of data blocks + * @handle: handle for this transaction + * @inode: inode we are dealing with + * @this_bh: indirect buffer_head which contains *@first and *@last + * @first: array of block numbers + * @last: points immediately past the end of array + * + * We are freeing all blocks referred from that array (numbers are stored as + * little-endian 32-bit) and updating @inode->i_blocks appropriately. + * + * We accumulate contiguous runs of blocks to free. Conveniently, if these + * blocks are contiguous then releasing them at one time will only affect one + * or two bitmap blocks (+ group descriptor(s) and superblock) and we won't + * actually use a lot of journal space. + * + * @this_bh will be %NULL if @first and @last point into the inode's direct + * block pointers. + */ +static void ext4_free_data(handle_t *handle, struct inode *inode, + struct buffer_head *this_bh, + __le32 *first, __le32 *last) +{ + ext4_fsblk_t block_to_free = 0; /* Starting block # of a run */ + unsigned long count = 0; /* Number of blocks in the run */ + __le32 *block_to_free_p = NULL; /* Pointer into inode/ind + corresponding to + block_to_free */ + ext4_fsblk_t nr; /* Current block # */ + __le32 *p; /* Pointer into inode/ind + for current block */ + int err = 0; + + if (this_bh) { /* For indirect block */ + BUFFER_TRACE(this_bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, this_bh); + /* Important: if we can't update the indirect pointers + * to the blocks, we can't free them. */ + if (err) + return; + } + + for (p = first; p < last; p++) { + nr = le32_to_cpu(*p); + if (nr) { + /* accumulate blocks to free if they're contiguous */ + if (count == 0) { + block_to_free = nr; + block_to_free_p = p; + count = 1; + } else if (nr == block_to_free + count) { + count++; + } else { + err = ext4_clear_blocks(handle, inode, this_bh, + block_to_free, count, + block_to_free_p, p); + if (err) + break; + block_to_free = nr; + block_to_free_p = p; + count = 1; + } + } + } + + if (!err && count > 0) + err = ext4_clear_blocks(handle, inode, this_bh, block_to_free, + count, block_to_free_p, p); + if (err < 0) + /* fatal error */ + return; + + if (this_bh) { + BUFFER_TRACE(this_bh, "call ext4_handle_dirty_metadata"); + + /* + * The buffer head should have an attached journal head at this + * point. However, if the data is corrupted and an indirect + * block pointed to itself, it would have been detached when + * the block was cleared. Check for this instead of OOPSing. + */ + if ((EXT4_JOURNAL(inode) == NULL) || bh2jh(this_bh)) + ext4_handle_dirty_metadata(handle, inode, this_bh); + else + EXT4_ERROR_INODE(inode, + "circular indirect block detected at " + "block %llu", + (unsigned long long) this_bh->b_blocknr); + } +} + +/** + * ext4_free_branches - free an array of branches + * @handle: JBD handle for this transaction + * @inode: inode we are dealing with + * @parent_bh: the buffer_head which contains *@first and *@last + * @first: array of block numbers + * @last: pointer immediately past the end of array + * @depth: depth of the branches to free + * + * We are freeing all blocks referred from these branches (numbers are + * stored as little-endian 32-bit) and updating @inode->i_blocks + * appropriately. + */ +static void ext4_free_branches(handle_t *handle, struct inode *inode, + struct buffer_head *parent_bh, + __le32 *first, __le32 *last, int depth) +{ + ext4_fsblk_t nr; + __le32 *p; + + if (ext4_handle_is_aborted(handle)) + return; + + if (depth--) { + struct buffer_head *bh; + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + p = last; + while (--p >= first) { + nr = le32_to_cpu(*p); + if (!nr) + continue; /* A hole */ + + if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), + nr, 1)) { + EXT4_ERROR_INODE(inode, + "invalid indirect mapped " + "block %lu (level %d)", + (unsigned long) nr, depth); + break; + } + + /* Go read the buffer for the next level down */ + bh = sb_bread(inode->i_sb, nr); + + /* + * A read failure? Report error and clear slot + * (should be rare). + */ + if (!bh) { + EXT4_ERROR_INODE_BLOCK(inode, nr, + "Read failure"); + continue; + } + + /* This zaps the entire block. Bottom up. */ + BUFFER_TRACE(bh, "free child branches"); + ext4_free_branches(handle, inode, bh, + (__le32 *) bh->b_data, + (__le32 *) bh->b_data + addr_per_block, + depth); + brelse(bh); + + /* + * Everything below this this pointer has been + * released. Now let this top-of-subtree go. + * + * We want the freeing of this indirect block to be + * atomic in the journal with the updating of the + * bitmap block which owns it. So make some room in + * the journal. + * + * We zero the parent pointer *after* freeing its + * pointee in the bitmaps, so if extend_transaction() + * for some reason fails to put the bitmap changes and + * the release into the same transaction, recovery + * will merely complain about releasing a free block, + * rather than leaking blocks. + */ + if (ext4_handle_is_aborted(handle)) + return; + if (try_to_extend_transaction(handle, inode)) { + ext4_mark_inode_dirty(handle, inode); + ext4_truncate_restart_trans(handle, inode, + ext4_blocks_for_truncate(inode)); + } + + /* + * The forget flag here is critical because if + * we are journaling (and not doing data + * journaling), we have to make sure a revoke + * record is written to prevent the journal + * replay from overwriting the (former) + * indirect block if it gets reallocated as a + * data block. This must happen in the same + * transaction where the data blocks are + * actually freed. + */ + ext4_free_blocks(handle, inode, NULL, nr, 1, + EXT4_FREE_BLOCKS_METADATA| + EXT4_FREE_BLOCKS_FORGET); + + if (parent_bh) { + /* + * The block which we have just freed is + * pointed to by an indirect block: journal it + */ + BUFFER_TRACE(parent_bh, "get_write_access"); + if (!ext4_journal_get_write_access(handle, + parent_bh)){ + *p = 0; + BUFFER_TRACE(parent_bh, + "call ext4_handle_dirty_metadata"); + ext4_handle_dirty_metadata(handle, + inode, + parent_bh); + } + } + } + } else { + /* We have reached the bottom of the tree. */ + BUFFER_TRACE(parent_bh, "free data blocks"); + ext4_free_data(handle, inode, parent_bh, first, last); + } +} + +void ext4_ind_truncate(struct inode *inode) +{ + handle_t *handle; + struct ext4_inode_info *ei = EXT4_I(inode); + __le32 *i_data = ei->i_data; + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + struct address_space *mapping = inode->i_mapping; + ext4_lblk_t offsets[4]; + Indirect chain[4]; + Indirect *partial; + __le32 nr = 0; + int n = 0; + ext4_lblk_t last_block, max_block; + unsigned blocksize = inode->i_sb->s_blocksize; + + handle = start_transaction(inode); + if (IS_ERR(handle)) + return; /* AKPM: return what? */ + + last_block = (inode->i_size + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + max_block = (EXT4_SB(inode->i_sb)->s_bitmap_maxbytes + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + + if (inode->i_size & (blocksize - 1)) + if (ext4_block_truncate_page(handle, mapping, inode->i_size)) + goto out_stop; + + if (last_block != max_block) { + n = ext4_block_to_path(inode, last_block, offsets, NULL); + if (n == 0) + goto out_stop; /* error */ + } + + /* + * OK. This truncate is going to happen. We add the inode to the + * orphan list, so that if this truncate spans multiple transactions, + * and we crash, we will resume the truncate when the filesystem + * recovers. It also marks the inode dirty, to catch the new size. + * + * Implication: the file must always be in a sane, consistent + * truncatable state while each transaction commits. + */ + if (ext4_orphan_add(handle, inode)) + goto out_stop; + + /* + * From here we block out all ext4_get_block() callers who want to + * modify the block allocation tree. + */ + down_write(&ei->i_data_sem); + + ext4_discard_preallocations(inode); + + /* + * The orphan list entry will now protect us from any crash which + * occurs before the truncate completes, so it is now safe to propagate + * the new, shorter inode size (held for now in i_size) into the + * on-disk inode. We do this via i_disksize, which is the value which + * ext4 *really* writes onto the disk inode. + */ + ei->i_disksize = inode->i_size; + + if (last_block == max_block) { + /* + * It is unnecessary to free any data blocks if last_block is + * equal to the indirect block limit. + */ + goto out_unlock; + } else if (n == 1) { /* direct blocks */ + ext4_free_data(handle, inode, NULL, i_data+offsets[0], + i_data + EXT4_NDIR_BLOCKS); + goto do_indirects; + } + + partial = ext4_find_shared(inode, n, offsets, chain, &nr); + /* Kill the top of shared branch (not detached) */ + if (nr) { + if (partial == chain) { + /* Shared branch grows from the inode */ + ext4_free_branches(handle, inode, NULL, + &nr, &nr+1, (chain+n-1) - partial); + *partial->p = 0; + /* + * We mark the inode dirty prior to restart, + * and prior to stop. No need for it here. + */ + } else { + /* Shared branch grows from an indirect block */ + BUFFER_TRACE(partial->bh, "get_write_access"); + ext4_free_branches(handle, inode, partial->bh, + partial->p, + partial->p+1, (chain+n-1) - partial); + } + } + /* Clear the ends of indirect blocks on the shared branch */ + while (partial > chain) { + ext4_free_branches(handle, inode, partial->bh, partial->p + 1, + (__le32*)partial->bh->b_data+addr_per_block, + (chain+n-1) - partial); + BUFFER_TRACE(partial->bh, "call brelse"); + brelse(partial->bh); + partial--; + } +do_indirects: + /* Kill the remaining (whole) subtrees */ + switch (offsets[0]) { + default: + nr = i_data[EXT4_IND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 1); + i_data[EXT4_IND_BLOCK] = 0; + } + case EXT4_IND_BLOCK: + nr = i_data[EXT4_DIND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 2); + i_data[EXT4_DIND_BLOCK] = 0; + } + case EXT4_DIND_BLOCK: + nr = i_data[EXT4_TIND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 3); + i_data[EXT4_TIND_BLOCK] = 0; + } + case EXT4_TIND_BLOCK: + ; + } + +out_unlock: + up_write(&ei->i_data_sem); + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ext4_mark_inode_dirty(handle, inode); + + /* + * In a multi-transaction truncate, we only make the final transaction + * synchronous + */ + if (IS_SYNC(inode)) + ext4_handle_sync(handle); +out_stop: + /* + * If this was a simple ftruncate(), and the file will remain alive + * then we need to clear up the orphan record which we created above. + * However, if this was a real unlink then we were called by + * ext4_delete_inode(), and we allow that function to clean up the + * orphan info for us. + */ + if (inode->i_nlink) + ext4_orphan_del(handle, inode); + + ext4_journal_stop(handle); + trace_ext4_truncate_exit(inode); +} + diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 3e5191f9f398..d47264cafee0 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -12,10 +12,6 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * - * Goal-directed block allocation by Stephen Tweedie - * (sct@redhat.com), 1993, 1998 - * Big-endian to little-endian byte-swapping/bitmaps by - * David S. Miller (davem@caip.rutgers.edu), 1995 * 64-bit file support on 64-bit platforms by Jakub Jelinek * (jj@sunsite.ms.mff.cuni.cz) * @@ -47,6 +43,7 @@ #include "xattr.h" #include "acl.h" #include "ext4_extents.h" +#include "truncate.h" #include @@ -88,72 +85,6 @@ static int ext4_inode_is_fast_symlink(struct inode *inode) return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0); } -/* - * Work out how many blocks we need to proceed with the next chunk of a - * truncate transaction. - */ -static unsigned long blocks_for_truncate(struct inode *inode) -{ - ext4_lblk_t needed; - - needed = inode->i_blocks >> (inode->i_sb->s_blocksize_bits - 9); - - /* Give ourselves just enough room to cope with inodes in which - * i_blocks is corrupt: we've seen disk corruptions in the past - * which resulted in random data in an inode which looked enough - * like a regular file for ext4 to try to delete it. Things - * will go a bit crazy if that happens, but at least we should - * try not to panic the whole kernel. */ - if (needed < 2) - needed = 2; - - /* But we need to bound the transaction so we don't overflow the - * journal. */ - if (needed > EXT4_MAX_TRANS_DATA) - needed = EXT4_MAX_TRANS_DATA; - - return EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + needed; -} - -/* - * Truncate transactions can be complex and absolutely huge. So we need to - * be able to restart the transaction at a conventient checkpoint to make - * sure we don't overflow the journal. - * - * start_transaction gets us a new handle for a truncate transaction, - * and extend_transaction tries to extend the existing one a bit. If - * extend fails, we need to propagate the failure up and restart the - * transaction in the top-level truncate loop. --sct - */ -static handle_t *start_transaction(struct inode *inode) -{ - handle_t *result; - - result = ext4_journal_start(inode, blocks_for_truncate(inode)); - if (!IS_ERR(result)) - return result; - - ext4_std_error(inode->i_sb, PTR_ERR(result)); - return result; -} - -/* - * Try to extend this transaction for the purposes of truncation. - * - * Returns 0 if we managed to create more room. If we can't create more - * room, and the transaction must be restarted we return 1. - */ -static int try_to_extend_transaction(handle_t *handle, struct inode *inode) -{ - if (!ext4_handle_valid(handle)) - return 0; - if (ext4_handle_has_enough_credits(handle, EXT4_RESERVE_TRANS_BLOCKS+1)) - return 0; - if (!ext4_journal_extend(handle, blocks_for_truncate(inode))) - return 0; - return 1; -} - /* * Restart the transaction associated with *handle. This does a commit, * so before we call here everything must be consistently dirtied against @@ -190,6 +121,33 @@ void ext4_evict_inode(struct inode *inode) trace_ext4_evict_inode(inode); if (inode->i_nlink) { + /* + * When journalling data dirty buffers are tracked only in the + * journal. So although mm thinks everything is clean and + * ready for reaping the inode might still have some pages to + * write in the running transaction or waiting to be + * checkpointed. Thus calling jbd2_journal_invalidatepage() + * (via truncate_inode_pages()) to discard these buffers can + * cause data loss. Also even if we did not discard these + * buffers, we would have no way to find them after the inode + * is reaped and thus user could see stale data if he tries to + * read them before the transaction is checkpointed. So be + * careful and force everything to disk here... We use + * ei->i_datasync_tid to store the newest transaction + * containing inode's data. + * + * Note that directories do not have this problem because they + * don't use page cache. + */ + if (ext4_should_journal_data(inode) && + (S_ISLNK(inode->i_mode) || S_ISREG(inode->i_mode))) { + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; + tid_t commit_tid = EXT4_I(inode)->i_datasync_tid; + + jbd2_log_start_commit(journal, commit_tid); + jbd2_log_wait_commit(journal, commit_tid); + filemap_write_and_wait(&inode->i_data); + } truncate_inode_pages(&inode->i_data, 0); goto no_delete; } @@ -204,7 +162,7 @@ void ext4_evict_inode(struct inode *inode) if (is_bad_inode(inode)) goto no_delete; - handle = ext4_journal_start(inode, blocks_for_truncate(inode)+3); + handle = ext4_journal_start(inode, ext4_blocks_for_truncate(inode)+3); if (IS_ERR(handle)) { ext4_std_error(inode->i_sb, PTR_ERR(handle)); /* @@ -277,793 +235,6 @@ no_delete: ext4_clear_inode(inode); /* We must guarantee clearing of inode... */ } -typedef struct { - __le32 *p; - __le32 key; - struct buffer_head *bh; -} Indirect; - -static inline void add_chain(Indirect *p, struct buffer_head *bh, __le32 *v) -{ - p->key = *(p->p = v); - p->bh = bh; -} - -/** - * ext4_block_to_path - parse the block number into array of offsets - * @inode: inode in question (we are only interested in its superblock) - * @i_block: block number to be parsed - * @offsets: array to store the offsets in - * @boundary: set this non-zero if the referred-to block is likely to be - * followed (on disk) by an indirect block. - * - * To store the locations of file's data ext4 uses a data structure common - * for UNIX filesystems - tree of pointers anchored in the inode, with - * data blocks at leaves and indirect blocks in intermediate nodes. - * This function translates the block number into path in that tree - - * return value is the path length and @offsets[n] is the offset of - * pointer to (n+1)th node in the nth one. If @block is out of range - * (negative or too large) warning is printed and zero returned. - * - * Note: function doesn't find node addresses, so no IO is needed. All - * we need to know is the capacity of indirect blocks (taken from the - * inode->i_sb). - */ - -/* - * Portability note: the last comparison (check that we fit into triple - * indirect block) is spelled differently, because otherwise on an - * architecture with 32-bit longs and 8Kb pages we might get into trouble - * if our filesystem had 8Kb blocks. We might use long long, but that would - * kill us on x86. Oh, well, at least the sign propagation does not matter - - * i_block would have to be negative in the very beginning, so we would not - * get there at all. - */ - -static int ext4_block_to_path(struct inode *inode, - ext4_lblk_t i_block, - ext4_lblk_t offsets[4], int *boundary) -{ - int ptrs = EXT4_ADDR_PER_BLOCK(inode->i_sb); - int ptrs_bits = EXT4_ADDR_PER_BLOCK_BITS(inode->i_sb); - const long direct_blocks = EXT4_NDIR_BLOCKS, - indirect_blocks = ptrs, - double_blocks = (1 << (ptrs_bits * 2)); - int n = 0; - int final = 0; - - if (i_block < direct_blocks) { - offsets[n++] = i_block; - final = direct_blocks; - } else if ((i_block -= direct_blocks) < indirect_blocks) { - offsets[n++] = EXT4_IND_BLOCK; - offsets[n++] = i_block; - final = ptrs; - } else if ((i_block -= indirect_blocks) < double_blocks) { - offsets[n++] = EXT4_DIND_BLOCK; - offsets[n++] = i_block >> ptrs_bits; - offsets[n++] = i_block & (ptrs - 1); - final = ptrs; - } else if (((i_block -= double_blocks) >> (ptrs_bits * 2)) < ptrs) { - offsets[n++] = EXT4_TIND_BLOCK; - offsets[n++] = i_block >> (ptrs_bits * 2); - offsets[n++] = (i_block >> ptrs_bits) & (ptrs - 1); - offsets[n++] = i_block & (ptrs - 1); - final = ptrs; - } else { - ext4_warning(inode->i_sb, "block %lu > max in inode %lu", - i_block + direct_blocks + - indirect_blocks + double_blocks, inode->i_ino); - } - if (boundary) - *boundary = final - 1 - (i_block & (ptrs - 1)); - return n; -} - -static int __ext4_check_blockref(const char *function, unsigned int line, - struct inode *inode, - __le32 *p, unsigned int max) -{ - struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es; - __le32 *bref = p; - unsigned int blk; - - while (bref < p+max) { - blk = le32_to_cpu(*bref++); - if (blk && - unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb), - blk, 1))) { - es->s_last_error_block = cpu_to_le64(blk); - ext4_error_inode(inode, function, line, blk, - "invalid block"); - return -EIO; - } - } - return 0; -} - - -#define ext4_check_indirect_blockref(inode, bh) \ - __ext4_check_blockref(__func__, __LINE__, inode, \ - (__le32 *)(bh)->b_data, \ - EXT4_ADDR_PER_BLOCK((inode)->i_sb)) - -#define ext4_check_inode_blockref(inode) \ - __ext4_check_blockref(__func__, __LINE__, inode, \ - EXT4_I(inode)->i_data, \ - EXT4_NDIR_BLOCKS) - -/** - * ext4_get_branch - read the chain of indirect blocks leading to data - * @inode: inode in question - * @depth: depth of the chain (1 - direct pointer, etc.) - * @offsets: offsets of pointers in inode/indirect blocks - * @chain: place to store the result - * @err: here we store the error value - * - * Function fills the array of triples and returns %NULL - * if everything went OK or the pointer to the last filled triple - * (incomplete one) otherwise. Upon the return chain[i].key contains - * the number of (i+1)-th block in the chain (as it is stored in memory, - * i.e. little-endian 32-bit), chain[i].p contains the address of that - * number (it points into struct inode for i==0 and into the bh->b_data - * for i>0) and chain[i].bh points to the buffer_head of i-th indirect - * block for i>0 and NULL for i==0. In other words, it holds the block - * numbers of the chain, addresses they were taken from (and where we can - * verify that chain did not change) and buffer_heads hosting these - * numbers. - * - * Function stops when it stumbles upon zero pointer (absent block) - * (pointer to last triple returned, *@err == 0) - * or when it gets an IO error reading an indirect block - * (ditto, *@err == -EIO) - * or when it reads all @depth-1 indirect blocks successfully and finds - * the whole chain, all way to the data (returns %NULL, *err == 0). - * - * Need to be called with - * down_read(&EXT4_I(inode)->i_data_sem) - */ -static Indirect *ext4_get_branch(struct inode *inode, int depth, - ext4_lblk_t *offsets, - Indirect chain[4], int *err) -{ - struct super_block *sb = inode->i_sb; - Indirect *p = chain; - struct buffer_head *bh; - - *err = 0; - /* i_data is not going away, no lock needed */ - add_chain(chain, NULL, EXT4_I(inode)->i_data + *offsets); - if (!p->key) - goto no_block; - while (--depth) { - bh = sb_getblk(sb, le32_to_cpu(p->key)); - if (unlikely(!bh)) - goto failure; - - if (!bh_uptodate_or_lock(bh)) { - if (bh_submit_read(bh) < 0) { - put_bh(bh); - goto failure; - } - /* validate block references */ - if (ext4_check_indirect_blockref(inode, bh)) { - put_bh(bh); - goto failure; - } - } - - add_chain(++p, bh, (__le32 *)bh->b_data + *++offsets); - /* Reader: end */ - if (!p->key) - goto no_block; - } - return NULL; - -failure: - *err = -EIO; -no_block: - return p; -} - -/** - * ext4_find_near - find a place for allocation with sufficient locality - * @inode: owner - * @ind: descriptor of indirect block. - * - * This function returns the preferred place for block allocation. - * It is used when heuristic for sequential allocation fails. - * Rules are: - * + if there is a block to the left of our position - allocate near it. - * + if pointer will live in indirect block - allocate near that block. - * + if pointer will live in inode - allocate in the same - * cylinder group. - * - * In the latter case we colour the starting block by the callers PID to - * prevent it from clashing with concurrent allocations for a different inode - * in the same block group. The PID is used here so that functionally related - * files will be close-by on-disk. - * - * Caller must make sure that @ind is valid and will stay that way. - */ -static ext4_fsblk_t ext4_find_near(struct inode *inode, Indirect *ind) -{ - struct ext4_inode_info *ei = EXT4_I(inode); - __le32 *start = ind->bh ? (__le32 *) ind->bh->b_data : ei->i_data; - __le32 *p; - ext4_fsblk_t bg_start; - ext4_fsblk_t last_block; - ext4_grpblk_t colour; - ext4_group_t block_group; - int flex_size = ext4_flex_bg_size(EXT4_SB(inode->i_sb)); - - /* Try to find previous block */ - for (p = ind->p - 1; p >= start; p--) { - if (*p) - return le32_to_cpu(*p); - } - - /* No such thing, so let's try location of indirect block */ - if (ind->bh) - return ind->bh->b_blocknr; - - /* - * It is going to be referred to from the inode itself? OK, just put it - * into the same cylinder group then. - */ - block_group = ei->i_block_group; - if (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) { - block_group &= ~(flex_size-1); - if (S_ISREG(inode->i_mode)) - block_group++; - } - bg_start = ext4_group_first_block_no(inode->i_sb, block_group); - last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1; - - /* - * If we are doing delayed allocation, we don't need take - * colour into account. - */ - if (test_opt(inode->i_sb, DELALLOC)) - return bg_start; - - if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block) - colour = (current->pid % 16) * - (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16); - else - colour = (current->pid % 16) * ((last_block - bg_start) / 16); - return bg_start + colour; -} - -/** - * ext4_find_goal - find a preferred place for allocation. - * @inode: owner - * @block: block we want - * @partial: pointer to the last triple within a chain - * - * Normally this function find the preferred place for block allocation, - * returns it. - * Because this is only used for non-extent files, we limit the block nr - * to 32 bits. - */ -static ext4_fsblk_t ext4_find_goal(struct inode *inode, ext4_lblk_t block, - Indirect *partial) -{ - ext4_fsblk_t goal; - - /* - * XXX need to get goal block from mballoc's data structures - */ - - goal = ext4_find_near(inode, partial); - goal = goal & EXT4_MAX_BLOCK_FILE_PHYS; - return goal; -} - -/** - * ext4_blks_to_allocate - Look up the block map and count the number - * of direct blocks need to be allocated for the given branch. - * - * @branch: chain of indirect blocks - * @k: number of blocks need for indirect blocks - * @blks: number of data blocks to be mapped. - * @blocks_to_boundary: the offset in the indirect block - * - * return the total number of blocks to be allocate, including the - * direct and indirect blocks. - */ -static int ext4_blks_to_allocate(Indirect *branch, int k, unsigned int blks, - int blocks_to_boundary) -{ - unsigned int count = 0; - - /* - * Simple case, [t,d]Indirect block(s) has not allocated yet - * then it's clear blocks on that path have not allocated - */ - if (k > 0) { - /* right now we don't handle cross boundary allocation */ - if (blks < blocks_to_boundary + 1) - count += blks; - else - count += blocks_to_boundary + 1; - return count; - } - - count++; - while (count < blks && count <= blocks_to_boundary && - le32_to_cpu(*(branch[0].p + count)) == 0) { - count++; - } - return count; -} - -/** - * ext4_alloc_blocks: multiple allocate blocks needed for a branch - * @handle: handle for this transaction - * @inode: inode which needs allocated blocks - * @iblock: the logical block to start allocated at - * @goal: preferred physical block of allocation - * @indirect_blks: the number of blocks need to allocate for indirect - * blocks - * @blks: number of desired blocks - * @new_blocks: on return it will store the new block numbers for - * the indirect blocks(if needed) and the first direct block, - * @err: on return it will store the error code - * - * This function will return the number of blocks allocated as - * requested by the passed-in parameters. - */ -static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, - ext4_lblk_t iblock, ext4_fsblk_t goal, - int indirect_blks, int blks, - ext4_fsblk_t new_blocks[4], int *err) -{ - struct ext4_allocation_request ar; - int target, i; - unsigned long count = 0, blk_allocated = 0; - int index = 0; - ext4_fsblk_t current_block = 0; - int ret = 0; - - /* - * Here we try to allocate the requested multiple blocks at once, - * on a best-effort basis. - * To build a branch, we should allocate blocks for - * the indirect blocks(if not allocated yet), and at least - * the first direct block of this branch. That's the - * minimum number of blocks need to allocate(required) - */ - /* first we try to allocate the indirect blocks */ - target = indirect_blks; - while (target > 0) { - count = target; - /* allocating blocks for indirect blocks and direct blocks */ - current_block = ext4_new_meta_blocks(handle, inode, goal, - 0, &count, err); - if (*err) - goto failed_out; - - if (unlikely(current_block + count > EXT4_MAX_BLOCK_FILE_PHYS)) { - EXT4_ERROR_INODE(inode, - "current_block %llu + count %lu > %d!", - current_block, count, - EXT4_MAX_BLOCK_FILE_PHYS); - *err = -EIO; - goto failed_out; - } - - target -= count; - /* allocate blocks for indirect blocks */ - while (index < indirect_blks && count) { - new_blocks[index++] = current_block++; - count--; - } - if (count > 0) { - /* - * save the new block number - * for the first direct block - */ - new_blocks[index] = current_block; - printk(KERN_INFO "%s returned more blocks than " - "requested\n", __func__); - WARN_ON(1); - break; - } - } - - target = blks - count ; - blk_allocated = count; - if (!target) - goto allocated; - /* Now allocate data blocks */ - memset(&ar, 0, sizeof(ar)); - ar.inode = inode; - ar.goal = goal; - ar.len = target; - ar.logical = iblock; - if (S_ISREG(inode->i_mode)) - /* enable in-core preallocation only for regular files */ - ar.flags = EXT4_MB_HINT_DATA; - - current_block = ext4_mb_new_blocks(handle, &ar, err); - if (unlikely(current_block + ar.len > EXT4_MAX_BLOCK_FILE_PHYS)) { - EXT4_ERROR_INODE(inode, - "current_block %llu + ar.len %d > %d!", - current_block, ar.len, - EXT4_MAX_BLOCK_FILE_PHYS); - *err = -EIO; - goto failed_out; - } - - if (*err && (target == blks)) { - /* - * if the allocation failed and we didn't allocate - * any blocks before - */ - goto failed_out; - } - if (!*err) { - if (target == blks) { - /* - * save the new block number - * for the first direct block - */ - new_blocks[index] = current_block; - } - blk_allocated += ar.len; - } -allocated: - /* total number of blocks allocated for direct blocks */ - ret = blk_allocated; - *err = 0; - return ret; -failed_out: - for (i = 0; i < index; i++) - ext4_free_blocks(handle, inode, NULL, new_blocks[i], 1, 0); - return ret; -} - -/** - * ext4_alloc_branch - allocate and set up a chain of blocks. - * @handle: handle for this transaction - * @inode: owner - * @indirect_blks: number of allocated indirect blocks - * @blks: number of allocated direct blocks - * @goal: preferred place for allocation - * @offsets: offsets (in the blocks) to store the pointers to next. - * @branch: place to store the chain in. - * - * This function allocates blocks, zeroes out all but the last one, - * links them into chain and (if we are synchronous) writes them to disk. - * In other words, it prepares a branch that can be spliced onto the - * inode. It stores the information about that chain in the branch[], in - * the same format as ext4_get_branch() would do. We are calling it after - * we had read the existing part of chain and partial points to the last - * triple of that (one with zero ->key). Upon the exit we have the same - * picture as after the successful ext4_get_block(), except that in one - * place chain is disconnected - *branch->p is still zero (we did not - * set the last link), but branch->key contains the number that should - * be placed into *branch->p to fill that gap. - * - * If allocation fails we free all blocks we've allocated (and forget - * their buffer_heads) and return the error value the from failed - * ext4_alloc_block() (normally -ENOSPC). Otherwise we set the chain - * as described above and return 0. - */ -static int ext4_alloc_branch(handle_t *handle, struct inode *inode, - ext4_lblk_t iblock, int indirect_blks, - int *blks, ext4_fsblk_t goal, - ext4_lblk_t *offsets, Indirect *branch) -{ - int blocksize = inode->i_sb->s_blocksize; - int i, n = 0; - int err = 0; - struct buffer_head *bh; - int num; - ext4_fsblk_t new_blocks[4]; - ext4_fsblk_t current_block; - - num = ext4_alloc_blocks(handle, inode, iblock, goal, indirect_blks, - *blks, new_blocks, &err); - if (err) - return err; - - branch[0].key = cpu_to_le32(new_blocks[0]); - /* - * metadata blocks and data blocks are allocated. - */ - for (n = 1; n <= indirect_blks; n++) { - /* - * Get buffer_head for parent block, zero it out - * and set the pointer to new one, then send - * parent to disk. - */ - bh = sb_getblk(inode->i_sb, new_blocks[n-1]); - if (unlikely(!bh)) { - err = -EIO; - goto failed; - } - - branch[n].bh = bh; - lock_buffer(bh); - BUFFER_TRACE(bh, "call get_create_access"); - err = ext4_journal_get_create_access(handle, bh); - if (err) { - /* Don't brelse(bh) here; it's done in - * ext4_journal_forget() below */ - unlock_buffer(bh); - goto failed; - } - - memset(bh->b_data, 0, blocksize); - branch[n].p = (__le32 *) bh->b_data + offsets[n]; - branch[n].key = cpu_to_le32(new_blocks[n]); - *branch[n].p = branch[n].key; - if (n == indirect_blks) { - current_block = new_blocks[n]; - /* - * End of chain, update the last new metablock of - * the chain to point to the new allocated - * data blocks numbers - */ - for (i = 1; i < num; i++) - *(branch[n].p + i) = cpu_to_le32(++current_block); - } - BUFFER_TRACE(bh, "marking uptodate"); - set_buffer_uptodate(bh); - unlock_buffer(bh); - - BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); - err = ext4_handle_dirty_metadata(handle, inode, bh); - if (err) - goto failed; - } - *blks = num; - return err; -failed: - /* Allocation failed, free what we already allocated */ - ext4_free_blocks(handle, inode, NULL, new_blocks[0], 1, 0); - for (i = 1; i <= n ; i++) { - /* - * branch[i].bh is newly allocated, so there is no - * need to revoke the block, which is why we don't - * need to set EXT4_FREE_BLOCKS_METADATA. - */ - ext4_free_blocks(handle, inode, NULL, new_blocks[i], 1, - EXT4_FREE_BLOCKS_FORGET); - } - for (i = n+1; i < indirect_blks; i++) - ext4_free_blocks(handle, inode, NULL, new_blocks[i], 1, 0); - - ext4_free_blocks(handle, inode, NULL, new_blocks[i], num, 0); - - return err; -} - -/** - * ext4_splice_branch - splice the allocated branch onto inode. - * @handle: handle for this transaction - * @inode: owner - * @block: (logical) number of block we are adding - * @chain: chain of indirect blocks (with a missing link - see - * ext4_alloc_branch) - * @where: location of missing link - * @num: number of indirect blocks we are adding - * @blks: number of direct blocks we are adding - * - * This function fills the missing link and does all housekeeping needed in - * inode (->i_blocks, etc.). In case of success we end up with the full - * chain to new block and return 0. - */ -static int ext4_splice_branch(handle_t *handle, struct inode *inode, - ext4_lblk_t block, Indirect *where, int num, - int blks) -{ - int i; - int err = 0; - ext4_fsblk_t current_block; - - /* - * If we're splicing into a [td]indirect block (as opposed to the - * inode) then we need to get write access to the [td]indirect block - * before the splice. - */ - if (where->bh) { - BUFFER_TRACE(where->bh, "get_write_access"); - err = ext4_journal_get_write_access(handle, where->bh); - if (err) - goto err_out; - } - /* That's it */ - - *where->p = where->key; - - /* - * Update the host buffer_head or inode to point to more just allocated - * direct blocks blocks - */ - if (num == 0 && blks > 1) { - current_block = le32_to_cpu(where->key) + 1; - for (i = 1; i < blks; i++) - *(where->p + i) = cpu_to_le32(current_block++); - } - - /* We are done with atomic stuff, now do the rest of housekeeping */ - /* had we spliced it onto indirect block? */ - if (where->bh) { - /* - * If we spliced it onto an indirect block, we haven't - * altered the inode. Note however that if it is being spliced - * onto an indirect block at the very end of the file (the - * file is growing) then we *will* alter the inode to reflect - * the new i_size. But that is not done here - it is done in - * generic_commit_write->__mark_inode_dirty->ext4_dirty_inode. - */ - jbd_debug(5, "splicing indirect only\n"); - BUFFER_TRACE(where->bh, "call ext4_handle_dirty_metadata"); - err = ext4_handle_dirty_metadata(handle, inode, where->bh); - if (err) - goto err_out; - } else { - /* - * OK, we spliced it into the inode itself on a direct block. - */ - ext4_mark_inode_dirty(handle, inode); - jbd_debug(5, "splicing direct\n"); - } - return err; - -err_out: - for (i = 1; i <= num; i++) { - /* - * branch[i].bh is newly allocated, so there is no - * need to revoke the block, which is why we don't - * need to set EXT4_FREE_BLOCKS_METADATA. - */ - ext4_free_blocks(handle, inode, where[i].bh, 0, 1, - EXT4_FREE_BLOCKS_FORGET); - } - ext4_free_blocks(handle, inode, NULL, le32_to_cpu(where[num].key), - blks, 0); - - return err; -} - -/* - * The ext4_ind_map_blocks() function handles non-extents inodes - * (i.e., using the traditional indirect/double-indirect i_blocks - * scheme) for ext4_map_blocks(). - * - * Allocation strategy is simple: if we have to allocate something, we will - * have to go the whole way to leaf. So let's do it before attaching anything - * to tree, set linkage between the newborn blocks, write them if sync is - * required, recheck the path, free and repeat if check fails, otherwise - * set the last missing link (that will protect us from any truncate-generated - * removals - all blocks on the path are immune now) and possibly force the - * write on the parent block. - * That has a nice additional property: no special recovery from the failed - * allocations is needed - we simply release blocks and do not touch anything - * reachable from inode. - * - * `handle' can be NULL if create == 0. - * - * return > 0, # of blocks mapped or allocated. - * return = 0, if plain lookup failed. - * return < 0, error case. - * - * The ext4_ind_get_blocks() function should be called with - * down_write(&EXT4_I(inode)->i_data_sem) if allocating filesystem - * blocks (i.e., flags has EXT4_GET_BLOCKS_CREATE set) or - * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system - * blocks. - */ -static int ext4_ind_map_blocks(handle_t *handle, struct inode *inode, - struct ext4_map_blocks *map, - int flags) -{ - int err = -EIO; - ext4_lblk_t offsets[4]; - Indirect chain[4]; - Indirect *partial; - ext4_fsblk_t goal; - int indirect_blks; - int blocks_to_boundary = 0; - int depth; - int count = 0; - ext4_fsblk_t first_block = 0; - - trace_ext4_ind_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); - J_ASSERT(!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))); - J_ASSERT(handle != NULL || (flags & EXT4_GET_BLOCKS_CREATE) == 0); - depth = ext4_block_to_path(inode, map->m_lblk, offsets, - &blocks_to_boundary); - - if (depth == 0) - goto out; - - partial = ext4_get_branch(inode, depth, offsets, chain, &err); - - /* Simplest case - block found, no allocation needed */ - if (!partial) { - first_block = le32_to_cpu(chain[depth - 1].key); - count++; - /*map more blocks*/ - while (count < map->m_len && count <= blocks_to_boundary) { - ext4_fsblk_t blk; - - blk = le32_to_cpu(*(chain[depth-1].p + count)); - - if (blk == first_block + count) - count++; - else - break; - } - goto got_it; - } - - /* Next simple case - plain lookup or failed read of indirect block */ - if ((flags & EXT4_GET_BLOCKS_CREATE) == 0 || err == -EIO) - goto cleanup; - - /* - * Okay, we need to do block allocation. - */ - goal = ext4_find_goal(inode, map->m_lblk, partial); - - /* the number of blocks need to allocate for [d,t]indirect blocks */ - indirect_blks = (chain + depth) - partial - 1; - - /* - * Next look up the indirect map to count the totoal number of - * direct blocks to allocate for this branch. - */ - count = ext4_blks_to_allocate(partial, indirect_blks, - map->m_len, blocks_to_boundary); - /* - * Block out ext4_truncate while we alter the tree - */ - err = ext4_alloc_branch(handle, inode, map->m_lblk, indirect_blks, - &count, goal, - offsets + (partial - chain), partial); - - /* - * The ext4_splice_branch call will free and forget any buffers - * on the new chain if there is a failure, but that risks using - * up transaction credits, especially for bitmaps where the - * credits cannot be returned. Can we handle this somehow? We - * may need to return -EAGAIN upwards in the worst case. --sct - */ - if (!err) - err = ext4_splice_branch(handle, inode, map->m_lblk, - partial, indirect_blks, count); - if (err) - goto cleanup; - - map->m_flags |= EXT4_MAP_NEW; - - ext4_update_inode_fsync_trans(handle, inode, 1); -got_it: - map->m_flags |= EXT4_MAP_MAPPED; - map->m_pblk = le32_to_cpu(chain[depth-1].key); - map->m_len = count; - if (count > blocks_to_boundary) - map->m_flags |= EXT4_MAP_BOUNDARY; - err = count; - /* Clean up and exit */ - partial = chain + depth - 1; /* the whole chain */ -cleanup: - while (partial > chain) { - BUFFER_TRACE(partial->bh, "call brelse"); - brelse(partial->bh); - partial--; - } -out: - trace_ext4_ind_map_blocks_exit(inode, map->m_lblk, - map->m_pblk, map->m_len, err); - return err; -} - #ifdef CONFIG_QUOTA qsize_t *ext4_get_reserved_space(struct inode *inode) { @@ -1071,33 +242,6 @@ qsize_t *ext4_get_reserved_space(struct inode *inode) } #endif -/* - * Calculate the number of metadata blocks need to reserve - * to allocate a new block at @lblocks for non extent file based file - */ -static int ext4_indirect_calc_metadata_amount(struct inode *inode, - sector_t lblock) -{ - struct ext4_inode_info *ei = EXT4_I(inode); - sector_t dind_mask = ~((sector_t)EXT4_ADDR_PER_BLOCK(inode->i_sb) - 1); - int blk_bits; - - if (lblock < EXT4_NDIR_BLOCKS) - return 0; - - lblock -= EXT4_NDIR_BLOCKS; - - if (ei->i_da_metadata_calc_len && - (lblock & dind_mask) == ei->i_da_metadata_calc_last_lblock) { - ei->i_da_metadata_calc_len++; - return 0; - } - ei->i_da_metadata_calc_last_lblock = lblock & dind_mask; - ei->i_da_metadata_calc_len = 1; - blk_bits = order_base_2(lblock); - return (blk_bits / EXT4_ADDR_PER_BLOCK_BITS(inode->i_sb)) + 1; -} - /* * Calculate the number of metadata blocks need to reserve * to allocate a block located at @lblock @@ -1107,7 +251,7 @@ static int ext4_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock) if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) return ext4_ext_calc_metadata_amount(inode, lblock); - return ext4_indirect_calc_metadata_amount(inode, lblock); + return ext4_ind_calc_metadata_amount(inode, lblock); } /* @@ -1589,16 +733,6 @@ static int do_journal_get_write_access(handle_t *handle, return ret; } -/* - * Truncate blocks that were not used by write. We have to truncate the - * pagecache as well so that corresponding buffers get properly unmapped. - */ -static void ext4_truncate_failed_write(struct inode *inode) -{ - truncate_inode_pages(inode->i_mapping, inode->i_size); - ext4_truncate(inode); -} - static int ext4_get_block_write(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create); static int ext4_write_begin(struct file *file, struct address_space *mapping, @@ -1863,6 +997,7 @@ static int ext4_journalled_write_end(struct file *file, if (new_i_size > inode->i_size) i_size_write(inode, pos+copied); ext4_set_inode_state(inode, EXT4_STATE_JDATA); + EXT4_I(inode)->i_datasync_tid = handle->h_transaction->t_tid; if (new_i_size > EXT4_I(inode)->i_disksize) { ext4_update_i_disksize(inode, new_i_size); ret2 = ext4_mark_inode_dirty(handle, inode); @@ -2571,6 +1706,7 @@ static int __ext4_journalled_writepage(struct page *page, write_end_fn); if (ret == 0) ret = err; + EXT4_I(inode)->i_datasync_tid = handle->h_transaction->t_tid; err = ext4_journal_stop(handle); if (!ret) ret = err; @@ -3449,112 +2585,6 @@ static int ext4_releasepage(struct page *page, gfp_t wait) return try_to_free_buffers(page); } -/* - * O_DIRECT for ext3 (or indirect map) based files - * - * If the O_DIRECT write will extend the file then add this inode to the - * orphan list. So recovery will truncate it back to the original size - * if the machine crashes during the write. - * - * If the O_DIRECT write is intantiating holes inside i_size and the machine - * crashes then stale disk data _may_ be exposed inside the file. But current - * VFS code falls back into buffered path in that case so we are safe. - */ -static ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) -{ - struct file *file = iocb->ki_filp; - struct inode *inode = file->f_mapping->host; - struct ext4_inode_info *ei = EXT4_I(inode); - handle_t *handle; - ssize_t ret; - int orphan = 0; - size_t count = iov_length(iov, nr_segs); - int retries = 0; - - if (rw == WRITE) { - loff_t final_size = offset + count; - - if (final_size > inode->i_size) { - /* Credits for sb + inode write */ - handle = ext4_journal_start(inode, 2); - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto out; - } - ret = ext4_orphan_add(handle, inode); - if (ret) { - ext4_journal_stop(handle); - goto out; - } - orphan = 1; - ei->i_disksize = inode->i_size; - ext4_journal_stop(handle); - } - } - -retry: - if (rw == READ && ext4_should_dioread_nolock(inode)) - ret = __blockdev_direct_IO(rw, iocb, inode, - inode->i_sb->s_bdev, iov, - offset, nr_segs, - ext4_get_block, NULL, NULL, 0); - else { - ret = blockdev_direct_IO(rw, iocb, inode, iov, - offset, nr_segs, ext4_get_block); - - if (unlikely((rw & WRITE) && ret < 0)) { - loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); - - if (end > isize) - ext4_truncate_failed_write(inode); - } - } - if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) - goto retry; - - if (orphan) { - int err; - - /* Credits for sb + inode write */ - handle = ext4_journal_start(inode, 2); - if (IS_ERR(handle)) { - /* This is really bad luck. We've written the data - * but cannot extend i_size. Bail out and pretend - * the write failed... */ - ret = PTR_ERR(handle); - if (inode->i_nlink) - ext4_orphan_del(NULL, inode); - - goto out; - } - if (inode->i_nlink) - ext4_orphan_del(handle, inode); - if (ret > 0) { - loff_t end = offset + ret; - if (end > inode->i_size) { - ei->i_disksize = end; - i_size_write(inode, end); - /* - * We're going to return a positive `ret' - * here due to non-zero-length I/O, so there's - * no way of reporting error returns from - * ext4_mark_inode_dirty() to userspace. So - * ignore it. - */ - ext4_mark_inode_dirty(handle, inode); - } - } - err = ext4_journal_stop(handle); - if (ret == 0) - ret = err; - } -out: - return ret; -} - /* * ext4_get_block used when preparing for a DIO write or buffer write. * We allocate an uinitialized extent if blocks haven't been allocated. @@ -4033,383 +3063,6 @@ unlock: return err; } -/* - * Probably it should be a library function... search for first non-zero word - * or memcmp with zero_page, whatever is better for particular architecture. - * Linus? - */ -static inline int all_zeroes(__le32 *p, __le32 *q) -{ - while (p < q) - if (*p++) - return 0; - return 1; -} - -/** - * ext4_find_shared - find the indirect blocks for partial truncation. - * @inode: inode in question - * @depth: depth of the affected branch - * @offsets: offsets of pointers in that branch (see ext4_block_to_path) - * @chain: place to store the pointers to partial indirect blocks - * @top: place to the (detached) top of branch - * - * This is a helper function used by ext4_truncate(). - * - * When we do truncate() we may have to clean the ends of several - * indirect blocks but leave the blocks themselves alive. Block is - * partially truncated if some data below the new i_size is referred - * from it (and it is on the path to the first completely truncated - * data block, indeed). We have to free the top of that path along - * with everything to the right of the path. Since no allocation - * past the truncation point is possible until ext4_truncate() - * finishes, we may safely do the latter, but top of branch may - * require special attention - pageout below the truncation point - * might try to populate it. - * - * We atomically detach the top of branch from the tree, store the - * block number of its root in *@top, pointers to buffer_heads of - * partially truncated blocks - in @chain[].bh and pointers to - * their last elements that should not be removed - in - * @chain[].p. Return value is the pointer to last filled element - * of @chain. - * - * The work left to caller to do the actual freeing of subtrees: - * a) free the subtree starting from *@top - * b) free the subtrees whose roots are stored in - * (@chain[i].p+1 .. end of @chain[i].bh->b_data) - * c) free the subtrees growing from the inode past the @chain[0]. - * (no partially truncated stuff there). */ - -static Indirect *ext4_find_shared(struct inode *inode, int depth, - ext4_lblk_t offsets[4], Indirect chain[4], - __le32 *top) -{ - Indirect *partial, *p; - int k, err; - - *top = 0; - /* Make k index the deepest non-null offset + 1 */ - for (k = depth; k > 1 && !offsets[k-1]; k--) - ; - partial = ext4_get_branch(inode, k, offsets, chain, &err); - /* Writer: pointers */ - if (!partial) - partial = chain + k-1; - /* - * If the branch acquired continuation since we've looked at it - - * fine, it should all survive and (new) top doesn't belong to us. - */ - if (!partial->key && *partial->p) - /* Writer: end */ - goto no_top; - for (p = partial; (p > chain) && all_zeroes((__le32 *) p->bh->b_data, p->p); p--) - ; - /* - * OK, we've found the last block that must survive. The rest of our - * branch should be detached before unlocking. However, if that rest - * of branch is all ours and does not grow immediately from the inode - * it's easier to cheat and just decrement partial->p. - */ - if (p == chain + k - 1 && p > chain) { - p->p--; - } else { - *top = *p->p; - /* Nope, don't do this in ext4. Must leave the tree intact */ -#if 0 - *p->p = 0; -#endif - } - /* Writer: end */ - - while (partial > p) { - brelse(partial->bh); - partial--; - } -no_top: - return partial; -} - -/* - * Zero a number of block pointers in either an inode or an indirect block. - * If we restart the transaction we must again get write access to the - * indirect block for further modification. - * - * We release `count' blocks on disk, but (last - first) may be greater - * than `count' because there can be holes in there. - * - * Return 0 on success, 1 on invalid block range - * and < 0 on fatal error. - */ -static int ext4_clear_blocks(handle_t *handle, struct inode *inode, - struct buffer_head *bh, - ext4_fsblk_t block_to_free, - unsigned long count, __le32 *first, - __le32 *last) -{ - __le32 *p; - int flags = EXT4_FREE_BLOCKS_FORGET | EXT4_FREE_BLOCKS_VALIDATED; - int err; - - if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) - flags |= EXT4_FREE_BLOCKS_METADATA; - - if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), block_to_free, - count)) { - EXT4_ERROR_INODE(inode, "attempt to clear invalid " - "blocks %llu len %lu", - (unsigned long long) block_to_free, count); - return 1; - } - - if (try_to_extend_transaction(handle, inode)) { - if (bh) { - BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); - err = ext4_handle_dirty_metadata(handle, inode, bh); - if (unlikely(err)) - goto out_err; - } - err = ext4_mark_inode_dirty(handle, inode); - if (unlikely(err)) - goto out_err; - err = ext4_truncate_restart_trans(handle, inode, - blocks_for_truncate(inode)); - if (unlikely(err)) - goto out_err; - if (bh) { - BUFFER_TRACE(bh, "retaking write access"); - err = ext4_journal_get_write_access(handle, bh); - if (unlikely(err)) - goto out_err; - } - } - - for (p = first; p < last; p++) - *p = 0; - - ext4_free_blocks(handle, inode, NULL, block_to_free, count, flags); - return 0; -out_err: - ext4_std_error(inode->i_sb, err); - return err; -} - -/** - * ext4_free_data - free a list of data blocks - * @handle: handle for this transaction - * @inode: inode we are dealing with - * @this_bh: indirect buffer_head which contains *@first and *@last - * @first: array of block numbers - * @last: points immediately past the end of array - * - * We are freeing all blocks referred from that array (numbers are stored as - * little-endian 32-bit) and updating @inode->i_blocks appropriately. - * - * We accumulate contiguous runs of blocks to free. Conveniently, if these - * blocks are contiguous then releasing them at one time will only affect one - * or two bitmap blocks (+ group descriptor(s) and superblock) and we won't - * actually use a lot of journal space. - * - * @this_bh will be %NULL if @first and @last point into the inode's direct - * block pointers. - */ -static void ext4_free_data(handle_t *handle, struct inode *inode, - struct buffer_head *this_bh, - __le32 *first, __le32 *last) -{ - ext4_fsblk_t block_to_free = 0; /* Starting block # of a run */ - unsigned long count = 0; /* Number of blocks in the run */ - __le32 *block_to_free_p = NULL; /* Pointer into inode/ind - corresponding to - block_to_free */ - ext4_fsblk_t nr; /* Current block # */ - __le32 *p; /* Pointer into inode/ind - for current block */ - int err = 0; - - if (this_bh) { /* For indirect block */ - BUFFER_TRACE(this_bh, "get_write_access"); - err = ext4_journal_get_write_access(handle, this_bh); - /* Important: if we can't update the indirect pointers - * to the blocks, we can't free them. */ - if (err) - return; - } - - for (p = first; p < last; p++) { - nr = le32_to_cpu(*p); - if (nr) { - /* accumulate blocks to free if they're contiguous */ - if (count == 0) { - block_to_free = nr; - block_to_free_p = p; - count = 1; - } else if (nr == block_to_free + count) { - count++; - } else { - err = ext4_clear_blocks(handle, inode, this_bh, - block_to_free, count, - block_to_free_p, p); - if (err) - break; - block_to_free = nr; - block_to_free_p = p; - count = 1; - } - } - } - - if (!err && count > 0) - err = ext4_clear_blocks(handle, inode, this_bh, block_to_free, - count, block_to_free_p, p); - if (err < 0) - /* fatal error */ - return; - - if (this_bh) { - BUFFER_TRACE(this_bh, "call ext4_handle_dirty_metadata"); - - /* - * The buffer head should have an attached journal head at this - * point. However, if the data is corrupted and an indirect - * block pointed to itself, it would have been detached when - * the block was cleared. Check for this instead of OOPSing. - */ - if ((EXT4_JOURNAL(inode) == NULL) || bh2jh(this_bh)) - ext4_handle_dirty_metadata(handle, inode, this_bh); - else - EXT4_ERROR_INODE(inode, - "circular indirect block detected at " - "block %llu", - (unsigned long long) this_bh->b_blocknr); - } -} - -/** - * ext4_free_branches - free an array of branches - * @handle: JBD handle for this transaction - * @inode: inode we are dealing with - * @parent_bh: the buffer_head which contains *@first and *@last - * @first: array of block numbers - * @last: pointer immediately past the end of array - * @depth: depth of the branches to free - * - * We are freeing all blocks referred from these branches (numbers are - * stored as little-endian 32-bit) and updating @inode->i_blocks - * appropriately. - */ -static void ext4_free_branches(handle_t *handle, struct inode *inode, - struct buffer_head *parent_bh, - __le32 *first, __le32 *last, int depth) -{ - ext4_fsblk_t nr; - __le32 *p; - - if (ext4_handle_is_aborted(handle)) - return; - - if (depth--) { - struct buffer_head *bh; - int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); - p = last; - while (--p >= first) { - nr = le32_to_cpu(*p); - if (!nr) - continue; /* A hole */ - - if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), - nr, 1)) { - EXT4_ERROR_INODE(inode, - "invalid indirect mapped " - "block %lu (level %d)", - (unsigned long) nr, depth); - break; - } - - /* Go read the buffer for the next level down */ - bh = sb_bread(inode->i_sb, nr); - - /* - * A read failure? Report error and clear slot - * (should be rare). - */ - if (!bh) { - EXT4_ERROR_INODE_BLOCK(inode, nr, - "Read failure"); - continue; - } - - /* This zaps the entire block. Bottom up. */ - BUFFER_TRACE(bh, "free child branches"); - ext4_free_branches(handle, inode, bh, - (__le32 *) bh->b_data, - (__le32 *) bh->b_data + addr_per_block, - depth); - brelse(bh); - - /* - * Everything below this this pointer has been - * released. Now let this top-of-subtree go. - * - * We want the freeing of this indirect block to be - * atomic in the journal with the updating of the - * bitmap block which owns it. So make some room in - * the journal. - * - * We zero the parent pointer *after* freeing its - * pointee in the bitmaps, so if extend_transaction() - * for some reason fails to put the bitmap changes and - * the release into the same transaction, recovery - * will merely complain about releasing a free block, - * rather than leaking blocks. - */ - if (ext4_handle_is_aborted(handle)) - return; - if (try_to_extend_transaction(handle, inode)) { - ext4_mark_inode_dirty(handle, inode); - ext4_truncate_restart_trans(handle, inode, - blocks_for_truncate(inode)); - } - - /* - * The forget flag here is critical because if - * we are journaling (and not doing data - * journaling), we have to make sure a revoke - * record is written to prevent the journal - * replay from overwriting the (former) - * indirect block if it gets reallocated as a - * data block. This must happen in the same - * transaction where the data blocks are - * actually freed. - */ - ext4_free_blocks(handle, inode, NULL, nr, 1, - EXT4_FREE_BLOCKS_METADATA| - EXT4_FREE_BLOCKS_FORGET); - - if (parent_bh) { - /* - * The block which we have just freed is - * pointed to by an indirect block: journal it - */ - BUFFER_TRACE(parent_bh, "get_write_access"); - if (!ext4_journal_get_write_access(handle, - parent_bh)){ - *p = 0; - BUFFER_TRACE(parent_bh, - "call ext4_handle_dirty_metadata"); - ext4_handle_dirty_metadata(handle, - inode, - parent_bh); - } - } - } - } else { - /* We have reached the bottom of the tree. */ - BUFFER_TRACE(parent_bh, "free data blocks"); - ext4_free_data(handle, inode, parent_bh, first, last); - } -} - int ext4_can_truncate(struct inode *inode) { if (S_ISREG(inode->i_mode)) @@ -4476,19 +3129,6 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length) */ void ext4_truncate(struct inode *inode) { - handle_t *handle; - struct ext4_inode_info *ei = EXT4_I(inode); - __le32 *i_data = ei->i_data; - int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); - struct address_space *mapping = inode->i_mapping; - ext4_lblk_t offsets[4]; - Indirect chain[4]; - Indirect *partial; - __le32 nr = 0; - int n = 0; - ext4_lblk_t last_block, max_block; - unsigned blocksize = inode->i_sb->s_blocksize; - trace_ext4_truncate_enter(inode); if (!ext4_can_truncate(inode)) @@ -4499,149 +3139,11 @@ void ext4_truncate(struct inode *inode) if (inode->i_size == 0 && !test_opt(inode->i_sb, NO_AUTO_DA_ALLOC)) ext4_set_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE); - if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) ext4_ext_truncate(inode); - trace_ext4_truncate_exit(inode); - return; - } + else + ext4_ind_truncate(inode); - handle = start_transaction(inode); - if (IS_ERR(handle)) - return; /* AKPM: return what? */ - - last_block = (inode->i_size + blocksize-1) - >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); - max_block = (EXT4_SB(inode->i_sb)->s_bitmap_maxbytes + blocksize-1) - >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); - - if (inode->i_size & (blocksize - 1)) - if (ext4_block_truncate_page(handle, mapping, inode->i_size)) - goto out_stop; - - if (last_block != max_block) { - n = ext4_block_to_path(inode, last_block, offsets, NULL); - if (n == 0) - goto out_stop; /* error */ - } - - /* - * OK. This truncate is going to happen. We add the inode to the - * orphan list, so that if this truncate spans multiple transactions, - * and we crash, we will resume the truncate when the filesystem - * recovers. It also marks the inode dirty, to catch the new size. - * - * Implication: the file must always be in a sane, consistent - * truncatable state while each transaction commits. - */ - if (ext4_orphan_add(handle, inode)) - goto out_stop; - - /* - * From here we block out all ext4_get_block() callers who want to - * modify the block allocation tree. - */ - down_write(&ei->i_data_sem); - - ext4_discard_preallocations(inode); - - /* - * The orphan list entry will now protect us from any crash which - * occurs before the truncate completes, so it is now safe to propagate - * the new, shorter inode size (held for now in i_size) into the - * on-disk inode. We do this via i_disksize, which is the value which - * ext4 *really* writes onto the disk inode. - */ - ei->i_disksize = inode->i_size; - - if (last_block == max_block) { - /* - * It is unnecessary to free any data blocks if last_block is - * equal to the indirect block limit. - */ - goto out_unlock; - } else if (n == 1) { /* direct blocks */ - ext4_free_data(handle, inode, NULL, i_data+offsets[0], - i_data + EXT4_NDIR_BLOCKS); - goto do_indirects; - } - - partial = ext4_find_shared(inode, n, offsets, chain, &nr); - /* Kill the top of shared branch (not detached) */ - if (nr) { - if (partial == chain) { - /* Shared branch grows from the inode */ - ext4_free_branches(handle, inode, NULL, - &nr, &nr+1, (chain+n-1) - partial); - *partial->p = 0; - /* - * We mark the inode dirty prior to restart, - * and prior to stop. No need for it here. - */ - } else { - /* Shared branch grows from an indirect block */ - BUFFER_TRACE(partial->bh, "get_write_access"); - ext4_free_branches(handle, inode, partial->bh, - partial->p, - partial->p+1, (chain+n-1) - partial); - } - } - /* Clear the ends of indirect blocks on the shared branch */ - while (partial > chain) { - ext4_free_branches(handle, inode, partial->bh, partial->p + 1, - (__le32*)partial->bh->b_data+addr_per_block, - (chain+n-1) - partial); - BUFFER_TRACE(partial->bh, "call brelse"); - brelse(partial->bh); - partial--; - } -do_indirects: - /* Kill the remaining (whole) subtrees */ - switch (offsets[0]) { - default: - nr = i_data[EXT4_IND_BLOCK]; - if (nr) { - ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 1); - i_data[EXT4_IND_BLOCK] = 0; - } - case EXT4_IND_BLOCK: - nr = i_data[EXT4_DIND_BLOCK]; - if (nr) { - ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 2); - i_data[EXT4_DIND_BLOCK] = 0; - } - case EXT4_DIND_BLOCK: - nr = i_data[EXT4_TIND_BLOCK]; - if (nr) { - ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 3); - i_data[EXT4_TIND_BLOCK] = 0; - } - case EXT4_TIND_BLOCK: - ; - } - -out_unlock: - up_write(&ei->i_data_sem); - inode->i_mtime = inode->i_ctime = ext4_current_time(inode); - ext4_mark_inode_dirty(handle, inode); - - /* - * In a multi-transaction truncate, we only make the final transaction - * synchronous - */ - if (IS_SYNC(inode)) - ext4_handle_sync(handle); -out_stop: - /* - * If this was a simple ftruncate(), and the file will remain alive - * then we need to clear up the orphan record which we created above. - * However, if this was a real unlink then we were called by - * ext4_delete_inode(), and we allow that function to clean up the - * orphan info for us. - */ - if (inode->i_nlink) - ext4_orphan_del(handle, inode); - - ext4_journal_stop(handle); trace_ext4_truncate_exit(inode); } @@ -5012,7 +3514,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) (S_ISLNK(inode->i_mode) && !ext4_inode_is_fast_symlink(inode))) { /* Validate block references which are part of inode */ - ret = ext4_check_inode_blockref(inode); + ret = ext4_ind_check_inode(inode); } if (ret) goto bad_inode; @@ -5459,34 +3961,10 @@ int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry, return 0; } -static int ext4_indirect_trans_blocks(struct inode *inode, int nrblocks, - int chunk) -{ - int indirects; - - /* if nrblocks are contiguous */ - if (chunk) { - /* - * With N contiguous data blocks, we need at most - * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) + 1 indirect blocks, - * 2 dindirect blocks, and 1 tindirect block - */ - return DIV_ROUND_UP(nrblocks, - EXT4_ADDR_PER_BLOCK(inode->i_sb)) + 4; - } - /* - * if nrblocks are not contiguous, worse case, each block touch - * a indirect block, and each indirect block touch a double indirect - * block, plus a triple indirect block - */ - indirects = nrblocks * 2 + 1; - return indirects; -} - static int ext4_index_trans_blocks(struct inode *inode, int nrblocks, int chunk) { if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) - return ext4_indirect_trans_blocks(inode, nrblocks, chunk); + return ext4_ind_trans_blocks(inode, nrblocks, chunk); return ext4_ext_index_trans_blocks(inode, nrblocks, chunk); } diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 808c554e773f..f18bfe37aff8 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -202,8 +202,9 @@ setversion_out: struct super_block *sb = inode->i_sb; int err, err2=0; - if (!capable(CAP_SYS_RESOURCE)) - return -EPERM; + err = ext4_resize_begin(sb); + if (err) + return err; if (get_user(n_blocks_count, (__u32 __user *)arg)) return -EFAULT; @@ -221,6 +222,7 @@ setversion_out: if (err == 0) err = err2; mnt_drop_write(filp->f_path.mnt); + ext4_resize_end(sb); return err; } @@ -271,8 +273,9 @@ mext_out: struct super_block *sb = inode->i_sb; int err, err2=0; - if (!capable(CAP_SYS_RESOURCE)) - return -EPERM; + err = ext4_resize_begin(sb); + if (err) + return err; if (copy_from_user(&input, (struct ext4_new_group_input __user *)arg, sizeof(input))) @@ -291,6 +294,7 @@ mext_out: if (err == 0) err = err2; mnt_drop_write(filp->f_path.mnt); + ext4_resize_end(sb); return err; } diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 6ed859d56850..17a5a57c415a 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -75,8 +75,8 @@ * * The inode preallocation space is used looking at the _logical_ start * block. If only the logical file block falls within the range of prealloc - * space we will consume the particular prealloc space. This make sure that - * that the we have contiguous physical blocks representing the file blocks + * space we will consume the particular prealloc space. This makes sure that + * we have contiguous physical blocks representing the file blocks * * The important thing to be noted in case of inode prealloc space is that * we don't modify the values associated to inode prealloc space except @@ -84,7 +84,7 @@ * * If we are not able to find blocks in the inode prealloc space and if we * have the group allocation flag set then we look at the locality group - * prealloc space. These are per CPU prealloc list repreasented as + * prealloc space. These are per CPU prealloc list represented as * * ext4_sb_info.s_locality_groups[smp_processor_id()] * @@ -128,12 +128,13 @@ * we are doing a group prealloc we try to normalize the request to * sbi->s_mb_group_prealloc. Default value of s_mb_group_prealloc is * 512 blocks. This can be tuned via - * /sys/fs/ext4//mb_group_prealloc. The value is represented in * terms of number of blocks. If we have mounted the file system with -O * stripe= option the group prealloc request is normalized to the - * stripe value (sbi->s_stripe) + * the smallest multiple of the stripe value (sbi->s_stripe) which is + * greater than the default mb_group_prealloc. * - * The regular allocator(using the buddy cache) supports few tunables. + * The regular allocator (using the buddy cache) supports a few tunables. * * /sys/fs/ext4//mb_min_to_scan * /sys/fs/ext4//mb_max_to_scan @@ -152,7 +153,7 @@ * best extent in the found extents. Searching for the blocks starts with * the group specified as the goal value in allocation context via * ac_g_ex. Each group is first checked based on the criteria whether it - * can used for allocation. ext4_mb_good_group explains how the groups are + * can be used for allocation. ext4_mb_good_group explains how the groups are * checked. * * Both the prealloc space are getting populated as above. So for the first @@ -492,10 +493,11 @@ static void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap) b2 = (unsigned char *) bitmap; for (i = 0; i < e4b->bd_sb->s_blocksize; i++) { if (b1[i] != b2[i]) { - printk(KERN_ERR "corruption in group %u " - "at byte %u(%u): %x in copy != %x " - "on disk/prealloc\n", - e4b->bd_group, i, i * 8, b1[i], b2[i]); + ext4_msg(e4b->bd_sb, KERN_ERR, + "corruption in group %u " + "at byte %u(%u): %x in copy != %x " + "on disk/prealloc", + e4b->bd_group, i, i * 8, b1[i], b2[i]); BUG(); } } @@ -1125,7 +1127,7 @@ ext4_mb_load_buddy(struct super_block *sb, ext4_group_t group, grp = ext4_get_group_info(sb, group); e4b->bd_blkbits = sb->s_blocksize_bits; - e4b->bd_info = ext4_get_group_info(sb, group); + e4b->bd_info = grp; e4b->bd_sb = sb; e4b->bd_group = group; e4b->bd_buddy_page = NULL; @@ -1281,7 +1283,7 @@ static void mb_clear_bits(void *bm, int cur, int len) } } -static void mb_set_bits(void *bm, int cur, int len) +void ext4_set_bits(void *bm, int cur, int len) { __u32 *addr; @@ -1510,7 +1512,7 @@ static int mb_mark_used(struct ext4_buddy *e4b, struct ext4_free_extent *ex) } mb_set_largest_free_order(e4b->bd_sb, e4b->bd_info); - mb_set_bits(EXT4_MB_BITMAP(e4b), ex->fe_start, len0); + ext4_set_bits(EXT4_MB_BITMAP(e4b), ex->fe_start, len0); mb_check_buddy(e4b); return ret; @@ -2223,8 +2225,8 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group, EXT4_DESC_PER_BLOCK_BITS(sb); meta_group_info = kmalloc(metalen, GFP_KERNEL); if (meta_group_info == NULL) { - printk(KERN_ERR "EXT4-fs: can't allocate mem for a " - "buddy group\n"); + ext4_msg(sb, KERN_ERR, "EXT4-fs: can't allocate mem " + "for a buddy group"); goto exit_meta_group_info; } sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)] = @@ -2237,7 +2239,7 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group, meta_group_info[i] = kmem_cache_alloc(cachep, GFP_KERNEL); if (meta_group_info[i] == NULL) { - printk(KERN_ERR "EXT4-fs: can't allocate buddy mem\n"); + ext4_msg(sb, KERN_ERR, "EXT4-fs: can't allocate buddy mem"); goto exit_group_info; } memset(meta_group_info[i], 0, kmem_cache_size(cachep)); @@ -2279,8 +2281,10 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group, exit_group_info: /* If a meta_group_info table has been allocated, release it now */ - if (group % EXT4_DESC_PER_BLOCK(sb) == 0) + if (group % EXT4_DESC_PER_BLOCK(sb) == 0) { kfree(sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)]); + sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)] = NULL; + } exit_meta_group_info: return -ENOMEM; } /* ext4_mb_add_groupinfo */ @@ -2328,23 +2332,26 @@ static int ext4_mb_init_backend(struct super_block *sb) /* An 8TB filesystem with 64-bit pointers requires a 4096 byte * kmalloc. A 128kb malloc should suffice for a 256TB filesystem. * So a two level scheme suffices for now. */ - sbi->s_group_info = kzalloc(array_size, GFP_KERNEL); + sbi->s_group_info = ext4_kvzalloc(array_size, GFP_KERNEL); if (sbi->s_group_info == NULL) { - printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n"); + ext4_msg(sb, KERN_ERR, "can't allocate buddy meta group"); return -ENOMEM; } sbi->s_buddy_cache = new_inode(sb); if (sbi->s_buddy_cache == NULL) { - printk(KERN_ERR "EXT4-fs: can't get new inode\n"); + ext4_msg(sb, KERN_ERR, "can't get new inode"); goto err_freesgi; } - sbi->s_buddy_cache->i_ino = get_next_ino(); + /* To avoid potentially colliding with an valid on-disk inode number, + * use EXT4_BAD_INO for the buddy cache inode number. This inode is + * not in the inode hash, so it should never be found by iget(), but + * this will avoid confusion if it ever shows up during debugging. */ + sbi->s_buddy_cache->i_ino = EXT4_BAD_INO; EXT4_I(sbi->s_buddy_cache)->i_disksize = 0; for (i = 0; i < ngroups; i++) { desc = ext4_get_group_desc(sb, i, NULL); if (desc == NULL) { - printk(KERN_ERR - "EXT4-fs: can't read descriptor %u\n", i); + ext4_msg(sb, KERN_ERR, "can't read descriptor %u", i); goto err_freebuddy; } if (ext4_mb_add_groupinfo(sb, i, desc) != 0) @@ -2362,7 +2369,7 @@ err_freebuddy: kfree(sbi->s_group_info[i]); iput(sbi->s_buddy_cache); err_freesgi: - kfree(sbi->s_group_info); + ext4_kvfree(sbi->s_group_info); return -ENOMEM; } @@ -2404,14 +2411,15 @@ static int ext4_groupinfo_create_slab(size_t size) slab_size, 0, SLAB_RECLAIM_ACCOUNT, NULL); + ext4_groupinfo_caches[cache_index] = cachep; + mutex_unlock(&ext4_grpinfo_slab_create_mutex); if (!cachep) { - printk(KERN_EMERG "EXT4: no memory for groupinfo slab cache\n"); + printk(KERN_EMERG + "EXT4-fs: no memory for groupinfo slab cache\n"); return -ENOMEM; } - ext4_groupinfo_caches[cache_index] = cachep; - return 0; } @@ -2457,12 +2465,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) i++; } while (i <= sb->s_blocksize_bits + 1); - /* init file for buddy data */ - ret = ext4_mb_init_backend(sb); - if (ret != 0) { - goto out; - } - spin_lock_init(&sbi->s_md_lock); spin_lock_init(&sbi->s_bal_lock); @@ -2472,6 +2474,18 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) sbi->s_mb_stream_request = MB_DEFAULT_STREAM_THRESHOLD; sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS; sbi->s_mb_group_prealloc = MB_DEFAULT_GROUP_PREALLOC; + /* + * If there is a s_stripe > 1, then we set the s_mb_group_prealloc + * to the lowest multiple of s_stripe which is bigger than + * the s_mb_group_prealloc as determined above. We want + * the preallocation size to be an exact multiple of the + * RAID stripe size so that preallocations don't fragment + * the stripes. + */ + if (sbi->s_stripe > 1) { + sbi->s_mb_group_prealloc = roundup( + sbi->s_mb_group_prealloc, sbi->s_stripe); + } sbi->s_locality_groups = alloc_percpu(struct ext4_locality_group); if (sbi->s_locality_groups == NULL) { @@ -2487,6 +2501,12 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) spin_lock_init(&lg->lg_prealloc_lock); } + /* init file for buddy data */ + ret = ext4_mb_init_backend(sb); + if (ret != 0) { + goto out; + } + if (sbi->s_proc) proc_create_data("mb_groups", S_IRUGO, sbi->s_proc, &ext4_mb_seq_groups_fops, sb); @@ -2544,32 +2564,32 @@ int ext4_mb_release(struct super_block *sb) EXT4_DESC_PER_BLOCK_BITS(sb); for (i = 0; i < num_meta_group_infos; i++) kfree(sbi->s_group_info[i]); - kfree(sbi->s_group_info); + ext4_kvfree(sbi->s_group_info); } kfree(sbi->s_mb_offsets); kfree(sbi->s_mb_maxs); if (sbi->s_buddy_cache) iput(sbi->s_buddy_cache); if (sbi->s_mb_stats) { - printk(KERN_INFO - "EXT4-fs: mballoc: %u blocks %u reqs (%u success)\n", + ext4_msg(sb, KERN_INFO, + "mballoc: %u blocks %u reqs (%u success)", atomic_read(&sbi->s_bal_allocated), atomic_read(&sbi->s_bal_reqs), atomic_read(&sbi->s_bal_success)); - printk(KERN_INFO - "EXT4-fs: mballoc: %u extents scanned, %u goal hits, " - "%u 2^N hits, %u breaks, %u lost\n", + ext4_msg(sb, KERN_INFO, + "mballoc: %u extents scanned, %u goal hits, " + "%u 2^N hits, %u breaks, %u lost", atomic_read(&sbi->s_bal_ex_scanned), atomic_read(&sbi->s_bal_goals), atomic_read(&sbi->s_bal_2orders), atomic_read(&sbi->s_bal_breaks), atomic_read(&sbi->s_mb_lost_chunks)); - printk(KERN_INFO - "EXT4-fs: mballoc: %lu generated and it took %Lu\n", - sbi->s_mb_buddies_generated++, + ext4_msg(sb, KERN_INFO, + "mballoc: %lu generated and it took %Lu", + sbi->s_mb_buddies_generated, sbi->s_mb_generation_time); - printk(KERN_INFO - "EXT4-fs: mballoc: %u preallocated, %u discarded\n", + ext4_msg(sb, KERN_INFO, + "mballoc: %u preallocated, %u discarded", atomic_read(&sbi->s_mb_preallocated), atomic_read(&sbi->s_mb_discarded)); } @@ -2628,6 +2648,15 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) rb_erase(&entry->node, &(db->bb_free_root)); mb_free_blocks(NULL, &e4b, entry->start_blk, entry->count); + /* + * Clear the trimmed flag for the group so that the next + * ext4_trim_fs can trim it. + * If the volume is mounted with -o discard, online discard + * is supported and the free blocks will be trimmed online. + */ + if (!test_opt(sb, DISCARD)) + EXT4_MB_GRP_CLEAR_TRIMMED(db); + if (!db->bb_free_root.rb_node) { /* No more items in the per group rb tree * balance refcounts from ext4_mb_free_metadata() @@ -2771,8 +2800,8 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, * We leak some of the blocks here. */ ext4_lock_group(sb, ac->ac_b_ex.fe_group); - mb_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start, - ac->ac_b_ex.fe_len); + ext4_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start, + ac->ac_b_ex.fe_len); ext4_unlock_group(sb, ac->ac_b_ex.fe_group); err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); if (!err) @@ -2790,7 +2819,8 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, } } #endif - mb_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start,ac->ac_b_ex.fe_len); + ext4_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start, + ac->ac_b_ex.fe_len); if (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) { gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT); ext4_free_blks_set(sb, gdp, @@ -2830,8 +2860,9 @@ out_err: /* * here we normalize request for locality group - * Group request are normalized to s_strip size if we set the same via mount - * option. If not we set it to s_mb_group_prealloc which can be configured via + * Group request are normalized to s_mb_group_prealloc, which goes to + * s_strip if we set the same via mount option. + * s_mb_group_prealloc can be configured via * /sys/fs/ext4//mb_group_prealloc * * XXX: should we try to preallocate more than the group has now? @@ -2842,10 +2873,7 @@ static void ext4_mb_normalize_group_request(struct ext4_allocation_context *ac) struct ext4_locality_group *lg = ac->ac_lg; BUG_ON(lg == NULL); - if (EXT4_SB(sb)->s_stripe) - ac->ac_g_ex.fe_len = EXT4_SB(sb)->s_stripe; - else - ac->ac_g_ex.fe_len = EXT4_SB(sb)->s_mb_group_prealloc; + ac->ac_g_ex.fe_len = EXT4_SB(sb)->s_mb_group_prealloc; mb_debug(1, "#%u: goal %u blocks for locality group\n", current->pid, ac->ac_g_ex.fe_len); } @@ -3001,9 +3029,10 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac, if (start + size <= ac->ac_o_ex.fe_logical && start > ac->ac_o_ex.fe_logical) { - printk(KERN_ERR "start %lu, size %lu, fe_logical %lu\n", - (unsigned long) start, (unsigned long) size, - (unsigned long) ac->ac_o_ex.fe_logical); + ext4_msg(ac->ac_sb, KERN_ERR, + "start %lu, size %lu, fe_logical %lu", + (unsigned long) start, (unsigned long) size, + (unsigned long) ac->ac_o_ex.fe_logical); } BUG_ON(start + size <= ac->ac_o_ex.fe_logical && start > ac->ac_o_ex.fe_logical); @@ -3262,7 +3291,7 @@ static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, while (n) { entry = rb_entry(n, struct ext4_free_data, node); - mb_set_bits(bitmap, entry->start_blk, entry->count); + ext4_set_bits(bitmap, entry->start_blk, entry->count); n = rb_next(n); } return; @@ -3304,7 +3333,7 @@ void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, if (unlikely(len == 0)) continue; BUG_ON(groupnr != group); - mb_set_bits(bitmap, start, len); + ext4_set_bits(bitmap, start, len); preallocated += len; count++; } @@ -3584,10 +3613,11 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, bit = next + 1; } if (free != pa->pa_free) { - printk(KERN_CRIT "pa %p: logic %lu, phys. %lu, len %lu\n", - pa, (unsigned long) pa->pa_lstart, - (unsigned long) pa->pa_pstart, - (unsigned long) pa->pa_len); + ext4_msg(e4b->bd_sb, KERN_CRIT, + "pa %p: logic %lu, phys. %lu, len %lu", + pa, (unsigned long) pa->pa_lstart, + (unsigned long) pa->pa_pstart, + (unsigned long) pa->pa_len); ext4_grp_locked_error(sb, group, 0, 0, "free %u, pa_free %u", free, pa->pa_free); /* @@ -3775,7 +3805,8 @@ repeat: * use preallocation while we're discarding it */ spin_unlock(&pa->pa_lock); spin_unlock(&ei->i_prealloc_lock); - printk(KERN_ERR "uh-oh! used pa while discarding\n"); + ext4_msg(sb, KERN_ERR, + "uh-oh! used pa while discarding"); WARN_ON(1); schedule_timeout_uninterruptible(HZ); goto repeat; @@ -3852,12 +3883,13 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac) (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED)) return; - printk(KERN_ERR "EXT4-fs: Can't allocate:" - " Allocation context details:\n"); - printk(KERN_ERR "EXT4-fs: status %d flags %d\n", + ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: Can't allocate:" + " Allocation context details:"); + ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: status %d flags %d", ac->ac_status, ac->ac_flags); - printk(KERN_ERR "EXT4-fs: orig %lu/%lu/%lu@%lu, goal %lu/%lu/%lu@%lu, " - "best %lu/%lu/%lu@%lu cr %d\n", + ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: orig %lu/%lu/%lu@%lu, " + "goal %lu/%lu/%lu@%lu, " + "best %lu/%lu/%lu@%lu cr %d", (unsigned long)ac->ac_o_ex.fe_group, (unsigned long)ac->ac_o_ex.fe_start, (unsigned long)ac->ac_o_ex.fe_len, @@ -3871,9 +3903,9 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac) (unsigned long)ac->ac_b_ex.fe_len, (unsigned long)ac->ac_b_ex.fe_logical, (int)ac->ac_criteria); - printk(KERN_ERR "EXT4-fs: %lu scanned, %d found\n", ac->ac_ex_scanned, - ac->ac_found); - printk(KERN_ERR "EXT4-fs: groups: \n"); + ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: %lu scanned, %d found", + ac->ac_ex_scanned, ac->ac_found); + ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: groups: "); ngroups = ext4_get_groups_count(sb); for (i = 0; i < ngroups; i++) { struct ext4_group_info *grp = ext4_get_group_info(sb, i); @@ -4637,7 +4669,7 @@ do_more: } ext4_mark_super_dirty(sb); error_return: - if (freed) + if (freed && !(flags & EXT4_FREE_BLOCKS_NO_QUOT_UPDATE)) dquot_free_block(inode, freed); brelse(bitmap_bh); ext4_std_error(sb, err); @@ -4645,7 +4677,7 @@ error_return: } /** - * ext4_add_groupblocks() -- Add given blocks to an existing group + * ext4_group_add_blocks() -- Add given blocks to an existing group * @handle: handle to this transaction * @sb: super block * @block: start physcial block to add to the block group @@ -4653,7 +4685,7 @@ error_return: * * This marks the blocks as free in the bitmap and buddy. */ -void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, +int ext4_group_add_blocks(handle_t *handle, struct super_block *sb, ext4_fsblk_t block, unsigned long count) { struct buffer_head *bitmap_bh = NULL; @@ -4666,25 +4698,35 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, struct ext4_buddy e4b; int err = 0, ret, blk_free_count; ext4_grpblk_t blocks_freed; - struct ext4_group_info *grp; ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1); + if (count == 0) + return 0; + ext4_get_group_no_and_offset(sb, block, &block_group, &bit); - grp = ext4_get_group_info(sb, block_group); /* * Check to see if we are freeing blocks across a group * boundary. */ - if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) + if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) { + ext4_warning(sb, "too much blocks added to group %u\n", + block_group); + err = -EINVAL; goto error_return; + } bitmap_bh = ext4_read_block_bitmap(sb, block_group); - if (!bitmap_bh) + if (!bitmap_bh) { + err = -EIO; goto error_return; + } + desc = ext4_get_group_desc(sb, block_group, &gd_bh); - if (!desc) + if (!desc) { + err = -EIO; goto error_return; + } if (in_range(ext4_block_bitmap(sb, desc), block, count) || in_range(ext4_inode_bitmap(sb, desc), block, count) || @@ -4694,6 +4736,7 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, ext4_error(sb, "Adding blocks in system zones - " "Block = %llu, count = %lu", block, count); + err = -EINVAL; goto error_return; } @@ -4762,7 +4805,7 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, error_return: brelse(bitmap_bh); ext4_std_error(sb, err); - return; + return err; } /** @@ -4782,6 +4825,8 @@ static void ext4_trim_extent(struct super_block *sb, int start, int count, { struct ext4_free_extent ex; + trace_ext4_trim_extent(sb, group, start, count); + assert_spin_locked(ext4_group_lock_ptr(sb, group)); ex.fe_start = start; @@ -4802,7 +4847,7 @@ static void ext4_trim_extent(struct super_block *sb, int start, int count, /** * ext4_trim_all_free -- function to trim all free space in alloc. group * @sb: super block for file system - * @e4b: ext4 buddy + * @group: group to be trimmed * @start: first group block to examine * @max: last group block to examine * @minblocks: minimum extent block count @@ -4823,10 +4868,12 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, ext4_grpblk_t minblocks) { void *bitmap; - ext4_grpblk_t next, count = 0; + ext4_grpblk_t next, count = 0, free_count = 0; struct ext4_buddy e4b; int ret; + trace_ext4_trim_all_free(sb, group, start, max); + ret = ext4_mb_load_buddy(sb, group, &e4b); if (ret) { ext4_error(sb, "Error in loading buddy " @@ -4836,6 +4883,10 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, bitmap = e4b.bd_bitmap; ext4_lock_group(sb, group); + if (EXT4_MB_GRP_WAS_TRIMMED(e4b.bd_info) && + minblocks >= atomic_read(&EXT4_SB(sb)->s_last_trim_minblks)) + goto out; + start = (e4b.bd_info->bb_first_free > start) ? e4b.bd_info->bb_first_free : start; @@ -4850,6 +4901,7 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, next - start, group, &e4b); count += next - start; } + free_count += next - start; start = next + 1; if (fatal_signal_pending(current)) { @@ -4863,9 +4915,13 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, ext4_lock_group(sb, group); } - if ((e4b.bd_info->bb_free - count) < minblocks) + if ((e4b.bd_info->bb_free - free_count) < minblocks) break; } + + if (!ret) + EXT4_MB_GRP_SET_TRIMMED(e4b.bd_info); +out: ext4_unlock_group(sb, group); ext4_mb_unload_buddy(&e4b); @@ -4904,6 +4960,8 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) if (unlikely(minlen > EXT4_BLOCKS_PER_GROUP(sb))) return -EINVAL; + if (start + len <= first_data_blk) + goto out; if (start < first_data_blk) { len -= first_data_blk - start; start = first_data_blk; @@ -4952,5 +5010,9 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) } range->len = trimmed * sb->s_blocksize; + if (!ret) + atomic_set(&EXT4_SB(sb)->s_last_trim_minblks, minlen); + +out: return ret; } diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index 20b5e7bfebd1..9d4a636b546c 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -187,7 +187,6 @@ struct ext4_allocation_context { __u16 ac_flags; /* allocation hints */ __u8 ac_status; __u8 ac_criteria; - __u8 ac_repeats; __u8 ac_2order; /* if request is to allocate 2^N blocks and * N > 0, the field stores N, otherwise 0 */ __u8 ac_op; /* operation, for history only */ diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 8c9babac43dc..565a154e22d4 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -289,7 +289,7 @@ static struct stats dx_show_leaf(struct dx_hash_info *hinfo, struct ext4_dir_ent while (len--) printk("%c", *name++); ext4fs_dirhash(de->name, de->name_len, &h); printk(":%x.%u ", h.hash, - ((char *) de - base)); + (unsigned) ((char *) de - base)); } space += EXT4_DIR_REC_LEN(de->name_len); names++; @@ -1013,7 +1013,7 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q *err = -ENOENT; errout: - dxtrace(printk(KERN_DEBUG "%s not found\n", name)); + dxtrace(printk(KERN_DEBUG "%s not found\n", d_name->name)); dx_release (frames); return NULL; } @@ -1985,18 +1985,11 @@ int ext4_orphan_add(handle_t *handle, struct inode *inode) if (!list_empty(&EXT4_I(inode)->i_orphan)) goto out_unlock; - /* Orphan handling is only valid for files with data blocks - * being truncated, or files being unlinked. */ - - /* @@@ FIXME: Observation from aviro: - * I think I can trigger J_ASSERT in ext4_orphan_add(). We block - * here (on s_orphan_lock), so race with ext4_link() which might bump - * ->i_nlink. For, say it, character device. Not a regular file, - * not a directory, not a symlink and ->i_nlink > 0. - * - * tytso, 4/25/2009: I'm not sure how that could happen; - * shouldn't the fs core protect us from these sort of - * unlink()/link() races? + /* + * Orphan handling is only valid for files with data blocks + * being truncated, or files being unlinked. Note that we either + * hold i_mutex, or the inode can not be referenced from outside, + * so i_nlink should not be bumped due to race */ J_ASSERT((S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) || inode->i_nlink == 0); diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 7bb8f76d470a..430c401d0895 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -285,11 +285,7 @@ static int io_submit_init(struct ext4_io_submit *io, io_end = ext4_init_io_end(inode, GFP_NOFS); if (!io_end) return -ENOMEM; - do { - bio = bio_alloc(GFP_NOIO, nvecs); - nvecs >>= 1; - } while (bio == NULL); - + bio = bio_alloc(GFP_NOIO, min(nvecs, BIO_MAX_PAGES)); bio->bi_sector = bh->b_blocknr * (bh->b_size >> 9); bio->bi_bdev = bh->b_bdev; bio->bi_private = io->io_end = io_end; diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 80bbc9c60c24..707d3f16f7ce 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -16,6 +16,35 @@ #include "ext4_jbd2.h" +int ext4_resize_begin(struct super_block *sb) +{ + int ret = 0; + + if (!capable(CAP_SYS_RESOURCE)) + return -EPERM; + + /* + * We are not allowed to do online-resizing on a filesystem mounted + * with error, because it can destroy the filesystem easily. + */ + if (EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS) { + ext4_warning(sb, "There are errors in the filesystem, " + "so online resizing is not allowed\n"); + return -EPERM; + } + + if (test_and_set_bit_lock(EXT4_RESIZING, &EXT4_SB(sb)->s_resize_flags)) + ret = -EBUSY; + + return ret; +} + +void ext4_resize_end(struct super_block *sb) +{ + clear_bit_unlock(EXT4_RESIZING, &EXT4_SB(sb)->s_resize_flags); + smp_mb__after_clear_bit(); +} + #define outside(b, first, last) ((b) < (first) || (b) >= (last)) #define inside(b, first, last) ((b) >= (first) && (b) < (last)) @@ -118,10 +147,8 @@ static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, brelse(bh); bh = ERR_PTR(err); } else { - lock_buffer(bh); memset(bh->b_data, 0, sb->s_blocksize); set_buffer_uptodate(bh); - unlock_buffer(bh); } return bh; @@ -132,8 +159,7 @@ static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, * If that fails, restart the transaction & regain write access for the * buffer head which is used for block_bitmap modifications. */ -static int extend_or_restart_transaction(handle_t *handle, int thresh, - struct buffer_head *bh) +static int extend_or_restart_transaction(handle_t *handle, int thresh) { int err; @@ -144,9 +170,8 @@ static int extend_or_restart_transaction(handle_t *handle, int thresh, if (err < 0) return err; if (err) { - if ((err = ext4_journal_restart(handle, EXT4_MAX_TRANS_DATA))) - return err; - if ((err = ext4_journal_get_write_access(handle, bh))) + err = ext4_journal_restart(handle, EXT4_MAX_TRANS_DATA); + if (err) return err; } @@ -181,21 +206,7 @@ static int setup_new_group_blocks(struct super_block *sb, if (IS_ERR(handle)) return PTR_ERR(handle); - mutex_lock(&sbi->s_resize_lock); - if (input->group != sbi->s_groups_count) { - err = -EBUSY; - goto exit_journal; - } - - if (IS_ERR(bh = bclean(handle, sb, input->block_bitmap))) { - err = PTR_ERR(bh); - goto exit_journal; - } - - if (ext4_bg_has_super(sb, input->group)) { - ext4_debug("mark backup superblock %#04llx (+0)\n", start); - ext4_set_bit(0, bh->b_data); - } + BUG_ON(input->group != sbi->s_groups_count); /* Copy all of the GDT blocks into the backup in this group */ for (i = 0, bit = 1, block = start + 1; @@ -203,29 +214,26 @@ static int setup_new_group_blocks(struct super_block *sb, struct buffer_head *gdb; ext4_debug("update backup group %#04llx (+%d)\n", block, bit); - - if ((err = extend_or_restart_transaction(handle, 1, bh))) - goto exit_bh; + err = extend_or_restart_transaction(handle, 1); + if (err) + goto exit_journal; gdb = sb_getblk(sb, block); if (!gdb) { err = -EIO; - goto exit_bh; + goto exit_journal; } if ((err = ext4_journal_get_write_access(handle, gdb))) { brelse(gdb); - goto exit_bh; + goto exit_journal; } - lock_buffer(gdb); memcpy(gdb->b_data, sbi->s_group_desc[i]->b_data, gdb->b_size); set_buffer_uptodate(gdb); - unlock_buffer(gdb); err = ext4_handle_dirty_metadata(handle, NULL, gdb); if (unlikely(err)) { brelse(gdb); - goto exit_bh; + goto exit_journal; } - ext4_set_bit(bit, bh->b_data); brelse(gdb); } @@ -235,9 +243,22 @@ static int setup_new_group_blocks(struct super_block *sb, err = sb_issue_zeroout(sb, gdblocks + start + 1, reserved_gdb, GFP_NOFS); if (err) - goto exit_bh; - for (i = 0, bit = gdblocks + 1; i < reserved_gdb; i++, bit++) - ext4_set_bit(bit, bh->b_data); + goto exit_journal; + + err = extend_or_restart_transaction(handle, 2); + if (err) + goto exit_journal; + + bh = bclean(handle, sb, input->block_bitmap); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + goto exit_journal; + } + + if (ext4_bg_has_super(sb, input->group)) { + ext4_debug("mark backup group tables %#04llx (+0)\n", start); + ext4_set_bits(bh->b_data, 0, gdblocks + reserved_gdb + 1); + } ext4_debug("mark block bitmap %#04llx (+%llu)\n", input->block_bitmap, input->block_bitmap - start); @@ -253,12 +274,9 @@ static int setup_new_group_blocks(struct super_block *sb, err = sb_issue_zeroout(sb, block, sbi->s_itb_per_group, GFP_NOFS); if (err) goto exit_bh; - for (i = 0, bit = input->inode_table - start; - i < sbi->s_itb_per_group; i++, bit++) - ext4_set_bit(bit, bh->b_data); + ext4_set_bits(bh->b_data, input->inode_table - start, + sbi->s_itb_per_group); - if ((err = extend_or_restart_transaction(handle, 2, bh))) - goto exit_bh; ext4_mark_bitmap_end(input->blocks_count, sb->s_blocksize * 8, bh->b_data); @@ -285,7 +303,6 @@ exit_bh: brelse(bh); exit_journal: - mutex_unlock(&sbi->s_resize_lock); if ((err2 = ext4_journal_stop(handle)) && !err) err = err2; @@ -377,15 +394,15 @@ static int verify_reserved_gdb(struct super_block *sb, * fail once we start modifying the data on disk, because JBD has no rollback. */ static int add_new_gdb(handle_t *handle, struct inode *inode, - struct ext4_new_group_data *input, - struct buffer_head **primary) + ext4_group_t group) { struct super_block *sb = inode->i_sb; struct ext4_super_block *es = EXT4_SB(sb)->s_es; - unsigned long gdb_num = input->group / EXT4_DESC_PER_BLOCK(sb); + unsigned long gdb_num = group / EXT4_DESC_PER_BLOCK(sb); ext4_fsblk_t gdblock = EXT4_SB(sb)->s_sbh->b_blocknr + 1 + gdb_num; struct buffer_head **o_group_desc, **n_group_desc; struct buffer_head *dind; + struct buffer_head *gdb_bh; int gdbackups; struct ext4_iloc iloc; __le32 *data; @@ -408,11 +425,12 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, return -EPERM; } - *primary = sb_bread(sb, gdblock); - if (!*primary) + gdb_bh = sb_bread(sb, gdblock); + if (!gdb_bh) return -EIO; - if ((gdbackups = verify_reserved_gdb(sb, *primary)) < 0) { + gdbackups = verify_reserved_gdb(sb, gdb_bh); + if (gdbackups < 0) { err = gdbackups; goto exit_bh; } @@ -427,7 +445,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, data = (__le32 *)dind->b_data; if (le32_to_cpu(data[gdb_num % EXT4_ADDR_PER_BLOCK(sb)]) != gdblock) { ext4_warning(sb, "new group %u GDT block %llu not reserved", - input->group, gdblock); + group, gdblock); err = -EINVAL; goto exit_dind; } @@ -436,7 +454,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, if (unlikely(err)) goto exit_dind; - err = ext4_journal_get_write_access(handle, *primary); + err = ext4_journal_get_write_access(handle, gdb_bh); if (unlikely(err)) goto exit_sbh; @@ -449,12 +467,13 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, if (unlikely(err)) goto exit_dindj; - n_group_desc = kmalloc((gdb_num + 1) * sizeof(struct buffer_head *), - GFP_NOFS); + n_group_desc = ext4_kvmalloc((gdb_num + 1) * + sizeof(struct buffer_head *), + GFP_NOFS); if (!n_group_desc) { err = -ENOMEM; - ext4_warning(sb, - "not enough memory for %lu groups", gdb_num + 1); + ext4_warning(sb, "not enough memory for %lu groups", + gdb_num + 1); goto exit_inode; } @@ -475,8 +494,8 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, } inode->i_blocks -= (gdbackups + 1) * sb->s_blocksize >> 9; ext4_mark_iloc_dirty(handle, inode, &iloc); - memset((*primary)->b_data, 0, sb->s_blocksize); - err = ext4_handle_dirty_metadata(handle, NULL, *primary); + memset(gdb_bh->b_data, 0, sb->s_blocksize); + err = ext4_handle_dirty_metadata(handle, NULL, gdb_bh); if (unlikely(err)) { ext4_std_error(sb, err); goto exit_inode; @@ -486,10 +505,10 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, o_group_desc = EXT4_SB(sb)->s_group_desc; memcpy(n_group_desc, o_group_desc, EXT4_SB(sb)->s_gdb_count * sizeof(struct buffer_head *)); - n_group_desc[gdb_num] = *primary; + n_group_desc[gdb_num] = gdb_bh; EXT4_SB(sb)->s_group_desc = n_group_desc; EXT4_SB(sb)->s_gdb_count++; - kfree(o_group_desc); + ext4_kvfree(o_group_desc); le16_add_cpu(&es->s_reserved_gdt_blocks, -1); err = ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh); @@ -499,6 +518,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, return err; exit_inode: + ext4_kvfree(n_group_desc); /* ext4_handle_release_buffer(handle, iloc.bh); */ brelse(iloc.bh); exit_dindj: @@ -508,7 +528,7 @@ exit_sbh: exit_dind: brelse(dind); exit_bh: - brelse(*primary); + brelse(gdb_bh); ext4_debug("leaving with error %d\n", err); return err; @@ -528,7 +548,7 @@ exit_bh: * backup GDT blocks are stored in their reserved primary GDT block. */ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, - struct ext4_new_group_data *input) + ext4_group_t group) { struct super_block *sb = inode->i_sb; int reserved_gdb =le16_to_cpu(EXT4_SB(sb)->s_es->s_reserved_gdt_blocks); @@ -599,7 +619,7 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, * Finally we can add each of the reserved backup GDT blocks from * the new group to its reserved primary GDT block. */ - blk = input->group * EXT4_BLOCKS_PER_GROUP(sb); + blk = group * EXT4_BLOCKS_PER_GROUP(sb); for (i = 0; i < reserved_gdb; i++) { int err2; data = (__le32 *)primary[i]->b_data; @@ -799,13 +819,6 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) goto exit_put; } - mutex_lock(&sbi->s_resize_lock); - if (input->group != sbi->s_groups_count) { - ext4_warning(sb, "multiple resizers run on filesystem!"); - err = -EBUSY; - goto exit_journal; - } - if ((err = ext4_journal_get_write_access(handle, sbi->s_sbh))) goto exit_journal; @@ -820,16 +833,25 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) if ((err = ext4_journal_get_write_access(handle, primary))) goto exit_journal; - if (reserved_gdb && ext4_bg_num_gdb(sb, input->group) && - (err = reserve_backup_gdb(handle, inode, input))) + if (reserved_gdb && ext4_bg_num_gdb(sb, input->group)) { + err = reserve_backup_gdb(handle, inode, input->group); + if (err) + goto exit_journal; + } + } else { + /* + * Note that we can access new group descriptor block safely + * only if add_new_gdb() succeeds. + */ + err = add_new_gdb(handle, inode, input->group); + if (err) goto exit_journal; - } else if ((err = add_new_gdb(handle, inode, input, &primary))) - goto exit_journal; + primary = sbi->s_group_desc[gdb_num]; + } /* * OK, now we've set up the new group. Time to make it active. * - * We do not lock all allocations via s_resize_lock * so we have to be safe wrt. concurrent accesses the group * data. So we need to be careful to set all of the relevant * group descriptor data etc. *before* we enable the group. @@ -886,13 +908,9 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) * * The precise rules we use are: * - * * Writers of s_groups_count *must* hold s_resize_lock - * AND * * Writers must perform a smp_wmb() after updating all dependent * data and before modifying the groups count * - * * Readers must hold s_resize_lock over the access - * OR * * Readers must perform an smp_rmb() after reading the groups count * and before reading any dependent data. * @@ -937,10 +955,9 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) ext4_handle_dirty_super(handle, sb); exit_journal: - mutex_unlock(&sbi->s_resize_lock); if ((err2 = ext4_journal_stop(handle)) && !err) err = err2; - if (!err) { + if (!err && primary) { update_backups(sb, sbi->s_sbh->b_blocknr, (char *)es, sizeof(struct ext4_super_block)); update_backups(sb, primary->b_blocknr, primary->b_data, @@ -969,16 +986,13 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, ext4_grpblk_t add; struct buffer_head *bh; handle_t *handle; - int err; + int err, err2; ext4_group_t group; - /* We don't need to worry about locking wrt other resizers just - * yet: we're going to revalidate es->s_blocks_count after - * taking the s_resize_lock below. */ o_blocks_count = ext4_blocks_count(es); if (test_opt(sb, DEBUG)) - printk(KERN_DEBUG "EXT4-fs: extending last group from %llu uto %llu blocks\n", + printk(KERN_DEBUG "EXT4-fs: extending last group from %llu to %llu blocks\n", o_blocks_count, n_blocks_count); if (n_blocks_count == 0 || n_blocks_count == o_blocks_count) @@ -995,7 +1009,7 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, if (n_blocks_count < o_blocks_count) { ext4_warning(sb, "can't shrink FS - resize aborted"); - return -EBUSY; + return -EINVAL; } /* Handle the remaining blocks in the last group only. */ @@ -1038,32 +1052,25 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, goto exit_put; } - mutex_lock(&EXT4_SB(sb)->s_resize_lock); - if (o_blocks_count != ext4_blocks_count(es)) { - ext4_warning(sb, "multiple resizers run on filesystem!"); - mutex_unlock(&EXT4_SB(sb)->s_resize_lock); - ext4_journal_stop(handle); - err = -EBUSY; - goto exit_put; - } - if ((err = ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh))) { ext4_warning(sb, "error %d on journal write access", err); - mutex_unlock(&EXT4_SB(sb)->s_resize_lock); ext4_journal_stop(handle); goto exit_put; } ext4_blocks_count_set(es, o_blocks_count + add); - mutex_unlock(&EXT4_SB(sb)->s_resize_lock); ext4_debug("freeing blocks %llu through %llu\n", o_blocks_count, o_blocks_count + add); /* We add the blocks to the bitmap and set the group need init bit */ - ext4_add_groupblocks(handle, sb, o_blocks_count, add); + err = ext4_group_add_blocks(handle, sb, o_blocks_count, add); ext4_handle_dirty_super(handle, sb); ext4_debug("freed blocks %llu through %llu\n", o_blocks_count, o_blocks_count + add); - if ((err = ext4_journal_stop(handle))) + err2 = ext4_journal_stop(handle); + if (!err && err2) + err = err2; + + if (err) goto exit_put; if (test_opt(sb, DEBUG)) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 9ea71aa864b3..e2d88baf91d3 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -110,6 +110,35 @@ static struct file_system_type ext3_fs_type = { #define IS_EXT3_SB(sb) (0) #endif +void *ext4_kvmalloc(size_t size, gfp_t flags) +{ + void *ret; + + ret = kmalloc(size, flags); + if (!ret) + ret = __vmalloc(size, flags, PAGE_KERNEL); + return ret; +} + +void *ext4_kvzalloc(size_t size, gfp_t flags) +{ + void *ret; + + ret = kmalloc(size, flags); + if (!ret) + ret = __vmalloc(size, flags | __GFP_ZERO, PAGE_KERNEL); + return ret; +} + +void ext4_kvfree(void *ptr) +{ + if (is_vmalloc_addr(ptr)) + vfree(ptr); + else + kfree(ptr); + +} + ext4_fsblk_t ext4_block_bitmap(struct super_block *sb, struct ext4_group_desc *bg) { @@ -269,6 +298,7 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) journal_t *journal; handle_t *handle; + trace_ext4_journal_start(sb, nblocks, _RET_IP_); if (sb->s_flags & MS_RDONLY) return ERR_PTR(-EROFS); @@ -789,11 +819,8 @@ static void ext4_put_super(struct super_block *sb) for (i = 0; i < sbi->s_gdb_count; i++) brelse(sbi->s_group_desc[i]); - kfree(sbi->s_group_desc); - if (is_vmalloc_addr(sbi->s_flex_groups)) - vfree(sbi->s_flex_groups); - else - kfree(sbi->s_flex_groups); + ext4_kvfree(sbi->s_group_desc); + ext4_kvfree(sbi->s_flex_groups); percpu_counter_destroy(&sbi->s_freeblocks_counter); percpu_counter_destroy(&sbi->s_freeinodes_counter); percpu_counter_destroy(&sbi->s_dirs_counter); @@ -1976,15 +2003,11 @@ static int ext4_fill_flex_info(struct super_block *sb) ((le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) + 1) << EXT4_DESC_PER_BLOCK_BITS(sb))) / groups_per_flex; size = flex_group_count * sizeof(struct flex_groups); - sbi->s_flex_groups = kzalloc(size, GFP_KERNEL); + sbi->s_flex_groups = ext4_kvzalloc(size, GFP_KERNEL); if (sbi->s_flex_groups == NULL) { - sbi->s_flex_groups = vzalloc(size); - if (sbi->s_flex_groups == NULL) { - ext4_msg(sb, KERN_ERR, - "not enough memory for %u flex groups", - flex_group_count); - goto failed; - } + ext4_msg(sb, KERN_ERR, "not enough memory for %u flex groups", + flex_group_count); + goto failed; } for (i = 0; i < sbi->s_groups_count; i++) { @@ -2383,17 +2406,25 @@ static unsigned long ext4_get_stripe_size(struct ext4_sb_info *sbi) unsigned long stride = le16_to_cpu(sbi->s_es->s_raid_stride); unsigned long stripe_width = le32_to_cpu(sbi->s_es->s_raid_stripe_width); + int ret; if (sbi->s_stripe && sbi->s_stripe <= sbi->s_blocks_per_group) - return sbi->s_stripe; + ret = sbi->s_stripe; + else if (stripe_width <= sbi->s_blocks_per_group) + ret = stripe_width; + else if (stride <= sbi->s_blocks_per_group) + ret = stride; + else + ret = 0; - if (stripe_width <= sbi->s_blocks_per_group) - return stripe_width; + /* + * If the stripe width is 1, this makes no sense and + * we set it to 0 to turn off stripe handling code. + */ + if (ret <= 1) + ret = 0; - if (stride <= sbi->s_blocks_per_group) - return stride; - - return 0; + return ret; } /* sysfs supprt */ @@ -3408,8 +3439,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) (EXT4_MAX_BLOCK_FILE_PHYS / EXT4_BLOCKS_PER_GROUP(sb))); db_count = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) / EXT4_DESC_PER_BLOCK(sb); - sbi->s_group_desc = kmalloc(db_count * sizeof(struct buffer_head *), - GFP_KERNEL); + sbi->s_group_desc = ext4_kvmalloc(db_count * + sizeof(struct buffer_head *), + GFP_KERNEL); if (sbi->s_group_desc == NULL) { ext4_msg(sb, KERN_ERR, "not enough memory"); goto failed_mount; @@ -3491,7 +3523,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */ mutex_init(&sbi->s_orphan_lock); - mutex_init(&sbi->s_resize_lock); + sbi->s_resize_flags = 0; sb->s_root = NULL; @@ -3741,12 +3773,8 @@ failed_mount_wq: } failed_mount3: del_timer(&sbi->s_err_report); - if (sbi->s_flex_groups) { - if (is_vmalloc_addr(sbi->s_flex_groups)) - vfree(sbi->s_flex_groups); - else - kfree(sbi->s_flex_groups); - } + if (sbi->s_flex_groups) + ext4_kvfree(sbi->s_flex_groups); percpu_counter_destroy(&sbi->s_freeblocks_counter); percpu_counter_destroy(&sbi->s_freeinodes_counter); percpu_counter_destroy(&sbi->s_dirs_counter); @@ -3756,7 +3784,7 @@ failed_mount3: failed_mount2: for (i = 0; i < db_count; i++) brelse(sbi->s_group_desc[i]); - kfree(sbi->s_group_desc); + ext4_kvfree(sbi->s_group_desc); failed_mount: if (sbi->s_proc) { remove_proc_entry(sb->s_id, ext4_proc_root); diff --git a/fs/ext4/truncate.h b/fs/ext4/truncate.h new file mode 100644 index 000000000000..011ba6670d99 --- /dev/null +++ b/fs/ext4/truncate.h @@ -0,0 +1,43 @@ +/* + * linux/fs/ext4/truncate.h + * + * Common inline functions needed for truncate support + */ + +/* + * Truncate blocks that were not used by write. We have to truncate the + * pagecache as well so that corresponding buffers get properly unmapped. + */ +static inline void ext4_truncate_failed_write(struct inode *inode) +{ + truncate_inode_pages(inode->i_mapping, inode->i_size); + ext4_truncate(inode); +} + +/* + * Work out how many blocks we need to proceed with the next chunk of a + * truncate transaction. + */ +static inline unsigned long ext4_blocks_for_truncate(struct inode *inode) +{ + ext4_lblk_t needed; + + needed = inode->i_blocks >> (inode->i_sb->s_blocksize_bits - 9); + + /* Give ourselves just enough room to cope with inodes in which + * i_blocks is corrupt: we've seen disk corruptions in the past + * which resulted in random data in an inode which looked enough + * like a regular file for ext4 to try to delete it. Things + * will go a bit crazy if that happens, but at least we should + * try not to panic the whole kernel. */ + if (needed < 2) + needed = 2; + + /* But we need to bound the transaction so we don't overflow the + * journal. */ + if (needed > EXT4_MAX_TRANS_DATA) + needed = EXT4_MAX_TRANS_DATA; + + return EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + needed; +} + diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 2c62c5aae82f..16a698bd906d 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -257,9 +257,12 @@ static void __flush_batch(journal_t *journal, int *batch_count) { int i; + struct blk_plug plug; + blk_start_plug(&plug); for (i = 0; i < *batch_count; i++) - write_dirty_buffer(journal->j_chkpt_bhs[i], WRITE); + write_dirty_buffer(journal->j_chkpt_bhs[i], WRITE_SYNC); + blk_finish_plug(&plug); for (i = 0; i < *batch_count; i++) { struct buffer_head *bh = journal->j_chkpt_bhs[i]; diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 0dfa5b598e68..f24df13adc4e 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -2390,73 +2390,6 @@ static void __exit journal_exit(void) jbd2_journal_destroy_caches(); } -/* - * jbd2_dev_to_name is a utility function used by the jbd2 and ext4 - * tracing infrastructure to map a dev_t to a device name. - * - * The caller should use rcu_read_lock() in order to make sure the - * device name stays valid until its done with it. We use - * rcu_read_lock() as well to make sure we're safe in case the caller - * gets sloppy, and because rcu_read_lock() is cheap and can be safely - * nested. - */ -struct devname_cache { - struct rcu_head rcu; - dev_t device; - char devname[BDEVNAME_SIZE]; -}; -#define CACHE_SIZE_BITS 6 -static struct devname_cache *devcache[1 << CACHE_SIZE_BITS]; -static DEFINE_SPINLOCK(devname_cache_lock); - -static void free_devcache(struct rcu_head *rcu) -{ - kfree(rcu); -} - -const char *jbd2_dev_to_name(dev_t device) -{ - int i = hash_32(device, CACHE_SIZE_BITS); - char *ret; - struct block_device *bd; - static struct devname_cache *new_dev; - - rcu_read_lock(); - if (devcache[i] && devcache[i]->device == device) { - ret = devcache[i]->devname; - rcu_read_unlock(); - return ret; - } - rcu_read_unlock(); - - new_dev = kmalloc(sizeof(struct devname_cache), GFP_KERNEL); - if (!new_dev) - return "NODEV-ALLOCFAILURE"; /* Something non-NULL */ - bd = bdget(device); - spin_lock(&devname_cache_lock); - if (devcache[i]) { - if (devcache[i]->device == device) { - kfree(new_dev); - bdput(bd); - ret = devcache[i]->devname; - spin_unlock(&devname_cache_lock); - return ret; - } - call_rcu(&devcache[i]->rcu, free_devcache); - } - devcache[i] = new_dev; - devcache[i]->device = device; - if (bd) { - bdevname(bd, devcache[i]->devname); - bdput(bd); - } else - __bdevname(device, devcache[i]->devname); - ret = devcache[i]->devname; - spin_unlock(&devname_cache_lock); - return ret; -} -EXPORT_SYMBOL(jbd2_dev_to_name); - MODULE_LICENSE("GPL"); module_init(journal_init); module_exit(journal_exit); diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index d087c2e7b2aa..38f307b8c334 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -1329,12 +1329,6 @@ extern int jbd_blocks_per_page(struct inode *inode); #define BUFFER_TRACE2(bh, bh2, info) do {} while (0) #define JBUFFER_TRACE(jh, info) do {} while (0) -/* - * jbd2_dev_to_name is a utility function used by the jbd2 and ext4 - * tracing infrastructure to map a dev_t to a device name. - */ -extern const char *jbd2_dev_to_name(dev_t device); - #endif /* __KERNEL__ */ #endif /* _LINUX_JBD2_H */ diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 6363193a3418..b50a54736242 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -23,7 +23,7 @@ TRACE_EVENT(ext4_free_inode, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( umode_t, mode ) + __field( __u16, mode ) __field( uid_t, uid ) __field( gid_t, gid ) __field( __u64, blocks ) @@ -52,7 +52,7 @@ TRACE_EVENT(ext4_request_inode, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, dir ) - __field( umode_t, mode ) + __field( __u16, mode ) ), TP_fast_assign( @@ -75,7 +75,7 @@ TRACE_EVENT(ext4_allocate_inode, __field( dev_t, dev ) __field( ino_t, ino ) __field( ino_t, dir ) - __field( umode_t, mode ) + __field( __u16, mode ) ), TP_fast_assign( @@ -725,7 +725,7 @@ TRACE_EVENT(ext4_free_blocks, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( umode_t, mode ) + __field( __u16, mode ) __field( __u64, block ) __field( unsigned long, count ) __field( int, flags ) @@ -1012,7 +1012,7 @@ TRACE_EVENT(ext4_forget, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( umode_t, mode ) + __field( __u16, mode ) __field( int, is_metadata ) __field( __u64, block ) ), @@ -1039,7 +1039,7 @@ TRACE_EVENT(ext4_da_update_reserve_space, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( umode_t, mode ) + __field( __u16, mode ) __field( __u64, i_blocks ) __field( int, used_blocks ) __field( int, reserved_data_blocks ) @@ -1076,7 +1076,7 @@ TRACE_EVENT(ext4_da_reserve_space, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( umode_t, mode ) + __field( __u16, mode ) __field( __u64, i_blocks ) __field( int, md_needed ) __field( int, reserved_data_blocks ) @@ -1110,7 +1110,7 @@ TRACE_EVENT(ext4_da_release_space, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( umode_t, mode ) + __field( __u16, mode ) __field( __u64, i_blocks ) __field( int, freed_blocks ) __field( int, reserved_data_blocks ) @@ -1518,6 +1518,77 @@ TRACE_EVENT(ext4_load_inode, (unsigned long) __entry->ino) ); +TRACE_EVENT(ext4_journal_start, + TP_PROTO(struct super_block *sb, int nblocks, unsigned long IP), + + TP_ARGS(sb, nblocks, IP), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, nblocks ) + __field(unsigned long, ip ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->nblocks = nblocks; + __entry->ip = IP; + ), + + TP_printk("dev %d,%d nblocks %d caller %pF", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->nblocks, (void *)__entry->ip) +); + +DECLARE_EVENT_CLASS(ext4__trim, + TP_PROTO(struct super_block *sb, + ext4_group_t group, + ext4_grpblk_t start, + ext4_grpblk_t len), + + TP_ARGS(sb, group, start, len), + + TP_STRUCT__entry( + __field( int, dev_major ) + __field( int, dev_minor ) + __field( __u32, group ) + __field( int, start ) + __field( int, len ) + ), + + TP_fast_assign( + __entry->dev_major = MAJOR(sb->s_dev); + __entry->dev_minor = MINOR(sb->s_dev); + __entry->group = group; + __entry->start = start; + __entry->len = len; + ), + + TP_printk("dev %d,%d group %u, start %d, len %d", + __entry->dev_major, __entry->dev_minor, + __entry->group, __entry->start, __entry->len) +); + +DEFINE_EVENT(ext4__trim, ext4_trim_extent, + + TP_PROTO(struct super_block *sb, + ext4_group_t group, + ext4_grpblk_t start, + ext4_grpblk_t len), + + TP_ARGS(sb, group, start, len) +); + +DEFINE_EVENT(ext4__trim, ext4_trim_all_free, + + TP_PROTO(struct super_block *sb, + ext4_group_t group, + ext4_grpblk_t start, + ext4_grpblk_t len), + + TP_ARGS(sb, group, start, len) +); + #endif /* _TRACE_EXT4_H */ /* This part must be outside protection */ diff --git a/include/trace/events/jbd2.h b/include/trace/events/jbd2.h index bf16545cc977..75964412ddbb 100644 --- a/include/trace/events/jbd2.h +++ b/include/trace/events/jbd2.h @@ -26,8 +26,8 @@ TRACE_EVENT(jbd2_checkpoint, __entry->result = result; ), - TP_printk("dev %s result %d", - jbd2_dev_to_name(__entry->dev), __entry->result) + TP_printk("dev %d,%d result %d", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->result) ); DECLARE_EVENT_CLASS(jbd2_commit, @@ -48,9 +48,9 @@ DECLARE_EVENT_CLASS(jbd2_commit, __entry->transaction = commit_transaction->t_tid; ), - TP_printk("dev %s transaction %d sync %d", - jbd2_dev_to_name(__entry->dev), __entry->transaction, - __entry->sync_commit) + TP_printk("dev %d,%d transaction %d sync %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->transaction, __entry->sync_commit) ); DEFINE_EVENT(jbd2_commit, jbd2_start_commit, @@ -100,9 +100,9 @@ TRACE_EVENT(jbd2_end_commit, __entry->head = journal->j_tail_sequence; ), - TP_printk("dev %s transaction %d sync %d head %d", - jbd2_dev_to_name(__entry->dev), __entry->transaction, - __entry->sync_commit, __entry->head) + TP_printk("dev %d,%d transaction %d sync %d head %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->transaction, __entry->sync_commit, __entry->head) ); TRACE_EVENT(jbd2_submit_inode_data, @@ -120,8 +120,9 @@ TRACE_EVENT(jbd2_submit_inode_data, __entry->ino = inode->i_ino; ), - TP_printk("dev %s ino %lu", - jbd2_dev_to_name(__entry->dev), (unsigned long) __entry->ino) + TP_printk("dev %d,%d ino %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino) ); TRACE_EVENT(jbd2_run_stats, @@ -156,9 +157,9 @@ TRACE_EVENT(jbd2_run_stats, __entry->blocks_logged = stats->rs_blocks_logged; ), - TP_printk("dev %s tid %lu wait %u running %u locked %u flushing %u " + TP_printk("dev %d,%d tid %lu wait %u running %u locked %u flushing %u " "logging %u handle_count %u blocks %u blocks_logged %u", - jbd2_dev_to_name(__entry->dev), __entry->tid, + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->tid, jiffies_to_msecs(__entry->wait), jiffies_to_msecs(__entry->running), jiffies_to_msecs(__entry->locked), @@ -192,9 +193,9 @@ TRACE_EVENT(jbd2_checkpoint_stats, __entry->dropped = stats->cs_dropped; ), - TP_printk("dev %s tid %lu chp_time %u forced_to_close %u " + TP_printk("dev %d,%d tid %lu chp_time %u forced_to_close %u " "written %u dropped %u", - jbd2_dev_to_name(__entry->dev), __entry->tid, + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->tid, jiffies_to_msecs(__entry->chp_time), __entry->forced_to_close, __entry->written, __entry->dropped) ); @@ -222,9 +223,10 @@ TRACE_EVENT(jbd2_cleanup_journal_tail, __entry->freed = freed; ), - TP_printk("dev %s from %u to %u offset %lu freed %lu", - jbd2_dev_to_name(__entry->dev), __entry->tail_sequence, - __entry->first_tid, __entry->block_nr, __entry->freed) + TP_printk("dev %d,%d from %u to %u offset %lu freed %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->tail_sequence, __entry->first_tid, + __entry->block_nr, __entry->freed) ); #endif /* _TRACE_JBD2_H */