diff options
Diffstat (limited to 'fs/ext4/inode.c')
| -rw-r--r-- | fs/ext4/inode.c | 5935 |
1 files changed, 3815 insertions, 2120 deletions
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 945cbf6cb1f..8a064734e6e 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -12,20 +12,14 @@ * * 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) * * Assorted race fixes, rewrite of ext4_get_block() by Al Viro, 2000 */ -#include <linux/module.h> #include <linux/fs.h> #include <linux/time.h> -#include <linux/ext4_jbd2.h> #include <linux/jbd2.h> #include <linux/highuid.h> #include <linux/pagemap.h> @@ -33,134 +27,131 @@ #include <linux/string.h> #include <linux/buffer_head.h> #include <linux/writeback.h> +#include <linux/pagevec.h> #include <linux/mpage.h> +#include <linux/namei.h> #include <linux/uio.h> #include <linux/bio.h> +#include <linux/workqueue.h> +#include <linux/kernel.h> +#include <linux/printk.h> +#include <linux/slab.h> +#include <linux/ratelimit.h> +#include <linux/aio.h> +#include <linux/bitops.h> + +#include "ext4_jbd2.h" #include "xattr.h" #include "acl.h" +#include "truncate.h" -/* - * Test whether an inode is a fast symlink. - */ -static int ext4_inode_is_fast_symlink(struct inode *inode) -{ - int ea_blocks = EXT4_I(inode)->i_file_acl ? - (inode->i_sb->s_blocksize >> 9) : 0; +#include <trace/events/ext4.h> - return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0); -} +#define MPAGE_DA_EXTENT_TAIL 0x01 -/* - * The ext4 forget function must perform a revoke if we are freeing data - * which has been journaled. Metadata (eg. indirect blocks) must be - * revoked in all cases. - * - * "bh" may be NULL: a metadata block may have been freed from memory - * but there may still be a record of it in the journal, and that record - * still needs to be revoked. - */ -int ext4_forget(handle_t *handle, int is_metadata, struct inode *inode, - struct buffer_head *bh, ext4_fsblk_t blocknr) +static __u32 ext4_inode_csum(struct inode *inode, struct ext4_inode *raw, + struct ext4_inode_info *ei) { - int err; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + __u16 csum_lo; + __u16 csum_hi = 0; + __u32 csum; + + csum_lo = le16_to_cpu(raw->i_checksum_lo); + raw->i_checksum_lo = 0; + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE && + EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi)) { + csum_hi = le16_to_cpu(raw->i_checksum_hi); + raw->i_checksum_hi = 0; + } - might_sleep(); + csum = ext4_chksum(sbi, ei->i_csum_seed, (__u8 *)raw, + EXT4_INODE_SIZE(inode->i_sb)); - BUFFER_TRACE(bh, "enter"); + raw->i_checksum_lo = cpu_to_le16(csum_lo); + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE && + EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi)) + raw->i_checksum_hi = cpu_to_le16(csum_hi); - jbd_debug(4, "forgetting bh %p: is_metadata = %d, mode %o, " - "data mode %lx\n", - bh, is_metadata, inode->i_mode, - test_opt(inode->i_sb, DATA_FLAGS)); + return csum; +} - /* Never use the revoke function if we are doing full data - * journaling: there is no need to, and a V1 superblock won't - * support it. Otherwise, only skip the revoke on un-journaled - * data blocks. */ +static int ext4_inode_csum_verify(struct inode *inode, struct ext4_inode *raw, + struct ext4_inode_info *ei) +{ + __u32 provided, calculated; - if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA || - (!is_metadata && !ext4_should_journal_data(inode))) { - if (bh) { - BUFFER_TRACE(bh, "call jbd2_journal_forget"); - return ext4_journal_forget(handle, bh); - } - return 0; - } + if (EXT4_SB(inode->i_sb)->s_es->s_creator_os != + cpu_to_le32(EXT4_OS_LINUX) || + !EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) + return 1; + + provided = le16_to_cpu(raw->i_checksum_lo); + calculated = ext4_inode_csum(inode, raw, ei); + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE && + EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi)) + provided |= ((__u32)le16_to_cpu(raw->i_checksum_hi)) << 16; + else + calculated &= 0xFFFF; - /* - * data!=journal && (is_metadata || should_journal_data(inode)) - */ - BUFFER_TRACE(bh, "call ext4_journal_revoke"); - err = ext4_journal_revoke(handle, blocknr, bh); - if (err) - ext4_abort(inode->i_sb, __FUNCTION__, - "error %d when attempting revoke", err); - BUFFER_TRACE(bh, "exit"); - return err; + return provided == calculated; } -/* - * 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) +static void ext4_inode_csum_set(struct inode *inode, struct ext4_inode *raw, + struct ext4_inode_info *ei) { - ext4_lblk_t needed; - - needed = inode->i_blocks >> (inode->i_sb->s_blocksize_bits - 9); + __u32 csum; - /* 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; + if (EXT4_SB(inode->i_sb)->s_es->s_creator_os != + cpu_to_le32(EXT4_OS_LINUX) || + !EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) + return; - return EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + needed; + csum = ext4_inode_csum(inode, raw, ei); + raw->i_checksum_lo = cpu_to_le16(csum & 0xFFFF); + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE && + EXT4_FITS_IN_INODE(raw, ei, i_checksum_hi)) + raw->i_checksum_hi = cpu_to_le16(csum >> 16); } -/* - * 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) +static inline int ext4_begin_ordered_truncate(struct inode *inode, + loff_t new_size) { - 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; + trace_ext4_begin_ordered_truncate(inode, new_size); + /* + * If jinode is zero, then we never opened the file for + * writing, so there's no need to call + * jbd2_journal_begin_ordered_truncate() since there's no + * outstanding writes we need to flush. + */ + if (!EXT4_I(inode)->jinode) + return 0; + return jbd2_journal_begin_ordered_truncate(EXT4_JOURNAL(inode), + EXT4_I(inode)->jinode, + new_size); } +static void ext4_invalidatepage(struct page *page, unsigned int offset, + unsigned int length); +static int __ext4_journalled_writepage(struct page *page, unsigned int len); +static int ext4_bh_delay_or_unwritten(handle_t *handle, struct buffer_head *bh); +static int ext4_meta_trans_blocks(struct inode *inode, int lblocks, + int pextents); + /* - * 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. + * Test whether an inode is a fast symlink. */ -static int try_to_extend_transaction(handle_t *handle, struct inode *inode) +static int ext4_inode_is_fast_symlink(struct inode *inode) { - if (handle->h_buffer_credits > EXT4_RESERVE_TRANS_BLOCKS) - return 0; - if (!ext4_journal_extend(handle, blocks_for_truncate(inode))) + int ea_blocks = EXT4_I(inode)->i_file_acl ? + EXT4_CLUSTER_SIZE(inode->i_sb) >> 9 : 0; + + if (ext4_has_inline_data(inode)) return 0; - return 1; + + return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0); } /* @@ -168,40 +159,134 @@ static int try_to_extend_transaction(handle_t *handle, struct inode *inode) * so before we call here everything must be consistently dirtied against * this transaction. */ -static int ext4_journal_test_restart(handle_t *handle, struct inode *inode) +int ext4_truncate_restart_trans(handle_t *handle, struct inode *inode, + int nblocks) { + int ret; + + /* + * Drop i_data_sem to avoid deadlock with ext4_map_blocks. At this + * moment, get_block can be called only for blocks inside i_size since + * page cache has been already dropped and writes are blocked by + * i_mutex. So we can safely drop the i_data_sem here. + */ + BUG_ON(EXT4_JOURNAL(inode) == NULL); jbd_debug(2, "restarting handle %p\n", handle); - return ext4_journal_restart(handle, blocks_for_truncate(inode)); + up_write(&EXT4_I(inode)->i_data_sem); + ret = ext4_journal_restart(handle, nblocks); + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); + + return ret; } /* * Called at the last iput() if i_nlink is zero. */ -void ext4_delete_inode (struct inode * inode) +void ext4_evict_inode(struct inode *inode) { handle_t *handle; + int err; - truncate_inode_pages(&inode->i_data, 0); + 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)) && + inode->i_ino != EXT4_JOURNAL_INO) { + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; + tid_t commit_tid = EXT4_I(inode)->i_datasync_tid; + + jbd2_complete_transaction(journal, commit_tid); + filemap_write_and_wait(&inode->i_data); + } + truncate_inode_pages_final(&inode->i_data); + + WARN_ON(atomic_read(&EXT4_I(inode)->i_ioend_count)); + goto no_delete; + } + + if (!is_bad_inode(inode)) + dquot_initialize(inode); + + if (ext4_should_order_data(inode)) + ext4_begin_ordered_truncate(inode, 0); + truncate_inode_pages_final(&inode->i_data); + WARN_ON(atomic_read(&EXT4_I(inode)->i_ioend_count)); if (is_bad_inode(inode)) goto no_delete; - handle = start_transaction(inode); + /* + * Protect us against freezing - iput() caller didn't have to have any + * protection against it + */ + sb_start_intwrite(inode->i_sb); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, + ext4_blocks_for_truncate(inode)+3); if (IS_ERR(handle)) { + ext4_std_error(inode->i_sb, PTR_ERR(handle)); /* * If we're going to skip the normal cleanup, we still need to * make sure that the in-core orphan linked list is properly * cleaned up. */ ext4_orphan_del(NULL, inode); + sb_end_intwrite(inode->i_sb); goto no_delete; } if (IS_SYNC(inode)) - handle->h_sync = 1; + ext4_handle_sync(handle); inode->i_size = 0; + err = ext4_mark_inode_dirty(handle, inode); + if (err) { + ext4_warning(inode->i_sb, + "couldn't mark inode dirty (err %d)", err); + goto stop_handle; + } if (inode->i_blocks) ext4_truncate(inode); + + /* + * ext4_ext_truncate() doesn't reserve any slop when it + * restarts journal transactions; therefore there may not be + * enough credits left in the handle to remove the inode from + * the orphan list and set the dtime field. + */ + if (!ext4_handle_has_enough_credits(handle, 3)) { + err = ext4_journal_extend(handle, 3); + if (err > 0) + err = ext4_journal_restart(handle, 3); + if (err != 0) { + ext4_warning(inode->i_sb, + "couldn't extend journal (err %d)", err); + stop_handle: + ext4_journal_stop(handle); + ext4_orphan_del(NULL, inode); + sb_end_intwrite(inode->i_sb); + goto no_delete; + } + } + /* * Kill off the orphan record which ext4_truncate created. * AKPM: I think this can be inside the above `if'. @@ -222,887 +307,543 @@ void ext4_delete_inode (struct inode * inode) */ if (ext4_mark_inode_dirty(handle, inode)) /* If that failed, just do the required in-core inode clear. */ - clear_inode(inode); + ext4_clear_inode(inode); else ext4_free_inode(handle, inode); ext4_journal_stop(handle); + sb_end_intwrite(inode->i_sb); return; no_delete: - clear_inode(inode); /* We must guarantee clearing of inode... */ + 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) +#ifdef CONFIG_QUOTA +qsize_t *ext4_get_reserved_space(struct inode *inode) { - p->key = *(p->p = v); - p->bh = bh; + return &EXT4_I(inode)->i_reserved_quota; } - -/** - * 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). - */ +#endif /* - * 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. + * Calculate the number of metadata blocks need to reserve + * to allocate a block located at @lblock */ - -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 < 0) { - ext4_warning (inode->i_sb, "ext4_block_to_path", "block < 0"); - } else 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, "ext4_block_to_path", - "block %lu > max", - i_block + direct_blocks + - indirect_blocks + double_blocks); - } - 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 <key, p, bh> 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) +static int ext4_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock) { - struct super_block *sb = inode->i_sb; - Indirect *p = chain; - struct buffer_head *bh; + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return ext4_ext_calc_metadata_amount(inode, lblock); - *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_bread(sb, le32_to_cpu(p->key)); - if (!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; + return ext4_ind_calc_metadata_amount(inode, lblock); } -/** - * ext4_find_near - find a place for allocation with sufficient locality - * @inode: owner - * @ind: descriptor of indirect block. - * - * This function returns the prefered 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. +/* + * Called with i_data_sem down, which is important since we can call + * ext4_discard_preallocations() from here. */ -static ext4_fsblk_t ext4_find_near(struct inode *inode, Indirect *ind) +void ext4_da_update_reserve_space(struct inode *inode, + int used, int quota_claim) { + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); 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; - /* Try to find previous block */ - for (p = ind->p - 1; p >= start; p--) { - if (*p) - return le32_to_cpu(*p); + spin_lock(&ei->i_block_reservation_lock); + trace_ext4_da_update_reserve_space(inode, used, quota_claim); + if (unlikely(used > ei->i_reserved_data_blocks)) { + ext4_warning(inode->i_sb, "%s: ino %lu, used %d " + "with only %d reserved data blocks", + __func__, inode->i_ino, used, + ei->i_reserved_data_blocks); + WARN_ON(1); + used = ei->i_reserved_data_blocks; } - /* 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. - */ - bg_start = ext4_group_first_block_no(inode->i_sb, ei->i_block_group); - last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1; - - 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 prefered place for allocation. - * @inode: owner - * @block: block we want - * @partial: pointer to the last triple within a chain - * - * Normally this function find the prefered place for block allocation, - * returns it. - */ -static ext4_fsblk_t ext4_find_goal(struct inode *inode, ext4_lblk_t block, - Indirect *partial) -{ - struct ext4_block_alloc_info *block_i; - - block_i = EXT4_I(inode)->i_block_alloc_info; - - /* - * try the heuristic for sequential allocation, - * failing that at least try to get decent locality. - */ - if (block_i && (block == block_i->last_alloc_logical_block + 1) - && (block_i->last_alloc_physical_block != 0)) { - return block_i->last_alloc_physical_block + 1; + if (unlikely(ei->i_allocated_meta_blocks > ei->i_reserved_meta_blocks)) { + ext4_warning(inode->i_sb, "ino %lu, allocated %d " + "with only %d reserved metadata blocks " + "(releasing %d blocks with reserved %d data blocks)", + inode->i_ino, ei->i_allocated_meta_blocks, + ei->i_reserved_meta_blocks, used, + ei->i_reserved_data_blocks); + WARN_ON(1); + ei->i_allocated_meta_blocks = ei->i_reserved_meta_blocks; } - return ext4_find_near(inode, partial); -} - -/** - * 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 long blks, - int blocks_to_boundary) -{ - unsigned long count = 0; + /* Update per-inode reservations */ + ei->i_reserved_data_blocks -= used; + ei->i_reserved_meta_blocks -= ei->i_allocated_meta_blocks; + percpu_counter_sub(&sbi->s_dirtyclusters_counter, + used + ei->i_allocated_meta_blocks); + ei->i_allocated_meta_blocks = 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; + if (ei->i_reserved_data_blocks == 0) { + /* + * We can release all of the reserved metadata blocks + * only when we have written all of the delayed + * allocation blocks. + */ + percpu_counter_sub(&sbi->s_dirtyclusters_counter, + ei->i_reserved_meta_blocks); + ei->i_reserved_meta_blocks = 0; + ei->i_da_metadata_calc_len = 0; } + spin_unlock(&EXT4_I(inode)->i_block_reservation_lock); - count++; - while (count < blks && count <= blocks_to_boundary && - le32_to_cpu(*(branch[0].p + count)) == 0) { - count++; + /* Update quota subsystem for data blocks */ + if (quota_claim) + dquot_claim_block(inode, EXT4_C2B(sbi, used)); + else { + /* + * We did fallocate with an offset that is already delayed + * allocated. So on delayed allocated writeback we should + * not re-claim the quota for fallocated blocks. + */ + dquot_release_reservation_block(inode, EXT4_C2B(sbi, used)); } - return count; -} - -/** - * ext4_alloc_blocks: multiple allocate blocks needed for a branch - * @indirect_blks: the number of blocks need to allocate for indirect - * blocks - * - * @new_blocks: on return it will store the new block numbers for - * the indirect blocks(if needed) and the first direct block, - * @blks: on return it will store the total number of allocated - * direct blocks - */ -static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, - ext4_fsblk_t goal, int indirect_blks, int blks, - ext4_fsblk_t new_blocks[4], int *err) -{ - int target, i; - unsigned long count = 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) + * If we have done all the pending block allocations and if + * there aren't any writers on the inode, we can discard the + * inode's preallocations. */ - target = blks + indirect_blks; - - while (1) { - count = target; - /* allocating blocks for indirect blocks and direct blocks */ - current_block = ext4_new_blocks(handle,inode,goal,&count,err); - if (*err) - goto failed_out; - - target -= count; - /* allocate blocks for indirect blocks */ - while (index < indirect_blks && count) { - new_blocks[index++] = current_block++; - count--; - } - - if (count > 0) - break; - } - - /* save the new block number for the first direct block */ - new_blocks[index] = current_block; - - /* total number of blocks allocated for direct blocks */ - ret = count; - *err = 0; - return ret; -failed_out: - for (i = 0; i <index; i++) - ext4_free_blocks(handle, inode, new_blocks[i], 1, 0); - return ret; + if ((ei->i_reserved_data_blocks == 0) && + (atomic_read(&inode->i_writecount) == 0)) + ext4_discard_preallocations(inode); } -/** - * ext4_alloc_branch - allocate and set up a chain of blocks. - * @inode: owner - * @indirect_blks: number of allocated indirect blocks - * @blks: number of allocated direct blocks - * @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, - int indirect_blks, int *blks, ext4_fsblk_t goal, - ext4_lblk_t *offsets, Indirect *branch) +static int __check_block_validity(struct inode *inode, const char *func, + unsigned int line, + struct ext4_map_blocks *map) { - 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, 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]); - branch[n].bh = bh; - lock_buffer(bh); - BUFFER_TRACE(bh, "call get_create_access"); - err = ext4_journal_get_create_access(handle, bh); - if (err) { - unlock_buffer(bh); - brelse(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_journal_dirty_metadata"); - err = ext4_journal_dirty_metadata(handle, bh); - if (err) - goto failed; - } - *blks = num; - return err; -failed: - /* Allocation failed, free what we already allocated */ - for (i = 1; i <= n ; i++) { - BUFFER_TRACE(branch[i].bh, "call jbd2_journal_forget"); - ext4_journal_forget(handle, branch[i].bh); + if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), map->m_pblk, + map->m_len)) { + ext4_error_inode(inode, func, line, map->m_pblk, + "lblock %lu mapped to illegal pblock " + "(length %d)", (unsigned long) map->m_lblk, + map->m_len); + return -EIO; } - for (i = 0; i <indirect_blks; i++) - ext4_free_blocks(handle, inode, new_blocks[i], 1, 0); - - ext4_free_blocks(handle, inode, new_blocks[i], num, 0); - - return err; + return 0; } -/** - * ext4_splice_branch - splice the allocated branch onto inode. - * @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; - struct ext4_block_alloc_info *block_i; - ext4_fsblk_t current_block; +#define check_block_validity(inode, map) \ + __check_block_validity((inode), __func__, __LINE__, (map)) - block_i = EXT4_I(inode)->i_block_alloc_info; - /* - * 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++); - } +#ifdef ES_AGGRESSIVE_TEST +static void ext4_map_blocks_es_recheck(handle_t *handle, + struct inode *inode, + struct ext4_map_blocks *es_map, + struct ext4_map_blocks *map, + int flags) +{ + int retval; + map->m_flags = 0; /* - * update the most recently allocated logical & physical block - * in i_block_alloc_info, to assist find the proper goal block for next - * allocation + * There is a race window that the result is not the same. + * e.g. xfstests #223 when dioread_nolock enables. The reason + * is that we lookup a block mapping in extent status tree with + * out taking i_data_sem. So at the time the unwritten extent + * could be converted. */ - if (block_i) { - block_i->last_alloc_logical_block = block + blks - 1; - block_i->last_alloc_physical_block = - le32_to_cpu(where[num].key) + blks - 1; - } - - /* We are done with atomic stuff, now do the rest of housekeeping */ - - inode->i_ctime = ext4_current_time(inode); - ext4_mark_inode_dirty(handle, inode); - - /* 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_journal_dirty_metadata"); - err = ext4_journal_dirty_metadata(handle, where->bh); - if (err) - goto err_out; + if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) + down_read(&EXT4_I(inode)->i_data_sem); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + retval = ext4_ext_map_blocks(handle, inode, map, flags & + EXT4_GET_BLOCKS_KEEP_SIZE); } else { - /* - * OK, we spliced it into the inode itself on a direct block. - * Inode was dirtied above. - */ - jbd_debug(5, "splicing direct\n"); + retval = ext4_ind_map_blocks(handle, inode, map, flags & + EXT4_GET_BLOCKS_KEEP_SIZE); } - return err; - -err_out: - for (i = 1; i <= num; i++) { - BUFFER_TRACE(where[i].bh, "call jbd2_journal_forget"); - ext4_journal_forget(handle, where[i].bh); - ext4_free_blocks(handle, inode, - le32_to_cpu(where[i-1].key), 1, 0); - } - ext4_free_blocks(handle, inode, le32_to_cpu(where[num].key), blks, 0); - - return err; -} - -/* - * 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. - * - * - * Need to be called with - * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block - * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem) - */ -int ext4_get_blocks_handle(handle_t *handle, struct inode *inode, - ext4_lblk_t iblock, unsigned long maxblocks, - struct buffer_head *bh_result, - int create, int extend_disksize) -{ - 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; - struct ext4_inode_info *ei = EXT4_I(inode); - int count = 0; - ext4_fsblk_t first_block = 0; - - - J_ASSERT(!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)); - J_ASSERT(handle != NULL || create == 0); - depth = ext4_block_to_path(inode, iblock, 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); - clear_buffer_new(bh_result); - count++; - /*map more blocks*/ - while (count < maxblocks && 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 (!create || err == -EIO) - goto cleanup; - + if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) + up_read((&EXT4_I(inode)->i_data_sem)); /* - * Okay, we need to do block allocation. Lazily initialize the block - * allocation info here if necessary - */ - if (S_ISREG(inode->i_mode) && (!ei->i_block_alloc_info)) - ext4_init_block_alloc_info(inode); - - goal = ext4_find_goal(inode, iblock, 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. + * Clear EXT4_MAP_FROM_CLUSTER and EXT4_MAP_BOUNDARY flag + * because it shouldn't be marked in es_map->m_flags. */ - count = ext4_blks_to_allocate(partial, indirect_blks, - maxblocks, blocks_to_boundary); - /* - * Block out ext4_truncate while we alter the tree - */ - err = ext4_alloc_branch(handle, inode, indirect_blks, &count, goal, - offsets + (partial - chain), partial); + map->m_flags &= ~(EXT4_MAP_FROM_CLUSTER | EXT4_MAP_BOUNDARY); /* - * 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 + * We don't check m_len because extent will be collpased in status + * tree. So the m_len might not equal. */ - if (!err) - err = ext4_splice_branch(handle, inode, iblock, - partial, indirect_blks, count); - /* - * i_disksize growing is protected by i_data_sem. Don't forget to - * protect it if you're about to implement concurrent - * ext4_get_block() -bzzz - */ - if (!err && extend_disksize && inode->i_size > ei->i_disksize) - ei->i_disksize = inode->i_size; - if (err) - goto cleanup; - - set_buffer_new(bh_result); -got_it: - map_bh(bh_result, inode->i_sb, le32_to_cpu(chain[depth-1].key)); - if (count > blocks_to_boundary) - set_buffer_boundary(bh_result); - 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--; - } - BUFFER_TRACE(bh_result, "returned"); -out: - return err; + if (es_map->m_lblk != map->m_lblk || + es_map->m_flags != map->m_flags || + es_map->m_pblk != map->m_pblk) { + printk("ES cache assertion failed for inode: %lu " + "es_cached ex [%d/%d/%llu/%x] != " + "found ex [%d/%d/%llu/%x] retval %d flags %x\n", + inode->i_ino, es_map->m_lblk, es_map->m_len, + es_map->m_pblk, es_map->m_flags, map->m_lblk, + map->m_len, map->m_pblk, map->m_flags, + retval, flags); + } } +#endif /* ES_AGGRESSIVE_TEST */ -/* Maximum number of blocks we map for direct IO at once. */ -#define DIO_MAX_BLOCKS 4096 /* - * Number of credits we need for writing DIO_MAX_BLOCKS: - * We need sb + group descriptor + bitmap + inode -> 4 - * For B blocks with A block pointers per block we need: - * 1 (triple ind.) + (B/A/A + 2) (doubly ind.) + (B/A + 2) (indirect). - * If we plug in 4096 for B and 256 for A (for 1KB block size), we get 25. - */ -#define DIO_CREDITS 25 - - -/* - * + * The ext4_map_blocks() function tries to look up the requested blocks, + * and returns if the blocks are already mapped. * - * ext4_ext4 get_block() wrapper function - * It will do a look up first, and returns if the blocks already mapped. * Otherwise it takes the write lock of the i_data_sem and allocate blocks * and store the allocated blocks in the result buffer head and mark it * mapped. * - * If file type is extents based, it will call ext4_ext_get_blocks(), - * Otherwise, call with ext4_get_blocks_handle() to handle indirect mapping + * If file type is extents based, it will call ext4_ext_map_blocks(), + * Otherwise, call with ext4_ind_map_blocks() to handle indirect mapping * based files * - * On success, it returns the number of blocks being mapped or allocate. - * if create==0 and the blocks are pre-allocated and uninitialized block, + * On success, it returns the number of blocks being mapped or allocated. + * if create==0 and the blocks are pre-allocated and unwritten block, * the result buffer head is unmapped. If the create ==1, it will make sure * the buffer head is mapped. * * It returns 0 if plain look up failed (blocks have not been allocated), in - * that casem, buffer head is unmapped + * that case, buffer head is unmapped * * It returns the error in case of allocation failure. */ -int ext4_get_blocks_wrap(handle_t *handle, struct inode *inode, sector_t block, - unsigned long max_blocks, struct buffer_head *bh, - int create, int extend_disksize) +int ext4_map_blocks(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, int flags) { + struct extent_status es; int retval; + int ret = 0; +#ifdef ES_AGGRESSIVE_TEST + struct ext4_map_blocks orig_map; + + memcpy(&orig_map, map, sizeof(*map)); +#endif - clear_buffer_mapped(bh); + map->m_flags = 0; + ext_debug("ext4_map_blocks(): inode %lu, flag %d, max_blocks %u," + "logical block %lu\n", inode->i_ino, flags, map->m_len, + (unsigned long) map->m_lblk); /* - * Try to see if we can get the block without requesting - * for new file system block. + * ext4_map_blocks returns an int, and m_len is an unsigned int */ - down_read((&EXT4_I(inode)->i_data_sem)); - if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) { - retval = ext4_ext_get_blocks(handle, inode, block, max_blocks, - bh, 0, 0); + if (unlikely(map->m_len > INT_MAX)) + map->m_len = INT_MAX; + + /* We can handle the block number less than EXT_MAX_BLOCKS */ + if (unlikely(map->m_lblk >= EXT_MAX_BLOCKS)) + return -EIO; + + /* Lookup extent status tree firstly */ + if (ext4_es_lookup_extent(inode, map->m_lblk, &es)) { + ext4_es_lru_add(inode); + if (ext4_es_is_written(&es) || ext4_es_is_unwritten(&es)) { + map->m_pblk = ext4_es_pblock(&es) + + map->m_lblk - es.es_lblk; + map->m_flags |= ext4_es_is_written(&es) ? + EXT4_MAP_MAPPED : EXT4_MAP_UNWRITTEN; + retval = es.es_len - (map->m_lblk - es.es_lblk); + if (retval > map->m_len) + retval = map->m_len; + map->m_len = retval; + } else if (ext4_es_is_delayed(&es) || ext4_es_is_hole(&es)) { + retval = 0; + } else { + BUG_ON(1); + } +#ifdef ES_AGGRESSIVE_TEST + ext4_map_blocks_es_recheck(handle, inode, map, + &orig_map, flags); +#endif + goto found; + } + + /* + * Try to see if we can get the block without requesting a new + * file system block. + */ + if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) + down_read(&EXT4_I(inode)->i_data_sem); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + retval = ext4_ext_map_blocks(handle, inode, map, flags & + EXT4_GET_BLOCKS_KEEP_SIZE); } else { - retval = ext4_get_blocks_handle(handle, - inode, block, max_blocks, bh, 0, 0); + retval = ext4_ind_map_blocks(handle, inode, map, flags & + EXT4_GET_BLOCKS_KEEP_SIZE); + } + if (retval > 0) { + unsigned int status; + + if (unlikely(retval != map->m_len)) { + ext4_warning(inode->i_sb, + "ES len assertion failed for inode " + "%lu: retval %d != map->m_len %d", + inode->i_ino, retval, map->m_len); + WARN_ON(1); + } + + status = map->m_flags & EXT4_MAP_UNWRITTEN ? + EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; + if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && + ext4_find_delalloc_range(inode, map->m_lblk, + map->m_lblk + map->m_len - 1)) + status |= EXTENT_STATUS_DELAYED; + ret = ext4_es_insert_extent(inode, map->m_lblk, + map->m_len, map->m_pblk, status); + if (ret < 0) + retval = ret; + } + if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) + up_read((&EXT4_I(inode)->i_data_sem)); + +found: + if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { + ret = check_block_validity(inode, map); + if (ret != 0) + return ret; } - up_read((&EXT4_I(inode)->i_data_sem)); /* If it is only a block(s) look up */ - if (!create) + if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) return retval; /* * Returns if the blocks have already allocated * * Note that if blocks have been preallocated - * ext4_ext_get_block() returns th create = 0 + * ext4_ext_get_block() returns the create = 0 * with buffer head unmapped. */ - if (retval > 0 && buffer_mapped(bh)) - return retval; + if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) + /* + * If we need to convert extent to unwritten + * we continue and do the actual work in + * ext4_ext_map_blocks() + */ + if (!(flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN)) + return retval; /* - * New blocks allocate and/or writing to uninitialized extent + * Here we clear m_flags because after allocating an new extent, + * it will be set again. + */ + map->m_flags &= ~EXT4_MAP_FLAGS; + + /* + * New blocks allocate and/or writing to unwritten extent * will possibly result in updating i_data, so we take * the write lock of i_data_sem, and call get_blocks() * with create == 1 flag. */ - down_write((&EXT4_I(inode)->i_data_sem)); + down_write(&EXT4_I(inode)->i_data_sem); + + /* + * if the caller is from delayed allocation writeout path + * we have already reserved fs blocks for allocation + * let the underlying get_block() function know to + * avoid double accounting + */ + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) + ext4_set_inode_state(inode, EXT4_STATE_DELALLOC_RESERVED); /* * We need to check for EXT4 here because migrate * could have changed the inode type in between */ - if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) { - retval = ext4_ext_get_blocks(handle, inode, block, max_blocks, - bh, create, extend_disksize); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + retval = ext4_ext_map_blocks(handle, inode, map, flags); } else { - retval = ext4_get_blocks_handle(handle, inode, block, - max_blocks, bh, create, extend_disksize); + retval = ext4_ind_map_blocks(handle, inode, map, flags); + + if (retval > 0 && map->m_flags & EXT4_MAP_NEW) { + /* + * We allocated new blocks which will result in + * i_data's format changing. Force the migrate + * to fail by clearing migrate flags + */ + ext4_clear_inode_state(inode, EXT4_STATE_EXT_MIGRATE); + } + + /* + * Update reserved blocks/metadata blocks after successful + * block allocation which had been deferred till now. We don't + * support fallocate for non extent files. So we can update + * reserve space here. + */ + if ((retval > 0) && + (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)) + ext4_da_update_reserve_space(inode, retval, 1); } + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) + ext4_clear_inode_state(inode, EXT4_STATE_DELALLOC_RESERVED); + + if (retval > 0) { + unsigned int status; + + if (unlikely(retval != map->m_len)) { + ext4_warning(inode->i_sb, + "ES len assertion failed for inode " + "%lu: retval %d != map->m_len %d", + inode->i_ino, retval, map->m_len); + WARN_ON(1); + } + + /* + * If the extent has been zeroed out, we don't need to update + * extent status tree. + */ + if ((flags & EXT4_GET_BLOCKS_PRE_IO) && + ext4_es_lookup_extent(inode, map->m_lblk, &es)) { + if (ext4_es_is_written(&es)) + goto has_zeroout; + } + status = map->m_flags & EXT4_MAP_UNWRITTEN ? + EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; + if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && + ext4_find_delalloc_range(inode, map->m_lblk, + map->m_lblk + map->m_len - 1)) + status |= EXTENT_STATUS_DELAYED; + ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + map->m_pblk, status); + if (ret < 0) + retval = ret; + } + +has_zeroout: up_write((&EXT4_I(inode)->i_data_sem)); + if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { + ret = check_block_validity(inode, map); + if (ret != 0) + return ret; + } return retval; } -static int ext4_get_block(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) +/* Maximum number of blocks we map for direct IO at once. */ +#define DIO_MAX_BLOCKS 4096 + +static int _ext4_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int flags) { handle_t *handle = ext4_journal_current_handle(); + struct ext4_map_blocks map; int ret = 0, started = 0; - unsigned max_blocks = bh_result->b_size >> inode->i_blkbits; + int dio_credits; + + if (ext4_has_inline_data(inode)) + return -ERANGE; + + map.m_lblk = iblock; + map.m_len = bh->b_size >> inode->i_blkbits; - if (create && !handle) { + if (flags && !(flags & EXT4_GET_BLOCKS_NO_LOCK) && !handle) { /* Direct IO write... */ - if (max_blocks > DIO_MAX_BLOCKS) - max_blocks = DIO_MAX_BLOCKS; - handle = ext4_journal_start(inode, DIO_CREDITS + - 2 * EXT4_QUOTA_TRANS_BLOCKS(inode->i_sb)); + if (map.m_len > DIO_MAX_BLOCKS) + map.m_len = DIO_MAX_BLOCKS; + dio_credits = ext4_chunk_trans_blocks(inode, map.m_len); + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, + dio_credits); if (IS_ERR(handle)) { ret = PTR_ERR(handle); - goto out; + return ret; } started = 1; } - ret = ext4_get_blocks_wrap(handle, inode, iblock, - max_blocks, bh_result, create, 0); + ret = ext4_map_blocks(handle, inode, &map, flags); if (ret > 0) { - bh_result->b_size = (ret << inode->i_blkbits); + ext4_io_end_t *io_end = ext4_inode_aio(inode); + + map_bh(bh, inode->i_sb, map.m_pblk); + bh->b_state = (bh->b_state & ~EXT4_MAP_FLAGS) | map.m_flags; + if (io_end && io_end->flag & EXT4_IO_END_UNWRITTEN) + set_buffer_defer_completion(bh); + bh->b_size = inode->i_sb->s_blocksize * map.m_len; ret = 0; } if (started) ext4_journal_stop(handle); -out: return ret; } +int ext4_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int create) +{ + return _ext4_get_block(inode, iblock, bh, + create ? EXT4_GET_BLOCKS_CREATE : 0); +} + /* * `handle' can be NULL if create is zero */ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode, ext4_lblk_t block, int create, int *errp) { - struct buffer_head dummy; + struct ext4_map_blocks map; + struct buffer_head *bh; int fatal = 0, err; J_ASSERT(handle != NULL || create == 0); - dummy.b_state = 0; - dummy.b_blocknr = -1000; - buffer_trace_init(&dummy.b_history); - err = ext4_get_blocks_wrap(handle, inode, block, 1, - &dummy, create, 1); - /* - * ext4_get_blocks_handle() returns number of blocks - * mapped. 0 in case of a HOLE. - */ - if (err > 0) { - if (err > 1) - WARN_ON(1); - err = 0; + map.m_lblk = block; + map.m_len = 1; + err = ext4_map_blocks(handle, inode, &map, + create ? EXT4_GET_BLOCKS_CREATE : 0); + + /* ensure we send some value back into *errp */ + *errp = 0; + + if (create && err == 0) + err = -ENOSPC; /* should never happen */ + if (err < 0) + *errp = err; + if (err <= 0) + return NULL; + + bh = sb_getblk(inode->i_sb, map.m_pblk); + if (unlikely(!bh)) { + *errp = -ENOMEM; + return NULL; } - *errp = err; - if (!err && buffer_mapped(&dummy)) { - struct buffer_head *bh; - bh = sb_getblk(inode->i_sb, dummy.b_blocknr); - if (!bh) { - *errp = -EIO; - goto err; - } - if (buffer_new(&dummy)) { - J_ASSERT(create != 0); - J_ASSERT(handle != NULL); + if (map.m_flags & EXT4_MAP_NEW) { + J_ASSERT(create != 0); + J_ASSERT(handle != NULL); - /* - * Now that we do not always journal data, we should - * keep in mind whether this should always journal the - * new buffer as metadata. For now, regular file - * writes use ext4_get_block instead, so it's not a - * problem. - */ - lock_buffer(bh); - BUFFER_TRACE(bh, "call get_create_access"); - fatal = ext4_journal_get_create_access(handle, bh); - if (!fatal && !buffer_uptodate(bh)) { - memset(bh->b_data,0,inode->i_sb->s_blocksize); - set_buffer_uptodate(bh); - } - unlock_buffer(bh); - BUFFER_TRACE(bh, "call ext4_journal_dirty_metadata"); - err = ext4_journal_dirty_metadata(handle, bh); - if (!fatal) - fatal = err; - } else { - BUFFER_TRACE(bh, "not a new buffer"); - } - if (fatal) { - *errp = fatal; - brelse(bh); - bh = NULL; + /* + * Now that we do not always journal data, we should + * keep in mind whether this should always journal the + * new buffer as metadata. For now, regular file + * writes use ext4_get_block instead, so it's not a + * problem. + */ + lock_buffer(bh); + BUFFER_TRACE(bh, "call get_create_access"); + fatal = ext4_journal_get_create_access(handle, bh); + if (!fatal && !buffer_uptodate(bh)) { + memset(bh->b_data, 0, inode->i_sb->s_blocksize); + set_buffer_uptodate(bh); } - return bh; + unlock_buffer(bh); + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (!fatal) + fatal = err; + } else { + BUFFER_TRACE(bh, "not a new buffer"); } -err: - return NULL; + if (fatal) { + *errp = fatal; + brelse(bh); + bh = NULL; + } + return bh; } struct buffer_head *ext4_bread(handle_t *handle, struct inode *inode, ext4_lblk_t block, int create, int *err) { - struct buffer_head * bh; + struct buffer_head *bh; bh = ext4_getblk(handle, inode, block, create, err); if (!bh) return bh; if (buffer_uptodate(bh)) return bh; - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META | REQ_PRIO, 1, &bh); wait_on_buffer(bh); if (buffer_uptodate(bh)) return bh; @@ -1111,13 +852,13 @@ struct buffer_head *ext4_bread(handle_t *handle, struct inode *inode, return NULL; } -static int walk_page_buffers( handle_t *handle, - struct buffer_head *head, - unsigned from, - unsigned to, - int *partial, - int (*fn)( handle_t *handle, - struct buffer_head *bh)) +int ext4_walk_page_buffers(handle_t *handle, + struct buffer_head *head, + unsigned from, + unsigned to, + int *partial, + int (*fn)(handle_t *handle, + struct buffer_head *bh)) { struct buffer_head *bh; unsigned block_start, block_end; @@ -1125,10 +866,9 @@ static int walk_page_buffers( handle_t *handle, int err, ret = 0; struct buffer_head *next; - for ( bh = head, block_start = 0; - ret == 0 && (bh != head || !block_start); - block_start = block_end, bh = next) - { + for (bh = head, block_start = 0; + ret == 0 && (bh != head || !block_start); + block_start = block_end, bh = next) { next = bh->b_this_page; block_end = block_start + blocksize; if (block_end <= from || block_start >= to) { @@ -1150,11 +890,10 @@ static int walk_page_buffers( handle_t *handle, * and the commit_write(). So doing the jbd2_journal_start at the start of * prepare_write() is the right place. * - * Also, this function can nest inside ext4_writepage() -> - * block_write_full_page(). In that case, we *know* that ext4_writepage() - * has generated enough buffer credits to do the whole page. So we won't - * block on the journal in that case, which is good, because the caller may - * be PF_MEMALLOC. + * Also, this function can nest inside ext4_writepage(). In that case, we + * *know* that ext4_writepage() has generated enough buffer credits to do the + * whole page. So we won't block on the journal in that case, which is good, + * because the caller may be PF_MEMALLOC. * * By accident, ext4 can be reentered when a transaction is open via * quota file writes. If we were to commit the transaction while thus @@ -1168,103 +907,153 @@ static int walk_page_buffers( handle_t *handle, * is elevated. We'll still have enough credits for the tiny quotafile * write. */ -static int do_journal_get_write_access(handle_t *handle, - struct buffer_head *bh) +int do_journal_get_write_access(handle_t *handle, + struct buffer_head *bh) { + int dirty = buffer_dirty(bh); + int ret; + if (!buffer_mapped(bh) || buffer_freed(bh)) return 0; - return ext4_journal_get_write_access(handle, bh); + /* + * __block_write_begin() could have dirtied some buffers. Clean + * the dirty bit as jbd2_journal_get_write_access() could complain + * otherwise about fs integrity issues. Setting of the dirty bit + * by __block_write_begin() isn't a real problem here as we clear + * the bit before releasing a page lock and thus writeback cannot + * ever write the buffer. + */ + if (dirty) + clear_buffer_dirty(bh); + BUFFER_TRACE(bh, "get write access"); + ret = ext4_journal_get_write_access(handle, bh); + if (!ret && dirty) + ret = ext4_handle_dirty_metadata(handle, NULL, bh); + return ret; } +static int ext4_get_block_write_nolock(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, - loff_t pos, unsigned len, unsigned flags, - struct page **pagep, void **fsdata) + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) { - struct inode *inode = mapping->host; - int ret, needed_blocks = ext4_writepage_trans_blocks(inode); + struct inode *inode = mapping->host; + int ret, needed_blocks; handle_t *handle; int retries = 0; - struct page *page; - pgoff_t index; - unsigned from, to; + struct page *page; + pgoff_t index; + unsigned from, to; - index = pos >> PAGE_CACHE_SHIFT; - from = pos & (PAGE_CACHE_SIZE - 1); - to = from + len; + trace_ext4_write_begin(inode, pos, len, flags); + /* + * Reserve one block more for addition to orphan list in case + * we allocate blocks but write fails for some reason + */ + needed_blocks = ext4_writepage_trans_blocks(inode) + 1; + index = pos >> PAGE_CACHE_SHIFT; + from = pos & (PAGE_CACHE_SIZE - 1); + to = from + len; -retry: - page = __grab_cache_page(mapping, index); - if (!page) - return -ENOMEM; - *pagep = page; + if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) { + ret = ext4_try_to_write_inline_data(mapping, inode, pos, len, + flags, pagep); + if (ret < 0) + return ret; + if (ret == 1) + return 0; + } - handle = ext4_journal_start(inode, needed_blocks); - if (IS_ERR(handle)) { - unlock_page(page); - page_cache_release(page); - ret = PTR_ERR(handle); - goto out; + /* + * grab_cache_page_write_begin() can take a long time if the + * system is thrashing due to memory pressure, or if the page + * is being written back. So grab it first before we start + * the transaction handle. This also allows us to allocate + * the page (if needed) without using GFP_NOFS. + */ +retry_grab: + page = grab_cache_page_write_begin(mapping, index, flags); + if (!page) + return -ENOMEM; + unlock_page(page); + +retry_journal: + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); + if (IS_ERR(handle)) { + page_cache_release(page); + return PTR_ERR(handle); } - ret = block_write_begin(file, mapping, pos, len, flags, pagep, fsdata, - ext4_get_block); + lock_page(page); + if (page->mapping != mapping) { + /* The page got truncated from under us */ + unlock_page(page); + page_cache_release(page); + ext4_journal_stop(handle); + goto retry_grab; + } + /* In case writeback began while the page was unlocked */ + wait_for_stable_page(page); + + if (ext4_should_dioread_nolock(inode)) + ret = __block_write_begin(page, pos, len, ext4_get_block_write); + else + ret = __block_write_begin(page, pos, len, ext4_get_block); if (!ret && ext4_should_journal_data(inode)) { - ret = walk_page_buffers(handle, page_buffers(page), - from, to, NULL, do_journal_get_write_access); + ret = ext4_walk_page_buffers(handle, page_buffers(page), + from, to, NULL, + do_journal_get_write_access); } if (ret) { + unlock_page(page); + /* + * __block_write_begin may have instantiated a few blocks + * outside i_size. Trim these off again. Don't need + * i_size_read because we hold i_mutex. + * + * Add inode to orphan list in case we crash before + * truncate finishes + */ + if (pos + len > inode->i_size && ext4_can_truncate(inode)) + ext4_orphan_add(handle, inode); + ext4_journal_stop(handle); - unlock_page(page); - page_cache_release(page); - } + if (pos + len > inode->i_size) { + ext4_truncate_failed_write(inode); + /* + * If truncate failed early the inode might + * still be on the orphan list; we need to + * make sure the inode is removed from the + * orphan list in that case. + */ + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + } - if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) - goto retry; -out: + if (ret == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry_journal; + page_cache_release(page); + return ret; + } + *pagep = page; return ret; } -int ext4_journal_dirty_data(handle_t *handle, struct buffer_head *bh) -{ - int err = jbd2_journal_dirty_data(handle, bh); - if (err) - ext4_journal_abort_handle(__FUNCTION__, __FUNCTION__, - bh, handle, err); - return err; -} - /* For write_end() in data=journal mode */ static int write_end_fn(handle_t *handle, struct buffer_head *bh) { + int ret; if (!buffer_mapped(bh) || buffer_freed(bh)) return 0; set_buffer_uptodate(bh); - return ext4_journal_dirty_metadata(handle, bh); -} - -/* - * Generic write_end handler for ordered and writeback ext4 journal modes. - * We can't use generic_write_end, because that unlocks the page and we need to - * unlock the page after ext4_journal_stop, but ext4_journal_stop must run - * after block_write_end. - */ -static int ext4_generic_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) -{ - struct inode *inode = file->f_mapping->host; - - copied = block_write_end(file, mapping, pos, len, copied, page, fsdata); - - if (pos+copied > inode->i_size) { - i_size_write(inode, pos+copied); - mark_inode_dirty(inode); - } - - return copied; + ret = ext4_handle_dirty_metadata(handle, NULL, bh); + clear_buffer_meta(bh); + clear_buffer_prio(bh); + return ret; } /* @@ -1274,168 +1063,660 @@ static int ext4_generic_write_end(struct file *file, * ext4 never places buffers on inode->i_mapping->private_list. metadata * buffers are managed internally. */ -static int ext4_ordered_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) +static int ext4_write_end(struct file *file, + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { handle_t *handle = ext4_journal_current_handle(); - struct inode *inode = file->f_mapping->host; - unsigned from, to; + struct inode *inode = mapping->host; int ret = 0, ret2; + int i_size_changed = 0; - from = pos & (PAGE_CACHE_SIZE - 1); - to = from + len; + trace_ext4_write_end(inode, pos, len, copied); + if (ext4_test_inode_state(inode, EXT4_STATE_ORDERED_MODE)) { + ret = ext4_jbd2_file_inode(handle, inode); + if (ret) { + unlock_page(page); + page_cache_release(page); + goto errout; + } + } - ret = walk_page_buffers(handle, page_buffers(page), - from, to, NULL, ext4_journal_dirty_data); + if (ext4_has_inline_data(inode)) { + ret = ext4_write_inline_data_end(inode, pos, len, + copied, page); + if (ret < 0) + goto errout; + copied = ret; + } else + copied = block_write_end(file, mapping, pos, + len, copied, page, fsdata); - if (ret == 0) { - /* - * generic_write_end() will run mark_inode_dirty() if i_size - * changes. So let's piggyback the i_disksize mark_inode_dirty - * into that. - */ - loff_t new_i_size; + /* + * No need to use i_size_read() here, the i_size + * cannot change under us because we hole i_mutex. + * + * But it's important to update i_size while still holding page lock: + * page writeout could otherwise come in and zero beyond i_size. + */ + if (pos + copied > inode->i_size) { + i_size_write(inode, pos + copied); + i_size_changed = 1; + } - new_i_size = pos + copied; - if (new_i_size > EXT4_I(inode)->i_disksize) - EXT4_I(inode)->i_disksize = new_i_size; - copied = ext4_generic_write_end(file, mapping, pos, len, copied, - page, fsdata); - if (copied < 0) - ret = copied; + if (pos + copied > EXT4_I(inode)->i_disksize) { + /* We need to mark inode dirty even if + * new_i_size is less that inode->i_size + * but greater than i_disksize. (hint delalloc) + */ + ext4_update_i_disksize(inode, (pos + copied)); + i_size_changed = 1; } - ret2 = ext4_journal_stop(handle); - if (!ret) - ret = ret2; unlock_page(page); page_cache_release(page); - return ret ? ret : copied; -} - -static int ext4_writeback_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) -{ - handle_t *handle = ext4_journal_current_handle(); - struct inode *inode = file->f_mapping->host; - int ret = 0, ret2; - loff_t new_i_size; - - new_i_size = pos + copied; - if (new_i_size > EXT4_I(inode)->i_disksize) - EXT4_I(inode)->i_disksize = new_i_size; - - copied = ext4_generic_write_end(file, mapping, pos, len, copied, - page, fsdata); - if (copied < 0) - ret = copied; + /* + * Don't mark the inode dirty under page lock. First, it unnecessarily + * makes the holding time of page lock longer. Second, it forces lock + * ordering of page lock and transaction start for journaling + * filesystems. + */ + if (i_size_changed) + ext4_mark_inode_dirty(handle, inode); + if (pos + len > inode->i_size && ext4_can_truncate(inode)) + /* if we have allocated more blocks and copied + * less. We will have blocks allocated outside + * inode->i_size. So truncate them + */ + ext4_orphan_add(handle, inode); +errout: ret2 = ext4_journal_stop(handle); if (!ret) ret = ret2; - unlock_page(page); - page_cache_release(page); + + if (pos + len > inode->i_size) { + ext4_truncate_failed_write(inode); + /* + * If truncate failed early the inode might still be + * on the orphan list; we need to make sure the inode + * is removed from the orphan list in that case. + */ + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + } return ret ? ret : copied; } static int ext4_journalled_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { handle_t *handle = ext4_journal_current_handle(); struct inode *inode = mapping->host; int ret = 0, ret2; int partial = 0; unsigned from, to; + loff_t new_i_size; + trace_ext4_journalled_write_end(inode, pos, len, copied); from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; - if (copied < len) { - if (!PageUptodate(page)) - copied = 0; - page_zero_new_buffers(page, from+copied, to); - } + BUG_ON(!ext4_handle_valid(handle)); - ret = walk_page_buffers(handle, page_buffers(page), from, - to, &partial, write_end_fn); - if (!partial) - SetPageUptodate(page); - if (pos+copied > inode->i_size) + if (ext4_has_inline_data(inode)) + copied = ext4_write_inline_data_end(inode, pos, len, + copied, page); + else { + if (copied < len) { + if (!PageUptodate(page)) + copied = 0; + page_zero_new_buffers(page, from+copied, to); + } + + ret = ext4_walk_page_buffers(handle, page_buffers(page), from, + to, &partial, write_end_fn); + if (!partial) + SetPageUptodate(page); + } + new_i_size = pos + copied; + if (new_i_size > inode->i_size) i_size_write(inode, pos+copied); - EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; - if (inode->i_size > EXT4_I(inode)->i_disksize) { - EXT4_I(inode)->i_disksize = inode->i_size; + 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); if (!ret) ret = ret2; } + unlock_page(page); + page_cache_release(page); + if (pos + len > inode->i_size && ext4_can_truncate(inode)) + /* if we have allocated more blocks and copied + * less. We will have blocks allocated outside + * inode->i_size. So truncate them + */ + ext4_orphan_add(handle, inode); + ret2 = ext4_journal_stop(handle); if (!ret) ret = ret2; - unlock_page(page); - page_cache_release(page); + if (pos + len > inode->i_size) { + ext4_truncate_failed_write(inode); + /* + * If truncate failed early the inode might still be + * on the orphan list; we need to make sure the inode + * is removed from the orphan list in that case. + */ + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + } return ret ? ret : copied; } /* - * bmap() is special. It gets used by applications such as lilo and by - * the swapper to find the on-disk block of a specific piece of data. - * - * Naturally, this is dangerous if the block concerned is still in the - * journal. If somebody makes a swapfile on an ext4 data-journaling - * filesystem and enables swap, then they may get a nasty shock when the - * data getting swapped to that swapfile suddenly gets overwritten by - * the original zero's written out previously to the journal and - * awaiting writeback in the kernel's buffer cache. - * - * So, if we see any bmap calls here on a modified, data-journaled file, - * take extra steps to flush any blocks which might be in the cache. + * Reserve a metadata for a single block located at lblock */ -static sector_t ext4_bmap(struct address_space *mapping, sector_t block) +static int ext4_da_reserve_metadata(struct inode *inode, ext4_lblk_t lblock) { - struct inode *inode = mapping->host; - journal_t *journal; - int err; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_inode_info *ei = EXT4_I(inode); + unsigned int md_needed; + ext4_lblk_t save_last_lblock; + int save_len; - if (EXT4_I(inode)->i_state & EXT4_STATE_JDATA) { + /* + * recalculate the amount of metadata blocks to reserve + * in order to allocate nrblocks + * worse case is one extent per block + */ + spin_lock(&ei->i_block_reservation_lock); + /* + * ext4_calc_metadata_amount() has side effects, which we have + * to be prepared undo if we fail to claim space. + */ + save_len = ei->i_da_metadata_calc_len; + save_last_lblock = ei->i_da_metadata_calc_last_lblock; + md_needed = EXT4_NUM_B2C(sbi, + ext4_calc_metadata_amount(inode, lblock)); + trace_ext4_da_reserve_space(inode, md_needed); + + /* + * We do still charge estimated metadata to the sb though; + * we cannot afford to run out of free blocks. + */ + if (ext4_claim_free_clusters(sbi, md_needed, 0)) { + ei->i_da_metadata_calc_len = save_len; + ei->i_da_metadata_calc_last_lblock = save_last_lblock; + spin_unlock(&ei->i_block_reservation_lock); + return -ENOSPC; + } + ei->i_reserved_meta_blocks += md_needed; + spin_unlock(&ei->i_block_reservation_lock); + + return 0; /* success */ +} + +/* + * Reserve a single cluster located at lblock + */ +static int ext4_da_reserve_space(struct inode *inode, ext4_lblk_t lblock) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_inode_info *ei = EXT4_I(inode); + unsigned int md_needed; + int ret; + ext4_lblk_t save_last_lblock; + int save_len; + + /* + * We will charge metadata quota at writeout time; this saves + * us from metadata over-estimation, though we may go over by + * a small amount in the end. Here we just reserve for data. + */ + ret = dquot_reserve_block(inode, EXT4_C2B(sbi, 1)); + if (ret) + return ret; + + /* + * recalculate the amount of metadata blocks to reserve + * in order to allocate nrblocks + * worse case is one extent per block + */ + spin_lock(&ei->i_block_reservation_lock); + /* + * ext4_calc_metadata_amount() has side effects, which we have + * to be prepared undo if we fail to claim space. + */ + save_len = ei->i_da_metadata_calc_len; + save_last_lblock = ei->i_da_metadata_calc_last_lblock; + md_needed = EXT4_NUM_B2C(sbi, + ext4_calc_metadata_amount(inode, lblock)); + trace_ext4_da_reserve_space(inode, md_needed); + + /* + * We do still charge estimated metadata to the sb though; + * we cannot afford to run out of free blocks. + */ + if (ext4_claim_free_clusters(sbi, md_needed + 1, 0)) { + ei->i_da_metadata_calc_len = save_len; + ei->i_da_metadata_calc_last_lblock = save_last_lblock; + spin_unlock(&ei->i_block_reservation_lock); + dquot_release_reservation_block(inode, EXT4_C2B(sbi, 1)); + return -ENOSPC; + } + ei->i_reserved_data_blocks++; + ei->i_reserved_meta_blocks += md_needed; + spin_unlock(&ei->i_block_reservation_lock); + + return 0; /* success */ +} + +static void ext4_da_release_space(struct inode *inode, int to_free) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_inode_info *ei = EXT4_I(inode); + + if (!to_free) + return; /* Nothing to release, exit */ + + spin_lock(&EXT4_I(inode)->i_block_reservation_lock); + + trace_ext4_da_release_space(inode, to_free); + if (unlikely(to_free > ei->i_reserved_data_blocks)) { /* - * This is a REALLY heavyweight approach, but the use of - * bmap on dirty files is expected to be extremely rare: - * only if we run lilo or swapon on a freshly made file - * do we expect this to happen. - * - * (bmap requires CAP_SYS_RAWIO so this does not - * represent an unprivileged user DOS attack --- we'd be - * in trouble if mortal users could trigger this path at - * will.) - * - * NB. EXT4_STATE_JDATA is not set on files other than - * regular files. If somebody wants to bmap a directory - * or symlink and gets confused because the buffer - * hasn't yet been flushed to disk, they deserve - * everything they get. + * if there aren't enough reserved blocks, then the + * counter is messed up somewhere. Since this + * function is called from invalidate page, it's + * harmless to return without any action. */ + ext4_warning(inode->i_sb, "ext4_da_release_space: " + "ino %lu, to_free %d with only %d reserved " + "data blocks", inode->i_ino, to_free, + ei->i_reserved_data_blocks); + WARN_ON(1); + to_free = ei->i_reserved_data_blocks; + } + ei->i_reserved_data_blocks -= to_free; - EXT4_I(inode)->i_state &= ~EXT4_STATE_JDATA; - journal = EXT4_JOURNAL(inode); - jbd2_journal_lock_updates(journal); - err = jbd2_journal_flush(journal); - jbd2_journal_unlock_updates(journal); + if (ei->i_reserved_data_blocks == 0) { + /* + * We can release all of the reserved metadata blocks + * only when we have written all of the delayed + * allocation blocks. + * Note that in case of bigalloc, i_reserved_meta_blocks, + * i_reserved_data_blocks, etc. refer to number of clusters. + */ + percpu_counter_sub(&sbi->s_dirtyclusters_counter, + ei->i_reserved_meta_blocks); + ei->i_reserved_meta_blocks = 0; + ei->i_da_metadata_calc_len = 0; + } - if (err) + /* update fs dirty data blocks counter */ + percpu_counter_sub(&sbi->s_dirtyclusters_counter, to_free); + + spin_unlock(&EXT4_I(inode)->i_block_reservation_lock); + + dquot_release_reservation_block(inode, EXT4_C2B(sbi, to_free)); +} + +static void ext4_da_page_release_reservation(struct page *page, + unsigned int offset, + unsigned int length) +{ + int to_release = 0; + struct buffer_head *head, *bh; + unsigned int curr_off = 0; + struct inode *inode = page->mapping->host; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + unsigned int stop = offset + length; + int num_clusters; + ext4_fsblk_t lblk; + + BUG_ON(stop > PAGE_CACHE_SIZE || stop < length); + + head = page_buffers(page); + bh = head; + do { + unsigned int next_off = curr_off + bh->b_size; + + if (next_off > stop) + break; + + if ((offset <= curr_off) && (buffer_delay(bh))) { + to_release++; + clear_buffer_delay(bh); + } + curr_off = next_off; + } while ((bh = bh->b_this_page) != head); + + if (to_release) { + lblk = page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); + ext4_es_remove_extent(inode, lblk, to_release); + } + + /* If we have released all the blocks belonging to a cluster, then we + * need to release the reserved space for that cluster. */ + num_clusters = EXT4_NUM_B2C(sbi, to_release); + while (num_clusters > 0) { + lblk = (page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits)) + + ((num_clusters - 1) << sbi->s_cluster_bits); + if (sbi->s_cluster_ratio == 1 || + !ext4_find_delalloc_cluster(inode, lblk)) + ext4_da_release_space(inode, 1); + + num_clusters--; + } +} + +/* + * Delayed allocation stuff + */ + +struct mpage_da_data { + struct inode *inode; + struct writeback_control *wbc; + + pgoff_t first_page; /* The first page to write */ + pgoff_t next_page; /* Current page to examine */ + pgoff_t last_page; /* Last page to examine */ + /* + * Extent to map - this can be after first_page because that can be + * fully mapped. We somewhat abuse m_flags to store whether the extent + * is delalloc or unwritten. + */ + struct ext4_map_blocks map; + struct ext4_io_submit io_submit; /* IO submission data */ +}; + +static void mpage_release_unused_pages(struct mpage_da_data *mpd, + bool invalidate) +{ + int nr_pages, i; + pgoff_t index, end; + struct pagevec pvec; + struct inode *inode = mpd->inode; + struct address_space *mapping = inode->i_mapping; + + /* This is necessary when next_page == 0. */ + if (mpd->first_page >= mpd->next_page) + return; + + index = mpd->first_page; + end = mpd->next_page - 1; + if (invalidate) { + ext4_lblk_t start, last; + start = index << (PAGE_CACHE_SHIFT - inode->i_blkbits); + last = end << (PAGE_CACHE_SHIFT - inode->i_blkbits); + ext4_es_remove_extent(inode, start, last - start + 1); + } + + pagevec_init(&pvec, 0); + while (index <= end) { + nr_pages = pagevec_lookup(&pvec, mapping, index, PAGEVEC_SIZE); + if (nr_pages == 0) + break; + for (i = 0; i < nr_pages; i++) { + struct page *page = pvec.pages[i]; + if (page->index > end) + break; + BUG_ON(!PageLocked(page)); + BUG_ON(PageWriteback(page)); + if (invalidate) { + block_invalidatepage(page, 0, PAGE_CACHE_SIZE); + ClearPageUptodate(page); + } + unlock_page(page); + } + index = pvec.pages[nr_pages - 1]->index + 1; + pagevec_release(&pvec); + } +} + +static void ext4_print_free_blocks(struct inode *inode) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct super_block *sb = inode->i_sb; + struct ext4_inode_info *ei = EXT4_I(inode); + + ext4_msg(sb, KERN_CRIT, "Total free blocks count %lld", + EXT4_C2B(EXT4_SB(inode->i_sb), + ext4_count_free_clusters(sb))); + ext4_msg(sb, KERN_CRIT, "Free/Dirty block details"); + ext4_msg(sb, KERN_CRIT, "free_blocks=%lld", + (long long) EXT4_C2B(EXT4_SB(sb), + percpu_counter_sum(&sbi->s_freeclusters_counter))); + ext4_msg(sb, KERN_CRIT, "dirty_blocks=%lld", + (long long) EXT4_C2B(EXT4_SB(sb), + percpu_counter_sum(&sbi->s_dirtyclusters_counter))); + ext4_msg(sb, KERN_CRIT, "Block reservation details"); + ext4_msg(sb, KERN_CRIT, "i_reserved_data_blocks=%u", + ei->i_reserved_data_blocks); + ext4_msg(sb, KERN_CRIT, "i_reserved_meta_blocks=%u", + ei->i_reserved_meta_blocks); + ext4_msg(sb, KERN_CRIT, "i_allocated_meta_blocks=%u", + ei->i_allocated_meta_blocks); + return; +} + +static int ext4_bh_delay_or_unwritten(handle_t *handle, struct buffer_head *bh) +{ + return (buffer_delay(bh) || buffer_unwritten(bh)) && buffer_dirty(bh); +} + +/* + * This function is grabs code from the very beginning of + * ext4_map_blocks, but assumes that the caller is from delayed write + * time. This function looks up the requested blocks and sets the + * buffer delay bit under the protection of i_data_sem. + */ +static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, + struct ext4_map_blocks *map, + struct buffer_head *bh) +{ + struct extent_status es; + int retval; + sector_t invalid_block = ~((sector_t) 0xffff); +#ifdef ES_AGGRESSIVE_TEST + struct ext4_map_blocks orig_map; + + memcpy(&orig_map, map, sizeof(*map)); +#endif + + if (invalid_block < ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es)) + invalid_block = ~0; + + map->m_flags = 0; + ext_debug("ext4_da_map_blocks(): inode %lu, max_blocks %u," + "logical block %lu\n", inode->i_ino, map->m_len, + (unsigned long) map->m_lblk); + + /* Lookup extent status tree firstly */ + if (ext4_es_lookup_extent(inode, iblock, &es)) { + ext4_es_lru_add(inode); + if (ext4_es_is_hole(&es)) { + retval = 0; + down_read(&EXT4_I(inode)->i_data_sem); + goto add_delayed; + } + + /* + * Delayed extent could be allocated by fallocate. + * So we need to check it. + */ + if (ext4_es_is_delayed(&es) && !ext4_es_is_unwritten(&es)) { + map_bh(bh, inode->i_sb, invalid_block); + set_buffer_new(bh); + set_buffer_delay(bh); return 0; + } + + map->m_pblk = ext4_es_pblock(&es) + iblock - es.es_lblk; + retval = es.es_len - (iblock - es.es_lblk); + if (retval > map->m_len) + retval = map->m_len; + map->m_len = retval; + if (ext4_es_is_written(&es)) + map->m_flags |= EXT4_MAP_MAPPED; + else if (ext4_es_is_unwritten(&es)) + map->m_flags |= EXT4_MAP_UNWRITTEN; + else + BUG_ON(1); + +#ifdef ES_AGGRESSIVE_TEST + ext4_map_blocks_es_recheck(NULL, inode, map, &orig_map, 0); +#endif + return retval; } - return generic_block_bmap(mapping,block,ext4_get_block); + /* + * Try to see if we can get the block without requesting a new + * file system block. + */ + down_read(&EXT4_I(inode)->i_data_sem); + if (ext4_has_inline_data(inode)) { + /* + * We will soon create blocks for this page, and let + * us pretend as if the blocks aren't allocated yet. + * In case of clusters, we have to handle the work + * of mapping from cluster so that the reserved space + * is calculated properly. + */ + if ((EXT4_SB(inode->i_sb)->s_cluster_ratio > 1) && + ext4_find_delalloc_cluster(inode, map->m_lblk)) + map->m_flags |= EXT4_MAP_FROM_CLUSTER; + retval = 0; + } else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + retval = ext4_ext_map_blocks(NULL, inode, map, + EXT4_GET_BLOCKS_NO_PUT_HOLE); + else + retval = ext4_ind_map_blocks(NULL, inode, map, + EXT4_GET_BLOCKS_NO_PUT_HOLE); + +add_delayed: + if (retval == 0) { + int ret; + /* + * XXX: __block_prepare_write() unmaps passed block, + * is it OK? + */ + /* + * If the block was allocated from previously allocated cluster, + * then we don't need to reserve it again. However we still need + * to reserve metadata for every block we're going to write. + */ + if (!(map->m_flags & EXT4_MAP_FROM_CLUSTER)) { + ret = ext4_da_reserve_space(inode, iblock); + if (ret) { + /* not enough space to reserve */ + retval = ret; + goto out_unlock; + } + } else { + ret = ext4_da_reserve_metadata(inode, iblock); + if (ret) { + /* not enough space to reserve */ + retval = ret; + goto out_unlock; + } + } + + ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + ~0, EXTENT_STATUS_DELAYED); + if (ret) { + retval = ret; + goto out_unlock; + } + + /* Clear EXT4_MAP_FROM_CLUSTER flag since its purpose is served + * and it should not appear on the bh->b_state. + */ + map->m_flags &= ~EXT4_MAP_FROM_CLUSTER; + + map_bh(bh, inode->i_sb, invalid_block); + set_buffer_new(bh); + set_buffer_delay(bh); + } else if (retval > 0) { + int ret; + unsigned int status; + + if (unlikely(retval != map->m_len)) { + ext4_warning(inode->i_sb, + "ES len assertion failed for inode " + "%lu: retval %d != map->m_len %d", + inode->i_ino, retval, map->m_len); + WARN_ON(1); + } + + status = map->m_flags & EXT4_MAP_UNWRITTEN ? + EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; + ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + map->m_pblk, status); + if (ret != 0) + retval = ret; + } + +out_unlock: + up_read((&EXT4_I(inode)->i_data_sem)); + + return retval; +} + +/* + * This is a special get_blocks_t callback which is used by + * ext4_da_write_begin(). It will either return mapped block or + * reserve space for a single block. + * + * For delayed buffer_head we have BH_Mapped, BH_New, BH_Delay set. + * We also have b_blocknr = -1 and b_bdev initialized properly + * + * For unwritten buffer_head we have BH_Mapped, BH_New, BH_Unwritten set. + * We also have b_blocknr = physicalblock mapping unwritten extent and b_bdev + * initialized properly. + */ +int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int create) +{ + struct ext4_map_blocks map; + int ret = 0; + + BUG_ON(create == 0); + BUG_ON(bh->b_size != inode->i_sb->s_blocksize); + + map.m_lblk = iblock; + map.m_len = 1; + + /* + * first, we need to know whether the block is allocated already + * preallocated blocks are unmapped but should treated + * the same as allocated blocks. + */ + ret = ext4_da_map_blocks(inode, iblock, &map, bh); + if (ret <= 0) + return ret; + + map_bh(bh, inode->i_sb, map.m_pblk); + bh->b_state = (bh->b_state & ~EXT4_MAP_FLAGS) | map.m_flags; + + if (buffer_unwritten(bh)) { + /* A delayed write to unwritten bh should be marked + * new and mapped. Mapped ensures that we don't do + * get_block multiple times when we write to the same + * offset and new ensures that we do proper zero out + * for partial write. + */ + set_buffer_new(bh); + set_buffer_mapped(bh); + } + return 0; } static int bget_one(handle_t *handle, struct buffer_head *bh) @@ -1450,338 +1731,1527 @@ static int bput_one(handle_t *handle, struct buffer_head *bh) return 0; } -static int jbd2_journal_dirty_data_fn(handle_t *handle, struct buffer_head *bh) +static int __ext4_journalled_writepage(struct page *page, + unsigned int len) { - if (buffer_mapped(bh)) - return ext4_journal_dirty_data(handle, bh); - return 0; + struct address_space *mapping = page->mapping; + struct inode *inode = mapping->host; + struct buffer_head *page_bufs = NULL; + handle_t *handle = NULL; + int ret = 0, err = 0; + int inline_data = ext4_has_inline_data(inode); + struct buffer_head *inode_bh = NULL; + + ClearPageChecked(page); + + if (inline_data) { + BUG_ON(page->index != 0); + BUG_ON(len > ext4_get_max_inline_size(inode)); + inode_bh = ext4_journalled_write_inline_data(inode, len, page); + if (inode_bh == NULL) + goto out; + } else { + page_bufs = page_buffers(page); + if (!page_bufs) { + BUG(); + goto out; + } + ext4_walk_page_buffers(handle, page_bufs, 0, len, + NULL, bget_one); + } + /* As soon as we unlock the page, it can go away, but we have + * references to buffers so we are safe */ + unlock_page(page); + + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, + ext4_writepage_trans_blocks(inode)); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out; + } + + BUG_ON(!ext4_handle_valid(handle)); + + if (inline_data) { + BUFFER_TRACE(inode_bh, "get write access"); + ret = ext4_journal_get_write_access(handle, inode_bh); + + err = ext4_handle_dirty_metadata(handle, inode, inode_bh); + + } else { + ret = ext4_walk_page_buffers(handle, page_bufs, 0, len, NULL, + do_journal_get_write_access); + + err = ext4_walk_page_buffers(handle, page_bufs, 0, len, NULL, + 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; + + if (!ext4_has_inline_data(inode)) + ext4_walk_page_buffers(NULL, page_bufs, 0, len, + NULL, bput_one); + ext4_set_inode_state(inode, EXT4_STATE_JDATA); +out: + brelse(inode_bh); + return ret; } /* - * Note that we always start a transaction even if we're not journalling - * data. This is to preserve ordering: any hole instantiation within - * __block_write_full_page -> ext4_get_block() should be journalled - * along with the data so we don't crash and then get metadata which - * refers to old data. + * Note that we don't need to start a transaction unless we're journaling data + * because we should have holes filled from ext4_page_mkwrite(). We even don't + * need to file the inode to the transaction's list in ordered mode because if + * we are writing back data added by write(), the inode is already there and if + * we are writing back data modified via mmap(), no one guarantees in which + * transaction the data will hit the disk. In case we are journaling data, we + * cannot start transaction directly because transaction start ranks above page + * lock so we have to do some magic. + * + * This function can get called via... + * - ext4_writepages after taking page lock (have journal handle) + * - journal_submit_inode_data_buffers (no journal handle) + * - shrink_page_list via the kswapd/direct reclaim (no journal handle) + * - grab_page_cache when doing write_begin (have journal handle) * - * In all journalling modes block_write_full_page() will start the I/O. + * We don't do any block allocation in this function. If we have page with + * multiple blocks we need to write those buffer_heads that are mapped. This + * is important for mmaped based write. So if we do with blocksize 1K + * truncate(f, 1024); + * a = mmap(f, 0, 4096); + * a[0] = 'a'; + * truncate(f, 4096); + * we have in the page first buffer_head mapped via page_mkwrite call back + * but other buffer_heads would be unmapped but dirty (dirty done via the + * do_wp_page). So writepage should write the first block. If we modify + * the mmap area beyond 1024 we will again get a page_fault and the + * page_mkwrite callback will do the block allocation and mark the + * buffer_heads mapped. * - * Problem: + * We redirty the page if we have any buffer_heads that is either delay or + * unwritten in the page. + * + * We can get recursively called as show below. * * ext4_writepage() -> kmalloc() -> __alloc_pages() -> page_launder() -> * ext4_writepage() * - * Similar for: + * But since we don't do any block allocation we should not deadlock. + * Page also have the dirty flag cleared so we don't get recurive page_lock. + */ +static int ext4_writepage(struct page *page, + struct writeback_control *wbc) +{ + int ret = 0; + loff_t size; + unsigned int len; + struct buffer_head *page_bufs = NULL; + struct inode *inode = page->mapping->host; + struct ext4_io_submit io_submit; + bool keep_towrite = false; + + trace_ext4_writepage(page); + size = i_size_read(inode); + if (page->index == size >> PAGE_CACHE_SHIFT) + len = size & ~PAGE_CACHE_MASK; + else + len = PAGE_CACHE_SIZE; + + page_bufs = page_buffers(page); + /* + * We cannot do block allocation or other extent handling in this + * function. If there are buffers needing that, we have to redirty + * the page. But we may reach here when we do a journal commit via + * journal_submit_inode_data_buffers() and in that case we must write + * allocated buffers to achieve data=ordered mode guarantees. + */ + if (ext4_walk_page_buffers(NULL, page_bufs, 0, len, NULL, + ext4_bh_delay_or_unwritten)) { + redirty_page_for_writepage(wbc, page); + if (current->flags & PF_MEMALLOC) { + /* + * For memory cleaning there's no point in writing only + * some buffers. So just bail out. Warn if we came here + * from direct reclaim. + */ + WARN_ON_ONCE((current->flags & (PF_MEMALLOC|PF_KSWAPD)) + == PF_MEMALLOC); + unlock_page(page); + return 0; + } + keep_towrite = true; + } + + if (PageChecked(page) && ext4_should_journal_data(inode)) + /* + * It's mmapped pagecache. Add buffers and journal it. There + * doesn't seem much point in redirtying the page here. + */ + return __ext4_journalled_writepage(page, len); + + ext4_io_submit_init(&io_submit, wbc); + io_submit.io_end = ext4_init_io_end(inode, GFP_NOFS); + if (!io_submit.io_end) { + redirty_page_for_writepage(wbc, page); + unlock_page(page); + return -ENOMEM; + } + ret = ext4_bio_write_page(&io_submit, page, len, wbc, keep_towrite); + ext4_io_submit(&io_submit); + /* Drop io_end reference we got from init */ + ext4_put_io_end_defer(io_submit.io_end); + return ret; +} + +static int mpage_submit_page(struct mpage_da_data *mpd, struct page *page) +{ + int len; + loff_t size = i_size_read(mpd->inode); + int err; + + BUG_ON(page->index != mpd->first_page); + if (page->index == size >> PAGE_CACHE_SHIFT) + len = size & ~PAGE_CACHE_MASK; + else + len = PAGE_CACHE_SIZE; + clear_page_dirty_for_io(page); + err = ext4_bio_write_page(&mpd->io_submit, page, len, mpd->wbc, false); + if (!err) + mpd->wbc->nr_to_write--; + mpd->first_page++; + + return err; +} + +#define BH_FLAGS ((1 << BH_Unwritten) | (1 << BH_Delay)) + +/* + * mballoc gives us at most this number of blocks... + * XXX: That seems to be only a limitation of ext4_mb_normalize_request(). + * The rest of mballoc seems to handle chunks up to full group size. + */ +#define MAX_WRITEPAGES_EXTENT_LEN 2048 + +/* + * mpage_add_bh_to_extent - try to add bh to extent of blocks to map + * + * @mpd - extent of blocks + * @lblk - logical number of the block in the file + * @bh - buffer head we want to add to the extent * - * ext4_file_write() -> generic_file_write() -> __alloc_pages() -> ... + * The function is used to collect contig. blocks in the same state. If the + * buffer doesn't require mapping for writeback and we haven't started the + * extent of buffers to map yet, the function returns 'true' immediately - the + * caller can write the buffer right away. Otherwise the function returns true + * if the block has been added to the extent, false if the block couldn't be + * added. + */ +static bool mpage_add_bh_to_extent(struct mpage_da_data *mpd, ext4_lblk_t lblk, + struct buffer_head *bh) +{ + struct ext4_map_blocks *map = &mpd->map; + + /* Buffer that doesn't need mapping for writeback? */ + if (!buffer_dirty(bh) || !buffer_mapped(bh) || + (!buffer_delay(bh) && !buffer_unwritten(bh))) { + /* So far no extent to map => we write the buffer right away */ + if (map->m_len == 0) + return true; + return false; + } + + /* First block in the extent? */ + if (map->m_len == 0) { + map->m_lblk = lblk; + map->m_len = 1; + map->m_flags = bh->b_state & BH_FLAGS; + return true; + } + + /* Don't go larger than mballoc is willing to allocate */ + if (map->m_len >= MAX_WRITEPAGES_EXTENT_LEN) + return false; + + /* Can we merge the block to our big extent? */ + if (lblk == map->m_lblk + map->m_len && + (bh->b_state & BH_FLAGS) == map->m_flags) { + map->m_len++; + return true; + } + return false; +} + +/* + * mpage_process_page_bufs - submit page buffers for IO or add them to extent * - * Same applies to ext4_get_block(). We will deadlock on various things like - * lock_journal and i_data_sem + * @mpd - extent of blocks for mapping + * @head - the first buffer in the page + * @bh - buffer we should start processing from + * @lblk - logical number of the block in the file corresponding to @bh * - * Setting PF_MEMALLOC here doesn't work - too many internal memory - * allocations fail. + * Walk through page buffers from @bh upto @head (exclusive) and either submit + * the page for IO if all buffers in this page were mapped and there's no + * accumulated extent of buffers to map or add buffers in the page to the + * extent of buffers to map. The function returns 1 if the caller can continue + * by processing the next page, 0 if it should stop adding buffers to the + * extent to map because we cannot extend it anymore. It can also return value + * < 0 in case of error during IO submission. + */ +static int mpage_process_page_bufs(struct mpage_da_data *mpd, + struct buffer_head *head, + struct buffer_head *bh, + ext4_lblk_t lblk) +{ + struct inode *inode = mpd->inode; + int err; + ext4_lblk_t blocks = (i_size_read(inode) + (1 << inode->i_blkbits) - 1) + >> inode->i_blkbits; + + do { + BUG_ON(buffer_locked(bh)); + + if (lblk >= blocks || !mpage_add_bh_to_extent(mpd, lblk, bh)) { + /* Found extent to map? */ + if (mpd->map.m_len) + return 0; + /* Everything mapped so far and we hit EOF */ + break; + } + } while (lblk++, (bh = bh->b_this_page) != head); + /* So far everything mapped? Submit the page for IO. */ + if (mpd->map.m_len == 0) { + err = mpage_submit_page(mpd, head->b_page); + if (err < 0) + return err; + } + return lblk < blocks; +} + +/* + * mpage_map_buffers - update buffers corresponding to changed extent and + * submit fully mapped pages for IO * - * 16May01: If we're reentered then journal_current_handle() will be - * non-zero. We simply *return*. + * @mpd - description of extent to map, on return next extent to map * - * 1 July 2001: @@@ FIXME: - * In journalled data mode, a data buffer may be metadata against the - * current transaction. But the same file is part of a shared mapping - * and someone does a writepage() on it. + * Scan buffers corresponding to changed extent (we expect corresponding pages + * to be already locked) and update buffer state according to new extent state. + * We map delalloc buffers to their physical location, clear unwritten bits, + * and mark buffers as uninit when we perform writes to unwritten extents + * and do extent conversion after IO is finished. If the last page is not fully + * mapped, we update @map to the next extent in the last page that needs + * mapping. Otherwise we submit the page for IO. + */ +static int mpage_map_and_submit_buffers(struct mpage_da_data *mpd) +{ + struct pagevec pvec; + int nr_pages, i; + struct inode *inode = mpd->inode; + struct buffer_head *head, *bh; + int bpp_bits = PAGE_CACHE_SHIFT - inode->i_blkbits; + pgoff_t start, end; + ext4_lblk_t lblk; + sector_t pblock; + int err; + + start = mpd->map.m_lblk >> bpp_bits; + end = (mpd->map.m_lblk + mpd->map.m_len - 1) >> bpp_bits; + lblk = start << bpp_bits; + pblock = mpd->map.m_pblk; + + pagevec_init(&pvec, 0); + while (start <= end) { + nr_pages = pagevec_lookup(&pvec, inode->i_mapping, start, + PAGEVEC_SIZE); + if (nr_pages == 0) + break; + for (i = 0; i < nr_pages; i++) { + struct page *page = pvec.pages[i]; + + if (page->index > end) + break; + /* Up to 'end' pages must be contiguous */ + BUG_ON(page->index != start); + bh = head = page_buffers(page); + do { + if (lblk < mpd->map.m_lblk) + continue; + if (lblk >= mpd->map.m_lblk + mpd->map.m_len) { + /* + * Buffer after end of mapped extent. + * Find next buffer in the page to map. + */ + mpd->map.m_len = 0; + mpd->map.m_flags = 0; + /* + * FIXME: If dioread_nolock supports + * blocksize < pagesize, we need to make + * sure we add size mapped so far to + * io_end->size as the following call + * can submit the page for IO. + */ + err = mpage_process_page_bufs(mpd, head, + bh, lblk); + pagevec_release(&pvec); + if (err > 0) + err = 0; + return err; + } + if (buffer_delay(bh)) { + clear_buffer_delay(bh); + bh->b_blocknr = pblock++; + } + clear_buffer_unwritten(bh); + } while (lblk++, (bh = bh->b_this_page) != head); + + /* + * FIXME: This is going to break if dioread_nolock + * supports blocksize < pagesize as we will try to + * convert potentially unmapped parts of inode. + */ + mpd->io_submit.io_end->size += PAGE_CACHE_SIZE; + /* Page fully mapped - let IO run! */ + err = mpage_submit_page(mpd, page); + if (err < 0) { + pagevec_release(&pvec); + return err; + } + start++; + } + pagevec_release(&pvec); + } + /* Extent fully mapped and matches with page boundary. We are done. */ + mpd->map.m_len = 0; + mpd->map.m_flags = 0; + return 0; +} + +static int mpage_map_one_extent(handle_t *handle, struct mpage_da_data *mpd) +{ + struct inode *inode = mpd->inode; + struct ext4_map_blocks *map = &mpd->map; + int get_blocks_flags; + int err, dioread_nolock; + + trace_ext4_da_write_pages_extent(inode, map); + /* + * Call ext4_map_blocks() to allocate any delayed allocation blocks, or + * to convert an unwritten extent to be initialized (in the case + * where we have written into one or more preallocated blocks). It is + * possible that we're going to need more metadata blocks than + * previously reserved. However we must not fail because we're in + * writeback and there is nothing we can do about it so it might result + * in data loss. So use reserved blocks to allocate metadata if + * possible. + * + * We pass in the magic EXT4_GET_BLOCKS_DELALLOC_RESERVE if the blocks + * in question are delalloc blocks. This affects functions in many + * different parts of the allocation call path. This flag exists + * primarily because we don't want to change *many* call functions, so + * ext4_map_blocks() will set the EXT4_STATE_DELALLOC_RESERVED flag + * once the inode's allocation semaphore is taken. + */ + get_blocks_flags = EXT4_GET_BLOCKS_CREATE | + EXT4_GET_BLOCKS_METADATA_NOFAIL; + dioread_nolock = ext4_should_dioread_nolock(inode); + if (dioread_nolock) + get_blocks_flags |= EXT4_GET_BLOCKS_IO_CREATE_EXT; + if (map->m_flags & (1 << BH_Delay)) + get_blocks_flags |= EXT4_GET_BLOCKS_DELALLOC_RESERVE; + + err = ext4_map_blocks(handle, inode, map, get_blocks_flags); + if (err < 0) + return err; + if (dioread_nolock && (map->m_flags & EXT4_MAP_UNWRITTEN)) { + if (!mpd->io_submit.io_end->handle && + ext4_handle_valid(handle)) { + mpd->io_submit.io_end->handle = handle->h_rsv_handle; + handle->h_rsv_handle = NULL; + } + ext4_set_io_unwritten_flag(inode, mpd->io_submit.io_end); + } + + BUG_ON(map->m_len == 0); + if (map->m_flags & EXT4_MAP_NEW) { + struct block_device *bdev = inode->i_sb->s_bdev; + int i; + + for (i = 0; i < map->m_len; i++) + unmap_underlying_metadata(bdev, map->m_pblk + i); + } + return 0; +} + +/* + * mpage_map_and_submit_extent - map extent starting at mpd->lblk of length + * mpd->len and submit pages underlying it for IO * - * We will move the buffer onto the async_data list, but *after* it has - * been dirtied. So there's a small window where we have dirty data on - * BJ_Metadata. + * @handle - handle for journal operations + * @mpd - extent to map + * @give_up_on_write - we set this to true iff there is a fatal error and there + * is no hope of writing the data. The caller should discard + * dirty pages to avoid infinite loops. * - * Note that this only applies to the last partial page in the file. The - * bit which block_write_full_page() uses prepare/commit for. (That's - * broken code anyway: it's wrong for msync()). + * The function maps extent starting at mpd->lblk of length mpd->len. If it is + * delayed, blocks are allocated, if it is unwritten, we may need to convert + * them to initialized or split the described range from larger unwritten + * extent. Note that we need not map all the described range since allocation + * can return less blocks or the range is covered by more unwritten extents. We + * cannot map more because we are limited by reserved transaction credits. On + * the other hand we always make sure that the last touched page is fully + * mapped so that it can be written out (and thus forward progress is + * guaranteed). After mapping we submit all mapped pages for IO. + */ +static int mpage_map_and_submit_extent(handle_t *handle, + struct mpage_da_data *mpd, + bool *give_up_on_write) +{ + struct inode *inode = mpd->inode; + struct ext4_map_blocks *map = &mpd->map; + int err; + loff_t disksize; + + mpd->io_submit.io_end->offset = + ((loff_t)map->m_lblk) << inode->i_blkbits; + do { + err = mpage_map_one_extent(handle, mpd); + if (err < 0) { + struct super_block *sb = inode->i_sb; + + if (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED) + goto invalidate_dirty_pages; + /* + * Let the uper layers retry transient errors. + * In the case of ENOSPC, if ext4_count_free_blocks() + * is non-zero, a commit should free up blocks. + */ + if ((err == -ENOMEM) || + (err == -ENOSPC && ext4_count_free_clusters(sb))) + return err; + ext4_msg(sb, KERN_CRIT, + "Delayed block allocation failed for " + "inode %lu at logical offset %llu with" + " max blocks %u with error %d", + inode->i_ino, + (unsigned long long)map->m_lblk, + (unsigned)map->m_len, -err); + ext4_msg(sb, KERN_CRIT, + "This should not happen!! Data will " + "be lost\n"); + if (err == -ENOSPC) + ext4_print_free_blocks(inode); + invalidate_dirty_pages: + *give_up_on_write = true; + return err; + } + /* + * Update buffer state, submit mapped pages, and get us new + * extent to map + */ + err = mpage_map_and_submit_buffers(mpd); + if (err < 0) + return err; + } while (map->m_len); + + /* + * Update on-disk size after IO is submitted. Races with + * truncate are avoided by checking i_size under i_data_sem. + */ + disksize = ((loff_t)mpd->first_page) << PAGE_CACHE_SHIFT; + if (disksize > EXT4_I(inode)->i_disksize) { + int err2; + loff_t i_size; + + down_write(&EXT4_I(inode)->i_data_sem); + i_size = i_size_read(inode); + if (disksize > i_size) + disksize = i_size; + if (disksize > EXT4_I(inode)->i_disksize) + EXT4_I(inode)->i_disksize = disksize; + err2 = ext4_mark_inode_dirty(handle, inode); + up_write(&EXT4_I(inode)->i_data_sem); + if (err2) + ext4_error(inode->i_sb, + "Failed to mark inode %lu dirty", + inode->i_ino); + if (!err) + err = err2; + } + return err; +} + +/* + * Calculate the total number of credits to reserve for one writepages + * iteration. This is called from ext4_writepages(). We map an extent of + * up to MAX_WRITEPAGES_EXTENT_LEN blocks and then we go on and finish mapping + * the last partial page. So in total we can map MAX_WRITEPAGES_EXTENT_LEN + + * bpp - 1 blocks in bpp different extents. + */ +static int ext4_da_writepages_trans_blocks(struct inode *inode) +{ + int bpp = ext4_journal_blocks_per_page(inode); + + return ext4_meta_trans_blocks(inode, + MAX_WRITEPAGES_EXTENT_LEN + bpp - 1, bpp); +} + +/* + * mpage_prepare_extent_to_map - find & lock contiguous range of dirty pages + * and underlying extent to map * - * It's a rare case: affects the final partial page, for journalled data - * where the file is subject to bith write() and writepage() in the same - * transction. To fix it we'll need a custom block_write_full_page(). - * We'll probably need that anyway for journalling writepage() output. + * @mpd - where to look for pages * - * We don't honour synchronous mounts for writepage(). That would be - * disastrous. Any write() or metadata operation will sync the fs for - * us. + * Walk dirty pages in the mapping. If they are fully mapped, submit them for + * IO immediately. When we find a page which isn't mapped we start accumulating + * extent of buffers underlying these pages that needs mapping (formed by + * either delayed or unwritten buffers). We also lock the pages containing + * these buffers. The extent found is returned in @mpd structure (starting at + * mpd->lblk with length mpd->len blocks). * - * AKPM2: if all the page's buffers are mapped to disk and !data=journal, - * we don't need to open a transaction here. + * Note that this function can attach bios to one io_end structure which are + * neither logically nor physically contiguous. Although it may seem as an + * unnecessary complication, it is actually inevitable in blocksize < pagesize + * case as we need to track IO to all buffers underlying a page in one io_end. */ -static int ext4_ordered_writepage(struct page *page, - struct writeback_control *wbc) +static int mpage_prepare_extent_to_map(struct mpage_da_data *mpd) { - struct inode *inode = page->mapping->host; - struct buffer_head *page_bufs; + struct address_space *mapping = mpd->inode->i_mapping; + struct pagevec pvec; + unsigned int nr_pages; + long left = mpd->wbc->nr_to_write; + pgoff_t index = mpd->first_page; + pgoff_t end = mpd->last_page; + int tag; + int i, err = 0; + int blkbits = mpd->inode->i_blkbits; + ext4_lblk_t lblk; + struct buffer_head *head; + + if (mpd->wbc->sync_mode == WB_SYNC_ALL || mpd->wbc->tagged_writepages) + tag = PAGECACHE_TAG_TOWRITE; + else + tag = PAGECACHE_TAG_DIRTY; + + pagevec_init(&pvec, 0); + mpd->map.m_len = 0; + mpd->next_page = index; + while (index <= end) { + nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, tag, + min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1); + if (nr_pages == 0) + goto out; + + for (i = 0; i < nr_pages; i++) { + struct page *page = pvec.pages[i]; + + /* + * At this point, the page may be truncated or + * invalidated (changing page->mapping to NULL), or + * even swizzled back from swapper_space to tmpfs file + * mapping. However, page->index will not change + * because we have a reference on the page. + */ + if (page->index > end) + goto out; + + /* + * Accumulated enough dirty pages? This doesn't apply + * to WB_SYNC_ALL mode. For integrity sync we have to + * keep going because someone may be concurrently + * dirtying pages, and we might have synced a lot of + * newly appeared dirty pages, but have not synced all + * of the old dirty pages. + */ + if (mpd->wbc->sync_mode == WB_SYNC_NONE && left <= 0) + goto out; + + /* If we can't merge this page, we are done. */ + if (mpd->map.m_len > 0 && mpd->next_page != page->index) + goto out; + + lock_page(page); + /* + * If the page is no longer dirty, or its mapping no + * longer corresponds to inode we are writing (which + * means it has been truncated or invalidated), or the + * page is already under writeback and we are not doing + * a data integrity writeback, skip the page + */ + if (!PageDirty(page) || + (PageWriteback(page) && + (mpd->wbc->sync_mode == WB_SYNC_NONE)) || + unlikely(page->mapping != mapping)) { + unlock_page(page); + continue; + } + + wait_on_page_writeback(page); + BUG_ON(PageWriteback(page)); + + if (mpd->map.m_len == 0) + mpd->first_page = page->index; + mpd->next_page = page->index + 1; + /* Add all dirty buffers to mpd */ + lblk = ((ext4_lblk_t)page->index) << + (PAGE_CACHE_SHIFT - blkbits); + head = page_buffers(page); + err = mpage_process_page_bufs(mpd, head, head, lblk); + if (err <= 0) + goto out; + err = 0; + left--; + } + pagevec_release(&pvec); + cond_resched(); + } + return 0; +out: + pagevec_release(&pvec); + return err; +} + +static int __writepage(struct page *page, struct writeback_control *wbc, + void *data) +{ + struct address_space *mapping = data; + int ret = ext4_writepage(page, wbc); + mapping_set_error(mapping, ret); + return ret; +} + +static int ext4_writepages(struct address_space *mapping, + struct writeback_control *wbc) +{ + pgoff_t writeback_index = 0; + long nr_to_write = wbc->nr_to_write; + int range_whole = 0; + int cycled = 1; handle_t *handle = NULL; - int ret = 0; - int err; + struct mpage_da_data mpd; + struct inode *inode = mapping->host; + int needed_blocks, rsv_blocks = 0, ret = 0; + struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); + bool done; + struct blk_plug plug; + bool give_up_on_write = false; - J_ASSERT(PageLocked(page)); + trace_ext4_writepages(inode, wbc); /* - * We give up here if we're reentered, because it might be for a - * different filesystem. + * No pages to write? This is mainly a kludge to avoid starting + * a transaction for special inodes like journal inode on last iput() + * because that could violate lock ordering on umount */ - if (ext4_journal_current_handle()) - goto out_fail; + if (!mapping->nrpages || !mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) + goto out_writepages; - handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); + if (ext4_should_journal_data(inode)) { + struct blk_plug plug; - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto out_fail; + blk_start_plug(&plug); + ret = write_cache_pages(mapping, wbc, __writepage, mapping); + blk_finish_plug(&plug); + goto out_writepages; } - if (!page_has_buffers(page)) { - create_empty_buffers(page, inode->i_sb->s_blocksize, - (1 << BH_Dirty)|(1 << BH_Uptodate)); + /* + * If the filesystem has aborted, it is read-only, so return + * right away instead of dumping stack traces later on that + * will obscure the real source of the problem. We test + * EXT4_MF_FS_ABORTED instead of sb->s_flag's MS_RDONLY because + * the latter could be true if the filesystem is mounted + * read-only, and in that case, ext4_writepages should + * *never* be called, so if that ever happens, we would want + * the stack trace. + */ + if (unlikely(sbi->s_mount_flags & EXT4_MF_FS_ABORTED)) { + ret = -EROFS; + goto out_writepages; } - page_bufs = page_buffers(page); - walk_page_buffers(handle, page_bufs, 0, - PAGE_CACHE_SIZE, NULL, bget_one); - ret = block_write_full_page(page, ext4_get_block, wbc); + if (ext4_should_dioread_nolock(inode)) { + /* + * We may need to convert up to one extent per block in + * the page and we may dirty the inode. + */ + rsv_blocks = 1 + (PAGE_CACHE_SIZE >> inode->i_blkbits); + } /* - * The page can become unlocked at any point now, and - * truncate can then come in and change things. So we - * can't touch *page from now on. But *page_bufs is - * safe due to elevated refcount. + * If we have inline data and arrive here, it means that + * we will soon create the block for the 1st page, so + * we'd better clear the inline data here. */ + if (ext4_has_inline_data(inode)) { + /* Just inode will be modified... */ + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out_writepages; + } + BUG_ON(ext4_test_inode_state(inode, + EXT4_STATE_MAY_INLINE_DATA)); + ext4_destroy_inline_data(handle, inode); + ext4_journal_stop(handle); + } + + if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX) + range_whole = 1; + + if (wbc->range_cyclic) { + writeback_index = mapping->writeback_index; + if (writeback_index) + cycled = 0; + mpd.first_page = writeback_index; + mpd.last_page = -1; + } else { + mpd.first_page = wbc->range_start >> PAGE_CACHE_SHIFT; + mpd.last_page = wbc->range_end >> PAGE_CACHE_SHIFT; + } + + mpd.inode = inode; + mpd.wbc = wbc; + ext4_io_submit_init(&mpd.io_submit, wbc); +retry: + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_writepages) + tag_pages_for_writeback(mapping, mpd.first_page, mpd.last_page); + done = false; + blk_start_plug(&plug); + while (!done && mpd.first_page <= mpd.last_page) { + /* For each extent of pages we use new io_end */ + mpd.io_submit.io_end = ext4_init_io_end(inode, GFP_KERNEL); + if (!mpd.io_submit.io_end) { + ret = -ENOMEM; + break; + } + + /* + * We have two constraints: We find one extent to map and we + * must always write out whole page (makes a difference when + * blocksize < pagesize) so that we don't block on IO when we + * try to write out the rest of the page. Journalled mode is + * not supported by delalloc. + */ + BUG_ON(ext4_should_journal_data(inode)); + needed_blocks = ext4_da_writepages_trans_blocks(inode); + + /* start a new transaction */ + handle = ext4_journal_start_with_reserve(inode, + EXT4_HT_WRITE_PAGE, needed_blocks, rsv_blocks); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + ext4_msg(inode->i_sb, KERN_CRIT, "%s: jbd2_start: " + "%ld pages, ino %lu; err %d", __func__, + wbc->nr_to_write, inode->i_ino, ret); + /* Release allocated io_end */ + ext4_put_io_end(mpd.io_submit.io_end); + break; + } + + trace_ext4_da_write_pages(inode, mpd.first_page, mpd.wbc); + ret = mpage_prepare_extent_to_map(&mpd); + if (!ret) { + if (mpd.map.m_len) + ret = mpage_map_and_submit_extent(handle, &mpd, + &give_up_on_write); + else { + /* + * We scanned the whole range (or exhausted + * nr_to_write), submitted what was mapped and + * didn't find anything needing mapping. We are + * done. + */ + done = true; + } + } + ext4_journal_stop(handle); + /* Submit prepared bio */ + ext4_io_submit(&mpd.io_submit); + /* Unlock pages we didn't use */ + mpage_release_unused_pages(&mpd, give_up_on_write); + /* Drop our io_end reference we got from init */ + ext4_put_io_end(mpd.io_submit.io_end); + + if (ret == -ENOSPC && sbi->s_journal) { + /* + * Commit the transaction which would + * free blocks released in the transaction + * and try again + */ + jbd2_journal_force_commit_nested(sbi->s_journal); + ret = 0; + continue; + } + /* Fatal error - ENOMEM, EIO... */ + if (ret) + break; + } + blk_finish_plug(&plug); + if (!ret && !cycled && wbc->nr_to_write > 0) { + cycled = 1; + mpd.last_page = writeback_index - 1; + mpd.first_page = 0; + goto retry; + } + + /* Update index */ + if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0)) + /* + * Set the writeback_index so that range_cyclic + * mode will write it back later + */ + mapping->writeback_index = mpd.first_page; + +out_writepages: + trace_ext4_writepages_result(inode, wbc, ret, + nr_to_write - wbc->nr_to_write); + return ret; +} + +static int ext4_nonda_switch(struct super_block *sb) +{ + s64 free_clusters, dirty_clusters; + struct ext4_sb_info *sbi = EXT4_SB(sb); /* - * And attach them to the current transaction. But only if - * block_write_full_page() succeeded. Otherwise they are unmapped, - * and generally junk. + * switch to non delalloc mode if we are running low + * on free block. The free block accounting via percpu + * counters can get slightly wrong with percpu_counter_batch getting + * accumulated on each CPU without updating global counters + * Delalloc need an accurate free block accounting. So switch + * to non delalloc when we are near to error range. */ - if (ret == 0) { - err = walk_page_buffers(handle, page_bufs, 0, PAGE_CACHE_SIZE, - NULL, jbd2_journal_dirty_data_fn); - if (!ret) - ret = err; + free_clusters = + percpu_counter_read_positive(&sbi->s_freeclusters_counter); + dirty_clusters = + percpu_counter_read_positive(&sbi->s_dirtyclusters_counter); + /* + * Start pushing delalloc when 1/2 of free blocks are dirty. + */ + if (dirty_clusters && (free_clusters < 2 * dirty_clusters)) + try_to_writeback_inodes_sb(sb, WB_REASON_FS_FREE_SPACE); + + if (2 * free_clusters < 3 * dirty_clusters || + free_clusters < (dirty_clusters + EXT4_FREECLUSTERS_WATERMARK)) { + /* + * free block count is less than 150% of dirty blocks + * or free blocks is less than watermark + */ + return 1; } - walk_page_buffers(handle, page_bufs, 0, - PAGE_CACHE_SIZE, NULL, bput_one); - err = ext4_journal_stop(handle); - if (!ret) - ret = err; - return ret; + return 0; +} -out_fail: - redirty_page_for_writepage(wbc, page); +static int ext4_da_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) +{ + int ret, retries = 0; + struct page *page; + pgoff_t index; + struct inode *inode = mapping->host; + handle_t *handle; + + index = pos >> PAGE_CACHE_SHIFT; + + if (ext4_nonda_switch(inode->i_sb)) { + *fsdata = (void *)FALL_BACK_TO_NONDELALLOC; + return ext4_write_begin(file, mapping, pos, + len, flags, pagep, fsdata); + } + *fsdata = (void *)0; + trace_ext4_da_write_begin(inode, pos, len, flags); + + if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) { + ret = ext4_da_write_inline_data_begin(mapping, inode, + pos, len, flags, + pagep, fsdata); + if (ret < 0) + return ret; + if (ret == 1) + return 0; + } + + /* + * grab_cache_page_write_begin() can take a long time if the + * system is thrashing due to memory pressure, or if the page + * is being written back. So grab it first before we start + * the transaction handle. This also allows us to allocate + * the page (if needed) without using GFP_NOFS. + */ +retry_grab: + page = grab_cache_page_write_begin(mapping, index, flags); + if (!page) + return -ENOMEM; unlock_page(page); + + /* + * With delayed allocation, we don't log the i_disksize update + * if there is delayed block allocation. But we still need + * to journalling the i_disksize update if writes to the end + * of file which has an already mapped buffer. + */ +retry_journal: + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, 1); + if (IS_ERR(handle)) { + page_cache_release(page); + return PTR_ERR(handle); + } + + lock_page(page); + if (page->mapping != mapping) { + /* The page got truncated from under us */ + unlock_page(page); + page_cache_release(page); + ext4_journal_stop(handle); + goto retry_grab; + } + /* In case writeback began while the page was unlocked */ + wait_for_stable_page(page); + + ret = __block_write_begin(page, pos, len, ext4_da_get_block_prep); + if (ret < 0) { + unlock_page(page); + ext4_journal_stop(handle); + /* + * block_write_begin may have instantiated a few blocks + * outside i_size. Trim these off again. Don't need + * i_size_read because we hold i_mutex. + */ + if (pos + len > inode->i_size) + ext4_truncate_failed_write(inode); + + if (ret == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry_journal; + + page_cache_release(page); + return ret; + } + + *pagep = page; return ret; } -static int ext4_writeback_writepage(struct page *page, - struct writeback_control *wbc) +/* + * Check if we should update i_disksize + * when write to the end of file but not require block allocation + */ +static int ext4_da_should_update_i_disksize(struct page *page, + unsigned long offset) { + struct buffer_head *bh; struct inode *inode = page->mapping->host; - handle_t *handle = NULL; - int ret = 0; - int err; + unsigned int idx; + int i; - if (ext4_journal_current_handle()) - goto out_fail; + bh = page_buffers(page); + idx = offset >> inode->i_blkbits; - handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto out_fail; + for (i = 0; i < idx; i++) + bh = bh->b_this_page; + + if (!buffer_mapped(bh) || (buffer_delay(bh)) || buffer_unwritten(bh)) + return 0; + return 1; +} + +static int ext4_da_write_end(struct file *file, + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) +{ + struct inode *inode = mapping->host; + int ret = 0, ret2; + handle_t *handle = ext4_journal_current_handle(); + loff_t new_i_size; + unsigned long start, end; + int write_mode = (int)(unsigned long)fsdata; + + if (write_mode == FALL_BACK_TO_NONDELALLOC) + return ext4_write_end(file, mapping, pos, + len, copied, page, fsdata); + + trace_ext4_da_write_end(inode, pos, len, copied); + start = pos & (PAGE_CACHE_SIZE - 1); + end = start + copied - 1; + + /* + * generic_write_end() will run mark_inode_dirty() if i_size + * changes. So let's piggyback the i_disksize mark_inode_dirty + * into that. + */ + new_i_size = pos + copied; + if (copied && new_i_size > EXT4_I(inode)->i_disksize) { + if (ext4_has_inline_data(inode) || + ext4_da_should_update_i_disksize(page, end)) { + down_write(&EXT4_I(inode)->i_data_sem); + if (new_i_size > EXT4_I(inode)->i_disksize) + EXT4_I(inode)->i_disksize = new_i_size; + up_write(&EXT4_I(inode)->i_data_sem); + /* We need to mark inode dirty even if + * new_i_size is less that inode->i_size + * bu greater than i_disksize.(hint delalloc) + */ + ext4_mark_inode_dirty(handle, inode); + } } - if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode)) - ret = nobh_writepage(page, ext4_get_block, wbc); + if (write_mode != CONVERT_INLINE_DATA && + ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA) && + ext4_has_inline_data(inode)) + ret2 = ext4_da_write_inline_data_end(inode, pos, len, copied, + page); else - ret = block_write_full_page(page, ext4_get_block, wbc); + ret2 = generic_write_end(file, mapping, pos, len, copied, + page, fsdata); - err = ext4_journal_stop(handle); + copied = ret2; + if (ret2 < 0) + ret = ret2; + ret2 = ext4_journal_stop(handle); if (!ret) - ret = err; - return ret; + ret = ret2; -out_fail: - redirty_page_for_writepage(wbc, page); - unlock_page(page); - return ret; + return ret ? ret : copied; } -static int ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) +static void ext4_da_invalidatepage(struct page *page, unsigned int offset, + unsigned int length) { - struct inode *inode = page->mapping->host; - handle_t *handle = NULL; - int ret = 0; - int err; + /* + * Drop reserved blocks + */ + BUG_ON(!PageLocked(page)); + if (!page_has_buffers(page)) + goto out; - if (ext4_journal_current_handle()) - goto no_write; + ext4_da_page_release_reservation(page, offset, length); - handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto no_write; - } +out: + ext4_invalidatepage(page, offset, length); + + return; +} + +/* + * Force all delayed allocation blocks to be allocated for a given inode. + */ +int ext4_alloc_da_blocks(struct inode *inode) +{ + trace_ext4_alloc_da_blocks(inode); + + if (!EXT4_I(inode)->i_reserved_data_blocks && + !EXT4_I(inode)->i_reserved_meta_blocks) + return 0; + + /* + * We do something simple for now. The filemap_flush() will + * also start triggering a write of the data blocks, which is + * not strictly speaking necessary (and for users of + * laptop_mode, not even desirable). However, to do otherwise + * would require replicating code paths in: + * + * ext4_writepages() -> + * write_cache_pages() ---> (via passed in callback function) + * __mpage_da_writepage() --> + * mpage_add_bh_to_extent() + * mpage_da_map_blocks() + * + * The problem is that write_cache_pages(), located in + * mm/page-writeback.c, marks pages clean in preparation for + * doing I/O, which is not desirable if we're not planning on + * doing I/O at all. + * + * We could call write_cache_pages(), and then redirty all of + * the pages by calling redirty_page_for_writepage() but that + * would be ugly in the extreme. So instead we would need to + * replicate parts of the code in the above functions, + * simplifying them because we wouldn't actually intend to + * write out the pages, but rather only collect contiguous + * logical block extents, call the multi-block allocator, and + * then update the buffer heads with the block allocations. + * + * For now, though, we'll cheat by calling filemap_flush(), + * which will map the blocks, and start the I/O, but not + * actually wait for the I/O to complete. + */ + return filemap_flush(inode->i_mapping); +} - if (!page_has_buffers(page) || PageChecked(page)) { +/* + * bmap() is special. It gets used by applications such as lilo and by + * the swapper to find the on-disk block of a specific piece of data. + * + * Naturally, this is dangerous if the block concerned is still in the + * journal. If somebody makes a swapfile on an ext4 data-journaling + * filesystem and enables swap, then they may get a nasty shock when the + * data getting swapped to that swapfile suddenly gets overwritten by + * the original zero's written out previously to the journal and + * awaiting writeback in the kernel's buffer cache. + * + * So, if we see any bmap calls here on a modified, data-journaled file, + * take extra steps to flush any blocks which might be in the cache. + */ +static sector_t ext4_bmap(struct address_space *mapping, sector_t block) +{ + struct inode *inode = mapping->host; + journal_t *journal; + int err; + + /* + * We can get here for an inline file via the FIBMAP ioctl + */ + if (ext4_has_inline_data(inode)) + return 0; + + if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY) && + test_opt(inode->i_sb, DELALLOC)) { /* - * It's mmapped pagecache. Add buffers and journal it. There - * doesn't seem much point in redirtying the page here. + * With delalloc we want to sync the file + * so that we can make sure we allocate + * blocks for file */ - ClearPageChecked(page); - ret = block_prepare_write(page, 0, PAGE_CACHE_SIZE, - ext4_get_block); - if (ret != 0) { - ext4_journal_stop(handle); - goto out_unlock; - } - ret = walk_page_buffers(handle, page_buffers(page), 0, - PAGE_CACHE_SIZE, NULL, do_journal_get_write_access); + filemap_write_and_wait(mapping); + } - err = walk_page_buffers(handle, page_buffers(page), 0, - PAGE_CACHE_SIZE, NULL, write_end_fn); - if (ret == 0) - ret = err; - EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; - unlock_page(page); - } else { + if (EXT4_JOURNAL(inode) && + ext4_test_inode_state(inode, EXT4_STATE_JDATA)) { /* - * It may be a page full of checkpoint-mode buffers. We don't - * really know unless we go poke around in the buffer_heads. - * But block_write_full_page will do the right thing. + * This is a REALLY heavyweight approach, but the use of + * bmap on dirty files is expected to be extremely rare: + * only if we run lilo or swapon on a freshly made file + * do we expect this to happen. + * + * (bmap requires CAP_SYS_RAWIO so this does not + * represent an unprivileged user DOS attack --- we'd be + * in trouble if mortal users could trigger this path at + * will.) + * + * NB. EXT4_STATE_JDATA is not set on files other than + * regular files. If somebody wants to bmap a directory + * or symlink and gets confused because the buffer + * hasn't yet been flushed to disk, they deserve + * everything they get. */ - ret = block_write_full_page(page, ext4_get_block, wbc); + + ext4_clear_inode_state(inode, EXT4_STATE_JDATA); + journal = EXT4_JOURNAL(inode); + jbd2_journal_lock_updates(journal); + err = jbd2_journal_flush(journal); + jbd2_journal_unlock_updates(journal); + + if (err) + return 0; } - err = ext4_journal_stop(handle); - if (!ret) - ret = err; -out: - return ret; -no_write: - redirty_page_for_writepage(wbc, page); -out_unlock: - unlock_page(page); - goto out; + return generic_block_bmap(mapping, block, ext4_get_block); } static int ext4_readpage(struct file *file, struct page *page) { - return mpage_readpage(page, ext4_get_block); + int ret = -EAGAIN; + struct inode *inode = page->mapping->host; + + trace_ext4_readpage(page); + + if (ext4_has_inline_data(inode)) + ret = ext4_readpage_inline(inode, page); + + if (ret == -EAGAIN) + return mpage_readpage(page, ext4_get_block); + + return ret; } static int ext4_readpages(struct file *file, struct address_space *mapping, struct list_head *pages, unsigned nr_pages) { + struct inode *inode = mapping->host; + + /* If the file has inline data, no need to do readpages. */ + if (ext4_has_inline_data(inode)) + return 0; + return mpage_readpages(mapping, pages, nr_pages, ext4_get_block); } -static void ext4_invalidatepage(struct page *page, unsigned long offset) +static void ext4_invalidatepage(struct page *page, unsigned int offset, + unsigned int length) +{ + trace_ext4_invalidatepage(page, offset, length); + + /* No journalling happens on data buffers when this function is used */ + WARN_ON(page_has_buffers(page) && buffer_jbd(page_buffers(page))); + + block_invalidatepage(page, offset, length); +} + +static int __ext4_journalled_invalidatepage(struct page *page, + unsigned int offset, + unsigned int length) { journal_t *journal = EXT4_JOURNAL(page->mapping->host); + trace_ext4_journalled_invalidatepage(page, offset, length); + /* * If it's a full truncate we just forget about the pending dirtying */ - if (offset == 0) + if (offset == 0 && length == PAGE_CACHE_SIZE) ClearPageChecked(page); - jbd2_journal_invalidatepage(journal, page, offset); + return jbd2_journal_invalidatepage(journal, page, offset, length); +} + +/* Wrapper for aops... */ +static void ext4_journalled_invalidatepage(struct page *page, + unsigned int offset, + unsigned int length) +{ + WARN_ON(__ext4_journalled_invalidatepage(page, offset, length) < 0); } static int ext4_releasepage(struct page *page, gfp_t wait) { journal_t *journal = EXT4_JOURNAL(page->mapping->host); - WARN_ON(PageChecked(page)); - if (!page_has_buffers(page)) + trace_ext4_releasepage(page); + + /* Page has dirty journalled data -> cannot release */ + if (PageChecked(page)) return 0; - return jbd2_journal_try_to_free_buffers(journal, page, wait); + if (journal) + return jbd2_journal_try_to_free_buffers(journal, page, wait); + else + return try_to_free_buffers(page); +} + +/* + * ext4_get_block used when preparing for a DIO write or buffer write. + * We allocate an uinitialized extent if blocks haven't been allocated. + * The extent will be converted to initialized after the IO is complete. + */ +int ext4_get_block_write(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + ext4_debug("ext4_get_block_write: inode %lu, create flag %d\n", + inode->i_ino, create); + return _ext4_get_block(inode, iblock, bh_result, + EXT4_GET_BLOCKS_IO_CREATE_EXT); +} + +static int ext4_get_block_write_nolock(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + ext4_debug("ext4_get_block_write_nolock: inode %lu, create flag %d\n", + inode->i_ino, create); + return _ext4_get_block(inode, iblock, bh_result, + EXT4_GET_BLOCKS_NO_LOCK); +} + +static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, + ssize_t size, void *private) +{ + ext4_io_end_t *io_end = iocb->private; + + /* if not async direct IO just return */ + if (!io_end) + return; + + ext_debug("ext4_end_io_dio(): io_end 0x%p " + "for inode %lu, iocb 0x%p, offset %llu, size %zd\n", + iocb->private, io_end->inode->i_ino, iocb, offset, + size); + + iocb->private = NULL; + io_end->offset = offset; + io_end->size = size; + ext4_put_io_end(io_end); } /* + * For ext4 extent files, ext4 will do direct-io write to holes, + * preallocated extents, and those write extend the file, no need to + * fall back to buffered IO. + * + * For holes, we fallocate those blocks, mark them as unwritten + * If those blocks were preallocated, we mark sure they are split, but + * still keep the range to write as unwritten. + * + * The unwritten extents will be converted to written when DIO is completed. + * For async direct IO, since the IO may still pending when return, we + * set up an end_io call back function, which will do the conversion + * when async direct IO completed. + * * 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_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) +static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, + struct iov_iter *iter, loff_t offset) { 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); + size_t count = iov_iter_count(iter); + int overwrite = 0; + get_block_t *get_block_func = NULL; + int dio_flags = 0; + loff_t final_size = offset + count; + ext4_io_end_t *io_end = NULL; - if (rw == WRITE) { - loff_t final_size = offset + count; + /* Use the old path for reads and writes beyond i_size. */ + if (rw != WRITE || final_size > inode->i_size) + return ext4_ind_direct_IO(rw, iocb, iter, offset); - 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); - } - } + BUG_ON(iocb->private == NULL); - ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, - ext4_get_block, NULL); + /* + * Make all waiters for direct IO properly wait also for extent + * conversion. This also disallows race between truncate() and + * overwrite DIO as i_dio_count needs to be incremented under i_mutex. + */ + if (rw == WRITE) + atomic_inc(&inode->i_dio_count); - if (orphan) { - int err; + /* If we do a overwrite dio, i_mutex locking can be released */ + overwrite = *((int *)iocb->private); - /* 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); - goto out; + if (overwrite) { + down_read(&EXT4_I(inode)->i_data_sem); + mutex_unlock(&inode->i_mutex); + } + + /* + * We could direct write to holes and fallocate. + * + * Allocated blocks to fill the hole are marked as + * unwritten to prevent parallel buffered read to expose + * the stale data before DIO complete the data IO. + * + * As to previously fallocated extents, ext4 get_block will + * just simply mark the buffer mapped but still keep the + * extents unwritten. + * + * For non AIO case, we will convert those unwritten extents + * to written after return back from blockdev_direct_IO. + * + * For async DIO, the conversion needs to be deferred when the + * IO is completed. The ext4 end_io callback function will be + * called to take care of the conversion work. Here for async + * case, we allocate an io_end structure to hook to the iocb. + */ + iocb->private = NULL; + ext4_inode_aio_set(inode, NULL); + if (!is_sync_kiocb(iocb)) { + io_end = ext4_init_io_end(inode, GFP_NOFS); + if (!io_end) { + ret = -ENOMEM; + goto retake_lock; } - 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); - } + /* + * Grab reference for DIO. Will be dropped in ext4_end_io_dio() + */ + iocb->private = ext4_get_io_end(io_end); + /* + * we save the io structure for current async direct + * IO, so that later ext4_map_blocks() could flag the + * io structure whether there is a unwritten extents + * needs to be converted when IO is completed. + */ + ext4_inode_aio_set(inode, io_end); + } + + if (overwrite) { + get_block_func = ext4_get_block_write_nolock; + } else { + get_block_func = ext4_get_block_write; + dio_flags = DIO_LOCKING; + } + ret = __blockdev_direct_IO(rw, iocb, inode, + inode->i_sb->s_bdev, iter, + offset, + get_block_func, + ext4_end_io_dio, + NULL, + dio_flags); + + /* + * Put our reference to io_end. This can free the io_end structure e.g. + * in sync IO case or in case of error. It can even perform extent + * conversion if all bios we submitted finished before we got here. + * Note that in that case iocb->private can be already set to NULL + * here. + */ + if (io_end) { + ext4_inode_aio_set(inode, NULL); + ext4_put_io_end(io_end); + /* + * When no IO was submitted ext4_end_io_dio() was not + * called so we have to put iocb's reference. + */ + if (ret <= 0 && ret != -EIOCBQUEUED && iocb->private) { + WARN_ON(iocb->private != io_end); + WARN_ON(io_end->flag & EXT4_IO_END_UNWRITTEN); + ext4_put_io_end(io_end); + iocb->private = NULL; } - err = ext4_journal_stop(handle); - if (ret == 0) + } + if (ret > 0 && !overwrite && ext4_test_inode_state(inode, + EXT4_STATE_DIO_UNWRITTEN)) { + int err; + /* + * for non AIO case, since the IO is already + * completed, we could do the conversion right here + */ + err = ext4_convert_unwritten_extents(NULL, inode, + offset, ret); + if (err < 0) ret = err; + ext4_clear_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); } -out: + +retake_lock: + if (rw == WRITE) + inode_dio_done(inode); + /* take i_mutex locking again if we do a ovewrite dio */ + if (overwrite) { + up_read(&EXT4_I(inode)->i_data_sem); + mutex_lock(&inode->i_mutex); + } + + return ret; +} + +static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, + struct iov_iter *iter, loff_t offset) +{ + struct file *file = iocb->ki_filp; + struct inode *inode = file->f_mapping->host; + size_t count = iov_iter_count(iter); + ssize_t ret; + + /* + * If we are doing data journalling we don't support O_DIRECT + */ + if (ext4_should_journal_data(inode)) + return 0; + + /* Let buffer I/O handle the inline data case. */ + if (ext4_has_inline_data(inode)) + return 0; + + trace_ext4_direct_IO_enter(inode, offset, count, rw); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + ret = ext4_ext_direct_IO(rw, iocb, iter, offset); + else + ret = ext4_ind_direct_IO(rw, iocb, iter, offset); + trace_ext4_direct_IO_exit(inode, offset, count, rw, ret); return ret; } @@ -1804,88 +3274,110 @@ static int ext4_journalled_set_page_dirty(struct page *page) return __set_page_dirty_nobuffers(page); } -static const struct address_space_operations ext4_ordered_aops = { - .readpage = ext4_readpage, - .readpages = ext4_readpages, - .writepage = ext4_ordered_writepage, - .sync_page = block_sync_page, - .write_begin = ext4_write_begin, - .write_end = ext4_ordered_write_end, - .bmap = ext4_bmap, - .invalidatepage = ext4_invalidatepage, - .releasepage = ext4_releasepage, - .direct_IO = ext4_direct_IO, - .migratepage = buffer_migrate_page, +static const struct address_space_operations ext4_aops = { + .readpage = ext4_readpage, + .readpages = ext4_readpages, + .writepage = ext4_writepage, + .writepages = ext4_writepages, + .write_begin = ext4_write_begin, + .write_end = ext4_write_end, + .bmap = ext4_bmap, + .invalidatepage = ext4_invalidatepage, + .releasepage = ext4_releasepage, + .direct_IO = ext4_direct_IO, + .migratepage = buffer_migrate_page, + .is_partially_uptodate = block_is_partially_uptodate, + .error_remove_page = generic_error_remove_page, }; -static const struct address_space_operations ext4_writeback_aops = { - .readpage = ext4_readpage, - .readpages = ext4_readpages, - .writepage = ext4_writeback_writepage, - .sync_page = block_sync_page, - .write_begin = ext4_write_begin, - .write_end = ext4_writeback_write_end, - .bmap = ext4_bmap, - .invalidatepage = ext4_invalidatepage, - .releasepage = ext4_releasepage, - .direct_IO = ext4_direct_IO, - .migratepage = buffer_migrate_page, +static const struct address_space_operations ext4_journalled_aops = { + .readpage = ext4_readpage, + .readpages = ext4_readpages, + .writepage = ext4_writepage, + .writepages = ext4_writepages, + .write_begin = ext4_write_begin, + .write_end = ext4_journalled_write_end, + .set_page_dirty = ext4_journalled_set_page_dirty, + .bmap = ext4_bmap, + .invalidatepage = ext4_journalled_invalidatepage, + .releasepage = ext4_releasepage, + .direct_IO = ext4_direct_IO, + .is_partially_uptodate = block_is_partially_uptodate, + .error_remove_page = generic_error_remove_page, }; -static const struct address_space_operations ext4_journalled_aops = { - .readpage = ext4_readpage, - .readpages = ext4_readpages, - .writepage = ext4_journalled_writepage, - .sync_page = block_sync_page, - .write_begin = ext4_write_begin, - .write_end = ext4_journalled_write_end, - .set_page_dirty = ext4_journalled_set_page_dirty, - .bmap = ext4_bmap, - .invalidatepage = ext4_invalidatepage, - .releasepage = ext4_releasepage, +static const struct address_space_operations ext4_da_aops = { + .readpage = ext4_readpage, + .readpages = ext4_readpages, + .writepage = ext4_writepage, + .writepages = ext4_writepages, + .write_begin = ext4_da_write_begin, + .write_end = ext4_da_write_end, + .bmap = ext4_bmap, + .invalidatepage = ext4_da_invalidatepage, + .releasepage = ext4_releasepage, + .direct_IO = ext4_direct_IO, + .migratepage = buffer_migrate_page, + .is_partially_uptodate = block_is_partially_uptodate, + .error_remove_page = generic_error_remove_page, }; void ext4_set_aops(struct inode *inode) { - if (ext4_should_order_data(inode)) - inode->i_mapping->a_ops = &ext4_ordered_aops; - else if (ext4_should_writeback_data(inode)) - inode->i_mapping->a_ops = &ext4_writeback_aops; - else + switch (ext4_inode_journal_mode(inode)) { + case EXT4_INODE_ORDERED_DATA_MODE: + ext4_set_inode_state(inode, EXT4_STATE_ORDERED_MODE); + break; + case EXT4_INODE_WRITEBACK_DATA_MODE: + ext4_clear_inode_state(inode, EXT4_STATE_ORDERED_MODE); + break; + case EXT4_INODE_JOURNAL_DATA_MODE: inode->i_mapping->a_ops = &ext4_journalled_aops; + return; + default: + BUG(); + } + if (test_opt(inode->i_sb, DELALLOC)) + inode->i_mapping->a_ops = &ext4_da_aops; + else + inode->i_mapping->a_ops = &ext4_aops; } /* - * ext4_block_truncate_page() zeroes out a mapping from file offset `from' - * up to the end of the block which corresponds to `from'. - * This required during truncate. We need to physically zero the tail end - * of that block so it doesn't yield old data if the file is later grown. + * ext4_block_zero_page_range() zeros out a mapping of length 'length' + * starting from file offset 'from'. The range to be zero'd must + * be contained with in one block. If the specified range exceeds + * the end of the block it will be shortened to end of the block + * that cooresponds to 'from' */ -int ext4_block_truncate_page(handle_t *handle, struct page *page, - struct address_space *mapping, loff_t from) +static int ext4_block_zero_page_range(handle_t *handle, + struct address_space *mapping, loff_t from, loff_t length) { ext4_fsblk_t index = from >> PAGE_CACHE_SHIFT; unsigned offset = from & (PAGE_CACHE_SIZE-1); - unsigned blocksize, length, pos; + unsigned blocksize, max, pos; ext4_lblk_t iblock; struct inode *inode = mapping->host; struct buffer_head *bh; + struct page *page; int err = 0; + page = find_or_create_page(mapping, from >> PAGE_CACHE_SHIFT, + mapping_gfp_mask(mapping) & ~__GFP_FS); + if (!page) + return -ENOMEM; + blocksize = inode->i_sb->s_blocksize; - length = blocksize - (offset & (blocksize - 1)); - iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits); + max = blocksize - (offset & (blocksize - 1)); /* - * For "nobh" option, we can only work if we don't need to - * read-in the page - otherwise we create buffers to do the IO. + * correct length if it does not fall between + * 'from' and the end of the block */ - if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) && - ext4_should_writeback_data(inode) && PageUptodate(page)) { - zero_user(page, offset, length); - set_page_dirty(page); - goto unlock; - } + if (length > max || length < 0) + length = max; + + iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits); if (!page_has_buffers(page)) create_empty_buffers(page, blocksize, 0); @@ -1898,13 +3390,10 @@ int ext4_block_truncate_page(handle_t *handle, struct page *page, iblock++; pos += blocksize; } - - err = 0; if (buffer_freed(bh)) { BUFFER_TRACE(bh, "freed: skip"); goto unlock; } - if (!buffer_mapped(bh)) { BUFFER_TRACE(bh, "unmapped"); ext4_get_block(inode, iblock, bh, 0); @@ -1927,25 +3416,22 @@ int ext4_block_truncate_page(handle_t *handle, struct page *page, if (!buffer_uptodate(bh)) goto unlock; } - if (ext4_should_journal_data(inode)) { BUFFER_TRACE(bh, "get write access"); err = ext4_journal_get_write_access(handle, bh); if (err) goto unlock; } - zero_user(page, offset, length); - BUFFER_TRACE(bh, "zeroed end of block"); - err = 0; if (ext4_should_journal_data(inode)) { - err = ext4_journal_dirty_metadata(handle, bh); + err = ext4_handle_dirty_metadata(handle, inode, bh); } else { - if (ext4_should_order_data(inode)) - err = ext4_journal_dirty_data(handle, bh); + err = 0; mark_buffer_dirty(bh); + if (ext4_test_inode_state(inode, EXT4_STATE_ORDERED_MODE)) + err = ext4_jbd2_file_inode(handle, inode); } unlock: @@ -1955,342 +3441,235 @@ unlock: } /* - * 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? + * ext4_block_truncate_page() zeroes out a mapping from file offset `from' + * up to the end of the block which corresponds to `from'. + * This required during truncate. We need to physically zero the tail end + * of that block so it doesn't yield old data if the file is later grown. */ -static inline int all_zeroes(__le32 *p, __le32 *q) +static int ext4_block_truncate_page(handle_t *handle, + struct address_space *mapping, loff_t from) { - while (p < q) - if (*p++) - return 0; - return 1; + unsigned offset = from & (PAGE_CACHE_SIZE-1); + unsigned length; + unsigned blocksize; + struct inode *inode = mapping->host; + + blocksize = inode->i_sb->s_blocksize; + length = blocksize - (offset & (blocksize - 1)); + + return ext4_block_zero_page_range(handle, mapping, from, length); } -/** - * 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 refered - * 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 offest + 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 */ +int ext4_zero_partial_blocks(handle_t *handle, struct inode *inode, + loff_t lstart, loff_t length) +{ + struct super_block *sb = inode->i_sb; + struct address_space *mapping = inode->i_mapping; + unsigned partial_start, partial_end; + ext4_fsblk_t start, end; + loff_t byte_end = (lstart + length - 1); + int err = 0; + + partial_start = lstart & (sb->s_blocksize - 1); + partial_end = byte_end & (sb->s_blocksize - 1); + + start = lstart >> sb->s_blocksize_bits; + end = byte_end >> sb->s_blocksize_bits; - while(partial > p) { - brelse(partial->bh); - partial--; + /* Handle partial zero within the single block */ + if (start == end && + (partial_start || (partial_end != sb->s_blocksize - 1))) { + err = ext4_block_zero_page_range(handle, mapping, + lstart, length); + return err; + } + /* Handle partial zero out on the start of the range */ + if (partial_start) { + err = ext4_block_zero_page_range(handle, mapping, + lstart, sb->s_blocksize); + if (err) + return err; } -no_top: - return partial; + /* Handle partial zero out on the end of the range */ + if (partial_end != sb->s_blocksize - 1) + err = ext4_block_zero_page_range(handle, mapping, + byte_end - partial_end, + partial_end + 1); + return err; +} + +int ext4_can_truncate(struct inode *inode) +{ + if (S_ISREG(inode->i_mode)) + return 1; + if (S_ISDIR(inode->i_mode)) + return 1; + if (S_ISLNK(inode->i_mode)) + return !ext4_inode_is_fast_symlink(inode); + return 0; } /* - * 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. + * ext4_punch_hole: punches a hole in a file by releaseing the blocks + * associated with the given offset and length + * + * @inode: File inode + * @offset: The offset where the hole will begin + * @len: The length of the hole * - * We release `count' blocks on disk, but (last - first) may be greater - * than `count' because there can be holes in there. + * Returns: 0 on success or negative on failure */ -static void 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; - if (try_to_extend_transaction(handle, inode)) { - if (bh) { - BUFFER_TRACE(bh, "call ext4_journal_dirty_metadata"); - ext4_journal_dirty_metadata(handle, bh); - } - ext4_mark_inode_dirty(handle, inode); - ext4_journal_test_restart(handle, inode); - if (bh) { - BUFFER_TRACE(bh, "retaking write access"); - ext4_journal_get_write_access(handle, bh); - } - } + +int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length) +{ + struct super_block *sb = inode->i_sb; + ext4_lblk_t first_block, stop_block; + struct address_space *mapping = inode->i_mapping; + loff_t first_block_offset, last_block_offset; + handle_t *handle; + unsigned int credits; + int ret = 0; + + if (!S_ISREG(inode->i_mode)) + return -EOPNOTSUPP; + + trace_ext4_punch_hole(inode, offset, length, 0); /* - * Any buffers which are on the journal will be in memory. We find - * them on the hash table so jbd2_journal_revoke() will run jbd2_journal_forget() - * on them. We've already detached each block from the file, so - * bforget() in jbd2_journal_forget() should be safe. - * - * AKPM: turn on bforget in jbd2_journal_forget()!!! + * Write out all dirty pages to avoid race conditions + * Then release them. */ - for (p = first; p < last; p++) { - u32 nr = le32_to_cpu(*p); - if (nr) { - struct buffer_head *tbh; - - *p = 0; - tbh = sb_find_get_block(inode->i_sb, nr); - ext4_forget(handle, 0, inode, tbh, nr); - } + if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { + ret = filemap_write_and_wait_range(mapping, offset, + offset + length - 1); + if (ret) + return ret; } - ext4_free_blocks(handle, inode, block_to_free, count, 0); -} + mutex_lock(&inode->i_mutex); -/** - * 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 refered 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; + /* No need to punch hole beyond i_size */ + if (offset >= inode->i_size) + goto out_mutex; - 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; + /* + * If the hole extends beyond i_size, set the hole + * to end after the page that contains i_size + */ + if (offset + length > inode->i_size) { + length = inode->i_size + + PAGE_CACHE_SIZE - (inode->i_size & (PAGE_CACHE_SIZE - 1)) - + offset; } - 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 { - ext4_clear_blocks(handle, inode, this_bh, - block_to_free, - count, block_to_free_p, p); - block_to_free = nr; - block_to_free_p = p; - count = 1; - } - } + if (offset & (sb->s_blocksize - 1) || + (offset + length) & (sb->s_blocksize - 1)) { + /* + * Attach jinode to inode for jbd2 if we do any zeroing of + * partial block + */ + ret = ext4_inode_attach_jinode(inode); + if (ret < 0) + goto out_mutex; + } - if (count > 0) - ext4_clear_blocks(handle, inode, this_bh, block_to_free, - count, block_to_free_p, p); + first_block_offset = round_up(offset, sb->s_blocksize); + last_block_offset = round_down((offset + length), sb->s_blocksize) - 1; + + /* Now release the pages and zero block aligned part of pages*/ + if (last_block_offset > first_block_offset) + truncate_pagecache_range(inode, first_block_offset, + last_block_offset); - if (this_bh) { - BUFFER_TRACE(this_bh, "call ext4_journal_dirty_metadata"); - ext4_journal_dirty_metadata(handle, this_bh); + /* Wait all existing dio workers, newcomers will block on i_mutex */ + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + credits = ext4_writepage_trans_blocks(inode); + else + credits = ext4_blocks_for_truncate(inode); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + ext4_std_error(sb, ret); + goto out_dio; } -} -/** - * 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 refered 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; + ret = ext4_zero_partial_blocks(handle, inode, offset, + length); + if (ret) + goto out_stop; - if (is_handle_aborted(handle)) - return; + first_block = (offset + sb->s_blocksize - 1) >> + EXT4_BLOCK_SIZE_BITS(sb); + stop_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb); - 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 there are no blocks to remove, return now */ + if (first_block >= stop_block) + goto out_stop; - /* Go read the buffer for the next level down */ - bh = sb_bread(inode->i_sb, nr); + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); - /* - * A read failure? Report error and clear slot - * (should be rare). - */ - if (!bh) { - ext4_error(inode->i_sb, "ext4_free_branches", - "Read failure, inode=%lu, block=%llu", - inode->i_ino, nr); - continue; - } + ret = ext4_es_remove_extent(inode, first_block, + stop_block - first_block); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } - /* 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); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + ret = ext4_ext_remove_space(inode, first_block, + stop_block - 1); + else + ret = ext4_free_hole_blocks(handle, inode, first_block, + stop_block); - /* - * We've probably journalled the indirect block several - * times during the truncate. But it's no longer - * needed and we now drop it from the transaction via - * jbd2_journal_revoke(). - * - * That's easy if it's exclusively part of this - * transaction. But if it's part of the committing - * transaction then jbd2_journal_forget() will simply - * brelse() it. That means that if the underlying - * block is reallocated in ext4_get_block(), - * unmap_underlying_metadata() will find this block - * and will try to get rid of it. damn, damn. - * - * If this block has already been committed to the - * journal, a revoke record will be written. And - * revoke records must be emitted *before* clearing - * this block's bit in the bitmaps. - */ - ext4_forget(handle, 1, inode, bh, bh->b_blocknr); + up_write(&EXT4_I(inode)->i_data_sem); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); - /* - * 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 (is_handle_aborted(handle)) - return; - if (try_to_extend_transaction(handle, inode)) { - ext4_mark_inode_dirty(handle, inode); - ext4_journal_test_restart(handle, inode); - } + /* Now release the pages again to reduce race window */ + if (last_block_offset > first_block_offset) + truncate_pagecache_range(inode, first_block_offset, + last_block_offset); - ext4_free_blocks(handle, inode, nr, 1, 1); + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ext4_mark_inode_dirty(handle, inode); +out_stop: + ext4_journal_stop(handle); +out_dio: + ext4_inode_resume_unlocked_dio(inode); +out_mutex: + mutex_unlock(&inode->i_mutex); + return ret; +} - 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_journal_dirty_metadata"); - ext4_journal_dirty_metadata(handle, - parent_bh); - } - } +int ext4_inode_attach_jinode(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct jbd2_inode *jinode; + + if (ei->jinode || !EXT4_SB(inode->i_sb)->s_journal) + return 0; + + jinode = jbd2_alloc_inode(GFP_KERNEL); + spin_lock(&inode->i_lock); + if (!ei->jinode) { + if (!jinode) { + spin_unlock(&inode->i_lock); + return -ENOMEM; } - } 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); + ei->jinode = jinode; + jbd2_journal_init_jbd_inode(ei->jinode, inode); + jinode = NULL; } + spin_unlock(&inode->i_lock); + if (unlikely(jinode != NULL)) + jbd2_free_inode(jinode); + return 0; } /* @@ -2300,7 +3679,7 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, * transaction, and VFS/VM ensures that ext4_truncate() cannot run * simultaneously on behalf of the same inode. * - * As we work through the truncate and commmit bits of it to the journal there + * As we work through the truncate and commit bits of it to the journal there * is one core, guiding principle: the file's tree must always be consistent on * disk. We must be able to restart the truncate after a crash. * @@ -2323,73 +3702,61 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, */ 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); + unsigned int credits; + handle_t *handle; struct address_space *mapping = inode->i_mapping; - ext4_lblk_t offsets[4]; - Indirect chain[4]; - Indirect *partial; - __le32 nr = 0; - int n; - ext4_lblk_t last_block; - unsigned blocksize = inode->i_sb->s_blocksize; - struct page *page; - - if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || - S_ISLNK(inode->i_mode))) - return; - if (ext4_inode_is_fast_symlink(inode)) - return; - if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) - return; /* - * We have to lock the EOF page here, because lock_page() nests - * outside jbd2_journal_start(). + * There is a possibility that we're either freeing the inode + * or it's a completely new inode. In those cases we might not + * have i_mutex locked because it's not necessary. */ - if ((inode->i_size & (blocksize - 1)) == 0) { - /* Block boundary? Nothing to do */ - page = NULL; - } else { - page = grab_cache_page(mapping, - inode->i_size >> PAGE_CACHE_SHIFT); - if (!page) - return; - } + if (!(inode->i_state & (I_NEW|I_FREEING))) + WARN_ON(!mutex_is_locked(&inode->i_mutex)); + trace_ext4_truncate_enter(inode); - if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) { - ext4_ext_truncate(inode, page); + if (!ext4_can_truncate(inode)) return; + + ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS); + + 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_has_inline_data(inode)) { + int has_inline = 1; + + ext4_inline_data_truncate(inode, &has_inline); + if (has_inline) + return; } - handle = start_transaction(inode); - if (IS_ERR(handle)) { - if (page) { - clear_highpage(page); - flush_dcache_page(page); - unlock_page(page); - page_cache_release(page); - } - return; /* AKPM: return what? */ + /* If we zero-out tail of the page, we have to create jinode for jbd2 */ + if (inode->i_size & (inode->i_sb->s_blocksize - 1)) { + if (ext4_inode_attach_jinode(inode) < 0) + return; } - last_block = (inode->i_size + blocksize-1) - >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + credits = ext4_writepage_trans_blocks(inode); + else + credits = ext4_blocks_for_truncate(inode); - if (page) - ext4_block_truncate_page(handle, page, mapping, inode->i_size); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); + if (IS_ERR(handle)) { + ext4_std_error(inode->i_sb, PTR_ERR(handle)); + return; + } - n = ext4_block_to_path(inode, last_block, offsets, NULL); - if (n == 0) - goto out_stop; /* error */ + if (inode->i_size & (inode->i_sb->s_blocksize - 1)) + ext4_block_truncate_page(handle, mapping, inode->i_size); /* - * 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. + * 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. @@ -2397,96 +3764,23 @@ void ext4_truncate(struct inode *inode) if (ext4_orphan_add(handle, inode)) goto out_stop; - /* - * 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; - - /* - * From here we block out all ext4_get_block() callers who want to - * modify the block allocation tree. - */ - down_write(&ei->i_data_sem); + down_write(&EXT4_I(inode)->i_data_sem); - if (n == 1) { /* direct blocks */ - ext4_free_data(handle, inode, NULL, i_data+offsets[0], - i_data + EXT4_NDIR_BLOCKS); - goto do_indirects; - } + ext4_discard_preallocations(inode); - 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: - ; - } - - ext4_discard_reservation(inode); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + ext4_ext_truncate(handle, inode); + else + ext4_ind_truncate(handle, inode); 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)) - handle->h_sync = 1; + ext4_handle_sync(handle); + out_stop: /* - * If this was a simple ftruncate(), and the file will remain alive + * 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 @@ -2495,56 +3789,11 @@ out_stop: if (inode->i_nlink) ext4_orphan_del(handle, inode); + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ext4_mark_inode_dirty(handle, inode); ext4_journal_stop(handle); -} -static ext4_fsblk_t ext4_get_inode_block(struct super_block *sb, - unsigned long ino, struct ext4_iloc *iloc) -{ - unsigned long desc, group_desc; - ext4_group_t block_group; - unsigned long offset; - ext4_fsblk_t block; - struct buffer_head *bh; - struct ext4_group_desc * gdp; - - if (!ext4_valid_inum(sb, ino)) { - /* - * This error is already checked for in namei.c unless we are - * looking at an NFS filehandle, in which case no error - * report is needed - */ - return 0; - } - - block_group = (ino - 1) / EXT4_INODES_PER_GROUP(sb); - if (block_group >= EXT4_SB(sb)->s_groups_count) { - ext4_error(sb,"ext4_get_inode_block","group >= groups count"); - return 0; - } - smp_rmb(); - group_desc = block_group >> EXT4_DESC_PER_BLOCK_BITS(sb); - desc = block_group & (EXT4_DESC_PER_BLOCK(sb) - 1); - bh = EXT4_SB(sb)->s_group_desc[group_desc]; - if (!bh) { - ext4_error (sb, "ext4_get_inode_block", - "Descriptor not loaded"); - return 0; - } - - gdp = (struct ext4_group_desc *)((__u8 *)bh->b_data + - desc * EXT4_DESC_SIZE(sb)); - /* - * Figure out the offset within the block group inode table - */ - offset = ((ino - 1) % EXT4_INODES_PER_GROUP(sb)) * - EXT4_INODE_SIZE(sb); - block = ext4_inode_table(sb, gdp) + - (offset >> EXT4_BLOCK_SIZE_BITS(sb)); - - iloc->block_group = block_group; - iloc->offset = offset & (EXT4_BLOCK_SIZE(sb) - 1); - return block; + trace_ext4_truncate_exit(inode); } /* @@ -2556,23 +3805,45 @@ static ext4_fsblk_t ext4_get_inode_block(struct super_block *sb, static int __ext4_get_inode_loc(struct inode *inode, struct ext4_iloc *iloc, int in_mem) { - ext4_fsblk_t block; - struct buffer_head *bh; - - block = ext4_get_inode_block(inode->i_sb, inode->i_ino, iloc); - if (!block) + struct ext4_group_desc *gdp; + struct buffer_head *bh; + struct super_block *sb = inode->i_sb; + ext4_fsblk_t block; + int inodes_per_block, inode_offset; + + iloc->bh = NULL; + if (!ext4_valid_inum(sb, inode->i_ino)) return -EIO; - bh = sb_getblk(inode->i_sb, block); - if (!bh) { - ext4_error (inode->i_sb, "ext4_get_inode_loc", - "unable to read inode block - " - "inode=%lu, block=%llu", - inode->i_ino, block); + iloc->block_group = (inode->i_ino - 1) / EXT4_INODES_PER_GROUP(sb); + gdp = ext4_get_group_desc(sb, iloc->block_group, NULL); + if (!gdp) return -EIO; - } + + /* + * Figure out the offset within the block group inode table + */ + inodes_per_block = EXT4_SB(sb)->s_inodes_per_block; + inode_offset = ((inode->i_ino - 1) % + EXT4_INODES_PER_GROUP(sb)); + block = ext4_inode_table(sb, gdp) + (inode_offset / inodes_per_block); + iloc->offset = (inode_offset % inodes_per_block) * EXT4_INODE_SIZE(sb); + + bh = sb_getblk(sb, block); + if (unlikely(!bh)) + return -ENOMEM; if (!buffer_uptodate(bh)) { lock_buffer(bh); + + /* + * If the buffer has the write error flag, we have failed + * to write out another inode in the same block. In this + * case, we don't have to read the block because we may + * read the old inode data successfully. + */ + if (buffer_write_io_error(bh) && !buffer_uptodate(bh)) + set_buffer_uptodate(bh); + if (buffer_uptodate(bh)) { /* someone brought it uptodate while we waited */ unlock_buffer(bh); @@ -2586,29 +3857,13 @@ static int __ext4_get_inode_loc(struct inode *inode, */ if (in_mem) { struct buffer_head *bitmap_bh; - struct ext4_group_desc *desc; - int inodes_per_buffer; - int inode_offset, i; - ext4_group_t block_group; - int start; - - block_group = (inode->i_ino - 1) / - EXT4_INODES_PER_GROUP(inode->i_sb); - inodes_per_buffer = bh->b_size / - EXT4_INODE_SIZE(inode->i_sb); - inode_offset = ((inode->i_ino - 1) % - EXT4_INODES_PER_GROUP(inode->i_sb)); - start = inode_offset & ~(inodes_per_buffer - 1); + int i, start; - /* Is the inode bitmap in cache? */ - desc = ext4_get_group_desc(inode->i_sb, - block_group, NULL); - if (!desc) - goto make_io; + start = inode_offset & ~(inodes_per_block - 1); - bitmap_bh = sb_getblk(inode->i_sb, - ext4_inode_bitmap(inode->i_sb, desc)); - if (!bitmap_bh) + /* Is the inode bitmap in cache? */ + bitmap_bh = sb_getblk(sb, ext4_inode_bitmap(sb, gdp)); + if (unlikely(!bitmap_bh)) goto make_io; /* @@ -2620,14 +3875,14 @@ static int __ext4_get_inode_loc(struct inode *inode, brelse(bitmap_bh); goto make_io; } - for (i = start; i < start + inodes_per_buffer; i++) { + for (i = start; i < start + inodes_per_block; i++) { if (i == inode_offset) continue; if (ext4_test_bit(i, bitmap_bh->b_data)) break; } brelse(bitmap_bh); - if (i == start + inodes_per_buffer) { + if (i == start + inodes_per_block) { /* all other inodes are free, so skip I/O */ memset(bh->b_data, 0, bh->b_size); set_buffer_uptodate(bh); @@ -2638,19 +3893,43 @@ static int __ext4_get_inode_loc(struct inode *inode, make_io: /* + * If we need to do any I/O, try to pre-readahead extra + * blocks from the inode table. + */ + if (EXT4_SB(sb)->s_inode_readahead_blks) { + ext4_fsblk_t b, end, table; + unsigned num; + __u32 ra_blks = EXT4_SB(sb)->s_inode_readahead_blks; + + table = ext4_inode_table(sb, gdp); + /* s_inode_readahead_blks is always a power of 2 */ + b = block & ~((ext4_fsblk_t) ra_blks - 1); + if (table > b) + b = table; + end = b + ra_blks; + num = EXT4_INODES_PER_GROUP(sb); + if (ext4_has_group_desc_csum(sb)) + num -= ext4_itable_unused_count(sb, gdp); + table += num / inodes_per_block; + if (end > table) + end = table; + while (b <= end) + sb_breadahead(sb, b++); + } + + /* * There are other valid inodes in the buffer, this inode * has in-inode xattrs, or we don't have this inode in memory. * Read the block from disk. */ + trace_ext4_load_inode(inode); get_bh(bh); bh->b_end_io = end_buffer_read_sync; - submit_bh(READ_META, bh); + submit_bh(READ | REQ_META | REQ_PRIO, bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) { - ext4_error(inode->i_sb, "ext4_get_inode_loc", - "unable to read inode block - " - "inode=%lu, block=%llu", - inode->i_ino, block); + EXT4_ERROR_INODE_BLOCK(inode, block, + "unable to read itable block"); brelse(bh); return -EIO; } @@ -2664,46 +3943,55 @@ int ext4_get_inode_loc(struct inode *inode, struct ext4_iloc *iloc) { /* We have all inode data except xattrs in memory here. */ return __ext4_get_inode_loc(inode, iloc, - !(EXT4_I(inode)->i_state & EXT4_STATE_XATTR)); + !ext4_test_inode_state(inode, EXT4_STATE_XATTR)); } void ext4_set_inode_flags(struct inode *inode) { unsigned int flags = EXT4_I(inode)->i_flags; + unsigned int new_fl = 0; - inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); if (flags & EXT4_SYNC_FL) - inode->i_flags |= S_SYNC; + new_fl |= S_SYNC; if (flags & EXT4_APPEND_FL) - inode->i_flags |= S_APPEND; + new_fl |= S_APPEND; if (flags & EXT4_IMMUTABLE_FL) - inode->i_flags |= S_IMMUTABLE; + new_fl |= S_IMMUTABLE; if (flags & EXT4_NOATIME_FL) - inode->i_flags |= S_NOATIME; + new_fl |= S_NOATIME; if (flags & EXT4_DIRSYNC_FL) - inode->i_flags |= S_DIRSYNC; + new_fl |= S_DIRSYNC; + inode_set_flags(inode, new_fl, + S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); } /* Propagate flags from i_flags to EXT4_I(inode)->i_flags */ void ext4_get_inode_flags(struct ext4_inode_info *ei) { - unsigned int flags = ei->vfs_inode.i_flags; - - ei->i_flags &= ~(EXT4_SYNC_FL|EXT4_APPEND_FL| - EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL|EXT4_DIRSYNC_FL); - if (flags & S_SYNC) - ei->i_flags |= EXT4_SYNC_FL; - if (flags & S_APPEND) - ei->i_flags |= EXT4_APPEND_FL; - if (flags & S_IMMUTABLE) - ei->i_flags |= EXT4_IMMUTABLE_FL; - if (flags & S_NOATIME) - ei->i_flags |= EXT4_NOATIME_FL; - if (flags & S_DIRSYNC) - ei->i_flags |= EXT4_DIRSYNC_FL; + unsigned int vfs_fl; + unsigned long old_fl, new_fl; + + do { + vfs_fl = ei->vfs_inode.i_flags; + old_fl = ei->i_flags; + new_fl = old_fl & ~(EXT4_SYNC_FL|EXT4_APPEND_FL| + EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL| + EXT4_DIRSYNC_FL); + if (vfs_fl & S_SYNC) + new_fl |= EXT4_SYNC_FL; + if (vfs_fl & S_APPEND) + new_fl |= EXT4_APPEND_FL; + if (vfs_fl & S_IMMUTABLE) + new_fl |= EXT4_IMMUTABLE_FL; + if (vfs_fl & S_NOATIME) + new_fl |= EXT4_NOATIME_FL; + if (vfs_fl & S_DIRSYNC) + new_fl |= EXT4_DIRSYNC_FL; + } while (cmpxchg(&ei->i_flags, old_fl, new_fl) != old_fl); } + static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, - struct ext4_inode_info *ei) + struct ext4_inode_info *ei) { blkcnt_t i_blocks ; struct inode *inode = &(ei->vfs_inode); @@ -2714,7 +4002,7 @@ static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, /* we are using combined 48 bit field */ i_blocks = ((u64)le16_to_cpu(raw_inode->i_blocks_high)) << 32 | le32_to_cpu(raw_inode->i_blocks_lo); - if (ei->i_flags & EXT4_HUGE_FILE_FL) { + if (ext4_test_inode_flag(inode, EXT4_INODE_HUGE_FILE)) { /* i_blocks represent file system block size */ return i_blocks << (inode->i_blkbits - 9); } else { @@ -2725,15 +4013,30 @@ static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, } } +static inline void ext4_iget_extra_inode(struct inode *inode, + struct ext4_inode *raw_inode, + struct ext4_inode_info *ei) +{ + __le32 *magic = (void *)raw_inode + + EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize; + if (*magic == cpu_to_le32(EXT4_XATTR_MAGIC)) { + ext4_set_inode_state(inode, EXT4_STATE_XATTR); + ext4_find_inline_data_nolock(inode); + } else + EXT4_I(inode)->i_inline_off = 0; +} + struct inode *ext4_iget(struct super_block *sb, unsigned long ino) { struct ext4_iloc iloc; struct ext4_inode *raw_inode; struct ext4_inode_info *ei; - struct buffer_head *bh; struct inode *inode; + journal_t *journal = EXT4_SB(sb)->s_journal; long ret; int block; + uid_t i_uid; + gid_t i_gid; inode = iget_locked(sb, ino); if (!inode) @@ -2742,27 +4045,58 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) return inode; ei = EXT4_I(inode); -#ifdef CONFIG_EXT4DEV_FS_POSIX_ACL - ei->i_acl = EXT4_ACL_NOT_CACHED; - ei->i_default_acl = EXT4_ACL_NOT_CACHED; -#endif - ei->i_block_alloc_info = NULL; + iloc.bh = NULL; ret = __ext4_get_inode_loc(inode, &iloc, 0); if (ret < 0) goto bad_inode; - bh = iloc.bh; raw_inode = ext4_raw_inode(&iloc); + + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) { + ei->i_extra_isize = le16_to_cpu(raw_inode->i_extra_isize); + if (EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize > + EXT4_INODE_SIZE(inode->i_sb)) { + EXT4_ERROR_INODE(inode, "bad extra_isize (%u != %u)", + EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize, + EXT4_INODE_SIZE(inode->i_sb)); + ret = -EIO; + goto bad_inode; + } + } else + ei->i_extra_isize = 0; + + /* Precompute checksum seed for inode metadata */ + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) { + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + __u32 csum; + __le32 inum = cpu_to_le32(inode->i_ino); + __le32 gen = raw_inode->i_generation; + csum = ext4_chksum(sbi, sbi->s_csum_seed, (__u8 *)&inum, + sizeof(inum)); + ei->i_csum_seed = ext4_chksum(sbi, csum, (__u8 *)&gen, + sizeof(gen)); + } + + if (!ext4_inode_csum_verify(inode, raw_inode, ei)) { + EXT4_ERROR_INODE(inode, "checksum invalid"); + ret = -EIO; + goto bad_inode; + } + inode->i_mode = le16_to_cpu(raw_inode->i_mode); - inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); - inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); - if(!(test_opt (inode->i_sb, NO_UID32))) { - inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; - inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; + i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); + i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); + if (!(test_opt(inode->i_sb, NO_UID32))) { + i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; + i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; } - inode->i_nlink = le16_to_cpu(raw_inode->i_links_count); + i_uid_write(inode, i_uid); + i_gid_write(inode, i_gid); + set_nlink(inode, le16_to_cpu(raw_inode->i_links_count)); - ei->i_state = 0; + ext4_clear_state_flags(ei); /* Only relevant on 32-bit archs */ + ei->i_inline_off = 0; ei->i_dir_start_lookup = 0; ei->i_dtime = le32_to_cpu(raw_inode->i_dtime); /* We now have enough fields to check if the inode was active or not. @@ -2771,30 +4105,34 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) * NeilBrown 1999oct15 */ if (inode->i_nlink == 0) { - if (inode->i_mode == 0 || - !(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_ORPHAN_FS)) { + if ((inode->i_mode == 0 || + !(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_ORPHAN_FS)) && + ino != EXT4_BOOT_LOADER_INO) { /* this inode is deleted */ - brelse (bh); ret = -ESTALE; goto bad_inode; } /* The only unlinked inodes we let through here have * valid i_mode and are being read by the orphan * recovery code: that's fine, we're about to complete - * the process of deleting those. */ + * the process of deleting those. + * OR it is the EXT4_BOOT_LOADER_INO which is + * not initialized on a new filesystem. */ } ei->i_flags = le32_to_cpu(raw_inode->i_flags); inode->i_blocks = ext4_inode_blocks(raw_inode, ei); ei->i_file_acl = le32_to_cpu(raw_inode->i_file_acl_lo); - if (EXT4_SB(inode->i_sb)->s_es->s_creator_os != - cpu_to_le32(EXT4_OS_HURD)) { + if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_64BIT)) ei->i_file_acl |= ((__u64)le16_to_cpu(raw_inode->i_file_acl_high)) << 32; - } inode->i_size = ext4_isize(raw_inode); ei->i_disksize = inode->i_size; +#ifdef CONFIG_QUOTA + ei->i_reserved_quota = 0; +#endif inode->i_generation = le32_to_cpu(raw_inode->i_generation); ei->i_block_group = iloc.block_group; + ei->i_last_alloc_group = ~0; /* * NOTE! The in-memory inode i_data array is in little-endian order * even on big-endian machines: we do NOT byteswap the block numbers! @@ -2803,40 +4141,79 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) ei->i_data[block] = raw_inode->i_block[block]; INIT_LIST_HEAD(&ei->i_orphan); + /* + * Set transaction id's of transactions that have to be committed + * to finish f[data]sync. We set them to currently running transaction + * as we cannot be sure that the inode or some of its metadata isn't + * part of the transaction - the inode could have been reclaimed and + * now it is reread from disk. + */ + if (journal) { + transaction_t *transaction; + tid_t tid; + + read_lock(&journal->j_state_lock); + if (journal->j_running_transaction) + transaction = journal->j_running_transaction; + else + transaction = journal->j_committing_transaction; + if (transaction) + tid = transaction->t_tid; + else + tid = journal->j_commit_sequence; + read_unlock(&journal->j_state_lock); + ei->i_sync_tid = tid; + ei->i_datasync_tid = tid; + } + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) { - ei->i_extra_isize = le16_to_cpu(raw_inode->i_extra_isize); - if (EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize > - EXT4_INODE_SIZE(inode->i_sb)) { - brelse (bh); - ret = -EIO; - goto bad_inode; - } if (ei->i_extra_isize == 0) { /* The extra space is currently unused. Use it. */ ei->i_extra_isize = sizeof(struct ext4_inode) - EXT4_GOOD_OLD_INODE_SIZE; } else { - __le32 *magic = (void *)raw_inode + - EXT4_GOOD_OLD_INODE_SIZE + - ei->i_extra_isize; - if (*magic == cpu_to_le32(EXT4_XATTR_MAGIC)) - ei->i_state |= EXT4_STATE_XATTR; + ext4_iget_extra_inode(inode, raw_inode, ei); } - } else - ei->i_extra_isize = 0; + } EXT4_INODE_GET_XTIME(i_ctime, inode, raw_inode); EXT4_INODE_GET_XTIME(i_mtime, inode, raw_inode); EXT4_INODE_GET_XTIME(i_atime, inode, raw_inode); EXT4_EINODE_GET_XTIME(i_crtime, ei, raw_inode); - inode->i_version = le32_to_cpu(raw_inode->i_disk_version); - if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) { - if (EXT4_FITS_IN_INODE(raw_inode, ei, i_version_hi)) - inode->i_version |= - (__u64)(le32_to_cpu(raw_inode->i_version_hi)) << 32; + if (likely(!test_opt2(inode->i_sb, HURD_COMPAT))) { + inode->i_version = le32_to_cpu(raw_inode->i_disk_version); + if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) { + if (EXT4_FITS_IN_INODE(raw_inode, ei, i_version_hi)) + inode->i_version |= + (__u64)(le32_to_cpu(raw_inode->i_version_hi)) << 32; + } } + ret = 0; + if (ei->i_file_acl && + !ext4_data_block_valid(EXT4_SB(sb), ei->i_file_acl, 1)) { + EXT4_ERROR_INODE(inode, "bad extended attribute block %llu", + ei->i_file_acl); + ret = -EIO; + goto bad_inode; + } else if (!ext4_has_inline_data(inode)) { + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + if ((S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + (S_ISLNK(inode->i_mode) && + !ext4_inode_is_fast_symlink(inode)))) + /* Validate extent which is part of inode */ + ret = ext4_ext_check_inode(inode); + } else if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + (S_ISLNK(inode->i_mode) && + !ext4_inode_is_fast_symlink(inode))) { + /* Validate block references which are part of inode */ + ret = ext4_ind_check_inode(inode); + } + } + if (ret) + goto bad_inode; + if (S_ISREG(inode->i_mode)) { inode->i_op = &ext4_file_inode_operations; inode->i_fop = &ext4_file_operations; @@ -2845,13 +4222,16 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) inode->i_op = &ext4_dir_inode_operations; inode->i_fop = &ext4_dir_operations; } else if (S_ISLNK(inode->i_mode)) { - if (ext4_inode_is_fast_symlink(inode)) + if (ext4_inode_is_fast_symlink(inode)) { inode->i_op = &ext4_fast_symlink_inode_operations; - else { + nd_terminate_link(ei->i_data, inode->i_size, + sizeof(ei->i_data) - 1); + } else { inode->i_op = &ext4_symlink_inode_operations; ext4_set_aops(inode); } - } else { + } else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode) || + S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) { inode->i_op = &ext4_special_inode_operations; if (raw_inode->i_block[0]) init_special_inode(inode, inode->i_mode, @@ -2859,13 +4239,20 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) else init_special_inode(inode, inode->i_mode, new_decode_dev(le32_to_cpu(raw_inode->i_block[1]))); + } else if (ino == EXT4_BOOT_LOADER_INO) { + make_bad_inode(inode); + } else { + ret = -EIO; + EXT4_ERROR_INODE(inode, "bogus i_mode (%o)", inode->i_mode); + goto bad_inode; } - brelse (iloc.bh); + brelse(iloc.bh); ext4_set_inode_flags(inode); unlock_new_inode(inode); return inode; bad_inode: + brelse(iloc.bh); iget_failed(inode); return ERR_PTR(ret); } @@ -2877,46 +4264,36 @@ static int ext4_inode_blocks_set(handle_t *handle, struct inode *inode = &(ei->vfs_inode); u64 i_blocks = inode->i_blocks; struct super_block *sb = inode->i_sb; - int err = 0; if (i_blocks <= ~0U) { /* - * i_blocks can be represnted in a 32 bit variable + * i_blocks can be represented in a 32 bit variable * as multiple of 512 bytes */ raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = 0; - ei->i_flags &= ~EXT4_HUGE_FILE_FL; - } else if (i_blocks <= 0xffffffffffffULL) { + ext4_clear_inode_flag(inode, EXT4_INODE_HUGE_FILE); + return 0; + } + if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_HUGE_FILE)) + return -EFBIG; + + if (i_blocks <= 0xffffffffffffULL) { /* * i_blocks can be represented in a 48 bit variable * as multiple of 512 bytes */ - err = ext4_update_rocompat_feature(handle, sb, - EXT4_FEATURE_RO_COMPAT_HUGE_FILE); - if (err) - goto err_out; - /* i_block is stored in the split 48 bit fields */ raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32); - ei->i_flags &= ~EXT4_HUGE_FILE_FL; + ext4_clear_inode_flag(inode, EXT4_INODE_HUGE_FILE); } else { - /* - * i_blocks should be represented in a 48 bit variable - * as multiple of file system block size - */ - err = ext4_update_rocompat_feature(handle, sb, - EXT4_FEATURE_RO_COMPAT_HUGE_FILE); - if (err) - goto err_out; - ei->i_flags |= EXT4_HUGE_FILE_FL; + ext4_set_inode_flag(inode, EXT4_INODE_HUGE_FILE); /* i_block is stored in file system block size */ i_blocks = i_blocks >> (inode->i_blkbits - 9); raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32); } -err_out: - return err; + return 0; } /* @@ -2933,36 +4310,42 @@ static int ext4_do_update_inode(handle_t *handle, struct ext4_inode *raw_inode = ext4_raw_inode(iloc); struct ext4_inode_info *ei = EXT4_I(inode); struct buffer_head *bh = iloc->bh; + struct super_block *sb = inode->i_sb; int err = 0, rc, block; + int need_datasync = 0, set_large_file = 0; + uid_t i_uid; + gid_t i_gid; + + spin_lock(&ei->i_raw_lock); - /* For fields not not tracking in the in-memory inode, + /* For fields not tracked in the in-memory inode, * initialise them to zero for new inodes. */ - if (ei->i_state & EXT4_STATE_NEW) + if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size); ext4_get_inode_flags(ei); raw_inode->i_mode = cpu_to_le16(inode->i_mode); - if(!(test_opt(inode->i_sb, NO_UID32))) { - raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid)); - raw_inode->i_gid_low = cpu_to_le16(low_16_bits(inode->i_gid)); + i_uid = i_uid_read(inode); + i_gid = i_gid_read(inode); + if (!(test_opt(inode->i_sb, NO_UID32))) { + raw_inode->i_uid_low = cpu_to_le16(low_16_bits(i_uid)); + raw_inode->i_gid_low = cpu_to_le16(low_16_bits(i_gid)); /* * Fix up interoperability with old kernels. Otherwise, old inodes get * re-used with the upper 16 bits of the uid/gid intact */ - if(!ei->i_dtime) { + if (!ei->i_dtime) { raw_inode->i_uid_high = - cpu_to_le16(high_16_bits(inode->i_uid)); + cpu_to_le16(high_16_bits(i_uid)); raw_inode->i_gid_high = - cpu_to_le16(high_16_bits(inode->i_gid)); + cpu_to_le16(high_16_bits(i_gid)); } else { raw_inode->i_uid_high = 0; raw_inode->i_gid_high = 0; } } else { - raw_inode->i_uid_low = - cpu_to_le16(fs_high2lowuid(inode->i_uid)); - raw_inode->i_gid_low = - cpu_to_le16(fs_high2lowgid(inode->i_gid)); + raw_inode->i_uid_low = cpu_to_le16(fs_high2lowuid(i_uid)); + raw_inode->i_gid_low = cpu_to_le16(fs_high2lowgid(i_gid)); raw_inode->i_uid_high = 0; raw_inode->i_gid_high = 0; } @@ -2973,37 +4356,26 @@ static int ext4_do_update_inode(handle_t *handle, EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode); EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode); - if (ext4_inode_blocks_set(handle, raw_inode, ei)) + if (ext4_inode_blocks_set(handle, raw_inode, ei)) { + spin_unlock(&ei->i_raw_lock); goto out_brelse; + } raw_inode->i_dtime = cpu_to_le32(ei->i_dtime); - raw_inode->i_flags = cpu_to_le32(ei->i_flags); - if (EXT4_SB(inode->i_sb)->s_es->s_creator_os != - cpu_to_le32(EXT4_OS_HURD)) + raw_inode->i_flags = cpu_to_le32(ei->i_flags & 0xFFFFFFFF); + if (likely(!test_opt2(inode->i_sb, HURD_COMPAT))) raw_inode->i_file_acl_high = cpu_to_le16(ei->i_file_acl >> 32); raw_inode->i_file_acl_lo = cpu_to_le32(ei->i_file_acl); - ext4_isize_set(raw_inode, ei->i_disksize); + if (ei->i_disksize != ext4_isize(raw_inode)) { + ext4_isize_set(raw_inode, ei->i_disksize); + need_datasync = 1; + } if (ei->i_disksize > 0x7fffffffULL) { - struct super_block *sb = inode->i_sb; if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_LARGE_FILE) || EXT4_SB(sb)->s_es->s_rev_level == - cpu_to_le32(EXT4_GOOD_OLD_REV)) { - /* If this is the first large file - * created, add a flag to the superblock. - */ - err = ext4_journal_get_write_access(handle, - EXT4_SB(sb)->s_sbh); - if (err) - goto out_brelse; - ext4_update_dynamic_rev(sb); - EXT4_SET_RO_COMPAT_FEATURE(sb, - EXT4_FEATURE_RO_COMPAT_LARGE_FILE); - sb->s_dirt = 1; - handle->h_sync = 1; - err = ext4_journal_dirty_metadata(handle, - EXT4_SB(sb)->s_sbh); - } + cpu_to_le32(EXT4_GOOD_OLD_REV)) + set_large_file = 1; } raw_inode->i_generation = cpu_to_le32(inode->i_generation); if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) { @@ -3017,26 +4389,45 @@ static int ext4_do_update_inode(handle_t *handle, cpu_to_le32(new_encode_dev(inode->i_rdev)); raw_inode->i_block[2] = 0; } - } else for (block = 0; block < EXT4_N_BLOCKS; block++) - raw_inode->i_block[block] = ei->i_data[block]; + } else if (!ext4_has_inline_data(inode)) { + for (block = 0; block < EXT4_N_BLOCKS; block++) + raw_inode->i_block[block] = ei->i_data[block]; + } - raw_inode->i_disk_version = cpu_to_le32(inode->i_version); - if (ei->i_extra_isize) { - if (EXT4_FITS_IN_INODE(raw_inode, ei, i_version_hi)) - raw_inode->i_version_hi = - cpu_to_le32(inode->i_version >> 32); - raw_inode->i_extra_isize = cpu_to_le16(ei->i_extra_isize); + if (likely(!test_opt2(inode->i_sb, HURD_COMPAT))) { + raw_inode->i_disk_version = cpu_to_le32(inode->i_version); + if (ei->i_extra_isize) { + if (EXT4_FITS_IN_INODE(raw_inode, ei, i_version_hi)) + raw_inode->i_version_hi = + cpu_to_le32(inode->i_version >> 32); + raw_inode->i_extra_isize = + cpu_to_le16(ei->i_extra_isize); + } } + ext4_inode_csum_set(inode, raw_inode, ei); + + spin_unlock(&ei->i_raw_lock); - BUFFER_TRACE(bh, "call ext4_journal_dirty_metadata"); - rc = ext4_journal_dirty_metadata(handle, bh); + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + rc = ext4_handle_dirty_metadata(handle, NULL, bh); if (!err) err = rc; - ei->i_state &= ~EXT4_STATE_NEW; - + ext4_clear_inode_state(inode, EXT4_STATE_NEW); + if (set_large_file) { + BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get write access"); + err = ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh); + if (err) + goto out_brelse; + ext4_update_dynamic_rev(sb); + EXT4_SET_RO_COMPAT_FEATURE(sb, + EXT4_FEATURE_RO_COMPAT_LARGE_FILE); + ext4_handle_sync(handle); + err = ext4_handle_dirty_super(handle, sb); + } + ext4_update_inode_fsync_trans(handle, inode, need_datasync); out_brelse: - brelse (bh); + brelse(bh); ext4_std_error(inode->i_sb, err); return err; } @@ -3046,21 +4437,20 @@ out_brelse: * * We are called from a few places: * - * - Within generic_file_write() for O_SYNC files. + * - Within generic_file_aio_write() -> generic_write_sync() for O_SYNC files. * Here, there will be no transaction running. We wait for any running - * trasnaction to commit. + * transaction to commit. * - * - Within sys_sync(), kupdate and such. - * We wait on commit, if tol to. + * - Within flush work (sys_sync(), kupdate and such). + * We wait on commit, if told to. * - * - Within prune_icache() (PF_MEMALLOC == true) - * Here we simply return. We can't afford to block kswapd on the - * journal commit. + * - Within iput_final() -> write_inode_now() + * We wait on commit, if told to. * * In all cases it is actually safe for us to return without doing anything, * because the inode has been copied into a raw inode buffer in - * ext4_mark_inode_dirty(). This is a correctness thing for O_SYNC and for - * knfsd. + * ext4_mark_inode_dirty(). This is a correctness thing for WB_SYNC_ALL + * writeback. * * Note that we are absolutely dependent upon all inode dirtiers doing the * right thing: they *must* call mark_inode_dirty() after dirtying info in @@ -3072,25 +4462,95 @@ out_brelse: * stuff(); * inode->i_size = expr; * - * is in error because a kswapd-driven write_inode() could occur while - * `stuff()' is running, and the new i_size will be lost. Plus the inode - * will no longer be on the superblock's dirty inode list. + * is in error because write_inode() could occur while `stuff()' is running, + * and the new i_size will be lost. Plus the inode will no longer be on the + * superblock's dirty inode list. */ -int ext4_write_inode(struct inode *inode, int wait) +int ext4_write_inode(struct inode *inode, struct writeback_control *wbc) { - if (current->flags & PF_MEMALLOC) + int err; + + if (WARN_ON_ONCE(current->flags & PF_MEMALLOC)) return 0; - if (ext4_journal_current_handle()) { - jbd_debug(1, "called recursively, non-PF_MEMALLOC!\n"); - dump_stack(); - return -EIO; + if (EXT4_SB(inode->i_sb)->s_journal) { + if (ext4_journal_current_handle()) { + jbd_debug(1, "called recursively, non-PF_MEMALLOC!\n"); + dump_stack(); + return -EIO; + } + + /* + * No need to force transaction in WB_SYNC_NONE mode. Also + * ext4_sync_fs() will force the commit after everything is + * written. + */ + if (wbc->sync_mode != WB_SYNC_ALL || wbc->for_sync) + return 0; + + err = ext4_force_commit(inode->i_sb); + } else { + struct ext4_iloc iloc; + + err = __ext4_get_inode_loc(inode, &iloc, 0); + if (err) + return err; + /* + * sync(2) will flush the whole buffer cache. No need to do + * it here separately for each inode. + */ + if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync) + sync_dirty_buffer(iloc.bh); + if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) { + EXT4_ERROR_INODE_BLOCK(inode, iloc.bh->b_blocknr, + "IO error syncing inode"); + err = -EIO; + } + brelse(iloc.bh); } + return err; +} - if (!wait) - return 0; +/* + * In data=journal mode ext4_journalled_invalidatepage() may fail to invalidate + * buffers that are attached to a page stradding i_size and are undergoing + * commit. In that case we have to wait for commit to finish and try again. + */ +static void ext4_wait_for_tail_page_commit(struct inode *inode) +{ + struct page *page; + unsigned offset; + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; + tid_t commit_tid = 0; + int ret; - return ext4_force_commit(inode->i_sb); + offset = inode->i_size & (PAGE_CACHE_SIZE - 1); + /* + * All buffers in the last page remain valid? Then there's nothing to + * do. We do the check mainly to optimize the common PAGE_CACHE_SIZE == + * blocksize case + */ + if (offset > PAGE_CACHE_SIZE - (1 << inode->i_blkbits)) + return; + while (1) { + page = find_lock_page(inode->i_mapping, + inode->i_size >> PAGE_CACHE_SHIFT); + if (!page) + return; + ret = __ext4_journalled_invalidatepage(page, offset, + PAGE_CACHE_SIZE - offset); + unlock_page(page); + page_cache_release(page); + if (ret != -EBUSY) + return; + commit_tid = 0; + read_lock(&journal->j_state_lock); + if (journal->j_committing_transaction) + commit_tid = journal->j_committing_transaction->t_tid; + read_unlock(&journal->j_state_lock); + if (commit_tid) + jbd2_log_wait_commit(journal, commit_tid); + } } /* @@ -3108,31 +4568,42 @@ int ext4_write_inode(struct inode *inode, int wait) * be freed, so we have a strong guarantee that no future commit will * leave these blocks visible to the user.) * - * Called with inode->sem down. + * Another thing we have to assure is that if we are in ordered mode + * and inode is still attached to the committing transaction, we must + * we start writeout of all the dirty pages which are being truncated. + * This way we are sure that all the data written in the previous + * transaction are already on disk (truncate waits for pages under + * writeback). + * + * Called with inode->i_mutex down. */ int ext4_setattr(struct dentry *dentry, struct iattr *attr) { struct inode *inode = dentry->d_inode; int error, rc = 0; + int orphan = 0; const unsigned int ia_valid = attr->ia_valid; error = inode_change_ok(inode, attr); if (error) return error; - if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || - (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { + if (is_quota_modification(inode, attr)) + dquot_initialize(inode); + if ((ia_valid & ATTR_UID && !uid_eq(attr->ia_uid, inode->i_uid)) || + (ia_valid & ATTR_GID && !gid_eq(attr->ia_gid, inode->i_gid))) { handle_t *handle; /* (user+group)*(old+new) structure, inode write (sb, * inode block, ? - but truncate inode update has it) */ - handle = ext4_journal_start(inode, 2*(EXT4_QUOTA_INIT_BLOCKS(inode->i_sb)+ - EXT4_QUOTA_DEL_BLOCKS(inode->i_sb))+3); + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, + (EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb) + + EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb)) + 3); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto err_out; } - error = DQUOT_TRANSFER(inode, attr) ? -EDQUOT : 0; + error = dquot_transfer(inode, attr); if (error) { ext4_journal_stop(handle); return error; @@ -3147,45 +4618,97 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) ext4_journal_stop(handle); } - if (attr->ia_valid & ATTR_SIZE) { - if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)) { + if (attr->ia_valid & ATTR_SIZE && attr->ia_size != inode->i_size) { + handle_t *handle; + + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - if (attr->ia_size > sbi->s_bitmap_maxbytes) { - error = -EFBIG; - goto err_out; - } + if (attr->ia_size > sbi->s_bitmap_maxbytes) + return -EFBIG; } - } - if (S_ISREG(inode->i_mode) && - attr->ia_valid & ATTR_SIZE && attr->ia_size < inode->i_size) { - handle_t *handle; + if (IS_I_VERSION(inode) && attr->ia_size != inode->i_size) + inode_inc_iversion(inode); - handle = ext4_journal_start(inode, 3); - if (IS_ERR(handle)) { - error = PTR_ERR(handle); - goto err_out; - } + if (S_ISREG(inode->i_mode) && + (attr->ia_size < inode->i_size)) { + if (ext4_should_order_data(inode)) { + error = ext4_begin_ordered_truncate(inode, + attr->ia_size); + if (error) + goto err_out; + } + handle = ext4_journal_start(inode, EXT4_HT_INODE, 3); + if (IS_ERR(handle)) { + error = PTR_ERR(handle); + goto err_out; + } + if (ext4_handle_valid(handle)) { + error = ext4_orphan_add(handle, inode); + orphan = 1; + } + down_write(&EXT4_I(inode)->i_data_sem); + EXT4_I(inode)->i_disksize = attr->ia_size; + rc = ext4_mark_inode_dirty(handle, inode); + if (!error) + error = rc; + /* + * We have to update i_size under i_data_sem together + * with i_disksize to avoid races with writeback code + * running ext4_wb_update_i_disksize(). + */ + if (!error) + i_size_write(inode, attr->ia_size); + up_write(&EXT4_I(inode)->i_data_sem); + ext4_journal_stop(handle); + if (error) { + ext4_orphan_del(NULL, inode); + goto err_out; + } + } else + i_size_write(inode, attr->ia_size); - error = ext4_orphan_add(handle, inode); - EXT4_I(inode)->i_disksize = attr->ia_size; - rc = ext4_mark_inode_dirty(handle, inode); - if (!error) - error = rc; - ext4_journal_stop(handle); + /* + * Blocks are going to be removed from the inode. Wait + * for dio in flight. Temporarily disable + * dioread_nolock to prevent livelock. + */ + if (orphan) { + if (!ext4_should_journal_data(inode)) { + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + ext4_inode_resume_unlocked_dio(inode); + } else + ext4_wait_for_tail_page_commit(inode); + } + /* + * Truncate pagecache after we've waited for commit + * in data=journal mode to make pages freeable. + */ + truncate_pagecache(inode, inode->i_size); } + /* + * We want to call ext4_truncate() even if attr->ia_size == + * inode->i_size for cases like truncation of fallocated space + */ + if (attr->ia_valid & ATTR_SIZE) + ext4_truncate(inode); - rc = inode_setattr(inode, attr); + if (!rc) { + setattr_copy(inode, attr); + mark_inode_dirty(inode); + } - /* If inode_setattr's call to ext4_truncate failed to get a - * transaction handle at all, we need to clean up the in-core - * orphan list manually. */ - if (inode->i_nlink) + /* + * If the call to ext4_truncate failed to get a transaction handle at + * all, we need to clean up the in-core orphan list manually. + */ + if (orphan && inode->i_nlink) ext4_orphan_del(NULL, inode); if (!rc && (ia_valid & ATTR_MODE)) - rc = ext4_acl_chmod(inode); + rc = posix_acl_chmod(inode, inode->i_mode); err_out: ext4_std_error(inode->i_sb, error); @@ -3194,67 +4717,142 @@ err_out: return error; } +int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) +{ + struct inode *inode; + unsigned long long delalloc_blocks; + + inode = dentry->d_inode; + generic_fillattr(inode, stat); + + /* + * If there is inline data in the inode, the inode will normally not + * have data blocks allocated (it may have an external xattr block). + * Report at least one sector for such files, so tools like tar, rsync, + * others doen't incorrectly think the file is completely sparse. + */ + if (unlikely(ext4_has_inline_data(inode))) + stat->blocks += (stat->size + 511) >> 9; + + /* + * We can't update i_blocks if the block allocation is delayed + * otherwise in the case of system crash before the real block + * allocation is done, we will have i_blocks inconsistent with + * on-disk file blocks. + * We always keep i_blocks updated together with real + * allocation. But to not confuse with user, stat + * will return the blocks that include the delayed allocation + * blocks for this file. + */ + delalloc_blocks = EXT4_C2B(EXT4_SB(inode->i_sb), + EXT4_I(inode)->i_reserved_data_blocks); + stat->blocks += delalloc_blocks << (inode->i_sb->s_blocksize_bits - 9); + return 0; +} + +static int ext4_index_trans_blocks(struct inode *inode, int lblocks, + int pextents) +{ + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) + return ext4_ind_trans_blocks(inode, lblocks); + return ext4_ext_index_trans_blocks(inode, pextents); +} /* - * How many blocks doth make a writepage()? - * - * With N blocks per page, it may be: - * N data blocks - * 2 indirect block - * 2 dindirect - * 1 tindirect - * N+5 bitmap blocks (from the above) - * N+5 group descriptor summary blocks - * 1 inode block - * 1 superblock. - * 2 * EXT4_SINGLEDATA_TRANS_BLOCKS for the quote files - * - * 3 * (N + 5) + 2 + 2 * EXT4_SINGLEDATA_TRANS_BLOCKS - * - * With ordered or writeback data it's the same, less the N data blocks. - * - * If the inode's direct blocks can hold an integral number of pages then a - * page cannot straddle two indirect blocks, and we can only touch one indirect - * and dindirect block, and the "5" above becomes "3". - * - * This still overestimates under most circumstances. If we were to pass the - * start and end offsets in here as well we could do block_to_path() on each - * block and work out the exact number of indirects which are touched. Pah. + * Account for index blocks, block groups bitmaps and block group + * descriptor blocks if modify datablocks and index blocks + * worse case, the indexs blocks spread over different block groups + * + * If datablocks are discontiguous, they are possible to spread over + * different block groups too. If they are contiguous, with flexbg, + * they could still across block group boundary. + * + * Also account for superblock, inode, quota and xattr blocks */ +static int ext4_meta_trans_blocks(struct inode *inode, int lblocks, + int pextents) +{ + ext4_group_t groups, ngroups = ext4_get_groups_count(inode->i_sb); + int gdpblocks; + int idxblocks; + int ret = 0; + + /* + * How many index blocks need to touch to map @lblocks logical blocks + * to @pextents physical extents? + */ + idxblocks = ext4_index_trans_blocks(inode, lblocks, pextents); + + ret = idxblocks; + /* + * Now let's see how many group bitmaps and group descriptors need + * to account + */ + groups = idxblocks + pextents; + gdpblocks = groups; + if (groups > ngroups) + groups = ngroups; + if (groups > EXT4_SB(inode->i_sb)->s_gdb_count) + gdpblocks = EXT4_SB(inode->i_sb)->s_gdb_count; + + /* bitmaps and block group descriptor blocks */ + ret += groups + gdpblocks; + + /* Blocks for super block, inode, quota and xattr blocks */ + ret += EXT4_META_TRANS_BLOCKS(inode->i_sb); + + return ret; +} + +/* + * Calculate the total number of credits to reserve to fit + * the modification of a single pages into a single transaction, + * which may include multiple chunks of block allocations. + * + * This could be called via ext4_write_begin() + * + * We need to consider the worse case, when + * one new block per extent. + */ int ext4_writepage_trans_blocks(struct inode *inode) { int bpp = ext4_journal_blocks_per_page(inode); - int indirects = (EXT4_NDIR_BLOCKS % bpp) ? 5 : 3; int ret; - if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) - return ext4_ext_writepage_trans_blocks(inode, bpp); + ret = ext4_meta_trans_blocks(inode, bpp, bpp); + /* Account for data blocks for journalled mode */ if (ext4_should_journal_data(inode)) - ret = 3 * (bpp + indirects) + 2; - else - ret = 2 * (bpp + indirects) + 2; - -#ifdef CONFIG_QUOTA - /* We know that structure was already allocated during DQUOT_INIT so - * we will be updating only the data blocks + inodes */ - ret += 2*EXT4_QUOTA_TRANS_BLOCKS(inode->i_sb); -#endif - + ret += bpp; return ret; } /* + * Calculate the journal credits for a chunk of data modification. + * + * This is called from DIO, fallocate or whoever calling + * ext4_map_blocks() to map/allocate a chunk of contiguous disk blocks. + * + * journal buffers for data blocks are not included here, as DIO + * and fallocate do no need to journal data buffers. + */ +int ext4_chunk_trans_blocks(struct inode *inode, int nrblocks) +{ + return ext4_meta_trans_blocks(inode, nrblocks, 1); +} + +/* * The caller must have previously called ext4_reserve_inode_write(). * Give this, we know that the caller already has write access to iloc->bh. */ int ext4_mark_iloc_dirty(handle_t *handle, - struct inode *inode, struct ext4_iloc *iloc) + struct inode *inode, struct ext4_iloc *iloc) { int err = 0; - if (test_opt(inode->i_sb, I_VERSION)) + if (IS_I_VERSION(inode)) inode_inc_iversion(inode); /* the do_update_inode consumes one bh->b_count */ @@ -3275,16 +4873,15 @@ int ext4_reserve_inode_write(handle_t *handle, struct inode *inode, struct ext4_iloc *iloc) { - int err = 0; - if (handle) { - err = ext4_get_inode_loc(inode, iloc); - if (!err) { - BUFFER_TRACE(iloc->bh, "get_write_access"); - err = ext4_journal_get_write_access(handle, iloc->bh); - if (err) { - brelse(iloc->bh); - iloc->bh = NULL; - } + int err; + + err = ext4_get_inode_loc(inode, iloc); + if (!err) { + BUFFER_TRACE(iloc->bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, iloc->bh); + if (err) { + brelse(iloc->bh); + iloc->bh = NULL; } } ext4_std_error(inode->i_sb, err); @@ -3302,7 +4899,6 @@ static int ext4_expand_extra_isize(struct inode *inode, { struct ext4_inode *raw_inode; struct ext4_xattr_ibody_header *header; - struct ext4_xattr_entry *entry; if (EXT4_I(inode)->i_extra_isize >= new_extra_isize) return 0; @@ -3310,11 +4906,10 @@ static int ext4_expand_extra_isize(struct inode *inode, raw_inode = ext4_raw_inode(&iloc); header = IHDR(inode, raw_inode); - entry = IFIRST(header); /* No extended attributes present */ - if (!(EXT4_I(inode)->i_state & EXT4_STATE_XATTR) || - header->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC)) { + if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR) || + header->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC)) { memset((void *)raw_inode + EXT4_GOOD_OLD_INODE_SIZE, 0, new_extra_isize); EXT4_I(inode)->i_extra_isize = new_extra_isize; @@ -3338,14 +4933,6 @@ static int ext4_expand_extra_isize(struct inode *inode, * inode out, but prune_icache isn't a user-visible syncing function. * Whenever the user wants stuff synced (sys_sync, sys_msync, sys_fsync) * we start and wait on commits. - * - * Is this efficient/effective? Well, we're being nice to the system - * by cleaning up our inodes proactively so they can be reaped - * without I/O. But we are potentially leaving up to five seconds' - * worth of inodes floating about which prune_icache wants us to - * write out. One way to fix that would be to get prune_icache() - * to do a write_super() to free up some memory. It has the desired - * effect. */ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode) { @@ -3355,9 +4942,11 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode) int err, ret; might_sleep(); + trace_ext4_mark_inode_dirty(inode, _RET_IP_); err = ext4_reserve_inode_write(handle, inode, &iloc); - if (EXT4_I(inode)->i_extra_isize < sbi->s_want_extra_isize && - !(EXT4_I(inode)->i_state & EXT4_STATE_NO_EXPAND)) { + if (ext4_handle_valid(handle) && + EXT4_I(inode)->i_extra_isize < sbi->s_want_extra_isize && + !ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND)) { /* * We need extra buffer credits since we may write into EA block * with this same handle. If journal_extend fails, then it will @@ -3371,10 +4960,11 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode) sbi->s_want_extra_isize, iloc, handle); if (ret) { - EXT4_I(inode)->i_state |= EXT4_STATE_NO_EXPAND; + ext4_set_inode_state(inode, + EXT4_STATE_NO_EXPAND); if (mnt_count != le16_to_cpu(sbi->s_es->s_mnt_count)) { - ext4_warning(inode->i_sb, __FUNCTION__, + ext4_warning(inode->i_sb, "Unable to expand inode %lu. Delete" " some EAs or run e2fsck.", inode->i_ino); @@ -3396,31 +4986,23 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode) * i_size has been changed by generic_commit_write() and we thus need * to include the updated inode in the current transaction. * - * Also, DQUOT_ALLOC_SPACE() will always dirty the inode when blocks + * Also, dquot_alloc_block() will always dirty the inode when blocks * are allocated to the file. * * If the inode is marked synchronous, we don't honour that here - doing * so would cause a commit on atime updates, which we don't bother doing. * We handle synchronous inodes at the highest possible level. */ -void ext4_dirty_inode(struct inode *inode) +void ext4_dirty_inode(struct inode *inode, int flags) { - handle_t *current_handle = ext4_journal_current_handle(); handle_t *handle; - handle = ext4_journal_start(inode, 2); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 2); if (IS_ERR(handle)) goto out; - if (current_handle && - current_handle->h_transaction != handle->h_transaction) { - /* This task has a transaction open against a different fs */ - printk(KERN_EMERG "%s: transactions do not match!\n", - __FUNCTION__); - } else { - jbd_debug(5, "marking dirty. outer handle=%p\n", - current_handle); - ext4_mark_inode_dirty(handle, inode); - } + + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); out: return; @@ -3445,8 +5027,9 @@ static int ext4_pin_inode(handle_t *handle, struct inode *inode) BUFFER_TRACE(iloc.bh, "get_write_access"); err = jbd2_journal_get_write_access(handle, iloc.bh); if (!err) - err = ext4_journal_dirty_metadata(handle, - iloc.bh); + err = ext4_handle_dirty_metadata(handle, + NULL, + iloc.bh); brelse(iloc.bh); } } @@ -3472,11 +5055,27 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) */ journal = EXT4_JOURNAL(inode); + if (!journal) + return 0; if (is_journal_aborted(journal)) return -EROFS; + /* We have to allocate physical blocks for delalloc blocks + * before flushing journal. otherwise delalloc blocks can not + * be allocated any more. even more truncate on delalloc blocks + * could trigger BUG by flushing delalloc blocks in journal. + * There is no delalloc block in non-journal data mode. + */ + if (val && test_opt(inode->i_sb, DELALLOC)) { + err = ext4_alloc_da_blocks(inode); + if (err < 0) + return err; + } + + /* Wait for all existing dio workers */ + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); jbd2_journal_lock_updates(journal); - jbd2_journal_flush(journal); /* * OK, there are no updates running now, and all cached data is @@ -3487,23 +5086,119 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) */ if (val) - EXT4_I(inode)->i_flags |= EXT4_JOURNAL_DATA_FL; - else - EXT4_I(inode)->i_flags &= ~EXT4_JOURNAL_DATA_FL; + ext4_set_inode_flag(inode, EXT4_INODE_JOURNAL_DATA); + else { + jbd2_journal_flush(journal); + ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA); + } ext4_set_aops(inode); jbd2_journal_unlock_updates(journal); + ext4_inode_resume_unlocked_dio(inode); /* Finally we can mark the inode as dirty. */ - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); if (IS_ERR(handle)) return PTR_ERR(handle); err = ext4_mark_inode_dirty(handle, inode); - handle->h_sync = 1; + ext4_handle_sync(handle); ext4_journal_stop(handle); ext4_std_error(inode->i_sb, err); return err; } + +static int ext4_bh_unmapped(handle_t *handle, struct buffer_head *bh) +{ + return !buffer_mapped(bh); +} + +int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + struct page *page = vmf->page; + loff_t size; + unsigned long len; + int ret; + struct file *file = vma->vm_file; + struct inode *inode = file_inode(file); + struct address_space *mapping = inode->i_mapping; + handle_t *handle; + get_block_t *get_block; + int retries = 0; + + sb_start_pagefault(inode->i_sb); + file_update_time(vma->vm_file); + /* Delalloc case is easy... */ + if (test_opt(inode->i_sb, DELALLOC) && + !ext4_should_journal_data(inode) && + !ext4_nonda_switch(inode->i_sb)) { + do { + ret = __block_page_mkwrite(vma, vmf, + ext4_da_get_block_prep); + } while (ret == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)); + goto out_ret; + } + + lock_page(page); + size = i_size_read(inode); + /* Page got truncated from under us? */ + if (page->mapping != mapping || page_offset(page) > size) { + unlock_page(page); + ret = VM_FAULT_NOPAGE; + goto out; + } + + if (page->index == size >> PAGE_CACHE_SHIFT) + len = size & ~PAGE_CACHE_MASK; + else + len = PAGE_CACHE_SIZE; + /* + * Return if we have all the buffers mapped. This avoids the need to do + * journal_start/journal_stop which can block and take a long time + */ + if (page_has_buffers(page)) { + if (!ext4_walk_page_buffers(NULL, page_buffers(page), + 0, len, NULL, + ext4_bh_unmapped)) { + /* Wait so that we don't change page under IO */ + wait_for_stable_page(page); + ret = VM_FAULT_LOCKED; + goto out; + } + } + unlock_page(page); + /* OK, we need to fill the hole... */ + if (ext4_should_dioread_nolock(inode)) + get_block = ext4_get_block_write; + else + get_block = ext4_get_block; +retry_alloc: + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, + ext4_writepage_trans_blocks(inode)); + if (IS_ERR(handle)) { + ret = VM_FAULT_SIGBUS; + goto out; + } + ret = __block_page_mkwrite(vma, vmf, get_block); + if (!ret && ext4_should_journal_data(inode)) { + if (ext4_walk_page_buffers(handle, page_buffers(page), 0, + PAGE_CACHE_SIZE, NULL, do_journal_get_write_access)) { + unlock_page(page); + ret = VM_FAULT_SIGBUS; + ext4_journal_stop(handle); + goto out; + } + ext4_set_inode_state(inode, EXT4_STATE_JDATA); + } + ext4_journal_stop(handle); + if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry_alloc; +out_ret: + ret = block_page_mkwrite_return(ret); +out: + sb_end_pagefault(inode->i_sb); + return ret; +} |