From a633f5a319cf4116d977e25fea2830dce23a8e74 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Sat, 22 Feb 2014 06:18:17 -0500 Subject: ext4: translate fallocate mode bits to strings Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" --- include/trace/events/ext4.h | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 197d3125df2..451e0202aa6 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -68,6 +68,11 @@ struct extent_status; { EXTENT_STATUS_DELAYED, "D" }, \ { EXTENT_STATUS_HOLE, "H" }) +#define show_falloc_mode(mode) __print_flags(mode, "|", \ + { FALLOC_FL_KEEP_SIZE, "KEEP_SIZE"}, \ + { FALLOC_FL_PUNCH_HOLE, "PUNCH_HOLE"}, \ + { FALLOC_FL_NO_HIDE_STALE, "NO_HIDE_STALE"}) + TRACE_EVENT(ext4_free_inode, TP_PROTO(struct inode *inode), @@ -1349,10 +1354,10 @@ TRACE_EVENT(ext4_fallocate_enter, __entry->mode = mode; ), - TP_printk("dev %d,%d ino %lu pos %lld len %lld mode %d", + TP_printk("dev %d,%d ino %lu pos %lld len %lld mode %s", MAJOR(__entry->dev), MINOR(__entry->dev), (unsigned long) __entry->ino, __entry->pos, - __entry->len, __entry->mode) + __entry->len, show_falloc_mode(__entry->mode)) ); TRACE_EVENT(ext4_fallocate_exit, -- cgit v1.2.3-70-g09d2 From 9eb79482a97152930b113b51dff530aba9e28c8e Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sun, 23 Feb 2014 15:18:59 -0500 Subject: ext4: Add support FALLOC_FL_COLLAPSE_RANGE for fallocate This patch implements fallocate's FALLOC_FL_COLLAPSE_RANGE for Ext4. The semantics of this flag are following: 1) It collapses the range lying between offset and length by removing any data blocks which are present in this range and than updates all the logical offsets of extents beyond "offset + len" to nullify the hole created by removing blocks. In short, it does not leave a hole. 2) It should be used exclusively. No other fallocate flag in combination. 3) Offset and length supplied to fallocate should be fs block size aligned in case of xfs and ext4. 4) Collaspe range does not work beyond i_size. Signed-off-by: Namjae Jeon Signed-off-by: Ashish Sangwan Tested-by: Dongsu Park Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 3 + fs/ext4/extents.c | 307 +++++++++++++++++++++++++++++++++++++++++++- fs/ext4/move_extent.c | 2 +- include/trace/events/ext4.h | 33 ++++- 4 files changed, 342 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index b7207db3107..beec42750a8 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2758,6 +2758,7 @@ extern int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk); extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); extern int ext4_ext_precache(struct inode *inode); +extern int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len); /* move_extent.c */ extern void ext4_double_down_write_data_sem(struct inode *first, @@ -2767,6 +2768,8 @@ extern void ext4_double_up_write_data_sem(struct inode *orig_inode, extern int ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 start_orig, __u64 start_donor, __u64 len, __u64 *moved_len); +extern int mext_next_extent(struct inode *inode, struct ext4_ext_path *path, + struct ext4_extent **extent); /* page-io.c */ extern int __init ext4_init_pageio(void); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 2e0608e3be6..bbba1ef5417 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4581,12 +4581,16 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) unsigned int credits, blkbits = inode->i_blkbits; /* Return error if mode is not supported */ - if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) + if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | + FALLOC_FL_COLLAPSE_RANGE)) return -EOPNOTSUPP; if (mode & FALLOC_FL_PUNCH_HOLE) return ext4_punch_hole(inode, offset, len); + if (mode & FALLOC_FL_COLLAPSE_RANGE) + return ext4_collapse_range(inode, offset, len); + ret = ext4_convert_inline_data(inode); if (ret) return ret; @@ -4885,3 +4889,304 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, ext4_es_lru_add(inode); return error; } + +/* + * ext4_access_path: + * Function to access the path buffer for marking it dirty. + * It also checks if there are sufficient credits left in the journal handle + * to update path. + */ +static int +ext4_access_path(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path) +{ + int credits, err; + + if (!ext4_handle_valid(handle)) + return 0; + + /* + * Check if need to extend journal credits + * 3 for leaf, sb, and inode plus 2 (bmap and group + * descriptor) for each block group; assume two block + * groups + */ + if (handle->h_buffer_credits < 7) { + credits = ext4_writepage_trans_blocks(inode); + err = ext4_ext_truncate_extend_restart(handle, inode, credits); + /* EAGAIN is success */ + if (err && err != -EAGAIN) + return err; + } + + err = ext4_ext_get_access(handle, inode, path); + return err; +} + +/* + * ext4_ext_shift_path_extents: + * Shift the extents of a path structure lying between path[depth].p_ext + * and EXT_LAST_EXTENT(path[depth].p_hdr) downwards, by subtracting shift + * from starting block for each extent. + */ +static int +ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift, + struct inode *inode, handle_t *handle, + ext4_lblk_t *start) +{ + int depth, err = 0; + struct ext4_extent *ex_start, *ex_last; + bool update = 0; + depth = path->p_depth; + + while (depth >= 0) { + if (depth == path->p_depth) { + ex_start = path[depth].p_ext; + if (!ex_start) + return -EIO; + + ex_last = EXT_LAST_EXTENT(path[depth].p_hdr); + if (!ex_last) + return -EIO; + + err = ext4_access_path(handle, inode, path + depth); + if (err) + goto out; + + if (ex_start == EXT_FIRST_EXTENT(path[depth].p_hdr)) + update = 1; + + *start = ex_last->ee_block + + ext4_ext_get_actual_len(ex_last); + + while (ex_start <= ex_last) { + ex_start->ee_block -= shift; + if (ex_start > + EXT_FIRST_EXTENT(path[depth].p_hdr)) { + if (ext4_ext_try_to_merge_right(inode, + path, ex_start - 1)) + ex_last--; + } + ex_start++; + } + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto out; + + if (--depth < 0 || !update) + break; + } + + /* Update index too */ + err = ext4_access_path(handle, inode, path + depth); + if (err) + goto out; + + path[depth].p_idx->ei_block -= shift; + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto out; + + /* we are done if current index is not a starting index */ + if (path[depth].p_idx != EXT_FIRST_INDEX(path[depth].p_hdr)) + break; + + depth--; + } + +out: + return err; +} + +/* + * ext4_ext_shift_extents: + * All the extents which lies in the range from start to the last allocated + * block for the file are shifted downwards by shift blocks. + * On success, 0 is returned, error otherwise. + */ +static int +ext4_ext_shift_extents(struct inode *inode, handle_t *handle, + ext4_lblk_t start, ext4_lblk_t shift) +{ + struct ext4_ext_path *path; + int ret = 0, depth; + struct ext4_extent *extent; + ext4_lblk_t stop_block, current_block; + ext4_lblk_t ex_start, ex_end; + + /* Let path point to the last extent */ + path = ext4_ext_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + + depth = path->p_depth; + extent = path[depth].p_ext; + if (!extent) { + ext4_ext_drop_refs(path); + kfree(path); + return ret; + } + + stop_block = extent->ee_block + ext4_ext_get_actual_len(extent); + ext4_ext_drop_refs(path); + kfree(path); + + /* Nothing to shift, if hole is at the end of file */ + if (start >= stop_block) + return ret; + + /* + * Don't start shifting extents until we make sure the hole is big + * enough to accomodate the shift. + */ + path = ext4_ext_find_extent(inode, start - 1, NULL, 0); + depth = path->p_depth; + extent = path[depth].p_ext; + ex_start = extent->ee_block; + ex_end = extent->ee_block + ext4_ext_get_actual_len(extent); + ext4_ext_drop_refs(path); + kfree(path); + + if ((start == ex_start && shift > ex_start) || + (shift > start - ex_end)) + return -EINVAL; + + /* Its safe to start updating extents */ + while (start < stop_block) { + path = ext4_ext_find_extent(inode, start, NULL, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + depth = path->p_depth; + extent = path[depth].p_ext; + current_block = extent->ee_block; + if (start > current_block) { + /* Hole, move to the next extent */ + ret = mext_next_extent(inode, path, &extent); + if (ret != 0) { + ext4_ext_drop_refs(path); + kfree(path); + if (ret == 1) + ret = 0; + break; + } + } + ret = ext4_ext_shift_path_extents(path, shift, inode, + handle, &start); + ext4_ext_drop_refs(path); + kfree(path); + if (ret) + break; + } + + return ret; +} + +/* + * ext4_collapse_range: + * This implements the fallocate's collapse range functionality for ext4 + * Returns: 0 and non-zero on error. + */ +int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) +{ + struct super_block *sb = inode->i_sb; + ext4_lblk_t punch_start, punch_stop; + handle_t *handle; + unsigned int credits; + loff_t new_size; + int ret; + + BUG_ON(offset + len > i_size_read(inode)); + + /* Collapse range works only on fs block size aligned offsets. */ + if (offset & (EXT4_BLOCK_SIZE(sb) - 1) || + len & (EXT4_BLOCK_SIZE(sb) - 1)) + return -EINVAL; + + if (!S_ISREG(inode->i_mode)) + return -EOPNOTSUPP; + + trace_ext4_collapse_range(inode, offset, len); + + punch_start = offset >> EXT4_BLOCK_SIZE_BITS(sb); + punch_stop = (offset + len) >> EXT4_BLOCK_SIZE_BITS(sb); + + /* Write out all dirty pages */ + ret = filemap_write_and_wait_range(inode->i_mapping, offset, -1); + if (ret) + return ret; + + /* Take mutex lock */ + mutex_lock(&inode->i_mutex); + + /* It's not possible punch hole on append only file */ + if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) { + ret = -EPERM; + goto out_mutex; + } + + if (IS_SWAPFILE(inode)) { + ret = -ETXTBSY; + goto out_mutex; + } + + /* Currently just for extent based files */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + ret = -EOPNOTSUPP; + goto out_mutex; + } + + truncate_pagecache_range(inode, offset, -1); + + /* Wait for existing dio to complete */ + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + + credits = ext4_writepage_trans_blocks(inode); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out_dio; + } + + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); + + ret = ext4_es_remove_extent(inode, punch_start, + EXT_MAX_BLOCKS - punch_start - 1); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } + + ret = ext4_ext_remove_space(inode, punch_start, punch_stop - 1); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } + + ret = ext4_ext_shift_extents(inode, handle, punch_stop, + punch_stop - punch_start); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } + + new_size = i_size_read(inode) - len; + truncate_setsize(inode, new_size); + EXT4_I(inode)->i_disksize = new_size; + + ext4_discard_preallocations(inode); + up_write(&EXT4_I(inode)->i_data_sem); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + 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; +} diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index f39a88abe32..58ee7dc8766 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -76,7 +76,7 @@ copy_extent_status(struct ext4_extent *src, struct ext4_extent *dest) * ext4_ext_path structure refers to the last extent, or a negative error * value on failure. */ -static int +int mext_next_extent(struct inode *inode, struct ext4_ext_path *path, struct ext4_extent **extent) { diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 451e0202aa6..e9d7ee77d3a 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -16,6 +16,11 @@ struct mpage_da_data; struct ext4_map_blocks; struct extent_status; +/* shim until we merge in the xfs_collapse_range branch */ +#ifndef FALLOC_FL_COLLAPSE_RANGE +#define FALLOC_FL_COLLAPSE_RANGE 0x08 +#endif + #define EXT4_I(inode) (container_of(inode, struct ext4_inode_info, vfs_inode)) #define show_mballoc_flags(flags) __print_flags(flags, "|", \ @@ -71,7 +76,8 @@ struct extent_status; #define show_falloc_mode(mode) __print_flags(mode, "|", \ { FALLOC_FL_KEEP_SIZE, "KEEP_SIZE"}, \ { FALLOC_FL_PUNCH_HOLE, "PUNCH_HOLE"}, \ - { FALLOC_FL_NO_HIDE_STALE, "NO_HIDE_STALE"}) + { FALLOC_FL_NO_HIDE_STALE, "NO_HIDE_STALE"}, \ + { FALLOC_FL_COLLAPSE_RANGE, "COLLAPSE_RANGE"}) TRACE_EVENT(ext4_free_inode, @@ -2415,6 +2421,31 @@ TRACE_EVENT(ext4_es_shrink_exit, __entry->shrunk_nr, __entry->cache_cnt) ); +TRACE_EVENT(ext4_collapse_range, + TP_PROTO(struct inode *inode, loff_t offset, loff_t len), + + TP_ARGS(inode, offset, len), + + TP_STRUCT__entry( + __field(dev_t, dev) + __field(ino_t, ino) + __field(loff_t, offset) + __field(loff_t, len) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->offset = offset; + __entry->len = len; + ), + + TP_printk("dev %d,%d ino %lu offset %lld len %lld", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->offset, __entry->len) +); + #endif /* _TRACE_EXT4_H */ /* This part must be outside protection */ -- cgit v1.2.3-70-g09d2 From b8a8684502a0fc852afa0056c6bb2a9273f6fcc0 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Tue, 18 Mar 2014 18:05:35 -0400 Subject: ext4: Introduce FALLOC_FL_ZERO_RANGE flag for fallocate Introduce new FALLOC_FL_ZERO_RANGE flag for fallocate. This has the same functionality as xfs ioctl XFS_IOC_ZERO_RANGE. It can be used to convert a range of file to zeros preferably without issuing data IO. Blocks should be preallocated for the regions that span holes in the file, and the entire range is preferable converted to unwritten extents This can be also used to preallocate blocks past EOF in the same way as with fallocate. Flag FALLOC_FL_KEEP_SIZE which should cause the inode size to remain the same. Also add appropriate tracepoints. Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 2 + fs/ext4/extents.c | 273 +++++++++++++++++++++++++++++++++++++++++--- fs/ext4/inode.c | 17 ++- include/trace/events/ext4.h | 68 +++++------ 4 files changed, 307 insertions(+), 53 deletions(-) (limited to 'include') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index beec42750a8..1b3cbf8cacf 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -568,6 +568,8 @@ enum { #define EXT4_GET_BLOCKS_NO_LOCK 0x0100 /* Do not put hole in extent cache */ #define EXT4_GET_BLOCKS_NO_PUT_HOLE 0x0200 + /* Convert written extents to unwritten */ +#define EXT4_GET_BLOCKS_CONVERT_UNWRITTEN 0x0400 /* * The bit position of these flags must not overlap with any of the diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 2db2d77769a..464e95da716 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3602,6 +3602,8 @@ out: * b> Splits in two extents: Write is happening at either end of the extent * c> Splits in three extents: Somone is writing in middle of the extent * + * This works the same way in the case of initialized -> unwritten conversion. + * * One of more index blocks maybe needed if the extent tree grow after * the uninitialized extent split. To prevent ENOSPC occur at the IO * complete, we need to split the uninitialized extent before DIO submit @@ -3612,7 +3614,7 @@ out: * * Returns the size of uninitialized extent to be written on success. */ -static int ext4_split_unwritten_extents(handle_t *handle, +static int ext4_split_convert_extents(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, struct ext4_ext_path *path, @@ -3624,9 +3626,9 @@ static int ext4_split_unwritten_extents(handle_t *handle, unsigned int ee_len; int split_flag = 0, depth; - ext_debug("ext4_split_unwritten_extents: inode %lu, logical" - "block %llu, max_blocks %u\n", inode->i_ino, - (unsigned long long)map->m_lblk, map->m_len); + ext_debug("%s: inode %lu, logical block %llu, max_blocks %u\n", + __func__, inode->i_ino, + (unsigned long long)map->m_lblk, map->m_len); eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >> inode->i_sb->s_blocksize_bits; @@ -3641,14 +3643,73 @@ static int ext4_split_unwritten_extents(handle_t *handle, ee_block = le32_to_cpu(ex->ee_block); ee_len = ext4_ext_get_actual_len(ex); - split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0; - split_flag |= EXT4_EXT_MARK_UNINIT2; - if (flags & EXT4_GET_BLOCKS_CONVERT) - split_flag |= EXT4_EXT_DATA_VALID2; + /* Convert to unwritten */ + if (flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN) { + split_flag |= EXT4_EXT_DATA_VALID1; + /* Convert to initialized */ + } else if (flags & EXT4_GET_BLOCKS_CONVERT) { + split_flag |= ee_block + ee_len <= eof_block ? + EXT4_EXT_MAY_ZEROOUT : 0; + split_flag |= (EXT4_EXT_MARK_UNINIT2 | EXT4_EXT_DATA_VALID2); + } flags |= EXT4_GET_BLOCKS_PRE_IO; return ext4_split_extent(handle, inode, path, map, split_flag, flags); } +static int ext4_convert_initialized_extents(handle_t *handle, + struct inode *inode, + struct ext4_map_blocks *map, + struct ext4_ext_path *path) +{ + struct ext4_extent *ex; + ext4_lblk_t ee_block; + unsigned int ee_len; + int depth; + int err = 0; + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + + ext_debug("%s: inode %lu, logical" + "block %llu, max_blocks %u\n", __func__, inode->i_ino, + (unsigned long long)ee_block, ee_len); + + if (ee_block != map->m_lblk || ee_len > map->m_len) { + err = ext4_split_convert_extents(handle, inode, map, path, + EXT4_GET_BLOCKS_CONVERT_UNWRITTEN); + if (err < 0) + goto out; + ext4_ext_drop_refs(path); + path = ext4_ext_find_extent(inode, map->m_lblk, path, 0); + if (IS_ERR(path)) { + err = PTR_ERR(path); + goto out; + } + depth = ext_depth(inode); + ex = path[depth].p_ext; + } + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto out; + /* first mark the extent as uninitialized */ + ext4_ext_mark_uninitialized(ex); + + /* note: ext4_ext_correct_indexes() isn't needed here because + * borders are not changed + */ + ext4_ext_try_to_merge(handle, inode, path, ex); + + /* Mark modified extent as dirty */ + err = ext4_ext_dirty(handle, inode, path + path->p_depth); +out: + ext4_ext_show_leaf(inode, path); + return err; +} + + static int ext4_convert_unwritten_extents_endio(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, @@ -3682,8 +3743,8 @@ static int ext4_convert_unwritten_extents_endio(handle_t *handle, inode->i_ino, (unsigned long long)ee_block, ee_len, (unsigned long long)map->m_lblk, map->m_len); #endif - err = ext4_split_unwritten_extents(handle, inode, map, path, - EXT4_GET_BLOCKS_CONVERT); + err = ext4_split_convert_extents(handle, inode, map, path, + EXT4_GET_BLOCKS_CONVERT); if (err < 0) goto out; ext4_ext_drop_refs(path); @@ -3883,6 +3944,38 @@ get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start, return allocated_clusters; } +static int +ext4_ext_convert_initialized_extent(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, + struct ext4_ext_path *path, int flags, + unsigned int allocated, ext4_fsblk_t newblock) +{ + int ret = 0; + int err = 0; + + /* + * Make sure that the extent is no bigger than we support with + * uninitialized extent + */ + if (map->m_len > EXT_UNINIT_MAX_LEN) + map->m_len = EXT_UNINIT_MAX_LEN / 2; + + ret = ext4_convert_initialized_extents(handle, inode, map, + path); + if (ret >= 0) { + ext4_update_inode_fsync_trans(handle, inode, 1); + err = check_eofblocks_fl(handle, inode, map->m_lblk, + path, map->m_len); + } else + err = ret; + map->m_flags |= EXT4_MAP_UNWRITTEN; + if (allocated > map->m_len) + allocated = map->m_len; + map->m_len = allocated; + + return err ? err : allocated; +} + static int ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, @@ -3910,8 +4003,8 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, /* get_block() before submit the IO, split the extent */ if ((flags & EXT4_GET_BLOCKS_PRE_IO)) { - ret = ext4_split_unwritten_extents(handle, inode, map, - path, flags); + ret = ext4_split_convert_extents(handle, inode, map, + path, flags | EXT4_GET_BLOCKS_CONVERT); if (ret <= 0) goto out; /* @@ -4199,6 +4292,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext4_fsblk_t ee_start = ext4_ext_pblock(ex); unsigned short ee_len; + /* * Uninitialized extents are treated as holes, except that * we split out initialized portions during a write. @@ -4215,7 +4309,17 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk, ee_block, ee_len, newblock); - if (!ext4_ext_is_uninitialized(ex)) + /* + * If the extent is initialized check whether the + * caller wants to convert it to unwritten. + */ + if ((!ext4_ext_is_uninitialized(ex)) && + (flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN)) { + allocated = ext4_ext_convert_initialized_extent( + handle, inode, map, path, flags, + allocated, newblock); + goto out2; + } else if (!ext4_ext_is_uninitialized(ex)) goto out; ret = ext4_ext_handle_uninitialized_extents( @@ -4604,6 +4708,144 @@ retry: return ret > 0 ? ret2 : ret; } +static long ext4_zero_range(struct file *file, loff_t offset, + loff_t len, int mode) +{ + struct inode *inode = file_inode(file); + handle_t *handle = NULL; + unsigned int max_blocks; + loff_t new_size = 0; + int ret = 0; + int flags; + int partial; + loff_t start, end; + ext4_lblk_t lblk; + struct address_space *mapping = inode->i_mapping; + unsigned int blkbits = inode->i_blkbits; + + trace_ext4_zero_range(inode, offset, len, mode); + + /* + * Write out all dirty pages to avoid race conditions + * Then release them. + */ + if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { + ret = filemap_write_and_wait_range(mapping, offset, + offset + len - 1); + if (ret) + return ret; + } + + /* + * Round up offset. This is not fallocate, we neet to zero out + * blocks, so convert interior block aligned part of the range to + * unwritten and possibly manually zero out unaligned parts of the + * range. + */ + start = round_up(offset, 1 << blkbits); + end = round_down((offset + len), 1 << blkbits); + + if (start < offset || end > offset + len) + return -EINVAL; + partial = (offset + len) & ((1 << blkbits) - 1); + + lblk = start >> blkbits; + max_blocks = (end >> blkbits); + if (max_blocks < lblk) + max_blocks = 0; + else + max_blocks -= lblk; + + flags = EXT4_GET_BLOCKS_CREATE_UNINIT_EXT | + EXT4_GET_BLOCKS_CONVERT_UNWRITTEN; + if (mode & FALLOC_FL_KEEP_SIZE) + flags |= EXT4_GET_BLOCKS_KEEP_SIZE; + + mutex_lock(&inode->i_mutex); + + /* + * Indirect files do not support unwritten extnets + */ + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { + ret = -EOPNOTSUPP; + goto out_mutex; + } + + if (!(mode & FALLOC_FL_KEEP_SIZE) && + offset + len > i_size_read(inode)) { + new_size = offset + len; + ret = inode_newsize_ok(inode, new_size); + if (ret) + goto out_mutex; + /* + * If we have a partial block after EOF we have to allocate + * the entire block. + */ + if (partial) + max_blocks += 1; + } + + if (max_blocks > 0) { + + /* Now release the pages and zero block aligned part of pages*/ + truncate_pagecache_range(inode, start, end - 1); + + /* Wait all existing dio workers, newcomers will block on i_mutex */ + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + + /* + * Remove entire range from the extent status tree. + */ + ret = ext4_es_remove_extent(inode, lblk, max_blocks); + if (ret) + goto out_dio; + + ret = ext4_alloc_file_blocks(file, lblk, max_blocks, flags, + mode); + if (ret) + goto out_dio; + } + + handle = ext4_journal_start(inode, EXT4_HT_MISC, 4); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + ext4_std_error(inode->i_sb, ret); + goto out_dio; + } + + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + + if (!ret && new_size) { + if (new_size > i_size_read(inode)) + i_size_write(inode, new_size); + if (new_size > EXT4_I(inode)->i_disksize) + ext4_update_i_disksize(inode, new_size); + } else if (!ret && !new_size) { + /* + * Mark that we allocate beyond EOF so the subsequent truncate + * can proceed even if the new size is the same as i_size. + */ + if ((offset + len) > i_size_read(inode)) + ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS); + } + + ext4_mark_inode_dirty(handle, inode); + + /* Zero out partial block at the edges of the range */ + ret = ext4_zero_partial_blocks(handle, inode, offset, len); + + if (file->f_flags & O_SYNC) + ext4_handle_sync(handle); + + ext4_journal_stop(handle); +out_dio: + ext4_inode_resume_unlocked_dio(inode); +out_mutex: + mutex_unlock(&inode->i_mutex); + return ret; +} + /* * preallocate space for a file. This implements ext4's fallocate file * operation, which gets called from sys_fallocate system call. @@ -4625,7 +4867,7 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) /* Return error if mode is not supported */ if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | - FALLOC_FL_COLLAPSE_RANGE)) + FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE)) return -EOPNOTSUPP; if (mode & FALLOC_FL_PUNCH_HOLE) @@ -4645,6 +4887,9 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) return -EOPNOTSUPP; + if (mode & FALLOC_FL_ZERO_RANGE) + return ext4_zero_range(file, offset, len, mode); + trace_ext4_fallocate_enter(inode, offset, len, mode); lblk = offset >> blkbits; /* diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index ab3e8357929..7cc24555eca 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -503,6 +503,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, { struct extent_status es; int retval; + int ret = 0; #ifdef ES_AGGRESSIVE_TEST struct ext4_map_blocks orig_map; @@ -558,7 +559,6 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, EXT4_GET_BLOCKS_KEEP_SIZE); } if (retval > 0) { - int ret; unsigned int status; if (unlikely(retval != map->m_len)) { @@ -585,7 +585,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, found: if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret = check_block_validity(inode, map); + ret = check_block_validity(inode, map); if (ret != 0) return ret; } @@ -602,7 +602,13 @@ found: * with buffer head unmapped. */ if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) - return retval; + /* + * 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; /* * Here we clear m_flags because after allocating an new extent, @@ -658,7 +664,6 @@ found: ext4_clear_inode_state(inode, EXT4_STATE_DELALLOC_RESERVED); if (retval > 0) { - int ret; unsigned int status; if (unlikely(retval != map->m_len)) { @@ -693,7 +698,7 @@ found: has_zeroout: up_write((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret = check_block_validity(inode, map); + ret = check_block_validity(inode, map); if (ret != 0) return ret; } @@ -3507,7 +3512,7 @@ int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length) if (!S_ISREG(inode->i_mode)) return -EOPNOTSUPP; - trace_ext4_punch_hole(inode, offset, length); + trace_ext4_punch_hole(inode, offset, length, 0); /* * Write out all dirty pages to avoid race conditions diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index e9d7ee77d3a..010ea89eeb0 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -21,6 +21,10 @@ struct extent_status; #define FALLOC_FL_COLLAPSE_RANGE 0x08 #endif +#ifndef FALLOC_FL_ZERO_RANGE +#define FALLOC_FL_ZERO_RANGE 0x10 +#endif + #define EXT4_I(inode) (container_of(inode, struct ext4_inode_info, vfs_inode)) #define show_mballoc_flags(flags) __print_flags(flags, "|", \ @@ -77,7 +81,8 @@ struct extent_status; { FALLOC_FL_KEEP_SIZE, "KEEP_SIZE"}, \ { FALLOC_FL_PUNCH_HOLE, "PUNCH_HOLE"}, \ { FALLOC_FL_NO_HIDE_STALE, "NO_HIDE_STALE"}, \ - { FALLOC_FL_COLLAPSE_RANGE, "COLLAPSE_RANGE"}) + { FALLOC_FL_COLLAPSE_RANGE, "COLLAPSE_RANGE"}, \ + { FALLOC_FL_ZERO_RANGE, "ZERO_RANGE"}) TRACE_EVENT(ext4_free_inode, @@ -1339,7 +1344,7 @@ TRACE_EVENT(ext4_direct_IO_exit, __entry->rw, __entry->ret) ); -TRACE_EVENT(ext4_fallocate_enter, +DECLARE_EVENT_CLASS(ext4__fallocate_mode, TP_PROTO(struct inode *inode, loff_t offset, loff_t len, int mode), TP_ARGS(inode, offset, len, mode), @@ -1347,23 +1352,45 @@ TRACE_EVENT(ext4_fallocate_enter, TP_STRUCT__entry( __field( dev_t, dev ) __field( ino_t, ino ) - __field( loff_t, pos ) - __field( loff_t, len ) + __field( loff_t, offset ) + __field( loff_t, len ) __field( int, mode ) ), TP_fast_assign( __entry->dev = inode->i_sb->s_dev; __entry->ino = inode->i_ino; - __entry->pos = offset; + __entry->offset = offset; __entry->len = len; __entry->mode = mode; ), - TP_printk("dev %d,%d ino %lu pos %lld len %lld mode %s", + TP_printk("dev %d,%d ino %lu offset %lld len %lld mode %s", MAJOR(__entry->dev), MINOR(__entry->dev), - (unsigned long) __entry->ino, __entry->pos, - __entry->len, show_falloc_mode(__entry->mode)) + (unsigned long) __entry->ino, + __entry->offset, __entry->len, + show_falloc_mode(__entry->mode)) +); + +DEFINE_EVENT(ext4__fallocate_mode, ext4_fallocate_enter, + + TP_PROTO(struct inode *inode, loff_t offset, loff_t len, int mode), + + TP_ARGS(inode, offset, len, mode) +); + +DEFINE_EVENT(ext4__fallocate_mode, ext4_punch_hole, + + TP_PROTO(struct inode *inode, loff_t offset, loff_t len, int mode), + + TP_ARGS(inode, offset, len, mode) +); + +DEFINE_EVENT(ext4__fallocate_mode, ext4_zero_range, + + TP_PROTO(struct inode *inode, loff_t offset, loff_t len, int mode), + + TP_ARGS(inode, offset, len, mode) ); TRACE_EVENT(ext4_fallocate_exit, @@ -1395,31 +1422,6 @@ TRACE_EVENT(ext4_fallocate_exit, __entry->ret) ); -TRACE_EVENT(ext4_punch_hole, - TP_PROTO(struct inode *inode, loff_t offset, loff_t len), - - TP_ARGS(inode, offset, len), - - TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) - __field( loff_t, offset ) - __field( loff_t, len ) - ), - - TP_fast_assign( - __entry->dev = inode->i_sb->s_dev; - __entry->ino = inode->i_ino; - __entry->offset = offset; - __entry->len = len; - ), - - TP_printk("dev %d,%d ino %lu offset %lld len %lld", - MAJOR(__entry->dev), MINOR(__entry->dev), - (unsigned long) __entry->ino, - __entry->offset, __entry->len) -); - TRACE_EVENT(ext4_unlink_enter, TP_PROTO(struct inode *parent, struct dentry *dentry), -- cgit v1.2.3-70-g09d2 From 3e037e5211252902a188a6a11aecd247409d0229 Mon Sep 17 00:00:00 2001 From: T Makphaibulchoke Date: Tue, 18 Mar 2014 19:19:41 -0400 Subject: fs/mbcache.c: change block and index hash chain to hlist_bl_node This patch changes each mb_cache's both block and index hash chains to use a hlist_bl_node, which contains a built-in lock. This is the first step in decoupling of locks serializing accesses to mb_cache global data and each mb_cache_entry local data. Signed-off-by: T. Makphaibulchoke Signed-off-by: "Theodore Ts'o" --- fs/mbcache.c | 117 ++++++++++++++++++++++++++++++++---------------- include/linux/mbcache.h | 12 ++--- 2 files changed, 85 insertions(+), 44 deletions(-) (limited to 'include') diff --git a/fs/mbcache.c b/fs/mbcache.c index e519e45bf67..55db0daaca7 100644 --- a/fs/mbcache.c +++ b/fs/mbcache.c @@ -34,9 +34,9 @@ #include #include #include -#include +#include #include - +#include #ifdef MB_CACHE_DEBUG # define mb_debug(f...) do { \ @@ -87,21 +87,38 @@ static LIST_HEAD(mb_cache_lru_list); static DEFINE_SPINLOCK(mb_cache_spinlock); static inline int -__mb_cache_entry_is_hashed(struct mb_cache_entry *ce) +__mb_cache_entry_is_block_hashed(struct mb_cache_entry *ce) { - return !list_empty(&ce->e_block_list); + return !hlist_bl_unhashed(&ce->e_block_list); } -static void -__mb_cache_entry_unhash(struct mb_cache_entry *ce) +static inline void +__mb_cache_entry_unhash_block(struct mb_cache_entry *ce) { - if (__mb_cache_entry_is_hashed(ce)) { - list_del_init(&ce->e_block_list); - list_del(&ce->e_index.o_list); - } + if (__mb_cache_entry_is_block_hashed(ce)) + hlist_bl_del_init(&ce->e_block_list); +} + +static inline int +__mb_cache_entry_is_index_hashed(struct mb_cache_entry *ce) +{ + return !hlist_bl_unhashed(&ce->e_index.o_list); } +static inline void +__mb_cache_entry_unhash_index(struct mb_cache_entry *ce) +{ + if (__mb_cache_entry_is_index_hashed(ce)) + hlist_bl_del_init(&ce->e_index.o_list); +} + +static inline void +__mb_cache_entry_unhash(struct mb_cache_entry *ce) +{ + __mb_cache_entry_unhash_index(ce); + __mb_cache_entry_unhash_block(ce); +} static void __mb_cache_entry_forget(struct mb_cache_entry *ce, gfp_t gfp_mask) @@ -125,7 +142,7 @@ __mb_cache_entry_release_unlock(struct mb_cache_entry *ce) ce->e_used -= MB_CACHE_WRITER; ce->e_used--; if (!(ce->e_used || ce->e_queued)) { - if (!__mb_cache_entry_is_hashed(ce)) + if (!__mb_cache_entry_is_block_hashed(ce)) goto forget; mb_assert(list_empty(&ce->e_lru_list)); list_add_tail(&ce->e_lru_list, &mb_cache_lru_list); @@ -221,18 +238,18 @@ mb_cache_create(const char *name, int bucket_bits) cache->c_name = name; atomic_set(&cache->c_entry_count, 0); cache->c_bucket_bits = bucket_bits; - cache->c_block_hash = kmalloc(bucket_count * sizeof(struct list_head), - GFP_KERNEL); + cache->c_block_hash = kmalloc(bucket_count * + sizeof(struct hlist_bl_head), GFP_KERNEL); if (!cache->c_block_hash) goto fail; for (n=0; nc_block_hash[n]); - cache->c_index_hash = kmalloc(bucket_count * sizeof(struct list_head), - GFP_KERNEL); + INIT_HLIST_BL_HEAD(&cache->c_block_hash[n]); + cache->c_index_hash = kmalloc(bucket_count * + sizeof(struct hlist_bl_head), GFP_KERNEL); if (!cache->c_index_hash) goto fail; for (n=0; nc_index_hash[n]); + INIT_HLIST_BL_HEAD(&cache->c_index_hash[n]); cache->c_entry_cache = kmem_cache_create(name, sizeof(struct mb_cache_entry), 0, SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD, NULL); @@ -364,10 +381,13 @@ mb_cache_entry_alloc(struct mb_cache *cache, gfp_t gfp_flags) return NULL; atomic_inc(&cache->c_entry_count); INIT_LIST_HEAD(&ce->e_lru_list); - INIT_LIST_HEAD(&ce->e_block_list); + INIT_HLIST_BL_NODE(&ce->e_block_list); + INIT_HLIST_BL_NODE(&ce->e_index.o_list); ce->e_cache = cache; ce->e_queued = 0; } + ce->e_block_hash_p = &cache->c_block_hash[0]; + ce->e_index_hash_p = &cache->c_index_hash[0]; ce->e_used = 1 + MB_CACHE_WRITER; return ce; } @@ -393,25 +413,32 @@ mb_cache_entry_insert(struct mb_cache_entry *ce, struct block_device *bdev, { struct mb_cache *cache = ce->e_cache; unsigned int bucket; - struct list_head *l; + struct hlist_bl_node *l; int error = -EBUSY; + struct hlist_bl_head *block_hash_p; + struct hlist_bl_head *index_hash_p; + struct mb_cache_entry *lce; + mb_assert(ce); bucket = hash_long((unsigned long)bdev + (block & 0xffffffff), cache->c_bucket_bits); + block_hash_p = &cache->c_block_hash[bucket]; spin_lock(&mb_cache_spinlock); - list_for_each_prev(l, &cache->c_block_hash[bucket]) { - struct mb_cache_entry *ce = - list_entry(l, struct mb_cache_entry, e_block_list); - if (ce->e_bdev == bdev && ce->e_block == block) + hlist_bl_for_each_entry(lce, l, block_hash_p, e_block_list) { + if (lce->e_bdev == bdev && lce->e_block == block) goto out; } + mb_assert(!__mb_cache_entry_is_block_hashed(ce)); __mb_cache_entry_unhash(ce); ce->e_bdev = bdev; ce->e_block = block; - list_add(&ce->e_block_list, &cache->c_block_hash[bucket]); + ce->e_block_hash_p = block_hash_p; ce->e_index.o_key = key; bucket = hash_long(key, cache->c_bucket_bits); - list_add(&ce->e_index.o_list, &cache->c_index_hash[bucket]); + index_hash_p = &cache->c_index_hash[bucket]; + ce->e_index_hash_p = index_hash_p; + hlist_bl_add_head(&ce->e_index.o_list, index_hash_p); + hlist_bl_add_head(&ce->e_block_list, block_hash_p); error = 0; out: spin_unlock(&mb_cache_spinlock); @@ -463,14 +490,16 @@ mb_cache_entry_get(struct mb_cache *cache, struct block_device *bdev, sector_t block) { unsigned int bucket; - struct list_head *l; + struct hlist_bl_node *l; struct mb_cache_entry *ce; + struct hlist_bl_head *block_hash_p; bucket = hash_long((unsigned long)bdev + (block & 0xffffffff), cache->c_bucket_bits); + block_hash_p = &cache->c_block_hash[bucket]; spin_lock(&mb_cache_spinlock); - list_for_each(l, &cache->c_block_hash[bucket]) { - ce = list_entry(l, struct mb_cache_entry, e_block_list); + hlist_bl_for_each_entry(ce, l, block_hash_p, e_block_list) { + mb_assert(ce->e_block_hash_p == block_hash_p); if (ce->e_bdev == bdev && ce->e_block == block) { DEFINE_WAIT(wait); @@ -489,7 +518,7 @@ mb_cache_entry_get(struct mb_cache *cache, struct block_device *bdev, finish_wait(&mb_cache_queue, &wait); ce->e_used += 1 + MB_CACHE_WRITER; - if (!__mb_cache_entry_is_hashed(ce)) { + if (!__mb_cache_entry_is_block_hashed(ce)) { __mb_cache_entry_release_unlock(ce); return NULL; } @@ -506,12 +535,14 @@ cleanup: #if !defined(MB_CACHE_INDEXES_COUNT) || (MB_CACHE_INDEXES_COUNT > 0) static struct mb_cache_entry * -__mb_cache_entry_find(struct list_head *l, struct list_head *head, +__mb_cache_entry_find(struct hlist_bl_node *l, struct hlist_bl_head *head, struct block_device *bdev, unsigned int key) { - while (l != head) { + while (l != NULL) { struct mb_cache_entry *ce = - list_entry(l, struct mb_cache_entry, e_index.o_list); + hlist_bl_entry(l, struct mb_cache_entry, + e_index.o_list); + mb_assert(ce->e_index_hash_p == head); if (ce->e_bdev == bdev && ce->e_index.o_key == key) { DEFINE_WAIT(wait); @@ -532,7 +563,7 @@ __mb_cache_entry_find(struct list_head *l, struct list_head *head, } finish_wait(&mb_cache_queue, &wait); - if (!__mb_cache_entry_is_hashed(ce)) { + if (!__mb_cache_entry_is_block_hashed(ce)) { __mb_cache_entry_release_unlock(ce); spin_lock(&mb_cache_spinlock); return ERR_PTR(-EAGAIN); @@ -562,12 +593,16 @@ mb_cache_entry_find_first(struct mb_cache *cache, struct block_device *bdev, unsigned int key) { unsigned int bucket = hash_long(key, cache->c_bucket_bits); - struct list_head *l; - struct mb_cache_entry *ce; + struct hlist_bl_node *l; + struct mb_cache_entry *ce = NULL; + struct hlist_bl_head *index_hash_p; + index_hash_p = &cache->c_index_hash[bucket]; spin_lock(&mb_cache_spinlock); - l = cache->c_index_hash[bucket].next; - ce = __mb_cache_entry_find(l, &cache->c_index_hash[bucket], bdev, key); + if (!hlist_bl_empty(index_hash_p)) { + l = hlist_bl_first(index_hash_p); + ce = __mb_cache_entry_find(l, index_hash_p, bdev, key); + } spin_unlock(&mb_cache_spinlock); return ce; } @@ -597,12 +632,16 @@ mb_cache_entry_find_next(struct mb_cache_entry *prev, { struct mb_cache *cache = prev->e_cache; unsigned int bucket = hash_long(key, cache->c_bucket_bits); - struct list_head *l; + struct hlist_bl_node *l; struct mb_cache_entry *ce; + struct hlist_bl_head *index_hash_p; + index_hash_p = &cache->c_index_hash[bucket]; + mb_assert(prev->e_index_hash_p == index_hash_p); spin_lock(&mb_cache_spinlock); + mb_assert(!hlist_bl_empty(index_hash_p)); l = prev->e_index.o_list.next; - ce = __mb_cache_entry_find(l, &cache->c_index_hash[bucket], bdev, key); + ce = __mb_cache_entry_find(l, index_hash_p, bdev, key); __mb_cache_entry_release_unlock(prev); return ce; } diff --git a/include/linux/mbcache.h b/include/linux/mbcache.h index 5525d370701..6a392e7a723 100644 --- a/include/linux/mbcache.h +++ b/include/linux/mbcache.h @@ -3,19 +3,21 @@ (C) 2001 by Andreas Gruenbacher, */ - struct mb_cache_entry { struct list_head e_lru_list; struct mb_cache *e_cache; unsigned short e_used; unsigned short e_queued; + atomic_t e_refcnt; struct block_device *e_bdev; sector_t e_block; - struct list_head e_block_list; + struct hlist_bl_node e_block_list; struct { - struct list_head o_list; + struct hlist_bl_node o_list; unsigned int o_key; } e_index; + struct hlist_bl_head *e_block_hash_p; + struct hlist_bl_head *e_index_hash_p; }; struct mb_cache { @@ -25,8 +27,8 @@ struct mb_cache { int c_max_entries; int c_bucket_bits; struct kmem_cache *c_entry_cache; - struct list_head *c_block_hash; - struct list_head *c_index_hash; + struct hlist_bl_head *c_block_hash; + struct hlist_bl_head *c_index_hash; }; /* Functions on caches */ -- cgit v1.2.3-70-g09d2 From 5f16f3225b06242a9ee876f07c1c9b6ed36a22b6 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 24 Mar 2014 14:43:12 -0400 Subject: ext4: atomically set inode->i_flags in ext4_set_inode_flags() Use cmpxchg() to atomically set i_flags instead of clearing out the S_IMMUTABLE, S_APPEND, etc. flags and then setting them from the EXT4_IMMUTABLE_FL, EXT4_APPEND_FL flags, since this opens up a race where an immutable file has the immutable flag cleared for a brief window of time. Reported-by: John Sullivan Signed-off-by: "Theodore Ts'o" Cc: stable@kernel.org --- fs/ext4/inode.c | 14 ++++++++------ fs/inode.c | 31 +++++++++++++++++++++++++++++++ include/linux/fs.h | 3 +++ 3 files changed, 42 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index b5e182acf9b..df067c3c6c9 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3938,18 +3938,20 @@ int ext4_get_inode_loc(struct inode *inode, struct ext4_iloc *iloc) 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 */ diff --git a/fs/inode.c b/fs/inode.c index 4bcdad3c936..26f95ceb625 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1899,3 +1899,34 @@ void inode_dio_done(struct inode *inode) wake_up_bit(&inode->i_state, __I_DIO_WAKEUP); } EXPORT_SYMBOL(inode_dio_done); + +/* + * inode_set_flags - atomically set some inode flags + * + * Note: the caller should be holding i_mutex, or else be sure that + * they have exclusive access to the inode structure (i.e., while the + * inode is being instantiated). The reason for the cmpxchg() loop + * --- which wouldn't be necessary if all code paths which modify + * i_flags actually followed this rule, is that there is at least one + * code path which doesn't today --- for example, + * __generic_file_aio_write() calls file_remove_suid() without holding + * i_mutex --- so we use cmpxchg() out of an abundance of caution. + * + * In the long run, i_mutex is overkill, and we should probably look + * at using the i_lock spinlock to protect i_flags, and then make sure + * it is so documented in include/linux/fs.h and that all code follows + * the locking convention!! + */ +void inode_set_flags(struct inode *inode, unsigned int flags, + unsigned int mask) +{ + unsigned int old_flags, new_flags; + + WARN_ON_ONCE(flags & ~mask); + do { + old_flags = ACCESS_ONCE(inode->i_flags); + new_flags = (old_flags & ~mask) | flags; + } while (unlikely(cmpxchg(&inode->i_flags, old_flags, + new_flags) != old_flags)); +} +EXPORT_SYMBOL(inode_set_flags); diff --git a/include/linux/fs.h b/include/linux/fs.h index 60829565e55..5d1f6fa8dae 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2556,6 +2556,9 @@ static inline ssize_t blockdev_direct_IO(int rw, struct kiocb *iocb, void inode_dio_wait(struct inode *inode); void inode_dio_done(struct inode *inode); +extern void inode_set_flags(struct inode *inode, unsigned int flags, + unsigned int mask); + extern const struct file_operations generic_ro_fops; #define special_file(m) (S_ISCHR(m)||S_ISBLK(m)||S_ISFIFO(m)||S_ISSOCK(m)) -- cgit v1.2.3-70-g09d2