diff options
39 files changed, 2271 insertions, 1301 deletions
diff --git a/Documentation/filesystems/ext4.txt b/Documentation/filesystems/ext4.txt index 174eaff7ded..cec829bc729 100644 --- a/Documentation/filesystems/ext4.txt +++ b/Documentation/filesystems/ext4.txt @@ -58,13 +58,22 @@ Note: More extensive information for getting started with ext4 can be # mount -t ext4 /dev/hda1 /wherever - - When comparing performance with other filesystems, remember that - ext3/4 by default offers higher data integrity guarantees than most. - So when comparing with a metadata-only journalling filesystem, such - as ext3, use `mount -o data=writeback'. And you might as well use - `mount -o nobh' too along with it. Making the journal larger than - the mke2fs default often helps performance with metadata-intensive - workloads. + - When comparing performance with other filesystems, it's always + important to try multiple workloads; very often a subtle change in a + workload parameter can completely change the ranking of which + filesystems do well compared to others. When comparing versus ext3, + note that ext4 enables write barriers by default, while ext3 does + not enable write barriers by default. So it is useful to use + explicitly specify whether barriers are enabled or not when via the + '-o barriers=[0|1]' mount option for both ext3 and ext4 filesystems + for a fair comparison. When tuning ext3 for best benchmark numbers, + it is often worthwhile to try changing the data journaling mode; '-o + data=writeback,nobh' can be faster for some workloads. (Note + however that running mounted with data=writeback can potentially + leave stale data exposed in recently written files in case of an + unclean shutdown, which could be a security exposure in some + situations.) Configuring the filesystem with a large journal can + also be helpful for metadata-intensive workloads. 2. Features =========== @@ -74,7 +83,7 @@ Note: More extensive information for getting started with ext4 can be * ability to use filesystems > 16TB (e2fsprogs support not available yet) * extent format reduces metadata overhead (RAM, IO for access, transactions) * extent format more robust in face of on-disk corruption due to magics, -* internal redunancy in tree +* internal redundancy in tree * improved file allocation (multi-block alloc) * fix 32000 subdirectory limit * nsec timestamps for mtime, atime, ctime, create time @@ -116,10 +125,11 @@ grouping of bitmaps and inode tables. Some test results available here: When mounting an ext4 filesystem, the following option are accepted: (*) == default -extents (*) ext4 will use extents to address file data. The - file system will no longer be mountable by ext3. - -noextents ext4 will not use extents for newly created files +ro Mount filesystem read only. Note that ext4 will + replay the journal (and thus write to the + partition) even when mounted "read only". The + mount options "ro,noload" can be used to prevent + writes to the filesystem. journal_checksum Enable checksumming of the journal transactions. This will allow the recovery code in e2fsck and the @@ -134,17 +144,17 @@ journal_async_commit Commit block can be written to disk without waiting journal=update Update the ext4 file system's journal to the current format. -journal=inum When a journal already exists, this option is ignored. - Otherwise, it specifies the number of the inode which - will represent the ext4 file system's journal file. - journal_dev=devnum When the external journal device's major/minor numbers have changed, this option allows the user to specify the new journal location. The journal device is identified through its new major/minor numbers encoded in devnum. -noload Don't load the journal on mounting. +noload Don't load the journal on mounting. Note that + if the filesystem was not unmounted cleanly, + skipping the journal replay will lead to the + filesystem containing inconsistencies that can + lead to any number of problems. data=journal All data are committed into the journal prior to being written into the main file system. @@ -219,9 +229,12 @@ minixdf Make 'df' act like Minix. debug Extra debugging information is sent to syslog. -errors=remount-ro(*) Remount the filesystem read-only on an error. +errors=remount-ro Remount the filesystem read-only on an error. errors=continue Keep going on a filesystem error. errors=panic Panic and halt the machine if an error occurs. + (These mount options override the errors behavior + specified in the superblock, which can be configured + using tune2fs) data_err=ignore(*) Just print an error message if an error occurs in a file data buffer in ordered mode. @@ -261,6 +274,42 @@ delalloc (*) Deferring block allocation until write-out time. nodelalloc Disable delayed allocation. Blocks are allocation when data is copied from user to page cache. +max_batch_time=usec Maximum amount of time ext4 should wait for + additional filesystem operations to be batch + together with a synchronous write operation. + Since a synchronous write operation is going to + force a commit and then a wait for the I/O + complete, it doesn't cost much, and can be a + huge throughput win, we wait for a small amount + of time to see if any other transactions can + piggyback on the synchronous write. The + algorithm used is designed to automatically tune + for the speed of the disk, by measuring the + amount of time (on average) that it takes to + finish committing a transaction. Call this time + the "commit time". If the time that the + transactoin has been running is less than the + commit time, ext4 will try sleeping for the + commit time to see if other operations will join + the transaction. The commit time is capped by + the max_batch_time, which defaults to 15000us + (15ms). This optimization can be turned off + entirely by setting max_batch_time to 0. + +min_batch_time=usec This parameter sets the commit time (as + described above) to be at least min_batch_time. + It defaults to zero microseconds. Increasing + this parameter may improve the throughput of + multi-threaded, synchronous workloads on very + fast disks, at the cost of increasing latency. + +journal_ioprio=prio The I/O priority (from 0 to 7, where 0 is the + highest priorty) which should be used for I/O + operations submitted by kjournald2 during a + commit operation. This defaults to 3, which is + a slightly higher priority than the default I/O + priority. + Data Mode ========= There are 3 different data modes: diff --git a/block/Kconfig b/block/Kconfig index ac0956f7778..0cbb3b88b59 100644 --- a/block/Kconfig +++ b/block/Kconfig @@ -36,6 +36,12 @@ config LBD This option also enables support for single files larger than 2TB. + The ext4 filesystem requires that this feature be enabled in + order to support filesystems that have the huge_file feature + enabled. Otherwise, it will refuse to mount any filesystems + that use the huge_file feature, which is enabled by default + by mke2fs.ext4. The GFS2 filesystem also requires this feature. + If unsure, say N. config BLK_DEV_IO_TRACE diff --git a/fs/block_dev.c b/fs/block_dev.c index 8ebbfdf708c..ac7031f12ea 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1234,6 +1234,20 @@ static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) return blkdev_ioctl(bdev, mode, cmd, arg); } +/* + * Try to release a page associated with block device when the system + * is under memory pressure. + */ +static int blkdev_releasepage(struct page *page, gfp_t wait) +{ + struct super_block *super = BDEV_I(page->mapping->host)->bdev.bd_super; + + if (super && super->s_op->bdev_try_to_free_page) + return super->s_op->bdev_try_to_free_page(super, page, wait); + + return try_to_free_buffers(page); +} + static const struct address_space_operations def_blk_aops = { .readpage = blkdev_readpage, .writepage = blkdev_writepage, @@ -1241,6 +1255,7 @@ static const struct address_space_operations def_blk_aops = { .write_begin = blkdev_write_begin, .write_end = blkdev_write_end, .writepages = generic_writepages, + .releasepage = blkdev_releasepage, .direct_IO = blkdev_direct_IO, }; diff --git a/fs/ext3/hash.c b/fs/ext3/hash.c index c30e149fbd2..7d215b4d4f2 100644 --- a/fs/ext3/hash.c +++ b/fs/ext3/hash.c @@ -35,23 +35,71 @@ static void TEA_transform(__u32 buf[4], __u32 const in[]) /* The old legacy hash */ -static __u32 dx_hack_hash (const char *name, int len) +static __u32 dx_hack_hash_unsigned(const char *name, int len) { - __u32 hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9; + __u32 hash, hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9; + const unsigned char *ucp = (const unsigned char *) name; + + while (len--) { + hash = hash1 + (hash0 ^ (((int) *ucp++) * 7152373)); + + if (hash & 0x80000000) + hash -= 0x7fffffff; + hash1 = hash0; + hash0 = hash; + } + return hash0 << 1; +} + +static __u32 dx_hack_hash_signed(const char *name, int len) +{ + __u32 hash, hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9; + const signed char *scp = (const signed char *) name; + while (len--) { - __u32 hash = hash1 + (hash0 ^ (*name++ * 7152373)); + hash = hash1 + (hash0 ^ (((int) *scp++) * 7152373)); - if (hash & 0x80000000) hash -= 0x7fffffff; + if (hash & 0x80000000) + hash -= 0x7fffffff; hash1 = hash0; hash0 = hash; } - return (hash0 << 1); + return hash0 << 1; } -static void str2hashbuf(const char *msg, int len, __u32 *buf, int num) +static void str2hashbuf_signed(const char *msg, int len, __u32 *buf, int num) { __u32 pad, val; int i; + const signed char *scp = (const signed char *) msg; + + pad = (__u32)len | ((__u32)len << 8); + pad |= pad << 16; + + val = pad; + if (len > num*4) + len = num * 4; + for (i = 0; i < len; i++) { + if ((i % 4) == 0) + val = pad; + val = ((int) scp[i]) + (val << 8); + if ((i % 4) == 3) { + *buf++ = val; + val = pad; + num--; + } + } + if (--num >= 0) + *buf++ = val; + while (--num >= 0) + *buf++ = pad; +} + +static void str2hashbuf_unsigned(const char *msg, int len, __u32 *buf, int num) +{ + __u32 pad, val; + int i; + const unsigned char *ucp = (const unsigned char *) msg; pad = (__u32)len | ((__u32)len << 8); pad |= pad << 16; @@ -62,7 +110,7 @@ static void str2hashbuf(const char *msg, int len, __u32 *buf, int num) for (i=0; i < len; i++) { if ((i % 4) == 0) val = pad; - val = msg[i] + (val << 8); + val = ((int) ucp[i]) + (val << 8); if ((i % 4) == 3) { *buf++ = val; val = pad; @@ -95,6 +143,8 @@ int ext3fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo) const char *p; int i; __u32 in[8], buf[4]; + void (*str2hashbuf)(const char *, int, __u32 *, int) = + str2hashbuf_signed; /* Initialize the default seed for the hash checksum functions */ buf[0] = 0x67452301; @@ -113,13 +163,18 @@ int ext3fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo) } switch (hinfo->hash_version) { + case DX_HASH_LEGACY_UNSIGNED: + hash = dx_hack_hash_unsigned(name, len); + break; case DX_HASH_LEGACY: - hash = dx_hack_hash(name, len); + hash = dx_hack_hash_signed(name, len); break; + case DX_HASH_HALF_MD4_UNSIGNED: + str2hashbuf = str2hashbuf_unsigned; case DX_HASH_HALF_MD4: p = name; while (len > 0) { - str2hashbuf(p, len, in, 8); + (*str2hashbuf)(p, len, in, 8); half_md4_transform(buf, in); len -= 32; p += 32; @@ -127,10 +182,12 @@ int ext3fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo) minor_hash = buf[2]; hash = buf[1]; break; + case DX_HASH_TEA_UNSIGNED: + str2hashbuf = str2hashbuf_unsigned; case DX_HASH_TEA: p = name; while (len > 0) { - str2hashbuf(p, len, in, 4); + (*str2hashbuf)(p, len, in, 4); TEA_transform(buf, in); len -= 16; p += 16; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 8d6f965e502..69a3d19ca9f 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -364,6 +364,8 @@ dx_probe(struct qstr *entry, struct inode *dir, goto fail; } hinfo->hash_version = root->info.hash_version; + if (hinfo->hash_version <= DX_HASH_TEA) + hinfo->hash_version += EXT3_SB(dir->i_sb)->s_hash_unsigned; hinfo->seed = EXT3_SB(dir->i_sb)->s_hash_seed; if (entry) ext3fs_dirhash(entry->name, entry->len, hinfo); @@ -632,6 +634,9 @@ int ext3_htree_fill_tree(struct file *dir_file, __u32 start_hash, dir = dir_file->f_path.dentry->d_inode; if (!(EXT3_I(dir)->i_flags & EXT3_INDEX_FL)) { hinfo.hash_version = EXT3_SB(dir->i_sb)->s_def_hash_version; + if (hinfo.hash_version <= DX_HASH_TEA) + hinfo.hash_version += + EXT3_SB(dir->i_sb)->s_hash_unsigned; hinfo.seed = EXT3_SB(dir->i_sb)->s_hash_seed; count = htree_dirblock_to_tree(dir_file, dir, 0, &hinfo, start_hash, start_minor_hash); @@ -1152,9 +1157,9 @@ static struct ext3_dir_entry_2 *do_split(handle_t *handle, struct inode *dir, u32 hash2; struct dx_map_entry *map; char *data1 = (*bh)->b_data, *data2; - unsigned split, move, size, i; + unsigned split, move, size; struct ext3_dir_entry_2 *de = NULL, *de2; - int err = 0; + int err = 0, i; bh2 = ext3_append (handle, dir, &newblock, &err); if (!(bh2)) { @@ -1394,6 +1399,8 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, /* Initialize as for dx_probe */ hinfo.hash_version = root->info.hash_version; + if (hinfo.hash_version <= DX_HASH_TEA) + hinfo.hash_version += EXT3_SB(dir->i_sb)->s_hash_unsigned; hinfo.seed = EXT3_SB(dir->i_sb)->s_hash_seed; ext3fs_dirhash(name, namelen, &hinfo); frame = frames; diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 01c235bc205..5d047a030a7 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -683,6 +683,26 @@ static struct dentry *ext3_fh_to_parent(struct super_block *sb, struct fid *fid, ext3_nfs_get_inode); } +/* + * Try to release metadata pages (indirect blocks, directories) which are + * mapped via the block device. Since these pages could have journal heads + * which would prevent try_to_free_buffers() from freeing them, we must use + * jbd layer's try_to_free_buffers() function to release them. + */ +static int bdev_try_to_free_page(struct super_block *sb, struct page *page, + gfp_t wait) +{ + journal_t *journal = EXT3_SB(sb)->s_journal; + + WARN_ON(PageChecked(page)); + if (!page_has_buffers(page)) + return 0; + if (journal) + return journal_try_to_free_buffers(journal, page, + wait & ~__GFP_WAIT); + return try_to_free_buffers(page); +} + #ifdef CONFIG_QUOTA #define QTYPE2NAME(t) ((t)==USRQUOTA?"user":"group") #define QTYPE2MOPT(on, t) ((t)==USRQUOTA?((on)##USRJQUOTA):((on)##GRPJQUOTA)) @@ -749,6 +769,7 @@ static const struct super_operations ext3_sops = { .quota_read = ext3_quota_read, .quota_write = ext3_quota_write, #endif + .bdev_try_to_free_page = bdev_try_to_free_page, }; static const struct export_operations ext3_export_ops = { @@ -1750,6 +1771,18 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) for (i=0; i < 4; i++) sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]); sbi->s_def_hash_version = es->s_def_hash_version; + i = le32_to_cpu(es->s_flags); + if (i & EXT2_FLAGS_UNSIGNED_HASH) + sbi->s_hash_unsigned = 3; + else if ((i & EXT2_FLAGS_SIGNED_HASH) == 0) { +#ifdef __CHAR_UNSIGNED__ + es->s_flags |= cpu_to_le32(EXT2_FLAGS_UNSIGNED_HASH); + sbi->s_hash_unsigned = 3; +#else + es->s_flags |= cpu_to_le32(EXT2_FLAGS_SIGNED_HASH); +#endif + sb->s_dirt = 1; + } if (sbi->s_blocks_per_group > blocksize * 8) { printk (KERN_ERR diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 38b3acf5683..6bba06b09dd 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -20,6 +20,7 @@ #include "ext4.h" #include "ext4_jbd2.h" #include "group.h" +#include "mballoc.h" /* * balloc.c contains the blocks allocation and deallocation routines @@ -100,10 +101,10 @@ unsigned ext4_init_block_bitmap(struct super_block *sb, struct buffer_head *bh, * essentially implementing a per-group read-only flag. */ if (!ext4_group_desc_csum_verify(sbi, block_group, gdp)) { ext4_error(sb, __func__, - "Checksum bad for group %lu\n", block_group); - gdp->bg_free_blocks_count = 0; - gdp->bg_free_inodes_count = 0; - gdp->bg_itable_unused = 0; + "Checksum bad for group %u", block_group); + ext4_free_blks_set(sb, gdp, 0); + ext4_free_inodes_set(sb, gdp, 0); + ext4_itable_unused_set(sb, gdp, 0); memset(bh->b_data, 0xff, sb->s_blocksize); return 0; } @@ -205,15 +206,15 @@ struct ext4_group_desc * ext4_get_group_desc(struct super_block *sb, ext4_group_t block_group, struct buffer_head **bh) { - unsigned long group_desc; - unsigned long offset; + unsigned int group_desc; + unsigned int offset; struct ext4_group_desc *desc; struct ext4_sb_info *sbi = EXT4_SB(sb); if (block_group >= sbi->s_groups_count) { ext4_error(sb, "ext4_get_group_desc", "block_group >= groups_count - " - "block_group = %lu, groups_count = %lu", + "block_group = %u, groups_count = %u", block_group, sbi->s_groups_count); return NULL; @@ -225,7 +226,7 @@ struct ext4_group_desc * ext4_get_group_desc(struct super_block *sb, if (!sbi->s_group_desc[group_desc]) { ext4_error(sb, "ext4_get_group_desc", "Group descriptor not loaded - " - "block_group = %lu, group_desc = %lu, desc = %lu", + "block_group = %u, group_desc = %u, desc = %u", block_group, group_desc, offset); return NULL; } @@ -315,29 +316,50 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group) if (unlikely(!bh)) { ext4_error(sb, __func__, "Cannot read block bitmap - " - "block_group = %lu, block_bitmap = %llu", + "block_group = %u, block_bitmap = %llu", block_group, bitmap_blk); return NULL; } - if (buffer_uptodate(bh) && - !(desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) + + if (bitmap_uptodate(bh)) return bh; lock_buffer(bh); + if (bitmap_uptodate(bh)) { + unlock_buffer(bh); + return bh; + } spin_lock(sb_bgl_lock(EXT4_SB(sb), block_group)); if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) { ext4_init_block_bitmap(sb, bh, block_group, desc); + set_bitmap_uptodate(bh); set_buffer_uptodate(bh); - unlock_buffer(bh); spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group)); + unlock_buffer(bh); return bh; } spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group)); + if (buffer_uptodate(bh)) { + /* + * if not uninit if bh is uptodate, + * bitmap is also uptodate + */ + set_bitmap_uptodate(bh); + unlock_buffer(bh); + return bh; + } + /* + * submit the buffer_head for read. We can + * safely mark the bitmap as uptodate now. + * We do it here so the bitmap uptodate bit + * get set with buffer lock held. + */ + set_bitmap_uptodate(bh); if (bh_submit_read(bh) < 0) { put_bh(bh); ext4_error(sb, __func__, "Cannot read block bitmap - " - "block_group = %lu, block_bitmap = %llu", + "block_group = %u, block_bitmap = %llu", block_group, bitmap_blk); return NULL; } @@ -350,62 +372,44 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group) } /** - * ext4_free_blocks_sb() -- Free given blocks and update quota + * ext4_add_groupblocks() -- Add given blocks to an existing group * @handle: handle to this transaction * @sb: super block - * @block: start physcial block to free + * @block: start physcial block to add to the block group * @count: number of blocks to free - * @pdquot_freed_blocks: pointer to quota * - * XXX This function is only used by the on-line resizing code, which - * should probably be fixed up to call the mballoc variant. There - * this needs to be cleaned up later; in fact, I'm not convinced this - * is 100% correct in the face of the mballoc code. The online resizing - * code needs to be fixed up to more tightly (and correctly) interlock - * with the mballoc code. + * This marks the blocks as free in the bitmap. We ask the + * mballoc to reload the buddy after this by setting group + * EXT4_GROUP_INFO_NEED_INIT_BIT flag */ -void ext4_free_blocks_sb(handle_t *handle, struct super_block *sb, - ext4_fsblk_t block, unsigned long count, - unsigned long *pdquot_freed_blocks) +void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, + ext4_fsblk_t block, unsigned long count) { struct buffer_head *bitmap_bh = NULL; struct buffer_head *gd_bh; ext4_group_t block_group; ext4_grpblk_t bit; - unsigned long i; - unsigned long overflow; + unsigned int i; struct ext4_group_desc *desc; struct ext4_super_block *es; struct ext4_sb_info *sbi; - int err = 0, ret; - ext4_grpblk_t group_freed; + int err = 0, ret, blk_free_count; + ext4_grpblk_t blocks_freed; + struct ext4_group_info *grp; - *pdquot_freed_blocks = 0; sbi = EXT4_SB(sb); es = sbi->s_es; - if (block < le32_to_cpu(es->s_first_data_block) || - block + count < block || - block + count > ext4_blocks_count(es)) { - ext4_error(sb, "ext4_free_blocks", - "Freeing blocks not in datazone - " - "block = %llu, count = %lu", block, count); - goto error_return; - } - - ext4_debug("freeing block(s) %llu-%llu\n", block, block + count - 1); + ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1); -do_more: - overflow = 0; ext4_get_group_no_and_offset(sb, block, &block_group, &bit); + grp = ext4_get_group_info(sb, block_group); /* * Check to see if we are freeing blocks across a group * boundary. */ if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) { - overflow = bit + count - EXT4_BLOCKS_PER_GROUP(sb); - count -= overflow; + goto error_return; } - brelse(bitmap_bh); bitmap_bh = ext4_read_block_bitmap(sb, block_group); if (!bitmap_bh) goto error_return; @@ -418,18 +422,17 @@ do_more: in_range(block, ext4_inode_table(sb, desc), sbi->s_itb_per_group) || in_range(block + count - 1, ext4_inode_table(sb, desc), sbi->s_itb_per_group)) { - ext4_error(sb, "ext4_free_blocks", - "Freeing blocks in system zones - " + ext4_error(sb, __func__, + "Adding blocks in system zones - " "Block = %llu, count = %lu", block, count); goto error_return; } /* - * We are about to start releasing blocks in the bitmap, + * We are about to add blocks to the bitmap, * so we need undo access. */ - /* @@@ check errors */ BUFFER_TRACE(bitmap_bh, "getting undo access"); err = ext4_journal_get_undo_access(handle, bitmap_bh); if (err) @@ -444,107 +447,55 @@ do_more: err = ext4_journal_get_write_access(handle, gd_bh); if (err) goto error_return; - - jbd_lock_bh_state(bitmap_bh); - - for (i = 0, group_freed = 0; i < count; i++) { - /* - * An HJ special. This is expensive... - */ -#ifdef CONFIG_JBD2_DEBUG - jbd_unlock_bh_state(bitmap_bh); - { - struct buffer_head *debug_bh; - debug_bh = sb_find_get_block(sb, block + i); - if (debug_bh) { - BUFFER_TRACE(debug_bh, "Deleted!"); - if (!bh2jh(bitmap_bh)->b_committed_data) - BUFFER_TRACE(debug_bh, - "No commited data in bitmap"); - BUFFER_TRACE2(debug_bh, bitmap_bh, "bitmap"); - __brelse(debug_bh); - } - } - jbd_lock_bh_state(bitmap_bh); -#endif - if (need_resched()) { - jbd_unlock_bh_state(bitmap_bh); - cond_resched(); - jbd_lock_bh_state(bitmap_bh); - } - /* @@@ This prevents newly-allocated data from being - * freed and then reallocated within the same - * transaction. - * - * Ideally we would want to allow that to happen, but to - * do so requires making jbd2_journal_forget() capable of - * revoking the queued write of a data block, which - * implies blocking on the journal lock. *forget() - * cannot block due to truncate races. - * - * Eventually we can fix this by making jbd2_journal_forget() - * return a status indicating whether or not it was able - * to revoke the buffer. On successful revoke, it is - * safe not to set the allocation bit in the committed - * bitmap, because we know that there is no outstanding - * activity on the buffer any more and so it is safe to - * reallocate it. - */ - BUFFER_TRACE(bitmap_bh, "set in b_committed_data"); - J_ASSERT_BH(bitmap_bh, - bh2jh(bitmap_bh)->b_committed_data != NULL); - ext4_set_bit_atomic(sb_bgl_lock(sbi, block_group), bit + i, - bh2jh(bitmap_bh)->b_committed_data); - - /* - * We clear the bit in the bitmap after setting the committed - * data bit, because this is the reverse order to that which - * the allocator uses. - */ + /* + * make sure we don't allow a parallel init on other groups in the + * same buddy cache + */ + down_write(&grp->alloc_sem); + for (i = 0, blocks_freed = 0; i < count; i++) { BUFFER_TRACE(bitmap_bh, "clear bit"); if (!ext4_clear_bit_atomic(sb_bgl_lock(sbi, block_group), bit + i, bitmap_bh->b_data)) { - jbd_unlock_bh_state(bitmap_bh); ext4_error(sb, __func__, "bit already cleared for block %llu", (ext4_fsblk_t)(block + i)); - jbd_lock_bh_state(bitmap_bh); BUFFER_TRACE(bitmap_bh, "bit already cleared"); } else { - group_freed++; + blocks_freed++; } } - jbd_unlock_bh_state(bitmap_bh); - spin_lock(sb_bgl_lock(sbi, block_group)); - le16_add_cpu(&desc->bg_free_blocks_count, group_freed); + blk_free_count = blocks_freed + ext4_free_blks_count(sb, desc); + ext4_free_blks_set(sb, desc, blk_free_count); desc->bg_checksum = ext4_group_desc_csum(sbi, block_group, desc); spin_unlock(sb_bgl_lock(sbi, block_group)); - percpu_counter_add(&sbi->s_freeblocks_counter, count); + percpu_counter_add(&sbi->s_freeblocks_counter, blocks_freed); if (sbi->s_log_groups_per_flex) { ext4_group_t flex_group = ext4_flex_group(sbi, block_group); spin_lock(sb_bgl_lock(sbi, flex_group)); - sbi->s_flex_groups[flex_group].free_blocks += count; + sbi->s_flex_groups[flex_group].free_blocks += blocks_freed; spin_unlock(sb_bgl_lock(sbi, flex_group)); } + /* + * request to reload the buddy with the + * new bitmap information + */ + set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &(grp->bb_state)); + ext4_mb_update_group_info(grp, blocks_freed); + up_write(&grp->alloc_sem); /* We dirtied the bitmap block */ BUFFER_TRACE(bitmap_bh, "dirtied bitmap block"); - err = ext4_journal_dirty_metadata(handle, bitmap_bh); + err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); /* And the group descriptor block */ BUFFER_TRACE(gd_bh, "dirtied group descriptor block"); - ret = ext4_journal_dirty_metadata(handle, gd_bh); - if (!err) err = ret; - *pdquot_freed_blocks += group_freed; - - if (overflow && !err) { - block += count; - count = overflow; - goto do_more; |