b2 = (unsigned char *) bitmap;
for (i = 0; i < e4b->bd_sb->s_blocksize; i++) {
if (b1[i] != b2[i]) {
- printk("corruption in group %lu at byte %u(%u):"
- " %x in copy != %x on disk/prealloc\n",
- e4b->bd_group, i, i * 8, b1[i], b2[i]);
+ printk(KERN_ERR "corruption in group %lu "
+ "at byte %u(%u): %x in copy != %x "
+ "on disk/prealloc\n",
+ e4b->bd_group, i, i * 8, b1[i], b2[i]);
BUG();
}
}
void *buddy;
void *buddy2;
- if (!test_opt(sb, MBALLOC))
- return 0;
-
{
static int mb_check_counter;
if (mb_check_counter++ % 100 != 0)
if (bh[i] == NULL)
goto out;
- if (bh_uptodate_or_lock(bh[i]))
+ if (buffer_uptodate(bh[i]) &&
+ !(desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))
continue;
+ lock_buffer(bh[i]);
+ spin_lock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
ext4_init_block_bitmap(sb, bh[i],
first_group + i, desc);
set_buffer_uptodate(bh[i]);
unlock_buffer(bh[i]);
+ spin_unlock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
continue;
}
+ spin_unlock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
get_bh(bh[i]);
bh[i]->b_end_io = end_buffer_read_sync;
submit_bh(READ, bh[i]);
}
}
-static int mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
+static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
int first, int count)
{
int block = 0;
blocknr += block;
blocknr +=
le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block);
-
+ ext4_unlock_group(sb, e4b->bd_group);
ext4_error(sb, __func__, "double-free of inode"
" %lu's block %llu(bit %u in group %lu)\n",
inode ? inode->i_ino : 0, blocknr, block,
e4b->bd_group);
+ ext4_lock_group(sb, e4b->bd_group);
}
mb_clear_bit(block, EXT4_MB_BITMAP(e4b));
e4b->bd_info->bb_counters[order]++;
} while (1);
}
mb_check_buddy(e4b);
-
- return 0;
}
static int mb_find_extent(struct ext4_buddy *e4b, int order, int block,
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
- remove_proc_entry("mb_groups", sbi->s_mb_proc);
- remove_proc_entry("mb_history", sbi->s_mb_proc);
-
+ if (sbi->s_proc != NULL) {
+ remove_proc_entry("mb_groups", sbi->s_proc);
+ remove_proc_entry("mb_history", sbi->s_proc);
+ }
kfree(sbi->s_mb_history);
}
struct ext4_sb_info *sbi = EXT4_SB(sb);
int i;
- if (sbi->s_mb_proc != NULL) {
- proc_create_data("mb_history", S_IRUGO, sbi->s_mb_proc,
+ if (sbi->s_proc != NULL) {
+ proc_create_data("mb_history", S_IRUGO, sbi->s_proc,
&ext4_mb_seq_history_fops, sb);
- proc_create_data("mb_groups", S_IRUGO, sbi->s_mb_proc,
+ proc_create_data("mb_groups", S_IRUGO, sbi->s_proc,
&ext4_mb_seq_groups_fops, sb);
}
#define ext4_mb_history_init(sb)
#endif
+
+/* Create and initialize ext4_group_info data for the given group. */
+int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
+ struct ext4_group_desc *desc)
+{
+ int i, len;
+ int metalen = 0;
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_group_info **meta_group_info;
+
+ /*
+ * First check if this group is the first of a reserved block.
+ * If it's true, we have to allocate a new table of pointers
+ * to ext4_group_info structures
+ */
+ if (group % EXT4_DESC_PER_BLOCK(sb) == 0) {
+ metalen = sizeof(*meta_group_info) <<
+ EXT4_DESC_PER_BLOCK_BITS(sb);
+ meta_group_info = kmalloc(metalen, GFP_KERNEL);
+ if (meta_group_info == NULL) {
+ printk(KERN_ERR "EXT4-fs: can't allocate mem for a "
+ "buddy group\n");
+ goto exit_meta_group_info;
+ }
+ sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)] =
+ meta_group_info;
+ }
+
+ /*
+ * calculate needed size. if change bb_counters size,
+ * don't forget about ext4_mb_generate_buddy()
+ */
+ len = offsetof(typeof(**meta_group_info),
+ bb_counters[sb->s_blocksize_bits + 2]);
+
+ meta_group_info =
+ sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)];
+ i = group & (EXT4_DESC_PER_BLOCK(sb) - 1);
+
+ meta_group_info[i] = kzalloc(len, GFP_KERNEL);
+ if (meta_group_info[i] == NULL) {
+ printk(KERN_ERR "EXT4-fs: can't allocate buddy mem\n");
+ goto exit_group_info;
+ }
+ set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT,
+ &(meta_group_info[i]->bb_state));
+
+ /*
+ * initialize bb_free to be able to skip
+ * empty groups without initialization
+ */
+ if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
+ meta_group_info[i]->bb_free =
+ ext4_free_blocks_after_init(sb, group, desc);
+ } else {
+ meta_group_info[i]->bb_free =
+ le16_to_cpu(desc->bg_free_blocks_count);
+ }
+
+ INIT_LIST_HEAD(&meta_group_info[i]->bb_prealloc_list);
+ meta_group_info[i]->bb_free_root.rb_node = NULL;;
+
+#ifdef DOUBLE_CHECK
+ {
+ struct buffer_head *bh;
+ meta_group_info[i]->bb_bitmap =
+ kmalloc(sb->s_blocksize, GFP_KERNEL);
+ BUG_ON(meta_group_info[i]->bb_bitmap == NULL);
+ bh = ext4_read_block_bitmap(sb, group);
+ BUG_ON(bh == NULL);
+ memcpy(meta_group_info[i]->bb_bitmap, bh->b_data,
+ sb->s_blocksize);
+ put_bh(bh);
+ }
+#endif
+
+ return 0;
+
+exit_group_info:
+ /* If a meta_group_info table has been allocated, release it now */
+ if (group % EXT4_DESC_PER_BLOCK(sb) == 0)
+ kfree(sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)]);
+exit_meta_group_info:
+ return -ENOMEM;
+} /* ext4_mb_add_groupinfo */
+
+/*
+ * Add a group to the existing groups.
+ * This function is used for online resize
+ */
+int ext4_mb_add_more_groupinfo(struct super_block *sb, ext4_group_t group,
+ struct ext4_group_desc *desc)
+{
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct inode *inode = sbi->s_buddy_cache;
+ int blocks_per_page;
+ int block;
+ int pnum;
+ struct page *page;
+ int err;
+
+ /* Add group based on group descriptor*/
+ err = ext4_mb_add_groupinfo(sb, group, desc);
+ if (err)
+ return err;
+
+ /*
+ * Cache pages containing dynamic mb_alloc datas (buddy and bitmap
+ * datas) are set not up to date so that they will be re-initilaized
+ * during the next call to ext4_mb_load_buddy
+ */
+
+ /* Set buddy page as not up to date */
+ blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize;
+ block = group * 2;
+ pnum = block / blocks_per_page;
+ page = find_get_page(inode->i_mapping, pnum);
+ if (page != NULL) {
+ ClearPageUptodate(page);
+ page_cache_release(page);
+ }
+
+ /* Set bitmap page as not up to date */
+ block++;
+ pnum = block / blocks_per_page;
+ page = find_get_page(inode->i_mapping, pnum);
+ if (page != NULL) {
+ ClearPageUptodate(page);
+ page_cache_release(page);
+ }
+
+ return 0;
+}
+
+/*
+ * Update an existing group.
+ * This function is used for online resize
+ */
+void ext4_mb_update_group_info(struct ext4_group_info *grp, ext4_grpblk_t add)
+{
+ grp->bb_free += add;
+}
+
static int ext4_mb_init_backend(struct super_block *sb)
{
ext4_group_t i;
- int j, len, metalen;
+ int metalen;
struct ext4_sb_info *sbi = EXT4_SB(sb);
- int num_meta_group_infos =
- (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) >>
- EXT4_DESC_PER_BLOCK_BITS(sb);
+ struct ext4_super_block *es = sbi->s_es;
+ int num_meta_group_infos;
+ int num_meta_group_infos_max;
+ int array_size;
struct ext4_group_info **meta_group_info;
+ struct ext4_group_desc *desc;
+
+ /* This is the number of blocks used by GDT */
+ num_meta_group_infos = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) -
+ 1) >> EXT4_DESC_PER_BLOCK_BITS(sb);
+
+ /*
+ * This is the total number of blocks used by GDT including
+ * the number of reserved blocks for GDT.
+ * The s_group_info array is allocated with this value
+ * to allow a clean online resize without a complex
+ * manipulation of pointer.
+ * The drawback is the unused memory when no resize
+ * occurs but it's very low in terms of pages
+ * (see comments below)
+ * Need to handle this properly when META_BG resizing is allowed
+ */
+ num_meta_group_infos_max = num_meta_group_infos +
+ le16_to_cpu(es->s_reserved_gdt_blocks);
+ /*
+ * array_size is the size of s_group_info array. We round it
+ * to the next power of two because this approximation is done
+ * internally by kmalloc so we can have some more memory
+ * for free here (e.g. may be used for META_BG resize).
+ */
+ array_size = 1;
+ while (array_size < sizeof(*sbi->s_group_info) *
+ num_meta_group_infos_max)
+ array_size = array_size << 1;
/* An 8TB filesystem with 64-bit pointers requires a 4096 byte
* kmalloc. A 128kb malloc should suffice for a 256TB filesystem.
* So a two level scheme suffices for now. */
- sbi->s_group_info = kmalloc(sizeof(*sbi->s_group_info) *
- num_meta_group_infos, GFP_KERNEL);
+ sbi->s_group_info = kmalloc(array_size, GFP_KERNEL);
if (sbi->s_group_info == NULL) {
printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n");
return -ENOMEM;
sbi->s_group_info[i] = meta_group_info;
}
- /*
- * calculate needed size. if change bb_counters size,
- * don't forget about ext4_mb_generate_buddy()
- */
- len = sizeof(struct ext4_group_info);
- len += sizeof(unsigned short) * (sb->s_blocksize_bits + 2);
for (i = 0; i < sbi->s_groups_count; i++) {
- struct ext4_group_desc *desc;
-
- meta_group_info =
- sbi->s_group_info[i >> EXT4_DESC_PER_BLOCK_BITS(sb)];
- j = i & (EXT4_DESC_PER_BLOCK(sb) - 1);
-
- meta_group_info[j] = kzalloc(len, GFP_KERNEL);
- if (meta_group_info[j] == NULL) {
- printk(KERN_ERR "EXT4-fs: can't allocate buddy mem\n");
- goto err_freebuddy;
- }
desc = ext4_get_group_desc(sb, i, NULL);
if (desc == NULL) {
printk(KERN_ERR
"EXT4-fs: can't read descriptor %lu\n", i);
- i++;
goto err_freebuddy;
}
- set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT,
- &(meta_group_info[j]->bb_state));
-
- /*
- * initialize bb_free to be able to skip
- * empty groups without initialization
- */
- if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
- meta_group_info[j]->bb_free =
- ext4_free_blocks_after_init(sb, i, desc);
- } else {
- meta_group_info[j]->bb_free =
- le16_to_cpu(desc->bg_free_blocks_count);
- }
-
- INIT_LIST_HEAD(&meta_group_info[j]->bb_prealloc_list);
-
-#ifdef DOUBLE_CHECK
- {
- struct buffer_head *bh;
- meta_group_info[j]->bb_bitmap =
- kmalloc(sb->s_blocksize, GFP_KERNEL);
- BUG_ON(meta_group_info[j]->bb_bitmap == NULL);
- bh = ext4_read_block_bitmap(sb, i);
- BUG_ON(bh == NULL);
- memcpy(meta_group_info[j]->bb_bitmap, bh->b_data,
- sb->s_blocksize);
- put_bh(bh);
- }
-#endif
-
+ if (ext4_mb_add_groupinfo(sb, i, desc) != 0)
+ goto err_freebuddy;
}
return 0;
int ext4_mb_init(struct super_block *sb, int needs_recovery)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
- unsigned i;
+ unsigned i, j;
unsigned offset;
unsigned max;
int ret;
- if (!test_opt(sb, MBALLOC))
- return 0;
-
i = (sb->s_blocksize_bits + 2) * sizeof(unsigned short);
sbi->s_mb_offsets = kmalloc(i, GFP_KERNEL);
if (sbi->s_mb_offsets == NULL) {
- clear_opt(sbi->s_mount_opt, MBALLOC);
return -ENOMEM;
}
sbi->s_mb_maxs = kmalloc(i, GFP_KERNEL);
if (sbi->s_mb_maxs == NULL) {
- clear_opt(sbi->s_mount_opt, MBALLOC);
kfree(sbi->s_mb_maxs);
return -ENOMEM;
}
/* init file for buddy data */
ret = ext4_mb_init_backend(sb);
if (ret != 0) {
- clear_opt(sbi->s_mount_opt, MBALLOC);
kfree(sbi->s_mb_offsets);
kfree(sbi->s_mb_maxs);
return ret;
}
spin_lock_init(&sbi->s_md_lock);
- INIT_LIST_HEAD(&sbi->s_active_transaction);
- INIT_LIST_HEAD(&sbi->s_closed_transaction);
- INIT_LIST_HEAD(&sbi->s_committed_transaction);
spin_lock_init(&sbi->s_bal_lock);
sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
sbi->s_mb_history_filter = EXT4_MB_HISTORY_DEFAULT;
sbi->s_mb_group_prealloc = MB_DEFAULT_GROUP_PREALLOC;
- i = sizeof(struct ext4_locality_group) * NR_CPUS;
- sbi->s_locality_groups = kmalloc(i, GFP_KERNEL);
+ sbi->s_locality_groups = alloc_percpu(struct ext4_locality_group);
if (sbi->s_locality_groups == NULL) {
- clear_opt(sbi->s_mount_opt, MBALLOC);
kfree(sbi->s_mb_offsets);
kfree(sbi->s_mb_maxs);
return -ENOMEM;
}
- for (i = 0; i < NR_CPUS; i++) {
+ for_each_possible_cpu(i) {
struct ext4_locality_group *lg;
- lg = &sbi->s_locality_groups[i];
+ lg = per_cpu_ptr(sbi->s_locality_groups, i);
mutex_init(&lg->lg_mutex);
- INIT_LIST_HEAD(&lg->lg_prealloc_list);
+ for (j = 0; j < PREALLOC_TB_SIZE; j++)
+ INIT_LIST_HEAD(&lg->lg_prealloc_list[j]);
spin_lock_init(&lg->lg_prealloc_lock);
}
ext4_mb_init_per_dev_proc(sb);
ext4_mb_history_init(sb);
- printk("EXT4-fs: mballoc enabled\n");
+ sbi->s_journal->j_commit_callback = release_blocks_on_commit;
+
+ printk(KERN_INFO "EXT4-fs: mballoc enabled\n");
return 0;
}
pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list);
list_del(&pa->pa_group_list);
count++;
- kfree(pa);
+ kmem_cache_free(ext4_pspace_cachep, pa);
}
if (count)
mb_debug("mballoc: %u PAs left\n", count);
struct ext4_group_info *grinfo;
struct ext4_sb_info *sbi = EXT4_SB(sb);
- if (!test_opt(sb, MBALLOC))
- return 0;
-
- /* release freed, non-committed blocks */
- spin_lock(&sbi->s_md_lock);
- list_splice_init(&sbi->s_closed_transaction,
- &sbi->s_committed_transaction);
- list_splice_init(&sbi->s_active_transaction,
- &sbi->s_committed_transaction);
- spin_unlock(&sbi->s_md_lock);
- ext4_mb_free_committed_blocks(sb);
-
if (sbi->s_group_info) {
for (i = 0; i < sbi->s_groups_count; i++) {
grinfo = ext4_get_group_info(sb, i);
atomic_read(&sbi->s_mb_discarded));
}
- kfree(sbi->s_locality_groups);
-
+ free_percpu(sbi->s_locality_groups);
ext4_mb_history_release(sb);
ext4_mb_destroy_per_dev_proc(sb);
return 0;
}
-static noinline_for_stack void
-ext4_mb_free_committed_blocks(struct super_block *sb)
+/*
+ * This function is called by the jbd2 layer once the commit has finished,
+ * so we know we can free the blocks that were released with that commit.
+ */
+static void release_blocks_on_commit(journal_t *journal, transaction_t *txn)
{
- struct ext4_sb_info *sbi = EXT4_SB(sb);
- int err;
- int i;
- int count = 0;
- int count2 = 0;
- struct ext4_free_metadata *md;
+ struct super_block *sb = journal->j_private;
struct ext4_buddy e4b;
+ struct ext4_group_info *db;
+ int err, count = 0, count2 = 0;
+ struct ext4_free_data *entry;
+ ext4_fsblk_t discard_block;
+ struct list_head *l, *ltmp;
- if (list_empty(&sbi->s_committed_transaction))
- return;
-
- /* there is committed blocks to be freed yet */
- do {
- /* get next array of blocks */
- md = NULL;
- spin_lock(&sbi->s_md_lock);
- if (!list_empty(&sbi->s_committed_transaction)) {
- md = list_entry(sbi->s_committed_transaction.next,
- struct ext4_free_metadata, list);
- list_del(&md->list);
- }
- spin_unlock(&sbi->s_md_lock);
-
- if (md == NULL)
- break;
+ list_for_each_safe(l, ltmp, &txn->t_private_list) {
+ entry = list_entry(l, struct ext4_free_data, list);
mb_debug("gonna free %u blocks in group %lu (0x%p):",
- md->num, md->group, md);
+ entry->count, entry->group, entry);
- err = ext4_mb_load_buddy(sb, md->group, &e4b);
+ err = ext4_mb_load_buddy(sb, entry->group, &e4b);
/* we expect to find existing buddy because it's pinned */
BUG_ON(err != 0);
+ db = e4b.bd_info;
/* there are blocks to put in buddy to make them really free */
- count += md->num;
+ count += entry->count;
count2++;
- ext4_lock_group(sb, md->group);
- for (i = 0; i < md->num; i++) {
- mb_debug(" %u", md->blocks[i]);
- err = mb_free_blocks(NULL, &e4b, md->blocks[i], 1);
- BUG_ON(err != 0);
+ ext4_lock_group(sb, entry->group);
+ /* Take it out of per group rb tree */
+ rb_erase(&entry->node, &(db->bb_free_root));
+ mb_free_blocks(NULL, &e4b, entry->start_blk, entry->count);
+
+ if (!db->bb_free_root.rb_node) {
+ /* No more items in the per group rb tree
+ * balance refcounts from ext4_mb_free_metadata()
+ */
+ page_cache_release(e4b.bd_buddy_page);
+ page_cache_release(e4b.bd_bitmap_page);
}
- mb_debug("\n");
- ext4_unlock_group(sb, md->group);
-
- /* balance refcounts from ext4_mb_free_metadata() */
- page_cache_release(e4b.bd_buddy_page);
- page_cache_release(e4b.bd_bitmap_page);
-
- kfree(md);
+ ext4_unlock_group(sb, entry->group);
+ discard_block = (ext4_fsblk_t) entry->group * EXT4_BLOCKS_PER_GROUP(sb)
+ + entry->start_blk
+ + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block);
+ trace_mark(ext4_discard_blocks, "dev %s blk %llu count %u", sb->s_id,
+ (unsigned long long) discard_block, entry->count);
+ sb_issue_discard(sb, discard_block, entry->count);
+
+ kmem_cache_free(ext4_free_ext_cachep, entry);
ext4_mb_release_desc(&e4b);
-
- } while (md);
+ }
mb_debug("freed %u blocks in %u structures\n", count, count2);
}
#define EXT4_MB_STREAM_REQ "stream_req"
#define EXT4_MB_GROUP_PREALLOC "group_prealloc"
-
-
-#define MB_PROC_FOPS(name) \
-static int ext4_mb_##name##_proc_show(struct seq_file *m, void *v) \
-{ \
- struct ext4_sb_info *sbi = m->private; \
- \
- seq_printf(m, "%ld\n", sbi->s_mb_##name); \
- return 0; \
-} \
- \
-static int ext4_mb_##name##_proc_open(struct inode *inode, struct file *file)\
-{ \
- return single_open(file, ext4_mb_##name##_proc_show, PDE(inode)->data);\
-} \
- \
-static ssize_t ext4_mb_##name##_proc_write(struct file *file, \
- const char __user *buf, size_t cnt, loff_t *ppos) \
-{ \
- struct ext4_sb_info *sbi = PDE(file->f_path.dentry->d_inode)->data;\
- char str[32]; \
- long value; \
- if (cnt >= sizeof(str)) \
- return -EINVAL; \
- if (copy_from_user(str, buf, cnt)) \
- return -EFAULT; \
- value = simple_strtol(str, NULL, 0); \
- if (value <= 0) \
- return -ERANGE; \
- sbi->s_mb_##name = value; \
- return cnt; \
-} \
- \
-static const struct file_operations ext4_mb_##name##_proc_fops = { \
- .owner = THIS_MODULE, \
- .open = ext4_mb_##name##_proc_open, \
- .read = seq_read, \
- .llseek = seq_lseek, \
- .release = single_release, \
- .write = ext4_mb_##name##_proc_write, \
-};
-
-MB_PROC_FOPS(stats);
-MB_PROC_FOPS(max_to_scan);
-MB_PROC_FOPS(min_to_scan);
-MB_PROC_FOPS(order2_reqs);
-MB_PROC_FOPS(stream_request);
-MB_PROC_FOPS(group_prealloc);
-
-#define MB_PROC_HANDLER(name, var) \
-do { \
- proc = proc_create_data(name, mode, sbi->s_mb_proc, \
- &ext4_mb_##var##_proc_fops, sbi); \
- if (proc == NULL) { \
- printk(KERN_ERR "EXT4-fs: can't to create %s\n", name); \
- goto err_out; \
- } \
-} while (0)
-
static int ext4_mb_init_per_dev_proc(struct super_block *sb)
{
+#ifdef CONFIG_PROC_FS
mode_t mode = S_IFREG | S_IRUGO | S_IWUSR;
struct ext4_sb_info *sbi = EXT4_SB(sb);
struct proc_dir_entry *proc;
- char devname[64];
-
- bdevname(sb->s_bdev, devname);
- sbi->s_mb_proc = proc_mkdir(devname, proc_root_ext4);
- MB_PROC_HANDLER(EXT4_MB_STATS_NAME, stats);
- MB_PROC_HANDLER(EXT4_MB_MAX_TO_SCAN_NAME, max_to_scan);
- MB_PROC_HANDLER(EXT4_MB_MIN_TO_SCAN_NAME, min_to_scan);
- MB_PROC_HANDLER(EXT4_MB_ORDER2_REQ, order2_reqs);
- MB_PROC_HANDLER(EXT4_MB_STREAM_REQ, stream_request);
- MB_PROC_HANDLER(EXT4_MB_GROUP_PREALLOC, group_prealloc);
+ if (sbi->s_proc == NULL)
+ return -EINVAL;
+ EXT4_PROC_HANDLER(EXT4_MB_STATS_NAME, mb_stats);
+ EXT4_PROC_HANDLER(EXT4_MB_MAX_TO_SCAN_NAME, mb_max_to_scan);
+ EXT4_PROC_HANDLER(EXT4_MB_MIN_TO_SCAN_NAME, mb_min_to_scan);
+ EXT4_PROC_HANDLER(EXT4_MB_ORDER2_REQ, mb_order2_reqs);
+ EXT4_PROC_HANDLER(EXT4_MB_STREAM_REQ, mb_stream_request);
+ EXT4_PROC_HANDLER(EXT4_MB_GROUP_PREALLOC, mb_group_prealloc);
return 0;
err_out:
- printk(KERN_ERR "EXT4-fs: Unable to create %s\n", devname);
- remove_proc_entry(EXT4_MB_GROUP_PREALLOC, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_STREAM_REQ, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_ORDER2_REQ, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_MIN_TO_SCAN_NAME, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_MAX_TO_SCAN_NAME, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_STATS_NAME, sbi->s_mb_proc);
- remove_proc_entry(devname, proc_root_ext4);
- sbi->s_mb_proc = NULL;
-
+ remove_proc_entry(EXT4_MB_GROUP_PREALLOC, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_STREAM_REQ, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_ORDER2_REQ, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_MIN_TO_SCAN_NAME, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_MAX_TO_SCAN_NAME, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_STATS_NAME, sbi->s_proc);
return -ENOMEM;
+#else
+ return 0;
+#endif
}
static int ext4_mb_destroy_per_dev_proc(struct super_block *sb)
{
+#ifdef CONFIG_PROC_FS
struct ext4_sb_info *sbi = EXT4_SB(sb);
- char devname[64];
- if (sbi->s_mb_proc == NULL)
+ if (sbi->s_proc == NULL)
return -EINVAL;
- bdevname(sb->s_bdev, devname);
- remove_proc_entry(EXT4_MB_GROUP_PREALLOC, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_STREAM_REQ, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_ORDER2_REQ, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_MIN_TO_SCAN_NAME, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_MAX_TO_SCAN_NAME, sbi->s_mb_proc);
- remove_proc_entry(EXT4_MB_STATS_NAME, sbi->s_mb_proc);
- remove_proc_entry(devname, proc_root_ext4);
-
+ remove_proc_entry(EXT4_MB_GROUP_PREALLOC, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_STREAM_REQ, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_ORDER2_REQ, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_MIN_TO_SCAN_NAME, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_MAX_TO_SCAN_NAME, sbi->s_proc);
+ remove_proc_entry(EXT4_MB_STATS_NAME, sbi->s_proc);
+#endif
return 0;
}
kmem_cache_destroy(ext4_pspace_cachep);
return -ENOMEM;
}
-#ifdef CONFIG_PROC_FS
- proc_root_ext4 = proc_mkdir("fs/ext4", NULL);
- if (proc_root_ext4 == NULL)
- printk(KERN_ERR "EXT4-fs: Unable to create fs/ext4\n");
-#endif
+
+ ext4_free_ext_cachep =
+ kmem_cache_create("ext4_free_block_extents",
+ sizeof(struct ext4_free_data),
+ 0, SLAB_RECLAIM_ACCOUNT, NULL);
+ if (ext4_free_ext_cachep == NULL) {
+ kmem_cache_destroy(ext4_pspace_cachep);
+ kmem_cache_destroy(ext4_ac_cachep);
+ return -ENOMEM;
+ }
return 0;
}
/* XXX: synchronize_rcu(); */
kmem_cache_destroy(ext4_pspace_cachep);
kmem_cache_destroy(ext4_ac_cachep);
-#ifdef CONFIG_PROC_FS
- remove_proc_entry("fs/ext4", NULL);
-#endif
+ kmem_cache_destroy(ext4_free_ext_cachep);
}
*/
static noinline_for_stack int
ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac,
- handle_t *handle)
+ handle_t *handle, unsigned long reserv_blks)
{
struct buffer_head *bitmap_bh = NULL;
struct ext4_super_block *es;
gdp->bg_checksum = ext4_group_desc_csum(sbi, ac->ac_b_ex.fe_group, gdp);
spin_unlock(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group));
percpu_counter_sub(&sbi->s_freeblocks_counter, ac->ac_b_ex.fe_len);
+ /*
+ * Now reduce the dirty block count also. Should not go negative
+ */
+ if (!(ac->ac_flags & EXT4_MB_DELALLOC_RESERVED))
+ /* release all the reserved blocks if non delalloc */
+ percpu_counter_sub(&sbi->s_dirtyblocks_counter, reserv_blks);
+ else
+ percpu_counter_sub(&sbi->s_dirtyblocks_counter,
+ ac->ac_b_ex.fe_len);
+
+ if (sbi->s_log_groups_per_flex) {
+ ext4_group_t flex_group = ext4_flex_group(sbi,
+ ac->ac_b_ex.fe_group);
+ spin_lock(sb_bgl_lock(sbi, flex_group));
+ sbi->s_flex_groups[flex_group].free_blocks -= ac->ac_b_ex.fe_len;
+ spin_unlock(sb_bgl_lock(sbi, flex_group));
+ }
err = ext4_journal_dirty_metadata(handle, bitmap_bh);
if (err)
struct ext4_prealloc_space *pa)
{
unsigned int len = ac->ac_o_ex.fe_len;
+
ext4_get_group_no_and_offset(ac->ac_sb, pa->pa_pstart,
&ac->ac_b_ex.fe_group,
&ac->ac_b_ex.fe_start);
}
/*
+ * Return the prealloc space that have minimal distance
+ * from the goal block. @cpa is the prealloc
+ * space that is having currently known minimal distance
+ * from the goal block.
+ */
+static struct ext4_prealloc_space *
+ext4_mb_check_group_pa(ext4_fsblk_t goal_block,
+ struct ext4_prealloc_space *pa,
+ struct ext4_prealloc_space *cpa)
+{
+ ext4_fsblk_t cur_distance, new_distance;
+
+ if (cpa == NULL) {
+ atomic_inc(&pa->pa_count);
+ return pa;
+ }
+ cur_distance = abs(goal_block - cpa->pa_pstart);
+ new_distance = abs(goal_block - pa->pa_pstart);
+
+ if (cur_distance < new_distance)
+ return cpa;
+
+ /* drop the previous reference */
+ atomic_dec(&cpa->pa_count);
+ atomic_inc(&pa->pa_count);
+ return pa;
+}
+
+/*
* search goal blocks in preallocated space
*/
static noinline_for_stack int
ext4_mb_use_preallocated(struct ext4_allocation_context *ac)
{
+ int order, i;
struct ext4_inode_info *ei = EXT4_I(ac->ac_inode);
struct ext4_locality_group *lg;
- struct ext4_prealloc_space *pa;
+ struct ext4_prealloc_space *pa, *cpa = NULL;
+ ext4_fsblk_t goal_block;
/* only data can be preallocated */
if (!(ac->ac_flags & EXT4_MB_HINT_DATA))
lg = ac->ac_lg;
if (lg == NULL)
return 0;
+ order = fls(ac->ac_o_ex.fe_len) - 1;
+ if (order > PREALLOC_TB_SIZE - 1)
+ /* The max size of hash table is PREALLOC_TB_SIZE */
+ order = PREALLOC_TB_SIZE - 1;
+
+ goal_block = ac->ac_g_ex.fe_group * EXT4_BLOCKS_PER_GROUP(ac->ac_sb) +
+ ac->ac_g_ex.fe_start +
+ le32_to_cpu(EXT4_SB(ac->ac_sb)->s_es->s_first_data_block);
+ /*
+ * search for the prealloc space that is having
+ * minimal distance from the goal block.
+ */
+ for (i = order; i < PREALLOC_TB_SIZE; i++) {
+ rcu_read_lock();
+ list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[i],
+ pa_inode_list) {
+ spin_lock(&pa->pa_lock);
+ if (pa->pa_deleted == 0 &&
+ pa->pa_free >= ac->ac_o_ex.fe_len) {
- rcu_read_lock();
- list_for_each_entry_rcu(pa, &lg->lg_prealloc_list, pa_inode_list) {
- spin_lock(&pa->pa_lock);
- if (pa->pa_deleted == 0 && pa->pa_free >= ac->ac_o_ex.fe_len) {
- atomic_inc(&pa->pa_count);
- ext4_mb_use_group_pa(ac, pa);
+ cpa = ext4_mb_check_group_pa(goal_block,
+ pa, cpa);
+ }
spin_unlock(&pa->pa_lock);
- ac->ac_criteria = 20;
- rcu_read_unlock();
- return 1;
}
- spin_unlock(&pa->pa_lock);
+ rcu_read_unlock();
+ }
+ if (cpa) {
+ ext4_mb_use_group_pa(ac, cpa);
+ ac->ac_criteria = 20;
+ return 1;
}
- rcu_read_unlock();
-
return 0;
}
pa->pa_free = pa->pa_len;
atomic_set(&pa->pa_count, 1);
spin_lock_init(&pa->pa_lock);
+ INIT_LIST_HEAD(&pa->pa_inode_list);
pa->pa_deleted = 0;
pa->pa_linear = 1;
list_add(&pa->pa_group_list, &grp->bb_prealloc_list);
ext4_unlock_group(sb, ac->ac_b_ex.fe_group);
- spin_lock(pa->pa_obj_lock);
- list_add_tail_rcu(&pa->pa_inode_list, &lg->lg_prealloc_list);
- spin_unlock(pa->pa_obj_lock);
-
+ /*
+ * We will later add the new pa to the right bucket
+ * after updating the pa_free in ext4_mb_release_context
+ */
return 0;
}
bitmap_bh = ext4_read_block_bitmap(sb, group);
if (bitmap_bh == NULL) {
- /* error handling here */
- ext4_mb_release_desc(&e4b);
- BUG_ON(bitmap_bh == NULL);
+ ext4_error(sb, __func__, "Error in reading block "
+ "bitmap for %lu\n", group);
+ return 0;
}
err = ext4_mb_load_buddy(sb, group, &e4b);
- BUG_ON(err != 0); /* error handling here */
+ if (err) {
+ ext4_error(sb, __func__, "Error in loading buddy "
+ "information for %lu\n", group);
+ put_bh(bitmap_bh);
+ return 0;
+ }
if (needed == 0)
needed = EXT4_BLOCKS_PER_GROUP(sb) + 1;
- grp = ext4_get_group_info(sb, group);
INIT_LIST_HEAD(&list);
-
ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS);
repeat:
ext4_lock_group(sb, group);
*
* FIXME!! Make sure it is valid at all the call sites
*/
-void ext4_mb_discard_inode_preallocations(struct inode *inode)
+void ext4_discard_preallocations(struct inode *inode)
{
struct ext4_inode_info *ei = EXT4_I(inode);
struct super_block *sb = inode->i_sb;
struct ext4_buddy e4b;
int err;
- if (!test_opt(sb, MBALLOC) || !S_ISREG(inode->i_mode)) {
+ if (!S_ISREG(inode->i_mode)) {
/*BUG_ON(!list_empty(&ei->i_prealloc_list));*/
return;
}
ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, NULL);
err = ext4_mb_load_buddy(sb, group, &e4b);
- BUG_ON(err != 0); /* error handling here */
+ if (err) {
+ ext4_error(sb, __func__, "Error in loading buddy "
+ "information for %lu\n", group);
+ continue;
+ }
bitmap_bh = ext4_read_block_bitmap(sb, group);
if (bitmap_bh == NULL) {
- /* error handling here */
+ ext4_error(sb, __func__, "Error in reading block "
+ "bitmap for %lu\n", group);
ext4_mb_release_desc(&e4b);
- BUG_ON(bitmap_bh == NULL);
+ continue;
}
ext4_lock_group(sb, group);
* per cpu locality group is to reduce the contention between block
* request from multiple CPUs.
*/
- ac->ac_lg = &sbi->s_locality_groups[get_cpu()];
- put_cpu();
+ ac->ac_lg = per_cpu_ptr(sbi->s_locality_groups, raw_smp_processor_id());
/* we're going to use group allocation */
ac->ac_flags |= EXT4_MB_HINT_GROUP_ALLOC;
}
+static noinline_for_stack void
+ext4_mb_discard_lg_preallocations(struct super_block *sb,
+ struct ext4_locality_group *lg,
+ int order, int total_entries)
+{
+ ext4_group_t group = 0;
+ struct ext4_buddy e4b;
+ struct list_head discard_list;
+ struct ext4_prealloc_space *pa, *tmp;
+ struct ext4_allocation_context *ac;
+
+ mb_debug("discard locality group preallocation\n");
+
+ INIT_LIST_HEAD(&discard_list);
+ ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS);
+
+ spin_lock(&lg->lg_prealloc_lock);
+ list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[order],
+ pa_inode_list) {
+ spin_lock(&pa->pa_lock);
+ if (atomic_read(&pa->pa_count)) {
+ /*
+ * This is the pa that we just used
+ * for block allocation. So don't
+ * free that
+ */
+ spin_unlock(&pa->pa_lock);
+ continue;
+ }
+ if (pa->pa_deleted) {
+ spin_unlock(&pa->pa_lock);
+ continue;
+ }
+ /* only lg prealloc space */
+ BUG_ON(!pa->pa_linear);
+
+ /* seems this one can be freed ... */
+ pa->pa_deleted = 1;
+ spin_unlock(&pa->pa_lock);
+
+ list_del_rcu(&pa->pa_inode_list);
+ list_add(&pa->u.pa_tmp_list, &discard_list);
+
+ total_entries--;
+ if (total_entries <= 5) {
+ /*
+ * we want to keep only 5 entries
+ * allowing it to grow to 8. This
+ * mak sure we don't call discard
+ * soon for this list.
+ */
+ break;
+ }
+ }
+ spin_unlock(&lg->lg_prealloc_lock);
+
+ list_for_each_entry_safe(pa, tmp, &discard_list, u.pa_tmp_list) {
+
+ ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, NULL);
+ if (ext4_mb_load_buddy(sb, group, &e4b)) {
+ ext4_error(sb, __func__, "Error in loading buddy "
+ "information for %lu\n", group);
+ continue;
+ }
+ ext4_lock_group(sb, group);
+ list_del(&pa->pa_group_list);
+ ext4_mb_release_group_pa(&e4b, pa, ac);
+ ext4_unlock_group(sb, group);
+
+ ext4_mb_release_desc(&e4b);
+ list_del(&pa->u.pa_tmp_list);
+ call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback);
+ }
+ if (ac)
+ kmem_cache_free(ext4_ac_cachep, ac);
+}
+
+/*
+ * We have incremented pa_count. So it cannot be freed at this
+ * point. Also we hold lg_mutex. So no parallel allocation is
+ * possible from this lg. That means pa_free cannot be updated.
+ *
+ * A parallel ext4_mb_discard_group_preallocations is possible.
+ * which can cause the lg_prealloc_list to be updated.
+ */
+
+static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac)
+{
+ int order, added = 0, lg_prealloc_count = 1;
+ struct super_block *sb = ac->ac_sb;
+ struct ext4_locality_group *lg = ac->ac_lg;
+ struct ext4_prealloc_space *tmp_pa, *pa = ac->ac_pa;
+
+ order = fls(pa->pa_free) - 1;
+ if (order > PREALLOC_TB_SIZE - 1)
+ /* The max size of hash table is PREALLOC_TB_SIZE */
+ order = PREALLOC_TB_SIZE - 1;
+ /* Add the prealloc space to lg */
+ rcu_read_lock();
+ list_for_each_entry_rcu(tmp_pa, &lg->lg_prealloc_list[order],
+ pa_inode_list) {
+ spin_lock(&tmp_pa->pa_lock);
+ if (tmp_pa->pa_deleted) {
+ spin_unlock(&pa->pa_lock);
+ continue;
+ }
+ if (!added && pa->pa_free < tmp_pa->pa_free) {
+ /* Add to the tail of the previous entry */
+ list_add_tail_rcu(&pa->pa_inode_list,
+ &tmp_pa->pa_inode_list);
+ added = 1;
+ /*
+ * we want to count the total
+ * number of entries in the list
+ */
+ }
+ spin_unlock(&tmp_pa->pa_lock);
+ lg_prealloc_count++;
+ }
+ if (!added)
+ list_add_tail_rcu(&pa->pa_inode_list,
+ &lg->lg_prealloc_list[order]);
+ rcu_read_unlock();
+
+ /* Now trim the list to be not more than 8 elements */
+ if (lg_prealloc_count > 8) {
+ ext4_mb_discard_lg_preallocations(sb, lg,
+ order, lg_prealloc_count);
+ return;
+ }
+ return ;
+}
+
/*
* release all resource we used in allocation
*/
static int ext4_mb_release_context(struct ext4_allocation_context *ac)
{
- if (ac->ac_pa) {
- if (ac->ac_pa->pa_linear) {
+ struct ext4_prealloc_space *pa = ac->ac_pa;
+ if (pa) {
+ if (pa->pa_linear) {
/* see comment in ext4_mb_use_group_pa() */
- spin_lock(&ac->ac_pa->pa_lock);
- ac->ac_pa->pa_pstart += ac->ac_b_ex.fe_len;
- ac->ac_pa->pa_lstart += ac->ac_b_ex.fe_len;
- ac->ac_pa->pa_free -= ac->ac_b_ex.fe_len;
- ac->ac_pa->pa_len -= ac->ac_b_ex.fe_len;
- spin_unlock(&ac->ac_pa->pa_lock);
+ spin_lock(&pa->pa_lock);
+ pa->pa_pstart += ac->ac_b_ex.fe_len;
+ pa->pa_lstart += ac->ac_b_ex.fe_len;
+ pa->pa_free -= ac->ac_b_ex.fe_len;
+ pa->pa_len -= ac->ac_b_ex.fe_len;
+ spin_unlock(&pa->pa_lock);
+ /*
+ * We want to add the pa to the right bucket.
+ * Remove it from the list and while adding
+ * make sure the list to which we are adding
+ * doesn't grow big.
+ */
+ if (likely(pa->pa_free)) {
+ spin_lock(pa->pa_obj_lock);
+ list_del_rcu(&pa->pa_inode_list);
+ spin_unlock(pa->pa_obj_lock);
+ ext4_mb_add_n_trim(ac);
+ }
}
- ext4_mb_put_pa(ac, ac->ac_sb, ac->ac_pa);
+ ext4_mb_put_pa(ac, ac->ac_sb, pa);
}
if (ac->ac_bitmap_page)
page_cache_release(ac->ac_bitmap_page);
ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
struct ext4_allocation_request *ar, int *errp)
{
+ int freed;
struct ext4_allocation_context *ac = NULL;
struct ext4_sb_info *sbi;
struct super_block *sb;
ext4_fsblk_t block = 0;
- int freed;
- int inquota;
+ unsigned long inquota;
+ unsigned long reserv_blks = 0;
sb = ar->inode->i_sb;
sbi = EXT4_SB(sb);
- if (!test_opt(sb, MBALLOC)) {
- block = ext4_new_blocks_old(handle, ar->inode, ar->goal,
- &(ar->len), errp);
- return block;
+ if (!EXT4_I(ar->inode)->i_delalloc_reserved_flag) {
+ /*
+ * With delalloc we already reserved the blocks
+ */
+ while (ar->len && ext4_claim_free_blocks(sbi, ar->len)) {
+ /* let others to free the space */
+ yield();
+ ar->len = ar->len >> 1;
+ }
+ if (!ar->len) {
+ *errp = -ENOSPC;
+ return 0;
+ }
+ reserv_blks = ar->len;
}
-
while (ar->len && DQUOT_ALLOC_BLOCK(ar->inode, ar->len)) {
ar->flags |= EXT4_MB_HINT_NOPREALLOC;
ar->len--;
}
inquota = ar->len;
+ if (EXT4_I(ar->inode)->i_delalloc_reserved_flag)
+ ar->flags |= EXT4_MB_DELALLOC_RESERVED;
+
ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS);
if (!ac) {
+ ar->len = 0;
*errp = -ENOMEM;
- return 0;
+ goto out1;
}
- ext4_mb_poll_new_transaction(sb, handle);
-
*errp = ext4_mb_initialize_context(ac, ar);
if (*errp) {
ar->len = 0;
- goto out;
+ goto out2;
}
ac->ac_op = EXT4_MB_HISTORY_PREALLOC;
if (!ext4_mb_use_preallocated(ac)) {
-
ac->ac_op = EXT4_MB_HISTORY_ALLOC;
ext4_mb_normalize_request(ac, ar);
repeat:
}
if (likely(ac->ac_status == AC_STATUS_FOUND)) {
- *errp = ext4_mb_mark_diskspace_used(ac, handle);
+ *errp = ext4_mb_mark_diskspace_used(ac, handle, reserv_blks);
if (*errp == -EAGAIN) {
ac->ac_b_ex.fe_group = 0;
ac->ac_b_ex.fe_start = 0;
ext4_mb_release_context(ac);
-out:
+out2:
+ kmem_cache_free(ext4_ac_cachep, ac);
+out1:
if (ar->len < inquota)
DQUOT_FREE_BLOCK(ar->inode, inquota - ar->len);
- kmem_cache_free(ext4_ac_cachep, ac);
return block;
}
-static void ext4_mb_poll_new_transaction(struct super_block *sb,
- handle_t *handle)
-{
- struct ext4_sb_info *sbi = EXT4_SB(sb);
-
- if (sbi->s_last_transaction == handle->h_transaction->t_tid)
- return;
-
- /* new transaction! time to close last one and free blocks for
- * committed transaction. we know that only transaction can be
- * active, so previos transaction can be being logged and we
- * know that transaction before previous is known to be already
- * logged. this means that now we may free blocks freed in all
- * transactions before previous one. hope I'm clear enough ... */
-
- spin_lock(&sbi->s_md_lock);
- if (sbi->s_last_transaction != handle->h_transaction->t_tid) {
- mb_debug("new transaction %lu, old %lu\n",
- (unsigned long) handle->h_transaction->t_tid,
- (unsigned long) sbi->s_last_transaction);
- list_splice_init(&sbi->s_closed_transaction,
- &sbi->s_committed_transaction);
- list_splice_init(&sbi->s_active_transaction,
- &sbi->s_closed_transaction);
- sbi->s_last_transaction = handle->h_transaction->t_tid;
- }
- spin_unlock(&sbi->s_md_lock);
- ext4_mb_free_committed_blocks(sb);
+/*
+ * We can merge two free data extents only if the physical blocks
+ * are contiguous, AND the extents were freed by the same transaction,
+ * AND the blocks are associated with the same group.
+ */
+static int can_merge(struct ext4_free_data *entry1,
+ struct ext4_free_data *entry2)
+{
+ if ((entry1->t_tid == entry2->t_tid) &&
+ (entry1->group == entry2->group) &&
+ ((entry1->start_blk + entry1->count) == entry2->start_blk))
+ return 1;
+ return 0;
}
static noinline_for_stack int
struct ext4_group_info *db = e4b->bd_info;
struct super_block *sb = e4b->bd_sb;
struct ext4_sb_info *sbi = EXT4_SB(sb);
- struct ext4_free_metadata *md;
- int i;
+ struct ext4_free_data *entry, *new_entry;
+ struct rb_node **n = &db->bb_free_root.rb_node, *node;
+ struct rb_node *parent = NULL, *new_node;
+
BUG_ON(e4b->bd_bitmap_page == NULL);
BUG_ON(e4b->bd_buddy_page == NULL);
- ext4_lock_group(sb, group);
- for (i = 0; i < count; i++) {
- md = db->bb_md_cur;
- if (md && db->bb_tid != handle->h_transaction->t_tid) {
- db->bb_md_cur = NULL;
- md = NULL;
- }
+ new_entry = kmem_cache_alloc(ext4_free_ext_cachep, GFP_NOFS);
+ new_entry->start_blk = block;
+ new_entry->group = group;
+ new_entry->count = count;
+ new_entry->t_tid = handle->h_transaction->t_tid;
+ new_node = &new_entry->node;
- if (md == NULL) {
+ ext4_lock_group(sb, group);
+ if (!*n) {
+ /* first free block exent. We need to
+ protect buddy cache from being freed,
+ * otherwise we'll refresh it from
+ * on-disk bitmap and lose not-yet-available
+ * blocks */
+ page_cache_get(e4b->bd_buddy_page);
+ page_cache_get(e4b->bd_bitmap_page);
+ }
+ while (*n) {
+ parent = *n;
+ entry = rb_entry(parent, struct ext4_free_data, node);
+ if (block < entry->start_blk)
+ n = &(*n)->rb_left;
+ else if (block >= (entry->start_blk + entry->count))
+ n = &(*n)->rb_right;
+ else {
ext4_unlock_group(sb, group);
- md = kmalloc(sizeof(*md), GFP_NOFS);
- if (md == NULL)
- return -ENOMEM;
- md->num = 0;
- md->group = group;
+ ext4_error(sb, __func__,
+ "Double free of blocks %d (%d %d)\n",
+ block, entry->start_blk, entry->count);
+ return 0;
+ }
+ }
- ext4_lock_group(sb, group);
- if (db->bb_md_cur == NULL) {
- spin_lock(&sbi->s_md_lock);
- list_add(&md->list, &sbi->s_active_transaction);
- spin_unlock(&sbi->s_md_lock);
- /* protect buddy cache from being freed,
- * otherwise we'll refresh it from
- * on-disk bitmap and lose not-yet-available
- * blocks */
- page_cache_get(e4b->bd_buddy_page);
- page_cache_get(e4b->bd_bitmap_page);
- db->bb_md_cur = md;
- db->bb_tid = handle->h_transaction->t_tid;
- mb_debug("new md 0x%p for group %lu\n",
- md, md->group);
- } else {
- kfree(md);
- md = db->bb_md_cur;
- }
+ rb_link_node(new_node, parent, n);
+ rb_insert_color(new_node, &db->bb_free_root);
+
+ /* Now try to see the extent can be merged to left and right */
+ node = rb_prev(new_node);
+ if (node) {
+ entry = rb_entry(node, struct ext4_free_data, node);
+ if (can_merge(entry, new_entry)) {
+ new_entry->start_blk = entry->start_blk;
+ new_entry->count += entry->count;
+ rb_erase(node, &(db->bb_free_root));
+ spin_lock(&sbi->s_md_lock);
+ list_del(&entry->list);
+ spin_unlock(&sbi->s_md_lock);
+ kmem_cache_free(ext4_free_ext_cachep, entry);
}
+ }
- BUG_ON(md->num >= EXT4_BB_MAX_BLOCKS);
- md->blocks[md->num] = block + i;
- md->num++;
- if (md->num == EXT4_BB_MAX_BLOCKS) {
- /* no more space, put full container on a sb's list */
- db->bb_md_cur = NULL;
+ node = rb_next(new_node);
+ if (node) {
+ entry = rb_entry(node, struct ext4_free_data, node);
+ if (can_merge(new_entry, entry)) {
+ new_entry->count += entry->count;
+ rb_erase(node, &(db->bb_free_root));
+ spin_lock(&sbi->s_md_lock);
+ list_del(&entry->list);
+ spin_unlock(&sbi->s_md_lock);
+ kmem_cache_free(ext4_free_ext_cachep, entry);
}
}
+ /* Add the extent to transaction's private list */
+ spin_lock(&sbi->s_md_lock);
+ list_add(&new_entry->list, &handle->h_transaction->t_private_list);
+ spin_unlock(&sbi->s_md_lock);
ext4_unlock_group(sb, group);
return 0;
}
*freed = 0;
- ext4_mb_poll_new_transaction(sb, handle);
-
sbi = EXT4_SB(sb);
es = EXT4_SB(sb)->s_es;
if (block < le32_to_cpu(es->s_first_data_block) ||
count -= overflow;
}
bitmap_bh = ext4_read_block_bitmap(sb, block_group);
- if (!bitmap_bh)
+ if (!bitmap_bh) {
+ err = -EIO;
goto error_return;
+ }
gdp = ext4_get_group_desc(sb, block_group, &gd_bh);
- if (!gdp)
+ if (!gdp) {
+ err = -EIO;
goto error_return;
+ }
if (in_range(ext4_block_bitmap(sb, gdp), block, count) ||
in_range(ext4_inode_bitmap(sb, gdp), block, count) ||
ext4_mb_free_metadata(handle, &e4b, block_group, bit, count);
} else {
ext4_lock_group(sb, block_group);
- err = mb_free_blocks(inode, &e4b, bit, count);
+ mb_free_blocks(inode, &e4b, bit, count);
ext4_mb_return_to_preallocation(inode, &e4b, block, count);
ext4_unlock_group(sb, block_group);
- BUG_ON(err != 0);
}
spin_lock(sb_bgl_lock(sbi, block_group));
spin_unlock(sb_bgl_lock(sbi, block_group));
percpu_counter_add(&sbi->s_freeblocks_counter, count);
+ 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;
+ spin_unlock(sb_bgl_lock(sbi, flex_group));
+ }
+
ext4_mb_release_desc(&e4b);
*freed += count;