ext4: teach the inode allocator to use a goal inode number

Enhance the inode allocator to take a goal inode number as a
paremeter; if it is specified, it takes precedence over Orlov or
parent directory inode allocation algorithms.

The extents migration function uses the goal inode number so that the
extent trees allocated the migration function use the correct flex_bg.
In the future, the goal inode functionality will also be used to
allocate an adjacent inode for the extended attributes.

Also, for testing purposes the goal inode number can be specified via
/sys/fs/{dev}/inode_goal.  This can be useful for testing inode
allocation beyond 2^32 blocks on very large filesystems.

Signed-off-by: Andreas Dilger <adilger@sun.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

6 files changed, 36 insertions, 10 deletions

diff --git a/Documentation/ABI/testing/sysfs-fs-ext4 b/Documentation/ABI/testing/sysfs-fs-ext4
index 4e79074de282..5fb709997d96 100644
--- a/Documentation/ABI/testing/sysfs-fs-ext4
+++ b/Documentation/ABI/testing/sysfs-fs-ext4
@@ -79,3 +79,13 @@ Description:
                 This file is read-only and shows the number of
                 kilobytes of data that have been written to this
                 filesystem since it was mounted.
+
+What:           /sys/fs/ext4/<disk>/inode_goal
+Date:           June 2008
+Contact:        "Theodore Ts'o" <tytso@mit.edu>
+Description:
+                Tuning parameter which (if non-zero) controls the goal
+                inode used by the inode allocator in p0reference to
+                all other allocation hueristics.  This is intended for
+                debugging use only, and should be 0 on production
+                systems.
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index d035cf149e0e..746cdcba969d 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -863,6 +863,7 @@ struct ext4_sb_info {
         int s_inode_size;
         int s_first_ino;
         unsigned int s_inode_readahead_blks;
+        unsigned int s_inode_goal;
         spinlock_t s_next_gen_lock;
         u32 s_next_generation;
         u32 s_hash_seed[4];
@@ -1316,7 +1317,7 @@ extern int ext4fs_dirhash(const char *name, int len, struct
 
 /* ialloc.c */
 extern struct inode *ext4_new_inode(handle_t *, struct inode *, int,
-                                    const struct qstr *qstr);
+                                    const struct qstr *qstr, __u32 goal);
 extern void ext4_free_inode(handle_t *, struct inode *);
 extern struct inode * ext4_orphan_get(struct super_block *, unsigned long);
 extern unsigned long ext4_count_free_inodes(struct super_block *);
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 3f98ee712ff4..2f645732e3b7 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -799,7 +799,7 @@ err_ret:
  * group to find a free inode.
  */
 struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode,
-                             const struct qstr *qstr)
+                             const struct qstr *qstr, __u32 goal)
 {
         struct super_block *sb;
         struct buffer_head *inode_bitmap_bh = NULL;
@@ -830,6 +830,16 @@ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode,
         ei = EXT4_I(inode);
         sbi = EXT4_SB(sb);
 
+        if (!goal)
+                goal = sbi->s_inode_goal;
+
+        if (goal && goal < le32_to_cpu(sbi->s_es->s_inodes_count)) {
+                group = (goal - 1) / EXT4_INODES_PER_GROUP(sb);
+                ino = (goal - 1) % EXT4_INODES_PER_GROUP(sb);
+                ret2 = 0;
+                goto got_group;
+        }
+
         if (sbi->s_log_groups_per_flex && test_opt(sb, OLDALLOC)) {
                 ret2 = find_group_flex(sb, dir, &group);
                 if (ret2 == -1) {
@@ -858,7 +868,7 @@ got_group:
         if (ret2 == -1)
                 goto out;
 
-        for (i = 0; i < ngroups; i++) {
+        for (i = 0; i < ngroups; i++, ino = 0) {
                 err = -EIO;
 
                 gdp = ext4_get_group_desc(sb, group, &group_desc_bh);
@@ -870,8 +880,6 @@ got_group:
                 if (!inode_bitmap_bh)
                         goto fail;
 
-                ino = 0;
-
 repeat_in_this_group:
                 ino = ext4_find_next_zero_bit((unsigned long *)
                                               inode_bitmap_bh->b_data,
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 80d075b8aeaf..313a50b39741 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -458,6 +458,7 @@ int ext4_ext_migrate(struct inode *inode)
         struct inode *tmp_inode = NULL;
         struct list_blocks_struct lb;
         unsigned long max_entries;
+        __u32 goal;
 
         /*
          * If the filesystem does not support extents, or the inode
@@ -483,8 +484,10 @@ int ext4_ext_migrate(struct inode *inode)
                 retval = PTR_ERR(handle);
                 return retval;
         }
+        goal = (((inode->i_ino - 1) / EXT4_INODES_PER_GROUP(inode->i_sb)) *
+                EXT4_INODES_PER_GROUP(inode->i_sb)) + 1;
         tmp_inode = ext4_new_inode(handle, inode->i_sb->s_root->d_inode,
-                                   S_IFREG, 0);
+                                   S_IFREG, 0, goal);
         if (IS_ERR(tmp_inode)) {
                 retval = -ENOMEM;
                 ext4_journal_stop(handle);
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 5f00d2418a83..de04013d16ff 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -1782,7 +1782,7 @@ retry:
         if (IS_DIRSYNC(dir))
                 ext4_handle_sync(handle);
 
-        inode = ext4_new_inode(handle, dir, mode, &dentry->d_name);
+        inode = ext4_new_inode(handle, dir, mode, &dentry->d_name, 0);
         err = PTR_ERR(inode);
         if (!IS_ERR(inode)) {
                 inode->i_op = &ext4_file_inode_operations;
@@ -1816,7 +1816,7 @@ retry:
         if (IS_DIRSYNC(dir))
                 ext4_handle_sync(handle);
 
-        inode = ext4_new_inode(handle, dir, mode, &dentry->d_name);
+        inode = ext4_new_inode(handle, dir, mode, &dentry->d_name, 0);
         err = PTR_ERR(inode);
         if (!IS_ERR(inode)) {
                 init_special_inode(inode, inode->i_mode, rdev);
@@ -1853,7 +1853,8 @@ retry:
         if (IS_DIRSYNC(dir))
                 ext4_handle_sync(handle);
 
-        inode = ext4_new_inode(handle, dir, S_IFDIR | mode, &dentry->d_name);
+        inode = ext4_new_inode(handle, dir, S_IFDIR | mode,
+                               &dentry->d_name, 0);
         err = PTR_ERR(inode);
         if (IS_ERR(inode))
                 goto out_stop;
@@ -2264,7 +2265,8 @@ retry:
         if (IS_DIRSYNC(dir))
                 ext4_handle_sync(handle);
 
-        inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO, &dentry->d_name);
+        inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO,
+                               &dentry->d_name, 0);
         err = PTR_ERR(inode);
         if (IS_ERR(inode))
                 goto out_stop;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 04486a53469f..23013d303f81 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2206,6 +2206,7 @@ EXT4_RO_ATTR(session_write_kbytes);
 EXT4_RO_ATTR(lifetime_write_kbytes);
 EXT4_ATTR_OFFSET(inode_readahead_blks, 0644, sbi_ui_show,
                  inode_readahead_blks_store, s_inode_readahead_blks);
+EXT4_RW_ATTR_SBI_UI(inode_goal, s_inode_goal);
 EXT4_RW_ATTR_SBI_UI(mb_stats, s_mb_stats);
 EXT4_RW_ATTR_SBI_UI(mb_max_to_scan, s_mb_max_to_scan);
 EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb_min_to_scan);
@@ -2218,6 +2219,7 @@ static struct attribute *ext4_attrs[] = {
         ATTR_LIST(session_write_kbytes),
         ATTR_LIST(lifetime_write_kbytes),
         ATTR_LIST(inode_readahead_blks),
+        ATTR_LIST(inode_goal),
         ATTR_LIST(mb_stats),
         ATTR_LIST(mb_max_to_scan),
         ATTR_LIST(mb_min_to_scan),