ext4: fix ext4_free_inode() vs. ext4_claim_inode() race

(cherry picked from commit 7ce9d5d1f3c8736511daa413c64985a05b2feee3)

I was seeing fsck errors on inode bitmaps after a 4 thread
dbench run on a 4 cpu machine:

Inode bitmap differences: -50736 -(50752--50753) etc...

I believe that this is because ext4_free_inode() uses atomic
bitops, and although ext4_new_inode() *used* to also use atomic
bitops for synchronization, commit
393418676a7602e1d7d3f6e560159c65c8cbd50e changed this to use
the sb_bgl_lock, so that we could also synchronize against
read_inode_bitmap and initialization of uninit inode tables.

However, that change left ext4_free_inode using atomic bitops,
which I think leaves no synchronization between setting &
unsetting bits in the inode table.

The below patch fixes it for me, although I wonder if we're
getting at all heavy-handed with this spinlock...

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index cce841fe44909226b7a1f652141176f9482cb27b..f9b9fad959dca9dc680d6b3b4984aaa82dd87d66 100644 (file)
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -188,7 +188,7 @@ void ext4_free_inode (handle_t *handle, struct inode * inode)
        struct ext4_group_desc * gdp;
        struct ext4_super_block * es;
        struct ext4_sb_info *sbi;
-       int fatal = 0, err;
+       int fatal = 0, err, cleared;
        ext4_group_t flex_group;
 
        if (atomic_read(&inode->i_count) > 1) {
@@ -242,10 +242,12 @@ void ext4_free_inode (handle_t *handle, struct inode * inode)
                goto error_return;
 
        /* Ok, now we can actually update the inode bitmaps.. */
-       if (!ext4_clear_bit_atomic(sb_bgl_lock(sbi, block_group),
-                                       bit, bitmap_bh->b_data))
-               ext4_error (sb, "ext4_free_inode",
-                             "bit already cleared for inode %lu", ino);
+       spin_lock(sb_bgl_lock(sbi, block_group));
+       cleared = ext4_clear_bit(bit, bitmap_bh->b_data);
+       spin_unlock(sb_bgl_lock(sbi, block_group));
+       if (!cleared)
+               ext4_error(sb, "ext4_free_inode",
+                          "bit already cleared for inode %lu", ino);
        else {
                gdp = ext4_get_group_desc (sb, block_group, &bh2);