[Ocfs2-devel] [PATCH 1/2] ocfs2: clear links count in ocfs2_mknod() if an error occurs

[Andrew Morton]

I've been sitting on this patch series for over two years!  I assume
the problem isn't terribly serious.

Can we please come to a decision on how to proceed?  Thanks.

The current patches from the mm tree are below.

From: Wangyan <wangyan122 at huawei.com>
Subject: ocfs2: clear links count in ocfs2_mknod() if an error occurs

In this condition, the inode can not be wiped when error happened.
ocfs2_mkdir()
  ->ocfs2_mknod()
    ->ocfs2_mknod_locked()
      ->__ocfs2_mknod_locked()
        ->ocfs2_set_links_count() // i_links_count is 2
    -> ... // an error accrue, goto roll_back or leave.
    ->ocfs2_commit_trans()
    ->iput(inode)
      ->evict()
        ->ocfs2_evict_inode()
          ->ocfs2_delete_inode()
            ->ocfs2_inode_lock()
              ->ocfs2_inode_lock_update()
                ->ocfs2_refresh_inode()
                  ->set_nlink();    // inode->i_nlink is 2 now.
            /* if wipe is 0, it will goto bail_unlock_inode */
            ->ocfs2_query_inode_wipe()
              ->if (inode->i_nlink) return; // wipe is 0.
            /* inode can not be wiped */
            ->ocfs2_wipe_inode()
So, we need clear links before the transaction committed.

Link: http://lkml.kernel.org/r/d8147c41-fb2b-bdf7-b660-1f3c8448c33f@huawei.com
Signed-off-by: Yan Wang <wangyan122 at huawei.com>
Reviewed-by: Jun Piao <piaojun at huawei.com>
Cc: Mark Fasheh <mark at fasheh.com>
Cc: Joel Becker <jlbec at evilplan.org>
Cc: Junxiao Bi <junxiao.bi at oracle.com>
Cc: Joseph Qi <jiangqi903 at gmail.com>
Cc: Changwei Ge <gechangwei at live.cn>
Cc: Gang He <ghe at suse.com>
Cc: <stable at vger.kernel.org>
Signed-off-by: Andrew Morton <akpm at linux-foundation.org>
---

 fs/ocfs2/namei.c |   14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

--- a/fs/ocfs2/namei.c~ocfs2-clear-links-count-in-ocfs2_mknod-if-an-error-occurs
+++ a/fs/ocfs2/namei.c
@@ -454,8 +454,12 @@ roll_back:
 leave:
 	if (status < 0 && did_quota_inode)
 		dquot_free_inode(inode);
-	if (handle)
+	if (handle) {
+		if (status < 0 && new_fe_bh != NULL)
+			ocfs2_set_links_count((struct ocfs2_dinode *)
+					new_fe_bh->b_data, 0);
 		ocfs2_commit_trans(osb, handle);
+	}
 
 	ocfs2_inode_unlock(dir, 1);
 	if (did_block_signals)
@@ -599,6 +603,8 @@ static int __ocfs2_mknod_locked(struct i
 leave:
 	if (status < 0) {
 		if (*new_fe_bh) {
+			if (fe)
+				ocfs2_set_links_count(fe, 0);
 			brelse(*new_fe_bh);
 			*new_fe_bh = NULL;
 		}
@@ -2028,8 +2034,12 @@ bail:
 					ocfs2_clusters_to_bytes(osb->sb, 1));
 	if (status < 0 && did_quota_inode)
 		dquot_free_inode(inode);
-	if (handle)
+	if (handle) {
+		if (status < 0 && new_fe_bh != NULL)
+			ocfs2_set_links_count((struct ocfs2_dinode *)
+					new_fe_bh->b_data, 0);
 		ocfs2_commit_trans(osb, handle);
+	}
 
 	ocfs2_inode_unlock(dir, 1);
 	if (did_block_signals)
_





From: Wangyan <wangyan122 at huawei.com>
Subject: ocfs2: fix ocfs2 corrupt when iputting an inode

In this condition, it will cause an bug on error.
ocfs2_mkdir()
  ->ocfs2_mknod()
    ->ocfs2_mknod_locked()
      ->__ocfs2_mknod_locked()
        //Assume inode->i_generation is genN.
        ->inode->i_generation = osb->s_next_generation++;
        // The inode lockres has been initialized.
        ->ocfs2_populate_inode()
        ->ocfs2_create_new_inode_locks()
            ->An error happened, returned value is non-zero
      // free the start_bit x in bg_blkno
      ->ocfs2_free_suballoc_bits()
    ->...  /* Another process execute mkdir success in this place,
              and it occupied the start_bit x in bg_blkno
              which has been freed before. Its inode->i_generation
              is genN + 1 */
    ->iput(inode)
      ->evict()
        ->ocfs2_evict_inode()
          ->ocfs2_delete_inode()
            ->ocfs2_inode_lock()
              ->ocfs2_inode_lock_update()
                /* Bug on here, genN != genN + 1 */
                ->mlog_bug_on_msg(inode->i_generation !=
                  le32_to_cpu(fe->i_generation))

So, we need not to reclaim the inode when the inode->ip_inode_lockres
has been initialized. It will be freed in iput().

Link: http://lkml.kernel.org/r/ef080ca3-5d74-e276-17a1-d9e7c7e662c9@huawei.com
Fixes: b1529a41f777 ("ocfs2: should reclaim the inode if '__ocfs2_mknod_locked' returns an error")
Signed-off-by: Yan Wang <wangyan122 at huawei.com>
Reviewed-by: Jun Piao <piaojun at huawei.com>
Cc: Mark Fasheh <mark at fasheh.com>
Cc: Joel Becker <jlbec at evilplan.org>
Cc: Junxiao Bi <junxiao.bi at oracle.com>
Cc: Joseph Qi <jiangqi903 at gmail.com>
Cc: Changwei Ge <gechangwei at live.cn>
Cc: Gang He <ghe at suse.com>
Cc: <stable at vger.kernel.org>
Signed-off-by: Andrew Morton <akpm at linux-foundation.org>
---

 fs/ocfs2/namei.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

--- a/fs/ocfs2/namei.c~ocfs2-fix-ocfs2-corrupt-when-iputting-an-inode
+++ a/fs/ocfs2/namei.c
@@ -641,7 +641,8 @@ static int ocfs2_mknod_locked(struct ocf
 	status = __ocfs2_mknod_locked(dir, inode, dev, new_fe_bh,
 				    parent_fe_bh, handle, inode_ac,
 				    fe_blkno, suballoc_loc, suballoc_bit);
-	if (status < 0) {
+	if (status < 0 && !(OCFS2_I(inode)->ip_inode_lockres.l_flags &
+				OCFS2_LOCK_INITIALIZED)) {
 		u64 bg_blkno = ocfs2_which_suballoc_group(fe_blkno, suballoc_bit);
 		int tmp = ocfs2_free_suballoc_bits(handle, inode_ac->ac_inode,
 				inode_ac->ac_bh, suballoc_bit, bg_blkno, 1);
_