Fix the nested PR lock calling issue in ACL

Hi,

Thanks a lot for all the review and comments so far;) I'd like to send
the improved (V4) version of this patch.

This patch fixes a deadlock in OCFS2 ACL. We found this bug in OCFS2
and Samba integration using scenario, the symptom is several smbd
processes will be hung under heavy workload. Finally we found out it
is the nested PR lock calling that leads to this deadlock:

 node1        node2
              gr PR
                |
                V
 PR(EX)---> BAST:OCFS2_LOCK_BLOCKED
                |
                V
              rq PR
                |
                V
              wait=1

After requesting the 2nd PR lock, the process "smbd" went into D
state. It can only be woken up when the 1st PR lock's RO holder equals
zero. There should be an ocfs2_inode_unlock in the calling path later
on, which can decrement the RO holder. But since it has been in
uninterruptible sleep, the unlock function has no chance to be called.

The related stack trace is:
smbd          D ffff8800013d0600     0  9522   5608 0x00000000
 ffff88002ca7fb18 0000000000000282 ffff88002f964500 ffff88002ca7fa98
 ffff8800013d0600 ffff88002ca7fae0 ffff88002f964340 ffff88002f964340
 ffff88002ca7ffd8 ffff88002ca7ffd8 ffff88002f964340 ffff88002f964340
Call Trace:
[<ffffffff80350425>] schedule_timeout+0x175/0x210
[<ffffffff8034f580>] wait_for_common+0xf0/0x210
[<ffffffffa03e12b9>] __ocfs2_cluster_lock+0x3b9/0xa90 [ocfs2]
[<ffffffffa03e7665>] ocfs2_inode_lock_full_nested+0x255/0xdb0 [ocfs2]
[<ffffffffa0446019>] ocfs2_get_acl+0x69/0x120 [ocfs2]
[<ffffffffa0446368>] ocfs2_check_acl+0x28/0x80 [ocfs2]
[<ffffffff800e3507>] acl_permission_check+0x57/0xb0
[<ffffffff800e357d>] generic_permission+0x1d/0xc0
[<ffffffffa03eecea>] ocfs2_permission+0x10a/0x1d0 [ocfs2]
[<ffffffff800e3f65>] inode_permission+0x45/0x100
[<ffffffff800d86b3>] sys_chdir+0x53/0x90
[<ffffffff80007458>] system_call_fastpath+0x16/0x1b
[<00007f34a4ef6927>] 0x7f34a4ef6927

For details, please see:
https://bugzilla.novell.com/show_bug.cgi?id=614332 and
http://oss.oracle.com/bugzilla/show_bug.cgi?id=1278

Signed-off-by: Jiaju Zhang <jjzhang@suse.de>
Acked-by: Mark Fasheh <mfasheh@suse.com>
Cc: stable@kernel.org
Signed-off-by: Joel Becker <joel.becker@oracle.com>

diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c
index 5a21778..a76e0aa 100644 (file)
--- a/fs/ocfs2/acl.c
+++ b/fs/ocfs2/acl.c
@@ -290,12 +290,30 @@ static int ocfs2_set_acl(handle_t *handle,
 
 int ocfs2_check_acl(struct inode *inode, int mask)
 {
-       struct posix_acl *acl = ocfs2_get_acl(inode, ACL_TYPE_ACCESS);
+       struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+       struct buffer_head *di_bh = NULL;
+       struct posix_acl *acl;
+       int ret = -EAGAIN;
 
-       if (IS_ERR(acl))
+       if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+               return ret;
+
+       ret = ocfs2_read_inode_block(inode, &di_bh);
+       if (ret < 0) {
+               mlog_errno(ret);
+               return ret;
+       }
+
+       acl = ocfs2_get_acl_nolock(inode, ACL_TYPE_ACCESS, di_bh);
+
+       brelse(di_bh);
+
+       if (IS_ERR(acl)) {
+               mlog_errno(PTR_ERR(acl));
                return PTR_ERR(acl);
+       }
        if (acl) {
-               int ret = posix_acl_permission(inode, acl, mask);
+               ret = posix_acl_permission(inode, acl, mask);
                posix_acl_release(acl);
                return ret;
        }