gfs2: Partially revert gfs2_inode_lookup change
authorAndreas Gruenbacher <agruenba@redhat.com>
Mon, 5 Dec 2022 21:27:28 +0000 (22:27 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 31 Dec 2022 12:32:45 +0000 (13:32 +0100)
[ Upstream commit 88f4a9f813c549f6b8a6fbf12030949b48a4d5a4 ]

Commit c412a97cf6c5 changed delete_work_func() to always perform an
inode lookup when gfs2_try_evict() fails.  This doesn't make sense as a
gfs2_try_evict() failure indicates that the inode is likely still in
use.  Revert that change.

Fixes: c412a97cf6c5 ("gfs2: Use TRY lock in gfs2_inode_lookup for UNLINKED inodes")
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
fs/gfs2/glock.c

index df335c2..235a094 100644 (file)
@@ -1039,6 +1039,7 @@ static void delete_work_func(struct work_struct *work)
                        if (gfs2_queue_delete_work(gl, 5 * HZ))
                                return;
                }
+               goto out;
        }
 
        inode = gfs2_lookup_by_inum(sdp, no_addr, gl->gl_no_formal_ino,
@@ -1051,6 +1052,7 @@ static void delete_work_func(struct work_struct *work)
                d_prune_aliases(inode);
                iput(inode);
        }
+out:
        gfs2_glock_put(gl);
 }