xfs: unset MS_ACTIVE if mount fails
authorDarrick J. Wong <darrick.wong@oracle.com>
Thu, 20 Oct 2016 04:45:40 +0000 (15:45 +1100)
committerDave Chinner <david@fromorbit.com>
Thu, 20 Oct 2016 04:45:40 +0000 (15:45 +1100)
commitd099245297e28fc9f8493edd9d5a1f0967a72511
treec871c85d824b3dccc91771f574db0128a11ef70d
parentfe23759eaf2f6540de20c1623f066aad967ff9c9
xfs: unset MS_ACTIVE if mount fails

As part of the inode block map intent log item recovery process, we
had to set the IRECOVERY flag to prevent an unlinked inode from
being truncated during the first iput call.  This required us to set
MS_ACTIVE so that iput puts the inode on the lru instead of
immediately evicting the inode.

Unfortunately, if the mount fails later on, the inodes that have
been loaded (root dir and realtime) actually need to be evicted
since we're aborting the mount.  If we don't clear MS_ACTIVE in the
failure step, those inodes are not evicted and therefore leak.   The
leak was found by running xfs/130 and rmmoding xfs immediately after
the test.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
fs/xfs/xfs_mount.c