xfs: remove dead ENODEV handling in xfs_destroy_ioend
authorChristoph Hellwig <hch@infradead.org>
Tue, 23 Aug 2011 08:28:09 +0000 (08:28 +0000)
committerAlex Elder <aelder@sgi.com>
Wed, 12 Oct 2011 02:15:00 +0000 (21:15 -0500)
No driver returns ENODEV from it bio completion handler, not has this
ever been documented.  Remove the dead code dealing with it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Alex Elder <aelder@sgi.com>
fs/xfs/xfs_aops.c

index 8c37dde4c521403cf9e91b6f0d5bcbbf94f04c12..d91564404abfa945a9d223f3b683ad9c5e626ced 100644 (file)
@@ -122,17 +122,6 @@ xfs_destroy_ioend(
                bh->b_end_io(bh, !ioend->io_error);
        }
 
-       /*
-        * Volume managers supporting multiple paths can send back ENODEV
-        * when the final path disappears.  In this case continuing to fill
-        * the page cache with dirty data which cannot be written out is
-        * evil, so prevent that.
-        */
-       if (unlikely(ioend->io_error == -ENODEV)) {
-               xfs_do_force_shutdown(ip->i_mount, SHUTDOWN_DEVICE_REQ,
-                                     __FILE__, __LINE__);
-       }
-
        xfs_ioend_wake(ip);
        mempool_free(ioend, xfs_ioend_pool);
 }