staging/lustre/nfs: writing to new files will return ENOENT
authorPatrick Farrell <paf@cray.com>
Tue, 26 Nov 2013 02:04:57 +0000 (10:04 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 26 Nov 2013 03:14:47 +0000 (19:14 -0800)
commitad8dbc93a464869f64365a2123b3491965df3b3e
treeaa58ec17c3449cee32c4d15c2634c808be8e5f0b
parent65f1c7816db85db02b2069f9d0e5111de52f5ee3
staging/lustre/nfs: writing to new files will return ENOENT

This happend with SLES11SP2 Lustre client, which in turn acts as an
NFS server, exporting a subtree of an Lustre fs through NFS.

We detected that whenever we are writing to a new file using, fx,
'echo blah > newfile', it will return ENOENT error. We found
out that this was caused by the anonymous dentry. In SLESS11SP2,
anonymous dentries are assigned '/' as the name, instead of an
empty string. When MDT handles the intent_open call, it will look
up the obj by the name if it is not an empty string, and thus
couldn't find it.

As MDS_OPEN_BY_FID is always set on this request, we never need
to send the name in this request.  The fid is already available
and should be used in case the file has been renamed.

Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3544
Lustre-change: http://review.whamcloud.com/6920
Signed-off-by: Cheng Shao <cheng_shao@xyratex.com>
Signed-off-by: Patrick Farrell <paf@cray.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Alexey Shvetsov <alexxy@gentoo.org>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Peng Tao <bergwolf@gmail.com>
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/staging/lustre/lustre/llite/file.c