NFSv3: Ensure that do_proc_get_root() reports errors correctly
authorTrond Myklebust <Trond.Myklebust@netapp.com>
Mon, 20 Aug 2012 16:42:15 +0000 (12:42 -0400)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 14 Sep 2012 17:00:38 +0000 (10:00 -0700)
commit 086600430493e04b802bee6e5b3ce0458e4eb77f upstream.

If the rpc call to NFS3PROC_FSINFO fails, then we need to report that
error so that the mount fails. Otherwise we can end up with a
superblock with completely unusable values for block sizes, maxfilesize,
etc.

Reported-by: Yuanming Chen <hikvision_linux@163.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/nfs/nfs3proc.c

index 771741f14799de09ac17636fe5658a934e9d5adf..edfca53059b7fe8df2ae11882d6c8086a4be0460 100644 (file)
@@ -68,7 +68,7 @@ do_proc_get_root(struct rpc_clnt *client, struct nfs_fh *fhandle,
        nfs_fattr_init(info->fattr);
        status = rpc_call_sync(client, &msg, 0);
        dprintk("%s: reply fsinfo: %d\n", __func__, status);
-       if (!(info->fattr->valid & NFS_ATTR_FATTR)) {
+       if (status == 0 && !(info->fattr->valid & NFS_ATTR_FATTR)) {
                msg.rpc_proc = &nfs3_procedures[NFS3PROC_GETATTR];
                msg.rpc_resp = info->fattr;
                status = rpc_call_sync(client, &msg, 0);