nfsd/blocklayout: pretend we can send deviceid notifications

Commit df52699e4fcef ("NFSv4.1: Don't cache deviceids that have no
notifications") causes the Linux NFS client to stop caching deviceid's
unless a server pretends to support deviceid notifications.  While this
behavior is stupid and the language around this area in rfc5661 is a
mess carified by an errata that I submittted, Trond insists on this
behavior.  Not caching deviceids degrades block layout performance
massively as a GETDEVICEINFO is fairly expensive.

So add this hack to make the Linux client happy again.

Cc: stable@vger.kernel.org
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>

1 files changed, 11 insertions, 0 deletions

diff --git a/fs/nfsd/blocklayout.c b/fs/nfsd/blocklayout.c
index 03d647bf195d..cdefaa331a07 100644
--- a/fs/nfsd/blocklayout.c
+++ b/fs/nfsd/blocklayout.c
@@ -181,6 +181,17 @@ nfsd4_block_proc_layoutcommit(struct inode *inode,
 }
 
 const struct nfsd4_layout_ops bl_layout_ops = {
+        /*
+         * Pretend that we send notification to the client.  This is a blatant
+         * lie to force recent Linux clients to cache our device IDs.
+         * We rarely ever change the device ID, so the harm of leaking deviceids
+         * for a while isn't too bad.  Unfortunately RFC5661 is a complete mess
+         * in this regard, but I filed errata 4119 for this a while ago, and
+         * hopefully the Linux client will eventually start caching deviceids
+         * without this again.
+         */
+        .notify_types           =
+                        NOTIFY_DEVICEID4_DELETE | NOTIFY_DEVICEID4_CHANGE,
         .proc_getdeviceinfo     = nfsd4_block_proc_getdeviceinfo,
         .encode_getdeviceinfo   = nfsd4_block_encode_getdeviceinfo,
         .proc_layoutget         = nfsd4_block_proc_layoutget,