NFSD: Add nfsd4_copy time-to-live

Keep async copy state alive for a few lease cycles after the copy
completes so that OFFLOAD_STATUS returns something meaningful.

This means that NFSD's client shutdown processing needs to purge
any of this state that happens to be waiting to die.

Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

1 files changed, 15 insertions, 0 deletions

diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index dcbebd53e5f4..8092894e8f3c 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -137,6 +137,21 @@ struct nfs4_cpntf_state {
 	time64_t		cpntf_time;	/* last time stateid used */
 };
 
+/*
+ * RFC 7862 Section 4.8 states:
+ *
+ * | A copy offload stateid will be valid until either (A) the client
+ * | or server restarts or (B) the client returns the resource by
+ * | issuing an OFFLOAD_CANCEL operation or the client replies to a
+ * | CB_OFFLOAD operation.
+ *
+ * Because a client might not reply to a CB_OFFLOAD, or a reply
+ * might get lost due to connection loss, NFSD purges async copy
+ * state after a short period to prevent it from accumulating
+ * over time.
+ */
+#define NFSD_COPY_INITIAL_TTL 10
+
 struct nfs4_cb_fattr {
 	struct nfsd4_callback ncf_getattr;
 	u32 ncf_cb_status;