NFSD: add delegation reaper to react to low memory condition

The delegation reaper is called by nfsd memory shrinker's on
the 'count' callback. It scans the client list and sends the
courtesy CB_RECALL_ANY to the clients that hold delegations.

To avoid flooding the clients with CB_RECALL_ANY requests, the
delegation reaper sends only one CB_RECALL_ANY request to each
client per 5 seconds.

Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
[ cel: moved definition of RCA4_TYPE_MASK_RDATA_DLG ]
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

3 files changed, 102 insertions, 4 deletions

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 00b3fa164fe8..c981704ec6a1 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -2144,6 +2144,7 @@ static void __free_client(struct kref *k)
 	kfree(clp->cl_nii_domain.data);
 	kfree(clp->cl_nii_name.data);
 	idr_destroy(&clp->cl_stateids);
+	kfree(clp->cl_ra);
 	kmem_cache_free(client_slab, clp);
 }
 
@@ -2871,6 +2872,36 @@ static const struct tree_descr client_files[] = {
 	[3] = {""},
 };
 
+static int
+nfsd4_cb_recall_any_done(struct nfsd4_callback *cb,
+				struct rpc_task *task)
+{
+	switch (task->tk_status) {
+	case -NFS4ERR_DELAY:
+		rpc_delay(task, 2 * HZ);
+		return 0;
+	default:
+		return 1;
+	}
+}
+
+static void
+nfsd4_cb_recall_any_release(struct nfsd4_callback *cb)
+{
+	struct nfs4_client *clp = cb->cb_clp;
+	struct nfsd_net *nn = net_generic(clp->net, nfsd_net_id);
+
+	spin_lock(&nn->client_lock);
+	clear_bit(NFSD4_CLIENT_CB_RECALL_ANY, &clp->cl_flags);
+	put_client_renew_locked(clp);
+	spin_unlock(&nn->client_lock);
+}
+
+static const struct nfsd4_callback_ops nfsd4_cb_recall_any_ops = {
+	.done		= nfsd4_cb_recall_any_done,
+	.release	= nfsd4_cb_recall_any_release,
+};
+
 static struct nfs4_client *create_client(struct xdr_netobj name,
 		struct svc_rqst *rqstp, nfs4_verifier *verf)
 {
@@ -2908,6 +2939,14 @@ static struct nfs4_client *create_client(struct xdr_netobj name,
 		free_client(clp);
 		return NULL;
 	}
+	clp->cl_ra = kzalloc(sizeof(*clp->cl_ra), GFP_KERNEL);
+	if (!clp->cl_ra) {
+		free_client(clp);
+		return NULL;
+	}
+	clp->cl_ra_time = 0;
+	nfsd4_init_cb(&clp->cl_ra->ra_cb, clp, &nfsd4_cb_recall_any_ops,
+			NFSPROC4_CLNT_CB_RECALL_ANY);
 	return clp;
 }
 
@@ -4363,14 +4402,16 @@ out:
 static unsigned long
 nfsd4_state_shrinker_count(struct shrinker *shrink, struct shrink_control *sc)
 {
-	int cnt;
+	int count;
 	struct nfsd_net *nn = container_of(shrink,
 			struct nfsd_net, nfsd_client_shrinker);
 
-	cnt = atomic_read(&nn->nfsd_courtesy_clients);
-	if (cnt > 0)
+	count = atomic_read(&nn->nfsd_courtesy_clients);
+	if (!count)
+		count = atomic_long_read(&num_delegations);
+	if (count)
 		mod_delayed_work(laundry_wq, &nn->nfsd_shrinker_work, 0);
-	return (unsigned long)cnt;
+	return (unsigned long)count;
 }
 
 static unsigned long
@@ -6160,6 +6201,44 @@ courtesy_client_reaper(struct nfsd_net *nn)
 }
 
 static void
+deleg_reaper(struct nfsd_net *nn)
+{
+	struct list_head *pos, *next;
+	struct nfs4_client *clp;
+	struct list_head cblist;
+
+	INIT_LIST_HEAD(&cblist);
+	spin_lock(&nn->client_lock);
+	list_for_each_safe(pos, next, &nn->client_lru) {
+		clp = list_entry(pos, struct nfs4_client, cl_lru);
+		if (clp->cl_state != NFSD4_ACTIVE ||
+			list_empty(&clp->cl_delegations) ||
+			atomic_read(&clp->cl_delegs_in_recall) ||
+			test_bit(NFSD4_CLIENT_CB_RECALL_ANY, &clp->cl_flags) ||
+			(ktime_get_boottime_seconds() -
+				clp->cl_ra_time < 5)) {
+			continue;
+		}
+		list_add(&clp->cl_ra_cblist, &cblist);
+
+		/* release in nfsd4_cb_recall_any_release */
+		atomic_inc(&clp->cl_rpc_users);
+		set_bit(NFSD4_CLIENT_CB_RECALL_ANY, &clp->cl_flags);
+		clp->cl_ra_time = ktime_get_boottime_seconds();
+	}
+	spin_unlock(&nn->client_lock);
+
+	while (!list_empty(&cblist)) {
+		clp = list_first_entry(&cblist, struct nfs4_client,
+					cl_ra_cblist);
+		list_del_init(&clp->cl_ra_cblist);
+		clp->cl_ra->ra_keep = 0;
+		clp->cl_ra->ra_bmval[0] = BIT(RCA4_TYPE_MASK_RDATA_DLG);
+		nfsd4_run_cb(&clp->cl_ra->ra_cb);
+	}
+}
+
+static void
 nfsd4_state_shrinker_worker(struct work_struct *work)
 {
 	struct delayed_work *dwork = to_delayed_work(work);
@@ -6167,6 +6246,7 @@ nfsd4_state_shrinker_worker(struct work_struct *work)
 				nfsd_shrinker_work);
 
 	courtesy_client_reaper(nn);
+	deleg_reaper(nn);
 }
 
 static inline __be32 nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stid *stp)
diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index e30882f8b851..e94634d30591 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -368,6 +368,7 @@ struct nfs4_client {
 #define NFSD4_CLIENT_UPCALL_LOCK	(5)	/* upcall serialization */
 #define NFSD4_CLIENT_CB_FLAG_MASK	(1 << NFSD4_CLIENT_CB_UPDATE | \
 					 1 << NFSD4_CLIENT_CB_KILL)
+#define NFSD4_CLIENT_CB_RECALL_ANY	(6)
 	unsigned long		cl_flags;
 	const struct cred	*cl_cb_cred;
 	struct rpc_clnt		*cl_cb_client;
@@ -411,6 +412,10 @@ struct nfs4_client {
 
 	unsigned int		cl_state;
 	atomic_t		cl_delegs_in_recall;
+
+	struct nfsd4_cb_recall_any	*cl_ra;
+	time64_t		cl_ra_time;
+	struct list_head	cl_ra_cblist;
 };
 
 /* struct nfs4_client_reset
diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h
index 8d04b6a5964c..730003c4f4af 100644
--- a/include/linux/nfs4.h
+++ b/include/linux/nfs4.h
@@ -732,4 +732,17 @@ enum nfs4_setxattr_options {
 	SETXATTR4_CREATE	= 1,
 	SETXATTR4_REPLACE	= 2,
 };
+
+enum {
+	RCA4_TYPE_MASK_RDATA_DLG	= 0,
+	RCA4_TYPE_MASK_WDATA_DLG	= 1,
+	RCA4_TYPE_MASK_DIR_DLG		= 2,
+	RCA4_TYPE_MASK_FILE_LAYOUT	= 3,
+	RCA4_TYPE_MASK_BLK_LAYOUT	= 4,
+	RCA4_TYPE_MASK_OBJ_LAYOUT_MIN	= 8,
+	RCA4_TYPE_MASK_OBJ_LAYOUT_MAX	= 9,
+	RCA4_TYPE_MASK_OTHER_LAYOUT_MIN	= 12,
+	RCA4_TYPE_MASK_OTHER_LAYOUT_MAX	= 15,
+};
+
 #endif