NFSD: add shrinker to reap courtesy clients on low memory condition

Add courtesy_client_reaper to react to low memory condition triggered
by the system memory shrinker.

The delayed_work for the courtesy_client_reaper is scheduled on
the shrinker's count callback using the laundry_wq.

The shrinker's scan callback is not used for expiring the courtesy
clients due to potential deadlocks.

Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

diff --git a/fs/nfsd/netns.h b/fs/nfsd/netns.h
index 55c7006d6109a709d2a70ebd4851651c5fbf717d..8c854ba3285bbcddfe62010dc6b93018fdc3927c 100644 (file)
--- a/fs/nfsd/netns.h
+++ b/fs/nfsd/netns.h
@@ -194,6 +194,8 @@ struct nfsd_net {
        int                     nfs4_max_clients;
 
        atomic_t                nfsd_courtesy_clients;
+       struct shrinker         nfsd_client_shrinker;
+       struct delayed_work     nfsd_shrinker_work;
 };
 
 /* Simple check to find out if a given net was properly initialized */

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index a248ecfe89c8ec7dc6ea7fadfcf7bb95d4a42b89..56b50210da38bb6ace471437e522c1e7cbf42759 100644 (file)
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -4349,7 +4349,27 @@ out:
        return -ENOMEM;
 }
 
-void nfsd4_init_leases_net(struct nfsd_net *nn)
+static unsigned long
+nfsd_courtesy_client_count(struct shrinker *shrink, struct shrink_control *sc)
+{
+       int cnt;
+       struct nfsd_net *nn = container_of(shrink,
+                       struct nfsd_net, nfsd_client_shrinker);
+
+       cnt = atomic_read(&nn->nfsd_courtesy_clients);
+       if (cnt > 0)
+               mod_delayed_work(laundry_wq, &nn->nfsd_shrinker_work, 0);
+       return (unsigned long)cnt;
+}
+
+static unsigned long
+nfsd_courtesy_client_scan(struct shrinker *shrink, struct shrink_control *sc)
+{
+       return SHRINK_STOP;
+}
+
+int
+nfsd4_init_leases_net(struct nfsd_net *nn)
 {
        struct sysinfo si;
        u64 max_clients;
@@ -4370,6 +4390,16 @@ void nfsd4_init_leases_net(struct nfsd_net *nn)
        nn->nfs4_max_clients = max_t(int, max_clients, NFS4_CLIENTS_PER_GB);
 
        atomic_set(&nn->nfsd_courtesy_clients, 0);
+       nn->nfsd_client_shrinker.scan_objects = nfsd_courtesy_client_scan;
+       nn->nfsd_client_shrinker.count_objects = nfsd_courtesy_client_count;
+       nn->nfsd_client_shrinker.seeks = DEFAULT_SEEKS;
+       return register_shrinker(&nn->nfsd_client_shrinker, "nfsd-client");
+}
+
+void
+nfsd4_leases_net_shutdown(struct nfsd_net *nn)
+{
+       unregister_shrinker(&nn->nfsd_client_shrinker);
 }
 
 static void init_nfs4_replay(struct nfs4_replay *rp)
@@ -5942,10 +5972,49 @@ exp_client:
        spin_unlock(&nn->client_lock);
 }
 
+static void
+nfs4_get_courtesy_client_reaplist(struct nfsd_net *nn,
+                               struct list_head *reaplist)
+{
+       unsigned int maxreap = 0, reapcnt = 0;
+       struct list_head *pos, *next;
+       struct nfs4_client *clp;
+
+       maxreap = NFSD_CLIENT_MAX_TRIM_PER_RUN;
+       INIT_LIST_HEAD(reaplist);
+
+       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)
+                       break;
+               if (reapcnt >= maxreap)
+                       break;
+               if (!mark_client_expired_locked(clp)) {
+                       list_add(&clp->cl_lru, reaplist);
+                       reapcnt++;
+               }
+       }
+       spin_unlock(&nn->client_lock);
+}
+
+static void
+nfs4_process_client_reaplist(struct list_head *reaplist)
+{
+       struct list_head *pos, *next;
+       struct nfs4_client *clp;
+
+       list_for_each_safe(pos, next, reaplist) {
+               clp = list_entry(pos, struct nfs4_client, cl_lru);
+               trace_nfsd_clid_purged(&clp->cl_clientid);
+               list_del_init(&clp->cl_lru);
+               expire_client(clp);
+       }
+}
+
 static time64_t
 nfs4_laundromat(struct nfsd_net *nn)
 {
-       struct nfs4_client *clp;
        struct nfs4_openowner *oo;
        struct nfs4_delegation *dp;
        struct nfs4_ol_stateid *stp;
@@ -5974,12 +6043,8 @@ nfs4_laundromat(struct nfsd_net *nn)
        }
        spin_unlock(&nn->s2s_cp_lock);
        nfs4_get_client_reaplist(nn, &reaplist, &lt);
-       list_for_each_safe(pos, next, &reaplist) {
-               clp = list_entry(pos, struct nfs4_client, cl_lru);
-               trace_nfsd_clid_purged(&clp->cl_clientid);
-               list_del_init(&clp->cl_lru);
-               expire_client(clp);
-       }
+       nfs4_process_client_reaplist(&reaplist);
+
        spin_lock(&state_lock);
        list_for_each_safe(pos, next, &nn->del_recall_lru) {
                dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru);
@@ -6062,6 +6127,18 @@ laundromat_main(struct work_struct *laundry)
        queue_delayed_work(laundry_wq, &nn->laundromat_work, t*HZ);
 }
 
+static void
+courtesy_client_reaper(struct work_struct *reaper)
+{
+       struct list_head reaplist;
+       struct delayed_work *dwork = to_delayed_work(reaper);
+       struct nfsd_net *nn = container_of(dwork, struct nfsd_net,
+                                       nfsd_shrinker_work);
+
+       nfs4_get_courtesy_client_reaplist(nn, &reaplist);
+       nfs4_process_client_reaplist(&reaplist);
+}
+
 static inline __be32 nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stid *stp)
 {
        if (!fh_match(&fhp->fh_handle, &stp->sc_file->fi_fhandle))
@@ -7879,6 +7956,7 @@ static int nfs4_state_create_net(struct net *net)
        INIT_LIST_HEAD(&nn->blocked_locks_lru);
 
        INIT_DELAYED_WORK(&nn->laundromat_work, laundromat_main);
+       INIT_DELAYED_WORK(&nn->nfsd_shrinker_work, courtesy_client_reaper);
        get_net(net);
 
        return 0;

diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 917fa1892fd2d5c12faea564499cbcff2514e4ec..597a26ad4183fb33dd5d36c2ada1f749c3e445a0 100644 (file)
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -1481,11 +1481,12 @@ static __net_init int nfsd_init_net(struct net *net)
                goto out_idmap_error;
        nn->nfsd_versions = NULL;
        nn->nfsd4_minorversions = NULL;
+       retval = nfsd4_init_leases_net(nn);
+       if (retval)
+               goto out_drc_error;
        retval = nfsd_reply_cache_init(nn);
        if (retval)
                goto out_drc_error;
-       nfsd4_init_leases_net(nn);
-
        get_random_bytes(&nn->siphash_key, sizeof(nn->siphash_key));
        seqlock_init(&nn->writeverf_lock);
 
@@ -1507,6 +1508,7 @@ static __net_exit void nfsd_exit_net(struct net *net)
        nfsd_idmap_shutdown(net);
        nfsd_export_shutdown(net);
        nfsd_netns_free_versions(net_generic(net, nfsd_net_id));
+       nfsd4_leases_net_shutdown(nn);
 }
 
 static struct pernet_operations nfsd_net_ops = {

diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
index 6ab4ad41ae84e0fb65bf429211970f9f4864a182..09726c5b9a317bbafe823c1c716e963d7a19f238 100644 (file)
--- a/fs/nfsd/nfsd.h
+++ b/fs/nfsd/nfsd.h
@@ -505,7 +505,8 @@ extern void unregister_cld_notifier(void);
 extern void nfsd4_ssc_init_umount_work(struct nfsd_net *nn);
 #endif
 
-extern void nfsd4_init_leases_net(struct nfsd_net *nn);
+extern int nfsd4_init_leases_net(struct nfsd_net *nn);
+extern void nfsd4_leases_net_shutdown(struct nfsd_net *nn);
 
 #else /* CONFIG_NFSD_V4 */
 static inline int nfsd4_is_junction(struct dentry *dentry)
@@ -513,7 +514,8 @@ static inline int nfsd4_is_junction(struct dentry *dentry)
        return 0;
 }
 
-static inline void nfsd4_init_leases_net(struct nfsd_net *nn) {};
+static inline int nfsd4_init_leases_net(struct nfsd_net *nn) { return 0; };
+static inline void nfsd4_leases_net_shutdown(struct nfsd_net *nn) {};
 
 #define register_cld_notifier() 0
 #define unregister_cld_notifier() do { } while(0)