From: J. Bruce Fields <bfields@xxxxxxxxxxxxxx>
Instead of creating the new rpc client from a regular server thread,
set a flag, kick off a null call, and allow the null call to do the work
of setting up the client on the callback workqueue.
Use a spinlock to ensure the callback work gets a consistent view of the
callback parameters.
This allows, for example, changing the callback from contexts where
sleeping is not allowed. I hope it will also keep the locking simple as
we add more session and trunking features, by serializing most of the
callback-specific work.
This also closes a small race where the the new cb_ident could be used
with an old connection (or vice-versa).
Signed-off-by: J. Bruce Fields <bfields@xxxxxxxxxxxxxx>
---
fs/nfsd/nfs4callback.c | 73 ++++++++++++++++++++++++++++++++++--------------
fs/nfsd/nfs4state.c | 7 ++--
fs/nfsd/state.h | 10 +++++-
3 files changed, 63 insertions(+), 27 deletions(-)
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 384a990..4019b9a 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -284,7 +284,7 @@ nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p,
struct xdr_stream xdr;
struct nfs4_delegation *args = cb->cb_op;
struct nfs4_cb_compound_hdr hdr = {
- .ident = args->dl_ident,
+ .ident = cb->cb_clp->cl_cb_ident,
.minorversion = cb->cb_minorversion,
};
@@ -505,7 +505,8 @@ int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
PTR_ERR(client));
return PTR_ERR(client);
}
- nfsd4_set_callback_client(clp, client);
+ clp->cl_cb_ident = conn->cb_ident;
+ clp->cl_cb_client = client;
return 0;
}
@@ -568,15 +569,12 @@ void do_probe_callback(struct nfs4_client *clp)
*/
void nfsd4_probe_callback(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
{
- int status;
-
BUG_ON(atomic_read(&clp->cl_cb_set));
- status = setup_callback_client(clp, conn);
- if (status) {
- warn_no_callback_path(clp, status);
- return;
- }
+ spin_lock(&clp->cl_lock);
+ memcpy(&clp->cl_cb_conn, conn, sizeof(struct nfs4_cb_conn));
+ set_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_cb_flags);
+ spin_unlock(&clp->cl_lock);
do_probe_callback(clp);
}
@@ -729,19 +727,16 @@ void nfsd4_destroy_callback_queue(void)
}
/* must be called under the state lock */
-void nfsd4_set_callback_client(struct nfs4_client *clp, struct rpc_clnt *new)
+void nfsd4_shutdown_callback(struct nfs4_client *clp)
{
- struct rpc_clnt *old = clp->cl_cb_client;
-
- clp->cl_cb_client = new;
+ set_bit(NFSD4_CLIENT_KILL, &clp->cl_cb_flags);
/*
- * After this, any work that saw the old value of cl_cb_client will
- * be gone:
+ * Note this won't actually result in a null callback;
+ * instead, nfsd4_do_callback_rpc() will detect the killed
+ * client, destroy the rpc client, and stop:
*/
+ do_probe_callback(clp);
flush_workqueue(callback_wq);
- /* So we can safely shut it down: */
- if (old)
- rpc_shutdown_client(old);
}
void nfsd4_release_cb(struct nfsd4_callback *cb)
@@ -750,15 +745,51 @@ void nfsd4_release_cb(struct nfsd4_callback *cb)
cb->cb_ops->rpc_release(cb);
}
+void nfsd4_process_cb_update(struct nfsd4_callback *cb)
+{
+ struct nfs4_cb_conn conn;
+ struct nfs4_client *clp = cb->cb_clp;
+ int err;
+
+ /*
+ * This is either an update, or the client dying; in either case,
+ * kill the old client:
+ */
+ if (clp->cl_cb_client) {
+ rpc_shutdown_client(clp->cl_cb_client);
+ clp->cl_cb_client = NULL;
+ }
+ if (test_bit(NFSD4_CLIENT_KILL, &clp->cl_cb_flags))
+ return;
+ spin_lock(&clp->cl_lock);
+ /*
+ * Only serialized callback code is allowed to clear these
+ * flags; main nfsd code can only set them:
+ */
+ BUG_ON(!clp->cl_cb_flags);
+ clear_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_cb_flags);
+ memcpy(&conn, &cb->cb_clp->cl_cb_conn, sizeof(struct nfs4_cb_conn));
+ spin_unlock(&clp->cl_lock);
+
+ err = setup_callback_client(clp, &conn);
+ if (err)
+ warn_no_callback_path(clp, err);
+}
+
void nfsd4_do_callback_rpc(struct work_struct *w)
{
struct nfsd4_callback *cb = container_of(w, struct nfsd4_callback, cb_work);
struct nfs4_client *clp = cb->cb_clp;
- struct rpc_clnt *clnt = clp->cl_cb_client;
+ struct rpc_clnt *clnt;
- if (clnt == NULL) {
+ if (clp->cl_cb_flags)
+ nfsd4_process_cb_update(cb);
+
+ clnt = clp->cl_cb_client;
+ if (!clnt) {
+ /* Callback channel broken, or client killed; give up: */
nfsd4_release_cb(cb);
- return; /* Client is shutting down; give up. */
+ return;
}
rpc_call_async(clnt, &cb->cb_msg, RPC_TASK_SOFT | RPC_TASK_SOFTCONN,
cb->cb_ops, cb);
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 2f464fb..d3f12dc 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -207,7 +207,6 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
{
struct nfs4_delegation *dp;
struct nfs4_file *fp = stp->st_file;
- struct nfs4_cb_conn *conn = &stp->st_stateowner->so_client->cl_cb_conn;
dprintk("NFSD alloc_init_deleg\n");
/*
@@ -234,7 +233,6 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
nfs4_file_get_access(fp, O_RDONLY);
dp->dl_flock = NULL;
dp->dl_type = type;
- dp->dl_ident = conn->cb_ident;
dp->dl_stateid.si_boot = boot_time;
dp->dl_stateid.si_stateownerid = current_delegid++;
dp->dl_stateid.si_fileid = 0;
@@ -875,7 +873,7 @@ expire_client(struct nfs4_client *clp)
sop = list_entry(clp->cl_openowners.next, struct nfs4_stateowner, so_perclient);
release_openowner(sop);
}
- nfsd4_set_callback_client(clp, NULL);
+ nfsd4_shutdown_callback(clp);
if (clp->cl_cb_conn.cb_xprt)
svc_xprt_put(clp->cl_cb_conn.cb_xprt);
list_del(&clp->cl_idhash);
@@ -978,6 +976,7 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir,
INIT_LIST_HEAD(&clp->cl_delegations);
INIT_LIST_HEAD(&clp->cl_sessions);
INIT_LIST_HEAD(&clp->cl_lru);
+ spin_lock_init(&clp->cl_lock);
INIT_WORK(&clp->cl_cb_null.cb_work, nfsd4_do_callback_rpc);
clp->cl_time = get_seconds();
clear_bit(0, &clp->cl_cb_slot_busy);
@@ -1547,7 +1546,7 @@ nfsd4_destroy_session(struct svc_rqst *r,
nfs4_lock_state();
/* wait for callbacks */
- nfsd4_set_callback_client(ses->se_client, NULL);
+ nfsd4_shutdown_callback(ses->se_client);
nfs4_unlock_state();
nfsd4_put_session(ses);
status = nfs_ok;
diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index 2ece6be..58bc2a6 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -84,7 +84,6 @@ struct nfs4_delegation {
u32 dl_type;
time_t dl_time;
/* For recall: */
- u32 dl_ident;
stateid_t dl_stateid;
struct knfsd_fh dl_fh;
int dl_retries;
@@ -217,10 +216,17 @@ struct nfs4_client {
/* for v4.0 and v4.1 callbacks: */
struct nfs4_cb_conn cl_cb_conn;
+#define NFSD4_CLIENT_CB_UPDATE 1
+#define NFSD4_CLIENT_KILL 2
+ unsigned long cl_cb_flags;
struct rpc_clnt *cl_cb_client;
+ u32 cl_cb_ident;
atomic_t cl_cb_set;
struct nfsd4_callback cl_cb_null;
+ /* for all client information that callback code might need: */
+ spinlock_t cl_lock;
+
/* for nfs41 */
struct list_head cl_sessions;
struct nfsd4_clid_slot cl_cs_slot; /* create_session slot */
@@ -439,7 +445,7 @@ extern void nfsd4_do_callback_rpc(struct work_struct *);
extern void nfsd4_cb_recall(struct nfs4_delegation *dp);
extern int nfsd4_create_callback_queue(void);
extern void nfsd4_destroy_callback_queue(void);
-extern void nfsd4_set_callback_client(struct nfs4_client *, struct rpc_clnt *);
+extern void nfsd4_shutdown_callback(struct nfs4_client *);
extern void nfs4_put_delegation(struct nfs4_delegation *dp);
extern __be32 nfs4_make_rec_clidname(char *clidname, struct xdr_netobj *clname);
extern void nfsd4_init_recdir(char *recdir_name);
--
1.7.0.4
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
