aboutsummaryrefslogtreecommitdiff
path: root/include
diff options
context:
space:
mode:
Diffstat (limited to 'include')
-rw-r--r--include/linux/nfs_fs.h4
-rw-r--r--include/linux/sunrpc/debug.h3
-rw-r--r--include/linux/sunrpc/msg_prot.h25
-rw-r--r--include/linux/sunrpc/xdr.h6
-rw-r--r--include/linux/sunrpc/xprt.h221
5 files changed, 176 insertions, 83 deletions
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 9a6047ff1b2..7bac2785c6e 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -41,6 +41,10 @@
#define NFS_MAX_FILE_IO_BUFFER_SIZE 32768
#define NFS_DEF_FILE_IO_BUFFER_SIZE 4096
+/* Default timeout values */
+#define NFS_MAX_UDP_TIMEOUT (60*HZ)
+#define NFS_MAX_TCP_TIMEOUT (600*HZ)
+
/*
* superblock magic number for NFS
*/
diff --git a/include/linux/sunrpc/debug.h b/include/linux/sunrpc/debug.h
index eadb31e3c19..1a42d902bc1 100644
--- a/include/linux/sunrpc/debug.h
+++ b/include/linux/sunrpc/debug.h
@@ -32,6 +32,7 @@
#define RPCDBG_AUTH 0x0010
#define RPCDBG_PMAP 0x0020
#define RPCDBG_SCHED 0x0040
+#define RPCDBG_TRANS 0x0080
#define RPCDBG_SVCSOCK 0x0100
#define RPCDBG_SVCDSP 0x0200
#define RPCDBG_MISC 0x0400
@@ -94,6 +95,8 @@ enum {
CTL_NLMDEBUG,
CTL_SLOTTABLE_UDP,
CTL_SLOTTABLE_TCP,
+ CTL_MIN_RESVPORT,
+ CTL_MAX_RESVPORT,
};
#endif /* _LINUX_SUNRPC_DEBUG_H_ */
diff --git a/include/linux/sunrpc/msg_prot.h b/include/linux/sunrpc/msg_prot.h
index 15f11533238..f43f237360a 100644
--- a/include/linux/sunrpc/msg_prot.h
+++ b/include/linux/sunrpc/msg_prot.h
@@ -76,5 +76,30 @@ enum rpc_auth_stat {
#define RPC_MAXNETNAMELEN 256
+/*
+ * From RFC 1831:
+ *
+ * "A record is composed of one or more record fragments. A record
+ * fragment is a four-byte header followed by 0 to (2**31) - 1 bytes of
+ * fragment data. The bytes encode an unsigned binary number; as with
+ * XDR integers, the byte order is from highest to lowest. The number
+ * encodes two values -- a boolean which indicates whether the fragment
+ * is the last fragment of the record (bit value 1 implies the fragment
+ * is the last fragment) and a 31-bit unsigned binary value which is the
+ * length in bytes of the fragment's data. The boolean value is the
+ * highest-order bit of the header; the length is the 31 low-order bits.
+ * (Note that this record specification is NOT in XDR standard form!)"
+ *
+ * The Linux RPC client always sends its requests in a single record
+ * fragment, limiting the maximum payload size for stream transports to
+ * 2GB.
+ */
+
+typedef u32 rpc_fraghdr;
+
+#define RPC_LAST_STREAM_FRAGMENT (1U << 31)
+#define RPC_FRAGMENT_SIZE_MASK (~RPC_LAST_STREAM_FRAGMENT)
+#define RPC_MAX_FRAGMENT_SIZE ((1U << 31) - 1)
+
#endif /* __KERNEL__ */
#endif /* _LINUX_SUNRPC_MSGPROT_H_ */
diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h
index 23448d0fb5b..5da968729cf 100644
--- a/include/linux/sunrpc/xdr.h
+++ b/include/linux/sunrpc/xdr.h
@@ -161,14 +161,10 @@ typedef struct {
typedef size_t (*skb_read_actor_t)(skb_reader_t *desc, void *to, size_t len);
+extern int csum_partial_copy_to_xdr(struct xdr_buf *, struct sk_buff *);
extern ssize_t xdr_partial_copy_from_skb(struct xdr_buf *, unsigned int,
skb_reader_t *, skb_read_actor_t);
-struct socket;
-struct sockaddr;
-extern int xdr_sendpages(struct socket *, struct sockaddr *, int,
- struct xdr_buf *, unsigned int, int);
-
extern int xdr_encode_word(struct xdr_buf *, int, u32);
extern int xdr_decode_word(struct xdr_buf *, int, u32 *);
diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index e618c164981..99cad3ead81 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -1,5 +1,5 @@
/*
- * linux/include/linux/sunrpc/clnt_xprt.h
+ * linux/include/linux/sunrpc/xprt.h
*
* Declarations for the RPC transport interface.
*
@@ -15,20 +15,6 @@
#include <linux/sunrpc/sched.h>
#include <linux/sunrpc/xdr.h>
-/*
- * The transport code maintains an estimate on the maximum number of out-
- * standing RPC requests, using a smoothed version of the congestion
- * avoidance implemented in 44BSD. This is basically the Van Jacobson
- * congestion algorithm: If a retransmit occurs, the congestion window is
- * halved; otherwise, it is incremented by 1/cwnd when
- *
- * - a reply is received and
- * - a full number of requests are outstanding and
- * - the congestion window hasn't been updated recently.
- *
- * Upper procedures may check whether a request would block waiting for
- * a free RPC slot by using the RPC_CONGESTED() macro.
- */
extern unsigned int xprt_udp_slot_table_entries;
extern unsigned int xprt_tcp_slot_table_entries;
@@ -36,34 +22,23 @@ extern unsigned int xprt_tcp_slot_table_entries;
#define RPC_DEF_SLOT_TABLE (16U)
#define RPC_MAX_SLOT_TABLE (128U)
-#define RPC_CWNDSHIFT (8U)
-#define RPC_CWNDSCALE (1U << RPC_CWNDSHIFT)
-#define RPC_INITCWND RPC_CWNDSCALE
-#define RPC_MAXCWND(xprt) ((xprt)->max_reqs << RPC_CWNDSHIFT)
-#define RPCXPRT_CONGESTED(xprt) ((xprt)->cong >= (xprt)->cwnd)
-
-/* Default timeout values */
-#define RPC_MAX_UDP_TIMEOUT (60*HZ)
-#define RPC_MAX_TCP_TIMEOUT (600*HZ)
-
/*
- * Wait duration for an RPC TCP connection to be established. Solaris
- * NFS over TCP uses 60 seconds, for example, which is in line with how
- * long a server takes to reboot.
+ * RPC call and reply header size as number of 32bit words (verifier
+ * size computed separately)
*/
-#define RPC_CONNECT_TIMEOUT (60*HZ)
+#define RPC_CALLHDRSIZE 6
+#define RPC_REPHDRSIZE 4
/*
- * Delay an arbitrary number of seconds before attempting to reconnect
- * after an error.
+ * Parameters for choosing a free port
*/
-#define RPC_REESTABLISH_TIMEOUT (15*HZ)
+extern unsigned int xprt_min_resvport;
+extern unsigned int xprt_max_resvport;
-/* RPC call and reply header size as number of 32bit words (verifier
- * size computed separately)
- */
-#define RPC_CALLHDRSIZE 6
-#define RPC_REPHDRSIZE 4
+#define RPC_MIN_RESVPORT (1U)
+#define RPC_MAX_RESVPORT (65535U)
+#define RPC_DEF_MIN_RESVPORT (650U)
+#define RPC_DEF_MAX_RESVPORT (1023U)
/*
* This describes a timeout strategy
@@ -76,6 +51,9 @@ struct rpc_timeout {
unsigned char to_exponential;
};
+struct rpc_task;
+struct rpc_xprt;
+
/*
* This describes a complete RPC request
*/
@@ -121,12 +99,21 @@ struct rpc_rqst {
#define rq_svec rq_snd_buf.head
#define rq_slen rq_snd_buf.len
-#define XPRT_LAST_FRAG (1 << 0)
-#define XPRT_COPY_RECM (1 << 1)
-#define XPRT_COPY_XID (1 << 2)
-#define XPRT_COPY_DATA (1 << 3)
+struct rpc_xprt_ops {
+ void (*set_buffer_size)(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize);
+ int (*reserve_xprt)(struct rpc_task *task);
+ void (*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
+ void (*connect)(struct rpc_task *task);
+ int (*send_request)(struct rpc_task *task);
+ void (*set_retrans_timeout)(struct rpc_task *task);
+ void (*timer)(struct rpc_task *task);
+ void (*release_request)(struct rpc_task *task);
+ void (*close)(struct rpc_xprt *xprt);
+ void (*destroy)(struct rpc_xprt *xprt);
+};
struct rpc_xprt {
+ struct rpc_xprt_ops * ops; /* transport methods */
struct socket * sock; /* BSD socket layer */
struct sock * inet; /* INET layer */
@@ -137,11 +124,13 @@ struct rpc_xprt {
unsigned long cong; /* current congestion */
unsigned long cwnd; /* congestion window */
- unsigned int rcvsize, /* socket receive buffer size */
- sndsize; /* socket send buffer size */
+ size_t rcvsize, /* transport rcv buffer size */
+ sndsize; /* transport send buffer size */
size_t max_payload; /* largest RPC payload size,
in bytes */
+ unsigned int tsh_size; /* size of transport specific
+ header */
struct rpc_wait_queue sending; /* requests waiting to send */
struct rpc_wait_queue resend; /* requests waiting to resend */
@@ -150,11 +139,9 @@ struct rpc_xprt {
struct list_head free; /* free slots */
struct rpc_rqst * slot; /* slot table storage */
unsigned int max_reqs; /* total slots */
- unsigned long sockstate; /* Socket state */
+ unsigned long state; /* transport state */
unsigned char shutdown : 1, /* being shut down */
- nocong : 1, /* no congestion control */
- resvport : 1, /* use a reserved port */
- stream : 1; /* TCP */
+ resvport : 1; /* use a reserved port */
/*
* XID
@@ -171,22 +158,27 @@ struct rpc_xprt {
unsigned long tcp_copied, /* copied to request */
tcp_flags;
/*
- * Connection of sockets
+ * Connection of transports
*/
- struct work_struct sock_connect;
+ unsigned long connect_timeout,
+ bind_timeout,
+ reestablish_timeout;
+ struct work_struct connect_worker;
unsigned short port;
+
/*
- * Disconnection of idle sockets
+ * Disconnection of idle transports
*/
struct work_struct task_cleanup;
struct timer_list timer;
- unsigned long last_used;
+ unsigned long last_used,
+ idle_timeout;
/*
* Send stuff
*/
- spinlock_t sock_lock; /* lock socket info */
- spinlock_t xprt_lock; /* lock xprt info */
+ spinlock_t transport_lock; /* lock transport info */
+ spinlock_t reserve_lock; /* lock slot table */
struct rpc_task * snd_task; /* Task blocked in send */
struct list_head recv;
@@ -195,37 +187,110 @@ struct rpc_xprt {
void (*old_data_ready)(struct sock *, int);
void (*old_state_change)(struct sock *);
void (*old_write_space)(struct sock *);
-
- wait_queue_head_t cong_wait;
};
+#define XPRT_LAST_FRAG (1 << 0)
+#define XPRT_COPY_RECM (1 << 1)
+#define XPRT_COPY_XID (1 << 2)
+#define XPRT_COPY_DATA (1 << 3)
+
#ifdef __KERNEL__
-struct rpc_xprt * xprt_create_proto(int proto, struct sockaddr_in *addr,
- struct rpc_timeout *toparms);
-int xprt_destroy(struct rpc_xprt *);
-void xprt_set_timeout(struct rpc_timeout *, unsigned int,
- unsigned long);
+/*
+ * Transport operations used by ULPs
+ */
+struct rpc_xprt * xprt_create_proto(int proto, struct sockaddr_in *addr, struct rpc_timeout *to);
+void xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr);
-void xprt_reserve(struct rpc_task *);
-int xprt_prepare_transmit(struct rpc_task *);
-void xprt_transmit(struct rpc_task *);
-void xprt_receive(struct rpc_task *);
+/*
+ * Generic internal transport functions
+ */
+void xprt_connect(struct rpc_task *task);
+void xprt_reserve(struct rpc_task *task);
+int xprt_reserve_xprt(struct rpc_task *task);
+int xprt_reserve_xprt_cong(struct rpc_task *task);
+int xprt_prepare_transmit(struct rpc_task *task);
+void xprt_transmit(struct rpc_task *task);
int xprt_adjust_timeout(struct rpc_rqst *req);
-void xprt_release(struct rpc_task *);
-void xprt_connect(struct rpc_task *);
-void xprt_sock_setbufsize(struct rpc_xprt *);
-
-#define XPRT_LOCKED 0
-#define XPRT_CONNECT 1
-#define XPRT_CONNECTING 2
-
-#define xprt_connected(xp) (test_bit(XPRT_CONNECT, &(xp)->sockstate))
-#define xprt_set_connected(xp) (set_bit(XPRT_CONNECT, &(xp)->sockstate))
-#define xprt_test_and_set_connected(xp) (test_and_set_bit(XPRT_CONNECT, &(xp)->sockstate))
-#define xprt_test_and_clear_connected(xp) \
- (test_and_clear_bit(XPRT_CONNECT, &(xp)->sockstate))
-#define xprt_clear_connected(xp) (clear_bit(XPRT_CONNECT, &(xp)->sockstate))
+void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task);
+void xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task);
+void xprt_release(struct rpc_task *task);
+int xprt_destroy(struct rpc_xprt *xprt);
+
+static inline u32 *xprt_skip_transport_header(struct rpc_xprt *xprt, u32 *p)
+{
+ return p + xprt->tsh_size;
+}
+
+/*
+ * Transport switch helper functions
+ */
+void xprt_set_retrans_timeout_def(struct rpc_task *task);
+void xprt_set_retrans_timeout_rtt(struct rpc_task *task);
+void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status);
+void xprt_wait_for_buffer_space(struct rpc_task *task);
+void xprt_write_space(struct rpc_xprt *xprt);
+void xprt_update_rtt(struct rpc_task *task);
+void xprt_adjust_cwnd(struct rpc_task *task, int result);
+struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid);
+void xprt_complete_rqst(struct rpc_task *task, int copied);
+void xprt_release_rqst_cong(struct rpc_task *task);
+void xprt_disconnect(struct rpc_xprt *xprt);
+
+/*
+ * Socket transport setup operations
+ */
+int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to);
+int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to);
+
+/*
+ * Reserved bit positions in xprt->state
+ */
+#define XPRT_LOCKED (0)
+#define XPRT_CONNECTED (1)
+#define XPRT_CONNECTING (2)
+
+static inline void xprt_set_connected(struct rpc_xprt *xprt)
+{
+ set_bit(XPRT_CONNECTED, &xprt->state);
+}
+
+static inline void xprt_clear_connected(struct rpc_xprt *xprt)
+{
+ clear_bit(XPRT_CONNECTED, &xprt->state);
+}
+
+static inline int xprt_connected(struct rpc_xprt *xprt)
+{
+ return test_bit(XPRT_CONNECTED, &xprt->state);
+}
+
+static inline int xprt_test_and_set_connected(struct rpc_xprt *xprt)
+{
+ return test_and_set_bit(XPRT_CONNECTED, &xprt->state);
+}
+
+static inline int xprt_test_and_clear_connected(struct rpc_xprt *xprt)
+{
+ return test_and_clear_bit(XPRT_CONNECTED, &xprt->state);
+}
+
+static inline void xprt_clear_connecting(struct rpc_xprt *xprt)
+{
+ smp_mb__before_clear_bit();
+ clear_bit(XPRT_CONNECTING, &xprt->state);
+ smp_mb__after_clear_bit();
+}
+
+static inline int xprt_connecting(struct rpc_xprt *xprt)
+{
+ return test_bit(XPRT_CONNECTING, &xprt->state);
+}
+
+static inline int xprt_test_and_set_connecting(struct rpc_xprt *xprt)
+{
+ return test_and_set_bit(XPRT_CONNECTING, &xprt->state);
+}
#endif /* __KERNEL__*/