/*
* net/dccp/output.c
- *
+ *
* An implementation of the DCCP protocol
* Arnaldo Carvalho de Melo <acme@conectiva.com.br>
*
* 2 of the License, or (at your option) any later version.
*/
-#include <linux/config.h>
#include <linux/dccp.h>
#include <linux/kernel.h>
#include <linux/skbuff.h>
inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
}
-static inline void dccp_skb_entail(struct sock *sk, struct sk_buff *skb)
+static void dccp_skb_entail(struct sock *sk, struct sk_buff *skb)
{
skb_set_owner_w(skb, sk);
WARN_ON(sk->sk_send_head);
struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
struct dccp_hdr *dh;
/* XXX For now we're using only 48 bits sequence numbers */
- const int dccp_header_size = sizeof(*dh) +
+ const u32 dccp_header_size = sizeof(*dh) +
sizeof(struct dccp_hdr_ext) +
dccp_packet_hdr_len(dcb->dccpd_type);
int err, set_ack = 1;
set_ack = 0;
/* fall through */
case DCCP_PKT_DATAACK:
+ case DCCP_PKT_RESET:
break;
+ case DCCP_PKT_REQUEST:
+ set_ack = 0;
+ /* fall through */
+
case DCCP_PKT_SYNC:
case DCCP_PKT_SYNCACK:
- ackno = dcb->dccpd_seq;
+ ackno = dcb->dccpd_ack_seq;
/* fall through */
default:
/*
- * Only data packets should come through with skb->sk
- * set.
+ * Set owner/destructor: some skbs are allocated via
+ * alloc_skb (e.g. when retransmission may happen).
+ * Only Data, DataAck, and Reset packets should come
+ * through here with skb->sk set.
*/
WARN_ON(skb->sk);
skb_set_owner_w(skb, sk);
}
dcb->dccpd_seq = dp->dccps_gss;
- dccp_insert_options(sk, skb);
-
- skb->h.raw = skb_push(skb, dccp_header_size);
- dh = dccp_hdr(skb);
+
+ if (dccp_insert_options(sk, skb)) {
+ kfree_skb(skb);
+ return -EPROTO;
+ }
+
/* Build DCCP header and checksum it. */
- memset(dh, 0, dccp_header_size);
+ dh = dccp_zeroed_hdr(skb, dccp_header_size);
dh->dccph_type = dcb->dccpd_type;
dh->dccph_sport = inet->sport;
dh->dccph_dport = inet->dport;
dh->dccph_doff = (dccp_header_size + dcb->dccpd_opt_len) / 4;
dh->dccph_ccval = dcb->dccpd_ccval;
+ dh->dccph_cscov = dp->dccps_pcslen;
/* XXX For now we're using only 48 bits sequence numbers */
dh->dccph_x = 1;
break;
}
- icsk->icsk_af_ops->send_check(sk, skb->len, skb);
+ icsk->icsk_af_ops->send_check(sk, 0, skb);
if (set_ack)
dccp_event_ack_sent(sk);
memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
err = icsk->icsk_af_ops->queue_xmit(skb, 0);
- if (err <= 0)
- return err;
-
- /* NET_XMIT_CN is special. It does not guarantee,
- * that this packet is lost. It tells that device
- * is about to start to drop packets or already
- * drops some packets of the same priority and
- * invokes us to send less aggressively.
- */
- return err == NET_XMIT_CN ? 0 : err;
+ return net_xmit_eval(err);
}
return -ENOBUFS;
}
/**
* dccp_wait_for_ccid - Wait for ccid to tell us we can send a packet
- * @sk: socket to wait for
- * @timeo: for how long
+ * @sk: socket to wait for
+ * @skb: current skb to pass on for waiting
+ * @delay: sleep timeout in milliseconds (> 0)
+ * This function is called by default when the socket is closed, and
+ * when a non-zero linger time is set on the socket. For consistency
*/
-static int dccp_wait_for_ccid(struct sock *sk, struct sk_buff *skb,
- long *timeo)
+static int dccp_wait_for_ccid(struct sock *sk, struct sk_buff *skb, int delay)
{
struct dccp_sock *dp = dccp_sk(sk);
DEFINE_WAIT(wait);
- long delay;
+ unsigned long jiffdelay;
int rc;
- while (1) {
- prepare_to_wait(sk->sk_sleep, &wait, TASK_INTERRUPTIBLE);
+ do {
+ dccp_pr_debug("delayed send by %d msec\n", delay);
+ jiffdelay = msecs_to_jiffies(delay);
- if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN))
- goto do_error;
- if (!*timeo)
- goto do_nonblock;
- if (signal_pending(current))
- goto do_interrupted;
-
- rc = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb,
- skb->len);
- if (rc <= 0)
- break;
- delay = msecs_to_jiffies(rc);
- if (delay > *timeo || delay < 0)
- goto do_nonblock;
+ prepare_to_wait(sk->sk_sleep, &wait, TASK_INTERRUPTIBLE);
sk->sk_write_pending++;
release_sock(sk);
- *timeo -= schedule_timeout(delay);
+ schedule_timeout(jiffdelay);
lock_sock(sk);
sk->sk_write_pending--;
- }
+
+ if (sk->sk_err)
+ goto do_error;
+ if (signal_pending(current))
+ goto do_interrupted;
+
+ rc = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb);
+ } while ((delay = rc) > 0);
out:
finish_wait(sk->sk_sleep, &wait);
return rc;
do_error:
rc = -EPIPE;
goto out;
-do_nonblock:
- rc = -EAGAIN;
- goto out;
do_interrupted:
- rc = sock_intr_errno(*timeo);
+ rc = -EINTR;
goto out;
}
-int dccp_write_xmit(struct sock *sk, struct sk_buff *skb, long *timeo)
+void dccp_write_xmit(struct sock *sk, int block)
{
- const struct dccp_sock *dp = dccp_sk(sk);
- int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb,
- skb->len);
+ struct dccp_sock *dp = dccp_sk(sk);
+ struct sk_buff *skb;
- if (err > 0)
- err = dccp_wait_for_ccid(sk, skb, timeo);
+ while ((skb = skb_peek(&sk->sk_write_queue))) {
+ int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb);
+
+ if (err > 0) {
+ if (!block) {
+ sk_reset_timer(sk, &dp->dccps_xmit_timer,
+ msecs_to_jiffies(err)+jiffies);
+ break;
+ } else
+ err = dccp_wait_for_ccid(sk, skb, err);
+ if (err && err != -EINTR)
+ DCCP_BUG("err=%d after dccp_wait_for_ccid", err);
+ }
- if (err == 0) {
- struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
- const int len = skb->len;
+ skb_dequeue(&sk->sk_write_queue);
+ if (err == 0) {
+ struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
+ const int len = skb->len;
- if (sk->sk_state == DCCP_PARTOPEN) {
- /* See 8.1.5. Handshake Completion */
- inet_csk_schedule_ack(sk);
- inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
+ if (sk->sk_state == DCCP_PARTOPEN) {
+ /* See 8.1.5. Handshake Completion */
+ inet_csk_schedule_ack(sk);
+ inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
inet_csk(sk)->icsk_rto,
DCCP_RTO_MAX);
- dcb->dccpd_type = DCCP_PKT_DATAACK;
- } else if (dccp_ack_pending(sk))
- dcb->dccpd_type = DCCP_PKT_DATAACK;
- else
- dcb->dccpd_type = DCCP_PKT_DATA;
-
- err = dccp_transmit_skb(sk, skb);
- ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len);
- } else
- kfree_skb(skb);
-
- return err;
+ dcb->dccpd_type = DCCP_PKT_DATAACK;
+ } else if (dccp_ack_pending(sk))
+ dcb->dccpd_type = DCCP_PKT_DATAACK;
+ else
+ dcb->dccpd_type = DCCP_PKT_DATA;
+
+ err = dccp_transmit_skb(sk, skb);
+ ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len);
+ if (err)
+ DCCP_BUG("err=%d after ccid_hc_tx_packet_sent",
+ err);
+ } else {
+ dccp_pr_debug("packet discarded due to err=%d\n", err);
+ kfree_skb(skb);
+ }
+ }
}
int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
{
struct dccp_hdr *dh;
struct dccp_request_sock *dreq;
- const int dccp_header_size = sizeof(struct dccp_hdr) +
+ const u32 dccp_header_size = sizeof(struct dccp_hdr) +
sizeof(struct dccp_hdr_ext) +
sizeof(struct dccp_hdr_response);
- struct sk_buff *skb = sock_wmalloc(sk, MAX_HEADER + DCCP_MAX_OPT_LEN +
- dccp_header_size, 1,
+ struct sk_buff *skb = sock_wmalloc(sk, sk->sk_prot->max_header, 1,
GFP_ATOMIC);
if (skb == NULL)
return NULL;
/* Reserve space for headers. */
- skb_reserve(skb, MAX_HEADER + DCCP_MAX_OPT_LEN + dccp_header_size);
+ skb_reserve(skb, sk->sk_prot->max_header);
skb->dst = dst_clone(dst);
- skb->csum = 0;
dreq = dccp_rsk(req);
+ if (inet_rsk(req)->acked) /* increase ISS upon retransmission */
+ dccp_inc_seqno(&dreq->dreq_iss);
DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_RESPONSE;
DCCP_SKB_CB(skb)->dccpd_seq = dreq->dreq_iss;
- dccp_insert_options(sk, skb);
- skb->h.raw = skb_push(skb, dccp_header_size);
+ if (dccp_insert_options(sk, skb)) {
+ kfree_skb(skb);
+ return NULL;
+ }
- dh = dccp_hdr(skb);
- memset(dh, 0, dccp_header_size);
+ /* Build and checksum header */
+ dh = dccp_zeroed_hdr(skb, dccp_header_size);
dh->dccph_sport = inet_sk(sk)->sport;
dh->dccph_dport = inet_rsk(req)->rmt_port;
dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dreq->dreq_isr);
dccp_hdr_response(skb)->dccph_resp_service = dreq->dreq_service;
- dh->dccph_checksum = dccp_v4_checksum(skb, inet_rsk(req)->loc_addr,
- inet_rsk(req)->rmt_addr);
+ dccp_csum_outgoing(skb);
+ /* We use `acked' to remember that a Response was already sent. */
+ inet_rsk(req)->acked = 1;
DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
return skb;
}
EXPORT_SYMBOL_GPL(dccp_make_response);
-struct sk_buff *dccp_make_reset(struct sock *sk, struct dst_entry *dst,
- const enum dccp_reset_codes code)
-
+/* answer offending packet in @rcv_skb with Reset from control socket @ctl */
+struct sk_buff *dccp_ctl_make_reset(struct socket *ctl, struct sk_buff *rcv_skb)
{
- struct dccp_hdr *dh;
- struct dccp_sock *dp = dccp_sk(sk);
- const int dccp_header_size = sizeof(struct dccp_hdr) +
- sizeof(struct dccp_hdr_ext) +
- sizeof(struct dccp_hdr_reset);
- struct sk_buff *skb = sock_wmalloc(sk, MAX_HEADER + DCCP_MAX_OPT_LEN +
- dccp_header_size, 1,
- GFP_ATOMIC);
+ struct dccp_hdr *rxdh = dccp_hdr(rcv_skb), *dh;
+ struct dccp_skb_cb *dcb = DCCP_SKB_CB(rcv_skb);
+ const u32 dccp_hdr_reset_len = sizeof(struct dccp_hdr) +
+ sizeof(struct dccp_hdr_ext) +
+ sizeof(struct dccp_hdr_reset);
+ struct dccp_hdr_reset *dhr;
+ struct sk_buff *skb;
+
+ skb = alloc_skb(ctl->sk->sk_prot->max_header, GFP_ATOMIC);
if (skb == NULL)
return NULL;
- /* Reserve space for headers. */
- skb_reserve(skb, MAX_HEADER + DCCP_MAX_OPT_LEN + dccp_header_size);
+ skb_reserve(skb, ctl->sk->sk_prot->max_header);
- skb->dst = dst_clone(dst);
- skb->csum = 0;
+ /* Swap the send and the receive. */
+ dh = dccp_zeroed_hdr(skb, dccp_hdr_reset_len);
+ dh->dccph_type = DCCP_PKT_RESET;
+ dh->dccph_sport = rxdh->dccph_dport;
+ dh->dccph_dport = rxdh->dccph_sport;
+ dh->dccph_doff = dccp_hdr_reset_len / 4;
+ dh->dccph_x = 1;
- dccp_inc_seqno(&dp->dccps_gss);
+ dhr = dccp_hdr_reset(skb);
+ dhr->dccph_reset_code = dcb->dccpd_reset_code;
+
+ switch (dcb->dccpd_reset_code) {
+ case DCCP_RESET_CODE_PACKET_ERROR:
+ dhr->dccph_reset_data[0] = rxdh->dccph_type;
+ break;
+ case DCCP_RESET_CODE_OPTION_ERROR: /* fall through */
+ case DCCP_RESET_CODE_MANDATORY_ERROR:
+ memcpy(dhr->dccph_reset_data, dcb->dccpd_reset_data, 3);
+ break;
+ }
+ /*
+ * From RFC 4340, 8.3.1:
+ * If P.ackno exists, set R.seqno := P.ackno + 1.
+ * Else set R.seqno := 0.
+ */
+ if (dcb->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ)
+ dccp_hdr_set_seq(dh, ADD48(dcb->dccpd_ack_seq, 1));
+ dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dcb->dccpd_seq);
- DCCP_SKB_CB(skb)->dccpd_reset_code = code;
- DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_RESET;
- DCCP_SKB_CB(skb)->dccpd_seq = dp->dccps_gss;
- dccp_insert_options(sk, skb);
+ dccp_csum_outgoing(skb);
+ return skb;
+}
- skb->h.raw = skb_push(skb, dccp_header_size);
+EXPORT_SYMBOL_GPL(dccp_ctl_make_reset);
- dh = dccp_hdr(skb);
- memset(dh, 0, dccp_header_size);
+/* send Reset on established socket, to close or abort the connection */
+int dccp_send_reset(struct sock *sk, enum dccp_reset_codes code)
+{
+ struct sk_buff *skb;
+ /*
+ * FIXME: what if rebuild_header fails?
+ * Should we be doing a rebuild_header here?
+ */
+ int err = inet_sk_rebuild_header(sk);
- dh->dccph_sport = inet_sk(sk)->sport;
- dh->dccph_dport = inet_sk(sk)->dport;
- dh->dccph_doff = (dccp_header_size +
- DCCP_SKB_CB(skb)->dccpd_opt_len) / 4;
- dh->dccph_type = DCCP_PKT_RESET;
- dh->dccph_x = 1;
- dccp_hdr_set_seq(dh, dp->dccps_gss);
- dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dp->dccps_gsr);
+ if (err != 0)
+ return err;
- dccp_hdr_reset(skb)->dccph_reset_code = code;
+ skb = sock_wmalloc(sk, sk->sk_prot->max_header, 1, GFP_ATOMIC);
+ if (skb == NULL)
+ return -ENOBUFS;
- dh->dccph_checksum = dccp_v4_checksum(skb, inet_sk(sk)->saddr,
- inet_sk(sk)->daddr);
+ /* Reserve space for headers and prepare control bits. */
+ skb_reserve(skb, sk->sk_prot->max_header);
+ DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_RESET;
+ DCCP_SKB_CB(skb)->dccpd_reset_code = code;
- DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
- return skb;
+ return dccp_transmit_skb(sk, skb);
}
/*
sk->sk_err = 0;
sock_reset_flag(sk, SOCK_DONE);
-
+
dccp_sync_mss(sk, dst_mtu(dst));
- dccp_update_gss(sk, dp->dccps_iss);
- /*
+ /*
* SWL and AWL are initially adjusted so that they are not less than
* the initial Sequence Numbers received and sent, respectively:
* SWL := max(GSR + 1 - floor(W/4), ISR),
* AWL := max(GSS - W' + 1, ISS).
* These adjustments MUST be applied only at the beginning of the
* connection.
- */
+ */
+ dccp_update_gss(sk, dp->dccps_iss);
dccp_set_seqno(&dp->dccps_awl, max48(dp->dccps_awl, dp->dccps_iss));
+ /* S.GAR - greatest valid acknowledgement number received on a non-Sync;
+ * initialized to S.ISS (sec. 8.5) */
+ dp->dccps_gar = dp->dccps_iss;
+
icsk->icsk_retransmits = 0;
}
dccp_connect_init(sk);
- skb = alloc_skb(MAX_DCCP_HEADER + 15, sk->sk_allocation);
+ skb = alloc_skb(sk->sk_prot->max_header, sk->sk_allocation);
if (unlikely(skb == NULL))
return -ENOBUFS;
/* Reserve space for headers. */
- skb_reserve(skb, MAX_DCCP_HEADER);
+ skb_reserve(skb, sk->sk_prot->max_header);
DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_REQUEST;
- skb->csum = 0;
dccp_skb_entail(sk, skb);
dccp_transmit_skb(sk, skb_clone(skb, GFP_KERNEL));
{
/* If we have been reset, we may not send again. */
if (sk->sk_state != DCCP_CLOSED) {
- struct sk_buff *skb = alloc_skb(MAX_DCCP_HEADER, GFP_ATOMIC);
+ struct sk_buff *skb = alloc_skb(sk->sk_prot->max_header,
+ GFP_ATOMIC);
if (skb == NULL) {
inet_csk_schedule_ack(sk);
}
/* Reserve space for headers */
- skb_reserve(skb, MAX_DCCP_HEADER);
- skb->csum = 0;
+ skb_reserve(skb, sk->sk_prot->max_header);
DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_ACK;
dccp_transmit_skb(sk, skb);
}
EXPORT_SYMBOL_GPL(dccp_send_ack);
+/* FIXME: Is this still necessary (11.3) - currently nowhere used by DCCP. */
void dccp_send_delayed_ack(struct sock *sk)
{
struct inet_connection_sock *icsk = inet_csk(sk);
sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout);
}
-void dccp_send_sync(struct sock *sk, const u64 seq,
+void dccp_send_sync(struct sock *sk, const u64 ackno,
const enum dccp_pkt_type pkt_type)
{
/*
* dccp_transmit_skb() will set the ownership to this
* sock.
*/
- struct sk_buff *skb = alloc_skb(MAX_DCCP_HEADER, GFP_ATOMIC);
+ struct sk_buff *skb = alloc_skb(sk->sk_prot->max_header, GFP_ATOMIC);
- if (skb == NULL)
+ if (skb == NULL) {
/* FIXME: how to make sure the sync is sent? */
+ DCCP_CRIT("could not send %s", dccp_packet_name(pkt_type));
return;
+ }
/* Reserve space for headers and prepare control bits. */
- skb_reserve(skb, MAX_DCCP_HEADER);
- skb->csum = 0;
+ skb_reserve(skb, sk->sk_prot->max_header);
DCCP_SKB_CB(skb)->dccpd_type = pkt_type;
- DCCP_SKB_CB(skb)->dccpd_seq = seq;
+ DCCP_SKB_CB(skb)->dccpd_ack_seq = ackno;
dccp_transmit_skb(sk, skb);
}
+EXPORT_SYMBOL_GPL(dccp_send_sync);
+
/*
* Send a DCCP_PKT_CLOSE/CLOSEREQ. The caller locks the socket for us. This
* cannot be allowed to fail queueing a DCCP_PKT_CLOSE/CLOSEREQ frame under
/* Reserve space for headers and prepare control bits. */
skb_reserve(skb, sk->sk_prot->max_header);
- skb->csum = 0;
DCCP_SKB_CB(skb)->dccpd_type = dp->dccps_role == DCCP_ROLE_CLIENT ?
DCCP_PKT_CLOSE : DCCP_PKT_CLOSEREQ;
if (active) {
+ dccp_write_xmit(sk, 1);
dccp_skb_entail(sk, skb);
dccp_transmit_skb(sk, skb_clone(skb, prio));
+ /* FIXME do we need a retransmit timer here? */
} else
dccp_transmit_skb(sk, skb);
}