summaryrefslogtreecommitdiffstats
path: root/net/dccp
diff options
context:
space:
mode:
Diffstat (limited to 'net/dccp')
-rw-r--r--net/dccp/ccids/ccid3.c14
-rw-r--r--net/dccp/ccids/lib/loss_interval.c10
-rw-r--r--net/dccp/ccids/lib/packet_history.c103
-rw-r--r--net/dccp/ccids/lib/packet_history.h30
-rw-r--r--net/dccp/dccp.h19
-rw-r--r--net/dccp/input.c2
-rw-r--r--net/dccp/ipv4.c30
-rw-r--r--net/dccp/ipv6.c33
-rw-r--r--net/dccp/options.c14
-rw-r--r--net/dccp/output.c53
-rw-r--r--net/dccp/proto.c8
-rw-r--r--net/dccp/timer.c24
12 files changed, 173 insertions, 167 deletions
diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c
index a1929f33d70..f6756e0c9e6 100644
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -794,7 +794,7 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb)
{
struct ccid3_hc_rx_sock *hcrx = ccid3_hc_rx_sk(sk);
enum ccid3_fback_type do_feedback = CCID3_FBACK_NONE;
- const u32 ndp = dccp_sk(sk)->dccps_options_received.dccpor_ndp;
+ const u64 ndp = dccp_sk(sk)->dccps_options_received.dccpor_ndp;
const bool is_data_packet = dccp_data_packet(skb);
if (unlikely(hcrx->ccid3hcrx_state == TFRC_RSTATE_NO_DATA)) {
@@ -825,18 +825,16 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb)
}
/*
- * Handle pending losses and otherwise check for new loss
+ * Perform loss detection and handle pending losses
*/
- if (tfrc_rx_hist_loss_pending(&hcrx->ccid3hcrx_hist) &&
- tfrc_rx_handle_loss(&hcrx->ccid3hcrx_hist,
- &hcrx->ccid3hcrx_li_hist,
- skb, ndp, ccid3_first_li, sk) ) {
+ if (tfrc_rx_handle_loss(&hcrx->ccid3hcrx_hist, &hcrx->ccid3hcrx_li_hist,
+ skb, ndp, ccid3_first_li, sk)) {
do_feedback = CCID3_FBACK_PARAM_CHANGE;
goto done_receiving;
}
- if (tfrc_rx_hist_new_loss_indicated(&hcrx->ccid3hcrx_hist, skb, ndp))
- goto update_records;
+ if (tfrc_rx_hist_loss_pending(&hcrx->ccid3hcrx_hist))
+ return; /* done receiving */
/*
* Handle data packets: RTT sampling and monitoring p
diff --git a/net/dccp/ccids/lib/loss_interval.c b/net/dccp/ccids/lib/loss_interval.c
index 849e181e698..bcd6ac415bb 100644
--- a/net/dccp/ccids/lib/loss_interval.c
+++ b/net/dccp/ccids/lib/loss_interval.c
@@ -90,14 +90,14 @@ u8 tfrc_lh_update_i_mean(struct tfrc_loss_hist *lh, struct sk_buff *skb)
{
struct tfrc_loss_interval *cur = tfrc_lh_peek(lh);
u32 old_i_mean = lh->i_mean;
- s64 length;
+ s64 len;
if (cur == NULL) /* not initialised */
return 0;
- length = dccp_delta_seqno(cur->li_seqno, DCCP_SKB_CB(skb)->dccpd_seq);
+ len = dccp_delta_seqno(cur->li_seqno, DCCP_SKB_CB(skb)->dccpd_seq) + 1;
- if (length - cur->li_length <= 0) /* duplicate or reordered */
+ if (len - (s64)cur->li_length <= 0) /* duplicate or reordered */
return 0;
if (SUB16(dccp_hdr(skb)->dccph_ccval, cur->li_ccval) > 4)
@@ -114,7 +114,7 @@ u8 tfrc_lh_update_i_mean(struct tfrc_loss_hist *lh, struct sk_buff *skb)
if (tfrc_lh_length(lh) == 1) /* due to RFC 3448, 6.3.1 */
return 0;
- cur->li_length = length;
+ cur->li_length = len;
tfrc_lh_calc_i_mean(lh);
return (lh->i_mean < old_i_mean);
@@ -159,7 +159,7 @@ int tfrc_lh_interval_add(struct tfrc_loss_hist *lh, struct tfrc_rx_hist *rh,
else {
cur->li_length = dccp_delta_seqno(cur->li_seqno, new->li_seqno);
new->li_length = dccp_delta_seqno(new->li_seqno,
- tfrc_rx_hist_last_rcv(rh)->tfrchrx_seqno);
+ tfrc_rx_hist_last_rcv(rh)->tfrchrx_seqno) + 1;
if (lh->counter > (2*LIH_SIZE))
lh->counter -= LIH_SIZE;
diff --git a/net/dccp/ccids/lib/packet_history.c b/net/dccp/ccids/lib/packet_history.c
index 20af1a69342..6cc108afdc3 100644
--- a/net/dccp/ccids/lib/packet_history.c
+++ b/net/dccp/ccids/lib/packet_history.c
@@ -153,7 +153,7 @@ void tfrc_rx_packet_history_exit(void)
static inline void tfrc_rx_hist_entry_from_skb(struct tfrc_rx_hist_entry *entry,
const struct sk_buff *skb,
- const u32 ndp)
+ const u64 ndp)
{
const struct dccp_hdr *dh = dccp_hdr(skb);
@@ -166,7 +166,7 @@ static inline void tfrc_rx_hist_entry_from_skb(struct tfrc_rx_hist_entry *entry,
void tfrc_rx_hist_add_packet(struct tfrc_rx_hist *h,
const struct sk_buff *skb,
- const u32 ndp)
+ const u64 ndp)
{
struct tfrc_rx_hist_entry *entry = tfrc_rx_hist_last_rcv(h);
@@ -206,31 +206,39 @@ static void tfrc_rx_hist_swap(struct tfrc_rx_hist *h, const u8 a, const u8 b)
*
* In the descriptions, `Si' refers to the sequence number of entry number i,
* whose NDP count is `Ni' (lower case is used for variables).
- * Note: All __after_loss functions expect that a test against duplicates has
- * been performed already: the seqno of the skb must not be less than the
- * seqno of loss_prev; and it must not equal that of any valid hist_entry.
+ * Note: All __xxx_loss functions expect that a test against duplicates has been
+ * performed already: the seqno of the skb must not be less than the seqno
+ * of loss_prev; and it must not equal that of any valid history entry.
*/
+static void __do_track_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u64 n1)
+{
+ u64 s0 = tfrc_rx_hist_loss_prev(h)->tfrchrx_seqno,
+ s1 = DCCP_SKB_CB(skb)->dccpd_seq;
+
+ if (!dccp_loss_free(s0, s1, n1)) { /* gap between S0 and S1 */
+ h->loss_count = 1;
+ tfrc_rx_hist_entry_from_skb(tfrc_rx_hist_entry(h, 1), skb, n1);
+ }
+}
+
static void __one_after_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u32 n2)
{
u64 s0 = tfrc_rx_hist_loss_prev(h)->tfrchrx_seqno,
s1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_seqno,
s2 = DCCP_SKB_CB(skb)->dccpd_seq;
- int n1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_ndp,
- d12 = dccp_delta_seqno(s1, s2), d2;
- if (d12 > 0) { /* S1 < S2 */
+ if (likely(dccp_delta_seqno(s1, s2) > 0)) { /* S1 < S2 */
h->loss_count = 2;
tfrc_rx_hist_entry_from_skb(tfrc_rx_hist_entry(h, 2), skb, n2);
return;
}
/* S0 < S2 < S1 */
- d2 = dccp_delta_seqno(s0, s2);
- if (d2 == 1 || n2 >= d2) { /* S2 is direct successor of S0 */
- int d21 = -d12;
+ if (dccp_loss_free(s0, s2, n2)) {
+ u64 n1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_ndp;
- if (d21 == 1 || n1 >= d21) {
+ if (dccp_loss_free(s2, s1, n1)) {
/* hole is filled: S0, S2, and S1 are consecutive */
h->loss_count = 0;
h->loss_start = tfrc_rx_hist_index(h, 1);
@@ -238,9 +246,9 @@ static void __one_after_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u32 n2
/* gap between S2 and S1: just update loss_prev */
tfrc_rx_hist_entry_from_skb(tfrc_rx_hist_loss_prev(h), skb, n2);
- } else { /* hole between S0 and S2 */
+ } else { /* gap between S0 and S2 */
/*
- * Reorder history to insert S2 between S0 and s1
+ * Reorder history to insert S2 between S0 and S1
*/
tfrc_rx_hist_swap(h, 0, 3);
h->loss_start = tfrc_rx_hist_index(h, 3);
@@ -256,22 +264,18 @@ static int __two_after_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u32 n3)
s1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_seqno,
s2 = tfrc_rx_hist_entry(h, 2)->tfrchrx_seqno,
s3 = DCCP_SKB_CB(skb)->dccpd_seq;
- int n1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_ndp,
- d23 = dccp_delta_seqno(s2, s3), d13, d3, d31;
- if (d23 > 0) { /* S2 < S3 */
+ if (likely(dccp_delta_seqno(s2, s3) > 0)) { /* S2 < S3 */
h->loss_count = 3;
tfrc_rx_hist_entry_from_skb(tfrc_rx_hist_entry(h, 3), skb, n3);
return 1;
}
/* S3 < S2 */
- d13 = dccp_delta_seqno(s1, s3);
- if (d13 > 0) {
+ if (dccp_delta_seqno(s1, s3) > 0) { /* S1 < S3 < S2 */
/*
- * The sequence number order is S1, S3, S2
- * Reorder history to insert entry between S1 and S2
+ * Reorder history to insert S3 between S1 and S2
*/
tfrc_rx_hist_swap(h, 2, 3);
tfrc_rx_hist_entry_from_skb(tfrc_rx_hist_entry(h, 2), skb, n3);
@@ -280,17 +284,15 @@ static int __two_after_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u32 n3)
}
/* S0 < S3 < S1 */
- d31 = -d13;
- d3 = dccp_delta_seqno(s0, s3);
- if (d3 == 1 || n3 >= d3) { /* S3 is a successor of S0 */
+ if (dccp_loss_free(s0, s3, n3)) {
+ u64 n1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_ndp;
- if (d31 == 1 || n1 >= d31) {
+ if (dccp_loss_free(s3, s1, n1)) {
/* hole between S0 and S1 filled by S3 */
- int d2 = dccp_delta_seqno(s1, s2),
- n2 = tfrc_rx_hist_entry(h, 2)->tfrchrx_ndp;
+ u64 n2 = tfrc_rx_hist_entry(h, 2)->tfrchrx_ndp;
- if (d2 == 1 || n2 >= d2) {
+ if (dccp_loss_free(s1, s2, n2)) {
/* entire hole filled by S0, S3, S1, S2 */
h->loss_start = tfrc_rx_hist_index(h, 2);
h->loss_count = 0;
@@ -307,8 +309,8 @@ static int __two_after_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u32 n3)
}
/*
- * The remaining case: S3 is not a successor of S0.
- * Sequence order is S0, S3, S1, S2; reorder to insert between S0 and S1
+ * The remaining case: S0 < S3 < S1 < S2; gap between S0 and S3
+ * Reorder history to insert S3 between S0 and S1.
*/
tfrc_rx_hist_swap(h, 0, 3);
h->loss_start = tfrc_rx_hist_index(h, 3);
@@ -318,33 +320,25 @@ static int __two_after_loss(struct tfrc_rx_hist *h, struct sk_buff *skb, u32 n3)
return 1;
}
-/* return the signed modulo-2^48 sequence number distance from entry e1 to e2 */
-static s64 tfrc_rx_hist_delta_seqno(struct tfrc_rx_hist *h, u8 e1, u8 e2)
-{
- DCCP_BUG_ON(e1 > h->loss_count || e2 > h->loss_count);
-
- return dccp_delta_seqno(tfrc_rx_hist_entry(h, e1)->tfrchrx_seqno,
- tfrc_rx_hist_entry(h, e2)->tfrchrx_seqno);
-}
-
/* recycle RX history records to continue loss detection if necessary */
static void __three_after_loss(struct tfrc_rx_hist *h)
{
/*
- * The distance between S0 and S1 is always greater than 1 and the NDP
- * count of S1 is smaller than this distance. Otherwise there would
- * have been no loss. Hence it is only necessary to see whether there
- * are further missing data packets between S1/S2 and S2/S3.
+ * At this stage we know already that there is a gap between S0 and S1
+ * (since S0 was the highest sequence number received before detecting
+ * the loss). To recycle the loss record, it is thus only necessary to
+ * check for other possible gaps between S1/S2 and between S2/S3.
*/
- int d2 = tfrc_rx_hist_delta_seqno(h, 1, 2),
- d3 = tfrc_rx_hist_delta_seqno(h, 2, 3),
- n2 = tfrc_rx_hist_entry(h, 2)->tfrchrx_ndp,
+ u64 s1 = tfrc_rx_hist_entry(h, 1)->tfrchrx_seqno,
+ s2 = tfrc_rx_hist_entry(h, 2)->tfrchrx_seqno,
+ s3 = tfrc_rx_hist_entry(h, 3)->tfrchrx_seqno;
+ u64 n2 = tfrc_rx_hist_entry(h, 2)->tfrchrx_ndp,
n3 = tfrc_rx_hist_entry(h, 3)->tfrchrx_ndp;
- if (d2 == 1 || n2 >= d2) { /* S2 is successor to S1 */
+ if (dccp_loss_free(s1, s2, n2)) {
- if (d3 == 1 || n3 >= d3) {
- /* S3 is successor of S2: entire hole is filled */
+ if (dccp_loss_free(s2, s3, n3)) {
+ /* no gap between S2 and S3: entire hole is filled */
h->loss_start = tfrc_rx_hist_index(h, 3);
h->loss_count = 0;
} else {
@@ -353,7 +347,7 @@ static void __three_after_loss(struct tfrc_rx_hist *h)
h->loss_count = 1;
}
- } else { /* gap between S1 and S2 */
+ } else { /* gap between S1 and S2 */
h->loss_start = tfrc_rx_hist_index(h, 1);
h->loss_count = 2;
}
@@ -370,15 +364,20 @@ static void __three_after_loss(struct tfrc_rx_hist *h)
* Chooses action according to pending loss, updates LI database when a new
* loss was detected, and does required post-processing. Returns 1 when caller
* should send feedback, 0 otherwise.
+ * Since it also takes care of reordering during loss detection and updates the
+ * records accordingly, the caller should not perform any more RX history
+ * operations when loss_count is greater than 0 after calling this function.
*/
int tfrc_rx_handle_loss(struct tfrc_rx_hist *h,
struct tfrc_loss_hist *lh,
- struct sk_buff *skb, u32 ndp,
+ struct sk_buff *skb, const u64 ndp,
u32 (*calc_first_li)(struct sock *), struct sock *sk)
{
int is_new_loss = 0;
- if (h->loss_count == 1) {
+ if (h->loss_count == 0) {
+ __do_track_loss(h, skb, ndp);
+ } else if (h->loss_count == 1) {
__one_after_loss(h, skb, ndp);
} else if (h->loss_count != 2) {
DCCP_BUG("invalid loss_count %d", h->loss_count);
diff --git a/net/dccp/ccids/lib/packet_history.h b/net/dccp/ccids/lib/packet_history.h
index c7eeda49cb2..461cc91cce8 100644
--- a/net/dccp/ccids/lib/packet_history.h
+++ b/net/dccp/ccids/lib/packet_history.h
@@ -64,7 +64,7 @@ struct tfrc_rx_hist_entry {
u64 tfrchrx_seqno:48,
tfrchrx_ccval:4,
tfrchrx_type:4;
- u32 tfrchrx_ndp; /* In fact it is from 8 to 24 bits */
+ u64 tfrchrx_ndp:48;
ktime_t tfrchrx_tstamp;
};
@@ -118,41 +118,21 @@ static inline struct tfrc_rx_hist_entry *
return h->ring[h->loss_start];
}
-/* initialise loss detection and disable RTT sampling */
-static inline void tfrc_rx_hist_loss_indicated(struct tfrc_rx_hist *h)
-{
- h->loss_count = 1;
-}
-
/* indicate whether previously a packet was detected missing */
-static inline int tfrc_rx_hist_loss_pending(const struct tfrc_rx_hist *h)
-{
- return h->loss_count;
-}
-
-/* any data packets missing between last reception and skb ? */
-static inline int tfrc_rx_hist_new_loss_indicated(struct tfrc_rx_hist *h,
- const struct sk_buff *skb,
- u32 ndp)
+static inline bool tfrc_rx_hist_loss_pending(const struct tfrc_rx_hist *h)
{
- int delta = dccp_delta_seqno(tfrc_rx_hist_last_rcv(h)->tfrchrx_seqno,
- DCCP_SKB_CB(skb)->dccpd_seq);
-
- if (delta > 1 && ndp < delta)
- tfrc_rx_hist_loss_indicated(h);
-
- return tfrc_rx_hist_loss_pending(h);
+ return h->loss_count > 0;
}
extern void tfrc_rx_hist_add_packet(struct tfrc_rx_hist *h,
- const struct sk_buff *skb, const u32 ndp);
+ const struct sk_buff *skb, const u64 ndp);
extern int tfrc_rx_hist_duplicate(struct tfrc_rx_hist *h, struct sk_buff *skb);
struct tfrc_loss_hist;
extern int tfrc_rx_handle_loss(struct tfrc_rx_hist *h,
struct tfrc_loss_hist *lh,
- struct sk_buff *skb, u32 ndp,
+ struct sk_buff *skb, const u64 ndp,
u32 (*first_li)(struct sock *sk),
struct sock *sk);
extern u32 tfrc_rx_hist_sample_rtt(struct tfrc_rx_hist *h,
diff --git a/net/dccp/dccp.h b/net/dccp/dccp.h
index f44d492d3b7..1c2e3ec2eb5 100644
--- a/net/dccp/dccp.h
+++ b/net/dccp/dccp.h
@@ -153,6 +153,21 @@ static inline u64 max48(const u64 seq1, const u64 seq2)
return after48(seq1, seq2) ? seq1 : seq2;
}
+/**
+ * dccp_loss_free - Evaluates condition for data loss from RFC 4340, 7.7.1
+ * @s1: start sequence number
+ * @s2: end sequence number
+ * @ndp: NDP count on packet with sequence number @s2
+ * Returns true if the sequence range s1...s2 has no data loss.
+ */
+static inline bool dccp_loss_free(const u64 s1, const u64 s2, const u64 ndp)
+{
+ s64 delta = dccp_delta_seqno(s1, s2);
+
+ WARN_ON(delta < 0);
+ return (u64)delta <= ndp + 1;
+}
+
enum {
DCCP_MIB_NUM = 0,
DCCP_MIB_ACTIVEOPENS, /* ActiveOpens */
@@ -211,7 +226,7 @@ static inline void dccp_csum_outgoing(struct sk_buff *skb)
extern void dccp_v4_send_check(struct sock *sk, int len, struct sk_buff *skb);
-extern int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb);
+extern int dccp_retransmit_skb(struct sock *sk);
extern void dccp_send_ack(struct sock *sk);
extern void dccp_reqsk_send_ack(struct sk_buff *sk, struct request_sock *rsk);
@@ -262,7 +277,7 @@ extern int dccp_rcv_established(struct sock *sk, struct sk_buff *skb,
const struct dccp_hdr *dh, const unsigned len);
extern int dccp_init_sock(struct sock *sk, const __u8 ctl_sock_initialized);
-extern int dccp_destroy_sock(struct sock *sk);
+extern void dccp_destroy_sock(struct sock *sk);
extern void dccp_close(struct sock *sk, long timeout);
extern struct sk_buff *dccp_make_response(struct sock *sk,
diff --git a/net/dccp/input.c b/net/dccp/input.c
index 08392ed86c2..df2f110df94 100644
--- a/net/dccp/input.c
+++ b/net/dccp/input.c
@@ -413,7 +413,7 @@ static int dccp_rcv_request_sent_state_process(struct sock *sk,
/* Stop the REQUEST timer */
inet_csk_clear_xmit_timer(sk, ICSK_TIME_RETRANS);
- BUG_TRAP(sk->sk_send_head != NULL);
+ WARN_ON(sk->sk_send_head == NULL);
__kfree_skb(sk->sk_send_head);
sk->sk_send_head = NULL;
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index 37d27bcb361..882c5c4de69 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -196,8 +196,8 @@ static inline void dccp_do_pmtu_discovery(struct sock *sk,
static void dccp_v4_err(struct sk_buff *skb, u32 info)
{
const struct iphdr *iph = (struct iphdr *)skb->data;
- const struct dccp_hdr *dh = (struct dccp_hdr *)(skb->data +
- (iph->ihl << 2));
+ const u8 offset = iph->ihl << 2;
+ const struct dccp_hdr *dh = (struct dccp_hdr *)(skb->data + offset);
struct dccp_sock *dp;
struct inet_sock *inet;
const int type = icmp_hdr(skb)->type;
@@ -205,17 +205,19 @@ static void dccp_v4_err(struct sk_buff *skb, u32 info)
struct sock *sk;
__u64 seq;
int err;
+ struct net *net = dev_net(skb->dev);
- if (skb->len < (iph->ihl << 2) + 8) {
- ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
+ if (skb->len < offset + sizeof(*dh) ||
+ skb->len < offset + __dccp_basic_hdr_len(dh)) {
+ ICMP_INC_STATS_BH(net, ICMP_MIB_INERRORS);
return;
}
- sk = inet_lookup(dev_net(skb->dev), &dccp_hashinfo,
+ sk = inet_lookup(net, &dccp_hashinfo,
iph->daddr, dh->dccph_dport,
iph->saddr, dh->dccph_sport, inet_iif(skb));
if (sk == NULL) {
- ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
+ ICMP_INC_STATS_BH(net, ICMP_MIB_INERRORS);
return;
}
@@ -229,7 +231,7 @@ static void dccp_v4_err(struct sk_buff *skb, u32 info)
* servers this needs to be solved differently.
*/
if (sock_owned_by_user(sk))
- NET_INC_STATS_BH(LINUX_MIB_LOCKDROPPEDICMPS);
+ NET_INC_STATS_BH(net, LINUX_MIB_LOCKDROPPEDICMPS);
if (sk->sk_state == DCCP_CLOSED)
goto out;
@@ -237,8 +239,8 @@ static void dccp_v4_err(struct sk_buff *skb, u32 info)
dp = dccp_sk(sk);
seq = dccp_hdr_seq(dh);
if ((1 << sk->sk_state) & ~(DCCPF_REQUESTING | DCCPF_LISTEN) &&
- !between48(seq, dp->dccps_swl, dp->dccps_swh)) {
- NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS);
+ !between48(seq, dp->dccps_awl, dp->dccps_awh)) {
+ NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
goto out;
}
@@ -282,10 +284,10 @@ static void dccp_v4_err(struct sk_buff *skb, u32 info)
* ICMPs are not backlogged, hence we cannot get an established
* socket here.
*/
- BUG_TRAP(!req->sk);
+ WARN_ON(req->sk);
if (seq != dccp_rsk(req)->dreq_iss) {
- NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS);
+ NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
goto out;
}
/*
@@ -408,9 +410,9 @@ struct sock *dccp_v4_request_recv_sock(struct sock *sk, struct sk_buff *skb,
return newsk;
exit_overflow:
- NET_INC_STATS_BH(LINUX_MIB_LISTENOVERFLOWS);
+ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
exit:
- NET_INC_STATS_BH(LINUX_MIB_LISTENDROPS);
+ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENDROPS);
dst_release(dst);
return NULL;
}
@@ -464,7 +466,7 @@ static struct dst_entry* dccp_v4_route_skb(struct net *net, struct sock *sk,
security_skb_classify_flow(skb, &fl);
if (ip_route_output_flow(net, &rt, &fl, sk, 0)) {
- IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
+ IP_INC_STATS_BH(net, IPSTATS_MIB_OUTNOROUTES);
return NULL;
}
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index f7fe2a572d7..5e1ee0da2c4 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -89,12 +89,20 @@ static void dccp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
{
struct ipv6hdr *hdr = (struct ipv6hdr *)skb->data;
const struct dccp_hdr *dh = (struct dccp_hdr *)(skb->data + offset);
+ struct dccp_sock *dp;
struct ipv6_pinfo *np;
struct sock *sk;
int err;
__u64 seq;
+ struct net *net = dev_net(skb->dev);
- sk = inet6_lookup(dev_net(skb->dev), &dccp_hashinfo,
+ if (skb->len < offset + sizeof(*dh) ||
+ skb->len < offset + __dccp_basic_hdr_len(dh)) {
+ ICMP6_INC_STATS_BH(__in6_dev_get(skb->dev), ICMP6_MIB_INERRORS);
+ return;
+ }
+
+ sk = inet6_lookup(net, &dccp_hashinfo,
&hdr->daddr, dh->dccph_dport,
&hdr->saddr, dh->dccph_sport, inet6_iif(skb));
@@ -110,11 +118,19 @@ static void dccp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
bh_lock_sock(sk);
if (sock_owned_by_user(sk))
- NET_INC_STATS_BH(LINUX_MIB_LOCKDROPPEDICMPS);
+ NET_INC_STATS_BH(net, LINUX_MIB_LOCKDROPPEDICMPS);
if (sk->sk_state == DCCP_CLOSED)
goto out;
+ dp = dccp_sk(sk);
+ seq = dccp_hdr_seq(dh);
+ if ((1 << sk->sk_state) & ~(DCCPF_REQUESTING | DCCPF_LISTEN) &&
+ !between48(seq, dp->dccps_awl, dp->dccps_awh)) {
+ NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
+ goto out;
+ }
+
np = inet6_sk(sk);
if (type == ICMPV6_PKT_TOOBIG) {
@@ -167,7 +183,6 @@ static void dccp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
icmpv6_err_convert(type, code, &err);
- seq = dccp_hdr_seq(dh);
/* Might be for an request_sock */
switch (sk->sk_state) {
struct request_sock *req, **prev;
@@ -185,10 +200,10 @@ static void dccp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
* ICMPs are not backlogged, hence we cannot get an established
* socket here.
*/
- BUG_TRAP(req->sk == NULL);
+ WARN_ON(req->sk != NULL);
if (seq != dccp_rsk(req)->dreq_iss) {
- NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS);
+ NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
goto out;
}
@@ -629,9 +644,9 @@ static struct sock *dccp_v6_request_recv_sock(struct sock *sk,
return newsk;
out_overflow:
- NET_INC_STATS_BH(LINUX_MIB_LISTENOVERFLOWS);
+ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
out:
- NET_INC_STATS_BH(LINUX_MIB_LISTENDROPS);
+ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENDROPS);
if (opt != NULL && opt != np->opt)
sock_kfree_s(sk, opt, opt->tot_len);
dst_release(dst);
@@ -1091,10 +1106,10 @@ static int dccp_v6_init_sock(struct sock *sk)
return err;
}
-static int dccp_v6_destroy_sock(struct sock *sk)
+static void dccp_v6_destroy_sock(struct sock *sk)
{
dccp_destroy_sock(sk);
- return inet6_destroy_sock(sk);
+ inet6_destroy_sock(sk);
}
static struct timewait_sock_ops dccp6_timewait_sock_ops = {
diff --git a/net/dccp/options.c b/net/dccp/options.c
index 43bc24e761d..dc7c158a2f4 100644
--- a/net/dccp/options.c
+++ b/net/dccp/options.c
@@ -124,12 +124,12 @@ int dccp_parse_options(struct sock *sk, struct dccp_request_sock *dreq,
mandatory = 1;
break;
case DCCPO_NDP_COUNT:
- if (len > 3)
+ if (len > 6)
goto out_invalid_option;
opt_recv->dccpor_ndp = dccp_decode_value_var(value, len);
- dccp_pr_debug("%s rx opt: NDP count=%d\n", dccp_role(sk),
- opt_recv->dccpor_ndp);
+ dccp_pr_debug("%s opt: NDP count=%llu\n", dccp_role(sk),
+ (unsigned long long)opt_recv->dccpor_ndp);
break;
case DCCPO_CHANGE_L:
/* fall through */
@@ -307,9 +307,11 @@ static void dccp_encode_value_var(const u32 value, unsigned char *to,
*to++ = (value & 0xFF);
}
-static inline int dccp_ndp_len(const int ndp)
+static inline u8 dccp_ndp_len(const u64 ndp)
{
- return likely(ndp <= 0xFF) ? 1 : ndp <= 0xFFFF ? 2 : 3;
+ if (likely(ndp <= 0xFF))
+ return 1;
+ return likely(ndp <= USHORT_MAX) ? 2 : (ndp <= UINT_MAX ? 4 : 6);
}
int dccp_insert_option(struct sock *sk, struct sk_buff *skb,
@@ -336,7 +338,7 @@ EXPORT_SYMBOL_GPL(dccp_insert_option);
static int dccp_insert_option_ndp(struct sock *sk, struct sk_buff *skb)
{
struct dccp_sock *dp = dccp_sk(sk);
- int ndp = dp->dccps_ndp_count;
+ u64 ndp = dp->dccps_ndp_count;
if (dccp_non_data_packet(skb))
++dp->dccps_ndp_count;
diff --git a/net/dccp/output.c b/net/dccp/output.c
index fe20068c5d8..d06945c7d3d 100644
--- a/net/dccp/output.c
+++ b/net/dccp/output.c
@@ -53,8 +53,11 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb)
dccp_packet_hdr_len(dcb->dccpd_type);
int err, set_ack = 1;
u64 ackno = dp->dccps_gsr;
-
- dccp_inc_seqno(&dp->dccps_gss);
+ /*
+ * Increment GSS here already in case the option code needs it.
+ * Update GSS for real only if option processing below succeeds.
+ */
+ dcb->dccpd_seq = ADD48(dp->dccps_gss, 1);
switch (dcb->dccpd_type) {
case DCCP_PKT_DATA:
@@ -66,6 +69,9 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb)
case DCCP_PKT_REQUEST:
set_ack = 0;
+ /* Use ISS on the first (non-retransmitted) Request. */
+ if (icsk->icsk_retransmits == 0)
+ dcb->dccpd_seq = dp->dccps_iss;
/* fall through */
case DCCP_PKT_SYNC:
@@ -84,8 +90,6 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb)
break;
}
- dcb->dccpd_seq = dp->dccps_gss;
-
if (dccp_insert_options(sk, skb)) {
kfree_skb(skb);
return -EPROTO;
@@ -103,7 +107,7 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb)
/* XXX For now we're using only 48 bits sequence numbers */
dh->dccph_x = 1;
- dp->dccps_awh = dp->dccps_gss;
+ dccp_update_gss(sk, dcb->dccpd_seq);
dccp_hdr_set_seq(dh, dp->dccps_gss);
if (set_ack)
dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), ackno);
@@ -112,6 +116,11 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb)
case DCCP_PKT_REQUEST:
dccp_hdr_request(skb)->dccph_req_service =
dp->dccps_service;
+ /*
+ * Limit Ack window to ISS <= P.ackno <= GSS, so that
+ * only Responses to Requests we sent are considered.
+ */
+ dp->dccps_awl = dp->dccps_iss;
break;
case DCCP_PKT_RESET:
dccp_hdr_reset(skb)->dccph_reset_code =
@@ -284,14 +293,26 @@ void dccp_write_xmit(struct sock *sk, int block)
}
}
-int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
+/**
+ * dccp_retransmit_skb - Retransmit Request, Close, or CloseReq packets
+ * There are only four retransmittable packet types in DCCP:
+ * - Request in client-REQUEST state (sec. 8.1.1),
+ * - CloseReq in server-CLOSEREQ state (sec. 8.3),
+ * - Close in node-CLOSING state (sec. 8.3),
+ * - Acks in client-PARTOPEN state (sec. 8.1.5, handled by dccp_delack_timer()).
+ * This function expects sk->sk_send_head to contain the original skb.
+ */
+int dccp_retransmit_skb(struct sock *sk)
{
+ WARN_ON(sk->sk_send_head == NULL);
+
if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk) != 0)
return -EHOSTUNREACH; /* Routing failure or similar. */
- return dccp_transmit_skb(sk, (skb_cloned(skb) ?
- pskb_copy(skb, GFP_ATOMIC):
- skb_clone(skb, GFP_ATOMIC)));
+ /* this count is used to distinguish original and retransmitted skb */
+ inet_csk(sk)->icsk_retransmits++;
+
+ return dccp_transmit_skb(sk, skb_clone(sk->sk_send_head, GFP_ATOMIC));
}
struct sk_buff *dccp_make_response(struct sock *sk, struct dst_entry *dst,
@@ -437,19 +458,7 @@ static inline void dccp_connect_init(struct sock *sk)
dccp_sync_mss(sk, dst_mtu(dst));
- /*
- * 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) */
+ /* Initialise GAR as per 8.5; AWL/AWH are set in dccp_transmit_skb() */
dp->dccps_gar = dp->dccps_iss;
icsk->icsk_retransmits = 0;
diff --git a/net/dccp/proto.c b/net/dccp/proto.c
index 9dfe2470962..b622d974485 100644
--- a/net/dccp/proto.c
+++ b/net/dccp/proto.c
@@ -237,7 +237,7 @@ int dccp_init_sock(struct sock *sk, const __u8 ctl_sock_initialized)
EXPORT_SYMBOL_GPL(dccp_init_sock);
-int dccp_destroy_sock(struct sock *sk)
+void dccp_destroy_sock(struct sock *sk)
{
struct dccp_sock *dp = dccp_sk(sk);
struct dccp_minisock *dmsk = dccp_msk(sk);
@@ -268,8 +268,6 @@ int dccp_destroy_sock(struct sock *sk)
/* clean up feature negotiation state */
dccp_feat_clean(dmsk);
-
- return 0;
}
EXPORT_SYMBOL_GPL(dccp_destroy_sock);
@@ -329,7 +327,7 @@ int dccp_disconnect(struct sock *sk, int flags)
inet_csk_delack_init(sk);
__sk_dst_reset(sk);
- BUG_TRAP(!inet->num || icsk->icsk_bind_hash);
+ WARN_ON(inet->num && !icsk->icsk_bind_hash);
sk->sk_error_report(sk);
return err;
@@ -983,7 +981,7 @@ adjudge_to_death:
*/
local_bh_disable();
bh_lock_sock(sk);
- BUG_TRAP(!sock_owned_by_user(sk));
+ WARN_ON(sock_owned_by_user(sk));
/* Have we already been destroyed by a softirq or backlog? */
if (state != DCCP_CLOSED && sk->sk_state == DCCP_CLOSED)
diff --git a/net/dccp/timer.c b/net/dccp/timer.c
index 8703a792b56..54b3c7e9e01 100644
--- a/net/dccp/timer.c
+++ b/net/dccp/timer.c
@@ -99,21 +99,11 @@ static void dccp_retransmit_timer(struct sock *sk)
}
/*
- * sk->sk_send_head has to have one skb with
- * DCCP_SKB_CB(skb)->dccpd_type set to one of the retransmittable DCCP
- * packet types. The only packets eligible for retransmission are:
- * -- Requests in client-REQUEST state (sec. 8.1.1)
- * -- Acks in client-PARTOPEN state (sec. 8.1.5)
- * -- CloseReq in server-CLOSEREQ state (sec. 8.3)
- * -- Close in node-CLOSING state (sec. 8.3) */
- BUG_TRAP(sk->sk_send_head != NULL);
-
- /*
* More than than 4MSL (8 minutes) has passed, a RESET(aborted) was
* sent, no need to retransmit, this sock is dead.
*/
if (dccp_write_timeout(sk))
- goto out;
+ return;
/*
* We want to know the number of packets retransmitted, not the
@@ -122,30 +112,28 @@ static void dccp_retransmit_timer(struct sock *sk)
if (icsk->icsk_retransmits == 0)
DCCP_INC_STATS_BH(DCCP_MIB_TIMEOUTS);
- if (dccp_retransmit_skb(sk, sk->sk_send_head) < 0) {
+ if (dccp_retransmit_skb(sk) != 0) {
/*
* Retransmission failed because of local congestion,
* do not backoff.
*/
- if (icsk->icsk_retransmits == 0)
+ if (--icsk->icsk_retransmits == 0)
icsk->icsk_retransmits = 1;
inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
min(icsk->icsk_rto,
TCP_RESOURCE_PROBE_INTERVAL),
DCCP_RTO_MAX);
- goto out;
+ return;
}
backoff:
icsk->icsk_backoff++;
- icsk->icsk_retransmits++;
icsk->icsk_rto = min(icsk->icsk_rto << 1, DCCP_RTO_MAX);
inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto,
DCCP_RTO_MAX);
if (icsk->icsk_retransmits > sysctl_dccp_retries1)
__sk_dst_reset(sk);
-out:;
}
static void dccp_write_timer(unsigned long data)
@@ -224,7 +212,7 @@ static void dccp_delack_timer(unsigned long data)
if (sock_owned_by_user(sk)) {
/* Try again later. */
icsk->icsk_ack.blocked = 1;
- NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKLOCKED);
+ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_DELAYEDACKLOCKED);
sk_reset_timer(sk, &icsk->icsk_delack_timer,
jiffies + TCP_DELACK_MIN);
goto out;
@@ -254,7 +242,7 @@ static void dccp_delack_timer(unsigned long data)
icsk->icsk_ack.ato = TCP_ATO_MIN;
}
dccp_send_ack(sk);
- NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKS);
+ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_DELAYEDACKS);
}
out:
bh_unlock_sock(sk);
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2024-09-29 07:32:34 +0000