diff options
Diffstat (limited to 'net/dccp')
-rw-r--r-- | net/dccp/Kconfig | 1 | ||||
-rw-r--r-- | net/dccp/ackvec.c | 163 | ||||
-rw-r--r-- | net/dccp/ackvec.h | 62 | ||||
-rw-r--r-- | net/dccp/ccid.c | 8 | ||||
-rw-r--r-- | net/dccp/ccid.h | 37 | ||||
-rw-r--r-- | net/dccp/ccids/Kconfig | 30 | ||||
-rw-r--r-- | net/dccp/ccids/ccid2.c | 228 | ||||
-rw-r--r-- | net/dccp/ccids/ccid2.h | 21 | ||||
-rw-r--r-- | net/dccp/ccids/ccid3.c | 710 | ||||
-rw-r--r-- | net/dccp/ccids/ccid3.h | 41 | ||||
-rw-r--r-- | net/dccp/ccids/lib/Makefile | 2 | ||||
-rw-r--r-- | net/dccp/ccids/lib/loss_interval.c | 352 | ||||
-rw-r--r-- | net/dccp/ccids/lib/loss_interval.h | 64 | ||||
-rw-r--r-- | net/dccp/ccids/lib/packet_history.c | 599 | ||||
-rw-r--r-- | net/dccp/ccids/lib/packet_history.h | 220 | ||||
-rw-r--r-- | net/dccp/ccids/lib/tfrc.c | 63 | ||||
-rw-r--r-- | net/dccp/ccids/lib/tfrc.h | 29 | ||||
-rw-r--r-- | net/dccp/dccp.h | 35 | ||||
-rw-r--r-- | net/dccp/feat.c | 29 | ||||
-rw-r--r-- | net/dccp/feat.h | 26 | ||||
-rw-r--r-- | net/dccp/input.c | 155 | ||||
-rw-r--r-- | net/dccp/ipv4.c | 12 | ||||
-rw-r--r-- | net/dccp/ipv6.c | 10 | ||||
-rw-r--r-- | net/dccp/minisocks.c | 33 | ||||
-rw-r--r-- | net/dccp/options.c | 139 | ||||
-rw-r--r-- | net/dccp/output.c | 55 | ||||
-rw-r--r-- | net/dccp/proto.c | 194 | ||||
-rw-r--r-- | net/dccp/sysctl.c | 36 | ||||
-rw-r--r-- | net/dccp/timer.c | 5 |
29 files changed, 1777 insertions, 1582 deletions
diff --git a/net/dccp/Kconfig b/net/dccp/Kconfig index 0549e4719b1..7aa2a7acc7e 100644 --- a/net/dccp/Kconfig +++ b/net/dccp/Kconfig @@ -1,6 +1,7 @@ menuconfig IP_DCCP tristate "The DCCP Protocol (EXPERIMENTAL)" depends on INET && EXPERIMENTAL + select IP_DCCP_CCID2 ---help--- Datagram Congestion Control Protocol (RFC 4340) diff --git a/net/dccp/ackvec.c b/net/dccp/ackvec.c index 83378f379f7..6de4bd195d2 100644 --- a/net/dccp/ackvec.c +++ b/net/dccp/ackvec.c @@ -30,7 +30,7 @@ static struct dccp_ackvec_record *dccp_ackvec_record_new(void) kmem_cache_alloc(dccp_ackvec_record_slab, GFP_ATOMIC); if (avr != NULL) - INIT_LIST_HEAD(&avr->dccpavr_node); + INIT_LIST_HEAD(&avr->avr_node); return avr; } @@ -40,7 +40,7 @@ static void dccp_ackvec_record_delete(struct dccp_ackvec_record *avr) if (unlikely(avr == NULL)) return; /* Check if deleting a linked record */ - WARN_ON(!list_empty(&avr->dccpavr_node)); + WARN_ON(!list_empty(&avr->avr_node)); kmem_cache_free(dccp_ackvec_record_slab, avr); } @@ -52,16 +52,15 @@ static void dccp_ackvec_insert_avr(struct dccp_ackvec *av, * just add the AVR at the head of the list. * -sorbo. */ - if (!list_empty(&av->dccpav_records)) { + if (!list_empty(&av->av_records)) { const struct dccp_ackvec_record *head = - list_entry(av->dccpav_records.next, + list_entry(av->av_records.next, struct dccp_ackvec_record, - dccpavr_node); - BUG_ON(before48(avr->dccpavr_ack_seqno, - head->dccpavr_ack_seqno)); + avr_node); + BUG_ON(before48(avr->avr_ack_seqno, head->avr_ack_seqno)); } - list_add(&avr->dccpavr_node, &av->dccpav_records); + list_add(&avr->avr_node, &av->av_records); } int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) @@ -69,9 +68,8 @@ int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) struct dccp_sock *dp = dccp_sk(sk); struct dccp_ackvec *av = dp->dccps_hc_rx_ackvec; /* Figure out how many options do we need to represent the ackvec */ - const u16 nr_opts = DIV_ROUND_UP(av->dccpav_vec_len, - DCCP_MAX_ACKVEC_OPT_LEN); - u16 len = av->dccpav_vec_len + 2 * nr_opts, i; + const u16 nr_opts = DIV_ROUND_UP(av->av_vec_len, DCCP_MAX_ACKVEC_OPT_LEN); + u16 len = av->av_vec_len + 2 * nr_opts, i; u32 elapsed_time; const unsigned char *tail, *from; unsigned char *to; @@ -81,7 +79,7 @@ int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) return -1; - delta = ktime_us_delta(ktime_get_real(), av->dccpav_time); + delta = ktime_us_delta(ktime_get_real(), av->av_time); elapsed_time = delta / 10; if (elapsed_time != 0 && @@ -95,9 +93,9 @@ int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) DCCP_SKB_CB(skb)->dccpd_opt_len += len; to = skb_push(skb, len); - len = av->dccpav_vec_len; - from = av->dccpav_buf + av->dccpav_buf_head; - tail = av->dccpav_buf + DCCP_MAX_ACKVEC_LEN; + len = av->av_vec_len; + from = av->av_buf + av->av_buf_head; + tail = av->av_buf + DCCP_MAX_ACKVEC_LEN; for (i = 0; i < nr_opts; ++i) { int copylen = len; @@ -116,7 +114,7 @@ int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) to += tailsize; len -= tailsize; copylen -= tailsize; - from = av->dccpav_buf; + from = av->av_buf; } memcpy(to, from, copylen); @@ -134,19 +132,19 @@ int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) * buf_head; ack_ackno will equal buf_ackno; and ack_nonce will * equal buf_nonce. */ - avr->dccpavr_ack_seqno = DCCP_SKB_CB(skb)->dccpd_seq; - avr->dccpavr_ack_ptr = av->dccpav_buf_head; - avr->dccpavr_ack_ackno = av->dccpav_buf_ackno; - avr->dccpavr_ack_nonce = av->dccpav_buf_nonce; - avr->dccpavr_sent_len = av->dccpav_vec_len; + avr->avr_ack_seqno = DCCP_SKB_CB(skb)->dccpd_seq; + avr->avr_ack_ptr = av->av_buf_head; + avr->avr_ack_ackno = av->av_buf_ackno; + avr->avr_ack_nonce = av->av_buf_nonce; + avr->avr_sent_len = av->av_vec_len; dccp_ackvec_insert_avr(av, avr); dccp_pr_debug("%s ACK Vector 0, len=%d, ack_seqno=%llu, " "ack_ackno=%llu\n", - dccp_role(sk), avr->dccpavr_sent_len, - (unsigned long long)avr->dccpavr_ack_seqno, - (unsigned long long)avr->dccpavr_ack_ackno); + dccp_role(sk), avr->avr_sent_len, + (unsigned long long)avr->avr_ack_seqno, + (unsigned long long)avr->avr_ack_ackno); return 0; } @@ -155,12 +153,12 @@ struct dccp_ackvec *dccp_ackvec_alloc(const gfp_t priority) struct dccp_ackvec *av = kmem_cache_alloc(dccp_ackvec_slab, priority); if (av != NULL) { - av->dccpav_buf_head = DCCP_MAX_ACKVEC_LEN - 1; - av->dccpav_buf_ackno = UINT48_MAX + 1; - av->dccpav_buf_nonce = av->dccpav_buf_nonce = 0; - av->dccpav_time = ktime_set(0, 0); - av->dccpav_vec_len = 0; - INIT_LIST_HEAD(&av->dccpav_records); + av->av_buf_head = DCCP_MAX_ACKVEC_LEN - 1; + av->av_buf_ackno = UINT48_MAX + 1; + av->av_buf_nonce = 0; + av->av_time = ktime_set(0, 0); + av->av_vec_len = 0; + INIT_LIST_HEAD(&av->av_records); } return av; @@ -171,12 +169,11 @@ void dccp_ackvec_free(struct dccp_ackvec *av) if (unlikely(av == NULL)) return; - if (!list_empty(&av->dccpav_records)) { + if (!list_empty(&av->av_records)) { struct dccp_ackvec_record *avr, *next; - list_for_each_entry_safe(avr, next, &av->dccpav_records, - dccpavr_node) { - list_del_init(&avr->dccpavr_node); + list_for_each_entry_safe(avr, next, &av->av_records, avr_node) { + list_del_init(&avr->avr_node); dccp_ackvec_record_delete(avr); } } @@ -187,13 +184,13 @@ void dccp_ackvec_free(struct dccp_ackvec *av) static inline u8 dccp_ackvec_state(const struct dccp_ackvec *av, const u32 index) { - return av->dccpav_buf[index] & DCCP_ACKVEC_STATE_MASK; + return av->av_buf[index] & DCCP_ACKVEC_STATE_MASK; } static inline u8 dccp_ackvec_len(const struct dccp_ackvec *av, const u32 index) { - return av->dccpav_buf[index] & DCCP_ACKVEC_LEN_MASK; + return av->av_buf[index] & DCCP_ACKVEC_LEN_MASK; } /* @@ -208,29 +205,29 @@ static inline int dccp_ackvec_set_buf_head_state(struct dccp_ackvec *av, unsigned int gap; long new_head; - if (av->dccpav_vec_len + packets > DCCP_MAX_ACKVEC_LEN) + if (av->av_vec_len + packets > DCCP_MAX_ACKVEC_LEN) return -ENOBUFS; gap = packets - 1; - new_head = av->dccpav_buf_head - packets; + new_head = av->av_buf_head - packets; if (new_head < 0) { if (gap > 0) { - memset(av->dccpav_buf, DCCP_ACKVEC_STATE_NOT_RECEIVED, + memset(av->av_buf, DCCP_ACKVEC_STATE_NOT_RECEIVED, gap + new_head + 1); gap = -new_head; } new_head += DCCP_MAX_ACKVEC_LEN; } - av->dccpav_buf_head = new_head; + av->av_buf_head = new_head; if (gap > 0) - memset(av->dccpav_buf + av->dccpav_buf_head + 1, + memset(av->av_buf + av->av_buf_head + 1, DCCP_ACKVEC_STATE_NOT_RECEIVED, gap); - av->dccpav_buf[av->dccpav_buf_head] = state; - av->dccpav_vec_len += packets; + av->av_buf[av->av_buf_head] = state; + av->av_vec_len += packets; return 0; } @@ -243,7 +240,7 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, /* * Check at the right places if the buffer is full, if it is, tell the * caller to start dropping packets till the HC-Sender acks our ACK - * vectors, when we will free up space in dccpav_buf. + * vectors, when we will free up space in av_buf. * * We may well decide to do buffer compression, etc, but for now lets * just drop. @@ -263,22 +260,20 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, */ /* See if this is the first ackno being inserted */ - if (av->dccpav_vec_len == 0) { - av->dccpav_buf[av->dccpav_buf_head] = state; - av->dccpav_vec_len = 1; - } else if (after48(ackno, av->dccpav_buf_ackno)) { - const u64 delta = dccp_delta_seqno(av->dccpav_buf_ackno, - ackno); + if (av->av_vec_len == 0) { + av->av_buf[av->av_buf_head] = state; + av->av_vec_len = 1; + } else if (after48(ackno, av->av_buf_ackno)) { + const u64 delta = dccp_delta_seqno(av->av_buf_ackno, ackno); /* * Look if the state of this packet is the same as the * previous ackno and if so if we can bump the head len. */ if (delta == 1 && - dccp_ackvec_state(av, av->dccpav_buf_head) == state && - (dccp_ackvec_len(av, av->dccpav_buf_head) < - DCCP_ACKVEC_LEN_MASK)) - av->dccpav_buf[av->dccpav_buf_head]++; + dccp_ackvec_state(av, av->av_buf_head) == state && + dccp_ackvec_len(av, av->av_buf_head) < DCCP_ACKVEC_LEN_MASK) + av->av_buf[av->av_buf_head]++; else if (dccp_ackvec_set_buf_head_state(av, delta, state)) return -ENOBUFS; } else { @@ -290,14 +285,14 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, * the byte corresponding to S. (Indexing structures * could reduce the complexity of this scan.) */ - u64 delta = dccp_delta_seqno(ackno, av->dccpav_buf_ackno); - u32 index = av->dccpav_buf_head; + u64 delta = dccp_delta_seqno(ackno, av->av_buf_ackno); + u32 index = av->av_buf_head; while (1) { const u8 len = dccp_ackvec_len(av, index); const u8 state = dccp_ackvec_state(av, index); /* - * valid packets not yet in dccpav_buf have a reserved + * valid packets not yet in av_buf have a reserved * entry, with a len equal to 0. */ if (state == DCCP_ACKVEC_STATE_NOT_RECEIVED && @@ -305,7 +300,7 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, reserved seat! */ dccp_pr_debug("Found %llu reserved seat!\n", (unsigned long long)ackno); - av->dccpav_buf[index] = state; + av->av_buf[index] = state; goto out; } /* len == 0 means one packet */ @@ -318,8 +313,8 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, } } - av->dccpav_buf_ackno = ackno; - av->dccpav_time = ktime_get_real(); + av->av_buf_ackno = ackno; + av->av_time = ktime_get_real(); out: return 0; @@ -349,9 +344,9 @@ void dccp_ackvector_print(const u64 ackno, const unsigned char *vector, int len) void dccp_ackvec_print(const struct dccp_ackvec *av) { - dccp_ackvector_print(av->dccpav_buf_ackno, - av->dccpav_buf + av->dccpav_buf_head, - av->dccpav_vec_len); + dccp_ackvector_print(av->av_buf_ackno, + av->av_buf + av->av_buf_head, + av->av_vec_len); } #endif @@ -361,17 +356,15 @@ static void dccp_ackvec_throw_record(struct dccp_ackvec *av, struct dccp_ackvec_record *next; /* sort out vector length */ - if (av->dccpav_buf_head <= avr->dccpavr_ack_ptr) - av->dccpav_vec_len = avr->dccpavr_ack_ptr - av->dccpav_buf_head; + if (av->av_buf_head <= avr->avr_ack_ptr) + av->av_vec_len = avr->avr_ack_ptr - av->av_buf_head; else - av->dccpav_vec_len = DCCP_MAX_ACKVEC_LEN - 1 - - av->dccpav_buf_head - + avr->dccpavr_ack_ptr; + av->av_vec_len = DCCP_MAX_ACKVEC_LEN - 1 - + av->av_buf_head + avr->avr_ack_ptr; /* free records */ - list_for_each_entry_safe_from(avr, next, &av->dccpav_records, - dccpavr_node) { - list_del_init(&avr->dccpavr_node); + list_for_each_entry_safe_from(avr, next, &av->av_records, avr_node) { + list_del_init(&avr->avr_node); dccp_ackvec_record_delete(avr); } } @@ -386,16 +379,16 @@ void dccp_ackvec_check_rcv_ackno(struct dccp_ackvec *av, struct sock *sk, * windows. We will be receiving ACKs for stuff we sent a while back * -sorbo. */ - list_for_each_entry_reverse(avr, &av->dccpav_records, dccpavr_node) { - if (ackno == avr->dccpavr_ack_seqno) { + list_for_each_entry_reverse(avr, &av->av_records, avr_node) { + if (ackno == avr->avr_ack_seqno) { dccp_pr_debug("%s ACK packet 0, len=%d, ack_seqno=%llu, " "ack_ackno=%llu, ACKED!\n", dccp_role(sk), 1, - (unsigned long long)avr->dccpavr_ack_seqno, - (unsigned long long)avr->dccpavr_ack_ackno); + (unsigned long long)avr->avr_ack_seqno, + (unsigned long long)avr->avr_ack_ackno); dccp_ackvec_throw_record(av, avr); break; - } else if (avr->dccpavr_ack_seqno > ackno) + } else if (avr->avr_ack_seqno > ackno) break; /* old news */ } } @@ -409,7 +402,7 @@ static void dccp_ackvec_check_rcv_ackvector(struct dccp_ackvec *av, struct dccp_ackvec_record *avr; /* Check if we actually sent an ACK vector */ - if (list_empty(&av->dccpav_records)) + if (list_empty(&av->av_records)) return; i = len; @@ -418,8 +411,7 @@ static void dccp_ackvec_check_rcv_ackvector(struct dccp_ackvec *av, * I think it might be more efficient to work backwards. See comment on * rcv_ackno. -sorbo. */ - avr = list_entry(av->dccpav_records.next, struct dccp_ackvec_record, - dccpavr_node); + avr = list_entry(av->av_records.next, struct dccp_ackvec_record, avr_node); while (i--) { const u8 rl = *vector & DCCP_ACKVEC_LEN_MASK; u64 ackno_end_rl; @@ -430,15 +422,14 @@ static void dccp_ackvec_check_rcv_ackvector(struct dccp_ackvec *av, * If our AVR sequence number is greater than the ack, go * forward in the AVR list until it is not so. */ - list_for_each_entry_from(avr, &av->dccpav_records, - dccpavr_node) { - if (!after48(avr->dccpavr_ack_seqno, *ackno)) + list_for_each_entry_from(avr, &av->av_records, avr_node) { + if (!after48(avr->avr_ack_seqno, *ackno)) goto found; } - /* End of the dccpav_records list, not found, exit */ + /* End of the av_records list, not found, exit */ break; found: - if (between48(avr->dccpavr_ack_seqno, ackno_end_rl, *ackno)) { + if (between48(avr->avr_ack_seqno, ackno_end_rl, *ackno)) { const u8 state = *vector & DCCP_ACKVEC_STATE_MASK; if (state != DCCP_ACKVEC_STATE_NOT_RECEIVED) { dccp_pr_debug("%s ACK vector 0, len=%d, " @@ -446,9 +437,9 @@ found: "ACKED!\n", dccp_role(sk), len, (unsigned long long) - avr->dccpavr_ack_seqno, + avr->avr_ack_seqno, (unsigned long long) - avr->dccpavr_ack_ackno); + avr->avr_ack_ackno); dccp_ackvec_throw_record(av, avr); break; } diff --git a/net/dccp/ackvec.h b/net/dccp/ackvec.h index 9671ecd17e0..bcb64fb4ace 100644 --- a/net/dccp/ackvec.h +++ b/net/dccp/ackvec.h @@ -32,54 +32,54 @@ * * This data structure is the one defined in RFC 4340, Appendix A. * - * @dccpav_buf_head - circular buffer head - * @dccpav_buf_tail - circular buffer tail - * @dccpav_buf_ackno - ack # of the most recent packet acknowledgeable in the - * buffer (i.e. %dccpav_buf_head) - * @dccpav_buf_nonce - the one-bit sum of the ECN Nonces on all packets acked + * @av_buf_head - circular buffer head + * @av_buf_tail - circular buffer tail + * @av_buf_ackno - ack # of the most recent packet acknowledgeable in the + * buffer (i.e. %av_buf_head) + * @av_buf_nonce - the one-bit sum of the ECN Nonces on all packets acked * by the buffer with State 0 * * Additionally, the HC-Receiver must keep some information about the * Ack Vectors it has recently sent. For each packet sent carrying an * Ack Vector, it remembers four variables: * - * @dccpav_records - list of dccp_ackvec_record - * @dccpav_ack_nonce - the one-bit sum of the ECN Nonces for all State 0. + * @av_records - list of dccp_ackvec_record + * @av_ack_nonce - the one-bit sum of the ECN Nonces for all State 0. * - * @dccpav_time - the time in usecs - * @dccpav_buf - circular buffer of acknowledgeable packets + * @av_time - the time in usecs + * @av_buf - circular buffer of acknowledgeable packets */ struct dccp_ackvec { - u64 dccpav_buf_ackno; - struct list_head dccpav_records; - ktime_t dccpav_time; - u16 dccpav_buf_head; - u16 dccpav_vec_len; - u8 dccpav_buf_nonce; - u8 dccpav_ack_nonce; - u8 dccpav_buf[DCCP_MAX_ACKVEC_LEN]; + u64 av_buf_ackno; + struct list_head av_records; + ktime_t av_time; + u16 av_buf_head; + u16 av_vec_len; + u8 av_buf_nonce; + u8 av_ack_nonce; + u8 av_buf[DCCP_MAX_ACKVEC_LEN]; }; /** struct dccp_ackvec_record - ack vector record * * ACK vector record as defined in Appendix A of spec. * - * The list is sorted by dccpavr_ack_seqno + * The list is sorted by avr_ack_seqno * - * @dccpavr_node - node in dccpav_records - * @dccpavr_ack_seqno - sequence number of the packet this record was sent on - * @dccpavr_ack_ackno - sequence number being acknowledged - * @dccpavr_ack_ptr - pointer into dccpav_buf where this record starts - * @dccpavr_ack_nonce - dccpav_ack_nonce at the time this record was sent - * @dccpavr_sent_len - length of the record in dccpav_buf + * @avr_node - node in av_records + * @avr_ack_seqno - sequence number of the packet this record was sent on + * @avr_ack_ackno - sequence number being acknowledged + * @avr_ack_ptr - pointer into av_buf where this record starts + * @avr_ack_nonce - av_ack_nonce at the time this record was sent + * @avr_sent_len - lenght of the record in av_buf */ struct dccp_ackvec_record { - struct list_head dccpavr_node; - u64 dccpavr_ack_seqno; - u64 dccpavr_ack_ackno; - u16 dccpavr_ack_ptr; - u16 dccpavr_sent_len; - u8 dccpavr_ack_nonce; + struct list_head avr_node; + u64 avr_ack_seqno; + u64 avr_ack_ackno; + u16 avr_ack_ptr; + u16 avr_sent_len; + u8 avr_ack_nonce; }; struct sock; @@ -105,7 +105,7 @@ extern int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb); static inline int dccp_ackvec_pending(const struct dccp_ackvec *av) { - return av->dccpav_vec_len; + return av->av_vec_len; } #else /* CONFIG_IP_DCCP_ACKVEC */ static inline int dccp_ackvec_init(void) diff --git a/net/dccp/ccid.c b/net/dccp/ccid.c index c45088b5e6f..4809753d12a 100644 --- a/net/dccp/ccid.c +++ b/net/dccp/ccid.c @@ -92,15 +92,15 @@ int ccid_register(struct ccid_operations *ccid_ops) ccid_ops->ccid_hc_rx_slab = ccid_kmem_cache_create(ccid_ops->ccid_hc_rx_obj_size, - "%s_hc_rx_sock", - ccid_ops->ccid_name); + "ccid%u_hc_rx_sock", + ccid_ops->ccid_id); if (ccid_ops->ccid_hc_rx_slab == NULL) goto out; ccid_ops->ccid_hc_tx_slab = ccid_kmem_cache_create(ccid_ops->ccid_hc_tx_obj_size, - "%s_hc_tx_sock", - ccid_ops->ccid_name); + "ccid%u_hc_tx_sock", + ccid_ops->ccid_id); if (ccid_ops->ccid_hc_tx_slab == NULL) goto out_free_rx_slab; diff --git a/net/dccp/ccid.h b/net/dccp/ccid.h index c65cb2453e4..fdeae7b5731 100644 --- a/net/dccp/ccid.h +++ b/net/dccp/ccid.h @@ -23,14 +23,37 @@ struct tcp_info; +/** + * struct ccid_operations - Interface to Congestion-Control Infrastructure + * + * @ccid_id: numerical CCID ID (up to %CCID_MAX, cf. table 5 in RFC 4340, 10.) + * @ccid_ccmps: the CCMPS including network/transport headers (0 when disabled) + * @ccid_name: alphabetical identifier string for @ccid_id + * @ccid_owner: module which implements/owns this CCID + * @ccid_hc_{r,t}x_slab: memory pool for the receiver/sender half-connection + * @ccid_hc_{r,t}x_obj_size: size of the receiver/sender half-connection socket + * + * @ccid_hc_{r,t}x_init: CCID-specific initialisation routine (before startup) + * @ccid_hc_{r,t}x_exit: CCID-specific cleanup routine (before destruction) + * @ccid_hc_rx_packet_recv: implements the HC-receiver side + * @ccid_hc_{r,t}x_parse_options: parsing routine for CCID/HC-specific options + * @ccid_hc_{r,t}x_insert_options: insert routine for CCID/HC-specific options + * @ccid_hc_tx_packet_recv: implements feedback processing for the HC-sender + * @ccid_hc_tx_send_packet: implements the sending part of the HC-sender + * @ccid_hc_tx_packet_sent: does accounting for packets in flight by HC-sender + * @ccid_hc_{r,t}x_get_info: INET_DIAG information for HC-receiver/sender + * @ccid_hc_{r,t}x_getsockopt: socket options specific to HC-receiver/sender + */ struct ccid_operations { - unsigned char ccid_id; - const char *ccid_name; - struct module *ccid_owner; - struct kmem_cache *ccid_hc_rx_slab; - __u32 ccid_hc_rx_obj_size; - struct kmem_cache *ccid_hc_tx_slab; - __u32 ccid_hc_tx_obj_size; + unsigned char ccid_id; + __u32 ccid_ccmps; + const char *ccid_name; + struct module *ccid_owner; + struct kmem_cache *ccid_hc_rx_slab, + *ccid_hc_tx_slab; + __u32 ccid_hc_rx_obj_size, + ccid_hc_tx_obj_size; + /* Interface Routines */ int (*ccid_hc_rx_init)(struct ccid *ccid, struct sock *sk); int (*ccid_hc_tx_init)(struct ccid *ccid, struct sock *sk); void (*ccid_hc_rx_exit)(struct sock *sk); diff --git a/net/dccp/ccids/Kconfig b/net/dccp/ccids/Kconfig index 80f46988769..12275943eab 100644 --- a/net/dccp/ccids/Kconfig +++ b/net/dccp/ccids/Kconfig @@ -1,9 +1,8 @@ menu "DCCP CCIDs Configuration (EXPERIMENTAL)" - depends on IP_DCCP && EXPERIMENTAL + depends on EXPERIMENTAL config IP_DCCP_CCID2 tristate "CCID2 (TCP-Like) (EXPERIMENTAL)" - depends on IP_DCCP def_tristate IP_DCCP select IP_DCCP_ACKVEC ---help--- @@ -20,18 +19,9 @@ config IP_DCCP_CCID2 to the user. For example, a hypothetical application that transferred files over DCCP, using application-level retransmissions for lost packets, would prefer CCID 2 to CCID 3. On-line games may - also prefer CCID 2. + also prefer CCID 2. See RFC 4341 for further details. - CCID 2 is further described in RFC 4341, - http://www.ietf.org/rfc/rfc4341.txt - - This text was extracted from RFC 4340 (sec. 10.1), - http://www.ietf.org/rfc/rfc4340.txt - - To compile this CCID as a module, choose M here: the module will be - called dccp_ccid2. - - If in doubt, say M. + CCID2 is the default CCID used by DCCP. config IP_DCCP_CCID2_DEBUG bool "CCID2 debugging messages" @@ -47,8 +37,8 @@ config IP_DCCP_CCID2_DEBUG config IP_DCCP_CCID3 tristate "CCID3 (TCP-Friendly) (EXPERIMENTAL)" - depends on IP_DCCP def_tristate IP_DCCP + select IP_DCCP_TFRC_LIB ---help--- CCID 3 denotes TCP-Friendly Rate Control (TFRC), an equation-based rate-controlled congestion control mechanism. TFRC is designed to @@ -74,10 +64,6 @@ config IP_DCCP_CCID3 If in doubt, say M. -config IP_DCCP_TFRC_LIB - depends on IP_DCCP_CCID3 - def_tristate IP_DCCP_CCID3 - config IP_DCCP_CCID3_DEBUG bool "CCID3 debugging messages" depends on IP_DCCP_CCID3 @@ -121,5 +107,13 @@ config IP_DCCP_CCID3_RTO is serious network congestion: experimenting with larger values should therefore not be performed on WANs. +config IP_DCCP_TFRC_LIB + tristate + default n + +config IP_DCCP_TFRC_DEBUG + bool + depends on IP_DCCP_TFRC_LIB + default y if IP_DCCP_CCID3_DEBUG endmenu diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c index d694656b880..b5b52ebb269 100644 --- a/net/dccp/ccids/ccid2.c +++ b/net/dccp/ccids/ccid2.c @@ -24,9 +24,6 @@ /* * This implementation should follow RFC 4341 - * - * BUGS: - * - sequence number wrapping */ #include "../ccid.h" @@ -129,50 +126,35 @@ static int ccid2_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb) { struct ccid2_hc_tx_sock *hctx = ccid2_hc_tx_sk(sk); - ccid2_pr_debug("pipe=%d cwnd=%d\n", hctx->ccid2hctx_pipe, - hctx->ccid2hctx_cwnd); - - if (hctx->ccid2hctx_pipe < hctx->ccid2hctx_cwnd) { - /* OK we can send... make sure previous packet was sent off */ - if (!hctx->ccid2hctx_sendwait) { - hctx->ccid2hctx_sendwait = 1; - return 0; - } - } + if (hctx->ccid2hctx_pipe < hctx->ccid2hctx_cwnd) + return 0; return 1; /* XXX CCID should dequeue when ready instead of polling */ } -static void ccid2_change_l_ack_ratio(struct sock *sk, int val) +static void ccid2_change_l_ack_ratio(struct sock *sk, u32 val) { struct dccp_sock *dp = dccp_sk(sk); + u32 max_ratio = DIV_ROUND_UP(ccid2_hc_tx_sk(sk)->ccid2hctx_cwnd, 2); + /* - * XXX I don't really agree with val != 2. If cwnd is 1, ack ratio - * should be 1... it shouldn't be allowed to become 2. - * -sorbo. + * Ensure that Ack Ratio does not exceed ceil(cwnd/2), which is (2) from + * RFC 4341, 6.1.2. We ignore the statement that Ack Ratio 2 is always + * acceptable since this causes starvation/deadlock whenever cwnd < 2. + * The same problem arises when Ack Ratio is 0 (ie. Ack Ratio disabled). */ - if (val != 2) { - const struct ccid2_hc_tx_sock *hctx = ccid2_hc_tx_sk(sk); - int max = hctx->ccid2hctx_cwnd / 2; - - /* round up */ - if (hctx->ccid2hctx_cwnd & 1) - max++; - - if (val > max) - val = max; + if (val == 0 || val > max_ratio) { + DCCP_WARN("Limiting Ack Ratio (%u) to %u\n", val, max_ratio); + val = max_ratio; } + if (val > 0xFFFF) /* RFC 4340, 11.3 */ + val = 0xFFFF; - ccid2_pr_debug("changing local ack ratio to %d\n", val); - WARN_ON(val <= 0); - dp->dccps_l_ack_ratio = val; -} + if (val == dp->dccps_l_ack_ratio) + return; -static void ccid2_change_cwnd(struct ccid2_hc_tx_sock *hctx, u32 val) -{ - /* XXX do we need to change ack ratio? */ - hctx->ccid2hctx_cwnd = val? : 1; - ccid2_pr_debug("changed cwnd to %u\n", hctx->ccid2hctx_cwnd); + ccid2_pr_debug("changing local ack ratio to %u\n", val); + dp->dccps_l_ack_ratio = val; } static void ccid2_change_srtt(struct ccid2_hc_tx_sock *hctx, long val) @@ -181,11 +163,6 @@ static void ccid2_change_srtt(struct ccid2_hc_tx_sock *hctx, long val) hctx->ccid2hctx_srtt = val; } -static void ccid2_change_pipe(struct ccid2_hc_tx_sock *hctx, long val) -{ - hctx->ccid2hctx_pipe = val; -} - static void ccid2_start_rto_timer(struct sock *sk); static void ccid2_hc_tx_rto_expire(unsigned long data) @@ -215,21 +192,17 @@ static void ccid2_hc_tx_rto_expire(unsigned long data) ccid2_start_rto_timer(sk); /* adjust pipe, cwnd etc */ - ccid2_change_pipe(hctx, 0); - hctx->ccid2hctx_ssthresh = hctx->ccid2hctx_cwnd >> 1; + hctx->ccid2hctx_ssthresh = hctx->ccid2hctx_cwnd / 2; if (hctx->ccid2hctx_ssthresh < 2) hctx->ccid2hctx_ssthresh = 2; - ccid2_change_cwnd(hctx, 1); + hctx->ccid2hctx_cwnd = 1; + hctx->ccid2hctx_pipe = 0; /* clear state about stuff we sent */ - hctx->ccid2hctx_seqt = hctx->ccid2hctx_seqh; - hctx->ccid2hctx_ssacks = 0; - hctx->ccid2hctx_acks = 0; - hctx->ccid2hctx_sent = 0; + hctx->ccid2hctx_seqt = hctx->ccid2hctx_seqh; + hctx->ccid2hctx_packets_acked = 0; /* clear ack ratio state. */ - hctx->ccid2hctx_arsent = 0; - hctx->ccid2hctx_ackloss = 0; hctx->ccid2hctx_rpseq = 0; hctx->ccid2hctx_rpdupack = -1; ccid2_change_l_ack_ratio(sk, 1); @@ -255,23 +228,10 @@ static void ccid2_hc_tx_packet_sent(struct sock *sk, int more, unsigned int len) struct dccp_sock *dp = dccp_sk(sk); struct ccid2_hc_tx_sock *hctx = ccid2_hc_tx_sk(sk); struct ccid2_seq *next; - u64 seq; - - ccid2_hc_tx_check_sanity(hctx); - BUG_ON(!hctx->ccid2hctx_sendwait); - hctx->ccid2hctx_sendwait = 0; - ccid2_change_pipe(hctx, hctx->ccid2hctx_pipe + 1); - BUG_ON(hctx->ccid2hctx_pipe < 0); + hctx->ccid2hctx_pipe++; - /* There is an issue. What if another packet is sent between - * packet_send() and packet_sent(). Then the sequence number would be - * wrong. - * -sorbo. - */ - seq = dp->dccps_gss; - - hctx->ccid2hctx_seqh->ccid2s_seq = seq; + hctx->ccid2hctx_seqh->ccid2s_seq = dp->dccps_gss; hctx->ccid2hctx_seqh->ccid2s_acked = 0; hctx->ccid2hctx_seqh->ccid2s_sent = jiffies; @@ -291,8 +251,26 @@ static void ccid2_hc_tx_packet_sent(struct sock *sk, int more, unsigned int len) ccid2_pr_debug("cwnd=%d pipe=%d\n", hctx->ccid2hctx_cwnd, hctx->ccid2hctx_pipe); - hctx->ccid2hctx_sent++; - + /* + * FIXME: The code below is broken and the variables have been removed + * from the socket struct. The `ackloss' variable was always set to 0, + * and with arsent there are several problems: + * (i) it doesn't just count the number of Acks, but all sent packets; + * (ii) it is expressed in # of packets, not # of windows, so the + * comparison below uses the wrong formula: Appendix A of RFC 4341 + * comes up with the number K = cwnd / (R^2 - R) of consecutive windows + * of data with no lost or marked Ack packets. If arsent were the # of + * consecutive Acks received without loss, then Ack Ratio needs to be + * decreased by 1 when + * arsent >= K * cwnd / R = cwnd^2 / (R^3 - R^2) + * where cwnd / R is the number of Acks received per window of data + * (cf. RFC 4341, App. A). The problems are that + * - arsent counts other packets as well; + * - the comparison uses a formula different from RFC 4341; + * - computing a cubic/quadratic equation each time is too complicated. + * Hence a different algorithm is needed. |