diff options
30 files changed, 657 insertions, 604 deletions
diff --git a/Documentation/networking/dccp.txt b/Documentation/networking/dccp.txt index dda15886bcb..387482e46c4 100644 --- a/Documentation/networking/dccp.txt +++ b/Documentation/networking/dccp.txt @@ -19,7 +19,8 @@ for real time and multimedia traffic. It has a base protocol and pluggable congestion control IDs (CCIDs). -It is at experimental RFC status and the homepage for DCCP as a protocol is at: +It is at proposed standard RFC status and the homepage for DCCP as a protocol +is at: http://www.read.cs.ucla.edu/dccp/ Missing features @@ -34,9 +35,6 @@ The known bugs are at: Socket options ============== -DCCP_SOCKOPT_PACKET_SIZE is used for CCID3 to set default packet size for -calculations. - DCCP_SOCKOPT_SERVICE sets the service. The specification mandates use of service codes (RFC 4340, sec. 8.1.2); if this socket option is not set, the socket will fall back to 0 (which means that no meaningful service code diff --git a/crypto/sha512.c b/crypto/sha512.c index 2dfe7f170b4..15eab9db9be 100644 --- a/crypto/sha512.c +++ b/crypto/sha512.c @@ -24,7 +24,7 @@ #define SHA384_DIGEST_SIZE 48 #define SHA512_DIGEST_SIZE 64 -#define SHA384_HMAC_BLOCK_SIZE 96 +#define SHA384_HMAC_BLOCK_SIZE 128 #define SHA512_HMAC_BLOCK_SIZE 128 struct sha512_ctx { diff --git a/drivers/atm/.gitignore b/drivers/atm/.gitignore index a165b716771..fc0ae5eb05d 100644 --- a/drivers/atm/.gitignore +++ b/drivers/atm/.gitignore @@ -2,4 +2,4 @@ fore200e_mkfirm fore200e_pca_fw.c pca200e.bin - +pca200e_ecd.bin2 diff --git a/drivers/net/hamradio/baycom_epp.c b/drivers/net/hamradio/baycom_epp.c index 8a83db0fb3b..153b6dc80af 100644 --- a/drivers/net/hamradio/baycom_epp.c +++ b/drivers/net/hamradio/baycom_epp.c @@ -1177,7 +1177,7 @@ static void baycom_probe(struct net_device *dev) dev->mtu = AX25_DEF_PACLEN; /* eth_mtu is the default */ dev->addr_len = AX25_ADDR_LEN; /* sizeof an ax.25 address */ memcpy(dev->broadcast, &ax25_bcast, AX25_ADDR_LEN); - memcpy(dev->dev_addr, &ax25_nocall, AX25_ADDR_LEN); + memcpy(dev->dev_addr, &null_ax25_address, AX25_ADDR_LEN); dev->tx_queue_len = 16; /* New style flags */ diff --git a/drivers/net/wan/Kconfig b/drivers/net/wan/Kconfig index d5ab9cf1325..21f76f51c95 100644 --- a/drivers/net/wan/Kconfig +++ b/drivers/net/wan/Kconfig @@ -382,7 +382,7 @@ config SDLA # Wan router core. config WAN_ROUTER_DRIVERS - bool "WAN router drivers" + tristate "WAN router drivers" depends on WAN && WAN_ROUTER ---help--- Connect LAN to WAN via Linux box. @@ -393,7 +393,8 @@ config WAN_ROUTER_DRIVERS <file:Documentation/networking/wan-router.txt>. Note that the answer to this question won't directly affect the - kernel: saying N will just cause the configurator to skip all + kernel except for how subordinate drivers may be built: + saying N will just cause the configurator to skip all the questions about WAN router drivers. If unsure, say N. diff --git a/include/linux/dccp.h b/include/linux/dccp.h index ed6cc8962d8..1cb054bd93f 100644 --- a/include/linux/dccp.h +++ b/include/linux/dccp.h @@ -176,20 +176,20 @@ enum { }; /* DCCP features (RFC 4340 section 6.4) */ - enum { - DCCPF_RESERVED = 0, - DCCPF_CCID = 1, +enum { + DCCPF_RESERVED = 0, + DCCPF_CCID = 1, DCCPF_SHORT_SEQNOS = 2, /* XXX: not yet implemented */ - DCCPF_SEQUENCE_WINDOW = 3, + DCCPF_SEQUENCE_WINDOW = 3, DCCPF_ECN_INCAPABLE = 4, /* XXX: not yet implemented */ - DCCPF_ACK_RATIO = 5, - DCCPF_SEND_ACK_VECTOR = 6, - DCCPF_SEND_NDP_COUNT = 7, + DCCPF_ACK_RATIO = 5, + DCCPF_SEND_ACK_VECTOR = 6, + DCCPF_SEND_NDP_COUNT = 7, DCCPF_MIN_CSUM_COVER = 8, DCCPF_DATA_CHECKSUM = 9, /* XXX: not yet implemented */ - /* 10-127 reserved */ - DCCPF_MIN_CCID_SPECIFIC = 128, - DCCPF_MAX_CCID_SPECIFIC = 255, + /* 10-127 reserved */ + DCCPF_MIN_CCID_SPECIFIC = 128, + DCCPF_MAX_CCID_SPECIFIC = 255, }; /* this structure is argument to DCCP_SOCKOPT_CHANGE_X */ @@ -427,7 +427,7 @@ struct dccp_service_list { }; #define DCCP_SERVICE_INVALID_VALUE htonl((__u32)-1) -#define DCCP_SERVICE_CODE_IS_ABSENT 0 +#define DCCP_SERVICE_CODE_IS_ABSENT 0 static inline int dccp_list_has_service(const struct dccp_service_list *sl, const __be32 service) @@ -436,7 +436,7 @@ static inline int dccp_list_has_service(const struct dccp_service_list *sl, u32 i = sl->dccpsl_nr; while (i--) if (sl->dccpsl_list[i] == service) - return 1; + return 1; } return 0; } @@ -511,7 +511,7 @@ struct dccp_sock { __u8 dccps_hc_tx_insert_options:1; struct timer_list dccps_xmit_timer; }; - + static inline struct dccp_sock *dccp_sk(const struct sock *sk) { return (struct dccp_sock *)sk; diff --git a/include/linux/tfrc.h b/include/linux/tfrc.h index 31a9b25276f..8a8462b4a4d 100644 --- a/include/linux/tfrc.h +++ b/include/linux/tfrc.h @@ -37,10 +37,14 @@ struct tfrc_rx_info { * @tfrctx_p: current loss event rate (5.4) * @tfrctx_rto: estimate of RTO, equals 4*RTT (4.3) * @tfrctx_ipi: inter-packet interval (4.6) + * + * Note: X and X_recv are both maintained in units of 64 * bytes/second. This + * enables a finer resolution of sending rates and avoids problems with + * integer arithmetic; u32 is not sufficient as scaling consumes 6 bits. */ struct tfrc_tx_info { - __u32 tfrctx_x; - __u32 tfrctx_x_recv; + __u64 tfrctx_x; + __u64 tfrctx_x_recv; __u32 tfrctx_x_calc; __u32 tfrctx_rtt; __u32 tfrctx_p; diff --git a/include/net/ax25.h b/include/net/ax25.h index e1d116f1192..14b72d868f0 100644 --- a/include/net/ax25.h +++ b/include/net/ax25.h @@ -285,6 +285,8 @@ extern struct sock *ax25_make_new(struct sock *, struct ax25_dev *); extern const ax25_address ax25_bcast; extern const ax25_address ax25_defaddr; extern const ax25_address null_ax25_address; +extern char *ax2asc(char *buf, const ax25_address *); +extern void asc2ax(ax25_address *addr, const char *callsign); extern int ax25cmp(const ax25_address *, const ax25_address *); extern int ax25digicmp(const ax25_digi *, const ax25_digi *); extern const unsigned char *ax25_addr_parse(const unsigned char *, int, diff --git a/net/ax25/ax25_addr.c b/net/ax25/ax25_addr.c index 21a0616152f..97a49c79c60 100644 --- a/net/ax25/ax25_addr.c +++ b/net/ax25/ax25_addr.c @@ -83,7 +83,7 @@ EXPORT_SYMBOL(ax2asc); */ void asc2ax(ax25_address *addr, const char *callsign) { - char *s; + const char *s; int n; for (s = callsign, n = 0; n < 6; n++) { diff --git a/net/core/netpoll.c b/net/core/netpoll.c index 8a271285f2f..a01abdd2d3e 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -242,22 +242,28 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb) /* don't get messages out of order, and no recursion */ if (skb_queue_len(&npinfo->txq) == 0 && - npinfo->poll_owner != smp_processor_id() && - netif_tx_trylock(dev)) { - /* try until next clock tick */ - for (tries = jiffies_to_usecs(1)/USEC_PER_POLL; tries > 0; --tries) { - if (!netif_queue_stopped(dev)) - status = dev->hard_start_xmit(skb, dev); + npinfo->poll_owner != smp_processor_id()) { + unsigned long flags; - if (status == NETDEV_TX_OK) - break; + local_irq_save(flags); + if (netif_tx_trylock(dev)) { + /* try until next clock tick */ + for (tries = jiffies_to_usecs(1)/USEC_PER_POLL; + tries > 0; --tries) { + if (!netif_queue_stopped(dev)) + status = dev->hard_start_xmit(skb, dev); - /* tickle device maybe there is some cleanup */ - netpoll_poll(np); + if (status == NETDEV_TX_OK) + break; - udelay(USEC_PER_POLL); + /* tickle device maybe there is some cleanup */ + netpoll_poll(np); + + udelay(USEC_PER_POLL); + } + netif_tx_unlock(dev); } - netif_tx_unlock(dev); + local_irq_restore(flags); } if (status != NETDEV_TX_OK) { diff --git a/net/dccp/ackvec.c b/net/dccp/ackvec.c index 1f4727ddbdb..a086c6312d3 100644 --- a/net/dccp/ackvec.c +++ b/net/dccp/ackvec.c @@ -223,7 +223,7 @@ static inline int dccp_ackvec_set_buf_head_state(struct dccp_ackvec *av, gap = -new_head; } new_head += DCCP_MAX_ACKVEC_LEN; - } + } av->dccpav_buf_head = new_head; @@ -336,7 +336,7 @@ out_duplicate: void dccp_ackvector_print(const u64 ackno, const unsigned char *vector, int len) { dccp_pr_debug_cat("ACK vector len=%d, ackno=%llu |", len, - (unsigned long long)ackno); + (unsigned long long)ackno); while (len--) { const u8 state = (*vector & DCCP_ACKVEC_STATE_MASK) >> 6; diff --git a/net/dccp/ccid.h b/net/dccp/ccid.h index bcc2d12ae81..c65cb2453e4 100644 --- a/net/dccp/ccid.h +++ b/net/dccp/ccid.h @@ -43,8 +43,6 @@ struct ccid_operations { unsigned char* value); int (*ccid_hc_rx_insert_options)(struct sock *sk, struct sk_buff *skb); - int (*ccid_hc_tx_insert_options)(struct sock *sk, - struct sk_buff *skb); void (*ccid_hc_tx_packet_recv)(struct sock *sk, struct sk_buff *skb); int (*ccid_hc_tx_parse_options)(struct sock *sk, @@ -146,14 +144,6 @@ static inline int ccid_hc_rx_parse_options(struct ccid *ccid, struct sock *sk, return rc; } -static inline int ccid_hc_tx_insert_options(struct ccid *ccid, struct sock *sk, - struct sk_buff *skb) -{ - if (ccid->ccid_ops->ccid_hc_tx_insert_options != NULL) - return ccid->ccid_ops->ccid_hc_tx_insert_options(sk, skb); - return 0; -} - static inline int ccid_hc_rx_insert_options(struct ccid *ccid, struct sock *sk, struct sk_buff *skb) { diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c index 2555be8f479..fd38b05d6f7 100644 --- a/net/dccp/ccids/ccid2.c +++ b/net/dccp/ccids/ccid2.c @@ -351,7 +351,7 @@ static void ccid2_hc_tx_packet_sent(struct sock *sk, int more, unsigned int len) while (seqp != hctx->ccid2hctx_seqh) { ccid2_pr_debug("out seq=%llu acked=%d time=%lu\n", - (unsigned long long)seqp->ccid2s_seq, + (unsigned long long)seqp->ccid2s_seq, seqp->ccid2s_acked, seqp->ccid2s_sent); seqp = seqp->ccid2s_next; } @@ -473,7 +473,7 @@ static inline void ccid2_new_ack(struct sock *sk, /* first measurement */ if (hctx->ccid2hctx_srtt == -1) { ccid2_pr_debug("R: %lu Time=%lu seq=%llu\n", - r, jiffies, + r, jiffies, (unsigned long long)seqp->ccid2s_seq); ccid2_change_srtt(hctx, r); hctx->ccid2hctx_rttvar = r >> 1; @@ -518,8 +518,8 @@ static inline void ccid2_new_ack(struct sock *sk, hctx->ccid2hctx_lastrtt = jiffies; ccid2_pr_debug("srtt: %ld rttvar: %ld rto: %ld (HZ=%d) R=%lu\n", - hctx->ccid2hctx_srtt, hctx->ccid2hctx_rttvar, - hctx->ccid2hctx_rto, HZ, r); + hctx->ccid2hctx_srtt, hctx->ccid2hctx_rttvar, + hctx->ccid2hctx_rto, HZ, r); hctx->ccid2hctx_sent = 0; } @@ -667,9 +667,9 @@ static void ccid2_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) /* new packet received or marked */ if (state != DCCP_ACKVEC_STATE_NOT_RECEIVED && !seqp->ccid2s_acked) { - if (state == + if (state == DCCP_ACKVEC_STATE_ECN_MARKED) { - ccid2_congestion_event(hctx, + ccid2_congestion_event(hctx, seqp); } else ccid2_new_ack(sk, seqp, diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c index 66a27b9688c..fa6b75372ed 100644 --- a/net/dccp/ccids/ccid3.c +++ b/net/dccp/ccids/ccid3.c @@ -41,27 +41,6 @@ #include "lib/tfrc.h" #include "ccid3.h" -/* - * Reason for maths here is to avoid 32 bit overflow when a is big. - * With this we get close to the limit. - */ -static u32 usecs_div(const u32 a, const u32 b) -{ - const u32 div = a < (UINT_MAX / (USEC_PER_SEC / 10)) ? 10 : - a < (UINT_MAX / (USEC_PER_SEC / 50)) ? 50 : - a < (UINT_MAX / (USEC_PER_SEC / 100)) ? 100 : - a < (UINT_MAX / (USEC_PER_SEC / 500)) ? 500 : - a < (UINT_MAX / (USEC_PER_SEC / 1000)) ? 1000 : - a < (UINT_MAX / (USEC_PER_SEC / 5000)) ? 5000 : - a < (UINT_MAX / (USEC_PER_SEC / 10000)) ? 10000 : - a < (UINT_MAX / (USEC_PER_SEC / 50000)) ? 50000 : - 100000; - const u32 tmp = a * (USEC_PER_SEC / div); - return (b >= 2 * div) ? tmp / (b / div) : tmp; -} - - - #ifdef CONFIG_IP_DCCP_CCID3_DEBUG static int ccid3_debug; #define ccid3_pr_debug(format, a...) DCCP_PR_DEBUG(ccid3_debug, format, ##a) @@ -108,8 +87,9 @@ static inline void ccid3_update_send_time(struct ccid3_hc_tx_sock *hctx) { timeval_sub_usecs(&hctx->ccid3hctx_t_nom, hctx->ccid3hctx_t_ipi); - /* Calculate new t_ipi (inter packet interval) by t_ipi = s / X_inst */ - hctx->ccid3hctx_t_ipi = usecs_div(hctx->ccid3hctx_s, hctx->ccid3hctx_x); + /* Calculate new t_ipi = s / X_inst (X_inst is in 64 * bytes/second) */ + hctx->ccid3hctx_t_ipi = scaled_div(hctx->ccid3hctx_s, + hctx->ccid3hctx_x >> 6); /* Update nominal send time with regard to the new t_ipi */ timeval_add_usecs(&hctx->ccid3hctx_t_nom, hctx->ccid3hctx_t_ipi); @@ -128,40 +108,44 @@ static inline void ccid3_update_send_time(struct ccid3_hc_tx_sock *hctx) * X = max(min(2 * X, 2 * X_recv), s / R); * tld = now; * + * Note: X and X_recv are both stored in units of 64 * bytes/second, to support + * fine-grained resolution of sending rates. This requires scaling by 2^6 + * throughout the code. Only X_calc is unscaled (in bytes/second). + * * If X has changed, we also update the scheduled send time t_now, * the inter-packet interval t_ipi, and the delta value. - */ + */ static void ccid3_hc_tx_update_x(struct sock *sk, struct timeval *now) { struct ccid3_hc_tx_sock *hctx = ccid3_hc_tx_sk(sk); - const __u32 old_x = hctx->ccid3hctx_x; + const __u64 old_x = hctx->ccid3hctx_x; if (hctx->ccid3hctx_p > 0) { - hctx->ccid3hctx_x_calc = tfrc_calc_x(hctx->ccid3hctx_s, - hctx->ccid3hctx_rtt, - hctx->ccid3hctx_p); - hctx->ccid3hctx_x = max_t(u32, min(hctx->ccid3hctx_x_calc, - hctx->ccid3hctx_x_recv * 2), - hctx->ccid3hctx_s / TFRC_T_MBI); - - } else if (timeval_delta(now, &hctx->ccid3hctx_t_ld) >= - hctx->ccid3hctx_rtt) { - hctx->ccid3hctx_x = max(min(hctx->ccid3hctx_x_recv, - hctx->ccid3hctx_x ) * 2, - usecs_div(hctx->ccid3hctx_s, - hctx->ccid3hctx_rtt) ); + + hctx->ccid3hctx_x = min(((__u64)hctx->ccid3hctx_x_calc) << 6, + hctx->ccid3hctx_x_recv * 2); + hctx->ccid3hctx_x = max(hctx->ccid3hctx_x, + (((__u64)hctx->ccid3hctx_s) << 6) / + TFRC_T_MBI); + + } else if (timeval_delta(now, &hctx->ccid3hctx_t_ld) - + (suseconds_t)hctx->ccid3hctx_rtt >= 0) { + + hctx->ccid3hctx_x = + max(2 * min(hctx->ccid3hctx_x, hctx->ccid3hctx_x_recv), + scaled_div(((__u64)hctx->ccid3hctx_s) << 6, + hctx->ccid3hctx_rtt)); hctx->ccid3hctx_t_ld = *now; - } else - ccid3_pr_debug("Not changing X\n"); + } if (hctx->ccid3hctx_x != old_x) ccid3_update_send_time(hctx); } /* - * Track the mean packet size `s' (cf. RFC 4342, 5.3 and RFC 3448, 4.1) - * @len: DCCP packet payload size in bytes + * Track the mean packet size `s' (cf. RFC 4342, 5.3 and RFC 3448, 4.1) + * @len: DCCP packet payload size in bytes */ static inline void ccid3_hc_tx_update_s(struct ccid3_hc_tx_sock *hctx, int len) { @@ -178,6 +162,33 @@ static inline void ccid3_hc_tx_update_s(struct ccid3_hc_tx_sock *hctx, int len) */ } +/* + * Update Window Counter using the algorithm from [RFC 4342, 8.1]. + * The algorithm is not applicable if RTT < 4 microseconds. + */ +static inline void ccid3_hc_tx_update_win_count(struct ccid3_hc_tx_sock *hctx, + struct timeval *now) +{ + suseconds_t delta; + u32 quarter_rtts; + + if (unlikely(hctx->ccid3hctx_rtt < 4)) /* avoid divide-by-zero */ + return; + + delta = timeval_delta(now, &hctx->ccid3hctx_t_last_win_count); + DCCP_BUG_ON(delta < 0); + + quarter_rtts = (u32)delta / (hctx->ccid3hctx_rtt / 4); + + if (quarter_rtts > 0) { + hctx->ccid3hctx_t_last_win_count = *now; + hctx->ccid3hctx_last_win_count += min_t(u32, quarter_rtts, 5); + hctx->ccid3hctx_last_win_count &= 0xF; /* mod 16 */ + + ccid3_pr_debug("now at %#X\n", hctx->ccid3hctx_last_win_count); + } +} + static void ccid3_hc_tx_no_feedback_timer(unsigned long data) { struct sock *sk = (struct sock *)data; @@ -191,20 +202,20 @@ static void ccid3_hc_tx_no_feedback_timer(unsigned long data) goto restart_timer; } - ccid3_pr_debug("%s, sk=%p, state=%s\n", dccp_role(sk), sk, + ccid3_pr_debug("%s(%p, state=%s) - entry \n", dccp_role(sk), sk, ccid3_tx_state_name(hctx->ccid3hctx_state)); - + switch (hctx->ccid3hctx_state) { case TFRC_SSTATE_NO_FBACK: /* RFC 3448, 4.4: Halve send rate directly */ - hctx->ccid3hctx_x = min_t(u32, hctx->ccid3hctx_x / 2, - hctx->ccid3hctx_s / TFRC_T_MBI); + hctx->ccid3hctx_x = max(hctx->ccid3hctx_x / 2, + (((__u64)hctx->ccid3hctx_s) << 6) / + TFRC_T_MBI); - ccid3_pr_debug("%s, sk=%p, state=%s, updated tx rate to %d " - "bytes/s\n", - dccp_role(sk), sk, + ccid3_pr_debug("%s(%p, state=%s), updated tx rate to %u " + "bytes/s\n", dccp_role(sk), sk, ccid3_tx_state_name(hctx->ccid3hctx_state), - hctx->ccid3hctx_x); + (unsigned)(hctx->ccid3hctx_x >> 6)); /* The value of R is still undefined and so we can not recompute * the timout value. Keep initial value as per [RFC 4342, 5]. */ t_nfb = TFRC_INITIAL_TIMEOUT; @@ -213,34 +224,46 @@ static void ccid3_hc_tx_no_feedback_timer(unsigned long data) case TFRC_SSTATE_FBACK: /* * Check if IDLE since last timeout and recv rate is less than - * 4 packets per RTT + * 4 packets (in units of 64*bytes/sec) per RTT */ if (!hctx->ccid3hctx_idle || - (hctx->ccid3hctx_x_recv >= - 4 * usecs_div(hctx->ccid3hctx_s, hctx->ccid3hctx_rtt))) { + (hctx->ccid3hctx_x_recv >= 4 * + scaled_div(((__u64)hctx->ccid3hctx_s) << 6, + hctx->ccid3hctx_rtt))) { struct timeval now; - ccid3_pr_debug("%s, sk=%p, state=%s, not idle\n", + ccid3_pr_debug("%s(%p, state=%s), not idle\n", dccp_role(sk), sk, - ccid3_tx_state_name(hctx->ccid3hctx_state)); - /* Halve sending rate */ + ccid3_tx_state_name(hctx->ccid3hctx_state)); - /* If (p == 0 || X_calc > 2 * X_recv) + /* + * Modify the cached value of X_recv [RFC 3448, 4.4] + * + * If (p == 0 || X_calc > 2 * X_recv) * X_recv = max(X_recv / 2, s / (2 * t_mbi)); * Else * X_recv = X_calc / 4; + * + * Note that X_recv is scaled by 2^6 while X_calc is not */ BUG_ON(hctx->ccid3hctx_p && !hctx->ccid3hctx_x_calc); if (hctx->ccid3hctx_p == 0 || - hctx->ccid3hctx_x_calc > 2 * hctx->ccid3hctx_x_recv) - hctx->ccid3hctx_x_recv = max_t(u32, hctx->ccid3hctx_x_recv / 2, - hctx->ccid3hctx_s / (2 * TFRC_T_MBI)); - else - hctx->ccid3hctx_x_recv = hctx->ccid3hctx_x_calc / 4; - - /* Update sending rate */ - dccp_timestamp(sk, &now); + (hctx->ccid3hctx_x_calc > + (hctx->ccid3hctx_x_recv >> 5))) { + + hctx->ccid3hctx_x_recv = + max(hctx->ccid3hctx_x_recv / 2, + (((__u64)hctx->ccid3hctx_s) << 6) / + (2 * TFRC_T_MBI)); + + if (hctx->ccid3hctx_p == 0) + dccp_timestamp(sk, &now); + } else { + hctx->ccid3hctx_x_recv = hctx->ccid3hctx_x_calc; + hctx->ccid3hctx_x_recv <<= 4; + } + /* Now recalculate X [RFC 3448, 4.3, step (4)] */ ccid3_hc_tx_update_x(sk, &now); } /* @@ -251,7 +274,7 @@ static void ccid3_hc_tx_no_feedback_timer(unsigned long data) t_nfb = max(hctx->ccid3hctx_t_rto, 2 * hctx->ccid3hctx_t_ipi); break; case TFRC_SSTATE_NO_SENT: - DCCP_BUG("Illegal %s state NO_SENT, sk=%p", dccp_role(sk), sk); + DCCP_BUG("%s(%p) - Illegal state NO_SENT", dccp_role(sk), sk); /* fall through */ case TFRC_SSTATE_TERM: goto out; @@ -277,9 +300,8 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb) { struct dccp_sock *dp = dccp_sk(sk); struct ccid3_hc_tx_sock *hctx = ccid3_hc_tx_sk(sk); - struct dccp_tx_hist_entry *new_packet; struct timeval now; - long delay; + suseconds_t delay; BUG_ON(hctx == NULL); @@ -291,34 +313,21 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb) if (unlikely(skb->len == 0)) return -EBADMSG; - /* See if last packet allocated was not sent */ - new_packet = dccp_tx_hist_head(&hctx->ccid3hctx_hist); - if (new_packet == NULL || new_packet->dccphtx_sent) { - new_packet = dccp_tx_hist_entry_new(ccid3_tx_hist, - GFP_ATOMIC); - - if (unlikely(new_packet == NULL)) { - DCCP_WARN("%s, sk=%p, not enough mem to add to history," - "send refused\n", dccp_role(sk), sk); - return -ENOBUFS; - } - - dccp_tx_hist_add_entry(&hctx->ccid3hctx_hist, new_packet); - } - dccp_timestamp(sk, &now); switch (hctx->ccid3hctx_state) { case TFRC_SSTATE_NO_SENT: sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, - jiffies + usecs_to_jiffies(TFRC_INITIAL_TIMEOUT)); + (jiffies + + usecs_to_jiffies(TFRC_INITIAL_TIMEOUT))); hctx->ccid3hctx_last_win_count = 0; hctx->ccid3hctx_t_last_win_count = now; ccid3_hc_tx_set_state(sk, TFRC_SSTATE_NO_FBACK); - /* Set initial sending rate to 1 packet per second */ + /* Set initial sending rate X/s to 1pps (X is scaled by 2^6) */ ccid3_hc_tx_update_s(hctx, skb->len); - hctx->ccid3hctx_x = hctx->ccid3hctx_s; + hctx->ccid3hctx_x = hctx->ccid3hctx_s; + hctx->ccid3hctx_x <<= 6; /* First timeout, according to [RFC 3448, 4.2], is 1 second */ hctx->ccid3hctx_t_ipi = USEC_PER_SEC; @@ -332,77 +341,57 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb) case TFRC_SSTATE_FBACK: delay = timeval_delta(&hctx->ccid3hctx_t_nom, &now); /* - * Scheduling of packet transmissions [RFC 3448, 4.6] + * Scheduling of packet transmissions [RFC 3448, 4.6] * * if (t_now > t_nom - delta) * // send the packet now * else * // send the packet in (t_nom - t_now) milliseconds. */ - if (delay - (long)hctx->ccid3hctx_delta >= 0) + if (delay - (suseconds_t)hctx->ccid3hctx_delta >= 0) return delay / 1000L; + + ccid3_hc_tx_update_win_count(hctx, &now); break; case TFRC_SSTATE_TERM: - DCCP_BUG("Illegal %s state TERM, sk=%p", dccp_role(sk), sk); + DCCP_BUG("%s(%p) - Illegal state TERM", dccp_role(sk), sk); return -EINVAL; } /* prepare to send now (add options etc.) */ dp->dccps_hc_tx_insert_options = 1; - new_packet->dccphtx_ccval = DCCP_SKB_CB(skb)->dccpd_ccval = - hctx->ccid3hctx_last_win_count; + DCCP_SKB_CB(skb)->dccpd_ccval = hctx->ccid3hctx_last_win_count; + + /* set the nominal send time for the next following packet */ timeval_add_usecs(&hctx->ccid3hctx_t_nom, hctx->ccid3hctx_t_ipi); return 0; } -static void ccid3_hc_tx_packet_sent(struct sock *sk, int more, unsigned int len) +static void ccid3_hc_tx_packet_sent(struct sock *sk, int more, + unsigned int len) { - const struct dccp_sock *dp = dccp_sk(sk); struct ccid3_hc_tx_sock *hctx = ccid3_hc_tx_sk(sk); struct timeval now; - unsigned long quarter_rtt; struct dccp_tx_hist_entry *packet; BUG_ON(hctx == NULL); - dccp_timestamp(sk, &now); - ccid3_hc_tx_update_s(hctx, len); - packet = dccp_tx_hist_head(&hctx->ccid3hctx_hist); + packet = dccp_tx_hist_entry_new(ccid3_tx_hist, GFP_ATOMIC); if (unlikely(packet == NULL)) { - DCCP_WARN("packet doesn't exist in history!\n"); - return; - } - if (unlikely(packet->dccphtx_sent)) { - DCCP_WARN("no unsent packet in history!\n"); + DCCP_CRIT("packet history - out of memory!"); return; } - packet->dccphtx_tstamp = now; - packet->dccphtx_seqno = dp->dccps_gss; - /* - * Check if win_count have changed - * Algorithm in "8.1. Window Counter Value" in RFC 4342. - */ - quarter_rtt = timeval_delta(&now, &hctx->ccid3hctx_t_last_win_count); - if (likely(hctx->ccid3hctx_rtt > 8)) - quarter_rtt /= hctx->ccid3hctx_rtt / 4; - - if (quarter_rtt > 0) { - hctx->ccid3hctx_t_last_win_count = now; - hctx->ccid3hctx_last_win_count = (hctx->ccid3hctx_last_win_count + - min_t(unsigned long, quarter_rtt, 5)) % 16; - ccid3_pr_debug("%s, sk=%p, window changed from " - "%u to %u!\n", - dccp_role(sk), sk, - packet->dccphtx_ccval, - hctx->ccid3hctx_last_win_count); - } + dccp_tx_hist_add_entry(&hctx->ccid3hctx_hist, packet); - hctx->ccid3hctx_idle = 0; - packet->dccphtx_rtt = hctx->ccid3hctx_rtt; - packet->dccphtx_sent = 1; + dccp_timestamp(sk, &now); + packet->dccphtx_tstamp = now; + packet->dccphtx_seqno = dccp_sk(sk)->dccps_gss; + packet->dccphtx_rtt = hctx->ccid3hctx_rtt; + packet->dccphtx_sent = 1; + hctx->ccid3hctx_idle = 0; } static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) @@ -414,7 +403,7 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) struct timeval now; unsigned long t_nfb; u32 pinv; - long r_sample, t_elapsed; + suseconds_t r_sample, t_elapsed; BUG_ON(hctx == NULL); @@ -430,44 +419,44 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) case TFRC_SSTATE_FBACK: /* get packet from history to look up t_recvdata */ packet = dccp_tx_hist_find_entry(&hctx->ccid3hctx_hist, - DCCP_SKB_CB(skb)->dccpd_ack_seq); + DCCP_SKB_CB(skb)->dccpd_ack_seq); if (unlikely(packet == NULL)) { DCCP_WARN("%s(%p), seqno %llu(%s) doesn't exist " "in history!\n", dccp_role(sk), sk, (unsigned long long)DCCP_SKB_CB(skb)->dccpd_ack_seq, - dccp_packet_name(DCCP_SKB_CB(skb)->dccpd_type)); + dccp_packet_name(DCCP_SKB_CB(skb)->dccpd_type)); return; } - /* Update receive rate */ + /* Update receive rate in units of 64 * bytes/second */ hctx->ccid3hctx_x_recv = opt_recv->ccid3or_receive_rate; + hctx->ccid3hctx_x_recv <<= 6; /* Update loss event rate */ pinv = opt_recv->ccid3or_loss_event_rate; - if (pinv == ~0U || pinv == 0) + if (pinv == ~0U || pinv == 0) /* see RFC 4342, 8.5 */ hctx->ccid3hctx_p = 0; - else - hctx->ccid3hctx_p = 1000000 / pinv; + else /* can not exceed 100% */ + hctx->ccid3hctx_p = 1000000 / pinv; dccp_timestamp(sk, &now); /* * Calculate new round trip sample as per [RFC 3448, 4.3] by - * R_sample = (now - t_recvdata) - t_elapsed + * R_sample = (now - t_recvdata) - t_elapsed */ r_sample = timeval_delta(&now, &packet->dccphtx_tstamp); t_elapsed = dp->dccps_options_received.dccpor_elapsed_time * 10; - if (unlikely(r_sample <= 0)) { - DCCP_WARN("WARNING: R_sample (%ld) <= 0!\n", r_sample); - r_sample = 0; - } else if (unlikely(r_sample <= t_elapsed)) - DCCP_WARN("WARNING: r_sample=%ldus <= t_elapsed=%ldus\n", - r_sample, t_elapsed); + DCCP_BUG_ON(r_sample < 0); + if (unlikely(r_sample <= t_elapsed)) + DCCP_WARN("WARNING: r_sample=%dus <= t_elapsed=%dus\n", + (int)r_sample, (int)t_elapsed); else r_sample -= t_elapsed; + CCID3_RTT_SANITY_CHECK(r_sample); - /* Update RTT estimate by + /* Update RTT estimate by * If (No feedback recv) * R = R_sample; * Else @@ -476,34 +465,45 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) * q is a constant, RFC 3448 recomments 0.9 */ if (hctx->ccid3hctx_state == TFRC_SSTATE_NO_FBACK) { - /* Use Larger Initial Windows [RFC 4342, sec. 5] - * We deviate in that we use `s' instead of `MSS'. */ - u16 w_init = max( 4 * hctx->ccid3hctx_s, - max(2 * hctx->ccid3hctx_s, 4380)); + /* + * Larger Initial Windows [RFC 4342, sec. 5] + * We deviate in that we use `s' instead of `MSS'. + */ + __u64 w_init = min(4 * hctx->ccid3hctx_s, + max(2 * hctx->ccid3hctx_s, 4380)); hctx->ccid3hctx_rtt = r_sample; - |