diff options
Diffstat (limited to 'net/dccp')
30 files changed, 2139 insertions, 1929 deletions
diff --git a/net/dccp/Kconfig b/net/dccp/Kconfig index ef8919cca74..b8a68dd4100 100644 --- a/net/dccp/Kconfig +++ b/net/dccp/Kconfig @@ -38,6 +38,9 @@ config IP_DCCP_DEBUG ---help--- Only use this if you're hacking DCCP. + When compiling DCCP as a module, this debugging output can be toggled + by setting the parameter dccp_debug of the `dccp' module to 0 or 1. + Just say N. config NET_DCCPPROBE @@ -49,7 +52,7 @@ config NET_DCCPPROBE DCCP congestion avoidance modules. If you don't understand what was just said, you don't need it: say N. - Documentation on how to use the packet generator can be found + Documentation on how to use DCCP connection probing can be found at http://linux-net.osdl.org/index.php/DccpProbe To compile this code as a module, choose M here: the diff --git a/net/dccp/Makefile b/net/dccp/Makefile index 17ed99c4661..f4f8793aaff 100644 --- a/net/dccp/Makefile +++ b/net/dccp/Makefile @@ -1,13 +1,13 @@ -obj-$(CONFIG_IPV6) += dccp_ipv6.o - -dccp_ipv6-y := ipv6.o - obj-$(CONFIG_IP_DCCP) += dccp.o dccp_ipv4.o dccp-y := ccid.o feat.o input.o minisocks.o options.o output.o proto.o timer.o dccp_ipv4-y := ipv4.o +# build dccp_ipv6 as module whenever either IPv6 or DCCP is a module +obj-$(subst y,$(CONFIG_IP_DCCP),$(CONFIG_IPV6)) += dccp_ipv6.o +dccp_ipv6-y := ipv6.o + dccp-$(CONFIG_IP_DCCP_ACKVEC) += ackvec.o obj-$(CONFIG_INET_DCCP_DIAG) += dccp_diag.o diff --git a/net/dccp/ackvec.c b/net/dccp/ackvec.c index f8208874ac7..a086c6312d3 100644 --- a/net/dccp/ackvec.c +++ b/net/dccp/ackvec.c @@ -21,8 +21,8 @@ #include <net/sock.h> -static kmem_cache_t *dccp_ackvec_slab; -static kmem_cache_t *dccp_ackvec_record_slab; +static struct kmem_cache *dccp_ackvec_slab; +static struct kmem_cache *dccp_ackvec_record_slab; static struct dccp_ackvec_record *dccp_ackvec_record_new(void) { @@ -67,15 +67,16 @@ static void dccp_ackvec_insert_avr(struct dccp_ackvec *av, int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) { struct dccp_sock *dp = dccp_sk(sk); -#ifdef CONFIG_IP_DCCP_DEBUG - const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? - "CLIENT tx: " : "server tx: "; -#endif struct dccp_ackvec *av = dp->dccps_hc_rx_ackvec; - int len = av->dccpav_vec_len + 2; + /* Figure out how many options do we need to represent the ackvec */ + const u16 nr_opts = (av->dccpav_vec_len + + DCCP_MAX_ACKVEC_OPT_LEN - 1) / + DCCP_MAX_ACKVEC_OPT_LEN; + u16 len = av->dccpav_vec_len + 2 * nr_opts, i; struct timeval now; u32 elapsed_time; - unsigned char *to, *from; + const unsigned char *tail, *from; + unsigned char *to; struct dccp_ackvec_record *avr; if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) @@ -94,24 +95,37 @@ 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); - *to++ = DCCPO_ACK_VECTOR_0; - *to++ = 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; + + for (i = 0; i < nr_opts; ++i) { + int copylen = len; + + if (len > DCCP_MAX_ACKVEC_OPT_LEN) + copylen = DCCP_MAX_ACKVEC_OPT_LEN; - /* Check if buf_head wraps */ - if ((int)av->dccpav_buf_head + len > DCCP_MAX_ACKVEC_LEN) { - const u32 tailsize = DCCP_MAX_ACKVEC_LEN - av->dccpav_buf_head; + *to++ = DCCPO_ACK_VECTOR_0; + *to++ = copylen + 2; - memcpy(to, from, tailsize); - to += tailsize; - len -= tailsize; - from = av->dccpav_buf; + /* Check if buf_head wraps */ + if (from + copylen > tail) { + const u16 tailsize = tail - from; + + memcpy(to, from, tailsize); + to += tailsize; + len -= tailsize; + copylen -= tailsize; + from = av->dccpav_buf; + } + + memcpy(to, from, copylen); + from += copylen; + to += copylen; + len -= copylen; } - memcpy(to, from, len); /* * From RFC 4340, A.2: * @@ -129,9 +143,9 @@ int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb) dccp_ackvec_insert_avr(av, avr); - dccp_pr_debug("%sACK Vector 0, len=%d, ack_seqno=%llu, " + dccp_pr_debug("%s ACK Vector 0, len=%d, ack_seqno=%llu, " "ack_ackno=%llu\n", - debug_prefix, avr->dccpavr_sent_len, + dccp_role(sk), avr->dccpavr_sent_len, (unsigned long long)avr->dccpavr_ack_seqno, (unsigned long long)avr->dccpavr_ack_ackno); return 0; @@ -145,7 +159,6 @@ struct dccp_ackvec *dccp_ackvec_alloc(const gfp_t priority) av->dccpav_buf_head = DCCP_MAX_ACKVEC_LEN - 1; av->dccpav_buf_ackno = DCCP_MAX_SEQNO + 1; av->dccpav_buf_nonce = av->dccpav_buf_nonce = 0; - av->dccpav_ack_ptr = 0; av->dccpav_time.tv_sec = 0; av->dccpav_time.tv_usec = 0; av->dccpav_vec_len = 0; @@ -174,13 +187,13 @@ void dccp_ackvec_free(struct dccp_ackvec *av) } static inline u8 dccp_ackvec_state(const struct dccp_ackvec *av, - const u8 index) + const u32 index) { return av->dccpav_buf[index] & DCCP_ACKVEC_STATE_MASK; } static inline u8 dccp_ackvec_len(const struct dccp_ackvec *av, - const u8 index) + const u32 index) { return av->dccpav_buf[index] & DCCP_ACKVEC_LEN_MASK; } @@ -210,7 +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; @@ -280,7 +293,7 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, * could reduce the complexity of this scan.) */ u64 delta = dccp_delta_seqno(ackno, av->dccpav_buf_ackno); - u8 index = av->dccpav_buf_head; + u32 index = av->dccpav_buf_head; while (1) { const u8 len = dccp_ackvec_len(av, index); @@ -322,21 +335,18 @@ out_duplicate: #ifdef CONFIG_IP_DCCP_DEBUG void dccp_ackvector_print(const u64 ackno, const unsigned char *vector, int len) { - if (!dccp_debug) - return; - - printk("ACK vector len=%d, ackno=%llu |", len, - (unsigned long long)ackno); + dccp_pr_debug_cat("ACK vector len=%d, ackno=%llu |", len, + (unsigned long long)ackno); while (len--) { const u8 state = (*vector & DCCP_ACKVEC_STATE_MASK) >> 6; const u8 rl = *vector & DCCP_ACKVEC_LEN_MASK; - printk("%d,%d|", state, rl); + dccp_pr_debug_cat("%d,%d|", state, rl); ++vector; } - printk("\n"); + dccp_pr_debug_cat("\n"); } void dccp_ackvec_print(const struct dccp_ackvec *av) @@ -380,24 +390,20 @@ void dccp_ackvec_check_rcv_ackno(struct dccp_ackvec *av, struct sock *sk, */ list_for_each_entry_reverse(avr, &av->dccpav_records, dccpavr_node) { if (ackno == avr->dccpavr_ack_seqno) { -#ifdef CONFIG_IP_DCCP_DEBUG - struct dccp_sock *dp = dccp_sk(sk); - const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? - "CLIENT rx ack: " : "server rx ack: "; -#endif - dccp_pr_debug("%sACK packet 0, len=%d, ack_seqno=%llu, " + dccp_pr_debug("%s ACK packet 0, len=%d, ack_seqno=%llu, " "ack_ackno=%llu, ACKED!\n", - debug_prefix, 1, + dccp_role(sk), 1, (unsigned long long)avr->dccpavr_ack_seqno, (unsigned long long)avr->dccpavr_ack_ackno); dccp_ackvec_throw_record(av, avr); break; - } + } else if (avr->dccpavr_ack_seqno > ackno) + break; /* old news */ } } static void dccp_ackvec_check_rcv_ackvector(struct dccp_ackvec *av, - struct sock *sk, u64 ackno, + struct sock *sk, u64 *ackno, const unsigned char len, const unsigned char *vector) { @@ -420,7 +426,7 @@ static void dccp_ackvec_check_rcv_ackvector(struct dccp_ackvec *av, const u8 rl = *vector & DCCP_ACKVEC_LEN_MASK; u64 ackno_end_rl; - dccp_set_seqno(&ackno_end_rl, ackno - rl); + dccp_set_seqno(&ackno_end_rl, *ackno - rl); /* * If our AVR sequence number is greater than the ack, go @@ -428,25 +434,19 @@ static void dccp_ackvec_check_rcv_ackvector(struct dccp_ackvec *av, */ list_for_each_entry_from(avr, &av->dccpav_records, dccpavr_node) { - if (!after48(avr->dccpavr_ack_seqno, ackno)) + if (!after48(avr->dccpavr_ack_seqno, *ackno)) goto found; } /* End of the dccpav_records list, not found, exit */ break; found: - if (between48(avr->dccpavr_ack_seqno, ackno_end_rl, ackno)) { + if (between48(avr->dccpavr_ack_seqno, ackno_end_rl, *ackno)) { const u8 state = *vector & DCCP_ACKVEC_STATE_MASK; if (state != DCCP_ACKVEC_STATE_NOT_RECEIVED) { -#ifdef CONFIG_IP_DCCP_DEBUG - struct dccp_sock *dp = dccp_sk(sk); - const char *debug_prefix = - dp->dccps_role == DCCP_ROLE_CLIENT ? - "CLIENT rx ack: " : "server rx ack: "; -#endif - dccp_pr_debug("%sACK vector 0, len=%d, " + dccp_pr_debug("%s ACK vector 0, len=%d, " "ack_seqno=%llu, ack_ackno=%llu, " "ACKED!\n", - debug_prefix, len, + dccp_role(sk), len, (unsigned long long) avr->dccpavr_ack_seqno, (unsigned long long) @@ -460,27 +460,23 @@ found: */ } - dccp_set_seqno(&ackno, ackno_end_rl - 1); + dccp_set_seqno(ackno, ackno_end_rl - 1); ++vector; } } int dccp_ackvec_parse(struct sock *sk, const struct sk_buff *skb, - const u8 opt, const u8 *value, const u8 len) + u64 *ackno, const u8 opt, const u8 *value, const u8 len) { - if (len > DCCP_MAX_ACKVEC_LEN) + if (len > DCCP_MAX_ACKVEC_OPT_LEN) return -1; /* dccp_ackvector_print(DCCP_SKB_CB(skb)->dccpd_ack_seq, value, len); */ dccp_ackvec_check_rcv_ackvector(dccp_sk(sk)->dccps_hc_rx_ackvec, sk, - DCCP_SKB_CB(skb)->dccpd_ack_seq, - len, value); + ackno, len, value); return 0; } -static char dccp_ackvec_slab_msg[] __initdata = - KERN_CRIT "DCCP: Unable to create ack vectors slab caches\n"; - int __init dccp_ackvec_init(void) { dccp_ackvec_slab = kmem_cache_create("dccp_ackvec", @@ -502,7 +498,7 @@ out_destroy_slab: kmem_cache_destroy(dccp_ackvec_slab); dccp_ackvec_slab = NULL; out_err: - printk(dccp_ackvec_slab_msg); + DCCP_CRIT("Unable to create Ack Vector slab cache"); return -ENOBUFS; } diff --git a/net/dccp/ackvec.h b/net/dccp/ackvec.h index cf8f20ce23a..96504a3b16e 100644 --- a/net/dccp/ackvec.h +++ b/net/dccp/ackvec.h @@ -17,7 +17,9 @@ #include <linux/types.h> /* Read about the ECN nonce to see why it is 253 */ -#define DCCP_MAX_ACKVEC_LEN 253 +#define DCCP_MAX_ACKVEC_OPT_LEN 253 +/* We can spread an ack vector across multiple options */ +#define DCCP_MAX_ACKVEC_LEN (DCCP_MAX_ACKVEC_OPT_LEN * 2) #define DCCP_ACKVEC_STATE_RECEIVED 0 #define DCCP_ACKVEC_STATE_ECN_MARKED (1 << 6) @@ -41,7 +43,6 @@ * Ack Vectors it has recently sent. For each packet sent carrying an * Ack Vector, it remembers four variables: * - * @dccpav_ack_ptr - the value of buf_head at the time of acknowledgement. * @dccpav_records - list of dccp_ackvec_record * @dccpav_ack_nonce - the one-bit sum of the ECN Nonces for all State 0. * @@ -52,9 +53,8 @@ struct dccp_ackvec { u64 dccpav_buf_ackno; struct list_head dccpav_records; struct timeval dccpav_time; - u8 dccpav_buf_head; - u8 dccpav_ack_ptr; - u8 dccpav_vec_len; + u16 dccpav_buf_head; + u16 dccpav_vec_len; u8 dccpav_buf_nonce; u8 dccpav_ack_nonce; u8 dccpav_buf[DCCP_MAX_ACKVEC_LEN]; @@ -77,9 +77,9 @@ struct dccp_ackvec_record { struct list_head dccpavr_node; u64 dccpavr_ack_seqno; u64 dccpavr_ack_ackno; - u8 dccpavr_ack_ptr; + u16 dccpavr_ack_ptr; + u16 dccpavr_sent_len; u8 dccpavr_ack_nonce; - u8 dccpavr_sent_len; }; struct sock; @@ -98,7 +98,8 @@ extern int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk, extern void dccp_ackvec_check_rcv_ackno(struct dccp_ackvec *av, struct sock *sk, const u64 ackno); extern int dccp_ackvec_parse(struct sock *sk, const struct sk_buff *skb, - const u8 opt, const u8 *value, const u8 len); + u64 *ackno, const u8 opt, + const u8 *value, const u8 len); extern int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb); @@ -137,7 +138,8 @@ static inline void dccp_ackvec_check_rcv_ackno(struct dccp_ackvec *av, } static inline int dccp_ackvec_parse(struct sock *sk, const struct sk_buff *skb, - const u8 opt, const u8 *value, const u8 len) + const u64 *ackno, const u8 opt, + const u8 *value, const u8 len) { return -1; } diff --git a/net/dccp/ccid.c b/net/dccp/ccid.c index ff05e59043c..d8cf92f09e6 100644 --- a/net/dccp/ccid.c +++ b/net/dccp/ccid.c @@ -55,9 +55,9 @@ static inline void ccids_read_unlock(void) #define ccids_read_unlock() do { } while(0) #endif -static kmem_cache_t *ccid_kmem_cache_create(int obj_size, const char *fmt,...) +static struct kmem_cache *ccid_kmem_cache_create(int obj_size, const char *fmt,...) { - kmem_cache_t *slab; + struct kmem_cache *slab; char slab_name_fmt[32], *slab_name; va_list args; @@ -75,7 +75,7 @@ static kmem_cache_t *ccid_kmem_cache_create(int obj_size, const char *fmt,...) return slab; } -static void ccid_kmem_cache_destroy(kmem_cache_t *slab) +static void ccid_kmem_cache_destroy(struct kmem_cache *slab) { if (slab != NULL) { const char *name = kmem_cache_name(slab); diff --git a/net/dccp/ccid.h b/net/dccp/ccid.h index f7eb6c61341..c65cb2453e4 100644 --- a/net/dccp/ccid.h +++ b/net/dccp/ccid.h @@ -27,9 +27,9 @@ struct ccid_operations { unsigned char ccid_id; const char *ccid_name; struct module *ccid_owner; - kmem_cache_t *ccid_hc_rx_slab; + struct kmem_cache *ccid_hc_rx_slab; __u32 ccid_hc_rx_obj_size; - kmem_cache_t *ccid_hc_tx_slab; + struct kmem_cache *ccid_hc_tx_slab; __u32 ccid_hc_tx_obj_size; int (*ccid_hc_rx_init)(struct ccid *ccid, struct sock *sk); int (*ccid_hc_tx_init)(struct ccid *ccid, struct sock *sk); @@ -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, @@ -52,9 +50,9 @@ struct ccid_operations { unsigned char len, u16 idx, unsigned char* value); int (*ccid_hc_tx_send_packet)(struct sock *sk, - struct sk_buff *skb, int len); - void (*ccid_hc_tx_packet_sent)(struct sock *sk, int more, - int len); + struct sk_buff *skb); + void (*ccid_hc_tx_packet_sent)(struct sock *sk, + int more, unsigned int len); void (*ccid_hc_rx_get_info)(struct sock *sk, struct tcp_info *info); void (*ccid_hc_tx_get_info)(struct sock *sk, @@ -94,16 +92,16 @@ extern void ccid_hc_rx_delete(struct ccid *ccid, struct sock *sk); extern void ccid_hc_tx_delete(struct ccid *ccid, struct sock *sk); static inline int ccid_hc_tx_send_packet(struct ccid *ccid, struct sock *sk, - struct sk_buff *skb, int len) + struct sk_buff *skb) { int rc = 0; if (ccid->ccid_ops->ccid_hc_tx_send_packet != NULL) - rc = ccid->ccid_ops->ccid_hc_tx_send_packet(sk, skb, len); + rc = ccid->ccid_ops->ccid_hc_tx_send_packet(sk, skb); return rc; } static inline void ccid_hc_tx_packet_sent(struct ccid *ccid, struct sock *sk, - int more, int len) + int more, unsigned int len) { if (ccid->ccid_ops->ccid_hc_tx_packet_sent != NULL) ccid->ccid_ops->ccid_hc_tx_packet_sent(sk, more, len); @@ -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/Kconfig b/net/dccp/ccids/Kconfig index 8533dabfb9f..80f46988769 100644 --- a/net/dccp/ccids/Kconfig +++ b/net/dccp/ccids/Kconfig @@ -28,13 +28,20 @@ config IP_DCCP_CCID2 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. config IP_DCCP_CCID2_DEBUG - bool "CCID2 debug" + bool "CCID2 debugging messages" depends on IP_DCCP_CCID2 ---help--- - Enable CCID2 debug messages. + Enable CCID2-specific debugging messages. + + When compiling CCID2 as a module, this debugging output can + additionally be toggled by setting the ccid2_debug module + parameter to 0 or 1. If in doubt, say N. @@ -62,10 +69,57 @@ config IP_DCCP_CCID3 This text was extracted from RFC 4340 (sec. 10.2), http://www.ietf.org/rfc/rfc4340.txt + To compile this CCID as a module, choose M here: the module will be + called 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 + ---help--- + Enable CCID3-specific debugging messages. + + When compiling CCID3 as a module, this debugging output can + additionally be toggled by setting the ccid3_debug module + parameter to 0 or 1. + + If in doubt, say N. + +config IP_DCCP_CCID3_RTO + int "Use higher bound for nofeedback timer" + default 100 + depends on IP_DCCP_CCID3 && EXPERIMENTAL + ---help--- + Use higher lower bound for nofeedback timer expiration. + + The TFRC nofeedback timer normally expires after the maximum of 4 + RTTs and twice the current send interval (RFC 3448, 4.3). On LANs + with a small RTT this can mean a high processing load and reduced + performance, since then the nofeedback timer is triggered very + frequently. + + This option enables to set a higher lower bound for the nofeedback + value. Values in units of milliseconds can be set here. + + A value of 0 disables this feature by enforcing the value specified + in RFC 3448. The following values have been suggested as bounds for + experimental use: + * 16-20ms to match the typical multimedia inter-frame interval + * 100ms as a reasonable compromise [default] + * 1000ms corresponds to the lower TCP RTO bound (RFC 2988, 2.4) + + The default of 100ms is a compromise between a large value for + efficient DCCP implementations, and a small value to avoid disrupting + the network in times of congestion. + + The purpose of the nofeedback timer is to slow DCCP down when there + is serious network congestion: experimenting with larger values should + therefore not be performed on WANs. + + endmenu diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c index 162032baeac..fd38b05d6f7 100644 --- a/net/dccp/ccids/ccid2.c +++ b/net/dccp/ccids/ccid2.c @@ -33,18 +33,11 @@ #include "../dccp.h" #include "ccid2.h" -static int ccid2_debug; #ifdef CONFIG_IP_DCCP_CCID2_DEBUG -#define ccid2_pr_debug(format, a...) \ - do { if (ccid2_debug) \ - printk(KERN_DEBUG "%s: " format, __FUNCTION__, ##a); \ - } while (0) -#else -#define ccid2_pr_debug(format, a...) -#endif +static int ccid2_debug; +#define ccid2_pr_debug(format, a...) DCCP_PR_DEBUG(ccid2_debug, format, ##a) -#ifdef CONFIG_IP_DCCP_CCID2_DEBUG static void ccid2_hc_tx_check_sanity(const struct ccid2_hc_tx_sock *hctx) { int len = 0; @@ -86,7 +79,8 @@ static void ccid2_hc_tx_check_sanity(const struct ccid2_hc_tx_sock *hctx) BUG_ON(len != hctx->ccid2hctx_seqbufc * CCID2_SEQBUF_LEN); } #else -#define ccid2_hc_tx_check_sanity(hctx) do {} while (0) +#define ccid2_pr_debug(format, a...) +#define ccid2_hc_tx_check_sanity(hctx) #endif static int ccid2_hc_tx_alloc_seq(struct ccid2_hc_tx_sock *hctx, int num, @@ -131,8 +125,7 @@ static int ccid2_hc_tx_alloc_seq(struct ccid2_hc_tx_sock *hctx, int num, return 0; } -static int ccid2_hc_tx_send_packet(struct sock *sk, - struct sk_buff *skb, int len) +static int ccid2_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb) { struct ccid2_hc_tx_sock *hctx; @@ -274,7 +267,7 @@ static void ccid2_start_rto_timer(struct sock *sk) jiffies + hctx->ccid2hctx_rto); } -static void ccid2_hc_tx_packet_sent(struct sock *sk, int more, int len) +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); @@ -358,7 +351,7 @@ static void ccid2_hc_tx_packet_sent(struct sock *sk, int more, 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; } @@ -426,7 +419,7 @@ static int ccid2_ackvector(struct sock *sk, struct sk_buff *skb, int offset, return -1; out_invalid_option: - BUG_ON(1); /* should never happen... options were previously parsed ! */ + DCCP_BUG("Invalid option - this should not happen (previous parsing)!"); return -1; } @@ -480,7 +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; @@ -525,8 +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; } @@ -619,7 +612,17 @@ static void ccid2_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) } ackno = DCCP_SKB_CB(skb)->dccpd_ack_seq; - seqp = hctx->ccid2hctx_seqh->ccid2s_prev; + if (after48(ackno, hctx->ccid2hctx_high_ack)) + hctx->ccid2hctx_high_ack = ackno; + + seqp = hctx->ccid2hctx_seqt; + while (before48(seqp->ccid2s_seq, ackno)) { + seqp = seqp->ccid2s_next; + if (seqp == hctx->ccid2hctx_seqh) { + seqp = hctx->ccid2hctx_seqh->ccid2s_prev; + break; + } + } /* If in slow-start, cwnd can increase at most Ack Ratio / 2 packets for * this single ack. I round up. @@ -664,9 +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, @@ -697,7 +700,14 @@ static void ccid2_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) /* The state about what is acked should be correct now * Check for NUMDUPACK */ - seqp = hctx->ccid2hctx_seqh->ccid2s_prev; + seqp = hctx->ccid2hctx_seqt; + while (before48(seqp->ccid2s_seq, hctx->ccid2hctx_high_ack)) { + seqp = seqp->ccid2s_next; + if (seqp == hctx->ccid2hctx_seqh) { + seqp = hctx->ccid2hctx_seqh->ccid2s_prev; + break; + } + } done = 0; while (1) { if (seqp->ccid2s_acked) { @@ -771,6 +781,7 @@ static int ccid2_hc_tx_init(struct ccid *ccid, struct sock *sk) hctx->ccid2hctx_lastrtt = 0; hctx->ccid2hctx_rpdupack = -1; hctx->ccid2hctx_last_cong = jiffies; + hctx->ccid2hctx_high_ack = 0; hctx->ccid2hctx_rtotimer.function = &ccid2_hc_tx_rto_expire; hctx->ccid2hctx_rtotimer.data = (unsigned long)sk; @@ -823,8 +834,10 @@ static struct ccid_operations ccid2 = { .ccid_hc_rx_packet_recv = ccid2_hc_rx_packet_recv, }; +#ifdef CONFIG_IP_DCCP_CCID2_DEBUG module_param(ccid2_debug, int, 0444); MODULE_PARM_DESC(ccid2_debug, "Enable debug messages"); +#endif static __init int ccid2_module_init(void) { diff --git a/net/dccp/ccids/ccid2.h b/net/dccp/ccids/ccid2.h index 5b2ef4acb30..ebd79499c85 100644 --- a/net/dccp/ccids/ccid2.h +++ b/net/dccp/ccids/ccid2.h @@ -35,7 +35,7 @@ struct ccid2_seq { struct ccid2_seq *ccid2s_next; }; -#define CCID2_SEQBUF_LEN 256 +#define CCID2_SEQBUF_LEN 1024 #define CCID2_SEQBUF_MAX 128 /** struct ccid2_hc_tx_sock - CCID2 TX half connection @@ -72,6 +72,7 @@ struct ccid2_hc_tx_sock { int ccid2hctx_rpdupack; int ccid2hctx_sendwait; unsigned long ccid2hctx_last_cong; + u64 ccid2hctx_high_ack; }; struct ccid2_hc_rx_sock { diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c index cec23ad286d..40402c59506 100644 --- a/net/dccp/ccids/ccid3.c +++ b/net/dccp/ccids/ccid3.c @@ -41,32 +41,9 @@ #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; - -#ifdef CCID3_DEBUG -#define ccid3_pr_debug(format, a...) \ - do { if (ccid3_debug) \ - printk(KERN_DEBUG "%s: " format, __FUNCTION__, ##a); \ - } while (0) +#define ccid3_pr_debug(format, a...) DCCP_PR_DEBUG(ccid3_debug, format, ##a) #else #define ccid3_pr_debug(format, a...) #endif @@ -75,15 +52,7 @@ static struct dccp_tx_hist *ccid3_tx_hist; static struct dccp_rx_hist *ccid3_rx_hist; static struct dccp_li_hist *ccid3_li_hist; -/* TFRC sender states */ -enum ccid3_hc_tx_states { - TFRC_SSTATE_NO_SENT = 1, - TFRC_SSTATE_NO_FBACK, - TFRC_SSTATE_FBACK, - TFRC_SSTATE_TERM, -}; - -#ifdef CCID3_DEBUG +#ifdef CONFIG_IP_DCCP_CCID3_DEBUG static const char *ccid3_tx_state_name(enum ccid3_hc_tx_states state) { static char *ccid3_state_names[] = { @@ -110,324 +79,319 @@ static void ccid3_hc_tx_set_state(struct sock *sk, hctx->ccid3hctx_state = state; } -/* Calculate new t_ipi (inter packet interval) by t_ipi = s / X_inst */ -static inline void ccid3_calc_new_t_ipi(struct ccid3_hc_tx_sock *hctx) +/* + * Recalculate scheduled nominal send time t_nom, inter-packet interval + * t_ipi, and delta value. Should be called after each change to X. + */ +static inline void ccid3_update_send_time(struct ccid3_hc_tx_sock *hctx) { - /* - * If no feedback spec says t_ipi is 1 second (set elsewhere and then - * doubles after every no feedback timer (separate function) - */ - if (hctx->ccid3hctx_state != TFRC_SSTATE_NO_FBACK) - hctx->ccid3hctx_t_ipi = usecs_div(hctx->ccid3hctx_s, - hctx->ccid3hctx_x); -} + timeval_sub_usecs(&hctx->ccid3hctx_t_nom, hctx->ccid3hctx_t_ipi); -/* Calculate new delta by delta = min(t_ipi / 2, t_gran / 2) */ -static inline void ccid3_calc_new_delta(struct ccid3_hc_tx_sock *hctx) -{ + /* 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); + + /* Calculate new delta by delta = min(t_ipi / 2, t_gran / 2) */ hctx->ccid3hctx_delta = min_t(u32, hctx->ccid3hctx_t_ipi / 2, TFRC_OPSYS_HALF_TIME_GRAN); } - /* * Update X by * If (p > 0) - * x_calc = calcX(s, R, p); + * X_calc = calcX(s, R, p); * X = max(min(X_calc, 2 * X_recv), s / t_mbi); * Else * If (now - tld >= R) * X = max(min(2 * X, 2 * X_recv), s / R); * tld = now; - */ -static void ccid3_hc_tx_update_x(struct sock *sk) + * + * 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 __u64 old_x = hctx->ccid3hctx_x; - /* To avoid large error in calcX */ - if (hctx->ccid3hctx_p >= TFRC_SMALLEST_P) { - hctx->ccid3hctx_x_calc = tfrc_calc_x(hctx->ccid3hctx_s, - hctx->ccid3hctx_rtt, - hctx->ccid3hctx_p); - hctx->ccid3hctx_x = max_t(u32, min_t(u32, hctx->ccid3hctx_x_calc, - 2 * hctx->ccid3hctx_x_recv), - (hctx->ccid3hctx_s / - TFRC_MAX_BACK_OFF_TIME)); - } else { - struct timeval now; + if (hctx->ccid3hctx_p > 0) { - dccp_timestamp(sk, &now); - if (timeval_delta(&now, &hctx->ccid3hctx_t_ld) >= - hctx->ccid3hctx_rtt) { - hctx->ccid3hctx_x = max_t(u32, min_t(u32, hctx->ccid3hctx_x_recv, - hctx->ccid3hctx_x) * 2, - usecs_div(hctx->ccid3hctx_s, - hctx->ccid3hctx_rtt)); - hctx->ccid3hctx_t_ld = now; - } + 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; + } + + 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 + */ +static inline void ccid3_hc_tx_update_s(struct ccid3_hc_tx_sock *hctx, int len) +{ + if (unlikely(len == 0)) + ccid3_pr_debug("Packet payload length is 0 - not updating\n"); + else + hctx->ccid3hctx_s = hctx->ccid3hctx_s == 0 ? len : + (9 * hctx->ccid3hctx_s + len) / 10; + /* + * Note: We could do a potential optimisation here - when `s' changes, + * recalculate sending rate and consequently t_ipi, t_delta, and + * t_now. This is however non-standard, and the benefits are not + * clear, so it is currently left out. + */ +} + +/* + * 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; - unsigned long next_tmout = 0; struct ccid3_hc_tx_sock *hctx = ccid3_hc_tx_sk(sk); + unsigned long t_nfb = USEC_PER_SEC / 5; bh_lock_sock(sk); if (sock_owned_by_user(sk)) { /* Try again later. */ /* XXX: set some sensible MIB */ - sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, - jiffies + HZ / 5); - goto out; + 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_TERM: - goto out; case TFRC_SSTATE_NO_FBACK: - /* Halve send rate */ - hctx->ccid3hctx_x /= 2; - if (hctx->ccid3hctx_x < (hctx->ccid3hctx_s / - TFRC_MAX_BACK_OFF_TIME)) - hctx->ccid3hctx_x = (hctx->ccid3hctx_s / - TFRC_MAX_BACK_OFF_TIME); - - ccid3_pr_debug("%s, sk=%p, state=%s, updated tx rate to %d " - "bytes/s\n", - dccp_role(sk), sk, + /* RFC 3448, 4.4: Halve send rate directly */ + hctx->ccid3hctx_x = max(hctx->ccid3hctx_x / 2, + (((__u64)hctx->ccid3hctx_s) << 6) / + TFRC_T_MBI); + + 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); - next_tmout = max_t(u32, 2 * usecs_div(hctx->ccid3hctx_s, - hctx->ccid3hctx_x), - TFRC_INITIAL_TIMEOUT); - /* - * FIXME - not sure above calculation is correct. See section - * 5 of CCID3 11 should adjust tx_t_ipi and double that to - * achieve it really - */ + (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; + ccid3_update_send_time(hctx); break; 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))) { - ccid3_pr_debug("%s, sk=%p, state=%s, not idle\n", + (hctx->ccid3hctx_x_recv >= 4 * + scaled_div(((__u64)hctx->ccid3hctx_s) << 6, + hctx->ccid3hctx_rtt))) { + struct timeval now; + + 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 (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 >= TFRC_SMALLEST_P && - hctx->ccid3hctx_x_calc == 0); - - /* check also if p is zero -> x_calc is infinity? */ - if (hctx->ccid3hctx_p < TFRC_SMALLEST_P || - 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_MAX_BACK_OFF_TIME)); - else - hctx->ccid3hctx_x_recv = hctx->ccid3hctx_x_calc / 4; - - /* Update sending rate */ - ccid3_hc_tx_update_x(sk); + BUG_ON(hctx->ccid3hctx_p && !hctx->ccid3hctx_x_calc); + + if (hctx->ccid3hctx_p == 0 || + (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); } /* * Schedule no feedback timer to expire in - * max(4 * R, 2 * s / X) + * max(t_RTO, 2 * s/X) = max(t_RTO, 2 * t_ipi) + * See comments in packet_recv() regarding the value of t_RTO. */ - next_tmout = max_t(u32, hctx->ccid3hctx_t_rto, - 2 * usecs_div(hctx->ccid3hctx_s, - hctx->ccid3hctx_x)); + t_nfb = max(hctx->ccid3hctx_t_rto, 2 * hctx->ccid3hctx_t_ipi); break; - default: - printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", - __FUNCTION__, dccp_role(sk), sk, hctx->ccid3hctx_state); - dump_stack(); + case TFRC_SSTATE_NO_SENT: + DCCP_BUG("%s(%p) - Illegal state NO_SENT", dccp_role(sk), sk); + /* fall through */ + case TFRC_SSTATE_TERM: goto out; } - sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, - jiffies + max_t(u32, 1, usecs_to_jiffies(next_tmout))); hctx->ccid3hctx_idle = 1; + +restart_timer: + sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, + jiffies + usecs_to_jiffies(t_nfb)); out: bh_unlock_sock(sk); sock_put(sk); } -static int ccid3_hc_tx_send_packet(struct sock *sk, - struct sk_buff *skb, int len) +/* + * returns + * > 0: delay (in msecs) that should pass before actually sending + * = 0: can send immediately + * < 0: error condition; do not send packet + */ +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; - int rc = -ENOTCONN; + suseconds_t delay; - BUG_ON(hctx == NULL || hctx->ccid3hctx_state == TFRC_SSTATE_TERM); + BUG_ON(hctx == NULL); - /* Check if pure ACK or Terminating*/ /* - * XXX: We only call this function for DATA and DATAACK, on, these - * packets can have zero length, but why the comment about "pure ACK"? + * This function is called only for Data and DataAck packets. Sending + * zero-sized Data(Ack)s is theoretically possible, but for congestion + * control this case is pathological - ignore it. */ - if (unlikely(len == 0)) - goto out; - - /* 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, - SLAB_ATOMIC); - - rc = -ENOBUFS; - if (unlikely(new_packet == NULL)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, not enough " - "mem to add to history, send refused\n", - __FUNCTION__, dccp_role(sk), sk); - goto out; - } - - dccp_tx_hist_add_entry(&hctx->ccid3hctx_hist, new_packet); - } + if (unlikely(skb->len == 0)) + return -EBADMSG; 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); - hctx->ccid3hctx_t_ipi = TFRC_INITIAL_IPI; - /* Set nominal send time for initial packet */ + /* 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 <<= 6; + + /* First timeout, according to [RFC 3448, 4.2], is 1 second */ + hctx->ccid3hctx_t_ipi = USEC_PER_SEC; + /* Initial delta: minimum of 0.5 sec and t_gran/2 */ + hctx->ccid3hctx_delta = TFRC_OPSYS_HALF_TIME_GRAN; + + /* Set t_0 for initial packet */ hctx->ccid3hctx_t_nom = now; - timeval_add_usecs(&hctx->ccid3hctx_t_nom, - hctx->ccid3hctx_t_ipi); - ccid3_calc_new_delta(hctx); - rc = 0; break; case TFRC_SSTATE_NO_FBACK: case TFRC_SSTATE_FBACK: - delay = (timeval_delta(&now, &hctx->ccid3hctx_t_nom) - - hctx->ccid3hctx_delta); - delay /= -1000; - /* divide by -1000 is to convert to ms and get sign right */ - rc = delay > 0 ? delay : 0; - break; - default: - printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", - __FUNCTION__, dccp_role(sk), sk, hctx->ccid3hctx_state); - dump_stack(); - rc = -EINVAL; + delay = timeval_delta(&hctx->ccid3hctx_t_nom, &now); + /* + * 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 - (suseconds_t)hctx->ccid3hctx_delta >= 0) + return delay / 1000L; + + ccid3_hc_tx_update_win_count(hctx, &now); break; + case TFRC_SSTATE_TERM: + DCCP_BUG("%s(%p) - Illegal state TERM", dccp_role(sk), sk); + return -EINVAL; } - /* Can we send? if so add options and add to packet history */ - if (rc == 0) { - dp->dccps_hc_tx_insert_options = 1; - new_packet->dccphtx_ccval = - DCCP_SKB_CB(skb)->dccpd_ccval = - hctx->ccid3hctx_last_win_count; - timeval_add_usecs(&hctx->ccid3hctx_t_nom, - hctx->ccid3hctx_t_ipi); - } -out: - return rc; + /* prepare to send now (add options etc.) */ + dp->dccps_hc_tx_insert_options = 1; + 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, 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; + struct dccp_tx_hist_entry *packet; - BUG_ON(hctx == NULL || hctx->ccid3hctx_state == TFRC_SSTATE_TERM); - - dccp_timestamp(sk, &now); - - /* check if we have sent a data packet */ - if (len > 0) { - unsigned long quarter_rtt; - struct dccp_tx_hist_entry *packet; - - packet = dccp_tx_hist_head(&hctx->ccid3hctx_hist); - if (unlikely(packet == NULL)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: packet doesn't " - "exists in history!\n", __FUNCTION__); - return; - } - if (unlikely(packet->dccphtx_sent)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: no unsent packet in " - "history!\n", __FUNCTION__); - 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); - } + BUG_ON(hctx == NULL); - hctx->ccid3hctx_idle = 0; - packet->dccphtx_rtt = hctx->ccid3hctx_rtt; - packet->dccphtx_sent = 1; - } else - ccid3_pr_debug("%s, sk=%p, seqno=%llu NOT inserted!\n", - dccp_role(sk), sk, dp->dccps_gss); + ccid3_hc_tx_update_s(hctx, len); - switch (hctx->ccid3hctx_state) { - case TFRC_SSTATE_NO_SENT: - /* if first wasn't pure ack */ - if (len != 0) - printk(KERN_CRIT "%s: %s, First packet sent is noted " - "as a data packet\n", - __FUNCTION__, dccp_role(sk)); + packet = dccp_tx_hist_entry_new(ccid3_tx_hist, GFP_ATOMIC); + if (unlikely(packet == NULL)) { + DCCP_CRIT("packet history - out of memory!"); return; - case TFRC_SSTATE_NO_FBACK: - case TFRC_SSTATE_FBACK: - if (len > 0) { - timeval_sub_usecs(&hctx->ccid3hctx_t_nom, - hctx->ccid3hctx_t_ipi); - ccid3_calc_new_t_ipi(hctx); - ccid3_calc_new_delta(hctx); - timeval_add_usecs(&hctx->ccid3hctx_t_nom, - hctx->ccid3hctx_t_ipi); - } - break; - default: - printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", - __FUNCTION__, dccp_role(sk), sk, hctx->ccid3hctx_state); - dump_stack(); - break; } + dccp_tx_hist_add_entry(&hctx->ccid3hctx_hist, packet); + + 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) @@ -437,13 +401,11 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) struct ccid3_options_received *opt_recv; struct dccp_tx_hist_entry *packet; struct timeval now; - unsigned long next_tmout; - u32 t_elapsed; + unsigned long t_nfb; u32 pinv; - u32 x_recv; - u32 r_sample; + suseconds_t r_sample, t_elapsed; - BUG_ON(hctx == NULL || hctx->ccid3hctx_state == TFRC_SSTATE_TERM); + BUG_ON(hctx == NULL); /* we are only interested in ACKs */ if (!(DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_ACK || @@ -452,41 +414,49 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) opt_recv = &hctx->ccid3hctx_options_received; - t_elapsed = dp->dccps_options_received.dccpor_elapsed_time * 10; - x_recv = opt_recv->ccid3or_receive_rate; - pinv = opt_recv->ccid3or_loss_event_rate; - switch (hctx->ccid3hctx_state) { - case TFRC_SSTATE_NO_SENT: - /* FIXME: what to do here? */ - return; case TFRC_SSTATE_NO_FBACK: case TFRC_SSTATE_FBACK: - /* Calculate new round trip sample by - * R_sample = (now - t_recvdata) - t_delay */ - /* get t_recvdata from history */ + /* 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)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, seqno " - "%llu(%s) does't exist in history!\n", - __FUNCTION__, dccp_role(sk), sk, + 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)); return; } - /* Update RTT */ + /* 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) /* see RFC 4342, 8.5 */ + hctx->ccid3hctx_p = 0; + else /* can not exceed 100% */ + hctx->ccid3hctx_p = 1000000 / pinv; + dccp_timestamp(sk, &now); - r_sample = timeval_delta(&now, &packet->dccphtx_tstamp); + + /* + * Calculate new round trip sample as per [RFC 3448, 4.3] by + * 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; + + DCCP_BUG_ON(r_sample < 0); if (unlikely(r_sample <= t_elapsed)) - LIMIT_NETDEBUG(KERN_WARNING "%s: r_sample=%uus, " - "t_elapsed=%uus\n", - __FUNCTION__, 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 @@ -495,97 +465,96 @@ 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) { - ccid3_hc_tx_set_state(sk, TFRC_SSTATE_FBACK); - hctx->ccid3hctx_rtt = r_sample; - } else - hctx->ccid3hctx_rtt = (hctx->ccid3hctx_rtt * 9) / 10 + - r_sample / 10; - - ccid3_pr_debug("%s, sk=%p, New RTT estimate=%uus, " - "r_sample=%us\n", dccp_role(sk), sk, - hctx->ccid3hctx_rtt, r_sample); - - /* Update timeout interval */ - hctx->ccid3hctx_t_rto = max_t(u32, 4 * hctx->ccid3hctx_rtt, - USEC_PER_SEC); + /* + * 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; + hctx->ccid3hctx_x = scaled_div(w_init << 6, r_sample); + hctx->ccid3hctx_t_ld = now; - /* Update receive rate */ - hctx->ccid3hctx_x_recv = x_recv;/* X_recv in bytes per sec */ + ccid3_update_send_time(hctx); - /* Update loss event rate */ - if (pinv == ~0 || pinv == 0) - hctx->ccid3hctx_p = 0; - else { - hctx->ccid3hctx_p = 1000000 / pinv; + ccid3_pr_debug("%s(%p), s=%u, w_init=%llu, " + "R_sample=%dus, X=%u\n", dccp_role(sk), + sk, hctx->ccid3hctx_s, w_init, + (int)r_sample, + (unsigned)(hctx->ccid3hctx_x >> 6)); - if (hctx->ccid3hctx_p < TFRC_SMALLEST_P) { - hctx->ccid3hctx_p = TFRC_SMALLEST_P; - ccid3_pr_debug("%s, sk=%p, Smallest p used!\n", - dccp_role(sk), sk); - } + ccid3_hc_tx_set_state(sk, TFRC_SSTATE_FBACK); + } else { + hctx->ccid3hctx_rtt = (9 * hctx->ccid3hctx_rtt + + (u32)r_sample) / 10; + + /* Update sending rate (step 4 of [RFC 3448, 4.3]) */ + if (hctx->ccid3hctx_p > 0) + hctx->ccid3hctx_x_calc = + tfrc_calc_x(hctx->ccid3hctx_s, + hctx->ccid3hctx_rtt, + hctx->ccid3hctx_p); + ccid3_hc_tx_update_x(sk, &now); + + ccid3_pr_debug("%s(%p), RTT=%uus (sample=%dus), s=%u, " + "p=%u, X_calc=%u, X_recv=%u, X=%u\n", + dccp_role(sk), + sk, hctx->ccid3hctx_rtt, (int)r_sample, + hctx->ccid3hctx_s, hctx->ccid3hctx_p, + hctx->ccid3hctx_x_calc, + (unsigned)(hctx->ccid3hctx_x_recv >> 6), + (unsigned)(hctx->ccid3hctx_x >> 6)); } /* unschedule no feedback timer */ sk_stop_timer(sk, &hctx->ccid3hctx_no_feedback_timer); - /* Update sending rate */ - ccid3_hc_tx_update_x(sk); - - /* Update next send time */ - timeval_sub_usecs(&hctx->ccid3hctx_t_nom, - hctx->ccid3hctx_t_ipi); - ccid3_calc_new_t_ipi(hctx); - timeval_add_usecs(&hctx->ccid3hctx_t_nom, - hctx->ccid3hctx_t_ipi); - ccid3_calc_new_delta(hctx); - /* remove all packets older than the one acked from history */ dccp_tx_hist_purge_older(ccid3_tx_hist, &hctx->ccid3hctx_hist, packet); /* - * As we have calculated new ipi, delta, t_nom it is possible that - * we now can send a packet, so wake up dccp_wait_for_ccids. + * As we have calculated new ipi, delta, t_nom it is possible + * that we now can send a packet, so wake up dccp_wait_for_ccid */ sk->sk_write_space(sk); /* + * Update timeout interval for the nofeedback timer. + * We use a configuration option to increase the lower bound. + * This can help avoid triggering the nofeedback timer too + * often ('spinning') on LANs with small RTTs. + */ + hctx->ccid3hctx_t_rto = max_t(u32, 4 * hctx->ccid3hctx_rtt, + CONFIG_IP_DCCP_CCID3_RTO * + (USEC_PER_SEC/1000)); + /* * Schedule no feedback timer to expire in - * max(4 * R, 2 * s / X) + * max(t_RTO, 2 * s/X) = max(t_RTO, 2 * t_ipi) */ - next_tmout = max(hctx->ccid3hctx_t_rto, - 2 * usecs_div(hctx->ccid3hctx_s, - hctx->ccid3hctx_x)); - - ccid3_pr_debug("%s, sk=%p, Scheduled no feedback timer to " + t_nfb = max(hctx->ccid3hctx_t_rto, 2 * hctx->ccid3hctx_t_ipi); + + ccid3_pr_debug("%s(%p), Scheduled no feedback timer to " "expire in %lu jiffies (%luus)\n", - dccp_role(sk), sk, - usecs_to_jiffies(next_tmout), next_tmout); + dccp_role(sk), + sk, usecs_to_jiffies(t_nfb), t_nfb); - sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, - jiffies + max_t(u32, 1, usecs_to_jiffies(next_tmout))); + sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, + jiffies + usecs_to_jiffies(t_nfb)); /* set idle flag */ - hctx->ccid3hctx_idle = 1; + hctx->ccid3hctx_idle = 1; break; - default: - printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", - __FUNCTION__, dccp_role(sk), sk, hctx->ccid3hctx_state); - dump_stack(); + case TFRC_SSTATE_NO_SENT: + /* + * XXX when implementing bidirectional rx/tx check this again + */ + DCCP_WARN("Illegal ACK received - no packet sent\n"); + /* fall through */ + case TFRC_SSTATE_TERM: /* ignore feedback when closing */ break; } } -static int ccid3_hc_tx_insert_options(struct sock *sk, struct sk_buff *skb) -{ - const struct ccid3_hc_tx_sock *hctx = ccid3_hc_tx_sk(sk); - - BUG_ON(hctx == NULL); - - if (sk->sk_state == DCCP_OPEN || sk->sk_state == DCCP_PARTOPEN) - DCCP_SKB_CB(skb)->dccpd_ccval = hctx->ccid3hctx_last_win_count; - return 0; -} - static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option, unsigned char len, u16 idx, unsigned char *value) @@ -610,13 +579,14 @@ static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option, switch (option) { case TFRC_OPT_LOSS_EVENT_RATE: if (unlikely(len != 4)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, invalid " - "len for TFRC_OPT_LOSS_EVENT_RATE\n", - __FUNCTION__, dccp_role(sk), sk); + DCCP_WARN("%s(%p), invalid len %d " + "for TFRC_OPT_LOSS_EVENT_RATE\n", + dccp_role(sk), sk, len); rc = -EINVAL; } else { - opt_recv->ccid3or_loss_event_rate = ntohl(*(__be32 *)value); - ccid3_pr_debug("%s, sk=%p, LOSS_EVENT_RATE=%u\n", + opt_recv->ccid3or_loss_event_rate = + ntohl(*(__be32 *)value); + ccid3_pr_debug("%s(%p), LOSS_EVENT_RATE=%u\n", dccp_role(sk), sk, opt_recv->ccid3or_loss_event_rate); } @@ -624,20 +594,21 @@ static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option, case TFRC_OPT_LOSS_INTERVALS: opt_recv->ccid3or_loss_intervals_idx = idx; opt_recv->ccid3or_loss_intervals_len = len; - ccid3_pr_debug("%s, sk=%p, LOSS_INTERVALS=(%u, %u)\n", + ccid3_pr_debug("%s(%p), LOSS_INTERVALS=(%u, %u)\n", dccp_role(sk), sk, opt_recv->ccid3or_loss_intervals_idx, opt_recv->ccid3or_loss_intervals_len); break; case TFRC_OPT_RECEIVE_RATE: if (unlikely(len != 4)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, invalid " - "len for TFRC_OPT_RECEIVE_RATE\n", - __FUNCTION__, dccp_role(sk), sk); + DCCP_WARN("%s(%p), invalid len %d " + "for TFRC_OPT_RECEIVE_RATE\n", + dccp_role(sk), sk, len); rc = -EINVAL; } else { - opt_recv->ccid3or_receive_rate = ntohl(*(__be32 *)value); - ccid3_pr_debug("%s, sk=%p, RECEIVE_RATE=%u\n", + opt_recv->ccid3or_receive_rate = + ntohl(*(__be32 *)value); + ccid3_pr_debug("%s(%p), RECEIVE_RATE=%u\n", dccp_role(sk), sk, opt_recv->ccid3or_receive_rate); } @@ -649,22 +620,15 @@ static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option, static int ccid3_hc_tx_init(struct ccid *ccid, struct sock *sk) { - struct dccp_sock *dp = dccp_sk(sk); struct ccid3_hc_tx_sock *hctx = ccid_priv(ccid); - if (dp->dccps_packet_size >= TFRC_MIN_PACKET_SIZE && - dp->dccps_packet_size <= TFRC_MAX_PACKET_SIZE) - hctx->ccid3hctx_s = dp->dccps_packet_size; - else - hctx->ccid3hctx_s = TFRC_STD_PACKET_SIZE; - - /* Set transmission rate to 1 packet per second */ - hctx->ccid3hctx_x = hctx->ccid3hctx_s; - hctx->ccid3hctx_t_rto = USEC_PER_SEC; + hctx->ccid3hctx_s = 0; + hctx->ccid3hctx_rtt = 0; hctx->ccid3hctx_state = TFRC_SSTATE_NO_SENT; INIT_LIST_HEAD(&hctx->ccid3hctx_hist); - hctx->ccid3hctx_no_feedback_timer.function = ccid3_hc_tx_no_feedback_timer; + hctx->ccid3hctx_no_feedback_timer.function = + ccid3_hc_tx_no_feedback_timer; hctx->ccid3hctx_no_feedback_timer.data = (unsigned long)sk; init_timer(&hctx->ccid3hctx_no_feedback_timer); @@ -688,14 +652,7 @@ static void ccid3_hc_tx_exit(struct sock *sk) * RX Half Connection methods */ -/* TFRC receiver states */ -enum ccid3_hc_rx_states { - TFRC_RSTATE_NO_DATA = 1, - TFRC_RSTATE_DATA, - TFRC_RSTATE_TERM = 127, -}; - -#ifdef CCID3_DEBUG +#ifdef CONFIG_IP_DCCP_CCID3_DEBUG static const char *ccid3_rx_state_name(enum ccid3_hc_rx_states state) { static char *ccid3_rx_state_names[] = { @@ -721,14 +678,24 @@ static void ccid3_hc_rx_set_state(struct sock *sk, hcrx->ccid3hcrx_state = state; } +static inline void ccid3_hc_rx_update_s(struct ccid3_hc_rx_sock *hcrx, int len) +{ + if (unlikely(len == 0)) /* don't update on empty packets (e.g. ACKs) */ + ccid3_pr_debug("Packet payload length is 0 - not updating\n"); + else + hcrx->ccid3hcrx_s = hcrx->ccid3hcrx_s == 0 ? len : + (9 * hcrx->ccid3hcrx_s + len) / 10; +} + static void ccid3_hc_rx_send_feedback(struct sock *sk) { struct ccid3_hc_rx_sock *hcrx = ccid3_hc_rx_sk(sk); struct dccp_sock *dp = dccp_sk(sk); struct dccp_rx_hist_entry *packet; struct timeval now; + suseconds_t delta; - ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + ccid3_pr_debug("%s(%p) - entry \n", dccp_role(sk), sk); dccp_timestamp(sk, &now); @@ -736,25 +703,22 @@ static void ccid3_hc_rx_send_feedback(struct sock *sk) case TFRC_RSTATE_NO_DATA: hcrx->ccid3hcrx_x_recv = 0; break; - case TFRC_RSTATE_DATA: { - const u32 delta = timeval_delta(&now, - &hcrx->ccid3hcrx_tstamp_last_feedback); - hcrx->ccid3hcrx_x_recv = usecs_div(hcrx->ccid3hcrx_bytes_recv, - delta); - } + case TFRC_RSTATE_DATA: + delta = timeval_delta(&now, + &hcrx->ccid3hcrx_tstamp_last_feedback); + DCCP_BUG_ON(delta < 0); + hcrx->ccid3hcrx_x_recv = + scaled_div32(hcrx->ccid3hcrx_bytes_recv, delta); break; - default: - printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", - __FUNCTION__, dccp_role(sk), sk, hcrx->ccid3hcrx_state); - dump_stack(); + case TFRC_RSTATE_TERM: + DCCP_BUG("%s(%p) - Illegal state TERM", dccp_role(sk), sk); return; } packet = dccp_rx_hist_find_data_packet(&hcrx->ccid3hcrx_hist); if (unlikely(packet == NULL)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, no data packet " - "in history!\n", - __FUNCTION__, dccp_role(sk), sk); + DCCP_WARN("%s(%p), no data packet in history!\n", + dccp_role(sk), sk); return; } @@ -762,13 +726,19 @@ static void ccid3_hc_rx_send_feedback(struct sock *sk) hcrx->ccid3hcrx_ccval_last_counter = packet->dccphrx_ccval; hcrx->ccid3hcrx_bytes_recv = 0; - /* Convert to multiples of 10us */ - hcrx->ccid3hcrx_elapsed_time = - timeval_delta(&now, &packet->dccphrx_tstamp) / 10; + /* Elapsed time information [RFC 4340, 13.2] in units of 10 * usecs */ + delta = timeval_delta(&now, &packet->dccphrx_tstamp); + DCCP_BUG_ON(delta < 0); + hcrx->ccid3hcrx_elapsed_time = delta / 10; + if (hcrx->ccid3hcrx_p == 0) - hcrx->ccid3hcrx_pinv = ~0; - else + hcrx->ccid3hcrx_pinv = ~0U; /* see RFC 4342, 8.5 */ + else if (hcrx->ccid3hcrx_p > 1000000) { + DCCP_WARN("p (%u) > 100%%\n", hcrx->ccid3hcrx_p); + hcrx->ccid3hcrx_pinv = 1; /* use 100% in this case */ + } else hcrx->ccid3hcrx_pinv = 1000000 / hcrx->ccid3hcrx_p; + dp->dccps_hc_rx_insert_options = 1; dccp_send_ack(sk); } @@ -796,9 +766,9 @@ static int ccid3_hc_rx_insert_options(struct sock *sk, struct sk_buff *skb) hcrx->ccid3hcrx_elapsed_time)) || dccp_insert_option_timestamp(sk, skb) || dccp_insert_option(sk, skb, TFRC_OPT_LOSS_EVENT_RATE, - &pinv, sizeof(pinv)) || + &pinv, sizeof(pinv)) || dccp_insert_option(sk, skb, TFRC_OPT_RECEIVE_RATE, - &x_recv, sizeof(x_recv))) + &x_recv, sizeof(x_recv))) return -1; return 0; @@ -812,12 +782,13 @@ static u32 ccid3_hc_rx_calc_first_li(struct sock *sk) { struct ccid3_hc_rx_sock *hcrx = ccid3_hc_rx_sk(sk); struct dccp_rx_hist_entry *entry, *next, *tail = NULL; - u32 rtt, delta, x_recv, fval, p, tmp2; + u32 x_recv, p; + suseconds_t rtt, delta; struct timeval tstamp = { 0, }; int interval = 0; int win_count = 0; int step = 0; - u64 tmp1; + u64 fval; list_for_each_entry_safe(entry, next, &hcrx->ccid3hcrx_hist, dccphrx_node) { @@ -842,58 +813,66 @@ static u32 ccid3_hc_rx_calc_first_li(struct sock *sk) } if (unlikely(step == 0)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, packet history " - "contains no data packets!\n", - __FUNCTION__, dccp_role(sk), sk); + DCCP_WARN("%s(%p), packet history has no data packets!\n", + dccp_role(sk), sk); return ~0; } if (unlikely(interval == 0)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, Could not find a " - "win_count interval > 0. Defaulting to 1\n", - __FUNCTION__, dccp_role(sk), sk); + DCCP_WARN("%s(%p), Could not find a win_count interval > 0." + "Defaulting to 1\n", dccp_role(sk), sk); interval = 1; } found: if (!tail) { - LIMIT_NETDEBUG(KERN_WARNING "%s: tail is null\n", - __FUNCTION__); + DCCP_CRIT("tail is null\n"); return ~0; } - rtt = timeval_delta(&tstamp, &tail->dccphrx_tstamp) * 4 / interval; - ccid3_pr_debug("%s, sk=%p, approximated RTT to %uus\n", - dccp_role(sk), sk, rtt); - if (rtt == 0) - rtt = 1; - - dccp_timestamp(sk, &tstamp); - delta = timeval_delta(&tstamp, &hcrx->ccid3hcrx_tstamp_last_feedback); - x_recv = usecs_div(hcrx->ccid3hcrx_bytes_recv, delta); - if (x_recv == 0) - x_recv = hcrx->ccid3hcrx_x_recv; + delta = timeval_delta(&tstamp, &tail->dccphrx_tstamp); + DCCP_BUG_ON(delta < 0); - tmp1 = (u64)x_recv * (u64)rtt; - do_div(tmp1,10000000); - tmp2 = (u32)tmp1; + rtt = delta * 4 / interval; + ccid3_pr_debug("%s(%p), approximated RTT to %dus\n", + dccp_role(sk), sk, (int)rtt); - if (!tmp2) { - LIMIT_NETDEBUG(KERN_WARNING "tmp2 = 0 " - "%s: x_recv = %u, rtt =%u\n", - __FUNCTION__, x_recv, rtt); + /* + * Determine the length of the first loss interval via inverse lookup. + * Assume that X_recv can be computed by the throughput equation + * s + * X_recv = -------- + * R * fval + * Find some p such that f(p) = fval; return 1/p [RFC 3448, 6.3.1]. + */ + if (rtt == 0) { /* would result in divide-by-zero */ + DCCP_WARN("RTT==0\n"); return ~0; } - fval = (hcrx->ccid3hcrx_s * 100000) / tmp2; - /* do not alter order above or you will get overflow on 32 bit */ + dccp_timestamp(sk, &tstamp); + delta = timeval_delta(&tstamp, &hcrx->ccid3hcrx_tstamp_last_feedback); + DCCP_BUG_ON(delta <= 0); + + x_recv = scaled_div32(hcrx->ccid3hcrx_bytes_recv, delta); + if (x_recv == 0) { /* would also trigger divide-by-zero */ + DCCP_WARN("X_recv==0\n"); + if ((x_recv = hcrx->ccid3hcrx_x_recv) == 0) { + DCCP_BUG("stored value of X_recv is zero"); + return ~0; + } + } + + fval = scaled_div(hcrx->ccid3hcrx_s, rtt); + fval = scaled_div32(fval, x_recv); p = tfrc_calc_x_reverse_lookup(fval); - ccid3_pr_debug("%s, sk=%p, receive rate=%u bytes/s, implied " + + ccid3_pr_debug("%s(%p), receive rate=%u bytes/s, implied " "loss rate=%u\n", dccp_role(sk), sk, x_recv, p); if (p == 0) return ~0; else - return 1000000 / p; + return 1000000 / p; } static void ccid3_hc_rx_update_li(struct sock *sk, u64 seq_loss, u8 win_loss) @@ -923,11 +902,10 @@ static void ccid3_hc_rx_update_li(struct sock *sk, u64 seq_loss, u8 win_loss) /* new loss event detected */ /* calculate last interval length */ seq_temp = dccp_delta_seqno(head->dccplih_seqno, seq_loss); - entry = dccp_li_hist_entry_new(ccid3_li_hist, SLAB_ATOMIC); + entry = dccp_li_hist_entry_new(ccid3_li_hist, GFP_ATOMIC); if (entry == NULL) { - printk(KERN_CRIT "%s: out of memory\n",__FUNCTION__); - dump_stack(); + DCCP_BUG("out of memory - can not allocate entry"); return; } @@ -948,7 +926,8 @@ static int ccid3_hc_rx_detect_loss(struct sock *sk, struct dccp_rx_hist_entry *packet) { struct ccid3_hc_rx_sock *hcrx = ccid3_hc_rx_sk(sk); - struct dccp_rx_hist_entry *rx_hist = dccp_rx_hist_head(&hcrx->ccid3hcrx_hist); + struct dccp_rx_hist_entry *rx_hist = + dccp_rx_hist_head(&hcrx->ccid3hcrx_hist); u64 seqno = packet->dccphrx_seqno; u64 tmp_seqno; int loss = 0; @@ -976,7 +955,7 @@ static int ccid3_hc_rx_detect_loss(struct sock *sk, dccp_inc_seqno(&tmp_seqno); while (dccp_rx_hist_find_entry(&hcrx->ccid3hcrx_hist, tmp_seqno, &ccval)) { - hcrx->ccid3hcrx_seqno_nonloss = tmp_seqno; + hcrx->ccid3hcrx_seqno_nonloss = tmp_seqno; hcrx->ccid3hcrx_ccval_nonloss = ccval; dccp_inc_seqno(&tmp_seqno); } @@ -1002,13 +981,11 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) const struct dccp_options_received *opt_recv; struct dccp_rx_hist_entry *packet; struct timeval now; - u8 win_count; - u32 p_prev, rtt_prev, r_sample, t_elapsed; - int loss; + u32 p_prev, rtt_prev; + suseconds_t r_sample, t_elapsed; + int loss, payload_size; - BUG_ON(hcrx == NULL || - !(hcrx->ccid3hcrx_state == TFRC_RSTATE_NO_DATA || - hcrx->ccid3hcrx_state == TFRC_RSTATE_DATA)); + BUG_ON(hcrx == NULL); opt_recv = &dccp_sk(sk)->dccps_options_received; @@ -1025,12 +1002,13 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) r_sample = timeval_usecs(&now); t_elapsed = opt_recv->dccpor_elapsed_time * 10; + DCCP_BUG_ON(r_sample < 0); if (unlikely(r_sample <= t_elapsed)) - LIMIT_NETDEBUG(KERN_WARNING "%s: r_sample=%uus, " - "t_elapsed=%uus\n", - __FUNCTION__, r_sample, t_elapsed); + DCCP_WARN("r_sample=%ldus, t_elapsed=%ldus\n", + r_sample, t_elapsed); else r_sample -= t_elapsed; + CCID3_RTT_SANITY_CHECK(r_sample); if (hcrx->ccid3hcrx_state == TFRC_RSTATE_NO_DATA) hcrx->ccid3hcrx_rtt = r_sample; @@ -1039,8 +1017,8 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) r_sample / 10; if (rtt_prev != hcrx->ccid3hcrx_rtt) - ccid3_pr_debug("%s, New RTT=%uus, elapsed time=%u\n", - dccp_role(sk), hcrx->ccid3hcrx_rtt, + ccid3_pr_debug("%s(%p), New RTT=%uus, elapsed time=%u\n", + dccp_role(sk), sk, hcrx->ccid3hcrx_rtt, opt_recv->dccpor_elapsed_time); break; case DCCP_PKT_DATA: @@ -1050,52 +1028,48 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) } packet = dccp_rx_hist_entry_new(ccid3_rx_hist, sk, opt_recv->dccpor_ndp, - skb, SLAB_ATOMIC); + skb, GFP_ATOMIC); if (unlikely(packet == NULL)) { - LIMIT_NETDEBUG(KERN_WARNING "%s: %s, sk=%p, Not enough mem to " - "add rx packet to history, consider it lost!\n", - __FUNCTION__, dccp_role(sk), sk); + DCCP_WARN("%s(%p), Not enough mem to add rx packet " + "to history, consider it lost!\n", dccp_role(sk), sk); return; } - win_count = packet->dccphrx_ccval; - loss = ccid3_hc_rx_detect_loss(sk, packet); if (DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_ACK) return; + payload_size = skb->len - dccp_hdr(skb)->dccph_doff * 4; + ccid3_hc_rx_update_s(hcrx, payload_size); + switch (hcrx->ccid3hcrx_state) { case TFRC_RSTATE_NO_DATA: - ccid3_pr_debug("%s, sk=%p(%s), skb=%p, sending initial " - "feedback\n", - dccp_role(sk), sk, + ccid3_pr_debug("%s(%p, state=%s), skb=%p, sending initial " + "feedback\n", dccp_role(sk), sk, dccp_state_name(sk->sk_state), skb); ccid3_hc_rx_send_feedback(sk); ccid3_hc_rx_set_state(sk, TFRC_RSTATE_DATA); return; case TFRC_RSTATE_DATA: - hcrx->ccid3hcrx_bytes_recv += skb->len - - dccp_hdr(skb)->dccph_doff * 4; + hcrx->ccid3hcrx_bytes_recv += payload_size; if (loss) break; dccp_timestamp(sk, &now); - if (timeval_delta(&now, &hcrx->ccid3hcrx_tstamp_last_ack) >= - hcrx->ccid3hcrx_rtt) { + if ((timeval_delta(&now, &hcrx->ccid3hcrx_tstamp_last_ack) - + (suseconds_t)hcrx->ccid3hcrx_rtt) >= 0) { hcrx->ccid3hcrx_tstamp_last_ack = now; ccid3_hc_rx_send_feedback(sk); } return; - default: - printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", - __FUNCTION__, dccp_role(sk), sk, hcrx->ccid3hcrx_state); - dump_stack(); + case TFRC_RSTATE_TERM: + DCCP_BUG("%s(%p) - Illegal state TERM", dccp_role(sk), sk); return; } /* Dealing with packet loss */ - ccid3_pr_debug("%s, sk=%p(%s), data loss! Reacting...\n", + ccid3_pr_debug("%s(%p, state=%s), data loss! Reacting...\n", dccp_role(sk), sk, dccp_state_name(sk->sk_state)); p_prev = hcrx->ccid3hcrx_p; @@ -1107,10 +1081,8 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) /* Scaling up by 1000000 as fixed decimal */ if (i_mean != 0) hcrx->ccid3hcrx_p = 1000000 / i_mean; - } else { - printk(KERN_CRIT "%s: empty loss hist\n",__FUNCTION__); - dump_stack(); - } + } else + DCCP_BUG("empty loss history"); if (hcrx->ccid3hcrx_p > p_prev) { ccid3_hc_rx_send_feedback(sk); @@ -1120,23 +1092,17 @@ static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) static int ccid3_hc_rx_init(struct ccid *ccid, struct sock *sk) { - struct dccp_sock *dp = dccp_sk(sk); struct ccid3_hc_rx_sock *hcrx = ccid_priv(ccid); - ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); - - if (dp->dccps_packet_size >= TFRC_MIN_PACKET_SIZE && - dp->dccps_packet_size <= TFRC_MAX_PACKET_SIZE) - hcrx->ccid3hcrx_s = dp->dccps_packet_size; - else - hcrx->ccid3hcrx_s = TFRC_STD_PACKET_SIZE; + ccid3_pr_debug("entry\n"); hcrx->ccid3hcrx_state = TFRC_RSTATE_NO_DATA; INIT_LIST_HEAD(&hcrx->ccid3hcrx_hist); INIT_LIST_HEAD(&hcrx->ccid3hcrx_li_hist); dccp_timestamp(sk, &hcrx->ccid3hcrx_tstamp_last_ack); hcrx->ccid3hcrx_tstamp_last_feedback = hcrx->ccid3hcrx_tstamp_last_ack; - hcrx->ccid3hcrx_rtt = 5000; /* XXX 5ms for now... */ + hcrx->ccid3hcrx_s = 0; + hcrx->ccid3hcrx_rtt = 0; return 0; } @@ -1165,9 +1131,9 @@ static void ccid3_hc_rx_get_info(struct sock *sk, struct tcp_info *info) BUG_ON(hcrx == NULL); - info->tcpi_ca_state = hcrx->ccid3hcrx_state; - info->tcpi_options |= TCPI_OPT_TIMESTAMPS; - info->tcpi_rcv_rtt = hcrx->ccid3hcrx_rtt; + info->tcpi_ca_state = hcrx->ccid3hcrx_state; + info->tcpi_options |= TCPI_OPT_TIMESTAMPS; + info->tcpi_rcv_rtt = hcrx->ccid3hcrx_rtt; } static void ccid3_hc_tx_get_info(struct sock *sk, struct tcp_info *info) @@ -1248,7 +1214,6 @@ static struct ccid_operations ccid3 = { .ccid_hc_tx_send_packet = ccid3_hc_tx_send_packet, .ccid_hc_tx_packet_sent = ccid3_hc_tx_packet_sent, .ccid_hc_tx_packet_recv = ccid3_hc_tx_packet_recv, - .ccid_hc_tx_insert_options = ccid3_hc_tx_insert_options, .ccid_hc_tx_parse_options = ccid3_hc_tx_parse_options, .ccid_hc_rx_obj_size = sizeof(struct ccid3_hc_rx_sock), .ccid_hc_rx_init = ccid3_hc_rx_init, @@ -1260,9 +1225,11 @@ static struct ccid_operations ccid3 = { .ccid_hc_rx_getsockopt = ccid3_hc_rx_getsockopt, .ccid_hc_tx_getsockopt = ccid3_hc_tx_getsockopt, }; - + +#ifdef CONFIG_IP_DCCP_CCID3_DEBUG module_param(ccid3_debug, int, 0444); MODULE_PARM_DESC(ccid3_debug, "Enable debug messages"); +#endif static __init int ccid3_module_init(void) { @@ -1281,7 +1248,7 @@ static __init int ccid3_module_init(void) goto out_free_tx; rc = ccid_register(&ccid3); - if (rc != 0) + if (rc != 0) goto out_free_loss_interval_history; out: return rc; diff --git a/net/dccp/ccids/ccid3.h b/net/dccp/ccids/ccid3.h index 0a2cb7536d2..15776a88c09 100644 --- a/net/dccp/ccids/ccid3.h +++ b/net/dccp/ccids/ccid3.h @@ -42,22 +42,24 @@ #include <linux/tfrc.h> #include "../ccid.h" -#define TFRC_MIN_PACKET_SIZE 16 -#define TFRC_STD_PACKET_SIZE 256 -#define TFRC_MAX_PACKET_SIZE 65535 - -/* Two seconds as per CCID3 spec */ +/* Two seconds as per RFC 3448 4.2 */ #define TFRC_INITIAL_TIMEOUT (2 * USEC_PER_SEC) -#define TFRC_INITIAL_IPI (USEC_PER_SEC / 4) - /* In usecs - half the scheduling granularity as per RFC3448 4.6 */ #define TFRC_OPSYS_HALF_TIME_GRAN (USEC_PER_SEC / (2 * HZ)) -/* In seconds */ -#define TFRC_MAX_BACK_OFF_TIME 64 +/* Parameter t_mbi from [RFC 3448, 4.3]: backoff interval in seconds */ +#define TFRC_T_MBI 64 + +/* What we think is a reasonable upper limit on RTT values */ +#define CCID3_SANE_RTT_MAX ((suseconds_t)(4 * USEC_PER_SEC)) -#define TFRC_SMALLEST_P 40 +#define CCID3_RTT_SANITY_CHECK(rtt) do { \ + if (rtt > CCID3_SANE_RTT_MAX) { \ + DCCP_CRIT("RTT (%d) too large, substituting %d", \ + (int)rtt, (int)CCID3_SANE_RTT_MAX); \ + rtt = CCID3_SANE_RTT_MAX; \ + } } while (0) enum ccid3_options { TFRC_OPT_LOSS_EVENT_RATE = 192, @@ -73,26 +75,36 @@ struct ccid3_options_received { u32 ccid3or_receive_rate; }; -/** struct ccid3_hc_tx_sock - CCID3 sender half connection sock +/* TFRC sender states */ +enum ccid3_hc_tx_states { + TFRC_SSTATE_NO_SENT = 1, + TFRC_SSTATE_NO_FBACK, + TFRC_SSTATE_FBACK, + TFRC_SSTATE_TERM, +}; + +/** struct ccid3_hc_tx_sock - CCID3 sender half-connection socket * - * @ccid3hctx_state - Sender state - * @ccid3hctx_x - Current sending rate - * @ccid3hctx_x_recv - Receive rate - * @ccid3hctx_x_calc - Calculated send (?) rate - * @ccid3hctx_s - Packet size - * @ccid3hctx_rtt - Estimate of current round trip time in usecs - * @@ccid3hctx_p - Current loss event rate (0-1) scaled by 1000000 - * @ccid3hctx_last_win_count - Last window counter sent - * @ccid3hctx_t_last_win_count - Timestamp of earliest packet - * with last_win_count value sent - * @ccid3hctx_no_feedback_timer - Handle to no feedback timer - * @ccid3hctx_idle - FIXME - * @ccid3hctx_t_ld - Time last doubled during slow start - * @ccid3hctx_t_nom - Nominal send time of next packet - * @ccid3hctx_t_ipi - Interpacket (send) interval - * @ccid3hctx_delta - Send timer delta - * @ccid3hctx_hist - Packet history - */ + * @ccid3hctx_x - Current sending rate in 64 * bytes per second + * @ccid3hctx_x_recv - Receive rate in 64 * bytes per second + * @ccid3hctx_x_calc - Calculated rate in bytes per second + * @ccid3hctx_rtt - Estimate of current round trip time in usecs + * @ccid3hctx_p - Current loss event rate (0-1) scaled by 1000000 + * @ccid3hctx_s - Packet size in bytes + * @ccid3hctx_t_rto - Nofeedback Timer setting in usecs + * @ccid3hctx_t_ipi - Interpacket (send) interval (RFC 3448, 4.6) in usecs + * @ccid3hctx_state - Sender state, one of %ccid3_hc_tx_states + * @ccid3hctx_last_win_count - Last window counter sent + * @ccid3hctx_t_last_win_count - Timestamp of earliest packet + * with last_win_count value sent + * @ccid3hctx_no_feedback_timer - Handle to no feedback timer + * @ccid3hctx_idle - Flag indicating that sender is idling + * @ccid3hctx_t_ld - Time last doubled during slow start + * @ccid3hctx_t_nom - Nominal send time of next packet + * @ccid3hctx_delta - Send timer delta (RFC 3448, 4.6) in usecs + * @ccid3hctx_hist - Packet history + * @ccid3hctx_options_received - Parsed set of retrieved options + */ struct ccid3_hc_tx_sock { struct tfrc_tx_info ccid3hctx_tfrc; #define ccid3hctx_x ccid3hctx_tfrc.tfrctx_x @@ -103,7 +115,7 @@ struct ccid3_hc_tx_sock { #define ccid3hctx_t_rto ccid3hctx_tfrc.tfrctx_rto #define ccid3hctx_t_ipi ccid3hctx_tfrc.tfrctx_ipi u16 ccid3hctx_s; - u8 ccid3hctx_state; + enum ccid3_hc_tx_states ccid3hctx_state:8; u8 ccid3hctx_last_win_count; u8 ccid3hctx_idle; struct timeval ccid3hctx_t_last_win_count; @@ -115,23 +127,48 @@ struct ccid3_hc_tx_sock { struct ccid3_options_received ccid3hctx_options_received; }; +/* TFRC receiver states */ +enum ccid3_hc_rx_states { + TFRC_RSTATE_NO_DATA = 1, + TFRC_RSTATE_DATA, + TFRC_RSTATE_TERM = 127, +}; + +/** struct ccid3_hc_rx_sock - CCID3 receiver half-connection socket + * + * @ccid3hcrx_x_recv - Receiver estimate of send rate (RFC 3448 4.3) + * @ccid3hcrx_rtt - Receiver estimate of rtt (non-standard) + * @ccid3hcrx_p - current loss event rate (RFC 3448 5.4) + * @ccid3hcrx_seqno_nonloss - Last received non-loss sequence number + * @ccid3hcrx_ccval_nonloss - Last received non-loss Window CCVal + * @ccid3hcrx_ccval_last_counter - Tracks window counter (RFC 4342, 8.1) + * @ccid3hcrx_state - receiver state, one of %ccid3_hc_rx_states + * @ccid3hcrx_bytes_recv - Total sum of DCCP payload bytes + * @ccid3hcrx_tstamp_last_feedback - Time at which last feedback was sent + * @ccid3hcrx_tstamp_last_ack - Time at which last feedback was sent + * @ccid3hcrx_hist - Packet history + * @ccid3hcrx_li_hist - Loss Interval History + * @ccid3hcrx_s - Received packet size in bytes + * @ccid3hcrx_pinv - Inverse of Loss Event Rate (RFC 4342, sec. 8.5) + * @ccid3hcrx_elapsed_time - Time since packet reception + */ struct ccid3_hc_rx_sock { - struct tfrc_rx_info ccid3hcrx_tfrc; -#define ccid3hcrx_x_recv ccid3hcrx_tfrc.tfrcrx_x_recv -#define ccid3hcrx_rtt ccid3hcrx_tfrc.tfrcrx_rtt -#define ccid3hcrx_p ccid3hcrx_tfrc.tfrcrx_p - u64 ccid3hcrx_seqno_nonloss:48, - ccid3hcrx_ccval_nonloss:4, - ccid3hcrx_state:8, - ccid3hcrx_ccval_last_counter:4; - u32 ccid3hcrx_bytes_recv; - struct timeval ccid3hcrx_tstamp_last_feedback; - struct timeval ccid3hcrx_tstamp_last_ack; - struct list_head ccid3hcrx_hist; - struct list_head ccid3hcrx_li_hist; - u16 ccid3hcrx_s; - u32 ccid3hcrx_pinv; - u32 ccid3hcrx_elapsed_time; + struct tfrc_rx_info ccid3hcrx_tfrc; +#define ccid3hcrx_x_recv ccid3hcrx_tfrc.tfrcrx_x_recv +#define ccid3hcrx_rtt ccid3hcrx_tfrc.tfrcrx_rtt +#define ccid3hcrx_p ccid3hcrx_tfrc.tfrcrx_p + u64 ccid3hcrx_seqno_nonloss:48, + ccid3hcrx_ccval_nonloss:4, + ccid3hcrx_ccval_last_counter:4; + enum ccid3_hc_rx_states ccid3hcrx_state:8; + u32 ccid3hcrx_bytes_recv; + struct timeval ccid3hcrx_tstamp_last_feedback; + struct timeval ccid3hcrx_tstamp_last_ack; + struct list_head ccid3hcrx_hist; + struct list_head ccid3hcrx_li_hist; + u16 ccid3hcrx_s; + u32 ccid3hcrx_pinv; + u32 ccid3hcrx_elapsed_time; }; static inline struct ccid3_hc_tx_sock *ccid3_hc_tx_sk(const struct sock *sk) diff --git a/net/dccp/ccids/lib/loss_interval.c b/net/dccp/ccids/lib/loss_interval.c index 906c81ab9d4..0a0baef16b3 100644 --- a/net/dccp/ccids/lib/loss_interval.c +++ b/net/dccp/ccids/lib/loss_interval.c @@ -13,7 +13,7 @@ #include <linux/module.h> #include <net/sock.h> - +#include "../../dccp.h" #include "loss_interval.h" struct dccp_li_hist *dccp_li_hist_new(const char *name) @@ -109,7 +109,7 @@ u32 dccp_li_hist_calc_i_mean(struct list_head *list) i_tot = max(i_tot0, i_tot1); if (!w_tot) { - LIMIT_NETDEBUG(KERN_WARNING "%s: w_tot = 0\n", __FUNCTION__); + DCCP_WARN("w_tot = 0\n"); return 1; } @@ -125,10 +125,10 @@ int dccp_li_hist_interval_new(struct dccp_li_hist *hist, int i; for (i = 0; i < DCCP_LI_HIST_IVAL_F_LENGTH; i++) { - entry = dccp_li_hist_entry_new(hist, SLAB_ATOMIC); + entry = dccp_li_hist_entry_new(hist, GFP_ATOMIC); if (entry == NULL) { dccp_li_hist_purge(hist, list); - dump_stack(); + DCCP_BUG("loss interval list entry is NULL"); return 0; } entry->dccplih_interval = ~0; diff --git a/net/dccp/ccids/lib/loss_interval.h b/net/dccp/ccids/lib/loss_interval.h index 0ae85f0340b..eb257014dd7 100644 --- a/net/dccp/ccids/lib/loss_interval.h +++ b/net/dccp/ccids/lib/loss_interval.h @@ -20,7 +20,7 @@ #define DCCP_LI_HIST_IVAL_F_LENGTH 8 struct dccp_li_hist { - kmem_cache_t *dccplih_slab; + struct kmem_cache *dccplih_slab; }; extern struct dccp_li_hist *dccp_li_hist_new(const char *name); diff --git a/net/dccp/ccids/lib/packet_history.c b/net/dccp/ccids/lib/packet_history.c index b876c9c81c6..2e8ef42721e 100644 --- a/net/dccp/ccids/lib/packet_history.c +++ b/net/dccp/ccids/lib/packet_history.c @@ -36,9 +36,100 @@ #include <linux/module.h> #include <linux/string.h> - #include "packet_history.h" +/* + * Transmitter History Routines + */ +struct dccp_tx_hist *dccp_tx_hist_new(const char *name) +{ + struct dccp_tx_hist *hist = kmalloc(sizeof(*hist), GFP_ATOMIC); + static const char dccp_tx_hist_mask[] = "tx_hist_%s"; + char *slab_name; + + if (hist == NULL) + goto out; + + slab_name = kmalloc(strlen(name) + sizeof(dccp_tx_hist_mask) - 1, + GFP_ATOMIC); + if (slab_name == NULL) + goto out_free_hist; + + sprintf(slab_name, dccp_tx_hist_mask, name); + hist->dccptxh_slab = kmem_cache_create(slab_name, + sizeof(struct dccp_tx_hist_entry), + 0, SLAB_HWCACHE_ALIGN, + NULL, NULL); + if (hist->dccptxh_slab == NULL) + goto out_free_slab_name; +out: + return hist; +out_free_slab_name: + kfree(slab_name); +out_free_hist: + kfree(hist); + hist = NULL; + goto out; +} + +EXPORT_SYMBOL_GPL(dccp_tx_hist_new); + +void dccp_tx_hist_delete(struct dccp_tx_hist *hist) +{ + const char* name = kmem_cache_name(hist->dccptxh_slab); + + kmem_cache_destroy(hist->dccptxh_slab); + kfree(name); + kfree(hist); +} + +EXPORT_SYMBOL_GPL(dccp_tx_hist_delete); + +struct dccp_tx_hist_entry * + dccp_tx_hist_find_entry(const struct list_head *list, const u64 seq) +{ + struct dccp_tx_hist_entry *packet = NULL, *entry; + + list_for_each_entry(entry, list, dccphtx_node) + if (entry->dccphtx_seqno == seq) { + packet = entry; + break; + } + + return packet; +} + +EXPORT_SYMBOL_GPL(dccp_tx_hist_find_entry); + +void dccp_tx_hist_purge(struct dccp_tx_hist *hist, struct list_head *list) +{ + struct dccp_tx_hist_entry *entry, *next; + + list_for_each_entry_safe(entry, next, list, dccphtx_node) { + list_del_init(&entry->dccphtx_node); + dccp_tx_hist_entry_delete(hist, entry); + } +} + +EXPORT_SYMBOL_GPL(dccp_tx_hist_purge); + +void dccp_tx_hist_purge_older(struct dccp_tx_hist *hist, + struct list_head *list, + struct dccp_tx_hist_entry *packet) +{ + struct dccp_tx_hist_entry *next; + + list_for_each_entry_safe_continue(packet, next, list, dccphtx_node) { + list_del_init(&packet->dccphtx_node); + dccp_tx_hist_entry_delete(hist, packet); + } +} + +EXPORT_SYMBOL_GPL(dccp_tx_hist_purge_older); + +/* + * Receiver History Routines + */ struct dccp_rx_hist *dccp_rx_hist_new(const char *name) { struct dccp_rx_hist *hist = kmalloc(sizeof(*hist), GFP_ATOMIC); @@ -83,18 +174,24 @@ void dccp_rx_hist_delete(struct dccp_rx_hist *hist) EXPORT_SYMBOL_GPL(dccp_rx_hist_delete); -void dccp_rx_hist_purge(struct dccp_rx_hist *hist, struct list_head *list) +int dccp_rx_hist_find_entry(const struct list_head *list, const u64 seq, + u8 *ccval) { - struct dccp_rx_hist_entry *entry, *next; + struct dccp_rx_hist_entry *packet = NULL, *entry; - list_for_each_entry_safe(entry, next, list, dccphrx_node) { - list_del_init(&entry->dccphrx_node); - kmem_cache_free(hist->dccprxh_slab, entry); - } -} + list_for_each_entry(entry, list, dccphrx_node) + if (entry->dccphrx_seqno == seq) { + packet = entry; + break; + } -EXPORT_SYMBOL_GPL(dccp_rx_hist_purge); + if (packet) + *ccval = packet->dccphrx_ccval; + return packet != NULL; +} + +EXPORT_SYMBOL_GPL(dccp_rx_hist_find_entry); struct dccp_rx_hist_entry * dccp_rx_hist_find_data_packet(const struct list_head *list) { @@ -184,110 +281,18 @@ void dccp_rx_hist_add_packet(struct dccp_rx_hist *hist, EXPORT_SYMBOL_GPL(dccp_rx_hist_add_packet); -struct dccp_tx_hist *dccp_tx_hist_new(const char *name) -{ - struct dccp_tx_hist *hist = kmalloc(sizeof(*hist), GFP_ATOMIC); - static const char dccp_tx_hist_mask[] = "tx_hist_%s"; - char *slab_name; - - if (hist == NULL) - goto out; - - slab_name = kmalloc(strlen(name) + sizeof(dccp_tx_hist_mask) - 1, - GFP_ATOMIC); - if (slab_name == NULL) - goto out_free_hist; - - sprintf(slab_name, dccp_tx_hist_mask, name); - hist->dccptxh_slab = kmem_cache_create(slab_name, - sizeof(struct dccp_tx_hist_entry), - 0, SLAB_HWCACHE_ALIGN, - NULL, NULL); - if (hist->dccptxh_slab == NULL) - goto out_free_slab_name; -out: - return hist; -out_free_slab_name: - kfree(slab_name); -out_free_hist: - kfree(hist); - hist = NULL; - goto out; -} - -EXPORT_SYMBOL_GPL(dccp_tx_hist_new); - -void dccp_tx_hist_delete(struct dccp_tx_hist *hist) -{ - const char* name = kmem_cache_name(hist->dccptxh_slab); - - kmem_cache_destroy(hist->dccptxh_slab); - kfree(name); - kfree(hist); -} - -EXPORT_SYMBOL_GPL(dccp_tx_hist_delete); - -struct dccp_tx_hist_entry * - dccp_tx_hist_find_entry(const struct list_head *list, const u64 seq) -{ - struct dccp_tx_hist_entry *packet = NULL, *entry; - - list_for_each_entry(entry, list, dccphtx_node) - if (entry->dccphtx_seqno == seq) { - packet = entry; - break; - } - - return packet; -} - -EXPORT_SYMBOL_GPL(dccp_tx_hist_find_entry); - -int dccp_rx_hist_find_entry(const struct list_head *list, const u64 seq, - u8 *ccval) -{ - struct dccp_rx_hist_entry *packet = NULL, *entry; - - list_for_each_entry(entry, list, dccphrx_node) - if (entry->dccphrx_seqno == seq) { - packet = entry; - break; - } - - if (packet) - *ccval = packet->dccphrx_ccval; - - return packet != NULL; -} - -EXPORT_SYMBOL_GPL(dccp_rx_hist_find_entry); - -void dccp_tx_hist_purge_older(struct dccp_tx_hist *hist, - struct list_head *list, - struct dccp_tx_hist_entry *packet) +void dccp_rx_hist_purge(struct dccp_rx_hist *hist, struct list_head *list) { - struct dccp_tx_hist_entry *next; + struct dccp_rx_hist_entry *entry, *next; - list_for_each_entry_safe_continue(packet, next, list, dccphtx_node) { - list_del_init(&packet->dccphtx_node); - dccp_tx_hist_entry_delete(hist, packet); + list_for_each_entry_safe(entry, next, list, dccphrx_node) { + list_del_init(&entry->dccphrx_node); + kmem_cache_free(hist->dccprxh_slab, entry); } } -EXPORT_SYMBOL_GPL(dccp_tx_hist_purge_older); - -void dccp_tx_hist_purge(struct dccp_tx_hist *hist, struct list_head *list) -{ - struct dccp_tx_hist_entry *entry, *next; - - list_for_each_entry_safe(entry, next, list, dccphtx_node) { - list_del_init(&entry->dccphtx_node); - dccp_tx_hist_entry_delete(hist, entry); - } -} +EXPORT_SYMBOL_GPL(dccp_rx_hist_purge); -EXPORT_SYMBOL_GPL(dccp_tx_hist_purge); MODULE_AUTHOR("Ian McDonald <ian.mcdonald@jandi.co.nz>, " "Arnaldo Carvalho de Melo <acme@ghostprotocols.net>"); diff --git a/net/dccp/ccids/lib/packet_history.h b/net/dccp/ccids/lib/packet_history.h index 067cf1c85a3..1f960c19ea1 100644 --- a/net/dccp/ccids/lib/packet_history.h +++ b/net/dccp/ccids/lib/packet_history.h @@ -49,43 +49,27 @@ #define TFRC_WIN_COUNT_PER_RTT 4 #define TFRC_WIN_COUNT_LIMIT 16 +/* + * Transmitter History data structures and declarations + */ struct dccp_tx_hist_entry { struct list_head dccphtx_node; u64 dccphtx_seqno:48, - dccphtx_ccval:4, dccphtx_sent:1; u32 dccphtx_rtt; struct timeval dccphtx_tstamp; }; -struct dccp_rx_hist_entry { - struct list_head dccphrx_node; - u64 dccphrx_seqno:48, - dccphrx_ccval:4, - dccphrx_type:4; - u32 dccphrx_ndp; /* In fact it is from 8 to 24 bits */ - struct timeval dccphrx_tstamp; -}; - struct dccp_tx_hist { - kmem_cache_t *dccptxh_slab; + struct kmem_cache *dccptxh_slab; }; extern struct dccp_tx_hist *dccp_tx_hist_new(const char *name); -extern void dccp_tx_hist_delete(struct dccp_tx_hist *hist); - -struct dccp_rx_hist { - kmem_cache_t *dccprxh_slab; -}; - -extern struct dccp_rx_hist *dccp_rx_hist_new(const char *name); -extern void dccp_rx_hist_delete(struct dccp_rx_hist *hist); -extern struct dccp_rx_hist_entry * - dccp_rx_hist_find_data_packet(const struct list_head *list); +extern void dccp_tx_hist_delete(struct dccp_tx_hist *hist); static inline struct dccp_tx_hist_entry * - dccp_tx_hist_entry_new(struct dccp_tx_hist *hist, - const gfp_t prio) + dccp_tx_hist_entry_new(struct dccp_tx_hist *hist, + const gfp_t prio) { struct dccp_tx_hist_entry *entry = kmem_cache_alloc(hist->dccptxh_slab, prio); @@ -96,18 +80,20 @@ static inline struct dccp_tx_hist_entry * return entry; } -static inline void dccp_tx_hist_entry_delete(struct dccp_tx_hist *hist, - struct dccp_tx_hist_entry *entry) +static inline struct dccp_tx_hist_entry * + dccp_tx_hist_head(struct list_head *list) { - if (entry != NULL) - kmem_cache_free(hist->dccptxh_slab, entry); + struct dccp_tx_hist_entry *head = NULL; + + if (!list_empty(list)) + head = list_entry(list->next, struct dccp_tx_hist_entry, + dccphtx_node); + return head; } extern struct dccp_tx_hist_entry * dccp_tx_hist_find_entry(const struct list_head *list, const u64 seq); -extern int dccp_rx_hist_find_entry(const struct list_head *list, const u64 seq, - u8 *ccval); static inline void dccp_tx_hist_add_entry(struct list_head *list, struct dccp_tx_hist_entry *entry) @@ -115,30 +101,45 @@ static inline void dccp_tx_hist_add_entry(struct list_head *list, list_add(&entry->dccphtx_node, list); } +static inline void dccp_tx_hist_entry_delete(struct dccp_tx_hist *hist, + struct dccp_tx_hist_entry *entry) +{ + if (entry != NULL) + kmem_cache_free(hist->dccptxh_slab, entry); +} + +extern void dccp_tx_hist_purge(struct dccp_tx_hist *hist, + struct list_head *list); + extern void dccp_tx_hist_purge_older(struct dccp_tx_hist *hist, struct list_head *list, struct dccp_tx_hist_entry *next); -extern void dccp_tx_hist_purge(struct dccp_tx_hist *hist, - struct list_head *list); +/* + * Receiver History data structures and declarations + */ +struct dccp_rx_hist_entry { + struct list_head dccphrx_node; + u64 dccphrx_seqno:48, + dccphrx_ccval:4, + dccphrx_type:4; + u32 dccphrx_ndp; /* In fact it is from 8 to 24 bits */ + struct timeval dccphrx_tstamp; +}; -static inline struct dccp_tx_hist_entry * - dccp_tx_hist_head(struct list_head *list) -{ - struct dccp_tx_hist_entry *head = NULL; +struct dccp_rx_hist { + struct kmem_cache *dccprxh_slab; +}; - if (!list_empty(list)) - head = list_entry(list->next, struct dccp_tx_hist_entry, - dccphtx_node); - return head; -} +extern struct dccp_rx_hist *dccp_rx_hist_new(const char *name); +extern void dccp_rx_hist_delete(struct dccp_rx_hist *hist); static inline struct dccp_rx_hist_entry * - dccp_rx_hist_entry_new(struct dccp_rx_hist *hist, - const struct sock *sk, - const u32 ndp, - const struct sk_buff *skb, - const gfp_t prio) + dccp_rx_hist_entry_new(struct dccp_rx_hist *hist, + const struct sock *sk, + const u32 ndp, + const struct sk_buff *skb, + const gfp_t prio) { struct dccp_rx_hist_entry *entry = kmem_cache_alloc(hist->dccprxh_slab, prio); @@ -156,18 +157,8 @@ static inline struct dccp_rx_hist_entry * return entry; } -static inline void dccp_rx_hist_entry_delete(struct dccp_rx_hist *hist, - struct dccp_rx_hist_entry *entry) -{ - if (entry != NULL) - kmem_cache_free(hist->dccprxh_slab, entry); -} - -extern void dccp_rx_hist_purge(struct dccp_rx_hist *hist, - struct list_head *list); - static inline struct dccp_rx_hist_entry * - dccp_rx_hist_head(struct list_head *list) + dccp_rx_hist_head(struct list_head *list) { struct dccp_rx_hist_entry *head = NULL; @@ -177,6 +168,27 @@ static inline struct dccp_rx_hist_entry * return head; } +extern int dccp_rx_hist_find_entry(const struct list_head *list, const u64 seq, + u8 *ccval); +extern struct dccp_rx_hist_entry * + dccp_rx_hist_find_data_packet(const struct list_head *list); + +extern void dccp_rx_hist_add_packet(struct dccp_rx_hist *hist, + struct list_head *rx_list, + struct list_head *li_list, + struct dccp_rx_hist_entry *packet, + u64 nonloss_seqno); + +static inline void dccp_rx_hist_entry_delete(struct dccp_rx_hist *hist, + struct dccp_rx_hist_entry *entry) +{ + if (entry != NULL) + kmem_cache_free(hist->dccprxh_slab, entry); +} + +extern void dccp_rx_hist_purge(struct dccp_rx_hist *hist, + struct list_head *list); + static inline int dccp_rx_hist_entry_data_packet(const struct dccp_rx_hist_entry *entry) { @@ -184,12 +196,6 @@ static inline int entry->dccphrx_type == DCCP_PKT_DATAACK; } -extern void dccp_rx_hist_add_packet(struct dccp_rx_hist *hist, - struct list_head *rx_list, - struct list_head *li_list, - struct dccp_rx_hist_entry *packet, - u64 nonloss_seqno); - extern u64 dccp_rx_hist_detect_loss(struct list_head *rx_list, struct list_head *li_list, u8 *win_loss); diff --git a/net/dccp/ccids/lib/tfrc.h b/net/dccp/ccids/lib/tfrc.h index 45f30f59ea2..faf5f7e219e 100644 --- a/net/dccp/ccids/lib/tfrc.h +++ b/net/dccp/ccids/lib/tfrc.h @@ -13,8 +13,29 @@ * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. */ - #include <linux/types.h> +#include <asm/div64.h> + +/* integer-arithmetic divisions of type (a * 1000000)/b */ +static inline u64 scaled_div(u64 a, u32 b) +{ + BUG_ON(b==0); + a *= 1000000; + do_div(a, b); + return a; +} + +static inline u32 scaled_div32(u64 a, u32 b) +{ + u64 result = scaled_div(a, b); + + if (result > UINT_MAX) { + DCCP_CRIT("Overflow: a(%llu)/b(%u) > ~0U", + (unsigned long long)a, b); + return UINT_MAX; + } + return result; +} extern u32 tfrc_calc_x(u16 s, u32 R, u32 p); extern u32 tfrc_calc_x_reverse_lookup(u32 fvalue); diff --git a/net/dccp/ccids/lib/tfrc_equation.c b/net/dccp/ccids/lib/tfrc_equation.c index 44076e0c659..90009fd77e1 100644 --- a/net/dccp/ccids/lib/tfrc_equation.c +++ b/net/dccp/ccids/lib/tfrc_equation.c @@ -13,16 +13,83 @@ */ #include <linux/module.h> - -#include <asm/div64.h> - +#include "../../dccp.h" #include "tfrc.h" #define TFRC_CALC_X_ARRSIZE 500 +#define TFRC_CALC_X_SPLIT 50000 /* 0.05 * 1000000, details below */ +#define TFRC_SMALLEST_P (TFRC_CALC_X_SPLIT/TFRC_CALC_X_ARRSIZE) -#define TFRC_CALC_X_SPLIT 50000 -/* equivalent to 0.05 */ - +/* + TFRC TCP Reno Throughput Equation Lookup Table for f(p) + + The following two-column lookup table implements a part of the TCP throughput + equation from [RFC 3448, sec. 3.1]: + + s + X_calc = -------------------------------------------------------------- + R * sqrt(2*b*p/3) + (3 * t_RTO * sqrt(3*b*p/8) * (p + 32*p^3)) + + Where: + X is the transmit rate in bytes/second + s is the packet size in bytes + R is the round trip time in seconds + p is the loss event rate, between 0 and 1.0, of the number of loss + events as a fraction of the number of packets transmitted + t_RTO is the TCP retransmission timeout value in seconds + b is the number of packets acknowledged by a single TCP ACK + + We can assume that b = 1 and t_RTO is 4 * R. The equation now becomes: + + s + X_calc = ------------------------------------------------------- + R * sqrt(p*2/3) + (12 * R * sqrt(p*3/8) * (p + 32*p^3)) + + which we can break down into: + + s + X_calc = --------- + R * f(p) + + where f(p) is given for 0 < p <= 1 by: + + f(p) = sqrt(2*p/3) + 12 * sqrt(3*p/8) * (p + 32*p^3) + + Since this is kernel code, floating-point arithmetic is avoided in favour of + integer arithmetic. This means that nearly all fractional parameters are + scaled by 1000000: + * the parameters p and R + * the return result f(p) + The lookup table therefore actually tabulates the following function g(q): + + g(q) = 1000000 * f(q/1000000) + + Hence, when p <= 1, q must be less than or equal to 1000000. To achieve finer + granularity for the practically more relevant case of small values of p (up to + 5%), the second column is used; the first one ranges up to 100%. This split + corresponds to the value of q = TFRC_CALC_X_SPLIT. At the same time this also + determines the smallest resolution possible with this lookup table: + + TFRC_SMALLEST_P = TFRC_CALC_X_SPLIT / TFRC_CALC_X_ARRSIZE + + The entire table is generated by: + for(i=0; i < TFRC_CALC_X_ARRSIZE; i++) { + lookup[i][0] = g((i+1) * 1000000/TFRC_CALC_X_ARRSIZE); + lookup[i][1] = g((i+1) * TFRC_CALC_X_SPLIT/TFRC_CALC_X_ARRSIZE); + } + + With the given configuration, we have, with M = TFRC_CALC_X_ARRSIZE-1, + lookup[0][0] = g(1000000/(M+1)) = 1000000 * f(0.2%) + lookup[M][0] = g(1000000) = 1000000 * f(100%) + lookup[0][1] = g(TFRC_SMALLEST_P) = 1000000 * f(0.01%) + lookup[M][1] = g(TFRC_CALC_X_SPLIT) = 1000000 * f(5%) + + In summary, the two columns represent f(p) for the following ranges: + * The first column is for 0.002 <= p <= 1.0 + * The second column is for 0.0001 <= p <= 0.05 + Where the columns overlap, the second (finer-grained) is given preference, + i.e. the first column is used only for p >= 0.05. + */ static const u32 tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE][2] = { { 37172, 8172 }, { 53499, 11567 }, @@ -526,117 +593,105 @@ static const u32 tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE][2] = { { 243315981, 271305 } }; -/* Calculate the send rate as per section 3.1 of RFC3448 - -Returns send rate in bytes per second - -Integer maths and lookups are used as not allowed floating point in kernel - -The function for Xcalc as per section 3.1 of RFC3448 is: - -X = s - ------------------------------------------------------------- - R*sqrt(2*b*p/3) + (t_RTO * (3*sqrt(3*b*p/8) * p * (1+32*p^2))) - -where -X is the trasmit rate in bytes/second -s is the packet size in bytes -R is the round trip time in seconds -p is the loss event rate, between 0 and 1.0, of the number of loss events - as a fraction of the number of packets transmitted -t_RTO is the TCP retransmission timeout value in seconds -b is the number of packets acknowledged by a single TCP acknowledgement - -we can assume that b = 1 and t_RTO is 4 * R. With this the equation becomes: - -X = s - ----------------------------------------------------------------------- - R * sqrt(2 * p / 3) + (12 * R * (sqrt(3 * p / 8) * p * (1 + 32 * p^2))) - - -which we can break down into: - -X = s - -------- - R * f(p) - -where f(p) = sqrt(2 * p / 3) + (12 * sqrt(3 * p / 8) * p * (1 + 32 * p * p)) - -Function parameters: -s - bytes -R - RTT in usecs -p - loss rate (decimal fraction multiplied by 1,000,000) - -Returns Xcalc in bytes per second - -DON'T alter this code unless you run test cases against it as the code -has been manipulated to stop underflow/overlow. +/* return largest index i such that fval <= lookup[i][small] */ +static inline u32 tfrc_binsearch(u32 fval, u8 small) +{ + u32 try, low = 0, high = TFRC_CALC_X_ARRSIZE - 1; + + while (low < high) { + try = (low + high) / 2; + if (fval <= tfrc_calc_x_lookup[try][small]) + high = try; + else + low = try + 1; + } + return high; +} -*/ +/** + * tfrc_calc_x - Calculate the send rate as per section 3.1 of RFC3448 + * + * @s: packet size in bytes + * @R: RTT scaled by 1000000 (i.e., microseconds) + * @p: loss ratio estimate scaled by 1000000 + * Returns X_calc in bytes per second (not scaled). + */ u32 tfrc_calc_x(u16 s, u32 R, u32 p) { - int index; + u16 index; u32 f; - u64 tmp1, tmp2; + u64 result; - if (p < TFRC_CALC_X_SPLIT) - index = (p / (TFRC_CALC_X_SPLIT / TFRC_CALC_X_ARRSIZE)) - 1; - else - index = (p / (1000000 / TFRC_CALC_X_ARRSIZE)) - 1; + /* check against invalid parameters and divide-by-zero */ + BUG_ON(p > 1000000); /* p must not exceed 100% */ + BUG_ON(p == 0); /* f(0) = 0, divide by zero */ + if (R == 0) { /* possible divide by zero */ + DCCP_CRIT("WARNING: RTT is 0, returning maximum X_calc."); + return ~0U; + } - if (index < 0) - /* p should be 0 unless there is a bug in my code */ - index = 0; + if (p <= TFRC_CALC_X_SPLIT) { /* 0.0000 < p <= 0.05 */ + if (p < TFRC_SMALLEST_P) { /* 0.0000 < p < 0.0001 */ + DCCP_WARN("Value of p (%d) below resolution. " + "Substituting %d\n", p, TFRC_SMALLEST_P); + index = 0; + } else /* 0.0001 <= p <= 0.05 */ + index = p/TFRC_SMALLEST_P - 1; - if (R == 0) - R = 1; /* RTT can't be zero or else divide by zero */ + f = tfrc_calc_x_lookup[index][1]; - BUG_ON(index >= TFRC_CALC_X_ARRSIZE); + } else { /* 0.05 < p <= 1.00 */ + index = p/(1000000/TFRC_CALC_X_ARRSIZE) - 1; - if (p >= TFRC_CALC_X_SPLIT) f = tfrc_calc_x_lookup[index][0]; - else - f = tfrc_calc_x_lookup[index][1]; - - tmp1 = ((u64)s * 100000000); - tmp2 = ((u64)R * (u64)f); - do_div(tmp2, 10000); - do_div(tmp1, tmp2); - /* Don't alter above math unless you test due to overflow on 32 bit */ - - return (u32)tmp1; + } + + /* + * Compute X = s/(R*f(p)) in bytes per second. + * Since f(p) and R are both scaled by 1000000, we need to multiply by + * 1000000^2. To avoid overflow, the result is computed in two stages. + * This works under almost all reasonable operational conditions, for a + * wide range of parameters. Yet, should some strange combination of + * parameters result in overflow, the use of scaled_div32 will catch + * this and return UINT_MAX - which is a logically adequate consequence. + */ + result = scaled_div(s, R); + return scaled_div32(result, f); } EXPORT_SYMBOL_GPL(tfrc_calc_x); /* - * args: fvalue - function value to match - * returns: p closest to that value + * tfrc_calc_x_reverse_lookup - try to find p given f(p) * - * both fvalue and p are multiplied by 1,000,000 to use ints + * @fvalue: function value to match, scaled by 1000000 + * Returns closest match for p, also scaled by 1000000 */ u32 tfrc_calc_x_reverse_lookup(u32 fvalue) { - int ctr = 0; - int small; + int index; - if (fvalue < tfrc_calc_x_lookup[0][1]) + if (fvalue == 0) /* f(p) = 0 whenever p = 0 */ return 0; - if (fvalue <= tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE - 1][1]) - small = 1; - else if (fvalue > tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE - 1][0]) + /* Error cases. */ + if (fvalue < tfrc_calc_x_lookup[0][1]) { + DCCP_WARN("fvalue %d smaller than resolution\n", fvalue); + return tfrc_calc_x_lookup[0][1]; + } + if (fvalue > tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE - 1][0]) { + DCCP_WARN("fvalue %d exceeds bounds!\n", fvalue); return 1000000; - else - small = 0; - - while (fvalue > tfrc_calc_x_lookup[ctr][small]) - ctr++; + } - if (small) - return TFRC_CALC_X_SPLIT * ctr / TFRC_CALC_X_ARRSIZE; - else - return 1000000 * ctr / TFRC_CALC_X_ARRSIZE; + if (fvalue <= tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE - 1][1]) { + index = tfrc_binsearch(fvalue, 1); + return (index + 1) * TFRC_CALC_X_SPLIT / TFRC_CALC_X_ARRSIZE; + } + + /* else ... it must be in the coarse-grained column */ + index = tfrc_binsearch(fvalue, 0); + return (index + 1) * 1000000 / TFRC_CALC_X_ARRSIZE; } EXPORT_SYMBOL_GPL(tfrc_calc_x_reverse_lookup); diff --git a/net/dccp/dccp.h b/net/dccp/dccp.h index 272e8584564..a0900bf98e6 100644 --- a/net/dccp/dccp.h +++ b/net/dccp/dccp.h @@ -18,15 +18,33 @@ #include <net/tcp.h> #include "ackvec.h" +/* + * DCCP - specific warning and debugging macros. + */ +#define DCCP_WARN(fmt, a...) LIMIT_NETDEBUG(KERN_WARNING "%s: " fmt, \ + __FUNCTION__, ##a) +#define DCCP_CRIT(fmt, a...) printk(KERN_CRIT fmt " at %s:%d/%s()\n", ##a, \ + __FILE__, __LINE__, __FUNCTION__) +#define DCCP_BUG(a...) do { DCCP_CRIT("BUG: " a); dump_stack(); } while(0) +#define DCCP_BUG_ON(cond) do { if (unlikely((cond) != 0)) \ + DCCP_BUG("\"%s\" holds (exception!)", \ + __stringify(cond)); \ + } while (0) + +#ifdef MODULE +#define DCCP_PRINTK(enable, fmt, args...) do { if (enable) \ + printk(fmt, ##args); \ + } while(0) +#else +#define DCCP_PRINTK(enable, fmt, args...) printk(fmt, ##args) +#endif +#define DCCP_PR_DEBUG(enable, fmt, a...) DCCP_PRINTK(enable, KERN_DEBUG \ + "%s: " fmt, __FUNCTION__, ##a) + #ifdef CONFIG_IP_DCCP_DEBUG extern int dccp_debug; - -#define dccp_pr_debug(format, a...) \ - do { if (dccp_debug) \ - printk(KERN_DEBUG "%s: " format, __FUNCTION__ , ##a); \ - } while (0) -#define dccp_pr_debug_cat(format, a...) do { if (dccp_debug) \ - printk(format, ##a); } while (0) +#define dccp_pr_debug(format, a...) DCCP_PR_DEBUG(dccp_debug, format, ##a) +#define dccp_pr_debug_cat(format, a...) DCCP_PRINTK(dccp_debug, format, ##a) #else #define dccp_pr_debug(format, a...) #define dccp_pr_debug_cat(format, a...) @@ -35,17 +53,21 @@ extern int dccp_debug; extern struct inet_hashinfo dccp_hashinfo; extern atomic_t dccp_orphan_count; -extern int dccp_tw_count; -extern void dccp_tw_deschedule(struct inet_timewait_sock *tw); extern void dccp_time_wait(struct sock *sk, int state, int timeo); -/* FIXME: Right size this */ -#define DCCP_MAX_OPT_LEN 128 - -#define DCCP_MAX_PACKET_HDR 32 - -#define MAX_DCCP_HEADER (DCCP_MAX_PACKET_HDR + DCCP_MAX_OPT_LEN + MAX_HEADER) +/* + * Set safe upper bounds for header and option length. Since Data Offset is 8 + * bits (RFC 4340, sec. 5.1), the total header length can never be more than + * 4 * 255 = 1020 bytes. The largest possible header length is 28 bytes (X=1): + * - DCCP-Response with ACK Subheader and 4 bytes of Service code OR + * - DCCP-Reset with ACK Subheader and 4 bytes of Reset Code fields + * Hence a safe upper bound for the maximum option length is 1020-28 = 992 + */ +#define MAX_DCCP_SPECIFIC_HEADER (255 * sizeof(int)) +#define DCCP_MAX_PACKET_HDR 28 +#define DCCP_MAX_OPT_LEN (MAX_DCCP_SPECIFIC_HEADER - DCCP_MAX_PACKET_HDR) +#define MAX_DCCP_HEADER (MAX_DCCP_SPECIFIC_HEADER + MAX_HEADER) #define DCCP_TIMEWAIT_LEN (60 * HZ) /* how long to wait to destroy TIME-WAIT * state, about 60 seconds */ @@ -58,6 +80,18 @@ extern void dccp_time_wait(struct sock *sk, int state, int timeo); #define DCCP_RTO_MAX ((unsigned)(120 * HZ)) /* FIXME: using TCP value */ +/* sysctl variables for DCCP */ +extern int sysctl_dccp_request_retries; +extern int sysctl_dccp_retries1; +extern int sysctl_dccp_retries2; +extern int sysctl_dccp_feat_sequence_window; +extern int sysctl_dccp_feat_rx_ccid; +extern int sysctl_dccp_feat_tx_ccid; +extern int sysctl_dccp_feat_ack_ratio; +extern int sysctl_dccp_feat_send_ack_vector; +extern int sysctl_dccp_feat_send_ndp_count; +extern int sysctl_dccp_tx_qlen; + /* is seq1 < seq2 ? */ static inline int before48(const u64 seq1, const u64 seq2) { @@ -123,10 +157,36 @@ DECLARE_SNMP_STAT(struct dccp_mib, dccp_statistics); #define DCCP_ADD_STATS_USER(field, val) \ SNMP_ADD_STATS_USER(dccp_statistics, field, val) +/* + * Checksumming routines + */ +static inline int dccp_csum_coverage(const struct sk_buff *skb) +{ + const struct dccp_hdr* dh = dccp_hdr(skb); + + if (dh->dccph_cscov == 0) + return skb->len; + return (dh->dccph_doff + dh->dccph_cscov - 1) * sizeof(u32); +} + +static inline void dccp_csum_outgoing(struct sk_buff *skb) +{ + int cov = dccp_csum_coverage(skb); + + if (cov >= skb->len) + dccp_hdr(skb)->dccph_cscov = 0; + + skb->csum = skb_checksum(skb, 0, (cov > skb->len)? skb->len : cov, 0); +} + +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 void dccp_send_ack(struct sock *sk); extern void dccp_send_delayed_ack(struct sock *sk); +extern void dccp_reqsk_send_ack(struct sk_buff *sk, struct request_sock *rsk); + extern void dccp_send_sync(struct sock *sk, const u64 seq, const enum dccp_pkt_type pkt_type); @@ -147,18 +207,7 @@ extern const char *dccp_state_name(const int state); extern void dccp_set_state(struct sock *sk, const int state); extern void dccp_done(struct sock *sk); -static inline void dccp_openreq_init(struct request_sock *req, - struct dccp_sock *dp, - struct sk_buff *skb) -{ - /* - * FIXME: fill in the other req fields from the DCCP options - * received - */ - inet_rsk(req)->rmt_port = dccp_hdr(skb)->dccph_sport; - inet_rsk(req)->acked = 0; - req->rcv_wnd = 0; -} +extern void dccp_reqsk_init(struct request_sock *req, struct sk_buff *skb); extern int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb); @@ -217,14 +266,9 @@ extern void dccp_shutdown(struct sock *sk, int how); extern int inet_dccp_listen(struct socket *sock, int backlog); extern unsigned int dccp_poll(struct file *file, struct socket *sock, poll_table *wait); -extern void dccp_v4_send_check(struct sock *sk, int len, - struct sk_buff *skb); extern int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); -extern int dccp_v4_checksum(const struct sk_buff *skb, - const __be32 saddr, const __be32 daddr); - extern int dccp_send_reset(struct sock *sk, enum dccp_reset_codes code); extern void dccp_send_close(struct sock *sk, const int active); extern int dccp_invalid_packet(struct sk_buff *skb); @@ -388,6 +432,7 @@ static inline void timeval_sub_usecs(struct timeval *tv, tv->tv_sec--; tv->tv_usec += USEC_PER_SEC; } + DCCP_BUG_ON(tv->tv_sec < 0); } #ifdef CONFIG_SYSCTL diff --git a/net/dccp/feat.c b/net/dccp/feat.c index a1b0682ee77..95b6927ec65 100644 --- a/net/dccp/feat.c +++ b/net/dccp/feat.c @@ -12,7 +12,6 @@ #include <linux/module.h> -#include "dccp.h" #include "ccid.h" #include "feat.h" @@ -23,9 +22,17 @@ int dccp_feat_change(struct dccp_minisock *dmsk, u8 type, u8 feature, { struct dccp_opt_pend *opt; - dccp_pr_debug("feat change type=%d feat=%d\n", type, feature); + dccp_feat_debug(type, feature, *val); - /* XXX sanity check feat change request */ + if (!dccp_feat_is_valid_type(type)) { + DCCP_WARN("option type %d invalid in negotiation\n", type); + return 1; + } + if (!dccp_feat_is_valid_length(type, feature, len)) { + DCCP_WARN("invalid length %d\n", len); + return 1; + } + /* XXX add further sanity checks */ /* check if that feature is already being negotiated */ list_for_each_entry(opt, &dmsk->dccpms_pending, dccpop_node) { @@ -95,14 +102,14 @@ static int dccp_feat_update_ccid(struct sock *sk, u8 type, u8 new_ccid_nr) /* XXX taking only u8 vals */ static int dccp_feat_update(struct sock *sk, u8 type, u8 feat, u8 val) { - dccp_pr_debug("changing [%d] feat %d to %d\n", type, feat, val); + dccp_feat_debug(type, feat, val); switch (feat) { case DCCPF_CCID: return dccp_feat_update_ccid(sk, type, val); default: - dccp_pr_debug("IMPLEMENT changing [%d] feat %d to %d\n", - type, feat, val); + dccp_pr_debug("UNIMPLEMENTED: %s(%d, ...)\n", + dccp_feat_typename(type), feat); break; } return 0; @@ -162,7 +169,8 @@ static int dccp_feat_reconcile(struct sock *sk, struct dccp_opt_pend *opt, break; default: - WARN_ON(1); /* XXX implement res */ + DCCP_BUG("Fell through, feat=%d", opt->dccpop_feat); + /* XXX implement res */ return -EFAULT; } @@ -265,10 +273,10 @@ static int dccp_feat_nn(struct sock *sk, u8 type, u8 feature, u8 *val, u8 len) u8 *copy; int rc; - /* NN features must be change L */ - if (type == DCCPO_CHANGE_R) { - dccp_pr_debug("received CHANGE_R %d for NN feat %d\n", - type, feature); + /* NN features must be Change L (sec. 6.3.2) */ + if (type != DCCPO_CHANGE_L) { + dccp_pr_debug("received %s for NN feature %d\n", + dccp_feat_typename(type), feature); return -EFAULT; } @@ -279,12 +287,11 @@ static int dccp_feat_nn(struct sock *sk, u8 type, u8 feature, u8 *val, u8 len) if (opt == NULL) return -ENOMEM; - copy = kmalloc(len, GFP_ATOMIC); + copy = kmemdup(val, len, GFP_ATOMIC); if (copy == NULL) { kfree(opt); return -ENOMEM; } - memcpy(copy, val, len); opt->dccpop_type = DCCPO_CONFIRM_R; /* NN can only confirm R */ opt->dccpop_feat = feature; @@ -299,7 +306,8 @@ static int dccp_feat_nn(struct sock *sk, u8 type, u8 feature, u8 *val, u8 len) return rc; } - dccp_pr_debug("Confirming NN feature %d (val=%d)\n", feature, *copy); + dccp_feat_debug(type, feature, *copy); + list_add_tail(&opt->dccpop_node, &dmsk->dccpms_conf); return 0; @@ -318,14 +326,19 @@ static void dccp_feat_empty_confirm(struct dccp_minisock *dmsk, return; } - opt->dccpop_type = type == DCCPO_CHANGE_L ? DCCPO_CONFIRM_R : - DCCPO_CONFIRM_L; + switch (type) { + case DCCPO_CHANGE_L: opt->dccpop_type = DCCPO_CONFIRM_R; break; + case DCCPO_CHANGE_R: opt->dccpop_type = DCCPO_CONFIRM_L; break; + default: DCCP_WARN("invalid type %d\n", type); return; + + } opt->dccpop_feat = feature; opt->dccpop_val = NULL; opt->dccpop_len = 0; /* change feature */ - dccp_pr_debug("Empty confirm feature %d type %d\n", feature, type); + dccp_pr_debug("Empty %s(%d)\n", dccp_feat_typename(type), feature); + list_add_tail(&opt->dccpop_node, &dmsk->dccpms_conf); } @@ -359,7 +372,7 @@ int dccp_feat_change_recv(struct sock *sk, u8 type, u8 feature, u8 *val, u8 len) { int rc; - dccp_pr_debug("got feat change type=%d feat=%d\n", type, feature); + dccp_feat_debug(type, feature, *val); /* figure out if it's SP or NN feature */ switch (feature) { @@ -375,6 +388,8 @@ int dccp_feat_change_recv(struct sock *sk, u8 type, u8 feature, u8 *val, u8 len) /* XXX implement other features */ default: + dccp_pr_debug("UNIMPLEMENTED: not handling %s(%d, ...)\n", + dccp_feat_typename(type), feature); rc = -EFAULT; break; } @@ -403,20 +418,27 @@ int dccp_feat_confirm_recv(struct sock *sk, u8 type, u8 feature, u8 t; struct dccp_opt_pend *opt; struct dccp_minisock *dmsk = dccp_msk(sk); - int rc = 1; + int found = 0; int all_confirmed = 1; - dccp_pr_debug("got feat confirm type=%d feat=%d\n", type, feature); - - /* XXX sanity check type & feat */ + dccp_feat_debug(type, feature, *val); /* locate our change request */ - t = type == DCCPO_CONFIRM_L ? DCCPO_CHANGE_R : DCCPO_CHANGE_L; + switch (type) { + case DCCPO_CONFIRM_L: t = DCCPO_CHANGE_R; break; + case DCCPO_CONFIRM_R: t = DCCPO_CHANGE_L; break; + default: DCCP_WARN("invalid type %d\n", type); + return 1; + + } + /* XXX sanity check feature value */ list_for_each_entry(opt, &dmsk->dccpms_pending, dccpop_node) { if (!opt->dccpop_conf && opt->dccpop_type == t && opt->dccpop_feat == feature) { - /* we found it */ + found = 1; + dccp_pr_debug("feature %d found\n", opt->dccpop_feat); + /* XXX do sanity check */ opt->dccpop_conf = 1; @@ -425,9 +447,7 @@ int dccp_feat_confirm_recv(struct sock *sk, u8 type, u8 feature, dccp_feat_update(sk, opt->dccpop_type, opt->dccpop_feat, *val); - dccp_pr_debug("feat %d type %d confirmed %d\n", - feature, type, *val); - rc = 0; + /* XXX check the return value of dccp_feat_update */ break; } @@ -446,9 +466,9 @@ int dccp_feat_confirm_recv(struct sock *sk, u8 type, u8 feature, inet_csk_clear_xmit_timer(sk, ICSK_TIME_RETRANS); } - if (rc) - dccp_pr_debug("feat %d type %d never requested\n", - feature, type); + if (!found) + dccp_pr_debug("%s(%d, ...) never requested\n", + dccp_feat_typename(type), feature); return 0; } @@ -501,20 +521,18 @@ int dccp_feat_clone(struct sock *oldsk, struct sock *newsk) list_for_each_entry(opt, &olddmsk->dccpms_pending, dccpop_node) { struct dccp_opt_pend *newopt; /* copy the value of the option */ - u8 *val = kmalloc(opt->dccpop_len, GFP_ATOMIC); + u8 *val = kmemdup(opt->dccpop_val, opt->dccpop_len, GFP_ATOMIC); if (val == NULL) goto out_clean; - memcpy(val, opt->dccpop_val, opt->dccpop_len); - newopt = kmalloc(sizeof(*newopt), GFP_ATOMIC); + newopt = kmemdup(opt, sizeof(*newopt), GFP_ATOMIC); if (newopt == NULL) { kfree(val); goto out_clean; } /* insert the option */ - memcpy(newopt, opt, sizeof(*newopt)); newopt->dccpop_val = val; list_add_tail(&newopt->dccpop_node, &newdmsk->dccpms_pending); @@ -545,10 +563,9 @@ static int __dccp_feat_init(struct dccp_minisock *dmsk, u8 type, u8 feat, u8 *val, u8 len) { int rc = -ENOMEM; - u8 *copy = kmalloc(len, GFP_KERNEL); + u8 *copy = kmemdup(val, len, GFP_KERNEL); if (copy != NULL) { - memcpy(copy, val, len); rc = dccp_feat_change(dmsk, type, feat, copy, len, GFP_KERNEL); if (rc) kfree(copy); @@ -583,3 +600,45 @@ out: } EXPORT_SYMBOL_GPL(dccp_feat_init); + +#ifdef CONFIG_IP_DCCP_DEBUG +const char *dccp_feat_typename(const u8 type) +{ + switch(type) { + case DCCPO_CHANGE_L: return("ChangeL"); + case DCCPO_CONFIRM_L: return("ConfirmL"); + case DCCPO_CHANGE_R: return("ChangeR"); + case DCCPO_CONFIRM_R: return("ConfirmR"); + /* the following case must not appear in feature negotation */ + default: dccp_pr_debug("unknown type %d [BUG!]\n", type); + } + return NULL; +} + +EXPORT_SYMBOL_GPL(dccp_feat_typename); + +const char *dccp_feat_name(const u8 feat) +{ + static const char *feature_names[] = { + [DCCPF_RESERVED] = "Reserved", + [DCCPF_CCID] = "CCID", + [DCCPF_SHORT_SEQNOS] = "Allow Short Seqnos", + [DCCPF_SEQUENCE_WINDOW] = "Sequence Window", + [DCCPF_ECN_INCAPABLE] = "ECN Incapable", + [DCCPF_ACK_RATIO] = "Ack Ratio", + [DCCPF_SEND_ACK_VECTOR] = "Send ACK Vector", + [DCCPF_SEND_NDP_COUNT] = "Send NDP Count", + [DCCPF_MIN_CSUM_COVER] = "Min. Csum Coverage", + [DCCPF_DATA_CHECKSUM] = "Send Data Checksum", + }; + if (feat >= DCCPF_MIN_CCID_SPECIFIC) + return "CCID-specific"; + + if (dccp_feat_is_reserved(feat)) + return feature_names[DCCPF_RESERVED]; + + return feature_names[feat]; +} + +EXPORT_SYMBOL_GPL(dccp_feat_name); +#endif /* CONFIG_IP_DCCP_DEBUG */ diff --git a/net/dccp/feat.h b/net/dccp/feat.h index cee553d416c..2c373ad7edc 100644 --- a/net/dccp/feat.h +++ b/net/dccp/feat.h @@ -12,9 +12,46 @@ */ #include <linux/types.h> +#include "dccp.h" -struct sock; -struct dccp_minisock; +static inline int dccp_feat_is_valid_length(u8 type, u8 feature, u8 len) +{ + /* sec. 6.1: Confirm has at least length 3, + * sec. 6.2: Change has at least length 4 */ + if (len < 3) + return 1; + if (len < 4 && (type == DCCPO_CHANGE_L || type == DCCPO_CHANGE_R)) + return 1; + /* XXX: add per-feature length validation (sec. 6.6.8) */ + return 0; +} + +static inline int dccp_feat_is_reserved(const u8 feat) +{ + return (feat > DCCPF_DATA_CHECKSUM && + feat < DCCPF_MIN_CCID_SPECIFIC) || + feat == DCCPF_RESERVED; +} + +/* feature negotiation knows only these four option types (RFC 4340, sec. 6) */ +static inline int dccp_feat_is_valid_type(const u8 optnum) +{ + return optnum >= DCCPO_CHANGE_L && optnum <= DCCPO_CONFIRM_R; + +} + +#ifdef CONFIG_IP_DCCP_DEBUG +extern const char *dccp_feat_typename(const u8 type); +extern const char *dccp_feat_name(const u8 feat); + +static inline void dccp_feat_debug(const u8 type, const u8 feat, const u8 val) +{ + dccp_pr_debug("%s(%s (%d), %d)\n", dccp_feat_typename(type), + dccp_feat_name(feat), feat, val); +} +#else +#define dccp_feat_debug(type, feat, val) +#endif /* CONFIG_IP_DCCP_DEBUG */ extern int dccp_feat_change(struct dccp_minisock *dmsk, u8 type, u8 feature, u8 *val, u8 len, gfp_t gfp); @@ -26,11 +63,4 @@ extern void dccp_feat_clean(struct dccp_minisock *dmsk); extern int dccp_feat_clone(struct sock *oldsk, struct sock *newsk); extern int dccp_feat_init(struct dccp_minisock *dmsk); -extern int dccp_feat_default_sequence_window; -extern int dccp_feat_default_rx_ccid; -extern int dccp_feat_default_tx_ccid; -extern int dccp_feat_default_ack_ratio; -extern int dccp_feat_default_send_ack_vector; -extern int dccp_feat_default_send_ndp_count; - #endif /* _DCCP_FEAT_H */ diff --git a/net/dccp/input.c b/net/dccp/input.c index 1d24881ac0a..565bc80557c 100644 --- a/net/dccp/input.c +++ b/net/dccp/input.c @@ -1,6 +1,6 @@ /* * net/dccp/input.c - * + * * An implementation of the DCCP protocol * Arnaldo Carvalho de Melo <acme@conectiva.com.br> * @@ -82,7 +82,7 @@ static int dccp_check_seqno(struct sock *sk, struct sk_buff *skb) * Otherwise, * Drop packet and return */ - if (dh->dccph_type == DCCP_PKT_SYNC || + if (dh->dccph_type == DCCP_PKT_SYNC || dh->dccph_type == DCCP_PKT_SYNCACK) { if (between48(DCCP_SKB_CB(skb)->dccpd_ack_seq, dp->dccps_awl, dp->dccps_awh) && @@ -128,21 +128,18 @@ static int dccp_check_seqno(struct sock *sk, struct sk_buff *skb) DCCP_PKT_WITHOUT_ACK_SEQ)) dp->dccps_gar = DCCP_SKB_CB(skb)->dccpd_ack_seq; } else { - LIMIT_NETDEBUG(KERN_WARNING "DCCP: Step 6 failed for %s packet, " - "(LSWL(%llu) <= P.seqno(%llu) <= S.SWH(%llu)) and " - "(P.ackno %s or LAWL(%llu) <= P.ackno(%llu) <= S.AWH(%llu), " - "sending SYNC...\n", - dccp_packet_name(dh->dccph_type), - (unsigned long long) lswl, - (unsigned long long) - DCCP_SKB_CB(skb)->dccpd_seq, - (unsigned long long) dp->dccps_swh, - (DCCP_SKB_CB(skb)->dccpd_ack_seq == + DCCP_WARN("DCCP: Step 6 failed for %s packet, " + "(LSWL(%llu) <= P.seqno(%llu) <= S.SWH(%llu)) and " + "(P.ackno %s or LAWL(%llu) <= P.ackno(%llu) <= S.AWH(%llu), " + "sending SYNC...\n", dccp_packet_name(dh->dccph_type), + (unsigned long long) lswl, + (unsigned long long) DCCP_SKB_CB(skb)->dccpd_seq, + (unsigned long long) dp->dccps_swh, + (DCCP_SKB_CB(skb)->dccpd_ack_seq == DCCP_PKT_WITHOUT_ACK_SEQ) ? "doesn't exist" : "exists", - (unsigned long long) lawl, - (unsigned long long) - DCCP_SKB_CB(skb)->dccpd_ack_seq, - (unsigned long long) dp->dccps_awh); + (unsigned long long) lawl, + (unsigned long long) DCCP_SKB_CB(skb)->dccpd_ack_seq, + (unsigned long long) dp->dccps_awh); dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq, DCCP_PKT_SYNC); return -1; } @@ -188,8 +185,8 @@ static int __dccp_rcv_established(struct sock *sk, struct sk_buff *skb, dccp_rcv_close(sk, skb); return 0; case DCCP_PKT_REQUEST: - /* Step 7 - * or (S.is_server and P.type == Response) + /* Step 7 + * or (S.is_server and P.type == Response) * or (S.is_client and P.type == Request) * or (S.state >= OPEN and P.type == Request * and P.seqno >= S.OSR) @@ -251,8 +248,18 @@ int dccp_rcv_established(struct sock *sk, struct sk_buff *skb, DCCP_ACKVEC_STATE_RECEIVED)) goto discard; - ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb); - ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb); + /* + * Deliver to the CCID module in charge. + * FIXME: Currently DCCP operates one-directional only, i.e. a listening + * server is not at the same time a connecting client. There is + * not much sense in delivering to both rx/tx sides at the moment + * (only one is active at a time); when moving to bidirectional + * service, this needs to be revised. + */ + if (dccp_sk(sk)->dccps_role == DCCP_ROLE_SERVER) + ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb); + else + ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb); return __dccp_rcv_established(sk, skb, dh, len); discard: @@ -267,7 +274,7 @@ static int dccp_rcv_request_sent_state_process(struct sock *sk, const struct dccp_hdr *dh, const unsigned len) { - /* + /* * Step 4: Prepare sequence numbers in REQUEST * If S.state == REQUEST, * If (P.type == Response or P.type == Reset) @@ -335,7 +342,7 @@ static int dccp_rcv_request_sent_state_process(struct sock *sk, * from the Response * / * S.state := PARTOPEN * Set PARTOPEN timer - * Continue with S.state == PARTOPEN + * Continue with S.state == PARTOPEN * / * Step 12 will send the Ack completing the * three-way handshake * / */ @@ -366,7 +373,7 @@ static int dccp_rcv_request_sent_state_process(struct sock *sk, */ __kfree_skb(skb); return 0; - } + } dccp_send_ack(sk); return -1; } @@ -374,7 +381,7 @@ static int dccp_rcv_request_sent_state_process(struct sock *sk, out_invalid_packet: /* dccp_v4_do_rcv will send a reset */ DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_PACKET_ERROR; - return 1; + return 1; } static int dccp_rcv_respond_partopen_state_process(struct sock *sk, @@ -431,29 +438,25 @@ int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb, /* * Step 3: Process LISTEN state - * (Continuing from dccp_v4_do_rcv and dccp_v6_do_rcv) * * If S.state == LISTEN, - * If P.type == Request or P contains a valid Init Cookie - * option, - * * Must scan the packet's options to check for an Init - * Cookie. Only the Init Cookie is processed here, - * however; other options are processed in Step 8. This - * scan need only be performed if the endpoint uses Init - * Cookies * - * * Generate a new socket and switch to that socket * - * Set S := new socket for this port pair - * S.state = RESPOND - * Choose S.ISS (initial seqno) or set from Init Cookie - * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie - * Continue with S.state == RESPOND - * * A Response packet will be generated in Step 11 * - * Otherwise, - * Generate Reset(No Connection) unless P.type == Reset - * Drop packet and return - * - * NOTE: the check for the packet types is done in - * dccp_rcv_state_process + * If P.type == Request or P contains a valid Init Cookie option, + * (* Must scan the packet's options to check for Init + * Cookies. Only Init Cookies are processed here, + * however; other options are processed in Step 8. This + * scan need only be performed if the endpoint uses Init + * Cookies *) + * (* Generate a new socket and switch to that socket *) + * Set S := new socket for this port pair + * S.state = RESPOND + * Choose S.ISS (initial seqno) or set from Init Cookies + * Initialize S.GAR := S.ISS + * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init + * Cookies Continue with S.state == RESPOND + * (* A Response packet will be generated in Step 11 *) + * Otherwise, + * Generate Reset(No Connection) unless P.type == Reset + * Drop packet and return */ if (sk->sk_state == DCCP_LISTEN) { if (dh->dccph_type == DCCP_PKT_REQUEST) { @@ -485,14 +488,17 @@ int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb, if (dcb->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ) dccp_event_ack_recv(sk, skb); - if (dccp_msk(sk)->dccpms_send_ack_vector && + if (dccp_msk(sk)->dccpms_send_ack_vector && dccp_ackvec_add(dp->dccps_hc_rx_ackvec, sk, - DCCP_SKB_CB(skb)->dccpd_seq, - DCCP_ACKVEC_STATE_RECEIVED)) - goto discard; + DCCP_SKB_CB(skb)->dccpd_seq, + DCCP_ACKVEC_STATE_RECEIVED)) + goto discard; - ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb); - ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb); + /* XXX see the comments in dccp_rcv_established about this */ + if (dccp_sk(sk)->dccps_role == DCCP_ROLE_SERVER) + ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb); + else + ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb); } /* @@ -574,7 +580,7 @@ int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb, } } - if (!queued) { + if (!queued) { discard: __kfree_skb(skb); } diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index e08e7688a26..90c74b4adb7 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -113,13 +113,8 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) /* OK, now commit destination to socket. */ sk_setup_caps(sk, &rt->u.dst); - dp->dccps_gar = - dp->dccps_iss = secure_dccp_sequence_number(inet->saddr, - inet->daddr, - inet->sport, - usin->sin_port); - dccp_update_gss(sk, dp->dccps_iss); - + dp->dccps_iss = secure_dccp_sequence_number(inet->saddr, inet->daddr, + inet->sport, inet->dport); inet->id = dp->dccps_iss ^ jiffies; err = dccp_connect(sk); @@ -162,7 +157,7 @@ static inline void dccp_do_pmtu_discovery(struct sock *sk, /* We don't check in the destentry if pmtu discovery is forbidden * on this route. We just assume that no packet_to_big packets * are send back when pmtu discovery is not active. - * There is a small race when the user changes this flag in the + * There is a small race when the user changes this flag in the * route, but I think that's acceptable. */ if ((dst = __sk_dst_check(sk, 0)) == NULL) @@ -193,86 +188,6 @@ static inline void dccp_do_pmtu_discovery(struct sock *sk, } /* else let the usual retransmit timer handle it */ } -static void dccp_v4_reqsk_send_ack(struct sk_buff *rxskb, - struct request_sock *req) -{ - int err; - struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh; - const u32 dccp_hdr_ack_len = sizeof(struct dccp_hdr) + - sizeof(struct dccp_hdr_ext) + - sizeof(struct dccp_hdr_ack_bits); - struct sk_buff *skb; - - if (((struct rtable *)rxskb->dst)->rt_type != RTN_LOCAL) - return; - - skb = alloc_skb(dccp_v4_ctl_socket->sk->sk_prot->max_header, GFP_ATOMIC); - if (skb == NULL) - return; - - /* Reserve space for headers. */ - skb_reserve(skb, dccp_v4_ctl_socket->sk->sk_prot->max_header); - - skb->dst = dst_clone(rxskb->dst); - - skb->h.raw = skb_push(skb, dccp_hdr_ack_len); - dh = dccp_hdr(skb); - memset(dh, 0, dccp_hdr_ack_len); - - /* Build DCCP header and checksum it. */ - dh->dccph_type = DCCP_PKT_ACK; - dh->dccph_sport = rxdh->dccph_dport; - dh->dccph_dport = rxdh->dccph_sport; - dh->dccph_doff = dccp_hdr_ack_len / 4; - dh->dccph_x = 1; - - dccp_hdr_set_seq(dh, DCCP_SKB_CB(rxskb)->dccpd_ack_seq); - dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), - DCCP_SKB_CB(rxskb)->dccpd_seq); - - bh_lock_sock(dccp_v4_ctl_socket->sk); - err = ip_build_and_send_pkt(skb, dccp_v4_ctl_socket->sk, - rxskb->nh.iph->daddr, - rxskb->nh.iph->saddr, NULL); - bh_unlock_sock(dccp_v4_ctl_socket->sk); - - if (err == NET_XMIT_CN || err == 0) { - DCCP_INC_STATS_BH(DCCP_MIB_OUTSEGS); - DCCP_INC_STATS_BH(DCCP_MIB_OUTRSTS); - } -} - -static int dccp_v4_send_response(struct sock *sk, struct request_sock *req, - struct dst_entry *dst) -{ - int err = -1; - struct sk_buff *skb; - - /* First, grab a route. */ - - if (dst == NULL && (dst = inet_csk_route_req(sk, req)) == NULL) - goto out; - - skb = dccp_make_response(sk, dst, req); - if (skb != NULL) { - const struct inet_request_sock *ireq = inet_rsk(req); - struct dccp_hdr *dh = dccp_hdr(skb); - - dh->dccph_checksum = dccp_v4_checksum(skb, ireq->loc_addr, - ireq->rmt_addr); - memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); - err = ip_build_and_send_pkt(skb, sk, ireq->loc_addr, - ireq->rmt_addr, - ireq->opt); - if (err == NET_XMIT_CN) - err = 0; - } - -out: - dst_release(dst); - return err; -} - /* * This routine is called by the ICMP module when it gets some sort of error * condition. If err < 0 then the socket should be closed and the error @@ -329,7 +244,7 @@ static void dccp_v4_err(struct sk_buff *skb, u32 info) seq = dccp_hdr_seq(skb); if (sk->sk_state != DCCP_LISTEN && !between48(seq, dp->dccps_swl, dp->dccps_swh)) { - NET_INC_STATS(LINUX_MIB_OUTOFWINDOWICMPS); + NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS); goto out; } @@ -429,19 +344,24 @@ out: sock_put(sk); } -/* This routine computes an IPv4 DCCP checksum. */ -void dccp_v4_send_check(struct sock *sk, int len, struct sk_buff *skb) +static inline __sum16 dccp_v4_csum_finish(struct sk_buff *skb, + __be32 src, __be32 dst) +{ + return csum_tcpudp_magic(src, dst, skb->len, IPPROTO_DCCP, skb->csum); +} + +void dccp_v4_send_check(struct sock *sk, int unused, struct sk_buff *skb) { const struct inet_sock *inet = inet_sk(sk); struct dccp_hdr *dh = dccp_hdr(skb); - dh->dccph_checksum = dccp_v4_checksum(skb, inet->saddr, inet->daddr); + dccp_csum_outgoing(skb); + dh->dccph_checksum = dccp_v4_csum_finish(skb, inet->saddr, inet->daddr); } EXPORT_SYMBOL_GPL(dccp_v4_send_check); -static inline u64 dccp_v4_init_sequence(const struct sock *sk, - const struct sk_buff *skb) +static inline u64 dccp_v4_init_sequence(const struct sk_buff *skb) { return secure_dccp_sequence_number(skb->nh.iph->daddr, skb->nh.iph->saddr, @@ -449,95 +369,6 @@ static inline u64 dccp_v4_init_sequence(const struct sock *sk, dccp_hdr(skb)->dccph_sport); } -static struct request_sock_ops dccp_request_sock_ops; - -int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb) -{ - struct inet_request_sock *ireq; - struct dccp_sock dp; - struct request_sock *req; - struct dccp_request_sock *dreq; - const __be32 saddr = skb->nh.iph->saddr; - const __be32 daddr = skb->nh.iph->daddr; - const __be32 service = dccp_hdr_request(skb)->dccph_req_service; - struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb); - __u8 reset_code = DCCP_RESET_CODE_TOO_BUSY; - - /* Never answer to DCCP_PKT_REQUESTs send to broadcast or multicast */ - if (((struct rtable *)skb->dst)->rt_flags & - (RTCF_BROADCAST | RTCF_MULTICAST)) { - reset_code = DCCP_RESET_CODE_NO_CONNECTION; - goto drop; - } - - if (dccp_bad_service_code(sk, service)) { - reset_code = DCCP_RESET_CODE_BAD_SERVICE_CODE; - goto drop; - } - /* - * TW buckets are converted to open requests without - * limitations, they conserve resources and peer is - * evidently real one. - */ - if (inet_csk_reqsk_queue_is_full(sk)) - goto drop; - - /* - * Accept backlog is full. If we have already queued enough - * of warm entries in syn queue, drop request. It is better than - * clogging syn queue with openreqs with exponentially increasing - * timeout. - */ - if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1) - goto drop; - - req = reqsk_alloc(&dccp_request_sock_ops); - if (req == NULL) - goto drop; - - if (dccp_parse_options(sk, skb)) - goto drop_and_free; - - dccp_openreq_init(req, &dp, skb); - - if (security_inet_conn_request(sk, skb, req)) - goto drop_and_free; - - ireq = inet_rsk(req); - ireq->loc_addr = daddr; - ireq->rmt_addr = saddr; - req->rcv_wnd = dccp_feat_default_sequence_window; - ireq->opt = NULL; - - /* - * Step 3: Process LISTEN state - * - * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie - * - * In fact we defer setting S.GSR, S.SWL, S.SWH to - * dccp_create_openreq_child. - */ - dreq = dccp_rsk(req); - dreq->dreq_isr = dcb->dccpd_seq; - dreq->dreq_iss = dccp_v4_init_sequence(sk, skb); - dreq->dreq_service = service; - - if (dccp_v4_send_response(sk, req, NULL)) - goto drop_and_free; - - inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT); - return 0; - -drop_and_free: - reqsk_free(req); -drop: - DCCP_INC_STATS_BH(DCCP_MIB_ATTEMPTFAILS); - dcb->dccpd_reset_code = reset_code; - return -1; -} - -EXPORT_SYMBOL_GPL(dccp_v4_conn_request); - /* * The three way handshake has completed - we got a valid ACK or DATAACK - * now create the new socket. @@ -623,47 +454,6 @@ static struct sock *dccp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) return sk; } -int dccp_v4_checksum(const struct sk_buff *skb, const __be32 saddr, - const __be32 daddr) -{ - const struct dccp_hdr* dh = dccp_hdr(skb); - int checksum_len; - u32 tmp; - - if (dh->dccph_cscov == 0) - checksum_len = skb->len; - else { - checksum_len = (dh->dccph_cscov + dh->dccph_x) * sizeof(u32); - checksum_len = checksum_len < skb->len ? checksum_len : - skb->len; - } - - tmp = csum_partial((unsigned char *)dh, checksum_len, 0); - return csum_tcpudp_magic(saddr, daddr, checksum_len, - IPPROTO_DCCP, tmp); -} - -EXPORT_SYMBOL_GPL(dccp_v4_checksum); - -static int dccp_v4_verify_checksum(struct sk_buff *skb, - const __be32 saddr, const __be32 daddr) -{ - struct dccp_hdr *dh = dccp_hdr(skb); - int checksum_len; - u32 tmp; - - if (dh->dccph_cscov == 0) - checksum_len = skb->len; - else { - checksum_len = (dh->dccph_cscov + dh->dccph_x) * sizeof(u32); - checksum_len = checksum_len < skb->len ? checksum_len : - skb->len; - } - tmp = csum_partial((unsigned char *)dh, checksum_len, 0); - return csum_tcpudp_magic(saddr, daddr, checksum_len, - IPPROTO_DCCP, tmp) == 0 ? 0 : -1; -} - static struct dst_entry* dccp_v4_route_skb(struct sock *sk, struct sk_buff *skb) { @@ -677,7 +467,7 @@ static struct dst_entry* dccp_v4_route_skb(struct sock *sk, .uli_u = { .ports = { .sport = dccp_hdr(skb)->dccph_dport, .dport = dccp_hdr(skb)->dccph_sport } - } + } }; security_skb_classify_flow(skb, &fl); @@ -689,7 +479,37 @@ static struct dst_entry* dccp_v4_route_skb(struct sock *sk, return &rt->u.dst; } -static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb) +static int dccp_v4_send_response(struct sock *sk, struct request_sock *req, + struct dst_entry *dst) +{ + int err = -1; + struct sk_buff *skb; + + /* First, grab a route. */ + + if (dst == NULL && (dst = inet_csk_route_req(sk, req)) == NULL) + goto out; + + skb = dccp_make_response(sk, dst, req); + if (skb != NULL) { + const struct inet_request_sock *ireq = inet_rsk(req); + struct dccp_hdr *dh = dccp_hdr(skb); + + dh->dccph_checksum = dccp_v4_csum_finish(skb, ireq->loc_addr, + ireq->rmt_addr); + memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); + err = ip_build_and_send_pkt(skb, sk, ireq->loc_addr, + ireq->rmt_addr, + ireq->opt); + err = net_xmit_eval(err); + } + +out: + dst_release(dst); + return err; +} + +static void dccp_v4_ctl_send_reset(struct sock *sk, struct sk_buff *rxskb) { int err; struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh; @@ -698,7 +518,7 @@ static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb) sizeof(struct dccp_hdr_reset); struct sk_buff *skb; struct dst_entry *dst; - u64 seqno; + u64 seqno = 0; /* Never send a reset in response to a reset. */ if (rxdh->dccph_type == DCCP_PKT_RESET) @@ -720,9 +540,7 @@ static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb) skb_reserve(skb, dccp_v4_ctl_socket->sk->sk_prot->max_header); skb->dst = dst_clone(dst); - skb->h.raw = skb_push(skb, dccp_hdr_reset_len); - dh = dccp_hdr(skb); - memset(dh, 0, dccp_hdr_reset_len); + dh = dccp_zeroed_hdr(skb, dccp_hdr_reset_len); /* Build DCCP header and checksum it. */ dh->dccph_type = DCCP_PKT_RESET; @@ -734,16 +552,15 @@ static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb) DCCP_SKB_CB(rxskb)->dccpd_reset_code; /* See "8.3.1. Abnormal Termination" in RFC 4340 */ - seqno = 0; if (DCCP_SKB_CB(rxskb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ) dccp_set_seqno(&seqno, DCCP_SKB_CB(rxskb)->dccpd_ack_seq + 1); dccp_hdr_set_seq(dh, seqno); - dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), - DCCP_SKB_CB(rxskb)->dccpd_seq); + dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), DCCP_SKB_CB(rxskb)->dccpd_seq); - dh->dccph_checksum = dccp_v4_checksum(skb, rxskb->nh.iph->saddr, - rxskb->nh.iph->daddr); + dccp_csum_outgoing(skb); + dh->dccph_checksum = dccp_v4_csum_finish(skb, rxskb->nh.iph->saddr, + rxskb->nh.iph->daddr); bh_lock_sock(dccp_v4_ctl_socket->sk); err = ip_build_and_send_pkt(skb, dccp_v4_ctl_socket->sk, @@ -751,7 +568,7 @@ static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb) rxskb->nh.iph->saddr, NULL); bh_unlock_sock(dccp_v4_ctl_socket->sk); - if (err == NET_XMIT_CN || err == 0) { + if (net_xmit_eval(err) == 0) { DCCP_INC_STATS_BH(DCCP_MIB_OUTSEGS); DCCP_INC_STATS_BH(DCCP_MIB_OUTRSTS); } @@ -759,6 +576,103 @@ out: dst_release(dst); } +static void dccp_v4_reqsk_destructor(struct request_sock *req) +{ + kfree(inet_rsk(req)->opt); +} + +static struct request_sock_ops dccp_request_sock_ops __read_mostly = { + .family = PF_INET, + .obj_size = sizeof(struct dccp_request_sock), + .rtx_syn_ack = dccp_v4_send_response, + .send_ack = dccp_reqsk_send_ack, + .destructor = dccp_v4_reqsk_destructor, + .send_reset = dccp_v4_ctl_send_reset, +}; + +int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb) +{ + struct inet_request_sock *ireq; + struct request_sock *req; + struct dccp_request_sock *dreq; + const __be32 service = dccp_hdr_request(skb)->dccph_req_service; + struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb); + __u8 reset_code = DCCP_RESET_CODE_TOO_BUSY; + + /* Never answer to DCCP_PKT_REQUESTs send to broadcast or multicast */ + if (((struct rtable *)skb->dst)->rt_flags & + (RTCF_BROADCAST | RTCF_MULTICAST)) { + reset_code = DCCP_RESET_CODE_NO_CONNECTION; + goto drop; + } + + if (dccp_bad_service_code(sk, service)) { + reset_code = DCCP_RESET_CODE_BAD_SERVICE_CODE; + goto drop; + } + /* + * TW buckets are converted to open requests without + * limitations, they conserve resources and peer is + * evidently real one. + */ + if (inet_csk_reqsk_queue_is_full(sk)) + goto drop; + + /* + * Accept backlog is full. If we have already queued enough + * of warm entries in syn queue, drop request. It is better than + * clogging syn queue with openreqs with exponentially increasing + * timeout. + */ + if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1) + goto drop; + + req = reqsk_alloc(&dccp_request_sock_ops); + if (req == NULL) + goto drop; + + if (dccp_parse_options(sk, skb)) + goto drop_and_free; + + dccp_reqsk_init(req, skb); + + if (security_inet_conn_request(sk, skb, req)) + goto drop_and_free; + + ireq = inet_rsk(req); + ireq->loc_addr = skb->nh.iph->daddr; + ireq->rmt_addr = skb->nh.iph->saddr; + ireq->opt = NULL; + + /* + * Step 3: Process LISTEN state + * + * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie + * + * In fact we defer setting S.GSR, S.SWL, S.SWH to + * dccp_create_openreq_child. + */ + dreq = dccp_rsk(req); + dreq->dreq_isr = dcb->dccpd_seq; + dreq->dreq_iss = dccp_v4_init_sequence(skb); + dreq->dreq_service = service; + + if (dccp_v4_send_response(sk, req, NULL)) + goto drop_and_free; + + inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT); + return 0; + +drop_and_free: + reqsk_free(req); +drop: + DCCP_INC_STATS_BH(DCCP_MIB_ATTEMPTFAILS); + dcb->dccpd_reset_code = reset_code; + return -1; +} + +EXPORT_SYMBOL_GPL(dccp_v4_conn_request); + int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) { struct dccp_hdr *dh = dccp_hdr(skb); @@ -771,24 +685,23 @@ int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) /* * Step 3: Process LISTEN state - * If S.state == LISTEN, - * If P.type == Request or P contains a valid Init Cookie - * option, - * * Must scan the packet's options to check for an Init - * Cookie. Only the Init Cookie is processed here, - * however; other options are processed in Step 8. This - * scan need only be performed if the endpoint uses Init - * Cookies * - * * Generate a new socket and switch to that socket * - * Set S := new socket for this port pair - * S.state = RESPOND - * Choose S.ISS (initial seqno) or set from Init Cookie - * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie - * Continue with S.state == RESPOND - * * A Response packet will be generated in Step 11 * - * Otherwise, - * Generate Reset(No Connection) unless P.type == Reset - * Drop packet and return + * If P.type == Request or P contains a valid Init Cookie option, + * (* Must scan the packet's options to check for Init + * Cookies. Only Init Cookies are processed here, + * however; other options are processed in Step 8. This + * scan need only be performed if the endpoint uses Init + * Cookies *) + * (* Generate a new socket and switch to that socket *) + * Set S := new socket for this port pair + * S.state = RESPOND + * Choose S.ISS (initial seqno) or set from Init Cookies + * Initialize S.GAR := S.ISS + * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookies + * Continue with S.state == RESPOND + * (* A Response packet will be generated in Step 11 *) + * Otherwise, + * Generate Reset(No Connection) unless P.type == Reset + * Drop packet and return * * NOTE: the check for the packet types is done in * dccp_rcv_state_process @@ -811,7 +724,7 @@ int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) return 0; reset: - dccp_v4_ctl_send_reset(skb); + dccp_v4_ctl_send_reset(sk, skb); discard: kfree_skb(skb); return 0; @@ -819,60 +732,74 @@ discard: EXPORT_SYMBOL_GPL(dccp_v4_do_rcv); +/** + * dccp_invalid_packet - check for malformed packets + * Implements RFC 4340, 8.5: Step 1: Check header basics + * Packets that fail these checks are ignored and do not receive Resets. + */ int dccp_invalid_packet(struct sk_buff *skb) { const struct dccp_hdr *dh; + unsigned int cscov; if (skb->pkt_type != PACKET_HOST) return 1; + /* If the packet is shorter than 12 bytes, drop packet and return */ if (!pskb_may_pull(skb, sizeof(struct dccp_hdr))) { - LIMIT_NETDEBUG(KERN_WARNING "DCCP: pskb_may_pull failed\n"); + DCCP_WARN("pskb_may_pull failed\n"); return 1; } dh = dccp_hdr(skb); - /* If the packet type is not understood, drop packet and return */ + /* If P.type is not understood, drop packet and return */ if (dh->dccph_type >= DCCP_PKT_INVALID) { - LIMIT_NETDEBUG(KERN_WARNING "DCCP: invalid packet type\n"); + DCCP_WARN("invalid packet type\n"); return 1; } /* - * If P.Data Offset is too small for packet type, or too large for - * packet, drop packet and return + * If P.Data Offset is too small for packet type, drop packet and return */ if (dh->dccph_doff < dccp_hdr_len(skb) / sizeof(u32)) { - LIMIT_NETDEBUG(KERN_WARNING "DCCP: P.Data Offset(%u) " - "too small 1\n", - dh->dccph_doff); + DCCP_WARN("P.Data Offset(%u) too small\n", dh->dccph_doff); return 1; } - + /* + * If P.Data Offset is too too large for packet, drop packet and return + */ if (!pskb_may_pull(skb, dh->dccph_doff * sizeof(u32))) { - LIMIT_NETDEBUG(KERN_WARNING "DCCP: P.Data Offset(%u) " - "too small 2\n", - dh->dccph_doff); + DCCP_WARN("P.Data Offset(%u) too large\n", dh->dccph_doff); return 1; } - dh = dccp_hdr(skb); - /* * If P.type is not Data, Ack, or DataAck and P.X == 0 (the packet * has short sequence numbers), drop packet and return */ - if (dh->dccph_x == 0 && - dh->dccph_type != DCCP_PKT_DATA && - dh->dccph_type != DCCP_PKT_ACK && - dh->dccph_type != DCCP_PKT_DATAACK) { - LIMIT_NETDEBUG(KERN_WARNING "DCCP: P.type (%s) not Data, Ack " - "nor DataAck and P.X == 0\n", - dccp_packet_name(dh->dccph_type)); + if (dh->dccph_type >= DCCP_PKT_DATA && + dh->dccph_type <= DCCP_PKT_DATAACK && dh->dccph_x == 0) { + DCCP_WARN("P.type (%s) not Data || [Data]Ack, while P.X == 0\n", + dccp_packet_name(dh->dccph_type)); return 1; } + /* + * If P.CsCov is too large for the packet size, drop packet and return. + * This must come _before_ checksumming (not as RFC 4340 suggests). + */ + cscov = dccp_csum_coverage(skb); + if (cscov > skb->len) { + DCCP_WARN("P.CsCov %u exceeds packet length %d\n", + dh->dccph_cscov, skb->len); + return 1; + } + + /* If header checksum is incorrect, drop packet and return. + * (This step is completed in the AF-dependent functions.) */ + skb->csum = skb_checksum(skb, 0, cscov, 0); + return 0; } @@ -883,17 +810,16 @@ static int dccp_v4_rcv(struct sk_buff *skb) { const struct dccp_hdr *dh; struct sock *sk; + int min_cov; - /* Step 1: Check header basics: */ + /* Step 1: Check header basics */ if (dccp_invalid_packet(skb)) goto discard_it; - /* If the header checksum is incorrect, drop packet and return */ - if (dccp_v4_verify_checksum(skb, skb->nh.iph->saddr, - skb->nh.iph->daddr) < 0) { - LIMIT_NETDEBUG(KERN_WARNING "%s: incorrect header checksum\n", - __FUNCTION__); + /* Step 1: If header checksum is incorrect, drop packet and return */ + if (dccp_v4_csum_finish(skb, skb->nh.iph->saddr, skb->nh.iph->daddr)) { + DCCP_WARN("dropped packet with invalid checksum\n"); goto discard_it; } @@ -915,23 +841,20 @@ static int dccp_v4_rcv(struct sk_buff *skb) dccp_pr_debug_cat("\n"); } else { DCCP_SKB_CB(skb)->dccpd_ack_seq = dccp_hdr_ack_seq(skb); - dccp_pr_debug_cat(", ack=%llu\n", - (unsigned long long) + dccp_pr_debug_cat(", ack=%llu\n", (unsigned long long) DCCP_SKB_CB(skb)->dccpd_ack_seq); } /* Step 2: - * Look up flow ID in table and get corresponding socket */ + * Look up flow ID in table and get corresponding socket */ sk = __inet_lookup(&dccp_hashinfo, skb->nh.iph->saddr, dh->dccph_sport, skb->nh.iph->daddr, dh->dccph_dport, inet_iif(skb)); - /* + /* * Step 2: - * If no socket ... - * Generate Reset(No Connection) unless P.type == Reset - * Drop packet and return + * If no socket ... */ if (sk == NULL) { dccp_pr_debug("failed to look up flow ID in table and " @@ -939,51 +862,61 @@ static int dccp_v4_rcv(struct sk_buff *skb) goto no_dccp_socket; } - /* + /* * Step 2: - * ... or S.state == TIMEWAIT, + * ... or S.state == TIMEWAIT, * Generate Reset(No Connection) unless P.type == Reset * Drop packet and return */ - if (sk->sk_state == DCCP_TIME_WAIT) { - dccp_pr_debug("sk->sk_state == DCCP_TIME_WAIT: " - "do_time_wait\n"); - goto do_time_wait; + dccp_pr_debug("sk->sk_state == DCCP_TIME_WAIT: do_time_wait\n"); + inet_twsk_put(inet_twsk(sk)); + goto no_dccp_socket; + } + + /* + * RFC 4340, sec. 9.2.1: Minimum Checksum Coverage + * o if MinCsCov = 0, only packets with CsCov = 0 are accepted + * o if MinCsCov > 0, also accept packets with CsCov >= MinCsCov + */ + min_cov = dccp_sk(sk)->dccps_pcrlen; + if (dh->dccph_cscov && (min_cov == 0 || dh->dccph_cscov < min_cov)) { + dccp_pr_debug("Packet CsCov %d does not satisfy MinCsCov %d\n", + dh->dccph_cscov, min_cov); + /* FIXME: "Such packets SHOULD be reported using Data Dropped + * options (Section 11.7) with Drop Code 0, Protocol + * Constraints." */ + goto discard_and_relse; } if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb)) goto discard_and_relse; nf_reset(skb); - return sk_receive_skb(sk, skb); + return sk_receive_skb(sk, skb, 1); no_dccp_socket: if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) goto discard_it; /* * Step 2: + * If no socket ... * Generate Reset(No Connection) unless P.type == Reset * Drop packet and return */ if (dh->dccph_type != DCCP_PKT_RESET) { DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_NO_CONNECTION; - dccp_v4_ctl_send_reset(skb); + dccp_v4_ctl_send_reset(sk, skb); } discard_it: - /* Discard frame. */ kfree_skb(skb); return 0; discard_and_relse: sock_put(sk); goto discard_it; - -do_time_wait: - inet_twsk_put(inet_twsk(sk)); - goto no_dccp_socket; } static struct inet_connection_sock_af_ops dccp_ipv4_af_ops = { @@ -1017,20 +950,6 @@ static int dccp_v4_init_sock(struct sock *sk) return err; } -static void dccp_v4_reqsk_destructor(struct request_sock *req) -{ - kfree(inet_rsk(req)->opt); -} - -static struct request_sock_ops dccp_request_sock_ops = { - .family = PF_INET, - .obj_size = sizeof(struct dccp_request_sock), - .rtx_syn_ack = dccp_v4_send_response, - .send_ack = dccp_v4_reqsk_send_ack, - .destructor = dccp_v4_reqsk_destructor, - .send_reset = dccp_v4_ctl_send_reset, -}; - static struct timewait_sock_ops dccp_timewait_sock_ops = { .twsk_obj_size = sizeof(struct inet_timewait_sock), }; diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index fc4242c0767..6b91a9dd041 100644 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@ -36,13 +36,6 @@ /* Socket used for sending RSTs and ACKs */ static struct socket *dccp_v6_ctl_socket; -static void dccp_v6_ctl_send_reset(struct sk_buff *skb); -static void dccp_v6_reqsk_send_ack(struct sk_buff *skb, - struct request_sock *req); -static void dccp_v6_send_check(struct sock *sk, int len, struct sk_buff *skb); - -static int dccp_v6_do_rcv(struct sock *sk, struct sk_buff *skb); - static struct inet_connection_sock_af_ops dccp_ipv6_mapped; static struct inet_connection_sock_af_ops dccp_ipv6_af_ops; @@ -65,205 +58,37 @@ static void dccp_v6_hash(struct sock *sk) } } -static inline u16 dccp_v6_check(struct dccp_hdr *dh, int len, - struct in6_addr *saddr, - struct in6_addr *daddr, - unsigned long base) +/* add pseudo-header to DCCP checksum stored in skb->csum */ +static inline __sum16 dccp_v6_csum_finish(struct sk_buff *skb, + struct in6_addr *saddr, + struct in6_addr *daddr) { - return csum_ipv6_magic(saddr, daddr, len, IPPROTO_DCCP, base); + return csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_DCCP, skb->csum); } -static __u32 dccp_v6_init_sequence(struct sock *sk, struct sk_buff *skb) +static inline void dccp_v6_send_check(struct sock *sk, int unused_value, + struct sk_buff *skb) { - const struct dccp_hdr *dh = dccp_hdr(skb); - - if (skb->protocol == htons(ETH_P_IPV6)) - return secure_tcpv6_sequence_number(skb->nh.ipv6h->daddr.s6_addr32, - skb->nh.ipv6h->saddr.s6_addr32, - dh->dccph_dport, - dh->dccph_sport); + struct ipv6_pinfo *np = inet6_sk(sk); + struct dccp_hdr *dh = dccp_hdr(skb); - return secure_dccp_sequence_number(skb->nh.iph->daddr, - skb->nh.iph->saddr, - dh->dccph_dport, - dh->dccph_sport); + dccp_csum_outgoing(skb); + dh->dccph_checksum = dccp_v6_csum_finish(skb, &np->saddr, &np->daddr); } -static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr, - int addr_len) +static inline __u32 secure_dccpv6_sequence_number(__be32 *saddr, __be32 *daddr, + __be16 sport, __be16 dport ) { - struct sockaddr_in6 *usin = (struct sockaddr_in6 *)uaddr; - struct inet_connection_sock *icsk = inet_csk(sk); - struct inet_sock *inet = inet_sk(sk); - struct ipv6_pinfo *np = inet6_sk(sk); - struct dccp_sock *dp = dccp_sk(sk); - struct in6_addr *saddr = NULL, *final_p = NULL, final; - struct flowi fl; - struct dst_entry *dst; - int addr_type; - int err; - - dp->dccps_role = DCCP_ROLE_CLIENT; - - if (addr_len < SIN6_LEN_RFC2133) - return -EINVAL; - - if (usin->sin6_family != AF_INET6) - return -EAFNOSUPPORT; - - memset(&fl, 0, sizeof(fl)); - - if (np->sndflow) { - fl.fl6_flowlabel = usin->sin6_flowinfo & IPV6_FLOWINFO_MASK; - IP6_ECN_flow_init(fl.fl6_flowlabel); - if (fl.fl6_flowlabel & IPV6_FLOWLABEL_MASK) { - struct ip6_flowlabel *flowlabel; - flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel); - if (flowlabel == NULL) - return -EINVAL; - ipv6_addr_copy(&usin->sin6_addr, &flowlabel->dst); - fl6_sock_release(flowlabel); - } - } - /* - * connect() to INADDR_ANY means loopback (BSD'ism). - */ - if (ipv6_addr_any(&usin->sin6_addr)) - usin->sin6_addr.s6_addr[15] = 1; - - addr_type = ipv6_addr_type(&usin->sin6_addr); - - if (addr_type & IPV6_ADDR_MULTICAST) - return -ENETUNREACH; - - if (addr_type & IPV6_ADDR_LINKLOCAL) { - if (addr_len >= sizeof(struct sockaddr_in6) && - usin->sin6_scope_id) { - /* If interface is set while binding, indices - * must coincide. - */ - if (sk->sk_bound_dev_if && - sk->sk_bound_dev_if != usin->sin6_scope_id) - return -EINVAL; - - sk->sk_bound_dev_if = usin->sin6_scope_id; - } - - /* Connect to link-local address requires an interface */ - if (!sk->sk_bound_dev_if) - return -EINVAL; - } - - ipv6_addr_copy(&np->daddr, &usin->sin6_addr); - np->flow_label = fl.fl6_flowlabel; - - /* - * DCCP over IPv4 - */ - if (addr_type == IPV6_ADDR_MAPPED) { - u32 exthdrlen = icsk->icsk_ext_hdr_len; - struct sockaddr_in sin; - - SOCK_DEBUG(sk, "connect: ipv4 mapped\n"); - - if (__ipv6_only_sock(sk)) - return -ENETUNREACH; - - sin.sin_family = AF_INET; - sin.sin_port = usin->sin6_port; - sin.sin_addr.s_addr = usin->sin6_addr.s6_addr32[3]; - - icsk->icsk_af_ops = &dccp_ipv6_mapped; - sk->sk_backlog_rcv = dccp_v4_do_rcv; - - err = dccp_v4_connect(sk, (struct sockaddr *)&sin, sizeof(sin)); - if (err) { - icsk->icsk_ext_hdr_len = exthdrlen; - icsk->icsk_af_ops = &dccp_ipv6_af_ops; - sk->sk_backlog_rcv = dccp_v6_do_rcv; - goto failure; - } else { - ipv6_addr_set(&np->saddr, 0, 0, htonl(0x0000FFFF), - inet->saddr); - ipv6_addr_set(&np->rcv_saddr, 0, 0, htonl(0x0000FFFF), - inet->rcv_saddr); - } - - return err; - } - - if (!ipv6_addr_any(&np->rcv_saddr)) - saddr = &np->rcv_saddr; - - fl.proto = IPPROTO_DCCP; - ipv6_addr_copy(&fl.fl6_dst, &np->daddr); - ipv6_addr_copy(&fl.fl6_src, saddr ? saddr : &np->saddr); - fl.oif = sk->sk_bound_dev_if; - fl.fl_ip_dport = usin->sin6_port; - fl.fl_ip_sport = inet->sport; - security_sk_classify_flow(sk, &fl); - - if (np->opt != NULL && np->opt->srcrt != NULL) { - const struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt; - - ipv6_addr_copy(&final, &fl.fl6_dst); - ipv6_addr_copy(&fl.fl6_dst, rt0->addr); - final_p = &final; - } - - err = ip6_dst_lookup(sk, &dst, &fl); - if (err) - goto failure; - - if (final_p) - ipv6_addr_copy(&fl.fl6_dst, final_p); - - err = xfrm_lookup(&dst, &fl, sk, 0); - if (err < 0) - goto failure; - - if (saddr == NULL) { - saddr = &fl.fl6_src; - ipv6_addr_copy(&np->rcv_saddr, saddr); - } - - /* set the source address */ - ipv6_addr_copy(&np->saddr, saddr); - inet->rcv_saddr = LOOPBACK4_IPV6; - - __ip6_dst_store(sk, dst, NULL, NULL); - - icsk->icsk_ext_hdr_len = 0; - if (np->opt != NULL) - icsk->icsk_ext_hdr_len = (np->opt->opt_flen + - np->opt->opt_nflen); - - inet->dport = usin->sin6_port; - - dccp_set_state(sk, DCCP_REQUESTING); - err = inet6_hash_connect(&dccp_death_row, sk); - if (err) - goto late_failure; - /* FIXME */ -#if 0 - dp->dccps_gar = secure_dccp_v6_sequence_number(np->saddr.s6_addr32, - np->daddr.s6_addr32, - inet->sport, - inet->dport); -#endif - err = dccp_connect(sk); - if (err) - goto late_failure; + return secure_tcpv6_sequence_number(saddr, daddr, sport, dport); +} - return 0; +static inline __u32 dccp_v6_init_sequence(struct sk_buff *skb) +{ + return secure_dccpv6_sequence_number(skb->nh.ipv6h->daddr.s6_addr32, + skb->nh.ipv6h->saddr.s6_addr32, + dccp_hdr(skb)->dccph_dport, + dccp_hdr(skb)->dccph_sport ); -late_failure: - dccp_set_state(sk, DCCP_CLOSED); - __sk_dst_reset(sk); -failure: - inet->dport = 0; - sk->sk_route_caps = 0; - return err; } static void dccp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, @@ -464,16 +289,12 @@ static int dccp_v6_send_response(struct sock *sk, struct request_sock *req, if (skb != NULL) { struct dccp_hdr *dh = dccp_hdr(skb); - dh->dccph_checksum = dccp_v6_check(dh, skb->len, - &ireq6->loc_addr, - &ireq6->rmt_addr, - csum_partial((char *)dh, - skb->len, - skb->csum)); + dh->dccph_checksum = dccp_v6_csum_finish(skb, + &ireq6->loc_addr, + &ireq6->rmt_addr); ipv6_addr_copy(&fl.fl6_dst, &ireq6->rmt_addr); err = ip6_xmit(sk, skb, &fl, opt, 0); - if (err == NET_XMIT_CN) - err = 0; + err = net_xmit_eval(err); } done: @@ -489,32 +310,7 @@ static void dccp_v6_reqsk_destructor(struct request_sock *req) kfree_skb(inet6_rsk(req)->pktopts); } -static struct request_sock_ops dccp6_request_sock_ops = { - .family = AF_INET6, - .obj_size = sizeof(struct dccp6_request_sock), - .rtx_syn_ack = dccp_v6_send_response, - .send_ack = dccp_v6_reqsk_send_ack, - .destructor = dccp_v6_reqsk_destructor, - .send_reset = dccp_v6_ctl_send_reset, -}; - -static struct timewait_sock_ops dccp6_timewait_sock_ops = { - .twsk_obj_size = sizeof(struct dccp6_timewait_sock), -}; - -static void dccp_v6_send_check(struct sock *sk, int len, struct sk_buff *skb) -{ - struct ipv6_pinfo *np = inet6_sk(sk); - struct dccp_hdr *dh = dccp_hdr(skb); - - dh->dccph_checksum = csum_ipv6_magic(&np->saddr, &np->daddr, - len, IPPROTO_DCCP, - csum_partial((char *)dh, - dh->dccph_doff << 2, - skb->csum)); -} - -static void dccp_v6_ctl_send_reset(struct sk_buff *rxskb) +static void dccp_v6_ctl_send_reset(struct sock *sk, struct sk_buff *rxskb) { struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh; const u32 dccp_hdr_reset_len = sizeof(struct dccp_hdr) + @@ -522,7 +318,7 @@ static void dccp_v6_ctl_send_reset(struct sk_buff *rxskb) sizeof(struct dccp_hdr_reset); struct sk_buff *skb; struct flowi fl; - u64 seqno; + u64 seqno = 0; if (rxdh->dccph_type == DCCP_PKT_RESET) return; @@ -533,13 +329,11 @@ static void dccp_v6_ctl_send_reset(struct sk_buff *rxskb) skb = alloc_skb(dccp_v6_ctl_socket->sk->sk_prot->max_header, GFP_ATOMIC); if (skb == NULL) - return; + return; skb_reserve(skb, dccp_v6_ctl_socket->sk->sk_prot->max_header); - skb->h.raw = skb_push(skb, dccp_hdr_reset_len); - dh = dccp_hdr(skb); - memset(dh, 0, dccp_hdr_reset_len); + dh = dccp_zeroed_hdr(skb, dccp_hdr_reset_len); /* Swap the send and the receive. */ dh->dccph_type = DCCP_PKT_RESET; @@ -551,20 +345,20 @@ static void dccp_v6_ctl_send_reset(struct sk_buff *rxskb) DCCP_SKB_CB(rxskb)->dccpd_reset_code; /* See "8.3.1. Abnormal Termination" in RFC 4340 */ - seqno = 0; if (DCCP_SKB_CB(rxskb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ) dccp_set_seqno(&seqno, DCCP_SKB_CB(rxskb)->dccpd_ack_seq + 1); dccp_hdr_set_seq(dh, seqno); - dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), - DCCP_SKB_CB(rxskb)->dccpd_seq); + dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), DCCP_SKB_CB(rxskb)->dccpd_seq); + + dccp_csum_outgoing(skb); + dh->dccph_checksum = dccp_v6_csum_finish(skb, &rxskb->nh.ipv6h->saddr, + &rxskb->nh.ipv6h->daddr); memset(&fl, 0, sizeof(fl)); ipv6_addr_copy(&fl.fl6_dst, &rxskb->nh.ipv6h->saddr); ipv6_addr_copy(&fl.fl6_src, &rxskb->nh.ipv6h->daddr); - dh->dccph_checksum = csum_ipv6_magic(&fl.fl6_src, &fl.fl6_dst, - sizeof(*dh), IPPROTO_DCCP, - skb->csum); + fl.proto = IPPROTO_DCCP; fl.oif = inet6_iif(rxskb); fl.fl_ip_dport = dh->dccph_dport; @@ -584,60 +378,14 @@ static void dccp_v6_ctl_send_reset(struct sk_buff *rxskb) kfree_skb(skb); } -static void dccp_v6_reqsk_send_ack(struct sk_buff *rxskb, - struct request_sock *req) -{ - struct flowi fl; - struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh; - const u32 dccp_hdr_ack_len = sizeof(struct dccp_hdr) + - sizeof(struct dccp_hdr_ext) + - sizeof(struct dccp_hdr_ack_bits); - struct sk_buff *skb; - - skb = alloc_skb(dccp_v6_ctl_socket->sk->sk_prot->max_header, - GFP_ATOMIC); - if (skb == NULL) - return; - - skb_reserve(skb, dccp_v6_ctl_socket->sk->sk_prot->max_header); - - skb->h.raw = skb_push(skb, dccp_hdr_ack_len); - dh = dccp_hdr(skb); - memset(dh, 0, dccp_hdr_ack_len); - - /* Build DCCP header and checksum it. */ - dh->dccph_type = DCCP_PKT_ACK; - dh->dccph_sport = rxdh->dccph_dport; - dh->dccph_dport = rxdh->dccph_sport; - dh->dccph_doff = dccp_hdr_ack_len / 4; - dh->dccph_x = 1; - - dccp_hdr_set_seq(dh, DCCP_SKB_CB(rxskb)->dccpd_ack_seq); - dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), - DCCP_SKB_CB(rxskb)->dccpd_seq); - - memset(&fl, 0, sizeof(fl)); - ipv6_addr_copy(&fl.fl6_dst, &rxskb->nh.ipv6h->saddr); - ipv6_addr_copy(&fl.fl6_src, &rxskb->nh.ipv6h->daddr); - - /* FIXME: calculate checksum, IPv4 also should... */ - - fl.proto = IPPROTO_DCCP; - fl.oif = inet6_iif(rxskb); - fl.fl_ip_dport = dh->dccph_dport; - fl.fl_ip_sport = dh->dccph_sport; - security_req_classify_flow(req, &fl); - - if (!ip6_dst_lookup(NULL, &skb->dst, &fl)) { - if (xfrm_lookup(&skb->dst, &fl, NULL, 0) >= 0) { - ip6_xmit(dccp_v6_ctl_socket->sk, skb, &fl, NULL, 0); - DCCP_INC_STATS_BH(DCCP_MIB_OUTSEGS); - return; - } - } - - kfree_skb(skb); -} +static struct request_sock_ops dccp6_request_sock_ops = { + .family = AF_INET6, + .obj_size = sizeof(struct dccp6_request_sock), + .rtx_syn_ack = dccp_v6_send_response, + .send_ack = dccp_reqsk_send_ack, + .destructor = dccp_v6_reqsk_destructor, + .send_reset = dccp_v6_ctl_send_reset, +}; static struct sock *dccp_v6_hnd_req(struct sock *sk,struct sk_buff *skb) { @@ -672,12 +420,11 @@ static struct sock *dccp_v6_hnd_req(struct sock *sk,struct sk_buff *skb) static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb) { - struct dccp_sock dp; struct request_sock *req; struct dccp_request_sock *dreq; struct inet6_request_sock *ireq6; struct ipv6_pinfo *np = inet6_sk(sk); - const __be32 service = dccp_hdr_request(skb)->dccph_req_service; + const __be32 service = dccp_hdr_request(skb)->dccph_req_service; struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb); __u8 reset_code = DCCP_RESET_CODE_TOO_BUSY; @@ -690,7 +437,7 @@ static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb) if (dccp_bad_service_code(sk, service)) { reset_code = DCCP_RESET_CODE_BAD_SERVICE_CODE; goto drop; - } + } /* * There are no SYN attacks on IPv6, yet... */ @@ -704,9 +451,10 @@ static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb) if (req == NULL) goto drop; - /* FIXME: process options */ + if (dccp_parse_options(sk, skb)) + goto drop_and_free; - dccp_openreq_init(req, &dp, skb); + dccp_reqsk_init(req, skb); if (security_inet_conn_request(sk, skb, req)) goto drop_and_free; @@ -714,7 +462,6 @@ static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb) ireq6 = inet6_rsk(req); ipv6_addr_copy(&ireq6->rmt_addr, &skb->nh.ipv6h->saddr); ipv6_addr_copy(&ireq6->loc_addr, &skb->nh.ipv6h->daddr); - req->rcv_wnd = dccp_feat_default_sequence_window; ireq6->pktopts = NULL; if (ipv6_opt_accepted(sk, skb) || @@ -733,14 +480,14 @@ static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb) /* * Step 3: Process LISTEN state * - * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie + * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie * - * In fact we defer setting S.GSR, S.SWL, S.SWH to - * dccp_create_openreq_child. + * In fact we defer setting S.GSR, S.SWL, S.SWH to + * dccp_create_openreq_child. */ dreq = dccp_rsk(req); dreq->dreq_isr = dcb->dccpd_seq; - dreq->dreq_iss = dccp_v6_init_sequence(sk, skb); + dreq->dreq_iss = dccp_v6_init_sequence(skb); dreq->dreq_service = service; if (dccp_v6_send_response(sk, req, NULL)) @@ -990,18 +737,46 @@ static int dccp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) --ANK (980728) */ if (np->rxopt.all) + /* + * FIXME: Add handling of IPV6_PKTOPTIONS skb. See the comments below + * (wrt ipv6_pktopions) and net/ipv6/tcp_ipv6.c for an example. + */ opt_skb = skb_clone(skb, GFP_ATOMIC); if (sk->sk_state == DCCP_OPEN) { /* Fast path */ if (dccp_rcv_established(sk, skb, dccp_hdr(skb), skb->len)) goto reset; if (opt_skb) { - /* This is where we would goto ipv6_pktoptions. */ + /* XXX This is where we would goto ipv6_pktoptions. */ __kfree_skb(opt_skb); } return 0; } + /* + * Step 3: Process LISTEN state + * If S.state == LISTEN, + * If P.type == Request or P contains a valid Init Cookie option, + * (* Must scan the packet's options to check for Init + * Cookies. Only Init Cookies are processed here, + * however; other options are processed in Step 8. This + * scan need only be performed if the endpoint uses Init + * Cookies *) + * (* Generate a new socket and switch to that socket *) + * Set S := new socket for this port pair + * S.state = RESPOND + * Choose S.ISS (initial seqno) or set from Init Cookies + * Initialize S.GAR := S.ISS + * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookies + * Continue with S.state == RESPOND + * (* A Response packet will be generated in Step 11 *) + * Otherwise, + * Generate Reset(No Connection) unless P.type == Reset + * Drop packet and return + * + * NOTE: the check for the packet types is done in + * dccp_rcv_state_process + */ if (sk->sk_state == DCCP_LISTEN) { struct sock *nsk = dccp_v6_hnd_req(sk, skb); @@ -1012,7 +787,7 @@ static int dccp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) * otherwise we just shortcircuit this and continue with * the new socket.. */ - if (nsk != sk) { + if (nsk != sk) { if (dccp_child_process(sk, nsk, skb)) goto reset; if (opt_skb != NULL) @@ -1024,13 +799,13 @@ static int dccp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) if (dccp_rcv_state_process(sk, skb, dccp_hdr(skb), skb->len)) goto reset; if (opt_skb) { - /* This is where we would goto ipv6_pktoptions. */ + /* XXX This is where we would goto ipv6_pktoptions. */ __kfree_skb(opt_skb); } return 0; reset: - dccp_v6_ctl_send_reset(skb); + dccp_v6_ctl_send_reset(sk, skb); discard: if (opt_skb != NULL) __kfree_skb(opt_skb); @@ -1043,12 +818,20 @@ static int dccp_v6_rcv(struct sk_buff **pskb) const struct dccp_hdr *dh; struct sk_buff *skb = *pskb; struct sock *sk; + int min_cov; - /* Step 1: Check header basics: */ + /* Step 1: Check header basics */ if (dccp_invalid_packet(skb)) goto discard_it; + /* Step 1: If header checksum is incorrect, drop packet and return. */ + if (dccp_v6_csum_finish(skb, &skb->nh.ipv6h->saddr, + &skb->nh.ipv6h->daddr)) { + DCCP_WARN("dropped packet with invalid checksum\n"); + goto discard_it; + } + dh = dccp_hdr(skb); DCCP_SKB_CB(skb)->dccpd_seq = dccp_hdr_seq(skb); @@ -1060,63 +843,247 @@ static int dccp_v6_rcv(struct sk_buff **pskb) DCCP_SKB_CB(skb)->dccpd_ack_seq = dccp_hdr_ack_seq(skb); /* Step 2: - * Look up flow ID in table and get corresponding socket */ + * Look up flow ID in table and get corresponding socket */ sk = __inet6_lookup(&dccp_hashinfo, &skb->nh.ipv6h->saddr, dh->dccph_sport, &skb->nh.ipv6h->daddr, ntohs(dh->dccph_dport), inet6_iif(skb)); /* * Step 2: - * If no socket ... - * Generate Reset(No Connection) unless P.type == Reset - * Drop packet and return + * If no socket ... */ - if (sk == NULL) + if (sk == NULL) { + dccp_pr_debug("failed to look up flow ID in table and " + "get corresponding socket\n"); goto no_dccp_socket; + } /* * Step 2: - * ... or S.state == TIMEWAIT, + * ... or S.state == TIMEWAIT, * Generate Reset(No Connection) unless P.type == Reset * Drop packet and return */ - if (sk->sk_state == DCCP_TIME_WAIT) - goto do_time_wait; + if (sk->sk_state == DCCP_TIME_WAIT) { + dccp_pr_debug("sk->sk_state == DCCP_TIME_WAIT: do_time_wait\n"); + inet_twsk_put(inet_twsk(sk)); + goto no_dccp_socket; + } + + /* + * RFC 4340, sec. 9.2.1: Minimum Checksum Coverage + * o if MinCsCov = 0, only packets with CsCov = 0 are accepted + * o if MinCsCov > 0, also accept packets with CsCov >= MinCsCov + */ + min_cov = dccp_sk(sk)->dccps_pcrlen; + if (dh->dccph_cscov && (min_cov == 0 || dh->dccph_cscov < min_cov)) { + dccp_pr_debug("Packet CsCov %d does not satisfy MinCsCov %d\n", + dh->dccph_cscov, min_cov); + /* FIXME: send Data Dropped option (see also dccp_v4_rcv) */ + goto discard_and_relse; + } if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb)) goto discard_and_relse; - return sk_receive_skb(sk, skb) ? -1 : 0; + return sk_receive_skb(sk, skb, 1) ? -1 : 0; no_dccp_socket: if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) goto discard_it; /* * Step 2: + * If no socket ... * Generate Reset(No Connection) unless P.type == Reset * Drop packet and return */ if (dh->dccph_type != DCCP_PKT_RESET) { DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_NO_CONNECTION; - dccp_v6_ctl_send_reset(skb); + dccp_v6_ctl_send_reset(sk, skb); } -discard_it: - - /* - * Discard frame - */ +discard_it: kfree_skb(skb); return 0; discard_and_relse: sock_put(sk); goto discard_it; +} -do_time_wait: - inet_twsk_put(inet_twsk(sk)); - goto no_dccp_socket; +static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr, + int addr_len) +{ + struct sockaddr_in6 *usin = (struct sockaddr_in6 *)uaddr; + struct inet_connection_sock *icsk = inet_csk(sk); + struct inet_sock *inet = inet_sk(sk); + struct ipv6_pinfo *np = inet6_sk(sk); + struct dccp_sock *dp = dccp_sk(sk); + struct in6_addr *saddr = NULL, *final_p = NULL, final; + struct flowi fl; + struct dst_entry *dst; + int addr_type; + int err; + + dp->dccps_role = DCCP_ROLE_CLIENT; + + if (addr_len < SIN6_LEN_RFC2133) + return -EINVAL; + + if (usin->sin6_family != AF_INET6) + return -EAFNOSUPPORT; + + memset(&fl, 0, sizeof(fl)); + + if (np->sndflow) { + fl.fl6_flowlabel = usin->sin6_flowinfo & IPV6_FLOWINFO_MASK; + IP6_ECN_flow_init(fl.fl6_flowlabel); + if (fl.fl6_flowlabel & IPV6_FLOWLABEL_MASK) { + struct ip6_flowlabel *flowlabel; + flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel); + if (flowlabel == NULL) + return -EINVAL; + ipv6_addr_copy(&usin->sin6_addr, &flowlabel->dst); + fl6_sock_release(flowlabel); + } + } + /* + * connect() to INADDR_ANY means loopback (BSD'ism). + */ + if (ipv6_addr_any(&usin->sin6_addr)) + usin->sin6_addr.s6_addr[15] = 1; + + addr_type = ipv6_addr_type(&usin->sin6_addr); + + if (addr_type & IPV6_ADDR_MULTICAST) + return -ENETUNREACH; + + if (addr_type & IPV6_ADDR_LINKLOCAL) { + if (addr_len >= sizeof(struct sockaddr_in6) && + usin->sin6_scope_id) { + /* If interface is set while binding, indices + * must coincide. + */ + if (sk->sk_bound_dev_if && + sk->sk_bound_dev_if != usin->sin6_scope_id) + return -EINVAL; + + sk->sk_bound_dev_if = usin->sin6_scope_id; + } + + /* Connect to link-local address requires an interface */ + if (!sk->sk_bound_dev_if) + return -EINVAL; + } + + ipv6_addr_copy(&np->daddr, &usin->sin6_addr); + np->flow_label = fl.fl6_flowlabel; + + /* + * DCCP over IPv4 + */ + if (addr_type == IPV6_ADDR_MAPPED) { + u32 exthdrlen = icsk->icsk_ext_hdr_len; + struct sockaddr_in sin; + + SOCK_DEBUG(sk, "connect: ipv4 mapped\n"); + + if (__ipv6_only_sock(sk)) + return -ENETUNREACH; + + sin.sin_family = AF_INET; + sin.sin_port = usin->sin6_port; + sin.sin_addr.s_addr = usin->sin6_addr.s6_addr32[3]; + + icsk->icsk_af_ops = &dccp_ipv6_mapped; + sk->sk_backlog_rcv = dccp_v4_do_rcv; + + err = dccp_v4_connect(sk, (struct sockaddr *)&sin, sizeof(sin)); + if (err) { + icsk->icsk_ext_hdr_len = exthdrlen; + icsk->icsk_af_ops = &dccp_ipv6_af_ops; + sk->sk_backlog_rcv = dccp_v6_do_rcv; + goto failure; + } else { + ipv6_addr_set(&np->saddr, 0, 0, htonl(0x0000FFFF), + inet->saddr); + ipv6_addr_set(&np->rcv_saddr, 0, 0, htonl(0x0000FFFF), + inet->rcv_saddr); + } + + return err; + } + + if (!ipv6_addr_any(&np->rcv_saddr)) + saddr = &np->rcv_saddr; + + fl.proto = IPPROTO_DCCP; + ipv6_addr_copy(&fl.fl6_dst, &np->daddr); + ipv6_addr_copy(&fl.fl6_src, saddr ? saddr : &np->saddr); + fl.oif = sk->sk_bound_dev_if; + fl.fl_ip_dport = usin->sin6_port; + fl.fl_ip_sport = inet->sport; + security_sk_classify_flow(sk, &fl); + + if (np->opt != NULL && np->opt->srcrt != NULL) { + const struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt; + + ipv6_addr_copy(&final, &fl.fl6_dst); + ipv6_addr_copy(&fl.fl6_dst, rt0->addr); + final_p = &final; + } + + err = ip6_dst_lookup(sk, &dst, &fl); + if (err) + goto failure; + + if (final_p) + ipv6_addr_copy(&fl.fl6_dst, final_p); + + err = xfrm_lookup(&dst, &fl, sk, 0); + if (err < 0) + goto failure; + + if (saddr == NULL) { + saddr = &fl.fl6_src; + ipv6_addr_copy(&np->rcv_saddr, saddr); + } + + /* set the source address */ + ipv6_addr_copy(&np->saddr, saddr); + inet->rcv_saddr = LOOPBACK4_IPV6; + + __ip6_dst_store(sk, dst, NULL, NULL); + + icsk->icsk_ext_hdr_len = 0; + if (np->opt != NULL) + icsk->icsk_ext_hdr_len = (np->opt->opt_flen + + np->opt->opt_nflen); + + inet->dport = usin->sin6_port; + + dccp_set_state(sk, DCCP_REQUESTING); + err = inet6_hash_connect(&dccp_death_row, sk); + if (err) + goto late_failure; + + dp->dccps_iss = secure_dccpv6_sequence_number(np->saddr.s6_addr32, + np->daddr.s6_addr32, + inet->sport, inet->dport); + err = dccp_connect(sk); + if (err) + goto late_failure; + + return 0; + +late_failure: + dccp_set_state(sk, DCCP_CLOSED); + __sk_dst_reset(sk); +failure: + inet->dport = 0; + sk->sk_route_caps = 0; + return err; } static struct inet_connection_sock_af_ops dccp_ipv6_af_ops = { @@ -1179,6 +1146,10 @@ static int dccp_v6_destroy_sock(struct sock *sk) return inet6_destroy_sock(sk); } +static struct timewait_sock_ops dccp6_timewait_sock_ops = { + .twsk_obj_size = sizeof(struct dccp6_timewait_sock), +}; + static struct proto dccp_v6_prot = { .name = "DCCPv6", .owner = THIS_MODULE, diff --git a/net/dccp/minisocks.c b/net/dccp/minisocks.c index 9045438d6b3..6656bb497c7 100644 --- a/net/dccp/minisocks.c +++ b/net/dccp/minisocks.c @@ -11,6 +11,7 @@ */ #include <linux/dccp.h> +#include <linux/kernel.h> #include <linux/skbuff.h> #include <linux/timer.h> @@ -31,8 +32,7 @@ struct inet_timewait_death_row dccp_death_row = { .tw_timer = TIMER_INITIALIZER(inet_twdr_hangman, 0, (unsigned long)&dccp_death_row), .twkill_work = __WORK_INITIALIZER(dccp_death_row.twkill_work, - inet_twdr_twkill_work, - &dccp_death_row), + inet_twdr_twkill_work), /* Short-time timewait calendar */ .twcal_hand = -1, @@ -83,8 +83,7 @@ void dccp_time_wait(struct sock *sk, int state, int timeo) * socket up. We've got bigger problems than * non-graceful socket closings. */ - LIMIT_NETDEBUG(KERN_INFO "DCCP: time wait bucket " - "table overflow\n"); + DCCP_WARN("time wait bucket table overflow\n"); } dccp_done(sk); @@ -97,8 +96,8 @@ struct sock *dccp_create_openreq_child(struct sock *sk, /* * Step 3: Process LISTEN state * - * // Generate a new socket and switch to that socket - * Set S := new socket for this port pair + * (* Generate a new socket and switch to that socket *) + * Set S := new socket for this port pair */ struct sock *newsk = inet_csk_clone(sk, req, GFP_ATOMIC); @@ -147,9 +146,9 @@ out_free: /* * Step 3: Process LISTEN state * - * Choose S.ISS (initial seqno) or set from Init Cookie - * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init - * Cookie + * Choose S.ISS (initial seqno) or set from Init Cookies + * Initialize S.GAR := S.ISS + * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookies */ /* See dccp_v4_conn_request */ @@ -183,7 +182,7 @@ out_free: EXPORT_SYMBOL_GPL(dccp_create_openreq_child); -/* +/* * Process an incoming packet for RESPOND sockets represented * as an request_sock. */ @@ -195,15 +194,17 @@ struct sock *dccp_check_req(struct sock *sk, struct sk_buff *skb, /* Check for retransmitted REQUEST */ if (dccp_hdr(skb)->dccph_type == DCCP_PKT_REQUEST) { - if (after48(DCCP_SKB_CB(skb)->dccpd_seq, - dccp_rsk(req)->dreq_isr)) { - struct dccp_request_sock *dreq = dccp_rsk(req); + struct dccp_request_sock *dreq = dccp_rsk(req); + if (after48(DCCP_SKB_CB(skb)->dccpd_seq, dreq->dreq_isr)) { dccp_pr_debug("Retransmitted REQUEST\n"); - /* Send another RESPONSE packet */ - dccp_set_seqno(&dreq->dreq_iss, dreq->dreq_iss + 1); - dccp_set_seqno(&dreq->dreq_isr, - DCCP_SKB_CB(skb)->dccpd_seq); + dreq->dreq_isr = DCCP_SKB_CB(skb)->dccpd_seq; + /* + * Send another RESPONSE packet + * To protect against Request floods, increment retrans + * counter (backoff, monitored by dccp_response_timer). + */ + req->retrans++; req->rsk_ops->rtx_syn_ack(sk, req, NULL); } /* Network Duplicate, discard packet */ @@ -243,7 +244,7 @@ listen_overflow: DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_TOO_BUSY; drop: if (dccp_hdr(skb)->dccph_type != DCCP_PKT_RESET) - req->rsk_ops->send_reset(skb); + req->rsk_ops->send_reset(sk, skb); inet_csk_reqsk_queue_drop(sk, req, prev); goto out; @@ -283,3 +284,19 @@ int dccp_child_process(struct sock *parent, struct sock *child, } EXPORT_SYMBOL_GPL(dccp_child_process); + +void dccp_reqsk_send_ack(struct sk_buff *skb, struct request_sock *rsk) +{ + DCCP_BUG("DCCP-ACK packets are never sent in LISTEN/RESPOND state"); +} + +EXPORT_SYMBOL_GPL(dccp_reqsk_send_ack); + +void dccp_reqsk_init(struct request_sock *req, struct sk_buff *skb) +{ + inet_rsk(req)->rmt_port = dccp_hdr(skb)->dccph_sport; + inet_rsk(req)->acked = 0; + req->rcv_wnd = sysctl_dccp_feat_sequence_window; +} + +EXPORT_SYMBOL_GPL(dccp_reqsk_init); diff --git a/net/dccp/options.c b/net/dccp/options.c index fb0db1f7cd7..c03ba61eb6d 100644 --- a/net/dccp/options.c +++ b/net/dccp/options.c @@ -22,23 +22,23 @@ #include "dccp.h" #include "feat.h" -int dccp_feat_default_sequence_window = DCCPF_INITIAL_SEQUENCE_WINDOW; -int dccp_feat_default_rx_ccid = DCCPF_INITIAL_CCID; -int dccp_feat_default_tx_ccid = DCCPF_INITIAL_CCID; -int dccp_feat_default_ack_ratio = DCCPF_INITIAL_ACK_RATIO; -int dccp_feat_default_send_ack_vector = DCCPF_INITIAL_SEND_ACK_VECTOR; -int dccp_feat_default_send_ndp_count = DCCPF_INITIAL_SEND_NDP_COUNT; +int sysctl_dccp_feat_sequence_window = DCCPF_INITIAL_SEQUENCE_WINDOW; +int sysctl_dccp_feat_rx_ccid = DCCPF_INITIAL_CCID; +int sysctl_dccp_feat_tx_ccid = DCCPF_INITIAL_CCID; +int sysctl_dccp_feat_ack_ratio = DCCPF_INITIAL_ACK_RATIO; +int sysctl_dccp_feat_send_ack_vector = DCCPF_INITIAL_SEND_ACK_VECTOR; +int sysctl_dccp_feat_send_ndp_count = DCCPF_INITIAL_SEND_NDP_COUNT; -EXPORT_SYMBOL_GPL(dccp_feat_default_sequence_window); +EXPORT_SYMBOL_GPL(sysctl_dccp_feat_sequence_window); void dccp_minisock_init(struct dccp_minisock *dmsk) { - dmsk->dccpms_sequence_window = dccp_feat_default_sequence_window; - dmsk->dccpms_rx_ccid = dccp_feat_default_rx_ccid; - dmsk->dccpms_tx_ccid = dccp_feat_default_tx_ccid; - dmsk->dccpms_ack_ratio = dccp_feat_default_ack_ratio; - dmsk->dccpms_send_ack_vector = dccp_feat_default_send_ack_vector; - dmsk->dccpms_send_ndp_count = dccp_feat_default_send_ndp_count; + dmsk->dccpms_sequence_window = sysctl_dccp_feat_sequence_window; + dmsk->dccpms_rx_ccid = sysctl_dccp_feat_rx_ccid; + dmsk->dccpms_tx_ccid = sysctl_dccp_feat_tx_ccid; + dmsk->dccpms_ack_ratio = sysctl_dccp_feat_ack_ratio; + dmsk->dccpms_send_ack_vector = sysctl_dccp_feat_send_ack_vector; + dmsk->dccpms_send_ndp_count = sysctl_dccp_feat_send_ndp_count; } static u32 dccp_decode_value_var(const unsigned char *bf, const u8 len) @@ -60,12 +60,9 @@ static u32 dccp_decode_value_var(const unsigned char *bf, const u8 len) int dccp_parse_options(struct sock *sk, struct sk_buff *skb) { struct dccp_sock *dp = dccp_sk(sk); -#ifdef CONFIG_IP_DCCP_DEBUG - const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? - "CLIENT rx opt: " : "server rx opt: "; -#endif const struct dccp_hdr *dh = dccp_hdr(skb); const u8 pkt_type = DCCP_SKB_CB(skb)->dccpd_type; + u64 ackno = DCCP_SKB_CB(skb)->dccpd_ack_seq; unsigned char *options = (unsigned char *)dh + dccp_hdr_len(skb); unsigned char *opt_ptr = options; const unsigned char *opt_end = (unsigned char *)dh + @@ -119,7 +116,7 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) goto out_invalid_option; opt_recv->dccpor_ndp = dccp_decode_value_var(value, len); - dccp_pr_debug("%sNDP count=%d\n", debug_prefix, + dccp_pr_debug("%s rx opt: NDP count=%d\n", dccp_role(sk), opt_recv->dccpor_ndp); break; case DCCPO_CHANGE_L: @@ -153,7 +150,7 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) break; if (dccp_msk(sk)->dccpms_send_ack_vector && - dccp_ackvec_parse(sk, skb, opt, value, len)) + dccp_ackvec_parse(sk, skb, &ackno, opt, value, len)) goto out_invalid_option; break; case DCCPO_TIMESTAMP: @@ -165,8 +162,8 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) dp->dccps_timestamp_echo = opt_recv->dccpor_timestamp; dccp_timestamp(sk, &dp->dccps_timestamp_time); - dccp_pr_debug("%sTIMESTAMP=%u, ackno=%llu\n", - debug_prefix, opt_recv->dccpor_timestamp, + dccp_pr_debug("%s rx opt: TIMESTAMP=%u, ackno=%llu\n", + dccp_role(sk), opt_recv->dccpor_timestamp, (unsigned long long) DCCP_SKB_CB(skb)->dccpd_ack_seq); break; @@ -176,8 +173,8 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) opt_recv->dccpor_timestamp_echo = ntohl(*(__be32 *)value); - dccp_pr_debug("%sTIMESTAMP_ECHO=%u, len=%d, ackno=%llu, ", - debug_prefix, + dccp_pr_debug("%s rx opt: TIMESTAMP_ECHO=%u, len=%d, " + "ackno=%llu, ", dccp_role(sk), opt_recv->dccpor_timestamp_echo, len + 2, (unsigned long long) @@ -211,8 +208,8 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) if (elapsed_time > opt_recv->dccpor_elapsed_time) opt_recv->dccpor_elapsed_time = elapsed_time; - dccp_pr_debug("%sELAPSED_TIME=%d\n", debug_prefix, - elapsed_time); + dccp_pr_debug("%s rx opt: ELAPSED_TIME=%d\n", + dccp_role(sk), elapsed_time); break; /* * From RFC 4340, sec. 10.3: @@ -242,9 +239,8 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) } break; default: - pr_info("DCCP(%p): option %d(len=%d) not " - "implemented, ignoring\n", - sk, opt, len); + DCCP_CRIT("DCCP(%p): option %d(len=%d) not " + "implemented, ignoring", sk, opt, len); break; } @@ -261,7 +257,7 @@ int dccp_parse_options(struct sock *sk, struct sk_buff *skb) out_invalid_option: DCCP_INC_STATS_BH(DCCP_MIB_INVALIDOPT); DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_OPTION_ERROR; - pr_info("DCCP(%p): invalid option %d, len=%d\n", sk, opt, len); + DCCP_WARN("DCCP(%p): invalid option %d, len=%d", sk, opt, len); return -1; } @@ -451,8 +447,7 @@ static int dccp_insert_feat_opt(struct sk_buff *skb, u8 type, u8 feat, u8 *to; if (DCCP_SKB_CB(skb)->dccpd_opt_len + len + 3 > DCCP_MAX_OPT_LEN) { - LIMIT_NETDEBUG(KERN_INFO "DCCP: packet too small" - " to insert feature %d option!\n", feat); + DCCP_WARN("packet too small for feature %d option!\n", feat); return -1; } @@ -465,8 +460,10 @@ static int dccp_insert_feat_opt(struct sk_buff *skb, u8 type, u8 feat, if (len) memcpy(to, val, len); - dccp_pr_debug("option %d feat %d len %d\n", type, feat, len); + dccp_pr_debug("%s(%s (%d), ...), length %d\n", + dccp_feat_typename(type), + dccp_feat_name(feat), feat, len); return 0; } @@ -560,11 +557,6 @@ int dccp_insert_options(struct sock *sk, struct sk_buff *skb) return -1; dp->dccps_hc_rx_insert_options = 0; } - if (dp->dccps_hc_tx_insert_options) { - if (ccid_hc_tx_insert_options(dp->dccps_hc_tx_ccid, sk, skb)) - return -1; - dp->dccps_hc_tx_insert_options = 0; - } /* Feature negotiation */ /* Data packets can't do feat negotiation */ diff --git a/net/dccp/output.c b/net/dccp/output.c index 7102e3aed4c..82456965908 100644 --- a/net/dccp/output.c +++ b/net/dccp/output.c @@ -1,6 +1,6 @@ /* * net/dccp/output.c - * + * * An implementation of the DCCP protocol * Arnaldo Carvalho de Melo <acme@conectiva.com.br> * @@ -88,16 +88,15 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb) return -EPROTO; } - skb->h.raw = skb_push(skb, dccp_header_size); - dh = dccp_hdr(skb); /* 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; @@ -117,7 +116,7 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb) 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); @@ -125,17 +124,8 @@ static int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb) DCCP_INC_STATS(DCCP_MIB_OUTSEGS); 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; + err = icsk->icsk_af_ops->queue_xmit(skb, sk, 0); + return net_xmit_eval(err); } return -ENOBUFS; } @@ -185,14 +175,12 @@ void dccp_write_space(struct sock *sk) /** * dccp_wait_for_ccid - Wait for ccid to tell us we can send a packet * @sk: socket to wait for - * @timeo: for how long */ -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) { struct dccp_sock *dp = dccp_sk(sk); DEFINE_WAIT(wait); - long delay; + unsigned long delay; int rc; while (1) { @@ -200,22 +188,16 @@ static int dccp_wait_for_ccid(struct sock *sk, struct sk_buff *skb, if (sk->sk_err) 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); + rc = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb); if (rc <= 0) break; delay = msecs_to_jiffies(rc); - if (delay > *timeo || delay < 0) - goto do_nonblock; - sk->sk_write_pending++; release_sock(sk); - *timeo -= schedule_timeout(delay); + schedule_timeout(delay); lock_sock(sk); sk->sk_write_pending--; } @@ -226,11 +208,8 @@ out: do_error: rc = -EPIPE; goto out; -do_nonblock: - rc = -EAGAIN; - goto out; do_interrupted: - rc = sock_intr_errno(*timeo); + rc = -EINTR; goto out; } @@ -251,12 +230,9 @@ void dccp_write_xmit(struct sock *sk, int block) { struct dccp_sock *dp = dccp_sk(sk); struct sk_buff *skb; - long timeo = 30000; /* If a packet is taking longer than 2 secs - we have other issues */ while ((skb = skb_peek(&sk->sk_write_queue))) { - int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb, - skb->len); + int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb); if (err > 0) { if (!block) { @@ -264,12 +240,9 @@ void dccp_write_xmit(struct sock *sk, int block) msecs_to_jiffies(err)+jiffies); break; } else - err = dccp_wait_for_ccid(sk, skb, &timeo); - if (err) { - printk(KERN_CRIT "%s:err at dccp_wait_for_ccid" - " %d\n", __FUNCTION__, err); - dump_stack(); - } + err = dccp_wait_for_ccid(sk, skb); + if (err && err != -EINTR) + DCCP_BUG("err=%d after dccp_wait_for_ccid", err); } skb_dequeue(&sk->sk_write_queue); @@ -291,14 +264,13 @@ void dccp_write_xmit(struct sock *sk, int block) err = dccp_transmit_skb(sk, skb); ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len); - if (err) { - printk(KERN_CRIT "%s:err from " - "ccid_hc_tx_packet_sent %d\n", - __FUNCTION__, err); - dump_stack(); - } - } else + if (err) + DCCP_BUG("err=%d after ccid_hc_tx_packet_sent", + err); + } else { + dccp_pr_debug("packet discarded\n"); kfree(skb); + } } } @@ -329,9 +301,10 @@ struct sk_buff *dccp_make_response(struct sock *sk, struct dst_entry *dst, 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; @@ -340,10 +313,8 @@ struct sk_buff *dccp_make_response(struct sock *sk, struct dst_entry *dst, return NULL; } - skb->h.raw = skb_push(skb, dccp_header_size); - - 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; @@ -355,6 +326,10 @@ struct sk_buff *dccp_make_response(struct sock *sk, struct dst_entry *dst, dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dreq->dreq_isr); dccp_hdr_response(skb)->dccph_resp_service = dreq->dreq_service; + 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; } @@ -363,7 +338,6 @@ EXPORT_SYMBOL_GPL(dccp_make_response); static struct sk_buff *dccp_make_reset(struct sock *sk, struct dst_entry *dst, const enum dccp_reset_codes code) - { struct dccp_hdr *dh; struct dccp_sock *dp = dccp_sk(sk); @@ -379,7 +353,6 @@ static struct sk_buff *dccp_make_reset(struct sock *sk, struct dst_entry *dst, skb_reserve(skb, sk->sk_prot->max_header); skb->dst = dst_clone(dst); - skb->csum = 0; dccp_inc_seqno(&dp->dccps_gss); @@ -392,10 +365,7 @@ static struct sk_buff *dccp_make_reset(struct sock *sk, struct dst_entry *dst, return NULL; } - skb->h.raw = skb_push(skb, dccp_header_size); - - dh = dccp_hdr(skb); - memset(dh, 0, dccp_header_size); + dh = dccp_zeroed_hdr(skb, dccp_header_size); dh->dccph_sport = inet_sk(sk)->sport; dh->dccph_dport = inet_sk(sk)->dport; @@ -407,7 +377,7 @@ static struct sk_buff *dccp_make_reset(struct sock *sk, struct dst_entry *dst, dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dp->dccps_gsr); dccp_hdr_reset(skb)->dccph_reset_code = code; - inet_csk(sk)->icsk_af_ops->send_check(sk, skb->len, skb); + inet_csk(sk)->icsk_af_ops->send_check(sk, 0, skb); DCCP_INC_STATS(DCCP_MIB_OUTSEGS); return skb; @@ -426,9 +396,8 @@ int dccp_send_reset(struct sock *sk, enum dccp_reset_codes code) code); if (skb != NULL) { memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); - err = inet_csk(sk)->icsk_af_ops->queue_xmit(skb, 0); - if (err == NET_XMIT_CN) - err = 0; + err = inet_csk(sk)->icsk_af_ops->queue_xmit(skb, sk, 0); + return net_xmit_eval(err); } } @@ -449,17 +418,21 @@ static inline void dccp_connect_init(struct sock *sk) 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; init_timer(&dp->dccps_xmit_timer); dp->dccps_xmit_timer.data = (unsigned long)sk; @@ -481,7 +454,6 @@ int dccp_connect(struct sock *sk) 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)); @@ -513,7 +485,6 @@ void dccp_send_ack(struct sock *sk) /* Reserve space for headers */ skb_reserve(skb, sk->sk_prot->max_header); - skb->csum = 0; DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_ACK; dccp_transmit_skb(sk, skb); } @@ -567,7 +538,6 @@ void dccp_send_sync(struct sock *sk, const u64 seq, /* 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 = pkt_type; DCCP_SKB_CB(skb)->dccpd_seq = seq; @@ -593,7 +563,6 @@ void dccp_send_close(struct sock *sk, const int active) /* 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; diff --git a/net/dccp/probe.c b/net/dccp/probe.c index fded1493c1d..f81e37de35d 100644 --- a/net/dccp/probe.c +++ b/net/dccp/probe.c @@ -106,8 +106,10 @@ static int jdccp_sendmsg(struct kiocb *iocb, struct sock *sk, } static struct jprobe dccp_send_probe = { - .kp = { .addr = (kprobe_opcode_t *)&dccp_sendmsg, }, - .entry = (kprobe_opcode_t *)&jdccp_sendmsg, + .kp = { + .symbol_name = "dccp_sendmsg", + }, + .entry = JPROBE_ENTRY(jdccp_sendmsg), }; static int dccpprobe_open(struct inode *inode, struct file *file) diff --git a/net/dccp/proto.c b/net/dccp/proto.c index 72cbdcfc2c6..63b3fa20e14 100644 --- a/net/dccp/proto.c +++ b/net/dccp/proto.c @@ -52,6 +52,9 @@ struct inet_hashinfo __cacheline_aligned dccp_hashinfo = { EXPORT_SYMBOL_GPL(dccp_hashinfo); +/* the maximum queue length for tx in packets. 0 is no limit */ +int sysctl_dccp_tx_qlen __read_mostly = 5; + void dccp_set_state(struct sock *sk, const int state) { const int oldstate = sk->sk_state; @@ -193,7 +196,7 @@ int dccp_init_sock(struct sock *sk, const __u8 ctl_sock_initialized) sk, GFP_KERNEL); dp->dccps_hc_tx_ccid = ccid_hc_tx_new(dmsk->dccpms_tx_ccid, sk, GFP_KERNEL); - if (unlikely(dp->dccps_hc_rx_ccid == NULL || + if (unlikely(dp->dccps_hc_rx_ccid == NULL || dp->dccps_hc_tx_ccid == NULL)) { ccid_hc_rx_delete(dp->dccps_hc_rx_ccid, sk); ccid_hc_tx_delete(dp->dccps_hc_tx_ccid, sk); @@ -212,6 +215,7 @@ int dccp_init_sock(struct sock *sk, const __u8 ctl_sock_initialized) dccp_init_xmit_timers(sk); icsk->icsk_rto = DCCP_TIMEOUT_INIT; + icsk->icsk_syn_retries = sysctl_dccp_request_retries; sk->sk_state = DCCP_CLOSED; sk->sk_write_space = dccp_write_space; icsk->icsk_sync_mss = dccp_sync_mss; @@ -262,12 +266,12 @@ int dccp_destroy_sock(struct sock *sk) EXPORT_SYMBOL_GPL(dccp_destroy_sock); -static inline int dccp_listen_start(struct sock *sk) +static inline int dccp_listen_start(struct sock *sk, int backlog) { struct dccp_sock *dp = dccp_sk(sk); dp->dccps_role = DCCP_ROLE_LISTEN; - return inet_csk_listen_start(sk, TCP_SYNQ_HSIZE); + return inet_csk_listen_start(sk, backlog); } int dccp_disconnect(struct sock *sk, int flags) @@ -386,7 +390,7 @@ static int dccp_setsockopt_service(struct sock *sk, const __be32 service, struct dccp_sock *dp = dccp_sk(sk); struct dccp_service_list *sl = NULL; - if (service == DCCP_SERVICE_INVALID_VALUE || + if (service == DCCP_SERVICE_INVALID_VALUE || optlen > DCCP_SERVICE_LIST_MAX_LEN * sizeof(u32)) return -EINVAL; @@ -451,9 +455,8 @@ out_free_val: static int do_dccp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, int optlen) { - struct dccp_sock *dp; - int err; - int val; + struct dccp_sock *dp = dccp_sk(sk); + int val, err = 0; if (optlen < sizeof(int)) return -EINVAL; @@ -465,14 +468,11 @@ static int do_dccp_setsockopt(struct sock *sk, int level, int optname, return dccp_setsockopt_service(sk, val, optval, optlen); lock_sock(sk); - dp = dccp_sk(sk); - err = 0; - switch (optname) { case DCCP_SOCKOPT_PACKET_SIZE: - dp->dccps_packet_size = val; + DCCP_WARN("sockopt(PACKET_SIZE) is deprecated: fix your app\n"); + err = 0; break; - case DCCP_SOCKOPT_CHANGE_L: if (optlen != sizeof(struct dccp_so_feat)) err = -EINVAL; @@ -481,7 +481,6 @@ static int do_dccp_setsockopt(struct sock *sk, int level, int optname, (struct dccp_so_feat __user *) optval); break; - case DCCP_SOCKOPT_CHANGE_R: if (optlen != sizeof(struct dccp_so_feat)) err = -EINVAL; @@ -490,12 +489,26 @@ static int do_dccp_setsockopt(struct sock *sk, int level, int optname, (struct dccp_so_feat __user *) optval); break; - + case DCCP_SOCKOPT_SEND_CSCOV: /* sender side, RFC 4340, sec. 9.2 */ + if (val < 0 || val > 15) + err = -EINVAL; + else + dp->dccps_pcslen = val; + break; + case DCCP_SOCKOPT_RECV_CSCOV: /* receiver side, RFC 4340 sec. 9.2.1 */ + if (val < 0 || val > 15) + err = -EINVAL; + else { + dp->dccps_pcrlen = val; + /* FIXME: add feature negotiation, + * ChangeL(MinimumChecksumCoverage, val) */ + } + break; default: err = -ENOPROTOOPT; break; } - + release_sock(sk); return err; } @@ -569,12 +582,17 @@ static int do_dccp_getsockopt(struct sock *sk, int level, int optname, switch (optname) { case DCCP_SOCKOPT_PACKET_SIZE: - val = dp->dccps_packet_size; - len = sizeof(dp->dccps_packet_size); - break; + DCCP_WARN("sockopt(PACKET_SIZE) is deprecated: fix your app\n"); + return 0; case DCCP_SOCKOPT_SERVICE: return dccp_getsockopt_service(sk, len, (__be32 __user *)optval, optlen); + case DCCP_SOCKOPT_SEND_CSCOV: + val = dp->dccps_pcslen; + break; + case DCCP_SOCKOPT_RECV_CSCOV: + val = dp->dccps_pcrlen; + break; case 128 ... 191: return ccid_hc_rx_getsockopt(dp->dccps_hc_rx_ccid, sk, optname, len, (u32 __user *)optval, optlen); @@ -630,6 +648,13 @@ int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, return -EMSGSIZE; lock_sock(sk); + + if (sysctl_dccp_tx_qlen && + (sk->sk_write_queue.qlen >= sysctl_dccp_tx_qlen)) { + rc = -EAGAIN; + goto out_release; + } + timeo = sock_sndtimeo(sk, noblock); /* @@ -788,7 +813,7 @@ int inet_dccp_listen(struct socket *sock, int backlog) * FIXME: here it probably should be sk->sk_prot->listen_start * see tcp_listen_start */ - err = dccp_listen_start(sk); + err = dccp_listen_start(sk, backlog); if (err) goto out; } @@ -805,7 +830,7 @@ EXPORT_SYMBOL_GPL(inet_dccp_listen); static const unsigned char dccp_new_state[] = { /* current state: new state: action: */ [0] = DCCP_CLOSED, - [DCCP_OPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, + [DCCP_OPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, [DCCP_REQUESTING] = DCCP_CLOSED, [DCCP_PARTOPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, [DCCP_LISTEN] = DCCP_CLOSED, @@ -1008,8 +1033,7 @@ static int __init dccp_init(void) } while (!dccp_hashinfo.ehash && --ehash_order > 0); if (!dccp_hashinfo.ehash) { - printk(KERN_CRIT "Failed to allocate DCCP " - "established hash table\n"); + DCCP_CRIT("Failed to allocate DCCP established hash table"); goto out_free_bind_bucket_cachep; } @@ -1031,7 +1055,7 @@ static int __init dccp_init(void) } while (!dccp_hashinfo.bhash && --bhash_order >= 0); if (!dccp_hashinfo.bhash) { - printk(KERN_CRIT "Failed to allocate DCCP bind hash table\n"); + DCCP_CRIT("Failed to allocate DCCP bind hash table"); goto out_free_dccp_ehash; } diff --git a/net/dccp/sysctl.c b/net/dccp/sysctl.c index 38bc157876f..fdcfca3e920 100644 --- a/net/dccp/sysctl.c +++ b/net/dccp/sysctl.c @@ -11,6 +11,7 @@ #include <linux/mm.h> #include <linux/sysctl.h> +#include "dccp.h" #include "feat.h" #ifndef CONFIG_SYSCTL @@ -19,53 +20,76 @@ static struct ctl_table dccp_default_table[] = { { - .ctl_name = NET_DCCP_DEFAULT_SEQ_WINDOW, .procname = "seq_window", - .data = &dccp_feat_default_sequence_window, - .maxlen = sizeof(dccp_feat_default_sequence_window), + .data = &sysctl_dccp_feat_sequence_window, + .maxlen = sizeof(sysctl_dccp_feat_sequence_window), .mode = 0644, .proc_handler = proc_dointvec, }, { - .ctl_name = NET_DCCP_DEFAULT_RX_CCID, .procname = "rx_ccid", - .data = &dccp_feat_default_rx_ccid, - .maxlen = sizeof(dccp_feat_default_rx_ccid), + .data = &sysctl_dccp_feat_rx_ccid, + .maxlen = sizeof(sysctl_dccp_feat_rx_ccid), .mode = 0644, .proc_handler = proc_dointvec, }, { - .ctl_name = NET_DCCP_DEFAULT_TX_CCID, .procname = "tx_ccid", - .data = &dccp_feat_default_tx_ccid, - .maxlen = sizeof(dccp_feat_default_tx_ccid), + .data = &sysctl_dccp_feat_tx_ccid, + .maxlen = sizeof(sysctl_dccp_feat_tx_ccid), .mode = 0644, .proc_handler = proc_dointvec, }, { - .ctl_name = NET_DCCP_DEFAULT_ACK_RATIO, .procname = "ack_ratio", - .data = &dccp_feat_default_ack_ratio, - .maxlen = sizeof(dccp_feat_default_ack_ratio), + .data = &sysctl_dccp_feat_ack_ratio, + .maxlen = sizeof(sysctl_dccp_feat_ack_ratio), .mode = 0644, .proc_handler = proc_dointvec, }, { - .ctl_name = NET_DCCP_DEFAULT_SEND_ACKVEC, .procname = "send_ackvec", - .data = &dccp_feat_default_send_ack_vector, - .maxlen = sizeof(dccp_feat_default_send_ack_vector), + .data = &sysctl_dccp_feat_send_ack_vector, + .maxlen = sizeof(sysctl_dccp_feat_send_ack_vector), .mode = 0644, .proc_handler = proc_dointvec, }, { - .ctl_name = NET_DCCP_DEFAULT_SEND_NDP, .procname = "send_ndp", - .data = &dccp_feat_default_send_ndp_count, - .maxlen = sizeof(dccp_feat_default_send_ndp_count), + .data = &sysctl_dccp_feat_send_ndp_count, + .maxlen = sizeof(sysctl_dccp_feat_send_ndp_count), .mode = 0644, .proc_handler = proc_dointvec, }, + { + .procname = "request_retries", + .data = &sysctl_dccp_request_retries, + .maxlen = sizeof(sysctl_dccp_request_retries), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "retries1", + .data = &sysctl_dccp_retries1, + .maxlen = sizeof(sysctl_dccp_retries1), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "retries2", + .data = &sysctl_dccp_retries2, + .maxlen = sizeof(sysctl_dccp_retries2), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "tx_qlen", + .data = &sysctl_dccp_tx_qlen, + .maxlen = sizeof(sysctl_dccp_tx_qlen), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { .ctl_name = 0, } }; diff --git a/net/dccp/timer.c b/net/dccp/timer.c index 8447742f561..e5348f369c6 100644 --- a/net/dccp/timer.c +++ b/net/dccp/timer.c @@ -1,6 +1,6 @@ /* * net/dccp/timer.c - * + * * An implementation of the DCCP protocol * Arnaldo Carvalho de Melo <acme@conectiva.com.br> * @@ -15,15 +15,10 @@ #include "dccp.h" -static void dccp_write_timer(unsigned long data); -static void dccp_keepalive_timer(unsigned long data); -static void dccp_delack_timer(unsigned long data); - -void dccp_init_xmit_timers(struct sock *sk) -{ - inet_csk_init_xmit_timers(sk, &dccp_write_timer, &dccp_delack_timer, - &dccp_keepalive_timer); -} +/* sysctl variables governing numbers of retransmission attempts */ +int sysctl_dccp_request_retries __read_mostly = TCP_SYN_RETRIES; +int sysctl_dccp_retries1 __read_mostly = TCP_RETR1; +int sysctl_dccp_retries2 __read_mostly = TCP_RETR2; static void dccp_write_err(struct sock *sk) { @@ -44,11 +39,10 @@ static int dccp_write_timeout(struct sock *sk) if (sk->sk_state == DCCP_REQUESTING || sk->sk_state == DCCP_PARTOPEN) { if (icsk->icsk_retransmits != 0) dst_negative_advice(&sk->sk_dst_cache); - retry_until = icsk->icsk_syn_retries ? : - /* FIXME! */ 3 /* FIXME! sysctl_tcp_syn_retries */; + retry_until = icsk->icsk_syn_retries ? + : sysctl_dccp_request_retries; } else { - if (icsk->icsk_retransmits >= - /* FIXME! sysctl_tcp_retries1 */ 5 /* FIXME! */) { + if (icsk->icsk_retransmits >= sysctl_dccp_retries1) { /* NOTE. draft-ietf-tcpimpl-pmtud-01.txt requires pmtu black hole detection. :-( @@ -72,7 +66,7 @@ static int dccp_write_timeout(struct sock *sk) dst_negative_advice(&sk->sk_dst_cache); } - retry_until = /* FIXME! */ 15 /* FIXME! sysctl_tcp_retries2 */; + retry_until = sysctl_dccp_retries2; /* * FIXME: see tcp_write_timout and tcp_out_of_resources */ @@ -86,53 +80,6 @@ static int dccp_write_timeout(struct sock *sk) return 0; } -/* This is the same as tcp_delack_timer, sans prequeue & mem_reclaim stuff */ -static void dccp_delack_timer(unsigned long data) -{ - struct sock *sk = (struct sock *)data; - struct inet_connection_sock *icsk = inet_csk(sk); - - bh_lock_sock(sk); - if (sock_owned_by_user(sk)) { - /* Try again later. */ - icsk->icsk_ack.blocked = 1; - NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKLOCKED); - sk_reset_timer(sk, &icsk->icsk_delack_timer, - jiffies + TCP_DELACK_MIN); - goto out; - } - - if (sk->sk_state == DCCP_CLOSED || - !(icsk->icsk_ack.pending & ICSK_ACK_TIMER)) - goto out; - if (time_after(icsk->icsk_ack.timeout, jiffies)) { - sk_reset_timer(sk, &icsk->icsk_delack_timer, - icsk->icsk_ack.timeout); - goto out; - } - - icsk->icsk_ack.pending &= ~ICSK_ACK_TIMER; - - if (inet_csk_ack_scheduled(sk)) { - if (!icsk->icsk_ack.pingpong) { - /* Delayed ACK missed: inflate ATO. */ - icsk->icsk_ack.ato = min(icsk->icsk_ack.ato << 1, - icsk->icsk_rto); - } else { - /* Delayed ACK missed: leave pingpong mode and - * deflate ATO. - */ - icsk->icsk_ack.pingpong = 0; - icsk->icsk_ack.ato = TCP_ATO_MIN; - } - dccp_send_ack(sk); - NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKS); - } -out: - bh_unlock_sock(sk); - sock_put(sk); -} - /* * The DCCP retransmit timer. */ @@ -142,7 +89,7 @@ static void dccp_retransmit_timer(struct sock *sk) /* retransmit timer is used for feature negotiation throughout * connection. In this case, no packet is re-transmitted, but rather an - * ack is generated and pending changes are splaced into its options. + * ack is generated and pending changes are placed into its options. */ if (sk->sk_send_head == NULL) { dccp_pr_debug("feat negotiation retransmit timeout %p\n", sk); @@ -154,12 +101,14 @@ 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 (REQUEST, RESPONSE, the ACK in the 3way handshake - * (PARTOPEN timer), etc). - */ + * 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. */ @@ -194,7 +143,7 @@ backoff: 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 > 3 /* FIXME: sysctl_dccp_retries1 */) + if (icsk->icsk_retransmits > sysctl_dccp_retries1) __sk_dst_reset(sk); out:; } @@ -251,7 +200,7 @@ static void dccp_keepalive_timer(unsigned long data) /* Only process if socket is not in use. */ bh_lock_sock(sk); if (sock_owned_by_user(sk)) { - /* Try again later. */ + /* Try again later. */ inet_csk_reset_keepalive_timer(sk, HZ / 20); goto out; } @@ -264,3 +213,56 @@ out: bh_unlock_sock(sk); sock_put(sk); } + +/* This is the same as tcp_delack_timer, sans prequeue & mem_reclaim stuff */ +static void dccp_delack_timer(unsigned long data) +{ + struct sock *sk = (struct sock *)data; + struct inet_connection_sock *icsk = inet_csk(sk); + + bh_lock_sock(sk); + if (sock_owned_by_user(sk)) { + /* Try again later. */ + icsk->icsk_ack.blocked = 1; + NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKLOCKED); + sk_reset_timer(sk, &icsk->icsk_delack_timer, + jiffies + TCP_DELACK_MIN); + goto out; + } + + if (sk->sk_state == DCCP_CLOSED || + !(icsk->icsk_ack.pending & ICSK_ACK_TIMER)) + goto out; + if (time_after(icsk->icsk_ack.timeout, jiffies)) { + sk_reset_timer(sk, &icsk->icsk_delack_timer, + icsk->icsk_ack.timeout); + goto out; + } + + icsk->icsk_ack.pending &= ~ICSK_ACK_TIMER; + + if (inet_csk_ack_scheduled(sk)) { + if (!icsk->icsk_ack.pingpong) { + /* Delayed ACK missed: inflate ATO. */ + icsk->icsk_ack.ato = min(icsk->icsk_ack.ato << 1, + icsk->icsk_rto); + } else { + /* Delayed ACK missed: leave pingpong mode and + * deflate ATO. + */ + icsk->icsk_ack.pingpong = 0; + icsk->icsk_ack.ato = TCP_ATO_MIN; + } + dccp_send_ack(sk); + NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKS); + } +out: + bh_unlock_sock(sk); + sock_put(sk); +} + +void dccp_init_xmit_timers(struct sock *sk) +{ + inet_csk_init_xmit_timers(sk, &dccp_write_timer, &dccp_delack_timer, + &dccp_keepalive_timer); +} |