From 2e6599cb899ba4b133f42cbf9d2b1883d2dc583a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 18 Jun 2005 22:46:52 -0700 Subject: [NET] Generalise TCP's struct open_request minisock infrastructure Kept this first changeset minimal, without changing existing names to ease peer review. Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn has two new members: ->slab, that replaces tcp_openreq_cachep ->obj_size, to inform the size of the openreq descendant for a specific protocol The protocol specific fields in struct open_request were moved to a class hierarchy, with the things that are common to all connection oriented PF_INET protocols in struct inet_request_sock, the TCP ones in tcp_request_sock, that is an inet_request_sock, that is an open_request. I.e. this uses the same approach used for the struct sock class hierarchy, with sk_prot indicating if the protocol wants to use the open_request infrastructure by filling in sk_prot->rsk_prot with an or_calltable. Results? Performance is improved and TCP v4 now uses only 64 bytes per open request minisock, down from 96 without this patch :-) Next changeset will rename some of the structs, fields and functions mentioned above, struct or_calltable is way unclear, better name it struct request_sock_ops, s/struct open_request/struct request_sock/g, etc. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 25 ++++++++++++++----------- 1 file changed, 14 insertions(+), 11 deletions(-) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index 8faa8948f75c..700ff2413588 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -458,6 +458,7 @@ static int tcpdiag_fill_req(struct sk_buff *skb, struct sock *sk, struct open_request *req, u32 pid, u32 seq) { + const struct inet_request_sock *ireq = inet_rsk(req); struct inet_sock *inet = inet_sk(sk); unsigned char *b = skb->tail; struct tcpdiagmsg *r; @@ -482,9 +483,9 @@ static int tcpdiag_fill_req(struct sk_buff *skb, struct sock *sk, tmo = 0; r->id.tcpdiag_sport = inet->sport; - r->id.tcpdiag_dport = req->rmt_port; - r->id.tcpdiag_src[0] = req->af.v4_req.loc_addr; - r->id.tcpdiag_dst[0] = req->af.v4_req.rmt_addr; + r->id.tcpdiag_dport = ireq->rmt_port; + r->id.tcpdiag_src[0] = ireq->loc_addr; + r->id.tcpdiag_dst[0] = ireq->rmt_addr; r->tcpdiag_expires = jiffies_to_msecs(tmo), r->tcpdiag_rqueue = 0; r->tcpdiag_wqueue = 0; @@ -493,9 +494,9 @@ static int tcpdiag_fill_req(struct sk_buff *skb, struct sock *sk, #ifdef CONFIG_IP_TCPDIAG_IPV6 if (r->tcpdiag_family == AF_INET6) { ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_src, - &req->af.v6_req.loc_addr); + &tcp6_rsk(req)->loc_addr); ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_dst, - &req->af.v6_req.rmt_addr); + &tcp6_rsk(req)->rmt_addr); } #endif nlh->nlmsg_len = skb->tail - b; @@ -545,9 +546,11 @@ static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, reqnum = 0; for (req = head; req; reqnum++, req = req->dl_next) { + struct inet_request_sock *ireq = inet_rsk(req); + if (reqnum < s_reqnum) continue; - if (r->id.tcpdiag_dport != req->rmt_port && + if (r->id.tcpdiag_dport != ireq->rmt_port && r->id.tcpdiag_dport) continue; @@ -555,16 +558,16 @@ static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, entry.saddr = #ifdef CONFIG_IP_TCPDIAG_IPV6 (entry.family == AF_INET6) ? - req->af.v6_req.loc_addr.s6_addr32 : + tcp6_rsk(req)->loc_addr.s6_addr32 : #endif - &req->af.v4_req.loc_addr; + &ireq->loc_addr; entry.daddr = #ifdef CONFIG_IP_TCPDIAG_IPV6 (entry.family == AF_INET6) ? - req->af.v6_req.rmt_addr.s6_addr32 : + tcp6_rsk(req)->rmt_addr.s6_addr32 : #endif - &req->af.v4_req.rmt_addr; - entry.dport = ntohs(req->rmt_port); + &ireq->rmt_addr; + entry.dport = ntohs(ireq->rmt_port); if (!tcpdiag_bc_run(RTA_DATA(bc), RTA_PAYLOAD(bc), &entry)) -- cgit v1.2.1 From 60236fdd08b2169045a3bbfc5ffe1576e6c3c17b Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 18 Jun 2005 22:47:21 -0700 Subject: [NET] Rename open_request to request_sock Ok, this one just renames some stuff to have a better namespace and to dissassociate it from TCP: struct open_request -> struct request_sock tcp_openreq_alloc -> reqsk_alloc tcp_openreq_free -> reqsk_free tcp_openreq_fastfree -> __reqsk_free With this most of the infrastructure closely resembles a struct sock methods subset. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index 700ff2413588..67277800d0c1 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -455,7 +455,7 @@ static int tcpdiag_dump_sock(struct sk_buff *skb, struct sock *sk, } static int tcpdiag_fill_req(struct sk_buff *skb, struct sock *sk, - struct open_request *req, + struct request_sock *req, u32 pid, u32 seq) { const struct inet_request_sock *ireq = inet_rsk(req); @@ -542,7 +542,7 @@ static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, } for (j = s_j; j < TCP_SYNQ_HSIZE; j++) { - struct open_request *req, *head = lopt->syn_table[j]; + struct request_sock *req, *head = lopt->syn_table[j]; reqnum = 0; for (req = head; req; reqnum++, req = req->dl_next) { -- cgit v1.2.1 From 0e87506fcc734647c7b2497eee4eb81e785c857a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 18 Jun 2005 22:47:59 -0700 Subject: [NET] Generalise tcp_listen_opt This chunks out the accept_queue and tcp_listen_opt code and moves them to net/core/request_sock.c and include/net/request_sock.h, to make it useful for other transport protocols, DCCP being the first one to use it. Next patches will rename tcp_listen_opt to accept_sock and remove the inline tcp functions that just call a reqsk_queue_ function. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index 67277800d0c1..c3328fa48837 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -529,9 +529,9 @@ static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, entry.family = sk->sk_family; - read_lock_bh(&tp->syn_wait_lock); + read_lock_bh(&tp->accept_queue.syn_wait_lock); - lopt = tp->listen_opt; + lopt = tp->accept_queue.listen_opt; if (!lopt || !lopt->qlen) goto out; @@ -588,7 +588,7 @@ static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, } out: - read_unlock_bh(&tp->syn_wait_lock); + read_unlock_bh(&tp->accept_queue.syn_wait_lock); return err; } -- cgit v1.2.1 From 2ad69c55a282315e6119cf7fd744f26a925bdfd2 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 18 Jun 2005 22:48:55 -0700 Subject: [NET] rename struct tcp_listen_opt to struct listen_sock Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index c3328fa48837..634befc07921 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -514,7 +514,7 @@ static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, struct tcpdiag_entry entry; struct tcpdiagreq *r = NLMSG_DATA(cb->nlh); struct tcp_sock *tp = tcp_sk(sk); - struct tcp_listen_opt *lopt; + struct listen_sock *lopt; struct rtattr *bc = NULL; struct inet_sock *inet = inet_sk(sk); int j, s_j; -- cgit v1.2.1 From 317a76f9a44b437d6301718f4e5d08bd93f98da7 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Thu, 23 Jun 2005 12:19:55 -0700 Subject: [TCP]: Add pluggable congestion control algorithm infrastructure. Allow TCP to have multiple pluggable congestion control algorithms. Algorithms are defined by a set of operations and can be built in or modules. The legacy "new RENO" algorithm is used as a starting point and fallback. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 20 ++------------------ 1 file changed, 2 insertions(+), 18 deletions(-) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index 634befc07921..867acc0f79d8 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -42,7 +42,6 @@ struct tcpdiag_entry static struct sock *tcpnl; - #define TCPDIAG_PUT(skb, attrtype, attrlen) \ ({ int rtalen = RTA_LENGTH(attrlen); \ struct rtattr *rta; \ @@ -61,7 +60,6 @@ static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, struct nlmsghdr *nlh; struct tcp_info *info = NULL; struct tcpdiag_meminfo *minfo = NULL; - struct tcpvegas_info *vinfo = NULL; unsigned char *b = skb->tail; nlh = NLMSG_PUT(skb, pid, seq, TCPDIAG_GETSOCK, sizeof(*r)); @@ -73,9 +71,6 @@ static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, if (ext & (1<<(TCPDIAG_INFO-1))) info = TCPDIAG_PUT(skb, TCPDIAG_INFO, sizeof(*info)); - if ((tcp_is_westwood(tp) || tcp_is_vegas(tp)) - && (ext & (1<<(TCPDIAG_VEGASINFO-1)))) - vinfo = TCPDIAG_PUT(skb, TCPDIAG_VEGASINFO, sizeof(*vinfo)); } r->tcpdiag_family = sk->sk_family; r->tcpdiag_state = sk->sk_state; @@ -166,19 +161,8 @@ static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, if (info) tcp_get_info(sk, info); - if (vinfo) { - if (tcp_is_vegas(tp)) { - vinfo->tcpv_enabled = tp->vegas.doing_vegas_now; - vinfo->tcpv_rttcnt = tp->vegas.cntRTT; - vinfo->tcpv_rtt = jiffies_to_usecs(tp->vegas.baseRTT); - vinfo->tcpv_minrtt = jiffies_to_usecs(tp->vegas.minRTT); - } else { - vinfo->tcpv_enabled = 0; - vinfo->tcpv_rttcnt = 0; - vinfo->tcpv_rtt = jiffies_to_usecs(tp->westwood.rtt); - vinfo->tcpv_minrtt = jiffies_to_usecs(tp->westwood.rtt_min); - } - } + if (sk->sk_state < TCP_TIME_WAIT && tp->ca_ops->get_info) + tp->ca_ops->get_info(tp, ext, skb); nlh->nlmsg_len = skb->tail - b; return skb->len; -- cgit v1.2.1 From 7c99c909fa69a183c1b80bd64fb9f0d11459aff3 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Thu, 23 Jun 2005 12:20:36 -0700 Subject: [TCP]: Change tcp_diag to use the existing __RTA_PUT() macro. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index 867acc0f79d8..a4e512036d88 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -43,13 +43,7 @@ struct tcpdiag_entry static struct sock *tcpnl; #define TCPDIAG_PUT(skb, attrtype, attrlen) \ -({ int rtalen = RTA_LENGTH(attrlen); \ - struct rtattr *rta; \ - if (skb_tailroom(skb) < RTA_ALIGN(rtalen)) goto nlmsg_failure; \ - rta = (void*)__skb_put(skb, RTA_ALIGN(rtalen)); \ - rta->rta_type = attrtype; \ - rta->rta_len = rtalen; \ - RTA_DATA(rta); }) + RTA_DATA(__RTA_PUT(skb, attrtype, attrlen)) static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, int ext, u32 pid, u32 seq, u16 nlmsg_flags) @@ -167,6 +161,7 @@ static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, nlh->nlmsg_len = skb->tail - b; return skb->len; +rtattr_failure: nlmsg_failure: skb_trim(skb, b - skb->data); return -1; -- cgit v1.2.1 From 056ede6cface66b400cd3b8e60ed077cc5b85c18 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Thu, 23 Jun 2005 12:21:28 -0700 Subject: [TCP]: Report congestion control algorithm in tcp_diag. Enhancement to the tcp_diag interface used by the iproute2 ss command to report the tcp congestion control being used by a socket. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/ipv4/tcp_diag.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'net/ipv4/tcp_diag.c') diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index a4e512036d88..f66945cb158f 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -65,6 +65,11 @@ static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, if (ext & (1<<(TCPDIAG_INFO-1))) info = TCPDIAG_PUT(skb, TCPDIAG_INFO, sizeof(*info)); + if (ext & (1<<(TCPDIAG_CONG-1))) { + size_t len = strlen(tp->ca_ops->name); + strcpy(TCPDIAG_PUT(skb, TCPDIAG_CONG, len+1), + tp->ca_ops->name); + } } r->tcpdiag_family = sk->sk_family; r->tcpdiag_state = sk->sk_state; -- cgit v1.2.1