diff options
Diffstat (limited to 'net')
292 files changed, 21634 insertions, 6391 deletions
diff --git a/net/802/fc.c b/net/802/fc.c index 640d34e026c2..282c4ab1abe6 100644 --- a/net/802/fc.c +++ b/net/802/fc.c @@ -87,7 +87,7 @@ static int fc_rebuild_header(struct sk_buff *skb) struct fch_hdr *fch=(struct fch_hdr *)skb->data; struct fcllc *fcllc=(struct fcllc *)(skb->data+sizeof(struct fch_hdr)); if(fcllc->ethertype != htons(ETH_P_IP)) { - printk("fc_rebuild_header: Don't know how to resolve type %04X addresses ?\n",(unsigned int)htons(fcllc->ethertype)); + printk("fc_rebuild_header: Don't know how to resolve type %04X addresses ?\n", ntohs(fcllc->ethertype)); return 0; } #ifdef CONFIG_INET diff --git a/net/802/fddi.c b/net/802/fddi.c index 5ce24c4bb840..ac242a4bc346 100644 --- a/net/802/fddi.c +++ b/net/802/fddi.c @@ -108,8 +108,8 @@ static int fddi_rebuild_header(struct sk_buff *skb) else #endif { - printk("%s: Don't know how to resolve type %02X addresses.\n", - skb->dev->name, htons(fddi->hdr.llc_snap.ethertype)); + printk("%s: Don't know how to resolve type %04X addresses.\n", + skb->dev->name, ntohs(fddi->hdr.llc_snap.ethertype)); return(0); } } diff --git a/net/802/hippi.c b/net/802/hippi.c index 051e8af56a77..6d7fed3dd99a 100644 --- a/net/802/hippi.c +++ b/net/802/hippi.c @@ -51,6 +51,7 @@ static int hippi_header(struct sk_buff *skb, struct net_device *dev, unsigned len) { struct hippi_hdr *hip = (struct hippi_hdr *)skb_push(skb, HIPPI_HLEN); + struct hippi_cb *hcb = (struct hippi_cb *) skb->cb; if (!len){ len = skb->len - HIPPI_HLEN; @@ -84,9 +85,10 @@ static int hippi_header(struct sk_buff *skb, struct net_device *dev, if (daddr) { memcpy(hip->le.dest_switch_addr, daddr + 3, 3); - memcpy(&skb->private.ifield, daddr + 2, 4); + memcpy(&hcb->ifield, daddr + 2, 4); return HIPPI_HLEN; } + hcb->ifield = 0; return -((int)HIPPI_HLEN); } @@ -122,7 +124,7 @@ static int hippi_rebuild_header(struct sk_buff *skb) * Determine the packet's protocol ID. */ -unsigned short hippi_type_trans(struct sk_buff *skb, struct net_device *dev) +__be16 hippi_type_trans(struct sk_buff *skb, struct net_device *dev) { struct hippi_hdr *hip; diff --git a/net/802/p8022.c b/net/802/p8022.c index 5ae63416df6d..b24817c63ca8 100644 --- a/net/802/p8022.c +++ b/net/802/p8022.c @@ -35,7 +35,8 @@ static int p8022_request(struct datalink_proto *dl, struct sk_buff *skb, struct datalink_proto *register_8022_client(unsigned char type, int (*func)(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt)) + struct packet_type *pt, + struct net_device *orig_dev)) { struct datalink_proto *proto; diff --git a/net/802/p8023.c b/net/802/p8023.c index a0b61b40225f..6368d3dce444 100644 --- a/net/802/p8023.c +++ b/net/802/p8023.c @@ -20,6 +20,7 @@ #include <linux/skbuff.h> #include <net/datalink.h> +#include <net/p8022.h> /* * Place an 802.3 header on a packet. The driver will do the mac diff --git a/net/802/psnap.c b/net/802/psnap.c index 1053821ddf93..ab80b1fab53c 100644 --- a/net/802/psnap.c +++ b/net/802/psnap.c @@ -47,7 +47,7 @@ static struct datalink_proto *find_snap_client(unsigned char *desc) * A SNAP packet has arrived */ static int snap_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt) + struct packet_type *pt, struct net_device *orig_dev) { int rc = 1; struct datalink_proto *proto; @@ -61,7 +61,7 @@ static int snap_rcv(struct sk_buff *skb, struct net_device *dev, /* Pass the frame on. */ skb->h.raw += 5; skb_pull(skb, 5); - rc = proto->rcvfunc(skb, dev, &snap_packet_type); + rc = proto->rcvfunc(skb, dev, &snap_packet_type, orig_dev); } else { skb->sk = NULL; kfree_skb(skb); @@ -118,7 +118,8 @@ module_exit(snap_exit); struct datalink_proto *register_snap_client(unsigned char *desc, int (*rcvfunc)(struct sk_buff *, struct net_device *, - struct packet_type *)) + struct packet_type *, + struct net_device *)) { struct datalink_proto *proto = NULL; diff --git a/net/802/sysctl_net_802.c b/net/802/sysctl_net_802.c index 36079630c49f..700129556c13 100644 --- a/net/802/sysctl_net_802.c +++ b/net/802/sysctl_net_802.c @@ -10,9 +10,10 @@ * 2 of the License, or (at your option) any later version. */ +#include <linux/config.h> #include <linux/mm.h> +#include <linux/if_tr.h> #include <linux/sysctl.h> -#include <linux/config.h> #ifdef CONFIG_TR extern int sysctl_tr_rif_timeout; diff --git a/net/8021q/vlan.h b/net/8021q/vlan.h index 508b1fa14546..9ae3a14dd016 100644 --- a/net/8021q/vlan.h +++ b/net/8021q/vlan.h @@ -51,7 +51,7 @@ struct net_device *__find_vlan_dev(struct net_device* real_dev, /* found in vlan_dev.c */ int vlan_dev_rebuild_header(struct sk_buff *skb); int vlan_skb_recv(struct sk_buff *skb, struct net_device *dev, - struct packet_type* ptype); + struct packet_type *ptype, struct net_device *orig_dev); int vlan_dev_hard_header(struct sk_buff *skb, struct net_device *dev, unsigned short type, void *daddr, void *saddr, unsigned len); diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 49c487413518..145f5cde96cf 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -113,7 +113,7 @@ static inline struct sk_buff *vlan_check_reorder_header(struct sk_buff *skb) * */ int vlan_skb_recv(struct sk_buff *skb, struct net_device *dev, - struct packet_type* ptype) + struct packet_type* ptype, struct net_device *orig_dev) { unsigned char *rawp = NULL; struct vlan_hdr *vhdr = (struct vlan_hdr *)(skb->data); diff --git a/net/Kconfig b/net/Kconfig index 40a31ba86d2c..c07aafb59a0f 100644 --- a/net/Kconfig +++ b/net/Kconfig @@ -147,6 +147,7 @@ source "net/bridge/netfilter/Kconfig" endif +source "net/dccp/Kconfig" source "net/sctp/Kconfig" source "net/atm/Kconfig" source "net/bridge/Kconfig" @@ -205,6 +206,8 @@ config NET_PKTGEN To compile this code as a module, choose M here: the module will be called pktgen. +source "net/netfilter/Kconfig" + endmenu endmenu diff --git a/net/Makefile b/net/Makefile index 8e2bdc025ab8..7e6eff206c81 100644 --- a/net/Makefile +++ b/net/Makefile @@ -16,6 +16,7 @@ obj-$(CONFIG_NET) += $(tmp-y) obj-$(CONFIG_LLC) += llc/ obj-$(CONFIG_NET) += ethernet/ 802/ sched/ netlink/ obj-$(CONFIG_INET) += ipv4/ +obj-$(CONFIG_NETFILTER) += netfilter/ obj-$(CONFIG_XFRM) += xfrm/ obj-$(CONFIG_UNIX) += unix/ ifneq ($(CONFIG_IPV6),) @@ -41,6 +42,7 @@ obj-$(CONFIG_ATM) += atm/ obj-$(CONFIG_DECNET) += decnet/ obj-$(CONFIG_ECONET) += econet/ obj-$(CONFIG_VLAN_8021Q) += 8021q/ +obj-$(CONFIG_IP_DCCP) += dccp/ obj-$(CONFIG_IP_SCTP) += sctp/ ifeq ($(CONFIG_NET),y) diff --git a/net/appletalk/aarp.c b/net/appletalk/aarp.c index c34614ea5fce..7076097debc2 100644 --- a/net/appletalk/aarp.c +++ b/net/appletalk/aarp.c @@ -698,7 +698,7 @@ static void __aarp_resolved(struct aarp_entry **list, struct aarp_entry *a, * frame. We currently only support Ethernet. */ static int aarp_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt) + struct packet_type *pt, struct net_device *orig_dev) { struct elapaarp *ea = aarp_hdr(skb); int hash, ret = 0; diff --git a/net/appletalk/ddp.c b/net/appletalk/ddp.c index 192b529f86a4..1d31b3a3f1e5 100644 --- a/net/appletalk/ddp.c +++ b/net/appletalk/ddp.c @@ -53,12 +53,12 @@ #include <linux/config.h> #include <linux/module.h> -#include <linux/tcp.h> #include <linux/if_arp.h> #include <linux/termios.h> /* For TIOCOUTQ/INQ */ #include <net/datalink.h> #include <net/psnap.h> #include <net/sock.h> +#include <net/tcp_states.h> #include <net/route.h> #include <linux/atalk.h> @@ -1390,7 +1390,7 @@ free_it: * [ie ARPHRD_ETHERTALK] */ static int atalk_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt) + struct packet_type *pt, struct net_device *orig_dev) { struct ddpehdr *ddp; struct sock *sock; @@ -1482,7 +1482,7 @@ freeit: * header and append a long one. */ static int ltalk_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt) + struct packet_type *pt, struct net_device *orig_dev) { /* Expand any short form frames */ if (skb->mac.raw[2] == 1) { @@ -1528,7 +1528,7 @@ static int ltalk_rcv(struct sk_buff *skb, struct net_device *dev, } skb->h.raw = skb->data; - return atalk_rcv(skb, dev, pt); + return atalk_rcv(skb, dev, pt, orig_dev); freeit: kfree_skb(skb); return 0; diff --git a/net/atm/ipcommon.c b/net/atm/ipcommon.c index 181a3002d8ad..4b1faca5013f 100644 --- a/net/atm/ipcommon.c +++ b/net/atm/ipcommon.c @@ -34,7 +34,6 @@ void skb_migrate(struct sk_buff_head *from,struct sk_buff_head *to) { - struct sk_buff *skb; unsigned long flags; struct sk_buff *skb_from = (struct sk_buff *) from; struct sk_buff *skb_to = (struct sk_buff *) to; @@ -47,8 +46,6 @@ void skb_migrate(struct sk_buff_head *from,struct sk_buff_head *to) prev->next = skb_to; to->prev->next = from->next; to->prev = from->prev; - for (skb = from->next; skb != skb_to; skb = skb->next) - skb->list = to; to->qlen += from->qlen; spin_unlock(&to->lock); from->prev = skb_from; diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c index a5c94f11547c..ea43dfb774e2 100644 --- a/net/ax25/af_ax25.c +++ b/net/ax25/af_ax25.c @@ -45,7 +45,7 @@ #include <linux/sysctl.h> #include <linux/init.h> #include <linux/spinlock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/ip.h> #include <net/arp.h> diff --git a/net/ax25/ax25_ds_in.c b/net/ax25/ax25_ds_in.c index 8adc0022cf58..edcaa897027c 100644 --- a/net/ax25/ax25_ds_in.c +++ b/net/ax25/ax25_ds_in.c @@ -22,8 +22,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/ip.h> /* For ip_rcv */ -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> diff --git a/net/ax25/ax25_ds_timer.c b/net/ax25/ax25_ds_timer.c index 3a8b67316fc3..061083efc1dc 100644 --- a/net/ax25/ax25_ds_timer.c +++ b/net/ax25/ax25_ds_timer.c @@ -18,7 +18,7 @@ #include <linux/string.h> #include <linux/sockios.h> #include <linux/net.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/ax25.h> #include <linux/inet.h> #include <linux/netdevice.h> diff --git a/net/ax25/ax25_in.c b/net/ax25/ax25_in.c index 3dc808fde33f..810c9c76c2e0 100644 --- a/net/ax25/ax25_in.c +++ b/net/ax25/ax25_in.c @@ -9,7 +9,6 @@ * Copyright (C) Joerg Reuter DL1BKE (jreuter@yaina.de) * Copyright (C) Hans-Joachim Hetscher DD8NE (dd8ne@bnv-bamberg.de) */ -#include <linux/config.h> #include <linux/errno.h> #include <linux/types.h> #include <linux/socket.h> @@ -26,9 +25,7 @@ #include <linux/skbuff.h> #include <linux/netfilter.h> #include <net/sock.h> -#include <net/ip.h> /* For ip_rcv */ -#include <net/tcp.h> -#include <net/arp.h> /* For arp_rcv */ +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> @@ -114,7 +111,6 @@ int ax25_rx_iframe(ax25_cb *ax25, struct sk_buff *skb) pid = *skb->data; -#ifdef CONFIG_INET if (pid == AX25_P_IP) { /* working around a TCP bug to keep additional listeners * happy. TCP re-uses the buffer and destroys the original @@ -132,10 +128,9 @@ int ax25_rx_iframe(ax25_cb *ax25, struct sk_buff *skb) skb->dev = ax25->ax25_dev->dev; skb->pkt_type = PACKET_HOST; skb->protocol = htons(ETH_P_IP); - ip_rcv(skb, skb->dev, NULL); /* Wrong ptype */ + netif_rx(skb); return 1; } -#endif if (pid == AX25_P_SEGMENT) { skb_pull(skb, 1); /* Remove PID */ return ax25_rx_fragment(ax25, skb); @@ -250,7 +245,6 @@ static int ax25_rcv(struct sk_buff *skb, struct net_device *dev, /* Now we are pointing at the pid byte */ switch (skb->data[1]) { -#ifdef CONFIG_INET case AX25_P_IP: skb_pull(skb,2); /* drop PID/CTRL */ skb->h.raw = skb->data; @@ -258,7 +252,7 @@ static int ax25_rcv(struct sk_buff *skb, struct net_device *dev, skb->dev = dev; skb->pkt_type = PACKET_HOST; skb->protocol = htons(ETH_P_IP); - ip_rcv(skb, dev, ptype); /* Note ptype here is the wrong one, fix me later */ + netif_rx(skb); break; case AX25_P_ARP: @@ -268,9 +262,8 @@ static int ax25_rcv(struct sk_buff *skb, struct net_device *dev, skb->dev = dev; skb->pkt_type = PACKET_HOST; skb->protocol = htons(ETH_P_ARP); - arp_rcv(skb, dev, ptype); /* Note ptype here is wrong... */ + netif_rx(skb); break; -#endif case AX25_P_TEXT: /* Now find a suitable dgram socket */ sk = ax25_get_socket(&dest, &src, SOCK_DGRAM); @@ -454,7 +447,7 @@ static int ax25_rcv(struct sk_buff *skb, struct net_device *dev, * Receive an AX.25 frame via a SLIP interface. */ int ax25_kiss_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *ptype) + struct packet_type *ptype, struct net_device *orig_dev) { skb->sk = NULL; /* Initially we don't know who it's for */ skb->destructor = NULL; /* Who initializes this, dammit?! */ diff --git a/net/ax25/ax25_std_in.c b/net/ax25/ax25_std_in.c index 7131873322c4..f6ed283e9de8 100644 --- a/net/ax25/ax25_std_in.c +++ b/net/ax25/ax25_std_in.c @@ -29,8 +29,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/ip.h> /* For ip_rcv */ -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> diff --git a/net/ax25/ax25_std_timer.c b/net/ax25/ax25_std_timer.c index 066897bc0749..a29c480a4dc1 100644 --- a/net/ax25/ax25_std_timer.c +++ b/net/ax25/ax25_std_timer.c @@ -24,7 +24,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> diff --git a/net/ax25/ax25_subr.c b/net/ax25/ax25_subr.c index 99694b57f6f5..c41dbe5fadee 100644 --- a/net/ax25/ax25_subr.c +++ b/net/ax25/ax25_subr.c @@ -24,7 +24,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> @@ -76,7 +76,7 @@ void ax25_requeue_frames(ax25_cb *ax25) if (skb_prev == NULL) skb_queue_head(&ax25->write_queue, skb); else - skb_append(skb_prev, skb); + skb_append(skb_prev, skb, &ax25->write_queue); skb_prev = skb; } } diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index ffa26c10bfe8..55dc42eac92c 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -191,7 +191,7 @@ static void hci_init_req(struct hci_dev *hdev, unsigned long opt) /* Special commands */ while ((skb = skb_dequeue(&hdev->driver_init))) { - skb->pkt_type = HCI_COMMAND_PKT; + bt_cb(skb)->pkt_type = HCI_COMMAND_PKT; skb->dev = (void *) hdev; skb_queue_tail(&hdev->cmd_q, skb); hci_sched_cmd(hdev); @@ -995,11 +995,11 @@ static int hci_send_frame(struct sk_buff *skb) return -ENODEV; } - BT_DBG("%s type %d len %d", hdev->name, skb->pkt_type, skb->len); + BT_DBG("%s type %d len %d", hdev->name, bt_cb(skb)->pkt_type, skb->len); if (atomic_read(&hdev->promisc)) { /* Time stamp */ - do_gettimeofday(&skb->stamp); + __net_timestamp(skb); hci_send_to_sock(hdev, skb); } @@ -1034,7 +1034,7 @@ int hci_send_cmd(struct hci_dev *hdev, __u16 ogf, __u16 ocf, __u32 plen, void *p BT_DBG("skb len %d", skb->len); - skb->pkt_type = HCI_COMMAND_PKT; + bt_cb(skb)->pkt_type = HCI_COMMAND_PKT; skb->dev = (void *) hdev; skb_queue_tail(&hdev->cmd_q, skb); hci_sched_cmd(hdev); @@ -1081,7 +1081,7 @@ int hci_send_acl(struct hci_conn *conn, struct sk_buff *skb, __u16 flags) BT_DBG("%s conn %p flags 0x%x", hdev->name, conn, flags); skb->dev = (void *) hdev; - skb->pkt_type = HCI_ACLDATA_PKT; + bt_cb(skb)->pkt_type = HCI_ACLDATA_PKT; hci_add_acl_hdr(skb, conn->handle, flags | ACL_START); if (!(list = skb_shinfo(skb)->frag_list)) { @@ -1103,7 +1103,7 @@ int hci_send_acl(struct hci_conn *conn, struct sk_buff *skb, __u16 flags) skb = list; list = list->next; skb->dev = (void *) hdev; - skb->pkt_type = HCI_ACLDATA_PKT; + bt_cb(skb)->pkt_type = HCI_ACLDATA_PKT; hci_add_acl_hdr(skb, conn->handle, flags | ACL_CONT); BT_DBG("%s frag %p len %d", hdev->name, skb, skb->len); @@ -1139,7 +1139,7 @@ int hci_send_sco(struct hci_conn *conn, struct sk_buff *skb) memcpy(skb->h.raw, &hdr, HCI_SCO_HDR_SIZE); skb->dev = (void *) hdev; - skb->pkt_type = HCI_SCODATA_PKT; + bt_cb(skb)->pkt_type = HCI_SCODATA_PKT; skb_queue_tail(&conn->data_q, skb); hci_sched_tx(hdev); return 0; @@ -1369,7 +1369,7 @@ void hci_rx_task(unsigned long arg) if (test_bit(HCI_INIT, &hdev->flags)) { /* Don't process data packets in this states. */ - switch (skb->pkt_type) { + switch (bt_cb(skb)->pkt_type) { case HCI_ACLDATA_PKT: case HCI_SCODATA_PKT: kfree_skb(skb); @@ -1378,7 +1378,7 @@ void hci_rx_task(unsigned long arg) } /* Process frame */ - switch (skb->pkt_type) { + switch (bt_cb(skb)->pkt_type) { case HCI_EVENT_PKT: hci_event_packet(hdev, skb); break; diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 46367bd129c3..d6da0939216d 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -484,14 +484,18 @@ static inline void hci_inquiry_complete_evt(struct hci_dev *hdev, struct sk_buff /* Inquiry Result */ static inline void hci_inquiry_result_evt(struct hci_dev *hdev, struct sk_buff *skb) { + struct inquiry_data data; struct inquiry_info *info = (struct inquiry_info *) (skb->data + 1); int num_rsp = *((__u8 *) skb->data); BT_DBG("%s num_rsp %d", hdev->name, num_rsp); + if (!num_rsp) + return; + hci_dev_lock(hdev); + for (; num_rsp; num_rsp--) { - struct inquiry_data data; bacpy(&data.bdaddr, &info->bdaddr); data.pscan_rep_mode = info->pscan_rep_mode; data.pscan_period_mode = info->pscan_period_mode; @@ -502,30 +506,55 @@ static inline void hci_inquiry_result_evt(struct hci_dev *hdev, struct sk_buff * info++; hci_inquiry_cache_update(hdev, &data); } + hci_dev_unlock(hdev); } /* Inquiry Result With RSSI */ static inline void hci_inquiry_result_with_rssi_evt(struct hci_dev *hdev, struct sk_buff *skb) { - struct inquiry_info_with_rssi *info = (struct inquiry_info_with_rssi *) (skb->data + 1); + struct inquiry_data data; int num_rsp = *((__u8 *) skb->data); BT_DBG("%s num_rsp %d", hdev->name, num_rsp); + if (!num_rsp) + return; + hci_dev_lock(hdev); - for (; num_rsp; num_rsp--) { - struct inquiry_data data; - bacpy(&data.bdaddr, &info->bdaddr); - data.pscan_rep_mode = info->pscan_rep_mode; - data.pscan_period_mode = info->pscan_period_mode; - data.pscan_mode = 0x00; - memcpy(data.dev_class, info->dev_class, 3); - data.clock_offset = info->clock_offset; - data.rssi = info->rssi; - info++; - hci_inquiry_cache_update(hdev, &data); + + if ((skb->len - 1) / num_rsp != sizeof(struct inquiry_info_with_rssi)) { + struct inquiry_info_with_rssi_and_pscan_mode *info = + (struct inquiry_info_with_rssi_and_pscan_mode *) (skb->data + 1); + + for (; num_rsp; num_rsp--) { + bacpy(&data.bdaddr, &info->bdaddr); + data.pscan_rep_mode = info->pscan_rep_mode; + data.pscan_period_mode = info->pscan_period_mode; + data.pscan_mode = info->pscan_mode; + memcpy(data.dev_class, info->dev_class, 3); + data.clock_offset = info->clock_offset; + data.rssi = info->rssi; + info++; + hci_inquiry_cache_update(hdev, &data); + } + } else { + struct inquiry_info_with_rssi *info = + (struct inquiry_info_with_rssi *) (skb->data + 1); + + for (; num_rsp; num_rsp--) { + bacpy(&data.bdaddr, &info->bdaddr); + data.pscan_rep_mode = info->pscan_rep_mode; + data.pscan_period_mode = info->pscan_period_mode; + data.pscan_mode = 0x00; + memcpy(data.dev_class, info->dev_class, 3); + data.clock_offset = info->clock_offset; + data.rssi = info->rssi; + info++; + hci_inquiry_cache_update(hdev, &data); + } } + hci_dev_unlock(hdev); } @@ -865,6 +894,24 @@ static inline void hci_clock_offset_evt(struct hci_dev *hdev, struct sk_buff *sk hci_dev_unlock(hdev); } +/* Page Scan Repetition Mode */ +static inline void hci_pscan_rep_mode_evt(struct hci_dev *hdev, struct sk_buff *skb) +{ + struct hci_ev_pscan_rep_mode *ev = (struct hci_ev_pscan_rep_mode *) skb->data; + struct inquiry_entry *ie; + + BT_DBG("%s", hdev->name); + + hci_dev_lock(hdev); + + if ((ie = hci_inquiry_cache_lookup(hdev, &ev->bdaddr))) { + ie->data.pscan_rep_mode = ev->pscan_rep_mode; + ie->timestamp = jiffies; + } + + hci_dev_unlock(hdev); +} + void hci_event_packet(struct hci_dev *hdev, struct sk_buff *skb) { struct hci_event_hdr *hdr = (struct hci_event_hdr *) skb->data; @@ -937,6 +984,10 @@ void hci_event_packet(struct hci_dev *hdev, struct sk_buff *skb) hci_clock_offset_evt(hdev, skb); break; + case HCI_EV_PSCAN_REP_MODE: + hci_pscan_rep_mode_evt(hdev, skb); + break; + case HCI_EV_CMD_STATUS: cs = (struct hci_ev_cmd_status *) skb->data; skb_pull(skb, sizeof(cs)); @@ -1036,9 +1087,9 @@ void hci_si_event(struct hci_dev *hdev, int type, int dlen, void *data) memcpy(ev->data, data, dlen); bt_cb(skb)->incoming = 1; - do_gettimeofday(&skb->stamp); + __net_timestamp(skb); - skb->pkt_type = HCI_EVENT_PKT; + bt_cb(skb)->pkt_type = HCI_EVENT_PKT; skb->dev = (void *) hdev; hci_send_to_sock(hdev, skb); kfree_skb(skb); diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index ebdcce5e7ca0..32ef7975a139 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -110,11 +110,11 @@ void hci_send_to_sock(struct hci_dev *hdev, struct sk_buff *skb) /* Apply filter */ flt = &hci_pi(sk)->filter; - if (!test_bit((skb->pkt_type == HCI_VENDOR_PKT) ? - 0 : (skb->pkt_type & HCI_FLT_TYPE_BITS), &flt->type_mask)) + if (!test_bit((bt_cb(skb)->pkt_type == HCI_VENDOR_PKT) ? + 0 : (bt_cb(skb)->pkt_type & HCI_FLT_TYPE_BITS), &flt->type_mask)) continue; - if (skb->pkt_type == HCI_EVENT_PKT) { + if (bt_cb(skb)->pkt_type == HCI_EVENT_PKT) { register int evt = (*(__u8 *)skb->data & HCI_FLT_EVENT_BITS); if (!hci_test_bit(evt, &flt->event_mask)) @@ -131,7 +131,7 @@ void hci_send_to_sock(struct hci_dev *hdev, struct sk_buff *skb) continue; /* Put type byte before the data */ - memcpy(skb_push(nskb, 1), &nskb->pkt_type, 1); + memcpy(skb_push(nskb, 1), &bt_cb(nskb)->pkt_type, 1); if (sock_queue_rcv_skb(sk, nskb)) kfree_skb(nskb); @@ -327,11 +327,17 @@ static inline void hci_sock_cmsg(struct sock *sk, struct msghdr *msg, struct sk_ { __u32 mask = hci_pi(sk)->cmsg_mask; - if (mask & HCI_CMSG_DIR) - put_cmsg(msg, SOL_HCI, HCI_CMSG_DIR, sizeof(int), &bt_cb(skb)->incoming); + if (mask & HCI_CMSG_DIR) { + int incoming = bt_cb(skb)->incoming; + put_cmsg(msg, SOL_HCI, HCI_CMSG_DIR, sizeof(incoming), &incoming); + } + + if (mask & HCI_CMSG_TSTAMP) { + struct timeval tv; - if (mask & HCI_CMSG_TSTAMP) - put_cmsg(msg, SOL_HCI, HCI_CMSG_TSTAMP, sizeof(skb->stamp), &skb->stamp); + skb_get_timestamp(skb, &tv); + put_cmsg(msg, SOL_HCI, HCI_CMSG_TSTAMP, sizeof(tv), &tv); + } } static int hci_sock_recvmsg(struct kiocb *iocb, struct socket *sock, @@ -405,11 +411,11 @@ static int hci_sock_sendmsg(struct kiocb *iocb, struct socket *sock, goto drop; } - skb->pkt_type = *((unsigned char *) skb->data); + bt_cb(skb)->pkt_type = *((unsigned char *) skb->data); skb_pull(skb, 1); skb->dev = (void *) hdev; - if (skb->pkt_type == HCI_COMMAND_PKT) { + if (bt_cb(skb)->pkt_type == HCI_COMMAND_PKT) { u16 opcode = __le16_to_cpu(get_unaligned((u16 *)skb->data)); u16 ogf = hci_opcode_ogf(opcode); u16 ocf = hci_opcode_ocf(opcode); diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c index 32fccfb5bfa5..d3d6bc547212 100644 --- a/net/bluetooth/l2cap.c +++ b/net/bluetooth/l2cap.c @@ -372,7 +372,7 @@ static struct proto l2cap_proto = { .obj_size = sizeof(struct l2cap_pinfo) }; -static struct sock *l2cap_sock_alloc(struct socket *sock, int proto, int prio) +static struct sock *l2cap_sock_alloc(struct socket *sock, int proto, unsigned int __nocast prio) { struct sock *sk; diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index 27bf5047cd33..173f46e8cdae 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -21,10 +21,6 @@ SOFTWARE IS DISCLAIMED. */ -/* - RPN support - Dirk Husemann <hud@zurich.ibm.com> -*/ - /* * Bluetooth RFCOMM core. * @@ -115,10 +111,10 @@ static void rfcomm_session_del(struct rfcomm_session *s); #define __get_mcc_len(b) ((b & 0xfe) >> 1) /* RPN macros */ -#define __rpn_line_settings(data, stop, parity) ((data & 0x3) | ((stop & 0x1) << 2) | ((parity & 0x3) << 3)) +#define __rpn_line_settings(data, stop, parity) ((data & 0x3) | ((stop & 0x1) << 2) | ((parity & 0x7) << 3)) #define __get_rpn_data_bits(line) ((line) & 0x3) #define __get_rpn_stop_bits(line) (((line) >> 2) & 0x1) -#define __get_rpn_parity(line) (((line) >> 3) & 0x3) +#define __get_rpn_parity(line) (((line) >> 3) & 0x7) static inline void rfcomm_schedule(uint event) { @@ -233,7 +229,7 @@ static void rfcomm_dlc_clear_state(struct rfcomm_dlc *d) d->rx_credits = RFCOMM_DEFAULT_CREDITS; } -struct rfcomm_dlc *rfcomm_dlc_alloc(int prio) +struct rfcomm_dlc *rfcomm_dlc_alloc(unsigned int __nocast prio) { struct rfcomm_dlc *d = kmalloc(sizeof(*d), prio); if (!d) @@ -780,10 +776,10 @@ static int rfcomm_send_pn(struct rfcomm_session *s, int cr, struct rfcomm_dlc *d return rfcomm_send_frame(s, buf, ptr - buf); } -static int rfcomm_send_rpn(struct rfcomm_session *s, int cr, u8 dlci, - u8 bit_rate, u8 data_bits, u8 stop_bits, - u8 parity, u8 flow_ctrl_settings, - u8 xon_char, u8 xoff_char, u16 param_mask) +int rfcomm_send_rpn(struct rfcomm_session *s, int cr, u8 dlci, + u8 bit_rate, u8 data_bits, u8 stop_bits, + u8 parity, u8 flow_ctrl_settings, + u8 xon_char, u8 xoff_char, u16 param_mask) { struct rfcomm_hdr *hdr; struct rfcomm_mcc *mcc; @@ -791,9 +787,9 @@ static int rfcomm_send_rpn(struct rfcomm_session *s, int cr, u8 dlci, u8 buf[16], *ptr = buf; BT_DBG("%p cr %d dlci %d bit_r 0x%x data_b 0x%x stop_b 0x%x parity 0x%x" - "flwc_s 0x%x xon_c 0x%x xoff_c 0x%x p_mask 0x%x", - s, cr, dlci, bit_rate, data_bits, stop_bits, parity, - flow_ctrl_settings, xon_char, xoff_char, param_mask); + " flwc_s 0x%x xon_c 0x%x xoff_c 0x%x p_mask 0x%x", + s, cr, dlci, bit_rate, data_bits, stop_bits, parity, + flow_ctrl_settings, xon_char, xoff_char, param_mask); hdr = (void *) ptr; ptr += sizeof(*hdr); hdr->addr = __addr(s->initiator, 0); @@ -1265,16 +1261,16 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ u8 xon_char = 0; u8 xoff_char = 0; u16 rpn_mask = RFCOMM_RPN_PM_ALL; - - BT_DBG("dlci %d cr %d len 0x%x bitr 0x%x line 0x%x flow 0x%x xonc 0x%x xoffc 0x%x pm 0x%x", - dlci, cr, len, rpn->bit_rate, rpn->line_settings, rpn->flow_ctrl, - rpn->xon_char, rpn->xoff_char, rpn->param_mask); - - if (!cr) + + BT_DBG("dlci %d cr %d len 0x%x bitr 0x%x line 0x%x flow 0x%x xonc 0x%x xoffc 0x%x pm 0x%x", + dlci, cr, len, rpn->bit_rate, rpn->line_settings, rpn->flow_ctrl, + rpn->xon_char, rpn->xoff_char, rpn->param_mask); + + if (!cr) return 0; - + if (len == 1) { - /* request: return default setting */ + /* This is a request, return default settings */ bit_rate = RFCOMM_RPN_BR_115200; data_bits = RFCOMM_RPN_DATA_8; stop_bits = RFCOMM_RPN_STOP_1; @@ -1282,11 +1278,12 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ flow_ctrl = RFCOMM_RPN_FLOW_NONE; xon_char = RFCOMM_RPN_XON_CHAR; xoff_char = RFCOMM_RPN_XOFF_CHAR; - goto rpn_out; } - /* check for sane values: ignore/accept bit_rate, 8 bits, 1 stop bit, no parity, - no flow control lines, normal XON/XOFF chars */ + + /* Check for sane values, ignore/accept bit_rate, 8 bits, 1 stop bit, + * no parity, no flow control lines, normal XON/XOFF chars */ + if (rpn->param_mask & RFCOMM_RPN_PM_BITRATE) { bit_rate = rpn->bit_rate; if (bit_rate != RFCOMM_RPN_BR_115200) { @@ -1295,6 +1292,7 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ rpn_mask ^= RFCOMM_RPN_PM_BITRATE; } } + if (rpn->param_mask & RFCOMM_RPN_PM_DATA) { data_bits = __get_rpn_data_bits(rpn->line_settings); if (data_bits != RFCOMM_RPN_DATA_8) { @@ -1303,6 +1301,7 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ rpn_mask ^= RFCOMM_RPN_PM_DATA; } } + if (rpn->param_mask & RFCOMM_RPN_PM_STOP) { stop_bits = __get_rpn_stop_bits(rpn->line_settings); if (stop_bits != RFCOMM_RPN_STOP_1) { @@ -1311,6 +1310,7 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ rpn_mask ^= RFCOMM_RPN_PM_STOP; } } + if (rpn->param_mask & RFCOMM_RPN_PM_PARITY) { parity = __get_rpn_parity(rpn->line_settings); if (parity != RFCOMM_RPN_PARITY_NONE) { @@ -1319,6 +1319,7 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ rpn_mask ^= RFCOMM_RPN_PM_PARITY; } } + if (rpn->param_mask & RFCOMM_RPN_PM_FLOW) { flow_ctrl = rpn->flow_ctrl; if (flow_ctrl != RFCOMM_RPN_FLOW_NONE) { @@ -1327,6 +1328,7 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ rpn_mask ^= RFCOMM_RPN_PM_FLOW; } } + if (rpn->param_mask & RFCOMM_RPN_PM_XON) { xon_char = rpn->xon_char; if (xon_char != RFCOMM_RPN_XON_CHAR) { @@ -1335,6 +1337,7 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ rpn_mask ^= RFCOMM_RPN_PM_XON; } } + if (rpn->param_mask & RFCOMM_RPN_PM_XOFF) { xoff_char = rpn->xoff_char; if (xoff_char != RFCOMM_RPN_XOFF_CHAR) { @@ -1345,9 +1348,8 @@ static int rfcomm_recv_rpn(struct rfcomm_session *s, int cr, int len, struct sk_ } rpn_out: - rfcomm_send_rpn(s, 0, dlci, - bit_rate, data_bits, stop_bits, parity, flow_ctrl, - xon_char, xoff_char, rpn_mask); + rfcomm_send_rpn(s, 0, dlci, bit_rate, data_bits, stop_bits, + parity, flow_ctrl, xon_char, xoff_char, rpn_mask); return 0; } @@ -1358,14 +1360,13 @@ static int rfcomm_recv_rls(struct rfcomm_session *s, int cr, struct sk_buff *skb u8 dlci = __get_dlci(rls->dlci); BT_DBG("dlci %d cr %d status 0x%x", dlci, cr, rls->status); - + if (!cr) return 0; - /* FIXME: We should probably do something with this - information here. But for now it's sufficient just - to reply -- Bluetooth 1.1 says it's mandatory to - recognise and respond to RLS */ + /* We should probably do something with this information here. But + * for now it's sufficient just to reply -- Bluetooth 1.1 says it's + * mandatory to recognise and respond to RLS */ rfcomm_send_rls(s, 0, dlci, rls->status); @@ -1381,7 +1382,7 @@ static int rfcomm_recv_msc(struct rfcomm_session *s, int cr, struct sk_buff *skb BT_DBG("dlci %d cr %d v24 0x%x", dlci, cr, msc->v24_sig); d = rfcomm_dlc_get(s, dlci); - if (!d) + if (!d) return 0; if (cr) { @@ -1389,7 +1390,7 @@ static int rfcomm_recv_msc(struct rfcomm_session *s, int cr, struct sk_buff *skb set_bit(RFCOMM_TX_THROTTLED, &d->flags); else clear_bit(RFCOMM_TX_THROTTLED, &d->flags); - + rfcomm_dlc_lock(d); if (d->modem_status) d->modem_status(d, msc->v24_sig); @@ -1398,7 +1399,7 @@ static int rfcomm_recv_msc(struct rfcomm_session *s, int cr, struct sk_buff *skb rfcomm_send_msc(s, 0, dlci, msc->v24_sig); d->mscex |= RFCOMM_MSCEX_RX; - } else + } else d->mscex |= RFCOMM_MSCEX_TX; return 0; diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index 63a123c5c41b..90e19eb6d3cc 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -284,7 +284,7 @@ static struct proto rfcomm_proto = { .obj_size = sizeof(struct rfcomm_pinfo) }; -static struct sock *rfcomm_sock_alloc(struct socket *sock, int proto, int prio) +static struct sock *rfcomm_sock_alloc(struct socket *sock, int proto, unsigned int __nocast prio) { struct rfcomm_dlc *d; struct sock *sk; diff --git a/net/bluetooth/rfcomm/tty.c b/net/bluetooth/rfcomm/tty.c index 6304590fd36a..1bca860a6109 100644 --- a/net/bluetooth/rfcomm/tty.c +++ b/net/bluetooth/rfcomm/tty.c @@ -286,7 +286,7 @@ static inline void rfcomm_set_owner_w(struct sk_buff *skb, struct rfcomm_dev *de skb->destructor = rfcomm_wfree; } -static struct sk_buff *rfcomm_wmalloc(struct rfcomm_dev *dev, unsigned long size, int priority) +static struct sk_buff *rfcomm_wmalloc(struct rfcomm_dev *dev, unsigned long size, unsigned int __nocast priority) { if (atomic_read(&dev->wmem_alloc) < rfcomm_room(dev->dlc)) { struct sk_buff *skb = alloc_skb(size, priority); @@ -528,9 +528,14 @@ static void rfcomm_dev_modem_status(struct rfcomm_dlc *dlc, u8 v24_sig) struct rfcomm_dev *dev = dlc->owner; if (!dev) return; - + BT_DBG("dlc %p dev %p v24_sig 0x%02x", dlc, dev, v24_sig); + if ((dev->modem_status & TIOCM_CD) && !(v24_sig & RFCOMM_V24_DV)) { + if (dev->tty && !C_CLOCAL(dev->tty)) + tty_hangup(dev->tty); + } + dev->modem_status = ((v24_sig & RFCOMM_V24_RTC) ? (TIOCM_DSR | TIOCM_DTR) : 0) | ((v24_sig & RFCOMM_V24_RTR) ? (TIOCM_RTS | TIOCM_CTS) : 0) | @@ -740,20 +745,143 @@ static int rfcomm_tty_ioctl(struct tty_struct *tty, struct file *filp, unsigned return -ENOIOCTLCMD; } -#define RELEVANT_IFLAG(iflag) (iflag & (IGNBRK|BRKINT|IGNPAR|PARMRK|INPCK)) - static void rfcomm_tty_set_termios(struct tty_struct *tty, struct termios *old) { - BT_DBG("tty %p", tty); + struct termios *new = (struct termios *) tty->termios; + int old_baud_rate = tty_termios_baud_rate(old); + int new_baud_rate = tty_termios_baud_rate(new); - if ((tty->termios->c_cflag == old->c_cflag) && - (RELEVANT_IFLAG(tty->termios->c_iflag) == RELEVANT_IFLAG(old->c_iflag))) - return; + u8 baud, data_bits, stop_bits, parity, x_on, x_off; + u16 changes = 0; + + struct rfcomm_dev *dev = (struct rfcomm_dev *) tty->driver_data; + + BT_DBG("tty %p termios %p", tty, old); + + /* Handle turning off CRTSCTS */ + if ((old->c_cflag & CRTSCTS) && !(new->c_cflag & CRTSCTS)) + BT_DBG("Turning off CRTSCTS unsupported"); + + /* Parity on/off and when on, odd/even */ + if (((old->c_cflag & PARENB) != (new->c_cflag & PARENB)) || + ((old->c_cflag & PARODD) != (new->c_cflag & PARODD)) ) { + changes |= RFCOMM_RPN_PM_PARITY; + BT_DBG("Parity change detected."); + } + + /* Mark and space parity are not supported! */ + if (new->c_cflag & PARENB) { + if (new->c_cflag & PARODD) { + BT_DBG("Parity is ODD"); + parity = RFCOMM_RPN_PARITY_ODD; + } else { + BT_DBG("Parity is EVEN"); + parity = RFCOMM_RPN_PARITY_EVEN; + } + } else { + BT_DBG("Parity is OFF"); + parity = RFCOMM_RPN_PARITY_NONE; + } + + /* Setting the x_on / x_off characters */ + if (old->c_cc[VSTOP] != new->c_cc[VSTOP]) { + BT_DBG("XOFF custom"); + x_on = new->c_cc[VSTOP]; + changes |= RFCOMM_RPN_PM_XON; + } else { + BT_DBG("XOFF default"); + x_on = RFCOMM_RPN_XON_CHAR; + } + + if (old->c_cc[VSTART] != new->c_cc[VSTART]) { + BT_DBG("XON custom"); + x_off = new->c_cc[VSTART]; + changes |= RFCOMM_RPN_PM_XOFF; + } else { + BT_DBG("XON default"); + x_off = RFCOMM_RPN_XOFF_CHAR; + } + + /* Handle setting of stop bits */ + if ((old->c_cflag & CSTOPB) != (new->c_cflag & CSTOPB)) + changes |= RFCOMM_RPN_PM_STOP; + + /* POSIX does not support 1.5 stop bits and RFCOMM does not + * support 2 stop bits. So a request for 2 stop bits gets + * translated to 1.5 stop bits */ + if (new->c_cflag & CSTOPB) { + stop_bits = RFCOMM_RPN_STOP_15; + } else { + stop_bits = RFCOMM_RPN_STOP_1; + } + + /* Handle number of data bits [5-8] */ + if ((old->c_cflag & CSIZE) != (new->c_cflag & CSIZE)) + changes |= RFCOMM_RPN_PM_DATA; + + switch (new->c_cflag & CSIZE) { + case CS5: + data_bits = RFCOMM_RPN_DATA_5; + break; + case CS6: + data_bits = RFCOMM_RPN_DATA_6; + break; + case CS7: + data_bits = RFCOMM_RPN_DATA_7; + break; + case CS8: + data_bits = RFCOMM_RPN_DATA_8; + break; + default: + data_bits = RFCOMM_RPN_DATA_8; + break; + } + + /* Handle baudrate settings */ + if (old_baud_rate != new_baud_rate) + changes |= RFCOMM_RPN_PM_BITRATE; - /* handle turning off CRTSCTS */ - if ((old->c_cflag & CRTSCTS) && !(tty->termios->c_cflag & CRTSCTS)) { - BT_DBG("turning off CRTSCTS"); + switch (new_baud_rate) { + case 2400: + baud = RFCOMM_RPN_BR_2400; + break; + case 4800: + baud = RFCOMM_RPN_BR_4800; + break; + case 7200: + baud = RFCOMM_RPN_BR_7200; + break; + case 9600: + baud = RFCOMM_RPN_BR_9600; + break; + case 19200: + baud = RFCOMM_RPN_BR_19200; + break; + case 38400: + baud = RFCOMM_RPN_BR_38400; + break; + case 57600: + baud = RFCOMM_RPN_BR_57600; + break; + case 115200: + baud = RFCOMM_RPN_BR_115200; + break; + case 230400: + baud = RFCOMM_RPN_BR_230400; + break; + default: + /* 9600 is standard accordinag to the RFCOMM specification */ + baud = RFCOMM_RPN_BR_9600; + break; + } + + if (changes) + rfcomm_send_rpn(dev->dlc->session, 1, dev->dlc->dlci, baud, + data_bits, stop_bits, parity, + RFCOMM_RPN_FLOW_NONE, x_on, x_off, changes); + + return; } static void rfcomm_tty_throttle(struct tty_struct *tty) @@ -761,7 +889,7 @@ static void rfcomm_tty_throttle(struct tty_struct *tty) struct rfcomm_dev *dev = (struct rfcomm_dev *) tty->driver_data; BT_DBG("tty %p dev %p", tty, dev); - + rfcomm_dlc_throttle(dev->dlc); } @@ -770,7 +898,7 @@ static void rfcomm_tty_unthrottle(struct tty_struct *tty) struct rfcomm_dev *dev = (struct rfcomm_dev *) tty->driver_data; BT_DBG("tty %p dev %p", tty, dev); - + rfcomm_dlc_unthrottle(dev->dlc); } @@ -841,35 +969,35 @@ static int rfcomm_tty_tiocmget(struct tty_struct *tty, struct file *filp) static int rfcomm_tty_tiocmset(struct tty_struct *tty, struct file *filp, unsigned int set, unsigned int clear) { - struct rfcomm_dev *dev = (struct rfcomm_dev *) tty->driver_data; - struct rfcomm_dlc *dlc = dev->dlc; - u8 v24_sig; + struct rfcomm_dev *dev = (struct rfcomm_dev *) tty->driver_data; + struct rfcomm_dlc *dlc = dev->dlc; + u8 v24_sig; BT_DBG("tty %p dev %p set 0x%02x clear 0x%02x", tty, dev, set, clear); - rfcomm_dlc_get_modem_status(dlc, &v24_sig); - - if (set & TIOCM_DSR || set & TIOCM_DTR) - v24_sig |= RFCOMM_V24_RTC; - if (set & TIOCM_RTS || set & TIOCM_CTS) - v24_sig |= RFCOMM_V24_RTR; - if (set & TIOCM_RI) - v24_sig |= RFCOMM_V24_IC; - if (set & TIOCM_CD) - v24_sig |= RFCOMM_V24_DV; - - if (clear & TIOCM_DSR || clear & TIOCM_DTR) - v24_sig &= ~RFCOMM_V24_RTC; - if (clear & TIOCM_RTS || clear & TIOCM_CTS) - v24_sig &= ~RFCOMM_V24_RTR; - if (clear & TIOCM_RI) - v24_sig &= ~RFCOMM_V24_IC; - if (clear & TIOCM_CD) - v24_sig &= ~RFCOMM_V24_DV; - - rfcomm_dlc_set_modem_status(dlc, v24_sig); - - return 0; + rfcomm_dlc_get_modem_status(dlc, &v24_sig); + + if (set & TIOCM_DSR || set & TIOCM_DTR) + v24_sig |= RFCOMM_V24_RTC; + if (set & TIOCM_RTS || set & TIOCM_CTS) + v24_sig |= RFCOMM_V24_RTR; + if (set & TIOCM_RI) + v24_sig |= RFCOMM_V24_IC; + if (set & TIOCM_CD) + v24_sig |= RFCOMM_V24_DV; + + if (clear & TIOCM_DSR || clear & TIOCM_DTR) + v24_sig &= ~RFCOMM_V24_RTC; + if (clear & TIOCM_RTS || clear & TIOCM_CTS) + v24_sig &= ~RFCOMM_V24_RTR; + if (clear & TIOCM_RI) + v24_sig &= ~RFCOMM_V24_IC; + if (clear & TIOCM_CD) + v24_sig &= ~RFCOMM_V24_DV; + + rfcomm_dlc_set_modem_status(dlc, v24_sig); + + return 0; } /* ---- TTY structure ---- */ diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index 746c11fc017e..ce7ab7dfa0b2 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -418,7 +418,7 @@ static struct proto sco_proto = { .obj_size = sizeof(struct sco_pinfo) }; -static struct sock *sco_sock_alloc(struct socket *sock, int proto, int prio) +static struct sock *sco_sock_alloc(struct socket *sock, int proto, unsigned int __nocast prio) { struct sock *sk; diff --git a/net/bridge/br_fdb.c b/net/bridge/br_fdb.c index e6c2200b7ca3..24396b914d11 100644 --- a/net/bridge/br_fdb.c +++ b/net/bridge/br_fdb.c @@ -23,7 +23,7 @@ #include <asm/atomic.h> #include "br_private.h" -static kmem_cache_t *br_fdb_cache; +static kmem_cache_t *br_fdb_cache __read_mostly; static int fdb_insert(struct net_bridge *br, struct net_bridge_port *source, const unsigned char *addr); diff --git a/net/bridge/netfilter/ebt_mark.c b/net/bridge/netfilter/ebt_mark.c index 02c632b4d325..c93d35ab95c0 100644 --- a/net/bridge/netfilter/ebt_mark.c +++ b/net/bridge/netfilter/ebt_mark.c @@ -23,10 +23,9 @@ static int ebt_target_mark(struct sk_buff **pskb, unsigned int hooknr, { struct ebt_mark_t_info *info = (struct ebt_mark_t_info *)data; - if ((*pskb)->nfmark != info->mark) { + if ((*pskb)->nfmark != info->mark) (*pskb)->nfmark = info->mark; - (*pskb)->nfcache |= NFC_ALTERED; - } + return info->target; } diff --git a/net/bridge/netfilter/ebt_ulog.c b/net/bridge/netfilter/ebt_ulog.c index 01af4fcef26d..aae26ae2e61f 100644 --- a/net/bridge/netfilter/ebt_ulog.c +++ b/net/bridge/netfilter/ebt_ulog.c @@ -78,8 +78,8 @@ static void ulog_send(unsigned int nlgroup) if (ub->qlen > 1) ub->lastnlh->nlmsg_type = NLMSG_DONE; - NETLINK_CB(ub->skb).dst_groups = 1 << nlgroup; - netlink_broadcast(ebtulognl, ub->skb, 0, 1 << nlgroup, GFP_ATOMIC); + NETLINK_CB(ub->skb).dst_group = nlgroup + 1; + netlink_broadcast(ebtulognl, ub->skb, 0, nlgroup + 1, GFP_ATOMIC); ub->qlen = 0; ub->skb = NULL; @@ -162,7 +162,7 @@ static void ebt_ulog(const struct sk_buff *skb, unsigned int hooknr, pm->version = EBT_ULOG_VERSION; do_gettimeofday(&pm->stamp); if (ub->qlen == 1) - ub->skb->stamp = pm->stamp; + skb_set_timestamp(ub->skb, &pm->stamp); pm->data_len = copy_len; pm->mark = skb->nfmark; pm->hook = hooknr; @@ -258,7 +258,8 @@ static int __init init(void) spin_lock_init(&ulog_buffers[i].lock); } - ebtulognl = netlink_kernel_create(NETLINK_NFLOG, NULL); + ebtulognl = netlink_kernel_create(NETLINK_NFLOG, EBT_ULOG_MAXNLGROUPS, + NULL, THIS_MODULE); if (!ebtulognl) ret = -ENOMEM; else if ((ret = ebt_register_watcher(&ulog))) diff --git a/net/core/Makefile b/net/core/Makefile index f5f5e58943e8..630da0f0579e 100644 --- a/net/core/Makefile +++ b/net/core/Makefile @@ -12,7 +12,6 @@ obj-y += dev.o ethtool.o dev_mcast.o dst.o \ obj-$(CONFIG_XFRM) += flow.o obj-$(CONFIG_SYSFS) += net-sysfs.o -obj-$(CONFIG_NETFILTER) += netfilter.o obj-$(CONFIG_NET_DIVERT) += dv.o obj-$(CONFIG_NET_PKTGEN) += pktgen.o obj-$(CONFIG_NET_RADIO) += wireless.o diff --git a/net/core/datagram.c b/net/core/datagram.c index fcee054b6f75..da9bf71421a7 100644 --- a/net/core/datagram.c +++ b/net/core/datagram.c @@ -43,7 +43,6 @@ #include <linux/errno.h> #include <linux/sched.h> #include <linux/inet.h> -#include <linux/tcp.h> #include <linux/netdevice.h> #include <linux/rtnetlink.h> #include <linux/poll.h> @@ -51,9 +50,10 @@ #include <net/protocol.h> #include <linux/skbuff.h> -#include <net/sock.h> -#include <net/checksum.h> +#include <net/checksum.h> +#include <net/sock.h> +#include <net/tcp_states.h> /* * Is a socket 'connection oriented' ? diff --git a/net/core/dev.c b/net/core/dev.c index faf59b02c4bf..c01511e3d0c1 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -267,10 +267,6 @@ void dev_add_pack(struct packet_type *pt) spin_unlock_bh(&ptype_lock); } -extern void linkwatch_run_queue(void); - - - /** * __dev_remove_pack - remove packet handler * @pt: packet type declaration @@ -1009,13 +1005,22 @@ void net_disable_timestamp(void) atomic_dec(&netstamp_needed); } -static inline void net_timestamp(struct timeval *stamp) +void __net_timestamp(struct sk_buff *skb) +{ + struct timeval tv; + + do_gettimeofday(&tv); + skb_set_timestamp(skb, &tv); +} +EXPORT_SYMBOL(__net_timestamp); + +static inline void net_timestamp(struct sk_buff *skb) { if (atomic_read(&netstamp_needed)) - do_gettimeofday(stamp); + __net_timestamp(skb); else { - stamp->tv_sec = 0; - stamp->tv_usec = 0; + skb->tstamp.off_sec = 0; + skb->tstamp.off_usec = 0; } } @@ -1027,7 +1032,8 @@ static inline void net_timestamp(struct timeval *stamp) void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev) { struct packet_type *ptype; - net_timestamp(&skb->stamp); + + net_timestamp(skb); rcu_read_lock(); list_for_each_entry_rcu(ptype, &ptype_all, list) { @@ -1058,7 +1064,7 @@ void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev) skb2->h.raw = skb2->nh.raw; skb2->pkt_type = PACKET_OUTGOING; - ptype->func(skb2, skb->dev, ptype); + ptype->func(skb2, skb->dev, ptype, skb->dev); } } rcu_read_unlock(); @@ -1123,8 +1129,6 @@ static inline int illegal_highdma(struct net_device *dev, struct sk_buff *skb) #define illegal_highdma(dev, skb) (0) #endif -extern void skb_release_data(struct sk_buff *); - /* Keep head the same: replace data */ int __skb_linearize(struct sk_buff *skb, unsigned int __nocast gfp_mask) { @@ -1379,8 +1383,8 @@ int netif_rx(struct sk_buff *skb) if (netpoll_rx(skb)) return NET_RX_DROP; - if (!skb->stamp.tv_sec) - net_timestamp(&skb->stamp); + if (!skb->tstamp.off_sec) + net_timestamp(skb); /* * The code is rearranged so that the path is the most @@ -1425,14 +1429,14 @@ int netif_rx_ni(struct sk_buff *skb) EXPORT_SYMBOL(netif_rx_ni); -static __inline__ void skb_bond(struct sk_buff *skb) +static inline struct net_device *skb_bond(struct sk_buff *skb) { struct net_device *dev = skb->dev; - if (dev->master) { - skb->real_dev = skb->dev; + if (dev->master) skb->dev = dev->master; - } + + return dev; } static void net_tx_action(struct softirq_action *h) @@ -1482,10 +1486,11 @@ static void net_tx_action(struct softirq_action *h) } static __inline__ int deliver_skb(struct sk_buff *skb, - struct packet_type *pt_prev) + struct packet_type *pt_prev, + struct net_device *orig_dev) { atomic_inc(&skb->users); - return pt_prev->func(skb, skb->dev, pt_prev); + return pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } #if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE) @@ -1496,7 +1501,8 @@ struct net_bridge_fdb_entry *(*br_fdb_get_hook)(struct net_bridge *br, void (*br_fdb_put_hook)(struct net_bridge_fdb_entry *ent); static __inline__ int handle_bridge(struct sk_buff **pskb, - struct packet_type **pt_prev, int *ret) + struct packet_type **pt_prev, int *ret, + struct net_device *orig_dev) { struct net_bridge_port *port; @@ -1505,14 +1511,14 @@ static __inline__ int handle_bridge(struct sk_buff **pskb, return 0; if (*pt_prev) { - *ret = deliver_skb(*pskb, *pt_prev); + *ret = deliver_skb(*pskb, *pt_prev, orig_dev); *pt_prev = NULL; } return br_handle_frame_hook(port, pskb); } #else -#define handle_bridge(skb, pt_prev, ret) (0) +#define handle_bridge(skb, pt_prev, ret, orig_dev) (0) #endif #ifdef CONFIG_NET_CLS_ACT @@ -1534,17 +1540,14 @@ static int ing_filter(struct sk_buff *skb) __u32 ttl = (__u32) G_TC_RTTL(skb->tc_verd); if (MAX_RED_LOOP < ttl++) { printk("Redir loop detected Dropping packet (%s->%s)\n", - skb->input_dev?skb->input_dev->name:"??",skb->dev->name); + skb->input_dev->name, skb->dev->name); return TC_ACT_SHOT; } skb->tc_verd = SET_TC_RTTL(skb->tc_verd,ttl); skb->tc_verd = SET_TC_AT(skb->tc_verd,AT_INGRESS); - if (NULL == skb->input_dev) { - skb->input_dev = skb->dev; - printk("ing_filter: fixed %s out %s\n",skb->input_dev->name,skb->dev->name); - } + spin_lock(&dev->ingress_lock); if ((q = dev->qdisc_ingress) != NULL) result = q->enqueue(skb, q); @@ -1559,6 +1562,7 @@ static int ing_filter(struct sk_buff *skb) int netif_receive_skb(struct sk_buff *skb) { struct packet_type *ptype, *pt_prev; + struct net_device *orig_dev; int ret = NET_RX_DROP; unsigned short type; @@ -1566,10 +1570,13 @@ int netif_receive_skb(struct sk_buff *skb) if (skb->dev->poll && netpoll_rx(skb)) return NET_RX_DROP; - if (!skb->stamp.tv_sec) - net_timestamp(&skb->stamp); + if (!skb->tstamp.off_sec) + net_timestamp(skb); + + if (!skb->input_dev) + skb->input_dev = skb->dev; - skb_bond(skb); + orig_dev = skb_bond(skb); __get_cpu_var(netdev_rx_stat).total++; @@ -1590,14 +1597,14 @@ int netif_receive_skb(struct sk_buff *skb) list_for_each_entry_rcu(ptype, &ptype_all, list) { if (!ptype->dev || ptype->dev == skb->dev) { if (pt_prev) - ret = deliver_skb(skb, pt_prev); + ret = deliver_skb(skb, pt_prev, orig_dev); pt_prev = ptype; } } #ifdef CONFIG_NET_CLS_ACT if (pt_prev) { - ret = deliver_skb(skb, pt_prev); + ret = deliver_skb(skb, pt_prev, orig_dev); pt_prev = NULL; /* noone else should process this after*/ } else { skb->tc_verd = SET_TC_OK2MUNGE(skb->tc_verd); @@ -1616,7 +1623,7 @@ ncls: handle_diverter(skb); - if (handle_bridge(&skb, &pt_prev, &ret)) + if (handle_bridge(&skb, &pt_prev, &ret, orig_dev)) goto out; type = skb->protocol; @@ -1624,13 +1631,13 @@ ncls: if (ptype->type == type && (!ptype->dev || ptype->dev == skb->dev)) { if (pt_prev) - ret = deliver_skb(skb, pt_prev); + ret = deliver_skb(skb, pt_prev, orig_dev); pt_prev = ptype; } } if (pt_prev) { - ret = pt_prev->func(skb, skb->dev, pt_prev); + ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } else { kfree_skb(skb); /* Jamal, now you will not able to escape explaining diff --git a/net/core/ethtool.c b/net/core/ethtool.c index a3eeb88e1c81..289c1b5a8e4a 100644 --- a/net/core/ethtool.c +++ b/net/core/ethtool.c @@ -81,6 +81,18 @@ int ethtool_op_set_tso(struct net_device *dev, u32 data) return 0; } +int ethtool_op_get_perm_addr(struct net_device *dev, struct ethtool_perm_addr *addr, u8 *data) +{ + unsigned char len = dev->addr_len; + if ( addr->size < len ) + return -ETOOSMALL; + + addr->size = len; + memcpy(data, dev->perm_addr, len); + return 0; +} + + /* Handlers for each ethtool command */ static int ethtool_get_settings(struct net_device *dev, void __user *useraddr) @@ -683,6 +695,39 @@ static int ethtool_get_stats(struct net_device *dev, void __user *useraddr) return ret; } +static int ethtool_get_perm_addr(struct net_device *dev, void *useraddr) +{ + struct ethtool_perm_addr epaddr; + u8 *data; + int ret; + + if (!dev->ethtool_ops->get_perm_addr) + return -EOPNOTSUPP; + + if (copy_from_user(&epaddr,useraddr,sizeof(epaddr))) + return -EFAULT; + + data = kmalloc(epaddr.size, GFP_USER); + if (!data) + return -ENOMEM; + + ret = dev->ethtool_ops->get_perm_addr(dev,&epaddr,data); + if (ret) + return ret; + + ret = -EFAULT; + if (copy_to_user(useraddr, &epaddr, sizeof(epaddr))) + goto out; + useraddr += sizeof(epaddr); + if (copy_to_user(useraddr, data, epaddr.size)) + goto out; + ret = 0; + + out: + kfree(data); + return ret; +} + /* The main entry point in this file. Called from net/core/dev.c */ int dev_ethtool(struct ifreq *ifr) @@ -806,6 +851,9 @@ int dev_ethtool(struct ifreq *ifr) case ETHTOOL_GSTATS: rc = ethtool_get_stats(dev, useraddr); break; + case ETHTOOL_GPERMADDR: + rc = ethtool_get_perm_addr(dev, useraddr); + break; default: rc = -EOPNOTSUPP; } @@ -826,6 +874,7 @@ int dev_ethtool(struct ifreq *ifr) EXPORT_SYMBOL(dev_ethtool); EXPORT_SYMBOL(ethtool_op_get_link); +EXPORT_SYMBOL_GPL(ethtool_op_get_perm_addr); EXPORT_SYMBOL(ethtool_op_get_sg); EXPORT_SYMBOL(ethtool_op_get_tso); EXPORT_SYMBOL(ethtool_op_get_tx_csum); diff --git a/net/core/flow.c b/net/core/flow.c index f289570b15a3..7e95b39de9fd 100644 --- a/net/core/flow.c +++ b/net/core/flow.c @@ -42,7 +42,7 @@ static DEFINE_PER_CPU(struct flow_cache_entry **, flow_tables) = { NULL }; #define flow_table(cpu) (per_cpu(flow_tables, cpu)) -static kmem_cache_t *flow_cachep; +static kmem_cache_t *flow_cachep __read_mostly; static int flow_lwm, flow_hwm; diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 1beb782ac41b..39fc55edf691 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -1217,7 +1217,7 @@ static void neigh_proxy_process(unsigned long arg) while (skb != (struct sk_buff *)&tbl->proxy_queue) { struct sk_buff *back = skb; - long tdif = back->stamp.tv_usec - now; + long tdif = NEIGH_CB(back)->sched_next - now; skb = skb->next; if (tdif <= 0) { @@ -1248,8 +1248,9 @@ void pneigh_enqueue(struct neigh_table *tbl, struct neigh_parms *p, kfree_skb(skb); return; } - skb->stamp.tv_sec = LOCALLY_ENQUEUED; - skb->stamp.tv_usec = sched_next; + + NEIGH_CB(skb)->sched_next = sched_next; + NEIGH_CB(skb)->flags |= LOCALLY_ENQUEUED; spin_lock(&tbl->proxy_queue.lock); if (del_timer(&tbl->proxy_timer)) { @@ -2342,8 +2343,8 @@ void neigh_app_ns(struct neighbour *n) } nlh = (struct nlmsghdr *)skb->data; nlh->nlmsg_flags = NLM_F_REQUEST; - NETLINK_CB(skb).dst_groups = RTMGRP_NEIGH; - netlink_broadcast(rtnl, skb, 0, RTMGRP_NEIGH, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = RTNLGRP_NEIGH; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_NEIGH, GFP_ATOMIC); } static void neigh_app_notify(struct neighbour *n) @@ -2360,8 +2361,8 @@ static void neigh_app_notify(struct neighbour *n) return; } nlh = (struct nlmsghdr *)skb->data; - NETLINK_CB(skb).dst_groups = RTMGRP_NEIGH; - netlink_broadcast(rtnl, skb, 0, RTMGRP_NEIGH, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = RTNLGRP_NEIGH; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_NEIGH, GFP_ATOMIC); } #endif /* CONFIG_ARPD */ diff --git a/net/core/netfilter.c b/net/core/netfilter.c deleted file mode 100644 index 076c156d5eda..000000000000 --- a/net/core/netfilter.c +++ /dev/null @@ -1,648 +0,0 @@ -/* netfilter.c: look after the filters for various protocols. - * Heavily influenced by the old firewall.c by David Bonn and Alan Cox. - * - * Thanks to Rob `CmdrTaco' Malda for not influencing this code in any - * way. - * - * Rusty Russell (C)2000 -- This code is GPL. - * - * February 2000: Modified by James Morris to have 1 queue per protocol. - * 15-Mar-2000: Added NF_REPEAT --RR. - * 08-May-2003: Internal logging interface added by Jozsef Kadlecsik. - */ -#include <linux/config.h> -#include <linux/kernel.h> -#include <linux/netfilter.h> -#include <net/protocol.h> -#include <linux/init.h> -#include <linux/skbuff.h> -#include <linux/wait.h> -#include <linux/module.h> -#include <linux/interrupt.h> -#include <linux/if.h> -#include <linux/netdevice.h> -#include <linux/inetdevice.h> -#include <linux/tcp.h> -#include <linux/udp.h> -#include <linux/icmp.h> -#include <net/sock.h> -#include <net/route.h> -#include <linux/ip.h> - -/* In this code, we can be waiting indefinitely for userspace to - * service a packet if a hook returns NF_QUEUE. We could keep a count - * of skbuffs queued for userspace, and not deregister a hook unless - * this is zero, but that sucks. Now, we simply check when the - * packets come back: if the hook is gone, the packet is discarded. */ -#ifdef CONFIG_NETFILTER_DEBUG -#define NFDEBUG(format, args...) printk(format , ## args) -#else -#define NFDEBUG(format, args...) -#endif - -/* Sockopts only registered and called from user context, so - net locking would be overkill. Also, [gs]etsockopt calls may - sleep. */ -static DECLARE_MUTEX(nf_sockopt_mutex); - -struct list_head nf_hooks[NPROTO][NF_MAX_HOOKS]; -static LIST_HEAD(nf_sockopts); -static DEFINE_SPINLOCK(nf_hook_lock); - -/* - * A queue handler may be registered for each protocol. Each is protected by - * long term mutex. The handler must provide an an outfn() to accept packets - * for queueing and must reinject all packets it receives, no matter what. - */ -static struct nf_queue_handler_t { - nf_queue_outfn_t outfn; - void *data; -} queue_handler[NPROTO]; -static DEFINE_RWLOCK(queue_handler_lock); - -int nf_register_hook(struct nf_hook_ops *reg) -{ - struct list_head *i; - - spin_lock_bh(&nf_hook_lock); - list_for_each(i, &nf_hooks[reg->pf][reg->hooknum]) { - if (reg->priority < ((struct nf_hook_ops *)i)->priority) - break; - } - list_add_rcu(®->list, i->prev); - spin_unlock_bh(&nf_hook_lock); - - synchronize_net(); - return 0; -} - -void nf_unregister_hook(struct nf_hook_ops *reg) -{ - spin_lock_bh(&nf_hook_lock); - list_del_rcu(®->list); - spin_unlock_bh(&nf_hook_lock); - - synchronize_net(); -} - -/* Do exclusive ranges overlap? */ -static inline int overlap(int min1, int max1, int min2, int max2) -{ - return max1 > min2 && min1 < max2; -} - -/* Functions to register sockopt ranges (exclusive). */ -int nf_register_sockopt(struct nf_sockopt_ops *reg) -{ - struct list_head *i; - int ret = 0; - - if (down_interruptible(&nf_sockopt_mutex) != 0) - return -EINTR; - - list_for_each(i, &nf_sockopts) { - struct nf_sockopt_ops *ops = (struct nf_sockopt_ops *)i; - if (ops->pf == reg->pf - && (overlap(ops->set_optmin, ops->set_optmax, - reg->set_optmin, reg->set_optmax) - || overlap(ops->get_optmin, ops->get_optmax, - reg->get_optmin, reg->get_optmax))) { - NFDEBUG("nf_sock overlap: %u-%u/%u-%u v %u-%u/%u-%u\n", - ops->set_optmin, ops->set_optmax, - ops->get_optmin, ops->get_optmax, - reg->set_optmin, reg->set_optmax, - reg->get_optmin, reg->get_optmax); - ret = -EBUSY; - goto out; - } - } - - list_add(®->list, &nf_sockopts); -out: - up(&nf_sockopt_mutex); - return ret; -} - -void nf_unregister_sockopt(struct nf_sockopt_ops *reg) -{ - /* No point being interruptible: we're probably in cleanup_module() */ - restart: - down(&nf_sockopt_mutex); - if (reg->use != 0) { - /* To be woken by nf_sockopt call... */ - /* FIXME: Stuart Young's name appears gratuitously. */ - set_current_state(TASK_UNINTERRUPTIBLE); - reg->cleanup_task = current; - up(&nf_sockopt_mutex); - schedule(); - goto restart; - } - list_del(®->list); - up(&nf_sockopt_mutex); -} - -/* Call get/setsockopt() */ -static int nf_sockopt(struct sock *sk, int pf, int val, - char __user *opt, int *len, int get) -{ - struct list_head *i; - struct nf_sockopt_ops *ops; - int ret; - - if (down_interruptible(&nf_sockopt_mutex) != 0) - return -EINTR; - - list_for_each(i, &nf_sockopts) { - ops = (struct nf_sockopt_ops *)i; - if (ops->pf == pf) { - if (get) { - if (val >= ops->get_optmin - && val < ops->get_optmax) { - ops->use++; - up(&nf_sockopt_mutex); - ret = ops->get(sk, val, opt, len); - goto out; - } - } else { - if (val >= ops->set_optmin - && val < ops->set_optmax) { - ops->use++; - up(&nf_sockopt_mutex); - ret = ops->set(sk, val, opt, *len); - goto out; - } - } - } - } - up(&nf_sockopt_mutex); - return -ENOPROTOOPT; - - out: - down(&nf_sockopt_mutex); - ops->use--; - if (ops->cleanup_task) - wake_up_process(ops->cleanup_task); - up(&nf_sockopt_mutex); - return ret; -} - -int nf_setsockopt(struct sock *sk, int pf, int val, char __user *opt, - int len) -{ - return nf_sockopt(sk, pf, val, opt, &len, 0); -} - -int nf_getsockopt(struct sock *sk, int pf, int val, char __user *opt, int *len) -{ - return nf_sockopt(sk, pf, val, opt, len, 1); -} - -static unsigned int nf_iterate(struct list_head *head, - struct sk_buff **skb, - int hook, - const struct net_device *indev, - const struct net_device *outdev, - struct list_head **i, - int (*okfn)(struct sk_buff *), - int hook_thresh) -{ - unsigned int verdict; - - /* - * The caller must not block between calls to this - * function because of risk of continuing from deleted element. - */ - list_for_each_continue_rcu(*i, head) { - struct nf_hook_ops *elem = (struct nf_hook_ops *)*i; - - if (hook_thresh > elem->priority) - continue; - - /* Optimization: we don't need to hold module - reference here, since function can't sleep. --RR */ - verdict = elem->hook(hook, skb, indev, outdev, okfn); - if (verdict != NF_ACCEPT) { -#ifdef CONFIG_NETFILTER_DEBUG - if (unlikely(verdict > NF_MAX_VERDICT)) { - NFDEBUG("Evil return from %p(%u).\n", - elem->hook, hook); - continue; - } -#endif - if (verdict != NF_REPEAT) - return verdict; - *i = (*i)->prev; - } - } - return NF_ACCEPT; -} - -int nf_register_queue_handler(int pf, nf_queue_outfn_t outfn, void *data) -{ - int ret; - - write_lock_bh(&queue_handler_lock); - if (queue_handler[pf].outfn) - ret = -EBUSY; - else { - queue_handler[pf].outfn = outfn; - queue_handler[pf].data = data; - ret = 0; - } - write_unlock_bh(&queue_handler_lock); - - return ret; -} - -/* The caller must flush their queue before this */ -int nf_unregister_queue_handler(int pf) -{ - write_lock_bh(&queue_handler_lock); - queue_handler[pf].outfn = NULL; - queue_handler[pf].data = NULL; - write_unlock_bh(&queue_handler_lock); - - return 0; -} - -/* - * Any packet that leaves via this function must come back - * through nf_reinject(). - */ -static int nf_queue(struct sk_buff *skb, - struct list_head *elem, - int pf, unsigned int hook, - struct net_device *indev, - struct net_device *outdev, - int (*okfn)(struct sk_buff *)) -{ - int status; - struct nf_info *info; -#ifdef CONFIG_BRIDGE_NETFILTER - struct net_device *physindev = NULL; - struct net_device *physoutdev = NULL; -#endif - - /* QUEUE == DROP if noone is waiting, to be safe. */ - read_lock(&queue_handler_lock); - if (!queue_handler[pf].outfn) { - read_unlock(&queue_handler_lock); - kfree_skb(skb); - return 1; - } - - info = kmalloc(sizeof(*info), GFP_ATOMIC); - if (!info) { - if (net_ratelimit()) - printk(KERN_ERR "OOM queueing packet %p\n", - skb); - read_unlock(&queue_handler_lock); - kfree_skb(skb); - return 1; - } - - *info = (struct nf_info) { - (struct nf_hook_ops *)elem, pf, hook, indev, outdev, okfn }; - - /* If it's going away, ignore hook. */ - if (!try_module_get(info->elem->owner)) { - read_unlock(&queue_handler_lock); - kfree(info); - return 0; - } - - /* Bump dev refs so they don't vanish while packet is out */ - if (indev) dev_hold(indev); - if (outdev) dev_hold(outdev); - -#ifdef CONFIG_BRIDGE_NETFILTER - if (skb->nf_bridge) { - physindev = skb->nf_bridge->physindev; - if (physindev) dev_hold(physindev); - physoutdev = skb->nf_bridge->physoutdev; - if (physoutdev) dev_hold(physoutdev); - } -#endif - - status = queue_handler[pf].outfn(skb, info, queue_handler[pf].data); - read_unlock(&queue_handler_lock); - - if (status < 0) { - /* James M doesn't say fuck enough. */ - if (indev) dev_put(indev); - if (outdev) dev_put(outdev); -#ifdef CONFIG_BRIDGE_NETFILTER - if (physindev) dev_put(physindev); - if (physoutdev) dev_put(physoutdev); -#endif - module_put(info->elem->owner); - kfree(info); - kfree_skb(skb); - return 1; - } - return 1; -} - -/* Returns 1 if okfn() needs to be executed by the caller, - * -EPERM for NF_DROP, 0 otherwise. */ -int nf_hook_slow(int pf, unsigned int hook, struct sk_buff **pskb, - struct net_device *indev, - struct net_device *outdev, - int (*okfn)(struct sk_buff *), - int hook_thresh) -{ - struct list_head *elem; - unsigned int verdict; - int ret = 0; - - /* We may already have this, but read-locks nest anyway */ - rcu_read_lock(); - - elem = &nf_hooks[pf][hook]; -next_hook: - verdict = nf_iterate(&nf_hooks[pf][hook], pskb, hook, indev, - outdev, &elem, okfn, hook_thresh); - if (verdict == NF_ACCEPT || verdict == NF_STOP) { - ret = 1; - goto unlock; - } else if (verdict == NF_DROP) { - kfree_skb(*pskb); - ret = -EPERM; - } else if (verdict == NF_QUEUE) { - NFDEBUG("nf_hook: Verdict = QUEUE.\n"); - if (!nf_queue(*pskb, elem, pf, hook, indev, outdev, okfn)) - goto next_hook; - } -unlock: - rcu_read_unlock(); - return ret; -} - -void nf_reinject(struct sk_buff *skb, struct nf_info *info, - unsigned int verdict) -{ - struct list_head *elem = &info->elem->list; - struct list_head *i; - - rcu_read_lock(); - - /* Release those devices we held, or Alexey will kill me. */ - if (info->indev) dev_put(info->indev); - if (info->outdev) dev_put(info->outdev); -#ifdef CONFIG_BRIDGE_NETFILTER - if (skb->nf_bridge) { - if (skb->nf_bridge->physindev) - dev_put(skb->nf_bridge->physindev); - if (skb->nf_bridge->physoutdev) - dev_put(skb->nf_bridge->physoutdev); - } -#endif - - /* Drop reference to owner of hook which queued us. */ - module_put(info->elem->owner); - - list_for_each_rcu(i, &nf_hooks[info->pf][info->hook]) { - if (i == elem) - break; - } - - if (elem == &nf_hooks[info->pf][info->hook]) { - /* The module which sent it to userspace is gone. */ - NFDEBUG("%s: module disappeared, dropping packet.\n", - __FUNCTION__); - verdict = NF_DROP; - } - - /* Continue traversal iff userspace said ok... */ - if (verdict == NF_REPEAT) { - elem = elem->prev; - verdict = NF_ACCEPT; - } - - if (verdict == NF_ACCEPT) { - next_hook: - verdict = nf_iterate(&nf_hooks[info->pf][info->hook], - &skb, info->hook, - info->indev, info->outdev, &elem, - info->okfn, INT_MIN); - } - - switch (verdict) { - case NF_ACCEPT: - info->okfn(skb); - break; - - case NF_QUEUE: - if (!nf_queue(skb, elem, info->pf, info->hook, - info->indev, info->outdev, info->okfn)) - goto next_hook; - break; - } - rcu_read_unlock(); - - if (verdict == NF_DROP) - kfree_skb(skb); - - kfree(info); - return; -} - -#ifdef CONFIG_INET -/* route_me_harder function, used by iptable_nat, iptable_mangle + ip_queue */ -int ip_route_me_harder(struct sk_buff **pskb) -{ - struct iphdr *iph = (*pskb)->nh.iph; - struct rtable *rt; - struct flowi fl = {}; - struct dst_entry *odst; - unsigned int hh_len; - - /* some non-standard hacks like ipt_REJECT.c:send_reset() can cause - * packets with foreign saddr to appear on the NF_IP_LOCAL_OUT hook. - */ - if (inet_addr_type(iph->saddr) == RTN_LOCAL) { - fl.nl_u.ip4_u.daddr = iph->daddr; - fl.nl_u.ip4_u.saddr = iph->saddr; - fl.nl_u.ip4_u.tos = RT_TOS(iph->tos); - fl.oif = (*pskb)->sk ? (*pskb)->sk->sk_bound_dev_if : 0; -#ifdef CONFIG_IP_ROUTE_FWMARK - fl.nl_u.ip4_u.fwmark = (*pskb)->nfmark; -#endif - fl.proto = iph->protocol; - if (ip_route_output_key(&rt, &fl) != 0) - return -1; - - /* Drop old route. */ - dst_release((*pskb)->dst); - (*pskb)->dst = &rt->u.dst; - } else { - /* non-local src, find valid iif to satisfy - * rp-filter when calling ip_route_input. */ - fl.nl_u.ip4_u.daddr = iph->saddr; - if (ip_route_output_key(&rt, &fl) != 0) - return -1; - - odst = (*pskb)->dst; - if (ip_route_input(*pskb, iph->daddr, iph->saddr, - RT_TOS(iph->tos), rt->u.dst.dev) != 0) { - dst_release(&rt->u.dst); - return -1; - } - dst_release(&rt->u.dst); - dst_release(odst); - } - - if ((*pskb)->dst->error) - return -1; - - /* Change in oif may mean change in hh_len. */ - hh_len = (*pskb)->dst->dev->hard_header_len; - if (skb_headroom(*pskb) < hh_len) { - struct sk_buff *nskb; - - nskb = skb_realloc_headroom(*pskb, hh_len); - if (!nskb) - return -1; - if ((*pskb)->sk) - skb_set_owner_w(nskb, (*pskb)->sk); - kfree_skb(*pskb); - *pskb = nskb; - } - - return 0; -} -EXPORT_SYMBOL(ip_route_me_harder); - -int skb_ip_make_writable(struct sk_buff **pskb, unsigned int writable_len) -{ - struct sk_buff *nskb; - - if (writable_len > (*pskb)->len) - return 0; - - /* Not exclusive use of packet? Must copy. */ - if (skb_shared(*pskb) || skb_cloned(*pskb)) - goto copy_skb; - - return pskb_may_pull(*pskb, writable_len); - -copy_skb: - nskb = skb_copy(*pskb, GFP_ATOMIC); - if (!nskb) - return 0; - BUG_ON(skb_is_nonlinear(nskb)); - - /* Rest of kernel will get very unhappy if we pass it a - suddenly-orphaned skbuff */ - if ((*pskb)->sk) - skb_set_owner_w(nskb, (*pskb)->sk); - kfree_skb(*pskb); - *pskb = nskb; - return 1; -} -EXPORT_SYMBOL(skb_ip_make_writable); -#endif /*CONFIG_INET*/ - -/* Internal logging interface, which relies on the real - LOG target modules */ - -#define NF_LOG_PREFIXLEN 128 - -static nf_logfn *nf_logging[NPROTO]; /* = NULL */ -static int reported = 0; -static DEFINE_SPINLOCK(nf_log_lock); - -int nf_log_register(int pf, nf_logfn *logfn) -{ - int ret = -EBUSY; - - /* Any setup of logging members must be done before - * substituting pointer. */ - spin_lock(&nf_log_lock); - if (!nf_logging[pf]) { - rcu_assign_pointer(nf_logging[pf], logfn); - ret = 0; - } - spin_unlock(&nf_log_lock); - return ret; -} - -void nf_log_unregister(int pf, nf_logfn *logfn) -{ - spin_lock(&nf_log_lock); - if (nf_logging[pf] == logfn) - nf_logging[pf] = NULL; - spin_unlock(&nf_log_lock); - - /* Give time to concurrent readers. */ - synchronize_net(); -} - -void nf_log_packet(int pf, - unsigned int hooknum, - const struct sk_buff *skb, - const struct net_device *in, - const struct net_device *out, - const char *fmt, ...) -{ - va_list args; - char prefix[NF_LOG_PREFIXLEN]; - nf_logfn *logfn; - - rcu_read_lock(); - logfn = rcu_dereference(nf_logging[pf]); - if (logfn) { - va_start(args, fmt); - vsnprintf(prefix, sizeof(prefix), fmt, args); - va_end(args); - /* We must read logging before nf_logfn[pf] */ - logfn(hooknum, skb, in, out, prefix); - } else if (!reported) { - printk(KERN_WARNING "nf_log_packet: can\'t log yet, " - "no backend logging module loaded in!\n"); - reported++; - } - rcu_read_unlock(); -} -EXPORT_SYMBOL(nf_log_register); -EXPORT_SYMBOL(nf_log_unregister); -EXPORT_SYMBOL(nf_log_packet); - -/* This does not belong here, but locally generated errors need it if connection - tracking in use: without this, connection may not be in hash table, and hence - manufactured ICMP or RST packets will not be associated with it. */ -void (*ip_ct_attach)(struct sk_buff *, struct sk_buff *); - -void nf_ct_attach(struct sk_buff *new, struct sk_buff *skb) -{ - void (*attach)(struct sk_buff *, struct sk_buff *); - - if (skb->nfct && (attach = ip_ct_attach) != NULL) { - mb(); /* Just to be sure: must be read before executing this */ - attach(new, skb); - } -} - -void __init netfilter_init(void) -{ - int i, h; - - for (i = 0; i < NPROTO; i++) { - for (h = 0; h < NF_MAX_HOOKS; h++) - INIT_LIST_HEAD(&nf_hooks[i][h]); - } -} - -EXPORT_SYMBOL(ip_ct_attach); -EXPORT_SYMBOL(nf_ct_attach); -EXPORT_SYMBOL(nf_getsockopt); -EXPORT_SYMBOL(nf_hook_slow); -EXPORT_SYMBOL(nf_hooks); -EXPORT_SYMBOL(nf_register_hook); -EXPORT_SYMBOL(nf_register_queue_handler); -EXPORT_SYMBOL(nf_register_sockopt); -EXPORT_SYMBOL(nf_reinject); -EXPORT_SYMBOL(nf_setsockopt); -EXPORT_SYMBOL(nf_unregister_hook); -EXPORT_SYMBOL(nf_unregister_queue_handler); -EXPORT_SYMBOL(nf_unregister_sockopt); diff --git a/net/core/request_sock.c b/net/core/request_sock.c index bb55675f0685..b8203de5ff07 100644 --- a/net/core/request_sock.c +++ b/net/core/request_sock.c @@ -32,7 +32,6 @@ * Further increasing requires to change hash table size. */ int sysctl_max_syn_backlog = 256; -EXPORT_SYMBOL(sysctl_max_syn_backlog); int reqsk_queue_alloc(struct request_sock_queue *queue, const int nr_table_entries) @@ -53,6 +52,8 @@ int reqsk_queue_alloc(struct request_sock_queue *queue, get_random_bytes(&lopt->hash_rnd, sizeof(lopt->hash_rnd)); rwlock_init(&queue->syn_wait_lock); queue->rskq_accept_head = queue->rskq_accept_head = NULL; + queue->rskq_defer_accept = 0; + lopt->nr_table_entries = nr_table_entries; write_lock_bh(&queue->syn_wait_lock); queue->listen_opt = lopt; @@ -62,3 +63,28 @@ int reqsk_queue_alloc(struct request_sock_queue *queue, } EXPORT_SYMBOL(reqsk_queue_alloc); + +void reqsk_queue_destroy(struct request_sock_queue *queue) +{ + /* make all the listen_opt local to us */ + struct listen_sock *lopt = reqsk_queue_yank_listen_sk(queue); + + if (lopt->qlen != 0) { + int i; + + for (i = 0; i < lopt->nr_table_entries; i++) { + struct request_sock *req; + + while ((req = lopt->syn_table[i]) != NULL) { + lopt->syn_table[i] = req->dl_next; + lopt->qlen--; + reqsk_free(req); + } + } + } + + BUG_TRAP(lopt->qlen == 0); + kfree(lopt); +} + +EXPORT_SYMBOL(reqsk_queue_destroy); diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 4b1bb30e6381..9bed7569ce3f 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -148,7 +148,7 @@ int rtnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo) { int err = 0; - NETLINK_CB(skb).dst_groups = group; + NETLINK_CB(skb).dst_group = group; if (echo) atomic_inc(&skb->users); netlink_broadcast(rtnl, skb, pid, group, GFP_KERNEL); @@ -458,8 +458,8 @@ void rtmsg_ifinfo(int type, struct net_device *dev, unsigned change) kfree_skb(skb); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_LINK; - netlink_broadcast(rtnl, skb, 0, RTMGRP_LINK, GFP_KERNEL); + NETLINK_CB(skb).dst_group = RTNLGRP_LINK; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_LINK, GFP_KERNEL); } static int rtnetlink_done(struct netlink_callback *cb) @@ -708,7 +708,8 @@ void __init rtnetlink_init(void) if (!rta_buf) panic("rtnetlink_init: cannot allocate rta_buf\n"); - rtnl = netlink_kernel_create(NETLINK_ROUTE, rtnetlink_rcv); + rtnl = netlink_kernel_create(NETLINK_ROUTE, RTNLGRP_MAX, rtnetlink_rcv, + THIS_MODULE); if (rtnl == NULL) panic("rtnetlink_init: cannot initialize rtnetlink\n"); netlink_set_nonroot(NETLINK_ROUTE, NL_NONROOT_RECV); diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 7eab867ede59..f80a28785610 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -68,7 +68,10 @@ #include <asm/uaccess.h> #include <asm/system.h> -static kmem_cache_t *skbuff_head_cache; +static kmem_cache_t *skbuff_head_cache __read_mostly; +static kmem_cache_t *skbuff_fclone_cache __read_mostly; + +struct timeval __read_mostly skb_tv_base; /* * Keep out-of-line to prevent kernel bloat. @@ -118,7 +121,7 @@ void skb_under_panic(struct sk_buff *skb, int sz, void *here) */ /** - * alloc_skb - allocate a network buffer + * __alloc_skb - allocate a network buffer * @size: size to allocate * @gfp_mask: allocation mask * @@ -129,14 +132,20 @@ void skb_under_panic(struct sk_buff *skb, int sz, void *here) * Buffers may only be allocated from interrupts using a @gfp_mask of * %GFP_ATOMIC. */ -struct sk_buff *alloc_skb(unsigned int size, unsigned int __nocast gfp_mask) +struct sk_buff *__alloc_skb(unsigned int size, unsigned int __nocast gfp_mask, + int fclone) { struct sk_buff *skb; u8 *data; /* Get the HEAD */ - skb = kmem_cache_alloc(skbuff_head_cache, - gfp_mask & ~__GFP_DMA); + if (fclone) + skb = kmem_cache_alloc(skbuff_fclone_cache, + gfp_mask & ~__GFP_DMA); + else + skb = kmem_cache_alloc(skbuff_head_cache, + gfp_mask & ~__GFP_DMA); + if (!skb) goto out; @@ -153,7 +162,15 @@ struct sk_buff *alloc_skb(unsigned int size, unsigned int __nocast gfp_mask) skb->data = data; skb->tail = data; skb->end = data + size; + if (fclone) { + struct sk_buff *child = skb + 1; + atomic_t *fclone_ref = (atomic_t *) (child + 1); + skb->fclone = SKB_FCLONE_ORIG; + atomic_set(fclone_ref, 1); + + child->fclone = SKB_FCLONE_UNAVAILABLE; + } atomic_set(&(skb_shinfo(skb)->dataref), 1); skb_shinfo(skb)->nr_frags = 0; skb_shinfo(skb)->tso_size = 0; @@ -266,8 +283,34 @@ void skb_release_data(struct sk_buff *skb) */ void kfree_skbmem(struct sk_buff *skb) { + struct sk_buff *other; + atomic_t *fclone_ref; + skb_release_data(skb); - kmem_cache_free(skbuff_head_cache, skb); + switch (skb->fclone) { + case SKB_FCLONE_UNAVAILABLE: + kmem_cache_free(skbuff_head_cache, skb); + break; + + case SKB_FCLONE_ORIG: + fclone_ref = (atomic_t *) (skb + 2); + if (atomic_dec_and_test(fclone_ref)) + kmem_cache_free(skbuff_fclone_cache, skb); + break; + + case SKB_FCLONE_CLONE: + fclone_ref = (atomic_t *) (skb + 1); + other = skb - 1; + + /* The clone portion is available for + * fast-cloning again. + */ + skb->fclone = SKB_FCLONE_UNAVAILABLE; + + if (atomic_dec_and_test(fclone_ref)) + kmem_cache_free(skbuff_fclone_cache, other); + break; + }; } /** @@ -281,8 +324,6 @@ void kfree_skbmem(struct sk_buff *skb) void __kfree_skb(struct sk_buff *skb) { - BUG_ON(skb->list != NULL); - dst_release(skb->dst); #ifdef CONFIG_XFRM secpath_put(skb->sp); @@ -302,7 +343,6 @@ void __kfree_skb(struct sk_buff *skb) skb->tc_index = 0; #ifdef CONFIG_NET_CLS_ACT skb->tc_verd = 0; - skb->tc_classid = 0; #endif #endif @@ -325,19 +365,27 @@ void __kfree_skb(struct sk_buff *skb) struct sk_buff *skb_clone(struct sk_buff *skb, unsigned int __nocast gfp_mask) { - struct sk_buff *n = kmem_cache_alloc(skbuff_head_cache, gfp_mask); - - if (!n) - return NULL; + struct sk_buff *n; + + n = skb + 1; + if (skb->fclone == SKB_FCLONE_ORIG && + n->fclone == SKB_FCLONE_UNAVAILABLE) { + atomic_t *fclone_ref = (atomic_t *) (n + 1); + n->fclone = SKB_FCLONE_CLONE; + atomic_inc(fclone_ref); + } else { + n = kmem_cache_alloc(skbuff_head_cache, gfp_mask); + if (!n) + return NULL; + n->fclone = SKB_FCLONE_UNAVAILABLE; + } #define C(x) n->x = skb->x n->next = n->prev = NULL; - n->list = NULL; n->sk = NULL; - C(stamp); + C(tstamp); C(dev); - C(real_dev); C(h); C(nh); C(mac); @@ -361,7 +409,6 @@ struct sk_buff *skb_clone(struct sk_buff *skb, unsigned int __nocast gfp_mask) n->destructor = NULL; #ifdef CONFIG_NETFILTER C(nfmark); - C(nfcache); C(nfct); nf_conntrack_get(skb->nfct); C(nfctinfo); @@ -370,9 +417,6 @@ struct sk_buff *skb_clone(struct sk_buff *skb, unsigned int __nocast gfp_mask) nf_bridge_get(skb->nf_bridge); #endif #endif /*CONFIG_NETFILTER*/ -#if defined(CONFIG_HIPPI) - C(private); -#endif #ifdef CONFIG_NET_SCHED C(tc_index); #ifdef CONFIG_NET_CLS_ACT @@ -380,7 +424,6 @@ struct sk_buff *skb_clone(struct sk_buff *skb, unsigned int __nocast gfp_mask) n->tc_verd = CLR_TC_OK2MUNGE(n->tc_verd); n->tc_verd = CLR_TC_MUNGED(n->tc_verd); C(input_dev); - C(tc_classid); #endif #endif @@ -404,10 +447,8 @@ static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old) */ unsigned long offset = new->data - old->data; - new->list = NULL; new->sk = NULL; new->dev = old->dev; - new->real_dev = old->real_dev; new->priority = old->priority; new->protocol = old->protocol; new->dst = dst_clone(old->dst); @@ -419,12 +460,12 @@ static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old) new->mac.raw = old->mac.raw + offset; memcpy(new->cb, old->cb, sizeof(old->cb)); new->local_df = old->local_df; + new->fclone = SKB_FCLONE_UNAVAILABLE; new->pkt_type = old->pkt_type; - new->stamp = old->stamp; + new->tstamp = old->tstamp; new->destructor = NULL; #ifdef CONFIG_NETFILTER new->nfmark = old->nfmark; - new->nfcache = old->nfcache; new->nfct = old->nfct; nf_conntrack_get(old->nfct); new->nfctinfo = old->nfctinfo; @@ -1344,50 +1385,43 @@ void skb_queue_tail(struct sk_buff_head *list, struct sk_buff *newsk) __skb_queue_tail(list, newsk); spin_unlock_irqrestore(&list->lock, flags); } + /** * skb_unlink - remove a buffer from a list * @skb: buffer to remove + * @list: list to use * - * Place a packet after a given packet in a list. The list locks are taken - * and this function is atomic with respect to other list locked calls + * Remove a packet from a list. The list locks are taken and this + * function is atomic with respect to other list locked calls * - * Works even without knowing the list it is sitting on, which can be - * handy at times. It also means that THE LIST MUST EXIST when you - * unlink. Thus a list must have its contents unlinked before it is - * destroyed. + * You must know what list the SKB is on. */ -void skb_unlink(struct sk_buff *skb) +void skb_unlink(struct sk_buff *skb, struct sk_buff_head *list) { - struct sk_buff_head *list = skb->list; - - if (list) { - unsigned long flags; + unsigned long flags; - spin_lock_irqsave(&list->lock, flags); - if (skb->list == list) - __skb_unlink(skb, skb->list); - spin_unlock_irqrestore(&list->lock, flags); - } + spin_lock_irqsave(&list->lock, flags); + __skb_unlink(skb, list); + spin_unlock_irqrestore(&list->lock, flags); } - /** * skb_append - append a buffer * @old: buffer to insert after * @newsk: buffer to insert + * @list: list to use * * Place a packet after a given packet in a list. The list locks are taken * and this function is atomic with respect to other list locked calls. * A buffer cannot be placed on two lists at the same time. */ - -void skb_append(struct sk_buff *old, struct sk_buff *newsk) +void skb_append(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list) { unsigned long flags; - spin_lock_irqsave(&old->list->lock, flags); - __skb_append(old, newsk); - spin_unlock_irqrestore(&old->list->lock, flags); + spin_lock_irqsave(&list->lock, flags); + __skb_append(old, newsk, list); + spin_unlock_irqrestore(&list->lock, flags); } @@ -1395,19 +1429,21 @@ void skb_append(struct sk_buff *old, struct sk_buff *newsk) * skb_insert - insert a buffer * @old: buffer to insert before * @newsk: buffer to insert + * @list: list to use + * + * Place a packet before a given packet in a list. The list locks are + * taken and this function is atomic with respect to other list locked + * calls. * - * Place a packet before a given packet in a list. The list locks are taken - * and this function is atomic with respect to other list locked calls * A buffer cannot be placed on two lists at the same time. */ - -void skb_insert(struct sk_buff *old, struct sk_buff *newsk) +void skb_insert(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list) { unsigned long flags; - spin_lock_irqsave(&old->list->lock, flags); - __skb_insert(newsk, old->prev, old, old->list); - spin_unlock_irqrestore(&old->list->lock, flags); + spin_lock_irqsave(&list->lock, flags); + __skb_insert(newsk, old->prev, old, list); + spin_unlock_irqrestore(&list->lock, flags); } #if 0 @@ -1663,12 +1699,23 @@ void __init skb_init(void) NULL, NULL); if (!skbuff_head_cache) panic("cannot create skbuff cache"); + + skbuff_fclone_cache = kmem_cache_create("skbuff_fclone_cache", + (2*sizeof(struct sk_buff)) + + sizeof(atomic_t), + 0, + SLAB_HWCACHE_ALIGN, + NULL, NULL); + if (!skbuff_fclone_cache) + panic("cannot create skbuff cache"); + + do_gettimeofday(&skb_tv_base); } EXPORT_SYMBOL(___pskb_trim); EXPORT_SYMBOL(__kfree_skb); EXPORT_SYMBOL(__pskb_pull_tail); -EXPORT_SYMBOL(alloc_skb); +EXPORT_SYMBOL(__alloc_skb); EXPORT_SYMBOL(pskb_copy); EXPORT_SYMBOL(pskb_expand_head); EXPORT_SYMBOL(skb_checksum); @@ -1696,3 +1743,4 @@ EXPORT_SYMBOL(skb_prepare_seq_read); EXPORT_SYMBOL(skb_seq_read); EXPORT_SYMBOL(skb_abort_seq_read); EXPORT_SYMBOL(skb_find_text); +EXPORT_SYMBOL(skb_tv_base); diff --git a/net/core/sock.c b/net/core/sock.c index 12f6d9a2a522..ccd10fd65682 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -260,7 +260,7 @@ int sock_setsockopt(struct socket *sock, int level, int optname, if (val > sysctl_wmem_max) val = sysctl_wmem_max; - +set_sndbuf: sk->sk_userlocks |= SOCK_SNDBUF_LOCK; if ((val * 2) < SOCK_MIN_SNDBUF) sk->sk_sndbuf = SOCK_MIN_SNDBUF; @@ -274,6 +274,13 @@ int sock_setsockopt(struct socket *sock, int level, int optname, sk->sk_write_space(sk); break; + case SO_SNDBUFFORCE: + if (!capable(CAP_NET_ADMIN)) { + ret = -EPERM; + break; + } + goto set_sndbuf; + case SO_RCVBUF: /* Don't error on this BSD doesn't and if you think about it this is right. Otherwise apps have to @@ -282,7 +289,7 @@ int sock_setsockopt(struct socket *sock, int level, int optname, if (val > sysctl_rmem_max) val = sysctl_rmem_max; - +set_rcvbuf: sk->sk_userlocks |= SOCK_RCVBUF_LOCK; /* FIXME: is this lower bound the right one? */ if ((val * 2) < SOCK_MIN_RCVBUF) @@ -291,6 +298,13 @@ int sock_setsockopt(struct socket *sock, int level, int optname, sk->sk_rcvbuf = val * 2; break; + case SO_RCVBUFFORCE: + if (!capable(CAP_NET_ADMIN)) { + ret = -EPERM; + break; + } + goto set_rcvbuf; + case SO_KEEPALIVE: #ifdef CONFIG_INET if (sk->sk_protocol == IPPROTO_TCP) @@ -686,6 +700,80 @@ void sk_free(struct sock *sk) module_put(owner); } +struct sock *sk_clone(const struct sock *sk, const unsigned int __nocast priority) +{ + struct sock *newsk = sk_alloc(sk->sk_family, priority, sk->sk_prot, 0); + + if (newsk != NULL) { + struct sk_filter *filter; + + memcpy(newsk, sk, sk->sk_prot->obj_size); + + /* SANITY */ + sk_node_init(&newsk->sk_node); + sock_lock_init(newsk); + bh_lock_sock(newsk); + + atomic_set(&newsk->sk_rmem_alloc, 0); + atomic_set(&newsk->sk_wmem_alloc, 0); + atomic_set(&newsk->sk_omem_alloc, 0); + skb_queue_head_init(&newsk->sk_receive_queue); + skb_queue_head_init(&newsk->sk_write_queue); + + rwlock_init(&newsk->sk_dst_lock); + rwlock_init(&newsk->sk_callback_lock); + + newsk->sk_dst_cache = NULL; + newsk->sk_wmem_queued = 0; + newsk->sk_forward_alloc = 0; + newsk->sk_send_head = NULL; + newsk->sk_backlog.head = newsk->sk_backlog.tail = NULL; + newsk->sk_userlocks = sk->sk_userlocks & ~SOCK_BINDPORT_LOCK; + + sock_reset_flag(newsk, SOCK_DONE); + skb_queue_head_init(&newsk->sk_error_queue); + + filter = newsk->sk_filter; + if (filter != NULL) + sk_filter_charge(newsk, filter); + + if (unlikely(xfrm_sk_clone_policy(newsk))) { + /* It is still raw copy of parent, so invalidate + * destructor and make plain sk_free() */ + newsk->sk_destruct = NULL; + sk_free(newsk); + newsk = NULL; + goto out; + } + + newsk->sk_err = 0; + newsk->sk_priority = 0; + atomic_set(&newsk->sk_refcnt, 2); + + /* + * Increment the counter in the same struct proto as the master + * sock (sk_refcnt_debug_inc uses newsk->sk_prot->socks, that + * is the same as sk->sk_prot->socks, as this field was copied + * with memcpy). + * + * This _changes_ the previous behaviour, where + * tcp_create_openreq_child always was incrementing the + * equivalent to tcp_prot->socks (inet_sock_nr), so this have + * to be taken into account in all callers. -acme + */ + sk_refcnt_debug_inc(newsk); + newsk->sk_socket = NULL; + newsk->sk_sleep = NULL; + + if (newsk->sk_prot->sockets_allocated) + atomic_inc(newsk->sk_prot->sockets_allocated); + } +out: + return newsk; +} + +EXPORT_SYMBOL_GPL(sk_clone); + void __init sk_init(void) { if (num_physpages <= 4096) { @@ -1353,11 +1441,7 @@ void sk_common_release(struct sock *sk) xfrm_sk_free_policy(sk); -#ifdef INET_REFCNT_DEBUG - if (atomic_read(&sk->sk_refcnt) != 1) - printk(KERN_DEBUG "Destruction of the socket %p delayed, c=%d\n", - sk, atomic_read(&sk->sk_refcnt)); -#endif + sk_refcnt_debug_release(sk); sock_put(sk); } @@ -1368,7 +1452,8 @@ static LIST_HEAD(proto_list); int proto_register(struct proto *prot, int alloc_slab) { - char *request_sock_slab_name; + char *request_sock_slab_name = NULL; + char *timewait_sock_slab_name; int rc = -ENOBUFS; if (alloc_slab) { @@ -1399,6 +1484,23 @@ int proto_register(struct proto *prot, int alloc_slab) goto out_free_request_sock_slab_name; } } + + if (prot->twsk_obj_size) { + static const char mask[] = "tw_sock_%s"; + + timewait_sock_slab_name = kmalloc(strlen(prot->name) + sizeof(mask) - 1, GFP_KERNEL); + + if (timewait_sock_slab_name == NULL) + goto out_free_request_sock_slab; + + sprintf(timewait_sock_slab_name, mask, prot->name); + prot->twsk_slab = kmem_cache_create(timewait_sock_slab_name, + prot->twsk_obj_size, + 0, SLAB_HWCACHE_ALIGN, + NULL, NULL); + if (prot->twsk_slab == NULL) + goto out_free_timewait_sock_slab_name; + } } write_lock(&proto_list_lock); @@ -1407,6 +1509,13 @@ int proto_register(struct proto *prot, int alloc_slab) rc = 0; out: return rc; +out_free_timewait_sock_slab_name: + kfree(timewait_sock_slab_name); +out_free_request_sock_slab: + if (prot->rsk_prot && prot->rsk_prot->slab) { + kmem_cache_destroy(prot->rsk_prot->slab); + prot->rsk_prot->slab = NULL; + } out_free_request_sock_slab_name: kfree(request_sock_slab_name); out_free_sock_slab: @@ -1434,6 +1543,14 @@ void proto_unregister(struct proto *prot) prot->rsk_prot->slab = NULL; } + if (prot->twsk_slab != NULL) { + const char *name = kmem_cache_name(prot->twsk_slab); + + kmem_cache_destroy(prot->twsk_slab); + kfree(name); + prot->twsk_slab = NULL; + } + list_del(&prot->node); write_unlock(&proto_list_lock); } diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index 8f817ad9f546..2f278c8e4743 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -9,23 +9,18 @@ #include <linux/sysctl.h> #include <linux/config.h> #include <linux/module.h> +#include <linux/socket.h> +#include <net/sock.h> #ifdef CONFIG_SYSCTL extern int netdev_max_backlog; -extern int netdev_budget; extern int weight_p; -extern int net_msg_cost; -extern int net_msg_burst; extern __u32 sysctl_wmem_max; extern __u32 sysctl_rmem_max; -extern __u32 sysctl_wmem_default; -extern __u32 sysctl_rmem_default; extern int sysctl_core_destroy_delay; -extern int sysctl_optmem_max; -extern int sysctl_somaxconn; #ifdef CONFIG_NET_DIVERT extern char sysctl_divert_version[]; diff --git a/net/core/utils.c b/net/core/utils.c index 88eb8b68e26b..7b5970fc9e40 100644 --- a/net/core/utils.c +++ b/net/core/utils.c @@ -16,7 +16,9 @@ #include <linux/module.h> #include <linux/jiffies.h> #include <linux/kernel.h> +#include <linux/inet.h> #include <linux/mm.h> +#include <linux/net.h> #include <linux/string.h> #include <linux/types.h> #include <linux/random.h> diff --git a/net/core/wireless.c b/net/core/wireless.c index 3ff5639c0b78..5caae2399f3a 100644 --- a/net/core/wireless.c +++ b/net/core/wireless.c @@ -571,10 +571,6 @@ static int wireless_seq_show(struct seq_file *seq, void *v) return 0; } -extern void *dev_seq_start(struct seq_file *seq, loff_t *pos); -extern void *dev_seq_next(struct seq_file *seq, void *v, loff_t *pos); -extern void dev_seq_stop(struct seq_file *seq, void *v); - static struct seq_operations wireless_seq_ops = { .start = dev_seq_start, .next = dev_seq_next, @@ -1144,8 +1140,8 @@ static inline void rtmsg_iwinfo(struct net_device * dev, kfree_skb(skb); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_LINK; - netlink_broadcast(rtnl, skb, 0, RTMGRP_LINK, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = RTNLGRP_LINK; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_LINK, GFP_ATOMIC); } #endif /* WE_EVENT_NETLINK */ diff --git a/net/dccp/Kconfig b/net/dccp/Kconfig new file mode 100644 index 000000000000..187ac182e24b --- /dev/null +++ b/net/dccp/Kconfig @@ -0,0 +1,50 @@ +menu "DCCP Configuration (EXPERIMENTAL)" + depends on INET && EXPERIMENTAL + +config IP_DCCP + tristate "The DCCP Protocol (EXPERIMENTAL)" + ---help--- + Datagram Congestion Control Protocol + + From draft-ietf-dccp-spec-11 <http://www.icir.org/kohler/dcp/draft-ietf-dccp-spec-11.txt>. + + The Datagram Congestion Control Protocol (DCCP) is a transport + protocol that implements bidirectional, unicast connections of + congestion-controlled, unreliable datagrams. It should be suitable + for use by applications such as streaming media, Internet telephony, + and on-line games + + To compile this protocol support as a module, choose M here: the + module will be called dccp. + + If in doubt, say N. + +config INET_DCCP_DIAG + depends on IP_DCCP && INET_DIAG + def_tristate y if (IP_DCCP = y && INET_DIAG = y) + def_tristate m + +source "net/dccp/ccids/Kconfig" + +menu "DCCP Kernel Hacking" + depends on IP_DCCP && DEBUG_KERNEL=y + +config IP_DCCP_DEBUG + bool "DCCP debug messages" + ---help--- + Only use this if you're hacking DCCP. + + Just say N. + +config IP_DCCP_UNLOAD_HACK + depends on IP_DCCP=m && IP_DCCP_CCID3=m + bool "DCCP control sock unload hack" + ---help--- + Enable this to be able to unload the dccp module when the it + has only one refcount held, the control sock one. Just execute + "rmmod dccp_ccid3 dccp" + + Just say N. +endmenu + +endmenu diff --git a/net/dccp/Makefile b/net/dccp/Makefile new file mode 100644 index 000000000000..fb97bb042455 --- /dev/null +++ b/net/dccp/Makefile @@ -0,0 +1,10 @@ +obj-$(CONFIG_IP_DCCP) += dccp.o + +dccp-y := ccid.o input.o ipv4.o minisocks.o options.o output.o proto.o \ + timer.o + +obj-$(CONFIG_INET_DCCP_DIAG) += dccp_diag.o + +dccp_diag-y := diag.o + +obj-y += ccids/ diff --git a/net/dccp/ccid.c b/net/dccp/ccid.c new file mode 100644 index 000000000000..9d8fc0e289ea --- /dev/null +++ b/net/dccp/ccid.c @@ -0,0 +1,139 @@ +/* + * net/dccp/ccid.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * CCID infrastructure + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include "ccid.h" + +static struct ccid *ccids[CCID_MAX]; +#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT) +static atomic_t ccids_lockct = ATOMIC_INIT(0); +static DEFINE_SPINLOCK(ccids_lock); + +/* + * The strategy is: modifications ccids vector are short, do not sleep and + * veeery rare, but read access should be free of any exclusive locks. + */ +static void ccids_write_lock(void) +{ + spin_lock(&ccids_lock); + while (atomic_read(&ccids_lockct) != 0) { + spin_unlock(&ccids_lock); + yield(); + spin_lock(&ccids_lock); + } +} + +static inline void ccids_write_unlock(void) +{ + spin_unlock(&ccids_lock); +} + +static inline void ccids_read_lock(void) +{ + atomic_inc(&ccids_lockct); + spin_unlock_wait(&ccids_lock); +} + +static inline void ccids_read_unlock(void) +{ + atomic_dec(&ccids_lockct); +} + +#else +#define ccids_write_lock() do { } while(0) +#define ccids_write_unlock() do { } while(0) +#define ccids_read_lock() do { } while(0) +#define ccids_read_unlock() do { } while(0) +#endif + +int ccid_register(struct ccid *ccid) +{ + int err; + + if (ccid->ccid_init == NULL) + return -1; + + ccids_write_lock(); + err = -EEXIST; + if (ccids[ccid->ccid_id] == NULL) { + ccids[ccid->ccid_id] = ccid; + err = 0; + } + ccids_write_unlock(); + if (err == 0) + pr_info("CCID: Registered CCID %d (%s)\n", + ccid->ccid_id, ccid->ccid_name); + return err; +} + +EXPORT_SYMBOL_GPL(ccid_register); + +int ccid_unregister(struct ccid *ccid) +{ + ccids_write_lock(); + ccids[ccid->ccid_id] = NULL; + ccids_write_unlock(); + pr_info("CCID: Unregistered CCID %d (%s)\n", + ccid->ccid_id, ccid->ccid_name); + return 0; +} + +EXPORT_SYMBOL_GPL(ccid_unregister); + +struct ccid *ccid_init(unsigned char id, struct sock *sk) +{ + struct ccid *ccid; + +#ifdef CONFIG_KMOD + if (ccids[id] == NULL) + request_module("net-dccp-ccid-%d", id); +#endif + ccids_read_lock(); + + ccid = ccids[id]; + if (ccid == NULL) + goto out; + + if (!try_module_get(ccid->ccid_owner)) + goto out_err; + + if (ccid->ccid_init(sk) != 0) + goto out_module_put; +out: + ccids_read_unlock(); + return ccid; +out_module_put: + module_put(ccid->ccid_owner); +out_err: + ccid = NULL; + goto out; +} + +EXPORT_SYMBOL_GPL(ccid_init); + +void ccid_exit(struct ccid *ccid, struct sock *sk) +{ + if (ccid == NULL) + return; + + ccids_read_lock(); + + if (ccids[ccid->ccid_id] != NULL) { + if (ccid->ccid_exit != NULL) + ccid->ccid_exit(sk); + module_put(ccid->ccid_owner); + } + + ccids_read_unlock(); +} + +EXPORT_SYMBOL_GPL(ccid_exit); diff --git a/net/dccp/ccid.h b/net/dccp/ccid.h new file mode 100644 index 000000000000..962f1e9e2f7e --- /dev/null +++ b/net/dccp/ccid.h @@ -0,0 +1,180 @@ +#ifndef _CCID_H +#define _CCID_H +/* + * net/dccp/ccid.h + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * CCID infrastructure + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <net/sock.h> +#include <linux/dccp.h> +#include <linux/list.h> +#include <linux/module.h> + +#define CCID_MAX 255 + +struct ccid { + unsigned char ccid_id; + const char *ccid_name; + struct module *ccid_owner; + int (*ccid_init)(struct sock *sk); + void (*ccid_exit)(struct sock *sk); + int (*ccid_hc_rx_init)(struct sock *sk); + int (*ccid_hc_tx_init)(struct sock *sk); + void (*ccid_hc_rx_exit)(struct sock *sk); + void (*ccid_hc_tx_exit)(struct sock *sk); + void (*ccid_hc_rx_packet_recv)(struct sock *sk, + struct sk_buff *skb); + int (*ccid_hc_rx_parse_options)(struct sock *sk, + unsigned char option, + unsigned char len, u16 idx, + unsigned char* value); + void (*ccid_hc_rx_insert_options)(struct sock *sk, + struct sk_buff *skb); + void (*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, + unsigned char option, + 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); + void (*ccid_hc_rx_get_info)(struct sock *sk, + struct tcp_info *info); + void (*ccid_hc_tx_get_info)(struct sock *sk, + struct tcp_info *info); +}; + +extern int ccid_register(struct ccid *ccid); +extern int ccid_unregister(struct ccid *ccid); + +extern struct ccid *ccid_init(unsigned char id, struct sock *sk); +extern void ccid_exit(struct ccid *ccid, struct sock *sk); + +static inline void __ccid_get(struct ccid *ccid) +{ + __module_get(ccid->ccid_owner); +} + +static inline int ccid_hc_tx_send_packet(struct ccid *ccid, struct sock *sk, + struct sk_buff *skb, int len) +{ + int rc = 0; + if (ccid->ccid_hc_tx_send_packet != NULL) + rc = ccid->ccid_hc_tx_send_packet(sk, skb, len); + return rc; +} + +static inline void ccid_hc_tx_packet_sent(struct ccid *ccid, struct sock *sk, + int more, int len) +{ + if (ccid->ccid_hc_tx_packet_sent != NULL) + ccid->ccid_hc_tx_packet_sent(sk, more, len); +} + +static inline int ccid_hc_rx_init(struct ccid *ccid, struct sock *sk) +{ + int rc = 0; + if (ccid->ccid_hc_rx_init != NULL) + rc = ccid->ccid_hc_rx_init(sk); + return rc; +} + +static inline int ccid_hc_tx_init(struct ccid *ccid, struct sock *sk) +{ + int rc = 0; + if (ccid->ccid_hc_tx_init != NULL) + rc = ccid->ccid_hc_tx_init(sk); + return rc; +} + +static inline void ccid_hc_rx_exit(struct ccid *ccid, struct sock *sk) +{ + if (ccid->ccid_hc_rx_exit != NULL && + dccp_sk(sk)->dccps_hc_rx_ccid_private != NULL) + ccid->ccid_hc_rx_exit(sk); +} + +static inline void ccid_hc_tx_exit(struct ccid *ccid, struct sock *sk) +{ + if (ccid->ccid_hc_tx_exit != NULL && + dccp_sk(sk)->dccps_hc_tx_ccid_private != NULL) + ccid->ccid_hc_tx_exit(sk); +} + +static inline void ccid_hc_rx_packet_recv(struct ccid *ccid, struct sock *sk, + struct sk_buff *skb) +{ + if (ccid->ccid_hc_rx_packet_recv != NULL) + ccid->ccid_hc_rx_packet_recv(sk, skb); +} + +static inline void ccid_hc_tx_packet_recv(struct ccid *ccid, struct sock *sk, + struct sk_buff *skb) +{ + if (ccid->ccid_hc_tx_packet_recv != NULL) + ccid->ccid_hc_tx_packet_recv(sk, skb); +} + +static inline int ccid_hc_tx_parse_options(struct ccid *ccid, struct sock *sk, + unsigned char option, + unsigned char len, u16 idx, + unsigned char* value) +{ + int rc = 0; + if (ccid->ccid_hc_tx_parse_options != NULL) + rc = ccid->ccid_hc_tx_parse_options(sk, option, len, idx, + value); + return rc; +} + +static inline int ccid_hc_rx_parse_options(struct ccid *ccid, struct sock *sk, + unsigned char option, + unsigned char len, u16 idx, + unsigned char* value) +{ + int rc = 0; + if (ccid->ccid_hc_rx_parse_options != NULL) + rc = ccid->ccid_hc_rx_parse_options(sk, option, len, idx, value); + return rc; +} + +static inline void ccid_hc_tx_insert_options(struct ccid *ccid, struct sock *sk, + struct sk_buff *skb) +{ + if (ccid->ccid_hc_tx_insert_options != NULL) + ccid->ccid_hc_tx_insert_options(sk, skb); +} + +static inline void ccid_hc_rx_insert_options(struct ccid *ccid, struct sock *sk, + struct sk_buff *skb) +{ + if (ccid->ccid_hc_rx_insert_options != NULL) + ccid->ccid_hc_rx_insert_options(sk, skb); +} + +static inline void ccid_hc_rx_get_info(struct ccid *ccid, struct sock *sk, + struct tcp_info *info) +{ + if (ccid->ccid_hc_rx_get_info != NULL) + ccid->ccid_hc_rx_get_info(sk, info); +} + +static inline void ccid_hc_tx_get_info(struct ccid *ccid, struct sock *sk, + struct tcp_info *info) +{ + if (ccid->ccid_hc_tx_get_info != NULL) + ccid->ccid_hc_tx_get_info(sk, info); +} +#endif /* _CCID_H */ diff --git a/net/dccp/ccids/Kconfig b/net/dccp/ccids/Kconfig new file mode 100644 index 000000000000..7684d83946a4 --- /dev/null +++ b/net/dccp/ccids/Kconfig @@ -0,0 +1,29 @@ +menu "DCCP CCIDs Configuration (EXPERIMENTAL)" + depends on IP_DCCP && EXPERIMENTAL + +config IP_DCCP_CCID3 + tristate "CCID3 (TFRC) (EXPERIMENTAL)" + depends on IP_DCCP + ---help--- + CCID 3 denotes TCP-Friendly Rate Control (TFRC), an equation-based + rate-controlled congestion control mechanism. TFRC is designed to + be reasonably fair when competing for bandwidth with TCP-like flows, + where a flow is "reasonably fair" if its sending rate is generally + within a factor of two of the sending rate of a TCP flow under the + same conditions. However, TFRC has a much lower variation of + throughput over time compared with TCP, which makes CCID 3 more + suitable than CCID 2 for applications such streaming media where a + relatively smooth sending rate is of importance. + + CCID 3 is further described in [CCID 3 PROFILE]. The TFRC + congestion control algorithms were initially described in RFC 3448. + + This text was extracted from draft-ietf-dccp-spec-11.txt. + + If in doubt, say M. + +config IP_DCCP_TFRC_LIB + depends on IP_DCCP_CCID3 + def_tristate IP_DCCP_CCID3 + +endmenu diff --git a/net/dccp/ccids/Makefile b/net/dccp/ccids/Makefile new file mode 100644 index 000000000000..956f79f50743 --- /dev/null +++ b/net/dccp/ccids/Makefile @@ -0,0 +1,5 @@ +obj-$(CONFIG_IP_DCCP_CCID3) += dccp_ccid3.o + +dccp_ccid3-y := ccid3.o + +obj-y += lib/ diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c new file mode 100644 index 000000000000..7bf3b3a91e97 --- /dev/null +++ b/net/dccp/ccids/ccid3.c @@ -0,0 +1,1221 @@ +/* + * net/dccp/ccids/ccid3.c + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * + * An implementation of the DCCP protocol + * + * This code has been developed by the University of Waikato WAND + * research group. For further information please see http://www.wand.net.nz/ + * + * This code also uses code from Lulea University, rereleased as GPL by its + * authors: + * Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon + * + * Changes to meet Linux coding standards, to make it meet latest ccid3 draft + * and to make it work as a loadable module in the DCCP stack written by + * Arnaldo Carvalho de Melo <acme@conectiva.com.br>. + * + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include <linux/config.h> +#include "../ccid.h" +#include "../dccp.h" +#include "lib/packet_history.h" +#include "lib/loss_interval.h" +#include "lib/tfrc.h" +#include "ccid3.h" + +/* + * Reason for maths with 10 here is to avoid 32 bit overflow when a is big. + */ +static inline u32 usecs_div(const u32 a, const u32 b) +{ + const u32 tmp = a * (USEC_PER_SEC / 10); + return b > 20 ? tmp / (b / 10) : tmp; +} + +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) +#else +#define ccid3_pr_debug(format, a...) +#endif + +static struct dccp_tx_hist *ccid3_tx_hist; +static struct dccp_rx_hist *ccid3_rx_hist; +static struct dccp_li_hist *ccid3_li_hist; + +static int ccid3_init(struct sock *sk) +{ + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + return 0; +} + +static void ccid3_exit(struct sock *sk) +{ + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); +} + +/* 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 +static const char *ccid3_tx_state_name(enum ccid3_hc_tx_states state) +{ + static char *ccid3_state_names[] = { + [TFRC_SSTATE_NO_SENT] = "NO_SENT", + [TFRC_SSTATE_NO_FBACK] = "NO_FBACK", + [TFRC_SSTATE_FBACK] = "FBACK", + [TFRC_SSTATE_TERM] = "TERM", + }; + + return ccid3_state_names[state]; +} +#endif + +static inline void ccid3_hc_tx_set_state(struct sock *sk, + enum ccid3_hc_tx_states state) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + enum ccid3_hc_tx_states oldstate = hctx->ccid3hctx_state; + + ccid3_pr_debug("%s(%p) %-8.8s -> %s\n", + dccp_role(sk), sk, ccid3_tx_state_name(oldstate), + ccid3_tx_state_name(state)); + WARN_ON(state == oldstate); + 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) +{ + /* + * 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); +} + +/* 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) +{ + 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 = 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) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + + /* 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; + + do_gettimeofday(&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; + } + } +} + +static void ccid3_hc_tx_no_feedback_timer(unsigned long data) +{ + struct sock *sk = (struct sock *)data; + struct dccp_sock *dp = dccp_sk(sk); + unsigned long next_tmout = 0; + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + + 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; + } + + ccid3_pr_debug("%s, sk=%p, state=%s\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, + 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 + */ + break; + case TFRC_SSTATE_FBACK: + /* + * Check if IDLE since last timeout and recv rate is less than + * 4 packets 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", + dccp_role(sk), sk, + ccid3_tx_state_name(hctx->ccid3hctx_state)); + /* Halve sending rate */ + + /* If (X_calc > 2 * X_recv) + * X_recv = max(X_recv / 2, s / (2 * t_mbi)); + * Else + * X_recv = X_calc / 4; + */ + 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); + } + /* + * Schedule no feedback timer to expire in + * max(4 * R, 2 * s / X) + */ + next_tmout = max_t(u32, hctx->ccid3hctx_t_rto, + 2 * usecs_div(hctx->ccid3hctx_s, + hctx->ccid3hctx_x)); + break; + default: + printk(KERN_CRIT "%s: %s, sk=%p, Illegal state (%d)!\n", + __FUNCTION__, dccp_role(sk), sk, hctx->ccid3hctx_state); + dump_stack(); + 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; +out: + bh_unlock_sock(sk); + sock_put(sk); +} + +static int ccid3_hc_tx_send_packet(struct sock *sk, + struct sk_buff *skb, int len) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + struct dccp_tx_hist_entry *new_packet; + struct timeval now; + long delay; + int rc = -ENOTCONN; + + /* 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"? + */ + if (hctx == NULL || len == 0 || + hctx->ccid3hctx_state == TFRC_SSTATE_TERM) + 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 (new_packet == NULL) { + ccid3_pr_debug("%s, sk=%p, not enough mem to add " + "to history, send refused\n", + dccp_role(sk), sk); + goto out; + } + + dccp_tx_hist_add_entry(&hctx->ccid3hctx_hist, new_packet); + } + + do_gettimeofday(&now); + + switch (hctx->ccid3hctx_state) { + case TFRC_SSTATE_NO_SENT: + ccid3_pr_debug("%s, sk=%p, first packet(%llu)\n", + dccp_role(sk), sk, dp->dccps_gss); + + hctx->ccid3hctx_no_feedback_timer.function = ccid3_hc_tx_no_feedback_timer; + hctx->ccid3hctx_no_feedback_timer.data = (unsigned long)sk; + sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, + 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_TIMEOUT; + + /* Set nominal send time 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); + ccid3_pr_debug("send_packet delay=%ld\n", delay); + 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; + break; + } + + /* Can we send? if so add options and add to packet history */ + if (rc == 0) + new_packet->dccphtx_ccval = + DCCP_SKB_CB(skb)->dccpd_ccval = + hctx->ccid3hctx_last_win_count; +out: + return rc; +} + +static void ccid3_hc_tx_packet_sent(struct sock *sk, int more, int len) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + struct timeval now; + + BUG_ON(hctx == NULL); + + if (hctx->ccid3hctx_state == TFRC_SSTATE_TERM) { + ccid3_pr_debug("%s, sk=%p, while state is TFRC_SSTATE_TERM!\n", + dccp_role(sk), sk); + return; + } + + do_gettimeofday(&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 (packet == NULL) { + printk(KERN_CRIT "%s: packet doesn't exists in " + "history!\n", __FUNCTION__); + return; + } + if (packet->dccphtx_sent) { + printk(KERN_CRIT "%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 Valuer" in + * draft-ietf-dccp-ccid3-11.txt + */ + 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); + } + + 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); + + 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)); + return; + case TFRC_SSTATE_NO_FBACK: + case TFRC_SSTATE_FBACK: + if (len > 0) { + hctx->ccid3hctx_t_nom = now; + 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; + } +} + +static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + struct ccid3_options_received *opt_recv; + struct dccp_tx_hist_entry *packet; + unsigned long next_tmout; + u32 t_elapsed; + u32 pinv; + u32 x_recv; + u32 r_sample; + + if (hctx == NULL) + return; + + if (hctx->ccid3hctx_state == TFRC_SSTATE_TERM) { + ccid3_pr_debug("%s, sk=%p, received a packet when " + "terminating!\n", dccp_role(sk), sk); + return; + } + + /* we are only interested in ACKs */ + if (!(DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_ACK || + DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_DATAACK)) + return; + + opt_recv = &hctx->ccid3hctx_options_received; + + t_elapsed = dp->dccps_options_received.dccpor_elapsed_time; + 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 */ + packet = dccp_tx_hist_find_entry(&hctx->ccid3hctx_hist, + DCCP_SKB_CB(skb)->dccpd_ack_seq); + if (packet == NULL) { + ccid3_pr_debug("%s, sk=%p, seqno %llu(%s) does't " + "exist in history!\n", + dccp_role(sk), sk, + DCCP_SKB_CB(skb)->dccpd_ack_seq, + dccp_packet_name(DCCP_SKB_CB(skb)->dccpd_type)); + return; + } + + /* Update RTT */ + r_sample = timeval_now_delta(&packet->dccphtx_tstamp); + /* FIXME: */ + // r_sample -= usecs_to_jiffies(t_elapsed * 10); + + /* Update RTT estimate by + * If (No feedback recv) + * R = R_sample; + * Else + * R = q * R + (1 - q) * R_sample; + * + * 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); + + /* Update receive rate */ + hctx->ccid3hctx_x_recv = x_recv;/* X_recv in bytes per sec */ + + /* Update loss event rate */ + if (pinv == ~0 || pinv == 0) + hctx->ccid3hctx_p = 0; + else { + hctx->ccid3hctx_p = 1000000 / pinv; + + 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); + } + } + + /* 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. + */ + sk->sk_write_space(sk); + + /* + * Schedule no feedback timer to expire in + * max(4 * R, 2 * s / X) + */ + 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 " + "expire in %lu jiffies (%luus)\n", + dccp_role(sk), sk, + usecs_to_jiffies(next_tmout), next_tmout); + + sk_reset_timer(sk, &hctx->ccid3hctx_no_feedback_timer, + jiffies + max_t(u32, 1, usecs_to_jiffies(next_tmout))); + + /* set idle flag */ + 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(); + break; + } +} + +static void ccid3_hc_tx_insert_options(struct sock *sk, struct sk_buff *skb) +{ + const struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + + if (hctx == NULL || !(sk->sk_state == DCCP_OPEN || + sk->sk_state == DCCP_PARTOPEN)) + return; + + DCCP_SKB_CB(skb)->dccpd_ccval = hctx->ccid3hctx_last_win_count; +} + +static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option, + unsigned char len, u16 idx, + unsigned char *value) +{ + int rc = 0; + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + struct ccid3_options_received *opt_recv; + + if (hctx == NULL) + return 0; + + opt_recv = &hctx->ccid3hctx_options_received; + + if (opt_recv->ccid3or_seqno != dp->dccps_gsr) { + opt_recv->ccid3or_seqno = dp->dccps_gsr; + opt_recv->ccid3or_loss_event_rate = ~0; + opt_recv->ccid3or_loss_intervals_idx = 0; + opt_recv->ccid3or_loss_intervals_len = 0; + opt_recv->ccid3or_receive_rate = 0; + } + + switch (option) { + case TFRC_OPT_LOSS_EVENT_RATE: + if (len != 4) { + ccid3_pr_debug("%s, sk=%p, invalid len for " + "TFRC_OPT_LOSS_EVENT_RATE\n", + dccp_role(sk), sk); + rc = -EINVAL; + } else { + opt_recv->ccid3or_loss_event_rate = ntohl(*(u32 *)value); + ccid3_pr_debug("%s, sk=%p, LOSS_EVENT_RATE=%u\n", + dccp_role(sk), sk, + opt_recv->ccid3or_loss_event_rate); + } + break; + 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", + dccp_role(sk), sk, + opt_recv->ccid3or_loss_intervals_idx, + opt_recv->ccid3or_loss_intervals_len); + break; + case TFRC_OPT_RECEIVE_RATE: + if (len != 4) { + ccid3_pr_debug("%s, sk=%p, invalid len for " + "TFRC_OPT_RECEIVE_RATE\n", + dccp_role(sk), sk); + rc = -EINVAL; + } else { + opt_recv->ccid3or_receive_rate = ntohl(*(u32 *)value); + ccid3_pr_debug("%s, sk=%p, RECEIVE_RATE=%u\n", + dccp_role(sk), sk, + opt_recv->ccid3or_receive_rate); + } + break; + } + + return rc; +} + +static int ccid3_hc_tx_init(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx; + + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + + hctx = dp->dccps_hc_tx_ccid_private = kmalloc(sizeof(*hctx), + gfp_any()); + if (hctx == NULL) + return -ENOMEM; + + memset(hctx, 0, sizeof(*hctx)); + + 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_state = TFRC_SSTATE_NO_SENT; + INIT_LIST_HEAD(&hctx->ccid3hctx_hist); + init_timer(&hctx->ccid3hctx_no_feedback_timer); + + return 0; +} + +static void ccid3_hc_tx_exit(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + BUG_ON(hctx == NULL); + + ccid3_hc_tx_set_state(sk, TFRC_SSTATE_TERM); + sk_stop_timer(sk, &hctx->ccid3hctx_no_feedback_timer); + + /* Empty packet history */ + dccp_tx_hist_purge(ccid3_tx_hist, &hctx->ccid3hctx_hist); + + kfree(dp->dccps_hc_tx_ccid_private); + dp->dccps_hc_tx_ccid_private = NULL; +} + +/* + * 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 +static const char *ccid3_rx_state_name(enum ccid3_hc_rx_states state) +{ + static char *ccid3_rx_state_names[] = { + [TFRC_RSTATE_NO_DATA] = "NO_DATA", + [TFRC_RSTATE_DATA] = "DATA", + [TFRC_RSTATE_TERM] = "TERM", + }; + + return ccid3_rx_state_names[state]; +} +#endif + +static inline void ccid3_hc_rx_set_state(struct sock *sk, + enum ccid3_hc_rx_states state) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + enum ccid3_hc_rx_states oldstate = hcrx->ccid3hcrx_state; + + ccid3_pr_debug("%s(%p) %-8.8s -> %s\n", + dccp_role(sk), sk, ccid3_rx_state_name(oldstate), + ccid3_rx_state_name(state)); + WARN_ON(state == oldstate); + hcrx->ccid3hcrx_state = state; +} + +static void ccid3_hc_rx_send_feedback(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + struct dccp_rx_hist_entry *packet; + struct timeval now; + + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + + do_gettimeofday(&now); + + switch (hcrx->ccid3hcrx_state) { + 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 = (hcrx->ccid3hcrx_bytes_recv * + USEC_PER_SEC); + if (likely(delta > 1)) + hcrx->ccid3hcrx_x_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(); + return; + } + + packet = dccp_rx_hist_find_data_packet(&hcrx->ccid3hcrx_hist); + if (packet == NULL) { + printk(KERN_CRIT "%s: %s, sk=%p, no data packet in history!\n", + __FUNCTION__, dccp_role(sk), sk); + dump_stack(); + return; + } + + hcrx->ccid3hcrx_tstamp_last_feedback = now; + hcrx->ccid3hcrx_last_counter = packet->dccphrx_ccval; + hcrx->ccid3hcrx_seqno_last_counter = packet->dccphrx_seqno; + hcrx->ccid3hcrx_bytes_recv = 0; + + /* Convert to multiples of 10us */ + hcrx->ccid3hcrx_elapsed_time = + timeval_delta(&now, &packet->dccphrx_tstamp) / 10; + if (hcrx->ccid3hcrx_p == 0) + hcrx->ccid3hcrx_pinv = ~0; + else + hcrx->ccid3hcrx_pinv = 1000000 / hcrx->ccid3hcrx_p; + dccp_send_ack(sk); +} + +static void ccid3_hc_rx_insert_options(struct sock *sk, struct sk_buff *skb) +{ + const struct dccp_sock *dp = dccp_sk(sk); + u32 x_recv, pinv; + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + + if (hcrx == NULL || !(sk->sk_state == DCCP_OPEN || + sk->sk_state == DCCP_PARTOPEN)) + return; + + DCCP_SKB_CB(skb)->dccpd_ccval = hcrx->ccid3hcrx_last_counter; + + if (dccp_packet_without_ack(skb)) + return; + + if (hcrx->ccid3hcrx_elapsed_time != 0) + dccp_insert_option_elapsed_time(sk, skb, + hcrx->ccid3hcrx_elapsed_time); + dccp_insert_option_timestamp(sk, skb); + x_recv = htonl(hcrx->ccid3hcrx_x_recv); + pinv = htonl(hcrx->ccid3hcrx_pinv); + dccp_insert_option(sk, skb, TFRC_OPT_LOSS_EVENT_RATE, + &pinv, sizeof(pinv)); + dccp_insert_option(sk, skb, TFRC_OPT_RECEIVE_RATE, + &x_recv, sizeof(x_recv)); +} + +/* calculate first loss interval + * + * returns estimated loss interval in usecs */ + +static u32 ccid3_hc_rx_calc_first_li(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + struct dccp_rx_hist_entry *entry, *next, *tail = NULL; + u32 rtt, delta, x_recv, fval, p, tmp2; + struct timeval tstamp = { 0, }; + int interval = 0; + int win_count = 0; + int step = 0; + u64 tmp1; + + list_for_each_entry_safe(entry, next, &hcrx->ccid3hcrx_hist, + dccphrx_node) { + if (dccp_rx_hist_entry_data_packet(entry)) { + tail = entry; + + switch (step) { + case 0: + tstamp = entry->dccphrx_tstamp; + win_count = entry->dccphrx_ccval; + step = 1; + break; + case 1: + interval = win_count - entry->dccphrx_ccval; + if (interval < 0) + interval += TFRC_WIN_COUNT_LIMIT; + if (interval > 4) + goto found; + break; + } + } + } + + if (step == 0) { + printk(KERN_CRIT "%s: %s, sk=%p, packet history contains no " + "data packets!\n", + __FUNCTION__, dccp_role(sk), sk); + return ~0; + } + + if (interval == 0) { + ccid3_pr_debug("%s, sk=%p, Could not find a win_count " + "interval > 0. Defaulting to 1\n", + dccp_role(sk), sk); + interval = 1; + } +found: + 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; + + delta = timeval_now_delta(&hcrx->ccid3hcrx_tstamp_last_feedback); + x_recv = hcrx->ccid3hcrx_bytes_recv * USEC_PER_SEC; + if (likely(delta > 1)) + x_recv /= delta; + + tmp1 = (u64)x_recv * (u64)rtt; + do_div(tmp1,10000000); + tmp2 = (u32)tmp1; + fval = (hcrx->ccid3hcrx_s * 100000) / tmp2; + /* do not alter order above or you will get overflow on 32 bit */ + p = tfrc_calc_x_reverse_lookup(fval); + ccid3_pr_debug("%s, sk=%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; +} + +static void ccid3_hc_rx_update_li(struct sock *sk, u64 seq_loss, u8 win_loss) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + + if (seq_loss != DCCP_MAX_SEQNO + 1 && + list_empty(&hcrx->ccid3hcrx_li_hist)) { + struct dccp_li_hist_entry *li_tail; + + li_tail = dccp_li_hist_interval_new(ccid3_li_hist, + &hcrx->ccid3hcrx_li_hist, + seq_loss, win_loss); + if (li_tail == NULL) + return; + li_tail->dccplih_interval = ccid3_hc_rx_calc_first_li(sk); + } + /* FIXME: find end of interval */ +} + +static void ccid3_hc_rx_detect_loss(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + u8 win_loss; + const u64 seq_loss = dccp_rx_hist_detect_loss(&hcrx->ccid3hcrx_hist, + &hcrx->ccid3hcrx_li_hist, + &win_loss); + + ccid3_hc_rx_update_li(sk, seq_loss, win_loss); +} + +static void ccid3_hc_rx_packet_recv(struct sock *sk, struct sk_buff *skb) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + const struct dccp_options_received *opt_recv; + struct dccp_rx_hist_entry *packet; + struct timeval now; + u8 win_count; + u32 p_prev; + int ins; + + if (hcrx == NULL) + return; + + BUG_ON(!(hcrx->ccid3hcrx_state == TFRC_RSTATE_NO_DATA || + hcrx->ccid3hcrx_state == TFRC_RSTATE_DATA)); + + opt_recv = &dp->dccps_options_received; + + switch (DCCP_SKB_CB(skb)->dccpd_type) { + case DCCP_PKT_ACK: + if (hcrx->ccid3hcrx_state == TFRC_RSTATE_NO_DATA) + return; + case DCCP_PKT_DATAACK: + if (opt_recv->dccpor_timestamp_echo == 0) + break; + p_prev = hcrx->ccid3hcrx_rtt; + do_gettimeofday(&now); + hcrx->ccid3hcrx_rtt = timeval_usecs(&now) - + (opt_recv->dccpor_timestamp_echo - + opt_recv->dccpor_elapsed_time) * 10; + if (p_prev != hcrx->ccid3hcrx_rtt) + ccid3_pr_debug("%s, New RTT=%luus, elapsed time=%u\n", + dccp_role(sk), hcrx->ccid3hcrx_rtt, + opt_recv->dccpor_elapsed_time); + break; + case DCCP_PKT_DATA: + break; + default: + ccid3_pr_debug("%s, sk=%p, not DATA/DATAACK/ACK packet(%s)\n", + dccp_role(sk), sk, + dccp_packet_name(DCCP_SKB_CB(skb)->dccpd_type)); + return; + } + + packet = dccp_rx_hist_entry_new(ccid3_rx_hist, opt_recv->dccpor_ndp, + skb, SLAB_ATOMIC); + if (packet == NULL) { + ccid3_pr_debug("%s, sk=%p, Not enough mem to add rx packet " + "to history (consider it lost)!", + dccp_role(sk), sk); + return; + } + + win_count = packet->dccphrx_ccval; + + ins = dccp_rx_hist_add_packet(ccid3_rx_hist, &hcrx->ccid3hcrx_hist, + &hcrx->ccid3hcrx_li_hist, packet); + + if (DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_ACK) + return; + + 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, + 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; + if (ins != 0) + break; + + do_gettimeofday(&now); + if (timeval_delta(&now, &hcrx->ccid3hcrx_tstamp_last_ack) >= + hcrx->ccid3hcrx_rtt) { + 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(); + return; + } + + /* Dealing with packet loss */ + ccid3_pr_debug("%s, sk=%p(%s), data loss! Reacting...\n", + dccp_role(sk), sk, dccp_state_name(sk->sk_state)); + + ccid3_hc_rx_detect_loss(sk); + p_prev = hcrx->ccid3hcrx_p; + + /* Calculate loss event rate */ + if (!list_empty(&hcrx->ccid3hcrx_li_hist)) + /* Scaling up by 1000000 as fixed decimal */ + hcrx->ccid3hcrx_p = 1000000 / dccp_li_hist_calc_i_mean(&hcrx->ccid3hcrx_li_hist); + + if (hcrx->ccid3hcrx_p > p_prev) { + ccid3_hc_rx_send_feedback(sk); + return; + } +} + +static int ccid3_hc_rx_init(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx; + + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + + hcrx = dp->dccps_hc_rx_ccid_private = kmalloc(sizeof(*hcrx), + gfp_any()); + if (hcrx == NULL) + return -ENOMEM; + + memset(hcrx, 0, sizeof(*hcrx)); + + 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; + + hcrx->ccid3hcrx_state = TFRC_RSTATE_NO_DATA; + INIT_LIST_HEAD(&hcrx->ccid3hcrx_hist); + INIT_LIST_HEAD(&hcrx->ccid3hcrx_li_hist); + /* + * XXX this seems to be paranoid, need to think more about this, for + * now start with something different than zero. -acme + */ + hcrx->ccid3hcrx_rtt = USEC_PER_SEC / 5; + return 0; +} + +static void ccid3_hc_rx_exit(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + + ccid3_pr_debug("%s, sk=%p\n", dccp_role(sk), sk); + + if (hcrx == NULL) + return; + + ccid3_hc_rx_set_state(sk, TFRC_RSTATE_TERM); + + /* Empty packet history */ + dccp_rx_hist_purge(ccid3_rx_hist, &hcrx->ccid3hcrx_hist); + + /* Empty loss interval history */ + dccp_li_hist_purge(ccid3_li_hist, &hcrx->ccid3hcrx_li_hist); + + kfree(dp->dccps_hc_rx_ccid_private); + dp->dccps_hc_rx_ccid_private = NULL; +} + +static void ccid3_hc_rx_get_info(struct sock *sk, struct tcp_info *info) +{ + const struct dccp_sock *dp = dccp_sk(sk); + const struct ccid3_hc_rx_sock *hcrx = dp->dccps_hc_rx_ccid_private; + + if (hcrx == NULL) + return; + + 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) +{ + const struct dccp_sock *dp = dccp_sk(sk); + const struct ccid3_hc_tx_sock *hctx = dp->dccps_hc_tx_ccid_private; + + if (hctx == NULL) + return; + + info->tcpi_rto = hctx->ccid3hctx_t_rto; + info->tcpi_rtt = hctx->ccid3hctx_rtt; +} + +static struct ccid ccid3 = { + .ccid_id = 3, + .ccid_name = "ccid3", + .ccid_owner = THIS_MODULE, + .ccid_init = ccid3_init, + .ccid_exit = ccid3_exit, + .ccid_hc_tx_init = ccid3_hc_tx_init, + .ccid_hc_tx_exit = ccid3_hc_tx_exit, + .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_init = ccid3_hc_rx_init, + .ccid_hc_rx_exit = ccid3_hc_rx_exit, + .ccid_hc_rx_insert_options = ccid3_hc_rx_insert_options, + .ccid_hc_rx_packet_recv = ccid3_hc_rx_packet_recv, + .ccid_hc_rx_get_info = ccid3_hc_rx_get_info, + .ccid_hc_tx_get_info = ccid3_hc_tx_get_info, +}; + +module_param(ccid3_debug, int, 0444); +MODULE_PARM_DESC(ccid3_debug, "Enable debug messages"); + +static __init int ccid3_module_init(void) +{ + int rc = -ENOBUFS; + + ccid3_rx_hist = dccp_rx_hist_new("ccid3"); + if (ccid3_rx_hist == NULL) + goto out; + + ccid3_tx_hist = dccp_tx_hist_new("ccid3"); + if (ccid3_tx_hist == NULL) + goto out_free_rx; + + ccid3_li_hist = dccp_li_hist_new("ccid3"); + if (ccid3_li_hist == NULL) + goto out_free_tx; + + rc = ccid_register(&ccid3); + if (rc != 0) + goto out_free_loss_interval_history; +out: + return rc; + +out_free_loss_interval_history: + dccp_li_hist_delete(ccid3_li_hist); + ccid3_li_hist = NULL; +out_free_tx: + dccp_tx_hist_delete(ccid3_tx_hist); + ccid3_tx_hist = NULL; +out_free_rx: + dccp_rx_hist_delete(ccid3_rx_hist); + ccid3_rx_hist = NULL; + goto out; +} +module_init(ccid3_module_init); + +static __exit void ccid3_module_exit(void) +{ +#ifdef CONFIG_IP_DCCP_UNLOAD_HACK + /* + * Hack to use while developing, so that we get rid of the control + * sock, that is what keeps a refcount on dccp.ko -acme + */ + extern void dccp_ctl_sock_exit(void); + + dccp_ctl_sock_exit(); +#endif + ccid_unregister(&ccid3); + + if (ccid3_tx_hist != NULL) { + dccp_tx_hist_delete(ccid3_tx_hist); + ccid3_tx_hist = NULL; + } + if (ccid3_rx_hist != NULL) { + dccp_rx_hist_delete(ccid3_rx_hist); + ccid3_rx_hist = NULL; + } + if (ccid3_li_hist != NULL) { + dccp_li_hist_delete(ccid3_li_hist); + ccid3_li_hist = NULL; + } +} +module_exit(ccid3_module_exit); + +MODULE_AUTHOR("Ian McDonald <iam4@cs.waikato.ac.nz>, " + "Arnaldo Carvalho de Melo <acme@ghostprotocols.net>"); +MODULE_DESCRIPTION("DCCP TFRC CCID3 CCID"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS("net-dccp-ccid-3"); diff --git a/net/dccp/ccids/ccid3.h b/net/dccp/ccids/ccid3.h new file mode 100644 index 000000000000..ee8cbace6630 --- /dev/null +++ b/net/dccp/ccids/ccid3.h @@ -0,0 +1,137 @@ +/* + * net/dccp/ccids/ccid3.h + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * + * An implementation of the DCCP protocol + * + * This code has been developed by the University of Waikato WAND + * research group. For further information please see http://www.wand.net.nz/ + * or e-mail Ian McDonald - iam4@cs.waikato.ac.nz + * + * This code also uses code from Lulea University, rereleased as GPL by its + * authors: + * Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon + * + * Changes to meet Linux coding standards, to make it meet latest ccid3 draft + * and to make it work as a loadable module in the DCCP stack written by + * Arnaldo Carvalho de Melo <acme@conectiva.com.br>. + * + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ +#ifndef _DCCP_CCID3_H_ +#define _DCCP_CCID3_H_ + +#include <linux/config.h> +#include <linux/list.h> +#include <linux/time.h> +#include <linux/types.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 */ +#define TFRC_INITIAL_TIMEOUT (2 * USEC_PER_SEC) + +/* 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 + +#define TFRC_SMALLEST_P 40 + +enum ccid3_options { + TFRC_OPT_LOSS_EVENT_RATE = 192, + TFRC_OPT_LOSS_INTERVALS = 193, + TFRC_OPT_RECEIVE_RATE = 194, +}; + +struct ccid3_options_received { + u64 ccid3or_seqno:48, + ccid3or_loss_intervals_idx:16; + u16 ccid3or_loss_intervals_len; + u32 ccid3or_loss_event_rate; + u32 ccid3or_receive_rate; +}; + +/** struct ccid3_hc_tx_sock - CCID3 sender half connection sock + * + * @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 + */ +struct ccid3_hc_tx_sock { + u32 ccid3hctx_x; + u32 ccid3hctx_x_recv; + u32 ccid3hctx_x_calc; + u16 ccid3hctx_s; + u32 ccid3hctx_rtt; + u32 ccid3hctx_p; + u8 ccid3hctx_state; + u8 ccid3hctx_last_win_count; + u8 ccid3hctx_idle; + struct timeval ccid3hctx_t_last_win_count; + struct timer_list ccid3hctx_no_feedback_timer; + struct timeval ccid3hctx_t_ld; + struct timeval ccid3hctx_t_nom; + u32 ccid3hctx_t_rto; + u32 ccid3hctx_t_ipi; + u32 ccid3hctx_delta; + struct list_head ccid3hctx_hist; + struct ccid3_options_received ccid3hctx_options_received; +}; + +struct ccid3_hc_rx_sock { + u64 ccid3hcrx_seqno_last_counter:48, + ccid3hcrx_state:8, + ccid3hcrx_last_counter:4; + unsigned long ccid3hcrx_rtt; + u32 ccid3hcrx_p; + 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; + u32 ccid3hcrx_x_recv; +}; + +#define ccid3_hc_tx_field(s,field) (s->dccps_hc_tx_ccid_private == NULL ? 0 : \ + ((struct ccid3_hc_tx_sock *)s->dccps_hc_tx_ccid_private)->ccid3hctx_##field) + +#define ccid3_hc_rx_field(s,field) (s->dccps_hc_rx_ccid_private == NULL ? 0 : \ + ((struct ccid3_hc_rx_sock *)s->dccps_hc_rx_ccid_private)->ccid3hcrx_##field) + +#endif /* _DCCP_CCID3_H_ */ diff --git a/net/dccp/ccids/lib/Makefile b/net/dccp/ccids/lib/Makefile new file mode 100644 index 000000000000..5f940a6cbaca --- /dev/null +++ b/net/dccp/ccids/lib/Makefile @@ -0,0 +1,3 @@ +obj-$(CONFIG_IP_DCCP_TFRC_LIB) += dccp_tfrc_lib.o + +dccp_tfrc_lib-y := loss_interval.o packet_history.o tfrc_equation.o diff --git a/net/dccp/ccids/lib/loss_interval.c b/net/dccp/ccids/lib/loss_interval.c new file mode 100644 index 000000000000..4c01a54143ad --- /dev/null +++ b/net/dccp/ccids/lib/loss_interval.c @@ -0,0 +1,144 @@ +/* + * net/dccp/ccids/lib/loss_interval.c + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/module.h> + +#include "loss_interval.h" + +struct dccp_li_hist *dccp_li_hist_new(const char *name) +{ + struct dccp_li_hist *hist = kmalloc(sizeof(*hist), GFP_ATOMIC); + static const char dccp_li_hist_mask[] = "li_hist_%s"; + char *slab_name; + + if (hist == NULL) + goto out; + + slab_name = kmalloc(strlen(name) + sizeof(dccp_li_hist_mask) - 1, + GFP_ATOMIC); + if (slab_name == NULL) + goto out_free_hist; + + sprintf(slab_name, dccp_li_hist_mask, name); + hist->dccplih_slab = kmem_cache_create(slab_name, + sizeof(struct dccp_li_hist_entry), + 0, SLAB_HWCACHE_ALIGN, + NULL, NULL); + if (hist->dccplih_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_li_hist_new); + +void dccp_li_hist_delete(struct dccp_li_hist *hist) +{ + const char* name = kmem_cache_name(hist->dccplih_slab); + + kmem_cache_destroy(hist->dccplih_slab); + kfree(name); + kfree(hist); +} + +EXPORT_SYMBOL_GPL(dccp_li_hist_delete); + +void dccp_li_hist_purge(struct dccp_li_hist *hist, struct list_head *list) +{ + struct dccp_li_hist_entry *entry, *next; + + list_for_each_entry_safe(entry, next, list, dccplih_node) { + list_del_init(&entry->dccplih_node); + kmem_cache_free(hist->dccplih_slab, entry); + } +} + +EXPORT_SYMBOL_GPL(dccp_li_hist_purge); + +/* Weights used to calculate loss event rate */ +/* + * These are integers as per section 8 of RFC3448. We can then divide by 4 * + * when we use it. + */ +static const int dccp_li_hist_w[DCCP_LI_HIST_IVAL_F_LENGTH] = { + 4, 4, 4, 4, 3, 2, 1, 1, +}; + +u32 dccp_li_hist_calc_i_mean(struct list_head *list) +{ + struct dccp_li_hist_entry *li_entry, *li_next; + int i = 0; + u32 i_tot; + u32 i_tot0 = 0; + u32 i_tot1 = 0; + u32 w_tot = 0; + + list_for_each_entry_safe(li_entry, li_next, list, dccplih_node) { + if (i < DCCP_LI_HIST_IVAL_F_LENGTH) { + i_tot0 += li_entry->dccplih_interval * dccp_li_hist_w[i]; + w_tot += dccp_li_hist_w[i]; + } + + if (i != 0) + i_tot1 += li_entry->dccplih_interval * dccp_li_hist_w[i - 1]; + + if (++i > DCCP_LI_HIST_IVAL_F_LENGTH) + break; + } + + if (i != DCCP_LI_HIST_IVAL_F_LENGTH) + return 0; + + i_tot = max(i_tot0, i_tot1); + + /* FIXME: Why do we do this? -Ian McDonald */ + if (i_tot * 4 < w_tot) + i_tot = w_tot * 4; + + return i_tot * 4 / w_tot; +} + +EXPORT_SYMBOL_GPL(dccp_li_hist_calc_i_mean); + +struct dccp_li_hist_entry *dccp_li_hist_interval_new(struct dccp_li_hist *hist, + struct list_head *list, + const u64 seq_loss, + const u8 win_loss) +{ + struct dccp_li_hist_entry *tail = NULL, *entry; + int i; + + for (i = 0; i <= DCCP_LI_HIST_IVAL_F_LENGTH; ++i) { + entry = dccp_li_hist_entry_new(hist, SLAB_ATOMIC); + if (entry == NULL) { + dccp_li_hist_purge(hist, list); + return NULL; + } + if (tail == NULL) + tail = entry; + list_add(&entry->dccplih_node, list); + } + + entry->dccplih_seqno = seq_loss; + entry->dccplih_win_count = win_loss; + return tail; +} + +EXPORT_SYMBOL_GPL(dccp_li_hist_interval_new); diff --git a/net/dccp/ccids/lib/loss_interval.h b/net/dccp/ccids/lib/loss_interval.h new file mode 100644 index 000000000000..13ad47ba1420 --- /dev/null +++ b/net/dccp/ccids/lib/loss_interval.h @@ -0,0 +1,61 @@ +#ifndef _DCCP_LI_HIST_ +#define _DCCP_LI_HIST_ +/* + * net/dccp/ccids/lib/loss_interval.h + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + */ + +#include <linux/config.h> +#include <linux/list.h> +#include <linux/slab.h> +#include <linux/time.h> + +#define DCCP_LI_HIST_IVAL_F_LENGTH 8 + +struct dccp_li_hist { + kmem_cache_t *dccplih_slab; +}; + +extern struct dccp_li_hist *dccp_li_hist_new(const char *name); +extern void dccp_li_hist_delete(struct dccp_li_hist *hist); + +struct dccp_li_hist_entry { + struct list_head dccplih_node; + u64 dccplih_seqno:48, + dccplih_win_count:4; + u32 dccplih_interval; +}; + +static inline struct dccp_li_hist_entry * + dccp_li_hist_entry_new(struct dccp_li_hist *hist, + const unsigned int __nocast prio) +{ + return kmem_cache_alloc(hist->dccplih_slab, prio); +} + +static inline void dccp_li_hist_entry_delete(struct dccp_li_hist *hist, + struct dccp_li_hist_entry *entry) +{ + if (entry != NULL) + kmem_cache_free(hist->dccplih_slab, entry); +} + +extern void dccp_li_hist_purge(struct dccp_li_hist *hist, + struct list_head *list); + +extern u32 dccp_li_hist_calc_i_mean(struct list_head *list); + +extern struct dccp_li_hist_entry * + dccp_li_hist_interval_new(struct dccp_li_hist *hist, + struct list_head *list, + const u64 seq_loss, + const u8 win_loss); +#endif /* _DCCP_LI_HIST_ */ diff --git a/net/dccp/ccids/lib/packet_history.c b/net/dccp/ccids/lib/packet_history.c new file mode 100644 index 000000000000..d3f9d2053830 --- /dev/null +++ b/net/dccp/ccids/lib/packet_history.c @@ -0,0 +1,398 @@ +/* + * net/dccp/packet_history.h + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * + * An implementation of the DCCP protocol + * + * This code has been developed by the University of Waikato WAND + * research group. For further information please see http://www.wand.net.nz/ + * or e-mail Ian McDonald - iam4@cs.waikato.ac.nz + * + * This code also uses code from Lulea University, rereleased as GPL by its + * authors: + * Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon + * + * Changes to meet Linux coding standards, to make it meet latest ccid3 draft + * and to make it work as a loadable module in the DCCP stack written by + * Arnaldo Carvalho de Melo <acme@conectiva.com.br>. + * + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/string.h> + +#include "packet_history.h" + +struct dccp_rx_hist *dccp_rx_hist_new(const char *name) +{ + struct dccp_rx_hist *hist = kmalloc(sizeof(*hist), GFP_ATOMIC); + static const char dccp_rx_hist_mask[] = "rx_hist_%s"; + char *slab_name; + + if (hist == NULL) + goto out; + + slab_name = kmalloc(strlen(name) + sizeof(dccp_rx_hist_mask) - 1, + GFP_ATOMIC); + if (slab_name == NULL) + goto out_free_hist; + + sprintf(slab_name, dccp_rx_hist_mask, name); + hist->dccprxh_slab = kmem_cache_create(slab_name, + sizeof(struct dccp_rx_hist_entry), + 0, SLAB_HWCACHE_ALIGN, + NULL, NULL); + if (hist->dccprxh_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_rx_hist_new); + +void dccp_rx_hist_delete(struct dccp_rx_hist *hist) +{ + const char* name = kmem_cache_name(hist->dccprxh_slab); + + kmem_cache_destroy(hist->dccprxh_slab); + kfree(name); + kfree(hist); +} + +EXPORT_SYMBOL_GPL(dccp_rx_hist_delete); + +void dccp_rx_hist_purge(struct dccp_rx_hist *hist, struct list_head *list) +{ + struct dccp_rx_hist_entry *entry, *next; + + 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_rx_hist_purge); + +struct dccp_rx_hist_entry * + dccp_rx_hist_find_data_packet(const struct list_head *list) +{ + struct dccp_rx_hist_entry *entry, *packet = NULL; + + list_for_each_entry(entry, list, dccphrx_node) + if (entry->dccphrx_type == DCCP_PKT_DATA || + entry->dccphrx_type == DCCP_PKT_DATAACK) { + packet = entry; + break; + } + + return packet; +} + +EXPORT_SYMBOL_GPL(dccp_rx_hist_find_data_packet); + +int 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) +{ + struct dccp_rx_hist_entry *entry, *next, *iter; + u8 num_later = 0; + + iter = dccp_rx_hist_head(rx_list); + if (iter == NULL) + dccp_rx_hist_add_entry(rx_list, packet); + else { + const u64 seqno = packet->dccphrx_seqno; + + if (after48(seqno, iter->dccphrx_seqno)) + dccp_rx_hist_add_entry(rx_list, packet); + else { + if (dccp_rx_hist_entry_data_packet(iter)) + num_later = 1; + + list_for_each_entry_continue(iter, rx_list, + dccphrx_node) { + if (after48(seqno, iter->dccphrx_seqno)) { + dccp_rx_hist_add_entry(&iter->dccphrx_node, + packet); + goto trim_history; + } + + if (dccp_rx_hist_entry_data_packet(iter)) + num_later++; + + if (num_later == TFRC_RECV_NUM_LATE_LOSS) { + dccp_rx_hist_entry_delete(hist, packet); + return 1; + } + } + + if (num_later < TFRC_RECV_NUM_LATE_LOSS) + dccp_rx_hist_add_entry(rx_list, packet); + /* + * FIXME: else what? should we destroy the packet + * like above? + */ + } + } + +trim_history: + /* + * Trim history (remove all packets after the NUM_LATE_LOSS + 1 + * data packets) + */ + num_later = TFRC_RECV_NUM_LATE_LOSS + 1; + + if (!list_empty(li_list)) { + list_for_each_entry_safe(entry, next, rx_list, dccphrx_node) { + if (num_later == 0) { + list_del_init(&entry->dccphrx_node); + dccp_rx_hist_entry_delete(hist, entry); + } else if (dccp_rx_hist_entry_data_packet(entry)) + --num_later; + } + } else { + int step = 0; + u8 win_count = 0; /* Not needed, but lets shut up gcc */ + int tmp; + /* + * We have no loss interval history so we need at least one + * rtt:s of data packets to approximate rtt. + */ + list_for_each_entry_safe(entry, next, rx_list, dccphrx_node) { + if (num_later == 0) { + switch (step) { + case 0: + step = 1; + /* OK, find next data packet */ + num_later = 1; + break; + case 1: + step = 2; + /* OK, find next data packet */ + num_later = 1; + win_count = entry->dccphrx_ccval; + break; + case 2: + tmp = win_count - entry->dccphrx_ccval; + if (tmp < 0) + tmp += TFRC_WIN_COUNT_LIMIT; + if (tmp > TFRC_WIN_COUNT_PER_RTT + 1) { + /* + * We have found a packet older + * than one rtt remove the rest + */ + step = 3; + } else /* OK, find next data packet */ + num_later = 1; + break; + case 3: + list_del_init(&entry->dccphrx_node); + dccp_rx_hist_entry_delete(hist, entry); + break; + } + } else if (dccp_rx_hist_entry_data_packet(entry)) + --num_later; + } + } + + return 0; +} + +EXPORT_SYMBOL_GPL(dccp_rx_hist_add_packet); + +u64 dccp_rx_hist_detect_loss(struct list_head *rx_list, + struct list_head *li_list, u8 *win_loss) +{ + struct dccp_rx_hist_entry *entry, *next, *packet; + struct dccp_rx_hist_entry *a_loss = NULL; + struct dccp_rx_hist_entry *b_loss = NULL; + u64 seq_loss = DCCP_MAX_SEQNO + 1; + u8 num_later = TFRC_RECV_NUM_LATE_LOSS; + + list_for_each_entry_safe(entry, next, rx_list, dccphrx_node) { + if (num_later == 0) { + b_loss = entry; + break; + } else if (dccp_rx_hist_entry_data_packet(entry)) + --num_later; + } + + if (b_loss == NULL) + goto out; + + num_later = 1; + list_for_each_entry_safe_continue(entry, next, rx_list, dccphrx_node) { + if (num_later == 0) { + a_loss = entry; + break; + } else if (dccp_rx_hist_entry_data_packet(entry)) + --num_later; + } + + if (a_loss == NULL) { + if (list_empty(li_list)) { + /* no loss event have occured yet */ + LIMIT_NETDEBUG("%s: TODO: find a lost data packet by " + "comparing to initial seqno\n", + __FUNCTION__); + goto out; + } else { + LIMIT_NETDEBUG("%s: Less than 4 data pkts in history!", + __FUNCTION__); + goto out; + } + } + + /* Locate a lost data packet */ + entry = packet = b_loss; + list_for_each_entry_safe_continue(entry, next, rx_list, dccphrx_node) { + u64 delta = dccp_delta_seqno(entry->dccphrx_seqno, + packet->dccphrx_seqno); + + if (delta != 0) { + if (dccp_rx_hist_entry_data_packet(packet)) + --delta; + /* + * FIXME: check this, probably this % usage is because + * in earlier drafts the ndp count was just 8 bits + * long, but now it cam be up to 24 bits long. + */ +#if 0 + if (delta % DCCP_NDP_LIMIT != + (packet->dccphrx_ndp - + entry->dccphrx_ndp) % DCCP_NDP_LIMIT) +#endif + if (delta != packet->dccphrx_ndp - entry->dccphrx_ndp) { + seq_loss = entry->dccphrx_seqno; + dccp_inc_seqno(&seq_loss); + } + } + packet = entry; + if (packet == a_loss) + break; + } +out: + if (seq_loss != DCCP_MAX_SEQNO + 1) + *win_loss = a_loss->dccphrx_ccval; + else + *win_loss = 0; /* Paranoia */ + + return seq_loss; +} + +EXPORT_SYMBOL_GPL(dccp_rx_hist_detect_loss); + +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_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); + +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); + +MODULE_AUTHOR("Ian McDonald <iam4@cs.waikato.ac.nz>, " + "Arnaldo Carvalho de Melo <acme@ghostprotocols.net>"); +MODULE_DESCRIPTION("DCCP TFRC library"); +MODULE_LICENSE("GPL"); diff --git a/net/dccp/ccids/lib/packet_history.h b/net/dccp/ccids/lib/packet_history.h new file mode 100644 index 000000000000..fb90a91aa93d --- /dev/null +++ b/net/dccp/ccids/lib/packet_history.h @@ -0,0 +1,199 @@ +/* + * net/dccp/packet_history.h + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * + * An implementation of the DCCP protocol + * + * This code has been developed by the University of Waikato WAND + * research group. For further information please see http://www.wand.net.nz/ + * or e-mail Ian McDonald - iam4@cs.waikato.ac.nz + * + * This code also uses code from Lulea University, rereleased as GPL by its + * authors: + * Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon + * + * Changes to meet Linux coding standards, to make it meet latest ccid3 draft + * and to make it work as a loadable module in the DCCP stack written by + * Arnaldo Carvalho de Melo <acme@conectiva.com.br>. + * + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#ifndef _DCCP_PKT_HIST_ +#define _DCCP_PKT_HIST_ + +#include <linux/config.h> +#include <linux/list.h> +#include <linux/slab.h> +#include <linux/time.h> + +#include "../../dccp.h" + +/* Number of later packets received before one is considered lost */ +#define TFRC_RECV_NUM_LATE_LOSS 3 + +#define TFRC_WIN_COUNT_PER_RTT 4 +#define TFRC_WIN_COUNT_LIMIT 16 + +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; +}; + +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); + +static inline struct dccp_tx_hist_entry * + dccp_tx_hist_entry_new(struct dccp_tx_hist *hist, + const unsigned int __nocast prio) +{ + struct dccp_tx_hist_entry *entry = kmem_cache_alloc(hist->dccptxh_slab, + prio); + + if (entry != NULL) + entry->dccphtx_sent = 0; + + return entry; +} + +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 struct dccp_tx_hist_entry * + dccp_tx_hist_find_entry(const struct list_head *list, + const u64 seq); + +static inline void dccp_tx_hist_add_entry(struct list_head *list, + struct dccp_tx_hist_entry *entry) +{ + list_add(&entry->dccphtx_node, 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); + +static inline struct dccp_tx_hist_entry * + dccp_tx_hist_head(struct list_head *list) +{ + 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; +} + +static inline struct dccp_rx_hist_entry * + dccp_rx_hist_entry_new(struct dccp_rx_hist *hist, + const u32 ndp, + const struct sk_buff *skb, + const unsigned int __nocast prio) +{ + struct dccp_rx_hist_entry *entry = kmem_cache_alloc(hist->dccprxh_slab, + prio); + + if (entry != NULL) { + const struct dccp_hdr *dh = dccp_hdr(skb); + + entry->dccphrx_seqno = DCCP_SKB_CB(skb)->dccpd_seq; + entry->dccphrx_ccval = dh->dccph_ccval; + entry->dccphrx_type = dh->dccph_type; + entry->dccphrx_ndp = ndp; + do_gettimeofday(&(entry->dccphrx_tstamp)); + } + + 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 void dccp_rx_hist_add_entry(struct list_head *list, + struct dccp_rx_hist_entry *entry) +{ + list_add(&entry->dccphrx_node, list); +} + +static inline struct dccp_rx_hist_entry * + dccp_rx_hist_head(struct list_head *list) +{ + struct dccp_rx_hist_entry *head = NULL; + + if (!list_empty(list)) + head = list_entry(list->next, struct dccp_rx_hist_entry, + dccphrx_node); + return head; +} + +static inline int + dccp_rx_hist_entry_data_packet(const struct dccp_rx_hist_entry *entry) +{ + return entry->dccphrx_type == DCCP_PKT_DATA || + entry->dccphrx_type == DCCP_PKT_DATAACK; +} + +extern int 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); + +extern u64 dccp_rx_hist_detect_loss(struct list_head *rx_list, + struct list_head *li_list, u8 *win_loss); + +#endif /* _DCCP_PKT_HIST_ */ diff --git a/net/dccp/ccids/lib/tfrc.h b/net/dccp/ccids/lib/tfrc.h new file mode 100644 index 000000000000..130c4c40cfe3 --- /dev/null +++ b/net/dccp/ccids/lib/tfrc.h @@ -0,0 +1,22 @@ +#ifndef _TFRC_H_ +#define _TFRC_H_ +/* + * net/dccp/ccids/lib/tfrc.h + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include <linux/types.h> + +extern u32 tfrc_calc_x(u16 s, u32 R, u32 p); +extern u32 tfrc_calc_x_reverse_lookup(u32 fvalue); + +#endif /* _TFRC_H_ */ diff --git a/net/dccp/ccids/lib/tfrc_equation.c b/net/dccp/ccids/lib/tfrc_equation.c new file mode 100644 index 000000000000..d2b5933b4510 --- /dev/null +++ b/net/dccp/ccids/lib/tfrc_equation.c @@ -0,0 +1,644 @@ +/* + * net/dccp/ccids/lib/tfrc_equation.c + * + * Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand. + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/module.h> + +#include <asm/bug.h> +#include <asm/div64.h> + +#include "tfrc.h" + +#define TFRC_CALC_X_ARRSIZE 500 + +#define TFRC_CALC_X_SPLIT 50000 +/* equivalent to 0.05 */ + +static const u32 tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE][2] = { + { 37172, 8172 }, + { 53499, 11567 }, + { 66664, 14180 }, + { 78298, 16388 }, + { 89021, 18339 }, + { 99147, 20108 }, + { 108858, 21738 }, + { 118273, 23260 }, + { 127474, 24693 }, + { 136520, 26052 }, + { 145456, 27348 }, + { 154316, 28589 }, + { 163130, 29783 }, + { 171919, 30935 }, + { 180704, 32049 }, + { 189502, 33130 }, + { 198328, 34180 }, + { 207194, 35202 }, + { 216114, 36198 }, + { 225097, 37172 }, + { 234153, 38123 }, + { 243294, 39055 }, + { 252527, 39968 }, + { 261861, 40864 }, + { 271305, 41743 }, + { 280866, 42607 }, + { 290553, 43457 }, + { 300372, 44293 }, + { 310333, 45117 }, + { 320441, 45929 }, + { 330705, 46729 }, + { 341131, 47518 }, + { 351728, 48297 }, + { 362501, 49066 }, + { 373460, 49826 }, + { 384609, 50577 }, + { 395958, 51320 }, + { 407513, 52054 }, + { 419281, 52780 }, + { 431270, 53499 }, + { 443487, 54211 }, + { 455940, 54916 }, + { 468635, 55614 }, + { 481581, 56306 }, + { 494785, 56991 }, + { 508254, 57671 }, + { 521996, 58345 }, + { 536019, 59014 }, + { 550331, 59677 }, + { 564939, 60335 }, + { 579851, 60988 }, + { 595075, 61636 }, + { 610619, 62279 }, + { 626491, 62918 }, + { 642700, 63553 }, + { 659253, 64183 }, + { 676158, 64809 }, + { 693424, 65431 }, + { 711060, 66050 }, + { 729073, 66664 }, + { 747472, 67275 }, + { 766266, 67882 }, + { 785464, 68486 }, + { 805073, 69087 }, + { 825103, 69684 }, + { 845562, 70278 }, + { 866460, 70868 }, + { 887805, 71456 }, + { 909606, 72041 }, + { 931873, 72623 }, + { 954614, 73202 }, + { 977839, 73778 }, + { 1001557, 74352 }, + { 1025777, 74923 }, + { 1050508, 75492 }, + { 1075761, 76058 }, + { 1101544, 76621 }, + { 1127867, 77183 }, + { 1154739, 77741 }, + { 1182172, 78298 }, + { 1210173, 78852 }, + { 1238753, 79405 }, + { 1267922, 79955 }, + { 1297689, 80503 }, + { 1328066, 81049 }, + { 1359060, 81593 }, + { 1390684, 82135 }, + { 1422947, 82675 }, + { 1455859, 83213 }, + { 1489430, 83750 }, + { 1523671, 84284 }, + { 1558593, 84817 }, + { 1594205, 85348 }, + { 1630518, 85878 }, + { 1667543, 86406 }, + { 1705290, 86932 }, + { 1743770, 87457 }, + { 1782994, 87980 }, + { 1822973, 88501 }, + { 1863717, 89021 }, + { 1905237, 89540 }, + { 1947545, 90057 }, + { 1990650, 90573 }, + { 2034566, 91087 }, + { 2079301, 91600 }, + { 2124869, 92111 }, + { 2171279, 92622 }, + { 2218543, 93131 }, + { 2266673, 93639 }, + { 2315680, 94145 }, + { 2365575, 94650 }, + { 2416371, 95154 }, + { 2468077, 95657 }, + { 2520707, 96159 }, + { 2574271, 96660 }, + { 2628782, 97159 }, + { 2684250, 97658 }, + { 2740689, 98155 }, + { 2798110, 98651 }, + { 2856524, 99147 }, + { 2915944, 99641 }, + { 2976382, 100134 }, + { 3037850, 100626 }, + { 3100360, 101117 }, + { 3163924, 101608 }, + { 3228554, 102097 }, + { 3294263, 102586 }, + { 3361063, 103073 }, + { 3428966, 103560 }, + { 3497984, 104045 }, + { 3568131, 104530 }, + { 3639419, 105014 }, + { 3711860, 105498 }, + { 3785467, 105980 }, + { 3860253, 106462 }, + { 3936229, 106942 }, + { 4013410, 107422 }, + { 4091808, 107902 }, + { 4171435, 108380 }, + { 4252306, 108858 }, + { 4334431, 109335 }, + { 4417825, 109811 }, + { 4502501, 110287 }, + { 4588472, 110762 }, + { 4675750, 111236 }, + { 4764349, 111709 }, + { 4854283, 112182 }, + { 4945564, 112654 }, + { 5038206, 113126 }, + { 5132223, 113597 }, + { 5227627, 114067 }, + { 5324432, 114537 }, + { 5422652, 115006 }, + { 5522299, 115474 }, + { 5623389, 115942 }, + { 5725934, 116409 }, + { 5829948, 116876 }, + { 5935446, 117342 }, + { 6042439, 117808 }, + { 6150943, 118273 }, + { 6260972, 118738 }, + { 6372538, 119202 }, + { 6485657, 119665 }, + { 6600342, 120128 }, + { 6716607, 120591 }, + { 6834467, 121053 }, + { 6953935, 121514 }, + { 7075025, 121976 }, + { 7197752, 122436 }, + { 7322131, 122896 }, + { 7448175, 123356 }, + { 7575898, 123815 }, + { 7705316, 124274 }, + { 7836442, 124733 }, + { 7969291, 125191 }, + { 8103877, 125648 }, + { 8240216, 126105 }, + { 8378321, 126562 }, + { 8518208, 127018 }, + { 8659890, 127474 }, + { 8803384, 127930 }, + { 8948702, 128385 }, + { 9095861, 128840 }, + { 9244875, 129294 }, + { 9395760, 129748 }, + { 9548529, 130202 }, + { 9703198, 130655 }, + { 9859782, 131108 }, + { 10018296, 131561 }, + { 10178755, 132014 }, + { 10341174, 132466 }, + { 10505569, 132917 }, + { 10671954, 133369 }, + { 10840345, 133820 }, + { 11010757, 134271 }, + { 11183206, 134721 }, + { 11357706, 135171 }, + { 11534274, 135621 }, + { 11712924, 136071 }, + { 11893673, 136520 }, + { 12076536, 136969 }, + { 12261527, 137418 }, + { 12448664, 137867 }, + { 12637961, 138315 }, + { 12829435, 138763 }, + { 13023101, 139211 }, + { 13218974, 139658 }, + { 13417071, 140106 }, + { 13617407, 140553 }, + { 13819999, 140999 }, + { 14024862, 141446 }, + { 14232012, 141892 }, + { 14441465, 142339 }, + { 14653238, 142785 }, + { 14867346, 143230 }, + { 15083805, 143676 }, + { 15302632, 144121 }, + { 15523842, 144566 }, + { 15747453, 145011 }, + { 15973479, 145456 }, + { 16201939, 145900 }, + { 16432847, 146345 }, + { 16666221, 146789 }, + { 16902076, 147233 }, + { 17140429, 147677 }, + { 17381297, 148121 }, + { 17624696, 148564 }, + { 17870643, 149007 }, + { 18119154, 149451 }, + { 18370247, 149894 }, + { 18623936, 150336 }, + { 18880241, 150779 }, + { 19139176, 151222 }, + { 19400759, 151664 }, + { 19665007, 152107 }, + { 19931936, 152549 }, + { 20201564, 152991 }, + { 20473907, 153433 }, + { 20748982, 153875 }, + { 21026807, 154316 }, + { 21307399, 154758 }, + { 21590773, 155199 }, + { 21876949, 155641 }, + { 22165941, 156082 }, + { 22457769, 156523 }, + { 22752449, 156964 }, + { 23049999, 157405 }, + { 23350435, 157846 }, + { 23653774, 158287 }, + { 23960036, 158727 }, + { 24269236, 159168 }, + { 24581392, 159608 }, + { 24896521, 160049 }, + { 25214642, 160489 }, + { 25535772, 160929 }, + { 25859927, 161370 }, + { 26187127, 161810 }, + { 26517388, 162250 }, + { 26850728, 162690 }, + { 27187165, 163130 }, + { 27526716, 163569 }, + { 27869400, 164009 }, + { 28215234, 164449 }, + { 28564236, 164889 }, + { 28916423, 165328 }, + { 29271815, 165768 }, + { 29630428, 166208 }, + { 29992281, 166647 }, + { 30357392, 167087 }, + { 30725779, 167526 }, + { 31097459, 167965 }, + { 31472452, 168405 }, + { 31850774, 168844 }, + { 32232445, 169283 }, + { 32617482, 169723 }, + { 33005904, 170162 }, + { 33397730, 170601 }, + { 33792976, 171041 }, + { 34191663, 171480 }, + { 34593807, 171919 }, + { 34999428, 172358 }, + { 35408544, 172797 }, + { 35821174, 173237 }, + { 36237335, 173676 }, + { 36657047, 174115 }, + { 37080329, 174554 }, + { 37507197, 174993 }, + { 37937673, 175433 }, + { 38371773, 175872 }, + { 38809517, 176311 }, + { 39250924, 176750 }, + { 39696012, 177190 }, + { 40144800, 177629 }, + { 40597308, 178068 }, + { 41053553, 178507 }, + { 41513554, 178947 }, + { 41977332, 179386 }, + { 42444904, 179825 }, + { 42916290, 180265 }, + { 43391509, 180704 }, + { 43870579, 181144 }, + { 44353520, 181583 }, + { 44840352, 182023 }, + { 45331092, 182462 }, + { 45825761, 182902 }, + { 46324378, 183342 }, + { 46826961, 183781 }, + { 47333531, 184221 }, + { 47844106, 184661 }, + { 48358706, 185101 }, + { 48877350, 185541 }, + { 49400058, 185981 }, + { 49926849, 186421 }, + { 50457743, 186861 }, + { 50992759, 187301 }, + { 51531916, 187741 }, + { 52075235, 188181 }, + { 52622735, 188622 }, + { 53174435, 189062 }, + { 53730355, 189502 }, + { 54290515, 189943 }, + { 54854935, 190383 }, + { 55423634, 190824 }, + { 55996633, 191265 }, + { 56573950, 191706 }, + { 57155606, 192146 }, + { 57741621, 192587 }, + { 58332014, 193028 }, + { 58926806, 193470 }, + { 59526017, 193911 }, + { 60129666, 194352 }, + { 60737774, 194793 }, + { 61350361, 195235 }, + { 61967446, 195677 }, + { 62589050, 196118 }, + { 63215194, 196560 }, + { 63845897, 197002 }, + { 64481179, 197444 }, + { 65121061, 197886 }, + { 65765563, 198328 }, + { 66414705, 198770 }, + { 67068508, 199213 }, + { 67726992, 199655 }, + { 68390177, 200098 }, + { 69058085, 200540 }, + { 69730735, 200983 }, + { 70408147, 201426 }, + { 71090343, 201869 }, + { 71777343, 202312 }, + { 72469168, 202755 }, + { 73165837, 203199 }, + { 73867373, 203642 }, + { 74573795, 204086 }, + { 75285124, 204529 }, + { 76001380, 204973 }, + { 76722586, 205417 }, + { 77448761, 205861 }, + { 78179926, 206306 }, + { 78916102, 206750 }, + { 79657310, 207194 }, + { 80403571, 207639 }, + { 81154906, 208084 }, + { 81911335, 208529 }, + { 82672880, 208974 }, + { 83439562, 209419 }, + { 84211402, 209864 }, + { 84988421, 210309 }, + { 85770640, 210755 }, + { 86558080, 211201 }, + { 87350762, 211647 }, + { 88148708, 212093 }, + { 88951938, 212539 }, + { 89760475, 212985 }, + { 90574339, 213432 }, + { 91393551, 213878 }, + { 92218133, 214325 }, + { 93048107, 214772 }, + { 93883493, 215219 }, + { 94724314, 215666 }, + { 95570590, 216114 }, + { 96422343, 216561 }, + { 97279594, 217009 }, + { 98142366, 217457 }, + { 99010679, 217905 }, + { 99884556, 218353 }, + { 100764018, 218801 }, + { 101649086, 219250 }, + { 102539782, 219698 }, + { 103436128, 220147 }, + { 104338146, 220596 }, + { 105245857, 221046 }, + { 106159284, 221495 }, + { 107078448, 221945 }, + { 108003370, 222394 }, + { 108934074, 222844 }, + { 109870580, 223294 }, + { 110812910, 223745 }, + { 111761087, 224195 }, + { 112715133, 224646 }, + { 113675069, 225097 }, + { 114640918, 225548 }, + { 115612702, 225999 }, + { 116590442, 226450 }, + { 117574162, 226902 }, + { 118563882, 227353 }, + { 119559626, 227805 }, + { 120561415, 228258 }, + { 121569272, 228710 }, + { 122583219, 229162 }, + { 123603278, 229615 }, + { 124629471, 230068 }, + { 125661822, 230521 }, + { 126700352, 230974 }, + { 127745083, 231428 }, + { 128796039, 231882 }, + { 129853241, 232336 }, + { 130916713, 232790 }, + { 131986475, 233244 }, + { 133062553, 233699 }, + { 134144966, 234153 }, + { 135233739, 234608 }, + { 136328894, 235064 }, + { 137430453, 235519 }, + { 138538440, 235975 }, + { 139652876, 236430 }, + { 140773786, 236886 }, + { 141901190, 237343 }, + { 143035113, 237799 }, + { 144175576, 238256 }, + { 145322604, 238713 }, + { 146476218, 239170 }, + { 147636442, 239627 }, + { 148803298, 240085 }, + { 149976809, 240542 }, + { 151156999, 241000 }, + { 152343890, 241459 }, + { 153537506, 241917 }, + { 154737869, 242376 }, + { 155945002, 242835 }, + { 157158929, 243294 }, + { 158379673, 243753 }, + { 159607257, 244213 }, + { 160841704, 244673 }, + { 162083037, 245133 }, + { 163331279, 245593 }, + { 164586455, 246054 }, + { 165848586, 246514 }, + { 167117696, 246975 }, + { 168393810, 247437 }, + { 169676949, 247898 }, + { 170967138, 248360 }, + { 172264399, 248822 }, + { 173568757, 249284 }, + { 174880235, 249747 }, + { 176198856, 250209 }, + { 177524643, 250672 }, + { 178857621, 251136 }, + { 180197813, 251599 }, + { 181545242, 252063 }, + { 182899933, 252527 }, + { 184261908, 252991 }, + { 185631191, 253456 }, + { 187007807, 253920 }, + { 188391778, 254385 }, + { 189783129, 254851 }, + { 191181884, 255316 }, + { 192588065, 255782 }, + { 194001698, 256248 }, + { 195422805, 256714 }, + { 196851411, 257181 }, + { 198287540, 257648 }, + { 199731215, 258115 }, + { 201182461, 258582 }, + { 202641302, 259050 }, + { 204107760, 259518 }, + { 205581862, 259986 }, + { 207063630, 260454 }, + { 208553088, 260923 }, + { 210050262, 261392 }, + { 211555174, 261861 }, + { 213067849, 262331 }, + { 214588312, 262800 }, + { 216116586, 263270 }, + { 217652696, 263741 }, + { 219196666, 264211 }, + { 220748520, 264682 }, + { 222308282, 265153 }, + { 223875978, 265625 }, + { 225451630, 266097 }, + { 227035265, 266569 }, + { 228626905, 267041 }, + { 230226576, 267514 }, + { 231834302, 267986 }, + { 233450107, 268460 }, + { 235074016, 268933 }, + { 236706054, 269407 }, + { 238346244, 269881 }, + { 239994613, 270355 }, + { 241651183, 270830 }, + { 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. + +*/ +u32 tfrc_calc_x(u16 s, u32 R, u32 p) +{ + int index; + u32 f; + u64 tmp1, tmp2; + + 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; + + if (index < 0) + /* p should be 0 unless there is a bug in my code */ + index = 0; + + if (R == 0) + R = 1; /* RTT can't be zero or else divide by zero */ + + BUG_ON(index >= TFRC_CALC_X_ARRSIZE); + + 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; +} + +EXPORT_SYMBOL_GPL(tfrc_calc_x); + +/* + * args: fvalue - function value to match + * returns: p closest to that value + * + * both fvalue and p are multiplied by 1,000,000 to use ints + */ +u32 tfrc_calc_x_reverse_lookup(u32 fvalue) +{ + int ctr = 0; + int small; + + if (fvalue < tfrc_calc_x_lookup[0][1]) + 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]) + 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; +} + +EXPORT_SYMBOL_GPL(tfrc_calc_x_reverse_lookup); diff --git a/net/dccp/dccp.h b/net/dccp/dccp.h new file mode 100644 index 000000000000..33456c0d5937 --- /dev/null +++ b/net/dccp/dccp.h @@ -0,0 +1,493 @@ +#ifndef _DCCP_H +#define _DCCP_H +/* + * net/dccp/dccp.h + * + * An implementation of the DCCP protocol + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <net/snmp.h> +#include <net/sock.h> +#include <net/tcp.h> + +#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) +#else +#define dccp_pr_debug(format, a...) +#define dccp_pr_debug_cat(format, a...) +#endif + +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) + +#define DCCP_TIMEWAIT_LEN (60 * HZ) /* how long to wait to destroy TIME-WAIT + * state, about 60 seconds */ + +/* draft-ietf-dccp-spec-11.txt initial RTO value */ +#define DCCP_TIMEOUT_INIT ((unsigned)(3 * HZ)) + +/* Maximal interval between probes for local resources. */ +#define DCCP_RESOURCE_PROBE_INTERVAL ((unsigned)(HZ / 2U)) + +#define DCCP_RTO_MAX ((unsigned)(120 * HZ)) /* FIXME: using TCP value */ + +extern struct proto dccp_v4_prot; + +/* is seq1 < seq2 ? */ +static inline int before48(const u64 seq1, const u64 seq2) +{ + return (s64)((seq1 << 16) - (seq2 << 16)) < 0; +} + +/* is seq1 > seq2 ? */ +static inline int after48(const u64 seq1, const u64 seq2) +{ + return (s64)((seq2 << 16) - (seq1 << 16)) < 0; +} + +/* is seq2 <= seq1 <= seq3 ? */ +static inline int between48(const u64 seq1, const u64 seq2, const u64 seq3) +{ + return (seq3 << 16) - (seq2 << 16) >= (seq1 << 16) - (seq2 << 16); +} + +static inline u64 max48(const u64 seq1, const u64 seq2) +{ + return after48(seq1, seq2) ? seq1 : seq2; +} + +enum { + DCCP_MIB_NUM = 0, + DCCP_MIB_ACTIVEOPENS, /* ActiveOpens */ + DCCP_MIB_ESTABRESETS, /* EstabResets */ + DCCP_MIB_CURRESTAB, /* CurrEstab */ + DCCP_MIB_OUTSEGS, /* OutSegs */ + DCCP_MIB_OUTRSTS, + DCCP_MIB_ABORTONTIMEOUT, + DCCP_MIB_TIMEOUTS, + DCCP_MIB_ABORTFAILED, + DCCP_MIB_PASSIVEOPENS, + DCCP_MIB_ATTEMPTFAILS, + DCCP_MIB_OUTDATAGRAMS, + DCCP_MIB_INERRS, + DCCP_MIB_OPTMANDATORYERROR, + DCCP_MIB_INVALIDOPT, + __DCCP_MIB_MAX +}; + +#define DCCP_MIB_MAX __DCCP_MIB_MAX +struct dccp_mib { + unsigned long mibs[DCCP_MIB_MAX]; +} __SNMP_MIB_ALIGN__; + +DECLARE_SNMP_STAT(struct dccp_mib, dccp_statistics); +#define DCCP_INC_STATS(field) SNMP_INC_STATS(dccp_statistics, field) +#define DCCP_INC_STATS_BH(field) SNMP_INC_STATS_BH(dccp_statistics, field) +#define DCCP_INC_STATS_USER(field) SNMP_INC_STATS_USER(dccp_statistics, field) +#define DCCP_DEC_STATS(field) SNMP_DEC_STATS(dccp_statistics, field) +#define DCCP_ADD_STATS_BH(field, val) \ + SNMP_ADD_STATS_BH(dccp_statistics, field, val) +#define DCCP_ADD_STATS_USER(field, val) \ + SNMP_ADD_STATS_USER(dccp_statistics, field, val) + +extern int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb); +extern int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb); + +extern int dccp_send_response(struct sock *sk); +extern void dccp_send_ack(struct sock *sk); +extern void dccp_send_delayed_ack(struct sock *sk); +extern void dccp_send_sync(struct sock *sk, const u64 seq, + const enum dccp_pkt_type pkt_type); + +extern int dccp_write_xmit(struct sock *sk, struct sk_buff *skb, long *timeo); +extern void dccp_write_space(struct sock *sk); + +extern void dccp_init_xmit_timers(struct sock *sk); +static inline void dccp_clear_xmit_timers(struct sock *sk) +{ + inet_csk_clear_xmit_timers(sk); +} + +extern unsigned int dccp_sync_mss(struct sock *sk, u32 pmtu); + +extern const char *dccp_packet_name(const int type); +extern const char *dccp_state_name(const int state); + +static inline void dccp_set_state(struct sock *sk, const int state) +{ + const int oldstate = sk->sk_state; + + dccp_pr_debug("%s(%p) %-10.10s -> %s\n", + dccp_role(sk), sk, + dccp_state_name(oldstate), dccp_state_name(state)); + WARN_ON(state == oldstate); + + switch (state) { + case DCCP_OPEN: + if (oldstate != DCCP_OPEN) + DCCP_INC_STATS(DCCP_MIB_CURRESTAB); + break; + + case DCCP_CLOSED: + if (oldstate == DCCP_CLOSING || oldstate == DCCP_OPEN) + DCCP_INC_STATS(DCCP_MIB_ESTABRESETS); + + sk->sk_prot->unhash(sk); + if (inet_csk(sk)->icsk_bind_hash != NULL && + !(sk->sk_userlocks & SOCK_BINDPORT_LOCK)) + inet_put_port(&dccp_hashinfo, sk); + /* fall through */ + default: + if (oldstate == DCCP_OPEN) + DCCP_DEC_STATS(DCCP_MIB_CURRESTAB); + } + + /* Change state AFTER socket is unhashed to avoid closed + * socket sitting in hash tables. + */ + sk->sk_state = state; +} + +static inline void dccp_done(struct sock *sk) +{ + dccp_set_state(sk, DCCP_CLOSED); + dccp_clear_xmit_timers(sk); + + sk->sk_shutdown = SHUTDOWN_MASK; + + if (!sock_flag(sk, SOCK_DEAD)) + sk->sk_state_change(sk); + else + inet_csk_destroy_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 int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb); + +extern struct sock *dccp_create_openreq_child(struct sock *sk, + const struct request_sock *req, + const struct sk_buff *skb); + +extern int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb); + +extern void dccp_v4_err(struct sk_buff *skb, u32); + +extern int dccp_v4_rcv(struct sk_buff *skb); + +extern struct sock *dccp_v4_request_recv_sock(struct sock *sk, + struct sk_buff *skb, + struct request_sock *req, + struct dst_entry *dst); +extern struct sock *dccp_check_req(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct request_sock **prev); + +extern int dccp_child_process(struct sock *parent, struct sock *child, + struct sk_buff *skb); +extern int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + struct dccp_hdr *dh, unsigned len); +extern int dccp_rcv_established(struct sock *sk, struct sk_buff *skb, + const struct dccp_hdr *dh, const unsigned len); + +extern void dccp_close(struct sock *sk, long timeout); +extern struct sk_buff *dccp_make_response(struct sock *sk, + struct dst_entry *dst, + struct request_sock *req); +extern struct sk_buff *dccp_make_reset(struct sock *sk, + struct dst_entry *dst, + enum dccp_reset_codes code); + +extern int dccp_connect(struct sock *sk); +extern int dccp_disconnect(struct sock *sk, int flags); +extern int dccp_getsockopt(struct sock *sk, int level, int optname, + char __user *optval, int __user *optlen); +extern int dccp_setsockopt(struct sock *sk, int level, int optname, + char __user *optval, int optlen); +extern int dccp_ioctl(struct sock *sk, int cmd, unsigned long arg); +extern int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, + struct msghdr *msg, size_t size); +extern int dccp_recvmsg(struct kiocb *iocb, struct sock *sk, + struct msghdr *msg, size_t len, int nonblock, + int flags, int *addr_len); +extern void dccp_shutdown(struct sock *sk, int how); + +extern int dccp_v4_checksum(const struct sk_buff *skb, + const u32 saddr, const u32 daddr); + +extern int dccp_v4_send_reset(struct sock *sk, + enum dccp_reset_codes code); +extern void dccp_send_close(struct sock *sk, const int active); + +struct dccp_skb_cb { + __u8 dccpd_type; + __u8 dccpd_reset_code; + __u8 dccpd_service; + __u8 dccpd_ccval; + __u64 dccpd_seq; + __u64 dccpd_ack_seq; + int dccpd_opt_len; +}; + +#define DCCP_SKB_CB(__skb) ((struct dccp_skb_cb *)&((__skb)->cb[0])) + +static inline int dccp_non_data_packet(const struct sk_buff *skb) +{ + const __u8 type = DCCP_SKB_CB(skb)->dccpd_type; + + return type == DCCP_PKT_ACK || + type == DCCP_PKT_CLOSE || + type == DCCP_PKT_CLOSEREQ || + type == DCCP_PKT_RESET || + type == DCCP_PKT_SYNC || + type == DCCP_PKT_SYNCACK; +} + +static inline int dccp_packet_without_ack(const struct sk_buff *skb) +{ + const __u8 type = DCCP_SKB_CB(skb)->dccpd_type; + + return type == DCCP_PKT_DATA || type == DCCP_PKT_REQUEST; +} + +#define DCCP_MAX_SEQNO ((((u64)1) << 48) - 1) +#define DCCP_PKT_WITHOUT_ACK_SEQ (DCCP_MAX_SEQNO << 2) + +static inline void dccp_set_seqno(u64 *seqno, u64 value) +{ + if (value > DCCP_MAX_SEQNO) + value -= DCCP_MAX_SEQNO + 1; + *seqno = value; +} + +static inline u64 dccp_delta_seqno(u64 seqno1, u64 seqno2) +{ + return ((seqno2 << 16) - (seqno1 << 16)) >> 16; +} + +static inline void dccp_inc_seqno(u64 *seqno) +{ + if (++*seqno > DCCP_MAX_SEQNO) + *seqno = 0; +} + +static inline void dccp_hdr_set_seq(struct dccp_hdr *dh, const u64 gss) +{ + struct dccp_hdr_ext *dhx = (struct dccp_hdr_ext *)((void *)dh + + sizeof(*dh)); + +#if defined(__LITTLE_ENDIAN_BITFIELD) + dh->dccph_seq = htonl((gss >> 32)) >> 8; +#elif defined(__BIG_ENDIAN_BITFIELD) + dh->dccph_seq = htonl((gss >> 32)); +#else +#error "Adjust your <asm/byteorder.h> defines" +#endif + dhx->dccph_seq_low = htonl(gss & 0xffffffff); +} + +static inline void dccp_hdr_set_ack(struct dccp_hdr_ack_bits *dhack, + const u64 gsr) +{ +#if defined(__LITTLE_ENDIAN_BITFIELD) + dhack->dccph_ack_nr_high = htonl((gsr >> 32)) >> 8; +#elif defined(__BIG_ENDIAN_BITFIELD) + dhack->dccph_ack_nr_high = htonl((gsr >> 32)); +#else +#error "Adjust your <asm/byteorder.h> defines" +#endif + dhack->dccph_ack_nr_low = htonl(gsr & 0xffffffff); +} + +static inline void dccp_update_gsr(struct sock *sk, u64 seq) +{ + struct dccp_sock *dp = dccp_sk(sk); + + dp->dccps_gsr = seq; + dccp_set_seqno(&dp->dccps_swl, + (dp->dccps_gsr + 1 - + (dp->dccps_options.dccpo_sequence_window / 4))); + dccp_set_seqno(&dp->dccps_swh, + (dp->dccps_gsr + + (3 * dp->dccps_options.dccpo_sequence_window) / 4)); +} + +static inline void dccp_update_gss(struct sock *sk, u64 seq) +{ + struct dccp_sock *dp = dccp_sk(sk); + + dp->dccps_awh = dp->dccps_gss = seq; + dccp_set_seqno(&dp->dccps_awl, + (dp->dccps_gss - + dp->dccps_options.dccpo_sequence_window + 1)); +} + +extern void dccp_insert_options(struct sock *sk, struct sk_buff *skb); +extern void dccp_insert_option_elapsed_time(struct sock *sk, + struct sk_buff *skb, + u32 elapsed_time); +extern void dccp_insert_option_timestamp(struct sock *sk, + struct sk_buff *skb); +extern void dccp_insert_option(struct sock *sk, struct sk_buff *skb, + unsigned char option, + const void *value, unsigned char len); + +extern struct socket *dccp_ctl_socket; + +#define DCCP_ACKPKTS_STATE_RECEIVED 0 +#define DCCP_ACKPKTS_STATE_ECN_MARKED (1 << 6) +#define DCCP_ACKPKTS_STATE_NOT_RECEIVED (3 << 6) + +#define DCCP_ACKPKTS_STATE_MASK 0xC0 /* 11000000 */ +#define DCCP_ACKPKTS_LEN_MASK 0x3F /* 00111111 */ + +/** struct dccp_ackpkts - acknowledgeable packets + * + * This data structure is the one defined in the DCCP draft + * Appendix A. + * + * @dccpap_buf_head - circular buffer head + * @dccpap_buf_tail - circular buffer tail + * @dccpap_buf_ackno - ack # of the most recent packet acknowledgeable in the + * buffer (i.e. %dccpap_buf_head) + * @dccpap_buf_nonce - the one-bit sum of the ECN Nonces on all packets acked + * by the buffer with State 0 + * + * Additionally, the HC-Receiver must keep some information about the + * Ack Vectors it has recently sent. For each packet sent carrying an + * Ack Vector, it remembers four variables: + * + * @dccpap_ack_seqno - the Sequence Number used for the packet + * (HC-Receiver seqno) + * @dccpap_ack_ptr - the value of buf_head at the time of acknowledgement. + * @dccpap_ack_ackno - the Acknowledgement Number used for the packet + * (HC-Sender seqno) + * @dccpap_ack_nonce - the one-bit sum of the ECN Nonces for all State 0. + * + * @dccpap_buf_len - circular buffer length + * @dccpap_time - the time in usecs + * @dccpap_buf - circular buffer of acknowledgeable packets + */ +struct dccp_ackpkts { + unsigned int dccpap_buf_head; + unsigned int dccpap_buf_tail; + u64 dccpap_buf_ackno; + u64 dccpap_ack_seqno; + u64 dccpap_ack_ackno; + unsigned int dccpap_ack_ptr; + unsigned int dccpap_buf_vector_len; + unsigned int dccpap_ack_vector_len; + unsigned int dccpap_buf_len; + struct timeval dccpap_time; + u8 dccpap_buf_nonce; + u8 dccpap_ack_nonce; + u8 dccpap_buf[0]; +}; + +extern struct dccp_ackpkts * + dccp_ackpkts_alloc(unsigned int len, + const unsigned int __nocast priority); +extern void dccp_ackpkts_free(struct dccp_ackpkts *ap); +extern int dccp_ackpkts_add(struct dccp_ackpkts *ap, u64 ackno, u8 state); +extern void dccp_ackpkts_check_rcv_ackno(struct dccp_ackpkts *ap, + struct sock *sk, u64 ackno); + +static inline suseconds_t timeval_usecs(const struct timeval *tv) +{ + return tv->tv_sec * USEC_PER_SEC + tv->tv_usec; +} + +static inline suseconds_t timeval_delta(const struct timeval *large, + const struct timeval *small) +{ + time_t secs = large->tv_sec - small->tv_sec; + suseconds_t usecs = large->tv_usec - small->tv_usec; + + if (usecs < 0) { + secs--; + usecs += USEC_PER_SEC; + } + return secs * USEC_PER_SEC + usecs; +} + +static inline void timeval_add_usecs(struct timeval *tv, + const suseconds_t usecs) +{ + tv->tv_usec += usecs; + while (tv->tv_usec >= USEC_PER_SEC) { + tv->tv_sec++; + tv->tv_usec -= USEC_PER_SEC; + } +} + +static inline void timeval_sub_usecs(struct timeval *tv, + const suseconds_t usecs) +{ + tv->tv_usec -= usecs; + while (tv->tv_usec < 0) { + tv->tv_sec--; + tv->tv_usec += USEC_PER_SEC; + } +} + +/* + * Returns the difference in usecs between timeval + * passed in and current time + */ +static inline suseconds_t timeval_now_delta(const struct timeval *tv) +{ + struct timeval now; + do_gettimeofday(&now); + return timeval_delta(&now, tv); +} + +#ifdef CONFIG_IP_DCCP_DEBUG +extern void dccp_ackvector_print(const u64 ackno, + const unsigned char *vector, int len); +extern void dccp_ackpkts_print(const struct dccp_ackpkts *ap); +#else +static inline void dccp_ackvector_print(const u64 ackno, + const unsigned char *vector, + int len) { } +static inline void dccp_ackpkts_print(const struct dccp_ackpkts *ap) { } +#endif + +#endif /* _DCCP_H */ diff --git a/net/dccp/diag.c b/net/dccp/diag.c new file mode 100644 index 000000000000..f675d8e642d3 --- /dev/null +++ b/net/dccp/diag.c @@ -0,0 +1,71 @@ +/* + * net/dccp/diag.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@mandriva.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/config.h> + +#include <linux/module.h> +#include <linux/inet_diag.h> + +#include "ccid.h" +#include "dccp.h" + +static void dccp_get_info(struct sock *sk, struct tcp_info *info) +{ + struct dccp_sock *dp = dccp_sk(sk); + const struct inet_connection_sock *icsk = inet_csk(sk); + + memset(info, 0, sizeof(*info)); + + info->tcpi_state = sk->sk_state; + info->tcpi_retransmits = icsk->icsk_retransmits; + info->tcpi_probes = icsk->icsk_probes_out; + info->tcpi_backoff = icsk->icsk_backoff; + info->tcpi_pmtu = dp->dccps_pmtu_cookie; + + if (dp->dccps_options.dccpo_send_ack_vector) + info->tcpi_options |= TCPI_OPT_SACK; + + ccid_hc_rx_get_info(dp->dccps_hc_rx_ccid, sk, info); + ccid_hc_tx_get_info(dp->dccps_hc_tx_ccid, sk, info); +} + +static void dccp_diag_get_info(struct sock *sk, struct inet_diag_msg *r, + void *_info) +{ + r->idiag_rqueue = r->idiag_wqueue = 0; + + if (_info != NULL) + dccp_get_info(sk, _info); +} + +static struct inet_diag_handler dccp_diag_handler = { + .idiag_hashinfo = &dccp_hashinfo, + .idiag_get_info = dccp_diag_get_info, + .idiag_type = DCCPDIAG_GETSOCK, + .idiag_info_size = sizeof(struct tcp_info), +}; + +static int __init dccp_diag_init(void) +{ + return inet_diag_register(&dccp_diag_handler); +} + +static void __exit dccp_diag_fini(void) +{ + inet_diag_unregister(&dccp_diag_handler); +} + +module_init(dccp_diag_init); +module_exit(dccp_diag_fini); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Arnaldo Carvalho de Melo <acme@mandriva.com>"); +MODULE_DESCRIPTION("DCCP inet_diag handler"); diff --git a/net/dccp/input.c b/net/dccp/input.c new file mode 100644 index 000000000000..ef29cef1dafe --- /dev/null +++ b/net/dccp/input.c @@ -0,0 +1,600 @@ +/* + * net/dccp/input.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/skbuff.h> + +#include <net/sock.h> + +#include "ccid.h" +#include "dccp.h" + +static void dccp_fin(struct sock *sk, struct sk_buff *skb) +{ + sk->sk_shutdown |= RCV_SHUTDOWN; + sock_set_flag(sk, SOCK_DONE); + __skb_pull(skb, dccp_hdr(skb)->dccph_doff * 4); + __skb_queue_tail(&sk->sk_receive_queue, skb); + skb_set_owner_r(skb, sk); + sk->sk_data_ready(sk, 0); +} + +static void dccp_rcv_close(struct sock *sk, struct sk_buff *skb) +{ + dccp_v4_send_reset(sk, DCCP_RESET_CODE_CLOSED); + dccp_fin(sk, skb); + dccp_set_state(sk, DCCP_CLOSED); + sk_wake_async(sk, 1, POLL_HUP); +} + +static void dccp_rcv_closereq(struct sock *sk, struct sk_buff *skb) +{ + /* + * Step 7: Check for unexpected packet types + * If (S.is_server and P.type == CloseReq) + * Send Sync packet acknowledging P.seqno + * Drop packet and return + */ + if (dccp_sk(sk)->dccps_role != DCCP_ROLE_CLIENT) { + dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq, DCCP_PKT_SYNC); + return; + } + + dccp_set_state(sk, DCCP_CLOSING); + dccp_send_close(sk, 0); +} + +static inline void dccp_event_ack_recv(struct sock *sk, struct sk_buff *skb) +{ + struct dccp_sock *dp = dccp_sk(sk); + + if (dp->dccps_options.dccpo_send_ack_vector) + dccp_ackpkts_check_rcv_ackno(dp->dccps_hc_rx_ackpkts, sk, + DCCP_SKB_CB(skb)->dccpd_ack_seq); +} + +static int dccp_check_seqno(struct sock *sk, struct sk_buff *skb) +{ + const struct dccp_hdr *dh = dccp_hdr(skb); + struct dccp_sock *dp = dccp_sk(sk); + u64 lswl, lawl; + + /* + * Step 5: Prepare sequence numbers for Sync + * If P.type == Sync or P.type == SyncAck, + * If S.AWL <= P.ackno <= S.AWH and P.seqno >= S.SWL, + * / * P is valid, so update sequence number variables + * accordingly. After this update, P will pass the tests + * in Step 6. A SyncAck is generated if necessary in + * Step 15 * / + * Update S.GSR, S.SWL, S.SWH + * Otherwise, + * Drop packet and return + */ + 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) && + !before48(DCCP_SKB_CB(skb)->dccpd_seq, dp->dccps_swl)) + dccp_update_gsr(sk, DCCP_SKB_CB(skb)->dccpd_seq); + else + return -1; + } + + /* + * Step 6: Check sequence numbers + * Let LSWL = S.SWL and LAWL = S.AWL + * If P.type == CloseReq or P.type == Close or P.type == Reset, + * LSWL := S.GSR + 1, LAWL := S.GAR + * If LSWL <= P.seqno <= S.SWH + * and (P.ackno does not exist or LAWL <= P.ackno <= S.AWH), + * Update S.GSR, S.SWL, S.SWH + * If P.type != Sync, + * Update S.GAR + * Otherwise, + * Send Sync packet acknowledging P.seqno + * Drop packet and return + */ + lswl = dp->dccps_swl; + lawl = dp->dccps_awl; + + if (dh->dccph_type == DCCP_PKT_CLOSEREQ || + dh->dccph_type == DCCP_PKT_CLOSE || + dh->dccph_type == DCCP_PKT_RESET) { + lswl = dp->dccps_gsr; + dccp_inc_seqno(&lswl); + lawl = dp->dccps_gar; + } + + if (between48(DCCP_SKB_CB(skb)->dccpd_seq, lswl, dp->dccps_swh) && + (DCCP_SKB_CB(skb)->dccpd_ack_seq == DCCP_PKT_WITHOUT_ACK_SEQ || + between48(DCCP_SKB_CB(skb)->dccpd_ack_seq, + lawl, dp->dccps_awh))) { + dccp_update_gsr(sk, DCCP_SKB_CB(skb)->dccpd_seq); + + if (dh->dccph_type != DCCP_PKT_SYNC && + (DCCP_SKB_CB(skb)->dccpd_ack_seq != + 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_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); + dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq, DCCP_PKT_SYNC); + return -1; + } + + return 0; +} + +int dccp_rcv_established(struct sock *sk, struct sk_buff *skb, + const struct dccp_hdr *dh, const unsigned len) +{ + struct dccp_sock *dp = dccp_sk(sk); + + if (dccp_check_seqno(sk, skb)) + goto discard; + + if (dccp_parse_options(sk, skb)) + goto discard; + + if (DCCP_SKB_CB(skb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ) + dccp_event_ack_recv(sk, skb); + + /* + * FIXME: check ECN to see if we should use + * DCCP_ACKPKTS_STATE_ECN_MARKED + */ + if (dp->dccps_options.dccpo_send_ack_vector) { + struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts; + + if (dccp_ackpkts_add(dp->dccps_hc_rx_ackpkts, + DCCP_SKB_CB(skb)->dccpd_seq, + DCCP_ACKPKTS_STATE_RECEIVED)) { + LIMIT_NETDEBUG(KERN_WARNING "DCCP: acknowledgeable " + "packets buffer full!\n"); + ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1; + inet_csk_schedule_ack(sk); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + TCP_DELACK_MIN, + DCCP_RTO_MAX); + goto discard; + } + + /* + * FIXME: this activation is probably wrong, have to study more + * TCP delack machinery and how it fits into DCCP draft, but + * for now it kinda "works" 8) + */ + if (!inet_csk_ack_scheduled(sk)) { + inet_csk_schedule_ack(sk); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, 5 * HZ, + DCCP_RTO_MAX); + } + } + + ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb); + ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb); + + switch (dccp_hdr(skb)->dccph_type) { + case DCCP_PKT_DATAACK: + case DCCP_PKT_DATA: + /* + * FIXME: check if sk_receive_queue is full, schedule DATA_DROPPED + * option if it is. + */ + __skb_pull(skb, dh->dccph_doff * 4); + __skb_queue_tail(&sk->sk_receive_queue, skb); + skb_set_owner_r(skb, sk); + sk->sk_data_ready(sk, 0); + return 0; + case DCCP_PKT_ACK: + goto discard; + case DCCP_PKT_RESET: + /* + * Step 9: Process Reset + * If P.type == Reset, + * Tear down connection + * S.state := TIMEWAIT + * Set TIMEWAIT timer + * Drop packet and return + */ + dccp_fin(sk, skb); + dccp_time_wait(sk, DCCP_TIME_WAIT, 0); + return 0; + case DCCP_PKT_CLOSEREQ: + dccp_rcv_closereq(sk, skb); + goto discard; + case DCCP_PKT_CLOSE: + dccp_rcv_close(sk, skb); + return 0; + case DCCP_PKT_REQUEST: + /* 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) + * or (S.state >= OPEN and P.type == Response + * and P.seqno >= S.OSR) + * or (S.state == RESPOND and P.type == Data), + * Send Sync packet acknowledging P.seqno + * Drop packet and return + */ + if (dp->dccps_role != DCCP_ROLE_LISTEN) + goto send_sync; + goto check_seq; + case DCCP_PKT_RESPONSE: + if (dp->dccps_role != DCCP_ROLE_CLIENT) + goto send_sync; +check_seq: + if (!before48(DCCP_SKB_CB(skb)->dccpd_seq, dp->dccps_osr)) { +send_sync: + dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq, + DCCP_PKT_SYNC); + } + break; + case DCCP_PKT_SYNC: + dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq, + DCCP_PKT_SYNCACK); + /* + * From the draft: + * + * As with DCCP-Ack packets, DCCP-Sync and DCCP-SyncAck packets + * MAY have non-zero-length application data areas, whose + * contents * receivers MUST ignore. + */ + goto discard; + } + + DCCP_INC_STATS_BH(DCCP_MIB_INERRS); +discard: + __kfree_skb(skb); + return 0; +} + +static int dccp_rcv_request_sent_state_process(struct sock *sk, + struct sk_buff *skb, + 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) + * and S.AWL <= P.ackno <= S.AWH, + * / * Set sequence number variables corresponding to the + * other endpoint, so P will pass the tests in Step 6 * / + * Set S.GSR, S.ISR, S.SWL, S.SWH + * / * Response processing continues in Step 10; Reset + * processing continues in Step 9 * / + */ + if (dh->dccph_type == DCCP_PKT_RESPONSE) { + const struct inet_connection_sock *icsk = inet_csk(sk); + struct dccp_sock *dp = dccp_sk(sk); + + /* Stop the REQUEST timer */ + inet_csk_clear_xmit_timer(sk, ICSK_TIME_RETRANS); + BUG_TRAP(sk->sk_send_head != NULL); + __kfree_skb(sk->sk_send_head); + sk->sk_send_head = NULL; + + if (!between48(DCCP_SKB_CB(skb)->dccpd_ack_seq, + dp->dccps_awl, dp->dccps_awh)) { + dccp_pr_debug("invalid ackno: S.AWL=%llu, " + "P.ackno=%llu, S.AWH=%llu \n", + (unsigned long long)dp->dccps_awl, + (unsigned long long)DCCP_SKB_CB(skb)->dccpd_ack_seq, + (unsigned long long)dp->dccps_awh); + goto out_invalid_packet; + } + + dp->dccps_isr = DCCP_SKB_CB(skb)->dccpd_seq; + dccp_update_gsr(sk, dp->dccps_isr); + /* + * 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. + * + * AWL was adjusted in dccp_v4_connect -acme + */ + dccp_set_seqno(&dp->dccps_swl, + max48(dp->dccps_swl, dp->dccps_isr)); + + if (ccid_hc_rx_init(dp->dccps_hc_rx_ccid, sk) != 0 || + ccid_hc_tx_init(dp->dccps_hc_tx_ccid, sk) != 0) { + ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk); + ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk); + /* FIXME: send appropriate RESET code */ + goto out_invalid_packet; + } + + dccp_sync_mss(sk, dp->dccps_pmtu_cookie); + + /* + * Step 10: Process REQUEST state (second part) + * If S.state == REQUEST, + * / * If we get here, P is a valid Response from the + * server (see Step 4), and we should move to + * PARTOPEN state. PARTOPEN means send an Ack, + * don't send Data packets, retransmit Acks + * periodically, and always include any Init Cookie + * from the Response * / + * S.state := PARTOPEN + * Set PARTOPEN timer + * Continue with S.state == PARTOPEN + * / * Step 12 will send the Ack completing the + * three-way handshake * / + */ + dccp_set_state(sk, DCCP_PARTOPEN); + + /* Make sure socket is routed, for correct metrics. */ + inet_sk_rebuild_header(sk); + + if (!sock_flag(sk, SOCK_DEAD)) { + sk->sk_state_change(sk); + sk_wake_async(sk, 0, POLL_OUT); + } + + if (sk->sk_write_pending || icsk->icsk_ack.pingpong || + icsk->icsk_accept_queue.rskq_defer_accept) { + /* Save one ACK. Data will be ready after + * several ticks, if write_pending is set. + * + * It may be deleted, but with this feature tcpdumps + * look so _wonderfully_ clever, that I was not able + * to stand against the temptation 8) --ANK + */ + /* + * OK, in DCCP we can as well do a similar trick, its + * even in the draft, but there is no need for us to + * schedule an ack here, as dccp_sendmsg does this for + * us, also stated in the draft. -acme + */ + __kfree_skb(skb); + return 0; + } + dccp_send_ack(sk); + return -1; + } + +out_invalid_packet: + return 1; /* dccp_v4_do_rcv will send a reset, but... + FIXME: the reset code should be + DCCP_RESET_CODE_PACKET_ERROR */ +} + +static int dccp_rcv_respond_partopen_state_process(struct sock *sk, + struct sk_buff *skb, + const struct dccp_hdr *dh, + const unsigned len) +{ + int queued = 0; + + switch (dh->dccph_type) { + case DCCP_PKT_RESET: + inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK); + break; + case DCCP_PKT_DATAACK: + case DCCP_PKT_ACK: + /* + * FIXME: we should be reseting the PARTOPEN (DELACK) timer + * here but only if we haven't used the DELACK timer for + * something else, like sending a delayed ack for a TIMESTAMP + * echo, etc, for now were not clearing it, sending an extra + * ACK when there is nothing else to do in DELACK is not a big + * deal after all. + */ + + /* Stop the PARTOPEN timer */ + if (sk->sk_state == DCCP_PARTOPEN) + inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK); + + dccp_sk(sk)->dccps_osr = DCCP_SKB_CB(skb)->dccpd_seq; + dccp_set_state(sk, DCCP_OPEN); + + if (dh->dccph_type == DCCP_PKT_DATAACK) { + dccp_rcv_established(sk, skb, dh, len); + queued = 1; /* packet was queued + (by dccp_rcv_established) */ + } + break; + } + + return queued; +} + +int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + struct dccp_hdr *dh, unsigned len) +{ + struct dccp_sock *dp = dccp_sk(sk); + const int old_state = sk->sk_state; + int queued = 0; + + /* + * 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 (sk->sk_state == DCCP_LISTEN) { + if (dh->dccph_type == DCCP_PKT_REQUEST) { + if (dccp_v4_conn_request(sk, skb) < 0) + return 1; + + /* FIXME: do congestion control initialization */ + goto discard; + } + if (dh->dccph_type == DCCP_PKT_RESET) + goto discard; + + /* Caller (dccp_v4_do_rcv) will send Reset(No Connection)*/ + return 1; + } + + if (sk->sk_state != DCCP_REQUESTING) { + if (dccp_check_seqno(sk, skb)) + goto discard; + + /* + * Step 8: Process options and mark acknowledgeable + */ + if (dccp_parse_options(sk, skb)) + goto discard; + + if (DCCP_SKB_CB(skb)->dccpd_ack_seq != + DCCP_PKT_WITHOUT_ACK_SEQ) + dccp_event_ack_recv(sk, skb); + + ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb); + ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb); + + /* + * FIXME: check ECN to see if we should use + * DCCP_ACKPKTS_STATE_ECN_MARKED + */ + if (dp->dccps_options.dccpo_send_ack_vector) { + if (dccp_ackpkts_add(dp->dccps_hc_rx_ackpkts, + DCCP_SKB_CB(skb)->dccpd_seq, + DCCP_ACKPKTS_STATE_RECEIVED)) + goto discard; + /* + * FIXME: this activation is probably wrong, have to + * study more TCP delack machinery and how it fits into + * DCCP draft, but for now it kinda "works" 8) + */ + if ((dp->dccps_hc_rx_ackpkts->dccpap_ack_seqno == + DCCP_MAX_SEQNO + 1) && + !inet_csk_ack_scheduled(sk)) { + inet_csk_schedule_ack(sk); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + TCP_DELACK_MIN, + DCCP_RTO_MAX); + } + } + } + + /* + * Step 9: Process Reset + * If P.type == Reset, + * Tear down connection + * S.state := TIMEWAIT + * Set TIMEWAIT timer + * Drop packet and return + */ + if (dh->dccph_type == DCCP_PKT_RESET) { + /* + * Queue the equivalent of TCP fin so that dccp_recvmsg + * exits the loop + */ + dccp_fin(sk, skb); + dccp_time_wait(sk, DCCP_TIME_WAIT, 0); + return 0; + /* + * Step 7: Check for unexpected packet types + * If (S.is_server and P.type == CloseReq) + * or (S.is_server and P.type == Response) + * or (S.is_client and P.type == Request) + * or (S.state == RESPOND and P.type == Data), + * Send Sync packet acknowledging P.seqno + * Drop packet and return + */ + } else if ((dp->dccps_role != DCCP_ROLE_CLIENT && + (dh->dccph_type == DCCP_PKT_RESPONSE || + dh->dccph_type == DCCP_PKT_CLOSEREQ)) || + (dp->dccps_role == DCCP_ROLE_CLIENT && + dh->dccph_type == DCCP_PKT_REQUEST) || + (sk->sk_state == DCCP_RESPOND && + dh->dccph_type == DCCP_PKT_DATA)) { + dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq, + DCCP_PKT_SYNC); + goto discard; + } else if (dh->dccph_type == DCCP_PKT_CLOSEREQ) { + dccp_rcv_closereq(sk, skb); + goto discard; + } else if (dh->dccph_type == DCCP_PKT_CLOSE) { + dccp_rcv_close(sk, skb); + return 0; + } + + switch (sk->sk_state) { + case DCCP_CLOSED: + return 1; + + case DCCP_REQUESTING: + /* FIXME: do congestion control initialization */ + + queued = dccp_rcv_request_sent_state_process(sk, skb, dh, len); + if (queued >= 0) + return queued; + + __kfree_skb(skb); + return 0; + + case DCCP_RESPOND: + case DCCP_PARTOPEN: + queued = dccp_rcv_respond_partopen_state_process(sk, skb, + dh, len); + break; + } + + if (dh->dccph_type == DCCP_PKT_ACK || + dh->dccph_type == DCCP_PKT_DATAACK) { + switch (old_state) { + case DCCP_PARTOPEN: + sk->sk_state_change(sk); + sk_wake_async(sk, 0, POLL_OUT); + break; + } + } + + if (!queued) { +discard: + __kfree_skb(skb); + } + return 0; +} diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c new file mode 100644 index 000000000000..3fc75dbee4b8 --- /dev/null +++ b/net/dccp/ipv4.c @@ -0,0 +1,1356 @@ +/* + * net/dccp/ipv4.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/icmp.h> +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/random.h> + +#include <net/icmp.h> +#include <net/inet_hashtables.h> +#include <net/sock.h> +#include <net/tcp_states.h> +#include <net/xfrm.h> + +#include "ccid.h" +#include "dccp.h" + +struct inet_hashinfo __cacheline_aligned dccp_hashinfo = { + .lhash_lock = RW_LOCK_UNLOCKED, + .lhash_users = ATOMIC_INIT(0), + .lhash_wait = __WAIT_QUEUE_HEAD_INITIALIZER(dccp_hashinfo.lhash_wait), + .portalloc_lock = SPIN_LOCK_UNLOCKED, + .port_rover = 1024 - 1, +}; + +EXPORT_SYMBOL_GPL(dccp_hashinfo); + +static int dccp_v4_get_port(struct sock *sk, const unsigned short snum) +{ + return inet_csk_get_port(&dccp_hashinfo, sk, snum); +} + +static void dccp_v4_hash(struct sock *sk) +{ + inet_hash(&dccp_hashinfo, sk); +} + +static void dccp_v4_unhash(struct sock *sk) +{ + inet_unhash(&dccp_hashinfo, sk); +} + +/* called with local bh disabled */ +static int __dccp_v4_check_established(struct sock *sk, const __u16 lport, + struct inet_timewait_sock **twp) +{ + struct inet_sock *inet = inet_sk(sk); + const u32 daddr = inet->rcv_saddr; + const u32 saddr = inet->daddr; + const int dif = sk->sk_bound_dev_if; + INET_ADDR_COOKIE(acookie, saddr, daddr) + const __u32 ports = INET_COMBINED_PORTS(inet->dport, lport); + const int hash = inet_ehashfn(daddr, lport, saddr, inet->dport, + dccp_hashinfo.ehash_size); + struct inet_ehash_bucket *head = &dccp_hashinfo.ehash[hash]; + const struct sock *sk2; + const struct hlist_node *node; + struct inet_timewait_sock *tw; + + write_lock(&head->lock); + + /* Check TIME-WAIT sockets first. */ + sk_for_each(sk2, node, &(head + dccp_hashinfo.ehash_size)->chain) { + tw = inet_twsk(sk2); + + if (INET_TW_MATCH(sk2, acookie, saddr, daddr, ports, dif)) + goto not_unique; + } + tw = NULL; + + /* And established part... */ + sk_for_each(sk2, node, &head->chain) { + if (INET_MATCH(sk2, acookie, saddr, daddr, ports, dif)) + goto not_unique; + } + + /* Must record num and sport now. Otherwise we will see + * in hash table socket with a funny identity. */ + inet->num = lport; + inet->sport = htons(lport); + sk->sk_hashent = hash; + BUG_TRAP(sk_unhashed(sk)); + __sk_add_node(sk, &head->chain); + sock_prot_inc_use(sk->sk_prot); + write_unlock(&head->lock); + + if (twp != NULL) { + *twp = tw; + NET_INC_STATS_BH(LINUX_MIB_TIMEWAITRECYCLED); + } else if (tw != NULL) { + /* Silly. Should hash-dance instead... */ + inet_twsk_deschedule(tw, &dccp_death_row); + NET_INC_STATS_BH(LINUX_MIB_TIMEWAITRECYCLED); + + inet_twsk_put(tw); + } + + return 0; + +not_unique: + write_unlock(&head->lock); + return -EADDRNOTAVAIL; +} + +/* + * Bind a port for a connect operation and hash it. + */ +static int dccp_v4_hash_connect(struct sock *sk) +{ + const unsigned short snum = inet_sk(sk)->num; + struct inet_bind_hashbucket *head; + struct inet_bind_bucket *tb; + int ret; + + if (snum == 0) { + int rover; + int low = sysctl_local_port_range[0]; + int high = sysctl_local_port_range[1]; + int remaining = (high - low) + 1; + struct hlist_node *node; + struct inet_timewait_sock *tw = NULL; + + local_bh_disable(); + + /* TODO. Actually it is not so bad idea to remove + * dccp_hashinfo.portalloc_lock before next submission to + * Linus. + * As soon as we touch this place at all it is time to think. + * + * Now it protects single _advisory_ variable + * dccp_hashinfo.port_rover, hence it is mostly useless. + * Code will work nicely if we just delete it, but + * I am afraid in contented case it will work not better or + * even worse: another cpu just will hit the same bucket + * and spin there. + * So some cpu salt could remove both contention and + * memory pingpong. Any ideas how to do this in a nice way? + */ + spin_lock(&dccp_hashinfo.portalloc_lock); + rover = dccp_hashinfo.port_rover; + + do { + rover++; + if ((rover < low) || (rover > high)) + rover = low; + head = &dccp_hashinfo.bhash[inet_bhashfn(rover, + dccp_hashinfo.bhash_size)]; + spin_lock(&head->lock); + + /* Does not bother with rcv_saddr checks, + * because the established check is already + * unique enough. + */ + inet_bind_bucket_for_each(tb, node, &head->chain) { + if (tb->port == rover) { + BUG_TRAP(!hlist_empty(&tb->owners)); + if (tb->fastreuse >= 0) + goto next_port; + if (!__dccp_v4_check_established(sk, + rover, + &tw)) + goto ok; + goto next_port; + } + } + + tb = inet_bind_bucket_create(dccp_hashinfo.bind_bucket_cachep, + head, rover); + if (tb == NULL) { + spin_unlock(&head->lock); + break; + } + tb->fastreuse = -1; + goto ok; + + next_port: + spin_unlock(&head->lock); + } while (--remaining > 0); + dccp_hashinfo.port_rover = rover; + spin_unlock(&dccp_hashinfo.portalloc_lock); + + local_bh_enable(); + + return -EADDRNOTAVAIL; + +ok: + /* All locks still held and bhs disabled */ + dccp_hashinfo.port_rover = rover; + spin_unlock(&dccp_hashinfo.portalloc_lock); + + inet_bind_hash(sk, tb, rover); + if (sk_unhashed(sk)) { + inet_sk(sk)->sport = htons(rover); + __inet_hash(&dccp_hashinfo, sk, 0); + } + spin_unlock(&head->lock); + + if (tw != NULL) { + inet_twsk_deschedule(tw, &dccp_death_row); + inet_twsk_put(tw); + } + + ret = 0; + goto out; + } + + head = &dccp_hashinfo.bhash[inet_bhashfn(snum, + dccp_hashinfo.bhash_size)]; + tb = inet_csk(sk)->icsk_bind_hash; + spin_lock_bh(&head->lock); + if (sk_head(&tb->owners) == sk && sk->sk_bind_node.next == NULL) { + __inet_hash(&dccp_hashinfo, sk, 0); + spin_unlock_bh(&head->lock); + return 0; + } else { + spin_unlock(&head->lock); + /* No definite answer... Walk to established hash table */ + ret = __dccp_v4_check_established(sk, snum, NULL); +out: + local_bh_enable(); + return ret; + } +} + +static int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, + int addr_len) +{ + struct inet_sock *inet = inet_sk(sk); + struct dccp_sock *dp = dccp_sk(sk); + const struct sockaddr_in *usin = (struct sockaddr_in *)uaddr; + struct rtable *rt; + u32 daddr, nexthop; + int tmp; + int err; + + dp->dccps_role = DCCP_ROLE_CLIENT; + + if (addr_len < sizeof(struct sockaddr_in)) + return -EINVAL; + + if (usin->sin_family != AF_INET) + return -EAFNOSUPPORT; + + nexthop = daddr = usin->sin_addr.s_addr; + if (inet->opt != NULL && inet->opt->srr) { + if (daddr == 0) + return -EINVAL; + nexthop = inet->opt->faddr; + } + + tmp = ip_route_connect(&rt, nexthop, inet->saddr, + RT_CONN_FLAGS(sk), sk->sk_bound_dev_if, + IPPROTO_DCCP, + inet->sport, usin->sin_port, sk); + if (tmp < 0) + return tmp; + + if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) { + ip_rt_put(rt); + return -ENETUNREACH; + } + + if (inet->opt == NULL || !inet->opt->srr) + daddr = rt->rt_dst; + + if (inet->saddr == 0) + inet->saddr = rt->rt_src; + inet->rcv_saddr = inet->saddr; + + inet->dport = usin->sin_port; + inet->daddr = daddr; + + dp->dccps_ext_header_len = 0; + if (inet->opt != NULL) + dp->dccps_ext_header_len = inet->opt->optlen; + /* + * Socket identity is still unknown (sport may be zero). + * However we set state to DCCP_REQUESTING and not releasing socket + * lock select source port, enter ourselves into the hash tables and + * complete initialization after this. + */ + dccp_set_state(sk, DCCP_REQUESTING); + err = dccp_v4_hash_connect(sk); + if (err != 0) + goto failure; + + err = ip_route_newports(&rt, inet->sport, inet->dport, sk); + if (err != 0) + goto failure; + + /* 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); + + /* + * 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_set_seqno(&dp->dccps_awl, max48(dp->dccps_awl, dp->dccps_iss)); + + inet->id = dp->dccps_iss ^ jiffies; + + err = dccp_connect(sk); + rt = NULL; + if (err != 0) + goto failure; +out: + return err; +failure: + /* + * This unhashes the socket and releases the local port, if necessary. + */ + dccp_set_state(sk, DCCP_CLOSED); + ip_rt_put(rt); + sk->sk_route_caps = 0; + inet->dport = 0; + goto out; +} + +/* + * This routine does path mtu discovery as defined in RFC1191. + */ +static inline void dccp_do_pmtu_discovery(struct sock *sk, + const struct iphdr *iph, + u32 mtu) +{ + struct dst_entry *dst; + const struct inet_sock *inet = inet_sk(sk); + const struct dccp_sock *dp = dccp_sk(sk); + + /* We are not interested in DCCP_LISTEN and request_socks (RESPONSEs + * send out by Linux are always < 576bytes so they should go through + * unfragmented). + */ + if (sk->sk_state == DCCP_LISTEN) + return; + + /* 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 + * route, but I think that's acceptable. + */ + if ((dst = __sk_dst_check(sk, 0)) == NULL) + return; + + dst->ops->update_pmtu(dst, mtu); + + /* Something is about to be wrong... Remember soft error + * for the case, if this connection will not able to recover. + */ + if (mtu < dst_mtu(dst) && ip_dont_fragment(sk, dst)) + sk->sk_err_soft = EMSGSIZE; + + mtu = dst_mtu(dst); + + if (inet->pmtudisc != IP_PMTUDISC_DONT && + dp->dccps_pmtu_cookie > mtu) { + dccp_sync_mss(sk, mtu); + + /* + * From: draft-ietf-dccp-spec-11.txt + * + * DCCP-Sync packets are the best choice for upward + * probing, since DCCP-Sync probes do not risk application + * data loss. + */ + dccp_send_sync(sk, dp->dccps_gsr, DCCP_PKT_SYNC); + } /* else let the usual retransmit timer handle it */ +} + +static void dccp_v4_ctl_send_ack(struct sk_buff *rxskb) +{ + int err; + struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh; + const int 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(MAX_DCCP_HEADER + 15, GFP_ATOMIC); + if (skb == NULL) + return; + + /* Reserve space for headers. */ + skb_reserve(skb, MAX_DCCP_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_ctl_socket->sk); + err = ip_build_and_send_pkt(skb, dccp_ctl_socket->sk, + rxskb->nh.iph->daddr, + rxskb->nh.iph->saddr, NULL); + bh_unlock_sock(dccp_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 void dccp_v4_reqsk_send_ack(struct sk_buff *skb, + struct request_sock *req) +{ + dccp_v4_ctl_send_ack(skb); +} + +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); + + 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 + * returned to the user. If err > 0 it's just the icmp type << 8 | icmp code. + * After adjustment header points to the first 8 bytes of the tcp header. We + * need to find the appropriate port. + * + * The locking strategy used here is very "optimistic". When someone else + * accesses the socket the ICMP is just dropped and for some paths there is no + * check at all. A more general error queue to queue errors for later handling + * is probably better. + */ +void dccp_v4_err(struct sk_buff *skb, u32 info) +{ + const struct iphdr *iph = (struct iphdr *)skb->data; + const struct dccp_hdr *dh = (struct dccp_hdr *)(skb->data + + (iph->ihl << 2)); + struct dccp_sock *dp; + struct inet_sock *inet; + const int type = skb->h.icmph->type; + const int code = skb->h.icmph->code; + struct sock *sk; + __u64 seq; + int err; + + if (skb->len < (iph->ihl << 2) + 8) { + ICMP_INC_STATS_BH(ICMP_MIB_INERRORS); + return; + } + + sk = inet_lookup(&dccp_hashinfo, iph->daddr, dh->dccph_dport, + iph->saddr, dh->dccph_sport, inet_iif(skb)); + if (sk == NULL) { + ICMP_INC_STATS_BH(ICMP_MIB_INERRORS); + return; + } + + if (sk->sk_state == DCCP_TIME_WAIT) { + inet_twsk_put((struct inet_timewait_sock *)sk); + return; + } + + bh_lock_sock(sk); + /* If too many ICMPs get dropped on busy + * servers this needs to be solved differently. + */ + if (sock_owned_by_user(sk)) + NET_INC_STATS_BH(LINUX_MIB_LOCKDROPPEDICMPS); + + if (sk->sk_state == DCCP_CLOSED) + goto out; + + dp = dccp_sk(sk); + 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); + goto out; + } + + switch (type) { + case ICMP_SOURCE_QUENCH: + /* Just silently ignore these. */ + goto out; + case ICMP_PARAMETERPROB: + err = EPROTO; + break; + case ICMP_DEST_UNREACH: + if (code > NR_ICMP_UNREACH) + goto out; + + if (code == ICMP_FRAG_NEEDED) { /* PMTU discovery (RFC1191) */ + if (!sock_owned_by_user(sk)) + dccp_do_pmtu_discovery(sk, iph, info); + goto out; + } + + err = icmp_err_convert[code].errno; + break; + case ICMP_TIME_EXCEEDED: + err = EHOSTUNREACH; + break; + default: + goto out; + } + + switch (sk->sk_state) { + struct request_sock *req , **prev; + case DCCP_LISTEN: + if (sock_owned_by_user(sk)) + goto out; + req = inet_csk_search_req(sk, &prev, dh->dccph_dport, + iph->daddr, iph->saddr); + if (!req) + goto out; + + /* + * ICMPs are not backlogged, hence we cannot get an established + * socket here. + */ + BUG_TRAP(!req->sk); + + if (seq != dccp_rsk(req)->dreq_iss) { + NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS); + goto out; + } + /* + * Still in RESPOND, just remove it silently. + * There is no good way to pass the error to the newly + * created socket, and POSIX does not want network + * errors returned from accept(). + */ + inet_csk_reqsk_queue_drop(sk, req, prev); + goto out; + + case DCCP_REQUESTING: + case DCCP_RESPOND: + if (!sock_owned_by_user(sk)) { + DCCP_INC_STATS_BH(DCCP_MIB_ATTEMPTFAILS); + sk->sk_err = err; + + sk->sk_error_report(sk); + + dccp_done(sk); + } else + sk->sk_err_soft = err; + goto out; + } + + /* If we've already connected we will keep trying + * until we time out, or the user gives up. + * + * rfc1122 4.2.3.9 allows to consider as hard errors + * only PROTO_UNREACH and PORT_UNREACH (well, FRAG_FAILED too, + * but it is obsoleted by pmtu discovery). + * + * Note, that in modern internet, where routing is unreliable + * and in each dark corner broken firewalls sit, sending random + * errors ordered by their masters even this two messages finally lose + * their original sense (even Linux sends invalid PORT_UNREACHs) + * + * Now we are in compliance with RFCs. + * --ANK (980905) + */ + + inet = inet_sk(sk); + if (!sock_owned_by_user(sk) && inet->recverr) { + sk->sk_err = err; + sk->sk_error_report(sk); + } else /* Only an error on timeout */ + sk->sk_err_soft = err; +out: + bh_unlock_sock(sk); + sock_put(sk); +} + +int dccp_v4_send_reset(struct sock *sk, enum dccp_reset_codes code) +{ + struct sk_buff *skb; + /* + * FIXME: what if rebuild_header fails? + * Should we be doing a rebuild_header here? + */ + int err = inet_sk_rebuild_header(sk); + + if (err != 0) + return err; + + skb = dccp_make_reset(sk, sk->sk_dst_cache, code); + if (skb != NULL) { + const struct dccp_sock *dp = dccp_sk(sk); + const struct inet_sock *inet = inet_sk(sk); + + err = ip_build_and_send_pkt(skb, sk, + inet->saddr, inet->daddr, NULL); + if (err == NET_XMIT_CN) + err = 0; + + ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk); + ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk); + } + + return err; +} + +static inline u64 dccp_v4_init_sequence(const struct sock *sk, + const struct sk_buff *skb) +{ + return secure_dccp_sequence_number(skb->nh.iph->daddr, + skb->nh.iph->saddr, + dccp_hdr(skb)->dccph_dport, + dccp_hdr(skb)->dccph_sport); +} + +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 __u32 saddr = skb->nh.iph->saddr; + const __u32 daddr = skb->nh.iph->daddr; + struct dst_entry *dst = NULL; + + /* Never answer to DCCP_PKT_REQUESTs send to broadcast or multicast */ + if (((struct rtable *)skb->dst)->rt_flags & + (RTCF_BROADCAST | RTCF_MULTICAST)) + 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(sk->sk_prot->rsk_prot); + if (req == NULL) + goto drop; + + /* FIXME: process options */ + + dccp_openreq_init(req, &dp, skb); + + ireq = inet_rsk(req); + ireq->loc_addr = daddr; + ireq->rmt_addr = saddr; + /* FIXME: Merge Aristeu's option parsing code when ready */ + req->rcv_wnd = 100; /* Fake, option parsing will get the + right value */ + 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 = DCCP_SKB_CB(skb)->dccpd_seq; + dreq->dreq_iss = dccp_v4_init_sequence(sk, skb); + dreq->dreq_service = dccp_hdr_request(skb)->dccph_req_service; + + if (dccp_v4_send_response(sk, req, dst)) + goto drop_and_free; + + inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT); + return 0; + +drop_and_free: + /* + * FIXME: should be reqsk_free after implementing req->rsk_ops + */ + __reqsk_free(req); +drop: + DCCP_INC_STATS_BH(DCCP_MIB_ATTEMPTFAILS); + return -1; +} + +/* + * The three way handshake has completed - we got a valid ACK or DATAACK - + * now create the new socket. + * + * This is the equivalent of TCP's tcp_v4_syn_recv_sock + */ +struct sock *dccp_v4_request_recv_sock(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct dst_entry *dst) +{ + struct inet_request_sock *ireq; + struct inet_sock *newinet; + struct dccp_sock *newdp; + struct sock *newsk; + + if (sk_acceptq_is_full(sk)) + goto exit_overflow; + + if (dst == NULL && (dst = inet_csk_route_req(sk, req)) == NULL) + goto exit; + + newsk = dccp_create_openreq_child(sk, req, skb); + if (newsk == NULL) + goto exit; + + sk_setup_caps(newsk, dst); + + newdp = dccp_sk(newsk); + newinet = inet_sk(newsk); + ireq = inet_rsk(req); + newinet->daddr = ireq->rmt_addr; + newinet->rcv_saddr = ireq->loc_addr; + newinet->saddr = ireq->loc_addr; + newinet->opt = ireq->opt; + ireq->opt = NULL; + newinet->mc_index = inet_iif(skb); + newinet->mc_ttl = skb->nh.iph->ttl; + newinet->id = jiffies; + + dccp_sync_mss(newsk, dst_mtu(dst)); + + __inet_hash(&dccp_hashinfo, newsk, 0); + __inet_inherit_port(&dccp_hashinfo, sk, newsk); + + return newsk; + +exit_overflow: + NET_INC_STATS_BH(LINUX_MIB_LISTENOVERFLOWS); +exit: + NET_INC_STATS_BH(LINUX_MIB_LISTENDROPS); + dst_release(dst); + return NULL; +} + +static struct sock *dccp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) +{ + const struct dccp_hdr *dh = dccp_hdr(skb); + const struct iphdr *iph = skb->nh.iph; + struct sock *nsk; + struct request_sock **prev; + /* Find possible connection requests. */ + struct request_sock *req = inet_csk_search_req(sk, &prev, + dh->dccph_sport, + iph->saddr, iph->daddr); + if (req != NULL) + return dccp_check_req(sk, skb, req, prev); + + nsk = __inet_lookup_established(&dccp_hashinfo, + iph->saddr, dh->dccph_sport, + iph->daddr, ntohs(dh->dccph_dport), + inet_iif(skb)); + if (nsk != NULL) { + if (nsk->sk_state != DCCP_TIME_WAIT) { + bh_lock_sock(nsk); + return nsk; + } + inet_twsk_put((struct inet_timewait_sock *)nsk); + return NULL; + } + + return sk; +} + +int dccp_v4_checksum(const struct sk_buff *skb, const u32 saddr, + const u32 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); +} + +static int dccp_v4_verify_checksum(struct sk_buff *skb, + const u32 saddr, const u32 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) +{ + struct rtable *rt; + struct flowi fl = { .oif = ((struct rtable *)skb->dst)->rt_iif, + .nl_u = { .ip4_u = + { .daddr = skb->nh.iph->saddr, + .saddr = skb->nh.iph->daddr, + .tos = RT_CONN_FLAGS(sk) } }, + .proto = sk->sk_protocol, + .uli_u = { .ports = + { .sport = dccp_hdr(skb)->dccph_dport, + .dport = dccp_hdr(skb)->dccph_sport } + } + }; + + if (ip_route_output_flow(&rt, &fl, sk, 0)) { + IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES); + return NULL; + } + + return &rt->u.dst; +} + +static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb) +{ + int err; + struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh; + const int dccp_hdr_reset_len = sizeof(struct dccp_hdr) + + sizeof(struct dccp_hdr_ext) + + sizeof(struct dccp_hdr_reset); + struct sk_buff *skb; + struct dst_entry *dst; + u64 seqno; + + /* Never send a reset in response to a reset. */ + if (rxdh->dccph_type == DCCP_PKT_RESET) + return; + + if (((struct rtable *)rxskb->dst)->rt_type != RTN_LOCAL) + return; + + dst = dccp_v4_route_skb(dccp_ctl_socket->sk, rxskb); + if (dst == NULL) + return; + + skb = alloc_skb(MAX_DCCP_HEADER + 15, GFP_ATOMIC); + if (skb == NULL) + goto out; + + /* Reserve space for headers. */ + skb_reserve(skb, MAX_DCCP_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); + + /* Build DCCP header and checksum it. */ + dh->dccph_type = DCCP_PKT_RESET; + dh->dccph_sport = rxdh->dccph_dport; + dh->dccph_dport = rxdh->dccph_sport; + dh->dccph_doff = dccp_hdr_reset_len / 4; + dh->dccph_x = 1; + dccp_hdr_reset(skb)->dccph_reset_code = + DCCP_SKB_CB(rxskb)->dccpd_reset_code; + + /* See "8.3.1. Abnormal Termination" in draft-ietf-dccp-spec-11 */ + 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); + + dh->dccph_checksum = dccp_v4_checksum(skb, rxskb->nh.iph->saddr, + rxskb->nh.iph->daddr); + + bh_lock_sock(dccp_ctl_socket->sk); + err = ip_build_and_send_pkt(skb, dccp_ctl_socket->sk, + rxskb->nh.iph->daddr, + rxskb->nh.iph->saddr, NULL); + bh_unlock_sock(dccp_ctl_socket->sk); + + if (err == NET_XMIT_CN || err == 0) { + DCCP_INC_STATS_BH(DCCP_MIB_OUTSEGS); + DCCP_INC_STATS_BH(DCCP_MIB_OUTRSTS); + } +out: + dst_release(dst); +} + +int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) +{ + struct dccp_hdr *dh = dccp_hdr(skb); + + if (sk->sk_state == DCCP_OPEN) { /* Fast path */ + if (dccp_rcv_established(sk, skb, dh, skb->len)) + goto reset; + 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 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 (sk->sk_state == DCCP_LISTEN) { + struct sock *nsk = dccp_v4_hnd_req(sk, skb); + + if (nsk == NULL) + goto discard; + + if (nsk != sk) { + if (dccp_child_process(sk, nsk, skb)) + goto reset; + return 0; + } + } + + if (dccp_rcv_state_process(sk, skb, dh, skb->len)) + goto reset; + return 0; + +reset: + DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_NO_CONNECTION; + dccp_v4_ctl_send_reset(skb); +discard: + kfree_skb(skb); + return 0; +} + +static inline int dccp_invalid_packet(struct sk_buff *skb) +{ + const struct dccp_hdr *dh; + + if (skb->pkt_type != PACKET_HOST) + return 1; + + if (!pskb_may_pull(skb, sizeof(struct dccp_hdr))) { + LIMIT_NETDEBUG(KERN_WARNING "DCCP: pskb_may_pull failed\n"); + return 1; + } + + dh = dccp_hdr(skb); + + /* If the packet type is not understood, drop packet and return */ + if (dh->dccph_type >= DCCP_PKT_INVALID) { + LIMIT_NETDEBUG(KERN_WARNING "DCCP: 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 (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); + return 1; + } + + 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); + 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)); + return 1; + } + + /* 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 "DCCP: header checksum is " + "incorrect\n"); + return 1; + } + + return 0; +} + +/* this is called when real data arrives */ +int dccp_v4_rcv(struct sk_buff *skb) +{ + const struct dccp_hdr *dh; + struct sock *sk; + int rc; + + /* Step 1: Check header basics: */ + + if (dccp_invalid_packet(skb)) + goto discard_it; + + dh = dccp_hdr(skb); +#if 0 + /* + * Use something like this to simulate some DATA/DATAACK loss to test + * dccp_ackpkts_add, you'll get something like this on a session that + * sends 10 DATA/DATAACK packets: + * + * ackpkts_print: 281473596467422 |0,0|3,0|0,0|3,0|0,0|3,0|0,0|3,0|0,1| + * + * 0, 0 means: DCCP_ACKPKTS_STATE_RECEIVED, RLE == just this packet + * 0, 1 means: DCCP_ACKPKTS_STATE_RECEIVED, RLE == two adjacent packets + * with the same state + * 3, 0 means: DCCP_ACKPKTS_STATE_NOT_RECEIVED, RLE == just this packet + * + * So... + * + * 281473596467422 was received + * 281473596467421 was not received + * 281473596467420 was received + * 281473596467419 was not received + * 281473596467418 was received + * 281473596467417 was not received + * 281473596467416 was received + * 281473596467415 was not received + * 281473596467414 was received + * 281473596467413 was received (this one was the 3way handshake + * RESPONSE) + * + */ + if (dh->dccph_type == DCCP_PKT_DATA || + dh->dccph_type == DCCP_PKT_DATAACK) { + static int discard = 0; + + if (discard) { + discard = 0; + goto discard_it; + } + discard = 1; + } +#endif + DCCP_SKB_CB(skb)->dccpd_seq = dccp_hdr_seq(skb); + DCCP_SKB_CB(skb)->dccpd_type = dh->dccph_type; + + dccp_pr_debug("%8.8s " + "src=%u.%u.%u.%u@%-5d " + "dst=%u.%u.%u.%u@%-5d seq=%llu", + dccp_packet_name(dh->dccph_type), + NIPQUAD(skb->nh.iph->saddr), ntohs(dh->dccph_sport), + NIPQUAD(skb->nh.iph->daddr), ntohs(dh->dccph_dport), + (unsigned long long) DCCP_SKB_CB(skb)->dccpd_seq); + + if (dccp_packet_without_ack(skb)) { + DCCP_SKB_CB(skb)->dccpd_ack_seq = DCCP_PKT_WITHOUT_ACK_SEQ; + 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_SKB_CB(skb)->dccpd_ack_seq); + } + + /* Step 2: + * 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, ntohs(dh->dccph_dport), + inet_iif(skb)); + + /* + * Step 2: + * If no socket ... + * Generate Reset(No Connection) unless P.type == Reset + * Drop packet and return + */ + 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, + * 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; + } + + if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb)) { + dccp_pr_debug("xfrm4_policy_check failed\n"); + goto discard_and_relse; + } + + if (sk_filter(sk, skb, 0)) { + dccp_pr_debug("sk_filter failed\n"); + goto discard_and_relse; + } + + skb->dev = NULL; + + bh_lock_sock(sk); + rc = 0; + if (!sock_owned_by_user(sk)) + rc = dccp_v4_do_rcv(sk, skb); + else + sk_add_backlog(sk, skb); + bh_unlock_sock(sk); + + sock_put(sk); + return rc; + +no_dccp_socket: + if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) + goto discard_it; + /* + * Step 2: + * 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); + } + +discard_it: + /* Discard frame. */ + kfree_skb(skb); + return 0; + +discard_and_relse: + sock_put(sk); + goto discard_it; + +do_time_wait: + inet_twsk_put((struct inet_timewait_sock *)sk); + goto no_dccp_socket; +} + +static int dccp_v4_init_sock(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + static int dccp_ctl_socket_init = 1; + + dccp_options_init(&dp->dccps_options); + + if (dp->dccps_options.dccpo_send_ack_vector) { + dp->dccps_hc_rx_ackpkts = + dccp_ackpkts_alloc(DCCP_MAX_ACK_VECTOR_LEN, + GFP_KERNEL); + + if (dp->dccps_hc_rx_ackpkts == NULL) + return -ENOMEM; + } + + /* + * FIXME: We're hardcoding the CCID, and doing this at this point makes + * the listening (master) sock get CCID control blocks, which is not + * necessary, but for now, to not mess with the test userspace apps, + * lets leave it here, later the real solution is to do this in a + * setsockopt(CCIDs-I-want/accept). -acme + */ + if (likely(!dccp_ctl_socket_init)) { + dp->dccps_hc_rx_ccid = ccid_init(dp->dccps_options.dccpo_ccid, + sk); + dp->dccps_hc_tx_ccid = ccid_init(dp->dccps_options.dccpo_ccid, + sk); + if (dp->dccps_hc_rx_ccid == NULL || + dp->dccps_hc_tx_ccid == NULL) { + ccid_exit(dp->dccps_hc_rx_ccid, sk); + ccid_exit(dp->dccps_hc_tx_ccid, sk); + dccp_ackpkts_free(dp->dccps_hc_rx_ackpkts); + dp->dccps_hc_rx_ackpkts = NULL; + dp->dccps_hc_rx_ccid = dp->dccps_hc_tx_ccid = NULL; + return -ENOMEM; + } + } else + dccp_ctl_socket_init = 0; + + dccp_init_xmit_timers(sk); + inet_csk(sk)->icsk_rto = DCCP_TIMEOUT_INIT; + sk->sk_state = DCCP_CLOSED; + sk->sk_write_space = dccp_write_space; + dp->dccps_mss_cache = 536; + dp->dccps_role = DCCP_ROLE_UNDEFINED; + + return 0; +} + +static int dccp_v4_destroy_sock(struct sock *sk) +{ + struct dccp_sock *dp = dccp_sk(sk); + + /* + * DCCP doesn't use sk_qrite_queue, just sk_send_head + * for retransmissions + */ + if (sk->sk_send_head != NULL) { + kfree_skb(sk->sk_send_head); + sk->sk_send_head = NULL; + } + + /* Clean up a referenced DCCP bind bucket. */ + if (inet_csk(sk)->icsk_bind_hash != NULL) + inet_put_port(&dccp_hashinfo, sk); + + ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk); + ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk); + dccp_ackpkts_free(dp->dccps_hc_rx_ackpkts); + dp->dccps_hc_rx_ackpkts = NULL; + ccid_exit(dp->dccps_hc_rx_ccid, sk); + ccid_exit(dp->dccps_hc_tx_ccid, sk); + dp->dccps_hc_rx_ccid = dp->dccps_hc_tx_ccid = NULL; + + return 0; +} + +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, +}; + +struct proto dccp_v4_prot = { + .name = "DCCP", + .owner = THIS_MODULE, + .close = dccp_close, + .connect = dccp_v4_connect, + .disconnect = dccp_disconnect, + .ioctl = dccp_ioctl, + .init = dccp_v4_init_sock, + .setsockopt = dccp_setsockopt, + .getsockopt = dccp_getsockopt, + .sendmsg = dccp_sendmsg, + .recvmsg = dccp_recvmsg, + .backlog_rcv = dccp_v4_do_rcv, + .hash = dccp_v4_hash, + .unhash = dccp_v4_unhash, + .accept = inet_csk_accept, + .get_port = dccp_v4_get_port, + .shutdown = dccp_shutdown, + .destroy = dccp_v4_destroy_sock, + .orphan_count = &dccp_orphan_count, + .max_header = MAX_DCCP_HEADER, + .obj_size = sizeof(struct dccp_sock), + .rsk_prot = &dccp_request_sock_ops, + .twsk_obj_size = sizeof(struct inet_timewait_sock), +}; diff --git a/net/dccp/minisocks.c b/net/dccp/minisocks.c new file mode 100644 index 000000000000..ce5dff4ac22e --- /dev/null +++ b/net/dccp/minisocks.c @@ -0,0 +1,264 @@ +/* + * net/dccp/minisocks.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/skbuff.h> +#include <linux/timer.h> + +#include <net/sock.h> +#include <net/xfrm.h> +#include <net/inet_timewait_sock.h> + +#include "ccid.h" +#include "dccp.h" + +struct inet_timewait_death_row dccp_death_row = { + .sysctl_max_tw_buckets = NR_FILE * 2, + .period = DCCP_TIMEWAIT_LEN / INET_TWDR_TWKILL_SLOTS, + .death_lock = SPIN_LOCK_UNLOCKED, + .hashinfo = &dccp_hashinfo, + .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), +/* Short-time timewait calendar */ + + .twcal_hand = -1, + .twcal_timer = TIMER_INITIALIZER(inet_twdr_twcal_tick, 0, + (unsigned long)&dccp_death_row), +}; + +void dccp_time_wait(struct sock *sk, int state, int timeo) +{ + struct inet_timewait_sock *tw = NULL; + + if (dccp_death_row.tw_count < dccp_death_row.sysctl_max_tw_buckets) + tw = inet_twsk_alloc(sk, state); + + if (tw != NULL) { + const struct inet_connection_sock *icsk = inet_csk(sk); + const int rto = (icsk->icsk_rto << 2) - (icsk->icsk_rto >> 1); + + /* Linkage updates. */ + __inet_twsk_hashdance(tw, sk, &dccp_hashinfo); + + /* Get the TIME_WAIT timeout firing. */ + if (timeo < rto) + timeo = rto; + + tw->tw_timeout = DCCP_TIMEWAIT_LEN; + if (state == DCCP_TIME_WAIT) + timeo = DCCP_TIMEWAIT_LEN; + + inet_twsk_schedule(tw, &dccp_death_row, timeo, + DCCP_TIMEWAIT_LEN); + inet_twsk_put(tw); + } else { + /* Sorry, if we're out of memory, just CLOSE this + * socket up. We've got bigger problems than + * non-graceful socket closings. + */ + LIMIT_NETDEBUG(KERN_INFO "DCCP: time wait bucket " + "table overflow\n"); + } + + dccp_done(sk); +} + +struct sock *dccp_create_openreq_child(struct sock *sk, + const struct request_sock *req, + const struct sk_buff *skb) +{ + /* + * Step 3: Process LISTEN state + * + * // 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); + + if (newsk != NULL) { + const struct dccp_request_sock *dreq = dccp_rsk(req); + struct inet_connection_sock *newicsk = inet_csk(sk); + struct dccp_sock *newdp = dccp_sk(newsk); + + newdp->dccps_hc_rx_ackpkts = NULL; + newdp->dccps_role = DCCP_ROLE_SERVER; + newicsk->icsk_rto = DCCP_TIMEOUT_INIT; + + if (newdp->dccps_options.dccpo_send_ack_vector) { + newdp->dccps_hc_rx_ackpkts = + dccp_ackpkts_alloc(DCCP_MAX_ACK_VECTOR_LEN, + GFP_ATOMIC); + /* + * XXX: We're using the same CCIDs set on the parent, + * i.e. sk_clone copied the master sock and left the + * CCID pointers for this child, that is why we do the + * __ccid_get calls. + */ + if (unlikely(newdp->dccps_hc_rx_ackpkts == NULL)) + goto out_free; + } + + if (unlikely(ccid_hc_rx_init(newdp->dccps_hc_rx_ccid, + newsk) != 0 || + ccid_hc_tx_init(newdp->dccps_hc_tx_ccid, + newsk) != 0)) { + dccp_ackpkts_free(newdp->dccps_hc_rx_ackpkts); + ccid_hc_rx_exit(newdp->dccps_hc_rx_ccid, newsk); + ccid_hc_tx_exit(newdp->dccps_hc_tx_ccid, newsk); +out_free: + /* It is still raw copy of parent, so invalidate + * destructor and make plain sk_free() */ + newsk->sk_destruct = NULL; + sk_free(newsk); + return NULL; + } + + __ccid_get(newdp->dccps_hc_rx_ccid); + __ccid_get(newdp->dccps_hc_tx_ccid); + + /* + * 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 + */ + + /* See dccp_v4_conn_request */ + newdp->dccps_options.dccpo_sequence_window = req->rcv_wnd; + + newdp->dccps_gar = newdp->dccps_isr = dreq->dreq_isr; + dccp_update_gsr(newsk, dreq->dreq_isr); + + newdp->dccps_iss = dreq->dreq_iss; + dccp_update_gss(newsk, dreq->dreq_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_set_seqno(&newdp->dccps_swl, + max48(newdp->dccps_swl, newdp->dccps_isr)); + dccp_set_seqno(&newdp->dccps_awl, + max48(newdp->dccps_awl, newdp->dccps_iss)); + + dccp_init_xmit_timers(newsk); + + DCCP_INC_STATS_BH(DCCP_MIB_PASSIVEOPENS); + } + return newsk; +} + +/* + * Process an incoming packet for RESPOND sockets represented + * as an request_sock. + */ +struct sock *dccp_check_req(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct request_sock **prev) +{ + struct sock *child = NULL; + + /* 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); + + 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); + req->rsk_ops->rtx_syn_ack(sk, req, NULL); + } + /* Network Duplicate, discard packet */ + return NULL; + } + + DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_PACKET_ERROR; + + if (dccp_hdr(skb)->dccph_type != DCCP_PKT_ACK && + dccp_hdr(skb)->dccph_type != DCCP_PKT_DATAACK) + goto drop; + + /* Invalid ACK */ + if (DCCP_SKB_CB(skb)->dccpd_ack_seq != dccp_rsk(req)->dreq_iss) { + dccp_pr_debug("Invalid ACK number: ack_seq=%llu, " + "dreq_iss=%llu\n", + (unsigned long long) + DCCP_SKB_CB(skb)->dccpd_ack_seq, + (unsigned long long) + dccp_rsk(req)->dreq_iss); + goto drop; + } + + child = dccp_v4_request_recv_sock(sk, skb, req, NULL); + if (child == NULL) + goto listen_overflow; + + /* FIXME: deal with options */ + + inet_csk_reqsk_queue_unlink(sk, req, prev); + inet_csk_reqsk_queue_removed(sk, req); + inet_csk_reqsk_queue_add(sk, req, child); +out: + return child; +listen_overflow: + dccp_pr_debug("listen_overflow!\n"); + 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); + + inet_csk_reqsk_queue_drop(sk, req, prev); + goto out; +} + +/* + * Queue segment on the new socket if the new socket is active, + * otherwise we just shortcircuit this and continue with + * the new socket. + */ +int dccp_child_process(struct sock *parent, struct sock *child, + struct sk_buff *skb) +{ + int ret = 0; + const int state = child->sk_state; + + if (!sock_owned_by_user(child)) { + ret = dccp_rcv_state_process(child, skb, dccp_hdr(skb), + skb->len); + + /* Wakeup parent, send SIGIO */ + if (state == DCCP_RESPOND && child->sk_state != state) + parent->sk_data_ready(parent, 0); + } else { + /* Alas, it is possible again, because we do lookup + * in main socket hash table and lock on listening + * socket does not protect us more. + */ + sk_add_backlog(child, skb); + } + + bh_unlock_sock(child); + sock_put(child); + return ret; +} diff --git a/net/dccp/options.c b/net/dccp/options.c new file mode 100644 index 000000000000..382c5894acb2 --- /dev/null +++ b/net/dccp/options.c @@ -0,0 +1,855 @@ +/* + * net/dccp/options.c + * + * An implementation of the DCCP protocol + * Copyright (c) 2005 Aristeu Sergio Rozanski Filho <aris@cathedrallabs.org> + * Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@ghostprotocols.net> + * Copyright (c) 2005 Ian McDonald <iam4@cs.waikato.ac.nz> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/module.h> +#include <linux/types.h> +#include <linux/kernel.h> +#include <linux/skbuff.h> + +#include "ccid.h" +#include "dccp.h" + +static void dccp_ackpkts_check_rcv_ackvector(struct dccp_ackpkts *ap, + struct sock *sk, + const u64 ackno, + const unsigned char len, + const unsigned char *vector); + +/* stores the default values for new connection. may be changed with sysctl */ +static const struct dccp_options dccpo_default_values = { + .dccpo_sequence_window = DCCPF_INITIAL_SEQUENCE_WINDOW, + .dccpo_ccid = DCCPF_INITIAL_CCID, + .dccpo_send_ack_vector = DCCPF_INITIAL_SEND_ACK_VECTOR, + .dccpo_send_ndp_count = DCCPF_INITIAL_SEND_NDP_COUNT, +}; + +void dccp_options_init(struct dccp_options *dccpo) +{ + memcpy(dccpo, &dccpo_default_values, sizeof(*dccpo)); +} + +static u32 dccp_decode_value_var(const unsigned char *bf, const u8 len) +{ + u32 value = 0; + + if (len > 3) + value += *bf++ << 24; + if (len > 2) + value += *bf++ << 16; + if (len > 1) + value += *bf++ << 8; + if (len > 0) + value += *bf; + + return value; +} + +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; + unsigned char *options = (unsigned char *)dh + dccp_hdr_len(skb); + unsigned char *opt_ptr = options; + const unsigned char *opt_end = (unsigned char *)dh + + (dh->dccph_doff * 4); + struct dccp_options_received *opt_recv = &dp->dccps_options_received; + unsigned char opt, len; + unsigned char *value; + + memset(opt_recv, 0, sizeof(*opt_recv)); + + while (opt_ptr != opt_end) { + opt = *opt_ptr++; + len = 0; + value = NULL; + + /* Check if this isn't a single byte option */ + if (opt > DCCPO_MAX_RESERVED) { + if (opt_ptr == opt_end) + goto out_invalid_option; + + len = *opt_ptr++; + if (len < 3) + goto out_invalid_option; + /* + * Remove the type and len fields, leaving + * just the value size + */ + len -= 2; + value = opt_ptr; + opt_ptr += len; + + if (opt_ptr > opt_end) + goto out_invalid_option; + } + + switch (opt) { + case DCCPO_PADDING: + break; + case DCCPO_NDP_COUNT: + if (len > 3) + goto out_invalid_option; + + opt_recv->dccpor_ndp = dccp_decode_value_var(value, len); + dccp_pr_debug("%sNDP count=%d\n", debug_prefix, + opt_recv->dccpor_ndp); + break; + case DCCPO_ACK_VECTOR_0: + if (len > DCCP_MAX_ACK_VECTOR_LEN) + goto out_invalid_option; + + if (pkt_type == DCCP_PKT_DATA) + continue; + + opt_recv->dccpor_ack_vector_len = len; + opt_recv->dccpor_ack_vector_idx = value - options; + + dccp_pr_debug("%sACK vector 0, len=%d, ack_ackno=%llu\n", + debug_prefix, len, + (unsigned long long) + DCCP_SKB_CB(skb)->dccpd_ack_seq); + dccp_ackvector_print(DCCP_SKB_CB(skb)->dccpd_ack_seq, + value, len); + dccp_ackpkts_check_rcv_ackvector(dp->dccps_hc_rx_ackpkts, + sk, + DCCP_SKB_CB(skb)->dccpd_ack_seq, + len, value); + break; + case DCCPO_TIMESTAMP: + if (len != 4) + goto out_invalid_option; + + opt_recv->dccpor_timestamp = ntohl(*(u32 *)value); + + dp->dccps_timestamp_echo = opt_recv->dccpor_timestamp; + do_gettimeofday(&dp->dccps_timestamp_time); + + dccp_pr_debug("%sTIMESTAMP=%u, ackno=%llu\n", + debug_prefix, opt_recv->dccpor_timestamp, + (unsigned long long) + DCCP_SKB_CB(skb)->dccpd_ack_seq); + break; + case DCCPO_TIMESTAMP_ECHO: + if (len != 4 && len != 6 && len != 8) + goto out_invalid_option; + + opt_recv->dccpor_timestamp_echo = ntohl(*(u32 *)value); + + dccp_pr_debug("%sTIMESTAMP_ECHO=%u, len=%d, ackno=%llu, ", + debug_prefix, + opt_recv->dccpor_timestamp_echo, + len + 2, + (unsigned long long) + DCCP_SKB_CB(skb)->dccpd_ack_seq); + + if (len > 4) { + if (len == 6) + opt_recv->dccpor_elapsed_time = + ntohs(*(u16 *)(value + 4)); + else + opt_recv->dccpor_elapsed_time = + ntohl(*(u32 *)(value + 4)); + + dccp_pr_debug("%sTIMESTAMP_ECHO ELAPSED_TIME=%d\n", + debug_prefix, + opt_recv->dccpor_elapsed_time); + } + break; + case DCCPO_ELAPSED_TIME: + if (len != 2 && len != 4) + goto out_invalid_option; + + if (pkt_type == DCCP_PKT_DATA) + continue; + + if (len == 2) + opt_recv->dccpor_elapsed_time = + ntohs(*(u16 *)value); + else + opt_recv->dccpor_elapsed_time = + ntohl(*(u32 *)value); + + dccp_pr_debug("%sELAPSED_TIME=%d\n", debug_prefix, + opt_recv->dccpor_elapsed_time); + break; + /* + * From draft-ietf-dccp-spec-11.txt: + * + * Option numbers 128 through 191 are for + * options sent from the HC-Sender to the + * HC-Receiver; option numbers 192 through 255 + * are for options sent from the HC-Receiver to + * the HC-Sender. + */ + case 128 ... 191: { + const u16 idx = value - options; + + if (ccid_hc_rx_parse_options(dp->dccps_hc_rx_ccid, sk, + opt, len, idx, + value) != 0) + goto out_invalid_option; + } + break; + case 192 ... 255: { + const u16 idx = value - options; + + if (ccid_hc_tx_parse_options(dp->dccps_hc_tx_ccid, sk, + opt, len, idx, + value) != 0) + goto out_invalid_option; + } + break; + default: + pr_info("DCCP(%p): option %d(len=%d) not " + "implemented, ignoring\n", + sk, opt, len); + break; + } + } + + return 0; + +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); + return -1; +} + +static void dccp_encode_value_var(const u32 value, unsigned char *to, + const unsigned int len) +{ + if (len > 3) + *to++ = (value & 0xFF000000) >> 24; + if (len > 2) + *to++ = (value & 0xFF0000) >> 16; + if (len > 1) + *to++ = (value & 0xFF00) >> 8; + if (len > 0) + *to++ = (value & 0xFF); +} + +static inline int dccp_ndp_len(const int ndp) +{ + return likely(ndp <= 0xFF) ? 1 : ndp <= 0xFFFF ? 2 : 3; +} + +void dccp_insert_option(struct sock *sk, struct sk_buff *skb, + const unsigned char option, + const void *value, const unsigned char len) +{ + unsigned char *to; + + if (DCCP_SKB_CB(skb)->dccpd_opt_len + len + 2 > DCCP_MAX_OPT_LEN) { + LIMIT_NETDEBUG(KERN_INFO "DCCP: packet too small to insert " + "%d option!\n", option); + return; + } + + DCCP_SKB_CB(skb)->dccpd_opt_len += len + 2; + + to = skb_push(skb, len + 2); + *to++ = option; + *to++ = len + 2; + + memcpy(to, value, len); +} + +EXPORT_SYMBOL_GPL(dccp_insert_option); + +static void dccp_insert_option_ndp(struct sock *sk, struct sk_buff *skb) +{ + struct dccp_sock *dp = dccp_sk(sk); + int ndp = dp->dccps_ndp_count; + + if (dccp_non_data_packet(skb)) + ++dp->dccps_ndp_count; + else + dp->dccps_ndp_count = 0; + + if (ndp > 0) { + unsigned char *ptr; + const int ndp_len = dccp_ndp_len(ndp); + const int len = ndp_len + 2; + + if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) + return; + + DCCP_SKB_CB(skb)->dccpd_opt_len += len; + + ptr = skb_push(skb, len); + *ptr++ = DCCPO_NDP_COUNT; + *ptr++ = len; + dccp_encode_value_var(ndp, ptr, ndp_len); + } +} + +static inline int dccp_elapsed_time_len(const u32 elapsed_time) +{ + return elapsed_time == 0 ? 0 : elapsed_time <= 0xFFFF ? 2 : 4; +} + +void dccp_insert_option_elapsed_time(struct sock *sk, + struct sk_buff *skb, + u32 elapsed_time) +{ +#ifdef CONFIG_IP_DCCP_DEBUG + struct dccp_sock *dp = dccp_sk(sk); + const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? + "CLIENT TX opt: " : "server TX opt: "; +#endif + const int elapsed_time_len = dccp_elapsed_time_len(elapsed_time); + const int len = 2 + elapsed_time_len; + unsigned char *to; + + if (elapsed_time_len == 0) + return; + + if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) { + LIMIT_NETDEBUG(KERN_INFO "DCCP: packet too small to " + "insert elapsed time!\n"); + return; + } + + DCCP_SKB_CB(skb)->dccpd_opt_len += len; + + to = skb_push(skb, len); + *to++ = DCCPO_ELAPSED_TIME; + *to++ = len; + + if (elapsed_time_len == 2) { + const u16 var16 = htons((u16)elapsed_time); + memcpy(to, &var16, 2); + } else { + const u32 var32 = htonl(elapsed_time); + memcpy(to, &var32, 4); + } + + dccp_pr_debug("%sELAPSED_TIME=%u, len=%d, seqno=%llu\n", + debug_prefix, elapsed_time, + len, + (unsigned long long) DCCP_SKB_CB(skb)->dccpd_seq); +} + +EXPORT_SYMBOL_GPL(dccp_insert_option_elapsed_time); + +static void dccp_insert_option_ack_vector(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 opt: " : "server TX opt: "; +#endif + struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts; + int len = ap->dccpap_buf_vector_len + 2; + const u32 elapsed_time = timeval_now_delta(&ap->dccpap_time) / 10; + unsigned char *to, *from; + + if (elapsed_time != 0) + dccp_insert_option_elapsed_time(sk, skb, elapsed_time); + + if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) { + LIMIT_NETDEBUG(KERN_INFO "DCCP: packet too small to " + "insert ACK Vector!\n"); + return; + } + + /* + * XXX: now we have just one ack vector sent record, so + * we have to wait for it to be cleared. + * + * Of course this is not acceptable, but this is just for + * basic testing now. + */ + if (ap->dccpap_ack_seqno != DCCP_MAX_SEQNO + 1) + return; + + DCCP_SKB_CB(skb)->dccpd_opt_len += len; + + to = skb_push(skb, len); + *to++ = DCCPO_ACK_VECTOR_0; + *to++ = len; + + len = ap->dccpap_buf_vector_len; + from = ap->dccpap_buf + ap->dccpap_buf_head; + + /* Check if buf_head wraps */ + if (ap->dccpap_buf_head + len > ap->dccpap_buf_len) { + const unsigned int tailsize = (ap->dccpap_buf_len - + ap->dccpap_buf_head); + + memcpy(to, from, tailsize); + to += tailsize; + len -= tailsize; + from = ap->dccpap_buf; + } + + memcpy(to, from, len); + /* + * From draft-ietf-dccp-spec-11.txt: + * + * For each acknowledgement it sends, the HC-Receiver will add an + * acknowledgement record. ack_seqno will equal the HC-Receiver + * sequence number it used for the ack packet; ack_ptr will equal + * buf_head; ack_ackno will equal buf_ackno; and ack_nonce will + * equal buf_nonce. + * + * This implemention uses just one ack record for now. + */ + ap->dccpap_ack_seqno = DCCP_SKB_CB(skb)->dccpd_seq; + ap->dccpap_ack_ptr = ap->dccpap_buf_head; + ap->dccpap_ack_ackno = ap->dccpap_buf_ackno; + ap->dccpap_ack_nonce = ap->dccpap_buf_nonce; + ap->dccpap_ack_vector_len = ap->dccpap_buf_vector_len; + + dccp_pr_debug("%sACK Vector 0, len=%d, ack_seqno=%llu, " + "ack_ackno=%llu\n", + debug_prefix, ap->dccpap_ack_vector_len, + (unsigned long long) ap->dccpap_ack_seqno, + (unsigned long long) ap->dccpap_ack_ackno); +} + +void dccp_insert_option_timestamp(struct sock *sk, struct sk_buff *skb) +{ + struct timeval tv; + u32 now; + + do_gettimeofday(&tv); + now = (tv.tv_sec * USEC_PER_SEC + tv.tv_usec) / 10; + /* yes this will overflow but that is the point as we want a + * 10 usec 32 bit timer which mean it wraps every 11.9 hours */ + + now = htonl(now); + dccp_insert_option(sk, skb, DCCPO_TIMESTAMP, &now, sizeof(now)); +} + +EXPORT_SYMBOL_GPL(dccp_insert_option_timestamp); + +static void dccp_insert_option_timestamp_echo(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 opt: " : "server TX opt: "; +#endif + u32 tstamp_echo; + const u32 elapsed_time = + timeval_now_delta(&dp->dccps_timestamp_time) / 10; + const int elapsed_time_len = dccp_elapsed_time_len(elapsed_time); + const int len = 6 + elapsed_time_len; + unsigned char *to; + + if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) { + LIMIT_NETDEBUG(KERN_INFO "DCCP: packet too small to insert " + "timestamp echo!\n"); + return; + } + + DCCP_SKB_CB(skb)->dccpd_opt_len += len; + + to = skb_push(skb, len); + *to++ = DCCPO_TIMESTAMP_ECHO; + *to++ = len; + + tstamp_echo = htonl(dp->dccps_timestamp_echo); + memcpy(to, &tstamp_echo, 4); + to += 4; + + if (elapsed_time_len == 2) { + const u16 var16 = htons((u16)elapsed_time); + memcpy(to, &var16, 2); + } else if (elapsed_time_len == 4) { + const u32 var32 = htonl(elapsed_time); + memcpy(to, &var32, 4); + } + + dccp_pr_debug("%sTIMESTAMP_ECHO=%u, len=%d, seqno=%llu\n", + debug_prefix, dp->dccps_timestamp_echo, + len, + (unsigned long long) DCCP_SKB_CB(skb)->dccpd_seq); + + dp->dccps_timestamp_echo = 0; + dp->dccps_timestamp_time.tv_sec = 0; + dp->dccps_timestamp_time.tv_usec = 0; +} + +void dccp_insert_options(struct sock *sk, struct sk_buff *skb) +{ + struct dccp_sock *dp = dccp_sk(sk); + + DCCP_SKB_CB(skb)->dccpd_opt_len = 0; + + if (dp->dccps_options.dccpo_send_ndp_count) + dccp_insert_option_ndp(sk, skb); + + if (!dccp_packet_without_ack(skb)) { + if (dp->dccps_options.dccpo_send_ack_vector && + (dp->dccps_hc_rx_ackpkts->dccpap_buf_ackno != + DCCP_MAX_SEQNO + 1)) + dccp_insert_option_ack_vector(sk, skb); + + if (dp->dccps_timestamp_echo != 0) + dccp_insert_option_timestamp_echo(sk, skb); + } + + ccid_hc_rx_insert_options(dp->dccps_hc_rx_ccid, sk, skb); + ccid_hc_tx_insert_options(dp->dccps_hc_tx_ccid, sk, skb); + + /* XXX: insert other options when appropriate */ + + if (DCCP_SKB_CB(skb)->dccpd_opt_len != 0) { + /* The length of all options has to be a multiple of 4 */ + int padding = DCCP_SKB_CB(skb)->dccpd_opt_len % 4; + + if (padding != 0) { + padding = 4 - padding; + memset(skb_push(skb, padding), 0, padding); + DCCP_SKB_CB(skb)->dccpd_opt_len += padding; + } + } +} + +struct dccp_ackpkts *dccp_ackpkts_alloc(const unsigned int len, + const unsigned int __nocast priority) +{ + struct dccp_ackpkts *ap = kmalloc(sizeof(*ap) + len, priority); + + if (ap != NULL) { +#ifdef CONFIG_IP_DCCP_DEBUG + memset(ap->dccpap_buf, 0xFF, len); +#endif + ap->dccpap_buf_len = len; + ap->dccpap_buf_head = + ap->dccpap_buf_tail = + ap->dccpap_buf_len - 1; + ap->dccpap_buf_ackno = + ap->dccpap_ack_ackno = + ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1; + ap->dccpap_buf_nonce = ap->dccpap_buf_nonce = 0; + ap->dccpap_ack_ptr = 0; + ap->dccpap_time.tv_sec = 0; + ap->dccpap_time.tv_usec = 0; + ap->dccpap_buf_vector_len = ap->dccpap_ack_vector_len = 0; + } + + return ap; +} + +void dccp_ackpkts_free(struct dccp_ackpkts *ap) +{ + if (ap != NULL) { +#ifdef CONFIG_IP_DCCP_DEBUG + memset(ap, 0xFF, sizeof(*ap) + ap->dccpap_buf_len); +#endif + kfree(ap); + } +} + +static inline u8 dccp_ackpkts_state(const struct dccp_ackpkts *ap, + const unsigned int index) +{ + return ap->dccpap_buf[index] & DCCP_ACKPKTS_STATE_MASK; +} + +static inline u8 dccp_ackpkts_len(const struct dccp_ackpkts *ap, + const unsigned int index) +{ + return ap->dccpap_buf[index] & DCCP_ACKPKTS_LEN_MASK; +} + +/* + * If several packets are missing, the HC-Receiver may prefer to enter multiple + * bytes with run length 0, rather than a single byte with a larger run length; + * this simplifies table updates if one of the missing packets arrives. + */ +static inline int dccp_ackpkts_set_buf_head_state(struct dccp_ackpkts *ap, + const unsigned int packets, + const unsigned char state) +{ + unsigned int gap; + signed long new_head; + + if (ap->dccpap_buf_vector_len + packets > ap->dccpap_buf_len) + return -ENOBUFS; + + gap = packets - 1; + new_head = ap->dccpap_buf_head - packets; + + if (new_head < 0) { + if (gap > 0) { + memset(ap->dccpap_buf, DCCP_ACKPKTS_STATE_NOT_RECEIVED, + gap + new_head + 1); + gap = -new_head; + } + new_head += ap->dccpap_buf_len; + } + + ap->dccpap_buf_head = new_head; + + if (gap > 0) + memset(ap->dccpap_buf + ap->dccpap_buf_head + 1, + DCCP_ACKPKTS_STATE_NOT_RECEIVED, gap); + + ap->dccpap_buf[ap->dccpap_buf_head] = state; + ap->dccpap_buf_vector_len += packets; + return 0; +} + +/* + * Implements the draft-ietf-dccp-spec-11.txt Appendix A + */ +int dccp_ackpkts_add(struct dccp_ackpkts *ap, u64 ackno, u8 state) +{ + /* + * Check at the right places if the buffer is full, if it is, tell the + * caller to start dropping packets till the HC-Sender acks our ACK + * vectors, when we will free up space in dccpap_buf. + * + * We may well decide to do buffer compression, etc, but for now lets + * just drop. + * + * From Appendix A: + * + * Of course, the circular buffer may overflow, either when the + * HC-Sender is sending data at a very high rate, when the + * HC-Receiver's acknowledgements are not reaching the HC-Sender, + * or when the HC-Sender is forgetting to acknowledge those acks + * (so the HC-Receiver is unable to clean up old state). In this + * case, the HC-Receiver should either compress the buffer (by + * increasing run lengths when possible), transfer its state to + * a larger buffer, or, as a last resort, drop all received + * packets, without processing them whatsoever, until its buffer + * shrinks again. + */ + + /* See if this is the first ackno being inserted */ + if (ap->dccpap_buf_vector_len == 0) { + ap->dccpap_buf[ap->dccpap_buf_head] = state; + ap->dccpap_buf_vector_len = 1; + } else if (after48(ackno, ap->dccpap_buf_ackno)) { + const u64 delta = dccp_delta_seqno(ap->dccpap_buf_ackno, + ackno); + + /* + * Look if the state of this packet is the same as the + * previous ackno and if so if we can bump the head len. + */ + if (delta == 1 && + dccp_ackpkts_state(ap, ap->dccpap_buf_head) == state && + (dccp_ackpkts_len(ap, ap->dccpap_buf_head) < + DCCP_ACKPKTS_LEN_MASK)) + ap->dccpap_buf[ap->dccpap_buf_head]++; + else if (dccp_ackpkts_set_buf_head_state(ap, delta, state)) + return -ENOBUFS; + } else { + /* + * A.1.2. Old Packets + * + * When a packet with Sequence Number S arrives, and + * S <= buf_ackno, the HC-Receiver will scan the table + * for the byte corresponding to S. (Indexing structures + * could reduce the complexity of this scan.) + */ + u64 delta = dccp_delta_seqno(ackno, ap->dccpap_buf_ackno); + unsigned int index = ap->dccpap_buf_head; + + while (1) { + const u8 len = dccp_ackpkts_len(ap, index); + const u8 state = dccp_ackpkts_state(ap, index); + /* + * valid packets not yet in dccpap_buf have a reserved + * entry, with a len equal to 0. + */ + if (state == DCCP_ACKPKTS_STATE_NOT_RECEIVED && + len == 0 && delta == 0) { /* Found our + reserved seat! */ + dccp_pr_debug("Found %llu reserved seat!\n", + (unsigned long long) ackno); + ap->dccpap_buf[index] = state; + goto out; + } + /* len == 0 means one packet */ + if (delta < len + 1) + goto out_duplicate; + + delta -= len + 1; + if (++index == ap->dccpap_buf_len) + index = 0; + } + } + + ap->dccpap_buf_ackno = ackno; + do_gettimeofday(&ap->dccpap_time); +out: + dccp_pr_debug(""); + dccp_ackpkts_print(ap); + return 0; + +out_duplicate: + /* Duplicate packet */ + dccp_pr_debug("Received a dup or already considered lost " + "packet: %llu\n", (unsigned long long) ackno); + return -EILSEQ; +} + +#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); + + while (len--) { + const u8 state = (*vector & DCCP_ACKPKTS_STATE_MASK) >> 6; + const u8 rl = (*vector & DCCP_ACKPKTS_LEN_MASK); + + printk("%d,%d|", state, rl); + ++vector; + } + + printk("\n"); +} + +void dccp_ackpkts_print(const struct dccp_ackpkts *ap) +{ + dccp_ackvector_print(ap->dccpap_buf_ackno, + ap->dccpap_buf + ap->dccpap_buf_head, + ap->dccpap_buf_vector_len); +} +#endif + +static void dccp_ackpkts_trow_away_ack_record(struct dccp_ackpkts *ap) +{ + /* + * As we're keeping track of the ack vector size + * (dccpap_buf_vector_len) and the sent ack vector size + * (dccpap_ack_vector_len) we don't need dccpap_buf_tail at all, but + * keep this code here as in the future we'll implement a vector of + * ack records, as suggested in draft-ietf-dccp-spec-11.txt + * Appendix A. -acme + */ +#if 0 + ap->dccpap_buf_tail = ap->dccpap_ack_ptr + 1; + if (ap->dccpap_buf_tail >= ap->dccpap_buf_len) + ap->dccpap_buf_tail -= ap->dccpap_buf_len; +#endif + ap->dccpap_buf_vector_len -= ap->dccpap_ack_vector_len; +} + +void dccp_ackpkts_check_rcv_ackno(struct dccp_ackpkts *ap, struct sock *sk, + u64 ackno) +{ + /* Check if we actually sent an ACK vector */ + if (ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1) + return; + + if (ackno == ap->dccpap_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, " + "ack_ackno=%llu, ACKED!\n", + debug_prefix, 1, + (unsigned long long) ap->dccpap_ack_seqno, + (unsigned long long) ap->dccpap_ack_ackno); + dccp_ackpkts_trow_away_ack_record(ap); + ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1; + } +} + +static void dccp_ackpkts_check_rcv_ackvector(struct dccp_ackpkts *ap, + struct sock *sk, u64 ackno, + const unsigned char len, + const unsigned char *vector) +{ + unsigned char i; + + /* Check if we actually sent an ACK vector */ + if (ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1) + return; + /* + * We're in the receiver half connection, so if the received an ACK + * vector ackno (e.g. 50) before dccpap_ack_seqno (e.g. 52), we're + * not interested. + * + * Extra explanation with example: + * + * if we received an ACK vector with ackno 50, it can only be acking + * 50, 49, 48, etc, not 52 (the seqno for the ACK vector we sent). + */ + /* dccp_pr_debug("is %llu < %llu? ", ackno, ap->dccpap_ack_seqno); */ + if (before48(ackno, ap->dccpap_ack_seqno)) { + /* dccp_pr_debug_cat("yes\n"); */ + return; + } + /* dccp_pr_debug_cat("no\n"); */ + + i = len; + while (i--) { + const u8 rl = (*vector & DCCP_ACKPKTS_LEN_MASK); + u64 ackno_end_rl; + + dccp_set_seqno(&ackno_end_rl, ackno - rl); + + /* + * dccp_pr_debug("is %llu <= %llu <= %llu? ", ackno_end_rl, + * ap->dccpap_ack_seqno, ackno); + */ + if (between48(ap->dccpap_ack_seqno, ackno_end_rl, ackno)) { + const u8 state = (*vector & + DCCP_ACKPKTS_STATE_MASK) >> 6; + /* dccp_pr_debug_cat("yes\n"); */ + + if (state != DCCP_ACKPKTS_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, " + "ack_seqno=%llu, ack_ackno=%llu, " + "ACKED!\n", + debug_prefix, len, + (unsigned long long) + ap->dccpap_ack_seqno, + (unsigned long long) + ap->dccpap_ack_ackno); + dccp_ackpkts_trow_away_ack_record(ap); + } + /* + * If dccpap_ack_seqno was not received, no problem + * we'll send another ACK vector. + */ + ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1; + break; + } + /* dccp_pr_debug_cat("no\n"); */ + + dccp_set_seqno(&ackno, ackno_end_rl - 1); + ++vector; + } +} diff --git a/net/dccp/output.c b/net/dccp/output.c new file mode 100644 index 000000000000..28de157a4326 --- /dev/null +++ b/net/dccp/output.c @@ -0,0 +1,528 @@ +/* + * net/dccp/output.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/skbuff.h> + +#include <net/sock.h> + +#include "ccid.h" +#include "dccp.h" + +static inline void dccp_event_ack_sent(struct sock *sk) +{ + inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK); +} + +/* + * All SKB's seen here are completely headerless. It is our + * job to build the DCCP header, and pass the packet down to + * IP so it can do the same plus pass the packet off to the + * device. + */ +int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb) +{ + if (likely(skb != NULL)) { + const struct inet_sock *inet = inet_sk(sk); + struct dccp_sock *dp = dccp_sk(sk); + struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb); + struct dccp_hdr *dh; + /* XXX For now we're using only 48 bits sequence numbers */ + const int dccp_header_size = sizeof(*dh) + + sizeof(struct dccp_hdr_ext) + + dccp_packet_hdr_len(dcb->dccpd_type); + int err, set_ack = 1; + u64 ackno = dp->dccps_gsr; + + dccp_inc_seqno(&dp->dccps_gss); + + switch (dcb->dccpd_type) { + case DCCP_PKT_DATA: + set_ack = 0; + break; + case DCCP_PKT_SYNC: + case DCCP_PKT_SYNCACK: + ackno = dcb->dccpd_seq; + break; + } + + dcb->dccpd_seq = dp->dccps_gss; + dccp_insert_options(sk, skb); + + skb->h.raw = skb_push(skb, dccp_header_size); + dh = dccp_hdr(skb); + /* + * Data packets are not cloned as they are never retransmitted + */ + if (skb_cloned(skb)) + skb_set_owner_w(skb, sk); + + /* Build DCCP header and checksum it. */ + memset(dh, 0, 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; + /* XXX For now we're using only 48 bits sequence numbers */ + dh->dccph_x = 1; + + dp->dccps_awh = dp->dccps_gss; + dccp_hdr_set_seq(dh, dp->dccps_gss); + if (set_ack) + dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), ackno); + + switch (dcb->dccpd_type) { + case DCCP_PKT_REQUEST: + dccp_hdr_request(skb)->dccph_req_service = + dcb->dccpd_service; + break; + case DCCP_PKT_RESET: + dccp_hdr_reset(skb)->dccph_reset_code = + dcb->dccpd_reset_code; + break; + } + + dh->dccph_checksum = dccp_v4_checksum(skb, inet->saddr, + inet->daddr); + + if (set_ack) + dccp_event_ack_sent(sk); + + DCCP_INC_STATS(DCCP_MIB_OUTSEGS); + + err = ip_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; + } + return -ENOBUFS; +} + +unsigned int dccp_sync_mss(struct sock *sk, u32 pmtu) +{ + struct dccp_sock *dp = dccp_sk(sk); + int mss_now; + + /* + * FIXME: we really should be using the af_specific thing to support + * IPv6. + * mss_now = pmtu - tp->af_specific->net_header_len - + * sizeof(struct dccp_hdr) - sizeof(struct dccp_hdr_ext); + */ + mss_now = pmtu - sizeof(struct iphdr) - sizeof(struct dccp_hdr) - + sizeof(struct dccp_hdr_ext); + + /* Now subtract optional transport overhead */ + mss_now -= dp->dccps_ext_header_len; + + /* + * FIXME: this should come from the CCID infrastructure, where, say, + * TFRC will say it wants TIMESTAMPS, ELAPSED time, etc, for now lets + * put a rough estimate for NDP + TIMESTAMP + TIMESTAMP_ECHO + ELAPSED + * TIME + TFRC_OPT_LOSS_EVENT_RATE + TFRC_OPT_RECEIVE_RATE + padding to + * make it a multiple of 4 + */ + + mss_now -= ((5 + 6 + 10 + 6 + 6 + 6 + 3) / 4) * 4; + + /* And store cached results */ + dp->dccps_pmtu_cookie = pmtu; + dp->dccps_mss_cache = mss_now; + + return mss_now; +} + +void dccp_write_space(struct sock *sk) +{ + read_lock(&sk->sk_callback_lock); + + if (sk->sk_sleep && waitqueue_active(sk->sk_sleep)) + wake_up_interruptible(sk->sk_sleep); + /* Should agree with poll, otherwise some programs break */ + if (sock_writeable(sk)) + sk_wake_async(sk, 2, POLL_OUT); + + read_unlock(&sk->sk_callback_lock); +} + +/** + * 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) +{ + struct dccp_sock *dp = dccp_sk(sk); + DEFINE_WAIT(wait); + long delay; + int rc; + + while (1) { + prepare_to_wait(sk->sk_sleep, &wait, TASK_INTERRUPTIBLE); + + if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN)) + 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); + 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); + lock_sock(sk); + sk->sk_write_pending--; + } +out: + finish_wait(sk->sk_sleep, &wait); + return rc; + +do_error: + rc = -EPIPE; + goto out; +do_nonblock: + rc = -EAGAIN; + goto out; +do_interrupted: + rc = sock_intr_errno(*timeo); + goto out; +} + +int dccp_write_xmit(struct sock *sk, struct sk_buff *skb, long *timeo) +{ + const struct dccp_sock *dp = dccp_sk(sk); + int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb, + skb->len); + + if (err > 0) + err = dccp_wait_for_ccid(sk, skb, timeo); + + if (err == 0) { + const struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts; + struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb); + const int len = skb->len; + + if (sk->sk_state == DCCP_PARTOPEN) { + /* See 8.1.5. Handshake Completion */ + inet_csk_schedule_ack(sk); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + inet_csk(sk)->icsk_rto, + DCCP_RTO_MAX); + dcb->dccpd_type = DCCP_PKT_DATAACK; + /* + * FIXME: we really should have a + * dccps_ack_pending or use icsk. + */ + } else if (inet_csk_ack_scheduled(sk) || + dp->dccps_timestamp_echo != 0 || + (dp->dccps_options.dccpo_send_ack_vector && + ap->dccpap_buf_ackno != DCCP_MAX_SEQNO + 1 && + ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1)) + dcb->dccpd_type = DCCP_PKT_DATAACK; + else + dcb->dccpd_type = DCCP_PKT_DATA; + + err = dccp_transmit_skb(sk, skb); + ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len); + } + + return err; +} + +int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb) +{ + if (inet_sk_rebuild_header(sk) != 0) + return -EHOSTUNREACH; /* Routing failure or similar. */ + + return dccp_transmit_skb(sk, (skb_cloned(skb) ? + pskb_copy(skb, GFP_ATOMIC): + skb_clone(skb, GFP_ATOMIC))); +} + +struct sk_buff *dccp_make_response(struct sock *sk, struct dst_entry *dst, + struct request_sock *req) +{ + struct dccp_hdr *dh; + const int dccp_header_size = sizeof(struct dccp_hdr) + + sizeof(struct dccp_hdr_ext) + + sizeof(struct dccp_hdr_response); + struct sk_buff *skb = sock_wmalloc(sk, MAX_HEADER + DCCP_MAX_OPT_LEN + + dccp_header_size, 1, + GFP_ATOMIC); + if (skb == NULL) + return NULL; + + /* Reserve space for headers. */ + skb_reserve(skb, MAX_HEADER + DCCP_MAX_OPT_LEN + dccp_header_size); + + skb->dst = dst_clone(dst); + skb->csum = 0; + + DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_RESPONSE; + DCCP_SKB_CB(skb)->dccpd_seq = dccp_rsk(req)->dreq_iss; + dccp_insert_options(sk, skb); + + skb->h.raw = skb_push(skb, dccp_header_size); + + dh = dccp_hdr(skb); + memset(dh, 0, dccp_header_size); + + dh->dccph_sport = inet_sk(sk)->sport; + dh->dccph_dport = inet_rsk(req)->rmt_port; + dh->dccph_doff = (dccp_header_size + + DCCP_SKB_CB(skb)->dccpd_opt_len) / 4; + dh->dccph_type = DCCP_PKT_RESPONSE; + dh->dccph_x = 1; + dccp_hdr_set_seq(dh, dccp_rsk(req)->dreq_iss); + dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dccp_rsk(req)->dreq_isr); + + dh->dccph_checksum = dccp_v4_checksum(skb, inet_rsk(req)->loc_addr, + inet_rsk(req)->rmt_addr); + + DCCP_INC_STATS(DCCP_MIB_OUTSEGS); + return skb; +} + +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); + const int dccp_header_size = sizeof(struct dccp_hdr) + + sizeof(struct dccp_hdr_ext) + + sizeof(struct dccp_hdr_reset); + struct sk_buff *skb = sock_wmalloc(sk, MAX_HEADER + DCCP_MAX_OPT_LEN + + dccp_header_size, 1, + GFP_ATOMIC); + if (skb == NULL) + return NULL; + + /* Reserve space for headers. */ + skb_reserve(skb, MAX_HEADER + DCCP_MAX_OPT_LEN + dccp_header_size); + + skb->dst = dst_clone(dst); + skb->csum = 0; + + dccp_inc_seqno(&dp->dccps_gss); + + DCCP_SKB_CB(skb)->dccpd_reset_code = code; + DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_RESET; + DCCP_SKB_CB(skb)->dccpd_seq = dp->dccps_gss; + dccp_insert_options(sk, skb); + + skb->h.raw = skb_push(skb, dccp_header_size); + + dh = dccp_hdr(skb); + memset(dh, 0, dccp_header_size); + + dh->dccph_sport = inet_sk(sk)->sport; + dh->dccph_dport = inet_sk(sk)->dport; + dh->dccph_doff = (dccp_header_size + + DCCP_SKB_CB(skb)->dccpd_opt_len) / 4; + dh->dccph_type = DCCP_PKT_RESET; + dh->dccph_x = 1; + dccp_hdr_set_seq(dh, dp->dccps_gss); + dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dp->dccps_gsr); + + dccp_hdr_reset(skb)->dccph_reset_code = code; + + dh->dccph_checksum = dccp_v4_checksum(skb, inet_sk(sk)->saddr, + inet_sk(sk)->daddr); + + DCCP_INC_STATS(DCCP_MIB_OUTSEGS); + return skb; +} + +/* + * Do all connect socket setups that can be done AF independent. + */ +static inline void dccp_connect_init(struct sock *sk) +{ + struct dst_entry *dst = __sk_dst_get(sk); + struct inet_connection_sock *icsk = inet_csk(sk); + + sk->sk_err = 0; + sock_reset_flag(sk, SOCK_DONE); + + dccp_sync_mss(sk, dst_mtu(dst)); + + /* + * FIXME: set dp->{dccps_swh,dccps_swl}, with + * something like dccp_inc_seq + */ + + icsk->icsk_retransmits = 0; +} + +int dccp_connect(struct sock *sk) +{ + struct sk_buff *skb; + struct inet_connection_sock *icsk = inet_csk(sk); + + dccp_connect_init(sk); + + skb = alloc_skb(MAX_DCCP_HEADER + 15, sk->sk_allocation); + if (unlikely(skb == NULL)) + return -ENOBUFS; + + /* Reserve space for headers. */ + skb_reserve(skb, MAX_DCCP_HEADER); + + DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_REQUEST; + /* FIXME: set service to something meaningful, coming + * from userspace*/ + DCCP_SKB_CB(skb)->dccpd_service = 0; + skb->csum = 0; + skb_set_owner_w(skb, sk); + + BUG_TRAP(sk->sk_send_head == NULL); + sk->sk_send_head = skb; + dccp_transmit_skb(sk, skb_clone(skb, GFP_KERNEL)); + DCCP_INC_STATS(DCCP_MIB_ACTIVEOPENS); + + /* Timer for repeating the REQUEST until an answer. */ + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + icsk->icsk_rto, DCCP_RTO_MAX); + return 0; +} + +void dccp_send_ack(struct sock *sk) +{ + /* If we have been reset, we may not send again. */ + if (sk->sk_state != DCCP_CLOSED) { + struct sk_buff *skb = alloc_skb(MAX_DCCP_HEADER, GFP_ATOMIC); + + if (skb == NULL) { + inet_csk_schedule_ack(sk); + inet_csk(sk)->icsk_ack.ato = TCP_ATO_MIN; + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + TCP_DELACK_MAX, + DCCP_RTO_MAX); + return; + } + + /* Reserve space for headers */ + skb_reserve(skb, MAX_DCCP_HEADER); + skb->csum = 0; + DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_ACK; + skb_set_owner_w(skb, sk); + dccp_transmit_skb(sk, skb); + } +} + +EXPORT_SYMBOL_GPL(dccp_send_ack); + +void dccp_send_delayed_ack(struct sock *sk) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + /* + * FIXME: tune this timer. elapsed time fixes the skew, so no problem + * with using 2s, and active senders also piggyback the ACK into a + * DATAACK packet, so this is really for quiescent senders. + */ + unsigned long timeout = jiffies + 2 * HZ; + + /* Use new timeout only if there wasn't a older one earlier. */ + if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) { + /* If delack timer was blocked or is about to expire, + * send ACK now. + * + * FIXME: check the "about to expire" part + */ + if (icsk->icsk_ack.blocked) { + dccp_send_ack(sk); + return; + } + + if (!time_before(timeout, icsk->icsk_ack.timeout)) + timeout = icsk->icsk_ack.timeout; + } + icsk->icsk_ack.pending |= ICSK_ACK_SCHED | ICSK_ACK_TIMER; + icsk->icsk_ack.timeout = timeout; + sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout); +} + +void dccp_send_sync(struct sock *sk, const u64 seq, + const enum dccp_pkt_type pkt_type) +{ + /* + * We are not putting this on the write queue, so + * dccp_transmit_skb() will set the ownership to this + * sock. + */ + struct sk_buff *skb = alloc_skb(MAX_DCCP_HEADER, GFP_ATOMIC); + + if (skb == NULL) + /* FIXME: how to make sure the sync is sent? */ + return; + + /* Reserve space for headers and prepare control bits. */ + skb_reserve(skb, MAX_DCCP_HEADER); + skb->csum = 0; + DCCP_SKB_CB(skb)->dccpd_type = pkt_type; + DCCP_SKB_CB(skb)->dccpd_seq = seq; + + skb_set_owner_w(skb, sk); + dccp_transmit_skb(sk, skb); +} + +/* + * Send a DCCP_PKT_CLOSE/CLOSEREQ. The caller locks the socket for us. This + * cannot be allowed to fail queueing a DCCP_PKT_CLOSE/CLOSEREQ frame under + * any circumstances. + */ +void dccp_send_close(struct sock *sk, const int active) +{ + struct dccp_sock *dp = dccp_sk(sk); + struct sk_buff *skb; + const unsigned int prio = active ? GFP_KERNEL : GFP_ATOMIC; + + skb = alloc_skb(sk->sk_prot->max_header, prio); + if (skb == NULL) + return; + + /* 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; + + skb_set_owner_w(skb, sk); + if (active) { + BUG_TRAP(sk->sk_send_head == NULL); + sk->sk_send_head = skb; + dccp_transmit_skb(sk, skb_clone(skb, prio)); + } else + dccp_transmit_skb(sk, skb); + + ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk); + ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk); +} diff --git a/net/dccp/proto.c b/net/dccp/proto.c new file mode 100644 index 000000000000..18a0e69c9dc7 --- /dev/null +++ b/net/dccp/proto.c @@ -0,0 +1,826 @@ +/* + * net/dccp/proto.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/module.h> +#include <linux/types.h> +#include <linux/sched.h> +#include <linux/kernel.h> +#include <linux/skbuff.h> +#include <linux/netdevice.h> +#include <linux/in.h> +#include <linux/if_arp.h> +#include <linux/init.h> +#include <linux/random.h> +#include <net/checksum.h> + +#include <net/inet_common.h> +#include <net/ip.h> +#include <net/protocol.h> +#include <net/sock.h> +#include <net/xfrm.h> + +#include <asm/semaphore.h> +#include <linux/spinlock.h> +#include <linux/timer.h> +#include <linux/delay.h> +#include <linux/poll.h> +#include <linux/dccp.h> + +#include "ccid.h" +#include "dccp.h" + +DEFINE_SNMP_STAT(struct dccp_mib, dccp_statistics) __read_mostly; + +atomic_t dccp_orphan_count = ATOMIC_INIT(0); + +static struct net_protocol dccp_protocol = { + .handler = dccp_v4_rcv, + .err_handler = dccp_v4_err, +}; + +const char *dccp_packet_name(const int type) +{ + static const char *dccp_packet_names[] = { + [DCCP_PKT_REQUEST] = "REQUEST", + [DCCP_PKT_RESPONSE] = "RESPONSE", + [DCCP_PKT_DATA] = "DATA", + [DCCP_PKT_ACK] = "ACK", + [DCCP_PKT_DATAACK] = "DATAACK", + [DCCP_PKT_CLOSEREQ] = "CLOSEREQ", + [DCCP_PKT_CLOSE] = "CLOSE", + [DCCP_PKT_RESET] = "RESET", + [DCCP_PKT_SYNC] = "SYNC", + [DCCP_PKT_SYNCACK] = "SYNCACK", + }; + + if (type >= DCCP_NR_PKT_TYPES) + return "INVALID"; + else + return dccp_packet_names[type]; +} + +EXPORT_SYMBOL_GPL(dccp_packet_name); + +const char *dccp_state_name(const int state) +{ + static char *dccp_state_names[] = { + [DCCP_OPEN] = "OPEN", + [DCCP_REQUESTING] = "REQUESTING", + [DCCP_PARTOPEN] = "PARTOPEN", + [DCCP_LISTEN] = "LISTEN", + [DCCP_RESPOND] = "RESPOND", + [DCCP_CLOSING] = "CLOSING", + [DCCP_TIME_WAIT] = "TIME_WAIT", + [DCCP_CLOSED] = "CLOSED", + }; + + if (state >= DCCP_MAX_STATES) + return "INVALID STATE!"; + else + return dccp_state_names[state]; +} + +EXPORT_SYMBOL_GPL(dccp_state_name); + +static inline int dccp_listen_start(struct sock *sk) +{ + dccp_sk(sk)->dccps_role = DCCP_ROLE_LISTEN; + return inet_csk_listen_start(sk, TCP_SYNQ_HSIZE); +} + +int dccp_disconnect(struct sock *sk, int flags) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + struct inet_sock *inet = inet_sk(sk); + int err = 0; + const int old_state = sk->sk_state; + + if (old_state != DCCP_CLOSED) + dccp_set_state(sk, DCCP_CLOSED); + + /* ABORT function of RFC793 */ + if (old_state == DCCP_LISTEN) { + inet_csk_listen_stop(sk); + /* FIXME: do the active reset thing */ + } else if (old_state == DCCP_REQUESTING) + sk->sk_err = ECONNRESET; + + dccp_clear_xmit_timers(sk); + __skb_queue_purge(&sk->sk_receive_queue); + if (sk->sk_send_head != NULL) { + __kfree_skb(sk->sk_send_head); + sk->sk_send_head = NULL; + } + + inet->dport = 0; + + if (!(sk->sk_userlocks & SOCK_BINDADDR_LOCK)) + inet_reset_saddr(sk); + + sk->sk_shutdown = 0; + sock_reset_flag(sk, SOCK_DONE); + + icsk->icsk_backoff = 0; + inet_csk_delack_init(sk); + __sk_dst_reset(sk); + + BUG_TRAP(!inet->num || icsk->icsk_bind_hash); + + sk->sk_error_report(sk); + return err; +} + +/* + * Wait for a DCCP event. + * + * Note that we don't need to lock the socket, as the upper poll layers + * take care of normal races (between the test and the event) and we don't + * go look at any of the socket buffers directly. + */ +static unsigned int dccp_poll(struct file *file, struct socket *sock, + poll_table *wait) +{ + unsigned int mask; + struct sock *sk = sock->sk; + + poll_wait(file, sk->sk_sleep, wait); + if (sk->sk_state == DCCP_LISTEN) + return inet_csk_listen_poll(sk); + + /* Socket is not locked. We are protected from async events + by poll logic and correct handling of state changes + made by another threads is impossible in any case. + */ + + mask = 0; + if (sk->sk_err) + mask = POLLERR; + + if (sk->sk_shutdown == SHUTDOWN_MASK || sk->sk_state == DCCP_CLOSED) + mask |= POLLHUP; + if (sk->sk_shutdown & RCV_SHUTDOWN) + mask |= POLLIN | POLLRDNORM; + + /* Connected? */ + if ((1 << sk->sk_state) & ~(DCCPF_REQUESTING | DCCPF_RESPOND)) { + if (atomic_read(&sk->sk_rmem_alloc) > 0) + mask |= POLLIN | POLLRDNORM; + + if (!(sk->sk_shutdown & SEND_SHUTDOWN)) { + if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk)) { + mask |= POLLOUT | POLLWRNORM; + } else { /* send SIGIO later */ + set_bit(SOCK_ASYNC_NOSPACE, + &sk->sk_socket->flags); + set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); + + /* Race breaker. If space is freed after + * wspace test but before the flags are set, + * IO signal will be lost. + */ + if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk)) + mask |= POLLOUT | POLLWRNORM; + } + } + } + return mask; +} + +int dccp_ioctl(struct sock *sk, int cmd, unsigned long arg) +{ + dccp_pr_debug("entry\n"); + return -ENOIOCTLCMD; +} + +int dccp_setsockopt(struct sock *sk, int level, int optname, + char __user *optval, int optlen) +{ + struct dccp_sock *dp; + int err; + int val; + + if (level != SOL_DCCP) + return ip_setsockopt(sk, level, optname, optval, optlen); + + if (optlen < sizeof(int)) + return -EINVAL; + + if (get_user(val, (int __user *)optval)) + return -EFAULT; + + lock_sock(sk); + + dp = dccp_sk(sk); + err = 0; + + switch (optname) { + case DCCP_SOCKOPT_PACKET_SIZE: + dp->dccps_packet_size = val; + break; + default: + err = -ENOPROTOOPT; + break; + } + + release_sock(sk); + return err; +} + +int dccp_getsockopt(struct sock *sk, int level, int optname, + char __user *optval, int __user *optlen) +{ + struct dccp_sock *dp; + int val, len; + + if (level != SOL_DCCP) + return ip_getsockopt(sk, level, optname, optval, optlen); + + if (get_user(len, optlen)) + return -EFAULT; + + len = min_t(unsigned int, len, sizeof(int)); + if (len < 0) + return -EINVAL; + + dp = dccp_sk(sk); + + switch (optname) { + case DCCP_SOCKOPT_PACKET_SIZE: + val = dp->dccps_packet_size; + break; + default: + return -ENOPROTOOPT; + } + + if (put_user(len, optlen) || copy_to_user(optval, &val, len)) + return -EFAULT; + + return 0; +} + +int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, + size_t len) +{ + const struct dccp_sock *dp = dccp_sk(sk); + const int flags = msg->msg_flags; + const int noblock = flags & MSG_DONTWAIT; + struct sk_buff *skb; + int rc, size; + long timeo; + + if (len > dp->dccps_mss_cache) + return -EMSGSIZE; + + lock_sock(sk); + timeo = sock_sndtimeo(sk, noblock); + + /* + * We have to use sk_stream_wait_connect here to set sk_write_pending, + * so that the trick in dccp_rcv_request_sent_state_process. + */ + /* Wait for a connection to finish. */ + if ((1 << sk->sk_state) & ~(DCCPF_OPEN | DCCPF_PARTOPEN | DCCPF_CLOSING)) + if ((rc = sk_stream_wait_connect(sk, &timeo)) != 0) + goto out_release; + + size = sk->sk_prot->max_header + len; + release_sock(sk); + skb = sock_alloc_send_skb(sk, size, noblock, &rc); + lock_sock(sk); + if (skb == NULL) + goto out_release; + + skb_reserve(skb, sk->sk_prot->max_header); + rc = memcpy_fromiovec(skb_put(skb, len), msg->msg_iov, len); + if (rc != 0) + goto out_discard; + + rc = dccp_write_xmit(sk, skb, &timeo); + /* + * XXX we don't use sk_write_queue, so just discard the packet. + * Current plan however is to _use_ sk_write_queue with + * an algorith similar to tcp_sendmsg, where the main difference + * is that in DCCP we have to respect packet boundaries, so + * no coalescing of skbs. + * + * This bug was _quickly_ found & fixed by just looking at an OSTRA + * generated callgraph 8) -acme + */ + if (rc != 0) + goto out_discard; +out_release: + release_sock(sk); + return rc ? : len; +out_discard: + kfree_skb(skb); + goto out_release; +} + +int dccp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, + size_t len, int nonblock, int flags, int *addr_len) +{ + const struct dccp_hdr *dh; + long timeo; + + lock_sock(sk); + + if (sk->sk_state == DCCP_LISTEN) { + len = -ENOTCONN; + goto out; + } + + timeo = sock_rcvtimeo(sk, nonblock); + + do { + struct sk_buff *skb = skb_peek(&sk->sk_receive_queue); + + if (skb == NULL) + goto verify_sock_status; + + dh = dccp_hdr(skb); + + if (dh->dccph_type == DCCP_PKT_DATA || + dh->dccph_type == DCCP_PKT_DATAACK) + goto found_ok_skb; + + if (dh->dccph_type == DCCP_PKT_RESET || + dh->dccph_type == DCCP_PKT_CLOSE) { + dccp_pr_debug("found fin ok!\n"); + len = 0; + goto found_fin_ok; + } + dccp_pr_debug("packet_type=%s\n", + dccp_packet_name(dh->dccph_type)); + sk_eat_skb(sk, skb); +verify_sock_status: + if (sock_flag(sk, SOCK_DONE)) { + len = 0; + break; + } + + if (sk->sk_err) { + len = sock_error(sk); + break; + } + + if (sk->sk_shutdown & RCV_SHUTDOWN) { + len = 0; + break; + } + + if (sk->sk_state == DCCP_CLOSED) { + if (!sock_flag(sk, SOCK_DONE)) { + /* This occurs when user tries to read + * from never connected socket. + */ + len = -ENOTCONN; + break; + } + len = 0; + break; + } + + if (!timeo) { + len = -EAGAIN; + break; + } + + if (signal_pending(current)) { + len = sock_intr_errno(timeo); + break; + } + + sk_wait_data(sk, &timeo); + continue; + found_ok_skb: + if (len > skb->len) + len = skb->len; + else if (len < skb->len) + msg->msg_flags |= MSG_TRUNC; + + if (skb_copy_datagram_iovec(skb, 0, msg->msg_iov, len)) { + /* Exception. Bailout! */ + len = -EFAULT; + break; + } + found_fin_ok: + if (!(flags & MSG_PEEK)) + sk_eat_skb(sk, skb); + break; + } while (1); +out: + release_sock(sk); + return len; +} + +static int inet_dccp_listen(struct socket *sock, int backlog) +{ + struct sock *sk = sock->sk; + unsigned char old_state; + int err; + + lock_sock(sk); + + err = -EINVAL; + if (sock->state != SS_UNCONNECTED || sock->type != SOCK_DCCP) + goto out; + + old_state = sk->sk_state; + if (!((1 << old_state) & (DCCPF_CLOSED | DCCPF_LISTEN))) + goto out; + + /* Really, if the socket is already in listen state + * we can only allow the backlog to be adjusted. + */ + if (old_state != DCCP_LISTEN) { + /* + * FIXME: here it probably should be sk->sk_prot->listen_start + * see tcp_listen_start + */ + err = dccp_listen_start(sk); + if (err) + goto out; + } + sk->sk_max_ack_backlog = backlog; + err = 0; + +out: + release_sock(sk); + return err; +} + +static const unsigned char dccp_new_state[] = { + /* current state: new state: action: */ + [0] = DCCP_CLOSED, + [DCCP_OPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, + [DCCP_REQUESTING] = DCCP_CLOSED, + [DCCP_PARTOPEN] = DCCP_CLOSING | DCCP_ACTION_FIN, + [DCCP_LISTEN] = DCCP_CLOSED, + [DCCP_RESPOND] = DCCP_CLOSED, + [DCCP_CLOSING] = DCCP_CLOSED, + [DCCP_TIME_WAIT] = DCCP_CLOSED, + [DCCP_CLOSED] = DCCP_CLOSED, +}; + +static int dccp_close_state(struct sock *sk) +{ + const int next = dccp_new_state[sk->sk_state]; + const int ns = next & DCCP_STATE_MASK; + + if (ns != sk->sk_state) + dccp_set_state(sk, ns); + + return next & DCCP_ACTION_FIN; +} + +void dccp_close(struct sock *sk, long timeout) +{ + struct sk_buff *skb; + + lock_sock(sk); + + sk->sk_shutdown = SHUTDOWN_MASK; + + if (sk->sk_state == DCCP_LISTEN) { + dccp_set_state(sk, DCCP_CLOSED); + + /* Special case. */ + inet_csk_listen_stop(sk); + + goto adjudge_to_death; + } + + /* + * We need to flush the recv. buffs. We do this only on the + * descriptor close, not protocol-sourced closes, because the + *reader process may not have drained the data yet! + */ + /* FIXME: check for unread data */ + while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) { + __kfree_skb(skb); + } + + if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) { + /* Check zero linger _after_ checking for unread data. */ + sk->sk_prot->disconnect(sk, 0); + } else if (dccp_close_state(sk)) { + dccp_send_close(sk, 1); + } + + sk_stream_wait_close(sk, timeout); + +adjudge_to_death: + /* + * It is the last release_sock in its life. It will remove backlog. + */ + release_sock(sk); + /* + * Now socket is owned by kernel and we acquire BH lock + * to finish close. No need to check for user refs. + */ + local_bh_disable(); + bh_lock_sock(sk); + BUG_TRAP(!sock_owned_by_user(sk)); + + sock_hold(sk); + sock_orphan(sk); + + /* + * The last release_sock may have processed the CLOSE or RESET + * packet moving sock to CLOSED state, if not we have to fire + * the CLOSE/CLOSEREQ retransmission timer, see "8.3. Termination" + * in draft-ietf-dccp-spec-11. -acme + */ + if (sk->sk_state == DCCP_CLOSING) { + /* FIXME: should start at 2 * RTT */ + /* Timer for repeating the CLOSE/CLOSEREQ until an answer. */ + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, + DCCP_RTO_MAX); +#if 0 + /* Yeah, we should use sk->sk_prot->orphan_count, etc */ + dccp_set_state(sk, DCCP_CLOSED); +#endif + } + + atomic_inc(sk->sk_prot->orphan_count); + if (sk->sk_state == DCCP_CLOSED) + inet_csk_destroy_sock(sk); + + /* Otherwise, socket is reprieved until protocol close. */ + + bh_unlock_sock(sk); + local_bh_enable(); + sock_put(sk); +} + +void dccp_shutdown(struct sock *sk, int how) +{ + dccp_pr_debug("entry\n"); +} + +static struct proto_ops inet_dccp_ops = { + .family = PF_INET, + .owner = THIS_MODULE, + .release = inet_release, + .bind = inet_bind, + .connect = inet_stream_connect, + .socketpair = sock_no_socketpair, + .accept = inet_accept, + .getname = inet_getname, + /* FIXME: work on tcp_poll to rename it to inet_csk_poll */ + .poll = dccp_poll, + .ioctl = inet_ioctl, + /* FIXME: work on inet_listen to rename it to sock_common_listen */ + .listen = inet_dccp_listen, + .shutdown = inet_shutdown, + .setsockopt = sock_common_setsockopt, + .getsockopt = sock_common_getsockopt, + .sendmsg = inet_sendmsg, + .recvmsg = sock_common_recvmsg, + .mmap = sock_no_mmap, + .sendpage = sock_no_sendpage, +}; + +extern struct net_proto_family inet_family_ops; + +static struct inet_protosw dccp_v4_protosw = { + .type = SOCK_DCCP, + .protocol = IPPROTO_DCCP, + .prot = &dccp_v4_prot, + .ops = &inet_dccp_ops, + .capability = -1, + .no_check = 0, + .flags = 0, +}; + +/* + * This is the global socket data structure used for responding to + * the Out-of-the-blue (OOTB) packets. A control sock will be created + * for this socket at the initialization time. + */ +struct socket *dccp_ctl_socket; + +static char dccp_ctl_socket_err_msg[] __initdata = + KERN_ERR "DCCP: Failed to create the control socket.\n"; + +static int __init dccp_ctl_sock_init(void) +{ + int rc = sock_create_kern(PF_INET, SOCK_DCCP, IPPROTO_DCCP, + &dccp_ctl_socket); + if (rc < 0) + printk(dccp_ctl_socket_err_msg); + else { + dccp_ctl_socket->sk->sk_allocation = GFP_ATOMIC; + inet_sk(dccp_ctl_socket->sk)->uc_ttl = -1; + + /* Unhash it so that IP input processing does not even + * see it, we do not wish this socket to see incoming + * packets. + */ + dccp_ctl_socket->sk->sk_prot->unhash(dccp_ctl_socket->sk); + } + + return rc; +} + +#ifdef CONFIG_IP_DCCP_UNLOAD_HACK +void dccp_ctl_sock_exit(void) +{ + if (dccp_ctl_socket != NULL) { + sock_release(dccp_ctl_socket); + dccp_ctl_socket = NULL; + } +} + +EXPORT_SYMBOL_GPL(dccp_ctl_sock_exit); +#endif + +static int __init init_dccp_v4_mibs(void) +{ + int rc = -ENOMEM; + + dccp_statistics[0] = alloc_percpu(struct dccp_mib); + if (dccp_statistics[0] == NULL) + goto out; + + dccp_statistics[1] = alloc_percpu(struct dccp_mib); + if (dccp_statistics[1] == NULL) + goto out_free_one; + + rc = 0; +out: + return rc; +out_free_one: + free_percpu(dccp_statistics[0]); + dccp_statistics[0] = NULL; + goto out; + +} + +static int thash_entries; +module_param(thash_entries, int, 0444); +MODULE_PARM_DESC(thash_entries, "Number of ehash buckets"); + +#ifdef CONFIG_IP_DCCP_DEBUG +int dccp_debug; +module_param(dccp_debug, int, 0444); +MODULE_PARM_DESC(dccp_debug, "Enable debug messages"); +#endif + +static int __init dccp_init(void) +{ + unsigned long goal; + int ehash_order, bhash_order, i; + int rc = proto_register(&dccp_v4_prot, 1); + + if (rc) + goto out; + + dccp_hashinfo.bind_bucket_cachep = + kmem_cache_create("dccp_bind_bucket", + sizeof(struct inet_bind_bucket), 0, + SLAB_HWCACHE_ALIGN, NULL, NULL); + if (!dccp_hashinfo.bind_bucket_cachep) + goto out_proto_unregister; + + /* + * Size and allocate the main established and bind bucket + * hash tables. + * + * The methodology is similar to that of the buffer cache. + */ + if (num_physpages >= (128 * 1024)) + goal = num_physpages >> (21 - PAGE_SHIFT); + else + goal = num_physpages >> (23 - PAGE_SHIFT); + + if (thash_entries) + goal = (thash_entries * + sizeof(struct inet_ehash_bucket)) >> PAGE_SHIFT; + for (ehash_order = 0; (1UL << ehash_order) < goal; ehash_order++) + ; + do { + dccp_hashinfo.ehash_size = (1UL << ehash_order) * PAGE_SIZE / + sizeof(struct inet_ehash_bucket); + dccp_hashinfo.ehash_size >>= 1; + while (dccp_hashinfo.ehash_size & + (dccp_hashinfo.ehash_size - 1)) + dccp_hashinfo.ehash_size--; + dccp_hashinfo.ehash = (struct inet_ehash_bucket *) + __get_free_pages(GFP_ATOMIC, ehash_order); + } while (!dccp_hashinfo.ehash && --ehash_order > 0); + + if (!dccp_hashinfo.ehash) { + printk(KERN_CRIT "Failed to allocate DCCP " + "established hash table\n"); + goto out_free_bind_bucket_cachep; + } + + for (i = 0; i < (dccp_hashinfo.ehash_size << 1); i++) { + rwlock_init(&dccp_hashinfo.ehash[i].lock); + INIT_HLIST_HEAD(&dccp_hashinfo.ehash[i].chain); + } + + bhash_order = ehash_order; + + do { + dccp_hashinfo.bhash_size = (1UL << bhash_order) * PAGE_SIZE / + sizeof(struct inet_bind_hashbucket); + if ((dccp_hashinfo.bhash_size > (64 * 1024)) && + bhash_order > 0) + continue; + dccp_hashinfo.bhash = (struct inet_bind_hashbucket *) + __get_free_pages(GFP_ATOMIC, bhash_order); + } while (!dccp_hashinfo.bhash && --bhash_order >= 0); + + if (!dccp_hashinfo.bhash) { + printk(KERN_CRIT "Failed to allocate DCCP bind hash table\n"); + goto out_free_dccp_ehash; + } + + for (i = 0; i < dccp_hashinfo.bhash_size; i++) { + spin_lock_init(&dccp_hashinfo.bhash[i].lock); + INIT_HLIST_HEAD(&dccp_hashinfo.bhash[i].chain); + } + + if (init_dccp_v4_mibs()) + goto out_free_dccp_bhash; + + rc = -EAGAIN; + if (inet_add_protocol(&dccp_protocol, IPPROTO_DCCP)) + goto out_free_dccp_v4_mibs; + + inet_register_protosw(&dccp_v4_protosw); + + rc = dccp_ctl_sock_init(); + if (rc) + goto out_unregister_protosw; +out: + return rc; +out_unregister_protosw: + inet_unregister_protosw(&dccp_v4_protosw); + inet_del_protocol(&dccp_protocol, IPPROTO_DCCP); +out_free_dccp_v4_mibs: + free_percpu(dccp_statistics[0]); + free_percpu(dccp_statistics[1]); + dccp_statistics[0] = dccp_statistics[1] = NULL; +out_free_dccp_bhash: + free_pages((unsigned long)dccp_hashinfo.bhash, bhash_order); + dccp_hashinfo.bhash = NULL; +out_free_dccp_ehash: + free_pages((unsigned long)dccp_hashinfo.ehash, ehash_order); + dccp_hashinfo.ehash = NULL; +out_free_bind_bucket_cachep: + kmem_cache_destroy(dccp_hashinfo.bind_bucket_cachep); + dccp_hashinfo.bind_bucket_cachep = NULL; +out_proto_unregister: + proto_unregister(&dccp_v4_prot); + goto out; +} + +static const char dccp_del_proto_err_msg[] __exitdata = + KERN_ERR "can't remove dccp net_protocol\n"; + +static void __exit dccp_fini(void) +{ + inet_unregister_protosw(&dccp_v4_protosw); + + if (inet_del_protocol(&dccp_protocol, IPPROTO_DCCP) < 0) + printk(dccp_del_proto_err_msg); + + free_percpu(dccp_statistics[0]); + free_percpu(dccp_statistics[1]); + free_pages((unsigned long)dccp_hashinfo.bhash, + get_order(dccp_hashinfo.bhash_size * + sizeof(struct inet_bind_hashbucket))); + free_pages((unsigned long)dccp_hashinfo.ehash, + get_order(dccp_hashinfo.ehash_size * + sizeof(struct inet_ehash_bucket))); + kmem_cache_destroy(dccp_hashinfo.bind_bucket_cachep); + proto_unregister(&dccp_v4_prot); +} + +module_init(dccp_init); +module_exit(dccp_fini); + +/* + * __stringify doesn't likes enums, so use SOCK_DCCP (6) and IPPROTO_DCCP (33) + * values directly, Also cover the case where the protocol is not specified, + * i.e. net-pf-PF_INET-proto-0-type-SOCK_DCCP + */ +MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-proto-33-type-6"); +MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-proto-0-type-6"); +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Arnaldo Carvalho de Melo <acme@conectiva.com.br>"); +MODULE_DESCRIPTION("DCCP - Datagram Congestion Controlled Protocol"); diff --git a/net/dccp/timer.c b/net/dccp/timer.c new file mode 100644 index 000000000000..aa34b576e228 --- /dev/null +++ b/net/dccp/timer.c @@ -0,0 +1,255 @@ +/* + * net/dccp/timer.c + * + * An implementation of the DCCP protocol + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/dccp.h> +#include <linux/skbuff.h> + +#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); +} + +static void dccp_write_err(struct sock *sk) +{ + sk->sk_err = sk->sk_err_soft ? : ETIMEDOUT; + sk->sk_error_report(sk); + + dccp_v4_send_reset(sk, DCCP_RESET_CODE_ABORTED); + dccp_done(sk); + DCCP_INC_STATS_BH(DCCP_MIB_ABORTONTIMEOUT); +} + +/* A write timeout has occurred. Process the after effects. */ +static int dccp_write_timeout(struct sock *sk) +{ + const struct inet_connection_sock *icsk = inet_csk(sk); + int retry_until; + + 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 */; + } else { + if (icsk->icsk_retransmits >= + /* FIXME! sysctl_tcp_retries1 */ 5 /* FIXME! */) { + /* NOTE. draft-ietf-tcpimpl-pmtud-01.txt requires pmtu + black hole detection. :-( + + It is place to make it. It is not made. I do not want + to make it. It is disguisting. It does not work in any + case. Let me to cite the same draft, which requires for + us to implement this: + + "The one security concern raised by this memo is that ICMP black holes + are often caused by over-zealous security administrators who block + all ICMP messages. It is vitally important that those who design and + deploy security systems understand the impact of strict filtering on + upper-layer protocols. The safest web site in the world is worthless + if most TCP implementations cannot transfer data from it. It would + be far nicer to have all of the black holes fixed rather than fixing + all of the TCP implementations." + + Golden words :-). + */ + + dst_negative_advice(&sk->sk_dst_cache); + } + + retry_until = /* FIXME! */ 15 /* FIXME! sysctl_tcp_retries2 */; + /* + * FIXME: see tcp_write_timout and tcp_out_of_resources + */ + } + + if (icsk->icsk_retransmits >= retry_until) { + /* Has it gone just too far? */ + dccp_write_err(sk); + return 1; + } + 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. + */ +static void dccp_retransmit_timer(struct sock *sk) +{ + struct inet_connection_sock *icsk = inet_csk(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). + */ + BUG_TRAP(sk->sk_send_head != NULL); + + /* + * More than than 4MSL (8 minutes) has passed, a RESET(aborted) was + * sent, no need to retransmit, this sock is dead. + */ + if (dccp_write_timeout(sk)) + goto out; + + /* + * We want to know the number of packets retransmitted, not the + * total number of retransmissions of clones of original packets. + */ + if (icsk->icsk_retransmits == 0) + DCCP_INC_STATS_BH(DCCP_MIB_TIMEOUTS); + + if (dccp_retransmit_skb(sk, sk->sk_send_head) < 0) { + /* + * Retransmission failed because of local congestion, + * do not backoff. + */ + if (icsk->icsk_retransmits == 0) + icsk->icsk_retransmits = 1; + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + min(icsk->icsk_rto, + TCP_RESOURCE_PROBE_INTERVAL), + DCCP_RTO_MAX); + goto out; + } + + icsk->icsk_backoff++; + icsk->icsk_retransmits++; + + icsk->icsk_rto = min(icsk->icsk_rto << 1, DCCP_RTO_MAX); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto, + DCCP_RTO_MAX); + if (icsk->icsk_retransmits > 3 /* FIXME: sysctl_dccp_retries1 */) + __sk_dst_reset(sk); +out:; +} + +static void dccp_write_timer(unsigned long data) +{ + struct sock *sk = (struct sock *)data; + struct inet_connection_sock *icsk = inet_csk(sk); + int event = 0; + + bh_lock_sock(sk); + if (sock_owned_by_user(sk)) { + /* Try again later */ + sk_reset_timer(sk, &icsk->icsk_retransmit_timer, + jiffies + (HZ / 20)); + goto out; + } + + if (sk->sk_state == DCCP_CLOSED || !icsk->icsk_pending) + goto out; + + if (time_after(icsk->icsk_timeout, jiffies)) { + sk_reset_timer(sk, &icsk->icsk_retransmit_timer, + icsk->icsk_timeout); + goto out; + } + + event = icsk->icsk_pending; + icsk->icsk_pending = 0; + + switch (event) { + case ICSK_TIME_RETRANS: + dccp_retransmit_timer(sk); + break; + } +out: + bh_unlock_sock(sk); + sock_put(sk); +} + +/* + * Timer for listening sockets + */ +static void dccp_response_timer(struct sock *sk) +{ + inet_csk_reqsk_queue_prune(sk, TCP_SYNQ_INTERVAL, DCCP_TIMEOUT_INIT, + DCCP_RTO_MAX); +} + +static void dccp_keepalive_timer(unsigned long data) +{ + struct sock *sk = (struct sock *)data; + + /* Only process if socket is not in use. */ + bh_lock_sock(sk); + if (sock_owned_by_user(sk)) { + /* Try again later. */ + inet_csk_reset_keepalive_timer(sk, HZ / 20); + goto out; + } + + if (sk->sk_state == DCCP_LISTEN) { + dccp_response_timer(sk); + goto out; + } +out: + bh_unlock_sock(sk); + sock_put(sk); +} diff --git a/net/decnet/af_decnet.c b/net/decnet/af_decnet.c index acdd18e6adb2..621680f127af 100644 --- a/net/decnet/af_decnet.c +++ b/net/decnet/af_decnet.c @@ -118,7 +118,7 @@ Version 0.0.6 2.1.110 07-aug-98 Eduardo Marcelo Serrat #include <linux/netfilter.h> #include <linux/seq_file.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/flow.h> #include <asm/system.h> #include <asm/ioctls.h> @@ -1763,7 +1763,7 @@ static int dn_recvmsg(struct kiocb *iocb, struct socket *sock, nskb = skb->next; if (skb->len == 0) { - skb_unlink(skb); + skb_unlink(skb, queue); kfree_skb(skb); /* * N.B. Don't refer to skb or cb after this point @@ -2064,7 +2064,7 @@ static struct notifier_block dn_dev_notifier = { .notifier_call = dn_device_event, }; -extern int dn_route_rcv(struct sk_buff *, struct net_device *, struct packet_type *); +extern int dn_route_rcv(struct sk_buff *, struct net_device *, struct packet_type *, struct net_device *); static struct packet_type dn_dix_packet_type = { .type = __constant_htons(ETH_P_DNA_RT), diff --git a/net/decnet/dn_dev.c b/net/decnet/dn_dev.c index 00233ecbc9cb..5610bb16dbf9 100644 --- a/net/decnet/dn_dev.c +++ b/net/decnet/dn_dev.c @@ -752,16 +752,16 @@ static void rtmsg_ifa(int event, struct dn_ifaddr *ifa) skb = alloc_skb(size, GFP_KERNEL); if (!skb) { - netlink_set_err(rtnl, 0, RTMGRP_DECnet_IFADDR, ENOBUFS); + netlink_set_err(rtnl, 0, RTNLGRP_DECnet_IFADDR, ENOBUFS); return; } if (dn_dev_fill_ifaddr(skb, ifa, 0, 0, event, 0) < 0) { kfree_skb(skb); - netlink_set_err(rtnl, 0, RTMGRP_DECnet_IFADDR, EINVAL); + netlink_set_err(rtnl, 0, RTNLGRP_DECnet_IFADDR, EINVAL); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_DECnet_IFADDR; - netlink_broadcast(rtnl, skb, 0, RTMGRP_DECnet_IFADDR, GFP_KERNEL); + NETLINK_CB(skb).dst_group = RTNLGRP_DECnet_IFADDR; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_DECnet_IFADDR, GFP_KERNEL); } static int dn_dev_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb) diff --git a/net/decnet/dn_nsp_in.c b/net/decnet/dn_nsp_in.c index 202dbde9850d..369f25b60f3f 100644 --- a/net/decnet/dn_nsp_in.c +++ b/net/decnet/dn_nsp_in.c @@ -60,7 +60,7 @@ #include <linux/inet.h> #include <linux/route.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/system.h> #include <linux/fcntl.h> #include <linux/mm.h> diff --git a/net/decnet/dn_nsp_out.c b/net/decnet/dn_nsp_out.c index 8cce1fdbda90..e0bebf4bbcad 100644 --- a/net/decnet/dn_nsp_out.c +++ b/net/decnet/dn_nsp_out.c @@ -479,7 +479,7 @@ int dn_nsp_check_xmit_queue(struct sock *sk, struct sk_buff *skb, struct sk_buff xmit_count = cb2->xmit_count; segnum = cb2->segnum; /* Remove and drop ack'ed packet */ - skb_unlink(ack); + skb_unlink(ack, q); kfree_skb(ack); ack = NULL; diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c index 2399fa8a3f86..2c915f305be3 100644 --- a/net/decnet/dn_route.c +++ b/net/decnet/dn_route.c @@ -572,7 +572,7 @@ static int dn_route_ptp_hello(struct sk_buff *skb) return NET_RX_SUCCESS; } -int dn_route_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +int dn_route_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct dn_skb_cb *cb; unsigned char flags = 0; diff --git a/net/decnet/dn_table.c b/net/decnet/dn_table.c index 28ba5777a25a..eeba56f99323 100644 --- a/net/decnet/dn_table.c +++ b/net/decnet/dn_table.c @@ -79,7 +79,7 @@ for( ; ((f) = *(fp)) != NULL && dn_key_eq((f)->fn_key, (key)); (fp) = &(f)->fn_n static DEFINE_RWLOCK(dn_fib_tables_lock); struct dn_fib_table *dn_fib_tables[RT_TABLE_MAX + 1]; -static kmem_cache_t *dn_hash_kmem; +static kmem_cache_t *dn_hash_kmem __read_mostly; static int dn_fib_hash_zombies; static inline dn_fib_idx_t dn_hash(dn_fib_key_t key, struct dn_zone *dz) @@ -349,10 +349,10 @@ static void dn_rtmsg_fib(int event, struct dn_fib_node *f, int z, int tb_id, kfree_skb(skb); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_DECnet_ROUTE; + NETLINK_CB(skb).dst_group = RTNLGRP_DECnet_ROUTE; if (nlh->nlmsg_flags & NLM_F_ECHO) atomic_inc(&skb->users); - netlink_broadcast(rtnl, skb, pid, RTMGRP_DECnet_ROUTE, GFP_KERNEL); + netlink_broadcast(rtnl, skb, pid, RTNLGRP_DECnet_ROUTE, GFP_KERNEL); if (nlh->nlmsg_flags & NLM_F_ECHO) netlink_unicast(rtnl, skb, pid, MSG_DONTWAIT); } diff --git a/net/decnet/netfilter/dn_rtmsg.c b/net/decnet/netfilter/dn_rtmsg.c index 284a9998e53d..1ab94c6e22ed 100644 --- a/net/decnet/netfilter/dn_rtmsg.c +++ b/net/decnet/netfilter/dn_rtmsg.c @@ -19,6 +19,7 @@ #include <linux/netfilter.h> #include <linux/spinlock.h> #include <linux/netlink.h> +#include <linux/netfilter_decnet.h> #include <net/sock.h> #include <net/flow.h> @@ -71,10 +72,10 @@ static void dnrmg_send_peer(struct sk_buff *skb) switch(flags & DN_RT_CNTL_MSK) { case DN_RT_PKT_L1RT: - group = DNRMG_L1_GROUP; + group = DNRNG_NLGRP_L1; break; case DN_RT_PKT_L2RT: - group = DNRMG_L2_GROUP; + group = DNRNG_NLGRP_L2; break; default: return; @@ -83,7 +84,7 @@ static void dnrmg_send_peer(struct sk_buff *skb) skb2 = dnrmg_build_message(skb, &status); if (skb2 == NULL) return; - NETLINK_CB(skb2).dst_groups = group; + NETLINK_CB(skb2).dst_group = group; netlink_broadcast(dnrmg, skb2, 0, group, GFP_ATOMIC); } @@ -138,7 +139,8 @@ static int __init init(void) { int rv = 0; - dnrmg = netlink_kernel_create(NETLINK_DNRTMSG, dnrmg_receive_user_sk); + dnrmg = netlink_kernel_create(NETLINK_DNRTMSG, DNRNG_NLGRP_MAX, + dnrmg_receive_user_sk, THIS_MODULE); if (dnrmg == NULL) { printk(KERN_ERR "dn_rtmsg: Cannot create netlink socket"); return -ENOMEM; @@ -162,6 +164,7 @@ static void __exit fini(void) MODULE_DESCRIPTION("DECnet Routing Message Grabulator"); MODULE_AUTHOR("Steven Whitehouse <steve@chygwyn.com>"); MODULE_LICENSE("GPL"); +MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_DNRTMSG); module_init(init); module_exit(fini); diff --git a/net/econet/af_econet.c b/net/econet/af_econet.c index de691e119e17..4a62093eb343 100644 --- a/net/econet/af_econet.c +++ b/net/econet/af_econet.c @@ -159,7 +159,7 @@ static int econet_recvmsg(struct kiocb *iocb, struct socket *sock, err = memcpy_toiovec(msg->msg_iov, skb->data, copied); if (err) goto out_free; - sk->sk_stamp = skb->stamp; + skb_get_timestamp(skb, &sk->sk_stamp); if (msg->msg_name) memcpy(msg->msg_name, skb->cb, msg->msg_namelen); @@ -869,7 +869,7 @@ static void aun_tx_ack(unsigned long seq, int result) foundit: tx_result(skb->sk, eb->cookie, result); - skb_unlink(skb); + skb_unlink(skb, &aun_queue); spin_unlock_irqrestore(&aun_queue_lock, flags); kfree_skb(skb); } @@ -947,7 +947,7 @@ static void ab_cleanup(unsigned long h) { tx_result(skb->sk, eb->cookie, ECTYPE_TRANSMIT_NOT_PRESENT); - skb_unlink(skb); + skb_unlink(skb, &aun_queue); kfree_skb(skb); } skb = newskb; @@ -1009,7 +1009,7 @@ release: * Receive an Econet frame from a device. */ -static int econet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +static int econet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct ec_framehdr *hdr; struct sock *sk; diff --git a/net/ethernet/eth.c b/net/ethernet/eth.c index f6dbfb99b14d..87a052a9a84f 100644 --- a/net/ethernet/eth.c +++ b/net/ethernet/eth.c @@ -62,8 +62,6 @@ #include <asm/system.h> #include <asm/checksum.h> -extern int __init netdev_boot_setup(char *str); - __setup("ether=", netdev_boot_setup); /* @@ -163,7 +161,6 @@ __be16 eth_type_trans(struct sk_buff *skb, struct net_device *dev) skb->mac.raw=skb->data; skb_pull(skb,ETH_HLEN); eth = eth_hdr(skb); - skb->input_dev = dev; if(*eth->h_dest&1) { diff --git a/net/ethernet/sysctl_net_ether.c b/net/ethernet/sysctl_net_ether.c index b81a6d532342..66b39fc342d2 100644 --- a/net/ethernet/sysctl_net_ether.c +++ b/net/ethernet/sysctl_net_ether.c @@ -7,6 +7,7 @@ #include <linux/mm.h> #include <linux/sysctl.h> +#include <linux/if_ether.h> ctl_table ether_table[] = { {0} diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig index 0b3d9f1d8069..e55136ae09f4 100644 --- a/net/ipv4/Kconfig +++ b/net/ipv4/Kconfig @@ -413,20 +413,19 @@ config INET_TUNNEL If unsure, say Y. -config IP_TCPDIAG - tristate "IP: TCP socket monitoring interface" +config INET_DIAG + tristate "INET: socket monitoring interface" default y ---help--- - Support for TCP socket monitoring interface used by native Linux - tools such as ss. ss is included in iproute2, currently downloadable - at <http://developer.osdl.org/dev/iproute2>. If you want IPv6 support - and have selected IPv6 as a module, you need to build this as a - module too. + Support for INET (TCP, DCCP, etc) socket monitoring interface used by + native Linux tools such as ss. ss is included in iproute2, currently + downloadable at <http://developer.osdl.org/dev/iproute2>. If unsure, say Y. -config IP_TCPDIAG_IPV6 - def_bool (IP_TCPDIAG=y && IPV6=y) || (IP_TCPDIAG=m && IPV6) +config INET_TCP_DIAG + depends on INET_DIAG + def_tristate INET_DIAG config TCP_CONG_ADVANCED bool "TCP: advanced congestion control" diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile index 55dc6cca1e7b..f0435d00db6b 100644 --- a/net/ipv4/Makefile +++ b/net/ipv4/Makefile @@ -4,11 +4,12 @@ obj-y := route.o inetpeer.o protocol.o \ ip_input.o ip_fragment.o ip_forward.o ip_options.o \ - ip_output.o ip_sockglue.o \ + ip_output.o ip_sockglue.o inet_hashtables.o \ + inet_timewait_sock.o inet_connection_sock.o \ tcp.o tcp_input.o tcp_output.o tcp_timer.o tcp_ipv4.o \ tcp_minisocks.o tcp_cong.o \ datagram.o raw.o udp.o arp.o icmp.o devinet.o af_inet.o igmp.o \ - sysctl_net_ipv4.o fib_frontend.o fib_semantics.o + sysctl_net_ipv4.o fib_frontend.o fib_semantics.o netfilter.o obj-$(CONFIG_IP_FIB_HASH) += fib_hash.o obj-$(CONFIG_IP_FIB_TRIE) += fib_trie.o @@ -29,8 +30,9 @@ obj-$(CONFIG_IP_ROUTE_MULTIPATH_WRANDOM) += multipath_wrandom.o obj-$(CONFIG_IP_ROUTE_MULTIPATH_DRR) += multipath_drr.o obj-$(CONFIG_NETFILTER) += netfilter/ obj-$(CONFIG_IP_VS) += ipvs/ -obj-$(CONFIG_IP_TCPDIAG) += tcp_diag.o +obj-$(CONFIG_INET_DIAG) += inet_diag.o obj-$(CONFIG_IP_ROUTE_MULTIPATH_CACHED) += multipath.o +obj-$(CONFIG_INET_TCP_DIAG) += tcp_diag.o obj-$(CONFIG_TCP_CONG_BIC) += tcp_bic.o obj-$(CONFIG_TCP_CONG_WESTWOOD) += tcp_westwood.o obj-$(CONFIG_TCP_CONG_HSTCP) += tcp_highspeed.o diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 163ae4068b5f..bf147f8db399 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -99,6 +99,7 @@ #include <net/arp.h> #include <net/route.h> #include <net/ip_fib.h> +#include <net/inet_connection_sock.h> #include <net/tcp.h> #include <net/udp.h> #include <linux/skbuff.h> @@ -112,11 +113,7 @@ #include <linux/mroute.h> #endif -DEFINE_SNMP_STAT(struct linux_mib, net_statistics); - -#ifdef INET_REFCNT_DEBUG -atomic_t inet_sock_nr; -#endif +DEFINE_SNMP_STAT(struct linux_mib, net_statistics) __read_mostly; extern void ip_mc_drop_socket(struct sock *sk); @@ -153,11 +150,7 @@ void inet_sock_destruct(struct sock *sk) if (inet->opt) kfree(inet->opt); dst_release(sk->sk_dst_cache); -#ifdef INET_REFCNT_DEBUG - atomic_dec(&inet_sock_nr); - printk(KERN_DEBUG "INET socket %p released, %d are still alive\n", - sk, atomic_read(&inet_sock_nr)); -#endif + sk_refcnt_debug_dec(sk); } /* @@ -210,7 +203,7 @@ int inet_listen(struct socket *sock, int backlog) * we can only allow the backlog to be adjusted. */ if (old_state != TCP_LISTEN) { - err = tcp_listen_start(sk); + err = inet_csk_listen_start(sk, TCP_SYNQ_HSIZE); if (err) goto out; } @@ -235,12 +228,14 @@ static int inet_create(struct socket *sock, int protocol) struct proto *answer_prot; unsigned char answer_flags; char answer_no_check; - int err; + int try_loading_module = 0; + int err = -ESOCKTNOSUPPORT; sock->state = SS_UNCONNECTED; /* Look for the requested type/protocol pair. */ answer = NULL; +lookup_protocol: rcu_read_lock(); list_for_each_rcu(p, &inetsw[sock->type]) { answer = list_entry(p, struct inet_protosw, list); @@ -261,9 +256,28 @@ static int inet_create(struct socket *sock, int protocol) answer = NULL; } - err = -ESOCKTNOSUPPORT; - if (!answer) - goto out_rcu_unlock; + if (unlikely(answer == NULL)) { + if (try_loading_module < 2) { + rcu_read_unlock(); + /* + * Be more specific, e.g. net-pf-2-proto-132-type-1 + * (net-pf-PF_INET-proto-IPPROTO_SCTP-type-SOCK_STREAM) + */ + if (++try_loading_module == 1) + request_module("net-pf-%d-proto-%d-type-%d", + PF_INET, protocol, sock->type); + /* + * Fall back to generic, e.g. net-pf-2-proto-132 + * (net-pf-PF_INET-proto-IPPROTO_SCTP) + */ + else + request_module("net-pf-%d-proto-%d", + PF_INET, protocol); + goto lookup_protocol; + } else + goto out_rcu_unlock; + } + err = -EPERM; if (answer->capability > 0 && !capable(answer->capability)) goto out_rcu_unlock; @@ -317,9 +331,7 @@ static int inet_create(struct socket *sock, int protocol) inet->mc_index = 0; inet->mc_list = NULL; -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet_sock_nr); -#endif + sk_refcnt_debug_inc(sk); if (inet->num) { /* It assumes that any protocol which allows @@ -847,10 +859,6 @@ static struct net_proto_family inet_family_ops = { .owner = THIS_MODULE, }; - -extern void tcp_init(void); -extern void tcp_v4_init(struct net_proto_family *); - /* Upon startup we insert all the elements in inetsw_array[] into * the linked list inetsw. */ @@ -961,6 +969,119 @@ void inet_unregister_protosw(struct inet_protosw *p) } } +/* + * Shall we try to damage output packets if routing dev changes? + */ + +int sysctl_ip_dynaddr; + +static int inet_sk_reselect_saddr(struct sock *sk) +{ + struct inet_sock *inet = inet_sk(sk); + int err; + struct rtable *rt; + __u32 old_saddr = inet->saddr; + __u32 new_saddr; + __u32 daddr = inet->daddr; + + if (inet->opt && inet->opt->srr) + daddr = inet->opt->faddr; + + /* Query new route. */ + err = ip_route_connect(&rt, daddr, 0, + RT_CONN_FLAGS(sk), + sk->sk_bound_dev_if, + sk->sk_protocol, + inet->sport, inet->dport, sk); + if (err) + return err; + + sk_setup_caps(sk, &rt->u.dst); + + new_saddr = rt->rt_src; + + if (new_saddr == old_saddr) + return 0; + + if (sysctl_ip_dynaddr > 1) { + printk(KERN_INFO "%s(): shifting inet->" + "saddr from %d.%d.%d.%d to %d.%d.%d.%d\n", + __FUNCTION__, + NIPQUAD(old_saddr), + NIPQUAD(new_saddr)); + } + + inet->saddr = inet->rcv_saddr = new_saddr; + + /* + * XXX The only one ugly spot where we need to + * XXX really change the sockets identity after + * XXX it has entered the hashes. -DaveM + * + * Besides that, it does not check for connection + * uniqueness. Wait for troubles. + */ + __sk_prot_rehash(sk); + return 0; +} + +int inet_sk_rebuild_header(struct sock *sk) +{ + struct inet_sock *inet = inet_sk(sk); + struct rtable *rt = (struct rtable *)__sk_dst_check(sk, 0); + u32 daddr; + int err; + + /* Route is OK, nothing to do. */ + if (rt) + return 0; + + /* Reroute. */ + daddr = inet->daddr; + if (inet->opt && inet->opt->srr) + daddr = inet->opt->faddr; +{ + struct flowi fl = { + .oif = sk->sk_bound_dev_if, + .nl_u = { + .ip4_u = { + .daddr = daddr, + .saddr = inet->saddr, + .tos = RT_CONN_FLAGS(sk), + }, + }, + .proto = sk->sk_protocol, + .uli_u = { + .ports = { + .sport = inet->sport, + .dport = inet->dport, + }, + }, + }; + + err = ip_route_output_flow(&rt, &fl, sk, 0); +} + if (!err) + sk_setup_caps(sk, &rt->u.dst); + else { + /* Routing failed... */ + sk->sk_route_caps = 0; + /* + * Other protocols have to map its equivalent state to TCP_SYN_SENT. + * DCCP maps its DCCP_REQUESTING state to TCP_SYN_SENT. -acme + */ + if (!sysctl_ip_dynaddr || + sk->sk_state != TCP_SYN_SENT || + (sk->sk_userlocks & SOCK_BINDADDR_LOCK) || + (err = inet_sk_reselect_saddr(sk)) != 0) + sk->sk_err_soft = -err; + } + + return err; +} + +EXPORT_SYMBOL(inet_sk_rebuild_header); + #ifdef CONFIG_IP_MULTICAST static struct net_protocol igmp_protocol = { .handler = igmp_rcv, @@ -1007,7 +1128,6 @@ static int __init init_ipv4_mibs(void) } static int ipv4_proc_init(void); -extern void ipfrag_init(void); /* * IP protocol layer initialiser @@ -1128,19 +1248,10 @@ module_init(inet_init); /* ------------------------------------------------------------------------ */ #ifdef CONFIG_PROC_FS -extern int fib_proc_init(void); -extern void fib_proc_exit(void); #ifdef CONFIG_IP_FIB_TRIE extern int fib_stat_proc_init(void); extern void fib_stat_proc_exit(void); #endif -extern int ip_misc_proc_init(void); -extern int raw_proc_init(void); -extern void raw_proc_exit(void); -extern int tcp4_proc_init(void); -extern void tcp4_proc_exit(void); -extern int udp4_proc_init(void); -extern void udp4_proc_exit(void); static int __init ipv4_proc_init(void) { @@ -1205,7 +1316,3 @@ EXPORT_SYMBOL(inet_stream_ops); EXPORT_SYMBOL(inet_unregister_protosw); EXPORT_SYMBOL(net_statistics); EXPORT_SYMBOL(sysctl_ip_nonlocal_bind); - -#ifdef INET_REFCNT_DEBUG -EXPORT_SYMBOL(inet_sock_nr); -#endif diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index a642fd612853..8bf312bdea13 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -700,7 +700,7 @@ void arp_send(int type, int ptype, u32 dest_ip, static void parp_redo(struct sk_buff *skb) { nf_reset(skb); - arp_rcv(skb, skb->dev, NULL); + arp_rcv(skb, skb->dev, NULL, skb->dev); } /* @@ -865,7 +865,7 @@ static int arp_process(struct sk_buff *skb) if (n) neigh_release(n); - if (skb->stamp.tv_sec == LOCALLY_ENQUEUED || + if (NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED || skb->pkt_type == PACKET_HOST || in_dev->arp_parms->proxy_delay == 0) { arp_send(ARPOP_REPLY,ETH_P_ARP,sip,dev,tip,sha,dev->dev_addr,sha); @@ -927,7 +927,7 @@ out: * Receive an arp request from the device layer. */ -int arp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +int arp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct arphdr *arp; @@ -948,6 +948,8 @@ int arp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL) goto out_of_mem; + memset(NEIGH_CB(skb), 0, sizeof(struct neighbour_cb)); + return NF_HOOK(NF_ARP, NF_ARP_IN, skb, dev, NULL, arp_process); freeskb: diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c index b1db561f2542..c1b42b5257f8 100644 --- a/net/ipv4/datagram.c +++ b/net/ipv4/datagram.c @@ -16,9 +16,10 @@ #include <linux/module.h> #include <linux/ip.h> #include <linux/in.h> +#include <net/ip.h> #include <net/sock.h> -#include <net/tcp.h> #include <net/route.h> +#include <net/tcp_states.h> int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) { diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index d8a10e3dd77d..ba2895ae8151 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1111,13 +1111,12 @@ static void rtmsg_ifa(int event, struct in_ifaddr* ifa) struct sk_buff *skb = alloc_skb(size, GFP_KERNEL); if (!skb) - netlink_set_err(rtnl, 0, RTMGRP_IPV4_IFADDR, ENOBUFS); + netlink_set_err(rtnl, 0, RTNLGRP_IPV4_IFADDR, ENOBUFS); else if (inet_fill_ifaddr(skb, ifa, current->pid, 0, event, 0) < 0) { kfree_skb(skb); - netlink_set_err(rtnl, 0, RTMGRP_IPV4_IFADDR, EINVAL); + netlink_set_err(rtnl, 0, RTNLGRP_IPV4_IFADDR, EINVAL); } else { - NETLINK_CB(skb).dst_groups = RTMGRP_IPV4_IFADDR; - netlink_broadcast(rtnl, skb, 0, RTMGRP_IPV4_IFADDR, GFP_KERNEL); + netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV4_IFADDR, GFP_KERNEL); } } diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c index ba57446d5d1f..b31ffc5053d2 100644 --- a/net/ipv4/esp4.c +++ b/net/ipv4/esp4.c @@ -331,8 +331,8 @@ static void esp4_err(struct sk_buff *skb, u32 info) x = xfrm_state_lookup((xfrm_address_t *)&iph->daddr, esph->spi, IPPROTO_ESP, AF_INET); if (!x) return; - NETDEBUG(printk(KERN_DEBUG "pmtu discovery on SA ESP/%08x/%08x\n", - ntohl(esph->spi), ntohl(iph->daddr))); + NETDEBUG(KERN_DEBUG "pmtu discovery on SA ESP/%08x/%08x\n", + ntohl(esph->spi), ntohl(iph->daddr)); xfrm_state_put(x); } @@ -395,10 +395,10 @@ static int esp_init_state(struct xfrm_state *x) if (aalg_desc->uinfo.auth.icv_fullbits/8 != crypto_tfm_alg_digestsize(esp->auth.tfm)) { - NETDEBUG(printk(KERN_INFO "ESP: %s digestsize %u != %hu\n", - x->aalg->alg_name, - crypto_tfm_alg_digestsize(esp->auth.tfm), - aalg_desc->uinfo.auth.icv_fullbits/8)); + NETDEBUG(KERN_INFO "ESP: %s digestsize %u != %hu\n", + x->aalg->alg_name, + crypto_tfm_alg_digestsize(esp->auth.tfm), + aalg_desc->uinfo.auth.icv_fullbits/8); goto error; } diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c index cd8e45ab9580..4e1379f71269 100644 --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -558,16 +558,15 @@ static void nl_fib_input(struct sock *sk, int len) nl_fib_lookup(frn, tb); pid = nlh->nlmsg_pid; /*pid of sending process */ - NETLINK_CB(skb).groups = 0; /* not in mcast group */ NETLINK_CB(skb).pid = 0; /* from kernel */ NETLINK_CB(skb).dst_pid = pid; - NETLINK_CB(skb).dst_groups = 0; /* unicast */ + NETLINK_CB(skb).dst_group = 0; /* unicast */ netlink_unicast(sk, skb, pid, MSG_DONTWAIT); } static void nl_fib_lookup_init(void) { - netlink_kernel_create(NETLINK_FIB_LOOKUP, nl_fib_input); + netlink_kernel_create(NETLINK_FIB_LOOKUP, 0, nl_fib_input, THIS_MODULE); } static void fib_disable_ip(struct net_device *dev, int force) @@ -662,5 +661,4 @@ void __init ip_fib_init(void) } EXPORT_SYMBOL(inet_addr_type); -EXPORT_SYMBOL(ip_dev_find); EXPORT_SYMBOL(ip_rt_ioctl); diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c index b10d6bb5ef3d..2a8c9afc3695 100644 --- a/net/ipv4/fib_hash.c +++ b/net/ipv4/fib_hash.c @@ -45,8 +45,8 @@ #include "fib_lookup.h" -static kmem_cache_t *fn_hash_kmem; -static kmem_cache_t *fn_alias_kmem; +static kmem_cache_t *fn_hash_kmem __read_mostly; +static kmem_cache_t *fn_alias_kmem __read_mostly; struct fib_node { struct hlist_node fn_hash; diff --git a/net/ipv4/fib_lookup.h b/net/ipv4/fib_lookup.h index b729d97cfa93..ef6609ea0eb7 100644 --- a/net/ipv4/fib_lookup.h +++ b/net/ipv4/fib_lookup.h @@ -7,6 +7,7 @@ struct fib_alias { struct list_head fa_list; + struct rcu_head rcu; struct fib_info *fa_info; u8 fa_tos; u8 fa_type; diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index e278cb9d0075..d41219e8037c 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -290,10 +290,10 @@ void rtmsg_fib(int event, u32 key, struct fib_alias *fa, kfree_skb(skb); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_IPV4_ROUTE; + NETLINK_CB(skb).dst_group = RTNLGRP_IPV4_ROUTE; if (n->nlmsg_flags&NLM_F_ECHO) atomic_inc(&skb->users); - netlink_broadcast(rtnl, skb, pid, RTMGRP_IPV4_ROUTE, GFP_KERNEL); + netlink_broadcast(rtnl, skb, pid, RTNLGRP_IPV4_ROUTE, GFP_KERNEL); if (n->nlmsg_flags&NLM_F_ECHO) netlink_unicast(rtnl, skb, pid, MSG_DONTWAIT); } @@ -854,6 +854,7 @@ failure: return NULL; } +/* Note! fib_semantic_match intentionally uses RCU list functions. */ int fib_semantic_match(struct list_head *head, const struct flowi *flp, struct fib_result *res, __u32 zone, __u32 mask, int prefixlen) @@ -861,7 +862,7 @@ int fib_semantic_match(struct list_head *head, const struct flowi *flp, struct fib_alias *fa; int nh_sel = 0; - list_for_each_entry(fa, head, fa_list) { + list_for_each_entry_rcu(fa, head, fa_list) { int err; if (fa->fa_tos && diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c index 45efd5f4741b..b2dea4e5da77 100644 --- a/net/ipv4/fib_trie.c +++ b/net/ipv4/fib_trie.c @@ -43,7 +43,7 @@ * 2 of the License, or (at your option) any later version. */ -#define VERSION "0.325" +#define VERSION "0.402" #include <linux/config.h> #include <asm/uaccess.h> @@ -62,6 +62,7 @@ #include <linux/netdevice.h> #include <linux/if_arp.h> #include <linux/proc_fs.h> +#include <linux/rcupdate.h> #include <linux/skbuff.h> #include <linux/netlink.h> #include <linux/init.h> @@ -77,56 +78,55 @@ #undef CONFIG_IP_FIB_TRIE_STATS #define MAX_CHILDS 16384 -#define EXTRACT(p, n, str) ((str)<<(p)>>(32-(n))) #define KEYLENGTH (8*sizeof(t_key)) #define MASK_PFX(k, l) (((l)==0)?0:(k >> (KEYLENGTH-l)) << (KEYLENGTH-l)) #define TKEY_GET_MASK(offset, bits) (((bits)==0)?0:((t_key)(-1) << (KEYLENGTH - bits) >> offset)) -static DEFINE_RWLOCK(fib_lock); - typedef unsigned int t_key; #define T_TNODE 0 #define T_LEAF 1 #define NODE_TYPE_MASK 0x1UL -#define NODE_PARENT(_node) \ - ((struct tnode *)((_node)->_parent & ~NODE_TYPE_MASK)) -#define NODE_SET_PARENT(_node, _ptr) \ - ((_node)->_parent = (((unsigned long)(_ptr)) | \ - ((_node)->_parent & NODE_TYPE_MASK))) -#define NODE_INIT_PARENT(_node, _type) \ - ((_node)->_parent = (_type)) -#define NODE_TYPE(_node) \ - ((_node)->_parent & NODE_TYPE_MASK) - -#define IS_TNODE(n) (!(n->_parent & T_LEAF)) -#define IS_LEAF(n) (n->_parent & T_LEAF) +#define NODE_PARENT(node) \ + ((struct tnode *)rcu_dereference(((node)->parent & ~NODE_TYPE_MASK))) + +#define NODE_TYPE(node) ((node)->parent & NODE_TYPE_MASK) + +#define NODE_SET_PARENT(node, ptr) \ + rcu_assign_pointer((node)->parent, \ + ((unsigned long)(ptr)) | NODE_TYPE(node)) + +#define IS_TNODE(n) (!(n->parent & T_LEAF)) +#define IS_LEAF(n) (n->parent & T_LEAF) struct node { - t_key key; - unsigned long _parent; + t_key key; + unsigned long parent; }; struct leaf { - t_key key; - unsigned long _parent; + t_key key; + unsigned long parent; struct hlist_head list; + struct rcu_head rcu; }; struct leaf_info { struct hlist_node hlist; + struct rcu_head rcu; int plen; struct list_head falh; }; struct tnode { - t_key key; - unsigned long _parent; - unsigned short pos:5; /* 2log(KEYLENGTH) bits needed */ - unsigned short bits:5; /* 2log(KEYLENGTH) bits needed */ - unsigned short full_children; /* KEYLENGTH bits needed */ - unsigned short empty_children; /* KEYLENGTH bits needed */ - struct node *child[0]; + t_key key; + unsigned long parent; + unsigned short pos:5; /* 2log(KEYLENGTH) bits needed */ + unsigned short bits:5; /* 2log(KEYLENGTH) bits needed */ + unsigned short full_children; /* KEYLENGTH bits needed */ + unsigned short empty_children; /* KEYLENGTH bits needed */ + struct rcu_head rcu; + struct node *child[0]; }; #ifdef CONFIG_IP_FIB_TRIE_STATS @@ -150,77 +150,45 @@ struct trie_stat { }; struct trie { - struct node *trie; + struct node *trie; #ifdef CONFIG_IP_FIB_TRIE_STATS struct trie_use_stats stats; #endif - int size; + int size; unsigned int revision; }; -static int trie_debug = 0; - -static int tnode_full(struct tnode *tn, struct node *n); static void put_child(struct trie *t, struct tnode *tn, int i, struct node *n); static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n, int wasfull); -static int tnode_child_length(struct tnode *tn); static struct node *resize(struct trie *t, struct tnode *tn); -static struct tnode *inflate(struct trie *t, struct tnode *tn, int *err); -static struct tnode *halve(struct trie *t, struct tnode *tn, int *err); +static struct tnode *inflate(struct trie *t, struct tnode *tn); +static struct tnode *halve(struct trie *t, struct tnode *tn); static void tnode_free(struct tnode *tn); static void trie_dump_seq(struct seq_file *seq, struct trie *t); -extern struct fib_alias *fib_find_alias(struct list_head *fah, u8 tos, u32 prio); -extern int fib_detect_death(struct fib_info *fi, int order, - struct fib_info **last_resort, int *last_idx, int *dflt); - -extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa, int z, int tb_id, - struct nlmsghdr *n, struct netlink_skb_parms *req); -static kmem_cache_t *fn_alias_kmem; +static kmem_cache_t *fn_alias_kmem __read_mostly; static struct trie *trie_local = NULL, *trie_main = NULL; -static void trie_bug(char *err) -{ - printk("Trie Bug: %s\n", err); - BUG(); -} + +/* rcu_read_lock needs to be hold by caller from readside */ static inline struct node *tnode_get_child(struct tnode *tn, int i) { - if (i >= 1<<tn->bits) - trie_bug("tnode_get_child"); + BUG_ON(i >= 1 << tn->bits); - return tn->child[i]; + return rcu_dereference(tn->child[i]); } -static inline int tnode_child_length(struct tnode *tn) +static inline int tnode_child_length(const struct tnode *tn) { - return 1<<tn->bits; + return 1 << tn->bits; } -/* - _________________________________________________________________ - | i | i | i | i | i | i | i | N | N | N | S | S | S | S | S | C | - ---------------------------------------------------------------- - 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 - - _________________________________________________________________ - | C | C | C | u | u | u | u | u | u | u | u | u | u | u | u | u | - ----------------------------------------------------------------- - 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 - - tp->pos = 7 - tp->bits = 3 - n->pos = 15 - n->bits=4 - KEYLENGTH=32 -*/ - static inline t_key tkey_extract_bits(t_key a, int offset, int bits) { - if (offset < KEYLENGTH) + if (offset < KEYLENGTH) return ((t_key)(a << offset)) >> (KEYLENGTH - bits); - else + else return 0; } @@ -233,8 +201,8 @@ static inline int tkey_sub_equals(t_key a, int offset, int bits, t_key b) { if (bits == 0 || offset >= KEYLENGTH) return 1; - bits = bits > KEYLENGTH ? KEYLENGTH : bits; - return ((a ^ b) << offset) >> (KEYLENGTH - bits) == 0; + bits = bits > KEYLENGTH ? KEYLENGTH : bits; + return ((a ^ b) << offset) >> (KEYLENGTH - bits) == 0; } static inline int tkey_mismatch(t_key a, int offset, t_key b) @@ -249,14 +217,6 @@ static inline int tkey_mismatch(t_key a, int offset, t_key b) return i; } -/* Candiate for fib_semantics */ - -static void fn_free_alias(struct fib_alias *fa) -{ - fib_release_info(fa->fa_info); - kmem_cache_free(fn_alias_kmem, fa); -} - /* To understand this stuff, an understanding of keys and all their bits is necessary. Every node in the trie has a key associated with it, but not @@ -295,7 +255,7 @@ static void fn_free_alias(struct fib_alias *fa) tp->pos = 7 tp->bits = 3 n->pos = 15 - n->bits=4 + n->bits = 4 First, let's just ignore the bits that come before the parent tp, that is the bits from 0 to (tp->pos-1). They are *known* but at this point we do @@ -320,60 +280,65 @@ static void fn_free_alias(struct fib_alias *fa) */ -static void check_tnode(struct tnode *tn) +static inline void check_tnode(const struct tnode *tn) { - if (tn && tn->pos+tn->bits > 32) { - printk("TNODE ERROR tn=%p, pos=%d, bits=%d\n", tn, tn->pos, tn->bits); - } + WARN_ON(tn && tn->pos+tn->bits > 32); } static int halve_threshold = 25; static int inflate_threshold = 50; -static struct leaf *leaf_new(void) + +static void __alias_free_mem(struct rcu_head *head) { - struct leaf *l = kmalloc(sizeof(struct leaf), GFP_KERNEL); - if (l) { - NODE_INIT_PARENT(l, T_LEAF); - INIT_HLIST_HEAD(&l->list); - } - return l; + struct fib_alias *fa = container_of(head, struct fib_alias, rcu); + kmem_cache_free(fn_alias_kmem, fa); } -static struct leaf_info *leaf_info_new(int plen) +static inline void alias_free_mem_rcu(struct fib_alias *fa) { - struct leaf_info *li = kmalloc(sizeof(struct leaf_info), GFP_KERNEL); - if (li) { - li->plen = plen; - INIT_LIST_HEAD(&li->falh); - } - return li; + call_rcu(&fa->rcu, __alias_free_mem); +} + +static void __leaf_free_rcu(struct rcu_head *head) +{ + kfree(container_of(head, struct leaf, rcu)); +} + +static inline void free_leaf(struct leaf *leaf) +{ + call_rcu(&leaf->rcu, __leaf_free_rcu); } -static inline void free_leaf(struct leaf *l) +static void __leaf_info_free_rcu(struct rcu_head *head) { - kfree(l); + kfree(container_of(head, struct leaf_info, rcu)); } -static inline void free_leaf_info(struct leaf_info *li) +static inline void free_leaf_info(struct leaf_info *leaf) { - kfree(li); + call_rcu(&leaf->rcu, __leaf_info_free_rcu); } static struct tnode *tnode_alloc(unsigned int size) { - if (size <= PAGE_SIZE) { - return kmalloc(size, GFP_KERNEL); - } else { - return (struct tnode *) - __get_free_pages(GFP_KERNEL, get_order(size)); - } + struct page *pages; + + if (size <= PAGE_SIZE) + return kcalloc(size, 1, GFP_KERNEL); + + pages = alloc_pages(GFP_KERNEL|__GFP_ZERO, get_order(size)); + if (!pages) + return NULL; + + return page_address(pages); } -static void __tnode_free(struct tnode *tn) +static void __tnode_free_rcu(struct rcu_head *head) { + struct tnode *tn = container_of(head, struct tnode, rcu); unsigned int size = sizeof(struct tnode) + - (1<<tn->bits) * sizeof(struct node *); + (1 << tn->bits) * sizeof(struct node *); if (size <= PAGE_SIZE) kfree(tn); @@ -381,15 +346,40 @@ static void __tnode_free(struct tnode *tn) free_pages((unsigned long)tn, get_order(size)); } +static inline void tnode_free(struct tnode *tn) +{ + call_rcu(&tn->rcu, __tnode_free_rcu); +} + +static struct leaf *leaf_new(void) +{ + struct leaf *l = kmalloc(sizeof(struct leaf), GFP_KERNEL); + if (l) { + l->parent = T_LEAF; + INIT_HLIST_HEAD(&l->list); + } + return l; +} + +static struct leaf_info *leaf_info_new(int plen) +{ + struct leaf_info *li = kmalloc(sizeof(struct leaf_info), GFP_KERNEL); + if (li) { + li->plen = plen; + INIT_LIST_HEAD(&li->falh); + } + return li; +} + static struct tnode* tnode_new(t_key key, int pos, int bits) { int nchildren = 1<<bits; int sz = sizeof(struct tnode) + nchildren * sizeof(struct node *); struct tnode *tn = tnode_alloc(sz); - if (tn) { + if (tn) { memset(tn, 0, sz); - NODE_INIT_PARENT(tn, T_TNODE); + tn->parent = T_TNODE; tn->pos = pos; tn->bits = bits; tn->key = key; @@ -397,38 +387,17 @@ static struct tnode* tnode_new(t_key key, int pos, int bits) tn->empty_children = 1<<bits; } - if (trie_debug > 0) - printk("AT %p s=%u %u\n", tn, (unsigned int) sizeof(struct tnode), - (unsigned int) (sizeof(struct node) * 1<<bits)); + pr_debug("AT %p s=%u %u\n", tn, (unsigned int) sizeof(struct tnode), + (unsigned int) (sizeof(struct node) * 1<<bits)); return tn; } -static void tnode_free(struct tnode *tn) -{ - if (!tn) { - trie_bug("tnode_free\n"); - } - if (IS_LEAF(tn)) { - free_leaf((struct leaf *)tn); - if (trie_debug > 0 ) - printk("FL %p \n", tn); - } - else if (IS_TNODE(tn)) { - __tnode_free(tn); - if (trie_debug > 0 ) - printk("FT %p \n", tn); - } - else { - trie_bug("tnode_free\n"); - } -} - /* * Check whether a tnode 'n' is "full", i.e. it is an internal node * and no bits are skipped. See discussion in dyntree paper p. 6 */ -static inline int tnode_full(struct tnode *tn, struct node *n) +static inline int tnode_full(const struct tnode *tn, const struct node *n) { if (n == NULL || IS_LEAF(n)) return 0; @@ -448,15 +417,11 @@ static inline void put_child(struct trie *t, struct tnode *tn, int i, struct nod static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n, int wasfull) { - struct node *chi; + struct node *chi = tn->child[i]; int isfull; - if (i >= 1<<tn->bits) { - printk("bits=%d, i=%d\n", tn->bits, i); - trie_bug("tnode_put_child_reorg bits"); - } - write_lock_bh(&fib_lock); - chi = tn->child[i]; + BUG_ON(i >= 1<<tn->bits); + /* update emptyChildren */ if (n == NULL && chi != NULL) @@ -465,33 +430,32 @@ static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n, int w tn->empty_children--; /* update fullChildren */ - if (wasfull == -1) + if (wasfull == -1) wasfull = tnode_full(tn, chi); isfull = tnode_full(tn, n); if (wasfull && !isfull) tn->full_children--; - else if (!wasfull && isfull) tn->full_children++; + if (n) NODE_SET_PARENT(n, tn); - tn->child[i] = n; - write_unlock_bh(&fib_lock); + rcu_assign_pointer(tn->child[i], n); } static struct node *resize(struct trie *t, struct tnode *tn) { int i; int err = 0; + struct tnode *old_tn; if (!tn) return NULL; - if (trie_debug) - printk("In tnode_resize %p inflate_threshold=%d threshold=%d\n", - tn, inflate_threshold, halve_threshold); + pr_debug("In tnode_resize %p inflate_threshold=%d threshold=%d\n", + tn, inflate_threshold, halve_threshold); /* No children */ if (tn->empty_children == tnode_child_length(tn)) { @@ -501,20 +465,16 @@ static struct node *resize(struct trie *t, struct tnode *tn) /* One child */ if (tn->empty_children == tnode_child_length(tn) - 1) for (i = 0; i < tnode_child_length(tn); i++) { + struct node *n; - write_lock_bh(&fib_lock); - if (tn->child[i] != NULL) { - - /* compress one level */ - struct node *n = tn->child[i]; - if (n) - NODE_INIT_PARENT(n, NODE_TYPE(n)); + n = tn->child[i]; + if (!n) + continue; - write_unlock_bh(&fib_lock); - tnode_free(tn); - return n; - } - write_unlock_bh(&fib_lock); + /* compress one level */ + NODE_SET_PARENT(n, NULL); + tnode_free(tn); + return n; } /* * Double as long as the resulting node has a number of @@ -566,16 +526,16 @@ static struct node *resize(struct trie *t, struct tnode *tn) * * expand not_to_be_doubled and to_be_doubled, and shorten: * 100 * (tnode_child_length(tn) - tn->empty_children + - * tn->full_children ) >= inflate_threshold * new_child_length + * tn->full_children) >= inflate_threshold * new_child_length * * expand new_child_length: * 100 * (tnode_child_length(tn) - tn->empty_children + - * tn->full_children ) >= + * tn->full_children) >= * inflate_threshold * tnode_child_length(tn) * 2 * * shorten again: * 50 * (tn->full_children + tnode_child_length(tn) - - * tn->empty_children ) >= inflate_threshold * + * tn->empty_children) >= inflate_threshold * * tnode_child_length(tn) * */ @@ -587,9 +547,10 @@ static struct node *resize(struct trie *t, struct tnode *tn) 50 * (tn->full_children + tnode_child_length(tn) - tn->empty_children) >= inflate_threshold * tnode_child_length(tn))) { - tn = inflate(t, tn, &err); - - if (err) { + old_tn = tn; + tn = inflate(t, tn); + if (IS_ERR(tn)) { + tn = old_tn; #ifdef CONFIG_IP_FIB_TRIE_STATS t->stats.resize_node_skipped++; #endif @@ -609,9 +570,10 @@ static struct node *resize(struct trie *t, struct tnode *tn) 100 * (tnode_child_length(tn) - tn->empty_children) < halve_threshold * tnode_child_length(tn)) { - tn = halve(t, tn, &err); - - if (err) { + old_tn = tn; + tn = halve(t, tn); + if (IS_ERR(tn)) { + tn = old_tn; #ifdef CONFIG_IP_FIB_TRIE_STATS t->stats.resize_node_skipped++; #endif @@ -621,44 +583,37 @@ static struct node *resize(struct trie *t, struct tnode *tn) /* Only one child remains */ - if (tn->empty_children == tnode_child_length(tn) - 1) for (i = 0; i < tnode_child_length(tn); i++) { - - write_lock_bh(&fib_lock); - if (tn->child[i] != NULL) { - /* compress one level */ - struct node *n = tn->child[i]; - - if (n) - NODE_INIT_PARENT(n, NODE_TYPE(n)); - - write_unlock_bh(&fib_lock); - tnode_free(tn); - return n; - } - write_unlock_bh(&fib_lock); + struct node *n; + + n = tn->child[i]; + if (!n) + continue; + + /* compress one level */ + + NODE_SET_PARENT(n, NULL); + tnode_free(tn); + return n; } return (struct node *) tn; } -static struct tnode *inflate(struct trie *t, struct tnode *tn, int *err) +static struct tnode *inflate(struct trie *t, struct tnode *tn) { struct tnode *inode; struct tnode *oldtnode = tn; int olen = tnode_child_length(tn); int i; - if (trie_debug) - printk("In inflate\n"); + pr_debug("In inflate\n"); tn = tnode_new(oldtnode->key, oldtnode->pos, oldtnode->bits + 1); - if (!tn) { - *err = -ENOMEM; - return oldtnode; - } + if (!tn) + return ERR_PTR(-ENOMEM); /* * Preallocate and store tnodes before the actual work so we @@ -666,8 +621,8 @@ static struct tnode *inflate(struct trie *t, struct tnode *tn, int *err) * fails. In case of failure we return the oldnode and inflate * of tnode is ignored. */ - - for(i = 0; i < olen; i++) { + + for (i = 0; i < olen; i++) { struct tnode *inode = (struct tnode *) tnode_get_child(oldtnode, i); if (inode && @@ -675,46 +630,30 @@ static struct tnode *inflate(struct trie *t, struct tnode *tn, int *err) inode->pos == oldtnode->pos + oldtnode->bits && inode->bits > 1) { struct tnode *left, *right; - t_key m = TKEY_GET_MASK(inode->pos, 1); left = tnode_new(inode->key&(~m), inode->pos + 1, inode->bits - 1); + if (!left) + goto nomem; - if (!left) { - *err = -ENOMEM; - break; - } - right = tnode_new(inode->key|m, inode->pos + 1, inode->bits - 1); - if (!right) { - *err = -ENOMEM; - break; - } + if (!right) { + tnode_free(left); + goto nomem; + } put_child(t, tn, 2*i, (struct node *) left); put_child(t, tn, 2*i+1, (struct node *) right); } } - if (*err) { - int size = tnode_child_length(tn); - int j; - - for(j = 0; j < size; j++) - if (tn->child[j]) - tnode_free((struct tnode *)tn->child[j]); - - tnode_free(tn); - - *err = -ENOMEM; - return oldtnode; - } - - for(i = 0; i < olen; i++) { + for (i = 0; i < olen; i++) { struct node *node = tnode_get_child(oldtnode, i); + struct tnode *left, *right; + int size, j; /* An empty child */ if (node == NULL) @@ -740,76 +679,82 @@ static struct tnode *inflate(struct trie *t, struct tnode *tn, int *err) put_child(t, tn, 2*i+1, inode->child[1]); tnode_free(inode); + continue; } - /* An internal node with more than two children */ - else { - struct tnode *left, *right; - int size, j; - - /* We will replace this node 'inode' with two new - * ones, 'left' and 'right', each with half of the - * original children. The two new nodes will have - * a position one bit further down the key and this - * means that the "significant" part of their keys - * (see the discussion near the top of this file) - * will differ by one bit, which will be "0" in - * left's key and "1" in right's key. Since we are - * moving the key position by one step, the bit that - * we are moving away from - the bit at position - * (inode->pos) - is the one that will differ between - * left and right. So... we synthesize that bit in the - * two new keys. - * The mask 'm' below will be a single "one" bit at - * the position (inode->pos) - */ - - /* Use the old key, but set the new significant - * bit to zero. - */ + /* An internal node with more than two children */ + + /* We will replace this node 'inode' with two new + * ones, 'left' and 'right', each with half of the + * original children. The two new nodes will have + * a position one bit further down the key and this + * means that the "significant" part of their keys + * (see the discussion near the top of this file) + * will differ by one bit, which will be "0" in + * left's key and "1" in right's key. Since we are + * moving the key position by one step, the bit that + * we are moving away from - the bit at position + * (inode->pos) - is the one that will differ between + * left and right. So... we synthesize that bit in the + * two new keys. + * The mask 'm' below will be a single "one" bit at + * the position (inode->pos) + */ - left = (struct tnode *) tnode_get_child(tn, 2*i); - put_child(t, tn, 2*i, NULL); + /* Use the old key, but set the new significant + * bit to zero. + */ - if (!left) - BUG(); + left = (struct tnode *) tnode_get_child(tn, 2*i); + put_child(t, tn, 2*i, NULL); - right = (struct tnode *) tnode_get_child(tn, 2*i+1); - put_child(t, tn, 2*i+1, NULL); + BUG_ON(!left); - if (!right) - BUG(); + right = (struct tnode *) tnode_get_child(tn, 2*i+1); + put_child(t, tn, 2*i+1, NULL); - size = tnode_child_length(left); - for(j = 0; j < size; j++) { - put_child(t, left, j, inode->child[j]); - put_child(t, right, j, inode->child[j + size]); - } - put_child(t, tn, 2*i, resize(t, left)); - put_child(t, tn, 2*i+1, resize(t, right)); + BUG_ON(!right); - tnode_free(inode); + size = tnode_child_length(left); + for (j = 0; j < size; j++) { + put_child(t, left, j, inode->child[j]); + put_child(t, right, j, inode->child[j + size]); } + put_child(t, tn, 2*i, resize(t, left)); + put_child(t, tn, 2*i+1, resize(t, right)); + + tnode_free(inode); } tnode_free(oldtnode); return tn; +nomem: + { + int size = tnode_child_length(tn); + int j; + + for (j = 0; j < size; j++) + if (tn->child[j]) + tnode_free((struct tnode *)tn->child[j]); + + tnode_free(tn); + + return ERR_PTR(-ENOMEM); + } } -static struct tnode *halve(struct trie *t, struct tnode *tn, int *err) +static struct tnode *halve(struct trie *t, struct tnode *tn) { struct tnode *oldtnode = tn; struct node *left, *right; int i; int olen = tnode_child_length(tn); - if (trie_debug) printk("In halve\n"); + pr_debug("In halve\n"); tn = tnode_new(oldtnode->key, oldtnode->pos, oldtnode->bits - 1); - if (!tn) { - *err = -ENOMEM; - return oldtnode; - } + if (!tn) + return ERR_PTR(-ENOMEM); /* * Preallocate and store tnodes before the actual work so we @@ -818,38 +763,27 @@ static struct tnode *halve(struct trie *t, struct tnode *tn, int *err) * of tnode is ignored. */ - for(i = 0; i < olen; i += 2) { + for (i = 0; i < olen; i += 2) { left = tnode_get_child(oldtnode, i); right = tnode_get_child(oldtnode, i+1); /* Two nonempty children */ - if (left && right) { - struct tnode *newBinNode = - tnode_new(left->key, tn->pos + tn->bits, 1); + if (left && right) { + struct tnode *newn; - if (!newBinNode) { - *err = -ENOMEM; - break; - } - put_child(t, tn, i/2, (struct node *)newBinNode); - } - } + newn = tnode_new(left->key, tn->pos + tn->bits, 1); - if (*err) { - int size = tnode_child_length(tn); - int j; + if (!newn) + goto nomem; - for(j = 0; j < size; j++) - if (tn->child[j]) - tnode_free((struct tnode *)tn->child[j]); + put_child(t, tn, i/2, (struct node *)newn); + } - tnode_free(tn); - - *err = -ENOMEM; - return oldtnode; } - for(i = 0; i < olen; i += 2) { + for (i = 0; i < olen; i += 2) { + struct tnode *newBinNode; + left = tnode_get_child(oldtnode, i); right = tnode_get_child(oldtnode, i+1); @@ -858,88 +792,99 @@ static struct tnode *halve(struct trie *t, struct tnode *tn, int *err) if (right == NULL) /* Both are empty */ continue; put_child(t, tn, i/2, right); - } else if (right == NULL) + continue; + } + + if (right == NULL) { put_child(t, tn, i/2, left); + continue; + } /* Two nonempty children */ - else { - struct tnode *newBinNode = - (struct tnode *) tnode_get_child(tn, i/2); - put_child(t, tn, i/2, NULL); - - if (!newBinNode) - BUG(); - - put_child(t, newBinNode, 0, left); - put_child(t, newBinNode, 1, right); - put_child(t, tn, i/2, resize(t, newBinNode)); - } + newBinNode = (struct tnode *) tnode_get_child(tn, i/2); + put_child(t, tn, i/2, NULL); + put_child(t, newBinNode, 0, left); + put_child(t, newBinNode, 1, right); + put_child(t, tn, i/2, resize(t, newBinNode)); } tnode_free(oldtnode); return tn; +nomem: + { + int size = tnode_child_length(tn); + int j; + + for (j = 0; j < size; j++) + if (tn->child[j]) + tnode_free((struct tnode *)tn->child[j]); + + tnode_free(tn); + + return ERR_PTR(-ENOMEM); + } } -static void *trie_init(struct trie *t) +static void trie_init(struct trie *t) { - if (t) { - t->size = 0; - t->trie = NULL; - t->revision = 0; + if (!t) + return; + + t->size = 0; + rcu_assign_pointer(t->trie, NULL); + t->revision = 0; #ifdef CONFIG_IP_FIB_TRIE_STATS - memset(&t->stats, 0, sizeof(struct trie_use_stats)); + memset(&t->stats, 0, sizeof(struct trie_use_stats)); #endif - } - return t; } +/* readside most use rcu_read_lock currently dump routines + via get_fa_head and dump */ + static struct leaf_info *find_leaf_info(struct hlist_head *head, int plen) { struct hlist_node *node; struct leaf_info *li; - hlist_for_each_entry(li, node, head, hlist) { + hlist_for_each_entry_rcu(li, node, head, hlist) if (li->plen == plen) return li; - } + return NULL; } static inline struct list_head * get_fa_head(struct leaf *l, int plen) { - struct list_head *fa_head = NULL; struct leaf_info *li = find_leaf_info(&l->list, plen); - if (li) - fa_head = &li->falh; + if (!li) + return NULL; - return fa_head; + return &li->falh; } static void insert_leaf_info(struct hlist_head *head, struct leaf_info *new) { - struct leaf_info *li = NULL, *last = NULL; - struct hlist_node *node, *tmp; - - write_lock_bh(&fib_lock); - - if (hlist_empty(head)) - hlist_add_head(&new->hlist, head); - else { - hlist_for_each_entry_safe(li, node, tmp, head, hlist) { - - if (new->plen > li->plen) - break; - - last = li; - } - if (last) - hlist_add_after(&last->hlist, &new->hlist); - else - hlist_add_before(&new->hlist, &li->hlist); - } - write_unlock_bh(&fib_lock); + struct leaf_info *li = NULL, *last = NULL; + struct hlist_node *node; + + if (hlist_empty(head)) { + hlist_add_head_rcu(&new->hlist, head); + } else { + hlist_for_each_entry(li, node, head, hlist) { + if (new->plen > li->plen) + break; + + last = li; + } + if (last) + hlist_add_after_rcu(&last->hlist, &new->hlist); + else + hlist_add_before_rcu(&new->hlist, &li->hlist); + } } +/* rcu_read_lock needs to be hold by caller from readside */ + static struct leaf * fib_find_node(struct trie *t, u32 key) { @@ -948,61 +893,43 @@ fib_find_node(struct trie *t, u32 key) struct node *n; pos = 0; - n = t->trie; + n = rcu_dereference(t->trie); while (n != NULL && NODE_TYPE(n) == T_TNODE) { tn = (struct tnode *) n; - + check_tnode(tn); - + if (tkey_sub_equals(tn->key, pos, tn->pos-pos, key)) { - pos=tn->pos + tn->bits; + pos = tn->pos + tn->bits; n = tnode_get_child(tn, tkey_extract_bits(key, tn->pos, tn->bits)); - } - else + } else break; } /* Case we have found a leaf. Compare prefixes */ - if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key)) { - struct leaf *l = (struct leaf *) n; - return l; - } + if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key)) + return (struct leaf *)n; + return NULL; } static struct node *trie_rebalance(struct trie *t, struct tnode *tn) { - int i = 0; int wasfull; t_key cindex, key; struct tnode *tp = NULL; - if (!tn) - BUG(); - key = tn->key; - i = 0; while (tn != NULL && NODE_PARENT(tn) != NULL) { - if (i > 10) { - printk("Rebalance tn=%p \n", tn); - if (tn) printk("tn->parent=%p \n", NODE_PARENT(tn)); - - printk("Rebalance tp=%p \n", tp); - if (tp) printk("tp->parent=%p \n", NODE_PARENT(tp)); - } - - if (i > 12) BUG(); - i++; - tp = NODE_PARENT(tn); cindex = tkey_extract_bits(key, tp->pos, tp->bits); wasfull = tnode_full(tp, tnode_get_child(tp, cindex)); tn = (struct tnode *) resize (t, (struct tnode *)tn); tnode_put_child_reorg((struct tnode *)tp, cindex,(struct node*)tn, wasfull); - + if (!NODE_PARENT(tn)) break; @@ -1015,6 +942,8 @@ static struct node *trie_rebalance(struct trie *t, struct tnode *tn) return (struct node*) tn; } +/* only used from updater-side */ + static struct list_head * fib_insert_node(struct trie *t, int *err, u32 key, int plen) { @@ -1050,20 +979,16 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen) while (n != NULL && NODE_TYPE(n) == T_TNODE) { tn = (struct tnode *) n; - + check_tnode(tn); - + if (tkey_sub_equals(tn->key, pos, tn->pos-pos, key)) { tp = tn; - pos=tn->pos + tn->bits; + pos = tn->pos + tn->bits; n = tnode_get_child(tn, tkey_extract_bits(key, tn->pos, tn->bits)); - if (n && NODE_PARENT(n) != tn) { - printk("BUG tn=%p, n->parent=%p\n", tn, NODE_PARENT(n)); - BUG(); - } - } - else + BUG_ON(n && NODE_PARENT(n) != tn); + } else break; } @@ -1073,17 +998,15 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen) * tp is n's (parent) ----> NULL or TNODE */ - if (tp && IS_LEAF(tp)) - BUG(); - + BUG_ON(tp && IS_LEAF(tp)); /* Case 1: n is a leaf. Compare prefixes */ if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key)) { - struct leaf *l = ( struct leaf *) n; - + struct leaf *l = (struct leaf *) n; + li = leaf_info_new(plen); - + if (!li) { *err = -ENOMEM; goto err; @@ -1113,35 +1036,29 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen) fa_head = &li->falh; insert_leaf_info(&l->list, li); - /* Case 2: n is NULL, and will just insert a new leaf */ if (t->trie && n == NULL) { + /* Case 2: n is NULL, and will just insert a new leaf */ NODE_SET_PARENT(l, tp); - - if (!tp) - BUG(); - else { - cindex = tkey_extract_bits(key, tp->pos, tp->bits); - put_child(t, (struct tnode *)tp, cindex, (struct node *)l); - } - } - /* Case 3: n is a LEAF or a TNODE and the key doesn't match. */ - else { + cindex = tkey_extract_bits(key, tp->pos, tp->bits); + put_child(t, (struct tnode *)tp, cindex, (struct node *)l); + } else { + /* Case 3: n is a LEAF or a TNODE and the key doesn't match. */ /* * Add a new tnode here * first tnode need some special handling */ if (tp) - pos=tp->pos+tp->bits; + pos = tp->pos+tp->bits; else - pos=0; + pos = 0; + if (n) { newpos = tkey_mismatch(key, pos, n->key); tn = tnode_new(n->key, newpos, 1); - } - else { + } else { newpos = 0; tn = tnode_new(key, newpos, 1); /* First tnode */ } @@ -1151,32 +1068,33 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen) tnode_free((struct tnode *) l); *err = -ENOMEM; goto err; - } - + } + NODE_SET_PARENT(tn, tp); - missbit=tkey_extract_bits(key, newpos, 1); + missbit = tkey_extract_bits(key, newpos, 1); put_child(t, tn, missbit, (struct node *)l); put_child(t, tn, 1-missbit, n); if (tp) { cindex = tkey_extract_bits(key, tp->pos, tp->bits); put_child(t, (struct tnode *)tp, cindex, (struct node *)tn); - } - else { - t->trie = (struct node*) tn; /* First tnode */ + } else { + rcu_assign_pointer(t->trie, (struct node *)tn); /* First tnode */ tp = tn; } } - if (tp && tp->pos+tp->bits > 32) { + + if (tp && tp->pos + tp->bits > 32) printk("ERROR tp=%p pos=%d, bits=%d, key=%0x plen=%d\n", tp, tp->pos, tp->bits, key, plen); - } + /* Rebalance the trie */ - t->trie = trie_rebalance(t, tp); + + rcu_assign_pointer(t->trie, trie_rebalance(t, tp)); done: t->revision++; -err:; +err: return fa_head; } @@ -1204,17 +1122,18 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, key = ntohl(key); - if (trie_debug) - printk("Insert table=%d %08x/%d\n", tb->tb_id, key, plen); + pr_debug("Insert table=%d %08x/%d\n", tb->tb_id, key, plen); - mask = ntohl( inet_make_mask(plen) ); + mask = ntohl(inet_make_mask(plen)); if (key & ~mask) return -EINVAL; key = key & mask; - if ((fi = fib_create_info(r, rta, nlhdr, &err)) == NULL) + fi = fib_create_info(r, rta, nlhdr, &err); + + if (!fi) goto err; l = fib_find_node(t, key); @@ -1236,8 +1155,7 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, * and we need to allocate a new one of those as well. */ - if (fa && - fa->fa_info->fib_priority == fi->fib_priority) { + if (fa && fa->fa_info->fib_priority == fi->fib_priority) { struct fib_alias *fa_orig; err = -EEXIST; @@ -1248,22 +1166,27 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, struct fib_info *fi_drop; u8 state; - write_lock_bh(&fib_lock); + err = -ENOBUFS; + new_fa = kmem_cache_alloc(fn_alias_kmem, SLAB_KERNEL); + if (new_fa == NULL) + goto out; fi_drop = fa->fa_info; - fa->fa_info = fi; - fa->fa_type = type; - fa->fa_scope = r->rtm_scope; + new_fa->fa_tos = fa->fa_tos; + new_fa->fa_info = fi; + new_fa->fa_type = type; + new_fa->fa_scope = r->rtm_scope; state = fa->fa_state; - fa->fa_state &= ~FA_S_ACCESSED; + new_fa->fa_state &= ~FA_S_ACCESSED; - write_unlock_bh(&fib_lock); + list_replace_rcu(&fa->fa_list, &new_fa->fa_list); + alias_free_mem_rcu(fa); fib_release_info(fi_drop); if (state & FA_S_ACCESSED) - rt_cache_flush(-1); + rt_cache_flush(-1); - goto succeeded; + goto succeeded; } /* Error if we find a perfect match which * uses the same scope, type, and nexthop @@ -1285,7 +1208,7 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, fa = fa_orig; } err = -ENOENT; - if (!(nlhdr->nlmsg_flags&NLM_F_CREATE)) + if (!(nlhdr->nlmsg_flags & NLM_F_CREATE)) goto out; err = -ENOBUFS; @@ -1298,9 +1221,6 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, new_fa->fa_type = type; new_fa->fa_scope = r->rtm_scope; new_fa->fa_state = 0; -#if 0 - new_fa->dst = NULL; -#endif /* * Insert new entry to the list. */ @@ -1312,12 +1232,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, goto out_free_new_fa; } - write_lock_bh(&fib_lock); - - list_add_tail(&new_fa->fa_list, - (fa ? &fa->fa_list : fa_head)); - - write_unlock_bh(&fib_lock); + list_add_tail_rcu(&new_fa->fa_list, + (fa ? &fa->fa_list : fa_head)); rt_cache_flush(-1); rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, nlhdr, req); @@ -1328,11 +1244,14 @@ out_free_new_fa: kmem_cache_free(fn_alias_kmem, new_fa); out: fib_release_info(fi); -err:; +err: return err; } -static inline int check_leaf(struct trie *t, struct leaf *l, t_key key, int *plen, const struct flowi *flp, + +/* should be clalled with rcu_read_lock */ +static inline int check_leaf(struct trie *t, struct leaf *l, + t_key key, int *plen, const struct flowi *flp, struct fib_result *res) { int err, i; @@ -1341,8 +1260,7 @@ static inline int check_leaf(struct trie *t, struct leaf *l, t_key key, int *pl struct hlist_head *hhead = &l->list; struct hlist_node *node; - hlist_for_each_entry(li, node, hhead, hlist) { - + hlist_for_each_entry_rcu(li, node, hhead, hlist) { i = li->plen; mask = ntohl(inet_make_mask(i)); if (l->key != (key & mask)) @@ -1370,13 +1288,17 @@ fn_trie_lookup(struct fib_table *tb, const struct flowi *flp, struct fib_result struct node *n; struct tnode *pn; int pos, bits; - t_key key=ntohl(flp->fl4_dst); + t_key key = ntohl(flp->fl4_dst); int chopped_off; t_key cindex = 0; int current_prefix_length = KEYLENGTH; - n = t->trie; + struct tnode *cn; + t_key node_prefix, key_prefix, pref_mismatch; + int mp; + + rcu_read_lock(); - read_lock(&fib_lock); + n = rcu_dereference(t->trie); if (!n) goto failed; @@ -1393,8 +1315,7 @@ fn_trie_lookup(struct fib_table *tb, const struct flowi *flp, struct fib_result pn = (struct tnode *) n; chopped_off = 0; - while (pn) { - + while (pn) { pos = pn->pos; bits = pn->bits; @@ -1410,130 +1331,129 @@ fn_trie_lookup(struct fib_table *tb, const struct flowi *flp, struct fib_result goto backtrace; } - if (IS_TNODE(n)) { + if (IS_LEAF(n)) { + if ((ret = check_leaf(t, (struct leaf *)n, key, &plen, flp, res)) <= 0) + goto found; + else + goto backtrace; + } + #define HL_OPTIMIZE #ifdef HL_OPTIMIZE - struct tnode *cn = (struct tnode *)n; - t_key node_prefix, key_prefix, pref_mismatch; - int mp; + cn = (struct tnode *)n; - /* - * It's a tnode, and we can do some extra checks here if we - * like, to avoid descending into a dead-end branch. - * This tnode is in the parent's child array at index - * key[p_pos..p_pos+p_bits] but potentially with some bits - * chopped off, so in reality the index may be just a - * subprefix, padded with zero at the end. - * We can also take a look at any skipped bits in this - * tnode - everything up to p_pos is supposed to be ok, - * and the non-chopped bits of the index (se previous - * paragraph) are also guaranteed ok, but the rest is - * considered unknown. - * - * The skipped bits are key[pos+bits..cn->pos]. - */ - - /* If current_prefix_length < pos+bits, we are already doing - * actual prefix matching, which means everything from - * pos+(bits-chopped_off) onward must be zero along some - * branch of this subtree - otherwise there is *no* valid - * prefix present. Here we can only check the skipped - * bits. Remember, since we have already indexed into the - * parent's child array, we know that the bits we chopped of - * *are* zero. - */ + /* + * It's a tnode, and we can do some extra checks here if we + * like, to avoid descending into a dead-end branch. + * This tnode is in the parent's child array at index + * key[p_pos..p_pos+p_bits] but potentially with some bits + * chopped off, so in reality the index may be just a + * subprefix, padded with zero at the end. + * We can also take a look at any skipped bits in this + * tnode - everything up to p_pos is supposed to be ok, + * and the non-chopped bits of the index (se previous + * paragraph) are also guaranteed ok, but the rest is + * considered unknown. + * + * The skipped bits are key[pos+bits..cn->pos]. + */ - /* NOTA BENE: CHECKING ONLY SKIPPED BITS FOR THE NEW NODE HERE */ - - if (current_prefix_length < pos+bits) { - if (tkey_extract_bits(cn->key, current_prefix_length, - cn->pos - current_prefix_length) != 0 || - !(cn->child[0])) - goto backtrace; - } + /* If current_prefix_length < pos+bits, we are already doing + * actual prefix matching, which means everything from + * pos+(bits-chopped_off) onward must be zero along some + * branch of this subtree - otherwise there is *no* valid + * prefix present. Here we can only check the skipped + * bits. Remember, since we have already indexed into the + * parent's child array, we know that the bits we chopped of + * *are* zero. + */ - /* - * If chopped_off=0, the index is fully validated and we - * only need to look at the skipped bits for this, the new, - * tnode. What we actually want to do is to find out if - * these skipped bits match our key perfectly, or if we will - * have to count on finding a matching prefix further down, - * because if we do, we would like to have some way of - * verifying the existence of such a prefix at this point. - */ + /* NOTA BENE: CHECKING ONLY SKIPPED BITS FOR THE NEW NODE HERE */ - /* The only thing we can do at this point is to verify that - * any such matching prefix can indeed be a prefix to our - * key, and if the bits in the node we are inspecting that - * do not match our key are not ZERO, this cannot be true. - * Thus, find out where there is a mismatch (before cn->pos) - * and verify that all the mismatching bits are zero in the - * new tnode's key. - */ + if (current_prefix_length < pos+bits) { + if (tkey_extract_bits(cn->key, current_prefix_length, + cn->pos - current_prefix_length) != 0 || + !(cn->child[0])) + goto backtrace; + } - /* Note: We aren't very concerned about the piece of the key - * that precede pn->pos+pn->bits, since these have already been - * checked. The bits after cn->pos aren't checked since these are - * by definition "unknown" at this point. Thus, what we want to - * see is if we are about to enter the "prefix matching" state, - * and in that case verify that the skipped bits that will prevail - * throughout this subtree are zero, as they have to be if we are - * to find a matching prefix. - */ + /* + * If chopped_off=0, the index is fully validated and we + * only need to look at the skipped bits for this, the new, + * tnode. What we actually want to do is to find out if + * these skipped bits match our key perfectly, or if we will + * have to count on finding a matching prefix further down, + * because if we do, we would like to have some way of + * verifying the existence of such a prefix at this point. + */ - node_prefix = MASK_PFX(cn->key, cn->pos); - key_prefix = MASK_PFX(key, cn->pos); - pref_mismatch = key_prefix^node_prefix; - mp = 0; + /* The only thing we can do at this point is to verify that + * any such matching prefix can indeed be a prefix to our + * key, and if the bits in the node we are inspecting that + * do not match our key are not ZERO, this cannot be true. + * Thus, find out where there is a mismatch (before cn->pos) + * and verify that all the mismatching bits are zero in the + * new tnode's key. + */ - /* In short: If skipped bits in this node do not match the search - * key, enter the "prefix matching" state.directly. - */ - if (pref_mismatch) { - while (!(pref_mismatch & (1<<(KEYLENGTH-1)))) { - mp++; - pref_mismatch = pref_mismatch <<1; - } - key_prefix = tkey_extract_bits(cn->key, mp, cn->pos-mp); - - if (key_prefix != 0) - goto backtrace; - - if (current_prefix_length >= cn->pos) - current_prefix_length=mp; - } -#endif - pn = (struct tnode *)n; /* Descend */ - chopped_off = 0; - continue; + /* Note: We aren't very concerned about the piece of the key + * that precede pn->pos+pn->bits, since these have already been + * checked. The bits after cn->pos aren't checked since these are + * by definition "unknown" at this point. Thus, what we want to + * see is if we are about to enter the "prefix matching" state, + * and in that case verify that the skipped bits that will prevail + * throughout this subtree are zero, as they have to be if we are + * to find a matching prefix. + */ + + node_prefix = MASK_PFX(cn->key, cn->pos); + key_prefix = MASK_PFX(key, cn->pos); + pref_mismatch = key_prefix^node_prefix; + mp = 0; + + /* In short: If skipped bits in this node do not match the search + * key, enter the "prefix matching" state.directly. + */ + if (pref_mismatch) { + while (!(pref_mismatch & (1<<(KEYLENGTH-1)))) { + mp++; + pref_mismatch = pref_mismatch <<1; + } + key_prefix = tkey_extract_bits(cn->key, mp, cn->pos-mp); + + if (key_prefix != 0) + goto backtrace; + + if (current_prefix_length >= cn->pos) + current_prefix_length = mp; } - if (IS_LEAF(n)) { - if ((ret = check_leaf(t, (struct leaf *)n, key, &plen, flp, res)) <= 0) - goto found; - } +#endif + pn = (struct tnode *)n; /* Descend */ + chopped_off = 0; + continue; + backtrace: chopped_off++; /* As zero don't change the child key (cindex) */ - while ((chopped_off <= pn->bits) && !(cindex & (1<<(chopped_off-1)))) { + while ((chopped_off <= pn->bits) && !(cindex & (1<<(chopped_off-1)))) chopped_off++; - } /* Decrease current_... with bits chopped off */ if (current_prefix_length > pn->pos + pn->bits - chopped_off) current_prefix_length = pn->pos + pn->bits - chopped_off; - + /* * Either we do the actual chop off according or if we have * chopped off all bits in this tnode walk up to our parent. */ - if (chopped_off <= pn->bits) + if (chopped_off <= pn->bits) { cindex &= ~(1 << (chopped_off-1)); - else { + } else { if (NODE_PARENT(pn) == NULL) goto failed; - + /* Get Child's index */ cindex = tkey_extract_bits(pn->key, NODE_PARENT(pn)->pos, NODE_PARENT(pn)->bits); pn = NODE_PARENT(pn); @@ -1548,10 +1468,11 @@ backtrace: failed: ret = 1; found: - read_unlock(&fib_lock); + rcu_read_unlock(); return ret; } +/* only called from updater side */ static int trie_leaf_remove(struct trie *t, t_key key) { t_key cindex; @@ -1559,24 +1480,20 @@ static int trie_leaf_remove(struct trie *t, t_key key) struct node *n = t->trie; struct leaf *l; - if (trie_debug) - printk("entering trie_leaf_remove(%p)\n", n); + pr_debug("entering trie_leaf_remove(%p)\n", n); /* Note that in the case skipped bits, those bits are *not* checked! * When we finish this, we will have NULL or a T_LEAF, and the * T_LEAF may or may not match our key. */ - while (n != NULL && IS_TNODE(n)) { + while (n != NULL && IS_TNODE(n)) { struct tnode *tn = (struct tnode *) n; check_tnode(tn); n = tnode_get_child(tn ,tkey_extract_bits(key, tn->pos, tn->bits)); - if (n && NODE_PARENT(n) != tn) { - printk("BUG tn=%p, n->parent=%p\n", tn, NODE_PARENT(n)); - BUG(); - } - } + BUG_ON(n && NODE_PARENT(n) != tn); + } l = (struct leaf *) n; if (!n || !tkey_equals(l->key, key)) @@ -1590,23 +1507,24 @@ static int trie_leaf_remove(struct trie *t, t_key key) t->revision++; t->size--; + preempt_disable(); tp = NODE_PARENT(n); tnode_free((struct tnode *) n); if (tp) { cindex = tkey_extract_bits(key, tp->pos, tp->bits); put_child(t, (struct tnode *)tp, cindex, NULL); - t->trie = trie_rebalance(t, tp); - } - else - t->trie = NULL; + rcu_assign_pointer(t->trie, trie_rebalance(t, tp)); + } else + rcu_assign_pointer(t->trie, NULL); + preempt_enable(); return 1; } static int fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *nlhdr, struct netlink_skb_parms *req) + struct nlmsghdr *nlhdr, struct netlink_skb_parms *req) { struct trie *t = (struct trie *) tb->tb_data; u32 key, mask; @@ -1615,6 +1533,8 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, struct fib_alias *fa, *fa_to_delete; struct list_head *fa_head; struct leaf *l; + struct leaf_info *li; + if (plen > 32) return -EINVAL; @@ -1624,7 +1544,7 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, memcpy(&key, rta->rta_dst, 4); key = ntohl(key); - mask = ntohl( inet_make_mask(plen) ); + mask = ntohl(inet_make_mask(plen)); if (key & ~mask) return -EINVAL; @@ -1641,11 +1561,11 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, if (!fa) return -ESRCH; - if (trie_debug) - printk("Deleting %08x/%d tos=%d t=%p\n", key, plen, tos, t); + pr_debug("Deleting %08x/%d tos=%d t=%p\n", key, plen, tos, t); fa_to_delete = NULL; fa_head = fa->fa_list.prev; + list_for_each_entry(fa, fa_head, fa_list) { struct fib_info *fi = fa->fa_info; @@ -1664,39 +1584,31 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, } } - if (fa_to_delete) { - int kill_li = 0; - struct leaf_info *li; - - fa = fa_to_delete; - rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req); + if (!fa_to_delete) + return -ESRCH; - l = fib_find_node(t, key); - li = find_leaf_info(&l->list, plen); + fa = fa_to_delete; + rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req); - write_lock_bh(&fib_lock); + l = fib_find_node(t, key); + li = find_leaf_info(&l->list, plen); - list_del(&fa->fa_list); + list_del_rcu(&fa->fa_list); - if (list_empty(fa_head)) { - hlist_del(&li->hlist); - kill_li = 1; - } - write_unlock_bh(&fib_lock); - - if (kill_li) - free_leaf_info(li); + if (list_empty(fa_head)) { + hlist_del_rcu(&li->hlist); + free_leaf_info(li); + } - if (hlist_empty(&l->list)) - trie_leaf_remove(t, key); + if (hlist_empty(&l->list)) + trie_leaf_remove(t, key); - if (fa->fa_state & FA_S_ACCESSED) - rt_cache_flush(-1); + if (fa->fa_state & FA_S_ACCESSED) + rt_cache_flush(-1); - fn_free_alias(fa); - return 0; - } - return -ESRCH; + fib_release_info(fa->fa_info); + alias_free_mem_rcu(fa); + return 0; } static int trie_flush_list(struct trie *t, struct list_head *head) @@ -1706,14 +1618,11 @@ static int trie_flush_list(struct trie *t, struct list_head *head) list_for_each_entry_safe(fa, fa_node, head, fa_list) { struct fib_info *fi = fa->fa_info; - - if (fi && (fi->fib_flags&RTNH_F_DEAD)) { - - write_lock_bh(&fib_lock); - list_del(&fa->fa_list); - write_unlock_bh(&fib_lock); - fn_free_alias(fa); + if (fi && (fi->fib_flags & RTNH_F_DEAD)) { + list_del_rcu(&fa->fa_list); + fib_release_info(fa->fa_info); + alias_free_mem_rcu(fa); found++; } } @@ -1728,37 +1637,34 @@ static int trie_flush_leaf(struct trie *t, struct leaf *l) struct leaf_info *li = NULL; hlist_for_each_entry_safe(li, node, tmp, lih, hlist) { - found += trie_flush_list(t, &li->falh); if (list_empty(&li->falh)) { - - write_lock_bh(&fib_lock); - hlist_del(&li->hlist); - write_unlock_bh(&fib_lock); - + hlist_del_rcu(&li->hlist); free_leaf_info(li); } } return found; } +/* rcu_read_lock needs to be hold by caller from readside */ + static struct leaf *nextleaf(struct trie *t, struct leaf *thisleaf) { struct node *c = (struct node *) thisleaf; struct tnode *p; int idx; + struct node *trie = rcu_dereference(t->trie); if (c == NULL) { - if (t->trie == NULL) + if (trie == NULL) return NULL; - if (IS_LEAF(t->trie)) /* trie w. just a leaf */ - return (struct leaf *) t->trie; + if (IS_LEAF(trie)) /* trie w. just a leaf */ + return (struct leaf *) trie; - p = (struct tnode*) t->trie; /* Start */ - } - else + p = (struct tnode*) trie; /* Start */ + } else p = (struct tnode *) NODE_PARENT(c); while (p) { @@ -1771,29 +1677,31 @@ static struct leaf *nextleaf(struct trie *t, struct leaf *thisleaf) pos = 0; last = 1 << p->bits; - for(idx = pos; idx < last ; idx++) { - if (p->child[idx]) { - - /* Decend if tnode */ - - while (IS_TNODE(p->child[idx])) { - p = (struct tnode*) p->child[idx]; - idx = 0; - - /* Rightmost non-NULL branch */ - if (p && IS_TNODE(p)) - while (p->child[idx] == NULL && idx < (1 << p->bits)) idx++; - - /* Done with this tnode? */ - if (idx >= (1 << p->bits) || p->child[idx] == NULL ) - goto up; - } - return (struct leaf*) p->child[idx]; + for (idx = pos; idx < last ; idx++) { + c = rcu_dereference(p->child[idx]); + + if (!c) + continue; + + /* Decend if tnode */ + while (IS_TNODE(c)) { + p = (struct tnode *) c; + idx = 0; + + /* Rightmost non-NULL branch */ + if (p && IS_TNODE(p)) + while (!(c = rcu_dereference(p->child[idx])) + && idx < (1<<p->bits)) idx++; + + /* Done with this tnode? */ + if (idx >= (1 << p->bits) || !c) + goto up; } + return (struct leaf *) c; } up: /* No more children go up one step */ - c = (struct node*) p; + c = (struct node *) p; p = (struct tnode *) NODE_PARENT(p); } return NULL; /* Ready. Root of trie */ @@ -1807,23 +1715,24 @@ static int fn_trie_flush(struct fib_table *tb) t->revision++; - for (h=0; (l = nextleaf(t, l)) != NULL; h++) { + rcu_read_lock(); + for (h = 0; (l = nextleaf(t, l)) != NULL; h++) { found += trie_flush_leaf(t, l); if (ll && hlist_empty(&ll->list)) trie_leaf_remove(t, ll->key); ll = l; } + rcu_read_unlock(); if (ll && hlist_empty(&ll->list)) trie_leaf_remove(t, ll->key); - if (trie_debug) - printk("trie_flush found=%d\n", found); + pr_debug("trie_flush found=%d\n", found); return found; } -static int trie_last_dflt=-1; +static int trie_last_dflt = -1; static void fn_trie_select_default(struct fib_table *tb, const struct flowi *flp, struct fib_result *res) @@ -1840,7 +1749,7 @@ fn_trie_select_default(struct fib_table *tb, const struct flowi *flp, struct fib last_resort = NULL; order = -1; - read_lock(&fib_lock); + rcu_read_lock(); l = fib_find_node(t, 0); if (!l) @@ -1853,20 +1762,20 @@ fn_trie_select_default(struct fib_table *tb, const struct flowi *flp, struct fib if (list_empty(fa_head)) goto out; - list_for_each_entry(fa, fa_head, fa_list) { + list_for_each_entry_rcu(fa, fa_head, fa_list) { struct fib_info *next_fi = fa->fa_info; - + if (fa->fa_scope != res->scope || fa->fa_type != RTN_UNICAST) continue; - + if (next_fi->fib_priority > res->fi->fib_priority) break; if (!next_fi->fib_nh[0].nh_gw || next_fi->fib_nh[0].nh_scope != RT_SCOPE_LINK) continue; fa->fa_state |= FA_S_ACCESSED; - + if (fi == NULL) { if (next_fi != res->fi) break; @@ -1904,7 +1813,7 @@ fn_trie_select_default(struct fib_table *tb, const struct flowi *flp, struct fib } trie_last_dflt = last_idx; out:; - read_unlock(&fib_lock); + rcu_read_unlock(); } static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fib_table *tb, @@ -1913,12 +1822,14 @@ static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fi int i, s_i; struct fib_alias *fa; - u32 xkey=htonl(key); + u32 xkey = htonl(key); - s_i=cb->args[3]; + s_i = cb->args[3]; i = 0; - list_for_each_entry(fa, fah, fa_list) { + /* rcu_read_lock is hold by caller */ + + list_for_each_entry_rcu(fa, fah, fa_list) { if (i < s_i) { i++; continue; @@ -1946,10 +1857,10 @@ static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fi fa->fa_info, 0) < 0) { cb->args[3] = i; return -1; - } + } i++; } - cb->args[3]=i; + cb->args[3] = i; return skb->len; } @@ -1959,10 +1870,10 @@ static int fn_trie_dump_plen(struct trie *t, int plen, struct fib_table *tb, str int h, s_h; struct list_head *fa_head; struct leaf *l = NULL; - s_h=cb->args[2]; - for (h=0; (l = nextleaf(t, l)) != NULL; h++) { + s_h = cb->args[2]; + for (h = 0; (l = nextleaf(t, l)) != NULL; h++) { if (h < s_h) continue; if (h > s_h) @@ -1970,7 +1881,7 @@ static int fn_trie_dump_plen(struct trie *t, int plen, struct fib_table *tb, str sizeof(cb->args) - 3*sizeof(cb->args[0])); fa_head = get_fa_head(l, plen); - + if (!fa_head) continue; @@ -1978,11 +1889,11 @@ static int fn_trie_dump_plen(struct trie *t, int plen, struct fib_table *tb, str continue; if (fn_trie_dump_fa(l->key, plen, fa_head, tb, skb, cb)<0) { - cb->args[2]=h; + cb->args[2] = h; return -1; } } - cb->args[2]=h; + cb->args[2] = h; return skb->len; } @@ -1993,25 +1904,24 @@ static int fn_trie_dump(struct fib_table *tb, struct sk_buff *skb, struct netlin s_m = cb->args[1]; - read_lock(&fib_lock); - for (m=0; m<=32; m++) { - + rcu_read_lock(); + for (m = 0; m <= 32; m++) { if (m < s_m) continue; if (m > s_m) memset(&cb->args[2], 0, - sizeof(cb->args) - 2*sizeof(cb->args[0])); + sizeof(cb->args) - 2*sizeof(cb->args[0])); if (fn_trie_dump_plen(t, 32-m, tb, skb, cb)<0) { cb->args[1] = m; goto out; } } - read_unlock(&fib_lock); + rcu_read_unlock(); cb->args[1] = m; return skb->len; - out: - read_unlock(&fib_lock); +out: + rcu_read_unlock(); return -1; } @@ -2051,9 +1961,9 @@ struct fib_table * __init fib_hash_init(int id) trie_init(t); if (id == RT_TABLE_LOCAL) - trie_local = t; + trie_local = t; else if (id == RT_TABLE_MAIN) - trie_main = t; + trie_main = t; if (id == RT_TABLE_LOCAL) printk("IPv4 FIB: Using LC-trie version %s\n", VERSION); @@ -2065,7 +1975,8 @@ struct fib_table * __init fib_hash_init(int id) static void putspace_seq(struct seq_file *seq, int n) { - while (n--) seq_printf(seq, " "); + while (n--) + seq_printf(seq, " "); } static void printbin_seq(struct seq_file *seq, unsigned int v, int bits) @@ -2086,29 +1997,22 @@ static void printnode_seq(struct seq_file *seq, int indent, struct node *n, seq_printf(seq, "%d/", cindex); printbin_seq(seq, cindex, bits); seq_printf(seq, ": "); - } - else + } else seq_printf(seq, "<root>: "); seq_printf(seq, "%s:%p ", IS_LEAF(n)?"Leaf":"Internal node", n); - if (IS_LEAF(n)) - seq_printf(seq, "key=%d.%d.%d.%d\n", - n->key >> 24, (n->key >> 16) % 256, (n->key >> 8) % 256, n->key % 256); - else { - int plen = ((struct tnode *)n)->pos; - t_key prf=MASK_PFX(n->key, plen); - seq_printf(seq, "key=%d.%d.%d.%d/%d\n", - prf >> 24, (prf >> 16) % 256, (prf >> 8) % 256, prf % 256, plen); - } if (IS_LEAF(n)) { - struct leaf *l=(struct leaf *)n; + struct leaf *l = (struct leaf *)n; struct fib_alias *fa; int i; - for (i=32; i>=0; i--) - if (find_leaf_info(&l->list, i)) { - + + seq_printf(seq, "key=%d.%d.%d.%d\n", + n->key >> 24, (n->key >> 16) % 256, (n->key >> 8) % 256, n->key % 256); + + for (i = 32; i >= 0; i--) + if (find_leaf_info(&l->list, i)) { struct list_head *fa_head = get_fa_head(l, i); - + if (!fa_head) continue; @@ -2118,17 +2022,16 @@ static void printnode_seq(struct seq_file *seq, int indent, struct node *n, putspace_seq(seq, indent+2); seq_printf(seq, "{/%d...dumping}\n", i); - - list_for_each_entry(fa, fa_head, fa_list) { + list_for_each_entry_rcu(fa, fa_head, fa_list) { putspace_seq(seq, indent+2); - if (fa->fa_info->fib_nh == NULL) { - seq_printf(seq, "Error _fib_nh=NULL\n"); - continue; - } if (fa->fa_info == NULL) { seq_printf(seq, "Error fa_info=NULL\n"); continue; } + if (fa->fa_info->fib_nh == NULL) { + seq_printf(seq, "Error _fib_nh=NULL\n"); + continue; + } seq_printf(seq, "{type=%d scope=%d TOS=%d}\n", fa->fa_type, @@ -2136,11 +2039,16 @@ static void printnode_seq(struct seq_file *seq, int indent, struct node *n, fa->fa_tos); } } - } - else if (IS_TNODE(n)) { + } else { struct tnode *tn = (struct tnode *)n; + int plen = ((struct tnode *)n)->pos; + t_key prf = MASK_PFX(n->key, plen); + + seq_printf(seq, "key=%d.%d.%d.%d/%d\n", + prf >> 24, (prf >> 16) % 256, (prf >> 8) % 256, prf % 256, plen); + putspace_seq(seq, indent); seq_printf(seq, "| "); - seq_printf(seq, "{key prefix=%08x/", tn->key&TKEY_GET_MASK(0, tn->pos)); + seq_printf(seq, "{key prefix=%08x/", tn->key & TKEY_GET_MASK(0, tn->pos)); printbin_seq(seq, tkey_extract_bits(tn->key, 0, tn->pos), tn->pos); seq_printf(seq, "}\n"); putspace_seq(seq, indent); seq_printf(seq, "| "); @@ -2154,194 +2062,196 @@ static void printnode_seq(struct seq_file *seq, int indent, struct node *n, static void trie_dump_seq(struct seq_file *seq, struct trie *t) { - struct node *n = t->trie; - int cindex=0; - int indent=1; - int pend=0; + struct node *n; + int cindex = 0; + int indent = 1; + int pend = 0; int depth = 0; + struct tnode *tn; - read_lock(&fib_lock); - + rcu_read_lock(); + n = rcu_dereference(t->trie); seq_printf(seq, "------ trie_dump of t=%p ------\n", t); - if (n) { - printnode_seq(seq, indent, n, pend, cindex, 0); - if (IS_TNODE(n)) { - struct tnode *tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - putspace_seq(seq, indent); seq_printf(seq, "\\--\n"); - indent += 3; - depth++; - - while (tn && cindex < (1 << tn->bits)) { - if (tn->child[cindex]) { - - /* Got a child */ - - printnode_seq(seq, indent, tn->child[cindex], pend, cindex, tn->bits); - if (IS_LEAF(tn->child[cindex])) { - cindex++; - - } - else { - /* - * New tnode. Decend one level - */ - - depth++; - n = tn->child[cindex]; - tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - putspace_seq(seq, indent); seq_printf(seq, "\\--\n"); - indent+=3; - cindex=0; - } - } - else - cindex++; + if (!n) { + seq_printf(seq, "------ trie is empty\n"); + + rcu_read_unlock(); + return; + } + + printnode_seq(seq, indent, n, pend, cindex, 0); + + if (!IS_TNODE(n)) { + rcu_read_unlock(); + return; + } + + tn = (struct tnode *)n; + pend = tn->pos+tn->bits; + putspace_seq(seq, indent); seq_printf(seq, "\\--\n"); + indent += 3; + depth++; + + while (tn && cindex < (1 << tn->bits)) { + struct node *child = rcu_dereference(tn->child[cindex]); + if (!child) + cindex++; + else { + /* Got a child */ + printnode_seq(seq, indent, child, pend, + cindex, tn->bits); + + if (IS_LEAF(child)) + cindex++; + + else { /* - * Test if we are done + * New tnode. Decend one level */ - - while (cindex >= (1 << tn->bits)) { - /* - * Move upwards and test for root - * pop off all traversed nodes - */ - - if (NODE_PARENT(tn) == NULL) { - tn = NULL; - n = NULL; - break; - } - else { - cindex = tkey_extract_bits(tn->key, NODE_PARENT(tn)->pos, NODE_PARENT(tn)->bits); - tn = NODE_PARENT(tn); - cindex++; - n = (struct node *)tn; - pend = tn->pos+tn->bits; - indent-=3; - depth--; - } - } + depth++; + n = child; + tn = (struct tnode *)n; + pend = tn->pos+tn->bits; + putspace_seq(seq, indent); + seq_printf(seq, "\\--\n"); + indent += 3; + cindex = 0; } } - else n = NULL; - } - else seq_printf(seq, "------ trie is empty\n"); - read_unlock(&fib_lock); + /* + * Test if we are done + */ + + while (cindex >= (1 << tn->bits)) { + /* + * Move upwards and test for root + * pop off all traversed nodes + */ + + if (NODE_PARENT(tn) == NULL) { + tn = NULL; + break; + } + + cindex = tkey_extract_bits(tn->key, NODE_PARENT(tn)->pos, NODE_PARENT(tn)->bits); + cindex++; + tn = NODE_PARENT(tn); + pend = tn->pos + tn->bits; + indent -= 3; + depth--; + } + } + rcu_read_unlock(); } static struct trie_stat *trie_stat_new(void) { - struct trie_stat *s = kmalloc(sizeof(struct trie_stat), GFP_KERNEL); + struct trie_stat *s; int i; - if (s) { - s->totdepth = 0; - s->maxdepth = 0; - s->tnodes = 0; - s->leaves = 0; - s->nullpointers = 0; - - for(i=0; i< MAX_CHILDS; i++) - s->nodesizes[i] = 0; - } + s = kmalloc(sizeof(struct trie_stat), GFP_KERNEL); + if (!s) + return NULL; + + s->totdepth = 0; + s->maxdepth = 0; + s->tnodes = 0; + s->leaves = 0; + s->nullpointers = 0; + + for (i = 0; i < MAX_CHILDS; i++) + s->nodesizes[i] = 0; + return s; } static struct trie_stat *trie_collect_stats(struct trie *t) { - struct node *n = t->trie; + struct node *n; struct trie_stat *s = trie_stat_new(); int cindex = 0; - int indent = 1; int pend = 0; int depth = 0; - read_lock(&fib_lock); + if (!s) + return NULL; - if (s) { - if (n) { - if (IS_TNODE(n)) { - struct tnode *tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - indent += 3; - s->nodesizes[tn->bits]++; - depth++; + rcu_read_lock(); + n = rcu_dereference(t->trie); - while (tn && cindex < (1 << tn->bits)) { - if (tn->child[cindex]) { - /* Got a child */ - - if (IS_LEAF(tn->child[cindex])) { - cindex++; - - /* stats */ - if (depth > s->maxdepth) - s->maxdepth = depth; - s->totdepth += depth; - s->leaves++; - } - - else { - /* - * New tnode. Decend one level - */ - - s->tnodes++; - s->nodesizes[tn->bits]++; - depth++; - - n = tn->child[cindex]; - tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - - indent += 3; - cindex = 0; - } - } - else { - cindex++; - s->nullpointers++; - } + if (!n) + return s; + + if (IS_TNODE(n)) { + struct tnode *tn = (struct tnode *)n; + pend = tn->pos+tn->bits; + s->nodesizes[tn->bits]++; + depth++; + + while (tn && cindex < (1 << tn->bits)) { + struct node *ch = rcu_dereference(tn->child[cindex]); + if (ch) { + /* Got a child */ + + if (IS_LEAF(tn->child[cindex])) { + cindex++; + + /* stats */ + if (depth > s->maxdepth) + s->maxdepth = depth; + s->totdepth += depth; + s->leaves++; + } else { /* - * Test if we are done + * New tnode. Decend one level */ - - while (cindex >= (1 << tn->bits)) { - - /* - * Move upwards and test for root - * pop off all traversed nodes - */ - - - if (NODE_PARENT(tn) == NULL) { - tn = NULL; - n = NULL; - break; - } - else { - cindex = tkey_extract_bits(tn->key, NODE_PARENT(tn)->pos, NODE_PARENT(tn)->bits); - tn = NODE_PARENT(tn); - cindex++; - n = (struct node *)tn; - pend = tn->pos+tn->bits; - indent -= 3; - depth--; - } - } + + s->tnodes++; + s->nodesizes[tn->bits]++; + depth++; + + n = ch; + tn = (struct tnode *)n; + pend = tn->pos+tn->bits; + + cindex = 0; } + } else { + cindex++; + s->nullpointers++; } - else n = NULL; + + /* + * Test if we are done + */ + + while (cindex >= (1 << tn->bits)) { + /* + * Move upwards and test for root + * pop off all traversed nodes + */ + + if (NODE_PARENT(tn) == NULL) { + tn = NULL; + n = NULL; + break; + } + + cindex = tkey_extract_bits(tn->key, NODE_PARENT(tn)->pos, NODE_PARENT(tn)->bits); + tn = NODE_PARENT(tn); + cindex++; + n = (struct node *)tn; + pend = tn->pos+tn->bits; + depth--; + } } } - read_unlock(&fib_lock); + rcu_read_unlock(); return s; } @@ -2359,17 +2269,22 @@ static struct fib_alias *fib_triestat_get_next(struct seq_file *seq) static void *fib_triestat_seq_start(struct seq_file *seq, loff_t *pos) { - void *v = NULL; + if (!ip_fib_main_table) + return NULL; - if (ip_fib_main_table) - v = *pos ? fib_triestat_get_next(seq) : SEQ_START_TOKEN; - return v; + if (*pos) + return fib_triestat_get_next(seq); + else + return SEQ_START_TOKEN; } static void *fib_triestat_seq_next(struct seq_file *seq, void *v, loff_t *pos) { ++*pos; - return v == SEQ_START_TOKEN ? fib_triestat_get_first(seq) : fib_triestat_get_next(seq); + if (v == SEQ_START_TOKEN) + return fib_triestat_get_first(seq); + else + return fib_triestat_get_next(seq); } static void fib_triestat_seq_stop(struct seq_file *seq, void *v) @@ -2388,22 +2303,22 @@ static void collect_and_show(struct trie *t, struct seq_file *seq) { int bytes = 0; /* How many bytes are used, a ref is 4 bytes */ int i, max, pointers; - struct trie_stat *stat; + struct trie_stat *stat; int avdepth; stat = trie_collect_stats(t); - bytes=0; + bytes = 0; seq_printf(seq, "trie=%p\n", t); if (stat) { if (stat->leaves) - avdepth=stat->totdepth*100 / stat->leaves; + avdepth = stat->totdepth*100 / stat->leaves; else - avdepth=0; - seq_printf(seq, "Aver depth: %d.%02d\n", avdepth / 100, avdepth % 100 ); + avdepth = 0; + seq_printf(seq, "Aver depth: %d.%02d\n", avdepth / 100, avdepth % 100); seq_printf(seq, "Max depth: %4d\n", stat->maxdepth); - + seq_printf(seq, "Leaves: %d\n", stat->leaves); bytes += sizeof(struct leaf) * stat->leaves; seq_printf(seq, "Internal nodes: %d\n", stat->tnodes); @@ -2455,11 +2370,9 @@ static int fib_triestat_seq_show(struct seq_file *seq, void *v) if (trie_main) collect_and_show(trie_main, seq); - } - else { - snprintf(bf, sizeof(bf), - "*\t%08X\t%08X", 200, 400); - + } else { + snprintf(bf, sizeof(bf), "*\t%08X\t%08X", 200, 400); + seq_printf(seq, "%-127s\n", bf); } return 0; @@ -2520,22 +2433,27 @@ static struct fib_alias *fib_trie_get_next(struct seq_file *seq) static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos) { - void *v = NULL; + if (!ip_fib_main_table) + return NULL; - if (ip_fib_main_table) - v = *pos ? fib_trie_get_next(seq) : SEQ_START_TOKEN; - return v; + if (*pos) + return fib_trie_get_next(seq); + else + return SEQ_START_TOKEN; } static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos) { ++*pos; - return v == SEQ_START_TOKEN ? fib_trie_get_first(seq) : fib_trie_get_next(seq); + if (v == SEQ_START_TOKEN) + return fib_trie_get_first(seq); + else + return fib_trie_get_next(seq); + } static void fib_trie_seq_stop(struct seq_file *seq, void *v) { - } /* @@ -2555,9 +2473,7 @@ static int fib_trie_seq_show(struct seq_file *seq, void *v) if (trie_main) trie_dump_seq(seq, trie_main); - } - - else { + } else { snprintf(bf, sizeof(bf), "*\t%08X\t%08X", 200, 400); seq_printf(seq, "%-127s\n", bf); diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index badfc5849973..24eb56ae1b5a 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -114,7 +114,7 @@ struct icmp_bxm { /* * Statistics */ -DEFINE_SNMP_STAT(struct icmp_mib, icmp_statistics); +DEFINE_SNMP_STAT(struct icmp_mib, icmp_statistics) __read_mostly; /* An array of errno for error messages from dest unreach. */ /* RFC 1122: 3.2.2.1 States that NET_UNREACH, HOST_UNREACH and SR_FAILED MUST be considered 'transient errs'. */ @@ -627,11 +627,10 @@ static void icmp_unreach(struct sk_buff *skb) break; case ICMP_FRAG_NEEDED: if (ipv4_config.no_pmtu_disc) { - LIMIT_NETDEBUG( - printk(KERN_INFO "ICMP: %u.%u.%u.%u: " + LIMIT_NETDEBUG(KERN_INFO "ICMP: %u.%u.%u.%u: " "fragmentation needed " "and DF set.\n", - NIPQUAD(iph->daddr))); + NIPQUAD(iph->daddr)); } else { info = ip_rt_frag_needed(iph, ntohs(icmph->un.frag.mtu)); @@ -640,10 +639,9 @@ static void icmp_unreach(struct sk_buff *skb) } break; case ICMP_SR_FAILED: - LIMIT_NETDEBUG( - printk(KERN_INFO "ICMP: %u.%u.%u.%u: Source " + LIMIT_NETDEBUG(KERN_INFO "ICMP: %u.%u.%u.%u: Source " "Route Failed.\n", - NIPQUAD(iph->daddr))); + NIPQUAD(iph->daddr)); break; default: break; @@ -936,7 +934,7 @@ int icmp_rcv(struct sk_buff *skb) case CHECKSUM_HW: if (!(u16)csum_fold(skb->csum)) break; - LIMIT_NETDEBUG(printk(KERN_DEBUG "icmp v4 hw csum failure\n")); + LIMIT_NETDEBUG(KERN_DEBUG "icmp v4 hw csum failure\n"); case CHECKSUM_NONE: if ((u16)csum_fold(skb_checksum(skb, 0, skb->len, 0))) goto error; diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 5088f90835ae..44607f4767b8 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -904,7 +904,7 @@ int igmp_rcv(struct sk_buff *skb) case IGMP_MTRACE_RESP: break; default: - NETDEBUG(printk(KERN_DEBUG "New IGMP type=%d, why we do not know about it?\n", ih->type)); + NETDEBUG(KERN_DEBUG "New IGMP type=%d, why we do not know about it?\n", ih->type); } in_dev_put(in_dev); kfree_skb(skb); diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c new file mode 100644 index 000000000000..fe3c6d3d0c91 --- /dev/null +++ b/net/ipv4/inet_connection_sock.c @@ -0,0 +1,641 @@ +/* + * INET An implementation of the TCP/IP protocol suite for the LINUX + * operating system. INET is implemented using the BSD Socket + * interface as the means of communication with the user level. + * + * Support for INET connection oriented protocols. + * + * Authors: See the TCP sources + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or(at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/jhash.h> + +#include <net/inet_connection_sock.h> +#include <net/inet_hashtables.h> +#include <net/inet_timewait_sock.h> +#include <net/ip.h> +#include <net/route.h> +#include <net/tcp_states.h> +#include <net/xfrm.h> + +#ifdef INET_CSK_DEBUG +const char inet_csk_timer_bug_msg[] = "inet_csk BUG: unknown timer value\n"; +EXPORT_SYMBOL(inet_csk_timer_bug_msg); +#endif + +/* + * This array holds the first and last local port number. + * For high-usage systems, use sysctl to change this to + * 32768-61000 + */ +int sysctl_local_port_range[2] = { 1024, 4999 }; + +static inline int inet_csk_bind_conflict(struct sock *sk, struct inet_bind_bucket *tb) +{ + const u32 sk_rcv_saddr = inet_rcv_saddr(sk); + struct sock *sk2; + struct hlist_node *node; + int reuse = sk->sk_reuse; + + sk_for_each_bound(sk2, node, &tb->owners) { + if (sk != sk2 && + !inet_v6_ipv6only(sk2) && + (!sk->sk_bound_dev_if || + !sk2->sk_bound_dev_if || + sk->sk_bound_dev_if == sk2->sk_bound_dev_if)) { + if (!reuse || !sk2->sk_reuse || + sk2->sk_state == TCP_LISTEN) { + const u32 sk2_rcv_saddr = inet_rcv_saddr(sk2); + if (!sk2_rcv_saddr || !sk_rcv_saddr || + sk2_rcv_saddr == sk_rcv_saddr) + break; + } + } + } + return node != NULL; +} + +/* Obtain a reference to a local port for the given sock, + * if snum is zero it means select any available local port. + */ +int inet_csk_get_port(struct inet_hashinfo *hashinfo, + struct sock *sk, unsigned short snum) +{ + struct inet_bind_hashbucket *head; + struct hlist_node *node; + struct inet_bind_bucket *tb; + int ret; + + local_bh_disable(); + if (!snum) { + int low = sysctl_local_port_range[0]; + int high = sysctl_local_port_range[1]; + int remaining = (high - low) + 1; + int rover; + + spin_lock(&hashinfo->portalloc_lock); + if (hashinfo->port_rover < low) + rover = low; + else + rover = hashinfo->port_rover; + do { + rover++; + if (rover > high) + rover = low; + head = &hashinfo->bhash[inet_bhashfn(rover, hashinfo->bhash_size)]; + spin_lock(&head->lock); + inet_bind_bucket_for_each(tb, node, &head->chain) + if (tb->port == rover) + goto next; + break; + next: + spin_unlock(&head->lock); + } while (--remaining > 0); + hashinfo->port_rover = rover; + spin_unlock(&hashinfo->portalloc_lock); + + /* Exhausted local port range during search? It is not + * possible for us to be holding one of the bind hash + * locks if this test triggers, because if 'remaining' + * drops to zero, we broke out of the do/while loop at + * the top level, not from the 'break;' statement. + */ + ret = 1; + if (remaining <= 0) + goto fail; + + /* OK, here is the one we will use. HEAD is + * non-NULL and we hold it's mutex. + */ + snum = rover; + } else { + head = &hashinfo->bhash[inet_bhashfn(snum, hashinfo->bhash_size)]; + spin_lock(&head->lock); + inet_bind_bucket_for_each(tb, node, &head->chain) + if (tb->port == snum) + goto tb_found; + } + tb = NULL; + goto tb_not_found; +tb_found: + if (!hlist_empty(&tb->owners)) { + if (sk->sk_reuse > 1) + goto success; + if (tb->fastreuse > 0 && + sk->sk_reuse && sk->sk_state != TCP_LISTEN) { + goto success; + } else { + ret = 1; + if (inet_csk_bind_conflict(sk, tb)) + goto fail_unlock; + } + } +tb_not_found: + ret = 1; + if (!tb && (tb = inet_bind_bucket_create(hashinfo->bind_bucket_cachep, head, snum)) == NULL) + goto fail_unlock; + if (hlist_empty(&tb->owners)) { + if (sk->sk_reuse && sk->sk_state != TCP_LISTEN) + tb->fastreuse = 1; + else + tb->fastreuse = 0; + } else if (tb->fastreuse && + (!sk->sk_reuse || sk->sk_state == TCP_LISTEN)) + tb->fastreuse = 0; +success: + if (!inet_csk(sk)->icsk_bind_hash) + inet_bind_hash(sk, tb, snum); + BUG_TRAP(inet_csk(sk)->icsk_bind_hash == tb); + ret = 0; + +fail_unlock: + spin_unlock(&head->lock); +fail: + local_bh_enable(); + return ret; +} + +EXPORT_SYMBOL_GPL(inet_csk_get_port); + +/* + * Wait for an incoming connection, avoid race conditions. This must be called + * with the socket locked. + */ +static int inet_csk_wait_for_connect(struct sock *sk, long timeo) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + DEFINE_WAIT(wait); + int err; + + /* + * True wake-one mechanism for incoming connections: only + * one process gets woken up, not the 'whole herd'. + * Since we do not 'race & poll' for established sockets + * anymore, the common case will execute the loop only once. + * + * Subtle issue: "add_wait_queue_exclusive()" will be added + * after any current non-exclusive waiters, and we know that + * it will always _stay_ after any new non-exclusive waiters + * because all non-exclusive waiters are added at the + * beginning of the wait-queue. As such, it's ok to "drop" + * our exclusiveness temporarily when we get woken up without + * having to remove and re-insert us on the wait queue. + */ + for (;;) { + prepare_to_wait_exclusive(sk->sk_sleep, &wait, + TASK_INTERRUPTIBLE); + release_sock(sk); + if (reqsk_queue_empty(&icsk->icsk_accept_queue)) + timeo = schedule_timeout(timeo); + lock_sock(sk); + err = 0; + if (!reqsk_queue_empty(&icsk->icsk_accept_queue)) + break; + err = -EINVAL; + if (sk->sk_state != TCP_LISTEN) + break; + err = sock_intr_errno(timeo); + if (signal_pending(current)) + break; + err = -EAGAIN; + if (!timeo) + break; + } + finish_wait(sk->sk_sleep, &wait); + return err; +} + +/* + * This will accept the next outstanding connection. + */ +struct sock *inet_csk_accept(struct sock *sk, int flags, int *err) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + struct sock *newsk; + int error; + + lock_sock(sk); + + /* We need to make sure that this socket is listening, + * and that it has something pending. + */ + error = -EINVAL; + if (sk->sk_state != TCP_LISTEN) + goto out_err; + + /* Find already established connection */ + if (reqsk_queue_empty(&icsk->icsk_accept_queue)) { + long timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + + /* If this is a non blocking socket don't sleep */ + error = -EAGAIN; + if (!timeo) + goto out_err; + + error = inet_csk_wait_for_connect(sk, timeo); + if (error) + goto out_err; + } + + newsk = reqsk_queue_get_child(&icsk->icsk_accept_queue, sk); + BUG_TRAP(newsk->sk_state != TCP_SYN_RECV); +out: + release_sock(sk); + return newsk; +out_err: + newsk = NULL; + *err = error; + goto out; +} + +EXPORT_SYMBOL(inet_csk_accept); + +/* + * Using different timers for retransmit, delayed acks and probes + * We may wish use just one timer maintaining a list of expire jiffies + * to optimize. + */ +void inet_csk_init_xmit_timers(struct sock *sk, + void (*retransmit_handler)(unsigned long), + void (*delack_handler)(unsigned long), + void (*keepalive_handler)(unsigned long)) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + + init_timer(&icsk->icsk_retransmit_timer); + init_timer(&icsk->icsk_delack_timer); + init_timer(&sk->sk_timer); + + icsk->icsk_retransmit_timer.function = retransmit_handler; + icsk->icsk_delack_timer.function = delack_handler; + sk->sk_timer.function = keepalive_handler; + + icsk->icsk_retransmit_timer.data = + icsk->icsk_delack_timer.data = + sk->sk_timer.data = (unsigned long)sk; + + icsk->icsk_pending = icsk->icsk_ack.pending = 0; +} + +EXPORT_SYMBOL(inet_csk_init_xmit_timers); + +void inet_csk_clear_xmit_timers(struct sock *sk) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + + icsk->icsk_pending = icsk->icsk_ack.pending = icsk->icsk_ack.blocked = 0; + + sk_stop_timer(sk, &icsk->icsk_retransmit_timer); + sk_stop_timer(sk, &icsk->icsk_delack_timer); + sk_stop_timer(sk, &sk->sk_timer); +} + +EXPORT_SYMBOL(inet_csk_clear_xmit_timers); + +void inet_csk_delete_keepalive_timer(struct sock *sk) +{ + sk_stop_timer(sk, &sk->sk_timer); +} + +EXPORT_SYMBOL(inet_csk_delete_keepalive_timer); + +void inet_csk_reset_keepalive_timer(struct sock *sk, unsigned long len) +{ + sk_reset_timer(sk, &sk->sk_timer, jiffies + len); +} + +EXPORT_SYMBOL(inet_csk_reset_keepalive_timer); + +struct dst_entry* inet_csk_route_req(struct sock *sk, + const struct request_sock *req) +{ + struct rtable *rt; + const struct inet_request_sock *ireq = inet_rsk(req); + struct ip_options *opt = inet_rsk(req)->opt; + struct flowi fl = { .oif = sk->sk_bound_dev_if, + .nl_u = { .ip4_u = + { .daddr = ((opt && opt->srr) ? + opt->faddr : + ireq->rmt_addr), + .saddr = ireq->loc_addr, + .tos = RT_CONN_FLAGS(sk) } }, + .proto = sk->sk_protocol, + .uli_u = { .ports = + { .sport = inet_sk(sk)->sport, + .dport = ireq->rmt_port } } }; + + if (ip_route_output_flow(&rt, &fl, sk, 0)) { + IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES); + return NULL; + } + if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway) { + ip_rt_put(rt); + IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES); + return NULL; + } + return &rt->u.dst; +} + +EXPORT_SYMBOL_GPL(inet_csk_route_req); + +static inline u32 inet_synq_hash(const u32 raddr, const u16 rport, + const u32 rnd, const u16 synq_hsize) +{ + return jhash_2words(raddr, (u32)rport, rnd) & (synq_hsize - 1); +} + +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) +#define AF_INET_FAMILY(fam) ((fam) == AF_INET) +#else +#define AF_INET_FAMILY(fam) 1 +#endif + +struct request_sock *inet_csk_search_req(const struct sock *sk, + struct request_sock ***prevp, + const __u16 rport, const __u32 raddr, + const __u32 laddr) +{ + const struct inet_connection_sock *icsk = inet_csk(sk); + struct listen_sock *lopt = icsk->icsk_accept_queue.listen_opt; + struct request_sock *req, **prev; + + for (prev = &lopt->syn_table[inet_synq_hash(raddr, rport, lopt->hash_rnd, + lopt->nr_table_entries)]; + (req = *prev) != NULL; + prev = &req->dl_next) { + const struct inet_request_sock *ireq = inet_rsk(req); + + if (ireq->rmt_port == rport && + ireq->rmt_addr == raddr && + ireq->loc_addr == laddr && + AF_INET_FAMILY(req->rsk_ops->family)) { + BUG_TRAP(!req->sk); + *prevp = prev; + break; + } + } + + return req; +} + +EXPORT_SYMBOL_GPL(inet_csk_search_req); + +void inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, + const unsigned timeout) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + struct listen_sock *lopt = icsk->icsk_accept_queue.listen_opt; + const u32 h = inet_synq_hash(inet_rsk(req)->rmt_addr, inet_rsk(req)->rmt_port, + lopt->hash_rnd, lopt->nr_table_entries); + + reqsk_queue_hash_req(&icsk->icsk_accept_queue, h, req, timeout); + inet_csk_reqsk_queue_added(sk, timeout); +} + +/* Only thing we need from tcp.h */ +extern int sysctl_tcp_synack_retries; + +EXPORT_SYMBOL_GPL(inet_csk_reqsk_queue_hash_add); + +void inet_csk_reqsk_queue_prune(struct sock *parent, + const unsigned long interval, + const unsigned long timeout, + const unsigned long max_rto) +{ + struct inet_connection_sock *icsk = inet_csk(parent); + struct request_sock_queue *queue = &icsk->icsk_accept_queue; + struct listen_sock *lopt = queue->listen_opt; + int max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries; + int thresh = max_retries; + unsigned long now = jiffies; + struct request_sock **reqp, *req; + int i, budget; + + if (lopt == NULL || lopt->qlen == 0) + return; + + /* Normally all the openreqs are young and become mature + * (i.e. converted to established socket) for first timeout. + * If synack was not acknowledged for 3 seconds, it means + * one of the following things: synack was lost, ack was lost, + * rtt is high or nobody planned to ack (i.e. synflood). + * When server is a bit loaded, queue is populated with old + * open requests, reducing effective size of queue. + * When server is well loaded, queue size reduces to zero + * after several minutes of work. It is not synflood, + * it is normal operation. The solution is pruning + * too old entries overriding normal timeout, when + * situation becomes dangerous. + * + * Essentially, we reserve half of room for young + * embrions; and abort old ones without pity, if old + * ones are about to clog our table. + */ + if (lopt->qlen>>(lopt->max_qlen_log-1)) { + int young = (lopt->qlen_young<<1); + + while (thresh > 2) { + if (lopt->qlen < young) + break; + thresh--; + young <<= 1; + } + } + + if (queue->rskq_defer_accept) + max_retries = queue->rskq_defer_accept; + + budget = 2 * (lopt->nr_table_entries / (timeout / interval)); + i = lopt->clock_hand; + + do { + reqp=&lopt->syn_table[i]; + while ((req = *reqp) != NULL) { + if (time_after_eq(now, req->expires)) { + if ((req->retrans < thresh || + (inet_rsk(req)->acked && req->retrans < max_retries)) + && !req->rsk_ops->rtx_syn_ack(parent, req, NULL)) { + unsigned long timeo; + + if (req->retrans++ == 0) + lopt->qlen_young--; + timeo = min((timeout << req->retrans), max_rto); + req->expires = now + timeo; + reqp = &req->dl_next; + continue; + } + + /* Drop this request */ + inet_csk_reqsk_queue_unlink(parent, req, reqp); + reqsk_queue_removed(queue, req); + reqsk_free(req); + continue; + } + reqp = &req->dl_next; + } + + i = (i + 1) & (lopt->nr_table_entries - 1); + + } while (--budget > 0); + + lopt->clock_hand = i; + + if (lopt->qlen) + inet_csk_reset_keepalive_timer(parent, interval); +} + +EXPORT_SYMBOL_GPL(inet_csk_reqsk_queue_prune); + +struct sock *inet_csk_clone(struct sock *sk, const struct request_sock *req, + const unsigned int __nocast priority) +{ + struct sock *newsk = sk_clone(sk, priority); + + if (newsk != NULL) { + struct inet_connection_sock *newicsk = inet_csk(newsk); + + newsk->sk_state = TCP_SYN_RECV; + newicsk->icsk_bind_hash = NULL; + + inet_sk(newsk)->dport = inet_rsk(req)->rmt_port; + newsk->sk_write_space = sk_stream_write_space; + + newicsk->icsk_retransmits = 0; + newicsk->icsk_backoff = 0; + newicsk->icsk_probes_out = 0; + + /* Deinitialize accept_queue to trap illegal accesses. */ + memset(&newicsk->icsk_accept_queue, 0, sizeof(newicsk->icsk_accept_queue)); + } + return newsk; +} + +EXPORT_SYMBOL_GPL(inet_csk_clone); + +/* + * At this point, there should be no process reference to this + * socket, and thus no user references at all. Therefore we + * can assume the socket waitqueue is inactive and nobody will + * try to jump onto it. + */ +void inet_csk_destroy_sock(struct sock *sk) +{ + BUG_TRAP(sk->sk_state == TCP_CLOSE); + BUG_TRAP(sock_flag(sk, SOCK_DEAD)); + + /* It cannot be in hash table! */ + BUG_TRAP(sk_unhashed(sk)); + + /* If it has not 0 inet_sk(sk)->num, it must be bound */ + BUG_TRAP(!inet_sk(sk)->num || inet_csk(sk)->icsk_bind_hash); + + sk->sk_prot->destroy(sk); + + sk_stream_kill_queues(sk); + + xfrm_sk_free_policy(sk); + + sk_refcnt_debug_release(sk); + + atomic_dec(sk->sk_prot->orphan_count); + sock_put(sk); +} + +EXPORT_SYMBOL(inet_csk_destroy_sock); + +int inet_csk_listen_start(struct sock *sk, const int nr_table_entries) +{ + struct inet_sock *inet = inet_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); + int rc = reqsk_queue_alloc(&icsk->icsk_accept_queue, nr_table_entries); + + if (rc != 0) + return rc; + + sk->sk_max_ack_backlog = 0; + sk->sk_ack_backlog = 0; + inet_csk_delack_init(sk); + + /* There is race window here: we announce ourselves listening, + * but this transition is still not validated by get_port(). + * It is OK, because this socket enters to hash table only + * after validation is complete. + */ + sk->sk_state = TCP_LISTEN; + if (!sk->sk_prot->get_port(sk, inet->num)) { + inet->sport = htons(inet->num); + + sk_dst_reset(sk); + sk->sk_prot->hash(sk); + + return 0; + } + + sk->sk_state = TCP_CLOSE; + __reqsk_queue_destroy(&icsk->icsk_accept_queue); + return -EADDRINUSE; +} + +EXPORT_SYMBOL_GPL(inet_csk_listen_start); + +/* + * This routine closes sockets which have been at least partially + * opened, but not yet accepted. + */ +void inet_csk_listen_stop(struct sock *sk) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + struct request_sock *acc_req; + struct request_sock *req; + + inet_csk_delete_keepalive_timer(sk); + + /* make all the listen_opt local to us */ + acc_req = reqsk_queue_yank_acceptq(&icsk->icsk_accept_queue); + + /* Following specs, it would be better either to send FIN + * (and enter FIN-WAIT-1, it is normal close) + * or to send active reset (abort). + * Certainly, it is pretty dangerous while synflood, but it is + * bad justification for our negligence 8) + * To be honest, we are not able to make either + * of the variants now. --ANK + */ + reqsk_queue_destroy(&icsk->icsk_accept_queue); + + while ((req = acc_req) != NULL) { + struct sock *child = req->sk; + + acc_req = req->dl_next; + + local_bh_disable(); + bh_lock_sock(child); + BUG_TRAP(!sock_owned_by_user(child)); + sock_hold(child); + + sk->sk_prot->disconnect(child, O_NONBLOCK); + + sock_orphan(child); + + atomic_inc(sk->sk_prot->orphan_count); + + inet_csk_destroy_sock(child); + + bh_unlock_sock(child); + local_bh_enable(); + sock_put(child); + + sk_acceptq_removed(sk); + __reqsk_free(req); + } + BUG_TRAP(!sk->sk_ack_backlog); +} + +EXPORT_SYMBOL_GPL(inet_csk_listen_stop); diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c new file mode 100644 index 000000000000..71f3c7350c6e --- /dev/null +++ b/net/ipv4/inet_diag.c @@ -0,0 +1,868 @@ +/* + * inet_diag.c Module for monitoring INET transport protocols sockets. + * + * Version: $Id: inet_diag.c,v 1.3 2002/02/01 22:01:04 davem Exp $ + * + * Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/types.h> +#include <linux/fcntl.h> +#include <linux/random.h> +#include <linux/cache.h> +#include <linux/init.h> +#include <linux/time.h> + +#include <net/icmp.h> +#include <net/tcp.h> +#include <net/ipv6.h> +#include <net/inet_common.h> +#include <net/inet_connection_sock.h> +#include <net/inet_hashtables.h> +#include <net/inet_timewait_sock.h> +#include <net/inet6_hashtables.h> + +#include <linux/inet.h> +#include <linux/stddef.h> + +#include <linux/inet_diag.h> + +static const struct inet_diag_handler **inet_diag_table; + +struct inet_diag_entry { + u32 *saddr; + u32 *daddr; + u16 sport; + u16 dport; + u16 family; + u16 userlocks; +}; + +static struct sock *idiagnl; + +#define INET_DIAG_PUT(skb, attrtype, attrlen) \ + RTA_DATA(__RTA_PUT(skb, attrtype, attrlen)) + +static int inet_diag_fill(struct sk_buff *skb, struct sock *sk, + int ext, u32 pid, u32 seq, u16 nlmsg_flags, + const struct nlmsghdr *unlh) +{ + const struct inet_sock *inet = inet_sk(sk); + const struct inet_connection_sock *icsk = inet_csk(sk); + struct inet_diag_msg *r; + struct nlmsghdr *nlh; + void *info = NULL; + struct inet_diag_meminfo *minfo = NULL; + unsigned char *b = skb->tail; + const struct inet_diag_handler *handler; + + handler = inet_diag_table[unlh->nlmsg_type]; + BUG_ON(handler == NULL); + + nlh = NLMSG_PUT(skb, pid, seq, unlh->nlmsg_type, sizeof(*r)); + nlh->nlmsg_flags = nlmsg_flags; + + r = NLMSG_DATA(nlh); + if (sk->sk_state != TCP_TIME_WAIT) { + if (ext & (1 << (INET_DIAG_MEMINFO - 1))) + minfo = INET_DIAG_PUT(skb, INET_DIAG_MEMINFO, + sizeof(*minfo)); + if (ext & (1 << (INET_DIAG_INFO - 1))) + info = INET_DIAG_PUT(skb, INET_DIAG_INFO, + handler->idiag_info_size); + + if ((ext & (1 << (INET_DIAG_CONG - 1))) && icsk->icsk_ca_ops) { + size_t len = strlen(icsk->icsk_ca_ops->name); + strcpy(INET_DIAG_PUT(skb, INET_DIAG_CONG, len + 1), + icsk->icsk_ca_ops->name); + } + } + r->idiag_family = sk->sk_family; + r->idiag_state = sk->sk_state; + r->idiag_timer = 0; + r->idiag_retrans = 0; + + r->id.idiag_if = sk->sk_bound_dev_if; + r->id.idiag_cookie[0] = (u32)(unsigned long)sk; + r->id.idiag_cookie[1] = (u32)(((unsigned long)sk >> 31) >> 1); + + if (r->idiag_state == TCP_TIME_WAIT) { + const struct inet_timewait_sock *tw = inet_twsk(sk); + long tmo = tw->tw_ttd - jiffies; + if (tmo < 0) + tmo = 0; + + r->id.idiag_sport = tw->tw_sport; + r->id.idiag_dport = tw->tw_dport; + r->id.idiag_src[0] = tw->tw_rcv_saddr; + r->id.idiag_dst[0] = tw->tw_daddr; + r->idiag_state = tw->tw_substate; + r->idiag_timer = 3; + r->idiag_expires = (tmo * 1000 + HZ - 1) / HZ; + r->idiag_rqueue = 0; + r->idiag_wqueue = 0; + r->idiag_uid = 0; + r->idiag_inode = 0; +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + if (r->idiag_family == AF_INET6) { + const struct tcp6_timewait_sock *tcp6tw = tcp6_twsk(sk); + + ipv6_addr_copy((struct in6_addr *)r->id.idiag_src, + &tcp6tw->tw_v6_rcv_saddr); + ipv6_addr_copy((struct in6_addr *)r->id.idiag_dst, + &tcp6tw->tw_v6_daddr); + } +#endif + nlh->nlmsg_len = skb->tail - b; + return skb->len; + } + + r->id.idiag_sport = inet->sport; + r->id.idiag_dport = inet->dport; + r->id.idiag_src[0] = inet->rcv_saddr; + r->id.idiag_dst[0] = inet->daddr; + +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + if (r->idiag_family == AF_INET6) { + struct ipv6_pinfo *np = inet6_sk(sk); + + ipv6_addr_copy((struct in6_addr *)r->id.idiag_src, + &np->rcv_saddr); + ipv6_addr_copy((struct in6_addr *)r->id.idiag_dst, + &np->daddr); + } +#endif + +#define EXPIRES_IN_MS(tmo) ((tmo - jiffies) * 1000 + HZ - 1) / HZ + + if (icsk->icsk_pending == ICSK_TIME_RETRANS) { + r->idiag_timer = 1; + r->idiag_retrans = icsk->icsk_retransmits; + r->idiag_expires = EXPIRES_IN_MS(icsk->icsk_timeout); + } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { + r->idiag_timer = 4; + r->idiag_retrans = icsk->icsk_probes_out; + r->idiag_expires = EXPIRES_IN_MS(icsk->icsk_timeout); + } else if (timer_pending(&sk->sk_timer)) { + r->idiag_timer = 2; + r->idiag_retrans = icsk->icsk_probes_out; + r->idiag_expires = EXPIRES_IN_MS(sk->sk_timer.expires); + } else { + r->idiag_timer = 0; + r->idiag_expires = 0; + } +#undef EXPIRES_IN_MS + + r->idiag_uid = sock_i_uid(sk); + r->idiag_inode = sock_i_ino(sk); + + if (minfo) { + minfo->idiag_rmem = atomic_read(&sk->sk_rmem_alloc); + minfo->idiag_wmem = sk->sk_wmem_queued; + minfo->idiag_fmem = sk->sk_forward_alloc; + minfo->idiag_tmem = atomic_read(&sk->sk_wmem_alloc); + } + + handler->idiag_get_info(sk, r, info); + + if (sk->sk_state < TCP_TIME_WAIT && + icsk->icsk_ca_ops && icsk->icsk_ca_ops->get_info) + icsk->icsk_ca_ops->get_info(sk, ext, skb); + + nlh->nlmsg_len = skb->tail - b; + return skb->len; + +rtattr_failure: +nlmsg_failure: + skb_trim(skb, b - skb->data); + return -1; +} + +static int inet_diag_get_exact(struct sk_buff *in_skb, const struct nlmsghdr *nlh) +{ + int err; + struct sock *sk; + struct inet_diag_req *req = NLMSG_DATA(nlh); + struct sk_buff *rep; + struct inet_hashinfo *hashinfo; + const struct inet_diag_handler *handler; + + handler = inet_diag_table[nlh->nlmsg_type]; + BUG_ON(handler == NULL); + hashinfo = handler->idiag_hashinfo; + + if (req->idiag_family == AF_INET) { + sk = inet_lookup(hashinfo, req->id.idiag_dst[0], + req->id.idiag_dport, req->id.idiag_src[0], + req->id.idiag_sport, req->id.idiag_if); + } +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + else if (req->idiag_family == AF_INET6) { + sk = inet6_lookup(hashinfo, + (struct in6_addr *)req->id.idiag_dst, + req->id.idiag_dport, + (struct in6_addr *)req->id.idiag_src, + req->id.idiag_sport, + req->id.idiag_if); + } +#endif + else { + return -EINVAL; + } + + if (sk == NULL) + return -ENOENT; + + err = -ESTALE; + if ((req->id.idiag_cookie[0] != INET_DIAG_NOCOOKIE || + req->id.idiag_cookie[1] != INET_DIAG_NOCOOKIE) && + ((u32)(unsigned long)sk != req->id.idiag_cookie[0] || + (u32)((((unsigned long)sk) >> 31) >> 1) != req->id.idiag_cookie[1])) + goto out; + + err = -ENOMEM; + rep = alloc_skb(NLMSG_SPACE((sizeof(struct inet_diag_msg) + + sizeof(struct inet_diag_meminfo) + + handler->idiag_info_size + 64)), + GFP_KERNEL); + if (!rep) + goto out; + + if (inet_diag_fill(rep, sk, req->idiag_ext, + NETLINK_CB(in_skb).pid, + nlh->nlmsg_seq, 0, nlh) <= 0) + BUG(); + + err = netlink_unicast(idiagnl, rep, NETLINK_CB(in_skb).pid, + MSG_DONTWAIT); + if (err > 0) + err = 0; + +out: + if (sk) { + if (sk->sk_state == TCP_TIME_WAIT) + inet_twsk_put((struct inet_timewait_sock *)sk); + else + sock_put(sk); + } + return err; +} + +static int bitstring_match(const u32 *a1, const u32 *a2, int bits) +{ + int words = bits >> 5; + + bits &= 0x1f; + + if (words) { + if (memcmp(a1, a2, words << 2)) + return 0; + } + if (bits) { + __u32 w1, w2; + __u32 mask; + + w1 = a1[words]; + w2 = a2[words]; + + mask = htonl((0xffffffff) << (32 - bits)); + + if ((w1 ^ w2) & mask) + return 0; + } + + return 1; +} + + +static int inet_diag_bc_run(const void *bc, int len, + const struct inet_diag_entry *entry) +{ + while (len > 0) { + int yes = 1; + const struct inet_diag_bc_op *op = bc; + + switch (op->code) { + case INET_DIAG_BC_NOP: + break; + case INET_DIAG_BC_JMP: + yes = 0; + break; + case INET_DIAG_BC_S_GE: + yes = entry->sport >= op[1].no; + break; + case INET_DIAG_BC_S_LE: + yes = entry->dport <= op[1].no; + break; + case INET_DIAG_BC_D_GE: + yes = entry->dport >= op[1].no; + break; + case INET_DIAG_BC_D_LE: + yes = entry->dport <= op[1].no; + break; + case INET_DIAG_BC_AUTO: + yes = !(entry->userlocks & SOCK_BINDPORT_LOCK); + break; + case INET_DIAG_BC_S_COND: + case INET_DIAG_BC_D_COND: { + struct inet_diag_hostcond *cond; + u32 *addr; + + cond = (struct inet_diag_hostcond *)(op + 1); + if (cond->port != -1 && + cond->port != (op->code == INET_DIAG_BC_S_COND ? + entry->sport : entry->dport)) { + yes = 0; + break; + } + + if (cond->prefix_len == 0) + break; + + if (op->code == INET_DIAG_BC_S_COND) + addr = entry->saddr; + else + addr = entry->daddr; + + if (bitstring_match(addr, cond->addr, cond->prefix_len)) + break; + if (entry->family == AF_INET6 && + cond->family == AF_INET) { + if (addr[0] == 0 && addr[1] == 0 && + addr[2] == htonl(0xffff) && + bitstring_match(addr + 3, cond->addr, + cond->prefix_len)) + break; + } + yes = 0; + break; + } + } + + if (yes) { + len -= op->yes; + bc += op->yes; + } else { + len -= op->no; + bc += op->no; + } + } + return (len == 0); +} + +static int valid_cc(const void *bc, int len, int cc) +{ + while (len >= 0) { + const struct inet_diag_bc_op *op = bc; + + if (cc > len) + return 0; + if (cc == len) + return 1; + if (op->yes < 4) + return 0; + len -= op->yes; + bc += op->yes; + } + return 0; +} + +static int inet_diag_bc_audit(const void *bytecode, int bytecode_len) +{ + const unsigned char *bc = bytecode; + int len = bytecode_len; + + while (len > 0) { + struct inet_diag_bc_op *op = (struct inet_diag_bc_op *)bc; + +//printk("BC: %d %d %d {%d} / %d\n", op->code, op->yes, op->no, op[1].no, len); + switch (op->code) { + case INET_DIAG_BC_AUTO: + case INET_DIAG_BC_S_COND: + case INET_DIAG_BC_D_COND: + case INET_DIAG_BC_S_GE: + case INET_DIAG_BC_S_LE: + case INET_DIAG_BC_D_GE: + case INET_DIAG_BC_D_LE: + if (op->yes < 4 || op->yes > len + 4) + return -EINVAL; + case INET_DIAG_BC_JMP: + if (op->no < 4 || op->no > len + 4) + return -EINVAL; + if (op->no < len && + !valid_cc(bytecode, bytecode_len, len - op->no)) + return -EINVAL; + break; + case INET_DIAG_BC_NOP: + if (op->yes < 4 || op->yes > len + 4) + return -EINVAL; + break; + default: + return -EINVAL; + } + bc += op->yes; + len -= op->yes; + } + return len == 0 ? 0 : -EINVAL; +} + +static int inet_diag_dump_sock(struct sk_buff *skb, struct sock *sk, + struct netlink_callback *cb) +{ + struct inet_diag_req *r = NLMSG_DATA(cb->nlh); + + if (cb->nlh->nlmsg_len > 4 + NLMSG_SPACE(sizeof(*r))) { + struct inet_diag_entry entry; + struct rtattr *bc = (struct rtattr *)(r + 1); + struct inet_sock *inet = inet_sk(sk); + + entry.family = sk->sk_family; +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + if (entry.family == AF_INET6) { + struct ipv6_pinfo *np = inet6_sk(sk); + + entry.saddr = np->rcv_saddr.s6_addr32; + entry.daddr = np->daddr.s6_addr32; + } else +#endif + { + entry.saddr = &inet->rcv_saddr; + entry.daddr = &inet->daddr; + } + entry.sport = inet->num; + entry.dport = ntohs(inet->dport); + entry.userlocks = sk->sk_userlocks; + + if (!inet_diag_bc_run(RTA_DATA(bc), RTA_PAYLOAD(bc), &entry)) + return 0; + } + + return inet_diag_fill(skb, sk, r->idiag_ext, NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, NLM_F_MULTI, cb->nlh); +} + +static int inet_diag_fill_req(struct sk_buff *skb, struct sock *sk, + struct request_sock *req, + u32 pid, u32 seq, + const struct nlmsghdr *unlh) +{ + const struct inet_request_sock *ireq = inet_rsk(req); + struct inet_sock *inet = inet_sk(sk); + unsigned char *b = skb->tail; + struct inet_diag_msg *r; + struct nlmsghdr *nlh; + long tmo; + + nlh = NLMSG_PUT(skb, pid, seq, unlh->nlmsg_type, sizeof(*r)); + nlh->nlmsg_flags = NLM_F_MULTI; + r = NLMSG_DATA(nlh); + + r->idiag_family = sk->sk_family; + r->idiag_state = TCP_SYN_RECV; + r->idiag_timer = 1; + r->idiag_retrans = req->retrans; + + r->id.idiag_if = sk->sk_bound_dev_if; + r->id.idiag_cookie[0] = (u32)(unsigned long)req; + r->id.idiag_cookie[1] = (u32)(((unsigned long)req >> 31) >> 1); + + tmo = req->expires - jiffies; + if (tmo < 0) + tmo = 0; + + r->id.idiag_sport = inet->sport; + r->id.idiag_dport = ireq->rmt_port; + r->id.idiag_src[0] = ireq->loc_addr; + r->id.idiag_dst[0] = ireq->rmt_addr; + r->idiag_expires = jiffies_to_msecs(tmo); + r->idiag_rqueue = 0; + r->idiag_wqueue = 0; + r->idiag_uid = sock_i_uid(sk); + r->idiag_inode = 0; +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + if (r->idiag_family == AF_INET6) { + ipv6_addr_copy((struct in6_addr *)r->id.idiag_src, + &tcp6_rsk(req)->loc_addr); + ipv6_addr_copy((struct in6_addr *)r->id.idiag_dst, + &tcp6_rsk(req)->rmt_addr); + } +#endif + nlh->nlmsg_len = skb->tail - b; + + return skb->len; + +nlmsg_failure: + skb_trim(skb, b - skb->data); + return -1; +} + +static int inet_diag_dump_reqs(struct sk_buff *skb, struct sock *sk, + struct netlink_callback *cb) +{ + struct inet_diag_entry entry; + struct inet_diag_req *r = NLMSG_DATA(cb->nlh); + struct inet_connection_sock *icsk = inet_csk(sk); + struct listen_sock *lopt; + struct rtattr *bc = NULL; + struct inet_sock *inet = inet_sk(sk); + int j, s_j; + int reqnum, s_reqnum; + int err = 0; + + s_j = cb->args[3]; + s_reqnum = cb->args[4]; + + if (s_j > 0) + s_j--; + + entry.family = sk->sk_family; + + read_lock_bh(&icsk->icsk_accept_queue.syn_wait_lock); + + lopt = icsk->icsk_accept_queue.listen_opt; + if (!lopt || !lopt->qlen) + goto out; + + if (cb->nlh->nlmsg_len > 4 + NLMSG_SPACE(sizeof(*r))) { + bc = (struct rtattr *)(r + 1); + entry.sport = inet->num; + entry.userlocks = sk->sk_userlocks; + } + + for (j = s_j; j < lopt->nr_table_entries; j++) { + struct request_sock *req, *head = lopt->syn_table[j]; + + 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.idiag_dport != ireq->rmt_port && + r->id.idiag_dport) + continue; + + if (bc) { + entry.saddr = +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + (entry.family == AF_INET6) ? + tcp6_rsk(req)->loc_addr.s6_addr32 : +#endif + &ireq->loc_addr; + entry.daddr = +#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) + (entry.family == AF_INET6) ? + tcp6_rsk(req)->rmt_addr.s6_addr32 : +#endif + &ireq->rmt_addr; + entry.dport = ntohs(ireq->rmt_port); + + if (!inet_diag_bc_run(RTA_DATA(bc), + RTA_PAYLOAD(bc), &entry)) + continue; + } + + err = inet_diag_fill_req(skb, sk, req, + NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, cb->nlh); + if (err < 0) { + cb->args[3] = j + 1; + cb->args[4] = reqnum; + goto out; + } + } + + s_reqnum = 0; + } + +out: + read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock); + + return err; +} + +static int inet_diag_dump(struct sk_buff *skb, struct netlink_callback *cb) +{ + int i, num; + int s_i, s_num; + struct inet_diag_req *r = NLMSG_DATA(cb->nlh); + const struct inet_diag_handler *handler; + struct inet_hashinfo *hashinfo; + + handler = inet_diag_table[cb->nlh->nlmsg_type]; + BUG_ON(handler == NULL); + hashinfo = handler->idiag_hashinfo; + + s_i = cb->args[1]; + s_num = num = cb->args[2]; + + if (cb->args[0] == 0) { + if (!(r->idiag_states & (TCPF_LISTEN | TCPF_SYN_RECV))) + goto skip_listen_ht; + + inet_listen_lock(hashinfo); + for (i = s_i; i < INET_LHTABLE_SIZE; i++) { + struct sock *sk; + struct hlist_node *node; + + num = 0; + sk_for_each(sk, node, &hashinfo->listening_hash[i]) { + struct inet_sock *inet = inet_sk(sk); + + if (num < s_num) { + num++; + continue; + } + + if (r->id.idiag_sport != inet->sport && + r->id.idiag_sport) + goto next_listen; + + if (!(r->idiag_states & TCPF_LISTEN) || + r->id.idiag_dport || + cb->args[3] > 0) + goto syn_recv; + + if (inet_diag_dump_sock(skb, sk, cb) < 0) { + inet_listen_unlock(hashinfo); + goto done; + } + +syn_recv: + if (!(r->idiag_states & TCPF_SYN_RECV)) + goto next_listen; + + if (inet_diag_dump_reqs(skb, sk, cb) < 0) { + inet_listen_unlock(hashinfo); + goto done; + } + +next_listen: + cb->args[3] = 0; + cb->args[4] = 0; + ++num; + } + + s_num = 0; + cb->args[3] = 0; + cb->args[4] = 0; + } + inet_listen_unlock(hashinfo); +skip_listen_ht: + cb->args[0] = 1; + s_i = num = s_num = 0; + } + + if (!(r->idiag_states & ~(TCPF_LISTEN | TCPF_SYN_RECV))) + return skb->len; + + for (i = s_i; i < hashinfo->ehash_size; i++) { + struct inet_ehash_bucket *head = &hashinfo->ehash[i]; + struct sock *sk; + struct hlist_node *node; + + if (i > s_i) + s_num = 0; + + read_lock_bh(&head->lock); + + num = 0; + sk_for_each(sk, node, &head->chain) { + struct inet_sock *inet = inet_sk(sk); + + if (num < s_num) + goto next_normal; + if (!(r->idiag_states & (1 << sk->sk_state))) + goto next_normal; + if (r->id.idiag_sport != inet->sport && + r->id.idiag_sport) + goto next_normal; + if (r->id.idiag_dport != inet->dport && r->id.idiag_dport) + goto next_normal; + if (inet_diag_dump_sock(skb, sk, cb) < 0) { + read_unlock_bh(&head->lock); + goto done; + } +next_normal: + ++num; + } + + if (r->idiag_states & TCPF_TIME_WAIT) { + sk_for_each(sk, node, + &hashinfo->ehash[i + hashinfo->ehash_size].chain) { + struct inet_sock *inet = inet_sk(sk); + + if (num < s_num) + goto next_dying; + if (r->id.idiag_sport != inet->sport && + r->id.idiag_sport) + goto next_dying; + if (r->id.idiag_dport != inet->dport && + r->id.idiag_dport) + goto next_dying; + if (inet_diag_dump_sock(skb, sk, cb) < 0) { + read_unlock_bh(&head->lock); + goto done; + } +next_dying: + ++num; + } + } + read_unlock_bh(&head->lock); + } + +done: + cb->args[1] = i; + cb->args[2] = num; + return skb->len; +} + +static int inet_diag_dump_done(struct netlink_callback *cb) +{ + return 0; +} + + +static __inline__ int +inet_diag_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) +{ + if (!(nlh->nlmsg_flags&NLM_F_REQUEST)) + return 0; + + if (nlh->nlmsg_type >= INET_DIAG_GETSOCK_MAX) + goto err_inval; + + if (inet_diag_table[nlh->nlmsg_type] == NULL) + return -ENOENT; + + if (NLMSG_LENGTH(sizeof(struct inet_diag_req)) > skb->len) + goto err_inval; + + if (nlh->nlmsg_flags&NLM_F_DUMP) { + if (nlh->nlmsg_len > + (4 + NLMSG_SPACE(sizeof(struct inet_diag_req)))) { + struct rtattr *rta = (void *)(NLMSG_DATA(nlh) + + sizeof(struct inet_diag_req)); + if (rta->rta_type != INET_DIAG_REQ_BYTECODE || + rta->rta_len < 8 || + rta->rta_len > + (nlh->nlmsg_len - + NLMSG_SPACE(sizeof(struct inet_diag_req)))) + goto err_inval; + if (inet_diag_bc_audit(RTA_DATA(rta), RTA_PAYLOAD(rta))) + goto err_inval; + } + return netlink_dump_start(idiagnl, skb, nlh, + inet_diag_dump, + inet_diag_dump_done); + } else { + return inet_diag_get_exact(skb, nlh); + } + +err_inval: + return -EINVAL; +} + + +static inline void inet_diag_rcv_skb(struct sk_buff *skb) +{ + int err; + struct nlmsghdr * nlh; + + if (skb->len >= NLMSG_SPACE(0)) { + nlh = (struct nlmsghdr *)skb->data; + if (nlh->nlmsg_len < sizeof(*nlh) || skb->len < nlh->nlmsg_len) + return; + err = inet_diag_rcv_msg(skb, nlh); + if (err || nlh->nlmsg_flags & NLM_F_ACK) + netlink_ack(skb, nlh, err); + } +} + +static void inet_diag_rcv(struct sock *sk, int len) +{ + struct sk_buff *skb; + unsigned int qlen = skb_queue_len(&sk->sk_receive_queue); + + while (qlen-- && (skb = skb_dequeue(&sk->sk_receive_queue))) { + inet_diag_rcv_skb(skb); + kfree_skb(skb); + } +} + +static DEFINE_SPINLOCK(inet_diag_register_lock); + +int inet_diag_register(const struct inet_diag_handler *h) +{ + const __u16 type = h->idiag_type; + int err = -EINVAL; + + if (type >= INET_DIAG_GETSOCK_MAX) + goto out; + + spin_lock(&inet_diag_register_lock); + err = -EEXIST; + if (inet_diag_table[type] == NULL) { + inet_diag_table[type] = h; + err = 0; + } + spin_unlock(&inet_diag_register_lock); +out: + return err; +} +EXPORT_SYMBOL_GPL(inet_diag_register); + +void inet_diag_unregister(const struct inet_diag_handler *h) +{ + const __u16 type = h->idiag_type; + + if (type >= INET_DIAG_GETSOCK_MAX) + return; + + spin_lock(&inet_diag_register_lock); + inet_diag_table[type] = NULL; + spin_unlock(&inet_diag_register_lock); + + synchronize_rcu(); +} +EXPORT_SYMBOL_GPL(inet_diag_unregister); + +static int __init inet_diag_init(void) +{ + const int inet_diag_table_size = (INET_DIAG_GETSOCK_MAX * + sizeof(struct inet_diag_handler *)); + int err = -ENOMEM; + + inet_diag_table = kmalloc(inet_diag_table_size, GFP_KERNEL); + if (!inet_diag_table) + goto out; + + memset(inet_diag_table, 0, inet_diag_table_size); + idiagnl = netlink_kernel_create(NETLINK_INET_DIAG, 0, inet_diag_rcv, + THIS_MODULE); + if (idiagnl == NULL) + goto out_free_table; + err = 0; +out: + return err; +out_free_table: + kfree(inet_diag_table); + goto out; +} + +static void __exit inet_diag_exit(void) +{ + sock_release(idiagnl->sk_socket); + kfree(inet_diag_table); +} + +module_init(inet_diag_init); +module_exit(inet_diag_exit); +MODULE_LICENSE("GPL"); diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c new file mode 100644 index 000000000000..e8d29fe736d2 --- /dev/null +++ b/net/ipv4/inet_hashtables.c @@ -0,0 +1,165 @@ +/* + * INET An implementation of the TCP/IP protocol suite for the LINUX + * operating system. INET is implemented using the BSD Socket + * interface as the means of communication with the user level. + * + * Generic INET transport hashtables + * + * Authors: Lotsa people, from code originally in tcp + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/sched.h> +#include <linux/slab.h> +#include <linux/wait.h> + +#include <net/inet_connection_sock.h> +#include <net/inet_hashtables.h> + +/* + * Allocate and initialize a new local port bind bucket. + * The bindhash mutex for snum's hash chain must be held here. + */ +struct inet_bind_bucket *inet_bind_bucket_create(kmem_cache_t *cachep, + struct inet_bind_hashbucket *head, + const unsigned short snum) +{ + struct inet_bind_bucket *tb = kmem_cache_alloc(cachep, SLAB_ATOMIC); + + if (tb != NULL) { + tb->port = snum; + tb->fastreuse = 0; + INIT_HLIST_HEAD(&tb->owners); + hlist_add_head(&tb->node, &head->chain); + } + return tb; +} + +EXPORT_SYMBOL(inet_bind_bucket_create); + +/* + * Caller must hold hashbucket lock for this tb with local BH disabled + */ +void inet_bind_bucket_destroy(kmem_cache_t *cachep, struct inet_bind_bucket *tb) +{ + if (hlist_empty(&tb->owners)) { + __hlist_del(&tb->node); + kmem_cache_free(cachep, tb); + } +} + +void inet_bind_hash(struct sock *sk, struct inet_bind_bucket *tb, + const unsigned short snum) +{ + inet_sk(sk)->num = snum; + sk_add_bind_node(sk, &tb->owners); + inet_csk(sk)->icsk_bind_hash = tb; +} + +EXPORT_SYMBOL(inet_bind_hash); + +/* + * Get rid of any references to a local port held by the given sock. + */ +static void __inet_put_port(struct inet_hashinfo *hashinfo, struct sock *sk) +{ + const int bhash = inet_bhashfn(inet_sk(sk)->num, hashinfo->bhash_size); + struct inet_bind_hashbucket *head = &hashinfo->bhash[bhash]; + struct inet_bind_bucket *tb; + + spin_lock(&head->lock); + tb = inet_csk(sk)->icsk_bind_hash; + __sk_del_bind_node(sk); + inet_csk(sk)->icsk_bind_hash = NULL; + inet_sk(sk)->num = 0; + inet_bind_bucket_destroy(hashinfo->bind_bucket_cachep, tb); + spin_unlock(&head->lock); +} + +void inet_put_port(struct inet_hashinfo *hashinfo, struct sock *sk) +{ + local_bh_disable(); + __inet_put_port(hashinfo, sk); + local_bh_enable(); +} + +EXPORT_SYMBOL(inet_put_port); + +/* + * This lock without WQ_FLAG_EXCLUSIVE is good on UP and it can be very bad on SMP. + * Look, when several writers sleep and reader wakes them up, all but one + * immediately hit write lock and grab all the cpus. Exclusive sleep solves + * this, _but_ remember, it adds useless work on UP machines (wake up each + * exclusive lock release). It should be ifdefed really. + */ +void inet_listen_wlock(struct inet_hashinfo *hashinfo) +{ + write_lock(&hashinfo->lhash_lock); + + if (atomic_read(&hashinfo->lhash_users)) { + DEFINE_WAIT(wait); + + for (;;) { + prepare_to_wait_exclusive(&hashinfo->lhash_wait, + &wait, TASK_UNINTERRUPTIBLE); + if (!atomic_read(&hashinfo->lhash_users)) + break; + write_unlock_bh(&hashinfo->lhash_lock); + schedule(); + write_lock_bh(&hashinfo->lhash_lock); + } + + finish_wait(&hashinfo->lhash_wait, &wait); + } +} + +EXPORT_SYMBOL(inet_listen_wlock); + +/* + * Don't inline this cruft. Here are some nice properties to exploit here. The + * BSD API does not allow a listening sock to specify the remote port nor the + * remote address for the connection. So always assume those are both + * wildcarded during the search since they can never be otherwise. + */ +struct sock *__inet_lookup_listener(const struct hlist_head *head, const u32 daddr, + const unsigned short hnum, const int dif) +{ + struct sock *result = NULL, *sk; + const struct hlist_node *node; + int hiscore = -1; + + sk_for_each(sk, node, head) { + const struct inet_sock *inet = inet_sk(sk); + + if (inet->num == hnum && !ipv6_only_sock(sk)) { + const __u32 rcv_saddr = inet->rcv_saddr; + int score = sk->sk_family == PF_INET ? 1 : 0; + + if (rcv_saddr) { + if (rcv_saddr != daddr) + continue; + score += 2; + } + if (sk->sk_bound_dev_if) { + if (sk->sk_bound_dev_if != dif) + continue; + score += 2; + } + if (score == 5) + return sk; + if (score > hiscore) { + hiscore = score; + result = sk; + } + } + } + return result; +} + +EXPORT_SYMBOL_GPL(__inet_lookup_listener); diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c new file mode 100644 index 000000000000..4d1502a49852 --- /dev/null +++ b/net/ipv4/inet_timewait_sock.c @@ -0,0 +1,384 @@ +/* + * INET An implementation of the TCP/IP protocol suite for the LINUX + * operating system. INET is implemented using the BSD Socket + * interface as the means of communication with the user level. + * + * Generic TIME_WAIT sockets functions + * + * From code orinally in TCP + */ + +#include <linux/config.h> + +#include <net/inet_hashtables.h> +#include <net/inet_timewait_sock.h> +#include <net/ip.h> + +/* Must be called with locally disabled BHs. */ +void __inet_twsk_kill(struct inet_timewait_sock *tw, struct inet_hashinfo *hashinfo) +{ + struct inet_bind_hashbucket *bhead; + struct inet_bind_bucket *tb; + /* Unlink from established hashes. */ + struct inet_ehash_bucket *ehead = &hashinfo->ehash[tw->tw_hashent]; + + write_lock(&ehead->lock); + if (hlist_unhashed(&tw->tw_node)) { + write_unlock(&ehead->lock); + return; + } + __hlist_del(&tw->tw_node); + sk_node_init(&tw->tw_node); + write_unlock(&ehead->lock); + + /* Disassociate with bind bucket. */ + bhead = &hashinfo->bhash[inet_bhashfn(tw->tw_num, hashinfo->bhash_size)]; + spin_lock(&bhead->lock); + tb = tw->tw_tb; + __hlist_del(&tw->tw_bind_node); + tw->tw_tb = NULL; + inet_bind_bucket_destroy(hashinfo->bind_bucket_cachep, tb); + spin_unlock(&bhead->lock); +#ifdef SOCK_REFCNT_DEBUG + if (atomic_read(&tw->tw_refcnt) != 1) { + printk(KERN_DEBUG "%s timewait_sock %p refcnt=%d\n", + tw->tw_prot->name, tw, atomic_read(&tw->tw_refcnt)); + } +#endif + inet_twsk_put(tw); +} + +EXPORT_SYMBOL_GPL(__inet_twsk_kill); + +/* + * Enter the time wait state. This is called with locally disabled BH. + * Essentially we whip up a timewait bucket, copy the relevant info into it + * from the SK, and mess with hash chains and list linkage. + */ +void __inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk, + struct inet_hashinfo *hashinfo) +{ + const struct inet_sock *inet = inet_sk(sk); + const struct inet_connection_sock *icsk = inet_csk(sk); + struct inet_ehash_bucket *ehead = &hashinfo->ehash[sk->sk_hashent]; + struct inet_bind_hashbucket *bhead; + /* Step 1: Put TW into bind hash. Original socket stays there too. + Note, that any socket with inet->num != 0 MUST be bound in + binding cache, even if it is closed. + */ + bhead = &hashinfo->bhash[inet_bhashfn(inet->num, hashinfo->bhash_size)]; + spin_lock(&bhead->lock); + tw->tw_tb = icsk->icsk_bind_hash; + BUG_TRAP(icsk->icsk_bind_hash); + inet_twsk_add_bind_node(tw, &tw->tw_tb->owners); + spin_unlock(&bhead->lock); + + write_lock(&ehead->lock); + + /* Step 2: Remove SK from established hash. */ + if (__sk_del_node_init(sk)) + sock_prot_dec_use(sk->sk_prot); + + /* Step 3: Hash TW into TIMEWAIT half of established hash table. */ + inet_twsk_add_node(tw, &(ehead + hashinfo->ehash_size)->chain); + atomic_inc(&tw->tw_refcnt); + + write_unlock(&ehead->lock); +} + +EXPORT_SYMBOL_GPL(__inet_twsk_hashdance); + +struct inet_timewait_sock *inet_twsk_alloc(const struct sock *sk, const int state) +{ + struct inet_timewait_sock *tw = kmem_cache_alloc(sk->sk_prot_creator->twsk_slab, + SLAB_ATOMIC); + if (tw != NULL) { + const struct inet_sock *inet = inet_sk(sk); + + /* Give us an identity. */ + tw->tw_daddr = inet->daddr; + tw->tw_rcv_saddr = inet->rcv_saddr; + tw->tw_bound_dev_if = sk->sk_bound_dev_if; + tw->tw_num = inet->num; + tw->tw_state = TCP_TIME_WAIT; + tw->tw_substate = state; + tw->tw_sport = inet->sport; + tw->tw_dport = inet->dport; + tw->tw_family = sk->sk_family; + tw->tw_reuse = sk->sk_reuse; + tw->tw_hashent = sk->sk_hashent; + tw->tw_ipv6only = 0; + tw->tw_prot = sk->sk_prot_creator; + atomic_set(&tw->tw_refcnt, 1); + inet_twsk_dead_node_init(tw); + } + + return tw; +} + +EXPORT_SYMBOL_GPL(inet_twsk_alloc); + +/* Returns non-zero if quota exceeded. */ +static int inet_twdr_do_twkill_work(struct inet_timewait_death_row *twdr, + const int slot) +{ + struct inet_timewait_sock *tw; + struct hlist_node *node; + unsigned int killed; + int ret; + + /* NOTE: compare this to previous version where lock + * was released after detaching chain. It was racy, + * because tw buckets are scheduled in not serialized context + * in 2.3 (with netfilter), and with softnet it is common, because + * soft irqs are not sequenced. + */ + killed = 0; + ret = 0; +rescan: + inet_twsk_for_each_inmate(tw, node, &twdr->cells[slot]) { + __inet_twsk_del_dead_node(tw); + spin_unlock(&twdr->death_lock); + __inet_twsk_kill(tw, twdr->hashinfo); + inet_twsk_put(tw); + killed++; + spin_lock(&twdr->death_lock); + if (killed > INET_TWDR_TWKILL_QUOTA) { + ret = 1; + break; + } + + /* While we dropped twdr->death_lock, another cpu may have + * killed off the next TW bucket in the list, therefore + * do a fresh re-read of the hlist head node with the + * lock reacquired. We still use the hlist traversal + * macro in order to get the prefetches. + */ + goto rescan; + } + + twdr->tw_count -= killed; + NET_ADD_STATS_BH(LINUX_MIB_TIMEWAITED, killed); + + return ret; +} + +void inet_twdr_hangman(unsigned long data) +{ + struct inet_timewait_death_row *twdr; + int unsigned need_timer; + + twdr = (struct inet_timewait_death_row *)data; + spin_lock(&twdr->death_lock); + + if (twdr->tw_count == 0) + goto out; + + need_timer = 0; + if (inet_twdr_do_twkill_work(twdr, twdr->slot)) { + twdr->thread_slots |= (1 << twdr->slot); + mb(); + schedule_work(&twdr->twkill_work); + need_timer = 1; + } else { + /* We purged the entire slot, anything left? */ + if (twdr->tw_count) + need_timer = 1; + } + twdr->slot = ((twdr->slot + 1) & (INET_TWDR_TWKILL_SLOTS - 1)); + if (need_timer) + mod_timer(&twdr->tw_timer, jiffies + twdr->period); +out: + spin_unlock(&twdr->death_lock); +} + +EXPORT_SYMBOL_GPL(inet_twdr_hangman); + +extern void twkill_slots_invalid(void); + +void inet_twdr_twkill_work(void *data) +{ + struct inet_timewait_death_row *twdr = data; + int i; + + if ((INET_TWDR_TWKILL_SLOTS - 1) > (sizeof(twdr->thread_slots) * 8)) + twkill_slots_invalid(); + + while (twdr->thread_slots) { + spin_lock_bh(&twdr->death_lock); + for (i = 0; i < INET_TWDR_TWKILL_SLOTS; i++) { + if (!(twdr->thread_slots & (1 << i))) + continue; + + while (inet_twdr_do_twkill_work(twdr, i) != 0) { + if (need_resched()) { + spin_unlock_bh(&twdr->death_lock); + schedule(); + spin_lock_bh(&twdr->death_lock); + } + } + + twdr->thread_slots &= ~(1 << i); + } + spin_unlock_bh(&twdr->death_lock); + } +} + +EXPORT_SYMBOL_GPL(inet_twdr_twkill_work); + +/* These are always called from BH context. See callers in + * tcp_input.c to verify this. + */ + +/* This is for handling early-kills of TIME_WAIT sockets. */ +void inet_twsk_deschedule(struct inet_timewait_sock *tw, + struct inet_timewait_death_row *twdr) +{ + spin_lock(&twdr->death_lock); + if (inet_twsk_del_dead_node(tw)) { + inet_twsk_put(tw); + if (--twdr->tw_count == 0) + del_timer(&twdr->tw_timer); + } + spin_unlock(&twdr->death_lock); + __inet_twsk_kill(tw, twdr->hashinfo); +} + +EXPORT_SYMBOL(inet_twsk_deschedule); + +void inet_twsk_schedule(struct inet_timewait_sock *tw, + struct inet_timewait_death_row *twdr, + const int timeo, const int timewait_len) +{ + struct hlist_head *list; + int slot; + + /* timeout := RTO * 3.5 + * + * 3.5 = 1+2+0.5 to wait for two retransmits. + * + * RATIONALE: if FIN arrived and we entered TIME-WAIT state, + * our ACK acking that FIN can be lost. If N subsequent retransmitted + * FINs (or previous seqments) are lost (probability of such event + * is p^(N+1), where p is probability to lose single packet and + * time to detect the loss is about RTO*(2^N - 1) with exponential + * backoff). Normal timewait length is calculated so, that we + * waited at least for one retransmitted FIN (maximal RTO is 120sec). + * [ BTW Linux. following BSD, violates this requirement waiting + * only for 60sec, we should wait at least for 240 secs. + * Well, 240 consumes too much of resources 8) + * ] + * This interval is not reduced to catch old duplicate and + * responces to our wandering segments living for two MSLs. + * However, if we use PAWS to detect + * old duplicates, we can reduce the interval to bounds required + * by RTO, rather than MSL. So, if peer understands PAWS, we + * kill tw bucket after 3.5*RTO (it is important that this number + * is greater than TS tick!) and detect old duplicates with help + * of PAWS. + */ + slot = (timeo + (1 << INET_TWDR_RECYCLE_TICK) - 1) >> INET_TWDR_RECYCLE_TICK; + + spin_lock(&twdr->death_lock); + + /* Unlink it, if it was scheduled */ + if (inet_twsk_del_dead_node(tw)) + twdr->tw_count--; + else + atomic_inc(&tw->tw_refcnt); + + if (slot >= INET_TWDR_RECYCLE_SLOTS) { + /* Schedule to slow timer */ + if (timeo >= timewait_len) { + slot = INET_TWDR_TWKILL_SLOTS - 1; + } else { + slot = (timeo + twdr->period - 1) / twdr->period; + if (slot >= INET_TWDR_TWKILL_SLOTS) + slot = INET_TWDR_TWKILL_SLOTS - 1; + } + tw->tw_ttd = jiffies + timeo; + slot = (twdr->slot + slot) & (INET_TWDR_TWKILL_SLOTS - 1); + list = &twdr->cells[slot]; + } else { + tw->tw_ttd = jiffies + (slot << INET_TWDR_RECYCLE_TICK); + + if (twdr->twcal_hand < 0) { + twdr->twcal_hand = 0; + twdr->twcal_jiffie = jiffies; + twdr->twcal_timer.expires = twdr->twcal_jiffie + + (slot << INET_TWDR_RECYCLE_TICK); + add_timer(&twdr->twcal_timer); + } else { + if (time_after(twdr->twcal_timer.expires, + jiffies + (slot << INET_TWDR_RECYCLE_TICK))) + mod_timer(&twdr->twcal_timer, + jiffies + (slot << INET_TWDR_RECYCLE_TICK)); + slot = (twdr->twcal_hand + slot) & (INET_TWDR_RECYCLE_SLOTS - 1); + } + list = &twdr->twcal_row[slot]; + } + + hlist_add_head(&tw->tw_death_node, list); + + if (twdr->tw_count++ == 0) + mod_timer(&twdr->tw_timer, jiffies + twdr->period); + spin_unlock(&twdr->death_lock); +} + +EXPORT_SYMBOL_GPL(inet_twsk_schedule); + +void inet_twdr_twcal_tick(unsigned long data) +{ + struct inet_timewait_death_row *twdr; + int n, slot; + unsigned long j; + unsigned long now = jiffies; + int killed = 0; + int adv = 0; + + twdr = (struct inet_timewait_death_row *)data; + + spin_lock(&twdr->death_lock); + if (twdr->twcal_hand < 0) + goto out; + + slot = twdr->twcal_hand; + j = twdr->twcal_jiffie; + + for (n = 0; n < INET_TWDR_RECYCLE_SLOTS; n++) { + if (time_before_eq(j, now)) { + struct hlist_node *node, *safe; + struct inet_timewait_sock *tw; + + inet_twsk_for_each_inmate_safe(tw, node, safe, + &twdr->twcal_row[slot]) { + __inet_twsk_del_dead_node(tw); + __inet_twsk_kill(tw, twdr->hashinfo); + inet_twsk_put(tw); + killed++; + } + } else { + if (!adv) { + adv = 1; + twdr->twcal_jiffie = j; + twdr->twcal_hand = slot; + } + + if (!hlist_empty(&twdr->twcal_row[slot])) { + mod_timer(&twdr->twcal_timer, j); + goto out; + } + } + j += 1 << INET_TWDR_RECYCLE_TICK; + slot = (slot + 1) & (INET_TWDR_RECYCLE_SLOTS - 1); + } + twdr->twcal_hand = -1; + +out: + if ((twdr->tw_count -= killed) == 0) + del_timer(&twdr->tw_timer); + NET_ADD_STATS_BH(LINUX_MIB_TIMEWAITKILLED, killed); + spin_unlock(&twdr->death_lock); +} + +EXPORT_SYMBOL_GPL(inet_twdr_twcal_tick); diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index ab18a853d7ce..f84ba9c96551 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -20,6 +20,7 @@ #include <linux/kernel.h> #include <linux/mm.h> #include <linux/net.h> +#include <net/ip.h> #include <net/inetpeer.h> /* @@ -72,7 +73,7 @@ /* Exported for inet_getid inline function. */ DEFINE_SPINLOCK(inet_peer_idlock); -static kmem_cache_t *peer_cachep; +static kmem_cache_t *peer_cachep __read_mostly; #define node_height(x) x->avl_height static struct inet_peer peer_fake_node = { @@ -459,5 +460,3 @@ static void peer_check_expire(unsigned long dummy) peer_total / inet_peer_threshold * HZ; add_timer(&peer_periodic_timer); } - -EXPORT_SYMBOL(inet_peer_idlock); diff --git a/net/ipv4/ip_forward.c b/net/ipv4/ip_forward.c index 77094aac6c28..0923add122b4 100644 --- a/net/ipv4/ip_forward.c +++ b/net/ipv4/ip_forward.c @@ -76,16 +76,12 @@ int ip_forward(struct sk_buff *skb) * that reaches zero, we must reply an ICMP control message telling * that the packet's lifetime expired. */ - - iph = skb->nh.iph; - - if (iph->ttl <= 1) + if (skb->nh.iph->ttl <= 1) goto too_many_hops; if (!xfrm4_route_forward(skb)) goto drop; - iph = skb->nh.iph; rt = (struct rtable*)skb->dst; if (opt->is_strictroute && rt->rt_dst != rt->rt_gateway) diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c index eb377ae15305..9e6e683cc34d 100644 --- a/net/ipv4/ip_fragment.c +++ b/net/ipv4/ip_fragment.c @@ -377,7 +377,7 @@ static struct ipq *ip_frag_create(unsigned hash, struct iphdr *iph, u32 user) return ip_frag_intern(hash, qp); out_nomem: - LIMIT_NETDEBUG(printk(KERN_ERR "ip_frag_create: no memory left !\n")); + LIMIT_NETDEBUG(KERN_ERR "ip_frag_create: no memory left !\n"); return NULL; } @@ -533,7 +533,7 @@ static void ip_frag_queue(struct ipq *qp, struct sk_buff *skb) if (skb->dev) qp->iif = skb->dev->ifindex; skb->dev = NULL; - qp->stamp = skb->stamp; + skb_get_timestamp(skb, &qp->stamp); qp->meat += skb->len; atomic_add(skb->truesize, &ip_frag_mem); if (offset == 0) @@ -615,7 +615,7 @@ static struct sk_buff *ip_frag_reasm(struct ipq *qp, struct net_device *dev) head->next = NULL; head->dev = dev; - head->stamp = qp->stamp; + skb_set_timestamp(head, &qp->stamp); iph = head->nh.iph; iph->frag_off = 0; @@ -625,8 +625,8 @@ static struct sk_buff *ip_frag_reasm(struct ipq *qp, struct net_device *dev) return head; out_nomem: - LIMIT_NETDEBUG(printk(KERN_ERR "IP: queue_glue: no memory for gluing " - "queue %p\n", qp)); + LIMIT_NETDEBUG(KERN_ERR "IP: queue_glue: no memory for gluing " + "queue %p\n", qp); goto out_fail; out_oversize: if (net_ratelimit()) diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c index c703528e0bcd..473d0f2b2e0d 100644 --- a/net/ipv4/ip_input.c +++ b/net/ipv4/ip_input.c @@ -150,7 +150,7 @@ * SNMP management statistics */ -DEFINE_SNMP_STAT(struct ipstats_mib, ip_statistics); +DEFINE_SNMP_STAT(struct ipstats_mib, ip_statistics) __read_mostly; /* * Process Router Attention IP option @@ -225,8 +225,8 @@ static inline int ip_local_deliver_finish(struct sk_buff *skb) /* If there maybe a raw socket we must check - if not we * don't care less */ - if (raw_sk) - raw_v4_input(skb, skb->nh.iph, hash); + if (raw_sk && !raw_v4_input(skb, skb->nh.iph, hash)) + raw_sk = NULL; if ((ipprot = rcu_dereference(inet_protos[hash])) != NULL) { int ret; @@ -279,18 +279,70 @@ int ip_local_deliver(struct sk_buff *skb) ip_local_deliver_finish); } -static inline int ip_rcv_finish(struct sk_buff *skb) +static inline int ip_rcv_options(struct sk_buff *skb) { + struct ip_options *opt; + struct iphdr *iph; struct net_device *dev = skb->dev; + + /* It looks as overkill, because not all + IP options require packet mangling. + But it is the easiest for now, especially taking + into account that combination of IP options + and running sniffer is extremely rare condition. + --ANK (980813) + */ + if (skb_cow(skb, skb_headroom(skb))) { + IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS); + goto drop; + } + + iph = skb->nh.iph; + + if (ip_options_compile(NULL, skb)) { + IP_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS); + goto drop; + } + + opt = &(IPCB(skb)->opt); + if (unlikely(opt->srr)) { + struct in_device *in_dev = in_dev_get(dev); + if (in_dev) { + if (!IN_DEV_SOURCE_ROUTE(in_dev)) { + if (IN_DEV_LOG_MARTIANS(in_dev) && + net_ratelimit()) + printk(KERN_INFO "source route option " + "%u.%u.%u.%u -> %u.%u.%u.%u\n", + NIPQUAD(iph->saddr), + NIPQUAD(iph->daddr)); + in_dev_put(in_dev); + goto drop; + } + + in_dev_put(in_dev); + } + + if (ip_options_rcv_srr(skb)) + goto drop; + } + + return 0; +drop: + return -1; +} + +static inline int ip_rcv_finish(struct sk_buff *skb) +{ struct iphdr *iph = skb->nh.iph; - int err; /* * Initialise the virtual path cache for the packet. It describes * how the packet travels inside Linux networking. */ - if (skb->dst == NULL) { - if ((err = ip_route_input(skb, iph->daddr, iph->saddr, iph->tos, dev))) { + if (likely(skb->dst == NULL)) { + int err = ip_route_input(skb, iph->daddr, iph->saddr, iph->tos, + skb->dev); + if (unlikely(err)) { if (err == -EHOSTUNREACH) IP_INC_STATS_BH(IPSTATS_MIB_INADDRERRORS); goto drop; @@ -298,7 +350,7 @@ static inline int ip_rcv_finish(struct sk_buff *skb) } #ifdef CONFIG_NET_CLS_ROUTE - if (skb->dst->tclassid) { + if (unlikely(skb->dst->tclassid)) { struct ip_rt_acct *st = ip_rt_acct + 256*smp_processor_id(); u32 idx = skb->dst->tclassid; st[idx&0xFF].o_packets++; @@ -308,48 +360,11 @@ static inline int ip_rcv_finish(struct sk_buff *skb) } #endif - if (iph->ihl > 5) { - struct ip_options *opt; - - /* It looks as overkill, because not all - IP options require packet mangling. - But it is the easiest for now, especially taking - into account that combination of IP options - and running sniffer is extremely rare condition. - --ANK (980813) - */ - - if (skb_cow(skb, skb_headroom(skb))) { - IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS); - goto drop; - } - iph = skb->nh.iph; - - if (ip_options_compile(NULL, skb)) - goto inhdr_error; - - opt = &(IPCB(skb)->opt); - if (opt->srr) { - struct in_device *in_dev = in_dev_get(dev); - if (in_dev) { - if (!IN_DEV_SOURCE_ROUTE(in_dev)) { - if (IN_DEV_LOG_MARTIANS(in_dev) && net_ratelimit()) - printk(KERN_INFO "source route option %u.%u.%u.%u -> %u.%u.%u.%u\n", - NIPQUAD(iph->saddr), NIPQUAD(iph->daddr)); - in_dev_put(in_dev); - goto drop; - } - in_dev_put(in_dev); - } - if (ip_options_rcv_srr(skb)) - goto drop; - } - } + if (iph->ihl > 5 && ip_rcv_options(skb)) + goto drop; return dst_input(skb); -inhdr_error: - IP_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS); drop: kfree_skb(skb); return NET_RX_DROP; @@ -358,9 +373,10 @@ drop: /* * Main IP Receive routine. */ -int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct iphdr *iph; + u32 len; /* When the interface is in promisc. mode, drop all the crap * that it receives, do not try to analyse it. @@ -392,29 +408,27 @@ int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) */ if (iph->ihl < 5 || iph->version != 4) - goto inhdr_error; + goto inhdr_error; if (!pskb_may_pull(skb, iph->ihl*4)) goto inhdr_error; iph = skb->nh.iph; - if (ip_fast_csum((u8 *)iph, iph->ihl) != 0) - goto inhdr_error; + if (unlikely(ip_fast_csum((u8 *)iph, iph->ihl))) + goto inhdr_error; - { - __u32 len = ntohs(iph->tot_len); - if (skb->len < len || len < (iph->ihl<<2)) - goto inhdr_error; + len = ntohs(iph->tot_len); + if (skb->len < len || len < (iph->ihl*4)) + goto inhdr_error; - /* Our transport medium may have padded the buffer out. Now we know it - * is IP we can trim to the true length of the frame. - * Note this now means skb->len holds ntohs(iph->tot_len). - */ - if (pskb_trim_rcsum(skb, len)) { - IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS); - goto drop; - } + /* Our transport medium may have padded the buffer out. Now we know it + * is IP we can trim to the true length of the frame. + * Note this now means skb->len holds ntohs(iph->tot_len). + */ + if (pskb_trim_rcsum(skb, len)) { + IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS); + goto drop; } return NF_HOOK(PF_INET, NF_IP_PRE_ROUTING, skb, dev, NULL, @@ -428,5 +442,4 @@ out: return NET_RX_DROP; } -EXPORT_SYMBOL(ip_rcv); EXPORT_SYMBOL(ip_statistics); diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c index 6d89f3f3e701..bce4e875193b 100644 --- a/net/ipv4/ip_options.c +++ b/net/ipv4/ip_options.c @@ -489,23 +489,18 @@ void ip_options_undo(struct ip_options * opt) } } -int ip_options_get(struct ip_options **optp, unsigned char *data, int optlen, int user) +static struct ip_options *ip_options_get_alloc(const int optlen) { - struct ip_options *opt; + struct ip_options *opt = kmalloc(sizeof(*opt) + ((optlen + 3) & ~3), + GFP_KERNEL); + if (opt) + memset(opt, 0, sizeof(*opt)); + return opt; +} - opt = kmalloc(sizeof(struct ip_options)+((optlen+3)&~3), GFP_KERNEL); - if (!opt) - return -ENOMEM; - memset(opt, 0, sizeof(struct ip_options)); - if (optlen) { - if (user) { - if (copy_from_user(opt->__data, data, optlen)) { - kfree(opt); - return -EFAULT; - } - } else - memcpy(opt->__data, data, optlen); - } +static int ip_options_get_finish(struct ip_options **optp, + struct ip_options *opt, int optlen) +{ while (optlen & 3) opt->__data[optlen++] = IPOPT_END; opt->optlen = optlen; @@ -521,6 +516,30 @@ int ip_options_get(struct ip_options **optp, unsigned char *data, int optlen, in return 0; } +int ip_options_get_from_user(struct ip_options **optp, unsigned char __user *data, int optlen) +{ + struct ip_options *opt = ip_options_get_alloc(optlen); + + if (!opt) + return -ENOMEM; + if (optlen && copy_from_user(opt->__data, data, optlen)) { + kfree(opt); + return -EFAULT; + } + return ip_options_get_finish(optp, opt, optlen); +} + +int ip_options_get(struct ip_options **optp, unsigned char *data, int optlen) +{ + struct ip_options *opt = ip_options_get_alloc(optlen); + + if (!opt) + return -ENOMEM; + if (optlen) + memcpy(opt->__data, data, optlen); + return ip_options_get_finish(optp, opt, optlen); +} + void ip_forward_options(struct sk_buff *skb) { struct ip_options * opt = &(IPCB(skb)->opt); @@ -620,6 +639,3 @@ int ip_options_rcv_srr(struct sk_buff *skb) } return 0; } - -EXPORT_SYMBOL(ip_options_compile); -EXPORT_SYMBOL(ip_options_undo); diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 80d13103b2b0..3f1a263e1249 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -69,13 +69,10 @@ #include <net/ip.h> #include <net/protocol.h> #include <net/route.h> -#include <net/tcp.h> -#include <net/udp.h> #include <linux/skbuff.h> #include <net/sock.h> #include <net/arp.h> #include <net/icmp.h> -#include <net/raw.h> #include <net/checksum.h> #include <net/inetpeer.h> #include <net/checksum.h> @@ -84,12 +81,8 @@ #include <linux/netfilter_bridge.h> #include <linux/mroute.h> #include <linux/netlink.h> +#include <linux/tcp.h> -/* - * Shall we try to damage output packets if routing dev changes? - */ - -int sysctl_ip_dynaddr; int sysctl_ip_default_ttl = IPDEFTTL; /* Generate a checksum for an outgoing IP datagram. */ @@ -165,6 +158,8 @@ int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk, dst_output); } +EXPORT_SYMBOL_GPL(ip_build_and_send_pkt); + static inline int ip_finish_output2(struct sk_buff *skb) { struct dst_entry *dst = skb->dst; @@ -205,7 +200,7 @@ static inline int ip_finish_output2(struct sk_buff *skb) return -EINVAL; } -int ip_finish_output(struct sk_buff *skb) +static inline int ip_finish_output(struct sk_buff *skb) { struct net_device *dev = skb->dst->dev; @@ -329,8 +324,7 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok) if (ip_route_output_flow(&rt, &fl, sk, 0)) goto no_route; } - __sk_dst_set(sk, &rt->u.dst); - tcp_v4_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->u.dst); } skb->dst = dst_clone(&rt->u.dst); @@ -392,7 +386,6 @@ static void ip_copy_metadata(struct sk_buff *to, struct sk_buff *from) #endif #ifdef CONFIG_NETFILTER to->nfmark = from->nfmark; - to->nfcache = from->nfcache; /* Connection association is same as pre-frag packet */ nf_conntrack_put(to->nfct); to->nfct = from->nfct; @@ -580,7 +573,7 @@ slow_path: */ if ((skb2 = alloc_skb(len+hlen+ll_rs, GFP_ATOMIC)) == NULL) { - NETDEBUG(printk(KERN_INFO "IP: frag: no memory for new fragment!\n")); + NETDEBUG(KERN_INFO "IP: frag: no memory for new fragment!\n"); err = -ENOMEM; goto fail; } @@ -1329,12 +1322,7 @@ void __init ip_init(void) #endif } -EXPORT_SYMBOL(ip_finish_output); EXPORT_SYMBOL(ip_fragment); EXPORT_SYMBOL(ip_generic_getfrag); EXPORT_SYMBOL(ip_queue_xmit); EXPORT_SYMBOL(ip_send_check); - -#ifdef CONFIG_SYSCTL -EXPORT_SYMBOL(sysctl_ip_default_ttl); -#endif diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index ff4bd067b397..2f0b47da5b37 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -153,7 +153,7 @@ int ip_cmsg_send(struct msghdr *msg, struct ipcm_cookie *ipc) switch (cmsg->cmsg_type) { case IP_RETOPTS: err = cmsg->cmsg_len - CMSG_ALIGN(sizeof(struct cmsghdr)); - err = ip_options_get(&ipc->opt, CMSG_DATA(cmsg), err < 40 ? err : 40, 0); + err = ip_options_get(&ipc->opt, CMSG_DATA(cmsg), err < 40 ? err : 40); if (err) return err; break; @@ -425,7 +425,7 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, struct ip_options * opt = NULL; if (optlen > 40 || optlen < 0) goto e_inval; - err = ip_options_get(&opt, optval, optlen, 1); + err = ip_options_get_from_user(&opt, optval, optlen); if (err) break; if (sk->sk_type == SOCK_STREAM) { @@ -614,7 +614,6 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, } case IP_MSFILTER: { - extern int sysctl_optmem_max; extern int sysctl_igmp_max_msf; struct ip_msfilter *msf; @@ -769,7 +768,6 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, } case MCAST_MSFILTER: { - extern int sysctl_optmem_max; extern int sysctl_igmp_max_msf; struct sockaddr_in *psin; struct ip_msfilter *msf = NULL; @@ -1090,7 +1088,5 @@ int ip_getsockopt(struct sock *sk, int level, int optname, char __user *optval, EXPORT_SYMBOL(ip_cmsg_recv); -#ifdef CONFIG_IP_SCTP_MODULE EXPORT_SYMBOL(ip_getsockopt); EXPORT_SYMBOL(ip_setsockopt); -#endif diff --git a/net/ipv4/ipcomp.c b/net/ipv4/ipcomp.c index 7ded6e60f43a..dcb7ee6c4858 100644 --- a/net/ipv4/ipcomp.c +++ b/net/ipv4/ipcomp.c @@ -214,8 +214,8 @@ static void ipcomp4_err(struct sk_buff *skb, u32 info) spi, IPPROTO_COMP, AF_INET); if (!x) return; - NETDEBUG(printk(KERN_DEBUG "pmtu discovery on SA IPCOMP/%08x/%u.%u.%u.%u\n", - spi, NIPQUAD(iph->daddr))); + NETDEBUG(KERN_DEBUG "pmtu discovery on SA IPCOMP/%08x/%u.%u.%u.%u\n", + spi, NIPQUAD(iph->daddr)); xfrm_state_put(x); } diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index d2bf8e1930a3..63e106605f28 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -393,7 +393,7 @@ static int __init ic_defaults(void) #ifdef IPCONFIG_RARP -static int ic_rarp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt); +static int ic_rarp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev); static struct packet_type rarp_packet_type __initdata = { .type = __constant_htons(ETH_P_RARP), @@ -414,7 +414,7 @@ static inline void ic_rarp_cleanup(void) * Process received RARP packet. */ static int __init -ic_rarp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +ic_rarp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct arphdr *rarp; unsigned char *rarp_ptr; @@ -555,7 +555,7 @@ struct bootp_pkt { /* BOOTP packet format */ #define DHCPRELEASE 7 #define DHCPINFORM 8 -static int ic_bootp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt); +static int ic_bootp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev); static struct packet_type bootp_packet_type __initdata = { .type = __constant_htons(ETH_P_IP), @@ -823,7 +823,7 @@ static void __init ic_do_bootp_ext(u8 *ext) /* * Receive BOOTP reply. */ -static int __init ic_bootp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +static int __init ic_bootp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct bootp_pkt *b; struct iphdr *h; diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c index dc806b578427..9dbf5909f3a6 100644 --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -103,7 +103,7 @@ static DEFINE_SPINLOCK(mfc_unres_lock); In this case data path is free of exclusive locks at all. */ -static kmem_cache_t *mrt_cachep; +static kmem_cache_t *mrt_cachep __read_mostly; static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local); static int ipmr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert); diff --git a/net/ipv4/ipvs/ip_vs_app.c b/net/ipv4/ipvs/ip_vs_app.c index d9212addd193..6e092dadb388 100644 --- a/net/ipv4/ipvs/ip_vs_app.c +++ b/net/ipv4/ipvs/ip_vs_app.c @@ -26,6 +26,7 @@ #include <linux/in.h> #include <linux/ip.h> #include <net/protocol.h> +#include <net/tcp.h> #include <asm/system.h> #include <linux/stat.h> #include <linux/proc_fs.h> diff --git a/net/ipv4/ipvs/ip_vs_conn.c b/net/ipv4/ipvs/ip_vs_conn.c index d0145a8b1551..e11952ea17af 100644 --- a/net/ipv4/ipvs/ip_vs_conn.c +++ b/net/ipv4/ipvs/ip_vs_conn.c @@ -40,7 +40,7 @@ static struct list_head *ip_vs_conn_tab; /* SLAB cache for IPVS connections */ -static kmem_cache_t *ip_vs_conn_cachep; +static kmem_cache_t *ip_vs_conn_cachep __read_mostly; /* counter for current IPVS connections */ static atomic_t ip_vs_conn_count = ATOMIC_INIT(0); diff --git a/net/ipv4/ipvs/ip_vs_core.c b/net/ipv4/ipvs/ip_vs_core.c index 5fb257dd07cb..3ac7eeca04ac 100644 --- a/net/ipv4/ipvs/ip_vs_core.c +++ b/net/ipv4/ipvs/ip_vs_core.c @@ -22,6 +22,7 @@ * * Changes: * Paul `Rusty' Russell properly handle non-linear skbs + * Harald Welte don't use nfcache * */ @@ -529,7 +530,7 @@ static unsigned int ip_vs_post_routing(unsigned int hooknum, const struct net_device *out, int (*okfn)(struct sk_buff *)) { - if (!((*pskb)->nfcache & NFC_IPVS_PROPERTY)) + if (!((*pskb)->ipvs_property)) return NF_ACCEPT; /* The packet was sent from IPVS, exit this chain */ @@ -701,7 +702,7 @@ static int ip_vs_out_icmp(struct sk_buff **pskb, int *related) /* do the statistics and put it back */ ip_vs_out_stats(cp, skb); - skb->nfcache |= NFC_IPVS_PROPERTY; + skb->ipvs_property = 1; verdict = NF_ACCEPT; out: @@ -739,7 +740,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff **pskb, EnterFunction(11); - if (skb->nfcache & NFC_IPVS_PROPERTY) + if (skb->ipvs_property) return NF_ACCEPT; iph = skb->nh.iph; @@ -821,7 +822,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff **pskb, ip_vs_set_state(cp, IP_VS_DIR_OUTPUT, skb, pp); ip_vs_conn_put(cp); - skb->nfcache |= NFC_IPVS_PROPERTY; + skb->ipvs_property = 1; LeaveFunction(11); return NF_ACCEPT; diff --git a/net/ipv4/ipvs/ip_vs_ctl.c b/net/ipv4/ipvs/ip_vs_ctl.c index 7d99ede2ef79..2d66848e7aa0 100644 --- a/net/ipv4/ipvs/ip_vs_ctl.c +++ b/net/ipv4/ipvs/ip_vs_ctl.c @@ -1598,7 +1598,7 @@ static ctl_table vs_table[] = { { .ctl_name = 0 } }; -static ctl_table ipv4_table[] = { +static ctl_table ipvs_ipv4_table[] = { { .ctl_name = NET_IPV4, .procname = "ipv4", @@ -1613,7 +1613,7 @@ static ctl_table vs_root_table[] = { .ctl_name = CTL_NET, .procname = "net", .mode = 0555, - .child = ipv4_table, + .child = ipvs_ipv4_table, }, { .ctl_name = 0 } }; diff --git a/net/ipv4/ipvs/ip_vs_lblc.c b/net/ipv4/ipvs/ip_vs_lblc.c index c035838b780a..561cda326fa8 100644 --- a/net/ipv4/ipvs/ip_vs_lblc.c +++ b/net/ipv4/ipvs/ip_vs_lblc.c @@ -131,7 +131,7 @@ static ctl_table vs_table[] = { { .ctl_name = 0 } }; -static ctl_table ipv4_table[] = { +static ctl_table ipvs_ipv4_table[] = { { .ctl_name = NET_IPV4, .procname = "ipv4", @@ -146,7 +146,7 @@ static ctl_table lblc_root_table[] = { .ctl_name = CTL_NET, .procname = "net", .mode = 0555, - .child = ipv4_table + .child = ipvs_ipv4_table }, { .ctl_name = 0 } }; diff --git a/net/ipv4/ipvs/ip_vs_lblcr.c b/net/ipv4/ipvs/ip_vs_lblcr.c index 22b5dd55d271..ce456dbf09a5 100644 --- a/net/ipv4/ipvs/ip_vs_lblcr.c +++ b/net/ipv4/ipvs/ip_vs_lblcr.c @@ -320,7 +320,7 @@ static ctl_table vs_table[] = { { .ctl_name = 0 } }; -static ctl_table ipv4_table[] = { +static ctl_table ipvs_ipv4_table[] = { { .ctl_name = NET_IPV4, .procname = "ipv4", @@ -335,7 +335,7 @@ static ctl_table lblcr_root_table[] = { .ctl_name = CTL_NET, .procname = "net", .mode = 0555, - .child = ipv4_table + .child = ipvs_ipv4_table }, { .ctl_name = 0 } }; diff --git a/net/ipv4/ipvs/ip_vs_proto_tcp.c b/net/ipv4/ipvs/ip_vs_proto_tcp.c index e65de675da74..c19408973c09 100644 --- a/net/ipv4/ipvs/ip_vs_proto_tcp.c +++ b/net/ipv4/ipvs/ip_vs_proto_tcp.c @@ -604,14 +604,14 @@ void ip_vs_tcp_conn_listen(struct ip_vs_conn *cp) } -static void tcp_init(struct ip_vs_protocol *pp) +static void ip_vs_tcp_init(struct ip_vs_protocol *pp) { IP_VS_INIT_HASH_TABLE(tcp_apps); pp->timeout_table = tcp_timeouts; } -static void tcp_exit(struct ip_vs_protocol *pp) +static void ip_vs_tcp_exit(struct ip_vs_protocol *pp) { } @@ -621,8 +621,8 @@ struct ip_vs_protocol ip_vs_protocol_tcp = { .protocol = IPPROTO_TCP, .dont_defrag = 0, .appcnt = ATOMIC_INIT(0), - .init = tcp_init, - .exit = tcp_exit, + .init = ip_vs_tcp_init, + .exit = ip_vs_tcp_exit, .register_app = tcp_register_app, .unregister_app = tcp_unregister_app, .conn_schedule = tcp_conn_schedule, diff --git a/net/ipv4/ipvs/ip_vs_xmit.c b/net/ipv4/ipvs/ip_vs_xmit.c index a8512a3fd08a..3b87482049cf 100644 --- a/net/ipv4/ipvs/ip_vs_xmit.c +++ b/net/ipv4/ipvs/ip_vs_xmit.c @@ -127,7 +127,7 @@ ip_vs_dst_reset(struct ip_vs_dest *dest) #define IP_VS_XMIT(skb, rt) \ do { \ - (skb)->nfcache |= NFC_IPVS_PROPERTY; \ + (skb)->ipvs_property = 1; \ (skb)->ip_summed = CHECKSUM_NONE; \ NF_HOOK(PF_INET, NF_IP_LOCAL_OUT, (skb), NULL, \ (rt)->u.dst.dev, dst_output); \ diff --git a/net/ipv4/multipath_drr.c b/net/ipv4/multipath_drr.c index c9cf8726051d..db67373f9b34 100644 --- a/net/ipv4/multipath_drr.c +++ b/net/ipv4/multipath_drr.c @@ -107,7 +107,7 @@ static int drr_dev_event(struct notifier_block *this, return NOTIFY_DONE; } -struct notifier_block drr_dev_notifier = { +static struct notifier_block drr_dev_notifier = { .notifier_call = drr_dev_event, }; diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c new file mode 100644 index 000000000000..ae0779d82c5d --- /dev/null +++ b/net/ipv4/netfilter.c @@ -0,0 +1,139 @@ +/* IPv4 specific functions of netfilter core */ + +#include <linux/config.h> +#ifdef CONFIG_NETFILTER + +#include <linux/kernel.h> +#include <linux/netfilter.h> +#include <linux/netfilter_ipv4.h> + +#include <linux/tcp.h> +#include <linux/udp.h> +#include <linux/icmp.h> +#include <net/route.h> +#include <linux/ip.h> + +/* route_me_harder function, used by iptable_nat, iptable_mangle + ip_queue */ +int ip_route_me_harder(struct sk_buff **pskb) +{ + struct iphdr *iph = (*pskb)->nh.iph; + struct rtable *rt; + struct flowi fl = {}; + struct dst_entry *odst; + unsigned int hh_len; + + /* some non-standard hacks like ipt_REJECT.c:send_reset() can cause + * packets with foreign saddr to appear on the NF_IP_LOCAL_OUT hook. + */ + if (inet_addr_type(iph->saddr) == RTN_LOCAL) { + fl.nl_u.ip4_u.daddr = iph->daddr; + fl.nl_u.ip4_u.saddr = iph->saddr; + fl.nl_u.ip4_u.tos = RT_TOS(iph->tos); + fl.oif = (*pskb)->sk ? (*pskb)->sk->sk_bound_dev_if : 0; +#ifdef CONFIG_IP_ROUTE_FWMARK + fl.nl_u.ip4_u.fwmark = (*pskb)->nfmark; +#endif + fl.proto = iph->protocol; + if (ip_route_output_key(&rt, &fl) != 0) + return -1; + + /* Drop old route. */ + dst_release((*pskb)->dst); + (*pskb)->dst = &rt->u.dst; + } else { + /* non-local src, find valid iif to satisfy + * rp-filter when calling ip_route_input. */ + fl.nl_u.ip4_u.daddr = iph->saddr; + if (ip_route_output_key(&rt, &fl) != 0) + return -1; + + odst = (*pskb)->dst; + if (ip_route_input(*pskb, iph->daddr, iph->saddr, + RT_TOS(iph->tos), rt->u.dst.dev) != 0) { + dst_release(&rt->u.dst); + return -1; + } + dst_release(&rt->u.dst); + dst_release(odst); + } + + if ((*pskb)->dst->error) + return -1; + + /* Change in oif may mean change in hh_len. */ + hh_len = (*pskb)->dst->dev->hard_header_len; + if (skb_headroom(*pskb) < hh_len) { + struct sk_buff *nskb; + + nskb = skb_realloc_headroom(*pskb, hh_len); + if (!nskb) + return -1; + if ((*pskb)->sk) + skb_set_owner_w(nskb, (*pskb)->sk); + kfree_skb(*pskb); + *pskb = nskb; + } + + return 0; +} +EXPORT_SYMBOL(ip_route_me_harder); + +/* + * Extra routing may needed on local out, as the QUEUE target never + * returns control to the table. + */ + +struct ip_rt_info { + u_int32_t daddr; + u_int32_t saddr; + u_int8_t tos; +}; + +static void queue_save(const struct sk_buff *skb, struct nf_info *info) +{ + struct ip_rt_info *rt_info = nf_info_reroute(info); + + if (info->hook == NF_IP_LOCAL_OUT) { + const struct iphdr *iph = skb->nh.iph; + + rt_info->tos = iph->tos; + rt_info->daddr = iph->daddr; + rt_info->saddr = iph->saddr; + } +} + +static int queue_reroute(struct sk_buff **pskb, const struct nf_info *info) +{ + const struct ip_rt_info *rt_info = nf_info_reroute(info); + + if (info->hook == NF_IP_LOCAL_OUT) { + struct iphdr *iph = (*pskb)->nh.iph; + + if (!(iph->tos == rt_info->tos + && iph->daddr == rt_info->daddr + && iph->saddr == rt_info->saddr)) + return ip_route_me_harder(pskb); + } + return 0; +} + +static struct nf_queue_rerouter ip_reroute = { + .rer_size = sizeof(struct ip_rt_info), + .save = queue_save, + .reroute = queue_reroute, +}; + +static int init(void) +{ + return nf_register_queue_rerouter(PF_INET, &ip_reroute); +} + +static void fini(void) +{ + nf_unregister_queue_rerouter(PF_INET); +} + +module_init(init); +module_exit(fini); + +#endif /* CONFIG_NETFILTER */ diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig index 46d4cb1c06f0..e046f5521814 100644 --- a/net/ipv4/netfilter/Kconfig +++ b/net/ipv4/netfilter/Kconfig @@ -40,6 +40,16 @@ config IP_NF_CONNTRACK_MARK of packets, but this mark value is kept in the conntrack session instead of the individual packets. +config IP_NF_CONNTRACK_EVENTS + bool "Connection tracking events" + depends on IP_NF_CONNTRACK + help + If this option is enabled, the connection tracking code will + provide a notifier chain that can be used by other kernel code + to get notified about changes in the connection tracking state. + + IF unsure, say `N'. + config IP_NF_CT_PROTO_SCTP tristate 'SCTP protocol connection tracking support (EXPERIMENTAL)' depends on IP_NF_CONNTRACK && EXPERIMENTAL @@ -100,11 +110,15 @@ config IP_NF_AMANDA To compile it as a module, choose M here. If unsure, say Y. config IP_NF_QUEUE - tristate "Userspace queueing via NETLINK" + tristate "IP Userspace queueing via NETLINK (OBSOLETE)" help Netfilter has the ability to queue packets to user space: the netlink device can be used to access them using this driver. + This option enables the old IPv4-only "ip_queue" implementation + which has been obsoleted by the new "nfnetlink_queue" code (see + CONFIG_NETFILTER_NETLINK_QUEUE). + To compile it as a module, choose M here. If unsure, say N. config IP_NF_IPTABLES @@ -340,6 +354,17 @@ config IP_NF_MATCH_SCTP If you want to compile it as a module, say M here and read <file:Documentation/modules.txt>. If unsure, say `N'. +config IP_NF_MATCH_DCCP + tristate 'DCCP protocol match support' + depends on IP_NF_IPTABLES + help + With this option enabled, you will be able to use the iptables + `dccp' match in order to match on DCCP source/destination ports + and DCCP flags. + + If you want to compile it as a module, say M here and read + <file:Documentation/modules.txt>. If unsure, say `N'. + config IP_NF_MATCH_COMMENT tristate 'comment match support' depends on IP_NF_IPTABLES @@ -361,6 +386,16 @@ config IP_NF_MATCH_CONNMARK <file:Documentation/modules.txt>. The module will be called ipt_connmark.o. If unsure, say `N'. +config IP_NF_MATCH_CONNBYTES + tristate 'Connection byte/packet counter match support' + depends on IP_NF_CT_ACCT && IP_NF_IPTABLES + help + This option adds a `connbytes' match, which allows you to match the + number of bytes and/or packets for each direction within a connection. + + If you want to compile it as a module, say M here and read + <file:Documentation/modules.txt>. If unsure, say `N'. + config IP_NF_MATCH_HASHLIMIT tristate 'hashlimit match support' depends on IP_NF_IPTABLES @@ -375,6 +410,19 @@ config IP_NF_MATCH_HASHLIMIT destination IP' or `500pps from any given source IP' with a single IPtables rule. +config IP_NF_MATCH_STRING + tristate 'string match support' + depends on IP_NF_IPTABLES + select TEXTSEARCH + select TEXTSEARCH_KMP + select TEXTSEARCH_BM + select TEXTSEARCH_FSM + help + This option adds a `string' match, which allows you to look for + pattern matchings in packets. + + To compile it as a module, choose M here. If unsure, say N. + # `filter', generic and specific targets config IP_NF_FILTER tristate "Packet filtering" @@ -616,6 +664,20 @@ config IP_NF_TARGET_CLASSIFY To compile it as a module, choose M here. If unsure, say N. +config IP_NF_TARGET_TTL + tristate 'TTL target support' + depends on IP_NF_MANGLE + help + This option adds a `TTL' target, which enables the user to modify + the TTL value of the IP header. + + While it is safe to decrement/lower the TTL, this target also enables + functionality to increment and set the TTL value of the IP header to + arbitrary values. This is EXTREMELY DANGEROUS since you can easily + create immortal packets that loop forever on the network. + + To compile it as a module, choose M here. If unsure, say N. + config IP_NF_TARGET_CONNMARK tristate 'CONNMARK target support' depends on IP_NF_CONNTRACK_MARK && IP_NF_MANGLE @@ -692,5 +754,11 @@ config IP_NF_ARP_MANGLE Allows altering the ARP packet payload: source and destination hardware and network addresses. +config IP_NF_CONNTRACK_NETLINK + tristate 'Connection tracking netlink interface' + depends on IP_NF_CONNTRACK && NETFILTER_NETLINK + help + This option enables support for a netlink-based userspace interface + endmenu diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile index 45796d5924dd..a7bd38f50522 100644 --- a/net/ipv4/netfilter/Makefile +++ b/net/ipv4/netfilter/Makefile @@ -9,6 +9,10 @@ iptable_nat-objs := ip_nat_standalone.o ip_nat_rule.o ip_nat_core.o ip_nat_helpe # connection tracking obj-$(CONFIG_IP_NF_CONNTRACK) += ip_conntrack.o +# conntrack netlink interface +obj-$(CONFIG_IP_NF_CONNTRACK_NETLINK) += ip_conntrack_netlink.o + + # SCTP protocol connection tracking obj-$(CONFIG_IP_NF_CT_PROTO_SCTP) += ip_conntrack_proto_sctp.o @@ -38,6 +42,7 @@ obj-$(CONFIG_IP_NF_MATCH_HELPER) += ipt_helper.o obj-$(CONFIG_IP_NF_MATCH_LIMIT) += ipt_limit.o obj-$(CONFIG_IP_NF_MATCH_HASHLIMIT) += ipt_hashlimit.o obj-$(CONFIG_IP_NF_MATCH_SCTP) += ipt_sctp.o +obj-$(CONFIG_IP_NF_MATCH_DCCP) += ipt_dccp.o obj-$(CONFIG_IP_NF_MATCH_MARK) += ipt_mark.o obj-$(CONFIG_IP_NF_MATCH_MAC) += ipt_mac.o obj-$(CONFIG_IP_NF_MATCH_IPRANGE) += ipt_iprange.o @@ -54,11 +59,13 @@ obj-$(CONFIG_IP_NF_MATCH_TTL) += ipt_ttl.o obj-$(CONFIG_IP_NF_MATCH_STATE) += ipt_state.o obj-$(CONFIG_IP_NF_MATCH_CONNMARK) += ipt_connmark.o obj-$(CONFIG_IP_NF_MATCH_CONNTRACK) += ipt_conntrack.o +obj-$(CONFIG_IP_NF_MATCH_CONNBYTES) += ipt_connbytes.o obj-$(CONFIG_IP_NF_MATCH_TCPMSS) += ipt_tcpmss.o obj-$(CONFIG_IP_NF_MATCH_REALM) += ipt_realm.o obj-$(CONFIG_IP_NF_MATCH_ADDRTYPE) += ipt_addrtype.o obj-$(CONFIG_IP_NF_MATCH_PHYSDEV) += ipt_physdev.o obj-$(CONFIG_IP_NF_MATCH_COMMENT) += ipt_comment.o +obj-$(CONFIG_IP_NF_MATCH_STRING) += ipt_string.o # targets obj-$(CONFIG_IP_NF_TARGET_REJECT) += ipt_REJECT.o @@ -78,6 +85,7 @@ obj-$(CONFIG_IP_NF_TARGET_ULOG) += ipt_ULOG.o obj-$(CONFIG_IP_NF_TARGET_TCPMSS) += ipt_TCPMSS.o obj-$(CONFIG_IP_NF_TARGET_NOTRACK) += ipt_NOTRACK.o obj-$(CONFIG_IP_NF_TARGET_CLUSTERIP) += ipt_CLUSTERIP.o +obj-$(CONFIG_IP_NF_TARGET_TTL) += ipt_TTL.o # generic ARP tables obj-$(CONFIG_IP_NF_ARPTABLES) += arp_tables.o @@ -87,3 +95,4 @@ obj-$(CONFIG_IP_NF_ARP_MANGLE) += arpt_mangle.o obj-$(CONFIG_IP_NF_ARPFILTER) += arptable_filter.o obj-$(CONFIG_IP_NF_QUEUE) += ip_queue.o +obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += ipt_NFQUEUE.o diff --git a/net/ipv4/netfilter/ip_conntrack_amanda.c b/net/ipv4/netfilter/ip_conntrack_amanda.c index 01e1b58322a9..be4c9eb3243f 100644 --- a/net/ipv4/netfilter/ip_conntrack_amanda.c +++ b/net/ipv4/netfilter/ip_conntrack_amanda.c @@ -40,7 +40,7 @@ MODULE_PARM_DESC(master_timeout, "timeout for the master connection"); static char *conns[] = { "DATA ", "MESG ", "INDEX " }; /* This is slow, but it's simple. --RR */ -static char amanda_buffer[65536]; +static char *amanda_buffer; static DEFINE_SPINLOCK(amanda_buffer_lock); unsigned int (*ip_nat_amanda_hook)(struct sk_buff **pskb, @@ -153,11 +153,25 @@ static struct ip_conntrack_helper amanda_helper = { static void __exit fini(void) { ip_conntrack_helper_unregister(&amanda_helper); + kfree(amanda_buffer); } static int __init init(void) { - return ip_conntrack_helper_register(&amanda_helper); + int ret; + + amanda_buffer = kmalloc(65536, GFP_KERNEL); + if (!amanda_buffer) + return -ENOMEM; + + ret = ip_conntrack_helper_register(&amanda_helper); + if (ret < 0) { + kfree(amanda_buffer); + return ret; + } + return 0; + + } module_init(init); diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c index a7f0c821a9b2..a0648600190e 100644 --- a/net/ipv4/netfilter/ip_conntrack_core.c +++ b/net/ipv4/netfilter/ip_conntrack_core.c @@ -37,6 +37,7 @@ #include <linux/err.h> #include <linux/percpu.h> #include <linux/moduleparam.h> +#include <linux/notifier.h> /* ip_conntrack_lock protects the main hash table, protocol/helper/expected registrations, conntrack timers*/ @@ -49,7 +50,7 @@ #include <linux/netfilter_ipv4/ip_conntrack_core.h> #include <linux/netfilter_ipv4/listhelp.h> -#define IP_CONNTRACK_VERSION "2.1" +#define IP_CONNTRACK_VERSION "2.3" #if 0 #define DEBUGP printk @@ -69,22 +70,81 @@ static LIST_HEAD(helpers); unsigned int ip_conntrack_htable_size = 0; int ip_conntrack_max; struct list_head *ip_conntrack_hash; -static kmem_cache_t *ip_conntrack_cachep; -static kmem_cache_t *ip_conntrack_expect_cachep; +static kmem_cache_t *ip_conntrack_cachep __read_mostly; +static kmem_cache_t *ip_conntrack_expect_cachep __read_mostly; struct ip_conntrack ip_conntrack_untracked; unsigned int ip_ct_log_invalid; static LIST_HEAD(unconfirmed); static int ip_conntrack_vmalloc; -DEFINE_PER_CPU(struct ip_conntrack_stat, ip_conntrack_stat); +static unsigned int ip_conntrack_next_id = 1; +static unsigned int ip_conntrack_expect_next_id = 1; +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS +struct notifier_block *ip_conntrack_chain; +struct notifier_block *ip_conntrack_expect_chain; + +DEFINE_PER_CPU(struct ip_conntrack_ecache, ip_conntrack_ecache); -void -ip_conntrack_put(struct ip_conntrack *ct) +/* deliver cached events and clear cache entry - must be called with locally + * disabled softirqs */ +static inline void +__ip_ct_deliver_cached_events(struct ip_conntrack_ecache *ecache) { - IP_NF_ASSERT(ct); - nf_conntrack_put(&ct->ct_general); + DEBUGP("ecache: delivering events for %p\n", ecache->ct); + if (is_confirmed(ecache->ct) && !is_dying(ecache->ct) && ecache->events) + notifier_call_chain(&ip_conntrack_chain, ecache->events, + ecache->ct); + ecache->events = 0; + ip_conntrack_put(ecache->ct); + ecache->ct = NULL; } +/* Deliver all cached events for a particular conntrack. This is called + * by code prior to async packet handling or freeing the skb */ +void ip_ct_deliver_cached_events(const struct ip_conntrack *ct) +{ + struct ip_conntrack_ecache *ecache; + + local_bh_disable(); + ecache = &__get_cpu_var(ip_conntrack_ecache); + if (ecache->ct == ct) + __ip_ct_deliver_cached_events(ecache); + local_bh_enable(); +} + +void __ip_ct_event_cache_init(struct ip_conntrack *ct) +{ + struct ip_conntrack_ecache *ecache; + + /* take care of delivering potentially old events */ + ecache = &__get_cpu_var(ip_conntrack_ecache); + BUG_ON(ecache->ct == ct); + if (ecache->ct) + __ip_ct_deliver_cached_events(ecache); + /* initialize for this conntrack/packet */ + ecache->ct = ct; + nf_conntrack_get(&ct->ct_general); +} + +/* flush the event cache - touches other CPU's data and must not be called while + * packets are still passing through the code */ +static void ip_ct_event_cache_flush(void) +{ + struct ip_conntrack_ecache *ecache; + int cpu; + + for_each_cpu(cpu) { + ecache = &per_cpu(ip_conntrack_ecache, cpu); + if (ecache->ct) + ip_conntrack_put(ecache->ct); + } +} +#else +static inline void ip_ct_event_cache_flush(void) {} +#endif /* CONFIG_IP_NF_CONNTRACK_EVENTS */ + +DEFINE_PER_CPU(struct ip_conntrack_stat, ip_conntrack_stat); + static int ip_conntrack_hash_rnd_initted; static unsigned int ip_conntrack_hash_rnd; @@ -144,6 +204,13 @@ static void unlink_expect(struct ip_conntrack_expect *exp) list_del(&exp->list); CONNTRACK_STAT_INC(expect_delete); exp->master->expecting--; + ip_conntrack_expect_put(exp); +} + +void __ip_ct_expect_unlink_destroy(struct ip_conntrack_expect *exp) +{ + unlink_expect(exp); + ip_conntrack_expect_put(exp); } static void expectation_timed_out(unsigned long ul_expect) @@ -156,6 +223,33 @@ static void expectation_timed_out(unsigned long ul_expect) ip_conntrack_expect_put(exp); } +struct ip_conntrack_expect * +__ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple) +{ + struct ip_conntrack_expect *i; + + list_for_each_entry(i, &ip_conntrack_expect_list, list) { + if (ip_ct_tuple_mask_cmp(tuple, &i->tuple, &i->mask)) { + atomic_inc(&i->use); + return i; + } + } + return NULL; +} + +/* Just find a expectation corresponding to a tuple. */ +struct ip_conntrack_expect * +ip_conntrack_expect_find_get(const struct ip_conntrack_tuple *tuple) +{ + struct ip_conntrack_expect *i; + + read_lock_bh(&ip_conntrack_lock); + i = __ip_conntrack_expect_find(tuple); + read_unlock_bh(&ip_conntrack_lock); + + return i; +} + /* If an expectation for this connection is found, it gets delete from * global list then returned. */ static struct ip_conntrack_expect * @@ -180,7 +274,7 @@ find_expectation(const struct ip_conntrack_tuple *tuple) } /* delete all expectations for this conntrack */ -static void remove_expectations(struct ip_conntrack *ct) +void ip_ct_remove_expectations(struct ip_conntrack *ct) { struct ip_conntrack_expect *i, *tmp; @@ -210,7 +304,7 @@ clean_from_lists(struct ip_conntrack *ct) LIST_DELETE(&ip_conntrack_hash[hr], &ct->tuplehash[IP_CT_DIR_REPLY]); /* Destroy all pending expectations */ - remove_expectations(ct); + ip_ct_remove_expectations(ct); } static void @@ -223,10 +317,13 @@ destroy_conntrack(struct nf_conntrack *nfct) IP_NF_ASSERT(atomic_read(&nfct->use) == 0); IP_NF_ASSERT(!timer_pending(&ct->timeout)); + ip_conntrack_event(IPCT_DESTROY, ct); + set_bit(IPS_DYING_BIT, &ct->status); + /* To make sure we don't get any weird locking issues here: * destroy_conntrack() MUST NOT be called with a write lock * to ip_conntrack_lock!!! -HW */ - proto = ip_ct_find_proto(ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum); + proto = __ip_conntrack_proto_find(ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum); if (proto && proto->destroy) proto->destroy(ct); @@ -238,7 +335,7 @@ destroy_conntrack(struct nf_conntrack *nfct) * except TFTP can create an expectation on the first packet, * before connection is in the list, so we need to clean here, * too. */ - remove_expectations(ct); + ip_ct_remove_expectations(ct); /* We overload first tuple to link into unconfirmed list. */ if (!is_confirmed(ct)) { @@ -253,8 +350,7 @@ destroy_conntrack(struct nf_conntrack *nfct) ip_conntrack_put(ct->master); DEBUGP("destroy_conntrack: returning ct=%p to slab\n", ct); - kmem_cache_free(ip_conntrack_cachep, ct); - atomic_dec(&ip_conntrack_count); + ip_conntrack_free(ct); } static void death_by_timeout(unsigned long ul_conntrack) @@ -280,7 +376,7 @@ conntrack_tuple_cmp(const struct ip_conntrack_tuple_hash *i, && ip_ct_tuple_equal(tuple, &i->tuple); } -static struct ip_conntrack_tuple_hash * +struct ip_conntrack_tuple_hash * __ip_conntrack_find(const struct ip_conntrack_tuple *tuple, const struct ip_conntrack *ignored_conntrack) { @@ -315,6 +411,29 @@ ip_conntrack_find_get(const struct ip_conntrack_tuple *tuple, return h; } +static void __ip_conntrack_hash_insert(struct ip_conntrack *ct, + unsigned int hash, + unsigned int repl_hash) +{ + ct->id = ++ip_conntrack_next_id; + list_prepend(&ip_conntrack_hash[hash], + &ct->tuplehash[IP_CT_DIR_ORIGINAL].list); + list_prepend(&ip_conntrack_hash[repl_hash], + &ct->tuplehash[IP_CT_DIR_REPLY].list); +} + +void ip_conntrack_hash_insert(struct ip_conntrack *ct) +{ + unsigned int hash, repl_hash; + + hash = hash_conntrack(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); + repl_hash = hash_conntrack(&ct->tuplehash[IP_CT_DIR_REPLY].tuple); + + write_lock_bh(&ip_conntrack_lock); + __ip_conntrack_hash_insert(ct, hash, repl_hash); + write_unlock_bh(&ip_conntrack_lock); +} + /* Confirm a connection given skb; places it in hash table */ int __ip_conntrack_confirm(struct sk_buff **pskb) @@ -361,10 +480,7 @@ __ip_conntrack_confirm(struct sk_buff **pskb) /* Remove from unconfirmed list */ list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list); - list_prepend(&ip_conntrack_hash[hash], - &ct->tuplehash[IP_CT_DIR_ORIGINAL]); - list_prepend(&ip_conntrack_hash[repl_hash], - &ct->tuplehash[IP_CT_DIR_REPLY]); + __ip_conntrack_hash_insert(ct, hash, repl_hash); /* Timer relative to confirmation time, not original setting time, otherwise we'd get timer wrap in weird delay cases. */ @@ -374,6 +490,16 @@ __ip_conntrack_confirm(struct sk_buff **pskb) set_bit(IPS_CONFIRMED_BIT, &ct->status); CONNTRACK_STAT_INC(insert); write_unlock_bh(&ip_conntrack_lock); + if (ct->helper) + ip_conntrack_event_cache(IPCT_HELPER, *pskb); +#ifdef CONFIG_IP_NF_NAT_NEEDED + if (test_bit(IPS_SRC_NAT_DONE_BIT, &ct->status) || + test_bit(IPS_DST_NAT_DONE_BIT, &ct->status)) + ip_conntrack_event_cache(IPCT_NATINFO, *pskb); +#endif + ip_conntrack_event_cache(master_ct(ct) ? + IPCT_RELATED : IPCT_NEW, *pskb); + return NF_ACCEPT; } @@ -438,34 +564,84 @@ static inline int helper_cmp(const struct ip_conntrack_helper *i, return ip_ct_tuple_mask_cmp(rtuple, &i->tuple, &i->mask); } -static struct ip_conntrack_helper *ip_ct_find_helper(const struct ip_conntrack_tuple *tuple) +static struct ip_conntrack_helper * +__ip_conntrack_helper_find( const struct ip_conntrack_tuple *tuple) { return LIST_FIND(&helpers, helper_cmp, struct ip_conntrack_helper *, tuple); } -/* Allocate a new conntrack: we return -ENOMEM if classification - failed due to stress. Otherwise it really is unclassifiable. */ -static struct ip_conntrack_tuple_hash * -init_conntrack(const struct ip_conntrack_tuple *tuple, - struct ip_conntrack_protocol *protocol, - struct sk_buff *skb) +struct ip_conntrack_helper * +ip_conntrack_helper_find_get( const struct ip_conntrack_tuple *tuple) +{ + struct ip_conntrack_helper *helper; + + /* need ip_conntrack_lock to assure that helper exists until + * try_module_get() is called */ + read_lock_bh(&ip_conntrack_lock); + + helper = __ip_conntrack_helper_find(tuple); + if (helper) { + /* need to increase module usage count to assure helper will + * not go away while the caller is e.g. busy putting a + * conntrack in the hash that uses the helper */ + if (!try_module_get(helper->me)) + helper = NULL; + } + + read_unlock_bh(&ip_conntrack_lock); + + return helper; +} + +void ip_conntrack_helper_put(struct ip_conntrack_helper *helper) +{ + module_put(helper->me); +} + +struct ip_conntrack_protocol * +__ip_conntrack_proto_find(u_int8_t protocol) +{ + return ip_ct_protos[protocol]; +} + +/* this is guaranteed to always return a valid protocol helper, since + * it falls back to generic_protocol */ +struct ip_conntrack_protocol * +ip_conntrack_proto_find_get(u_int8_t protocol) +{ + struct ip_conntrack_protocol *p; + + preempt_disable(); + p = __ip_conntrack_proto_find(protocol); + if (p) { + if (!try_module_get(p->me)) + p = &ip_conntrack_generic_protocol; + } + preempt_enable(); + + return p; +} + +void ip_conntrack_proto_put(struct ip_conntrack_protocol *p) +{ + module_put(p->me); +} + +struct ip_conntrack *ip_conntrack_alloc(struct ip_conntrack_tuple *orig, + struct ip_conntrack_tuple *repl) { struct ip_conntrack *conntrack; - struct ip_conntrack_tuple repl_tuple; - size_t hash; - struct ip_conntrack_expect *exp; if (!ip_conntrack_hash_rnd_initted) { get_random_bytes(&ip_conntrack_hash_rnd, 4); ip_conntrack_hash_rnd_initted = 1; } - hash = hash_conntrack(tuple); - if (ip_conntrack_max && atomic_read(&ip_conntrack_count) >= ip_conntrack_max) { + unsigned int hash = hash_conntrack(orig); /* Try dropping from this hash chain. */ if (!early_drop(&ip_conntrack_hash[hash])) { if (net_ratelimit()) @@ -476,11 +652,6 @@ init_conntrack(const struct ip_conntrack_tuple *tuple, } } - if (!ip_ct_invert_tuple(&repl_tuple, tuple, protocol)) { - DEBUGP("Can't invert tuple.\n"); - return NULL; - } - conntrack = kmem_cache_alloc(ip_conntrack_cachep, GFP_ATOMIC); if (!conntrack) { DEBUGP("Can't allocate conntrack.\n"); @@ -490,17 +661,50 @@ init_conntrack(const struct ip_conntrack_tuple *tuple, memset(conntrack, 0, sizeof(*conntrack)); atomic_set(&conntrack->ct_general.use, 1); conntrack->ct_general.destroy = destroy_conntrack; - conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple = *tuple; - conntrack->tuplehash[IP_CT_DIR_REPLY].tuple = repl_tuple; - if (!protocol->new(conntrack, skb)) { - kmem_cache_free(ip_conntrack_cachep, conntrack); - return NULL; - } + conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple = *orig; + conntrack->tuplehash[IP_CT_DIR_REPLY].tuple = *repl; /* Don't set timer yet: wait for confirmation */ init_timer(&conntrack->timeout); conntrack->timeout.data = (unsigned long)conntrack; conntrack->timeout.function = death_by_timeout; + atomic_inc(&ip_conntrack_count); + + return conntrack; +} + +void +ip_conntrack_free(struct ip_conntrack *conntrack) +{ + atomic_dec(&ip_conntrack_count); + kmem_cache_free(ip_conntrack_cachep, conntrack); +} + +/* Allocate a new conntrack: we return -ENOMEM if classification + * failed due to stress. Otherwise it really is unclassifiable */ +static struct ip_conntrack_tuple_hash * +init_conntrack(struct ip_conntrack_tuple *tuple, + struct ip_conntrack_protocol *protocol, + struct sk_buff *skb) +{ + struct ip_conntrack *conntrack; + struct ip_conntrack_tuple repl_tuple; + struct ip_conntrack_expect *exp; + + if (!ip_ct_invert_tuple(&repl_tuple, tuple, protocol)) { + DEBUGP("Can't invert tuple.\n"); + return NULL; + } + + conntrack = ip_conntrack_alloc(tuple, &repl_tuple); + if (conntrack == NULL || IS_ERR(conntrack)) + return (struct ip_conntrack_tuple_hash *)conntrack; + + if (!protocol->new(conntrack, skb)) { + ip_conntrack_free(conntrack); + return NULL; + } + write_lock_bh(&ip_conntrack_lock); exp = find_expectation(tuple); @@ -521,7 +725,7 @@ init_conntrack(const struct ip_conntrack_tuple *tuple, nf_conntrack_get(&conntrack->master->ct_general); CONNTRACK_STAT_INC(expect_new); } else { - conntrack->helper = ip_ct_find_helper(&repl_tuple); + conntrack->helper = __ip_conntrack_helper_find(&repl_tuple); CONNTRACK_STAT_INC(new); } @@ -529,7 +733,6 @@ init_conntrack(const struct ip_conntrack_tuple *tuple, /* Overload tuple linked list to put us in unconfirmed list. */ list_add(&conntrack->tuplehash[IP_CT_DIR_ORIGINAL].list, &unconfirmed); - atomic_inc(&ip_conntrack_count); write_unlock_bh(&ip_conntrack_lock); if (exp) { @@ -607,7 +810,7 @@ unsigned int ip_conntrack_in(unsigned int hooknum, struct ip_conntrack *ct; enum ip_conntrack_info ctinfo; struct ip_conntrack_protocol *proto; - int set_reply; + int set_reply = 0; int ret; /* Previously seen (loopback or untracked)? Ignore. */ @@ -625,9 +828,6 @@ unsigned int ip_conntrack_in(unsigned int hooknum, return NF_DROP; } - /* FIXME: Do this right please. --RR */ - (*pskb)->nfcache |= NFC_UNKNOWN; - /* Doesn't cover locally-generated broadcast, so not worth it. */ #if 0 /* Ignore broadcast: no `connection'. */ @@ -643,7 +843,7 @@ unsigned int ip_conntrack_in(unsigned int hooknum, } #endif - proto = ip_ct_find_proto((*pskb)->nh.iph->protocol); + proto = __ip_conntrack_proto_find((*pskb)->nh.iph->protocol); /* It may be an special packet, error, unclean... * inverse of the return code tells to the netfilter @@ -679,8 +879,8 @@ unsigned int ip_conntrack_in(unsigned int hooknum, return -ret; } - if (set_reply) - set_bit(IPS_SEEN_REPLY_BIT, &ct->status); + if (set_reply && !test_and_set_bit(IPS_SEEN_REPLY_BIT, &ct->status)) + ip_conntrack_event_cache(IPCT_STATUS, *pskb); return ret; } @@ -689,7 +889,7 @@ int invert_tuplepr(struct ip_conntrack_tuple *inverse, const struct ip_conntrack_tuple *orig) { return ip_ct_invert_tuple(inverse, orig, - ip_ct_find_proto(orig->dst.protonum)); + __ip_conntrack_proto_find(orig->dst.protonum)); } /* Would two expected things clash? */ @@ -769,6 +969,8 @@ static void ip_conntrack_expect_insert(struct ip_conntrack_expect *exp) exp->timeout.expires = jiffies + exp->master->helper->timeout * HZ; add_timer(&exp->timeout); + exp->id = ++ip_conntrack_expect_next_id; + atomic_inc(&exp->use); CONNTRACK_STAT_INC(expect_create); } @@ -827,6 +1029,7 @@ int ip_conntrack_expect_related(struct ip_conntrack_expect *expect) evict_oldest_expect(expect->master); ip_conntrack_expect_insert(expect); + ip_conntrack_expect_event(IPEXP_NEW, expect); ret = 0; out: write_unlock_bh(&ip_conntrack_lock); @@ -847,7 +1050,7 @@ void ip_conntrack_alter_reply(struct ip_conntrack *conntrack, conntrack->tuplehash[IP_CT_DIR_REPLY].tuple = *newreply; if (!conntrack->master && conntrack->expecting == 0) - conntrack->helper = ip_ct_find_helper(newreply); + conntrack->helper = __ip_conntrack_helper_find(newreply); write_unlock_bh(&ip_conntrack_lock); } @@ -861,11 +1064,26 @@ int ip_conntrack_helper_register(struct ip_conntrack_helper *me) return 0; } +struct ip_conntrack_helper * +__ip_conntrack_helper_find_byname(const char *name) +{ + struct ip_conntrack_helper *h; + + list_for_each_entry(h, &helpers, list) { + if (!strcmp(h->name, name)) + return h; + } + + return NULL; +} + static inline int unhelp(struct ip_conntrack_tuple_hash *i, const struct ip_conntrack_helper *me) { - if (tuplehash_to_ctrack(i)->helper == me) + if (tuplehash_to_ctrack(i)->helper == me) { + ip_conntrack_event(IPCT_HELPER, tuplehash_to_ctrack(i)); tuplehash_to_ctrack(i)->helper = NULL; + } return 0; } @@ -927,12 +1145,46 @@ void ip_ct_refresh_acct(struct ip_conntrack *ct, if (del_timer(&ct->timeout)) { ct->timeout.expires = jiffies + extra_jiffies; add_timer(&ct->timeout); + ip_conntrack_event_cache(IPCT_REFRESH, skb); } ct_add_counters(ct, ctinfo, skb); write_unlock_bh(&ip_conntrack_lock); } } +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) +/* Generic function for tcp/udp/sctp/dccp and alike. This needs to be + * in ip_conntrack_core, since we don't want the protocols to autoload + * or depend on ctnetlink */ +int ip_ct_port_tuple_to_nfattr(struct sk_buff *skb, + const struct ip_conntrack_tuple *tuple) +{ + NFA_PUT(skb, CTA_PROTO_SRC_PORT, sizeof(u_int16_t), + &tuple->src.u.tcp.port); + NFA_PUT(skb, CTA_PROTO_DST_PORT, sizeof(u_int16_t), + &tuple->dst.u.tcp.port); + return 0; + +nfattr_failure: + return -1; +} + +int ip_ct_port_nfattr_to_tuple(struct nfattr *tb[], + struct ip_conntrack_tuple *t) +{ + if (!tb[CTA_PROTO_SRC_PORT-1] || !tb[CTA_PROTO_DST_PORT-1]) + return -EINVAL; + + t->src.u.tcp.port = + *(u_int16_t *)NFA_DATA(tb[CTA_PROTO_SRC_PORT-1]); + t->dst.u.tcp.port = + *(u_int16_t *)NFA_DATA(tb[CTA_PROTO_DST_PORT-1]); + + return 0; +} +#endif + /* Returns new sk_buff, or NULL */ struct sk_buff * ip_ct_gather_frags(struct sk_buff *skb, u_int32_t user) @@ -943,10 +1195,8 @@ ip_ct_gather_frags(struct sk_buff *skb, u_int32_t user) skb = ip_defrag(skb, user); local_bh_enable(); - if (skb) { + if (skb) ip_send_check(skb->nh.iph); - skb->nfcache |= NFC_ALTERED; - } return skb; } @@ -1096,16 +1346,14 @@ static void free_conntrack_hash(void) * ip_conntrack_htable_size)); } -/* Mishearing the voices in his head, our hero wonders how he's - supposed to kill the mall. */ -void ip_conntrack_cleanup(void) +void ip_conntrack_flush() { - ip_ct_attach = NULL; /* This makes sure all current packets have passed through netfilter framework. Roll on, two-stage module delete... */ synchronize_net(); - + + ip_ct_event_cache_flush(); i_see_dead_people: ip_ct_iterate_cleanup(kill_all, NULL); if (atomic_read(&ip_conntrack_count) != 0) { @@ -1115,7 +1363,14 @@ void ip_conntrack_cleanup(void) /* wait until all references to ip_conntrack_untracked are dropped */ while (atomic_read(&ip_conntrack_untracked.ct_general.use) > 1) schedule(); +} +/* Mishearing the voices in his head, our hero wonders how he's + supposed to kill the mall. */ +void ip_conntrack_cleanup(void) +{ + ip_ct_attach = NULL; + ip_conntrack_flush(); kmem_cache_destroy(ip_conntrack_cachep); kmem_cache_destroy(ip_conntrack_expect_cachep); free_conntrack_hash(); diff --git a/net/ipv4/netfilter/ip_conntrack_ftp.c b/net/ipv4/netfilter/ip_conntrack_ftp.c index 7a3b773be3f9..3a2627db1729 100644 --- a/net/ipv4/netfilter/ip_conntrack_ftp.c +++ b/net/ipv4/netfilter/ip_conntrack_ftp.c @@ -25,8 +25,7 @@ MODULE_AUTHOR("Rusty Russell <rusty@rustcorp.com.au>"); MODULE_DESCRIPTION("ftp connection tracking helper"); /* This is slow, but it's simple. --RR */ -static char ftp_buffer[65536]; - +static char *ftp_buffer; static DEFINE_SPINLOCK(ip_ftp_lock); #define MAX_PORTS 8 @@ -262,7 +261,8 @@ static int find_nl_seq(u32 seq, const struct ip_ct_ftp_master *info, int dir) } /* We don't update if it's older than what we have. */ -static void update_nl_seq(u32 nl_seq, struct ip_ct_ftp_master *info, int dir) +static void update_nl_seq(u32 nl_seq, struct ip_ct_ftp_master *info, int dir, + struct sk_buff *skb) { unsigned int i, oldest = NUM_SEQ_TO_REMEMBER; @@ -276,10 +276,13 @@ static void update_nl_seq(u32 nl_seq, struct ip_ct_ftp_master *info, int dir) oldest = i; } - if (info->seq_aft_nl_num[dir] < NUM_SEQ_TO_REMEMBER) + if (info->seq_aft_nl_num[dir] < NUM_SEQ_TO_REMEMBER) { info->seq_aft_nl[dir][info->seq_aft_nl_num[dir]++] = nl_seq; - else if (oldest != NUM_SEQ_TO_REMEMBER) + ip_conntrack_event_cache(IPCT_HELPINFO_VOLATILE, skb); + } else if (oldest != NUM_SEQ_TO_REMEMBER) { info->seq_aft_nl[dir][oldest] = nl_seq; + ip_conntrack_event_cache(IPCT_HELPINFO_VOLATILE, skb); + } } static int help(struct sk_buff **pskb, @@ -439,7 +442,7 @@ out_update_nl: /* Now if this ends in \n, update ftp info. Seq may have been * adjusted by NAT code. */ if (ends_in_nl) - update_nl_seq(seq, ct_ftp_info,dir); + update_nl_seq(seq, ct_ftp_info,dir, *pskb); out: spin_unlock_bh(&ip_ftp_lock); return ret; @@ -457,6 +460,8 @@ static void fini(void) ports[i]); ip_conntrack_helper_unregister(&ftp[i]); } + + kfree(ftp_buffer); } static int __init init(void) @@ -464,6 +469,10 @@ static int __init init(void) int i, ret; char *tmpname; + ftp_buffer = kmalloc(65536, GFP_KERNEL); + if (!ftp_buffer) + return -ENOMEM; + if (ports_c == 0) ports[ports_c++] = FTP_PORT; diff --git a/net/ipv4/netfilter/ip_conntrack_irc.c b/net/ipv4/netfilter/ip_conntrack_irc.c index 4a28f297d502..25438eec21a1 100644 --- a/net/ipv4/netfilter/ip_conntrack_irc.c +++ b/net/ipv4/netfilter/ip_conntrack_irc.c @@ -39,7 +39,7 @@ static int ports_c; static int max_dcc_channels = 8; static unsigned int dcc_timeout = 300; /* This is slow, but it's simple. --RR */ -static char irc_buffer[65536]; +static char *irc_buffer; static DEFINE_SPINLOCK(irc_buffer_lock); unsigned int (*ip_nat_irc_hook)(struct sk_buff **pskb, @@ -257,6 +257,10 @@ static int __init init(void) printk("ip_conntrack_irc: dcc_timeout must be a positive integer\n"); return -EBUSY; } + + irc_buffer = kmalloc(65536, GFP_KERNEL); + if (!irc_buffer) + return -ENOMEM; /* If no port given, default to standard irc port */ if (ports_c == 0) @@ -304,6 +308,7 @@ static void fini(void) ports[i]); ip_conntrack_helper_unregister(&irc_helpers[i]); } + kfree(irc_buffer); } module_init(init); diff --git a/net/ipv4/netfilter/ip_conntrack_netlink.c b/net/ipv4/netfilter/ip_conntrack_netlink.c new file mode 100644 index 000000000000..a4e9278db4ed --- /dev/null +++ b/net/ipv4/netfilter/ip_conntrack_netlink.c @@ -0,0 +1,1579 @@ +/* Connection tracking via netlink socket. Allows for user space + * protocol helpers and general trouble making from userspace. + * + * (C) 2001 by Jay Schulist <jschlst@samba.org> + * (C) 2002-2005 by Harald Welte <laforge@gnumonks.org> + * (C) 2003 by Patrick Mchardy <kaber@trash.net> + * (C) 2005 by Pablo Neira Ayuso <pablo@eurodev.net> + * + * I've reworked this stuff to use attributes instead of conntrack + * structures. 5.44 am. I need more tea. --pablo 05/07/11. + * + * Initial connection tracking via netlink development funded and + * generally made possible by Network Robots, Inc. (www.networkrobots.com) + * + * Further development of this code funded by Astaro AG (http://www.astaro.com) + * + * This software may be used and distributed according to the terms + * of the GNU General Public License, incorporated herein by reference. + */ + +#include <linux/init.h> +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/timer.h> +#include <linux/skbuff.h> +#include <linux/errno.h> +#include <linux/netlink.h> +#include <linux/spinlock.h> +#include <linux/notifier.h> +#include <linux/rtnetlink.h> + +#include <linux/netfilter.h> +#include <linux/netfilter_ipv4.h> +#include <linux/netfilter_ipv4/ip_tables.h> +#include <linux/netfilter_ipv4/ip_conntrack.h> +#include <linux/netfilter_ipv4/ip_conntrack_core.h> +#include <linux/netfilter_ipv4/ip_conntrack_helper.h> +#include <linux/netfilter_ipv4/ip_conntrack_protocol.h> +#include <linux/netfilter_ipv4/ip_nat_protocol.h> + +#include <linux/netfilter/nfnetlink.h> +#include <linux/netfilter/nfnetlink_conntrack.h> + +MODULE_LICENSE("GPL"); + +static char __initdata version[] = "0.90"; + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + + +static inline int +ctnetlink_dump_tuples_proto(struct sk_buff *skb, + const struct ip_conntrack_tuple *tuple) +{ + struct ip_conntrack_protocol *proto; + + NFA_PUT(skb, CTA_PROTO_NUM, sizeof(u_int8_t), &tuple->dst.protonum); + + proto = ip_conntrack_proto_find_get(tuple->dst.protonum); + if (proto && proto->tuple_to_nfattr) + return proto->tuple_to_nfattr(skb, tuple); + + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_dump_tuples(struct sk_buff *skb, + const struct ip_conntrack_tuple *tuple) +{ + struct nfattr *nest_parms; + + nest_parms = NFA_NEST(skb, CTA_TUPLE_IP); + NFA_PUT(skb, CTA_IP_V4_SRC, sizeof(u_int32_t), &tuple->src.ip); + NFA_PUT(skb, CTA_IP_V4_DST, sizeof(u_int32_t), &tuple->dst.ip); + NFA_NEST_END(skb, nest_parms); + + nest_parms = NFA_NEST(skb, CTA_TUPLE_PROTO); + ctnetlink_dump_tuples_proto(skb, tuple); + NFA_NEST_END(skb, nest_parms); + + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_dump_status(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + u_int32_t status = htonl((u_int32_t) ct->status); + NFA_PUT(skb, CTA_STATUS, sizeof(status), &status); + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_dump_timeout(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + long timeout_l = ct->timeout.expires - jiffies; + u_int32_t timeout; + + if (timeout_l < 0) + timeout = 0; + else + timeout = htonl(timeout_l / HZ); + + NFA_PUT(skb, CTA_TIMEOUT, sizeof(timeout), &timeout); + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_dump_protoinfo(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + struct ip_conntrack_protocol *proto = ip_conntrack_proto_find_get(ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum); + + struct nfattr *nest_proto; + int ret; + + if (!proto || !proto->to_nfattr) + return 0; + + nest_proto = NFA_NEST(skb, CTA_PROTOINFO); + + ret = proto->to_nfattr(skb, nest_proto, ct); + + ip_conntrack_proto_put(proto); + + NFA_NEST_END(skb, nest_proto); + + return ret; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_dump_helpinfo(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + struct nfattr *nest_helper; + + if (!ct->helper) + return 0; + + nest_helper = NFA_NEST(skb, CTA_HELP); + NFA_PUT(skb, CTA_HELP_NAME, CTA_HELP_MAXNAMESIZE, &ct->helper->name); + + if (ct->helper->to_nfattr) + ct->helper->to_nfattr(skb, ct); + + NFA_NEST_END(skb, nest_helper); + + return 0; + +nfattr_failure: + return -1; +} + +#ifdef CONFIG_IP_NF_CT_ACCT +static inline int +ctnetlink_dump_counters(struct sk_buff *skb, const struct ip_conntrack *ct, + enum ip_conntrack_dir dir) +{ + enum ctattr_type type = dir ? CTA_COUNTERS_REPLY: CTA_COUNTERS_ORIG; + struct nfattr *nest_count = NFA_NEST(skb, type); + u_int64_t tmp; + + tmp = cpu_to_be64(ct->counters[dir].packets); + NFA_PUT(skb, CTA_COUNTERS_PACKETS, sizeof(u_int64_t), &tmp); + + tmp = cpu_to_be64(ct->counters[dir].bytes); + NFA_PUT(skb, CTA_COUNTERS_BYTES, sizeof(u_int64_t), &tmp); + + NFA_NEST_END(skb, nest_count); + + return 0; + +nfattr_failure: + return -1; +} +#else +#define ctnetlink_dump_counters(a, b, c) (0) +#endif + +#ifdef CONFIG_IP_NF_CONNTRACK_MARK +static inline int +ctnetlink_dump_mark(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + u_int32_t mark = htonl(ct->mark); + + NFA_PUT(skb, CTA_MARK, sizeof(u_int32_t), &mark); + return 0; + +nfattr_failure: + return -1; +} +#else +#define ctnetlink_dump_mark(a, b) (0) +#endif + +static inline int +ctnetlink_dump_id(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + u_int32_t id = htonl(ct->id); + NFA_PUT(skb, CTA_ID, sizeof(u_int32_t), &id); + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_dump_use(struct sk_buff *skb, const struct ip_conntrack *ct) +{ + unsigned int use = htonl(atomic_read(&ct->ct_general.use)); + + NFA_PUT(skb, CTA_USE, sizeof(u_int32_t), &use); + return 0; + +nfattr_failure: + return -1; +} + +#define tuple(ct, dir) (&(ct)->tuplehash[dir].tuple) + +static int +ctnetlink_fill_info(struct sk_buff *skb, u32 pid, u32 seq, + int event, int nowait, + const struct ip_conntrack *ct) +{ + struct nlmsghdr *nlh; + struct nfgenmsg *nfmsg; + struct nfattr *nest_parms; + unsigned char *b; + + b = skb->tail; + + event |= NFNL_SUBSYS_CTNETLINK << 8; + nlh = NLMSG_PUT(skb, pid, seq, event, sizeof(struct nfgenmsg)); + nfmsg = NLMSG_DATA(nlh); + + nlh->nlmsg_flags = (nowait && pid) ? NLM_F_MULTI : 0; + nfmsg->nfgen_family = AF_INET; + nfmsg->version = NFNETLINK_V0; + nfmsg->res_id = 0; + + nest_parms = NFA_NEST(skb, CTA_TUPLE_ORIG); + if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_ORIGINAL)) < 0) + goto nfattr_failure; + NFA_NEST_END(skb, nest_parms); + + nest_parms = NFA_NEST(skb, CTA_TUPLE_REPLY); + if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_REPLY)) < 0) + goto nfattr_failure; + NFA_NEST_END(skb, nest_parms); + + if (ctnetlink_dump_status(skb, ct) < 0 || + ctnetlink_dump_timeout(skb, ct) < 0 || + ctnetlink_dump_counters(skb, ct, IP_CT_DIR_ORIGINAL) < 0 || + ctnetlink_dump_counters(skb, ct, IP_CT_DIR_REPLY) < 0 || + ctnetlink_dump_protoinfo(skb, ct) < 0 || + ctnetlink_dump_helpinfo(skb, ct) < 0 || + ctnetlink_dump_mark(skb, ct) < 0 || + ctnetlink_dump_id(skb, ct) < 0 || + ctnetlink_dump_use(skb, ct) < 0) + goto nfattr_failure; + + nlh->nlmsg_len = skb->tail - b; + return skb->len; + +nlmsg_failure: +nfattr_failure: + skb_trim(skb, b - skb->data); + return -1; +} + +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS +static int ctnetlink_conntrack_event(struct notifier_block *this, + unsigned long events, void *ptr) +{ + struct nlmsghdr *nlh; + struct nfgenmsg *nfmsg; + struct nfattr *nest_parms; + struct ip_conntrack *ct = (struct ip_conntrack *)ptr; + struct sk_buff *skb; + unsigned int type; + unsigned char *b; + unsigned int flags = 0, group; + + /* ignore our fake conntrack entry */ + if (ct == &ip_conntrack_untracked) + return NOTIFY_DONE; + + if (events & IPCT_DESTROY) { + type = IPCTNL_MSG_CT_DELETE; + group = NFNLGRP_CONNTRACK_DESTROY; + goto alloc_skb; + } + if (events & (IPCT_NEW | IPCT_RELATED)) { + type = IPCTNL_MSG_CT_NEW; + flags = NLM_F_CREATE|NLM_F_EXCL; + /* dump everything */ + events = ~0UL; + group = NFNLGRP_CONNTRACK_NEW; + goto alloc_skb; + } + if (events & (IPCT_STATUS | + IPCT_PROTOINFO | + IPCT_HELPER | + IPCT_HELPINFO | + IPCT_NATINFO)) { + type = IPCTNL_MSG_CT_NEW; + group = NFNLGRP_CONNTRACK_UPDATE; + goto alloc_skb; + } + + return NOTIFY_DONE; + +alloc_skb: + /* FIXME: Check if there are any listeners before, don't hurt performance */ + + skb = alloc_skb(NLMSG_GOODSIZE, GFP_ATOMIC); + if (!skb) + return NOTIFY_DONE; + + b = skb->tail; + + type |= NFNL_SUBSYS_CTNETLINK << 8; + nlh = NLMSG_PUT(skb, 0, 0, type, sizeof(struct nfgenmsg)); + nfmsg = NLMSG_DATA(nlh); + + nlh->nlmsg_flags = flags; + nfmsg->nfgen_family = AF_INET; + nfmsg->version = NFNETLINK_V0; + nfmsg->res_id = 0; + + nest_parms = NFA_NEST(skb, CTA_TUPLE_ORIG); + if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_ORIGINAL)) < 0) + goto nfattr_failure; + NFA_NEST_END(skb, nest_parms); + + nest_parms = NFA_NEST(skb, CTA_TUPLE_REPLY); + if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_REPLY)) < 0) + goto nfattr_failure; + NFA_NEST_END(skb, nest_parms); + + /* NAT stuff is now a status flag */ + if ((events & IPCT_STATUS || events & IPCT_NATINFO) + && ctnetlink_dump_status(skb, ct) < 0) + goto nfattr_failure; + if (events & IPCT_REFRESH + && ctnetlink_dump_timeout(skb, ct) < 0) + goto nfattr_failure; + if (events & IPCT_PROTOINFO + && ctnetlink_dump_protoinfo(skb, ct) < 0) + goto nfattr_failure; + if (events & IPCT_HELPINFO + && ctnetlink_dump_helpinfo(skb, ct) < 0) + goto nfattr_failure; + + if (ctnetlink_dump_counters(skb, ct, IP_CT_DIR_ORIGINAL) < 0 || + ctnetlink_dump_counters(skb, ct, IP_CT_DIR_REPLY) < 0) + goto nfattr_failure; + + nlh->nlmsg_len = skb->tail - b; + nfnetlink_send(skb, 0, group, 0); + return NOTIFY_DONE; + +nlmsg_failure: +nfattr_failure: + kfree_skb(skb); + return NOTIFY_DONE; +} +#endif /* CONFIG_IP_NF_CONNTRACK_EVENTS */ + +static int ctnetlink_done(struct netlink_callback *cb) +{ + DEBUGP("entered %s\n", __FUNCTION__); + return 0; +} + +static int +ctnetlink_dump_table(struct sk_buff *skb, struct netlink_callback *cb) +{ + struct ip_conntrack *ct = NULL; + struct ip_conntrack_tuple_hash *h; + struct list_head *i; + u_int32_t *id = (u_int32_t *) &cb->args[1]; + + DEBUGP("entered %s, last bucket=%lu id=%u\n", __FUNCTION__, + cb->args[0], *id); + + read_lock_bh(&ip_conntrack_lock); + for (; cb->args[0] < ip_conntrack_htable_size; cb->args[0]++, *id = 0) { + list_for_each_prev(i, &ip_conntrack_hash[cb->args[0]]) { + h = (struct ip_conntrack_tuple_hash *) i; + if (DIRECTION(h) != IP_CT_DIR_ORIGINAL) + continue; + ct = tuplehash_to_ctrack(h); + if (ct->id <= *id) + continue; + if (ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, + IPCTNL_MSG_CT_NEW, + 1, ct) < 0) + goto out; + *id = ct->id; + } + } +out: + read_unlock_bh(&ip_conntrack_lock); + + DEBUGP("leaving, last bucket=%lu id=%u\n", cb->args[0], *id); + + return skb->len; +} + +#ifdef CONFIG_IP_NF_CT_ACCT +static int +ctnetlink_dump_table_w(struct sk_buff *skb, struct netlink_callback *cb) +{ + struct ip_conntrack *ct = NULL; + struct ip_conntrack_tuple_hash *h; + struct list_head *i; + u_int32_t *id = (u_int32_t *) &cb->args[1]; + + DEBUGP("entered %s, last bucket=%u id=%u\n", __FUNCTION__, + cb->args[0], *id); + + write_lock_bh(&ip_conntrack_lock); + for (; cb->args[0] < ip_conntrack_htable_size; cb->args[0]++, *id = 0) { + list_for_each_prev(i, &ip_conntrack_hash[cb->args[0]]) { + h = (struct ip_conntrack_tuple_hash *) i; + if (DIRECTION(h) != IP_CT_DIR_ORIGINAL) + continue; + ct = tuplehash_to_ctrack(h); + if (ct->id <= *id) + continue; + if (ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, + IPCTNL_MSG_CT_NEW, + 1, ct) < 0) + goto out; + *id = ct->id; + + memset(&ct->counters, 0, sizeof(ct->counters)); + } + } +out: + write_unlock_bh(&ip_conntrack_lock); + + DEBUGP("leaving, last bucket=%lu id=%u\n", cb->args[0], *id); + + return skb->len; +} +#endif + +static const int cta_min_ip[CTA_IP_MAX] = { + [CTA_IP_V4_SRC-1] = sizeof(u_int32_t), + [CTA_IP_V4_DST-1] = sizeof(u_int32_t), +}; + +static inline int +ctnetlink_parse_tuple_ip(struct nfattr *attr, struct ip_conntrack_tuple *tuple) +{ + struct nfattr *tb[CTA_IP_MAX]; + + DEBUGP("entered %s\n", __FUNCTION__); + + + if (nfattr_parse_nested(tb, CTA_IP_MAX, attr) < 0) + goto nfattr_failure; + + if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip)) + return -EINVAL; + + if (!tb[CTA_IP_V4_SRC-1]) + return -EINVAL; + tuple->src.ip = *(u_int32_t *)NFA_DATA(tb[CTA_IP_V4_SRC-1]); + + if (!tb[CTA_IP_V4_DST-1]) + return -EINVAL; + tuple->dst.ip = *(u_int32_t *)NFA_DATA(tb[CTA_IP_V4_DST-1]); + + DEBUGP("leaving\n"); + + return 0; + +nfattr_failure: + return -1; +} + +static const int cta_min_proto[CTA_PROTO_MAX] = { + [CTA_PROTO_NUM-1] = sizeof(u_int16_t), + [CTA_PROTO_SRC_PORT-1] = sizeof(u_int16_t), + [CTA_PROTO_DST_PORT-1] = sizeof(u_int16_t), + [CTA_PROTO_ICMP_TYPE-1] = sizeof(u_int8_t), + [CTA_PROTO_ICMP_CODE-1] = sizeof(u_int8_t), + [CTA_PROTO_ICMP_ID-1] = sizeof(u_int16_t), +}; + +static inline int +ctnetlink_parse_tuple_proto(struct nfattr *attr, + struct ip_conntrack_tuple *tuple) +{ + struct nfattr *tb[CTA_PROTO_MAX]; + struct ip_conntrack_protocol *proto; + int ret = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (nfattr_parse_nested(tb, CTA_PROTO_MAX, attr) < 0) + goto nfattr_failure; + + if (nfattr_bad_size(tb, CTA_PROTO_MAX, cta_min_proto)) + return -EINVAL; + + if (!tb[CTA_PROTO_NUM-1]) + return -EINVAL; + tuple->dst.protonum = *(u_int16_t *)NFA_DATA(tb[CTA_PROTO_NUM-1]); + + proto = ip_conntrack_proto_find_get(tuple->dst.protonum); + + if (likely(proto && proto->nfattr_to_tuple)) { + ret = proto->nfattr_to_tuple(tb, tuple); + ip_conntrack_proto_put(proto); + } + + return ret; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_parse_tuple(struct nfattr *cda[], struct ip_conntrack_tuple *tuple, + enum ctattr_tuple type) +{ + struct nfattr *tb[CTA_TUPLE_MAX]; + int err; + + DEBUGP("entered %s\n", __FUNCTION__); + + memset(tuple, 0, sizeof(*tuple)); + + if (nfattr_parse_nested(tb, CTA_TUPLE_MAX, cda[type-1]) < 0) + goto nfattr_failure; + + if (!tb[CTA_TUPLE_IP-1]) + return -EINVAL; + + err = ctnetlink_parse_tuple_ip(tb[CTA_TUPLE_IP-1], tuple); + if (err < 0) + return err; + + if (!tb[CTA_TUPLE_PROTO-1]) + return -EINVAL; + + err = ctnetlink_parse_tuple_proto(tb[CTA_TUPLE_PROTO-1], tuple); + if (err < 0) + return err; + + /* orig and expect tuples get DIR_ORIGINAL */ + if (type == CTA_TUPLE_REPLY) + tuple->dst.dir = IP_CT_DIR_REPLY; + else + tuple->dst.dir = IP_CT_DIR_ORIGINAL; + + DUMP_TUPLE(tuple); + + DEBUGP("leaving\n"); + + return 0; + +nfattr_failure: + return -1; +} + +#ifdef CONFIG_IP_NF_NAT_NEEDED +static const int cta_min_protonat[CTA_PROTONAT_MAX] = { + [CTA_PROTONAT_PORT_MIN-1] = sizeof(u_int16_t), + [CTA_PROTONAT_PORT_MAX-1] = sizeof(u_int16_t), +}; + +static int ctnetlink_parse_nat_proto(struct nfattr *attr, + const struct ip_conntrack *ct, + struct ip_nat_range *range) +{ + struct nfattr *tb[CTA_PROTONAT_MAX]; + struct ip_nat_protocol *npt; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (nfattr_parse_nested(tb, CTA_PROTONAT_MAX, attr) < 0) + goto nfattr_failure; + + if (nfattr_bad_size(tb, CTA_PROTONAT_MAX, cta_min_protonat)) + goto nfattr_failure; + + npt = ip_nat_proto_find_get(ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum); + if (!npt) + return 0; + + if (!npt->nfattr_to_range) { + ip_nat_proto_put(npt); + return 0; + } + + /* nfattr_to_range returns 1 if it parsed, 0 if not, neg. on error */ + if (npt->nfattr_to_range(tb, range) > 0) + range->flags |= IP_NAT_RANGE_PROTO_SPECIFIED; + + ip_nat_proto_put(npt); + + DEBUGP("leaving\n"); + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_parse_nat(struct nfattr *cda[], + const struct ip_conntrack *ct, struct ip_nat_range *range) +{ + struct nfattr *tb[CTA_NAT_MAX]; + int err; + + DEBUGP("entered %s\n", __FUNCTION__); + + memset(range, 0, sizeof(*range)); + + if (nfattr_parse_nested(tb, CTA_NAT_MAX, cda[CTA_NAT-1]) < 0) + goto nfattr_failure; + + if (tb[CTA_NAT_MINIP-1]) + range->min_ip = *(u_int32_t *)NFA_DATA(tb[CTA_NAT_MINIP-1]); + + if (!tb[CTA_NAT_MAXIP-1]) + range->max_ip = range->min_ip; + else + range->max_ip = *(u_int32_t *)NFA_DATA(tb[CTA_NAT_MAXIP-1]); + + if (range->min_ip) + range->flags |= IP_NAT_RANGE_MAP_IPS; + + if (!tb[CTA_NAT_PROTO-1]) + return 0; + + err = ctnetlink_parse_nat_proto(tb[CTA_NAT_PROTO-1], ct, range); + if (err < 0) + return err; + + DEBUGP("leaving\n"); + return 0; + +nfattr_failure: + return -1; +} +#endif + +static inline int +ctnetlink_parse_help(struct nfattr *attr, char **helper_name) +{ + struct nfattr *tb[CTA_HELP_MAX]; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (nfattr_parse_nested(tb, CTA_HELP_MAX, attr) < 0) + goto nfattr_failure; + + if (!tb[CTA_HELP_NAME-1]) + return -EINVAL; + + *helper_name = NFA_DATA(tb[CTA_HELP_NAME-1]); + + return 0; + +nfattr_failure: + return -1; +} + +static int +ctnetlink_del_conntrack(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *cda[], int *errp) +{ + struct ip_conntrack_tuple_hash *h; + struct ip_conntrack_tuple tuple; + struct ip_conntrack *ct; + int err = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (cda[CTA_TUPLE_ORIG-1]) + err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_ORIG); + else if (cda[CTA_TUPLE_REPLY-1]) + err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_REPLY); + else { + /* Flush the whole table */ + ip_conntrack_flush(); + return 0; + } + + if (err < 0) + return err; + + h = ip_conntrack_find_get(&tuple, NULL); + if (!h) { + DEBUGP("tuple not found in conntrack hash\n"); + return -ENOENT; + } + + ct = tuplehash_to_ctrack(h); + + if (cda[CTA_ID-1]) { + u_int32_t id = ntohl(*(u_int32_t *)NFA_DATA(cda[CTA_ID-1])); + if (ct->id != id) { + ip_conntrack_put(ct); + return -ENOENT; + } + } + if (del_timer(&ct->timeout)) { + ip_conntrack_put(ct); + ct->timeout.function((unsigned long)ct); + return 0; + } + ip_conntrack_put(ct); + DEBUGP("leaving\n"); + + return 0; +} + +static int +ctnetlink_get_conntrack(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *cda[], int *errp) +{ + struct ip_conntrack_tuple_hash *h; + struct ip_conntrack_tuple tuple; + struct ip_conntrack *ct; + struct sk_buff *skb2 = NULL; + int err = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (nlh->nlmsg_flags & NLM_F_DUMP) { + struct nfgenmsg *msg = NLMSG_DATA(nlh); + u32 rlen; + + if (msg->nfgen_family != AF_INET) + return -EAFNOSUPPORT; + + if (NFNL_MSG_TYPE(nlh->nlmsg_type) == + IPCTNL_MSG_CT_GET_CTRZERO) { +#ifdef CONFIG_IP_NF_CT_ACCT + if ((*errp = netlink_dump_start(ctnl, skb, nlh, + ctnetlink_dump_table_w, + ctnetlink_done)) != 0) + return -EINVAL; +#else + return -ENOTSUPP; +#endif + } else { + if ((*errp = netlink_dump_start(ctnl, skb, nlh, + ctnetlink_dump_table, + ctnetlink_done)) != 0) + return -EINVAL; + } + + rlen = NLMSG_ALIGN(nlh->nlmsg_len); + if (rlen > skb->len) + rlen = skb->len; + skb_pull(skb, rlen); + return 0; + } + + if (cda[CTA_TUPLE_ORIG-1]) + err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_ORIG); + else if (cda[CTA_TUPLE_REPLY-1]) + err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_REPLY); + else + return -EINVAL; + + if (err < 0) + return err; + + h = ip_conntrack_find_get(&tuple, NULL); + if (!h) { + DEBUGP("tuple not found in conntrack hash"); + return -ENOENT; + } + DEBUGP("tuple found\n"); + ct = tuplehash_to_ctrack(h); + + err = -ENOMEM; + skb2 = alloc_skb(NLMSG_GOODSIZE, GFP_ATOMIC); + if (!skb2) { + ip_conntrack_put(ct); + return -ENOMEM; + } + NETLINK_CB(skb2).dst_pid = NETLINK_CB(skb).pid; + + err = ctnetlink_fill_info(skb2, NETLINK_CB(skb).pid, nlh->nlmsg_seq, + IPCTNL_MSG_CT_NEW, 1, ct); + ip_conntrack_put(ct); + if (err <= 0) + goto out; + + err = netlink_unicast(ctnl, skb2, NETLINK_CB(skb).pid, MSG_DONTWAIT); + if (err < 0) + goto out; + + DEBUGP("leaving\n"); + return 0; + +out: + if (skb2) + kfree_skb(skb2); + return -1; +} + +static inline int +ctnetlink_change_status(struct ip_conntrack *ct, struct nfattr *cda[]) +{ + unsigned long d, status = *(u_int32_t *)NFA_DATA(cda[CTA_STATUS-1]); + d = ct->status ^ status; + + if (d & (IPS_EXPECTED|IPS_CONFIRMED|IPS_DYING)) + /* unchangeable */ + return -EINVAL; + + if (d & IPS_SEEN_REPLY && !(status & IPS_SEEN_REPLY)) + /* SEEN_REPLY bit can only be set */ + return -EINVAL; + + + if (d & IPS_ASSURED && !(status & IPS_ASSURED)) + /* ASSURED bit can only be set */ + return -EINVAL; + + if (cda[CTA_NAT-1]) { +#ifndef CONFIG_IP_NF_NAT_NEEDED + return -EINVAL; +#else + unsigned int hooknum; + struct ip_nat_range range; + + if (ctnetlink_parse_nat(cda, ct, &range) < 0) + return -EINVAL; + + DEBUGP("NAT: %u.%u.%u.%u-%u.%u.%u.%u:%u-%u\n", + NIPQUAD(range.min_ip), NIPQUAD(range.max_ip), + htons(range.min.all), htons(range.max.all)); + + /* This is tricky but it works. ip_nat_setup_info needs the + * hook number as parameter, so let's do the correct + * conversion and run away */ + if (status & IPS_SRC_NAT_DONE) + hooknum = NF_IP_POST_ROUTING; /* IP_NAT_MANIP_SRC */ + else if (status & IPS_DST_NAT_DONE) + hooknum = NF_IP_PRE_ROUTING; /* IP_NAT_MANIP_DST */ + else + return -EINVAL; /* Missing NAT flags */ + + DEBUGP("NAT status: %lu\n", + status & (IPS_NAT_MASK | IPS_NAT_DONE_MASK)); + + if (ip_nat_initialized(ct, hooknum)) + return -EEXIST; + ip_nat_setup_info(ct, &range, hooknum); + + DEBUGP("NAT status after setup_info: %lu\n", + ct->status & (IPS_NAT_MASK | IPS_NAT_DONE_MASK)); +#endif + } + + /* Be careful here, modifying NAT bits can screw up things, + * so don't let users modify them directly if they don't pass + * ip_nat_range. */ + ct->status |= status & ~(IPS_NAT_DONE_MASK | IPS_NAT_MASK); + return 0; +} + + +static inline int +ctnetlink_change_helper(struct ip_conntrack *ct, struct nfattr *cda[]) +{ + struct ip_conntrack_helper *helper; + char *helpname; + int err; + + DEBUGP("entered %s\n", __FUNCTION__); + + /* don't change helper of sibling connections */ + if (ct->master) + return -EINVAL; + + err = ctnetlink_parse_help(cda[CTA_HELP-1], &helpname); + if (err < 0) + return err; + + helper = __ip_conntrack_helper_find_byname(helpname); + if (!helper) { + if (!strcmp(helpname, "")) + helper = NULL; + else + return -EINVAL; + } + + if (ct->helper) { + if (!helper) { + /* we had a helper before ... */ + ip_ct_remove_expectations(ct); + ct->helper = NULL; + } else { + /* need to zero data of old helper */ + memset(&ct->help, 0, sizeof(ct->help)); + } + } + + ct->helper = helper; + + return 0; +} + +static inline int +ctnetlink_change_timeout(struct ip_conntrack *ct, struct nfattr *cda[]) +{ + u_int32_t timeout = ntohl(*(u_int32_t *)NFA_DATA(cda[CTA_TIMEOUT-1])); + + if (!del_timer(&ct->timeout)) + return -ETIME; + + ct->timeout.expires = jiffies + timeout * HZ; + add_timer(&ct->timeout); + + return 0; +} + +static int +ctnetlink_change_conntrack(struct ip_conntrack *ct, struct nfattr *cda[]) +{ + int err; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (cda[CTA_HELP-1]) { + err = ctnetlink_change_helper(ct, cda); + if (err < 0) + return err; + } + + if (cda[CTA_TIMEOUT-1]) { + err = ctnetlink_change_timeout(ct, cda); + if (err < 0) + return err; + } + + if (cda[CTA_STATUS-1]) { + err = ctnetlink_change_status(ct, cda); + if (err < 0) + return err; + } + + DEBUGP("all done\n"); + return 0; +} + +static int +ctnetlink_create_conntrack(struct nfattr *cda[], + struct ip_conntrack_tuple *otuple, + struct ip_conntrack_tuple *rtuple) +{ + struct ip_conntrack *ct; + int err = -EINVAL; + + DEBUGP("entered %s\n", __FUNCTION__); + + ct = ip_conntrack_alloc(otuple, rtuple); + if (ct == NULL || IS_ERR(ct)) + return -ENOMEM; + + if (!cda[CTA_TIMEOUT-1]) + goto err; + ct->timeout.expires = ntohl(*(u_int32_t *)NFA_DATA(cda[CTA_TIMEOUT-1])); + + ct->timeout.expires = jiffies + ct->timeout.expires * HZ; + ct->status |= IPS_CONFIRMED; + + err = ctnetlink_change_status(ct, cda); + if (err < 0) + goto err; + + ct->helper = ip_conntrack_helper_find_get(rtuple); + + add_timer(&ct->timeout); + ip_conntrack_hash_insert(ct); + + if (ct->helper) + ip_conntrack_helper_put(ct->helper); + + DEBUGP("conntrack with id %u inserted\n", ct->id); + return 0; + +err: + ip_conntrack_free(ct); + return err; +} + +static int +ctnetlink_new_conntrack(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *cda[], int *errp) +{ + struct ip_conntrack_tuple otuple, rtuple; + struct ip_conntrack_tuple_hash *h = NULL; + int err = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (cda[CTA_TUPLE_ORIG-1]) { + err = ctnetlink_parse_tuple(cda, &otuple, CTA_TUPLE_ORIG); + if (err < 0) + return err; + } + + if (cda[CTA_TUPLE_REPLY-1]) { + err = ctnetlink_parse_tuple(cda, &rtuple, CTA_TUPLE_REPLY); + if (err < 0) + return err; + } + + write_lock_bh(&ip_conntrack_lock); + if (cda[CTA_TUPLE_ORIG-1]) + h = __ip_conntrack_find(&otuple, NULL); + else if (cda[CTA_TUPLE_REPLY-1]) + h = __ip_conntrack_find(&rtuple, NULL); + + if (h == NULL) { + write_unlock_bh(&ip_conntrack_lock); + DEBUGP("no such conntrack, create new\n"); + err = -ENOENT; + if (nlh->nlmsg_flags & NLM_F_CREATE) + err = ctnetlink_create_conntrack(cda, &otuple, &rtuple); + return err; + } + /* implicit 'else' */ + + /* we only allow nat config for new conntracks */ + if (cda[CTA_NAT-1]) { + err = -EINVAL; + goto out_unlock; + } + + /* We manipulate the conntrack inside the global conntrack table lock, + * so there's no need to increase the refcount */ + DEBUGP("conntrack found\n"); + err = -EEXIST; + if (!(nlh->nlmsg_flags & NLM_F_EXCL)) + err = ctnetlink_change_conntrack(tuplehash_to_ctrack(h), cda); + +out_unlock: + write_unlock_bh(&ip_conntrack_lock); + return err; +} + +/*********************************************************************** + * EXPECT + ***********************************************************************/ + +static inline int +ctnetlink_exp_dump_tuple(struct sk_buff *skb, + const struct ip_conntrack_tuple *tuple, + enum ctattr_expect type) +{ + struct nfattr *nest_parms = NFA_NEST(skb, type); + + if (ctnetlink_dump_tuples(skb, tuple) < 0) + goto nfattr_failure; + + NFA_NEST_END(skb, nest_parms); + + return 0; + +nfattr_failure: + return -1; +} + +static inline int +ctnetlink_exp_dump_expect(struct sk_buff *skb, + const struct ip_conntrack_expect *exp) +{ + struct ip_conntrack *master = exp->master; + u_int32_t timeout = htonl((exp->timeout.expires - jiffies) / HZ); + u_int32_t id = htonl(exp->id); + + if (ctnetlink_exp_dump_tuple(skb, &exp->tuple, CTA_EXPECT_TUPLE) < 0) + goto nfattr_failure; + if (ctnetlink_exp_dump_tuple(skb, &exp->mask, CTA_EXPECT_MASK) < 0) + goto nfattr_failure; + if (ctnetlink_exp_dump_tuple(skb, + &master->tuplehash[IP_CT_DIR_ORIGINAL].tuple, + CTA_EXPECT_MASTER) < 0) + goto nfattr_failure; + + NFA_PUT(skb, CTA_EXPECT_TIMEOUT, sizeof(timeout), &timeout); + NFA_PUT(skb, CTA_EXPECT_ID, sizeof(u_int32_t), &id); + + return 0; + +nfattr_failure: + return -1; +} + +static int +ctnetlink_exp_fill_info(struct sk_buff *skb, u32 pid, u32 seq, + int event, + int nowait, + const struct ip_conntrack_expect *exp) +{ + struct nlmsghdr *nlh; + struct nfgenmsg *nfmsg; + unsigned char *b; + + b = skb->tail; + + event |= NFNL_SUBSYS_CTNETLINK_EXP << 8; + nlh = NLMSG_PUT(skb, pid, seq, event, sizeof(struct nfgenmsg)); + nfmsg = NLMSG_DATA(nlh); + + nlh->nlmsg_flags = (nowait && pid) ? NLM_F_MULTI : 0; + nfmsg->nfgen_family = AF_INET; + nfmsg->version = NFNETLINK_V0; + nfmsg->res_id = 0; + + if (ctnetlink_exp_dump_expect(skb, exp) < 0) + goto nfattr_failure; + + nlh->nlmsg_len = skb->tail - b; + return skb->len; + +nlmsg_failure: +nfattr_failure: + skb_trim(skb, b - skb->data); + return -1; +} + +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS +static int ctnetlink_expect_event(struct notifier_block *this, + unsigned long events, void *ptr) +{ + struct nlmsghdr *nlh; + struct nfgenmsg *nfmsg; + struct ip_conntrack_expect *exp = (struct ip_conntrack_expect *)ptr; + struct sk_buff *skb; + unsigned int type; + unsigned char *b; + int flags = 0; + u16 proto; + + if (events & IPEXP_NEW) { + type = IPCTNL_MSG_EXP_NEW; + flags = NLM_F_CREATE|NLM_F_EXCL; + } else + return NOTIFY_DONE; + + skb = alloc_skb(NLMSG_GOODSIZE, GFP_ATOMIC); + if (!skb) + return NOTIFY_DONE; + + b = skb->tail; + + type |= NFNL_SUBSYS_CTNETLINK << 8; + nlh = NLMSG_PUT(skb, 0, 0, type, sizeof(struct nfgenmsg)); + nfmsg = NLMSG_DATA(nlh); + + nlh->nlmsg_flags = flags; + nfmsg->nfgen_family = AF_INET; + nfmsg->version = NFNETLINK_V0; + nfmsg->res_id = 0; + + if (ctnetlink_exp_dump_expect(skb, exp) < 0) + goto nfattr_failure; + + nlh->nlmsg_len = skb->tail - b; + proto = exp->tuple.dst.protonum; + nfnetlink_send(skb, 0, NFNLGRP_CONNTRACK_EXP_NEW, 0); + return NOTIFY_DONE; + +nlmsg_failure: +nfattr_failure: + kfree_skb(skb); + return NOTIFY_DONE; +} +#endif + +static int +ctnetlink_exp_dump_table(struct sk_buff *skb, struct netlink_callback *cb) +{ + struct ip_conntrack_expect *exp = NULL; + struct list_head *i; + u_int32_t *id = (u_int32_t *) &cb->args[0]; + + DEBUGP("entered %s, last id=%llu\n", __FUNCTION__, *id); + + read_lock_bh(&ip_conntrack_lock); + list_for_each_prev(i, &ip_conntrack_expect_list) { + exp = (struct ip_conntrack_expect *) i; + if (exp->id <= *id) + continue; + if (ctnetlink_exp_fill_info(skb, NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, + IPCTNL_MSG_EXP_NEW, + 1, exp) < 0) + goto out; + *id = exp->id; + } +out: + read_unlock_bh(&ip_conntrack_lock); + + DEBUGP("leaving, last id=%llu\n", *id); + + return skb->len; +} + +static int +ctnetlink_get_expect(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *cda[], int *errp) +{ + struct ip_conntrack_tuple tuple; + struct ip_conntrack_expect *exp; + struct sk_buff *skb2; + int err = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (nlh->nlmsg_flags & NLM_F_DUMP) { + struct nfgenmsg *msg = NLMSG_DATA(nlh); + u32 rlen; + + if (msg->nfgen_family != AF_INET) + return -EAFNOSUPPORT; + + if ((*errp = netlink_dump_start(ctnl, skb, nlh, + ctnetlink_exp_dump_table, + ctnetlink_done)) != 0) + return -EINVAL; + rlen = NLMSG_ALIGN(nlh->nlmsg_len); + if (rlen > skb->len) + rlen = skb->len; + skb_pull(skb, rlen); + return 0; + } + + if (cda[CTA_EXPECT_MASTER-1]) + err = ctnetlink_parse_tuple(cda, &tuple, CTA_EXPECT_MASTER); + else + return -EINVAL; + + if (err < 0) + return err; + + exp = ip_conntrack_expect_find_get(&tuple); + if (!exp) + return -ENOENT; + + err = -ENOMEM; + skb2 = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL); + if (!skb2) + goto out; + NETLINK_CB(skb2).dst_pid = NETLINK_CB(skb).pid; + + err = ctnetlink_exp_fill_info(skb2, NETLINK_CB(skb).pid, + nlh->nlmsg_seq, IPCTNL_MSG_EXP_NEW, + 1, exp); + if (err <= 0) + goto out; + + ip_conntrack_expect_put(exp); + + err = netlink_unicast(ctnl, skb2, NETLINK_CB(skb).pid, MSG_DONTWAIT); + if (err < 0) + goto free; + + return err; + +out: + ip_conntrack_expect_put(exp); +free: + if (skb2) + kfree_skb(skb2); + return err; +} + +static int +ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *cda[], int *errp) +{ + struct ip_conntrack_expect *exp, *tmp; + struct ip_conntrack_tuple tuple; + struct ip_conntrack_helper *h; + int err; + + if (cda[CTA_EXPECT_TUPLE-1]) { + /* delete a single expect by tuple */ + err = ctnetlink_parse_tuple(cda, &tuple, CTA_EXPECT_TUPLE); + if (err < 0) + return err; + + /* bump usage count to 2 */ + exp = ip_conntrack_expect_find_get(&tuple); + if (!exp) + return -ENOENT; + + if (cda[CTA_EXPECT_ID-1]) { + u_int32_t id = + *(u_int32_t *)NFA_DATA(cda[CTA_EXPECT_ID-1]); + if (exp->id != ntohl(id)) { + ip_conntrack_expect_put(exp); + return -ENOENT; + } + } + + /* after list removal, usage count == 1 */ + ip_conntrack_unexpect_related(exp); + /* have to put what we 'get' above. + * after this line usage count == 0 */ + ip_conntrack_expect_put(exp); + } else if (cda[CTA_EXPECT_HELP_NAME-1]) { + char *name = NFA_DATA(cda[CTA_EXPECT_HELP_NAME-1]); + + /* delete all expectations for this helper */ + write_lock_bh(&ip_conntrack_lock); + h = __ip_conntrack_helper_find_byname(name); + if (!h) { + write_unlock_bh(&ip_conntrack_lock); + return -EINVAL; + } + list_for_each_entry_safe(exp, tmp, &ip_conntrack_expect_list, + list) { + if (exp->master->helper == h + && del_timer(&exp->timeout)) + __ip_ct_expect_unlink_destroy(exp); + } + write_unlock(&ip_conntrack_lock); + } else { + /* This basically means we have to flush everything*/ + write_lock_bh(&ip_conntrack_lock); + list_for_each_entry_safe(exp, tmp, &ip_conntrack_expect_list, + list) { + if (del_timer(&exp->timeout)) + __ip_ct_expect_unlink_destroy(exp); + } + write_unlock_bh(&ip_conntrack_lock); + } + + return 0; +} +static int +ctnetlink_change_expect(struct ip_conntrack_expect *x, struct nfattr *cda[]) +{ + return -EOPNOTSUPP; +} + +static int +ctnetlink_create_expect(struct nfattr *cda[]) +{ + struct ip_conntrack_tuple tuple, mask, master_tuple; + struct ip_conntrack_tuple_hash *h = NULL; + struct ip_conntrack_expect *exp; + struct ip_conntrack *ct; + int err = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + /* caller guarantees that those three CTA_EXPECT_* exist */ + err = ctnetlink_parse_tuple(cda, &tuple, CTA_EXPECT_TUPLE); + if (err < 0) + return err; + err = ctnetlink_parse_tuple(cda, &mask, CTA_EXPECT_MASK); + if (err < 0) + return err; + err = ctnetlink_parse_tuple(cda, &master_tuple, CTA_EXPECT_MASTER); + if (err < 0) + return err; + + /* Look for master conntrack of this expectation */ + h = ip_conntrack_find_get(&master_tuple, NULL); + if (!h) + return -ENOENT; + ct = tuplehash_to_ctrack(h); + + if (!ct->helper) { + /* such conntrack hasn't got any helper, abort */ + err = -EINVAL; + goto out; + } + + exp = ip_conntrack_expect_alloc(ct); + if (!exp) { + err = -ENOMEM; + goto out; + } + + exp->expectfn = NULL; + exp->master = ct; + memcpy(&exp->tuple, &tuple, sizeof(struct ip_conntrack_tuple)); + memcpy(&exp->mask, &mask, sizeof(struct ip_conntrack_tuple)); + + err = ip_conntrack_expect_related(exp); + ip_conntrack_expect_put(exp); + +out: + ip_conntrack_put(tuplehash_to_ctrack(h)); + return err; +} + +static int +ctnetlink_new_expect(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *cda[], int *errp) +{ + struct ip_conntrack_tuple tuple; + struct ip_conntrack_expect *exp; + int err = 0; + + DEBUGP("entered %s\n", __FUNCTION__); + + if (!cda[CTA_EXPECT_TUPLE-1] + || !cda[CTA_EXPECT_MASK-1] + || !cda[CTA_EXPECT_MASTER-1]) + return -EINVAL; + + err = ctnetlink_parse_tuple(cda, &tuple, CTA_EXPECT_TUPLE); + if (err < 0) + return err; + + write_lock_bh(&ip_conntrack_lock); + exp = __ip_conntrack_expect_find(&tuple); + + if (!exp) { + write_unlock_bh(&ip_conntrack_lock); + err = -ENOENT; + if (nlh->nlmsg_flags & NLM_F_CREATE) + err = ctnetlink_create_expect(cda); + return err; + } + + err = -EEXIST; + if (!(nlh->nlmsg_flags & NLM_F_EXCL)) + err = ctnetlink_change_expect(exp, cda); + write_unlock_bh(&ip_conntrack_lock); + + DEBUGP("leaving\n"); + + return err; +} + +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS +static struct notifier_block ctnl_notifier = { + .notifier_call = ctnetlink_conntrack_event, +}; + +static struct notifier_block ctnl_notifier_exp = { + .notifier_call = ctnetlink_expect_event, +}; +#endif + +static struct nfnl_callback ctnl_cb[IPCTNL_MSG_MAX] = { + [IPCTNL_MSG_CT_NEW] = { .call = ctnetlink_new_conntrack, + .attr_count = CTA_MAX, + .cap_required = CAP_NET_ADMIN }, + [IPCTNL_MSG_CT_GET] = { .call = ctnetlink_get_conntrack, + .attr_count = CTA_MAX, + .cap_required = CAP_NET_ADMIN }, + [IPCTNL_MSG_CT_DELETE] = { .call = ctnetlink_del_conntrack, + .attr_count = CTA_MAX, + .cap_required = CAP_NET_ADMIN }, + [IPCTNL_MSG_CT_GET_CTRZERO] = { .call = ctnetlink_get_conntrack, + .attr_count = CTA_MAX, + .cap_required = CAP_NET_ADMIN }, +}; + +static struct nfnl_callback ctnl_exp_cb[IPCTNL_MSG_EXP_MAX] = { + [IPCTNL_MSG_EXP_GET] = { .call = ctnetlink_get_expect, + .attr_count = CTA_EXPECT_MAX, + .cap_required = CAP_NET_ADMIN }, + [IPCTNL_MSG_EXP_NEW] = { .call = ctnetlink_new_expect, + .attr_count = CTA_EXPECT_MAX, + .cap_required = CAP_NET_ADMIN }, + [IPCTNL_MSG_EXP_DELETE] = { .call = ctnetlink_del_expect, + .attr_count = CTA_EXPECT_MAX, + .cap_required = CAP_NET_ADMIN }, +}; + +static struct nfnetlink_subsystem ctnl_subsys = { + .name = "conntrack", + .subsys_id = NFNL_SUBSYS_CTNETLINK, + .cb_count = IPCTNL_MSG_MAX, + .cb = ctnl_cb, +}; + +static struct nfnetlink_subsystem ctnl_exp_subsys = { + .name = "conntrack_expect", + .subsys_id = NFNL_SUBSYS_CTNETLINK_EXP, + .cb_count = IPCTNL_MSG_EXP_MAX, + .cb = ctnl_exp_cb, +}; + +static int __init ctnetlink_init(void) +{ + int ret; + + printk("ctnetlink v%s: registering with nfnetlink.\n", version); + ret = nfnetlink_subsys_register(&ctnl_subsys); + if (ret < 0) { + printk("ctnetlink_init: cannot register with nfnetlink.\n"); + goto err_out; + } + + ret = nfnetlink_subsys_register(&ctnl_exp_subsys); + if (ret < 0) { + printk("ctnetlink_init: cannot register exp with nfnetlink.\n"); + goto err_unreg_subsys; + } + +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS + ret = ip_conntrack_register_notifier(&ctnl_notifier); + if (ret < 0) { + printk("ctnetlink_init: cannot register notifier.\n"); + goto err_unreg_exp_subsys; + } + + ret = ip_conntrack_expect_register_notifier(&ctnl_notifier_exp); + if (ret < 0) { + printk("ctnetlink_init: cannot expect register notifier.\n"); + goto err_unreg_notifier; + } +#endif + + return 0; + +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS +err_unreg_notifier: + ip_conntrack_unregister_notifier(&ctnl_notifier); +err_unreg_exp_subsys: + nfnetlink_subsys_unregister(&ctnl_exp_subsys); +#endif +err_unreg_subsys: + nfnetlink_subsys_unregister(&ctnl_subsys); +err_out: + return ret; +} + +static void __exit ctnetlink_exit(void) +{ + printk("ctnetlink: unregistering from nfnetlink.\n"); + +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS + ip_conntrack_unregister_notifier(&ctnl_notifier_exp); + ip_conntrack_unregister_notifier(&ctnl_notifier); +#endif + + nfnetlink_subsys_unregister(&ctnl_exp_subsys); + nfnetlink_subsys_unregister(&ctnl_subsys); + return; +} + +module_init(ctnetlink_init); +module_exit(ctnetlink_exit); diff --git a/net/ipv4/netfilter/ip_conntrack_proto_icmp.c b/net/ipv4/netfilter/ip_conntrack_proto_icmp.c index 602c74db3252..838d1d69b36e 100644 --- a/net/ipv4/netfilter/ip_conntrack_proto_icmp.c +++ b/net/ipv4/netfilter/ip_conntrack_proto_icmp.c @@ -102,22 +102,24 @@ static int icmp_packet(struct ip_conntrack *ct, ct->timeout.function((unsigned long)ct); } else { atomic_inc(&ct->proto.icmp.count); + ip_conntrack_event_cache(IPCT_PROTOINFO_VOLATILE, skb); ip_ct_refresh_acct(ct, ctinfo, skb, ip_ct_icmp_timeout); } return NF_ACCEPT; } +static u_int8_t valid_new[] = { + [ICMP_ECHO] = 1, + [ICMP_TIMESTAMP] = 1, + [ICMP_INFO_REQUEST] = 1, + [ICMP_ADDRESS] = 1 +}; + /* Called when a new connection for this protocol found. */ static int icmp_new(struct ip_conntrack *conntrack, const struct sk_buff *skb) { - static u_int8_t valid_new[] - = { [ICMP_ECHO] = 1, - [ICMP_TIMESTAMP] = 1, - [ICMP_INFO_REQUEST] = 1, - [ICMP_ADDRESS] = 1 }; - if (conntrack->tuplehash[0].tuple.dst.u.icmp.type >= sizeof(valid_new) || !valid_new[conntrack->tuplehash[0].tuple.dst.u.icmp.type]) { /* Can't create a new ICMP `conn' with this. */ @@ -158,11 +160,12 @@ icmp_error_message(struct sk_buff *skb, return NF_ACCEPT; } - innerproto = ip_ct_find_proto(inside->ip.protocol); + innerproto = ip_conntrack_proto_find_get(inside->ip.protocol); dataoff = skb->nh.iph->ihl*4 + sizeof(inside->icmp) + inside->ip.ihl*4; /* Are they talking about one of our connections? */ if (!ip_ct_get_tuple(&inside->ip, skb, dataoff, &origtuple, innerproto)) { DEBUGP("icmp_error: ! get_tuple p=%u", inside->ip.protocol); + ip_conntrack_proto_put(innerproto); return NF_ACCEPT; } @@ -170,8 +173,10 @@ icmp_error_message(struct sk_buff *skb, been preserved inside the ICMP. */ if (!ip_ct_invert_tuple(&innertuple, &origtuple, innerproto)) { DEBUGP("icmp_error_track: Can't invert tuple\n"); + ip_conntrack_proto_put(innerproto); return NF_ACCEPT; } + ip_conntrack_proto_put(innerproto); *ctinfo = IP_CT_RELATED; @@ -212,7 +217,7 @@ icmp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo, icmph = skb_header_pointer(skb, skb->nh.iph->ihl*4, sizeof(_ih), &_ih); if (icmph == NULL) { if (LOG_INVALID(IPPROTO_ICMP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_icmp: short packet "); return -NF_ACCEPT; } @@ -226,13 +231,13 @@ icmp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo, if (!(u16)csum_fold(skb->csum)) break; if (LOG_INVALID(IPPROTO_ICMP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_icmp: bad HW ICMP checksum "); return -NF_ACCEPT; case CHECKSUM_NONE: if ((u16)csum_fold(skb_checksum(skb, 0, skb->len, 0))) { if (LOG_INVALID(IPPROTO_ICMP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_icmp: bad ICMP checksum "); return -NF_ACCEPT; } @@ -249,7 +254,7 @@ checksum_skipped: */ if (icmph->type > NR_ICMP_TYPES) { if (LOG_INVALID(IPPROTO_ICMP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_icmp: invalid ICMP type "); return -NF_ACCEPT; } @@ -265,6 +270,47 @@ checksum_skipped: return icmp_error_message(skb, ctinfo, hooknum); } +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) +static int icmp_tuple_to_nfattr(struct sk_buff *skb, + const struct ip_conntrack_tuple *t) +{ + NFA_PUT(skb, CTA_PROTO_ICMP_ID, sizeof(u_int16_t), + &t->src.u.icmp.id); + NFA_PUT(skb, CTA_PROTO_ICMP_TYPE, sizeof(u_int8_t), + &t->dst.u.icmp.type); + NFA_PUT(skb, CTA_PROTO_ICMP_CODE, sizeof(u_int8_t), + &t->dst.u.icmp.code); + + if (t->dst.u.icmp.type >= sizeof(valid_new) + || !valid_new[t->dst.u.icmp.type]) + return -EINVAL; + + return 0; + +nfattr_failure: + return -1; +} + +static int icmp_nfattr_to_tuple(struct nfattr *tb[], + struct ip_conntrack_tuple *tuple) +{ + if (!tb[CTA_PROTO_ICMP_TYPE-1] + || !tb[CTA_PROTO_ICMP_CODE-1] + || !tb[CTA_PROTO_ICMP_ID-1]) + return -1; + + tuple->dst.u.icmp.type = + *(u_int8_t *)NFA_DATA(tb[CTA_PROTO_ICMP_TYPE-1]); + tuple->dst.u.icmp.code = + *(u_int8_t *)NFA_DATA(tb[CTA_PROTO_ICMP_CODE-1]); + tuple->src.u.icmp.id = + *(u_int8_t *)NFA_DATA(tb[CTA_PROTO_ICMP_ID-1]); + + return 0; +} +#endif + struct ip_conntrack_protocol ip_conntrack_protocol_icmp = { .proto = IPPROTO_ICMP, @@ -276,4 +322,9 @@ struct ip_conntrack_protocol ip_conntrack_protocol_icmp = .packet = icmp_packet, .new = icmp_new, .error = icmp_error, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .tuple_to_nfattr = icmp_tuple_to_nfattr, + .nfattr_to_tuple = icmp_nfattr_to_tuple, +#endif }; diff --git a/net/ipv4/netfilter/ip_conntrack_proto_sctp.c b/net/ipv4/netfilter/ip_conntrack_proto_sctp.c index 31d75390bf12..a875f35e576d 100644 --- a/net/ipv4/netfilter/ip_conntrack_proto_sctp.c +++ b/net/ipv4/netfilter/ip_conntrack_proto_sctp.c @@ -404,6 +404,8 @@ static int sctp_packet(struct ip_conntrack *conntrack, } conntrack->proto.sctp.state = newconntrack; + if (oldsctpstate != newconntrack) + ip_conntrack_event_cache(IPCT_PROTOINFO, skb); write_unlock_bh(&sctp_lock); } @@ -503,7 +505,12 @@ static struct ip_conntrack_protocol ip_conntrack_protocol_sctp = { .packet = sctp_packet, .new = sctp_new, .destroy = NULL, - .me = THIS_MODULE + .me = THIS_MODULE, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .tuple_to_nfattr = ip_ct_port_tuple_to_nfattr, + .nfattr_to_tuple = ip_ct_port_nfattr_to_tuple, +#endif }; #ifdef CONFIG_SYSCTL diff --git a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c index 809dfed766d4..f23ef1f88c46 100644 --- a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c +++ b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c @@ -336,6 +336,23 @@ static int tcp_print_conntrack(struct seq_file *s, return seq_printf(s, "%s ", tcp_conntrack_names[state]); } +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) +static int tcp_to_nfattr(struct sk_buff *skb, struct nfattr *nfa, + const struct ip_conntrack *ct) +{ + read_lock_bh(&tcp_lock); + NFA_PUT(skb, CTA_PROTOINFO_TCP_STATE, sizeof(u_int8_t), + &ct->proto.tcp.state); + read_unlock_bh(&tcp_lock); + + return 0; + +nfattr_failure: + return -1; +} +#endif + static unsigned int get_conntrack_index(const struct tcphdr *tcph) { if (tcph->rst) return TCP_RST_SET; @@ -699,7 +716,7 @@ static int tcp_in_window(struct ip_ct_tcp *state, res = 1; } else { if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: %s ", before(seq, sender->td_maxend + 1) ? after(end, sender->td_end - receiver->td_maxwin - 1) ? @@ -798,7 +815,7 @@ static int tcp_error(struct sk_buff *skb, sizeof(_tcph), &_tcph); if (th == NULL) { if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: short packet "); return -NF_ACCEPT; } @@ -806,7 +823,7 @@ static int tcp_error(struct sk_buff *skb, /* Not whole TCP header or malformed packet */ if (th->doff*4 < sizeof(struct tcphdr) || tcplen < th->doff*4) { if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: truncated/malformed packet "); return -NF_ACCEPT; } @@ -823,7 +840,7 @@ static int tcp_error(struct sk_buff *skb, skb->ip_summed == CHECKSUM_HW ? skb->csum : skb_checksum(skb, iph->ihl*4, tcplen, 0))) { if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: bad TCP checksum "); return -NF_ACCEPT; } @@ -832,7 +849,7 @@ static int tcp_error(struct sk_buff *skb, tcpflags = (((u_int8_t *)th)[13] & ~(TH_ECE|TH_CWR)); if (!tcp_valid_flags[tcpflags]) { if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: invalid TCP flag combination "); return -NF_ACCEPT; } @@ -880,8 +897,9 @@ static int tcp_packet(struct ip_conntrack *conntrack, */ write_unlock_bh(&tcp_lock); if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, - "ip_ct_tcp: killing out of sync session "); + nf_log_packet(PF_INET, 0, skb, NULL, NULL, + NULL, "ip_ct_tcp: " + "killing out of sync session "); if (del_timer(&conntrack->timeout)) conntrack->timeout.function((unsigned long) conntrack); @@ -895,7 +913,7 @@ static int tcp_packet(struct ip_conntrack *conntrack, write_unlock_bh(&tcp_lock); if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: invalid packet ignored "); return NF_ACCEPT; case TCP_CONNTRACK_MAX: @@ -905,7 +923,7 @@ static int tcp_packet(struct ip_conntrack *conntrack, old_state); write_unlock_bh(&tcp_lock); if (LOG_INVALID(IPPROTO_TCP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_tcp: invalid state "); return -NF_ACCEPT; case TCP_CONNTRACK_SYN_SENT: @@ -926,7 +944,7 @@ static int tcp_packet(struct ip_conntrack *conntrack, write_unlock_bh(&tcp_lock); if (LOG_INVALID(IPPROTO_TCP)) nf_log_packet(PF_INET, 0, skb, NULL, NULL, - "ip_ct_tcp: invalid SYN"); + NULL, "ip_ct_tcp: invalid SYN"); return -NF_ACCEPT; } case TCP_CONNTRACK_CLOSE: @@ -973,6 +991,10 @@ static int tcp_packet(struct ip_conntrack *conntrack, ? ip_ct_tcp_timeout_max_retrans : *tcp_timeouts[new_state]; write_unlock_bh(&tcp_lock); + ip_conntrack_event_cache(IPCT_PROTOINFO_VOLATILE, skb); + if (new_state != old_state) + ip_conntrack_event_cache(IPCT_PROTOINFO, skb); + if (!test_bit(IPS_SEEN_REPLY_BIT, &conntrack->status)) { /* If only reply is a RST, we can consider ourselves not to have an established connection: this is a fairly common @@ -1096,4 +1118,10 @@ struct ip_conntrack_protocol ip_conntrack_protocol_tcp = .packet = tcp_packet, .new = tcp_new, .error = tcp_error, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .to_nfattr = tcp_to_nfattr, + .tuple_to_nfattr = ip_ct_port_tuple_to_nfattr, + .nfattr_to_tuple = ip_ct_port_nfattr_to_tuple, +#endif }; diff --git a/net/ipv4/netfilter/ip_conntrack_proto_udp.c b/net/ipv4/netfilter/ip_conntrack_proto_udp.c index 8c1eaba098d4..f2dcac7c7660 100644 --- a/net/ipv4/netfilter/ip_conntrack_proto_udp.c +++ b/net/ipv4/netfilter/ip_conntrack_proto_udp.c @@ -73,7 +73,8 @@ static int udp_packet(struct ip_conntrack *conntrack, ip_ct_refresh_acct(conntrack, ctinfo, skb, ip_ct_udp_timeout_stream); /* Also, more likely to be important, and not a probe */ - set_bit(IPS_ASSURED_BIT, &conntrack->status); + if (!test_and_set_bit(IPS_ASSURED_BIT, &conntrack->status)) + ip_conntrack_event_cache(IPCT_STATUS, skb); } else ip_ct_refresh_acct(conntrack, ctinfo, skb, ip_ct_udp_timeout); @@ -97,7 +98,7 @@ static int udp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo, hdr = skb_header_pointer(skb, iph->ihl*4, sizeof(_hdr), &_hdr); if (hdr == NULL) { if (LOG_INVALID(IPPROTO_UDP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_udp: short packet "); return -NF_ACCEPT; } @@ -105,7 +106,7 @@ static int udp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo, /* Truncated/malformed packets */ if (ntohs(hdr->len) > udplen || ntohs(hdr->len) < sizeof(*hdr)) { if (LOG_INVALID(IPPROTO_UDP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_udp: truncated/malformed packet "); return -NF_ACCEPT; } @@ -125,7 +126,7 @@ static int udp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo, skb->ip_summed == CHECKSUM_HW ? skb->csum : skb_checksum(skb, iph->ihl*4, udplen, 0))) { if (LOG_INVALID(IPPROTO_UDP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_udp: bad UDP checksum "); return -NF_ACCEPT; } @@ -144,4 +145,9 @@ struct ip_conntrack_protocol ip_conntrack_protocol_udp = .packet = udp_packet, .new = udp_new, .error = udp_error, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .tuple_to_nfattr = ip_ct_port_tuple_to_nfattr, + .nfattr_to_tuple = ip_ct_port_nfattr_to_tuple, +#endif }; diff --git a/net/ipv4/netfilter/ip_conntrack_standalone.c b/net/ipv4/netfilter/ip_conntrack_standalone.c index 61798c46e91d..ee5895afd0c3 100644 --- a/net/ipv4/netfilter/ip_conntrack_standalone.c +++ b/net/ipv4/netfilter/ip_conntrack_standalone.c @@ -5,7 +5,7 @@ */ /* (C) 1999-2001 Paul `Rusty' Russell - * (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org> + * (C) 2002-2005 Netfilter Core Team <coreteam@netfilter.org> * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -147,8 +147,7 @@ static int ct_seq_show(struct seq_file *s, void *v) if (DIRECTION(hash)) return 0; - proto = ip_ct_find_proto(conntrack->tuplehash[IP_CT_DIR_ORIGINAL] - .tuple.dst.protonum); + proto = __ip_conntrack_proto_find(conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum); IP_NF_ASSERT(proto); if (seq_printf(s, "%-8s %u %ld ", @@ -185,7 +184,7 @@ static int ct_seq_show(struct seq_file *s, void *v) return -ENOSPC; #if defined(CONFIG_IP_NF_CONNTRACK_MARK) - if (seq_printf(s, "mark=%lu ", conntrack->mark)) + if (seq_printf(s, "mark=%u ", conntrack->mark)) return -ENOSPC; #endif @@ -283,7 +282,7 @@ static int exp_seq_show(struct seq_file *s, void *v) seq_printf(s, "proto=%u ", expect->tuple.dst.protonum); print_tuple(s, &expect->tuple, - ip_ct_find_proto(expect->tuple.dst.protonum)); + __ip_conntrack_proto_find(expect->tuple.dst.protonum)); return seq_putc(s, '\n'); } @@ -889,6 +888,7 @@ static int init_or_cleanup(int init) return ret; cleanup: + synchronize_net(); #ifdef CONFIG_SYSCTL unregister_sysctl_table(ip_ct_sysctl_header); cleanup_localinops: @@ -971,6 +971,14 @@ void need_ip_conntrack(void) { } +#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS +EXPORT_SYMBOL_GPL(ip_conntrack_chain); +EXPORT_SYMBOL_GPL(ip_conntrack_expect_chain); +EXPORT_SYMBOL_GPL(ip_conntrack_register_notifier); +EXPORT_SYMBOL_GPL(ip_conntrack_unregister_notifier); +EXPORT_SYMBOL_GPL(__ip_ct_event_cache_init); +EXPORT_PER_CPU_SYMBOL_GPL(ip_conntrack_ecache); +#endif EXPORT_SYMBOL(ip_conntrack_protocol_register); EXPORT_SYMBOL(ip_conntrack_protocol_unregister); EXPORT_SYMBOL(ip_ct_get_tuple); @@ -982,12 +990,16 @@ EXPORT_SYMBOL(ip_conntrack_helper_register); EXPORT_SYMBOL(ip_conntrack_helper_unregister); EXPORT_SYMBOL(ip_ct_iterate_cleanup); EXPORT_SYMBOL(ip_ct_refresh_acct); -EXPORT_SYMBOL(ip_ct_protos); -EXPORT_SYMBOL(ip_ct_find_proto); + EXPORT_SYMBOL(ip_conntrack_expect_alloc); EXPORT_SYMBOL(ip_conntrack_expect_put); +EXPORT_SYMBOL_GPL(ip_conntrack_expect_find_get); EXPORT_SYMBOL(ip_conntrack_expect_related); EXPORT_SYMBOL(ip_conntrack_unexpect_related); +EXPORT_SYMBOL_GPL(ip_conntrack_expect_list); +EXPORT_SYMBOL_GPL(__ip_conntrack_expect_find); +EXPORT_SYMBOL_GPL(__ip_ct_expect_unlink_destroy); + EXPORT_SYMBOL(ip_conntrack_tuple_taken); EXPORT_SYMBOL(ip_ct_gather_frags); EXPORT_SYMBOL(ip_conntrack_htable_size); @@ -995,7 +1007,28 @@ EXPORT_SYMBOL(ip_conntrack_lock); EXPORT_SYMBOL(ip_conntrack_hash); EXPORT_SYMBOL(ip_conntrack_untracked); EXPORT_SYMBOL_GPL(ip_conntrack_find_get); -EXPORT_SYMBOL_GPL(ip_conntrack_put); #ifdef CONFIG_IP_NF_NAT_NEEDED EXPORT_SYMBOL(ip_conntrack_tcp_update); #endif + +EXPORT_SYMBOL_GPL(ip_conntrack_flush); +EXPORT_SYMBOL_GPL(__ip_conntrack_find); + +EXPORT_SYMBOL_GPL(ip_conntrack_alloc); +EXPORT_SYMBOL_GPL(ip_conntrack_free); +EXPORT_SYMBOL_GPL(ip_conntrack_hash_insert); + +EXPORT_SYMBOL_GPL(ip_ct_remove_expectations); + +EXPORT_SYMBOL_GPL(ip_conntrack_helper_find_get); +EXPORT_SYMBOL_GPL(ip_conntrack_helper_put); +EXPORT_SYMBOL_GPL(__ip_conntrack_helper_find_byname); + +EXPORT_SYMBOL_GPL(ip_conntrack_proto_find_get); +EXPORT_SYMBOL_GPL(ip_conntrack_proto_put); +EXPORT_SYMBOL_GPL(__ip_conntrack_proto_find); +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) +EXPORT_SYMBOL_GPL(ip_ct_port_tuple_to_nfattr); +EXPORT_SYMBOL_GPL(ip_ct_port_nfattr_to_tuple); +#endif diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c index 739b6dde1c82..1adedb743f60 100644 --- a/net/ipv4/netfilter/ip_nat_core.c +++ b/net/ipv4/netfilter/ip_nat_core.c @@ -47,8 +47,39 @@ DEFINE_RWLOCK(ip_nat_lock); static unsigned int ip_nat_htable_size; static struct list_head *bysource; + +#define MAX_IP_NAT_PROTO 256 struct ip_nat_protocol *ip_nat_protos[MAX_IP_NAT_PROTO]; +static inline struct ip_nat_protocol * +__ip_nat_proto_find(u_int8_t protonum) +{ + return ip_nat_protos[protonum]; +} + +struct ip_nat_protocol * +ip_nat_proto_find_get(u_int8_t protonum) +{ + struct ip_nat_protocol *p; + + /* we need to disable preemption to make sure 'p' doesn't get + * removed until we've grabbed the reference */ + preempt_disable(); + p = __ip_nat_proto_find(protonum); + if (p) { + if (!try_module_get(p->me)) + p = &ip_nat_unknown_protocol; + } + preempt_enable(); + + return p; +} + +void +ip_nat_proto_put(struct ip_nat_protocol *p) +{ + module_put(p->me); +} /* We keep an extra hash for each conntrack, for fast searching. */ static inline unsigned int @@ -103,7 +134,8 @@ static int in_range(const struct ip_conntrack_tuple *tuple, const struct ip_nat_range *range) { - struct ip_nat_protocol *proto = ip_nat_find_proto(tuple->dst.protonum); + struct ip_nat_protocol *proto = + __ip_nat_proto_find(tuple->dst.protonum); /* If we are supposed to map IPs, then we must be in the range specified, otherwise let this drag us onto a new src IP. */ @@ -216,8 +248,7 @@ get_unique_tuple(struct ip_conntrack_tuple *tuple, struct ip_conntrack *conntrack, enum ip_nat_manip_type maniptype) { - struct ip_nat_protocol *proto - = ip_nat_find_proto(orig_tuple->dst.protonum); + struct ip_nat_protocol *proto; /* 1) If this srcip/proto/src-proto-part is currently mapped, and that same mapping gives a unique tuple within the given @@ -242,14 +273,20 @@ get_unique_tuple(struct ip_conntrack_tuple *tuple, /* 3) The per-protocol part of the manip is made to map into the range to make a unique tuple. */ + proto = ip_nat_proto_find_get(orig_tuple->dst.protonum); + /* Only bother mapping if it's not already in range and unique */ if ((!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED) || proto->in_range(tuple, maniptype, &range->min, &range->max)) - && !ip_nat_used_tuple(tuple, conntrack)) + && !ip_nat_used_tuple(tuple, conntrack)) { + ip_nat_proto_put(proto); return; + } /* Last change: get protocol to try to obtain unique tuple. */ proto->unique_tuple(tuple, range, maniptype, conntrack); + + ip_nat_proto_put(proto); } unsigned int @@ -320,17 +357,20 @@ manip_pkt(u_int16_t proto, enum ip_nat_manip_type maniptype) { struct iphdr *iph; + struct ip_nat_protocol *p; - (*pskb)->nfcache |= NFC_ALTERED; - if (!skb_ip_make_writable(pskb, iphdroff + sizeof(*iph))) + if (!skb_make_writable(pskb, iphdroff + sizeof(*iph))) return 0; iph = (void *)(*pskb)->data + iphdroff; /* Manipulate protcol part. */ - if (!ip_nat_find_proto(proto)->manip_pkt(pskb, iphdroff, - target, maniptype)) + p = ip_nat_proto_find_get(proto); + if (!p->manip_pkt(pskb, iphdroff, target, maniptype)) { + ip_nat_proto_put(p); return 0; + } + ip_nat_proto_put(p); iph = (void *)(*pskb)->data + iphdroff; @@ -391,7 +431,7 @@ int icmp_reply_translation(struct sk_buff **pskb, struct ip_conntrack_tuple inner, target; int hdrlen = (*pskb)->nh.iph->ihl * 4; - if (!skb_ip_make_writable(pskb, hdrlen + sizeof(*inside))) + if (!skb_make_writable(pskb, hdrlen + sizeof(*inside))) return 0; inside = (void *)(*pskb)->data + (*pskb)->nh.iph->ihl*4; @@ -426,7 +466,8 @@ int icmp_reply_translation(struct sk_buff **pskb, if (!ip_ct_get_tuple(&inside->ip, *pskb, (*pskb)->nh.iph->ihl*4 + sizeof(struct icmphdr) + inside->ip.ihl*4, - &inner, ip_ct_find_proto(inside->ip.protocol))) + &inner, + __ip_conntrack_proto_find(inside->ip.protocol))) return 0; /* Change inner back to look like incoming packet. We do the @@ -496,6 +537,49 @@ void ip_nat_protocol_unregister(struct ip_nat_protocol *proto) synchronize_net(); } +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) +int +ip_nat_port_range_to_nfattr(struct sk_buff *skb, + const struct ip_nat_range *range) +{ + NFA_PUT(skb, CTA_PROTONAT_PORT_MIN, sizeof(u_int16_t), + &range->min.tcp.port); + NFA_PUT(skb, CTA_PROTONAT_PORT_MAX, sizeof(u_int16_t), + &range->max.tcp.port); + + return 0; + +nfattr_failure: + return -1; +} + +int +ip_nat_port_nfattr_to_range(struct nfattr *tb[], struct ip_nat_range *range) +{ + int ret = 0; + + /* we have to return whether we actually parsed something or not */ + + if (tb[CTA_PROTONAT_PORT_MIN-1]) { + ret = 1; + range->min.tcp.port = + *(u_int16_t *)NFA_DATA(tb[CTA_PROTONAT_PORT_MIN-1]); + } + + if (!tb[CTA_PROTONAT_PORT_MAX-1]) { + if (ret) + range->max.tcp.port = range->min.tcp.port; + } else { + ret = 1; + range->max.tcp.port = + *(u_int16_t *)NFA_DATA(tb[CTA_PROTONAT_PORT_MAX-1]); + } + + return ret; +} +#endif + int __init ip_nat_init(void) { size_t i; diff --git a/net/ipv4/netfilter/ip_nat_helper.c b/net/ipv4/netfilter/ip_nat_helper.c index 158f34f32c04..d2dd5d313556 100644 --- a/net/ipv4/netfilter/ip_nat_helper.c +++ b/net/ipv4/netfilter/ip_nat_helper.c @@ -168,7 +168,7 @@ ip_nat_mangle_tcp_packet(struct sk_buff **pskb, struct tcphdr *tcph; int datalen; - if (!skb_ip_make_writable(pskb, (*pskb)->len)) + if (!skb_make_writable(pskb, (*pskb)->len)) return 0; if (rep_len > match_len @@ -228,7 +228,7 @@ ip_nat_mangle_udp_packet(struct sk_buff **pskb, match_offset + match_len) return 0; - if (!skb_ip_make_writable(pskb, (*pskb)->len)) + if (!skb_make_writable(pskb, (*pskb)->len)) return 0; if (rep_len > match_len @@ -315,7 +315,7 @@ ip_nat_sack_adjust(struct sk_buff **pskb, optoff = (*pskb)->nh.iph->ihl*4 + sizeof(struct tcphdr); optend = (*pskb)->nh.iph->ihl*4 + tcph->doff*4; - if (!skb_ip_make_writable(pskb, optend)) + if (!skb_make_writable(pskb, optend)) return 0; dir = CTINFO2DIR(ctinfo); @@ -363,7 +363,7 @@ ip_nat_seq_adjust(struct sk_buff **pskb, this_way = &ct->nat.info.seq[dir]; other_way = &ct->nat.info.seq[!dir]; - if (!skb_ip_make_writable(pskb, (*pskb)->nh.iph->ihl*4+sizeof(*tcph))) + if (!skb_make_writable(pskb, (*pskb)->nh.iph->ihl*4+sizeof(*tcph))) return 0; tcph = (void *)(*pskb)->data + (*pskb)->nh.iph->ihl*4; diff --git a/net/ipv4/netfilter/ip_nat_proto_icmp.c b/net/ipv4/netfilter/ip_nat_proto_icmp.c index 6596c9ee1655..938719043999 100644 --- a/net/ipv4/netfilter/ip_nat_proto_icmp.c +++ b/net/ipv4/netfilter/ip_nat_proto_icmp.c @@ -62,7 +62,7 @@ icmp_manip_pkt(struct sk_buff **pskb, struct icmphdr *hdr; unsigned int hdroff = iphdroff + iph->ihl*4; - if (!skb_ip_make_writable(pskb, hdroff + sizeof(*hdr))) + if (!skb_make_writable(pskb, hdroff + sizeof(*hdr))) return 0; hdr = (struct icmphdr *)((*pskb)->data + hdroff); @@ -106,11 +106,18 @@ icmp_print_range(char *buffer, const struct ip_nat_range *range) else return 0; } -struct ip_nat_protocol ip_nat_protocol_icmp -= { "ICMP", IPPROTO_ICMP, - icmp_manip_pkt, - icmp_in_range, - icmp_unique_tuple, - icmp_print, - icmp_print_range +struct ip_nat_protocol ip_nat_protocol_icmp = { + .name = "ICMP", + .protonum = IPPROTO_ICMP, + .me = THIS_MODULE, + .manip_pkt = icmp_manip_pkt, + .in_range = icmp_in_range, + .unique_tuple = icmp_unique_tuple, + .print = icmp_print, + .print_range = icmp_print_range, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .range_to_nfattr = ip_nat_port_range_to_nfattr, + .nfattr_to_range = ip_nat_port_nfattr_to_range, +#endif }; diff --git a/net/ipv4/netfilter/ip_nat_proto_tcp.c b/net/ipv4/netfilter/ip_nat_proto_tcp.c index a98e36d2b3c6..1d381bf68574 100644 --- a/net/ipv4/netfilter/ip_nat_proto_tcp.c +++ b/net/ipv4/netfilter/ip_nat_proto_tcp.c @@ -12,6 +12,7 @@ #include <linux/ip.h> #include <linux/tcp.h> #include <linux/if.h> +#include <linux/netfilter/nfnetlink_conntrack.h> #include <linux/netfilter_ipv4/ip_nat.h> #include <linux/netfilter_ipv4/ip_nat_rule.h> #include <linux/netfilter_ipv4/ip_nat_protocol.h> @@ -102,7 +103,7 @@ tcp_manip_pkt(struct sk_buff **pskb, if ((*pskb)->len >= hdroff + sizeof(struct tcphdr)) hdrsize = sizeof(struct tcphdr); - if (!skb_ip_make_writable(pskb, hdroff + hdrsize)) + if (!skb_make_writable(pskb, hdroff + hdrsize)) return 0; iph = (struct iphdr *)((*pskb)->data + iphdroff); @@ -169,11 +170,18 @@ tcp_print_range(char *buffer, const struct ip_nat_range *range) else return 0; } -struct ip_nat_protocol ip_nat_protocol_tcp -= { "TCP", IPPROTO_TCP, - tcp_manip_pkt, - tcp_in_range, - tcp_unique_tuple, - tcp_print, - tcp_print_range +struct ip_nat_protocol ip_nat_protocol_tcp = { + .name = "TCP", + .protonum = IPPROTO_TCP, + .me = THIS_MODULE, + .manip_pkt = tcp_manip_pkt, + .in_range = tcp_in_range, + .unique_tuple = tcp_unique_tuple, + .print = tcp_print, + .print_range = tcp_print_range, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .range_to_nfattr = ip_nat_port_range_to_nfattr, + .nfattr_to_range = ip_nat_port_nfattr_to_range, +#endif }; diff --git a/net/ipv4/netfilter/ip_nat_proto_udp.c b/net/ipv4/netfilter/ip_nat_proto_udp.c index 9f66e5625664..c4906e1aa24a 100644 --- a/net/ipv4/netfilter/ip_nat_proto_udp.c +++ b/net/ipv4/netfilter/ip_nat_proto_udp.c @@ -94,7 +94,7 @@ udp_manip_pkt(struct sk_buff **pskb, u32 oldip, newip; u16 *portptr, newport; - if (!skb_ip_make_writable(pskb, hdroff + sizeof(*hdr))) + if (!skb_make_writable(pskb, hdroff + sizeof(*hdr))) return 0; iph = (struct iphdr *)((*pskb)->data + iphdroff); @@ -156,11 +156,18 @@ udp_print_range(char *buffer, const struct ip_nat_range *range) else return 0; } -struct ip_nat_protocol ip_nat_protocol_udp -= { "UDP", IPPROTO_UDP, - udp_manip_pkt, - udp_in_range, - udp_unique_tuple, - udp_print, - udp_print_range +struct ip_nat_protocol ip_nat_protocol_udp = { + .name = "UDP", + .protonum = IPPROTO_UDP, + .me = THIS_MODULE, + .manip_pkt = udp_manip_pkt, + .in_range = udp_in_range, + .unique_tuple = udp_unique_tuple, + .print = udp_print, + .print_range = udp_print_range, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .range_to_nfattr = ip_nat_port_range_to_nfattr, + .nfattr_to_range = ip_nat_port_nfattr_to_range, +#endif }; diff --git a/net/ipv4/netfilter/ip_nat_proto_unknown.c b/net/ipv4/netfilter/ip_nat_proto_unknown.c index f5525bd58d16..99bbef56f84e 100644 --- a/net/ipv4/netfilter/ip_nat_proto_unknown.c +++ b/net/ipv4/netfilter/ip_nat_proto_unknown.c @@ -61,10 +61,11 @@ unknown_print_range(char *buffer, const struct ip_nat_range *range) } struct ip_nat_protocol ip_nat_unknown_protocol = { - "unknown", 0, - unknown_manip_pkt, - unknown_in_range, - unknown_unique_tuple, - unknown_print, - unknown_print_range + .name = "unknown", + .me = THIS_MODULE, + .manip_pkt = unknown_manip_pkt, + .in_range = unknown_in_range, + .unique_tuple = unknown_unique_tuple, + .print = unknown_print, + .print_range = unknown_print_range }; diff --git a/net/ipv4/netfilter/ip_nat_snmp_basic.c b/net/ipv4/netfilter/ip_nat_snmp_basic.c index 2a48b6e635ae..93b2c5111bb2 100644 --- a/net/ipv4/netfilter/ip_nat_snmp_basic.c +++ b/net/ipv4/netfilter/ip_nat_snmp_basic.c @@ -1275,7 +1275,7 @@ static int help(struct sk_buff **pskb, return NF_DROP; } - if (!skb_ip_make_writable(pskb, (*pskb)->len)) + if (!skb_make_writable(pskb, (*pskb)->len)) return NF_DROP; spin_lock_bh(&snmp_lock); diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c index 91d5ea1dbbc9..89db052add81 100644 --- a/net/ipv4/netfilter/ip_nat_standalone.c +++ b/net/ipv4/netfilter/ip_nat_standalone.c @@ -73,8 +73,6 @@ ip_nat_fn(unsigned int hooknum, IP_NF_ASSERT(!((*pskb)->nh.iph->frag_off & htons(IP_MF|IP_OFFSET))); - (*pskb)->nfcache |= NFC_UNKNOWN; - /* If we had a hardware checksum before, it's now invalid */ if ((*pskb)->ip_summed == CHECKSUM_HW) if (skb_checksum_help(*pskb, (out == NULL))) @@ -396,6 +394,8 @@ module_exit(fini); EXPORT_SYMBOL(ip_nat_setup_info); EXPORT_SYMBOL(ip_nat_protocol_register); EXPORT_SYMBOL(ip_nat_protocol_unregister); +EXPORT_SYMBOL_GPL(ip_nat_proto_find_get); +EXPORT_SYMBOL_GPL(ip_nat_proto_put); EXPORT_SYMBOL(ip_nat_cheat_check); EXPORT_SYMBOL(ip_nat_mangle_tcp_packet); EXPORT_SYMBOL(ip_nat_mangle_udp_packet); diff --git a/net/ipv4/netfilter/ip_queue.c b/net/ipv4/netfilter/ip_queue.c index c6baa8174389..d54f14d926f6 100644 --- a/net/ipv4/netfilter/ip_queue.c +++ b/net/ipv4/netfilter/ip_queue.c @@ -43,17 +43,10 @@ #define NET_IPQ_QMAX 2088 #define NET_IPQ_QMAX_NAME "ip_queue_maxlen" -struct ipq_rt_info { - __u8 tos; - __u32 daddr; - __u32 saddr; -}; - struct ipq_queue_entry { struct list_head list; struct nf_info *info; struct sk_buff *skb; - struct ipq_rt_info rt_info; }; typedef int (*ipq_cmpfn)(struct ipq_queue_entry *, unsigned long); @@ -247,8 +240,8 @@ ipq_build_packet_message(struct ipq_queue_entry *entry, int *errp) pmsg->packet_id = (unsigned long )entry; pmsg->data_len = data_len; - pmsg->timestamp_sec = entry->skb->stamp.tv_sec; - pmsg->timestamp_usec = entry->skb->stamp.tv_usec; + pmsg->timestamp_sec = skb_tv_base.tv_sec + entry->skb->tstamp.off_sec; + pmsg->timestamp_usec = skb_tv_base.tv_usec + entry->skb->tstamp.off_usec; pmsg->mark = entry->skb->nfmark; pmsg->hook = entry->info->hook; pmsg->hw_protocol = entry->skb->protocol; @@ -287,7 +280,8 @@ nlmsg_failure: } static int -ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, void *data) +ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, + unsigned int queuenum, void *data) { int status = -EINVAL; struct sk_buff *nskb; @@ -305,14 +299,6 @@ ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, void *data) entry->info = info; entry->skb = skb; - if (entry->info->hook == NF_IP_LOCAL_OUT) { - struct iphdr *iph = skb->nh.iph; - - entry->rt_info.tos = iph->tos; - entry->rt_info.daddr = iph->daddr; - entry->rt_info.saddr = iph->saddr; - } - nskb = ipq_build_packet_message(entry, &status); if (nskb == NULL) goto err_out_free; @@ -388,24 +374,11 @@ ipq_mangle_ipv4(ipq_verdict_msg_t *v, struct ipq_queue_entry *e) } skb_put(e->skb, diff); } - if (!skb_ip_make_writable(&e->skb, v->data_len)) + if (!skb_make_writable(&e->skb, v->data_len)) return -ENOMEM; memcpy(e->skb->data, v->payload, v->data_len); e->skb->ip_summed = CHECKSUM_NONE; - e->skb->nfcache |= NFC_ALTERED; - - /* - * Extra routing may needed on local out, as the QUEUE target never - * returns control to the table. - */ - if (e->info->hook == NF_IP_LOCAL_OUT) { - struct iphdr *iph = e->skb->nh.iph; - - if (!(iph->tos == e->rt_info.tos - && iph->daddr == e->rt_info.daddr - && iph->saddr == e->rt_info.saddr)) - return ip_route_me_harder(&e->skb); - } + return 0; } @@ -683,6 +656,11 @@ ipq_get_info(char *buffer, char **start, off_t offset, int length) } #endif /* CONFIG_PROC_FS */ +static struct nf_queue_handler nfqh = { + .name = "ip_queue", + .outfn = &ipq_enqueue_packet, +}; + static int init_or_cleanup(int init) { @@ -693,7 +671,8 @@ init_or_cleanup(int init) goto cleanup; netlink_register_notifier(&ipq_nl_notifier); - ipqnl = netlink_kernel_create(NETLINK_FIREWALL, ipq_rcv_sk); + ipqnl = netlink_kernel_create(NETLINK_FIREWALL, 0, ipq_rcv_sk, + THIS_MODULE); if (ipqnl == NULL) { printk(KERN_ERR "ip_queue: failed to create netlink socket\n"); goto cleanup_netlink_notifier; @@ -710,7 +689,7 @@ init_or_cleanup(int init) register_netdevice_notifier(&ipq_dev_notifier); ipq_sysctl_header = register_sysctl_table(ipq_root_table, 0); - status = nf_register_queue_handler(PF_INET, ipq_enqueue_packet, NULL); + status = nf_register_queue_handler(PF_INET, &nfqh); if (status < 0) { printk(KERN_ERR "ip_queue: failed to register queue handler\n"); goto cleanup_sysctl; @@ -718,7 +697,7 @@ init_or_cleanup(int init) return status; cleanup: - nf_unregister_queue_handler(PF_INET); + nf_unregister_queue_handlers(&nfqh); synchronize_net(); ipq_flush(NF_DROP); diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c index c88dfcd38c56..eef99a1b5de6 100644 --- a/net/ipv4/netfilter/ip_tables.c +++ b/net/ipv4/netfilter/ip_tables.c @@ -312,7 +312,6 @@ ipt_do_table(struct sk_buff **pskb, do { IP_NF_ASSERT(e); IP_NF_ASSERT(back); - (*pskb)->nfcache |= e->nfcache; if (ip_packet_match(ip, indev, outdev, &e->ip, offset)) { struct ipt_entry_target *t; @@ -341,8 +340,8 @@ ipt_do_table(struct sk_buff **pskb, back->comefrom); continue; } - if (table_base + v - != (void *)e + e->next_offset) { + if (table_base + v != (void *)e + e->next_offset + && !(e->ip.flags & IPT_F_GOTO)) { /* Save old back ptr in next entry */ struct ipt_entry *next = (void *)e + e->next_offset; diff --git a/net/ipv4/netfilter/ipt_CLASSIFY.c b/net/ipv4/netfilter/ipt_CLASSIFY.c index 9842e6e23184..dab78d8bd494 100644 --- a/net/ipv4/netfilter/ipt_CLASSIFY.c +++ b/net/ipv4/netfilter/ipt_CLASSIFY.c @@ -32,10 +32,8 @@ target(struct sk_buff **pskb, { const struct ipt_classify_target_info *clinfo = targinfo; - if((*pskb)->priority != clinfo->priority) { + if((*pskb)->priority != clinfo->priority) (*pskb)->priority = clinfo->priority; - (*pskb)->nfcache |= NFC_ALTERED; - } return IPT_CONTINUE; } diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c index 6706d3a1bc4f..2d05cafec221 100644 --- a/net/ipv4/netfilter/ipt_CLUSTERIP.c +++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c @@ -367,7 +367,7 @@ target(struct sk_buff **pskb, #ifdef DEBUG_CLUSTERP DUMP_TUPLE(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); #endif - DEBUGP("hash=%u ct_hash=%lu ", hash, ct->mark); + DEBUGP("hash=%u ct_hash=%u ", hash, ct->mark); if (!clusterip_responsible(cipinfo->config, hash)) { DEBUGP("not responsible\n"); return NF_DROP; diff --git a/net/ipv4/netfilter/ipt_CONNMARK.c b/net/ipv4/netfilter/ipt_CONNMARK.c index 30ddd3e18eb7..134638021339 100644 --- a/net/ipv4/netfilter/ipt_CONNMARK.c +++ b/net/ipv4/netfilter/ipt_CONNMARK.c @@ -40,9 +40,9 @@ target(struct sk_buff **pskb, void *userinfo) { const struct ipt_connmark_target_info *markinfo = targinfo; - unsigned long diff; - unsigned long nfmark; - unsigned long newmark; + u_int32_t diff; + u_int32_t nfmark; + u_int32_t newmark; enum ip_conntrack_info ctinfo; struct ip_conntrack *ct = ip_conntrack_get((*pskb), &ctinfo); @@ -61,10 +61,8 @@ target(struct sk_buff **pskb, case IPT_CONNMARK_RESTORE: nfmark = (*pskb)->nfmark; diff = (ct->mark ^ nfmark) & markinfo->mask; - if (diff != 0) { + if (diff != 0) (*pskb)->nfmark = nfmark ^ diff; - (*pskb)->nfcache |= NFC_ALTERED; - } break; } } @@ -94,6 +92,11 @@ checkentry(const char *tablename, } } + if (matchinfo->mark > 0xffffffff || matchinfo->mask > 0xffffffff) { + printk(KERN_WARNING "CONNMARK: Only supports 32bit mark\n"); + return 0; + } + return 1; } diff --git a/net/ipv4/netfilter/ipt_DSCP.c b/net/ipv4/netfilter/ipt_DSCP.c index 3ea4509099f9..6e319570a28c 100644 --- a/net/ipv4/netfilter/ipt_DSCP.c +++ b/net/ipv4/netfilter/ipt_DSCP.c @@ -39,7 +39,7 @@ target(struct sk_buff **pskb, if (((*pskb)->nh.iph->tos & IPT_DSCP_MASK) != sh_dscp) { u_int16_t diffs[2]; - if (!skb_ip_make_writable(pskb, sizeof(struct iphdr))) + if (!skb_make_writable(pskb, sizeof(struct iphdr))) return NF_DROP; diffs[0] = htons((*pskb)->nh.iph->tos) ^ 0xFFFF; @@ -51,7 +51,6 @@ target(struct sk_buff **pskb, sizeof(diffs), (*pskb)->nh.iph->check ^ 0xFFFF)); - (*pskb)->nfcache |= NFC_ALTERED; } return IPT_CONTINUE; } diff --git a/net/ipv4/netfilter/ipt_ECN.c b/net/ipv4/netfilter/ipt_ECN.c index 94a0ce1c1c9d..a1319693f648 100644 --- a/net/ipv4/netfilter/ipt_ECN.c +++ b/net/ipv4/netfilter/ipt_ECN.c @@ -31,7 +31,7 @@ set_ect_ip(struct sk_buff **pskb, const struct ipt_ECN_info *einfo) != (einfo->ip_ect & IPT_ECN_IP_MASK)) { u_int16_t diffs[2]; - if (!skb_ip_make_writable(pskb, sizeof(struct iphdr))) + if (!skb_make_writable(pskb, sizeof(struct iphdr))) return 0; diffs[0] = htons((*pskb)->nh.iph->tos) ^ 0xFFFF; @@ -43,7 +43,6 @@ set_ect_ip(struct sk_buff **pskb, const struct ipt_ECN_info *einfo) sizeof(diffs), (*pskb)->nh.iph->check ^0xFFFF)); - (*pskb)->nfcache |= NFC_ALTERED; } return 1; } @@ -67,7 +66,7 @@ set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo, int inward) tcph->cwr == einfo->proto.tcp.cwr))) return 1; - if (!skb_ip_make_writable(pskb, (*pskb)->nh.iph->ihl*4+sizeof(*tcph))) + if (!skb_make_writable(pskb, (*pskb)->nh.iph->ihl*4+sizeof(*tcph))) return 0; tcph = (void *)(*pskb)->nh.iph + (*pskb)->nh.iph->ihl*4; @@ -87,7 +86,6 @@ set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo, int inward) tcph->check = csum_fold(csum_partial((char *)diffs, sizeof(diffs), tcph->check^0xFFFF)); - (*pskb)->nfcache |= NFC_ALTERED; return 1; } diff --git a/net/ipv4/netfilter/ipt_LOG.c b/net/ipv4/netfilter/ipt_LOG.c index ef08733d26da..92ed050fac69 100644 --- a/net/ipv4/netfilter/ipt_LOG.c +++ b/net/ipv4/netfilter/ipt_LOG.c @@ -27,10 +27,6 @@ MODULE_LICENSE("GPL"); MODULE_AUTHOR("Netfilter Core Team <coreteam@netfilter.org>"); MODULE_DESCRIPTION("iptables syslog logging module"); -static unsigned int nflog = 1; -module_param(nflog, int, 0400); -MODULE_PARM_DESC(nflog, "register as internal netfilter logging module"); - #if 0 #define DEBUGP printk #else @@ -41,11 +37,17 @@ MODULE_PARM_DESC(nflog, "register as internal netfilter logging module"); static DEFINE_SPINLOCK(log_lock); /* One level of recursion won't kill us */ -static void dump_packet(const struct ipt_log_info *info, +static void dump_packet(const struct nf_loginfo *info, const struct sk_buff *skb, unsigned int iphoff) { struct iphdr _iph, *ih; + unsigned int logflags; + + if (info->type == NF_LOG_TYPE_LOG) + logflags = info->u.log.logflags; + else + logflags = NF_LOG_MASK; ih = skb_header_pointer(skb, iphoff, sizeof(_iph), &_iph); if (ih == NULL) { @@ -76,7 +78,7 @@ static void dump_packet(const struct ipt_log_info *info, if (ntohs(ih->frag_off) & IP_OFFSET) printk("FRAG:%u ", ntohs(ih->frag_off) & IP_OFFSET); - if ((info->logflags & IPT_LOG_IPOPT) + if ((logflags & IPT_LOG_IPOPT) && ih->ihl * 4 > sizeof(struct iphdr)) { unsigned char _opt[4 * 15 - sizeof(struct iphdr)], *op; unsigned int i, optsize; @@ -119,7 +121,7 @@ static void dump_packet(const struct ipt_log_info *info, printk("SPT=%u DPT=%u ", ntohs(th->source), ntohs(th->dest)); /* Max length: 30 "SEQ=4294967295 ACK=4294967295 " */ - if (info->logflags & IPT_LOG_TCPSEQ) + if (logflags & IPT_LOG_TCPSEQ) printk("SEQ=%u ACK=%u ", ntohl(th->seq), ntohl(th->ack_seq)); /* Max length: 13 "WINDOW=65535 " */ @@ -146,7 +148,7 @@ static void dump_packet(const struct ipt_log_info *info, /* Max length: 11 "URGP=65535 " */ printk("URGP=%u ", ntohs(th->urg_ptr)); - if ((info->logflags & IPT_LOG_TCPOPT) + if ((logflags & IPT_LOG_TCPOPT) && th->doff * 4 > sizeof(struct tcphdr)) { unsigned char _opt[4 * 15 - sizeof(struct tcphdr)]; unsigned char *op; @@ -328,7 +330,7 @@ static void dump_packet(const struct ipt_log_info *info, } /* Max length: 15 "UID=4294967295 " */ - if ((info->logflags & IPT_LOG_UID) && !iphoff && skb->sk) { + if ((logflags & IPT_LOG_UID) && !iphoff && skb->sk) { read_lock_bh(&skb->sk->sk_callback_lock); if (skb->sk->sk_socket && skb->sk->sk_socket->file) printk("UID=%u ", skb->sk->sk_socket->file->f_uid); @@ -349,19 +351,31 @@ static void dump_packet(const struct ipt_log_info *info, /* maxlen = 230+ 91 + 230 + 252 = 803 */ } +struct nf_loginfo default_loginfo = { + .type = NF_LOG_TYPE_LOG, + .u = { + .log = { + .level = 0, + .logflags = NF_LOG_MASK, + }, + }, +}; + static void -ipt_log_packet(unsigned int hooknum, +ipt_log_packet(unsigned int pf, + unsigned int hooknum, const struct sk_buff *skb, const struct net_device *in, const struct net_device *out, - const struct ipt_log_info *loginfo, - const char *level_string, + const struct nf_loginfo *loginfo, const char *prefix) { + if (!loginfo) + loginfo = &default_loginfo; + spin_lock_bh(&log_lock); - printk(level_string); - printk("%sIN=%s OUT=%s ", - prefix == NULL ? loginfo->prefix : prefix, + printk("<%d>%sIN=%s OUT=%s ", loginfo->u.log.level, + prefix, in ? in->name : "", out ? out->name : ""); #ifdef CONFIG_BRIDGE_NETFILTER @@ -405,28 +419,15 @@ ipt_log_target(struct sk_buff **pskb, void *userinfo) { const struct ipt_log_info *loginfo = targinfo; - char level_string[4] = "< >"; + struct nf_loginfo li; - level_string[1] = '0' + (loginfo->level % 8); - ipt_log_packet(hooknum, *pskb, in, out, loginfo, level_string, NULL); + li.type = NF_LOG_TYPE_LOG; + li.u.log.level = loginfo->level; + li.u.log.logflags = loginfo->logflags; - return IPT_CONTINUE; -} + nf_log_packet(PF_INET, hooknum, *pskb, in, out, &li, loginfo->prefix); -static void -ipt_logfn(unsigned int hooknum, - const struct sk_buff *skb, - const struct net_device *in, - const struct net_device *out, - const char *prefix) -{ - struct ipt_log_info loginfo = { - .level = 0, - .logflags = IPT_LOG_MASK, - .prefix = "" - }; - - ipt_log_packet(hooknum, skb, in, out, &loginfo, KERN_WARNING, prefix); + return IPT_CONTINUE; } static int ipt_log_checkentry(const char *tablename, @@ -464,20 +465,29 @@ static struct ipt_target ipt_log_reg = { .me = THIS_MODULE, }; +static struct nf_logger ipt_log_logger ={ + .name = "ipt_LOG", + .logfn = &ipt_log_packet, + .me = THIS_MODULE, +}; + static int __init init(void) { if (ipt_register_target(&ipt_log_reg)) return -EINVAL; - if (nflog) - nf_log_register(PF_INET, &ipt_logfn); + if (nf_log_register(PF_INET, &ipt_log_logger) < 0) { + printk(KERN_WARNING "ipt_LOG: not logging via system console " + "since somebody else already registered for PF_INET\n"); + /* we cannot make module load fail here, since otherwise + * iptables userspace would abort */ + } return 0; } static void __exit fini(void) { - if (nflog) - nf_log_unregister(PF_INET, &ipt_logfn); + nf_log_unregister_logger(&ipt_log_logger); ipt_unregister_target(&ipt_log_reg); } diff --git a/net/ipv4/netfilter/ipt_MARK.c b/net/ipv4/netfilter/ipt_MARK.c index 33c6f9b63b8d..52b4f2c296bf 100644 --- a/net/ipv4/netfilter/ipt_MARK.c +++ b/net/ipv4/netfilter/ipt_MARK.c @@ -29,10 +29,9 @@ target_v0(struct sk_buff **pskb, { const struct ipt_mark_target_info *markinfo = targinfo; - if((*pskb)->nfmark != markinfo->mark) { + if((*pskb)->nfmark != markinfo->mark) (*pskb)->nfmark = markinfo->mark; - (*pskb)->nfcache |= NFC_ALTERED; - } + return IPT_CONTINUE; } @@ -61,10 +60,9 @@ target_v1(struct sk_buff **pskb, break; } - if((*pskb)->nfmark != mark) { + if((*pskb)->nfmark != mark) (*pskb)->nfmark = mark; - (*pskb)->nfcache |= NFC_ALTERED; - } + return IPT_CONTINUE; } @@ -76,6 +74,8 @@ checkentry_v0(const char *tablename, unsigned int targinfosize, unsigned int hook_mask) { + struct ipt_mark_target_info *markinfo = targinfo; + if (targinfosize != IPT_ALIGN(sizeof(struct ipt_mark_target_info))) { printk(KERN_WARNING "MARK: targinfosize %u != %Zu\n", targinfosize, @@ -88,6 +88,11 @@ checkentry_v0(const char *tablename, return 0; } + if (markinfo->mark > 0xffffffff) { + printk(KERN_WARNING "MARK: Only supports 32bit wide mark\n"); + return 0; + } + return 1; } @@ -120,6 +125,11 @@ checkentry_v1(const char *tablename, return 0; } + if (markinfo->mark > 0xffffffff) { + printk(KERN_WARNING "MARK: Only supports 32bit wide mark\n"); + return 0; + } + return 1; } diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c index 91e74502c3d3..2f3e181c8e97 100644 --- a/net/ipv4/netfilter/ipt_MASQUERADE.c +++ b/net/ipv4/netfilter/ipt_MASQUERADE.c @@ -86,11 +86,6 @@ masquerade_target(struct sk_buff **pskb, IP_NF_ASSERT(hooknum == NF_IP_POST_ROUTING); - /* FIXME: For the moment, don't do local packets, breaks - testsuite for 2.3.49 --RR */ - if ((*pskb)->sk) - return NF_ACCEPT; - ct = ip_conntrack_get(*pskb, &ctinfo); IP_NF_ASSERT(ct && (ctinfo == IP_CT_NEW || ctinfo == IP_CT_RELATED || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)); diff --git a/net/ipv4/netfilter/ipt_NETMAP.c b/net/ipv4/netfilter/ipt_NETMAP.c index 06254b29d034..e6e7b6095363 100644 --- a/net/ipv4/netfilter/ipt_NETMAP.c +++ b/net/ipv4/netfilter/ipt_NETMAP.c @@ -46,7 +46,8 @@ check(const char *tablename, DEBUGP(MODULENAME":check: size %u.\n", targinfosize); return 0; } - if (hook_mask & ~((1 << NF_IP_PRE_ROUTING) | (1 << NF_IP_POST_ROUTING))) { + if (hook_mask & ~((1 << NF_IP_PRE_ROUTING) | (1 << NF_IP_POST_ROUTING) | + (1 << NF_IP_LOCAL_OUT))) { DEBUGP(MODULENAME":check: bad hooks %x.\n", hook_mask); return 0; } @@ -76,12 +77,13 @@ target(struct sk_buff **pskb, struct ip_nat_range newrange; IP_NF_ASSERT(hooknum == NF_IP_PRE_ROUTING - || hooknum == NF_IP_POST_ROUTING); + || hooknum == NF_IP_POST_ROUTING + || hooknum == NF_IP_LOCAL_OUT); ct = ip_conntrack_get(*pskb, &ctinfo); netmask = ~(mr->range[0].min_ip ^ mr->range[0].max_ip); - if (hooknum == NF_IP_PRE_ROUTING) + if (hooknum == NF_IP_PRE_ROUTING || hooknum == NF_IP_LOCAL_OUT) new_ip = (*pskb)->nh.iph->daddr & ~netmask; else new_ip = (*pskb)->nh.iph->saddr & ~netmask; diff --git a/net/ipv4/netfilter/ipt_NFQUEUE.c b/net/ipv4/netfilter/ipt_NFQUEUE.c new file mode 100644 index 000000000000..3cedc9be8807 --- /dev/null +++ b/net/ipv4/netfilter/ipt_NFQUEUE.c @@ -0,0 +1,70 @@ +/* iptables module for using new netfilter netlink queue + * + * (C) 2005 by Harald Welte <laforge@netfilter.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include <linux/module.h> +#include <linux/skbuff.h> + +#include <linux/netfilter.h> +#include <linux/netfilter_ipv4/ip_tables.h> +#include <linux/netfilter_ipv4/ipt_NFQUEUE.h> + +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_DESCRIPTION("iptables NFQUEUE target"); +MODULE_LICENSE("GPL"); + +static unsigned int +target(struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + unsigned int hooknum, + const void *targinfo, + void *userinfo) +{ + const struct ipt_NFQ_info *tinfo = targinfo; + + return NF_QUEUE_NR(tinfo->queuenum); +} + +static int +checkentry(const char *tablename, + const struct ipt_entry *e, + void *targinfo, + unsigned int targinfosize, + unsigned int hook_mask) +{ + if (targinfosize != IPT_ALIGN(sizeof(struct ipt_NFQ_info))) { + printk(KERN_WARNING "NFQUEUE: targinfosize %u != %Zu\n", + targinfosize, + IPT_ALIGN(sizeof(struct ipt_NFQ_info))); + return 0; + } + + return 1; +} + +static struct ipt_target ipt_NFQ_reg = { + .name = "NFQUEUE", + .target = target, + .checkentry = checkentry, + .me = THIS_MODULE, +}; + +static int __init init(void) +{ + return ipt_register_target(&ipt_NFQ_reg); +} + +static void __exit fini(void) +{ + ipt_unregister_target(&ipt_NFQ_reg); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c index 915696446020..f115a84a4ac6 100644 --- a/net/ipv4/netfilter/ipt_REJECT.c +++ b/net/ipv4/netfilter/ipt_REJECT.c @@ -156,7 +156,6 @@ static void send_reset(struct sk_buff *oldskb, int hook) /* This packet will not be the same as the other: clear nf fields */ nf_reset(nskb); - nskb->nfcache = 0; nskb->nfmark = 0; #ifdef CONFIG_BRIDGE_NETFILTER nf_bridge_put(nskb->nf_bridge); diff --git a/net/ipv4/netfilter/ipt_TCPMSS.c b/net/ipv4/netfilter/ipt_TCPMSS.c index 7b84a254440e..8db70d6908c3 100644 --- a/net/ipv4/netfilter/ipt_TCPMSS.c +++ b/net/ipv4/netfilter/ipt_TCPMSS.c @@ -58,7 +58,7 @@ ipt_tcpmss_target(struct sk_buff **pskb, unsigned int i; u_int8_t *opt; - if (!skb_ip_make_writable(pskb, (*pskb)->len)) + if (!skb_make_writable(pskb, (*pskb)->len)) return NF_DROP; if ((*pskb)->ip_summed == CHECKSUM_HW && @@ -190,7 +190,6 @@ ipt_tcpmss_target(struct sk_buff **pskb, newmss); retmodified: - (*pskb)->nfcache |= NFC_UNKNOWN | NFC_ALTERED; return IPT_CONTINUE; } diff --git a/net/ipv4/netfilter/ipt_TOS.c b/net/ipv4/netfilter/ipt_TOS.c index 85c70d240f8b..deadb36d4428 100644 --- a/net/ipv4/netfilter/ipt_TOS.c +++ b/net/ipv4/netfilter/ipt_TOS.c @@ -33,7 +33,7 @@ target(struct sk_buff **pskb, if (((*pskb)->nh.iph->tos & IPTOS_TOS_MASK) != tosinfo->tos) { u_int16_t diffs[2]; - if (!skb_ip_make_writable(pskb, sizeof(struct iphdr))) + if (!skb_make_writable(pskb, sizeof(struct iphdr))) return NF_DROP; diffs[0] = htons((*pskb)->nh.iph->tos) ^ 0xFFFF; @@ -46,7 +46,6 @@ target(struct sk_buff **pskb, sizeof(diffs), (*pskb)->nh.iph->check ^0xFFFF)); - (*pskb)->nfcache |= NFC_ALTERED; } return IPT_CONTINUE; } diff --git a/net/ipv4/netfilter/ipt_TTL.c b/net/ipv4/netfilter/ipt_TTL.c new file mode 100644 index 000000000000..b9ae6a9382f3 --- /dev/null +++ b/net/ipv4/netfilter/ipt_TTL.c @@ -0,0 +1,119 @@ +/* TTL modification target for IP tables + * (C) 2000,2005 by Harald Welte <laforge@netfilter.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/ip.h> +#include <net/checksum.h> + +#include <linux/netfilter_ipv4/ip_tables.h> +#include <linux/netfilter_ipv4/ipt_TTL.h> + +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_DESCRIPTION("IP tables TTL modification module"); +MODULE_LICENSE("GPL"); + +static unsigned int +ipt_ttl_target(struct sk_buff **pskb, const struct net_device *in, + const struct net_device *out, unsigned int hooknum, + const void *targinfo, void *userinfo) +{ + struct iphdr *iph; + const struct ipt_TTL_info *info = targinfo; + u_int16_t diffs[2]; + int new_ttl; + + if (!skb_make_writable(pskb, (*pskb)->len)) + return NF_DROP; + + iph = (*pskb)->nh.iph; + + switch (info->mode) { + case IPT_TTL_SET: + new_ttl = info->ttl; + break; + case IPT_TTL_INC: + new_ttl = iph->ttl + info->ttl; + if (new_ttl > 255) + new_ttl = 255; + break; + case IPT_TTL_DEC: + new_ttl = iph->ttl - info->ttl; + if (new_ttl < 0) + new_ttl = 0; + break; + default: + new_ttl = iph->ttl; + break; + } + + if (new_ttl != iph->ttl) { + diffs[0] = htons(((unsigned)iph->ttl) << 8) ^ 0xFFFF; + iph->ttl = new_ttl; + diffs[1] = htons(((unsigned)iph->ttl) << 8); + iph->check = csum_fold(csum_partial((char *)diffs, + sizeof(diffs), + iph->check^0xFFFF)); + } + + return IPT_CONTINUE; +} + +static int ipt_ttl_checkentry(const char *tablename, + const struct ipt_entry *e, + void *targinfo, + unsigned int targinfosize, + unsigned int hook_mask) +{ + struct ipt_TTL_info *info = targinfo; + + if (targinfosize != IPT_ALIGN(sizeof(struct ipt_TTL_info))) { + printk(KERN_WARNING "ipt_TTL: targinfosize %u != %Zu\n", + targinfosize, + IPT_ALIGN(sizeof(struct ipt_TTL_info))); + return 0; + } + + if (strcmp(tablename, "mangle")) { + printk(KERN_WARNING "ipt_TTL: can only be called from " + "\"mangle\" table, not \"%s\"\n", tablename); + return 0; + } + + if (info->mode > IPT_TTL_MAXMODE) { + printk(KERN_WARNING "ipt_TTL: invalid or unknown Mode %u\n", + info->mode); + return 0; + } + + if ((info->mode != IPT_TTL_SET) && (info->ttl == 0)) + return 0; + + return 1; +} + +static struct ipt_target ipt_TTL = { + .name = "TTL", + .target = ipt_ttl_target, + .checkentry = ipt_ttl_checkentry, + .me = THIS_MODULE, +}; + +static int __init init(void) +{ + return ipt_register_target(&ipt_TTL); +} + +static void __exit fini(void) +{ + ipt_unregister_target(&ipt_TTL); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/netfilter/ipt_ULOG.c b/net/ipv4/netfilter/ipt_ULOG.c index 52a0076302a7..e2c14f3cb2fc 100644 --- a/net/ipv4/netfilter/ipt_ULOG.c +++ b/net/ipv4/netfilter/ipt_ULOG.c @@ -62,6 +62,7 @@ MODULE_LICENSE("GPL"); MODULE_AUTHOR("Harald Welte <laforge@gnumonks.org>"); MODULE_DESCRIPTION("iptables userspace logging module"); +MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NFLOG); #define ULOG_NL_EVENT 111 /* Harald's favorite number */ #define ULOG_MAXNLGROUPS 32 /* numer of nlgroups */ @@ -115,10 +116,10 @@ static void ulog_send(unsigned int nlgroupnum) if (ub->qlen > 1) ub->lastnlh->nlmsg_type = NLMSG_DONE; - NETLINK_CB(ub->skb).dst_groups = (1 << nlgroupnum); - DEBUGP("ipt_ULOG: throwing %d packets to netlink mask %u\n", - ub->qlen, nlgroupnum); - netlink_broadcast(nflognl, ub->skb, 0, (1 << nlgroupnum), GFP_ATOMIC); + NETLINK_CB(ub->skb).dst_group = nlgroupnum + 1; + DEBUGP("ipt_ULOG: throwing %d packets to netlink group %u\n", + ub->qlen, nlgroupnum + 1); + netlink_broadcast(nflognl, ub->skb, 0, nlgroupnum + 1, GFP_ATOMIC); ub->qlen = 0; ub->skb = NULL; @@ -219,13 +220,13 @@ static void ipt_ulog_packet(unsigned int hooknum, pm = NLMSG_DATA(nlh); /* We might not have a timestamp, get one */ - if (skb->stamp.tv_sec == 0) - do_gettimeofday((struct timeval *)&skb->stamp); + if (skb->tstamp.off_sec == 0) + __net_timestamp((struct sk_buff *)skb); /* copy hook, prefix, timestamp, payload, etc. */ pm->data_len = copy_len; - pm->timestamp_sec = skb->stamp.tv_sec; - pm->timestamp_usec = skb->stamp.tv_usec; + pm->timestamp_sec = skb_tv_base.tv_sec + skb->tstamp.off_sec; + pm->timestamp_usec = skb_tv_base.tv_usec + skb->tstamp.off_usec; pm->mark = skb->nfmark; pm->hook = hooknum; if (prefix != NULL) @@ -303,18 +304,27 @@ static unsigned int ipt_ulog_target(struct sk_buff **pskb, return IPT_CONTINUE; } -static void ipt_logfn(unsigned int hooknum, +static void ipt_logfn(unsigned int pf, + unsigned int hooknum, const struct sk_buff *skb, const struct net_device *in, const struct net_device *out, + const struct nf_loginfo *li, const char *prefix) { - struct ipt_ulog_info loginfo = { - .nl_group = ULOG_DEFAULT_NLGROUP, - .copy_range = 0, - .qthreshold = ULOG_DEFAULT_QTHRESHOLD, - .prefix = "" - }; + struct ipt_ulog_info loginfo; + + if (!li || li->type != NF_LOG_TYPE_ULOG) { + loginfo.nl_group = ULOG_DEFAULT_NLGROUP; + loginfo.copy_range = 0; + loginfo.qthreshold = ULOG_DEFAULT_QTHRESHOLD; + loginfo.prefix[0] = '\0'; + } else { + loginfo.nl_group = li->u.ulog.group; + loginfo.copy_range = li->u.ulog.copy_len; + loginfo.qthreshold = li->u.ulog.qthreshold; + strlcpy(loginfo.prefix, prefix, sizeof(loginfo.prefix)); + } ipt_ulog_packet(hooknum, skb, in, out, &loginfo, prefix); } @@ -354,6 +364,12 @@ static struct ipt_target ipt_ulog_reg = { .me = THIS_MODULE, }; +static struct nf_logger ipt_ulog_logger = { + .name = "ipt_ULOG", + .logfn = &ipt_logfn, + .me = THIS_MODULE, +}; + static int __init init(void) { int i; @@ -372,7 +388,8 @@ static int __init init(void) ulog_buffers[i].timer.data = i; } - nflognl = netlink_kernel_create(NETLINK_NFLOG, NULL); + nflognl = netlink_kernel_create(NETLINK_NFLOG, ULOG_MAXNLGROUPS, NULL, + THIS_MODULE); if (!nflognl) return -ENOMEM; @@ -381,7 +398,7 @@ static int __init init(void) return -EINVAL; } if (nflog) - nf_log_register(PF_INET, &ipt_logfn); + nf_log_register(PF_INET, &ipt_ulog_logger); return 0; } @@ -394,7 +411,7 @@ static void __exit fini(void) DEBUGP("ipt_ULOG: cleanup_module\n"); if (nflog) - nf_log_unregister(PF_INET, &ipt_logfn); + nf_log_unregister_logger(&ipt_ulog_logger); ipt_unregister_target(&ipt_ulog_reg); sock_release(nflognl->sk_socket); diff --git a/net/ipv4/netfilter/ipt_connbytes.c b/net/ipv4/netfilter/ipt_connbytes.c new file mode 100644 index 000000000000..df4a42c6da22 --- /dev/null +++ b/net/ipv4/netfilter/ipt_connbytes.c @@ -0,0 +1,162 @@ +/* Kernel module to match connection tracking byte counter. + * GPL (C) 2002 Martin Devera (devik@cdi.cz). + * + * 2004-07-20 Harald Welte <laforge@netfilter.org> + * - reimplemented to use per-connection accounting counters + * - add functionality to match number of packets + * - add functionality to match average packet size + * - add support to match directions seperately + * + */ +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/netfilter_ipv4/ip_conntrack.h> +#include <linux/netfilter_ipv4/ip_tables.h> +#include <linux/netfilter_ipv4/ipt_connbytes.h> + +#include <asm/div64.h> +#include <asm/bitops.h> + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_DESCRIPTION("iptables match for matching number of pkts/bytes per connection"); + +/* 64bit divisor, dividend and result. dynamic precision */ +static u_int64_t div64_64(u_int64_t dividend, u_int64_t divisor) +{ + u_int32_t d = divisor; + + if (divisor > 0xffffffffULL) { + unsigned int shift = fls(divisor >> 32); + + d = divisor >> shift; + dividend >>= shift; + } + + do_div(dividend, d); + return dividend; +} + +static int +match(const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const void *matchinfo, + int offset, + int *hotdrop) +{ + const struct ipt_connbytes_info *sinfo = matchinfo; + enum ip_conntrack_info ctinfo; + struct ip_conntrack *ct; + u_int64_t what = 0; /* initialize to make gcc happy */ + + if (!(ct = ip_conntrack_get((struct sk_buff *)skb, &ctinfo))) + return 0; /* no match */ + + switch (sinfo->what) { + case IPT_CONNBYTES_PKTS: + switch (sinfo->direction) { + case IPT_CONNBYTES_DIR_ORIGINAL: + what = ct->counters[IP_CT_DIR_ORIGINAL].packets; + break; + case IPT_CONNBYTES_DIR_REPLY: + what = ct->counters[IP_CT_DIR_REPLY].packets; + break; + case IPT_CONNBYTES_DIR_BOTH: + what = ct->counters[IP_CT_DIR_ORIGINAL].packets; + what += ct->counters[IP_CT_DIR_REPLY].packets; + break; + } + break; + case IPT_CONNBYTES_BYTES: + switch (sinfo->direction) { + case IPT_CONNBYTES_DIR_ORIGINAL: + what = ct->counters[IP_CT_DIR_ORIGINAL].bytes; + break; + case IPT_CONNBYTES_DIR_REPLY: + what = ct->counters[IP_CT_DIR_REPLY].bytes; + break; + case IPT_CONNBYTES_DIR_BOTH: + what = ct->counters[IP_CT_DIR_ORIGINAL].bytes; + what += ct->counters[IP_CT_DIR_REPLY].bytes; + break; + } + break; + case IPT_CONNBYTES_AVGPKT: + switch (sinfo->direction) { + case IPT_CONNBYTES_DIR_ORIGINAL: + what = div64_64(ct->counters[IP_CT_DIR_ORIGINAL].bytes, + ct->counters[IP_CT_DIR_ORIGINAL].packets); + break; + case IPT_CONNBYTES_DIR_REPLY: + what = div64_64(ct->counters[IP_CT_DIR_REPLY].bytes, + ct->counters[IP_CT_DIR_REPLY].packets); + break; + case IPT_CONNBYTES_DIR_BOTH: + { + u_int64_t bytes; + u_int64_t pkts; + bytes = ct->counters[IP_CT_DIR_ORIGINAL].bytes + + ct->counters[IP_CT_DIR_REPLY].bytes; + pkts = ct->counters[IP_CT_DIR_ORIGINAL].packets+ + ct->counters[IP_CT_DIR_REPLY].packets; + + /* FIXME_THEORETICAL: what to do if sum + * overflows ? */ + + what = div64_64(bytes, pkts); + } + break; + } + break; + } + + if (sinfo->count.to) + return (what <= sinfo->count.to && what >= sinfo->count.from); + else + return (what >= sinfo->count.from); +} + +static int check(const char *tablename, + const struct ipt_ip *ip, + void *matchinfo, + unsigned int matchsize, + unsigned int hook_mask) +{ + const struct ipt_connbytes_info *sinfo = matchinfo; + + if (matchsize != IPT_ALIGN(sizeof(struct ipt_connbytes_info))) + return 0; + + if (sinfo->what != IPT_CONNBYTES_PKTS && + sinfo->what != IPT_CONNBYTES_BYTES && + sinfo->what != IPT_CONNBYTES_AVGPKT) + return 0; + + if (sinfo->direction != IPT_CONNBYTES_DIR_ORIGINAL && + sinfo->direction != IPT_CONNBYTES_DIR_REPLY && + sinfo->direction != IPT_CONNBYTES_DIR_BOTH) + return 0; + + return 1; +} + +static struct ipt_match state_match = { + .name = "connbytes", + .match = &match, + .checkentry = &check, + .me = THIS_MODULE +}; + +static int __init init(void) +{ + return ipt_register_match(&state_match); +} + +static void __exit fini(void) +{ + ipt_unregister_match(&state_match); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/netfilter/ipt_connmark.c b/net/ipv4/netfilter/ipt_connmark.c index 2706f96cea55..bf8de47ce004 100644 --- a/net/ipv4/netfilter/ipt_connmark.c +++ b/net/ipv4/netfilter/ipt_connmark.c @@ -54,9 +54,16 @@ checkentry(const char *tablename, unsigned int matchsize, unsigned int hook_mask) { + struct ipt_connmark_info *cm = + (struct ipt_connmark_info *)matchinfo; if (matchsize != IPT_ALIGN(sizeof(struct ipt_connmark_info))) return 0; + if (cm->mark > 0xffffffff || cm->mask > 0xffffffff) { + printk(KERN_WARNING "connmark: only support 32bit mark\n"); + return 0; + } + return 1; } diff --git a/net/ipv4/netfilter/ipt_dccp.c b/net/ipv4/netfilter/ipt_dccp.c new file mode 100644 index 000000000000..ad3278bba6c1 --- /dev/null +++ b/net/ipv4/netfilter/ipt_dccp.c @@ -0,0 +1,176 @@ +/* + * iptables module for DCCP protocol header matching + * + * (C) 2005 by Harald Welte <laforge@netfilter.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/spinlock.h> +#include <net/ip.h> +#include <linux/dccp.h> + +#include <linux/netfilter_ipv4/ip_tables.h> +#include <linux/netfilter_ipv4/ipt_dccp.h> + +#define DCCHECK(cond, option, flag, invflag) (!((flag) & (option)) \ + || (!!((invflag) & (option)) ^ (cond))) + +static unsigned char *dccp_optbuf; +static DEFINE_SPINLOCK(dccp_buflock); + +static inline int +dccp_find_option(u_int8_t option, + const struct sk_buff *skb, + const struct dccp_hdr *dh, + int *hotdrop) +{ + /* tcp.doff is only 4 bits, ie. max 15 * 4 bytes */ + unsigned char *op; + unsigned int optoff = __dccp_hdr_len(dh); + unsigned int optlen = dh->dccph_doff*4 - __dccp_hdr_len(dh); + unsigned int i; + + if (dh->dccph_doff * 4 < __dccp_hdr_len(dh)) { + *hotdrop = 1; + return 0; + } + + if (!optlen) + return 0; + + spin_lock_bh(&dccp_buflock); + op = skb_header_pointer(skb, + skb->nh.iph->ihl*4 + optoff, + optlen, dccp_optbuf); + if (op == NULL) { + /* If we don't have the whole header, drop packet. */ + spin_unlock_bh(&dccp_buflock); + *hotdrop = 1; + return 0; + } + + for (i = 0; i < optlen; ) { + if (op[i] == option) { + spin_unlock_bh(&dccp_buflock); + return 1; + } + + if (op[i] < 2) + i++; + else + i += op[i+1]?:1; + } + + spin_unlock_bh(&dccp_buflock); + return 0; +} + + +static inline int +match_types(const struct dccp_hdr *dh, u_int16_t typemask) +{ + return (typemask & (1 << dh->dccph_type)); +} + +static inline int +match_option(u_int8_t option, const struct sk_buff *skb, + const struct dccp_hdr *dh, int *hotdrop) +{ + return dccp_find_option(option, skb, dh, hotdrop); +} + +static int +match(const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const void *matchinfo, + int offset, + int *hotdrop) +{ + const struct ipt_dccp_info *info = + (const struct ipt_dccp_info *)matchinfo; + struct dccp_hdr _dh, *dh; + + if (offset) + return 0; + + dh = skb_header_pointer(skb, skb->nh.iph->ihl*4, sizeof(_dh), &_dh); + if (dh == NULL) { + *hotdrop = 1; + return 0; + } + + return DCCHECK(((ntohs(dh->dccph_sport) >= info->spts[0]) + && (ntohs(dh->dccph_sport) <= info->spts[1])), + IPT_DCCP_SRC_PORTS, info->flags, info->invflags) + && DCCHECK(((ntohs(dh->dccph_dport) >= info->dpts[0]) + && (ntohs(dh->dccph_dport) <= info->dpts[1])), + IPT_DCCP_DEST_PORTS, info->flags, info->invflags) + && DCCHECK(match_types(dh, info->typemask), + IPT_DCCP_TYPE, info->flags, info->invflags) + && DCCHECK(match_option(info->option, skb, dh, hotdrop), + IPT_DCCP_OPTION, info->flags, info->invflags); +} + +static int +checkentry(const char *tablename, + const struct ipt_ip *ip, + void *matchinfo, + unsigned int matchsize, + unsigned int hook_mask) +{ + const struct ipt_dccp_info *info; + + info = (const struct ipt_dccp_info *)matchinfo; + + return ip->proto == IPPROTO_DCCP + && !(ip->invflags & IPT_INV_PROTO) + && matchsize == IPT_ALIGN(sizeof(struct ipt_dccp_info)) + && !(info->flags & ~IPT_DCCP_VALID_FLAGS) + && !(info->invflags & ~IPT_DCCP_VALID_FLAGS) + && !(info->invflags & ~info->flags); +} + +static struct ipt_match dccp_match = +{ + .name = "dccp", + .match = &match, + .checkentry = &checkentry, + .me = THIS_MODULE, +}; + +static int __init init(void) +{ + int ret; + + /* doff is 8 bits, so the maximum option size is (4*256). Don't put + * this in BSS since DaveM is worried about locked TLB's for kernel + * BSS. */ + dccp_optbuf = kmalloc(256 * 4, GFP_KERNEL); + if (!dccp_optbuf) + return -ENOMEM; + ret = ipt_register_match(&dccp_match); + if (ret) + kfree(dccp_optbuf); + + return ret; +} + +static void __exit fini(void) +{ + ipt_unregister_match(&dccp_match); + kfree(dccp_optbuf); +} + +module_init(init); +module_exit(fini); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_DESCRIPTION("Match for DCCP protocol packets"); + diff --git a/net/ipv4/netfilter/ipt_hashlimit.c b/net/ipv4/netfilter/ipt_hashlimit.c index 564b49bfebcf..2dd1cccbdab9 100644 --- a/net/ipv4/netfilter/ipt_hashlimit.c +++ b/net/ipv4/netfilter/ipt_hashlimit.c @@ -94,7 +94,7 @@ struct ipt_hashlimit_htable { static DEFINE_SPINLOCK(hashlimit_lock); /* protects htables list */ static DECLARE_MUTEX(hlimit_mutex); /* additional checkentry protection */ static HLIST_HEAD(hashlimit_htables); -static kmem_cache_t *hashlimit_cachep; +static kmem_cache_t *hashlimit_cachep __read_mostly; static inline int dst_cmp(const struct dsthash_ent *ent, struct dsthash_dst *b) { diff --git a/net/ipv4/netfilter/ipt_mark.c b/net/ipv4/netfilter/ipt_mark.c index 8955728127b9..00bef6cdd3f8 100644 --- a/net/ipv4/netfilter/ipt_mark.c +++ b/net/ipv4/netfilter/ipt_mark.c @@ -37,9 +37,16 @@ checkentry(const char *tablename, unsigned int matchsize, unsigned int hook_mask) { + struct ipt_mark_info *minfo = (struct ipt_mark_info *) matchinfo; + if (matchsize != IPT_ALIGN(sizeof(struct ipt_mark_info))) return 0; + if (minfo->mark > 0xffffffff || minfo->mask > 0xffffffff) { + printk(KERN_WARNING "mark: only supports 32bit mark\n"); + return 0; + } + return 1; } diff --git a/net/ipv4/netfilter/ipt_owner.c b/net/ipv4/netfilter/ipt_owner.c index 3b9065e06381..c1889f88262b 100644 --- a/net/ipv4/netfilter/ipt_owner.c +++ b/net/ipv4/netfilter/ipt_owner.c @@ -21,106 +21,6 @@ MODULE_AUTHOR("Marc Boucher <marc@mbsi.ca>"); MODULE_DESCRIPTION("iptables owner match"); static int -match_comm(const struct sk_buff *skb, const char *comm) -{ - struct task_struct *g, *p; - struct files_struct *files; - int i; - - read_lock(&tasklist_lock); - do_each_thread(g, p) { - if(strncmp(p->comm, comm, sizeof(p->comm))) - continue; - - task_lock(p); - files = p->files; - if(files) { - spin_lock(&files->file_lock); - for (i=0; i < files->max_fds; i++) { - if (fcheck_files(files, i) == - skb->sk->sk_socket->file) { - spin_unlock(&files->file_lock); - task_unlock(p); - read_unlock(&tasklist_lock); - return 1; - } - } - spin_unlock(&files->file_lock); - } - task_unlock(p); - } while_each_thread(g, p); - read_unlock(&tasklist_lock); - return 0; -} - -static int -match_pid(const struct sk_buff *skb, pid_t pid) -{ - struct task_struct *p; - struct files_struct *files; - int i; - - read_lock(&tasklist_lock); - p = find_task_by_pid(pid); - if (!p) - goto out; - task_lock(p); - files = p->files; - if(files) { - spin_lock(&files->file_lock); - for (i=0; i < files->max_fds; i++) { - if (fcheck_files(files, i) == - skb->sk->sk_socket->file) { - spin_unlock(&files->file_lock); - task_unlock(p); - read_unlock(&tasklist_lock); - return 1; - } - } - spin_unlock(&files->file_lock); - } - task_unlock(p); -out: - read_unlock(&tasklist_lock); - return 0; -} - -static int -match_sid(const struct sk_buff *skb, pid_t sid) -{ - struct task_struct *g, *p; - struct file *file = skb->sk->sk_socket->file; - int i, found=0; - - read_lock(&tasklist_lock); - do_each_thread(g, p) { - struct files_struct *files; - if (p->signal->session != sid) - continue; - - task_lock(p); - files = p->files; - if (files) { - spin_lock(&files->file_lock); - for (i=0; i < files->max_fds; i++) { - if (fcheck_files(files, i) == file) { - found = 1; - break; - } - } - spin_unlock(&files->file_lock); - } - task_unlock(p); - if (found) - goto out; - } while_each_thread(g, p); -out: - read_unlock(&tasklist_lock); - - return found; -} - -static int match(const struct sk_buff *skb, const struct net_device *in, const struct net_device *out, @@ -145,24 +45,6 @@ match(const struct sk_buff *skb, return 0; } - if(info->match & IPT_OWNER_PID) { - if (!match_pid(skb, info->pid) ^ - !!(info->invert & IPT_OWNER_PID)) - return 0; - } - - if(info->match & IPT_OWNER_SID) { - if (!match_sid(skb, info->sid) ^ - !!(info->invert & IPT_OWNER_SID)) - return 0; - } - - if(info->match & IPT_OWNER_COMM) { - if (!match_comm(skb, info->comm) ^ - !!(info->invert & IPT_OWNER_COMM)) - return 0; - } - return 1; } @@ -173,6 +55,8 @@ checkentry(const char *tablename, unsigned int matchsize, unsigned int hook_mask) { + const struct ipt_owner_info *info = matchinfo; + if (hook_mask & ~((1 << NF_IP_LOCAL_OUT) | (1 << NF_IP_POST_ROUTING))) { printk("ipt_owner: only valid for LOCAL_OUT or POST_ROUTING.\n"); @@ -184,15 +68,13 @@ checkentry(const char *tablename, IPT_ALIGN(sizeof(struct ipt_owner_info))); return 0; } -#ifdef CONFIG_SMP - /* files->file_lock can not be used in a BH */ - if (((struct ipt_owner_info *)matchinfo)->match - & (IPT_OWNER_PID|IPT_OWNER_SID|IPT_OWNER_COMM)) { - printk("ipt_owner: pid, sid and command matching is broken " - "on SMP.\n"); + + if (info->match & (IPT_OWNER_PID|IPT_OWNER_SID|IPT_OWNER_COMM)) { + printk("ipt_owner: pid, sid and command matching " + "not supported anymore\n"); return 0; } -#endif + return 1; } diff --git a/net/ipv4/netfilter/ipt_string.c b/net/ipv4/netfilter/ipt_string.c new file mode 100644 index 000000000000..b5def204d798 --- /dev/null +++ b/net/ipv4/netfilter/ipt_string.c @@ -0,0 +1,91 @@ +/* String matching match for iptables + * + * (C) 2005 Pablo Neira Ayuso <pablo@eurodev.net> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/init.h> +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/skbuff.h> +#include <linux/netfilter_ipv4/ip_tables.h> +#include <linux/netfilter_ipv4/ipt_string.h> +#include <linux/textsearch.h> + +MODULE_AUTHOR("Pablo Neira Ayuso <pablo@eurodev.net>"); +MODULE_DESCRIPTION("IP tables string match module"); +MODULE_LICENSE("GPL"); + +static int match(const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const void *matchinfo, + int offset, + int *hotdrop) +{ + struct ts_state state; + struct ipt_string_info *conf = (struct ipt_string_info *) matchinfo; + + memset(&state, 0, sizeof(struct ts_state)); + + return (skb_find_text((struct sk_buff *)skb, conf->from_offset, + conf->to_offset, conf->config, &state) + != UINT_MAX) && !conf->invert; +} + +#define STRING_TEXT_PRIV(m) ((struct ipt_string_info *) m) + +static int checkentry(const char *tablename, + const struct ipt_ip *ip, + void *matchinfo, + unsigned int matchsize, + unsigned int hook_mask) +{ + struct ipt_string_info *conf = matchinfo; + struct ts_config *ts_conf; + + if (matchsize != IPT_ALIGN(sizeof(struct ipt_string_info))) + return 0; + + /* Damn, can't handle this case properly with iptables... */ + if (conf->from_offset > conf->to_offset) + return 0; + + ts_conf = textsearch_prepare(conf->algo, conf->pattern, conf->patlen, + GFP_KERNEL, TS_AUTOLOAD); + if (IS_ERR(ts_conf)) + return 0; + + conf->config = ts_conf; + + return 1; +} + +static void destroy(void *matchinfo, unsigned int matchsize) +{ + textsearch_destroy(STRING_TEXT_PRIV(matchinfo)->config); +} + +static struct ipt_match string_match = { + .name = "string", + .match = match, + .checkentry = checkentry, + .destroy = destroy, + .me = THIS_MODULE +}; + +static int __init init(void) +{ + return ipt_register_match(&string_match); +} + +static void __exit fini(void) +{ + ipt_unregister_match(&string_match); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c index 912bbcc7f415..f7943ba1f43c 100644 --- a/net/ipv4/proc.c +++ b/net/ipv4/proc.c @@ -59,13 +59,10 @@ static int fold_prot_inuse(struct proto *proto) */ static int sockstat_seq_show(struct seq_file *seq, void *v) { - /* From net/socket.c */ - extern void socket_seq_show(struct seq_file *seq); - socket_seq_show(seq); seq_printf(seq, "TCP: inuse %d orphan %d tw %d alloc %d mem %d\n", fold_prot_inuse(&tcp_prot), atomic_read(&tcp_orphan_count), - tcp_tw_count, atomic_read(&tcp_sockets_allocated), + tcp_death_row.tw_count, atomic_read(&tcp_sockets_allocated), atomic_read(&tcp_memory_allocated)); seq_printf(seq, "UDP: inuse %d\n", fold_prot_inuse(&udp_prot)); seq_printf(seq, "RAW: inuse %d\n", fold_prot_inuse(&raw_prot)); diff --git a/net/ipv4/protocol.c b/net/ipv4/protocol.c index 0db405a869f2..291831e792af 100644 --- a/net/ipv4/protocol.c +++ b/net/ipv4/protocol.c @@ -40,7 +40,6 @@ #include <linux/timer.h> #include <net/ip.h> #include <net/protocol.h> -#include <net/tcp.h> #include <linux/skbuff.h> #include <net/sock.h> #include <net/icmp.h> diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index d1835b1bc8c4..304bb0a1d4f0 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -59,7 +59,6 @@ #include <linux/netdevice.h> #include <linux/in_route.h> #include <linux/route.h> -#include <linux/tcp.h> #include <linux/skbuff.h> #include <net/dst.h> #include <net/sock.h> @@ -71,6 +70,7 @@ #include <net/udp.h> #include <net/raw.h> #include <net/snmp.h> +#include <net/tcp_states.h> #include <net/inet_common.h> #include <net/checksum.h> #include <net/xfrm.h> @@ -150,10 +150,11 @@ static __inline__ int icmp_filter(struct sock *sk, struct sk_buff *skb) * RFC 1122: SHOULD pass TOS value up to the transport layer. * -> It does. And not only TOS, but all IP header. */ -void raw_v4_input(struct sk_buff *skb, struct iphdr *iph, int hash) +int raw_v4_input(struct sk_buff *skb, struct iphdr *iph, int hash) { struct sock *sk; struct hlist_head *head; + int delivered = 0; read_lock(&raw_v4_lock); head = &raw_v4_htable[hash]; @@ -164,6 +165,7 @@ void raw_v4_input(struct sk_buff *skb, struct iphdr *iph, int hash) skb->dev->ifindex); while (sk) { + delivered = 1; if (iph->protocol != IPPROTO_ICMP || !icmp_filter(sk, skb)) { struct sk_buff *clone = skb_clone(skb, GFP_ATOMIC); @@ -177,6 +179,7 @@ void raw_v4_input(struct sk_buff *skb, struct iphdr *iph, int hash) } out: read_unlock(&raw_v4_lock); + return delivered; } void raw_err (struct sock *sk, struct sk_buff *skb, u32 info) diff --git a/net/ipv4/route.c b/net/ipv4/route.c index d675ff80b04d..8c0b14e3beec 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -240,7 +240,9 @@ static unsigned rt_hash_mask; static int rt_hash_log; static unsigned int rt_hash_rnd; -struct rt_cache_stat *rt_cache_stat; +static struct rt_cache_stat *rt_cache_stat; +#define RT_CACHE_STAT_INC(field) \ + (per_cpu_ptr(rt_cache_stat, raw_smp_processor_id())->field++) static int rt_intern_hash(unsigned hash, struct rtable *rth, struct rtable **res); @@ -2600,6 +2602,8 @@ int __ip_route_output_key(struct rtable **rp, const struct flowi *flp) return ip_route_output_slow(rp, flp); } +EXPORT_SYMBOL_GPL(__ip_route_output_key); + int ip_route_output_flow(struct rtable **rp, struct flowi *flp, struct sock *sk, int flags) { int err; @@ -2618,6 +2622,8 @@ int ip_route_output_flow(struct rtable **rp, struct flowi *flp, struct sock *sk, return 0; } +EXPORT_SYMBOL_GPL(ip_route_output_flow); + int ip_route_output_key(struct rtable **rp, struct flowi *flp) { return ip_route_output_flow(rp, flp, NULL, 0); diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c index 72d014442185..a34e60ea48a1 100644 --- a/net/ipv4/syncookies.c +++ b/net/ipv4/syncookies.c @@ -169,8 +169,6 @@ static inline int cookie_check(struct sk_buff *skb, __u32 cookie) return mssind < NUM_MSS ? msstab[mssind] + 1 : 0; } -extern struct request_sock_ops tcp_request_sock_ops; - static inline struct sock *get_cookie_sock(struct sock *sk, struct sk_buff *skb, struct request_sock *req, struct dst_entry *dst) @@ -180,7 +178,7 @@ static inline struct sock *get_cookie_sock(struct sock *sk, struct sk_buff *skb, child = tp->af_specific->syn_recv_sock(sk, skb, req, dst); if (child) - tcp_acceptq_queue(sk, req, child); + inet_csk_reqsk_queue_add(sk, req, child); else reqsk_free(req); diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index e32894532416..652685623519 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -11,7 +11,9 @@ #include <linux/module.h> #include <linux/sysctl.h> #include <linux/config.h> +#include <linux/igmp.h> #include <net/snmp.h> +#include <net/icmp.h> #include <net/ip.h> #include <net/route.h> #include <net/tcp.h> @@ -19,36 +21,6 @@ /* From af_inet.c */ extern int sysctl_ip_nonlocal_bind; -/* From icmp.c */ -extern int sysctl_icmp_echo_ignore_all; -extern int sysctl_icmp_echo_ignore_broadcasts; -extern int sysctl_icmp_ignore_bogus_error_responses; -extern int sysctl_icmp_errors_use_inbound_ifaddr; - -/* From ip_fragment.c */ -extern int sysctl_ipfrag_low_thresh; -extern int sysctl_ipfrag_high_thresh; -extern int sysctl_ipfrag_time; -extern int sysctl_ipfrag_secret_interval; - -/* From ip_output.c */ -extern int sysctl_ip_dynaddr; - -/* From icmp.c */ -extern int sysctl_icmp_ratelimit; -extern int sysctl_icmp_ratemask; - -/* From igmp.c */ -extern int sysctl_igmp_max_memberships; -extern int sysctl_igmp_max_msf; - -/* From inetpeer.c */ -extern int inet_peer_threshold; -extern int inet_peer_minttl; -extern int inet_peer_maxttl; -extern int inet_peer_gc_mintime; -extern int inet_peer_gc_maxtime; - #ifdef CONFIG_SYSCTL static int tcp_retr1_max = 255; static int ip_local_port_range_min[] = { 1, 1 }; @@ -57,8 +29,6 @@ static int ip_local_port_range_max[] = { 65535, 65535 }; struct ipv4_config ipv4_config; -extern ctl_table ipv4_route_table[]; - #ifdef CONFIG_SYSCTL static @@ -136,10 +106,11 @@ static int proc_tcp_congestion_control(ctl_table *ctl, int write, struct file * return ret; } -int sysctl_tcp_congestion_control(ctl_table *table, int __user *name, int nlen, - void __user *oldval, size_t __user *oldlenp, - void __user *newval, size_t newlen, - void **context) +static int sysctl_tcp_congestion_control(ctl_table *table, int __user *name, + int nlen, void __user *oldval, + size_t __user *oldlenp, + void __user *newval, size_t newlen, + void **context) { char val[TCP_CA_NAME_MAX]; ctl_table tbl = { @@ -259,7 +230,7 @@ ctl_table ipv4_table[] = { { .ctl_name = NET_TCP_MAX_TW_BUCKETS, .procname = "tcp_max_tw_buckets", - .data = &sysctl_tcp_max_tw_buckets, + .data = &tcp_death_row.sysctl_max_tw_buckets, .maxlen = sizeof(int), .mode = 0644, .proc_handler = &proc_dointvec @@ -363,7 +334,7 @@ ctl_table ipv4_table[] = { { .ctl_name = NET_TCP_TW_RECYCLE, .procname = "tcp_tw_recycle", - .data = &sysctl_tcp_tw_recycle, + .data = &tcp_death_row.sysctl_tw_recycle, .maxlen = sizeof(int), .mode = 0644, .proc_handler = &proc_dointvec diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 69b1fcf70077..02fdda68718d 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -269,13 +269,12 @@ int sysctl_tcp_fin_timeout = TCP_FIN_TIMEOUT; -DEFINE_SNMP_STAT(struct tcp_mib, tcp_statistics); - -kmem_cache_t *tcp_bucket_cachep; -kmem_cache_t *tcp_timewait_cachep; +DEFINE_SNMP_STAT(struct tcp_mib, tcp_statistics) __read_mostly; atomic_t tcp_orphan_count = ATOMIC_INIT(0); +EXPORT_SYMBOL_GPL(tcp_orphan_count); + int sysctl_tcp_mem[3]; int sysctl_tcp_wmem[3] = { 4 * 1024, 16 * 1024, 128 * 1024 }; int sysctl_tcp_rmem[3] = { 4 * 1024, 87380, 87380 * 2 }; @@ -311,15 +310,6 @@ void tcp_enter_memory_pressure(void) EXPORT_SYMBOL(tcp_enter_memory_pressure); /* - * LISTEN is a special case for poll.. - */ -static __inline__ unsigned int tcp_listen_poll(struct sock *sk, - poll_table *wait) -{ - return !reqsk_queue_empty(&tcp_sk(sk)->accept_queue) ? (POLLIN | POLLRDNORM) : 0; -} - -/* * Wait for a TCP event. * * Note that we don't need to lock the socket, as the upper poll layers @@ -334,7 +324,7 @@ unsigned int tcp_poll(struct file *file, struct socket *sock, poll_table *wait) poll_wait(file, sk->sk_sleep, wait); if (sk->sk_state == TCP_LISTEN) - return tcp_listen_poll(sk, wait); + return inet_csk_listen_poll(sk); /* Socket is not locked. We are protected from async events by poll logic and correct handling of state changes @@ -457,109 +447,6 @@ int tcp_ioctl(struct sock *sk, int cmd, unsigned long arg) return put_user(answ, (int __user *)arg); } - -int tcp_listen_start(struct sock *sk) -{ - struct inet_sock *inet = inet_sk(sk); - struct tcp_sock *tp = tcp_sk(sk); - int rc = reqsk_queue_alloc(&tp->accept_queue, TCP_SYNQ_HSIZE); - - if (rc != 0) - return rc; - - sk->sk_max_ack_backlog = 0; - sk->sk_ack_backlog = 0; - tcp_delack_init(tp); - - /* There is race window here: we announce ourselves listening, - * but this transition is still not validated by get_port(). - * It is OK, because this socket enters to hash table only - * after validation is complete. - */ - sk->sk_state = TCP_LISTEN; - if (!sk->sk_prot->get_port(sk, inet->num)) { - inet->sport = htons(inet->num); - - sk_dst_reset(sk); - sk->sk_prot->hash(sk); - - return 0; - } - - sk->sk_state = TCP_CLOSE; - reqsk_queue_destroy(&tp->accept_queue); - return -EADDRINUSE; -} - -/* - * This routine closes sockets which have been at least partially - * opened, but not yet accepted. - */ - -static void tcp_listen_stop (struct sock *sk) -{ - struct tcp_sock *tp = tcp_sk(sk); - struct listen_sock *lopt; - struct request_sock *acc_req; - struct request_sock *req; - int i; - - tcp_delete_keepalive_timer(sk); - - /* make all the listen_opt local to us */ - lopt = reqsk_queue_yank_listen_sk(&tp->accept_queue); - acc_req = reqsk_queue_yank_acceptq(&tp->accept_queue); - - if (lopt->qlen) { - for (i = 0; i < TCP_SYNQ_HSIZE; i++) { - while ((req = lopt->syn_table[i]) != NULL) { - lopt->syn_table[i] = req->dl_next; - lopt->qlen--; - reqsk_free(req); - - /* Following specs, it would be better either to send FIN - * (and enter FIN-WAIT-1, it is normal close) - * or to send active reset (abort). - * Certainly, it is pretty dangerous while synflood, but it is - * bad justification for our negligence 8) - * To be honest, we are not able to make either - * of the variants now. --ANK - */ - } - } - } - BUG_TRAP(!lopt->qlen); - - kfree(lopt); - - while ((req = acc_req) != NULL) { - struct sock *child = req->sk; - - acc_req = req->dl_next; - - local_bh_disable(); - bh_lock_sock(child); - BUG_TRAP(!sock_owned_by_user(child)); - sock_hold(child); - - tcp_disconnect(child, O_NONBLOCK); - - sock_orphan(child); - - atomic_inc(&tcp_orphan_count); - - tcp_destroy_sock(child); - - bh_unlock_sock(child); - local_bh_enable(); - sock_put(child); - - sk_acceptq_removed(sk); - __reqsk_free(req); - } - BUG_TRAP(!sk->sk_ack_backlog); -} - static inline void tcp_mark_push(struct tcp_sock *tp, struct sk_buff *skb) { TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH; @@ -975,7 +862,7 @@ do_fault: if (!skb->len) { if (sk->sk_send_head == skb) sk->sk_send_head = NULL; - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &sk->sk_write_queue); sk_stream_free_skb(sk, skb); } @@ -1057,20 +944,21 @@ static void cleanup_rbuf(struct sock *sk, int copied) BUG_TRAP(!skb || before(tp->copied_seq, TCP_SKB_CB(skb)->end_seq)); #endif - if (tcp_ack_scheduled(tp)) { + if (inet_csk_ack_scheduled(sk)) { + const struct inet_connection_sock *icsk = inet_csk(sk); /* Delayed ACKs frequently hit locked sockets during bulk * receive. */ - if (tp->ack.blocked || + if (icsk->icsk_ack.blocked || /* Once-per-two-segments ACK was not sent by tcp_input.c */ - tp->rcv_nxt - tp->rcv_wup > tp->ack.rcv_mss || + tp->rcv_nxt - tp->rcv_wup > icsk->icsk_ack.rcv_mss || /* * If this read emptied read buffer, we send ACK, if * connection is not bidirectional, user drained * receive buffer and there was a small segment * in queue. */ - (copied > 0 && (tp->ack.pending & TCP_ACK_PUSHED) && - !tp->ack.pingpong && !atomic_read(&sk->sk_rmem_alloc))) + (copied > 0 && (icsk->icsk_ack.pending & ICSK_ACK_PUSHED) && + !icsk->icsk_ack.pingpong && !atomic_read(&sk->sk_rmem_alloc))) time_to_ack = 1; } @@ -1572,40 +1460,6 @@ void tcp_shutdown(struct sock *sk, int how) } } -/* - * At this point, there should be no process reference to this - * socket, and thus no user references at all. Therefore we - * can assume the socket waitqueue is inactive and nobody will - * try to jump onto it. - */ -void tcp_destroy_sock(struct sock *sk) -{ - BUG_TRAP(sk->sk_state == TCP_CLOSE); - BUG_TRAP(sock_flag(sk, SOCK_DEAD)); - - /* It cannot be in hash table! */ - BUG_TRAP(sk_unhashed(sk)); - - /* If it has not 0 inet_sk(sk)->num, it must be bound */ - BUG_TRAP(!inet_sk(sk)->num || tcp_sk(sk)->bind_hash); - - sk->sk_prot->destroy(sk); - - sk_stream_kill_queues(sk); - - xfrm_sk_free_policy(sk); - -#ifdef INET_REFCNT_DEBUG - if (atomic_read(&sk->sk_refcnt) != 1) { - printk(KERN_DEBUG "Destruction TCP %p delayed, c=%d\n", - sk, atomic_read(&sk->sk_refcnt)); - } -#endif - - atomic_dec(&tcp_orphan_count); - sock_put(sk); -} - void tcp_close(struct sock *sk, long timeout) { struct sk_buff *skb; @@ -1618,7 +1472,7 @@ void tcp_close(struct sock *sk, long timeout) tcp_set_state(sk, TCP_CLOSE); /* Special case. */ - tcp_listen_stop(sk); + inet_csk_listen_stop(sk); goto adjudge_to_death; } @@ -1721,12 +1575,12 @@ adjudge_to_death: tcp_send_active_reset(sk, GFP_ATOMIC); NET_INC_STATS_BH(LINUX_MIB_TCPABORTONLINGER); } else { - int tmo = tcp_fin_time(tp); + const int tmo = tcp_fin_time(sk); if (tmo > TCP_TIMEWAIT_LEN) { - tcp_reset_keepalive_timer(sk, tcp_fin_time(tp)); + inet_csk_reset_keepalive_timer(sk, tcp_fin_time(sk)); } else { - atomic_inc(&tcp_orphan_count); + atomic_inc(sk->sk_prot->orphan_count); tcp_time_wait(sk, TCP_FIN_WAIT2, tmo); goto out; } @@ -1734,7 +1588,7 @@ adjudge_to_death: } if (sk->sk_state != TCP_CLOSE) { sk_stream_mem_reclaim(sk); - if (atomic_read(&tcp_orphan_count) > sysctl_tcp_max_orphans || + if (atomic_read(sk->sk_prot->orphan_count) > sysctl_tcp_max_orphans || (sk->sk_wmem_queued > SOCK_MIN_SNDBUF && atomic_read(&tcp_memory_allocated) > sysctl_tcp_mem[2])) { if (net_ratelimit()) @@ -1745,10 +1599,10 @@ adjudge_to_death: NET_INC_STATS_BH(LINUX_MIB_TCPABORTONMEMORY); } } - atomic_inc(&tcp_orphan_count); + atomic_inc(sk->sk_prot->orphan_count); if (sk->sk_state == TCP_CLOSE) - tcp_destroy_sock(sk); + inet_csk_destroy_sock(sk); /* Otherwise, socket is reprieved until protocol close. */ out: @@ -1769,6 +1623,7 @@ static inline int tcp_need_reset(int state) int tcp_disconnect(struct sock *sk, int flags) { struct inet_sock *inet = inet_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); int err = 0; int old_state = sk->sk_state; @@ -1778,7 +1633,7 @@ int tcp_disconnect(struct sock *sk, int flags) /* ABORT function of RFC793 */ if (old_state == TCP_LISTEN) { - tcp_listen_stop(sk); + inet_csk_listen_stop(sk); } else if (tcp_need_reset(old_state) || (tp->snd_nxt != tp->write_seq && (1 << old_state) & (TCPF_CLOSING | TCPF_LAST_ACK))) { @@ -1805,125 +1660,34 @@ int tcp_disconnect(struct sock *sk, int flags) tp->srtt = 0; if ((tp->write_seq += tp->max_window + 2) == 0) tp->write_seq = 1; - tp->backoff = 0; + icsk->icsk_backoff = 0; tp->snd_cwnd = 2; - tp->probes_out = 0; + icsk->icsk_probes_out = 0; tp->packets_out = 0; tp->snd_ssthresh = 0x7fffffff; tp->snd_cwnd_cnt = 0; - tcp_set_ca_state(tp, TCP_CA_Open); + tcp_set_ca_state(sk, TCP_CA_Open); tcp_clear_retrans(tp); - tcp_delack_init(tp); + inet_csk_delack_init(sk); sk->sk_send_head = NULL; tp->rx_opt.saw_tstamp = 0; tcp_sack_reset(&tp->rx_opt); __sk_dst_reset(sk); - BUG_TRAP(!inet->num || tp->bind_hash); + BUG_TRAP(!inet->num || icsk->icsk_bind_hash); sk->sk_error_report(sk); return err; } /* - * Wait for an incoming connection, avoid race - * conditions. This must be called with the socket locked. - */ -static int wait_for_connect(struct sock *sk, long timeo) -{ - struct tcp_sock *tp = tcp_sk(sk); - DEFINE_WAIT(wait); - int err; - - /* - * True wake-one mechanism for incoming connections: only - * one process gets woken up, not the 'whole herd'. - * Since we do not 'race & poll' for established sockets - * anymore, the common case will execute the loop only once. - * - * Subtle issue: "add_wait_queue_exclusive()" will be added - * after any current non-exclusive waiters, and we know that - * it will always _stay_ after any new non-exclusive waiters - * because all non-exclusive waiters are added at the - * beginning of the wait-queue. As such, it's ok to "drop" - * our exclusiveness temporarily when we get woken up without - * having to remove and re-insert us on the wait queue. - */ - for (;;) { - prepare_to_wait_exclusive(sk->sk_sleep, &wait, - TASK_INTERRUPTIBLE); - release_sock(sk); - if (reqsk_queue_empty(&tp->accept_queue)) - timeo = schedule_timeout(timeo); - lock_sock(sk); - err = 0; - if (!reqsk_queue_empty(&tp->accept_queue)) - break; - err = -EINVAL; - if (sk->sk_state != TCP_LISTEN) - break; - err = sock_intr_errno(timeo); - if (signal_pending(current)) - break; - err = -EAGAIN; - if (!timeo) - break; - } - finish_wait(sk->sk_sleep, &wait); - return err; -} - -/* - * This will accept the next outstanding connection. - */ - -struct sock *tcp_accept(struct sock *sk, int flags, int *err) -{ - struct tcp_sock *tp = tcp_sk(sk); - struct sock *newsk; - int error; - - lock_sock(sk); - - /* We need to make sure that this socket is listening, - * and that it has something pending. - */ - error = -EINVAL; - if (sk->sk_state != TCP_LISTEN) - goto out_err; - - /* Find already established connection */ - if (reqsk_queue_empty(&tp->accept_queue)) { - long timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); - - /* If this is a non blocking socket don't sleep */ - error = -EAGAIN; - if (!timeo) - goto out_err; - - error = wait_for_connect(sk, timeo); - if (error) - goto out_err; - } - - newsk = reqsk_queue_get_child(&tp->accept_queue, sk); - BUG_TRAP(newsk->sk_state != TCP_SYN_RECV); -out: - release_sock(sk); - return newsk; -out_err: - newsk = NULL; - *err = error; - goto out; -} - -/* * Socket option code for TCP. */ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, int optlen) { struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); int val; int err = 0; @@ -1945,7 +1709,7 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, name[val] = 0; lock_sock(sk); - err = tcp_set_congestion_control(tp, name); + err = tcp_set_congestion_control(sk, name); release_sock(sk); return err; } @@ -2022,7 +1786,7 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, elapsed = tp->keepalive_time - elapsed; else elapsed = 0; - tcp_reset_keepalive_timer(sk, elapsed); + inet_csk_reset_keepalive_timer(sk, elapsed); } } break; @@ -2042,7 +1806,7 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, if (val < 1 || val > MAX_TCP_SYNCNT) err = -EINVAL; else - tp->syn_retries = val; + icsk->icsk_syn_retries = val; break; case TCP_LINGER2: @@ -2055,15 +1819,15 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, break; case TCP_DEFER_ACCEPT: - tp->defer_accept = 0; + icsk->icsk_accept_queue.rskq_defer_accept = 0; if (val > 0) { /* Translate value in seconds to number of * retransmits */ - while (tp->defer_accept < 32 && + while (icsk->icsk_accept_queue.rskq_defer_accept < 32 && val > ((TCP_TIMEOUT_INIT / HZ) << - tp->defer_accept)) - tp->defer_accept++; - tp->defer_accept++; + icsk->icsk_accept_queue.rskq_defer_accept)) + icsk->icsk_accept_queue.rskq_defer_accept++; + icsk->icsk_accept_queue.rskq_defer_accept++; } break; @@ -2081,16 +1845,16 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, case TCP_QUICKACK: if (!val) { - tp->ack.pingpong = 1; + icsk->icsk_ack.pingpong = 1; } else { - tp->ack.pingpong = 0; + icsk->icsk_ack.pingpong = 0; if ((1 << sk->sk_state) & (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT) && - tcp_ack_scheduled(tp)) { - tp->ack.pending |= TCP_ACK_PUSHED; + inet_csk_ack_scheduled(sk)) { + icsk->icsk_ack.pending |= ICSK_ACK_PUSHED; cleanup_rbuf(sk, 1); if (!(val & 1)) - tp->ack.pingpong = 1; + icsk->icsk_ack.pingpong = 1; } } break; @@ -2107,15 +1871,16 @@ int tcp_setsockopt(struct sock *sk, int level, int optname, char __user *optval, void tcp_get_info(struct sock *sk, struct tcp_info *info) { struct tcp_sock *tp = tcp_sk(sk); + const struct inet_connection_sock *icsk = inet_csk(sk); u32 now = tcp_time_stamp; memset(info, 0, sizeof(*info)); info->tcpi_state = sk->sk_state; - info->tcpi_ca_state = tp->ca_state; - info->tcpi_retransmits = tp->retransmits; - info->tcpi_probes = tp->probes_out; - info->tcpi_backoff = tp->backoff; + info->tcpi_ca_state = icsk->icsk_ca_state; + info->tcpi_retransmits = icsk->icsk_retransmits; + info->tcpi_probes = icsk->icsk_probes_out; + info->tcpi_backoff = icsk->icsk_backoff; if (tp->rx_opt.tstamp_ok) info->tcpi_options |= TCPI_OPT_TIMESTAMPS; @@ -2130,10 +1895,10 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info) if (tp->ecn_flags&TCP_ECN_OK) info->tcpi_options |= TCPI_OPT_ECN; - info->tcpi_rto = jiffies_to_usecs(tp->rto); - info->tcpi_ato = jiffies_to_usecs(tp->ack.ato); + info->tcpi_rto = jiffies_to_usecs(icsk->icsk_rto); + info->tcpi_ato = jiffies_to_usecs(icsk->icsk_ack.ato); info->tcpi_snd_mss = tp->mss_cache; - info->tcpi_rcv_mss = tp->ack.rcv_mss; + info->tcpi_rcv_mss = icsk->icsk_ack.rcv_mss; info->tcpi_unacked = tp->packets_out; info->tcpi_sacked = tp->sacked_out; @@ -2142,7 +1907,7 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info) info->tcpi_fackets = tp->fackets_out; info->tcpi_last_data_sent = jiffies_to_msecs(now - tp->lsndtime); - info->tcpi_last_data_recv = jiffies_to_msecs(now - tp->ack.lrcvtime); + info->tcpi_last_data_recv = jiffies_to_msecs(now - icsk->icsk_ack.lrcvtime); info->tcpi_last_ack_recv = jiffies_to_msecs(now - tp->rcv_tstamp); info->tcpi_pmtu = tp->pmtu_cookie; @@ -2165,6 +1930,7 @@ EXPORT_SYMBOL_GPL(tcp_get_info); int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); int val, len; @@ -2202,7 +1968,7 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, val = tp->keepalive_probes ? : sysctl_tcp_keepalive_probes; break; case TCP_SYNCNT: - val = tp->syn_retries ? : sysctl_tcp_syn_retries; + val = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; break; case TCP_LINGER2: val = tp->linger2; @@ -2210,8 +1976,8 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, val = (val ? : sysctl_tcp_fin_timeout) / HZ; break; case TCP_DEFER_ACCEPT: - val = !tp->defer_accept ? 0 : ((TCP_TIMEOUT_INIT / HZ) << - (tp->defer_accept - 1)); + val = !icsk->icsk_accept_queue.rskq_defer_accept ? 0 : + ((TCP_TIMEOUT_INIT / HZ) << (icsk->icsk_accept_queue.rskq_defer_accept - 1)); break; case TCP_WINDOW_CLAMP: val = tp->window_clamp; @@ -2232,7 +1998,7 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, return 0; } case TCP_QUICKACK: - val = !tp->ack.pingpong; + val = !icsk->icsk_ack.pingpong; break; case TCP_CONGESTION: @@ -2241,7 +2007,7 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, len = min_t(unsigned int, len, TCP_CA_NAME_MAX); if (put_user(len, optlen)) return -EFAULT; - if (copy_to_user(optval, tp->ca_ops->name, len)) + if (copy_to_user(optval, icsk->icsk_ca_ops->name, len)) return -EFAULT; return 0; default: @@ -2278,79 +2044,72 @@ void __init tcp_init(void) __skb_cb_too_small_for_tcp(sizeof(struct tcp_skb_cb), sizeof(skb->cb)); - tcp_bucket_cachep = kmem_cache_create("tcp_bind_bucket", - sizeof(struct tcp_bind_bucket), - 0, SLAB_HWCACHE_ALIGN, - NULL, NULL); - if (!tcp_bucket_cachep) + tcp_hashinfo.bind_bucket_cachep = + kmem_cache_create("tcp_bind_bucket", + sizeof(struct inet_bind_bucket), 0, + SLAB_HWCACHE_ALIGN, NULL, NULL); + if (!tcp_hashinfo.bind_bucket_cachep) panic("tcp_init: Cannot alloc tcp_bind_bucket cache."); - tcp_timewait_cachep = kmem_cache_create("tcp_tw_bucket", - sizeof(struct tcp_tw_bucket), - 0, SLAB_HWCACHE_ALIGN, - NULL, NULL); - if (!tcp_timewait_cachep) - panic("tcp_init: Cannot alloc tcp_tw_bucket cache."); - /* Size and allocate the main established and bind bucket * hash tables. * * The methodology is similar to that of the buffer cache. */ - tcp_ehash = (struct tcp_ehash_bucket *) + tcp_hashinfo.ehash = alloc_large_system_hash("TCP established", - sizeof(struct tcp_ehash_bucket), + sizeof(struct inet_ehash_bucket), thash_entries, (num_physpages >= 128 * 1024) ? (25 - PAGE_SHIFT) : (27 - PAGE_SHIFT), HASH_HIGHMEM, - &tcp_ehash_size, + &tcp_hashinfo.ehash_size, NULL, 0); - tcp_ehash_size = (1 << tcp_ehash_size) >> 1; - for (i = 0; i < (tcp_ehash_size << 1); i++) { - rwlock_init(&tcp_ehash[i].lock); - INIT_HLIST_HEAD(&tcp_ehash[i].chain); + tcp_hashinfo.ehash_size = (1 << tcp_hashinfo.ehash_size) >> 1; + for (i = 0; i < (tcp_hashinfo.ehash_size << 1); i++) { + rwlock_init(&tcp_hashinfo.ehash[i].lock); + INIT_HLIST_HEAD(&tcp_hashinfo.ehash[i].chain); } - tcp_bhash = (struct tcp_bind_hashbucket *) + tcp_hashinfo.bhash = alloc_large_system_hash("TCP bind", - sizeof(struct tcp_bind_hashbucket), - tcp_ehash_size, + sizeof(struct inet_bind_hashbucket), + tcp_hashinfo.ehash_size, (num_physpages >= 128 * 1024) ? (25 - PAGE_SHIFT) : (27 - PAGE_SHIFT), HASH_HIGHMEM, - &tcp_bhash_size, + &tcp_hashinfo.bhash_size, NULL, 64 * 1024); - tcp_bhash_size = 1 << tcp_bhash_size; - for (i = 0; i < tcp_bhash_size; i++) { - spin_lock_init(&tcp_bhash[i].lock); - INIT_HLIST_HEAD(&tcp_bhash[i].chain); + tcp_hashinfo.bhash_size = 1 << tcp_hashinfo.bhash_size; + for (i = 0; i < tcp_hashinfo.bhash_size; i++) { + spin_lock_init(&tcp_hashinfo.bhash[i].lock); + INIT_HLIST_HEAD(&tcp_hashinfo.bhash[i].chain); } /* Try to be a bit smarter and adjust defaults depending * on available memory. */ for (order = 0; ((1 << order) << PAGE_SHIFT) < - (tcp_bhash_size * sizeof(struct tcp_bind_hashbucket)); + (tcp_hashinfo.bhash_size * sizeof(struct inet_bind_hashbucket)); order++) ; if (order >= 4) { sysctl_local_port_range[0] = 32768; sysctl_local_port_range[1] = 61000; - sysctl_tcp_max_tw_buckets = 180000; + tcp_death_row.sysctl_max_tw_buckets = 180000; sysctl_tcp_max_orphans = 4096 << (order - 4); sysctl_max_syn_backlog = 1024; } else if (order < 3) { sysctl_local_port_range[0] = 1024 * (3 - order); - sysctl_tcp_max_tw_buckets >>= (3 - order); + tcp_death_row.sysctl_max_tw_buckets >>= (3 - order); sysctl_tcp_max_orphans >>= (3 - order); sysctl_max_syn_backlog = 128; } - tcp_port_rover = sysctl_local_port_range[0] - 1; + tcp_hashinfo.port_rover = sysctl_local_port_range[0] - 1; sysctl_tcp_mem[0] = 768 << order; sysctl_tcp_mem[1] = 1024 << order; @@ -2365,14 +2124,12 @@ void __init tcp_init(void) printk(KERN_INFO "TCP: Hash tables configured " "(established %d bind %d)\n", - tcp_ehash_size << 1, tcp_bhash_size); + tcp_hashinfo.ehash_size << 1, tcp_hashinfo.bhash_size); tcp_register_congestion_control(&tcp_reno); } -EXPORT_SYMBOL(tcp_accept); EXPORT_SYMBOL(tcp_close); -EXPORT_SYMBOL(tcp_destroy_sock); EXPORT_SYMBOL(tcp_disconnect); EXPORT_SYMBOL(tcp_getsockopt); EXPORT_SYMBOL(tcp_ioctl); @@ -2384,4 +2141,3 @@ EXPORT_SYMBOL(tcp_sendpage); EXPORT_SYMBOL(tcp_setsockopt); EXPORT_SYMBOL(tcp_shutdown); EXPORT_SYMBOL(tcp_statistics); -EXPORT_SYMBOL(tcp_timewait_cachep); diff --git a/net/ipv4/tcp_bic.c b/net/ipv4/tcp_bic.c index ec38d45d6649..b940346de4e7 100644 --- a/net/ipv4/tcp_bic.c +++ b/net/ipv4/tcp_bic.c @@ -86,11 +86,11 @@ static inline void bictcp_reset(struct bictcp *ca) ca->delayed_ack = 2 << ACK_RATIO_SHIFT; } -static void bictcp_init(struct tcp_sock *tp) +static void bictcp_init(struct sock *sk) { - bictcp_reset(tcp_ca(tp)); + bictcp_reset(inet_csk_ca(sk)); if (initial_ssthresh) - tp->snd_ssthresh = initial_ssthresh; + tcp_sk(sk)->snd_ssthresh = initial_ssthresh; } /* @@ -156,9 +156,10 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd) /* Detect low utilization in congestion avoidance */ -static inline void bictcp_low_utilization(struct tcp_sock *tp, int flag) +static inline void bictcp_low_utilization(struct sock *sk, int flag) { - struct bictcp *ca = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + struct bictcp *ca = inet_csk_ca(sk); u32 dist, delay; /* No time stamp */ @@ -208,12 +209,13 @@ static inline void bictcp_low_utilization(struct tcp_sock *tp, int flag) } -static void bictcp_cong_avoid(struct tcp_sock *tp, u32 ack, +static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 seq_rtt, u32 in_flight, int data_acked) { - struct bictcp *ca = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct bictcp *ca = inet_csk_ca(sk); - bictcp_low_utilization(tp, data_acked); + bictcp_low_utilization(sk, data_acked); if (in_flight < tp->snd_cwnd) return; @@ -242,9 +244,10 @@ static void bictcp_cong_avoid(struct tcp_sock *tp, u32 ack, * behave like Reno until low_window is reached, * then increase congestion window slowly */ -static u32 bictcp_recalc_ssthresh(struct tcp_sock *tp) +static u32 bictcp_recalc_ssthresh(struct sock *sk) { - struct bictcp *ca = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + struct bictcp *ca = inet_csk_ca(sk); ca->epoch_start = 0; /* end of epoch */ @@ -269,31 +272,34 @@ static u32 bictcp_recalc_ssthresh(struct tcp_sock *tp) return max((tp->snd_cwnd * beta) / BICTCP_BETA_SCALE, 2U); } -static u32 bictcp_undo_cwnd(struct tcp_sock *tp) +static u32 bictcp_undo_cwnd(struct sock *sk) { - struct bictcp *ca = tcp_ca(tp); - + const struct tcp_sock *tp = tcp_sk(sk); + const struct bictcp *ca = inet_csk_ca(sk); return max(tp->snd_cwnd, ca->last_max_cwnd); } -static u32 bictcp_min_cwnd(struct tcp_sock *tp) +static u32 bictcp_min_cwnd(struct sock *sk) { + const struct tcp_sock *tp = tcp_sk(sk); return tp->snd_ssthresh; } -static void bictcp_state(struct tcp_sock *tp, u8 new_state) +static void bictcp_state(struct sock *sk, u8 new_state) { if (new_state == TCP_CA_Loss) - bictcp_reset(tcp_ca(tp)); + bictcp_reset(inet_csk_ca(sk)); } /* Track delayed acknowledgement ratio using sliding window * ratio = (15*ratio + sample) / 16 */ -static void bictcp_acked(struct tcp_sock *tp, u32 cnt) +static void bictcp_acked(struct sock *sk, u32 cnt) { - if (cnt > 0 && tp->ca_state == TCP_CA_Open) { - struct bictcp *ca = tcp_ca(tp); + const struct inet_connection_sock *icsk = inet_csk(sk); + + if (cnt > 0 && icsk->icsk_ca_state == TCP_CA_Open) { + struct bictcp *ca = inet_csk_ca(sk); cnt -= ca->delayed_ack >> ACK_RATIO_SHIFT; ca->delayed_ack += cnt; } @@ -314,7 +320,7 @@ static struct tcp_congestion_ops bictcp = { static int __init bictcp_register(void) { - BUG_ON(sizeof(struct bictcp) > TCP_CA_PRIV_SIZE); + BUG_ON(sizeof(struct bictcp) > ICSK_CA_PRIV_SIZE); return tcp_register_congestion_control(&bictcp); } diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c index 4970d10a7785..bbf2d6624e89 100644 --- a/net/ipv4/tcp_cong.c +++ b/net/ipv4/tcp_cong.c @@ -73,33 +73,36 @@ void tcp_unregister_congestion_control(struct tcp_congestion_ops *ca) EXPORT_SYMBOL_GPL(tcp_unregister_congestion_control); /* Assign choice of congestion control. */ -void tcp_init_congestion_control(struct tcp_sock *tp) +void tcp_init_congestion_control(struct sock *sk) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_congestion_ops *ca; - if (tp->ca_ops != &tcp_init_congestion_ops) + if (icsk->icsk_ca_ops != &tcp_init_congestion_ops) return; rcu_read_lock(); list_for_each_entry_rcu(ca, &tcp_cong_list, list) { if (try_module_get(ca->owner)) { - tp->ca_ops = ca; + icsk->icsk_ca_ops = ca; break; } } rcu_read_unlock(); - if (tp->ca_ops->init) - tp->ca_ops->init(tp); + if (icsk->icsk_ca_ops->init) + icsk->icsk_ca_ops->init(sk); } /* Manage refcounts on socket close. */ -void tcp_cleanup_congestion_control(struct tcp_sock *tp) +void tcp_cleanup_congestion_control(struct sock *sk) { - if (tp->ca_ops->release) - tp->ca_ops->release(tp); - module_put(tp->ca_ops->owner); + struct inet_connection_sock *icsk = inet_csk(sk); + + if (icsk->icsk_ca_ops->release) + icsk->icsk_ca_ops->release(sk); + module_put(icsk->icsk_ca_ops->owner); } /* Used by sysctl to change default congestion control */ @@ -143,14 +146,15 @@ void tcp_get_default_congestion_control(char *name) } /* Change congestion control for socket */ -int tcp_set_congestion_control(struct tcp_sock *tp, const char *name) +int tcp_set_congestion_control(struct sock *sk, const char *name) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_congestion_ops *ca; int err = 0; rcu_read_lock(); ca = tcp_ca_find(name); - if (ca == tp->ca_ops) + if (ca == icsk->icsk_ca_ops) goto out; if (!ca) @@ -160,10 +164,10 @@ int tcp_set_congestion_control(struct tcp_sock *tp, const char *name) err = -EBUSY; else { - tcp_cleanup_congestion_control(tp); - tp->ca_ops = ca; - if (tp->ca_ops->init) - tp->ca_ops->init(tp); + tcp_cleanup_congestion_control(sk); + icsk->icsk_ca_ops = ca; + if (icsk->icsk_ca_ops->init) + icsk->icsk_ca_ops->init(sk); } out: rcu_read_unlock(); @@ -177,9 +181,11 @@ int tcp_set_congestion_control(struct tcp_sock *tp, const char *name) /* This is Jacobson's slow start and congestion avoidance. * SIGCOMM '88, p. 328. */ -void tcp_reno_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, u32 in_flight, +void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 rtt, u32 in_flight, int flag) { + struct tcp_sock *tp = tcp_sk(sk); + if (in_flight < tp->snd_cwnd) return; @@ -202,15 +208,17 @@ void tcp_reno_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, u32 in_flight, EXPORT_SYMBOL_GPL(tcp_reno_cong_avoid); /* Slow start threshold is half the congestion window (min 2) */ -u32 tcp_reno_ssthresh(struct tcp_sock *tp) +u32 tcp_reno_ssthresh(struct sock *sk) { + const struct tcp_sock *tp = tcp_sk(sk); return max(tp->snd_cwnd >> 1U, 2U); } EXPORT_SYMBOL_GPL(tcp_reno_ssthresh); /* Lower bound on congestion window. */ -u32 tcp_reno_min_cwnd(struct tcp_sock *tp) +u32 tcp_reno_min_cwnd(struct sock *sk) { + const struct tcp_sock *tp = tcp_sk(sk); return tp->snd_ssthresh/2; } EXPORT_SYMBOL_GPL(tcp_reno_min_cwnd); diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index f66945cb158f..c148c1081880 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -1,5 +1,5 @@ /* - * tcp_diag.c Module for monitoring TCP sockets. + * tcp_diag.c Module for monitoring TCP transport protocols sockets. * * Version: $Id: tcp_diag.c,v 1.3 2002/02/01 22:01:04 davem Exp $ * @@ -12,779 +12,43 @@ */ #include <linux/config.h> -#include <linux/module.h> -#include <linux/types.h> -#include <linux/fcntl.h> -#include <linux/random.h> -#include <linux/cache.h> -#include <linux/init.h> -#include <linux/time.h> - -#include <net/icmp.h> -#include <net/tcp.h> -#include <net/ipv6.h> -#include <net/inet_common.h> - -#include <linux/inet.h> -#include <linux/stddef.h> - -#include <linux/tcp_diag.h> -struct tcpdiag_entry -{ - u32 *saddr; - u32 *daddr; - u16 sport; - u16 dport; - u16 family; - u16 userlocks; -}; +#include <linux/module.h> +#include <linux/inet_diag.h> -static struct sock *tcpnl; +#include <linux/tcp.h> -#define TCPDIAG_PUT(skb, attrtype, attrlen) \ - RTA_DATA(__RTA_PUT(skb, attrtype, attrlen)) +#include <net/tcp.h> -static int tcpdiag_fill(struct sk_buff *skb, struct sock *sk, - int ext, u32 pid, u32 seq, u16 nlmsg_flags) +static void tcp_diag_get_info(struct sock *sk, struct inet_diag_msg *r, + void *_info) { - struct inet_sock *inet = inet_sk(sk); - struct tcp_sock *tp = tcp_sk(sk); - struct tcpdiagmsg *r; - struct nlmsghdr *nlh; - struct tcp_info *info = NULL; - struct tcpdiag_meminfo *minfo = NULL; - unsigned char *b = skb->tail; - - nlh = NLMSG_PUT(skb, pid, seq, TCPDIAG_GETSOCK, sizeof(*r)); - nlh->nlmsg_flags = nlmsg_flags; - r = NLMSG_DATA(nlh); - if (sk->sk_state != TCP_TIME_WAIT) { - if (ext & (1<<(TCPDIAG_MEMINFO-1))) - minfo = TCPDIAG_PUT(skb, TCPDIAG_MEMINFO, sizeof(*minfo)); - 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; - r->tcpdiag_timer = 0; - r->tcpdiag_retrans = 0; - - r->id.tcpdiag_if = sk->sk_bound_dev_if; - r->id.tcpdiag_cookie[0] = (u32)(unsigned long)sk; - r->id.tcpdiag_cookie[1] = (u32)(((unsigned long)sk >> 31) >> 1); - - if (r->tcpdiag_state == TCP_TIME_WAIT) { - struct tcp_tw_bucket *tw = (struct tcp_tw_bucket*)sk; - long tmo = tw->tw_ttd - jiffies; - if (tmo < 0) - tmo = 0; - - r->id.tcpdiag_sport = tw->tw_sport; - r->id.tcpdiag_dport = tw->tw_dport; - r->id.tcpdiag_src[0] = tw->tw_rcv_saddr; - r->id.tcpdiag_dst[0] = tw->tw_daddr; - r->tcpdiag_state = tw->tw_substate; - r->tcpdiag_timer = 3; - r->tcpdiag_expires = (tmo*1000+HZ-1)/HZ; - r->tcpdiag_rqueue = 0; - r->tcpdiag_wqueue = 0; - r->tcpdiag_uid = 0; - r->tcpdiag_inode = 0; -#ifdef CONFIG_IP_TCPDIAG_IPV6 - if (r->tcpdiag_family == AF_INET6) { - ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_src, - &tw->tw_v6_rcv_saddr); - ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_dst, - &tw->tw_v6_daddr); - } -#endif - nlh->nlmsg_len = skb->tail - b; - return skb->len; - } - - r->id.tcpdiag_sport = inet->sport; - r->id.tcpdiag_dport = inet->dport; - r->id.tcpdiag_src[0] = inet->rcv_saddr; - r->id.tcpdiag_dst[0] = inet->daddr; - -#ifdef CONFIG_IP_TCPDIAG_IPV6 - if (r->tcpdiag_family == AF_INET6) { - struct ipv6_pinfo *np = inet6_sk(sk); - - ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_src, - &np->rcv_saddr); - ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_dst, - &np->daddr); - } -#endif - -#define EXPIRES_IN_MS(tmo) ((tmo-jiffies)*1000+HZ-1)/HZ - - if (tp->pending == TCP_TIME_RETRANS) { - r->tcpdiag_timer = 1; - r->tcpdiag_retrans = tp->retransmits; - r->tcpdiag_expires = EXPIRES_IN_MS(tp->timeout); - } else if (tp->pending == TCP_TIME_PROBE0) { - r->tcpdiag_timer = 4; - r->tcpdiag_retrans = tp->probes_out; - r->tcpdiag_expires = EXPIRES_IN_MS(tp->timeout); - } else if (timer_pending(&sk->sk_timer)) { - r->tcpdiag_timer = 2; - r->tcpdiag_retrans = tp->probes_out; - r->tcpdiag_expires = EXPIRES_IN_MS(sk->sk_timer.expires); - } else { - r->tcpdiag_timer = 0; - r->tcpdiag_expires = 0; - } -#undef EXPIRES_IN_MS + const struct tcp_sock *tp = tcp_sk(sk); + struct tcp_info *info = _info; - r->tcpdiag_rqueue = tp->rcv_nxt - tp->copied_seq; - r->tcpdiag_wqueue = tp->write_seq - tp->snd_una; - r->tcpdiag_uid = sock_i_uid(sk); - r->tcpdiag_inode = sock_i_ino(sk); - - if (minfo) { - minfo->tcpdiag_rmem = atomic_read(&sk->sk_rmem_alloc); - minfo->tcpdiag_wmem = sk->sk_wmem_queued; - minfo->tcpdiag_fmem = sk->sk_forward_alloc; - minfo->tcpdiag_tmem = atomic_read(&sk->sk_wmem_alloc); - } - - if (info) + r->idiag_rqueue = tp->rcv_nxt - tp->copied_seq; + r->idiag_wqueue = tp->write_seq - tp->snd_una; + if (info != NULL) tcp_get_info(sk, info); - - 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; - -rtattr_failure: -nlmsg_failure: - skb_trim(skb, b - skb->data); - return -1; -} - -extern struct sock *tcp_v4_lookup(u32 saddr, u16 sport, u32 daddr, u16 dport, - int dif); -#ifdef CONFIG_IP_TCPDIAG_IPV6 -extern struct sock *tcp_v6_lookup(struct in6_addr *saddr, u16 sport, - struct in6_addr *daddr, u16 dport, - int dif); -#else -static inline struct sock *tcp_v6_lookup(struct in6_addr *saddr, u16 sport, - struct in6_addr *daddr, u16 dport, - int dif) -{ - return NULL; -} -#endif - -static int tcpdiag_get_exact(struct sk_buff *in_skb, const struct nlmsghdr *nlh) -{ - int err; - struct sock *sk; - struct tcpdiagreq *req = NLMSG_DATA(nlh); - struct sk_buff *rep; - - if (req->tcpdiag_family == AF_INET) { - sk = tcp_v4_lookup(req->id.tcpdiag_dst[0], req->id.tcpdiag_dport, - req->id.tcpdiag_src[0], req->id.tcpdiag_sport, - req->id.tcpdiag_if); - } -#ifdef CONFIG_IP_TCPDIAG_IPV6 - else if (req->tcpdiag_family == AF_INET6) { - sk = tcp_v6_lookup((struct in6_addr*)req->id.tcpdiag_dst, req->id.tcpdiag_dport, - (struct in6_addr*)req->id.tcpdiag_src, req->id.tcpdiag_sport, - req->id.tcpdiag_if); - } -#endif - else { - return -EINVAL; - } - - if (sk == NULL) - return -ENOENT; - - err = -ESTALE; - if ((req->id.tcpdiag_cookie[0] != TCPDIAG_NOCOOKIE || - req->id.tcpdiag_cookie[1] != TCPDIAG_NOCOOKIE) && - ((u32)(unsigned long)sk != req->id.tcpdiag_cookie[0] || - (u32)((((unsigned long)sk) >> 31) >> 1) != req->id.tcpdiag_cookie[1])) - goto out; - - err = -ENOMEM; - rep = alloc_skb(NLMSG_SPACE(sizeof(struct tcpdiagmsg)+ - sizeof(struct tcpdiag_meminfo)+ - sizeof(struct tcp_info)+64), GFP_KERNEL); - if (!rep) - goto out; - - if (tcpdiag_fill(rep, sk, req->tcpdiag_ext, - NETLINK_CB(in_skb).pid, - nlh->nlmsg_seq, 0) <= 0) - BUG(); - - err = netlink_unicast(tcpnl, rep, NETLINK_CB(in_skb).pid, MSG_DONTWAIT); - if (err > 0) - err = 0; - -out: - if (sk) { - if (sk->sk_state == TCP_TIME_WAIT) - tcp_tw_put((struct tcp_tw_bucket*)sk); - else - sock_put(sk); - } - return err; -} - -static int bitstring_match(const u32 *a1, const u32 *a2, int bits) -{ - int words = bits >> 5; - - bits &= 0x1f; - - if (words) { - if (memcmp(a1, a2, words << 2)) - return 0; - } - if (bits) { - __u32 w1, w2; - __u32 mask; - - w1 = a1[words]; - w2 = a2[words]; - - mask = htonl((0xffffffff) << (32 - bits)); - - if ((w1 ^ w2) & mask) - return 0; - } - - return 1; -} - - -static int tcpdiag_bc_run(const void *bc, int len, - const struct tcpdiag_entry *entry) -{ - while (len > 0) { - int yes = 1; - const struct tcpdiag_bc_op *op = bc; - - switch (op->code) { - case TCPDIAG_BC_NOP: - break; - case TCPDIAG_BC_JMP: - yes = 0; - break; - case TCPDIAG_BC_S_GE: - yes = entry->sport >= op[1].no; - break; - case TCPDIAG_BC_S_LE: - yes = entry->dport <= op[1].no; - break; - case TCPDIAG_BC_D_GE: - yes = entry->dport >= op[1].no; - break; - case TCPDIAG_BC_D_LE: - yes = entry->dport <= op[1].no; - break; - case TCPDIAG_BC_AUTO: - yes = !(entry->userlocks & SOCK_BINDPORT_LOCK); - break; - case TCPDIAG_BC_S_COND: - case TCPDIAG_BC_D_COND: - { - struct tcpdiag_hostcond *cond = (struct tcpdiag_hostcond*)(op+1); - u32 *addr; - - if (cond->port != -1 && - cond->port != (op->code == TCPDIAG_BC_S_COND ? - entry->sport : entry->dport)) { - yes = 0; - break; - } - - if (cond->prefix_len == 0) - break; - - if (op->code == TCPDIAG_BC_S_COND) - addr = entry->saddr; - else - addr = entry->daddr; - - if (bitstring_match(addr, cond->addr, cond->prefix_len)) - break; - if (entry->family == AF_INET6 && - cond->family == AF_INET) { - if (addr[0] == 0 && addr[1] == 0 && - addr[2] == htonl(0xffff) && - bitstring_match(addr+3, cond->addr, cond->prefix_len)) - break; - } - yes = 0; - break; - } - } - - if (yes) { - len -= op->yes; - bc += op->yes; - } else { - len -= op->no; - bc += op->no; - } - } - return (len == 0); -} - -static int valid_cc(const void *bc, int len, int cc) -{ - while (len >= 0) { - const struct tcpdiag_bc_op *op = bc; - - if (cc > len) - return 0; - if (cc == len) - return 1; - if (op->yes < 4) - return 0; - len -= op->yes; - bc += op->yes; - } - return 0; -} - -static int tcpdiag_bc_audit(const void *bytecode, int bytecode_len) -{ - const unsigned char *bc = bytecode; - int len = bytecode_len; - - while (len > 0) { - struct tcpdiag_bc_op *op = (struct tcpdiag_bc_op*)bc; - -//printk("BC: %d %d %d {%d} / %d\n", op->code, op->yes, op->no, op[1].no, len); - switch (op->code) { - case TCPDIAG_BC_AUTO: - case TCPDIAG_BC_S_COND: - case TCPDIAG_BC_D_COND: - case TCPDIAG_BC_S_GE: - case TCPDIAG_BC_S_LE: - case TCPDIAG_BC_D_GE: - case TCPDIAG_BC_D_LE: - if (op->yes < 4 || op->yes > len+4) - return -EINVAL; - case TCPDIAG_BC_JMP: - if (op->no < 4 || op->no > len+4) - return -EINVAL; - if (op->no < len && - !valid_cc(bytecode, bytecode_len, len-op->no)) - return -EINVAL; - break; - case TCPDIAG_BC_NOP: - if (op->yes < 4 || op->yes > len+4) - return -EINVAL; - break; - default: - return -EINVAL; - } - bc += op->yes; - len -= op->yes; - } - return len == 0 ? 0 : -EINVAL; -} - -static int tcpdiag_dump_sock(struct sk_buff *skb, struct sock *sk, - struct netlink_callback *cb) -{ - struct tcpdiagreq *r = NLMSG_DATA(cb->nlh); - - if (cb->nlh->nlmsg_len > 4 + NLMSG_SPACE(sizeof(*r))) { - struct tcpdiag_entry entry; - struct rtattr *bc = (struct rtattr *)(r + 1); - struct inet_sock *inet = inet_sk(sk); - - entry.family = sk->sk_family; -#ifdef CONFIG_IP_TCPDIAG_IPV6 - if (entry.family == AF_INET6) { - struct ipv6_pinfo *np = inet6_sk(sk); - - entry.saddr = np->rcv_saddr.s6_addr32; - entry.daddr = np->daddr.s6_addr32; - } else -#endif - { - entry.saddr = &inet->rcv_saddr; - entry.daddr = &inet->daddr; - } - entry.sport = inet->num; - entry.dport = ntohs(inet->dport); - entry.userlocks = sk->sk_userlocks; - - if (!tcpdiag_bc_run(RTA_DATA(bc), RTA_PAYLOAD(bc), &entry)) - return 0; - } - - return tcpdiag_fill(skb, sk, r->tcpdiag_ext, NETLINK_CB(cb->skb).pid, - cb->nlh->nlmsg_seq, NLM_F_MULTI); } -static int tcpdiag_fill_req(struct sk_buff *skb, struct sock *sk, - struct request_sock *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; - struct nlmsghdr *nlh; - long tmo; - - nlh = NLMSG_PUT(skb, pid, seq, TCPDIAG_GETSOCK, sizeof(*r)); - nlh->nlmsg_flags = NLM_F_MULTI; - r = NLMSG_DATA(nlh); - - r->tcpdiag_family = sk->sk_family; - r->tcpdiag_state = TCP_SYN_RECV; - r->tcpdiag_timer = 1; - r->tcpdiag_retrans = req->retrans; - - r->id.tcpdiag_if = sk->sk_bound_dev_if; - r->id.tcpdiag_cookie[0] = (u32)(unsigned long)req; - r->id.tcpdiag_cookie[1] = (u32)(((unsigned long)req >> 31) >> 1); - - tmo = req->expires - jiffies; - if (tmo < 0) - tmo = 0; - - r->id.tcpdiag_sport = inet->sport; - 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; - r->tcpdiag_uid = sock_i_uid(sk); - r->tcpdiag_inode = 0; -#ifdef CONFIG_IP_TCPDIAG_IPV6 - if (r->tcpdiag_family == AF_INET6) { - ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_src, - &tcp6_rsk(req)->loc_addr); - ipv6_addr_copy((struct in6_addr *)r->id.tcpdiag_dst, - &tcp6_rsk(req)->rmt_addr); - } -#endif - nlh->nlmsg_len = skb->tail - b; - - return skb->len; - -nlmsg_failure: - skb_trim(skb, b - skb->data); - return -1; -} - -static int tcpdiag_dump_reqs(struct sk_buff *skb, struct sock *sk, - struct netlink_callback *cb) -{ - struct tcpdiag_entry entry; - struct tcpdiagreq *r = NLMSG_DATA(cb->nlh); - struct tcp_sock *tp = tcp_sk(sk); - struct listen_sock *lopt; - struct rtattr *bc = NULL; - struct inet_sock *inet = inet_sk(sk); - int j, s_j; - int reqnum, s_reqnum; - int err = 0; - - s_j = cb->args[3]; - s_reqnum = cb->args[4]; - - if (s_j > 0) - s_j--; - - entry.family = sk->sk_family; - - read_lock_bh(&tp->accept_queue.syn_wait_lock); - - lopt = tp->accept_queue.listen_opt; - if (!lopt || !lopt->qlen) - goto out; - - if (cb->nlh->nlmsg_len > 4 + NLMSG_SPACE(sizeof(*r))) { - bc = (struct rtattr *)(r + 1); - entry.sport = inet->num; - entry.userlocks = sk->sk_userlocks; - } - - for (j = s_j; j < TCP_SYNQ_HSIZE; j++) { - struct request_sock *req, *head = lopt->syn_table[j]; - - 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 != ireq->rmt_port && - r->id.tcpdiag_dport) - continue; - - if (bc) { - entry.saddr = -#ifdef CONFIG_IP_TCPDIAG_IPV6 - (entry.family == AF_INET6) ? - tcp6_rsk(req)->loc_addr.s6_addr32 : -#endif - &ireq->loc_addr; - entry.daddr = -#ifdef CONFIG_IP_TCPDIAG_IPV6 - (entry.family == AF_INET6) ? - tcp6_rsk(req)->rmt_addr.s6_addr32 : -#endif - &ireq->rmt_addr; - entry.dport = ntohs(ireq->rmt_port); - - if (!tcpdiag_bc_run(RTA_DATA(bc), - RTA_PAYLOAD(bc), &entry)) - continue; - } - - err = tcpdiag_fill_req(skb, sk, req, - NETLINK_CB(cb->skb).pid, - cb->nlh->nlmsg_seq); - if (err < 0) { - cb->args[3] = j + 1; - cb->args[4] = reqnum; - goto out; - } - } - - s_reqnum = 0; - } - -out: - read_unlock_bh(&tp->accept_queue.syn_wait_lock); - - return err; -} - -static int tcpdiag_dump(struct sk_buff *skb, struct netlink_callback *cb) -{ - int i, num; - int s_i, s_num; - struct tcpdiagreq *r = NLMSG_DATA(cb->nlh); - - s_i = cb->args[1]; - s_num = num = cb->args[2]; - - if (cb->args[0] == 0) { - if (!(r->tcpdiag_states&(TCPF_LISTEN|TCPF_SYN_RECV))) - goto skip_listen_ht; - tcp_listen_lock(); - for (i = s_i; i < TCP_LHTABLE_SIZE; i++) { - struct sock *sk; - struct hlist_node *node; - - num = 0; - sk_for_each(sk, node, &tcp_listening_hash[i]) { - struct inet_sock *inet = inet_sk(sk); - - if (num < s_num) { - num++; - continue; - } - - if (r->id.tcpdiag_sport != inet->sport && - r->id.tcpdiag_sport) - goto next_listen; - - if (!(r->tcpdiag_states&TCPF_LISTEN) || - r->id.tcpdiag_dport || - cb->args[3] > 0) - goto syn_recv; - - if (tcpdiag_dump_sock(skb, sk, cb) < 0) { - tcp_listen_unlock(); - goto done; - } - -syn_recv: - if (!(r->tcpdiag_states&TCPF_SYN_RECV)) - goto next_listen; - - if (tcpdiag_dump_reqs(skb, sk, cb) < 0) { - tcp_listen_unlock(); - goto done; - } - -next_listen: - cb->args[3] = 0; - cb->args[4] = 0; - ++num; - } - - s_num = 0; - cb->args[3] = 0; - cb->args[4] = 0; - } - tcp_listen_unlock(); -skip_listen_ht: - cb->args[0] = 1; - s_i = num = s_num = 0; - } - - if (!(r->tcpdiag_states&~(TCPF_LISTEN|TCPF_SYN_RECV))) - return skb->len; - - for (i = s_i; i < tcp_ehash_size; i++) { - struct tcp_ehash_bucket *head = &tcp_ehash[i]; - struct sock *sk; - struct hlist_node *node; - - if (i > s_i) - s_num = 0; - - read_lock_bh(&head->lock); - - num = 0; - sk_for_each(sk, node, &head->chain) { - struct inet_sock *inet = inet_sk(sk); - - if (num < s_num) - goto next_normal; - if (!(r->tcpdiag_states & (1 << sk->sk_state))) - goto next_normal; - if (r->id.tcpdiag_sport != inet->sport && - r->id.tcpdiag_sport) - goto next_normal; - if (r->id.tcpdiag_dport != inet->dport && r->id.tcpdiag_dport) - goto next_normal; - if (tcpdiag_dump_sock(skb, sk, cb) < 0) { - read_unlock_bh(&head->lock); - goto done; - } -next_normal: - ++num; - } - - if (r->tcpdiag_states&TCPF_TIME_WAIT) { - sk_for_each(sk, node, - &tcp_ehash[i + tcp_ehash_size].chain) { - struct inet_sock *inet = inet_sk(sk); - - if (num < s_num) - goto next_dying; - if (r->id.tcpdiag_sport != inet->sport && - r->id.tcpdiag_sport) - goto next_dying; - if (r->id.tcpdiag_dport != inet->dport && - r->id.tcpdiag_dport) - goto next_dying; - if (tcpdiag_dump_sock(skb, sk, cb) < 0) { - read_unlock_bh(&head->lock); - goto done; - } -next_dying: - ++num; - } - } - read_unlock_bh(&head->lock); - } - -done: - cb->args[1] = i; - cb->args[2] = num; - return skb->len; -} - -static int tcpdiag_dump_done(struct netlink_callback *cb) -{ - return 0; -} - - -static __inline__ int -tcpdiag_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) -{ - if (!(nlh->nlmsg_flags&NLM_F_REQUEST)) - return 0; - - if (nlh->nlmsg_type != TCPDIAG_GETSOCK) - goto err_inval; - - if (NLMSG_LENGTH(sizeof(struct tcpdiagreq)) > skb->len) - goto err_inval; - - if (nlh->nlmsg_flags&NLM_F_DUMP) { - if (nlh->nlmsg_len > 4 + NLMSG_SPACE(sizeof(struct tcpdiagreq))) { - struct rtattr *rta = (struct rtattr*)(NLMSG_DATA(nlh) + sizeof(struct tcpdiagreq)); - if (rta->rta_type != TCPDIAG_REQ_BYTECODE || - rta->rta_len < 8 || - rta->rta_len > nlh->nlmsg_len - NLMSG_SPACE(sizeof(struct tcpdiagreq))) - goto err_inval; - if (tcpdiag_bc_audit(RTA_DATA(rta), RTA_PAYLOAD(rta))) - goto err_inval; - } - return netlink_dump_start(tcpnl, skb, nlh, - tcpdiag_dump, - tcpdiag_dump_done); - } else { - return tcpdiag_get_exact(skb, nlh); - } - -err_inval: - return -EINVAL; -} - - -static inline void tcpdiag_rcv_skb(struct sk_buff *skb) -{ - int err; - struct nlmsghdr * nlh; - - if (skb->len >= NLMSG_SPACE(0)) { - nlh = (struct nlmsghdr *)skb->data; - if (nlh->nlmsg_len < sizeof(*nlh) || skb->len < nlh->nlmsg_len) - return; - err = tcpdiag_rcv_msg(skb, nlh); - if (err || nlh->nlmsg_flags & NLM_F_ACK) - netlink_ack(skb, nlh, err); - } -} - -static void tcpdiag_rcv(struct sock *sk, int len) -{ - struct sk_buff *skb; - unsigned int qlen = skb_queue_len(&sk->sk_receive_queue); - - while (qlen-- && (skb = skb_dequeue(&sk->sk_receive_queue))) { - tcpdiag_rcv_skb(skb); - kfree_skb(skb); - } -} +static struct inet_diag_handler tcp_diag_handler = { + .idiag_hashinfo = &tcp_hashinfo, + .idiag_get_info = tcp_diag_get_info, + .idiag_type = TCPDIAG_GETSOCK, + .idiag_info_size = sizeof(struct tcp_info), +}; -static int __init tcpdiag_init(void) +static int __init tcp_diag_init(void) { - tcpnl = netlink_kernel_create(NETLINK_TCPDIAG, tcpdiag_rcv); - if (tcpnl == NULL) - return -ENOMEM; - return 0; + return inet_diag_register(&tcp_diag_handler); } -static void __exit tcpdiag_exit(void) +static void __exit tcp_diag_exit(void) { - sock_release(tcpnl->sk_socket); + inet_diag_unregister(&tcp_diag_handler); } -module_init(tcpdiag_init); -module_exit(tcpdiag_exit); +module_init(tcp_diag_init); +module_exit(tcp_diag_exit); MODULE_LICENSE("GPL"); diff --git a/net/ipv4/tcp_highspeed.c b/net/ipv4/tcp_highspeed.c index 36c51f8136bf..6acc04bde080 100644 --- a/net/ipv4/tcp_highspeed.c +++ b/net/ipv4/tcp_highspeed.c @@ -98,9 +98,10 @@ struct hstcp { u32 ai; }; -static void hstcp_init(struct tcp_sock *tp) +static void hstcp_init(struct sock *sk) { - struct hstcp *ca = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct hstcp *ca = inet_csk_ca(sk); ca->ai = 0; @@ -109,10 +110,11 @@ static void hstcp_init(struct tcp_sock *tp) tp->snd_cwnd_clamp = min_t(u32, tp->snd_cwnd_clamp, 0xffffffff/128); } -static void hstcp_cong_avoid(struct tcp_sock *tp, u32 adk, u32 rtt, +static void hstcp_cong_avoid(struct sock *sk, u32 adk, u32 rtt, u32 in_flight, int good) { - struct hstcp *ca = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct hstcp *ca = inet_csk_ca(sk); if (in_flight < tp->snd_cwnd) return; @@ -143,9 +145,10 @@ static void hstcp_cong_avoid(struct tcp_sock *tp, u32 adk, u32 rtt, } } -static u32 hstcp_ssthresh(struct tcp_sock *tp) +static u32 hstcp_ssthresh(struct sock *sk) { - struct hstcp *ca = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + const struct hstcp *ca = inet_csk_ca(sk); /* Do multiplicative decrease */ return max(tp->snd_cwnd - ((tp->snd_cwnd * hstcp_aimd_vals[ca->ai].md) >> 8), 2U); @@ -164,7 +167,7 @@ static struct tcp_congestion_ops tcp_highspeed = { static int __init hstcp_register(void) { - BUG_ON(sizeof(struct hstcp) > TCP_CA_PRIV_SIZE); + BUG_ON(sizeof(struct hstcp) > ICSK_CA_PRIV_SIZE); return tcp_register_congestion_control(&tcp_highspeed); } diff --git a/net/ipv4/tcp_htcp.c b/net/ipv4/tcp_htcp.c index 40168275acf9..e47b37984e95 100644 --- a/net/ipv4/tcp_htcp.c +++ b/net/ipv4/tcp_htcp.c @@ -55,18 +55,21 @@ static inline void htcp_reset(struct htcp *ca) ca->snd_cwnd_cnt2 = 0; } -static u32 htcp_cwnd_undo(struct tcp_sock *tp) +static u32 htcp_cwnd_undo(struct sock *sk) { - struct htcp *ca = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + struct htcp *ca = inet_csk_ca(sk); ca->ccount = ca->undo_ccount; ca->maxRTT = ca->undo_maxRTT; ca->old_maxB = ca->undo_old_maxB; return max(tp->snd_cwnd, (tp->snd_ssthresh<<7)/ca->beta); } -static inline void measure_rtt(struct tcp_sock *tp) +static inline void measure_rtt(struct sock *sk) { - struct htcp *ca = tcp_ca(tp); + const struct inet_connection_sock *icsk = inet_csk(sk); + const struct tcp_sock *tp = tcp_sk(sk); + struct htcp *ca = inet_csk_ca(sk); u32 srtt = tp->srtt>>3; /* keep track of minimum RTT seen so far, minRTT is zero at first */ @@ -74,7 +77,7 @@ static inline void measure_rtt(struct tcp_sock *tp) ca->minRTT = srtt; /* max RTT */ - if (tp->ca_state == TCP_CA_Open && tp->snd_ssthresh < 0xFFFF && ca->ccount > 3) { + if (icsk->icsk_ca_state == TCP_CA_Open && tp->snd_ssthresh < 0xFFFF && ca->ccount > 3) { if (ca->maxRTT < ca->minRTT) ca->maxRTT = ca->minRTT; if (ca->maxRTT < srtt && srtt <= ca->maxRTT+HZ/50) @@ -82,13 +85,16 @@ static inline void measure_rtt(struct tcp_sock *tp) } } -static void measure_achieved_throughput(struct tcp_sock *tp, u32 pkts_acked) +static void measure_achieved_throughput(struct sock *sk, u32 pkts_acked) { - struct htcp *ca = tcp_ca(tp); + const struct inet_connection_sock *icsk = inet_csk(sk); + const struct tcp_sock *tp = tcp_sk(sk); + struct htcp *ca = inet_csk_ca(sk); u32 now = tcp_time_stamp; /* achieved throughput calculations */ - if (tp->ca_state != TCP_CA_Open && tp->ca_state != TCP_CA_Disorder) { + if (icsk->icsk_ca_state != TCP_CA_Open && + icsk->icsk_ca_state != TCP_CA_Disorder) { ca->packetcount = 0; ca->lasttime = now; return; @@ -173,9 +179,9 @@ static inline void htcp_alpha_update(struct htcp *ca) * that point do we really have a real sense of maxRTT (the queues en route * were getting just too full now). */ -static void htcp_param_update(struct tcp_sock *tp) +static void htcp_param_update(struct sock *sk) { - struct htcp *ca = tcp_ca(tp); + struct htcp *ca = inet_csk_ca(sk); u32 minRTT = ca->minRTT; u32 maxRTT = ca->maxRTT; @@ -187,17 +193,19 @@ static void htcp_param_update(struct tcp_sock *tp) ca->maxRTT = minRTT + ((maxRTT-minRTT)*95)/100; } -static u32 htcp_recalc_ssthresh(struct tcp_sock *tp) +static u32 htcp_recalc_ssthresh(struct sock *sk) { - struct htcp *ca = tcp_ca(tp); - htcp_param_update(tp); + const struct tcp_sock *tp = tcp_sk(sk); + const struct htcp *ca = inet_csk_ca(sk); + htcp_param_update(sk); return max((tp->snd_cwnd * ca->beta) >> 7, 2U); } -static void htcp_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, +static void htcp_cong_avoid(struct sock *sk, u32 ack, u32 rtt, u32 in_flight, int data_acked) { - struct htcp *ca = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct htcp *ca = inet_csk_ca(sk); if (in_flight < tp->snd_cwnd) return; @@ -207,7 +215,7 @@ static void htcp_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, if (tp->snd_cwnd < tp->snd_cwnd_clamp) tp->snd_cwnd++; } else { - measure_rtt(tp); + measure_rtt(sk); /* keep track of number of round-trip times since last backoff event */ if (ca->snd_cwnd_cnt2++ > tp->snd_cwnd) { @@ -229,28 +237,29 @@ static void htcp_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, } /* Lower bound on congestion window. */ -static u32 htcp_min_cwnd(struct tcp_sock *tp) +static u32 htcp_min_cwnd(struct sock *sk) { + const struct tcp_sock *tp = tcp_sk(sk); return tp->snd_ssthresh; } -static void htcp_init(struct tcp_sock *tp) +static void htcp_init(struct sock *sk) { - struct htcp *ca = tcp_ca(tp); + struct htcp *ca = inet_csk_ca(sk); memset(ca, 0, sizeof(struct htcp)); ca->alpha = ALPHA_BASE; ca->beta = BETA_MIN; } -static void htcp_state(struct tcp_sock *tp, u8 new_state) +static void htcp_state(struct sock *sk, u8 new_state) { switch (new_state) { case TCP_CA_CWR: case TCP_CA_Recovery: case TCP_CA_Loss: - htcp_reset(tcp_ca(tp)); + htcp_reset(inet_csk_ca(sk)); break; } } @@ -269,7 +278,7 @@ static struct tcp_congestion_ops htcp = { static int __init htcp_register(void) { - BUG_ON(sizeof(struct htcp) > TCP_CA_PRIV_SIZE); + BUG_ON(sizeof(struct htcp) > ICSK_CA_PRIV_SIZE); BUILD_BUG_ON(BETA_MIN >= BETA_MAX); if (!use_bandwidth_switch) htcp.pkts_acked = NULL; diff --git a/net/ipv4/tcp_hybla.c b/net/ipv4/tcp_hybla.c index 13a66342c304..77add63623df 100644 --- a/net/ipv4/tcp_hybla.c +++ b/net/ipv4/tcp_hybla.c @@ -33,19 +33,20 @@ MODULE_PARM_DESC(rtt0, "reference rout trip time (ms)"); /* This is called to refresh values for hybla parameters */ -static inline void hybla_recalc_param (struct tcp_sock *tp) +static inline void hybla_recalc_param (struct sock *sk) { - struct hybla *ca = tcp_ca(tp); + struct hybla *ca = inet_csk_ca(sk); - ca->rho_3ls = max_t(u32, tp->srtt / msecs_to_jiffies(rtt0), 8); + ca->rho_3ls = max_t(u32, tcp_sk(sk)->srtt / msecs_to_jiffies(rtt0), 8); ca->rho = ca->rho_3ls >> 3; ca->rho2_7ls = (ca->rho_3ls * ca->rho_3ls) << 1; ca->rho2 = ca->rho2_7ls >>7; } -static void hybla_init(struct tcp_sock *tp) +static void hybla_init(struct sock *sk) { - struct hybla *ca = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct hybla *ca = inet_csk_ca(sk); ca->rho = 0; ca->rho2 = 0; @@ -57,17 +58,16 @@ static void hybla_init(struct tcp_sock *tp) tp->snd_cwnd_clamp = 65535; /* 1st Rho measurement based on initial srtt */ - hybla_recalc_param(tp); + hybla_recalc_param(sk); /* set minimum rtt as this is the 1st ever seen */ ca->minrtt = tp->srtt; tp->snd_cwnd = ca->rho; } -static void hybla_state(struct tcp_sock *tp, u8 ca_state) +static void hybla_state(struct sock *sk, u8 ca_state) { - struct hybla *ca = tcp_ca(tp); - + struct hybla *ca = inet_csk_ca(sk); ca->hybla_en = (ca_state == TCP_CA_Open); } @@ -86,27 +86,28 @@ static inline u32 hybla_fraction(u32 odds) * o Give cwnd a new value based on the model proposed * o remember increments <1 */ -static void hybla_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, +static void hybla_cong_avoid(struct sock *sk, u32 ack, u32 rtt, u32 in_flight, int flag) { - struct hybla *ca = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct hybla *ca = inet_csk_ca(sk); u32 increment, odd, rho_fractions; int is_slowstart = 0; /* Recalculate rho only if this srtt is the lowest */ if (tp->srtt < ca->minrtt){ - hybla_recalc_param(tp); + hybla_recalc_param(sk); ca->minrtt = tp->srtt; } if (!ca->hybla_en) - return tcp_reno_cong_avoid(tp, ack, rtt, in_flight, flag); + return tcp_reno_cong_avoid(sk, ack, rtt, in_flight, flag); if (in_flight < tp->snd_cwnd) return; if (ca->rho == 0) - hybla_recalc_param(tp); + hybla_recalc_param(sk); rho_fractions = ca->rho_3ls - (ca->rho << 3); @@ -170,7 +171,7 @@ static struct tcp_congestion_ops tcp_hybla = { static int __init hybla_register(void) { - BUG_ON(sizeof(struct hybla) > TCP_CA_PRIV_SIZE); + BUG_ON(sizeof(struct hybla) > ICSK_CA_PRIV_SIZE); return tcp_register_congestion_control(&tcp_hybla); } diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 53a8a5399f1e..1afb080bdf0c 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -114,20 +114,21 @@ int sysctl_tcp_moderate_rcvbuf = 1; /* Adapt the MSS value used to make delayed ack decision to the * real world. */ -static inline void tcp_measure_rcv_mss(struct tcp_sock *tp, - struct sk_buff *skb) +static inline void tcp_measure_rcv_mss(struct sock *sk, + const struct sk_buff *skb) { - unsigned int len, lss; + struct inet_connection_sock *icsk = inet_csk(sk); + const unsigned int lss = icsk->icsk_ack.last_seg_size; + unsigned int len; - lss = tp->ack.last_seg_size; - tp->ack.last_seg_size = 0; + icsk->icsk_ack.last_seg_size = 0; /* skb->len may jitter because of SACKs, even if peer * sends good full-sized frames. */ len = skb->len; - if (len >= tp->ack.rcv_mss) { - tp->ack.rcv_mss = len; + if (len >= icsk->icsk_ack.rcv_mss) { + icsk->icsk_ack.rcv_mss = len; } else { /* Otherwise, we make more careful check taking into account, * that SACKs block is variable. @@ -147,41 +148,44 @@ static inline void tcp_measure_rcv_mss(struct tcp_sock *tp, * tcp header plus fixed timestamp option length. * Resulting "len" is MSS free of SACK jitter. */ - len -= tp->tcp_header_len; - tp->ack.last_seg_size = len; + len -= tcp_sk(sk)->tcp_header_len; + icsk->icsk_ack.last_seg_size = len; if (len == lss) { - tp->ack.rcv_mss = len; + icsk->icsk_ack.rcv_mss = len; return; } } - tp->ack.pending |= TCP_ACK_PUSHED; + icsk->icsk_ack.pending |= ICSK_ACK_PUSHED; } } -static void tcp_incr_quickack(struct tcp_sock *tp) +static void tcp_incr_quickack(struct sock *sk) { - unsigned quickacks = tp->rcv_wnd/(2*tp->ack.rcv_mss); + struct inet_connection_sock *icsk = inet_csk(sk); + unsigned quickacks = tcp_sk(sk)->rcv_wnd / (2 * icsk->icsk_ack.rcv_mss); if (quickacks==0) quickacks=2; - if (quickacks > tp->ack.quick) - tp->ack.quick = min(quickacks, TCP_MAX_QUICKACKS); + if (quickacks > icsk->icsk_ack.quick) + icsk->icsk_ack.quick = min(quickacks, TCP_MAX_QUICKACKS); } -void tcp_enter_quickack_mode(struct tcp_sock *tp) +void tcp_enter_quickack_mode(struct sock *sk) { - tcp_incr_quickack(tp); - tp->ack.pingpong = 0; - tp->ack.ato = TCP_ATO_MIN; + struct inet_connection_sock *icsk = inet_csk(sk); + tcp_incr_quickack(sk); + icsk->icsk_ack.pingpong = 0; + icsk->icsk_ack.ato = TCP_ATO_MIN; } /* Send ACKs quickly, if "quick" count is not exhausted * and the session is not interactive. */ -static __inline__ int tcp_in_quickack_mode(struct tcp_sock *tp) +static inline int tcp_in_quickack_mode(const struct sock *sk) { - return (tp->ack.quick && !tp->ack.pingpong); + const struct inet_connection_sock *icsk = inet_csk(sk); + return icsk->icsk_ack.quick && !icsk->icsk_ack.pingpong; } /* Buffer size and advertised window tuning. @@ -224,8 +228,8 @@ static void tcp_fixup_sndbuf(struct sock *sk) */ /* Slow part of check#2. */ -static int __tcp_grow_window(struct sock *sk, struct tcp_sock *tp, - struct sk_buff *skb) +static int __tcp_grow_window(const struct sock *sk, struct tcp_sock *tp, + const struct sk_buff *skb) { /* Optimize this! */ int truesize = tcp_win_from_space(skb->truesize)/2; @@ -233,7 +237,7 @@ static int __tcp_grow_window(struct sock *sk, struct tcp_sock *tp, while (tp->rcv_ssthresh <= window) { if (truesize <= skb->len) - return 2*tp->ack.rcv_mss; + return 2 * inet_csk(sk)->icsk_ack.rcv_mss; truesize >>= 1; window >>= 1; @@ -260,7 +264,7 @@ static inline void tcp_grow_window(struct sock *sk, struct tcp_sock *tp, if (incr) { tp->rcv_ssthresh = min(tp->rcv_ssthresh + incr, tp->window_clamp); - tp->ack.quick |= 1; + inet_csk(sk)->icsk_ack.quick |= 1; } } } @@ -321,11 +325,12 @@ static void tcp_init_buffer_space(struct sock *sk) /* 5. Recalculate window clamp after socket hit its memory bounds. */ static void tcp_clamp_window(struct sock *sk, struct tcp_sock *tp) { + struct inet_connection_sock *icsk = inet_csk(sk); struct sk_buff *skb; unsigned int app_win = tp->rcv_nxt - tp->copied_seq; int ofo_win = 0; - tp->ack.quick = 0; + icsk->icsk_ack.quick = 0; skb_queue_walk(&tp->out_of_order_queue, skb) { ofo_win += skb->len; @@ -346,8 +351,8 @@ static void tcp_clamp_window(struct sock *sk, struct tcp_sock *tp) app_win += ofo_win; if (atomic_read(&sk->sk_rmem_alloc) >= 2 * sk->sk_rcvbuf) app_win >>= 1; - if (app_win > tp->ack.rcv_mss) - app_win -= tp->ack.rcv_mss; + if (app_win > icsk->icsk_ack.rcv_mss) + app_win -= icsk->icsk_ack.rcv_mss; app_win = max(app_win, 2U*tp->advmss); if (!ofo_win) @@ -415,11 +420,12 @@ new_measure: tp->rcv_rtt_est.time = tcp_time_stamp; } -static inline void tcp_rcv_rtt_measure_ts(struct tcp_sock *tp, struct sk_buff *skb) +static inline void tcp_rcv_rtt_measure_ts(struct sock *sk, const struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); if (tp->rx_opt.rcv_tsecr && (TCP_SKB_CB(skb)->end_seq - - TCP_SKB_CB(skb)->seq >= tp->ack.rcv_mss)) + TCP_SKB_CB(skb)->seq >= inet_csk(sk)->icsk_ack.rcv_mss)) tcp_rcv_rtt_update(tp, tcp_time_stamp - tp->rx_opt.rcv_tsecr, 0); } @@ -492,41 +498,42 @@ new_measure: */ static void tcp_event_data_recv(struct sock *sk, struct tcp_sock *tp, struct sk_buff *skb) { + struct inet_connection_sock *icsk = inet_csk(sk); u32 now; - tcp_schedule_ack(tp); + inet_csk_schedule_ack(sk); - tcp_measure_rcv_mss(tp, skb); + tcp_measure_rcv_mss(sk, skb); tcp_rcv_rtt_measure(tp); now = tcp_time_stamp; - if (!tp->ack.ato) { + if (!icsk->icsk_ack.ato) { /* The _first_ data packet received, initialize * delayed ACK engine. */ - tcp_incr_quickack(tp); - tp->ack.ato = TCP_ATO_MIN; + tcp_incr_quickack(sk); + icsk->icsk_ack.ato = TCP_ATO_MIN; } else { - int m = now - tp->ack.lrcvtime; + int m = now - icsk->icsk_ack.lrcvtime; if (m <= TCP_ATO_MIN/2) { /* The fastest case is the first. */ - tp->ack.ato = (tp->ack.ato>>1) + TCP_ATO_MIN/2; - } else if (m < tp->ack.ato) { - tp->ack.ato = (tp->ack.ato>>1) + m; - if (tp->ack.ato > tp->rto) - tp->ack.ato = tp->rto; - } else if (m > tp->rto) { + icsk->icsk_ack.ato = (icsk->icsk_ack.ato >> 1) + TCP_ATO_MIN / 2; + } else if (m < icsk->icsk_ack.ato) { + icsk->icsk_ack.ato = (icsk->icsk_ack.ato >> 1) + m; + if (icsk->icsk_ack.ato > icsk->icsk_rto) + icsk->icsk_ack.ato = icsk->icsk_rto; + } else if (m > icsk->icsk_rto) { /* Too long gap. Apparently sender falled to * restart window, so that we send ACKs quickly. */ - tcp_incr_quickack(tp); + tcp_incr_quickack(sk); sk_stream_mem_reclaim(sk); } } - tp->ack.lrcvtime = now; + icsk->icsk_ack.lrcvtime = now; TCP_ECN_check_ce(tp, skb); @@ -543,8 +550,10 @@ static void tcp_event_data_recv(struct sock *sk, struct tcp_sock *tp, struct sk_ * To save cycles in the RFC 1323 implementation it was better to break * it up into three procedures. -- erics */ -static void tcp_rtt_estimator(struct tcp_sock *tp, __u32 mrtt, u32 *usrtt) +static void tcp_rtt_estimator(struct sock *sk, const __u32 mrtt, u32 *usrtt) { + struct tcp_sock *tp = tcp_sk(sk); + const struct inet_connection_sock *icsk = inet_csk(sk); long m = mrtt; /* RTT */ /* The following amusing code comes from Jacobson's @@ -604,15 +613,16 @@ static void tcp_rtt_estimator(struct tcp_sock *tp, __u32 mrtt, u32 *usrtt) tp->rtt_seq = tp->snd_nxt; } - if (tp->ca_ops->rtt_sample) - tp->ca_ops->rtt_sample(tp, *usrtt); + if (icsk->icsk_ca_ops->rtt_sample) + icsk->icsk_ca_ops->rtt_sample(sk, *usrtt); } /* Calculate rto without backoff. This is the second half of Van Jacobson's * routine referred to above. */ -static inline void tcp_set_rto(struct tcp_sock *tp) +static inline void tcp_set_rto(struct sock *sk) { + const struct tcp_sock *tp = tcp_sk(sk); /* Old crap is replaced with new one. 8) * * More seriously: @@ -623,7 +633,7 @@ static inline void tcp_set_rto(struct tcp_sock *tp) * is invisible. Actually, Linux-2.4 also generates erratic * ACKs in some curcumstances. */ - tp->rto = (tp->srtt >> 3) + tp->rttvar; + inet_csk(sk)->icsk_rto = (tp->srtt >> 3) + tp->rttvar; /* 2. Fixups made earlier cannot be right. * If we do not estimate RTO correctly without them, @@ -635,10 +645,10 @@ static inline void tcp_set_rto(struct tcp_sock *tp) /* NOTE: clamping at TCP_RTO_MIN is not required, current algo * guarantees that rto is higher. */ -static inline void tcp_bound_rto(struct tcp_sock *tp) +static inline void tcp_bound_rto(struct sock *sk) { - if (tp->rto > TCP_RTO_MAX) - tp->rto = TCP_RTO_MAX; + if (inet_csk(sk)->icsk_rto > TCP_RTO_MAX) + inet_csk(sk)->icsk_rto = TCP_RTO_MAX; } /* Save metrics learned by this TCP session. @@ -656,9 +666,10 @@ void tcp_update_metrics(struct sock *sk) dst_confirm(dst); if (dst && (dst->flags&DST_HOST)) { + const struct inet_connection_sock *icsk = inet_csk(sk); int m; - if (tp->backoff || !tp->srtt) { + if (icsk->icsk_backoff || !tp->srtt) { /* This session failed to estimate rtt. Why? * Probably, no packets returned in time. * Reset our results. @@ -707,7 +718,7 @@ void tcp_update_metrics(struct sock *sk) tp->snd_cwnd > dst_metric(dst, RTAX_CWND)) dst->metrics[RTAX_CWND-1] = tp->snd_cwnd; } else if (tp->snd_cwnd > tp->snd_ssthresh && - tp->ca_state == TCP_CA_Open) { + icsk->icsk_ca_state == TCP_CA_Open) { /* Cong. avoidance phase, cwnd is reliable. */ if (!dst_metric_locked(dst, RTAX_SSTHRESH)) dst->metrics[RTAX_SSTHRESH-1] = @@ -801,9 +812,9 @@ static void tcp_init_metrics(struct sock *sk) tp->mdev = dst_metric(dst, RTAX_RTTVAR); tp->mdev_max = tp->rttvar = max(tp->mdev, TCP_RTO_MIN); } - tcp_set_rto(tp); - tcp_bound_rto(tp); - if (tp->rto < TCP_TIMEOUT_INIT && !tp->rx_opt.saw_tstamp) + tcp_set_rto(sk); + tcp_bound_rto(sk); + if (inet_csk(sk)->icsk_rto < TCP_TIMEOUT_INIT && !tp->rx_opt.saw_tstamp) goto reset; tp->snd_cwnd = tcp_init_cwnd(tp, dst); tp->snd_cwnd_stamp = tcp_time_stamp; @@ -817,12 +828,14 @@ reset: if (!tp->rx_opt.saw_tstamp && tp->srtt) { tp->srtt = 0; tp->mdev = tp->mdev_max = tp->rttvar = TCP_TIMEOUT_INIT; - tp->rto = TCP_TIMEOUT_INIT; + inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT; } } -static void tcp_update_reordering(struct tcp_sock *tp, int metric, int ts) +static void tcp_update_reordering(struct sock *sk, const int metric, + const int ts) { + struct tcp_sock *tp = tcp_sk(sk); if (metric > tp->reordering) { tp->reordering = min(TCP_MAX_REORDERING, metric); @@ -837,7 +850,7 @@ static void tcp_update_reordering(struct tcp_sock *tp, int metric, int ts) NET_INC_STATS_BH(LINUX_MIB_TCPSACKREORDER); #if FASTRETRANS_DEBUG > 1 printk(KERN_DEBUG "Disorder%d %d %u f%u s%u rr%d\n", - tp->rx_opt.sack_ok, tp->ca_state, + tp->rx_opt.sack_ok, inet_csk(sk)->icsk_ca_state, tp->reordering, tp->fackets_out, tp->sacked_out, @@ -899,6 +912,7 @@ static void tcp_update_reordering(struct tcp_sock *tp, int metric, int ts) static int tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_una) { + const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); unsigned char *ptr = ack_skb->h.raw + TCP_SKB_CB(ack_skb)->sacked; struct tcp_sack_block *sp = (struct tcp_sack_block *)(ptr+2); @@ -1064,7 +1078,7 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ * we have to account for reordering! Ugly, * but should help. */ - if (lost_retrans && tp->ca_state == TCP_CA_Recovery) { + if (lost_retrans && icsk->icsk_ca_state == TCP_CA_Recovery) { struct sk_buff *skb; sk_stream_for_retrans_queue(skb, sk) { @@ -1093,8 +1107,8 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ tp->left_out = tp->sacked_out + tp->lost_out; - if ((reord < tp->fackets_out) && tp->ca_state != TCP_CA_Loss) - tcp_update_reordering(tp, ((tp->fackets_out + 1) - reord), 0); + if ((reord < tp->fackets_out) && icsk->icsk_ca_state != TCP_CA_Loss) + tcp_update_reordering(sk, ((tp->fackets_out + 1) - reord), 0); #if FASTRETRANS_DEBUG > 0 BUG_TRAP((int)tp->sacked_out >= 0); @@ -1111,17 +1125,18 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ */ void tcp_enter_frto(struct sock *sk) { + const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; tp->frto_counter = 1; - if (tp->ca_state <= TCP_CA_Disorder || + if (icsk->icsk_ca_state <= TCP_CA_Disorder || tp->snd_una == tp->high_seq || - (tp->ca_state == TCP_CA_Loss && !tp->retransmits)) { - tp->prior_ssthresh = tcp_current_ssthresh(tp); - tp->snd_ssthresh = tp->ca_ops->ssthresh(tp); - tcp_ca_event(tp, CA_EVENT_FRTO); + (icsk->icsk_ca_state == TCP_CA_Loss && !icsk->icsk_retransmits)) { + tp->prior_ssthresh = tcp_current_ssthresh(sk); + tp->snd_ssthresh = icsk->icsk_ca_ops->ssthresh(sk); + tcp_ca_event(sk, CA_EVENT_FRTO); } /* Have to clear retransmission markers here to keep the bookkeeping @@ -1138,7 +1153,7 @@ void tcp_enter_frto(struct sock *sk) } tcp_sync_left_out(tp); - tcp_set_ca_state(tp, TCP_CA_Open); + tcp_set_ca_state(sk, TCP_CA_Open); tp->frto_highmark = tp->snd_nxt; } @@ -1184,7 +1199,7 @@ static void tcp_enter_frto_loss(struct sock *sk) tp->reordering = min_t(unsigned int, tp->reordering, sysctl_tcp_reordering); - tcp_set_ca_state(tp, TCP_CA_Loss); + tcp_set_ca_state(sk, TCP_CA_Loss); tp->high_seq = tp->frto_highmark; TCP_ECN_queue_cwr(tp); } @@ -1208,16 +1223,17 @@ void tcp_clear_retrans(struct tcp_sock *tp) */ void tcp_enter_loss(struct sock *sk, int how) { + const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; int cnt = 0; /* Reduce ssthresh if it has not yet been made inside this window. */ - if (tp->ca_state <= TCP_CA_Disorder || tp->snd_una == tp->high_seq || - (tp->ca_state == TCP_CA_Loss && !tp->retransmits)) { - tp->prior_ssthresh = tcp_current_ssthresh(tp); - tp->snd_ssthresh = tp->ca_ops->ssthresh(tp); - tcp_ca_event(tp, CA_EVENT_LOSS); + if (icsk->icsk_ca_state <= TCP_CA_Disorder || tp->snd_una == tp->high_seq || + (icsk->icsk_ca_state == TCP_CA_Loss && !icsk->icsk_retransmits)) { + tp->prior_ssthresh = tcp_current_ssthresh(sk); + tp->snd_ssthresh = icsk->icsk_ca_ops->ssthresh(sk); + tcp_ca_event(sk, CA_EVENT_LOSS); } tp->snd_cwnd = 1; tp->snd_cwnd_cnt = 0; @@ -1248,12 +1264,12 @@ void tcp_enter_loss(struct sock *sk, int how) tp->reordering = min_t(unsigned int, tp->reordering, sysctl_tcp_reordering); - tcp_set_ca_state(tp, TCP_CA_Loss); + tcp_set_ca_state(sk, TCP_CA_Loss); tp->high_seq = tp->snd_nxt; TCP_ECN_queue_cwr(tp); } -static int tcp_check_sack_reneging(struct sock *sk, struct tcp_sock *tp) +static int tcp_check_sack_reneging(struct sock *sk) { struct sk_buff *skb; @@ -1265,12 +1281,14 @@ static int tcp_check_sack_reneging(struct sock *sk, struct tcp_sock *tp) */ if ((skb = skb_peek(&sk->sk_write_queue)) != NULL && (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)) { + struct inet_connection_sock *icsk = inet_csk(sk); NET_INC_STATS_BH(LINUX_MIB_TCPSACKRENEGING); tcp_enter_loss(sk, 1); - tp->retransmits++; + icsk->icsk_retransmits++; tcp_retransmit_skb(sk, skb_peek(&sk->sk_write_queue)); - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, tp->rto); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + icsk->icsk_rto, TCP_RTO_MAX); return 1; } return 0; @@ -1281,15 +1299,15 @@ static inline int tcp_fackets_out(struct tcp_sock *tp) return IsReno(tp) ? tp->sacked_out+1 : tp->fackets_out; } -static inline int tcp_skb_timedout(struct tcp_sock *tp, struct sk_buff *skb) +static inline int tcp_skb_timedout(struct sock *sk, struct sk_buff *skb) { - return (tcp_time_stamp - TCP_SKB_CB(skb)->when > tp->rto); + return (tcp_time_stamp - TCP_SKB_CB(skb)->when > inet_csk(sk)->icsk_rto); } static inline int tcp_head_timedout(struct sock *sk, struct tcp_sock *tp) { return tp->packets_out && - tcp_skb_timedout(tp, skb_peek(&sk->sk_write_queue)); + tcp_skb_timedout(sk, skb_peek(&sk->sk_write_queue)); } /* Linux NewReno/SACK/FACK/ECN state machine. @@ -1423,8 +1441,9 @@ static int tcp_time_to_recover(struct sock *sk, struct tcp_sock *tp) * in assumption of absent reordering, interpret this as reordering. * The only another reason could be bug in receiver TCP. */ -static void tcp_check_reno_reordering(struct tcp_sock *tp, int addend) +static void tcp_check_reno_reordering(struct sock *sk, const int addend) { + struct tcp_sock *tp = tcp_sk(sk); u32 holes; holes = max(tp->lost_out, 1U); @@ -1432,16 +1451,17 @@ static void tcp_check_reno_reordering(struct tcp_sock *tp, int addend) if ((tp->sacked_out + holes) > tp->packets_out) { tp->sacked_out = tp->packets_out - holes; - tcp_update_reordering(tp, tp->packets_out+addend, 0); + tcp_update_reordering(sk, tp->packets_out + addend, 0); } } /* Emulate SACKs for SACKless connection: account for a new dupack. */ -static void tcp_add_reno_sack(struct tcp_sock *tp) +static void tcp_add_reno_sack(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); tp->sacked_out++; - tcp_check_reno_reordering(tp, 0); + tcp_check_reno_reordering(sk, 0); tcp_sync_left_out(tp); } @@ -1456,7 +1476,7 @@ static void tcp_remove_reno_sacks(struct sock *sk, struct tcp_sock *tp, int acke else tp->sacked_out -= acked-1; } - tcp_check_reno_reordering(tp, acked); + tcp_check_reno_reordering(sk, acked); tcp_sync_left_out(tp); } @@ -1509,7 +1529,7 @@ static void tcp_update_scoreboard(struct sock *sk, struct tcp_sock *tp) struct sk_buff *skb; sk_stream_for_retrans_queue(skb, sk) { - if (tcp_skb_timedout(tp, skb) && + if (tcp_skb_timedout(sk, skb) && !(TCP_SKB_CB(skb)->sacked&TCPCB_TAGBITS)) { TCP_SKB_CB(skb)->sacked |= TCPCB_LOST; tp->lost_out += tcp_skb_pcount(skb); @@ -1530,14 +1550,16 @@ static inline void tcp_moderate_cwnd(struct tcp_sock *tp) } /* Decrease cwnd each second ack. */ -static void tcp_cwnd_down(struct tcp_sock *tp) +static void tcp_cwnd_down(struct sock *sk) { + const struct inet_connection_sock *icsk = inet_csk(sk); + struct tcp_sock *tp = tcp_sk(sk); int decr = tp->snd_cwnd_cnt + 1; tp->snd_cwnd_cnt = decr&1; decr >>= 1; - if (decr && tp->snd_cwnd > tp->ca_ops->min_cwnd(tp)) + if (decr && tp->snd_cwnd > icsk->icsk_ca_ops->min_cwnd(sk)) tp->snd_cwnd -= decr; tp->snd_cwnd = min(tp->snd_cwnd, tcp_packets_in_flight(tp)+1); @@ -1571,11 +1593,15 @@ static void DBGUNDO(struct sock *sk, struct tcp_sock *tp, const char *msg) #define DBGUNDO(x...) do { } while (0) #endif -static void tcp_undo_cwr(struct tcp_sock *tp, int undo) +static void tcp_undo_cwr(struct sock *sk, const int undo) { + struct tcp_sock *tp = tcp_sk(sk); + if (tp->prior_ssthresh) { - if (tp->ca_ops->undo_cwnd) - tp->snd_cwnd = tp->ca_ops->undo_cwnd(tp); + const struct inet_connection_sock *icsk = inet_csk(sk); + + if (icsk->icsk_ca_ops->undo_cwnd) + tp->snd_cwnd = icsk->icsk_ca_ops->undo_cwnd(sk); else tp->snd_cwnd = max(tp->snd_cwnd, tp->snd_ssthresh<<1); @@ -1603,9 +1629,9 @@ static int tcp_try_undo_recovery(struct sock *sk, struct tcp_sock *tp) /* Happy end! We did not retransmit anything * or our original transmission succeeded. */ - DBGUNDO(sk, tp, tp->ca_state == TCP_CA_Loss ? "loss" : "retrans"); - tcp_undo_cwr(tp, 1); - if (tp->ca_state == TCP_CA_Loss) + DBGUNDO(sk, tp, inet_csk(sk)->icsk_ca_state == TCP_CA_Loss ? "loss" : "retrans"); + tcp_undo_cwr(sk, 1); + if (inet_csk(sk)->icsk_ca_state == TCP_CA_Loss) NET_INC_STATS_BH(LINUX_MIB_TCPLOSSUNDO); else NET_INC_STATS_BH(LINUX_MIB_TCPFULLUNDO); @@ -1618,7 +1644,7 @@ static int tcp_try_undo_recovery(struct sock *sk, struct tcp_sock *tp) tcp_moderate_cwnd(tp); return 1; } - tcp_set_ca_state(tp, TCP_CA_Open); + tcp_set_ca_state(sk, TCP_CA_Open); return 0; } @@ -1627,7 +1653,7 @@ static void tcp_try_undo_dsack(struct sock *sk, struct tcp_sock *tp) { if (tp->undo_marker && !tp->undo_retrans) { DBGUNDO(sk, tp, "D-SACK"); - tcp_undo_cwr(tp, 1); + tcp_undo_cwr(sk, 1); tp->undo_marker = 0; NET_INC_STATS_BH(LINUX_MIB_TCPDSACKUNDO); } @@ -1648,10 +1674,10 @@ static int tcp_try_undo_partial(struct sock *sk, struct tcp_sock *tp, if (tp->retrans_out == 0) tp->retrans_stamp = 0; - tcp_update_reordering(tp, tcp_fackets_out(tp)+acked, 1); + tcp_update_reordering(sk, tcp_fackets_out(tp) + acked, 1); DBGUNDO(sk, tp, "Hoe"); - tcp_undo_cwr(tp, 0); + tcp_undo_cwr(sk, 0); NET_INC_STATS_BH(LINUX_MIB_TCPPARTIALUNDO); /* So... Do not make Hoe's retransmit yet. @@ -1674,22 +1700,23 @@ static int tcp_try_undo_loss(struct sock *sk, struct tcp_sock *tp) DBGUNDO(sk, tp, "partial loss"); tp->lost_out = 0; tp->left_out = tp->sacked_out; - tcp_undo_cwr(tp, 1); + tcp_undo_cwr(sk, 1); NET_INC_STATS_BH(LINUX_MIB_TCPLOSSUNDO); - tp->retransmits = 0; + inet_csk(sk)->icsk_retransmits = 0; tp->undo_marker = 0; if (!IsReno(tp)) - tcp_set_ca_state(tp, TCP_CA_Open); + tcp_set_ca_state(sk, TCP_CA_Open); return 1; } return 0; } -static inline void tcp_complete_cwr(struct tcp_sock *tp) +static inline void tcp_complete_cwr(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); tp->snd_cwnd = min(tp->snd_cwnd, tp->snd_ssthresh); tp->snd_cwnd_stamp = tcp_time_stamp; - tcp_ca_event(tp, CA_EVENT_COMPLETE_CWR); + tcp_ca_event(sk, CA_EVENT_COMPLETE_CWR); } static void tcp_try_to_open(struct sock *sk, struct tcp_sock *tp, int flag) @@ -1700,21 +1727,21 @@ static void tcp_try_to_open(struct sock *sk, struct tcp_sock *tp, int flag) tp->retrans_stamp = 0; if (flag&FLAG_ECE) - tcp_enter_cwr(tp); + tcp_enter_cwr(sk); - if (tp->ca_state != TCP_CA_CWR) { + if (inet_csk(sk)->icsk_ca_state != TCP_CA_CWR) { int state = TCP_CA_Open; if (tp->left_out || tp->retrans_out || tp->undo_marker) state = TCP_CA_Disorder; - if (tp->ca_state != state) { - tcp_set_ca_state(tp, state); + if (inet_csk(sk)->icsk_ca_state != state) { + tcp_set_ca_state(sk, state); tp->high_seq = tp->snd_nxt; } tcp_moderate_cwnd(tp); } else { - tcp_cwnd_down(tp); + tcp_cwnd_down(sk); } } @@ -1733,6 +1760,7 @@ static void tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, int prior_packets, int flag) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); int is_dupack = (tp->snd_una == prior_snd_una && !(flag&FLAG_NOT_DUP)); @@ -1750,13 +1778,13 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, tp->prior_ssthresh = 0; /* B. In all the states check for reneging SACKs. */ - if (tp->sacked_out && tcp_check_sack_reneging(sk, tp)) + if (tp->sacked_out && tcp_check_sack_reneging(sk)) return; /* C. Process data loss notification, provided it is valid. */ if ((flag&FLAG_DATA_LOST) && before(tp->snd_una, tp->high_seq) && - tp->ca_state != TCP_CA_Open && + icsk->icsk_ca_state != TCP_CA_Open && tp->fackets_out > tp->reordering) { tcp_mark_head_lost(sk, tp, tp->fackets_out-tp->reordering, tp->high_seq); NET_INC_STATS_BH(LINUX_MIB_TCPLOSS); @@ -1767,14 +1795,14 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, /* E. Check state exit conditions. State can be terminated * when high_seq is ACKed. */ - if (tp->ca_state == TCP_CA_Open) { + if (icsk->icsk_ca_state == TCP_CA_Open) { if (!sysctl_tcp_frto) BUG_TRAP(tp->retrans_out == 0); tp->retrans_stamp = 0; } else if (!before(tp->snd_una, tp->high_seq)) { - switch (tp->ca_state) { + switch (icsk->icsk_ca_state) { case TCP_CA_Loss: - tp->retransmits = 0; + icsk->icsk_retransmits = 0; if (tcp_try_undo_recovery(sk, tp)) return; break; @@ -1783,8 +1811,8 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, /* CWR is to be held something *above* high_seq * is ACKed for CWR bit to reach receiver. */ if (tp->snd_una != tp->high_seq) { - tcp_complete_cwr(tp); - tcp_set_ca_state(tp, TCP_CA_Open); + tcp_complete_cwr(sk); + tcp_set_ca_state(sk, TCP_CA_Open); } break; @@ -1795,7 +1823,7 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, * catching for all duplicate ACKs. */ IsReno(tp) || tp->snd_una != tp->high_seq) { tp->undo_marker = 0; - tcp_set_ca_state(tp, TCP_CA_Open); + tcp_set_ca_state(sk, TCP_CA_Open); } break; @@ -1804,17 +1832,17 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, tcp_reset_reno_sack(tp); if (tcp_try_undo_recovery(sk, tp)) return; - tcp_complete_cwr(tp); + tcp_complete_cwr(sk); break; } } /* F. Process state. */ - switch (tp->ca_state) { + switch (icsk->icsk_ca_state) { case TCP_CA_Recovery: if (prior_snd_una == tp->snd_una) { if (IsReno(tp) && is_dupack) - tcp_add_reno_sack(tp); + tcp_add_reno_sack(sk); } else { int acked = prior_packets - tp->packets_out; if (IsReno(tp)) @@ -1824,13 +1852,13 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, break; case TCP_CA_Loss: if (flag&FLAG_DATA_ACKED) - tp->retransmits = 0; + icsk->icsk_retransmits = 0; if (!tcp_try_undo_loss(sk, tp)) { tcp_moderate_cwnd(tp); tcp_xmit_retransmit_queue(sk); return; } - if (tp->ca_state != TCP_CA_Open) + if (icsk->icsk_ca_state != TCP_CA_Open) return; /* Loss is undone; fall through to processing in Open state. */ default: @@ -1838,10 +1866,10 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, if (tp->snd_una != prior_snd_una) tcp_reset_reno_sack(tp); if (is_dupack) - tcp_add_reno_sack(tp); + tcp_add_reno_sack(sk); } - if (tp->ca_state == TCP_CA_Disorder) + if (icsk->icsk_ca_state == TCP_CA_Disorder) tcp_try_undo_dsack(sk, tp); if (!tcp_time_to_recover(sk, tp)) { @@ -1861,30 +1889,28 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, tp->undo_marker = tp->snd_una; tp->undo_retrans = tp->retrans_out; - if (tp->ca_state < TCP_CA_CWR) { + if (icsk->icsk_ca_state < TCP_CA_CWR) { if (!(flag&FLAG_ECE)) - tp->prior_ssthresh = tcp_current_ssthresh(tp); - tp->snd_ssthresh = tp->ca_ops->ssthresh(tp); + tp->prior_ssthresh = tcp_current_ssthresh(sk); + tp->snd_ssthresh = icsk->icsk_ca_ops->ssthresh(sk); TCP_ECN_queue_cwr(tp); } tp->snd_cwnd_cnt = 0; - tcp_set_ca_state(tp, TCP_CA_Recovery); + tcp_set_ca_state(sk, TCP_CA_Recovery); } if (is_dupack || tcp_head_timedout(sk, tp)) tcp_update_scoreboard(sk, tp); - tcp_cwnd_down(tp); + tcp_cwnd_down(sk); tcp_xmit_retransmit_queue(sk); } /* Read draft-ietf-tcplw-high-performance before mucking * with this code. (Superceeds RFC1323) */ -static void tcp_ack_saw_tstamp(struct tcp_sock *tp, u32 *usrtt, int flag) +static void tcp_ack_saw_tstamp(struct sock *sk, u32 *usrtt, int flag) { - __u32 seq_rtt; - /* RTTM Rule: A TSecr value received in a segment is used to * update the averaged RTT measurement only if the segment * acknowledges some new data, i.e., only if it advances the @@ -1900,14 +1926,15 @@ static void tcp_ack_saw_tstamp(struct tcp_sock *tp, u32 *usrtt, int flag) * answer arrives rto becomes 120 seconds! If at least one of segments * in window is lost... Voila. --ANK (010210) */ - seq_rtt = tcp_time_stamp - tp->rx_opt.rcv_tsecr; - tcp_rtt_estimator(tp, seq_rtt, usrtt); - tcp_set_rto(tp); - tp->backoff = 0; - tcp_bound_rto(tp); + struct tcp_sock *tp = tcp_sk(sk); + const __u32 seq_rtt = tcp_time_stamp - tp->rx_opt.rcv_tsecr; + tcp_rtt_estimator(sk, seq_rtt, usrtt); + tcp_set_rto(sk); + inet_csk(sk)->icsk_backoff = 0; + tcp_bound_rto(sk); } -static void tcp_ack_no_tstamp(struct tcp_sock *tp, u32 seq_rtt, u32 *usrtt, int flag) +static void tcp_ack_no_tstamp(struct sock *sk, u32 seq_rtt, u32 *usrtt, int flag) { /* We don't have a timestamp. Can only use * packets that are not retransmitted to determine @@ -1921,27 +1948,29 @@ static void tcp_ack_no_tstamp(struct tcp_sock *tp, u32 seq_rtt, u32 *usrtt, int if (flag & FLAG_RETRANS_DATA_ACKED) return; - tcp_rtt_estimator(tp, seq_rtt, usrtt); - tcp_set_rto(tp); - tp->backoff = 0; - tcp_bound_rto(tp); + tcp_rtt_estimator(sk, seq_rtt, usrtt); + tcp_set_rto(sk); + inet_csk(sk)->icsk_backoff = 0; + tcp_bound_rto(sk); } -static inline void tcp_ack_update_rtt(struct tcp_sock *tp, - int flag, s32 seq_rtt, u32 *usrtt) +static inline void tcp_ack_update_rtt(struct sock *sk, const int flag, + const s32 seq_rtt, u32 *usrtt) { + const struct tcp_sock *tp = tcp_sk(sk); /* Note that peer MAY send zero echo. In this case it is ignored. (rfc1323) */ if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr) - tcp_ack_saw_tstamp(tp, usrtt, flag); + tcp_ack_saw_tstamp(sk, usrtt, flag); else if (seq_rtt >= 0) - tcp_ack_no_tstamp(tp, seq_rtt, usrtt, flag); + tcp_ack_no_tstamp(sk, seq_rtt, usrtt, flag); } -static inline void tcp_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, +static inline void tcp_cong_avoid(struct sock *sk, u32 ack, u32 rtt, u32 in_flight, int good) { - tp->ca_ops->cong_avoid(tp, ack, rtt, in_flight, good); - tp->snd_cwnd_stamp = tcp_time_stamp; + const struct inet_connection_sock *icsk = inet_csk(sk); + icsk->icsk_ca_ops->cong_avoid(sk, ack, rtt, in_flight, good); + tcp_sk(sk)->snd_cwnd_stamp = tcp_time_stamp; } /* Restart timer after forward progress on connection. @@ -1951,9 +1980,9 @@ static inline void tcp_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, static inline void tcp_ack_packets_out(struct sock *sk, struct tcp_sock *tp) { if (!tp->packets_out) { - tcp_clear_xmit_timer(sk, TCP_TIME_RETRANS); + inet_csk_clear_xmit_timer(sk, ICSK_TIME_RETRANS); } else { - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, tp->rto); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, inet_csk(sk)->icsk_rto, TCP_RTO_MAX); } } @@ -2068,9 +2097,13 @@ static int tcp_clean_rtx_queue(struct sock *sk, __s32 *seq_rtt_p, s32 *seq_usrtt seq_rtt = -1; } else if (seq_rtt < 0) seq_rtt = now - scb->when; - if (seq_usrtt) - *seq_usrtt = (usnow.tv_sec - skb->stamp.tv_sec) * 1000000 - + (usnow.tv_usec - skb->stamp.tv_usec); + if (seq_usrtt) { + struct timeval tv; + + skb_get_timestamp(skb, &tv); + *seq_usrtt = (usnow.tv_sec - tv.tv_sec) * 1000000 + + (usnow.tv_usec - tv.tv_usec); + } if (sacked & TCPCB_SACKED_ACKED) tp->sacked_out -= tcp_skb_pcount(skb); @@ -2085,16 +2118,17 @@ static int tcp_clean_rtx_queue(struct sock *sk, __s32 *seq_rtt_p, s32 *seq_usrtt seq_rtt = now - scb->when; tcp_dec_pcount_approx(&tp->fackets_out, skb); tcp_packets_out_dec(tp, skb); - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &sk->sk_write_queue); sk_stream_free_skb(sk, skb); } if (acked&FLAG_ACKED) { - tcp_ack_update_rtt(tp, acked, seq_rtt, seq_usrtt); + const struct inet_connection_sock *icsk = inet_csk(sk); + tcp_ack_update_rtt(sk, acked, seq_rtt, seq_usrtt); tcp_ack_packets_out(sk, tp); - if (tp->ca_ops->pkts_acked) - tp->ca_ops->pkts_acked(tp, pkts_acked); + if (icsk->icsk_ca_ops->pkts_acked) + icsk->icsk_ca_ops->pkts_acked(sk, pkts_acked); } #if FASTRETRANS_DEBUG > 0 @@ -2102,19 +2136,20 @@ static int tcp_clean_rtx_queue(struct sock *sk, __s32 *seq_rtt_p, s32 *seq_usrtt BUG_TRAP((int)tp->lost_out >= 0); BUG_TRAP((int)tp->retrans_out >= 0); if (!tp->packets_out && tp->rx_opt.sack_ok) { + const struct inet_connection_sock *icsk = inet_csk(sk); if (tp->lost_out) { printk(KERN_DEBUG "Leak l=%u %d\n", - tp->lost_out, tp->ca_state); + tp->lost_out, icsk->icsk_ca_state); tp->lost_out = 0; } if (tp->sacked_out) { printk(KERN_DEBUG "Leak s=%u %d\n", - tp->sacked_out, tp->ca_state); + tp->sacked_out, icsk->icsk_ca_state); tp->sacked_out = 0; } if (tp->retrans_out) { printk(KERN_DEBUG "Leak r=%u %d\n", - tp->retrans_out, tp->ca_state); + tp->retrans_out, icsk->icsk_ca_state); tp->retrans_out = 0; } } @@ -2125,40 +2160,43 @@ static int tcp_clean_rtx_queue(struct sock *sk, __s32 *seq_rtt_p, s32 *seq_usrtt static void tcp_ack_probe(struct sock *sk) { - struct tcp_sock *tp = tcp_sk(sk); + const struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); /* Was it a usable window open? */ if (!after(TCP_SKB_CB(sk->sk_send_head)->end_seq, tp->snd_una + tp->snd_wnd)) { - tp->backoff = 0; - tcp_clear_xmit_timer(sk, TCP_TIME_PROBE0); + icsk->icsk_backoff = 0; + inet_csk_clear_xmit_timer(sk, ICSK_TIME_PROBE0); /* Socket must be waked up by subsequent tcp_data_snd_check(). * This function is not for random using! */ } else { - tcp_reset_xmit_timer(sk, TCP_TIME_PROBE0, - min(tp->rto << tp->backoff, TCP_RTO_MAX)); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, + min(icsk->icsk_rto << icsk->icsk_backoff, TCP_RTO_MAX), + TCP_RTO_MAX); } } -static inline int tcp_ack_is_dubious(struct tcp_sock *tp, int flag) +static inline int tcp_ack_is_dubious(const struct sock *sk, const int flag) { return (!(flag & FLAG_NOT_DUP) || (flag & FLAG_CA_ALERT) || - tp->ca_state != TCP_CA_Open); + inet_csk(sk)->icsk_ca_state != TCP_CA_Open); } -static inline int tcp_may_raise_cwnd(struct tcp_sock *tp, int flag) +static inline int tcp_may_raise_cwnd(const struct sock *sk, const int flag) { + const struct tcp_sock *tp = tcp_sk(sk); return (!(flag & FLAG_ECE) || tp->snd_cwnd < tp->snd_ssthresh) && - !((1<<tp->ca_state)&(TCPF_CA_Recovery|TCPF_CA_CWR)); + !((1 << inet_csk(sk)->icsk_ca_state) & (TCPF_CA_Recovery | TCPF_CA_CWR)); } /* Check that window update is acceptable. * The function assumes that snd_una<=ack<=snd_next. */ -static inline int tcp_may_update_window(struct tcp_sock *tp, u32 ack, - u32 ack_seq, u32 nwin) +static inline int tcp_may_update_window(const struct tcp_sock *tp, const u32 ack, + const u32 ack_seq, const u32 nwin) { return (after(ack, tp->snd_una) || after(ack_seq, tp->snd_wl1) || @@ -2241,6 +2279,7 @@ static void tcp_process_frto(struct sock *sk, u32 prior_snd_una) /* This routine deals with incoming acks, but not outgoing ones. */ static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); u32 prior_snd_una = tp->snd_una; u32 ack_seq = TCP_SKB_CB(skb)->seq; @@ -2268,7 +2307,7 @@ static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) tp->snd_una = ack; flag |= FLAG_WIN_UPDATE; - tcp_ca_event(tp, CA_EVENT_FAST_ACK); + tcp_ca_event(sk, CA_EVENT_FAST_ACK); NET_INC_STATS_BH(LINUX_MIB_TCPHPACKS); } else { @@ -2285,7 +2324,7 @@ static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) if (TCP_ECN_rcv_ecn_echo(tp, skb->h.th)) flag |= FLAG_ECE; - tcp_ca_event(tp, CA_EVENT_SLOW_ACK); + tcp_ca_event(sk, CA_EVENT_SLOW_ACK); } /* We passed data and got it acked, remove any soft error @@ -2301,19 +2340,19 @@ static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) /* See if we can take anything off of the retransmit queue. */ flag |= tcp_clean_rtx_queue(sk, &seq_rtt, - tp->ca_ops->rtt_sample ? &seq_usrtt : NULL); + icsk->icsk_ca_ops->rtt_sample ? &seq_usrtt : NULL); if (tp->frto_counter) tcp_process_frto(sk, prior_snd_una); - if (tcp_ack_is_dubious(tp, flag)) { + if (tcp_ack_is_dubious(sk, flag)) { /* Advanve CWND, if state allows this. */ - if ((flag & FLAG_DATA_ACKED) && tcp_may_raise_cwnd(tp, flag)) - tcp_cong_avoid(tp, ack, seq_rtt, prior_in_flight, 0); + if ((flag & FLAG_DATA_ACKED) && tcp_may_raise_cwnd(sk, flag)) + tcp_cong_avoid(sk, ack, seq_rtt, prior_in_flight, 0); tcp_fastretrans_alert(sk, prior_snd_una, prior_packets, flag); } else { if ((flag & FLAG_DATA_ACKED)) - tcp_cong_avoid(tp, ack, seq_rtt, prior_in_flight, 1); + tcp_cong_avoid(sk, ack, seq_rtt, prior_in_flight, 1); } if ((flag & FLAG_FORWARD_PROGRESS) || !(flag&FLAG_NOT_DUP)) @@ -2322,7 +2361,7 @@ static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) return 1; no_queue: - tp->probes_out = 0; + icsk->icsk_probes_out = 0; /* If this ack opens up a zero window, clear backoff. It was * being used to time the probes, and is probably far higher than @@ -2500,8 +2539,9 @@ static inline void tcp_replace_ts_recent(struct tcp_sock *tp, u32 seq) * up to bandwidth of 18Gigabit/sec. 8) ] */ -static int tcp_disordered_ack(struct tcp_sock *tp, struct sk_buff *skb) +static int tcp_disordered_ack(const struct sock *sk, const struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); struct tcphdr *th = skb->h.th; u32 seq = TCP_SKB_CB(skb)->seq; u32 ack = TCP_SKB_CB(skb)->ack_seq; @@ -2516,14 +2556,15 @@ static int tcp_disordered_ack(struct tcp_sock *tp, struct sk_buff *skb) !tcp_may_update_window(tp, ack, seq, ntohs(th->window) << tp->rx_opt.snd_wscale) && /* 4. ... and sits in replay window. */ - (s32)(tp->rx_opt.ts_recent - tp->rx_opt.rcv_tsval) <= (tp->rto*1024)/HZ); + (s32)(tp->rx_opt.ts_recent - tp->rx_opt.rcv_tsval) <= (inet_csk(sk)->icsk_rto * 1024) / HZ); } -static inline int tcp_paws_discard(struct tcp_sock *tp, struct sk_buff *skb) +static inline int tcp_paws_discard(const struct sock *sk, const struct sk_buff *skb) { + const struct tcp_sock *tp = tcp_sk(sk); return ((s32)(tp->rx_opt.ts_recent - tp->rx_opt.rcv_tsval) > TCP_PAWS_WINDOW && xtime.tv_sec < tp->rx_opt.ts_recent_stamp + TCP_PAWS_24DAYS && - !tcp_disordered_ack(tp, skb)); + !tcp_disordered_ack(sk, skb)); } /* Check segment sequence number for validity. @@ -2586,7 +2627,7 @@ static void tcp_fin(struct sk_buff *skb, struct sock *sk, struct tcphdr *th) { struct tcp_sock *tp = tcp_sk(sk); - tcp_schedule_ack(tp); + inet_csk_schedule_ack(sk); sk->sk_shutdown |= RCV_SHUTDOWN; sock_set_flag(sk, SOCK_DONE); @@ -2596,7 +2637,7 @@ static void tcp_fin(struct sk_buff *skb, struct sock *sk, struct tcphdr *th) case TCP_ESTABLISHED: /* Move to CLOSE_WAIT */ tcp_set_state(sk, TCP_CLOSE_WAIT); - tp->ack.pingpong = 1; + inet_csk(sk)->icsk_ack.pingpong = 1; break; case TCP_CLOSE_WAIT: @@ -2694,7 +2735,7 @@ static void tcp_send_dupack(struct sock *sk, struct sk_buff *skb) if (TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq && before(TCP_SKB_CB(skb)->seq, tp->rcv_nxt)) { NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKLOST); - tcp_enter_quickack_mode(tp); + tcp_enter_quickack_mode(sk); if (tp->rx_opt.sack_ok && sysctl_tcp_dsack) { u32 end_seq = TCP_SKB_CB(skb)->end_seq; @@ -2853,7 +2894,7 @@ static void tcp_ofo_queue(struct sock *sk) if (!after(TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt)) { SOCK_DEBUG(sk, "ofo packet was already received \n"); - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &tp->out_of_order_queue); __kfree_skb(skb); continue; } @@ -2861,7 +2902,7 @@ static void tcp_ofo_queue(struct sock *sk) tp->rcv_nxt, TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq); - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &tp->out_of_order_queue); __skb_queue_tail(&sk->sk_receive_queue, skb); tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq; if(skb->h.th->fin) @@ -2942,7 +2983,7 @@ queue_and_out: * gap in queue is filled. */ if (skb_queue_empty(&tp->out_of_order_queue)) - tp->ack.pingpong = 0; + inet_csk(sk)->icsk_ack.pingpong = 0; } if (tp->rx_opt.num_sacks) @@ -2963,8 +3004,8 @@ queue_and_out: tcp_dsack_set(tp, TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq); out_of_window: - tcp_enter_quickack_mode(tp); - tcp_schedule_ack(tp); + tcp_enter_quickack_mode(sk); + inet_csk_schedule_ack(sk); drop: __kfree_skb(skb); return; @@ -2974,7 +3015,7 @@ drop: if (!before(TCP_SKB_CB(skb)->seq, tp->rcv_nxt + tcp_receive_window(tp))) goto out_of_window; - tcp_enter_quickack_mode(tp); + tcp_enter_quickack_mode(sk); if (before(TCP_SKB_CB(skb)->seq, tp->rcv_nxt)) { /* Partial packet, seq < rcv_next < end_seq */ @@ -3003,7 +3044,7 @@ drop: /* Disable header prediction. */ tp->pred_flags = 0; - tcp_schedule_ack(tp); + inet_csk_schedule_ack(sk); SOCK_DEBUG(sk, "out of order segment: rcv_next %X seq %X - %X\n", tp->rcv_nxt, TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq); @@ -3027,7 +3068,7 @@ drop: u32 end_seq = TCP_SKB_CB(skb)->end_seq; if (seq == TCP_SKB_CB(skb1)->end_seq) { - __skb_append(skb1, skb); + __skb_append(skb1, skb, &tp->out_of_order_queue); if (!tp->rx_opt.num_sacks || tp->selective_acks[0].end_seq != seq) @@ -3071,7 +3112,7 @@ drop: tcp_dsack_extend(tp, TCP_SKB_CB(skb1)->seq, end_seq); break; } - __skb_unlink(skb1, skb1->list); + __skb_unlink(skb1, &tp->out_of_order_queue); tcp_dsack_extend(tp, TCP_SKB_CB(skb1)->seq, TCP_SKB_CB(skb1)->end_seq); __kfree_skb(skb1); } @@ -3088,8 +3129,9 @@ add_sack: * simplifies code) */ static void -tcp_collapse(struct sock *sk, struct sk_buff *head, - struct sk_buff *tail, u32 start, u32 end) +tcp_collapse(struct sock *sk, struct sk_buff_head *list, + struct sk_buff *head, struct sk_buff *tail, + u32 start, u32 end) { struct sk_buff *skb; @@ -3099,7 +3141,7 @@ tcp_collapse(struct sock *sk, struct sk_buff *head, /* No new bits? It is possible on ofo queue. */ if (!before(start, TCP_SKB_CB(skb)->end_seq)) { struct sk_buff *next = skb->next; - __skb_unlink(skb, skb->list); + __skb_unlink(skb, list); __kfree_skb(skb); NET_INC_STATS_BH(LINUX_MIB_TCPRCVCOLLAPSED); skb = next; @@ -3145,7 +3187,7 @@ tcp_collapse(struct sock *sk, struct sk_buff *head, nskb->mac.raw = nskb->head + (skb->mac.raw-skb->head); memcpy(nskb->cb, skb->cb, sizeof(skb->cb)); TCP_SKB_CB(nskb)->seq = TCP_SKB_CB(nskb)->end_seq = start; - __skb_insert(nskb, skb->prev, skb, skb->list); + __skb_insert(nskb, skb->prev, skb, list); sk_stream_set_owner_r(nskb, sk); /* Copy data, releasing collapsed skbs. */ @@ -3164,7 +3206,7 @@ tcp_collapse(struct sock *sk, struct sk_buff *head, } if (!before(start, TCP_SKB_CB(skb)->end_seq)) { struct sk_buff *next = skb->next; - __skb_unlink(skb, skb->list); + __skb_unlink(skb, list); __kfree_skb(skb); NET_INC_STATS_BH(LINUX_MIB_TCPRCVCOLLAPSED); skb = next; @@ -3200,7 +3242,8 @@ static void tcp_collapse_ofo_queue(struct sock *sk) if (skb == (struct sk_buff *)&tp->out_of_order_queue || after(TCP_SKB_CB(skb)->seq, end) || before(TCP_SKB_CB(skb)->end_seq, start)) { - tcp_collapse(sk, head, skb, start, end); + tcp_collapse(sk, &tp->out_of_order_queue, + head, skb, start, end); head = skb; if (skb == (struct sk_buff *)&tp->out_of_order_queue) break; @@ -3237,7 +3280,8 @@ static int tcp_prune_queue(struct sock *sk) tp->rcv_ssthresh = min(tp->rcv_ssthresh, 4U * tp->advmss); tcp_collapse_ofo_queue(sk); - tcp_collapse(sk, sk->sk_receive_queue.next, + tcp_collapse(sk, &sk->sk_receive_queue, + sk->sk_receive_queue.next, (struct sk_buff*)&sk->sk_receive_queue, tp->copied_seq, tp->rcv_nxt); sk_stream_mem_reclaim(sk); @@ -3286,12 +3330,12 @@ void tcp_cwnd_application_limited(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); - if (tp->ca_state == TCP_CA_Open && + if (inet_csk(sk)->icsk_ca_state == TCP_CA_Open && sk->sk_socket && !test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) { /* Limited by application or receiver window. */ u32 win_used = max(tp->snd_cwnd_used, 2U); if (win_used < tp->snd_cwnd) { - tp->snd_ssthresh = tcp_current_ssthresh(tp); + tp->snd_ssthresh = tcp_current_ssthresh(sk); tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1; } tp->snd_cwnd_used = 0; @@ -3370,13 +3414,13 @@ static void __tcp_ack_snd_check(struct sock *sk, int ofo_possible) struct tcp_sock *tp = tcp_sk(sk); /* More than one full frame received... */ - if (((tp->rcv_nxt - tp->rcv_wup) > tp->ack.rcv_mss + if (((tp->rcv_nxt - tp->rcv_wup) > inet_csk(sk)->icsk_ack.rcv_mss /* ... and right edge of window advances far enough. * (tcp_recvmsg() will send ACK otherwise). Or... */ && __tcp_select_window(sk) >= tp->rcv_wnd) || /* We ACK each frame or... */ - tcp_in_quickack_mode(tp) || + tcp_in_quickack_mode(sk) || /* We have out of order data. */ (ofo_possible && skb_peek(&tp->out_of_order_queue))) { @@ -3390,8 +3434,7 @@ static void __tcp_ack_snd_check(struct sock *sk, int ofo_possible) static __inline__ void tcp_ack_snd_check(struct sock *sk) { - struct tcp_sock *tp = tcp_sk(sk); - if (!tcp_ack_scheduled(tp)) { + if (!inet_csk_ack_scheduled(sk)) { /* We sent a data segment already. */ return; } @@ -3462,7 +3505,7 @@ static void tcp_check_urg(struct sock * sk, struct tcphdr * th) struct sk_buff *skb = skb_peek(&sk->sk_receive_queue); tp->copied_seq++; if (skb && !before(tp->copied_seq, TCP_SKB_CB(skb)->end_seq)) { - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &sk->sk_receive_queue); __kfree_skb(skb); } } @@ -3645,7 +3688,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, tp->rcv_nxt == tp->rcv_wup) tcp_store_ts_recent(tp); - tcp_rcv_rtt_measure_ts(tp, skb); + tcp_rcv_rtt_measure_ts(sk, skb); /* We know that such packets are checksummed * on entry. @@ -3678,7 +3721,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, tp->rcv_nxt == tp->rcv_wup) tcp_store_ts_recent(tp); - tcp_rcv_rtt_measure_ts(tp, skb); + tcp_rcv_rtt_measure_ts(sk, skb); __skb_pull(skb, tcp_header_len); tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq; @@ -3699,7 +3742,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, tp->rcv_nxt == tp->rcv_wup) tcp_store_ts_recent(tp); - tcp_rcv_rtt_measure_ts(tp, skb); + tcp_rcv_rtt_measure_ts(sk, skb); if ((int)skb->truesize > sk->sk_forward_alloc) goto step5; @@ -3719,7 +3762,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, /* Well, only one small jumplet in fast path... */ tcp_ack(sk, skb, FLAG_DATA); tcp_data_snd_check(sk, tp); - if (!tcp_ack_scheduled(tp)) + if (!inet_csk_ack_scheduled(sk)) goto no_ack; } @@ -3741,7 +3784,7 @@ slow_path: * RFC1323: H1. Apply PAWS check first. */ if (tcp_fast_parse_options(skb, th, tp) && tp->rx_opt.saw_tstamp && - tcp_paws_discard(tp, skb)) { + tcp_paws_discard(sk, skb)) { if (!th->rst) { NET_INC_STATS_BH(LINUX_MIB_PAWSESTABREJECTED); tcp_send_dupack(sk, skb); @@ -3788,7 +3831,7 @@ step5: if(th->ack) tcp_ack(sk, skb, FLAG_SLOWPATH); - tcp_rcv_rtt_measure_ts(tp, skb); + tcp_rcv_rtt_measure_ts(sk, skb); /* Process urgent data. */ tcp_urg(sk, skb, th); @@ -3817,6 +3860,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, tcp_parse_options(skb, &tp->rx_opt, 0); if (th->ack) { + struct inet_connection_sock *icsk; /* rfc793: * "If the state is SYN-SENT then * first check the ACK bit @@ -3920,7 +3964,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, tcp_init_metrics(sk); - tcp_init_congestion_control(tp); + tcp_init_congestion_control(sk); /* Prevent spurious tcp_cwnd_restart() on first data * packet. @@ -3930,7 +3974,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, tcp_init_buffer_space(sk); if (sock_flag(sk, SOCK_KEEPOPEN)) - tcp_reset_keepalive_timer(sk, keepalive_time_when(tp)); + inet_csk_reset_keepalive_timer(sk, keepalive_time_when(tp)); if (!tp->rx_opt.snd_wscale) __tcp_fast_path_on(tp, tp->snd_wnd); @@ -3942,7 +3986,11 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, sk_wake_async(sk, 0, POLL_OUT); } - if (sk->sk_write_pending || tp->defer_accept || tp->ack.pingpong) { + icsk = inet_csk(sk); + + if (sk->sk_write_pending || + icsk->icsk_accept_queue.rskq_defer_accept || + icsk->icsk_ack.pingpong) { /* Save one ACK. Data will be ready after * several ticks, if write_pending is set. * @@ -3950,12 +3998,13 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, * look so _wonderfully_ clever, that I was not able * to stand against the temptation 8) --ANK */ - tcp_schedule_ack(tp); - tp->ack.lrcvtime = tcp_time_stamp; - tp->ack.ato = TCP_ATO_MIN; - tcp_incr_quickack(tp); - tcp_enter_quickack_mode(tp); - tcp_reset_xmit_timer(sk, TCP_TIME_DACK, TCP_DELACK_MAX); + inet_csk_schedule_ack(sk); + icsk->icsk_ack.lrcvtime = tcp_time_stamp; + icsk->icsk_ack.ato = TCP_ATO_MIN; + tcp_incr_quickack(sk); + tcp_enter_quickack_mode(sk); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + TCP_DELACK_MAX, TCP_RTO_MAX); discard: __kfree_skb(skb); @@ -4111,7 +4160,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, } if (tcp_fast_parse_options(skb, th, tp) && tp->rx_opt.saw_tstamp && - tcp_paws_discard(tp, skb)) { + tcp_paws_discard(sk, skb)) { if (!th->rst) { NET_INC_STATS_BH(LINUX_MIB_PAWSESTABREJECTED); tcp_send_dupack(sk, skb); @@ -4180,7 +4229,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, */ if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr && !tp->srtt) - tcp_ack_saw_tstamp(tp, 0, 0); + tcp_ack_saw_tstamp(sk, NULL, 0); if (tp->rx_opt.tstamp_ok) tp->advmss -= TCPOLEN_TSTAMP_ALIGNED; @@ -4192,7 +4241,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, tcp_init_metrics(sk); - tcp_init_congestion_control(tp); + tcp_init_congestion_control(sk); /* Prevent spurious tcp_cwnd_restart() on * first data packet. @@ -4227,9 +4276,9 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, return 1; } - tmo = tcp_fin_time(tp); + tmo = tcp_fin_time(sk); if (tmo > TCP_TIMEWAIT_LEN) { - tcp_reset_keepalive_timer(sk, tmo - TCP_TIMEWAIT_LEN); + inet_csk_reset_keepalive_timer(sk, tmo - TCP_TIMEWAIT_LEN); } else if (th->fin || sock_owned_by_user(sk)) { /* Bad case. We could lose such FIN otherwise. * It is not a big problem, but it looks confusing @@ -4237,7 +4286,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, * if it spins in bh_lock_sock(), but it is really * marginal case. */ - tcp_reset_keepalive_timer(sk, tmo); + inet_csk_reset_keepalive_timer(sk, tmo); } else { tcp_time_wait(sk, TCP_FIN_WAIT2, tmo); goto discard; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 67c670886c1f..13dfb391cdf1 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -64,7 +64,9 @@ #include <linux/times.h> #include <net/icmp.h> +#include <net/inet_hashtables.h> #include <net/tcp.h> +#include <net/transp_v6.h> #include <net/ipv6.h> #include <net/inet_common.h> #include <net/xfrm.h> @@ -75,7 +77,6 @@ #include <linux/proc_fs.h> #include <linux/seq_file.h> -extern int sysctl_ip_dynaddr; int sysctl_tcp_tw_reuse; int sysctl_tcp_low_latency; @@ -88,463 +89,29 @@ static struct socket *tcp_socket; void tcp_v4_send_check(struct sock *sk, struct tcphdr *th, int len, struct sk_buff *skb); -struct tcp_hashinfo __cacheline_aligned tcp_hashinfo = { - .__tcp_lhash_lock = RW_LOCK_UNLOCKED, - .__tcp_lhash_users = ATOMIC_INIT(0), - .__tcp_lhash_wait - = __WAIT_QUEUE_HEAD_INITIALIZER(tcp_hashinfo.__tcp_lhash_wait), - .__tcp_portalloc_lock = SPIN_LOCK_UNLOCKED +struct inet_hashinfo __cacheline_aligned tcp_hashinfo = { + .lhash_lock = RW_LOCK_UNLOCKED, + .lhash_users = ATOMIC_INIT(0), + .lhash_wait = __WAIT_QUEUE_HEAD_INITIALIZER(tcp_hashinfo.lhash_wait), + .portalloc_lock = SPIN_LOCK_UNLOCKED, + .port_rover = 1024 - 1, }; -/* - * This array holds the first and last local port number. - * For high-usage systems, use sysctl to change this to - * 32768-61000 - */ -int sysctl_local_port_range[2] = { 1024, 4999 }; -int tcp_port_rover = 1024 - 1; - -static __inline__ int tcp_hashfn(__u32 laddr, __u16 lport, - __u32 faddr, __u16 fport) -{ - int h = (laddr ^ lport) ^ (faddr ^ fport); - h ^= h >> 16; - h ^= h >> 8; - return h & (tcp_ehash_size - 1); -} - -static __inline__ int tcp_sk_hashfn(struct sock *sk) -{ - struct inet_sock *inet = inet_sk(sk); - __u32 laddr = inet->rcv_saddr; - __u16 lport = inet->num; - __u32 faddr = inet->daddr; - __u16 fport = inet->dport; - - return tcp_hashfn(laddr, lport, faddr, fport); -} - -/* Allocate and initialize a new TCP local port bind bucket. - * The bindhash mutex for snum's hash chain must be held here. - */ -struct tcp_bind_bucket *tcp_bucket_create(struct tcp_bind_hashbucket *head, - unsigned short snum) -{ - struct tcp_bind_bucket *tb = kmem_cache_alloc(tcp_bucket_cachep, - SLAB_ATOMIC); - if (tb) { - tb->port = snum; - tb->fastreuse = 0; - INIT_HLIST_HEAD(&tb->owners); - hlist_add_head(&tb->node, &head->chain); - } - return tb; -} - -/* Caller must hold hashbucket lock for this tb with local BH disabled */ -void tcp_bucket_destroy(struct tcp_bind_bucket *tb) -{ - if (hlist_empty(&tb->owners)) { - __hlist_del(&tb->node); - kmem_cache_free(tcp_bucket_cachep, tb); - } -} - -/* Caller must disable local BH processing. */ -static __inline__ void __tcp_inherit_port(struct sock *sk, struct sock *child) -{ - struct tcp_bind_hashbucket *head = - &tcp_bhash[tcp_bhashfn(inet_sk(child)->num)]; - struct tcp_bind_bucket *tb; - - spin_lock(&head->lock); - tb = tcp_sk(sk)->bind_hash; - sk_add_bind_node(child, &tb->owners); - tcp_sk(child)->bind_hash = tb; - spin_unlock(&head->lock); -} - -inline void tcp_inherit_port(struct sock *sk, struct sock *child) -{ - local_bh_disable(); - __tcp_inherit_port(sk, child); - local_bh_enable(); -} - -void tcp_bind_hash(struct sock *sk, struct tcp_bind_bucket *tb, - unsigned short snum) -{ - inet_sk(sk)->num = snum; - sk_add_bind_node(sk, &tb->owners); - tcp_sk(sk)->bind_hash = tb; -} - -static inline int tcp_bind_conflict(struct sock *sk, struct tcp_bind_bucket *tb) -{ - const u32 sk_rcv_saddr = tcp_v4_rcv_saddr(sk); - struct sock *sk2; - struct hlist_node *node; - int reuse = sk->sk_reuse; - - sk_for_each_bound(sk2, node, &tb->owners) { - if (sk != sk2 && - !tcp_v6_ipv6only(sk2) && - (!sk->sk_bound_dev_if || - !sk2->sk_bound_dev_if || - sk->sk_bound_dev_if == sk2->sk_bound_dev_if)) { - if (!reuse || !sk2->sk_reuse || - sk2->sk_state == TCP_LISTEN) { - const u32 sk2_rcv_saddr = tcp_v4_rcv_saddr(sk2); - if (!sk2_rcv_saddr || !sk_rcv_saddr || - sk2_rcv_saddr == sk_rcv_saddr) - break; - } - } - } - return node != NULL; -} - -/* Obtain a reference to a local port for the given sock, - * if snum is zero it means select any available local port. - */ static int tcp_v4_get_port(struct sock *sk, unsigned short snum) { - struct tcp_bind_hashbucket *head; - struct hlist_node *node; - struct tcp_bind_bucket *tb; - int ret; - - local_bh_disable(); - if (!snum) { - int low = sysctl_local_port_range[0]; - int high = sysctl_local_port_range[1]; - int remaining = (high - low) + 1; - int rover; - - spin_lock(&tcp_portalloc_lock); - if (tcp_port_rover < low) - rover = low; - else - rover = tcp_port_rover; - do { - rover++; - if (rover > high) - rover = low; - head = &tcp_bhash[tcp_bhashfn(rover)]; - spin_lock(&head->lock); - tb_for_each(tb, node, &head->chain) - if (tb->port == rover) - goto next; - break; - next: - spin_unlock(&head->lock); - } while (--remaining > 0); - tcp_port_rover = rover; - spin_unlock(&tcp_portalloc_lock); - - /* Exhausted local port range during search? It is not - * possible for us to be holding one of the bind hash - * locks if this test triggers, because if 'remaining' - * drops to zero, we broke out of the do/while loop at - * the top level, not from the 'break;' statement. - */ - ret = 1; - if (unlikely(remaining <= 0)) - goto fail; - - /* OK, here is the one we will use. HEAD is - * non-NULL and we hold it's mutex. - */ - snum = rover; - } else { - head = &tcp_bhash[tcp_bhashfn(snum)]; - spin_lock(&head->lock); - tb_for_each(tb, node, &head->chain) - if (tb->port == snum) - goto tb_found; - } - tb = NULL; - goto tb_not_found; -tb_found: - if (!hlist_empty(&tb->owners)) { - if (sk->sk_reuse > 1) - goto success; - if (tb->fastreuse > 0 && - sk->sk_reuse && sk->sk_state != TCP_LISTEN) { - goto success; - } else { - ret = 1; - if (tcp_bind_conflict(sk, tb)) - goto fail_unlock; - } - } -tb_not_found: - ret = 1; - if (!tb && (tb = tcp_bucket_create(head, snum)) == NULL) - goto fail_unlock; - if (hlist_empty(&tb->owners)) { - if (sk->sk_reuse && sk->sk_state != TCP_LISTEN) - tb->fastreuse = 1; - else - tb->fastreuse = 0; - } else if (tb->fastreuse && - (!sk->sk_reuse || sk->sk_state == TCP_LISTEN)) - tb->fastreuse = 0; -success: - if (!tcp_sk(sk)->bind_hash) - tcp_bind_hash(sk, tb, snum); - BUG_TRAP(tcp_sk(sk)->bind_hash == tb); - ret = 0; - -fail_unlock: - spin_unlock(&head->lock); -fail: - local_bh_enable(); - return ret; -} - -/* Get rid of any references to a local port held by the - * given sock. - */ -static void __tcp_put_port(struct sock *sk) -{ - struct inet_sock *inet = inet_sk(sk); - struct tcp_bind_hashbucket *head = &tcp_bhash[tcp_bhashfn(inet->num)]; - struct tcp_bind_bucket *tb; - - spin_lock(&head->lock); - tb = tcp_sk(sk)->bind_hash; - __sk_del_bind_node(sk); - tcp_sk(sk)->bind_hash = NULL; - inet->num = 0; - tcp_bucket_destroy(tb); - spin_unlock(&head->lock); -} - -void tcp_put_port(struct sock *sk) -{ - local_bh_disable(); - __tcp_put_port(sk); - local_bh_enable(); -} - -/* This lock without WQ_FLAG_EXCLUSIVE is good on UP and it can be very bad on SMP. - * Look, when several writers sleep and reader wakes them up, all but one - * immediately hit write lock and grab all the cpus. Exclusive sleep solves - * this, _but_ remember, it adds useless work on UP machines (wake up each - * exclusive lock release). It should be ifdefed really. - */ - -void tcp_listen_wlock(void) -{ - write_lock(&tcp_lhash_lock); - - if (atomic_read(&tcp_lhash_users)) { - DEFINE_WAIT(wait); - - for (;;) { - prepare_to_wait_exclusive(&tcp_lhash_wait, - &wait, TASK_UNINTERRUPTIBLE); - if (!atomic_read(&tcp_lhash_users)) - break; - write_unlock_bh(&tcp_lhash_lock); - schedule(); - write_lock_bh(&tcp_lhash_lock); - } - - finish_wait(&tcp_lhash_wait, &wait); - } -} - -static __inline__ void __tcp_v4_hash(struct sock *sk, const int listen_possible) -{ - struct hlist_head *list; - rwlock_t *lock; - - BUG_TRAP(sk_unhashed(sk)); - if (listen_possible && sk->sk_state == TCP_LISTEN) { - list = &tcp_listening_hash[tcp_sk_listen_hashfn(sk)]; - lock = &tcp_lhash_lock; - tcp_listen_wlock(); - } else { - list = &tcp_ehash[(sk->sk_hashent = tcp_sk_hashfn(sk))].chain; - lock = &tcp_ehash[sk->sk_hashent].lock; - write_lock(lock); - } - __sk_add_node(sk, list); - sock_prot_inc_use(sk->sk_prot); - write_unlock(lock); - if (listen_possible && sk->sk_state == TCP_LISTEN) - wake_up(&tcp_lhash_wait); + return inet_csk_get_port(&tcp_hashinfo, sk, snum); } static void tcp_v4_hash(struct sock *sk) { - if (sk->sk_state != TCP_CLOSE) { - local_bh_disable(); - __tcp_v4_hash(sk, 1); - local_bh_enable(); - } + inet_hash(&tcp_hashinfo, sk); } void tcp_unhash(struct sock *sk) { - rwlock_t *lock; - - if (sk_unhashed(sk)) - goto ende; - - if (sk->sk_state == TCP_LISTEN) { - local_bh_disable(); - tcp_listen_wlock(); - lock = &tcp_lhash_lock; - } else { - struct tcp_ehash_bucket *head = &tcp_ehash[sk->sk_hashent]; - lock = &head->lock; - write_lock_bh(&head->lock); - } - - if (__sk_del_node_init(sk)) - sock_prot_dec_use(sk->sk_prot); - write_unlock_bh(lock); - - ende: - if (sk->sk_state == TCP_LISTEN) - wake_up(&tcp_lhash_wait); -} - -/* Don't inline this cruft. Here are some nice properties to - * exploit here. The BSD API does not allow a listening TCP - * to specify the remote port nor the remote address for the - * connection. So always assume those are both wildcarded - * during the search since they can never be otherwise. - */ -static struct sock *__tcp_v4_lookup_listener(struct hlist_head *head, u32 daddr, - unsigned short hnum, int dif) -{ - struct sock *result = NULL, *sk; - struct hlist_node *node; - int score, hiscore; - - hiscore=-1; - sk_for_each(sk, node, head) { - struct inet_sock *inet = inet_sk(sk); - - if (inet->num == hnum && !ipv6_only_sock(sk)) { - __u32 rcv_saddr = inet->rcv_saddr; - - score = (sk->sk_family == PF_INET ? 1 : 0); - if (rcv_saddr) { - if (rcv_saddr != daddr) - continue; - score+=2; - } - if (sk->sk_bound_dev_if) { - if (sk->sk_bound_dev_if != dif) - continue; - score+=2; - } - if (score == 5) - return sk; - if (score > hiscore) { - hiscore = score; - result = sk; - } - } - } - return result; -} - -/* Optimize the common listener case. */ -static inline struct sock *tcp_v4_lookup_listener(u32 daddr, - unsigned short hnum, int dif) -{ - struct sock *sk = NULL; - struct hlist_head *head; - - read_lock(&tcp_lhash_lock); - head = &tcp_listening_hash[tcp_lhashfn(hnum)]; - if (!hlist_empty(head)) { - struct inet_sock *inet = inet_sk((sk = __sk_head(head))); - - if (inet->num == hnum && !sk->sk_node.next && - (!inet->rcv_saddr || inet->rcv_saddr == daddr) && - (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) && - !sk->sk_bound_dev_if) - goto sherry_cache; - sk = __tcp_v4_lookup_listener(head, daddr, hnum, dif); - } - if (sk) { -sherry_cache: - sock_hold(sk); - } - read_unlock(&tcp_lhash_lock); - return sk; + inet_unhash(&tcp_hashinfo, sk); } -/* Sockets in TCP_CLOSE state are _always_ taken out of the hash, so - * we need not check it for TCP lookups anymore, thanks Alexey. -DaveM - * - * Local BH must be disabled here. - */ - -static inline struct sock *__tcp_v4_lookup_established(u32 saddr, u16 sport, - u32 daddr, u16 hnum, - int dif) -{ - struct tcp_ehash_bucket *head; - TCP_V4_ADDR_COOKIE(acookie, saddr, daddr) - __u32 ports = TCP_COMBINED_PORTS(sport, hnum); - struct sock *sk; - struct hlist_node *node; - /* Optimize here for direct hit, only listening connections can - * have wildcards anyways. - */ - int hash = tcp_hashfn(daddr, hnum, saddr, sport); - head = &tcp_ehash[hash]; - read_lock(&head->lock); - sk_for_each(sk, node, &head->chain) { - if (TCP_IPV4_MATCH(sk, acookie, saddr, daddr, ports, dif)) - goto hit; /* You sunk my battleship! */ - } - - /* Must check for a TIME_WAIT'er before going to listener hash. */ - sk_for_each(sk, node, &(head + tcp_ehash_size)->chain) { - if (TCP_IPV4_TW_MATCH(sk, acookie, saddr, daddr, ports, dif)) - goto hit; - } - sk = NULL; -out: - read_unlock(&head->lock); - return sk; -hit: - sock_hold(sk); - goto out; -} - -static inline struct sock *__tcp_v4_lookup(u32 saddr, u16 sport, - u32 daddr, u16 hnum, int dif) -{ - struct sock *sk = __tcp_v4_lookup_established(saddr, sport, - daddr, hnum, dif); - - return sk ? : tcp_v4_lookup_listener(daddr, hnum, dif); -} - -inline struct sock *tcp_v4_lookup(u32 saddr, u16 sport, u32 daddr, - u16 dport, int dif) -{ - struct sock *sk; - - local_bh_disable(); - sk = __tcp_v4_lookup(saddr, sport, daddr, ntohs(dport), dif); - local_bh_enable(); - - return sk; -} - -EXPORT_SYMBOL_GPL(tcp_v4_lookup); - static inline __u32 tcp_v4_init_sequence(struct sock *sk, struct sk_buff *skb) { return secure_tcp_sequence_number(skb->nh.iph->daddr, @@ -555,27 +122,28 @@ static inline __u32 tcp_v4_init_sequence(struct sock *sk, struct sk_buff *skb) /* called with local bh disabled */ static int __tcp_v4_check_established(struct sock *sk, __u16 lport, - struct tcp_tw_bucket **twp) + struct inet_timewait_sock **twp) { struct inet_sock *inet = inet_sk(sk); u32 daddr = inet->rcv_saddr; u32 saddr = inet->daddr; int dif = sk->sk_bound_dev_if; - TCP_V4_ADDR_COOKIE(acookie, saddr, daddr) - __u32 ports = TCP_COMBINED_PORTS(inet->dport, lport); - int hash = tcp_hashfn(daddr, lport, saddr, inet->dport); - struct tcp_ehash_bucket *head = &tcp_ehash[hash]; + INET_ADDR_COOKIE(acookie, saddr, daddr) + const __u32 ports = INET_COMBINED_PORTS(inet->dport, lport); + const int hash = inet_ehashfn(daddr, lport, saddr, inet->dport, tcp_hashinfo.ehash_size); + struct inet_ehash_bucket *head = &tcp_hashinfo.ehash[hash]; struct sock *sk2; - struct hlist_node *node; - struct tcp_tw_bucket *tw; + const struct hlist_node *node; + struct inet_timewait_sock *tw; write_lock(&head->lock); /* Check TIME-WAIT sockets first. */ - sk_for_each(sk2, node, &(head + tcp_ehash_size)->chain) { - tw = (struct tcp_tw_bucket *)sk2; + sk_for_each(sk2, node, &(head + tcp_hashinfo.ehash_size)->chain) { + tw = inet_twsk(sk2); - if (TCP_IPV4_TW_MATCH(sk2, acookie, saddr, daddr, ports, dif)) { + if (INET_TW_MATCH(sk2, acookie, saddr, daddr, ports, dif)) { + const struct tcp_timewait_sock *tcptw = tcp_twsk(sk2); struct tcp_sock *tp = tcp_sk(sk); /* With PAWS, it is safe from the viewpoint @@ -592,15 +160,15 @@ static int __tcp_v4_check_established(struct sock *sk, __u16 lport, fall back to VJ's scheme and use initial timestamp retrieved from peer table. */ - if (tw->tw_ts_recent_stamp && + if (tcptw->tw_ts_recent_stamp && (!twp || (sysctl_tcp_tw_reuse && xtime.tv_sec - - tw->tw_ts_recent_stamp > 1))) { - if ((tp->write_seq = - tw->tw_snd_nxt + 65535 + 2) == 0) + tcptw->tw_ts_recent_stamp > 1))) { + tp->write_seq = tcptw->tw_snd_nxt + 65535 + 2; + if (tp->write_seq == 0) tp->write_seq = 1; - tp->rx_opt.ts_recent = tw->tw_ts_recent; - tp->rx_opt.ts_recent_stamp = tw->tw_ts_recent_stamp; + tp->rx_opt.ts_recent = tcptw->tw_ts_recent; + tp->rx_opt.ts_recent_stamp = tcptw->tw_ts_recent_stamp; sock_hold(sk2); goto unique; } else @@ -611,7 +179,7 @@ static int __tcp_v4_check_established(struct sock *sk, __u16 lport, /* And established part... */ sk_for_each(sk2, node, &head->chain) { - if (TCP_IPV4_MATCH(sk2, acookie, saddr, daddr, ports, dif)) + if (INET_MATCH(sk2, acookie, saddr, daddr, ports, dif)) goto not_unique; } @@ -631,10 +199,10 @@ unique: NET_INC_STATS_BH(LINUX_MIB_TIMEWAITRECYCLED); } else if (tw) { /* Silly. Should hash-dance instead... */ - tcp_tw_deschedule(tw); + inet_twsk_deschedule(tw, &tcp_death_row); NET_INC_STATS_BH(LINUX_MIB_TIMEWAITRECYCLED); - tcp_tw_put(tw); + inet_twsk_put(tw); } return 0; @@ -657,9 +225,9 @@ static inline u32 connect_port_offset(const struct sock *sk) */ static inline int tcp_v4_hash_connect(struct sock *sk) { - unsigned short snum = inet_sk(sk)->num; - struct tcp_bind_hashbucket *head; - struct tcp_bind_bucket *tb; + const unsigned short snum = inet_sk(sk)->num; + struct inet_bind_hashbucket *head; + struct inet_bind_bucket *tb; int ret; if (!snum) { @@ -671,19 +239,19 @@ static inline int tcp_v4_hash_connect(struct sock *sk) static u32 hint; u32 offset = hint + connect_port_offset(sk); struct hlist_node *node; - struct tcp_tw_bucket *tw = NULL; + struct inet_timewait_sock *tw = NULL; local_bh_disable(); for (i = 1; i <= range; i++) { port = low + (i + offset) % range; - head = &tcp_bhash[tcp_bhashfn(port)]; + head = &tcp_hashinfo.bhash[inet_bhashfn(port, tcp_hashinfo.bhash_size)]; spin_lock(&head->lock); /* Does not bother with rcv_saddr checks, * because the established check is already * unique enough. */ - tb_for_each(tb, node, &head->chain) { + inet_bind_bucket_for_each(tb, node, &head->chain) { if (tb->port == port) { BUG_TRAP(!hlist_empty(&tb->owners)); if (tb->fastreuse >= 0) @@ -696,7 +264,7 @@ static inline int tcp_v4_hash_connect(struct sock *sk) } } - tb = tcp_bucket_create(head, port); + tb = inet_bind_bucket_create(tcp_hashinfo.bind_bucket_cachep, head, port); if (!tb) { spin_unlock(&head->lock); break; @@ -715,27 +283,27 @@ ok: hint += i; /* Head lock still held and bh's disabled */ - tcp_bind_hash(sk, tb, port); + inet_bind_hash(sk, tb, port); if (sk_unhashed(sk)) { inet_sk(sk)->sport = htons(port); - __tcp_v4_hash(sk, 0); + __inet_hash(&tcp_hashinfo, sk, 0); } spin_unlock(&head->lock); if (tw) { - tcp_tw_deschedule(tw); - tcp_tw_put(tw); + inet_twsk_deschedule(tw, &tcp_death_row);; + inet_twsk_put(tw); } ret = 0; goto out; } - head = &tcp_bhash[tcp_bhashfn(snum)]; - tb = tcp_sk(sk)->bind_hash; + head = &tcp_hashinfo.bhash[inet_bhashfn(snum, tcp_hashinfo.bhash_size)]; + tb = inet_csk(sk)->icsk_bind_hash; spin_lock_bh(&head->lock); if (sk_head(&tb->owners) == sk && !sk->sk_bind_node.next) { - __tcp_v4_hash(sk, 0); + __inet_hash(&tcp_hashinfo, sk, 0); spin_unlock_bh(&head->lock); return 0; } else { @@ -798,7 +366,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) tp->write_seq = 0; } - if (sysctl_tcp_tw_recycle && + if (tcp_death_row.sysctl_tw_recycle && !tp->rx_opt.ts_recent_stamp && rt->rt_dst == daddr) { struct inet_peer *peer = rt_get_peer(rt); @@ -837,8 +405,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) goto failure; /* OK, now commit destination to socket. */ - __sk_dst_set(sk, &rt->u.dst); - tcp_v4_setup_caps(sk, &rt->u.dst); + sk_setup_caps(sk, &rt->u.dst); if (!tp->write_seq) tp->write_seq = secure_tcp_sequence_number(inet->saddr, @@ -864,53 +431,6 @@ failure: return err; } -static __inline__ int tcp_v4_iif(struct sk_buff *skb) -{ - return ((struct rtable *)skb->dst)->rt_iif; -} - -static __inline__ u32 tcp_v4_synq_hash(u32 raddr, u16 rport, u32 rnd) -{ - return (jhash_2words(raddr, (u32) rport, rnd) & (TCP_SYNQ_HSIZE - 1)); -} - -static struct request_sock *tcp_v4_search_req(struct tcp_sock *tp, - struct request_sock ***prevp, - __u16 rport, - __u32 raddr, __u32 laddr) -{ - struct listen_sock *lopt = tp->accept_queue.listen_opt; - struct request_sock *req, **prev; - - for (prev = &lopt->syn_table[tcp_v4_synq_hash(raddr, rport, lopt->hash_rnd)]; - (req = *prev) != NULL; - prev = &req->dl_next) { - const struct inet_request_sock *ireq = inet_rsk(req); - - if (ireq->rmt_port == rport && - ireq->rmt_addr == raddr && - ireq->loc_addr == laddr && - TCP_INET_FAMILY(req->rsk_ops->family)) { - BUG_TRAP(!req->sk); - *prevp = prev; - break; - } - } - - return req; -} - -static void tcp_v4_synq_add(struct sock *sk, struct request_sock *req) -{ - struct tcp_sock *tp = tcp_sk(sk); - struct listen_sock *lopt = tp->accept_queue.listen_opt; - u32 h = tcp_v4_synq_hash(inet_rsk(req)->rmt_addr, inet_rsk(req)->rmt_port, lopt->hash_rnd); - - reqsk_queue_hash_req(&tp->accept_queue, h, req, TCP_TIMEOUT_INIT); - tcp_synq_added(sk); -} - - /* * This routine does path mtu discovery as defined in RFC1191. */ @@ -993,14 +513,14 @@ void tcp_v4_err(struct sk_buff *skb, u32 info) return; } - sk = tcp_v4_lookup(iph->daddr, th->dest, iph->saddr, - th->source, tcp_v4_iif(skb)); + sk = inet_lookup(&tcp_hashinfo, iph->daddr, th->dest, iph->saddr, + th->source, inet_iif(skb)); if (!sk) { ICMP_INC_STATS_BH(ICMP_MIB_INERRORS); return; } if (sk->sk_state == TCP_TIME_WAIT) { - tcp_tw_put((struct tcp_tw_bucket *)sk); + inet_twsk_put((struct inet_timewait_sock *)sk); return; } @@ -1054,8 +574,8 @@ void tcp_v4_err(struct sk_buff *skb, u32 info) if (sock_owned_by_user(sk)) goto out; - req = tcp_v4_search_req(tp, &prev, th->dest, - iph->daddr, iph->saddr); + req = inet_csk_search_req(sk, &prev, th->dest, + iph->daddr, iph->saddr); if (!req) goto out; @@ -1075,7 +595,7 @@ void tcp_v4_err(struct sk_buff *skb, u32 info) * created socket, and POSIX does not want network * errors returned from accept(). */ - tcp_synq_drop(sk, req, prev); + inet_csk_reqsk_queue_drop(sk, req, prev); goto out; case TCP_SYN_SENT: @@ -1245,12 +765,13 @@ static void tcp_v4_send_ack(struct sk_buff *skb, u32 seq, u32 ack, static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb) { - struct tcp_tw_bucket *tw = (struct tcp_tw_bucket *)sk; + struct inet_timewait_sock *tw = inet_twsk(sk); + const struct tcp_timewait_sock *tcptw = tcp_twsk(sk); - tcp_v4_send_ack(skb, tw->tw_snd_nxt, tw->tw_rcv_nxt, - tw->tw_rcv_wnd >> tw->tw_rcv_wscale, tw->tw_ts_recent); + tcp_v4_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt, + tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale, tcptw->tw_ts_recent); - tcp_tw_put(tw); + inet_twsk_put(tw); } static void tcp_v4_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req) @@ -1259,36 +780,6 @@ static void tcp_v4_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req) req->ts_recent); } -static struct dst_entry* tcp_v4_route_req(struct sock *sk, - struct request_sock *req) -{ - struct rtable *rt; - const struct inet_request_sock *ireq = inet_rsk(req); - struct ip_options *opt = inet_rsk(req)->opt; - struct flowi fl = { .oif = sk->sk_bound_dev_if, - .nl_u = { .ip4_u = - { .daddr = ((opt && opt->srr) ? - opt->faddr : - ireq->rmt_addr), - .saddr = ireq->loc_addr, - .tos = RT_CONN_FLAGS(sk) } }, - .proto = IPPROTO_TCP, - .uli_u = { .ports = - { .sport = inet_sk(sk)->sport, - .dport = ireq->rmt_port } } }; - - if (ip_route_output_flow(&rt, &fl, sk, 0)) { - IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES); - return NULL; - } - if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway) { - ip_rt_put(rt); - IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES); - return NULL; - } - return &rt->u.dst; -} - /* * Send a SYN-ACK after having received an ACK. * This still operates on a request_sock only, not on a big @@ -1302,7 +793,7 @@ static int tcp_v4_send_synack(struct sock *sk, struct request_sock *req, struct sk_buff * skb; /* First, grab a route. */ - if (!dst && (dst = tcp_v4_route_req(sk, req)) == NULL) + if (!dst && (dst = inet_csk_route_req(sk, req)) == NULL) goto out; skb = tcp_make_synack(sk, dst, req); @@ -1404,7 +895,7 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) * limitations, they conserve resources and peer is * evidently real one. */ - if (tcp_synq_is_full(sk) && !isn) { + if (inet_csk_reqsk_queue_is_full(sk) && !isn) { #ifdef CONFIG_SYN_COOKIES if (sysctl_tcp_syncookies) { want_cookie = 1; @@ -1418,7 +909,7 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) * clogging syn queue with openreqs with exponentially increasing * timeout. */ - if (sk_acceptq_is_full(sk) && tcp_synq_young(sk) > 1) + if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1) goto drop; req = reqsk_alloc(&tcp_request_sock_ops); @@ -1474,8 +965,8 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) * are made in the function processing timewait state. */ if (tmp_opt.saw_tstamp && - sysctl_tcp_tw_recycle && - (dst = tcp_v4_route_req(sk, req)) != NULL && + tcp_death_row.sysctl_tw_recycle && + (dst = inet_csk_route_req(sk, req)) != NULL && (peer = rt_get_peer((struct rtable *)dst)) != NULL && peer->v4daddr == saddr) { if (xtime.tv_sec < peer->tcp_ts_stamp + TCP_PAWS_MSL && @@ -1488,7 +979,7 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) } /* Kill the following clause, if you dislike this way. */ else if (!sysctl_tcp_syncookies && - (sysctl_max_syn_backlog - tcp_synq_len(sk) < + (sysctl_max_syn_backlog - inet_csk_reqsk_queue_len(sk) < (sysctl_max_syn_backlog >> 2)) && (!peer || !peer->tcp_ts_stamp) && (!dst || !dst_metric(dst, RTAX_RTT))) { @@ -1499,11 +990,10 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) * to destinations, already remembered * to the moment of synflood. */ - LIMIT_NETDEBUG(printk(KERN_DEBUG "TCP: drop open " - "request from %u.%u." - "%u.%u/%u\n", - NIPQUAD(saddr), - ntohs(skb->h.th->source))); + LIMIT_NETDEBUG(KERN_DEBUG "TCP: drop open " + "request from %u.%u.%u.%u/%u\n", + NIPQUAD(saddr), + ntohs(skb->h.th->source)); dst_release(dst); goto drop_and_free; } @@ -1518,7 +1008,7 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) if (want_cookie) { reqsk_free(req); } else { - tcp_v4_synq_add(sk, req); + inet_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT); } return 0; @@ -1546,15 +1036,14 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb, if (sk_acceptq_is_full(sk)) goto exit_overflow; - if (!dst && (dst = tcp_v4_route_req(sk, req)) == NULL) + if (!dst && (dst = inet_csk_route_req(sk, req)) == NULL) goto exit; newsk = tcp_create_openreq_child(sk, req, skb); if (!newsk) goto exit; - newsk->sk_dst_cache = dst; - tcp_v4_setup_caps(newsk, dst); + sk_setup_caps(newsk, dst); newtp = tcp_sk(newsk); newinet = inet_sk(newsk); @@ -1564,7 +1053,7 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb, newinet->saddr = ireq->loc_addr; newinet->opt = ireq->opt; ireq->opt = NULL; - newinet->mc_index = tcp_v4_iif(skb); + newinet->mc_index = inet_iif(skb); newinet->mc_ttl = skb->nh.iph->ttl; newtp->ext_header_len = 0; if (newinet->opt) @@ -1575,8 +1064,8 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb, newtp->advmss = dst_metric(dst, RTAX_ADVMSS); tcp_initialize_rcv_mss(newsk); - __tcp_v4_hash(newsk, 0); - __tcp_inherit_port(sk, newsk); + __inet_hash(&tcp_hashinfo, newsk, 0); + __inet_inherit_port(&tcp_hashinfo, sk, newsk); return newsk; @@ -1592,27 +1081,24 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) { struct tcphdr *th = skb->h.th; struct iphdr *iph = skb->nh.iph; - struct tcp_sock *tp = tcp_sk(sk); struct sock *nsk; struct request_sock **prev; /* Find possible connection requests. */ - struct request_sock *req = tcp_v4_search_req(tp, &prev, th->source, - iph->saddr, iph->daddr); + struct request_sock *req = inet_csk_search_req(sk, &prev, th->source, + iph->saddr, iph->daddr); if (req) return tcp_check_req(sk, skb, req, prev); - nsk = __tcp_v4_lookup_established(skb->nh.iph->saddr, - th->source, - skb->nh.iph->daddr, - ntohs(th->dest), - tcp_v4_iif(skb)); + nsk = __inet_lookup_established(&tcp_hashinfo, skb->nh.iph->saddr, + th->source, skb->nh.iph->daddr, + ntohs(th->dest), inet_iif(skb)); if (nsk) { if (nsk->sk_state != TCP_TIME_WAIT) { bh_lock_sock(nsk); return nsk; } - tcp_tw_put((struct tcp_tw_bucket *)nsk); + inet_twsk_put((struct inet_timewait_sock *)nsk); return NULL; } @@ -1631,7 +1117,7 @@ static int tcp_v4_checksum_init(struct sk_buff *skb) skb->nh.iph->daddr, skb->csum)) return 0; - LIMIT_NETDEBUG(printk(KERN_DEBUG "hw tcp v4 csum failed\n")); + LIMIT_NETDEBUG(KERN_DEBUG "hw tcp v4 csum failed\n"); skb->ip_summed = CHECKSUM_NONE; } if (skb->len <= 76) { @@ -1747,9 +1233,9 @@ int tcp_v4_rcv(struct sk_buff *skb) TCP_SKB_CB(skb)->flags = skb->nh.iph->tos; TCP_SKB_CB(skb)->sacked = 0; - sk = __tcp_v4_lookup(skb->nh.iph->saddr, th->source, - skb->nh.iph->daddr, ntohs(th->dest), - tcp_v4_iif(skb)); + sk = __inet_lookup(&tcp_hashinfo, skb->nh.iph->saddr, th->source, + skb->nh.iph->daddr, ntohs(th->dest), + inet_iif(skb)); if (!sk) goto no_tcp_socket; @@ -1801,24 +1287,26 @@ discard_and_relse: do_time_wait: if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) { - tcp_tw_put((struct tcp_tw_bucket *) sk); + inet_twsk_put((struct inet_timewait_sock *) sk); goto discard_it; } if (skb->len < (th->doff << 2) || tcp_checksum_complete(skb)) { TCP_INC_STATS_BH(TCP_MIB_INERRS); - tcp_tw_put((struct tcp_tw_bucket *) sk); + inet_twsk_put((struct inet_timewait_sock *) sk); goto discard_it; } - switch (tcp_timewait_state_process((struct tcp_tw_bucket *)sk, - skb, th, skb->len)) { + switch (tcp_timewait_state_process((struct inet_timewait_sock *)sk, + skb, th)) { case TCP_TW_SYN: { - struct sock *sk2 = tcp_v4_lookup_listener(skb->nh.iph->daddr, - ntohs(th->dest), - tcp_v4_iif(skb)); + struct sock *sk2 = inet_lookup_listener(&tcp_hashinfo, + skb->nh.iph->daddr, + ntohs(th->dest), + inet_iif(skb)); if (sk2) { - tcp_tw_deschedule((struct tcp_tw_bucket *)sk); - tcp_tw_put((struct tcp_tw_bucket *)sk); + inet_twsk_deschedule((struct inet_timewait_sock *)sk, + &tcp_death_row); + inet_twsk_put((struct inet_timewait_sock *)sk); sk = sk2; goto process; } @@ -1834,112 +1322,6 @@ do_time_wait: goto discard_it; } -/* With per-bucket locks this operation is not-atomic, so that - * this version is not worse. - */ -static void __tcp_v4_rehash(struct sock *sk) -{ - sk->sk_prot->unhash(sk); - sk->sk_prot->hash(sk); -} - -static int tcp_v4_reselect_saddr(struct sock *sk) -{ - struct inet_sock *inet = inet_sk(sk); - int err; - struct rtable *rt; - __u32 old_saddr = inet->saddr; - __u32 new_saddr; - __u32 daddr = inet->daddr; - - if (inet->opt && inet->opt->srr) - daddr = inet->opt->faddr; - - /* Query new route. */ - err = ip_route_connect(&rt, daddr, 0, - RT_CONN_FLAGS(sk), - sk->sk_bound_dev_if, - IPPROTO_TCP, - inet->sport, inet->dport, sk); - if (err) - return err; - - __sk_dst_set(sk, &rt->u.dst); - tcp_v4_setup_caps(sk, &rt->u.dst); - - new_saddr = rt->rt_src; - - if (new_saddr == old_saddr) - return 0; - - if (sysctl_ip_dynaddr > 1) { - printk(KERN_INFO "tcp_v4_rebuild_header(): shifting inet->" - "saddr from %d.%d.%d.%d to %d.%d.%d.%d\n", - NIPQUAD(old_saddr), - NIPQUAD(new_saddr)); - } - - inet->saddr = new_saddr; - inet->rcv_saddr = new_saddr; - - /* XXX The only one ugly spot where we need to - * XXX really change the sockets identity after - * XXX it has entered the hashes. -DaveM - * - * Besides that, it does not check for connection - * uniqueness. Wait for troubles. - */ - __tcp_v4_rehash(sk); - return 0; -} - -int tcp_v4_rebuild_header(struct sock *sk) -{ - struct inet_sock *inet = inet_sk(sk); - struct rtable *rt = (struct rtable *)__sk_dst_check(sk, 0); - u32 daddr; - int err; - - /* Route is OK, nothing to do. */ - if (rt) - return 0; - - /* Reroute. */ - daddr = inet->daddr; - if (inet->opt && inet->opt->srr) - daddr = inet->opt->faddr; - - { - struct flowi fl = { .oif = sk->sk_bound_dev_if, - .nl_u = { .ip4_u = - { .daddr = daddr, - .saddr = inet->saddr, - .tos = RT_CONN_FLAGS(sk) } }, - .proto = IPPROTO_TCP, - .uli_u = { .ports = - { .sport = inet->sport, - .dport = inet->dport } } }; - - err = ip_route_output_flow(&rt, &fl, sk, 0); - } - if (!err) { - __sk_dst_set(sk, &rt->u.dst); - tcp_v4_setup_caps(sk, &rt->u.dst); - return 0; - } - - /* Routing failed... */ - sk->sk_route_caps = 0; - - if (!sysctl_ip_dynaddr || - sk->sk_state != TCP_SYN_SENT || - (sk->sk_userlocks & SOCK_BINDADDR_LOCK) || - (err = tcp_v4_reselect_saddr(sk)) != 0) - sk->sk_err_soft = -err; - - return err; -} - static void v4_addr2sockaddr(struct sock *sk, struct sockaddr * uaddr) { struct sockaddr_in *sin = (struct sockaddr_in *) uaddr; @@ -1988,18 +1370,18 @@ int tcp_v4_remember_stamp(struct sock *sk) return 0; } -int tcp_v4_tw_remember_stamp(struct tcp_tw_bucket *tw) +int tcp_v4_tw_remember_stamp(struct inet_timewait_sock *tw) { - struct inet_peer *peer = NULL; - - peer = inet_getpeer(tw->tw_daddr, 1); + struct inet_peer *peer = inet_getpeer(tw->tw_daddr, 1); if (peer) { - if ((s32)(peer->tcp_ts - tw->tw_ts_recent) <= 0 || + const struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); + + if ((s32)(peer->tcp_ts - tcptw->tw_ts_recent) <= 0 || (peer->tcp_ts_stamp + TCP_PAWS_MSL < xtime.tv_sec && - peer->tcp_ts_stamp <= tw->tw_ts_recent_stamp)) { - peer->tcp_ts_stamp = tw->tw_ts_recent_stamp; - peer->tcp_ts = tw->tw_ts_recent; + peer->tcp_ts_stamp <= tcptw->tw_ts_recent_stamp)) { + peer->tcp_ts_stamp = tcptw->tw_ts_recent_stamp; + peer->tcp_ts = tcptw->tw_ts_recent; } inet_putpeer(peer); return 1; @@ -2011,7 +1393,7 @@ int tcp_v4_tw_remember_stamp(struct tcp_tw_bucket *tw) struct tcp_func ipv4_specific = { .queue_xmit = ip_queue_xmit, .send_check = tcp_v4_send_check, - .rebuild_header = tcp_v4_rebuild_header, + .rebuild_header = inet_sk_rebuild_header, .conn_request = tcp_v4_conn_request, .syn_recv_sock = tcp_v4_syn_recv_sock, .remember_stamp = tcp_v4_remember_stamp, @@ -2027,13 +1409,14 @@ struct tcp_func ipv4_specific = { */ static int tcp_v4_init_sock(struct sock *sk) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); skb_queue_head_init(&tp->out_of_order_queue); tcp_init_xmit_timers(sk); tcp_prequeue_init(tp); - tp->rto = TCP_TIMEOUT_INIT; + icsk->icsk_rto = TCP_TIMEOUT_INIT; tp->mdev = TCP_TIMEOUT_INIT; /* So many TCP implementations out there (incorrectly) count the @@ -2051,7 +1434,7 @@ static int tcp_v4_init_sock(struct sock *sk) tp->mss_cache = 536; tp->reordering = sysctl_tcp_reordering; - tp->ca_ops = &tcp_init_congestion_ops; + icsk->icsk_ca_ops = &tcp_init_congestion_ops; sk->sk_state = TCP_CLOSE; @@ -2074,7 +1457,7 @@ int tcp_v4_destroy_sock(struct sock *sk) tcp_clear_xmit_timers(sk); - tcp_cleanup_congestion_control(tp); + tcp_cleanup_congestion_control(sk); /* Cleanup up the write buffer. */ sk_stream_writequeue_purge(sk); @@ -2086,8 +1469,8 @@ int tcp_v4_destroy_sock(struct sock *sk) __skb_queue_purge(&tp->ucopy.prequeue); /* Clean up a referenced TCP bind bucket. */ - if (tp->bind_hash) - tcp_put_port(sk); + if (inet_csk(sk)->icsk_bind_hash) + inet_put_port(&tcp_hashinfo, sk); /* * If sendmsg cached page exists, toss it. @@ -2107,13 +1490,13 @@ EXPORT_SYMBOL(tcp_v4_destroy_sock); #ifdef CONFIG_PROC_FS /* Proc filesystem TCP sock list dumping. */ -static inline struct tcp_tw_bucket *tw_head(struct hlist_head *head) +static inline struct inet_timewait_sock *tw_head(struct hlist_head *head) { return hlist_empty(head) ? NULL : - list_entry(head->first, struct tcp_tw_bucket, tw_node); + list_entry(head->first, struct inet_timewait_sock, tw_node); } -static inline struct tcp_tw_bucket *tw_next(struct tcp_tw_bucket *tw) +static inline struct inet_timewait_sock *tw_next(struct inet_timewait_sock *tw) { return tw->tw_node.next ? hlist_entry(tw->tw_node.next, typeof(*tw), tw_node) : NULL; @@ -2121,14 +1504,14 @@ static inline struct tcp_tw_bucket *tw_next(struct tcp_tw_bucket *tw) static void *listening_get_next(struct seq_file *seq, void *cur) { - struct tcp_sock *tp; + struct inet_connection_sock *icsk; struct hlist_node *node; struct sock *sk = cur; struct tcp_iter_state* st = seq->private; if (!sk) { st->bucket = 0; - sk = sk_head(&tcp_listening_hash[0]); + sk = sk_head(&tcp_hashinfo.listening_hash[0]); goto get_sk; } @@ -2137,7 +1520,7 @@ static void *listening_get_next(struct seq_file *seq, void *cur) if (st->state == TCP_SEQ_STATE_OPENREQ) { struct request_sock *req = cur; - tp = tcp_sk(st->syn_wait_sk); + icsk = inet_csk(st->syn_wait_sk); req = req->dl_next; while (1) { while (req) { @@ -2150,17 +1533,17 @@ static void *listening_get_next(struct seq_file *seq, void *cur) if (++st->sbucket >= TCP_SYNQ_HSIZE) break; get_req: - req = tp->accept_queue.listen_opt->syn_table[st->sbucket]; + req = icsk->icsk_accept_queue.listen_opt->syn_table[st->sbucket]; } sk = sk_next(st->syn_wait_sk); st->state = TCP_SEQ_STATE_LISTENING; - read_unlock_bh(&tp->accept_queue.syn_wait_lock); + read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock); } else { - tp = tcp_sk(sk); - read_lock_bh(&tp->accept_queue.syn_wait_lock); - if (reqsk_queue_len(&tp->accept_queue)) + icsk = inet_csk(sk); + read_lock_bh(&icsk->icsk_accept_queue.syn_wait_lock); + if (reqsk_queue_len(&icsk->icsk_accept_queue)) goto start_req; - read_unlock_bh(&tp->accept_queue.syn_wait_lock); + read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock); sk = sk_next(sk); } get_sk: @@ -2169,9 +1552,9 @@ get_sk: cur = sk; goto out; } - tp = tcp_sk(sk); - read_lock_bh(&tp->accept_queue.syn_wait_lock); - if (reqsk_queue_len(&tp->accept_queue)) { + icsk = inet_csk(sk); + read_lock_bh(&icsk->icsk_accept_queue.syn_wait_lock); + if (reqsk_queue_len(&icsk->icsk_accept_queue)) { start_req: st->uid = sock_i_uid(sk); st->syn_wait_sk = sk; @@ -2179,10 +1562,10 @@ start_req: st->sbucket = 0; goto get_req; } - read_unlock_bh(&tp->accept_queue.syn_wait_lock); + read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock); } - if (++st->bucket < TCP_LHTABLE_SIZE) { - sk = sk_head(&tcp_listening_hash[st->bucket]); + if (++st->bucket < INET_LHTABLE_SIZE) { + sk = sk_head(&tcp_hashinfo.listening_hash[st->bucket]); goto get_sk; } cur = NULL; @@ -2206,16 +1589,16 @@ static void *established_get_first(struct seq_file *seq) struct tcp_iter_state* st = seq->private; void *rc = NULL; - for (st->bucket = 0; st->bucket < tcp_ehash_size; ++st->bucket) { + for (st->bucket = 0; st->bucket < tcp_hashinfo.ehash_size; ++st->bucket) { struct sock *sk; struct hlist_node *node; - struct tcp_tw_bucket *tw; + struct inet_timewait_sock *tw; /* We can reschedule _before_ having picked the target: */ cond_resched_softirq(); - read_lock(&tcp_ehash[st->bucket].lock); - sk_for_each(sk, node, &tcp_ehash[st->bucket].chain) { + read_lock(&tcp_hashinfo.ehash[st->bucket].lock); + sk_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) { if (sk->sk_family != st->family) { continue; } @@ -2223,15 +1606,15 @@ static void *established_get_first(struct seq_file *seq) goto out; } st->state = TCP_SEQ_STATE_TIME_WAIT; - tw_for_each(tw, node, - &tcp_ehash[st->bucket + tcp_ehash_size].chain) { + inet_twsk_for_each(tw, node, + &tcp_hashinfo.ehash[st->bucket + tcp_hashinfo.ehash_size].chain) { if (tw->tw_family != st->family) { continue; } rc = tw; goto out; } - read_unlock(&tcp_ehash[st->bucket].lock); + read_unlock(&tcp_hashinfo.ehash[st->bucket].lock); st->state = TCP_SEQ_STATE_ESTABLISHED; } out: @@ -2241,7 +1624,7 @@ out: static void *established_get_next(struct seq_file *seq, void *cur) { struct sock *sk = cur; - struct tcp_tw_bucket *tw; + struct inet_timewait_sock *tw; struct hlist_node *node; struct tcp_iter_state* st = seq->private; @@ -2258,15 +1641,15 @@ get_tw: cur = tw; goto out; } - read_unlock(&tcp_ehash[st->bucket].lock); + read_unlock(&tcp_hashinfo.ehash[st->bucket].lock); st->state = TCP_SEQ_STATE_ESTABLISHED; /* We can reschedule between buckets: */ cond_resched_softirq(); - if (++st->bucket < tcp_ehash_size) { - read_lock(&tcp_ehash[st->bucket].lock); - sk = sk_head(&tcp_ehash[st->bucket].chain); + if (++st->bucket < tcp_hashinfo.ehash_size) { + read_lock(&tcp_hashinfo.ehash[st->bucket].lock); + sk = sk_head(&tcp_hashinfo.ehash[st->bucket].chain); } else { cur = NULL; goto out; @@ -2280,7 +1663,7 @@ get_tw: } st->state = TCP_SEQ_STATE_TIME_WAIT; - tw = tw_head(&tcp_ehash[st->bucket + tcp_ehash_size].chain); + tw = tw_head(&tcp_hashinfo.ehash[st->bucket + tcp_hashinfo.ehash_size].chain); goto get_tw; found: cur = sk; @@ -2304,12 +1687,12 @@ static void *tcp_get_idx(struct seq_file *seq, loff_t pos) void *rc; struct tcp_iter_state* st = seq->private; - tcp_listen_lock(); + inet_listen_lock(&tcp_hashinfo); st->state = TCP_SEQ_STATE_LISTENING; rc = listening_get_idx(seq, &pos); if (!rc) { - tcp_listen_unlock(); + inet_listen_unlock(&tcp_hashinfo); local_bh_disable(); st->state = TCP_SEQ_STATE_ESTABLISHED; rc = established_get_idx(seq, pos); @@ -2342,7 +1725,7 @@ static void *tcp_seq_next(struct seq_file *seq, void *v, loff_t *pos) case TCP_SEQ_STATE_LISTENING: rc = listening_get_next(seq, v); if (!rc) { - tcp_listen_unlock(); + inet_listen_unlock(&tcp_hashinfo); local_bh_disable(); st->state = TCP_SEQ_STATE_ESTABLISHED; rc = established_get_first(seq); @@ -2365,17 +1748,17 @@ static void tcp_seq_stop(struct seq_file *seq, void *v) switch (st->state) { case TCP_SEQ_STATE_OPENREQ: if (v) { - struct tcp_sock *tp = tcp_sk(st->syn_wait_sk); - read_unlock_bh(&tp->accept_queue.syn_wait_lock); + struct inet_connection_sock *icsk = inet_csk(st->syn_wait_sk); + read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock); } case TCP_SEQ_STATE_LISTENING: if (v != SEQ_START_TOKEN) - tcp_listen_unlock(); + inet_listen_unlock(&tcp_hashinfo); break; case TCP_SEQ_STATE_TIME_WAIT: case TCP_SEQ_STATE_ESTABLISHED: if (v) - read_unlock(&tcp_ehash[st->bucket].lock); + read_unlock(&tcp_hashinfo.ehash[st->bucket].lock); local_bh_enable(); break; } @@ -2472,18 +1855,19 @@ static void get_tcp4_sock(struct sock *sp, char *tmpbuf, int i) int timer_active; unsigned long timer_expires; struct tcp_sock *tp = tcp_sk(sp); + const struct inet_connection_sock *icsk = inet_csk(sp); struct inet_sock *inet = inet_sk(sp); unsigned int dest = inet->daddr; unsigned int src = inet->rcv_saddr; __u16 destp = ntohs(inet->dport); __u16 srcp = ntohs(inet->sport); - if (tp->pending == TCP_TIME_RETRANS) { + if (icsk->icsk_pending == ICSK_TIME_RETRANS) { timer_active = 1; - timer_expires = tp->timeout; - } else if (tp->pending == TCP_TIME_PROBE0) { + timer_expires = icsk->icsk_timeout; + } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; - timer_expires = tp->timeout; + timer_expires = icsk->icsk_timeout; } else if (timer_pending(&sp->sk_timer)) { timer_active = 2; timer_expires = sp->sk_timer.expires; @@ -2498,17 +1882,19 @@ static void get_tcp4_sock(struct sock *sp, char *tmpbuf, int i) tp->write_seq - tp->snd_una, tp->rcv_nxt - tp->copied_seq, timer_active, jiffies_to_clock_t(timer_expires - jiffies), - tp->retransmits, + icsk->icsk_retransmits, sock_i_uid(sp), - tp->probes_out, + icsk->icsk_probes_out, sock_i_ino(sp), atomic_read(&sp->sk_refcnt), sp, - tp->rto, tp->ack.ato, (tp->ack.quick << 1) | tp->ack.pingpong, + icsk->icsk_rto, + icsk->icsk_ack.ato, + (icsk->icsk_ack.quick << 1) | icsk->icsk_ack.pingpong, tp->snd_cwnd, tp->snd_ssthresh >= 0xFFFF ? -1 : tp->snd_ssthresh); } -static void get_timewait4_sock(struct tcp_tw_bucket *tw, char *tmpbuf, int i) +static void get_timewait4_sock(struct inet_timewait_sock *tw, char *tmpbuf, int i) { unsigned int dest, src; __u16 destp, srcp; @@ -2588,7 +1974,7 @@ struct proto tcp_prot = { .close = tcp_close, .connect = tcp_v4_connect, .disconnect = tcp_disconnect, - .accept = tcp_accept, + .accept = inet_csk_accept, .ioctl = tcp_ioctl, .init = tcp_v4_init_sock, .destroy = tcp_v4_destroy_sock, @@ -2603,6 +1989,7 @@ struct proto tcp_prot = { .get_port = tcp_v4_get_port, .enter_memory_pressure = tcp_enter_memory_pressure, .sockets_allocated = &tcp_sockets_allocated, + .orphan_count = &tcp_orphan_count, .memory_allocated = &tcp_memory_allocated, .memory_pressure = &tcp_memory_pressure, .sysctl_mem = sysctl_tcp_mem, @@ -2610,6 +1997,7 @@ struct proto tcp_prot = { .sysctl_rmem = sysctl_tcp_rmem, .max_header = MAX_TCP_HEADER, .obj_size = sizeof(struct tcp_sock), + .twsk_obj_size = sizeof(struct tcp_timewait_sock), .rsk_prot = &tcp_request_sock_ops, }; @@ -2631,19 +2019,13 @@ void __init tcp_v4_init(struct net_proto_family *ops) } EXPORT_SYMBOL(ipv4_specific); -EXPORT_SYMBOL(tcp_bind_hash); -EXPORT_SYMBOL(tcp_bucket_create); +EXPORT_SYMBOL(inet_bind_bucket_create); EXPORT_SYMBOL(tcp_hashinfo); -EXPORT_SYMBOL(tcp_inherit_port); -EXPORT_SYMBOL(tcp_listen_wlock); -EXPORT_SYMBOL(tcp_port_rover); EXPORT_SYMBOL(tcp_prot); -EXPORT_SYMBOL(tcp_put_port); EXPORT_SYMBOL(tcp_unhash); EXPORT_SYMBOL(tcp_v4_conn_request); EXPORT_SYMBOL(tcp_v4_connect); EXPORT_SYMBOL(tcp_v4_do_rcv); -EXPORT_SYMBOL(tcp_v4_rebuild_header); EXPORT_SYMBOL(tcp_v4_remember_stamp); EXPORT_SYMBOL(tcp_v4_send_check); EXPORT_SYMBOL(tcp_v4_syn_recv_sock); diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index f42a284164b7..a88db28b0af7 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -35,13 +35,27 @@ #define SYNC_INIT 1 #endif -int sysctl_tcp_tw_recycle; -int sysctl_tcp_max_tw_buckets = NR_FILE*2; - int sysctl_tcp_syncookies = SYNC_INIT; int sysctl_tcp_abort_on_overflow; -static void tcp_tw_schedule(struct tcp_tw_bucket *tw, int timeo); +struct inet_timewait_death_row tcp_death_row = { + .sysctl_max_tw_buckets = NR_FILE * 2, + .period = TCP_TIMEWAIT_LEN / INET_TWDR_TWKILL_SLOTS, + .death_lock = SPIN_LOCK_UNLOCKED, + .hashinfo = &tcp_hashinfo, + .tw_timer = TIMER_INITIALIZER(inet_twdr_hangman, 0, + (unsigned long)&tcp_death_row), + .twkill_work = __WORK_INITIALIZER(tcp_death_row.twkill_work, + inet_twdr_twkill_work, + &tcp_death_row), +/* Short-time timewait calendar */ + + .twcal_hand = -1, + .twcal_timer = TIMER_INITIALIZER(inet_twdr_twcal_tick, 0, + (unsigned long)&tcp_death_row), +}; + +EXPORT_SYMBOL_GPL(tcp_death_row); static __inline__ int tcp_in_window(u32 seq, u32 end_seq, u32 s_win, u32 e_win) { @@ -52,47 +66,6 @@ static __inline__ int tcp_in_window(u32 seq, u32 end_seq, u32 s_win, u32 e_win) return (seq == e_win && seq == end_seq); } -/* New-style handling of TIME_WAIT sockets. */ - -int tcp_tw_count; - - -/* Must be called with locally disabled BHs. */ -static void tcp_timewait_kill(struct tcp_tw_bucket *tw) -{ - struct tcp_ehash_bucket *ehead; - struct tcp_bind_hashbucket *bhead; - struct tcp_bind_bucket *tb; - - /* Unlink from established hashes. */ - ehead = &tcp_ehash[tw->tw_hashent]; - write_lock(&ehead->lock); - if (hlist_unhashed(&tw->tw_node)) { - write_unlock(&ehead->lock); - return; - } - __hlist_del(&tw->tw_node); - sk_node_init(&tw->tw_node); - write_unlock(&ehead->lock); - - /* Disassociate with bind bucket. */ - bhead = &tcp_bhash[tcp_bhashfn(tw->tw_num)]; - spin_lock(&bhead->lock); - tb = tw->tw_tb; - __hlist_del(&tw->tw_bind_node); - tw->tw_tb = NULL; - tcp_bucket_destroy(tb); - spin_unlock(&bhead->lock); - -#ifdef INET_REFCNT_DEBUG - if (atomic_read(&tw->tw_refcnt) != 1) { - printk(KERN_DEBUG "tw_bucket %p refcnt=%d\n", tw, - atomic_read(&tw->tw_refcnt)); - } -#endif - tcp_tw_put(tw); -} - /* * * Main purpose of TIME-WAIT state is to close connection gracefully, * when one of ends sits in LAST-ACK or CLOSING retransmitting FIN @@ -122,19 +95,20 @@ static void tcp_timewait_kill(struct tcp_tw_bucket *tw) * to avoid misread sequence numbers, states etc. --ANK */ enum tcp_tw_status -tcp_timewait_state_process(struct tcp_tw_bucket *tw, struct sk_buff *skb, - struct tcphdr *th, unsigned len) +tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb, + const struct tcphdr *th) { + struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); struct tcp_options_received tmp_opt; int paws_reject = 0; tmp_opt.saw_tstamp = 0; - if (th->doff > (sizeof(struct tcphdr) >> 2) && tw->tw_ts_recent_stamp) { + if (th->doff > (sizeof(*th) >> 2) && tcptw->tw_ts_recent_stamp) { tcp_parse_options(skb, &tmp_opt, 0); if (tmp_opt.saw_tstamp) { - tmp_opt.ts_recent = tw->tw_ts_recent; - tmp_opt.ts_recent_stamp = tw->tw_ts_recent_stamp; + tmp_opt.ts_recent = tcptw->tw_ts_recent; + tmp_opt.ts_recent_stamp = tcptw->tw_ts_recent_stamp; paws_reject = tcp_paws_check(&tmp_opt, th->rst); } } @@ -145,20 +119,20 @@ tcp_timewait_state_process(struct tcp_tw_bucket *tw, struct sk_buff *skb, /* Out of window, send ACK */ if (paws_reject || !tcp_in_window(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq, - tw->tw_rcv_nxt, - tw->tw_rcv_nxt + tw->tw_rcv_wnd)) + tcptw->tw_rcv_nxt, + tcptw->tw_rcv_nxt + tcptw->tw_rcv_wnd)) return TCP_TW_ACK; if (th->rst) goto kill; - if (th->syn && !before(TCP_SKB_CB(skb)->seq, tw->tw_rcv_nxt)) + if (th->syn && !before(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt)) goto kill_with_rst; /* Dup ACK? */ - if (!after(TCP_SKB_CB(skb)->end_seq, tw->tw_rcv_nxt) || + if (!after(TCP_SKB_CB(skb)->end_seq, tcptw->tw_rcv_nxt) || TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq) { - tcp_tw_put(tw); + inet_twsk_put(tw); return TCP_TW_SUCCESS; } @@ -166,19 +140,19 @@ tcp_timewait_state_process(struct tcp_tw_bucket *tw, struct sk_buff *skb, * reset. */ if (!th->fin || - TCP_SKB_CB(skb)->end_seq != tw->tw_rcv_nxt + 1) { + TCP_SKB_CB(skb)->end_seq != tcptw->tw_rcv_nxt + 1) { kill_with_rst: - tcp_tw_deschedule(tw); - tcp_tw_put(tw); + inet_twsk_deschedule(tw, &tcp_death_row); + inet_twsk_put(tw); return TCP_TW_RST; } /* FIN arrived, enter true time-wait state. */ - tw->tw_substate = TCP_TIME_WAIT; - tw->tw_rcv_nxt = TCP_SKB_CB(skb)->end_seq; + tw->tw_substate = TCP_TIME_WAIT; + tcptw->tw_rcv_nxt = TCP_SKB_CB(skb)->end_seq; if (tmp_opt.saw_tstamp) { - tw->tw_ts_recent_stamp = xtime.tv_sec; - tw->tw_ts_recent = tmp_opt.rcv_tsval; + tcptw->tw_ts_recent_stamp = xtime.tv_sec; + tcptw->tw_ts_recent = tmp_opt.rcv_tsval; } /* I am shamed, but failed to make it more elegant. @@ -187,11 +161,13 @@ kill_with_rst: * do not undertsnad recycling in any case, it not * a big problem in practice. --ANK */ if (tw->tw_family == AF_INET && - sysctl_tcp_tw_recycle && tw->tw_ts_recent_stamp && + tcp_death_row.sysctl_tw_recycle && tcptw->tw_ts_recent_stamp && tcp_v4_tw_remember_stamp(tw)) - tcp_tw_schedule(tw, tw->tw_timeout); + inet_twsk_schedule(tw, &tcp_death_row, tw->tw_timeout, + TCP_TIMEWAIT_LEN); else - tcp_tw_schedule(tw, TCP_TIMEWAIT_LEN); + inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN, + TCP_TIMEWAIT_LEN); return TCP_TW_ACK; } @@ -213,7 +189,7 @@ kill_with_rst: */ if (!paws_reject && - (TCP_SKB_CB(skb)->seq == tw->tw_rcv_nxt && + (TCP_SKB_CB(skb)->seq == tcptw->tw_rcv_nxt && (TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq || th->rst))) { /* In window segment, it may be only reset or bare ack. */ @@ -224,19 +200,20 @@ kill_with_rst: */ if (sysctl_tcp_rfc1337 == 0) { kill: - tcp_tw_deschedule(tw); - tcp_tw_put(tw); + inet_twsk_deschedule(tw, &tcp_death_row); + inet_twsk_put(tw); return TCP_TW_SUCCESS; } } - tcp_tw_schedule(tw, TCP_TIMEWAIT_LEN); + inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN, + TCP_TIMEWAIT_LEN); if (tmp_opt.saw_tstamp) { - tw->tw_ts_recent = tmp_opt.rcv_tsval; - tw->tw_ts_recent_stamp = xtime.tv_sec; + tcptw->tw_ts_recent = tmp_opt.rcv_tsval; + tcptw->tw_ts_recent_stamp = xtime.tv_sec; } - tcp_tw_put(tw); + inet_twsk_put(tw); return TCP_TW_SUCCESS; } @@ -258,9 +235,10 @@ kill: */ if (th->syn && !th->rst && !th->ack && !paws_reject && - (after(TCP_SKB_CB(skb)->seq, tw->tw_rcv_nxt) || - (tmp_opt.saw_tstamp && (s32)(tw->tw_ts_recent - tmp_opt.rcv_tsval) < 0))) { - u32 isn = tw->tw_snd_nxt + 65535 + 2; + (after(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt) || + (tmp_opt.saw_tstamp && + (s32)(tcptw->tw_ts_recent - tmp_opt.rcv_tsval) < 0))) { + u32 isn = tcptw->tw_snd_nxt + 65535 + 2; if (isn == 0) isn++; TCP_SKB_CB(skb)->when = isn; @@ -278,107 +256,57 @@ kill: * Do not reschedule in the last case. */ if (paws_reject || th->ack) - tcp_tw_schedule(tw, TCP_TIMEWAIT_LEN); + inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN, + TCP_TIMEWAIT_LEN); /* Send ACK. Note, we do not put the bucket, * it will be released by caller. */ return TCP_TW_ACK; } - tcp_tw_put(tw); + inet_twsk_put(tw); return TCP_TW_SUCCESS; } -/* Enter the time wait state. This is called with locally disabled BH. - * Essentially we whip up a timewait bucket, copy the - * relevant info into it from the SK, and mess with hash chains - * and list linkage. - */ -static void __tcp_tw_hashdance(struct sock *sk, struct tcp_tw_bucket *tw) -{ - struct tcp_ehash_bucket *ehead = &tcp_ehash[sk->sk_hashent]; - struct tcp_bind_hashbucket *bhead; - - /* Step 1: Put TW into bind hash. Original socket stays there too. - Note, that any socket with inet_sk(sk)->num != 0 MUST be bound in - binding cache, even if it is closed. - */ - bhead = &tcp_bhash[tcp_bhashfn(inet_sk(sk)->num)]; - spin_lock(&bhead->lock); - tw->tw_tb = tcp_sk(sk)->bind_hash; - BUG_TRAP(tcp_sk(sk)->bind_hash); - tw_add_bind_node(tw, &tw->tw_tb->owners); - spin_unlock(&bhead->lock); - - write_lock(&ehead->lock); - - /* Step 2: Remove SK from established hash. */ - if (__sk_del_node_init(sk)) - sock_prot_dec_use(sk->sk_prot); - - /* Step 3: Hash TW into TIMEWAIT half of established hash table. */ - tw_add_node(tw, &(ehead + tcp_ehash_size)->chain); - atomic_inc(&tw->tw_refcnt); - - write_unlock(&ehead->lock); -} - /* * Move a socket to time-wait or dead fin-wait-2 state. */ void tcp_time_wait(struct sock *sk, int state, int timeo) { - struct tcp_tw_bucket *tw = NULL; - struct tcp_sock *tp = tcp_sk(sk); + struct inet_timewait_sock *tw = NULL; + const struct tcp_sock *tp = tcp_sk(sk); int recycle_ok = 0; - if (sysctl_tcp_tw_recycle && tp->rx_opt.ts_recent_stamp) + if (tcp_death_row.sysctl_tw_recycle && tp->rx_opt.ts_recent_stamp) recycle_ok = tp->af_specific->remember_stamp(sk); - if (tcp_tw_count < sysctl_tcp_max_tw_buckets) - tw = kmem_cache_alloc(tcp_timewait_cachep, SLAB_ATOMIC); - - if(tw != NULL) { - struct inet_sock *inet = inet_sk(sk); - int rto = (tp->rto<<2) - (tp->rto>>1); - - /* Give us an identity. */ - tw->tw_daddr = inet->daddr; - tw->tw_rcv_saddr = inet->rcv_saddr; - tw->tw_bound_dev_if = sk->sk_bound_dev_if; - tw->tw_num = inet->num; - tw->tw_state = TCP_TIME_WAIT; - tw->tw_substate = state; - tw->tw_sport = inet->sport; - tw->tw_dport = inet->dport; - tw->tw_family = sk->sk_family; - tw->tw_reuse = sk->sk_reuse; - tw->tw_rcv_wscale = tp->rx_opt.rcv_wscale; - atomic_set(&tw->tw_refcnt, 1); + if (tcp_death_row.tw_count < tcp_death_row.sysctl_max_tw_buckets) + tw = inet_twsk_alloc(sk, state); - tw->tw_hashent = sk->sk_hashent; - tw->tw_rcv_nxt = tp->rcv_nxt; - tw->tw_snd_nxt = tp->snd_nxt; - tw->tw_rcv_wnd = tcp_receive_window(tp); - tw->tw_ts_recent = tp->rx_opt.ts_recent; - tw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp; - tw_dead_node_init(tw); + if (tw != NULL) { + struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); + const struct inet_connection_sock *icsk = inet_csk(sk); + const int rto = (icsk->icsk_rto << 2) - (icsk->icsk_rto >> 1); + + tw->tw_rcv_wscale = tp->rx_opt.rcv_wscale; + tcptw->tw_rcv_nxt = tp->rcv_nxt; + tcptw->tw_snd_nxt = tp->snd_nxt; + tcptw->tw_rcv_wnd = tcp_receive_window(tp); + tcptw->tw_ts_recent = tp->rx_opt.ts_recent; + tcptw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp; #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) if (tw->tw_family == PF_INET6) { struct ipv6_pinfo *np = inet6_sk(sk); + struct tcp6_timewait_sock *tcp6tw = tcp6_twsk((struct sock *)tw); - ipv6_addr_copy(&tw->tw_v6_daddr, &np->daddr); - ipv6_addr_copy(&tw->tw_v6_rcv_saddr, &np->rcv_saddr); - tw->tw_v6_ipv6only = np->ipv6only; - } else { - memset(&tw->tw_v6_daddr, 0, sizeof(tw->tw_v6_daddr)); - memset(&tw->tw_v6_rcv_saddr, 0, sizeof(tw->tw_v6_rcv_saddr)); - tw->tw_v6_ipv6only = 0; + ipv6_addr_copy(&tcp6tw->tw_v6_daddr, &np->daddr); + ipv6_addr_copy(&tcp6tw->tw_v6_rcv_saddr, &np->rcv_saddr); + tw->tw_ipv6only = np->ipv6only; } #endif /* Linkage updates. */ - __tcp_tw_hashdance(sk, tw); + __inet_twsk_hashdance(tw, sk, &tcp_hashinfo); /* Get the TIME_WAIT timeout firing. */ if (timeo < rto) @@ -392,8 +320,9 @@ void tcp_time_wait(struct sock *sk, int state, int timeo) timeo = TCP_TIMEWAIT_LEN; } - tcp_tw_schedule(tw, timeo); - tcp_tw_put(tw); + inet_twsk_schedule(tw, &tcp_death_row, timeo, + TCP_TIMEWAIT_LEN); + inet_twsk_put(tw); } else { /* Sorry, if we're out of memory, just CLOSE this * socket up. We've got bigger problems than @@ -407,277 +336,6 @@ void tcp_time_wait(struct sock *sk, int state, int timeo) tcp_done(sk); } -/* Kill off TIME_WAIT sockets once their lifetime has expired. */ -static int tcp_tw_death_row_slot; - -static void tcp_twkill(unsigned long); - -/* TIME_WAIT reaping mechanism. */ -#define TCP_TWKILL_SLOTS 8 /* Please keep this a power of 2. */ -#define TCP_TWKILL_PERIOD (TCP_TIMEWAIT_LEN/TCP_TWKILL_SLOTS) - -#define TCP_TWKILL_QUOTA 100 - -static struct hlist_head tcp_tw_death_row[TCP_TWKILL_SLOTS]; -static DEFINE_SPINLOCK(tw_death_lock); -static struct timer_list tcp_tw_timer = TIMER_INITIALIZER(tcp_twkill, 0, 0); -static void twkill_work(void *); -static DECLARE_WORK(tcp_twkill_work, twkill_work, NULL); -static u32 twkill_thread_slots; - -/* Returns non-zero if quota exceeded. */ -static int tcp_do_twkill_work(int slot, unsigned int quota) -{ - struct tcp_tw_bucket *tw; - struct hlist_node *node; - unsigned int killed; - int ret; - - /* NOTE: compare this to previous version where lock - * was released after detaching chain. It was racy, - * because tw buckets are scheduled in not serialized context - * in 2.3 (with netfilter), and with softnet it is common, because - * soft irqs are not sequenced. - */ - killed = 0; - ret = 0; -rescan: - tw_for_each_inmate(tw, node, &tcp_tw_death_row[slot]) { - __tw_del_dead_node(tw); - spin_unlock(&tw_death_lock); - tcp_timewait_kill(tw); - tcp_tw_put(tw); - killed++; - spin_lock(&tw_death_lock); - if (killed > quota) { - ret = 1; - break; - } - - /* While we dropped tw_death_lock, another cpu may have - * killed off the next TW bucket in the list, therefore - * do a fresh re-read of the hlist head node with the - * lock reacquired. We still use the hlist traversal - * macro in order to get the prefetches. - */ - goto rescan; - } - - tcp_tw_count -= killed; - NET_ADD_STATS_BH(LINUX_MIB_TIMEWAITED, killed); - - return ret; -} - -static void tcp_twkill(unsigned long dummy) -{ - int need_timer, ret; - - spin_lock(&tw_death_lock); - - if (tcp_tw_count == 0) - goto out; - - need_timer = 0; - ret = tcp_do_twkill_work(tcp_tw_death_row_slot, TCP_TWKILL_QUOTA); - if (ret) { - twkill_thread_slots |= (1 << tcp_tw_death_row_slot); - mb(); - schedule_work(&tcp_twkill_work); - need_timer = 1; - } else { - /* We purged the entire slot, anything left? */ - if (tcp_tw_count) - need_timer = 1; - } - tcp_tw_death_row_slot = - ((tcp_tw_death_row_slot + 1) & (TCP_TWKILL_SLOTS - 1)); - if (need_timer) - mod_timer(&tcp_tw_timer, jiffies + TCP_TWKILL_PERIOD); -out: - spin_unlock(&tw_death_lock); -} - -extern void twkill_slots_invalid(void); - -static void twkill_work(void *dummy) -{ - int i; - - if ((TCP_TWKILL_SLOTS - 1) > (sizeof(twkill_thread_slots) * 8)) - twkill_slots_invalid(); - - while (twkill_thread_slots) { - spin_lock_bh(&tw_death_lock); - for (i = 0; i < TCP_TWKILL_SLOTS; i++) { - if (!(twkill_thread_slots & (1 << i))) - continue; - - while (tcp_do_twkill_work(i, TCP_TWKILL_QUOTA) != 0) { - if (need_resched()) { - spin_unlock_bh(&tw_death_lock); - schedule(); - spin_lock_bh(&tw_death_lock); - } - } - - twkill_thread_slots &= ~(1 << i); - } - spin_unlock_bh(&tw_death_lock); - } -} - -/* These are always called from BH context. See callers in - * tcp_input.c to verify this. - */ - -/* This is for handling early-kills of TIME_WAIT sockets. */ -void tcp_tw_deschedule(struct tcp_tw_bucket *tw) -{ - spin_lock(&tw_death_lock); - if (tw_del_dead_node(tw)) { - tcp_tw_put(tw); - if (--tcp_tw_count == 0) - del_timer(&tcp_tw_timer); - } - spin_unlock(&tw_death_lock); - tcp_timewait_kill(tw); -} - -/* Short-time timewait calendar */ - -static int tcp_twcal_hand = -1; -static int tcp_twcal_jiffie; -static void tcp_twcal_tick(unsigned long); -static struct timer_list tcp_twcal_timer = - TIMER_INITIALIZER(tcp_twcal_tick, 0, 0); -static struct hlist_head tcp_twcal_row[TCP_TW_RECYCLE_SLOTS]; - -static void tcp_tw_schedule(struct tcp_tw_bucket *tw, int timeo) -{ - struct hlist_head *list; - int slot; - - /* timeout := RTO * 3.5 - * - * 3.5 = 1+2+0.5 to wait for two retransmits. - * - * RATIONALE: if FIN arrived and we entered TIME-WAIT state, - * our ACK acking that FIN can be lost. If N subsequent retransmitted - * FINs (or previous seqments) are lost (probability of such event - * is p^(N+1), where p is probability to lose single packet and - * time to detect the loss is about RTO*(2^N - 1) with exponential - * backoff). Normal timewait length is calculated so, that we - * waited at least for one retransmitted FIN (maximal RTO is 120sec). - * [ BTW Linux. following BSD, violates this requirement waiting - * only for 60sec, we should wait at least for 240 secs. - * Well, 240 consumes too much of resources 8) - * ] - * This interval is not reduced to catch old duplicate and - * responces to our wandering segments living for two MSLs. - * However, if we use PAWS to detect - * old duplicates, we can reduce the interval to bounds required - * by RTO, rather than MSL. So, if peer understands PAWS, we - * kill tw bucket after 3.5*RTO (it is important that this number - * is greater than TS tick!) and detect old duplicates with help - * of PAWS. - */ - slot = (timeo + (1<<TCP_TW_RECYCLE_TICK) - 1) >> TCP_TW_RECYCLE_TICK; - - spin_lock(&tw_death_lock); - - /* Unlink it, if it was scheduled */ - if (tw_del_dead_node(tw)) - tcp_tw_count--; - else - atomic_inc(&tw->tw_refcnt); - - if (slot >= TCP_TW_RECYCLE_SLOTS) { - /* Schedule to slow timer */ - if (timeo >= TCP_TIMEWAIT_LEN) { - slot = TCP_TWKILL_SLOTS-1; - } else { - slot = (timeo + TCP_TWKILL_PERIOD-1) / TCP_TWKILL_PERIOD; - if (slot >= TCP_TWKILL_SLOTS) - slot = TCP_TWKILL_SLOTS-1; - } - tw->tw_ttd = jiffies + timeo; - slot = (tcp_tw_death_row_slot + slot) & (TCP_TWKILL_SLOTS - 1); - list = &tcp_tw_death_row[slot]; - } else { - tw->tw_ttd = jiffies + (slot << TCP_TW_RECYCLE_TICK); - - if (tcp_twcal_hand < 0) { - tcp_twcal_hand = 0; - tcp_twcal_jiffie = jiffies; - tcp_twcal_timer.expires = tcp_twcal_jiffie + (slot<<TCP_TW_RECYCLE_TICK); - add_timer(&tcp_twcal_timer); - } else { - if (time_after(tcp_twcal_timer.expires, jiffies + (slot<<TCP_TW_RECYCLE_TICK))) - mod_timer(&tcp_twcal_timer, jiffies + (slot<<TCP_TW_RECYCLE_TICK)); - slot = (tcp_twcal_hand + slot)&(TCP_TW_RECYCLE_SLOTS-1); - } - list = &tcp_twcal_row[slot]; - } - - hlist_add_head(&tw->tw_death_node, list); - - if (tcp_tw_count++ == 0) - mod_timer(&tcp_tw_timer, jiffies+TCP_TWKILL_PERIOD); - spin_unlock(&tw_death_lock); -} - -void tcp_twcal_tick(unsigned long dummy) -{ - int n, slot; - unsigned long j; - unsigned long now = jiffies; - int killed = 0; - int adv = 0; - - spin_lock(&tw_death_lock); - if (tcp_twcal_hand < 0) - goto out; - - slot = tcp_twcal_hand; - j = tcp_twcal_jiffie; - - for (n=0; n<TCP_TW_RECYCLE_SLOTS; n++) { - if (time_before_eq(j, now)) { - struct hlist_node *node, *safe; - struct tcp_tw_bucket *tw; - - tw_for_each_inmate_safe(tw, node, safe, - &tcp_twcal_row[slot]) { - __tw_del_dead_node(tw); - tcp_timewait_kill(tw); - tcp_tw_put(tw); - killed++; - } - } else { - if (!adv) { - adv = 1; - tcp_twcal_jiffie = j; - tcp_twcal_hand = slot; - } - - if (!hlist_empty(&tcp_twcal_row[slot])) { - mod_timer(&tcp_twcal_timer, j); - goto out; - } - } - j += (1<<TCP_TW_RECYCLE_TICK); - slot = (slot+1)&(TCP_TW_RECYCLE_SLOTS-1); - } - tcp_twcal_hand = -1; - -out: - if ((tcp_tw_count -= killed) == 0) - del_timer(&tcp_tw_timer); - NET_ADD_STATS_BH(LINUX_MIB_TIMEWAITKILLED, killed); - spin_unlock(&tw_death_lock); -} - /* This is not only more efficient than what we used to do, it eliminates * a lot of code duplication between IPv4/IPv6 SYN recv processing. -DaveM * @@ -686,75 +344,27 @@ out: */ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, struct sk_buff *skb) { - /* allocate the newsk from the same slab of the master sock, - * if not, at sk_free time we'll try to free it from the wrong - * slabcache (i.e. is it TCPv4 or v6?), this is handled thru sk->sk_prot -acme */ - struct sock *newsk = sk_alloc(PF_INET, GFP_ATOMIC, sk->sk_prot, 0); + struct sock *newsk = inet_csk_clone(sk, req, GFP_ATOMIC); - if(newsk != NULL) { - struct inet_request_sock *ireq = inet_rsk(req); + if (newsk != NULL) { + const struct inet_request_sock *ireq = inet_rsk(req); struct tcp_request_sock *treq = tcp_rsk(req); + struct inet_connection_sock *newicsk = inet_csk(sk); struct tcp_sock *newtp; - struct sk_filter *filter; - - memcpy(newsk, sk, sizeof(struct tcp_sock)); - newsk->sk_state = TCP_SYN_RECV; - - /* SANITY */ - sk_node_init(&newsk->sk_node); - tcp_sk(newsk)->bind_hash = NULL; - - /* Clone the TCP header template */ - inet_sk(newsk)->dport = ireq->rmt_port; - - sock_lock_init(newsk); - bh_lock_sock(newsk); - - rwlock_init(&newsk->sk_dst_lock); - atomic_set(&newsk->sk_rmem_alloc, 0); - skb_queue_head_init(&newsk->sk_receive_queue); - atomic_set(&newsk->sk_wmem_alloc, 0); - skb_queue_head_init(&newsk->sk_write_queue); - atomic_set(&newsk->sk_omem_alloc, 0); - newsk->sk_wmem_queued = 0; - newsk->sk_forward_alloc = 0; - - sock_reset_flag(newsk, SOCK_DONE); - newsk->sk_userlocks = sk->sk_userlocks & ~SOCK_BINDPORT_LOCK; - newsk->sk_backlog.head = newsk->sk_backlog.tail = NULL; - newsk->sk_send_head = NULL; - rwlock_init(&newsk->sk_callback_lock); - skb_queue_head_init(&newsk->sk_error_queue); - newsk->sk_write_space = sk_stream_write_space; - - if ((filter = newsk->sk_filter) != NULL) - sk_filter_charge(newsk, filter); - - if (unlikely(xfrm_sk_clone_policy(newsk))) { - /* It is still raw copy of parent, so invalidate - * destructor and make plain sk_free() */ - newsk->sk_destruct = NULL; - sk_free(newsk); - return NULL; - } /* Now setup tcp_sock */ newtp = tcp_sk(newsk); newtp->pred_flags = 0; newtp->rcv_nxt = treq->rcv_isn + 1; - newtp->snd_nxt = treq->snt_isn + 1; - newtp->snd_una = treq->snt_isn + 1; - newtp->snd_sml = treq->snt_isn + 1; + newtp->snd_nxt = newtp->snd_una = newtp->snd_sml = treq->snt_isn + 1; tcp_prequeue_init(newtp); tcp_init_wl(newtp, treq->snt_isn, treq->rcv_isn); - newtp->retransmits = 0; - newtp->backoff = 0; newtp->srtt = 0; newtp->mdev = TCP_TIMEOUT_INIT; - newtp->rto = TCP_TIMEOUT_INIT; + newicsk->icsk_rto = TCP_TIMEOUT_INIT; newtp->packets_out = 0; newtp->left_out = 0; @@ -774,9 +384,9 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, newtp->frto_counter = 0; newtp->frto_highmark = 0; - newtp->ca_ops = &tcp_reno; + newicsk->icsk_ca_ops = &tcp_reno; - tcp_set_ca_state(newtp, TCP_CA_Open); + tcp_set_ca_state(newsk, TCP_CA_Open); tcp_init_xmit_timers(newsk); skb_queue_head_init(&newtp->out_of_order_queue); newtp->rcv_wup = treq->rcv_isn + 1; @@ -789,26 +399,12 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, newtp->rx_opt.dsack = 0; newtp->rx_opt.eff_sacks = 0; - newtp->probes_out = 0; newtp->rx_opt.num_sacks = 0; newtp->urg_data = 0; - /* Deinitialize accept_queue to trap illegal accesses. */ - memset(&newtp->accept_queue, 0, sizeof(newtp->accept_queue)); - - /* Back to base struct sock members. */ - newsk->sk_err = 0; - newsk->sk_priority = 0; - atomic_set(&newsk->sk_refcnt, 2); -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet_sock_nr); -#endif - atomic_inc(&tcp_sockets_allocated); if (sock_flag(newsk, SOCK_KEEPOPEN)) - tcp_reset_keepalive_timer(newsk, - keepalive_time_when(newtp)); - newsk->sk_socket = NULL; - newsk->sk_sleep = NULL; + inet_csk_reset_keepalive_timer(newsk, + keepalive_time_when(newtp)); newtp->rx_opt.tstamp_ok = ireq->tstamp_ok; if((newtp->rx_opt.sack_ok = ireq->sack_ok) != 0) { @@ -838,7 +434,7 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, newtp->tcp_header_len = sizeof(struct tcphdr); } if (skb->len >= TCP_MIN_RCVMSS+newtp->tcp_header_len) - newtp->ack.last_seg_size = skb->len-newtp->tcp_header_len; + newicsk->icsk_ack.last_seg_size = skb->len - newtp->tcp_header_len; newtp->rx_opt.mss_clamp = req->mss; TCP_ECN_openreq_child(newtp, req); if (newtp->ecn_flags&TCP_ECN_OK) @@ -934,9 +530,10 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb, does sequence test, SYN is truncated, and thus we consider it a bare ACK. - If tp->defer_accept, we silently drop this bare ACK. Otherwise, - we create an established connection. Both ends (listening sockets) - accept the new incoming connection and try to talk to each other. 8-) + If icsk->icsk_accept_queue.rskq_defer_accept, we silently drop this + bare ACK. Otherwise, we create an established connection. Both + ends (listening sockets) accept the new incoming connection and try + to talk to each other. 8-) Note: This case is both harmless, and rare. Possibility is about the same as us discovering intelligent life on another plant tomorrow. @@ -1003,7 +600,8 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb, return NULL; /* If TCP_DEFER_ACCEPT is set, drop bare ACK. */ - if (tp->defer_accept && TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) { + if (inet_csk(sk)->icsk_accept_queue.rskq_defer_accept && + TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) { inet_rsk(req)->acked = 1; return NULL; } @@ -1018,10 +616,10 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb, if (child == NULL) goto listen_overflow; - tcp_synq_unlink(tp, req, prev); - tcp_synq_removed(sk, req); + inet_csk_reqsk_queue_unlink(sk, req, prev); + inet_csk_reqsk_queue_removed(sk, req); - tcp_acceptq_queue(sk, req, child); + inet_csk_reqsk_queue_add(sk, req, child); return child; listen_overflow: @@ -1035,7 +633,7 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb, if (!(flg & TCP_FLAG_RST)) req->rsk_ops->send_reset(skb); - tcp_synq_drop(sk, req, prev); + inet_csk_reqsk_queue_drop(sk, req, prev); return NULL; } @@ -1074,4 +672,3 @@ EXPORT_SYMBOL(tcp_check_req); EXPORT_SYMBOL(tcp_child_process); EXPORT_SYMBOL(tcp_create_openreq_child); EXPORT_SYMBOL(tcp_timewait_state_process); -EXPORT_SYMBOL(tcp_tw_deschedule); diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index dd30dd137b74..75b68116682a 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -105,18 +105,19 @@ static __u16 tcp_advertise_mss(struct sock *sk) /* RFC2861. Reset CWND after idle period longer RTO to "restart window". * This is the first part of cwnd validation mechanism. */ -static void tcp_cwnd_restart(struct tcp_sock *tp, struct dst_entry *dst) +static void tcp_cwnd_restart(struct sock *sk, struct dst_entry *dst) { + struct tcp_sock *tp = tcp_sk(sk); s32 delta = tcp_time_stamp - tp->lsndtime; u32 restart_cwnd = tcp_init_cwnd(tp, dst); u32 cwnd = tp->snd_cwnd; - tcp_ca_event(tp, CA_EVENT_CWND_RESTART); + tcp_ca_event(sk, CA_EVENT_CWND_RESTART); - tp->snd_ssthresh = tcp_current_ssthresh(tp); + tp->snd_ssthresh = tcp_current_ssthresh(sk); restart_cwnd = min(restart_cwnd, cwnd); - while ((delta -= tp->rto) > 0 && cwnd > restart_cwnd) + while ((delta -= inet_csk(sk)->icsk_rto) > 0 && cwnd > restart_cwnd) cwnd >>= 1; tp->snd_cwnd = max(cwnd, restart_cwnd); tp->snd_cwnd_stamp = tcp_time_stamp; @@ -126,26 +127,25 @@ static void tcp_cwnd_restart(struct tcp_sock *tp, struct dst_entry *dst) static inline void tcp_event_data_sent(struct tcp_sock *tp, struct sk_buff *skb, struct sock *sk) { - u32 now = tcp_time_stamp; + struct inet_connection_sock *icsk = inet_csk(sk); + const u32 now = tcp_time_stamp; - if (!tp->packets_out && (s32)(now - tp->lsndtime) > tp->rto) - tcp_cwnd_restart(tp, __sk_dst_get(sk)); + if (!tp->packets_out && (s32)(now - tp->lsndtime) > icsk->icsk_rto) + tcp_cwnd_restart(sk, __sk_dst_get(sk)); tp->lsndtime = now; /* If it is a reply for ato after last received * packet, enter pingpong mode. */ - if ((u32)(now - tp->ack.lrcvtime) < tp->ack.ato) - tp->ack.pingpong = 1; + if ((u32)(now - icsk->icsk_ack.lrcvtime) < icsk->icsk_ack.ato) + icsk->icsk_ack.pingpong = 1; } static __inline__ void tcp_event_ack_sent(struct sock *sk, unsigned int pkts) { - struct tcp_sock *tp = tcp_sk(sk); - - tcp_dec_quickack_mode(tp, pkts); - tcp_clear_xmit_timer(sk, TCP_TIME_DACK); + tcp_dec_quickack_mode(sk, pkts); + inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK); } /* Determine a window scaling and initial window to offer. @@ -265,6 +265,7 @@ static __inline__ u16 tcp_select_window(struct sock *sk) static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb) { if (skb != NULL) { + const struct inet_connection_sock *icsk = inet_csk(sk); struct inet_sock *inet = inet_sk(sk); struct tcp_sock *tp = tcp_sk(sk); struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); @@ -280,8 +281,8 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb) #define SYSCTL_FLAG_SACK 0x4 /* If congestion control is doing timestamping */ - if (tp->ca_ops->rtt_sample) - do_gettimeofday(&skb->stamp); + if (icsk->icsk_ca_ops->rtt_sample) + __net_timestamp(skb); sysctl_flags = 0; if (tcb->flags & TCPCB_FLAG_SYN) { @@ -308,7 +309,7 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb) } if (tcp_packets_in_flight(tp) == 0) - tcp_ca_event(tp, CA_EVENT_TX_START); + tcp_ca_event(sk, CA_EVENT_TX_START); th = (struct tcphdr *) skb_push(skb, tcp_header_size); skb->h.th = th; @@ -366,7 +367,7 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb) if (err <= 0) return err; - tcp_enter_cwr(tp); + tcp_enter_cwr(sk); /* NET_XMIT_CN is special. It does not guarantee, * that this packet is lost. It tells that device @@ -482,7 +483,7 @@ static int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned * skbs, which it never sent before. --ANK */ TCP_SKB_CB(buff)->when = TCP_SKB_CB(skb)->when; - buff->stamp = skb->stamp; + buff->tstamp = skb->tstamp; if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST) { tp->lost_out -= tcp_skb_pcount(skb); @@ -505,7 +506,7 @@ static int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned /* Link BUFF into the send queue. */ skb_header_release(buff); - __skb_append(skb, buff); + __skb_append(skb, buff, &sk->sk_write_queue); return 0; } @@ -696,7 +697,7 @@ static inline void tcp_cwnd_validate(struct sock *sk, struct tcp_sock *tp) if (tp->packets_out > tp->snd_cwnd_used) tp->snd_cwnd_used = tp->packets_out; - if ((s32)(tcp_time_stamp - tp->snd_cwnd_stamp) >= tp->rto) + if ((s32)(tcp_time_stamp - tp->snd_cwnd_stamp) >= inet_csk(sk)->icsk_rto) tcp_cwnd_application_limited(sk); } } @@ -893,7 +894,7 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, /* Link BUFF into the send queue. */ skb_header_release(buff); - __skb_append(skb, buff); + __skb_append(skb, buff, &sk->sk_write_queue); return 0; } @@ -905,12 +906,13 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, */ static int tcp_tso_should_defer(struct sock *sk, struct tcp_sock *tp, struct sk_buff *skb) { + const struct inet_connection_sock *icsk = inet_csk(sk); u32 send_win, cong_win, limit, in_flight; if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN) return 0; - if (tp->ca_state != TCP_CA_Open) + if (icsk->icsk_ca_state != TCP_CA_Open) return 0; in_flight = tcp_packets_in_flight(tp); @@ -1147,6 +1149,7 @@ void tcp_push_one(struct sock *sk, unsigned int mss_now) */ u32 __tcp_select_window(struct sock *sk) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); /* MSS for the peer's data. Previous verions used mss_clamp * here. I don't know if the value based on our guesses @@ -1154,7 +1157,7 @@ u32 __tcp_select_window(struct sock *sk) * but may be worse for the performance because of rcv_mss * fluctuations. --SAW 1998/11/1 */ - int mss = tp->ack.rcv_mss; + int mss = icsk->icsk_ack.rcv_mss; int free_space = tcp_space(sk); int full_space = min_t(int, tp->window_clamp, tcp_full_space(sk)); int window; @@ -1163,7 +1166,7 @@ u32 __tcp_select_window(struct sock *sk) mss = full_space; if (free_space < full_space/2) { - tp->ack.quick = 0; + icsk->icsk_ack.quick = 0; if (tcp_memory_pressure) tp->rcv_ssthresh = min(tp->rcv_ssthresh, 4U*tp->advmss); @@ -1238,7 +1241,7 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int m tcp_skb_pcount(next_skb) != 1); /* Ok. We will be able to collapse the packet. */ - __skb_unlink(next_skb, next_skb->list); + __skb_unlink(next_skb, &sk->sk_write_queue); memcpy(skb_put(skb, next_skb_size), next_skb->data, next_skb_size); @@ -1286,6 +1289,7 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int m */ void tcp_simple_retransmit(struct sock *sk) { + const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; unsigned int mss = tcp_current_mss(sk, 0); @@ -1316,12 +1320,12 @@ void tcp_simple_retransmit(struct sock *sk) * in network, but units changed and effective * cwnd/ssthresh really reduced now. */ - if (tp->ca_state != TCP_CA_Loss) { + if (icsk->icsk_ca_state != TCP_CA_Loss) { tp->high_seq = tp->snd_nxt; - tp->snd_ssthresh = tcp_current_ssthresh(tp); + tp->snd_ssthresh = tcp_current_ssthresh(sk); tp->prior_ssthresh = 0; tp->undo_marker = 0; - tcp_set_ca_state(tp, TCP_CA_Loss); + tcp_set_ca_state(sk, TCP_CA_Loss); } tcp_xmit_retransmit_queue(sk); } @@ -1461,6 +1465,7 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb) */ void tcp_xmit_retransmit_queue(struct sock *sk) { + const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; int packet_cnt = tp->lost_out; @@ -1484,14 +1489,16 @@ void tcp_xmit_retransmit_queue(struct sock *sk) if (!(sacked&(TCPCB_SACKED_ACKED|TCPCB_SACKED_RETRANS))) { if (tcp_retransmit_skb(sk, skb)) return; - if (tp->ca_state != TCP_CA_Loss) + if (icsk->icsk_ca_state != TCP_CA_Loss) NET_INC_STATS_BH(LINUX_MIB_TCPFASTRETRANS); else NET_INC_STATS_BH(LINUX_MIB_TCPSLOWSTARTRETRANS); if (skb == skb_peek(&sk->sk_write_queue)) - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, tp->rto); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, + TCP_RTO_MAX); } packet_cnt -= tcp_skb_pcount(skb); @@ -1504,7 +1511,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk) /* OK, demanded retransmission is finished. */ /* Forward retransmissions are possible only during Recovery. */ - if (tp->ca_state != TCP_CA_Recovery) + if (icsk->icsk_ca_state != TCP_CA_Recovery) return; /* No forward retransmissions in Reno are possible. */ @@ -1544,7 +1551,9 @@ void tcp_xmit_retransmit_queue(struct sock *sk) break; if (skb == skb_peek(&sk->sk_write_queue)) - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, tp->rto); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, + TCP_RTO_MAX); NET_INC_STATS_BH(LINUX_MIB_TCPFORWARDRETRANS); } @@ -1573,7 +1582,7 @@ void tcp_send_fin(struct sock *sk) } else { /* Socket is locked, keep trying until memory is available. */ for (;;) { - skb = alloc_skb(MAX_TCP_HEADER, GFP_KERNEL); + skb = alloc_skb_fclone(MAX_TCP_HEADER, GFP_KERNEL); if (skb) break; yield(); @@ -1780,8 +1789,8 @@ static inline void tcp_connect_init(struct sock *sk) tp->rcv_wup = 0; tp->copied_seq = 0; - tp->rto = TCP_TIMEOUT_INIT; - tp->retransmits = 0; + inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT; + inet_csk(sk)->icsk_retransmits = 0; tcp_clear_retrans(tp); } @@ -1795,7 +1804,7 @@ int tcp_connect(struct sock *sk) tcp_connect_init(sk); - buff = alloc_skb(MAX_TCP_HEADER + 15, sk->sk_allocation); + buff = alloc_skb_fclone(MAX_TCP_HEADER + 15, sk->sk_allocation); if (unlikely(buff == NULL)) return -ENOBUFS; @@ -1824,7 +1833,8 @@ int tcp_connect(struct sock *sk) TCP_INC_STATS(TCP_MIB_ACTIVEOPENS); /* Timer for repeating the SYN until an answer. */ - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, tp->rto); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, TCP_RTO_MAX); return 0; } @@ -1834,20 +1844,21 @@ int tcp_connect(struct sock *sk) */ void tcp_send_delayed_ack(struct sock *sk) { - struct tcp_sock *tp = tcp_sk(sk); - int ato = tp->ack.ato; + struct inet_connection_sock *icsk = inet_csk(sk); + int ato = icsk->icsk_ack.ato; unsigned long timeout; if (ato > TCP_DELACK_MIN) { + const struct tcp_sock *tp = tcp_sk(sk); int max_ato = HZ/2; - if (tp->ack.pingpong || (tp->ack.pending&TCP_ACK_PUSHED)) + if (icsk->icsk_ack.pingpong || (icsk->icsk_ack.pending & ICSK_ACK_PUSHED)) max_ato = TCP_DELACK_MAX; /* Slow path, intersegment interval is "high". */ /* If some rtt estimate is known, use it to bound delayed ack. - * Do not use tp->rto here, use results of rtt measurements + * Do not use inet_csk(sk)->icsk_rto here, use results of rtt measurements * directly. */ if (tp->srtt) { @@ -1864,21 +1875,22 @@ void tcp_send_delayed_ack(struct sock *sk) timeout = jiffies + ato; /* Use new timeout only if there wasn't a older one earlier. */ - if (tp->ack.pending&TCP_ACK_TIMER) { + if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) { /* If delack timer was blocked or is about to expire, * send ACK now. */ - if (tp->ack.blocked || time_before_eq(tp->ack.timeout, jiffies+(ato>>2))) { + if (icsk->icsk_ack.blocked || + time_before_eq(icsk->icsk_ack.timeout, jiffies + (ato >> 2))) { tcp_send_ack(sk); return; } - if (!time_before(timeout, tp->ack.timeout)) - timeout = tp->ack.timeout; + if (!time_before(timeout, icsk->icsk_ack.timeout)) + timeout = icsk->icsk_ack.timeout; } - tp->ack.pending |= TCP_ACK_SCHED|TCP_ACK_TIMER; - tp->ack.timeout = timeout; - sk_reset_timer(sk, &tp->delack_timer, timeout); + icsk->icsk_ack.pending |= ICSK_ACK_SCHED | ICSK_ACK_TIMER; + icsk->icsk_ack.timeout = timeout; + sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout); } /* This routine sends an ack and also updates the window. */ @@ -1895,9 +1907,10 @@ void tcp_send_ack(struct sock *sk) */ buff = alloc_skb(MAX_TCP_HEADER, GFP_ATOMIC); if (buff == NULL) { - tcp_schedule_ack(tp); - tp->ack.ato = TCP_ATO_MIN; - tcp_reset_xmit_timer(sk, TCP_TIME_DACK, TCP_DELACK_MAX); + inet_csk_schedule_ack(sk); + inet_csk(sk)->icsk_ack.ato = TCP_ATO_MIN; + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + TCP_DELACK_MAX, TCP_RTO_MAX); return; } @@ -2011,6 +2024,7 @@ int tcp_write_wakeup(struct sock *sk) */ void tcp_send_probe0(struct sock *sk) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); int err; @@ -2018,28 +2032,31 @@ void tcp_send_probe0(struct sock *sk) if (tp->packets_out || !sk->sk_send_head) { /* Cancel probe timer, if it is not required. */ - tp->probes_out = 0; - tp->backoff = 0; + icsk->icsk_probes_out = 0; + icsk->icsk_backoff = 0; return; } if (err <= 0) { - if (tp->backoff < sysctl_tcp_retries2) - tp->backoff++; - tp->probes_out++; - tcp_reset_xmit_timer (sk, TCP_TIME_PROBE0, - min(tp->rto << tp->backoff, TCP_RTO_MAX)); + if (icsk->icsk_backoff < sysctl_tcp_retries2) + icsk->icsk_backoff++; + icsk->icsk_probes_out++; + inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, + min(icsk->icsk_rto << icsk->icsk_backoff, TCP_RTO_MAX), + TCP_RTO_MAX); } else { /* If packet was not sent due to local congestion, - * do not backoff and do not remember probes_out. + * do not backoff and do not remember icsk_probes_out. * Let local senders to fight for local resources. * * Use accumulated backoff yet. */ - if (!tp->probes_out) - tp->probes_out=1; - tcp_reset_xmit_timer (sk, TCP_TIME_PROBE0, - min(tp->rto << tp->backoff, TCP_RESOURCE_PROBE_INTERVAL)); + if (!icsk->icsk_probes_out) + icsk->icsk_probes_out = 1; + inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, + min(icsk->icsk_rto << icsk->icsk_backoff, + TCP_RESOURCE_PROBE_INTERVAL), + TCP_RTO_MAX); } } diff --git a/net/ipv4/tcp_scalable.c b/net/ipv4/tcp_scalable.c index 70e108e15c71..327770bf5522 100644 --- a/net/ipv4/tcp_scalable.c +++ b/net/ipv4/tcp_scalable.c @@ -16,9 +16,10 @@ #define TCP_SCALABLE_AI_CNT 50U #define TCP_SCALABLE_MD_SCALE 3 -static void tcp_scalable_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, +static void tcp_scalable_cong_avoid(struct sock *sk, u32 ack, u32 rtt, u32 in_flight, int flag) { + struct tcp_sock *tp = tcp_sk(sk); if (in_flight < tp->snd_cwnd) return; @@ -35,8 +36,9 @@ static void tcp_scalable_cong_avoid(struct tcp_sock *tp, u32 ack, u32 rtt, tp->snd_cwnd_stamp = tcp_time_stamp; } -static u32 tcp_scalable_ssthresh(struct tcp_sock *tp) +static u32 tcp_scalable_ssthresh(struct sock *sk) { + const struct tcp_sock *tp = tcp_sk(sk); return max(tp->snd_cwnd - (tp->snd_cwnd>>TCP_SCALABLE_MD_SCALE), 2U); } diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index 0084227438c2..415ee47ac1c5 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -36,49 +36,13 @@ static void tcp_write_timer(unsigned long); static void tcp_delack_timer(unsigned long); static void tcp_keepalive_timer (unsigned long data); -#ifdef TCP_DEBUG -const char tcp_timer_bug_msg[] = KERN_DEBUG "tcpbug: unknown timer value\n"; -EXPORT_SYMBOL(tcp_timer_bug_msg); -#endif - -/* - * Using different timers for retransmit, delayed acks and probes - * We may wish use just one timer maintaining a list of expire jiffies - * to optimize. - */ - void tcp_init_xmit_timers(struct sock *sk) { - struct tcp_sock *tp = tcp_sk(sk); - - init_timer(&tp->retransmit_timer); - tp->retransmit_timer.function=&tcp_write_timer; - tp->retransmit_timer.data = (unsigned long) sk; - tp->pending = 0; - - init_timer(&tp->delack_timer); - tp->delack_timer.function=&tcp_delack_timer; - tp->delack_timer.data = (unsigned long) sk; - tp->ack.pending = 0; - - init_timer(&sk->sk_timer); - sk->sk_timer.function = &tcp_keepalive_timer; - sk->sk_timer.data = (unsigned long)sk; + inet_csk_init_xmit_timers(sk, &tcp_write_timer, &tcp_delack_timer, + &tcp_keepalive_timer); } -void tcp_clear_xmit_timers(struct sock *sk) -{ - struct tcp_sock *tp = tcp_sk(sk); - - tp->pending = 0; - sk_stop_timer(sk, &tp->retransmit_timer); - - tp->ack.pending = 0; - tp->ack.blocked = 0; - sk_stop_timer(sk, &tp->delack_timer); - - sk_stop_timer(sk, &sk->sk_timer); -} +EXPORT_SYMBOL(tcp_init_xmit_timers); static void tcp_write_err(struct sock *sk) { @@ -155,15 +119,15 @@ static int tcp_orphan_retries(struct sock *sk, int alive) /* A write timeout has occurred. Process the after effects. */ static int tcp_write_timeout(struct sock *sk) { - struct tcp_sock *tp = tcp_sk(sk); + const struct inet_connection_sock *icsk = inet_csk(sk); int retry_until; if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) { - if (tp->retransmits) + if (icsk->icsk_retransmits) dst_negative_advice(&sk->sk_dst_cache); - retry_until = tp->syn_retries ? : sysctl_tcp_syn_retries; + retry_until = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; } else { - if (tp->retransmits >= sysctl_tcp_retries1) { + if (icsk->icsk_retransmits >= sysctl_tcp_retries1) { /* NOTE. draft-ietf-tcpimpl-pmtud-01.txt requires pmtu black hole detection. :-( @@ -189,16 +153,16 @@ static int tcp_write_timeout(struct sock *sk) retry_until = sysctl_tcp_retries2; if (sock_flag(sk, SOCK_DEAD)) { - int alive = (tp->rto < TCP_RTO_MAX); + const int alive = (icsk->icsk_rto < TCP_RTO_MAX); retry_until = tcp_orphan_retries(sk, alive); - if (tcp_out_of_resources(sk, alive || tp->retransmits < retry_until)) + if (tcp_out_of_resources(sk, alive || icsk->icsk_retransmits < retry_until)) return 1; } } - if (tp->retransmits >= retry_until) { + if (icsk->icsk_retransmits >= retry_until) { /* Has it gone just too far? */ tcp_write_err(sk); return 1; @@ -210,26 +174,27 @@ static void tcp_delack_timer(unsigned long data) { struct sock *sk = (struct sock*)data; struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); bh_lock_sock(sk); if (sock_owned_by_user(sk)) { /* Try again later. */ - tp->ack.blocked = 1; + icsk->icsk_ack.blocked = 1; NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKLOCKED); - sk_reset_timer(sk, &tp->delack_timer, jiffies + TCP_DELACK_MIN); + sk_reset_timer(sk, &icsk->icsk_delack_timer, jiffies + TCP_DELACK_MIN); goto out_unlock; } sk_stream_mem_reclaim(sk); - if (sk->sk_state == TCP_CLOSE || !(tp->ack.pending & TCP_ACK_TIMER)) + if (sk->sk_state == TCP_CLOSE || !(icsk->icsk_ack.pending & ICSK_ACK_TIMER)) goto out; - if (time_after(tp->ack.timeout, jiffies)) { - sk_reset_timer(sk, &tp->delack_timer, tp->ack.timeout); + if (time_after(icsk->icsk_ack.timeout, jiffies)) { + sk_reset_timer(sk, &icsk->icsk_delack_timer, icsk->icsk_ack.timeout); goto out; } - tp->ack.pending &= ~TCP_ACK_TIMER; + icsk->icsk_ack.pending &= ~ICSK_ACK_TIMER; if (!skb_queue_empty(&tp->ucopy.prequeue)) { struct sk_buff *skb; @@ -242,16 +207,16 @@ static void tcp_delack_timer(unsigned long data) tp->ucopy.memory = 0; } - if (tcp_ack_scheduled(tp)) { - if (!tp->ack.pingpong) { + if (inet_csk_ack_scheduled(sk)) { + if (!icsk->icsk_ack.pingpong) { /* Delayed ACK missed: inflate ATO. */ - tp->ack.ato = min(tp->ack.ato << 1, tp->rto); + icsk->icsk_ack.ato = min(icsk->icsk_ack.ato << 1, icsk->icsk_rto); } else { /* Delayed ACK missed: leave pingpong mode and * deflate ATO. */ - tp->ack.pingpong = 0; - tp->ack.ato = TCP_ATO_MIN; + icsk->icsk_ack.pingpong = 0; + icsk->icsk_ack.ato = TCP_ATO_MIN; } tcp_send_ack(sk); NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKS); @@ -268,11 +233,12 @@ out_unlock: static void tcp_probe_timer(struct sock *sk) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); int max_probes; if (tp->packets_out || !sk->sk_send_head) { - tp->probes_out = 0; + icsk->icsk_probes_out = 0; return; } @@ -283,7 +249,7 @@ static void tcp_probe_timer(struct sock *sk) * FIXME: We ought not to do it, Solaris 2.5 actually has fixing * this behaviour in Solaris down as a bug fix. [AC] * - * Let me to explain. probes_out is zeroed by incoming ACKs + * Let me to explain. icsk_probes_out is zeroed by incoming ACKs * even if they advertise zero window. Hence, connection is killed only * if we received no ACKs for normal connection timeout. It is not killed * only because window stays zero for some time, window may be zero @@ -294,15 +260,15 @@ static void tcp_probe_timer(struct sock *sk) max_probes = sysctl_tcp_retries2; if (sock_flag(sk, SOCK_DEAD)) { - int alive = ((tp->rto<<tp->backoff) < TCP_RTO_MAX); + const int alive = ((icsk->icsk_rto << icsk->icsk_backoff) < TCP_RTO_MAX); max_probes = tcp_orphan_retries(sk, alive); - if (tcp_out_of_resources(sk, alive || tp->probes_out <= max_probes)) + if (tcp_out_of_resources(sk, alive || icsk->icsk_probes_out <= max_probes)) return; } - if (tp->probes_out > max_probes) { + if (icsk->icsk_probes_out > max_probes) { tcp_write_err(sk); } else { /* Only send another probe if we didn't close things up. */ @@ -317,6 +283,7 @@ static void tcp_probe_timer(struct sock *sk) static void tcp_retransmit_timer(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); if (!tp->packets_out) goto out; @@ -351,20 +318,21 @@ static void tcp_retransmit_timer(struct sock *sk) if (tcp_write_timeout(sk)) goto out; - if (tp->retransmits == 0) { - if (tp->ca_state == TCP_CA_Disorder || tp->ca_state == TCP_CA_Recovery) { + if (icsk->icsk_retransmits == 0) { + if (icsk->icsk_ca_state == TCP_CA_Disorder || + icsk->icsk_ca_state == TCP_CA_Recovery) { if (tp->rx_opt.sack_ok) { - if (tp->ca_state == TCP_CA_Recovery) + if (icsk->icsk_ca_state == TCP_CA_Recovery) NET_INC_STATS_BH(LINUX_MIB_TCPSACKRECOVERYFAIL); else NET_INC_STATS_BH(LINUX_MIB_TCPSACKFAILURES); } else { - if (tp->ca_state == TCP_CA_Recovery) + if (icsk->icsk_ca_state == TCP_CA_Recovery) NET_INC_STATS_BH(LINUX_MIB_TCPRENORECOVERYFAIL); else NET_INC_STATS_BH(LINUX_MIB_TCPRENOFAILURES); } - } else if (tp->ca_state == TCP_CA_Loss) { + } else if (icsk->icsk_ca_state == TCP_CA_Loss) { NET_INC_STATS_BH(LINUX_MIB_TCPLOSSFAILURES); } else { NET_INC_STATS_BH(LINUX_MIB_TCPTIMEOUTS); @@ -381,10 +349,11 @@ static void tcp_retransmit_timer(struct sock *sk) /* Retransmission failed because of local congestion, * do not backoff. */ - if (!tp->retransmits) - tp->retransmits=1; - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, - min(tp->rto, TCP_RESOURCE_PROBE_INTERVAL)); + if (!icsk->icsk_retransmits) + icsk->icsk_retransmits = 1; + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + min(icsk->icsk_rto, TCP_RESOURCE_PROBE_INTERVAL), + TCP_RTO_MAX); goto out; } @@ -403,13 +372,13 @@ static void tcp_retransmit_timer(struct sock *sk) * implemented ftp to mars will work nicely. We will have to fix * the 120 second clamps though! */ - tp->backoff++; - tp->retransmits++; + icsk->icsk_backoff++; + icsk->icsk_retransmits++; out_reset_timer: - tp->rto = min(tp->rto << 1, TCP_RTO_MAX); - tcp_reset_xmit_timer(sk, TCP_TIME_RETRANS, tp->rto); - if (tp->retransmits > sysctl_tcp_retries1) + icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto, TCP_RTO_MAX); + if (icsk->icsk_retransmits > sysctl_tcp_retries1) __sk_dst_reset(sk); out:; @@ -418,32 +387,32 @@ out:; static void tcp_write_timer(unsigned long data) { struct sock *sk = (struct sock*)data; - struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); int event; bh_lock_sock(sk); if (sock_owned_by_user(sk)) { /* Try again later */ - sk_reset_timer(sk, &tp->retransmit_timer, jiffies + (HZ / 20)); + sk_reset_timer(sk, &icsk->icsk_retransmit_timer, jiffies + (HZ / 20)); goto out_unlock; } - if (sk->sk_state == TCP_CLOSE || !tp->pending) + if (sk->sk_state == TCP_CLOSE || !icsk->icsk_pending) goto out; - if (time_after(tp->timeout, jiffies)) { - sk_reset_timer(sk, &tp->retransmit_timer, tp->timeout); + if (time_after(icsk->icsk_timeout, jiffies)) { + sk_reset_timer(sk, &icsk->icsk_retransmit_timer, icsk->icsk_timeout); goto out; } - event = tp->pending; - tp->pending = 0; + event = icsk->icsk_pending; + icsk->icsk_pending = 0; switch (event) { - case TCP_TIME_RETRANS: + case ICSK_TIME_RETRANS: tcp_retransmit_timer(sk); break; - case TCP_TIME_PROBE0: + case ICSK_TIME_PROBE0: tcp_probe_timer(sk); break; } @@ -462,96 +431,8 @@ out_unlock: static void tcp_synack_timer(struct sock *sk) { - struct tcp_sock *tp = tcp_sk(sk); - struct listen_sock *lopt = tp->accept_queue.listen_opt; - int max_retries = tp->syn_retries ? : sysctl_tcp_synack_retries; - int thresh = max_retries; - unsigned long now = jiffies; - struct request_sock **reqp, *req; - int i, budget; - - if (lopt == NULL || lopt->qlen == 0) - return; - - /* Normally all the openreqs are young and become mature - * (i.e. converted to established socket) for first timeout. - * If synack was not acknowledged for 3 seconds, it means - * one of the following things: synack was lost, ack was lost, - * rtt is high or nobody planned to ack (i.e. synflood). - * When server is a bit loaded, queue is populated with old - * open requests, reducing effective size of queue. - * When server is well loaded, queue size reduces to zero - * after several minutes of work. It is not synflood, - * it is normal operation. The solution is pruning - * too old entries overriding normal timeout, when - * situation becomes dangerous. - * - * Essentially, we reserve half of room for young - * embrions; and abort old ones without pity, if old - * ones are about to clog our table. - */ - if (lopt->qlen>>(lopt->max_qlen_log-1)) { - int young = (lopt->qlen_young<<1); - - while (thresh > 2) { - if (lopt->qlen < young) - break; - thresh--; - young <<= 1; - } - } - - if (tp->defer_accept) - max_retries = tp->defer_accept; - - budget = 2*(TCP_SYNQ_HSIZE/(TCP_TIMEOUT_INIT/TCP_SYNQ_INTERVAL)); - i = lopt->clock_hand; - - do { - reqp=&lopt->syn_table[i]; - while ((req = *reqp) != NULL) { - if (time_after_eq(now, req->expires)) { - if ((req->retrans < thresh || - (inet_rsk(req)->acked && req->retrans < max_retries)) - && !req->rsk_ops->rtx_syn_ack(sk, req, NULL)) { - unsigned long timeo; - - if (req->retrans++ == 0) - lopt->qlen_young--; - timeo = min((TCP_TIMEOUT_INIT << req->retrans), - TCP_RTO_MAX); - req->expires = now + timeo; - reqp = &req->dl_next; - continue; - } - - /* Drop this request */ - tcp_synq_unlink(tp, req, reqp); - reqsk_queue_removed(&tp->accept_queue, req); - reqsk_free(req); - continue; - } - reqp = &req->dl_next; - } - - i = (i+1)&(TCP_SYNQ_HSIZE-1); - - } while (--budget > 0); - - lopt->clock_hand = i; - - if (lopt->qlen) - tcp_reset_keepalive_timer(sk, TCP_SYNQ_INTERVAL); -} - -void tcp_delete_keepalive_timer (struct sock *sk) -{ - sk_stop_timer(sk, &sk->sk_timer); -} - -void tcp_reset_keepalive_timer (struct sock *sk, unsigned long len) -{ - sk_reset_timer(sk, &sk->sk_timer, jiffies + len); + inet_csk_reqsk_queue_prune(sk, TCP_SYNQ_INTERVAL, + TCP_TIMEOUT_INIT, TCP_RTO_MAX); } void tcp_set_keepalive(struct sock *sk, int val) @@ -560,15 +441,16 @@ void tcp_set_keepalive(struct sock *sk, int val) return; if (val && !sock_flag(sk, SOCK_KEEPOPEN)) - tcp_reset_keepalive_timer(sk, keepalive_time_when(tcp_sk(sk))); + inet_csk_reset_keepalive_timer(sk, keepalive_time_when(tcp_sk(sk))); else if (!val) - tcp_delete_keepalive_timer(sk); + inet_csk_delete_keepalive_timer(sk); } static void tcp_keepalive_timer (unsigned long data) { struct sock *sk = (struct sock *) data; + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); __u32 elapsed; @@ -576,7 +458,7 @@ static void tcp_keepalive_timer (unsigned long data) bh_lock_sock(sk); if (sock_owned_by_user(sk)) { /* Try again later. */ - tcp_reset_keepalive_timer (sk, HZ/20); + inet_csk_reset_keepalive_timer (sk, HZ/20); goto out; } @@ -587,7 +469,7 @@ static void tcp_keepalive_timer (unsigned long data) if (sk->sk_state == TCP_FIN_WAIT2 && sock_flag(sk, SOCK_DEAD)) { if (tp->linger2 >= 0) { - int tmo = tcp_fin_time(tp) - TCP_TIMEWAIT_LEN; + const int tmo = tcp_fin_time(sk) - TCP_TIMEWAIT_LEN; if (tmo > 0) { tcp_time_wait(sk, TCP_FIN_WAIT2, tmo); @@ -610,14 +492,14 @@ static void tcp_keepalive_timer (unsigned long data) elapsed = tcp_time_stamp - tp->rcv_tstamp; if (elapsed >= keepalive_time_when(tp)) { - if ((!tp->keepalive_probes && tp->probes_out >= sysctl_tcp_keepalive_probes) || - (tp->keepalive_probes && tp->probes_out >= tp->keepalive_probes)) { + if ((!tp->keepalive_probes && icsk->icsk_probes_out >= sysctl_tcp_keepalive_probes) || + (tp->keepalive_probes && icsk->icsk_probes_out >= tp->keepalive_probes)) { tcp_send_active_reset(sk, GFP_ATOMIC); tcp_write_err(sk); goto out; } if (tcp_write_wakeup(sk) <= 0) { - tp->probes_out++; + icsk->icsk_probes_out++; elapsed = keepalive_intvl_when(tp); } else { /* If keepalive was lost due to local congestion, @@ -634,7 +516,7 @@ static void tcp_keepalive_timer (unsigned long data) sk_stream_mem_reclaim(sk); resched: - tcp_reset_keepalive_timer (sk, elapsed); + inet_csk_reset_keepalive_timer (sk, elapsed); goto out; death: @@ -644,8 +526,3 @@ out: bh_unlock_sock(sk); sock_put(sk); } - -EXPORT_SYMBOL(tcp_clear_xmit_timers); -EXPORT_SYMBOL(tcp_delete_keepalive_timer); -EXPORT_SYMBOL(tcp_init_xmit_timers); -EXPORT_SYMBOL(tcp_reset_keepalive_timer); diff --git a/net/ipv4/tcp_vegas.c b/net/ipv4/tcp_vegas.c index 9bd443db5193..93c5f92070f9 100644 --- a/net/ipv4/tcp_vegas.c +++ b/net/ipv4/tcp_vegas.c @@ -35,7 +35,7 @@ #include <linux/mm.h> #include <linux/module.h> #include <linux/skbuff.h> -#include <linux/tcp_diag.h> +#include <linux/inet_diag.h> #include <net/tcp.h> @@ -82,9 +82,10 @@ struct vegas { * Instead we must wait until the completion of an RTT during * which we actually receive ACKs. */ -static inline void vegas_enable(struct tcp_sock *tp) +static inline void vegas_enable(struct sock *sk) { - struct vegas *vegas = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + struct vegas *vegas = inet_csk_ca(sk); /* Begin taking Vegas samples next time we send something. */ vegas->doing_vegas_now = 1; @@ -97,19 +98,19 @@ static inline void vegas_enable(struct tcp_sock *tp) } /* Stop taking Vegas samples for now. */ -static inline void vegas_disable(struct tcp_sock *tp) +static inline void vegas_disable(struct sock *sk) { - struct vegas *vegas = tcp_ca(tp); + struct vegas *vegas = inet_csk_ca(sk); vegas->doing_vegas_now = 0; } -static void tcp_vegas_init(struct tcp_sock *tp) +static void tcp_vegas_init(struct sock *sk) { - struct vegas *vegas = tcp_ca(tp); + struct vegas *vegas = inet_csk_ca(sk); vegas->baseRTT = 0x7fffffff; - vegas_enable(tp); + vegas_enable(sk); } /* Do RTT sampling needed for Vegas. @@ -120,9 +121,9 @@ static void tcp_vegas_init(struct tcp_sock *tp) * o min-filter RTT samples from a much longer window (forever for now) * to find the propagation delay (baseRTT) */ -static void tcp_vegas_rtt_calc(struct tcp_sock *tp, u32 usrtt) +static void tcp_vegas_rtt_calc(struct sock *sk, u32 usrtt) { - struct vegas *vegas = tcp_ca(tp); + struct vegas *vegas = inet_csk_ca(sk); u32 vrtt = usrtt + 1; /* Never allow zero rtt or baseRTT */ /* Filter to find propagation delay: */ @@ -136,13 +137,13 @@ static void tcp_vegas_rtt_calc(struct tcp_sock *tp, u32 usrtt) vegas->cntRTT++; } -static void tcp_vegas_state(struct tcp_sock *tp, u8 ca_state) +static void tcp_vegas_state(struct sock *sk, u8 ca_state) { if (ca_state == TCP_CA_Open) - vegas_enable(tp); + vegas_enable(sk); else - vegas_disable(tp); + vegas_disable(sk); } /* @@ -154,20 +155,21 @@ static void tcp_vegas_state(struct tcp_sock *tp, u8 ca_state) * packets, _then_ we can make Vegas calculations * again. */ -static void tcp_vegas_cwnd_event(struct tcp_sock *tp, enum tcp_ca_event event) +static void tcp_vegas_cwnd_event(struct sock *sk, enum tcp_ca_event event) { if (event == CA_EVENT_CWND_RESTART || event == CA_EVENT_TX_START) - tcp_vegas_init(tp); + tcp_vegas_init(sk); } -static void tcp_vegas_cong_avoid(struct tcp_sock *tp, u32 ack, +static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 seq_rtt, u32 in_flight, int flag) { - struct vegas *vegas = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct vegas *vegas = inet_csk_ca(sk); if (!vegas->doing_vegas_now) - return tcp_reno_cong_avoid(tp, ack, seq_rtt, in_flight, flag); + return tcp_reno_cong_avoid(sk, ack, seq_rtt, in_flight, flag); /* The key players are v_beg_snd_una and v_beg_snd_nxt. * @@ -219,7 +221,7 @@ static void tcp_vegas_cong_avoid(struct tcp_sock *tp, u32 ack, * but that's not too awful, since we're taking the min, * rather than averaging. */ - tcp_vegas_rtt_calc(tp, seq_rtt*1000); + tcp_vegas_rtt_calc(sk, seq_rtt * 1000); /* We do the Vegas calculations only if we got enough RTT * samples that we can be reasonably sure that we got @@ -359,14 +361,14 @@ static void tcp_vegas_cong_avoid(struct tcp_sock *tp, u32 ack, } /* Extract info for Tcp socket info provided via netlink. */ -static void tcp_vegas_get_info(struct tcp_sock *tp, u32 ext, +static void tcp_vegas_get_info(struct sock *sk, u32 ext, struct sk_buff *skb) { - const struct vegas *ca = tcp_ca(tp); - if (ext & (1<<(TCPDIAG_VEGASINFO-1))) { + const struct vegas *ca = inet_csk_ca(sk); + if (ext & (1 << (INET_DIAG_VEGASINFO - 1))) { struct tcpvegas_info *info; - info = RTA_DATA(__RTA_PUT(skb, TCPDIAG_VEGASINFO, + info = RTA_DATA(__RTA_PUT(skb, INET_DIAG_VEGASINFO, sizeof(*info))); info->tcpv_enabled = ca->doing_vegas_now; @@ -393,7 +395,7 @@ static struct tcp_congestion_ops tcp_vegas = { static int __init tcp_vegas_register(void) { - BUG_ON(sizeof(struct vegas) > TCP_CA_PRIV_SIZE); + BUG_ON(sizeof(struct vegas) > ICSK_CA_PRIV_SIZE); tcp_register_congestion_control(&tcp_vegas); return 0; } diff --git a/net/ipv4/tcp_westwood.c b/net/ipv4/tcp_westwood.c index ef827242c940..0c340c3756c2 100644 --- a/net/ipv4/tcp_westwood.c +++ b/net/ipv4/tcp_westwood.c @@ -8,7 +8,7 @@ #include <linux/mm.h> #include <linux/module.h> #include <linux/skbuff.h> -#include <linux/tcp_diag.h> +#include <linux/inet_diag.h> #include <net/tcp.h> /* TCP Westwood structure */ @@ -40,9 +40,9 @@ struct westwood { * way as soon as possible. It will reasonably happen within the first * RTT period of the connection lifetime. */ -static void tcp_westwood_init(struct tcp_sock *tp) +static void tcp_westwood_init(struct sock *sk) { - struct westwood *w = tcp_ca(tp); + struct westwood *w = inet_csk_ca(sk); w->bk = 0; w->bw_ns_est = 0; @@ -51,7 +51,7 @@ static void tcp_westwood_init(struct tcp_sock *tp) w->cumul_ack = 0; w->rtt_min = w->rtt = TCP_WESTWOOD_INIT_RTT; w->rtt_win_sx = tcp_time_stamp; - w->snd_una = tp->snd_una; + w->snd_una = tcp_sk(sk)->snd_una; } /* @@ -74,11 +74,11 @@ static inline void westwood_filter(struct westwood *w, u32 delta) * Called after processing group of packets. * but all westwood needs is the last sample of srtt. */ -static void tcp_westwood_pkts_acked(struct tcp_sock *tp, u32 cnt) +static void tcp_westwood_pkts_acked(struct sock *sk, u32 cnt) { - struct westwood *w = tcp_ca(tp); + struct westwood *w = inet_csk_ca(sk); if (cnt > 0) - w->rtt = tp->srtt >> 3; + w->rtt = tcp_sk(sk)->srtt >> 3; } /* @@ -86,9 +86,9 @@ static void tcp_westwood_pkts_acked(struct tcp_sock *tp, u32 cnt) * It updates RTT evaluation window if it is the right moment to do * it. If so it calls filter for evaluating bandwidth. */ -static void westwood_update_window(struct tcp_sock *tp) +static void westwood_update_window(struct sock *sk) { - struct westwood *w = tcp_ca(tp); + struct westwood *w = inet_csk_ca(sk); s32 delta = tcp_time_stamp - w->rtt_win_sx; /* @@ -114,11 +114,12 @@ static void westwood_update_window(struct tcp_sock *tp) * header prediction is successful. In such case in fact update is * straight forward and doesn't need any particular care. */ -static inline void westwood_fast_bw(struct tcp_sock *tp) +static inline void westwood_fast_bw(struct sock *sk) { - struct westwood *w = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + struct westwood *w = inet_csk_ca(sk); - westwood_update_window(tp); + westwood_update_window(sk); w->bk += tp->snd_una - w->snd_una; w->snd_una = tp->snd_una; @@ -130,9 +131,10 @@ static inline void westwood_fast_bw(struct tcp_sock *tp) * This function evaluates cumul_ack for evaluating bk in case of * delayed or partial acks. */ -static inline u32 westwood_acked_count(struct tcp_sock *tp) +static inline u32 westwood_acked_count(struct sock *sk) { - struct westwood *w = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + struct westwood *w = inet_csk_ca(sk); w->cumul_ack = tp->snd_una - w->snd_una; @@ -160,9 +162,10 @@ static inline u32 westwood_acked_count(struct tcp_sock *tp) return w->cumul_ack; } -static inline u32 westwood_bw_rttmin(const struct tcp_sock *tp) +static inline u32 westwood_bw_rttmin(const struct sock *sk) { - struct westwood *w = tcp_ca(tp); + const struct tcp_sock *tp = tcp_sk(sk); + const struct westwood *w = inet_csk_ca(sk); return max_t(u32, (w->bw_est * w->rtt_min) / tp->mss_cache, 2); } @@ -172,31 +175,32 @@ static inline u32 westwood_bw_rttmin(const struct tcp_sock *tp) * in packets we use mss_cache). Rttmin is guaranteed to be >= 2 * so avoids ever returning 0. */ -static u32 tcp_westwood_cwnd_min(struct tcp_sock *tp) +static u32 tcp_westwood_cwnd_min(struct sock *sk) { - return westwood_bw_rttmin(tp); + return westwood_bw_rttmin(sk); } -static void tcp_westwood_event(struct tcp_sock *tp, enum tcp_ca_event event) +static void tcp_westwood_event(struct sock *sk, enum tcp_ca_event event) { - struct westwood *w = tcp_ca(tp); + struct tcp_sock *tp = tcp_sk(sk); + struct westwood *w = inet_csk_ca(sk); switch(event) { case CA_EVENT_FAST_ACK: - westwood_fast_bw(tp); + westwood_fast_bw(sk); break; case CA_EVENT_COMPLETE_CWR: - tp->snd_cwnd = tp->snd_ssthresh = westwood_bw_rttmin(tp); + tp->snd_cwnd = tp->snd_ssthresh = westwood_bw_rttmin(sk); break; case CA_EVENT_FRTO: - tp->snd_ssthresh = westwood_bw_rttmin(tp); + tp->snd_ssthresh = westwood_bw_rttmin(sk); break; case CA_EVENT_SLOW_ACK: - westwood_update_window(tp); - w->bk += westwood_acked_count(tp); + westwood_update_window(sk); + w->bk += westwood_acked_count(sk); w->rtt_min = min(w->rtt, w->rtt_min); break; @@ -208,15 +212,15 @@ static void tcp_westwood_event(struct tcp_sock *tp, enum tcp_ca_event event) /* Extract info for Tcp socket info provided via netlink. */ -static void tcp_westwood_info(struct tcp_sock *tp, u32 ext, +static void tcp_westwood_info(struct sock *sk, u32 ext, struct sk_buff *skb) { - const struct westwood *ca = tcp_ca(tp); - if (ext & (1<<(TCPDIAG_VEGASINFO-1))) { + const struct westwood *ca = inet_csk_ca(sk); + if (ext & (1 << (INET_DIAG_VEGASINFO - 1))) { struct rtattr *rta; struct tcpvegas_info *info; - rta = __RTA_PUT(skb, TCPDIAG_VEGASINFO, sizeof(*info)); + rta = __RTA_PUT(skb, INET_DIAG_VEGASINFO, sizeof(*info)); info = RTA_DATA(rta); info->tcpv_enabled = 1; info->tcpv_rttcnt = 0; @@ -242,7 +246,7 @@ static struct tcp_congestion_ops tcp_westwood = { static int __init tcp_westwood_register(void) { - BUG_ON(sizeof(struct westwood) > TCP_CA_PRIV_SIZE); + BUG_ON(sizeof(struct westwood) > ICSK_CA_PRIV_SIZE); return tcp_register_congestion_control(&tcp_westwood); } diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index dc4d07357e3a..e5beca7de86c 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -95,7 +95,8 @@ #include <linux/ipv6.h> #include <linux/netdevice.h> #include <net/snmp.h> -#include <net/tcp.h> +#include <net/ip.h> +#include <net/tcp_states.h> #include <net/protocol.h> #include <linux/skbuff.h> #include <linux/proc_fs.h> @@ -112,7 +113,7 @@ * Snmp MIB for the UDP layer */ -DEFINE_SNMP_STAT(struct udp_mib, udp_statistics); +DEFINE_SNMP_STAT(struct udp_mib, udp_statistics) __read_mostly; struct hlist_head udp_hash[UDP_HTABLE_SIZE]; DEFINE_RWLOCK(udp_hash_lock); @@ -628,7 +629,7 @@ back_from_confirm: /* ... which is an evident application bug. --ANK */ release_sock(sk); - LIMIT_NETDEBUG(printk(KERN_DEBUG "udp cork app bug 2\n")); + LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n"); err = -EINVAL; goto out; } @@ -693,7 +694,7 @@ static int udp_sendpage(struct sock *sk, struct page *page, int offset, if (unlikely(!up->pending)) { release_sock(sk); - LIMIT_NETDEBUG(printk(KERN_DEBUG "udp cork app bug 3\n")); + LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 3\n"); return -EINVAL; } @@ -1102,7 +1103,7 @@ static int udp_checksum_init(struct sk_buff *skb, struct udphdr *uh, skb->ip_summed = CHECKSUM_UNNECESSARY; if (!udp_check(uh, ulen, saddr, daddr, skb->csum)) return 0; - LIMIT_NETDEBUG(printk(KERN_DEBUG "udp v4 hw csum failure.\n")); + LIMIT_NETDEBUG(KERN_DEBUG "udp v4 hw csum failure.\n"); skb->ip_summed = CHECKSUM_NONE; } if (skb->ip_summed != CHECKSUM_UNNECESSARY) @@ -1181,13 +1182,13 @@ int udp_rcv(struct sk_buff *skb) return(0); short_packet: - LIMIT_NETDEBUG(printk(KERN_DEBUG "UDP: short packet: From %u.%u.%u.%u:%u %d/%d to %u.%u.%u.%u:%u\n", - NIPQUAD(saddr), - ntohs(uh->source), - ulen, - len, - NIPQUAD(daddr), - ntohs(uh->dest))); + LIMIT_NETDEBUG(KERN_DEBUG "UDP: short packet: From %u.%u.%u.%u:%u %d/%d to %u.%u.%u.%u:%u\n", + NIPQUAD(saddr), + ntohs(uh->source), + ulen, + len, + NIPQUAD(daddr), + ntohs(uh->dest)); no_header: UDP_INC_STATS_BH(UDP_MIB_INERRORS); kfree_skb(skb); @@ -1198,12 +1199,12 @@ csum_error: * RFC1122: OK. Discards the bad packet silently (as far as * the network is concerned, anyway) as per 4.1.3.4 (MUST). */ - LIMIT_NETDEBUG(printk(KERN_DEBUG "UDP: bad checksum. From %d.%d.%d.%d:%d to %d.%d.%d.%d:%d ulen %d\n", - NIPQUAD(saddr), - ntohs(uh->source), - NIPQUAD(daddr), - ntohs(uh->dest), - ulen)); + LIMIT_NETDEBUG(KERN_DEBUG "UDP: bad checksum. From %d.%d.%d.%d:%d to %d.%d.%d.%d:%d ulen %d\n", + NIPQUAD(saddr), + ntohs(uh->source), + NIPQUAD(daddr), + ntohs(uh->dest), + ulen); drop: UDP_INC_STATS_BH(UDP_MIB_INERRORS); kfree_skb(skb); diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c index 050611d7a967..d23e07fc81fa 100644 --- a/net/ipv4/xfrm4_state.c +++ b/net/ipv4/xfrm4_state.c @@ -128,8 +128,10 @@ void __init xfrm4_state_init(void) xfrm_state_register_afinfo(&xfrm4_state_afinfo); } +#if 0 void __exit xfrm4_state_fini(void) { xfrm_state_unregister_afinfo(&xfrm4_state_afinfo); } +#endif /* 0 */ diff --git a/net/ipv6/Makefile b/net/ipv6/Makefile index b39e04940590..6460eec834b7 100644 --- a/net/ipv6/Makefile +++ b/net/ipv6/Makefile @@ -8,7 +8,7 @@ ipv6-objs := af_inet6.o anycast.o ip6_output.o ip6_input.o addrconf.o sit.o \ route.o ip6_fib.o ipv6_sockglue.o ndisc.o udp.o raw.o \ protocol.o icmp.o mcast.o reassembly.o tcp_ipv6.o \ exthdrs.o sysctl_net_ipv6.o datagram.o proc.o \ - ip6_flowlabel.o ipv6_syms.o + ip6_flowlabel.o ipv6_syms.o netfilter.o ipv6-$(CONFIG_XFRM) += xfrm6_policy.o xfrm6_state.o xfrm6_input.o \ xfrm6_output.o @@ -23,3 +23,5 @@ obj-$(CONFIG_NETFILTER) += netfilter/ obj-$(CONFIG_IPV6_TUNNEL) += ip6_tunnel.o obj-y += exthdrs_core.o + +obj-$(subst m,y,$(CONFIG_IPV6)) += inet6_hashtables.o diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index 77004b9456c0..937ad32db77c 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -1041,9 +1041,9 @@ int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2) const struct in6_addr *sk_rcv_saddr6 = &inet6_sk(sk)->rcv_saddr; const struct in6_addr *sk2_rcv_saddr6 = tcp_v6_rcv_saddr(sk2); u32 sk_rcv_saddr = inet_sk(sk)->rcv_saddr; - u32 sk2_rcv_saddr = tcp_v4_rcv_saddr(sk2); + u32 sk2_rcv_saddr = inet_rcv_saddr(sk2); int sk_ipv6only = ipv6_only_sock(sk); - int sk2_ipv6only = tcp_v6_ipv6only(sk2); + int sk2_ipv6only = inet_v6_ipv6only(sk2); int addr_type = ipv6_addr_type(sk_rcv_saddr6); int addr_type2 = sk2_rcv_saddr6 ? ipv6_addr_type(sk2_rcv_saddr6) : IPV6_ADDR_MAPPED; @@ -1126,7 +1126,7 @@ void addrconf_leave_solict(struct inet6_dev *idev, struct in6_addr *addr) __ipv6_dev_mc_dec(idev, &maddr); } -void addrconf_join_anycast(struct inet6_ifaddr *ifp) +static void addrconf_join_anycast(struct inet6_ifaddr *ifp) { struct in6_addr addr; ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len); @@ -1135,7 +1135,7 @@ void addrconf_join_anycast(struct inet6_ifaddr *ifp) ipv6_dev_ac_inc(ifp->idev->dev, &addr); } -void addrconf_leave_anycast(struct inet6_ifaddr *ifp) +static void addrconf_leave_anycast(struct inet6_ifaddr *ifp) { struct in6_addr addr; ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len); @@ -2858,16 +2858,16 @@ static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa) skb = alloc_skb(size, GFP_ATOMIC); if (!skb) { - netlink_set_err(rtnl, 0, RTMGRP_IPV6_IFADDR, ENOBUFS); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_IFADDR, ENOBUFS); return; } if (inet6_fill_ifaddr(skb, ifa, current->pid, 0, event, 0) < 0) { kfree_skb(skb); - netlink_set_err(rtnl, 0, RTMGRP_IPV6_IFADDR, EINVAL); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_IFADDR, EINVAL); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_IPV6_IFADDR; - netlink_broadcast(rtnl, skb, 0, RTMGRP_IPV6_IFADDR, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = RTNLGRP_IPV6_IFADDR; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV6_IFADDR, GFP_ATOMIC); } static void inline ipv6_store_devconf(struct ipv6_devconf *cnf, @@ -2994,16 +2994,16 @@ void inet6_ifinfo_notify(int event, struct inet6_dev *idev) skb = alloc_skb(size, GFP_ATOMIC); if (!skb) { - netlink_set_err(rtnl, 0, RTMGRP_IPV6_IFINFO, ENOBUFS); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_IFINFO, ENOBUFS); return; } if (inet6_fill_ifinfo(skb, idev, current->pid, 0, event, 0) < 0) { kfree_skb(skb); - netlink_set_err(rtnl, 0, RTMGRP_IPV6_IFINFO, EINVAL); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_IFINFO, EINVAL); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_IPV6_IFINFO; - netlink_broadcast(rtnl, skb, 0, RTMGRP_IPV6_IFINFO, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = RTNLGRP_IPV6_IFINFO; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV6_IFINFO, GFP_ATOMIC); } static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev, @@ -3054,16 +3054,16 @@ static void inet6_prefix_notify(int event, struct inet6_dev *idev, skb = alloc_skb(size, GFP_ATOMIC); if (!skb) { - netlink_set_err(rtnl, 0, RTMGRP_IPV6_PREFIX, ENOBUFS); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_PREFIX, ENOBUFS); return; } if (inet6_fill_prefix(skb, idev, pinfo, current->pid, 0, event, 0) < 0) { kfree_skb(skb); - netlink_set_err(rtnl, 0, RTMGRP_IPV6_PREFIX, EINVAL); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_PREFIX, EINVAL); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_IPV6_PREFIX; - netlink_broadcast(rtnl, skb, 0, RTMGRP_IPV6_PREFIX, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = RTNLGRP_IPV6_PREFIX; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV6_PREFIX, GFP_ATOMIC); } static struct rtnetlink_link inet6_rtnetlink_table[RTM_NR_MSGTYPES] = { diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 28d9bcab0970..4f8795af2edb 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -44,6 +44,7 @@ #include <linux/netdevice.h> #include <linux/icmpv6.h> #include <linux/smp_lock.h> +#include <linux/netfilter_ipv6.h> #include <net/ip.h> #include <net/ipv6.h> @@ -66,45 +67,14 @@ MODULE_AUTHOR("Cast of dozens"); MODULE_DESCRIPTION("IPv6 protocol stack for Linux"); MODULE_LICENSE("GPL"); -/* IPv6 procfs goodies... */ - -#ifdef CONFIG_PROC_FS -extern int raw6_proc_init(void); -extern void raw6_proc_exit(void); -extern int tcp6_proc_init(void); -extern void tcp6_proc_exit(void); -extern int udp6_proc_init(void); -extern void udp6_proc_exit(void); -extern int ipv6_misc_proc_init(void); -extern void ipv6_misc_proc_exit(void); -extern int ac6_proc_init(void); -extern void ac6_proc_exit(void); -extern int if6_proc_init(void); -extern void if6_proc_exit(void); -#endif - int sysctl_ipv6_bindv6only; -#ifdef INET_REFCNT_DEBUG -atomic_t inet6_sock_nr; -EXPORT_SYMBOL(inet6_sock_nr); -#endif - /* The inetsw table contains everything that inet_create needs to * build a new socket. */ static struct list_head inetsw6[SOCK_MAX]; static DEFINE_SPINLOCK(inetsw6_lock); -static void inet6_sock_destruct(struct sock *sk) -{ - inet_sock_destruct(sk); - -#ifdef INET_REFCNT_DEBUG - atomic_dec(&inet6_sock_nr); -#endif -} - static __inline__ struct ipv6_pinfo *inet6_sk_generic(struct sock *sk) { const int offset = sk->sk_prot->obj_size - sizeof(struct ipv6_pinfo); @@ -185,7 +155,7 @@ static int inet6_create(struct socket *sock, int protocol) inet->hdrincl = 1; } - sk->sk_destruct = inet6_sock_destruct; + sk->sk_destruct = inet_sock_destruct; sk->sk_family = PF_INET6; sk->sk_protocol = protocol; @@ -212,12 +182,17 @@ static int inet6_create(struct socket *sock, int protocol) inet->pmtudisc = IP_PMTUDISC_DONT; else inet->pmtudisc = IP_PMTUDISC_WANT; + /* + * Increment only the relevant sk_prot->socks debug field, this changes + * the previous behaviour of incrementing both the equivalent to + * answer->prot->socks (inet6_sock_nr) and inet_sock_nr. + * + * This allows better debug granularity as we'll know exactly how many + * UDPv6, TCPv6, etc socks were allocated, not the sum of all IPv6 + * transport protocol socks. -acme + */ + sk_refcnt_debug_inc(sk); - -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet6_sock_nr); - atomic_inc(&inet_sock_nr); -#endif if (inet->num) { /* It assumes that any protocol which allows * the user to assign a number at socket @@ -513,11 +488,6 @@ static struct net_proto_family inet6_family_ops = { .owner = THIS_MODULE, }; -#ifdef CONFIG_SYSCTL -extern void ipv6_sysctl_register(void); -extern void ipv6_sysctl_unregister(void); -#endif - /* Same as inet6_dgram_ops, sans udp_poll. */ static struct proto_ops inet6_sockraw_ops = { .family = PF_INET6, @@ -684,8 +654,6 @@ static void cleanup_ipv6_mibs(void) snmp6_mib_free((void **)udp_stats_in6); } -extern int ipv6_misc_proc_init(void); - static int __init inet6_init(void) { struct sk_buff *dummy_skb; @@ -757,6 +725,9 @@ static int __init inet6_init(void) err = igmp6_init(&inet6_family_ops); if (err) goto igmp_fail; + err = ipv6_netfilter_init(); + if (err) + goto netfilter_fail; /* Create /proc/foo6 entries. */ #ifdef CONFIG_PROC_FS err = -ENOMEM; @@ -813,6 +784,8 @@ proc_tcp6_fail: raw6_proc_exit(); proc_raw6_fail: #endif + ipv6_netfilter_fini(); +netfilter_fail: igmp6_cleanup(); igmp_fail: ndisc_cleanup(); @@ -852,6 +825,7 @@ static void __exit inet6_exit(void) ip6_route_cleanup(); ipv6_packet_cleanup(); igmp6_cleanup(); + ipv6_netfilter_fini(); ndisc_cleanup(); icmpv6_cleanup(); #ifdef CONFIG_SYSCTL diff --git a/net/ipv6/ah6.c b/net/ipv6/ah6.c index 986fdfdccbcd..0ebfad907a03 100644 --- a/net/ipv6/ah6.c +++ b/net/ipv6/ah6.c @@ -131,10 +131,10 @@ static int ipv6_clear_mutable_options(struct ipv6hdr *iph, int len) case NEXTHDR_HOP: case NEXTHDR_DEST: if (!zero_out_mutable_opts(exthdr.opth)) { - LIMIT_NETDEBUG(printk( + LIMIT_NETDEBUG( KERN_WARNING "overrun %sopts\n", nexthdr == NEXTHDR_HOP ? - "hop" : "dest")); + "hop" : "dest"); return -EINVAL; } break; @@ -293,8 +293,7 @@ static int ah6_input(struct xfrm_state *x, struct xfrm_decap_state *decap, struc skb_push(skb, skb->data - skb->nh.raw); ahp->icv(ahp, skb, ah->auth_data); if (memcmp(ah->auth_data, auth_data, ahp->icv_trunc_len)) { - LIMIT_NETDEBUG( - printk(KERN_WARNING "ipsec ah authentication error\n")); + LIMIT_NETDEBUG(KERN_WARNING "ipsec ah authentication error\n"); x->stats.integrity_failed++; goto free_out; } @@ -332,9 +331,9 @@ static void ah6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, if (!x) return; - NETDEBUG(printk(KERN_DEBUG "pmtu discovery on SA AH/%08x/" - "%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x\n", - ntohl(ah->spi), NIP6(iph->daddr))); + NETDEBUG(KERN_DEBUG "pmtu discovery on SA AH/%08x/" + "%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x\n", + ntohl(ah->spi), NIP6(iph->daddr)); xfrm_state_put(x); } diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index 5229365cd8b4..01468fab3d3d 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -29,6 +29,7 @@ #include <net/addrconf.h> #include <net/transp_v6.h> #include <net/ip6_route.h> +#include <net/tcp_states.h> #include <linux/errqueue.h> #include <asm/uaccess.h> @@ -588,8 +589,8 @@ int datagram_send_ctl(struct msghdr *msg, struct flowi *fl, break; default: - LIMIT_NETDEBUG( - printk(KERN_DEBUG "invalid cmsg type: %d\n", cmsg->cmsg_type)); + LIMIT_NETDEBUG(KERN_DEBUG "invalid cmsg type: %d\n", + cmsg->cmsg_type); err = -EINVAL; break; }; diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c index 324db62515a2..e8bff9d3d96c 100644 --- a/net/ipv6/esp6.c +++ b/net/ipv6/esp6.c @@ -212,8 +212,7 @@ static int esp6_input(struct xfrm_state *x, struct xfrm_decap_state *decap, stru padlen = nexthdr[0]; if (padlen+2 >= elen) { - LIMIT_NETDEBUG( - printk(KERN_WARNING "ipsec esp packet is garbage padlen=%d, elen=%d\n", padlen+2, elen)); + LIMIT_NETDEBUG(KERN_WARNING "ipsec esp packet is garbage padlen=%d, elen=%d\n", padlen+2, elen); ret = -EINVAL; goto out; } diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c index e0839eafc3a9..5be6da2584ee 100644 --- a/net/ipv6/exthdrs.c +++ b/net/ipv6/exthdrs.c @@ -424,8 +424,8 @@ static int ipv6_hop_ra(struct sk_buff *skb, int optoff) IP6CB(skb)->ra = optoff; return 1; } - LIMIT_NETDEBUG( - printk(KERN_DEBUG "ipv6_hop_ra: wrong RA length %d\n", skb->nh.raw[optoff+1])); + LIMIT_NETDEBUG(KERN_DEBUG "ipv6_hop_ra: wrong RA length %d\n", + skb->nh.raw[optoff+1]); kfree_skb(skb); return 0; } @@ -437,8 +437,8 @@ static int ipv6_hop_jumbo(struct sk_buff *skb, int optoff) u32 pkt_len; if (skb->nh.raw[optoff+1] != 4 || (optoff&3) != 2) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "ipv6_hop_jumbo: wrong jumbo opt length/alignment %d\n", skb->nh.raw[optoff+1])); + LIMIT_NETDEBUG(KERN_DEBUG "ipv6_hop_jumbo: wrong jumbo opt length/alignment %d\n", + skb->nh.raw[optoff+1]); IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS); goto drop; } diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c index ff3ec9822e36..5176fc655ea9 100644 --- a/net/ipv6/icmp.c +++ b/net/ipv6/icmp.c @@ -67,7 +67,7 @@ #include <asm/uaccess.h> #include <asm/system.h> -DEFINE_SNMP_STAT(struct icmpv6_mib, icmpv6_statistics); +DEFINE_SNMP_STAT(struct icmpv6_mib, icmpv6_statistics) __read_mostly; /* * The ICMP socket(s). This is the most convenient way to flow control @@ -332,8 +332,7 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info, * for now we don't know that. */ if ((addr_type == IPV6_ADDR_ANY) || (addr_type & IPV6_ADDR_MULTICAST)) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "icmpv6_send: addr_any/mcast source\n")); + LIMIT_NETDEBUG(KERN_DEBUG "icmpv6_send: addr_any/mcast source\n"); return; } @@ -341,8 +340,7 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info, * Never answer to a ICMP packet. */ if (is_ineligible(skb)) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "icmpv6_send: no reply to icmp error\n")); + LIMIT_NETDEBUG(KERN_DEBUG "icmpv6_send: no reply to icmp error\n"); return; } @@ -393,8 +391,7 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info, len = skb->len - msg.offset; len = min_t(unsigned int, len, IPV6_MIN_MTU - sizeof(struct ipv6hdr) -sizeof(struct icmp6hdr)); if (len < 0) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "icmp: len problem\n")); + LIMIT_NETDEBUG(KERN_DEBUG "icmp: len problem\n"); goto out_dst_release; } @@ -551,7 +548,8 @@ static void icmpv6_notify(struct sk_buff *skb, int type, int code, u32 info) read_lock(&raw_v6_lock); if ((sk = sk_head(&raw_v6_htable[hash])) != NULL) { - while((sk = __raw_v6_lookup(sk, nexthdr, daddr, saddr))) { + while((sk = __raw_v6_lookup(sk, nexthdr, daddr, saddr, + skb->dev->ifindex))) { rawv6_err(sk, skb, NULL, type, code, inner_offset, info); sk = sk_next(sk); } @@ -583,17 +581,15 @@ static int icmpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) skb->ip_summed = CHECKSUM_UNNECESSARY; if (csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6, skb->csum)) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "ICMPv6 hw checksum failed\n")); + LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 hw checksum failed\n"); skb->ip_summed = CHECKSUM_NONE; } } if (skb->ip_summed == CHECKSUM_NONE) { if (csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6, skb_checksum(skb, 0, skb->len, 0))) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "ICMPv6 checksum failed [%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x > %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x]\n", - NIP6(*saddr), NIP6(*daddr))); + LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 checksum failed [%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x > %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x]\n", + NIP6(*saddr), NIP6(*daddr)); goto discard_it; } } @@ -669,8 +665,7 @@ static int icmpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) break; default: - LIMIT_NETDEBUG( - printk(KERN_DEBUG "icmpv6: msg of unknown type\n")); + LIMIT_NETDEBUG(KERN_DEBUG "icmpv6: msg of unknown type\n"); /* informational */ if (type & ICMPV6_INFOMSG_MASK) diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c new file mode 100644 index 000000000000..01d5f46d4e40 --- /dev/null +++ b/net/ipv6/inet6_hashtables.c @@ -0,0 +1,81 @@ +/* + * INET An implementation of the TCP/IP protocol suite for the LINUX + * operating system. INET is implemented using the BSD Socket + * interface as the means of communication with the user level. + * + * Generic INET6 transport hashtables + * + * Authors: Lotsa people, from code originally in tcp + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> + +#include <linux/module.h> + +#include <net/inet_connection_sock.h> +#include <net/inet_hashtables.h> +#include <net/inet6_hashtables.h> + +struct sock *inet6_lookup_listener(struct inet_hashinfo *hashinfo, + const struct in6_addr *daddr, + const unsigned short hnum, const int dif) +{ + struct sock *sk; + const struct hlist_node *node; + struct sock *result = NULL; + int score, hiscore = 0; + + read_lock(&hashinfo->lhash_lock); + sk_for_each(sk, node, &hashinfo->listening_hash[inet_lhashfn(hnum)]) { + if (inet_sk(sk)->num == hnum && sk->sk_family == PF_INET6) { + const struct ipv6_pinfo *np = inet6_sk(sk); + + score = 1; + if (!ipv6_addr_any(&np->rcv_saddr)) { + if (!ipv6_addr_equal(&np->rcv_saddr, daddr)) + continue; + score++; + } + if (sk->sk_bound_dev_if) { + if (sk->sk_bound_dev_if != dif) + continue; + score++; + } + if (score == 3) { + result = sk; + break; + } + if (score > hiscore) { + hiscore = score; + result = sk; + } + } + } + if (result) + sock_hold(result); + read_unlock(&hashinfo->lhash_lock); + return result; +} + +EXPORT_SYMBOL_GPL(inet6_lookup_listener); + +struct sock *inet6_lookup(struct inet_hashinfo *hashinfo, + const struct in6_addr *saddr, const u16 sport, + const struct in6_addr *daddr, const u16 dport, + const int dif) +{ + struct sock *sk; + + local_bh_disable(); + sk = __inet6_lookup(hashinfo, saddr, sport, daddr, ntohs(dport), dif); + local_bh_enable(); + + return sk; +} + +EXPORT_SYMBOL_GPL(inet6_lookup); diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index 1b354aa97934..16af874c9e8f 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -49,7 +49,7 @@ struct rt6_statistics rt6_stats; -static kmem_cache_t * fib6_node_kmem; +static kmem_cache_t * fib6_node_kmem __read_mostly; enum fib_walk_state_t { diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c index 10fbb50daea4..6e3480426939 100644 --- a/net/ipv6/ip6_input.c +++ b/net/ipv6/ip6_input.c @@ -56,7 +56,7 @@ static inline int ip6_rcv_finish( struct sk_buff *skb) return dst_input(skb); } -int ipv6_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +int ipv6_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct ipv6hdr *hdr; u32 pkt_len; @@ -166,8 +166,8 @@ resubmit: nexthdr = skb->nh.raw[nhoff]; raw_sk = sk_head(&raw_v6_htable[nexthdr & (MAX_INET_PROTOS - 1)]); - if (raw_sk) - ipv6_raw_deliver(skb, nexthdr); + if (raw_sk && !ipv6_raw_deliver(skb, nexthdr)) + raw_sk = NULL; hash = nexthdr & (MAX_INET_PROTOS - 1); if ((ipprot = rcu_dereference(inet6_protos[hash])) != NULL) { diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index ae652ca14bc9..01ef94f7c7f1 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -153,51 +153,6 @@ int ip6_output(struct sk_buff *skb) return ip6_output2(skb); } -#ifdef CONFIG_NETFILTER -int ip6_route_me_harder(struct sk_buff *skb) -{ - struct ipv6hdr *iph = skb->nh.ipv6h; - struct dst_entry *dst; - struct flowi fl = { - .oif = skb->sk ? skb->sk->sk_bound_dev_if : 0, - .nl_u = - { .ip6_u = - { .daddr = iph->daddr, - .saddr = iph->saddr, } }, - .proto = iph->nexthdr, - }; - - dst = ip6_route_output(skb->sk, &fl); - - if (dst->error) { - IP6_INC_STATS(IPSTATS_MIB_OUTNOROUTES); - LIMIT_NETDEBUG( - printk(KERN_DEBUG "ip6_route_me_harder: No more route.\n")); - dst_release(dst); - return -EINVAL; - } - - /* Drop old route. */ - dst_release(skb->dst); - - skb->dst = dst; - return 0; -} -#endif - -static inline int ip6_maybe_reroute(struct sk_buff *skb) -{ -#ifdef CONFIG_NETFILTER - if (skb->nfcache & NFC_ALTERED){ - if (ip6_route_me_harder(skb) != 0){ - kfree_skb(skb); - return -EINVAL; - } - } -#endif /* CONFIG_NETFILTER */ - return dst_output(skb); -} - /* * xmit an sk_buff (used by TCP) */ @@ -266,7 +221,8 @@ int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl, mtu = dst_mtu(dst); if ((skb->len <= mtu) || ipfragok) { IP6_INC_STATS(IPSTATS_MIB_OUTREQUESTS); - return NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev, ip6_maybe_reroute); + return NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev, + dst_output); } if (net_ratelimit()) @@ -321,7 +277,9 @@ static int ip6_call_ra_chain(struct sk_buff *skb, int sel) read_lock(&ip6_ra_lock); for (ra = ip6_ra_chain; ra; ra = ra->next) { struct sock *sk = ra->sk; - if (sk && ra->sel == sel) { + if (sk && ra->sel == sel && + (!sk->sk_bound_dev_if || + sk->sk_bound_dev_if == skb->dev->ifindex)) { if (last) { struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC); if (skb2) @@ -667,7 +625,7 @@ slow_path: */ if ((frag = alloc_skb(len+hlen+sizeof(struct frag_hdr)+LL_RESERVED_SPACE(rt->u.dst.dev), GFP_ATOMIC)) == NULL) { - NETDEBUG(printk(KERN_INFO "IPv6: frag: no memory for new fragment!\n")); + NETDEBUG(KERN_INFO "IPv6: frag: no memory for new fragment!\n"); IP6_INC_STATS(IPSTATS_MIB_FRAGFAILS); err = -ENOMEM; goto fail; diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c index 3bc144a79fa5..76466af8331e 100644 --- a/net/ipv6/ipv6_sockglue.c +++ b/net/ipv6/ipv6_sockglue.c @@ -55,7 +55,7 @@ #include <asm/uaccess.h> -DEFINE_SNMP_STAT(struct ipstats_mib, ipv6_statistics); +DEFINE_SNMP_STAT(struct ipstats_mib, ipv6_statistics) __read_mostly; static struct packet_type ipv6_packet_type = { .type = __constant_htons(ETH_P_IPV6), @@ -109,13 +109,6 @@ int ip6_ra_control(struct sock *sk, int sel, void (*destructor)(struct sock *)) return 0; } -extern int ip6_mc_source(int add, int omode, struct sock *sk, - struct group_source_req *pgsr); -extern int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf); -extern int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf, - struct group_filter __user *optval, int __user *optlen); - - int ipv6_setsockopt(struct sock *sk, int level, int optname, char __user *optval, int optlen) { @@ -163,6 +156,13 @@ int ipv6_setsockopt(struct sock *sk, int level, int optname, fl6_free_socklist(sk); ipv6_sock_mc_close(sk); + /* + * Sock is moving from IPv6 to IPv4 (sk_prot), so + * remove it from the refcnt debug socks count in the + * original family... + */ + sk_refcnt_debug_dec(sk); + if (sk->sk_protocol == IPPROTO_TCP) { struct tcp_sock *tp = tcp_sk(sk); @@ -192,9 +192,11 @@ int ipv6_setsockopt(struct sock *sk, int level, int optname, kfree_skb(pktopt); sk->sk_destruct = inet_sock_destruct; -#ifdef INET_REFCNT_DEBUG - atomic_dec(&inet6_sock_nr); -#endif + /* + * ... and add it to the refcnt debug socks count + * in the new family. -acme + */ + sk_refcnt_debug_inc(sk); module_put(THIS_MODULE); retv = 0; break; @@ -437,7 +439,6 @@ done: } case MCAST_MSFILTER: { - extern int sysctl_optmem_max; extern int sysctl_mld_max_msf; struct group_filter *gsf; diff --git a/net/ipv6/ipv6_syms.c b/net/ipv6/ipv6_syms.c index 5ade5a5d1990..37a4a99c9fe9 100644 --- a/net/ipv6/ipv6_syms.c +++ b/net/ipv6/ipv6_syms.c @@ -15,9 +15,6 @@ EXPORT_SYMBOL(ndisc_mc_map); EXPORT_SYMBOL(register_inet6addr_notifier); EXPORT_SYMBOL(unregister_inet6addr_notifier); EXPORT_SYMBOL(ip6_route_output); -#ifdef CONFIG_NETFILTER -EXPORT_SYMBOL(ip6_route_me_harder); -#endif EXPORT_SYMBOL(addrconf_lock); EXPORT_SYMBOL(ipv6_setsockopt); EXPORT_SYMBOL(ipv6_getsockopt); diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index 7ae72d4c9bd2..a7eae30f4554 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -812,7 +812,7 @@ static void ndisc_recv_ns(struct sk_buff *skb) if (ipv6_chk_acast_addr(dev, &msg->target) || (idev->cnf.forwarding && pneigh_lookup(&nd_tbl, &msg->target, dev, 0))) { - if (skb->stamp.tv_sec != LOCALLY_ENQUEUED && + if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) && skb->pkt_type != PACKET_HOST && inc != 0 && idev->nd_parms->proxy_delay != 0) { @@ -1487,6 +1487,8 @@ int ndisc_rcv(struct sk_buff *skb) return 0; } + memset(NEIGH_CB(skb), 0, sizeof(struct neighbour_cb)); + switch (msg->icmph.icmp6_type) { case NDISC_NEIGHBOUR_SOLICITATION: ndisc_recv_ns(skb); diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c new file mode 100644 index 000000000000..f8626ebf90fd --- /dev/null +++ b/net/ipv6/netfilter.c @@ -0,0 +1,104 @@ +#include <linux/config.h> +#include <linux/init.h> + +#ifdef CONFIG_NETFILTER + +#include <linux/kernel.h> +#include <linux/ipv6.h> +#include <linux/netfilter.h> +#include <linux/netfilter_ipv6.h> +#include <net/dst.h> +#include <net/ipv6.h> +#include <net/ip6_route.h> + +int ip6_route_me_harder(struct sk_buff *skb) +{ + struct ipv6hdr *iph = skb->nh.ipv6h; + struct dst_entry *dst; + struct flowi fl = { + .oif = skb->sk ? skb->sk->sk_bound_dev_if : 0, + .nl_u = + { .ip6_u = + { .daddr = iph->daddr, + .saddr = iph->saddr, } }, + .proto = iph->nexthdr, + }; + + dst = ip6_route_output(skb->sk, &fl); + + if (dst->error) { + IP6_INC_STATS(IPSTATS_MIB_OUTNOROUTES); + LIMIT_NETDEBUG(KERN_DEBUG "ip6_route_me_harder: No more route.\n"); + dst_release(dst); + return -EINVAL; + } + + /* Drop old route. */ + dst_release(skb->dst); + + skb->dst = dst; + return 0; +} +EXPORT_SYMBOL(ip6_route_me_harder); + +/* + * Extra routing may needed on local out, as the QUEUE target never + * returns control to the table. + */ + +struct ip6_rt_info { + struct in6_addr daddr; + struct in6_addr saddr; +}; + +static void save(const struct sk_buff *skb, struct nf_info *info) +{ + struct ip6_rt_info *rt_info = nf_info_reroute(info); + + if (info->hook == NF_IP6_LOCAL_OUT) { + struct ipv6hdr *iph = skb->nh.ipv6h; + + rt_info->daddr = iph->daddr; + rt_info->saddr = iph->saddr; + } +} + +static int reroute(struct sk_buff **pskb, const struct nf_info *info) +{ + struct ip6_rt_info *rt_info = nf_info_reroute(info); + + if (info->hook == NF_IP6_LOCAL_OUT) { + struct ipv6hdr *iph = (*pskb)->nh.ipv6h; + if (!ipv6_addr_equal(&iph->daddr, &rt_info->daddr) || + !ipv6_addr_equal(&iph->saddr, &rt_info->saddr)) + return ip6_route_me_harder(*pskb); + } + return 0; +} + +static struct nf_queue_rerouter ip6_reroute = { + .rer_size = sizeof(struct ip6_rt_info), + .save = &save, + .reroute = &reroute, +}; + +int __init ipv6_netfilter_init(void) +{ + return nf_register_queue_rerouter(PF_INET6, &ip6_reroute); +} + +void ipv6_netfilter_fini(void) +{ + nf_unregister_queue_rerouter(PF_INET6); +} + +#else /* CONFIG_NETFILTER */ +int __init ipv6_netfilter_init(void) +{ + return 0; +} + +void ipv6_netfilter_fini(void) +{ +} +#endif /* CONFIG_NETFILTER */ diff --git a/net/ipv6/netfilter/Kconfig b/net/ipv6/netfilter/Kconfig index 77ec704c9ee3..216fbe1ac65c 100644 --- a/net/ipv6/netfilter/Kconfig +++ b/net/ipv6/netfilter/Kconfig @@ -10,13 +10,16 @@ menu "IPv6: Netfilter Configuration (EXPERIMENTAL)" # dep_tristate ' FTP protocol support' CONFIG_IP6_NF_FTP $CONFIG_IP6_NF_CONNTRACK #fi config IP6_NF_QUEUE - tristate "Userspace queueing via NETLINK" + tristate "IP6 Userspace queueing via NETLINK (OBSOLETE)" ---help--- This option adds a queue handler to the kernel for IPv6 - packets which lets us to receive the filtered packets - with QUEUE target using libiptc as we can do with - the IPv4 now. + packets which enables users to receive the filtered packets + with QUEUE target using libipq. + + THis option enables the old IPv6-only "ip6_queue" implementation + which has been obsoleted by the new "nfnetlink_queue" code (see + CONFIG_NETFILTER_NETLINK_QUEUE). (C) Fernando Anton 2001 IPv64 Project - Work based in IPv64 draft by Arturo Azcorra. @@ -196,6 +199,16 @@ config IP6_NF_TARGET_LOG To compile it as a module, choose M here. If unsure, say N. +config IP6_NF_TARGET_REJECT + tristate "REJECT target support" + depends on IP6_NF_FILTER + help + The REJECT target allows a filtering rule to specify that an ICMPv6 + error should be issued in response to an incoming packet, rather + than silently being dropped. + + To compile it as a module, choose M here. If unsure, say N. + # if [ "$CONFIG_IP6_NF_FILTER" != "n" ]; then # dep_tristate ' REJECT target support' CONFIG_IP6_NF_TARGET_REJECT $CONFIG_IP6_NF_FILTER # if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then @@ -226,6 +239,22 @@ config IP6_NF_TARGET_MARK To compile it as a module, choose M here. If unsure, say N. +config IP6_NF_TARGET_HL + tristate 'HL (hoplimit) target support' + depends on IP6_NF_MANGLE + help + This option adds a `HL' target, which enables the user to decrement + the hoplimit value of the IPv6 header or set it to a given (lower) + value. + + While it is safe to decrement the hoplimit value, this option also + enables functionality to increment and set the hoplimit value of the + IPv6 header to arbitrary values. This is EXTREMELY DANGEROUS since + you can easily create immortal packets that loop forever on the + network. + + To compile it as a module, choose M here. If unsure, say N. + #dep_tristate ' LOG target support' CONFIG_IP6_NF_TARGET_LOG $CONFIG_IP6_NF_IPTABLES config IP6_NF_RAW tristate 'raw table support (required for TRACE)' diff --git a/net/ipv6/netfilter/Makefile b/net/ipv6/netfilter/Makefile index 2e51714953b6..bd9a16a5cbba 100644 --- a/net/ipv6/netfilter/Makefile +++ b/net/ipv6/netfilter/Makefile @@ -20,7 +20,10 @@ obj-$(CONFIG_IP6_NF_MATCH_PHYSDEV) += ip6t_physdev.o obj-$(CONFIG_IP6_NF_FILTER) += ip6table_filter.o obj-$(CONFIG_IP6_NF_MANGLE) += ip6table_mangle.o obj-$(CONFIG_IP6_NF_TARGET_MARK) += ip6t_MARK.o +obj-$(CONFIG_IP6_NF_TARGET_HL) += ip6t_HL.o obj-$(CONFIG_IP6_NF_QUEUE) += ip6_queue.o obj-$(CONFIG_IP6_NF_TARGET_LOG) += ip6t_LOG.o obj-$(CONFIG_IP6_NF_RAW) += ip6table_raw.o obj-$(CONFIG_IP6_NF_MATCH_HL) += ip6t_hl.o +obj-$(CONFIG_IP6_NF_TARGET_REJECT) += ip6t_REJECT.o +obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += ip6t_NFQUEUE.o diff --git a/net/ipv6/netfilter/ip6_queue.c b/net/ipv6/netfilter/ip6_queue.c index a16df5b27c84..aa11cf366efa 100644 --- a/net/ipv6/netfilter/ip6_queue.c +++ b/net/ipv6/netfilter/ip6_queue.c @@ -47,16 +47,10 @@ #define NET_IPQ_QMAX 2088 #define NET_IPQ_QMAX_NAME "ip6_queue_maxlen" -struct ipq_rt_info { - struct in6_addr daddr; - struct in6_addr saddr; -}; - struct ipq_queue_entry { struct list_head list; struct nf_info *info; struct sk_buff *skb; - struct ipq_rt_info rt_info; }; typedef int (*ipq_cmpfn)(struct ipq_queue_entry *, unsigned long); @@ -244,8 +238,8 @@ ipq_build_packet_message(struct ipq_queue_entry *entry, int *errp) pmsg->packet_id = (unsigned long )entry; pmsg->data_len = data_len; - pmsg->timestamp_sec = entry->skb->stamp.tv_sec; - pmsg->timestamp_usec = entry->skb->stamp.tv_usec; + pmsg->timestamp_sec = skb_tv_base.tv_sec + entry->skb->tstamp.off_sec; + pmsg->timestamp_usec = skb_tv_base.tv_usec + entry->skb->tstamp.off_usec; pmsg->mark = entry->skb->nfmark; pmsg->hook = entry->info->hook; pmsg->hw_protocol = entry->skb->protocol; @@ -284,7 +278,8 @@ nlmsg_failure: } static int -ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, void *data) +ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, + unsigned int queuenum, void *data) { int status = -EINVAL; struct sk_buff *nskb; @@ -302,13 +297,6 @@ ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, void *data) entry->info = info; entry->skb = skb; - if (entry->info->hook == NF_IP_LOCAL_OUT) { - struct ipv6hdr *iph = skb->nh.ipv6h; - - entry->rt_info.daddr = iph->daddr; - entry->rt_info.saddr = iph->saddr; - } - nskb = ipq_build_packet_message(entry, &status); if (nskb == NULL) goto err_out_free; @@ -384,23 +372,11 @@ ipq_mangle_ipv6(ipq_verdict_msg_t *v, struct ipq_queue_entry *e) } skb_put(e->skb, diff); } - if (!skb_ip_make_writable(&e->skb, v->data_len)) + if (!skb_make_writable(&e->skb, v->data_len)) return -ENOMEM; memcpy(e->skb->data, v->payload, v->data_len); e->skb->ip_summed = CHECKSUM_NONE; - e->skb->nfcache |= NFC_ALTERED; - - /* - * Extra routing may needed on local out, as the QUEUE target never - * returns control to the table. - * Not a nice way to cmp, but works - */ - if (e->info->hook == NF_IP_LOCAL_OUT) { - struct ipv6hdr *iph = e->skb->nh.ipv6h; - if (!ipv6_addr_equal(&iph->daddr, &e->rt_info.daddr) || - !ipv6_addr_equal(&iph->saddr, &e->rt_info.saddr)) - return ip6_route_me_harder(e->skb); - } + return 0; } @@ -676,6 +652,11 @@ ipq_get_info(char *buffer, char **start, off_t offset, int length) return len; } +static struct nf_queue_handler nfqh = { + .name = "ip6_queue", + .outfn = &ipq_enqueue_packet, +}; + static int init_or_cleanup(int init) { @@ -686,7 +667,8 @@ init_or_cleanup(int init) goto cleanup; netlink_register_notifier(&ipq_nl_notifier); - ipqnl = netlink_kernel_create(NETLINK_IP6_FW, ipq_rcv_sk); + ipqnl = netlink_kernel_create(NETLINK_IP6_FW, 0, ipq_rcv_sk, + THIS_MODULE); if (ipqnl == NULL) { printk(KERN_ERR "ip6_queue: failed to create netlink socket\n"); goto cleanup_netlink_notifier; @@ -703,7 +685,7 @@ init_or_cleanup(int init) register_netdevice_notifier(&ipq_dev_notifier); ipq_sysctl_header = register_sysctl_table(ipq_root_table, 0); - status = nf_register_queue_handler(PF_INET6, ipq_enqueue_packet, NULL); + status = nf_register_queue_handler(PF_INET6, &nfqh); if (status < 0) { printk(KERN_ERR "ip6_queue: failed to register queue handler\n"); goto cleanup_sysctl; @@ -711,7 +693,7 @@ init_or_cleanup(int init) return status; cleanup: - nf_unregister_queue_handler(PF_INET6); + nf_unregister_queue_handlers(&nfqh); synchronize_net(); ipq_flush(NF_DROP); diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c index 73034511c8db..1cb8adb2787f 100644 --- a/net/ipv6/netfilter/ip6_tables.c +++ b/net/ipv6/netfilter/ip6_tables.c @@ -401,7 +401,6 @@ ip6t_do_table(struct sk_buff **pskb, do { IP_NF_ASSERT(e); IP_NF_ASSERT(back); - (*pskb)->nfcache |= e->nfcache; if (ip6_packet_match(*pskb, indev, outdev, &e->ipv6, &protoff, &offset)) { struct ip6t_entry_target *t; @@ -434,8 +433,8 @@ ip6t_do_table(struct sk_buff **pskb, back->comefrom); continue; } - if (table_base + v - != (void *)e + e->next_offset) { + if (table_base + v != (void *)e + e->next_offset + && !(e->ipv6.flags & IP6T_F_GOTO)) { /* Save old back ptr in next entry */ struct ip6t_entry *next = (void *)e + e->next_offset; diff --git a/net/ipv6/netfilter/ip6t_HL.c b/net/ipv6/netfilter/ip6t_HL.c new file mode 100644 index 000000000000..8f5549b72720 --- /dev/null +++ b/net/ipv6/netfilter/ip6t_HL.c @@ -0,0 +1,118 @@ +/* + * Hop Limit modification target for ip6tables + * Maciej Soltysiak <solt@dns.toxicfilms.tv> + * Based on HW's TTL module + * + * This software is distributed under the terms of GNU GPL + */ + +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/ip.h> + +#include <linux/netfilter_ipv6/ip6_tables.h> +#include <linux/netfilter_ipv6/ip6t_HL.h> + +MODULE_AUTHOR("Maciej Soltysiak <solt@dns.toxicfilms.tv>"); +MODULE_DESCRIPTION("IP tables Hop Limit modification module"); +MODULE_LICENSE("GPL"); + +static unsigned int ip6t_hl_target(struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + unsigned int hooknum, + const void *targinfo, void *userinfo) +{ + struct ipv6hdr *ip6h; + const struct ip6t_HL_info *info = targinfo; + u_int16_t diffs[2]; + int new_hl; + + if (!skb_make_writable(pskb, (*pskb)->len)) + return NF_DROP; + + ip6h = (*pskb)->nh.ipv6h; + + switch (info->mode) { + case IP6T_HL_SET: + new_hl = info->hop_limit; + break; + case IP6T_HL_INC: + new_hl = ip6h->hop_limit + info->hop_limit; + if (new_hl > 255) + new_hl = 255; + break; + case IP6T_HL_DEC: + new_hl = ip6h->hop_limit - info->hop_limit; + if (new_hl < 0) + new_hl = 0; + break; + default: + new_hl = ip6h->hop_limit; + break; + } + + if (new_hl != ip6h->hop_limit) { + diffs[0] = htons(((unsigned)ip6h->hop_limit) << 8) ^ 0xFFFF; + ip6h->hop_limit = new_hl; + diffs[1] = htons(((unsigned)ip6h->hop_limit) << 8); + } + + return IP6T_CONTINUE; +} + +static int ip6t_hl_checkentry(const char *tablename, + const struct ip6t_entry *e, + void *targinfo, + unsigned int targinfosize, + unsigned int hook_mask) +{ + struct ip6t_HL_info *info = targinfo; + + if (targinfosize != IP6T_ALIGN(sizeof(struct ip6t_HL_info))) { + printk(KERN_WARNING "ip6t_HL: targinfosize %u != %Zu\n", + targinfosize, + IP6T_ALIGN(sizeof(struct ip6t_HL_info))); + return 0; + } + + if (strcmp(tablename, "mangle")) { + printk(KERN_WARNING "ip6t_HL: can only be called from " + "\"mangle\" table, not \"%s\"\n", tablename); + return 0; + } + + if (info->mode > IP6T_HL_MAXMODE) { + printk(KERN_WARNING "ip6t_HL: invalid or unknown Mode %u\n", + info->mode); + return 0; + } + + if ((info->mode != IP6T_HL_SET) && (info->hop_limit == 0)) { + printk(KERN_WARNING "ip6t_HL: increment/decrement doesn't " + "make sense with value 0\n"); + return 0; + } + + return 1; +} + +static struct ip6t_target ip6t_HL = { + .name = "HL", + .target = ip6t_hl_target, + .checkentry = ip6t_hl_checkentry, + .me = THIS_MODULE +}; + +static int __init init(void) +{ + return ip6t_register_target(&ip6t_HL); +} + +static void __exit fini(void) +{ + ip6t_unregister_target(&ip6t_HL); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv6/netfilter/ip6t_LOG.c b/net/ipv6/netfilter/ip6t_LOG.c index a692e26a4fa3..0cd1d1bd9033 100644 --- a/net/ipv6/netfilter/ip6t_LOG.c +++ b/net/ipv6/netfilter/ip6t_LOG.c @@ -26,10 +26,6 @@ MODULE_AUTHOR("Jan Rekorajski <baggins@pld.org.pl>"); MODULE_DESCRIPTION("IP6 tables LOG target module"); MODULE_LICENSE("GPL"); -static unsigned int nflog = 1; -module_param(nflog, int, 0400); -MODULE_PARM_DESC(nflog, "register as internal netfilter logging module"); - struct in_device; #include <net/route.h> #include <linux/netfilter_ipv6/ip6t_LOG.h> @@ -44,7 +40,7 @@ struct in_device; static DEFINE_SPINLOCK(log_lock); /* One level of recursion won't kill us */ -static void dump_packet(const struct ip6t_log_info *info, +static void dump_packet(const struct nf_loginfo *info, const struct sk_buff *skb, unsigned int ip6hoff, int recurse) { @@ -53,6 +49,12 @@ static void dump_packet(const struct ip6t_log_info *info, struct ipv6hdr _ip6h, *ih; unsigned int ptr; unsigned int hdrlen = 0; + unsigned int logflags; + + if (info->type == NF_LOG_TYPE_LOG) + logflags = info->u.log.logflags; + else + logflags = NF_LOG_MASK; ih = skb_header_pointer(skb, ip6hoff, sizeof(_ip6h), &_ip6h); if (ih == NULL) { @@ -84,7 +86,7 @@ static void dump_packet(const struct ip6t_log_info *info, } /* Max length: 48 "OPT (...) " */ - if (info->logflags & IP6T_LOG_IPOPT) + if (logflags & IP6T_LOG_IPOPT) printk("OPT ( "); switch (currenthdr) { @@ -119,7 +121,7 @@ static void dump_packet(const struct ip6t_log_info *info, case IPPROTO_ROUTING: case IPPROTO_HOPOPTS: if (fragment) { - if (info->logflags & IP6T_LOG_IPOPT) + if (logflags & IP6T_LOG_IPOPT) printk(")"); return; } @@ -127,7 +129,7 @@ static void dump_packet(const struct ip6t_log_info *info, break; /* Max Length */ case IPPROTO_AH: - if (info->logflags & IP6T_LOG_IPOPT) { + if (logflags & IP6T_LOG_IPOPT) { struct ip_auth_hdr _ahdr, *ah; /* Max length: 3 "AH " */ @@ -158,7 +160,7 @@ static void dump_packet(const struct ip6t_log_info *info, hdrlen = (hp->hdrlen+2)<<2; break; case IPPROTO_ESP: - if (info->logflags & IP6T_LOG_IPOPT) { + if (logflags & IP6T_LOG_IPOPT) { struct ip_esp_hdr _esph, *eh; /* Max length: 4 "ESP " */ @@ -190,7 +192,7 @@ static void dump_packet(const struct ip6t_log_info *info, printk("Unknown Ext Hdr %u", currenthdr); return; } - if (info->logflags & IP6T_LOG_IPOPT) + if (logflags & IP6T_LOG_IPOPT) printk(") "); currenthdr = hp->nexthdr; @@ -218,7 +220,7 @@ static void dump_packet(const struct ip6t_log_info *info, printk("SPT=%u DPT=%u ", ntohs(th->source), ntohs(th->dest)); /* Max length: 30 "SEQ=4294967295 ACK=4294967295 " */ - if (info->logflags & IP6T_LOG_TCPSEQ) + if (logflags & IP6T_LOG_TCPSEQ) printk("SEQ=%u ACK=%u ", ntohl(th->seq), ntohl(th->ack_seq)); /* Max length: 13 "WINDOW=65535 " */ @@ -245,7 +247,7 @@ static void dump_packet(const struct ip6t_log_info *info, /* Max length: 11 "URGP=65535 " */ printk("URGP=%u ", ntohs(th->urg_ptr)); - if ((info->logflags & IP6T_LOG_TCPOPT) + if ((logflags & IP6T_LOG_TCPOPT) && th->doff * 4 > sizeof(struct tcphdr)) { u_int8_t _opt[60 - sizeof(struct tcphdr)], *op; unsigned int i; @@ -349,7 +351,7 @@ static void dump_packet(const struct ip6t_log_info *info, } /* Max length: 15 "UID=4294967295 " */ - if ((info->logflags & IP6T_LOG_UID) && recurse && skb->sk) { + if ((logflags & IP6T_LOG_UID) && recurse && skb->sk) { read_lock_bh(&skb->sk->sk_callback_lock); if (skb->sk->sk_socket && skb->sk->sk_socket->file) printk("UID=%u ", skb->sk->sk_socket->file->f_uid); @@ -357,19 +359,31 @@ static void dump_packet(const struct ip6t_log_info *info, } } +static struct nf_loginfo default_loginfo = { + .type = NF_LOG_TYPE_LOG, + .u = { + .log = { + .level = 0, + .logflags = NF_LOG_MASK, + }, + }, +}; + static void -ip6t_log_packet(unsigned int hooknum, +ip6t_log_packet(unsigned int pf, + unsigned int hooknum, const struct sk_buff *skb, const struct net_device *in, const struct net_device *out, - const struct ip6t_log_info *loginfo, - const char *level_string, + const struct nf_loginfo *loginfo, const char *prefix) { + if (!loginfo) + loginfo = &default_loginfo; + spin_lock_bh(&log_lock); - printk(level_string); - printk("%sIN=%s OUT=%s ", - prefix == NULL ? loginfo->prefix : prefix, + printk("<%d>%sIN=%s OUT=%s ", loginfo->u.log.level, + prefix, in ? in->name : "", out ? out->name : ""); if (in && !out) { @@ -416,29 +430,17 @@ ip6t_log_target(struct sk_buff **pskb, void *userinfo) { const struct ip6t_log_info *loginfo = targinfo; - char level_string[4] = "< >"; + struct nf_loginfo li; + + li.type = NF_LOG_TYPE_LOG; + li.u.log.level = loginfo->level; + li.u.log.logflags = loginfo->logflags; - level_string[1] = '0' + (loginfo->level % 8); - ip6t_log_packet(hooknum, *pskb, in, out, loginfo, level_string, NULL); + nf_log_packet(PF_INET6, hooknum, *pskb, in, out, &li, loginfo->prefix); return IP6T_CONTINUE; } -static void -ip6t_logfn(unsigned int hooknum, - const struct sk_buff *skb, - const struct net_device *in, - const struct net_device *out, - const char *prefix) -{ - struct ip6t_log_info loginfo = { - .level = 0, - .logflags = IP6T_LOG_MASK, - .prefix = "" - }; - - ip6t_log_packet(hooknum, skb, in, out, &loginfo, KERN_WARNING, prefix); -} static int ip6t_log_checkentry(const char *tablename, const struct ip6t_entry *e, @@ -475,20 +477,29 @@ static struct ip6t_target ip6t_log_reg = { .me = THIS_MODULE, }; +static struct nf_logger ip6t_logger = { + .name = "ip6t_LOG", + .logfn = &ip6t_log_packet, + .me = THIS_MODULE, +}; + static int __init init(void) { if (ip6t_register_target(&ip6t_log_reg)) return -EINVAL; - if (nflog) - nf_log_register(PF_INET6, &ip6t_logfn); + if (nf_log_register(PF_INET6, &ip6t_logger) < 0) { + printk(KERN_WARNING "ip6t_LOG: not logging via system console " + "since somebody else already registered for PF_INET6\n"); + /* we cannot make module load fail here, since otherwise + * ip6tables userspace would abort */ + } return 0; } static void __exit fini(void) { - if (nflog) - nf_log_unregister(PF_INET6, &ip6t_logfn); + nf_log_unregister_logger(&ip6t_logger); ip6t_unregister_target(&ip6t_log_reg); } diff --git a/net/ipv6/netfilter/ip6t_MARK.c b/net/ipv6/netfilter/ip6t_MARK.c index d09ceb05013a..81924fcc5857 100644 --- a/net/ipv6/netfilter/ip6t_MARK.c +++ b/net/ipv6/netfilter/ip6t_MARK.c @@ -28,10 +28,9 @@ target(struct sk_buff **pskb, { const struct ip6t_mark_target_info *markinfo = targinfo; - if((*pskb)->nfmark != markinfo->mark) { + if((*pskb)->nfmark != markinfo->mark) (*pskb)->nfmark = markinfo->mark; - (*pskb)->nfcache |= NFC_ALTERED; - } + return IP6T_CONTINUE; } diff --git a/net/ipv6/netfilter/ip6t_NFQUEUE.c b/net/ipv6/netfilter/ip6t_NFQUEUE.c new file mode 100644 index 000000000000..c6e3730e7409 --- /dev/null +++ b/net/ipv6/netfilter/ip6t_NFQUEUE.c @@ -0,0 +1,70 @@ +/* ip6tables module for using new netfilter netlink queue + * + * (C) 2005 by Harald Welte <laforge@netfilter.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include <linux/module.h> +#include <linux/skbuff.h> + +#include <linux/netfilter.h> +#include <linux/netfilter_ipv6/ip6_tables.h> +#include <linux/netfilter_ipv4/ipt_NFQUEUE.h> + +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_DESCRIPTION("ip6tables NFQUEUE target"); +MODULE_LICENSE("GPL"); + +static unsigned int +target(struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + unsigned int hooknum, + const void *targinfo, + void *userinfo) +{ + const struct ipt_NFQ_info *tinfo = targinfo; + + return NF_QUEUE_NR(tinfo->queuenum); +} + +static int +checkentry(const char *tablename, + const struct ip6t_entry *e, + void *targinfo, + unsigned int targinfosize, + unsigned int hook_mask) +{ + if (targinfosize != IP6T_ALIGN(sizeof(struct ipt_NFQ_info))) { + printk(KERN_WARNING "NFQUEUE: targinfosize %u != %Zu\n", + targinfosize, + IP6T_ALIGN(sizeof(struct ipt_NFQ_info))); + return 0; + } + + return 1; +} + +static struct ip6t_target ipt_NFQ_reg = { + .name = "NFQUEUE", + .target = target, + .checkentry = checkentry, + .me = THIS_MODULE, +}; + +static int __init init(void) +{ + return ip6t_register_target(&ipt_NFQ_reg); +} + +static void __exit fini(void) +{ + ip6t_unregister_target(&ipt_NFQ_reg); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv6/netfilter/ip6t_REJECT.c b/net/ipv6/netfilter/ip6t_REJECT.c new file mode 100644 index 000000000000..14316c3ebde4 --- /dev/null +++ b/net/ipv6/netfilter/ip6t_REJECT.c @@ -0,0 +1,284 @@ +/* + * IP6 tables REJECT target module + * Linux INET6 implementation + * + * Copyright (C)2003 USAGI/WIDE Project + * + * Authors: + * Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp> + * + * Based on net/ipv4/netfilter/ipt_REJECT.c + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/icmpv6.h> +#include <linux/netdevice.h> +#include <net/ipv6.h> +#include <net/tcp.h> +#include <net/icmp.h> +#include <net/ip6_checksum.h> +#include <net/ip6_fib.h> +#include <net/ip6_route.h> +#include <net/flow.h> +#include <linux/netfilter_ipv6/ip6_tables.h> +#include <linux/netfilter_ipv6/ip6t_REJECT.h> + +MODULE_AUTHOR("Yasuyuki KOZAKAI <yasuyuki.kozakai@toshiba.co.jp>"); +MODULE_DESCRIPTION("IP6 tables REJECT target module"); +MODULE_LICENSE("GPL"); + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +/* Send RST reply */ +static void send_reset(struct sk_buff *oldskb) +{ + struct sk_buff *nskb; + struct tcphdr otcph, *tcph; + unsigned int otcplen, hh_len; + int tcphoff, needs_ack; + struct ipv6hdr *oip6h = oldskb->nh.ipv6h, *ip6h; + struct dst_entry *dst = NULL; + u8 proto; + struct flowi fl; + + if ((!(ipv6_addr_type(&oip6h->saddr) & IPV6_ADDR_UNICAST)) || + (!(ipv6_addr_type(&oip6h->daddr) & IPV6_ADDR_UNICAST))) { + DEBUGP("ip6t_REJECT: addr is not unicast.\n"); + return; + } + + proto = oip6h->nexthdr; + tcphoff = ipv6_skip_exthdr(oldskb, ((u8*)(oip6h+1) - oldskb->data), &proto); + + if ((tcphoff < 0) || (tcphoff > oldskb->len)) { + DEBUGP("ip6t_REJECT: Can't get TCP header.\n"); + return; + } + + otcplen = oldskb->len - tcphoff; + + /* IP header checks: fragment, too short. */ + if ((proto != IPPROTO_TCP) || (otcplen < sizeof(struct tcphdr))) { + DEBUGP("ip6t_REJECT: proto(%d) != IPPROTO_TCP, or too short. otcplen = %d\n", + proto, otcplen); + return; + } + + if (skb_copy_bits(oldskb, tcphoff, &otcph, sizeof(struct tcphdr))) + BUG(); + + /* No RST for RST. */ + if (otcph.rst) { + DEBUGP("ip6t_REJECT: RST is set\n"); + return; + } + + /* Check checksum. */ + if (csum_ipv6_magic(&oip6h->saddr, &oip6h->daddr, otcplen, IPPROTO_TCP, + skb_checksum(oldskb, tcphoff, otcplen, 0))) { + DEBUGP("ip6t_REJECT: TCP checksum is invalid\n"); + return; + } + + memset(&fl, 0, sizeof(fl)); + fl.proto = IPPROTO_TCP; + ipv6_addr_copy(&fl.fl6_src, &oip6h->daddr); + ipv6_addr_copy(&fl.fl6_dst, &oip6h->saddr); + fl.fl_ip_sport = otcph.dest; + fl.fl_ip_dport = otcph.source; + dst = ip6_route_output(NULL, &fl); + if (dst == NULL) + return; + if (dst->error || + xfrm_lookup(&dst, &fl, NULL, 0)) { + dst_release(dst); + return; + } + + hh_len = (dst->dev->hard_header_len + 15)&~15; + nskb = alloc_skb(hh_len + 15 + dst->header_len + sizeof(struct ipv6hdr) + + sizeof(struct tcphdr) + dst->trailer_len, + GFP_ATOMIC); + + if (!nskb) { + if (net_ratelimit()) + printk("ip6t_REJECT: Can't alloc skb\n"); + dst_release(dst); + return; + } + + nskb->dst = dst; + + skb_reserve(nskb, hh_len + dst->header_len); + + ip6h = nskb->nh.ipv6h = (struct ipv6hdr *) + skb_put(nskb, sizeof(struct ipv6hdr)); + ip6h->version = 6; + ip6h->hop_limit = dst_metric(dst, RTAX_HOPLIMIT); + ip6h->nexthdr = IPPROTO_TCP; + ip6h->payload_len = htons(sizeof(struct tcphdr)); + ipv6_addr_copy(&ip6h->saddr, &oip6h->daddr); + ipv6_addr_copy(&ip6h->daddr, &oip6h->saddr); + + tcph = (struct tcphdr *)skb_put(nskb, sizeof(struct tcphdr)); + /* Truncate to length (no data) */ + tcph->doff = sizeof(struct tcphdr)/4; + tcph->source = otcph.dest; + tcph->dest = otcph.source; + + if (otcph.ack) { + needs_ack = 0; + tcph->seq = otcph.ack_seq; + tcph->ack_seq = 0; + } else { + needs_ack = 1; + tcph->ack_seq = htonl(ntohl(otcph.seq) + otcph.syn + otcph.fin + + otcplen - (otcph.doff<<2)); + tcph->seq = 0; + } + + /* Reset flags */ + ((u_int8_t *)tcph)[13] = 0; + tcph->rst = 1; + tcph->ack = needs_ack; + tcph->window = 0; + tcph->urg_ptr = 0; + tcph->check = 0; + + /* Adjust TCP checksum */ + tcph->check = csum_ipv6_magic(&nskb->nh.ipv6h->saddr, + &nskb->nh.ipv6h->daddr, + sizeof(struct tcphdr), IPPROTO_TCP, + csum_partial((char *)tcph, + sizeof(struct tcphdr), 0)); + + NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, nskb, NULL, nskb->dst->dev, + dst_output); +} + +static inline void +send_unreach(struct sk_buff *skb_in, unsigned char code, unsigned int hooknum) +{ + if (hooknum == NF_IP6_LOCAL_OUT && skb_in->dev == NULL) + skb_in->dev = &loopback_dev; + + icmpv6_send(skb_in, ICMPV6_DEST_UNREACH, code, 0, NULL); +} + +static unsigned int reject6_target(struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + unsigned int hooknum, + const void *targinfo, + void *userinfo) +{ + const struct ip6t_reject_info *reject = targinfo; + + DEBUGP(KERN_DEBUG "%s: medium point\n", __FUNCTION__); + /* WARNING: This code causes reentry within ip6tables. + This means that the ip6tables jump stack is now crap. We + must return an absolute verdict. --RR */ + switch (reject->with) { + case IP6T_ICMP6_NO_ROUTE: + send_unreach(*pskb, ICMPV6_NOROUTE, hooknum); + break; + case IP6T_ICMP6_ADM_PROHIBITED: + send_unreach(*pskb, ICMPV6_ADM_PROHIBITED, hooknum); + break; + case IP6T_ICMP6_NOT_NEIGHBOUR: + send_unreach(*pskb, ICMPV6_NOT_NEIGHBOUR, hooknum); + break; + case IP6T_ICMP6_ADDR_UNREACH: + send_unreach(*pskb, ICMPV6_ADDR_UNREACH, hooknum); + break; + case IP6T_ICMP6_PORT_UNREACH: + send_unreach(*pskb, ICMPV6_PORT_UNREACH, hooknum); + break; + case IP6T_ICMP6_ECHOREPLY: + /* Do nothing */ + break; + case IP6T_TCP_RESET: + send_reset(*pskb); + break; + default: + if (net_ratelimit()) + printk(KERN_WARNING "ip6t_REJECT: case %u not handled yet\n", reject->with); + break; + } + + return NF_DROP; +} + +static int check(const char *tablename, + const struct ip6t_entry *e, + void *targinfo, + unsigned int targinfosize, + unsigned int hook_mask) +{ + const struct ip6t_reject_info *rejinfo = targinfo; + + if (targinfosize != IP6T_ALIGN(sizeof(struct ip6t_reject_info))) { + DEBUGP("ip6t_REJECT: targinfosize %u != 0\n", targinfosize); + return 0; + } + + /* Only allow these for packet filtering. */ + if (strcmp(tablename, "filter") != 0) { + DEBUGP("ip6t_REJECT: bad table `%s'.\n", tablename); + return 0; + } + + if ((hook_mask & ~((1 << NF_IP6_LOCAL_IN) + | (1 << NF_IP6_FORWARD) + | (1 << NF_IP6_LOCAL_OUT))) != 0) { + DEBUGP("ip6t_REJECT: bad hook mask %X\n", hook_mask); + return 0; + } + + if (rejinfo->with == IP6T_ICMP6_ECHOREPLY) { + printk("ip6t_REJECT: ECHOREPLY is not supported.\n"); + return 0; + } else if (rejinfo->with == IP6T_TCP_RESET) { + /* Must specify that it's a TCP packet */ + if (e->ipv6.proto != IPPROTO_TCP + || (e->ipv6.invflags & IP6T_INV_PROTO)) { + DEBUGP("ip6t_REJECT: TCP_RESET illegal for non-tcp\n"); + return 0; + } + } + + return 1; +} + +static struct ip6t_target ip6t_reject_reg = { + .name = "REJECT", + .target = reject6_target, + .checkentry = check, + .me = THIS_MODULE +}; + +static int __init init(void) +{ + if (ip6t_register_target(&ip6t_reject_reg)) + return -EINVAL; + return 0; +} + +static void __exit fini(void) +{ + ip6t_unregister_target(&ip6t_reject_reg); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv6/netfilter/ip6t_owner.c b/net/ipv6/netfilter/ip6t_owner.c index ab0e32d3de46..9b91decbfddb 100644 --- a/net/ipv6/netfilter/ip6t_owner.c +++ b/net/ipv6/netfilter/ip6t_owner.c @@ -20,71 +20,6 @@ MODULE_AUTHOR("Marc Boucher <marc@mbsi.ca>"); MODULE_DESCRIPTION("IP6 tables owner matching module"); MODULE_LICENSE("GPL"); -static int -match_pid(const struct sk_buff *skb, pid_t pid) -{ - struct task_struct *p; - struct files_struct *files; - int i; - - read_lock(&tasklist_lock); - p = find_task_by_pid(pid); - if (!p) - goto out; - task_lock(p); - files = p->files; - if(files) { - spin_lock(&files->file_lock); - for (i=0; i < files->max_fds; i++) { - if (fcheck_files(files, i) == skb->sk->sk_socket->file) { - spin_unlock(&files->file_lock); - task_unlock(p); - read_unlock(&tasklist_lock); - return 1; - } - } - spin_unlock(&files->file_lock); - } - task_unlock(p); -out: - read_unlock(&tasklist_lock); - return 0; -} - -static int -match_sid(const struct sk_buff *skb, pid_t sid) -{ - struct task_struct *g, *p; - struct file *file = skb->sk->sk_socket->file; - int i, found=0; - - read_lock(&tasklist_lock); - do_each_thread(g, p) { - struct files_struct *files; - if (p->signal->session != sid) - continue; - - task_lock(p); - files = p->files; - if (files) { - spin_lock(&files->file_lock); - for (i=0; i < files->max_fds; i++) { - if (fcheck_files(files, i) == file) { - found = 1; - break; - } - } - spin_unlock(&files->file_lock); - } - task_unlock(p); - if (found) - goto out; - } while_each_thread(g, p); -out: - read_unlock(&tasklist_lock); - - return found; -} static int match(const struct sk_buff *skb, @@ -112,18 +47,6 @@ match(const struct sk_buff *skb, return 0; } - if(info->match & IP6T_OWNER_PID) { - if (!match_pid(skb, info->pid) ^ - !!(info->invert & IP6T_OWNER_PID)) - return 0; - } - - if(info->match & IP6T_OWNER_SID) { - if (!match_sid(skb, info->sid) ^ - !!(info->invert & IP6T_OWNER_SID)) - return 0; - } - return 1; } @@ -134,6 +57,8 @@ checkentry(const char *tablename, unsigned int matchsize, unsigned int hook_mask) { + const struct ip6t_owner_info *info = matchinfo; + if (hook_mask & ~((1 << NF_IP6_LOCAL_OUT) | (1 << NF_IP6_POST_ROUTING))) { printk("ip6t_owner: only valid for LOCAL_OUT or POST_ROUTING.\n"); @@ -142,14 +67,13 @@ checkentry(const char *tablename, if (matchsize != IP6T_ALIGN(sizeof(struct ip6t_owner_info))) return 0; -#ifdef CONFIG_SMP - /* files->file_lock can not be used in a BH */ - if (((struct ip6t_owner_info *)matchinfo)->match - & (IP6T_OWNER_PID|IP6T_OWNER_SID)) { - printk("ip6t_owner: pid and sid matching is broken on SMP.\n"); + + if (info->match & (IP6T_OWNER_PID|IP6T_OWNER_SID)) { + printk("ipt_owner: pid and sid matching " + "not supported anymore\n"); return 0; } -#endif + return 1; } diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 1d4d75b34d32..7a5863298f3f 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -49,6 +49,7 @@ #include <net/transp_v6.h> #include <net/udp.h> #include <net/inet_common.h> +#include <net/tcp_states.h> #include <net/rawv6.h> #include <net/xfrm.h> @@ -81,7 +82,8 @@ static void raw_v6_unhash(struct sock *sk) /* Grumble... icmp and ip_input want to get at this... */ struct sock *__raw_v6_lookup(struct sock *sk, unsigned short num, - struct in6_addr *loc_addr, struct in6_addr *rmt_addr) + struct in6_addr *loc_addr, struct in6_addr *rmt_addr, + int dif) { struct hlist_node *node; int is_multicast = ipv6_addr_is_multicast(loc_addr); @@ -94,6 +96,9 @@ struct sock *__raw_v6_lookup(struct sock *sk, unsigned short num, !ipv6_addr_equal(&np->daddr, rmt_addr)) continue; + if (sk->sk_bound_dev_if && sk->sk_bound_dev_if != dif) + continue; + if (!ipv6_addr_any(&np->rcv_saddr)) { if (ipv6_addr_equal(&np->rcv_saddr, loc_addr)) goto found; @@ -137,11 +142,12 @@ static __inline__ int icmpv6_filter(struct sock *sk, struct sk_buff *skb) * * Caller owns SKB so we must make clones. */ -void ipv6_raw_deliver(struct sk_buff *skb, int nexthdr) +int ipv6_raw_deliver(struct sk_buff *skb, int nexthdr) { struct in6_addr *saddr; struct in6_addr *daddr; struct sock *sk; + int delivered = 0; __u8 hash; saddr = &skb->nh.ipv6h->saddr; @@ -160,9 +166,10 @@ void ipv6_raw_deliver(struct sk_buff *skb, int nexthdr) if (sk == NULL) goto out; - sk = __raw_v6_lookup(sk, nexthdr, daddr, saddr); + sk = __raw_v6_lookup(sk, nexthdr, daddr, saddr, skb->dev->ifindex); while (sk) { + delivered = 1; if (nexthdr != IPPROTO_ICMPV6 || !icmpv6_filter(sk, skb)) { struct sk_buff *clone = skb_clone(skb, GFP_ATOMIC); @@ -170,10 +177,12 @@ void ipv6_raw_deliver(struct sk_buff *skb, int nexthdr) if (clone) rawv6_rcv(sk, clone); } - sk = __raw_v6_lookup(sk_next(sk), nexthdr, daddr, saddr); + sk = __raw_v6_lookup(sk_next(sk), nexthdr, daddr, saddr, + skb->dev->ifindex); } out: read_unlock(&raw_v6_lock); + return delivered; } /* This cleans up af_inet6 a bit. -DaveM */ @@ -334,8 +343,7 @@ int rawv6_rcv(struct sock *sk, struct sk_buff *skb) if (csum_ipv6_magic(&skb->nh.ipv6h->saddr, &skb->nh.ipv6h->daddr, skb->len, inet->num, skb->csum)) { - LIMIT_NETDEBUG( - printk(KERN_DEBUG "raw v6 hw csum failure.\n")); + LIMIT_NETDEBUG(KERN_DEBUG "raw v6 hw csum failure.\n"); skb->ip_summed = CHECKSUM_NONE; } } diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c index 59e7c6317872..9d9e04344c77 100644 --- a/net/ipv6/reassembly.c +++ b/net/ipv6/reassembly.c @@ -562,7 +562,7 @@ static void ip6_frag_queue(struct frag_queue *fq, struct sk_buff *skb, if (skb->dev) fq->iif = skb->dev->ifindex; skb->dev = NULL; - fq->stamp = skb->stamp; + skb_get_timestamp(skb, &fq->stamp); fq->meat += skb->len; atomic_add(skb->truesize, &ip6_frag_mem); @@ -664,7 +664,7 @@ static int ip6_frag_reasm(struct frag_queue *fq, struct sk_buff **skb_in, head->next = NULL; head->dev = dev; - head->stamp = fq->stamp; + skb_set_timestamp(head, &fq->stamp); head->nh.ipv6h->payload_len = htons(payload_len); *skb_in = head; diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 878789b3122d..5d5bbb49ec78 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1372,7 +1372,7 @@ int ipv6_route_ioctl(unsigned int cmd, void __user *arg) * Drop the packet on the floor */ -int ip6_pkt_discard(struct sk_buff *skb) +static int ip6_pkt_discard(struct sk_buff *skb) { IP6_INC_STATS(IPSTATS_MIB_OUTNOROUTES); icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_NOROUTE, 0, skb->dev); @@ -1380,7 +1380,7 @@ int ip6_pkt_discard(struct sk_buff *skb) return 0; } -int ip6_pkt_discard_out(struct sk_buff *skb) +static int ip6_pkt_discard_out(struct sk_buff *skb) { skb->dev = skb->dst->dev; return ip6_pkt_discard(skb); @@ -1850,16 +1850,16 @@ void inet6_rt_notify(int event, struct rt6_info *rt, struct nlmsghdr *nlh, skb = alloc_skb(size, gfp_any()); if (!skb) { - netlink_set_err(rtnl, 0, RTMGRP_IPV6_ROUTE, ENOBUFS); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_ROUTE, ENOBUFS); return; } if (rt6_fill_node(skb, rt, NULL, NULL, 0, event, pid, seq, 0, 0) < 0) { kfree_skb(skb); - netlink_set_err(rtnl, 0, RTMGRP_IPV6_ROUTE, EINVAL); + netlink_set_err(rtnl, 0, RTNLGRP_IPV6_ROUTE, EINVAL); return; } - NETLINK_CB(skb).dst_groups = RTMGRP_IPV6_ROUTE; - netlink_broadcast(rtnl, skb, 0, RTMGRP_IPV6_ROUTE, gfp_any()); + NETLINK_CB(skb).dst_group = RTNLGRP_IPV6_ROUTE; + netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV6_ROUTE, gfp_any()); } /* @@ -1960,8 +1960,6 @@ static int rt6_proc_info(char *buffer, char **start, off_t offset, int length) return arg.len; } -extern struct rt6_statistics rt6_stats; - static int rt6_stats_seq_show(struct seq_file *seq, void *v) { seq_printf(seq, "%04x %04x %04x %04x %04x %04x %04x\n", diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c index e553e5b80d6e..c3123c9e1a8e 100644 --- a/net/ipv6/sit.c +++ b/net/ipv6/sit.c @@ -770,7 +770,7 @@ static int ipip6_tunnel_init(struct net_device *dev) return 0; } -int __init ipip6_fb_tunnel_init(struct net_device *dev) +static int __init ipip6_fb_tunnel_init(struct net_device *dev) { struct ip_tunnel *tunnel = dev->priv; struct iphdr *iph = &tunnel->parms.iph; diff --git a/net/ipv6/sysctl_net_ipv6.c b/net/ipv6/sysctl_net_ipv6.c index 3a18e0e6ffed..8eff9fa1e983 100644 --- a/net/ipv6/sysctl_net_ipv6.c +++ b/net/ipv6/sysctl_net_ipv6.c @@ -14,9 +14,6 @@ #include <net/ipv6.h> #include <net/addrconf.h> -extern ctl_table ipv6_route_table[]; -extern ctl_table ipv6_icmp_table[]; - #ifdef CONFIG_SYSCTL static ctl_table ipv6_table[] = { diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index ef29cfd936d3..794734f1d230 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -47,6 +47,7 @@ #include <net/tcp.h> #include <net/ndisc.h> +#include <net/inet6_hashtables.h> #include <net/ipv6.h> #include <net/transp_v6.h> #include <net/addrconf.h> @@ -75,34 +76,11 @@ static int tcp_v6_xmit(struct sk_buff *skb, int ipfragok); static struct tcp_func ipv6_mapped; static struct tcp_func ipv6_specific; -/* I have no idea if this is a good hash for v6 or not. -DaveM */ -static __inline__ int tcp_v6_hashfn(struct in6_addr *laddr, u16 lport, - struct in6_addr *faddr, u16 fport) +static inline int tcp_v6_bind_conflict(const struct sock *sk, + const struct inet_bind_bucket *tb) { - int hashent = (lport ^ fport); - - hashent ^= (laddr->s6_addr32[3] ^ faddr->s6_addr32[3]); - hashent ^= hashent>>16; - hashent ^= hashent>>8; - return (hashent & (tcp_ehash_size - 1)); -} - -static __inline__ int tcp_v6_sk_hashfn(struct sock *sk) -{ - struct inet_sock *inet = inet_sk(sk); - struct ipv6_pinfo *np = inet6_sk(sk); - struct in6_addr *laddr = &np->rcv_saddr; - struct in6_addr *faddr = &np->daddr; - __u16 lport = inet->num; - __u16 fport = inet->dport; - return tcp_v6_hashfn(laddr, lport, faddr, fport); -} - -static inline int tcp_v6_bind_conflict(struct sock *sk, - struct tcp_bind_bucket *tb) -{ - struct sock *sk2; - struct hlist_node *node; + const struct sock *sk2; + const struct hlist_node *node; /* We must walk the whole port owner list in this case. -DaveM */ sk_for_each_bound(sk2, node, &tb->owners) { @@ -126,8 +104,8 @@ static inline int tcp_v6_bind_conflict(struct sock *sk, */ static int tcp_v6_get_port(struct sock *sk, unsigned short snum) { - struct tcp_bind_hashbucket *head; - struct tcp_bind_bucket *tb; + struct inet_bind_hashbucket *head; + struct inet_bind_bucket *tb; struct hlist_node *node; int ret; @@ -138,25 +116,25 @@ static int tcp_v6_get_port(struct sock *sk, unsigned short snum) int remaining = (high - low) + 1; int rover; - spin_lock(&tcp_portalloc_lock); - if (tcp_port_rover < low) + spin_lock(&tcp_hashinfo.portalloc_lock); + if (tcp_hashinfo.port_rover < low) rover = low; else - rover = tcp_port_rover; + rover = tcp_hashinfo.port_rover; do { rover++; if (rover > high) rover = low; - head = &tcp_bhash[tcp_bhashfn(rover)]; + head = &tcp_hashinfo.bhash[inet_bhashfn(rover, tcp_hashinfo.bhash_size)]; spin_lock(&head->lock); - tb_for_each(tb, node, &head->chain) + inet_bind_bucket_for_each(tb, node, &head->chain) if (tb->port == rover) goto next; break; next: spin_unlock(&head->lock); } while (--remaining > 0); - tcp_port_rover = rover; - spin_unlock(&tcp_portalloc_lock); + tcp_hashinfo.port_rover = rover; + spin_unlock(&tcp_hashinfo.portalloc_lock); /* Exhausted local port range during search? It is not * possible for us to be holding one of the bind hash @@ -171,9 +149,9 @@ static int tcp_v6_get_port(struct sock *sk, unsigned short snum) /* OK, here is the one we will use. */ snum = rover; } else { - head = &tcp_bhash[tcp_bhashfn(snum)]; + head = &tcp_hashinfo.bhash[inet_bhashfn(snum, tcp_hashinfo.bhash_size)]; spin_lock(&head->lock); - tb_for_each(tb, node, &head->chain) + inet_bind_bucket_for_each(tb, node, &head->chain) if (tb->port == snum) goto tb_found; } @@ -192,8 +170,11 @@ tb_found: } tb_not_found: ret = 1; - if (!tb && (tb = tcp_bucket_create(head, snum)) == NULL) - goto fail_unlock; + if (tb == NULL) { + tb = inet_bind_bucket_create(tcp_hashinfo.bind_bucket_cachep, head, snum); + if (tb == NULL) + goto fail_unlock; + } if (hlist_empty(&tb->owners)) { if (sk->sk_reuse && sk->sk_state != TCP_LISTEN) tb->fastreuse = 1; @@ -204,9 +185,9 @@ tb_not_found: tb->fastreuse = 0; success: - if (!tcp_sk(sk)->bind_hash) - tcp_bind_hash(sk, tb, snum); - BUG_TRAP(tcp_sk(sk)->bind_hash == tb); + if (!inet_csk(sk)->icsk_bind_hash) + inet_bind_hash(sk, tb, snum); + BUG_TRAP(inet_csk(sk)->icsk_bind_hash == tb); ret = 0; fail_unlock: @@ -224,13 +205,13 @@ static __inline__ void __tcp_v6_hash(struct sock *sk) BUG_TRAP(sk_unhashed(sk)); if (sk->sk_state == TCP_LISTEN) { - list = &tcp_listening_hash[tcp_sk_listen_hashfn(sk)]; - lock = &tcp_lhash_lock; - tcp_listen_wlock(); + list = &tcp_hashinfo.listening_hash[inet_sk_listen_hashfn(sk)]; + lock = &tcp_hashinfo.lhash_lock; + inet_listen_wlock(&tcp_hashinfo); } else { - sk->sk_hashent = tcp_v6_sk_hashfn(sk); - list = &tcp_ehash[sk->sk_hashent].chain; - lock = &tcp_ehash[sk->sk_hashent].lock; + sk->sk_hashent = inet6_sk_ehashfn(sk, tcp_hashinfo.ehash_size); + list = &tcp_hashinfo.ehash[sk->sk_hashent].chain; + lock = &tcp_hashinfo.ehash[sk->sk_hashent].lock; write_lock(lock); } @@ -255,131 +236,11 @@ static void tcp_v6_hash(struct sock *sk) } } -static struct sock *tcp_v6_lookup_listener(struct in6_addr *daddr, unsigned short hnum, int dif) -{ - struct sock *sk; - struct hlist_node *node; - struct sock *result = NULL; - int score, hiscore; - - hiscore=0; - read_lock(&tcp_lhash_lock); - sk_for_each(sk, node, &tcp_listening_hash[tcp_lhashfn(hnum)]) { - if (inet_sk(sk)->num == hnum && sk->sk_family == PF_INET6) { - struct ipv6_pinfo *np = inet6_sk(sk); - - score = 1; - if (!ipv6_addr_any(&np->rcv_saddr)) { - if (!ipv6_addr_equal(&np->rcv_saddr, daddr)) - continue; - score++; - } - if (sk->sk_bound_dev_if) { - if (sk->sk_bound_dev_if != dif) - continue; - score++; - } - if (score == 3) { - result = sk; - break; - } - if (score > hiscore) { - hiscore = score; - result = sk; - } - } - } - if (result) - sock_hold(result); - read_unlock(&tcp_lhash_lock); - return result; -} - -/* Sockets in TCP_CLOSE state are _always_ taken out of the hash, so - * we need not check it for TCP lookups anymore, thanks Alexey. -DaveM - * - * The sockhash lock must be held as a reader here. - */ - -static inline struct sock *__tcp_v6_lookup_established(struct in6_addr *saddr, u16 sport, - struct in6_addr *daddr, u16 hnum, - int dif) -{ - struct tcp_ehash_bucket *head; - struct sock *sk; - struct hlist_node *node; - __u32 ports = TCP_COMBINED_PORTS(sport, hnum); - int hash; - - /* Optimize here for direct hit, only listening connections can - * have wildcards anyways. - */ - hash = tcp_v6_hashfn(daddr, hnum, saddr, sport); - head = &tcp_ehash[hash]; - read_lock(&head->lock); - sk_for_each(sk, node, &head->chain) { - /* For IPV6 do the cheaper port and family tests first. */ - if(TCP_IPV6_MATCH(sk, saddr, daddr, ports, dif)) - goto hit; /* You sunk my battleship! */ - } - /* Must check for a TIME_WAIT'er before going to listener hash. */ - sk_for_each(sk, node, &(head + tcp_ehash_size)->chain) { - /* FIXME: acme: check this... */ - struct tcp_tw_bucket *tw = (struct tcp_tw_bucket *)sk; - - if(*((__u32 *)&(tw->tw_dport)) == ports && - sk->sk_family == PF_INET6) { - if(ipv6_addr_equal(&tw->tw_v6_daddr, saddr) && - ipv6_addr_equal(&tw->tw_v6_rcv_saddr, daddr) && - (!sk->sk_bound_dev_if || sk->sk_bound_dev_if == dif)) - goto hit; - } - } - read_unlock(&head->lock); - return NULL; - -hit: - sock_hold(sk); - read_unlock(&head->lock); - return sk; -} - - -static inline struct sock *__tcp_v6_lookup(struct in6_addr *saddr, u16 sport, - struct in6_addr *daddr, u16 hnum, - int dif) -{ - struct sock *sk; - - sk = __tcp_v6_lookup_established(saddr, sport, daddr, hnum, dif); - - if (sk) - return sk; - - return tcp_v6_lookup_listener(daddr, hnum, dif); -} - -inline struct sock *tcp_v6_lookup(struct in6_addr *saddr, u16 sport, - struct in6_addr *daddr, u16 dport, - int dif) -{ - struct sock *sk; - - local_bh_disable(); - sk = __tcp_v6_lookup(saddr, sport, daddr, ntohs(dport), dif); - local_bh_enable(); - - return sk; -} - -EXPORT_SYMBOL_GPL(tcp_v6_lookup); - - /* * Open request hash tables. */ -static u32 tcp_v6_synq_hash(struct in6_addr *raddr, u16 rport, u32 rnd) +static u32 tcp_v6_synq_hash(const struct in6_addr *raddr, const u16 rport, const u32 rnd) { u32 a, b, c; @@ -399,14 +260,15 @@ static u32 tcp_v6_synq_hash(struct in6_addr *raddr, u16 rport, u32 rnd) return c & (TCP_SYNQ_HSIZE - 1); } -static struct request_sock *tcp_v6_search_req(struct tcp_sock *tp, +static struct request_sock *tcp_v6_search_req(const struct sock *sk, struct request_sock ***prevp, __u16 rport, struct in6_addr *raddr, struct in6_addr *laddr, int iif) { - struct listen_sock *lopt = tp->accept_queue.listen_opt; + const struct inet_connection_sock *icsk = inet_csk(sk); + struct listen_sock *lopt = icsk->icsk_accept_queue.listen_opt; struct request_sock *req, **prev; for (prev = &lopt->syn_table[tcp_v6_synq_hash(raddr, rport, lopt->hash_rnd)]; @@ -451,44 +313,48 @@ static __u32 tcp_v6_init_sequence(struct sock *sk, struct sk_buff *skb) } } -static int __tcp_v6_check_established(struct sock *sk, __u16 lport, - struct tcp_tw_bucket **twp) +static int __tcp_v6_check_established(struct sock *sk, const __u16 lport, + struct inet_timewait_sock **twp) { struct inet_sock *inet = inet_sk(sk); - struct ipv6_pinfo *np = inet6_sk(sk); - struct in6_addr *daddr = &np->rcv_saddr; - struct in6_addr *saddr = &np->daddr; - int dif = sk->sk_bound_dev_if; - u32 ports = TCP_COMBINED_PORTS(inet->dport, lport); - int hash = tcp_v6_hashfn(daddr, inet->num, saddr, inet->dport); - struct tcp_ehash_bucket *head = &tcp_ehash[hash]; + const struct ipv6_pinfo *np = inet6_sk(sk); + const struct in6_addr *daddr = &np->rcv_saddr; + const struct in6_addr *saddr = &np->daddr; + const int dif = sk->sk_bound_dev_if; + const u32 ports = INET_COMBINED_PORTS(inet->dport, lport); + const int hash = inet6_ehashfn(daddr, inet->num, saddr, inet->dport, + tcp_hashinfo.ehash_size); + struct inet_ehash_bucket *head = &tcp_hashinfo.ehash[hash]; struct sock *sk2; - struct hlist_node *node; - struct tcp_tw_bucket *tw; + const struct hlist_node *node; + struct inet_timewait_sock *tw; write_lock(&head->lock); /* Check TIME-WAIT sockets first. */ - sk_for_each(sk2, node, &(head + tcp_ehash_size)->chain) { - tw = (struct tcp_tw_bucket*)sk2; + sk_for_each(sk2, node, &(head + tcp_hashinfo.ehash_size)->chain) { + const struct tcp6_timewait_sock *tcp6tw = tcp6_twsk(sk2); + + tw = inet_twsk(sk2); if(*((__u32 *)&(tw->tw_dport)) == ports && sk2->sk_family == PF_INET6 && - ipv6_addr_equal(&tw->tw_v6_daddr, saddr) && - ipv6_addr_equal(&tw->tw_v6_rcv_saddr, daddr) && + ipv6_addr_equal(&tcp6tw->tw_v6_daddr, saddr) && + ipv6_addr_equal(&tcp6tw->tw_v6_rcv_saddr, daddr) && sk2->sk_bound_dev_if == sk->sk_bound_dev_if) { + const struct tcp_timewait_sock *tcptw = tcp_twsk(sk2); struct tcp_sock *tp = tcp_sk(sk); - if (tw->tw_ts_recent_stamp && - (!twp || (sysctl_tcp_tw_reuse && - xtime.tv_sec - - tw->tw_ts_recent_stamp > 1))) { + if (tcptw->tw_ts_recent_stamp && + (!twp || + (sysctl_tcp_tw_reuse && + xtime.tv_sec - tcptw->tw_ts_recent_stamp > 1))) { /* See comment in tcp_ipv4.c */ - tp->write_seq = tw->tw_snd_nxt + 65535 + 2; + tp->write_seq = tcptw->tw_snd_nxt + 65535 + 2; if (!tp->write_seq) tp->write_seq = 1; - tp->rx_opt.ts_recent = tw->tw_ts_recent; - tp->rx_opt.ts_recent_stamp = tw->tw_ts_recent_stamp; + tp->rx_opt.ts_recent = tcptw->tw_ts_recent; + tp->rx_opt.ts_recent_stamp = tcptw->tw_ts_recent_stamp; sock_hold(sk2); goto unique; } else @@ -499,7 +365,7 @@ static int __tcp_v6_check_established(struct sock *sk, __u16 lport, /* And established part... */ sk_for_each(sk2, node, &head->chain) { - if(TCP_IPV6_MATCH(sk2, saddr, daddr, ports, dif)) + if (INET6_MATCH(sk2, saddr, daddr, ports, dif)) goto not_unique; } @@ -515,10 +381,10 @@ unique: NET_INC_STATS_BH(LINUX_MIB_TIMEWAITRECYCLED); } else if (tw) { /* Silly. Should hash-dance instead... */ - tcp_tw_deschedule(tw); + inet_twsk_deschedule(tw, &tcp_death_row); NET_INC_STATS_BH(LINUX_MIB_TIMEWAITRECYCLED); - tcp_tw_put(tw); + inet_twsk_put(tw); } return 0; @@ -540,8 +406,8 @@ static inline u32 tcpv6_port_offset(const struct sock *sk) static int tcp_v6_hash_connect(struct sock *sk) { unsigned short snum = inet_sk(sk)->num; - struct tcp_bind_hashbucket *head; - struct tcp_bind_bucket *tb; + struct inet_bind_hashbucket *head; + struct inet_bind_bucket *tb; int ret; if (!snum) { @@ -553,19 +419,19 @@ static int tcp_v6_hash_connect(struct sock *sk) static u32 hint; u32 offset = hint + tcpv6_port_offset(sk); struct hlist_node *node; - struct tcp_tw_bucket *tw = NULL; + struct inet_timewait_sock *tw = NULL; local_bh_disable(); for (i = 1; i <= range; i++) { port = low + (i + offset) % range; - head = &tcp_bhash[tcp_bhashfn(port)]; + head = &tcp_hashinfo.bhash[inet_bhashfn(port, tcp_hashinfo.bhash_size)]; spin_lock(&head->lock); /* Does not bother with rcv_saddr checks, * because the established check is already * unique enough. */ - tb_for_each(tb, node, &head->chain) { + inet_bind_bucket_for_each(tb, node, &head->chain) { if (tb->port == port) { BUG_TRAP(!hlist_empty(&tb->owners)); if (tb->fastreuse >= 0) @@ -578,7 +444,7 @@ static int tcp_v6_hash_connect(struct sock *sk) } } - tb = tcp_bucket_create(head, port); + tb = inet_bind_bucket_create(tcp_hashinfo.bind_bucket_cachep, head, port); if (!tb) { spin_unlock(&head->lock); break; @@ -597,7 +463,7 @@ ok: hint += i; /* Head lock still held and bh's disabled */ - tcp_bind_hash(sk, tb, port); + inet_bind_hash(sk, tb, port); if (sk_unhashed(sk)) { inet_sk(sk)->sport = htons(port); __tcp_v6_hash(sk); @@ -605,16 +471,16 @@ ok: spin_unlock(&head->lock); if (tw) { - tcp_tw_deschedule(tw); - tcp_tw_put(tw); + inet_twsk_deschedule(tw, &tcp_death_row); + inet_twsk_put(tw); } ret = 0; goto out; } - head = &tcp_bhash[tcp_bhashfn(snum)]; - tb = tcp_sk(sk)->bind_hash; + head = &tcp_hashinfo.bhash[inet_bhashfn(snum, tcp_hashinfo.bhash_size)]; + tb = inet_csk(sk)->icsk_bind_hash; spin_lock_bh(&head->lock); if (sk_head(&tb->owners) == sk && !sk->sk_bind_node.next) { @@ -631,11 +497,6 @@ out: } } -static __inline__ int tcp_v6_iif(struct sk_buff *skb) -{ - return IP6CB(skb)->iif; -} - static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) { @@ -827,14 +688,15 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, int type, int code, int offset, __u32 info) { struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data; - struct tcphdr *th = (struct tcphdr *)(skb->data+offset); + const struct tcphdr *th = (struct tcphdr *)(skb->data+offset); struct ipv6_pinfo *np; struct sock *sk; int err; struct tcp_sock *tp; __u32 seq; - sk = tcp_v6_lookup(&hdr->daddr, th->dest, &hdr->saddr, th->source, skb->dev->ifindex); + sk = inet6_lookup(&tcp_hashinfo, &hdr->daddr, th->dest, &hdr->saddr, + th->source, skb->dev->ifindex); if (sk == NULL) { ICMP6_INC_STATS_BH(__in6_dev_get(skb->dev), ICMP6_MIB_INERRORS); @@ -842,7 +704,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, } if (sk->sk_state == TCP_TIME_WAIT) { - tcp_tw_put((struct tcp_tw_bucket*)sk); + inet_twsk_put((struct inet_timewait_sock *)sk); return; } @@ -920,8 +782,8 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, if (sock_owned_by_user(sk)) goto out; - req = tcp_v6_search_req(tp, &prev, th->dest, &hdr->daddr, - &hdr->saddr, tcp_v6_iif(skb)); + req = tcp_v6_search_req(sk, &prev, th->dest, &hdr->daddr, + &hdr->saddr, inet6_iif(skb)); if (!req) goto out; @@ -935,7 +797,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, goto out; } - tcp_synq_drop(sk, req, prev); + inet_csk_reqsk_queue_drop(sk, req, prev); goto out; case TCP_SYN_SENT: @@ -1132,7 +994,7 @@ static void tcp_v6_send_reset(struct sk_buff *skb) buff->csum); fl.proto = IPPROTO_TCP; - fl.oif = tcp_v6_iif(skb); + fl.oif = inet6_iif(skb); fl.fl_ip_dport = t1->dest; fl.fl_ip_sport = t1->source; @@ -1201,7 +1063,7 @@ static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32 buff->csum); fl.proto = IPPROTO_TCP; - fl.oif = tcp_v6_iif(skb); + fl.oif = inet6_iif(skb); fl.fl_ip_dport = t1->dest; fl.fl_ip_sport = t1->source; @@ -1220,12 +1082,14 @@ static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32 static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb) { - struct tcp_tw_bucket *tw = (struct tcp_tw_bucket *)sk; + struct inet_timewait_sock *tw = inet_twsk(sk); + const struct tcp_timewait_sock *tcptw = tcp_twsk(sk); - tcp_v6_send_ack(skb, tw->tw_snd_nxt, tw->tw_rcv_nxt, - tw->tw_rcv_wnd >> tw->tw_rcv_wscale, tw->tw_ts_recent); + tcp_v6_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt, + tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale, + tcptw->tw_ts_recent); - tcp_tw_put(tw); + inet_twsk_put(tw); } static void tcp_v6_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req) @@ -1237,28 +1101,25 @@ static void tcp_v6_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req) static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb) { struct request_sock *req, **prev; - struct tcphdr *th = skb->h.th; - struct tcp_sock *tp = tcp_sk(sk); + const struct tcphdr *th = skb->h.th; struct sock *nsk; /* Find possible connection requests. */ - req = tcp_v6_search_req(tp, &prev, th->source, &skb->nh.ipv6h->saddr, - &skb->nh.ipv6h->daddr, tcp_v6_iif(skb)); + req = tcp_v6_search_req(sk, &prev, th->source, &skb->nh.ipv6h->saddr, + &skb->nh.ipv6h->daddr, inet6_iif(skb)); if (req) return tcp_check_req(sk, skb, req, prev); - nsk = __tcp_v6_lookup_established(&skb->nh.ipv6h->saddr, - th->source, - &skb->nh.ipv6h->daddr, - ntohs(th->dest), - tcp_v6_iif(skb)); + nsk = __inet6_lookup_established(&tcp_hashinfo, &skb->nh.ipv6h->saddr, + th->source, &skb->nh.ipv6h->daddr, + ntohs(th->dest), inet6_iif(skb)); if (nsk) { if (nsk->sk_state != TCP_TIME_WAIT) { bh_lock_sock(nsk); return nsk; } - tcp_tw_put((struct tcp_tw_bucket*)nsk); + inet_twsk_put((struct inet_timewait_sock *)nsk); return NULL; } @@ -1271,12 +1132,12 @@ static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb) static void tcp_v6_synq_add(struct sock *sk, struct request_sock *req) { - struct tcp_sock *tp = tcp_sk(sk); - struct listen_sock *lopt = tp->accept_queue.listen_opt; - u32 h = tcp_v6_synq_hash(&tcp6_rsk(req)->rmt_addr, inet_rsk(req)->rmt_port, lopt->hash_rnd); + struct inet_connection_sock *icsk = inet_csk(sk); + struct listen_sock *lopt = icsk->icsk_accept_queue.listen_opt; + const u32 h = tcp_v6_synq_hash(&tcp6_rsk(req)->rmt_addr, inet_rsk(req)->rmt_port, lopt->hash_rnd); - reqsk_queue_hash_req(&tp->accept_queue, h, req, TCP_TIMEOUT_INIT); - tcp_synq_added(sk); + reqsk_queue_hash_req(&icsk->icsk_accept_queue, h, req, TCP_TIMEOUT_INIT); + inet_csk_reqsk_queue_added(sk, TCP_TIMEOUT_INIT); } @@ -1301,13 +1162,13 @@ static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb) /* * There are no SYN attacks on IPv6, yet... */ - if (tcp_synq_is_full(sk) && !isn) { + if (inet_csk_reqsk_queue_is_full(sk) && !isn) { if (net_ratelimit()) printk(KERN_INFO "TCPv6: dropping request, synflood is possible\n"); goto drop; } - if (sk_acceptq_is_full(sk) && tcp_synq_young(sk) > 1) + if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1) goto drop; req = reqsk_alloc(&tcp6_request_sock_ops); @@ -1339,7 +1200,7 @@ static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb) /* So that link locals have meaning */ if (!sk->sk_bound_dev_if && ipv6_addr_type(&treq->rmt_addr) & IPV6_ADDR_LINKLOCAL) - treq->iif = tcp_v6_iif(skb); + treq->iif = inet6_iif(skb); if (isn == 0) isn = tcp_v6_init_sequence(sk,skb); @@ -1404,15 +1265,14 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, newsk->sk_backlog_rcv = tcp_v4_do_rcv; newnp->pktoptions = NULL; newnp->opt = NULL; - newnp->mcast_oif = tcp_v6_iif(skb); + newnp->mcast_oif = inet6_iif(skb); newnp->mcast_hops = skb->nh.ipv6h->hop_limit; - /* Charge newly allocated IPv6 socket. Though it is mapped, - * it is IPv6 yet. + /* + * No need to charge this sock to the relevant IPv6 refcnt debug socks count + * here, tcp_create_openreq_child now does this for us, see the comment in + * that function for the gory details. -acme */ -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet6_sock_nr); -#endif /* It is tricky place. Until this moment IPv4 tcp worked with IPv6 af_tcp.af_specific. @@ -1467,10 +1327,11 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, if (newsk == NULL) goto out; - /* Charge newly allocated IPv6 socket */ -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet6_sock_nr); -#endif + /* + * No need to charge this sock to the relevant IPv6 refcnt debug socks + * count here, tcp_create_openreq_child now does this for us, see the + * comment in that function for the gory details. -acme + */ ip6_dst_store(newsk, dst, NULL); newsk->sk_route_caps = dst->dev->features & @@ -1509,7 +1370,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, skb_set_owner_r(newnp->pktoptions, newsk); } newnp->opt = NULL; - newnp->mcast_oif = tcp_v6_iif(skb); + newnp->mcast_oif = inet6_iif(skb); newnp->mcast_hops = skb->nh.ipv6h->hop_limit; /* Clone native IPv6 options from listening socket (if any) @@ -1536,7 +1397,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, newinet->daddr = newinet->saddr = newinet->rcv_saddr = LOOPBACK4_IPV6; __tcp_v6_hash(newsk); - tcp_inherit_port(sk, newsk); + inet_inherit_port(&tcp_hashinfo, sk, newsk); return newsk; @@ -1557,7 +1418,7 @@ static int tcp_v6_checksum_init(struct sk_buff *skb) if (!tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr, &skb->nh.ipv6h->daddr,skb->csum)) return 0; - LIMIT_NETDEBUG(printk(KERN_DEBUG "hw tcp v6 csum failed\n")); + LIMIT_NETDEBUG(KERN_DEBUG "hw tcp v6 csum failed\n"); } if (skb->len <= 76) { if (tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr, @@ -1684,7 +1545,7 @@ ipv6_pktoptions: if (TCP_SKB_CB(opt_skb)->end_seq == tp->rcv_nxt && !((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))) { if (np->rxopt.bits.rxinfo) - np->mcast_oif = tcp_v6_iif(opt_skb); + np->mcast_oif = inet6_iif(opt_skb); if (np->rxopt.bits.rxhlim) np->mcast_hops = opt_skb->nh.ipv6h->hop_limit; if (ipv6_opt_accepted(sk, opt_skb)) { @@ -1739,8 +1600,9 @@ static int tcp_v6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) TCP_SKB_CB(skb)->flags = ipv6_get_dsfield(skb->nh.ipv6h); TCP_SKB_CB(skb)->sacked = 0; - sk = __tcp_v6_lookup(&skb->nh.ipv6h->saddr, th->source, - &skb->nh.ipv6h->daddr, ntohs(th->dest), tcp_v6_iif(skb)); + sk = __inet6_lookup(&tcp_hashinfo, &skb->nh.ipv6h->saddr, th->source, + &skb->nh.ipv6h->daddr, ntohs(th->dest), + inet6_iif(skb)); if (!sk) goto no_tcp_socket; @@ -1795,26 +1657,29 @@ discard_and_relse: do_time_wait: if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) { - tcp_tw_put((struct tcp_tw_bucket *) sk); + inet_twsk_put((struct inet_timewait_sock *)sk); goto discard_it; } if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) { TCP_INC_STATS_BH(TCP_MIB_INERRS); - tcp_tw_put((struct tcp_tw_bucket *) sk); + inet_twsk_put((struct inet_timewait_sock *)sk); goto discard_it; } - switch(tcp_timewait_state_process((struct tcp_tw_bucket *)sk, - skb, th, skb->len)) { + switch (tcp_timewait_state_process((struct inet_timewait_sock *)sk, + skb, th)) { case TCP_TW_SYN: { struct sock *sk2; - sk2 = tcp_v6_lookup_listener(&skb->nh.ipv6h->daddr, ntohs(th->dest), tcp_v6_iif(skb)); + sk2 = inet6_lookup_listener(&tcp_hashinfo, + &skb->nh.ipv6h->daddr, + ntohs(th->dest), inet6_iif(skb)); if (sk2 != NULL) { - tcp_tw_deschedule((struct tcp_tw_bucket *)sk); - tcp_tw_put((struct tcp_tw_bucket *)sk); + struct inet_timewait_sock *tw = inet_twsk(sk); + inet_twsk_deschedule(tw, &tcp_death_row); + inet_twsk_put(tw); sk = sk2; goto process; } @@ -1983,7 +1848,7 @@ static struct tcp_func ipv6_specific = { static struct tcp_func ipv6_mapped = { .queue_xmit = ip_queue_xmit, .send_check = tcp_v4_send_check, - .rebuild_header = tcp_v4_rebuild_header, + .rebuild_header = inet_sk_rebuild_header, .conn_request = tcp_v6_conn_request, .syn_recv_sock = tcp_v6_syn_recv_sock, .remember_stamp = tcp_v4_remember_stamp, @@ -2002,13 +1867,14 @@ static struct tcp_func ipv6_mapped = { */ static int tcp_v6_init_sock(struct sock *sk) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); skb_queue_head_init(&tp->out_of_order_queue); tcp_init_xmit_timers(sk); tcp_prequeue_init(tp); - tp->rto = TCP_TIMEOUT_INIT; + icsk->icsk_rto = TCP_TIMEOUT_INIT; tp->mdev = TCP_TIMEOUT_INIT; /* So many TCP implementations out there (incorrectly) count the @@ -2030,7 +1896,7 @@ static int tcp_v6_init_sock(struct sock *sk) sk->sk_state = TCP_CLOSE; tp->af_specific = &ipv6_specific; - tp->ca_ops = &tcp_init_congestion_ops; + icsk->icsk_ca_ops = &tcp_init_congestion_ops; sk->sk_write_space = sk_stream_write_space; sock_set_flag(sk, SOCK_USE_WRITE_QUEUE); @@ -2044,8 +1910,6 @@ static int tcp_v6_init_sock(struct sock *sk) static int tcp_v6_destroy_sock(struct sock *sk) { - extern int tcp_v4_destroy_sock(struct sock *sk); - tcp_v4_destroy_sock(sk); return inet6_destroy_sock(sk); } @@ -2091,18 +1955,20 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) unsigned long timer_expires; struct inet_sock *inet = inet_sk(sp); struct tcp_sock *tp = tcp_sk(sp); + const struct inet_connection_sock *icsk = inet_csk(sp); struct ipv6_pinfo *np = inet6_sk(sp); dest = &np->daddr; src = &np->rcv_saddr; destp = ntohs(inet->dport); srcp = ntohs(inet->sport); - if (tp->pending == TCP_TIME_RETRANS) { + + if (icsk->icsk_pending == ICSK_TIME_RETRANS) { timer_active = 1; - timer_expires = tp->timeout; - } else if (tp->pending == TCP_TIME_PROBE0) { + timer_expires = icsk->icsk_timeout; + } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; - timer_expires = tp->timeout; + timer_expires = icsk->icsk_timeout; } else if (timer_pending(&sp->sk_timer)) { timer_active = 2; timer_expires = sp->sk_timer.expires; @@ -2123,28 +1989,31 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) tp->write_seq-tp->snd_una, tp->rcv_nxt-tp->copied_seq, timer_active, jiffies_to_clock_t(timer_expires - jiffies), - tp->retransmits, + icsk->icsk_retransmits, sock_i_uid(sp), - tp->probes_out, + icsk->icsk_probes_out, sock_i_ino(sp), atomic_read(&sp->sk_refcnt), sp, - tp->rto, tp->ack.ato, (tp->ack.quick<<1)|tp->ack.pingpong, + icsk->icsk_rto, + icsk->icsk_ack.ato, + (icsk->icsk_ack.quick << 1 ) | icsk->icsk_ack.pingpong, tp->snd_cwnd, tp->snd_ssthresh>=0xFFFF?-1:tp->snd_ssthresh ); } static void get_timewait6_sock(struct seq_file *seq, - struct tcp_tw_bucket *tw, int i) + struct inet_timewait_sock *tw, int i) { struct in6_addr *dest, *src; __u16 destp, srcp; + struct tcp6_timewait_sock *tcp6tw = tcp6_twsk((struct sock *)tw); int ttd = tw->tw_ttd - jiffies; if (ttd < 0) ttd = 0; - dest = &tw->tw_v6_daddr; - src = &tw->tw_v6_rcv_saddr; + dest = &tcp6tw->tw_v6_daddr; + src = &tcp6tw->tw_v6_rcv_saddr; destp = ntohs(tw->tw_dport); srcp = ntohs(tw->tw_sport); @@ -2219,7 +2088,7 @@ struct proto tcpv6_prot = { .close = tcp_close, .connect = tcp_v6_connect, .disconnect = tcp_disconnect, - .accept = tcp_accept, + .accept = inet_csk_accept, .ioctl = tcp_ioctl, .init = tcp_v6_init_sock, .destroy = tcp_v6_destroy_sock, @@ -2236,11 +2105,13 @@ struct proto tcpv6_prot = { .sockets_allocated = &tcp_sockets_allocated, .memory_allocated = &tcp_memory_allocated, .memory_pressure = &tcp_memory_pressure, + .orphan_count = &tcp_orphan_count, .sysctl_mem = sysctl_tcp_mem, .sysctl_wmem = sysctl_tcp_wmem, .sysctl_rmem = sysctl_tcp_rmem, .max_header = MAX_TCP_HEADER, .obj_size = sizeof(struct tcp6_sock), + .twsk_obj_size = sizeof(struct tcp6_timewait_sock), .rsk_prot = &tcp6_request_sock_ops, }; @@ -2250,8 +2121,6 @@ static struct inet6_protocol tcpv6_protocol = { .flags = INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL, }; -extern struct proto_ops inet6_stream_ops; - static struct inet_protosw tcpv6_protosw = { .type = SOCK_STREAM, .protocol = IPPROTO_TCP, diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index eff050ac7049..390d750449ce 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -51,6 +51,7 @@ #include <net/udp.h> #include <net/raw.h> #include <net/inet_common.h> +#include <net/tcp_states.h> #include <net/ip6_checksum.h> #include <net/xfrm.h> @@ -58,7 +59,7 @@ #include <linux/proc_fs.h> #include <linux/seq_file.h> -DEFINE_SNMP_STAT(struct udp_mib, udp_stats_in6); +DEFINE_SNMP_STAT(struct udp_mib, udp_stats_in6) __read_mostly; /* Grrr, addr_type already calculated by caller, but I don't want * to add some silly "cookie" argument to this method just for that. @@ -477,8 +478,7 @@ static int udpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) /* RFC 2460 section 8.1 says that we SHOULD log this error. Well, it is reasonable. */ - LIMIT_NETDEBUG( - printk(KERN_INFO "IPv6: udp checksum is 0\n")); + LIMIT_NETDEBUG(KERN_INFO "IPv6: udp checksum is 0\n"); goto discard; } @@ -493,7 +493,7 @@ static int udpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) if (skb->ip_summed==CHECKSUM_HW) { skb->ip_summed = CHECKSUM_UNNECESSARY; if (csum_ipv6_magic(saddr, daddr, ulen, IPPROTO_UDP, skb->csum)) { - LIMIT_NETDEBUG(printk(KERN_DEBUG "udp v6 hw csum failure.\n")); + LIMIT_NETDEBUG(KERN_DEBUG "udp v6 hw csum failure.\n"); skb->ip_summed = CHECKSUM_NONE; } } @@ -825,7 +825,7 @@ back_from_confirm: /* ... which is an evident application bug. --ANK */ release_sock(sk); - LIMIT_NETDEBUG(printk(KERN_DEBUG "udp cork app bug 2\n")); + LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n"); err = -EINVAL; goto out; } @@ -1054,8 +1054,6 @@ struct proto udpv6_prot = { .obj_size = sizeof(struct udp6_sock), }; -extern struct proto_ops inet6_dgram_ops; - static struct inet_protosw udpv6_protosw = { .type = SOCK_DGRAM, .protocol = IPPROTO_UDP, diff --git a/net/ipv6/xfrm6_tunnel.c b/net/ipv6/xfrm6_tunnel.c index 60c26c87277e..fbef7826a74f 100644 --- a/net/ipv6/xfrm6_tunnel.c +++ b/net/ipv6/xfrm6_tunnel.c @@ -79,7 +79,7 @@ static u32 xfrm6_tunnel_spi; #define XFRM6_TUNNEL_SPI_MIN 1 #define XFRM6_TUNNEL_SPI_MAX 0xffffffff -static kmem_cache_t *xfrm6_tunnel_spi_kmem; +static kmem_cache_t *xfrm6_tunnel_spi_kmem __read_mostly; #define XFRM6_TUNNEL_SPI_BYADDR_HSIZE 256 #define XFRM6_TUNNEL_SPI_BYSPI_HSIZE 256 diff --git a/net/ipx/af_ipx.c b/net/ipx/af_ipx.c index 5a27e5df5886..34b3bb868409 100644 --- a/net/ipx/af_ipx.c +++ b/net/ipx/af_ipx.c @@ -44,7 +44,6 @@ #include <linux/socket.h> #include <linux/sockios.h> #include <linux/string.h> -#include <linux/tcp.h> #include <linux/types.h> #include <linux/termios.h> @@ -52,6 +51,7 @@ #include <net/p8022.h> #include <net/psnap.h> #include <net/sock.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> @@ -1627,7 +1627,7 @@ out: return rc; } -static int ipx_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +static int ipx_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { /* NULL here for pt means the packet was looped back */ struct ipx_interface *intrfc; @@ -1796,8 +1796,8 @@ static int ipx_recvmsg(struct kiocb *iocb, struct socket *sock, copied); if (rc) goto out_free; - if (skb->stamp.tv_sec) - sk->sk_stamp = skb->stamp; + if (skb->tstamp.off_sec) + skb_get_timestamp(skb, &sk->sk_stamp); msg->msg_namelen = sizeof(*sipx); @@ -1940,9 +1940,7 @@ static struct notifier_block ipx_dev_notifier = { }; extern struct datalink_proto *make_EII_client(void); -extern struct datalink_proto *make_8023_client(void); extern void destroy_EII_client(struct datalink_proto *); -extern void destroy_8023_client(struct datalink_proto *); static unsigned char ipx_8022_type = 0xE0; static unsigned char ipx_snap_id[5] = { 0x0, 0x0, 0x0, 0x81, 0x37 }; diff --git a/net/ipx/ipx_proc.c b/net/ipx/ipx_proc.c index b6761913445a..1f73d9ea434d 100644 --- a/net/ipx/ipx_proc.c +++ b/net/ipx/ipx_proc.c @@ -10,7 +10,7 @@ #include <linux/proc_fs.h> #include <linux/spinlock.h> #include <linux/seq_file.h> -#include <linux/tcp.h> +#include <net/tcp_states.h> #include <net/ipx.h> static __inline__ struct ipx_interface *ipx_get_interface_idx(loff_t pos) diff --git a/net/irda/af_irda.c b/net/irda/af_irda.c index 92c6e8d4e731..6f92f9c62990 100644 --- a/net/irda/af_irda.c +++ b/net/irda/af_irda.c @@ -56,7 +56,7 @@ #include <asm/uaccess.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/irda/af_irda.h> diff --git a/net/irda/irlap_frame.c b/net/irda/irlap_frame.c index 6dafbb43b529..3e9a06abbdd0 100644 --- a/net/irda/irlap_frame.c +++ b/net/irda/irlap_frame.c @@ -988,9 +988,6 @@ void irlap_resend_rejected_frames(struct irlap_cb *self, int command) IRDA_DEBUG(0, "%s(), unable to copy\n", __FUNCTION__); return; } - /* Unlink tx_skb from list */ - tx_skb->next = tx_skb->prev = NULL; - tx_skb->list = NULL; /* Clear old Nr field + poll bit */ tx_skb->data[1] &= 0x0f; @@ -1063,9 +1060,6 @@ void irlap_resend_rejected_frame(struct irlap_cb *self, int command) IRDA_DEBUG(0, "%s(), unable to copy\n", __FUNCTION__); return; } - /* Unlink tx_skb from list */ - tx_skb->next = tx_skb->prev = NULL; - tx_skb->list = NULL; /* Clear old Nr field + poll bit */ tx_skb->data[1] &= 0x0f; @@ -1309,7 +1303,7 @@ static void irlap_recv_test_frame(struct irlap_cb *self, struct sk_buff *skb, * Jean II */ int irlap_driver_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *ptype) + struct packet_type *ptype, struct net_device *orig_dev) { struct irlap_info info; struct irlap_cb *self; diff --git a/net/irda/irlmp.c b/net/irda/irlmp.c index 7a4a4d7fbe66..c19e9ce05a3a 100644 --- a/net/irda/irlmp.c +++ b/net/irda/irlmp.c @@ -53,7 +53,6 @@ struct irlmp_cb *irlmp = NULL; /* These can be altered by the sysctl interface */ int sysctl_discovery = 0; int sysctl_discovery_timeout = 3; /* 3 seconds by default */ -EXPORT_SYMBOL(sysctl_discovery_timeout); int sysctl_discovery_slots = 6; /* 6 slots by default */ int sysctl_lap_keepalive_time = LM_IDLE_TIMEOUT * 1000 / HZ; char sysctl_devname[65]; @@ -67,7 +66,6 @@ const char *irlmp_reasons[] = { "LM_INIT_DISCONNECT", "ERROR, NOT USED", }; -EXPORT_SYMBOL(irlmp_reasons); /* * Function irlmp_init (void) @@ -675,7 +673,6 @@ struct lsap_cb *irlmp_dup(struct lsap_cb *orig, void *instance) return new; } -EXPORT_SYMBOL(irlmp_dup); /* * Function irlmp_disconnect_request (handle, userdata) diff --git a/net/irda/irmod.c b/net/irda/irmod.c index 6ffaed4544e9..634901dd156f 100644 --- a/net/irda/irmod.c +++ b/net/irda/irmod.c @@ -54,7 +54,7 @@ extern int irsock_init(void); extern void irsock_cleanup(void); /* irlap_frame.c */ extern int irlap_driver_rcv(struct sk_buff *, struct net_device *, - struct packet_type *); + struct packet_type *, struct net_device *); /* * Module parameters diff --git a/net/irda/irnet/irnet.h b/net/irda/irnet/irnet.h index 9004f7349a76..b391cb3893d4 100644 --- a/net/irda/irnet/irnet.h +++ b/net/irda/irnet/irnet.h @@ -517,9 +517,6 @@ extern int irda_irnet_init(void); /* Initialise IrDA part of IrNET */ extern void irda_irnet_cleanup(void); /* Teardown IrDA part of IrNET */ -/* ---------------------------- MODULE ---------------------------- */ -extern int - irnet_init(void); /* Initialise IrNET module */ /**************************** VARIABLES ****************************/ diff --git a/net/irda/irnet/irnet_ppp.c b/net/irda/irnet/irnet_ppp.c index f8f984bb9922..e53bf9e0053e 100644 --- a/net/irda/irnet/irnet_ppp.c +++ b/net/irda/irnet/irnet_ppp.c @@ -1107,7 +1107,7 @@ ppp_irnet_cleanup(void) /* * Module main entry point */ -int __init +static int __init irnet_init(void) { int err; diff --git a/net/irda/irqueue.c b/net/irda/irqueue.c index b0dd3ea35999..1ba8c7106639 100644 --- a/net/irda/irqueue.c +++ b/net/irda/irqueue.c @@ -822,7 +822,6 @@ void* hashbin_find_next( hashbin_t* hashbin, long hashv, const char* name, return entry; } -EXPORT_SYMBOL(hashbin_find_next); /* * Function hashbin_get_first (hashbin) diff --git a/net/lapb/lapb_subr.c b/net/lapb/lapb_subr.c index 5de05a0bc0ff..8b5eefd70f03 100644 --- a/net/lapb/lapb_subr.c +++ b/net/lapb/lapb_subr.c @@ -78,7 +78,7 @@ void lapb_requeue_frames(struct lapb_cb *lapb) if (!skb_prev) skb_queue_head(&lapb->write_queue, skb); else - skb_append(skb_prev, skb); + skb_append(skb_prev, skb, &lapb->write_queue); skb_prev = skb; } } diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c index 20b4cfebd74c..66f55e514b56 100644 --- a/net/llc/af_llc.c +++ b/net/llc/af_llc.c @@ -23,13 +23,13 @@ #include <linux/config.h> #include <linux/kernel.h> #include <linux/module.h> -#include <linux/tcp.h> #include <linux/rtnetlink.h> #include <linux/init.h> #include <net/llc.h> #include <net/llc_sap.h> #include <net/llc_pdu.h> #include <net/llc_conn.h> +#include <net/tcp_states.h> /* remember: uninitialized global data is zeroed because its in .bss */ static u16 llc_ui_sap_last_autoport = LLC_SAP_DYN_START; @@ -714,7 +714,7 @@ static int llc_ui_recvmsg(struct kiocb *iocb, struct socket *sock, if (uaddr) memcpy(uaddr, llc_ui_skb_cb(skb), sizeof(*uaddr)); msg->msg_namelen = sizeof(*uaddr); - if (!skb->list) { + if (!skb->next) { dgram_free: kfree_skb(skb); } diff --git a/net/llc/llc_conn.c b/net/llc/llc_conn.c index eba812a9c69c..4c644bc70eae 100644 --- a/net/llc/llc_conn.c +++ b/net/llc/llc_conn.c @@ -16,7 +16,7 @@ #include <net/llc_sap.h> #include <net/llc_conn.h> #include <net/sock.h> -#include <linux/tcp.h> +#include <net/tcp_states.h> #include <net/llc_c_ev.h> #include <net/llc_c_ac.h> #include <net/llc_c_st.h> @@ -71,7 +71,11 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb) if (!ev->ind_prim && !ev->cfm_prim) { /* indicate or confirm not required */ - if (!skb->list) + /* XXX this is not very pretty, perhaps we should store + * XXX indicate/confirm-needed state in the llc_conn_state_ev + * XXX control block of the SKB instead? -DaveM + */ + if (!skb->next) goto out_kfree_skb; goto out_skb_put; } diff --git a/net/llc/llc_core.c b/net/llc/llc_core.c index 5ff02c080a0b..9727455bf0e7 100644 --- a/net/llc/llc_core.c +++ b/net/llc/llc_core.c @@ -103,7 +103,8 @@ out: struct llc_sap *llc_sap_open(unsigned char lsap, int (*func)(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt)) + struct packet_type *pt, + struct net_device *orig_dev)) { struct llc_sap *sap = llc_sap_find(lsap); diff --git a/net/llc/llc_if.c b/net/llc/llc_if.c index 0f9fc48aeaf9..0f84f66018e4 100644 --- a/net/llc/llc_if.c +++ b/net/llc/llc_if.c @@ -15,7 +15,6 @@ #include <linux/module.h> #include <linux/kernel.h> #include <linux/netdevice.h> -#include <linux/tcp.h> #include <asm/errno.h> #include <net/llc_if.h> #include <net/llc_sap.h> @@ -25,6 +24,7 @@ #include <net/llc_c_ev.h> #include <net/llc_c_ac.h> #include <net/llc_c_st.h> +#include <net/tcp_states.h> u8 llc_mac_null_var[IFHWADDRLEN]; diff --git a/net/llc/llc_input.c b/net/llc/llc_input.c index 4da6976efc9c..13b46240b7a1 100644 --- a/net/llc/llc_input.c +++ b/net/llc/llc_input.c @@ -132,7 +132,7 @@ static inline int llc_fixup_skb(struct sk_buff *skb) * data now), it queues this frame in the connection's backlog. */ int llc_rcv(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt) + struct packet_type *pt, struct net_device *orig_dev) { struct llc_sap *sap; struct llc_pdu_sn *pdu; @@ -165,7 +165,7 @@ int llc_rcv(struct sk_buff *skb, struct net_device *dev, * LLC functionality */ if (sap->rcv_func) { - sap->rcv_func(skb, dev, pt); + sap->rcv_func(skb, dev, pt, orig_dev); goto out; } dest = llc_pdu_type(skb); diff --git a/net/llc/llc_sap.c b/net/llc/llc_sap.c index 965c94eb4bbc..34228ef14985 100644 --- a/net/llc/llc_sap.c +++ b/net/llc/llc_sap.c @@ -21,7 +21,7 @@ #include <net/llc_s_ev.h> #include <net/llc_s_st.h> #include <net/sock.h> -#include <linux/tcp.h> +#include <net/tcp_states.h> #include <linux/llc.h> /** diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig new file mode 100644 index 000000000000..8296b38bf270 --- /dev/null +++ b/net/netfilter/Kconfig @@ -0,0 +1,24 @@ +config NETFILTER_NETLINK + tristate "Netfilter netlink interface" + help + If this option is enabled, the kernel will include support + for the new netfilter netlink interface. + +config NETFILTER_NETLINK_QUEUE + tristate "Netfilter NFQUEUE over NFNETLINK interface" + depends on NETFILTER_NETLINK + help + If this option isenabled, the kernel will include support + for queueing packets via NFNETLINK. + +config NETFILTER_NETLINK_LOG + tristate "Netfilter LOG over NFNETLINK interface" + depends on NETFILTER_NETLINK + help + If this option is enabled, the kernel will include support + for logging packets via NFNETLINK. + + This obsoletes the existing ipt_ULOG and ebg_ulog mechanisms, + and is also scheduled to replace the old syslog-based ipt_LOG + and ip6t_LOG modules. + diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile new file mode 100644 index 000000000000..b3b44f8b415a --- /dev/null +++ b/net/netfilter/Makefile @@ -0,0 +1,7 @@ +netfilter-objs := core.o nf_log.o nf_queue.o nf_sockopt.o + +obj-$(CONFIG_NETFILTER) = netfilter.o + +obj-$(CONFIG_NETFILTER_NETLINK) += nfnetlink.o +obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += nfnetlink_queue.o +obj-$(CONFIG_NETFILTER_NETLINK_LOG) += nfnetlink_log.o diff --git a/net/netfilter/core.c b/net/netfilter/core.c new file mode 100644 index 000000000000..1ceb1a6c254b --- /dev/null +++ b/net/netfilter/core.c @@ -0,0 +1,216 @@ +/* netfilter.c: look after the filters for various protocols. + * Heavily influenced by the old firewall.c by David Bonn and Alan Cox. + * + * Thanks to Rob `CmdrTaco' Malda for not influencing this code in any + * way. + * + * Rusty Russell (C)2000 -- This code is GPL. + * + * February 2000: Modified by James Morris to have 1 queue per protocol. + * 15-Mar-2000: Added NF_REPEAT --RR. + * 08-May-2003: Internal logging interface added by Jozsef Kadlecsik. + */ +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/netfilter.h> +#include <net/protocol.h> +#include <linux/init.h> +#include <linux/skbuff.h> +#include <linux/wait.h> +#include <linux/module.h> +#include <linux/interrupt.h> +#include <linux/if.h> +#include <linux/netdevice.h> +#include <linux/inetdevice.h> +#include <linux/proc_fs.h> +#include <net/sock.h> + +#include "nf_internals.h" + +/* In this code, we can be waiting indefinitely for userspace to + * service a packet if a hook returns NF_QUEUE. We could keep a count + * of skbuffs queued for userspace, and not deregister a hook unless + * this is zero, but that sucks. Now, we simply check when the + * packets come back: if the hook is gone, the packet is discarded. */ +struct list_head nf_hooks[NPROTO][NF_MAX_HOOKS]; +EXPORT_SYMBOL(nf_hooks); +static DEFINE_SPINLOCK(nf_hook_lock); + +int nf_register_hook(struct nf_hook_ops *reg) +{ + struct list_head *i; + + spin_lock_bh(&nf_hook_lock); + list_for_each(i, &nf_hooks[reg->pf][reg->hooknum]) { + if (reg->priority < ((struct nf_hook_ops *)i)->priority) + break; + } + list_add_rcu(®->list, i->prev); + spin_unlock_bh(&nf_hook_lock); + + synchronize_net(); + return 0; +} +EXPORT_SYMBOL(nf_register_hook); + +void nf_unregister_hook(struct nf_hook_ops *reg) +{ + spin_lock_bh(&nf_hook_lock); + list_del_rcu(®->list); + spin_unlock_bh(&nf_hook_lock); + + synchronize_net(); +} +EXPORT_SYMBOL(nf_unregister_hook); + +unsigned int nf_iterate(struct list_head *head, + struct sk_buff **skb, + int hook, + const struct net_device *indev, + const struct net_device *outdev, + struct list_head **i, + int (*okfn)(struct sk_buff *), + int hook_thresh) +{ + unsigned int verdict; + + /* + * The caller must not block between calls to this + * function because of risk of continuing from deleted element. + */ + list_for_each_continue_rcu(*i, head) { + struct nf_hook_ops *elem = (struct nf_hook_ops *)*i; + + if (hook_thresh > elem->priority) + continue; + + /* Optimization: we don't need to hold module + reference here, since function can't sleep. --RR */ + verdict = elem->hook(hook, skb, indev, outdev, okfn); + if (verdict != NF_ACCEPT) { +#ifdef CONFIG_NETFILTER_DEBUG + if (unlikely((verdict & NF_VERDICT_MASK) + > NF_MAX_VERDICT)) { + NFDEBUG("Evil return from %p(%u).\n", + elem->hook, hook); + continue; + } +#endif + if (verdict != NF_REPEAT) + return verdict; + *i = (*i)->prev; + } + } + return NF_ACCEPT; +} + + +/* Returns 1 if okfn() needs to be executed by the caller, + * -EPERM for NF_DROP, 0 otherwise. */ +int nf_hook_slow(int pf, unsigned int hook, struct sk_buff **pskb, + struct net_device *indev, + struct net_device *outdev, + int (*okfn)(struct sk_buff *), + int hook_thresh) +{ + struct list_head *elem; + unsigned int verdict; + int ret = 0; + + /* We may already have this, but read-locks nest anyway */ + rcu_read_lock(); + + elem = &nf_hooks[pf][hook]; +next_hook: + verdict = nf_iterate(&nf_hooks[pf][hook], pskb, hook, indev, + outdev, &elem, okfn, hook_thresh); + if (verdict == NF_ACCEPT || verdict == NF_STOP) { + ret = 1; + goto unlock; + } else if (verdict == NF_DROP) { + kfree_skb(*pskb); + ret = -EPERM; + } else if ((verdict & NF_VERDICT_MASK) == NF_QUEUE) { + NFDEBUG("nf_hook: Verdict = QUEUE.\n"); + if (!nf_queue(pskb, elem, pf, hook, indev, outdev, okfn, + verdict >> NF_VERDICT_BITS)) + goto next_hook; + } +unlock: + rcu_read_unlock(); + return ret; +} +EXPORT_SYMBOL(nf_hook_slow); + + +int skb_make_writable(struct sk_buff **pskb, unsigned int writable_len) +{ + struct sk_buff *nskb; + + if (writable_len > (*pskb)->len) + return 0; + + /* Not exclusive use of packet? Must copy. */ + if (skb_shared(*pskb) || skb_cloned(*pskb)) + goto copy_skb; + + return pskb_may_pull(*pskb, writable_len); + +copy_skb: + nskb = skb_copy(*pskb, GFP_ATOMIC); + if (!nskb) + return 0; + BUG_ON(skb_is_nonlinear(nskb)); + + /* Rest of kernel will get very unhappy if we pass it a + suddenly-orphaned skbuff */ + if ((*pskb)->sk) + skb_set_owner_w(nskb, (*pskb)->sk); + kfree_skb(*pskb); + *pskb = nskb; + return 1; +} +EXPORT_SYMBOL(skb_make_writable); + + +/* This does not belong here, but locally generated errors need it if connection + tracking in use: without this, connection may not be in hash table, and hence + manufactured ICMP or RST packets will not be associated with it. */ +void (*ip_ct_attach)(struct sk_buff *, struct sk_buff *); +EXPORT_SYMBOL(ip_ct_attach); + +void nf_ct_attach(struct sk_buff *new, struct sk_buff *skb) +{ + void (*attach)(struct sk_buff *, struct sk_buff *); + + if (skb->nfct && (attach = ip_ct_attach) != NULL) { + mb(); /* Just to be sure: must be read before executing this */ + attach(new, skb); + } +} +EXPORT_SYMBOL(nf_ct_attach); + +#ifdef CONFIG_PROC_FS +struct proc_dir_entry *proc_net_netfilter; +EXPORT_SYMBOL(proc_net_netfilter); +#endif + +void __init netfilter_init(void) +{ + int i, h; + for (i = 0; i < NPROTO; i++) { + for (h = 0; h < NF_MAX_HOOKS; h++) + INIT_LIST_HEAD(&nf_hooks[i][h]); + } + +#ifdef CONFIG_PROC_FS + proc_net_netfilter = proc_mkdir("netfilter", proc_net); + if (!proc_net_netfilter) + panic("cannot create netfilter proc entry"); +#endif + + if (netfilter_queue_init() < 0) + panic("cannot initialize nf_queue"); + if (netfilter_log_init() < 0) + panic("cannot initialize nf_log"); +} diff --git a/net/netfilter/nf_internals.h b/net/netfilter/nf_internals.h new file mode 100644 index 000000000000..6bdee2910617 --- /dev/null +++ b/net/netfilter/nf_internals.h @@ -0,0 +1,39 @@ +#ifndef _NF_INTERNALS_H +#define _NF_INTERNALS_H + +#include <linux/config.h> +#include <linux/list.h> +#include <linux/skbuff.h> +#include <linux/netdevice.h> + +#ifdef CONFIG_NETFILTER_DEBUG +#define NFDEBUG(format, args...) printk(format , ## args) +#else +#define NFDEBUG(format, args...) +#endif + + +/* core.c */ +extern unsigned int nf_iterate(struct list_head *head, + struct sk_buff **skb, + int hook, + const struct net_device *indev, + const struct net_device *outdev, + struct list_head **i, + int (*okfn)(struct sk_buff *), + int hook_thresh); + +/* nf_queue.c */ +extern int nf_queue(struct sk_buff **skb, + struct list_head *elem, + int pf, unsigned int hook, + struct net_device *indev, + struct net_device *outdev, + int (*okfn)(struct sk_buff *), + unsigned int queuenum); +extern int __init netfilter_queue_init(void); + +/* nf_log.c */ +extern int __init netfilter_log_init(void); + +#endif diff --git a/net/netfilter/nf_log.c b/net/netfilter/nf_log.c new file mode 100644 index 000000000000..3e76bd0824a2 --- /dev/null +++ b/net/netfilter/nf_log.c @@ -0,0 +1,178 @@ +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/proc_fs.h> +#include <linux/skbuff.h> +#include <linux/netfilter.h> +#include <linux/seq_file.h> +#include <net/protocol.h> + +#include "nf_internals.h" + +/* Internal logging interface, which relies on the real + LOG target modules */ + +#define NF_LOG_PREFIXLEN 128 + +static struct nf_logger *nf_logging[NPROTO]; /* = NULL */ +static DEFINE_SPINLOCK(nf_log_lock); + +/* return EBUSY if somebody else is registered, EEXIST if the same logger + * is registred, 0 on success. */ +int nf_log_register(int pf, struct nf_logger *logger) +{ + int ret = -EBUSY; + + if (pf >= NPROTO) + return -EINVAL; + + /* Any setup of logging members must be done before + * substituting pointer. */ + spin_lock(&nf_log_lock); + if (!nf_logging[pf]) { + rcu_assign_pointer(nf_logging[pf], logger); + ret = 0; + } else if (nf_logging[pf] == logger) + ret = -EEXIST; + + spin_unlock(&nf_log_lock); + return ret; +} +EXPORT_SYMBOL(nf_log_register); + +int nf_log_unregister_pf(int pf) +{ + if (pf >= NPROTO) + return -EINVAL; + + spin_lock(&nf_log_lock); + nf_logging[pf] = NULL; + spin_unlock(&nf_log_lock); + + /* Give time to concurrent readers. */ + synchronize_net(); + + return 0; +} +EXPORT_SYMBOL(nf_log_unregister_pf); + +void nf_log_unregister_logger(struct nf_logger *logger) +{ + int i; + + spin_lock(&nf_log_lock); + for (i = 0; i < NPROTO; i++) { + if (nf_logging[i] == logger) + nf_logging[i] = NULL; + } + spin_unlock(&nf_log_lock); + + synchronize_net(); +} +EXPORT_SYMBOL(nf_log_unregister_logger); + +void nf_log_packet(int pf, + unsigned int hooknum, + const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + struct nf_loginfo *loginfo, + const char *fmt, ...) +{ + va_list args; + char prefix[NF_LOG_PREFIXLEN]; + struct nf_logger *logger; + + rcu_read_lock(); + logger = rcu_dereference(nf_logging[pf]); + if (logger) { + va_start(args, fmt); + vsnprintf(prefix, sizeof(prefix), fmt, args); + va_end(args); + /* We must read logging before nf_logfn[pf] */ + logger->logfn(pf, hooknum, skb, in, out, loginfo, prefix); + } else if (net_ratelimit()) { + printk(KERN_WARNING "nf_log_packet: can\'t log since " + "no backend logging module loaded in! Please either " + "load one, or disable logging explicitly\n"); + } + rcu_read_unlock(); +} +EXPORT_SYMBOL(nf_log_packet); + +#ifdef CONFIG_PROC_FS +static void *seq_start(struct seq_file *seq, loff_t *pos) +{ + rcu_read_lock(); + + if (*pos >= NPROTO) + return NULL; + + return pos; +} + +static void *seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + (*pos)++; + + if (*pos >= NPROTO) + return NULL; + + return pos; +} + +static void seq_stop(struct seq_file *s, void *v) +{ + rcu_read_unlock(); +} + +static int seq_show(struct seq_file *s, void *v) +{ + loff_t *pos = v; + const struct nf_logger *logger; + + logger = rcu_dereference(nf_logging[*pos]); + + if (!logger) + return seq_printf(s, "%2lld NONE\n", *pos); + + return seq_printf(s, "%2lld %s\n", *pos, logger->name); +} + +static struct seq_operations nflog_seq_ops = { + .start = seq_start, + .next = seq_next, + .stop = seq_stop, + .show = seq_show, +}; + +static int nflog_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &nflog_seq_ops); +} + +static struct file_operations nflog_file_ops = { + .owner = THIS_MODULE, + .open = nflog_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +#endif /* PROC_FS */ + + +int __init netfilter_log_init(void) +{ +#ifdef CONFIG_PROC_FS + struct proc_dir_entry *pde; + + pde = create_proc_entry("nf_log", S_IRUGO, proc_net_netfilter); + if (!pde) + return -1; + + pde->proc_fops = &nflog_file_ops; +#endif + return 0; +} diff --git a/net/netfilter/nf_queue.c b/net/netfilter/nf_queue.c new file mode 100644 index 000000000000..d10d552d9c40 --- /dev/null +++ b/net/netfilter/nf_queue.c @@ -0,0 +1,343 @@ +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/proc_fs.h> +#include <linux/skbuff.h> +#include <linux/netfilter.h> +#include <linux/seq_file.h> +#include <net/protocol.h> + +#include "nf_internals.h" + +/* + * A queue handler may be registered for each protocol. Each is protected by + * long term mutex. The handler must provide an an outfn() to accept packets + * for queueing and must reinject all packets it receives, no matter what. + */ +static struct nf_queue_handler *queue_handler[NPROTO]; +static struct nf_queue_rerouter *queue_rerouter; + +static DEFINE_RWLOCK(queue_handler_lock); + +/* return EBUSY when somebody else is registered, return EEXIST if the + * same handler is registered, return 0 in case of success. */ +int nf_register_queue_handler(int pf, struct nf_queue_handler *qh) +{ + int ret; + + if (pf >= NPROTO) + return -EINVAL; + + write_lock_bh(&queue_handler_lock); + if (queue_handler[pf] == qh) + ret = -EEXIST; + else if (queue_handler[pf]) + ret = -EBUSY; + else { + queue_handler[pf] = qh; + ret = 0; + } + write_unlock_bh(&queue_handler_lock); + + return ret; +} +EXPORT_SYMBOL(nf_register_queue_handler); + +/* The caller must flush their queue before this */ +int nf_unregister_queue_handler(int pf) +{ + if (pf >= NPROTO) + return -EINVAL; + + write_lock_bh(&queue_handler_lock); + queue_handler[pf] = NULL; + write_unlock_bh(&queue_handler_lock); + + return 0; +} +EXPORT_SYMBOL(nf_unregister_queue_handler); + +int nf_register_queue_rerouter(int pf, struct nf_queue_rerouter *rer) +{ + if (pf >= NPROTO) + return -EINVAL; + + write_lock_bh(&queue_handler_lock); + memcpy(&queue_rerouter[pf], rer, sizeof(queue_rerouter[pf])); + write_unlock_bh(&queue_handler_lock); + + return 0; +} +EXPORT_SYMBOL_GPL(nf_register_queue_rerouter); + +int nf_unregister_queue_rerouter(int pf) +{ + if (pf >= NPROTO) + return -EINVAL; + + write_lock_bh(&queue_handler_lock); + memset(&queue_rerouter[pf], 0, sizeof(queue_rerouter[pf])); + write_unlock_bh(&queue_handler_lock); + return 0; +} +EXPORT_SYMBOL_GPL(nf_unregister_queue_rerouter); + +void nf_unregister_queue_handlers(struct nf_queue_handler *qh) +{ + int pf; + + write_lock_bh(&queue_handler_lock); + for (pf = 0; pf < NPROTO; pf++) { + if (queue_handler[pf] == qh) + queue_handler[pf] = NULL; + } + write_unlock_bh(&queue_handler_lock); +} +EXPORT_SYMBOL_GPL(nf_unregister_queue_handlers); + +/* + * Any packet that leaves via this function must come back + * through nf_reinject(). + */ +int nf_queue(struct sk_buff **skb, + struct list_head *elem, + int pf, unsigned int hook, + struct net_device *indev, + struct net_device *outdev, + int (*okfn)(struct sk_buff *), + unsigned int queuenum) +{ + int status; + struct nf_info *info; +#ifdef CONFIG_BRIDGE_NETFILTER + struct net_device *physindev = NULL; + struct net_device *physoutdev = NULL; +#endif + + /* QUEUE == DROP if noone is waiting, to be safe. */ + read_lock(&queue_handler_lock); + if (!queue_handler[pf]->outfn) { + read_unlock(&queue_handler_lock); + kfree_skb(*skb); + return 1; + } + + info = kmalloc(sizeof(*info)+queue_rerouter[pf].rer_size, GFP_ATOMIC); + if (!info) { + if (net_ratelimit()) + printk(KERN_ERR "OOM queueing packet %p\n", + *skb); + read_unlock(&queue_handler_lock); + kfree_skb(*skb); + return 1; + } + + *info = (struct nf_info) { + (struct nf_hook_ops *)elem, pf, hook, indev, outdev, okfn }; + + /* If it's going away, ignore hook. */ + if (!try_module_get(info->elem->owner)) { + read_unlock(&queue_handler_lock); + kfree(info); + return 0; + } + + /* Bump dev refs so they don't vanish while packet is out */ + if (indev) dev_hold(indev); + if (outdev) dev_hold(outdev); + +#ifdef CONFIG_BRIDGE_NETFILTER + if ((*skb)->nf_bridge) { + physindev = (*skb)->nf_bridge->physindev; + if (physindev) dev_hold(physindev); + physoutdev = (*skb)->nf_bridge->physoutdev; + if (physoutdev) dev_hold(physoutdev); + } +#endif + if (queue_rerouter[pf].save) + queue_rerouter[pf].save(*skb, info); + + status = queue_handler[pf]->outfn(*skb, info, queuenum, + queue_handler[pf]->data); + + if (status >= 0 && queue_rerouter[pf].reroute) + status = queue_rerouter[pf].reroute(skb, info); + + read_unlock(&queue_handler_lock); + + if (status < 0) { + /* James M doesn't say fuck enough. */ + if (indev) dev_put(indev); + if (outdev) dev_put(outdev); +#ifdef CONFIG_BRIDGE_NETFILTER + if (physindev) dev_put(physindev); + if (physoutdev) dev_put(physoutdev); +#endif + module_put(info->elem->owner); + kfree(info); + kfree_skb(*skb); + + return 1; + } + + return 1; +} + +void nf_reinject(struct sk_buff *skb, struct nf_info *info, + unsigned int verdict) +{ + struct list_head *elem = &info->elem->list; + struct list_head *i; + + rcu_read_lock(); + + /* Release those devices we held, or Alexey will kill me. */ + if (info->indev) dev_put(info->indev); + if (info->outdev) dev_put(info->outdev); +#ifdef CONFIG_BRIDGE_NETFILTER + if (skb->nf_bridge) { + if (skb->nf_bridge->physindev) + dev_put(skb->nf_bridge->physindev); + if (skb->nf_bridge->physoutdev) + dev_put(skb->nf_bridge->physoutdev); + } +#endif + + /* Drop reference to owner of hook which queued us. */ + module_put(info->elem->owner); + + list_for_each_rcu(i, &nf_hooks[info->pf][info->hook]) { + if (i == elem) + break; + } + + if (elem == &nf_hooks[info->pf][info->hook]) { + /* The module which sent it to userspace is gone. */ + NFDEBUG("%s: module disappeared, dropping packet.\n", + __FUNCTION__); + verdict = NF_DROP; + } + + /* Continue traversal iff userspace said ok... */ + if (verdict == NF_REPEAT) { + elem = elem->prev; + verdict = NF_ACCEPT; + } + + if (verdict == NF_ACCEPT) { + next_hook: + verdict = nf_iterate(&nf_hooks[info->pf][info->hook], + &skb, info->hook, + info->indev, info->outdev, &elem, + info->okfn, INT_MIN); + } + + switch (verdict & NF_VERDICT_MASK) { + case NF_ACCEPT: + info->okfn(skb); + break; + + case NF_QUEUE: + if (!nf_queue(&skb, elem, info->pf, info->hook, + info->indev, info->outdev, info->okfn, + verdict >> NF_VERDICT_BITS)) + goto next_hook; + break; + } + rcu_read_unlock(); + + if (verdict == NF_DROP) + kfree_skb(skb); + + kfree(info); + return; +} +EXPORT_SYMBOL(nf_reinject); + +#ifdef CONFIG_PROC_FS +static void *seq_start(struct seq_file *seq, loff_t *pos) +{ + if (*pos >= NPROTO) + return NULL; + + return pos; +} + +static void *seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + (*pos)++; + + if (*pos >= NPROTO) + return NULL; + + return pos; +} + +static void seq_stop(struct seq_file *s, void *v) +{ + +} + +static int seq_show(struct seq_file *s, void *v) +{ + int ret; + loff_t *pos = v; + struct nf_queue_handler *qh; + + read_lock_bh(&queue_handler_lock); + qh = queue_handler[*pos]; + if (!qh) + ret = seq_printf(s, "%2lld NONE\n", *pos); + else + ret = seq_printf(s, "%2lld %s\n", *pos, qh->name); + read_unlock_bh(&queue_handler_lock); + + return ret; +} + +static struct seq_operations nfqueue_seq_ops = { + .start = seq_start, + .next = seq_next, + .stop = seq_stop, + .show = seq_show, +}; + +static int nfqueue_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &nfqueue_seq_ops); +} + +static struct file_operations nfqueue_file_ops = { + .owner = THIS_MODULE, + .open = nfqueue_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; +#endif /* PROC_FS */ + + +int __init netfilter_queue_init(void) +{ +#ifdef CONFIG_PROC_FS + struct proc_dir_entry *pde; +#endif + queue_rerouter = kmalloc(NPROTO * sizeof(struct nf_queue_rerouter), + GFP_KERNEL); + if (!queue_rerouter) + return -ENOMEM; + +#ifdef CONFIG_PROC_FS + pde = create_proc_entry("nf_queue", S_IRUGO, proc_net_netfilter); + if (!pde) { + kfree(queue_rerouter); + return -1; + } + pde->proc_fops = &nfqueue_file_ops; +#endif + memset(queue_rerouter, 0, NPROTO * sizeof(struct nf_queue_rerouter)); + + return 0; +} + diff --git a/net/netfilter/nf_sockopt.c b/net/netfilter/nf_sockopt.c new file mode 100644 index 000000000000..61a833a9caa6 --- /dev/null +++ b/net/netfilter/nf_sockopt.c @@ -0,0 +1,132 @@ +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/netfilter.h> +#include <net/sock.h> + +#include "nf_internals.h" + +/* Sockopts only registered and called from user context, so + net locking would be overkill. Also, [gs]etsockopt calls may + sleep. */ +static DECLARE_MUTEX(nf_sockopt_mutex); +static LIST_HEAD(nf_sockopts); + +/* Do exclusive ranges overlap? */ +static inline int overlap(int min1, int max1, int min2, int max2) +{ + return max1 > min2 && min1 < max2; +} + +/* Functions to register sockopt ranges (exclusive). */ +int nf_register_sockopt(struct nf_sockopt_ops *reg) +{ + struct list_head *i; + int ret = 0; + + if (down_interruptible(&nf_sockopt_mutex) != 0) + return -EINTR; + + list_for_each(i, &nf_sockopts) { + struct nf_sockopt_ops *ops = (struct nf_sockopt_ops *)i; + if (ops->pf == reg->pf + && (overlap(ops->set_optmin, ops->set_optmax, + reg->set_optmin, reg->set_optmax) + || overlap(ops->get_optmin, ops->get_optmax, + reg->get_optmin, reg->get_optmax))) { + NFDEBUG("nf_sock overlap: %u-%u/%u-%u v %u-%u/%u-%u\n", + ops->set_optmin, ops->set_optmax, + ops->get_optmin, ops->get_optmax, + reg->set_optmin, reg->set_optmax, + reg->get_optmin, reg->get_optmax); + ret = -EBUSY; + goto out; + } + } + + list_add(®->list, &nf_sockopts); +out: + up(&nf_sockopt_mutex); + return ret; +} +EXPORT_SYMBOL(nf_register_sockopt); + +void nf_unregister_sockopt(struct nf_sockopt_ops *reg) +{ + /* No point being interruptible: we're probably in cleanup_module() */ + restart: + down(&nf_sockopt_mutex); + if (reg->use != 0) { + /* To be woken by nf_sockopt call... */ + /* FIXME: Stuart Young's name appears gratuitously. */ + set_current_state(TASK_UNINTERRUPTIBLE); + reg->cleanup_task = current; + up(&nf_sockopt_mutex); + schedule(); + goto restart; + } + list_del(®->list); + up(&nf_sockopt_mutex); +} +EXPORT_SYMBOL(nf_unregister_sockopt); + +/* Call get/setsockopt() */ +static int nf_sockopt(struct sock *sk, int pf, int val, + char __user *opt, int *len, int get) +{ + struct list_head *i; + struct nf_sockopt_ops *ops; + int ret; + + if (down_interruptible(&nf_sockopt_mutex) != 0) + return -EINTR; + + list_for_each(i, &nf_sockopts) { + ops = (struct nf_sockopt_ops *)i; + if (ops->pf == pf) { + if (get) { + if (val >= ops->get_optmin + && val < ops->get_optmax) { + ops->use++; + up(&nf_sockopt_mutex); + ret = ops->get(sk, val, opt, len); + goto out; + } + } else { + if (val >= ops->set_optmin + && val < ops->set_optmax) { + ops->use++; + up(&nf_sockopt_mutex); + ret = ops->set(sk, val, opt, *len); + goto out; + } + } + } + } + up(&nf_sockopt_mutex); + return -ENOPROTOOPT; + + out: + down(&nf_sockopt_mutex); + ops->use--; + if (ops->cleanup_task) + wake_up_process(ops->cleanup_task); + up(&nf_sockopt_mutex); + return ret; +} + +int nf_setsockopt(struct sock *sk, int pf, int val, char __user *opt, + int len) +{ + return nf_sockopt(sk, pf, val, opt, &len, 0); +} +EXPORT_SYMBOL(nf_setsockopt); + +int nf_getsockopt(struct sock *sk, int pf, int val, char __user *opt, int *len) +{ + return nf_sockopt(sk, pf, val, opt, len, 1); +} +EXPORT_SYMBOL(nf_getsockopt); + diff --git a/net/netfilter/nfnetlink.c b/net/netfilter/nfnetlink.c new file mode 100644 index 000000000000..e089f17bb803 --- /dev/null +++ b/net/netfilter/nfnetlink.c @@ -0,0 +1,376 @@ +/* Netfilter messages via netlink socket. Allows for user space + * protocol helpers and general trouble making from userspace. + * + * (C) 2001 by Jay Schulist <jschlst@samba.org>, + * (C) 2002-2005 by Harald Welte <laforge@gnumonks.org> + * (C) 2005 by Pablo Neira Ayuso <pablo@eurodev.net> + * + * Initial netfilter messages via netlink development funded and + * generally made possible by Network Robots, Inc. (www.networkrobots.com) + * + * Further development of this code funded by Astaro AG (http://www.astaro.com) + * + * This software may be used and distributed according to the terms + * of the GNU General Public License, incorporated herein by reference. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/types.h> +#include <linux/socket.h> +#include <linux/kernel.h> +#include <linux/major.h> +#include <linux/sched.h> +#include <linux/timer.h> +#include <linux/string.h> +#include <linux/sockios.h> +#include <linux/net.h> +#include <linux/fcntl.h> +#include <linux/skbuff.h> +#include <asm/uaccess.h> +#include <asm/system.h> +#include <net/sock.h> +#include <linux/init.h> +#include <linux/spinlock.h> + +#include <linux/netfilter.h> +#include <linux/netlink.h> +#include <linux/netfilter/nfnetlink.h> + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER); + +static char __initdata nfversion[] = "0.30"; + +#if 0 +#define DEBUGP(format, args...) \ + printk(KERN_DEBUG "%s(%d):%s(): " format, __FILE__, \ + __LINE__, __FUNCTION__, ## args) +#else +#define DEBUGP(format, args...) +#endif + +static struct sock *nfnl = NULL; +static struct nfnetlink_subsystem *subsys_table[NFNL_SUBSYS_COUNT]; +DECLARE_MUTEX(nfnl_sem); + +void nfnl_lock(void) +{ + nfnl_shlock(); +} + +void nfnl_unlock(void) +{ + nfnl_shunlock(); +} + +int nfnetlink_subsys_register(struct nfnetlink_subsystem *n) +{ + DEBUGP("registering subsystem ID %u\n", n->subsys_id); + + nfnl_lock(); + if (subsys_table[n->subsys_id]) { + nfnl_unlock(); + return -EBUSY; + } + subsys_table[n->subsys_id] = n; + nfnl_unlock(); + + return 0; +} + +int nfnetlink_subsys_unregister(struct nfnetlink_subsystem *n) +{ + DEBUGP("unregistering subsystem ID %u\n", n->subsys_id); + + nfnl_lock(); + subsys_table[n->subsys_id] = NULL; + nfnl_unlock(); + + return 0; +} + +static inline struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type) +{ + u_int8_t subsys_id = NFNL_SUBSYS_ID(type); + + if (subsys_id >= NFNL_SUBSYS_COUNT + || subsys_table[subsys_id] == NULL) + return NULL; + + return subsys_table[subsys_id]; +} + +static inline struct nfnl_callback * +nfnetlink_find_client(u_int16_t type, struct nfnetlink_subsystem *ss) +{ + u_int8_t cb_id = NFNL_MSG_TYPE(type); + + if (cb_id >= ss->cb_count) { + DEBUGP("msgtype %u >= %u, returning\n", type, ss->cb_count); + return NULL; + } + + return &ss->cb[cb_id]; +} + +void __nfa_fill(struct sk_buff *skb, int attrtype, int attrlen, + const void *data) +{ + struct nfattr *nfa; + int size = NFA_LENGTH(attrlen); + + nfa = (struct nfattr *)skb_put(skb, NFA_ALIGN(size)); + nfa->nfa_type = attrtype; + nfa->nfa_len = size; + memcpy(NFA_DATA(nfa), data, attrlen); + memset(NFA_DATA(nfa) + attrlen, 0, NFA_ALIGN(size) - size); +} + +int nfattr_parse(struct nfattr *tb[], int maxattr, struct nfattr *nfa, int len) +{ + memset(tb, 0, sizeof(struct nfattr *) * maxattr); + + while (NFA_OK(nfa, len)) { + unsigned flavor = nfa->nfa_type; + if (flavor && flavor <= maxattr) + tb[flavor-1] = nfa; + nfa = NFA_NEXT(nfa, len); + } + + return 0; +} + +/** + * nfnetlink_check_attributes - check and parse nfnetlink attributes + * + * subsys: nfnl subsystem for which this message is to be parsed + * nlmsghdr: netlink message to be checked/parsed + * cda: array of pointers, needs to be at least subsys->attr_count big + * + */ +static int +nfnetlink_check_attributes(struct nfnetlink_subsystem *subsys, + struct nlmsghdr *nlh, struct nfattr *cda[]) +{ + int min_len; + u_int16_t attr_count; + u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type); + + if (unlikely(cb_id >= subsys->cb_count)) { + DEBUGP("msgtype %u >= %u, returning\n", + cb_id, subsys->cb_count); + return -EINVAL; + } + + min_len = NLMSG_ALIGN(sizeof(struct nfgenmsg)); + if (unlikely(nlh->nlmsg_len < min_len)) + return -EINVAL; + + attr_count = subsys->cb[cb_id].attr_count; + memset(cda, 0, sizeof(struct nfattr *) * attr_count); + + /* check attribute lengths. */ + if (likely(nlh->nlmsg_len > min_len)) { + struct nfattr *attr = NFM_NFA(NLMSG_DATA(nlh)); + int attrlen = nlh->nlmsg_len - NLMSG_ALIGN(min_len); + + while (NFA_OK(attr, attrlen)) { + unsigned flavor = attr->nfa_type; + if (flavor) { + if (flavor > attr_count) + return -EINVAL; + cda[flavor - 1] = attr; + } + attr = NFA_NEXT(attr, attrlen); + } + } + + /* implicit: if nlmsg_len == min_len, we return 0, and an empty + * (zeroed) cda[] array. The message is valid, but empty. */ + + return 0; +} + +int nfnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo) +{ + int allocation = in_interrupt() ? GFP_ATOMIC : GFP_KERNEL; + int err = 0; + + NETLINK_CB(skb).dst_group = group; + if (echo) + atomic_inc(&skb->users); + netlink_broadcast(nfnl, skb, pid, group, allocation); + if (echo) + err = netlink_unicast(nfnl, skb, pid, MSG_DONTWAIT); + + return err; +} + +int nfnetlink_unicast(struct sk_buff *skb, u_int32_t pid, int flags) +{ + return netlink_unicast(nfnl, skb, pid, flags); +} + +/* Process one complete nfnetlink message. */ +static inline int nfnetlink_rcv_msg(struct sk_buff *skb, + struct nlmsghdr *nlh, int *errp) +{ + struct nfnl_callback *nc; + struct nfnetlink_subsystem *ss; + int type, err = 0; + + DEBUGP("entered; subsys=%u, msgtype=%u\n", + NFNL_SUBSYS_ID(nlh->nlmsg_type), + NFNL_MSG_TYPE(nlh->nlmsg_type)); + + /* Only requests are handled by kernel now. */ + if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) { + DEBUGP("received non-request message\n"); + return 0; + } + + /* All the messages must at least contain nfgenmsg */ + if (nlh->nlmsg_len < + NLMSG_LENGTH(NLMSG_ALIGN(sizeof(struct nfgenmsg)))) { + DEBUGP("received message was too short\n"); + return 0; + } + + type = nlh->nlmsg_type; + ss = nfnetlink_get_subsys(type); + if (!ss) { +#ifdef CONFIG_KMOD + /* don't call nfnl_shunlock, since it would reenter + * with further packet processing */ + up(&nfnl_sem); + request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type)); + nfnl_shlock(); + ss = nfnetlink_get_subsys(type); + if (!ss) +#endif + goto err_inval; + } + + nc = nfnetlink_find_client(type, ss); + if (!nc) { + DEBUGP("unable to find client for type %d\n", type); + goto err_inval; + } + + if (nc->cap_required && + !cap_raised(NETLINK_CB(skb).eff_cap, nc->cap_required)) { + DEBUGP("permission denied for type %d\n", type); + *errp = -EPERM; + return -1; + } + + { + u_int16_t attr_count = + ss->cb[NFNL_MSG_TYPE(nlh->nlmsg_type)].attr_count; + struct nfattr *cda[attr_count]; + + memset(cda, 0, sizeof(struct nfattr *) * attr_count); + + err = nfnetlink_check_attributes(ss, nlh, cda); + if (err < 0) + goto err_inval; + + DEBUGP("calling handler\n"); + err = nc->call(nfnl, skb, nlh, cda, errp); + *errp = err; + return err; + } + +err_inval: + DEBUGP("returning -EINVAL\n"); + *errp = -EINVAL; + return -1; +} + +/* Process one packet of messages. */ +static inline int nfnetlink_rcv_skb(struct sk_buff *skb) +{ + int err; + struct nlmsghdr *nlh; + + while (skb->len >= NLMSG_SPACE(0)) { + u32 rlen; + + nlh = (struct nlmsghdr *)skb->data; + if (nlh->nlmsg_len < sizeof(struct nlmsghdr) + || skb->len < nlh->nlmsg_len) + return 0; + rlen = NLMSG_ALIGN(nlh->nlmsg_len); + if (rlen > skb->len) + rlen = skb->len; + if (nfnetlink_rcv_msg(skb, nlh, &err)) { + if (!err) + return -1; + netlink_ack(skb, nlh, err); + } else + if (nlh->nlmsg_flags & NLM_F_ACK) + netlink_ack(skb, nlh, 0); + skb_pull(skb, rlen); + } + + return 0; +} + +static void nfnetlink_rcv(struct sock *sk, int len) +{ + do { + struct sk_buff *skb; + + if (nfnl_shlock_nowait()) + return; + + while ((skb = skb_dequeue(&sk->sk_receive_queue)) != NULL) { + if (nfnetlink_rcv_skb(skb)) { + if (skb->len) + skb_queue_head(&sk->sk_receive_queue, + skb); + else + kfree_skb(skb); + break; + } + kfree_skb(skb); + } + + /* don't call nfnl_shunlock, since it would reenter + * with further packet processing */ + up(&nfnl_sem); + } while(nfnl && nfnl->sk_receive_queue.qlen); +} + +void __exit nfnetlink_exit(void) +{ + printk("Removing netfilter NETLINK layer.\n"); + sock_release(nfnl->sk_socket); + return; +} + +int __init nfnetlink_init(void) +{ + printk("Netfilter messages via NETLINK v%s.\n", nfversion); + + nfnl = netlink_kernel_create(NETLINK_NETFILTER, NFNLGRP_MAX, + nfnetlink_rcv, THIS_MODULE); + if (!nfnl) { + printk(KERN_ERR "cannot initialize nfnetlink!\n"); + return -1; + } + + return 0; +} + +module_init(nfnetlink_init); +module_exit(nfnetlink_exit); + +EXPORT_SYMBOL_GPL(nfnetlink_subsys_register); +EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister); +EXPORT_SYMBOL_GPL(nfnetlink_send); +EXPORT_SYMBOL_GPL(nfnetlink_unicast); +EXPORT_SYMBOL_GPL(nfattr_parse); +EXPORT_SYMBOL_GPL(__nfa_fill); diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c new file mode 100644 index 000000000000..ff5601ceedcb --- /dev/null +++ b/net/netfilter/nfnetlink_log.c @@ -0,0 +1,1055 @@ +/* + * This is a module which is used for logging packets to userspace via + * nfetlink. + * + * (C) 2005 by Harald Welte <laforge@netfilter.org> + * + * Based on the old ipv4-only ipt_ULOG.c: + * (C) 2000-2004 by Harald Welte <laforge@netfilter.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/init.h> +#include <linux/ip.h> +#include <linux/ipv6.h> +#include <linux/netdevice.h> +#include <linux/netfilter.h> +#include <linux/netlink.h> +#include <linux/netfilter/nfnetlink.h> +#include <linux/netfilter/nfnetlink_log.h> +#include <linux/spinlock.h> +#include <linux/sysctl.h> +#include <linux/proc_fs.h> +#include <linux/security.h> +#include <linux/list.h> +#include <linux/jhash.h> +#include <linux/random.h> +#include <net/sock.h> + +#include <asm/atomic.h> + +#ifdef CONFIG_BRIDGE_NETFILTER +#include "../bridge/br_private.h" +#endif + +#define NFULNL_NLBUFSIZ_DEFAULT 4096 +#define NFULNL_TIMEOUT_DEFAULT 100 /* every second */ +#define NFULNL_QTHRESH_DEFAULT 100 /* 100 packets */ + +#define PRINTR(x, args...) do { if (net_ratelimit()) \ + printk(x, ## args); } while (0); + +#if 0 +#define UDEBUG(x, args ...) printk(KERN_DEBUG "%s(%d):%s(): " x, \ + __FILE__, __LINE__, __FUNCTION__, \ + ## args) +#else +#define UDEBUG(x, ...) +#endif + +struct nfulnl_instance { + struct hlist_node hlist; /* global list of instances */ + spinlock_t lock; + atomic_t use; /* use count */ + + unsigned int qlen; /* number of nlmsgs in skb */ + struct sk_buff *skb; /* pre-allocatd skb */ + struct nlmsghdr *lastnlh; /* netlink header of last msg in skb */ + struct timer_list timer; + int peer_pid; /* PID of the peer process */ + + /* configurable parameters */ + unsigned int flushtimeout; /* timeout until queue flush */ + unsigned int nlbufsiz; /* netlink buffer allocation size */ + unsigned int qthreshold; /* threshold of the queue */ + u_int32_t copy_range; + u_int16_t group_num; /* number of this queue */ + u_int8_t copy_mode; +}; + +static DEFINE_RWLOCK(instances_lock); + +#define INSTANCE_BUCKETS 16 +static struct hlist_head instance_table[INSTANCE_BUCKETS]; +static unsigned int hash_init; + +static inline u_int8_t instance_hashfn(u_int16_t group_num) +{ + return ((group_num & 0xff) % INSTANCE_BUCKETS); +} + +static struct nfulnl_instance * +__instance_lookup(u_int16_t group_num) +{ + struct hlist_head *head; + struct hlist_node *pos; + struct nfulnl_instance *inst; + + UDEBUG("entering (group_num=%u)\n", group_num); + + head = &instance_table[instance_hashfn(group_num)]; + hlist_for_each_entry(inst, pos, head, hlist) { + if (inst->group_num == group_num) + return inst; + } + return NULL; +} + +static inline void +instance_get(struct nfulnl_instance *inst) +{ + atomic_inc(&inst->use); +} + +static struct nfulnl_instance * +instance_lookup_get(u_int16_t group_num) +{ + struct nfulnl_instance *inst; + + read_lock_bh(&instances_lock); + inst = __instance_lookup(group_num); + if (inst) + instance_get(inst); + read_unlock_bh(&instances_lock); + + return inst; +} + +static void +instance_put(struct nfulnl_instance *inst) +{ + if (inst && atomic_dec_and_test(&inst->use)) { + UDEBUG("kfree(inst=%p)\n", inst); + kfree(inst); + } +} + +static void nfulnl_timer(unsigned long data); + +static struct nfulnl_instance * +instance_create(u_int16_t group_num, int pid) +{ + struct nfulnl_instance *inst; + + UDEBUG("entering (group_num=%u, pid=%d)\n", group_num, + pid); + + write_lock_bh(&instances_lock); + if (__instance_lookup(group_num)) { + inst = NULL; + UDEBUG("aborting, instance already exists\n"); + goto out_unlock; + } + + inst = kmalloc(sizeof(*inst), GFP_ATOMIC); + if (!inst) + goto out_unlock; + + memset(inst, 0, sizeof(*inst)); + INIT_HLIST_NODE(&inst->hlist); + inst->lock = SPIN_LOCK_UNLOCKED; + /* needs to be two, since we _put() after creation */ + atomic_set(&inst->use, 2); + + init_timer(&inst->timer); + inst->timer.function = nfulnl_timer; + inst->timer.data = (unsigned long)inst; + /* don't start timer yet. (re)start it with every packet */ + + inst->peer_pid = pid; + inst->group_num = group_num; + + inst->qthreshold = NFULNL_QTHRESH_DEFAULT; + inst->flushtimeout = NFULNL_TIMEOUT_DEFAULT; + inst->nlbufsiz = NFULNL_NLBUFSIZ_DEFAULT; + inst->copy_mode = NFULNL_COPY_PACKET; + inst->copy_range = 0xffff; + + if (!try_module_get(THIS_MODULE)) + goto out_free; + + hlist_add_head(&inst->hlist, + &instance_table[instance_hashfn(group_num)]); + + UDEBUG("newly added node: %p, next=%p\n", &inst->hlist, + inst->hlist.next); + + write_unlock_bh(&instances_lock); + + return inst; + +out_free: + instance_put(inst); +out_unlock: + write_unlock_bh(&instances_lock); + return NULL; +} + +static int __nfulnl_send(struct nfulnl_instance *inst); + +static void +_instance_destroy2(struct nfulnl_instance *inst, int lock) +{ + /* first pull it out of the global list */ + if (lock) + write_lock_bh(&instances_lock); + + UDEBUG("removing instance %p (queuenum=%u) from hash\n", + inst, inst->group_num); + + hlist_del(&inst->hlist); + + if (lock) + write_unlock_bh(&instances_lock); + + /* then flush all pending packets from skb */ + + spin_lock_bh(&inst->lock); + if (inst->skb) { + if (inst->qlen) + __nfulnl_send(inst); + if (inst->skb) { + kfree_skb(inst->skb); + inst->skb = NULL; + } + } + spin_unlock_bh(&inst->lock); + + /* and finally put the refcount */ + instance_put(inst); + + module_put(THIS_MODULE); +} + +static inline void +__instance_destroy(struct nfulnl_instance *inst) +{ + _instance_destroy2(inst, 0); +} + +static inline void +instance_destroy(struct nfulnl_instance *inst) +{ + _instance_destroy2(inst, 1); +} + +static int +nfulnl_set_mode(struct nfulnl_instance *inst, u_int8_t mode, + unsigned int range) +{ + int status = 0; + + spin_lock_bh(&inst->lock); + + switch (mode) { + case NFULNL_COPY_NONE: + case NFULNL_COPY_META: + inst->copy_mode = mode; + inst->copy_range = 0; + break; + + case NFULNL_COPY_PACKET: + inst->copy_mode = mode; + /* we're using struct nfattr which has 16bit nfa_len */ + if (range > 0xffff) + inst->copy_range = 0xffff; + else + inst->copy_range = range; + break; + + default: + status = -EINVAL; + break; + } + + spin_unlock_bh(&inst->lock); + + return status; +} + +static int +nfulnl_set_nlbufsiz(struct nfulnl_instance *inst, u_int32_t nlbufsiz) +{ + int status; + + spin_lock_bh(&inst->lock); + if (nlbufsiz < NFULNL_NLBUFSIZ_DEFAULT) + status = -ERANGE; + else if (nlbufsiz > 131072) + status = -ERANGE; + else { + inst->nlbufsiz = nlbufsiz; + status = 0; + } + spin_unlock_bh(&inst->lock); + + return status; +} + +static int +nfulnl_set_timeout(struct nfulnl_instance *inst, u_int32_t timeout) +{ + spin_lock_bh(&inst->lock); + inst->flushtimeout = timeout; + spin_unlock_bh(&inst->lock); + + return 0; +} + +static int +nfulnl_set_qthresh(struct nfulnl_instance *inst, u_int32_t qthresh) +{ + spin_lock_bh(&inst->lock); + inst->qthreshold = qthresh; + spin_unlock_bh(&inst->lock); + + return 0; +} + +static struct sk_buff *nfulnl_alloc_skb(unsigned int inst_size, + unsigned int pkt_size) +{ + struct sk_buff *skb; + + UDEBUG("entered (%u, %u)\n", inst_size, pkt_size); + + /* alloc skb which should be big enough for a whole multipart + * message. WARNING: has to be <= 128k due to slab restrictions */ + + skb = alloc_skb(inst_size, GFP_ATOMIC); + if (!skb) { + PRINTR("nfnetlink_log: can't alloc whole buffer (%u bytes)\n", + inst_size); + + /* try to allocate only as much as we need for current + * packet */ + + skb = alloc_skb(pkt_size, GFP_ATOMIC); + if (!skb) + PRINTR("nfnetlink_log: can't even alloc %u bytes\n", + pkt_size); + } + + return skb; +} + +static int +__nfulnl_send(struct nfulnl_instance *inst) +{ + int status; + + if (timer_pending(&inst->timer)) + del_timer(&inst->timer); + + if (inst->qlen > 1) + inst->lastnlh->nlmsg_type = NLMSG_DONE; + + status = nfnetlink_unicast(inst->skb, inst->peer_pid, MSG_DONTWAIT); + if (status < 0) { + UDEBUG("netlink_unicast() failed\n"); + /* FIXME: statistics */ + } + + inst->qlen = 0; + inst->skb = NULL; + inst->lastnlh = NULL; + + return status; +} + +static void nfulnl_timer(unsigned long data) +{ + struct nfulnl_instance *inst = (struct nfulnl_instance *)data; + + UDEBUG("timer function called, flushing buffer\n"); + + spin_lock_bh(&inst->lock); + __nfulnl_send(inst); + instance_put(inst); + spin_unlock_bh(&inst->lock); +} + +static inline int +__build_packet_message(struct nfulnl_instance *inst, + const struct sk_buff *skb, + unsigned int data_len, + unsigned int pf, + unsigned int hooknum, + const struct net_device *indev, + const struct net_device *outdev, + const struct nf_loginfo *li, + const char *prefix) +{ + unsigned char *old_tail; + struct nfulnl_msg_packet_hdr pmsg; + struct nlmsghdr *nlh; + struct nfgenmsg *nfmsg; + u_int32_t tmp_uint; + + UDEBUG("entered\n"); + + old_tail = inst->skb->tail; + nlh = NLMSG_PUT(inst->skb, 0, 0, + NFNL_SUBSYS_ULOG << 8 | NFULNL_MSG_PACKET, + sizeof(struct nfgenmsg)); + nfmsg = NLMSG_DATA(nlh); + nfmsg->nfgen_family = pf; + nfmsg->version = NFNETLINK_V0; + nfmsg->res_id = htons(inst->group_num); + + pmsg.hw_protocol = htons(skb->protocol); + pmsg.hook = hooknum; + + NFA_PUT(inst->skb, NFULA_PACKET_HDR, sizeof(pmsg), &pmsg); + + if (prefix) { + int slen = strlen(prefix); + if (slen > NFULNL_PREFIXLEN) + slen = NFULNL_PREFIXLEN; + NFA_PUT(inst->skb, NFULA_PREFIX, slen, prefix); + } + + if (indev) { + tmp_uint = htonl(indev->ifindex); +#ifndef CONFIG_BRIDGE_NETFILTER + NFA_PUT(inst->skb, NFULA_IFINDEX_INDEV, sizeof(tmp_uint), + &tmp_uint); +#else + if (pf == PF_BRIDGE) { + /* Case 1: outdev is physical input device, we need to + * look for bridge group (when called from + * netfilter_bridge) */ + NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSINDEV, + sizeof(tmp_uint), &tmp_uint); + /* this is the bridge group "brX" */ + tmp_uint = htonl(indev->br_port->br->dev->ifindex); + NFA_PUT(inst->skb, NFULA_IFINDEX_INDEV, + sizeof(tmp_uint), &tmp_uint); + } else { + /* Case 2: indev is bridge group, we need to look for + * physical device (when called from ipv4) */ + NFA_PUT(inst->skb, NFULA_IFINDEX_INDEV, + sizeof(tmp_uint), &tmp_uint); + if (skb->nf_bridge && skb->nf_bridge->physindev) { + tmp_uint = + htonl(skb->nf_bridge->physindev->ifindex); + NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSINDEV, + sizeof(tmp_uint), &tmp_uint); + } + } +#endif + } + + if (outdev) { + tmp_uint = htonl(outdev->ifindex); +#ifndef CONFIG_BRIDGE_NETFILTER + NFA_PUT(inst->skb, NFULA_IFINDEX_OUTDEV, sizeof(tmp_uint), + &tmp_uint); +#else + if (pf == PF_BRIDGE) { + /* Case 1: outdev is physical output device, we need to + * look for bridge group (when called from + * netfilter_bridge) */ + NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSOUTDEV, + sizeof(tmp_uint), &tmp_uint); + /* this is the bridge group "brX" */ + tmp_uint = htonl(outdev->br_port->br->dev->ifindex); + NFA_PUT(inst->skb, NFULA_IFINDEX_OUTDEV, + sizeof(tmp_uint), &tmp_uint); + } else { + /* Case 2: indev is a bridge group, we need to look + * for physical device (when called from ipv4) */ + NFA_PUT(inst->skb, NFULA_IFINDEX_OUTDEV, + sizeof(tmp_uint), &tmp_uint); + if (skb->nf_bridge) { + tmp_uint = + htonl(skb->nf_bridge->physoutdev->ifindex); + NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSOUTDEV, + sizeof(tmp_uint), &tmp_uint); + } + } +#endif + } + + if (skb->nfmark) { + tmp_uint = htonl(skb->nfmark); + NFA_PUT(inst->skb, NFULA_MARK, sizeof(tmp_uint), &tmp_uint); + } + + if (indev && skb->dev && skb->dev->hard_header_parse) { + struct nfulnl_msg_packet_hw phw; + + phw.hw_addrlen = + skb->dev->hard_header_parse((struct sk_buff *)skb, + phw.hw_addr); + phw.hw_addrlen = htons(phw.hw_addrlen); + NFA_PUT(inst->skb, NFULA_HWADDR, sizeof(phw), &phw); + } + + if (skb->tstamp.off_sec) { + struct nfulnl_msg_packet_timestamp ts; + + ts.sec = cpu_to_be64(skb_tv_base.tv_sec + skb->tstamp.off_sec); + ts.usec = cpu_to_be64(skb_tv_base.tv_usec + skb->tstamp.off_usec); + + NFA_PUT(inst->skb, NFULA_TIMESTAMP, sizeof(ts), &ts); + } + + /* UID */ + if (skb->sk) { + read_lock_bh(&skb->sk->sk_callback_lock); + if (skb->sk->sk_socket && skb->sk->sk_socket->file) { + u_int32_t uid = htonl(skb->sk->sk_socket->file->f_uid); + /* need to unlock here since NFA_PUT may goto */ + read_unlock_bh(&skb->sk->sk_callback_lock); + NFA_PUT(inst->skb, NFULA_UID, sizeof(uid), &uid); + } else + read_unlock_bh(&skb->sk->sk_callback_lock); + } + + if (data_len) { + struct nfattr *nfa; + int size = NFA_LENGTH(data_len); + + if (skb_tailroom(inst->skb) < (int)NFA_SPACE(data_len)) { + printk(KERN_WARNING "nfnetlink_log: no tailroom!\n"); + goto nlmsg_failure; + } + + nfa = (struct nfattr *)skb_put(inst->skb, NFA_ALIGN(size)); + nfa->nfa_type = NFULA_PAYLOAD; + nfa->nfa_len = size; + + if (skb_copy_bits(skb, 0, NFA_DATA(nfa), data_len)) + BUG(); + } + + nlh->nlmsg_len = inst->skb->tail - old_tail; + return 0; + +nlmsg_failure: + UDEBUG("nlmsg_failure\n"); +nfattr_failure: + PRINTR(KERN_ERR "nfnetlink_log: error creating log nlmsg\n"); + return -1; +} + +#define RCV_SKB_FAIL(err) do { netlink_ack(skb, nlh, (err)); return; } while (0) + +static struct nf_loginfo default_loginfo = { + .type = NF_LOG_TYPE_ULOG, + .u = { + .ulog = { + .copy_len = 0xffff, + .group = 0, + .qthreshold = 1, + }, + }, +}; + +/* log handler for internal netfilter logging api */ +static void +nfulnl_log_packet(unsigned int pf, + unsigned int hooknum, + const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const struct nf_loginfo *li_user, + const char *prefix) +{ + unsigned int size, data_len; + struct nfulnl_instance *inst; + const struct nf_loginfo *li; + unsigned int qthreshold; + unsigned int nlbufsiz; + + if (li_user && li_user->type == NF_LOG_TYPE_ULOG) + li = li_user; + else + li = &default_loginfo; + + inst = instance_lookup_get(li->u.ulog.group); + if (!inst) + inst = instance_lookup_get(0); + if (!inst) { + PRINTR("nfnetlink_log: trying to log packet, " + "but no instance for group %u\n", li->u.ulog.group); + return; + } + + /* all macros expand to constant values at compile time */ + /* FIXME: do we want to make the size calculation conditional based on + * what is actually present? way more branches and checks, but more + * memory efficient... */ + size = NLMSG_SPACE(sizeof(struct nfgenmsg)) + + NFA_SPACE(sizeof(struct nfulnl_msg_packet_hdr)) + + NFA_SPACE(sizeof(u_int32_t)) /* ifindex */ + + NFA_SPACE(sizeof(u_int32_t)) /* ifindex */ +#ifdef CONFIG_BRIDGE_NETFILTER + + NFA_SPACE(sizeof(u_int32_t)) /* ifindex */ + + NFA_SPACE(sizeof(u_int32_t)) /* ifindex */ +#endif + + NFA_SPACE(sizeof(u_int32_t)) /* mark */ + + NFA_SPACE(sizeof(u_int32_t)) /* uid */ + + NFA_SPACE(NFULNL_PREFIXLEN) /* prefix */ + + NFA_SPACE(sizeof(struct nfulnl_msg_packet_hw)) + + NFA_SPACE(sizeof(struct nfulnl_msg_packet_timestamp)); + + UDEBUG("initial size=%u\n", size); + + spin_lock_bh(&inst->lock); + + qthreshold = inst->qthreshold; + /* per-rule qthreshold overrides per-instance */ + if (qthreshold > li->u.ulog.qthreshold) + qthreshold = li->u.ulog.qthreshold; + + switch (inst->copy_mode) { + case NFULNL_COPY_META: + case NFULNL_COPY_NONE: + data_len = 0; + break; + + case NFULNL_COPY_PACKET: + if (inst->copy_range == 0 + || inst->copy_range > skb->len) + data_len = skb->len; + else + data_len = inst->copy_range; + + size += NFA_SPACE(data_len); + UDEBUG("copy_packet, therefore size now %u\n", size); + break; + + default: + spin_unlock_bh(&inst->lock); + instance_put(inst); + return; + } + + if (size > inst->nlbufsiz) + nlbufsiz = size; + else + nlbufsiz = inst->nlbufsiz; + + if (!inst->skb) { + if (!(inst->skb = nfulnl_alloc_skb(nlbufsiz, size))) { + UDEBUG("error in nfulnl_alloc_skb(%u, %u)\n", + inst->nlbufsiz, size); + goto alloc_failure; + } + } else if (inst->qlen >= qthreshold || + size > skb_tailroom(inst->skb)) { + /* either the queue len is too high or we don't have + * enough room in the skb left. flush to userspace. */ + UDEBUG("flushing old skb\n"); + + __nfulnl_send(inst); + + if (!(inst->skb = nfulnl_alloc_skb(nlbufsiz, size))) { + UDEBUG("error in nfulnl_alloc_skb(%u, %u)\n", + inst->nlbufsiz, size); + goto alloc_failure; + } + } + + UDEBUG("qlen %d, qthreshold %d\n", inst->qlen, qthreshold); + inst->qlen++; + + __build_packet_message(inst, skb, data_len, pf, + hooknum, in, out, li, prefix); + + /* timer_pending always called within inst->lock, so there + * is no chance of a race here */ + if (!timer_pending(&inst->timer)) { + instance_get(inst); + inst->timer.expires = jiffies + (inst->flushtimeout*HZ/100); + add_timer(&inst->timer); + } + spin_unlock_bh(&inst->lock); + + return; + +alloc_failure: + spin_unlock_bh(&inst->lock); + instance_put(inst); + UDEBUG("error allocating skb\n"); + /* FIXME: statistics */ +} + +static int +nfulnl_rcv_nl_event(struct notifier_block *this, + unsigned long event, void *ptr) +{ + struct netlink_notify *n = ptr; + + if (event == NETLINK_URELEASE && + n->protocol == NETLINK_NETFILTER && n->pid) { + int i; + + /* destroy all instances for this pid */ + write_lock_bh(&instances_lock); + for (i = 0; i < INSTANCE_BUCKETS; i++) { + struct hlist_node *tmp, *t2; + struct nfulnl_instance *inst; + struct hlist_head *head = &instance_table[i]; + + hlist_for_each_entry_safe(inst, tmp, t2, head, hlist) { + UDEBUG("node = %p\n", inst); + if (n->pid == inst->peer_pid) + __instance_destroy(inst); + } + } + write_unlock_bh(&instances_lock); + } + return NOTIFY_DONE; +} + +static struct notifier_block nfulnl_rtnl_notifier = { + .notifier_call = nfulnl_rcv_nl_event, +}; + +static int +nfulnl_recv_unsupp(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *nfqa[], int *errp) +{ + return -ENOTSUPP; +} + +static struct nf_logger nfulnl_logger = { + .name = "nfnetlink_log", + .logfn = &nfulnl_log_packet, + .me = THIS_MODULE, +}; + +static const int nfula_min[NFULA_MAX] = { + [NFULA_PACKET_HDR-1] = sizeof(struct nfulnl_msg_packet_hdr), + [NFULA_MARK-1] = sizeof(u_int32_t), + [NFULA_TIMESTAMP-1] = sizeof(struct nfulnl_msg_packet_timestamp), + [NFULA_IFINDEX_INDEV-1] = sizeof(u_int32_t), + [NFULA_IFINDEX_OUTDEV-1]= sizeof(u_int32_t), + [NFULA_HWADDR-1] = sizeof(struct nfulnl_msg_packet_hw), + [NFULA_PAYLOAD-1] = 0, + [NFULA_PREFIX-1] = 0, + [NFULA_UID-1] = sizeof(u_int32_t), +}; + +static const int nfula_cfg_min[NFULA_CFG_MAX] = { + [NFULA_CFG_CMD-1] = sizeof(struct nfulnl_msg_config_cmd), + [NFULA_CFG_MODE-1] = sizeof(struct nfulnl_msg_config_mode), + [NFULA_CFG_TIMEOUT-1] = sizeof(u_int32_t), + [NFULA_CFG_QTHRESH-1] = sizeof(u_int32_t), + [NFULA_CFG_NLBUFSIZ-1] = sizeof(u_int32_t), +}; + +static int +nfulnl_recv_config(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *nfula[], int *errp) +{ + struct nfgenmsg *nfmsg = NLMSG_DATA(nlh); + u_int16_t group_num = ntohs(nfmsg->res_id); + struct nfulnl_instance *inst; + int ret = 0; + + UDEBUG("entering for msg %u\n", NFNL_MSG_TYPE(nlh->nlmsg_type)); + + if (nfattr_bad_size(nfula, NFULA_CFG_MAX, nfula_cfg_min)) { + UDEBUG("bad attribute size\n"); + return -EINVAL; + } + + inst = instance_lookup_get(group_num); + if (nfula[NFULA_CFG_CMD-1]) { + u_int8_t pf = nfmsg->nfgen_family; + struct nfulnl_msg_config_cmd *cmd; + cmd = NFA_DATA(nfula[NFULA_CFG_CMD-1]); + UDEBUG("found CFG_CMD for\n"); + + switch (cmd->command) { + case NFULNL_CFG_CMD_BIND: + if (inst) { + ret = -EBUSY; + goto out_put; + } + + inst = instance_create(group_num, + NETLINK_CB(skb).pid); + if (!inst) { + ret = -EINVAL; + goto out_put; + } + break; + case NFULNL_CFG_CMD_UNBIND: + if (!inst) { + ret = -ENODEV; + goto out_put; + } + + if (inst->peer_pid != NETLINK_CB(skb).pid) { + ret = -EPERM; + goto out_put; + } + + instance_destroy(inst); + break; + case NFULNL_CFG_CMD_PF_BIND: + UDEBUG("registering log handler for pf=%u\n", pf); + ret = nf_log_register(pf, &nfulnl_logger); + break; + case NFULNL_CFG_CMD_PF_UNBIND: + UDEBUG("unregistering log handler for pf=%u\n", pf); + /* This is a bug and a feature. We cannot unregister + * other handlers, like nfnetlink_inst can */ + nf_log_unregister_pf(pf); + break; + default: + ret = -EINVAL; + break; + } + } else { + if (!inst) { + UDEBUG("no config command, and no instance for " + "group=%u pid=%u =>ENOENT\n", + group_num, NETLINK_CB(skb).pid); + ret = -ENOENT; + goto out_put; + } + + if (inst->peer_pid != NETLINK_CB(skb).pid) { + UDEBUG("no config command, and wrong pid\n"); + ret = -EPERM; + goto out_put; + } + } + + if (nfula[NFULA_CFG_MODE-1]) { + struct nfulnl_msg_config_mode *params; + params = NFA_DATA(nfula[NFULA_CFG_MODE-1]); + + nfulnl_set_mode(inst, params->copy_mode, + ntohs(params->copy_range)); + } + + if (nfula[NFULA_CFG_TIMEOUT-1]) { + u_int32_t timeout = + *(u_int32_t *)NFA_DATA(nfula[NFULA_CFG_TIMEOUT-1]); + + nfulnl_set_timeout(inst, ntohl(timeout)); + } + + if (nfula[NFULA_CFG_NLBUFSIZ-1]) { + u_int32_t nlbufsiz = + *(u_int32_t *)NFA_DATA(nfula[NFULA_CFG_NLBUFSIZ-1]); + + nfulnl_set_nlbufsiz(inst, ntohl(nlbufsiz)); + } + + if (nfula[NFULA_CFG_QTHRESH-1]) { + u_int32_t qthresh = + *(u_int16_t *)NFA_DATA(nfula[NFULA_CFG_QTHRESH-1]); + + nfulnl_set_qthresh(inst, ntohl(qthresh)); + } + +out_put: + instance_put(inst); + return ret; +} + +static struct nfnl_callback nfulnl_cb[NFULNL_MSG_MAX] = { + [NFULNL_MSG_PACKET] = { .call = nfulnl_recv_unsupp, + .attr_count = NFULA_MAX, + .cap_required = CAP_NET_ADMIN, }, + [NFULNL_MSG_CONFIG] = { .call = nfulnl_recv_config, + .attr_count = NFULA_CFG_MAX, + .cap_required = CAP_NET_ADMIN }, +}; + +static struct nfnetlink_subsystem nfulnl_subsys = { + .name = "log", + .subsys_id = NFNL_SUBSYS_ULOG, + .cb_count = NFULNL_MSG_MAX, + .cb = nfulnl_cb, +}; + +#ifdef CONFIG_PROC_FS +struct iter_state { + unsigned int bucket; +}; + +static struct hlist_node *get_first(struct seq_file *seq) +{ + struct iter_state *st = seq->private; + + if (!st) + return NULL; + + for (st->bucket = 0; st->bucket < INSTANCE_BUCKETS; st->bucket++) { + if (!hlist_empty(&instance_table[st->bucket])) + return instance_table[st->bucket].first; + } + return NULL; +} + +static struct hlist_node *get_next(struct seq_file *seq, struct hlist_node *h) +{ + struct iter_state *st = seq->private; + + h = h->next; + while (!h) { + if (++st->bucket >= INSTANCE_BUCKETS) + return NULL; + + h = instance_table[st->bucket].first; + } + return h; +} + +static struct hlist_node *get_idx(struct seq_file *seq, loff_t pos) +{ + struct hlist_node *head; + head = get_first(seq); + + if (head) + while (pos && (head = get_next(seq, head))) + pos--; + return pos ? NULL : head; +} + +static void *seq_start(struct seq_file *seq, loff_t *pos) +{ + read_lock_bh(&instances_lock); + return get_idx(seq, *pos); +} + +static void *seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + (*pos)++; + return get_next(s, v); +} + +static void seq_stop(struct seq_file *s, void *v) +{ + read_unlock_bh(&instances_lock); +} + +static int seq_show(struct seq_file *s, void *v) +{ + const struct nfulnl_instance *inst = v; + + return seq_printf(s, "%5d %6d %5d %1d %5d %6d %2d\n", + inst->group_num, + inst->peer_pid, inst->qlen, + inst->copy_mode, inst->copy_range, + inst->flushtimeout, atomic_read(&inst->use)); +} + +static struct seq_operations nful_seq_ops = { + .start = seq_start, + .next = seq_next, + .stop = seq_stop, + .show = seq_show, +}; + +static int nful_open(struct inode *inode, struct file *file) +{ + struct seq_file *seq; + struct iter_state *is; + int ret; + + is = kmalloc(sizeof(*is), GFP_KERNEL); + if (!is) + return -ENOMEM; + memset(is, 0, sizeof(*is)); + ret = seq_open(file, &nful_seq_ops); + if (ret < 0) + goto out_free; + seq = file->private_data; + seq->private = is; + return ret; +out_free: + kfree(is); + return ret; +} + +static struct file_operations nful_file_ops = { + .owner = THIS_MODULE, + .open = nful_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +#endif /* PROC_FS */ + +static int +init_or_cleanup(int init) +{ + int i, status = -ENOMEM; +#ifdef CONFIG_PROC_FS + struct proc_dir_entry *proc_nful; +#endif + + if (!init) + goto cleanup; + + for (i = 0; i < INSTANCE_BUCKETS; i++) + INIT_HLIST_HEAD(&instance_table[i]); + + /* it's not really all that important to have a random value, so + * we can do this from the init function, even if there hasn't + * been that much entropy yet */ + get_random_bytes(&hash_init, sizeof(hash_init)); + + netlink_register_notifier(&nfulnl_rtnl_notifier); + status = nfnetlink_subsys_register(&nfulnl_subsys); + if (status < 0) { + printk(KERN_ERR "log: failed to create netlink socket\n"); + goto cleanup_netlink_notifier; + } + +#ifdef CONFIG_PROC_FS + proc_nful = create_proc_entry("nfnetlink_log", 0440, + proc_net_netfilter); + if (!proc_nful) + goto cleanup_subsys; + proc_nful->proc_fops = &nful_file_ops; +#endif + + return status; + +cleanup: + nf_log_unregister_logger(&nfulnl_logger); +#ifdef CONFIG_PROC_FS + remove_proc_entry("nfnetlink_log", proc_net_netfilter); +cleanup_subsys: +#endif + nfnetlink_subsys_unregister(&nfulnl_subsys); +cleanup_netlink_notifier: + netlink_unregister_notifier(&nfulnl_rtnl_notifier); + return status; +} + +static int __init init(void) +{ + + return init_or_cleanup(1); +} + +static void __exit fini(void) +{ + init_or_cleanup(0); +} + +MODULE_DESCRIPTION("netfilter userspace logging"); +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS_NFNL_SUBSYS(NFNL_SUBSYS_ULOG); + +module_init(init); +module_exit(fini); diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c new file mode 100644 index 000000000000..e3a5285329af --- /dev/null +++ b/net/netfilter/nfnetlink_queue.c @@ -0,0 +1,1132 @@ +/* + * This is a module which is used for queueing packets and communicating with + * userspace via nfetlink. + * + * (C) 2005 by Harald Welte <laforge@netfilter.org> + * + * Based on the old ipv4-only ip_queue.c: + * (C) 2000-2002 James Morris <jmorris@intercode.com.au> + * (C) 2003-2005 Netfilter Core Team <coreteam@netfilter.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ +#include <linux/module.h> +#include <linux/skbuff.h> +#include <linux/init.h> +#include <linux/spinlock.h> +#include <linux/notifier.h> +#include <linux/netdevice.h> +#include <linux/netfilter.h> +#include <linux/proc_fs.h> +#include <linux/netfilter_ipv4.h> +#include <linux/netfilter_ipv6.h> +#include <linux/netfilter/nfnetlink.h> +#include <linux/netfilter/nfnetlink_queue.h> +#include <linux/list.h> +#include <net/sock.h> + +#include <asm/atomic.h> + +#ifdef CONFIG_BRIDGE_NETFILTER +#include "../bridge/br_private.h" +#endif + +#define NFQNL_QMAX_DEFAULT 1024 + +#if 0 +#define QDEBUG(x, args ...) printk(KERN_DEBUG "%s(%d):%s(): " x, \ + __FILE__, __LINE__, __FUNCTION__, \ + ## args) +#else +#define QDEBUG(x, ...) +#endif + +struct nfqnl_queue_entry { + struct list_head list; + struct nf_info *info; + struct sk_buff *skb; + unsigned int id; +}; + +struct nfqnl_instance { + struct hlist_node hlist; /* global list of queues */ + atomic_t use; + + int peer_pid; + unsigned int queue_maxlen; + unsigned int copy_range; + unsigned int queue_total; + unsigned int queue_dropped; + unsigned int queue_user_dropped; + + atomic_t id_sequence; /* 'sequence' of pkt ids */ + + u_int16_t queue_num; /* number of this queue */ + u_int8_t copy_mode; + + spinlock_t lock; + + struct list_head queue_list; /* packets in queue */ +}; + +typedef int (*nfqnl_cmpfn)(struct nfqnl_queue_entry *, unsigned long); + +static DEFINE_RWLOCK(instances_lock); + +u_int64_t htonll(u_int64_t in) +{ + u_int64_t out; + int i; + + for (i = 0; i < sizeof(u_int64_t); i++) + ((u_int8_t *)&out)[sizeof(u_int64_t)-1] = ((u_int8_t *)&in)[i]; + + return out; +} + +#define INSTANCE_BUCKETS 16 +static struct hlist_head instance_table[INSTANCE_BUCKETS]; + +static inline u_int8_t instance_hashfn(u_int16_t queue_num) +{ + return ((queue_num >> 8) | queue_num) % INSTANCE_BUCKETS; +} + +static struct nfqnl_instance * +__instance_lookup(u_int16_t queue_num) +{ + struct hlist_head *head; + struct hlist_node *pos; + struct nfqnl_instance *inst; + + head = &instance_table[instance_hashfn(queue_num)]; + hlist_for_each_entry(inst, pos, head, hlist) { + if (inst->queue_num == queue_num) + return inst; + } + return NULL; +} + +static struct nfqnl_instance * +instance_lookup_get(u_int16_t queue_num) +{ + struct nfqnl_instance *inst; + + read_lock_bh(&instances_lock); + inst = __instance_lookup(queue_num); + if (inst) + atomic_inc(&inst->use); + read_unlock_bh(&instances_lock); + + return inst; +} + +static void +instance_put(struct nfqnl_instance *inst) +{ + if (inst && atomic_dec_and_test(&inst->use)) { + QDEBUG("kfree(inst=%p)\n", inst); + kfree(inst); + } +} + +static struct nfqnl_instance * +instance_create(u_int16_t queue_num, int pid) +{ + struct nfqnl_instance *inst; + + QDEBUG("entering for queue_num=%u, pid=%d\n", queue_num, pid); + + write_lock_bh(&instances_lock); + if (__instance_lookup(queue_num)) { + inst = NULL; + QDEBUG("aborting, instance already exists\n"); + goto out_unlock; + } + + inst = kmalloc(sizeof(*inst), GFP_ATOMIC); + if (!inst) + goto out_unlock; + + memset(inst, 0, sizeof(*inst)); + inst->queue_num = queue_num; + inst->peer_pid = pid; + inst->queue_maxlen = NFQNL_QMAX_DEFAULT; + inst->copy_range = 0xfffff; + inst->copy_mode = NFQNL_COPY_NONE; + atomic_set(&inst->id_sequence, 0); + /* needs to be two, since we _put() after creation */ + atomic_set(&inst->use, 2); + inst->lock = SPIN_LOCK_UNLOCKED; + INIT_LIST_HEAD(&inst->queue_list); + + if (!try_module_get(THIS_MODULE)) + goto out_free; + + hlist_add_head(&inst->hlist, + &instance_table[instance_hashfn(queue_num)]); + + write_unlock_bh(&instances_lock); + + QDEBUG("successfully created new instance\n"); + + return inst; + +out_free: + kfree(inst); +out_unlock: + write_unlock_bh(&instances_lock); + return NULL; +} + +static void nfqnl_flush(struct nfqnl_instance *queue, int verdict); + +static void +_instance_destroy2(struct nfqnl_instance *inst, int lock) +{ + /* first pull it out of the global list */ + if (lock) + write_lock_bh(&instances_lock); + + QDEBUG("removing instance %p (queuenum=%u) from hash\n", + inst, inst->queue_num); + hlist_del(&inst->hlist); + + if (lock) + write_unlock_bh(&instances_lock); + + /* then flush all pending skbs from the queue */ + nfqnl_flush(inst, NF_DROP); + + /* and finally put the refcount */ + instance_put(inst); + + module_put(THIS_MODULE); +} + +static inline void +__instance_destroy(struct nfqnl_instance *inst) +{ + _instance_destroy2(inst, 0); +} + +static inline void +instance_destroy(struct nfqnl_instance *inst) +{ + _instance_destroy2(inst, 1); +} + + + +static void +issue_verdict(struct nfqnl_queue_entry *entry, int verdict) +{ + QDEBUG("entering for entry %p, verdict %u\n", entry, verdict); + + /* TCP input path (and probably other bits) assume to be called + * from softirq context, not from syscall, like issue_verdict is + * called. TCP input path deadlocks with locks taken from timer + * softirq, e.g. We therefore emulate this by local_bh_disable() */ + + local_bh_disable(); + nf_reinject(entry->skb, entry->info, verdict); + local_bh_enable(); + + kfree(entry); +} + +static inline void +__enqueue_entry(struct nfqnl_instance *queue, + struct nfqnl_queue_entry *entry) +{ + list_add(&entry->list, &queue->queue_list); + queue->queue_total++; +} + +/* + * Find and return a queued entry matched by cmpfn, or return the last + * entry if cmpfn is NULL. + */ +static inline struct nfqnl_queue_entry * +__find_entry(struct nfqnl_instance *queue, nfqnl_cmpfn cmpfn, + unsigned long data) +{ + struct list_head *p; + + list_for_each_prev(p, &queue->queue_list) { + struct nfqnl_queue_entry *entry = (struct nfqnl_queue_entry *)p; + + if (!cmpfn || cmpfn(entry, data)) + return entry; + } + return NULL; +} + +static inline void +__dequeue_entry(struct nfqnl_instance *q, struct nfqnl_queue_entry *entry) +{ + list_del(&entry->list); + q->queue_total--; +} + +static inline struct nfqnl_queue_entry * +__find_dequeue_entry(struct nfqnl_instance *queue, + nfqnl_cmpfn cmpfn, unsigned long data) +{ + struct nfqnl_queue_entry *entry; + + entry = __find_entry(queue, cmpfn, data); + if (entry == NULL) + return NULL; + + __dequeue_entry(queue, entry); + return entry; +} + + +static inline void +__nfqnl_flush(struct nfqnl_instance *queue, int verdict) +{ + struct nfqnl_queue_entry *entry; + + while ((entry = __find_dequeue_entry(queue, NULL, 0))) + issue_verdict(entry, verdict); +} + +static inline int +__nfqnl_set_mode(struct nfqnl_instance *queue, + unsigned char mode, unsigned int range) +{ + int status = 0; + + switch (mode) { + case NFQNL_COPY_NONE: + case NFQNL_COPY_META: + queue->copy_mode = mode; + queue->copy_range = 0; + break; + + case NFQNL_COPY_PACKET: + queue->copy_mode = mode; + /* we're using struct nfattr which has 16bit nfa_len */ + if (range > 0xffff) + queue->copy_range = 0xffff; + else + queue->copy_range = range; + break; + + default: + status = -EINVAL; + + } + return status; +} + +static struct nfqnl_queue_entry * +find_dequeue_entry(struct nfqnl_instance *queue, + nfqnl_cmpfn cmpfn, unsigned long data) +{ + struct nfqnl_queue_entry *entry; + + spin_lock_bh(&queue->lock); + entry = __find_dequeue_entry(queue, cmpfn, data); + spin_unlock_bh(&queue->lock); + + return entry; +} + +static void +nfqnl_flush(struct nfqnl_instance *queue, int verdict) +{ + spin_lock_bh(&queue->lock); + __nfqnl_flush(queue, verdict); + spin_unlock_bh(&queue->lock); +} + +static struct sk_buff * +nfqnl_build_packet_message(struct nfqnl_instance *queue, + struct nfqnl_queue_entry *entry, int *errp) +{ + unsigned char *old_tail; + size_t size; + size_t data_len = 0; + struct sk_buff *skb; + struct nfqnl_msg_packet_hdr pmsg; + struct nlmsghdr *nlh; + struct nfgenmsg *nfmsg; + unsigned int tmp_uint; + + QDEBUG("entered\n"); + + /* all macros expand to constant values at compile time */ + size = NLMSG_SPACE(sizeof(struct nfqnl_msg_packet_hdr)) + + NLMSG_SPACE(sizeof(u_int32_t)) /* ifindex */ + + NLMSG_SPACE(sizeof(u_int32_t)) /* ifindex */ +#ifdef CONFIG_BRIDGE_NETFILTER + + NLMSG_SPACE(sizeof(u_int32_t)) /* ifindex */ + + NLMSG_SPACE(sizeof(u_int32_t)) /* ifindex */ +#endif + + NLMSG_SPACE(sizeof(u_int32_t)) /* mark */ + + NLMSG_SPACE(sizeof(struct nfqnl_msg_packet_hw)) + + NLMSG_SPACE(sizeof(struct nfqnl_msg_packet_timestamp)); + + spin_lock_bh(&queue->lock); + + switch (queue->copy_mode) { + case NFQNL_COPY_META: + case NFQNL_COPY_NONE: + data_len = 0; + break; + + case NFQNL_COPY_PACKET: + if (queue->copy_range == 0 + || queue->copy_range > entry->skb->len) + data_len = entry->skb->len; + else + data_len = queue->copy_range; + + size += NLMSG_SPACE(data_len); + break; + + default: + *errp = -EINVAL; + spin_unlock_bh(&queue->lock); + return NULL; + } + + spin_unlock_bh(&queue->lock); + + skb = alloc_skb(size, GFP_ATOMIC); + if (!skb) + goto nlmsg_failure; + + old_tail= skb->tail; + nlh = NLMSG_PUT(skb, 0, 0, + NFNL_SUBSYS_QUEUE << 8 | NFQNL_MSG_PACKET, + sizeof(struct nfgenmsg)); + nfmsg = NLMSG_DATA(nlh); + nfmsg->nfgen_family = entry->info->pf; + nfmsg->version = NFNETLINK_V0; + nfmsg->res_id = htons(queue->queue_num); + + pmsg.packet_id = htonl(entry->id); + pmsg.hw_protocol = htons(entry->skb->protocol); + pmsg.hook = entry->info->hook; + + NFA_PUT(skb, NFQA_PACKET_HDR, sizeof(pmsg), &pmsg); + + if (entry->info->indev) { + tmp_uint = htonl(entry->info->indev->ifindex); +#ifndef CONFIG_BRIDGE_NETFILTER + NFA_PUT(skb, NFQA_IFINDEX_INDEV, sizeof(tmp_uint), &tmp_uint); +#else + if (entry->info->pf == PF_BRIDGE) { + /* Case 1: indev is physical input device, we need to + * look for bridge group (when called from + * netfilter_bridge) */ + NFA_PUT(skb, NFQA_IFINDEX_PHYSINDEV, sizeof(tmp_uint), + &tmp_uint); + /* this is the bridge group "brX" */ + tmp_uint = htonl(entry->info->indev->br_port->br->dev->ifindex); + NFA_PUT(skb, NFQA_IFINDEX_INDEV, sizeof(tmp_uint), + &tmp_uint); + } else { + /* Case 2: indev is bridge group, we need to look for + * physical device (when called from ipv4) */ + NFA_PUT(skb, NFQA_IFINDEX_INDEV, sizeof(tmp_uint), + &tmp_uint); + if (entry->skb->nf_bridge + && entry->skb->nf_bridge->physindev) { + tmp_uint = htonl(entry->skb->nf_bridge->physindev->ifindex); + NFA_PUT(skb, NFQA_IFINDEX_PHYSINDEV, + sizeof(tmp_uint), &tmp_uint); + } + } +#endif + } + + if (entry->info->outdev) { + tmp_uint = htonl(entry->info->outdev->ifindex); +#ifndef CONFIG_BRIDGE_NETFILTER + NFA_PUT(skb, NFQA_IFINDEX_OUTDEV, sizeof(tmp_uint), &tmp_uint); +#else + if (entry->info->pf == PF_BRIDGE) { + /* Case 1: outdev is physical output device, we need to + * look for bridge group (when called from + * netfilter_bridge) */ + NFA_PUT(skb, NFQA_IFINDEX_PHYSOUTDEV, sizeof(tmp_uint), + &tmp_uint); + /* this is the bridge group "brX" */ + tmp_uint = htonl(entry->info->outdev->br_port->br->dev->ifindex); + NFA_PUT(skb, NFQA_IFINDEX_OUTDEV, sizeof(tmp_uint), + &tmp_uint); + } else { + /* Case 2: outdev is bridge group, we need to look for + * physical output device (when called from ipv4) */ + NFA_PUT(skb, NFQA_IFINDEX_OUTDEV, sizeof(tmp_uint), + &tmp_uint); + if (entry->skb->nf_bridge + && entry->skb->nf_bridge->physoutdev) { + tmp_uint = htonl(entry->skb->nf_bridge->physoutdev->ifindex); + NFA_PUT(skb, NFQA_IFINDEX_PHYSOUTDEV, + sizeof(tmp_uint), &tmp_uint); + } + } +#endif + } + + if (entry->skb->nfmark) { + tmp_uint = htonl(entry->skb->nfmark); + NFA_PUT(skb, NFQA_MARK, sizeof(u_int32_t), &tmp_uint); + } + + if (entry->info->indev && entry->skb->dev + && entry->skb->dev->hard_header_parse) { + struct nfqnl_msg_packet_hw phw; + + phw.hw_addrlen = + entry->skb->dev->hard_header_parse(entry->skb, + phw.hw_addr); + phw.hw_addrlen = htons(phw.hw_addrlen); + NFA_PUT(skb, NFQA_HWADDR, sizeof(phw), &phw); + } + + if (entry->skb->tstamp.off_sec) { + struct nfqnl_msg_packet_timestamp ts; + + ts.sec = htonll(skb_tv_base.tv_sec + entry->skb->tstamp.off_sec); + ts.usec = htonll(skb_tv_base.tv_usec + entry->skb->tstamp.off_usec); + + NFA_PUT(skb, NFQA_TIMESTAMP, sizeof(ts), &ts); + } + + if (data_len) { + struct nfattr *nfa; + int size = NFA_LENGTH(data_len); + + if (skb_tailroom(skb) < (int)NFA_SPACE(data_len)) { + printk(KERN_WARNING "nf_queue: no tailroom!\n"); + goto nlmsg_failure; + } + + nfa = (struct nfattr *)skb_put(skb, NFA_ALIGN(size)); + nfa->nfa_type = NFQA_PAYLOAD; + nfa->nfa_len = size; + + if (skb_copy_bits(entry->skb, 0, NFA_DATA(nfa), data_len)) + BUG(); + } + + nlh->nlmsg_len = skb->tail - old_tail; + return skb; + +nlmsg_failure: +nfattr_failure: + if (skb) + kfree_skb(skb); + *errp = -EINVAL; + if (net_ratelimit()) + printk(KERN_ERR "nf_queue: error creating packet message\n"); + return NULL; +} + +static int +nfqnl_enqueue_packet(struct sk_buff *skb, struct nf_info *info, + unsigned int queuenum, void *data) +{ + int status = -EINVAL; + struct sk_buff *nskb; + struct nfqnl_instance *queue; + struct nfqnl_queue_entry *entry; + + QDEBUG("entered\n"); + + queue = instance_lookup_get(queuenum); + if (!queue) { + QDEBUG("no queue instance matching\n"); + return -EINVAL; + } + + if (queue->copy_mode == NFQNL_COPY_NONE) { + QDEBUG("mode COPY_NONE, aborting\n"); + status = -EAGAIN; + goto err_out_put; + } + + entry = kmalloc(sizeof(*entry), GFP_ATOMIC); + if (entry == NULL) { + if (net_ratelimit()) + printk(KERN_ERR + "nf_queue: OOM in nfqnl_enqueue_packet()\n"); + status = -ENOMEM; + goto err_out_put; + } + + entry->info = info; + entry->skb = skb; + entry->id = atomic_inc_return(&queue->id_sequence); + + nskb = nfqnl_build_packet_message(queue, entry, &status); + if (nskb == NULL) + goto err_out_free; + + spin_lock_bh(&queue->lock); + + if (!queue->peer_pid) + goto err_out_free_nskb; + + if (queue->queue_total >= queue->queue_maxlen) { + queue->queue_dropped++; + status = -ENOSPC; + if (net_ratelimit()) + printk(KERN_WARNING "ip_queue: full at %d entries, " + "dropping packets(s). Dropped: %d\n", + queue->queue_total, queue->queue_dropped); + goto err_out_free_nskb; + } + + /* nfnetlink_unicast will either free the nskb or add it to a socket */ + status = nfnetlink_unicast(nskb, queue->peer_pid, MSG_DONTWAIT); + if (status < 0) { + queue->queue_user_dropped++; + goto err_out_unlock; + } + + __enqueue_entry(queue, entry); + + spin_unlock_bh(&queue->lock); + instance_put(queue); + return status; + +err_out_free_nskb: + kfree_skb(nskb); + +err_out_unlock: + spin_unlock_bh(&queue->lock); + +err_out_free: + kfree(entry); +err_out_put: + instance_put(queue); + return status; +} + +static int +nfqnl_mangle(void *data, int data_len, struct nfqnl_queue_entry *e) +{ + int diff; + + diff = data_len - e->skb->len; + if (diff < 0) + skb_trim(e->skb, data_len); + else if (diff > 0) { + if (data_len > 0xFFFF) + return -EINVAL; + if (diff > skb_tailroom(e->skb)) { + struct sk_buff *newskb; + + newskb = skb_copy_expand(e->skb, + skb_headroom(e->skb), + diff, + GFP_ATOMIC); + if (newskb == NULL) { + printk(KERN_WARNING "ip_queue: OOM " + "in mangle, dropping packet\n"); + return -ENOMEM; + } + if (e->skb->sk) + skb_set_owner_w(newskb, e->skb->sk); + kfree_skb(e->skb); + e->skb = newskb; + } + skb_put(e->skb, diff); + } + if (!skb_make_writable(&e->skb, data_len)) + return -ENOMEM; + memcpy(e->skb->data, data, data_len); + + return 0; +} + +static inline int +id_cmp(struct nfqnl_queue_entry *e, unsigned long id) +{ + return (id == e->id); +} + +static int +nfqnl_set_mode(struct nfqnl_instance *queue, + unsigned char mode, unsigned int range) +{ + int status; + + spin_lock_bh(&queue->lock); + status = __nfqnl_set_mode(queue, mode, range); + spin_unlock_bh(&queue->lock); + + return status; +} + +static int +dev_cmp(struct nfqnl_queue_entry *entry, unsigned long ifindex) +{ + if (entry->info->indev) + if (entry->info->indev->ifindex == ifindex) + return 1; + + if (entry->info->outdev) + if (entry->info->outdev->ifindex == ifindex) + return 1; + + return 0; +} + +/* drop all packets with either indev or outdev == ifindex from all queue + * instances */ +static void +nfqnl_dev_drop(int ifindex) +{ + int i; + + QDEBUG("entering for ifindex %u\n", ifindex); + + /* this only looks like we have to hold the readlock for a way too long + * time, issue_verdict(), nf_reinject(), ... - but we always only + * issue NF_DROP, which is processed directly in nf_reinject() */ + read_lock_bh(&instances_lock); + + for (i = 0; i < INSTANCE_BUCKETS; i++) { + struct hlist_node *tmp; + struct nfqnl_instance *inst; + struct hlist_head *head = &instance_table[i]; + + hlist_for_each_entry(inst, tmp, head, hlist) { + struct nfqnl_queue_entry *entry; + while ((entry = find_dequeue_entry(inst, dev_cmp, + ifindex)) != NULL) + issue_verdict(entry, NF_DROP); + } + } + + read_unlock_bh(&instances_lock); +} + +#define RCV_SKB_FAIL(err) do { netlink_ack(skb, nlh, (err)); return; } while (0) + +static int +nfqnl_rcv_dev_event(struct notifier_block *this, + unsigned long event, void *ptr) +{ + struct net_device *dev = ptr; + + /* Drop any packets associated with the downed device */ + if (event == NETDEV_DOWN) + nfqnl_dev_drop(dev->ifindex); + return NOTIFY_DONE; +} + +static struct notifier_block nfqnl_dev_notifier = { + .notifier_call = nfqnl_rcv_dev_event, +}; + +static int +nfqnl_rcv_nl_event(struct notifier_block *this, + unsigned long event, void *ptr) +{ + struct netlink_notify *n = ptr; + + if (event == NETLINK_URELEASE && + n->protocol == NETLINK_NETFILTER && n->pid) { + int i; + + /* destroy all instances for this pid */ + write_lock_bh(&instances_lock); + for (i = 0; i < INSTANCE_BUCKETS; i++) { + struct hlist_node *tmp, *t2; + struct nfqnl_instance *inst; + struct hlist_head *head = &instance_table[i]; + + hlist_for_each_entry_safe(inst, tmp, t2, head, hlist) { + if (n->pid == inst->peer_pid) + __instance_destroy(inst); + } + } + write_unlock_bh(&instances_lock); + } + return NOTIFY_DONE; +} + +static struct notifier_block nfqnl_rtnl_notifier = { + .notifier_call = nfqnl_rcv_nl_event, +}; + +static const int nfqa_verdict_min[NFQA_MAX] = { + [NFQA_VERDICT_HDR-1] = sizeof(struct nfqnl_msg_verdict_hdr), + [NFQA_MARK-1] = sizeof(u_int32_t), + [NFQA_PAYLOAD-1] = 0, +}; + +static int +nfqnl_recv_verdict(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *nfqa[], int *errp) +{ + struct nfgenmsg *nfmsg = NLMSG_DATA(nlh); + u_int16_t queue_num = ntohs(nfmsg->res_id); + + struct nfqnl_msg_verdict_hdr *vhdr; + struct nfqnl_instance *queue; + unsigned int verdict; + struct nfqnl_queue_entry *entry; + int err; + + if (nfattr_bad_size(nfqa, NFQA_MAX, nfqa_verdict_min)) { + QDEBUG("bad attribute size\n"); + return -EINVAL; + } + + queue = instance_lookup_get(queue_num); + if (!queue) + return -ENODEV; + + if (queue->peer_pid != NETLINK_CB(skb).pid) { + err = -EPERM; + goto err_out_put; + } + + if (!nfqa[NFQA_VERDICT_HDR-1]) { + err = -EINVAL; + goto err_out_put; + } + + vhdr = NFA_DATA(nfqa[NFQA_VERDICT_HDR-1]); + verdict = ntohl(vhdr->verdict); + + if ((verdict & NF_VERDICT_MASK) > NF_MAX_VERDICT) { + err = -EINVAL; + goto err_out_put; + } + + entry = find_dequeue_entry(queue, id_cmp, ntohl(vhdr->id)); + if (entry == NULL) { + err = -ENOENT; + goto err_out_put; + } + + if (nfqa[NFQA_PAYLOAD-1]) { + if (nfqnl_mangle(NFA_DATA(nfqa[NFQA_PAYLOAD-1]), + NFA_PAYLOAD(nfqa[NFQA_PAYLOAD-1]), entry) < 0) + verdict = NF_DROP; + } + + if (nfqa[NFQA_MARK-1]) + skb->nfmark = ntohl(*(u_int32_t *)NFA_DATA(nfqa[NFQA_MARK-1])); + + issue_verdict(entry, verdict); + instance_put(queue); + return 0; + +err_out_put: + instance_put(queue); + return err; +} + +static int +nfqnl_recv_unsupp(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *nfqa[], int *errp) +{ + return -ENOTSUPP; +} + +static const int nfqa_cfg_min[NFQA_CFG_MAX] = { + [NFQA_CFG_CMD-1] = sizeof(struct nfqnl_msg_config_cmd), + [NFQA_CFG_PARAMS-1] = sizeof(struct nfqnl_msg_config_params), +}; + +static struct nf_queue_handler nfqh = { + .name = "nf_queue", + .outfn = &nfqnl_enqueue_packet, +}; + +static int +nfqnl_recv_config(struct sock *ctnl, struct sk_buff *skb, + struct nlmsghdr *nlh, struct nfattr *nfqa[], int *errp) +{ + struct nfgenmsg *nfmsg = NLMSG_DATA(nlh); + u_int16_t queue_num = ntohs(nfmsg->res_id); + struct nfqnl_instance *queue; + int ret = 0; + + QDEBUG("entering for msg %u\n", NFNL_MSG_TYPE(nlh->nlmsg_type)); + + if (nfattr_bad_size(nfqa, NFQA_CFG_MAX, nfqa_cfg_min)) { + QDEBUG("bad attribute size\n"); + return -EINVAL; + } + + queue = instance_lookup_get(queue_num); + if (nfqa[NFQA_CFG_CMD-1]) { + struct nfqnl_msg_config_cmd *cmd; + cmd = NFA_DATA(nfqa[NFQA_CFG_CMD-1]); + QDEBUG("found CFG_CMD\n"); + + switch (cmd->command) { + case NFQNL_CFG_CMD_BIND: + if (queue) + return -EBUSY; + + queue = instance_create(queue_num, NETLINK_CB(skb).pid); + if (!queue) + return -EINVAL; + break; + case NFQNL_CFG_CMD_UNBIND: + if (!queue) + return -ENODEV; + + if (queue->peer_pid != NETLINK_CB(skb).pid) { + ret = -EPERM; + goto out_put; + } + + instance_destroy(queue); + break; + case NFQNL_CFG_CMD_PF_BIND: + QDEBUG("registering queue handler for pf=%u\n", + ntohs(cmd->pf)); + ret = nf_register_queue_handler(ntohs(cmd->pf), &nfqh); + break; + case NFQNL_CFG_CMD_PF_UNBIND: + QDEBUG("unregistering queue handler for pf=%u\n", + ntohs(cmd->pf)); + /* This is a bug and a feature. We can unregister + * other handlers(!) */ + ret = nf_unregister_queue_handler(ntohs(cmd->pf)); + break; + default: + ret = -EINVAL; + break; + } + } else { + if (!queue) { + QDEBUG("no config command, and no instance ENOENT\n"); + ret = -ENOENT; + goto out_put; + } + + if (queue->peer_pid != NETLINK_CB(skb).pid) { + QDEBUG("no config command, and wrong pid\n"); + ret = -EPERM; + goto out_put; + } + } + + if (nfqa[NFQA_CFG_PARAMS-1]) { + struct nfqnl_msg_config_params *params; + params = NFA_DATA(nfqa[NFQA_CFG_PARAMS-1]); + + nfqnl_set_mode(queue, params->copy_mode, + ntohl(params->copy_range)); + } + +out_put: + instance_put(queue); + return ret; +} + +static struct nfnl_callback nfqnl_cb[NFQNL_MSG_MAX] = { + [NFQNL_MSG_PACKET] = { .call = nfqnl_recv_unsupp, + .attr_count = NFQA_MAX, + .cap_required = CAP_NET_ADMIN }, + [NFQNL_MSG_VERDICT] = { .call = nfqnl_recv_verdict, + .attr_count = NFQA_MAX, + .cap_required = CAP_NET_ADMIN }, + [NFQNL_MSG_CONFIG] = { .call = nfqnl_recv_config, + .attr_count = NFQA_CFG_MAX, + .cap_required = CAP_NET_ADMIN }, +}; + +static struct nfnetlink_subsystem nfqnl_subsys = { + .name = "nf_queue", + .subsys_id = NFNL_SUBSYS_QUEUE, + .cb_count = NFQNL_MSG_MAX, + .cb = nfqnl_cb, +}; + +#ifdef CONFIG_PROC_FS +struct iter_state { + unsigned int bucket; +}; + +static struct hlist_node *get_first(struct seq_file *seq) +{ + struct iter_state *st = seq->private; + + if (!st) + return NULL; + + for (st->bucket = 0; st->bucket < INSTANCE_BUCKETS; st->bucket++) { + if (!hlist_empty(&instance_table[st->bucket])) + return instance_table[st->bucket].first; + } + return NULL; +} + +static struct hlist_node *get_next(struct seq_file *seq, struct hlist_node *h) +{ + struct iter_state *st = seq->private; + + h = h->next; + while (!h) { + if (++st->bucket >= INSTANCE_BUCKETS) + return NULL; + + h = instance_table[st->bucket].first; + } + return h; +} + +static struct hlist_node *get_idx(struct seq_file *seq, loff_t pos) +{ + struct hlist_node *head; + head = get_first(seq); + + if (head) + while (pos && (head = get_next(seq, head))) + pos--; + return pos ? NULL : head; +} + +static void *seq_start(struct seq_file *seq, loff_t *pos) +{ + read_lock_bh(&instances_lock); + return get_idx(seq, *pos); +} + +static void *seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + (*pos)++; + return get_next(s, v); +} + +static void seq_stop(struct seq_file *s, void *v) +{ + read_unlock_bh(&instances_lock); +} + +static int seq_show(struct seq_file *s, void *v) +{ + const struct nfqnl_instance *inst = v; + + return seq_printf(s, "%5d %6d %5d %1d %5d %5d %5d %8d %2d\n", + inst->queue_num, + inst->peer_pid, inst->queue_total, + inst->copy_mode, inst->copy_range, + inst->queue_dropped, inst->queue_user_dropped, + atomic_read(&inst->id_sequence), + atomic_read(&inst->use)); +} + +static struct seq_operations nfqnl_seq_ops = { + .start = seq_start, + .next = seq_next, + .stop = seq_stop, + .show = seq_show, +}; + +static int nfqnl_open(struct inode *inode, struct file *file) +{ + struct seq_file *seq; + struct iter_state *is; + int ret; + + is = kmalloc(sizeof(*is), GFP_KERNEL); + if (!is) + return -ENOMEM; + memset(is, 0, sizeof(*is)); + ret = seq_open(file, &nfqnl_seq_ops); + if (ret < 0) + goto out_free; + seq = file->private_data; + seq->private = is; + return ret; +out_free: + kfree(is); + return ret; +} + +static struct file_operations nfqnl_file_ops = { + .owner = THIS_MODULE, + .open = nfqnl_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +#endif /* PROC_FS */ + +static int +init_or_cleanup(int init) +{ + int i, status = -ENOMEM; +#ifdef CONFIG_PROC_FS + struct proc_dir_entry *proc_nfqueue; +#endif + + if (!init) + goto cleanup; + + for (i = 0; i < INSTANCE_BUCKETS; i++) + INIT_HLIST_HEAD(&instance_table[i]); + + netlink_register_notifier(&nfqnl_rtnl_notifier); + status = nfnetlink_subsys_register(&nfqnl_subsys); + if (status < 0) { + printk(KERN_ERR "nf_queue: failed to create netlink socket\n"); + goto cleanup_netlink_notifier; + } + +#ifdef CONFIG_PROC_FS + proc_nfqueue = create_proc_entry("nfnetlink_queue", 0440, + proc_net_netfilter); + if (!proc_nfqueue) + goto cleanup_subsys; + proc_nfqueue->proc_fops = &nfqnl_file_ops; +#endif + + register_netdevice_notifier(&nfqnl_dev_notifier); + + return status; + +cleanup: + nf_unregister_queue_handlers(&nfqh); + unregister_netdevice_notifier(&nfqnl_dev_notifier); +#ifdef CONFIG_PROC_FS + remove_proc_entry("nfnetlink_queue", proc_net_netfilter); +cleanup_subsys: +#endif + nfnetlink_subsys_unregister(&nfqnl_subsys); +cleanup_netlink_notifier: + netlink_unregister_notifier(&nfqnl_rtnl_notifier); + return status; +} + +static int __init init(void) +{ + + return init_or_cleanup(1); +} + +static void __exit fini(void) +{ + init_or_cleanup(0); +} + +MODULE_DESCRIPTION("netfilter packet queue handler"); +MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS_NFNL_SUBSYS(NFNL_SUBSYS_QUEUE); + +module_init(init); +module_exit(fini); diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index ff774a06c89d..62435ffc6184 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -13,7 +13,12 @@ * added netlink_proto_exit * Tue Jan 22 18:32:44 BRST 2002 Arnaldo C. de Melo <acme@conectiva.com.br> * use nlk_sk, as sk->protinfo is on a diet 8) - * + * Fri Jul 22 19:51:12 MEST 2005 Harald Welte <laforge@gnumonks.org> + * - inc module use count of module that owns + * the kernel socket in case userspace opens + * socket of same protocol + * - remove all module support, since netlink is + * mandatory if CONFIG_NET=y these days */ #include <linux/config.h> @@ -55,21 +60,29 @@ #include <net/scm.h> #define Nprintk(a...) +#define NLGRPSZ(x) (ALIGN(x, sizeof(unsigned long) * 8) / 8) struct netlink_sock { /* struct sock has to be the first member of netlink_sock */ struct sock sk; u32 pid; - unsigned int groups; u32 dst_pid; - unsigned int dst_groups; + u32 dst_group; + u32 flags; + u32 subscriptions; + u32 ngroups; + unsigned long *groups; unsigned long state; wait_queue_head_t wait; struct netlink_callback *cb; spinlock_t cb_lock; void (*data_ready)(struct sock *sk, int bytes); + struct module *module; }; +#define NETLINK_KERNEL_SOCKET 0x1 +#define NETLINK_RECV_PKTINFO 0x2 + static inline struct netlink_sock *nlk_sk(struct sock *sk) { return (struct netlink_sock *)sk; @@ -92,6 +105,9 @@ struct netlink_table { struct nl_pid_hash hash; struct hlist_head mc_list; unsigned int nl_nonroot; + unsigned int groups; + struct module *module; + int registered; }; static struct netlink_table *nl_table; @@ -106,6 +122,11 @@ static atomic_t nl_table_users = ATOMIC_INIT(0); static struct notifier_block *netlink_chain; +static u32 netlink_group_mask(u32 group) +{ + return group ? 1 << (group - 1) : 0; +} + static struct hlist_head *nl_pid_hashfn(struct nl_pid_hash *hash, u32 pid) { return &hash->table[jhash_1word(pid, hash->rnd) & hash->mask]; @@ -122,6 +143,7 @@ static void netlink_sock_destruct(struct sock *sk) BUG_TRAP(!atomic_read(&sk->sk_rmem_alloc)); BUG_TRAP(!atomic_read(&sk->sk_wmem_alloc)); BUG_TRAP(!nlk_sk(sk)->cb); + BUG_TRAP(!nlk_sk(sk)->groups); } /* This lock without WQ_FLAG_EXCLUSIVE is good on UP and it is _very_ bad on SMP. @@ -317,7 +339,7 @@ static void netlink_remove(struct sock *sk) netlink_table_grab(); if (sk_del_node_init(sk)) nl_table[sk->sk_protocol].hash.entries--; - if (nlk_sk(sk)->groups) + if (nlk_sk(sk)->subscriptions) __sk_del_bind_node(sk); netlink_table_ungrab(); } @@ -328,19 +350,11 @@ static struct proto netlink_proto = { .obj_size = sizeof(struct netlink_sock), }; -static int netlink_create(struct socket *sock, int protocol) +static int __netlink_create(struct socket *sock, int protocol) { struct sock *sk; struct netlink_sock *nlk; - sock->state = SS_UNCONNECTED; - - if (sock->type != SOCK_RAW && sock->type != SOCK_DGRAM) - return -ESOCKTNOSUPPORT; - - if (protocol<0 || protocol >= MAX_LINKS) - return -EPROTONOSUPPORT; - sock->ops = &netlink_ops; sk = sk_alloc(PF_NETLINK, GFP_KERNEL, &netlink_proto, 1); @@ -350,15 +364,67 @@ static int netlink_create(struct socket *sock, int protocol) sock_init_data(sock, sk); nlk = nlk_sk(sk); - spin_lock_init(&nlk->cb_lock); init_waitqueue_head(&nlk->wait); - sk->sk_destruct = netlink_sock_destruct; + sk->sk_destruct = netlink_sock_destruct; sk->sk_protocol = protocol; return 0; } +static int netlink_create(struct socket *sock, int protocol) +{ + struct module *module = NULL; + struct netlink_sock *nlk; + unsigned int groups; + int err = 0; + + sock->state = SS_UNCONNECTED; + + if (sock->type != SOCK_RAW && sock->type != SOCK_DGRAM) + return -ESOCKTNOSUPPORT; + + if (protocol<0 || protocol >= MAX_LINKS) + return -EPROTONOSUPPORT; + + netlink_lock_table(); +#ifdef CONFIG_KMOD + if (!nl_table[protocol].registered) { + netlink_unlock_table(); + request_module("net-pf-%d-proto-%d", PF_NETLINK, protocol); + netlink_lock_table(); + } +#endif + if (nl_table[protocol].registered && + try_module_get(nl_table[protocol].module)) + module = nl_table[protocol].module; + else + err = -EPROTONOSUPPORT; + groups = nl_table[protocol].groups; + netlink_unlock_table(); + + if (err || (err = __netlink_create(sock, protocol) < 0)) + goto out_module; + + nlk = nlk_sk(sock->sk); + + nlk->groups = kmalloc(NLGRPSZ(groups), GFP_KERNEL); + if (nlk->groups == NULL) { + err = -ENOMEM; + goto out_module; + } + memset(nlk->groups, 0, NLGRPSZ(groups)); + nlk->ngroups = groups; + + nlk->module = module; +out: + return err; + +out_module: + module_put(module); + goto out; +} + static int netlink_release(struct socket *sock) { struct sock *sk = sock->sk; @@ -387,14 +453,27 @@ static int netlink_release(struct socket *sock) skb_queue_purge(&sk->sk_write_queue); - if (nlk->pid && !nlk->groups) { + if (nlk->pid && !nlk->subscriptions) { struct netlink_notify n = { .protocol = sk->sk_protocol, .pid = nlk->pid, }; notifier_call_chain(&netlink_chain, NETLINK_URELEASE, &n); } - + + if (nlk->module) + module_put(nlk->module); + + if (nlk->flags & NETLINK_KERNEL_SOCKET) { + netlink_table_grab(); + nl_table[sk->sk_protocol].module = NULL; + nl_table[sk->sk_protocol].registered = 0; + netlink_table_ungrab(); + } + + kfree(nlk->groups); + nlk->groups = NULL; + sock_put(sk); return 0; } @@ -443,6 +522,18 @@ static inline int netlink_capable(struct socket *sock, unsigned int flag) capable(CAP_NET_ADMIN); } +static void +netlink_update_subscriptions(struct sock *sk, unsigned int subscriptions) +{ + struct netlink_sock *nlk = nlk_sk(sk); + + if (nlk->subscriptions && !subscriptions) + __sk_del_bind_node(sk); + else if (!nlk->subscriptions && subscriptions) + sk_add_bind_node(sk, &nl_table[sk->sk_protocol].mc_list); + nlk->subscriptions = subscriptions; +} + static int netlink_bind(struct socket *sock, struct sockaddr *addr, int addr_len) { struct sock *sk = sock->sk; @@ -468,15 +559,14 @@ static int netlink_bind(struct socket *sock, struct sockaddr *addr, int addr_len return err; } - if (!nladdr->nl_groups && !nlk->groups) + if (!nladdr->nl_groups && !(u32)nlk->groups[0]) return 0; netlink_table_grab(); - if (nlk->groups && !nladdr->nl_groups) - __sk_del_bind_node(sk); - else if (!nlk->groups && nladdr->nl_groups) - sk_add_bind_node(sk, &nl_table[sk->sk_protocol].mc_list); - nlk->groups = nladdr->nl_groups; + netlink_update_subscriptions(sk, nlk->subscriptions + + hweight32(nladdr->nl_groups) - + hweight32(nlk->groups[0])); + nlk->groups[0] = (nlk->groups[0] & ~0xffffffffUL) | nladdr->nl_groups; netlink_table_ungrab(); return 0; @@ -493,7 +583,7 @@ static int netlink_connect(struct socket *sock, struct sockaddr *addr, if (addr->sa_family == AF_UNSPEC) { sk->sk_state = NETLINK_UNCONNECTED; nlk->dst_pid = 0; - nlk->dst_groups = 0; + nlk->dst_group = 0; return 0; } if (addr->sa_family != AF_NETLINK) @@ -509,7 +599,7 @@ static int netlink_connect(struct socket *sock, struct sockaddr *addr, if (err == 0) { sk->sk_state = NETLINK_CONNECTED; nlk->dst_pid = nladdr->nl_pid; - nlk->dst_groups = nladdr->nl_groups; + nlk->dst_group = ffs(nladdr->nl_groups); } return err; @@ -527,10 +617,10 @@ static int netlink_getname(struct socket *sock, struct sockaddr *addr, int *addr if (peer) { nladdr->nl_pid = nlk->dst_pid; - nladdr->nl_groups = nlk->dst_groups; + nladdr->nl_groups = netlink_group_mask(nlk->dst_group); } else { nladdr->nl_pid = nlk->pid; - nladdr->nl_groups = nlk->groups; + nladdr->nl_groups = nlk->groups[0]; } return 0; } @@ -731,7 +821,8 @@ static inline int do_one_broadcast(struct sock *sk, if (p->exclude_sk == sk) goto out; - if (nlk->pid == p->pid || !(nlk->groups & p->group)) + if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups || + !test_bit(p->group - 1, nlk->groups)) goto out; if (p->failure) { @@ -770,7 +861,7 @@ out: } int netlink_broadcast(struct sock *ssk, struct sk_buff *skb, u32 pid, - u32 group, int allocation) + u32 group, unsigned int __nocast allocation) { struct netlink_broadcast_data info; struct hlist_node *node; @@ -827,7 +918,8 @@ static inline int do_one_set_err(struct sock *sk, if (sk == p->exclude_sk) goto out; - if (nlk->pid == p->pid || !(nlk->groups & p->group)) + if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups || + !test_bit(p->group - 1, nlk->groups)) goto out; sk->sk_err = p->code; @@ -855,6 +947,94 @@ void netlink_set_err(struct sock *ssk, u32 pid, u32 group, int code) read_unlock(&nl_table_lock); } +static int netlink_setsockopt(struct socket *sock, int level, int optname, + char __user *optval, int optlen) +{ + struct sock *sk = sock->sk; + struct netlink_sock *nlk = nlk_sk(sk); + int val = 0, err; + + if (level != SOL_NETLINK) + return -ENOPROTOOPT; + + if (optlen >= sizeof(int) && + get_user(val, (int __user *)optval)) + return -EFAULT; + + switch (optname) { + case NETLINK_PKTINFO: + if (val) + nlk->flags |= NETLINK_RECV_PKTINFO; + else + nlk->flags &= ~NETLINK_RECV_PKTINFO; + err = 0; + break; + case NETLINK_ADD_MEMBERSHIP: + case NETLINK_DROP_MEMBERSHIP: { + unsigned int subscriptions; + int old, new = optname == NETLINK_ADD_MEMBERSHIP ? 1 : 0; + + if (!netlink_capable(sock, NL_NONROOT_RECV)) + return -EPERM; + if (!val || val - 1 >= nlk->ngroups) + return -EINVAL; + netlink_table_grab(); + old = test_bit(val - 1, nlk->groups); + subscriptions = nlk->subscriptions - old + new; + if (new) + __set_bit(val - 1, nlk->groups); + else + __clear_bit(val - 1, nlk->groups); + netlink_update_subscriptions(sk, subscriptions); + netlink_table_ungrab(); + err = 0; + break; + } + default: + err = -ENOPROTOOPT; + } + return err; +} + +static int netlink_getsockopt(struct socket *sock, int level, int optname, + char __user *optval, int __user *optlen) +{ + struct sock *sk = sock->sk; + struct netlink_sock *nlk = nlk_sk(sk); + int len, val, err; + + if (level != SOL_NETLINK) + return -ENOPROTOOPT; + + if (get_user(len, optlen)) + return -EFAULT; + if (len < 0) + return -EINVAL; + + switch (optname) { + case NETLINK_PKTINFO: + if (len < sizeof(int)) + return -EINVAL; + len = sizeof(int); + val = nlk->flags & NETLINK_RECV_PKTINFO ? 1 : 0; + put_user(len, optlen); + put_user(val, optval); + err = 0; + break; + default: + err = -ENOPROTOOPT; + } + return err; +} + +static void netlink_cmsg_recv_pktinfo(struct msghdr *msg, struct sk_buff *skb) +{ + struct nl_pktinfo info; + + info.group = NETLINK_CB(skb).dst_group; + put_cmsg(msg, SOL_NETLINK, NETLINK_PKTINFO, sizeof(info), &info); +} + static inline void netlink_rcv_wake(struct sock *sk) { struct netlink_sock *nlk = nlk_sk(sk); @@ -873,7 +1053,7 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock, struct netlink_sock *nlk = nlk_sk(sk); struct sockaddr_nl *addr=msg->msg_name; u32 dst_pid; - u32 dst_groups; + u32 dst_group; struct sk_buff *skb; int err; struct scm_cookie scm; @@ -891,12 +1071,12 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock, if (addr->nl_family != AF_NETLINK) return -EINVAL; dst_pid = addr->nl_pid; - dst_groups = addr->nl_groups; - if (dst_groups && !netlink_capable(sock, NL_NONROOT_SEND)) + dst_group = ffs(addr->nl_groups); + if (dst_group && !netlink_capable(sock, NL_NONROOT_SEND)) return -EPERM; } else { dst_pid = nlk->dst_pid; - dst_groups = nlk->dst_groups; + dst_group = nlk->dst_group; } if (!nlk->pid) { @@ -914,9 +1094,8 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock, goto out; NETLINK_CB(skb).pid = nlk->pid; - NETLINK_CB(skb).groups = nlk->groups; NETLINK_CB(skb).dst_pid = dst_pid; - NETLINK_CB(skb).dst_groups = dst_groups; + NETLINK_CB(skb).dst_group = dst_group; NETLINK_CB(skb).loginuid = audit_get_loginuid(current->audit_context); memcpy(NETLINK_CREDS(skb), &siocb->scm->creds, sizeof(struct ucred)); @@ -938,9 +1117,9 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock, goto out; } - if (dst_groups) { + if (dst_group) { atomic_inc(&skb->users); - netlink_broadcast(sk, skb, dst_pid, dst_groups, GFP_KERNEL); + netlink_broadcast(sk, skb, dst_pid, dst_group, GFP_KERNEL); } err = netlink_unicast(sk, skb, dst_pid, msg->msg_flags&MSG_DONTWAIT); @@ -986,7 +1165,7 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock, addr->nl_family = AF_NETLINK; addr->nl_pad = 0; addr->nl_pid = NETLINK_CB(skb).pid; - addr->nl_groups = NETLINK_CB(skb).dst_groups; + addr->nl_groups = netlink_group_mask(NETLINK_CB(skb).dst_group); msg->msg_namelen = sizeof(*addr); } @@ -1001,6 +1180,8 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock, netlink_dump(sk); scm_recv(sock, msg, siocb->scm, flags); + if (nlk->flags & NETLINK_RECV_PKTINFO) + netlink_cmsg_recv_pktinfo(msg, skb); out: netlink_rcv_wake(sk); @@ -1023,10 +1204,13 @@ static void netlink_data_ready(struct sock *sk, int len) */ struct sock * -netlink_kernel_create(int unit, void (*input)(struct sock *sk, int len)) +netlink_kernel_create(int unit, unsigned int groups, + void (*input)(struct sock *sk, int len), + struct module *module) { struct socket *sock; struct sock *sk; + struct netlink_sock *nlk; if (!nl_table) return NULL; @@ -1037,20 +1221,31 @@ netlink_kernel_create(int unit, void (*input)(struct sock *sk, int len)) if (sock_create_lite(PF_NETLINK, SOCK_DGRAM, unit, &sock)) return NULL; - if (netlink_create(sock, unit) < 0) { - sock_release(sock); - return NULL; - } + if (__netlink_create(sock, unit) < 0) + goto out_sock_release; + sk = sock->sk; sk->sk_data_ready = netlink_data_ready; if (input) nlk_sk(sk)->data_ready = input; - if (netlink_insert(sk, 0)) { - sock_release(sock); - return NULL; - } + if (netlink_insert(sk, 0)) + goto out_sock_release; + + nlk = nlk_sk(sk); + nlk->flags |= NETLINK_KERNEL_SOCKET; + + netlink_table_grab(); + nl_table[unit].groups = groups < 32 ? 32 : groups; + nl_table[unit].module = module; + nl_table[unit].registered = 1; + netlink_table_ungrab(); + return sk; + +out_sock_release: + sock_release(sock); + return NULL; } void netlink_set_nonroot(int protocol, unsigned int flags) @@ -1288,7 +1483,8 @@ static int netlink_seq_show(struct seq_file *seq, void *v) s, s->sk_protocol, nlk->pid, - nlk->groups, + nlk->flags & NETLINK_KERNEL_SOCKET ? + 0 : (unsigned int)nlk->groups[0], atomic_read(&s->sk_rmem_alloc), atomic_read(&s->sk_wmem_alloc), nlk->cb, @@ -1362,8 +1558,8 @@ static struct proto_ops netlink_ops = { .ioctl = sock_no_ioctl, .listen = sock_no_listen, .shutdown = sock_no_shutdown, - .setsockopt = sock_no_setsockopt, - .getsockopt = sock_no_getsockopt, + .setsockopt = netlink_setsockopt, + .getsockopt = netlink_getsockopt, .sendmsg = netlink_sendmsg, .recvmsg = netlink_recvmsg, .mmap = sock_no_mmap, @@ -1438,21 +1634,7 @@ out: return err; } -static void __exit netlink_proto_exit(void) -{ - sock_unregister(PF_NETLINK); - proc_net_remove("netlink"); - kfree(nl_table); - nl_table = NULL; - proto_unregister(&netlink_proto); -} - core_initcall(netlink_proto_init); -module_exit(netlink_proto_exit); - -MODULE_LICENSE("GPL"); - -MODULE_ALIAS_NETPROTO(PF_NETLINK); EXPORT_SYMBOL(netlink_ack); EXPORT_SYMBOL(netlink_broadcast); diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c index 162a85fed150..4b53de982114 100644 --- a/net/netrom/af_netrom.c +++ b/net/netrom/af_netrom.c @@ -39,7 +39,7 @@ #include <linux/proc_fs.h> #include <linux/seq_file.h> #include <net/ip.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/arp.h> #include <linux/init.h> @@ -858,17 +858,16 @@ int nr_rx_frame(struct sk_buff *skb, struct net_device *dev) frametype = skb->data[19] & 0x0F; flags = skb->data[19] & 0xF0; -#ifdef CONFIG_INET /* * Check for an incoming IP over NET/ROM frame. */ - if (frametype == NR_PROTOEXT && circuit_index == NR_PROTO_IP && circuit_id == NR_PROTO_IP) { + if (frametype == NR_PROTOEXT && + circuit_index == NR_PROTO_IP && circuit_id == NR_PROTO_IP) { skb_pull(skb, NR_NETWORK_LEN + NR_TRANSPORT_LEN); skb->h.raw = skb->data; return nr_rx_ip(skb, dev); } -#endif /* * Find an existing socket connection, based on circuit ID, if it's diff --git a/net/netrom/nr_dev.c b/net/netrom/nr_dev.c index 220bf7494f71..263da4c26494 100644 --- a/net/netrom/nr_dev.c +++ b/net/netrom/nr_dev.c @@ -38,8 +38,6 @@ #include <net/ax25.h> #include <net/netrom.h> -#ifdef CONFIG_INET - /* * Only allow IP over NET/ROM frames through if the netrom device is up. */ @@ -64,11 +62,12 @@ int nr_rx_ip(struct sk_buff *skb, struct net_device *dev) skb->nh.raw = skb->data; skb->pkt_type = PACKET_HOST; - ip_rcv(skb, skb->dev, NULL); + netif_rx(skb); return 1; } +#ifdef CONFIG_INET static int nr_rebuild_header(struct sk_buff *skb) { diff --git a/net/netrom/nr_in.c b/net/netrom/nr_in.c index 9c44b3794126..64b81a796907 100644 --- a/net/netrom/nr_in.c +++ b/net/netrom/nr_in.c @@ -22,8 +22,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> -#include <net/ip.h> /* For ip_rcv */ +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> diff --git a/net/netrom/nr_subr.c b/net/netrom/nr_subr.c index 0627347b14b8..587bed2674bf 100644 --- a/net/netrom/nr_subr.c +++ b/net/netrom/nr_subr.c @@ -21,7 +21,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> @@ -77,7 +77,7 @@ void nr_requeue_frames(struct sock *sk) if (skb_prev == NULL) skb_queue_head(&sk->sk_write_queue, skb); else - skb_append(skb_prev, skb); + skb_append(skb_prev, skb, &sk->sk_write_queue); skb_prev = skb; } } diff --git a/net/netrom/nr_timer.c b/net/netrom/nr_timer.c index faabda8088be..75b72d389ba9 100644 --- a/net/netrom/nr_timer.c +++ b/net/netrom/nr_timer.c @@ -22,7 +22,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/system.h> #include <linux/fcntl.h> diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index c9d5980aa4de..ba997095f08f 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -241,7 +241,7 @@ static struct proto_ops packet_ops; #ifdef CONFIG_SOCK_PACKET static struct proto_ops packet_ops_spkt; -static int packet_rcv_spkt(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +static int packet_rcv_spkt(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct sock *sk; struct sockaddr_pkt *spkt; @@ -441,7 +441,7 @@ static inline unsigned run_filter(struct sk_buff *skb, struct sock *sk, unsigned we will not harm anyone. */ -static int packet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +static int packet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct sock *sk; struct sockaddr_ll *sll; @@ -546,7 +546,7 @@ drop: } #ifdef CONFIG_PACKET_MMAP -static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt) +static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { struct sock *sk; struct packet_sock *po; @@ -635,12 +635,12 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct pack h->tp_snaplen = snaplen; h->tp_mac = macoff; h->tp_net = netoff; - if (skb->stamp.tv_sec == 0) { - do_gettimeofday(&skb->stamp); + if (skb->tstamp.off_sec == 0) { + __net_timestamp(skb); sock_enable_timestamp(sk); } - h->tp_sec = skb->stamp.tv_sec; - h->tp_usec = skb->stamp.tv_usec; + h->tp_sec = skb_tv_base.tv_sec + skb->tstamp.off_sec; + h->tp_usec = skb_tv_base.tv_usec + skb->tstamp.off_usec; sll = (struct sockaddr_ll*)((u8*)h + TPACKET_ALIGN(sizeof(*h))); sll->sll_halen = 0; diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c index 5480caf8ccc2..c6e59f84c3ae 100644 --- a/net/rose/af_rose.c +++ b/net/rose/af_rose.c @@ -41,7 +41,7 @@ #include <net/rose.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/ip.h> #include <net/arp.h> diff --git a/net/rose/rose_in.c b/net/rose/rose_in.c index ef475a1bb1ba..8348d33f1efe 100644 --- a/net/rose/rose_in.c +++ b/net/rose/rose_in.c @@ -26,8 +26,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/ip.h> /* For ip_rcv */ -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/system.h> #include <linux/fcntl.h> #include <linux/mm.h> diff --git a/net/rose/rose_route.c b/net/rose/rose_route.c index 25da6f699fd0..4510cd7613ec 100644 --- a/net/rose/rose_route.c +++ b/net/rose/rose_route.c @@ -24,7 +24,7 @@ #include <linux/if_arp.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/system.h> #include <asm/uaccess.h> #include <linux/fcntl.h> diff --git a/net/rose/rose_subr.c b/net/rose/rose_subr.c index 7db7e1cedc3a..a29a3a960fd6 100644 --- a/net/rose/rose_subr.c +++ b/net/rose/rose_subr.c @@ -21,7 +21,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/system.h> #include <linux/fcntl.h> #include <linux/mm.h> @@ -74,7 +74,7 @@ void rose_requeue_frames(struct sock *sk) if (skb_prev == NULL) skb_queue_head(&sk->sk_write_queue, skb); else - skb_append(skb_prev, skb); + skb_append(skb_prev, skb, &sk->sk_write_queue); skb_prev = skb; } } diff --git a/net/rose/rose_timer.c b/net/rose/rose_timer.c index 84dd4403f792..50ae0371dab8 100644 --- a/net/rose/rose_timer.c +++ b/net/rose/rose_timer.c @@ -22,7 +22,7 @@ #include <linux/netdevice.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/system.h> #include <linux/fcntl.h> #include <linux/mm.h> diff --git a/net/rxrpc/transport.c b/net/rxrpc/transport.c index 9bce7794130a..122c086ee2db 100644 --- a/net/rxrpc/transport.c +++ b/net/rxrpc/transport.c @@ -330,7 +330,7 @@ static int rxrpc_incoming_msg(struct rxrpc_transport *trans, msg->trans = trans; msg->state = RXRPC_MSG_RECEIVED; - msg->stamp = pkt->stamp; + skb_get_timestamp(pkt, &msg->stamp); if (msg->stamp.tv_sec == 0) { do_gettimeofday(&msg->stamp); if (pkt->sk) diff --git a/net/sched/Kconfig b/net/sched/Kconfig index 59d3e71f8b85..45d3bc0812c8 100644 --- a/net/sched/Kconfig +++ b/net/sched/Kconfig @@ -491,6 +491,7 @@ config NET_EMATCH_TEXT depends on NET_EMATCH select TEXTSEARCH select TEXTSEARCH_KMP + select TEXTSEARCH_BM select TEXTSEARCH_FSM ---help--- Say Y here if you want to be ablt to classify packets based on diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 249c61936ea0..8aebe8f6d271 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -165,7 +165,7 @@ int tcf_action_exec(struct sk_buff *skb, struct tc_action *act, while ((a = act) != NULL) { repeat: if (a->ops && a->ops->act) { - ret = a->ops->act(&skb, a); + ret = a->ops->act(&skb, a, res); if (TC_MUNGED & skb->tc_verd) { /* copied already, allow trampling */ skb->tc_verd = SET_TC_OK2MUNGE(skb->tc_verd); @@ -179,11 +179,6 @@ repeat: act = a->next; } exec_done: - if (skb->tc_classid > 0) { - res->classid = skb->tc_classid; - res->class = 0; - skb->tc_classid = 0; - } return ret; } @@ -598,7 +593,7 @@ static int tca_action_flush(struct rtattr *rta, struct nlmsghdr *n, u32 pid) nlh->nlmsg_flags |= NLM_F_ROOT; module_put(a->ops->owner); kfree(a); - err = rtnetlink_send(skb, pid, RTMGRP_TC, n->nlmsg_flags&NLM_F_ECHO); + err = rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO); if (err > 0) return 0; @@ -661,7 +656,7 @@ tca_action_gd(struct rtattr *rta, struct nlmsghdr *n, u32 pid, int event) /* now do the delete */ tcf_action_destroy(head, 0); - ret = rtnetlink_send(skb, pid, RTMGRP_TC, + ret = rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO); if (ret > 0) return 0; @@ -703,9 +698,9 @@ static int tcf_add_notify(struct tc_action *a, u32 pid, u32 seq, int event, x->rta_len = skb->tail - (u8*)x; nlh->nlmsg_len = skb->tail - b; - NETLINK_CB(skb).dst_groups = RTMGRP_TC; + NETLINK_CB(skb).dst_group = RTNLGRP_TC; - err = rtnetlink_send(skb, pid, RTMGRP_TC, flags&NLM_F_ECHO); + err = rtnetlink_send(skb, pid, RTNLGRP_TC, flags&NLM_F_ECHO); if (err > 0) err = 0; return err; diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index 3b5714ef4d1a..b4d89fbb3782 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -367,7 +367,7 @@ static int tfilter_notify(struct sk_buff *oskb, struct nlmsghdr *n, return -EINVAL; } - return rtnetlink_send(skb, pid, RTMGRP_TC, n->nlmsg_flags&NLM_F_ECHO); + return rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO); } struct tcf_dump_args diff --git a/net/sched/gact.c b/net/sched/gact.c index a811c89fef7f..d1c6d542912a 100644 --- a/net/sched/gact.c +++ b/net/sched/gact.c @@ -135,7 +135,7 @@ tcf_gact_cleanup(struct tc_action *a, int bind) } static int -tcf_gact(struct sk_buff **pskb, struct tc_action *a) +tcf_gact(struct sk_buff **pskb, struct tc_action *a, struct tcf_result *res) { struct tcf_gact *p = PRIV(a, gact); struct sk_buff *skb = *pskb; diff --git a/net/sched/ipt.c b/net/sched/ipt.c index b114d994d523..f50136eed211 100644 --- a/net/sched/ipt.c +++ b/net/sched/ipt.c @@ -201,7 +201,7 @@ tcf_ipt_cleanup(struct tc_action *a, int bind) } static int -tcf_ipt(struct sk_buff **pskb, struct tc_action *a) +tcf_ipt(struct sk_buff **pskb, struct tc_action *a, struct tcf_result *res) { int ret = 0, result = 0; struct tcf_ipt *p = PRIV(a, ipt); diff --git a/net/sched/mirred.c b/net/sched/mirred.c index f309ce336803..20d06916dc0b 100644 --- a/net/sched/mirred.c +++ b/net/sched/mirred.c @@ -158,7 +158,7 @@ tcf_mirred_cleanup(struct tc_action *a, int bind) } static int -tcf_mirred(struct sk_buff **pskb, struct tc_action *a) +tcf_mirred(struct sk_buff **pskb, struct tc_action *a, struct tcf_result *res) { struct tcf_mirred *p = PRIV(a, mirred); struct net_device *dev; diff --git a/net/sched/pedit.c b/net/sched/pedit.c index 678be6a645fb..767d24f4610e 100644 --- a/net/sched/pedit.c +++ b/net/sched/pedit.c @@ -130,7 +130,7 @@ tcf_pedit_cleanup(struct tc_action *a, int bind) } static int -tcf_pedit(struct sk_buff **pskb, struct tc_action *a) +tcf_pedit(struct sk_buff **pskb, struct tc_action *a, struct tcf_result *res) { struct tcf_pedit *p = PRIV(a, pedit); struct sk_buff *skb = *pskb; diff --git a/net/sched/police.c b/net/sched/police.c index c03545faf523..eb39fb2f39b6 100644 --- a/net/sched/police.c +++ b/net/sched/police.c @@ -284,7 +284,8 @@ static int tcf_act_police_cleanup(struct tc_action *a, int bind) return 0; } -static int tcf_act_police(struct sk_buff **pskb, struct tc_action *a) +static int tcf_act_police(struct sk_buff **pskb, struct tc_action *a, + struct tcf_result *res) { psched_time_t now; struct sk_buff *skb = *pskb; diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index b9a069af4a02..737681cb9a92 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -816,7 +816,7 @@ static int qdisc_notify(struct sk_buff *oskb, struct nlmsghdr *n, } if (skb->len) - return rtnetlink_send(skb, pid, RTMGRP_TC, n->nlmsg_flags&NLM_F_ECHO); + return rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO); err_out: kfree_skb(skb); @@ -1040,7 +1040,7 @@ static int tclass_notify(struct sk_buff *oskb, struct nlmsghdr *n, return -EINVAL; } - return rtnetlink_send(skb, pid, RTMGRP_TC, n->nlmsg_flags&NLM_F_ECHO); + return rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO); } struct qdisc_dump_args diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 0d066c965342..99ceb91f0150 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -238,6 +238,20 @@ static void dev_watchdog_down(struct net_device *dev) spin_unlock_bh(&dev->xmit_lock); } +void netif_carrier_on(struct net_device *dev) +{ + if (test_and_clear_bit(__LINK_STATE_NOCARRIER, &dev->state)) + linkwatch_fire_event(dev); + if (netif_running(dev)) + __netdev_watchdog_up(dev); +} + +void netif_carrier_off(struct net_device *dev) +{ + if (!test_and_set_bit(__LINK_STATE_NOCARRIER, &dev->state)) + linkwatch_fire_event(dev); +} + /* "NOOP" scheduler: the best scheduler, recommended for all interfaces under all circumstances. It is difficult to invent anything faster or cheaper. @@ -600,6 +614,8 @@ void dev_shutdown(struct net_device *dev) } EXPORT_SYMBOL(__netdev_watchdog_up); +EXPORT_SYMBOL(netif_carrier_on); +EXPORT_SYMBOL(netif_carrier_off); EXPORT_SYMBOL(noop_qdisc); EXPORT_SYMBOL(noop_qdisc_ops); EXPORT_SYMBOL(qdisc_create_dflt); diff --git a/net/sched/simple.c b/net/sched/simple.c index 3ab4c675ab5d..8a6ae4f491e8 100644 --- a/net/sched/simple.c +++ b/net/sched/simple.c @@ -44,7 +44,7 @@ static DEFINE_RWLOCK(simp_lock); #include <net/pkt_act.h> #include <net/act_generic.h> -static int tcf_simp(struct sk_buff **pskb, struct tc_action *a) +static int tcf_simp(struct sk_buff **pskb, struct tc_action *a, struct tcf_result *res) { struct sk_buff *skb = *pskb; struct tcf_defact *p = PRIV(a, defact); diff --git a/net/sctp/input.c b/net/sctp/input.c index 742be9171b7d..28f32243397f 100644 --- a/net/sctp/input.c +++ b/net/sctp/input.c @@ -236,8 +236,8 @@ int sctp_rcv(struct sk_buff *skb) } /* SCTP seems to always need a timestamp right now (FIXME) */ - if (skb->stamp.tv_sec == 0) { - do_gettimeofday(&skb->stamp); + if (skb->tstamp.off_sec == 0) { + __net_timestamp(skb); sock_enable_timestamp(sk); } diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index e9b2fd480d61..fa3be2b8fb5f 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -66,8 +66,8 @@ #include <linux/seq_file.h> #include <net/protocol.h> -#include <net/tcp.h> #include <net/ndisc.h> +#include <net/ip.h> #include <net/ipv6.h> #include <net/transp_v6.h> #include <net/addrconf.h> @@ -641,10 +641,7 @@ static struct sock *sctp_v6_create_accept_sk(struct sock *sk, else newinet->pmtudisc = IP_PMTUDISC_WANT; -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet6_sock_nr); - atomic_inc(&inet_sock_nr); -#endif + sk_refcnt_debug_inc(newsk); if (newsk->sk_prot->init(newsk)) { sk_common_release(newsk); diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index ce9245e71fca..e7025be77691 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -62,7 +62,7 @@ /* Global data structures. */ struct sctp_globals sctp_globals; struct proc_dir_entry *proc_net_sctp; -DEFINE_SNMP_STAT(struct sctp_mib, sctp_statistics); +DEFINE_SNMP_STAT(struct sctp_mib, sctp_statistics) __read_mostly; struct idr sctp_assocs_id; DEFINE_SPINLOCK(sctp_assocs_id_lock); @@ -78,8 +78,8 @@ static struct sctp_pf *sctp_pf_inet_specific; static struct sctp_af *sctp_af_v4_specific; static struct sctp_af *sctp_af_v6_specific; -kmem_cache_t *sctp_chunk_cachep; -kmem_cache_t *sctp_bucket_cachep; +kmem_cache_t *sctp_chunk_cachep __read_mostly; +kmem_cache_t *sctp_bucket_cachep __read_mostly; extern int sctp_snmp_proc_init(void); extern int sctp_snmp_proc_exit(void); @@ -593,9 +593,7 @@ static struct sock *sctp_v4_create_accept_sk(struct sock *sk, newinet->mc_index = 0; newinet->mc_list = NULL; -#ifdef INET_REFCNT_DEBUG - atomic_inc(&inet_sock_nr); -#endif + sk_refcnt_debug_inc(newsk); if (newsk->sk_prot->init(newsk)) { sk_common_release(newsk); @@ -1244,6 +1242,10 @@ SCTP_STATIC __exit void sctp_exit(void) module_init(sctp_init); module_exit(sctp_exit); +/* + * __stringify doesn't likes enums, so use IPPROTO_SCTP value (132) directly. + */ +MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-proto-132"); MODULE_AUTHOR("Linux Kernel SCTP developers <lksctp-developers@lists.sourceforge.net>"); MODULE_DESCRIPTION("Support for the SCTP protocol (RFC2960)"); MODULE_LICENSE("GPL"); diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index 00d32b7c8266..3868a8d70cc0 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -1362,6 +1362,7 @@ struct sctp_association *sctp_unpack_cookie( char *key; sctp_scope_t scope; struct sk_buff *skb = chunk->skb; + struct timeval tv; headersize = sizeof(sctp_chunkhdr_t) + SCTP_SECRET_SIZE; bodysize = ntohs(chunk->chunk_hdr->length) - headersize; @@ -1434,7 +1435,8 @@ no_hmac: * an association, there is no need to check cookie's expiration * for init collision case of lost COOKIE ACK. */ - if (!asoc && tv_lt(bear_cookie->expiration, skb->stamp)) { + skb_get_timestamp(skb, &tv); + if (!asoc && tv_lt(bear_cookie->expiration, tv)) { __u16 len; /* * Section 3.3.10.3 Stale Cookie Error (3) @@ -1447,10 +1449,9 @@ no_hmac: len = ntohs(chunk->chunk_hdr->length); *errp = sctp_make_op_error_space(asoc, chunk, len); if (*errp) { - suseconds_t usecs = (skb->stamp.tv_sec - + suseconds_t usecs = (tv.tv_sec - bear_cookie->expiration.tv_sec) * 1000000L + - skb->stamp.tv_usec - - bear_cookie->expiration.tv_usec; + tv.tv_usec - bear_cookie->expiration.tv_usec; usecs = htonl(usecs); sctp_init_cause(*errp, SCTP_ERROR_STALE_COOKIE, diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 091a66f06a35..4454afe4727e 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4892,7 +4892,7 @@ static void sctp_sock_migrate(struct sock *oldsk, struct sock *newsk, sctp_skb_for_each(skb, &oldsk->sk_receive_queue, tmp) { event = sctp_skb2event(skb); if (event->asoc == assoc) { - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &oldsk->sk_receive_queue); __skb_queue_tail(&newsk->sk_receive_queue, skb); } } @@ -4921,7 +4921,7 @@ static void sctp_sock_migrate(struct sock *oldsk, struct sock *newsk, sctp_skb_for_each(skb, &oldsp->pd_lobby, tmp) { event = sctp_skb2event(skb); if (event->asoc == assoc) { - __skb_unlink(skb, skb->list); + __skb_unlink(skb, &oldsp->pd_lobby); __skb_queue_tail(queue, skb); } } diff --git a/net/sctp/ulpqueue.c b/net/sctp/ulpqueue.c index 8bbc279d6c99..ec2c857eae7f 100644 --- a/net/sctp/ulpqueue.c +++ b/net/sctp/ulpqueue.c @@ -50,9 +50,9 @@ /* Forward declarations for internal helpers. */ static struct sctp_ulpevent * sctp_ulpq_reasm(struct sctp_ulpq *ulpq, - struct sctp_ulpevent *); + struct sctp_ulpevent *); static struct sctp_ulpevent * sctp_ulpq_order(struct sctp_ulpq *, - struct sctp_ulpevent *); + struct sctp_ulpevent *); /* 1st Level Abstractions */ @@ -125,7 +125,9 @@ int sctp_ulpq_tail_data(struct sctp_ulpq *ulpq, struct sctp_chunk *chunk, event = sctp_ulpq_order(ulpq, event); } - /* Send event to the ULP. */ + /* Send event to the ULP. 'event' is the sctp_ulpevent for + * very first SKB on the 'temp' list. + */ if (event) sctp_ulpq_tail_event(ulpq, event); @@ -158,14 +160,18 @@ static int sctp_ulpq_clear_pd(struct sctp_ulpq *ulpq) return sctp_clear_pd(ulpq->asoc->base.sk); } - - +/* If the SKB of 'event' is on a list, it is the first such member + * of that list. + */ int sctp_ulpq_tail_event(struct sctp_ulpq *ulpq, struct sctp_ulpevent *event) { struct sock *sk = ulpq->asoc->base.sk; - struct sk_buff_head *queue; + struct sk_buff_head *queue, *skb_list; + struct sk_buff *skb = sctp_event2skb(event); int clear_pd = 0; + skb_list = (struct sk_buff_head *) skb->prev; + /* If the socket is just going to throw this away, do not * even try to deliver it. */ @@ -197,10 +203,10 @@ int sctp_ulpq_tail_event(struct sctp_ulpq *ulpq, struct sctp_ulpevent *event) /* If we are harvesting multiple skbs they will be * collected on a list. */ - if (sctp_event2skb(event)->list) - sctp_skb_list_tail(sctp_event2skb(event)->list, queue); + if (skb_list) + sctp_skb_list_tail(skb_list, queue); else - __skb_queue_tail(queue, sctp_event2skb(event)); + __skb_queue_tail(queue, skb); /* Did we just complete partial delivery and need to get * rolling again? Move pending data to the receive @@ -214,10 +220,11 @@ int sctp_ulpq_tail_event(struct sctp_ulpq *ulpq, struct sctp_ulpevent *event) return 1; out_free: - if (sctp_event2skb(event)->list) - sctp_queue_purge_ulpevents(sctp_event2skb(event)->list); + if (skb_list) + sctp_queue_purge_ulpevents(skb_list); else sctp_ulpevent_free(event); + return 0; } @@ -269,7 +276,7 @@ static inline void sctp_ulpq_store_reasm(struct sctp_ulpq *ulpq, * payload was fragmented on the way and ip had to reassemble them. * We add the rest of skb's to the first skb's fraglist. */ -static struct sctp_ulpevent *sctp_make_reassembled_event(struct sk_buff *f_frag, struct sk_buff *l_frag) +static struct sctp_ulpevent *sctp_make_reassembled_event(struct sk_buff_head *queue, struct sk_buff *f_frag, struct sk_buff *l_frag) { struct sk_buff *pos; struct sctp_ulpevent *event; @@ -294,7 +301,7 @@ static struct sctp_ulpevent *sctp_make_reassembled_event(struct sk_buff *f_frag, skb_shinfo(f_frag)->frag_list = pos; /* Remove the first fragment from the reassembly queue. */ - __skb_unlink(f_frag, f_frag->list); + __skb_unlink(f_frag, queue); while (pos) { pnext = pos->next; @@ -304,7 +311,7 @@ static struct sctp_ulpevent *sctp_make_reassembled_event(struct sk_buff *f_frag, f_frag->data_len += pos->len; /* Remove the fragment from the reassembly queue. */ - __skb_unlink(pos, pos->list); + __skb_unlink(pos, queue); /* Break if we have reached the last fragment. */ if (pos == l_frag) @@ -375,7 +382,7 @@ static inline struct sctp_ulpevent *sctp_ulpq_retrieve_reassembled(struct sctp_u done: return retval; found: - retval = sctp_make_reassembled_event(first_frag, pos); + retval = sctp_make_reassembled_event(&ulpq->reasm, first_frag, pos); if (retval) retval->msg_flags |= MSG_EOR; goto done; @@ -435,7 +442,7 @@ static inline struct sctp_ulpevent *sctp_ulpq_retrieve_partial(struct sctp_ulpq * further. */ done: - retval = sctp_make_reassembled_event(first_frag, last_frag); + retval = sctp_make_reassembled_event(&ulpq->reasm, first_frag, last_frag); if (retval && is_last) retval->msg_flags |= MSG_EOR; @@ -527,7 +534,7 @@ static inline struct sctp_ulpevent *sctp_ulpq_retrieve_first(struct sctp_ulpq *u * further. */ done: - retval = sctp_make_reassembled_event(first_frag, last_frag); + retval = sctp_make_reassembled_event(&ulpq->reasm, first_frag, last_frag); return retval; } @@ -537,6 +544,7 @@ done: static inline void sctp_ulpq_retrieve_ordered(struct sctp_ulpq *ulpq, struct sctp_ulpevent *event) { + struct sk_buff_head *event_list; struct sk_buff *pos, *tmp; struct sctp_ulpevent *cevent; struct sctp_stream *in; @@ -547,6 +555,8 @@ static inline void sctp_ulpq_retrieve_ordered(struct sctp_ulpq *ulpq, ssn = event->ssn; in = &ulpq->asoc->ssnmap->in; + event_list = (struct sk_buff_head *) sctp_event2skb(event)->prev; + /* We are holding the chunks by stream, by SSN. */ sctp_skb_for_each(pos, &ulpq->lobby, tmp) { cevent = (struct sctp_ulpevent *) pos->cb; @@ -567,10 +577,10 @@ static inline void sctp_ulpq_retrieve_ordered(struct sctp_ulpq *ulpq, /* Found it, so mark in the ssnmap. */ sctp_ssn_next(in, sid); - __skb_unlink(pos, pos->list); + __skb_unlink(pos, &ulpq->lobby); /* Attach all gathered skbs to the event. */ - __skb_queue_tail(sctp_event2skb(event)->list, pos); + __skb_queue_tail(event_list, pos); } } @@ -626,7 +636,7 @@ static inline void sctp_ulpq_store_ordered(struct sctp_ulpq *ulpq, } static struct sctp_ulpevent *sctp_ulpq_order(struct sctp_ulpq *ulpq, - struct sctp_ulpevent *event) + struct sctp_ulpevent *event) { __u16 sid, ssn; struct sctp_stream *in; @@ -667,7 +677,7 @@ static inline void sctp_ulpq_reap_ordered(struct sctp_ulpq *ulpq) { struct sk_buff *pos, *tmp; struct sctp_ulpevent *cevent; - struct sctp_ulpevent *event = NULL; + struct sctp_ulpevent *event; struct sctp_stream *in; struct sk_buff_head temp; __u16 csid, cssn; @@ -675,6 +685,8 @@ static inline void sctp_ulpq_reap_ordered(struct sctp_ulpq *ulpq) in = &ulpq->asoc->ssnmap->in; /* We are holding the chunks by stream, by SSN. */ + skb_queue_head_init(&temp); + event = NULL; sctp_skb_for_each(pos, &ulpq->lobby, tmp) { cevent = (struct sctp_ulpevent *) pos->cb; csid = cevent->stream; @@ -686,19 +698,20 @@ static inline void sctp_ulpq_reap_ordered(struct sctp_ulpq *ulpq) /* Found it, so mark in the ssnmap. */ sctp_ssn_next(in, csid); - __skb_unlink(pos, pos->list); + __skb_unlink(pos, &ulpq->lobby); if (!event) { /* Create a temporary list to collect chunks on. */ event = sctp_skb2event(pos); - skb_queue_head_init(&temp); __skb_queue_tail(&temp, sctp_event2skb(event)); } else { /* Attach all gathered skbs to the event. */ - __skb_queue_tail(sctp_event2skb(event)->list, pos); + __skb_queue_tail(&temp, pos); } } - /* Send event to the ULP. */ + /* Send event to the ULP. 'event' is the sctp_ulpevent for + * very first SKB on the 'temp' list. + */ if (event) sctp_ulpq_tail_event(ulpq, event); } diff --git a/net/socket.c b/net/socket.c index 6f2a17881972..94fe638b4d72 100644 --- a/net/socket.c +++ b/net/socket.c @@ -70,6 +70,8 @@ #include <linux/seq_file.h> #include <linux/wanrouter.h> #include <linux/if_bridge.h> +#include <linux/if_frad.h> +#include <linux/if_vlan.h> #include <linux/init.h> #include <linux/poll.h> #include <linux/cache.h> @@ -272,7 +274,7 @@ int move_addr_to_user(void *kaddr, int klen, void __user *uaddr, int __user *ule #define SOCKFS_MAGIC 0x534F434B -static kmem_cache_t * sock_inode_cachep; +static kmem_cache_t * sock_inode_cachep __read_mostly; static struct inode *sock_alloc_inode(struct super_block *sb) { @@ -331,7 +333,7 @@ static struct super_block *sockfs_get_sb(struct file_system_type *fs_type, return get_sb_pseudo(fs_type, "socket:", &sockfs_ops, SOCKFS_MAGIC); } -static struct vfsmount *sock_mnt; +static struct vfsmount *sock_mnt __read_mostly; static struct file_system_type sock_fs_type = { .name = "sockfs", @@ -404,6 +406,7 @@ int sock_map_fd(struct socket *sock) file->f_mode = FMODE_READ | FMODE_WRITE; file->f_flags = O_RDWR; file->f_pos = 0; + file->private_data = sock; fd_install(fd, file); } @@ -436,6 +439,9 @@ struct socket *sockfd_lookup(int fd, int *err) return NULL; } + if (file->f_op == &socket_file_ops) + return file->private_data; /* set in sock_map_fd */ + inode = file->f_dentry->d_inode; if (!S_ISSOCK(inode->i_mode)) { *err = -ENOTSOCK; @@ -720,8 +726,8 @@ static ssize_t sock_aio_write(struct kiocb *iocb, const char __user *ubuf, return __sock_sendmsg(iocb, sock, &x->async_msg, size); } -ssize_t sock_sendpage(struct file *file, struct page *page, - int offset, size_t size, loff_t *ppos, int more) +static ssize_t sock_sendpage(struct file *file, struct page *page, + int offset, size_t size, loff_t *ppos, int more) { struct socket *sock; int flags; @@ -944,7 +950,7 @@ static int sock_mmap(struct file * file, struct vm_area_struct * vma) return sock->ops->mmap(file, sock, vma); } -int sock_close(struct inode *inode, struct file *filp) +static int sock_close(struct inode *inode, struct file *filp) { /* * It was possible the inode is NULL we were @@ -2023,9 +2029,6 @@ int sock_unregister(int family) return 0; } - -extern void sk_init(void); - void __init sock_init(void) { /* diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 554f224c0445..fe1a73ce6cff 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -28,13 +28,13 @@ #include <linux/workqueue.h> #include <linux/sunrpc/rpc_pipe_fs.h> -static struct vfsmount *rpc_mount; +static struct vfsmount *rpc_mount __read_mostly; static int rpc_mount_count; static struct file_system_type rpc_pipe_fs_type; -static kmem_cache_t *rpc_inode_cachep; +static kmem_cache_t *rpc_inode_cachep __read_mostly; #define RPC_UPCALL_TIMEOUT (30*HZ) diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index 2d9eb7fbd521..f3104035e35d 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -34,10 +34,10 @@ static int rpc_task_id; #define RPC_BUFFER_MAXSIZE (2048) #define RPC_BUFFER_POOLSIZE (8) #define RPC_TASK_POOLSIZE (8) -static kmem_cache_t *rpc_task_slabp; -static kmem_cache_t *rpc_buffer_slabp; -static mempool_t *rpc_task_mempool; -static mempool_t *rpc_buffer_mempool; +static kmem_cache_t *rpc_task_slabp __read_mostly; +static kmem_cache_t *rpc_buffer_slabp __read_mostly; +static mempool_t *rpc_task_mempool __read_mostly; +static mempool_t *rpc_buffer_mempool __read_mostly; static void __rpc_default_timer(struct rpc_task *task); static void rpciod_killall(void); diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index d0c3120d0233..05fe2e735538 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -34,7 +34,7 @@ #include <net/sock.h> #include <net/checksum.h> #include <net/ip.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <asm/ioctls.h> @@ -584,13 +584,16 @@ svc_udp_recvfrom(struct svc_rqst *rqstp) /* possibly an icmp error */ dprintk("svc: recvfrom returned error %d\n", -err); } - if (skb->stamp.tv_sec == 0) { - skb->stamp.tv_sec = xtime.tv_sec; - skb->stamp.tv_usec = xtime.tv_nsec / NSEC_PER_USEC; + if (skb->tstamp.off_sec == 0) { + struct timeval tv; + + tv.tv_sec = xtime.tv_sec; + tv.tv_usec = xtime.tv_nsec * 1000; + skb_set_timestamp(skb, &tv); /* Don't enable netstamp, sunrpc doesn't need that much accuracy */ } - svsk->sk_sk->sk_stamp = skb->stamp; + skb_get_timestamp(skb, &svsk->sk_sk->sk_stamp); set_bit(SK_DATA, &svsk->sk_flags); /* there may be more data... */ /* diff --git a/net/sysctl_net.c b/net/sysctl_net.c index 3f6e31069c54..c5241fcbb966 100644 --- a/net/sysctl_net.c +++ b/net/sysctl_net.c @@ -17,17 +17,15 @@ #include <linux/sysctl.h> #ifdef CONFIG_INET -extern struct ctl_table ipv4_table[]; +#include <net/ip.h> #endif -extern struct ctl_table core_table[]; - #ifdef CONFIG_NET -extern struct ctl_table ether_table[]; +#include <linux/if_ether.h> #endif #ifdef CONFIG_TR -extern struct ctl_table tr_table[]; +#include <linux/if_tr.h> #endif struct ctl_table net_table[] = { diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index d403e34088ad..41feca3bef86 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -105,7 +105,7 @@ #include <linux/skbuff.h> #include <linux/netdevice.h> #include <net/sock.h> -#include <linux/tcp.h> +#include <net/tcp_states.h> #include <net/af_unix.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> @@ -2026,14 +2026,6 @@ static struct net_proto_family unix_family_ops = { .owner = THIS_MODULE, }; -#ifdef CONFIG_SYSCTL -extern void unix_sysctl_register(void); -extern void unix_sysctl_unregister(void); -#else -static inline void unix_sysctl_register(void) {} -static inline void unix_sysctl_unregister(void) {} -#endif - static int __init af_unix_init(void) { int rc = -1; diff --git a/net/unix/garbage.c b/net/unix/garbage.c index 4bd95c8f5934..6ffc64e1712d 100644 --- a/net/unix/garbage.c +++ b/net/unix/garbage.c @@ -76,11 +76,11 @@ #include <linux/netdevice.h> #include <linux/file.h> #include <linux/proc_fs.h> -#include <linux/tcp.h> #include <net/sock.h> #include <net/af_unix.h> #include <net/scm.h> +#include <net/tcp_states.h> /* Internal data structures and random procedures: */ @@ -286,16 +286,16 @@ void unix_gc(void) skb = skb_peek(&s->sk_receive_queue); while (skb && skb != (struct sk_buff *)&s->sk_receive_queue) { - nextsk=skb->next; + nextsk = skb->next; /* * Do we have file descriptors ? */ - if(UNIXCB(skb).fp) - { - __skb_unlink(skb, skb->list); - __skb_queue_tail(&hitlist,skb); + if (UNIXCB(skb).fp) { + __skb_unlink(skb, + &s->sk_receive_queue); + __skb_queue_tail(&hitlist, skb); } - skb=nextsk; + skb = nextsk; } spin_unlock(&s->sk_receive_queue.lock); } diff --git a/net/unix/sysctl_net_unix.c b/net/unix/sysctl_net_unix.c index c974dac4580a..690ffa5d5bfb 100644 --- a/net/unix/sysctl_net_unix.c +++ b/net/unix/sysctl_net_unix.c @@ -12,7 +12,7 @@ #include <linux/mm.h> #include <linux/sysctl.h> -extern int sysctl_unix_max_dgram_qlen; +#include <net/af_unix.h> static ctl_table unix_table[] = { { diff --git a/net/wanrouter/af_wanpipe.c b/net/wanrouter/af_wanpipe.c index d93b19faaab7..596cb96e5f47 100644 --- a/net/wanrouter/af_wanpipe.c +++ b/net/wanrouter/af_wanpipe.c @@ -57,7 +57,7 @@ #include <linux/wanpipe.h> #include <linux/if_wanpipe.h> #include <linux/pkt_sched.h> -#include <linux/tcp.h> +#include <linux/tcp_states.h> #include <linux/if_wanpipe_common.h> #include <linux/sdla_x25.h> diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c index 04bec047fa9a..020d73cc8414 100644 --- a/net/x25/af_x25.c +++ b/net/x25/af_x25.c @@ -47,7 +47,7 @@ #include <linux/if_arp.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <asm/uaccess.h> #include <linux/fcntl.h> #include <linux/termios.h> /* For TIOCINQ/OUTQ */ diff --git a/net/x25/x25_dev.c b/net/x25/x25_dev.c index 36fc3bf6d882..adfe7b8df355 100644 --- a/net/x25/x25_dev.c +++ b/net/x25/x25_dev.c @@ -81,7 +81,7 @@ static int x25_receive_data(struct sk_buff *skb, struct x25_neigh *nb) } int x25_lapb_receive_frame(struct sk_buff *skb, struct net_device *dev, - struct packet_type *ptype) + struct packet_type *ptype, struct net_device *orig_dev) { struct sk_buff *nskb; struct x25_neigh *nb; diff --git a/net/x25/x25_in.c b/net/x25/x25_in.c index b0197c70a9fc..26146874b839 100644 --- a/net/x25/x25_in.c +++ b/net/x25/x25_in.c @@ -28,7 +28,7 @@ #include <linux/string.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/x25.h> static int x25_queue_rx_frame(struct sock *sk, struct sk_buff *skb, int more) diff --git a/net/x25/x25_subr.c b/net/x25/x25_subr.c index 7fd872ad0c20..8be9b8fbc24d 100644 --- a/net/x25/x25_subr.c +++ b/net/x25/x25_subr.c @@ -27,7 +27,7 @@ #include <linux/string.h> #include <linux/skbuff.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/x25.h> /* @@ -80,7 +80,7 @@ void x25_requeue_frames(struct sock *sk) if (!skb_prev) skb_queue_head(&sk->sk_write_queue, skb); else - skb_append(skb_prev, skb); + skb_append(skb_prev, skb, &sk->sk_write_queue); skb_prev = skb; } } diff --git a/net/x25/x25_timer.c b/net/x25/x25_timer.c index d6a21a3ad80e..0a92e1da3922 100644 --- a/net/x25/x25_timer.c +++ b/net/x25/x25_timer.c @@ -23,7 +23,7 @@ #include <linux/jiffies.h> #include <linux/timer.h> #include <net/sock.h> -#include <net/tcp.h> +#include <net/tcp_states.h> #include <net/x25.h> static void x25_heartbeat_expiry(unsigned long); diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c index c58a6f05a0b6..2407a7072327 100644 --- a/net/xfrm/xfrm_input.c +++ b/net/xfrm/xfrm_input.c @@ -12,7 +12,7 @@ #include <net/ip.h> #include <net/xfrm.h> -static kmem_cache_t *secpath_cachep; +static kmem_cache_t *secpath_cachep __read_mostly; void __secpath_destroy(struct sec_path *sp) { diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index d65ed8684fc1..83c8135e1764 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -37,7 +37,7 @@ EXPORT_SYMBOL(xfrm_policy_list); static DEFINE_RWLOCK(xfrm_policy_afinfo_lock); static struct xfrm_policy_afinfo *xfrm_policy_afinfo[NPROTO]; -static kmem_cache_t *xfrm_dst_cache; +static kmem_cache_t *xfrm_dst_cache __read_mostly; static struct work_struct xfrm_policy_gc_work; static struct list_head xfrm_policy_gc_list = diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 8da3e25b2c4c..c35336a0f71b 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -1125,9 +1125,8 @@ static int xfrm_exp_state_notify(struct xfrm_state *x, struct km_event *c) if (build_expire(skb, x, c->data.hard) < 0) BUG(); - NETLINK_CB(skb).dst_groups = XFRMGRP_EXPIRE; - - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_EXPIRE, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_EXPIRE; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_EXPIRE, GFP_ATOMIC); } static int xfrm_notify_sa_flush(struct km_event *c) @@ -1152,7 +1151,8 @@ static int xfrm_notify_sa_flush(struct km_event *c) nlh->nlmsg_len = skb->tail - b; - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_SA, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_SA; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_SA, GFP_ATOMIC); nlmsg_failure: kfree_skb(skb); @@ -1226,7 +1226,8 @@ static int xfrm_notify_sa(struct xfrm_state *x, struct km_event *c) nlh->nlmsg_len = skb->tail - b; - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_SA, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_SA; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_SA, GFP_ATOMIC); nlmsg_failure: rtattr_failure: @@ -1304,9 +1305,8 @@ static int xfrm_send_acquire(struct xfrm_state *x, struct xfrm_tmpl *xt, if (build_acquire(skb, x, xt, xp, dir) < 0) BUG(); - NETLINK_CB(skb).dst_groups = XFRMGRP_ACQUIRE; - - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_ACQUIRE, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_ACQUIRE; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_ACQUIRE, GFP_ATOMIC); } /* User gives us xfrm_user_policy_info followed by an array of 0 @@ -1405,9 +1405,8 @@ static int xfrm_exp_policy_notify(struct xfrm_policy *xp, int dir, struct km_eve if (build_polexpire(skb, xp, dir, c->data.hard) < 0) BUG(); - NETLINK_CB(skb).dst_groups = XFRMGRP_EXPIRE; - - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_EXPIRE, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_EXPIRE; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_EXPIRE, GFP_ATOMIC); } static int xfrm_notify_policy(struct xfrm_policy *xp, int dir, struct km_event *c) @@ -1455,7 +1454,8 @@ static int xfrm_notify_policy(struct xfrm_policy *xp, int dir, struct km_event * nlh->nlmsg_len = skb->tail - b; - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_POLICY, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_POLICY; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_POLICY, GFP_ATOMIC); nlmsg_failure: rtattr_failure: @@ -1480,7 +1480,8 @@ static int xfrm_notify_policy_flush(struct km_event *c) nlh->nlmsg_len = skb->tail - b; - return netlink_broadcast(xfrm_nl, skb, 0, XFRMGRP_POLICY, GFP_ATOMIC); + NETLINK_CB(skb).dst_group = XFRMNLGRP_POLICY; + return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_POLICY, GFP_ATOMIC); nlmsg_failure: kfree_skb(skb); @@ -1519,7 +1520,8 @@ static int __init xfrm_user_init(void) { printk(KERN_INFO "Initializing IPsec netlink socket\n"); - xfrm_nl = netlink_kernel_create(NETLINK_XFRM, xfrm_netlink_rcv); + xfrm_nl = netlink_kernel_create(NETLINK_XFRM, XFRMNLGRP_MAX, + xfrm_netlink_rcv, THIS_MODULE); if (xfrm_nl == NULL) return -ENOMEM; @@ -1537,3 +1539,4 @@ static void __exit xfrm_user_exit(void) module_init(xfrm_user_init); module_exit(xfrm_user_exit); MODULE_LICENSE("GPL"); +MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_XFRM); |