diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2022-08-04 02:29:08 +0300 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2022-08-04 02:29:08 +0300 |
commit | f86d1fbbe7858884d6754534a0afbb74fc30bc26 (patch) | |
tree | f61796870edefbe77d495e9d719c68af1d14275b /net/ipv6/ip6mr.c | |
parent | 526942b8134cc34d25d27f95dfff98b8ce2f6fcd (diff) | |
parent | 7c6327c77d509e78bff76f2a4551fcfee851682e (diff) | |
download | linux-f86d1fbbe7858884d6754534a0afbb74fc30bc26.tar.xz |
Merge tag 'net-next-6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next
Pull networking changes from Paolo Abeni:
"Core:
- Refactor the forward memory allocation to better cope with memory
pressure with many open sockets, moving from a per socket cache to
a per-CPU one
- Replace rwlocks with RCU for better fairness in ping, raw sockets
and IP multicast router.
- Network-side support for IO uring zero-copy send.
- A few skb drop reason improvements, including codegen the source
file with string mapping instead of using macro magic.
- Rename reference tracking helpers to a more consistent netdev_*
schema.
- Adapt u64_stats_t type to address load/store tearing issues.
- Refine debug helper usage to reduce the log noise caused by bots.
BPF:
- Improve socket map performance, avoiding skb cloning on read
operation.
- Add support for 64 bits enum, to match types exposed by kernel.
- Introduce support for sleepable uprobes program.
- Introduce support for enum textual representation in libbpf.
- New helpers to implement synproxy with eBPF/XDP.
- Improve loop performances, inlining indirect calls when possible.
- Removed all the deprecated libbpf APIs.
- Implement new eBPF-based LSM flavor.
- Add type match support, which allow accurate queries to the eBPF
used types.
- A few TCP congetsion control framework usability improvements.
- Add new infrastructure to manipulate CT entries via eBPF programs.
- Allow for livepatch (KLP) and BPF trampolines to attach to the same
kernel function.
Protocols:
- Introduce per network namespace lookup tables for unix sockets,
increasing scalability and reducing contention.
- Preparation work for Wi-Fi 7 Multi-Link Operation (MLO) support.
- Add support to forciby close TIME_WAIT TCP sockets via user-space
tools.
- Significant performance improvement for the TLS 1.3 receive path,
both for zero-copy and not-zero-copy.
- Support for changing the initial MTPCP subflow priority/backup
status
- Introduce virtually contingus buffers for sockets over RDMA, to
cope better with memory pressure.
- Extend CAN ethtool support with timestamping capabilities
- Refactor CAN build infrastructure to allow building only the needed
features.
Driver API:
- Remove devlink mutex to allow parallel commands on multiple links.
- Add support for pause stats in distributed switch.
- Implement devlink helpers to query and flash line cards.
- New helper for phy mode to register conversion.
New hardware / drivers:
- Ethernet DSA driver for the rockchip mt7531 on BPI-R2 Pro.
- Ethernet DSA driver for the Renesas RZ/N1 A5PSW switch.
- Ethernet DSA driver for the Microchip LAN937x switch.
- Ethernet PHY driver for the Aquantia AQR113C EPHY.
- CAN driver for the OBD-II ELM327 interface.
- CAN driver for RZ/N1 SJA1000 CAN controller.
- Bluetooth: Infineon CYW55572 Wi-Fi plus Bluetooth combo device.
Drivers:
- Intel Ethernet NICs:
- i40e: add support for vlan pruning
- i40e: add support for XDP framented packets
- ice: improved vlan offload support
- ice: add support for PPPoE offload
- Mellanox Ethernet (mlx5)
- refactor packet steering offload for performance and scalability
- extend support for TC offload
- refactor devlink code to clean-up the locking schema
- support stacked vlans for bridge offloads
- use TLS objects pool to improve connection rate
- Netronome Ethernet NICs (nfp):
- extend support for IPv6 fields mangling offload
- add support for vepa mode in HW bridge
- better support for virtio data path acceleration (VDPA)
- enable TSO by default
- Microsoft vNIC driver (mana)
- add support for XDP redirect
- Others Ethernet drivers:
- bonding: add per-port priority support
- microchip lan743x: extend phy support
- Fungible funeth: support UDP segmentation offload and XDP xmit
- Solarflare EF100: add support for virtual function representors
- MediaTek SoC: add XDP support
- Mellanox Ethernet/IB switch (mlxsw):
- dropped support for unreleased H/W (XM router).
- improved stats accuracy
- unified bridge model coversion improving scalability (parts 1-6)
- support for PTP in Spectrum-2 asics
- Broadcom PHYs
- add PTP support for BCM54210E
- add support for the BCM53128 internal PHY
- Marvell Ethernet switches (prestera):
- implement support for multicast forwarding offload
- Embedded Ethernet switches:
- refactor OcteonTx MAC filter for better scalability
- improve TC H/W offload for the Felix driver
- refactor the Microchip ksz8 and ksz9477 drivers to share the
probe code (parts 1, 2), add support for phylink mac
configuration
- Other WiFi:
- Microchip wilc1000: diable WEP support and enable WPA3
- Atheros ath10k: encapsulation offload support
Old code removal:
- Neterion vxge ethernet driver: this is untouched since more than 10 years"
* tag 'net-next-6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1890 commits)
doc: sfp-phylink: Fix a broken reference
wireguard: selftests: support UML
wireguard: allowedips: don't corrupt stack when detecting overflow
wireguard: selftests: update config fragments
wireguard: ratelimiter: use hrtimer in selftest
net/mlx5e: xsk: Discard unaligned XSK frames on striding RQ
net: usb: ax88179_178a: Bind only to vendor-specific interface
selftests: net: fix IOAM test skip return code
net: usb: make USB_RTL8153_ECM non user configurable
net: marvell: prestera: remove reduntant code
octeontx2-pf: Reduce minimum mtu size to 60
net: devlink: Fix missing mutex_unlock() call
net/tls: Remove redundant workqueue flush before destroy
net: txgbe: Fix an error handling path in txgbe_probe()
net: dsa: Fix spelling mistakes and cleanup code
Documentation: devlink: add add devlink-selftests to the table of contents
dccp: put dccp_qpolicy_full() and dccp_qpolicy_push() in the same lock
net: ionic: fix error check for vlan flags in ionic_set_nic_features()
net: ice: fix error NETIF_F_HW_VLAN_CTAG_FILTER check in ice_vsi_sync_fltr()
nfp: flower: add support for tunnel offload without key ID
...
Diffstat (limited to 'net/ipv6/ip6mr.c')
-rw-r--r-- | net/ipv6/ip6mr.c | 301 |
1 files changed, 196 insertions, 105 deletions
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c index 4e74bc61a3db..a9ba41648e36 100644 --- a/net/ipv6/ip6mr.c +++ b/net/ipv6/ip6mr.c @@ -62,7 +62,12 @@ struct ip6mr_result { Note that the changes are semaphored via rtnl_lock. */ -static DEFINE_RWLOCK(mrt_lock); +static DEFINE_SPINLOCK(mrt_lock); + +static struct net_device *vif_dev_read(const struct vif_device *vif) +{ + return rcu_dereference(vif->dev); +} /* Multicast router control variables */ @@ -85,11 +90,13 @@ static void ip6mr_free_table(struct mr_table *mrt); static void ip6_mr_forward(struct net *net, struct mr_table *mrt, struct net_device *dev, struct sk_buff *skb, struct mfc6_cache *cache); -static int ip6mr_cache_report(struct mr_table *mrt, struct sk_buff *pkt, +static int ip6mr_cache_report(const struct mr_table *mrt, struct sk_buff *pkt, mifi_t mifi, int assert); static void mr6_netlink_event(struct mr_table *mrt, struct mfc6_cache *mfc, int cmd); -static void mrt6msg_netlink_event(struct mr_table *mrt, struct sk_buff *pkt); +static void mrt6msg_netlink_event(const struct mr_table *mrt, struct sk_buff *pkt); +static int ip6mr_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, + struct netlink_ext_ack *extack); static int ip6mr_rtm_dumproute(struct sk_buff *skb, struct netlink_callback *cb); static void mroute_clean_tables(struct mr_table *mrt, int flags); @@ -398,7 +405,7 @@ static void ip6mr_free_table(struct mr_table *mrt) */ static void *ip6mr_vif_seq_start(struct seq_file *seq, loff_t *pos) - __acquires(mrt_lock) + __acquires(RCU) { struct mr_vif_iter *iter = seq->private; struct net *net = seq_file_net(seq); @@ -410,14 +417,14 @@ static void *ip6mr_vif_seq_start(struct seq_file *seq, loff_t *pos) iter->mrt = mrt; - read_lock(&mrt_lock); + rcu_read_lock(); return mr_vif_seq_start(seq, pos); } static void ip6mr_vif_seq_stop(struct seq_file *seq, void *v) - __releases(mrt_lock) + __releases(RCU) { - read_unlock(&mrt_lock); + rcu_read_unlock(); } static int ip6mr_vif_seq_show(struct seq_file *seq, void *v) @@ -430,7 +437,11 @@ static int ip6mr_vif_seq_show(struct seq_file *seq, void *v) "Interface BytesIn PktsIn BytesOut PktsOut Flags\n"); } else { const struct vif_device *vif = v; - const char *name = vif->dev ? vif->dev->name : "none"; + const struct net_device *vif_dev; + const char *name; + + vif_dev = vif_dev_read(vif); + name = vif_dev ? vif_dev->name : "none"; seq_printf(seq, "%2td %-10s %8ld %7ld %8ld %7ld %05X\n", @@ -549,13 +560,11 @@ static int pim6_rcv(struct sk_buff *skb) if (ip6mr_fib_lookup(net, &fl6, &mrt) < 0) goto drop; - reg_vif_num = mrt->mroute_reg_vif_num; - read_lock(&mrt_lock); + /* Pairs with WRITE_ONCE() in mif6_add()/mif6_delete() */ + reg_vif_num = READ_ONCE(mrt->mroute_reg_vif_num); if (reg_vif_num >= 0) - reg_dev = mrt->vif_table[reg_vif_num].dev; - dev_hold(reg_dev); - read_unlock(&mrt_lock); + reg_dev = vif_dev_read(&mrt->vif_table[reg_vif_num]); if (!reg_dev) goto drop; @@ -570,7 +579,6 @@ static int pim6_rcv(struct sk_buff *skb) netif_rx(skb); - dev_put(reg_dev); return 0; drop: kfree_skb(skb); @@ -600,11 +608,12 @@ static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, if (ip6mr_fib_lookup(net, &fl6, &mrt) < 0) goto tx_err; - read_lock(&mrt_lock); dev->stats.tx_bytes += skb->len; dev->stats.tx_packets++; - ip6mr_cache_report(mrt, skb, mrt->mroute_reg_vif_num, MRT6MSG_WHOLEPKT); - read_unlock(&mrt_lock); + rcu_read_lock(); + ip6mr_cache_report(mrt, skb, READ_ONCE(mrt->mroute_reg_vif_num), + MRT6MSG_WHOLEPKT); + rcu_read_unlock(); kfree_skb(skb); return NETDEV_TX_OK; @@ -670,10 +679,11 @@ failure: static int call_ip6mr_vif_entry_notifiers(struct net *net, enum fib_event_type event_type, struct vif_device *vif, + struct net_device *vif_dev, mifi_t vif_index, u32 tb_id) { return mr_call_vif_notifiers(net, RTNL_FAMILY_IP6MR, event_type, - vif, vif_index, tb_id, + vif, vif_dev, vif_index, tb_id, &net->ipv6.ipmr_seq); } @@ -698,23 +708,21 @@ static int mif6_delete(struct mr_table *mrt, int vifi, int notify, v = &mrt->vif_table[vifi]; - if (VIF_EXISTS(mrt, vifi)) - call_ip6mr_vif_entry_notifiers(read_pnet(&mrt->net), - FIB_EVENT_VIF_DEL, v, vifi, - mrt->id); - - write_lock_bh(&mrt_lock); - dev = v->dev; - v->dev = NULL; - - if (!dev) { - write_unlock_bh(&mrt_lock); + dev = rtnl_dereference(v->dev); + if (!dev) return -EADDRNOTAVAIL; - } + + call_ip6mr_vif_entry_notifiers(read_pnet(&mrt->net), + FIB_EVENT_VIF_DEL, v, dev, + vifi, mrt->id); + spin_lock(&mrt_lock); + RCU_INIT_POINTER(v->dev, NULL); #ifdef CONFIG_IPV6_PIMSM_V2 - if (vifi == mrt->mroute_reg_vif_num) - mrt->mroute_reg_vif_num = -1; + if (vifi == mrt->mroute_reg_vif_num) { + /* Pairs with READ_ONCE() in ip6mr_cache_report() and reg_vif_xmit() */ + WRITE_ONCE(mrt->mroute_reg_vif_num, -1); + } #endif if (vifi + 1 == mrt->maxvif) { @@ -723,10 +731,10 @@ static int mif6_delete(struct mr_table *mrt, int vifi, int notify, if (VIF_EXISTS(mrt, tmp)) break; } - mrt->maxvif = tmp + 1; + WRITE_ONCE(mrt->maxvif, tmp + 1); } - write_unlock_bh(&mrt_lock); + spin_unlock(&mrt_lock); dev_set_allmulti(dev, -1); @@ -741,7 +749,7 @@ static int mif6_delete(struct mr_table *mrt, int vifi, int notify, if ((v->flags & MIFF_REGISTER) && !notify) unregister_netdevice_queue(dev, head); - dev_put_track(dev, &v->dev_tracker); + netdev_put(dev, &v->dev_tracker); return 0; } @@ -826,7 +834,7 @@ static void ipmr_expire_process(struct timer_list *t) spin_unlock(&mfc_unres_lock); } -/* Fill oifs list. It is called under write locked mrt_lock. */ +/* Fill oifs list. It is called under locked mrt_lock. */ static void ip6mr_update_thresholds(struct mr_table *mrt, struct mr_mfc *cache, @@ -912,18 +920,18 @@ static int mif6_add(struct net *net, struct mr_table *mrt, MIFF_REGISTER); /* And finish update writing critical data */ - write_lock_bh(&mrt_lock); - v->dev = dev; + spin_lock(&mrt_lock); + rcu_assign_pointer(v->dev, dev); netdev_tracker_alloc(dev, &v->dev_tracker, GFP_ATOMIC); #ifdef CONFIG_IPV6_PIMSM_V2 if (v->flags & MIFF_REGISTER) - mrt->mroute_reg_vif_num = vifi; + WRITE_ONCE(mrt->mroute_reg_vif_num, vifi); #endif if (vifi + 1 > mrt->maxvif) - mrt->maxvif = vifi + 1; - write_unlock_bh(&mrt_lock); + WRITE_ONCE(mrt->maxvif, vifi + 1); + spin_unlock(&mrt_lock); call_ip6mr_vif_entry_notifiers(net, FIB_EVENT_VIF_ADD, - v, vifi, mrt->id); + v, dev, vifi, mrt->id); return 0; } @@ -1028,10 +1036,10 @@ static void ip6mr_cache_resolve(struct net *net, struct mr_table *mrt, /* * Bounce a cache query up to pim6sd and netlink. * - * Called under mrt_lock. + * Called under rcu_read_lock() */ -static int ip6mr_cache_report(struct mr_table *mrt, struct sk_buff *pkt, +static int ip6mr_cache_report(const struct mr_table *mrt, struct sk_buff *pkt, mifi_t mifi, int assert) { struct sock *mroute6_sk; @@ -1072,7 +1080,7 @@ static int ip6mr_cache_report(struct mr_table *mrt, struct sk_buff *pkt, if (assert == MRT6MSG_WRMIFWHOLE) msg->im6_mif = mifi; else - msg->im6_mif = mrt->mroute_reg_vif_num; + msg->im6_mif = READ_ONCE(mrt->mroute_reg_vif_num); msg->im6_pad = 0; msg->im6_src = ipv6_hdr(pkt)->saddr; msg->im6_dst = ipv6_hdr(pkt)->daddr; @@ -1107,10 +1115,8 @@ static int ip6mr_cache_report(struct mr_table *mrt, struct sk_buff *pkt, skb->ip_summed = CHECKSUM_UNNECESSARY; } - rcu_read_lock(); mroute6_sk = rcu_dereference(mrt->mroute_sk); if (!mroute6_sk) { - rcu_read_unlock(); kfree_skb(skb); return -EINVAL; } @@ -1119,7 +1125,7 @@ static int ip6mr_cache_report(struct mr_table *mrt, struct sk_buff *pkt, /* Deliver to user space multicast routing algorithms */ ret = sock_queue_rcv_skb(mroute6_sk, skb); - rcu_read_unlock(); + if (ret < 0) { net_warn_ratelimited("mroute6: pending queue full, dropping entries\n"); kfree_skb(skb); @@ -1243,7 +1249,7 @@ static int ip6mr_device_event(struct notifier_block *this, ip6mr_for_each_table(mrt, net) { v = &mrt->vif_table[0]; for (ct = 0; ct < mrt->maxvif; ct++, v++) { - if (v->dev == dev) + if (rcu_access_pointer(v->dev) == dev) mif6_delete(mrt, ct, 1, NULL); } } @@ -1262,7 +1268,7 @@ static int ip6mr_dump(struct net *net, struct notifier_block *nb, struct netlink_ext_ack *extack) { return mr_dump(net, nb, RTNL_FAMILY_IP6MR, ip6mr_rules_dump, - ip6mr_mr_table_iter, &mrt_lock, extack); + ip6mr_mr_table_iter, extack); } static struct notifier_block ip6_mr_notifier = { @@ -1386,7 +1392,7 @@ int __init ip6_mr_init(void) } #endif err = rtnl_register_module(THIS_MODULE, RTNL_FAMILY_IP6MR, RTM_GETROUTE, - NULL, ip6mr_rtm_dumproute, 0); + ip6mr_rtm_getroute, ip6mr_rtm_dumproute, 0); if (err == 0) return 0; @@ -1437,12 +1443,12 @@ static int ip6mr_mfc_add(struct net *net, struct mr_table *mrt, &mfc->mf6cc_mcastgrp.sin6_addr, parent); rcu_read_unlock(); if (c) { - write_lock_bh(&mrt_lock); + spin_lock(&mrt_lock); c->_c.mfc_parent = mfc->mf6cc_parent; ip6mr_update_thresholds(mrt, &c->_c, ttls); if (!mrtsock) c->_c.mfc_flags |= MFC_STATIC; - write_unlock_bh(&mrt_lock); + spin_unlock(&mrt_lock); call_ip6mr_mfc_entry_notifiers(net, FIB_EVENT_ENTRY_REPLACE, c, mrt->id); mr6_netlink_event(mrt, c, RTM_NEWROUTE); @@ -1560,7 +1566,7 @@ static int ip6mr_sk_init(struct mr_table *mrt, struct sock *sk) struct net *net = sock_net(sk); rtnl_lock(); - write_lock_bh(&mrt_lock); + spin_lock(&mrt_lock); if (rtnl_dereference(mrt->mroute_sk)) { err = -EADDRINUSE; } else { @@ -1568,7 +1574,7 @@ static int ip6mr_sk_init(struct mr_table *mrt, struct sock *sk) sock_set_flag(sk, SOCK_RCU_FREE); atomic_inc(&net->ipv6.devconf_all->mc_forwarding); } - write_unlock_bh(&mrt_lock); + spin_unlock(&mrt_lock); if (!err) inet6_netconf_notify_devconf(net, RTM_NEWNETCONF, @@ -1598,14 +1604,14 @@ int ip6mr_sk_done(struct sock *sk) rtnl_lock(); ip6mr_for_each_table(mrt, net) { if (sk == rtnl_dereference(mrt->mroute_sk)) { - write_lock_bh(&mrt_lock); + spin_lock(&mrt_lock); RCU_INIT_POINTER(mrt->mroute_sk, NULL); /* Note that mroute_sk had SOCK_RCU_FREE set, * so the RCU grace period before sk freeing * is guaranteed by sk_destruct() */ atomic_dec(&devconf->mc_forwarding); - write_unlock_bh(&mrt_lock); + spin_unlock(&mrt_lock); inet6_netconf_notify_devconf(net, RTM_NEWNETCONF, NETCONFA_MC_FORWARDING, NETCONFA_IFINDEX_ALL, @@ -1891,20 +1897,20 @@ int ip6mr_ioctl(struct sock *sk, int cmd, void __user *arg) if (vr.mifi >= mrt->maxvif) return -EINVAL; vr.mifi = array_index_nospec(vr.mifi, mrt->maxvif); - read_lock(&mrt_lock); + rcu_read_lock(); vif = &mrt->vif_table[vr.mifi]; if (VIF_EXISTS(mrt, vr.mifi)) { - vr.icount = vif->pkt_in; - vr.ocount = vif->pkt_out; - vr.ibytes = vif->bytes_in; - vr.obytes = vif->bytes_out; - read_unlock(&mrt_lock); + vr.icount = READ_ONCE(vif->pkt_in); + vr.ocount = READ_ONCE(vif->pkt_out); + vr.ibytes = READ_ONCE(vif->bytes_in); + vr.obytes = READ_ONCE(vif->bytes_out); + rcu_read_unlock(); if (copy_to_user(arg, &vr, sizeof(vr))) return -EFAULT; return 0; } - read_unlock(&mrt_lock); + rcu_read_unlock(); return -EADDRNOTAVAIL; case SIOCGETSGCNT_IN6: if (copy_from_user(&sr, arg, sizeof(sr))) @@ -1966,20 +1972,20 @@ int ip6mr_compat_ioctl(struct sock *sk, unsigned int cmd, void __user *arg) if (vr.mifi >= mrt->maxvif) return -EINVAL; vr.mifi = array_index_nospec(vr.mifi, mrt->maxvif); - read_lock(&mrt_lock); + rcu_read_lock(); vif = &mrt->vif_table[vr.mifi]; if (VIF_EXISTS(mrt, vr.mifi)) { - vr.icount = vif->pkt_in; - vr.ocount = vif->pkt_out; - vr.ibytes = vif->bytes_in; - vr.obytes = vif->bytes_out; - read_unlock(&mrt_lock); + vr.icount = READ_ONCE(vif->pkt_in); + vr.ocount = READ_ONCE(vif->pkt_out); + vr.ibytes = READ_ONCE(vif->bytes_in); + vr.obytes = READ_ONCE(vif->bytes_out); + rcu_read_unlock(); if (copy_to_user(arg, &vr, sizeof(vr))) return -EFAULT; return 0; } - read_unlock(&mrt_lock); + rcu_read_unlock(); return -EADDRNOTAVAIL; case SIOCGETSGCNT_IN6: if (copy_from_user(&sr, arg, sizeof(sr))) @@ -2021,21 +2027,22 @@ static inline int ip6mr_forward2_finish(struct net *net, struct sock *sk, struct static int ip6mr_forward2(struct net *net, struct mr_table *mrt, struct sk_buff *skb, int vifi) { - struct ipv6hdr *ipv6h; struct vif_device *vif = &mrt->vif_table[vifi]; - struct net_device *dev; + struct net_device *vif_dev; + struct ipv6hdr *ipv6h; struct dst_entry *dst; struct flowi6 fl6; - if (!vif->dev) + vif_dev = vif_dev_read(vif); + if (!vif_dev) goto out_free; #ifdef CONFIG_IPV6_PIMSM_V2 if (vif->flags & MIFF_REGISTER) { - vif->pkt_out++; - vif->bytes_out += skb->len; - vif->dev->stats.tx_bytes += skb->len; - vif->dev->stats.tx_packets++; + WRITE_ONCE(vif->pkt_out, vif->pkt_out + 1); + WRITE_ONCE(vif->bytes_out, vif->bytes_out + skb->len); + vif_dev->stats.tx_bytes += skb->len; + vif_dev->stats.tx_packets++; ip6mr_cache_report(mrt, skb, vifi, MRT6MSG_WHOLEPKT); goto out_free; } @@ -2068,14 +2075,13 @@ static int ip6mr_forward2(struct net *net, struct mr_table *mrt, * not mrouter) cannot join to more than one interface - it will * result in receiving multiple packets. */ - dev = vif->dev; - skb->dev = dev; - vif->pkt_out++; - vif->bytes_out += skb->len; + skb->dev = vif_dev; + WRITE_ONCE(vif->pkt_out, vif->pkt_out + 1); + WRITE_ONCE(vif->bytes_out, vif->bytes_out + skb->len); /* We are about to write */ /* XXX: extension headers? */ - if (skb_cow(skb, sizeof(*ipv6h) + LL_RESERVED_SPACE(dev))) + if (skb_cow(skb, sizeof(*ipv6h) + LL_RESERVED_SPACE(vif_dev))) goto out_free; ipv6h = ipv6_hdr(skb); @@ -2084,7 +2090,7 @@ static int ip6mr_forward2(struct net *net, struct mr_table *mrt, IP6CB(skb)->flags |= IP6SKB_FORWARDED; return NF_HOOK(NFPROTO_IPV6, NF_INET_FORWARD, - net, NULL, skb, skb->dev, dev, + net, NULL, skb, skb->dev, vif_dev, ip6mr_forward2_finish); out_free: @@ -2092,17 +2098,20 @@ out_free: return 0; } +/* Called with rcu_read_lock() */ static int ip6mr_find_vif(struct mr_table *mrt, struct net_device *dev) { int ct; - for (ct = mrt->maxvif - 1; ct >= 0; ct--) { - if (mrt->vif_table[ct].dev == dev) + /* Pairs with WRITE_ONCE() in mif6_delete()/mif6_add() */ + for (ct = READ_ONCE(mrt->maxvif) - 1; ct >= 0; ct--) { + if (rcu_access_pointer(mrt->vif_table[ct].dev) == dev) break; } return ct; } +/* Called under rcu_read_lock() */ static void ip6_mr_forward(struct net *net, struct mr_table *mrt, struct net_device *dev, struct sk_buff *skb, struct mfc6_cache *c) @@ -2122,20 +2131,16 @@ static void ip6_mr_forward(struct net *net, struct mr_table *mrt, /* For an (*,G) entry, we only check that the incoming * interface is part of the static tree. */ - rcu_read_lock(); cache_proxy = mr_mfc_find_any_parent(mrt, vif); if (cache_proxy && - cache_proxy->_c.mfc_un.res.ttls[true_vifi] < 255) { - rcu_read_unlock(); + cache_proxy->_c.mfc_un.res.ttls[true_vifi] < 255) goto forward; - } - rcu_read_unlock(); } /* * Wrong interface: drop packet and (maybe) send PIM assert. */ - if (mrt->vif_table[vif].dev != dev) { + if (rcu_access_pointer(mrt->vif_table[vif].dev) != dev) { c->_c.mfc_un.res.wrong_if++; if (true_vifi >= 0 && mrt->mroute_do_assert && @@ -2159,8 +2164,10 @@ static void ip6_mr_forward(struct net *net, struct mr_table *mrt, } forward: - mrt->vif_table[vif].pkt_in++; - mrt->vif_table[vif].bytes_in += skb->len; + WRITE_ONCE(mrt->vif_table[vif].pkt_in, + mrt->vif_table[vif].pkt_in + 1); + WRITE_ONCE(mrt->vif_table[vif].bytes_in, + mrt->vif_table[vif].bytes_in + skb->len); /* * Forward the frame @@ -2238,7 +2245,6 @@ int ip6_mr_input(struct sk_buff *skb) return err; } - read_lock(&mrt_lock); cache = ip6mr_cache_find(mrt, &ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr); if (!cache) { @@ -2259,19 +2265,15 @@ int ip6_mr_input(struct sk_buff *skb) vif = ip6mr_find_vif(mrt, dev); if (vif >= 0) { int err = ip6mr_cache_unresolved(mrt, vif, skb, dev); - read_unlock(&mrt_lock); return err; } - read_unlock(&mrt_lock); kfree_skb(skb); return -ENODEV; } ip6_mr_forward(net, mrt, dev, skb, cache); - read_unlock(&mrt_lock); - return 0; } @@ -2287,7 +2289,7 @@ int ip6mr_get_route(struct net *net, struct sk_buff *skb, struct rtmsg *rtm, if (!mrt) return -ENOENT; - read_lock(&mrt_lock); + rcu_read_lock(); cache = ip6mr_cache_find(mrt, &rt->rt6i_src.addr, &rt->rt6i_dst.addr); if (!cache && skb->dev) { int vif = ip6mr_find_vif(mrt, skb->dev); @@ -2305,14 +2307,14 @@ int ip6mr_get_route(struct net *net, struct sk_buff *skb, struct rtmsg *rtm, dev = skb->dev; if (!dev || (vif = ip6mr_find_vif(mrt, dev)) < 0) { - read_unlock(&mrt_lock); + rcu_read_unlock(); return -ENODEV; } /* really correct? */ skb2 = alloc_skb(sizeof(struct ipv6hdr), GFP_ATOMIC); if (!skb2) { - read_unlock(&mrt_lock); + rcu_read_unlock(); return -ENOMEM; } @@ -2335,13 +2337,13 @@ int ip6mr_get_route(struct net *net, struct sk_buff *skb, struct rtmsg *rtm, iph->daddr = rt->rt6i_dst.addr; err = ip6mr_cache_unresolved(mrt, vif, skb2, dev); - read_unlock(&mrt_lock); + rcu_read_unlock(); return err; } err = mr_fill_mroute(mrt, skb, &cache->_c, rtm); - read_unlock(&mrt_lock); + rcu_read_unlock(); return err; } @@ -2460,7 +2462,7 @@ static size_t mrt6msg_netlink_msgsize(size_t payloadlen) return len; } -static void mrt6msg_netlink_event(struct mr_table *mrt, struct sk_buff *pkt) +static void mrt6msg_netlink_event(const struct mr_table *mrt, struct sk_buff *pkt) { struct net *net = read_pnet(&mrt->net); struct nlmsghdr *nlh; @@ -2508,6 +2510,95 @@ errout: rtnl_set_sk_err(net, RTNLGRP_IPV6_MROUTE_R, -ENOBUFS); } +static const struct nla_policy ip6mr_getroute_policy[RTA_MAX + 1] = { + [RTA_SRC] = NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)), + [RTA_DST] = NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)), + [RTA_TABLE] = { .type = NLA_U32 }, +}; + +static int ip6mr_rtm_valid_getroute_req(struct sk_buff *skb, + const struct nlmsghdr *nlh, + struct nlattr **tb, + struct netlink_ext_ack *extack) +{ + struct rtmsg *rtm; + int err; + + err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, ip6mr_getroute_policy, + extack); + if (err) + return err; + + rtm = nlmsg_data(nlh); + if ((rtm->rtm_src_len && rtm->rtm_src_len != 128) || + (rtm->rtm_dst_len && rtm->rtm_dst_len != 128) || + rtm->rtm_tos || rtm->rtm_table || rtm->rtm_protocol || + rtm->rtm_scope || rtm->rtm_type || rtm->rtm_flags) { + NL_SET_ERR_MSG_MOD(extack, + "Invalid values in header for multicast route get request"); + return -EINVAL; + } + + if ((tb[RTA_SRC] && !rtm->rtm_src_len) || + (tb[RTA_DST] && !rtm->rtm_dst_len)) { + NL_SET_ERR_MSG_MOD(extack, "rtm_src_len and rtm_dst_len must be 128 for IPv6"); + return -EINVAL; + } + + return 0; +} + +static int ip6mr_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, + struct netlink_ext_ack *extack) +{ + struct net *net = sock_net(in_skb->sk); + struct in6_addr src = {}, grp = {}; + struct nlattr *tb[RTA_MAX + 1]; + struct mfc6_cache *cache; + struct mr_table *mrt; + struct sk_buff *skb; + u32 tableid; + int err; + + err = ip6mr_rtm_valid_getroute_req(in_skb, nlh, tb, extack); + if (err < 0) + return err; + + if (tb[RTA_SRC]) + src = nla_get_in6_addr(tb[RTA_SRC]); + if (tb[RTA_DST]) + grp = nla_get_in6_addr(tb[RTA_DST]); + tableid = tb[RTA_TABLE] ? nla_get_u32(tb[RTA_TABLE]) : 0; + + mrt = ip6mr_get_table(net, tableid ?: RT_TABLE_DEFAULT); + if (!mrt) { + NL_SET_ERR_MSG_MOD(extack, "MR table does not exist"); + return -ENOENT; + } + + /* entries are added/deleted only under RTNL */ + rcu_read_lock(); + cache = ip6mr_cache_find(mrt, &src, &grp); + rcu_read_unlock(); + if (!cache) { + NL_SET_ERR_MSG_MOD(extack, "MR cache entry not found"); + return -ENOENT; + } + + skb = nlmsg_new(mr6_msgsize(false, mrt->maxvif), GFP_KERNEL); + if (!skb) + return -ENOBUFS; + + err = ip6mr_fill_mroute(mrt, skb, NETLINK_CB(in_skb).portid, + nlh->nlmsg_seq, cache, RTM_NEWROUTE, 0); + if (err < 0) { + kfree_skb(skb); + return err; + } + + return rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid); +} + static int ip6mr_rtm_dumproute(struct sk_buff *skb, struct netlink_callback *cb) { const struct nlmsghdr *nlh = cb->nlh; |