- Patch 1: better handle DSS corruptions from a bugged peer: reducing warnings, doing a fallback or a reset depending on the subflow state. For >= v5.7.
- Patch 2: fix DSS corruption due to large pmtu xmit, where MPTCP was not taken into account. For >= v5.6.
- Patch 3: fallback when MPTCP opts are dropped after the first data packet, instead of resetting the connection. For >= v5.6.
- Patch 4: restrict the removal of a subflow to other closing states, a better fix, for a recent one. For >= v5.10.
Signed-off-by: Matthieu Baerts (NGI0) matttbe@kernel.org --- Matthieu Baerts (NGI0) (2): mptcp: fallback when MPTCP opts are dropped after 1st data mptcp: pm: do not remove closing subflows
Paolo Abeni (2): mptcp: handle consistently DSS corruption tcp: fix mptcp DSS corruption due to large pmtu xmit
net/ipv4/tcp_output.c | 5 +---- net/mptcp/mib.c | 2 ++ net/mptcp/mib.h | 2 ++ net/mptcp/pm_netlink.c | 3 ++- net/mptcp/protocol.c | 24 +++++++++++++++++++++--- net/mptcp/subflow.c | 6 ++++-- 6 files changed, 32 insertions(+), 10 deletions(-) --- base-commit: f15b8d6eb63874230e36a45dd24239050a6f6250 change-id: 20241008-net-mptcp-fallback-fixes-16a9afee238e
Best regards,
From: Paolo Abeni pabeni@redhat.com
Bugged peer implementation can send corrupted DSS options, consistently hitting a few warning in the data path. Use DEBUG_NET assertions, to avoid the splat on some builds and handle consistently the error, dumping related MIBs and performing fallback and/or reset according to the subflow type.
Fixes: 6771bfd9ee24 ("mptcp: update mptcp ack sequence from work queue") Cc: stable@vger.kernel.org Signed-off-by: Paolo Abeni pabeni@redhat.com Reviewed-by: Matthieu Baerts (NGI0) matttbe@kernel.org Signed-off-by: Matthieu Baerts (NGI0) matttbe@kernel.org --- net/mptcp/mib.c | 2 ++ net/mptcp/mib.h | 2 ++ net/mptcp/protocol.c | 24 +++++++++++++++++++++--- net/mptcp/subflow.c | 4 +++- 4 files changed, 28 insertions(+), 4 deletions(-)
diff --git a/net/mptcp/mib.c b/net/mptcp/mib.c index 38c2efc82b948d9afd35c4d5bcd45d9e5422a88d..ad88bd3c58dffed8335eedb43ca6290418e3c4f4 100644 --- a/net/mptcp/mib.c +++ b/net/mptcp/mib.c @@ -32,6 +32,8 @@ static const struct snmp_mib mptcp_snmp_list[] = { SNMP_MIB_ITEM("MPJoinSynTxBindErr", MPTCP_MIB_JOINSYNTXBINDERR), SNMP_MIB_ITEM("MPJoinSynTxConnectErr", MPTCP_MIB_JOINSYNTXCONNECTERR), SNMP_MIB_ITEM("DSSNotMatching", MPTCP_MIB_DSSNOMATCH), + SNMP_MIB_ITEM("DSSCorruptionFallback", MPTCP_MIB_DSSCORRUPTIONFALLBACK), + SNMP_MIB_ITEM("DSSCorruptionReset", MPTCP_MIB_DSSCORRUPTIONRESET), SNMP_MIB_ITEM("InfiniteMapTx", MPTCP_MIB_INFINITEMAPTX), SNMP_MIB_ITEM("InfiniteMapRx", MPTCP_MIB_INFINITEMAPRX), SNMP_MIB_ITEM("DSSNoMatchTCP", MPTCP_MIB_DSSTCPMISMATCH), diff --git a/net/mptcp/mib.h b/net/mptcp/mib.h index c8ffe18a872217afa24e3af212fe90a3fb8d1c7f..3206cdda8bb1067f9a8354fd45deed86b67ac7da 100644 --- a/net/mptcp/mib.h +++ b/net/mptcp/mib.h @@ -27,6 +27,8 @@ enum linux_mptcp_mib_field { MPTCP_MIB_JOINSYNTXBINDERR, /* Not able to bind() the address when sending a SYN + MP_JOIN */ MPTCP_MIB_JOINSYNTXCONNECTERR, /* Not able to connect() when sending a SYN + MP_JOIN */ MPTCP_MIB_DSSNOMATCH, /* Received a new mapping that did not match the previous one */ + MPTCP_MIB_DSSCORRUPTIONFALLBACK,/* DSS corruption detected, fallback */ + MPTCP_MIB_DSSCORRUPTIONRESET, /* DSS corruption detected, MPJ subflow reset */ MPTCP_MIB_INFINITEMAPTX, /* Sent an infinite mapping */ MPTCP_MIB_INFINITEMAPRX, /* Received an infinite mapping */ MPTCP_MIB_DSSTCPMISMATCH, /* DSS-mapping did not map with TCP's sequence numbers */ diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index c2317919fc148a67a81ded795359bd613c9b0dff..6d0e201c3eb26aa6ca0ff27e5a65cb6f911012f2 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -620,6 +620,18 @@ static bool mptcp_check_data_fin(struct sock *sk) return ret; }
+static void mptcp_dss_corruption(struct mptcp_sock *msk, struct sock *ssk) +{ + if (READ_ONCE(msk->allow_infinite_fallback)) { + MPTCP_INC_STATS(sock_net(ssk), + MPTCP_MIB_DSSCORRUPTIONFALLBACK); + mptcp_do_fallback(ssk); + } else { + MPTCP_INC_STATS(sock_net(ssk), MPTCP_MIB_DSSCORRUPTIONRESET); + mptcp_subflow_reset(ssk); + } +} + static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk, struct sock *ssk, unsigned int *bytes) @@ -692,10 +704,16 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk, moved += len; seq += len;
- if (WARN_ON_ONCE(map_remaining < len)) - break; + if (unlikely(map_remaining < len)) { + DEBUG_NET_WARN_ON_ONCE(1); + mptcp_dss_corruption(msk, ssk); + } } else { - WARN_ON_ONCE(!fin); + if (unlikely(!fin)) { + DEBUG_NET_WARN_ON_ONCE(1); + mptcp_dss_corruption(msk, ssk); + } + sk_eat_skb(ssk, skb); done = true; } diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index 1040b3b9696b74b12c1f8c027e5a323c558900f0..e1046a696ab5c79a2cef79870eb79637b432fcd5 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -975,8 +975,10 @@ static bool skb_is_fully_mapped(struct sock *ssk, struct sk_buff *skb) unsigned int skb_consumed;
skb_consumed = tcp_sk(ssk)->copied_seq - TCP_SKB_CB(skb)->seq; - if (WARN_ON_ONCE(skb_consumed >= skb->len)) + if (unlikely(skb_consumed >= skb->len)) { + DEBUG_NET_WARN_ON_ONCE(1); return true; + }
return skb->len - skb_consumed <= subflow->map_data_len - mptcp_subflow_get_map_offset(subflow);
From: Paolo Abeni pabeni@redhat.com
Syzkaller was able to trigger a DSS corruption:
TCP: request_sock_subflow_v4: Possible SYN flooding on port [::]:20002. Sending cookies. ------------[ cut here ]------------ WARNING: CPU: 0 PID: 5227 at net/mptcp/protocol.c:695 __mptcp_move_skbs_from_subflow+0x20a9/0x21f0 net/mptcp/protocol.c:695 Modules linked in: CPU: 0 UID: 0 PID: 5227 Comm: syz-executor350 Not tainted 6.11.0-syzkaller-08829-gaf9c191ac2a0 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/06/2024 RIP: 0010:__mptcp_move_skbs_from_subflow+0x20a9/0x21f0 net/mptcp/protocol.c:695 Code: 0f b6 dc 31 ff 89 de e8 b5 dd ea f5 89 d8 48 81 c4 50 01 00 00 5b 41 5c 41 5d 41 5e 41 5f 5d c3 cc cc cc cc e8 98 da ea f5 90 <0f> 0b 90 e9 47 ff ff ff e8 8a da ea f5 90 0f 0b 90 e9 99 e0 ff ff RSP: 0018:ffffc90000006db8 EFLAGS: 00010246 RAX: ffffffff8ba9df18 RBX: 00000000000055f0 RCX: ffff888030023c00 RDX: 0000000000000100 RSI: 00000000000081e5 RDI: 00000000000055f0 RBP: 1ffff110062bf1ae R08: ffffffff8ba9cf12 R09: 1ffff110062bf1b8 R10: dffffc0000000000 R11: ffffed10062bf1b9 R12: 0000000000000000 R13: dffffc0000000000 R14: 00000000700cec61 R15: 00000000000081e5 FS: 000055556679c380(0000) GS:ffff8880b8600000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000020287000 CR3: 0000000077892000 CR4: 00000000003506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <IRQ> move_skbs_to_msk net/mptcp/protocol.c:811 [inline] mptcp_data_ready+0x29c/0xa90 net/mptcp/protocol.c:854 subflow_data_ready+0x34a/0x920 net/mptcp/subflow.c:1490 tcp_data_queue+0x20fd/0x76c0 net/ipv4/tcp_input.c:5283 tcp_rcv_established+0xfba/0x2020 net/ipv4/tcp_input.c:6237 tcp_v4_do_rcv+0x96d/0xc70 net/ipv4/tcp_ipv4.c:1915 tcp_v4_rcv+0x2dc0/0x37f0 net/ipv4/tcp_ipv4.c:2350 ip_protocol_deliver_rcu+0x22e/0x440 net/ipv4/ip_input.c:205 ip_local_deliver_finish+0x341/0x5f0 net/ipv4/ip_input.c:233 NF_HOOK+0x3a4/0x450 include/linux/netfilter.h:314 NF_HOOK+0x3a4/0x450 include/linux/netfilter.h:314 __netif_receive_skb_one_core net/core/dev.c:5662 [inline] __netif_receive_skb+0x2bf/0x650 net/core/dev.c:5775 process_backlog+0x662/0x15b0 net/core/dev.c:6107 __napi_poll+0xcb/0x490 net/core/dev.c:6771 napi_poll net/core/dev.c:6840 [inline] net_rx_action+0x89b/0x1240 net/core/dev.c:6962 handle_softirqs+0x2c5/0x980 kernel/softirq.c:554 do_softirq+0x11b/0x1e0 kernel/softirq.c:455 </IRQ> <TASK> __local_bh_enable_ip+0x1bb/0x200 kernel/softirq.c:382 local_bh_enable include/linux/bottom_half.h:33 [inline] rcu_read_unlock_bh include/linux/rcupdate.h:919 [inline] __dev_queue_xmit+0x1764/0x3e80 net/core/dev.c:4451 dev_queue_xmit include/linux/netdevice.h:3094 [inline] neigh_hh_output include/net/neighbour.h:526 [inline] neigh_output include/net/neighbour.h:540 [inline] ip_finish_output2+0xd41/0x1390 net/ipv4/ip_output.c:236 ip_local_out net/ipv4/ip_output.c:130 [inline] __ip_queue_xmit+0x118c/0x1b80 net/ipv4/ip_output.c:536 __tcp_transmit_skb+0x2544/0x3b30 net/ipv4/tcp_output.c:1466 tcp_transmit_skb net/ipv4/tcp_output.c:1484 [inline] tcp_mtu_probe net/ipv4/tcp_output.c:2547 [inline] tcp_write_xmit+0x641d/0x6bf0 net/ipv4/tcp_output.c:2752 __tcp_push_pending_frames+0x9b/0x360 net/ipv4/tcp_output.c:3015 tcp_push_pending_frames include/net/tcp.h:2107 [inline] tcp_data_snd_check net/ipv4/tcp_input.c:5714 [inline] tcp_rcv_established+0x1026/0x2020 net/ipv4/tcp_input.c:6239 tcp_v4_do_rcv+0x96d/0xc70 net/ipv4/tcp_ipv4.c:1915 sk_backlog_rcv include/net/sock.h:1113 [inline] __release_sock+0x214/0x350 net/core/sock.c:3072 release_sock+0x61/0x1f0 net/core/sock.c:3626 mptcp_push_release net/mptcp/protocol.c:1486 [inline] __mptcp_push_pending+0x6b5/0x9f0 net/mptcp/protocol.c:1625 mptcp_sendmsg+0x10bb/0x1b10 net/mptcp/protocol.c:1903 sock_sendmsg_nosec net/socket.c:730 [inline] __sock_sendmsg+0x1a6/0x270 net/socket.c:745 ____sys_sendmsg+0x52a/0x7e0 net/socket.c:2603 ___sys_sendmsg net/socket.c:2657 [inline] __sys_sendmsg+0x2aa/0x390 net/socket.c:2686 do_syscall_x64 arch/x86/entry/common.c:52 [inline] do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7fb06e9317f9 Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007ffe2cfd4f98 EFLAGS: 00000246 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 00007fb06e97f468 RCX: 00007fb06e9317f9 RDX: 0000000000000000 RSI: 0000000020000080 RDI: 0000000000000005 RBP: 00007fb06e97f446 R08: 0000555500000000 R09: 0000555500000000 R10: 0000555500000000 R11: 0000000000000246 R12: 00007fb06e97f406 R13: 0000000000000001 R14: 00007ffe2cfd4fe0 R15: 0000000000000003 </TASK>
Additionally syzkaller provided a nice reproducer. The repro enables pmtu on the loopback device, leading to tcp_mtu_probe() generating very large probe packets.
tcp_can_coalesce_send_queue_head() currently does not check for mptcp-level invariants, and allowed the creation of cross-DSS probes, leading to the mentioned corruption.
Address the issue teaching tcp_can_coalesce_send_queue_head() about mptcp using the tcp_skb_can_collapse(), also reducing the code duplication.
Fixes: 85712484110d ("tcp: coalesce/collapse must respect MPTCP extensions") Cc: stable@vger.kernel.org Reported-by: syzbot+d1bff73460e33101f0e7@syzkaller.appspotmail.com Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/513 Signed-off-by: Paolo Abeni pabeni@redhat.com Acked-by: Matthieu Baerts (NGI0) matttbe@kernel.org Signed-off-by: Matthieu Baerts (NGI0) matttbe@kernel.org --- net/ipv4/tcp_output.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 4fd746bd4d54f621601b20c3821e71370a4a615a..68804fd01dafc48101ca8c3f15991dbe02a0dd6f 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2342,10 +2342,7 @@ static bool tcp_can_coalesce_send_queue_head(struct sock *sk, int len) if (len <= skb->len) break;
- if (unlikely(TCP_SKB_CB(skb)->eor) || - tcp_has_tx_tstamp(skb) || - !skb_pure_zcopy_same(skb, next) || - skb_frags_readable(skb) != skb_frags_readable(next)) + if (tcp_has_tx_tstamp(skb) || !tcp_skb_can_collapse(skb, next)) return false;
len -= skb->len;
As reported by Christoph [1], before this patch, an MPTCP connection was wrongly reset when a host received a first data packet with MPTCP options after the 3wHS, but got the next ones without.
According to the MPTCP v1 specs [2], a fallback should happen in this case, because the host didn't receive a DATA_ACK from the other peer, nor receive data for more than the initial window which implies a DATA_ACK being received by the other peer.
The patch here re-uses the same logic as the one used in other places: by looking at allow_infinite_fallback, which is disabled at the creation of an additional subflow. It's not looking at the first DATA_ACK (or implying one received from the other side) as suggested by the RFC, but it is in continuation with what was already done, which is safer, and it fixes the reported issue. The next step, looking at this first DATA_ACK, is tracked in [4].
This patch has been validated using the following Packetdrill script:
0 socket(..., SOCK_STREAM, IPPROTO_MPTCP) = 3 +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +0 bind(3, ..., ...) = 0 +0 listen(3, 1) = 0
// 3WHS is OK +0.0 < S 0:0(0) win 65535 <mss 1460, sackOK, nop, nop, nop, wscale 6, mpcapable v1 flags[flag_h] nokey> +0.0 > S. 0:0(0) ack 1 <mss 1460, nop, nop, sackOK, nop, wscale 8, mpcapable v1 flags[flag_h] key[skey]> +0.1 < . 1:1(0) ack 1 win 2048 <mpcapable v1 flags[flag_h] key[ckey=2, skey]> +0 accept(3, ..., ...) = 4
// Data from the client with valid MPTCP options (no DATA_ACK: normal) +0.1 < P. 1:501(500) ack 1 win 2048 <mpcapable v1 flags[flag_h] key[skey, ckey] mpcdatalen 500, nop, nop> // From here, the MPTCP options will be dropped by a middlebox +0.0 > . 1:1(0) ack 501 <dss dack8=501 dll=0 nocs>
+0.1 read(4, ..., 500) = 500 +0 write(4, ..., 100) = 100
// The server replies with data, still thinking MPTCP is being used +0.0 > P. 1:101(100) ack 501 <dss dack8=501 dsn8=1 ssn=1 dll=100 nocs, nop, nop> // But the client already did a fallback to TCP, because the two previous packets have been received without MPTCP options +0.1 < . 501:501(0) ack 101 win 2048
+0.0 < P. 501:601(100) ack 101 win 2048 // The server should fallback to TCP, not reset: it didn't get a DATA_ACK, nor data for more than the initial window +0.0 > . 101:101(0) ack 601
Note that this script requires Packetdrill with MPTCP support, see [3].
Fixes: dea2b1ea9c70 ("mptcp: do not reset MP_CAPABLE subflow on mapping errors") Cc: stable@vger.kernel.org Reported-by: Christoph Paasch cpaasch@apple.com Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/518 [1] Link: https://datatracker.ietf.org/doc/html/rfc8684#name-fallback [2] Link: https://github.com/multipath-tcp/packetdrill [3] Link: https://github.com/multipath-tcp/mptcp_net-next/issues/519 [4] Reviewed-by: Paolo Abeni pabeni@redhat.com Signed-off-by: Matthieu Baerts (NGI0) matttbe@kernel.org --- net/mptcp/subflow.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index e1046a696ab5c79a2cef79870eb79637b432fcd5..25dde81bcb7575958635aaf14a5b8e9a5005e05f 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -1282,7 +1282,7 @@ static bool subflow_can_fallback(struct mptcp_subflow_context *subflow) else if (READ_ONCE(msk->csum_enabled)) return !subflow->valid_csum_seen; else - return !subflow->fully_established; + return READ_ONCE(msk->allow_infinite_fallback); }
static void mptcp_subflow_fail(struct mptcp_sock *msk, struct sock *ssk)
In a previous fix, the in-kernel path-manager has been modified not to retrigger the removal of a subflow if it was already closed, e.g. when the initial subflow is removed, but kept in the subflows list.
To be complete, this fix should also skip the subflows that are in any closing state: mptcp_close_ssk() will initiate the closure, but the switch to the TCP_CLOSE state depends on the other peer.
Fixes: 58e1b66b4e4b ("mptcp: pm: do not remove already closed subflows") Cc: stable@vger.kernel.org Suggested-by: Paolo Abeni pabeni@redhat.com Acked-by: Paolo Abeni pabeni@redhat.com Signed-off-by: Matthieu Baerts (NGI0) matttbe@kernel.org --- net/mptcp/pm_netlink.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c index 64fe0e7d87d7323583ef9ee5909c833a78e727c2..f6f0a38a0750f82bc909f02a75beec980d951f1f 100644 --- a/net/mptcp/pm_netlink.c +++ b/net/mptcp/pm_netlink.c @@ -860,7 +860,8 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk, int how = RCV_SHUTDOWN | SEND_SHUTDOWN; u8 id = subflow_get_local_id(subflow);
- if (inet_sk_state_load(ssk) == TCP_CLOSE) + if ((1 << inet_sk_state_load(ssk)) & + (TCPF_FIN_WAIT1 | TCPF_FIN_WAIT2 | TCPF_CLOSING | TCPF_CLOSE)) continue; if (rm_type == MPTCP_MIB_RMADDR && remote_id != rm_id) continue;
Hello:
This series was applied to netdev/net.git (main) by Jakub Kicinski kuba@kernel.org:
On Tue, 08 Oct 2024 13:04:51 +0200 you wrote:
Patch 1: better handle DSS corruptions from a bugged peer: reducing warnings, doing a fallback or a reset depending on the subflow state. For >= v5.7.
Patch 2: fix DSS corruption due to large pmtu xmit, where MPTCP was not taken into account. For >= v5.6.
[...]
Here is the summary with links: - [net,1/4] mptcp: handle consistently DSS corruption https://git.kernel.org/netdev/net/c/e32d262c89e2 - [net,2/4] tcp: fix mptcp DSS corruption due to large pmtu xmit https://git.kernel.org/netdev/net/c/4dabcdf58121 - [net,3/4] mptcp: fallback when MPTCP opts are dropped after 1st data https://git.kernel.org/netdev/net/c/119d51e225fe - [net,4/4] mptcp: pm: do not remove closing subflows https://git.kernel.org/netdev/net/c/db0a37b7ac27
You are awesome, thank you!
linux-stable-mirror@lists.linaro.org
-
Matthieu Baerts (NGI0) -
patchwork-bot+netdevbpf＠kernel.org