mptcp: refactor incoming options parsing
Subsystems (3)
Paolo Abeni
cover mptcp: refactor incoming options parsing
This is an alternative solution for the pending option parsing solution, previous attempt here:
https://lists.01.org/hyperkitty/list/mptcp(a)lists.01.org/thread/7FW4C3W4WS42GWTRWFQWCTCUYGKLIHFJ/
instead of adding (many) explicitly 'clear MPTCP options' operation to the TCP code, it moves completely the options parsing into the MPTCP code - see patch 2/6.
This addresses also the problems with the previois attempt for no-mptcp-opt pkts.
patch 1/6 is a needed pre-requisites, while patch 3,4,5 are smaller somewhat releated fixes.
patch 6/6 is just a minor clean-up and likely could be squashed into 6/6.
Stress test with syzkaller repro survived a few hours run without any splash, while I can hit WARN_ON() on top of current export branch + pending -net fixes in a few mins with the same test.
Overall I'm personally more confident in this approach than in the previous one. Any comments more than welcome! Specifically I'm wondering if this should target -net or net-next...
Paolo Abeni (6):
mptcp: consolidate synack processing.
mptcp: move option parsing into mptcp_incoming_options()
mptcp: avoid a WARN on bad input.
mptcp: fix 'use_ack' option access.
mptcp: initialize data_fin field for mpc packets
mptcp: cleanup option parsing
include/linux/tcp.h | 51 ------------------------ include/net/mptcp.h | 3 -- net/ipv4/tcp_input.c | 7 ---- net/mptcp/options.c | 95 +++++++++++++++++++------------------------- net/mptcp/protocol.c | 6 +-- net/mptcp/protocol.h | 43 +++++++++++++++++++- net/mptcp/subflow.c | 81 +++++++++++++++++++++++-------------- 7 files changed, 137 insertions(+), 149 deletions(-)
Signature · 1 line
2.21.1
Paolo Abeni
RFC PATCH 1/6 mptcp: consolidate synack processing.
Currently the MPTCP code uses 2 hooks to process syn-ack packets, mptcp_rcv_synsent() and the sk_rx_dst_set() callback.
We can drop the first, moving the relevant code into the latter, reducing the hooking into the TCP code. This is also needed by the next patch.
--- include/net/mptcp.h | 1 - net/ipv4/tcp_input.c | 3 --- net/mptcp/options.c | 22 ---------------------- net/mptcp/subflow.c | 27 ++++++++++++++++++++++++--- 4 files changed, 24 insertions(+), 29 deletions(-)
include/net/mptcp.h +0 -1
@@ -74,7 +74,6 @@ void mptcp_parse_option(const struct sk_buff *skb, const unsigned char *ptr, int opsize, struct tcp_options_received *opt_rx);bool mptcp_syn_options (struct sock *sk,const struct sk_buff *skb,unsigned int *size,struct mptcp_out_options *opts); -void mptcp_rcv_synsent (struct sock *sk);bool mptcp_synack_options (const struct request_sock *req,unsigned int *size,struct mptcp_out_options *opts); bool mptcp_established_options(struct sock *sk, struct sk_buff *skb,
net/ipv4/tcp_input.c +0 -3
@@ -5990,9 +5990,6 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb,tcp_sync_mss (sk, icsk->icsk_pmtu_cookie);tcp_initialize_rcv_mss (sk); -if (sk_is_mptcp (sk)) -mptcp_rcv_synsent (sk); -/* Remember, tcp_poll() does not lock socket! * Change state from SYN-SENT only after copied_seq * is initialized. */
net/mptcp/options.c +0 -22
@@ -344,28 +344,6 @@ bool mptcp_syn_options(struct sock *sk, const struct sk_buff *skb,return false; } -void mptcp_rcv_synsent (struct sock *sk) -{ -struct mptcp_subflow_context *subflow =mptcp_subflow_ctx (sk); -struct tcp_sock *tp =tcp_sk (sk); - -if (subflow->request_mptcp && tp->rx_opt.mptcp.mp_capable) { - subflow->mp_capable =1 ; - subflow->can_ack =1 ; - subflow->remote_key = tp->rx_opt.mptcp.sndr_key; -pr_debug ("subflow=%p, remote_key=%llu" , subflow, - subflow->remote_key); - }else if (subflow->request_join && tp->rx_opt.mptcp.mp_join) { - subflow->mp_join =1 ; - subflow->thmac = tp->rx_opt.mptcp.thmac; - subflow->remote_nonce = tp->rx_opt.mptcp.nonce; -pr_debug ("subflow=%p, thmac=%llu, remote_nonce=%u" , subflow, - subflow->thmac, subflow->remote_nonce); - }else if (subflow->request_mptcp) { -tcp_sk (sk)->is_mptcp =0 ; - } -} -/* MP_JOIN client subflow must wait for 4th ack before sending any data: * TCP can't schedule delack timer before the subflow is fully established. * MPTCP uses the delack timer to do 3rd ack retransmissions
net/mptcp/subflow.c +24 -3
@@ -222,6 +222,7 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb) {struct mptcp_subflow_context *subflow =mptcp_subflow_ctx (sk);struct sock *parent = subflow->conn; +struct tcp_sock *tp =tcp_sk (sk); subflow->icsk_af_ops->sk_rx_dst_set (sk, skb);
@@ -230,14 +231,35 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb) parent->sk_state_change (parent); } -if (subflow->conn_finished || !tcp_sk (sk)->is_mptcp) +/* be sure no special action on any packet other than syn-ack */ +if (subflow->conn_finished) +return ; + + subflow->conn_finished =1 ; + +if (subflow->request_mptcp && tp->rx_opt.mptcp.mp_capable) { + subflow->mp_capable =1 ; + subflow->can_ack =1 ; + subflow->remote_key = tp->rx_opt.mptcp.sndr_key; +pr_debug ("subflow=%p, remote_key=%llu" , subflow, + subflow->remote_key); + }else if (subflow->request_join && tp->rx_opt.mptcp.mp_join) { + subflow->mp_join =1 ; + subflow->thmac = tp->rx_opt.mptcp.thmac; + subflow->remote_nonce = tp->rx_opt.mptcp.nonce; +pr_debug ("subflow=%p, thmac=%llu, remote_nonce=%u" , subflow, + subflow->thmac, subflow->remote_nonce); + }else if (subflow->request_mptcp) { +tcp_sk (sk)->is_mptcp =0 ; + } + +if (!tcp_sk (sk)->is_mptcp)return ;if (subflow->mp_capable) {pr_debug ("subflow=%p, remote_key=%llu" ,mptcp_subflow_ctx (sk), subflow->remote_key);mptcp_finish_connect (sk); - subflow->conn_finished =1 ;if (skb) {pr_debug ("synack seq=%u" ,TCP_SKB_CB (skb)->seq);
@@ -264,7 +286,6 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)if (!mptcp_finish_join (sk)) goto do_reset; - subflow->conn_finished =1 ;MPTCP_INC_STATS (sock_net (sk),MPTCP_MIB_JOINSYNACKRX ); }else { do_reset:
Signature · 1 line
2.21.1
Paolo Abeni
RFC PATCH 2/6 mptcp: move option parsing into mptcp_incoming_options()
In reply to Paolo AbeniThe mptcp_options_received structure carries several per packet flags (mp_capable, mp_join, etc.). Such field must be cleared on each packet, even on dropped ones or packet not carrying any MPTCP options, but the current mptcp code clear them only on TCP option reset.
On several races/corner cases we end-up with stray bits in incoming options, leading to WARN_ON splats. e.g.:
[ 171.164906] Bad mapping: ssn=32714 map_seq=1 map_data_len=32713 [ 171.165006] WARNING: CPU: 1 PID: 5026 at net/mptcp/subflow.c:533 warn_bad_map (linux-mptcp/net/mptcp/subflow.c:533 linux-mptcp/net/mptcp/subflow.c:531) [ 171.167632] Modules linked in: ip6_vti ip_vti ip_gre ipip sit tunnel4 ip_tunnel geneve ip6_udp_tunnel udp_tunnel macsec macvtap tap ipvlan macvlan 8021q garp mrp xfrm_interface veth netdevsim nlmon dummy team bonding vcan bridge stp llc ip6_gre gre ip6_tunnel tunnel6 tun binfmt_misc intel_rapl_msr intel_rapl_common rfkill kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel joydev virtio_balloon pcspkr i2c_piix4 sunrpc ip_tables xfs libcrc32c crc32c_intel serio_raw virtio_console ata_generic virtio_blk virtio_net net_failover failover ata_piix libata [ 171.199464] CPU: 1 PID: 5026 Comm: repro Not tainted 5.7.0-rc1.mptcp_f227fdf5d388+ #95 [ 171.200886] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-2.fc30 04/01/2014 [ 171.202546] RIP: 0010:warn_bad_map (linux-mptcp/net/mptcp/subflow.c:533 linux-mptcp/net/mptcp/subflow.c:531) [ 171.206537] Code: c1 ea 03 0f b6 14 02 48 89 f8 83 e0 07 83 c0 03 38 d0 7c 04 84 d2 75 1d 8b 55 3c 44 89 e6 48 c7 c7 20 51 13 95 e8 37 8b 22 fe <0f> 0b 48 83 c4 08 5b 5d 41 5c c3 89 4c 24 04 e8 db d6 94 fe 8b 4c [ 171.220473] RSP: 0018:ffffc90000150560 EFLAGS: 00010282 [ 171.221639] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 [ 171.223108] RDX: 0000000000000000 RSI: 0000000000000008 RDI: fffff5200002a09e [ 171.224388] RBP: ffff8880aa6e3c00 R08: 0000000000000001 R09: fffffbfff2ec9955 [ 171.225706] R10: ffffffff9764caa7 R11: fffffbfff2ec9954 R12: 0000000000007fca [ 171.227211] R13: ffff8881066f4a7f R14: ffff8880aa6e3c00 R15: 0000000000000020 [ 171.228460] FS: 00007f8623719740(0000) GS:ffff88810be00000(0000) knlGS:0000000000000000 [ 171.230065] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 171.231303] CR2: 00007ffdab190a50 CR3: 00000001038ea006 CR4: 0000000000160ee0 [ 171.232586] Call Trace: [ 171.233109] <IRQ> [ 171.233531] get_mapping_status (linux-mptcp/net/mptcp/subflow.c:691) [ 171.234371] mptcp_subflow_data_available (linux-mptcp/net/mptcp/subflow.c:736 linux-mptcp/net/mptcp/subflow.c:832) [ 171.238181] subflow_state_change (linux-mptcp/net/mptcp/subflow.c:1085 (discriminator 1)) [ 171.239066] tcp_fin (linux-mptcp/net/ipv4/tcp_input.c:4217) [ 171.240123] tcp_data_queue (linux-mptcp/./include/linux/compiler.h:199 linux-mptcp/net/ipv4/tcp_input.c:4822) [ 171.245083] tcp_rcv_established (linux-mptcp/./include/linux/skbuff.h:1785 linux-mptcp/./include/net/tcp.h:1774 linux-mptcp/./include/net/tcp.h:1847 linux-mptcp/net/ipv4/tcp_input.c:5238 linux-mptcp/net/ipv4/tcp_input.c:5730) [ 171.254089] tcp_v4_rcv (linux-mptcp/./include/linux/spinlock.h:393 linux-mptcp/net/ipv4/tcp_ipv4.c:2009) [ 171.258969] ip_protocol_deliver_rcu (linux-mptcp/net/ipv4/ip_input.c:204 (discriminator 1)) [ 171.260214] ip_local_deliver_finish (linux-mptcp/./include/linux/rcupdate.h:651 linux-mptcp/net/ipv4/ip_input.c:232) [ 171.261389] ip_local_deliver (linux-mptcp/./include/linux/netfilter.h:307 linux-mptcp/./include/linux/netfilter.h:301 linux-mptcp/net/ipv4/ip_input.c:252) [ 171.265884] ip_rcv (linux-mptcp/./include/linux/netfilter.h:307 linux-mptcp/./include/linux/netfilter.h:301 linux-mptcp/net/ipv4/ip_input.c:539) [ 171.273666] process_backlog (linux-mptcp/./include/linux/rcupdate.h:651 linux-mptcp/net/core/dev.c:6135) [ 171.275328] net_rx_action (linux-mptcp/net/core/dev.c:6572 linux-mptcp/net/core/dev.c:6640) [ 171.280472] __do_softirq (linux-mptcp/./arch/x86/include/asm/jump_label.h:25 linux-mptcp/./include/linux/jump_label.h:200 linux-mptcp/./include/trace/events/irq.h:142 linux-mptcp/kernel/softirq.c:293) [ 171.281379] do_softirq_own_stack (linux-mptcp/arch/x86/entry/entry_64.S:1083) [ 171.282358] </IRQ>
We could address the issue clearing explicitly the relevant fields in several places - tcp_parse_option, tcp_fast_parse_options, possibly others.
Instead we move mptcp option parsing into the already existing mptcp ingress hook, so that we need to clear the fields in a single place.
This allows us dropping an mptcp hook from the TCP code and removing the quite large mptcp_options_received from the tcp_sock struct. As the filp side the MPTCP sockets will traverse the option space twice (in tcp_parse_option() and in mptcp_incoming_options(). That looks acceptable: we already do that for syn and 3rd ack packets, plain TCP socket will benefit from it, and even MPTCP sockets will experience better code locality, reducing the jumps between TCP and MPTCP code.
--- include/linux/tcp.h | 51 ---------------------------------- include/net/mptcp.h | 2 -- net/ipv4/tcp_input.c | 4 --- net/mptcp/options.c | 66 +++++++++++++++++++++++++------------------- net/mptcp/protocol.c | 6 ++-- net/mptcp/protocol.h | 43 +++++++++++++++++++++++++++-- net/mptcp/subflow.c | 65 ++++++++++++++++++++++--------------------- 7 files changed, 114 insertions(+), 123 deletions(-)
include/linux/tcp.h +0 -51
@@ -78,47 +78,6 @@ struct tcp_sack_block {#define TCP_SACK_SEEN (1 << 0)/*1 = peer is SACK capable, */ #define TCP_DSACK_SEEN (1 << 2)/*1 = DSACK was received from peer*/ -#if IS_ENABLED (CONFIG_MPTCP ) -struct mptcp_options_received { -u64 sndr_key; -u64 rcvr_key; -u64 data_ack; -u64 data_seq; -u32 subflow_seq; -u16 data_len; -u16 mp_capable :1 , - mp_join :1 , - dss :1 , - add_addr :1 , - rm_addr :1 , - family :4 , - echo :1 , - backup :1 ; -u32 token; -u32 nonce; -u64 thmac; -u8 hmac[20 ]; -u8 join_id; -u8 use_map:1 , - dsn64:1 , - data_fin:1 , - use_ack:1 , - ack64:1 , - mpc_map:1 , - __unused:2 ; -u8 addr_id; -u8 rm_id; -union { -struct in_addr addr; -#if IS_ENABLED (CONFIG_MPTCP_IPV6 ) -struct in6_addr addr6; -#endif - }; -u64 ahmac; -u16 port; -}; -#endif -struct tcp_options_received {/* PAWS/RTTM data */ int ts_recent_stamp;/* Time we stored ts_recent (for aging) */
@@ -136,9 +95,6 @@ struct tcp_options_received {u8 num_sacks;/* Number of SACK blocks */ u16 user_mss;/* mss requested by user in ioctl */ u16 mss_clamp;/* Maximal mss, negotiated at connection setup */ -#if IS_ENABLED (CONFIG_MPTCP ) -struct mptcp_options_received mptcp; -#endif };static inline void tcp_clear_options (struct tcp_options_received *rx_opt)
@@ -148,13 +104,6 @@ static inline void tcp_clear_options(struct tcp_options_received *rx_opt)#if IS_ENABLED (CONFIG_SMC ) rx_opt->smc_ok =0 ;#endif -#if IS_ENABLED (CONFIG_MPTCP ) - rx_opt->mptcp.mp_capable =0 ; - rx_opt->mptcp.mp_join =0 ; - rx_opt->mptcp.add_addr =0 ; - rx_opt->mptcp.rm_addr =0 ; - rx_opt->mptcp.dss =0 ; -#endif }/* This is the max number of SACKS that we'll generate and process. It's safe
include/net/mptcp.h +0 -2
@@ -70,8 +70,6 @@ static inline bool rsk_is_mptcp(const struct request_sock *req)void mptcp_space (const struct sock *ssk,int *space,int *full_space); -void mptcp_parse_option (const struct sk_buff *skb,const unsigned char *ptr, -int opsize,struct tcp_options_received *opt_rx);bool mptcp_syn_options (struct sock *sk,const struct sk_buff *skb,unsigned int *size,struct mptcp_out_options *opts); bool mptcp_synack_options(const struct request_sock *req, unsigned int *size,
net/ipv4/tcp_input.c +0 -4
@@ -3926,10 +3926,6 @@ void tcp_parse_options(const struct net *net,*/ break ;#endif -case TCPOPT_MPTCP : -mptcp_parse_option (skb, ptr, opsize, opt_rx); -break ; -case TCPOPT_FASTOPEN : tcp_parse_fastopen_option( opsize - TCPOLEN_FASTOPEN_BASE,
net/mptcp/options.c +37 -29
127 lines not on this page · raw message
net/mptcp/protocol.c +3 -3
@@ -1342,7 +1342,7 @@ static struct ipv6_pinfo *mptcp_inet6_sk(const struct sock *sk)#endif struct sock *mptcp_sk_clone (const struct sock *sk, -const struct tcp_options_received *opt_rx, +const struct mptcp_options_received *mp_opt,struct request_sock *req) {struct mptcp_subflow_request_sock *subflow_req =mptcp_subflow_rsk (req);
@@ -1381,9 +1381,9 @@ struct sock *mptcp_sk_clone(const struct sock *sk, msk->write_seq = subflow_req->idsn +1 ;atomic64_set (&msk->snd_una, msk->write_seq); -if (opt_rx->mptcp.mp_capable) { +if (mp_opt->mp_capable) { msk->can_ack = true; - msk->remote_key = opt_rx->mptcp.sndr_key; + msk->remote_key = mp_opt->sndr_key;mptcp_crypto_key_sha (msk->remote_key,NULL , &ack_seq); ack_seq++; msk->ack_seq = ack_seq;
net/mptcp/protocol.h +41 -2
57 lines not on this page · raw message
net/mptcp/subflow.c +33 -32
157 lines not on this page · raw message
Signature · 1 line
2.21.1
Paolo Abeni
RFC PATCH 3/6 mptcp: avoid a WARN on bad input.
In reply to Paolo AbeniSyzcaller has found a way to trigger the WARN_ON_ONCE condition in check_fully_established().
The root cause is actually a legit fallback to TCP scenario, so replace the WARN with a plain message on a more strict condition.
--- net/mptcp/options.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
net/mptcp/options.c +2 -2
@@ -703,8 +703,6 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *sk, goto fully_established; } -WARN_ON_ONCE (subflow->can_ack); -/* If the first established packet does not contain MP_CAPABLE + data * then fallback to TCP */
@@ -714,6 +712,8 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *sk,return false; } +if (unlikely (!READ_ONCE (msk->pm.server_side))) +pr_warn_once ("bogus mpc option on established client sk" ); subflow->fully_established =1 ; subflow->remote_key = mp_opt->sndr_key; subflow->can_ack =1 ;
Signature · 1 line
2.21.1
Paolo Abeni
RFC PATCH 4/6 mptcp: fix 'use_ack' option access.
In reply to Paolo AbeniThe mentioned RX option field is initialized only for DSS packet, we must access it only if 'dss' is set too, or the subflow will end-up in a bad status, leading to RFC violations.
--- net/mptcp/options.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/mptcp/options.c +1 -1
@@ -695,7 +695,7 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *sk,if (TCP_SKB_CB (skb)->seq != subflow->ssn_offset +1 )return subflow->mp_capable; -if (mp_opt->use_ack) { +if (mp_opt->dss && mp_opt->use_ack) {/* subflows are fully established as soon as we get any * additional ack. */
Signature · 1 line
2.21.1
Paolo Abeni
RFC PATCH 5/6 mptcp: initialize data_fin field for mpc packets
In reply to Paolo AbeniWhen parsing MPC+data packets we set the dss field, so we must also initialize the data_fin, or we can find stray value there.
--- net/mptcp/options.c | 1 + 1 file changed, 1 insertion(+)
net/mptcp/options.c +1 -0
@@ -858,6 +858,7 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb, mpext->subflow_seq =1 ; mpext->dsn64 =1 ; mpext->mpc_map =1 ; + mpext->data_fin =0 ; }else { mpext->data_seq = mp_opt.data_seq; mpext->subflow_seq = mp_opt.subflow_seq;
Signature · 1 line
2.21.1
Paolo Abeni
RFC PATCH 6/6 mptcp: cleanup option parsing
In reply to Paolo AbeniThe mptcp option status is always initialized by mptcp_get_options(), we can clean-up a couple of call sites open codying such task.
---
this could be squashed into
"mptcp: move option parsing into mptcp_incoming_options()", kept here to since the latter is already pretty big, but no strong opinion against squashing
--- net/mptcp/subflow.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
net/mptcp/subflow.c +3 -2
@@ -128,7 +128,6 @@ static void subflow_init_req(struct request_sock *req,pr_debug ("subflow_req=%p, listener=%p" , subflow_req, listener); -memset (&mp_opt,0 ,sizeof (mp_opt));mptcp_get_options (skb, &mp_opt); subflow_req->mp_capable =0 ;
@@ -425,6 +424,9 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,pr_debug ("listener=%p, req=%p, conn=%p" , listener, req, listener->conn); +/* we need later a valid 'mp_capable' value even when options are not +* parsed +*/ mp_opt.mp_capable =0 ;if (tcp_rsk (req)->is_mptcp ==0 ) goto create_child;
@@ -452,7 +454,6 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk, fallback = true; }else if (subflow_req->mp_join) { fallback_is_fatal = true; - mp_opt.mp_join =0 ;mptcp_get_options (skb, &mp_opt);if (!mp_opt.mp_join || !subflow_hmac_valid (req, &mp_opt)) {
Signature · 1 line
2.21.1
Mat Martineau
[MPTCP] Re: [RFC PATCH 2/6] mptcp: move option parsing into mptcp_incoming_options()
In reply to Paolo AbeniQuoted 52 lines · Paolo Abeni
The
mptcp_options_receivedstructure carries several per packet flags (mp_capable,mp_join, etc.). Such field must be cleared on each packet, even on dropped ones or packet not carrying any MPTCP options, but the current mptcp code clear them only on TCP option reset.On several races/corner cases we end-up with stray bits in incoming options, leading to
WARN_ONsplats. e.g.:[ 171.164906] Bad mapping: ssn=32714 map_seq=1 map_data_len=32713 [ 171.165006] WARNING: CPU: 1 PID: 5026 at net/mptcp/subflow.c:533 warn_bad_map (linux-mptcp/net/mptcp/subflow.c:533 linux-mptcp/net/mptcp/subflow.c:531) [ 171.167632] Modules linked in: ip6_vti ip_vti ip_gre ipip sit tunnel4 ip_tunnel geneve ip6_udp_tunnel udp_tunnel macsec macvtap tap ipvlan macvlan 8021q garp mrp xfrm_interface veth netdevsim nlmon dummy team bonding vcan bridge stp llc ip6_gre gre ip6_tunnel tunnel6 tun binfmt_misc intel_rapl_msr intel_rapl_common rfkill kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel joydev virtio_balloon pcspkr i2c_piix4 sunrpc ip_tables xfs libcrc32c crc32c_intel serio_raw virtio_console ata_generic virtio_blk virtio_net net_failover failover ata_piix libata [ 171.199464] CPU: 1 PID: 5026 Comm: repro Not tainted 5.7.0-rc1.mptcp_f227fdf5d388+ #95 [ 171.200886] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-2.fc30 04/01/2014 [ 171.202546] RIP: 0010:warn_bad_map (linux-mptcp/net/mptcp/subflow.c:533 linux-mptcp/net/mptcp/subflow.c:531) [ 171.206537] Code: c1 ea 03 0f b6 14 02 48 89 f8 83 e0 07 83 c0 03 38 d0 7c 04 84 d2 75 1d 8b 55 3c 44 89 e6 48 c7 c7 20 51 13 95 e8 37 8b 22 fe <0f> 0b 48 83 c4 08 5b 5d 41 5c c3 89 4c 24 04 e8 db d6 94 fe 8b 4c [ 171.220473] RSP: 0018:ffffc90000150560 EFLAGS: 00010282 [ 171.221639] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 [ 171.223108] RDX: 0000000000000000 RSI: 0000000000000008 RDI: fffff5200002a09e [ 171.224388] RBP: ffff8880aa6e3c00 R08: 0000000000000001 R09: fffffbfff2ec9955 [ 171.225706] R10: ffffffff9764caa7 R11: fffffbfff2ec9954 R12: 0000000000007fca [ 171.227211] R13: ffff8881066f4a7f R14: ffff8880aa6e3c00 R15: 0000000000000020 [ 171.228460] FS: 00007f8623719740(0000) GS:ffff88810be00000(0000) knlGS:0000000000000000 [ 171.230065] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 171.231303] CR2: 00007ffdab190a50 CR3: 00000001038ea006 CR4: 0000000000160ee0 [ 171.232586] Call Trace: [ 171.233109] <IRQ> [ 171.233531] get_mapping_status (linux-mptcp/net/mptcp/subflow.c:691) [ 171.234371] mptcp_subflow_data_available (linux-mptcp/net/mptcp/subflow.c:736 linux-mptcp/net/mptcp/subflow.c:832) [ 171.238181] subflow_state_change (linux-mptcp/net/mptcp/subflow.c:1085 (discriminator 1)) [ 171.239066] tcp_fin (linux-mptcp/net/ipv4/tcp_input.c:4217) [ 171.240123] tcp_data_queue (linux-mptcp/./include/linux/compiler.h:199 linux-mptcp/net/ipv4/tcp_input.c:4822) [ 171.245083] tcp_rcv_established (linux-mptcp/./include/linux/skbuff.h:1785 linux-mptcp/./include/net/tcp.h:1774 linux-mptcp/./include/net/tcp.h:1847 linux-mptcp/net/ipv4/tcp_input.c:5238 linux-mptcp/net/ipv4/tcp_input.c:5730) [ 171.254089] tcp_v4_rcv (linux-mptcp/./include/linux/spinlock.h:393 linux-mptcp/net/ipv4/tcp_ipv4.c:2009) [ 171.258969] ip_protocol_deliver_rcu (linux-mptcp/net/ipv4/ip_input.c:204 (discriminator 1)) [ 171.260214] ip_local_deliver_finish (linux-mptcp/./include/linux/rcupdate.h:651 linux-mptcp/net/ipv4/ip_input.c:232) [ 171.261389] ip_local_deliver (linux-mptcp/./include/linux/netfilter.h:307 linux-mptcp/./include/linux/netfilter.h:301 linux-mptcp/net/ipv4/ip_input.c:252) [ 171.265884] ip_rcv (linux-mptcp/./include/linux/netfilter.h:307 linux-mptcp/./include/linux/netfilter.h:301 linux-mptcp/net/ipv4/ip_input.c:539) [ 171.273666] process_backlog (linux-mptcp/./include/linux/rcupdate.h:651 linux-mptcp/net/core/dev.c:6135) [ 171.275328] net_rx_action (linux-mptcp/net/core/dev.c:6572 linux-mptcp/net/core/dev.c:6640) [ 171.280472] __do_softirq (linux-mptcp/./arch/x86/include/asm/jump_label.h:25 linux-mptcp/./include/linux/jump_label.h:200 linux-mptcp/./include/trace/events/irq.h:142 linux-mptcp/kernel/softirq.c:293) [ 171.281379] do_softirq_own_stack (linux-mptcp/arch/x86/entry/entry_64.S:1083) [ 171.282358] </IRQ>We could address the issue clearing explicitly the relevant fields in several places -
tcp_parse_option,tcp_fast_parse_options, possibly others.Instead we move mptcp option parsing into the already existing mptcp ingress hook, so that we need to clear the fields in a single place.
Quoted 9 lines
This allows us dropping an mptcp hook from the TCP code and removing the quite large
mptcp_options_receivedfrom thetcp_sockstruct. As the filp side the MPTCP sockets will traverse the option space twice (intcp_parse_option()and inmptcp_incoming_options(). That looks acceptable: we already do that for syn and 3rd ack packets, plain TCP socket will benefit from it, and even MPTCP sockets will experience better code locality, reducing the jumps between TCP and MPTCP code.
At first glance it seems unconventional to move the option parsing out of tcp_parse_option, but the parsed mptcp information isn't used at the TCP layer anyway. I agree that the tradeoffs look acceptable.
I think it would be good to target -net.
Thanks,
--
Mat Martineau
Intel