From 263e1201a2c324b60b15ecda5de9ebf1e7293e31 Mon Sep 17 00:00:00 2001
From: Paolo Abeni <pabeni@redhat.com>
Date: Thu, 30 Apr 2020 15:01:51 +0200
Subject: 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.

v1 -> v2:
 - use local tcp sock ptr instead of casting the sk variable
   several times - DaveM

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp_input.c | 3 ---
 1 file changed, 3 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index bf4ced9273e8..81425542da44 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -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. */
-- 
cgit 


From cfde141ea3faa30e362bbdb5c28001bbbdb0b8e0 Mon Sep 17 00:00:00 2001
From: Paolo Abeni <pabeni@redhat.com>
Date: Thu, 30 Apr 2020 15:01:52 +0200
Subject: mptcp: move option parsing into mptcp_incoming_options()

The mptcp_options_received structure carries several per
packet flags (mp_capable, mp_join, etc.). Such fields must
be cleared on each packet, even on dropped ones or packet
not carrying any MPTCP options, but the current mptcp
code clears 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 the 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. On the flip 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.

v1 -> v2:
 - rebased on current '-net' tree

Fixes: 648ef4b88673 ("mptcp: Implement MPTCP receive path")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 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, 115 insertions(+), 122 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 421c99c12291..4f8159e90ce1 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -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
diff --git a/include/net/mptcp.h b/include/net/mptcp.h
index 4ecfa7d5e0c7..3bce2019e4da 100644
--- a/include/net/mptcp.h
+++ b/include/net/mptcp.h
@@ -68,8 +68,6 @@ static inline bool rsk_is_mptcp(const struct request_sock *req)
 	return tcp_rsk(req)->is_mptcp;
 }
 
-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,
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 81425542da44..b996dc1069c5 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -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,
diff --git a/net/mptcp/options.c b/net/mptcp/options.c
index 8fea686a5562..eadbd59586e4 100644
--- a/net/mptcp/options.c
+++ b/net/mptcp/options.c
@@ -16,10 +16,10 @@ static bool mptcp_cap_flag_sha256(u8 flags)
 	return (flags & MPTCP_CAP_FLAG_MASK) == MPTCP_CAP_HMAC_SHA256;
 }
 
-void mptcp_parse_option(const struct sk_buff *skb, const unsigned char *ptr,
-			int opsize, struct tcp_options_received *opt_rx)
+static void mptcp_parse_option(const struct sk_buff *skb,
+			       const unsigned char *ptr, int opsize,
+			       struct mptcp_options_received *mp_opt)
 {
-	struct mptcp_options_received *mp_opt = &opt_rx->mptcp;
 	u8 subtype = *ptr >> 4;
 	int expected_opsize;
 	u8 version;
@@ -283,12 +283,20 @@ void mptcp_parse_option(const struct sk_buff *skb, const unsigned char *ptr,
 }
 
 void mptcp_get_options(const struct sk_buff *skb,
-		       struct tcp_options_received *opt_rx)
+		       struct mptcp_options_received *mp_opt)
 {
-	const unsigned char *ptr;
 	const struct tcphdr *th = tcp_hdr(skb);
-	int length = (th->doff * 4) - sizeof(struct tcphdr);
+	const unsigned char *ptr;
+	int length;
 
+	/* initialize option status */
+	mp_opt->mp_capable = 0;
+	mp_opt->mp_join = 0;
+	mp_opt->add_addr = 0;
+	mp_opt->rm_addr = 0;
+	mp_opt->dss = 0;
+
+	length = (th->doff * 4) - sizeof(struct tcphdr);
 	ptr = (const unsigned char *)(th + 1);
 
 	while (length > 0) {
@@ -308,7 +316,7 @@ void mptcp_get_options(const struct sk_buff *skb,
 			if (opsize > length)
 				return;	/* don't parse partial options */
 			if (opcode == TCPOPT_MPTCP)
-				mptcp_parse_option(skb, ptr, opsize, opt_rx);
+				mptcp_parse_option(skb, ptr, opsize, mp_opt);
 			ptr += opsize - 2;
 			length -= opsize;
 		}
@@ -797,41 +805,41 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb,
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
 	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
-	struct mptcp_options_received *mp_opt;
+	struct mptcp_options_received mp_opt;
 	struct mptcp_ext *mpext;
 
-	mp_opt = &opt_rx->mptcp;
-	if (!check_fully_established(msk, sk, subflow, skb, mp_opt))
+	mptcp_get_options(skb, &mp_opt);
+	if (!check_fully_established(msk, sk, subflow, skb, &mp_opt))
 		return;
 
-	if (mp_opt->add_addr && add_addr_hmac_valid(msk, mp_opt)) {
+	if (mp_opt.add_addr && add_addr_hmac_valid(msk, &mp_opt)) {
 		struct mptcp_addr_info addr;
 
-		addr.port = htons(mp_opt->port);
-		addr.id = mp_opt->addr_id;
-		if (mp_opt->family == MPTCP_ADDR_IPVERSION_4) {
+		addr.port = htons(mp_opt.port);
+		addr.id = mp_opt.addr_id;
+		if (mp_opt.family == MPTCP_ADDR_IPVERSION_4) {
 			addr.family = AF_INET;
-			addr.addr = mp_opt->addr;
+			addr.addr = mp_opt.addr;
 		}
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
-		else if (mp_opt->family == MPTCP_ADDR_IPVERSION_6) {
+		else if (mp_opt.family == MPTCP_ADDR_IPVERSION_6) {
 			addr.family = AF_INET6;
-			addr.addr6 = mp_opt->addr6;
+			addr.addr6 = mp_opt.addr6;
 		}
 #endif
-		if (!mp_opt->echo)
+		if (!mp_opt.echo)
 			mptcp_pm_add_addr_received(msk, &addr);
-		mp_opt->add_addr = 0;
+		mp_opt.add_addr = 0;
 	}
 
-	if (!mp_opt->dss)
+	if (!mp_opt.dss)
 		return;
 
 	/* we can't wait for recvmsg() to update the ack_seq, otherwise
 	 * monodirectional flows will stuck
 	 */
-	if (mp_opt->use_ack)
-		update_una(msk, mp_opt);
+	if (mp_opt.use_ack)
+		update_una(msk, &mp_opt);
 
 	mpext = skb_ext_add(skb, SKB_EXT_MPTCP);
 	if (!mpext)
@@ -839,8 +847,8 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb,
 
 	memset(mpext, 0, sizeof(*mpext));
 
-	if (mp_opt->use_map) {
-		if (mp_opt->mpc_map) {
+	if (mp_opt.use_map) {
+		if (mp_opt.mpc_map) {
 			/* this is an MP_CAPABLE carrying MPTCP data
 			 * we know this map the first chunk of data
 			 */
@@ -851,12 +859,12 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb,
 			mpext->dsn64 = 1;
 			mpext->mpc_map = 1;
 		} else {
-			mpext->data_seq = mp_opt->data_seq;
-			mpext->subflow_seq = mp_opt->subflow_seq;
-			mpext->dsn64 = mp_opt->dsn64;
-			mpext->data_fin = mp_opt->data_fin;
+			mpext->data_seq = mp_opt.data_seq;
+			mpext->subflow_seq = mp_opt.subflow_seq;
+			mpext->dsn64 = mp_opt.dsn64;
+			mpext->data_fin = mp_opt.data_fin;
 		}
-		mpext->data_len = mp_opt->data_len;
+		mpext->data_len = mp_opt.data_len;
 		mpext->use_map = 1;
 	}
 }
diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 6e0188f5d3f3..e1f23016ed3f 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -1334,7 +1334,7 @@ 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);
@@ -1373,9 +1373,9 @@ 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;
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index a2b3048037d0..e4ca6320ce76 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -91,6 +91,45 @@
 #define MPTCP_WORK_RTX		2
 #define MPTCP_WORK_EOF		3
 
+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;
+};
+
 static inline __be32 mptcp_option(u8 subopt, u8 len, u8 nib, u8 field)
 {
 	return htonl((TCPOPT_MPTCP << 24) | (len << 16) | (subopt << 12) |
@@ -331,10 +370,10 @@ int mptcp_proto_v6_init(void);
 #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);
 void mptcp_get_options(const struct sk_buff *skb,
-		       struct tcp_options_received *opt_rx);
+		       struct mptcp_options_received *mp_opt);
 
 void mptcp_finish_connect(struct sock *sk);
 void mptcp_data_ready(struct sock *sk, struct sock *ssk);
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index 84f6408594c9..bad998529767 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -124,12 +124,11 @@ static void subflow_init_req(struct request_sock *req,
 {
 	struct mptcp_subflow_context *listener = mptcp_subflow_ctx(sk_listener);
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
-	struct tcp_options_received rx_opt;
+	struct mptcp_options_received mp_opt;
 
 	pr_debug("subflow_req=%p, listener=%p", subflow_req, listener);
 
-	memset(&rx_opt.mptcp, 0, sizeof(rx_opt.mptcp));
-	mptcp_get_options(skb, &rx_opt);
+	mptcp_get_options(skb, &mp_opt);
 
 	subflow_req->mp_capable = 0;
 	subflow_req->mp_join = 0;
@@ -142,16 +141,16 @@ static void subflow_init_req(struct request_sock *req,
 		return;
 #endif
 
-	if (rx_opt.mptcp.mp_capable) {
+	if (mp_opt.mp_capable) {
 		SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_MPCAPABLEPASSIVE);
 
-		if (rx_opt.mptcp.mp_join)
+		if (mp_opt.mp_join)
 			return;
-	} else if (rx_opt.mptcp.mp_join) {
+	} else if (mp_opt.mp_join) {
 		SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_JOINSYNRX);
 	}
 
-	if (rx_opt.mptcp.mp_capable && listener->request_mptcp) {
+	if (mp_opt.mp_capable && listener->request_mptcp) {
 		int err;
 
 		err = mptcp_token_new_request(req);
@@ -159,13 +158,13 @@ static void subflow_init_req(struct request_sock *req,
 			subflow_req->mp_capable = 1;
 
 		subflow_req->ssn_offset = TCP_SKB_CB(skb)->seq;
-	} else if (rx_opt.mptcp.mp_join && listener->request_mptcp) {
+	} else if (mp_opt.mp_join && listener->request_mptcp) {
 		subflow_req->ssn_offset = TCP_SKB_CB(skb)->seq;
 		subflow_req->mp_join = 1;
-		subflow_req->backup = rx_opt.mptcp.backup;
-		subflow_req->remote_id = rx_opt.mptcp.join_id;
-		subflow_req->token = rx_opt.mptcp.token;
-		subflow_req->remote_nonce = rx_opt.mptcp.nonce;
+		subflow_req->backup = mp_opt.backup;
+		subflow_req->remote_id = mp_opt.join_id;
+		subflow_req->token = mp_opt.token;
+		subflow_req->remote_nonce = mp_opt.nonce;
 		pr_debug("token=%u, remote_nonce=%u", subflow_req->token,
 			 subflow_req->remote_nonce);
 		if (!subflow_token_join_request(req, skb)) {
@@ -221,6 +220,7 @@ static bool subflow_thmac_valid(struct mptcp_subflow_context *subflow)
 static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
+	struct mptcp_options_received mp_opt;
 	struct sock *parent = subflow->conn;
 	struct tcp_sock *tp = tcp_sk(sk);
 
@@ -237,16 +237,17 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 
 	subflow->conn_finished = 1;
 
-	if (subflow->request_mptcp && tp->rx_opt.mptcp.mp_capable) {
+	mptcp_get_options(skb, &mp_opt);
+	if (subflow->request_mptcp && mp_opt.mp_capable) {
 		subflow->mp_capable = 1;
 		subflow->can_ack = 1;
-		subflow->remote_key = tp->rx_opt.mptcp.sndr_key;
+		subflow->remote_key = mp_opt.sndr_key;
 		pr_debug("subflow=%p, remote_key=%llu", subflow,
 			 subflow->remote_key);
-	} else if (subflow->request_join && tp->rx_opt.mptcp.mp_join) {
+	} else if (subflow->request_join && mp_opt.mp_join) {
 		subflow->mp_join = 1;
-		subflow->thmac = tp->rx_opt.mptcp.thmac;
-		subflow->remote_nonce = tp->rx_opt.mptcp.nonce;
+		subflow->thmac = mp_opt.thmac;
+		subflow->remote_nonce = mp_opt.nonce;
 		pr_debug("subflow=%p, thmac=%llu, remote_nonce=%u", subflow,
 			 subflow->thmac, subflow->remote_nonce);
 	} else if (subflow->request_mptcp) {
@@ -343,7 +344,7 @@ drop:
 
 /* validate hmac received in third ACK */
 static bool subflow_hmac_valid(const struct request_sock *req,
-			       const struct tcp_options_received *rx_opt)
+			       const struct mptcp_options_received *mp_opt)
 {
 	const struct mptcp_subflow_request_sock *subflow_req;
 	u8 hmac[MPTCPOPT_HMAC_LEN];
@@ -360,7 +361,7 @@ static bool subflow_hmac_valid(const struct request_sock *req,
 			      subflow_req->local_nonce, hmac);
 
 	ret = true;
-	if (crypto_memneq(hmac, rx_opt->mptcp.hmac, sizeof(hmac)))
+	if (crypto_memneq(hmac, mp_opt->hmac, sizeof(hmac)))
 		ret = false;
 
 	sock_put((struct sock *)msk);
@@ -416,7 +417,7 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 {
 	struct mptcp_subflow_context *listener = mptcp_subflow_ctx(sk);
 	struct mptcp_subflow_request_sock *subflow_req;
-	struct tcp_options_received opt_rx;
+	struct mptcp_options_received mp_opt;
 	bool fallback_is_fatal = false;
 	struct sock *new_msk = NULL;
 	bool fallback = false;
@@ -424,7 +425,10 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 
 	pr_debug("listener=%p, req=%p, conn=%p", listener, req, listener->conn);
 
-	opt_rx.mptcp.mp_capable = 0;
+	/* 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;
 
@@ -439,22 +443,21 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 			goto create_msk;
 		}
 
-		mptcp_get_options(skb, &opt_rx);
-		if (!opt_rx.mptcp.mp_capable) {
+		mptcp_get_options(skb, &mp_opt);
+		if (!mp_opt.mp_capable) {
 			fallback = true;
 			goto create_child;
 		}
 
 create_msk:
-		new_msk = mptcp_sk_clone(listener->conn, &opt_rx, req);
+		new_msk = mptcp_sk_clone(listener->conn, &mp_opt, req);
 		if (!new_msk)
 			fallback = true;
 	} else if (subflow_req->mp_join) {
 		fallback_is_fatal = true;
-		opt_rx.mptcp.mp_join = 0;
-		mptcp_get_options(skb, &opt_rx);
-		if (!opt_rx.mptcp.mp_join ||
-		    !subflow_hmac_valid(req, &opt_rx)) {
+		mptcp_get_options(skb, &mp_opt);
+		if (!mp_opt.mp_join ||
+		    !subflow_hmac_valid(req, &mp_opt)) {
 			SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_JOINACKMAC);
 			return NULL;
 		}
@@ -494,9 +497,9 @@ create_child:
 			/* with OoO packets we can reach here without ingress
 			 * mpc option
 			 */
-			ctx->remote_key = opt_rx.mptcp.sndr_key;
-			ctx->fully_established = opt_rx.mptcp.mp_capable;
-			ctx->can_ack = opt_rx.mptcp.mp_capable;
+			ctx->remote_key = mp_opt.sndr_key;
+			ctx->fully_established = mp_opt.mp_capable;
+			ctx->can_ack = mp_opt.mp_capable;
 		} else if (ctx->mp_join) {
 			struct mptcp_sock *owner;
 
-- 
cgit 


From 2b195850128f5bafde177b12489d9fa27962cc1e Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Thu, 30 Apr 2020 10:35:41 -0700
Subject: tcp: add tp->dup_ack_counter

In commit 86de5921a3d5 ("tcp: defer SACK compression after DupThresh")
I added a TCP_FASTRETRANS_THRESH bias to tp->compressed_ack in order
to enable sack compression only after 3 dupacks.

Since we plan to relax this rule for flows that involve
stacks not requiring this old rule, this patch adds
a distinct tp->dup_ack_counter.

This means the TCP_FASTRETRANS_THRESH value is now used
in a single location that a future patch can adjust:

	if (tp->dup_ack_counter < TCP_FASTRETRANS_THRESH) {
		tp->dup_ack_counter++;
		goto send_now;
	}

This patch also introduces tcp_sack_compress_send_ack()
helper to ease following patch comprehension.

This patch refines LINUX_MIB_TCPACKCOMPRESSED to not
count the acks that we had to send if the timer expires
or tcp_sack_compress_send_ack() is sending an ack.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/tcp.h   |  1 +
 net/ipv4/tcp_input.c  | 36 +++++++++++++++++++++++++++---------
 net/ipv4/tcp_output.c |  6 +++---
 net/ipv4/tcp_timer.c  |  8 +++++++-
 4 files changed, 38 insertions(+), 13 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 421c99c12291..2c6f87e9f0cf 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -268,6 +268,7 @@ struct tcp_sock {
 	} rack;
 	u16	advmss;		/* Advertised MSS			*/
 	u8	compressed_ack;
+	u8	dup_ack_counter;
 	u32	chrono_start;	/* Start time in jiffies of a TCP chrono */
 	u32	chrono_stat[3];	/* Time in jiffies for chrono_stat stats */
 	u8	chrono_type:2,	/* current chronograph type */
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index bf4ced9273e8..da777df0a0ba 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4327,6 +4327,27 @@ static void tcp_sack_maybe_coalesce(struct tcp_sock *tp)
 	}
 }
 
+static void tcp_sack_compress_send_ack(struct sock *sk)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+
+	if (!tp->compressed_ack)
+		return;
+
+	if (hrtimer_try_to_cancel(&tp->compressed_ack_timer) == 1)
+		__sock_put(sk);
+
+	/* Since we have to send one ack finally,
+	 * substract one from tp->compressed_ack to keep
+	 * LINUX_MIB_TCPACKCOMPRESSED accurate.
+	 */
+	NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED,
+		      tp->compressed_ack - 1);
+
+	tp->compressed_ack = 0;
+	tcp_send_ack(sk);
+}
+
 static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
@@ -4355,8 +4376,7 @@ static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq)
 	 * If the sack array is full, forget about the last one.
 	 */
 	if (this_sack >= TCP_NUM_SACKS) {
-		if (tp->compressed_ack > TCP_FASTRETRANS_THRESH)
-			tcp_send_ack(sk);
+		tcp_sack_compress_send_ack(sk);
 		this_sack--;
 		tp->rx_opt.num_sacks--;
 		sp--;
@@ -5275,15 +5295,13 @@ send_now:
 
 	if (tp->compressed_ack_rcv_nxt != tp->rcv_nxt) {
 		tp->compressed_ack_rcv_nxt = tp->rcv_nxt;
-		if (tp->compressed_ack > TCP_FASTRETRANS_THRESH)
-			NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED,
-				      tp->compressed_ack - TCP_FASTRETRANS_THRESH);
-		tp->compressed_ack = 0;
+		tp->dup_ack_counter = 0;
 	}
-
-	if (++tp->compressed_ack <= TCP_FASTRETRANS_THRESH)
+	if (tp->dup_ack_counter < TCP_FASTRETRANS_THRESH) {
+		tp->dup_ack_counter++;
 		goto send_now;
-
+	}
+	tp->compressed_ack++;
 	if (hrtimer_is_queued(&tp->compressed_ack_timer))
 		return;
 
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index ba4482130f08..c414aeb1efa9 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -184,10 +184,10 @@ static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts,
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	if (unlikely(tp->compressed_ack > TCP_FASTRETRANS_THRESH)) {
+	if (unlikely(tp->compressed_ack)) {
 		NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED,
-			      tp->compressed_ack - TCP_FASTRETRANS_THRESH);
-		tp->compressed_ack = TCP_FASTRETRANS_THRESH;
+			      tp->compressed_ack);
+		tp->compressed_ack = 0;
 		if (hrtimer_try_to_cancel(&tp->compressed_ack_timer) == 1)
 			__sock_put(sk);
 	}
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index c3f26dcd6704..ada046f425d2 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -753,8 +753,14 @@ static enum hrtimer_restart tcp_compressed_ack_kick(struct hrtimer *timer)
 
 	bh_lock_sock(sk);
 	if (!sock_owned_by_user(sk)) {
-		if (tp->compressed_ack > TCP_FASTRETRANS_THRESH)
+		if (tp->compressed_ack) {
+			/* Since we have to send one ack finally,
+			 * substract one from tp->compressed_ack to keep
+			 * LINUX_MIB_TCPACKCOMPRESSED accurate.
+			 */
+			tp->compressed_ack--;
 			tcp_send_ack(sk);
+		}
 	} else {
 		if (!test_and_set_bit(TCP_DELACK_TIMER_DEFERRED,
 				      &sk->sk_tsq_flags))
-- 
cgit 


From ccd0628fca440268711560a1dbacc727b4f9e214 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Thu, 30 Apr 2020 10:35:42 -0700
Subject: tcp: tcp_sack_new_ofo_skb() should be more conservative

Currently, tcp_sack_new_ofo_skb() sends an ack if prior
acks were 'compressed', if room has to be made in tp->selective_acks[]

But there is no guarantee all four sack ranges can be included
in SACK option. As a matter of fact, when TCP timestamps option
is used, only three SACK ranges can be included.

Lets assume only two ranges can be included, and force the ack:

- When we touch more than 2 ranges in the reordering
  done if tcp_sack_extend() could be done.

- If we have at least 2 ranges when adding a new one.

This enforces that before a range is in third or fourth
position, at least one ACK packet included it in first/second
position.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp_input.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index da777df0a0ba..ef921ecba415 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4348,6 +4348,12 @@ static void tcp_sack_compress_send_ack(struct sock *sk)
 	tcp_send_ack(sk);
 }
 
+/* Reasonable amount of sack blocks included in TCP SACK option
+ * The max is 4, but this becomes 3 if TCP timestamps are there.
+ * Given that SACK packets might be lost, be conservative and use 2.
+ */
+#define TCP_SACK_BLOCKS_EXPECTED 2
+
 static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
@@ -4360,6 +4366,8 @@ static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq)
 
 	for (this_sack = 0; this_sack < cur_sacks; this_sack++, sp++) {
 		if (tcp_sack_extend(sp, seq, end_seq)) {
+			if (this_sack >= TCP_SACK_BLOCKS_EXPECTED)
+				tcp_sack_compress_send_ack(sk);
 			/* Rotate this_sack to the first one. */
 			for (; this_sack > 0; this_sack--, sp--)
 				swap(*sp, *(sp - 1));
@@ -4369,6 +4377,9 @@ static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq)
 		}
 	}
 
+	if (this_sack >= TCP_SACK_BLOCKS_EXPECTED)
+		tcp_sack_compress_send_ack(sk);
+
 	/* Could not find an adjacent existing SACK, build a new one,
 	 * put it at the front, and shift everyone else down.  We
 	 * always know there is at least one SACK present already here.
@@ -4376,7 +4387,6 @@ static void tcp_sack_new_ofo_skb(struct sock *sk, u32 seq, u32 end_seq)
 	 * If the sack array is full, forget about the last one.
 	 */
 	if (this_sack >= TCP_NUM_SACKS) {
-		tcp_sack_compress_send_ack(sk);
 		this_sack--;
 		tp->rx_opt.num_sacks--;
 		sp--;
-- 
cgit 


From a70437cc09a11771870e9f6bfc0ba1237161daa8 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Thu, 30 Apr 2020 10:35:43 -0700
Subject: tcp: add hrtimer slack to sack compression

Add a sysctl to control hrtimer slack, default of 100 usec.

This gives the opportunity to reduce system overhead,
and help very short RTT flows.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 Documentation/networking/ip-sysctl.rst | 8 ++++++++
 include/net/netns/ipv4.h               | 1 +
 net/ipv4/sysctl_net_ipv4.c             | 7 +++++++
 net/ipv4/tcp_input.c                   | 5 +++--
 net/ipv4/tcp_ipv4.c                    | 1 +
 5 files changed, 20 insertions(+), 2 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst
index 3266aee9e052..50b440d29a13 100644
--- a/Documentation/networking/ip-sysctl.rst
+++ b/Documentation/networking/ip-sysctl.rst
@@ -651,6 +651,14 @@ tcp_comp_sack_delay_ns - LONG INTEGER
 
 	Default : 1,000,000 ns (1 ms)
 
+tcp_comp_sack_slack_ns - LONG INTEGER
+	This sysctl control the slack used when arming the
+	timer used by SACK compression. This gives extra time
+	for small RTT flows, and reduces system overhead by allowing
+	opportunistic reduction of timer interrupts.
+
+	Default : 100,000 ns (100 us)
+
 tcp_comp_sack_nr - INTEGER
 	Max number of SACK that can be compressed.
 	Using 0 disables SACK compression.
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index 5acdb4d414c4..9e36738c1fe1 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -173,6 +173,7 @@ struct netns_ipv4 {
 	int sysctl_tcp_rmem[3];
 	int sysctl_tcp_comp_sack_nr;
 	unsigned long sysctl_tcp_comp_sack_delay_ns;
+	unsigned long sysctl_tcp_comp_sack_slack_ns;
 	struct inet_timewait_death_row tcp_death_row;
 	int sysctl_max_syn_backlog;
 	int sysctl_tcp_fastopen;
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index 95ad71e76cc3..3a628423d27b 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -1329,6 +1329,13 @@ static struct ctl_table ipv4_net_table[] = {
 		.mode		= 0644,
 		.proc_handler	= proc_doulongvec_minmax,
 	},
+	{
+		.procname	= "tcp_comp_sack_slack_ns",
+		.data		= &init_net.ipv4.sysctl_tcp_comp_sack_slack_ns,
+		.maxlen		= sizeof(unsigned long),
+		.mode		= 0644,
+		.proc_handler	= proc_doulongvec_minmax,
+	},
 	{
 		.procname	= "tcp_comp_sack_nr",
 		.data		= &init_net.ipv4.sysctl_tcp_comp_sack_nr,
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index ef921ecba415..d68128a672ab 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -5324,8 +5324,9 @@ send_now:
 	delay = min_t(unsigned long, sock_net(sk)->ipv4.sysctl_tcp_comp_sack_delay_ns,
 		      rtt * (NSEC_PER_USEC >> 3)/20);
 	sock_hold(sk);
-	hrtimer_start(&tp->compressed_ack_timer, ns_to_ktime(delay),
-		      HRTIMER_MODE_REL_PINNED_SOFT);
+	hrtimer_start_range_ns(&tp->compressed_ack_timer, ns_to_ktime(delay),
+			       sock_net(sk)->ipv4.sysctl_tcp_comp_sack_slack_ns,
+			       HRTIMER_MODE_REL_PINNED_SOFT);
 }
 
 static inline void tcp_ack_snd_check(struct sock *sk)
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 83a5d24e13b8..6c05f1ceb538 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -2780,6 +2780,7 @@ static int __net_init tcp_sk_init(struct net *net)
 		       sizeof(init_net.ipv4.sysctl_tcp_wmem));
 	}
 	net->ipv4.sysctl_tcp_comp_sack_delay_ns = NSEC_PER_MSEC;
+	net->ipv4.sysctl_tcp_comp_sack_slack_ns = 100 * NSEC_PER_USEC;
 	net->ipv4.sysctl_tcp_comp_sack_nr = 44;
 	net->ipv4.sysctl_tcp_fastopen = TFO_CLIENT_ENABLE;
 	spin_lock_init(&net->ipv4.tcp_fastopen_ctx_lock);
-- 
cgit 


From 8dc242ad661c2694a582541c2264ffc0e7c4d27d Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Mon, 4 May 2020 11:27:49 -0700
Subject: tcp: refine tcp_pacing_delay() for very low pacing rates

With the addition of horizon feature to sch_fq, we noticed some
suboptimal behavior of extremely low pacing rate TCP flows, especially
when TCP is not aware of a drop happening in lower stacks.

Back in commit 3f80e08f40cd ("tcp: add tcp_reset_xmit_timer() helper"),
tcp_pacing_delay() was added to estimate an extra delay to add to standard
rto timers.

This patch removes the skb argument from this helper and
tcp_reset_xmit_timer() because it makes more sense to simply
consider the time at which next packet is allowed to be sent,
instead of the time of whatever packet has been sent.

This avoids arming RTO timer too soon and removes
spurious horizon drops.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/tcp.h     | 21 ++++++++-------------
 net/ipv4/tcp_input.c  |  4 ++--
 net/ipv4/tcp_output.c |  8 +++-----
 3 files changed, 13 insertions(+), 20 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 1beed50522b1..43b87a8d4790 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1289,26 +1289,22 @@ static inline bool tcp_needs_internal_pacing(const struct sock *sk)
 	return smp_load_acquire(&sk->sk_pacing_status) == SK_PACING_NEEDED;
 }
 
-/* Return in jiffies the delay before one skb is sent.
- * If @skb is NULL, we look at EDT for next packet being sent on the socket.
+/* Estimates in how many jiffies next packet for this flow can be sent.
+ * Scheduling a retransmit timer too early would be silly.
  */
-static inline unsigned long tcp_pacing_delay(const struct sock *sk,
-					     const struct sk_buff *skb)
+static inline unsigned long tcp_pacing_delay(const struct sock *sk)
 {
-	s64 pacing_delay = skb ? skb->tstamp : tcp_sk(sk)->tcp_wstamp_ns;
+	s64 delay = tcp_sk(sk)->tcp_wstamp_ns - tcp_sk(sk)->tcp_clock_cache;
 
-	pacing_delay -= tcp_sk(sk)->tcp_clock_cache;
-
-	return pacing_delay > 0 ? nsecs_to_jiffies(pacing_delay) : 0;
+	return delay > 0 ? nsecs_to_jiffies(delay) : 0;
 }
 
 static inline void tcp_reset_xmit_timer(struct sock *sk,
 					const int what,
 					unsigned long when,
-					const unsigned long max_when,
-					const struct sk_buff *skb)
+					const unsigned long max_when)
 {
-	inet_csk_reset_xmit_timer(sk, what, when + tcp_pacing_delay(sk, skb),
+	inet_csk_reset_xmit_timer(sk, what, when + tcp_pacing_delay(sk),
 				  max_when);
 }
 
@@ -1336,8 +1332,7 @@ static inline void tcp_check_probe_timer(struct sock *sk)
 {
 	if (!tcp_sk(sk)->packets_out && !inet_csk(sk)->icsk_pending)
 		tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
-				     tcp_probe0_base(sk), TCP_RTO_MAX,
-				     NULL);
+				     tcp_probe0_base(sk), TCP_RTO_MAX);
 }
 
 static inline void tcp_init_wl(struct tcp_sock *tp, u32 seq)
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index d68128a672ab..7d205b2a733c 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3014,7 +3014,7 @@ void tcp_rearm_rto(struct sock *sk)
 			rto = usecs_to_jiffies(max_t(int, delta_us, 1));
 		}
 		tcp_reset_xmit_timer(sk, ICSK_TIME_RETRANS, rto,
-				     TCP_RTO_MAX, tcp_rtx_queue_head(sk));
+				     TCP_RTO_MAX);
 	}
 }
 
@@ -3291,7 +3291,7 @@ static void tcp_ack_probe(struct sock *sk)
 		unsigned long when = tcp_probe0_when(sk, TCP_RTO_MAX);
 
 		tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
-				     when, TCP_RTO_MAX, NULL);
+				     when, TCP_RTO_MAX);
 	}
 }
 
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index c414aeb1efa9..32c9db902f18 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2593,8 +2593,7 @@ bool tcp_schedule_loss_probe(struct sock *sk, bool advancing_rto)
 	if (rto_delta_us > 0)
 		timeout = min_t(u32, timeout, usecs_to_jiffies(rto_delta_us));
 
-	tcp_reset_xmit_timer(sk, ICSK_TIME_LOSS_PROBE, timeout,
-			     TCP_RTO_MAX, NULL);
+	tcp_reset_xmit_timer(sk, ICSK_TIME_LOSS_PROBE, timeout, TCP_RTO_MAX);
 	return true;
 }
 
@@ -3174,8 +3173,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk)
 		    icsk->icsk_pending != ICSK_TIME_REO_TIMEOUT)
 			tcp_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
 					     inet_csk(sk)->icsk_rto,
-					     TCP_RTO_MAX,
-					     skb);
+					     TCP_RTO_MAX);
 	}
 }
 
@@ -3907,7 +3905,7 @@ void tcp_send_probe0(struct sock *sk)
 		 */
 		timeout = TCP_RESOURCE_PROBE_INTERVAL;
 	}
-	tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, timeout, TCP_RTO_MAX, NULL);
+	tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, timeout, TCP_RTO_MAX);
 }
 
 int tcp_rtx_synack(const struct sock *sk, struct request_sock *req)
-- 
cgit 


From 636ef28d6e4d174e424102466caf572b0406fb0e Mon Sep 17 00:00:00 2001
From: zhang kai <zhangkaiheb@126.com>
Date: Thu, 7 May 2020 11:08:30 +0800
Subject: tcp: tcp_mark_head_lost is only valid for sack-tcp

so tcp_is_sack/reno checks are removed from tcp_mark_head_lost.

Signed-off-by: zhang kai <zhangkaiheb@126.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp_input.c | 32 +++++++-------------------------
 1 file changed, 7 insertions(+), 25 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 66e55e51c550..7529c2816f2f 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -2183,8 +2183,7 @@ static bool tcp_time_to_recover(struct sock *sk, int flag)
 }
 
 /* Detect loss in event "A" above by marking head of queue up as lost.
- * For non-SACK(Reno) senders, the first "packets" number of segments
- * are considered lost. For RFC3517 SACK, a segment is considered lost if it
+ * For RFC3517 SACK, a segment is considered lost if it
  * has at least tp->reordering SACKed seqments above it; "packets" refers to
  * the maximum SACKed segments to pass before reaching this limit.
  */
@@ -2192,10 +2191,9 @@ static void tcp_mark_head_lost(struct sock *sk, int packets, int mark_head)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
-	int cnt, oldcnt, lost;
-	unsigned int mss;
+	int cnt;
 	/* Use SACK to deduce losses of new sequences sent during recovery */
-	const u32 loss_high = tcp_is_sack(tp) ?  tp->snd_nxt : tp->high_seq;
+	const u32 loss_high = tp->snd_nxt;
 
 	WARN_ON(packets > tp->packets_out);
 	skb = tp->lost_skb_hint;
@@ -2218,26 +2216,11 @@ static void tcp_mark_head_lost(struct sock *sk, int packets, int mark_head)
 		if (after(TCP_SKB_CB(skb)->end_seq, loss_high))
 			break;
 
-		oldcnt = cnt;
-		if (tcp_is_reno(tp) ||
-		    (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED))
+		if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)
 			cnt += tcp_skb_pcount(skb);
 
-		if (cnt > packets) {
-			if (tcp_is_sack(tp) ||
-			    (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED) ||
-			    (oldcnt >= packets))
-				break;
-
-			mss = tcp_skb_mss(skb);
-			/* If needed, chop off the prefix to mark as lost. */
-			lost = (packets - oldcnt) * mss;
-			if (lost < skb->len &&
-			    tcp_fragment(sk, TCP_FRAG_IN_RTX_QUEUE, skb,
-					 lost, mss, GFP_ATOMIC) < 0)
-				break;
-			cnt = packets;
-		}
+		if (cnt > packets)
+			break;
 
 		tcp_skb_mark_lost(tp, skb);
 
@@ -2849,8 +2832,7 @@ static void tcp_fastretrans_alert(struct sock *sk, const u32 prior_snd_una,
 			if (tcp_try_undo_partial(sk, prior_snd_una))
 				return;
 			/* Partial ACK arrived. Force fast retransmit. */
-			do_lost = tcp_is_reno(tp) ||
-				  tcp_force_fast_retransmit(sk);
+			do_lost = tcp_force_fast_retransmit(sk);
 		}
 		if (tcp_try_undo_dsack(sk)) {
 			tcp_try_keep_open(sk);
-- 
cgit 


From 24adbc1676af4e134e709ddc7f34cf2adc2131e4 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Tue, 12 May 2020 06:54:30 -0700
Subject: tcp: fix SO_RCVLOWAT hangs with fat skbs

We autotune rcvbuf whenever SO_RCVLOWAT is set to account for 100%
overhead in tcp_set_rcvlowat()

This works well when skb->len/skb->truesize ratio is bigger than 0.5

But if we receive packets with small MSS, we can end up in a situation
where not enough bytes are available in the receive queue to satisfy
RCVLOWAT setting.
As our sk_rcvbuf limit is hit, we send zero windows in ACK packets,
preventing remote peer from sending more data.

Even autotuning does not help, because it only triggers at the time
user process drains the queue. If no EPOLLIN is generated, this
can not happen.

Note poll() has a similar issue, after commit
c7004482e8dc ("tcp: Respect SO_RCVLOWAT in tcp_poll().")

Fixes: 03f45c883c6f ("tcp: avoid extra wakeups for SO_RCVLOWAT users")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/tcp.h    | 13 +++++++++++++
 net/ipv4/tcp.c       | 14 +++++++++++---
 net/ipv4/tcp_input.c |  3 ++-
 3 files changed, 26 insertions(+), 4 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 64f84683feae..6f8e60c6fbc7 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1420,6 +1420,19 @@ static inline int tcp_full_space(const struct sock *sk)
 	return tcp_win_from_space(sk, READ_ONCE(sk->sk_rcvbuf));
 }
 
+/* We provision sk_rcvbuf around 200% of sk_rcvlowat.
+ * If 87.5 % (7/8) of the space has been consumed, we want to override
+ * SO_RCVLOWAT constraint, since we are receiving skbs with too small
+ * len/truesize ratio.
+ */
+static inline bool tcp_rmem_pressure(const struct sock *sk)
+{
+	int rcvbuf = READ_ONCE(sk->sk_rcvbuf);
+	int threshold = rcvbuf - (rcvbuf >> 3);
+
+	return atomic_read(&sk->sk_rmem_alloc) > threshold;
+}
+
 extern void tcp_openreq_init_rwin(struct request_sock *req,
 				  const struct sock *sk_listener,
 				  const struct dst_entry *dst);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index e72bd651d21a..a385fcaaa03b 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -476,9 +476,17 @@ static void tcp_tx_timestamp(struct sock *sk, u16 tsflags)
 static inline bool tcp_stream_is_readable(const struct tcp_sock *tp,
 					  int target, struct sock *sk)
 {
-	return (READ_ONCE(tp->rcv_nxt) - READ_ONCE(tp->copied_seq) >= target) ||
-		(sk->sk_prot->stream_memory_read ?
-		sk->sk_prot->stream_memory_read(sk) : false);
+	int avail = READ_ONCE(tp->rcv_nxt) - READ_ONCE(tp->copied_seq);
+
+	if (avail > 0) {
+		if (avail >= target)
+			return true;
+		if (tcp_rmem_pressure(sk))
+			return true;
+	}
+	if (sk->sk_prot->stream_memory_read)
+		return sk->sk_prot->stream_memory_read(sk);
+	return false;
 }
 
 /*
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index b996dc1069c5..29c6fc8c7716 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4757,7 +4757,8 @@ void tcp_data_ready(struct sock *sk)
 	const struct tcp_sock *tp = tcp_sk(sk);
 	int avail = tp->rcv_nxt - tp->copied_seq;
 
-	if (avail < sk->sk_rcvlowat && !sock_flag(sk, SOCK_DONE))
+	if (avail < sk->sk_rcvlowat && !tcp_rmem_pressure(sk) &&
+	    !sock_flag(sk, SOCK_DONE))
 		return;
 
 	sk->sk_data_ready(sk);
-- 
cgit 


From bc183dec08f9cb177cf5206a010b7a9e7b22e567 Mon Sep 17 00:00:00 2001
From: Florian Westphal <fw@strlen.de>
Date: Fri, 29 May 2020 00:01:52 +0200
Subject: tcp: tcp_init_buffer_space can be static

As of commit 98fa6271cfcb
("tcp: refactor setting the initial congestion window") this is called
only from tcp_input.c, so it can be static.

Signed-off-by: Florian Westphal <fw@strlen.de>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/tcp.h    | 1 -
 net/ipv4/tcp_input.c | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

(limited to 'net/ipv4/tcp_input.c')

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 66e4b8331850..bca761ffa25f 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -662,7 +662,6 @@ void tcp_initialize_rcv_mss(struct sock *sk);
 int tcp_mtu_to_mss(struct sock *sk, int pmtu);
 int tcp_mss_to_mtu(struct sock *sk, int mss);
 void tcp_mtup_init(struct sock *sk);
-void tcp_init_buffer_space(struct sock *sk);
 
 static inline void tcp_bound_rto(const struct sock *sk)
 {
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index ad90102f5dfb..83330a6cb242 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -437,7 +437,7 @@ static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb)
 /* 3. Try to fixup all. It is made immediately after connection enters
  *    established state.
  */
-void tcp_init_buffer_space(struct sock *sk)
+static void tcp_init_buffer_space(struct sock *sk)
 {
 	int tcp_app_win = sock_net(sk)->ipv4.sysctl_tcp_app_win;
 	struct tcp_sock *tp = tcp_sk(sk);
-- 
cgit