diff mbox series

[net] tcp: Retain ECT bits for tos reflection

Message ID 20201208175508.1793520-1-weiwan@google.com (mailing list archive)
State Accepted
Delegated to: Netdev Maintainers
Headers show
Series [net] tcp: Retain ECT bits for tos reflection | expand

Checks

Context Check Description
netdev/cover_letter success Link
netdev/fixes_present success Link
netdev/patch_count success Link
netdev/tree_selection success Clearly marked for net
netdev/subject_prefix success Link
netdev/source_inline success Was 0 now: 0
netdev/verify_signedoff success Link
netdev/module_param success Was 0 now: 0
netdev/build_32bit success Errors and warnings before: 17 this patch: 17
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/verify_fixes success Link
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 38 lines checked
netdev/build_allmodconfig_warn success Errors and warnings before: 17 this patch: 17
netdev/header_inline success Link
netdev/stable success Stable not CCed

Commit Message

Wei Wang Dec. 8, 2020, 5:55 p.m. UTC
For DCTCP, we have to retain the ECT bits set by the congestion control
algorithm on the socket when reflecting syn TOS in syn-ack, in order to
make ECN work properly.

Fixes: ac8f1710c12b ("tcp: reflect tos value received in SYN to the socket")
Reported-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Wei Wang <weiwan@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
---
 net/ipv4/tcp_ipv4.c | 7 +++++--
 net/ipv6/tcp_ipv6.c | 7 +++++--
 2 files changed, 10 insertions(+), 4 deletions(-)

Comments

David Miller Dec. 9, 2020, 12:30 a.m. UTC | #1
From: Wei Wang <weiwan@google.com>
Date: Tue,  8 Dec 2020 09:55:08 -0800

> For DCTCP, we have to retain the ECT bits set by the congestion control
> algorithm on the socket when reflecting syn TOS in syn-ack, in order to
> make ECN work properly.
> 
> Fixes: ac8f1710c12b ("tcp: reflect tos value received in SYN to the socket")
> Reported-by: Alexander Duyck <alexanderduyck@fb.com>
> Signed-off-by: Wei Wang <weiwan@google.com>
> Reviewed-by: Eric Dumazet <edumazet@google.com>

Applied, thanks.
diff mbox series

Patch

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 8391aa29e7a4..595dcc3afac5 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -984,7 +984,8 @@  static int tcp_v4_send_synack(const struct sock *sk, struct dst_entry *dst,
 		__tcp_v4_send_check(skb, ireq->ir_loc_addr, ireq->ir_rmt_addr);
 
 		tos = sock_net(sk)->ipv4.sysctl_tcp_reflect_tos ?
-				tcp_rsk(req)->syn_tos & ~INET_ECN_MASK :
+				(tcp_rsk(req)->syn_tos & ~INET_ECN_MASK) |
+				(inet_sk(sk)->tos & INET_ECN_MASK) :
 				inet_sk(sk)->tos;
 
 		if (!INET_ECN_is_capable(tos) &&
@@ -1541,7 +1542,9 @@  struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb,
 		inet_csk(newsk)->icsk_ext_hdr_len = inet_opt->opt.optlen;
 	newinet->inet_id = prandom_u32();
 
-	/* Set ToS of the new socket based upon the value of incoming SYN. */
+	/* Set ToS of the new socket based upon the value of incoming SYN.
+	 * ECT bits are set later in tcp_init_transfer().
+	 */
 	if (sock_net(sk)->ipv4.sysctl_tcp_reflect_tos)
 		newinet->tos = tcp_rsk(req)->syn_tos & ~INET_ECN_MASK;
 
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 992cbf3eb9e3..f676f176eb6b 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -528,7 +528,8 @@  static int tcp_v6_send_synack(const struct sock *sk, struct dst_entry *dst,
 			fl6->flowlabel = ip6_flowlabel(ipv6_hdr(ireq->pktopts));
 
 		tclass = sock_net(sk)->ipv4.sysctl_tcp_reflect_tos ?
-				tcp_rsk(req)->syn_tos & ~INET_ECN_MASK :
+				(tcp_rsk(req)->syn_tos & ~INET_ECN_MASK) |
+				(np->tclass & INET_ECN_MASK) :
 				np->tclass;
 
 		if (!INET_ECN_is_capable(tclass) &&
@@ -1320,7 +1321,9 @@  static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff *
 	if (np->repflow)
 		newnp->flow_label = ip6_flowlabel(ipv6_hdr(skb));
 
-	/* Set ToS of the new socket based upon the value of incoming SYN. */
+	/* Set ToS of the new socket based upon the value of incoming SYN.
+	 * ECT bits are set later in tcp_init_transfer().
+	 */
 	if (sock_net(sk)->ipv4.sysctl_tcp_reflect_tos)
 		newnp->tclass = tcp_rsk(req)->syn_tos & ~INET_ECN_MASK;