tcp: use dctcp if enabled on the route to the initiator

Currently, the following case doesn't use DCTCP, even if it should:
A responder has f.e. Cubic as system wide default, but for a specific
route to the initiating host, DCTCP is being set in RTAX_CC_ALGO. The
initiating host then uses DCTCP as congestion control, but since the
initiator sets ECT(0), tcp_ecn_create_request() doesn't set ecn_ok,
and we have to fall back to Reno after 3WHS completes.

We were thinking on how to solve this in a minimal, non-intrusive
way without bloating tcp_ecn_create_request() needlessly: lets cache
the CA ecn option flag in RTAX_FEATURES. In other words, when ECT(0)
is set on the SYN packet, set ecn_ok=1 iff route RTAX_FEATURES
contains the unexposed (internal-only) DST_FEATURE_ECN_CA. This allows
to only do a single metric feature lookup inside tcp_ecn_create_request().

Joint work with Florian Westphal.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/dst.h

@@ -207,6 +207,12 @@ static inline void dst_metric_set(struct dst_entry *dst, int metric, u32 val)
 		p[metric-1] = val;
 }
 
+/* Kernel-internal feature bits that are unallocated in user space. */
+#define DST_FEATURE_ECN_CA	(1 << 31)
+
+#define DST_FEATURE_MASK	(DST_FEATURE_ECN_CA)
+#define DST_FEATURE_ECN_MASK	(DST_FEATURE_ECN_CA | RTAX_FEATURE_ECN)
+
 static inline u32
 dst_feature(const struct dst_entry *dst, u32 feature)
 {

include/net/tcp.h

@@ -888,7 +888,7 @@ void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked);
 extern struct tcp_congestion_ops tcp_reno;
 
 struct tcp_congestion_ops *tcp_ca_find_key(u32 key);
-u32 tcp_ca_get_key_by_name(const char *name);
+u32 tcp_ca_get_key_by_name(const char *name, bool *ecn_ca);
 #ifdef CONFIG_INET
 char *tcp_ca_get_name_by_key(u32 key, char *buffer);
 #else