bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jakub Sitnicki <jakub@cloudflare.com>
To: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org, kernel-team@cloudflare.com,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	"David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	Marek Majkowski <marek@cloudflare.com>
Subject: [PATCH bpf-next v4 06/16] inet6: Run SK_LOOKUP BPF program on socket lookup
Date: Mon, 13 Jul 2020 19:46:44 +0200	[thread overview]
Message-ID: <20200713174654.642628-7-jakub@cloudflare.com> (raw)
In-Reply-To: <20200713174654.642628-1-jakub@cloudflare.com>

Following ipv4 stack changes, run a BPF program attached to netns before
looking up a listening socket. Program can return a listening socket to use
as result of socket lookup, fail the lookup, or take no action.

Suggested-by: Marek Majkowski <marek@cloudflare.com>
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
---

Notes:
    v4:
    - Adapt to changes in BPF prog return codes.
    - Invert return value from bpf_sk_lookup_run_v6 to true on skip reuseport.
    
    v3:
    - Use a static_key to minimize the hook overhead when not used. (Alexei)
    - Don't copy struct in6_addr when populating BPF prog context. (Martin)
    - Adapt for running an array of attached programs. (Alexei)
    - Adapt for optionally skipping reuseport selection. (Martin)

 include/linux/filter.h      | 44 +++++++++++++++++++++++++++++++++++++
 net/ipv6/inet6_hashtables.c | 35 +++++++++++++++++++++++++++++
 2 files changed, 79 insertions(+)

diff --git a/include/linux/filter.h b/include/linux/filter.h
index b9ad0fdabca5..900b71af5580 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -1397,4 +1397,48 @@ static inline bool bpf_sk_lookup_run_v4(struct net *net, int protocol,
 	return no_reuseport;
 }
 
+#if IS_ENABLED(CONFIG_IPV6)
+static inline bool bpf_sk_lookup_run_v6(struct net *net, int protocol,
+					const struct in6_addr *saddr,
+					const __be16 sport,
+					const struct in6_addr *daddr,
+					const u16 dport,
+					struct sock **psk)
+{
+	struct bpf_prog_array *run_array;
+	struct sock *selected_sk = NULL;
+	bool no_reuseport = false;
+
+	rcu_read_lock();
+	run_array = rcu_dereference(net->bpf.run_array[NETNS_BPF_SK_LOOKUP]);
+	if (run_array) {
+		struct bpf_sk_lookup_kern ctx = {
+			.family		= AF_INET6,
+			.protocol	= protocol,
+			.v6.saddr	= saddr,
+			.v6.daddr	= daddr,
+			.sport		= sport,
+			.dport		= dport,
+		};
+		u32 act;
+
+		act = BPF_PROG_SK_LOOKUP_RUN_ARRAY(run_array, ctx, BPF_PROG_RUN);
+		if (act == SK_PASS) {
+			selected_sk = ctx.selected_sk;
+			no_reuseport = ctx.no_reuseport;
+			goto unlock;
+		}
+
+		selected_sk = ERR_PTR(-ECONNREFUSED);
+		WARN_ONCE(act != SK_DROP,
+			  "Illegal BPF SK_LOOKUP return value %u, expect packet loss!\n",
+			  act);
+	}
+unlock:
+	rcu_read_unlock();
+	*psk = selected_sk;
+	return no_reuseport;
+}
+#endif /* IS_ENABLED(CONFIG_IPV6) */
+
 #endif /* __LINUX_FILTER_H__ */
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index 03942eef8ab6..2d3add9e6116 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -21,6 +21,8 @@
 #include <net/ip.h>
 #include <net/sock_reuseport.h>
 
+extern struct inet_hashinfo tcp_hashinfo;
+
 u32 inet6_ehashfn(const struct net *net,
 		  const struct in6_addr *laddr, const u16 lport,
 		  const struct in6_addr *faddr, const __be16 fport)
@@ -159,6 +161,31 @@ static struct sock *inet6_lhash2_lookup(struct net *net,
 	return result;
 }
 
+static inline struct sock *inet6_lookup_run_bpf(struct net *net,
+						struct inet_hashinfo *hashinfo,
+						struct sk_buff *skb, int doff,
+						const struct in6_addr *saddr,
+						const __be16 sport,
+						const struct in6_addr *daddr,
+						const u16 hnum)
+{
+	struct sock *sk, *reuse_sk;
+	bool no_reuseport;
+
+	if (hashinfo != &tcp_hashinfo)
+		return NULL; /* only TCP is supported */
+
+	no_reuseport = bpf_sk_lookup_run_v6(net, IPPROTO_TCP,
+					    saddr, sport, daddr, hnum, &sk);
+	if (no_reuseport || IS_ERR_OR_NULL(sk))
+		return sk;
+
+	reuse_sk = lookup_reuseport(net, sk, skb, doff, saddr, sport, daddr, hnum);
+	if (reuse_sk)
+		sk = reuse_sk;
+	return sk;
+}
+
 struct sock *inet6_lookup_listener(struct net *net,
 		struct inet_hashinfo *hashinfo,
 		struct sk_buff *skb, int doff,
@@ -170,6 +197,14 @@ struct sock *inet6_lookup_listener(struct net *net,
 	struct sock *result = NULL;
 	unsigned int hash2;
 
+	/* Lookup redirect from BPF */
+	if (static_branch_unlikely(&bpf_sk_lookup_enabled)) {
+		result = inet6_lookup_run_bpf(net, hashinfo, skb, doff,
+					      saddr, sport, daddr, hnum);
+		if (result)
+			goto done;
+	}
+
 	hash2 = ipv6_portaddr_hash(net, daddr, hnum);
 	ilb2 = inet_lhash2_bucket(hashinfo, hash2);
 
-- 
2.25.4


  parent reply	other threads:[~2020-07-13 17:47 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-13 17:46 [PATCH bpf-next v4 00/16] Run a " Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 01/16] bpf, netns: Handle multiple link attachments Jakub Sitnicki
2020-07-15 21:30   ` Andrii Nakryiko
2020-07-13 17:46 ` [PATCH bpf-next v4 02/16] bpf: Introduce SK_LOOKUP program type with a dedicated attach point Jakub Sitnicki
2020-07-16  1:41   ` Andrii Nakryiko
2020-07-16 12:17     ` Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 03/16] inet: Extract helper for selecting socket from reuseport group Jakub Sitnicki
2020-07-16  1:44   ` Andrii Nakryiko
2020-07-13 17:46 ` [PATCH bpf-next v4 04/16] inet: Run SK_LOOKUP BPF program on socket lookup Jakub Sitnicki
2020-07-16  2:23   ` Andrii Nakryiko
2020-07-16 12:32     ` Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 05/16] inet6: Extract helper for selecting socket from reuseport group Jakub Sitnicki
2020-07-13 17:46 ` Jakub Sitnicki [this message]
2020-07-13 17:46 ` [PATCH bpf-next v4 07/16] udp: " Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 08/16] udp: Run SK_LOOKUP BPF program on socket lookup Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 09/16] udp6: Extract helper for selecting socket from reuseport group Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 10/16] udp6: Run SK_LOOKUP BPF program on socket lookup Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 11/16] bpf: Sync linux/bpf.h to tools/ Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 12/16] libbpf: Add support for SK_LOOKUP program type Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 13/16] tools/bpftool: Add name mappings for SK_LOOKUP prog and attach type Jakub Sitnicki
2020-07-16  2:10   ` Andrii Nakryiko
2020-07-13 17:46 ` [PATCH bpf-next v4 14/16] selftests/bpf: Add verifier tests for bpf_sk_lookup context access Jakub Sitnicki
2020-07-16  2:13   ` Andrii Nakryiko
2020-07-13 17:46 ` [PATCH bpf-next v4 15/16] selftests/bpf: Rename test_sk_lookup_kern.c to test_ref_track_kern.c Jakub Sitnicki
2020-07-13 17:46 ` [PATCH bpf-next v4 16/16] selftests/bpf: Tests for BPF_SK_LOOKUP attach point Jakub Sitnicki
2020-07-16  2:19   ` Andrii Nakryiko
2020-07-16  2:25 ` [PATCH bpf-next v4 00/16] Run a BPF program on socket lookup Andrii Nakryiko

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200713174654.642628-7-jakub@cloudflare.com \
    --to=jakub@cloudflare.com \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=kernel-team@cloudflare.com \
    --cc=kuba@kernel.org \
    --cc=marek@cloudflare.com \
    --cc=netdev@vger.kernel.org \
    --subject='Re: [PATCH bpf-next v4 06/16] inet6: Run SK_LOOKUP BPF program on socket lookup' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).