Merge branch 'add-bpf_xdp_get_xfrm_state-kfunc'

Alexei Starovoitov · Alexei Starovoitov · commit 403f3e8fda60 · 2023-12-14T17:12:56.000-08:00
Daniel Xu says: ==================== Add bpf_xdp_get_xfrm_state() kfunc This patchset adds two kfunc helpers, bpf_xdp_get_xfrm_state() and bpf_xdp_xfrm_state_release() that wrap xfrm_state_lookup() and xfrm_state_put(). The intent is to support software RSS (via XDP) for the ongoing/upcoming ipsec pcpu work [0]. Recent experiments performed on (hopefully) reproducible AWS testbeds indicate that single tunnel pcpu ipsec can reach line rate on 100G ENA nics. Note this patchset only tests/shows generic xfrm_state access. The "secret sauce" (if you can really even call it that) involves accessing a soon-to-be-upstreamed pcpu_num field in xfrm_state. Early example is available here [1]. [0]: https://datatracker.ietf.org/doc/draft-ietf-ipsecme-multi-sa-performance/03/ [1]: https://github.com/danobi/xdp-tools/blob/e89a1c617aba3b50d990f779357d6ce2863ecb27/xdp-bench/xdp_redirect_cpumap.bpf.c#L385-L406 Changes from v5: * Improve kfunc doc comments * Remove extraneous replay-window setting on selftest reverse path * Squash two kfunc commits into one * Rebase to bpf-next to pick up bitfield write patches * Remove testing of opts.error in selftest prog Changes from v4: * Fixup commit message for selftest * Set opts->error -ENOENT for !x * Revert single file xfrm + bpf Changes from v3: * Place all xfrm bpf integrations in xfrm_bpf.c * Avoid using nval as a temporary * Rebase to bpf-next * Remove extraneous __failure_unpriv annotation for verifier tests Changes from v2: * Fix/simplify BPF_CORE_WRITE_BITFIELD() algorithm * Added verifier tests for bitfield writes * Fix state leakage across test_tunnel subtests Changes from v1: * Move xfrm tunnel tests to test_progs * Fix writing to opts->error when opts is invalid * Use __bpf_kfunc_start_defs() * Remove unused vxlanhdr definition * Add and use BPF_CORE_WRITE_BITFIELD() macro * Make series bisect clean Changes from RFCv2: * Rebased to ipsec-next * Fix netns leak Changes from RFCv1: * Add Antony's commit tags * Add KF_ACQUIRE and KF_RELEASE semantics ==================== Reviewed-by: Eyal Birger <eyal.birger@gmail.com> Link: https://lore.kernel.org/r/cover.1702593901.git.dxu@dxuuu.xyz Signed-off-by: Alexei Starovoitov <ast@kernel.org>
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
@@ -2190,4 +2190,13 @@ static inline int register_xfrm_interface_bpf(void)
 
 #endif
 
+#if IS_ENABLED(CONFIG_DEBUG_INFO_BTF)
+int register_xfrm_state_bpf(void);
+#else
+static inline int register_xfrm_state_bpf(void)
+{
+	return 0;
+}
+#endif
+
 #endif	/* _NET_XFRM_H */
diff --git a/net/xfrm/Makefile b/net/xfrm/Makefile
@@ -21,3 +21,4 @@ obj-$(CONFIG_XFRM_USER_COMPAT) += xfrm_compat.o
 obj-$(CONFIG_XFRM_IPCOMP) += xfrm_ipcomp.o
 obj-$(CONFIG_XFRM_INTERFACE) += xfrm_interface.o
 obj-$(CONFIG_XFRM_ESPINTCP) += espintcp.o
+obj-$(CONFIG_DEBUG_INFO_BTF) += xfrm_state_bpf.o
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
@@ -4218,6 +4218,8 @@ void __init xfrm_init(void)
 #ifdef CONFIG_XFRM_ESPINTCP
 	espintcp_init();
 #endif
+
+	register_xfrm_state_bpf();
 }
 
 #ifdef CONFIG_AUDITSYSCALL
diff --git a/net/xfrm/xfrm_state_bpf.c b/net/xfrm/xfrm_state_bpf.c
@@ -0,0 +1,134 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Unstable XFRM state BPF helpers.
+ *
+ * Note that it is allowed to break compatibility for these functions since the
+ * interface they are exposed through to BPF programs is explicitly unstable.
+ */
+
+#include <linux/bpf.h>
+#include <linux/btf.h>
+#include <linux/btf_ids.h>
+#include <net/xdp.h>
+#include <net/xfrm.h>
+
+/* bpf_xfrm_state_opts - Options for XFRM state lookup helpers
+ *
+ * Members:
+ * @error      - Out parameter, set for any errors encountered
+ *		 Values:
+ *		   -EINVAL - netns_id is less than -1
+ *		   -EINVAL - opts__sz isn't BPF_XFRM_STATE_OPTS_SZ
+ *		   -ENONET - No network namespace found for netns_id
+ *		   -ENOENT - No xfrm_state found
+ * @netns_id	- Specify the network namespace for lookup
+ *		 Values:
+ *		   BPF_F_CURRENT_NETNS (-1)
+ *		     Use namespace associated with ctx
+ *		   [0, S32_MAX]
+ *		     Network Namespace ID
+ * @mark	- XFRM mark to match on
+ * @daddr	- Destination address to match on
+ * @spi		- Security parameter index to match on
+ * @proto	- IP protocol to match on (eg. IPPROTO_ESP)
+ * @family	- Protocol family to match on (AF_INET/AF_INET6)
+ */
+struct bpf_xfrm_state_opts {
+	s32 error;
+	s32 netns_id;
+	u32 mark;
+	xfrm_address_t daddr;
+	__be32 spi;
+	u8 proto;
+	u16 family;
+};
+
+enum {
+	BPF_XFRM_STATE_OPTS_SZ = sizeof(struct bpf_xfrm_state_opts),
+};
+
+__bpf_kfunc_start_defs();
+
+/* bpf_xdp_get_xfrm_state - Get XFRM state
+ *
+ * A `struct xfrm_state *`, if found, must be released with a corresponding
+ * bpf_xdp_xfrm_state_release.
+ *
+ * Parameters:
+ * @ctx	- Pointer to ctx (xdp_md) in XDP program
+ *		    Cannot be NULL
+ * @opts	- Options for lookup (documented above)
+ *		    Cannot be NULL
+ * @opts__sz	- Length of the bpf_xfrm_state_opts structure
+ *		    Must be BPF_XFRM_STATE_OPTS_SZ
+ */
+__bpf_kfunc struct xfrm_state *
+bpf_xdp_get_xfrm_state(struct xdp_md *ctx, struct bpf_xfrm_state_opts *opts, u32 opts__sz)
+{
+	struct xdp_buff *xdp = (struct xdp_buff *)ctx;
+	struct net *net = dev_net(xdp->rxq->dev);
+	struct xfrm_state *x;
+
+	if (!opts || opts__sz < sizeof(opts->error))
+		return NULL;
+
+	if (opts__sz != BPF_XFRM_STATE_OPTS_SZ) {
+		opts->error = -EINVAL;
+		return NULL;
+	}
+
+	if (unlikely(opts->netns_id < BPF_F_CURRENT_NETNS)) {
+		opts->error = -EINVAL;
+		return NULL;
+	}
+
+	if (opts->netns_id >= 0) {
+		net = get_net_ns_by_id(net, opts->netns_id);
+		if (unlikely(!net)) {
+			opts->error = -ENONET;
+			return NULL;
+		}
+	}
+
+	x = xfrm_state_lookup(net, opts->mark, &opts->daddr, opts->spi,
+			      opts->proto, opts->family);
+
+	if (opts->netns_id >= 0)
+		put_net(net);
+	if (!x)
+		opts->error = -ENOENT;
+
+	return x;
+}
+
+/* bpf_xdp_xfrm_state_release - Release acquired xfrm_state object
+ *
+ * This must be invoked for referenced PTR_TO_BTF_ID, and the verifier rejects
+ * the program if any references remain in the program in all of the explored
+ * states.
+ *
+ * Parameters:
+ * @x		- Pointer to referenced xfrm_state object, obtained using
+ *		  bpf_xdp_get_xfrm_state.
+ */
+__bpf_kfunc void bpf_xdp_xfrm_state_release(struct xfrm_state *x)
+{
+	xfrm_state_put(x);
+}
+
+__bpf_kfunc_end_defs();
+
+BTF_SET8_START(xfrm_state_kfunc_set)
+BTF_ID_FLAGS(func, bpf_xdp_get_xfrm_state, KF_RET_NULL | KF_ACQUIRE)
+BTF_ID_FLAGS(func, bpf_xdp_xfrm_state_release, KF_RELEASE)
+BTF_SET8_END(xfrm_state_kfunc_set)
+
+static const struct btf_kfunc_id_set xfrm_state_xdp_kfunc_set = {
+	.owner = THIS_MODULE,
+	.set   = &xfrm_state_kfunc_set,
+};
+
+int __init register_xfrm_state_bpf(void)
+{
+	return register_btf_kfunc_id_set(BPF_PROG_TYPE_XDP,
+					 &xfrm_state_xdp_kfunc_set);
+}
diff --git a/tools/testing/selftests/bpf/prog_tests/test_tunnel.c b/tools/testing/selftests/bpf/prog_tests/test_tunnel.c
@@ -50,6 +50,7 @@
  */
 
 #include <arpa/inet.h>
+#include <linux/if_link.h>
 #include <linux/if_tun.h>
 #include <linux/limits.h>
 #include <linux/sysctl.h>
@@ -92,6 +93,11 @@
 #define IPIP_TUNL_DEV0 "ipip00"
 #define IPIP_TUNL_DEV1 "ipip11"
 
+#define XFRM_AUTH "0x1111111111111111111111111111111111111111"
+#define XFRM_ENC "0x22222222222222222222222222222222"
+#define XFRM_SPI_IN_TO_OUT 0x1
+#define XFRM_SPI_OUT_TO_IN 0x2
+
 #define PING_ARGS "-i 0.01 -c 3 -w 10 -q"
 
 static int config_device(void)
@@ -264,6 +270,92 @@ static void delete_ipip_tunnel(void)
 	SYS_NOFAIL("ip fou del port 5555 2> /dev/null");
 }
 
+static int add_xfrm_tunnel(void)
+{
+	/* at_ns0 namespace
+	 * at_ns0 -> root
+	 */
+	SYS(fail,
+	    "ip netns exec at_ns0 "
+		"ip xfrm state add src %s dst %s proto esp "
+			"spi %d reqid 1 mode tunnel replay-window 42 "
+			"auth-trunc 'hmac(sha1)' %s 96 enc 'cbc(aes)' %s",
+	    IP4_ADDR_VETH0, IP4_ADDR1_VETH1, XFRM_SPI_IN_TO_OUT, XFRM_AUTH, XFRM_ENC);
+	SYS(fail,
+	    "ip netns exec at_ns0 "
+		"ip xfrm policy add src %s/32 dst %s/32 dir out "
+			"tmpl src %s dst %s proto esp reqid 1 "
+			"mode tunnel",
+	    IP4_ADDR_TUNL_DEV0, IP4_ADDR_TUNL_DEV1, IP4_ADDR_VETH0, IP4_ADDR1_VETH1);
+
+	/* root -> at_ns0 */
+	SYS(fail,
+	    "ip netns exec at_ns0 "
+		"ip xfrm state add src %s dst %s proto esp "
+			"spi %d reqid 2 mode tunnel "
+			"auth-trunc 'hmac(sha1)' %s 96 enc 'cbc(aes)' %s",
+	    IP4_ADDR1_VETH1, IP4_ADDR_VETH0, XFRM_SPI_OUT_TO_IN, XFRM_AUTH, XFRM_ENC);
+	SYS(fail,
+	    "ip netns exec at_ns0 "
+		"ip xfrm policy add src %s/32 dst %s/32 dir in "
+			"tmpl src %s dst %s proto esp reqid 2 "
+			"mode tunnel",
+	    IP4_ADDR_TUNL_DEV1, IP4_ADDR_TUNL_DEV0, IP4_ADDR1_VETH1, IP4_ADDR_VETH0);
+
+	/* address & route */
+	SYS(fail, "ip netns exec at_ns0 ip addr add dev veth0 %s/32",
+	    IP4_ADDR_TUNL_DEV0);
+	SYS(fail, "ip netns exec at_ns0 ip route add %s dev veth0 via %s src %s",
+	    IP4_ADDR_TUNL_DEV1, IP4_ADDR1_VETH1, IP4_ADDR_TUNL_DEV0);
+
+	/* root namespace
+	 * at_ns0 -> root
+	 */
+	SYS(fail,
+	    "ip xfrm state add src %s dst %s proto esp "
+		    "spi %d reqid 1 mode tunnel replay-window 42 "
+		    "auth-trunc 'hmac(sha1)' %s 96  enc 'cbc(aes)' %s",
+	    IP4_ADDR_VETH0, IP4_ADDR1_VETH1, XFRM_SPI_IN_TO_OUT, XFRM_AUTH, XFRM_ENC);
+	SYS(fail,
+	    "ip xfrm policy add src %s/32 dst %s/32 dir in "
+		    "tmpl src %s dst %s proto esp reqid 1 "
+		    "mode tunnel",
+	    IP4_ADDR_TUNL_DEV0, IP4_ADDR_TUNL_DEV1, IP4_ADDR_VETH0, IP4_ADDR1_VETH1);
+
+	/* root -> at_ns0 */
+	SYS(fail,
+	    "ip xfrm state add src %s dst %s proto esp "
+		    "spi %d reqid 2 mode tunnel "
+		    "auth-trunc 'hmac(sha1)' %s 96  enc 'cbc(aes)' %s",
+	    IP4_ADDR1_VETH1, IP4_ADDR_VETH0, XFRM_SPI_OUT_TO_IN, XFRM_AUTH, XFRM_ENC);
+	SYS(fail,
+	    "ip xfrm policy add src %s/32 dst %s/32 dir out "
+		    "tmpl src %s dst %s proto esp reqid 2 "
+		    "mode tunnel",
+	    IP4_ADDR_TUNL_DEV1, IP4_ADDR_TUNL_DEV0, IP4_ADDR1_VETH1, IP4_ADDR_VETH0);
+
+	/* address & route */
+	SYS(fail, "ip addr add dev veth1 %s/32", IP4_ADDR_TUNL_DEV1);
+	SYS(fail, "ip route add %s dev veth1 via %s src %s",
+	    IP4_ADDR_TUNL_DEV0, IP4_ADDR_VETH0, IP4_ADDR_TUNL_DEV1);
+
+	return 0;
+fail:
+	return -1;
+}
+
+static void delete_xfrm_tunnel(void)
+{
+	SYS_NOFAIL("ip xfrm policy delete dir out src %s/32 dst %s/32 2> /dev/null",
+		   IP4_ADDR_TUNL_DEV1, IP4_ADDR_TUNL_DEV0);
+	SYS_NOFAIL("ip xfrm policy delete dir in src %s/32 dst %s/32 2> /dev/null",
+		   IP4_ADDR_TUNL_DEV0, IP4_ADDR_TUNL_DEV1);
+	SYS_NOFAIL("ip xfrm state delete src %s dst %s proto esp spi %d 2> /dev/null",
+		   IP4_ADDR_VETH0, IP4_ADDR1_VETH1, XFRM_SPI_IN_TO_OUT);
+	SYS_NOFAIL("ip xfrm state delete src %s dst %s proto esp spi %d 2> /dev/null",
+		   IP4_ADDR1_VETH1, IP4_ADDR_VETH0, XFRM_SPI_OUT_TO_IN);
+}
+
 static int test_ping(int family, const char *addr)
 {
 	SYS(fail, "%s %s %s > /dev/null", ping_command(family), PING_ARGS, addr);
@@ -532,25 +624,85 @@ static void test_ipip_tunnel(enum ipip_encap encap)
 		test_tunnel_kern__destroy(skel);
 }
 
+static void test_xfrm_tunnel(void)
+{
+	DECLARE_LIBBPF_OPTS(bpf_tc_hook, tc_hook,
+			    .attach_point = BPF_TC_INGRESS);
+	LIBBPF_OPTS(bpf_xdp_attach_opts, opts);
+	struct test_tunnel_kern *skel = NULL;
+	struct nstoken *nstoken;
+	int xdp_prog_fd;
+	int tc_prog_fd;
+	int ifindex;
+	int err;
+
+	err = add_xfrm_tunnel();
+	if (!ASSERT_OK(err, "add_xfrm_tunnel"))
+		return;
+
+	skel = test_tunnel_kern__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "test_tunnel_kern__open_and_load"))
+		goto done;
+
+	ifindex = if_nametoindex("veth1");
+	if (!ASSERT_NEQ(ifindex, 0, "veth1 ifindex"))
+		goto done;
+
+	/* attach tc prog to tunnel dev */
+	tc_hook.ifindex = ifindex;
+	tc_prog_fd = bpf_program__fd(skel->progs.xfrm_get_state);
+	if (!ASSERT_GE(tc_prog_fd, 0, "bpf_program__fd"))
+		goto done;
+	if (attach_tc_prog(&tc_hook, tc_prog_fd, -1))
+		goto done;
+
+	/* attach xdp prog to tunnel dev */
+	xdp_prog_fd = bpf_program__fd(skel->progs.xfrm_get_state_xdp);
+	if (!ASSERT_GE(xdp_prog_fd, 0, "bpf_program__fd"))
+		goto done;
+	err = bpf_xdp_attach(ifindex, xdp_prog_fd, XDP_FLAGS_REPLACE, &opts);
+	if (!ASSERT_OK(err, "bpf_xdp_attach"))
+		goto done;
+
+	/* ping from at_ns0 namespace test */
+	nstoken = open_netns("at_ns0");
+	err = test_ping(AF_INET, IP4_ADDR_TUNL_DEV1);
+	close_netns(nstoken);
+	if (!ASSERT_OK(err, "test_ping"))
+		goto done;
+
+	if (!ASSERT_EQ(skel->bss->xfrm_reqid, 1, "req_id"))
+		goto done;
+	if (!ASSERT_EQ(skel->bss->xfrm_spi, XFRM_SPI_IN_TO_OUT, "spi"))
+		goto done;
+	if (!ASSERT_EQ(skel->bss->xfrm_remote_ip, 0xac100164, "remote_ip"))
+		goto done;
+	if (!ASSERT_EQ(skel->bss->xfrm_replay_window, 42, "replay_window"))
+		goto done;
+
+done:
+	delete_xfrm_tunnel();
+	if (skel)
+		test_tunnel_kern__destroy(skel);
+}
+
 #define RUN_TEST(name, ...)						\
 	({								\
 		if (test__start_subtest(#name)) {			\
+			config_device();				\
 			test_ ## name(__VA_ARGS__);			\
+			cleanup();					\
 		}							\
 	})
 
 static void *test_tunnel_run_tests(void *arg)
 {
-	cleanup();
-	config_device();
-
 	RUN_TEST(vxlan_tunnel);
 	RUN_TEST(ip6vxlan_tunnel);
 	RUN_TEST(ipip_tunnel, NONE);
 	RUN_TEST(ipip_tunnel, FOU);
 	RUN_TEST(ipip_tunnel, GUE);
-
-	cleanup();
+	RUN_TEST(xfrm_tunnel);
 
 	return NULL;
 }
diff --git a/tools/testing/selftests/bpf/progs/bpf_tracing_net.h b/tools/testing/selftests/bpf/progs/bpf_tracing_net.h
@@ -26,6 +26,7 @@
 #define IPV6_AUTOFLOWLABEL	70
 
 #define TC_ACT_UNSPEC		(-1)
+#define TC_ACT_OK		0
 #define TC_ACT_SHOT		2
 
 #define SOL_TCP			6
diff --git a/tools/testing/selftests/bpf/progs/test_tunnel_kern.c b/tools/testing/selftests/bpf/progs/test_tunnel_kern.c
diff --git a/tools/testing/selftests/bpf/test_tunnel.sh b/tools/testing/selftests/bpf/test_tunnel.sh

Original file line number	Diff line number	Diff line change
`@@ -4218,6 +4218,8 @@ void __init xfrm_init(void)`
`4218`	`4218`	`#ifdef CONFIG_XFRM_ESPINTCP`
`4219`	`4219`	`espintcp_init();`
`4220`	`4220`	`#endif`
	`4221`	`+`
	`4222`	`+ register_xfrm_state_bpf();`
`4221`	`4223`	`}`
`4222`	`4224`
`4223`	`4225`	`#ifdef CONFIG_AUDITSYSCALL`