Merge branch 'Introduce unstable CT lookup helpers'

Kumar Kartikeya says:

====================

This series adds unstable conntrack lookup helpers using BPF kfunc support.  The
patch adding the lookup helper is based off of Maxim's recent patch to aid in
rebasing their series on top of this, all adjusted to work with module kfuncs [0].

  [0]: https://lore.kernel.org/bpf/20211019144655.3483197-8-maximmi@nvidia.com

To enable returning a reference to struct nf_conn, the verifier is extended to
support reference tracking for PTR_TO_BTF_ID, and kfunc is extended with support
for working as acquire/release functions, similar to existing BPF helpers. kfunc
returning pointer (limited to PTR_TO_BTF_ID in the kernel) can also return a
PTR_TO_BTF_ID_OR_NULL now, typically needed when acquiring a resource can fail.
kfunc can also receive PTR_TO_CTX and PTR_TO_MEM (with some limitations) as
arguments now. There is also support for passing a mem, len pair as argument
to kfunc now. In such cases, passing pointer to unsized type (void) is also
permitted.

Please see individual commits for details.

Changelog:
----------
v7 -> v8:
v7: https://lore.kernel.org/bpf/20220111180428.931466-1-memxor@gmail.com

 * Move enum btf_kfunc_hook to btf.c (Alexei)
 * Drop verbose log for unlikely failure case in __find_kfunc_desc_btf (Alexei)
 * Remove unnecessary barrier in register_btf_kfunc_id_set (Alexei)
 * Switch macro in bpf_nf test to __always_inline function (Alexei)

v6 -> v7:
v6: https://lore.kernel.org/bpf/20220102162115.1506833-1-memxor@gmail.com

 * Drop try_module_get_live patch, use flag in btf_module struct (Alexei)
 * Add comments and expand commit message detailing why we have to concatenate
   and sort vmlinux kfunc BTF ID sets (Alexei)
 * Use bpf_testmod for testing btf_try_get_module race (Alexei)
 * Use bpf_prog_type for both btf_kfunc_id_set_contains and
   register_btf_kfunc_id_set calls (Alexei)
 * In case of module set registration, directly assign set (Alexei)
 * Add CONFIG_USERFAULTFD=y to selftest config
 * Fix other nits

v5 -> v6:
v5: https://lore.kernel.org/bpf/20211230023705.3860970-1-memxor@gmail.com

 * Fix for a bug in btf_try_get_module leading to use-after-free
 * Drop *kallsyms_on_each_symbol loop, reinstate register_btf_kfunc_id_set (Alexei)
 * btf_free_kfunc_set_tab now takes struct btf, and handles resetting tab to NULL
 * Check return value btf_name_by_offset for param_name
 * Instead of using tmp_set, use btf->kfunc_set_tab directly, and simplify cleanup

v4 -> v5:
v4: https://lore.kernel.org/bpf/20211217015031.1278167-1-memxor@gmail.com

 * Move nf_conntrack helpers code to its own separate file (Toke, Pablo)
 * Remove verifier callbacks, put btf_id_sets in struct btf (Alexei)
  * Convert the in-kernel users away from the old API
 * Change len__ prefix convention to __sz suffix (Alexei)
 * Drop parent_ref_obj_id patch (Alexei)

v3 -> v4:
v3: https://lore.kernel.org/bpf/20211210130230.4128676-1-memxor@gmail.com

 * Guard unstable CT helpers with CONFIG_DEBUG_INFO_BTF_MODULES
 * Move addition of prog_test test kfuncs to selftest commit
 * Move negative kfunc tests to test_verifier suite
 * Limit struct nesting depth to 4, which should be enough for now

v2 -> v3:
v2: https://lore.kernel.org/bpf/20211209170929.3485242-1-memxor@gmail.com

 * Fix build error for !CONFIG_BPF_SYSCALL (Patchwork)

RFC v1 -> v2:
v1: https://lore.kernel.org/bpf/20211030144609.263572-1-memxor@gmail.com

 * Limit PTR_TO_MEM support to pointer to scalar, or struct with scalars (Alexei)
 * Use btf_id_set for checking acquire, release, ret type null (Alexei)
 * Introduce opts struct for CT helpers, move int err parameter to it
 * Add l4proto as parameter to CT helper's opts, remove separate tcp/udp helpers
 * Add support for mem, len argument pair to kfunc
 * Allow void * as pointer type for mem, len argument pair
 * Extend selftests to cover new additions to kfuncs
 * Copy ref_obj_id to PTR_TO_BTF_ID dst_reg on btf_struct_access, test it
 * Fix other misc nits, bugs, and expand commit messages
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>

10 files changed, 454 insertions, 43 deletions

diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c
index 46dd95755967..3a5bf8ad834d 100644
--- a/net/bpf/test_run.c
+++ b/net/bpf/test_run.c
@@ -5,6 +5,7 @@
 #include <linux/btf.h>
 #include <linux/btf_ids.h>
 #include <linux/slab.h>
+#include <linux/init.h>
 #include <linux/vmalloc.h>
 #include <linux/etherdevice.h>
 #include <linux/filter.h>
@@ -171,6 +172,8 @@ int noinline bpf_fentry_test1(int a)
 {
 	return a + 1;
 }
+EXPORT_SYMBOL_GPL(bpf_fentry_test1);
+ALLOW_ERROR_INJECTION(bpf_fentry_test1, ERRNO);
 
 int noinline bpf_fentry_test2(int a, u64 b)
 {
@@ -232,22 +235,137 @@ struct sock * noinline bpf_kfunc_call_test3(struct sock *sk)
 	return sk;
 }
 
+struct prog_test_ref_kfunc {
+	int a;
+	int b;
+	struct prog_test_ref_kfunc *next;
+};
+
+static struct prog_test_ref_kfunc prog_test_struct = {
+	.a = 42,
+	.b = 108,
+	.next = &prog_test_struct,
+};
+
+noinline struct prog_test_ref_kfunc *
+bpf_kfunc_call_test_acquire(unsigned long *scalar_ptr)
+{
+	/* randomly return NULL */
+	if (get_jiffies_64() % 2)
+		return NULL;
+	return &prog_test_struct;
+}
+
+noinline void bpf_kfunc_call_test_release(struct prog_test_ref_kfunc *p)
+{
+}
+
+struct prog_test_pass1 {
+	int x0;
+	struct {
+		int x1;
+		struct {
+			int x2;
+			struct {
+				int x3;
+			};
+		};
+	};
+};
+
+struct prog_test_pass2 {
+	int len;
+	short arr1[4];
+	struct {
+		char arr2[4];
+		unsigned long arr3[8];
+	} x;
+};
+
+struct prog_test_fail1 {
+	void *p;
+	int x;
+};
+
+struct prog_test_fail2 {
+	int x8;
+	struct prog_test_pass1 x;
+};
+
+struct prog_test_fail3 {
+	int len;
+	char arr1[2];
+	char arr2[0];
+};
+
+noinline void bpf_kfunc_call_test_pass_ctx(struct __sk_buff *skb)
+{
+}
+
+noinline void bpf_kfunc_call_test_pass1(struct prog_test_pass1 *p)
+{
+}
+
+noinline void bpf_kfunc_call_test_pass2(struct prog_test_pass2 *p)
+{
+}
+
+noinline void bpf_kfunc_call_test_fail1(struct prog_test_fail1 *p)
+{
+}
+
+noinline void bpf_kfunc_call_test_fail2(struct prog_test_fail2 *p)
+{
+}
+
+noinline void bpf_kfunc_call_test_fail3(struct prog_test_fail3 *p)
+{
+}
+
+noinline void bpf_kfunc_call_test_mem_len_pass1(void *mem, int mem__sz)
+{
+}
+
+noinline void bpf_kfunc_call_test_mem_len_fail1(void *mem, int len)
+{
+}
+
+noinline void bpf_kfunc_call_test_mem_len_fail2(u64 *mem, int len)
+{
+}
+
 __diag_pop();
 
 ALLOW_ERROR_INJECTION(bpf_modify_return_test, ERRNO);
 
-BTF_SET_START(test_sk_kfunc_ids)
+BTF_SET_START(test_sk_check_kfunc_ids)
 BTF_ID(func, bpf_kfunc_call_test1)
 BTF_ID(func, bpf_kfunc_call_test2)
 BTF_ID(func, bpf_kfunc_call_test3)
-BTF_SET_END(test_sk_kfunc_ids)
-
-bool bpf_prog_test_check_kfunc_call(u32 kfunc_id, struct module *owner)
-{
-	if (btf_id_set_contains(&test_sk_kfunc_ids, kfunc_id))
-		return true;
-	return bpf_check_mod_kfunc_call(&prog_test_kfunc_list, kfunc_id, owner);
-}
+BTF_ID(func, bpf_kfunc_call_test_acquire)
+BTF_ID(func, bpf_kfunc_call_test_release)
+BTF_ID(func, bpf_kfunc_call_test_pass_ctx)
+BTF_ID(func, bpf_kfunc_call_test_pass1)
+BTF_ID(func, bpf_kfunc_call_test_pass2)
+BTF_ID(func, bpf_kfunc_call_test_fail1)
+BTF_ID(func, bpf_kfunc_call_test_fail2)
+BTF_ID(func, bpf_kfunc_call_test_fail3)
+BTF_ID(func, bpf_kfunc_call_test_mem_len_pass1)
+BTF_ID(func, bpf_kfunc_call_test_mem_len_fail1)
+BTF_ID(func, bpf_kfunc_call_test_mem_len_fail2)
+BTF_SET_END(test_sk_check_kfunc_ids)
+
+BTF_SET_START(test_sk_acquire_kfunc_ids)
+BTF_ID(func, bpf_kfunc_call_test_acquire)
+BTF_SET_END(test_sk_acquire_kfunc_ids)
+
+BTF_SET_START(test_sk_release_kfunc_ids)
+BTF_ID(func, bpf_kfunc_call_test_release)
+BTF_SET_END(test_sk_release_kfunc_ids)
+
+BTF_SET_START(test_sk_ret_null_kfunc_ids)
+BTF_ID(func, bpf_kfunc_call_test_acquire)
+BTF_SET_END(test_sk_ret_null_kfunc_ids)
 
 static void *bpf_test_init(const union bpf_attr *kattr, u32 size,
 			   u32 headroom, u32 tailroom)
@@ -1067,3 +1185,17 @@ out:
 	kfree(ctx);
 	return err;
 }
+
+static const struct btf_kfunc_id_set bpf_prog_test_kfunc_set = {
+	.owner        = THIS_MODULE,
+	.check_set    = &test_sk_check_kfunc_ids,
+	.acquire_set  = &test_sk_acquire_kfunc_ids,
+	.release_set  = &test_sk_release_kfunc_ids,
+	.ret_null_set = &test_sk_ret_null_kfunc_ids,
+};
+
+static int __init bpf_prog_test_run_init(void)
+{
+	return register_btf_kfunc_id_set(BPF_PROG_TYPE_SCHED_CLS, &bpf_prog_test_kfunc_set);
+}
+late_initcall(bpf_prog_test_run_init);
diff --git a/net/core/filter.c b/net/core/filter.c
index 4603b7cd3cd1..f73a84c75970 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -10062,7 +10062,6 @@ const struct bpf_verifier_ops tc_cls_act_verifier_ops = {
 	.convert_ctx_access	= tc_cls_act_convert_ctx_access,
 	.gen_prologue		= tc_cls_act_prologue,
 	.gen_ld_abs		= bpf_gen_ld_abs,
-	.check_kfunc_call	= bpf_prog_test_check_kfunc_call,
 };
 
 const struct bpf_prog_ops tc_cls_act_prog_ops = {
diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c
index 9b7171c40434..3b471781327f 100644
--- a/net/core/net_namespace.c
+++ b/net/core/net_namespace.c
@@ -299,6 +299,7 @@ struct net *get_net_ns_by_id(const struct net *net, int id)
 
 	return peer;
 }
+EXPORT_SYMBOL_GPL(get_net_ns_by_id);
 
 /*
  * setup_net runs the initializers for the network namespace object.
diff --git a/net/ipv4/bpf_tcp_ca.c b/net/ipv4/bpf_tcp_ca.c
index de610cb83694..b60c9fd7147e 100644
--- a/net/ipv4/bpf_tcp_ca.c
+++ b/net/ipv4/bpf_tcp_ca.c
@@ -1,6 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 /* Copyright (c) 2019 Facebook  */
 
+#include <linux/init.h>
 #include <linux/types.h>
 #include <linux/bpf_verifier.h>
 #include <linux/bpf.h>
@@ -212,26 +213,23 @@ bpf_tcp_ca_get_func_proto(enum bpf_func_id func_id,
 	}
 }
 
-BTF_SET_START(bpf_tcp_ca_kfunc_ids)
+BTF_SET_START(bpf_tcp_ca_check_kfunc_ids)
 BTF_ID(func, tcp_reno_ssthresh)
 BTF_ID(func, tcp_reno_cong_avoid)
 BTF_ID(func, tcp_reno_undo_cwnd)
 BTF_ID(func, tcp_slow_start)
 BTF_ID(func, tcp_cong_avoid_ai)
-BTF_SET_END(bpf_tcp_ca_kfunc_ids)
+BTF_SET_END(bpf_tcp_ca_check_kfunc_ids)
 
-static bool bpf_tcp_ca_check_kfunc_call(u32 kfunc_btf_id, struct module *owner)
-{
-	if (btf_id_set_contains(&bpf_tcp_ca_kfunc_ids, kfunc_btf_id))
-		return true;
-	return bpf_check_mod_kfunc_call(&bpf_tcp_ca_kfunc_list, kfunc_btf_id, owner);
-}
+static const struct btf_kfunc_id_set bpf_tcp_ca_kfunc_set = {
+	.owner     = THIS_MODULE,
+	.check_set = &bpf_tcp_ca_check_kfunc_ids,
+};
 
 static const struct bpf_verifier_ops bpf_tcp_ca_verifier_ops = {
 	.get_func_proto		= bpf_tcp_ca_get_func_proto,
 	.is_valid_access	= bpf_tcp_ca_is_valid_access,
 	.btf_struct_access	= bpf_tcp_ca_btf_struct_access,
-	.check_kfunc_call	= bpf_tcp_ca_check_kfunc_call,
 };
 
 static int bpf_tcp_ca_init_member(const struct btf_type *t,
@@ -300,3 +298,9 @@ struct bpf_struct_ops bpf_tcp_congestion_ops = {
 	.init = bpf_tcp_ca_init,
 	.name = "tcp_congestion_ops",
 };
+
+static int __init bpf_tcp_ca_kfunc_init(void)
+{
+	return register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, &bpf_tcp_ca_kfunc_set);
+}
+late_initcall(bpf_tcp_ca_kfunc_init);
diff --git a/net/ipv4/tcp_bbr.c b/net/ipv4/tcp_bbr.c
index ec5550089b4d..02e8626ccb27 100644
--- a/net/ipv4/tcp_bbr.c
+++ b/net/ipv4/tcp_bbr.c
@@ -1154,7 +1154,7 @@ static struct tcp_congestion_ops tcp_bbr_cong_ops __read_mostly = {
 	.set_state	= bbr_set_state,
 };
 
-BTF_SET_START(tcp_bbr_kfunc_ids)
+BTF_SET_START(tcp_bbr_check_kfunc_ids)
 #ifdef CONFIG_X86
 #ifdef CONFIG_DYNAMIC_FTRACE
 BTF_ID(func, bbr_init)
@@ -1167,25 +1167,27 @@ BTF_ID(func, bbr_min_tso_segs)
 BTF_ID(func, bbr_set_state)
 #endif
 #endif
-BTF_SET_END(tcp_bbr_kfunc_ids)
+BTF_SET_END(tcp_bbr_check_kfunc_ids)
 
-static DEFINE_KFUNC_BTF_ID_SET(&tcp_bbr_kfunc_ids, tcp_bbr_kfunc_btf_set);
+static const struct btf_kfunc_id_set tcp_bbr_kfunc_set = {
+	.owner     = THIS_MODULE,
+	.check_set = &tcp_bbr_check_kfunc_ids,
+};
 
 static int __init bbr_register(void)
 {
 	int ret;
 
 	BUILD_BUG_ON(sizeof(struct bbr) > ICSK_CA_PRIV_SIZE);
-	ret = tcp_register_congestion_control(&tcp_bbr_cong_ops);
-	if (ret)
+
+	ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, &tcp_bbr_kfunc_set);
+	if (ret < 0)
 		return ret;
-	register_kfunc_btf_id_set(&bpf_tcp_ca_kfunc_list, &tcp_bbr_kfunc_btf_set);
-	return 0;
+	return tcp_register_congestion_control(&tcp_bbr_cong_ops);
 }
 
 static void __exit bbr_unregister(void)
 {
-	unregister_kfunc_btf_id_set(&bpf_tcp_ca_kfunc_list, &tcp_bbr_kfunc_btf_set);
 	tcp_unregister_congestion_control(&tcp_bbr_cong_ops);
 }
 
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index e07837e23b3f..24d562dd6225 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -485,7 +485,7 @@ static struct tcp_congestion_ops cubictcp __read_mostly = {
 	.name		= "cubic",
 };
 
-BTF_SET_START(tcp_cubic_kfunc_ids)
+BTF_SET_START(tcp_cubic_check_kfunc_ids)
 #ifdef CONFIG_X86
 #ifdef CONFIG_DYNAMIC_FTRACE
 BTF_ID(func, cubictcp_init)
@@ -496,9 +496,12 @@ BTF_ID(func, cubictcp_cwnd_event)
 BTF_ID(func, cubictcp_acked)
 #endif
 #endif
-BTF_SET_END(tcp_cubic_kfunc_ids)
+BTF_SET_END(tcp_cubic_check_kfunc_ids)
 
-static DEFINE_KFUNC_BTF_ID_SET(&tcp_cubic_kfunc_ids, tcp_cubic_kfunc_btf_set);
+static const struct btf_kfunc_id_set tcp_cubic_kfunc_set = {
+	.owner     = THIS_MODULE,
+	.check_set = &tcp_cubic_check_kfunc_ids,
+};
 
 static int __init cubictcp_register(void)
 {
@@ -534,16 +537,14 @@ static int __init cubictcp_register(void)
 	/* divide by bic_scale and by constant Srtt (100ms) */
 	do_div(cube_factor, bic_scale * 10);
 
-	ret = tcp_register_congestion_control(&cubictcp);
-	if (ret)
+	ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, &tcp_cubic_kfunc_set);
+	if (ret < 0)
 		return ret;
-	register_kfunc_btf_id_set(&bpf_tcp_ca_kfunc_list, &tcp_cubic_kfunc_btf_set);
-	return 0;
+	return tcp_register_congestion_control(&cubictcp);
 }
 
 static void __exit cubictcp_unregister(void)
 {
-	unregister_kfunc_btf_id_set(&bpf_tcp_ca_kfunc_list, &tcp_cubic_kfunc_btf_set);
 	tcp_unregister_congestion_control(&cubictcp);
 }
 
diff --git a/net/ipv4/tcp_dctcp.c b/net/ipv4/tcp_dctcp.c
index 0d7ab3cc7b61..1943a6630341 100644
--- a/net/ipv4/tcp_dctcp.c
+++ b/net/ipv4/tcp_dctcp.c
@@ -238,7 +238,7 @@ static struct tcp_congestion_ops dctcp_reno __read_mostly = {
 	.name		= "dctcp-reno",
 };
 
-BTF_SET_START(tcp_dctcp_kfunc_ids)
+BTF_SET_START(tcp_dctcp_check_kfunc_ids)
 #ifdef CONFIG_X86
 #ifdef CONFIG_DYNAMIC_FTRACE
 BTF_ID(func, dctcp_init)
@@ -249,25 +249,27 @@ BTF_ID(func, dctcp_cwnd_undo)
 BTF_ID(func, dctcp_state)
 #endif
 #endif
-BTF_SET_END(tcp_dctcp_kfunc_ids)
+BTF_SET_END(tcp_dctcp_check_kfunc_ids)
 
-static DEFINE_KFUNC_BTF_ID_SET(&tcp_dctcp_kfunc_ids, tcp_dctcp_kfunc_btf_set);
+static const struct btf_kfunc_id_set tcp_dctcp_kfunc_set = {
+	.owner     = THIS_MODULE,
+	.check_set = &tcp_dctcp_check_kfunc_ids,
+};
 
 static int __init dctcp_register(void)
 {
 	int ret;
 
 	BUILD_BUG_ON(sizeof(struct dctcp) > ICSK_CA_PRIV_SIZE);
-	ret = tcp_register_congestion_control(&dctcp);
-	if (ret)
+
+	ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, &tcp_dctcp_kfunc_set);
+	if (ret < 0)
 		return ret;
-	register_kfunc_btf_id_set(&bpf_tcp_ca_kfunc_list, &tcp_dctcp_kfunc_btf_set);
-	return 0;
+	return tcp_register_congestion_control(&dctcp);
 }
 
 static void __exit dctcp_unregister(void)
 {
-	unregister_kfunc_btf_id_set(&bpf_tcp_ca_kfunc_list, &tcp_dctcp_kfunc_btf_set);
 	tcp_unregister_congestion_control(&dctcp);
 }
 
diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile
index a135b1a46014..238b6a620e88 100644
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -14,6 +14,11 @@ nf_conntrack-$(CONFIG_NF_CONNTRACK_LABELS) += nf_conntrack_labels.o
 nf_conntrack-$(CONFIG_NF_CT_PROTO_DCCP) += nf_conntrack_proto_dccp.o
 nf_conntrack-$(CONFIG_NF_CT_PROTO_SCTP) += nf_conntrack_proto_sctp.o
 nf_conntrack-$(CONFIG_NF_CT_PROTO_GRE) += nf_conntrack_proto_gre.o
+ifeq ($(CONFIG_NF_CONNTRACK),m)
+nf_conntrack-$(CONFIG_DEBUG_INFO_BTF_MODULES) += nf_conntrack_bpf.o
+else ifeq ($(CONFIG_NF_CONNTRACK),y)
+nf_conntrack-$(CONFIG_DEBUG_INFO_BTF) += nf_conntrack_bpf.o
+endif
 
 obj-$(CONFIG_NETFILTER) = netfilter.o
 
diff --git a/net/netfilter/nf_conntrack_bpf.c b/net/netfilter/nf_conntrack_bpf.c
new file mode 100644
index 000000000000..8ad3f52579f3
--- /dev/null
+++ b/net/netfilter/nf_conntrack_bpf.c
@@ -0,0 +1,257 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Unstable Conntrack Helpers for XDP and TC-BPF hook
+ *
+ * These are called from the XDP and SCHED_CLS BPF programs. 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/types.h>
+#include <linux/btf_ids.h>
+#include <linux/net_namespace.h>
+#include <net/netfilter/nf_conntrack.h>
+#include <net/netfilter/nf_conntrack_core.h>
+
+/* bpf_ct_opts - Options for CT lookup helpers
+ *
+ * Members:
+ * @netns_id   - Specify the network namespace for lookup
+ *		 Values:
+ *		   BPF_F_CURRENT_NETNS (-1)
+ *		     Use namespace associated with ctx (xdp_md, __sk_buff)
+ *		   [0, S32_MAX]
+ *		     Network Namespace ID
+ * @error      - Out parameter, set for any errors encountered
+ *		 Values:
+ *		   -EINVAL - Passed NULL for bpf_tuple pointer
+ *		   -EINVAL - opts->reserved is not 0
+ *		   -EINVAL - netns_id is less than -1
+ *		   -EINVAL - opts__sz isn't NF_BPF_CT_OPTS_SZ (12)
+ *		   -EPROTO - l4proto isn't one of IPPROTO_TCP or IPPROTO_UDP
+ *		   -ENONET - No network namespace found for netns_id
+ *		   -ENOENT - Conntrack lookup could not find entry for tuple
+ *		   -EAFNOSUPPORT - tuple__sz isn't one of sizeof(tuple->ipv4)
+ *				   or sizeof(tuple->ipv6)
+ * @l4proto    - Layer 4 protocol
+ *		 Values:
+ *		   IPPROTO_TCP, IPPROTO_UDP
+ * @reserved   - Reserved member, will be reused for more options in future
+ *		 Values:
+ *		   0
+ */
+struct bpf_ct_opts {
+	s32 netns_id;
+	s32 error;
+	u8 l4proto;
+	u8 reserved[3];
+};
+
+enum {
+	NF_BPF_CT_OPTS_SZ = 12,
+};
+
+static struct nf_conn *__bpf_nf_ct_lookup(struct net *net,
+					  struct bpf_sock_tuple *bpf_tuple,
+					  u32 tuple_len, u8 protonum,
+					  s32 netns_id)
+{
+	struct nf_conntrack_tuple_hash *hash;
+	struct nf_conntrack_tuple tuple;
+
+	if (unlikely(protonum != IPPROTO_TCP && protonum != IPPROTO_UDP))
+		return ERR_PTR(-EPROTO);
+	if (unlikely(netns_id < BPF_F_CURRENT_NETNS))
+		return ERR_PTR(-EINVAL);
+
+	memset(&tuple, 0, sizeof(tuple));
+	switch (tuple_len) {
+	case sizeof(bpf_tuple->ipv4):
+		tuple.src.l3num = AF_INET;
+		tuple.src.u3.ip = bpf_tuple->ipv4.saddr;
+		tuple.src.u.tcp.port = bpf_tuple->ipv4.sport;
+		tuple.dst.u3.ip = bpf_tuple->ipv4.daddr;
+		tuple.dst.u.tcp.port = bpf_tuple->ipv4.dport;
+		break;
+	case sizeof(bpf_tuple->ipv6):
+		tuple.src.l3num = AF_INET6;
+		memcpy(tuple.src.u3.ip6, bpf_tuple->ipv6.saddr, sizeof(bpf_tuple->ipv6.saddr));
+		tuple.src.u.tcp.port = bpf_tuple->ipv6.sport;
+		memcpy(tuple.dst.u3.ip6, bpf_tuple->ipv6.daddr, sizeof(bpf_tuple->ipv6.daddr));
+		tuple.dst.u.tcp.port = bpf_tuple->ipv6.dport;
+		break;
+	default:
+		return ERR_PTR(-EAFNOSUPPORT);
+	}
+
+	tuple.dst.protonum = protonum;
+
+	if (netns_id >= 0) {
+		net = get_net_ns_by_id(net, netns_id);
+		if (unlikely(!net))
+			return ERR_PTR(-ENONET);
+	}
+
+	hash = nf_conntrack_find_get(net, &nf_ct_zone_dflt, &tuple);
+	if (netns_id >= 0)
+		put_net(net);
+	if (!hash)
+		return ERR_PTR(-ENOENT);
+	return nf_ct_tuplehash_to_ctrack(hash);
+}
+
+__diag_push();
+__diag_ignore(GCC, 8, "-Wmissing-prototypes",
+	      "Global functions as their definitions will be in nf_conntrack BTF");
+
+/* bpf_xdp_ct_lookup - Lookup CT entry for the given tuple, and acquire a
+ *		       reference to it
+ *
+ * Parameters:
+ * @xdp_ctx	- Pointer to ctx (xdp_md) in XDP program
+ *		    Cannot be NULL
+ * @bpf_tuple	- Pointer to memory representing the tuple to look up
+ *		    Cannot be NULL
+ * @tuple__sz	- Length of the tuple structure
+ *		    Must be one of sizeof(bpf_tuple->ipv4) or
+ *		    sizeof(bpf_tuple->ipv6)
+ * @opts	- Additional options for lookup (documented above)
+ *		    Cannot be NULL
+ * @opts__sz	- Length of the bpf_ct_opts structure
+ *		    Must be NF_BPF_CT_OPTS_SZ (12)
+ */
+struct nf_conn *
+bpf_xdp_ct_lookup(struct xdp_md *xdp_ctx, struct bpf_sock_tuple *bpf_tuple,
+		  u32 tuple__sz, struct bpf_ct_opts *opts, u32 opts__sz)
+{
+	struct xdp_buff *ctx = (struct xdp_buff *)xdp_ctx;
+	struct net *caller_net;
+	struct nf_conn *nfct;
+
+	BUILD_BUG_ON(sizeof(struct bpf_ct_opts) != NF_BPF_CT_OPTS_SZ);
+
+	if (!opts)
+		return NULL;
+	if (!bpf_tuple || opts->reserved[0] || opts->reserved[1] ||
+	    opts->reserved[2] || opts__sz != NF_BPF_CT_OPTS_SZ) {
+		opts->error = -EINVAL;
+		return NULL;
+	}
+	caller_net = dev_net(ctx->rxq->dev);
+	nfct = __bpf_nf_ct_lookup(caller_net, bpf_tuple, tuple__sz, opts->l4proto,
+				  opts->netns_id);
+	if (IS_ERR(nfct)) {
+		opts->error = PTR_ERR(nfct);
+		return NULL;
+	}
+	return nfct;
+}
+
+/* bpf_skb_ct_lookup - Lookup CT entry for the given tuple, and acquire a
+ *		       reference to it
+ *
+ * Parameters:
+ * @skb_ctx	- Pointer to ctx (__sk_buff) in TC program
+ *		    Cannot be NULL
+ * @bpf_tuple	- Pointer to memory representing the tuple to look up
+ *		    Cannot be NULL
+ * @tuple__sz	- Length of the tuple structure
+ *		    Must be one of sizeof(bpf_tuple->ipv4) or
+ *		    sizeof(bpf_tuple->ipv6)
+ * @opts	- Additional options for lookup (documented above)
+ *		    Cannot be NULL
+ * @opts__sz	- Length of the bpf_ct_opts structure
+ *		    Must be NF_BPF_CT_OPTS_SZ (12)
+ */
+struct nf_conn *
+bpf_skb_ct_lookup(struct __sk_buff *skb_ctx, struct bpf_sock_tuple *bpf_tuple,
+		  u32 tuple__sz, struct bpf_ct_opts *opts, u32 opts__sz)
+{
+	struct sk_buff *skb = (struct sk_buff *)skb_ctx;
+	struct net *caller_net;
+	struct nf_conn *nfct;
+
+	BUILD_BUG_ON(sizeof(struct bpf_ct_opts) != NF_BPF_CT_OPTS_SZ);
+
+	if (!opts)
+		return NULL;
+	if (!bpf_tuple || opts->reserved[0] || opts->reserved[1] ||
+	    opts->reserved[2] || opts__sz != NF_BPF_CT_OPTS_SZ) {
+		opts->error = -EINVAL;
+		return NULL;
+	}
+	caller_net = skb->dev ? dev_net(skb->dev) : sock_net(skb->sk);
+	nfct = __bpf_nf_ct_lookup(caller_net, bpf_tuple, tuple__sz, opts->l4proto,
+				  opts->netns_id);
+	if (IS_ERR(nfct)) {
+		opts->error = PTR_ERR(nfct);
+		return NULL;
+	}
+	return nfct;
+}
+
+/* bpf_ct_release - Release acquired nf_conn 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:
+ * @nf_conn	 - Pointer to referenced nf_conn object, obtained using
+ *		   bpf_xdp_ct_lookup or bpf_skb_ct_lookup.
+ */
+void bpf_ct_release(struct nf_conn *nfct)
+{
+	if (!nfct)
+		return;
+	nf_ct_put(nfct);
+}
+
+__diag_pop()
+
+BTF_SET_START(nf_ct_xdp_check_kfunc_ids)
+BTF_ID(func, bpf_xdp_ct_lookup)
+BTF_ID(func, bpf_ct_release)
+BTF_SET_END(nf_ct_xdp_check_kfunc_ids)
+
+BTF_SET_START(nf_ct_tc_check_kfunc_ids)
+BTF_ID(func, bpf_skb_ct_lookup)
+BTF_ID(func, bpf_ct_release)
+BTF_SET_END(nf_ct_tc_check_kfunc_ids)
+
+BTF_SET_START(nf_ct_acquire_kfunc_ids)
+BTF_ID(func, bpf_xdp_ct_lookup)
+BTF_ID(func, bpf_skb_ct_lookup)
+BTF_SET_END(nf_ct_acquire_kfunc_ids)
+
+BTF_SET_START(nf_ct_release_kfunc_ids)
+BTF_ID(func, bpf_ct_release)
+BTF_SET_END(nf_ct_release_kfunc_ids)
+
+/* Both sets are identical */
+#define nf_ct_ret_null_kfunc_ids nf_ct_acquire_kfunc_ids
+
+static const struct btf_kfunc_id_set nf_conntrack_xdp_kfunc_set = {
+	.owner        = THIS_MODULE,
+	.check_set    = &nf_ct_xdp_check_kfunc_ids,
+	.acquire_set  = &nf_ct_acquire_kfunc_ids,
+	.release_set  = &nf_ct_release_kfunc_ids,
+	.ret_null_set = &nf_ct_ret_null_kfunc_ids,
+};
+
+static const struct btf_kfunc_id_set nf_conntrack_tc_kfunc_set = {
+	.owner        = THIS_MODULE,
+	.check_set    = &nf_ct_tc_check_kfunc_ids,
+	.acquire_set  = &nf_ct_acquire_kfunc_ids,
+	.release_set  = &nf_ct_release_kfunc_ids,
+	.ret_null_set = &nf_ct_ret_null_kfunc_ids,
+};
+
+int register_nf_conntrack_bpf(void)
+{
+	int ret;
+
+	ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_XDP, &nf_conntrack_xdp_kfunc_set);
+	return ret ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_SCHED_CLS, &nf_conntrack_tc_kfunc_set);
+}
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 894a325d39f2..3b60fca61920 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -34,6 +34,7 @@
 #include <linux/rculist_nulls.h>
 
 #include <net/netfilter/nf_conntrack.h>
+#include <net/netfilter/nf_conntrack_bpf.h>
 #include <net/netfilter/nf_conntrack_l4proto.h>
 #include <net/netfilter/nf_conntrack_expect.h>
 #include <net/netfilter/nf_conntrack_helper.h>
@@ -2748,8 +2749,15 @@ int nf_conntrack_init_start(void)
 	conntrack_gc_work_init(&conntrack_gc_work);
 	queue_delayed_work(system_power_efficient_wq, &conntrack_gc_work.dwork, HZ);
 
+	ret = register_nf_conntrack_bpf();
+	if (ret < 0)
+		goto err_kfunc;
+
 	return 0;
 
+err_kfunc:
+	cancel_delayed_work_sync(&conntrack_gc_work.dwork);
+	nf_conntrack_proto_fini();
 err_proto:
 	nf_conntrack_seqadj_fini();
 err_seqadj: