bpf: Support bpf program calling kernel function

This patch adds support to BPF verifier to allow bpf program calling
kernel function directly.

The use case included in this set is to allow bpf-tcp-cc to directly
call some tcp-cc helper functions (e.g. "tcp_cong_avoid_ai()").  Those
functions have already been used by some kernel tcp-cc implementations.

This set will also allow the bpf-tcp-cc program to directly call the
kernel tcp-cc implementation,  For example, a bpf_dctcp may only want to
implement its own dctcp_cwnd_event() and reuse other dctcp_*() directly
from the kernel tcp_dctcp.c instead of reimplementing (or
copy-and-pasting) them.

The tcp-cc kernel functions mentioned above will be white listed
for the struct_ops bpf-tcp-cc programs to use in a later patch.
The white listed functions are not bounded to a fixed ABI contract.
Those functions have already been used by the existing kernel tcp-cc.
If any of them has changed, both in-tree and out-of-tree kernel tcp-cc
implementations have to be changed.  The same goes for the struct_ops
bpf-tcp-cc programs which have to be adjusted accordingly.

This patch is to make the required changes in the bpf verifier.

First change is in btf.c, it adds a case in "btf_check_func_arg_match()".
When the passed in "btf->kernel_btf == true", it means matching the
verifier regs' states with a kernel function.  This will handle the
PTR_TO_BTF_ID reg.  It also maps PTR_TO_SOCK_COMMON, PTR_TO_SOCKET,
and PTR_TO_TCP_SOCK to its kernel's btf_id.

In the later libbpf patch, the insn calling a kernel function will
look like:

insn->code == (BPF_JMP | BPF_CALL)
insn->src_reg == BPF_PSEUDO_KFUNC_CALL /* <- new in this patch */
insn->imm == func_btf_id /* btf_id of the running kernel */

[ For the future calling function-in-kernel-module support, an array
  of module btf_fds can be passed at the load time and insn->off
  can be used to index into this array. ]

At the early stage of verifier, the verifier will collect all kernel
function calls into "struct bpf_kfunc_desc".  Those
descriptors are stored in "prog->aux->kfunc_tab" and will
be available to the JIT.  Since this "add" operation is similar
to the current "add_subprog()" and looking for the same insn->code,
they are done together in the new "add_subprog_and_kfunc()".

In the "do_check()" stage, the new "check_kfunc_call()" is added
to verify the kernel function call instruction:
1. Ensure the kernel function can be used by a particular BPF_PROG_TYPE.
   A new bpf_verifier_ops "check_kfunc_call" is added to do that.
   The bpf-tcp-cc struct_ops program will implement this function in
   a later patch.
2. Call "btf_check_kfunc_args_match()" to ensure the regs can be
   used as the args of a kernel function.
3. Mark the regs' type, subreg_def, and zext_dst.

At the later do_misc_fixups() stage, the new fixup_kfunc_call()
will replace the insn->imm with the function address (relative
to __bpf_call_base).  If needed, the jit can find the btf_func_model
by calling the new bpf_jit_find_kfunc_model(prog, insn).
With the imm set to the function address, "bpftool prog dump xlated"
will be able to display the kernel function calls the same way as
it displays other bpf helper calls.

gpl_compatible program is required to call kernel function.

This feature currently requires JIT.

The verifier selftests are adjusted because of the changes in
the verbose log in add_subprog_and_kfunc().

Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210325015142.1544736-1-kafai@fb.com

13 files changed, 480 insertions, 46 deletions

diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index b35fc8023884..9eead60f0301 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -2346,3 +2346,8 @@ out:
 					   tmp : orig_prog);
 	return prog;
 }
+
+bool bpf_jit_supports_kfunc_call(void)
+{
+	return true;
+}
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index eaae618a90b5..b5b7967e3ff3 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -427,6 +427,7 @@ enum bpf_reg_type {
 	PTR_TO_PERCPU_BTF_ID,	 /* reg points to a percpu kernel variable */
 	PTR_TO_FUNC,		 /* reg points to a bpf program function */
 	PTR_TO_MAP_KEY,		 /* reg points to a map element key */
+	__BPF_REG_TYPE_MAX,
 };
 
 /* The information passed from prog-specific *_is_valid_access
@@ -480,6 +481,7 @@ struct bpf_verifier_ops {
 				 const struct btf_type *t, int off, int size,
 				 enum bpf_access_type atype,
 				 u32 *next_btf_id);
+	bool (*check_kfunc_call)(u32 kfunc_btf_id);
 };
 
 struct bpf_prog_offload_ops {
@@ -796,6 +798,8 @@ struct btf_mod_pair {
 	struct module *module;
 };
 
+struct bpf_kfunc_desc_tab;
+
 struct bpf_prog_aux {
 	atomic64_t refcnt;
 	u32 used_map_cnt;
@@ -832,6 +836,7 @@ struct bpf_prog_aux {
 	struct bpf_prog **func;
 	void *jit_data; /* JIT specific data. arch dependent */
 	struct bpf_jit_poke_descriptor *poke_tab;
+	struct bpf_kfunc_desc_tab *kfunc_tab;
 	u32 size_poke_tab;
 	struct bpf_ksym ksym;
 	const struct bpf_prog_ops *ops;
@@ -1547,6 +1552,9 @@ int btf_distill_func_proto(struct bpf_verifier_log *log,
 struct bpf_reg_state;
 int btf_check_subprog_arg_match(struct bpf_verifier_env *env, int subprog,
 				struct bpf_reg_state *regs);
+int btf_check_kfunc_arg_match(struct bpf_verifier_env *env,
+			      const struct btf *btf, u32 func_id,
+			      struct bpf_reg_state *regs);
 int btf_prepare_func_args(struct bpf_verifier_env *env, int subprog,
 			  struct bpf_reg_state *reg);
 int btf_check_type_match(struct bpf_verifier_log *log, const struct bpf_prog *prog,
@@ -1557,6 +1565,10 @@ struct bpf_link *bpf_link_by_id(u32 id);
 
 const struct bpf_func_proto *bpf_base_func_proto(enum bpf_func_id func_id);
 void bpf_task_storage_free(struct task_struct *task);
+bool bpf_prog_has_kfunc_call(const struct bpf_prog *prog);
+const struct btf_func_model *
+bpf_jit_find_kfunc_model(const struct bpf_prog *prog,
+			 const struct bpf_insn *insn);
 #else /* !CONFIG_BPF_SYSCALL */
 static inline struct bpf_prog *bpf_prog_get(u32 ufd)
 {
@@ -1737,6 +1749,18 @@ bpf_base_func_proto(enum bpf_func_id func_id)
 static inline void bpf_task_storage_free(struct task_struct *task)
 {
 }
+
+static inline bool bpf_prog_has_kfunc_call(const struct bpf_prog *prog)
+{
+	return false;
+}
+
+static inline const struct btf_func_model *
+bpf_jit_find_kfunc_model(const struct bpf_prog *prog,
+			 const struct bpf_insn *insn)
+{
+	return NULL;
+}
 #endif /* CONFIG_BPF_SYSCALL */
 
 void __bpf_free_used_btfs(struct bpf_prog_aux *aux,
diff --git a/include/linux/btf.h b/include/linux/btf.h
index 8a05687a4ee2..3bac66e0183a 100644
--- a/include/linux/btf.h
+++ b/include/linux/btf.h
@@ -110,6 +110,7 @@ const struct btf_type *btf_type_resolve_func_ptr(const struct btf *btf,
 const struct btf_type *
 btf_resolve_size(const struct btf *btf, const struct btf_type *type,
 		 u32 *type_size);
+const char *btf_type_str(const struct btf_type *t);
 
 #define for_each_member(i, struct_type, member)			\
 	for (i = 0, member = btf_type_member(struct_type);	\
diff --git a/include/linux/filter.h b/include/linux/filter.h
index 0d9c710eb050..eecfd82db648 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -918,6 +918,7 @@ u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
 struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog);
 void bpf_jit_compile(struct bpf_prog *prog);
 bool bpf_jit_needs_zext(void);
+bool bpf_jit_supports_kfunc_call(void);
 bool bpf_helper_changes_pkt_data(void *func);
 
 static inline bool bpf_dump_raw_ok(const struct cred *cred)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 008edc1dc8c1..598716742593 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -1117,6 +1117,10 @@ enum bpf_link_type {
  * offset to another bpf function
  */
 #define BPF_PSEUDO_CALL		1
+/* when bpf_call->src_reg == BPF_PSEUDO_KFUNC_CALL,
+ * bpf_call->imm == btf_id of a BTF_KIND_FUNC in the running kernel
+ */
+#define BPF_PSEUDO_KFUNC_CALL	2
 
 /* flags for BPF_MAP_UPDATE_ELEM command */
 enum {
diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
index 3c489adacf3b..ec8afc4bc560 100644
--- a/kernel/bpf/btf.c
+++ b/kernel/bpf/btf.c
@@ -283,7 +283,7 @@ static const char * const btf_kind_str[NR_BTF_KINDS] = {
 	[BTF_KIND_FLOAT]	= "FLOAT",
 };
 
-static const char *btf_type_str(const struct btf_type *t)
+const char *btf_type_str(const struct btf_type *t)
 {
 	return btf_kind_str[BTF_INFO_KIND(t->info)];
 }
@@ -5362,6 +5362,14 @@ int btf_check_type_match(struct bpf_verifier_log *log, const struct bpf_prog *pr
 	return btf_check_func_type_match(log, btf1, t1, btf2, t2);
 }
 
+static u32 *reg2btf_ids[__BPF_REG_TYPE_MAX] = {
+#ifdef CONFIG_NET
+	[PTR_TO_SOCKET] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK],
+	[PTR_TO_SOCK_COMMON] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK_COMMON],
+	[PTR_TO_TCP_SOCK] = &btf_sock_ids[BTF_SOCK_TYPE_TCP],
+#endif
+};
+
 static int btf_check_func_arg_match(struct bpf_verifier_env *env,
 				    const struct btf *btf, u32 func_id,
 				    struct bpf_reg_state *regs,
@@ -5371,12 +5379,12 @@ static int btf_check_func_arg_match(struct bpf_verifier_env *env,
 	const char *func_name, *ref_tname;
 	const struct btf_type *t, *ref_t;
 	const struct btf_param *args;
-	u32 i, nargs;
+	u32 i, nargs, ref_id;
 
 	t = btf_type_by_id(btf, func_id);
 	if (!t || !btf_type_is_func(t)) {
 		/* These checks were already done by the verifier while loading
-		 * struct bpf_func_info
+		 * struct bpf_func_info or in add_kfunc_call().
 		 */
 		bpf_log(log, "BTF of func_id %u doesn't point to KIND_FUNC\n",
 			func_id);
@@ -5418,9 +5426,49 @@ static int btf_check_func_arg_match(struct bpf_verifier_env *env,
 			return -EINVAL;
 		}
 
-		ref_t = btf_type_skip_modifiers(btf, t->type, NULL);
+		ref_t = btf_type_skip_modifiers(btf, t->type, &ref_id);
 		ref_tname = btf_name_by_offset(btf, ref_t->name_off);
-		if (btf_get_prog_ctx_type(log, btf, t, env->prog->type, i)) {
+		if (btf_is_kernel(btf)) {
+			const struct btf_type *reg_ref_t;
+			const struct btf *reg_btf;
+			const char *reg_ref_tname;
+			u32 reg_ref_id;
+
+			if (!btf_type_is_struct(ref_t)) {
+				bpf_log(log, "kernel function %s args#%d pointer type %s %s is not supported\n",
+					func_name, i, btf_type_str(ref_t),
+					ref_tname);
+				return -EINVAL;
+			}
+
+			if (reg->type == PTR_TO_BTF_ID) {
+				reg_btf = reg->btf;
+				reg_ref_id = reg->btf_id;
+			} else if (reg2btf_ids[reg->type]) {
+				reg_btf = btf_vmlinux;
+				reg_ref_id = *reg2btf_ids[reg->type];
+			} else {
+				bpf_log(log, "kernel function %s args#%d expected pointer to %s %s but R%d is not a pointer to btf_id\n",
+					func_name, i,
+					btf_type_str(ref_t), ref_tname, regno);
+				return -EINVAL;
+			}
+
+			reg_ref_t = btf_type_skip_modifiers(reg_btf, reg_ref_id,
+							    &reg_ref_id);
+			reg_ref_tname = btf_name_by_offset(reg_btf,
+							   reg_ref_t->name_off);
+			if (!btf_struct_ids_match(log, reg_btf, reg_ref_id,
+						  reg->off, btf, ref_id)) {
+				bpf_log(log, "kernel function %s args#%d expected pointer to %s %s but R%d has a pointer to %s %s\n",
+					func_name, i,
+					btf_type_str(ref_t), ref_tname,
+					regno, btf_type_str(reg_ref_t),
+					reg_ref_tname);
+				return -EINVAL;
+			}
+		} else if (btf_get_prog_ctx_type(log, btf, t,
+						 env->prog->type, i)) {
 			/* If function expects ctx type in BTF check that caller
 			 * is passing PTR_TO_CTX.
 			 */
@@ -5493,6 +5541,13 @@ int btf_check_subprog_arg_match(struct bpf_verifier_env *env, int subprog,
 	return err;
 }
 
+int btf_check_kfunc_arg_match(struct bpf_verifier_env *env,
+			      const struct btf *btf, u32 func_id,
+			      struct bpf_reg_state *regs)
+{
+	return btf_check_func_arg_match(env, btf, func_id, regs, false);
+}
+
 /* Convert BTF of a function into bpf_reg_state if possible
  * Returns:
  * EFAULT - there is a verifier bug. Abort verification.
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index a35eb3d7b126..f5423251c118 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -159,6 +159,9 @@ void bpf_prog_jit_attempt_done(struct bpf_prog *prog)
 		kvfree(prog->aux->jited_linfo);
 		prog->aux->jited_linfo = NULL;
 	}
+
+	kfree(prog->aux->kfunc_tab);
+	prog->aux->kfunc_tab = NULL;
 }
 
 /* The jit engine is responsible to provide an array
@@ -1840,9 +1843,15 @@ struct bpf_prog *bpf_prog_select_runtime(struct bpf_prog *fp, int *err)
 	/* In case of BPF to BPF calls, verifier did all the prep
 	 * work with regards to JITing, etc.
 	 */
+	bool jit_needed = false;
+
 	if (fp->bpf_func)
 		goto finalize;
 
+	if (IS_ENABLED(CONFIG_BPF_JIT_ALWAYS_ON) ||
+	    bpf_prog_has_kfunc_call(fp))
+		jit_needed = true;
+
 	bpf_prog_select_func(fp);
 
 	/* eBPF JITs can rewrite the program in case constant
@@ -1858,12 +1867,10 @@ struct bpf_prog *bpf_prog_select_runtime(struct bpf_prog *fp, int *err)
 
 		fp = bpf_int_jit_compile(fp);
 		bpf_prog_jit_attempt_done(fp);
-#ifdef CONFIG_BPF_JIT_ALWAYS_ON
-		if (!fp->jited) {
+		if (!fp->jited && jit_needed) {
 			*err = -ENOTSUPP;
 			return fp;
 		}
-#endif
 	} else {
 		*err = bpf_prog_offload_compile(fp);
 		if (*err)
@@ -2343,6 +2350,11 @@ bool __weak bpf_jit_needs_zext(void)
 	return false;
 }
 
+bool __weak bpf_jit_supports_kfunc_call(void)
+{
+	return false;
+}
+
 /* To execute LD_ABS/LD_IND instructions __bpf_prog_run() may call
  * skb_copy_bits(), so provide a weak definition of it for NET-less config.
  */
diff --git a/kernel/bpf/disasm.c b/kernel/bpf/disasm.c
index 3acc7e0b6916..dad821c8ecd0 100644
--- a/kernel/bpf/disasm.c
+++ b/kernel/bpf/disasm.c
@@ -19,16 +19,23 @@ static const char *__func_get_name(const struct bpf_insn_cbs *cbs,
 {
 	BUILD_BUG_ON(ARRAY_SIZE(func_id_str) != __BPF_FUNC_MAX_ID);
 
-	if (insn->src_reg != BPF_PSEUDO_CALL &&
+	if (!insn->src_reg &&
 	    insn->imm >= 0 && insn->imm < __BPF_FUNC_MAX_ID &&
 	    func_id_str[insn->imm])
 		return func_id_str[insn->imm];
 
-	if (cbs && cbs->cb_call)
-		return cbs->cb_call(cbs->private_data, insn);
+	if (cbs && cbs->cb_call) {
+		const char *res;
+
+		res = cbs->cb_call(cbs->private_data, insn);
+		if (res)
+			return res;
+	}
 
 	if (insn->src_reg == BPF_PSEUDO_CALL)
 		snprintf(buff, len, "%+d", insn->imm);
+	else if (insn->src_reg == BPF_PSEUDO_KFUNC_CALL)
+		snprintf(buff, len, "kernel-function");
 
 	return buff;
 }
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index eaf85bf51c5a..9603de81811a 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1696,6 +1696,7 @@ static void __bpf_prog_put_noref(struct bpf_prog *prog, bool deferred)
 	btf_put(prog->aux->btf);
 	kvfree(prog->aux->jited_linfo);
 	kvfree(prog->aux->linfo);
+	kfree(prog->aux->kfunc_tab);
 	if (prog->aux->attach_btf)
 		btf_put(prog->aux->attach_btf);
 
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index b31e62daafbd..852541a435ef 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -234,6 +234,12 @@ static bool bpf_pseudo_call(const struct bpf_insn *insn)
 	       insn->src_reg == BPF_PSEUDO_CALL;
 }
 
+static bool bpf_pseudo_kfunc_call(const struct bpf_insn *insn)
+{
+	return insn->code == (BPF_JMP | BPF_CALL) &&
+	       insn->src_reg == BPF_PSEUDO_KFUNC_CALL;
+}
+
 static bool bpf_pseudo_func(const struct bpf_insn *insn)
 {
 	return insn->code == (BPF_LD | BPF_IMM | BPF_DW) &&
@@ -1554,47 +1560,205 @@ static int add_subprog(struct bpf_verifier_env *env, int off)
 		verbose(env, "too many subprograms\n");
 		return -E2BIG;
 	}
+	/* determine subprog starts. The end is one before the next starts */
 	env->subprog_info[env->subprog_cnt++].start = off;
 	sort(env->subprog_info, env->subprog_cnt,
 	     sizeof(env->subprog_info[0]), cmp_subprogs, NULL);
 	return env->subprog_cnt - 1;
 }
 
-static int check_subprogs(struct bpf_verifier_env *env)
+struct bpf_kfunc_desc {
+	struct btf_func_model func_model;
+	u32 func_id;
+	s32 imm;
+};
+
+#define MAX_KFUNC_DESCS 256
+struct bpf_kfunc_desc_tab {
+	struct bpf_kfunc_desc descs[MAX_KFUNC_DESCS];
+	u32 nr_descs;
+};
+
+static int kfunc_desc_cmp_by_id(const void *a, const void *b)
+{
+	const struct bpf_kfunc_desc *d0 = a;
+	const struct bpf_kfunc_desc *d1 = b;
+
+	/* func_id is not greater than BTF_MAX_TYPE */
+	return d0->func_id - d1->func_id;
+}
+
+static const struct bpf_kfunc_desc *
+find_kfunc_desc(const struct bpf_prog *prog, u32 func_id)
+{
+	struct bpf_kfunc_desc desc = {
+		.func_id = func_id,
+	};
+	struct bpf_kfunc_desc_tab *tab;
+
+	tab = prog->aux->kfunc_tab;
+	return bsearch(&desc, tab->descs, tab->nr_descs,
+		       sizeof(tab->descs[0]), kfunc_desc_cmp_by_id);
+}
+
+static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id)
+{
+	const struct btf_type *func, *func_proto;
+	struct bpf_kfunc_desc_tab *tab;
+	struct bpf_prog_aux *prog_aux;
+	struct bpf_kfunc_desc *desc;
+	const char *func_name;
+	unsigned long addr;
+	int err;
+
+	prog_aux = env->prog->aux;
+	tab = prog_aux->kfunc_tab;
+	if (!tab) {
+		if (!btf_vmlinux) {
+			verbose(env, "calling kernel function is not supported without CONFIG_DEBUG_INFO_BTF\n");
+			return -ENOTSUPP;
+		}
+
+		if (!env->prog->jit_requested) {
+			verbose(env, "JIT is required for calling kernel function\n");
+			return -ENOTSUPP;
+		}
+
+		if (!bpf_jit_supports_kfunc_call()) {
+			verbose(env, "JIT does not support calling kernel function\n");
+			return -ENOTSUPP;
+		}
+
+		if (!env->prog->gpl_compatible) {
+			verbose(env, "cannot call kernel function from non-GPL compatible program\n");
+			return -EINVAL;
+		}
+
+		tab = kzalloc(sizeof(*tab), GFP_KERNEL);
+		if (!tab)
+			return -ENOMEM;
+		prog_aux->kfunc_tab = tab;
+	}
+
+	if (find_kfunc_desc(env->prog, func_id))
+		return 0;
+
+	if (tab->nr_descs == MAX_KFUNC_DESCS) {
+		verbose(env, "too many different kernel function calls\n");
+		return -E2BIG;
+	}
+
+	func = btf_type_by_id(btf_vmlinux, func_id);
+	if (!func || !btf_type_is_func(func)) {
+		verbose(env, "kernel btf_id %u is not a function\n",
+			func_id);
+		return -EINVAL;
+	}
+	func_proto = btf_type_by_id(btf_vmlinux, func->type);
+	if (!func_proto || !btf_type_is_func_proto(func_proto)) {
+		verbose(env, "kernel function btf_id %u does not have a valid func_proto\n",
+			func_id);
+		return -EINVAL;
+	}
+
+	func_name = btf_name_by_offset(btf_vmlinux, func->name_off);
+	addr = kallsyms_lookup_name(func_name);
+	if (!addr) {
+		verbose(env, "cannot find address for kernel function %s\n",
+			func_name);
+		return -EINVAL;
+	}
+
+	desc = &tab->descs[tab->nr_descs++];
+	desc->func_id = func_id;
+	desc->imm = BPF_CAST_CALL(addr) - __bpf_call_base;
+	err = btf_distill_func_proto(&env->log, btf_vmlinux,
+				     func_proto, func_name,
+				     &desc->func_model);
+	if (!err)
+		sort(tab->descs, tab->nr_descs, sizeof(tab->descs[0]),
+		     kfunc_desc_cmp_by_id, NULL);
+	return err;
+}
+
+static int kfunc_desc_cmp_by_imm(const void *a, const void *b)
+{
+	const struct bpf_kfunc_desc *d0 = a;
+	const struct bpf_kfunc_desc *d1 = b;
+
+	if (d0->imm > d1->imm)
+		return 1;
+	else if (d0->imm < d1->imm)
+		return -1;
+	return 0;
+}
+
+static void sort_kfunc_descs_by_imm(struct bpf_prog *prog)
+{
+	struct bpf_kfunc_desc_tab *tab;
+
+	tab = prog->aux->kfunc_tab;
+	if (!tab)
+		return;
+
+	sort(tab->descs, tab->nr_descs, sizeof(tab->descs[0]),
+	     kfunc_desc_cmp_by_imm, NULL);
+}
+
+bool bpf_prog_has_kfunc_call(const struct bpf_prog *prog)
+{
+	return !!prog->aux->kfunc_tab;
+}
+
+const struct btf_func_model *
+bpf_jit_find_kfunc_model(const struct bpf_prog *prog,
+			 const struct bpf_insn *insn)
+{
+	const struct bpf_kfunc_desc desc = {
+		.imm = insn->imm,
+	};
+	const struct bpf_kfunc_desc *res;
+	struct bpf_kfunc_desc_tab *tab;
+
+	tab = prog->aux->kfunc_tab;
+	res = bsearch(&desc, tab->descs, tab->nr_descs,
+		      sizeof(tab->descs[0]), kfunc_desc_cmp_by_imm);
+
+	return res ? &res->func_model : NULL;
+}
+
+static int add_subprog_and_kfunc(struct bpf_verifier_env *env)
 {
-	int i, ret, subprog_start, subprog_end, off, cur_subprog = 0;
 	struct bpf_subprog_info *subprog = env->subprog_info;
 	struct bpf_insn *insn = env->prog->insnsi;
-	int insn_cnt = env->prog->len;
+	int i, ret, insn_cnt = env->prog->len;
 
 	/* Add entry function. */
 	ret = add_subprog(env, 0);
-	if (ret < 0)
+	if (ret)
 		return ret;
 
-	/* determine subprog starts. The end is one before the next starts */
-	for (i = 0; i < insn_cnt; i++) {
-		if (bpf_pseudo_func(insn + i)) {
-			if (!env->bpf_capable) {
-				verbose(env,
-					"function pointers are allowed for CAP_BPF and CAP_SYS_ADMIN\n");
-				return -EPERM;
-			}
-			ret = add_subprog(env, i + insn[i].imm + 1);
-			if (ret < 0)
-				return ret;
-			/* remember subprog */
-			insn[i + 1].imm = ret;
-			continue;
-		}
-		if (!bpf_pseudo_call(insn + i))
+	for (i = 0; i < insn_cnt; i++, insn++) {
+		if (!bpf_pseudo_func(insn) && !bpf_pseudo_call(insn) &&
+		    !bpf_pseudo_kfunc_call(insn))
 			continue;
+
 		if (!env->bpf_capable) {
-			verbose(env,
-				"function calls to other bpf functions are allowed for CAP_BPF and CAP_SYS_ADMIN\n");
+			verbose(env, "loading/calling other bpf or kernel functions are allowed for CAP_BPF and CAP_SYS_ADMIN\n");
 			return -EPERM;
 		}
-		ret = add_subprog(env, i + insn[i].imm + 1);
+
+		if (bpf_pseudo_func(insn)) {
+			ret = add_subprog(env, i + insn->imm + 1);
+			if (ret >= 0)
+				/* remember subprog */
+				insn[1].imm = ret;
+		} else if (bpf_pseudo_call(insn)) {
+			ret = add_subprog(env, i + insn->imm + 1);
+		} else {
+			ret = add_kfunc_call(env, insn->imm);
+		}
+
 		if (ret < 0)
 			return ret;
 	}
@@ -1608,6 +1772,16 @@ static int check_subprogs(struct bpf_verifier_env *env)
 		for (i = 0; i < env->subprog_cnt; i++)
 			verbose(env, "func#%d @%d\n", i, subprog[i].start);
 
+	return 0;
+}
+
+static int check_subprogs(struct bpf_verifier_env *env)
+{
+	int i, subprog_start, subprog_end, off, cur_subprog = 0;
+	struct bpf_subprog_info *subprog = env->subprog_info;
+	struct bpf_insn *insn = env->prog->insnsi;
+	int insn_cnt = env->prog->len;
+
 	/* now check that all jumps are within the same subprog */
 	subprog_start = subprog[cur_subprog].start;
 	subprog_end = subprog[cur_subprog + 1].start;
@@ -1916,6 +2090,17 @@ static int get_prev_insn_idx(struct bpf_verifier_state *st, int i,
 	return i;
 }
 
+static const char *disasm_kfunc_name(void *data, const struct bpf_insn *insn)
+{
+	const struct btf_type *func;
+
+	if (insn->src_reg != BPF_PSEUDO_KFUNC_CALL)
+		return NULL;
+
+	func = btf_type_by_id(btf_vmlinux, insn->imm);
+	return btf_name_by_offset(btf_vmlinux, func->name_off);
+}
+
 /* For given verifier state backtrack_insn() is called from the last insn to
  * the first insn. Its purpose is to compute a bitmask of registers and
  * stack slots that needs precision in the parent verifier state.
@@ -1924,6 +2109,7 @@ static int backtrack_insn(struct bpf_verifier_env *env, int idx,
 			  u32 *reg_mask, u64 *stack_mask)
 {
 	const struct bpf_insn_cbs cbs = {
+		.cb_call	= disasm_kfunc_name,
 		.cb_print	= verbose,
 		.private_data	= env,
 	};
@@ -5960,6 +6146,98 @@ static int check_helper_call(struct bpf_verifier_env *env, struct bpf_insn *insn
 	return 0;
 }
 
+/* mark_btf_func_reg_size() is used when the reg size is determined by
+ * the BTF func_proto's return value size and argument.
+ */
+static void mark_btf_func_reg_size(struct bpf_verifier_env *env, u32 regno,
+				   size_t reg_size)
+{
+	struct bpf_reg_state *reg = &cur_regs(env)[regno];
+
+	if (regno == BPF_REG_0) {
+		/* Function return value */
+		reg->live |= REG_LIVE_WRITTEN;
+		reg->subreg_def = reg_size == sizeof(u64) ?
+			DEF_NOT_SUBREG : env->insn_idx + 1;
+	} else {
+		/* Function argument */
+		if (reg_size == sizeof(u64)) {
+			mark_insn_zext(env, reg);
+			mark_reg_read(env, reg, reg->parent, REG_LIVE_READ64);
+		} else {
+			mark_reg_read(env, reg, reg->parent, REG_LIVE_READ32);
+		}
+	}
+}
+
+static int check_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn)
+{
+	const struct btf_type *t, *func, *func_proto, *ptr_type;
+	struct bpf_reg_state *regs = cur_regs(env);
+	const char *func_name, *ptr_type_name;
+	u32 i, nargs, func_id, ptr_type_id;
+	const struct btf_param *args;
+	int err;
+
+	func_id = insn->imm;
+	func = btf_type_by_id(btf_vmlinux, func_id);
+	func_name = btf_name_by_offset(btf_vmlinux, func->name_off);
+	func_proto = btf_type_by_id(btf_vmlinux, func->type);
+
+	if (!env->ops->check_kfunc_call ||
+	    !env->ops->check_kfunc_call(func_id)) {
+		verbose(env, "calling kernel function %s is not allowed\n",
+			func_name);
+		return -EACCES;
+	}
+
+	/* Check the arguments */
+	err = btf_check_kfunc_arg_match(env, btf_vmlinux, func_id, regs);
+	if (err)
+		return err;
+
+	for (i = 0; i < CALLER_SAVED_REGS; i++)
+		mark_reg_not_init(env, regs, caller_saved[i]);
+
+	/* Check return type */
+	t = btf_type_skip_modifiers(btf_vmlinux, func_proto->type, NULL);
+	if (btf_type_is_scalar(t)) {
+		mark_reg_unknown(env, regs, BPF_REG_0);
+		mark_btf_func_reg_size(env, BPF_REG_0, t->size);
+	} else if (btf_type_is_ptr(t)) {
+		ptr_type = btf_type_skip_modifiers(btf_vmlinux, t->type,
+						   &ptr_type_id);
+		if (!btf_type_is_struct(ptr_type)) {
+			ptr_type_name = btf_name_by_offset(btf_vmlinux,
+							   ptr_type->name_off);
+			verbose(env, "kernel function %s returns pointer type %s %s is not supported\n",
+				func_name, btf_type_str(ptr_type),
+				ptr_type_name);
+			return -EINVAL;
+		}
+		mark_reg_known_zero(env, regs, BPF_REG_0);
+		regs[BPF_REG_0].btf = btf_vmlinux;
+		regs[BPF_REG_0].type = PTR_TO_BTF_ID;
+		regs[BPF_REG_0].btf_id = ptr_type_id;
+		mark_btf_func_reg_size(env, BPF_REG_0, sizeof(void *));
+	} /* else { add_kfunc_call() ensures it is btf_type_is_void(t) } */
+
+	nargs = btf_type_vlen(func_proto);
+	args = (const struct btf_param *)(func_proto + 1);
+	for (i = 0; i < nargs; i++) {
+		u32 regno = i + 1;
+
+		t = btf_type_skip_modifiers(btf_vmlinux, args[i].type, NULL);
+		if (btf_type_is_ptr(t))
+			mark_btf_func_reg_size(env, regno, sizeof(void *));
+		else
+			/* scalar. ensured by btf_check_kfunc_arg_match() */
+			mark_btf_func_reg_size(env, regno, t->size);
+	}
+
+	return 0;
+}
+
 static bool signed_add_overflows(s64 a, s64 b)
 {
 	/* Do the add in u64, where overflow is well-defined */
@@ -10162,6 +10440,7 @@ static int do_check(struct bpf_verifier_env *env)
 
 		if (env->log.level & BPF_LOG_LEVEL) {
 			const struct bpf_insn_cbs cbs = {
+				.cb_call	= disasm_kfunc_name,
 				.cb_print	= verbose,
 				.private_data	= env,
 			};
@@ -10309,7 +10588,8 @@ static int do_check(struct bpf_verifier_env *env)
 				if (BPF_SRC(insn->code) != BPF_K ||
 				    insn->off != 0 ||
 				    (insn->src_reg != BPF_REG_0 &&
-				     insn->src_reg != BPF_PSEUDO_CALL) ||
+				     insn->src_reg != BPF_PSEUDO_CALL &&
+				     insn->src_reg != BPF_PSEUDO_KFUNC_CALL) ||
 				    insn->dst_reg != BPF_REG_0 ||
 				    class == BPF_JMP32) {
 					verbose(env, "BPF_CALL uses reserved fields\n");
@@ -10324,6 +10604,8 @@ static int do_check(struct bpf_verifier_env *env)
 				}
 				if (insn->src_reg == BPF_PSEUDO_CALL)
 					err = check_func_call(env, insn, &env->insn_idx);
+				else if (insn->src_reg == BPF_PSEUDO_KFUNC_CALL)
+					err = check_kfunc_call(env, insn);
 				else
 					err = check_helper_call(env, insn, &env->insn_idx);
 				if (err)
@@ -11634,6 +11916,7 @@ static int jit_subprogs(struct bpf_verifier_env *env)
 		func[i]->aux->name[0] = 'F';
 		func[i]->aux->stack_depth = env->subprog_info[i].stack_depth;
 		func[i]->jit_requested = 1;
+		func[i]->aux->kfunc_tab = prog->aux->kfunc_tab;
 		func[i]->aux->linfo = prog->aux->linfo;
 		func[i]->aux->nr_linfo = prog->aux->nr_linfo;
 		func[i]->aux->jited_linfo = prog->aux->jited_linfo;
@@ -11773,6 +12056,7 @@ static int fixup_call_args(struct bpf_verifier_env *env)
 #ifndef CONFIG_BPF_JIT_ALWAYS_ON
 	struct bpf_prog *prog = env->prog;
 	struct bpf_insn *insn = prog->insnsi;
+	bool has_kfunc_call = bpf_prog_has_kfunc_call(prog);
 	int i, depth;
 #endif
 	int err = 0;
@@ -11786,6 +12070,10 @@ static int fixup_call_args(struct bpf_verifier_env *env)
 			return err;
 	}
 #ifndef CONFIG_BPF_JIT_ALWAYS_ON
+	if (has_kfunc_call) {
+		verbose(env, "calling kernel functions are not allowed in non-JITed programs\n");
+		return -EINVAL;
+	}
 	if (env->subprog_cnt > 1 && env->prog->aux->tail_call_reachable) {
 		/* When JIT fails the progs with bpf2bpf calls and tail_calls
 		 * have to be rejected, since interpreter doesn't support them yet.
@@ -11814,6 +12102,26 @@ static int fixup_call_args(struct bpf_verifier_env *env)
 	return err;
 }
 
+static int fixup_kfunc_call(struct bpf_verifier_env *env,
+			    struct bpf_insn *insn)
+{
+	const struct bpf_kfunc_desc *desc;
+
+	/* insn->imm has the btf func_id. Replace it with
+	 * an address (relative to __bpf_base_call).
+	 */
+	desc = find_kfunc_desc(env->prog, insn->imm);
+	if (!desc) {
+		verbose(env, "verifier internal error: kernel function descriptor not found for func_id %u\n",
+			insn->imm);
+		return -EFAULT;
+	}
+
+	insn->imm = desc->imm;
+
+	return 0;
+}
+
 /* Do various post-verification rewrites in a single program pass.
  * These rewrites simplify JIT and interpreter implementations.
  */
@@ -11949,6 +12257,12 @@ static int do_misc_fixups(struct bpf_verifier_env *env)
 			continue;
 		if (insn->src_reg == BPF_PSEUDO_CALL)
 			continue;
+		if (insn->src_reg == BPF_PSEUDO_KFUNC_CALL) {
+			ret = fixup_kfunc_call(env, insn);
+			if (ret)
+				return ret;
+			continue;
+		}
 
 		if (insn->imm == BPF_FUNC_get_route_realm)
 			prog->dst_needed = 1;
@@ -12178,6 +12492,8 @@ patch_call_imm:
 		}
 	}
 
+	sort_kfunc_descs_by_imm(env->prog);
+
 	return 0;
 }
 
@@ -12883,6 +13199,10 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr,
 	if (!env->explored_states)
 		goto skip_full_check;
 
+	ret = add_subprog_and_kfunc(env);
+	if (ret < 0)
+		goto skip_full_check;
+
 	ret = check_subprogs(env);
 	if (ret < 0)
 		goto skip_full_check;
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 2d3036e292a9..ab9f2233607c 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -1117,6 +1117,10 @@ enum bpf_link_type {
  * offset to another bpf function
  */
 #define BPF_PSEUDO_CALL		1
+/* when bpf_call->src_reg == BPF_PSEUDO_KFUNC_CALL,
+ * bpf_call->imm == btf_id of a BTF_KIND_FUNC in the running kernel
+ */
+#define BPF_PSEUDO_KFUNC_CALL	2
 
 /* flags for BPF_MAP_UPDATE_ELEM command */
 enum {
diff --git a/tools/testing/selftests/bpf/verifier/calls.c b/tools/testing/selftests/bpf/verifier/calls.c
index eb888c8479c3..336a749673d1 100644
--- a/tools/testing/selftests/bpf/verifier/calls.c
+++ b/tools/testing/selftests/bpf/verifier/calls.c
@@ -19,7 +19,7 @@
 	BPF_MOV64_IMM(BPF_REG_0, 2),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 	.retval = 1,
@@ -136,7 +136,7 @@
 {
 	"calls: wrong src reg",
 	.insns = {
-	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 2, 0, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 3, 0, 0),
 	BPF_MOV64_IMM(BPF_REG_0, 1),
 	BPF_EXIT_INSN(),
 	},
@@ -397,7 +397,7 @@
 	BPF_MOV64_IMM(BPF_REG_0, 1),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.fixup_map_hash_48b = { 3 },
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
@@ -1977,7 +1977,7 @@
 	BPF_EXIT_INSN(),
 	},
 	.prog_type = BPF_PROG_TYPE_SOCKET_FILTER,
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 },
@@ -2003,7 +2003,7 @@
 	BPF_EXIT_INSN(),
 	},
 	.prog_type = BPF_PROG_TYPE_SOCKET_FILTER,
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.errstr = "!read_ok",
 	.result = REJECT,
 },
@@ -2028,7 +2028,7 @@
 	BPF_EXIT_INSN(),
 	},
 	.prog_type = BPF_PROG_TYPE_SOCKET_FILTER,
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.errstr = "!read_ok",
 	.result = REJECT,
 },
diff --git a/tools/testing/selftests/bpf/verifier/dead_code.c b/tools/testing/selftests/bpf/verifier/dead_code.c
index 5cf361d8eb1c..17fe33a75034 100644
--- a/tools/testing/selftests/bpf/verifier/dead_code.c
+++ b/tools/testing/selftests/bpf/verifier/dead_code.c
@@ -85,7 +85,7 @@
 	BPF_MOV64_IMM(BPF_REG_0, 12),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 	.retval = 7,
@@ -103,7 +103,7 @@
 	BPF_MOV64_IMM(BPF_REG_0, 12),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 	.retval = 7,
@@ -121,7 +121,7 @@
 	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 1, 0, -5),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 	.retval = 7,
@@ -137,7 +137,7 @@
 	BPF_MOV64_REG(BPF_REG_0, BPF_REG_1),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 	.retval = 2,
@@ -152,7 +152,7 @@
 	BPF_MOV64_REG(BPF_REG_0, BPF_REG_1),
 	BPF_EXIT_INSN(),
 	},
-	.errstr_unpriv = "function calls to other bpf functions are allowed for",
+	.errstr_unpriv = "loading/calling other bpf or kernel functions are allowed for",
 	.result_unpriv = REJECT,
 	.result = ACCEPT,
 	.retval = 2,