bpf: Enable bpf_iter targets registering ctx argument types

Commit b121b341e5 ("bpf: Add PTR_TO_BTF_ID_OR_NULL
support") adds a field btf_id_or_null_non0_off to
bpf_prog->aux structure to indicate that the
first ctx argument is PTR_TO_BTF_ID reg_type and
all others are PTR_TO_BTF_ID_OR_NULL.
This approach does not really scale if we have
other different reg types in the future, e.g.,
a pointer to a buffer.

This patch enables bpf_iter targets registering ctx argument
reg types which may be different from the default one.
For example, for pointers to structures, the default reg_type
is PTR_TO_BTF_ID for tracing program. The target can register
a particular pointer type as PTR_TO_BTF_ID_OR_NULL which can
be used by the verifier to enforce accesses.

Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20200513180221.2949882-1-yhs@fb.com

include/linux/bpf.h

@@ -643,6 +643,12 @@ struct bpf_jit_poke_descriptor {
 	u16 reason;
 };
 
+/* reg_type info for ctx arguments */
+struct bpf_ctx_arg_aux {
+	u32 offset;
+	enum bpf_reg_type reg_type;
+};
+
 struct bpf_prog_aux {
 	atomic64_t refcnt;
 	u32 used_map_cnt;
@@ -654,12 +660,13 @@ struct bpf_prog_aux {
 	u32 func_cnt; /* used by non-func prog as the number of func progs */
 	u32 func_idx; /* 0 for non-func prog, the index in func array for func prog */
 	u32 attach_btf_id; /* in-kernel BTF type id to attach to */
+	u32 ctx_arg_info_size;
+	const struct bpf_ctx_arg_aux *ctx_arg_info;
 	struct bpf_prog *linked_prog;
 	bool verifier_zext; /* Zero extensions has been inserted by verifier. */
 	bool offload_requested;
 	bool attach_btf_trace; /* true if attaching to BTF-enabled raw tp */
 	bool func_proto_unreliable;
-	bool btf_id_or_null_non0_off;
 	enum bpf_tramp_prog_type trampoline_prog_type;
 	struct bpf_trampoline *trampoline;
 	struct hlist_node tramp_hlist;
@@ -1139,12 +1146,15 @@ int bpf_obj_get_user(const char __user *pathname, int flags);
 typedef int (*bpf_iter_init_seq_priv_t)(void *private_data);
 typedef void (*bpf_iter_fini_seq_priv_t)(void *private_data);
 
+#define BPF_ITER_CTX_ARG_MAX 2
 struct bpf_iter_reg {
 	const char *target;
 	const struct seq_operations *seq_ops;
 	bpf_iter_init_seq_priv_t init_seq_private;
 	bpf_iter_fini_seq_priv_t fini_seq_private;
 	u32 seq_priv_size;
+	u32 ctx_arg_info_size;
+	struct bpf_ctx_arg_aux ctx_arg_info[BPF_ITER_CTX_ARG_MAX];
 };
 
 struct bpf_iter_meta {

include/net/ip6_fib.h

@@ -540,6 +540,13 @@ static inline bool fib6_metric_locked(struct fib6_info *f6i, int metric)
 	return !!(f6i->fib6_metrics->metrics[RTAX_LOCK - 1] & (1 << metric));
 }
 
+#if IS_BUILTIN(CONFIG_IPV6) && defined(CONFIG_BPF_SYSCALL)
+struct bpf_iter__ipv6_route {
+	__bpf_md_ptr(struct bpf_iter_meta *, meta);
+	__bpf_md_ptr(struct fib6_info *, rt);
+};
+#endif
+
 #ifdef CONFIG_IPV6_MULTIPLE_TABLES
 static inline bool fib6_has_custom_rules(const struct net *net)
 {

kernel/bpf/bpf_iter.c

@@ -308,6 +308,11 @@ bool bpf_iter_prog_supported(struct bpf_prog *prog)
 	}
 	mutex_unlock(&targets_mutex);
 
+	if (supported) {
+		prog->aux->ctx_arg_info_size = tinfo->reg_info->ctx_arg_info_size;
+		prog->aux->ctx_arg_info = tinfo->reg_info->ctx_arg_info;
+	}
+
 	return supported;
 }
 

kernel/bpf/btf.c

@@ -3694,7 +3694,7 @@ bool btf_ctx_access(int off, int size, enum bpf_access_type type,
 	struct bpf_verifier_log *log = info->log;
 	const struct btf_param *args;
 	u32 nr_args, arg;
-	int ret;
+	int i, ret;
 
 	if (off % 8) {
 		bpf_log(log, "func '%s' offset %d is not multiple of 8\n",
@@ -3790,10 +3790,15 @@ bool btf_ctx_access(int off, int size, enum bpf_access_type type,
 		return true;
 
 	/* this is a pointer to another type */
-	if (off != 0 && prog->aux->btf_id_or_null_non0_off)
-		info->reg_type = PTR_TO_BTF_ID_OR_NULL;
-	else
-		info->reg_type = PTR_TO_BTF_ID;
+	info->reg_type = PTR_TO_BTF_ID;
+	for (i = 0; i < prog->aux->ctx_arg_info_size; i++) {
+		const struct bpf_ctx_arg_aux *ctx_arg_info = &prog->aux->ctx_arg_info[i];
+
+		if (ctx_arg_info->offset == off) {
+			info->reg_type = ctx_arg_info->reg_type;
+			break;
+		}
+	}
 
 	if (tgt_prog) {
 		ret = btf_translate_to_vmlinux(log, btf, t, tgt_prog->type, arg);

kernel/bpf/map_iter.c

@@ -87,6 +87,11 @@ static const struct bpf_iter_reg bpf_map_reg_info = {
 	.init_seq_private	= NULL,
 	.fini_seq_private	= NULL,
 	.seq_priv_size		= sizeof(struct bpf_iter_seq_map_info),
+	.ctx_arg_info_size	= 1,
+	.ctx_arg_info		= {
+		{ offsetof(struct bpf_iter__bpf_map, map),
+		  PTR_TO_BTF_ID_OR_NULL },
+	},
 };
 
 static int __init bpf_map_iter_init(void)

kernel/bpf/task_iter.c

@@ -312,6 +312,11 @@ static const struct bpf_iter_reg task_reg_info = {
 	.init_seq_private	= init_seq_pidns,
 	.fini_seq_private	= fini_seq_pidns,
 	.seq_priv_size		= sizeof(struct bpf_iter_seq_task_info),
+	.ctx_arg_info_size	= 1,
+	.ctx_arg_info		= {
+		{ offsetof(struct bpf_iter__task, task),
+		  PTR_TO_BTF_ID_OR_NULL },
+	},
 };
 
 static const struct bpf_iter_reg task_file_reg_info = {
@@ -320,6 +325,13 @@ static const struct bpf_iter_reg task_file_reg_info = {
 	.init_seq_private	= init_seq_pidns,
 	.fini_seq_private	= fini_seq_pidns,
 	.seq_priv_size		= sizeof(struct bpf_iter_seq_task_file_info),
+	.ctx_arg_info_size	= 2,
+	.ctx_arg_info		= {
+		{ offsetof(struct bpf_iter__task_file, task),
+		  PTR_TO_BTF_ID_OR_NULL },
+		{ offsetof(struct bpf_iter__task_file, file),
+		  PTR_TO_BTF_ID_OR_NULL },
+	},
 };
 
 static int __init task_iter_init(void)

kernel/bpf/verifier.c

@@ -10652,7 +10652,6 @@ static int check_attach_btf_id(struct bpf_verifier_env *env)
 		prog->aux->attach_func_proto = t;
 		if (!bpf_iter_prog_supported(prog))
 			return -EINVAL;
-		prog->aux->btf_id_or_null_non0_off = true;
 		ret = btf_distill_func_proto(&env->log, btf, t,
 					     tname, &fmodel);
 		return ret;

net/ipv6/ip6_fib.c

@@ -2638,11 +2638,6 @@ static void ipv6_route_native_seq_stop(struct seq_file *seq, void *v)
 }
 
 #if IS_BUILTIN(CONFIG_IPV6) && defined(CONFIG_BPF_SYSCALL)
-struct bpf_iter__ipv6_route {
-	__bpf_md_ptr(struct bpf_iter_meta *, meta);
-	__bpf_md_ptr(struct fib6_info *, rt);
-};
-
 static int ipv6_route_prog_seq_show(struct bpf_prog *prog,
 				    struct bpf_iter_meta *meta,
 				    void *v)

net/ipv6/route.c

@@ -6403,6 +6403,11 @@ static const struct bpf_iter_reg ipv6_route_reg_info = {
 	.init_seq_private	= bpf_iter_init_seq_net,
 	.fini_seq_private	= bpf_iter_fini_seq_net,
 	.seq_priv_size		= sizeof(struct ipv6_route_iter),
+	.ctx_arg_info_size	= 1,
+	.ctx_arg_info		= {
+		{ offsetof(struct bpf_iter__ipv6_route, rt),
+		  PTR_TO_BTF_ID_OR_NULL },
+	},
 };
 
 static int __init bpf_iter_register(void)

net/netlink/af_netlink.c

@@ -2809,6 +2809,11 @@ static const struct bpf_iter_reg netlink_reg_info = {
 	.init_seq_private	= bpf_iter_init_seq_net,
 	.fini_seq_private	= bpf_iter_fini_seq_net,
 	.seq_priv_size		= sizeof(struct nl_seq_iter),
+	.ctx_arg_info_size	= 1,
+	.ctx_arg_info		= {
+		{ offsetof(struct bpf_iter__netlink, sk),
+		  PTR_TO_BTF_ID_OR_NULL },
+	},
 };
 
 static int __init bpf_iter_register(void)