libbpf: Add safer high-level wrappers for map operations

Add high-level API wrappers for most common and typical BPF map
operations that works directly on instances of struct bpf_map * (so
you don't have to call bpf_map__fd()) and validate key/value size
expectations.

These helpers require users to specify key (and value, where
appropriate) sizes when performing lookup/update/delete/etc. This forces
user to actually think and validate (for themselves) those. This is
a good thing as user is expected by kernel to implicitly provide correct
key/value buffer sizes and kernel will just read/write necessary amount
of data. If it so happens that user doesn't set up buffers correctly
(which bit people for per-CPU maps especially) kernel either randomly
overwrites stack data or return -EFAULT, depending on user's luck and
circumstances. These high-level APIs are meant to prevent such
unpleasant and hard to debug bugs.

This patch also adds bpf_map_delete_elem_flags() low-level API and
requires passing flags to bpf_map__delete_elem() API for consistency
across all similar APIs, even though currently kernel doesn't expect
any extra flags for BPF_MAP_DELETE_ELEM operation.

List of map operations that get these high-level APIs:

  - bpf_map_lookup_elem;
  - bpf_map_update_elem;
  - bpf_map_delete_elem;
  - bpf_map_lookup_and_delete_elem;
  - bpf_map_get_next_key.

Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20220512220713.2617964-1-andrii@kernel.org

src/bpf.c

@@ -639,6 +639,20 @@ int bpf_map_delete_elem(int fd, const void *key)
 	return libbpf_err_errno(ret);
 }
 
+int bpf_map_delete_elem_flags(int fd, const void *key, __u64 flags)
+{
+	union bpf_attr attr;
+	int ret;
+
+	memset(&attr, 0, sizeof(attr));
+	attr.map_fd = fd;
+	attr.key = ptr_to_u64(key);
+	attr.flags = flags;
+
+	ret = sys_bpf(BPF_MAP_DELETE_ELEM, &attr, sizeof(attr));
+	return libbpf_err_errno(ret);
+}
+
 int bpf_map_get_next_key(int fd, const void *key, void *next_key)
 {
 	union bpf_attr attr;

src/bpf.h

@@ -244,6 +244,7 @@ LIBBPF_API int bpf_map_lookup_and_delete_elem(int fd, const void *key,
 LIBBPF_API int bpf_map_lookup_and_delete_elem_flags(int fd, const void *key,
 						    void *value, __u64 flags);
 LIBBPF_API int bpf_map_delete_elem(int fd, const void *key);
+LIBBPF_API int bpf_map_delete_elem_flags(int fd, const void *key, __u64 flags);
 LIBBPF_API int bpf_map_get_next_key(int fd, const void *key, void *next_key);
 LIBBPF_API int bpf_map_freeze(int fd);
 

src/libbpf.c

@@ -9949,6 +9949,110 @@ bpf_object__find_map_by_offset(struct bpf_object *obj, size_t offset)
 	return libbpf_err_ptr(-ENOTSUP);
 }
 
+static int validate_map_op(const struct bpf_map *map, size_t key_sz,
+			   size_t value_sz, bool check_value_sz)
+{
+	if (map->fd <= 0)
+		return -ENOENT;
+
+	if (map->def.key_size != key_sz) {
+		pr_warn("map '%s': unexpected key size %zu provided, expected %u\n",
+			map->name, key_sz, map->def.key_size);
+		return -EINVAL;
+	}
+
+	if (!check_value_sz)
+		return 0;
+
+	switch (map->def.type) {
+	case BPF_MAP_TYPE_PERCPU_ARRAY:
+	case BPF_MAP_TYPE_PERCPU_HASH:
+	case BPF_MAP_TYPE_LRU_PERCPU_HASH:
+	case BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE: {
+		int num_cpu = libbpf_num_possible_cpus();
+		size_t elem_sz = roundup(map->def.value_size, 8);
+
+		if (value_sz != num_cpu * elem_sz) {
+			pr_warn("map '%s': unexpected value size %zu provided for per-CPU map, expected %d * %zu = %zd\n",
+				map->name, value_sz, num_cpu, elem_sz, num_cpu * elem_sz);
+			return -EINVAL;
+		}
+		break;
+	}
+	default:
+		if (map->def.value_size != value_sz) {
+			pr_warn("map '%s': unexpected value size %zu provided, expected %u\n",
+				map->name, value_sz, map->def.value_size);
+			return -EINVAL;
+		}
+		break;
+	}
+	return 0;
+}
+
+int bpf_map__lookup_elem(const struct bpf_map *map,
+			 const void *key, size_t key_sz,
+			 void *value, size_t value_sz, __u64 flags)
+{
+	int err;
+
+	err = validate_map_op(map, key_sz, value_sz, true);
+	if (err)
+		return libbpf_err(err);
+
+	return bpf_map_lookup_elem_flags(map->fd, key, value, flags);
+}
+
+int bpf_map__update_elem(const struct bpf_map *map,
+			 const void *key, size_t key_sz,
+			 const void *value, size_t value_sz, __u64 flags)
+{
+	int err;
+
+	err = validate_map_op(map, key_sz, value_sz, true);
+	if (err)
+		return libbpf_err(err);
+
+	return bpf_map_update_elem(map->fd, key, value, flags);
+}
+
+int bpf_map__delete_elem(const struct bpf_map *map,
+			 const void *key, size_t key_sz, __u64 flags)
+{
+	int err;
+
+	err = validate_map_op(map, key_sz, 0, false /* check_value_sz */);
+	if (err)
+		return libbpf_err(err);
+
+	return bpf_map_delete_elem_flags(map->fd, key, flags);
+}
+
+int bpf_map__lookup_and_delete_elem(const struct bpf_map *map,
+				    const void *key, size_t key_sz,
+				    void *value, size_t value_sz, __u64 flags)
+{
+	int err;
+
+	err = validate_map_op(map, key_sz, value_sz, true);
+	if (err)
+		return libbpf_err(err);
+
+	return bpf_map_lookup_and_delete_elem_flags(map->fd, key, value, flags);
+}
+
+int bpf_map__get_next_key(const struct bpf_map *map,
+			  const void *cur_key, void *next_key, size_t key_sz)
+{
+	int err;
+
+	err = validate_map_op(map, key_sz, 0, false /* check_value_sz */);
+	if (err)
+		return libbpf_err(err);
+
+	return bpf_map_get_next_key(map->fd, cur_key, next_key);
+}
+
 long libbpf_get_error(const void *ptr)
 {
 	if (!IS_ERR_OR_NULL(ptr))

src/libbpf.h

@@ -990,6 +990,110 @@ LIBBPF_API int bpf_map__unpin(struct bpf_map *map, const char *path);
 LIBBPF_API int bpf_map__set_inner_map_fd(struct bpf_map *map, int fd);
 LIBBPF_API struct bpf_map *bpf_map__inner_map(struct bpf_map *map);
 
+/**
+ * @brief **bpf_map__lookup_elem()** allows to lookup BPF map value
+ * corresponding to provided key.
+ * @param map BPF map to lookup element in
+ * @param key pointer to memory containing bytes of the key used for lookup
+ * @param key_sz size in bytes of key data, needs to match BPF map definition's **key_size**
+ * @param value pointer to memory in which looked up value will be stored
+ * @param value_sz size in byte of value data memory; it has to match BPF map
+ * definition's **value_size**. For per-CPU BPF maps value size has to be
+ * a product of BPF map value size and number of possible CPUs in the system
+ * (could be fetched with **libbpf_num_possible_cpus()**). Note also that for
+ * per-CPU values value size has to be aligned up to closest 8 bytes for
+ * alignment reasons, so expected size is: `round_up(value_size, 8)
+ * * libbpf_num_possible_cpus()`.
+ * @flags extra flags passed to kernel for this operation
+ * @return 0, on success; negative error, otherwise
+ *
+ * **bpf_map__lookup_elem()** is high-level equivalent of
+ * **bpf_map_lookup_elem()** API with added check for key and value size.
+ */
+LIBBPF_API int bpf_map__lookup_elem(const struct bpf_map *map,
+				    const void *key, size_t key_sz,
+				    void *value, size_t value_sz, __u64 flags);
+
+/**
+ * @brief **bpf_map__update_elem()** allows to insert or update value in BPF
+ * map that corresponds to provided key.
+ * @param map BPF map to insert to or update element in
+ * @param key pointer to memory containing bytes of the key
+ * @param key_sz size in bytes of key data, needs to match BPF map definition's **key_size**
+ * @param value pointer to memory containing bytes of the value
+ * @param value_sz size in byte of value data memory; it has to match BPF map
+ * definition's **value_size**. For per-CPU BPF maps value size has to be
+ * a product of BPF map value size and number of possible CPUs in the system
+ * (could be fetched with **libbpf_num_possible_cpus()**). Note also that for
+ * per-CPU values value size has to be aligned up to closest 8 bytes for
+ * alignment reasons, so expected size is: `round_up(value_size, 8)
+ * * libbpf_num_possible_cpus()`.
+ * @flags extra flags passed to kernel for this operation
+ * @return 0, on success; negative error, otherwise
+ *
+ * **bpf_map__update_elem()** is high-level equivalent of
+ * **bpf_map_update_elem()** API with added check for key and value size.
+ */
+LIBBPF_API int bpf_map__update_elem(const struct bpf_map *map,
+				    const void *key, size_t key_sz,
+				    const void *value, size_t value_sz, __u64 flags);
+
+/**
+ * @brief **bpf_map__delete_elem()** allows to delete element in BPF map that
+ * corresponds to provided key.
+ * @param map BPF map to delete element from
+ * @param key pointer to memory containing bytes of the key
+ * @param key_sz size in bytes of key data, needs to match BPF map definition's **key_size**
+ * @flags extra flags passed to kernel for this operation
+ * @return 0, on success; negative error, otherwise
+ *
+ * **bpf_map__delete_elem()** is high-level equivalent of
+ * **bpf_map_delete_elem()** API with added check for key size.
+ */
+LIBBPF_API int bpf_map__delete_elem(const struct bpf_map *map,
+				    const void *key, size_t key_sz, __u64 flags);
+
+/**
+ * @brief **bpf_map__lookup_and_delete_elem()** allows to lookup BPF map value
+ * corresponding to provided key and atomically delete it afterwards.
+ * @param map BPF map to lookup element in
+ * @param key pointer to memory containing bytes of the key used for lookup
+ * @param key_sz size in bytes of key data, needs to match BPF map definition's **key_size**
+ * @param value pointer to memory in which looked up value will be stored
+ * @param value_sz size in byte of value data memory; it has to match BPF map
+ * definition's **value_size**. For per-CPU BPF maps value size has to be
+ * a product of BPF map value size and number of possible CPUs in the system
+ * (could be fetched with **libbpf_num_possible_cpus()**). Note also that for
+ * per-CPU values value size has to be aligned up to closest 8 bytes for
+ * alignment reasons, so expected size is: `round_up(value_size, 8)
+ * * libbpf_num_possible_cpus()`.
+ * @flags extra flags passed to kernel for this operation
+ * @return 0, on success; negative error, otherwise
+ *
+ * **bpf_map__lookup_and_delete_elem()** is high-level equivalent of
+ * **bpf_map_lookup_and_delete_elem()** API with added check for key and value size.
+ */
+LIBBPF_API int bpf_map__lookup_and_delete_elem(const struct bpf_map *map,
+					       const void *key, size_t key_sz,
+					       void *value, size_t value_sz, __u64 flags);
+
+/**
+ * @brief **bpf_map__get_next_key()** allows to iterate BPF map keys by
+ * fetching next key that follows current key.
+ * @param map BPF map to fetch next key from
+ * @param cur_key pointer to memory containing bytes of current key or NULL to
+ * fetch the first key
+ * @param next_key pointer to memory to write next key into
+ * @param key_sz size in bytes of key data, needs to match BPF map definition's **key_size**
+ * @return 0, on success; -ENOENT if **cur_key** is the last key in BPF map;
+ * negative error, otherwise
+ *
+ * **bpf_map__get_next_key()** is high-level equivalent of
+ * **bpf_map_get_next_key()** API with added check for key size.
+ */
+LIBBPF_API int bpf_map__get_next_key(const struct bpf_map *map,
+				     const void *cur_key, void *next_key, size_t key_sz);
+
 /**
  * @brief **libbpf_get_error()** extracts the error code from the passed
  * pointer

src/libbpf.map

@@ -443,7 +443,13 @@ LIBBPF_0.7.0 {
 LIBBPF_0.8.0 {
 	global:
 		bpf_map__autocreate;
+		bpf_map__get_next_key;
+		bpf_map__delete_elem;
+		bpf_map__lookup_and_delete_elem;
+		bpf_map__lookup_elem;
 		bpf_map__set_autocreate;
+		bpf_map__update_elem;
+		bpf_map_delete_elem_flags;
 		bpf_object__destroy_subskeleton;
 		bpf_object__open_subskeleton;
 		bpf_program__attach_kprobe_multi_opts;