Merge branch 'bpf-array-map-offload-and-tests'

Jakub Kicinski says:

====================
This set brings in the rest of map offload code held up by urgent
fixes and improvements to the BPF arrays.

The first 3 patches take care of array map offload, similarly to
hash maps the attribute validation is split out to a separate map
op, and used for both offloaded and non-offloaded case (allocation
only happens if map is on the host).  Offload support comes down
to allowing this map type through the offload check in the core.
NFP driver also rejects the delete operation in case of array maps.

Subsequent patches add reporting of target device in a very similar
way target device of programs is reported (ifindex+netns dev/ino).
Netdevsim is extended with a trivial map implementation allowing us
to test the offload in test_offload.py.

Last patch adds a small busy wait to NFP map IO, this improves the
response times which is especially useful for map dumps.
====================

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>

3 files changed, 92 insertions, 21 deletions

diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c
index ab94d304a634..b1f66480135b 100644
--- a/kernel/bpf/arraymap.c
+++ b/kernel/bpf/arraymap.c
@@ -49,27 +49,35 @@ static int bpf_array_alloc_percpu(struct bpf_array *array)
 }
 
 /* Called from syscall */
-static struct bpf_map *array_map_alloc(union bpf_attr *attr)
+static int array_map_alloc_check(union bpf_attr *attr)
 {
 	bool percpu = attr->map_type == BPF_MAP_TYPE_PERCPU_ARRAY;
 	int numa_node = bpf_map_attr_numa_node(attr);
-	u32 elem_size, index_mask, max_entries;
-	bool unpriv = !capable(CAP_SYS_ADMIN);
-	struct bpf_array *array;
-	u64 array_size, mask64;
 
 	/* check sanity of attributes */
 	if (attr->max_entries == 0 || attr->key_size != 4 ||
 	    attr->value_size == 0 ||
 	    attr->map_flags & ~ARRAY_CREATE_FLAG_MASK ||
 	    (percpu && numa_node != NUMA_NO_NODE))
-		return ERR_PTR(-EINVAL);
+		return -EINVAL;
 
 	if (attr->value_size > KMALLOC_MAX_SIZE)
 		/* if value_size is bigger, the user space won't be able to
 		 * access the elements.
 		 */
-		return ERR_PTR(-E2BIG);
+		return -E2BIG;
+
+	return 0;
+}
+
+static struct bpf_map *array_map_alloc(union bpf_attr *attr)
+{
+	bool percpu = attr->map_type == BPF_MAP_TYPE_PERCPU_ARRAY;
+	int numa_node = bpf_map_attr_numa_node(attr);
+	u32 elem_size, index_mask, max_entries;
+	bool unpriv = !capable(CAP_SYS_ADMIN);
+	struct bpf_array *array;
+	u64 array_size, mask64;
 
 	elem_size = round_up(attr->value_size, 8);
 
@@ -112,12 +120,7 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr)
 	array->map.unpriv_array = unpriv;
 
 	/* copy mandatory map attributes */
-	array->map.map_type = attr->map_type;
-	array->map.key_size = attr->key_size;
-	array->map.value_size = attr->value_size;
-	array->map.max_entries = attr->max_entries;
-	array->map.map_flags = attr->map_flags;
-	array->map.numa_node = numa_node;
+	bpf_map_init_from_attr(&array->map, attr);
 	array->elem_size = elem_size;
 
 	if (!percpu)
@@ -327,6 +330,7 @@ static void array_map_free(struct bpf_map *map)
 }
 
 const struct bpf_map_ops array_map_ops = {
+	.map_alloc_check = array_map_alloc_check,
 	.map_alloc = array_map_alloc,
 	.map_free = array_map_free,
 	.map_get_next_key = array_map_get_next_key,
@@ -337,6 +341,7 @@ const struct bpf_map_ops array_map_ops = {
 };
 
 const struct bpf_map_ops percpu_array_map_ops = {
+	.map_alloc_check = array_map_alloc_check,
 	.map_alloc = array_map_alloc,
 	.map_free = array_map_free,
 	.map_get_next_key = array_map_get_next_key,
@@ -345,12 +350,12 @@ const struct bpf_map_ops percpu_array_map_ops = {
 	.map_delete_elem = array_map_delete_elem,
 };
 
-static struct bpf_map *fd_array_map_alloc(union bpf_attr *attr)
+static int fd_array_map_alloc_check(union bpf_attr *attr)
 {
 	/* only file descriptors can be stored in this type of map */
 	if (attr->value_size != sizeof(u32))
-		return ERR_PTR(-EINVAL);
-	return array_map_alloc(attr);
+		return -EINVAL;
+	return array_map_alloc_check(attr);
 }
 
 static void fd_array_map_free(struct bpf_map *map)
@@ -474,7 +479,8 @@ void bpf_fd_array_map_clear(struct bpf_map *map)
 }
 
 const struct bpf_map_ops prog_array_map_ops = {
-	.map_alloc = fd_array_map_alloc,
+	.map_alloc_check = fd_array_map_alloc_check,
+	.map_alloc = array_map_alloc,
 	.map_free = fd_array_map_free,
 	.map_get_next_key = array_map_get_next_key,
 	.map_lookup_elem = fd_array_map_lookup_elem,
@@ -561,7 +567,8 @@ static void perf_event_fd_array_release(struct bpf_map *map,
 }
 
 const struct bpf_map_ops perf_event_array_map_ops = {
-	.map_alloc = fd_array_map_alloc,
+	.map_alloc_check = fd_array_map_alloc_check,
+	.map_alloc = array_map_alloc,
 	.map_free = fd_array_map_free,
 	.map_get_next_key = array_map_get_next_key,
 	.map_lookup_elem = fd_array_map_lookup_elem,
@@ -592,7 +599,8 @@ static void cgroup_fd_array_free(struct bpf_map *map)
 }
 
 const struct bpf_map_ops cgroup_array_map_ops = {
-	.map_alloc = fd_array_map_alloc,
+	.map_alloc_check = fd_array_map_alloc_check,
+	.map_alloc = array_map_alloc,
 	.map_free = cgroup_fd_array_free,
 	.map_get_next_key = array_map_get_next_key,
 	.map_lookup_elem = fd_array_map_lookup_elem,
@@ -610,7 +618,7 @@ static struct bpf_map *array_of_map_alloc(union bpf_attr *attr)
 	if (IS_ERR(inner_map_meta))
 		return inner_map_meta;
 
-	map = fd_array_map_alloc(attr);
+	map = array_map_alloc(attr);
 	if (IS_ERR(map)) {
 		bpf_map_meta_free(inner_map_meta);
 		return map;
@@ -673,6 +681,7 @@ static u32 array_of_map_gen_lookup(struct bpf_map *map,
 }
 
 const struct bpf_map_ops array_of_maps_map_ops = {
+	.map_alloc_check = fd_array_map_alloc_check,
 	.map_alloc = array_of_map_alloc,
 	.map_free = array_of_map_free,
 	.map_get_next_key = array_map_get_next_key,
diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c
index 6c0baa1cf8f8..c9401075b58c 100644
--- a/kernel/bpf/offload.c
+++ b/kernel/bpf/offload.c
@@ -299,7 +299,8 @@ struct bpf_map *bpf_map_offload_map_alloc(union bpf_attr *attr)
 
 	if (!capable(CAP_SYS_ADMIN))
 		return ERR_PTR(-EPERM);
-	if (attr->map_type != BPF_MAP_TYPE_HASH)
+	if (attr->map_type != BPF_MAP_TYPE_ARRAY &&
+	    attr->map_type != BPF_MAP_TYPE_HASH)
 		return ERR_PTR(-EINVAL);
 
 	offmap = kzalloc(sizeof(*offmap), GFP_USER);
@@ -412,6 +413,61 @@ int bpf_map_offload_get_next_key(struct bpf_map *map, void *key, void *next_key)
 	return ret;
 }
 
+struct ns_get_path_bpf_map_args {
+	struct bpf_offloaded_map *offmap;
+	struct bpf_map_info *info;
+};
+
+static struct ns_common *bpf_map_offload_info_fill_ns(void *private_data)
+{
+	struct ns_get_path_bpf_map_args *args = private_data;
+	struct ns_common *ns;
+	struct net *net;
+
+	rtnl_lock();
+	down_read(&bpf_devs_lock);
+
+	if (args->offmap->netdev) {
+		args->info->ifindex = args->offmap->netdev->ifindex;
+		net = dev_net(args->offmap->netdev);
+		get_net(net);
+		ns = &net->ns;
+	} else {
+		args->info->ifindex = 0;
+		ns = NULL;
+	}
+
+	up_read(&bpf_devs_lock);
+	rtnl_unlock();
+
+	return ns;
+}
+
+int bpf_map_offload_info_fill(struct bpf_map_info *info, struct bpf_map *map)
+{
+	struct ns_get_path_bpf_map_args args = {
+		.offmap	= map_to_offmap(map),
+		.info	= info,
+	};
+	struct inode *ns_inode;
+	struct path ns_path;
+	void *res;
+
+	res = ns_get_path_cb(&ns_path, bpf_map_offload_info_fill_ns, &args);
+	if (IS_ERR(res)) {
+		if (!info->ifindex)
+			return -ENODEV;
+		return PTR_ERR(res);
+	}
+
+	ns_inode = ns_path.dentry->d_inode;
+	info->netns_dev = new_encode_dev(ns_inode->i_sb->s_dev);
+	info->netns_ino = ns_inode->i_ino;
+	path_put(&ns_path);
+
+	return 0;
+}
+
 bool bpf_offload_dev_match(struct bpf_prog *prog, struct bpf_map *map)
 {
 	struct bpf_offloaded_map *offmap;
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 97a825ffc763..5bdb0cc84ad2 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1801,6 +1801,12 @@ static int bpf_map_get_info_by_fd(struct bpf_map *map,
 	info.map_flags = map->map_flags;
 	memcpy(info.name, map->name, sizeof(map->name));
 
+	if (bpf_map_is_dev_bound(map)) {
+		err = bpf_map_offload_info_fill(&info, map);
+		if (err)
+			return err;
+	}
+
 	if (copy_to_user(uinfo, &info, info_len) ||
 	    put_user(info_len, &uattr->info.info_len))
 		return -EFAULT;