[PATCH v3 bpf-next 0/2] bpf: adding support for mapinmap in libbpf

[PATCH v3 bpf-next 0/2] bpf: adding support for mapinmap in libbpf

[Nikita V. Shirokov]

in this patch series i'm adding a helper for libbpf which would allow
it to load map-in-map(BPF_MAP_TYPE_ARRAY_OF_MAPS and
BPF_MAP_TYPE_HASH_OF_MAPS).
first patch contains new helper + explains proposed workflow
second patch contains tests which also could be used as example of usage

v2->v3:
 - fixing typo in patch description
 - initializing inner_map_fd to -1 by default

v1->v2:
 - addressing nits
 - removing const identifier from fd in new helper
 - starting to check return val for bpf_map_update_elem

Nikita V. Shirokov (2):
  bpf: adding support for map in map in libbpf
  bpf: adding tests for mapinmap helpber in libbpf

 tools/lib/bpf/libbpf.c                      | 11 +++-
 tools/lib/bpf/libbpf.h                      |  2 +
 tools/testing/selftests/bpf/Makefile        |  3 +-
 tools/testing/selftests/bpf/test_mapinmap.c | 49 +++++++++++++++++
 tools/testing/selftests/bpf/test_maps.c     | 82 +++++++++++++++++++++++++++++
 5 files changed, 145 insertions(+), 2 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/test_mapinmap.c

-- 
2.15.1

[PATCH v3 bpf-next 1/2] bpf: adding support for map in map in libbpf

[Nikita V. Shirokov]

idea is pretty simple. for specified map (pointed by struct bpf_map)
we would provide descriptor of already loaded map, which is going to be
used as a prototype for inner map. proposed workflow:
1) open bpf's object (bpf_object__open)
2) create bpf's map which is going to be used as a prototype
3) find (by name) map-in-map which you want to load and update w/
descriptor of inner map w/ a new helper from this patch
4) load bpf program w/ bpf_object__load

inner_map_fd is ignored by any other maps aside from (hash|array) of
maps

Signed-off-by: Nikita V. Shirokov <tehnerd@tehnerd.com>
Acked-by: Yonghong Song <yhs@fb.com>
---
 tools/lib/bpf/libbpf.c | 11 ++++++++++-
 tools/lib/bpf/libbpf.h |  2 ++
 2 files changed, 12 insertions(+), 1 deletion(-)

diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index a01eb9584e52..7e130e0c8fc9 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -163,6 +163,7 @@ struct bpf_map {
 	char *name;
 	size_t offset;
 	int map_ifindex;
+	int inner_map_fd;
 	struct bpf_map_def def;
 	__u32 btf_key_type_id;
 	__u32 btf_value_type_id;
@@ -653,8 +654,10 @@ bpf_object__init_maps(struct bpf_object *obj, int flags)
 	 * fd (fd=0 is stdin) when failure (zclose won't close
 	 * negative fd)).
 	 */
-	for (i = 0; i < nr_maps; i++)
+	for (i = 0; i < nr_maps; i++) {
 		obj->maps[i].fd = -1;
+		obj->maps[i].inner_map_fd = -1;
+	}
 
 	/*
 	 * Fill obj->maps using data in "maps" section.
@@ -1146,6 +1149,7 @@ bpf_object__create_maps(struct bpf_object *obj)
 		create_attr.btf_fd = 0;
 		create_attr.btf_key_type_id = 0;
 		create_attr.btf_value_type_id = 0;
+		create_attr.inner_map_fd = map->inner_map_fd;
 
 		if (obj->btf && !bpf_map_find_btf_info(map, obj->btf)) {
 			create_attr.btf_fd = btf__fd(obj->btf);
@@ -2562,6 +2566,11 @@ void bpf_map__set_ifindex(struct bpf_map *map, __u32 ifindex)
 	map->map_ifindex = ifindex;
 }
 
+void bpf_map__add_inner_map_fd(struct bpf_map *map, int fd)
+{
+	map->inner_map_fd = fd;
+}
+
 static struct bpf_map *
 __bpf_map__iter(struct bpf_map *m, struct bpf_object *obj, int i)
 {
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index b1686a787102..0a0b7e0ed554 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -293,6 +293,8 @@ LIBBPF_API void bpf_map__set_ifindex(struct bpf_map *map, __u32 ifindex);
 LIBBPF_API int bpf_map__pin(struct bpf_map *map, const char *path);
 LIBBPF_API int bpf_map__unpin(struct bpf_map *map, const char *path);
 
+LIBBPF_API void bpf_map__add_inner_map_fd(struct bpf_map *map, int fd);
+
 LIBBPF_API long libbpf_get_error(const void *ptr);
 
 struct bpf_prog_load_attr {
-- 
2.15.1

[PATCH v3 bpf-next 2/2] bpf: adding tests for mapinmap helpber in libbpf

[Nikita V. Shirokov]

adding test/example of bpf_map__add_inner_map_fd usage

Signed-off-by: Nikita V. Shirokov <tehnerd@tehnerd.com>
Acked-by: Yonghong Song <yhs@fb.com>
---
 tools/testing/selftests/bpf/Makefile        |  3 +-
 tools/testing/selftests/bpf/test_mapinmap.c | 49 +++++++++++++++++
 tools/testing/selftests/bpf/test_maps.c     | 82 +++++++++++++++++++++++++++++
 3 files changed, 133 insertions(+), 1 deletion(-)
 create mode 100644 tools/testing/selftests/bpf/test_mapinmap.c

diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
index 57b4712a6276..a3ea69dc9bdf 100644
--- a/tools/testing/selftests/bpf/Makefile
+++ b/tools/testing/selftests/bpf/Makefile
@@ -38,7 +38,8 @@ TEST_GEN_FILES = test_pkt_access.o test_xdp.o test_l4lb.o test_tcp_estats.o test
 	test_lwt_seg6local.o sendmsg4_prog.o sendmsg6_prog.o test_lirc_mode2_kern.o \
 	get_cgroup_id_kern.o socket_cookie_prog.o test_select_reuseport_kern.o \
 	test_skb_cgroup_id_kern.o bpf_flow.o netcnt_prog.o \
-	test_sk_lookup_kern.o test_xdp_vlan.o test_queue_map.o test_stack_map.o
+	test_sk_lookup_kern.o test_xdp_vlan.o test_queue_map.o test_stack_map.o \
+	test_mapinmap.o
 
 # Order correspond to 'make run_tests' order
 TEST_PROGS := test_kmod.sh \
diff --git a/tools/testing/selftests/bpf/test_mapinmap.c b/tools/testing/selftests/bpf/test_mapinmap.c
new file mode 100644
index 000000000000..ce923e67e08e
--- /dev/null
+++ b/tools/testing/selftests/bpf/test_mapinmap.c
@@ -0,0 +1,49 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2018 Facebook */
+#include <stddef.h>
+#include <linux/bpf.h>
+#include <linux/types.h>
+#include "bpf_helpers.h"
+
+struct bpf_map_def SEC("maps") mim_array = {
+	.type = BPF_MAP_TYPE_ARRAY_OF_MAPS,
+	.key_size = sizeof(int),
+	/* must be sizeof(__u32) for map in map */
+	.value_size = sizeof(__u32),
+	.max_entries = 1,
+	.map_flags = 0,
+};
+
+struct bpf_map_def SEC("maps") mim_hash = {
+	.type = BPF_MAP_TYPE_HASH_OF_MAPS,
+	.key_size = sizeof(int),
+	/* must be sizeof(__u32) for map in map */
+	.value_size = sizeof(__u32),
+	.max_entries = 1,
+	.map_flags = 0,
+};
+
+SEC("xdp_mimtest")
+int xdp_mimtest0(struct xdp_md *ctx)
+{
+	int value = 123;
+	int key = 0;
+	void *map;
+
+	map = bpf_map_lookup_elem(&mim_array, &key);
+	if (!map)
+		return XDP_DROP;
+
+	bpf_map_update_elem(map, &key, &value, 0);
+
+	map = bpf_map_lookup_elem(&mim_hash, &key);
+	if (!map)
+		return XDP_DROP;
+
+	bpf_map_update_elem(map, &key, &value, 0);
+
+	return XDP_PASS;
+}
+
+int _version SEC("version") = 1;
+char _license[] SEC("license") = "GPL";
diff --git a/tools/testing/selftests/bpf/test_maps.c b/tools/testing/selftests/bpf/test_maps.c
index 4db2116e52be..6bbcb3ba5787 100644
--- a/tools/testing/selftests/bpf/test_maps.c
+++ b/tools/testing/selftests/bpf/test_maps.c
@@ -1080,6 +1080,86 @@ static void test_sockmap(int tasks, void *data)
 	exit(1);
 }
 
+#define MAPINMAP_PROG "./test_mapinmap.o"
+static void test_mapinmap(void)
+{
+	struct bpf_program *prog;
+	struct bpf_object *obj;
+	struct bpf_map *map;
+	int mim_fd, fd, err;
+	int pos = 0;
+
+	obj = bpf_object__open(MAPINMAP_PROG);
+
+	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(int), sizeof(int),
+			    2, 0);
+	if (fd < 0) {
+		printf("Failed to create hashmap '%s'!\n", strerror(errno));
+		exit(1);
+	}
+
+	map = bpf_object__find_map_by_name(obj, "mim_array");
+	if (IS_ERR(map)) {
+		printf("Failed to load array of maps from test prog\n");
+		goto out_mapinmap;
+	}
+	bpf_map__add_inner_map_fd(map, fd);
+
+	map = bpf_object__find_map_by_name(obj, "mim_hash");
+	if (IS_ERR(map)) {
+		printf("Failed to load hash of maps from test prog\n");
+		goto out_mapinmap;
+	}
+	bpf_map__add_inner_map_fd(map, fd);
+
+	bpf_object__for_each_program(prog, obj) {
+		bpf_program__set_xdp(prog);
+	}
+	bpf_object__load(obj);
+
+	map = bpf_object__find_map_by_name(obj, "mim_array");
+	if (IS_ERR(map)) {
+		printf("Failed to load array of maps from test prog\n");
+		goto out_mapinmap;
+	}
+	mim_fd = bpf_map__fd(map);
+	if (mim_fd < 0) {
+		printf("Failed to get descriptor for array of maps\n");
+		goto out_mapinmap;
+	}
+
+	err = bpf_map_update_elem(mim_fd, &pos, &fd, 0);
+	if (err) {
+		printf("Failed to update array of maps\n");
+		goto out_mapinmap;
+	}
+
+	map = bpf_object__find_map_by_name(obj, "mim_hash");
+	if (IS_ERR(map)) {
+		printf("Failed to load hash of maps from test prog\n");
+		goto out_mapinmap;
+	}
+	mim_fd = bpf_map__fd(map);
+	if (mim_fd < 0) {
+		printf("Failed to get descriptor for hash of maps\n");
+		goto out_mapinmap;
+	}
+
+	err = bpf_map_update_elem(mim_fd, &pos, &fd, 0);
+	if (err) {
+		printf("Failed to update hash of maps\n");
+		goto out_mapinmap;
+	}
+
+	close(fd);
+	bpf_object__close(obj);
+	return;
+
+out_mapinmap:
+	close(fd);
+	exit(1);
+}
+
 #define MAP_SIZE (32 * 1024)
 
 static void test_map_large(void)
@@ -1554,6 +1634,8 @@ static void run_all_tests(void)
 
 	test_queuemap(0, NULL);
 	test_stackmap(0, NULL);
+
+	test_mapinmap();
 }
 
 int main(void)
-- 
2.15.1

Re: [PATCH v3 bpf-next 1/2] bpf: adding support for map in map in libbpf

[Alexei Starovoitov]

On Mon, Nov 19, 2018 at 10:42:21PM -0800, Nikita V. Shirokov wrote:
> idea is pretty simple. for specified map (pointed by struct bpf_map)
> we would provide descriptor of already loaded map, which is going to be
> used as a prototype for inner map. proposed workflow:
> 1) open bpf's object (bpf_object__open)
> 2) create bpf's map which is going to be used as a prototype
> 3) find (by name) map-in-map which you want to load and update w/
> descriptor of inner map w/ a new helper from this patch
> 4) load bpf program w/ bpf_object__load
> 
> inner_map_fd is ignored by any other maps aside from (hash|array) of
> maps
> 
> Signed-off-by: Nikita V. Shirokov <tehnerd@tehnerd.com>
> Acked-by: Yonghong Song <yhs@fb.com>
> ---
>  tools/lib/bpf/libbpf.c | 11 ++++++++++-
>  tools/lib/bpf/libbpf.h |  2 ++
>  2 files changed, 12 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> index a01eb9584e52..7e130e0c8fc9 100644
> --- a/tools/lib/bpf/libbpf.c
> +++ b/tools/lib/bpf/libbpf.c
> @@ -163,6 +163,7 @@ struct bpf_map {
>  	char *name;
>  	size_t offset;
>  	int map_ifindex;
> +	int inner_map_fd;
>  	struct bpf_map_def def;
>  	__u32 btf_key_type_id;
>  	__u32 btf_value_type_id;
> @@ -653,8 +654,10 @@ bpf_object__init_maps(struct bpf_object *obj, int flags)
>  	 * fd (fd=0 is stdin) when failure (zclose won't close
>  	 * negative fd)).
>  	 */
> -	for (i = 0; i < nr_maps; i++)
> +	for (i = 0; i < nr_maps; i++) {
>  		obj->maps[i].fd = -1;
> +		obj->maps[i].inner_map_fd = -1;
> +	}
>  
>  	/*
>  	 * Fill obj->maps using data in "maps" section.
> @@ -1146,6 +1149,7 @@ bpf_object__create_maps(struct bpf_object *obj)
>  		create_attr.btf_fd = 0;
>  		create_attr.btf_key_type_id = 0;
>  		create_attr.btf_value_type_id = 0;
> +		create_attr.inner_map_fd = map->inner_map_fd;
>  
>  		if (obj->btf && !bpf_map_find_btf_info(map, obj->btf)) {
>  			create_attr.btf_fd = btf__fd(obj->btf);
> @@ -2562,6 +2566,11 @@ void bpf_map__set_ifindex(struct bpf_map *map, __u32 ifindex)
>  	map->map_ifindex = ifindex;
>  }
>  
> +void bpf_map__add_inner_map_fd(struct bpf_map *map, int fd)
> +{
> +	map->inner_map_fd = fd;

I think the name bpf_map__set_inner_map_fd() would be more appropriate
and it should check that map->def->type == map-in-map && map->inner_map_fd == -1
before assigning new one.

Also the behavior of bpf_object__create_maps() is not great.
If nothing is set the function will be passing
create_attr.inner_map_fd == -1 to the kernel.
For regular maps that field is sadly ignored by kernel.
Only for map-in-map the value of -1 will be triggering map_create error.
Imo bpf_object__create_maps() should be doing:
if (create_attr.map_type == map-in-map && map->inner_map_fd >= 0)
 create_attr.inner_map_fd = map->inner_map_fd;
// otherwise keep it zero inited