Merge branch 'bpf_iter-uapi-fix'

Yonghong Song says:

====================
Andrii raised a concern that current uapi for bpf iterator map
element is a little restrictive and not suitable for future potential
complex customization. This is a valid suggestion, considering people
may indeed add more complex custimization to the iterator, e.g.,
cgroup_id + user_id, etc. for task or task_file. Another example might
be map_id plus additional control so that the bpf iterator may bail
out a bucket earlier if a bucket has too many elements which may hold
lock too long and impact other parts of systems.

Patch freedreno-zz#1 modified uapi with kernel changes. Patch freedreno-zz#2
adjusted libbpf api accordingly.

Changelogs:
  v3 -> v4:
    . add a forward declaration of bpf_iter_link_info in
      tools/lib/bpf/bpf.h in case that libbpf is built against
      not-latest uapi bpf.h.
    . target the patch set to "bpf" instead of "bpf-next"
  v2 -> v3:
    . undo "not reject iter_info.map.map_fd == 0" from v1.
      In the future map_fd may become optional, so let us use map_fd == 0
      indicating the map_fd is not set by user space.
    . add link_info_len to bpf_iter_attach_opts to ensure always correct
      link_info_len from user. Otherwise, libbpf may deduce incorrect
      link_info_len if it uses different uapi header than the user app.
  v1 -> v2:
    . ensure link_create target_fd/flags == 0 since they are not used. (Andrii)
    . if either of iter_info ptr == 0 or iter_info_len == 0, but not both,
      return error to user space. (Andrii)
    . do not reject iter_info.map.map_fd == 0, go ahead to use it trying to
      get a map reference since the map_fd is required for map_elem iterator.
    . use bpf_iter_link_info in bpf_iter_attach_opts instead of map_fd.
      this way, user space is responsible to set up bpf_iter_link_info and
      libbpf just passes the data to the kernel, simplifying libbpf design.
      (Andrii)
====================

Signed-off-by: Alexei Starovoitov <[email protected]>

include/linux/bpf.h

@@ -1214,15 +1214,17 @@ struct bpf_iter_aux_info {
 	struct bpf_map *map;
 };
 
-typedef int (*bpf_iter_check_target_t)(struct bpf_prog *prog,
-				       struct bpf_iter_aux_info *aux);
+typedef int (*bpf_iter_attach_target_t)(struct bpf_prog *prog,
+					union bpf_iter_link_info *linfo,
+					struct bpf_iter_aux_info *aux);
+typedef void (*bpf_iter_detach_target_t)(struct bpf_iter_aux_info *aux);
 
 #define BPF_ITER_CTX_ARG_MAX 2
 struct bpf_iter_reg {
 	const char *target;
-	bpf_iter_check_target_t check_target;
+	bpf_iter_attach_target_t attach_target;
+	bpf_iter_detach_target_t detach_target;
 	u32 ctx_arg_info_size;
-	enum bpf_iter_link_info req_linfo;
 	struct bpf_ctx_arg_aux ctx_arg_info[BPF_ITER_CTX_ARG_MAX];
 	const struct bpf_iter_seq_info *seq_info;
 };

include/uapi/linux/bpf.h

@@ -81,6 +81,12 @@ struct bpf_cgroup_storage_key {
 	__u32	attach_type;		/* program attach type */
 };
 
+union bpf_iter_link_info {
+	struct {
+		__u32	map_fd;
+	} map;
+};
+
 /* BPF syscall commands, see bpf(2) man-page for details. */
 enum bpf_cmd {
 	BPF_MAP_CREATE,
@@ -249,13 +255,6 @@ enum bpf_link_type {
 	MAX_BPF_LINK_TYPE,
 };
 
-enum bpf_iter_link_info {
-	BPF_ITER_LINK_UNSPEC = 0,
-	BPF_ITER_LINK_MAP_FD = 1,
-
-	MAX_BPF_ITER_LINK_INFO,
-};
-
 /* cgroup-bpf attach flags used in BPF_PROG_ATTACH command
  *
  * NONE(default): No further bpf programs allowed in the subtree.
@@ -623,6 +622,8 @@ union bpf_attr {
 		};
 		__u32		attach_type;	/* attach type */
 		__u32		flags;		/* extra flags */
+		__aligned_u64	iter_info;	/* extra bpf_iter_link_info */
+		__u32		iter_info_len;	/* iter_info length */
 	} link_create;
 
 	struct { /* struct used by BPF_LINK_UPDATE command */

kernel/bpf/bpf_iter.c

@@ -338,8 +338,8 @@ static void bpf_iter_link_release(struct bpf_link *link)
 	struct bpf_iter_link *iter_link =
 		container_of(link, struct bpf_iter_link, link);
 
-	if (iter_link->aux.map)
-		bpf_map_put_with_uref(iter_link->aux.map);
+	if (iter_link->tinfo->reg_info->detach_target)
+		iter_link->tinfo->reg_info->detach_target(&iter_link->aux);
 }
 
 static void bpf_iter_link_dealloc(struct bpf_link *link)
@@ -390,15 +390,35 @@ bool bpf_link_is_iter(struct bpf_link *link)
 
 int bpf_iter_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
 {
+	union bpf_iter_link_info __user *ulinfo;
 	struct bpf_link_primer link_primer;
 	struct bpf_iter_target_info *tinfo;
-	struct bpf_iter_aux_info aux = {};
+	union bpf_iter_link_info linfo;
 	struct bpf_iter_link *link;
-	u32 prog_btf_id, target_fd;
+	u32 prog_btf_id, linfo_len;
 	bool existed = false;
-	struct bpf_map *map;
 	int err;
 
+	if (attr->link_create.target_fd || attr->link_create.flags)
+		return -EINVAL;
+
+	memset(&linfo, 0, sizeof(union bpf_iter_link_info));
+
+	ulinfo = u64_to_user_ptr(attr->link_create.iter_info);
+	linfo_len = attr->link_create.iter_info_len;
+	if (!ulinfo ^ !linfo_len)
+		return -EINVAL;
+
+	if (ulinfo) {
+		err = bpf_check_uarg_tail_zero(ulinfo, sizeof(linfo),
+					       linfo_len);
+		if (err)
+			return err;
+		linfo_len = min_t(u32, linfo_len, sizeof(linfo));
+		if (copy_from_user(&linfo, ulinfo, linfo_len))
+			return -EFAULT;
+	}
+
 	prog_btf_id = prog->aux->attach_btf_id;
 	mutex_lock(&targets_mutex);
 	list_for_each_entry(tinfo, &targets, list) {
@@ -411,13 +431,6 @@ int bpf_iter_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
 	if (!existed)
 		return -ENOENT;
 
-	/* Make sure user supplied flags are target expected. */
-	target_fd = attr->link_create.target_fd;
-	if (attr->link_create.flags != tinfo->reg_info->req_linfo)
-		return -EINVAL;
-	if (!attr->link_create.flags && target_fd)
-		return -EINVAL;
-
 	link = kzalloc(sizeof(*link), GFP_USER | __GFP_NOWARN);
 	if (!link)
 		return -ENOMEM;
@@ -431,28 +444,15 @@ int bpf_iter_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
 		return err;
 	}
 
-	if (tinfo->reg_info->req_linfo == BPF_ITER_LINK_MAP_FD) {
-		map = bpf_map_get_with_uref(target_fd);
-		if (IS_ERR(map)) {
-			err = PTR_ERR(map);
-			goto cleanup_link;
-		}
-
-		aux.map = map;
-		err = tinfo->reg_info->check_target(prog, &aux);
+	if (tinfo->reg_info->attach_target) {
+		err = tinfo->reg_info->attach_target(prog, &linfo, &link->aux);
 		if (err) {
-			bpf_map_put_with_uref(map);
-			goto cleanup_link;
+			bpf_link_cleanup(&link_primer);
+			return err;
 		}
-
-		link->aux.map = map;
 	}
 
 	return bpf_link_settle(&link_primer);
-
-cleanup_link:
-	bpf_link_cleanup(&link_primer);
-	return err;
 }
 
 static void init_seq_meta(struct bpf_iter_priv_data *priv_data,

kernel/bpf/map_iter.c

@@ -98,12 +98,21 @@ static struct bpf_iter_reg bpf_map_reg_info = {
 	.seq_info		= &bpf_map_seq_info,
 };
 
-static int bpf_iter_check_map(struct bpf_prog *prog,
-			      struct bpf_iter_aux_info *aux)
+static int bpf_iter_attach_map(struct bpf_prog *prog,
+			       union bpf_iter_link_info *linfo,
+			       struct bpf_iter_aux_info *aux)
 {
 	u32 key_acc_size, value_acc_size, key_size, value_size;
-	struct bpf_map *map = aux->map;
+	struct bpf_map *map;
 	bool is_percpu = false;
+	int err = -EINVAL;
+
+	if (!linfo->map.map_fd)
+		return -EBADF;
+
+	map = bpf_map_get_with_uref(linfo->map.map_fd);
+	if (IS_ERR(map))
+		return PTR_ERR(map);
 
 	if (map->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
 	    map->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH ||
@@ -112,7 +121,7 @@ static int bpf_iter_check_map(struct bpf_prog *prog,
 	else if (map->map_type != BPF_MAP_TYPE_HASH &&
 		 map->map_type != BPF_MAP_TYPE_LRU_HASH &&
 		 map->map_type != BPF_MAP_TYPE_ARRAY)
-		return -EINVAL;
+		goto put_map;
 
 	key_acc_size = prog->aux->max_rdonly_access;
 	value_acc_size = prog->aux->max_rdwr_access;
@@ -122,19 +131,31 @@ static int bpf_iter_check_map(struct bpf_prog *prog,
 	else
 		value_size = round_up(map->value_size, 8) * num_possible_cpus();
 
-	if (key_acc_size > key_size || value_acc_size > value_size)
-		return -EACCES;
+	if (key_acc_size > key_size || value_acc_size > value_size) {
+		err = -EACCES;
+		goto put_map;
+	}
 
+	aux->map = map;
 	return 0;
+
+put_map:
+	bpf_map_put_with_uref(map);
+	return err;
+}
+
+static void bpf_iter_detach_map(struct bpf_iter_aux_info *aux)
+{
+	bpf_map_put_with_uref(aux->map);
 }
 
 DEFINE_BPF_ITER_FUNC(bpf_map_elem, struct bpf_iter_meta *meta,
 		     struct bpf_map *map, void *key, void *value)
 
 static const struct bpf_iter_reg bpf_map_elem_reg_info = {
 	.target			= "bpf_map_elem",
-	.check_target		= bpf_iter_check_map,
-	.req_linfo		= BPF_ITER_LINK_MAP_FD,
+	.attach_target		= bpf_iter_attach_map,
+	.detach_target		= bpf_iter_detach_map,
 	.ctx_arg_info_size	= 2,
 	.ctx_arg_info		= {
 		{ offsetof(struct bpf_iter__bpf_map_elem, key),

kernel/bpf/syscall.c

@@ -3883,7 +3883,7 @@ static int tracing_bpf_link_attach(const union bpf_attr *attr, struct bpf_prog *
 	return -EINVAL;
 }
 
-#define BPF_LINK_CREATE_LAST_FIELD link_create.flags
+#define BPF_LINK_CREATE_LAST_FIELD link_create.iter_info_len
 static int link_create(union bpf_attr *attr)
 {
 	enum bpf_prog_type ptype;

net/core/bpf_sk_storage.c

@@ -1384,18 +1384,39 @@ static int bpf_iter_init_sk_storage_map(void *priv_data,
 	return 0;
 }
 
-static int bpf_iter_check_map(struct bpf_prog *prog,
-			      struct bpf_iter_aux_info *aux)
+static int bpf_iter_attach_map(struct bpf_prog *prog,
+			       union bpf_iter_link_info *linfo,
+			       struct bpf_iter_aux_info *aux)
 {
-	struct bpf_map *map = aux->map;
+	struct bpf_map *map;
+	int err = -EINVAL;
+
+	if (!linfo->map.map_fd)
+		return -EBADF;
+
+	map = bpf_map_get_with_uref(linfo->map.map_fd);
+	if (IS_ERR(map))
+		return PTR_ERR(map);
 
 	if (map->map_type != BPF_MAP_TYPE_SK_STORAGE)
-		return -EINVAL;
+		goto put_map;
 
-	if (prog->aux->max_rdonly_access > map->value_size)
-		return -EACCES;
+	if (prog->aux->max_rdonly_access > map->value_size) {
+		err = -EACCES;
+		goto put_map;
+	}
 
+	aux->map = map;
 	return 0;
+
+put_map:
+	bpf_map_put_with_uref(map);
+	return err;
+}
+
+static void bpf_iter_detach_map(struct bpf_iter_aux_info *aux)
+{
+	bpf_map_put_with_uref(aux->map);
 }
 
 static const struct seq_operations bpf_sk_storage_map_seq_ops = {
@@ -1414,8 +1435,8 @@ static const struct bpf_iter_seq_info iter_seq_info = {
 
 static struct bpf_iter_reg bpf_sk_storage_map_reg_info = {
 	.target			= "bpf_sk_storage_map",
-	.check_target		= bpf_iter_check_map,
-	.req_linfo		= BPF_ITER_LINK_MAP_FD,
+	.attach_target		= bpf_iter_attach_map,
+	.detach_target		= bpf_iter_detach_map,
 	.ctx_arg_info_size	= 2,
 	.ctx_arg_info		= {
 		{ offsetof(struct bpf_iter__bpf_sk_storage_map, sk),

tools/bpf/bpftool/iter.c

@@ -11,6 +11,7 @@
 static int do_pin(int argc, char **argv)
 {
 	DECLARE_LIBBPF_OPTS(bpf_iter_attach_opts, iter_opts);
+	union bpf_iter_link_info linfo;
 	const char *objfile, *path;
 	struct bpf_program *prog;
 	struct bpf_object *obj;
@@ -36,6 +37,11 @@ static int do_pin(int argc, char **argv)
 			map_fd = map_parse_fd(&argc, &argv);
 			if (map_fd < 0)
 				return -1;
+
+			memset(&linfo, 0, sizeof(linfo));
+			linfo.map.map_fd = map_fd;
+			iter_opts.link_info = &linfo;
+			iter_opts.link_info_len = sizeof(linfo);
 		}
 	}
 
@@ -57,9 +63,6 @@ static int do_pin(int argc, char **argv)
 		goto close_obj;
 	}
 
-	if (map_fd >= 0)
-		iter_opts.map_fd = map_fd;
-
 	link = bpf_program__attach_iter(prog, &iter_opts);
 	if (IS_ERR(link)) {
 		err = PTR_ERR(link);

tools/include/uapi/linux/bpf.h

@@ -81,6 +81,12 @@ struct bpf_cgroup_storage_key {
 	__u32	attach_type;		/* program attach type */
 };
 
+union bpf_iter_link_info {
+	struct {
+		__u32	map_fd;
+	} map;
+};
+
 /* BPF syscall commands, see bpf(2) man-page for details. */
 enum bpf_cmd {
 	BPF_MAP_CREATE,
@@ -249,13 +255,6 @@ enum bpf_link_type {
 	MAX_BPF_LINK_TYPE,
 };
 
-enum bpf_iter_link_info {
-	BPF_ITER_LINK_UNSPEC = 0,
-	BPF_ITER_LINK_MAP_FD = 1,
-
-	MAX_BPF_ITER_LINK_INFO,
-};
-
 /* cgroup-bpf attach flags used in BPF_PROG_ATTACH command
  *
  * NONE(default): No further bpf programs allowed in the subtree.
@@ -623,6 +622,8 @@ union bpf_attr {
 		};
 		__u32		attach_type;	/* attach type */
 		__u32		flags;		/* extra flags */
+		__aligned_u64	iter_info;	/* extra bpf_iter_link_info */
+		__u32		iter_info_len;	/* iter_info length */
 	} link_create;
 
 	struct { /* struct used by BPF_LINK_UPDATE command */

tools/lib/bpf/bpf.c

@@ -599,6 +599,9 @@ int bpf_link_create(int prog_fd, int target_fd,
 	attr.link_create.target_fd = target_fd;
 	attr.link_create.attach_type = attach_type;
 	attr.link_create.flags = OPTS_GET(opts, flags, 0);
+	attr.link_create.iter_info =
+		ptr_to_u64(OPTS_GET(opts, iter_info, (void *)0));
+	attr.link_create.iter_info_len = OPTS_GET(opts, iter_info_len, 0);
 
 	return sys_bpf(BPF_LINK_CREATE, &attr, sizeof(attr));
 }

tools/lib/bpf/bpf.h

@@ -168,11 +168,14 @@ LIBBPF_API int bpf_prog_detach(int attachable_fd, enum bpf_attach_type type);
 LIBBPF_API int bpf_prog_detach2(int prog_fd, int attachable_fd,
 				enum bpf_attach_type type);
 
+union bpf_iter_link_info; /* defined in up-to-date linux/bpf.h */
 struct bpf_link_create_opts {
 	size_t sz; /* size of this struct for forward/backward compatibility */
 	__u32 flags;
+	union bpf_iter_link_info *iter_info;
+	__u32 iter_info_len;
 };
-#define bpf_link_create_opts__last_field flags
+#define bpf_link_create_opts__last_field iter_info_len
 
 LIBBPF_API int bpf_link_create(int prog_fd, int target_fd,
 			       enum bpf_attach_type attach_type,

tools/lib/bpf/libbpf.c

@@ -8306,10 +8306,8 @@ bpf_program__attach_iter(struct bpf_program *prog,
 	if (!OPTS_VALID(opts, bpf_iter_attach_opts))
 		return ERR_PTR(-EINVAL);
 
-	if (OPTS_HAS(opts, map_fd)) {
-		target_fd = opts->map_fd;
-		link_create_opts.flags = BPF_ITER_LINK_MAP_FD;
-	}
+	link_create_opts.iter_info = OPTS_GET(opts, link_info, (void *)0);
+	link_create_opts.iter_info_len = OPTS_GET(opts, link_info_len, 0);
 
 	prog_fd = bpf_program__fd(prog);
 	if (prog_fd < 0) {