1 files changed, 155 insertions, 16 deletions

diff --git a/kernel/bpf/inode.c b/kernel/bpf/inode.c
index 5e40e7fccc21..4f841e16779e 100644
--- a/kernel/bpf/inode.c
+++ b/kernel/bpf/inode.c
@@ -20,11 +20,13 @@
 #include <linux/filter.h>
 #include <linux/bpf.h>
 #include <linux/bpf_trace.h>
+#include "preload/bpf_preload.h"
 
 enum bpf_type {
 	BPF_TYPE_UNSPEC	= 0,
 	BPF_TYPE_PROG,
 	BPF_TYPE_MAP,
+	BPF_TYPE_LINK,
 };
 
 static void *bpf_any_get(void *raw, enum bpf_type type)
@@ -36,6 +38,9 @@ static void *bpf_any_get(void *raw, enum bpf_type type)
 	case BPF_TYPE_MAP:
 		bpf_map_inc_with_uref(raw);
 		break;
+	case BPF_TYPE_LINK:
+		bpf_link_inc(raw);
+		break;
 	default:
 		WARN_ON_ONCE(1);
 		break;
@@ -53,6 +58,9 @@ static void bpf_any_put(void *raw, enum bpf_type type)
 	case BPF_TYPE_MAP:
 		bpf_map_put_with_uref(raw);
 		break;
+	case BPF_TYPE_LINK:
+		bpf_link_put(raw);
+		break;
 	default:
 		WARN_ON_ONCE(1);
 		break;
@@ -63,20 +71,32 @@ static void *bpf_fd_probe_obj(u32 ufd, enum bpf_type *type)
 {
 	void *raw;
 
-	*type = BPF_TYPE_MAP;
 	raw = bpf_map_get_with_uref(ufd);
-	if (IS_ERR(raw)) {
+	if (!IS_ERR(raw)) {
+		*type = BPF_TYPE_MAP;
+		return raw;
+	}
+
+	raw = bpf_prog_get(ufd);
+	if (!IS_ERR(raw)) {
 		*type = BPF_TYPE_PROG;
-		raw = bpf_prog_get(ufd);
+		return raw;
 	}
 
-	return raw;
+	raw = bpf_link_get_from_fd(ufd);
+	if (!IS_ERR(raw)) {
+		*type = BPF_TYPE_LINK;
+		return raw;
+	}
+
+	return ERR_PTR(-EINVAL);
 }
 
 static const struct inode_operations bpf_dir_iops;
 
 static const struct inode_operations bpf_prog_iops = { };
 static const struct inode_operations bpf_map_iops  = { };
+static const struct inode_operations bpf_link_iops  = { };
 
 static struct inode *bpf_get_inode(struct super_block *sb,
 				   const struct inode *dir,
@@ -102,7 +122,7 @@ static struct inode *bpf_get_inode(struct super_block *sb,
 	inode->i_mtime = inode->i_atime;
 	inode->i_ctime = inode->i_atime;
 
-	inode_init_owner(inode, dir, mode);
+	inode_init_owner(&init_user_ns, inode, dir, mode);
 
 	return inode;
 }
@@ -114,6 +134,8 @@ static int bpf_inode_type(const struct inode *inode, enum bpf_type *type)
 		*type = BPF_TYPE_PROG;
 	else if (inode->i_op == &bpf_map_iops)
 		*type = BPF_TYPE_MAP;
+	else if (inode->i_op == &bpf_link_iops)
+		*type = BPF_TYPE_LINK;
 	else
 		return -EACCES;
 
@@ -130,7 +152,8 @@ static void bpf_dentry_finalize(struct dentry *dentry, struct inode *inode,
 	dir->i_ctime = dir->i_mtime;
 }
 
-static int bpf_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
+static int bpf_mkdir(struct user_namespace *mnt_userns, struct inode *dir,
+		     struct dentry *dentry, umode_t mode)
 {
 	struct inode *inode;
 
@@ -205,10 +228,12 @@ static void *map_seq_next(struct seq_file *m, void *v, loff_t *pos)
 	else
 		prev_key = key;
 
+	rcu_read_lock();
 	if (map->ops->map_get_next_key(map, prev_key, key)) {
 		map_iter(m)->done = true;
-		return NULL;
+		key = NULL;
 	}
+	rcu_read_unlock();
 	return key;
 }
 
@@ -335,20 +360,30 @@ static int bpf_mkmap(struct dentry *dentry, umode_t mode, void *arg)
 			     &bpffs_map_fops : &bpffs_obj_fops);
 }
 
+static int bpf_mklink(struct dentry *dentry, umode_t mode, void *arg)
+{
+	struct bpf_link *link = arg;
+
+	return bpf_mkobj_ops(dentry, mode, arg, &bpf_link_iops,
+			     bpf_link_is_iter(link) ?
+			     &bpf_iter_fops : &bpffs_obj_fops);
+}
+
 static struct dentry *
 bpf_lookup(struct inode *dir, struct dentry *dentry, unsigned flags)
 {
 	/* Dots in names (e.g. "/sys/fs/bpf/foo.bar") are reserved for future
-	 * extensions.
+	 * extensions. That allows popoulate_bpffs() create special files.
 	 */
-	if (strchr(dentry->d_name.name, '.'))
+	if ((dir->i_mode & S_IALLUGO) &&
+	    strchr(dentry->d_name.name, '.'))
 		return ERR_PTR(-EPERM);
 
 	return simple_lookup(dir, dentry, flags);
 }
 
-static int bpf_symlink(struct inode *dir, struct dentry *dentry,
-		       const char *target)
+static int bpf_symlink(struct user_namespace *mnt_userns, struct inode *dir,
+		       struct dentry *dentry, const char *target)
 {
 	char *link = kstrdup(target, GFP_USER | __GFP_NOWARN);
 	struct inode *inode;
@@ -379,6 +414,27 @@ static const struct inode_operations bpf_dir_iops = {
 	.unlink		= simple_unlink,
 };
 
+/* pin iterator link into bpffs */
+static int bpf_iter_link_pin_kernel(struct dentry *parent,
+				    const char *name, struct bpf_link *link)
+{
+	umode_t mode = S_IFREG | S_IRUSR;
+	struct dentry *dentry;
+	int ret;
+
+	inode_lock(parent->d_inode);
+	dentry = lookup_one_len(name, parent, strlen(name));
+	if (IS_ERR(dentry)) {
+		inode_unlock(parent->d_inode);
+		return PTR_ERR(dentry);
+	}
+	ret = bpf_mkobj_ops(dentry, mode, link, &bpf_link_iops,
+			    &bpf_iter_fops);
+	dput(dentry);
+	inode_unlock(parent->d_inode);
+	return ret;
+}
+
 static int bpf_obj_do_pin(const char __user *pathname, void *raw,
 			  enum bpf_type type)
 {
@@ -411,6 +467,9 @@ static int bpf_obj_do_pin(const char __user *pathname, void *raw,
 	case BPF_TYPE_MAP:
 		ret = vfs_mkobj(dentry, mode, bpf_mkmap, raw);
 		break;
+	case BPF_TYPE_LINK:
+		ret = vfs_mkobj(dentry, mode, bpf_mklink, raw);
+		break;
 	default:
 		ret = -EPERM;
 	}
@@ -449,7 +508,7 @@ static void *bpf_obj_do_get(const char __user *pathname,
 		return ERR_PTR(ret);
 
 	inode = d_backing_inode(path.dentry);
-	ret = inode_permission(inode, ACC_MODE(flags));
+	ret = path_permission(&path, ACC_MODE(flags));
 	if (ret)
 		goto out;
 
@@ -487,6 +546,8 @@ int bpf_obj_get_user(const char __user *pathname, int flags)
 		ret = bpf_prog_new_fd(raw);
 	else if (type == BPF_TYPE_MAP)
 		ret = bpf_map_new_fd(raw, f_flags);
+	else if (type == BPF_TYPE_LINK)
+		ret = (f_flags != O_RDWR) ? -EINVAL : bpf_link_new_fd(raw);
 	else
 		return -ENOENT;
 
@@ -498,12 +559,14 @@ int bpf_obj_get_user(const char __user *pathname, int flags)
 static struct bpf_prog *__get_prog_inode(struct inode *inode, enum bpf_prog_type type)
 {
 	struct bpf_prog *prog;
-	int ret = inode_permission(inode, MAY_READ);
+	int ret = inode_permission(&init_user_ns, inode, MAY_READ);
 	if (ret)
 		return ERR_PTR(ret);
 
 	if (inode->i_op == &bpf_map_iops)
 		return ERR_PTR(-EINVAL);
+	if (inode->i_op == &bpf_link_iops)
+		return ERR_PTR(-EINVAL);
 	if (inode->i_op != &bpf_prog_iops)
 		return ERR_PTR(-EACCES);
 
@@ -585,12 +648,22 @@ static int bpf_parse_param(struct fs_context *fc, struct fs_parameter *param)
 	int opt;
 
 	opt = fs_parse(fc, bpf_fs_parameters, param, &result);
-	if (opt < 0)
+	if (opt < 0) {
 		/* We might like to report bad mount options here, but
 		 * traditionally we've ignored all mount options, so we'd
 		 * better continue to ignore non-existing options for bpf.
 		 */
-		return opt == -ENOPARAM ? 0 : opt;
+		if (opt == -ENOPARAM) {
+			opt = vfs_parse_fs_param_source(fc, param);
+			if (opt != -ENOPARAM)
+				return opt;
+
+			return 0;
+		}
+
+		if (opt < 0)
+			return opt;
+	}
 
 	switch (opt) {
 	case OPT_MODE:
@@ -601,6 +674,72 @@ static int bpf_parse_param(struct fs_context *fc, struct fs_parameter *param)
 	return 0;
 }
 
+struct bpf_preload_ops *bpf_preload_ops;
+EXPORT_SYMBOL_GPL(bpf_preload_ops);
+
+static bool bpf_preload_mod_get(void)
+{
+	/* If bpf_preload.ko wasn't loaded earlier then load it now.
+	 * When bpf_preload is built into vmlinux the module's __init
+	 * function will populate it.
+	 */
+	if (!bpf_preload_ops) {
+		request_module("bpf_preload");
+		if (!bpf_preload_ops)
+			return false;
+	}
+	/* And grab the reference, so the module doesn't disappear while the
+	 * kernel is interacting with the kernel module and its UMD.
+	 */
+	if (!try_module_get(bpf_preload_ops->owner)) {
+		pr_err("bpf_preload module get failed.\n");
+		return false;
+	}
+	return true;
+}
+
+static void bpf_preload_mod_put(void)
+{
+	if (bpf_preload_ops)
+		/* now user can "rmmod bpf_preload" if necessary */
+		module_put(bpf_preload_ops->owner);
+}
+
+static DEFINE_MUTEX(bpf_preload_lock);
+
+static int populate_bpffs(struct dentry *parent)
+{
+	struct bpf_preload_info objs[BPF_PRELOAD_LINKS] = {};
+	int err = 0, i;
+
+	/* grab the mutex to make sure the kernel interactions with bpf_preload
+	 * are serialized
+	 */
+	mutex_lock(&bpf_preload_lock);
+
+	/* if bpf_preload.ko wasn't built into vmlinux then load it */
+	if (!bpf_preload_mod_get())
+		goto out;
+
+	err = bpf_preload_ops->preload(objs);
+	if (err)
+		goto out_put;
+	for (i = 0; i < BPF_PRELOAD_LINKS; i++) {
+		bpf_link_inc(objs[i].link);
+		err = bpf_iter_link_pin_kernel(parent,
+					       objs[i].link_name, objs[i].link);
+		if (err) {
+			bpf_link_put(objs[i].link);
+			goto out_put;
+		}
+	}
+out_put:
+	bpf_preload_mod_put();
+out:
+	mutex_unlock(&bpf_preload_lock);
+	return err;
+}
+
 static int bpf_fill_super(struct super_block *sb, struct fs_context *fc)
 {
 	static const struct tree_descr bpf_rfiles[] = { { "" } };
@@ -617,8 +756,8 @@ static int bpf_fill_super(struct super_block *sb, struct fs_context *fc)
 	inode = sb->s_root->d_inode;
 	inode->i_op = &bpf_dir_iops;
 	inode->i_mode &= ~S_IALLUGO;
+	populate_bpffs(sb->s_root);
 	inode->i_mode |= S_ISVTX | opts->mode;
-
 	return 0;
 }