Re: [PATCH v2] binder: implement binderfs

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Fri, Dec 14, 2018 at 10:11:23AM +0300, Dan Carpenter wrote:
> On Thu, Dec 13, 2018 at 10:59:11PM +0100, Christian Brauner wrote:
> > +/**
> > + * binderfs_new_inode - allocate inode from super block of a binderfs mount
> > + * @ref_inode: inode from wich the super block will be taken
> > + * @userp:     buffer to copy information about new device for userspace to
> > + * @device:    binder device for which the new inode will be allocated
> > + * @req:       struct binderfs_device as copied from userspace
> > + *
> > + * This function will allocate a new inode from the super block of the
> > + * filesystem mount and attach a dentry to that inode.
> > + * Minor numbers are limited and tracked globally in binderfs_minors.
> > + * The function will stash a struct binder_device for the specific binder
> > + * device in i_private of the inode.
> > + *
> > + * Return: 0 on success, negative errno on failure
> > + */
> > +static int binderfs_new_inode(struct inode *ref_inode,
> > +			      struct binder_device *device,
> > +			      struct binderfs_device __user *userp,
> > +			      struct binderfs_device *req)
> > +{
> > +	int minor, ret;
> > +	struct dentry *dentry, *dup, *root;
> > +	size_t name_len = BINDERFS_MAX_NAME + 1;
> > +	char *name = NULL;
> > +	struct inode *inode = NULL;
> > +	struct super_block *sb = ref_inode->i_sb;
> > +	struct binderfs_info *info = sb->s_fs_info;
> > +
> > +	/* Reserve new minor number for the new device. */
> > +	mutex_lock(&binderfs_minors_mutex);
> > +	minor = ida_alloc_max(&binderfs_minors, BINDERFS_MAX_MINOR, GFP_KERNEL);
> > +	mutex_unlock(&binderfs_minors_mutex);
> > +	if (minor < 0)
> > +		return minor;
> > +
> > +	ret = -ENOMEM;
> > +	inode = new_inode(sb);
> > +	if (!inode)
> > +		goto err;
> > +
> > +	inode->i_ino = minor + INODE_OFFSET;
> > +	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
> > +	init_special_inode(inode, S_IFCHR | 0600,
> > +			   MKDEV(MAJOR(binderfs_dev), minor));
> > +	inode->i_fop = &binder_fops;
> > +	inode->i_uid = info->root_uid;
> > +	inode->i_gid = info->root_gid;
> > +	inode->i_private = device;
> > +
> > +	name = kmalloc(name_len, GFP_KERNEL);
> > +	if (!name)
> > +		goto err;
> > +
> > +	ret = snprintf(name, name_len, "%s", req->name);
> > +	if (ret < 0 || (size_t)ret >= name_len) {
> 
> kernel snprintf() doesn't return negatives and the cast isn't required
> either.

Good point. But I'd rather replace it with:
strscpy(name, req->name, name_len);
which is syntactically and semantically cleaner.

> 
> > +		ret = -EINVAL;
> > +		goto err;
> > +	}
> > +
> > +	device->binderfs_inode = inode;
> > +	device->context.binder_context_mgr_uid = INVALID_UID;
> > +	device->context.name = name;
> > +	device->miscdev.name = name;
> > +	device->miscdev.minor = minor;
> > +	mutex_init(&device->context.context_mgr_node_lock);
> > +
> > +	req->major = MAJOR(binderfs_dev);
> > +	req->minor = minor;
> > +
> > +	ret = copy_to_user(userp, req, sizeof(*req));
> > +	if (ret)
> > +		goto err;
> 
> copy_to_user() returns the number of bytes remaining.
> 
> 	ret = -EFAULT;
> 	if (copy_to_user(userp, req, sizeof(*req)))
> 		goto err;
> 
> Also if this copy_to_user() fails, then does the kfree(name) and the
> iput(inode) lead to a double free of name in binderfs_evict_inode()?

I'm going to defer setting inode->i_private in all codepaths right
before d_add() after which no error handling occurs any more.

> 
> > +
> > +	root = sb->s_root;
> > +	inode_lock(d_inode(root));
> > +	dentry = d_alloc_name(root, name);
> > +	if (!dentry) {
> > +		inode_unlock(d_inode(root));
> > +		ret = -ENOMEM;
> > +		goto err;
> > +	}
> > +
> > +	/* Verify that the name userspace gave us is not already in use. */
> > +	dup = d_lookup(root, &dentry->d_name);
> > +	if (dup) {
> > +		if (d_really_is_positive(dup)) {
> > +			dput(dup);
> > +			dput(dentry);
> > +			inode_unlock(d_inode(root));
> > +			/*
> > +			 * Prevent double free since iput() calls
> > +			 * binderfs_evict_inode().
> > +			 */
> > +			inode->i_private = NULL;
> > +			ret = -EEXIST;
> > +			goto err;
> > +		}
> > +		dput(dup);
> > +	}
> > +
> > +	d_add(dentry, inode);
> > +	fsnotify_create(root->d_inode, dentry);
> > +	inode_unlock(d_inode(root));
> > +
> > +	return 0;
> > +
> > +err:
> > +	kfree(name);
> > +	mutex_lock(&binderfs_minors_mutex);
> > +	ida_free(&binderfs_minors, minor);
> > +	mutex_unlock(&binderfs_minors_mutex);
> > +	iput(inode);
> > +
> > +	return ret;
> > +}
> > +
> > +static int binderfs_binder_device_create(struct inode *inode,
> > +					 struct binderfs_device __user *userp,
> > +					 struct binderfs_device *req)
> > +{
> > +	struct binder_device *device;
> > +	int ret;
> > +
> > +	device = kzalloc(sizeof(*device), GFP_KERNEL);
> > +	if (!device)
> > +		return -ENOMEM;
> 
> Just move this allocation into binderfs_new_inode() and get rid of this
> function.

Agreed.

> 
> > +
> > +	ret = binderfs_new_inode(inode, device, userp, req);
> > +	if (ret < 0) {
> > +		kfree(device);
> > +		return ret;
> > +	}
> > +
> > +	return 0;
> > +}
> > +
> > +/**
> > + * binderfs_ctl_ioctl - handle binder device node allocation requests
> > + *
> > + * The request handler for the binder-control device. All requests operate on
> > + * the binderfs mount the binder-control device resides in:
> > + * - BINDER_CTL_ADD
> > + *   Allocate a new binder device.
> > + *
> > + * Return: 0 on success, negative errno on failure
> > + */
> > +static long binder_ctl_ioctl(struct file *file, unsigned int cmd,
> > +			     unsigned long arg)
> > +{
> > +	struct binderfs_info *info;
> > +	int ret = -EINVAL;
> > +	struct inode *inode = file_inode(file);
> > +	struct binderfs_device __user *device = (struct binderfs_device __user *)arg;
> > +	struct binderfs_device device_req;
> > +
> > +	info = BINDERFS_I(inode);
> 
> info is not used.

Removed.

> 
> > +	switch (cmd) {
> > +	case BINDER_CTL_ADD:
> > +		ret = copy_from_user(&device_req, device, sizeof(device_req));
> 
> if (copy_from_user(&device_req, device, sizeof(device_req))) {
> 	ret = -EFAULT;
> 	break;
> }

Done.

> 
> > +		if (ret)
> > +			break;
> > +
> > +		ret = binderfs_binder_device_create(inode, device, &device_req);
> > +		break;
> > +	default:
> > +		break;
> > +	}
> > +
> > +	return ret;
> > +}
> > +
> > +static void binderfs_evict_inode(struct inode *inode)
> > +{
> > +	struct binder_device *device = inode->i_private;
> > +
> > +	clear_inode(inode);
> > +
> > +	if (!device)
> > +		return;
> > +
> > +	mutex_lock(&binderfs_minors_mutex);
> > +	ida_free(&binderfs_minors, device->miscdev.minor);
> > +	mutex_unlock(&binderfs_minors_mutex);
> > +
> > +	kfree(device->context.name);
> > +	kfree(device);
> > +}
> > +
> > +static const struct super_operations binderfs_super_ops = {
> > +	.statfs = simple_statfs,
> > +	.evict_inode = binderfs_evict_inode,
> > +};
> > +
> > +static int binderfs_rename(struct inode *old_dir, struct dentry *old_dentry,
> > +			   struct inode *new_dir, struct dentry *new_dentry,
> > +			   unsigned int flags)
> > +{
> > +	struct inode *inode = d_inode(old_dentry);
> > +
> > +	/* binderfs doesn't support directories. */
> > +	if (d_is_dir(old_dentry))
> > +		return -EPERM;
> > +
> > +	if (flags & ~RENAME_NOREPLACE)
> > +		return -EINVAL;
> > +
> > +	if (!simple_empty(new_dentry))
> > +		return -ENOTEMPTY;
> > +
> > +	if (d_really_is_positive(new_dentry))
> > +		simple_unlink(new_dir, new_dentry);
> > +
> > +	old_dir->i_ctime = old_dir->i_mtime = new_dir->i_ctime =
> > +		new_dir->i_mtime = inode->i_ctime = current_time(old_dir);
> > +
> > +	return 0;
> > +}
> > +
> > +static int binderfs_unlink(struct inode *dir, struct dentry *dentry)
> > +{
> > +	/*
> > +	 * The control dentry is only ever touched during mount so checking it
> > +	 * here should not require us to take lock.
> > +	 */
> > +	if (BINDERFS_I(dir)->control_dentry == dentry)
> > +		return -EPERM;
> > +
> > +	return simple_unlink(dir, dentry);
> > +}
> > +
> > +static const struct file_operations binder_ctl_fops = {
> > +	.owner		= THIS_MODULE,
> > +	.open		= nonseekable_open,
> > +	.unlocked_ioctl	= binder_ctl_ioctl,
> > +	.compat_ioctl	= binder_ctl_ioctl,
> > +	.llseek		= noop_llseek,
> > +};
> > +
> > +/**
> > + * binderfs_binder_ctl_create - create a new binder-control device
> > + * @sb: super block of the binderfs mount
> > + *
> > + * This function creates a new binder-control device node in the binderfs mount
> > + * referred to by @sb.
> > + *
> > + * Return: 0 on success, negative errno on failure
> > + */
> > +static int binderfs_binder_ctl_create(struct super_block *sb)
> > +{
> > +	int minor;
> > +	struct dentry *dentry;
> > +	struct binder_device *device;
> > +	int ret = 0;
> > +	struct inode *inode = NULL;
> > +	struct dentry *root = sb->s_root;
> > +	struct binderfs_info *info = sb->s_fs_info;
> > +
> > +	device = kzalloc(sizeof(*device), GFP_KERNEL);
> > +	if (!device)
> > +		return -ENOMEM;
> > +
> > +	inode_lock(d_inode(root));
> > +
> > +	if (info->control_dentry)
> > +		goto out;
> 
> I have a private static checker warning for gotos like this where we
> don't set the error code before the goto out.  Apparently here this
> is a success path.  I'm not an fs dev, so I don't know why this is a
> success path.  :(  Maybe you could add a comment?

Added a comment. See also fs/devpts/inode.c for ptmx_dentry that has the
same logic (checked right now).

> 
> 
> > +
> > +	ret = -ENOMEM;
> > +	inode = new_inode(sb);
> > +	if (!inode)
> > +		goto out;
> > +
> > +	/* Reserve a new minor number for the new device. */
> > +	mutex_lock(&binderfs_minors_mutex);
> > +	minor = ida_alloc_max(&binderfs_minors, BINDERFS_MAX_MINOR, GFP_KERNEL);
> > +	mutex_unlock(&binderfs_minors_mutex);
> > +	if (minor < 0) {
> > +		ret = minor;
> > +		goto out;
> > +	}
> > +
> > +	inode->i_ino = SECOND_INODE;
> > +	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
> > +	init_special_inode(inode, S_IFCHR | 0600,
> > +			   MKDEV(MAJOR(binderfs_dev), minor));
> > +	inode->i_fop = &binder_ctl_fops;
> > +	inode->i_uid = info->root_uid;
> > +	inode->i_gid = info->root_gid;
> > +	inode->i_private = device;
> > +
> > +	device->binderfs_inode = inode;
> > +	device->miscdev.minor = minor;
> > +
> > +	dentry = d_alloc_name(root, "binder-control");
> > +	if (!dentry)
> > +		goto out;
> > +
> > +	info->control_dentry = dentry;
> > +	d_add(dentry, inode);
> > +	inode_unlock(d_inode(root));
> > +
> > +	return 0;
> > +
> > +out:
> > +	inode_unlock(d_inode(root));
> > +	kfree(device);
> > +	if (inode) {
> > +		inode->i_private = NULL;
> > +		iput(inode);
> > +	}
> > +
> > +	return ret;
> > +}
> > +
> > +static const struct inode_operations binderfs_dir_inode_operations = {
> > +	.lookup = simple_lookup,
> > +	.rename = binderfs_rename,
> > +	.unlink = binderfs_unlink,
> > +};
> > +
> > +static int binderfs_fill_super(struct super_block *sb, void *data, int silent)
> > +{
> > +	struct binderfs_info *info;
> > +	int ret = -ENOMEM;
> > +	struct inode *inode = NULL;
> > +	struct ipc_namespace *ipc_ns = sb->s_fs_info;
> > +
> > +	get_ipc_ns(ipc_ns);
> > +
> > +	sb->s_blocksize = PAGE_SIZE;
> > +	sb->s_blocksize_bits = PAGE_SHIFT;
> > +
> > +	/*
> > +	 * The binderfs filesystem can be mounted by userns root in a
> > +	 * non-initial userns. By default such mounts have the SB_I_NODEV flag
> > +	 * set in s_iflags to prevent security issues where userns root can
> > +	 * just create random device nodes via mknod() since it owns the
> > +	 * filesystem mount. But binderfs does not allow to create any files
> > +	 * including devices nodes. The only way to create binder devices nodes
> > +	 * is through the binder-control device which userns root is explicitly
> > +	 * allowed to do. So removing the SB_I_NODEV flag from s_iflags is both
> > +	 * necessary and safe.
> > +	 */
> > +	sb->s_iflags &= ~SB_I_NODEV;
> > +	sb->s_iflags |= SB_I_NOEXEC;
> > +	sb->s_magic = BINDERFS_SUPER_MAGIC;
> > +	sb->s_op = &binderfs_super_ops;
> > +	sb->s_time_gran = 1;
> > +
> > +	info = kzalloc(sizeof(struct binderfs_info), GFP_KERNEL);
> > +	if (!info)
> > +		return ret;
> 
> You can't return directly here, you need to goto the put_ipc_ns()
> function in the error handling.

This can directly jump to err_without_dentry since iput() and kfree()
handle NULL pointers.

> 
> > +
> > +	info->ipc_ns = ipc_ns;
> > +	info->root_gid = make_kgid(sb->s_user_ns, 0);
> > +	if (!gid_valid(info->root_gid))
> > +		info->root_gid = GLOBAL_ROOT_GID;
> > +	info->root_uid = make_kuid(sb->s_user_ns, 0);
> > +	if (!uid_valid(info->root_uid))
> > +		info->root_uid = GLOBAL_ROOT_UID;
> > +
> > +	sb->s_fs_info = info;
> > +
> > +	inode = new_inode(sb);
> > +	if (!inode)
> > +		goto err_without_dentry;
> > +
> > +	inode->i_ino = FIRST_INODE;
> > +	inode->i_fop = &simple_dir_operations;
> > +	inode->i_mode = S_IFDIR | 0755;
> > +	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
> > +	inode->i_op = &binderfs_dir_inode_operations;
> > +	set_nlink(inode, 2);
> > +
> > +	sb->s_root = d_make_root(inode);
> > +	if (!sb->s_root)
> > +		goto err_without_dentry;
> > +
> > +	ret = binderfs_binder_ctl_create(sb);
> > +	if (ret)
> > +		goto err_with_dentry;
> > +
> > +	return 0;
> > +
> > +err_with_dentry:
> > +	dput(sb->s_root);
> > +	sb->s_root = NULL;
> > +
> > +err_without_dentry:
> > +	if (inode)
> > +		iput(inode);
> 
> iput accepts NULL pointers.

Yes, all if() checks are removed.

> 
> > +	kfree(info);
> > +	put_ipc_ns(ipc_ns);
> > +
> > +	return ret;
> > +}
> > +
> > +static int binderfs_test_super(struct super_block *sb, void *data)
> > +{
> > +	struct binderfs_info *info = sb->s_fs_info;
> > +
> > +	if (info)
> > +		return info->ipc_ns == data;
> > +
> > +	return 0;
> > +}
> > +
> > +static int binderfs_set_super(struct super_block *sb, void *data)
> > +{
> > +	sb->s_fs_info = data;
> > +	return set_anon_super(sb, NULL);
> > +}
> > +
> > +static struct dentry *binderfs_mount(struct file_system_type *fs_type,
> > +				     int flags, const char *dev_name,
> > +				     void *data)
> > +{
> > +	struct super_block *sb;
> > +	struct ipc_namespace *ipc_ns = current->nsproxy->ipc_ns;
> > +
> > +	if (!ns_capable(ipc_ns->user_ns, CAP_SYS_ADMIN))
> > +		return ERR_PTR(-EPERM);
> > +
> > +	sb = sget_userns(fs_type, binderfs_test_super, binderfs_set_super,
> > +			 flags, ipc_ns->user_ns, ipc_ns);
> > +	if (IS_ERR(sb))
> > +		return ERR_CAST(sb);
> > +
> > +	if (!sb->s_root) {
> > +		int ret = binderfs_fill_super(sb, data, flags & SB_SILENT ? 1 : 0);
> > +		if (ret) {
> > +			deactivate_locked_super(sb);
> > +			return ERR_PTR(ret);
> > +		}
> > +
> > +		sb->s_flags |= SB_ACTIVE;
> > +	}
> > +
> > +	return dget(sb->s_root);
> > +}
> > +
> > +static void binderfs_kill_super(struct super_block *sb)
> > +{
> > +	struct binderfs_info *info = sb->s_fs_info;
> > +
> > +	if (info && info->ipc_ns)
> > +		put_ipc_ns(info->ipc_ns);
> > +
> > +	kfree(info);
> > +	kill_litter_super(sb);
> > +}
> > +
> > +static struct file_system_type binder_fs_type = {
> > +	.name		= "binder",
> > +	.mount		= binderfs_mount,
> > +	.kill_sb	= binderfs_kill_super,
> > +	.fs_flags	= FS_USERNS_MOUNT,
> > +};
> > +
> > +static int __init init_binderfs(void)
> > +{
> > +	int ret;
> > +
> > +	/* Allocate new major number for binderfs. */
> > +	ret = alloc_chrdev_region(&binderfs_dev, 0, BINDERFS_MAX_MINOR,
> > +				  "binder");
> > +	if (ret < 0)
>             ^^^^^^^
> > +		return ret;
> > +
> > +	ret = register_filesystem(&binder_fs_type);
> > +	if (ret) {
>             ^^^
> These functions both return negative on error.  From my static checker
> perspective it's easiest if everyone checks "if (ret) " but either way
> let's make it consistent.

if (ret) it is.

> 
> > +		unregister_chrdev_region(binderfs_dev, BINDERFS_MAX_MINOR);
> > +		return ret;
> > +	}
> > +
> > +	binderfs_mnt = kern_mount(&binder_fs_type);
> > +	if (IS_ERR(binderfs_mnt)) {
> > +		ret = PTR_ERR(binderfs_mnt);
> > +		binderfs_mnt = NULL;
> > +		unregister_filesystem(&binder_fs_type);
> > +		unregister_chrdev_region(binderfs_dev, BINDERFS_MAX_MINOR);
> 
> Use gotos to unwind.

The function is so short that it doesn't really make sense to use labels
and gotos.

Thanks for the review! Much appreciated.

Christian

> 
> > +	}
> > +
> > +	return ret;
> > +}
> 
> regards,
> dan carpenter
> 
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel



[Index of Archives]     [Linux Driver Backports]     [DMA Engine]     [Linux GPIO]     [Linux SPI]     [Video for Linux]     [Linux USB Devel]     [Linux Coverity]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]
  Powered by Linux