pidfs: register pid in pidfs

Add simple helpers that allow a struct pid to be pinned via a pidfs
dentry/inode. If no pidfs dentry exists a new one will be allocated for
it. A reference is taken by pidfs on @pid. The reference must be
released via pidfs_put_pid().

This will allow AF_UNIX sockets to allocate a dentry for the peer
credentials pid at the time they are recorded where we know the task is
still alive. When the task gets reaped its exit status is guaranteed to
be recorded and a pidfd can be handed out for the reaped task.

Link: https://lore.kernel.org/[email protected]
Reviewed-by: Oleg Nesterov <[email protected]>
Reviewed-by: David Rheinsberg <[email protected]>
Signed-off-by: Christian Brauner <[email protected]>

Author: brauner

fs/pidfs.c

@@ -896,6 +896,65 @@ struct file *pidfs_alloc_file(struct pid *pid, unsigned int flags)
 	return pidfd_file;
 }
 
+/**
+ * pidfs_register_pid - register a struct pid in pidfs
+ * @pid: pid to pin
+ *
+ * Register a struct pid in pidfs. Needs to be paired with
+ * pidfs_put_pid() to not risk leaking the pidfs dentry and inode.
+ *
+ * Return: On success zero, on error a negative error code is returned.
+ */
+int pidfs_register_pid(struct pid *pid)
+{
+	struct path path __free(path_put) = {};
+	int ret;
+
+	might_sleep();
+
+	if (!pid)
+		return 0;
+
+	ret = path_from_stashed(&pid->stashed, pidfs_mnt, get_pid(pid), &path);
+	if (unlikely(ret))
+		return ret;
+	/* Keep the dentry and only put the reference to the mount. */
+	path.dentry = NULL;
+	return 0;
+}
+
+/**
+ * pidfs_get_pid - pin a struct pid through pidfs
+ * @pid: pid to pin
+ *
+ * Similar to pidfs_register_pid() but only valid if the caller knows
+ * there's a reference to the @pid through a dentry already that can't
+ * go away.
+ */
+void pidfs_get_pid(struct pid *pid)
+{
+	if (!pid)
+		return;
+	WARN_ON_ONCE(!stashed_dentry_get(&pid->stashed));
+}
+
+/**
+ * pidfs_put_pid - drop a pidfs reference
+ * @pid: pid to drop
+ *
+ * Drop a reference to @pid via pidfs. This is only safe if the
+ * reference has been taken via pidfs_get_pid().
+ */
+void pidfs_put_pid(struct pid *pid)
+{
+	might_sleep();
+
+	if (!pid)
+		return;
+	VFS_WARN_ON_ONCE(!pid->stashed);
+	dput(pid->stashed);
+}
+
 static void pidfs_inode_init_once(void *data)
 {
 	struct pidfs_inode *pi = data;

include/linux/pidfs.h

@@ -8,5 +8,8 @@ void pidfs_add_pid(struct pid *pid);
 void pidfs_remove_pid(struct pid *pid);
 void pidfs_exit(struct task_struct *tsk);
 extern const struct dentry_operations pidfs_dentry_operations;
+int pidfs_register_pid(struct pid *pid);
+void pidfs_get_pid(struct pid *pid);
+void pidfs_put_pid(struct pid *pid);
 
 #endif /* _LINUX_PID_FS_H */