Re: [PATCH -next 5/5] md: protect md_thread with a new disk level spin lock

[Yu Kuai <yukuai3@huawei.com>]

Hi, song!

在 2023/03/11 17:31, Yu Kuai 写道:
> From: Yu Kuai <yukuai3@huawei.com>
> 
> Our test reports a uaf for 'mddev->sync_thread':
> 
> T1                      T2
> md_start_sync
>   md_register_thread
> 			raid1d
> 			 md_check_recovery
> 			  md_reap_sync_thread
> 			   md_unregister_thread
> 			    kfree
> 
>   md_wakeup_thread
>    wake_up
>    ->sync_thread was freed
> 
> Currently, a global spinlock 'pers_lock' is borrowed to protect
> 'mddev->thread', this problem can be fixed likewise, however, there might
> be similar problem for other md_thread, and I really don't like the idea to
> borrow a global lock.
> 
> This patch use a disk level spinlock to protect md_thread in relevant apis.
> 
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
> ---
>   drivers/md/md.c | 23 ++++++++++-------------
>   drivers/md/md.h |  1 +
>   2 files changed, 11 insertions(+), 13 deletions(-)
> 
> diff --git a/drivers/md/md.c b/drivers/md/md.c
> index ab9299187cfe..a952978884a5 100644
> --- a/drivers/md/md.c
> +++ b/drivers/md/md.c
> @@ -663,6 +663,7 @@ void mddev_init(struct mddev *mddev)
>   	atomic_set(&mddev->active, 1);
>   	atomic_set(&mddev->openers, 0);
>   	spin_lock_init(&mddev->lock);
> +	spin_lock_init(&mddev->thread_lock);
>   	atomic_set(&mddev->flush_pending, 0);
>   	init_waitqueue_head(&mddev->sb_wait);
>   	init_waitqueue_head(&mddev->recovery_wait);
> @@ -801,13 +802,8 @@ void mddev_unlock(struct mddev *mddev)
>   	} else
>   		mutex_unlock(&mddev->reconfig_mutex);
>   
> -	/* As we've dropped the mutex we need a spinlock to
> -	 * make sure the thread doesn't disappear
> -	 */
> -	spin_lock(&pers_lock);
>   	md_wakeup_thread(&mddev->thread, mddev);
>   	wake_up(&mddev->sb_wait);
> -	spin_unlock(&pers_lock);
>   }
>   EXPORT_SYMBOL_GPL(mddev_unlock);
>   
> @@ -7895,13 +7891,16 @@ static int md_thread(void *arg)
>   
>   void md_wakeup_thread(struct md_thread **threadp, struct mddev *mddev)
>   {
> -	struct md_thread *thread = *threadp;
> +	struct md_thread *thread;
>   
> +	spin_lock(&mddev->thread_lock);
> +	thread = *threadp;
>   	if (thread) {
>   		pr_debug("md: waking up MD thread %s.\n", thread->tsk->comm);
>   		set_bit(THREAD_WAKEUP, &thread->flags);
>   		wake_up(&thread->wqueue);
>   	}
> +	spin_unlock(&mddev->thread_lock);

I just found that md_wakeup_thread can be called from irq context:

md_safemode_timeout
  md_wakeup_thread

And I need to use irq safe spinlock apis here.

Can you drop this verion from md-next? I'll send a new version after I
verified that there are no new regression, at least for mdadm tests.

Thanks,
Kuai
>   }
>   EXPORT_SYMBOL(md_wakeup_thread);
>   
> @@ -7929,7 +7928,9 @@ int md_register_thread(struct md_thread **threadp,
>   		return err;
>   	}
>   
> +	spin_lock(&mddev->thread_lock);
>   	*threadp = thread;
> +	spin_unlock(&mddev->thread_lock);
>   	return 0;
>   }
>   EXPORT_SYMBOL(md_register_thread);
> @@ -7938,18 +7939,14 @@ void md_unregister_thread(struct md_thread **threadp, struct mddev *mddev)
>   {
>   	struct md_thread *thread;
>   
> -	/*
> -	 * Locking ensures that mddev_unlock does not wake_up a
> -	 * non-existent thread
> -	 */
> -	spin_lock(&pers_lock);
> +	spin_lock(&mddev->thread_lock);
>   	thread = *threadp;
>   	if (!thread) {
> -		spin_unlock(&pers_lock);
> +		spin_unlock(&mddev->thread_lock);
>   		return;
>   	}
>   	*threadp = NULL;
> -	spin_unlock(&pers_lock);
> +	spin_unlock(&mddev->thread_lock);
>   
>   	pr_debug("interrupting MD-thread pid %d\n", task_pid_nr(thread->tsk));
>   	kthread_stop(thread->tsk);
> diff --git a/drivers/md/md.h b/drivers/md/md.h
> index 8f4137ad2dde..ca182d21dd8d 100644
> --- a/drivers/md/md.h
> +++ b/drivers/md/md.h
> @@ -367,6 +367,7 @@ struct mddev {
>   	int				new_chunk_sectors;
>   	int				reshape_backwards;
>   
> +	spinlock_t			thread_lock;
>   	struct md_thread		*thread;	/* management thread */
>   	struct md_thread		*sync_thread;	/* doing resync or reconstruct */
>   
>