Re: [PATCH 3/4] fanotify: fix fsnotify_prepare_user_wait() failure

[Amir Goldstein]

On Thu, Oct 19, 2017 at 4:58 PM, Miklos Szeredi <mszeredi@xxxxxxxxxx> wrote:
> If fsnotify_prepare_user_wait() fails, we leave the event on the
> notification list.  Which will result in a warning in
> fsnotify_destroy_event() and later use-after-free.
>
> Instead of adding a new helper to remove the event from the list in this
> case, I opted to move the prepare/finish up into fanotify_handle_event().
>
> This will allow these to be moved further out into the generic code later,
> and perhaps let us move to non-sleeping RCU.

Interesting.
Because all marks are ordered by group priority and there are no permission
events in priority 0, as soon as we see group prio 0 on both inode and vfs mark,
we can maybe change the mark iteration?

>
> Signed-off-by: Miklos Szeredi <mszeredi@xxxxxxxxxx>
> Fixes: 05f0e38724e8 ("fanotify: Release SRCU lock when waiting for userspace response")
> Cc: <stable@xxxxxxxxxxxxxxx> # v4.12
> ---
>  fs/notify/fanotify/fanotify.c | 33 ++++++++++++++++++++-------------
>  1 file changed, 20 insertions(+), 13 deletions(-)
>
> diff --git a/fs/notify/fanotify/fanotify.c b/fs/notify/fanotify/fanotify.c
> index 2fa99aeaa095..fb7a1339982c 100644
> --- a/fs/notify/fanotify/fanotify.c
> +++ b/fs/notify/fanotify/fanotify.c
> @@ -64,19 +64,8 @@ static int fanotify_get_response(struct fsnotify_group *group,
>
>         pr_debug("%s: group=%p event=%p\n", __func__, group, event);
>
> -       /*
> -        * fsnotify_prepare_user_wait() fails if we race with mark deletion.
> -        * Just let the operation pass in that case.
> -        */
> -       if (!fsnotify_prepare_user_wait(iter_info)) {
> -               event->response = FAN_ALLOW;
> -               goto out;
> -       }
> -
>         wait_event(group->fanotify_data.access_waitq, event->response);
>
> -       fsnotify_finish_user_wait(iter_info);
> -out:
>         /* userspace responded, convert to something usable */
>         switch (event->response) {
>         case FAN_ALLOW:
> @@ -211,9 +200,21 @@ static int fanotify_handle_event(struct fsnotify_group *group,
>         pr_debug("%s: group=%p inode=%p mask=%x\n", __func__, group, inode,
>                  mask);
>
> +#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
> +       if (mask & FAN_ALL_PERM_EVENTS) {
> +               /*
> +                * fsnotify_prepare_user_wait() fails if we race with mark deletion.
> +                * Just let the operation pass in that case.
> +                */
> +               if (!fsnotify_prepare_user_wait(iter_info))
> +                       return 0;
> +       }
> +#endif

Well you are not the one to introduce ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
to this code, but I have to wonder, what are all those ifdefs doing in
this code?
What are they trying to save?
FAN_ALL_PERM_EVENTS is masked out from valid event flags in fanotify_mark()
I don't really see a reason for any other ifdef in non headers here.

> +
>         event = fanotify_alloc_event(inode, mask, data);
> +       ret = -ENOMEM;
>         if (unlikely(!event))
> -               return -ENOMEM;
> +               goto finish;
>
>         fsn_event = &event->fse;
>         ret = fsnotify_add_event(group, fsn_event, fanotify_merge);
> @@ -223,7 +224,8 @@ static int fanotify_handle_event(struct fsnotify_group *group,
>                 /* Our event wasn't used in the end. Free it. */
>                 fsnotify_destroy_event(group, fsn_event);
>
> -               return 0;
> +               ret = 0;
> +               goto finish;
>         }
>
>  #ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
> @@ -232,6 +234,11 @@ static int fanotify_handle_event(struct fsnotify_group *group,
>                                             iter_info);
>                 fsnotify_destroy_event(group, fsn_event);
>         }
> +finish:
> +       if (mask & FAN_ALL_PERM_EVENTS)
> +               fsnotify_finish_user_wait(iter_info);
> +#else
> +finish:
>  #endif
>         return ret;
>  }
> --
> 2.5.5
>