Message-ID: <20230804-siegen-moralisieren-dd3dc2595ee2@brauner>
Date:   Fri, 4 Aug 2023 13:14:07 +0200
From:   Christian Brauner <brauner@...nel.org>
To:     Ian Kent <raven@...maw.net>
Cc:     Al Viro <viro@...IV.linux.org.uk>,
        autofs mailing list <autofs@...r.kernel.org>,
        linux-fsdevel <linux-fsdevel@...r.kernel.org>,
        Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Fedor Pchelkin <pchelkin@...ras.ru>,
        Takeshi Misawa <jeliantsurux@...il.com>,
        Alexey Khoroshilov <khoroshilov@...ras.ru>,
        Matthew Wilcox <willy@...radead.org>,
        Andrey Vagin <avagin@...nvz.org>
Subject: Re: [PATCH 1/2] autofs: fix memory leak of waitqueues in
 autofs_catatonic_mode

On Fri, Aug 04, 2023 at 01:33:12PM +0800, Ian Kent wrote:
> From: Fedor Pchelkin <pchelkin@...ras.ru>
> 
> Syzkaller reports a memory leak:
> 
> BUG: memory leak
> unreferenced object 0xffff88810b279e00 (size 96):
>   comm "syz-executor399", pid 3631, jiffies 4294964921 (age 23.870s)
>   hex dump (first 32 bytes):
>     00 00 00 00 00 00 00 00 08 9e 27 0b 81 88 ff ff  ..........'.....
>     08 9e 27 0b 81 88 ff ff 00 00 00 00 00 00 00 00  ..'.............
>   backtrace:
>     [<ffffffff814cfc90>] kmalloc_trace+0x20/0x90 mm/slab_common.c:1046
>     [<ffffffff81bb75ca>] kmalloc include/linux/slab.h:576 [inline]
>     [<ffffffff81bb75ca>] autofs_wait+0x3fa/0x9a0 fs/autofs/waitq.c:378
>     [<ffffffff81bb88a7>] autofs_do_expire_multi+0xa7/0x3e0 fs/autofs/expire.c:593
>     [<ffffffff81bb8c33>] autofs_expire_multi+0x53/0x80 fs/autofs/expire.c:619
>     [<ffffffff81bb6972>] autofs_root_ioctl_unlocked+0x322/0x3b0 fs/autofs/root.c:897
>     [<ffffffff81bb6a95>] autofs_root_ioctl+0x25/0x30 fs/autofs/root.c:910
>     [<ffffffff81602a9c>] vfs_ioctl fs/ioctl.c:51 [inline]
>     [<ffffffff81602a9c>] __do_sys_ioctl fs/ioctl.c:870 [inline]
>     [<ffffffff81602a9c>] __se_sys_ioctl fs/ioctl.c:856 [inline]
>     [<ffffffff81602a9c>] __x64_sys_ioctl+0xfc/0x140 fs/ioctl.c:856
>     [<ffffffff84608225>] do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>     [<ffffffff84608225>] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
>     [<ffffffff84800087>] entry_SYSCALL_64_after_hwframe+0x63/0xcd
> 
> autofs_wait_queue structs should be freed if their wait_ctr becomes zero.
> Otherwise they will be lost.
> 
> In this case an AUTOFS_IOC_EXPIRE_MULTI ioctl is done, then a new
> waitqueue struct is allocated in autofs_wait(), its initial wait_ctr
> equals 2. After that wait_event_killable() is interrupted (it returns
> -ERESTARTSYS), so that 'wq->name.name == NULL' condition may be not
> satisfied. Actually, this condition can be satisfied when
> autofs_wait_release() or autofs_catatonic_mode() is called and, what is
> also important, wait_ctr is decremented in those places. Upon the exit of
> autofs_wait(), wait_ctr is decremented to 1. Then the unmounting process
> begins: kill_sb calls autofs_catatonic_mode(), which should have freed the
> waitqueues, but it only decrements its usage counter to zero which is not
> a correct behaviour.
> 
> edit:imk
> This description is of course not correct. The umount performed as a result
> of an expire is a umount of a mount that has been automounted, it's not the
> autofs mount itself. They happen independently, usually after everything
> mounted within the autofs file system has been expired away. If everything
> hasn't been expired away the automount daemon can still exit leaving mounts
> in place. But expires done in both cases will result in a notification that
> calls autofs_wait_release() with a result status. The problem case is the
> summary execution of of the automount daemon. In this case any waiting
> processes won't be woken up until either they are terminated or the mount
> is umounted.
> end edit: imk
> 
> So in catatonic mode we should free waitqueues which counter becomes zero.
> 
> edit: imk
> Initially I was concerned that the calling of autofs_wait_release() and
> autofs_catatonic_mode() was not mutually exclusive but that can't be the
> case (obviously) because the queue entry (or entries) is removed from the
> list when either of these two functions are called. Consequently the wait
> entry will be freed by only one of these functions or by the woken process
> in autofs_wait() depending on the order of the calls.
> end edit: imk
> 
> Reported-by: syzbot+5e53f70e69ff0c0a1c0c@...kaller.appspotmail.com
> Suggested-by: Takeshi Misawa <jeliantsurux@...il.com>
> Signed-off-by: Fedor Pchelkin <pchelkin@...ras.ru>
> Signed-off-by: Alexey Khoroshilov <khoroshilov@...ras.ru>
> Signed-off-by: Ian Kent <raven@...maw.net>
> Cc: Matthew Wilcox <willy@...radead.org>
> Cc: Andrei Vagin <avagin@...il.com>
> Cc: autofs@...r.kernel.org
> Cc: linux-kernel@...r.kernel.org
> ---
>  fs/autofs/waitq.c |    3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/autofs/waitq.c b/fs/autofs/waitq.c
> index 54c1f8b8b075..efdc76732fae 100644
> --- a/fs/autofs/waitq.c
> +++ b/fs/autofs/waitq.c
> @@ -32,8 +32,9 @@ void autofs_catatonic_mode(struct autofs_sb_info *sbi)
>  		wq->status = -ENOENT; /* Magic is gone - report failure */
>  		kfree(wq->name.name - wq->offset);
>  		wq->name.name = NULL;
> -		wq->wait_ctr--;
>  		wake_up_interruptible(&wq->queue);
> +		if (!--wq->wait_ctr)
> +			kfree(wq);

The only thing that peeked my interest was:

autofs_wait()
-> if (!wq)
   -> wq->wait_ctr = 2;
   -> autofs_notify_daemon()

Let's say autofs_write() fails with -EIO or for whatever reason and so
we end up calling:

      -> autofs_catatonic_mode()

If wait_ctr can be decremented in between so that
autofs_catatonic_mode() frees it and then autofs_wait() would cause a
UAF when it tries to much with wq again. But afaict, this can't happen
because and would also affect autofs_notify_daemon() then.

Powered by blists - more mailing lists