Re: [PATCH net] net: usb: rtl8150: fix rx tasklet use-after-free on disconnect

[Oliver Neukum <oneukum@suse.com>]

[-- Attachment #1: Type: text/plain, Size: 1703 bytes --]

On 10.11.26 05:30, Myeonghun Pak wrote:

Hi,

> Kill the tasklet again after unlink_all_urbs() has waited for all
> completion handlers. tasklet_kill() drains any late tasklet and prevents
> self-rescheduling; disabled endpoints prevent it from leaving an RX
> URB in flight. Keep the first kill to stop an existing reschedule loop
> before unregister_netdev().
> 
> Found by a static analysis tool under development.

Good catch; nice tool.
But the fix is not correct.

> Fixes: eff674a514bd ("USB: rtl8150_disconnect() needs tasklet_kill()")
> Cc: stable@vger.kernel.org
> Co-developed-by: Ijae Kim <ae878000@gmail.com>
> Signed-off-by: Ijae Kim <ae878000@gmail.com>
> Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
> 
> ---
> 
>   drivers/net/usb/rtl8150.c | 2 ++
>   1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/net/usb/rtl8150.c b/drivers/net/usb/rtl8150.c
> index d51e43170e03..98c36c377956 100644
> --- a/drivers/net/usb/rtl8150.c
> +++ b/drivers/net/usb/rtl8150.c
> @@ -971,6 +971,8 @@ static void rtl8150_disconnect(struct usb_interface *intf)
>   		tasklet_kill(&dev->tl);
>   		unregister_netdev(dev->netdev);
>   		unlink_all_urbs(dev);
> +		/* A completion may have scheduled it after the first kill. */
> +		tasklet_kill(&dev->tl);

The tasklet, however, can also submit the URB again, which in turn
can submit the tasklet and so on. There is an unavoidable window
in that sequence, as tasklet_kill() will kill the tasklet, but it may
have run in the window.

Hence you need to also, in addition to your patch, actually prevent the
resubmission of the URBs. The attached fix will do so.

How do we do this? Shall I submit? Do you want a joint submission?

	Regards
		Oliver

[-- Attachment #2: 0001-net-usb-rtl8150-break-cycle-between-tasklet-and-URB.patch --]
[-- Type: text/x-patch, Size: 1907 bytes --]

From c8b88d45caef0c2fdfa0b1a51f77c2db567658bf Mon Sep 17 00:00:00 2001
From: Oliver Neukum <oneukum@suse.com>
Date: Tue, 10 Nov 2026 11:26:34 +0100
Subject: [PATCH] net: usb: rtl8150: break cycle between tasklet and URB

The tasklet can resubmit the URB, which in turn can
resubmit the tasklet. The cycle needs to be broken
the simplest way to do so is to prevent the resubmission
of the URBs by poisoning them.

Signed-off-by: Oliver Neukum <oneukum@suse.com>
---
 drivers/net/usb/rtl8150.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/drivers/net/usb/rtl8150.c b/drivers/net/usb/rtl8150.c
index d51e43170e03..635e65a0de43 100644
--- a/drivers/net/usb/rtl8150.c
+++ b/drivers/net/usb/rtl8150.c
@@ -353,11 +353,18 @@ static void free_all_urbs(rtl8150_t * dev)
 	usb_free_urb(dev->intr_urb);
 }
 
+static void reenable_all_urbs(rtl8150_t * dev)
+{
+	usb_unpoison_urb(dev->rx_urb);
+	usb_unpoison_urb(dev->tx_urb);
+	usb_unpoison_urb(dev->intr_urb);
+}
+
 static void unlink_all_urbs(rtl8150_t * dev)
 {
-	usb_kill_urb(dev->rx_urb);
-	usb_kill_urb(dev->tx_urb);
-	usb_kill_urb(dev->intr_urb);
+	usb_poison_urb(dev->rx_urb);
+	usb_poison_urb(dev->tx_urb);
+	usb_poison_urb(dev->intr_urb);
 }
 
 static inline struct sk_buff *pull_skb(rtl8150_t *dev)
@@ -518,7 +525,7 @@ static void intr_callback(struct urb *urb)
 	res = usb_submit_urb (urb, GFP_ATOMIC);
 	if (res == -ENODEV)
 		netif_device_detach(dev->netdev);
-	else if (res)
+	else if (res != -EPERM)
 		dev_err(&dev->udev->dev,
 			"can't resubmit intr, %s-%s/input0, status %d\n",
 			dev->udev->bus->bus_name, dev->udev->devpath, res);
@@ -751,6 +758,7 @@ static int rtl8150_open(struct net_device *netdev)
 	if (!dev->rx_skb)
 		return -ENOMEM;
 
+	reenable_all_urbs(dev);
 	set_registers(dev, IDR, 6, netdev->dev_addr);
 
 	usb_fill_bulk_urb(dev->rx_urb, dev->udev, usb_rcvbulkpipe(dev->udev, 1),
-- 
2.56.0