In ns83820_init_one, dev->tq_refill was bound with queue_refill.
If irq happens, it will call ns83820_irq->ns83820_do_isr.
Then it invokes tasklet_schedule(&dev->rx_tasklet) to start
rx_action function. And rx_action will call ns83820_rx_kick
and finally start queue_refill function.
If we remove the driver without finishing the work, there
may be a race condition between ndev, which may cause UAF
bug.
CPU0 CPU1
|queue_refill
ns83820_remove_one |
free_netdev |
put_device |
free ndev |
|rx_refill
|//use ndev
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Zheng Wang <[email protected]>
---
v3:
- add tasklet_kill to stop more task scheduling suggested by
Horatiu Vultur
v2:
- cancel the work after unregister_netdev to make sure there
is no more request suggested by Jakub Kicinski
---
drivers/net/ethernet/natsemi/ns83820.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/drivers/net/ethernet/natsemi/ns83820.c b/drivers/net/ethernet/natsemi/ns83820.c
index 998586872599..af597719795d 100644
--- a/drivers/net/ethernet/natsemi/ns83820.c
+++ b/drivers/net/ethernet/natsemi/ns83820.c
@@ -2208,8 +2208,14 @@ static void ns83820_remove_one(struct pci_dev *pci_dev)
ns83820_disable_interrupts(dev); /* paranoia */
+ netif_carrier_off(ndev);
+ netif_tx_disable(ndev);
+
unregister_netdev(ndev);
free_irq(dev->pci_dev->irq, ndev);
+ tasklet_kill(&dev->rx_tasklet);
+ cancel_work_sync(&dev->tq_refill);
+
iounmap(dev->base);
dma_free_coherent(&dev->pci_dev->dev, 4 * DESC_SIZE * NR_TX_DESC,
dev->tx_descs, dev->tx_phy_descs);
--
2.25.1
The 04/17/2023 09:31, Zheng Wang wrote:
>
> In ns83820_init_one, dev->tq_refill was bound with queue_refill.
>
> If irq happens, it will call ns83820_irq->ns83820_do_isr.
> Then it invokes tasklet_schedule(&dev->rx_tasklet) to start
> rx_action function. And rx_action will call ns83820_rx_kick
> and finally start queue_refill function.
>
> If we remove the driver without finishing the work, there
> may be a race condition between ndev, which may cause UAF
> bug.
>
> CPU0 CPU1
>
> |queue_refill
> ns83820_remove_one |
> free_netdev |
> put_device |
> free ndev |
> |rx_refill
> |//use ndev
>
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> Signed-off-by: Zheng Wang <[email protected]>
I think it looks OK:
Reviewed-by: Horatiu Vultur <[email protected]>
> ---
> v3:
> - add tasklet_kill to stop more task scheduling suggested by
> Horatiu Vultur
> v2:
> - cancel the work after unregister_netdev to make sure there
> is no more request suggested by Jakub Kicinski
> ---
> drivers/net/ethernet/natsemi/ns83820.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/net/ethernet/natsemi/ns83820.c b/drivers/net/ethernet/natsemi/ns83820.c
> index 998586872599..af597719795d 100644
> --- a/drivers/net/ethernet/natsemi/ns83820.c
> +++ b/drivers/net/ethernet/natsemi/ns83820.c
> @@ -2208,8 +2208,14 @@ static void ns83820_remove_one(struct pci_dev *pci_dev)
>
> ns83820_disable_interrupts(dev); /* paranoia */
>
> + netif_carrier_off(ndev);
> + netif_tx_disable(ndev);
> +
> unregister_netdev(ndev);
> free_irq(dev->pci_dev->irq, ndev);
> + tasklet_kill(&dev->rx_tasklet);
> + cancel_work_sync(&dev->tq_refill);
> +
> iounmap(dev->base);
> dma_free_coherent(&dev->pci_dev->dev, 4 * DESC_SIZE * NR_TX_DESC,
> dev->tx_descs, dev->tx_phy_descs);
> --
> 2.25.1
>
--
/Horatiu
On Mon, 17 Apr 2023 09:31:07 +0800 Zheng Wang wrote:
> + netif_carrier_off(ndev);
> + netif_tx_disable(ndev);
> +
> unregister_netdev(ndev
It's not immediately obvious to me why disabling carrier and tx
are supposed to help. Please elaborate in the commit message if
you're confident that this is right.
--
pw-bot: cr
Jakub Kicinski <[email protected]> 于2023年4月19日周三 12:33写道:
>
> On Mon, 17 Apr 2023 09:31:07 +0800 Zheng Wang wrote:
> > + netif_carrier_off(ndev);
> > + netif_tx_disable(ndev);
> > +
> > unregister_netdev(ndev
>
> It's not immediately obvious to me why disabling carrier and tx
> are supposed to help. Please elaborate in the commit message if
> you're confident that this is right.
>
Hi Jakub,
Thanks for your reply. I'll figure it out to see if it's really necessary.
Best regards,
Zheng
> --
> pw-bot: cr