2023-03-10 10:59:08

by Zheng Wang

Subject: [PATCH net] net: qcom/emac: Fix use after free bug in emac_remove due to race condition

In emac_probe, &adpt->work_thread is bound with
emac_work_thread. Then it will be started by timeout
handler emac_tx_timeout or a IRQ handler emac_isr.

If we remove the driver which will call emac_remove
to make cleanup, there may be a unfinished work.

The possible sequence is as follows:

Fix it by finishing the work before cleanup in the emac_remove
and disable timeout response.

CPU0 CPU1

|emac_work_thread
emac_remove |
free_netdev |
kfree(netdev); |
|emac_reinit_locked
|emac_mac_down
|//use netdev
Fixes: b9b17debc69d ("net: emac: emac gigabit ethernet controller driver")
Signed-off-by: Zheng Wang <[email protected]>
---
drivers/net/ethernet/qualcomm/emac/emac.c | 3 +++
1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/qualcomm/emac/emac.c b/drivers/net/ethernet/qualcomm/emac/emac.c
index 3115b2c12898..ddc328f7b96a 100644
--- a/drivers/net/ethernet/qualcomm/emac/emac.c
+++ b/drivers/net/ethernet/qualcomm/emac/emac.c
@@ -724,6 +724,9 @@ static int emac_remove(struct platform_device *pdev)
struct net_device *netdev = dev_get_drvdata(&pdev->dev);
struct emac_adapter *adpt = netdev_priv(netdev);

+ netif_carrier_off(netdev);
+ netif_tx_disable(netdev);
+ cancel_work_sync(&adpt->work_thread);
unregister_netdev(netdev);
netif_napi_del(&adpt->rx_q.napi);

--
2.25.1

2023-03-14 00:00:55

by Jakub Kicinski

Subject: Re: [PATCH net] net: qcom/emac: Fix use after free bug in emac_remove due to race condition

On Fri, 10 Mar 2023 18:57:34 +0800 Zheng Wang wrote:
> + netif_carrier_off(netdev);
> + netif_tx_disable(netdev);
> + cancel_work_sync(&adpt->work_thread);
> unregister_netdev(netdev);
> netif_napi_del(&adpt->rx_q.napi);

same problem as in the natsemi driver

2023-03-14 02:12:01

by Zheng Hacker

Subject: Re: [PATCH net] net: qcom/emac: Fix use after free bug in emac_remove due to race condition

Jakub Kicinski <[email protected]> 于2023年3月14日周二 08:00写道：
>
> On Fri, 10 Mar 2023 18:57:34 +0800 Zheng Wang wrote:
> > + netif_carrier_off(netdev);
> > + netif_tx_disable(netdev);
> > + cancel_work_sync(&adpt->work_thread);
> > unregister_netdev(netdev);
> > netif_napi_del(&adpt->rx_q.napi);
>
> same problem as in the natsemi driver

Hi Jakub,

Thanks for your reply. Here we schedule the work when handling IRQ
request, So I think we should add free_irq after netif_napi_del and move
the cancel_work_sync after that like:

netif_carrier_off(netdev);
netif_tx_disable(netdev);

unregister_netdev(netdev);
netif_napi_del(&adpt->rx_q.napi);

free_irq(adpt->irq.irq, &adpt->irq);
cancel_work_sync(&adpt->work_thread);

Apprecaite for your advice about the fix.

Best regards,
Zheng