diff mbox series

[net,2/2] ravb: Fix use-after-free issue in ravb_remove and ravb_tx_timeout_work

Message ID 20231004091253.4194205-3-yoshihiro.shimoda.uh@renesas.com (mailing list archive)
State Changes Requested
Delegated to: Netdev Maintainers
Headers show
Series [net,1/2] ravb: Fix dma_free_coherent() of desc_bat timing in ravb_remove() | expand

Checks

Context Check Description
netdev/series_format success Single patches do not need cover letters
netdev/tree_selection success Clearly marked for net
netdev/fixes_present success Fixes tag present in non-next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 1340 this patch: 1340
netdev/cc_maintainers fail 2 blamed authors not CCed: sergei.shtylyov@cogentembedded.com mitsuhiro.kimura.kc@renesas.com; 3 maintainers not CCed: sergei.shtylyov@cogentembedded.com wsa+renesas@sang-engineering.com mitsuhiro.kimura.kc@renesas.com
netdev/build_clang success Errors and warnings before: 1363 this patch: 1363
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success Fixes tag looks correct
netdev/build_allmodconfig_warn success Errors and warnings before: 1363 this patch: 1363
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 8 lines checked
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/source_inline success Was 0 now: 0

Commit Message

Yoshihiro Shimoda Oct. 4, 2023, 9:12 a.m. UTC
The ravb_stop() should call cancel_work_sync(). Otherwise,
ravb_tx_timeout_work() is possible to use the freed priv after
ravb_remove() was called like below:

CPU0			CPU1
			ravb_tx_timeout()
ravb_remove()
unregister_netdev()
free_netdev(ndev)
// free priv
			ravb_tx_timeout_work()
			// use priv

unregister_netdev() will call .ndo_stop() so that ravb_stop() is
called. And, after phy_stop() was called, netif_carrier_off()
is also called. So that .ndo_tx_timeout() will be not called
after phy_stop().

Link: https://lore.kernel.org/netdev/872cf8d7-3bd6-b11a-82ac-a9f4c82d0a02@omp.ru/
Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
Reported-by: Zheng Wang <zyytlz.wz@163.com>
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
---
 drivers/net/ethernet/renesas/ravb_main.c | 2 ++
 1 file changed, 2 insertions(+)

Comments

Sergey Shtylyov Oct. 4, 2023, 6:29 p.m. UTC | #1
Hello!

   Hm, concerning the subject: don't we actually have use-after-free in ravb_tx_timeout()
only? Also, you place () after the function names in patch #1 but not in this patch, why?

On 10/4/23 12:12 PM, Yoshihiro Shimoda wrote:

> The ravb_stop() should call cancel_work_sync(). Otherwise,
> ravb_tx_timeout_work() is possible to use the freed priv after
> ravb_remove() was called like below:
> 
> CPU0			CPU1
> 			ravb_tx_timeout()
> ravb_remove()
> unregister_netdev()
> free_netdev(ndev)
> // free priv
> 			ravb_tx_timeout_work()
> 			// use priv
> 
> unregister_netdev() will call .ndo_stop() so that ravb_stop() is
> called. And, after phy_stop() was called, netif_carrier_off()

   s/was/is/?

> is also called. So that .ndo_tx_timeout() will be not called

   Will not be...

> after phy_stop().
> 
> Link: https://lore.kernel.org/netdev/872cf8d7-3bd6-b11a-82ac-a9f4c82d0a02@omp.ru/
> Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
> Reported-by: Zheng Wang <zyytlz.wz@163.com>
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>

   Otherwise:

Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru>

[...]

MBR, Sergey
Yoshihiro Shimoda Oct. 5, 2023, 12:25 a.m. UTC | #2
Hello Sergey,

> From: Sergey Shtylyov, Sent: Thursday, October 5, 2023 3:30 AM
> 
> Hello!
> 
>    Hm, concerning the subject: don't we actually have use-after-free in ravb_tx_timeout()
> only?

IIUC, the issue causes ravb_remove(), and is in ravb_tx_timeout_work().

> Also, you place () after the function names in patch #1 but not in this patch, why?

I thought that the subject was long so that remove the ()...

So, I'll fix the subject as the follow:

ravb: Fix use-after-free issue in ravb_tx_timeout_work()

> On 10/4/23 12:12 PM, Yoshihiro Shimoda wrote:
> 
> > The ravb_stop() should call cancel_work_sync(). Otherwise,
> > ravb_tx_timeout_work() is possible to use the freed priv after
> > ravb_remove() was called like below:
> >
> > CPU0			CPU1
> > 			ravb_tx_timeout()
> > ravb_remove()
> > unregister_netdev()
> > free_netdev(ndev)
> > // free priv
> > 			ravb_tx_timeout_work()
> > 			// use priv
> >
> > unregister_netdev() will call .ndo_stop() so that ravb_stop() is
> > called. And, after phy_stop() was called, netif_carrier_off()
> 
>    s/was/is/?

I'll fix it.

> > is also called. So that .ndo_tx_timeout() will be not called
> 
>    Will not be...

Oops. I'll fix it.

> > after phy_stop().
> >
> > Link:
<snip URL>
> > Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
> > Reported-by: Zheng Wang <zyytlz.wz@163.com>
> > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> 
>    Otherwise:
> 
> Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru>

Thank you for your review!

Best regards,
Yoshihiro Shimoda

> [...]
> 
> MBR, Sergey
diff mbox series

Patch

diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
index 9e2e801049cc..0ef0b88b7145 100644
--- a/drivers/net/ethernet/renesas/ravb_main.c
+++ b/drivers/net/ethernet/renesas/ravb_main.c
@@ -2167,6 +2167,8 @@  static int ravb_close(struct net_device *ndev)
 			of_phy_deregister_fixed_link(np);
 	}
 
+	cancel_work_sync(&priv->work);
+
 	if (info->multi_irqs) {
 		free_irq(priv->tx_irqs[RAVB_NC], ndev);
 		free_irq(priv->rx_irqs[RAVB_NC], ndev);