diff mbox series

[net,v4] ravb: Fix races between ravb_tx_timeout_work() and net related ops

Message ID 20231127122420.3706751-1-yoshihiro.shimoda.uh@renesas.com (mailing list archive)
State Mainlined
Commit 9870257a0a338cd8d6c1cddab74e703f490f6779
Delegated to: Geert Uytterhoeven
Headers show
Series [net,v4] ravb: Fix races between ravb_tx_timeout_work() and net related ops | expand

Commit Message

Yoshihiro Shimoda Nov. 27, 2023, 12:24 p.m. UTC
Fix races between ravb_tx_timeout_work() and functions of net_device_ops
and ethtool_ops by using rtnl_trylock() and rtnl_unlock(). Note that
since ravb_close() is under the rtnl lock and calls cancel_work_sync(),
ravb_tx_timeout_work() should calls rtnl_trylock(). Otherwise, a deadlock
may happen in ravb_tx_timeout_work() like below:

CPU0			CPU1
			ravb_tx_timeout()
			schedule_work()
...
__dev_close_many()
// Under rtnl lock
ravb_close()
cancel_work_sync()
// Waiting
			ravb_tx_timeout_work()
			rtnl_lock()
			// This is possible to cause a deadlock

If rtnl_trylock() fails, rescheduling the work with sleep for 1 msec.

Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
---
Changes from v3:
https://lore.kernel.org/all/20231115022644.2316961-1-yoshihiro.shimoda.uh@renesas.com/
 - Based on v2 patch. This means that delayed work is not used.
 - Add rescheduling the work if rtnl_trylock() fails.
 - Drop Reviewed-by tags because implementation was changed.

Changes from v2:
https://lore.kernel.org/netdev/20231019113308.1133944-1-yoshihiro.shimoda.uh@renesas.com/
 - Add rescheduling if rtnl_trylock() fails and the netif is still running
   and update commit description for it.
 - Add Reviewed-by tags.

 drivers/net/ethernet/renesas/ravb_main.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

Comments

Sergey Shtylyov Nov. 27, 2023, 4:27 p.m. UTC | #1
On 11/27/23 3:24 PM, Yoshihiro Shimoda wrote:

> Fix races between ravb_tx_timeout_work() and functions of net_device_ops
> and ethtool_ops by using rtnl_trylock() and rtnl_unlock(). Note that
> since ravb_close() is under the rtnl lock and calls cancel_work_sync(),
> ravb_tx_timeout_work() should calls rtnl_trylock(). Otherwise, a deadlock
> may happen in ravb_tx_timeout_work() like below:
> 
> CPU0			CPU1
> 			ravb_tx_timeout()
> 			schedule_work()
> ...
> __dev_close_many()
> // Under rtnl lock
> ravb_close()
> cancel_work_sync()
> // Waiting
> 			ravb_tx_timeout_work()
> 			rtnl_lock()
> 			// This is possible to cause a deadlock
> 
> If rtnl_trylock() fails, rescheduling the work with sleep for 1 msec.
> 
> Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>

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

[...]

> diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
> index c70cff80cc99..7c007ecd3ff6 100644
> --- a/drivers/net/ethernet/renesas/ravb_main.c
> +++ b/drivers/net/ethernet/renesas/ravb_main.c
> @@ -1874,6 +1874,12 @@ static void ravb_tx_timeout_work(struct work_struct *work)
>  	struct net_device *ndev = priv->ndev;
>  	int error;
>  
> +	if (!rtnl_trylock()) {
> +		usleep_range(1000, 2000);

   Why not msleep(1) though?

> +		schedule_work(&priv->work);
> +		return;
> +	}
> +
>  	netif_tx_stop_all_queues(ndev);
>  
>  	/* Stop PTP Clock driver */
[...]

MBR, Sergey
Yoshihiro Shimoda Nov. 27, 2023, 11:58 p.m. UTC | #2
Hi Sergey,

> From: Sergey Shtylyov, Sent: Tuesday, November 28, 2023 1:28 AM
> 
> On 11/27/23 3:24 PM, Yoshihiro Shimoda wrote:
> 
> > Fix races between ravb_tx_timeout_work() and functions of net_device_ops
> > and ethtool_ops by using rtnl_trylock() and rtnl_unlock(). Note that
> > since ravb_close() is under the rtnl lock and calls cancel_work_sync(),
> > ravb_tx_timeout_work() should calls rtnl_trylock(). Otherwise, a deadlock
> > may happen in ravb_tx_timeout_work() like below:
> >
> > CPU0			CPU1
> > 			ravb_tx_timeout()
> > 			schedule_work()
> > ...
> > __dev_close_many()
> > // Under rtnl lock
> > ravb_close()
> > cancel_work_sync()
> > // Waiting
> > 			ravb_tx_timeout_work()
> > 			rtnl_lock()
> > 			// This is possible to cause a deadlock
> >
> > If rtnl_trylock() fails, rescheduling the work with sleep for 1 msec.
> >
> > Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper")
> > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> 
> Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru>

Thank you very much for your review!

> [...]
> 
> > diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
> > index c70cff80cc99..7c007ecd3ff6 100644
> > --- a/drivers/net/ethernet/renesas/ravb_main.c
> > +++ b/drivers/net/ethernet/renesas/ravb_main.c
> > @@ -1874,6 +1874,12 @@ static void ravb_tx_timeout_work(struct work_struct *work)
> >  	struct net_device *ndev = priv->ndev;
> >  	int error;
> >
> > +	if (!rtnl_trylock()) {
> > +		usleep_range(1000, 2000);
> 
>    Why not msleep(1) though?

It's for the following guideline:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/timers/timers-howto.rst?h=v6.7-rc3
-----
	SLEEPING FOR ~USECS OR SMALL MSECS ( 10us - 20ms):
		* Use usleep_range
-----

Best regards,
Yoshihiro Shimoda

> > +		schedule_work(&priv->work);
> > +		return;
> > +	}
> > +
> >  	netif_tx_stop_all_queues(ndev);
> >
> >  	/* Stop PTP Clock driver */
> [...]
> 
> MBR, Sergey
patchwork-bot+netdevbpf@kernel.org Nov. 29, 2023, 4:10 a.m. UTC | #3
Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Mon, 27 Nov 2023 21:24:20 +0900 you wrote:
> Fix races between ravb_tx_timeout_work() and functions of net_device_ops
> and ethtool_ops by using rtnl_trylock() and rtnl_unlock(). Note that
> since ravb_close() is under the rtnl lock and calls cancel_work_sync(),
> ravb_tx_timeout_work() should calls rtnl_trylock(). Otherwise, a deadlock
> may happen in ravb_tx_timeout_work() like below:
> 
> CPU0			CPU1
> 			ravb_tx_timeout()
> 			schedule_work()
> ...
> __dev_close_many()
> // Under rtnl lock
> ravb_close()
> cancel_work_sync()
> // Waiting
> 			ravb_tx_timeout_work()
> 			rtnl_lock()
> 			// This is possible to cause a deadlock
> 
> [...]

Here is the summary with links:
  - [net,v4] ravb: Fix races between ravb_tx_timeout_work() and net related ops
    https://git.kernel.org/netdev/net/c/9870257a0a33

You are awesome, thank you!
diff mbox series

Patch

diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
index c70cff80cc99..7c007ecd3ff6 100644
--- a/drivers/net/ethernet/renesas/ravb_main.c
+++ b/drivers/net/ethernet/renesas/ravb_main.c
@@ -1874,6 +1874,12 @@  static void ravb_tx_timeout_work(struct work_struct *work)
 	struct net_device *ndev = priv->ndev;
 	int error;
 
+	if (!rtnl_trylock()) {
+		usleep_range(1000, 2000);
+		schedule_work(&priv->work);
+		return;
+	}
+
 	netif_tx_stop_all_queues(ndev);
 
 	/* Stop PTP Clock driver */
@@ -1907,7 +1913,7 @@  static void ravb_tx_timeout_work(struct work_struct *work)
 		 */
 		netdev_err(ndev, "%s: ravb_dmac_init() failed, error %d\n",
 			   __func__, error);
-		return;
+		goto out_unlock;
 	}
 	ravb_emac_init(ndev);
 
@@ -1917,6 +1923,9 @@  static void ravb_tx_timeout_work(struct work_struct *work)
 		ravb_ptp_init(ndev, priv->pdev);
 
 	netif_tx_start_all_queues(ndev);
+
+out_unlock:
+	rtnl_unlock();
 }
 
 /* Packet transmit function for Ethernet AVB */