NACK: [SRU][F/J][PATCH 1/1] can: hi311x: hi3110_can_ist(): fix potential use-after-free

[Koichiro Den]

On Wed, Feb 12, 2025 at 07:03:13PM GMT, Massimiliano Pellizzer wrote:
> From: Dario Binacchi <dario.binacchi at amarulasolutions.com>
> 
> The commit a22bd630cfff ("can: hi311x: do not report txerr and rxerr
> during bus-off") removed the reporting of rxerr and txerr even in case
> of correct operation (i. e. not bus-off).
> 
> The error count information added to the CAN frame after netif_rx() is
> a potential use after free, since there is no guarantee that the skb
> is in the same state. It might be freed or reused.
> 
> Fix the issue by postponing the netif_rx() call in case of txerr and
> rxerr reporting.
> 
> Fixes: a22bd630cfff ("can: hi311x: do not report txerr and rxerr during bus-off")
> Signed-off-by: Dario Binacchi <dario.binacchi at amarulasolutions.com>
> Link: https://patch.msgid.link/20241122221650.633981-5-dario.binacchi@amarulasolutions.com
> Signed-off-by: Marc Kleine-Budde <mkl at pengutronix.de>
> (backported from commit 9ad86d377ef4a19c75a9c639964879a5b25a433b)
> [mpellizzer: backported considering differences in context due to
> missing commit:
> - 3e5c291c7942d can: add CAN_ERR_CNT flag to notify availability of error counter
> which do not affect the patch]
> CVE-2024-56651
> Signed-off-by: Massimiliano Pellizzer <massimiliano.pellizzer at canonical.com>
> ---
>  drivers/net/can/spi/hi311x.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/can/spi/hi311x.c b/drivers/net/can/spi/hi311x.c
> index 28273e84171a2..660c4fe4614f5 100644
> --- a/drivers/net/can/spi/hi311x.c
> +++ b/drivers/net/can/spi/hi311x.c
> @@ -673,9 +673,9 @@ static irqreturn_t hi3110_can_ist(int irq, void *dev_id)
>  			tx_state = txerr >= rxerr ? new_state : 0;
>  			rx_state = txerr <= rxerr ? new_state : 0;
>  			can_change_state(net, cf, tx_state, rx_state);
> -			netif_rx_ni(skb);
>  
>  			if (new_state == CAN_STATE_BUS_OFF) {
> +				netif_rx(skb);

J/F is quite old and netif_rx_ni() is still alive. This handler runs in an
irq handler thread, so it should be kept as-is. Even if that weren't the
case, making a silent change should be avoided.

>  				can_bus_off(net);
>  				if (priv->can.restart_ms == 0) {
>  					priv->force_quit = 1;
> @@ -685,6 +685,7 @@ static irqreturn_t hi3110_can_ist(int irq, void *dev_id)
>  			} else {
>  				cf->data[6] = txerr;
>  				cf->data[7] = rxerr;
> +				netif_rx(skb);
>  			}
>  		}
>  
> -- 
> 2.43.0
> 
> 
> -- 
> kernel-team mailing list
> kernel-team at lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team