[PATCH] igb: Fix NULL pointer dereference in ethtool loopback test

[PATCH] igb: Fix NULL pointer dereference in ethtool loopback test

[Tianyu Xu]

The igb driver currently causes a NULL pointer dereference
when executing the ethtool loopback test. This occurs because
there is no associated q_vector for the test ring when it is
set up, as interrupts are typically not added to the test rings.

Since commit 5ef44b3cb43b removed the napi_id assignment in
__xdp_rxq_info_reg(), there is no longer a need to pass a napi_id.
Therefore, simply use 0 as the final parameter.

Signed-off-by: Tianyu Xu <tianyxu@cisco.com>
---
 drivers/net/ethernet/intel/igb/igb_main.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index a9a7a94ae..453deb6d1 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -4453,8 +4453,7 @@ int igb_setup_rx_resources(struct igb_ring *rx_ring)
 	if (xdp_rxq_info_is_reg(&rx_ring->xdp_rxq))
 		xdp_rxq_info_unreg(&rx_ring->xdp_rxq);
 	res = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev,
-			       rx_ring->queue_index,
-			       rx_ring->q_vector->napi.napi_id);
+			       rx_ring->queue_index, 0);
 	if (res < 0) {
 		dev_err(dev, "Failed to register xdp_rxq index %u\n",
 			rx_ring->queue_index);
-- 
2.39.5 (Apple Git-154)

RE: [Intel-wired-lan] [PATCH] igb: Fix NULL pointer dereference in ethtool loopback test

[Loktionov, Aleksandr]

> -----Original Message-----
> From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf
> Of Tianyu Xu
> Sent: Monday, August 11, 2025 1:42 PM
> To: Nguyen, Anthony L <anthony.l.nguyen@intel.com>
> Cc: Kitszel, Przemyslaw <przemyslaw.kitszel@intel.com>;
> andrew+netdev@lunn.ch; kuba@kernel.org; sdf@fomichev.me; intel-wired-
> lan@lists.osuosl.org; netdev@vger.kernel.org; linux-
> kernel@vger.kernel.org; tianyxu@cisco.com
> Subject: [Intel-wired-lan] [PATCH] igb: Fix NULL pointer dereference
> in ethtool loopback test
> 
> The igb driver currently causes a NULL pointer dereference
> when executing the ethtool loopback test. This occurs because
> there is no associated q_vector for the test ring when it is
> set up, as interrupts are typically not added to the test rings.
> 
> Since commit 5ef44b3cb43b removed the napi_id assignment in
> __xdp_rxq_info_reg(), there is no longer a need to pass a napi_id.
> Therefore, simply use 0 as the final parameter.
> 
Good day,  Tianyu Xu,

Thanks for the patch. I've reviewed the commit message and have a few minor suggestions to align it with kernel style guidelines:

Can you consider wrapping body lines at 75 characters for better readability?
If applicable, please include Fixes: tag in the commit message. This helps maintainers and automated tools track regressions and understand the context of the fix. Other than that, the change itself looks good and clearly improves the accuracy of statistics reporting.

Best regards,
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>

> Signed-off-by: Tianyu Xu <tianyxu@cisco.com>
> ---
>  drivers/net/ethernet/intel/igb/igb_main.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/net/ethernet/intel/igb/igb_main.c
> b/drivers/net/ethernet/intel/igb/igb_main.c
> index a9a7a94ae..453deb6d1 100644
> --- a/drivers/net/ethernet/intel/igb/igb_main.c
> +++ b/drivers/net/ethernet/intel/igb/igb_main.c
> @@ -4453,8 +4453,7 @@ int igb_setup_rx_resources(struct igb_ring
> *rx_ring)
>  	if (xdp_rxq_info_is_reg(&rx_ring->xdp_rxq))
>  		xdp_rxq_info_unreg(&rx_ring->xdp_rxq);
>  	res = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev,
> -			       rx_ring->queue_index,
> -			       rx_ring->q_vector->napi.napi_id);
> +			       rx_ring->queue_index, 0);
>  	if (res < 0) {
>  		dev_err(dev, "Failed to register xdp_rxq index %u\n",
>  			rx_ring->queue_index);
> --
> 2.39.5 (Apple Git-154)

Re: [PATCH] igb: Fix NULL pointer dereference in ethtool loopback test

[Joe Damato]

On Mon, Aug 11, 2025 at 07:41:53PM +0800, Tianyu Xu wrote:
> The igb driver currently causes a NULL pointer dereference
> when executing the ethtool loopback test. This occurs because
> there is no associated q_vector for the test ring when it is
> set up, as interrupts are typically not added to the test rings.
> 
> Since commit 5ef44b3cb43b removed the napi_id assignment in
> __xdp_rxq_info_reg(), there is no longer a need to pass a napi_id.
> Therefore, simply use 0 as the final parameter.
> 
> Signed-off-by: Tianyu Xu <tianyxu@cisco.com>
> ---
>  drivers/net/ethernet/intel/igb/igb_main.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
> index a9a7a94ae..453deb6d1 100644
> --- a/drivers/net/ethernet/intel/igb/igb_main.c
> +++ b/drivers/net/ethernet/intel/igb/igb_main.c
> @@ -4453,8 +4453,7 @@ int igb_setup_rx_resources(struct igb_ring *rx_ring)
>  	if (xdp_rxq_info_is_reg(&rx_ring->xdp_rxq))
>  		xdp_rxq_info_unreg(&rx_ring->xdp_rxq);
>  	res = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev,
> -			       rx_ring->queue_index,
> -			       rx_ring->q_vector->napi.napi_id);
> +			       rx_ring->queue_index, 0);
>  	if (res < 0) {
>  		dev_err(dev, "Failed to register xdp_rxq index %u\n",
>  			rx_ring->queue_index);

This LGTM but will probably need to be re-sent with a Fixes tag.

See: https://www.kernel.org/doc/html/v6.16/process/maintainer-netdev.html#tl-dr

If you repost, feel free to add:

Reviewed-by: Joe Damato <joe@dama.to>