Re: [Intel-wired-lan] [PATCH net] ice: fix memory leak in ice_lbtest_prepare_rings()

[Marcin Szycik <marcin.szycik@linux.intel.com>]

On 09.06.2026 14:50, Dawei Feng wrote:
> While ice_lbtest_prepare_rings() correctly frees Rx rings if
> ice_vsi_start_all_rx_rings() fails, the earlier error paths for
> ice_vsi_setup_rx_rings() and ice_vsi_cfg_lan() jump past this cleanup.
> If Rx ring setup or LAN configuration fails, the function leaks the
> initialized Rx resources.
> 
> Fix this by routing these earlier failures to the existing
> err_start_rx_ring label. This ensures the Rx rings are properly freed
> before tearing down the Tx state.
> 
> The bug was first flagged by an experimental analysis tool we are
> developing for kernel memory-management bugs while analyzing
> v6.13-rc1. The tool is still under development and is not yet publicly
> available. Manual inspection confirms that the bug is still
> present in v7.1-rc5.
> 
> An x86_64 allyesconfig build showed no new warnings. As we do not have an
> Intel E800 Series adapter available to run the ethtool offline loopback
> selftest, no runtime testing was able to be performed.

IMO last two paragraphs should not be included in commit message,
rather after ---.

> Fixes: 0e674aeb0b77 ("ice: Add handler for ethtool selftest")
> Cc: stable@vger.kernel.org
> Signed-off-by: Zilin Guan <zilin@seu.edu.cn>
> Signed-off-by: Dawei Feng <dawei.feng@seu.edu.cn>
> ---
>  drivers/net/ethernet/intel/ice/ice_ethtool.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
> index f28416a707d7..7c81ca313645 100644
> --- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
> +++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
> @@ -1065,11 +1065,11 @@ static int ice_lbtest_prepare_rings(struct ice_vsi *vsi)
>  
>  	status = ice_vsi_setup_rx_rings(vsi);
>  	if (status)
> -		goto err_setup_rx_ring;
> +		goto err_start_rx_ring;
>  
>  	status = ice_vsi_cfg_lan(vsi);
>  	if (status)
> -		goto err_setup_rx_ring;
> +		goto err_start_rx_ring;
>  
>  	status = ice_vsi_start_all_rx_rings(vsi);
>  	if (status)
> @@ -1079,7 +1079,6 @@ static int ice_lbtest_prepare_rings(struct ice_vsi *vsi)
>  
>  err_start_rx_ring:
>  	ice_vsi_free_rx_rings(vsi);
> -err_setup_rx_ring:
>  	ice_vsi_stop_lan_tx_rings(vsi, ICE_NO_RESET, 0);

Correct me if I'm wrong, but looks like unroll order is reversed:
ice_vsi_stop_lan_tx_rings() unrolls ice_vsi_cfg_lan()
ice_vsi_free_rx_rings() unrolls ice_vsi_setup_rx_rings()
(was reversed before this patch too, but since we're fixing it, might as well)

>  err_setup_tx_ring:
>  	ice_vsi_free_tx_rings(vsi);

Thanks,
Marcin