[PATCH] net/ice: fix null pointer dereferences

Bruce Richardson bruce.richardson at intel.com
Fri Mar 1 11:34:48 CET 2024


On Fri, Mar 01, 2024 at 01:20:29PM +0800, Wenwu Ma wrote:
> This patch fixes two null pointer dereferences detected by
> coverity scan.
> 
> Coverity issue: 414096
> Fixes: 6ccef90ff5d3 ("net/ice: support VSI level bandwidth config")
> 
> Signed-off-by: Wenwu Ma <wenwux.ma at intel.com>
> ---
>  drivers/net/ice/ice_tm.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/ice/ice_tm.c b/drivers/net/ice/ice_tm.c
> index fbab0b8808..e10ac855f9 100644
> --- a/drivers/net/ice/ice_tm.c
> +++ b/drivers/net/ice/ice_tm.c
> @@ -616,7 +616,10 @@ static int ice_set_node_rate(struct ice_hw *hw,
>  					   ICE_MAX_BW,
>  					   rate);
>  	if (status) {
> -		PMD_DRV_LOG(ERR, "Failed to set max bandwidth for node %u", tm_node->id);
> +		if (tm_node != NULL)
> +			PMD_DRV_LOG(ERR, "Failed to set max bandwidth for node %u", tm_node->id);
> +		else
> +			PMD_DRV_LOG(ERR, "Failed to set max bandwidth");
>  		return -EINVAL;
>  	}
>  
> @@ -630,7 +633,10 @@ static int ice_set_node_rate(struct ice_hw *hw,
>  					   ICE_MIN_BW,
>  					   rate);
>  	if (status) {
> -		PMD_DRV_LOG(ERR, "Failed to set min bandwidth for node %u", tm_node->id);
> +		if (tm_node != NULL)
> +			PMD_DRV_LOG(ERR, "Failed to set min bandwidth for node %u", tm_node->id);
> +		else
> +			PMD_DRV_LOG(ERR, "Failed to set min bandwidth");
>  		return -EINVAL;
>  	}
>  
Hi Wenwu,

I'm not sure that this is the best fix here, since the error message
doesn't seem particularly useful without the node id. Looking at the code,
this is a static function, so non-public, and only called in three places in
rte_tm.c: from ice_cfg_hw_node, ice_do_hierarchy_commit and
ice_reset_nolead_nodes. In all three cases, failure of this function is
immediately followed by a more specific error message from the calling
function. Therefore, I think we can solve the coverity problem by just
deleting the error prints from here completely, and let the callers manage
error reporting.

What do you think?

/Bruce


More information about the dev mailing list