Re: [PATCH 5.10] ibmvnic: Add tx check to prevent skb leak

[Jacob Keller <jacob.e.keller@intel.com>]

On 1/20/2025 4:46 AM, Denis Arefev wrote:
> From: Nick Child <nnac123@linux.ibm.com>
> 
> From: Nick Child <nnac123@linux.ibm.com>
> 
> commit 0983d288caf984de0202c66641577b739caad561 upstream.
> 
> Below is a summary of how the driver stores a reference to an skb during
> transmit:
>     tx_buff[free_map[consumer_index]]->skb = new_skb;
>     free_map[consumer_index] = IBMVNIC_INVALID_MAP;
>     consumer_index ++;
> Where variable data looks like this:
>     free_map == [4, IBMVNIC_INVALID_MAP, IBMVNIC_INVALID_MAP, 0, 3]
>                                                	consumer_index^
>     tx_buff == [skb=null, skb=<ptr>, skb=<ptr>, skb=null, skb=null]
> 
> The driver has checks to ensure that free_map[consumer_index] pointed to
> a valid index but there was no check to ensure that this index pointed
> to an unused/null skb address. So, if, by some chance, our free_map and
> tx_buff lists become out of sync then we were previously risking an
> skb memory leak. This could then cause tcp congestion control to stop
> sending packets, eventually leading to ETIMEDOUT.
> 
> Therefore, add a conditional to ensure that the skb address is null. If
> not then warn the user (because this is still a bug that should be
> patched) and free the old pointer to prevent memleak/tcp problems.
> 
> Signed-off-by: Nick Child <nnac123@linux.ibm.com>
> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
> [Denis: minor fix to resolve merge conflict.]
> Signed-off-by: Denis Arefev <arefev@swemel.ru>
> ---

I thought the process asked to have the stable tag, i.e.

Cc: <stable@vger.kernel.org> # 5.10.x

Anyways, this looks good to me, and seems like a good candidate for
backporting.

Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>

Thanks,
Jake

> Backport fix for CVE-2024-41066
> Link: https://nvd.nist.gov/vuln/detail/CVE-2024-41066
> ---
>  drivers/net/ethernet/ibm/ibmvnic.c | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/drivers/net/ethernet/ibm/ibmvnic.c b/drivers/net/ethernet/ibm/ibmvnic.c
> index 84da6ccaf339..439796975cbf 100644
> --- a/drivers/net/ethernet/ibm/ibmvnic.c
> +++ b/drivers/net/ethernet/ibm/ibmvnic.c
> @@ -1625,6 +1625,18 @@ static netdev_tx_t ibmvnic_xmit(struct sk_buff *skb, struct net_device *netdev)
>  	    (tx_pool->consumer_index + 1) % tx_pool->num_buffers;
>  
>  	tx_buff = &tx_pool->tx_buff[index];
> +
> +	/* Sanity checks on our free map to make sure it points to an index
> +	 * that is not being occupied by another skb. If skb memory is
> +	 * not freed then we see congestion control kick in and halt tx.
> +	 */
> +	if (unlikely(tx_buff->skb)) {
> +		dev_warn_ratelimited(dev, "TX free map points to untracked skb (%s %d idx=%d)\n",
> +				     skb_is_gso(skb) ? "tso_pool" : "tx_pool",
> +				     queue_num, bufidx);
> +		dev_kfree_skb_any(tx_buff->skb);
> +	}
> +
>  	tx_buff->skb = skb;
>  	tx_buff->data_dma[0] = data_dma_addr;
>  	tx_buff->data_len[0] = skb->len;