Re: [PATCH net] be2net: fix potential memory leak in be_xmit()

[Simon Horman <horms@kernel.org>]

On Mon, Oct 14, 2024 at 10:47:58PM +0800, Wang Hai wrote:
> The be_xmit() returns NETDEV_TX_OK without freeing skb
> in case of be_xmit_enqueue() fails, add dev_kfree_skb_any() to fix it.
> 
> Fixes: 760c295e0e8d ("be2net: Support for OS2BMC.")
> Signed-off-by: Wang Hai <wanghai38@huawei.com>

Thanks,

I agree that this is correct an din keeping with other
drop handling within this function.

I do, however, wonder if this logic could be expressed both
more clearly and idiomatically like this (completely untested!):

diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index a8596ebcdfd6..d171acb6544f 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -1381,10 +1381,8 @@ static netdev_tx_t be_xmit(struct sk_buff *skb, struct net_device *netdev)
 	be_get_wrb_params_from_skb(adapter, skb, &wrb_params);
 
 	wrb_cnt = be_xmit_enqueue(adapter, txo, skb, &wrb_params);
-	if (unlikely(!wrb_cnt)) {
-		dev_kfree_skb_any(skb);
-		goto drop;
-	}
+	if (unlikely(!wrb_cnt))
+		goto drop_skb;
 
 	/* if os2bmc is enabled and if the pkt is destined to bmc,
 	 * enqueue the pkt a 2nd time with mgmt bit set.
@@ -1393,7 +1391,7 @@ static netdev_tx_t be_xmit(struct sk_buff *skb, struct net_device *netdev)
 		BE_WRB_F_SET(wrb_params.features, OS2BMC, 1);
 		wrb_cnt = be_xmit_enqueue(adapter, txo, skb, &wrb_params);
 		if (unlikely(!wrb_cnt))
-			goto drop;
+			goto drop_skb;
 		else
 			skb_get(skb);
 	}
@@ -1407,6 +1405,9 @@ static netdev_tx_t be_xmit(struct sk_buff *skb, struct net_device *netdev)
 		be_xmit_flush(adapter, txo);
 
 	return NETDEV_TX_OK;
+
+drop_skb:
+	dev_kfree_skb_any(skb);
 drop:
 	tx_stats(txo)->tx_drv_drops++;
 	/* Flush the already enqueued tx requests */