Skip to contentSkip to conversation

igb: remove skb_orphan calls

PATCH Applied Jeff Kirsher 2 participants 2 messages Mar 2, 2009 – Mar 3, 2009

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 igb: remove skb_orphan calls

From: Alexander Duyck <[email protected]>

Remove skb_orphan call from igb driver as it can cause multiple issues due to the fact that it is calling the desctructor and removing the skb from the socket prior to transmission.

The call was added to improve performance but did so by allowing the skb to be removed from the socket which gave the socket more window space to transmit. The performance gain is not worth the extra problems that this kind of workaround can introduce as this could lead to a potential DoS if a UDP stream decided to monopolize the transmit path.

---

 drivers/net/igb/igb_main.c |   11 -----------
 1 files changed, 0 insertions(+), 11 deletions(-)
drivers/net/igb/igb_main.c +0 -11
index 6cae258..78558f8 100644
@@ -3258,14 +3258,6 @@ static int igb_xmit_frame_ring_adv(struct sk_buff *skb,
 	if (unlikely(shtx->hardware)) {
 		shtx->in_progress = 1;
 		tx_flags |= IGB_TX_FLAGS_TSTAMP;
-	} else if (likely(!shtx->software)) {
-		/*
-		 * TODO: can this be solved in dev.c:dev_hard_start_xmit()?
-		 * There are probably unmodified driver which do something
-		 * like this and thus don't work in combination with
-		 * SOF_TIMESTAMPING_TX_SOFTWARE.
-		 */
-		skb_orphan(skb);
 	}
 
 	if (adapter->vlgrp && vlan_tx_tag_present(skb)) {
@@ -4253,9 +4245,6 @@ static void igb_tx_hwtstamp(struct igb_adapter *adapter, struct sk_buff *skb)
 				timecompare_transform(&adapter->compare, ns);
 			skb_tstamp_tx(skb, &shhwtstamps);
 		}
-
-		/* delayed orphaning: skb_tstamp_tx() needs the socket */
-		skb_orphan(skb);
 	}
 }
 
maintainer Applied
#1 lore raw

From: Jeff Kirsher <[email protected]>
Date: Mon, 02 Mar 2009 15:44:04 -0800

Quoted 12 lines

From: Alexander Duyck <[email protected]>

Remove skb_orphan call from igb driver as it can cause multiple issues due to the fact that it is calling the desctructor and removing the skb from the socket prior to transmission.

The call was added to improve performance but did so by allowing the skb to be removed from the socket which gave the socket more window space to transmit. The performance gain is not worth the extra problems that this kind of workaround can introduce as this could lead to a potential DoS if a UDP stream decided to monopolize the transmit path.

Quoted 3 lines

Applied.