Skip to contentSkip to conversation

sky2: Avoid race in sky2_change_mtu

PATCH In review Mike McCormack 3 participants 4 messages May 3, 2010

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

PATCH 1/1 sky2: Avoid race in sky2_change_mtu

netif_stop_queue does not ensure all in-progress transmits are complete,
so use netif_tx_disable() instead.

Make sure NAPI polls are disabled, otherwise NAPI might trigger a TX
restart between when we stop the queue and NAPI is disabled.

---
 drivers/net/sky2.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
drivers/net/sky2.c +1 -1
index 088c797..b839bae 100644
@@ -2236,8 +2236,8 @@ static int sky2_change_mtu(struct net_device *dev, int new_mtu)
 	sky2_write32(hw, B0_IMSK, 0);
 
 	dev->trans_start = jiffies;	/* prevent tx timeout */
-	netif_stop_queue(dev);
 	napi_disable(&hw->napi);
+	netif_tx_disable(dev);
 
 	synchronize_irq(hw->pdev->irq);
 
Signature · 1 line
1.5.6.5
maintainer
#1 lore raw

From: Mike McCormack <[email protected]>
Date: Mon, 03 May 2010 23:18:56 +0900

Quoted 7 lines

netif_stop_queue does not ensure all in-progress transmits are complete,
so use netif_tx_disable() instead.

Make sure NAPI polls are disabled, otherwise NAPI might trigger a TX
restart between when we stop the queue and NAPI is disabled.

This looks quite reasonable, Stephen please review.

#2 lore raw
In reply to Mike McCormack
Quoted 5 lines · On Mon, 03 May 2010 23:18:56 +0900 Mike McCormack <[email protected]> wrote:

netif_stop_queue does not ensure all in-progress transmits are complete,
so use netif_tx_disable() instead.

Make sure NAPI polls are disabled, otherwise NAPI might trigger a TX
restart between when we stop the queue and NAPI is disabled.

How could NAPI trigger a TX restart? Restart is a timer, not controlled by NAPI. There is a different race that your patch isn't covering.

If a large MTU send is in progress while mtu is changing, the changes to the GMAC store and forward (sky2_set_tx_stfwd) are not synchronized with hardware queued transmits.

So I think a there needs to be another spinloop.

drivers/net/sky2.c +5 -1
@@ -2274,8 +2274,12 @@ static int sky2_change_mtu(struct net_de
 	imask = sky2_read32(hw, B0_IMSK);
 	sky2_write32(hw, B0_IMSK, 0);
 
-	dev->trans_start = jiffies;	/* prevent tx timeout */
 	netif_stop_queue(dev);
+
+	/* Wait for ongoing Tx to complete */
+	while (sky2->tx_cons != sky2->tx_prod)
+		udelay(10);
+
 	napi_disable(&hw->napi);
 
 	synchronize_irq(hw->pdev->irq);
maintainer
#3 lore raw

From: Stephen Hemminger <[email protected]>
Date: Mon, 3 May 2010 16:26:34 -0700

Quoted 11 lines
Quoted 5 lines · On Mon, 03 May 2010 23:18:56 +0900 Mike McCormack <[email protected]> wrote:

netif_stop_queue does not ensure all in-progress transmits are complete,
so use netif_tx_disable() instead.

Make sure NAPI polls are disabled, otherwise NAPI might trigger a TX
restart between when we stop the queue and NAPI is disabled.

How could NAPI trigger a TX restart? Restart is a timer, not controlled by NAPI. There is a different race that your patch isn't covering.

I think he meant to say that NAPI can trigger a TX queue wake, which indeed it can, and it will if there is space in the TX queue and we happen to see the TXIDLE status at this time.