netdev - Re: [PATCH v2 06/11] net: calxedaxgmac: fix race with tx queue stop/wake

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 30 Aug 2013 18:12:46 +0100
From:	Ben Hutchings <bhutchings@...arflare.com>
To:	Rob Herring <robherring2@...il.com>
CC:	<netdev@...r.kernel.org>,
	Lennert Buytenhek <buytenh@...tstofly.org>,
	"Rob Herring" <rob.herring@...xeda.com>
Subject: Re: [PATCH v2 06/11] net: calxedaxgmac: fix race with tx queue
 stop/wake

On Thu, 2013-08-29 at 22:13 -0500, Rob Herring wrote:
> From: Rob Herring <rob.herring@...xeda.com>
> 
> Since the xgmac transmit start and completion work locklessly, it is
> possible for xgmac_xmit to stop the tx queue after the xgmac_tx_complete
> has run resulting in the tx queue never being woken up. Fix this by
> ensuring that ring buffer index updates are visible and recheck the ring
> space after stopping the queue.
> 
> The implementation used here was copied from
> drivers/net/ethernet/broadcom/tg3.c.
> 
> Signed-off-by: Rob Herring <rob.herring@...xeda.com>
> ---
> v2:
> - drop netif_tx_lock
> - use netif_start_queue instead of netif_wake_queue
[...]
> @@ -1125,10 +1130,15 @@ static netdev_tx_t xgmac_xmit(struct sk_buff *skb, struct net_device *dev)
>  
>  	priv->tx_head = dma_ring_incr(entry, DMA_TX_RING_SZ);
>  
> -	if (dma_ring_space(priv->tx_head, priv->tx_tail, DMA_TX_RING_SZ) <
> -	    MAX_SKB_FRAGS)
> +	/* Ensure tx_head update is visible to tx completion */
> +	smp_mb();
> +	if (unlikely(tx_dma_ring_space(priv) < MAX_SKB_FRAGS)) {
[...]

On re-reading, I see there's an off-by-one bug lingering here.  You use
up to 1 + MAX_SKB_FRAGS descriptors per skb.  So you must stop the queue
when the ring space is <= MAX_SKB_FRAGS.

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html