linux-kernel - Re: [PATCH net] net: xilinx: axienet: Fix Tx skb circular buffer occupancy check in dmaengine xmit

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]

Message-ID: <b35cdeb3-399e-4ca0-a9a4-66bab45bd7e4@linux.dev>
Date: Fri, 23 May 2025 12:21:10 -0400
From: Sean Anderson <sean.anderson@...ux.dev>
To: Suraj Gupta <suraj.gupta2@....com>, andrew+netdev@...n.ch,
 davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org,
 pabeni@...hat.com, michal.simek@....com, radhey.shyam.pandey@....com,
 horms@...nel.org
Cc: netdev@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
 linux-kernel@...r.kernel.org, git@....com, harini.katakam@....com
Subject: Re: [PATCH net] net: xilinx: axienet: Fix Tx skb circular buffer
 occupancy check in dmaengine xmit

On 5/23/25 12:20, Sean Anderson wrote:
> On 5/21/25 14:16, Suraj Gupta wrote:
>> In Dmaengine flow, driver maintains struct skbuf_dma_descriptor rings each
>> element of which corresponds to a skb. In Tx datapath, compare available
>> space in skb ring with number of skbs instead of skb fragments.
>> Replace x * (MAX_SKB_FRAGS) in netif_txq_completed_wake() and
>> netif_txq_maybe_stop() with x * (1 skb) to fix the comparison.
>> 
>> Fixes: 6a91b846af85 ("net: axienet: Introduce dmaengine support")
>> Signed-off-by: Suraj Gupta <suraj.gupta2@....com>
>> ---
>>  drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 6 +++---
>>  1 file changed, 3 insertions(+), 3 deletions(-)
>> 
>> diff --git a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c
>> index 1b7a653c1f4e..6011d7eae0c7 100644
>> --- a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c
>> +++ b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c
>> @@ -880,7 +880,7 @@ static void axienet_dma_tx_cb(void *data, const struct dmaengine_result *result)
>>  	dev_consume_skb_any(skbuf_dma->skb);
>>  	netif_txq_completed_wake(txq, 1, len,
>>  				 CIRC_SPACE(lp->tx_ring_head, lp->tx_ring_tail, TX_BD_NUM_MAX),
>> -				 2 * MAX_SKB_FRAGS);
>> +				 2);
>>  }
>>  
>>  /**
>> @@ -914,7 +914,7 @@ axienet_start_xmit_dmaengine(struct sk_buff *skb, struct net_device *ndev)
>>  
>>  	dma_dev = lp->tx_chan->device;
>>  	sg_len = skb_shinfo(skb)->nr_frags + 1;
>> -	if (CIRC_SPACE(lp->tx_ring_head, lp->tx_ring_tail, TX_BD_NUM_MAX) <= sg_len) {
>> +	if (CIRC_SPACE(lp->tx_ring_head, lp->tx_ring_tail, TX_BD_NUM_MAX) <= 1) {
>>  		netif_stop_queue(ndev);
>>  		if (net_ratelimit())
>>  			netdev_warn(ndev, "TX ring unexpectedly full\n");
>> @@ -964,7 +964,7 @@ axienet_start_xmit_dmaengine(struct sk_buff *skb, struct net_device *ndev)
>>  	txq = skb_get_tx_queue(lp->ndev, skb);
>>  	netdev_tx_sent_queue(txq, skb->len);
>>  	netif_txq_maybe_stop(txq, CIRC_SPACE(lp->tx_ring_head, lp->tx_ring_tail, TX_BD_NUM_MAX),
>> -			     MAX_SKB_FRAGS + 1, 2 * MAX_SKB_FRAGS);
>> +			     1, 2);
>>  
>>  	dmaengine_submit(dma_tx_desc);
>>  	dma_async_issue_pending(lp->tx_chan);
> 
> Reviwed-by: Sean Anderson <sean.anderson@...ux.dev>

Reviewed-by: Sean Anderson <sean.anderson@...ux.dev>

*spelling