Messages in this thread | | | Date | Wed, 16 Nov 2022 00:04:29 +0100 | From | Francois Romieu <> | Subject | Re: [PATCH net v2 3/3] net: nixge: fix tx queue handling |
| |
Zhang Changzhong <zhangchangzhong@huawei.com> : > Currently the driver check for available space at the beginning of > nixge_start_xmit(), and when there is not enough space for this packet, > it returns NETDEV_TX_OK, which casues packet loss and memory leak. > > Instead the queue should be stopped after the packet is added to the BD > when there may not be enough space for next packet. In addition, the > queue should be wakeup only if there is enough space for a packet with > max frags. > > Fixes: 492caffa8a1a ("net: ethernet: nixge: Add support for National Instruments XGE netdev") > Signed-off-by: Zhang Changzhong <zhangchangzhong@huawei.com> > --- > drivers/net/ethernet/ni/nixge.c | 54 +++++++++++++++++++++++++++++------------ > 1 file changed, 38 insertions(+), 16 deletions(-) > > diff --git a/drivers/net/ethernet/ni/nixge.c b/drivers/net/ethernet/ni/nixge.c > index 91b7ebc..3776a03 100644 > --- a/drivers/net/ethernet/ni/nixge.c > +++ b/drivers/net/ethernet/ni/nixge.c [...] > static netdev_tx_t nixge_start_xmit(struct sk_buff *skb, > @@ -518,10 +523,15 @@ static netdev_tx_t nixge_start_xmit(struct sk_buff *skb, > cur_p = &priv->tx_bd_v[priv->tx_bd_tail]; > tx_skb = &priv->tx_skb[priv->tx_bd_tail]; > > - if (nixge_check_tx_bd_space(priv, num_frag + 1)) { > - if (!netif_queue_stopped(ndev)) > - netif_stop_queue(ndev); > - return NETDEV_TX_OK; > + if (unlikely(nixge_check_tx_bd_space(priv, num_frag + 1))) { > + /* Should not happen as last start_xmit call should have > + * checked for sufficient space and queue should only be > + * woken when sufficient space is available. > + */
Almost. IRQ triggering after nixge_start_xmit::netif_stop_queue and before nixge_start_xmit::smp_mb may wrongly wake queue.
Call me timorous but I would feel more confortable if this code could be tested on real hardware before being fed into -net.
-- Ueimor
| |