On Tue, Sep 11, 2018 at 11:35 PM Jason Wang <jasow...@redhat.com> wrote: > > > > On 2018年09月11日 09:14, Willem de Bruijn wrote: > >>>> I cook a fixup, and it looks works in my setup: > >>>> > >>>> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c > >>>> index b320b6b14749..9181c3f2f832 100644 > >>>> --- a/drivers/net/virtio_net.c > >>>> +++ b/drivers/net/virtio_net.c > >>>> @@ -2204,10 +2204,17 @@ static int virtnet_set_coalesce(struct > >>>> net_device *dev, > >>>> return -EINVAL; > >>>> > >>>> if (napi_weight ^ vi->sq[0].napi.weight) { > >>>> - if (dev->flags & IFF_UP) > >>>> - return -EBUSY; > >>>> - for (i = 0; i < vi->max_queue_pairs; i++) > >>>> + for (i = 0; i < vi->max_queue_pairs; i++) { > >>>> + struct netdev_queue *txq = > >>>> + netdev_get_tx_queue(vi->dev, i); > >>>> + > >>>> + virtnet_napi_tx_disable(&vi->sq[i].napi); > >>>> + __netif_tx_lock_bh(txq); > >>>> vi->sq[i].napi.weight = napi_weight; > >>>> + __netif_tx_unlock_bh(txq); > >>>> + virtnet_napi_tx_enable(vi, vi->sq[i].vq, > >>>> + &vi->sq[i].napi); > >>>> + } > >>>> } > >>>> > >>>> return 0; > >>> Thanks! It passes my simple stress test, too. Which consists of two > >>> concurrent loops, one toggling the ethtool option, another running > >>> TCP_RR. > >>> > >>>> The only left case is the speculative tx polling in RX NAPI. I think we > >>>> don't need to care in this case since it was not a must for correctness. > >>> As long as the txq lock is held that will be a noop, anyway. The other > >>> concurrent action is skb_xmit_done. It looks correct to me, but need > >>> to think about it a bit. The tricky transition is coming out of napi > >>> without > >>> having >= 2 + MAX_SKB_FRAGS clean descriptors. If the queue is > >>> stopped it may deadlock transmission in no-napi mode. > >> Yes, maybe we can enable tx queue when napi weight is zero in > >> virtnet_poll_tx(). > > Yes, that precaution should resolve that edge case. > > > > I've done a stress test and it passes. The test contains: > > - vm with 2 queues > - a bash script to enable and disable tx napi > - two netperf UDP_STREAM sessions to send small packets
Great. That matches my results. Do you want to send the v2?