Re: [PATCH 1/3] kvm tools: Make virtio-net kvm__irq_line thread safe

[Pekka Enberg <penberg@kernel.org>]

On Fri, Apr 29, 2011 at 9:36 AM, Asias He <asias.hejun@gmail.com> wrote:
> This patch fixes virtio-net randmom stall
>
> host $ scp guest:/root/big.guest .
> big.guest 42%  440MB  67.7KB/s - stalled -
>
> Signed-off-by: Asias He <asias.hejun@gmail.com>
> ---
>  tools/kvm/virtio-net.c |    9 +++++++--
>  1 files changed, 7 insertions(+), 2 deletions(-)
>
> diff --git a/tools/kvm/virtio-net.c b/tools/kvm/virtio-net.c
> index 58b3de4..efe06cb 100644
> --- a/tools/kvm/virtio-net.c
> +++ b/tools/kvm/virtio-net.c
> @@ -77,9 +77,12 @@ static void virtio_net_rx_callback(struct kvm *self, void *param)
>                head = virt_queue__get_iov(vq, iov, &out, &in, self);
>                len = readv(net_device.tap_fd, iov, in);
>                virt_queue__set_used_elem(vq, head, len);
> -       }
>
> -       kvm__irq_line(self, VIRTIO_NET_IRQ, 1);
> +               /* We should interrupt guest right now, otherwise latency is huge. */
> +               mutex_lock(&net_device.mutex);
> +               kvm__irq_line(self, VIRTIO_NET_IRQ, 1);
> +               mutex_unlock(&net_device.mutex);
> +       }
>  }
>
>  static void virtio_net_tx_callback(struct kvm *self, void *param)

Please make that IRQ latency fix a separate patch. Don't we need to do
it for TX path as well, btw?