Re: [net-next 4/4] gve: Add ethtool support

[Florian Fainelli]

On 6/26/2019 11:52 AM, Catherine Sullivan wrote:

[snip]

> +static void
> +gve_get_ethtool_stats(struct net_device *netdev,
> +                   struct ethtool_stats *stats, u64 *data)
> +{
> +     struct gve_priv *priv = netdev_priv(netdev);
> +     u64 rx_pkts, rx_bytes, tx_pkts, tx_bytes;
> +     int ring;
> +     int i;
> +
> +     ASSERT_RTNL();
> +
> +     if (!netif_carrier_ok(netdev))
> +             return;
> +
> +     for (rx_pkts = 0, rx_bytes = 0, ring = 0;
> +          ring < priv->rx_cfg.num_queues; ring++) {
> +             rx_pkts += priv->rx[ring].rpackets;
> +             rx_bytes += priv->rx[ring].rbytes;
> +     }
> +     for (tx_pkts = 0, tx_bytes = 0, ring = 0;
> +          ring < priv->tx_cfg.num_queues; ring++) {
> +             tx_pkts += priv->tx[ring].pkt_done;
> +             tx_bytes += priv->tx[ring].bytes_done;
> +     }

Maybe you do not need to support 32-bit guests with that driver, but you
might as well be correct and use the include/linux/u64_stats_sync.h
primitives to help return consistent 64-bit stats on 32-bit machines.

[snip]

> +int gve_adjust_queues(struct gve_priv *priv,
> +                   struct gve_queue_config new_rx_config,
> +                   struct gve_queue_config new_tx_config)
> +{
> +     int err;
> +
> +     if (netif_carrier_ok(priv->dev)) {

Should not that be netif_running()?
-- 
Florian