On Fri, Feb 6, 2026 at 4:35 PM Jakub Kicinski <[email protected]> wrote:
>
> Count and report HW-GRO stats as seen by the kernel.
> The device stats for GRO seem to not reflect the reality,
> perhaps they count sessions which did not actually result
> in any aggregation. Also they count wire packets, so we
> have to count super-frames, anyway.
>
> Signed-off-by: Jakub Kicinski <[email protected]>

Reviewed-by: Michael Chan <[email protected]>

I'm suggesting a minor naming change below if you need to do v3.

> @@ -13492,6 +13497,8 @@ static void bnxt_get_one_ring_err_stats(struct bnxt 
> *bp,

With the new GRO counters, this function is no longer limited to error
stats.  So maybe rename it to something like
bnxt_get_one_ring_misc_stats()?

>         stats->rx_total_netpoll_discards += sw_stats->rx.rx_netpoll_discards;
>         stats->rx_total_ring_discards +=
>                 BNXT_GET_RING_STATS64(hw_stats, rx_discard_pkts);
> +       stats->rx_total_hw_gro_packets += sw_stats->rx.rx_hw_gro_packets;
> +       stats->rx_total_hw_gro_wire_packets += 
> sw_stats->rx.rx_hw_gro_wire_packets;
>         stats->tx_total_resets += sw_stats->tx.tx_resets;
>         stats->tx_total_ring_discards +=
>                 BNXT_GET_RING_STATS64(hw_stats, tx_discard_pkts);

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

Reply via email to