Re: [PATCH net-next] sfc: trace_xdp_exception on XDP failure

[Arthur Fabre <afabre@cloudflare.com>]

On Mon, Nov 11, 2019 at 5:27 PM Edward Cree <ecree@solarflare.com> wrote:
>
> On 11/11/2019 10:51, Arthur Fabre wrote:
> > diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
> > index a7d9841105d8..5bfe1f6112a1 100644
> > --- a/drivers/net/ethernet/sfc/rx.c
> > +++ b/drivers/net/ethernet/sfc/rx.c
> > @@ -678,6 +678,7 @@ static bool efx_do_xdp(struct efx_nic *efx, struct efx_channel *channel,
> >                                 "XDP is not possible with multiple receive fragments (%d)\n",
> >                                 channel->rx_pkt_n_frags);
> >               channel->n_rx_xdp_bad_drops++;
> > +             trace_xdp_exception(efx->net_dev, xdp_prog, xdp_act);
> >               return false;
> >       }
> AIUI trace_xdp_exception() is improper here as we have not run
>  the XDP program (and xdp_act is thus uninitialised).
>
> The other three, below, appear to be correct.
> -Ed
>

Good point. Do you know under what conditions we'd end up with
"fragmented" packets? As far as I can tell this isn't IP
fragmentation?

On Mon, Nov 11, 2019 at 5:27 PM Edward Cree <ecree@solarflare.com> wrote:
>
> On 11/11/2019 10:51, Arthur Fabre wrote:
> > The sfc driver can drop packets processed with XDP, notably when running
> > out of buffer space on XDP_TX, or returning an unknown XDP action.
> > This increments the rx_xdp_bad_drops ethtool counter.
> >
> > Call trace_xdp_exception everywhere rx_xdp_bad_drops is incremented to
> > easily monitor this from userspace.
> >
> > This mirrors the behavior of other drivers.
> >
> > Signed-off-by: Arthur Fabre <afabre@cloudflare.com>
> > ---
> >  drivers/net/ethernet/sfc/rx.c | 4 ++++
> >  1 file changed, 4 insertions(+)
> >
> > diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
> > index a7d9841105d8..5bfe1f6112a1 100644
> > --- a/drivers/net/ethernet/sfc/rx.c
> > +++ b/drivers/net/ethernet/sfc/rx.c
> > @@ -678,6 +678,7 @@ static bool efx_do_xdp(struct efx_nic *efx, struct efx_channel *channel,
> >                                 "XDP is not possible with multiple receive fragments (%d)\n",
> >                                 channel->rx_pkt_n_frags);
> >               channel->n_rx_xdp_bad_drops++;
> > +             trace_xdp_exception(efx->net_dev, xdp_prog, xdp_act);
> >               return false;
> >       }
> AIUI trace_xdp_exception() is improper here as we have not run
>  the XDP program (and xdp_act is thus uninitialised).
>
> The other three, below, appear to be correct.
> -Ed
>
> >
> > @@ -724,6 +725,7 @@ static bool efx_do_xdp(struct efx_nic *efx, struct efx_channel *channel,
> >                               netif_err(efx, rx_err, efx->net_dev,
> >                                         "XDP TX failed (%d)\n", err);
> >                       channel->n_rx_xdp_bad_drops++;
> > +                     trace_xdp_exception(efx->net_dev, xdp_prog, xdp_act);
> >               } else {
> >                       channel->n_rx_xdp_tx++;
> >               }
> > @@ -737,6 +739,7 @@ static bool efx_do_xdp(struct efx_nic *efx, struct efx_channel *channel,
> >                               netif_err(efx, rx_err, efx->net_dev,
> >                                         "XDP redirect failed (%d)\n", err);
> >                       channel->n_rx_xdp_bad_drops++;
> > +                     trace_xdp_exception(efx->net_dev, xdp_prog, xdp_act);
> >               } else {
> >                       channel->n_rx_xdp_redirect++;
> >               }
> > @@ -746,6 +749,7 @@ static bool efx_do_xdp(struct efx_nic *efx, struct efx_channel *channel,
> >               bpf_warn_invalid_xdp_action(xdp_act);
> >               efx_free_rx_buffers(rx_queue, rx_buf, 1);
> >               channel->n_rx_xdp_bad_drops++;
> > +             trace_xdp_exception(efx->net_dev, xdp_prog, xdp_act);
> >               break;
> >
> >       case XDP_ABORTED:
>