[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f65d3ce1-2ff6-180b-805a-dbab1d53772f@solarflare.com>
Date: Fri, 28 Sep 2018 08:30:18 +0100
From: Bert Kenward <bkenward@...arflare.com>
To: Eric Dumazet <edumazet@...gle.com>,
"David S . Miller" <davem@...emloft.net>
CC: netdev <netdev@...r.kernel.org>,
Michael Chan <michael.chan@...adcom.com>,
Aviad Krawczyk <aviad.krawczyk@...wei.com>,
Song Liu <songliubraving@...com>,
Douglas Miller <dougmill@...ux.vnet.ibm.com>,
Yisen Zhuang <yisen.zhuang@...wei.com>,
"Michael S . Tsirkin" <mst@...hat.com>,
Jason Wang <jasowang@...hat.com>,
Harish Patil <harish.patil@...ium.com>,
Manish Chopra <manish.chopra@...ium.com>,
Netanel Belgazal <netanel@...zon.com>,
"Solarflare linux maintainers" <linux-net-drivers@...arflare.com>,
Thomas Falcon <tlfalcon@...ux.vnet.ibm.com>,
Edward Cree <ecree@...arflare.com>
Subject: Re: [PATCH net 10/11] sfc-falcon: remove ndo_poll_controller
On 27/09/18 17:32, Eric Dumazet wrote:
> As diagnosed by Song Liu, ndo_poll_controller() can
> be very dangerous on loaded hosts, since the cpu
> calling ndo_poll_controller() might steal all NAPI
> contexts (for all RX/TX queues of the NIC). This capture
> can last for unlimited amount of time, since one
> cpu is generally not able to drain all the queues under load.
>
> sfc-falcon uses NAPI for TX completions, so we better let core
> networking stack call the napi->poll() to avoid the capture.
>
> Signed-off-by: Eric Dumazet <edumazet@...gle.com>
> Cc: Solarflare linux maintainers <linux-net-drivers@...arflare.com>
> Cc: Edward Cree <ecree@...arflare.com>
> Cc: Bert Kenward <bkenward@...arflare.com>
Acked-By: Bert Kenward <bkenward@...arflare.com>
> ---
> drivers/net/ethernet/sfc/falcon/efx.c | 26 --------------------------
> 1 file changed, 26 deletions(-)
>
> diff --git a/drivers/net/ethernet/sfc/falcon/efx.c b/drivers/net/ethernet/sfc/falcon/efx.c
> index dd5530a4f8c8936868aed7171bd9481f93730d76..03e2455c502eacd9a4fd5c7fd320a9edcf265f77 100644
> --- a/drivers/net/ethernet/sfc/falcon/efx.c
> +++ b/drivers/net/ethernet/sfc/falcon/efx.c
> @@ -2052,29 +2052,6 @@ static void ef4_fini_napi(struct ef4_nic *efx)
> ef4_fini_napi_channel(channel);
> }
>
> -/**************************************************************************
> - *
> - * Kernel netpoll interface
> - *
> - *************************************************************************/
> -
> -#ifdef CONFIG_NET_POLL_CONTROLLER
> -
> -/* Although in the common case interrupts will be disabled, this is not
> - * guaranteed. However, all our work happens inside the NAPI callback,
> - * so no locking is required.
> - */
> -static void ef4_netpoll(struct net_device *net_dev)
> -{
> - struct ef4_nic *efx = netdev_priv(net_dev);
> - struct ef4_channel *channel;
> -
> - ef4_for_each_channel(channel, efx)
> - ef4_schedule_channel(channel);
> -}
> -
> -#endif
> -
> /**************************************************************************
> *
> * Kernel net device interface
> @@ -2250,9 +2227,6 @@ static const struct net_device_ops ef4_netdev_ops = {
> .ndo_set_mac_address = ef4_set_mac_address,
> .ndo_set_rx_mode = ef4_set_rx_mode,
> .ndo_set_features = ef4_set_features,
> -#ifdef CONFIG_NET_POLL_CONTROLLER
> - .ndo_poll_controller = ef4_netpoll,
> -#endif
> .ndo_setup_tc = ef4_setup_tc,
> #ifdef CONFIG_RFS_ACCEL
> .ndo_rx_flow_steer = ef4_filter_rfs,
>
Powered by blists - more mailing lists