[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <bb985b09-fe29-4fc9-9c06-310ff3abfad1@nvidia.com>
Date: Thu, 22 Jan 2026 16:06:40 +0100
From: Dragos Tatulea <dtatulea@...dia.com>
To: Jakub Kicinski <kuba@...nel.org>, davem@...emloft.net
Cc: netdev@...r.kernel.org, edumazet@...gle.com, pabeni@...hat.com,
andrew+netdev@...n.ch, horms@...nel.org, michael.chan@...adcom.com,
sdf@...ichev.me, almasrymina@...gle.com, asml.silence@...il.com,
dw@...idwei.uk, daniel@...earbox.net
Subject: Re: [PATCH net-next 4/6] net: use netdev_queue_config() for mp
restart
On 22.01.26 01:51, Jakub Kicinski wrote:
[...]
> @@ -154,8 +154,11 @@ int __net_mp_open_rxq(struct net_device *dev, unsigned int rxq_idx,
> }
> #endif
>
> + netdev_queue_config(dev, rxq_idx, &qcfg[0]);
> rxq->mp_params = *p;
> - ret = netdev_rx_queue_restart(dev, rxq_idx);
> + netdev_queue_config(dev, rxq_idx, &qcfg[1]);
> +
> + ret = netdev_rx_queue_reconfig(dev, rxq_idx, &qcfg[0], &qcfg[1]);
> if (ret)
> memset(&rxq->mp_params, 0, sizeof(rxq->mp_params));
>
> @@ -176,6 +179,7 @@ int net_mp_open_rxq(struct net_device *dev, unsigned int rxq_idx,
> void __net_mp_close_rxq(struct net_device *dev, unsigned int ifq_idx,
> const struct pp_memory_provider_params *old_p)
> {
> + struct netdev_queue_config qcfg[2];
> struct netdev_rx_queue *rxq;
> int err;
>
> @@ -195,8 +199,11 @@ void __net_mp_close_rxq(struct net_device *dev, unsigned int ifq_idx,
> rxq->mp_params.mp_priv != old_p->mp_priv))
> return;
>
> + netdev_queue_config(dev, ifq_idx, &qcfg[0]);
> memset(&rxq->mp_params, 0, sizeof(rxq->mp_params));
> - err = netdev_rx_queue_restart(dev, ifq_idx);
> + netdev_queue_config(dev, ifq_idx, &qcfg[1]);
> +
Is it ok to assume that on close we always resume to the default?
For now yes but maybe in the future we might want to save qcfg to the
state before mp_open.
With the very first rx-buf-len series it was possible to set a
rx-buf-len via YNL for a normal queue, switch to a MP queue and then
on MP queue close the configuration got switched to the default value
of rx-buf-len instead of what the user had configured. This was
not convenient.
Thanks,
Dragos
Powered by blists - more mailing lists