All of lore.kernel.org
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: "Michael S. Tsirkin" <mst@redhat.com>, linux-kernel@vger.kernel.org
Cc: Jakub Kicinski <kuba@kernel.org>, Wei Wang <weiwan@google.com>,
	David Miller <davem@davemloft.net>,
	netdev@vger.kernel.org, Willem de Bruijn <willemb@google.com>,
	virtualization@lists.linux-foundation.org,
	Jason Wang <jasowang@redhat.com>
Subject: Re: [PATCH v3 4/4] virtio_net: disable cb aggressively
Date: Wed, 26 May 2021 17:15:08 +0200	[thread overview]
Message-ID: <18b47c22-8c8a-7699-ffaf-ccfdcbf39d37@gmail.com> (raw)
In-Reply-To: <20210526082423.47837-5-mst@redhat.com>



On 5/26/21 10:24 AM, Michael S. Tsirkin wrote:
> There are currently two cases where we poll TX vq not in response to a
> callback: start xmit and rx napi.  We currently do this with callbacks
> enabled which can cause extra interrupts from the card.  Used not to be
> a big issue as we run with interrupts disabled but that is no longer the
> case, and in some cases the rate of spurious interrupts is so high
> linux detects this and actually kills the interrupt.
> 
> Fix up by disabling the callbacks before polling the tx vq.


It is not clear why we want to poll TX completions from ndo_start_xmit() in napi mode ?

This seems not needed, adding costs to sender thread, this might
reduce the ability to use a different cpu for tx completions.

Also this will likely conflict with BQL model if we want to use BQL at some point.

> 

This probably needs a Fixes: tag 

> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> ---
>  drivers/net/virtio_net.c | 16 ++++++++++++----
>  1 file changed, 12 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index c29f42d1e04f..a83dc038d8af 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -1433,7 +1433,10 @@ static void virtnet_poll_cleantx(struct receive_queue *rq)
>  		return;
>  
>  	if (__netif_tx_trylock(txq)) {
> -		free_old_xmit_skbs(sq, true);
> +		do {
> +			virtqueue_disable_cb(sq->vq);
> +			free_old_xmit_skbs(sq, true);
> +		} while (unlikely(!virtqueue_enable_cb_delayed(sq->vq)));
>  
>  		if (sq->vq->num_free >= 2 + MAX_SKB_FRAGS)
>  			netif_tx_wake_queue(txq);
> @@ -1605,12 +1608,17 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev)
>  	struct netdev_queue *txq = netdev_get_tx_queue(dev, qnum);
>  	bool kick = !netdev_xmit_more();
>  	bool use_napi = sq->napi.weight;
> +	unsigned int bytes = skb->len;
>  
>  	/* Free up any pending old buffers before queueing new ones. */
> -	free_old_xmit_skbs(sq, false);
> +	do {
> +		if (use_napi)
> +			virtqueue_disable_cb(sq->vq);
>  
> -	if (use_napi && kick)
> -		virtqueue_enable_cb_delayed(sq->vq);
> +		free_old_xmit_skbs(sq, false);
> +
> +	} while (use_napi && kick &&
> +	       unlikely(!virtqueue_enable_cb_delayed(sq->vq)));
>  
>  	/* timestamp packet in software */
>  	skb_tx_timestamp(skb);
> 

WARNING: multiple messages have this Message-ID (diff)
From: Eric Dumazet <eric.dumazet@gmail.com>
To: "Michael S. Tsirkin" <mst@redhat.com>, linux-kernel@vger.kernel.org
Cc: Willem de Bruijn <willemb@google.com>,
	netdev@vger.kernel.org,
	virtualization@lists.linux-foundation.org,
	Jakub Kicinski <kuba@kernel.org>, Wei Wang <weiwan@google.com>,
	David Miller <davem@davemloft.net>
Subject: Re: [PATCH v3 4/4] virtio_net: disable cb aggressively
Date: Wed, 26 May 2021 17:15:08 +0200	[thread overview]
Message-ID: <18b47c22-8c8a-7699-ffaf-ccfdcbf39d37@gmail.com> (raw)
In-Reply-To: <20210526082423.47837-5-mst@redhat.com>



On 5/26/21 10:24 AM, Michael S. Tsirkin wrote:
> There are currently two cases where we poll TX vq not in response to a
> callback: start xmit and rx napi.  We currently do this with callbacks
> enabled which can cause extra interrupts from the card.  Used not to be
> a big issue as we run with interrupts disabled but that is no longer the
> case, and in some cases the rate of spurious interrupts is so high
> linux detects this and actually kills the interrupt.
> 
> Fix up by disabling the callbacks before polling the tx vq.


It is not clear why we want to poll TX completions from ndo_start_xmit() in napi mode ?

This seems not needed, adding costs to sender thread, this might
reduce the ability to use a different cpu for tx completions.

Also this will likely conflict with BQL model if we want to use BQL at some point.

> 

This probably needs a Fixes: tag 

> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> ---
>  drivers/net/virtio_net.c | 16 ++++++++++++----
>  1 file changed, 12 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index c29f42d1e04f..a83dc038d8af 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -1433,7 +1433,10 @@ static void virtnet_poll_cleantx(struct receive_queue *rq)
>  		return;
>  
>  	if (__netif_tx_trylock(txq)) {
> -		free_old_xmit_skbs(sq, true);
> +		do {
> +			virtqueue_disable_cb(sq->vq);
> +			free_old_xmit_skbs(sq, true);
> +		} while (unlikely(!virtqueue_enable_cb_delayed(sq->vq)));
>  
>  		if (sq->vq->num_free >= 2 + MAX_SKB_FRAGS)
>  			netif_tx_wake_queue(txq);
> @@ -1605,12 +1608,17 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev)
>  	struct netdev_queue *txq = netdev_get_tx_queue(dev, qnum);
>  	bool kick = !netdev_xmit_more();
>  	bool use_napi = sq->napi.weight;
> +	unsigned int bytes = skb->len;
>  
>  	/* Free up any pending old buffers before queueing new ones. */
> -	free_old_xmit_skbs(sq, false);
> +	do {
> +		if (use_napi)
> +			virtqueue_disable_cb(sq->vq);
>  
> -	if (use_napi && kick)
> -		virtqueue_enable_cb_delayed(sq->vq);
> +		free_old_xmit_skbs(sq, false);
> +
> +	} while (use_napi && kick &&
> +	       unlikely(!virtqueue_enable_cb_delayed(sq->vq)));
>  
>  	/* timestamp packet in software */
>  	skb_tx_timestamp(skb);
> 
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

  reply	other threads:[~2021-05-26 15:15 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-26  8:24 [PATCH v3 0/4] virtio net: spurious interrupt related fixes Michael S. Tsirkin
2021-05-26  8:24 ` Michael S. Tsirkin
2021-05-26  8:24 ` [PATCH v3 1/4] virtio_net: move tx vq operation under tx queue lock Michael S. Tsirkin
2021-05-26  8:24   ` Michael S. Tsirkin
2021-05-27  3:41   ` Jason Wang
2021-05-27  3:41     ` Jason Wang
2021-05-28 22:25     ` Willem de Bruijn
2021-05-28 22:25       ` Willem de Bruijn
2021-06-09 22:03       ` Michael S. Tsirkin
2021-06-09 22:03         ` Michael S. Tsirkin
2021-05-26  8:24 ` [PATCH v3 2/4] virtio_net: move txq wakeups under tx q lock Michael S. Tsirkin
2021-05-26  8:24   ` Michael S. Tsirkin
2021-05-27  3:48   ` Jason Wang
2021-05-27  3:48     ` Jason Wang
2021-05-26  8:24 ` [PATCH v3 3/4] virtio: fix up virtio_disable_cb Michael S. Tsirkin
2021-05-26  8:24   ` Michael S. Tsirkin
2021-05-27  4:01   ` Jason Wang
2021-05-27  4:01     ` Jason Wang
2023-03-30  6:07   ` Xuan Zhuo
2023-03-30  6:07     ` Xuan Zhuo
2023-03-30  6:44     ` Michael S. Tsirkin
2023-03-30  6:44       ` Michael S. Tsirkin
2023-03-30  6:54       ` Xuan Zhuo
2023-03-30  6:54         ` Xuan Zhuo
2023-03-30 14:04         ` Michael S. Tsirkin
2023-03-30 14:04           ` Michael S. Tsirkin
2023-03-31  3:38           ` Xuan Zhuo
2023-03-31  3:38             ` Xuan Zhuo
2021-05-26  8:24 ` [PATCH v3 4/4] virtio_net: disable cb aggressively Michael S. Tsirkin
2021-05-26  8:24   ` Michael S. Tsirkin
2021-05-26 15:15   ` Eric Dumazet [this message]
2021-05-26 15:15     ` Eric Dumazet
2021-05-26 21:22     ` Willem de Bruijn
2021-05-26 21:22       ` Willem de Bruijn
2021-05-26 19:39   ` Jakub Kicinski
2021-05-27  4:09   ` Jason Wang
2021-05-27  4:09     ` Jason Wang
2023-01-16 13:41   ` Laurent Vivier
2023-01-16 13:41     ` Laurent Vivier
2023-01-17  3:48     ` Jason Wang
2023-01-17  3:48       ` Jason Wang
2021-05-26 15:34 ` [PATCH v3 0/4] virtio net: spurious interrupt related fixes Willem de Bruijn
2021-05-26 15:34   ` Willem de Bruijn
2021-06-01  2:53   ` Willem de Bruijn
2021-06-01  2:53     ` Willem de Bruijn
2021-06-09 21:36     ` Willem de Bruijn
2021-06-09 21:36       ` Willem de Bruijn
2021-06-09 22:59       ` Willem de Bruijn
2021-06-09 22:59         ` Willem de Bruijn

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=18b47c22-8c8a-7699-ffaf-ccfdcbf39d37@gmail.com \
    --to=eric.dumazet@gmail.com \
    --cc=davem@davemloft.net \
    --cc=jasowang@redhat.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=netdev@vger.kernel.org \
    --cc=virtualization@lists.linux-foundation.org \
    --cc=weiwan@google.com \
    --cc=willemb@google.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.