Virtualization - Jul 2023 - [PATCH net-next V4 2/3] virtio_net: support per queue interrupt coalesce command

On Tue, 2023-07-25 at 16:07 +0300, Gavin Li wrote:
> Add interrupt_coalesce config in send_queue and receive_queue to cache user
> config.
> 
> Send per virtqueue interrupt moderation config to underlying device in
> order to have more efficient interrupt moderation and cpu utilization of
> guest VM.
> 
> Additionally, address all the VQs when updating the global configuration,
> as now the individual VQs configuration can diverge from the global
> configuration.
> 
> Signed-off-by: Gavin Li <gavinl at nvidia.com>
> Reviewed-by: Dragos Tatulea <dtatulea at nvidia.com>
> Reviewed-by: Jiri Pirko <jiri at nvidia.com>
> Acked-by: Michael S. Tsirkin <mst at redhat.com>
FTR, this patch is significantly different from the version previously
acked/reviewed, I'm unsure if all the reviewers are ok with the new
one.

[...]
>  static int virtnet_set_coalesce(struct net_device *dev,
>  				struct ethtool_coalesce *ec,
>  				struct kernel_ethtool_coalesce *kernel_coal,
>  				struct netlink_ext_ack *extack)
>  {
>  	struct virtnet_info *vi = netdev_priv(dev);
> -	int ret, i, napi_weight;
> +	int ret, queue_number, napi_weight;
>  	bool update_napi = false;
>  
>  	/* Can't change NAPI weight if the link is up */
>  	napi_weight = ec->tx_max_coalesced_frames ? NAPI_POLL_WEIGHT : 0;
> -	if (napi_weight ^ vi->sq[0].napi.weight) {
> -		if (dev->flags & IFF_UP)
> -			return -EBUSY;
> -		else
> -			update_napi = true;
> +	for (queue_number = 0; queue_number < vi->max_queue_pairs;
queue_number++) {
> +		ret = virtnet_should_update_vq_weight(dev->flags, napi_weight,
> +						      vi->sq[queue_number].napi.weight,
> +						      &update_napi);
> +		if (ret)
> +			return ret;
> +
> +		if (update_napi) {
> +			/* All queues that belong to [queue_number, queue_count] will be
> +			 * updated for the sake of simplicity, which might not be necessary
It looks like the comment above still refers to the old code. Should
be:
	[queue_number, vi->max_queue_pairs]
			
Otherwise LGTM, thanks!

Paolo

On Thu, Jul 27, 2023 at 9:28?PM Paolo Abeni <pabeni at redhat.com>
wrote:
>
> On Tue, 2023-07-25 at 16:07 +0300, Gavin Li wrote:
> > Add interrupt_coalesce config in send_queue and receive_queue to cache
user
> > config.
> >
> > Send per virtqueue interrupt moderation config to underlying device in
> > order to have more efficient interrupt moderation and cpu utilization
of
> > guest VM.
> >
> > Additionally, address all the VQs when updating the global
configuration,
> > as now the individual VQs configuration can diverge from the global
> > configuration.
> >
> > Signed-off-by: Gavin Li <gavinl at nvidia.com>
> > Reviewed-by: Dragos Tatulea <dtatulea at nvidia.com>
> > Reviewed-by: Jiri Pirko <jiri at nvidia.com>
> > Acked-by: Michael S. Tsirkin <mst at redhat.com>
>
> FTR, this patch is significantly different from the version previously
> acked/reviewed, I'm unsure if all the reviewers are ok with the new
> one.
Good point, and I plan to review this no later than next Monday and
offer my ack if necessary. Please hold this series now.

Thanks
>
> [...]
>
> >  static int virtnet_set_coalesce(struct net_device *dev,
> >                               struct ethtool_coalesce *ec,
> >                               struct kernel_ethtool_coalesce
*kernel_coal,
> >                               struct netlink_ext_ack *extack)
> >  {
> >       struct virtnet_info *vi = netdev_priv(dev);
> > -     int ret, i, napi_weight;
> > +     int ret, queue_number, napi_weight;
> >       bool update_napi = false;
> >
> >       /* Can't change NAPI weight if the link is up */
> >       napi_weight = ec->tx_max_coalesced_frames ? NAPI_POLL_WEIGHT
: 0;
> > -     if (napi_weight ^ vi->sq[0].napi.weight) {
> > -             if (dev->flags & IFF_UP)
> > -                     return -EBUSY;
> > -             else
> > -                     update_napi = true;
> > +     for (queue_number = 0; queue_number < vi->max_queue_pairs;
queue_number++) {
> > +             ret = virtnet_should_update_vq_weight(dev->flags,
napi_weight,
> > +                                                  
vi->sq[queue_number].napi.weight,
> > +                                                   &update_napi);
> > +             if (ret)
> > +                     return ret;
> > +
> > +             if (update_napi) {
> > +                     /* All queues that belong to [queue_number,
queue_count] will be
> > +                      * updated for the sake of simplicity, which
might not be necessary
>
> It looks like the comment above still refers to the old code. Should
> be:
>         [queue_number, vi->max_queue_pairs]
>
> Otherwise LGTM, thanks!
>
> Paolo
>

On Thu, Jul 27, 2023 at 03:28:32PM +0200, Paolo Abeni
wrote:
> On Tue, 2023-07-25 at 16:07 +0300, Gavin Li wrote:
> > Add interrupt_coalesce config in send_queue and receive_queue to cache
user
> > config.
> > 
> > Send per virtqueue interrupt moderation config to underlying device in
> > order to have more efficient interrupt moderation and cpu utilization
of
> > guest VM.
> > 
> > Additionally, address all the VQs when updating the global
configuration,
> > as now the individual VQs configuration can diverge from the global
> > configuration.
> > 
> > Signed-off-by: Gavin Li <gavinl at nvidia.com>
> > Reviewed-by: Dragos Tatulea <dtatulea at nvidia.com>
> > Reviewed-by: Jiri Pirko <jiri at nvidia.com>
> > Acked-by: Michael S. Tsirkin <mst at redhat.com>
> 
> FTR, this patch is significantly different from the version previously
> acked/reviewed, I'm unsure if all the reviewers are ok with the new
> one.
> 
> [...]
still ok by me

Acked-by: Michael S. Tsirkin <mst at redhat.com>

let's wait for Jason too.
> >  static int virtnet_set_coalesce(struct net_device *dev,
> >  				struct ethtool_coalesce *ec,
> >  				struct kernel_ethtool_coalesce *kernel_coal,
> >  				struct netlink_ext_ack *extack)
> >  {
> >  	struct virtnet_info *vi = netdev_priv(dev);
> > -	int ret, i, napi_weight;
> > +	int ret, queue_number, napi_weight;
> >  	bool update_napi = false;
> >  
> >  	/* Can't change NAPI weight if the link is up */
> >  	napi_weight = ec->tx_max_coalesced_frames ? NAPI_POLL_WEIGHT : 0;
> > -	if (napi_weight ^ vi->sq[0].napi.weight) {
> > -		if (dev->flags & IFF_UP)
> > -			return -EBUSY;
> > -		else
> > -			update_napi = true;
> > +	for (queue_number = 0; queue_number < vi->max_queue_pairs;
queue_number++) {
> > +		ret = virtnet_should_update_vq_weight(dev->flags, napi_weight,
> > +						      vi->sq[queue_number].napi.weight,
> > +						      &update_napi);
> > +		if (ret)
> > +			return ret;
> > +
> > +		if (update_napi) {
> > +			/* All queues that belong to [queue_number, queue_count] will be
> > +			 * updated for the sake of simplicity, which might not be
necessary
> 
> It looks like the comment above still refers to the old code. Should
> be:
> 	[queue_number, vi->max_queue_pairs]
> 			
> Otherwise LGTM, thanks!
> 
> Paolo