All of lore.kernel.org
 help / color / mirror / Atom feed
From: Tiwei Bie <tiwei.bie@intel.com>
To: Jason Wang <jasowang@redhat.com>
Cc: mst@redhat.com, virtualization@lists.linux-foundation.org,
	linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
	wexu@redhat.com, jfreimann@redhat.com
Subject: Re: [RFC v4 4/5] virtio_ring: add event idx support in packed ring
Date: Wed, 16 May 2018 20:58:25 +0800	[thread overview]
Message-ID: <20180516125825.GA4171@debian> (raw)
In-Reply-To: <64f5085b-4b62-d545-36e0-0951c11b29a8@redhat.com>

On Wed, May 16, 2018 at 08:17:21PM +0800, Jason Wang wrote:
> On 2018年05月16日 16:37, Tiwei Bie wrote:
[...]
> > @@ -1160,15 +1186,27 @@ static void virtqueue_disable_cb_packed(struct virtqueue *_vq)
> >   static unsigned virtqueue_enable_cb_prepare_packed(struct virtqueue *_vq)
> >   {
> >   	struct vring_virtqueue *vq = to_vvq(_vq);
> > +	u16 wrap_counter;
> >   	START_USE(vq);
> >   	/* We optimistically turn back on interrupts, then check if there was
> >   	 * more to do. */
> > +	/* Depending on the VIRTIO_RING_F_USED_EVENT_IDX feature, we need to
> > +	 * either clear the flags bit or point the event index at the next
> > +	 * entry. Always update the event index to keep code simple. */
> > +
> > +	wrap_counter = vq->wrap_counter;
> > +	if (vq->last_used_idx > vq->next_avail_idx)
> 
> Should this be ">=" consider rx refill may try to completely fill the ring?

It seems that there are two cases that last_used_idx
equals to next_avail_idx. The first one is that the
ring is empty. And the second one is that the ring
is full. Although in the first case, most probably,
the driver won't enable the interrupt.

Maybe I really should track the used_wrap_counter
instead of calculating it each time I need it.. I'll
give it a try..

> 
> > +		wrap_counter ^= 1;
> > +
> > +	vq->vring_packed.driver->off_wrap = cpu_to_virtio16(_vq->vdev,
> > +			vq->last_used_idx | (wrap_counter << 15));
> >   	if (vq->event_flags_shadow == VRING_EVENT_F_DISABLE) {
> >   		virtio_wmb(vq->weak_barriers);
> > -		vq->event_flags_shadow = VRING_EVENT_F_ENABLE;
> > +		vq->event_flags_shadow = vq->event ? VRING_EVENT_F_DESC :
> > +						     VRING_EVENT_F_ENABLE;
> >   		vq->vring_packed.driver->flags = cpu_to_virtio16(_vq->vdev,
> >   							vq->event_flags_shadow);
> >   	}
> > @@ -1194,15 +1232,40 @@ static bool virtqueue_poll_packed(struct virtqueue *_vq, unsigned last_used_idx)
> >   static bool virtqueue_enable_cb_delayed_packed(struct virtqueue *_vq)
> >   {
> >   	struct vring_virtqueue *vq = to_vvq(_vq);
> > +	u16 bufs, used_idx, wrap_counter;
> >   	START_USE(vq);
> >   	/* We optimistically turn back on interrupts, then check if there was
> >   	 * more to do. */
> > +	/* Depending on the VIRTIO_RING_F_USED_EVENT_IDX feature, we need to
> > +	 * either clear the flags bit or point the event index at the next
> > +	 * entry. Always update the event index to keep code simple. */
> > +
> > +	/* TODO: tune this threshold */
> > +	if (vq->next_avail_idx < vq->last_used_idx)
> > +		bufs = (vq->vring_packed.num + vq->next_avail_idx -
> > +				vq->last_used_idx) * 3 / 4;
> > +	else
> > +		bufs = (vq->next_avail_idx - vq->last_used_idx) * 3 / 4;
> > +
> > +	wrap_counter = vq->wrap_counter;
> > +	if (vq->last_used_idx > vq->next_avail_idx)
> > +		wrap_counter ^= 1;
> > +
> > +	used_idx = vq->last_used_idx + bufs;
> > +	if (used_idx >= vq->vring_packed.num) {
> > +		used_idx -= vq->vring_packed.num;
> > +		wrap_counter ^= 1;
> > +	}
> 
> Looks correct but maybe it's better to add some comments for such logic.

Make sense.

> 
> > +
> > +	vq->vring_packed.driver->off_wrap = cpu_to_virtio16(_vq->vdev,
> > +			used_idx | (wrap_counter << 15));
> >   	if (vq->event_flags_shadow == VRING_EVENT_F_DISABLE) {
> >   		virtio_wmb(vq->weak_barriers);
> > -		vq->event_flags_shadow = VRING_EVENT_F_ENABLE;
> > +		vq->event_flags_shadow = vq->event ? VRING_EVENT_F_DESC :
> > +						     VRING_EVENT_F_ENABLE;
> >   		vq->vring_packed.driver->flags = cpu_to_virtio16(_vq->vdev,
> >   							vq->event_flags_shadow);
> >   	}
> > @@ -1869,8 +1932,10 @@ void vring_transport_features(struct virtio_device *vdev)
> >   		switch (i) {
> >   		case VIRTIO_RING_F_INDIRECT_DESC:
> >   			break;
> > +#if 0
> >   		case VIRTIO_RING_F_EVENT_IDX:
> >   			break;
> > +#endif
> 
> Maybe it's time to remove this #if 0.

Will do it.

Thanks for the review!

Best regards,
Tiwei Bie

  reply	other threads:[~2018-05-16 12:58 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-05-16  8:37 [RFC v4 0/5] virtio: support packed ring Tiwei Bie
2018-05-16  8:37 ` [RFC v4 1/5] virtio: add packed ring definitions Tiwei Bie
2018-05-16  8:37   ` Tiwei Bie
2018-05-16  8:37 ` [RFC v4 2/5] virtio_ring: support creating packed ring Tiwei Bie
2018-05-16  8:37 ` Tiwei Bie
2018-05-16  8:37 ` [RFC v4 3/5] virtio_ring: add packed ring support Tiwei Bie
2018-05-16  8:37 ` Tiwei Bie
2018-05-16 11:50   ` Jason Wang
2018-05-16 12:39     ` Tiwei Bie
2018-05-16 12:51       ` Jason Wang
2018-05-16 13:45         ` Tiwei Bie
2018-05-16 13:45         ` Tiwei Bie
2018-05-16 14:05           ` Jason Wang
2018-05-16 14:05           ` Jason Wang
2018-05-16 14:33             ` Tiwei Bie
2018-05-17 12:01               ` Jason Wang
2018-05-17 12:01               ` Jason Wang
2018-05-18 11:29                 ` Tiwei Bie
2018-05-18 13:17                   ` Jason Wang
2018-05-18 13:17                   ` Jason Wang
2018-05-18 14:33                     ` Tiwei Bie
2018-05-19  1:12                       ` Jason Wang
2018-05-19  2:29                         ` Tiwei Bie
2018-05-21  2:30                           ` Jason Wang
2018-05-21  2:39                             ` Tiwei Bie
2018-05-21  2:39                             ` Tiwei Bie
2018-05-21  2:30                           ` Jason Wang
2018-05-19  2:29                         ` Tiwei Bie
2018-05-19  1:12                       ` Jason Wang
2018-05-18 14:33                     ` Tiwei Bie
2018-05-18 11:29                 ` Tiwei Bie
2018-05-16 14:33             ` Tiwei Bie
2018-05-16 12:51       ` Jason Wang
2018-05-16 12:39     ` Tiwei Bie
2018-05-16 11:50   ` Jason Wang
2018-05-16  8:37 ` [RFC v4 4/5] virtio_ring: add event idx support in packed ring Tiwei Bie
2018-05-16  8:37   ` Tiwei Bie
2018-05-16 12:17   ` Jason Wang
2018-05-16 12:58     ` Tiwei Bie [this message]
2018-05-16 13:31       ` Jason Wang
2018-05-16 13:31       ` Jason Wang
2018-05-16 12:58     ` Tiwei Bie
2018-05-16 12:17   ` Jason Wang
2018-05-16  8:37 ` [RFC v4 5/5] virtio_ring: enable " Tiwei Bie
2018-05-16  8:37 ` Tiwei Bie
2018-05-16 10:15   ` Sergei Shtylyov
2018-05-16 10:15     ` Sergei Shtylyov
2018-05-16 10:21     ` Tiwei Bie
2018-05-16 11:42       ` Sergei Shtylyov
2018-05-16 11:42         ` Sergei Shtylyov
2018-05-16 12:26         ` Tiwei Bie
2018-05-16 12:26         ` Tiwei Bie
2018-05-16 10:21     ` Tiwei Bie

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=20180516125825.GA4171@debian \
    --to=tiwei.bie@intel.com \
    --cc=jasowang@redhat.com \
    --cc=jfreimann@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=netdev@vger.kernel.org \
    --cc=virtualization@lists.linux-foundation.org \
    --cc=wexu@redhat.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.