From: Jason Wang <jasowang@redhat.com> To: Tiwei Bie <tiwei.bie@intel.com>, "Michael S. Tsirkin" <mst@redhat.com> Cc: virtualization@lists.linux-foundation.org, linux-kernel@vger.kernel.org, netdev@vger.kernel.org, virtio-dev@lists.oasis-open.org, wexu@redhat.com, jfreimann@redhat.com Subject: Re: [PATCH net-next v2 3/5] virtio_ring: add packed ring support Date: Thu, 8 Nov 2018 16:18:25 +0800 [thread overview] Message-ID: <2d46a41e-bc00-276a-e19a-105c9dffc75a@redhat.com> (raw) In-Reply-To: <20181108013759.GA20591@debian> On 2018/11/8 上午9:38, Tiwei Bie wrote: >>> + >>> + if (vq->vq.num_free < descs_used) { >>> + pr_debug("Can't add buf len %i - avail = %i\n", >>> + descs_used, vq->vq.num_free); >>> + /* FIXME: for historical reasons, we force a notify here if >>> + * there are outgoing parts to the buffer. Presumably the >>> + * host should service the ring ASAP. */ >> I don't think we have a reason to do this for packed ring. >> No historical baggage there, right? > Based on the original commit log, it seems that the notify here > is just an "optimization". But I don't quite understand what does > the "the heuristics which KVM uses" refer to. If it's safe to drop > this in packed ring, I'd like to do it. According to the commit log, it seems like a workaround of lguest networking backend. I agree to drop it, we should not have such burden. But we should notice that, with this removed, the compare between packed vs split is kind of unfair. Consider the removal of lguest support recently, maybe we can drop this for split ring as well? Thanks > > commit 44653eae1407f79dff6f52fcf594ae84cb165ec4 > Author: Rusty Russell<rusty@rustcorp.com.au> > Date: Fri Jul 25 12:06:04 2008 -0500 > > virtio: don't always force a notification when ring is full > > We force notification when the ring is full, even if the host has > indicated it doesn't want to know. This seemed like a good idea at > the time: if we fill the transmit ring, we should tell the host > immediately. > > Unfortunately this logic also applies to the receiving ring, which is > refilled constantly. We should introduce real notification thesholds > to replace this logic. Meanwhile, removing the logic altogether breaks > the heuristics which KVM uses, so we use a hack: only notify if there are > outgoing parts of the new buffer. > > Here are the number of exits with lguest's crappy network implementation: > Before: > network xmit 7859051 recv 236420 > After: > network xmit 7858610 recv 118136 > > Signed-off-by: Rusty Russell<rusty@rustcorp.com.au> > > diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c > index 72bf8bc09014..21d9a62767af 100644 > --- a/drivers/virtio/virtio_ring.c > +++ b/drivers/virtio/virtio_ring.c > @@ -87,8 +87,11 @@ static int vring_add_buf(struct virtqueue *_vq, > if (vq->num_free < out + in) { > pr_debug("Can't add buf len %i - avail = %i\n", > out + in, vq->num_free); > - /* We notify*even if* VRING_USED_F_NO_NOTIFY is set here. */ > - vq->notify(&vq->vq); > + /* FIXME: for historical reasons, we force a notify here if > + * there are outgoing parts to the buffer. Presumably the > + * host should service the ring ASAP. */ > + if (out) > + vq->notify(&vq->vq); > END_USE(vq); > return -ENOSPC; > } > >
WARNING: multiple messages have this Message-ID (diff)
From: Jason Wang <jasowang@redhat.com> To: Tiwei Bie <tiwei.bie@intel.com>, "Michael S. Tsirkin" <mst@redhat.com> Cc: virtualization@lists.linux-foundation.org, linux-kernel@vger.kernel.org, netdev@vger.kernel.org, virtio-dev@lists.oasis-open.org, wexu@redhat.com, jfreimann@redhat.com Subject: [virtio-dev] Re: [PATCH net-next v2 3/5] virtio_ring: add packed ring support Date: Thu, 8 Nov 2018 16:18:25 +0800 [thread overview] Message-ID: <2d46a41e-bc00-276a-e19a-105c9dffc75a@redhat.com> (raw) In-Reply-To: <20181108013759.GA20591@debian> On 2018/11/8 上午9:38, Tiwei Bie wrote: >>> + >>> + if (vq->vq.num_free < descs_used) { >>> + pr_debug("Can't add buf len %i - avail = %i\n", >>> + descs_used, vq->vq.num_free); >>> + /* FIXME: for historical reasons, we force a notify here if >>> + * there are outgoing parts to the buffer. Presumably the >>> + * host should service the ring ASAP. */ >> I don't think we have a reason to do this for packed ring. >> No historical baggage there, right? > Based on the original commit log, it seems that the notify here > is just an "optimization". But I don't quite understand what does > the "the heuristics which KVM uses" refer to. If it's safe to drop > this in packed ring, I'd like to do it. According to the commit log, it seems like a workaround of lguest networking backend. I agree to drop it, we should not have such burden. But we should notice that, with this removed, the compare between packed vs split is kind of unfair. Consider the removal of lguest support recently, maybe we can drop this for split ring as well? Thanks > > commit 44653eae1407f79dff6f52fcf594ae84cb165ec4 > Author: Rusty Russell<rusty@rustcorp.com.au> > Date: Fri Jul 25 12:06:04 2008 -0500 > > virtio: don't always force a notification when ring is full > > We force notification when the ring is full, even if the host has > indicated it doesn't want to know. This seemed like a good idea at > the time: if we fill the transmit ring, we should tell the host > immediately. > > Unfortunately this logic also applies to the receiving ring, which is > refilled constantly. We should introduce real notification thesholds > to replace this logic. Meanwhile, removing the logic altogether breaks > the heuristics which KVM uses, so we use a hack: only notify if there are > outgoing parts of the new buffer. > > Here are the number of exits with lguest's crappy network implementation: > Before: > network xmit 7859051 recv 236420 > After: > network xmit 7858610 recv 118136 > > Signed-off-by: Rusty Russell<rusty@rustcorp.com.au> > > diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c > index 72bf8bc09014..21d9a62767af 100644 > --- a/drivers/virtio/virtio_ring.c > +++ b/drivers/virtio/virtio_ring.c > @@ -87,8 +87,11 @@ static int vring_add_buf(struct virtqueue *_vq, > if (vq->num_free < out + in) { > pr_debug("Can't add buf len %i - avail = %i\n", > out + in, vq->num_free); > - /* We notify*even if* VRING_USED_F_NO_NOTIFY is set here. */ > - vq->notify(&vq->vq); > + /* FIXME: for historical reasons, we force a notify here if > + * there are outgoing parts to the buffer. Presumably the > + * host should service the ring ASAP. */ > + if (out) > + vq->notify(&vq->vq); > END_USE(vq); > return -ENOSPC; > } > > --------------------------------------------------------------------- To unsubscribe, e-mail: virtio-dev-unsubscribe@lists.oasis-open.org For additional commands, e-mail: virtio-dev-help@lists.oasis-open.org
next prev parent reply other threads:[~2018-11-08 8:18 UTC|newest] Thread overview: 167+ messages / expand[flat|nested] mbox.gz Atom feed top 2018-07-11 2:27 [PATCH net-next v2 0/5] virtio: support packed ring Tiwei Bie 2018-07-11 2:27 ` [virtio-dev] " Tiwei Bie 2018-07-11 2:27 ` Tiwei Bie 2018-07-11 2:27 ` [PATCH net-next v2 1/5] virtio: add packed ring definitions Tiwei Bie 2018-07-11 2:27 ` [virtio-dev] " Tiwei Bie 2018-09-07 13:51 ` Michael S. Tsirkin 2018-09-07 13:51 ` [virtio-dev] " Michael S. Tsirkin 2018-09-10 2:13 ` Tiwei Bie 2018-09-10 2:13 ` [virtio-dev] " Tiwei Bie 2018-09-12 12:53 ` Michael S. Tsirkin 2018-09-12 12:53 ` Michael S. Tsirkin 2018-09-12 12:53 ` [virtio-dev] " Michael S. Tsirkin 2018-09-10 2:13 ` Tiwei Bie 2018-09-07 13:51 ` Michael S. Tsirkin 2018-07-11 2:27 ` Tiwei Bie 2018-07-11 2:27 ` [PATCH net-next v2 2/5] virtio_ring: support creating packed ring Tiwei Bie 2018-07-11 2:27 ` [virtio-dev] " Tiwei Bie 2018-09-07 14:03 ` Michael S. Tsirkin 2018-09-07 14:03 ` Michael S. Tsirkin 2018-09-07 14:03 ` [virtio-dev] " Michael S. Tsirkin 2018-09-10 2:28 ` Tiwei Bie 2018-09-10 2:28 ` [virtio-dev] " Tiwei Bie 2018-09-12 7:51 ` Tiwei Bie 2018-09-12 7:51 ` Tiwei Bie 2018-09-12 7:51 ` [virtio-dev] " Tiwei Bie 2018-09-12 16:12 ` Michael S. Tsirkin 2018-09-12 16:12 ` Michael S. Tsirkin 2018-09-12 16:12 ` [virtio-dev] " Michael S. Tsirkin 2018-09-12 12:45 ` Michael S. Tsirkin 2018-09-12 12:45 ` [virtio-dev] " Michael S. Tsirkin 2018-09-12 12:45 ` Michael S. Tsirkin 2018-09-10 2:28 ` Tiwei Bie 2018-07-11 2:27 ` Tiwei Bie 2018-07-11 2:27 ` [PATCH net-next v2 3/5] virtio_ring: add packed ring support Tiwei Bie 2018-07-11 2:27 ` [virtio-dev] " Tiwei Bie 2018-09-07 13:49 ` Michael S. Tsirkin 2018-09-07 13:49 ` Michael S. Tsirkin 2018-09-07 13:49 ` [virtio-dev] " Michael S. Tsirkin 2018-09-10 2:03 ` Tiwei Bie 2018-09-10 2:03 ` [virtio-dev] " Tiwei Bie 2018-09-10 2:03 ` Tiwei Bie 2018-09-12 16:22 ` Michael S. Tsirkin 2018-09-12 16:22 ` [virtio-dev] " Michael S. Tsirkin 2018-09-12 16:22 ` Michael S. Tsirkin 2018-11-07 17:48 ` Michael S. Tsirkin 2018-11-07 17:48 ` [virtio-dev] " Michael S. Tsirkin 2018-11-08 1:38 ` Tiwei Bie 2018-11-08 1:38 ` Tiwei Bie 2018-11-08 1:38 ` [virtio-dev] " Tiwei Bie 2018-11-08 8:18 ` Jason Wang 2018-11-08 8:18 ` Jason Wang [this message] 2018-11-08 8:18 ` [virtio-dev] " Jason Wang 2018-11-08 11:51 ` Tiwei Bie 2018-11-08 11:51 ` [virtio-dev] " Tiwei Bie 2018-11-08 15:56 ` Michael S. Tsirkin 2018-11-08 15:56 ` Michael S. Tsirkin 2018-11-08 15:56 ` [virtio-dev] " Michael S. Tsirkin 2018-11-09 1:50 ` Tiwei Bie 2018-11-09 1:50 ` Tiwei Bie 2018-11-09 1:50 ` [virtio-dev] " Tiwei Bie 2018-11-09 2:30 ` Jason Wang 2018-11-09 2:30 ` Jason Wang 2018-11-09 2:30 ` [virtio-dev] " Jason Wang 2018-11-09 4:00 ` Michael S. Tsirkin 2018-11-09 4:00 ` [virtio-dev] " Michael S. Tsirkin 2018-11-09 10:05 ` Jason Wang 2018-11-09 10:05 ` [virtio-dev] " Jason Wang 2018-11-09 10:05 ` Jason Wang 2018-11-09 4:00 ` Michael S. Tsirkin 2018-11-08 11:51 ` Tiwei Bie 2018-11-08 14:14 ` Michael S. Tsirkin 2018-11-08 14:14 ` Michael S. Tsirkin 2018-11-08 14:14 ` [virtio-dev] " Michael S. Tsirkin 2018-11-09 2:25 ` Jason Wang 2018-11-09 2:25 ` [virtio-dev] " Jason Wang 2018-11-09 3:58 ` Michael S. Tsirkin 2018-11-09 3:58 ` Michael S. Tsirkin 2018-11-09 3:58 ` [virtio-dev] " Michael S. Tsirkin 2018-11-09 10:04 ` Jason Wang 2018-11-09 10:04 ` [virtio-dev] " Jason Wang 2018-11-09 10:04 ` Jason Wang 2018-11-09 2:25 ` Jason Wang 2018-11-07 17:48 ` Michael S. Tsirkin 2018-07-11 2:27 ` Tiwei Bie 2018-07-11 2:27 ` [PATCH net-next v2 4/5] virtio_ring: add event idx support in packed ring Tiwei Bie 2018-07-11 2:27 ` [virtio-dev] " Tiwei Bie 2018-09-07 14:10 ` Michael S. Tsirkin 2018-09-07 14:10 ` Michael S. Tsirkin 2018-09-07 14:10 ` [virtio-dev] " Michael S. Tsirkin 2018-09-10 2:35 ` Tiwei Bie 2018-09-10 2:35 ` Tiwei Bie 2018-09-10 2:35 ` Tiwei Bie 2018-09-10 2:35 ` Tiwei Bie 2018-07-11 2:27 ` Tiwei Bie 2018-07-11 2:27 ` [PATCH net-next v2 5/5] virtio_ring: enable " Tiwei Bie 2018-07-11 2:27 ` Tiwei Bie 2018-07-11 2:27 ` [virtio-dev] " Tiwei Bie 2018-07-11 2:52 ` [PATCH net-next v2 0/5] virtio: support " Jason Wang 2018-07-11 2:52 ` Jason Wang 2018-07-11 2:52 ` [virtio-dev] " Jason Wang 2018-07-12 21:44 ` David Miller 2018-07-12 21:44 ` David Miller 2018-07-13 0:52 ` Jason Wang 2018-07-13 0:52 ` Jason Wang 2018-07-13 0:52 ` [virtio-dev] " Jason Wang 2018-07-13 3:26 ` Michael S. Tsirkin 2018-07-13 3:26 ` [virtio-dev] " Michael S. Tsirkin 2018-07-13 3:26 ` Michael S. Tsirkin 2018-08-27 14:00 ` Michael S. Tsirkin 2018-08-27 14:00 ` Michael S. Tsirkin 2018-08-27 14:00 ` [virtio-dev] " Michael S. Tsirkin 2018-08-28 5:51 ` Jens Freimann 2018-09-07 1:22 ` Tiwei Bie 2018-09-07 1:22 ` Tiwei Bie 2018-09-07 1:22 ` Tiwei Bie 2018-09-07 1:22 ` Tiwei Bie 2018-09-07 13:00 ` [virtio-dev] " Michael S. Tsirkin 2018-09-07 13:00 ` Michael S. Tsirkin 2018-09-07 13:00 ` Michael S. Tsirkin 2018-09-07 13:00 ` Michael S. Tsirkin 2018-09-10 3:00 ` [virtio-dev] " Tiwei Bie 2018-09-10 3:00 ` Tiwei Bie 2018-09-10 3:00 ` Tiwei Bie 2018-09-10 3:00 ` Tiwei Bie 2018-09-10 3:33 ` [virtio-dev] " Jason Wang 2018-09-10 3:33 ` Jason Wang 2018-09-11 5:37 ` Tiwei Bie 2018-09-11 5:37 ` Tiwei Bie 2018-09-11 5:37 ` Tiwei Bie 2018-09-11 5:37 ` Tiwei Bie 2018-09-12 16:16 ` [virtio-dev] " Michael S. Tsirkin 2018-09-12 16:16 ` Michael S. Tsirkin 2018-09-13 8:59 ` Tiwei Bie 2018-09-13 8:59 ` Tiwei Bie 2018-09-13 8:59 ` Tiwei Bie 2018-09-13 9:47 ` [virtio-dev] " Jason Wang 2018-09-13 9:47 ` Jason Wang 2018-09-13 9:47 ` Jason Wang 2018-10-10 14:36 ` [virtio-dev] " Michael S. Tsirkin 2018-10-10 14:36 ` Michael S. Tsirkin 2018-10-11 12:12 ` Tiwei Bie 2018-10-11 12:12 ` Tiwei Bie 2018-10-11 12:12 ` Tiwei Bie 2018-10-11 12:12 ` Tiwei Bie 2018-10-11 13:48 ` [virtio-dev] " Michael S. Tsirkin 2018-10-11 13:48 ` Michael S. Tsirkin 2018-10-11 13:48 ` Michael S. Tsirkin 2018-10-11 14:13 ` Tiwei Bie 2018-10-11 14:13 ` Tiwei Bie 2018-10-11 14:17 ` Michael S. Tsirkin 2018-10-11 14:17 ` Michael S. Tsirkin 2018-10-11 14:17 ` Michael S. Tsirkin 2018-10-11 14:34 ` [virtio-dev] " Tiwei Bie 2018-10-11 14:34 ` Tiwei Bie 2018-10-11 14:34 ` Tiwei Bie 2018-10-11 14:34 ` [virtio-dev] " Tiwei Bie 2018-10-11 14:17 ` Michael S. Tsirkin 2018-10-11 14:13 ` Tiwei Bie 2018-10-10 14:36 ` Michael S. Tsirkin 2018-09-13 9:47 ` Jason Wang 2018-09-13 8:59 ` Tiwei Bie 2018-09-12 16:16 ` Michael S. Tsirkin 2018-09-10 3:33 ` Jason Wang 2018-09-12 13:06 ` Michael S. Tsirkin 2018-09-12 13:06 ` Michael S. Tsirkin 2018-09-12 13:06 ` Michael S. Tsirkin 2018-09-12 13:06 ` Michael S. Tsirkin
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=2d46a41e-bc00-276a-e19a-105c9dffc75a@redhat.com \ --to=jasowang@redhat.com \ --cc=jfreimann@redhat.com \ --cc=linux-kernel@vger.kernel.org \ --cc=mst@redhat.com \ --cc=netdev@vger.kernel.org \ --cc=tiwei.bie@intel.com \ --cc=virtio-dev@lists.oasis-open.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: linkBe 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.