From: "Hu, Jiayu" <jiayu.hu@intel.com>
To: "Bie, Tiwei" <tiwei.bie@intel.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>,
"maxime.coquelin@redhat.com" <maxime.coquelin@redhat.com>,
"stable@dpdk.org" <stable@dpdk.org>
Subject: Re: [dpdk-stable] [PATCH] vhost: fix interrupt suppression for the split ring
Date: Wed, 27 Feb 2019 05:38:28 +0000 [thread overview]
Message-ID: <ED946F0BEFE0A141B63BABBD629A2A9B3CF5D471@shsmsx102.ccr.corp.intel.com> (raw)
In-Reply-To: <20190225070654.GA22485@dpdk-tbie.sh.intel.com>
> -----Original Message-----
> From: Bie, Tiwei
> Sent: Monday, February 25, 2019 3:07 PM
> To: Hu, Jiayu <jiayu.hu@intel.com>
> Cc: dev@dpdk.org; maxime.coquelin@redhat.com; stable@dpdk.org
> Subject: Re: [PATCH] vhost: fix interrupt suppression for the split ring
>
> On Fri, Feb 22, 2019 at 04:13:50PM +0800, Jiayu Hu wrote:
> > The VIRTIO_RING_F_EVENT_IDX feature of split ring might
> > be broken, as the value of signalled_used is unpredictable
> > after live migration or start up. This patch fixes it by
> > using signalled_used_valid.
>
> Seems you also need to invalidate the signalled_used in e.g.
> set_vring_base, as signalled_used may not match last_used_idx
> anymore e.g. when guest driver is reloaded.
To handle the virtio device reload case, how about setting
signalled_used_valid to false in get_vring_base()?
Thanks,
Jiayu
>
> >
> > In addition, this patch makes the VIRTIO_RING_F_EVENT_IDX
> > implementation of split ring match kernel backend to suppress
> > more interrupts.
> >
> > Fixes: e37ff954405a ("vhost: support virtqueue interrupt/notification
> suppression")
> > Cc: stable@dpdk.org
> >
> > Signed-off-by: Jiayu Hu <jiayu.hu@intel.com>
> > ---
> > lib/librte_vhost/vhost.h | 12 ++++++++----
> > 1 file changed, 8 insertions(+), 4 deletions(-)
> >
> > diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h
> > index fc31796..c4e5e34 100644
> > --- a/lib/librte_vhost/vhost.h
> > +++ b/lib/librte_vhost/vhost.h
> > @@ -633,16 +633,20 @@ vhost_vring_call_split(struct virtio_net *dev,
> struct vhost_virtqueue *vq)
> > if (dev->features & (1ULL << VIRTIO_RING_F_EVENT_IDX)) {
> > uint16_t old = vq->signalled_used;
> > uint16_t new = vq->last_used_idx;
> > + bool signalled_used_valid = vq->signalled_used_valid;
> > +
> > + vq->signalled_used = new;
> > + vq->signalled_used_valid = true;
> >
> > VHOST_LOG_DEBUG(VHOST_DATA, "%s:
> used_event_idx=%d, old=%d, new=%d\n",
> > __func__,
> > vhost_used_event(vq),
> > old, new);
> > - if (vhost_need_event(vhost_used_event(vq), new, old)
> > - && (vq->callfd >= 0)) {
> > - vq->signalled_used = vq->last_used_idx;
> > +
> > + if ((vhost_need_event(vhost_used_event(vq), new, old) &&
> > + (vq->callfd >= 0)) ||
> > + unlikely(!signalled_used_valid))
> > eventfd_write(vq->callfd, (eventfd_t) 1);
> > - }
> > } else {
> > /* Kick the guest if necessary. */
> > if (!(vq->avail->flags & VRING_AVAIL_F_NO_INTERRUPT)
> > --
> > 2.7.4
> >
next prev parent reply other threads:[~2019-02-27 5:38 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-02-22 8:13 Jiayu Hu
2019-02-25 7:06 ` Tiwei Bie
2019-02-27 5:38 ` Hu, Jiayu [this message]
2019-02-27 10:55 ` Tiwei Bie
2019-03-17 6:38 ` [dpdk-stable] [PATCH v2] " Jiayu Hu
2019-03-20 0:45 ` Wang, Yinan
2019-03-20 4:56 ` Tiwei Bie
2019-03-20 7:36 ` Maxime Coquelin
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=ED946F0BEFE0A141B63BABBD629A2A9B3CF5D471@shsmsx102.ccr.corp.intel.com \
--to=jiayu.hu@intel.com \
--cc=dev@dpdk.org \
--cc=maxime.coquelin@redhat.com \
--cc=stable@dpdk.org \
--cc=tiwei.bie@intel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).