All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Shannon Nelson <shannon.nelson@amd.com>
Cc: xuanzhuo@linux.alibaba.com, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	virtualization@lists.linux-foundation.org, edumazet@google.com,
	kuba@kernel.org, maxime.coquelin@redhat.com, pabeni@redhat.com,
	davem@davemloft.net
Subject: Re: [PATCH net-next v4 2/2] virtio-net: add cond_resched() to the command waiting loop
Date: Thu, 20 Jul 2023 16:58:18 -0400	[thread overview]
Message-ID: <20230720164930-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <b949697e-319a-7cc1-84d8-1391713fa645@amd.com>

On Thu, Jul 20, 2023 at 08:31:13AM -0700, Shannon Nelson wrote:
> On 7/20/23 1:38 AM, Jason Wang wrote:
> > 
> > Adding cond_resched() to the command waiting loop for a better
> > co-operation with the scheduler. This allows to give CPU a breath to
> > run other task(workqueue) instead of busy looping when preemption is
> > not allowed on a device whose CVQ might be slow.
> > 
> > Signed-off-by: Jason Wang <jasowang@redhat.com>
> > ---
> >   drivers/net/virtio_net.c | 4 +++-
> >   1 file changed, 3 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> > index 9f3b1d6ac33d..e7533f29b219 100644
> > --- a/drivers/net/virtio_net.c
> > +++ b/drivers/net/virtio_net.c
> > @@ -2314,8 +2314,10 @@ static bool virtnet_send_command(struct virtnet_info *vi, u8 class, u8 cmd,
> >           * into the hypervisor, so the request should be handled immediately.
> >           */
> >          while (!virtqueue_get_buf(vi->cvq, &tmp) &&
> > -              !virtqueue_is_broken(vi->cvq))
> > +              !virtqueue_is_broken(vi->cvq)) {
> > +               cond_resched();
> >                  cpu_relax();
> > +       }
> 
> The cover letter suggests that this addresses the infinite poll for buggy
> devices, but I don't see how that is resolved here.  This should make it a
> little nicer to the system, but it still is going to poll forever on a
> device that has gone catatonic.  Is there a reason that I'm missing that we
> don't have a polling limit here?
> 
> sln

we don't know what the limit would be. but given it's a workqueue
now, why does it still have to poll as opposed to blocking?


> > 
> >          return vi->ctrl->status == VIRTIO_NET_OK;
> >   }
> > --
> > 2.39.3
> > 
> > _______________________________________________
> > Virtualization mailing list
> > Virtualization@lists.linux-foundation.org
> > https://lists.linuxfoundation.org/mailman/listinfo/virtualization

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

WARNING: multiple messages have this Message-ID (diff)
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Shannon Nelson <shannon.nelson@amd.com>
Cc: Jason Wang <jasowang@redhat.com>,
	xuanzhuo@linux.alibaba.com, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	virtualization@lists.linux-foundation.org, edumazet@google.com,
	maxime.coquelin@redhat.com, kuba@kernel.org, pabeni@redhat.com,
	davem@davemloft.net
Subject: Re: [PATCH net-next v4 2/2] virtio-net: add cond_resched() to the command waiting loop
Date: Thu, 20 Jul 2023 16:58:18 -0400	[thread overview]
Message-ID: <20230720164930-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <b949697e-319a-7cc1-84d8-1391713fa645@amd.com>

On Thu, Jul 20, 2023 at 08:31:13AM -0700, Shannon Nelson wrote:
> On 7/20/23 1:38 AM, Jason Wang wrote:
> > 
> > Adding cond_resched() to the command waiting loop for a better
> > co-operation with the scheduler. This allows to give CPU a breath to
> > run other task(workqueue) instead of busy looping when preemption is
> > not allowed on a device whose CVQ might be slow.
> > 
> > Signed-off-by: Jason Wang <jasowang@redhat.com>
> > ---
> >   drivers/net/virtio_net.c | 4 +++-
> >   1 file changed, 3 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> > index 9f3b1d6ac33d..e7533f29b219 100644
> > --- a/drivers/net/virtio_net.c
> > +++ b/drivers/net/virtio_net.c
> > @@ -2314,8 +2314,10 @@ static bool virtnet_send_command(struct virtnet_info *vi, u8 class, u8 cmd,
> >           * into the hypervisor, so the request should be handled immediately.
> >           */
> >          while (!virtqueue_get_buf(vi->cvq, &tmp) &&
> > -              !virtqueue_is_broken(vi->cvq))
> > +              !virtqueue_is_broken(vi->cvq)) {
> > +               cond_resched();
> >                  cpu_relax();
> > +       }
> 
> The cover letter suggests that this addresses the infinite poll for buggy
> devices, but I don't see how that is resolved here.  This should make it a
> little nicer to the system, but it still is going to poll forever on a
> device that has gone catatonic.  Is there a reason that I'm missing that we
> don't have a polling limit here?
> 
> sln

we don't know what the limit would be. but given it's a workqueue
now, why does it still have to poll as opposed to blocking?


> > 
> >          return vi->ctrl->status == VIRTIO_NET_OK;
> >   }
> > --
> > 2.39.3
> > 
> > _______________________________________________
> > Virtualization mailing list
> > Virtualization@lists.linux-foundation.org
> > https://lists.linuxfoundation.org/mailman/listinfo/virtualization


  reply	other threads:[~2023-07-20 20:58 UTC|newest]

Thread overview: 82+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-20  8:38 [PATCH net-next v4 0/2] virtio-net: don't busy poll for cvq command Jason Wang
2023-07-20  8:38 ` Jason Wang
2023-07-20  8:38 ` [PATCH net-next v4 1/2] virtio-net: convert rx mode setting to use workqueue Jason Wang
2023-07-20  8:38   ` Jason Wang
2023-07-20 20:25   ` Shannon Nelson via Virtualization
2023-07-20 20:25     ` Shannon Nelson
2023-07-20  8:38 ` [PATCH net-next v4 2/2] virtio-net: add cond_resched() to the command waiting loop Jason Wang
2023-07-20  8:38   ` Jason Wang
2023-07-20 15:31   ` Shannon Nelson via Virtualization
2023-07-20 15:31     ` Shannon Nelson
2023-07-20 20:58     ` Michael S. Tsirkin [this message]
2023-07-20 20:58       ` Michael S. Tsirkin
2023-07-20 20:26   ` Shannon Nelson via Virtualization
2023-07-20 20:26     ` Shannon Nelson
2023-07-20 21:02     ` Michael S. Tsirkin
2023-07-20 21:02       ` Michael S. Tsirkin
2023-07-21 14:37       ` Maxime Coquelin
2023-07-21 14:37         ` Maxime Coquelin
2023-07-21 14:45         ` Michael S. Tsirkin
2023-07-21 14:45           ` Michael S. Tsirkin
2023-07-21 14:58           ` Maxime Coquelin
2023-07-21 14:58             ` Maxime Coquelin
2023-07-21 15:10             ` Michael S. Tsirkin
2023-07-21 15:10               ` Michael S. Tsirkin
2023-07-21 20:18               ` Maxime Coquelin
2023-07-21 20:18                 ` Maxime Coquelin
2023-07-24  6:46                 ` Michael S. Tsirkin
2023-07-24  6:46                   ` Michael S. Tsirkin
2023-07-24  6:52                   ` Jason Wang
2023-07-24  6:52                     ` Jason Wang
2023-07-24  7:18                     ` Michael S. Tsirkin
2023-07-24  7:18                       ` Michael S. Tsirkin
2023-07-25  3:03                       ` Jason Wang
2023-07-25  3:03                         ` Jason Wang
2024-02-22 19:21                         ` Michael S. Tsirkin
2024-02-26  5:08                           ` Jason Wang
2023-10-05 19:35                   ` Feng Liu via Virtualization
2023-10-05 19:35                     ` Feng Liu
2023-10-08  5:27                     ` Jason Wang
2023-10-08  5:27                       ` Jason Wang
2023-07-24  6:52                 ` Jason Wang
2023-07-24  6:52                   ` Jason Wang
2023-07-24  7:17                   ` Michael S. Tsirkin
2023-07-24  7:17                     ` Michael S. Tsirkin
2023-07-25  3:07                     ` Jason Wang
2023-07-25  3:07                       ` Jason Wang
2023-07-25  7:36                       ` Michael S. Tsirkin
2023-07-25  7:36                         ` Michael S. Tsirkin
2023-07-26  1:55                         ` Jason Wang
2023-07-26  1:55                           ` Jason Wang
2023-07-26 11:37                           ` Michael S. Tsirkin
2023-07-26 11:37                             ` Michael S. Tsirkin
2023-07-27  6:03                             ` Jason Wang
2023-07-27  6:03                               ` Jason Wang
2023-07-27  6:10                               ` Michael S. Tsirkin
2023-07-27  6:10                                 ` Michael S. Tsirkin
2023-07-27  8:59                                 ` Jason Wang
2023-07-27  8:59                                   ` Jason Wang
2023-07-27  9:46                                   ` Michael S. Tsirkin
2023-07-27  9:46                                     ` Michael S. Tsirkin
2023-07-31  6:30                                     ` Jason Wang
2023-07-31  6:30                                       ` Jason Wang
2023-08-08  2:30                                       ` Jason Wang
2023-08-08  2:30                                         ` Jason Wang
2023-08-10 19:41                                         ` Michael S. Tsirkin
2023-08-10 19:41                                           ` Michael S. Tsirkin
2023-08-11  2:23                                           ` Jason Wang
2023-08-11  2:23                                             ` Jason Wang
2023-08-11  5:42                                             ` Michael S. Tsirkin
2023-08-11  5:42                                               ` Michael S. Tsirkin
2023-08-11  9:18                                               ` Jason Wang
2023-08-11  9:18                                                 ` Jason Wang
2023-08-11  9:21                                                 ` Michael S. Tsirkin
2023-08-11  9:21                                                   ` Michael S. Tsirkin
2023-08-11  9:43                                                   ` Jason Wang
2023-08-11  9:43                                                     ` Jason Wang
2023-08-11  9:51                                                     ` Michael S. Tsirkin
2023-08-11  9:51                                                       ` Michael S. Tsirkin
2023-08-11  9:54                                                       ` Jason Wang
2023-08-11  9:54                                                         ` Jason Wang
2023-08-11 10:12                                                         ` Michael S. Tsirkin
2023-08-11 10:12                                                           ` 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=20230720164930-mutt-send-email-mst@kernel.org \
    --to=mst@redhat.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=maxime.coquelin@redhat.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=shannon.nelson@amd.com \
    --cc=virtualization@lists.linux-foundation.org \
    --cc=xuanzhuo@linux.alibaba.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.