From: Shirley Ma <mashirle@us.ibm.com>
To: Jason Wang <jasowang@redhat.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>,
eric.dumazet@gmail.com, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, ebiederm@xmission.com,
davem@davemloft.net
Subject: Re: [V2 PATCH 9/9] vhost: zerocopy: poll vq in zerocopy callback
Date: Mon, 21 May 2012 08:42:52 -0700 [thread overview]
Message-ID: <1337614972.12999.56.camel@oc3660625478.ibm.com> (raw)
In-Reply-To: <4FB9DB24.2040408@redhat.com>
On Mon, 2012-05-21 at 14:05 +0800, Jason Wang wrote:
> >> - tx polling depends on skb_orphan() which is often called by
> device
> >> driver when it place the packet into the queue of the devices
> instead
> >> of when the packets were sent. So it was too early for vhost to be
> >> notified.
> > Then do you think it's better to replace with vhost_poll_queue here
> > instead?
>
> Just like what does this patch do - calling vhost_poll_queue() in
> vhost_zerocopy_callback().
> >> - it only works when the pending DMAs exceeds VHOST_MAX_PEND, it's
> >> highly possible that guest needs to be notified when the pending
> >> packets
> >> isn't so much.
> > In which situation the guest needs to be notified when there is no
> TX
> > besides buffers run out?
>
> Consider guest call virtqueue_enable_cb_delayed() which means it only
> need to be notified when 3/4 of pending buffers ( about 178 buffers
> (256-MAX_SKB_FRAGS-2)*3/4 ) were sent by host. So vhost_net would
> notify
> guest when about 60 buffers were pending. Since tx polling is only
> enabled when pending packets exceeds VHOST_MAX_PEND 128, so tx work
> would not be notified to run and guest would never get the interrupt
> it
> expected to re-enable the queue.
So it seems we still need vhost_enable_notify() in handle_tx when there
is no tx in zerocopy case.
Do you know which one is more expensive: the cost of vhost_poll_queue()
in each zerocopy callback or calling vhost_enable_notify()?
Have you compared the results by removing below code in handle_tx()?
- if (unlikely(num_pends > VHOST_MAX_PEND)) {
- tx_poll_start(net, sock);
- set_bit(SOCK_ASYNC_NOSPACE, &sock->flags);
- break;
- }
>
> And just like what we've discussed, tx polling based adding and
> signaling is too early for vhost.
Thanks
Shirley
next prev parent reply other threads:[~2012-05-21 15:42 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-05-02 3:41 [V2 PATCH 0/9] vhost/macvtap zeropcopy fixes Jason Wang
2012-05-02 3:41 ` [V2 PATCH 1/9] macvtap: zerocopy: fix offset calculation when building skb Jason Wang
2012-05-15 17:17 ` Shirley Ma
2012-05-02 3:41 ` [V2 PATCH 2/9] macvtap: zerocopy: fix truesize underestimation Jason Wang
2012-05-15 17:26 ` Shirley Ma
2012-05-16 3:04 ` Jason Wang
2012-05-16 15:03 ` Shirley Ma
2012-05-17 2:59 ` Jason Wang
2012-05-17 15:28 ` Shirley Ma
2012-05-18 10:10 ` Jason Wang
2012-05-18 15:22 ` Shirley Ma
2012-05-21 6:15 ` Jason Wang
2012-05-02 3:41 ` [V2 PATCH 3/9] macvtap: zerocopy: put page when fail to get all requested user pages Jason Wang
2012-05-15 17:33 ` Shirley Ma
2012-05-02 3:42 ` [V2 PATCH 4/9] macvtap: zerocopy: set SKBTX_DEV_ZEROCOPY only when skb is built successfully Jason Wang
2012-05-15 17:44 ` Shirley Ma
2012-05-16 3:17 ` Jason Wang
2012-05-02 3:42 ` [V2 PATCH 5/9] macvtap: zerocopy: validate vectors before building skb Jason Wang
2012-05-02 3:42 ` [V2 PATCH 6/9] vhost_net: zerocopy: fix possible NULL pointer dereference of vq->bufs Jason Wang
2012-05-02 3:42 ` [V2 PATCH 7/9] vhost_net: re-poll only on EAGAIN or ENOBUFS Jason Wang
2012-05-02 3:42 ` [V2 PATCH 8/9] vhost_net: zerocopy: adding and signalling immediately when fully copied Jason Wang
2012-05-02 3:42 ` [V2 PATCH 9/9] vhost: zerocopy: poll vq in zerocopy callback Jason Wang
2012-05-15 16:50 ` Shirley Ma
2012-05-16 2:58 ` Jason Wang
2012-05-16 15:10 ` Shirley Ma
2012-05-16 15:14 ` Michael S. Tsirkin
2012-05-16 17:32 ` Shirley Ma
2012-05-16 18:36 ` Michael S. Tsirkin
2012-05-16 19:08 ` Shirley Ma
2012-05-21 5:22 ` Jason Wang
2012-05-17 2:50 ` Jason Wang
2012-05-17 15:34 ` Shirley Ma
2012-05-18 9:58 ` Jason Wang
2012-05-18 15:29 ` Shirley Ma
2012-05-21 6:05 ` Jason Wang
2012-05-21 15:42 ` Shirley Ma [this message]
2012-05-21 16:12 ` Shirley Ma
2012-05-22 10:13 ` Jason Wang
2012-05-22 10:05 ` Jason Wang
2012-05-22 15:55 ` Shirley Ma
2012-05-23 10:31 ` Jason Wang
2012-05-02 5:50 ` [V2 PATCH 0/9] vhost/macvtap zeropcopy fixes Michael S. Tsirkin
2012-05-02 6:44 ` David Miller
2012-05-02 8:11 ` Michael S. Tsirkin
2012-05-02 19:40 ` Eric W. Biederman
2012-05-02 21:31 ` Michael S. Tsirkin
2012-05-02 21:54 ` Eric W. Biederman
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=1337614972.12999.56.camel@oc3660625478.ibm.com \
--to=mashirle@us.ibm.com \
--cc=davem@davemloft.net \
--cc=ebiederm@xmission.com \
--cc=eric.dumazet@gmail.com \
--cc=jasowang@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mst@redhat.com \
--cc=netdev@vger.kernel.org \
/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).