Linux virtualization list
 help / color / mirror / Atom feed
From: Will Deacon <will@kernel.org>
To: Jason Wang <jasowang@redhat.com>
Cc: linux-kernel@lists.infradead.org, paulmck@kernel.org,
	mst@redhat.com, peterz@infradead.org,
	virtualization@lists.linux-foundation.org,
	stern@rowland.harvard.edu
Subject: Re: read_barrier_depends() usage in vhost.c
Date: Wed, 30 Oct 2019 16:13:17 +0000	[thread overview]
Message-ID: <20191030161317.GA20094@willie-the-truck> (raw)
In-Reply-To: <d0219a18-b7be-1e43-559f-ab6c58483115@redhat.com>

Hi Jason,

On Mon, Oct 21, 2019 at 01:48:53PM +0800, Jason Wang wrote:
> On 2019/10/19 上午4:58, Will Deacon wrote:
> > Staring at the code some more, my best bet at the moment
> > is that the load of 'indirect->addr' is probably the one to worry about,
> > since it's part of the vring and can be updated concurrently.
> 
> I'm also confused about the barrier here, basically in driver side we did:
> 
> 1) allocate pages
> 
> 2) store pages in indirect->addr
> 
> 3) smp_wmb()
> 
> 4) increase the avail idx (somehow a tail pointer of vring)
> 
> in vhost we did:
> 
> 1) read avail idx
> 
> 2) smp_rmb()
> 
> 3) read indirect->addr
> 
> 4) read from indirect->addr
> 
> It looks to me even the data dependency barrier is not necessary since we
> have rmb() which is sufficient for us to the correct indirect->addr and
> driver are not expected to do any writing to indirect->addr after avail idx
> is increased ?

That makes sense to me: the synchronization is done via vq->avail_idx()
and so the subsequent access of the indirect pages doesn't need additional
barriers, even on Alpha. Thanks.

I'll write up a patch and adopt your explanation above.

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

  reply	other threads:[~2019-10-30 16:13 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-16 23:33 read_barrier_depends() usage in vhost.c Will Deacon
2019-10-16 23:36 ` Will Deacon
2019-10-17  2:17 ` Jason Wang
2019-10-18 20:58   ` Will Deacon
2019-10-21  5:48     ` Jason Wang
2019-10-30 16:13       ` Will Deacon [this message]
     [not found] <20191016233602.i2afxb5mb465laq6@willie-the-truck>
2019-12-18  9:19 ` Herbert Xu
     [not found] <20191218091906.cmzgqnwyekak5dzv@gondor.apana.org.au>
2019-12-20  3:40 ` Jason Wang

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=20191030161317.GA20094@willie-the-truck \
    --to=will@kernel.org \
    --cc=jasowang@redhat.com \
    --cc=linux-kernel@lists.infradead.org \
    --cc=mst@redhat.com \
    --cc=paulmck@kernel.org \
    --cc=peterz@infradead.org \
    --cc=stern@rowland.harvard.edu \
    --cc=virtualization@lists.linux-foundation.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