linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Will Deacon <will@kernel.org>
To: linux-kernel@vger.kernel.org
Cc: Will Deacon <will@kernel.org>, Yunjae Lee <lyj7694@gmail.com>,
	SeongJae Park <sj38.park@gmail.com>,
	"Paul E. McKenney" <paulmck@kernel.org>,
	Josh Triplett <josh@joshtriplett.org>,
	Matt Turner <mattst88@gmail.com>,
	Ivan Kokshaysky <ink@jurassic.park.msu.ru>,
	Richard Henderson <rth@twiddle.net>,
	Peter Zijlstra <peterz@infradead.org>,
	Alan Stern <stern@rowland.harvard.edu>,
	Michael Ellerman <mpe@ellerman.id.au>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	Jason Wang <jasowang@redhat.com>, Arnd Bergmann <arnd@arndb.de>,
	Joe Perches <joe@perches.com>, Boqun Feng <boqun.feng@gmail.com>,
	linux-alpha@vger.kernel.org,
	virtualization@lists.linux-foundation.org
Subject: [PATCH 04/13] vhost: Remove redundant use of read_barrier_depends() barrier
Date: Fri,  8 Nov 2019 17:01:11 +0000	[thread overview]
Message-ID: <20191108170120.22331-5-will@kernel.org> (raw)
In-Reply-To: <20191108170120.22331-1-will@kernel.org>

Since commit 76ebbe78f739 ("locking/barriers: Add implicit
smp_read_barrier_depends() to READ_ONCE()"), there is no need to use
'smp_read_barrier_depends()' outside of the Alpha architecture code.

Unfortunately, there is precisely >one< user in the vhost code, and
there isn't an obvious 'READ_ONCE()' access making the barrier
redundant. However, on closer inspection (thanks, Jason), it appears
that vring synchronisation between the producer and consumer occurs via
the 'avail_idx' field, which is followed up by an 'rmb()' in
'vhost_get_vq_desc()', making the 'read_barrier_depends()' redundant on
Alpha.

Jason says:

  | 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

Remove the redundant barrier invocation.

Suggested-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
---
 drivers/vhost/vhost.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 36ca2cf419bf..865bc91b783c 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -2128,11 +2128,6 @@ static int get_indirect(struct vhost_virtqueue *vq,
 		return ret;
 	}
 	iov_iter_init(&from, READ, vq->indirect, ret, len);
-
-	/* We will use the result as an address to read from, so most
-	 * architectures only need a compiler barrier here. */
-	read_barrier_depends();
-
 	count = len / sizeof desc;
 	/* Buffers are chained via a 16 bit next field, so
 	 * we can have at most 2^16 of these. */
-- 
2.24.0.rc1.363.gb1bccd3e3d-goog


  parent reply	other threads:[~2019-11-08 17:01 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-08 17:01 [PATCH 00/13] Finish off [smp_]read_barrier_depends() Will Deacon
2019-11-08 17:01 ` [PATCH 01/13] compiler.h: Split {READ,WRITE}_ONCE definitions out into rwonce.h Will Deacon
2019-11-08 19:57   ` Arnd Bergmann
2019-11-11  8:10     ` Christian Borntraeger
2019-11-11  9:32       ` Arnd Bergmann
2019-11-12 11:36         ` Will Deacon
2019-11-08 17:01 ` [PATCH 02/13] READ_ONCE: Undefine internal __READ_ONCE_SIZE macro after use Will Deacon
2019-11-08 17:01 ` [PATCH 03/13] READ_ONCE: Allow __READ_ONCE_SIZE cases to be overridden by the architecture Will Deacon
2019-11-08 17:01 ` Will Deacon [this message]
2019-11-08 17:01 ` [PATCH 05/13] alpha: Override READ_ONCE() with barriered implementation Will Deacon
2019-11-08 17:01 ` [PATCH 06/13] READ_ONCE: Remove smp_read_barrier_depends() invocation Will Deacon
2019-11-08 17:01 ` [PATCH 07/13] alpha: Replace smp_read_barrier_depends() usage with smp_[r]mb() Will Deacon
2019-11-08 17:01 ` [PATCH 08/13] locking/barriers: Remove definitions for [smp_]read_barrier_depends() Will Deacon
2019-11-08 17:01 ` [PATCH 09/13] Documentation/barriers: Remove references to [smp_]read_barrier_depends() Will Deacon
2019-11-21 19:32   ` [PATCH] Documentation/barriers/kokr: " SeongJae Park
2019-11-26 22:20     ` Paul E. McKenney
2019-11-29 18:08     ` [PATCH v2] " SeongJae Park
2019-12-06 17:20       ` SeongJae Park
2019-12-06 20:44         ` Paul E. McKenney
2019-12-06 21:29           ` SeongJae Park
2019-12-06 22:08             ` Paul E. McKenney
2019-12-06 22:38               ` SeongJae Park
2019-12-06 22:51                 ` Paul E. McKenney
2019-12-09  9:44                   ` Will Deacon
2019-12-09 17:00                     ` Paul E. McKenney
2019-12-09 17:06                       ` Will Deacon
2019-12-09 17:43                         ` SeongJae Park
2019-11-08 17:01 ` [PATCH 10/13] tools/memory-model: Remove smp_read_barrier_depends() from informal doc Will Deacon
2019-11-08 17:42   ` Alan Stern
2019-11-08 17:01 ` [PATCH 11/13] powerpc: Remove comment about read_barrier_depends() Will Deacon
2019-11-20 10:37   ` Michael Ellerman
2019-11-26 22:24     ` Paul E. McKenney
2019-11-08 17:01 ` [PATCH 12/13] include/linux: Remove smp_read_barrier_depends() from comments Will Deacon
2019-11-08 17:01 ` [PATCH 13/13] checkpatch: Remove checks relating to [smp_]read_barrier_depends() Will Deacon
2019-11-08 18:50 ` [PATCH 00/13] Finish off [smp_]read_barrier_depends() Paul E. McKenney

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=20191108170120.22331-5-will@kernel.org \
    --to=will@kernel.org \
    --cc=arnd@arndb.de \
    --cc=boqun.feng@gmail.com \
    --cc=ink@jurassic.park.msu.ru \
    --cc=jasowang@redhat.com \
    --cc=joe@perches.com \
    --cc=josh@joshtriplett.org \
    --cc=linux-alpha@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lyj7694@gmail.com \
    --cc=mattst88@gmail.com \
    --cc=mpe@ellerman.id.au \
    --cc=mst@redhat.com \
    --cc=paulmck@kernel.org \
    --cc=peterz@infradead.org \
    --cc=rth@twiddle.net \
    --cc=sj38.park@gmail.com \
    --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;
as well as URLs for NNTP newsgroup(s).