From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: David Howells <dhowells@redhat.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>,
Trond.Myklebust@netapp.com, linux-nfs@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH] NFS: Fix RCU warnings in nfs_inode_return_delegation_noreclaim() [ver #2]
Date: Mon, 5 Apr 2010 10:57:54 -0700 [thread overview]
Message-ID: <20100405175754.GE2525@linux.vnet.ibm.com> (raw)
In-Reply-To: <4161.1270133211@redhat.com>
On Thu, Apr 01, 2010 at 03:46:51PM +0100, David Howells wrote:
> Paul E. McKenney <paulmck@linux.vnet.ibm.com> wrote:
>
> > So you have objected to needless memory barriers. How do you feel
> > about possibly needless ACCESS_ONCE() calls?
>
> That would work here since it shouldn't emit any excess instructions.
And here is the corresponding patch. Seem reasonable?
Thanx, Paul
------------------------------------------------------------------------
rcupdate.h | 19 +++++++++++++++++++
1 file changed, 19 insertions(+)
commit 61ad1405cd442fe54f87ff97febf52610817903e
Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Date: Mon Apr 5 10:52:53 2010 -0700
rcu: add rcu_dereference_protect to avoid smp_read_barrier_depends overhead
This patch adds a variant of rcu_dereference() that handles situations
where the RCU-protected data structure cannot change, perhaps due to
our holding the update-side lock, or where the RCU-protected pointer is
only to be tested, not dereferenced.
Suggested-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
index 872a98e..123b834 100644
--- a/include/linux/rcupdate.h
+++ b/include/linux/rcupdate.h
@@ -209,9 +209,28 @@ static inline int rcu_read_lock_sched_held(void)
rcu_dereference_raw(p); \
})
+/**
+ * rcu_dereference_protected - fetch RCU pointer when protected by algorithm
+ *
+ * Return the value of the specified RCU-protected pointer, but omit
+ * the smp_read_barrier_depends() and keep the ACCESS_ONCE(). This
+ * is useful in cases where update-side locks prevent the value of the
+ * pointer from changing, and is also useful in cases where the value
+ * of this pointer is accessed, but the pointer is not dereferenced.
+ * An example of this latter occurs when testing an RCU-protected
+ * pointer against NULL.
+ */
+#define rcu_dereference_protected(p, c) \
+ ({ \
+ if (debug_lockdep_rcu_enabled() && !(c)) \
+ lockdep_rcu_dereference(__FILE__, __LINE__); \
+ ACCESS_ONCE(p); \
+ })
+
#else /* #ifdef CONFIG_PROVE_RCU */
#define rcu_dereference_check(p, c) rcu_dereference_raw(p)
+#define rcu_dereference_protected(p, c) ACCESS_ONCE(p)
#endif /* #else #ifdef CONFIG_PROVE_RCU */
next prev parent reply other threads:[~2010-04-05 17:58 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-03-18 13:33 [PATCH] NFS: Fix RCU warnings in nfs_inode_return_delegation_noreclaim() [ver #2] David Howells
2010-03-19 2:25 ` Paul E. McKenney
2010-03-29 19:02 ` David Howells
2010-03-29 19:21 ` Paul E. McKenney
2010-03-29 20:15 ` David Howells
2010-03-29 20:26 ` Eric Dumazet
2010-03-29 21:05 ` Paul E. McKenney
2010-03-29 22:22 ` David Howells
2010-03-29 22:36 ` Paul E. McKenney
2010-03-29 22:59 ` David Howells
2010-03-29 23:26 ` Paul E. McKenney
2010-03-30 15:40 ` Paul E. McKenney
2010-03-30 16:39 ` David Howells
2010-03-30 16:49 ` Paul E. McKenney
2010-03-30 17:04 ` Eric Dumazet
2010-03-30 17:25 ` Paul E. McKenney
2010-03-30 23:51 ` David Howells
2010-03-31 0:08 ` Paul E. McKenney
2010-03-31 14:04 ` David Howells
2010-03-31 15:16 ` Paul E. McKenney
2010-03-31 17:37 ` David Howells
2010-03-31 18:30 ` Paul E. McKenney
2010-03-31 18:32 ` Eric Dumazet
2010-03-31 22:53 ` David Howells
2010-04-01 1:29 ` Paul E. McKenney
2010-04-01 11:45 ` David Howells
2010-04-01 14:39 ` Paul E. McKenney
2010-04-01 14:46 ` David Howells
2010-04-05 17:57 ` Paul E. McKenney [this message]
2010-04-06 9:30 ` David Howells
2010-04-06 16:14 ` David Howells
2010-04-06 17:29 ` Paul E. McKenney
2010-04-06 19:34 ` David Howells
2010-04-07 0:02 ` Paul E. McKenney
2010-04-07 13:22 ` David Howells
2010-04-07 15:57 ` Paul E. McKenney
2010-04-07 16:35 ` RCU condition checks David Howells
2010-04-07 17:10 ` Paul E. McKenney
2010-04-11 22:57 ` Trond Myklebust
2010-04-12 16:47 ` Paul E. McKenney
2010-03-30 16:37 ` [PATCH] NFS: Fix RCU warnings in nfs_inode_return_delegation_noreclaim() [ver #2] David Howells
2010-03-30 17:01 ` 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=20100405175754.GE2525@linux.vnet.ibm.com \
--to=paulmck@linux.vnet.ibm.com \
--cc=Trond.Myklebust@netapp.com \
--cc=dhowells@redhat.com \
--cc=eric.dumazet@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nfs@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).