From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: Eric Dumazet <dada1@cosmosbay.com>
Cc: Andrew Morton <akpm@osdl.org>,
Dipankar Sarma <dipankar@in.ibm.com>,
linux-kernel@vger.kernel.org
Subject: Re: [RCU] adds a prefetch() in rcu_do_batch()
Date: Wed, 29 Nov 2006 17:25:28 -0800 [thread overview]
Message-ID: <20061130012528.GJ2335@us.ibm.com> (raw)
In-Reply-To: <200611221602.29597.dada1@cosmosbay.com>
On Wed, Nov 22, 2006 at 04:02:29PM +0100, Eric Dumazet wrote:
> On some workloads, (for example when lot of close() syscalls are done), RCU
> qlen can be quite large, and RCU heads are no longer in cpu cache when
> rcu_do_batch() is called.
>
> This patches adds a prefetch() in rcu_do_batch() to give CPU a hint to bring
> back cache lines containing 'struct rcu_head's.
>
> Most list manipulations macros include prefetch(), but not open coded ones (at
> least with current C compilers :) )
>
> I got a nice speedup on a trivial benchmark (3.48 us per iteration instead of
> 3.95 us on a 1.6 GHz Pentium-M)
> while (1) { pipe(p); close(fd[0]); close(fd[1]);}
Interesting! How much of the speedup was due to the prefetch() and how
much to removing the extra store to rdp->donelist?
Thanx, Paul
> Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
> --- linux-2.6.19-rc6/kernel/rcupdate.c 2006-11-16 05:03:40.000000000 +0100
> +++ linux-2.6.19-rc6-ed/kernel/rcupdate.c 2006-11-22 15:12:09.000000000 +0100
> @@ -235,12 +235,14 @@ static void rcu_do_batch(struct rcu_data
>
> list = rdp->donelist;
> while (list) {
> - next = rdp->donelist = list->next;
> + next = list->next;
> + prefetch(next);
> list->func(list);
> list = next;
> if (++count >= rdp->blimit)
> break;
> }
> + rdp->donelist = list;
>
> local_irq_disable();
> rdp->qlen -= count;
next prev parent reply other threads:[~2006-11-30 1:24 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-11-21 20:33 [PATCH] snd-hda-intel: fix insufficient memory wbrana
2006-11-22 6:46 ` Andrew Morton
2006-11-22 15:02 ` [RCU] adds a prefetch() in rcu_do_batch() Eric Dumazet
2006-11-22 17:48 ` [PATCH] dont insert pipe dentries into dentry_hashtable Eric Dumazet
2006-11-22 21:36 ` Andrew Morton
2006-11-22 21:40 ` Al Viro
2006-11-23 4:12 ` David Miller
2006-11-30 1:25 ` Paul E. McKenney [this message]
2006-11-30 8:55 ` [RCU] adds a prefetch() in rcu_do_batch() Eric Dumazet
2006-11-30 17:15 ` Paul E. McKenney
2006-11-22 17:19 ` [PATCH] snd-hda-intel: fix insufficient memory wbrana
2006-11-22 20:04 ` Andrew Morton
2006-11-23 20:17 ` wbrana
2006-11-23 21:10 ` Andrew Morton
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=20061130012528.GJ2335@us.ibm.com \
--to=paulmck@linux.vnet.ibm.com \
--cc=akpm@osdl.org \
--cc=dada1@cosmosbay.com \
--cc=dipankar@in.ibm.com \
--cc=linux-kernel@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