From: Tejun Heo <tj@kernel.org>
To: Christoph Lameter <cl@linux.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>,
akpm@linux-foundation.org, linux-mm@kvack.org,
shaohua.li@intel.com
Subject: Re: [PATCH] percpu: preemptless __per_cpu_counter_add
Date: Thu, 21 Apr 2011 16:43:00 +0200 [thread overview]
Message-ID: <20110421144300.GA22898@htj.dyndns.org> (raw)
In-Reply-To: <alpine.DEB.2.00.1104180930580.23207@router.home>
Hello, Christoph.
On Mon, Apr 18, 2011 at 09:38:03AM -0500, Christoph Lameter wrote:
> Preemption flips are not cheap since enabling preemption may mean a call
> into the scheduler. On RT things get more expensive.
>
> Preempt_enable means at least one additional branch. We are saving a
> branch by not using preempt.
It is cheap. The cost of preempt_enable() regarding scheduler call is
TIF_NEED_RESCHED check. The scheduler() call occurring afterwards is
not the overhead of preemption check, but the overhead of preemption
itself. Also, in cases where the preemption check doesn't make sense
(I don't think that's the case here), the right thing to do is using
preempt_enable_no_resched().
> In order to make it simple I avoided an preempt enable/disable. With
> Shaohua's patches there will be a simple atomic_add within the last if
> cluase. I was able to consolidate multiple code paths into the cmpxchg
> loop with this approach.
>
> The one below avoids the #ifdef that is ugly...
That said, combined with Shaohua's patch, maybe it's better this way.
Let's see...
Thanks.
--
tejun
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2011-04-21 14:43 UTC|newest]
Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-04-13 14:45 percpu: preemptless __per_cpu_counter_add Christoph Lameter
2011-04-13 16:49 ` Christoph Lameter
2011-04-13 18:56 ` Tejun Heo
2011-04-13 20:22 ` [PATCH] " Christoph Lameter
2011-04-13 21:50 ` Tejun Heo
2011-04-13 22:17 ` Christoph Lameter
2011-04-13 22:23 ` Christoph Lameter
2011-04-13 23:55 ` Tejun Heo
2011-04-14 2:00 ` Eric Dumazet
2011-04-14 2:14 ` Eric Dumazet
2011-04-14 21:10 ` Christoph Lameter
2011-04-14 21:15 ` Tejun Heo
2011-04-15 17:37 ` Christoph Lameter
2011-04-15 18:27 ` Tejun Heo
2011-04-15 19:43 ` Christoph Lameter
2011-04-15 23:52 ` Tejun Heo
2011-04-18 14:38 ` Christoph Lameter
2011-04-21 14:43 ` Tejun Heo [this message]
2011-04-21 14:58 ` Tejun Heo
2011-04-21 17:50 ` Christoph Lameter
2011-04-21 18:01 ` Tejun Heo
2011-04-21 18:20 ` Christoph Lameter
2011-04-21 18:37 ` Tejun Heo
2011-04-21 18:54 ` Christoph Lameter
2011-04-21 19:08 ` Tejun Heo
2011-04-22 2:33 ` Shaohua Li
2011-04-26 12:10 ` Tejun Heo
2011-04-26 19:02 ` Hugh Dickins
2011-04-27 10:28 ` Tejun Heo
2011-04-27 5:43 ` Shaohua Li
2011-04-27 10:20 ` Tejun Heo
2011-04-28 3:28 ` Shaohua Li
2011-04-28 10:09 ` Tejun Heo
2011-04-28 14:11 ` Christoph Lameter
2011-04-28 14:23 ` Tejun Heo
2011-04-28 14:30 ` Tejun Heo
2011-04-28 14:58 ` Christoph Lameter
2011-04-28 14:42 ` Christoph Lameter
2011-04-28 14:44 ` Tejun Heo
2011-04-28 14:52 ` Christoph Lameter
2011-04-28 14:56 ` Tejun Heo
2011-04-28 15:05 ` Christoph Lameter
2011-04-28 15:12 ` Tejun Heo
2011-04-28 15:22 ` Christoph Lameter
2011-04-28 15:31 ` Tejun Heo
2011-04-28 15:40 ` Tejun Heo
2011-04-28 15:47 ` Christoph Lameter
2011-04-28 15:48 ` Eric Dumazet
2011-04-28 15:59 ` Eric Dumazet
2011-04-28 16:17 ` Christoph Lameter
2011-04-28 16:35 ` Eric Dumazet
2011-04-28 16:52 ` Christoph Lameter
2011-04-28 16:59 ` Eric Dumazet
2011-04-29 8:52 ` Tejun Heo
2011-04-29 8:32 ` Shaohua Li
2011-04-29 8:19 ` Shaohua Li
2011-04-29 8:44 ` Tejun Heo
2011-04-29 14:02 ` Christoph Lameter
2011-04-29 14:03 ` Christoph Lameter
2011-04-29 14:18 ` Tejun Heo
2011-04-29 14:25 ` Christoph Lameter
2011-04-29 14:43 ` Tejun Heo
2011-04-29 14:55 ` Christoph Lameter
2011-05-05 4:08 ` Shaohua Li
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=20110421144300.GA22898@htj.dyndns.org \
--to=tj@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=cl@linux.com \
--cc=eric.dumazet@gmail.com \
--cc=linux-mm@kvack.org \
--cc=shaohua.li@intel.com \
/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).