From: Tejun Heo <tj@kernel.org>
To: Borislav Petkov <bp@alien8.de>,
Andrew Morton <akpm@linux-foundation.org>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 25/25] ipc: don't use [delayed_]work_pending()
Date: Mon, 24 Dec 2012 11:07:23 -0800 [thread overview]
Message-ID: <20121224190723.GE11817@htj.dyndns.org> (raw)
In-Reply-To: <20121224185555.GA5344@liondog.tnic>
Hello, Borislav.
On Mon, Dec 24, 2012 at 07:55:55PM +0100, Borislav Petkov wrote:
> Basically, with the amount of bloat we're adding to the kernel, the
> couple of instructions we're adding here and there and think they won't
> harm us, tends to crop up with time until we're too damn slow to do
> anything.
Yeah, I think we need them for the operations to hold w/o outer
synchronization. If you can think of ways to optimize them, please be
my guest.
> Now, you're saying that optimizing cold paths is something that calls
> for serious reconsideration of a person's priorities, and I'm saying
> just don't do anything in code that's not necessary. Fullstop. If this
Not doing anything that's not necessary is an extreme and not
necessary what you want in cold paths. If doing something extra makes
your code much simpler and more maintainable and the path is cold
enough for the extra to not matter, then that's the right trade-off.
> is leading to convolutions, then the design is suboptimal and needs
> adjustment. If [delayed_]work_pending is being abused, then fix it or
> add a new primitive which only checks for pending work and doesn't
> unconditionally toggle interrupts.
"only checks for pending work" in a way which is race-free and usable
from any context is a tricky thing to do.
> > I don't think we have cases where this actually matters but it could
> > be that we can add work_pending() tests to queue_work_on(). I *think*
> > that shouldn't break work scheduling semantics. Not completely sure
> > tho. Need to think about it more.
>
> Yes, something like that.
And that's broken. It seems trivial but it really isn't and trying to
optimize things like that in cold paths is just a bad idea. Not
enough people will pay attention to them and they will stay subtly
broken for a very long time. So, having "not doing anything in code
which isn't necessary in code" as priority in cold paths is likely to
hurt you. A better one would be "doing straight-forward and simple
thing with acceptable overhead".
Thanks.
--
tejun
next prev parent reply other threads:[~2012-12-24 19:07 UTC|newest]
Thread overview: 83+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-12-22 1:56 [PATCHSET] workqueue: " Tejun Heo
2012-12-22 1:56 ` [PATCH 01/25] charger_manager: " Tejun Heo
2013-01-05 22:11 ` Anton Vorontsov
2012-12-22 1:56 ` [PATCH 02/25] ab8500_charger: " Tejun Heo
[not found] ` <CACRpkdYE3F22rnW+ZRhOWT4VGRhG2CSnb7Rj3gdCosTK8wLQmA@mail.gmail.com>
[not found] ` <50EAA53F.1000304@stericsson.com>
2013-01-07 10:48 ` Arun MURTHY
2013-01-08 14:30 ` Linus Walleij
2012-12-22 1:56 ` [PATCH 03/25] sja1000: " Tejun Heo
2012-12-22 8:01 ` David Miller
2012-12-28 21:40 ` Tejun Heo
2012-12-22 1:56 ` [PATCH 04/25] ipw2x00: simplify scan_event handling Tejun Heo
2013-01-27 21:02 ` Stanislav Yakovlev
2013-02-09 19:31 ` Tejun Heo
2012-12-22 1:56 ` [PATCH 05/25] devfreq: don't use [delayed_]work_pending() Tejun Heo
2012-12-22 1:56 ` [PATCH 06/25] libertas: " Tejun Heo
2012-12-22 1:56 ` [PATCH 07/25] mwifiex: " Tejun Heo
2012-12-22 22:29 ` Bing Zhao
2012-12-28 21:41 ` Tejun Heo
2012-12-22 1:56 ` [PATCH 08/25] thinkpad_acpi: " Tejun Heo
2012-12-22 23:55 ` Henrique de Moraes Holschuh
2012-12-28 21:41 ` Tejun Heo
2012-12-22 1:56 ` [PATCH 09/25] wl1251: " Tejun Heo
2012-12-22 14:14 ` Luciano Coelho
2012-12-28 21:42 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 10/25] kprobes: fix wait_for_kprobe_optimizer() Tejun Heo
2012-12-25 3:51 ` Masami Hiramatsu
2013-01-28 19:49 ` Tejun Heo
2013-01-29 11:53 ` Masami Hiramatsu
2013-02-09 19:33 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 11/25] pm: don't use [delayed_]work_pending() Tejun Heo
2012-12-22 11:53 ` Rafael J. Wysocki
2012-12-25 16:44 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 12/25] bluetooth/l2cap: " Tejun Heo
2013-01-03 22:27 ` Gustavo Padovan
2012-12-22 1:57 ` [PATCH 13/25] sound/wm8350: " Tejun Heo
2012-12-24 16:11 ` Mark Brown
2012-12-22 1:57 ` [PATCH 14/25] rfkill: " Tejun Heo
2012-12-22 20:22 ` Johannes Berg
2012-12-28 21:42 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 15/25] x86/mce: " Tejun Heo
2012-12-25 11:07 ` Borislav Petkov
2012-12-28 21:44 ` [PATCH v2 " Tejun Heo
2012-12-22 1:57 ` [PATCH 16/25] PM / Domains: " Tejun Heo
2012-12-22 11:57 ` Rafael J. Wysocki
2012-12-25 17:03 ` Tejun Heo
2012-12-25 20:33 ` Rafael J. Wysocki
2012-12-26 1:23 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 17/25] wm97xx: " Tejun Heo
2012-12-23 9:54 ` Dmitry Torokhov
2012-12-24 16:18 ` Mark Brown
2013-03-09 23:53 ` Dmitry Torokhov
2013-03-12 18:49 ` Mark Brown
2012-12-24 18:25 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 18/25] TMIO MMC: " Tejun Heo
2012-12-24 22:31 ` Guennadi Liakhovetski
2012-12-22 1:57 ` [PATCH 19/25] net/caif: " Tejun Heo
2012-12-22 1:57 ` [PATCH 20/25] wimax/i2400m: fix i2400m->wake_tx_skb handling Tejun Heo
2012-12-22 15:28 ` Perez-Gonzalez, Inaky
2013-01-04 21:19 ` Dan Williams
2013-02-09 19:35 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 21/25] tty/max3100: don't use [delayed_]work_pending() Tejun Heo
2012-12-22 4:21 ` Greg Kroah-Hartman
2012-12-28 21:44 ` Tejun Heo
2012-12-22 1:57 ` [PATCH 22/25] usb/at91_udc: " Tejun Heo
2013-01-07 16:25 ` Nicolas Ferre
2012-12-22 1:57 ` [PATCH 23/25] video/exynos: " Tejun Heo
2012-12-22 3:05 ` Kukjin Kim
2012-12-26 4:04 ` Jingoo Han
2012-12-28 21:44 ` 'Tejun Heo'
2012-12-22 1:57 ` [PATCH 24/25] debugobjects: " Tejun Heo
2012-12-22 1:57 ` [PATCH 25/25] ipc: " Tejun Heo
2012-12-22 2:15 ` Andrew Morton
2012-12-22 2:22 ` Tejun Heo
2012-12-22 11:09 ` Borislav Petkov
2012-12-24 18:33 ` Tejun Heo
2012-12-24 18:45 ` Tejun Heo
2012-12-24 19:41 ` Borislav Petkov
2012-12-25 3:29 ` Tejun Heo
2012-12-25 10:46 ` Borislav Petkov
2012-12-25 16:35 ` Tejun Heo
2012-12-24 18:55 ` Borislav Petkov
2012-12-24 19:07 ` Tejun Heo [this message]
2012-12-24 19:32 ` Borislav Petkov
2012-12-25 3:18 ` Tejun Heo
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=20121224190723.GE11817@htj.dyndns.org \
--to=tj@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=bp@alien8.de \
--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
all inboxes | Powered by JetHome®