From: Dmitry Torokhov <dmitry.torokhov@gmail.com>
To: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
Alan Stern <stern@rowland.harvard.edu>,
Zhang Rui <rui.zhang@intel.com>,
LKML <linux-kernel@vger.kernel.org>,
ACPI Devel Maling List <linux-acpi@vger.kernel.org>,
pm list <linux-pm@lists.linux-foundation.org>
Subject: Re: Async suspend-resume patch w/ completions (was: Re: Async suspend-resume patch w/ rwsems)
Date: Sat, 19 Dec 2009 15:22:20 -0800 [thread overview]
Message-ID: <43A402BB-6AB3-4127-A441-D53EDE09F22E@gmail.com> (raw)
In-Reply-To: <200912200010.19899.rjw@sisk.pl>
On Dec 19, 2009, at 3:10 PM, "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> On Saturday 19 December 2009, Dmitry Torokhov wrote:
>> On Dec 19, 2009, at 1:33 PM, "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
>>
>>> On Saturday 19 December 2009, Dmitry Torokhov wrote:
>>>> On Fri, Dec 18, 2009 at 11:43:29PM +0100, Rafael J. Wysocki wrote:
>>>>> On Wednesday 16 December 2009, Dmitry Torokhov wrote:
>>>>>> On Wed, Dec 16, 2009 at 03:11:05AM +0100, Rafael J. Wysocki
>>>>>> wrote:
>>>>>>> On Tuesday 15 December 2009, Linus Torvalds wrote:
>>>>>>>>
>>>>>>>> On Tue, 15 Dec 2009, Rafael J. Wysocki wrote:
>>>>>>>>>>
>>>>>>>>>> Give a real example that matters.
>>>>>>>>>
>>>>>>>>> I'll try. Let -> denote child-parent relationships and assume
>>>>>>>>> dpm_list looks
>>>>>>>>> like this:
>>>>>>>>
>>>>>>>> No.
>>>>>>>>
>>>>>>>> I mean something real - something like
>>>>>>>>
>>>>>>>> - if you run on a non-PC with two USB buses behind non-PCI
>>>>>>>> controllers.
>>>>>>>>
>>>>>>>> - device xyz.
>>>>>>>>
>>>>>>>>> If this applies to _resume_ only, then I agree, but the
>>>>>>>>> Arjan's data clearly
>>>>>>>>> show that serio devices take much more time to suspend than
>>>>>>>>> USB.
>>>>>>>>
>>>>>>>> I mean in general - something where you actually have hard data
>>>>>>>> that some
>>>>>>>> device really needs anythign more than my one-liner, and really
>>>>>>>> _needs_
>>>>>>>> some complex infrastructure.
>>>>>>>>
>>>>>>>> Not "let's imagine a case like xyz".
>>>>>>>
>>>>>>> As I said I would, I made some measurements.
>>>>>>>
>>>>>>> I measured the total time of suspending and resuming devices as
>>>>>>> shown by the
>>>>>>> code added by this patch:
>>>>>>> http://git.kernel.org/?p=linux/kernel/git/rafael/suspend-2.6.git;a=commitdiff_plain;h=c1b8fc0a8bff7707c10f31f3d26bfa88e18ccd94;hp=087dbf5f079f1b55cbd3964c9ce71268473d5b67
>>>>>>> on two boxes, HP nx6325 and MSI Wind U100 (hardware-wise they
>>>>>>> are quite
>>>>>>> different and the HP was running 64-bit kernel and user space).
>>>>>>>
>>>>>>> I took four cases into consideration:
>>>>>>> (1) synchronous suspend and resume (/sys/power/pm_async = 0)
>>>>>>> (2) asynchronous suspend and resume as introduced by the async
>>>>>>> branch at:
>>>>>>> http://git.kernel.org/?p=linux/kernel/git/rafael/suspend-2.6.git;a=shortlog;h=refs/heads/async
>>>>>>> (3) asynchronous suspend and resume like in (2), but with your
>>>>>>> one-liner setting
>>>>>>> the power.async_suspend flag for PCI bridges on top
>>>>>>> (4) asynchronous suspend and resume like in (2), but with an
>>>>>>> extra patch that
>>>>>>> is appended on top
>>>>>>>
>>>>>>> For those tests I set power.async_suspend for all USB devices,
>>>>>>> all serio input
>>>>>>> devices, the ACPI battery and the USB PCI controllers (to see
>>>>>>> the impact of the
>>>>>>> one-liner, if any).
>>>>>>>
>>>>>>> I carried out 5 consecutive suspend-resume cycles (started from
>>>>>>> under X) on
>>>>>>> each box in each case, and the raw data are here (all times in
>>>>>>> milliseconds):
>>>>>>> http://www.sisk.pl/kernel/data/async-suspend.pdf
>>>>>>>
>>>>>>> The summarized data are below (the "big" numbers are averages
>>>>>>> and the +/-
>>>>>>> numbers are standard deviations, all in milliseconds):
>>>>>>>
>>>>>>> HP nx6325 MSI Wind U100
>>>>>>>
>>>>>>> sync suspend 1482 (+/- 40) 1180 (+/- 24)
>>>>>>> sync resume 2955 (+/- 2) 3597 (+/- 25)
>>>>>>>
>>>>>>> async suspend 1553 (+/- 49) 1177 (+/- 32)
>>>>>>> async resume 2692 (+/- 326) 3556 (+/- 33)
>>>>>>>
>>>>>>> async+one-liner suspend 1600 (+/- 39) 1212 (+/- 41)
>>>>>>> async+one-liner resume 2692 (+/- 324) 3579 (+/- 24)
>>>>>>>
>>>>>>> async+extra suspend 1496 (+/- 37) 1217 (+/- 38)
>>>>>>> async+extra resume 1859 (+/- 114) 1923 (+/- 35)
>>>>>>>
>>>>>>> So, in my opinion, with the above set of "async" devices, it
>>>>>>> doesn't
>>>>>>> make sense to do async suspend at all, because the sync suspend
>>>>>>> is actually
>>>>>>> the fastest on both machines.
>>>>>>
>>>>>> I think the async suspend is not asynchronous enough then - what
>>>>>> kind of
>>>>>> time do you get if you simply comment out call to psmouse_reset()
>>>>>> in
>>>>>> drivers/input/mouse/psmouse-base.c:psmouse_cleanup()? (Just for
>>>>>> testing
>>>>>> purposes only, I don't think we want to do that by default.)
>>>>>
>>>>> The problem apparently is that the i8042 suspend/resume is
>>>>> synchronous.
>>>>>
>>>>> Do you think it's safe to mark it as asynchronous?
>>>>>
>>>>
>>>> Umm.. there lie dragons. There is an implicit relationship between
>>>> i8042
>>>> and PNP/ACPI devices representing keyboard and mouse ports, and I
>>>> am not
>>>> sure how happy i8042 (and most importantly the BIOS) will be if
>>>> they get
>>>> shut down before i8042. Also there is EC which is in theory
>>>> independent
>>>> but in practice not so much.
>>>
>>> I see.
>>>
>>> Is this possible to identify ACPI devices that should wait for the
>>> i8042
>>> suspend and that should be waited for by it on resume?
>>
>> We could try to add some dependencies while discovering PNP to get
>> KBC
>> addresses in i8042 but we need tomake sure we do it even in presence
>> of i8042.nopnp.
>
> Well, I guess this is the example of the off-tree dependencies that
> actually
> matter Linus wanted. :-)
>
> I guess there are quite a few devices that can depend on the i8042 in
> principle, is this correct?
The devices that depend on i8042 are serio ports that are it's
children. I8042 itself may have indirect dependency on a couple of PNP
devices.
>
I hope this answers your question...
--
Dmitry
next prev parent reply other threads:[~2009-12-19 23:22 UTC|newest]
Thread overview: 235+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-12-05 21:16 [GIT PULL] PM updates for 2.6.33 Rafael J. Wysocki
2009-12-05 21:43 ` Linus Torvalds
2009-12-05 21:58 ` Linus Torvalds
2009-12-05 23:55 ` Rafael J. Wysocki
2009-12-06 0:45 ` Arjan van de Ven
2009-12-06 1:26 ` Rafael J. Wysocki
2009-12-06 1:58 ` Arjan van de Ven
2009-12-06 8:39 ` Ingo Molnar
2009-12-06 0:48 ` Linus Torvalds
2009-12-06 1:54 ` Rafael J. Wysocki
2009-12-06 1:57 ` Rafael J. Wysocki
2009-12-06 2:05 ` Linus Torvalds
2009-12-06 2:36 ` Rafael J. Wysocki
2009-12-06 15:23 ` Alan Stern
2009-12-06 19:04 ` [linux-pm] " Victor Lowther
2009-12-07 3:57 ` Zhang Rui
2009-12-07 5:57 ` Linus Torvalds
2009-12-07 6:15 ` Linus Torvalds
2009-12-17 23:28 ` Benjamin Herrenschmidt
2009-12-07 6:37 ` Arjan van de Ven
2009-12-07 15:13 ` Alan Stern
2009-12-07 16:31 ` Linus Torvalds
2009-12-07 16:55 ` Linus Torvalds
2009-12-07 17:52 ` Alan Stern
2009-12-07 18:05 ` Linus Torvalds
2009-12-07 20:37 ` Alan Stern
2009-12-07 20:48 ` Linus Torvalds
2009-12-07 21:32 ` Alan Stern
2009-12-07 21:41 ` Linus Torvalds
2009-12-07 21:47 ` Rafael J. Wysocki
2009-12-07 22:01 ` Alan Stern
2009-12-07 22:06 ` Linus Torvalds
2009-12-07 22:21 ` Alan Stern
2009-12-07 22:26 ` Linus Torvalds
2009-12-07 23:16 ` Alan Stern
2009-12-07 22:02 ` Rafael J. Wysocki
2009-12-07 22:16 ` Linus Torvalds
2009-12-07 23:51 ` Rafael J. Wysocki
2009-12-08 3:27 ` Alan Stern
2009-12-08 12:23 ` Async resume patch (was: Re: [GIT PULL] PM updates for 2.6.33) Rafael J. Wysocki
2009-12-08 12:35 ` Rafael J. Wysocki
2009-12-08 15:35 ` Linus Torvalds
2009-12-08 15:55 ` Alan Stern
2009-12-08 16:42 ` Linus Torvalds
2009-12-08 18:08 ` Alan Stern
2009-12-08 18:41 ` Linus Torvalds
2009-12-08 18:52 ` Linus Torvalds
2009-12-08 19:34 ` Alan Stern
2009-12-08 19:30 ` Alan Stern
2009-12-08 20:48 ` Linus Torvalds
2009-12-08 21:32 ` Alan Stern
2009-12-08 21:52 ` Christian Borntraeger
2009-12-08 22:16 ` Linus Torvalds
2009-12-09 19:06 ` Alan Stern
2009-12-09 21:52 ` Linus Torvalds
2009-12-08 19:44 ` Rafael J. Wysocki
2009-12-08 20:16 ` Alan Stern
2009-12-08 20:30 ` Rafael J. Wysocki
2009-12-08 20:44 ` Alan Stern
2009-12-08 20:52 ` Rafael J. Wysocki
2009-12-08 21:40 ` Alan Stern
2009-12-08 21:48 ` spinlock in completion_done() (was: Re: Async resume patch (was: Re: [GIT PULL] PM updates for 2.6.33)) Rafael J. Wysocki
2009-12-09 9:29 ` Ingo Molnar
2009-12-09 22:37 ` Rafael J. Wysocki
2009-12-10 7:59 ` Ingo Molnar
2009-12-11 4:10 ` Dave Chinner
2009-12-11 7:54 ` Ingo Molnar
2009-12-12 23:07 ` [PATCH] sched: Make wakeup side variants of completion API irq safe (was: Re: spinlock in completion_done()) Rafael J. Wysocki
2009-12-13 7:36 ` [tip:sched/urgent] sched: Make wakeup side and atomic variants of completion API irq safe tip-bot for Rafael J. Wysocki
2009-12-08 22:18 ` Async resume patch (was: Re: [GIT PULL] PM updates for 2.6.33) Linus Torvalds
2009-12-09 2:11 ` Alan Stern
2009-12-08 21:08 ` Linus Torvalds
2009-12-08 21:13 ` Linus Torvalds
2009-12-08 22:07 ` Alan Stern
2009-12-08 22:30 ` Rafael J. Wysocki
2009-12-09 2:23 ` Alan Stern
2009-12-09 21:56 ` Rafael J. Wysocki
2009-12-09 22:27 ` Alan Stern
2009-12-08 22:32 ` Linus Torvalds
2009-12-09 2:35 ` Alan Stern
2009-12-09 2:54 ` Linus Torvalds
2009-12-09 15:24 ` Alan Stern
2009-12-09 15:38 ` Linus Torvalds
2009-12-09 15:57 ` Alan Stern
2009-12-25 17:09 ` Pavel Machek
2009-12-09 13:38 ` Mark Brown
2009-12-09 15:49 ` Alan Stern
2009-12-09 16:02 ` Mark Brown
2009-12-09 16:23 ` Alan Stern
2009-12-09 16:46 ` Mark Brown
2009-12-09 16:57 ` Linus Torvalds
2009-12-09 17:45 ` Mark Brown
2009-12-09 17:57 ` Linus Torvalds
2009-12-09 18:27 ` Mark Brown
2009-12-09 17:10 ` Alan Stern
2009-12-09 17:19 ` Linus Torvalds
2009-12-09 18:08 ` Mark Brown
2009-12-08 21:04 ` Linus Torvalds
2009-12-08 21:40 ` Rafael J. Wysocki
2009-12-08 22:03 ` Rafael J. Wysocki
2009-12-08 22:55 ` Async suspend-resume patch w/ rwsems " Rafael J. Wysocki
2009-12-08 23:24 ` Rafael J. Wysocki
2009-12-09 20:15 ` Alan Stern
2009-12-09 22:18 ` Rafael J. Wysocki
2009-12-09 22:38 ` Alan Stern
2009-12-09 23:18 ` Async suspend-resume patch w/ completions (was: Re: Async suspend-resume patch w/ rwsems) Rafael J. Wysocki
2009-12-10 2:51 ` Linus Torvalds
2009-12-10 19:40 ` Rafael J. Wysocki
2009-12-10 23:30 ` Linus Torvalds
2009-12-11 1:02 ` Rafael J. Wysocki
2009-12-11 1:25 ` Linus Torvalds
2009-12-11 3:42 ` Alan Stern
2009-12-11 22:17 ` Rafael J. Wysocki
2009-12-12 0:38 ` Alan Stern
2009-12-11 22:11 ` Rafael J. Wysocki
2009-12-11 22:31 ` Linus Torvalds
2009-12-11 23:48 ` Rafael J. Wysocki
2009-12-11 23:53 ` Linus Torvalds
2009-12-12 17:48 ` Rafael J. Wysocki
2009-12-12 18:54 ` Linus Torvalds
2009-12-12 22:34 ` Rafael J. Wysocki
2009-12-12 22:40 ` Rafael J. Wysocki
2009-12-14 18:21 ` Linus Torvalds
2009-12-14 22:11 ` Rafael J. Wysocki
2009-12-14 22:41 ` Linus Torvalds
2009-12-14 22:43 ` Linus Torvalds
2009-12-14 23:18 ` Rafael J. Wysocki
2009-12-15 0:10 ` Linus Torvalds
2009-12-15 0:11 ` Linus Torvalds
2009-12-15 11:14 ` Rafael J. Wysocki
2009-12-15 15:31 ` Linus Torvalds
2009-12-15 11:03 ` Rafael J. Wysocki
2009-12-15 15:26 ` Linus Torvalds
2009-12-15 15:55 ` Alan Stern
2009-12-15 16:28 ` Linus Torvalds
2009-12-15 18:57 ` Linus Torvalds
2009-12-15 20:26 ` Alan Stern
2009-12-15 21:26 ` Rafael J. Wysocki
2009-12-15 22:01 ` Alan Stern
2009-12-15 21:54 ` Linus Torvalds
2009-12-15 22:27 ` Alan Stern
2009-12-16 2:11 ` Rafael J. Wysocki
2009-12-16 6:40 ` Dmitry Torokhov
2009-12-18 22:43 ` Rafael J. Wysocki
2009-12-19 19:59 ` Dmitry Torokhov
2009-12-19 21:33 ` Rafael J. Wysocki
2009-12-19 22:29 ` Rafael J. Wysocki
2009-12-19 22:43 ` Dmitry Torokhov
2009-12-19 22:47 ` Dmitry Torokhov
2009-12-19 23:10 ` Rafael J. Wysocki
2009-12-19 23:22 ` Dmitry Torokhov [this message]
2009-12-19 23:33 ` Rafael J. Wysocki
2009-12-19 23:23 ` Linus Torvalds
2009-12-19 23:40 ` Rafael J. Wysocki
2009-12-19 23:46 ` Linus Torvalds
2009-12-19 23:47 ` Linus Torvalds
2009-12-19 23:54 ` Rafael J. Wysocki
2009-12-19 23:53 ` Rafael J. Wysocki
2009-12-20 0:09 ` Linus Torvalds
2009-12-20 0:35 ` Rafael J. Wysocki
2009-12-20 2:41 ` Dmitry Torokhov
2009-12-20 19:25 ` [linux-pm] " Rafael J. Wysocki
2009-12-21 7:39 ` [linux-pm] Async suspend-resume patch w/ completions (was: Re: Async?suspend-resume " Dmitry Torokhov
2009-12-21 11:20 ` Vojtech Pavlik
2009-12-20 2:45 ` Async suspend-resume patch w/ completions (was: Re: Async suspend-resume " Dmitry Torokhov
2009-12-20 3:59 ` Alan Stern
2009-12-20 12:52 ` Rafael J. Wysocki
2009-12-20 17:12 ` Alan Stern
2009-12-20 18:10 ` Rafael J. Wysocki
2009-12-20 19:38 ` Alan Stern
2009-12-20 19:51 ` Rafael J. Wysocki
2009-12-16 15:22 ` Alan Stern
2009-12-16 19:26 ` Rafael J. Wysocki
2009-12-16 15:47 ` Linus Torvalds
2009-12-16 19:27 ` Rafael J. Wysocki
2009-12-16 20:59 ` Linus Torvalds
2009-12-16 21:57 ` Rafael J. Wysocki
2009-12-16 22:11 ` Linus Torvalds
2009-12-16 22:33 ` Rafael J. Wysocki
2009-12-16 23:04 ` Alan Stern
2009-12-16 23:18 ` Rafael J. Wysocki
2009-12-17 1:30 ` Rafael J. Wysocki
2009-12-17 1:49 ` Rafael J. Wysocki
2009-12-17 20:06 ` Alan Stern
2009-12-17 20:36 ` Rafael J. Wysocki
2009-12-18 1:51 ` Rafael J. Wysocki
2009-12-18 17:26 ` Alan Stern
2009-12-19 21:41 ` Rafael J. Wysocki
2009-12-20 3:48 ` Alan Stern
2009-12-20 12:55 ` Rafael J. Wysocki
2009-12-18 23:42 ` Rafael J. Wysocki
2009-12-13 13:08 ` Rafael J. Wysocki
2009-12-13 17:30 ` Alan Stern
2009-12-13 19:02 ` [linux-pm] " Alan Stern
2009-12-12 0:43 ` Alan Stern
2009-12-12 17:35 ` Rafael J. Wysocki
2009-12-10 15:31 ` Alan Stern
2009-12-10 15:45 ` Linus Torvalds
2009-12-10 18:37 ` Alan Stern
2009-12-10 23:51 ` Linus Torvalds
2009-12-10 21:14 ` Rafael J. Wysocki
2009-12-10 22:17 ` Alan Stern
2009-12-10 23:45 ` Rafael J. Wysocki
2009-12-07 15:15 ` [GIT PULL] PM updates for 2.6.33 Rafael J. Wysocki
2009-12-07 16:37 ` Linus Torvalds
2009-12-07 20:47 ` Rafael J. Wysocki
2009-12-07 20:56 ` Linus Torvalds
2009-12-07 5:20 ` Linus Torvalds
2009-12-07 15:42 ` Alan Stern
2009-12-06 19:35 ` Arjan van de Ven
2009-12-06 19:58 ` Linus Torvalds
2009-12-06 20:18 ` Arjan van de Ven
2009-12-06 21:08 ` Linus Torvalds
2009-12-06 22:54 ` Dmitry Torokhov
2009-12-07 0:55 ` Arjan van de Ven
2009-12-07 2:27 ` Dmitry Torokhov
2009-12-07 5:26 ` Arjan van de Ven
2009-12-07 6:00 ` Dmitry Torokhov
2009-12-21 9:01 ` Pavel Machek
2009-12-07 1:18 ` Arjan van de Ven
2009-12-07 2:27 ` Dmitry Torokhov
2009-12-07 5:31 ` Arjan van de Ven
2009-12-07 6:15 ` Dmitry Torokhov
2009-12-07 6:31 ` Arjan van de Ven
2009-12-07 6:32 ` Dmitry Torokhov
2009-12-07 15:17 ` Rafael J. Wysocki
2009-12-06 20:36 ` Alan Stern
2009-12-06 21:17 ` Arjan van de Ven
2009-12-06 21:46 ` Alan Stern
2009-12-06 21:57 ` Arjan van de Ven
2009-12-06 22:04 ` Alan Stern
2009-12-06 0:29 ` Rafael J. Wysocki
2009-12-06 0:52 ` Linus Torvalds
2009-12-06 1:24 ` Rafael J. Wysocki
2009-12-06 1:50 ` Linus Torvalds
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=43A402BB-6AB3-4127-A441-D53EDE09F22E@gmail.com \
--to=dmitry.torokhov@gmail.com \
--cc=linux-acpi@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@lists.linux-foundation.org \
--cc=rjw@sisk.pl \
--cc=rui.zhang@intel.com \
--cc=stern@rowland.harvard.edu \
--cc=torvalds@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).