From: Phillip Susi <psusi@ubuntu.com>
To: todd.e.brandt@linux.intel.com
Cc: tj@kernel.org, JBottomley@parallels.com,
linux-ide@vger.kernel.org, linux-scsi@vger.kernel.org
Subject: Re: [PATCH/RESEND v2 0/2] SATA disk resume time optimization
Date: Sat, 09 Nov 2013 15:59:04 -0500 [thread overview]
Message-ID: <527EA218.6010604@ubuntu.com> (raw)
In-Reply-To: <20131109012023.GA8978@linux.intel.com>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 11/08/2013 08:20 PM, Todd E Brandt wrote:
> I tested your patches and they do function. We tried a similar
> approach a few months back where instead of waking the scsi disks
> we just set them all to runtime_suspended and skipped the resume.
> Then we let them be awakened later by read/write access just as you
> have. It's a really tempting approach, in theory, since you're
> saving both time and power by only waking those disks you know you
> need. But in practice I've found that userspace doesn't play nice.
I've just about gotten rid of all instances of user space waking the
disks on my system. The one I have left to do is to fake the IDENTIFY
DEVICE command using the cached identity info to satisfy a script that
tries to set the APM mode of the disk after resume. If I disable that
script, my extra disks stay asleep and Xwindows fires up nice and fast.
> In my experience the user layer almost always manages to wake up
> every mounted disk after resume, even if you didn't deliberately
> use them prior to suspend. The accesses can come from the file
> manager doing a scan after resume, or from any number of apps
> running on the system that decide they need to get even the
> smallest piece of information from the disks. A simple space check
> will wake them up.
By simple space check I assume you mean df? The superblocks easily
fit in the cache so that shouldn't generate any IO.
> Thus when you leave all the disks stopped, user space ends up
> triggering a traffic jam when the OS wakes back up, which makes
> disk access take even longer.
>
> My patch works very similarly to yours but just triggers an
> asynchronous wakeup to all the disks in anticipation of userspace's
> needs. We've tested it pretty heavily on ubuntu machines of all
> types and it's done well.
I don't think the wakeup is needed since ( ata ) disks normally wake
up on their own unless you enable power up in standby and pro-actively
initiating the wakeup only buys you maybe 500ms or less? What is that
compared to a typical 10 second startup time?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.14 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iQEcBAEBCgAGBQJSfqIUAAoJEJrBOlT6nu75v3sH/R374d/Sgx3DB0DVGgDch3jA
ZlR4eb5x5umw2CApGE0jbbj91/330Z5uxgr76tn6/nSRftDJ5ZgLc6dBTF1VwX4q
fqxKgNY1euIARiCL4jLxiK9JfX7hB0GtknJaMRvG4JHaSP4d0Cvhr0sbd5mpmJp7
P0TMVslJJHyIFVk0QjvisDBcFgo1onBkbVnX6B5Z6mPZXhAd+WCA3CJfiHnAK7t+
mINmlTBXnZQFXLXY2rDrmZEUCLFfTqtlprkAuGdlfXsMVYBTD31notuZ74Xbv7C7
vBJLiQ6b7dyF8eqcHoc49qqNO1n38nhRmYhIOSYgsyRFhECjVms5/mfEj9UBkiY=
=iDTY
-----END PGP SIGNATURE-----
next prev parent reply other threads:[~2013-11-09 20:59 UTC|newest]
Thread overview: 165+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-10-17 19:33 [PATCH/RESEND v2 0/2] SATA disk resume time optimization Todd E Brandt
2013-11-07 1:53 ` Phillip Susi
2013-11-07 1:57 ` [PATCH 1/3] sd: don't bother spinning up disks on resume Phillip Susi
2013-11-07 18:21 ` Douglas Gilbert
2013-11-07 21:16 ` Phillip Susi
2013-11-16 18:20 ` James Bottomley
2013-11-17 3:50 ` Phillip Susi
2013-11-17 6:43 ` James Bottomley
2013-11-17 16:15 ` Phillip Susi
2013-11-17 23:54 ` Douglas Gilbert
2013-11-18 1:06 ` Phillip Susi
2013-11-18 15:54 ` Phillip Susi
2013-11-20 14:23 ` Mark Lord
2013-11-20 14:48 ` Phillip Susi
2013-11-28 1:39 ` Phillip Susi
2013-11-18 0:09 ` James Bottomley
2013-11-18 1:11 ` Phillip Susi
2013-11-16 5:23 ` Mark Lord
2013-11-16 14:52 ` Phillip Susi
2013-11-07 1:57 ` [PATCH 2/3] libata: resume in the background Phillip Susi
2013-11-07 1:57 ` [PATCH 3/3] libata: don't start disks on resume Phillip Susi
2013-11-09 1:20 ` [PATCH/RESEND v2 0/2] SATA disk resume time optimization Todd E Brandt
2013-11-09 20:59 ` Phillip Susi [this message]
2013-11-09 21:03 ` [PATCH 0/6] Let sleeping disks lie Phillip Susi
2013-12-16 23:30 ` Phillip Susi
2013-12-16 23:30 ` [PATCH 1/6] libata: use sleep instead of standby command Phillip Susi
2013-12-16 23:30 ` [PATCH 2/6] libata: avoid waking disk for several commands Phillip Susi
2013-12-16 23:30 ` [PATCH 3/6] libata: resume in the background Phillip Susi
2014-01-10 22:26 ` Dan Williams
2014-01-11 2:25 ` Phillip Susi
2014-01-11 3:11 ` Dan Williams
2013-12-16 23:30 ` [PATCH 4/6] libata: don't start disks on resume Phillip Susi
2013-12-16 23:30 ` [PATCH 5/6] sd: don't start disks on system resume Phillip Susi
2013-12-17 6:43 ` James Bottomley
2013-12-17 15:01 ` Phillip Susi
2013-12-17 17:48 ` James Bottomley
2013-12-17 18:30 ` Phillip Susi
2013-12-16 23:30 ` [PATCH 6/6] libata: return power status in REQUEST SENSE command Phillip Susi
2013-12-17 18:02 ` Sergei Shtylyov
2013-12-17 18:35 ` Phillip Susi
2014-01-06 2:14 ` REQ_PM vs REQ_TYPE_PM_RESUME Phillip Susi
2014-01-06 7:36 ` Sujit Reddy Thumma
2014-01-06 9:15 ` Aaron Lu
2014-01-06 14:40 ` Phillip Susi
2014-01-06 15:38 ` Alan Stern
2014-01-07 2:44 ` Phillip Susi
2014-01-07 15:20 ` Alan Stern
2014-01-07 15:40 ` Phillip Susi
2014-01-07 15:56 ` Alan Stern
2014-01-09 18:29 ` Douglas Gilbert
2014-01-09 19:20 ` Phillip Susi
2014-01-10 5:23 ` Douglas Gilbert
2014-01-10 14:29 ` Phillip Susi
2014-01-07 7:49 ` Aaron Lu
2014-01-07 14:50 ` Phillip Susi
2014-01-08 1:03 ` Aaron Lu
2014-01-08 1:16 ` Phillip Susi
2014-01-08 1:32 ` Aaron Lu
2014-01-08 1:53 ` Phillip Susi
2014-01-08 2:11 ` Aaron Lu
2014-01-08 2:19 ` Phillip Susi
2014-01-08 2:36 ` Aaron Lu
2014-01-08 5:24 ` Phillip Susi
2014-01-08 7:00 ` Aaron Lu
2014-01-08 19:30 ` Phillip Susi
2014-01-07 15:25 ` Alan Stern
2014-01-07 15:43 ` Phillip Susi
2014-01-07 16:08 ` Alan Stern
2014-01-07 16:37 ` Phillip Susi
2014-01-07 18:05 ` Alan Stern
2014-01-07 18:43 ` Phillip Susi
2014-01-07 19:18 ` Alan Stern
2014-01-07 23:47 ` Phillip Susi
2014-01-08 17:46 ` Alan Stern
2014-01-08 18:31 ` Alan Stern
2014-01-08 20:44 ` Allow runtime suspend during system resume Alan Stern
2014-01-08 21:17 ` Phillip Susi
2014-01-08 21:34 ` Alan Stern
2014-01-09 10:14 ` Ulf Hansson
2014-01-09 15:41 ` Alan Stern
2014-01-08 22:55 ` Rafael J. Wysocki
2014-01-08 23:24 ` Alan Stern
2014-01-09 0:05 ` Rafael J. Wysocki
2014-01-09 15:32 ` Alan Stern
2014-01-09 15:50 ` Phillip Susi
2014-01-09 16:08 ` Alan Stern
2014-01-09 16:30 ` Phillip Susi
2014-01-09 17:04 ` Alan Stern
2014-01-10 1:25 ` Rafael J. Wysocki
2014-01-10 1:55 ` Phillip Susi
2014-01-10 13:35 ` Rafael J. Wysocki
2014-01-10 14:46 ` Phillip Susi
2014-01-10 15:25 ` Alan Stern
2014-01-10 23:02 ` Rafael J. Wysocki
2014-01-11 2:08 ` Phillip Susi
2014-01-11 22:50 ` Alan Stern
2014-01-12 1:50 ` Phillip Susi
2014-01-11 22:34 ` Alan Stern
2014-01-08 20:20 ` REQ_PM vs REQ_TYPE_PM_RESUME Phillip Susi
2014-01-08 21:21 ` Alan Stern
2014-01-08 21:50 ` Phillip Susi
2014-01-09 1:29 ` Aaron Lu
2014-01-09 12:17 ` Rafael J. Wysocki
2014-01-09 13:18 ` Rafael J. Wysocki
2014-01-09 15:40 ` Alan Stern
2014-01-09 15:53 ` Phillip Susi
2014-01-09 16:14 ` Alan Stern
2014-01-09 16:34 ` Phillip Susi
2014-01-09 17:06 ` Alan Stern
2014-01-16 16:59 ` Disk spin-up optimization during system resume Alan Stern
2014-01-16 18:04 ` Todd E Brandt
2014-01-16 18:33 ` Phillip Susi
2014-01-16 20:05 ` Alan Stern
2014-01-16 22:04 ` Todd E Brandt
2014-01-17 14:57 ` Alan Stern
2014-01-17 19:31 ` Dan Williams
2014-01-17 20:15 ` Alan Stern
2014-01-17 20:24 ` Tejun Heo
2014-01-17 20:55 ` Phillip Susi
2014-01-18 14:04 ` Tejun Heo
2014-01-18 1:36 ` Todd E Brandt
2014-01-18 1:41 ` Alan Stern
2014-01-18 2:15 ` Dan Williams
2014-01-18 2:33 ` Alan Stern
2014-01-18 3:24 ` Phillip Susi
2014-01-18 13:59 ` Tejun Heo
2014-01-17 20:49 ` Phillip Susi
2014-01-17 22:17 ` Dan Williams
2014-01-18 3:18 ` Phillip Susi
2014-01-18 4:09 ` Dan Williams
2014-01-18 14:08 ` Alan Stern
2014-01-21 10:12 ` Dan Williams
2014-01-21 14:37 ` Phillip Susi
2014-01-21 16:03 ` Alan Stern
2014-01-21 16:18 ` Phillip Susi
2014-01-21 16:49 ` Alan Stern
2014-01-21 15:44 ` Alan Stern
2014-01-21 16:18 ` Dan Williams
2014-01-21 16:55 ` Dan Williams
2014-01-20 12:38 ` CrashPlan Pro
2014-01-20 14:30 ` Phillip Susi
2014-01-18 1:24 ` Todd E Brandt
2014-01-18 3:20 ` Phillip Susi
2014-01-16 18:39 ` Phillip Susi
2014-01-16 20:08 ` Alan Stern
2014-01-17 10:16 ` Oliver Neukum
2014-01-17 10:21 ` Tejun Heo
2014-01-17 18:18 ` Alan Stern
2014-01-17 18:39 ` James Bottomley
2014-01-17 19:07 ` Tejun Heo
2013-11-09 21:03 ` [PATCH 1/6] libata: use sleep instead of standby command Phillip Susi
2013-11-09 21:03 ` [PATCH 2/6] libata: avoid waking disk to check power Phillip Susi
2013-11-11 13:05 ` Sergei Shtylyov
2013-11-09 21:03 ` [PATCH 3/6] sd: don't bother spinning up disks on resume Phillip Susi
2013-11-11 13:08 ` Sergei Shtylyov
2013-11-11 14:28 ` Phillip Susi
2013-11-09 21:03 ` [PATCH 4/6] libata: resume in the background Phillip Susi
2013-11-11 13:10 ` Sergei Shtylyov
2013-11-09 21:03 ` [PATCH 5/6] libata: don't start disks on resume Phillip Susi
2013-11-09 21:03 ` [PATCH 6/6] libata: fake some more commands when drive is sleeping Phillip Susi
2013-11-11 16:59 ` [PATCH/RESEND v2 0/2] SATA disk resume time optimization Todd E Brandt
2013-11-11 17:08 ` Phillip Susi
2013-12-17 12:15 ` Tejun Heo
2013-12-17 14:24 ` Phillip Susi
2013-11-27 16:18 ` Phillip Susi
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=527EA218.6010604@ubuntu.com \
--to=psusi@ubuntu.com \
--cc=JBottomley@parallels.com \
--cc=linux-ide@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=tj@kernel.org \
--cc=todd.e.brandt@linux.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.