From: bugzilla-daemon@bugzilla.kernel.org
To: cpufreq@vger.kernel.org
Subject: [Bug 49231] New: Single CPU bound process results in non-optimal turbo boost configuration
Date: Sun, 21 Oct 2012 23:56:46 +0000 (UTC) [thread overview]
Message-ID: <bug-49231-12968@https.bugzilla.kernel.org/> (raw)
https://bugzilla.kernel.org/show_bug.cgi?id=49231
Summary: Single CPU bound process results in non-optimal turbo
boost configuration
Product: Power Management
Version: 2.5
Kernel Version: 3.4.4+
Platform: All
OS/Version: Linux
Tree: Mainline
Status: NEW
Severity: normal
Priority: P1
Component: cpufreq
AssignedTo: cpufreq@vger.kernel.org
ReportedBy: ras243-korg@yahoo.com
Regression: No
Created an attachment (id=84251)
--> (https://bugzilla.kernel.org/attachment.cgi?id=84251)
3.4.13 config file
Firstly I hope I've classified this issue correctly as it's a bit difficult to
nail down exactly what's at fault.
I have a new i7-3930k CPU on an Intel DX79SR motherboard which doesn't appear
to be handling the turbo boost function correctly under Linux. This is a 6
core CPU which is nominally 3.2GHz but can boost itself to somewhere between
3.5GHz and 3.8GHz. With 6 processes running it'll run at 3.5GHz which is fine.
The problem is that when only a single process is active it'll still only run
at 3.5GHz despite being able to go to 3.8GHz. The problem appears to be that a
single job is pulling all the cores out of C7 state and into at least C1 state
and once there, they're all are considered active and hence the fully utilised
CPU is unable to go beyond 3.5GHz.
I am unsure if the problem is with the scheduler, ACPI, or something in my
configuration. I can confirm that a 3.4.4 kernel configuration which works
correctly on a Xeon X5660 doesn't work correctly with the i7-3930k. I've tried
various kernels between 3.4.4 and 3.6.2 but all result in the same behaviour.
With gritted teeth I installed Windows 7 and it does appear to work correctly
there so I don't believe it's anything to do with BIOS settings.
Idle turbostat results:
core CPU %c0 GHz TSC %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6
%pc7
2.73 1.20 3.20 7.06 0.00 0.02 90.19 20.20 0.00 68.20
0.00
0 0 2.68 1.20 3.20 6.92 0.00 0.00 90.40 20.20 0.00 68.20
0.00
1 1 2.75 1.20 3.20 6.92 0.00 0.00 90.33 20.20 0.00 68.20
0.00
2 2 2.78 1.20 3.20 6.99 0.00 0.00 90.23 20.20 0.00 68.20
0.00
3 3 2.71 1.20 3.20 7.10 0.00 0.00 90.19 20.20 0.00 68.20
0.00
4 4 2.71 1.20 3.20 7.14 0.00 0.07 90.07 20.20 0.00 68.20
0.00
5 5 2.76 1.20 3.20 7.27 0.00 0.07 89.90 20.20 0.00 68.20
0.00
Single job turbostat results:
core CPU %c0 GHz TSC %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6
%pc7
18.46 3.50 3.20 81.54 0.00 0.00 0.00 0.00 0.00 0.00
0.00
0 0 2.02 3.50 3.20 97.98 0.00 0.00 0.00 0.00 0.00 0.00
0.00
1 1 2.19 3.50 3.20 97.81 0.00 0.00 0.00 0.00 0.00 0.00
0.00
2 2 2.20 3.50 3.20 97.80 0.00 0.00 0.00 0.00 0.00 0.00
0.00
3 3 2.17 3.50 3.20 97.83 0.00 0.00 0.00 0.00 0.00 0.00
0.00
4 4 100.00 3.50 3.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00
5 5 2.18 3.50 3.20 97.82 0.00 0.00 0.00 0.00 0.00 0.00
0.00
--
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
next reply other threads:[~2012-10-21 23:56 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-10-21 23:56 bugzilla-daemon [this message]
2012-10-22 13:44 ` [Bug 49231] Single CPU bound process results in non-optimal turbo boost configuration bugzilla-daemon
2012-10-22 13:57 ` bugzilla-daemon
2012-10-22 23:02 ` bugzilla-daemon
2012-10-23 2:09 ` bugzilla-daemon
2012-10-23 10:20 ` bugzilla-daemon
2012-10-24 4:21 ` bugzilla-daemon
2012-10-24 11:44 ` bugzilla-daemon
2013-01-28 23:57 ` bugzilla-daemon
2013-01-28 23:57 ` bugzilla-daemon
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=bug-49231-12968@https.bugzilla.kernel.org/ \
--to=bugzilla-daemon@bugzilla.kernel.org \
--cc=cpufreq@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;
as well as URLs for NNTP newsgroup(s).