Announcement

Collapse
No announcement yet.

Linux Kernel Power Bug Now High Importance In Ubuntu

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Linux Kernel Power Bug Now High Importance In Ubuntu

    Phoronix: Linux Kernel Power Bug Now High Importance In Ubuntu

    The bug introduced during the development of the Linux 2.6.38 kernel causing excessive power consumption is very real, is occurring on many different hardware platforms, and has just been deemed a bug of high importance by the Ubuntu Kernel Team. This serious regression was just made widely known on Friday in my Mobile Users Beware: Linux Has Major Power Regression article and then further detailed in The Tests Showing Ubuntu 11.04 On A Power Consumption Binge...

    http://www.phoronix.com/vr.php?view=OTM3NQ

  • #2
    scheduler

    Maybe its because of recent changes in scheduler. All in all its a shame!

    Comment


    • #3
      Ugh .. Still no evidence but the sky is certainly falling.

      Lenovo T400
      Fuduntu 14.9 (64bit)
      Kernel 2.6.38.4
      Completely idle on battery for ~15 minutes.
      6.36 Watts, 108 wakeups per second.



      Asus Eee PC 1015PEM
      Fuduntu 14.9 (32bit)
      Kernel 2.6.38.4
      Completely idle on battery (as before)
      7.31 Watts, 135.8 wakeups per second.



      As you can see from the images, even my Thinkpad is using only 6ish watts, and well under 200 wakeups per second (closer to 100 really!)..

      The only reason there are 33 reports is because you asked people in your forum to report it!

      http://phoronix.com/forums/showthrea...105#post201105

      Lets take a look at the data supplied in that bug report

      Code:
      Wakeups-from-idle per second : 542.9    interval: 10.0s
      Power usage (ACPI estimate): 14.1W (2.4 hours) (long term: 14.7W,/2.3h)
      
      Top causes for wakeups:
        25.7% (155.8)   [i915] <interrupt>
        13.3% ( 80.3)D  chromium-browse
        17.9% (108.5)   [extra timer interrupt]
         5.7% ( 34.7)   [iwlagn] <interrupt>
         5.1% ( 31.2)   PS/2 keyboard/mouse/touchpad interrupt
         5.0% ( 30.2)   [kernel scheduler] Load balancing tick
         4.9% ( 29.9)   kworker/0:0
         4.4% ( 26.5)   [Rescheduling interrupts] <kernel IPI>
         3.3% ( 20.1)   compiz
         0.0% (  0.0)D  flush-8:0
         3.2% ( 19.6)   desktopcouch-se
         0.1% (  0.7)D  ntop
         1.2% (  7.3)D  thunderbird-bin
         1.7% ( 10.0)   gwibber-service
         1.6% (  9.9)   ubuntuone-syncd
         1.4% (  8.6)   [TLB shootdowns] <kernel IPI>
         1.4% (  8.3)   [acpi] <interrupt>
         0.6% (  3.9)   [ahci] <interrupt>
      The program 'flush-8:0' is writing to file 'ECRYPTFS_FNEK_ENCRYPTED.FWY-tcU' on /dev/sda1.
      This prevents the disk from going to powersave mode.
      - https://launchpadlibrarian.net/69300044/hardcopy.0

      Code:
        25.7% (155.8)   [i915] <interrupt>
        13.3% ( 80.3)D  chromium-browse
        17.9% (108.5)   [extra timer interrupt]
         5.1% ( 31.2)   PS/2 keyboard/mouse/touchpad interrupt
      These are all related.

      375.8 wakeups per second contribute to applications

      What's not reported in the bug though (and probably the root cause), top output.

      Comment


      • #4
        Oh, and .. PowerTOP version 1.13

        PowerTOP 1.97 is required to take report on and take advantage of the new powersaving features of kernel 2.6.37 and newer.

        http://lists.lesswatts.org/pipermail...ry/000509.html

        The most common of these have been built into Jupiter.

        Has the Phoronix test suite been updated to use the kernel "perf" infrastructure?

        Comment


        • #5
          So is this a bug that affects all kernels or just Ubuntu patched ones? I don't understand this fixation on Ubuntu here (and generally). If this happens in vanilla kernels it should not just be reported to Ubuntu but to upstream Kernel development and maybe other distributors. Or did I miss something?

          Comment


          • #6
            Originally posted by fewt View Post
            Ugh .. Still no evidence but the sky is certainly falling.
            Did you see the plots?

            Comment


            • #7
              Originally posted by yotambien View Post
              Did you see the plots?
              Yes. It doesn't change my opinion (or my evidence) that this is not a kernel problem.

              Comment


              • #8
                Originally posted by fewt View Post
                Yes. It doesn't change my opinion (or my evidence) that this is not a kernel problem.
                It was rhetorical. I assume you read the article and saw the plots. How do you explain them?

                Comment


                • #9
                  Originally posted by yotambien View Post
                  It was rhetorical. I assume you read the article and saw the plots. How do you explain them?
                  I can't explain them (there isn't enough data), however I believe that the problem could be:

                  A) The build options between kernel versions may be different
                  B) Something in userspace
                  C) The Phoronix test suite itself
                  (2.6.37 pretty much required a new version of Powertop, Phoronix test suite probably also needs an update)
                  D) The kernel doesn't tune itself for power / battery (requiring a userspace tool like Jupiter)

                  It is most likely that this is just a misconfiguration blown grossly out of proportion.

                  Comment


                  • #10
                    Originally posted by fewt View Post
                    C) The Phoronix test suite itself
                    (2.6.37 pretty much required a new version of Powertop, Phoronix test suite probably also needs an update)
                    Nope, it doesn't need any update. And users have reported the issue independent of PTS - hence the original Canonical bug report completely independent of that, etc. Most users don't even know how to use the power monitoring features in PTS.
                    Michael Larabel
                    http://www.michaellarabel.com/

                    Comment


                    • #11
                      Originally posted by fewt View Post
                      It is most likely that this is just a misconfiguration blown grossly out of proportion.
                      So the mainline kernel PPA, Ubuntu kernel config, Arch Linux config, openSUSE config, Fedora config, and other peoples configurations, etc all went out of whack? There's been independent tests in all of those cases.
                      Michael Larabel
                      http://www.michaellarabel.com/

                      Comment


                      • #12
                        Originally posted by Michael View Post
                        Nope, it doesn't need any update. And users have reported the issue independent of PTS - hence the original Canonical bug report completely independent of that, etc. Most users don't even know how to use the power monitoring features in PTS.
                        Are you sure the users reporting the issue have correctly identified the problem? How are you confirming verification of the bug, what does the dataset look like? The Launchpad bug report I saw is questionable at best.

                        Comment


                        • #13
                          Originally posted by Michael View Post
                          So the mainline kernel PPA, Ubuntu kernel config, Arch Linux config, openSUSE config, Fedora config, and other peoples configurations, etc all went out of whack? There's been independent tests in all of those cases.
                          What data has been collected to verify these claims though? I don't mean any disrespect, I would just like to see a flood of data to analyze. The dataset that I posted in my initial comment seems to prove you wrong. If nothing else, it could be a good indicator of where to look for the root cause if it is a real problem by analyzing the content of /usr/lib/jupiter/kernel/{battery,power} since the issue does not appear to exist in Fuduntu with kernel 2.6.38.4.

                          Comment


                          • #14
                            Originally posted by fewt View Post
                            I can't explain them (there isn't enough data), however I believe that the problem could be:

                            A) The build options between kernel versions may be different
                            B) Something in userspace
                            C) The Phoronix test suite itself
                            (2.6.37 pretty much required a new version of Powertop, Phoronix test suite probably also needs an update)
                            D) The kernel doesn't tune itself for power / battery (requiring a userspace tool like Jupiter)

                            It is most likely that this is just a misconfiguration blown grossly out of proportion.

                            A) OK. In that case instead of a regression in the linux kernel itself we'd have a regression in the way one (more?) distribution are configuring their kernels. From the point of view of the users, it makes little practical difference.

                            B) I don't know. Maybe, I guess? It would have to be quite an important component of the system, and again, for users it would make not much difference. The way to test this would have to install 2 kernel versions. Some user in the forum did this and found a difference. I did and found none (but I can't measure anything).

                            C) It's not like 2.6.37 required a new version of PowerTop. PowerTop required a recent kernel for some of its features to work. I don't see how this is relevant.

                            D) That the kernel doesn't tweak itself is a constant in these tests, therefore it can't explain the observed differences.

                            We'll know soon enough, I guess.

                            Comment


                            • #15
                              I'd just like to stress that thanks to the Phoronix Test Suite and how its widely spread, this problem can be narrowed down to when exactly the issue was introduced much faster. Without it the amount of code to inspect would have been drastically more important, IMHO.

                              Thanks to Michael for inspecting this issue!

                              Comment

                              Working...
                              X