Page 6 of 11 FirstFirst ... 45678 ... LastLast
Results 51 to 60 of 101

Thread: Mobile Users Beware: Linux Has Major Power Regression

  1. #51
    Join Date
    Feb 2011
    Posts
    11

    Default

    Quote Originally Posted by Michael View Post
    On a clean install you won't ever see that prompt unless your boot fails. Add a new line of 'set timeout=5' to /boot/grub/grub.cfg as the easiest quick way to get the menu back.

    Okay, another Sandy Bridge system... I've seen reports earlier of the power issue under SNB too.
    Thx.

    on an interesting note, I was seeing 29W idle / 109-110W load with the B2 board but I don't remember what kernel it was.

  2. #52

    Default

    BTW, those users who have the power problem under Ubuntu, click the 'This bug affects me' on https://bugs.launchpad.net/ubuntu/+s...ux/+bug/760131

  3. #53
    Join Date
    Jan 2008
    Location
    Have a good day.
    Posts
    678

    Default

    Pentium-M, Debian. It's been a while my machine seems to be running hotter. I'm 100% sure about it, but I don't have any figures for the temperature sensor is hardwired to 2 states: 75 C (which means 'fine') and 85 C (which means 'bad'). I know. Anyway, I found out that some updates to laptop-mode and cpufreq-utils overwrote their config files and the CPU was set to 'ondemand' rather than 'conservative'. In any case, that wasn't the culprit. The fan still kicks in way more often than before. Before I installed 2.6.38? I can't tell, but I guess so from the article. Too bad I must've forgotten to enable detailed power statistics on my last compile, for I don't get the wakeups count in PowerTop.

  4. #54
    Join Date
    Nov 2009
    Posts
    328

    Default

    If all this is true this is a serious problem!! But even more serious would be the fact that quality tests done by linux foundation didn't see that regression during the release cycle (WTF ).

    Increasing 10-20% power consumption on stable kernels on all linux machines around the world is a painful problem that should be avoided.

  5. #55
    Join Date
    Feb 2011
    Posts
    11

    Default

    anyways, I looked in the grub menu; only have 2.6.38-x's in it


    wonder if there's a 2.6.37.6 .deb that can be used on 11.04

  6. #56

    Default

    Quote Originally Posted by Michael View Post
    BTW, those users who have the power problem under Ubuntu, click the 'This bug affects me' on https://bugs.launchpad.net/ubuntu/+s...ux/+bug/760131
    If you look at the powertop output in that bug report, most of the interrupts are from the WIFI card, Chromium, and mouse / keyboard input.

    https://launchpadlibrarian.net/69300044/hardcopy.0

    Of course it's registering over 500 wakeups per second, it isn't idle.

  7. #57
    Join Date
    May 2008
    Posts
    92

    Default

    Power problem does NOT appear to affect me.

    My setup:

    Software: Ubuntu 11.04 Beta 2 fresh install, updated as of 2011-04-25, using Nouveau and Unity in compositing mode.

    Hardware: Compaq Presario V3010AU laptop (AMD Turion X2 1.6GHz, nVidia GeForce 6100 IGP, nVidia chipset, 2Gb DDR2-800, 100Gb 2.5" Seagate HD, purchased in 2006).

    I haven't done battery tests yet, will do in the coming days, but under normal usage, CPU fan does not sound unusual compared to Ubuntu 10.10. Wakeups seem to be the same as Ubuntu 10.10.

  8. #58
    Join Date
    Jan 2008
    Location
    Have a good day.
    Posts
    678

    Default

    I've installed the 2.6.38 kernel available from Debian repositories and I don't see anything out of the blue in PowerTop either. Also, I installed 2.6.37 from snapshot.debian.org and I have to say that the fan still tends to run quite often too. I think I overlooked the influence of the weather: it's already getting warmer where I live, so probably the effectiveness of the cooling system is degraded. In any case, without proper measurements (and an isolated environment) I can't really tell whether there's something going on or not.

  9. #59
    Join Date
    Oct 2009
    Location
    .ca
    Posts
    375

    Default

    Quote Originally Posted by yotambien View Post
    In any case, without proper measurements (and an isolated environment) I can't really tell whether there's something going on or not.
    That's why it's so dicey to find/report power regressions (or take reported ones seriously).

    I only noticed because for a couple of months now I have collectd running (.deb available) and can look at load/sensors/other system info versus time.

  10. #60
    Join Date
    Apr 2011
    Posts
    1

    Default

    Finally a reason to put my trusty old watt-meter through some use again...

    Code:
    kernel 2.6.38.4
    idle: 50W  load: 95W
    
    kernel 2.6.37.6
    idle: 45W  load: 87W
    On a i720QM which is running much more quiet now that I've switched back to 2.6.37. So it wasn't the weather after all

    Great find phoronix/Michael, tnx!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •