Re: High CPU temp on Athlon MP w/ recent 2.6 kernels

From: Pasi Savolainen (psavo_at_iki.fi)
Date: 04/07/04

  • Next message: Sergiy Lozovsky: "Re: kernel stack challenge"
    To: linux-kernel@vger.kernel.org
    Date:	Tue, 6 Apr 2004 23:27:37 +0000 (UTC)
    
    

    * Joerg Sommrey <jo@sommrey.de>:
    > On Tue, Apr 06, 2004 at 04:26:37PM -0400, Gene Heskett wrote:
    >>
    >> But join the 70C club, that AMD athlon keeps itself at a medium simmer
    >> full time. Mine has been running 67-72C for 3 years now. Strangly,
    >> shutting down setiathome doesn't cool it by more than a couple
    >> degrees C. And, its got a $50 all copper Glaciator cooler on it,
    >> heavy heavy heavy.
    >
    > That's not quite my point. I am not afraid of running my athlons at
    > 70C. I just don't want to. With Debian Woody they ran at <40C, which
    > is impressing IMHO. An upgrade to Sarge raised the temp for about 5K,
    > which is still very cool. This temperature didn't change when I
    > upgraded to an early 2.6 kernel. Just after 2.6.3-mm4 there was this
    > jump for 10K that I just do not understand. It doesn't hurt the athlons
    > but seems unnecessary to me.

    At least on A7M266-D lmsensors read thermal sensors very wrong. I
    haven't got time to contact devs with that, but I do know for sure that
    amd76x_pm really does make cooling calls, even in 2.6.5-rc3-mm3
    (There should be /sys/devices/pci0000\:00/0000\:00\:00.0/C2_cnt file,
    which tells how many times has amd76x_pm really made the disconnection call).

    One issue is that from some kernel version amd76x_pm's idle() is called
    upto 3.5x times more often when there's some audio activity. So in
    effect number of calls to default_idle() jumps from 1100Hz to 3800Hz.
    (this is reproducible with 'rhytmbox' -application, but not with xmms.
    AFAIK my xmms uses OSS emulation and rhytmbox is native alsa.)

    Ahem. Could you actually try:
    echo 3 > /sys/devices/pci0000\:00/0000\:00\:00.0/lazy_idle

    This could help gaining 5-8°C. HZ changed from 100 to 1000 in 2.6, so
    amd76x_pm old default doesn't apply overly well here.

    There's some funniness going on with this tunable. It doesn't really
    affect how many times/second we call amd76x_pm.idle(), but rather how
    easily we go into sleep (no sleep if both CPU's aren't idle).
    With lazy_idle at 3 I get bad distortions with bttv card. with 3000 they
    disappear, but so does the thermal throttling :)

    (Sorry for lack of coherence right now)

    -- 
       Psi -- <http://www.iki.fi/pasi.savolainen>
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at  http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at  http://www.tux.org/lkml/
    

  • Next message: Sergiy Lozovsky: "Re: kernel stack challenge"

    Relevant Pages