kernel-packages team mailing list archive

[madbiologist <370173@xxxxxxxxxxxxxxxxxx>]

I neglected to mention that to use the new power management feature on R700 and newer hardware (other than APUs) requires installation of the latest AMD graphics microcode (ucode) files to /lib/firmware/radeon
These are available at http://people.freedesktop.org/~agd5f/radeon_ucode/
Get the version ending in "smc".

R700 basically means Radeon HD 4000 series and newer. However note that
according to Wikipedia and
http://xorg.freedesktop.org/wiki/RadeonFeature/#index5h2 the Mobility
Radeon HD 4225/4250 is a RV620 chip, so anyone with one of those
shouldn't need the updated firmware files.

-- 
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/370173

Title:
  laptop overheats and suddenly shuts down/off

Status in Linux ACPI client:
  New
Status in acpid:
  New
Status in cpufreqd:
  New
Status in The Linux Kernel:
  Expired
Status in lm-sensors - Utilities to read temperature/voltage/fan sensors:
  New
Status in “linux” package in Ubuntu:
  Won't Fix
Status in “acpi” package in Debian:
  New

Bug description:
  Ubuntu 9.04, fresh install on Acer Aspire 5005 / Toshiba Satellite.
  (confirmed also with 9.10 and 10.04)

  Not sure about package, but it seems the problem is:
    * with CPU frequency adjustment,
    * fan control,
    * thermal sensors,
    * control of cooling itself.

  Laptop shuts down right in the middle of CPU-greedy operation overheated. I haven't seen it before (since 7.10).
  The issue is discussed at several forums (eg. http://ubuntuforums.org/showthread.php?p=7399158 )

       *-cpu
            product: AMD Turion(tm) 64 Mobile Technology ML-37
            vendor: Advanced Micro Devices [AMD]
            physical id: 1
            bus info: cpu@0
            version: 15.4.2
            size: 2GHz
            capacity: 2GHz
            width: 64 bits
            capabilities: fpu fpu_exception wp vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt x86-64 3dnowext 3dnow up pni lahf_lm cpufreq

  also: AMD Turion Ultra 64 X2 RM-80 @ 2.1 GHz, AMD Turion 64 X2 TL-60,
  Intel(R) Celeron(R) CPU 540 @ 1.86GHz

  -----
  Replication:

  You can try:
  # sudo apt-get install stress
  # stress --cpu 16 --vm 2 --vm-bytes 128M

  But it may be the graphics card (gpu), connected to the same heat sink
  as the cpu, that causes the shutdown. And the gpu will only overheat
  when cpu is not getting very stressed/hot and thus the fan is not
  running fast enough to keep the gpu from overheating.

  It may also be reproducable by installing the ubunu alternate CD into
  virtualbox (without guest additions).

  -----
  Workarounds:

  Hint 1: Make sure the heatsink/vent system is not dusty. (Do clean it with pressured air/vacuum at own risk.)
  Hint 2: Using your laptop standing on its side or front edge (even if opened), i.e. when it is connected to external monitor etc., can change the heat flow and make parts like the gpu overheat which won't get so hot in normal orientation.

  In some cases using some additional kernel modules (or other cpu
  throttling packages) seem to work around the problem.

  Comment #308 points out how with ati graphics cards supported by the
  fglrx driver you can disable its acpi functions as a workaround. Bug
  #488152 and Bug #570589 deal with the many ATI cards that overheat.

  An example /etc/X11/xorg.conf to keep the gpu cooler in general is the
  following:

          Section "Screen"
     Identifier	"Configured Screen Device"
         Device	"Configured Video Device"
          EndSection

          Section "Device"
     Identifier	"Configured Video Device"
     Option "ClockGating" "true"
     Option "DynamicPM" "true"
     Option "DynamicClocks" "on"
          EndSection

  Comment #327 points to the "server" edition which does not use acpi
  and does not suffer from this bug.

  Comment #328 hinted that adding the "acpi.power_nocheck=1" AND
  "acpi_osi=linux" boot options can help, if your (updated) BIOS
  supports it. (Making nocheck=1 the default was proposed
  https://patchwork.kernel.org/patch/23156/ but it is not upstream as of
  april 2010.)

  ---

  If your fan does not start at all, but in emergency overheating
  conditions, you may be seeing an (source code wise) unrelated
  issue/bug, that can be worked around however by the same
  acpi.power_nocheck=1 boot option as well, though. (Or by staying 10
  seconds or so in the grub boot menu until the fan starts controlled by
  the bios can help.)

To manage notifications about this bug go to:
https://bugs.launchpad.net/acpi/+bug/370173/+subscriptions