Bug 67178 - discrete radeon HD 6400M/7400M intense heating even though it's "off"
Summary: discrete radeon HD 6400M/7400M intense heating even though it's "off"
Status: RESOLVED MOVED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: XOrg git
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-07-22 15:04 UTC by Tomas Pospisek
Modified: 2019-11-19 08:36 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
dmesg (77.22 KB, text/plain)
2013-07-22 17:03 UTC, Tomas Pospisek
no flags Details

Description Tomas Pospisek 2013-07-22 15:04:33 UTC
kernel 3.11-rc1 (also tested with 3.10 from Debian, 3.7, 3,2)

# lspci -vnnn | grep VGA
00:02.0 VGA compatible controller [0300]: Intel Corporation 2nd
  Generation Core Processor Family Integrated Graphics Controller
  [8086:0116] (rev 09) (prog-if 00 [VGA controller])
01:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc.
  [AMD/ATI] Seymour [Radeon HD 6400M/7400M Series] [1002:6760] (rev ff)
  (prog-if ff)

This is a HP Pavillon dv7 6130ez

Sometimes, at random, the radeon card goes into full heat production mode. This seems to occur a while (a few minutes?) after wakeup from resume, however I'm not sure about that. Suspending and resuming the laptop a few times (1 <= x <= ~4) seems to calm the card down again.

Power consumption according to powertop goes from ~14W up to ~50W. This happens even though the card should be in theory switched off. Aka:

  # cat /sys/kernel/debug/vgaswitcheroo/switch
  0:IGD:+:Pwr:0000:00:02.0
  1:DIS: :Off:0000:01:00.0

  # echo IGD > /sys/kernel/debug/vgaswitcheroo/switch
  # echo OFF > /sys/kernel/debug/vgaswitcheroo/switch

doesn't seem to have no effect whatsoever.

It *seems* that setting "radeon.dpm=1" kernel boot options and then:

  # echo "battery" > /sys/class/drm/card$no/device/power_dpm_state

puts the card in an "active" but "relatively low" power consumption mode. It is then using about 20W (instead of the "normal" ~14W) so it's continuously faning out slightly warm air. And it "seems that when the card is running in this mode, then it won't get so often, or ever (I'm not very sure) into the berserk heating mode.

The lowest observed power consumption of the laptop is ~7W. It'd be very nice if it'd be possible to shut the card down electrically via software and stay in this 7W mode.
Comment 1 Tomas Pospisek 2013-07-22 17:03:20 UTC
Created attachment 82828 [details]
dmesg

dmesg with wlan stuff edited out
Comment 2 Tomas Pospisek 2013-07-22 20:43:00 UTC
So if I boot with "radeon.dpm=1" then, as allready described, the laptop uses "relatively little" power, and is faning off some heat continuously, even when I set:

  # echo "battery" > /sys/class/drm/card1/device/power_dpm_state

Whether I do the following does *not* seem to matter:

  # echo IGD > /sys/kernel/debug/vgaswitcheroo/switch
  # echo OFF > /sys/kernel/debug/vgaswitcheroo/switch

But if I boot *without* "radeon.dpm=1" then the laptop will stay cool without fans running out of the box using around 14W.

The various settings will show this then:

  # cat /sys/class/drm/card1/device/power_method
  dpm
  # cat /sys/class/drm/card1/device/power_profile
  default
  # cat /sys/class/drm/card1/device/power_dpm_state
  balanced

Weird that power_method would show "dpm" even though I have booted *without* "radeon.dpm=1" ?!

When I then set:

  # echo "battery" > /sys/class/drm/card1/device/power_dpm_state

Power consumption will drop even further to ~11W.

The overal logic of all of this is evidently completely escaping me... 8-o
Comment 3 Martin Peres 2019-11-19 08:36:31 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/358.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.