Bug 90263 - GPU card fans run full speed at all times
Summary: GPU card fans run full speed at all times
Status: RESOLVED MOVED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-05-01 01:16 UTC by Todd
Modified: 2019-11-19 09:03 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
DMESG output (114.91 KB, text/plain)
2015-05-01 13:56 UTC, Todd
no flags Details
XORG log (45.60 KB, text/plain)
2015-05-01 17:53 UTC, Todd
no flags Details

Description Todd 2015-05-01 01:16:22 UTC
Arch w/ Linux 4. AMD "6770" "BARTS" GPU.

GPU fans run at full speed at all times.

Lots of promises on the forums that the fix would "just work" in Linux 4. It did not.

Not an issue when dual-booted to Windows 7. Fans run properly.
Comment 1 Michel Dänzer 2015-05-01 01:33:40 UTC
Please attach the dmesg output.
Comment 2 Todd 2015-05-01 13:56:06 UTC
Created attachment 115499 [details]
DMESG output

DMESG output as requested.
Comment 3 Alex Deucher 2015-05-01 14:20:43 UTC
Please attach your xorg log as well.  Are you using multiple monitors?
Comment 4 Todd 2015-05-01 17:53:57 UTC
Created attachment 115503 [details]
XORG log

XORG log as requested.

Yes; using 2 displays, each connected via DVI.
Comment 5 Alex Deucher 2015-05-01 18:27:51 UTC
Does it work properly with one display attached?  If so, it might be a duplicate of:
https://bugzilla.kernel.org/show_bug.cgi?id=60523
Comment 6 Todd 2015-05-01 18:59:04 UTC
Problem persists regardless of connection of only display A, only display B or both A & B. Tested by unplugging each DVI cable and starting computer from power-off state.
Comment 7 Todd 2015-05-18 02:25:09 UTC
What additional information is required to diagnose and solve this problem?
Comment 8 Alex Deucher 2015-05-18 13:09:37 UTC
Apparently you have a newer barts card that has limited default fan profile.  Support for SMC fan control would need to be written for pre-SI cards and reviewed for release.
Comment 9 Todd 2015-05-18 15:55:45 UTC
"Newer" is an odd description. The card is from 2011. For reference, it is a Gigabyte GV-R687OC-1GD.

http://www.gigabyte.com/products/product-page.aspx?pid=3663

If you need additional information, just ask.
Comment 10 Alex Deucher 2015-05-18 17:59:18 UTC
(In reply to Todd from comment #9)
> "Newer" is an odd description. The card is from 2011.

Newer relative to the original Barts cards.  IIRC, smc fan control was introduced sometime after the original barts launch.  Most cards just used the fan profile set up in the vbios.
Comment 11 Todd 2015-05-20 00:42:34 UTC
I was expecting this problem to be solved through the miracle of "dynamic power management support" brought down from the mountain of Linux 4.

I had assumed that the GPU was running at full power all the time under the Linux Radeon driver, hence the loud fans. When booted to Windows 7, the card is quiet.

Is fan speed control a different issue? I'm losing track of what this issue is a symptom of and where to correctly report the bug.
Comment 12 Alex Deucher 2015-05-20 13:50:29 UTC
(In reply to Todd from comment #11)
> I was expecting this problem to be solved through the miracle of "dynamic
> power management support" brought down from the mountain of Linux 4.
> 
> I had assumed that the GPU was running at full power all the time under the
> Linux Radeon driver, hence the loud fans. When booted to Windows 7, the card
> is quiet.
> 
> Is fan speed control a different issue? I'm losing track of what this issue
> is a symptom of and where to correctly report the bug.

Power management support has been available and enabled by default for your card for a while now (since 3.12 or so IIRC).  The driver properly controls the clocks and voltages on your card.  The fan control is somewhat independent.  For most cards of this generation, the vbios sets up the fan profile to scale the fan based on GPU temperature and the driver just uses that.  You happen to have a card with a limited vbios fan profile that requires the driver to set up smc fan control.
Comment 14 Alex Deucher 2015-05-22 13:33:45 UTC
(In reply to Todd from comment #13)
> The following may or may not be relevant to this issue.
> 
> "Fixes for SI fan handling"
> 
> http://cgit.freedesktop.org/%7Eagd5f/linux/commit/?h=drm-next-3.20-
> wip&id=e06b000d6619ac9b60b0724c82dc58141a86a3dc
> 
> 
> "Enable smc fan control on SI"
> 
> http://cgit.freedesktop.org/%7Eagd5f/linux/commit/?h=drm-next-3.20-
> wip&id=bdf40b0fa6588ea64ccb069de63d16f9348aeef6

Yes, the same thing need to be done for older asics.
Comment 15 Todd 2015-06-26 11:43:24 UTC
Continues to be an issue in 4.06. Anyone working on this? Can I help test or provide more information?
Comment 16 Todd 2015-08-18 18:28:16 UTC
Any hope for a fix of this bug in Linux 4.2?
Comment 17 Todd 2015-09-28 12:11:39 UTC
Problem remains in 4.2.1.

Is any attention being given to this bug?
Comment 18 Chernovsky Oleg 2015-09-28 22:24:48 UTC
Hmm let me lay it out for you:

I was like you a year ago when I had my fan speed too high. I mmiotraced proprietary fglrx driver and proposed a patch for this. Alex said that time that reverse-engineered patches are not welcome upstream (various licensing issues, I suppose) and suggested to pull a set of initial patches for handling fan control through AMD IP bureaucracy. I agreed.

This took around month or even more, IIRC. These shinies are not sort of things so easy to get from large corporation. In the end Alex handed me over these patches and I was able to fix typos and implement fan control for SI and CI cards.

I can do it for r600, but again, I will need patches of same kind and moreover, working BARTS GPU in St. Petersburg where I live to test and debug it.
Comment 19 Todd 2015-09-29 01:07:40 UTC
St. Petersburg USA or Russia?
Comment 20 Chernovsky Oleg 2015-09-29 22:25:13 UTC
(In reply to Todd from comment #19)
> St. Petersburg USA or Russia?

Russia.

P.S. If you're thinking about parcel, forget it, post service here is horrible. I asked local LUG for PITCAIRN GPU last time and can do it once more for BARTS.
Comment 21 Todd 2015-09-30 01:40:22 UTC
Russia is too far; agree.
Comment 22 Martin Peres 2019-11-19 09:03:59 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/606.


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.