Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
[solved] amdgpu - broken power management in new kernels?
View unanswered posts
View posts from last 24 hours

 
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 6920

PostPosted: Sun Oct 14, 2018 9:26 pm    Post subject: [solved] amdgpu - broken power management in new kernels? Reply with quote

Lately I'm seeing numbers like this, on an almost totally idle system:
/sys/kernel/debug/dri/0/amdgpu_pm_info:
Clock Gating Flags Mask: 0x37bcf
        …

GFX Clocks and Power:
        1500 MHz (MCLK)
        1206 MHz (SCLK)
        214 MHz (PSTATE_SCLK)
        625 MHz (PSTATE_MCLK)
        1000 mV (VDDGFX)
        13.65 W (average GPU)

GPU Temperature: 52 C
GPU Load: 0 %

UVD: Disabled

VCE: Disabled

13W/52°C is the sort of numbers I'd expect under load (compare to this previous post and this one). The idle numbers should be much lower. I'm not sure how to read the *CLK lines, but if the top two are current speed then that's not a good thing…

This is on kernel 4.18.11, I'm not yet sure when this first started but it was relatively recent. Anyone else seeing unusally high temps?


Last edited by Ant P. on Mon Oct 15, 2018 9:01 pm; edited 1 time in total
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sun Oct 14, 2018 10:18 pm    Post subject: Reply with quote

Don't have /sys/kernel/debug but seeing idle 99F with CPU at 75F with 4.17.19
Code:
$ sudo lspci|grep VGA
21:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Caicos [Radeon HD 6450/7450/8450 / R5 230 OEM]
This is a low power board running the radeon module.
What kernel config do I need to enable /sys/kernel/debug?
I can try both 4.17 and 4.18

I left 4.18 because the screen was updating strangely making a hash of nano.
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 6920

PostPosted: Sun Oct 14, 2018 10:53 pm    Post subject: Reply with quote

Tony0945 wrote:
What kernel config do I need to enable /sys/kernel/debug?

I think enabling CONFIG_DEBUG_FS should be enough, the AMD stuff within doesn't seem to have a separate "off" switch.

I'm not sure if the radeon driver will show anywhere near as much info, or the same bug for that matter; on my other (circa 2011) R600 card all it has is this:
/sys/kernel/debug/dri/0/radeon_pm_info:
uvd    vclk: 0 dclk: 0
power level 0    sclk: 27827 vddc: 900
Back to top
View user's profile Send private message
xdarma
l33t
l33t


Joined: 08 Dec 2003
Posts: 719
Location: tra veneto e friuli (italy)

PostPosted: Mon Oct 15, 2018 6:37 am    Post subject: Reply with quote

Maybe this is involved?
Some AMD GPUs Affected By A Nasty Power Regression That Snuck Into Linux 4.18 Stable
_________________
proud user of faKeDE-4.7.3 -> back to windowmaker -> moved to LXQt
Back to top
View user's profile Send private message
Goverp
Veteran
Veteran


Joined: 07 Mar 2007
Posts: 1997

PostPosted: Mon Oct 15, 2018 7:40 am    Post subject: Reply with quote

On a different but related subject, has anyone got hibernation working with AMDGPU? My hp laptop will suspend about 95% of the time, but not hibernate.
_________________
Greybeard
Back to top
View user's profile Send private message
tholin
Apprentice
Apprentice


Joined: 04 Oct 2008
Posts: 203

PostPosted: Mon Oct 15, 2018 12:01 pm    Post subject: Reply with quote

xdarma wrote:
Maybe this is involved?
Some AMD GPUs Affected By A Nasty Power Regression That Snuck Into Linux 4.18 Stable
AUTOSEL strikes again.

https://www.phoronix.com/forums/forum/linux-graphics-x-org-drivers/open-source-amd-linux/1052938-the-linux-4-18-power-regression-affecting-some-amd-graphics-cards-should-be-reverted?p=1052977#post1052977
"The patch was not intended for 4.18 and I did not flag it for stable. Sasha's patch auto select tool picked it up for 4.18, but it shouldn't have been applied. I've asked that it be reverted from 4.18."

The auto select tool is a machine learning tool that auto identifies bug fixing patches. Unfortunately the stable maintainers don't always check if the bug exist in older branches and just blindly backports the patch anyway. Since Torvalds is on hiatus gregkh is pulling double duty by maintaining both the mainline and the stable trees so don't expect the stable series to get a lot of attention.

EDIT: the faulty patch was added in v4.18.10 and reverted in v4.18.13
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 6920

PostPosted: Mon Oct 15, 2018 7:30 pm    Post subject: Reply with quote

xdarma wrote:
Maybe this is involved?
Some AMD GPUs Affected By A Nasty Power Regression That Snuck Into Linux 4.18 Stable

Looks spot-on. Thanks, will try updating past 4.18.13 and see if that fixes it.

EDIT: yep, 4.18.14 works fine. Temps are sane once more.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum