[Bug 41762] New: radeon default power_profile "default" makes laptop overheat (Mobility Radeon HD 3650)
https://bugs.freedesktop.org/show_bug.cgi?id=41762
Summary: radeon default power_profile "default" makes laptop overheat (Mobility Radeon HD 3650) Product: DRI Version: unspecified Platform: x86-64 (AMD64) OS/Version: Linux (All) Status: NEW Severity: major Priority: medium Component: DRM/Radeon AssignedTo: dri-devel@lists.freedesktop.org ReportedBy: pasik@iki.fi
I'm using a laptop with Mobility Radeon HD 3650, and I've had overheating problems for a quite some time, with multiple kernel versions.
The issue happens at least with Fedora 14, Fedora 15 and latest Fedora 16 (beta).
As a default radeon power_profile "default" is being used, and it causes laptop overheating and emergency thermal shutdowns.. on a completely idle system! It seems the power management doesn't work with the opensource radeon driver on this laptop (HP EliteBook 8530p). See below for analysis and more info.
When I manually switch the radeon power_profile to "low" (see below) the temperature of the laptop decreases over 10 degrees celsius, making the laptop much cooler and also the fans will immediately slow down because the temperature goes down. Any idea why the "default" power_profile doesn't automatically run the card at lower engine/memory clock speeds?
Here's more information from Fedota 16 beta (x86_64) liveCD.. notice how the temperature goes down from 84C to 72C when switching to power_profile "low":
# cat /etc/redhat-release Fedora release 16 (Verne)
# uname -a Linux localhost.localdomain 3.1.0-0.rc6.git0.3.fc16.x86_64 #1 SMP Fri Sep 16 12:26:22 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux
# lspci|grep -i vga 01:00.0 VGA compatible controller: ATI Technologies Inc Mobility Radeon HD 3650
$ cat /sys/kernel/debug/dri/0/radeon_pm_info default engine clock: 600000 kHz current engine clock: 594000 kHz default memory clock: 700000 kHz current memory clock: 693000 kHz voltage: 1100 mV PCIE lanes: 16
# cat /sys/devices/virtual/thermal/thermal_zone3/device/path _TZ_.CPUZ
$ cat /sys/devices/virtual/thermal/thermal_zone3/temp 84000
# cat /sys/class/drm/card0/device/power_profile default
# echo low > /sys/class/drm/card0/device/power_profile
(wait for 5 mins.. and notice how fans slow down)
$ cat /sys/devices/virtual/thermal/thermal_zone3/temp 72000
# cat /sys/kernel/debug/dri/0/radeon_pm_info default engine clock: 600000 kHz current engine clock: 334120 kHz default memory clock: 700000 kHz current memory clock: 405000 kHz voltage: 900 mV PCIE lanes: 16
dmesg for the f16beta 3.1.0-rc6 kernel available here: http://pasik.reaktio.net/fedora/f16/radeon/f16-beta-laptop-radeon-dmesg.txt
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #1 from Pasi Kärkkäinen pasik@iki.fi 2011-10-20 14:19:14 PDT --- Created attachment 52608 --> https://bugs.freedesktop.org/attachment.cgi?id=52608 dmesg for Linux 3.1.0-rc9
dmesg for Linux 3.1.0-0.rc9.git0.0.fc16.x86_64 (Fedora 16 TC1 kernel).
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #2 from Pasi Kärkkäinen pasik@iki.fi 2011-12-26 13:14:15 UTC --- Created attachment 54834 --> https://bugs.freedesktop.org/attachment.cgi?id=54834 dmesg from F16 livecd, Linux kernel 3.1.0
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #3 from Pasi Kärkkäinen pasik@iki.fi 2011-12-26 13:14:50 UTC --- Created attachment 54835 --> https://bugs.freedesktop.org/attachment.cgi?id=54835 dmesg from F16 livecd, Linux kernel 3.1.0, with drm.debug=0x2
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #4 from Pasi Kärkkäinen pasik@iki.fi 2011-12-26 13:18:18 PST ---
I verified the bug is still in Fedora 16 final, which uses Linux kernel 3.1.0-7.fc16.x86_64 .
Here's some output about radeon power states with "drm.debug=0x2" boot cmdline parameter:
[ 3.543677] [drm:radeon_pm_print_states], 7 Power State(s) [ 3.543679] [drm:radeon_pm_print_states], State 0: Default [ 3.543681] [drm:radeon_pm_print_states], Default [ 3.543683] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543685] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543687] [drm:radeon_pm_print_states], 0 e: 600000 m: 700000 v: 1100 No display only [ 3.543689] [drm:radeon_pm_print_states], 1 e: 600000 m: 700000 v: 1100 [ 3.543691] [drm:radeon_pm_print_states], 2 e: 600000 m: 700000 v: 1100 [ 3.543694] [drm:radeon_pm_print_states], State 1: Performance [ 3.543695] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543697] [drm:radeon_pm_print_states], Single display only [ 3.543699] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543701] [drm:radeon_pm_print_states], 0 e: 110000 m: 405000 v: 900 No display only [ 3.543703] [drm:radeon_pm_print_states], 1 e: 337000 m: 405000 v: 900 [ 3.543705] [drm:radeon_pm_print_states], 2 e: 600000 m: 700000 v: 1100 [ 3.543708] [drm:radeon_pm_print_states], State 2: Battery [ 3.543709] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543711] [drm:radeon_pm_print_states], Single display only [ 3.543713] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543714] [drm:radeon_pm_print_states], 0 e: 110000 m: 405000 v: 900 No display only [ 3.543717] [drm:radeon_pm_print_states], 1 e: 337000 m: 405000 v: 900 [ 3.543719] [drm:radeon_pm_print_states], 2 e: 337000 m: 405000 v: 900 [ 3.543721] [drm:radeon_pm_print_states], State 3: Default [ 3.543723] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543725] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543727] [drm:radeon_pm_print_states], 0 e: 300000 m: 405000 v: 1100 No display only [ 3.543729] [drm:radeon_pm_print_states], 1 e: 450000 m: 700000 v: 1100 [ 3.543731] [drm:radeon_pm_print_states], 2 e: 450000 m: 700000 v: 1100 [ 3.543733] [drm:radeon_pm_print_states], State 4: Performance [ 3.543735] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543737] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543739] [drm:radeon_pm_print_states], 0 e: 337000 m: 700000 v: 1100 No display only [ 3.543741] [drm:radeon_pm_print_states], 1 e: 337000 m: 700000 v: 1100 [ 3.543743] [drm:radeon_pm_print_states], 2 e: 600000 m: 700000 v: 1100 [ 3.543746] [drm:radeon_pm_print_states], State 5: Battery [ 3.543747] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543749] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543751] [drm:radeon_pm_print_states], 0 e: 337000 m: 405000 v: 900 No display only [ 3.543753] [drm:radeon_pm_print_states], 1 e: 337000 m: 405000 v: 900 [ 3.543756] [drm:radeon_pm_print_states], 2 e: 337000 m: 405000 v: 900 [ 3.543758] [drm:radeon_pm_print_states], State 6: Default [ 3.543760] [drm:radeon_pm_print_states], 16 PCIE Lanes [ 3.543761] [drm:radeon_pm_print_states], 3 Clock Mode(s) [ 3.543763] [drm:radeon_pm_print_states], 0 e: 337000 m: 405000 v: 900 No display only [ 3.543765] [drm:radeon_pm_print_states], 1 e: 337000 m: 405000 v: 900 [ 3.543768] [drm:radeon_pm_print_states], 2 e: 337000 m: 405000 v: 900 [ 3.543797] [drm] radeon: power management initialized
Do those power states look OK ? Any comments?
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #5 from Pasi Kärkkäinen pasik@iki.fi 2011-12-26 13:24:05 PST ---
Sorry, the paste is badly line wrapped, but the dmesg attachment has the whole proper kernel log with drm.debug=0x2
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #6 from Pasi Kärkkäinen pasik@iki.fi 2012-01-29 07:39:36 PST ---
As I understand "Mobility Radeon HD3650" is based on R600 chipset, so the feature matrix in http://wiki.x.org/wiki/RadeonFeature says all the power saving/management features are implemented for this card.
Is this an error in the matrix, or a bug in the radeon driver?
https://bugs.freedesktop.org/show_bug.cgi?id=41762
Tadej Janež tadej.janez@tadej.hicsalta.si changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |tadej.janez@tadej.hicsalta. | |si
--- Comment #7 from Tadej Janež tadej.janez@tadej.hicsalta.si 2012-05-13 07:09:40 PDT --- Hi, Pasi!
I have the same laptop (HP EliteBook 8530p) and I'm also experiencing the same overheating issues.
Thanks to your tip on how to switch to "low" radeon power_profile, I've managed to cool my laptop for ~15 degrees. The temperatures were around 67 - 69 degrees Celsius before and now they are around 54 - 56 degrees Celsius.
Does any DRM/Radeon developer have any thoughts on this issue? Should we consider changing the default behaviour to "low" power_profile? Or should we try to raise the awareness of users about different power profiles?
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #8 from Pasi Kärkkäinen pasik@iki.fi 2012-05-13 08:44:47 PDT ---
Yeah, many people have the overheating problem with radeon driver and laptops..
In my opinion the default radeon power profile is *broken*, and it should be switched to "low", or then the default profile should be fixed to work properly..
https://bugs.freedesktop.org/show_bug.cgi?id=41762
aleritty aleritty@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Hardware|x86-64 (AMD64) |All
--- Comment #9 from aleritty aleritty@gmail.com --- I tested this on a lot of kernels and distribution both iX86 or x64
My card is a Mobility Radeon HD 4500, RV710
Actually I think that the profile have to be set on low by default, to prevent hardware damages.
But this is not enough, because with proprietary drivers the card run around 55°C on my system, but with the radeon and low profile is around 70°C, and with the default the temperature (idle system) go around 82°C for most of the time.
This can easily harm a system.
Actually this bug assume a big importance because most of the affected chip aren't supported anymore from proprietary drivers, so a user can unconsciously run into troubles and hardware damage after a simply upgrade.
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #10 from renich@woralelandia.com renich@woralelandia.com --- Created attachment 104077 --> https://bugs.freedesktop.org/attachment.cgi?id=104077&action=edit journalctl -b
output of journalctl -b
https://bugs.freedesktop.org/show_bug.cgi?id=41762
--- Comment #11 from renich@woralelandia.com renich@woralelandia.com --- I am suffering of the same thing on Fedora 20. Even during install.
https://bugs.freedesktop.org/show_bug.cgi?id=41762
Dmitri Kolobov disjunctor@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |disjunctor@gmail.com
--- Comment #12 from Dmitri Kolobov disjunctor@gmail.com --- I have exactly the same problem as described by previous posts (Radeon 3650, Asus F8Va laptop. OpenSuSE 12-13.2, Fedora). With low profile the temperature is about 70°C, and with the default one the temperature (idle system) goes to 80-85°C. DPM mode shows the same behavior.
https://bugs.freedesktop.org/show_bug.cgi?id=41762
mondane.woodworker@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mondane.woodworker@gmail.co | |m
--- Comment #13 from mondane.woodworker@gmail.com --- This is still happening on Ubuntu 15.10 running the default kernel. My problems when away after setting the profile to 'low'.
https://bugs.freedesktop.org/show_bug.cgi?id=41762
Martin Peres martin.peres@free.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |MOVED Status|NEW |RESOLVED
--- Comment #14 from Martin Peres martin.peres@free.fr --- -- 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/222.
dri-devel@lists.freedesktop.org
-
bugzilla-daemon＠freedesktop.org