How to keep the GPU (and CPU) cooler

nouveau errors co-occur with high temperatures (seen from journalctl), therefore, as an experiment, I want to try to keep the GPU cool (and see whether the errors will stop occuring).

My graphical errors lead to almost complete freezing of X (say, during Skype video calls).

My system is ALT p7 with linux 3.10.21-std-def-alt1 on an HP laptop.

   # lspci | fgrep -i vga
   02:00.0 VGA compatible controller: NVIDIA Corporation C77 [GeForce 8200M G] (rev a2)
   #

Keep the Nvidia GPU cooler

By reading documentation on the web and by experiments, I have found out that setting the threshold temperatures (in sysfs) to lower values does cause the GPU to become cooler!

Current temperatures (and the threshold values except for "fanboost") are shown by sensors command.

I use sysfsutils package for setting the values in sysfs on boot.^[1] (service sysfs restart will reset the parameters to the new values afterwards, when needed, but not service sysfs start!)

   #
   # /etc/sysfs.conf - Configuration file for setting sysfs attributes.
   #

   # (temperature is in millidegrees)

   # "fanboost"; default: 90 degrees C
   class/drm/card0/device/hwmon/hwmon0/temp1_auto_point1_temp = 85000
   # "downclock"; default: 95
   class/drm/card0/device/hwmon/hwmon0/temp1_max = 83000
   # "critical"; default: 105
   class/drm/card0/device/hwmon/hwmon0/temp1_crit = 95000
   # "emergency"; default: 135
   # class/drm/card0/device/hwmon/hwmon0/temp1_emergency = 135000

"downclock" seems to bring most effect. Because I want the fan to be silent sometimes, and "fanboost" turned out not to be a really effective way to cool down the GPU, I raised the "fanboost" threshold above the "downclock" threshold (unlike in the default settings).

resume after suspend

It has been observed that sysfs utils are not run when coming from a suspend. Therefore additional care must be taken to set the parameters on resume!

Monitoring output examples

sensors output example

   # sensors
   nouveau-pci-0200
   Adapter: PCI adapter
   temp1:        +81.0°C  (high = +83.0°C, hyst =  +3.0°C)
   (crit = +95.0°C, hyst =  +5.0°C)
   (emerg = +135.0°C, hyst =  +5.0°C)

   k10temp-pci-00c3
   Adapter: PCI adapter
   temp1:        +53.6°C  (high = +70.0°C)

example of temperature threshold hits from journalctl

   # journalctl | fgrep nouveau | fgrep -10 critical
   дек 15 23:28:07 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:07 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:08 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:08 host-17.localdomain kernel[1506]: [ 6914.386641] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:08 host-17.localdomain kernel[1506]: [ 6914.387656] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:08 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:10 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:10 host-17.localdomain kernel[1506]: [ 6916.421280] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:10 host-17.localdomain kernel[1506]: [ 6916.422291] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:10 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:28:40 host-17.localdomain kernel: nouveau  [  PTHERM][0000:02:00.0] temperature (106 C) hit the 'critical' threshold
   дек 15 23:28:40 host-17.localdomain kernel[1506]: [ 6945.541057] nouveau  [  PTHERM][0000:02:00.0] temperature (106 C) hit the 'critical' threshold
   дек 15 23:29:16 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:16 host-17.localdomain kernel[1506]: [ 6981.487662] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:16 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:16 host-17.localdomain kernel[1506]: [ 6981.488677] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:17 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:17 host-17.localdomain kernel[1506]: [ 6982.722574] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:17 host-17.localdomain kernel[1506]: [ 6982.723589] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:17 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:17 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 15 23:29:17 host-17.localdomain kernel[1506]: [ 6983.152643] nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   --
   дек 18 18:32:25 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:25 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:25 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:25 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:25 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:25 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:27 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:27 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:27 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:27 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:32:53 host-17.localdomain kernel: nouveau  [  PTHERM][0000:02:00.0] temperature (106 C) hit the 'critical' threshold
   дек 18 18:45:10 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001
   дек 18 18:45:10 host-17.localdomain kernel: nouveau E[  PTHERM][0000:02:00.0] unhandled intr 0x00000001

And keep the CPU cooler

The heat from CPU goes to GPU, too, therefore cooling down the GPU without cooling down some really high CPU temperatures is not effective.

I do not know yet of a similar downclock-policy setting mechanism for CPU which would be based on temperature thresholds, therefore I simply use the brute force of a governor.

I'm unhappy that it will be active all the time though, of course. Usually I'm happy with the "ondemend" governor, and only when the temperature rises, I want to scale down...

TODO: find out how to do this! (Have a look at this, via [1]: "Additional control for modern Intel CPUs is available with the Linux Thermal Daemon (available as thermald in the AUR), which proactively controls thermal using P-states, T-states, and the Intel power clamp driver." And what about AMD?)

   #
   # /etc/sysfs.conf - Configuration file for setting sysfs attributes.
   #

   devices/system/cpu/cpu0/cpufreq/scaling_governor = powersave
   devices/system/cpu/cpu1/cpufreq/scaling_governor = powersave

FYI: more temperature attributes in sysfs

Perhaps we can make some use of them.

      # find /sys/ -name '*temp*'
      /sys/bus/pci/drivers/k10temp
      /sys/devices/pci0000:00/0000:00:18.3/temp1_input
      /sys/devices/pci0000:00/0000:00:18.3/temp1_max
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_crit
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_crit_hyst
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_auto_point1_pwm
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_max_hyst
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_auto_point1_temp
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_emergency_hyst
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_auto_point1_temp_hyst
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_emergency
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_input
      /sys/devices/pci0000:00/0000:00:0b.0/0000:02:00.0/hwmon/hwmon0/temp1_max
      /sys/devices/virtual/thermal/thermal_zone0/temp
      /sys/devices/virtual/thermal/thermal_zone0/emul_temp
      /sys/devices/virtual/thermal/thermal_zone0/trip_point_0_temp
      /sys/devices/virtual/thermal/thermal_zone0/trip_point_1_temp
      /sys/devices/virtual/thermal/thermal_zone0/trip_point_2_temp
      /sys/devices/virtual/thermal/thermal_zone1/temp
      /sys/devices/virtual/thermal/thermal_zone1/emul_temp
      /sys/devices/virtual/thermal/thermal_zone1/trip_point_0_temp
      /sys/devices/virtual/thermal/thermal_zone1/trip_point_1_temp
      /sys/devices/platform/hp-wmi/hddtemp
      /sys/module/hwmon/holders/k10temp
      /sys/module/k10temp
      /sys/module/k10temp/drivers/pci:k10temp

Notes and refs

↑ After the advice from ru:cpufreq.

[1] After the advice from ru:cpufreq.

[1]