Comparing two Debian kernels:
--------------------------------------------------------------------------------
$ uname -rv
3.2.0-4-powerpc64 #1 SMP Debian 3.2.41-2
$ lsmod | grep windfarm
windfarm_cpufreq_clamp 4265 1
windfarm_max6690_sensor 6030 1
windfarm_lm75_sensor 6521 1
windfarm_smu_sensors 7587 1
windfarm_smu_controls 7770 7
windfarm_pm112 14636 0
windfarm_smu_sat 8271 5 windfarm_pm112
windfarm_pid 4095 1 windfarm_pm112
windfarm_core 12117 7
windfarm_smu_sat,windfarm_pm112,windfarm_smu_controls,windfarm_smu_sensors,windfarm_lm75_sensor,windfarm_max6690_sensor,windfarm_cpufreq_clamp
$ dmesg | grep windfarm
[ 5.184978] windfarm: initializing for dual-core desktop G5
[ 10.488713] windfarm: CPUs control loops started.
[ 13.914914] windfarm: Backside control loop started.
[ 13.965432] windfarm: Slots control loop started.
[ 14.070479] windfarm: Drive bay control loop started.
$ ls /sys/devices/platform/windfarm.0/
backside-fan cpufreq-clamp cpu-power-1 cpu-temp-1 driver
slots-fan
backside-temp cpu-front-fan-0 cpu-rear-fan-0 cpu-voltage-0 hd-temp
slots-power
cpu-current-0 cpu-front-fan-1 cpu-rear-fan-1 cpu-voltage-1 modalias
subsystem
cpu-current-1 cpu-power-0 cpu-temp-0 drive-bay-fan power
uevent
$ cat /sys/devices/platform/windfarm.0/cpu-front-fan-*
524
524
$ cat /sys/devices/platform/windfarm.0/cpu-rear-fan-*
562
562
--------------------------------------------------------------------------------
$ uname -rv
3.10-1-powerpc64 #1 SMP Debian 3.10.3-1 (2013-07-27)
$ lsmod | grep windfarm
windfarm_cpufreq_clamp 4001 1
windfarm_max6690_sensor 4843 0
windfarm_lm75_sensor 5305 0
windfarm_smu_sensors 7691 1
windfarm_smu_controls 7954 7
windfarm_pm112 14588 0
windfarm_smu_sat 8348 1 windfarm_pm112
windfarm_pid 3975 1 windfarm_pm112
windfarm_core 12029 7
windfarm_cpufreq_clamp,windfarm_pm112,windfarm_smu_sensors,windfarm_smu_sat,windfarm_lm75_sensor,windfarm_max6690_sensor,windfarm_smu_controls
$ dmesg | grep windfarm
[ 5.130698] windfarm: initializing for dual-core desktop G5
$ ls /sys/devices/platform/windfarm.0/
backside-fan cpu-front-fan-0 cpu-rear-fan-0 drive-bay-fan modalias
slots-fan subsystem
cpufreq-clamp cpu-front-fan-1 cpu-rear-fan-1 driver power
slots-power uevent
$ cat /sys/devices/platform/windfarm.0/cpu-front-fan-*
0 RPM
0 RPM
$ cat /sys/devices/platform/windfarm.0/cpu-rear-fan-*
0 RPM
0 RPM
--------------------------------------------------------------------------------
You can help a lot, if you can build and upload kernel packages for vanilla
3.5.x, 3.6.x, 3.7.x, and 3.8.x. That's how we can find a working one
("good") kernel very fast. Then we can bisect kernel git source repo for
first "bad" commit.
On Sun, Jul 28, 2013 at 4:45 PM, Ben Hutchings <[email protected]> wrote:
> On Sun, 2013-07-28 at 12:25 -0400, Douglas Mencken wrote:
> > >Right, we have this big stockpile of hardware we could test on
> >
> > I can provide SSH access to the G5 box.
> >
> > P.S. Sadly, I have no microphone to show you how noisy it is on 6000
> > rpm of 8 fans :)
>
> Is anyone able to investigate this, or suggest how to proceeed? It's
> apparently a regression between 3.8 and 3.9, but the windfarm driver
> itself appears unchanged between those versions.
>
> Ben.
>
> --
> Ben Hutchings
> All extremists should be taken out and shot.
>
