[2.6.0-test1] ACPI slowdown

[2.6.0-test1] ACPI slowdown

[Florian Weimer]

If I enable ACPI on my box (Athlon XP at 1.6 GHz, Epox EP-8KHa+
mainboard), it becomes very slow (so slow that it's unusable).

Is this a known issue?  Maybe the thermal limits are misconfigured,
and the CPU clock is throttled unnecessarily (if something like this
is supported at all).

Re: [2.6.0-test1] ACPI slowdown

[Dominik Brugger]

On Wed, 23 Jul 2003 09:57:08 +0200
Florian Weimer <fw@deneb.enyo.de> wrote:

> If I enable ACPI on my box (Athlon XP at 1.6 GHz, Epox EP-8KHa+
> mainboard), it becomes very slow (so slow that it's unusable).
> 
> Is this a known issue?  Maybe the thermal limits are misconfigured,
> and the CPU clock is throttled unnecessarily (if something like this
> is supported at all).

I use the same board with ACPI enabled, no slow down.
The only problem I experience is that USB doesnt wakeup after resume from S3, reloading all of the related modules doesnt help, the port seems to be unpowered (usb optical mouse blinks for a few miliseconds after reloading uhci_hcd, maybe there is some little power left).

-Dominik Brugger

Re: [2.6.0-test1] ACPI slowdown

[bill davidsen]

In article <878yqpptez.fsf@deneb.enyo.de>,
Florian Weimer  <fw@deneb.enyo.de> wrote:
| If I enable ACPI on my box (Athlon XP at 1.6 GHz, Epox EP-8KHa+
| mainboard), it becomes very slow (so slow that it's unusable).
| 
| Is this a known issue?  Maybe the thermal limits are misconfigured,
| and the CPU clock is throttled unnecessarily (if something like this
| is supported at all).

There have been reports before, check the archives. I seem to remember
that the solution involved changing some unobvious kernel feature, but
others have had similar problems.
-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

Re: [2.6.0-test1] ACPI slowdown

[Bryan D. Stine]

[-- Attachment #1: Type: text/plain, Size: 860 bytes --]

I had that problem with my old Athlon TBird. Changing config to make
thermal a module and not loading it solved my problem. I don't know how
to change the thermal limits from within the system using ACPI.


On Wed, 2003-07-23 at 18:56, bill davidsen wrote:
> In article <878yqpptez.fsf@deneb.enyo.de>,
> Florian Weimer  <fw@deneb.enyo.de> wrote:
> | If I enable ACPI on my box (Athlon XP at 1.6 GHz, Epox EP-8KHa+
> | mainboard), it becomes very slow (so slow that it's unusable).
> | 
> | Is this a known issue?  Maybe the thermal limits are misconfigured,
> | and the CPU clock is throttled unnecessarily (if something like this
> | is supported at all).
> 
> There have been reports before, check the archives. I seem to remember
> that the solution involved changing some unobvious kernel feature, but
> others have had similar problems.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [2.6.0-test1] ACPI slowdown

[Florian Weimer]

Dominik Brugger <dominik83@gmx.net> writes:

> On Wed, 23 Jul 2003 09:57:08 +0200
> Florian Weimer <fw@deneb.enyo.de> wrote:
>
>> If I enable ACPI on my box (Athlon XP at 1.6 GHz, Epox EP-8KHa+
>> mainboard), it becomes very slow (so slow that it's unusable).
>> 
>> Is this a known issue?  Maybe the thermal limits are misconfigured,
>> and the CPU clock is throttled unnecessarily (if something like this
>> is supported at all).
>
> I use the same board with ACPI enabled, no slow down.

Could you post the output of "tail /proc/acpi/thermal_zone/THRM/*"?
Thanks.

Re: [2.6.0-test1] ACPI slowdown

[Dominik Brugger]

On Thu, 24 Jul 2003 08:33:42 +0200
Florian Weimer <fw@deneb.enyo.de> wrote:

> Could you post the output of "tail /proc/acpi/thermal_zone/THRM/*"?
> Thanks.

$ tail /proc/acpi/thermal_zone/THRM/*
==> /proc/acpi/thermal_zone/THRM/cooling_mode <==
cooling mode:            active

==> /proc/acpi/thermal_zone/THRM/polling_frequency <==
<polling disabled>

==> /proc/acpi/thermal_zone/THRM/state <==
state:                   ok

==> /proc/acpi/thermal_zone/THRM/temperature <==
temperature:             37 C

==> /proc/acpi/thermal_zone/THRM/trip_points <==
critical (S5):           70 C
passive:                 70 C: tc1=4 tc2=3 tsp=60 devices=0xdff6cde8
active[0]:               70 C: devices=0xdff64d68

-- 
-Dominik Brugger

Re: [2.6.0-test1] ACPI slowdown

[Florian Weimer]

Dominik Brugger <ml.dominik83@gmx.net> writes:

> ==> /proc/acpi/thermal_zone/THRM/temperature <==
> temperature:             37 C

This is at 57 C for me. 8-/

Probably that's why cooling measures kick in.

Re: [2.6.0-test1] ACPI slowdown

[Dominik Brugger]

Yes because I turned on the box only a few minutes ago, now it is at 49 C.

On Thu, 24 Jul 2003 11:35:11 +0200
Florian Weimer <fw@deneb.enyo.de> wrote:

> This is at 57 C for me. 8-/

-Dominik Brugger

Re: [2.6.0-test1] ACPI slowdown

[Florian Weimer]

"Bryan D. Stine" <admin@kentonet.net> writes:

> I had that problem with my old Athlon TBird. Changing config to make
> thermal a module and not loading it solved my problem. I don't know how
> to change the thermal limits from within the system using ACPI.

Maybe it's related to the BIOS configuration.  You can configure
limits there, and they seem to match those that are displayed under
/proc.

However, I still wonder if the 57 °C figure is something I should
worry about (although the machine has been running stable for years).

Re: [2.6.0-test1] ACPI slowdown

[Pavel Machek]

Hi!

> I had that problem with my old Athlon TBird. Changing config to make
> thermal a module and not loading it solved my problem. I don't know how
> to change the thermal limits from within the system using ACPI.

echo "83:83:55:70:0" > /proc/acpi/thermal_zone/THRM/trip_points

								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]

Re: [2.6.0-test1] ACPI slowdown

[Pavel Machek]

Hi!

> > Could you post the output of "tail /proc/acpi/thermal_zone/THRM/*"?
> > Thanks.
> 
> $ tail /proc/acpi/thermal_zone/THRM/*
> ==> /proc/acpi/thermal_zone/THRM/cooling_mode <==
> cooling mode:            active
> 
> ==> /proc/acpi/thermal_zone/THRM/polling_frequency <==
> <polling disabled>
> 
> ==> /proc/acpi/thermal_zone/THRM/state <==
> state:                   ok
> 
> ==> /proc/acpi/thermal_zone/THRM/temperature <==
> temperature:             37 C
> 
> ==> /proc/acpi/thermal_zone/THRM/trip_points <==
> critical (S5):           70 C
> passive:                 70 C: tc1=4 tc2=3 tsp=60 devices=0xdff6cde8
> active[0]:               70 C: devices=0xdff64d68

This is clearly wrong.... But it is probably is not causing your
slowdown as 37 < 70.

								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]

Re: [2.6.0-test1] ACPI slowdown

[Bryan D. Stine]

[-- Attachment #1: Type: text/plain, Size: 443 bytes --]

Thanks for that. Good to know I can control the termal limits now.

On Thu, 2003-07-24 at 19:41, Pavel Machek wrote:
> Hi!
> 
> > I had that problem with my old Athlon TBird. Changing config to make
> > thermal a module and not loading it solved my problem. I don't know how
> > to change the thermal limits from within the system using ACPI.
> 
> echo "83:83:55:70:0" > /proc/acpi/thermal_zone/THRM/trip_points
> 
> 								Pavel

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]