[ltp] Re: Trouble setting ACPI on a A21e 2628-J1G

Colin Macdonald linux-thinkpad@linux-thinkpad.org
Sat, 26 Mar 2005 13:58:00 -0800 (PST)


On Sat, 26 Mar 2005 linux-thinkpad-request@linux-thinkpad.org wrote:

> Ok, i've installed ubuntu warty on this A21e, recompiled 2.6.11.5 kernel
> and start configuring it.
> Plain 2.6.11.5 kernel, so i'm using ibm-acpi 0.8.

I have a A21e 2655, running Gentoo and 2.6.11.

> 1) there's no way to load cpu clock modulation, if i do:
>
> 	modprobe speedstep-smi
>
> i got error, even if i load speedstep-lib with relaxed-check=1 and
> kernel compiled with CONFIG_X86_SPEEDSTEP_RELAXED_CAP_CHECK=y, as
> thinkwiki.org suggest.

Same here.  I'm reasonable sure that my laptop supports it because
Windows used to switch between two (or maybe 3) speeds and there
are also various options in the bios.  Note the other guy who
commented on it said something about a PII...  but my A21e has a
Coppermine (crippled PIII).  Whats in yours?

One small note, instead of "Rebooting with cpufreq.debug=7" you
can just "echo 7 > /sys/module/cpufreq/parameters/debug" to
achieve the same thing if your kernel has sysfs support.

For reference, here's what I get in dmesg when I try to "modprobe
speedstep-smi" with "options speedstep-lib relaxed_check=1" in my
/etc/modules.conf:

speedstep-lib: x86: 6, model: 8
speedstep-lib: Coppermine: MSR_IA32_EBL_CR_POWERON is 0x46480020,
0x0
speedstep-lib: Coppermine: MSR_IA32_PLATFORM ID is 0x0, 0x4c340000
speedstep-smi: signature:0x47534943, command:0x008200b2,
event:0x00c000b3, perf_level:0x07d00000.
cpufreq-core: trying to register driver speedstep-smi
cpufreq-core: adding CPU 0
speedstep-smi: trying to obtain ownership with command 47534982 at
port b2
speedstep-smi: result is 2
speedstep-smi: fails in aquiring ownership of a SMI interface.
cpufreq-core: initialization failed
cpufreq-core: no CPU initialized for driver speedstep-smi
cpufreq-core: unregistering CPU 0

Perhaps you can send the equivalent msgs from your machine?

> 2) hem... someone have just done the acpi script for ``standard
> things'' (suspend if lid closed and battery, hibernate if battery goes
> too low, ...) and can give to me? ;-)))
> I've installed acpi-support and ibm-acpi packages, but both provide
> only minimal set of scripts...

Here's what I'm using for Fn-F3:

belladonna events # pwd
/etc/acpi/events
belladonna events # cat blank
event=ibm/hotkey HKEY 00000080 00001003
action=/etc/acpi/actions/blank-gentoo.sh
belladonna events # cd /etc/acpi/actions/
belladonna actions # cat blank-gentoo.sh
#!/bin/bash

export XAUTHORITY=/home/nilok/.Xauthority

/usr/bin/xset -display :0 dpms force off
belladonna actions #

For Fn-12:
belladonna events # cat hibernate
#event=(button/power|ibm/hotkey HKEY 00000080 0000100c)
event=ibm/hotkey HKEY 00000080 0000100c
action=/etc/acpi/actions/hibernate.sh
belladonna events # cat /etc/acpi/actions/hibernate.sh
#!/bin/bash

/bin/sync
/usr/sbin/hibernate
belladonna events #

(/usr/sbin/hibernate is from software suspend 2)

> 3) i can do:
>
> 	echo standby > /sys/power/state
>
> and the system stanby, but the lcd remains lit, all things seems work
> because if i close the lid lcd shut off, and if i open the lid the
> system come back from standby.
> It is correct?

Same behavior here.  I tried to get the machine to turn off the
LCD before running going into standby (using the xset -dpms off
stuff like above) but "echo 1 > /proc/acpi/sleep" then switches
the machine to a virtual console and wakes up my LCD :(

If you or anyone else figures out a solution to this, please let
me know.

> 4) hybernate seems work correctly

good, me too with suspend2, never really tried the in-kernel
suspend support.

My biggest complaint about this machine is lack of S3
(suspend-to-ram) support:

dmesg | grep supports
ACPI: (supports S0 S1 S4 S5)

Can anyone offer any clues as to why there isn't a S3 there?  My
DSDT is avaiable here:
http://acpi.sourceforge.net/dsdt/tables/IBM/ThinkPad_A21e_2655/

Cheers,
Colin

--
Colin Macdonald
Graduate Student, Dept. of Mathematics, Simon Fraser University
mail: cbm[at]sfu[dot]ca, phone: 604.268.6689 (office)

Please do not send me Word or PowerPoint attachments
See http://www.gnu.org/philosophy/no-word-attachments.html