current ACPI status in 2.6.7 kernel

current ACPI status in 2.6.7 kernel

[Alexander Sirotkin]

Hello.

I could not find this information in neither mailing list archive nor 
the web site.
What is the current status of ACPI ? I'm mainly interested in one simple 
question -
does suspend work ? If so, how stable it is.

Thanks.

Suggestion - put current status message on the web site.

-- 
Alexander Sirotkin
SW Engineer

Texas Instruments
Broadband Communications Israel (BCIL)
Tel:  +972-9-9706587
________________________________________________________________________
"Those who do not understand Unix are condemned to reinvent it, poorly."
      -- Henry Spencer 




-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Nigel Cunningham]

Hi Alexander.

On Tue, 2004-08-10 at 18:05, Alexander Sirotkin wrote:
> Hello.
> 
> I could not find this information in neither mailing list archive nor 
> the web site.
> What is the current status of ACPI ? I'm mainly interested in one simple 
> question -
> does suspend work ? If so, how stable it is.

Are you meaning suspend to ram, or to disk?

Regards,

Nigel



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Alexander Sirotkin]

[-- Attachment #1: Type: text/plain, Size: 740 bytes --]

Suspend to RAM, mostly...

Nigel Cunningham wrote:

>Hi Alexander.
>
>On Tue, 2004-08-10 at 18:05, Alexander Sirotkin wrote:
>  
>
>>Hello.
>>
>>I could not find this information in neither mailing list archive nor 
>>the web site.
>>What is the current status of ACPI ? I'm mainly interested in one simple 
>>question -
>>does suspend work ? If so, how stable it is.
>>    
>>
>
>Are you meaning suspend to ram, or to disk?
>
>Regards,
>
>Nigel
>
>  
>

-- 
Alexander Sirotkin
SW Engineer

Texas Instruments
Broadband Communications Israel (BCIL)
Tel:  +972-9-9706587
________________________________________________________________________
"Those who do not understand Unix are condemned to reinvent it, poorly."
      -- Henry Spencer 


[-- Attachment #2: Type: text/html, Size: 1237 bytes --]

Re: current ACPI status in 2.6.7 kernel

[Nathan Bryant]

Alexander Sirotkin wrote:
>>>I could not find this information in neither mailing list archive nor 
>>>the web site.
>>>What is the current status of ACPI ? I'm mainly interested in one simple 
>>>question -
>>>does suspend work ? If so, how stable it is.

Don't try it if you use any SCSI disks. IDE should be OK.


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Nathan Bryant]

Nathan Bryant wrote:
> Alexander Sirotkin wrote:
> 
>>>> I could not find this information in neither mailing list archive 
>>>> nor the web site.
>>>> What is the current status of ACPI ? I'm mainly interested in one 
>>>> simple question -
>>>> does suspend work ? If so, how stable it is.
> 
> 
> Don't try it if you use any SCSI disks. IDE should be OK.

Oh, and add many SATA adapters to that list. I think SATA is based on 
the SCSI midlayer.


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Alexander Sirotkin]

Nathan Bryant wrote:

> Nathan Bryant wrote:
>
>> Alexander Sirotkin wrote:
>>
>>>>> I could not find this information in neither mailing list archive 
>>>>> nor the web site.
>>>>> What is the current status of ACPI ? I'm mainly interested in one 
>>>>> simple question -
>>>>> does suspend work ? If so, how stable it is.
>>>>
>>
>>
>> Don't try it if you use any SCSI disks. IDE should be OK.
>
>
> Oh, and add many SATA adapters to that list. I think SATA is based on 
> the SCSI midlayer.

Thanks for the tip, fortunately I don't have any SCSI or SATA disks. 
Unfortunately, I still don't
know how to do this. Is there an utility to put PC into sleep/suspend ? 
I know I can switch
to various S modes using /proc entry, but what about all the peripherals 
and user mode apps ?
I would venture that there has to be some kind of subsystem in user mode 
to take care of
all this. I probably did not phrase my question right, but when I Asked 
if ACPI suspend/sleep
status is stable enough I was actually referring to the whole subsystem, 
not just kernel
support.

Thanks a lot.

-- 
Alexander Sirotkin
SW Engineer

Texas Instruments
Broadband Communications Israel (BCIL)
Tel:  +972-9-9706587
________________________________________________________________________
"Those who do not understand Unix are condemned to reinvent it, poorly."
      -- Henry Spencer 




-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Nathan Bryant]

> Thanks for the tip, fortunately I don't have any SCSI or SATA disks. 
> Unfortunately, I still don't
> know how to do this. Is there an utility to put PC into sleep/suspend 
> ? I know I can switch
> to various S modes using /proc entry, but what about all the 
> peripherals and user mode apps ?
> I would venture that there has to be some kind of subsystem in user 
> mode to take care of
> all this. I probably did not phrase my question right, but when I 
> Asked if ACPI suspend/sleep
> status is stable enough I was actually referring to the whole 
> subsystem, not just kernel
> support.

It's not that big of a deal for userspace, you just need scripts to 
unload any problem device drives before suspend and reload them after 
resume. You would also need to kill and restart any programs that will 
be broken by those devices going away. APM already has config files that 
specify a lot of this which could probably be adapted to ACPI. But the 
ultimate goal is to get devices that don't even need to be unloaded, 
since we now have proper suspend/resume hooks. ;)

I think there is a "pmsuspend" program floating around out there, don't 
know how well it has been integrated with existing suspend/resume 
scripts though. Otherwise you can just talk to the kernel directly...

Nathan

Oh, and there's that annoying little problem where acpid sees the button 
press that I have to use to resume, as a shutdown event...


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Bernard Blackham]

On Wed, Aug 11, 2004 at 09:19:01AM -0400, Nathan Bryant wrote:
> > know how to do this. Is there an utility to put PC into
> > sleep/suspend ? I know I can switch to various S modes using
> > /proc entry, but what about all the peripherals and user mode
> > apps ?
> > I would venture that there has to be some kind of subsystem in
> > user mode to take care of all this. I probably did not phrase my
> > question right, but when I Asked if ACPI suspend/sleep status is
> > stable enough I was actually referring to the whole subsystem,
> > not just kernel support.
> 
> It's not that big of a deal for userspace, you just need scripts to 
> unload any problem device drives before suspend and reload them after 
> resume. You would also need to kill and restart any programs that will 
> be broken by those devices going away. APM already has config files that 
> specify a lot of this which could probably be adapted to ACPI. But the 
> ultimate goal is to get devices that don't even need to be unloaded, 
> since we now have proper suspend/resume hooks. ;)

For Software Suspend 2, I've written a package called hibernate
which can be used to activate Software Suspend 2, or alternately
utilise the /sys/power/state interface. (It doesn't manipulate
/proc/acpi/sleep, as I was under an impression it was getting
deprecated - on the other hand if it isn't, writing a plugin to do
so isn't hard).

http://developer.berlios.de/project/showfiles.php?group_id=1412
under "Current Scripts".

> I think there is a "pmsuspend" program floating around out there, don't 
> know how well it has been integrated with existing suspend/resume 
> scripts though.

Seems a Mandrake thing, but covers much the same ground as
hibernate. hibernate tries to cover most distributions.

> Oh, and there's that annoying little problem where acpid sees the button 
> press that I have to use to resume, as a shutdown event...

Does "rmmod button" just before suspending help?

Bernard.

-- 
 Bernard Blackham <bernard at blackham dot com dot au>


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Herman Sheremetyev]

On Wed, 2004-08-11 at 09:19, Nathan Bryant wrote:
<snip>
> Oh, and there's that annoying little problem where acpid sees the button 
> press that I have to use to resume, as a shutdown event...
</snip>

I don't have this issue myself since hitting any button on my laptop
wakes it up but I think it would be possible to get around this problem
by putting something like this into your acpid script that handles the
button press:

if test ! -e /tmp/button_lock
then
	/sbin/shutdown
fi

Then in your suspend script add:

touch /tmp/button_lock
echo -n mem > /sys/power/state
rm /tmp/button_lock 

It would need to be tweaked a bit since the rm may get executed before
the suspend completes and maybe acpid will get the button event some
time after resume so there might need to be a sleep statement but in
general I think it would solve the problem.

-Herman
	



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Nathan Bryant]

Nathan Bryant wrote:
> Alexander Sirotkin wrote:
> 
>>>> I could not find this information in neither mailing list archive 
>>>> nor the web site.
>>>> What is the current status of ACPI ? I'm mainly interested in one 
>>>> simple question -
>>>> does suspend work ? If so, how stable it is.
> 
> 
> Don't try it if you use any SCSI disks. IDE should be OK.
> 

I've just fixed S3 suspend for SCSI disks. See 
http://marc.theaimsgroup.com/?l=linux-scsi&m=109218292625235&w=2

Nathan


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Len Brown]

Alexander,
The status of ACPI S3 suspend/resume in 2.6.7 is...

Run the latest ACPI patch on top of 2.6.7
(indeed, you still need the latest patch for 2.6.8)

build EHCI as a module, and unload it first

echo 3 >/proc/acpi/sleep

(yes, it is on the list to move to sysfs...)

some systems get a double event if wakeup with power button.
I use the LID to wake up my system and that works fine.

many systems come back up from S3, but don't re-enable
the LCD, or some re-enable it, but in the wrong mode.
Sometimes you can wake it up by running X.
Sometiems you can wake it up by switching
virtual consoles.
There is a lot of experimentation in this area, but
we don't really have fixes for this yet that are
universally positive -- ie fix platforms w/o breaking others.

in summary, YMMV;-)

cheers,
-Len




-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Karol Kozimor]

Thus wrote Len Brown:
> some systems get a double event if wakeup with power button.
> I use the LID to wake up my system and that works fine.

Note I do get a spurious PWRF event even when waking with the keyboard.
Best regards,

-- 
Karol 'sziwan' Kozimor
sziwan-DETuoxkZsSqrDJvtcaxF/A@public.gmane.org


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

Re: current ACPI status in 2.6.7 kernel

[Ow Mun Heng]

On Wed, 2004-08-11 at 20:48, Len Brown wrote:
> Alexander,
> The status of ACPI S3 suspend/resume in 2.6.7 is...
> 
> Run the latest ACPI patch on top of 2.6.7
> (indeed, you still need the latest patch for 2.6.8)
> 
> build EHCI as a module, and unload it first
> 
> echo 3 >/proc/acpi/sleep
> 
> (yes, it is on the list to move to sysfs...)
> 

Here's a top Snipped portion of my acpid handle.pl script. Notice It's
on S4?

Unload all the USB modules/ network and stops MySql.

S4 works. S3 No. Screen does not come back. POST issue I guess.

Have not experimented with S3_bios

D600 Radeon 9000

#! /usr/bin/perl -w 

use strict;

        my $first_arg = shift (@ARGV);

        if ($first_arg =~ /^button\/lid$/) {
                system ('/usr/X11R6/bin/xset -display :0 dpms force off');
        }

        if ($first_arg =~ /^button\/sleep$/) {
                shift (@ARGV);
                shift (@ARGV);
                if (check_ignore ('/etc/acpi/flags/suspend')) {
                        clear_ignore ('/etc/acpi/flags/suspend');
                } else {
                        set_ignore ('/etc/acpi/flags/suspend');
                        system ('/usr/X11R6/bin/xset -display :0 dpms force off');
                        system ('rmmod tg3');
                        system ('rmmod usb_storage usbhid uhci_hcd ehci_hcd');
                        system ('/sbin/service mysqld stop');
                        system ('echo 4 > /proc/acpi/sleep && hwclock --hctosys');
                        system ('echo "Resuming proceses....');
                        system ('modprobe tg3');
                        system ('modprobe usb_storage');
                        system ('modprobe ehci_hcd');
                        system ('modprobe uhci_hcd');
                        system ('modprobe usbhid');
                        system ('/sbin/service mysqld start');
                        }






> some systems get a double event if wakeup with power button.
> I use the LID to wake up my system and that works fine.
> 
> many systems come back up from S3, but don't re-enable
> the LCD, or some re-enable it, but in the wrong mode.
> Sometimes you can wake it up by running X.
> Sometiems you can wake it up by switching
> virtual consoles.
> There is a lot of experimentation in this area, but
> we don't really have fixes for this yet that are
> universally positive -- ie fix platforms w/o breaking others.
> 
> in summary, YMMV;-)
> 
> cheers,
> -Len
> 
> 
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
> 100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
> Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
> http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel

-- 
Ow Mun Heng
Fedora GNU/Linux Core 2 on D600 1.4Ghz CPU kernel
2.6.7-2.jul1-interactive 
Neuromancer 04:32:53 up 4:34, 4 users, load average: 0.61, 0.54, 0.37 


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285