ASUS M3N and sleep states

ASUS M3N and sleep states

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

I made some more tests about the sleep states on an ASUS M3N (Centrino
based). Even if I never got a complete resume, here the results:

- - S1
obviously, if I leave the USB module on, the resume process hangs trying
to set the IRQ for the USB controllers. So, I unload the USB modules and
the resume process hangs again. I found that the other problem are the
ALSA modules: if I unload them (and the USB as well), the resume process
goes well, but it automatically enters in INIT 0 (so it reboots). I
don't know how I could get a resume to my previous INIT...

- - S3
without USB and ALSA, the fan turns on suddendly after the 'echo 3 >
/proc/acpi/sleep', as when my OS hangs for other reasons. IMHO this is a
problem with the BIOS (and the DSDT), as once I shut down the PC
pressing 5sec the power-button, I need to press 2 times the power-button
to reach again a boot (the first time I press the power-button, the
laptop powers on and suddenly powers off).

- - S4
without USB and ALSA, I got a clear power-down and for the resume
(passing to the kernel the option 'resume=/dev/hda6' which is my swap
partition), the OS seems to resume correctly, it sees my resume command,
reads the resume partition, but then it reboots and my swap partition is
corrupted (I need a mkswap on my swap partition).

I use a kernel 2.6.0, without any ACPI patch, and I enable all the
'Power Management' options, as in my .config here:
	#
	# Power management options (ACPI, APM)
	#
	CONFIG_PM=y
	CONFIG_SOFTWARE_SUSPEND=y
	CONFIG_PM_DISK=y
	CONFIG_PM_DISK_PARTITION="/dev/hda6"

	#
	# ACPI (Advanced Configuration and Power Interface) Support
	#
	CONFIG_ACPI=y
	CONFIG_ACPI_BOOT=y
	CONFIG_ACPI_INTERPRETER=y
	CONFIG_ACPI_SLEEP=y
	CONFIG_ACPI_SLEEP_PROC_FS=y
	CONFIG_ACPI_AC=m
	CONFIG_ACPI_BATTERY=m
	CONFIG_ACPI_BUTTON=m
	CONFIG_ACPI_FAN=m
	CONFIG_ACPI_PROCESSOR=m
	CONFIG_ACPI_THERMAL=m
	CONFIG_ACPI_ASUS=m
	# CONFIG_ACPI_TOSHIBA is not set
	CONFIG_ACPI_DEBUG=y
	CONFIG_ACPI_BUS=y
	CONFIG_ACPI_EC=y
	CONFIG_ACPI_POWER=y
	CONFIG_ACPI_PCI=y
	CONFIG_ACPI_SYSTEM=y
	CONFIG_ACPI_RELAXED_AML=y

If someone could point me the problems/bugs...

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/7XKIVAp7Xm10JmkRAuxlAJ9QfenGXhgCBnm2htmoBztQ7yqzawCcCWbP
Hjxd5LkTCuBSUgokIBGqkZs=
=EAkm
-----END PGP SIGNATURE-----



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Jean-Michel POURE]

Le Samedi 27 Décembre 2003 12:52, Luca Capello a écrit :
> I made some more tests about the sleep states on an ASUS M3N (Centrino
> based). Even if I never got a complete resume, here the results:

I can confirm the very same problems on my ASUS 3800LC laptop, same 
configuration.

By the way, can you turn off the computer using "shutdown now -h"? On my 
laptop, init 0 hangs on without turning off the computer. I have to press the 
power button during 4 seconds or so.

Cheers,
Jean-Michel Pouré



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id\x1278&alloc_id371&op=click

Re: ASUS M3N and sleep states

[Pavel Machek]

Hi!

> I use a kernel 2.6.0, without any ACPI patch, and I enable all the
> 'Power Management' options, as in my .config here:
> 	#
> 	# Power management options (ACPI, APM)
> 	#
> 	CONFIG_PM=y
> 	CONFIG_SOFTWARE_SUSPEND=y
> 	CONFIG_PM_DISK=y
> 	CONFIG_PM_DISK_PARTITION="/dev/hda6"

Select either PM_DISK or SOFTWARE_SUSPEND but not both. Thanks.
								Pavel

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


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 12/27/03 18:33, Pavel Machek wrote:
| Select either PM_DISK or SOFTWARE_SUSPEND but not both. Thanks.
ok, I tried the following without the AGP/DRM modules (specifically,
'agpgart', 'intel-agp' and 'i830'):

- - PM_DISK > the OS enters ok in S4, but the resume process doesn't work
(not with resume=/dev/hda6, not with pmdisk=/dev/hda6 neither without).
I got a resume on my ext3 partitions, but the swap isn't lost

- - SWSUSP > it works! But, obviously, as I wrote above, without the
AGP/DRM modules. Anyway, I got a perfect S4, a perfect resume and after
the resume I can successfully resume or shutdown :-)

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/8cnLVAp7Xm10JmkRAnBsAKCFwN1+Ne4u3wd8HgVZB07z80r5aQCgjhVw
xLuaLzjm87MZleKdnsOKZrg=
=LeHG
-----END PGP SIGNATURE-----




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Karol Kozimor]

Thus wrote Luca Capello:
> - - S1
> obviously, if I leave the USB module on, the resume process hangs trying
> to set the IRQ for the USB controllers. So, I unload the USB modules and
> the resume process hangs again. I found that the other problem are the
> ALSA modules: if I unload them (and the USB as well), the resume process

[the following is for ASUS L3800C]
Same here for USB (actually, unloading {usb-,}uhci suffices, 2.6 doesn't
hang though). ALSA doesn't seem to care in my case.
About the shutdown: it is likely that a PWRF event is issued after resume
like in my case, kill acpid prior to suspending and see if it makes any
difference (I reported the PWRF problem here a while ago, but have
received no feedback so far).

> - - S4
> without USB and ALSA, I got a clear power-down and for the resume
> (passing to the kernel the option 'resume=/dev/hda6' which is my swap
> partition), the OS seems to resume correctly, it sees my resume command,
> reads the resume partition, but then it reboots and my swap partition is
> corrupted (I need a mkswap on my swap partition).

Kill X and unload AGP or use Nigel's patches from swsusp.sf.net.
Best regards,

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


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 12/27/03 22:51, Karol Kozimor wrote:
|>- - S1
<cut>
| [the following is for ASUS L3800C]
| Same here for USB (actually, unloading {usb-,}uhci suffices, 2.6 doesn't
| hang though). ALSA doesn't seem to care in my case.
| About the shutdown: it is likely that a PWRF event is issued after resume
| like in my case, kill acpid prior to suspending and see if it makes any
| difference (I reported the PWRF problem here a while ago, but have
| received no feedback so far).
sometimes I'm a bit stupid... You're right, there's a PWRF event once
you resume from S1 and this is the cause for the entering in 'INIT 0'.
Stopping 'acpid' or renaming the action associated to the PWRF event all
goes well: I got a complete S1 and a complete resume. There're 2 new
problems, now: after resuming from S1, the fan turns on loudly and I
can't reboot (but the 'shutdown -h now' works).

About the PWRF event, I sent this mail to the 'acpid' author: it seems
that any button you press to resume from S1 (so even not the
power-button), the 'acpid' generates a PWRF event. This should be avoid,
I mean, it seems that this happens only in S1 state and not in S3.

About the ALSA problem, maybe the *real* problem is related to the fact
that I've an 'osl-0892/semaphore' error in my 'dmesg' and at boot when
trying to load ALSA drivers (but they work without any problems): IIRC
this is a known problem on ASUS Centrino laptops (I'm sure at least on
ASUS M3N) and AFAIK there's no solution ATM.

|>- - S4
<cut>
| Kill X and unload AGP or use Nigel's patches from swsusp.sf.net.
I answered in the Pavel's mail about AGP/DRM modules. Question: the
Nigel's patches let you leave the AGP modules on?

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/8c0MVAp7Xm10JmkRAo2+AJ4wCjY6ywGHpldiGlJdh1Pwvrp+KgCfQsYK
LyUlxjrdDn5rcngRBdh0rOQ=
=Spls
-----END PGP SIGNATURE-----




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Karol Kozimor]

Thus wrote Luca Capello:
> About the PWRF event, I sent this mail to the 'acpid' author: it seems
> that any button you press to resume from S1 (so even not the
> power-button), the 'acpid' generates a PWRF event. This should be avoid,
> I mean, it seems that this happens only in S1 state and not in S3.

It's got nothing to do with acpid, it's just a userspace daemon, anyway.
The real problem is either somewhere in BIOS (and no, it's not in the AML
either) or in the ACPI code. For the record, I get the PWRF also after S3.

And no, I'm not pressing the power button to resume.

> About the ALSA problem, maybe the *real* problem is related to the fact
> that I've an 'osl-0892/semaphore' error in my 'dmesg' and at boot when
> trying to load ALSA drivers (but they work without any problems): IIRC
> this is a known problem on ASUS Centrino laptops (I'm sure at least on
> ASUS M3N) and AFAIK there's no solution ATM.

ALSA's suspend / resume support is not really complete, judging on my
experience. My impression is that it currently "works" by chance on my
machine (ICH3M - CS4299).

> | Kill X and unload AGP or use Nigel's patches from swsusp.sf.net.
> I answered in the Pavel's mail about AGP/DRM modules. Question: the
> Nigel's patches let you leave the AGP modules on?

Well, basically, yes, i.e. I'm able to suspend and resume with a DRI app
running, but that's with Radeon-specific patches to the DRM and DRI
modules. On the other hand, even with those patches both pmdisk and Pavel's
swsusp would reboot at resume if X was running during suspend.
Best regards,

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


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Tim Hockin]

On Tue, Dec 30, 2003 at 08:07:57PM +0100, Luca Capello wrote:
> sometimes I'm a bit stupid... You're right, there's a PWRF event once
> you resume from S1 and this is the cause for the entering in 'INIT 0'.
> Stopping 'acpid' or renaming the action associated to the PWRF event all
> goes well: I got a complete S1 and a complete resume. There're 2 new
> problems, now: after resuming from S1, the fan turns on loudly and I
> can't reboot (but the 'shutdown -h now' works).

I don't know any of the backstory on these - what's up?  Anything I need to
look at?

> About the PWRF event, I sent this mail to the 'acpid' author: it seems
> that any button you press to resume from S1 (so even not the
> power-button), the 'acpid' generates a PWRF event. This should be avoid,
> I mean, it seems that this happens only in S1 state and not in S3.

I *promise* you there is no PWRF event that acpid generates.  All acpid does
is relay the stuff your kernel sends.  And the kernel mostly relays BIOS
stuff.

acpid does not create anything on it's own.


-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 01/05/04 19:19, Tim Hockin wrote:
| I don't know any of the backstory on these - what's up?  Anything I
need to
| look at?
sorry Tim, my mystake: here the complete thread
http://sourceforge.net/mailarchive/forum.php?thread_id=3676335&forum_id=6102
(I can't find my starting post, it seems SourceForge.net archive isn't
complete or I've no other reasons. Anyway, the 'incriminated' messages
are there)

| I *promise* you there is no PWRF event that acpid generates.  All
acpid does
| is relay the stuff your kernel sends.  And the kernel mostly relays BIOS
| stuff.
|
| acpid does not create anything on it's own.
Another mystake of mine, after my post I forgot to send you the reponse
from Karol Kozimor, who better explained my problem. Here his words:

on 12/30/03 21:56, Karol Kozimor wrote:
| It's got nothing to do with acpid, it's just a userspace daemon, anyway.
| The real problem is either somewhere in BIOS (and no, it's not in the AML
| either) or in the ACPI code. For the record, I get the PWRF also after S3.
|
| And no, I'm not pressing the power button to resume.

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/+a8HVAp7Xm10JmkRAnAVAJ9lL9fGyZTCDuCweg+oPaaR9lt5ngCffQn8
QrAthBcIRDMC0MBjMMfXGw0=
=XLrx
-----END PGP SIGNATURE-----



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

Re: ASUS M3N and sleep states

[Luca Capello]

[-- Attachment #1: Type: text/plain, Size: 1020 bytes --]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 12/30/03 20:07, Luca Capello wrote:
| sometimes I'm a bit stupid... You're right, there's a PWRF event once
| you resume from S1 and this is the cause for the entering in 'INIT 0'.
| Stopping 'acpid' or renaming the action associated to the PWRF event all
| goes well: I got a complete S1 and a complete resume. There're 2 new
I hacked a little the general 'power-button action' included in the
'acpid', to act correctly even when resuming from S1. I attached it :-)
On my box it works without any problems, even if I don't use a lot S1.

If someone wants to use it, you should modify /etc/acpi/events/power (or
whatever you have associated to 'power-button event') to call this script.

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/+bisVAp7Xm10JmkRAs9+AJ9POpmd6AAVk+wYHCv8bjz44axxUgCgg4gc
DKPgh2Gw76+OAqNjspJJhUU=
=fnlB
-----END PGP SIGNATURE-----

[-- Attachment #2: btn_power.sh --]
[-- Type: application/x-sh, Size: 628 bytes --]

Re: ASUS M3N and sleep states

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 12/27/03 12:52, Luca Capello wrote:
| - S3
| without USB and ALSA, the fan turns on suddendly after the 'echo 3 >
| /proc/acpi/sleep', as when my OS hangs for other reasons. IMHO this is a
| problem with the BIOS (and the DSDT), as once I shut down the PC
| pressing 5sec the power-button, I need to press 2 times the power-button
| to reach again a boot (the first time I press the power-button, the
| laptop powers on and suddenly powers off).
I should say that this is not exactly: I made some other tests and
actually resuming from S3 is ok in 90% but the screen remains blank.
Sometimes the fan turns on and I can't shutdown (but this happens when
the temperature is too high). The other times, I can launch any
commands: I see the HD light blinking and I heard the HD working. More,
'shutdown -h now' works and I get a clean shutdown (but 'reboot' doesn't
work.

Here the complete situation: I can get S3 and a complete clear resume
even with the ALSA modules on and AGP. I tried it once I was logged in X
via GDM and it worked :-) This didn't with the USB modules on (as it was
supposed to be).

Another striking thing is that while normally the screen remains blank
during shutdown, if I enter in S3 from an X terminal and then I start
the shutdown (always without seeing my typing), then I can see the first
console (tty1) appearing and showing some messaged about the shutdown
process (just the "Shutdown called from..." and "acpid_power_off_called"
or something similar). This doesn't happen if I try from tty2, I mean, I
can't see tty1 in any other ways.

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/81lJVAp7Xm10JmkRAgDNAJ9Om3RmoAbTv11Bw68v3ImhEUyaegCggyi5
IiADUOXuEXYKUKnfq1NAy1k=
=k5+z
-----END PGP SIGNATURE-----



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click

RE: ASUS M3N and sleep states

[Yu, Luming]

>- - S1
>obviously, if I leave the USB module on, the resume process hangs
trying
>to set the IRQ for the USB controllers. So, I unload the USB modules
and
>the resume process hangs again. I found that the other problem are the
>ALSA modules: if I unload them (and the USB as well), the resume
process
>goes well, but it automatically enters in INIT 0 (so it reboots). I
>don't know how I could get a resume to my previous INIT...

I have 2 tracker about S1 issues of ASUS notebook.

http://bugzilla.kernel.org/show_bug.cgi?id=1557 

http://bugzilla.kernel.org/show_bug.cgi?id=1588 



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id\x1278&alloc_id371&op=click

Re: ASUS M3N and sleep states

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 12/29/03 02:35, Yu, Luming wrote:
| I have 2 tracker about S1 issues of ASUS notebook.
|
| http://bugzilla.kernel.org/show_bug.cgi?id=1557
this is a common problem with USB modules loaded before entering in S1.
AFAIK there's no solution for this but unloading USB modules before
entering in S1 (and if the modules are compiled in, this isn't possible,
right?).

| http://bugzilla.kernel.org/show_bug.cgi?id=1588
I tried your patch, but it doesn't solve my problem: I got a shutdown
without a resume and at reboot I had a recovery for my ext3 partition.

As I posted before, I solved the problems related to S1: it's the
'acpid', because as reported by Karol resuming from S1 generates a PWFR
events and so the 'acpid' acts as for a "normal" PWFR (not from a
resuming one). I hope I was clear :-)

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQE/8eBPVAp7Xm10JmkRAiN4AJ9qvowalxG1usOeZ7n9f67r6F0j2ACcCULs
dcW8isqcKQVrk6wInzNrcKA=
=j8Yf
-----END PGP SIGNATURE-----



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click