[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: can't restore from suspend



On Wed, Apr 16, 2003 at 01:53:36PM -0300, Derek Broughton wrote:
> That would be because it isn't part of the user's interface to ACPI.  How are
> you _getting_ to S3?  

This is the problem. I have absolutely no idea! I haven't configured
anything, it just does it. I know that S3 is the problem because I
manually typed in each of the Sx states until I got to the one that froze
my machine. I used:
	cd /proc/acpi
	echo 1 > sleep
	(nothing)
	echo 2 > sleep
	(nothing)
	echo 3 > sleep
	(frozen)

I did install acpid, but I haven't configured anything...

from /var/log/acpid
debian:/home/emmajane# more /var/log/acpid 
[Mon Apr 14 23:32:58 2003] starting up
[Mon Apr 14 23:32:58 2003] 1 rule loaded
[Tue Apr 15 01:13:54 2003] exiting
[Tue Apr 15 09:49:10 2003] starting up
[Tue Apr 15 09:49:10 2003] 1 rule loaded
[Tue Apr 15 10:56:08 2003] exiting
[Tue Apr 15 13:55:34 2003] starting up
[Tue Apr 15 13:55:34 2003] 1 rule loaded
[Tue Apr 15 19:03:27 2003] starting up
[Tue Apr 15 19:03:27 2003] 1 rule loaded
[Tue Apr 15 21:09:50 2003] exiting
[Wed Apr 16 09:39:18 2003] starting up
[Wed Apr 16 09:39:18 2003] 1 rule loaded
[Wed Apr 16 12:07:15 2003] starting up
[Wed Apr 16 12:07:15 2003] 1 rule loaded
[Wed Apr 16 12:32:10 2003] exiting
[Wed Apr 16 12:34:15 2003] starting up
[Wed Apr 16 12:34:15 2003] 1 rule loaded

I don't know where (or what) 1 rule is though...and I don't have the
/proc/acpi/events directory that's referred to in the man pages.

I tried doing dpkg-reconfigure acpid but all it seems to do is stop and
then start the daemon...it doesn't actually give me a configuration
screen.

> It isn't ACPI that tries to put the computer to sleep in
> the first place, though you can have set up acpid (which isn't at all the same
> thing) to do it.  ACPI merely generates or responds to events - some external
> process has to write to /proc/acpi/sleep to put it into S3.

Ok.

> You could disable sleep states in ACPI, but it requires rewriting the DSDT table
> in the BIOS.  The DSDT essentially tells ACPI what hardware options are
> supported.

I'm more afraid of the BIOS than I am of the kernel. ;) I did reboot into
Windows and set all my power save functions to "never." I saw a reference
to it somewhere...

> It sounds to me much more likely that your computer has _never_ gone to S3 but
> that your BIOS might be attempting to do something outside of ACPI. What
> evidence do you have?  

I'm not sure where to look.. this is from dmesg | grep ACPI

 BIOS-e820: 000000000fef0000 - 000000000feff000 (ACPI data)
 BIOS-e820: 000000000feff000 - 000000000ff00000 (ACPI NVS)
ACPI: Core Subsystem version [20011018]
ACPI: Subsystem enabled
ACPI: System firmware supports S0 S3 S4 S5
ACPI: Battery socket found, battery present
ACPI: AC Adapter found
ACPI: Thermal Zone found
ACPI: Thermal Zone found
ACPI: Power Button (FF) found
ACPI: Multiple power buttons detected, ignoring fixed-feature
ACPI: Power Button (CM) found
ACPI: Sleep Button (CM) found
ACPI: Lid Switch (CM) found

(/var/log/messages has only the equivalent date stamped lines for "ACPI")

/var/log/syslog has only the following information:
Apr 16 12:34:15 debian kernel: ACPI: Core Subsystem version [20011018]
Apr 16 12:34:15 debian kernel: ACPI: Subsystem enabled
Apr 16 12:34:15 debian kernel: ACPI: System firmware supports S0 S3 S4 S5
Apr 16 12:34:15 debian kernel: EC: found, GPE 34
Apr 16 12:34:15 debian kernel: ACPI: Battery socket found, battery present
Apr 16 12:34:15 debian kernel: ACPI: AC Adapter found
Apr 16 12:34:15 debian kernel: ACPI: Thermal Zone found
Apr 16 12:34:15 debian kernel: ACPI: Thermal Zone found
Apr 16 12:34:15 debian kernel: Processor[0]: C0 C1 C2 C3, 8 throttling
states
Apr 16 12:34:15 debian kernel: ACPI: Power Button (FF) found
Apr 16 12:34:15 debian kernel: ACPI: Multiple power buttons detected,
ignoring fixed-feature
Apr 16 12:34:15 debian kernel: ACPI: Power Button (CM) found
Apr 16 12:34:15 debian kernel: ACPI: Sleep Button (CM) found
Apr 16 12:34:15 debian kernel: ACPI: Lid Switch (CM) found


> If I really thought this was an ACPI issue, I'd suggest
> that acpi-support@lists.sourceforge.net would be a more appropriate place to ask

I'm already subscribed but I'm not even sure if my question is for them
yet. I was hoping their list would be a little higher volume and that
someone else would ask first. ;)

emma

-- 
Emma Jane Hogbin
[[ 416 417 2868 ][ www.xtrinsic.com ]]



Reply to: