You are not logged in.

#1 2012-10-06 21:46:49

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Kernel doesn't log ACPI events: ACPI module or Bios corruptions

Well, this is very strange and i dunno where to look at to solve this:

My new NP530U3C (Samusung ultrabook)is running Arch and i'm currently using acpid to trigger the acpi events couse i have no DE (fluxbox).

I can't understand the way my kernel and acpid log the LID event:


Kernel.log

Oct  6 20:06:46 localhost kernel: [    2.105259] ACPI: Lid Switch [LID0]
Oct  6 20:44:31 localhost kernel: [    2.111136] ACPI: Lid Switch [LID0]
Oct  6 23:03:18 localhost kernel: [    2.111244] ACPI: Lid Switch [LID0]
Oct  6 23:17:14 localhost kernel: [    2.088524] ACPI: Lid Switch [LID0]

Messages.log

Oct  6 20:06:46 localhost kernel: [    2.105259] ACPI: Lid Switch [LID0]
Oct  6 20:44:31 localhost kernel: [    2.111136] ACPI: Lid Switch [LID0]
Oct  6 20:45:29 localhost logger: LID closed
Oct  6 20:45:48 localhost logger: LID opened
Oct  6 20:53:58 localhost logger: LID closed
Oct  6 20:54:07 localhost logger: LID opened
Oct  6 23:00:04 localhost logger: LID closed
Oct  6 23:00:16 localhost logger: LID opened
Oct  6 23:03:18 localhost kernel: [    2.111244] ACPI: Lid Switch [LID0]
Oct  6 23:17:14 localhost kernel: [    2.088524] ACPI: Lid Switch [LID0]
Oct  6 23:18:32 localhost logger: LID closed
Oct  6 23:18:40 localhost logger: LID opened
Oct  6 23:24:33 localhost logger: LID closed
Oct  6 23:24:40 localhost logger: LID opened
Oct  6 23:27:18 localhost logger: LID closed
Oct  6 23:27:28 localhost logger: LID opened
Oct  6 23:33:40 localhost logger: LID closed
Oct  6 23:33:48 localhost logger: LID opened

When i close the lid, i have to wait 10 sec approx to trigger the LID close event and start the suspend action!
on wakeup i don't know because my samsung can wakeup only with the power button (no lid in wakeup events).

So somebody can enlighten me?

*EDIT <-----

I can summarize it is a corruption of the Bios/ACPI module that affect this laptop after a certain suspentions circle (I mean not every first suspentions).
I would like to debug in a better way but i have no instruments to investigate:
If someone can point me to a way to debug this stuff like Bios corruptions and/or Kernel module failure...

Any advide is appreciated

Last edited by j0lly (2012-10-29 14:29:44)


...why not

Offline

#2 2012-10-06 22:00:09

WonderWoofy
Member
From: Los Gatos, CA
Registered: 2012-05-19
Posts: 8,414

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

NP530U3C?  We may be a bunch of techie geeks here, but seriously, just a model number?

Offline

#3 2012-10-06 22:21:31

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

well, is a samsung ultrabook! i'm doing some m ore tests and still 10 sec delayed!
a note:
when the kernel log the LID state i can suspend instantly after i really close the LID.
after a reboot, and halt/start thigs doesn't changes....
The kernel logger resume after some reboots or *resetting* (using a needle to click a button above the laptop close to the battery) the laptop <-- this method was whrote somewhere and i tried it, and solved sometinms the issue for 2/3 days.
is starnge because it appears without any regularity, but i'm not so pro and is difficult for me to figure out what can trigger the event!

EDIT

Oct  7 00:21:58 localhost kernel: [  686.653890] psmouse serio1: Touchpad at isa0060/serio1/input0 lost sync at byte 6
Oct  7 00:21:59 localhost kernel: [  686.668774] psmouse serio1: Touchpad at isa0060/serio1/input0 - driver resynced.

dunno if is important, but after a resume from suspend, i got this messege. i thought it was only after it (like a little error in the resume hooks) but no i'm noticing it in  the messages.log every how

Last edited by j0lly (2012-10-06 22:24:03)


...why not

Offline

#4 2012-10-06 22:24:52

WonderWoofy
Member
From: Los Gatos, CA
Registered: 2012-05-19
Posts: 8,414

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

Are you using the regular acpi module?  Because I am pretty sure there is a samsung_laptop module as well.  Maybe that would suit you better?

Offline

#5 2012-10-06 22:41:35

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

well, i have both i think: i'm pretty sure i've the samsung_laptop installed

[root@0x0000 j0lly]# lsmod | grep samsung
samsung_laptop          8743  0 

no mention in lsmod for acpi module: how could i look if it is loaded?

anyway, isn't the acpi module compiled directly in the kernel? (standard arch)


...why not

Offline

#6 2012-10-06 23:08:11

WonderWoofy
Member
From: Los Gatos, CA
Registered: 2012-05-19
Posts: 8,414

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

I guess it is... in any case, you are in fact using the samsung_acpi module.  This has been going on with the 3.5.4 kernel right?  Because there was a kernel update today (maybe yesterday), so I wonder if you have upgraded or not yet?

Offline

#7 2012-10-06 23:21:25

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

[root@0x0000 j0lly]# uname -r
3.5.4-1-ARCH

but this is never changed from the first ime i installed arch on the laptop, but, as i was fixing other stuff and i was not so experienced, i thought could be related to other things and had to tweiak a lot so now i can say is acpi related and is from at least last 3/4 kernel release...
i gonna try this new kernel and see if fix something


...why not

Offline

#8 2012-10-06 23:26:08

WonderWoofy
Member
From: Los Gatos, CA
Registered: 2012-05-19
Posts: 8,414

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

As of today

> uname -r
3.5.5-1-ARCH

Offline

#9 2012-10-07 00:10:02

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

rebooted with 3.5.5-1 but nothing chnges...
i think i need something more specific... i'm looking at the dsdt decompiled and in dmesg:
there's a LID0 switch in both but i really can,t investigate further without some help!
at now i still have 10 sec delay before i can suspend with systemd and log in messages.log with acpid.


...why not

Offline

#10 2012-10-07 13:03:05

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

I probably isolated the cause of it: when laptop suspend, on wakeup the kernel stop logging events:
the powerbutton, volume, monitor switch and some more are logged at least in acpid but, as i say, the plug, unplug of the ac adapter, the battery changing state and the lid is not logged at all!

udev not logging eather, while was before....

I'm little bit frustrated for this and i would like to dig in it, but need to understand the kernel/acpi/udev linking and the way works....
i mean, why the kernel log sometimes the events but acpid log these same that the kernel not log???
have udev got the signals? from kernel? and acpid?
please, i really need some teaching class to understand the problem and see if is a acpi bug or a misconfiguration in, for example, my suspend/wakeup procedure (that is default from pm-utils and systemd.

every little help is really appreciated!


...why not

Offline

#11 2012-10-12 07:16:11

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

Still can't figure out what prevent acpid/udev to log porperly the events.
Also the Kernel stop log in. In sys/class/power_supply ADP1 is right, BAT not... and /proc/acpi/button/LID also is wrong and always say open!
Is this an ACPI bug? a bugged BIOS or a Kernel/config issue?


...why not

Offline

#12 2012-10-29 14:33:11

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

No clue?


...why not

Offline

#13 2012-11-08 18:44:00

IcePik
Member
From: Canterbury, UK
Registered: 2008-12-14
Posts: 31

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

Hey,

I have the Samsung NP530U3C too and am facing the same problems, the kernel fails to log the lid and power adapter events.

After a bit of digging I found this bug on the kernel bug tracker

https://bugzilla.kernel.org/show_bug.cgi?id=44161

Hope it helps explain something.

Do let me know if you get it working


IcePik
Samsung Series 5 Ultrabook NP530U3C-A01US
Intel® Core™ i5-3317U CPU @ 1.70GHz  |  4 GB DDR3  |  Intel® HD Graphics 4000 (Intel i915)  |  Intel® Centrino Advanced-N 6235

Offline

#14 2012-11-08 22:16:58

j0lly
Member
From: ~Milan
Registered: 2011-08-16
Posts: 49

Re: Kernel doesn't log ACPI events: ACPI module or Bios corruptions

Thanks, i'm already there!
If you need to reenable ACPI logging event you have to use a needle as described few posts above.
Hope the guy that discovered the core problem will aslo solve it smile

Let's see what's happen, but is definitly out of my knowledge..


...why not

Offline

Board footer

Powered by FluxBB