Bug 6767 - No lid events - Toshiba M50
Summary: No lid events - Toshiba M50
Status: REJECTED INSUFFICIENT_DATA
Alias: None
Product: ACPI
Classification: Unclassified
Component: Config-Interrupts (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: Zhang Rui
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-06-29 08:54 UTC by Scott Murphy
Modified: 2007-10-29 18:19 UTC (History)
3 users (show)

See Also:
Kernel Version: 2.6.17.1
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
Decompiled dsdt.dsl file (144.41 KB, text/plain)
2006-06-29 08:56 UTC, Scott Murphy
Details
ACPI dump (80.02 KB, text/plain)
2006-06-29 10:30 UTC, Scott Murphy
Details
add debug messages and remove two errors (142.86 KB, application/octet-stream)
2006-10-18 03:27 UTC, Zhang Rui
Details

Description Scott Murphy 2006-06-29 08:54:18 UTC
Most recent kernel where this bug did not occur:
None

Distribution:
Ubuntu 6.06 LTS Dapper Drake

Software environment:
Gnome 2.14.2
Gnome-power-manager 2.15.4
hald-0.5.4


Problem Description:

Laptop screen always down:
Bringing the screen up does not create an event. 
 
cat /proc/acpi/button/lid/LID/state
state:      closed

tilly@tilly-laptop:~$ sudo cat /proc/acpi/event
button/lid LID 00000080 00000002

First event is when I close the lid, but there isn't a second one when I open it.

--------------------------------------------
Suspend to mem:

Shuts down, power button flashes red. When I press the power button again, the
screen just goes to black and doesn't recover.

--------------------------------------------
Suspend to disk:

Goes to black and never recovers.


Steps to reproduce:
Comment 1 Scott Murphy 2006-06-29 08:56:28 UTC
Created attachment 8454 [details]
Decompiled dsdt.dsl file
Comment 2 Daniel Drake 2006-06-29 10:27:57 UTC
I have suggested to Scott that we use this bug solely for the lid state issue,
we'll file another bug for suspend problems.
Comment 3 Scott Murphy 2006-06-29 10:30:10 UTC
Created attachment 8455 [details]
ACPI dump
Comment 4 Len Brown 2006-07-01 10:11:05 UTC
Please try this test

kill acpid if it is running
# cat /proc/acpi/event

press the power button a few times -- do you see events?
do you see incrementing acpi entry in /proc/interrupts?

same for sleep button (if you have one)
same for lid button (similar to what you'veu done already,
but check /proc/interrupts)
Comment 5 Scott Murphy 2006-07-01 10:54:18 UTC
Pressing the power button is creating an event:
button/power PWRF 00000080 00000001

However, it doesn't seem to be upping the interrupts in /proc/interrupts:

169:      74952   IO-APIC-level  acpi, wifi0

It is staying on 169 when I press the power button and lower & lift the lid.
If I'm not reading that correctly, please reply.

Thanks

Comment 6 Scott Murphy 2006-07-01 12:14:43 UTC
Right, just discovered I was reading it incorrectly, so ignore last comment.

I've now killed my wifi0 which was sharing the IRQ (madwifi driver).

The interrupts are going up one at a time. One for power button, and one for
lowering and lifting (which I assume is actually meant to be two)

Thanks
Comment 7 Len Brown 2006-07-25 21:29:58 UTC
so the problem is that when you close the lid you get
an event, but you get no event when you open the lid?
Comment 8 Scott Murphy 2006-08-02 09:06:56 UTC
Yes, that is exactly right.
Comment 9 Zhang Rui 2006-10-18 03:27:58 UTC
Created attachment 9293 [details]
add debug messages and remove two errors

Please help me to try the test:
override your dsdt with the one I attched
kill acpid
#cat /proc/aacpi/event
open and close your Lid several times,
and attach the _dmesg_ output. :)
Comment 10 Zhang Rui 2007-09-27 20:27:34 UTC
Hi, Scott,

It seems that no GPE is generated when you open the lid.
Do you still have the problem in a later kernel? say 2.6.22?
If so, please ignore comment #9 and try the following test:
echo 0x04 >/sys/module/acpi/parameters/debug_layer
echo 0xffffffff > /sys/module/acpi/paramters/debug_level
close and open the lid
attach the dmesg output.
Comment 11 Fu Michael 2007-10-29 18:19:53 UTC
mark bug as unreproducible as no response from bug reporter. If reopen, please come with needed info.

Note You need to log in before you can comment on or make changes to this bug.