Skip to contentSkip to conversation

2.6.9-rc1 : Weirdness after shutdown - ACPI or Suspend bug?

DISCUSSION Stuart Young 3 participants 5 messages Sep 1, 2004 – Sep 4, 2004

lore
Share & tools
lore
#0 lore raw

OK, this one is weirding me out.

Note that when using 2.6.8.1 all is fine. The following situation only happens in 2.6.9-rc1 or later.

If I shutdown my laptop (ie: halt) it goes through the motions and everything goes off. If the lid switch changes state AFTER powerdown, the laptop starts up. Removing AC power, or with AC power connected and removing the battery does not trigger this, just seemingly the lid switch. This works on lid close, AND lid open.

I noticed it and thought my laptop was dying, and consequently made sure I shut the lid before the machine actually powered off, and didn't correlate it to 2.6.9-rc1. But when I was trying to compile 2.6.9-rc1-mm2 I had some other issues (general weirdness), and decided to boot back to 2.6.8.1 in case 2.6.9-rc1 was at fault. When I shut down I noticed it didn't have this behaviour, and went and did a little reboot test at random between 2.6.9-rc1 & 2.6.8.1 that showed this as consistent behaviour.

Note that once the laptop restarts, if I shut it down using the power switch at the lilo prompt, the machine stays off, regardless of the lid switch state.

Any ideas or suggestions? Going to start going back thru csets when I get a chance if no one has an ideas.

Signature · 2 lines
 Stuart Young (aka Cef)
 [email protected] is for LKML and related email only
#1 lore raw
Quoted 10 lines · Stuart Young

OK, this one is weirding me out.

Note that when using 2.6.8.1 all is fine. The following situation only happens in 2.6.9-rc1 or later.

If I shutdown my laptop (ie: halt) it goes through the motions and everything goes off. If the lid switch changes state AFTER powerdown, the laptop starts up. Removing AC power, or with AC power connected and removing the battery does not trigger this, just seemingly the lid switch. This works on lid close, AND lid open.

Len, I've tentatively traced this down to the addition of the ACPI wakeup_devices module that went into the kernel via ACPI 20040715.

Quoted 1 line

From a quick look at the code, the wakeup devices get set at boot, but on

shutdown, they don't get unset. Is this intentional?

Any clues, ideas, or suggestions?

Notes:
Asus L7300/L7200 series laptop
Latest BIOS from the Asus website
PIII-600 on Intel 82440MX chipset

Signature · 2 lines
 Stuart Young (aka Cef)
 [email protected] is for LKML and related email only
#2 lore raw
Quoted 4 lines · Stuart Young
Quoted 2 lines · Stuart Young

OK, this one is weirding me out.

Quoted 9 lines
Quoted 8 lines

Note that when using 2.6.8.1 all is fine. The following situation only happens in 2.6.9-rc1 or later.

If I shutdown my laptop (ie: halt) it goes through the motions and everything goes off. If the lid switch changes state AFTER powerdown, the laptop starts up. Removing AC power, or with AC power connected and removing the battery does not trigger this, just seemingly the lid switch. This works on lid close, AND lid open.

Interesting - I'm playing with a Sony VAIO S170p, and when you use shutdown -h, it turns off and turning it back on appears to be fine.

HOWEVER, using swsusp2, hibernating the machine works, but it refuses to wake back up until you pull the AC power and battery out.

thomas

#3 lore raw
Quoted 2 lines · Thomas Davis

HOWEVER, using swsusp2, hibernating the machine works, but it refuses to wake back up until you pull the AC power and battery out.

on my thinkpad t40 after suspending to ram the machine wakes up but then it shuts off. shutdown works fine though.

--alex--

Signature · 4 lines
| I believe the moment is at hand when, by a paranoiac and active |
|  advance of the mind, it will be possible (simultaneously with  |
|  automatism and other passive states) to systematize confusion  |
|  and thus to help to discredit completely the world of reality. |
#4 lore raw
In reply to Stuart Young
Quoted 18 lines · Stuart Young
Quoted 10 lines · Stuart Young

OK, this one is weirding me out.

Note that when using 2.6.8.1 all is fine. The following situation only happens in 2.6.9-rc1 or later.

If I shutdown my laptop (ie: halt) it goes through the motions and everything goes off. If the lid switch changes state AFTER powerdown, the laptop starts up. Removing AC power, or with AC power connected and removing the battery does not trigger this, just seemingly the lid switch. This works on lid close, AND lid open.

Len, I've tentatively traced this down to the addition of the ACPI wakeup_devices module that went into the kernel via ACPI 20040715.

From a quick look at the code, the wakeup devices get set at boot, but on shutdown, they don't get unset. Is this intentional?

Quoted 6 lines

Any clues, ideas, or suggestions?

Notes:
Asus L7300/L7200 series laptop
Latest BIOS from the Asus website
PIII-600 on Intel 82440MX chipset

Further info. Using 2.6.9-rc1-mm3 does not change this behaviour at all. Seems that 2.6.9-rc1-mm3 contains 20040816, so this is still a running issue.