2.6 won't poweroff

DISCUSSION Emma Jane Hogbin 7 participants 18 messages May 25, 2004 – May 31, 2004 12 messages not in the archive ACPI

Share & tools

Emma Jane Hogbin

author

2004-05-25 04:45 UTC #0 lore raw

I have two big irritants for the 2.6.5/2.6.6 (patched for ACPI) kernels with ACPI.

1. My computer no longer powers off. acpi_power_off is called, but the computer doesn't actually turn itself off. In fact when I use "poweroff" directly I get a message flashing across my screen saying:
syntax error on 658 of stdin

2. Random freezes. It's not just the touch pad or the keyboard, but rather the entire monitor seems to freeze from time to time and for no apparent reason. Gaim and xmms are usually (always?) open when it happens, but I can't believe that either of them would cause my system to freeze.

I have an AcerTM 634 that did not have either of these two problems with 2.4.22. As requested on acpi.sf.net I have also made available:

www.xtrinsic.com/geek/config/2.6.6/dmidecode.txt
www.xtrinsic.com/geek/config/2.6.6/interrupts.txt
www.xtrinsic.com/geek/config/2.6.6/lscpi.txt
www.xtrinsic.com/geek/config/2.6.6/dmesg.txt

(and 2.6.5 by changing the kernel number) I installed the PMTools Debian package, but it doesn't seem to include acpidmp. Is there a Debian package for it?

Any suggestions would be much appreciated.

thanks,
emma

Signature · 9 lines

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


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click

A message that is not in the archive (sent off-list or not yet received): [email protected]

Fred Labrosse

2004-05-25 08:14 UTC #2 lore raw

Quoted 8 lines · Emma Jane Hogbin

I have two big irritants for the 2.6.5/2.6.6 (patched for ACPI) kernels with ACPI.
1. My computer no longer powers off. acpi_power_off is called, but the computer doesn't actually turn itself off. In fact when I use "poweroff" directly I get a message flashing across my screen saying:
syntax error on 658 of stdin

Sorry, no solution, but exactly the same problem with an Acer TM636 (although I don't have any syntax error, only "power down").

This is a really annoying problem, even dangerous for the laptop as it sometimes, randomly, "shuts down" because of the erratic temperature reading but doesn't power off so stays there, still powered, still heating but not cooling down (fan off)...

I did try various combinations (APIC off, on, no compiled in, went back to 2.4.26, 2.4.25-gentoo-r2, 2.6.4, 2.6.6-rc2(?)), to no avail. It used to work fine when I was running RH8 with (I think) 2.4.24 (I now run gentoo).

Any suggestions would be more than appreciated.

TIA,

Fred

-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click

A message that is not in the archive (sent off-list or not yet received): [email protected]

Nico Sabbi

2004-05-25 08:59 UTC #4 lore raw

Quoted 21 lines · Fred Labrosse

Quoted 7 lines · Emma Jane Hogbin
I have two big irritants for the 2.6.5/2.6.6 (patched for ACPI) kernels with ACPI.
1. My computer no longer powers off. acpi_power_off is called, but the computer doesn't actually turn itself off. In fact when I use "poweroff" directly I get a message flashing across my screen saying:
syntax error on 658 of stdin
Sorry, no solution, but exactly the same problem with an Acer TM636 (although I don't have any syntax error, only "power down").
This is a really annoying problem, even dangerous for the laptop as it sometimes, randomly, "shuts down" because of the erratic temperature reading but doesn't power off so stays there, still powered, still heating but not cooling down (fan off)...
I did try various combinations (APIC off, on, no compiled in, went back to 2.4.26, 2.4.25-gentoo-r2, 2.6.4, 2.6.6-rc2(?)), to no avail. It used to work fine when I was running RH8 with (I think) 2.4.24 (I now run gentoo).

Quoted 9 lines

Any suggestions would be more than appreciated.
TIA,
Fred

same unresolved problem; the only way to shut down my Abit motherboard in this case is to use AltGR-Sysrq-O, which first repeats for the second time "acpi_power_off called" and then finally shuts down the pc.

Is there a way to make poweroff *really* power-off ?

Thanks,
Nico

Emma Jane Hogbin

author

2004-05-26 04:32 UTC #5 lore raw

In reply to a message not in the archive

Quoted 4 lines · Fred Labrosse

Quoted 3 lines · Emma Jane Hogbin
I have two big irritants for the 2.6.5/2.6.6 (patched for ACPI) kernels with ACPI.

Quoted 8 lines

Quoted 4 lines
1. My computer no longer powers off. acpi_power_off is called, but the computer doesn't actually turn itself off. In fact when I use "poweroff" directly I get a message flashing across my screen saying:
syntax error on 658 of stdin
I did try various combinations (APIC off, on, no compiled in, went back to 2.4.26, 2.4.25-gentoo-r2, 2.6.4, 2.6.6-rc2(?)), to no avail. It used to work fine when I was running RH8 with (I think) 2.4.24 (I now run gentoo).

I turned CONFIG_X86_GOOD_APIC=y to =n and I still can't shut down properly. I'm not sure why it did work in 2.4.22 but doesn't now. :(

Signature · 9 lines

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


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click

A message that is not in the archive (sent off-list or not yet received): [email protected]

Sérgio Monteiro Basto

2004-05-27 01:40 UTC #7 lore raw

Hi,

Quoted 10 lines · Emma Jane Hogbin

Quoted 9 lines · Fred Labrosse
Quoted 7 lines · Emma Jane Hogbin
I have two big irritants for the 2.6.5/2.6.6 (patched for ACPI) kernels with ACPI.
1. My computer no longer powers off. acpi_power_off is called, but the computer doesn't actually turn itself off. In fact when I use "poweroff" directly I get a message flashing across my screen saying:
syntax error on 658 of stdin

Quoted 6 lines

Quoted 3 lines
I did try various combinations (APIC off, on, no compiled in, went back to 2.4.26, 2.4.25-gentoo-r2, 2.6.4, 2.6.6-rc2(?)), to no avail. It used to work fine when I was running RH8 with (I think) 2.4.24 (I now run gentoo).
I turned CONFIG_X86_GOOD_APIC=y to =n and I still can't shut down properly. I'm not sure why it did work in 2.4.22 but doesn't now. :(

1- Have you configure Power Management support on yes (CONFIG_PM=y) ? 2- you use poweroff command or shutdown -h now ? poweroff can have better results (I think).
3- no longer have to worry about compile w/ or w/o APIC because you should have same results if you use "nolapic" on boot options, or not use.

best regards,

Signature · 9 lines

Sérgio M. B.



-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g.
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id149&alloc_id66&op=click

A message that is not in the archive (sent off-list or not yet received): 1085622000.3503.10.camel-4/[email protected]

Emma Jane Hogbin

author

2004-05-27 01:57 UTC #9 lore raw

Quoted 1 line · Sérgio Monteiro Basto

1- Have you configure Power Management support on yes (CONFIG_PM=y) ?

Yes.

Quoted 2 lines

2- you use poweroff command or shutdown -h now ? poweroff can have better results (I think).

I've tried both.

Quoted 3 lines

3- no longer have to worry about compile w/ or w/o APIC because you should have same results if you use "nolapic" on boot options, or not use.

I'm not sure if it's because it's been cooler out, or because I recompiled without APIC however...I'm no longer getting the random freezes that I was before. Either way I won't be putting APIC back in, just in case. :)

emma

Signature · 9 lines

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


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click

Fred Labrosse

2004-05-27 08:12 UTC #10 lore raw

In reply to a message not in the archive

Quoted 1 line · Sérgio Monteiro

1- Have you configure Power Management support on yes (CONFIG_PM=3Dy) ?

Yes, otherwise you can't even get to acpi, can you (in menuconfig I mean)

Quoted 2 lines

2- you use poweroff command or shutdown -h now ? poweroff can have better results (I think).

Have tried both.

Quoted 3 lines

3- no longer have to worry about compile w/ or w/o APIC because you should have same results if you use "nolapic" on boot options, or not use.

Have tried all the possible combinations for having and not having apic (at least the ones mentioned in the kernel options file.

Thanks anyway.

Fred

A message that is not in the archive (sent off-list or not yet received): 16565.41690.262681.183380-4CLdheC3Iye1Qrn1Bg8BZw@public.gmane.org

SérgioM. Basto

2004-05-27 12:33 UTC #12 lore raw

Quoted 5 lines · Fred Labrosse

Quoted 1 line · Sérgio Monteiro
1- Have you configure Power Management support on yes (CONFIG_PM=3Dy) ?
Yes, otherwise you can't even get to acpi, can you (in menuconfig I mean)

yes, you can get acpi w/o Power Management support, at least in kernel 2.4.26 and I kernel 2.6.6.

Sergio M. B.

[email protected]

2004-05-25 09:17 UTC #13 lore raw

In reply to a message not in the archive

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

Salut Emma,
I also use a TM634. You probably have compiled your kernel with APIC support
CONFIG_X86_UP_APIC
As the TM63x does not have an APIC, it is best, to make a kernel without that. Then it will power-off. I use a stock 2.6.6-kernel (as the ACPI-patches go into mainline so quickly...)

Your syntax error seams to be something, you normally do not see, but this sounds like a bash-thing to me...

As I do not use gaim and xmms, I do not encounter freezes...

Quoted 8 lines · Emma Jane Hogbin

I have two big irritants for the 2.6.5/2.6.6 (patched for ACPI) kernels with ACPI.
1. My computer no longer powers off. acpi_power_off is called, but the computer doesn't actually turn itself off. In fact when I use "poweroff" directly I get a message flashing across my screen saying:
syntax error on 658 of stdin

Quoted 8 lines

2. Random freezes. It's not just the touch pad or the keyboard, but rather the entire monitor seems to freeze from time to time and for no apparent reason. Gaim and xmms are usually (always?) open when it happens, but I can't believe that either of them would cause my system to freeze.
I have an AcerTM 634 that did not have either of these two problems with 2.4.22. As requested on acpi.sf.net I have also made available:

cheers
hartwig felger

Hartwig Felger informatics

- -- 1024D/339FD693 Hartwig Felger <[email protected]> Key fingerprint = FB2F 3EE9 345A D55B 6FF2 0EC1 F5B0 684F 339F D693 For the pulic keys, please visit my page.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)

iD8DBQFAsw9K9bBoTzOf1pMRAgYWAJ45mmieDXlUviVMC7PPnMdKWYPB8wCgiYhQ G6RXHgeFCf5bvX0a0V/wbjs=
=T1e0
-----END PGP SIGNATURE-----

A message that is not in the archive (sent off-list or not yet received): Pine.LNX.4.60.0405251112580.1337-KnfdeQs3A3X/[email protected]

Federico Di Gregorio

2004-05-25 17:31 UTC #15 lore raw

Lì martedì, 2004/05/25 alle 11:17, +0200, [email protected] ha scritto:

Quoted 10 lines

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Salut Emma,
I also use a TM634. You probably have compiled your kernel with APIC support
CONFIG_X86_UP_APIC
As the TM63x does not have an APIC, it is best, to make a kernel without that. Then it will power-off. I use a stock 2.6.6-kernel (as the ACPI-patches go into mainline so quickly...)

same problem here with sony vaio tr2: no power off unless apic is removed. sorry for the dumb question but if the apic is not there why does acpi say something like "apic found: re-enabling bios"?

federico

Signature · 5 lines

Federico Di Gregorio                         http://people.initd.org/fog
Debian GNU/Linux Developer                                [email protected]
INIT.D Developer                                           fog-NGVKUo/i/[email protected]
  We are all dust, Saqi, so play the lute
                    We are all wind, Saqi, so bring wine. -- Omar Khayam

A message that is not in the archive (sent off-list or not yet received): [email protected]

Len Brown

2004-05-27 06:19 UTC #17 lore raw

Quoted 3 lines · Federico Di Gregorio

same problem here with sony vaio tr2: no power off unless apic is removed. sorry for the dumb question but if the apic is not there why does acpi say something like "apic found: re-enabling bios"?

it isn't ACPI that does this, it is the base OS (and I think it is a bad idea). You'll see this message even with acpi=off.

"nolapic" will make it go away.

-Len

Len Brown

2004-05-27 06:18 UTC #18 lore raw

In reply to a message not in the archive

Quoted 10 lines · On Tue, 2004-05-25 at 05:17, [email protected]…

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Salut Emma,
I also use a TM634. You probably have compiled your kernel with APIC support
CONFIG_X86_UP_APIC
As the TM63x does not have an APIC, it is best, to make a kernel without that. Then it will power-off. I use a stock 2.6.6-kernel (as the ACPI-patches go into mainline so quickly...)

I think every processor since Pentium 1 has a Local APIC built in -- at least all the processors from Intel.

If you build in support for it, you can disable it with "nolapic".

Note that LAPIC is required to support IOAPIC -- which is a motherboard device in the chipset. This on can be disabled with "noapic".

-Len

A message that is not in the archive (sent off-list or not yet received): [email protected]

[email protected]

2004-05-27 15:22 UTC #20 lore raw

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

Salut Len & Emma,

Quoted 2 lines · Len Brown

Quoted 1 line · On Tue, 2004-05-25 at 05:17, [email protected]…

...

Quoted 6 lines

Quoted 3 lines
As the TM63x does not have an APIC, it is best, to make a kernel without that. Then it will power-off. I use a stock 2.6.6-kernel (as the ACPI-patches go into mainline so quickly...)
I think every processor since Pentium 1 has a Local APIC built in -- at least all the processors from Intel.

Ok, I was not precise about that. You are right, that I wanted to say, that the TM63x does not have an IOAPIC... Anyway, I know, what the IOAPIC is good for, but what do I need an LAPIC in an single CPU-system? May somebody elaborate on that?

My problem with Emma is, that I have the same TM634 and no problem with vanilla 2.6.6 kernel to power-off. There was something around 2.6.4 IIRC, but it was worked-around by not compiling in APIC-support. But I do not know, if this issue still persists.

Emma, does your TM634 has the latest BIOS from Acer, i.e. R01-C0O?

Cheers
hartwig felger

Hartwig Felger informatics

iD8DBQFAtgey9bBoTzOf1pMRAj9yAJ9bizWujezfU12AliUfOUD6tJtrQACfcQcC +Do5XPclH346voZ5zdBedK8=
=yWKz
-----END PGP SIGNATURE-----

A message that is not in the archive (sent off-list or not yet received): Pine.LNX.4.60.0405271714330.4143-KnfdeQs3A3X/[email protected]

Emma Jane Hogbin

author

2004-05-27 16:04 UTC #22 lore raw

Quoted 4 lines · On Thu, May 27, 2004 at 05:22:22PM +0200…

My problem with Emma is, that I have the same TM634 and no problem with vanilla 2.6.6 kernel to power-off. There was something around 2.6.4 IIRC, but it was worked-around by not compiling in APIC-support. But I do not know, if this issue still persists.

I was thinking some more about this...I read somewhere that the 2.4 kernel uses /proc events and the 2.6 kernel uses /sys events and that if you're using old scripts with a new kernel "halt" won't work because it's trying to turn off /proc events. Does that make sense? How would I know what my system is doing? (Debian unstable.)

Quoted 1 line

Emma, does your TM634 has the latest BIOS from Acer, i.e. R01-C0O?

I'm not sure how I'd check. :/

thanks for all of your ideas about my problem! I appreciate it. :)

emma

Signature · 9 lines

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


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click

A message that is not in the archive (sent off-list or not yet received): [email protected]

[email protected]

2004-05-27 17:07 UTC #24 lore raw

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

Salut Emma & Fred,

Quoted 1 line · Emma Jane Hogbin

I was thinking some more about this...I read somewhere that the 2.4 kernel uses /proc events and the 2.6 kernel uses /sys events and that if you're using old scripts with a new kernel "halt" won't work because it's trying to turn off /proc events. Does that make sense? How would I know what my system is doing? (Debian unstable.)

I would check the software as a last resort...

Quoted 2 lines

Quoted 1 line
Emma, does your TM634 has the latest BIOS from Acer, i.e. R01-C0O?
I'm not sure how I'd check. :/

On startup press F2 to enter the CMOS-Setup. There you will be on the page Information, where you can find on the line, starting with
SystemBIOS version: V4.R6.1-C0O
So no need to hurry Fred ;-))))

Cheers
hartwig felger

Hartwig Felger informatics

iD8DBQFAtiB29bBoTzOf1pMRAm1BAKCzhn1cif/dB0NxOMacD+6g1FqcQACgk4Y6 pC6bvpwg9ZQSe+cvzny9CiA=
=AlCZ
-----END PGP SIGNATURE-----

A message that is not in the archive (sent off-list or not yet received): Pine.LNX.4.60.0405271903070.4655-KnfdeQs3A3X/[email protected]

Fred Labrosse

2004-05-31 09:39 UTC #26 lore raw

Quoted 10 lines · [email protected] writes:

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Salut Emma & Fred,
Quoted 5 lines · Emma Jane Hogbin
I was thinking some more about this...I read somewhere that the 2.4 kernel uses /proc events and the 2.6 kernel uses /sys events and that if you're using old scripts with a new kernel "halt" won't work because it's trying to turn off /proc events. Does that make sense? How would I know what my system is doing? (Debian unstable.)

Quoted 9 lines

I would check the software as a last resort...
Quoted 2 lines
Quoted 1 line
Emma, does your TM634 has the latest BIOS from Acer, i.e. R01-C0O?
I'm not sure how I'd check. :/
On startup press F2 to enter the CMOS-Setup. There you will be on the page Information, where you can find on the line, starting with
SystemBIOS version: V4.R6.1-C0O
So no need to hurry Fred ;-))))

Here's my version: V4R6.1 R01-C0O (same as everybody else apparently).

I installed a vanilla 2.6.6 with Hartwig's config file (after a "make oldconfig", no other modification), and it doesn't power off.

So it's not the kernel. What can it be?

For reference, I run gentoo.

TIA,

Fred

Fred Labrosse

2004-05-27 16:13 UTC #27 lore raw

In reply to a message not in the archive

Quoted 5 lines · [email protected] writes:

My problem with Emma is, that I have the same TM634 and no problem with vanilla 2.6.6 kernel to power-off. There was something around 2.6.4 IIRC, but it was worked-around by not compiling in APIC-support. But I do not know, if this issue still persists.

I have tried 2.6.4 without APIC support in, and that did not solve the problem. I will try 2.6.6 asap and will report. It could be useful if you could send your .config for comparison with mine and Emma's as we both have the same problem with very similar machines (Emma's 634 and my 636).

Quoted 3 lines

Emma, does your TM634 has the latest BIOS from Acer, i.e. R01-C0O?

I will look at that. However, I've never been able to see the version as it is too fast. Is there any way to see it using some software or stopping the startup very early?

Cheers,

Fred

A message that is not in the archive (sent off-list or not yet received): [email protected]

Emma Jane Hogbin

author

2004-05-27 17:07 UTC #29 lore raw

Quoted 3 lines · Fred Labrosse

I will look at that. However, I've never been able to see the version as it is too fast. Is there any way to see it using some software or stopping the startup very early?

F2 or F12 at the green screen at the very beginning to get into the BIOS.

System BIOS: V4R6.1 R01-C0N (Phoenix)
VGA BIOS: 3.11.1.53.1

Signature · 9 lines

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


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click