ubuntu won't power-off

win32sux · 07-31-2006, 02:53 AM

i've installed dapper drake on my AMD athlon box, and it powers-off fine with the i386 kernel... the thing is that when i switch to the i686 or K7 kernels, it loses the ability to power-off properly... it does the whole shutdown sequence and then stays at "Will now halt".

so i'm wondering if you could tell me what i need to do in order to get the powering-off to work with the i686 kernel just like it does with the i386 one...

thanks in advance...

BTW, my board uses APM, not ACPI...

-=Graz=- · 07-31-2006, 03:09 AM

try do a:
$lsmod |grep apm
in the i386 kernel... if you get some output then try do the same thing with the i686 kernel. Maybe 686 does not have APM running.
I have not used APM before - only ACPI and i have always compiled it as modules. I can only assume APM would be the same (although i dont know for sure).

-=Graz=- · 07-31-2006, 03:12 AM

Also you if the apm module is not loaded you can see whether it is indeed available. it may be in /sys/modules
if it is you can just $modprobe <modulename>

dunno if this helps =)

win32sux · 07-31-2006, 03:50 AM

thanks for the reply... right now all i have is the i686 kernel, but i'll install the i386 again to run tests if you want... it looks like apm is indeed running here:

Code:

win32sux@carly:~$ uname -a
Linux carly 2.6.15-26-686 #1 SMP PREEMPT Mon Jul 17 20:14:14 UTC 2006 i686 GNU/Linux
win32sux@carly:~$ lsmod | grep apm
apm                    22768  1
win32sux@carly:~$ ps aux | grep apm | grep -v grep
root      4258  0.0  0.0      0     0 ?        S    02:05   0:00 [kapmd]
root      4269  0.0  0.0   1564   452 ?        Ss   02:05   0:00 /usr/sbin/apmd -P /etc/apm/apmd_proxy --proxy-timeout 30
win32sux@carly:~$

any ideas??

win32sux · 07-31-2006, 04:32 AM

i did some googling and it seems there's TONS of people experiencing this...

for example:

http://www.ubuntuforums.org/showthread.php?t=184749

https://launchpad.net/distros/ubuntu....15/+bug/31993

https://launchpad.net/distros/ubuntu....15/+bug/43961

https://launchpad.net/distros/ubuntu....15/+bug/36158

looks like they're waiting for this to get fixed via a kernel package update...

on that last thread i linked, there's a post which mentions a workaround:

https://launchpad.net/distros/ubuntu...58/comments/17

he says that for those like me with APM, the workaround is to boot with acpi=force - i'm not sure how much sense that makes, but i guess i might as well try it, right??

win32sux · 07-31-2006, 05:16 AM

well, as expected, the acpi=force did NOT help at all...

anyone got any ideas for things i could try?? or do you think my best option is to go back to the i386 kernel and just wait for ubuntu to fix the i686 and K7 kernels in a future package update??

having my PC properly turn-off is rather important to me, probably more important than the small performance gains i get from the optimized kernels...

win32sux · 07-31-2006, 06:45 AM

well i'm back with the i386 kernel now... i guess i'll just have to wait and see if future ubuntu i686/K7 kernels fix this issue... for the record, this is what the commands above look like now with the i386 kernel (basically the same):

Code:

win32sux@carly:~$ uname -a
Linux carly 2.6.15-26-386 #1 PREEMPT Mon Jul 17 19:52:53 UTC 2006 i686 GNU/Linux
win32sux@carly:~$ lsmod | grep apm
apm                    21228  1
win32sux@carly:~$ ps aux | grep apm | grep -v grep
root      4189  0.0  0.0      0     0 ?        S    06:39   0:00 [kapmd]
root      4200  0.0  0.0   1568   456 ?        Ss   06:39   0:00 /usr/sbin/apmd -P /etc/apm/apmd_proxy --proxy-timeout 30
win32sux@carly:~$

anyways, thanks for the help...

-=Graz=- · 08-01-2006, 05:45 AM

hey win32, yes those outputs are interesting indeed - basically the same...

I think you did the right thing to revert back to the i386 kernel and wait for the team to fix the bug

If other people are also having the same problem then there is prob not that much we can do untill they update the kernel.

Just on a side note - if i do a lsmod i get the below (many more modules) .. using ACPI of coarse.

pcc_acpi 12416 0
dev_acpi 11140 0
acpi_sbs 19980 0
battery 9988 1 acpi_sbs
ac 5252 1 acpi_sbs
i2c_acpi_ec 5120 1 acpi_sbs
i2c_core 21904 1 i2c_acpi_ec