Jonathan Berry wrote:
Please be aware that this is not a "fix" but rather is a "workaround." The problem is still there, the "acpi=off" just forces a different set of code to do the turning off (apm). Also, this workaround has some pontentially undesirable consequences. For instance, my laptop uses acpi for processor frequency scaling and power management. If I turn it off, these won't work, which would be fairly bad. If you have a desktop, I'd say the chance that it needs acpi is much less. I guess you just have to weigh whether you need acpi and if you can deal with manually powering off your system.
Thanks Jonathan. I am using a desktop, and turning ACPI off does not seem to have hurt me.
The big problem was that the power button actually wouldn't even respond. I have an Emachines box, it has a power button with a blue ring around it that lights up when it's on - I would press that button after the Power Down message and nothing would happen. I couldn't CTRL-ALT-DEL either. I had to pull the power plug the first time it happened.
What I did as a workaround first was choosing to restart instead, and then I was able to power the box off once it started the reboot sequence.
So, the acpi=off workaround is a lot cleaner for me :) I'm sure it's different for some.
Still haven't found a good way to suspend, though, nothing as clean as the suspend in XP. When I do "apm -s" or "apm -S", the box beeps a few times and it sounds like the drive spins down, but everything else is still going. I can still hear the CPU fan. The suspend in XP pretty much powers off the whole box, then that blue light blinks until I hit a key or move the mouse and it wakes again.
I have used Linux for a long time but haven't had to deal with power management much, I usually use desktops and don't worry about it...
Thanks,
Rich
