Matthew Garrett of Red Hat today is blogging about the different ways to reset an x86 system and the limitations of these five different methods available on Linux. These methods include using kbd (rebooting via the keyboard controller), triple (generating a triple fault), PCI, EFI, and ACPI.
Matthew recently began exploring how Microsoft Windows handles the reboot process. "A while back I did some tests with Windows running on top of qemu. This is a great way to evaluate OS behaviour, because you've got complete control of what's handed to the OS and what the OS tries to do to the hardware. And what I discovered was a little surprising. In the absence of an ACPI reboot vector, Windows will hit the keyboard controller, wait a while, hit it again and then give up. If an ACPI reboot vector is present, windows will poke it, try the keyboard controller, poke the ACPI vector again and try the keyboard controller one more time."
Knowing this ended up being very useful. "This turns out to be important. The first thing it means is that it generates two writes to the ACPI reboot vector. The second is that it leaves a gap between them while it's fiddling with the keyboard controller. And, shockingly, it turns out that on most systems the ACPI reboot vector points at 0xcf9 in system IO space. Even though most implementations nominally require two different values be written, it seems that this isn't a strict requirement and the ACPI method works."
Now knowing this, Matthew adapted this reboot method to the Linux kernel and it's been integrated into the 3.0 kernel. Not only is it in this next kernel release, but it will be the default behavior. This blog post by Matthew has all of the details and he notes that this should fix the Linux reboot support for various Apple computers and ThinkPads, in particular.
The Linux 3.0 kernel may also solve the Linux kernel power regressions.