EPIA-M, linuxbios + power button

[Dave Ashley]

Under award bios the power button did an instant off. Under linuxbios I
have to hold the power button down for 4 seconds to get the power to go
off. I vaguely remember reading about controlling the power button, but I can't
find it in the docs. Does anyone know how to fix this?

[SONE Takeshi]

On Mon, Jul 21, 2003 at 07:35:11AM -0700, Dave Ashley wrote:
> Under award bios the power button did an instant off. Under linuxbios I
> have to hold the power button down for 4 seconds to get the power to go
> off. I vaguely remember reading about controlling the power button, but I can't
> find it in the docs. Does anyone know how to fix this?

On EPIA, I neither found the bit to switch the functionality of
power button.
I thought that "Instant Off" feature is implemented by SMM BIOS.

So I wrote a small daemon program to poll the status of power button,
and invokes "/sbin/init 0" when hit.
Polling in the user space will not miss a short-time button press
(in my application it is simulated by PIC micro)
because the status bit will be stay ON once it is pressed.

Probably the same will apply to EPIA-M.

[SONE Takeshi]

On Mon, Jul 21, 2003 at 11:57:51PM +0900, SONE Takeshi wrote:
> So I wrote a small daemon program to poll the status of power button,
> and invokes "/sbin/init 0" when hit.

Of course this starts normal shutdown sequence rather than 
Instant Off.
You will again have to deal with PM I/O space to power off.

Handling SCI interrupt in kernel space would be better solution
but userspace program is almost trivial.

EPIA-M, linuxbios + power button [PMX:#]

[ron minnich]

On Mon, 21 Jul 2003, Dave Ashley wrote:
> Under award bios the power button did an instant off. Under linuxbios I
> have to hold the power button down for 4 seconds to get the power to go
> off. I vaguely remember reading about controlling the power button, but I can't
> find it in the docs. Does anyone know how to fix this?

what mobo? Is this the 440bx one again?

you can make the power switch event cause an interrupt, I guess, and catch 
it.

Linux power handling could use help, if anybody wants to take that on.

ron