q2d08A firmware anywhere?

Richard A. Smith richard at laptop.org
Mon Jan 14 00:53:40 EST 2008


Chris Ball wrote:

>    > This is already possible.  You set the SCI mask such that
>    > non-critical SCI's are masked.  I don't think we have a generic way
>    > yet for userspace to tell the kernel what mask bits to set though.
> 
> Yep, that's most of it.  Do we have any possible way to disable keyboard
> wakeups while still allowing power button wakeups, though?  I know we
> see power button presses in the Southbridge -- can the EC see them too?

Yes.  If you set the mask to 0x00 then only the power button will wake 
the machine up.  The power button is routed through the EC.  So the EC 
does the actual sensing and debouncing of the button and then asserts an 
output to match the state of the power button.  That output is connected 
to a power management pin on the southbridge.

So the EC _could_ inhibit the power button if it chose to but right now 
that only happens when the SCI inhibit (EC cmd 0x32) has been issued. 
So it blocks all SCI's + power button for a brief period before and 
after MAIN_ON going low.

Other than that its just a pass through and doesn't issue an EC SCI. 
The kernel sees an SCI but it comes from the power management system in 
the southbridge rather than from the EC.

-- 
Richard Smith  <richard at laptop.org>
One Laptop Per Child


More information about the Devel mailing list