https://bugzilla.kernel.org/show_bug.cgi?id=106941

--- Comment #14 from Aaron Lu <[email protected]> ---
(In reply to Lv Zheng from comment #13)
> Hi, Mika
> 
> Sorry for interrupting.
> 
> (In reply to Mika Westerberg from comment #12)
> > OK, I see. However, I'm not sure how we can fix this from the kernel side.
> > If Windows waits for the initial event and only then decides what to do, it
> > explains why it works. At least from the ACPI ASL code I'm not able to find
> > out any place where the initial value would be set to anything else than 0.
> 
> There simply won't be the initial event.
> On Samsung platforms, Linux drain events using a quirk. So there is no worry
> that the event will be lost. We can see that there is no such a event on
> some Samsung platforms.
> 
> The LID event should have already been handled by the BIOS, and OSPM was
> just invoked via the waking vector.
> We can also see,
> 1. There are many BIOS tables resetting LID state in _WAK because of this.
> 2. There are BIOS tables never initiating LID open event.

Hi Lv,

The problem here is about things after system boot, not system resume :-)

> 
> > 
> > One option is to modify systemd-sleep so that it does not read the initial
> > status (why would it suspend the device in the first place if it did not
> > receive an event about lid being closed).
> 
> If above mentioned platforms can all work with Windows, I don't know how
> this can be fixed.
> Just an idea that, userspace may not suspend system by reading the LID
> status, it probably should only wait for the "LID close" event.

I think this is a good idea, we probably need to raise this to systemd.

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

------------------------------------------------------------------------------
_______________________________________________
acpi-bugzilla mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-bugzilla

Reply via email to