Skip to contentSkip to conversation

Ghost Lid switch on Baytrail tablet

DISCUSSION Bastien Nocera 3 participants 4 messages Jul 19, 2015 – Jul 21, 2015 ACPI

lore
Share & tools
lore
#0 lore raw

Hey,

This tablet:
http://www.amazon.com/WinBook-10-1-Inch-Windows-full-size-Display/dp/B00N9ZG5U2/
doesn't have a physical lid switch unless you plug the keyboard accessory to it.

But it thinks that there is a lid, and that it's closed, making logind suspend the machine after about 30 seconds.

I'm currently testing this on a 4.0 kernel, as there have been regressions on Baytrail with 4.1 and 4.2rcs.

The DSDT for that tablet:
https://people.gnome.org/~hadess/Winbook%20TW100%20DSDT.dsl

Any ideas how to handle this sort of device?

Cheers

#1 lore raw

Hi,

The _LID returns LIDS:

        Device (LID)
        {
            Name (_HID, EisaId ("PNP0C0D") /* Lid Device */)  // _HID: Hardware ID
            Name (LIDS, Zero)
            Method (_LID, 0, NotSerialized)  // _LID: Lid Status
            {
                Return (LIDS) /* \_SB_.LID_.LIDS */
            }
        }
All LIDS related code can be found:
	Line 132:         LIDS,   8, 
	Line 5400:                         PARM |= (LIDS << 0x10)
	Line 5566:                         PARM = (LIDS << 0x08)
	Line 11792:             ^^LID.LIDS = Zero
*	Line 11802:             ^^LID.LIDS = One
	Line 11906:             Name (LIDS, Zero)
	Line 11909:                 Return (LIDS) /* \_SB_.LID_.LIDS */
The value is only set by Q14 (marked by * above):
        Method (Q14, 0, NotSerialized)
        {
            P8XH (One, 0xBB)
            P8XH (Zero, 0x14)
            ^^LID.LIDS = One
            Notify (LID, 0x80) // Status Change
            P8XH (One, 0xBB)
            P8XH (Zero, 0xA4)
        }

If QR_EC(0x14) cannot arrive to the OSPMs as early as possible, then the default LIDS=0 value may cause problem. We need EC debug log to track if QR_EC(0x14) can arrive earlier.

You can uncomment the following line from drivers/acpi/ec.c and rebuild the kernel:
/* #define DEBUG */

Send the dmesg output (before suspending) to us.

Thanks and best regards
-Lv

Quoted 25 lines

-----Original Message-----
From: Bastien Nocera [mailto:[email protected]]
Sent: Sunday, July 19, 2015 8:43 AM
To: Rafael J. Wysocki; Len Brown; [email protected]; [email protected]
Cc: [email protected]
Subject: Ghost Lid switch on Baytrail tablet

Hey,

This tablet:
http://www.amazon.com/WinBook-10-1-Inch-Windows-full-size-Display/dp/B00N9ZG5U2/
doesn't have a physical lid switch unless you plug the keyboard accessory to it.

But it thinks that there is a lid, and that it's closed, making logind suspend the machine after about 30 seconds.

I'm currently testing this on a 4.0 kernel, as there have been regressions on Baytrail with 4.1 and 4.2rcs.

The DSDT for that tablet:
https://people.gnome.org/~hadess/Winbook%20TW100%20DSDT.dsl

Any ideas how to handle this sort of device?

Cheers

#2 lore raw
In reply to Bastien Nocera

Hi,

If the issue is already identified and confirmed:
For example, you've confirmed the QR_EC(0x14) can only arrive when the keyboard accessory is plugged in. Then please ignore my previous reply.
I don't have idea about handling a LID that always returns "closed" when the keyboard accessory is not plugged in.

Thanks
-Lv

Quoted 69 lines

From: Zheng, Lv
Sent: Tuesday, July 21, 2015 8:23 AM

Hi,

The _LID returns LIDS:

        Device (LID)
        {
            Name (_HID, EisaId ("PNP0C0D") /* Lid Device */)  // _HID: Hardware ID
            Name (LIDS, Zero)
            Method (_LID, 0, NotSerialized)  // _LID: Lid Status
            {
                Return (LIDS) /* \_SB_.LID_.LIDS */
            }
        }
All LIDS related code can be found:
	Line 132:         LIDS,   8,
	Line 5400:                         PARM |= (LIDS << 0x10)
	Line 5566:                         PARM = (LIDS << 0x08)
	Line 11792:             ^^LID.LIDS = Zero
*	Line 11802:             ^^LID.LIDS = One
	Line 11906:             Name (LIDS, Zero)
	Line 11909:                 Return (LIDS) /* \_SB_.LID_.LIDS */
The value is only set by Q14 (marked by * above):
        Method (Q14, 0, NotSerialized)
        {
            P8XH (One, 0xBB)
            P8XH (Zero, 0x14)
            ^^LID.LIDS = One
            Notify (LID, 0x80) // Status Change
            P8XH (One, 0xBB)
            P8XH (Zero, 0xA4)
        }

If QR_EC(0x14) cannot arrive to the OSPMs as early as possible, then the default LIDS=0 value may cause problem. We need EC debug log to track if QR_EC(0x14) can arrive earlier.

You can uncomment the following line from drivers/acpi/ec.c and rebuild the kernel:
/* #define DEBUG */

Send the dmesg output (before suspending) to us.

Thanks and best regards
-Lv

Quoted 25 lines

-----Original Message-----
From: Bastien Nocera [mailto:[email protected]]
Sent: Sunday, July 19, 2015 8:43 AM
To: Rafael J. Wysocki; Len Brown; [email protected]; [email protected]
Cc: [email protected]
Subject: Ghost Lid switch on Baytrail tablet

Hey,

This tablet:
http://www.amazon.com/WinBook-10-1-Inch-Windows-full-size-Display/dp/B00N9ZG5U2/
doesn't have a physical lid switch unless you plug the keyboard accessory to it.

But it thinks that there is a lid, and that it's closed, making logind suspend the machine after about 30 seconds.

I'm currently testing this on a 4.0 kernel, as there have been regressions on Baytrail with 4.1 and 4.2rcs.

The DSDT for that tablet:
https://people.gnome.org/~hadess/Winbook%20TW100%20DSDT.dsl

Any ideas how to handle this sort of device?

Cheers

#3 lore raw
Quoted 6 lines · Lv

Hi,

If the issue is already identified and confirmed:
For example, you've confirmed the QR_EC(0x14) can only arrive when the keyboard accessory is plugged in. Then please ignore my previous reply.
I don't have idea about handling a LID that always returns "closed" when the keyboard accessory is not plugged in.

Perhaps the LID status should be ignored when the device is in tablet mode?

In the meantime, I took a look at two tablets here: ASUS T100 and Onda V975W, both are baytrail-T tablet while the T100 has a keyboard dock. For T100, its _LID return the value of GPO2.LIDZ, which has an initial value of 1: "Name(LIDZ, One)" and on LID events, LIDZ will be assigned proper value; and for Onda V975W, it doesn't define the LID acpi device. So these two devices don't have this problem.

-Aaron

Quoted 67 lines

Thanks
-Lv

Quoted 63 lines

From: Zheng, Lv
Sent: Tuesday, July 21, 2015 8:23 AM

Hi,

The _LID returns LIDS:

        Device (LID)
        {
            Name (_HID, EisaId ("PNP0C0D") /* Lid Device */)  // _HID: Hardware ID
            Name (LIDS, Zero)
            Method (_LID, 0, NotSerialized)  // _LID: Lid Status
            {
                Return (LIDS) /* \_SB_.LID_.LIDS */
            }
        }
All LIDS related code can be found:
	Line 132:         LIDS,   8,
	Line 5400:                         PARM |= (LIDS << 0x10)
	Line 5566:                         PARM = (LIDS << 0x08)
	Line 11792:             ^^LID.LIDS = Zero
*	Line 11802:             ^^LID.LIDS = One
	Line 11906:             Name (LIDS, Zero)
	Line 11909:                 Return (LIDS) /* \_SB_.LID_.LIDS */
The value is only set by Q14 (marked by * above):
        Method (Q14, 0, NotSerialized)
        {
            P8XH (One, 0xBB)
            P8XH (Zero, 0x14)
            ^^LID.LIDS = One
            Notify (LID, 0x80) // Status Change
            P8XH (One, 0xBB)
            P8XH (Zero, 0xA4)
        }

If QR_EC(0x14) cannot arrive to the OSPMs as early as possible, then the default LIDS=0 value may cause problem. We need EC debug log to track if QR_EC(0x14) can arrive earlier.

You can uncomment the following line from drivers/acpi/ec.c and rebuild the kernel:
/* #define DEBUG */

Send the dmesg output (before suspending) to us.

Thanks and best regards
-Lv

Quoted 19 lines

-----Original Message-----
From: Bastien Nocera [mailto:[email protected]]
Sent: Sunday, July 19, 2015 8:43 AM
To: Rafael J. Wysocki; Len Brown; [email protected]; [email protected]
Cc: [email protected]
Subject: Ghost Lid switch on Baytrail tablet

Hey,

This tablet:
http://www.amazon.com/WinBook-10-1-Inch-Windows-full-size-Display/dp/B00N9ZG5U2/
doesn't have a physical lid switch unless you plug the keyboard accessory to it.

But it thinks that there is a lid, and that it's closed, making logind suspend the machine after about 30 seconds.

I'm currently testing this on a 4.0 kernel, as there have been regressions on Baytrail with 4.1 and 4.2rcs.

Quoted 8 lines
Quoted 6 lines
Quoted 6 lines

The DSDT for that tablet:
https://people.gnome.org/~hadess/Winbook%20TW100%20DSDT.dsl

Any ideas how to handle this sort of device?

Cheers

N�����r��y���b�X��ǧv�^�)޺{.n�+����{�i�b�{ay�ʇڙ�,j��f���h���z��w��� ���j:+v���w�j�m��������zZ+�����ݢj"��!�i

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html