Bugzilla Bug 139852

With kernel 2.6.17-gentoo-r2, suspend/resume does not work on my HP/Compaq
nc6120 notebook.


Steps to reproduce:
1. Boot kernel 2.6.17-gentoo-r2 (config see attachment)
2. Suspend the system with "echo -n disk >/sys/power/state"
3. Boot again and try to resume

Actual results:
The following messages appear on the console...

ACPI: write EC, IB not empty
ACPI Exception (evregion-0409): AE_TIME, Returned by Handler for
[EmbeddedControl] [20060127]
ACPI Error (psparse-0517): Method parse/execution failed
[\_SB_.C002.C003.C004.C153] (Node c778e440), AE_TIME
ACPI Error (psparse-0517): Method parse/execution failed [\_TZ_.C23E] (Node
c7793f80), AE_TIME
ACPI Error (psparse-0517): Method parse/execution failed [\_TZ_.C245._ON_]
(Node c7793e00), AE_TIME
Synaptics Touchpad, model: 1, fw: 6.2, id: 0x25a0b1, caps: 0xa04793/0x300000
serio: Synaptics pass-through port at isa0060/serio4/input0
input: SynPS/2 Synaptics TouchPad as /class/input/input0

... and then the system hangs and must be switched off by pressing the power
switch for 5 seconds.

Expected results:
System should resume properly.


Additional observations:

1. Suspend/resume just works fine with vanilla-sources-2.6.17.4 (identical
kernel configuration, as far as possible)

2. If I reverse attached hunk from 4400_speakup-20060618.patch, the resulting
kernel suspends/resumes properly, too.

Created an attachment (id=91340) [edit]
Kernel .config

Created an attachment (id=91341) [edit]
Hunk of 4400_speakup-20060618.patch that causes (triggers?) the bug

Does this happen every time?

Please boot with the "initcall_debug" and "debug" parameters and examine the
hang messages again.

(In reply to comment #3)
> Does this happen every time?

Yes, it does.

> Please boot with the "initcall_debug" and "debug" parameters and examine the
> hang messages again.

Sorry, no additional hang messages. (The screen is cleared before resuming, so
I don't see what comes before. Maybe I can check with the serial console
later.)


However, I have further investigated: If I move "input/serio/" along with
"input/gameport/" and "input/" in drivers/Makefile, then the problem goes away.

See attached patch. It used to be like this in the 2.6.16 series, BTW.

Created an attachment (id=91352) [edit]
Proposed bugfix for drivers/Makefile

The problem still exists in 2.6.17-r3.

Fixed in gentoo-sources-2.6.17-r4

(In reply to comment #7)
> Fixed in gentoo-sources-2.6.17-r4

Yes, looks good.