Bug 12681

Summary:	s2ram: fails to wake up on Acer Extensa 4220 (SMP disabled)
Product:	ACPI	Reporter:	Orivej Desh (smpuj)
Component:	EC	Assignee:	Alexey Starikovskiy (astarikovskiy)
Status:	CLOSED UNREPRODUCIBLE
Severity:	high	CC:	acpi-bugzilla, astarikovskiy, lenb, rjw, rui.zhang, smpuj
Priority:	P1
Hardware:	All
OS:	Linux
Kernel Version:	2.6.28-rc4-00012-g1cfe62c	Subsystem:
Regression:	Yes	Bisected commit-id:
Bug Depends on:
Bug Blocks:	7216, 12398
Attachments:	acpidump merge irq and poll modes merge irq and poll modes #5

Description Orivej Desh 2009-02-09 13:01:56 UTC

Latest working kernel version: 2.6.28-rc4-00011-g3ad4f59
Earliest failing kernel version: 2.6.28-rc4-00012-g1cfe62c
Distribution: ArchLinux (testing repository enabled)
Hardware Environment: Acer Extensa 4220 (Intel Celeron 550; Intel 965GM)

1. SMP disabled

Prior to 2.6.28-rc4-00012-g1cfe62c 's2ram -f' was working just fine. Since 1cfe62c8010ac56e1bd3827e30386a87cc2f3594 when trying to resume the laptop lightens up for a moment, turns off for ~10 seconds, turns on for ~10 seconds with the screen completely blank and HDD LID blinking, quickly turns off and on, showing BIOS POST screen.
Reverting this 1cfe62c... allowed me to restore s2ram functionality on both 2.6.28.3 and the latest git (it was 45c82b5...).

Probably important note is that when I just had applied the patch (git diff -R 1cfe62c~1 1cfe62c >acpi-s2ram.patch) to my already built 2.6.28.3 and rebuild the kernel, it didn't work. I had to 'make clean' the source tree in order for the patch to take effect.

2. SMP enabled

With SMP enabled s2ram ceased to work long ago. Somewhere between May 28 (
a939098~1 was the last compilable and working version) and Jul 11 (d9fc3fd fixed the last issue causing kernel panics). After reapplying d9fc3fd, 1a98fd1 and 55f2623 I managed to narrow down the interval to May 28-Jun 21 (f580366), but this is still very large. Anyway, this is another bug.

Comment 1 ykzhao 2009-02-09 17:03:55 UTC

Hi, Orivej
    Will you please attach the output of acpidump?
    Thanks.

Comment 2 Orivej Desh 2009-02-09 23:43:21 UTC

Created attachment 20173 [details]
acpidump

Comment 3 Orivej Desh 2009-02-09 23:49:34 UTC

Bug 12642 seems to be the same as this: rmmod'ing psmouse allows my laptop to wakeup (tested on today's git only with SMP disabled).

Comment 4 Rafael J. Wysocki 2009-02-10 05:42:58 UTC

There appears to be a conflict between the EC and the psmouse driver.  Alex?

Comment 5 Rafael J. Wysocki 2009-02-10 05:44:06 UTC

Caused by:

commit 1cfe62c8010ac56e1bd3827e30386a87cc2f3594
Author: Alexey Starikovskiy <astarikovskiy@suse.de>
Date:   Tue Oct 28 00:35:30 2008 +0300

    ACPI: EC: revert msleep patch

    Signed-off-by: Alexey Starikovskiy <astarikovskiy@suse.de>
    Signed-off-by: Len Brown <len.brown@intel.com>

First-Bad-Commit : 1cfe62c8010ac56e1bd3827e30386a87cc2f3594

Comment 6 Zhang Rui 2009-02-12 01:48:32 UTC

re-assign to Alexey.

Comment 7 Rafael J. Wysocki 2009-03-14 13:15:25 UTC

Notify-Also : Linux ACPI <linux-acpi@vger.kernel.org>

Comment 8 Alexey Starikovskiy 2009-04-06 21:56:55 UTC

Created attachment 20838 [details]
merge irq and poll modes

Please check if this patch helps.

Comment 9 Orivej Desh 2009-04-10 09:20:29 UTC

Alexey, I've tested your patch against kernel 2.6.29.1 with gentoo-sources patches (ArchLinux's kernel26-ice 2.6.29-3) only, and your patch doesn't help: laptop fails to wakeup as described in the first post. Rmmod'ing psmouse makes it wakeup just fine.

Comment 10 Rafael J. Wysocki 2009-04-26 11:04:44 UTC

Handled-By : Alexey Starikovskiy <astarikovskiy@suse.de>
Notify-Also : Zhang Rui <rui.zhang@intel.com>

Comment 11 Zhang Rui 2009-06-17 07:13:34 UTC

hi, Alexey,
what's the status of this bug?

Comment 12 Alexey Starikovskiy 2009-06-17 08:09:05 UTC

Created attachment 21960 [details]
merge irq and poll modes #5

please check if this patch helps.

Comment 13 Orivej Desh 2009-07-02 01:34:01 UTC

Unfortunately, it doesn't help.

Comment 14 Len Brown 2009-08-13 03:17:09 UTC

closed due to inactivity for 1 month.
please re-open if this is still a problem in 2.6.30.stable