Bug 1372

Summary:	sleep/wakeup fails with ehci and ALSA i810 on Dell Inspiron 600m
Product:	Drivers	Reporter:	Pat Suwalski (pat)
Component:	USB	Assignee:	Greg Kroah-Hartman (greg)
Status:	CLOSED PATCH_ALREADY_AVAILABLE
Severity:	high	CC:	acpi-bugzilla, dopey
Priority:	P2
Hardware:	i386
OS:	Linux
Kernel Version:	2.6.0-2.6.7	Subsystem:
Regression:	---	Bisected commit-id:
Attachments:	acpidmp Output dmesg -s 40000 Output dmidecode Output (cool!) lspci -vv Output /proc/interrupts

Description Pat Suwalski 2003-10-18 15:34:35 UTC

Distribution:
Gentoo

Hardware Environment:
Dell Latitude 600m (centrino).

Software Environment:
Linux... removed PCMCIA and AGP. Left ALSA, USB and a few others. No X.

Problem Description:
The laptop successfully goes to sleep, but then becomes comatose. This is so for
modes "standby", "mem", and "disk". I assume that disk depends on mem which
depends on sleep.

Steps to reproduce:
echo -n "standby" > /sys/power/state

Comment 1 Pat Suwalski 2003-10-18 15:44:34 UTC

Created attachment 1093 [details]
acpidmp Output

Comment 2 Pat Suwalski 2003-10-18 15:45:04 UTC

Created attachment 1094 [details]
dmesg -s 40000 Output

Comment 3 Pat Suwalski 2003-10-18 15:45:30 UTC

Created attachment 1095 [details]
dmidecode Output (cool!)

Comment 4 Pat Suwalski 2003-10-18 15:45:49 UTC

Created attachment 1096 [details]
lspci -vv Output

Comment 5 Pat Suwalski 2003-10-18 15:46:44 UTC

Created attachment 1097 [details]
/proc/interrupts

I might also say that I've read through and tried all the useful steps from but
1169.

Comment 6 Pat Suwalski 2003-10-18 18:25:15 UTC

As another step to testing whether this is ACPI or some other device preventing
resuming, I built the most minimalistic kernel I could. All there is is IDE,
devfs, and pty support.

With this minimal kernel, I can "standby" properly, with resume.

The other two still do not work. Any ideas?

Comment 7 Pat Suwalski 2003-11-02 21:20:17 UTC

I took a few hours this weekend and tracked down the problems.

I can, in fact, do a standby with a fullblown system, as long as ehci (USB2),
and the ALSA i810 modules are not loaded.

Suspend to mem works too. However, as noted above, no screen upon return. The
system is still responsive, but headless.

Is this actually a bug of those specific modules, or is this your jurisdiction?
What can I do to help?

Comment 8 Len Brown 2003-11-03 09:34:17 UTC

good hunting Pat.  Yes, this one may require ACPI, but may not be an ACPI failure;-)

Comment 9 Pat Suwalski 2003-11-14 06:25:38 UTC

Actually, I'm more concerned with the sound + ACPI than EHCI + ACPI. Assuming
there is a Category/Component for that.

Comment 10 Pat Suwalski 2004-07-23 06:10:29 UTC

Did not work with 2.6.7 vanilla, but works with 2.6.7+acpi20040715.

ehci now refuses to go into a sleep state, and the i810 module seems to handle
it well.

Removing the ehci module allows for successful sleep and resume (minus video,
but that is another issue).