Bug 7813

Summary:	Slow boot if shutdown w/ psmouse loaded - HP Compaq nc6320
Product:	Drivers	Reporter:	Peter Clifton (pcjc2)
Component:	Input Devices	Assignee:	acpi_bios
Status:	CLOSED CODE_FIX
Severity:	normal	CC:	acpi-bugzilla
Priority:	P2
Hardware:	i386
OS:	Linux
Kernel Version:	Linux version 2.6.17 (Ubuntu 2.6.17-10.34-generic)	Subsystem:
Regression:	---	Bisected commit-id:
Bug Depends on:
Bug Blocks:	56331
Attachments:	dmesg output from the bad bootup dmesg output from the good bootup acpidump output from the bad bootup acpidump output from the good bootup

Description Peter Clifton 2007-01-12 08:25:56 UTC

Most recent kernel where this bug did *NOT* occur:
Unknown

Distribution:
Ubuntu

Hardware Environment:
HP Compaq nc6320

Software Environment:
Any

Problem Description:
After a shutdown or reboot in linux with psmouse loaded, the system is very slow
to boot, and when booted does not present the second SSDT table needed for CPU
frequency scaling.

Rebooting (Urgh...) Windows, executing "rmmod psmouse" before reboot or shutdown
, or unplugging the battery for a minute or so, gives a nice quick bootup and
proper CPU scaling on the next boot.

Steps to reproduce:
Shutdown linux with psmouse loaded, reboot.

Probably unrelated, but when booting the laptop on Battery, the laptop often
reports:

BUG: soft lockup detected on CPU#0

Most reliable cure is plugging in the AC. I will file a separate bug if someone
can suggest a category for it.

Comment 1 Zhang Rui 2007-01-16 00:57:28 UTC

Please attach the acpidump && dmesg output of the good state and bad state.

Comment 2 Peter Clifton 2007-01-16 04:00:15 UTC

Created attachment 10094 [details]
dmesg output from the bad bootup

Comment 3 Peter Clifton 2007-01-16 04:00:54 UTC

Created attachment 10095 [details]
dmesg output from the good bootup

Comment 4 Peter Clifton 2007-01-16 04:01:43 UTC

Created attachment 10096 [details]
acpidump output from the bad bootup

Comment 5 Peter Clifton 2007-01-16 04:02:20 UTC

Created attachment 10097 [details]
acpidump output from the good bootup

Comment 6 Peter Clifton 2007-01-16 04:05:45 UTC

Just to confirm a point in my original posting, the AC / Battery status does not
update when booted in a "bad" mode (psmouse loaded before the reboot).

Can someone suggest where to open the bug about the Softlockup issue? Seems that
ACPI is the real difference when on AC / Battery, but it is so hard to persuede
the machine to boot on Battery - I can't get a dmesg / acpidump to compare.

Comment 7 Luming Yu 2007-01-16 19:02:16 UTC

Please also test the latest base kernel, and post dmesg of that kernel.

Comment 8 Len Brown 2007-01-17 19:22:30 UTC

The failure to boot on battery issue is moved to bug 7840.
So lets focus this bug only on issues booting in A/C mode.

Does the system have any USB devices connected?
If yes, can you disconnect them to see if that helps?

Also, see if BIOS Setup has any options related to
Legacy emulation for USB devices, can you see if
disabling them helps?

Also, see if building the kernel with CONFIG_CPU_FREQ=n
makes a difference.

Comment 9 Fu Michael 2007-11-06 05:29:43 UTC

Peter, do you still have this issue?  Does the "slow boot" problem exist when using AC adapter? thanks.

Comment 10 Peter Clifton 2007-11-06 05:48:51 UTC

Appolosies for not following up, I believe this was fixed in the psmouse driver, and I no longer see the problem. I guess the bug should be closed.


commit a1cec06177386ecc320af643de11cfa77e8945bd
Author: Dmitry Torokhov <dtor@insightbb.com>
Date:   Sun Feb 18 01:40:24 2007 -0500

    Input: psmouse - properly reset mouse on shutdown/suspend
    
    Some people report that they need psmouse module unloaded
    for suspend to ram/disk to work properly. Let's make port
    cleanup behave the same way as driver unload.
    
    This fixes "bad state" roblem on various HP laptops, such
    as nx7400.
    
    Signed-off-by: Dmitry Torokhov <dtor@mail.ru>

Comment 11 Peter Clifton 2007-11-06 05:50:11 UTC

The (it turns out unrelated) issue, with:

BUG: soft lockup detected on CPU#0

Was due to a bug in the ipw3945 driver trying to access some thermal sensor on the wireless card which isn't powered up at boot if running on battery.