Bug 355506

Summary:	system freezes on Asus M2A HDMI Board when entering X (ACPI related)
Product:	[openSUSE] openSUSE 10.3	Reporter:	Tom Dörsam <postmaster>
Component:	Kernel	Assignee:	Thomas Renninger <trenn>
Status:	RESOLVED DUPLICATE	QA Contact:	E-mail List <qa-bugs>
Severity:	Major
Priority:	P5 - None	CC:	postmaster
Version:	Final
Target Milestone:	---
Hardware:	x86-64
OS:	openSUSE 10.3
Whiteboard:
Found By:	Other	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---
Attachments:	dmeg log hwinfo log dmesg log of untainted kernel xorg log

Description Tom Dörsam 2008-01-22 23:58:39 UTC

Created attachment 191416 [details]
dmeg log

This *could* be a bug of Bug #350017 but am not shure (same kernel oops message, see dmesg log) since the board is not an nvidia/intel chipset (hwinfo attached).

Using the openSUSE 10.3 default kernel as well a recent vailla (2.6.24rc8) exposes the same problem, X freezes after a few seconds after starting (usually i can log in, click a few times, than completely freeze). 
This only happens when acpi is enabled, if passing acpi=off to the kernel the the system runs stable.
If booting into non-X runlevel, powersaved crashes exposing the kernel messages:

[...]
Stack:  ffffffff80333d62 ffff81007880e608 ffff810074ae3e78 000000005053545f
 ffff8100777fab40 ffffffff8808045b ffffffff8033f4cc 0000000000000003
 ffff810079345000 0000000000000000 ffffffff8807f6a8 0000000000100073
Call Trace:
 [<ffffffff80333d62>] acpi_get_data+0x3e/0x6e
 [<ffffffff8033f4cc>] acpi_bus_get_device+0x25/0x68
 [<ffffffff8807f6a8>] :thermal:acpi_thermal_trip_seq_show+0x12b/0x257
 [<ffffffff802a29e6>] seq_read+0x105/0x28b
 [<ffffffff8028a42d>] vfs_read+0xcb/0x153
 [<ffffffff8028a7c1>] sys_read+0x45/0x6e
 [<ffffffff80209c2e>] system_call+0x7e/0x83

(see attached full dmesg log)

However, this is non-fatal and the system runs stable as long as in console mode. Starting X leads to the mentioned freeze (tried using Xorg fb driver, radeonhd and fglrx, all the same result). I dont know if this is related to the complete free as i cannot collect debug info after freeze.
The Problem also occured before BIOS update (was Version 1201, now its 1603).

Comment 1 Tom Dörsam 2008-01-22 23:59:44 UTC

Created attachment 191417 [details]
hwinfo log

Comment 2 Greg Kroah-Hartman 2008-02-05 16:21:37 UTC

Can you duplicate this without the closed source ATI driver loaded?

Comment 3 Tom Dörsam 2008-02-07 09:28:42 UTC

Created attachment 193559 [details]
dmesg log of untainted kernel

Comment 4 Tom Dörsam 2008-02-07 09:32:16 UTC

sorry, i forgot the fglrx module gets loaded even when not entering X. this time i'Ve blacklisted the fglrx module and double checked that "lsmod | grep fglrx" does not return a line.

The dmesg looks quite the same (except for the tainted flag). However, this time i was able to freeze up the system upon boot before entering X (right after acpid got loaded). Regretably, this only happend once and i was not able to reproduce it a second time, even so i rebooted a couple of time. But still, the system ALWAYS freezes a few seconds after entering X (this time with vesa driver, see xorg log). Again, as long as i disable acpi via kernel param, the system runs stable.

Comment 5 Tom Dörsam 2008-02-07 09:33:07 UTC

Created attachment 193563 [details]
xorg log

Comment 6 Thomas Renninger 2008-02-08 11:39:13 UTC

Seems as if ASUS copy and pasted this BIOS bug into several BIOS updates...
Please contact Asus about it, this is a BIOS bug which got introduced in their latest BIOS updates.

*** This bug has been marked as a duplicate of bug 350017 ***