Bug 10119 (panasonic-cf30f) - boot hangs at "ACPI: EC non-query interrupt received. switching to interrupt mode" - Panasonic CF-30F
Summary: boot hangs at "ACPI: EC non-query interrupt received. switching to interrupt ...
Status: CLOSED DUPLICATE of bug 9915
Alias: panasonic-cf30f
Product: ACPI
Classification: Unclassified
Component: EC (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Zhang Rui
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-02-27 03:08 UTC by Ph. Marek
Modified: 2008-10-17 14:41 UTC (History)
4 users (show)

See Also:
Kernel Version: 2.6.23, 2.6.24.3, 2.6.25rc3
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
dmesg (20.91 KB, text/plain)
2008-02-27 03:08 UTC, Ph. Marek
Details
dmidecode (11.79 KB, text/plain)
2008-02-27 03:09 UTC, Ph. Marek
Details
lspci (2.20 KB, text/plain)
2008-02-27 03:09 UTC, Ph. Marek
Details
acpidump.txt.bz2 (46.98 KB, application/x-bzip2)
2008-02-27 03:49 UTC, Ph. Marek
Details
try the custom DSDT (310.31 KB, patch)
2008-02-29 05:53 UTC, ykzhao
Details | Diff
patch for 2.6.23 - acpi-blacklist (825 bytes, text/x-diff)
2008-03-07 03:36 UTC, Ph. Marek
Details
dmesg fedora9 - plain boot (6.29 KB, text/plain)
2008-05-13 17:39 UTC, John Floyd
Details
fedora 9 - acpi=off (8.53 KB, text/plain)
2008-05-13 17:39 UTC, John Floyd
Details
vanilla kernel 2.6.25 - plain boot (5.93 KB, text/plain)
2008-05-13 17:42 UTC, John Floyd
Details

Description Ph. Marek 2008-02-27 03:08:03 UTC
Latest working kernel version: ?
Earliest failing kernel version: 2.6.25rc3
Distribution: Standard kernel from kernel.org

Hardware Environment: Panasonic Toughbook CF-30F
  Intel Core2 Duo L7500, 1.6GHz
  2GB RAM
  Intel GM965 rev. C0
  BIOS Phoenix V2.00L11
Software Environment: custom initrd, doesn't get started.
  Machine hangs on boot.


Problem Description:

The notebook doesn't boot; it hangs after these messages:
2.6.23:
  ACPI: EC: GPE = 0x1c, I/O: command/status = 0x66, data = 0x62
2.6.24.3:
  ACPI: EC non-query interrupt received. switching to interrupt mode
2.6.25rc3:
  ACPI: EC non-query interrupt received. switching to interrupt mode
  --[cut here]--
  WARNING: at arch/x86/mm/ioremap.c:137 ()
  Modules linked in:
  Pid 1, comm: swapper: [lots of addresses...]
  -- end trace ca143223eefdc828
  (looks similar to http://lkml.org/lkml/2008/2/24/265)


We've tried several boot options; acpi=ht and acpi=off works.
The others don't help:
  acpi_osi=!Linux (as suggested by kernel messages)
  pci=nommconf
  pci=conf1

I'll attach a demsg, lspci and acpidump from acpi=off.
Comment 1 Ph. Marek 2008-02-27 03:08:45 UTC
Created attachment 15021 [details]
dmesg

with acpi=off
Comment 2 Ph. Marek 2008-02-27 03:09:11 UTC
Created attachment 15022 [details]
dmidecode

with acpi=off
Comment 3 Ph. Marek 2008-02-27 03:09:28 UTC
Created attachment 15023 [details]
lspci

with acpi=off
Comment 4 Ph. Marek 2008-02-27 03:49:42 UTC
Created attachment 15024 [details]
acpidump.txt.bz2

with acpi=off
Comment 5 ykzhao 2008-02-27 17:43:19 UTC
Thanks for the info.
From the log it seems that it has the same failure with the bug9915. So it is marked as the duplicated bug of 9915.
Thanks.
Comment 6 ykzhao 2008-02-27 17:43:33 UTC

*** This bug has been marked as a duplicate of bug 9915 ***
Comment 7 Zhang Rui 2008-02-27 18:52:58 UTC
there are some conflicts in the bug report.
"version: 2.6.23, 2.6.24.3, 2.6.25rc3" means it can work in these kernel releases, but you also mentions that "Earliest failing kernel version: 2.6.25rc3" which means it can work before 2.6.25-rc3...
so did this laptop work on any kernel release?
If so, this is a regression and please tell us the latest kernel release that this laptop can work on. :)
Comment 8 Ph. Marek 2008-02-29 02:08:38 UTC
No, it doesn't work with any kernel. Just the messages differ.
Comment 9 ykzhao 2008-02-29 05:51:14 UTC
This bug is similar with the bug 9915. And it is caused by broken BIOS.
Will you please try the custom DSDT?
Comment 10 ykzhao 2008-02-29 05:53:46 UTC
Created attachment 15082 [details]
try the custom DSDT

Will you please try the custom DSDT and confirm whether the system can be booted?
How to use the custom DSDT can be found:
http://www.lesswatts.org/projects/acpi/faq.php
Thanks.
Comment 11 Ph. Marek 2008-03-06 06:55:22 UTC
Sorry for the delay.

Yes, with the custom DSDT the kernel 2.6.23 boots.

Will that be some kind of greylisting then? "Ignore/Patch the DSDT if this machine/this version"?
Comment 12 Ph. Marek 2008-03-07 03:35:16 UTC
As a short-time fix I came up with this patch.
Comment 13 Ph. Marek 2008-03-07 03:36:12 UTC
Created attachment 15172 [details]
patch for 2.6.23 - acpi-blacklist
Comment 14 Len Brown 2008-03-10 22:29:45 UTC
I'd rather not blacklist this box to disable ACPI
until we are positive that we can't implement a workaround.

are both of these changes necessary?

575,578d574
<             If (LEqual (Arg0, 0x0112))
<             {
<                 Return (\_SB.PCI0.LPCB.EC0.EC05 (Arg1, Arg2))
<             }
3300c3296
<                 OperationRegion (IGDM, SystemMemory, ASLB, 0x2000)
---
>                 OperationRegion (IGDM, SystemMemory, \ASLB, 0x2000)

Do any older kernels work, such as 2.6.21 or 2.6.22?
Comment 15 Zhang Rui 2008-03-10 23:06:44 UTC
(In reply to comment #14)
> I'd rather not blacklist this box to disable ACPI
I agree.

> until we are positive that we can't implement a workaround.
> are both of these changes necessary?
All we need is a one-line change in DSDT,
and it can not be fixed in the kernel.
I think the best way is to ping Panasonic and wait for a BIOS update.
Comment 16 Zhang Rui 2008-03-11 02:09:49 UTC

*** This bug has been marked as a duplicate of bug 9915 ***
Comment 17 Thomas Renninger 2008-04-29 07:00:22 UTC
I quickly browsed the two bugs..., what exactly is the problem?
It's the EC call?

Has Intel contacts to Panasonic?
If not at least the bug reporter could try to ask Panasonic to fix it (if someone could explain the problem detailed), AFAIK they care about Linux and may just fix this in a BIOS update.
Comment 18 Thomas Renninger 2008-04-29 07:07:41 UTC
I wonder whether one of Alexey's EC patches for 2.6.26 could fix this?
Comment 19 John Floyd 2008-05-13 17:37:38 UTC
Have a new CF30F, and have attempted a number of different kernels but hangs on all of them.  

Following this bug, I also got a copy of the patched DSDT, and included it in new kernels.  Still no luck.  Also og note, I cannot get the system to boot even with acpi=off.  This is strange, compared to the experiences described above.  Looking at the dmidecode output from above, I appear to have exactly the same machine.

I have tried kernels from OpenSuse-10.3 (2.6.22), Fedora-8 (2.6.22,4) ,9 (2.6.25), gparted-live (2.6.22?) etc.  Non boot.

Attached are dmesg outputs for fedora-9-preview, with/without acpi=off (original DSDT) and vanilla kernel 2.6.25 with patched DSDT, with/without acpi=off.  

I cannot get into the machine with linux under any circumstances.  So I dont know how to provide you with any other information.

John
Comment 20 John Floyd 2008-05-13 17:39:00 UTC
Created attachment 16130 [details]
dmesg fedora9 - plain boot
Comment 21 John Floyd 2008-05-13 17:39:46 UTC
Created attachment 16131 [details]
fedora 9 - acpi=off
Comment 22 John Floyd 2008-05-13 17:42:55 UTC
Created attachment 16132 [details]
vanilla kernel 2.6.25 - plain boot
Comment 23 John Floyd 2008-05-13 18:06:34 UTC
I am having trouble attaching the vanilla (patched dsdt) with acpi=off output.  It is basically the same, hanging at exactly the same point.  
Comment 24 John Floyd 2008-06-27 00:24:24 UTC
I have had this and other problems on the CF30 (mk2) and have just received a bios update from panasonic.  Its taken a month!  It fixes my problem - couldnt boot off a USB device (optic or mem-stick) and also the DSDT problems discussed here and in Bug 10119.  The bios version is now at V14.

Note You need to log in before you can comment on or make changes to this bug.