Bug 72941 - ACPI: Marking method _OSC as Serialized because of AE_ALREADY_EXISTS error
Summary: ACPI: Marking method _OSC as Serialized because of AE_ALREADY_EXISTS error
Status: CLOSED PATCH_ALREADY_AVAILABLE
Alias: None
Product: ACPI
Classification: Unclassified
Component: ACPICA-Core (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Lv Zheng
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-03-25 10:59 UTC by Aroon Mishra
Modified: 2014-06-23 08:39 UTC (History)
5 users (show)

See Also:
Kernel Version: since 2.6.32
Subsystem:
Regression: No
Bisected commit-id:


Attachments
output of dmesg, dmidecode, acpidump , messages (159.92 KB, application/zip)
2014-03-25 10:59 UTC, Aroon Mishra
Details

Description Aroon Mishra 2014-03-25 10:59:31 UTC
Created attachment 130651 [details]
output of dmesg, dmidecode, acpidump , messages

Hello All,

I am ruining Centos 6.2 final on IBM System x3200 M2 -[4368I1S].

BOIS : IBM BIOS Version 1.43 -[GUE143AUS-1.43]

CPU : Intel(R) Xeon(R) CPU E3110 @ 3.00GHz

I am getting below mentioned error in messages when system boot.

Mar 25 11:07:15 kernel: ACPI: bus type pci registered
Mar 25 11:07:15 kernel: PCI: MCFG configuration 0: base f0000000 segment 0 buses 0 - 12
Mar 25 11:07:15 kernel: PCI: MCFG area at f0000000 reserved in E820
Mar 25 11:07:15 kernel: PCI: Using MMCONFIG at f0000000 - f0cfffff
Mar 25 11:07:15 kernel: PCI: Using configuration type 1 for base access
Mar 25 11:07:15 kernel: bio: create slab <bio-0> at 0
Mar 25 11:07:15 kernel: ACPI Warning for \_SB_._OSC: Return type mismatch - found Integer, expected Buffer (20090903/nspredef-1018)
Mar 25 11:07:15 kernel: ACPI: Interpreter enabled
Mar 25 11:07:15 kernel: ACPI: (supports S0 S4 S5)
Mar 25 11:07:15 kernel: ACPI: Using IOAPIC for interrupt routing
Mar 25 11:07:15 kernel: ACPI: No dock devices found.
Mar 25 11:07:15 kernel: PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
Mar 25 11:07:15 kernel: ACPI Error (dsfield-0143): [CAPB] Namespace lookup failure, AE_ALREADY_EXISTS
Mar 25 11:07:15 kernel: ACPI Error (psparse-0537): Method parse/execution failed [\_SB_.PCI0._OSC] (Node ffff88011e6aa970), AE_ALREADY_EXI
STS
kernel: ACPI: Marking method _OSC as Serialized because of AE_ALREADY_EXISTS error
Mar 25 11:07:15 kernel: ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-3f])




Server is working fine but it rebooting automatically. We have checked the server
power supply, RAM and both are good. There is no any power fluctuation.

It seems the above error is causing the reboot issue.

I have updated kernel to latest one kernel /vmlinuz-2.6.32-431.5.1.el6.x86_64, previous it was kernel /vmlinuz-2.6.32-220.13.1.el6.

On new kernel same error is coming.

I have also boot the server with pci=nocrs and got same error.




Mar 25 15:00:27 kernel: ACPI: bus type pci registered
Mar 25 15:00:27 kernel: PCI: MCFG configuration 0: base f0000000 segment 0 buses 0 - 12
Mar 25 15:00:27 kernel: PCI: MCFG area at f0000000 reserved in E820
Mar 25 15:00:27 kernel: PCI: Using MMCONFIG at f0000000 - f0cfffff
Mar 25 15:00:27 kernel: PCI: Using configuration type 1 for base access
Mar 25 15:00:27 kernel: bio: create slab <bio-0> at 0
Mar 25 15:00:27 kernel: ACPI Warning for \_SB_._OSC: Return type mismatch - found Integer, expected Buffer (20090903/nspredef-1018)
Mar 25 15:00:27 kernel: ACPI: Interpreter enabled
Mar 25 15:00:27 kernel: ACPI: (supports S0 S4 S5)
Mar 25 15:00:27 kernel: ACPI: Using IOAPIC for interrupt routing
Mar 25 15:00:27 kernel: ACPI: No dock devices found.
Mar 25 15:00:27 kernel: PCI: Ignoring host bridge windows from ACPI; if necessary, use "pci=use_crs" and report a bug
Mar 25 15:00:27 kernel: ACPI Error (dsfield-0143): [CAPB] Namespace lookup failure, AE_ALREADY_EXISTS
Mar 25 15:00:27 kernel: ACPI Error (psparse-0537): Method parse/execution failed [\_SB_.PCI0._OSC] (Node ffff88011e6aa970), AE_ALREADY_EXI
STS
Mar 25 15:00:27 kernel: ACPI: Marking method _OSC as Serialized because of AE_ALREADY_EXISTS error
Mar 25 15:00:27 kernel: ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-3f])




I have searched on google and found bug 23442 listed on kernel bugzila. I have also checked the https://bugs.acpica.org/show_bug.cgi?id=946 and on bugs.acpica.org it is mentioned that there is no any issue with ACPI.

Please suggest where is the problem, in BIOS, ACPI,  in kernel or hardware.

Many many thanks in advance.
Comment 1 Aaron Lu 2014-03-27 09:05:06 UTC
(In reply to Aroon Mishra from comment #0)
> Server is working fine but it rebooting automatically. We have checked the
> server
> power supply, RAM and both are good. There is no any power fluctuation.

It gets rebooted in some un-determined time?

> 
> It seems the above error is causing the reboot issue.
> 
> I have updated kernel to latest one kernel
> /vmlinuz-2.6.32-431.5.1.el6.x86_64, previous it was kernel
> /vmlinuz-2.6.32-220.13.1.el6.
> 
> On new kernel same error is coming.

Which new kernel to be specific? Did you try any upstream kernels, like v3.14-rc8?
Comment 2 Alan 2014-04-08 10:31:29 UTC
2.6.32 is obsolete for years as far as upstream is concerned
Comment 3 Lv Zheng 2014-04-10 02:26:32 UTC
Hi,

Please try the lastest upstream kernel (https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git)
And post the dmesg/acpidump here.

Thanks in advance.
Comment 4 Robert Moore 2014-04-10 19:47:43 UTC
The patch in comment #3 implements the new "auto-serialize" feature, which should help in these ways:

1) The AE_ALREADY_EXISTS error should go away
2) The \_SB_.PCI0._OSC method should now execute properly, without aborting

So, please try the patch.
thanks.
Comment 5 Aroon Mishra 2014-04-12 04:49:57 UTC
Thanks to all for your suggestion.

Actually this issue was one my customer server , customer replaced motherboard of server and now everything is working fine.
Comment 6 Lv Zheng 2014-04-15 00:09:47 UTC
According to ACPICA BZ946 (https://bugs.acpica.org/show_bug.cgi?id=946), this seems to be a BIOS bug.
It will always be great if you can test the workaround shipped in the latest kernel and post result here.

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