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.
(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?
2.6.32 is obsolete for years as far as upstream is concerned
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.
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.
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.
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.