[arch-general] mce after linux-3.11.5-1 on NP900X3C

[Rasmus Liland]

Dear @arch-general readers,

I'm experiencing machine check exceptions since every kernel after package
linux-3.11.5-1 (Oct 14 2013). I hope some nice people will be able to assist
me or perhaps point me in a direction of something fruitful. First here goes
some kernel panic output I was able to snap on May 25th (I also made it an
attachment):
> [19367.116180] Disabling lock debugging due to kernel taint
> [19367.116196] mce: [Hardware Error]: CPU 1: Machine Check Exception: 5 Bank 4: b200000000100402
> [19367.116202] mce: [Hardware Error]: RIP !INEXACT! 33:<00007f8b4934c8b7>
> [19367.116205] mce: [Hardware Error]: TSC 2824672b8e7
> [19367.116211] mce: [Hardware Error]: PROCESSOR 0:306a9 TIME 14010118857 SOCKET 0 APIC 1 microcode 12
> [19367.116213] mce: [Hardware Error]: Run the above through 'mcelog --ascii'
> [19367.116216] mce: [Hardware Error]: Some CPUs didn't answer in synchronization
> [19367.116218] mce: [Hardware Error]: Machine check: Invalid
> [19367.116220] Kernel panic - not syncing: Fatal machine check on current CPU 
> [19368.211815] Shutting down cpus with NMI 
> [19368.222834] Kernel Offset: 0x0 from 0xffffffff81000000  0000 (relocation range: 0xffffffff80000000-0xffffffff9fffffff)
> [19368.222942] drm_kms_helper: panic occurred, switching back to text console
> [19368.245774] Rebooting in 30 seconds
> [19398.323579] ACPI RECOVERY or RESET_REG.

Assuming I did a complete system update about 5.4 hours earlier, this implies
kernel output for linux kernel version 3.14.4.

This has been the case for every kernel after version 3.11.5; from what I can 
trace in the `/var/cache/pacman/pkg' directory this also calls for _at least_
version 3.15.8; I've removed all other archived versions as they used
valuable space.

In dmesg output, my good version 3.11.5-1 is called:
> Linux version 3.11.5-1-ARCH (tobias at T-POWA-LX) (gcc version 4.8.1 20130725 (prerelease) (GCC)) #1 SMP PREEMPT Mon Oct 14 08:31:43 CEST 2013

The hardware of my system is perhaps relevant. I've got a Samsung
NP900X3C-A01SE (Jun.2012) laptop:
> Intel Core i5-3317U CPU @ 1.70GHz stepping 9 microcode 0x12
> 4GB RAM (3743M/3888M available, 2048M available to graphics)
> SanDisk SSD U100 128GB, 10.01.04, max UDMA/133
> Intel Centrino Advanced-N 6235 AGN, REV=0xB0

Presuming this is a good starting point, if more specific hardware
information is needed in this thread in the future I can add the output of
e.g. lscpi -vvnn.

--Rasmus

-- 
Rasmus Liland, jrl at jrl.dyndns.dk
-------------- next part --------------
[19367.116180] Disabling lock debugging due to kernel taint
[19367.116196] mce: [Hardware Error]: CPU 1: Machine Check Exception: 5 Bank 4: b200000000100402
[19367.116202] mce: [Hardware Error]: RIP !INEXACT! 33:<00007f8b4934c8b7>
[19367.116205] mce: [Hardware Error]: TSC 2824672b8e7
[19367.116211] mce: [Hardware Error]: PROCESSOR 0:306a9 TIME 14010118857 SOCKET 0 APIC 1 microcode 12
[19367.116213] mce: [Hardware Error]: Run the above through 'mcelog --ascii'
[19367.116216] mce: [Hardware Error]: Some CPUs didn't answer in synchronization
[19367.116218] mce: [Hardware Error]: Machine check: Invalid
[19367.116220] Kernel panic - not syncing: Fatal machine check on current CPU
[19368.211815] Shutting down cpus with NMI
[19368.222834] Kernel Offset: 0x0 from 0xffffffff81000000  0000 (relocation range: 0xffffffff80000000-0xffffffff9fffffff)
[19368.222942] drm_kms_helper: panic occurred, switching back to text console
[19368.245774] Rebooting in 30 seconds
[19398.323579] ACPI RECOVERY or RESET_REG.

-- 

Linux version 3.11.5-1-ARCH (tobias at T-POWA-LX) (gcc version 4.8.1 20130725 (prerelease) (GCC)) #1 SMP PREEMPT Mon Oct 14 08:31:43 CEST 2013

-- 

Samsung NP900X3C-A01SE (Jun.2012) laptop:
> Intel Core i5-3317U CPU @ 1.70GHz stepping 9 microcode 0x12
> 4GB RAM (3743M/3888M available, 2048M available to graphics)
> SanDisk SSD U100 128GB, 10.01.04, max UDMA/133
> Intel Centrino Advanced-N 6235 AGN, REV=0xB0
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <https://lists.archlinux.org/pipermail/arch-general/attachments/20141115/68272b9d/attachment.bin>