Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#31659 - [linux] 3.5.4 machine check exceptions when booting on i5-3317u in UEFI mode from SSD

Attached to Project: Arch Linux
Opened by Michael (c0da) - Sunday, 23 September 2012, 13:31 GMT
Last edited by Tobias Powalowski (tpowa) - Saturday, 17 November 2012, 09:02 GMT
Task Type Support Request
Category Packages: Core
Status Closed
Assigned To Tobias Powalowski (tpowa)
Thomas Bächler (brain0)
Architecture x86_64
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:

The trouble is following. I have Samsung 900x3c notebook with i5-3317u CPU. Windows 7 and Windows 8 works fine on the machine (? I don't know, for real. I am using now my second unit, the first one brought down under windows 8 control, so, I don't know, if this is really software error. But the new unit works fine under win-8, and behaves bad under Linux 3.5.4). Ok. The bad behavior show itself like that.

When I install Linux (EFIStub) in the UEFI boot mode, it boots, but in a lot of cases the boot process results in Machine Check Exception (something about incorrect processor state, see the attached picture). It looks like that some CPU state structures are corrupted in early boot. The system boots too fast and something bad happens. I'am thinking so because booting the same kernel, with the exactly same parameters and environment from USB stick results in MCE with a much less frequently. And on my previous unit with 3.4.x kernel (which booted slower from SSD) the MCE emerged less frequently.

And I've got no error, when was booting in legacy mode or in UEFI without systemd (which is aggressively parallel).

Additional info:
* package version(s)

Linux 3.5.4-1-ARCH
Systemd 189

I am using disk with GPT.
BIOS: PO4AAC (if it will help).
I've also tried to boot with GRUB, the effect is the same, stability depends only on drive speed.

* config and/or log files etc.

I cannot provide them, the system is completely unbootable. I can only provide some photos. Kernel panic, shown on picture, occurred during execution of some functions, but the functions are irrelevant, I was getting the same MCE (with the same code) during execution of other functions.

And I repeat again, that this is not clear hardware error, because it is repeated on the second device, and does not emerge on Windows 7/8.

Steps to reproduce:

Huh. Install Linux, try to boot it in UEFI mode on Samsung 900x3c-A02RU
   015.jpg (1.9 MiB)
This task depends upon

Closed by  Tobias Powalowski (tpowa)
Saturday, 17 November 2012, 09:02 GMT
Reason for closing:  Won't fix
Additional comments about closing:  See upstream report.
Comment by Michael (c0da) - Sunday, 23 September 2012, 18:26 GMT
May be this is relevant http://stackoverflow.com/questions/11254023/how-to-do-mmap-for-cacheable-pcie-bar -- the guy doing something with MTRRs. May be my problem can be fixed from kernel command line?
Comment by Michael (c0da) - Thursday, 01 November 2012, 09:21 GMT
The problem is within samsung-laptop module: https://bugzilla.kernel.org/show_bug.cgi?id=47121
Comment by Greg (dolby) - Saturday, 17 November 2012, 08:59 GMT
Upstream report is closed as wont fix.

Loading...