suspend/resume with "Gigabyte GA-EG45M-DS2H" board causes dmesg WARNING: at /build/buildd/linux-2.6.28/arch/x86/kernel/setup.c:719 check_for_bios_corruption+0xdd/0xe0()

[Martin Olsson]

fresh boot, did "suspend" right away and after resume apport reported this kernel error

ProblemType: KernelOops
Annotation: Your system might become unstable now and might need to be restarted.
Architecture: amd64
DistroRelease: Ubuntu 9.04
Failure: oops
MachineType: Gigabyte Technology Co., Ltd. EG45M-DS2H
Package: linux-image-2.6.28-11-generic 2.6.28-11.38
ProcCmdLine: root=UUID=0ceb2ba0-df2e-4da2-a28c-3be483363631 ro quiet splash crashkernel=384M-2G:64M@16M,2G-:128M@16M
ProcVersionSignature: Ubuntu 2.6.28-11.38-generic
SourcePackage: linux
Title: WARNING: at /build/buildd/linux-2.6.28/arch/x86/kernel/setup.c:719 check_for_bios_corruption+0xdd/0xe0()

[Paul Dufresne (paulduf)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. Please answer these questions:
Is this reproducible?
How does your BIOS identify itself (mark and version)?

I also note that you have the following kernel parameter:
crashkernel=384M-2G:64M@16M,2G-:128M@16M
You did add this yourself? Why was this needed?

According to:
http://lkml.indiana.edu/hypermail/linux/kernel/0808.3/1869.html
this check was added to check for BIOS corruption, but I am unsure how much abnormal it is on a resume.

Just a note about an other bug with similar warning:
https://bugs.launchpad.net/bugs/350704

[Martin Olsson (mnemo)]

I did not add that kernel param myself (maybe it was added when I installed the kerneloops package? just a guess)
Also a question: How can I find out how my BIOS identifies itself? What command to run? Where to look?

I'm just about to go to vacation but I will look into this when I get back (in 2 weeks maybe). I'm pretty sure it was reproducible yes, will confirm later though. Bye.

[Paul Dufresne (paulduf)]

>How can I find out how my BIOS identifies itself?
When you boot (or reboot), often the first lines that appears on the screen is the BIOS 'identifying itself'.
English is not my native language, it might not be the more appropriate expression.

In your case, I now know it is a BIOS made by AWARD.
And often, there is a version shown... that was part of what I was asking.
As it is shown for very short time, pressing 'pause' key while booting may be helpfull.

For the developer reading this bug, the general info of this motherboard is at:
http://www.gigabyte.com.tw/Products/Motherboard/Products_Overview.aspx?ProductID=2877

The specs links says about BIOS:
   1. 2 x 8 Mbit flash
   2. Use of licensed AWARD BIOS
   3. Support for DualBIOS™
   4. PnP 1.0a, DMI 2.0, SM BIOS 2.4, ACPI 1.0b

I note the following: "Revolution energy saving design with DES advanced technology featuring hardware based Dynamic 4-Gear switching."

[Paul Dufresne (paulduf)]

Now that the BIOS type is mostly answered... and that I don't see what else to ask...
I am confirming this bug.

[Martin Olsson (mnemo)]

Hi Paul... I see you figured out most the BIOS info already. Indeed it's an AWARD Modular 6.00PG BIOS and the exact version tag printed at the bottom of the boot screen says "07/18/2009-G45-ICH10-7A89PG03C-00

This motherboard was introduced into the market in aug 2008 and I ordered it right away so that seems like pretty likely values. I've not updated (flashed) the BIOS on this machine so far (not sure if there is a newer version available).

I have also installed 2.6.30-rc2 from the ubuntu mainline kernel builds and unfortunately that kernel fails to resume at all. Basically the LEDs light up again and I can hear the fans starting etc but I can't SSH into the machine and the display never shows any graphics even after waiting 2-3 minutes for the resume to finish. The machine did respond to SysRq-B though so the kernel was definitely alive post-resume but someway along the resume path it got stuck or something.

[Paul Dufresne (paulduf)]

Hi Martin.
I now note that in your CurrentDmesg.txt file you have:
[ 239.816016] Corrupted low memory at ffff88000000d348 (d348 phys) = 400000000000
just before the oops.

Don't really knows what it is... I searched for "d348 address BIOS" in Google, and saw a Linux boot sector with this address in it:
http://hanzubon.jp/mirrors/kernel.org/scm/linux/kernel/bkcvs/linux-2.4/arch/i386/boot/bootsect.S,v
I am probably just pointing to a totally unrelated stuff.

It could be good to know if this is always this same address, and value (400000000000).

Does the system continue to work fine after the oops?

[Thomas Novin (thomasn80)]

I found a workaround for this here, https://bbs.archlinux.org/viewtopic.php?pid=649022

Edit /etc/default/grub and replace

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"

with

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash memory_corruption_check=0"

This disables this memory check which goes wrong. I did a check of my memory (booted memtest86 from grub) and found nothing wrong so I don't think this check is very much needed.

After changing that file run 'sudo update-grub'

[Brad Figg (brad-figg)]

This bug was filed against a series that is no longer supported and so is being marked as Won't Fix. If this issue still exists in a supported series, please file a new bug.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.