Memory corruption detected with linux 3.9.2-1

Question

I have this message with dmesg and kernel 3.9.2-1, it disappears with kernel downgrade (3.8.11), what can i do ??
p.s.
the system seems to be stable.
[   61.584649] Corrupted low memory at ffff880000006598 (6598 phys) = ff0000000000
[   61.584695] ------------[ cut here ]------------
[   61.584709] WARNING: at arch/x86/kernel/check.c:140 check_for_bios_corruption+0x10f/0x120()
[   61.584712] Hardware name: HP Pavilion dv6 Notebook PC
[   61.584714] Memory corruption detected in low memory
[   61.584719] Modules linked in: fuse uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_core joydev videodev media hp_wmi sparse_keymap iTCO_wdt iTCO_vendor_support arc4 snd_hda_codec_hdmi intel_powerclamp coretemp microcode psmouse serio_raw pcspkr i2c_i801 intel_ips acpi_cpufreq mperf ath9k ath9k_common ath9k_hw ath mac80211 cfg80211 r8168(O) rfkill thermal wmi snd_hda_codec_idt battery snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_page_alloc hp_accel snd_timer lis3lv02d input_polldev snd ac evdev mei soundcore lpc_ich processor radeon ttm nfs lockd sunrpc fscache ext4 crc16 mbcache jbd2 sr_mod cdrom sd_mod ahci libahci ehci_pci ehci_hcd libata scsi_mod usbcore usb_common i915 video button i2c_algo_bit intel_agp intel_gtt drm_kms_helper drm i2c_core
[   61.584832] Pid: 23, comm: kworker/0:1 Tainted: G           O 3.9.2-1-ARCH #1
[   61.584835] Call Trace:
[   61.584845]  [<ffffffff81058090>] warn_slowpath_common+0x70/0xa0
[   61.584851]  [<ffffffff8105810c>] warn_slowpath_fmt+0x4c/0x50
[   61.584858]  [<ffffffff8104543f>] check_for_bios_corruption+0x10f/0x120
[   61.584865]  [<ffffffff8104545e>] check_corruption+0xe/0x40
[   61.584873]  [<ffffffff81075e30>] process_one_work+0x170/0x440
[   61.584880]  [<ffffffff81076905>] worker_thread+0x115/0x3d0
[   61.584886]  [<ffffffff810767f0>] ? manage_workers+0x340/0x340
[   61.584892]  [<ffffffff8107b5d0>] kthread+0xc0/0xd0
[   61.584898]  [<ffffffff8107b510>] ? kthread_create_on_node+0x120/0x120
[   61.584905]  [<ffffffff814d96ac>] ret_from_fork+0x7c/0xb0
[   61.584910]  [<ffffffff8107b510>] ? kthread_create_on_node+0x120/0x120
[   61.584914] ---[ end trace 1563130596fe98e7 ]---

Answer 1

Nekos wrote:
I have this message with dmesg and kernel 3.9.2-1, it disappears with kernel downgrade (3.8.11), what can i do ??
p.s.
the system seems to be stable.
Well, the pithy answer would be to buy a computer from a vendor who supports free operating systems.
Of course, I own and love my DV4.
What does it mean?  Some BIOS implementations use areas of low memory without properly allocating it (letting the operating system know it is in use).  Windows just does not use that area of memory, and some vendors do a sloppy job in their software quality; it only has to work with Redmond code, so why fix it?
Linux is forced to implement a bunch of ridiculous hacks to work around low quality hardware and firmware.  You found one.  The kernel was watching that memory to see if it could be trusted.  Something changed that memory when nothing (in the kernel's view) should have touched it.  Since it changed, the kernel has taken note that that memory should not be trusted.  Kind of analogous to finding a bad block on a disk drive and subsequently ignoring it by adding it to a list of bad blocks.
So, aside from sucking it up because you lost some usable memory, you need do nothing.
EDIT: BTW, it probably means the new kernel is more stable.
Last edited by ewaller (2013-05-16 15:03:00)