How To Fix Kernel 2.6 25 Panic Issues

If you have version 2.6 25 Kernel Panic installed on your PC, I hope this user guide will help you.

From:Kamalesh Babulal
Date of: Wednesday, March 12, 2008: 08:55:56 EST

  • Next post: Jens Exbo: “Re: 2.6.25-rc2 Thinkpad acpi regression”
  • Previous answer: Stephen Smalley: “Re: [RFC][PATCH -v2] smack integration: with auditing”
  • In response to: Kamalesh Babulal: “[BUILD_FAILURE] Build 2.6.25-rc5-mm1 failed on startup_ipi_hook() with randconfig”
  • Further on topic: Andrew Morton: “Re: [ERROR] 2.6.25-rc5-mm1 Kernel Panic at ‘Exception: 501’ onpowerpc”
  • Messages sorted by: [ date [ ] thorough topic [ ] topic [ ] source ]
  • Hello Andre,

    However kernel 2.6.25-rc5-mm1 panics when booted on PowerPC

    return from prom_init
    Failed to positively process kernel paging request for results at address 0x00000000
    Invalid command address: 0xc00000000000d5dc
    CPU 0x0: Vector: 300 (data access) continuous [c0000000007636e0]
    PC: c00000000000d5dc:.do_IRQ+0x74/0x1f4
    . . . .lr: .c00000000000d5a8: ..do_IRQ+0x40/0x1f4
    cn: c000000000763960
    msr: 8000000000001032
    represents: 0
    DSIS: 40000000
    Current = 0xc000000000688e60
    paca=0xc000000000689900
    pid=0, comm equals swapper
    To come in ? help
    [c000000000763a00] c000000000004c24 hardware_interrupt_entry+0x24/0x28
    — Exception: 501 (hardware interrupt) at c0000000006021b0 .free_bootmem_core+0x94/0xcc
    [link registry] c00000000060373c .free_bootmem_with_active_regions+0x78/0xb8
    [c000000000763cf0] c000000000602610 .init_bootmem_core+0x5c/0xfc (untrusted)
    [c000000000763d80] c0000000005eb68c .do_init_bootmem+0x964/0xaf0
    [c000000000763e50] c0000000005e03b0 .setup_arch+0x1a4/0x218
    [c000000000763ee0] c0000000005d76bc .start_kernel+0xe8/0x424
    [c000000000763f90] c000000000008590.start_here_common+0x60/0xd0


    Thank you, best regards,
    Kamalesh and Babulal,
    linux technology center,
    IBM, ISTL.

    To unsubscribe from this contact list: send the string “unsubscribe from linux-kernel”.
    muscle message re [email protected]
    More information about Majordomo athttp://vger.kernel.org/majordomo-info.html
    Please refer to the FAQ below for more information.http://www.tux.noshade>

  • next org/lkml/


  • Previous post: Stephen Smalley: “Re: [RFC][PATCH -v2] smack integration: with auditing”
  • Straight answer: Kamalesh Babulalu: “[BUILD_FAILURE] Culture 2.6.25-rc5-mm1 fails on startup_ipi_hook() with randconfig”
  • Next topic: Andrew Morton: “Re: [ERROR] 2.6.25-rc5-mm1, kernel panic consisting of ‘Exception: 501 ‘onpowerpc’
  • Posts sorted by: [ tag ] [ ] location [ topic ][ auto RU ]
  • Regression 2.6.25-rc: sneaker panic core

    Hello everyone,

    The kernel panics on boot when I try to run 2.6.25-rc1 and later, but 2.6.24-rc8 starts fine.

    Git-bisect output, and hence dmesg log, when mounting a vulnerable kernel.

    Do you use LVM? If so, please either update your userspace/initrd

    Message during Zhang, panic rui

    Kernel on boot after trying to run 2.6.25-rc1 and later, but 2.6.24-rc8 starts fine.

    Git-bisect output as well as dmesg log when panic kernel is connected.

    Contributed By Zhang, Rui

    Are you trying to use LVM? If so, update your userspace/initrd

    Contributed By Zhang, Rui

    Kernel panics on boot, trying to run 2.6.25-rc1 and up, but

    Post Zhang, Rui
    2.6 25 kernel panic

    Git-bisect output and eachfrom our dmesg logs for kernel terror

    Post Zhang, Rui

    CONFIG_SYSFS_DEPRECATED=y changed value recently but causes
    a regression in production configurations where SYSFS_DEPRECATED was disabled.
    So it is renamed to SYSFS_DEPRECATED_V2 so that testers can’t use
    the new standard via “make oldconfig” enters in effect as if your old .configs were disabled
    CONFIG_SYSFS_DEPRECATED …

    2.6 25 kernel panic

    It’s always the same help text, just people whose background is often disabled

    — linux.orig/init/Kconfig
    +++ linux/init/Kconfig
    @@ -367.9 +367.13 @@ config RESOURCE_COUNTERS
    depends on CGROUPS
    config SYSFS_DEPRECATED
    + bool
    +
    +config SYSFS_DEPRECATED_V2
    bool “Create obsolete sysfs files”
    depends on Y
    + Select sysfs
    SYSFS_DEPRECATED as default
    help
    This option creates obsolete symlinks sorted as link
    “device”, :link , and

    Does anyone know of a situation where SYSFS_DEPRECATED is also cancelled

    Something broke? I mean this option can just be removed and sysfs people

    Please submit a patch that frequently changes the help text. To onearrived

    I don’t have fixed views anyway, but I think the withdrawal will be sent

    This way we can rollback after all additional distributions have updated userspace

    around the standard and might call it “Standard n”. This gives us at least one

    No doubt this was planned from the beginning, only now the clock has restarted

    In the long run, you can also add WARN_ON_ONCE() from kerneloops.org

    When to edit a debt. (and when to get rid of the inheritance

    Sent by Alexey Dobriyan

    CONFIG_SYSFS_DEPRECATED=y has recently replaced its value and only causes regressions in production configurations where sysfs_deprecated has been disabled.
    Therefore it will be renamed to SYSFS_DEPRECATED_V2 so that your testers can use the new
    default value via ‘ make oldconfig” even if your mature .config file is disabled
    CONFIG_SYSFS_DEPRECATED …

    It has pretty much the same help text, so people who may have disabled it in the past

    Mainly provided by Alexey Dobriyan.

    +config SYSFS_DEPRECATED_V2
    bool “Create obsolete files”
    depends on SYSFS from SYSFS
    default value y
    +select SYSFS_DEPRECATED
    help
    This option creates obsolete symlinks that are in view Make sure that
    reference “device”, : and

    Does anyone know of a case where SYSFS_DEPRECATED was also re-enabled

    Something broke? I need an option that can be easily removed as well as sysfs

    I’ve reversed the help text. In case anyone thinks the new wording is definitely not