Bug 13975

Created attachment 22706 [details]
lspci

(switched to email.  Please respond via emailed reply-to-all, not via the
bugzilla web interface).

On Thu, 13 Aug 2009 14:02:01 GMT
bugzilla-daemon@bugzilla.kernel.org wrote:

> http://bugzilla.kernel.org/show_bug.cgi?id=13975
> 
>            Summary: BUG: unable to handle kernel paging request at
>                     000d7a0b
>            Product: Memory Management
>            Version: 2.5
>     Kernel Version: 2.6.30.4
>           Platform: All
>         OS/Version: Linux
>               Tree: Mainline
>             Status: NEW
>           Severity: normal
>           Priority: P1
>          Component: Page Allocator
>         AssignedTo: akpm@linux-foundation.org
>         ReportedBy: tonywhite100@googlemail.com
>         Regression: Yes
> 

Interesting.

> Last kernel version tried was 2.6.30.1 and I did not experience this bug
> using
> that kernel.
> 
> The crash happens randomly at boot and requires an fsck to fix. The kernel
> locks up.
> 
> Here's the interesting part of the log :
> 
> Aug 13 07:54:02 pentium-three kernel: BUG: unable to handle kernel paging
> request at 000d7a0b
> Aug 13 07:54:02 pentium-three kernel: IP: [<c01596d4>] m_show+0x94/0x190
> Aug 13 07:54:02 pentium-three kernel: *pde = 00000000 
> Aug 13 07:54:02 pentium-three kernel: Oops: 0000 [#1] PREEMPT SMP 
> Aug 13 07:54:02 pentium-three kernel: last sysfs file:
> /sys/module/nf_conntrack_ftp/initstate
> Aug 13 07:54:02 pentium-three kernel: Modules linked in: fuse af_packet
> xt_tcpudp xt_limit nf_conntrack_ipv4 nf_defrag_ipv4 xt_state ipt_LOG
> ipt_REJECT
> nf_conntrack_ftp nf_conntrack iptable_filter ip_tables x_tables dm_crypt
> rt2500usb arc4 snd_cmipci ecb gameport snd_pcm_oss snd_mixer_oss snd_pcm
> snd_page_alloc snd_opl3_lib rt73usb snd_hwdep crc_itu_t snd_mpu401_uart
> snd_seq_dummy snd_seq_oss rt2x00usb snd_seq_midi rt2x00lib snd_rawmidi
> led_class input_polldev snd_seq_midi_event snd_seq mac80211 snd_timer
> snd_seq_device cfg80211 snd evdev parport_pc rtc_cmos soundcore pcspkr
> rtc_core
> parport i2c_piix4 rtc_lib i2c_core shpchp pci_hotplug ext4 mbcache jbd2 crc16
> dm_mirror dm_region_hash dm_log dm_snapshot dm_mod usbhid hid sg sr_mod cdrom
> sd_mod pata_acpi ata_generic ata_piix uhci_hcd libata 8139too ehci_hcd floppy
> 8139cp scsi_mod usbcore mii intel_agp
> Aug 13 07:54:02 pentium-three kernel:
> Aug 13 07:54:02 pentium-three kernel: Pid: 2489, comm: lsmod Not tainted
> (2.6.30-4.slh.2-sidux-686 #1) To   be   Filled
> Aug 13 07:54:02 pentium-three kernel: EIP: 0060:[<c01596d4>] EFLAGS: 00010246
> CPU: 0
> Aug 13 07:54:02 pentium-three kernel: EIP is at m_show+0x94/0x190
> Aug 13 07:54:02 pentium-three kernel: EAX: 00000000 EBX: 000d7a0b ECX:
> de10fef4
> EDX: 00000000
> Aug 13 07:54:02 pentium-three kernel: ESI: e100fc98 EDI: df169660 EBP:
> e100fb60
> ESP: de10fee8
> Aug 13 07:54:02 pentium-three kernel: DS: 007b ES: 007b FS: 00d8 GS: 0033 SS:
> 0068
> Aug 13 07:54:02 pentium-three kernel: Process lsmod (pid: 2489, ti=de10e000
> task=ddd0c840 task.ti=de10e000)
> Aug 13 07:54:02 pentium-three kernel: Stack:
> Aug 13 07:54:02 pentium-three kernel: df169660 c041097e e11ebc4c 0000e764
> e100fb64 c102a2e0 00010246 00000000
> Aug 13 07:54:02 pentium-three kernel: df169660 e100fb64 00000035 c01d3d56
> 00000200 00000000 000003fa b7f95006
> Aug 13 07:54:02 pentium-three kernel: df1aca80 df169688 00000006 00000061
> 0000002b 00000000 0000002a 00000000
> Aug 13 07:54:02 pentium-three kernel: Call Trace:
> Aug 13 07:54:02 pentium-three kernel: [<c01d3d56>] ? seq_read+0x206/0x3d0
> Aug 13 07:54:02 pentium-three kernel: [<c01d3b50>] ? seq_read+0x0/0x3d0
> Aug 13 07:54:02 pentium-three kernel: [<c01f84e4>] ? proc_reg_read+0x64/0xa0
> Aug 13 07:54:02 pentium-three kernel: [<c01f8480>] ? proc_reg_read+0x0/0xa0
> Aug 13 07:54:02 pentium-three kernel: [<c01bb59d>] ? vfs_read+0x9d/0x160
> Aug 13 07:54:02 pentium-three kernel: [<c01bb731>] ? sys_read+0x41/0x80
> Aug 13 07:54:02 pentium-three kernel: [<c010394c>] ?
> sysenter_do_call+0x12/0x28
> Aug 13 07:54:02 pentium-three kernel: Code: c6 34 01 00 00 39 f3 74 31 8d b6
> 00
> 00 00 00 8b 43 08 89 3c 24 c7 44 24 04 7e 09 41 c0 83 c0 0c 89 44 24 08 e8 ce
> a2 07 00 8b 1b <8b> 03 0f 18 00 90 39 f3 75 da b8 01 00 00 00 8b 8d d4 00 00
> 00 
> Aug 13 07:54:02 pentium-three kernel: EIP: [<c01596d4>] m_show+0x94/0x190
> SS:ESP 0068:de10fee8
> Aug 13 07:54:02 pentium-three kernel: CR2: 00000000000d7a0b
> Aug 13 07:54:02 pentium-three kernel: ---[ end trace 1e50990a393b4447 ]---
> 
> I'll attach lspci and if there is any other information required please just
> ask and I'll post it.
> 

We died running `lsmod'.

We died in kernel/module.c:m_show().

The last sysfs file which userspace accessed was
/sys/module/nf_conntrack_ftp/initstate.

From which I surmise that nf_conntrack_ftp has somehow done something
bad to the module-related metadata which kernel/module.c presents to
userspace via /sys/module/nf_conntrack_ftp/initstate.  Or something like that.

Tell me: had you done any `rmmod's on that machine?  Perhaps of
netfilter modules?  If so, perhaps netfilter didn't properly clean up
after itself or something.

Oh.  "The crash happens randomly at boot".  That makes it hard.

The only thing I did before rebooting was uninstall sshd (Secure shell daemon process) Which was starting every boot and I never use. As such, I deemed it a good idea to remove it for better security.
No rmmods were done to my knowledge.

I can clear up the random at boot part now.
It happens everytime I cold boot the machine. After the crash, if I login to the rescue shell with the root password, issue fsck -a /dev/sda1 and then reboot, the machine boots up fine and the crash doesn't occur.

Andrew, could this message earlier on in the file :

Aug 13 07:54:02 pentium-three kernel: ip_tables: (C) 2000-2006 Netfilter Core Team
Aug 13 07:54:02 pentium-three kernel: nf_conntrack version 0.5.0 (8192 buckets, 32768 max)
Aug 13 07:54:02 pentium-three kernel: CONFIG_NF_CT_ACCT is deprecated and will be removed soon. Please use
Aug 13 07:54:02 pentium-three kernel: nf_conntrack.acct=1 kernel paramater, acct=1 nf_conntrack module option or
Aug 13 07:54:02 pentium-three kernel: sysctl net.netfilter.nf_conntrack_acct=1 to enable it.

Be a reason for this problem? And it's being caused by a depreciated iptables configuration, of which I need to change and make the debian sid packagers aware of?

Thanks,
Tony

argh, pelase don't update this bug vai the bugzilla web interface - the net guys dont' read it.

That's why I asked "(switched to email.  Please respond via emailed reply-to-all, not via the
bugzilla web interface)."

So please resend all this, via email so the netdev people get to see it, thanks.

2009/8/14  <bugzilla-daemon@bugzilla.kernel.org>:
> http://bugzilla.kernel.org/show_bug.cgi?id=13975
>
>
>
>
>
> --- Comment #4 from Andrew Morton <akpm@linux-foundation.org>  2009-08-13
> 23:24:21 ---
> argh, pelase don't update this bug vai the bugzilla web interface - the net
> guys dont' read it.
>
> That's why I asked "(switched to email.  Please respond via emailed
> reply-to-all, not via the
> bugzilla web interface)."
>
> So please resend all this, via email so the netdev people get to see it,
> thanks.
>
> --
> Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
> ------- You are receiving this mail because: -------
> You reported the bug.
>

Hi Andrew,
So send the entire report to bugzilla-daemon@bugzilla.kernel.org???
The address this mail has come from/gone to?
Have I reported the bug in the wrong place? Should it have gone here :
http://bugzilla.netfilter.org/ ?
Why am I being asked to repeat myself please? I like to keep my
bugzilla inbox spam to a minimum and that, apart from I never usually
report by email, was why I didn't reply using email.

Thanks,
Tony

Just do a reply-to-all to my initial email.  The one which said

(switched to email.  Please respond via emailed reply-to-all, not via the
bugzilla web interface).

*** This bug has been marked as a duplicate of bug 13341 ***