I am using Ubuntu 7.04 on an AMD64 machine. I encounter a serious kernel panic problem frequently. All of a sudden, the following message appears, and my computer stops completely:

[115890.814068] invalid opcode: 0000 [1] SMP
[115890.814061] ------------[ cut here ]------------

And then I have to cold restart the computer.


The following is the syslog after I restart the computer:

Jun 19 17:54:19 scg11 kernel: [115890.814061] ------------[ cut here ]------------
Jun 19 17:54:19 scg11 kernel: [115890.814066] kernel BUG at fs/buffer.c:2648!
Jun 19 17:54:19 scg11 kernel: [115890.814068] invalid opcode: 0000 [1] SMP
Jun 19 17:54:19 scg11 kernel: [115890.814070] CPU 1
Jun 19 17:54:19 scg11 kernel: [115890.814072] Modules linked in: xt_multiport iptable_filter ip_tables x_tables binfmt_misc rfcomm l2cap bluetooth ppdev cpufreq_userspace cpufreq_ondemand cpufreq_stats cpufreq_conservative freq_table cpufreq_powersave sony_acpi dev_acpi pcc_acpi tc1100_wmi sbs dock asus_acpi battery container backlight i2c_ec ac video button ipv6 af_packet lp fuse snd_hda_intel snd_hda_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_dummy snd_seq_oss snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq snd_timer snd_seq_device atl1 snd serio_raw nvidia(P) mii psmouse k8temp soundcore parport_pc parport i2c_core shpchp pci_hotplug snd_page_alloc pcspkr tsdev evdev ext3 jbd mbcache sg ide_cd cdrom sd_mod ata_generic floppy sata_nv ehci_hcd libata scsi_mod ohci_hcd amd74xx generic usbcore thermal processor fan fbcon tileblit font bitblit softcursor vesafb cfbcopyarea cfbimgblt cfbfillrect capability commoncap
Jun 19 17:54:19 scg11 kernel: [115890.814109] Pid: 180, comm: pdflush Tainted: P 2.6.20-16-generic #2
Jun 19 17:54:19 scg11 kernel: [115890.814112] RIP: 0010:[submit_bh+15/272] [submit_bh+15/272] submit_bh+0xf/0x110
Jun 19 17:54:19 scg11 kernel: [115890.814118] RSP: 0018:ffff81013b645bb0 EFLAGS: 00010246
Jun 19 17:54:19 scg11 kernel: [115890.814121] RAX: 0000000008ed0bc0 RBX: ffff81002ec5c000 RCX: ffff810097385d70
Jun 19 17:54:19 scg11 kernel: [115890.814124] RDX: 0000000000000000 RSI: ffff81002ec5c000 RDI: 0000000000000001
Jun 19 17:54:19 scg11 kernel: [115890.814126] RBP: 0000000000000001 R08: 0000000000000000 R09: ffff81013b645b00
Jun 19 17:54:19 scg11 kernel: [115890.814128] R10: ffff81013d6aaf40 R11: 0000000000000000 R12: ffff810108ed0bc0
Jun 19 17:54:19 scg11 kernel: [115890.814131] R13: 0000000000001073 R14: ffff81013d6aaf40 R15: ffff81007aed7a38
Jun 19 17:54:19 scg11 kernel: [115890.814134] FS: 00002b8ec3081df0(0000) GS:ffff81013b852f40(0000) knlGS:0000000000000000
Jun 19 17:54:19 scg11 kernel: [115890.814136] CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b
Jun 19 17:54:19 scg11 kernel: [115890.814138] CR2: 00000000006e38c8 CR3: 000000007c342000 CR4: 00000000000006e0
Jun 19 17:54:19 scg11 kernel: [115890.814141] Process pdflush (pid: 180, threadinfo ffff81013b644000, task ffff81013a1c4100)
Jun 19 17:54:19 scg11 kernel: [115890.814142] Stack: ffff810108ed0bc0 0000000000000001 ffff810108ed0bc0 ffffffff8021c859
Jun 19 17:54:19 scg11 kernel: [115890.814147] ffff81013b645e60 ffffffff88136420 000010006e999900 ffff81013d6aaf40
Jun 19 17:54:19 scg11 kernel: [115890.814151] ffff81006e999900 000000006e999900 ffff810108ed0bc0 ffff81013b645e60
Jun 19 17:54:19 scg11 kernel: [115890.814155] Call Trace:
Jun 19 17:54:19 scg11 kernel: [115890.814161] [__block_write_full_page+457/720] __block_write_full_page+0x1c9/0x2d0
Jun 19 17:54:19 scg11 kernel: [115890.814175] [_end+128976016/2130485360] :ext3:ext3_get_block+0x0/0x120
Jun 19 17:54:19 scg11 kernel: [115890.814186] [_end+128982629/2130485360] :ext3:ext3_ordered_writepage+0xe5/0x190
Jun 19 17:54:19 scg11 kernel: [115890.814193] [generic_writepages+442/800] generic_writepages+0x1ba/0x320
Jun 19 17:54:19 scg11 kernel: [115890.814202] [_end+128982400/2130485360] :ext3:ext3_ordered_writepage+0x0/0x190
Jun 19 17:54:19 scg11 kernel: [115890.814220] [do_writepages+39/64] do_writepages+0x27/0x40
Jun 19 17:54:19 scg11 kernel: [115890.814224] [__writeback_single_inode+470/912] __writeback_single_inode+0x1d6/0x390
Jun 19 17:54:19 scg11 kernel: [115890.814236] [sync_sb_inodes+478/704] sync_sb_inodes+0x1de/0x2c0
Jun 19 17:54:19 scg11 kernel: [115890.814242] [keventd_create_kthread+0/144] keventd_create_kthread+0x0/0x90
Jun 19 17:54:19 scg11 kernel: [115890.814247] [writeback_inodes+152/240] writeback_inodes+0x98/0xf0
Jun 19 17:54:19 scg11 kernel: [115890.814252] [pdflush+0/512] pdflush+0x0/0x200
Jun 19 17:54:19 scg11 kernel: [115890.814254] [wb_kupdate+166/288] wb_kupdate+0xa6/0x120
Jun 19 17:54:19 scg11 kernel: [115890.814264] [pdflush+340/512] pdflush+0x154/0x200
Jun 19 17:54:19 scg11 kernel: [115890.814267] [wb_kupdate+0/288] wb_kupdate+0x0/0x120
Jun 19 17:54:19 scg11 kernel: [115890.814273] [kthread+217/288] kthread+0xd9/0x120
Jun 19 17:54:19 scg11 kernel: [115890.814280] [child_rip+10/18] child_rip+0xa/0x12
Jun 19 17:54:19 scg11 kernel: [115890.814284] [keventd_create_kthread+0/144] keventd_create_kthread+0x0/0x90
Jun 19 17:54:19 scg11 kernel: [115890.814289] [physflat_send_IPI_mask+0/128] physflat_send_IPI_mask+0x0/0x80
Jun 19 17:54:19 scg11 kernel: [115890.814295] [kthread+0/288] kthread+0x0/0x120
Jun 19 17:54:19 scg11 kernel: [115890.814299] [child_rip+0/18] child_rip+0x0/0x12
Jun 19 17:54:19 scg11 kernel: [115890.814302]
Jun 19 17:54:19 scg11 kernel: [115890.814303]
Jun 19 17:54:19 scg11 kernel: [115890.814303] Code: 0f 0b eb fe 8b 06 a8 20 75 09 0f 0b 0f 1f 44 00 00 eb fe 48
Jun 19 17:54:19 scg11 kernel: [115890.814310] RIP [submit_bh+15/272] submit_bh+0xf/0x110
Jun 19 17:54:19 scg11 kernel: [115890.814313] RSP <ffff81013b645bb0>

Can anyone help?

Are you using an HT processor or dual core?
Does this happen always?

Install kernel without smp, amd64 unless dual core
is not supposed to support smp.

Does HT refer to Hyper transport?

I'm using AMD Athlon 64 x2 5600+ CPU. Yes, this is dual core.

This does not ALWAYS happen, but SEEMS to happen whenever I run long jobs. For example, when I used rsync or scp to get 14G files from another machine, the problem happen. But it did not happen immediately after the start of rsync or scp, but about a few hours after I start the rsync or scp.

How can I install the kernel without smp?

It looks like it's doing a paging-operation of some kind, and it may be that the problem only surfaces when the system is running out of memory... is under extreme stress.

To start with, I would do a clean recompile of the kernel, just to make sure, and make sure that you have properly defined [a sufficient amount of] the swap-space. Also, run smart tests to be sure that the drives have not logged any I/O errors.

But that's a bit strange, because I have 4 GB of memory. It should be enough in most purpose.

Although I'm not a newbie in using Linux, I had very little knowledge about Linux kernel, and how to recompile the kernel. What should I do in order to do that? Besides, will my installed programs and libraries or other data be erased in recompiling the kernel?

And what is the smart test? I type smart in the term and got the command not found result.

I am not familiar with ubuntu, but most distros
give you something called kernel-sources, or vanilla-sources..

install it and then goto /usr/src/linux/ or the corresponding
directory in your distro.

type make menuconfig

set all values properly, if you want delete existing config
by deleting .config file.

make && make modules_install && make install

(you should've mounted your /boot, LILO or GRUB should exist)

Lookout for specially patched kernels for amd64. For
gentoo, it comes automatically when gentoo-sources is emerged.
There should be something for ubuntu also.