Our production-server is crashed, and we don't know what this mean in the logging (messages.log):

Jan 6 04:39:39 kernel: ------------[ cut here ]------------
Jan 6 04:39:39 kernel: kernel BUG at lib/radix-tree.c:575!
Jan 6 04:39:39 kernel: invalid operand: 0000 [#1]
Jan 6 04:39:39 kernel: SMP
Jan 6 04:39:39 kernel: Modules linked in: ipt_REDIRECT ipt_REJECT ipt_state iptable_mangle iptable_nat ip_conntrack iptable_filter ip_tables parport_pc lp parport autofs4 rfcomm l2cap bluetooth sunrpc md5 ipv6 dm_mod video button battery ac uhci_hcd ehci_hcd i2c_i801 i2c_core shpchp e1000 floppy ext3 jbd raid1 ata_piix libata sd_mod scsi_mod
Jan 6 04:39:39 kernel: CPU: 0
Jan 6 04:39:39 kernel: EIP: 0060:[<c01d433a>] Not tainted VLI
Jan 6 04:39:39 kernel: EFLAGS: 00010046 (2.6.11-1.1369_FC4smp)
Jan 6 04:39:39 kernel: EIP is at __lookup_tag+0x15c/0x173
Jan 6 04:39:39 kernel: eax: 00000000 ebx: 00000031 ecx: 00000000 edx: 00000031
Jan 6 04:39:39 kernel: esi: c9683e38 edi: c9683f3c ebp: 00000001 esp: f7db9d98
Jan 6 04:39:39 kernel: ds: 007b es: 007b ss: 0068
Jan 6 04:39:39 kernel: Process pdflush (pid: 178, threadinfo=f7db9000 task=f7db8a80)
Jan 6 04:39:39 kernel: Stack: f7db9d94 00000000 00000000 f7db9e68 00000000 00000000 c9683e38 00000000
Jan 6 04:39:39 kernel: 00000001 00000000 0000000e 0000003f c2cf6208 c01d43a6 0000000e f7db9de0
Jan 6 04:39:39 kernel: 00000000 f7db9e68 00000040 c2cf6208 00000000 c2cf6214 f7db9e68 c0145085
Jan 6 04:39:39 kernel: Call Trace:
Jan 6 04:39:39 kernel: [<c01d43a6>] radix_tree_gang_lookup_tag+0x55/0x71
Jan 6 04:39:39 kernel: [<c0145085>] find_get_pages_tag+0x3a/0x92
Jan 6 04:39:39 kernel: [<c014e043>] pagevec_lookup_tag+0x24/0x2c
Jan 6 04:39:39 kernel: [<c01832bd>] mpage_writepages+0x166/0x3f0
Jan 6 04:39:39 kernel: [<f88c8c2d>] ext3_ordered_writepage+0x0/0x183 [ext3]
Jan 6 04:39:39 kernel: [<c0106641>] do_IRQ+0x55/0x86
Jan 6 04:39:39 kernel: [<c0104a5a>] common_interrupt+0x1a/0x20
Jan 6 04:39:39 kernel: [<c0307375>] schedule+0x995/0xc5e
Jan 6 04:39:39 kernel: [<c01817e9>] __sync_single_inode+0x57/0x217
Jan 6 04:39:39 kernel: [<c01819d0>] __writeback_single_inode+0x27/0x105
Jan 6 04:39:39 kernel: [<c0181c62>] sync_sb_inodes+0x1b4/0x2ad
Jan 6 04:39:39 kernel: [<c014ade4>] pdflush+0x0/0x1e
Jan 6 04:39:39 kernel: [<c0181dd0>] writeback_inodes+0x75/0xf3
Jan 6 04:39:39 kernel: [<c014a5cd>] wb_kupdate+0x85/0xf0
Jan 6 04:39:39 kernel: [<c014ad18>] __pdflush+0xc8/0x194
Jan 6 04:39:39 kernel: [<c014adfe>] pdflush+0x1a/0x1e
Jan 6 04:39:39 kernel: [<c014a548>] wb_kupdate+0x0/0xf0
Jan 6 04:39:39 kernel: [<c0134759>] kthread+0x93/0x97
Jan 6 04:39:39 kernel: [<c01346c6>] kthread+0x0/0x97
Jan 6 04:39:39 kernel: [<c01023d1>] kernel_thread_helper+0x5/0xb
Jan 6 04:39:39 kernel: Code: 83 44 24 04 04 83 c6 04 89 74 24 08 8b 44 24 38 39 44 24 10 75 b3 e9 3e ff ff ff 8b 74 24 18 8b 44 96 04 85 c0 0f 85 0a ff ff ff <0f> 0b 3f 02 ec 76 32 c0 e9 fd fe ff ff 0f 0b 50 02 ec 76 32 c0

We are running Fedora 4 - 2.6.11-1.1369_FC4smp
Could someone explain whats happened?

Jan 6 04:39:39 kernel: kernel BUG at lib/radix-tree.c:575!
Jan 6 04:39:39 kernel: invalid operand: 0000 [#1]
Jan 6 04:39:39 kernel: SMP
Jan 6 04:39:39 kernel: CPU: 0

A bug in lib/radix-tree.c:575
An SMP kernel and only 1 CPU reported? Is this an Multi-proccessor rig?
That may point to a problem. A proccessor gone down?

Try http://www.Linux.org ask if young Mr Torvalds can shed some light for you?

Also try posting your log file to Redhat, they should know more.

Actually use Mandriva on my server, an older Athlon 1.6GB single proccessor rig.
Sorry I can't be more helpfull.

Shorty943.

I wonder if it could be a damaged/corrupted kernel image.

It looks like the crash occurred in pdflush which is a kernel thread responsible for pushing modified pages back to disk. Based on the call trace, I'd guess that it happened when looking for pages to writeback. This may be caused by bad memory or some other hardware problem (overheating, bad CPU cache, disk corruption, etc.) or a legitimate bug in FC's kernel. Has this happened a lot? Are you running Fedora's latest kernel update? If not you should upgrade to make sure you have the latest version which should have no known bugs. What kind of hardware are you running on?