Description of Problem: Kernel oops with 2.4.9-31smp. Hardware details sketchy at this point. Have requested additional information. Unable to handle kernel NULL pointer dereference at virtual address 00000004 c01583a3 *pde = 00000000 Oops: 0002 CPU: 1 EIP: 0010:[prune_dqcache+115/192] Not tainted EIP: 0010:[<c01583a3>] Not tainted Using defaults from ksymoops -t elf32-i386 -a i386 EFLAGS: 00010206 eax: 00000000 ebx: c4c5fe20 ecx: c4c5fe28 edx: 00000000 esi: 00000011 edi: 000000c0 ebp: 0008e000 esp: eff8bf8c ds: 0018 es: 0018 ss: 0018 Process kswapd (pid: 7, stackpage=eff8b000) Stack: c771c000 c20ca470 eff8a000 0008e000 0000024d 000000c0 c0158405 00000052 c013605b 00000006 000000c0 c20ca470 000000c0 c20ca37c 000000c0 000000c0 00000000 eff8a000 00000006 c01360d5 000000c0 00000000 00010f00 efffbfb8 Call Trace: [shrink_dqcache_memory+21/48] [<c0158405>] shrink_dqcache_memory [kernel] 0x15 [<c013605b>] do_try_to_free_pages [kernel] 0x2b [<c0105866>] kernel_thread [kernel] 0x26 [kswapd+0/240] kswapd [kernel] 0x0 [<c0136080>] kswapd [kernel] 0x0 Code: 89 50 04 89 02 c7 41 04 00 00 00 00 c7 43 08 00 00 00 00 8b >>EIP; c01583a3 <prune_dqcache+73/c0> <===== Trace; c0158405 <shrink_dqcache_memory+15/30> Trace; c013605b <do_try_to_free_pages+2b/50> Trace; c0105866 <kernel_thread+26/30> Trace; c0136080 <kswapd+0/f0> Code; c01583a3 <prune_dqcache+73/c0> 00000000 <_EIP>: Code; c01583a3 <prune_dqcache+73/c0> <===== 0: 89 50 04 mov %edx,0x4(%eax) <===== Code; c01583a6 <prune_dqcache+76/c0> 3: 89 02 mov %eax,(%edx) Code; c01583a8 <prune_dqcache+78/c0> 5: c7 41 04 00 00 00 00 movl $0x0,0x4(%ecx) Code; c01583af <prune_dqcache+7f/c0> c: c7 43 08 00 00 00 00 movl $0x0,0x8(%ebx) Code; c01583b6 <prune_dqcache+86/c0> 13: 8b 00 mov (%eax),%eax
Created attachment 56047 [details] lspci, lsmod, oops, sysrq
Created attachment 57244 [details] Just another case
Created attachment 61125 [details] Just another case #2
sbrohs: what kind of kernel modules are you using? Please stop using insmod -f for one! That is asking for problems!
We are using Compaq Management Health and Wellness Driver cpqrid-2.2.0-8 and cpqhealth-2.3.0-20 (since the machine is a ProLiant DL3802G). In fact these modules (cpqhealth and cpqrid) are by default inserted with -f. We have now disabled them in order to find out if they confuse the kernel. to be continued...
Thanks for the bug report. However, Red Hat no longer maintains this version of the product. Please upgrade to the latest version and open a new bug if the problem persists. The Fedora Legacy project (http://fedoralegacy.org/) maintains some older releases, and if you believe this bug is interesting to them, please report the problem in the bug tracker at: http://bugzilla.fedora.us/