From Bugzilla Helper: User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en-US; rv:1.0rc3) Gecko/20020523 Description of problem: 2.4.9-31smp on my old-but-trusty i586 SMP system (Dec Celebris XP 5133) dies with an OOps in IO_APIC_get_PCI_irq_vector. The full text is: Oops: 0000 Kernel 2.4.9-34smp CPU: 0 EIP: 0010:[<c0114457> Not tainted EFLAGS: 00010246 EIP is at IO_APIC_get_PCI_itq_vectpr [kernel] 0x17 eax: 00000000 ebx: c1376c00 ecx: 00000000 edx: 00000000 esi: 00000010 edi: 000186a1 ebp: c12d9fab esp: c12d9f80 ds: 0018 es: 0018 ss: 0018 Process swapper (pid: 1, stackpage=c12d9000) Stack: 000186a1 ffffffff c1376c00 00000010 000186a1 c12d9fab c02f49ad 00000000 00000001 00000000 0008e000 c12d8000 c02effc0 c0105000 0008e000 c02f4621 00000000 c02d4cac 00000000 c0249cff c01b0e76 c12d8000 c02f092b c0105288 Call Trace: [<c010500>] stext [kernel] 0x0 [<c0249cff>] .rodata.str1.1. [kernel] 0x11fa [<c01b0e76>] pci_init [kernel] 0x6 [<c0105288>] init [kernel] 0x28 ... etc ... Code: 83 3c 90 ff 75 23 52 68 00 24 23 c0 e8 e8 52 00 00 8b 44 24 Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1.Boot with the 2.4.9-34smp kernel 2.Sit back and watch Actual Results: Oops! Expected Results: Whee! (aka, actual booting) Additional info:
Can you attach the output of dmesg to this bug of the successful boot with 2.4.9-31? (eg "dmesg > filename" and then "Create new attachement on this page) That way I can see how things are supposed to be
Created attachment 60184 [details] 2.4.9-31smp boot log
Note, BTW, that the 2.4.9-34 (non-smp) also boots fine on this hardware.
I do have a "suspect" for the failure, eg we added a patch to enable some bigger systems (with more than 32 PCI slots); this is the only change between -31 and -34 that touches such early boot stuff. I'll investigate your dmesg output to try to find out why it breaks.
From the ChangeLog, I made a wild guess that it could be: "- Add I/O SAPIC IA64 patch, fixes #63952" since that was the only thing I saw that mentioned the APIC stuff, although as an IA64 patch, it didn't seem all that likely (?). Glad you have a possible area to look. If you'd like me to try out a candidate fix for the bug, I'd be happy to do that. Just point me at an RPM that's ready to go.
*sigh* I won't be able to help you track down the problem after all. The machine concerned just died. :-( I guess all the recent reboots just didn't sit well with it or something. :-( :-( :-(
As it seems, I am having this *exact* same problem with kernel-smp-2.4.9-34.i586.rpm on my Dual P100-based Dell Optiplex DGX 5100 system. Is there something I can try?
please try "noapic" as kernel option...
Thanks! That did it!
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/