Hi Ingo,

I just found this oops from last night on my home audio/midi system
(AthlonXP running -48-37).  I was burning a CD at the time, and the scsi
error precedes the oops trace by 46 seconds.  The system is still up and 
fully functional, BTW.

scsi: unknown opcode 0x01
BUG: Unable to handle kernel NULL pointer dereference at virtual address 
00000009
 printing eip:
c044d001
*pde = 00000000
Oops: 0002 [#1]
PREEMPT 
Modules linked in: snd_seq_dummy snd_seq_oss snd_seq_midi it87 eeprom i2c_sensor i2c_isa i2c_viapro i2c_dev i2c_core realtime usbhid ohci1394 ieee1394 uhci_hcd ehci_hcd usbcore snd_emu10k1_synth snd_emux_synth snd_seq_virmidi snd_seq_midi_event snd_seq_midi_emul snd_seq snd_emu10k1 snd_util_mem snd_hwdep snd_mpu401 snd_via82xx snd_mpu401_uart snd_cs46xx gameport snd_rawmidi snd_seq_device snd_ac97_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc
CPU:    0
EIP:    0060:[<c044d001>]    Not tainted VLI
EFLAGS: 00010007   (2.6.12-rc6-RT-V0.7.48-37) 
EIP is at setup_IO_APIC+0x4d1/0x4f0
eax: 00000200   ebx: c04372c0   ecx: 00000009   edx: 00024201
esi: 00000010   edi: df096000   ebp: 00000000   esp: df097f84
ds: 007b   es: 007b   ss: 0068   preempt: 00000002
Process IRQ 16 (pid: 2163, threadinfo=df096000 task=df7e5230)
Stack: c01149db c04372c0 00000010 c013ee34 ffffff10 dfbd58e0 df096000 c04372c0 
       c04372c0 c013eee0 c013ef44 00000873 00000001 df097fbc 00000025 df096000 
       df203e3c c0131265 fffffffc ffffffff ffffffff c01311d0 00000000 00000000 
Call Trace:
 [<c01149db>] end_level_ioapic_irq+0x4b/0xa0 (4)
 [<c013ee34>] do_hardirq+0x94/0x140 (12)
 [<c013eee0>] do_irqd+0x0/0xa0 (24)
 [<c013ef44>] do_irqd+0x64/0xa0 (4)
 [<c0131265>] kthread+0x95/0xd0 (28)
 [<c01311d0>] kthread+0x0/0xd0 (16)
 [<c010138d>] kernel_thread_helper+0x5/0x18 (16)
Code: e6 00 e6 e6 e6 00 e6 e6 e6 00 e6 e6 e6 00 e6 e6 e6 00 e6 e6 e6 00 e6 
e6 e6 00 e6 e6 e6 00 e6 e6 e6 00 e6 e6 e6 00 e6 e6 e6 00 fe <31> 09 00 0c 
00 01 02 2e 00 00 00 55 24 10 00 0c 00 02 02 2e 2e 
 <6>note: IRQ 16[2163] exited with preempt_count 1
BUG: scheduling while atomic: IRQ 16/0x00000001/2163
caller is do_exit+0x211/0x3b0
 [<c030e8d5>] __schedule+0x555/0x710 (8)
 [<c011ee2c>] exit_notify+0x54c/0x8a0 (24)
 [<c0117c19>] wake_up_process+0x19/0x20 (4)
 [<c011f391>] do_exit+0x211/0x3b0 (52)
 [<c010443b>] die+0x16b/0x170 (40)
 [<c01162bd>] do_page_fault+0x34d/0x540 (64)
 [<c0117948>] recalc_task_prio+0x88/0x160 (16)
 [<c030e6e0>] __schedule+0x360/0x710 (36)
 [<c0117bef>] try_to_wake_up+0x11f/0x130 (12)
 [<c0117c19>] wake_up_process+0x19/0x20 (20)
 [<c030ec8f>] preempt_schedule_irq+0x4f/0x70 (48)
 [<c010313c>] need_resched+0x20/0x24 (20)
 [<c0115f70>] do_page_fault+0x0/0x540 (48)
 [<c0103c83>] error_code+0x4f/0x54 (8)
 [<c044d001>] setup_IO_APIC+0x4d1/0x4f0 (44)
 [<c01149db>] end_level_ioapic_irq+0x4b/0xa0 (12)
 [<c013ee34>] do_hardirq+0x94/0x140 (12)
 [<c013eee0>] do_irqd+0x0/0xa0 (24)
 [<c013ef44>] do_irqd+0x64/0xa0 (4)
 [<c0131265>] kthread+0x95/0xd0 (28)
 [<c01311d0>] kthread+0x0/0xd0 (16)
 [<c010138d>] kernel_thread_helper+0x5/0x18 (16)
prev->state: 2 != TASK_RUNNING??
IRQ 16/2163: BUG in __schedule at kernel/sched.c:2968
 [<c030e85c>] __schedule+0x4dc/0x710 (8)
 [<c0117c19>] wake_up_process+0x19/0x20 (28)
 [<c011f391>] do_exit+0x211/0x3b0 (52)
 [<c010443b>] die+0x16b/0x170 (40)
 [<c01162bd>] do_page_fault+0x34d/0x540 (64)
 [<c0117948>] recalc_task_prio+0x88/0x160 (16)
 [<c030e6e0>] __schedule+0x360/0x710 (36)
 [<c0117bef>] try_to_wake_up+0x11f/0x130 (12)
 [<c0117c19>] wake_up_process+0x19/0x20 (20)
 [<c030ec8f>] preempt_schedule_irq+0x4f/0x70 (48)
 [<c010313c>] need_resched+0x20/0x24 (20)
 [<c0115f70>] do_page_fault+0x0/0x540 (48)
 [<c0103c83>] error_code+0x4f/0x54 (8)
 [<c044d001>] setup_IO_APIC+0x4d1/0x4f0 (44)
 [<c01149db>] end_level_ioapic_irq+0x4b/0xa0 (12)
 [<c013ee34>] do_hardirq+0x94/0x140 (12)
 [<c013eee0>] do_irqd+0x0/0xa0 (24)
 [<c013ef44>] do_irqd+0x64/0xa0 (4)
 [<c0131265>] kthread+0x95/0xd0 (28)
 [<c01311d0>] kthread+0x0/0xd0 (16)
 [<c010138d>] kernel_thread_helper+0x5/0x18 (16)

Is this something you've already fixed in the -50-xx series?  My build of
-50-15 just finished, so I'll reboot and hammer on the cd burner with the 
latest, anyway.

--ww