Just got this using 2.6.9-rc2-mm1 patched with staircase scheduler. I don't think
this has anything to do with staircase though.
What boggles me is the complete lack of backtrace. CR2 & RIP look weird.
Debian pure 64-bit user space, 2-way cpu. Hmm?
<1>Unable to handle kernel NULL pointer dereference at 0000000000000286 RIP:
[<0000000000000286>]
PML4 3b551067 PGD 3b553067 PMD 0
Oops: 0010 [1] PREEMPT SMP
CPU 1
Modules linked in: snd_intel8x0 snd_ac97_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc 8139too mii crc32 e1000
Pid: 596, comm: xfce-mcs-manage Not tainted 2.6.9-rc2-mm1
RIP: 0010:[<0000000000000286>] [<0000000000000286>]
RSP: 0018:000001003bedf4c0 EFLAGS: 2000000040
RAX: 0000000000000020 RBX: 00000000000003e8 RCX: 0000007fbffff790
RDX: 0000000000000020 RSI: 0000007fbffff790 RDI: 0000000000000020
RBP: 000003e800000238 R08: 000001003b5b1e18 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000000 R12: 7fffffffffffffff
R13: 0000000000000004 R14: 0000000000000000 R15: 0000000000000008
FS: 0000002a98f82380(0000) GS:ffffffff803ba040(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000286 CR3: 000000003ffb8000 CR4: 00000000000006e0
Process xfce-mcs-manage (pid: 596, threadinfo 000001003b5b0000, task 000001003f9557e0)
Stack: 0000000000000003 0000000000000000 ffffffff803480e0 0000000000000000
000001003baa7dc0 000001003bb11c80 0000000100000001 000001003bedf4f8
000001003bedf4f8 0000000000000001
Call Trace:
Code: Bad RIP value.
RIP [<0000000000000286>] RSP <000001003bedf4c0>
CR2: 0000000000000286
NMI Watchdog detected LOCKUP on CPU1, registers:
CPU 1
Modules linked in: snd_intel8x0 snd_ac97_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc 8139too mii crc32 e1000
Pid: 596, comm: xfce-mcs-manage Not tainted 2.6.9-rc2-mm1
RIP: 0010:[<ffffffff802962d9>] <ffffffff802962d9>{_spin_lock+41}
RSP: 0018:000001003b5b17d8 EFLAGS: 00201046
RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000001
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffffff8031b288
RBP: ffffffff8031b288 R08: 0000000000000000 R09: 000001003b5b1898
R10: 0000000000000001 R11: 0000000000000000 R12: 000001003b5b1938
R13: 0000000000000000 R14: 000001003b5b1898 R15: 0000000000000006
FS: 0000002a98f82380(0000) GS:ffffffff803ba040(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000286 CR3: 000000003ffb8000 CR4: 00000000000006e0
Process xfce-mcs-manage (pid: 596, threadinfo 000001003b5b0000, task 000001003f9557e0)
Stack: 0000000000201046 0000000000000001 0000000000000000 ffffffff80110c25
ffffffff802a78c1 ffffffff80110da0 000001003b5b1938 0000000000000000
0000000000000004 ffffffff80110f1e
Call Trace:<ffffffff80110c25>{oops_begin+101} <ffffffff80110da0>{die+32}
<ffffffff80110f1e>{do_trap+334} <ffffffff80111201>{do_invalid_op+145}
<ffffffff8012eed4>{resched_task+36} <ffffffff802138ef>{serial8250_console_write+415}
<ffffffff8010feb1>{error_exit+0} <ffffffff8012eed4>{resched_task+36}
<ffffffff8012f262>{try_to_wake_up+514} <ffffffff80149e19>{autoremove_wake_function+9}
<ffffffff80130680>{__wake_up_common+64} <ffffffff801306f3>{__wake_up+67}
<ffffffff80134c10>{vprintk+432} <ffffffff80134cbd>{printk+141}
<ffffffff80134cbd>{printk+141} <ffffffff80110ae1>{show_registers+273}
<ffffffff80206987>{do_unblank_screen+119} <ffffffff8011f88e>{bust_spinlocks+62}
<ffffffff80110c65>{oops_end+21} <ffffffff80120086>{do_page_fault+1158}
<ffffffff80248ff7>{memcpy_toiovec+55} <ffffffff802474da>{__kfree_skb+154}
<ffffffff802920c9>{unix_stream_recvmsg+1081} <ffffffff8010feb1>{error_exit+0}
<ffffffff802430b6>{sock_aio_read+278} <ffffffff80184413>{poll_freewait+67}
<ffffffff8018491e>{do_select+1038} <ffffffff8017076d>{do_sync_read+173}
<ffffffff80184420>{__pollwait+0} <ffffffff80149e10>{autoremove_wake_function+0}
<ffffffff80184e38>{sys_select+1256} <ffffffff80170881>{vfs_read+225}
<ffffffff80170b13>{sys_read+83} <ffffffff8010f3c6>{system_call+126}
Code: 7f 48 65 48 8b 04 25 18 00 00 00 ff 88 44 e0 ff ff 65 48 8b
console shuts up ...
> <1>Unable to handle kernel NULL pointer dereference at 0000000000000286 RIP:
> [<0000000000000286>]
> PML4 3b551067 PGD 3b553067 PMD 0
> Oops: 0010 [1] PREEMPT SMP
> CPU 1
> Modules linked in: snd_intel8x0 snd_ac97_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc 8139too mii crc32 e1000
> Pid: 596, comm: xfce-mcs-manage Not tainted 2.6.9-rc2-mm1
> RIP: 0010:[<0000000000000286>] [<0000000000000286>]
Someone corrupted the stack most likely, overwriting the return
address. When the function returned it jumped to nirvana.
I would look for any new functions that manipulate arrays on the stack
and double check them. Suggest you revert the staircase patch and try again.
-Andi