Re: NMI watchdog detected lockup

["Randy.Dunlap" <rddunlap@osdl.org>]

Marc Bevand wrote:
> On 2004-10-17, Randy.Dunlap <rddunlap@osdl.org> wrote:
> | 
> |  I'm seeing this often during a kernel build on AIC79xx.
> |  I did one kernel build on SATA without seeing this.
> |  This is on a dual-Opteron IBM Workstation A with
> |  2 GB RAM, SATA, & SCSI.
> |  [...]
> |  NMI Watchdog detected LOCKUP on CPU0, registers:
> |  [...]
> 
> You are not the first one to observe frequent watchdog timeout
> lockup on dual Opteron systems during intense I/O operations,
> see this thread:
> 
>   http://thread.gmane.org/gmane.linux.ide/1933
> 
> Note: this does *not* seem to be SATA-related.

Hi,

Zwane suspected NMI spikes and advised me to disable nmi_watchdog
(nmi_watchdog=0).  After doing that, a kernel build completes
successfully, although with many messages like these:

Uhhuh. NMI received for unknown reason 21.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Uhhuh. NMI received for unknown reason 31.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Uhhuh. NMI received for unknown reason 31.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Uhhuh. NMI received for unknown reason 31.
Uhhuh. NMI received for unknown reason 31.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Uhhuh. NMI received for unknown reason 21.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Uhhuh. NMI received for unknown reason 21.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
Uhhuh. NMI received for unknown reason 21.
Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?


I've also seen reason == 20.

This is on 2.6.9-rc4.

Andi, any ideas?

I've had several hundred of these messages, with only 1 dazed &
confused that did not continue OK.

Adding show_registers(regs); in the NMI handler points to
default_idle():

Dazed and confused, but trying to continue
Do you have a strange power saving mode enabled?
CPU 0
Modules linked in: aic79xx usbserial aic7xxx ohci1394 ieee1394
Pid: 0, comm: swapper Not tainted 2.6.9-rc4
RIP: 0010:[<ffffffff8010f5f0>] <ffffffff8010f5f0>{default_idle+32}
RSP: 0018:ffffffff805e3fb8  EFLAGS: 00000246
RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000018
RDX: ffffffff8010f5d0 RSI: ffffffff80472b00 RDI: 0000010001e11b20
RBP: 0000000000000000 R08: 00000000ffffffff R09: 0000000000000001
R10: 0000000000000080 R11: ffffffff80562ae0 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
FS:  0000002a95b2e4c0(0000) GS:ffffffff805de800(0000) 
knlGS:0000000000000000
CS:  0010 DS: 0018 ES: 0018 CR0: 000000008005003b
CR2: 0000002a955a6000 CR3: 0000000000101000 CR4: 00000000000006e0
Process swapper (pid: 0, threadinfo ffffffff805e2000, task 
ffffffff80472b00)
Stack: ffffffff8010f9fd 0000000000000000 ffffffff805e56e5 0000000000000000
        ffffffff8055fbe0 0000000000000800 ffffffff805e51e0 
0000000000000404
        0000000000000000
Call Trace:<ffffffff8010f9fd>{cpu_idle+29} 
<ffffffff805e56e5>{start_kernel+421}
        <ffffffff805e51e0>{_sinittext+480}

Code: c3 fb f3 c3 66 66 66 90 66 66 66 90 66 66 66 90 48 83 ec 38

-- 
~Randy