Bug 27393

Summary:	NMI Watchdog detected LOCKUP on CPU1 -- mouse dies
Product:	[Retired] Red Hat Linux	Reporter:	Chris Runge <crunge>
Component:	kernel	Assignee:	Pete Zaitcev <zaitcev>
Status:	CLOSED WORKSFORME	QA Contact:	Brock Organ <borgan>
Severity:	high	Docs Contact:
Priority:	high
Version:	7.1	CC:	nalin
Target Milestone:	---
Target Release:	---
Hardware:	i386
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2001-03-19 15:40:30 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Chris Runge 2001-02-13 15:28:58 UTC

Fisher

if I start X, then exit or switch to VC I see this:

NMI Watchdog detected LOCKUP on CPU1, registers:
CPU:    1
EIP:    0010:[<e0890c16>]
EFLAGS: 00000056
eax: 00000000   ebx: 00000020   ecx: 00000002   edx: 00000002
esi: 00000006   edi: 00000001   ebp: 00000003   esp: de229e7c
ds: 0018   es: 0018   ss: 0018
Process gpm (pid: 860, stackpage=de229000)
.
.
.


The mouse then dies, both under X and under a VC. service gpm restart does
not revive it; I must reboot. 

The system :

SuperMicro 370DE6 motherboard (based on ServerWorks III HE-SL chipset)
2x 1Ghz Pentium III
512 MB RAM
MS Intellimouse Explorer connected to USB port

cat'ing /proc/interrupts displays a non 0 value for NMI

Marking as high since a reboot is req'd to get the mouse to work again

Comment 1 Glen Foster 2001-02-13 20:22:30 UTC

This defect is considered MUST-FIX for Florence Gold release

Comment 2 Michael K. Johnson 2001-02-14 00:49:40 UTC

Please try the next kernel we release and see if it fixes the
problem.  If not, we need a decoded oops messages -- you need
to run the whole oops message through the ksymoops program so
that it is symbolically decoded.

Comment 3 Chris Runge 2001-02-15 16:30:04 UTC

interesting.... I had a MS Force Feedback 2 joystick connected to one of the USB
ports in addition to the mouse. If I remove the joystick the problem appears to
go away.

Comment 4 Chris Runge 2001-02-15 16:33:53 UTC

see 25849 for an example of how the joystick caused problems with Xconfigurator

Comment 5 Michael K. Johnson 2001-02-15 17:07:07 UTC

Nalin, does this resemble what you saw?

Comment 6 Michael K. Johnson 2001-02-20 17:17:21 UTC

We still need that ksymoops output...

Comment 7 Nalin Dahyabhai 2001-02-20 17:25:15 UTC

Yes, this resembles what I saw.  I've upgraded to a newer kernel since then, so
the fixed ksymoops wouldn't give you any useful information.  Will try to
reproduce again today.

Comment 8 Michael K. Johnson 2001-03-01 03:44:35 UTC

Seems to resemble latest behavior in bug #23670

Comment 9 Pete Zaitcev 2001-03-05 23:52:08 UTC

So, where is a meaningful oops report?

I do not have a joystick and everything
works fine without one (I have RIO-600
and floppy on the same chain).
I _suspect_ that something may be up with
having several HID devices, or a joystick
specifically. But I need solid information
before I can try fixing anything.