>Number:         18890
>Category:       port-i386
>Synopsis:       MP bug: panic: not splsched() in cpu_switch!
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    port-i386-maintainer
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sat Nov 02 06:17:00 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     Havard Eidnes
>Release:        NetBSD 1.6K as of ~ 23:30 Nov 1 2002 UTC
>Organization:
	Unorganized
>Environment:
System: NetBSD stegg.urc.uninett.no 1.6K NetBSD 1.6K (STEGG.MPDEBUG) #3: Sat Nov  2 14:07:31 CET 2002     he@stegg.urc.uninett.no:/usr/src/sys/arch/i386/compile/STEGG.MPDEBUG i386
Architecture: i386
Machine: i386
>Description:
	When starting a (kernel) compile with

		# (nice make; time) >& list.1 & tail -f list.1

	the above mentioned panic appears to be quite reproducible.
	The stack backtrace in one of these crashes were (copied by
	hand):

panic: not splsched() in cpu_switch!
Stopped in pid 14933 (tail) at  cpu_Debugger+0x4:	leave
db{0}> trace
cpu_Debugger(c0339e00,0,d44c37f8,d44c37f8,c01fac40) at cpu_Debugger+0x4
panic(c01008c1,d44c37f8,0,c0339e00,c01b0854) at panic+0xad
cpu_switch(d44c37f8,0,195,c01e227c) at cpu_switch+0x19
bpendtsleep(c0c07000,118,c02b56c5,0,0) at bpendtsleep
kqueue_scan(d3b7e270,1bfbfb590,0,d44c37f8) at kqueue_scan+0x234
sys_kevent(d44c37f8,d47ddf80,d47ddf78,c0219553,d44c37f8) at sys_kevent+0x1d0
syscall_plain(1f,1f,1f,1f,bfbfb5c8) at syscall_plain+0xbc
db{0}> machine cpu 1
using cpu 1
db{0}> trace
db{0}> 

	I was unable to get a crash dump.

>How-To-Repeat:
	See above.

>Fix:
	Don't know.
>Release-Note:
>Audit-Trail:
>Unformatted: