panic: thread's turnstile has pending threads

[Mark Atkinson]

John Baldwin wrote:
> On Thursday 19 April 2007 04:17:11 pm Mark Atkinson wrote:
>> Mark Atkinson wrote:
>> 
>> > 7.0-CURRENT FreeBSD 7.0-CURRENT #6: Fri Mar 30 07:14:23 PDT 2007
>> > 
>> > Has anyone seen this?   This machine has dual CPUs plus two shared
>> > interrupt 8 port serial I/O cards under uart/puc.
>> 
>> Additional info:
>> 
>> # kgdb ./kernel.debug /var/crash/vmcore.7
>> [GDB will not be able to debug user-mode threads:
>> [/usr/lib/libthread_db.so:
>> Undefined symbol "ps_pglobal_lookup"]
>> GNU gdb 6.1.1 [FreeBSD]
>> Copyright 2004 Free Software Foundation, Inc.
>> GDB is free software, covered by the GNU General Public License, and you
>> are welcome to change it and/or distribute copies of it under certain
>> conditions.
>> Type "show copying" to see the conditions.
>> There is absolutely no warranty for GDB.  Type "show warranty" for
>> details. This GDB was configured as "i386-marcel-freebsd".
>> 
>> Unread portion of the kernel message buffer:
>> panic: thread's turnstile has pending threads
> 
> This is a "can't happen" panic.  Can you go up to the frame 10 and do 'p
> *ts'?

Thanks for looking at this, this appears to related to acpi/interrupts, and
some check-in between Mar 30 and Apr 24 fixed the interrupt routing for
this machine (dell poweredge 2450).

Before this machine had uhci and one of the puc cards sharing and interrupt,
I would see occasional corruption on the serial line and this panic would
only occur every one or two days with two CPUs enabled.  After upgrading
to -current around Apr 24 this problem has never reared it's head again.   

I was so pleased that I enabled SHED_ULE and have gotten even better
responsiveness out of the old system, but I have noted one strange
application thread or serial lockup in conserver-com that has only occured
once.

-- 
Mark Atkinson
atkin901 at yahoo.com
(!wired)?(coffee++):(wired);