umass causes panic on 7 amd64

[Paul Schmehl]

--On Thursday, April 17, 2008 12:33:10 +0300 Andriy Gapon <avg at icyb.net.ua> 
wrote:

> on 16/04/2008 19:19 Roland Smith said the following:
>> On Wed, Apr 16, 2008 at 09:10:23AM -0700, Steve Franks wrote:
>>> Fatal trap 12: page fault while in kernel mode
>>> cpuid = 0; apic id = 00
>>> fault virtual address   = 0x0
>>> fault code              = supervisor read instruction, page not present
>>> instruction pointer     = 0x8:0x0
>>> stack pointer           = 0x10:0xffffffffa0208570
>>> frame pointer           = 0x10:0xffffff0001e1ca00
>>> code segment            = base 0x0, limit 0xfffff, type 0x1b
>>>                         = DPL 0, pres 1, long 1, def32 0, gran 1
>>> processor eflags        = interrupt enabled, resume, IOPL = 0
>>> current process         = 12 (swi4: clock sio)
>>
>> Odd. This doesn't seem to have anything to do with usb. It is in a
>> kernel thread that runs the clock and serial port.
>
> It is not as odd if you consider that "clock" is actually softclock that
> executes timeouts/callouts and I have no doubts that we do use those in
> USB subsystem.
>
>>
>> If you give the 'bt' command (backtrace) here, what does it say?
>
> I think 'bt' should be a default command in kgdb init or at least there
> should be a big advice to the users: "if you managed to run kgdb on a
> core file, then at the very least execute bt command" :-)

I wish I had a core file to analyze.  *Every* time I reboot my machine, I have 
to disconnect my usb drive.  Then I have to remount it after I'm back up and 
running.  If I leave it connected during the reboot, I get the same kind of 
errors that were posted by Steve.  After the system is up and running, umass is 
detected normally and I can mount and use the drive with no problems.  I'm on 
i386, so it doesn't look like an AMD-specific problem.

# uname -a
FreeBSD utd65257.utdallas.edu 7.0-STABLE FreeBSD 7.0-STABLE #6: Wed Apr 16 
17:14:28 CDT 2008     utd65257.utdallas.edu:/usr/obj/usr/src/sys/GENERIC  i386

I've rebuilt kernel and world six times in the hopes that recent src updates 
would fix the problem.

Unfortunately, since the error occurs during boot, I know of no way to capture 
the error message.  If I log console would that do it?  I doubt the console is 
logging at that point.  I don't think syslogd is even running yet.

-- 
Paul Schmehl (pauls at utdallas.edu)
Senior Information Security Analyst
The University of Texas at Dallas
http://www.utdallas.edu/ir/security/