kern/142728: Panic: Fatal trap 12 in g_io_request

Henry Wong HWong at Lumeta.com
Tue Jan 12 02:56:32 UTC 2010


I apologize for the unreadable section.  Having problems sending the
MIME encoded.  Let me just retype in plain text:

-------------------------------------------------------------------

Fatal trap 12: page fault while in kernel mode
cpuid = 1; apic id = 00
fault virtual address   = 0x50006
fault code              = supervisor read data, page not present
instruction pointer     = 0x20:0xffffffff805276a7
stack pointer           = 0x28:0xffffff81d8190750
frame pointer           = 0x28:0xffffff81d81907a0
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         = 7601 (ls)
trap number             = 12
panic: page fault
cpuid = 1


---------------------------------------------------------------------

Additional information since the submission of the bug report:

The original netboot/system build was done on disks that had FreeBSD
file systems from an earlier release of FreeBSD.  During the system
installation process, the disk slices and partitions were not redefined
but were reused.  The file systems on them were remade.

Since then, I have found that if, during the installation process, I
delete all the slices (and partitions) and create new slices and
partitions, I don't seem to be having the problem.  I've only done
limited testing with this procedure.

This presents a problem for us.  We need a way of not losing the data
that is on the FreeBSD slices that are not affected by the sysinstall
(i.e. all the data that are on user only partitions and in the
PostgreSQL database).  Is there a way of doing this, or do we have to
reslice and repartition all the disks.

Thank you for any help that you can provide.

Henry Wong




More information about the freebsd-bugs mailing list