UP kernel, panic during installworld (mtree)

Matthew Dillon dillon at apollo.backplane.com
Sat Jul 11 08:43:16 PDT 2009


:Hi,
:
:Single user mode during installworld, UP kernel (100% panic):
:
:mtree -deU -f /usr/src/etc/mtree/BSD.root.dist -p /
:
:Fatal trap 12: page fault while in kernel mode
:...
:...
:
:Stopped at hashdev+0x2f: cmpl 0x24(%esi),%ebx
:hashdev()
:get_dev()
:spec_getattr()
:ufsspec_getattr()
:ufs_vnoperatespec()
:vop_getattr()
:vn_stat()
:kern_stat()
:sys_lstat()
:syscalls()
:
:May1st kernel works as expected (I didn't have other old versions of
:kernel on this box).
:
:Coredump is uploaded to leaf/~sephe/crash/{kernel,vmcore}.0
:
:Best Regards,
:sephe

    If the problem is repeatable please reproduce with
    sysctl debug.dev_refs=1

    It's very odd, the cdev_t in the hash list is full of
    garbage, but it isn't deadc0de... it's like corruption of some
    sort instead.  There is a sanity check in cdev_terminate that
    checks to make sure the device is not on the list before
    destroying it which prints something and I don't see that output
    either.  Which means something is probably stomping the structure.

					-Matt
					Matthew Dillon 
					<dillon at backplane.com>





More information about the Bugs mailing list