Andrew Gabriel <andrew@cucumber.demon.co.uk> wrote:
> In article <476A7B8E.5060104@comcast.net>,
> "Richard B. Gilbert" <rgilbert88@comcast.net> writes:

> umount(1M) is a user level program but this crashed in the kernel.
> I haven't got a sparc system handy I can disassemble dounmount,
> but it looks to me like it might have branched from dounmount()
> to a garbage location instead of one of the relevant filesystem's
> entrypoints (such a VFS_UNMOUNT).

Hm. Before I rebooted, I noticed, that I was unable to access NFS
shares that this system exports (ie. this system also acts as a
NFS server). Might be related, don't you think?

> It might be interesting to see if mdb on the crashdump could
> give a better backtrace...
>
> cd /pool/savecore/; echo '$C' | mdb -k 0

000002a100b4f0a1 fop_inactive+0x54(300407b9700, 60000c03d98, 1205800,
657700646204ed, 180c000, 60014875980)
000002a100b4f151 dounmount+0xbc(60009ca95c0, 0, 0, 2, 300407b9700, 1)
000002a100b4f201 umount2+0x140(26df8, 0, ffbfff38, 11aac, 0, 0)
000002a100b4f2e1 syscall_trap32+0xcc(26df8, 0, ffbfff38, 11aac, ff36e308, 2)

Is that better?

Best regards,

Alexander