On Tue, Apr 12, 2005 at 12:45:00AM +0200, Martin S. Weber wrote:
> "Mine too", but only if I have concurrent lp's; i.e. if I don't wait before
> job #1 is finished printing and start another lp there's ddb> waiting for
> me. I plan to write a bt here for quite some time but crashing that machine
> doesn't come handy all of the time...

Just right now it came handy:

#0  0xc0ae0000 in ?? ()
#1  0xc048aa80 in cpu_reboot ()
#2  0xc038a9c3 in db_sync_cmd ()
#3  0xc038a40e in db_command ()
#4  0xc038a118 in db_command_loop ()
#5  0xc038d521 in db_trap ()
#6  0xc048764e in kdb_trap ()
#7  0xc04978c2 in trap ()
#8  0xc010be5b in calltrap ()
#9  0xc02c73ec in uhci_abort_xfer ()
#10 0xc02c7204 in uhci_device_bulk_abort ()
#11 0xc0607039 in usbd_ar_pipe ()
#12 0xc0606d36 in usbd_abort_pipe ()
#13 0xc0618491 in ulptclose ()
#14 0xc041c2cf in spec_close ()
#15 0xc0388cb8 in ufsspec_close ()
#16 0xc041521b in VOP_CLOSE ()
#17 0xc0413d0e in vn_close ()
#18 0xc04147f1 in vn_closefile ()
#19 0xc03b2a7b in closef ()
#20 0xc03b2898 in fdfree ()
#21 0xc03b7dbe in exit1 ()
#22 0xc03b7bc6 in sys_exit ()
#23 0xc04972f0 in syscall_plain ()

ddb said it died at

"netbsd:usb_transfer_complete + 0x163: movl 0x48(%eax), %eax" (hand copy)
That enough of a hint, or should I go updating, compiling with -g and recrash ?

(I have the core lying around if further poking is wished for).

-Martin