yakpf

[Robert Watson]

On Sun, 31 May 2009, Randy Bush wrote:

> and one more, this from serial console
>
> Fatal trap 12: page fault while in kernel mode
> cpuid = 0; apic id = 00
> fault virtual address   = 0x0
> fault code              = supervisor write data, page not present
> instruction pointer     = 0x20:0xffffffff8047c1da
> stack pointer           = 0x28:0xffffff807a156630
> frame pointer           = 0x28:0xffffff807a1566f0
> 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         = 1385 (nfcapd)
> trap number             = 12
> panic: page fault
> cpuid = 0
>
> due to when it happens, about 01:50 when there are no cron jobs or anything 
> obvious, i am suspecting this is not the same as kern/134011
>
> it did not leave a dump.  and i could not force one as it required a power 
> cycle to get the box's attention.
>
> nfcapd is a flow data capture from a couple of lighly loaded (10-20Mbps) 
> routers.

Is it possible to use options KDB_TRACE and KDB_PANIC to dump a stack trace to 
the serial console to use as a starting point for debugging?  And/or DDB 
scripting to run various debugging commands which log to serial console or a 
textdump?

Robert N M Watson
Computer Laboratory
University of Cambridge