kernel trap 12 with interrupts disabled [bge0 on 7.2R]

[Chris Timmons]

(kgdb) list *0xc07a4dac
0xc07a4dac is in devvn_refthread (/usr/src/sys/kern/kern_conf.c:209).
204		struct cdev_priv *cdp;
205
206		mtx_assert(&devmtx, MA_NOTOWNED);
207		csw = NULL;
208		dev_lock();
209		*devp = vp->v_rdev;
210		if (*devp != NULL) {
211			cdp = (*devp)->si_priv;
212			if ((cdp->cdp_flags & CDP_SCHED_DTR) == 0) {
213				csw = (*devp)->si_devsw;


On Thu, 14 May 2009, Chris Timmons wrote:

>
> Yesterday I updated a rock-solid machine (uptime hundreds of days) from 
> 7-stable circa July, 2008, to the latest stable.   I run Nessus on this 
> machine, with about 60 concurrent scans.  It pushes the load average up as 
> high as 20 for short periods of time, but overall is reasonably efficient.
>
> I have never had the box become unresponsive, let alone crash, under any load 
> scenario.
>
> This morning, I ran my first scan on 7.2-stable, with Nessus 4.0.  It lasted 
> about 30 seconds before:
>
>
> Fatal trap 12: page fault while in kernel mode
> cpuid = 2; apic id = 06
> fault virtual address	= 0x1c
> fault code		= supervisor read, page not present
> instruction pointer	= 0x20:0xc07a4dac
> stack pointer	        = 0x28:0xee156ad4
> frame pointer	        = 0x28:0xee156ad8
> code segment		= base 0x0, limit 0xfffff, type 0x1b
> 			= DPL 0, pres 1, def32 1, gran 1
> processor eflags	= interrupt enabled, resume, IOPL = 0
> current process		= 5263 (nessusd)
> trap number		= 12
> panic: page fault
> cpuid = 3