7.0 CURRENT kernel's ath driver causes page fault, kernel panic (debugging kernel)

[Robert Watson]

On Fri, 18 Jul 2008, Edward Ruggeri wrote:

> I can launch lynx (my favorite non-graphical browser) or firefox and load 
> google.  I can make a search and get results, but the kernel always panics 
> by the time I try to load a third webpage (or earlier). This is what I get:
>
> Fatal trap 12: page fault while in kernel mode
> cpuid = 1; apic id = 01
> fault virtual address = 0x0
> fault code = supervisor read, page not present
> instruction pointer = 0x20:0xc0484aa6
> stack pointer = 0x28:0xe7ffe8bc
> frame pointer = 0x28:0xe7ffe928
> 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 = 1427 (lynx) [I've also seen this from ath0 taskq]
> trap number = 12
> panic: page fault
> cpuid = 1
> Uptime: 51m37s
> Physical memory: 2014 MB
> Dumping 109 MB:
> Snyncing disks, vndoes remaining...[Zeros]
>
> I'm pretty much a freeBSD novice, so I don't know what to do as it prints 
> out line after line of zeros (slowly).  Normally I just shut the sucker 
> down, because it seems content to print the zeros forever...  But as FreeBSD 
> starts up, it says "savecore: no dumps found."  But I have compiled the 
> kernel with debugging symbols (makeoptions DEBUG=-g), so maybe you can 
> advise.

This is a NULL pointer dereference in the kernel, which almost certainly means 
a kernel bug.  Per e-mail you've already received from Garrett Cooper, we need 
some further debugging information, such as a kernel astack trace.  If you're 
unable to get a dump, a DDB stack trace would probably help quite a bit 
(options DDB, when it drops into the debugger, save the results of the command 
"trace").  Please do a file a PR with those results.

I'm a bit puzzled that your box is syncing after a panic -- normally the 
kernel is configured not to do that -- could kern.sync_on_panic be set in your 
loader.conf or sysctl.conf?

Robert N M Watson
Computer Laboratory
University of Cambridge

>
> I hope someone is able to help.  Do you think I should file a problem report?
>
> Thanks for reading.  Good night!
>
> Sincerely,
>
> -- Ned Ruggeri
> _______________________________________________
> freebsd-current at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscribe at freebsd.org"
>