Fatal trap 12 on live migrate

[Adrian Chadd]

Ah, I need to see if I can dig up the stuff Kip told me a while ago
about bringing up gdb to actually debug FreeBSD DomU's over the
network.

Basically, there's a quick hack that can be done to build a
gdb-xenserver process which runs in the Dom0 and exports the DomU as a
remote gdb debug stub. You can then attach to it from a debugger on a
FreeBSD machine and do things like set breakpoints, check process
state, etc.

It makes debugging so very much easier.

I suggest to start - grab the xen 4.0.1 source tarball and see if
there's a gdb-xenserver directory (or something gdb'y and xen'y in
there.) If so, figure out how to build it. :-) Then email the list
with your results and I'll see what I can help you with via email.

Note though - I can't really look at any of this stuff until after mid
November due to university studies and exams (which is primarily why
I've been quiet about the whole Xen thing this year.)



Adrian

On 23 October 2010 03:31, Linda Messerschmidt
<linda.messerschmidt at gmail.com> wrote:
> Hello,
>
> With 8.1-STABLE (r214217) up and running under Xen 4.0.1 on the
> 2.6.32.24 pvops kernel, this command:
>
> xm migrate freebsd8s localhost
>
> Produces the following crash:
>
> kernel trap 12 with interrupts disabled
> Kernel page fault with the following non-sleepable locks held:
> exclusive sleep mutex intr sources (intr sources) r = 0 (0xc051c540)
> locked @ /usr/src/sys/i386/i386/intr_machdep.c:190
> KDB: stack backtrace:
> X_db_sym_numargs(c0333f0d,c00fbc50,d4667ad4,a,d4667b10,...) at
> X_db_sym_numargs+0x146
> kdb_backtrace(be,3,ffffffff,c04e24d4,d4667b64,...) at kdb_backtrace+0x2a
> witness_display_spinlock(c0336345,d4667b78,4,1,0,...) at
> witness_display_spinlock+0x75
> witness_warn(5,0,c035da1d,c03b4e7c,c234d2a8,...) at witness_warn+0x1fe
> trap(d4667bfc) at trap+0x16a
> alltraps(c2366cc0,f,d4667c78,c02dc08e,83,...) at alltraps+0x1b
> unbind_from_irqhandler(83,d4667cd8,c02d74d3,c22f7780,c22db03c,...) at
> unbind_from_irqhandler+0x21
> xencons_suspend(c22f7780,c22db03c,c036c3b8,b5,d4667cc8,...) at
> xencons_suspend+0x1e
> evtchn_device_upcall(c038d474,c262dc80,2,301,a,...) at
> evtchn_device_upcall+0x4e3
> xs_suspend(0,d4667d38,c032bf9f,345,c234d2a8,...) at xs_suspend+0x1fe
> fork_exit(c02d9e10,0,d4667d38) at fork_exit+0xb8
> fork_trampoline() at fork_trampoline+0x8
> --- trap 0, eip = 0, esp = 0xd4667d70, ebp = 0 ---
>
>
> Fatal trap 12: page fault while in kernel mode
> cpuid = 0; apic id = 00
> fault virtual address   = 0x0
> fault code              = supervisor read, page not present
> instruction pointer     = 0x21:0x0
> stack pointer           = 0x29:0xd4667c3c
> frame pointer           = 0x29:0xd4667c5c
> code segment            = base 0x0, limit 0xfffff, type 0x1b
>                        = DPL 1, pres 1, def32 1, gran 1
> processor eflags        = resume, IOPL = 0
> current process         = 14 (xenwatch)
> [thread pid 14 tid 100017 ]
> Stopped at      0:      *** error reading from address 0 ***
>
> I am no expert but it looks like some sort of null pointer problem?
> Maybe it tried to jump to a null address?  (eip = 0)
>
> I would like to learn more about the inner workings of DomU FreeBSD.
> How could I help debug this?
>
> Thanks!
> _______________________________________________
> freebsd-xen at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-xen
> To unsubscribe, send any mail to "freebsd-xen-unsubscribe at freebsd.org"
>