hi
Am 24.03.2015 um 08:20 schrieb Claudio Jeker:
On Tue, Mar 24, 2015 at 05:41:11AM +0100, Holger Glaess wrote:
hi
i play a little bit with rdomain
if add my pppoe device to rdomain 1
# cat /etc/hostname.pppoe0
rdomain 1
rtlabel netcologne
inet6 autoconf
inet 0.0.0.0 255.255.255.255 NONE \
pppoedev msk0 authproto pap \
authname 'foo@bar' authkey 'xyz' up
dest 0.0.0.1
!/sbin/route -T 1 add default -ifp pppoe0 0.0.0.1
!/sbin/route -T 1 add -inet6 default -ifp pppoe0 ::0.0.0.1
the pf lines
pass in on $lan_if from any to $myserver rtable 1
pass in on rdomain 1 proto tcp from <allow_ssh> to ($nc_pppoe:0) port 22
rdr-to 192.168.131.250 port 22 rtable 0
match out on $nc_pppoe nat-to ($nc_pppoe:0)
so , all works fine but if i connect to the system by ssh and do , for
example , an tcpdump -n- r /var/log/pflog
the whole machine crash to an kernel panik.
Please include the ddb output of show panic and tr so that we have an idea
where to start looking.
this are the last line on the monitor
UHCI0-3 Host controller halted
UHCI0-3 host controller proccess error
uvm_fault(0xd0bb3f60, 0x0, 0, 1) ->e
kernel:page fault trap,code=0
stopped at trap+0x21: movl 0x2c(%edx),%edi
after enter "trace"
trap (number -1)
uvm_fault(0xd0bb3f60, 0x0, 0, 1) ->e
kernel:page fault trap,code=0
faulted in DDB; continuing
if you need more , output please let me know and how to generate.
Holger
this is reproducable and , looks like , just when you have an inbound
connection ( rdr ) .
the kernel-pppoe works well inside the rdomain he discover properly the ip4
and the ip6 address.
