Re: [9fans] fossil+venti install and network?

[erik quanstrom]

i'm not quite sure what your problem is. but it looks like myipaddr() is bombing out.
you don't even seem to have a loopback address configured.

what does "ns | grep /net" give you? you should have at least this:

bind /net /net
bind -a '#l' /net
bind -a '#I' /net
mount -a '#s/cs' /net

(assuming you have an ethernet card. ;-))

- erik

[Eoghan Sherry]

> Why was the test working fine with a fossil only system, but not with
> fossil+venti?

The tests are failing because there is an ip device bound to /net/ipifc/0
before termrc is run. In this case it's a loopback device started by the
kernel for venti to listen on. In a fossil only system the loopback device
is not started and the tests succeed.

eoghan

[ge...@collyer.net]

My cpurc (based on one from the labs) does this:
---
# site-specific startup
if (test -e /rc/bin/cpurc.local)
. /rc/bin/cpurc.local

# [...]

# cpu-specific early startup, notably cron
if (test -e /cfg/$sysname/cpurc)
. /cfg/$sysname/cpurc

# [...]

# start up internet if we don't already have an address
if(! grep u /net/ipselftab | grep -sv 127.0.0.1)
ip/ipconfig
if(! grep -s 127.0.0.1 /net/ipselftab)
ip/ipconfig loopback /dev/null 127.1
---

which should be more robust.

One of my current tasks is to simplify administration. I'll be back
at the labs tomorrow (Monday) and will look into why we don't have
a similar cpurc nor /cfg on sources. At minimum, one wants /cfg/pxe for
pxe booting.