bug-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Making DHCP renew work in pfinet


From: Christian Dietrich
Subject: Re: Making DHCP renew work in pfinet
Date: Mon, 22 Oct 2007 23:03:49 +0200
User-agent: Mutt/1.5.14 (2007-02-12)

* Stefan Siegl <stesie@brokenpipe.de> [Oct 15 2007 23:57] wrote:
> > The whitespaces beetween -a and the ip are missing because -a, -g, -m,
> > -p, -A, -G have only optional arguments. If the argument is no passed
> > the value will be unset e.g. the gateway.
> 
> I'm afraid, but I don't think that's a good idea.  People are used to
> passing `-a 192.168.7.5' or similar and the provided IP addresses can be
> distinguished quite easily.
> 
> If a user provides the afore mentioned option, parse_opt is called with
> key == `a' and arg == NULL.  However ``state->argv[state->next]''
> provides a pointer to the IP address which should be examined in turn.
> Unless you consider the bit another option, i.e. you cannot consume it
> as some IP address, simply increment state->next.

now '-a 192.168.100.1' '-a192.168.100.1' '-a' are possible
so i didn't have to touch /etc/dhclient-script at all

> >  if [ x$reason = xPREINIT ]; then
> > -  settrans -afg /servers/socket/2 /hurd/pfinet --dhcp -i $interface
> >    exit_with_hooks 0
> >  fi
> 
> I'm unsure concerning this, but I think you should check whether there
> is some pfinet installed and set it in case not.

At this point the settrans call will never be called, when there is no
pfinet running. There will be the error 'Can't create addrlist socket'.

> > +  {"dhcp", 'd', 0        , 0, "Prepare pfinet for dhcp"},
> I don't think you should introduce the `--dhcp' option after all.

Removed.
 
> > +    /* Look for IPv6 default route (we use the first ifa->addr as
> > +       source address), but don't yet push it to the option stack. */
> 
> This comment doesn't fit here, at least not the latter part of the
> sentence.

Yep, that's right. ( Memo to myself: Copy and paste is evil)

> Last but not least I'd like to ask you to review your patch regarding
> whitespace changes.  There are quite a few hunks in it, resulting from
> those.  Especially there were no changes to timer-emul.c at all, apart
> from whitespace ones.

I hope it is a little bit better now.

> Thanks for your work, you know that lease-renew'ing dhcp-support has
> often been asked for.  And of course welcome on board :-)

Oh Captaion, my captain.

regards,
  didi

-- 
MSCHEISS:
  MicroSoft Certified Home Engineer for Intranet System Solutions

Attachment: pfinet_dhcp2.patch
Description: Text Data

Attachment: ifconfig
Description: Text document


reply via email to

[Prev in Thread] Current Thread [Next in Thread]