This is the mail archive of the ecos-discuss@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Re: Recent RedBoot/BOOTP change


On Tue, 2003-07-15 at 12:02, Jonathan Larmour wrote:
> Gary Thomas wrote:
> > Jonathan,
> > 
> > The change you recently made to BOOTP in RedBoot 
> > (redboot/current/src/net/bootp.c version 1.10) has broken
> > BOOTP - at least when using the stock DHCP server from
> > Red Hat (8.0 and newer).
> > 
> > I think we should revert this change, or find out why it was 
> > made and what the differences are between different servers, etc.
> 
> Hmm... I verified that change against the RFCs: 
> http://www.faqs.org/rfcs/rfc2132.html section 9.6. Clearly it is correct 
> (52 is option overload which is N/A for a client request and so should be 
> ignored by a server) so perhaps there is something else going on or indeed 
> servers out there are broken (although this seems unlikely). Assuming the 
> option is required at all for some reason - I haven't looked at what's 
> required for DHCP gatewaying.

It's obvious now that this option is being used just to make the
request look like a DHCP request - by specifying an innocuous option.
When the code 53 (which is supposed to indicate a DHCP request) is
used, at least the Red Hat servers balk.

Since it works with the old value (at least on some servers) and maybe
needs the new value for some others (aren't standards great?), perhaps
we should try both in the order old-way(52) new-way(53) in hopes that
one way or the other succeeds.

-- 
Gary Thomas <gary@mlbassoc.com>
MLB Associates


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]