[prev in list] [next in list] [prev in thread] [next in thread] 

List:       quagga-dev
Subject:    [quagga-dev 6752] Re: [PATCH] [zebra] Static route does not reach
From:       Joakim Tjernlund <joakim.tjernlund () transmode ! se>
Date:       2009-06-27 17:21:11
Message-ID: OF02E0042B.36F0E36D-ONC12575E2.0059B626-C12575E2.005F52BC () transmode ! se
[Download RAW message or body]

paul@jakma.org wrote on 26/06/2009 17:24:30:
>
> On Fri, 26 Jun 2009, Joakim Tjernlund wrote:
>
> >> I'm forming the opinion we need to step back a bit and perhaps
> >> impress some kind of design on it..
>
> > What to do meanwhile? Redesign will probably take too long time.
> > Apply it and redesign in the background.
>
> I didn't say 'redesign'. I said impress a design onto it. I.e.:
>
> - specify what the code /should/ do
> - try clean-up the code to follow that design
>
> I'll try get some patches together after all this patch merging
> quietens down.

I see, I will wait for these then.

>
> >> So, really, we need to refcount struct ospf_interface, don't we?
> >
> > hmm, that could work too. But exactly when to delete the interface?
>
> When the refcount hits 0.

I suspect refcounting OIs will be major surgery, but I havn't looked at it yet

>
> Also TODO: need a more better object system (there surely must be
> some existing, small library with good data-types (good object API
> particularly).
>
> >> - execute the ase calculation immediately after SPF, to eliminate the
> >>    window for interface events to come in.
> >
> > hmm, ase might only be the first user that SEGV there may be others too.
>
> What do you think, want to try?

No time ATM :( I got a new board coming Monday, our first with PCI so
I guess I will be quite busy getting that up and running.

 Jocke

_______________________________________________
Quagga-dev mailing list
Quagga-dev@lists.quagga.net
http://lists.quagga.net/mailman/listinfo/quagga-dev

[prev in list] [next in list] [prev in thread] [next in thread]