Re: [B.A.T.M.A.N.] [PATCH] batman-adv: roaming re-routing mechanism redesign

[Antonio Quartulli <ordex@autistici.org>]

[-- Attachment #1: Type: text/plain, Size: 3034 bytes --]

On Sun, Sep 09, 2012 at 04:29:57PM +0800, Marek Lindner wrote:
> On Thursday, August 30, 2012 23:04:24 Antonio Quartulli wrote:
> > The roaming re-routing procedure has been slightly revised in order to get
> > rid of the tt_poss_change variable that was not clearly representing the
> > node/client states. Now the code directly relies on the TT_CLIENT_ROAM
> > flag that can be set on the involved local/global clients.
> 
> This seems to explain what another patch is doing ?

uhm, after having dived the two patches I probably forgot to change this. Noted.

> 
> 
> > - This patch is based on top of:
> >   "batman-adv: substitute tt_poss_change with a per-tt_entry flag"
> 
> Where is that patch ? I am unable to find it.

yep, was in my queue but I never sent it. done a few minutes ago.

> 
> 
> > +static void
> > +batadv_tt_global_del_struct(struct batadv_priv *bat_priv,
> > +			    struct batadv_tt_global_entry *tt_global_entry,
> > +			    const char *message)
> 
> Can we come up with a better name than batadv_tt_global_del_struct ? "struct" 
> is a rather generic term.


well, this function is already in the code. I just moved it.
But ok, I'll try to come up with a better idea and send a patch to rename it.

> 
> 
> >  void batadv_tt_local_add(struct net_device *soft_iface, const uint8_t
> > *addr, int ifindex)
> >  {
> >  	struct batadv_priv *bat_priv = netdev_priv(soft_iface);
> > -	struct batadv_tt_local_entry *tt_local_entry = NULL;
> > -	struct batadv_tt_global_entry *tt_global_entry = NULL;
> > +	struct batadv_tt_local_entry *tt_local = NULL;
> > +	struct batadv_tt_global_entry *tt_global = NULL;
> 
> Please don't mix code changes with renames. That makes it hard to spot changes 
> in the behavior. Split these changes.

this was needed because the code change introduced a long line and I had to
shorten the variables name in order to make it checkpatch clean.
I will send a patch to shorten the names first and then resend this patch.

> 
> 
> > +		if (tt_local->common.flags & BATADV_TT_CLIENT_PENDING) {
> > +			/* if this client is marked as pending it means that it
> > +			 * was purged out before.
> > +			 */
> 
> Scratch "out".

ops, thanks.

> 
> 
> > +			batadv_dbg(BATADV_DBG_TT, bat_priv,
> > +				   "Readding pending client %pM\n", addr);
> 
> Can you change it to "Re-adding" ?

sure.

> 
> 
> > +		/* if the entry is marked as NEW, it means that the node is not
> > +		 * the original owner of this client (in this orig-interval),
> > +		 * therefore it does need to  eventually send a roaming
> > +		 * advertisement
> > +		 */
> > +		/*if (tt_local->common.flags & BATADV_TT_CLIENT_NEW)
> > +			goto out;*/
> 
> Dead code ?

yes. I left it there, but I forgot to remote it after the tests.


I will fix all these things and send v2 together with the renaming patches.

Thank you very much.
Cheers,

-- 
Antonio Quartulli

..each of us alone is worth nothing..
Ernesto "Che" Guevara

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]