netdev - Re: [PATCH net-next 2/2] vxlan: allow specifying multiple default destinations

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20130427225712.GA6772@zed>
Date:	Sun, 28 Apr 2013 01:57:12 +0300
From:	Mike Rapoport <mike.rapoport@...ellosystems.com>
To:	Atzm Watanabe <atzm@...atosphere.co.jp>
Cc:	netdev@...r.kernel.org
Subject: Re: [PATCH net-next 2/2] vxlan: allow specifying multiple default
 destinations

On Fri, Apr 26, 2013 at 03:59:49PM +0900, Atzm Watanabe wrote:
> At Thu, 25 Apr 2013 14:03:49 +0300,
> Mike Rapoport wrote:
> > +/* Add remote to default destinations list */
> > +static int vxlan_remote_add(struct vxlan_dev *vxlan, struct nlattr *attr)
> > +{

...

> > +}
> 
> I think the default destinations should be used for not only sending
> but receiving, so when multicast address was added, it should be
> joined to the group, if the interface state is up.
> (Forbidding the change on the running interface may make it easy.)
> 
> Also vxlan_open() and vxlan_stop() will need to control the
> membership of groups in the default destination list.

My original idea was to simulate multicast group with a list of unicast
destinations for cases when multicast is impossible. I'd prefer to limit
the additional destinations for unicast addresses for now. The support
for several multicast groups with the same VNI can be added later on,
IMHO. 
 
> 
> >  /* See if multicast group is already in use by other ID */
> >  static bool vxlan_group_used(struct vxlan_net *vn,
> > @@ -1500,6 +1600,14 @@ static void vxlan_flush(struct vxlan_dev *vxlan)
> >  	spin_unlock_bh(&vxlan->hash_lock);
> >  }
> >  
> > +static void vxlan_remotes_flush(struct vxlan_dev *vxlan)
> > +{
> > +	struct vxlan_rdst *rd;
> > +
> > +	for (rd = vxlan->default_dst.remote_next; rd; rd = rd->remote_next)
> > +		vxlan_remote_destroy(vxlan, rd);
> > +}
> > +
> >  /* Cleanup timer and forwarding table on shutdown */
> >  static int vxlan_stop(struct net_device *dev)
> >  {
> > @@ -1511,6 +1619,7 @@ static int vxlan_stop(struct net_device *dev)
> >  	del_timer_sync(&vxlan->age_timer);
> >  
> >  	vxlan_flush(vxlan);
> > +	vxlan_remotes_flush(vxlan);
> >  
> >  	return 0;
> >  }
> 
> vxlan_stop() is called when interface state changes to down.
> I think the default destinations should not be flushed at this timing,
> and this should be done at dellink instead.

Agree, will fix.

> 
> Thanks.

--
Sincerely yours,
Mike.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html