Xen devel - May 2006 - PATCH: Multicast support for linux-2.6.16-xen

I would really like this patch to be merged in. Linux''s bridging code
treats
multicasts as broadcasts, so this very simple code enables linux to run stuff 
like vrrpd inside xen, allowing various kinds of failover scenarios not 
possible without this patch.

oelewapperke@talin:~/src/xen/xen-3.0.2/linux-2.6.16-xen/drivers/xen/netfront$ 
diff netfront.c netfront-new.c
1098a1099,1106
> /** Set the multicast list
>  * since this interface is going to be connected by the linux bridging
>  * code, which broadcasts any and all multicasts are broadcasted by the
>  * linux bridge, this does not actually the need to do anything
>  */
> static void network_set_multicast_list(struct net_device *dev) {
> }
>
1167a1176
>       netdev->set_multicast_list = network_set_multicast_list;
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 2 May 2006, at 14:24, Christophe Devriese wrote:
> I would really like this patch to be merged in. Linux''s bridging
code
> treats
> multicasts as broadcasts, so this very simple code enables linux to 
> run stuff
> like vrrpd inside xen, allowing various kinds of failover scenarios not
> possible without this patch.
Send us a unified diff (diff -u) as an attachment or plain-text at the 
end of your email, and we''ll consider it.

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On Tuesday 02 May 2006 16:11, Keir Fraser wrote:
> On 2 May 2006, at 14:24, Christophe Devriese wrote:
> > I would really like this patch to be merged in. Linux''s
bridging code
> > treats
> > multicasts as broadcasts, so this very simple code enables linux to
> > run stuff
> > like vrrpd inside xen, allowing various kinds of failover scenarios
not
> > possible without this patch.
>
> Send us a unified diff (diff -u) as an attachment or plain-text at the
> end of your email, and we''ll consider it.
Here you go :

--- netfront.c  2006-04-10 00:05:52.000000000 +0200
+++ netfront-new.c      2006-05-02 15:21:17.000000000 +0200
@@ -1096,6 +1096,14 @@
        .set_tx_csum = ethtool_op_set_tx_csum,
 };

+/** Set the multicast list
+ * since this interface is going to be connected by the linux bridging
+ * code, which broadcasts any and all multicasts are broadcasted by the
+ * linux bridge, this does not actually the need to do anything
+ */
+static void network_set_multicast_list(struct net_device *dev) {
+}
+
 /** Create a network device.
  * @param handle device handle
  * @param val return parameter for created device
@@ -1165,6 +1173,7 @@
        netdev->stop            = network_close;
        netdev->get_stats       = network_get_stats;
        netdev->poll            = netif_poll;
+       netdev->set_multicast_list = network_set_multicast_list;
        netdev->uninit          = netif_uninit;
        netdev->weight          = 64;
        netdev->features        = NETIF_F_IP_CSUM;


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Christophe Devriese wrote:
> On Tuesday 02 May 2006 16:11, Keir Fraser wrote:
>   
>> On 2 May 2006, at 14:24, Christophe Devriese wrote:
>>     
>>> I would really like this patch to be merged in. Linux''s
bridging code
>>> treats
>>> multicasts as broadcasts, so this very simple code enables linux to
>>> run stuff
>>> like vrrpd inside xen, allowing various kinds of failover scenarios
not
>>> possible without this patch.
>>>       
>> Send us a unified diff (diff -u) as an attachment or plain-text at the
>> end of your email, and we''ll consider it.
>>     
>
> Here you go :
>   
You need a Signed-off-by line.  Also, when creating patches, it''s best 
to do it from the very top of the tree.  Better yet, use hg diff.

Regards,

Anthony Liguori
> --- netfront.c  2006-04-10 00:05:52.000000000 +0200
> +++ netfront-new.c      2006-05-02 15:21:17.000000000 +0200
> @@ -1096,6 +1096,14 @@
>         .set_tx_csum = ethtool_op_set_tx_csum,
>  };
>
> +/** Set the multicast list
> + * since this interface is going to be connected by the linux bridging
> + * code, which broadcasts any and all multicasts are broadcasted by the
> + * linux bridge, this does not actually the need to do anything
> + */
> +static void network_set_multicast_list(struct net_device *dev) {
> +}
> +
>  /** Create a network device.
>   * @param handle device handle
>   * @param val return parameter for created device
> @@ -1165,6 +1173,7 @@
>         netdev->stop            = network_close;
>         netdev->get_stats       = network_get_stats;
>         netdev->poll            = netif_poll;
> +       netdev->set_multicast_list = network_set_multicast_list;
>         netdev->uninit          = netif_uninit;
>         netdev->weight          = 64;
>         netdev->features        = NETIF_F_IP_CSUM;
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>   

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On Tuesday 02 May 2006 17:27, Anthony Liguori wrote:
> Christophe Devriese wrote:
> > On Tuesday 02 May 2006 16:11, Keir Fraser wrote:
> >> On 2 May 2006, at 14:24, Christophe Devriese wrote:
> >>> I would really like this patch to be merged in.
Linux''s bridging code
> >>> treats
> >>> multicasts as broadcasts, so this very simple code enables
linux to
> >>> run stuff
> >>> like vrrpd inside xen, allowing various kinds of failover
scenarios not
> >>> possible without this patch.
> >>
> >> Send us a unified diff (diff -u) as an attachment or plain-text at
the
> >> end of your email, and we''ll consider it.
> >
> > Here you go :
>
> You need a Signed-off-by line.  Also, when creating patches, it''s
best
> to do it from the very top of the tree.  Better yet, use hg diff.
What is a signed-off line ? What format does it take ? Do you want me to patch 
it in the mercurial repository ?
>
> Regards,
>
> Anthony Liguori
>
> > --- netfront.c  2006-04-10 00:05:52.000000000 +0200
> > +++ netfront-new.c      2006-05-02 15:21:17.000000000 +0200
> > @@ -1096,6 +1096,14 @@
> >         .set_tx_csum = ethtool_op_set_tx_csum,
> >  };
> >
> > +/** Set the multicast list
> > + * since this interface is going to be connected by the linux
bridging
> > + * code, which broadcasts any and all multicasts are broadcasted by
the
> > + * linux bridge, this does not actually the need to do anything
> > + */
> > +static void network_set_multicast_list(struct net_device *dev) {
> > +}
> > +
> >  /** Create a network device.
> >   * @param handle device handle
> >   * @param val return parameter for created device
> > @@ -1165,6 +1173,7 @@
> >         netdev->stop            = network_close;
> >         netdev->get_stats       = network_get_stats;
> >         netdev->poll            = netif_poll;
> > +       netdev->set_multicast_list = network_set_multicast_list;
> >         netdev->uninit          = netif_uninit;
> >         netdev->weight          = 64;
> >         netdev->features        = NETIF_F_IP_CSUM;
> >
> >
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@lists.xensource.com
> > http://lists.xensource.com/xen-devel
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 3 May 2006, at 09:47, Christophe Devriese wrote:
>> You need a Signed-off-by line.  Also, when creating patches,
it''s best
>> to do it from the very top of the tree.  Better yet, use hg diff.
>
> What is a signed-off line ? What format does it take ? Do you want me 
> to patch
> it in the mercurial repository ?
Signed-off-by: Christophe Devriese <christophe.devriese@eurid.eu>

You just put the line above somewhere in your email and that says that 
the patch is okay for open source.

In the case of your last patch, it was so trivial that I simply signed 
off on it myself, so don''t worry about it this time round.

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Christophe Devriese wrote:
>>You need a Signed-off-by line.  Also, when creating patches,
it''s best
>>to do it from the very top of the tree.  Better yet, use hg diff.
> 
> 
> What is a signed-off line ? What format does it take ? Do you want me to
patch
> it in the mercurial repository ?
> 

http://wiki.xensource.com/xenwiki/SubmittingXenPatches

thanks,
Nivedita


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel