[B.A.T.M.A.N.] [PATCH] batman-adv: use rx_handler_data pointer to store net_bridge_port pointer
From: Jiri Pirko jpirko@redhat.com
Register net_bridge_port pointer as rx_handler data pointer. As br_port is removed from struct net_device, another netdev priv_flag is added to indicate the device serves as a bridge port. Also rcuized pointers are now correctly dereferenced in br_fdb.c and in netfilter parts.
Signed-off-by: Jiri Pirko jpirko@redhat.com Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Sven Eckelmann sven.eckelmann@gmx.de --- batman-adv-kernelland/hard-interface.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/batman-adv-kernelland/hard-interface.c b/batman-adv-kernelland/hard-interface.c index 70fd314..0773420 100644 --- a/batman-adv-kernelland/hard-interface.c +++ b/batman-adv-kernelland/hard-interface.c @@ -74,7 +74,7 @@ static int is_valid_iface(struct net_device *net_dev) #endif
/* Device is being bridged */ - /* if (net_dev->br_port != NULL) + /* if (net_dev->priv_flags & IFF_BRIDGE_PORT) return 0; */
return 1;
Sven Eckelmann wrote:
From: Jiri Pirko jpirko@redhat.com
Register net_bridge_port pointer as rx_handler data pointer. As br_port is removed from struct net_device, another netdev priv_flag is added to indicate the device serves as a bridge port. Also rcuized pointers are now correctly dereferenced in br_fdb.c and in netfilter parts.
Signed-off-by: Jiri Pirko jpirko@redhat.com Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Sven Eckelmann sven.eckelmann@gmx.de
This patch was already merged by David S. Miller davem@davemloft.net on Tue Jun 15 20:48:58 CEST 2010 (probably for 2.6.36).
I have merged it into maint (not into trunk), but somebody has to think about a way to to get it working on older kernels. This patch has to made on top of that patch to get it integrated into the next kernel merge window.
Best regards, Sven
On Thursday 17 June 2010 00:42:11 Sven Eckelmann wrote:
I have merged it into maint (not into trunk), but somebody has to think about a way to to get it working on older kernels. This patch has to made on top of that patch to get it integrated into the next kernel merge window.
The code in question has not been enabled yet. It is just there as a possibility and could be enabled if it makes sense. I would say the one activating this code has to deal with the backward compatibility when it happens, ok ? :-)
Cheers, Marek
Marek Lindner wrote:
On Thursday 17 June 2010 00:42:11 Sven Eckelmann wrote:
I have merged it into maint (not into trunk), but somebody has to think about a way to to get it working on older kernels. This patch has to made on top of that patch to get it integrated into the next kernel merge window.
The code in question has not been enabled yet. It is just there as a possibility and could be enabled if it makes sense. I would say the one activating this code has to deal with the backward compatibility when it happens, ok ? :-)
Ok, let me rewrite it:
I have merged it into maint (not in trunk) but somebody has to think about a way to get it working on olders kernel _when he wants to enable it the future_. This patch has to made on top of that patch to get it integrated in _one of the next kernel merge windows_.
Better? :P
Personally I would do it with something like that:
diff --git a/batman-adv-kernelland/compat.h b/batman-adv-kernelland/compat.h index 8302a58..c503a51 100644 --- a/batman-adv-kernelland/compat.h +++ b/batman-adv-kernelland/compat.h @@ -232,3 +232,14 @@ static inline struct net_device_stats *dev_get_stats(struct net_device *dev) }
#endif /* < KERNEL_VERSION(2, 6, 29) */ + +#if !defined(IFF_BRIDGE_PORT) + +#ifdef netdev_isbridged +#undef netdev_isbridged +#endif + +#define netdev_isbridged(netdev) \ + ((net_dev)->br_port != NULL) + +#endif /* defined(IFF_BRIDGE_PORT) */ diff --git a/batman-adv-kernelland/hard-interface.c b/batman-adv-kernelland/hard-interface.c index 0773420..183d4c9 100644 --- a/batman-adv-kernelland/hard-interface.c +++ b/batman-adv-kernelland/hard-interface.c @@ -74,7 +74,7 @@ static int is_valid_iface(struct net_device *net_dev) #endif
/* Device is being bridged */ - /* if (net_dev->priv_flags & IFF_BRIDGE_PORT) + /* if (netdev_isbridged(net_dev)) return 0; */
return 1; diff --git a/batman-adv-kernelland/hard-interface.h b/batman-adv-kernelland/hard-interface.h index 1e5fc3e..9bd1837 100644 --- a/batman-adv-kernelland/hard-interface.h +++ b/batman-adv-kernelland/hard-interface.h @@ -38,3 +38,8 @@ int batman_skb_recv(struct sk_buff *skb, struct net_device *orig_dev); int hardif_min_mtu(void); void update_min_mtu(void); + +#ifndef netdev_isbridged +#define netdev_isbridged(netdev) \ + ((net_dev)->priv_flags & IFF_BRIDGE_PORT) +#endif
Best regards, Sven
--- a/batman-adv-kernelland/hard-interface.h +++ b/batman-adv-kernelland/hard-interface.h @@ -38,3 +38,8 @@ int batman_skb_recv(struct sk_buff *skb, struct net_device *orig_dev); int hardif_min_mtu(void); void update_min_mtu(void);
+#ifndef netdev_isbridged +#define netdev_isbridged(netdev) \
- ((net_dev)->priv_flags & IFF_BRIDGE_PORT)
+#endif
Maybe this could be pushed upstream? Seems like a useful bit of code in general?
Andrew
Andrew Lunn wrote:
--- a/batman-adv-kernelland/hard-interface.h +++ b/batman-adv-kernelland/hard-interface.h @@ -38,3 +38,8 @@ int batman_skb_recv(struct sk_buff *skb,
struct net_device *orig_dev);int hardif_min_mtu(void); void update_min_mtu(void);
+#ifndef netdev_isbridged +#define netdev_isbridged(netdev) \
- ((net_dev)->priv_flags & IFF_BRIDGE_PORT)
+#endif
Maybe this could be pushed upstream? Seems like a useful bit of code in general?
I don't think so. It is only used in ~7 places in the whole kernel. If you want then please feel free to add something like this and submit it to David S. Miller.
Best regards, Sven
On Thursday 17 June 2010 01:10:17 Sven Eckelmann wrote:
I have merged it into maint (not in trunk) but somebody has to think about a way to get it working on olders kernel _when he wants to enable it the future_. This patch has to made on top of that patch to get it integrated in _one of the next kernel merge windows_.
Ok, I think we are on the same page then. :-) I applied the patch (revision 1699).
Thanks, Marek
b.a.t.m.a.n@lists.open-mesh.org
-
Andrew Lunn -
Marek Lindner -
Sven Eckelmann