| Message ID | 1469722498-7450-1-git-send-email-cascardo@redhat.com |
|---|---|
| State | Accepted |
| Headers | show |
On Thu, Jul 28, 2016 at 01:14:58PM -0300, Thadeu Lima de Souza Cascardo wrote: > When the route table is reset, the addresses list may be out of date, as we race > for the many netlink socket notifications. > > A quick fix for this is flushing the addresses list, before dumping the routes > and gathering source addresses for them. > > That way, instead of using invalid source addresses or preventing an entry to be > added because of missing source addresses, repeated tests showed the correct > entry is always added. > > As route-table.c is only built for Linux, we don't need to be concerned that > Windows does not have netdev_get_addrs_list_flush, since it uses > route-table-stub.c instead. > > Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@redhat.com> I am sorry. Missed it again. Fixes: a8704b502785 ("tunneling: Handle multiple ip address for given device.") > --- > lib/route-table.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/lib/route-table.c b/lib/route-table.c > index 58e7f62..0457de3 100644 > --- a/lib/route-table.c > +++ b/lib/route-table.c > @@ -25,6 +25,7 @@ > #include <net/if.h> > > #include "hash.h" > +#include "netdev.h" > #include "netlink.h" > #include "netlink-notifier.h" > #include "netlink-socket.h" > @@ -151,6 +152,7 @@ route_table_reset(void) > struct ofpbuf request, reply, buf; > > route_map_clear(); > + netdev_get_addrs_list_flush(); > route_table_valid = true; > rt_change_seq++; > > -- > 2.7.4 > > _______________________________________________ > dev mailing list > dev@openvswitch.org > http://openvswitch.org/mailman/listinfo/dev
On Thu, Jul 28, 2016 at 01:19:11PM -0300, Thadeu Lima de Souza Cascardo wrote: > On Thu, Jul 28, 2016 at 01:14:58PM -0300, Thadeu Lima de Souza Cascardo wrote: > > When the route table is reset, the addresses list may be out of date, as we race > > for the many netlink socket notifications. > > > > A quick fix for this is flushing the addresses list, before dumping the routes > > and gathering source addresses for them. > > > > That way, instead of using invalid source addresses or preventing an entry to be > > added because of missing source addresses, repeated tests showed the correct > > entry is always added. > > > > As route-table.c is only built for Linux, we don't need to be concerned that > > Windows does not have netdev_get_addrs_list_flush, since it uses > > route-table-stub.c instead. > > > > Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@redhat.com> > > I am sorry. Missed it again. > > Fixes: a8704b502785 ("tunneling: Handle multiple ip address for given device.") No problem. I applied this to master, with that added.
diff --git a/lib/route-table.c b/lib/route-table.c index 58e7f62..0457de3 100644 --- a/lib/route-table.c +++ b/lib/route-table.c @@ -25,6 +25,7 @@ #include <net/if.h> #include "hash.h" +#include "netdev.h" #include "netlink.h" #include "netlink-notifier.h" #include "netlink-socket.h" @@ -151,6 +152,7 @@ route_table_reset(void) struct ofpbuf request, reply, buf; route_map_clear(); + netdev_get_addrs_list_flush(); route_table_valid = true; rt_change_seq++;
When the route table is reset, the addresses list may be out of date, as we race for the many netlink socket notifications. A quick fix for this is flushing the addresses list, before dumping the routes and gathering source addresses for them. That way, instead of using invalid source addresses or preventing an entry to be added because of missing source addresses, repeated tests showed the correct entry is always added. As route-table.c is only built for Linux, we don't need to be concerned that Windows does not have netdev_get_addrs_list_flush, since it uses route-table-stub.c instead. Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@redhat.com> --- lib/route-table.c | 2 ++ 1 file changed, 2 insertions(+)