search for: added_by_user

...| 1 + net/bridge/br_switchdev.c | 1 + 2 files changed, 2 insertions(+) diff --git a/include/net/switchdev.h b/include/net/switchdev.h index ca0312b78294..aaf918d4ba67 100644 --- a/include/net/switchdev.h +++ b/include/net/switchdev.h @@ -249,6 +249,7 @@ struct switchdev_notifier_fdb_info { u8 added_by_user:1, is_local:1, locked:1, + is_dyn:1, offloaded:1; }; diff --git a/net/bridge/br_switchdev.c b/net/bridge/br_switchdev.c index de18e9c1d7a7..9707d3fdb396 100644 --- a/net/bridge/br_switchdev.c +++ b/net/bridge/br_switchdev.c @@ -134,6 +134,7 @@ static void br_switchdev_fdb_popul...

...> 2 files changed, 2 insertions(+) > > diff --git a/include/net/switchdev.h b/include/net/switchdev.h > index ca0312b78294..aaf918d4ba67 100644 > --- a/include/net/switchdev.h > +++ b/include/net/switchdev.h > @@ -249,6 +249,7 @@ struct switchdev_notifier_fdb_info { > u8 added_by_user:1, > is_local:1, > locked:1, > + is_dyn:1, > offloaded:1; > }; > > diff --git a/net/bridge/br_switchdev.c b/net/bridge/br_switchdev.c > index 7eb6fd5bb917..60c05a00a1df 100644 > --- a/net/bridge/br_switchdev.c > +++ b/net/bridge/br_switchdev.c >...

...u_port feature ("if (switchdev_fdb_is_dynamically_learned(fdb_info))"). The reason being that a "dynamically learned" FDB entry (defined as this): static inline bool switchdev_fdb_is_dynamically_learned(const struct switchdev_notifier_fdb_info *fdb_info) { return !fdb_info->added_by_user && !fdb_info->is_local; } is also dynamic in the DSA_FDB_FLAG_DYNAMIC sense. But we install a static FDB entry for it on the CPU port. And in your follow-up patch 3/5, you make all drivers except mv88e6xxx ignore all DSA_FDB_FLAG_DYNAMIC entries (including the ones snooped from address...

...you have a suggestion, feel free. Didn't I explain what I would do from the first reply on this thread? https://patchwork.kernel.org/project/netdevbpf/patch/20230318141010.513424-3-netdev at kapio-technology.com/#25270613 As a bug fix, stop reporting to switchdev those FDB entries with BR_FDB_ADDED_BY_USER && !BR_FDB_STATIC. Then, after "net" is merged into "net-next" next Thursday (the ship has sailed for today), add "bool static" to the switchdev notifier info, and make all switchdev drivers (everywhere where a SWITCHDEV_FDB_ADD_TO_DEVICE handler appears) ignor...

On Thu, Mar 30, 2023 at 18:07, Vladimir Oltean <olteanv at gmail.com> wrote: > > Then, make DSA decide whether to handle the "added_by_user && !is_static" > combination or not, based on the presence of the DSA_FDB_FLAG_DYNAMIC > flag, which will be set in ds->supported_fdb_flags only for the mv88e6xxx > driver. Okay, so this will require a new function in the DSA layer that sets which flags are supported and t...

...ertions(+) >> >> diff --git a/include/net/switchdev.h b/include/net/switchdev.h >> index ca0312b78294..aaf918d4ba67 100644 >> --- a/include/net/switchdev.h >> +++ b/include/net/switchdev.h >> @@ -249,6 +249,7 @@ struct switchdev_notifier_fdb_info { >> u8 added_by_user:1, >> is_local:1, >> locked:1, >> + is_dyn:1, >> offloaded:1; >> }; >> >> diff --git a/net/bridge/br_switchdev.c b/net/bridge/br_switchdev.c >> index 7eb6fd5bb917..60c05a00a1df 100644 >> --- a/net/bridge/br_switchdev.c >&g...

On Thu, Mar 30, 2023 at 16:09, Vladimir Oltean <olteanv at gmail.com> wrote: > On Thu, Mar 30, 2023 at 02:59:04PM +0200, Hans Schultz wrote: >> On Thu, Mar 30, 2023 at 15:43, Vladimir Oltean <olteanv at gmail.com> wrote: >> > On Tue, Mar 28, 2023 at 09:45:26PM +0200, Hans Schultz wrote: >> >> So the solution would be to not let the DSA layer send the

...is_dynamically_learned(fdb_info))"). The > reason being > that a "dynamically learned" FDB entry (defined as this): > > static inline bool > switchdev_fdb_is_dynamically_learned(const struct > switchdev_notifier_fdb_info *fdb_info) > { > return !fdb_info->added_by_user && !fdb_info->is_local; > } > > is also dynamic in the DSA_FDB_FLAG_DYNAMIC sense. But we install a > static FDB entry for it on the CPU port. > > And in your follow-up patch 3/5, you make all drivers except mv88e6xxx > ignore all DSA_FDB_FLAG_DYNAMIC entries (incl...

There is a structural problem in switchdev, where the flag bits in struct switchdev_notifier_fdb_info (added_by_user, is_local etc) only represent a simplified / denatured view of what's in struct net_bridge_fdb_entry :: flags (BR_FDB_ADDED_BY_USER, BR_FDB_LOCAL etc). Each time we want to pass more information about struct net_bridge_fdb_entry :: flags to struct switchdev_notifier_fdb_info (here, BR_FDB_STATI...

...t; of multiple bits on both sides. Do you have any in mind? > >>> + >>> return fdb; >>> } >>> @@ -894,7 +940,7 @@ void br_fdb_update(struct net_bridge *br, struct net_bridge_port *source, >>> } >>> if (unlikely(test_bit(BR_FDB_ADDED_BY_USER, &flags))) >>> - set_bit(BR_FDB_ADDED_BY_USER, &fdb->flags); >>> + fdb_set_flag_not_learned(br, fdb, BR_FDB_ADDED_BY_USER); >> >> Unacceptable to take hash_lock and block all learning here, eventual >> consistency is ok or some other method that...

On Mon, Apr 10, 2023 at 11:49:51PM +0300, Vladimir Oltean wrote: > There is a structural problem in switchdev, where the flag bits in > struct switchdev_notifier_fdb_info (added_by_user, is_local etc) only > represent a simplified / denatured view of what's in struct > net_bridge_fdb_entry :: flags (BR_FDB_ADDED_BY_USER, BR_FDB_LOCAL etc). > Each time we want to pass more information about struct > net_bridge_fdb_entry :: flags to struct switchdev_notifier_fdb_info...

...me test the intersection of multiple bits on both sides. Do you have any in mind? > > + > > return fdb; > > } > > @@ -894,7 +940,7 @@ void br_fdb_update(struct net_bridge *br, struct net_bridge_port *source, > > } > > if (unlikely(test_bit(BR_FDB_ADDED_BY_USER, &flags))) > > - set_bit(BR_FDB_ADDED_BY_USER, &fdb->flags); > > + fdb_set_flag_not_learned(br, fdb, BR_FDB_ADDED_BY_USER); > > Unacceptable to take hash_lock and block all learning here, eventual > consistency is ok or some other method that is much lighter a...