Skip to contentSkip to conversation

rocker: Two small changes

PATCH Applied Ido Schimmel 3 participants 6 messages Nov 1, 2022 – Nov 3, 2022

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

v2 cover rocker: Two small changes

Patch #1 avoids allocating and scheduling a work item when it is not doing any work.

Patch #2 aligns rocker with other switchdev drivers to explicitly mark FDB entries as offloaded. Needed for upcoming MAB offload [1].

[1] https://lore.kernel.org/netdev/[email protected]/

Ido Schimmel (2):
rocker: Avoid unnecessary scheduling of work item
rocker: Explicitly mark learned FDB entries as offloaded

 drivers/net/ethernet/rocker/rocker_ofdpa.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
Signature · 1 line
2.37.3
#1 lore raw

PATCH v2 1/2 rocker: Avoid unnecessary scheduling of work item

The work item function ofdpa_port_fdb_learn_work() does not do anything when 'OFDPA_OP_FLAG_LEARNED' is not set in the work item's flags.

Therefore, do not allocate and do not schedule the work item when the flag is not set.

---

Notes:
v2:
* New patch

 drivers/net/ethernet/rocker/rocker_ofdpa.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
drivers/net/ethernet/rocker/rocker_ofdpa.c +7 -7
index 58cf7cc54f40..77ad09ad8304 100644
@@ -1821,19 +1821,16 @@ static void ofdpa_port_fdb_learn_work(struct work_struct *work)
 	const struct ofdpa_fdb_learn_work *lw =
 		container_of(work, struct ofdpa_fdb_learn_work, work);
 	bool removing = (lw->flags & OFDPA_OP_FLAG_REMOVE);
-	bool learned = (lw->flags & OFDPA_OP_FLAG_LEARNED);
 	struct switchdev_notifier_fdb_info info = {};
+	enum switchdev_notifier_type event;
 
 	info.addr = lw->addr;
 	info.vid = lw->vid;
+	event = removing ? SWITCHDEV_FDB_DEL_TO_BRIDGE :
+			   SWITCHDEV_FDB_ADD_TO_BRIDGE;
 
 	rtnl_lock();
-	if (learned && removing)
-		call_switchdev_notifiers(SWITCHDEV_FDB_DEL_TO_BRIDGE,
-					 lw->ofdpa_port->dev, &info.info, NULL);
-	else if (learned && !removing)
-		call_switchdev_notifiers(SWITCHDEV_FDB_ADD_TO_BRIDGE,
-					 lw->ofdpa_port->dev, &info.info, NULL);
+	call_switchdev_notifiers(event, lw->ofdpa_port->dev, &info.info, NULL);
 	rtnl_unlock();
 
 	kfree(work);
@@ -1865,6 +1862,9 @@ static int ofdpa_port_fdb_learn(struct ofdpa_port *ofdpa_port,
 	if (!ofdpa_port_is_bridged(ofdpa_port))
 		return 0;
 
+	if (!(flags & OFDPA_OP_FLAG_LEARNED))
+		return 0;
+
 	lw = kzalloc(sizeof(*lw), GFP_ATOMIC);
 	if (!lw)
 		return -ENOMEM;
Signature · 1 line
2.37.3
Reviewed-by on v2 1/2
#2 lore raw

Re: [PATCH net-next v2 1/2] rocker: Avoid unnecessary scheduling of work item

Quoted 9 lines · Ido Schimmel

The work item function ofdpa_port_fdb_learn_work() does not do anything when 'OFDPA_OP_FLAG_LEARNED' is not set in the work item's flags.

Therefore, do not allocate and do not schedule the work item when the flag is not set.

---

#3 lore raw

PATCH v2 2/2 rocker: Explicitly mark learned FDB entries as offloaded

In reply to Ido Schimmel

Currently, FDB entries that are notified to the bridge driver via 'SWITCHDEV_FDB_ADD_TO_BRIDGE' are always marked as offloaded by the bridge. With MAB enabled, this will no longer be universally true. Device drivers will report locked FDB entries to the bridge to let it know that the corresponding hosts required authorization, but it does not mean that these entries are necessarily programmed in the underlying hardware.

We would like to solve it by having the bridge driver determine the offload indication based of the 'offloaded' bit in the FDB notification [1].

Prepare for that change by having rocker explicitly mark learned FDB entries as offloaded. This is consistent with all the other switchdev drivers.

[1] https://lore.kernel.org/netdev/[email protected]/

---

Notes:
v2:
* Simplify condition.

 drivers/net/ethernet/rocker/rocker_ofdpa.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/rocker/rocker_ofdpa.c +1 -0
index 77ad09ad8304..826990459fa4 100644
@@ -1826,6 +1826,7 @@ static void ofdpa_port_fdb_learn_work(struct work_struct *work)
 
 	info.addr = lw->addr;
 	info.vid = lw->vid;
+	info.offloaded = !removing;
 	event = removing ? SWITCHDEV_FDB_DEL_TO_BRIDGE :
 			   SWITCHDEV_FDB_ADD_TO_BRIDGE;
 
Signature · 1 line
2.37.3
Reviewed-by on v2 2/2
#4 lore raw

Re: [PATCH net-next v2 2/2] rocker: Explicitly mark learned FDB entries as offloaded

Quoted 12 lines · Ido Schimmel

Currently, FDB entries that are notified to the bridge driver via 'SWITCHDEV_FDB_ADD_TO_BRIDGE' are always marked as offloaded by the bridge. With MAB enabled, this will no longer be universally true. Device drivers will report locked FDB entries to the bridge to let it know that the corresponding hosts required authorization, but it does not mean that these entries are necessarily programmed in the underlying hardware.

We would like to solve it by having the bridge driver determine the offload indication based of the 'offloaded' bit in the FDB notification [1].

Quoted 8 lines

Prepare for that change by having rocker explicitly mark learned FDB entries as offloaded. This is consistent with all the other switchdev drivers.

[1] https://lore.kernel.org/netdev/[email protected]/

---

#5 lore raw
In reply to Ido Schimmel
This series was applied to netdev/net-next.git (master) by Jakub Kicinski <[email protected]>:

Hello:

This series was applied to netdev/net-next.git (master) by Jakub Kicinski <[email protected]>:

Quoted 9 lines · On Tue, 1 Nov 2022 14:39:34 +0200 you wrote:

Patch #1 avoids allocating and scheduling a work item when it is not doing any work.

Patch #2 aligns rocker with other switchdev drivers to explicitly mark FDB entries as offloaded. Needed for upcoming MAB offload [1].

[1] https://lore.kernel.org/netdev/[email protected]/

[...]

Here is the summary with links:
- [net-next,v2,1/2] rocker: Avoid unnecessary scheduling of work item

    https://git.kernel.org/netdev/net-next/c/42e51de97cb4
  - [net-next,v2,2/2] rocker: Explicitly mark learned FDB entries as offloaded
    https://git.kernel.org/netdev/net-next/c/386b4174827c

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html