[PATCH net] net: dsa: felix: fix tagging protocol changes with multiple CPU ports

[PATCH net] net: dsa: felix: fix tagging protocol changes with multiple CPU ports

[Vladimir Oltean]

When the device tree has 2 CPU ports defined, a single one is active
(has any dp->cpu_dp pointers point to it). Yet the second one is still a
CPU port, and DSA still calls ->change_tag_protocol on it.

On the NXP LS1028A, the CPU ports are ports 4 and 5. Port 4 is the
active CPU port and port 5 is inactive.

After the following commands:

 # Initial setting
 cat /sys/class/net/eno2/dsa/tagging
 ocelot
 echo ocelot-8021q > /sys/class/net/eno2/dsa/tagging
 echo ocelot > /sys/class/net/eno2/dsa/tagging

traffic is now broken, because the driver has moved the NPI port from
port 4 to port 5, unbeknown to DSA.

The problem can be avoided by detecting that the second CPU port is
unused, and not doing anything for it. Further rework will be needed
when proper support for multiple CPU ports is added.

Treat this as a bug and prepare current kernels to work in single-CPU
mode with multiple-CPU DT blobs.

Fixes: adb3dccf090b ("net: dsa: felix: convert to the new .change_tag_protocol DSA API")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
---
 drivers/net/dsa/ocelot/felix.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/drivers/net/dsa/ocelot/felix.c b/drivers/net/dsa/ocelot/felix.c
index 413b0006e9a2..9e28219b223d 100644
--- a/drivers/net/dsa/ocelot/felix.c
+++ b/drivers/net/dsa/ocelot/felix.c
@@ -670,6 +670,8 @@ static int felix_change_tag_protocol(struct dsa_switch *ds, int cpu,
 	struct ocelot *ocelot = ds->priv;
 	struct felix *felix = ocelot_to_felix(ocelot);
 	enum dsa_tag_protocol old_proto = felix->tag_proto;
+	bool cpu_port_active = false;
+	struct dsa_port *dp;
 	int err;
 
 	if (proto != DSA_TAG_PROTO_SEVILLE &&
@@ -677,6 +679,27 @@ static int felix_change_tag_protocol(struct dsa_switch *ds, int cpu,
 	    proto != DSA_TAG_PROTO_OCELOT_8021Q)
 		return -EPROTONOSUPPORT;
 
+	/* We don't support multiple CPU ports, yet the DT blob may have
+	 * multiple CPU ports defined. The first CPU port is the active one,
+	 * the others are inactive. In this case, DSA will call
+	 * ->change_tag_protocol() multiple times, once per CPU port.
+	 * Since we implement the tagging protocol change towards "ocelot" or
+	 * "seville" as effectively initializing the NPI port, what we are
+	 * doing is effectively changing who the NPI port is to the last @cpu
+	 * argument passed, which is an unused DSA CPU port and not the one
+	 * that should actively pass traffic.
+	 * Suppress DSA's calls on CPU ports that are inactive.
+	 */
+	dsa_switch_for_each_user_port(dp, ds) {
+		if (dp->cpu_dp->index == cpu) {
+			cpu_port_active = true;
+			break;
+		}
+	}
+
+	if (!cpu_port_active)
+		return 0;
+
 	felix_del_tag_protocol(ds, cpu, old_proto);
 
 	err = felix_set_tag_protocol(ds, cpu, proto);
-- 
2.25.1

Re: [PATCH net] net: dsa: felix: fix tagging protocol changes with multiple CPU ports

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net.git (master)
by Paolo Abeni <pabeni@redhat.com>:

On Tue, 12 Apr 2022 20:22:09 +0300 you wrote:
> When the device tree has 2 CPU ports defined, a single one is active
> (has any dp->cpu_dp pointers point to it). Yet the second one is still a
> CPU port, and DSA still calls ->change_tag_protocol on it.
> 
> On the NXP LS1028A, the CPU ports are ports 4 and 5. Port 4 is the
> active CPU port and port 5 is inactive.
> 
> [...]

Here is the summary with links:
  - [net] net: dsa: felix: fix tagging protocol changes with multiple CPU ports
    https://git.kernel.org/netdev/net/c/00fa91bc9cc2

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html