Those who have been following part 1:
https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
part 2:
https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
and part 3:
https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
will know that I am trying to enable the second internal port pair from
the NXP LS1028A Felix switch for DSA-tagged traffic via "ocelot-8021q".
This series represents the final part of that effort. We have:
- the introduction of new UAPI in the form of IFLA_DSA_MASTER, the
iproute2 patch for which is here:
https://patchwork.kernel.org/project/netdevbpf/patch/[email protected]/
- preparation for LAG DSA masters in terms of suppressing some
operations for masters in the DSA core that simply don't make sense
when those masters are a bonding/team interface
- handling all the net device events that occur between DSA and a
LAG DSA master, including migration to a different DSA master when the
current master joins a LAG, or the LAG gets destroyed
- updating documentation
- adding an implementation for NXP LS1028A, where things are insanely
complicated due to hardware limitations. We have 2 tagging protocols:
* the native "ocelot" protocol (NPI port mode). This does not support
CPU ports in a LAG, and supports a single DSA master. The DSA master
can be changed between eno2 (2.5G) and eno3 (1G), but all ports must
be down during the changing process, and user ports assigned to the
old DSA master will refuse to come up if the user requests that
during a "transient" state.
* the "ocelot-8021q" software-defined protocol, where the Ethernet
ports connected to the CPU are not actually "god mode" ports as far
as the hardware is concerned. So here, static assignment between
user and CPU ports is possible by editing the PGID_SRC masks for
the port-based forwarding matrix, and "CPU ports in a LAG" simply
means "a LAG like any other".
The series was regression-tested on LS1028A using the local_termination.sh
kselftest, in most of the possible operating modes and tagging protocols.
I have not done a detailed performance evaluation yet, but using LAG, is
possible to exceed the termination bandwidth of a single CPU port in an
iperf3 test with multiple senders and multiple receivers.
Changes in v2:
- remove unused variable in dsa_port_assign_master
- add restriction that LAG slaves must be DSA masters of the same switch
tree
- move DSA master eligibility restriction earlier, from
dsa_master_lag_setup() to dsa_master_prechangeupper_sanity_check()
- set master_setup = true in dsa_master_lag_setup(), so that
dsa_master_teardown() would actually get called
- new patch to propagate extack to ds->ops->port_lag_join()
- don't overwrite extack in dsa_master_lag_setup() if
dsa_port_lag_join() provided a more specific one
- add restriction that interfaces which aren't DSA masters cannot join a
LAG DSA master
- provide alternate description of how to put CPU ports in a LAG
- pass extack to felix_port_change_master() from felix_lag_join()
v1 at:
https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
Previous (older) RFC at:
https://lore.kernel.org/netdev/[email protected]/
Vladimir Oltean (10):
net: introduce iterators over synced hw addresses
net: dsa: introduce dsa_port_get_master()
net: dsa: allow the DSA master to be seen and changed through
rtnetlink
net: dsa: don't keep track of admin/oper state on LAG DSA masters
net: dsa: suppress appending ethtool stats to LAG DSA masters
net: dsa: suppress device links to LAG DSA masters
net: dsa: propagate extack to port_lag_join
net: dsa: allow masters to join a LAG
docs: net: dsa: update information about multiple CPU ports
net: dsa: felix: add support for changing DSA master
.../networking/dsa/configuration.rst | 96 +++++
Documentation/networking/dsa/dsa.rst | 38 +-
drivers/net/dsa/bcm_sf2.c | 4 +-
drivers/net/dsa/bcm_sf2_cfp.c | 4 +-
drivers/net/dsa/lan9303-core.c | 4 +-
drivers/net/dsa/mv88e6xxx/chip.c | 27 +-
drivers/net/dsa/ocelot/felix.c | 121 +++++-
drivers/net/dsa/ocelot/felix.h | 3 +
drivers/net/dsa/qca/qca8k-common.c | 23 +-
drivers/net/dsa/qca/qca8k.h | 3 +-
.../net/ethernet/mediatek/mtk_ppe_offload.c | 2 +-
drivers/net/ethernet/mscc/ocelot.c | 11 +-
drivers/net/ethernet/mscc/ocelot_net.c | 7 +-
include/linux/netdevice.h | 6 +
include/net/dsa.h | 31 +-
include/soc/mscc/ocelot.h | 4 +-
include/uapi/linux/if_link.h | 10 +
net/dsa/Makefile | 10 +-
net/dsa/dsa.c | 9 +
net/dsa/dsa2.c | 34 +-
net/dsa/dsa_priv.h | 18 +-
net/dsa/master.c | 72 +++-
net/dsa/netlink.c | 62 +++
net/dsa/port.c | 160 +++++++-
net/dsa/slave.c | 362 +++++++++++++++++-
net/dsa/switch.c | 26 +-
net/dsa/tag_8021q.c | 4 +-
27 files changed, 1064 insertions(+), 87 deletions(-)
create mode 100644 net/dsa/netlink.c
--
2.34.1
Similar to the discussion about tracking the admin/oper state of LAG DSA
masters, we have the problem here that struct dsa_port *cpu_dp caches a
single pair of orig_ethtool_ops and netdev_ops pointers.
So if we call dsa_master_setup(bond0, cpu_dp) where cpu_dp is also the
dev->dsa_ptr of one of the physical DSA masters, we'd effectively
overwrite what we cached from that physical netdev with what replaced
from the bonding interface.
We don't need DSA ethtool stats on the bonding interface when used as
DSA master, it's good enough to have them just on the physical DSA
masters, so suppress this logic.
Signed-off-by: Vladimir Oltean <[email protected]>
---
v1->v2: none
net/dsa/master.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/net/dsa/master.c b/net/dsa/master.c
index fb810edc8281..99d773b24223 100644
--- a/net/dsa/master.c
+++ b/net/dsa/master.c
@@ -226,6 +226,9 @@ static int dsa_master_ethtool_setup(struct net_device *dev)
struct dsa_switch *ds = cpu_dp->ds;
struct ethtool_ops *ops;
+ if (netif_is_lag_master(dev))
+ return 0;
+
ops = devm_kzalloc(ds->dev, sizeof(*ops), GFP_KERNEL);
if (!ops)
return -ENOMEM;
@@ -250,6 +253,9 @@ static void dsa_master_ethtool_teardown(struct net_device *dev)
{
struct dsa_port *cpu_dp = dev->dsa_ptr;
+ if (netif_is_lag_master(dev))
+ return;
+
dev->ethtool_ops = cpu_dp->orig_ethtool_ops;
cpu_dp->orig_ethtool_ops = NULL;
}
@@ -257,6 +263,9 @@ static void dsa_master_ethtool_teardown(struct net_device *dev)
static void dsa_netdev_ops_set(struct net_device *dev,
const struct dsa_netdevice_ops *ops)
{
+ if (netif_is_lag_master(dev))
+ return;
+
dev->dsa_ptr->netdev_ops = ops;
}
--
2.34.1
We store information about the DSA master's state in
cpu_dp->master_admin_up and cpu_dp->master_oper_up, and this assumes a
bijective association between a CPU port and a DSA master.
However, when we have CPU ports in a LAG (and DSA masters in a LAG too),
the way in which we set up things is that the physical DSA masters still
have dev->dsa_ptr pointing to our cpu_dp, but the bonding/team device
itself also has its dev->dsa_ptr pointing towards one of the CPU port
structures (the first one).
So logically speaking, that first cpu_dp can't keep track of both the
physical master's admin/oper state, and of the bonding master's state.
This isn't even needed; the reason why we keep track of the DSA master's
state is to know when it is available for Ethernet-based register access.
For that use case, we don't even need LAG; we just need to decide upon
one of the physical DSA masters (if there is more than 1 available) and
use that.
This change suppresses dsa_tree_master_{admin,oper}_state_change() calls
on LAG DSA masters (which will be supported in a future change), to
allow the tracking of just physical DSA masters.
Link: https://lore.kernel.org/netdev/[email protected]/
Suggested-by: Christian Marangi <[email protected]>
Signed-off-by: Vladimir Oltean <[email protected]>
---
v1->v2: none
net/dsa/dsa2.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
index 44207d848722..2b324db8ea88 100644
--- a/net/dsa/dsa2.c
+++ b/net/dsa/dsa2.c
@@ -1336,6 +1336,12 @@ void dsa_tree_master_admin_state_change(struct dsa_switch_tree *dst,
struct dsa_port *cpu_dp = master->dsa_ptr;
bool notify = false;
+ /* Don't keep track of admin state on LAG DSA masters,
+ * but rather just of physical DSA masters
+ */
+ if (netif_is_lag_master(master))
+ return;
+
if ((dsa_port_master_is_operational(cpu_dp)) !=
(up && cpu_dp->master_oper_up))
notify = true;
@@ -1353,6 +1359,12 @@ void dsa_tree_master_oper_state_change(struct dsa_switch_tree *dst,
struct dsa_port *cpu_dp = master->dsa_ptr;
bool notify = false;
+ /* Don't keep track of oper state on LAG DSA masters,
+ * but rather just of physical DSA masters
+ */
+ if (netif_is_lag_master(master))
+ return;
+
if ((dsa_port_master_is_operational(cpu_dp)) !=
(cpu_dp->master_admin_up && up))
notify = true;
--
2.34.1
Changing the DSA master means different things depending on the tagging
protocol in use.
For NPI mode ("ocelot" and "seville"), there is a single port which can
be configured as NPI, but DSA only permits changing the CPU port
affinity of user ports one by one. So changing a user port to a
different NPI port globally changes what the NPI port is, and breaks the
user ports still using the old one.
To address this while still permitting the change of the NPI port,
require that the user ports which are still affine to the old NPI port
are down, and cannot be brought up until they are all affine to the same
NPI port.
The tag_8021q mode ("ocelot-8021q") is more flexible, in that each user
port can be freely assigned to one CPU port or to the other. This works
by filtering host addresses towards both tag_8021q CPU ports, and then
restricting the forwarding from a certain user port only to one of the
two tag_8021q CPU ports.
Additionally, the 2 tag_8021q CPU ports can be placed in a LAG. This
works by enabling forwarding via PGID_SRC from a certain user port
towards the logical port ID containing both tag_8021q CPU ports, but
then restricting forwarding per packet, via the LAG hash codes in
PGID_AGGR, to either one or the other.
When we change the DSA master to a LAG device, DSA guarantees us that
the LAG has at least one lower interface as a physical DSA master.
But DSA masters can come and go as lowers of that LAG, and
ds->ops->port_change_master() will not get called, because the DSA
master is still the same (the LAG). So we need to hook into the
ds->ops->port_lag_{join,leave} calls on the CPU ports and update the
logical port ID of the LAG that user ports are assigned to.
Signed-off-by: Vladimir Oltean <[email protected]>
---
v1->v2: pass extack to felix_port_change_master() from felix_lag_join()
drivers/net/dsa/ocelot/felix.c | 118 ++++++++++++++++++++++++++++-
drivers/net/dsa/ocelot/felix.h | 3 +
drivers/net/ethernet/mscc/ocelot.c | 3 +-
include/soc/mscc/ocelot.h | 1 +
4 files changed, 122 insertions(+), 3 deletions(-)
diff --git a/drivers/net/dsa/ocelot/felix.c b/drivers/net/dsa/ocelot/felix.c
index 82dcc21a7172..d2a9d292160c 100644
--- a/drivers/net/dsa/ocelot/felix.c
+++ b/drivers/net/dsa/ocelot/felix.c
@@ -42,6 +42,25 @@ static struct net_device *felix_classify_db(struct dsa_db db)
}
}
+static int felix_cpu_port_for_master(struct dsa_switch *ds,
+ struct net_device *master)
+{
+ struct ocelot *ocelot = ds->priv;
+ struct dsa_port *cpu_dp;
+ int lag;
+
+ if (netif_is_lag_master(master)) {
+ mutex_lock(&ocelot->fwd_domain_lock);
+ lag = ocelot_bond_get_id(ocelot, master);
+ mutex_unlock(&ocelot->fwd_domain_lock);
+
+ return lag;
+ }
+
+ cpu_dp = master->dsa_ptr;
+ return cpu_dp->index;
+}
+
/* Set up VCAP ES0 rules for pushing a tag_8021q VLAN towards the CPU such that
* the tagger can perform RX source port identification.
*/
@@ -422,6 +441,40 @@ static unsigned long felix_tag_npi_get_host_fwd_mask(struct dsa_switch *ds)
return BIT(ocelot->num_phys_ports);
}
+static int felix_tag_npi_change_master(struct dsa_switch *ds, int port,
+ struct net_device *master,
+ struct netlink_ext_ack *extack)
+{
+ struct dsa_port *dp = dsa_to_port(ds, port), *other_dp;
+ struct ocelot *ocelot = ds->priv;
+
+ if (netif_is_lag_master(master)) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "LAG DSA master only supported using ocelot-8021q");
+ return -EOPNOTSUPP;
+ }
+
+ /* Changing the NPI port breaks user ports still assigned to the old
+ * one, so only allow it while they're down, and don't allow them to
+ * come back up until they're all changed to the new one.
+ */
+ dsa_switch_for_each_user_port(other_dp, ds) {
+ struct net_device *slave = other_dp->slave;
+
+ if (other_dp != dp && (slave->flags & IFF_UP) &&
+ dsa_port_to_master(other_dp) != master) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Cannot change while old master still has users");
+ return -EOPNOTSUPP;
+ }
+ }
+
+ felix_npi_port_deinit(ocelot, ocelot->npi);
+ felix_npi_port_init(ocelot, felix_cpu_port_for_master(ds, master));
+
+ return 0;
+}
+
/* Alternatively to using the NPI functionality, that same hardware MAC
* connected internally to the enetc or fman DSA master can be configured to
* use the software-defined tag_8021q frame format. As far as the hardware is
@@ -433,6 +486,7 @@ static const struct felix_tag_proto_ops felix_tag_npi_proto_ops = {
.setup = felix_tag_npi_setup,
.teardown = felix_tag_npi_teardown,
.get_host_fwd_mask = felix_tag_npi_get_host_fwd_mask,
+ .change_master = felix_tag_npi_change_master,
};
static int felix_tag_8021q_setup(struct dsa_switch *ds)
@@ -507,10 +561,24 @@ static unsigned long felix_tag_8021q_get_host_fwd_mask(struct dsa_switch *ds)
return dsa_cpu_ports(ds);
}
+static int felix_tag_8021q_change_master(struct dsa_switch *ds, int port,
+ struct net_device *master,
+ struct netlink_ext_ack *extack)
+{
+ int cpu = felix_cpu_port_for_master(ds, master);
+ struct ocelot *ocelot = ds->priv;
+
+ ocelot_port_unassign_dsa_8021q_cpu(ocelot, port);
+ ocelot_port_assign_dsa_8021q_cpu(ocelot, port, cpu);
+
+ return felix_update_trapping_destinations(ds, true);
+}
+
static const struct felix_tag_proto_ops felix_tag_8021q_proto_ops = {
.setup = felix_tag_8021q_setup,
.teardown = felix_tag_8021q_teardown,
.get_host_fwd_mask = felix_tag_8021q_get_host_fwd_mask,
+ .change_master = felix_tag_8021q_change_master,
};
static void felix_set_host_flood(struct dsa_switch *ds, unsigned long mask,
@@ -673,6 +741,16 @@ static void felix_port_set_host_flood(struct dsa_switch *ds, int port,
!!felix->host_flood_mc_mask, true);
}
+static int felix_port_change_master(struct dsa_switch *ds, int port,
+ struct net_device *master,
+ struct netlink_ext_ack *extack)
+{
+ struct ocelot *ocelot = ds->priv;
+ struct felix *felix = ocelot_to_felix(ocelot);
+
+ return felix->tag_proto_ops->change_master(ds, port, master, extack);
+}
+
static int felix_set_ageing_time(struct dsa_switch *ds,
unsigned int ageing_time)
{
@@ -865,8 +943,17 @@ static int felix_lag_join(struct dsa_switch *ds, int port,
struct netlink_ext_ack *extack)
{
struct ocelot *ocelot = ds->priv;
+ int err;
- return ocelot_port_lag_join(ocelot, port, lag.dev, info, extack);
+ err = ocelot_port_lag_join(ocelot, port, lag.dev, info, extack);
+ if (err)
+ return err;
+
+ /* Update the logical LAG port that serves as tag_8021q CPU port */
+ if (!dsa_is_cpu_port(ds, port))
+ return 0;
+
+ return felix_port_change_master(ds, port, lag.dev, extack);
}
static int felix_lag_leave(struct dsa_switch *ds, int port,
@@ -876,7 +963,11 @@ static int felix_lag_leave(struct dsa_switch *ds, int port,
ocelot_port_lag_leave(ocelot, port, lag.dev);
- return 0;
+ /* Update the logical LAG port that serves as tag_8021q CPU port */
+ if (!dsa_is_cpu_port(ds, port))
+ return 0;
+
+ return felix_port_change_master(ds, port, lag.dev, NULL);
}
static int felix_lag_change(struct dsa_switch *ds, int port)
@@ -1014,6 +1105,27 @@ static void felix_phylink_mac_link_up(struct dsa_switch *ds, int port,
felix->info->port_sched_speed_set(ocelot, port, speed);
}
+static int felix_port_enable(struct dsa_switch *ds, int port,
+ struct phy_device *phydev)
+{
+ struct dsa_port *dp = dsa_to_port(ds, port);
+ struct ocelot *ocelot = ds->priv;
+
+ if (!dsa_port_is_user(dp))
+ return 0;
+
+ if (ocelot->npi >= 0) {
+ struct net_device *master = dsa_port_to_master(dp);
+
+ if (felix_cpu_port_for_master(ds, master) != ocelot->npi) {
+ dev_err(ds->dev, "Multiple masters are not allowed\n");
+ return -EINVAL;
+ }
+ }
+
+ return 0;
+}
+
static void felix_port_qos_map_init(struct ocelot *ocelot, int port)
{
int i;
@@ -1913,6 +2025,7 @@ const struct dsa_switch_ops felix_switch_ops = {
.phylink_mac_select_pcs = felix_phylink_mac_select_pcs,
.phylink_mac_link_down = felix_phylink_mac_link_down,
.phylink_mac_link_up = felix_phylink_mac_link_up,
+ .port_enable = felix_port_enable,
.port_fast_age = felix_port_fast_age,
.port_fdb_dump = felix_fdb_dump,
.port_fdb_add = felix_fdb_add,
@@ -1968,6 +2081,7 @@ const struct dsa_switch_ops felix_switch_ops = {
.port_add_dscp_prio = felix_port_add_dscp_prio,
.port_del_dscp_prio = felix_port_del_dscp_prio,
.port_set_host_flood = felix_port_set_host_flood,
+ .port_change_master = felix_port_change_master,
};
struct net_device *felix_port_to_netdev(struct ocelot *ocelot, int port)
diff --git a/drivers/net/dsa/ocelot/felix.h b/drivers/net/dsa/ocelot/felix.h
index deb8dde1fc19..e4fd5eef57a0 100644
--- a/drivers/net/dsa/ocelot/felix.h
+++ b/drivers/net/dsa/ocelot/felix.h
@@ -71,6 +71,9 @@ struct felix_tag_proto_ops {
int (*setup)(struct dsa_switch *ds);
void (*teardown)(struct dsa_switch *ds);
unsigned long (*get_host_fwd_mask)(struct dsa_switch *ds);
+ int (*change_master)(struct dsa_switch *ds, int port,
+ struct net_device *master,
+ struct netlink_ext_ack *extack);
};
extern const struct dsa_switch_ops felix_switch_ops;
diff --git a/drivers/net/ethernet/mscc/ocelot.c b/drivers/net/ethernet/mscc/ocelot.c
index 5c18f8986975..7a613b52787d 100644
--- a/drivers/net/ethernet/mscc/ocelot.c
+++ b/drivers/net/ethernet/mscc/ocelot.c
@@ -1382,7 +1382,7 @@ static u32 ocelot_get_bond_mask(struct ocelot *ocelot, struct net_device *bond)
/* The logical port number of a LAG is equal to the lowest numbered physical
* port ID present in that LAG. It may change if that port ever leaves the LAG.
*/
-static int ocelot_bond_get_id(struct ocelot *ocelot, struct net_device *bond)
+int ocelot_bond_get_id(struct ocelot *ocelot, struct net_device *bond)
{
int bond_mask = ocelot_get_bond_mask(ocelot, bond);
@@ -1391,6 +1391,7 @@ static int ocelot_bond_get_id(struct ocelot *ocelot, struct net_device *bond)
return __ffs(bond_mask);
}
+EXPORT_SYMBOL_GPL(ocelot_bond_get_id);
/* Returns the mask of user ports assigned to this DSA tag_8021q CPU port.
* Note that when CPU ports are in a LAG, the user ports are assigned to the
diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h
index ea19e8ef1f61..967ba30ea636 100644
--- a/include/soc/mscc/ocelot.h
+++ b/include/soc/mscc/ocelot.h
@@ -1234,6 +1234,7 @@ int ocelot_port_lag_join(struct ocelot *ocelot, int port,
void ocelot_port_lag_leave(struct ocelot *ocelot, int port,
struct net_device *bond);
void ocelot_port_lag_change(struct ocelot *ocelot, int port, bool lag_tx_active);
+int ocelot_bond_get_id(struct ocelot *ocelot, struct net_device *bond);
int ocelot_devlink_sb_register(struct ocelot *ocelot);
void ocelot_devlink_sb_unregister(struct ocelot *ocelot);
--
2.34.1
Some network drivers use __dev_mc_sync()/__dev_uc_sync() and therefore
program the hardware only with addresses with a non-zero sync_cnt.
Some of the above drivers also need to save/restore the address
filtering lists when certain events happen, and they need to walk
through the struct net_device :: uc and struct net_device :: mc lists.
But these lists contain unsynced addresses too.
To keep the appearance of an elementary form of data encapsulation,
provide iterators through these lists that only look at entries with a
non-zero sync_cnt, instead of filtering entries out from device drivers.
Signed-off-by: Vladimir Oltean <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
---
v1->v2: none
include/linux/netdevice.h | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index f0068c1ff1df..9f42fc871c3b 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -253,11 +253,17 @@ struct netdev_hw_addr_list {
#define netdev_uc_empty(dev) netdev_hw_addr_list_empty(&(dev)->uc)
#define netdev_for_each_uc_addr(ha, dev) \
netdev_hw_addr_list_for_each(ha, &(dev)->uc)
+#define netdev_for_each_synced_uc_addr(_ha, _dev) \
+ netdev_for_each_uc_addr((_ha), (_dev)) \
+ if ((_ha)->sync_cnt)
#define netdev_mc_count(dev) netdev_hw_addr_list_count(&(dev)->mc)
#define netdev_mc_empty(dev) netdev_hw_addr_list_empty(&(dev)->mc)
#define netdev_for_each_mc_addr(ha, dev) \
netdev_hw_addr_list_for_each(ha, &(dev)->mc)
+#define netdev_for_each_synced_mc_addr(_ha, _dev) \
+ netdev_for_each_mc_addr((_ha), (_dev)) \
+ if ((_ha)->sync_cnt)
struct hh_cache {
unsigned int hh_len;
--
2.34.1
There are 2 ways in which a DSA user port may become handled by 2 CPU
ports in a LAG:
(1) its current DSA master joins a LAG
ip link del bond0 && ip link add bond0 type bond mode 802.3ad
ip link set eno2 master bond0
When this happens, all user ports with "eno2" as DSA master get
automatically migrated to "bond0" as DSA master.
(2) it is explicitly configured as such by the user
# Before, the DSA master was eno3
ip link set swp0 type dsa master bond0
The design of this configuration is that the LAG device dynamically
becomes a DSA master through dsa_master_setup() when the first physical
DSA master becomes a LAG slave, and stops being so through
dsa_master_teardown() when the last physical DSA master leaves.
A LAG interface is considered as a valid DSA master only if it contains
existing DSA masters, and no other lower interfaces. Therefore, we
mainly rely on method (1) to enter this configuration.
Each physical DSA master (LAG slave) retains its dev->dsa_ptr for when
it becomes a standalone DSA master again. But the LAG master also has a
dev->dsa_ptr, and this is actually duplicated from one of the physical
LAG slaves, and therefore needs to be balanced when LAG slaves come and
go.
To the switch driver, putting DSA masters in a LAG is seen as putting
their associated CPU ports in a LAG.
We need to prepare cross-chip host FDB notifiers for CPU ports in a LAG,
by calling the driver's ->lag_fdb_add method rather than ->port_fdb_add.
Signed-off-by: Vladimir Oltean <[email protected]>
---
v1->v2:
- add restriction that LAG slaves must be DSA masters of the same switch
tree
- move DSA master eligibility restriction earlier, from
dsa_master_lag_setup() to dsa_master_prechangeupper_sanity_check()
- set master_setup = true in dsa_master_lag_setup(), so that
dsa_master_teardown() would actually get called
- don't overwrite extack in dsa_master_lag_setup() if
dsa_port_lag_join() provided a more specific one
- add restriction that interfaces which aren't DSA masters cannot join a
LAG DSA master
include/net/dsa.h | 12 +++
net/dsa/dsa_priv.h | 5 +
net/dsa/master.c | 49 ++++++++++
net/dsa/port.c | 1 +
net/dsa/slave.c | 231 +++++++++++++++++++++++++++++++++++++++++++--
net/dsa/switch.c | 22 ++++-
6 files changed, 310 insertions(+), 10 deletions(-)
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 1c80e75b3cad..d777eac5694f 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -300,6 +300,9 @@ struct dsa_port {
u8 master_admin_up:1;
u8 master_oper_up:1;
+ /* Valid only on user ports */
+ u8 cpu_port_in_lag:1;
+
u8 setup:1;
struct device_node *dn;
@@ -724,6 +727,9 @@ static inline bool dsa_port_offloads_lag(struct dsa_port *dp,
static inline struct net_device *dsa_port_to_master(const struct dsa_port *dp)
{
+ if (dp->cpu_port_in_lag)
+ return dsa_port_lag_dev_get(dp->cpu_dp);
+
return dp->cpu_dp->master;
}
@@ -811,6 +817,12 @@ dsa_tree_offloads_bridge_dev(struct dsa_switch_tree *dst,
return false;
}
+static inline bool dsa_port_tree_same(const struct dsa_port *a,
+ const struct dsa_port *b)
+{
+ return a->ds->dst == b->ds->dst;
+}
+
typedef int dsa_fdb_dump_cb_t(const unsigned char *addr, u16 vid,
bool is_static, void *data);
struct dsa_switch_ops {
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index f0ae54d0435e..129e4a649c7e 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -185,6 +185,11 @@ static inline int dsa_tag_protocol_overhead(const struct dsa_device_ops *ops)
/* master.c */
int dsa_master_setup(struct net_device *dev, struct dsa_port *cpu_dp);
void dsa_master_teardown(struct net_device *dev);
+int dsa_master_lag_setup(struct net_device *lag_dev, struct dsa_port *cpu_dp,
+ struct netdev_lag_upper_info *uinfo,
+ struct netlink_ext_ack *extack);
+void dsa_master_lag_teardown(struct net_device *lag_dev,
+ struct dsa_port *cpu_dp);
static inline struct net_device *dsa_master_find_slave(struct net_device *dev,
int device, int port)
diff --git a/net/dsa/master.c b/net/dsa/master.c
index 2176c14b97a8..40367ab41cf8 100644
--- a/net/dsa/master.c
+++ b/net/dsa/master.c
@@ -428,3 +428,52 @@ void dsa_master_teardown(struct net_device *dev)
*/
wmb();
}
+
+int dsa_master_lag_setup(struct net_device *lag_dev, struct dsa_port *cpu_dp,
+ struct netdev_lag_upper_info *uinfo,
+ struct netlink_ext_ack *extack)
+{
+ bool master_setup = false;
+ int err;
+
+ if (!netdev_uses_dsa(lag_dev)) {
+ err = dsa_master_setup(lag_dev, cpu_dp);
+ if (err)
+ return err;
+
+ master_setup = true;
+ }
+
+ err = dsa_port_lag_join(cpu_dp, lag_dev, uinfo, extack);
+ if (err) {
+ if (extack && !extack->_msg)
+ NL_SET_ERR_MSG_MOD(extack,
+ "CPU port failed to join LAG");
+ goto out_master_teardown;
+ }
+
+ return 0;
+
+out_master_teardown:
+ if (master_setup)
+ dsa_master_teardown(lag_dev);
+ return err;
+}
+
+/* Tear down a master if there isn't any other user port on it,
+ * optionally also destroying LAG information.
+ */
+void dsa_master_lag_teardown(struct net_device *lag_dev,
+ struct dsa_port *cpu_dp)
+{
+ struct net_device *upper;
+ struct list_head *iter;
+
+ dsa_port_lag_leave(cpu_dp, lag_dev);
+
+ netdev_for_each_upper_dev_rcu(lag_dev, upper, iter)
+ if (dsa_slave_dev_check(upper))
+ return;
+
+ dsa_master_teardown(lag_dev);
+}
diff --git a/net/dsa/port.c b/net/dsa/port.c
index 98f7fa0cdd5c..e6289a1db0a0 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -1393,6 +1393,7 @@ static int dsa_port_assign_master(struct dsa_port *dp,
return err;
dp->cpu_dp = master->dsa_ptr;
+ dp->cpu_port_in_lag = netif_is_lag_master(master);
return 0;
}
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index 00df6cf07866..aa47ddc19fdf 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -2818,11 +2818,45 @@ dsa_slave_prechangeupper_sanity_check(struct net_device *dev,
return NOTIFY_DONE;
}
+/* To be eligible as a DSA master, a LAG must have all lower interfaces be
+ * eligible DSA masters. Additionally, all LAG slaves must be DSA masters of
+ * switches in the same switch tree.
+ */
+static int dsa_lag_master_validate(struct net_device *lag_dev,
+ struct netlink_ext_ack *extack)
+{
+ struct net_device *lower1, *lower2;
+ struct list_head *iter1, *iter2;
+
+ netdev_for_each_lower_dev(lag_dev, lower1, iter1) {
+ netdev_for_each_lower_dev(lag_dev, lower2, iter2) {
+ if (!netdev_uses_dsa(lower1) ||
+ !netdev_uses_dsa(lower2)) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "All LAG ports must be eligible as DSA masters");
+ return notifier_from_errno(-EINVAL);
+ }
+
+ if (lower1 == lower2)
+ continue;
+
+ if (!dsa_port_tree_same(lower1->dsa_ptr,
+ lower2->dsa_ptr)) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "LAG contains DSA masters of disjoint switch trees");
+ return notifier_from_errno(-EINVAL);
+ }
+ }
+ }
+
+ return NOTIFY_DONE;
+}
+
static int
dsa_master_prechangeupper_sanity_check(struct net_device *master,
struct netdev_notifier_changeupper_info *info)
{
- struct netlink_ext_ack *extack;
+ struct netlink_ext_ack *extack = netdev_notifier_info_to_extack(&info->info);
if (!netdev_uses_dsa(master))
return NOTIFY_DONE;
@@ -2840,13 +2874,51 @@ dsa_master_prechangeupper_sanity_check(struct net_device *master,
if (netif_is_bridge_master(info->upper_dev))
return NOTIFY_DONE;
- extack = netdev_notifier_info_to_extack(&info->info);
+ /* Allow LAG uppers, subject to further restrictions in
+ * dsa_lag_master_prechangelower_sanity_check()
+ */
+ if (netif_is_lag_master(info->upper_dev))
+ return dsa_lag_master_validate(info->upper_dev, extack);
NL_SET_ERR_MSG_MOD(extack,
"DSA master cannot join unknown upper interfaces");
return notifier_from_errno(-EBUSY);
}
+static int
+dsa_lag_master_prechangelower_sanity_check(struct net_device *dev,
+ struct netdev_notifier_changeupper_info *info)
+{
+ struct netlink_ext_ack *extack = netdev_notifier_info_to_extack(&info->info);
+ struct net_device *lag_dev = info->upper_dev;
+ struct net_device *lower;
+ struct list_head *iter;
+
+ if (!netdev_uses_dsa(lag_dev) || !netif_is_lag_master(lag_dev))
+ return NOTIFY_DONE;
+
+ if (!info->linking)
+ return NOTIFY_DONE;
+
+ if (!netdev_uses_dsa(dev)) {
+ NL_SET_ERR_MSG(extack,
+ "Only DSA masters can join a LAG DSA master");
+ return notifier_from_errno(-EINVAL);
+ }
+
+ netdev_for_each_lower_dev(lag_dev, lower, iter) {
+ if (!dsa_port_tree_same(dev->dsa_ptr, lower->dsa_ptr)) {
+ NL_SET_ERR_MSG(extack,
+ "Interface is DSA master for a different switch tree than this LAG");
+ return notifier_from_errno(-EINVAL);
+ }
+
+ break;
+ }
+
+ return NOTIFY_DONE;
+}
+
/* Don't allow bridging of DSA masters, since the bridge layer rx_handler
* prevents the DSA fake ethertype handler to be invoked, so we don't get the
* chance to strip off and parse the DSA switch tag protocol header (the bridge
@@ -2887,6 +2959,136 @@ dsa_bridge_prechangelower_sanity_check(struct net_device *new_lower,
return NOTIFY_DONE;
}
+static void dsa_tree_migrate_ports_from_lag_master(struct dsa_switch_tree *dst,
+ struct net_device *lag_dev)
+{
+ struct net_device *new_master = dsa_tree_find_first_master(dst);
+ struct dsa_port *dp;
+ int err;
+
+ dsa_tree_for_each_user_port(dp, dst) {
+ if (dsa_port_to_master(dp) != lag_dev)
+ continue;
+
+ err = dsa_slave_change_master(dp->slave, new_master, NULL);
+ if (err) {
+ netdev_err(dp->slave,
+ "failed to restore master to %s: %pe\n",
+ new_master->name, ERR_PTR(err));
+ }
+ }
+}
+
+static int dsa_master_lag_join(struct net_device *master,
+ struct net_device *lag_dev,
+ struct netdev_lag_upper_info *uinfo,
+ struct netlink_ext_ack *extack)
+{
+ struct dsa_port *cpu_dp = master->dsa_ptr;
+ struct dsa_switch_tree *dst = cpu_dp->dst;
+ struct dsa_port *dp;
+ int err;
+
+ err = dsa_master_lag_setup(lag_dev, cpu_dp, uinfo, extack);
+ if (err)
+ return err;
+
+ dsa_tree_for_each_user_port(dp, dst) {
+ if (dsa_port_to_master(dp) != master)
+ continue;
+
+ err = dsa_slave_change_master(dp->slave, lag_dev, extack);
+ if (err)
+ goto restore;
+ }
+
+ return 0;
+
+restore:
+ dsa_tree_for_each_user_port_continue_reverse(dp, dst) {
+ if (dsa_port_to_master(dp) != lag_dev)
+ continue;
+
+ err = dsa_slave_change_master(dp->slave, master, NULL);
+ if (err) {
+ netdev_err(dp->slave,
+ "failed to restore master to %s: %pe\n",
+ master->name, ERR_PTR(err));
+ }
+ }
+
+ dsa_master_lag_teardown(lag_dev, master->dsa_ptr);
+
+ return err;
+}
+
+static void dsa_master_lag_leave(struct net_device *master,
+ struct net_device *lag_dev)
+{
+ struct dsa_port *dp, *cpu_dp = lag_dev->dsa_ptr;
+ struct dsa_switch_tree *dst = cpu_dp->dst;
+ struct dsa_port *new_cpu_dp = NULL;
+ struct net_device *lower;
+ struct list_head *iter;
+
+ netdev_for_each_lower_dev(lag_dev, lower, iter) {
+ if (netdev_uses_dsa(lower)) {
+ new_cpu_dp = lower->dsa_ptr;
+ break;
+ }
+ }
+
+ if (new_cpu_dp) {
+ /* Update the CPU port of the user ports still under the LAG
+ * so that dsa_port_to_master() continues to work properly
+ */
+ dsa_tree_for_each_user_port(dp, dst)
+ if (dsa_port_to_master(dp) == lag_dev)
+ dp->cpu_dp = new_cpu_dp;
+
+ /* Update the index of the virtual CPU port to match the lowest
+ * physical CPU port
+ */
+ lag_dev->dsa_ptr = new_cpu_dp;
+ wmb();
+ } else {
+ /* If the LAG DSA master has no ports left, migrate back all
+ * user ports to the first physical CPU port
+ */
+ dsa_tree_migrate_ports_from_lag_master(dst, lag_dev);
+ }
+
+ /* This DSA master has left its LAG in any case, so let
+ * the CPU port leave the hardware LAG as well
+ */
+ dsa_master_lag_teardown(lag_dev, master->dsa_ptr);
+}
+
+static int dsa_master_changeupper(struct net_device *dev,
+ struct netdev_notifier_changeupper_info *info)
+{
+ struct netlink_ext_ack *extack;
+ int err = NOTIFY_DONE;
+
+ if (!netdev_uses_dsa(dev))
+ return err;
+
+ extack = netdev_notifier_info_to_extack(&info->info);
+
+ if (netif_is_lag_master(info->upper_dev)) {
+ if (info->linking) {
+ err = dsa_master_lag_join(dev, info->upper_dev,
+ info->upper_info, extack);
+ err = notifier_from_errno(err);
+ } else {
+ dsa_master_lag_leave(dev, info->upper_dev);
+ err = NOTIFY_OK;
+ }
+ }
+
+ return err;
+}
+
static int dsa_slave_netdevice_event(struct notifier_block *nb,
unsigned long event, void *ptr)
{
@@ -2905,6 +3107,10 @@ static int dsa_slave_netdevice_event(struct notifier_block *nb,
if (notifier_to_errno(err))
return err;
+ err = dsa_lag_master_prechangelower_sanity_check(dev, info);
+ if (notifier_to_errno(err))
+ return err;
+
err = dsa_bridge_prechangelower_sanity_check(dev, info);
if (notifier_to_errno(err))
return err;
@@ -2930,6 +3136,10 @@ static int dsa_slave_netdevice_event(struct notifier_block *nb,
if (notifier_to_errno(err))
return err;
+ err = dsa_master_changeupper(dev, ptr);
+ if (notifier_to_errno(err))
+ return err;
+
break;
}
case NETDEV_CHANGELOWERSTATE: {
@@ -2937,12 +3147,21 @@ static int dsa_slave_netdevice_event(struct notifier_block *nb,
struct dsa_port *dp;
int err;
- if (!dsa_slave_dev_check(dev))
- break;
+ if (dsa_slave_dev_check(dev)) {
+ dp = dsa_slave_to_port(dev);
+
+ err = dsa_port_lag_change(dp, info->lower_state_info);
+ }
- dp = dsa_slave_to_port(dev);
+ /* Mirror LAG port events on DSA masters that are in
+ * a LAG towards their respective switch CPU ports
+ */
+ if (netdev_uses_dsa(dev)) {
+ dp = dev->dsa_ptr;
+
+ err = dsa_port_lag_change(dp, info->lower_state_info);
+ }
- err = dsa_port_lag_change(dp, info->lower_state_info);
return notifier_from_errno(err);
}
case NETDEV_CHANGE:
diff --git a/net/dsa/switch.c b/net/dsa/switch.c
index c2cb15e21324..ce56acdba203 100644
--- a/net/dsa/switch.c
+++ b/net/dsa/switch.c
@@ -398,8 +398,15 @@ static int dsa_switch_host_fdb_add(struct dsa_switch *ds,
dsa_switch_for_each_port(dp, ds) {
if (dsa_port_host_address_match(dp, info->dp)) {
- err = dsa_port_do_fdb_add(dp, info->addr, info->vid,
- info->db);
+ if (dsa_port_is_cpu(dp) && info->dp->cpu_port_in_lag) {
+ err = dsa_switch_do_lag_fdb_add(ds, dp->lag,
+ info->addr,
+ info->vid,
+ info->db);
+ } else {
+ err = dsa_port_do_fdb_add(dp, info->addr,
+ info->vid, info->db);
+ }
if (err)
break;
}
@@ -419,8 +426,15 @@ static int dsa_switch_host_fdb_del(struct dsa_switch *ds,
dsa_switch_for_each_port(dp, ds) {
if (dsa_port_host_address_match(dp, info->dp)) {
- err = dsa_port_do_fdb_del(dp, info->addr, info->vid,
- info->db);
+ if (dsa_port_is_cpu(dp) && info->dp->cpu_port_in_lag) {
+ err = dsa_switch_do_lag_fdb_del(ds, dp->lag,
+ info->addr,
+ info->vid,
+ info->db);
+ } else {
+ err = dsa_port_do_fdb_del(dp, info->addr,
+ info->vid, info->db);
+ }
if (err)
break;
}
--
2.34.1
Drivers could refuse to offload a LAG configuration for a variety of
reasons, mainly having to do with its TX type. Additionally, since DSA
masters may now also be LAG interfaces, and this will translate into a
call to port_lag_join on the CPU ports, there may be extra restrictions
there. Propagate the netlink extack to this DSA method in order for
drivers to give a meaningful error message back to the user.
Signed-off-by: Vladimir Oltean <[email protected]>
---
v1->v2: patch is new
drivers/net/dsa/mv88e6xxx/chip.c | 27 ++++++++++++++++++--------
drivers/net/dsa/ocelot/felix.c | 5 +++--
drivers/net/dsa/qca/qca8k-common.c | 23 ++++++++++++++++------
drivers/net/dsa/qca/qca8k.h | 3 ++-
drivers/net/ethernet/mscc/ocelot.c | 8 ++++++--
drivers/net/ethernet/mscc/ocelot_net.c | 7 +++----
include/net/dsa.h | 6 ++++--
include/soc/mscc/ocelot.h | 3 ++-
net/dsa/dsa_priv.h | 1 +
net/dsa/port.c | 1 +
net/dsa/switch.c | 4 ++--
11 files changed, 60 insertions(+), 28 deletions(-)
diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 6f4ea39ab466..5f178faa110f 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -6593,14 +6593,17 @@ static int mv88e6xxx_port_bridge_flags(struct dsa_switch *ds, int port,
static bool mv88e6xxx_lag_can_offload(struct dsa_switch *ds,
struct dsa_lag lag,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
struct mv88e6xxx_chip *chip = ds->priv;
struct dsa_port *dp;
int members = 0;
- if (!mv88e6xxx_has_lag(chip))
+ if (!mv88e6xxx_has_lag(chip)) {
+ NL_SET_ERR_MSG_MOD(extack, "Chip does not support LAG offload");
return false;
+ }
if (!lag.id)
return false;
@@ -6609,14 +6612,20 @@ static bool mv88e6xxx_lag_can_offload(struct dsa_switch *ds,
/* Includes the port joining the LAG */
members++;
- if (members > 8)
+ if (members > 8) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Cannot offload more than 8 LAG ports");
return false;
+ }
/* We could potentially relax this to include active
* backup in the future.
*/
- if (info->tx_type != NETDEV_LAG_TX_TYPE_HASH)
+ if (info->tx_type != NETDEV_LAG_TX_TYPE_HASH) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Can only offload LAG using hash TX type");
return false;
+ }
/* Ideally we would also validate that the hash type matches
* the hardware. Alas, this is always set to unknown on team
@@ -6769,12 +6778,13 @@ static int mv88e6xxx_port_lag_change(struct dsa_switch *ds, int port)
static int mv88e6xxx_port_lag_join(struct dsa_switch *ds, int port,
struct dsa_lag lag,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
struct mv88e6xxx_chip *chip = ds->priv;
int err, id;
- if (!mv88e6xxx_lag_can_offload(ds, lag, info))
+ if (!mv88e6xxx_lag_can_offload(ds, lag, info, extack))
return -EOPNOTSUPP;
/* DSA LAG IDs are one-based */
@@ -6827,12 +6837,13 @@ static int mv88e6xxx_crosschip_lag_change(struct dsa_switch *ds, int sw_index,
static int mv88e6xxx_crosschip_lag_join(struct dsa_switch *ds, int sw_index,
int port, struct dsa_lag lag,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
struct mv88e6xxx_chip *chip = ds->priv;
int err;
- if (!mv88e6xxx_lag_can_offload(ds, lag, info))
+ if (!mv88e6xxx_lag_can_offload(ds, lag, info, extack))
return -EOPNOTSUPP;
mv88e6xxx_reg_lock(chip);
diff --git a/drivers/net/dsa/ocelot/felix.c b/drivers/net/dsa/ocelot/felix.c
index c73ef5f7aa64..82dcc21a7172 100644
--- a/drivers/net/dsa/ocelot/felix.c
+++ b/drivers/net/dsa/ocelot/felix.c
@@ -861,11 +861,12 @@ static void felix_bridge_leave(struct dsa_switch *ds, int port,
static int felix_lag_join(struct dsa_switch *ds, int port,
struct dsa_lag lag,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
struct ocelot *ocelot = ds->priv;
- return ocelot_port_lag_join(ocelot, port, lag.dev, info);
+ return ocelot_port_lag_join(ocelot, port, lag.dev, info, extack);
}
static int felix_lag_leave(struct dsa_switch *ds, int port,
diff --git a/drivers/net/dsa/qca/qca8k-common.c b/drivers/net/dsa/qca/qca8k-common.c
index bba95613e218..fb45b598847b 100644
--- a/drivers/net/dsa/qca/qca8k-common.c
+++ b/drivers/net/dsa/qca/qca8k-common.c
@@ -1017,7 +1017,8 @@ int qca8k_port_vlan_del(struct dsa_switch *ds, int port,
static bool qca8k_lag_can_offload(struct dsa_switch *ds,
struct dsa_lag lag,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
struct dsa_port *dp;
int members = 0;
@@ -1029,15 +1030,24 @@ static bool qca8k_lag_can_offload(struct dsa_switch *ds,
/* Includes the port joining the LAG */
members++;
- if (members > QCA8K_NUM_PORTS_FOR_LAG)
+ if (members > QCA8K_NUM_PORTS_FOR_LAG) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Cannot offload more than 4 LAG ports");
return false;
+ }
- if (info->tx_type != NETDEV_LAG_TX_TYPE_HASH)
+ if (info->tx_type != NETDEV_LAG_TX_TYPE_HASH) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Can only offload LAG using hash TX type");
return false;
+ }
if (info->hash_type != NETDEV_LAG_HASH_L2 &&
- info->hash_type != NETDEV_LAG_HASH_L23)
+ info->hash_type != NETDEV_LAG_HASH_L23) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Can only offload L2 or L2+L3 TX hash");
return false;
+ }
return true;
}
@@ -1160,11 +1170,12 @@ static int qca8k_lag_refresh_portmap(struct dsa_switch *ds, int port,
}
int qca8k_port_lag_join(struct dsa_switch *ds, int port, struct dsa_lag lag,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
int ret;
- if (!qca8k_lag_can_offload(ds, lag, info))
+ if (!qca8k_lag_can_offload(ds, lag, info, extack))
return -EOPNOTSUPP;
ret = qca8k_lag_setup_hash(ds, lag, info);
diff --git a/drivers/net/dsa/qca/qca8k.h b/drivers/net/dsa/qca/qca8k.h
index e36ecc9777f4..0b7a5cb12321 100644
--- a/drivers/net/dsa/qca/qca8k.h
+++ b/drivers/net/dsa/qca/qca8k.h
@@ -512,7 +512,8 @@ int qca8k_port_vlan_del(struct dsa_switch *ds, int port,
/* Common port LAG function */
int qca8k_port_lag_join(struct dsa_switch *ds, int port, struct dsa_lag lag,
- struct netdev_lag_upper_info *info);
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack);
int qca8k_port_lag_leave(struct dsa_switch *ds, int port,
struct dsa_lag lag);
diff --git a/drivers/net/ethernet/mscc/ocelot.c b/drivers/net/ethernet/mscc/ocelot.c
index 874fb2a5874e..5c18f8986975 100644
--- a/drivers/net/ethernet/mscc/ocelot.c
+++ b/drivers/net/ethernet/mscc/ocelot.c
@@ -2132,10 +2132,14 @@ static void ocelot_migrate_lag_fdbs(struct ocelot *ocelot,
int ocelot_port_lag_join(struct ocelot *ocelot, int port,
struct net_device *bond,
- struct netdev_lag_upper_info *info)
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack)
{
- if (info->tx_type != NETDEV_LAG_TX_TYPE_HASH)
+ if (info->tx_type != NETDEV_LAG_TX_TYPE_HASH) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Can only offload LAG using hash TX type");
return -EOPNOTSUPP;
+ }
mutex_lock(&ocelot->fwd_domain_lock);
diff --git a/drivers/net/ethernet/mscc/ocelot_net.c b/drivers/net/ethernet/mscc/ocelot_net.c
index 2979fb1ba0f7..50858cc10fef 100644
--- a/drivers/net/ethernet/mscc/ocelot_net.c
+++ b/drivers/net/ethernet/mscc/ocelot_net.c
@@ -1412,11 +1412,10 @@ static int ocelot_netdevice_lag_join(struct net_device *dev,
int port = priv->port.index;
int err;
- err = ocelot_port_lag_join(ocelot, port, bond, info);
- if (err == -EOPNOTSUPP) {
- NL_SET_ERR_MSG_MOD(extack, "Offloading not supported");
+ err = ocelot_port_lag_join(ocelot, port, bond, info, extack);
+ if (err == -EOPNOTSUPP)
+ /* Offloading not supported, fall back to software LAG */
return 0;
- }
bridge_dev = netdev_master_upper_dev_get(bond);
if (!bridge_dev || !netif_is_bridge_master(bridge_dev))
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 3f717c3fcba0..1c80e75b3cad 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -1094,7 +1094,8 @@ struct dsa_switch_ops {
int port);
int (*crosschip_lag_join)(struct dsa_switch *ds, int sw_index,
int port, struct dsa_lag lag,
- struct netdev_lag_upper_info *info);
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack);
int (*crosschip_lag_leave)(struct dsa_switch *ds, int sw_index,
int port, struct dsa_lag lag);
@@ -1169,7 +1170,8 @@ struct dsa_switch_ops {
int (*port_lag_change)(struct dsa_switch *ds, int port);
int (*port_lag_join)(struct dsa_switch *ds, int port,
struct dsa_lag lag,
- struct netdev_lag_upper_info *info);
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack);
int (*port_lag_leave)(struct dsa_switch *ds, int port,
struct dsa_lag lag);
diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h
index 355cfdedc43b..ea19e8ef1f61 100644
--- a/include/soc/mscc/ocelot.h
+++ b/include/soc/mscc/ocelot.h
@@ -1229,7 +1229,8 @@ int ocelot_port_mdb_del(struct ocelot *ocelot, int port,
const struct net_device *bridge);
int ocelot_port_lag_join(struct ocelot *ocelot, int port,
struct net_device *bond,
- struct netdev_lag_upper_info *info);
+ struct netdev_lag_upper_info *info,
+ struct netlink_ext_ack *extack);
void ocelot_port_lag_leave(struct ocelot *ocelot, int port,
struct net_device *bond);
void ocelot_port_lag_change(struct ocelot *ocelot, int port, bool lag_tx_active);
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index d252a04ed725..f0ae54d0435e 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -88,6 +88,7 @@ struct dsa_notifier_lag_info {
const struct dsa_port *dp;
struct dsa_lag lag;
struct netdev_lag_upper_info *info;
+ struct netlink_ext_ack *extack;
};
/* DSA_NOTIFIER_VLAN_* */
diff --git a/net/dsa/port.c b/net/dsa/port.c
index e557b42de9f2..98f7fa0cdd5c 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -635,6 +635,7 @@ int dsa_port_lag_join(struct dsa_port *dp, struct net_device *lag_dev,
struct dsa_notifier_lag_info info = {
.dp = dp,
.info = uinfo,
+ .extack = extack,
};
struct net_device *bridge_dev;
int err;
diff --git a/net/dsa/switch.c b/net/dsa/switch.c
index 4dfd68cf61c5..c2cb15e21324 100644
--- a/net/dsa/switch.c
+++ b/net/dsa/switch.c
@@ -507,12 +507,12 @@ static int dsa_switch_lag_join(struct dsa_switch *ds,
{
if (info->dp->ds == ds && ds->ops->port_lag_join)
return ds->ops->port_lag_join(ds, info->dp->index, info->lag,
- info->info);
+ info->info, info->extack);
if (info->dp->ds != ds && ds->ops->crosschip_lag_join)
return ds->ops->crosschip_lag_join(ds, info->dp->ds->index,
info->dp->index, info->lag,
- info->info);
+ info->info, info->extack);
return -EOPNOTSUPP;
}
--
2.34.1
There is a desire to support for DSA masters in a LAG.
That configuration is intended to work by simply enslaving the master to
a bonding/team device. But the physical DSA master (the LAG slave) still
has a dev->dsa_ptr, and that cpu_dp still corresponds to the physical
CPU port.
However, we would like to be able to retrieve the LAG that's the upper
of the physical DSA master. In preparation for that, introduce a helper
called dsa_port_get_master() that replaces all occurrences of the
dp->cpu_dp->master pattern. The distinction between LAG and non-LAG will
be made later within the helper itself.
Signed-off-by: Vladimir Oltean <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
---
v1->v2: none
drivers/net/dsa/bcm_sf2.c | 4 +--
drivers/net/dsa/bcm_sf2_cfp.c | 4 +--
drivers/net/dsa/lan9303-core.c | 4 +--
.../net/ethernet/mediatek/mtk_ppe_offload.c | 2 +-
include/net/dsa.h | 5 ++++
net/dsa/dsa2.c | 8 +++---
net/dsa/dsa_priv.h | 2 +-
net/dsa/port.c | 28 +++++++++----------
net/dsa/slave.c | 11 ++++----
net/dsa/tag_8021q.c | 4 +--
10 files changed, 38 insertions(+), 34 deletions(-)
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 572f7450b527..6507663f35e5 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -983,7 +983,7 @@ static int bcm_sf2_sw_resume(struct dsa_switch *ds)
static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
struct ethtool_wolinfo *wol)
{
- struct net_device *p = dsa_to_port(ds, port)->cpu_dp->master;
+ struct net_device *p = dsa_port_to_master(dsa_to_port(ds, port));
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
struct ethtool_wolinfo pwol = { };
@@ -1007,7 +1007,7 @@ static void bcm_sf2_sw_get_wol(struct dsa_switch *ds, int port,
static int bcm_sf2_sw_set_wol(struct dsa_switch *ds, int port,
struct ethtool_wolinfo *wol)
{
- struct net_device *p = dsa_to_port(ds, port)->cpu_dp->master;
+ struct net_device *p = dsa_port_to_master(dsa_to_port(ds, port));
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
s8 cpu_port = dsa_to_port(ds, port)->cpu_dp->index;
struct ethtool_wolinfo pwol = { };
diff --git a/drivers/net/dsa/bcm_sf2_cfp.c b/drivers/net/dsa/bcm_sf2_cfp.c
index 22bc295bebdb..c4010b7bf089 100644
--- a/drivers/net/dsa/bcm_sf2_cfp.c
+++ b/drivers/net/dsa/bcm_sf2_cfp.c
@@ -1102,7 +1102,7 @@ static int bcm_sf2_cfp_rule_get_all(struct bcm_sf2_priv *priv,
int bcm_sf2_get_rxnfc(struct dsa_switch *ds, int port,
struct ethtool_rxnfc *nfc, u32 *rule_locs)
{
- struct net_device *p = dsa_to_port(ds, port)->cpu_dp->master;
+ struct net_device *p = dsa_port_to_master(dsa_to_port(ds, port));
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
int ret = 0;
@@ -1145,7 +1145,7 @@ int bcm_sf2_get_rxnfc(struct dsa_switch *ds, int port,
int bcm_sf2_set_rxnfc(struct dsa_switch *ds, int port,
struct ethtool_rxnfc *nfc)
{
- struct net_device *p = dsa_to_port(ds, port)->cpu_dp->master;
+ struct net_device *p = dsa_port_to_master(dsa_to_port(ds, port));
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
int ret = 0;
diff --git a/drivers/net/dsa/lan9303-core.c b/drivers/net/dsa/lan9303-core.c
index 9e04541c3144..438e46af03e9 100644
--- a/drivers/net/dsa/lan9303-core.c
+++ b/drivers/net/dsa/lan9303-core.c
@@ -1092,7 +1092,7 @@ static int lan9303_port_enable(struct dsa_switch *ds, int port,
if (!dsa_port_is_user(dp))
return 0;
- vlan_vid_add(dp->cpu_dp->master, htons(ETH_P_8021Q), port);
+ vlan_vid_add(dsa_port_to_master(dp), htons(ETH_P_8021Q), port);
return lan9303_enable_processing_port(chip, port);
}
@@ -1105,7 +1105,7 @@ static void lan9303_port_disable(struct dsa_switch *ds, int port)
if (!dsa_port_is_user(dp))
return;
- vlan_vid_del(dp->cpu_dp->master, htons(ETH_P_8021Q), port);
+ vlan_vid_del(dsa_port_to_master(dp), htons(ETH_P_8021Q), port);
lan9303_disable_processing_port(chip, port);
lan9303_phy_write(ds, chip->phy_addr_base + port, MII_BMCR, BMCR_PDOWN);
diff --git a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c
index 25dc3c3aa31d..5a1fc4bcd7a5 100644
--- a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c
+++ b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c
@@ -173,7 +173,7 @@ mtk_flow_get_dsa_port(struct net_device **dev)
if (dp->cpu_dp->tag_ops->proto != DSA_TAG_PROTO_MTK)
return -ENODEV;
- *dev = dp->cpu_dp->master;
+ *dev = dsa_port_to_master(dp);
return dp->index;
#else
diff --git a/include/net/dsa.h b/include/net/dsa.h
index f2ce12860546..23eac1bda843 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -718,6 +718,11 @@ static inline bool dsa_port_offloads_lag(struct dsa_port *dp,
return dsa_port_lag_dev_get(dp) == lag->dev;
}
+static inline struct net_device *dsa_port_to_master(const struct dsa_port *dp)
+{
+ return dp->cpu_dp->master;
+}
+
static inline
struct net_device *dsa_port_to_bridge_port(const struct dsa_port *dp)
{
diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
index 4bb0a203b85c..db34240bd0c5 100644
--- a/net/dsa/dsa2.c
+++ b/net/dsa/dsa2.c
@@ -1273,11 +1273,11 @@ int dsa_tree_change_tag_proto(struct dsa_switch_tree *dst,
* attempts to change the tagging protocol. If we ever lift the IFF_UP
* restriction, there needs to be another mutex which serializes this.
*/
- list_for_each_entry(dp, &dst->ports, list) {
- if (dsa_port_is_cpu(dp) && (dp->master->flags & IFF_UP))
+ dsa_tree_for_each_user_port(dp, dst) {
+ if (dsa_port_to_master(dp)->flags & IFF_UP)
goto out_unlock;
- if (dsa_port_is_user(dp) && (dp->slave->flags & IFF_UP))
+ if (dp->slave->flags & IFF_UP)
goto out_unlock;
}
@@ -1807,7 +1807,7 @@ void dsa_switch_shutdown(struct dsa_switch *ds)
rtnl_lock();
dsa_switch_for_each_user_port(dp, ds) {
- master = dp->cpu_dp->master;
+ master = dsa_port_to_master(dp);
slave_dev = dp->slave;
netdev_upper_dev_unlink(master, slave_dev);
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 614fbba8fe39..c48c5c8ba790 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -322,7 +322,7 @@ dsa_slave_to_master(const struct net_device *dev)
{
struct dsa_port *dp = dsa_slave_to_port(dev);
- return dp->cpu_dp->master;
+ return dsa_port_to_master(dp);
}
/* If under a bridge with vlan_filtering=0, make sure to send pvid-tagged
diff --git a/net/dsa/port.c b/net/dsa/port.c
index 7afc35db0c29..4183e60db4f9 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -1026,7 +1026,7 @@ int dsa_port_standalone_host_fdb_add(struct dsa_port *dp,
int dsa_port_bridge_host_fdb_add(struct dsa_port *dp,
const unsigned char *addr, u16 vid)
{
- struct dsa_port *cpu_dp = dp->cpu_dp;
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_db db = {
.type = DSA_DB_BRIDGE,
.bridge = *dp->bridge,
@@ -1037,8 +1037,8 @@ int dsa_port_bridge_host_fdb_add(struct dsa_port *dp,
* requires rtnl_lock(), since we can't guarantee that is held here,
* and we can't take it either.
*/
- if (cpu_dp->master->priv_flags & IFF_UNICAST_FLT) {
- err = dev_uc_add(cpu_dp->master, addr);
+ if (master->priv_flags & IFF_UNICAST_FLT) {
+ err = dev_uc_add(master, addr);
if (err)
return err;
}
@@ -1077,15 +1077,15 @@ int dsa_port_standalone_host_fdb_del(struct dsa_port *dp,
int dsa_port_bridge_host_fdb_del(struct dsa_port *dp,
const unsigned char *addr, u16 vid)
{
- struct dsa_port *cpu_dp = dp->cpu_dp;
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_db db = {
.type = DSA_DB_BRIDGE,
.bridge = *dp->bridge,
};
int err;
- if (cpu_dp->master->priv_flags & IFF_UNICAST_FLT) {
- err = dev_uc_del(cpu_dp->master, addr);
+ if (master->priv_flags & IFF_UNICAST_FLT) {
+ err = dev_uc_del(master, addr);
if (err)
return err;
}
@@ -1208,14 +1208,14 @@ int dsa_port_standalone_host_mdb_add(const struct dsa_port *dp,
int dsa_port_bridge_host_mdb_add(const struct dsa_port *dp,
const struct switchdev_obj_port_mdb *mdb)
{
- struct dsa_port *cpu_dp = dp->cpu_dp;
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_db db = {
.type = DSA_DB_BRIDGE,
.bridge = *dp->bridge,
};
int err;
- err = dev_mc_add(cpu_dp->master, mdb->addr);
+ err = dev_mc_add(master, mdb->addr);
if (err)
return err;
@@ -1252,14 +1252,14 @@ int dsa_port_standalone_host_mdb_del(const struct dsa_port *dp,
int dsa_port_bridge_host_mdb_del(const struct dsa_port *dp,
const struct switchdev_obj_port_mdb *mdb)
{
- struct dsa_port *cpu_dp = dp->cpu_dp;
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_db db = {
.type = DSA_DB_BRIDGE,
.bridge = *dp->bridge,
};
int err;
- err = dev_mc_del(cpu_dp->master, mdb->addr);
+ err = dev_mc_del(master, mdb->addr);
if (err)
return err;
@@ -1294,19 +1294,19 @@ int dsa_port_host_vlan_add(struct dsa_port *dp,
const struct switchdev_obj_port_vlan *vlan,
struct netlink_ext_ack *extack)
{
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_notifier_vlan_info info = {
.dp = dp,
.vlan = vlan,
.extack = extack,
};
- struct dsa_port *cpu_dp = dp->cpu_dp;
int err;
err = dsa_port_notify(dp, DSA_NOTIFIER_HOST_VLAN_ADD, &info);
if (err && err != -EOPNOTSUPP)
return err;
- vlan_vid_add(cpu_dp->master, htons(ETH_P_8021Q), vlan->vid);
+ vlan_vid_add(master, htons(ETH_P_8021Q), vlan->vid);
return err;
}
@@ -1314,18 +1314,18 @@ int dsa_port_host_vlan_add(struct dsa_port *dp,
int dsa_port_host_vlan_del(struct dsa_port *dp,
const struct switchdev_obj_port_vlan *vlan)
{
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_notifier_vlan_info info = {
.dp = dp,
.vlan = vlan,
};
- struct dsa_port *cpu_dp = dp->cpu_dp;
int err;
err = dsa_port_notify(dp, DSA_NOTIFIER_HOST_VLAN_DEL, &info);
if (err && err != -EOPNOTSUPP)
return err;
- vlan_vid_del(cpu_dp->master, htons(ETH_P_8021Q), vlan->vid);
+ vlan_vid_del(master, htons(ETH_P_8021Q), vlan->vid);
return err;
}
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index 345106b1ed78..55094b94a5ae 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -1503,8 +1503,7 @@ static int dsa_slave_setup_tc_block(struct net_device *dev,
static int dsa_slave_setup_ft_block(struct dsa_switch *ds, int port,
void *type_data)
{
- struct dsa_port *cpu_dp = dsa_to_port(ds, port)->cpu_dp;
- struct net_device *master = cpu_dp->master;
+ struct net_device *master = dsa_port_to_master(dsa_to_port(ds, port));
if (!master->netdev_ops->ndo_setup_tc)
return -EOPNOTSUPP;
@@ -2147,13 +2146,14 @@ static int dsa_slave_fill_forward_path(struct net_device_path_ctx *ctx,
struct net_device_path *path)
{
struct dsa_port *dp = dsa_slave_to_port(ctx->dev);
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_port *cpu_dp = dp->cpu_dp;
path->dev = ctx->dev;
path->type = DEV_PATH_DSA;
path->dsa.proto = cpu_dp->tag_ops->proto;
path->dsa.port = dp->index;
- ctx->dev = cpu_dp->master;
+ ctx->dev = master;
return 0;
}
@@ -2271,9 +2271,9 @@ static int dsa_slave_phy_setup(struct net_device *slave_dev)
void dsa_slave_setup_tagger(struct net_device *slave)
{
struct dsa_port *dp = dsa_slave_to_port(slave);
+ struct net_device *master = dsa_port_to_master(dp);
struct dsa_slave_priv *p = netdev_priv(slave);
const struct dsa_port *cpu_dp = dp->cpu_dp;
- struct net_device *master = cpu_dp->master;
const struct dsa_switch *ds = dp->ds;
slave->needed_headroom = cpu_dp->tag_ops->needed_headroom;
@@ -2330,8 +2330,7 @@ int dsa_slave_resume(struct net_device *slave_dev)
int dsa_slave_create(struct dsa_port *port)
{
- const struct dsa_port *cpu_dp = port->cpu_dp;
- struct net_device *master = cpu_dp->master;
+ struct net_device *master = dsa_port_to_master(port);
struct dsa_switch *ds = port->ds;
const char *name = port->name;
struct net_device *slave_dev;
diff --git a/net/dsa/tag_8021q.c b/net/dsa/tag_8021q.c
index b5f80bc45ceb..34e5ec5d3e23 100644
--- a/net/dsa/tag_8021q.c
+++ b/net/dsa/tag_8021q.c
@@ -330,7 +330,7 @@ static int dsa_tag_8021q_port_setup(struct dsa_switch *ds, int port)
if (!dsa_port_is_user(dp))
return 0;
- master = dp->cpu_dp->master;
+ master = dsa_port_to_master(dp);
err = dsa_port_tag_8021q_vlan_add(dp, vid, false);
if (err) {
@@ -359,7 +359,7 @@ static void dsa_tag_8021q_port_teardown(struct dsa_switch *ds, int port)
if (!dsa_port_is_user(dp))
return;
- master = dp->cpu_dp->master;
+ master = dsa_port_to_master(dp);
dsa_port_tag_8021q_vlan_del(dp, vid, false);
--
2.34.1
DSA now supports multiple CPU ports, explain the use cases that are
covered, the new UAPI, the permitted degrees of freedom, the driver API,
and remove some old "hanging fruits".
Signed-off-by: Vladimir Oltean <[email protected]>
---
v1->v2: provide alternate description of how to put CPU ports in a LAG
.../networking/dsa/configuration.rst | 96 +++++++++++++++++++
Documentation/networking/dsa/dsa.rst | 38 ++++++--
2 files changed, 128 insertions(+), 6 deletions(-)
diff --git a/Documentation/networking/dsa/configuration.rst b/Documentation/networking/dsa/configuration.rst
index 2b08f1a772d3..827701f8cbfe 100644
--- a/Documentation/networking/dsa/configuration.rst
+++ b/Documentation/networking/dsa/configuration.rst
@@ -49,6 +49,9 @@ In this documentation the following Ethernet interfaces are used:
*eth0*
the master interface
+*eth1*
+ another master interface
+
*lan1*
a slave interface
@@ -360,3 +363,96 @@ the ``self`` flag) has been removed. This results in the following changes:
Script writers are therefore encouraged to use the ``master static`` set of
flags when working with bridge FDB entries on DSA switch interfaces.
+
+Affinity of user ports to CPU ports
+-----------------------------------
+
+Typically, DSA switches are attached to the host via a single Ethernet
+interface, but in cases where the switch chip is discrete, the hardware design
+may permit the use of 2 or more ports connected to the host, for an increase in
+termination throughput.
+
+DSA can make use of multiple CPU ports in two ways. First, it is possible to
+statically assign the termination traffic associated with a certain user port
+to be processed by a certain CPU port. This way, user space can implement
+custom policies of static load balancing between user ports, by spreading the
+affinities according to the available CPU ports.
+
+Secondly, it is possible to perform load balancing between CPU ports on a per
+packet basis, rather than statically assigning user ports to CPU ports.
+This can be achieved by placing the DSA masters under a LAG interface (bonding
+or team). DSA monitors this operation and creates a mirror of this software LAG
+on the CPU ports facing the physical DSA masters that constitute the LAG slave
+devices.
+
+To make use of multiple CPU ports, the firmware (device tree) description of
+the switch must mark all the links between CPU ports and their DSA masters
+using the ``ethernet`` reference/phandle. At startup, only a single CPU port
+and DSA master will be used - the numerically first port from the firmware
+description which has an ``ethernet`` property. It is up to the user to
+configure the system for the switch to use other masters.
+
+DSA uses the ``rtnl_link_ops`` mechanism (with a "dsa" ``kind``) to allow
+changing the DSA master of a user port. The ``IFLA_DSA_MASTER`` u32 netlink
+attribute contains the ifindex of the master device that handles each slave
+device. The DSA master must be a valid candidate based on firmware node
+information, or a LAG interface which contains only slaves which are valid
+candidates.
+
+Using iproute2, the following manipulations are possible:
+
+ .. code-block:: sh
+
+ # See the DSA master in current use
+ ip -d link show dev swp0
+ (...)
+ dsa master eth0
+
+ # Static CPU port distribution
+ ip link set swp0 type dsa master eth1
+ ip link set swp1 type dsa master eth0
+ ip link set swp2 type dsa master eth1
+ ip link set swp3 type dsa master eth0
+
+ # CPU ports in LAG, using explicit assignment of the DSA master
+ ip link add bond0 type bond mode balance-xor && ip link set bond0 up
+ ip link set eth1 down && ip link set eth1 master bond0
+ ip link set swp0 type dsa master bond0
+ ip link set swp1 type dsa master bond0
+ ip link set swp2 type dsa master bond0
+ ip link set swp3 type dsa master bond0
+ ip link set eth0 down && ip link set eth0 master bond0
+ ip -d link show dev swp0
+ (...)
+ dsa master bond0
+
+ # CPU ports in LAG, relying on implicit migration of the DSA master
+ ip link add bond0 type bond mode balance-xor && ip link set bond0 up
+ ip link set eth0 down && ip link set eth0 master bond0
+ ip link set eth1 down && ip link set eth1 master bond0
+ ip -d link show dev swp0
+ (...)
+ dsa master bond0
+
+Notice that in the case of CPU ports under a LAG, the use of the
+``IFLA_DSA_MASTER`` netlink attribute is not strictly needed, but rather, DSA
+reacts to the ``IFLA_MASTER`` attribute change of its present master (``eth0``)
+and migrates all user ports to the new upper of ``eth0``, ``bond0``. Similarly,
+when ``bond0`` is destroyed using ``RTM_DELLINK``, DSA migrates the user ports
+that were assigned to this interface to the first physical DSA master which is
+eligible, based on the firmware description (it effectively reverts to the
+startup configuration).
+
+In a setup with more than 2 physical CPU ports, it is therefore possible to mix
+static user to CPU port assignment with LAG between DSA masters. It is not
+possible to statically assign a user port towards a DSA master that has any
+upper interfaces (this includes LAG devices - the master must always be the LAG
+in this case).
+
+Live changing of the DSA master (and thus CPU port) affinity of a user port is
+permitted, in order to allow dynamic redistribution in response to traffic.
+
+Physical DSA masters are allowed to join and leave at any time a LAG interface
+used as a DSA master; however, DSA will reject a LAG interface as a valid
+candidate for being a DSA master unless it has at least one physical DSA master
+as a slave device.
diff --git a/Documentation/networking/dsa/dsa.rst b/Documentation/networking/dsa/dsa.rst
index d742ba6bd211..a94ddf83348a 100644
--- a/Documentation/networking/dsa/dsa.rst
+++ b/Documentation/networking/dsa/dsa.rst
@@ -303,6 +303,20 @@ These frames are then queued for transmission using the master network device
Ethernet switch will be able to process these incoming frames from the
management interface and deliver them to the physical switch port.
+When using multiple CPU ports, it is possible to stack a LAG (bonding/team)
+device between the DSA slave devices and the physical DSA masters. The LAG
+device is thus also a DSA master, but the LAG slave devices continue to be DSA
+masters as well (just with no user port assigned to them; this is needed for
+recovery in case the LAG DSA master disappears). Thus, the data path of the LAG
+DSA master is used asymmetrically. On RX, the ``ETH_P_XDSA`` handler, which
+calls ``dsa_switch_rcv()``, is invoked early (on the physical DSA master;
+LAG slave). Therefore, the RX data path of the LAG DSA master is not used.
+On the other hand, TX takes place linearly: ``dsa_slave_xmit`` calls
+``dsa_enqueue_skb``, which calls ``dev_queue_xmit`` towards the LAG DSA master.
+The latter calls ``dev_queue_xmit`` towards one physical DSA master or the
+other, and in both cases, the packet exits the system through a hardware path
+towards the switch.
+
Graphical representation
------------------------
@@ -629,6 +643,24 @@ Switch configuration
PHY cannot be found. In this case, probing of the DSA switch continues
without that particular port.
+- ``port_change_master``: method through which the affinity (association used
+ for traffic termination purposes) between a user port and a CPU port can be
+ changed. By default all user ports from a tree are assigned to the first
+ available CPU port that makes sense for them (most of the times this means
+ the user ports of a tree are all assigned to the same CPU port, except for H
+ topologies as described in commit 2c0b03258b8b). The ``port`` argument
+ represents the index of the user port, and the ``master`` argument represents
+ the new DSA master ``net_device``. The CPU port associated with the new
+ master can be retrieved by looking at ``struct dsa_port *cpu_dp =
+ master->dsa_ptr``. Additionally, the master can also be a LAG device where
+ all the slave devices are physical DSA masters. LAG DSA masters also have a
+ valid ``master->dsa_ptr`` pointer, however this is not unique, but rather a
+ duplicate of the first physical DSA master's (LAG slave) ``dsa_ptr``. In case
+ of a LAG DSA master, a further call to ``port_lag_join`` will be emitted
+ separately for the physical CPU ports associated with the physical DSA
+ masters, requesting them to create a hardware LAG associated with the LAG
+ interface.
+
PHY devices and link management
-------------------------------
@@ -1095,9 +1127,3 @@ capable hardware, but does not enforce a strict switch device driver model. On
the other DSA enforces a fairly strict device driver model, and deals with most
of the switch specific. At some point we should envision a merger between these
two subsystems and get the best of both worlds.
-
-Other hanging fruits
---------------------
-
-- allowing more than one CPU/management interface:
- http://comments.gmane.org/gmane.linux.network/365657
--
2.34.1
Hello:
This series was applied to netdev/net-next.git (master)
by Paolo Abeni <[email protected]>:
On Sun, 11 Sep 2022 04:06:56 +0300 you wrote:
> Those who have been following part 1:
> https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
> part 2:
> https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
> and part 3:
> https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
> will know that I am trying to enable the second internal port pair from
> the NXP LS1028A Felix switch for DSA-tagged traffic via "ocelot-8021q".
>
> [...]
Here is the summary with links:
- [v2,net-next,01/10] net: introduce iterators over synced hw addresses
https://git.kernel.org/netdev/net-next/c/db01868bf2e9
- [v2,net-next,02/10] net: dsa: introduce dsa_port_get_master()
https://git.kernel.org/netdev/net-next/c/8f6a19c0316d
- [v2,net-next,03/10] net: dsa: allow the DSA master to be seen and changed through rtnetlink
https://git.kernel.org/netdev/net-next/c/95f510d0b792
- [v2,net-next,04/10] net: dsa: don't keep track of admin/oper state on LAG DSA masters
https://git.kernel.org/netdev/net-next/c/6e61b55c6d7f
- [v2,net-next,05/10] net: dsa: suppress appending ethtool stats to LAG DSA masters
https://git.kernel.org/netdev/net-next/c/cfeb84a52fcb
- [v2,net-next,06/10] net: dsa: suppress device links to LAG DSA masters
https://git.kernel.org/netdev/net-next/c/13eccc1bbb2e
- [v2,net-next,07/10] net: dsa: propagate extack to port_lag_join
https://git.kernel.org/netdev/net-next/c/2e359b00a117
- [v2,net-next,08/10] net: dsa: allow masters to join a LAG
https://git.kernel.org/netdev/net-next/c/acc43b7bf52a
- [v2,net-next,09/10] docs: net: dsa: update information about multiple CPU ports
https://git.kernel.org/netdev/net-next/c/0773e3a851c8
- [v2,net-next,10/10] net: dsa: felix: add support for changing DSA master
https://git.kernel.org/netdev/net-next/c/eca70102cfb1
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html