Mon, Oct 12, 2015 at 02:41:08PM IDT, ra...@blackwall.org wrote:
>From: Nikolay Aleksandrov <niko...@cumulusnetworks.com>
>
Hi,
>Ido Schimmel reported a problem with switchdev devices because of the
>order change of del_nbp operations, more specifically the move of
>nbp_vlan_flush() which deletes all vlans and frees vlgrp after the
>rx_handler has been unregistered. So in order to fix this break
>vlan_flush in two phases:
>1. delete all of vlan_group's vlans
>2. destroy rhtable and free vlgrp
>We execute phase I (free_rht == false) in the same place as before so the
>vlans can be cleared and free the vlgrp after the rx_handler has been
>unregistered in phase II (free_rht == true).
I don't fully understand the reason for the two-phase cleanup. Please
see below.
>
>Reported-by: Ido Schimmel <ido...@mellanox.com>
>Signed-off-by: Nikolay Aleksandrov <niko...@cumulusnetworks.com>
>---
>Ido: I hope this fixes it for your case, a test would be much appreciated.
This indeed fixes our switchdev issue. Thanks for the fix!
>
> net/bridge/br_if.c | 11 ++++++++---
> net/bridge/br_private.h | 8 ++++----
> net/bridge/br_vlan.c | 17 ++++++++++-------
> 3 files changed, 22 insertions(+), 14 deletions(-)
>
>diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
>index 934cae9fa317..74a03c0a4e5f 100644
>--- a/net/bridge/br_if.c
>+++ b/net/bridge/br_if.c
>@@ -248,6 +248,8 @@ static void del_nbp(struct net_bridge_port *p)
>
> list_del_rcu(&p->list);
>
>+ /* vlan_flush phase I: remove vlans */
>+ nbp_vlan_flush(p, false);
> br_fdb_delete_by_port(br, p, 0, 1);
> nbp_update_port_count(br);
>
>@@ -256,8 +258,10 @@ static void del_nbp(struct net_bridge_port *p)
> dev->priv_flags &= ~IFF_BRIDGE_PORT;
>
> netdev_rx_handler_unregister(dev);
>- /* use the synchronize_rcu done by netdev_rx_handler_unregister */
>- nbp_vlan_flush(p);
>+ /* use the synchronize_rcu done by netdev_rx_handler_unregister
>+ * vlan_flush phase II: free rht and vlgrp
>+ */
>+ nbp_vlan_flush(p, true);
>
> br_multicast_del_port(p);
>
>@@ -281,7 +285,8 @@ void br_dev_delete(struct net_device *dev, struct
>list_head *head)
>
> br_fdb_delete_by_port(br, NULL, 0, 1);
>
>- br_vlan_flush(br);
>+ /* vlan_flush execute both phases (see del_nbp) */
>+ br_vlan_flush(br, true);
> br_multicast_dev_del(br);
> del_timer_sync(&br->gc_timer);
>
>diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
>index 7d14ba93bba4..3938a976417f 100644
>--- a/net/bridge/br_private.h
>+++ b/net/bridge/br_private.h
>@@ -682,7 +682,7 @@ struct sk_buff *br_handle_vlan(struct net_bridge *br,
> struct sk_buff *skb);
> int br_vlan_add(struct net_bridge *br, u16 vid, u16 flags);
> int br_vlan_delete(struct net_bridge *br, u16 vid);
>-void br_vlan_flush(struct net_bridge *br);
>+void br_vlan_flush(struct net_bridge *br, bool free_rht);
> struct net_bridge_vlan *br_vlan_find(struct net_bridge_vlan_group *vg, u16
> vid);
> void br_recalculate_fwd_mask(struct net_bridge *br);
> int __br_vlan_filter_toggle(struct net_bridge *br, unsigned long val);
>@@ -694,7 +694,7 @@ int br_vlan_set_default_pvid(struct net_bridge *br,
>unsigned long val);
> int __br_vlan_set_default_pvid(struct net_bridge *br, u16 pvid);
> int nbp_vlan_add(struct net_bridge_port *port, u16 vid, u16 flags);
> int nbp_vlan_delete(struct net_bridge_port *port, u16 vid);
>-void nbp_vlan_flush(struct net_bridge_port *port);
>+void nbp_vlan_flush(struct net_bridge_port *port, bool free_rht);
> int nbp_vlan_init(struct net_bridge_port *port);
> int nbp_get_num_vlan_infos(struct net_bridge_port *p, u32 filter_mask);
>
>@@ -790,7 +790,7 @@ static inline int br_vlan_delete(struct net_bridge *br,
>u16 vid)
> return -EOPNOTSUPP;
> }
>
>-static inline void br_vlan_flush(struct net_bridge *br)
>+static inline void br_vlan_flush(struct net_bridge *br, bool free_rht)
> {
> }
>
>@@ -813,7 +813,7 @@ static inline int nbp_vlan_delete(struct net_bridge_port
>*port, u16 vid)
> return -EOPNOTSUPP;
> }
>
>-static inline void nbp_vlan_flush(struct net_bridge_port *port)
>+static inline void nbp_vlan_flush(struct net_bridge_port *port, bool free_rht)
> {
> }
>
>diff --git a/net/bridge/br_vlan.c b/net/bridge/br_vlan.c
>index a212979257b1..4fb9b23c9838 100644
>--- a/net/bridge/br_vlan.c
>+++ b/net/bridge/br_vlan.c
>@@ -307,15 +307,18 @@ out:
> return err;
> }
>
>-static void __vlan_flush(struct net_bridge_vlan_group *vlgrp)
>+static void __vlan_flush(struct net_bridge_vlan_group *vlgrp, bool free_rht)
> {
> struct net_bridge_vlan *vlan, *tmp;
>
> __vlan_delete_pvid(vlgrp, vlgrp->pvid);
> list_for_each_entry_safe(vlan, tmp, &vlgrp->vlan_list, vlist)
> __vlan_del(vlan);
>- rhashtable_destroy(&vlgrp->vlan_hash);
>- kfree_rcu(vlgrp, rcu);
>+
Why not just issue a synchronize_rcu here and remove the if statement? I
believe that would also be better for when we remove the bridge device
itself. It's fully symmetric with init that way.
>+ if (free_rht) {
>+ rhashtable_destroy(&vlgrp->vlan_hash);
>+ kfree_rcu(vlgrp, rcu);
>+ }
> }
>
> struct sk_buff *br_handle_vlan(struct net_bridge *br,
>@@ -569,11 +572,11 @@ int br_vlan_delete(struct net_bridge *br, u16 vid)
> return __vlan_del(v);
> }
>
>-void br_vlan_flush(struct net_bridge *br)
>+void br_vlan_flush(struct net_bridge *br, bool free_rht)
> {
> ASSERT_RTNL();
>
>- __vlan_flush(br_vlan_group(br));
>+ __vlan_flush(br_vlan_group(br), free_rht);
> }
>
> struct net_bridge_vlan *br_vlan_find(struct net_bridge_vlan_group *vg, u16
> vid)
>@@ -957,7 +960,7 @@ int nbp_vlan_delete(struct net_bridge_port *port, u16 vid)
> return __vlan_del(v);
> }
>
>-void nbp_vlan_flush(struct net_bridge_port *port)
>+void nbp_vlan_flush(struct net_bridge_port *port, bool free_rht)
> {
> struct net_bridge_vlan_group *vg;
> struct net_bridge_vlan *vlan;
>@@ -968,5 +971,5 @@ void nbp_vlan_flush(struct net_bridge_port *port)
> list_for_each_entry(vlan, &vg->vlan_list, vlist)
> vlan_vid_del(port->dev, port->br->vlan_proto, vlan->vid);
>
>- __vlan_flush(vg);
>+ __vlan_flush(vg, free_rht);
> }
>--
>2.4.3
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
