On Thu, 29 Oct 2020 15:53:21 +0100 Eelco Chaudron wrote:
> Silence suspicious RCU usage warning in ovs_flow_tbl_masks_cache_resize()
> by replacing rcu_dereference() with rcu_dereference_ovsl().
> 
> In addition, when creating a new datapath, make sure it's configured under
> the ovs_lock.
> 
> Fixes: 9bf24f594c6a ("net: openvswitch: make masks cache size configurable")
> Reported-by: syzbot+9a8f8bfcc56e85780...@syzkaller.appspotmail.com
> Signed-off-by: Eelco Chaudron <echau...@redhat.com>
> ---
>  net/openvswitch/datapath.c   |    8 ++++----
>  net/openvswitch/flow_table.c |    2 +-
>  2 files changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
> index 832f898edb6a..020f8539fede 100644
> --- a/net/openvswitch/datapath.c
> +++ b/net/openvswitch/datapath.c
> @@ -1695,6 +1695,9 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct 
> genl_info *info)
>       if (err)
>               goto err_destroy_ports;
>  
> +     /* So far only local changes have been made, now need the lock. */
> +     ovs_lock();

Should we move the lock below assignments to param?

Looks a little strange to protect stack variables with a global lock.

>       /* Set up our datapath device. */
>       parms.name = nla_data(a[OVS_DP_ATTR_NAME]);
>       parms.type = OVS_VPORT_TYPE_INTERNAL;
> @@ -1707,9 +1710,6 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct 
> genl_info *info)
>       if (err)
>               goto err_destroy_meters;
>  
> -     /* So far only local changes have been made, now need the lock. */
> -     ovs_lock();
> -
>       vport = new_vport(&parms);
>       if (IS_ERR(vport)) {
>               err = PTR_ERR(vport);
> @@ -1725,7 +1725,6 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct 
> genl_info *info)
>                               ovs_dp_reset_user_features(skb, info);
>               }
>  
> -             ovs_unlock();
>               goto err_destroy_meters;
>       }
>  
> @@ -1742,6 +1741,7 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct 
> genl_info *info)
>       return 0;
>  
>  err_destroy_meters:

Let's update the name of the label.

> +     ovs_unlock();
>       ovs_meters_exit(dp);
>  err_destroy_ports:
>       kfree(dp->ports);
> diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
> index f3486a37361a..c89c8da99f1a 100644
> --- a/net/openvswitch/flow_table.c
> +++ b/net/openvswitch/flow_table.c
> @@ -390,7 +390,7 @@ static struct mask_cache *tbl_mask_cache_alloc(u32 size)
>  }
>  int ovs_flow_tbl_masks_cache_resize(struct flow_table *table, u32 size)
>  {
> -     struct mask_cache *mc = rcu_dereference(table->mask_cache);
> +     struct mask_cache *mc = rcu_dereference_ovsl(table->mask_cache);
>       struct mask_cache *new;
>  
>       if (size == mc->cache_size)
> 

_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to