netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Vlad Yasevich <vyasevic@redhat.com>
To: "Michał Mirosław" <mirqus@gmail.com>
Cc: shemminger@vyatta.com, bridge@lists.linux-foundation.org,
	davem@davemloft.net, netdev@vger.kernel.org
Subject: Re: [PATCH v9 net-next 01/12] bridge: Add vlan filtering infrastructure
Date: Fri, 01 Feb 2013 21:11:39 -0500	[thread overview]
Message-ID: <510C75DB.5010801@redhat.com> (raw)
In-Reply-To: <CAHXqBFJfPYwnxbeDR6AfG0-PQJ__=rKMxQWRcFJGseGiXyuYGg@mail.gmail.com>

On 02/01/2013 08:04 PM, Michał Mirosław wrote:
> 2013/2/1 Vlad Yasevich <vyasevic@redhat.com>:
>> Adds an optional infrustructure component to bridge that would allow
>> native vlan filtering in the bridge.  Each bridge port (as well
>> as the bridge device) now get a VLAN bitmap.  Each bit in the bitmap
>> is associated with a vlan id.  This way if the bit corresponding to
>> the vid is set in the bitmap that the packet with vid is allowed to
>> enter and exit the port.
> [...]
> +struct net_port_vlans {
> +       u16                             port_idx;
> +       void                            *parent;
> +       struct rcu_head                 rcu;
> +       unsigned long                   vlan_bitmap[BR_VLAN_BITMAP_LEN];
> +};
> [later, in br_vlan.c]
> +#define vlans_to_parent(type, pv) ((type *)(pv)->parent)
>
> A really don't like this pointer casting. It's easy to get it wrong
> (and you did in __vlan_del).
>
> 1. union { struct net_bridge *br; struct net_bridge_port *port; } parent;
>
> 2. inline net_port_vlans __rcu **br_vlan_info_parent_ptr(struct
> net_port_vlans *v)
> {
> 	return v->port_idx ? &v->parent.port->vlan_info : &v->parent.br->vlan_info;
> }
>
> (I'm not insisting on #2. Just think about it. ;)

I need the parent, not the vlan_info.  Sometimes I need to get to the 
device and I really don't want to then use container_of().

I suppose I could do the union, but that's really not much different 
then the type casts.  Just as easy to goof.

>
>> @@ -156,6 +166,7 @@ struct net_bridge_port
>>   #ifdef CONFIG_NET_POLL_CONTROLLER
>>          struct netpoll                  *np;
>>   #endif
>> +       struct net_port_vlans __rcu     *vlan_info;
>>   };
>
> Missing #ifdef CONFIG_BRIDGE_VLAN_FILTERING?

Yep.  Its in net_bridge, but not here.  Will fix.

>
>> +static int __vlan_del(struct net_port_vlans *v, u16 vid)
>> +{
>> +       unsigned long first_bit;
>> +       unsigned long last_bit;
>> +
>> +       if (!test_bit(vid, v->vlan_bitmap))
>> +               return -EINVAL;
>> +
>> +       /* Check to see if any other vlans are in this table.  If this
>> +        * is the last vlan, delete the whole structure.  If this is not the
>> +        * last vlan, just clear the bit.
>> +        */
>> +       first_bit = find_first_bit(v->vlan_bitmap, BR_VLAN_BITMAP_LEN);
>> +       last_bit = find_last_bit(v->vlan_bitmap, BR_VLAN_BITMAP_LEN);
>> +
>> +       if (v->port_idx && vid) {
>> +               struct net_device *dev =
>> +                       vlans_to_parent(struct net_bridge, v)->dev;
>
> struct net_bridge_port (or union and v->parent.port)

Yep.  Typo when doing conversions.

>
>> +
>> +               if (dev->features & NETIF_F_HW_VLAN_FILTER)
>> +                       dev->netdev_ops->ndo_vlan_rx_kill_vid(dev, vid);
>> +       }
>> +
>> +       clear_bit(vid, v->vlan_bitmap);
>> +       if (first_bit == last_bit) {
>
> bitmap_empty() is faster than two times find_xxx_bit().
>

WTH?  I did this change and now its gone...  Time to see what else I 
somehow lost...

>> +               if (v->port_idx) {
>> +                       struct net_bridge_port *p =
>> +                               vlans_to_parent(struct net_bridge_port, v);
>> +                       rcu_assign_pointer(p->vlan_info, NULL);
>> +               } else {
>> +                       struct net_bridge *br =
>> +                               vlans_to_parent(struct net_bridge, v);
>> +                       rcu_assign_pointer(br->vlan_info, NULL);
>> +               }
>
> With inline function above:
>
> rcu_assign_pointer(*br_vlan_info_parent_ptr(v), NULL);
>
>> +               kfree_rcu(v, rcu);
>> +       }
>> +       return 0;
>> +}
>> +
>> +static void __vlan_flush(struct net_port_vlans *v)
>> +{
>> +       bitmap_zero(v->vlan_bitmap, BR_VLAN_BITMAP_LEN);
>> +       if (v->port_idx) {
>> +               struct net_bridge_port *p =
>> +                               vlans_to_parent(struct net_bridge_port, v);
>> +               rcu_assign_pointer(p->vlan_info, NULL);
>> +       } else {
>> +               struct net_bridge *br =
>> +                               vlans_to_parent(struct net_bridge, v);
>> +               rcu_assign_pointer(br->vlan_info, NULL);
>> +       }
>
> Same here.

Thanks.  I'll probably go with the union for direct reference.

-vlad
>
>> +       kfree_rcu(v, rcu);
>> +}
>
> Best Regards,
> Michał Mirosław
>

  reply	other threads:[~2013-02-02  2:11 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-02-01 20:01 [PATCH v9 net-next 00/12] Add basic VLAN support to bridges Vlad Yasevich
2013-02-01 20:01 ` [PATCH v9 net-next 01/12] bridge: Add vlan filtering infrastructure Vlad Yasevich
2013-02-02  1:04   ` Michał Mirosław
2013-02-02  2:11     ` Vlad Yasevich [this message]
2013-02-01 20:02 ` [PATCH v9 net-next 02/12] bridge: Validate that vlan is permitted on ingress Vlad Yasevich
2013-02-02  1:04   ` Michał Mirosław
2013-02-02  2:13     ` Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 03/12] bridge: Verify that a vlan is allowed to egress on give port Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 04/12] bridge: Add netlink interface to configure vlans on bridge ports Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 05/12] bridge: Add the ability to configure pvid Vlad Yasevich
2013-02-02  1:04   ` Michał Mirosław
2013-02-02  1:15     ` Michał Mirosław
2013-02-02  2:22       ` Vlad Yasevich
2013-02-03 14:49         ` Michał Mirosław
2013-02-03 15:20           ` Michał Mirosław
2013-02-04 16:59           ` Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 06/12] bridge: Implement vlan ingress/egress policy Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 07/12] bridge: Add vlan to unicast fdb entries Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 08/12] bridge: Add vlan id to multicast groups Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 09/12] bridge: Add vlan support to static neighbors Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 10/12] bridge: Add vlan support for local fdb entries Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 11/12] bridge: Dump vlan information from a bridge port Vlad Yasevich
2013-02-01 20:02 ` [PATCH v9 net-next 12/12] bridge: Separate egress policy bitmap Vlad Yasevich
2013-02-04 16:24 ` [PATCH v9 net-next 00/12] Add basic VLAN support to bridges Stephen Hemminger
2013-02-04 16:58   ` Vlad Yasevich
2013-02-07 22:48     ` Vlad Yasevich
2013-02-07 22:57       ` Stephen Hemminger
2013-02-07 23:00         ` Vlad Yasevich

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=510C75DB.5010801@redhat.com \
    --to=vyasevic@redhat.com \
    --cc=bridge@lists.linux-foundation.org \
    --cc=davem@davemloft.net \
    --cc=mirqus@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=shemminger@vyatta.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).