[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210629113557.eu5it6lfsnnaioej@skbuf>
Date: Tue, 29 Jun 2021 14:35:57 +0300
From: Vladimir Oltean <olteanv@...il.com>
To: Nikolay Aleksandrov <nikolay@...dia.com>
Cc: netdev@...r.kernel.org, Andrew Lunn <andrew@...n.ch>,
Florian Fainelli <f.fainelli@...il.com>,
Vivien Didelot <vivien.didelot@...il.com>,
Jiri Pirko <jiri@...nulli.us>,
Ido Schimmel <idosch@...sch.org>,
Tobias Waldekranz <tobias@...dekranz.com>,
Roopa Prabhu <roopa@...dia.com>,
Vladimir Oltean <vladimir.oltean@....com>
Subject: Re: [PATCH v4 net-next 01/14] net: bridge: switchdev: send FDB
notifications for host addresses
On Tue, Jun 29, 2021 at 01:40:20PM +0300, Nikolay Aleksandrov wrote:
> > @@ -117,18 +118,16 @@ br_switchdev_fdb_notify(const struct net_bridge_fdb_entry *fdb, int type)
> > .is_local = test_bit(BR_FDB_LOCAL, &fdb->flags),
> > .offloaded = test_bit(BR_FDB_OFFLOADED, &fdb->flags),
> > };
> > -
> > - if (!fdb->dst)
> > - return;
> > + struct net_device *dev = fdb->dst ? fdb->dst->dev : br->dev;
>
> you should use READ_ONCE() for fdb->dst here to make sure it's read only once,
> to be fair the old code had the same issue :)
Thanks for the comment. I still have budget for one patch until I hit
the 15 limit, so I guess I'll do that separately before this one.
Just trying to make sure I get it right. You want me to annotate
fdb_create(), br_fdb_update(), fdb_add_entry() and
br_fdb_external_learn_add() with WRITE_ONCE() too, right?
Can I resend right away or did you notice other issues in the other
patches?
Powered by blists - more mailing lists