[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <aAjbLNUErWRPa03b@shredder>
Date: Wed, 23 Apr 2025 15:21:00 +0300
From: Ido Schimmel <idosch@...dia.com>
To: Paolo Abeni <pabeni@...hat.com>
Cc: netdev@...r.kernel.org, davem@...emloft.net, kuba@...nel.org,
edumazet@...gle.com, andrew+netdev@...n.ch, horms@...nel.org,
petrm@...dia.com, razor@...ckwall.org
Subject: Re: [PATCH net-next 12/15] vxlan: Create wrappers for FDB lookup
On Tue, Apr 22, 2025 at 10:46:01AM +0200, Paolo Abeni wrote:
> On 4/15/25 2:11 PM, Ido Schimmel wrote:
> > @@ -1286,7 +1300,7 @@ int __vxlan_fdb_delete(struct vxlan_dev *vxlan,
> > struct vxlan_fdb *f;
> > int err = -ENOENT;
> >
> > - f = __vxlan_find_mac(vxlan, addr, src_vni);
> > + f = vxlan_find_mac(vxlan, addr, src_vni);
>
> Minor note for a possible follow-up (not blocking this series): AFAICS
> the above is safe because the caller held the hash lock (otherwise f can
> be released as soon as vxlan_find_mac() returns). It could be possibly
> useful to make the code more readable a vxlan_find_mac_locked() variant
> do to this lookup without the RCU lock and with the proper lockdep
> annotation.
Thanks for the feedback, but I'm not sure I understand what you are
asking for. vxlan_find_mac() expects to be called with the hash lock
held and there's a lockdep annotation there to make sure the lock is
held. You want me to rename vxlan_find_mac() to vxlan_find_mac_locked()?
>
> I think even the previous lookup could use a similar helper.
>
> Cheers,
>
> Paolo
>
Powered by blists - more mailing lists