[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250411141847.6dba6987@kernel.org>
Date: Fri, 11 Apr 2025 14:18:47 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Sabrina Dubroca <sd@...asysnail.net>
Cc: Antonio Quartulli <antonio@...nvpn.net>, netdev@...r.kernel.org, Eric
Dumazet <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>, Donald
Hunter <donald.hunter@...il.com>, Shuah Khan <shuah@...nel.org>,
ryazanov.s.a@...il.com, Andrew Lunn <andrew+netdev@...n.ch>, Simon Horman
<horms@...nel.org>, linux-kernel@...r.kernel.org,
linux-kselftest@...r.kernel.org, Xiao Liang <shaw.leon@...il.com>,
steffen.klassert@...unet.com, antony.antony@...unet.com
Subject: Re: [PATCH net-next v25 01/23] net: introduce OpenVPN Data Channel
Offload (ovpn)
On Fri, 11 Apr 2025 15:50:49 +0200 Sabrina Dubroca wrote:
> > My understanding is that this is the standard approach to:
> > 1) hook in the middle of registration/deregistration;
> > 2) handle events generated by other components/routines.
> >
> > I see in /drivers/net/ almost every driver registers a notifier for their
> > own device.
>
> I think most of them register a notifier for their lower device
> (bridge port, real device under a vlan, or similar).
>
> I've mentioned at some point that it would be more usual to replace
> this notifier with a custom dellink, and that ovpn->registered could
> likely be replaced with checking for NETREG_REGISTERED. I just thought
> it could be cleaned up a bit later, but it seems Jakub wants it done
> before taking the patches :)
Ideally, yes. One fewer place for us to check when trying to figure
out if we will break anything with the locking changes :(
Notifiers are very powerful but that comes at high maintenance cost.
Powered by blists - more mailing lists