[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230116090837.105b42a9@fixe.home>
Date: Mon, 16 Jan 2023 09:08:37 +0100
From: Clément Léger <clement.leger@...tlin.com>
To: Vladimir Oltean <olteanv@...il.com>
Cc: Andrew Lunn <andrew@...n.ch>, Jakub Kicinski <kuba@...nel.org>,
Florian Fainelli <f.fainelli@...il.com>,
"David S. Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>,
Paolo Abeni <pabeni@...hat.com>,
Russell King <linux@...linux.org.uk>,
linux-renesas-soc@...r.kernel.org, netdev@...r.kernel.org,
linux-kernel@...r.kernel.org,
Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
Herve Codina <herve.codina@...tlin.com>,
Miquèl Raynal <miquel.raynal@...tlin.com>,
Milan Stevanovic <milan.stevanovic@...com>,
Jimmy Lalande <jimmy.lalande@...com>,
Pascal Eberhard <pascal.eberhard@...com>
Subject: Re: [PATCH net-next] net: dsa: rzn1-a5psw: Add vlan support
Le Fri, 13 Jan 2023 17:37:30 +0200,
Vladimir Oltean <olteanv@...il.com> a écrit :
> On Fri, Jan 13, 2023 at 03:12:40PM +0100, Andrew Lunn wrote:
> > On Thu, Jan 12, 2023 at 09:37:55PM -0800, Jakub Kicinski wrote:
> > > On Wed, 11 Jan 2023 12:56:07 +0100 Clément Léger wrote:
> > > > Add support for vlan operation (add, del, filtering) on the RZN1
> > > > driver. The a5psw switch supports up to 32 VLAN IDs with filtering,
> > > > tagged/untagged VLANs and PVID for each ports.
> > >
> > > noob question - do you need that mutex?
> > > aren't those ops all under rtnl_lock?
> >
> > Hi Jakub
> >
> > Not commenting about this specific patch, but not everything in DSA is
> > done under RTNL. So you need to deal with some parallel API calls. But
> > they tend to be in different areas. I would not expect to see two VLAN
> > changes as the same time, but maybe VLAN and polling in a workqueue to
> > update the statistics for example could happen. Depending on the
> > switch, some protect might be needed to stop these operations
> > interfering with each other. And DSA drivers in general tend to KISS
> > and over lock. Nothing here is particularly hot path, the switch
> > itself is on the end of a slow bus, so the overhead of locks are
> > minimum.
>
> That being said, port_vlan_add(), port_vlan_del() and port_vlan_filtering()
> are all serialized by the rtnl_lock().
Ok then, I'll remove this lock and rely on the RTNL lock.
Thanks,
--
Clément Léger,
Embedded Linux and Kernel engineer at Bootlin
https://bootlin.com
Powered by blists - more mailing lists