netdev - Re: [PATCH net-next 09/13] net: dsa: lantiq_gswip: Forbid gswip_add_single_port

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Fri, 7 Jun 2024 16:49:35 +0300
From: Vladimir Oltean <olteanv@...il.com>
To: Martin Schiller <ms@....tdt.de>
Cc: martin.blumenstingl@...glemail.com, hauke@...ke-m.de, andrew@...n.ch,
	f.fainelli@...il.com, davem@...emloft.net, edumazet@...gle.com,
	kuba@...nel.org, pabeni@...hat.com, robh@...nel.org,
	krzk+dt@...nel.org, conor+dt@...nel.org, netdev@...r.kernel.org,
	devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next 09/13] net: dsa: lantiq_gswip: Forbid
 gswip_add_single_port_br on the CPU port

On Fri, Jun 07, 2024 at 03:31:57PM +0200, Martin Schiller wrote:
> On 2024-06-07 13:26, Vladimir Oltean wrote:
> > On Thu, Jun 06, 2024 at 10:52:30AM +0200, Martin Schiller wrote:
> > > From: Martin Blumenstingl <martin.blumenstingl@...glemail.com>
> > > 
> > > Calling gswip_add_single_port_br() with the CPU port would be a bug
> > > because then only the CPU port could talk to itself. Add the CPU
> > > port to
> > > the validation at the beginning of gswip_add_single_port_br().
> > > 
> > > Signed-off-by: Martin Blumenstingl
> > > <martin.blumenstingl@...glemail.com>
> > > ---
> > >  drivers/net/dsa/lantiq_gswip.c | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/net/dsa/lantiq_gswip.c
> > > b/drivers/net/dsa/lantiq_gswip.c
> > > index ee8296d5b901..d2195271ffe9 100644
> > > --- a/drivers/net/dsa/lantiq_gswip.c
> > > +++ b/drivers/net/dsa/lantiq_gswip.c
> > > @@ -657,7 +657,7 @@ static int gswip_add_single_port_br(struct
> > > gswip_priv *priv, int port, bool add)
> > >  	unsigned int max_ports = priv->hw_info->max_ports;
> > >  	int err;
> > > 
> > > -	if (port >= max_ports) {
> > > +	if (port >= max_ports || dsa_is_cpu_port(priv->ds, port)) {
> > >  		dev_err(priv->dev, "single port for %i supported\n", port);
> > >  		return -EIO;
> > >  	}
> > > --
> > > 2.39.2
> > > 
> > 
> > Isn't the new check effectively dead code?
> 
> As long as the dsa_switch_ops .port_bridge_join and .port_bridge_leave are not
> executed for the cpu port, I agree with you.

They aren't. The primary trigger for dsa_port_bridge_join() is dsa_user_changeupper(),
along with other code paths that replay the operation in certain circumstances,
again only for user ports.