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]
Message-ID: <64c2bff9.050a0220.49b7e.cb84@mx.google.com>
Date:   Thu, 27 Jul 2023 21:05:28 +0200
From:   Christian Marangi <ansuelsmth@...il.com>
To:     Vladimir Oltean <olteanv@...il.com>
Cc:     Andrew Lunn <andrew@...n.ch>,
        Florian Fainelli <f.fainelli@...il.com>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>, Atin Bainada <hi@...nb.me>,
        netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [net-next PATCH 2/3] net: dsa: qca8k: make learning configurable
 and keep off if standalone

On Wed, Jul 26, 2023 at 03:12:13PM +0300, Vladimir Oltean wrote:
> On Mon, Jul 24, 2023 at 05:30:57AM +0200, Christian Marangi wrote:
> > Address learning should initially be turned off by the driver for port
> > operation in standalone mode, then the DSA core handles changes to it
> > via ds->ops->port_bridge_flags().
> > 
> > Currently this is not the case for qca8k where learning is enabled
> > unconditionally in qca8k_setup for every user port.
> > 
> > Handle ports configured in standalone mode by making the learning
> > configurable and not enabling it by default.
> > 
> > Implement .port_pre_bridge_flags and .port_bridge_flags dsa ops to
> > enable learning for bridge that request it and tweak
> > .port_stp_state_set to correctly disable learning when port is
> > configured in standalone mode.
> > 
> > Signed-off-by: Christian Marangi <ansuelsmth@...il.com>
> > ---
> 
> Reviewed-by: Vladimir Oltean <olteanv@...il.com>
> 
> Small nitpick below.
> 
> >  drivers/net/dsa/qca/qca8k-8xxx.c   |  7 +++--
> >  drivers/net/dsa/qca/qca8k-common.c | 44 ++++++++++++++++++++++++++++++
> >  drivers/net/dsa/qca/qca8k.h        |  6 ++++
> >  3 files changed, 54 insertions(+), 3 deletions(-)
> > 
> > diff --git a/drivers/net/dsa/qca/qca8k-8xxx.c b/drivers/net/dsa/qca/qca8k-8xxx.c
> > index ae088a4df794..31552853fdd4 100644
> > --- a/drivers/net/dsa/qca/qca8k-8xxx.c
> > +++ b/drivers/net/dsa/qca/qca8k-8xxx.c
> > @@ -1870,9 +1870,8 @@ qca8k_setup(struct dsa_switch *ds)
> >  			if (ret)
> >  				return ret;
> >  
> > -			/* Enable ARP Auto-learning by default */
> > -			ret = regmap_set_bits(priv->regmap, QCA8K_PORT_LOOKUP_CTRL(i),
> > -					      QCA8K_PORT_LOOKUP_LEARN);
> > +			ret = regmap_clear_bits(priv->regmap, QCA8K_PORT_LOOKUP_CTRL(i),
> > +						QCA8K_PORT_LOOKUP_LEARN);
> >  			if (ret)
> >  				return ret;
> >  
> > @@ -1978,6 +1977,8 @@ static const struct dsa_switch_ops qca8k_switch_ops = {
> >  	.port_change_mtu	= qca8k_port_change_mtu,
> >  	.port_max_mtu		= qca8k_port_max_mtu,
> >  	.port_stp_state_set	= qca8k_port_stp_state_set,
> > +	.port_pre_bridge_flags	= qca8k_port_pre_bridge_flags,
> > +	.port_bridge_flags	= qca8k_port_bridge_flags,
> >  	.port_bridge_join	= qca8k_port_bridge_join,
> >  	.port_bridge_leave	= qca8k_port_bridge_leave,
> >  	.port_fast_age		= qca8k_port_fast_age,
> > diff --git a/drivers/net/dsa/qca/qca8k-common.c b/drivers/net/dsa/qca/qca8k-common.c
> > index 13b8452ce5b2..e53694d2852a 100644
> > --- a/drivers/net/dsa/qca/qca8k-common.c
> > +++ b/drivers/net/dsa/qca/qca8k-common.c
> > @@ -565,9 +565,26 @@ int qca8k_get_mac_eee(struct dsa_switch *ds, int port,
> >  	return 0;
> >  }
> >  
> > +static int qca8k_port_configure_learning(struct dsa_switch *ds, int port,
> > +					 bool learning)
> > +{
> > +	struct qca8k_priv *priv = ds->priv;
> > +
> > +	if (learning)
> > +		return regmap_set_bits(priv->regmap,
> > +				       QCA8K_PORT_LOOKUP_CTRL(port),
> > +				       QCA8K_PORT_LOOKUP_LEARN);
> > +	else
> > +		return regmap_clear_bits(priv->regmap,
> > +					 QCA8K_PORT_LOOKUP_CTRL(port),
> > +					 QCA8K_PORT_LOOKUP_LEARN);
> > +}
> > +
> >  void qca8k_port_stp_state_set(struct dsa_switch *ds, int port, u8 state)
> >  {
> > +	struct dsa_port *dp = dsa_to_port(ds, port);
> >  	struct qca8k_priv *priv = ds->priv;
> > +	bool learning = false;
> >  	u32 stp_state;
> >  
> >  	switch (state) {
> > @@ -582,8 +599,11 @@ void qca8k_port_stp_state_set(struct dsa_switch *ds, int port, u8 state)
> >  		break;
> >  	case BR_STATE_LEARNING:
> >  		stp_state = QCA8K_PORT_LOOKUP_STATE_LEARNING;
> > +		learning = dp->learning;
> >  		break;
> >  	case BR_STATE_FORWARDING:
> > +		learning = dp->learning;
> > +		fallthrough;
> >  	default:
> >  		stp_state = QCA8K_PORT_LOOKUP_STATE_FORWARD;
> >  		break;
> > @@ -591,6 +611,30 @@ void qca8k_port_stp_state_set(struct dsa_switch *ds, int port, u8 state)
> >  
> >  	qca8k_rmw(priv, QCA8K_PORT_LOOKUP_CTRL(port),
> >  		  QCA8K_PORT_LOOKUP_STATE_MASK, stp_state);
> > +
> > +	qca8k_port_configure_learning(ds, port, learning);
> > +}
> > +
> > +int qca8k_port_pre_bridge_flags(struct dsa_switch *ds, int port,
> > +				struct switchdev_brport_flags flags,
> > +				struct netlink_ext_ack *extack)
> > +{
> > +	if (flags.mask & ~BR_LEARNING)
> > +		return -EINVAL;
> > +
> > +	return 0;
> > +}
> > +
> > +int qca8k_port_bridge_flags(struct dsa_switch *ds, int port,
> > +			    struct switchdev_brport_flags flags,
> > +			    struct netlink_ext_ack *extack)
> > +{
> > +	int ret;
> > +
> > +	ret = qca8k_port_configure_learning(ds, port,
> > +					    flags.val & BR_LEARNING);
> > +
> > +	return ret;
> 
> I worry that the way in this is formulated will attract patches from
> kernel janitors to simplify it to:
> 
> 	return qca8k_port_configure_learning(...)
> 
> I agree that it's not strictly necessary to check flags.mask when
> port_pre_bridge_flags supports a single flag, but if you did that and
> structured the code for more future flags, you could avoid that.
> 
> 	int ret;
> 
> 	if (flags.mask & BR_LEARNING) {
> 		ret = qca8k_port_configure_learning(...,
> 						    flags.val & BR_LEARNING);
> 		if (ret)
> 			return ret;
> 	}
> 
> 	return 0;
> 
> Anyway, probably not a big deal.
> 
> >  }

I hope and expect to send fbd isolation later with FLOOD flags so I will
send v2 of this with the suggested format.

> >  
> >  int qca8k_port_bridge_join(struct dsa_switch *ds, int port,
> > diff --git a/drivers/net/dsa/qca/qca8k.h b/drivers/net/dsa/qca/qca8k.h
> > index c5cc8a172d65..8f88b7db384d 100644
> > --- a/drivers/net/dsa/qca/qca8k.h
> > +++ b/drivers/net/dsa/qca/qca8k.h
> > @@ -522,6 +522,12 @@ int qca8k_get_mac_eee(struct dsa_switch *ds, int port, struct ethtool_eee *e);
> >  
> >  /* Common bridge function */
> >  void qca8k_port_stp_state_set(struct dsa_switch *ds, int port, u8 state);
> > +int qca8k_port_pre_bridge_flags(struct dsa_switch *ds, int port,
> > +				struct switchdev_brport_flags flags,
> > +				struct netlink_ext_ack *extack);
> > +int qca8k_port_bridge_flags(struct dsa_switch *ds, int port,
> > +			    struct switchdev_brport_flags flags,
> > +			    struct netlink_ext_ack *extack);
> >  int qca8k_port_bridge_join(struct dsa_switch *ds, int port,
> >  			   struct dsa_bridge bridge,
> >  			   bool *tx_fwd_offload,
> > -- 
> > 2.40.1
> > 
> 

-- 
	Ansuel

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ