[PATCH] net: bcmgenet: use promisc for unsupported filters

[PATCH] net: bcmgenet: use promisc for unsupported filters

[justinpopo6]

From: Justin Chen <justinpopo6@gmail.com>

Currently we silently ignore filters if we cannot meet the filter
requirements. This will lead to the MAC dropping packets that are
expected to pass. A better solution would be to set the NIC to promisc
mode when the required filters cannot be met.

Also correct the number of MDF filters supported. It should be 17,
not 16.

Signed-off-by: Justin Chen <justinpopo6@gmail.com>
---
 drivers/net/ethernet/broadcom/genet/bcmgenet.c | 57 ++++++++++++--------------
 1 file changed, 26 insertions(+), 31 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
index 34466b8..a2b5780 100644
--- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c
+++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
@@ -3083,39 +3083,42 @@ static void bcmgenet_timeout(struct net_device *dev)
 	netif_tx_wake_all_queues(dev);
 }
 
-#define MAX_MC_COUNT	16
+#define MAX_MDF_FILTER	17
 
 static inline void bcmgenet_set_mdf_addr(struct bcmgenet_priv *priv,
 					 unsigned char *addr,
-					 int *i,
-					 int *mc)
+					 int *i)
 {
-	u32 reg;
-
 	bcmgenet_umac_writel(priv, addr[0] << 8 | addr[1],
 			     UMAC_MDF_ADDR + (*i * 4));
 	bcmgenet_umac_writel(priv, addr[2] << 24 | addr[3] << 16 |
 			     addr[4] << 8 | addr[5],
 			     UMAC_MDF_ADDR + ((*i + 1) * 4));
-	reg = bcmgenet_umac_readl(priv, UMAC_MDF_CTRL);
-	reg |= (1 << (MAX_MC_COUNT - *mc));
-	bcmgenet_umac_writel(priv, reg, UMAC_MDF_CTRL);
 	*i += 2;
-	(*mc)++;
 }
 
 static void bcmgenet_set_rx_mode(struct net_device *dev)
 {
 	struct bcmgenet_priv *priv = netdev_priv(dev);
 	struct netdev_hw_addr *ha;
-	int i, mc;
+	int i, nfilter;
 	u32 reg;
 
 	netif_dbg(priv, hw, dev, "%s: %08X\n", __func__, dev->flags);
 
-	/* Promiscuous mode */
+	/* Number of filters needed */
+	nfilter = netdev_uc_count(dev) + netdev_mc_count(dev) + 2;
+
+	/*
+	 * Turn on promicuous mode for three scenarios
+	 * 1. IFF_PROMISC flag is set
+	 * 2. IFF_ALLMULTI flag is set
+	 * 3. The number of filters needed exceeds the number filters
+	 *    supported by the hardware.
+	*/
 	reg = bcmgenet_umac_readl(priv, UMAC_CMD);
-	if (dev->flags & IFF_PROMISC) {
+	if ((dev->flags & (IFF_PROMISC | IFF_ALLMULTI)) ||
+	    (nfilter > MAX_MDF_FILTER)) {
 		reg |= CMD_PROMISC;
 		bcmgenet_umac_writel(priv, reg, UMAC_CMD);
 		bcmgenet_umac_writel(priv, 0, UMAC_MDF_CTRL);
@@ -3125,32 +3128,24 @@ static void bcmgenet_set_rx_mode(struct net_device *dev)
 		bcmgenet_umac_writel(priv, reg, UMAC_CMD);
 	}
 
-	/* UniMac doesn't support ALLMULTI */
-	if (dev->flags & IFF_ALLMULTI) {
-		netdev_warn(dev, "ALLMULTI is not supported\n");
-		return;
-	}
-
 	/* update MDF filter */
 	i = 0;
-	mc = 0;
 	/* Broadcast */
-	bcmgenet_set_mdf_addr(priv, dev->broadcast, &i, &mc);
+	bcmgenet_set_mdf_addr(priv, dev->broadcast, &i);
 	/* my own address.*/
-	bcmgenet_set_mdf_addr(priv, dev->dev_addr, &i, &mc);
-	/* Unicast list*/
-	if (netdev_uc_count(dev) > (MAX_MC_COUNT - mc))
-		return;
+	bcmgenet_set_mdf_addr(priv, dev->dev_addr, &i);
 
-	if (!netdev_uc_empty(dev))
-		netdev_for_each_uc_addr(ha, dev)
-			bcmgenet_set_mdf_addr(priv, ha->addr, &i, &mc);
-	/* Multicast */
-	if (netdev_mc_empty(dev) || netdev_mc_count(dev) >= (MAX_MC_COUNT - mc))
-		return;
+	/* Unicast */
+	netdev_for_each_uc_addr(ha, dev)
+		bcmgenet_set_mdf_addr(priv, ha->addr, &i);
 
+	/* Multicast */
 	netdev_for_each_mc_addr(ha, dev)
-		bcmgenet_set_mdf_addr(priv, ha->addr, &i, &mc);
+		bcmgenet_set_mdf_addr(priv, ha->addr, &i);
+
+	/* Enable filters */
+	reg = GENMASK(MAX_MDF_FILTER - 1, MAX_MDF_FILTER - nfilter);
+	bcmgenet_umac_writel(priv, reg, UMAC_MDF_CTRL);
 }
 
 /* Set the hardware MAC address. */
-- 
2.7.4

Re: [PATCH] net: bcmgenet: use promisc for unsupported filters

[Florian Fainelli]

On 7/17/2019 2:58 PM, justinpopo6@gmail.com wrote:
> From: Justin Chen <justinpopo6@gmail.com>
> 
> Currently we silently ignore filters if we cannot meet the filter
> requirements. This will lead to the MAC dropping packets that are
> expected to pass. A better solution would be to set the NIC to promisc
> mode when the required filters cannot be met.
> 
> Also correct the number of MDF filters supported. It should be 17,
> not 16.
> 
> Signed-off-by: Justin Chen <justinpopo6@gmail.com>

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>

Thanks Justin!
-- 
Florian

Re: [PATCH] net: bcmgenet: use promisc for unsupported filters

[David Miller]

From: Florian Fainelli <f.fainelli@gmail.com>
Date: Wed, 17 Jul 2019 20:38:47 -0700

> 
> 
> On 7/17/2019 2:58 PM, justinpopo6@gmail.com wrote:
>> From: Justin Chen <justinpopo6@gmail.com>
>> 
>> Currently we silently ignore filters if we cannot meet the filter
>> requirements. This will lead to the MAC dropping packets that are
>> expected to pass. A better solution would be to set the NIC to promisc
>> mode when the required filters cannot be met.
>> 
>> Also correct the number of MDF filters supported. It should be 17,
>> not 16.
>> 
>> Signed-off-by: Justin Chen <justinpopo6@gmail.com>
> 
> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>

Applied and queued up for -stable.

Re: [PATCH] net: bcmgenet: use promisc for unsupported filters

[Doug Berger]

On 7/17/19 2:58 PM, justinpopo6@gmail.com wrote:
> From: Justin Chen <justinpopo6@gmail.com>
> 
> Currently we silently ignore filters if we cannot meet the filter
> requirements. This will lead to the MAC dropping packets that are
> expected to pass. A better solution would be to set the NIC to promisc
> mode when the required filters cannot be met.
> 
> Also correct the number of MDF filters supported. It should be 17,
> not 16.
> 
> Signed-off-by: Justin Chen <justinpopo6@gmail.com>

Acked-by: Doug Berger <opendmb@gmail.com>

Thanks Justin :)
    Doug