macvlan: Use software path for offloaded local, broadcast, and multicast traffic

This change makes it so that we use a software path for packets that are
going to be locally switched between two macvlan interfaces on the same
device. In addition we resort to software replication of broadcast and
multicast packets instead of offloading that to hardware.

The general idea is that using the device for east/west traffic local to
the system is extremely inefficient. We can only support up to whatever the
PCIe limit is for any given device so this caps us at somewhere around 20G
for devices supported by ixgbe. This is compounded even further when you
take broadcast and multicast into account as a single 10G port can come to
a crawl as a packet is replicated up to 60+ times in some cases. In order
to get away from that I am implementing changes so that we handle
broadcast/multicast replication and east/west local traffic all in
software.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

1 files changed, 10 insertions, 29 deletions

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index 0fbeb2f48711..4d4fa46a1e7c 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -4219,7 +4219,8 @@ static void ixgbe_setup_psrtype(struct ixgbe_adapter *adapter)
 static void ixgbe_configure_virtualization(struct ixgbe_adapter *adapter)
 {
 	struct ixgbe_hw *hw = &adapter->hw;
-	u32 reg_offset, vf_shift;
+	u16 pool = adapter->num_rx_pools;
+	u32 reg_offset, vf_shift, vmolr;
 	u32 gcr_ext, vmdctl;
 	int i;
 
@@ -4233,6 +4234,13 @@ static void ixgbe_configure_virtualization(struct ixgbe_adapter *adapter)
 	vmdctl |= IXGBE_VT_CTL_REPLEN;
 	IXGBE_WRITE_REG(hw, IXGBE_VT_CTL, vmdctl);
 
+	/* accept untagged packets until a vlan tag is
+	 * specifically set for the VMDQ queue/pool
+	 */
+	vmolr = IXGBE_VMOLR_AUPE;
+	while (pool--)
+		IXGBE_WRITE_REG(hw, IXGBE_VMOLR(VMDQ_P(pool)), vmolr);
+
 	vf_shift = VMDQ_P(0) % 32;
 	reg_offset = (VMDQ_P(0) >= 32) ? 1 : 0;
 
@@ -5279,28 +5287,6 @@ static void ixgbe_fdir_filter_restore(struct ixgbe_adapter *adapter)
 	spin_unlock(&adapter->fdir_perfect_lock);
 }
 
-static void ixgbe_macvlan_set_rx_mode(struct net_device *dev, unsigned int pool,
-				      struct ixgbe_adapter *adapter)
-{
-	struct ixgbe_hw *hw = &adapter->hw;
-	u32 vmolr;
-
-	/* No unicast promiscuous support for VMDQ devices. */
-	vmolr = IXGBE_READ_REG(hw, IXGBE_VMOLR(pool));
-	vmolr |= (IXGBE_VMOLR_ROMPE | IXGBE_VMOLR_BAM | IXGBE_VMOLR_AUPE);
-
-	/* clear the affected bit */
-	vmolr &= ~IXGBE_VMOLR_MPE;
-
-	if (dev->flags & IFF_ALLMULTI) {
-		vmolr |= IXGBE_VMOLR_MPE;
-	} else {
-		vmolr |= IXGBE_VMOLR_ROMPE;
-		hw->mac.ops.update_mc_addr_list(hw, dev);
-	}
-	IXGBE_WRITE_REG(hw, IXGBE_VMOLR(pool), vmolr);
-}
-
 /**
  * ixgbe_clean_rx_ring - Free Rx Buffers per Queue
  * @rx_ring: ring to free buffers from
@@ -5384,10 +5370,8 @@ static int ixgbe_fwd_ring_up(struct net_device *vdev,
 	 */
 	err = ixgbe_add_mac_filter(adapter, vdev->dev_addr,
 				   VMDQ_P(accel->pool));
-	if (err >= 0) {
-		ixgbe_macvlan_set_rx_mode(vdev, accel->pool, adapter);
+	if (err >= 0)
 		return 0;
-	}
 
 	for (i = 0; i < adapter->num_rx_queues_per_pool; i++)
 		adapter->rx_ring[baseq + i]->netdev = NULL;
@@ -9832,9 +9816,6 @@ static void ixgbe_fwd_del(struct net_device *pdev, void *priv)
 	ixgbe_del_mac_filter(adapter, accel->netdev->dev_addr,
 			     VMDQ_P(accel->pool));
 
-	/* disable ability to receive packets for this pool */
-	IXGBE_WRITE_REG(&adapter->hw, IXGBE_VMOLR(accel->pool), 0);
-
 	/* Allow remaining Rx packets to get flushed out of the
 	 * Rx FIFO before we drop the netdev for the ring.
 	 */