bonding/xfrm: use real_dev instead of slave_dev
authorJarod Wilson <jarod@redhat.com>
Tue, 23 Jun 2020 20:40:01 +0000 (16:40 -0400)
committerDavid S. Miller <davem@davemloft.net>
Tue, 23 Jun 2020 22:19:55 +0000 (15:19 -0700)
Rather than requiring every hw crypto capable NIC driver to do a check for
slave_dev being set, set real_dev in the xfrm layer and xso init time, and
then override it in the bonding driver as needed. Then NIC drivers can
always use real_dev, and at the same time, we eliminate the use of a
variable name that probably shouldn't have been used in the first place,
particularly given recent current events.

CC: Boris Pismenny <borisp@mellanox.com>
CC: Saeed Mahameed <saeedm@mellanox.com>
CC: Leon Romanovsky <leon@kernel.org>
CC: Jay Vosburgh <j.vosburgh@gmail.com>
CC: Veaceslav Falico <vfalico@gmail.com>
CC: Andy Gospodarek <andy@greyhouse.net>
CC: "David S. Miller" <davem@davemloft.net>
CC: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
CC: Jakub Kicinski <kuba@kernel.org>
CC: Steffen Klassert <steffen.klassert@secunet.com>
CC: Herbert Xu <herbert@gondor.apana.org.au>
CC: netdev@vger.kernel.org
Suggested-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Jarod Wilson <jarod@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/bonding/bond_main.c
drivers/net/ethernet/intel/ixgbe/ixgbe_ipsec.c
drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec.c
include/net/xfrm.h
net/xfrm/xfrm_device.c

index 90939cc..4ef99ef 100644 (file)
@@ -386,7 +386,7 @@ static int bond_ipsec_add_sa(struct xfrm_state *xs)
        struct bonding *bond = netdev_priv(bond_dev);
        struct slave *slave = rtnl_dereference(bond->curr_active_slave);
 
-       xs->xso.slave_dev = slave->dev;
+       xs->xso.real_dev = slave->dev;
        bond->xs = xs;
 
        if (!(slave->dev->xfrmdev_ops
@@ -411,7 +411,7 @@ static void bond_ipsec_del_sa(struct xfrm_state *xs)
        if (!slave)
                return;
 
-       xs->xso.slave_dev = slave->dev;
+       xs->xso.real_dev = slave->dev;
 
        if (!(slave->dev->xfrmdev_ops
              && slave->dev->xfrmdev_ops->xdo_dev_state_delete)) {
@@ -440,7 +440,7 @@ static bool bond_ipsec_offload_ok(struct sk_buff *skb, struct xfrm_state *xs)
                return false;
        }
 
-       xs->xso.slave_dev = slave_dev;
+       xs->xso.real_dev = slave_dev;
        return slave_dev->xfrmdev_ops->xdo_dev_offload_ok(skb, xs);
 }
 
index 26b0a58..6516980 100644 (file)
@@ -427,14 +427,11 @@ static struct xfrm_state *ixgbe_ipsec_find_rx_state(struct ixgbe_ipsec *ipsec,
 static int ixgbe_ipsec_parse_proto_keys(struct xfrm_state *xs,
                                        u32 *mykey, u32 *mysalt)
 {
-       struct net_device *dev = xs->xso.dev;
+       struct net_device *dev = xs->xso.real_dev;
        unsigned char *key_data;
        char *alg_name = NULL;
        int key_len;
 
-       if (xs->xso.slave_dev)
-               dev = xs->xso.slave_dev;
-
        if (!xs->aead) {
                netdev_err(dev, "Unsupported IPsec algorithm\n");
                return -EINVAL;
@@ -480,9 +477,9 @@ static int ixgbe_ipsec_parse_proto_keys(struct xfrm_state *xs,
  **/
 static int ixgbe_ipsec_check_mgmt_ip(struct xfrm_state *xs)
 {
-       struct net_device *dev = xs->xso.dev;
-       struct ixgbe_adapter *adapter;
-       struct ixgbe_hw *hw;
+       struct net_device *dev = xs->xso.real_dev;
+       struct ixgbe_adapter *adapter = netdev_priv(dev);
+       struct ixgbe_hw *hw = &adapter->hw;
        u32 mfval, manc, reg;
        int num_filters = 4;
        bool manc_ipv4;
@@ -500,12 +497,6 @@ static int ixgbe_ipsec_check_mgmt_ip(struct xfrm_state *xs)
 #define BMCIP_V6                 0x3
 #define BMCIP_MASK               0x3
 
-       if (xs->xso.slave_dev)
-               dev = xs->xso.slave_dev;
-
-       adapter = netdev_priv(dev);
-       hw = &adapter->hw;
-
        manc = IXGBE_READ_REG(hw, IXGBE_MANC);
        manc_ipv4 = !!(manc & MANC_EN_IPV4_FILTER);
        mfval = IXGBE_READ_REG(hw, IXGBE_MFVAL);
@@ -569,22 +560,15 @@ static int ixgbe_ipsec_check_mgmt_ip(struct xfrm_state *xs)
  **/
 static int ixgbe_ipsec_add_sa(struct xfrm_state *xs)
 {
-       struct net_device *dev = xs->xso.dev;
-       struct ixgbe_adapter *adapter;
-       struct ixgbe_ipsec *ipsec;
-       struct ixgbe_hw *hw;
+       struct net_device *dev = xs->xso.real_dev;
+       struct ixgbe_adapter *adapter = netdev_priv(dev);
+       struct ixgbe_ipsec *ipsec = adapter->ipsec;
+       struct ixgbe_hw *hw = &adapter->hw;
        int checked, match, first;
        u16 sa_idx;
        int ret;
        int i;
 
-       if (xs->xso.slave_dev)
-               dev = xs->xso.slave_dev;
-
-       adapter = netdev_priv(dev);
-       ipsec = adapter->ipsec;
-       hw = &adapter->hw;
-
        if (xs->id.proto != IPPROTO_ESP && xs->id.proto != IPPROTO_AH) {
                netdev_err(dev, "Unsupported protocol 0x%04x for ipsec offload\n",
                           xs->id.proto);
@@ -761,20 +745,13 @@ static int ixgbe_ipsec_add_sa(struct xfrm_state *xs)
  **/
 static void ixgbe_ipsec_del_sa(struct xfrm_state *xs)
 {
-       struct net_device *dev = xs->xso.dev;
-       struct ixgbe_adapter *adapter;
-       struct ixgbe_ipsec *ipsec;
-       struct ixgbe_hw *hw;
+       struct net_device *dev = xs->xso.real_dev;
+       struct ixgbe_adapter *adapter = netdev_priv(dev);
+       struct ixgbe_ipsec *ipsec = adapter->ipsec;
+       struct ixgbe_hw *hw = &adapter->hw;
        u32 zerobuf[4] = {0, 0, 0, 0};
        u16 sa_idx;
 
-       if (xs->xso.slave_dev)
-               dev = xs->xso.slave_dev;
-
-       adapter = netdev_priv(dev);
-       ipsec = adapter->ipsec;
-       hw = &adapter->hw;
-
        if (xs->xso.flags & XFRM_OFFLOAD_INBOUND) {
                struct rx_sa *rsa;
                u8 ipi;
index 72ad666..bc55c82 100644 (file)
@@ -207,12 +207,9 @@ mlx5e_ipsec_build_accel_xfrm_attrs(struct mlx5e_ipsec_sa_entry *sa_entry,
 
 static inline int mlx5e_xfrm_validate_state(struct xfrm_state *x)
 {
-       struct net_device *netdev = x->xso.dev;
+       struct net_device *netdev = x->xso.real_dev;
        struct mlx5e_priv *priv;
 
-       if (x->xso.slave_dev)
-               netdev = x->xso.slave_dev;
-
        priv = netdev_priv(netdev);
 
        if (x->props.aalgo != SADB_AALG_NONE) {
@@ -288,15 +285,12 @@ static inline int mlx5e_xfrm_validate_state(struct xfrm_state *x)
 static int mlx5e_xfrm_add_state(struct xfrm_state *x)
 {
        struct mlx5e_ipsec_sa_entry *sa_entry = NULL;
-       struct net_device *netdev = x->xso.dev;
+       struct net_device *netdev = x->xso.real_dev;
        struct mlx5_accel_esp_xfrm_attrs attrs;
        struct mlx5e_priv *priv;
        unsigned int sa_handle;
        int err;
 
-       if (x->xso.slave_dev)
-               netdev = x->xso.slave_dev;
-
        priv = netdev_priv(netdev);
 
        err = mlx5e_xfrm_validate_state(x);
index e20b2b2..e648c9e 100644 (file)
@@ -127,7 +127,7 @@ struct xfrm_state_walk {
 
 struct xfrm_state_offload {
        struct net_device       *dev;
-       struct net_device       *slave_dev;
+       struct net_device       *real_dev;
        unsigned long           offload_handle;
        unsigned int            num_exthdrs;
        u8                      flags;
index b8918fc..7b64bb6 100644 (file)
@@ -120,8 +120,8 @@ struct sk_buff *validate_xmit_xfrm(struct sk_buff *skb, netdev_features_t featur
        if (xo->flags & XFRM_GRO || x->xso.flags & XFRM_OFFLOAD_INBOUND)
                return skb;
 
-       /* This skb was already validated on the master dev */
-       if ((x->xso.dev != dev) && (x->xso.slave_dev == dev))
+       /* This skb was already validated on the upper/virtual dev */
+       if ((x->xso.dev != dev) && (x->xso.real_dev == dev))
                return skb;
 
        local_irq_save(flags);
@@ -259,6 +259,7 @@ int xfrm_dev_state_add(struct net *net, struct xfrm_state *x,
        }
 
        xso->dev = dev;
+       xso->real_dev = dev;
        xso->num_exthdrs = 1;
        xso->flags = xuo->flags;