From: Saeed Mahameed <saeedm@mellanox.com>
To: "David S. Miller" <davem@davemloft.net>, kuba@kernel.org
Cc: netdev@vger.kernel.org, Vu Pham <vuhuong@mellanox.com>,
Parav Pandit <parav@mellanox.com>, Roi Dayan <roid@mellanox.com>,
Or Gerlitz <ogerlitz@mellanox.com>,
Saeed Mahameed <saeedm@mellanox.com>
Subject: [net-next V3 05/15] net/mlx5e: Support tc block sharing for representors
Date: Wed, 27 May 2020 18:16:46 -0700 [thread overview]
Message-ID: <20200528011656.559914-6-saeedm@mellanox.com> (raw)
In-Reply-To: <20200528011656.559914-1-saeedm@mellanox.com>
From: Vu Pham <vuhuong@mellanox.com>
Currently offloading a rule over a tc block shared by multiple
representors fails because an e-switch global hashtable to keep
the mapping from tc cookies to mlx5e flow instances is used, and
tc block sharing offloads the same rule/cookie multiple times,
each time for different representor sharing the tc block.
Changing the implementation and behavior by acknowledging and returning
success if the same rule/cookie is offloaded again to other slave
representor sharing the tc block by setting, checking and comparing
the netdev that added the rule first.
Signed-off-by: Vu Pham <vuhuong@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 23 +++++++++++++++++++
1 file changed, 23 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 571da14809fe..f3e65a15c950 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -145,6 +145,7 @@ struct mlx5e_tc_flow {
struct list_head hairpin; /* flows sharing the same hairpin */
struct list_head peer; /* flows with peer flow */
struct list_head unready; /* flows not ready to be offloaded (e.g due to missing route) */
+ struct net_device *orig_dev; /* netdev adding flow first */
int tmp_efi_index;
struct list_head tmp_list; /* temporary flow list used by neigh update */
refcount_t refcnt;
@@ -4624,11 +4625,21 @@ mlx5e_tc_add_flow(struct mlx5e_priv *priv,
return err;
}
+static bool is_flow_rule_duplicate_allowed(struct net_device *dev,
+ struct mlx5e_rep_priv *rpriv)
+{
+ /* Offloaded flow rule is allowed to duplicate on non-uplink representor
+ * sharing tc block with other slaves of a lag device.
+ */
+ return netif_is_lag_port(dev) && rpriv->rep->vport != MLX5_VPORT_UPLINK;
+}
+
int mlx5e_configure_flower(struct net_device *dev, struct mlx5e_priv *priv,
struct flow_cls_offload *f, unsigned long flags)
{
struct netlink_ext_ack *extack = f->common.extack;
struct rhashtable *tc_ht = get_tc_ht(priv, flags);
+ struct mlx5e_rep_priv *rpriv = priv->ppriv;
struct mlx5e_tc_flow *flow;
int err = 0;
@@ -4636,6 +4647,12 @@ int mlx5e_configure_flower(struct net_device *dev, struct mlx5e_priv *priv,
flow = rhashtable_lookup(tc_ht, &f->cookie, tc_ht_params);
rcu_read_unlock();
if (flow) {
+ /* Same flow rule offloaded to non-uplink representor sharing tc block,
+ * just return 0.
+ */
+ if (is_flow_rule_duplicate_allowed(dev, rpriv) && flow->orig_dev != dev)
+ goto out;
+
NL_SET_ERR_MSG_MOD(extack,
"flow cookie already exists, ignoring");
netdev_warn_once(priv->netdev,
@@ -4650,6 +4667,12 @@ int mlx5e_configure_flower(struct net_device *dev, struct mlx5e_priv *priv,
if (err)
goto out;
+ /* Flow rule offloaded to non-uplink representor sharing tc block,
+ * set the flow's owner dev.
+ */
+ if (is_flow_rule_duplicate_allowed(dev, rpriv))
+ flow->orig_dev = dev;
+
err = rhashtable_lookup_insert_fast(tc_ht, &flow->node, tc_ht_params);
if (err)
goto err_free;
--
2.26.2
next prev parent reply other threads:[~2020-05-28 1:17 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-28 1:16 [pull request][net-next V3 00/15] mlx5 updates 2020-05-26 Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 01/15] net/mlx5: E-Switch, Refactor eswitch egress acl codes Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 02/15] net/mlx5: E-Switch, Refactor eswitch ingress " Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 03/15] net/mlx5: E-Switch, Introduce APIs to enable egress acl forward-to-vport rule Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 04/15] net/mlx5e: Use netdev events to set/del " Saeed Mahameed
2020-05-28 1:16 ` Saeed Mahameed [this message]
2020-05-28 1:16 ` [net-next V3 06/15] net/mlx5e: Offload flow rules to active lower representor Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 07/15] net/mlx5e: Add bond_metadata and its slave entries Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 08/15] net/mlx5: E-Switch, Alloc and free unique metadata for match Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 09/15] net/mlx5e: Slave representors sharing " Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 10/15] net/mlx5e: Use change upper event to setup representors' bond_metadata Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 11/15] net/mlx5: Add missing mutex destroy Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 12/15] net/mlx5e: Helper function to set ethertype Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 13/15] net/mlx5e: Optimize performance for IPv4/IPv6 ethertype Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 14/15] net/mlx5: DR, Add a spinlock to protect the send ring Saeed Mahameed
2020-05-28 1:16 ` [net-next V3 15/15] net/mlx5: DR, Split RX and TX lock for parallel insertion Saeed Mahameed
2020-05-28 18:04 ` [pull request][net-next V3 00/15] mlx5 updates 2020-05-26 David Miller
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200528011656.559914-6-saeedm@mellanox.com \
--to=saeedm@mellanox.com \
--cc=davem@davemloft.net \
--cc=kuba@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=ogerlitz@mellanox.com \
--cc=parav@mellanox.com \
--cc=roid@mellanox.com \
--cc=vuhuong@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).