[pull request][net-next 00/11] Mellanox, mlx5 updates 2018-12-19

[pull request][net-next 00/11] Mellanox, mlx5 updates 2018-12-19

[Saeed Mahameed]

Hi Dave,

This series includes some misc updates and vlan encap tunnel support.
For more information please see tag log below.

Please pull and let me know if there is any problem.

Please note that the series starts with a merge of mlx5-next branch,
to resolve and avoid dependency with rdma tree.

=======================
Conflict note: when merged with net branch this series adds 2 trivial
conflicts in mlx5.
1)
+<<<<<<< HEAD
 +#define MLX5E_REP_PARAMS_DEF_LOG_SQ_SIZE \
 +      max(0x7, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE)
++=======
+ #define MLX5E_REP_PARAMS_LOG_SQ_SIZE \
+       max(0x6, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE)
+ #define MLX5E_REP_PARAMS_DEF_NUM_CHANNELS 1
++>>>>>>> dave/master

Resolution:

define MLX5E_REP_PARAMS_DEF_LOG_SQ_SIZE \
	max(0x7, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE)
#define MLX5E_REP_PARAMS_DEF_NUM_CHANNELS 1

2)
++<<<<<<< HEAD
 +      priv->channels.params.num_channels =
 +                      mlx5e_rep_get_default_num_channels(rpriv->rep, netdev);
++=======
+       priv->channels.params.num_channels = MLX5E_REP_PARAMS_DEF_NUM_CHANNELS;
++>>>>>>> dave/master

Resolution: Use the hunk from net-next (1st hunk).
=======================

You will see other conflicts in mlx5 but i already notified you about
those in pull request: Mellanox, mlx5 fixes 2018-12-13
https://patchwork.ozlabs.org/patch/1012968/

Thanks,
Saeed.

---
The following changes since commit 0056ce4865713162e32619dbfcca21ff2ba8b8dc:

  Merge branch 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux (2018-12-19 12:12:18 -0800)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5-updates-2018-12-19

for you to fetch changes up to a8836756493290d3b156d63050d203f95d691a32:

  net/mlx5e: Support tunnel encap over tagged Ethernet (2018-12-19 12:12:36 -0800)

----------------------------------------------------------------
mlx5-updates-2018-12-19

This series adds some misc updates and the support for tunnels over VLAN
tc offloads.

>From Miroslav Lichvar, patches #1,2
1) Update timecounter at least twice per counter overflow
2) Extend PTP gettime function to read system clock

>From Gavi Teitz, patch #3
3) Increase VF representors' SQ size to 128

>From Eli Britstein and Or Gerlitz, patches #4-10
4) Adds the capability to support tunnels over VLAN device.

Patch 4 avoids crash for TC flow with egress upper devices

Patch 5 refactors tunnel routing devs into a helper function

Patch 6 avoids crash for TC encap flows with vlan on underlay

Patches 7-8 refactor encap tunnel header preparing code.

Patch 9 adds support for building VLAN tagged ETH header.

Patch 10 adds support for tunnel routing to VLAN device.

----------------------------------------------------------------
Eli Britstein (7):
      net/mlx5e: Fail attempt to offload e-switch TC flows with egress upper devices
      net/mlx5e: Tunnel routing output devs helper function
      net/mlx5e: Fail attempt to offload e-switch TC encap flows with vlan on underlay
      net/mlx5e: Tunnel encap ETH header helper function
      net/mlx5e: Re-order route and encap header memory allocation
      net/mlx5e: Support VLAN encap ETH header generation
      net/mlx5e: Support tunnel encap over tagged Ethernet

Gavi Teitz (1):
      net/mlx5e: Increase VF representors' SQ size to 128

Miroslav Lichvar (2):
      mlx5: update timecounter at least twice per counter overflow
      mlx5: extend PTP gettime function to read system clock

Or Gerlitz (1):
      net/mlx5e: Allow vlans on e-switch uplink reps

 .../net/ethernet/mellanox/mlx5/core/en/tc_tun.c    | 186 ++++++++++++---------
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c   |  20 ++-
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.h   |   3 +
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c    |   3 +
 .../net/ethernet/mellanox/mlx5/core/lib/clock.c    |  24 +--
 drivers/net/ethernet/mellanox/mlx5/core/main.c     |  11 +-
 .../net/ethernet/mellanox/mlx5/core/mlx5_core.h    |   4 +-
 7 files changed, 156 insertions(+), 95 deletions(-)

[net-next 01/11] mlx5: update timecounter at least twice per counter overflow

[Saeed Mahameed]

From: Miroslav Lichvar <mlichvar@redhat.com>

The timecounter needs to be updated at least once in half of the
cyclecounter interval to prevent timecounter_cyc2time() interpreting a
new timestamp as an old value and causing a backward jump.

This would be an issue if the timecounter multiplier was so small that
the update interval would not be limited by the 64-bit overflow in
multiplication.

Shorten the calculated interval to make sure the timecounter is updated
in time even when the system clock is slowed down by up to 10%, the
multiplier is increased by up to 10%, and the scheduled overflow check
is late by 15%.

Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Ariel Levkovich <lariel@mellanox.com>
Cc: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
index d27c239e7d6c..53f021df4934 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
@@ -517,14 +517,14 @@ void mlx5_init_clock(struct mlx5_core_dev *mdev)
 			 ktime_to_ns(ktime_get_real()));
 
 	/* Calculate period in seconds to call the overflow watchdog - to make
-	 * sure counter is checked at least once every wrap around.
+	 * sure counter is checked at least twice every wrap around.
 	 * The period is calculated as the minimum between max HW cycles count
 	 * (The clock source mask) and max amount of cycles that can be
 	 * multiplied by clock multiplier where the result doesn't exceed
 	 * 64bits.
 	 */
 	overflow_cycles = div64_u64(~0ULL >> 1, clock->cycles.mult);
-	overflow_cycles = min(overflow_cycles, clock->cycles.mask >> 1);
+	overflow_cycles = min(overflow_cycles, div_u64(clock->cycles.mask, 3));
 
 	ns = cyclecounter_cyc2ns(&clock->cycles, overflow_cycles,
 				 frac, &frac);
-- 
2.19.2

[net-next 02/11] mlx5: extend PTP gettime function to read system clock

[Saeed Mahameed]

From: Miroslav Lichvar <mlichvar@redhat.com>

Read the system time right before and immediately after reading the low
register of the internal timer. This adds support for the
PTP_SYS_OFFSET_EXTENDED ioctl.

Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/lib/clock.c   | 20 +++++++++++--------
 .../net/ethernet/mellanox/mlx5/core/main.c    | 11 ++++++++--
 .../ethernet/mellanox/mlx5/core/mlx5_core.h   |  4 +++-
 3 files changed, 24 insertions(+), 11 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
index 53f021df4934..ca0ee9916e9e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
@@ -72,7 +72,7 @@ static u64 read_internal_timer(const struct cyclecounter *cc)
 	struct mlx5_core_dev *mdev = container_of(clock, struct mlx5_core_dev,
 						  clock);
 
-	return mlx5_read_internal_timer(mdev) & cc->mask;
+	return mlx5_read_internal_timer(mdev, NULL) & cc->mask;
 }
 
 static void mlx5_update_clock_info_page(struct mlx5_core_dev *mdev)
@@ -156,15 +156,19 @@ static int mlx5_ptp_settime(struct ptp_clock_info *ptp,
 	return 0;
 }
 
-static int mlx5_ptp_gettime(struct ptp_clock_info *ptp, struct timespec64 *ts)
+static int mlx5_ptp_gettimex(struct ptp_clock_info *ptp, struct timespec64 *ts,
+			     struct ptp_system_timestamp *sts)
 {
 	struct mlx5_clock *clock = container_of(ptp, struct mlx5_clock,
 						ptp_info);
-	u64 ns;
+	struct mlx5_core_dev *mdev = container_of(clock, struct mlx5_core_dev,
+						  clock);
 	unsigned long flags;
+	u64 cycles, ns;
 
 	write_seqlock_irqsave(&clock->lock, flags);
-	ns = timecounter_read(&clock->tc);
+	cycles = mlx5_read_internal_timer(mdev, sts);
+	ns = timecounter_cyc2time(&clock->tc, cycles);
 	write_sequnlock_irqrestore(&clock->lock, flags);
 
 	*ts = ns_to_timespec64(ns);
@@ -307,7 +311,7 @@ static int mlx5_perout_configure(struct ptp_clock_info *ptp,
 		ts.tv_sec = rq->perout.start.sec;
 		ts.tv_nsec = rq->perout.start.nsec;
 		ns = timespec64_to_ns(&ts);
-		cycles_now = mlx5_read_internal_timer(mdev);
+		cycles_now = mlx5_read_internal_timer(mdev, NULL);
 		write_seqlock_irqsave(&clock->lock, flags);
 		nsec_now = timecounter_cyc2time(&clock->tc, cycles_now);
 		nsec_delta = ns - nsec_now;
@@ -384,7 +388,7 @@ static const struct ptp_clock_info mlx5_ptp_clock_info = {
 	.pps		= 0,
 	.adjfreq	= mlx5_ptp_adjfreq,
 	.adjtime	= mlx5_ptp_adjtime,
-	.gettime64	= mlx5_ptp_gettime,
+	.gettimex64	= mlx5_ptp_gettimex,
 	.settime64	= mlx5_ptp_settime,
 	.enable		= NULL,
 	.verify		= NULL,
@@ -469,8 +473,8 @@ static int mlx5_pps_event(struct notifier_block *nb,
 		ptp_clock_event(clock->ptp, &ptp_event);
 		break;
 	case PTP_PF_PEROUT:
-		mlx5_ptp_gettime(&clock->ptp_info, &ts);
-		cycles_now = mlx5_read_internal_timer(mdev);
+		mlx5_ptp_gettimex(&clock->ptp_info, &ts, NULL);
+		cycles_now = mlx5_read_internal_timer(mdev, NULL);
 		ts.tv_sec += 1;
 		ts.tv_nsec = 0;
 		ns = timespec64_to_ns(&ts);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index 7e44d6f180b0..be81b319b0dc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -580,15 +580,22 @@ int mlx5_core_disable_hca(struct mlx5_core_dev *dev, u16 func_id)
 	return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
 }
 
-u64 mlx5_read_internal_timer(struct mlx5_core_dev *dev)
+u64 mlx5_read_internal_timer(struct mlx5_core_dev *dev,
+			     struct ptp_system_timestamp *sts)
 {
 	u32 timer_h, timer_h1, timer_l;
 
 	timer_h = ioread32be(&dev->iseg->internal_timer_h);
+	ptp_read_system_prets(sts);
 	timer_l = ioread32be(&dev->iseg->internal_timer_l);
+	ptp_read_system_postts(sts);
 	timer_h1 = ioread32be(&dev->iseg->internal_timer_h);
-	if (timer_h != timer_h1) /* wrap around */
+	if (timer_h != timer_h1) {
+		/* wrap around */
+		ptp_read_system_prets(sts);
 		timer_l = ioread32be(&dev->iseg->internal_timer_l);
+		ptp_read_system_postts(sts);
+	}
 
 	return (u64)timer_l | (u64)timer_h1 << 32;
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
index 73bf46599ec6..c68dcea5985b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
@@ -38,6 +38,7 @@
 #include <linux/sched.h>
 #include <linux/if_link.h>
 #include <linux/firmware.h>
+#include <linux/ptp_clock_kernel.h>
 #include <linux/mlx5/cq.h>
 #include <linux/mlx5/fs.h>
 
@@ -121,7 +122,8 @@ int mlx5_modify_scheduling_element_cmd(struct mlx5_core_dev *dev, u8 hierarchy,
 int mlx5_destroy_scheduling_element_cmd(struct mlx5_core_dev *dev, u8 hierarchy,
 					u32 element_id);
 int mlx5_wait_for_vf_pages(struct mlx5_core_dev *dev);
-u64 mlx5_read_internal_timer(struct mlx5_core_dev *dev);
+u64 mlx5_read_internal_timer(struct mlx5_core_dev *dev,
+			     struct ptp_system_timestamp *sts);
 
 void mlx5_cmd_trigger_completions(struct mlx5_core_dev *dev);
 int mlx5_cq_debugfs_init(struct mlx5_core_dev *dev);
-- 
2.19.2

[net-next 03/11] net/mlx5e: Increase VF representors' SQ size to 128

[Saeed Mahameed]

From: Gavi Teitz <gavi@mellanox.com>

The default size for the VF representors' SQ was too small to handle high
packet rates. Doubling the size from 64 to 128 drastically improves the
packet rate under stress (by about 50%), whereas increasing the size
beyond 128 has not shown to make any further difference.

The impact of the SQ size was measured with UDP traffic, in the following
topology: TG <-> PF <-> TC forwarding <-> VF representor <-> VF in VM
over a single core processing bi-directional traffic, with the following
results:

                                  SQ size of 64:     SQ size of 128:
Packet rate for 64B UDP packets:    860 [Kpps]         1280 [Kpps]
Packet rate for 114B VxLan
encapsulated UDP packets:           320 [Kpps]          500 [Kpps]

Signed-off-by: Gavi Teitz <gavi@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index ed1158b58798..18d0f4f7537b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -45,8 +45,8 @@
 #include "en/tc_tun.h"
 #include "fs_core.h"
 
-#define MLX5E_REP_PARAMS_LOG_SQ_SIZE \
-	max(0x6, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE)
+#define MLX5E_REP_PARAMS_DEF_LOG_SQ_SIZE \
+	max(0x7, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE)
 
 static const char mlx5e_rep_driver_name[] = "mlx5e_rep";
 
@@ -1336,7 +1336,7 @@ static void mlx5e_build_rep_params(struct net_device *netdev)
 	if (rep->vport == FDB_UPLINK_VPORT)
 		params->log_sq_size = MLX5E_PARAMS_DEFAULT_LOG_SQ_SIZE;
 	else
-		params->log_sq_size = MLX5E_REP_PARAMS_LOG_SQ_SIZE;
+		params->log_sq_size = MLX5E_REP_PARAMS_DEF_LOG_SQ_SIZE;;
 
 	/* RQ */
 	mlx5e_build_rq_params(mdev, params);
-- 
2.19.2

[net-next 04/11] net/mlx5e: Allow vlans on e-switch uplink reps

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

There are cases (e.g tunneling with vlan on underlay and potentially
more) where this makes sense, so allow that.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 18d0f4f7537b..91c3eb85f32e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -1381,7 +1381,7 @@ static void mlx5e_build_rep_netdev(struct net_device *netdev)
 
 	netdev->switchdev_ops = &mlx5e_rep_switchdev_ops;
 
-	netdev->features	 |= NETIF_F_VLAN_CHALLENGED | NETIF_F_HW_TC | NETIF_F_NETNS_LOCAL;
+	netdev->features	 |= NETIF_F_HW_TC | NETIF_F_NETNS_LOCAL;
 	netdev->hw_features      |= NETIF_F_HW_TC;
 
 	netdev->hw_features    |= NETIF_F_SG;
@@ -1392,6 +1392,9 @@ static void mlx5e_build_rep_netdev(struct net_device *netdev)
 	netdev->hw_features    |= NETIF_F_TSO6;
 	netdev->hw_features    |= NETIF_F_RXCSUM;
 
+	if (rep->vport != FDB_UPLINK_VPORT)
+		netdev->features |= NETIF_F_VLAN_CHALLENGED;
+
 	netdev->features |= netdev->hw_features;
 }
 
-- 
2.19.2

[net-next 05/11] net/mlx5e: Fail attempt to offload e-switch TC flows with egress upper devices

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

We use the switchdev parent HW id helper to identify if the mirred device
shares the same ASIC/port with the ingress device. This can get us wrong in
the presence of upper devices (e.g vlan, team, etc) set over the HW devices
(VF or uplink representors), b/c the switchdev ID is retrieved recursively.

To fail offload attempts in such cases, we condition the check on the egress
device to have not only the same switchdev ID but also the relevant mlx5 netdev ops.

Fixes: 03a9d11e6eeb ('net/mlx5e: Add TC drop and mirred/redirect action parsing for SRIOV offloads')
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 9 +++++++++
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.h | 3 +++
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c  | 3 +++
 3 files changed, 15 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 91c3eb85f32e..f414f19c1159 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -1316,6 +1316,15 @@ static const struct net_device_ops mlx5e_netdev_ops_uplink_rep = {
 	.ndo_get_vf_stats        = mlx5e_get_vf_stats,
 };
 
+bool mlx5e_eswitch_rep(struct net_device *netdev)
+{
+	if (netdev->netdev_ops == &mlx5e_netdev_ops_vf_rep ||
+	    netdev->netdev_ops == &mlx5e_netdev_ops_uplink_rep)
+		return true;
+
+	return false;
+}
+
 static void mlx5e_build_rep_params(struct net_device *netdev)
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
index 5645d3cef1bb..edd722824697 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
@@ -176,6 +176,9 @@ void mlx5e_rep_encap_entry_detach(struct mlx5e_priv *priv,
 				  struct mlx5e_encap_entry *e);
 
 void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv);
+
+bool mlx5e_eswitch_rep(struct net_device *netdev);
+
 #else /* CONFIG_MLX5_ESWITCH */
 static inline bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv) { return false; }
 static inline int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) { return 0; }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index c1a9120412b8..a4c6287dfd65 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -2578,6 +2578,9 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 				struct net_device *uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
 				struct net_device *uplink_upper = netdev_master_upper_dev_get(uplink_dev);
 
+				if (!mlx5e_eswitch_rep(out_dev))
+					return -EOPNOTSUPP;
+
 				if (uplink_upper &&
 				    netif_is_lag_master(uplink_upper) &&
 				    uplink_upper == out_dev)
-- 
2.19.2

[net-next 06/11] net/mlx5e: Tunnel routing output devs helper function

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

For tunnel we determine the output devs for IPv4/6 cases, in two
separate functions, with a duplicated code.

Move that code from IPv4/6 functions to a helper function, with no
functional change.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en/tc_tun.c   | 70 +++++++++----------
 1 file changed, 34 insertions(+), 36 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index b92f8b3ff6b2..c90a0f9cb6e1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -6,6 +6,34 @@
 #include "lib/vxlan.h"
 #include "en/tc_tun.h"
 
+static int get_route_and_out_devs(struct mlx5e_priv *priv,
+				  struct net_device *dev,
+				  struct net_device **route_dev,
+				  struct net_device **out_dev)
+{
+	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
+	struct net_device *uplink_dev, *uplink_upper;
+	bool dst_is_lag_dev;
+
+	uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
+	uplink_upper = netdev_master_upper_dev_get(uplink_dev);
+	dst_is_lag_dev = (uplink_upper &&
+			  netif_is_lag_master(uplink_upper) &&
+			  dev == uplink_upper &&
+			  mlx5_lag_is_sriov(priv->mdev));
+
+	/* if the egress device isn't on the same HW e-switch or
+	 * it's a LAG device, use the uplink
+	 */
+	if (!switchdev_port_same_parent_id(priv->netdev, dev) ||
+	    dst_is_lag_dev)
+		*out_dev = uplink_dev;
+	else
+		*out_dev = dev;
+
+	return 0;
+}
+
 static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv,
 				   struct net_device *mirred_dev,
 				   struct net_device **out_dev,
@@ -13,9 +41,6 @@ static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv,
 				   struct neighbour **out_n,
 				   u8 *out_ttl)
 {
-	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
-	struct net_device *uplink_dev, *uplink_upper;
-	bool dst_is_lag_dev;
 	struct rtable *rt;
 	struct neighbour *n = NULL;
 
@@ -30,21 +55,9 @@ static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv,
 	return -EOPNOTSUPP;
 #endif
 
-	uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
-	uplink_upper = netdev_master_upper_dev_get(uplink_dev);
-	dst_is_lag_dev = (uplink_upper &&
-			  netif_is_lag_master(uplink_upper) &&
-			  rt->dst.dev == uplink_upper &&
-			  mlx5_lag_is_sriov(priv->mdev));
-
-	/* if the egress device isn't on the same HW e-switch or
-	 * it's a LAG device, use the uplink
-	 */
-	if (!switchdev_port_same_parent_id(priv->netdev, rt->dst.dev) ||
-	    dst_is_lag_dev)
-		*out_dev = uplink_dev;
-	else
-		*out_dev = rt->dst.dev;
+	ret = get_route_and_out_devs(priv, rt->dst.dev, NULL, out_dev);
+	if (ret < 0)
+		return ret;
 
 	if (!(*out_ttl))
 		*out_ttl = ip4_dst_hoplimit(&rt->dst);
@@ -76,9 +89,6 @@ static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv,
 	struct dst_entry *dst;
 
 #if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6)
-	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
-	struct net_device *uplink_dev, *uplink_upper;
-	bool dst_is_lag_dev;
 	int ret;
 
 	ret = ipv6_stub->ipv6_dst_lookup(dev_net(mirred_dev), NULL, &dst,
@@ -89,21 +99,9 @@ static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv,
 	if (!(*out_ttl))
 		*out_ttl = ip6_dst_hoplimit(dst);
 
-	uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
-	uplink_upper = netdev_master_upper_dev_get(uplink_dev);
-	dst_is_lag_dev = (uplink_upper &&
-			  netif_is_lag_master(uplink_upper) &&
-			  dst->dev == uplink_upper &&
-			  mlx5_lag_is_sriov(priv->mdev));
-
-	/* if the egress device isn't on the same HW e-switch or
-	 * it's a LAG device, use the uplink
-	 */
-	if (!switchdev_port_same_parent_id(priv->netdev, dst->dev) ||
-	    dst_is_lag_dev)
-		*out_dev = uplink_dev;
-	else
-		*out_dev = dst->dev;
+	ret = get_route_and_out_devs(priv, dst->dev, NULL, out_dev);
+	if (ret < 0)
+		return ret;
 #else
 	return -EOPNOTSUPP;
 #endif
-- 
2.19.2

[net-next 07/11] net/mlx5e: Fail attempt to offload e-switch TC encap flows with vlan on underlay

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

Currently we don't support nor fail attempts to offload encap flows routed
to vlan device on the underlay network. We wrongly consider a vlan underlay
device to be on the same e-switch b/c the switchdev ID is retrieved recursively.

Add explicit check for that and fail such attempts.

Also align to a more strict check for the ingress and the underlay devices
to practically be on the same eswitch.

Fixes: ce99f6b97fcd ('net/mlx5e: Support SRIOV TC encapsulation offloads for IPv6 tunnels')
Fixes: 3e621b19b0bb ('net/mlx5e: Support TC encapsulation offloads with upper devices')
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index c90a0f9cb6e1..0d24e9648c21 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -15,6 +15,10 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv,
 	struct net_device *uplink_dev, *uplink_upper;
 	bool dst_is_lag_dev;
 
+	/* we currently don't offload vlan on underlay */
+	if (is_vlan_dev(dev))
+		return -EOPNOTSUPP;
+
 	uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
 	uplink_upper = netdev_master_upper_dev_get(uplink_dev);
 	dst_is_lag_dev = (uplink_upper &&
@@ -28,6 +32,8 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv,
 	if (!switchdev_port_same_parent_id(priv->netdev, dev) ||
 	    dst_is_lag_dev)
 		*out_dev = uplink_dev;
+	else if (!mlx5e_eswitch_rep(dev))
+		return -EOPNOTSUPP;
 	else
 		*out_dev = dev;
 
-- 
2.19.2

[net-next 08/11] net/mlx5e: Tunnel encap ETH header helper function

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

In tunnel encap we prepare the encap header for IPv4/6 cases, in two
separate functions. For ETH header generation the code is almost
duplicated.

Move the ETH header generation code from IPv4/6 functions to a helper
function, with no functional change.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en/tc_tun.c   | 29 +++++++++++--------
 1 file changed, 17 insertions(+), 12 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index 0d24e9648c21..f90c67c9add0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -180,6 +180,19 @@ static int mlx5e_gen_ip_tunnel_header(char buf[], __u8 *ip_proto,
 	return err;
 }
 
+static char *gen_eth_tnl_hdr(char *buf, struct net_device *dev,
+			     struct mlx5e_encap_entry *e,
+			     u16 proto)
+{
+	struct ethhdr *eth = (struct ethhdr *)buf;
+
+	ether_addr_copy(eth->h_dest, e->h_dest);
+	ether_addr_copy(eth->h_source, dev->dev_addr);
+	eth->h_proto = htons(proto);
+
+	return (char *)eth + ETH_HLEN;
+}
+
 int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 				    struct net_device *mirred_dev,
 				    struct mlx5e_encap_entry *e)
@@ -193,7 +206,6 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 	struct neighbour *n = NULL;
 	struct flowi4 fl4 = {};
 	char *encap_header;
-	struct ethhdr *eth;
 	u8 nud_state, ttl;
 	struct iphdr *ip;
 	int err;
@@ -242,13 +254,10 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 	read_unlock_bh(&n->lock);
 
 	/* add ethernet header */
-	eth = (struct ethhdr *)encap_header;
-	ether_addr_copy(eth->h_dest, e->h_dest);
-	ether_addr_copy(eth->h_source, out_dev->dev_addr);
-	eth->h_proto = htons(ETH_P_IP);
+	ip = (struct iphdr *)gen_eth_tnl_hdr(encap_header, out_dev, e,
+					     ETH_P_IP);
 
 	/* add ip header */
-	ip = (struct iphdr *)((char *)eth + sizeof(struct ethhdr));
 	ip->tos = tun_key->tos;
 	ip->version = 0x4;
 	ip->ihl = 0x5;
@@ -308,7 +317,6 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 	struct flowi6 fl6 = {};
 	struct ipv6hdr *ip6h;
 	char *encap_header;
-	struct ethhdr *eth;
 	u8 nud_state, ttl;
 	int err;
 
@@ -356,13 +364,10 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 	read_unlock_bh(&n->lock);
 
 	/* add ethernet header */
-	eth = (struct ethhdr *)encap_header;
-	ether_addr_copy(eth->h_dest, e->h_dest);
-	ether_addr_copy(eth->h_source, out_dev->dev_addr);
-	eth->h_proto = htons(ETH_P_IPV6);
+	ip6h = (struct ipv6hdr *)gen_eth_tnl_hdr(encap_header, out_dev, e,
+						 ETH_P_IPV6);
 
 	/* add ip header */
-	ip6h = (struct ipv6hdr *)((char *)eth + sizeof(struct ethhdr));
 	ip6_flow_hdr(ip6h, tun_key->tos, 0);
 	/* the HW fills up ipv6 payload len */
 	ip6h->hop_limit   = ttl;
-- 
2.19.2

[net-next 09/11] net/mlx5e: Re-order route and encap header memory allocation

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

Change the order to first route IPv4/6 and return if error. Only after
successful route continue to allocate an encap header, with no
functional change.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en/tc_tun.c   | 62 ++++++++++---------
 1 file changed, 34 insertions(+), 28 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index f90c67c9add0..bcf08f5abc2f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -198,28 +198,16 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 				    struct mlx5e_encap_entry *e)
 {
 	int max_encap_size = MLX5_CAP_ESW(priv->mdev, max_encap_header_size);
-	int ipv4_encap_size = ETH_HLEN +
-			      sizeof(struct iphdr) +
-			      e->tunnel_hlen;
 	struct ip_tunnel_key *tun_key = &e->tun_info.key;
 	struct net_device *out_dev;
 	struct neighbour *n = NULL;
 	struct flowi4 fl4 = {};
+	int ipv4_encap_size;
 	char *encap_header;
 	u8 nud_state, ttl;
 	struct iphdr *ip;
 	int err;
 
-	if (max_encap_size < ipv4_encap_size) {
-		mlx5_core_warn(priv->mdev, "encap size %d too big, max supported is %d\n",
-			       ipv4_encap_size, max_encap_size);
-		return -EOPNOTSUPP;
-	}
-
-	encap_header = kzalloc(ipv4_encap_size, GFP_KERNEL);
-	if (!encap_header)
-		return -ENOMEM;
-
 	/* add the IP fields */
 	fl4.flowi4_tos = tun_key->tos;
 	fl4.daddr = tun_key->u.ipv4.dst;
@@ -229,7 +217,22 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 	err = mlx5e_route_lookup_ipv4(priv, mirred_dev, &out_dev,
 				      &fl4, &n, &ttl);
 	if (err)
-		goto free_encap;
+		return err;
+
+	ipv4_encap_size =
+		ETH_HLEN +
+		sizeof(struct iphdr) +
+		e->tunnel_hlen;
+
+	if (max_encap_size < ipv4_encap_size) {
+		mlx5_core_warn(priv->mdev, "encap size %d too big, max supported is %d\n",
+			       ipv4_encap_size, max_encap_size);
+		return -EOPNOTSUPP;
+	}
+
+	encap_header = kzalloc(ipv4_encap_size, GFP_KERNEL);
+	if (!encap_header)
+		return -ENOMEM;
 
 	/* used by mlx5e_detach_encap to lookup a neigh hash table
 	 * entry in the neigh hash table when a user deletes a rule
@@ -308,28 +311,16 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 				    struct mlx5e_encap_entry *e)
 {
 	int max_encap_size = MLX5_CAP_ESW(priv->mdev, max_encap_header_size);
-	int ipv6_encap_size = ETH_HLEN +
-			      sizeof(struct ipv6hdr) +
-			      e->tunnel_hlen;
 	struct ip_tunnel_key *tun_key = &e->tun_info.key;
 	struct net_device *out_dev;
 	struct neighbour *n = NULL;
 	struct flowi6 fl6 = {};
 	struct ipv6hdr *ip6h;
+	int ipv6_encap_size;
 	char *encap_header;
 	u8 nud_state, ttl;
 	int err;
 
-	if (max_encap_size < ipv6_encap_size) {
-		mlx5_core_warn(priv->mdev, "encap size %d too big, max supported is %d\n",
-			       ipv6_encap_size, max_encap_size);
-		return -EOPNOTSUPP;
-	}
-
-	encap_header = kzalloc(ipv6_encap_size, GFP_KERNEL);
-	if (!encap_header)
-		return -ENOMEM;
-
 	ttl = tun_key->ttl;
 
 	fl6.flowlabel = ip6_make_flowinfo(RT_TOS(tun_key->tos), tun_key->label);
@@ -339,7 +330,22 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 	err = mlx5e_route_lookup_ipv6(priv, mirred_dev, &out_dev,
 				      &fl6, &n, &ttl);
 	if (err)
-		goto free_encap;
+		return err;
+
+	ipv6_encap_size =
+		ETH_HLEN +
+		sizeof(struct ipv6hdr) +
+		e->tunnel_hlen;
+
+	if (max_encap_size < ipv6_encap_size) {
+		mlx5_core_warn(priv->mdev, "encap size %d too big, max supported is %d\n",
+			       ipv6_encap_size, max_encap_size);
+		return -EOPNOTSUPP;
+	}
+
+	encap_header = kzalloc(ipv6_encap_size, GFP_KERNEL);
+	if (!encap_header)
+		return -ENOMEM;
 
 	/* used by mlx5e_detach_encap to lookup a neigh hash table
 	 * entry in the neigh hash table when a user deletes a rule
-- 
2.19.2

[net-next 10/11] net/mlx5e: Support VLAN encap ETH header generation

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

Support generation of native or tagged Ethernet header for encap
header, depending on provided net device.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/tc_tun.c   | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index bcf08f5abc2f..a6990744730a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -185,12 +185,23 @@ static char *gen_eth_tnl_hdr(char *buf, struct net_device *dev,
 			     u16 proto)
 {
 	struct ethhdr *eth = (struct ethhdr *)buf;
+	char *ip;
 
 	ether_addr_copy(eth->h_dest, e->h_dest);
 	ether_addr_copy(eth->h_source, dev->dev_addr);
-	eth->h_proto = htons(proto);
+	if (is_vlan_dev(dev)) {
+		struct vlan_hdr *vlan = (struct vlan_hdr *)
+					((char *)eth + ETH_HLEN);
+		ip = (char *)vlan + VLAN_HLEN;
+		eth->h_proto = vlan_dev_vlan_proto(dev);
+		vlan->h_vlan_TCI = htons(vlan_dev_vlan_id(dev));
+		vlan->h_vlan_encapsulated_proto = htons(proto);
+	} else {
+		eth->h_proto = htons(proto);
+		ip = (char *)eth + ETH_HLEN;
+	}
 
-	return (char *)eth + ETH_HLEN;
+	return ip;
 }
 
 int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
-- 
2.19.2

[net-next 11/11] net/mlx5e: Support tunnel encap over tagged Ethernet

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

Generate encap header depending on the routed device to support
native/tagged Ethernet header.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en/tc_tun.c   | 44 ++++++++++---------
 1 file changed, 24 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index a6990744730a..046948ead152 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -15,10 +15,6 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv,
 	struct net_device *uplink_dev, *uplink_upper;
 	bool dst_is_lag_dev;
 
-	/* we currently don't offload vlan on underlay */
-	if (is_vlan_dev(dev))
-		return -EOPNOTSUPP;
-
 	uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
 	uplink_upper = netdev_master_upper_dev_get(uplink_dev);
 	dst_is_lag_dev = (uplink_upper &&
@@ -30,12 +26,18 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv,
 	 * it's a LAG device, use the uplink
 	 */
 	if (!switchdev_port_same_parent_id(priv->netdev, dev) ||
-	    dst_is_lag_dev)
-		*out_dev = uplink_dev;
-	else if (!mlx5e_eswitch_rep(dev))
-		return -EOPNOTSUPP;
-	else
-		*out_dev = dev;
+	    dst_is_lag_dev) {
+		*route_dev = uplink_dev;
+		*out_dev = *route_dev;
+	} else {
+		*route_dev = dev;
+		if (is_vlan_dev(*route_dev))
+			*out_dev = uplink_dev;
+		else if (mlx5e_eswitch_rep(dev))
+			*out_dev = *route_dev;
+		else
+			return -EOPNOTSUPP;
+	}
 
 	return 0;
 }
@@ -43,6 +45,7 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv,
 static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv,
 				   struct net_device *mirred_dev,
 				   struct net_device **out_dev,
+				   struct net_device **route_dev,
 				   struct flowi4 *fl4,
 				   struct neighbour **out_n,
 				   u8 *out_ttl)
@@ -61,7 +64,7 @@ static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv,
 	return -EOPNOTSUPP;
 #endif
 
-	ret = get_route_and_out_devs(priv, rt->dst.dev, NULL, out_dev);
+	ret = get_route_and_out_devs(priv, rt->dst.dev, route_dev, out_dev);
 	if (ret < 0)
 		return ret;
 
@@ -87,6 +90,7 @@ static const char *mlx5e_netdev_kind(struct net_device *dev)
 static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv,
 				   struct net_device *mirred_dev,
 				   struct net_device **out_dev,
+				   struct net_device **route_dev,
 				   struct flowi6 *fl6,
 				   struct neighbour **out_n,
 				   u8 *out_ttl)
@@ -105,7 +109,7 @@ static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv,
 	if (!(*out_ttl))
 		*out_ttl = ip6_dst_hoplimit(dst);
 
-	ret = get_route_and_out_devs(priv, dst->dev, NULL, out_dev);
+	ret = get_route_and_out_devs(priv, dst->dev, route_dev, out_dev);
 	if (ret < 0)
 		return ret;
 #else
@@ -210,7 +214,7 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 {
 	int max_encap_size = MLX5_CAP_ESW(priv->mdev, max_encap_header_size);
 	struct ip_tunnel_key *tun_key = &e->tun_info.key;
-	struct net_device *out_dev;
+	struct net_device *out_dev, *route_dev;
 	struct neighbour *n = NULL;
 	struct flowi4 fl4 = {};
 	int ipv4_encap_size;
@@ -225,13 +229,13 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 	fl4.saddr = tun_key->u.ipv4.src;
 	ttl = tun_key->ttl;
 
-	err = mlx5e_route_lookup_ipv4(priv, mirred_dev, &out_dev,
+	err = mlx5e_route_lookup_ipv4(priv, mirred_dev, &out_dev, &route_dev,
 				      &fl4, &n, &ttl);
 	if (err)
 		return err;
 
 	ipv4_encap_size =
-		ETH_HLEN +
+		(is_vlan_dev(route_dev) ? VLAN_ETH_HLEN : ETH_HLEN) +
 		sizeof(struct iphdr) +
 		e->tunnel_hlen;
 
@@ -268,7 +272,7 @@ int mlx5e_tc_tun_create_header_ipv4(struct mlx5e_priv *priv,
 	read_unlock_bh(&n->lock);
 
 	/* add ethernet header */
-	ip = (struct iphdr *)gen_eth_tnl_hdr(encap_header, out_dev, e,
+	ip = (struct iphdr *)gen_eth_tnl_hdr(encap_header, route_dev, e,
 					     ETH_P_IP);
 
 	/* add ip header */
@@ -323,7 +327,7 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 {
 	int max_encap_size = MLX5_CAP_ESW(priv->mdev, max_encap_header_size);
 	struct ip_tunnel_key *tun_key = &e->tun_info.key;
-	struct net_device *out_dev;
+	struct net_device *out_dev, *route_dev;
 	struct neighbour *n = NULL;
 	struct flowi6 fl6 = {};
 	struct ipv6hdr *ip6h;
@@ -338,13 +342,13 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 	fl6.daddr = tun_key->u.ipv6.dst;
 	fl6.saddr = tun_key->u.ipv6.src;
 
-	err = mlx5e_route_lookup_ipv6(priv, mirred_dev, &out_dev,
+	err = mlx5e_route_lookup_ipv6(priv, mirred_dev, &out_dev, &route_dev,
 				      &fl6, &n, &ttl);
 	if (err)
 		return err;
 
 	ipv6_encap_size =
-		ETH_HLEN +
+		(is_vlan_dev(route_dev) ? VLAN_ETH_HLEN : ETH_HLEN) +
 		sizeof(struct ipv6hdr) +
 		e->tunnel_hlen;
 
@@ -381,7 +385,7 @@ int mlx5e_tc_tun_create_header_ipv6(struct mlx5e_priv *priv,
 	read_unlock_bh(&n->lock);
 
 	/* add ethernet header */
-	ip6h = (struct ipv6hdr *)gen_eth_tnl_hdr(encap_header, out_dev, e,
+	ip6h = (struct ipv6hdr *)gen_eth_tnl_hdr(encap_header, route_dev, e,
 						 ETH_P_IPV6);
 
 	/* add ip header */
-- 
2.19.2

Re: [net-next 05/11] net/mlx5e: Fail attempt to offload e-switch TC flows with egress upper devices

[Or Gerlitz]

On Thu, Dec 20, 2018 at 12:16 AM Saeed Mahameed <saeedm@mellanox.com> wrote:
>
> From: Eli Britstein <elibr@mellanox.com>
>
> We use the switchdev parent HW id helper to identify if the mirred device
> shares the same ASIC/port with the ingress device. This can get us wrong in
> the presence of upper devices (e.g vlan, team, etc) set over the HW devices
> (VF or uplink representors), b/c the switchdev ID is retrieved recursively.
>
> To fail offload attempts in such cases, we condition the check on the egress
> device to have not only the same switchdev ID but also the relevant mlx5 netdev ops.

The patch was made before we pushed vf lag.. need to fix it up a bit:

net/mlx5e: Fail attempt to offload e-switch TC flows with egress vlan devices

We use the switchdev parent HW id helper to identify if the mirred
device shares the same ASIC/port with the ingress device. This can get
us wrong in the presence of vlan upper devices set over the HW devices
(VF or uplink representors), b/c the switchdev ID is retrieved
recursively. To fail offload attempts in such cases, we condition the
check on the egress device to have not only the same switchdev ID but
also the relevant mlx5 netdev ops.


>
> Fixes: 03a9d11e6eeb ('net/mlx5e: Add TC drop and mirred/redirect action parsing for SRIOV offloads')
> Signed-off-by: Eli Britstein <elibr@mellanox.com>
> Reviewed-by: Roi Dayan <roid@mellanox.com>
> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
> ---
>  drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 9 +++++++++
>  drivers/net/ethernet/mellanox/mlx5/core/en_rep.h | 3 +++
>  drivers/net/ethernet/mellanox/mlx5/core/en_tc.c  | 3 +++
>  3 files changed, 15 insertions(+)
>
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
> index 91c3eb85f32e..f414f19c1159 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
> @@ -1316,6 +1316,15 @@ static const struct net_device_ops mlx5e_netdev_ops_uplink_rep = {
>         .ndo_get_vf_stats        = mlx5e_get_vf_stats,
>  };
>
> +bool mlx5e_eswitch_rep(struct net_device *netdev)
> +{
> +       if (netdev->netdev_ops == &mlx5e_netdev_ops_vf_rep ||
> +           netdev->netdev_ops == &mlx5e_netdev_ops_uplink_rep)
> +               return true;
> +
> +       return false;
> +}
> +
>  static void mlx5e_build_rep_params(struct net_device *netdev)
>  {
>         struct mlx5e_priv *priv = netdev_priv(netdev);
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
> index 5645d3cef1bb..edd722824697 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
> @@ -176,6 +176,9 @@ void mlx5e_rep_encap_entry_detach(struct mlx5e_priv *priv,
>                                   struct mlx5e_encap_entry *e);
>
>  void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv);
> +
> +bool mlx5e_eswitch_rep(struct net_device *netdev);
> +
>  #else /* CONFIG_MLX5_ESWITCH */
>  static inline bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv) { return false; }
>  static inline int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) { return 0; }
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
> index c1a9120412b8..a4c6287dfd65 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
> @@ -2578,6 +2578,9 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
>                                 struct net_device *uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
>                                 struct net_device *uplink_upper = netdev_master_upper_dev_get(uplink_dev);
>
> +                               if (!mlx5e_eswitch_rep(out_dev))
> +                                       return -EOPNOTSUPP;
> +

no, out_dev can be the team/bond device and this way we break vf lag

>                                 if (uplink_upper &&
>                                     netif_is_lag_master(uplink_upper) &&
>                                     uplink_upper == out_dev)

should move it here, after we realize that and potentially made
out_dev be the uplink rep

Otherwise, Acked- by: Or Gerlitz <ogerlitz@mellanox.com>

Re: [pull request][net-next 00/11] Mellanox, mlx5 updates 2018-12-19

[David Miller]

From: Saeed Mahameed <saeedm@mellanox.com>
Date: Wed, 19 Dec 2018 13:02:33 -0800

> This series includes some misc updates and vlan encap tunnel support.
> For more information please see tag log below.

I'll wait for the respin as you deal with the feedback from Or on patch #5.