[pull request][net-next 00/12] mlx5 updates 2020-07-02

[pull request][net-next 00/12] mlx5 updates 2020-07-02

[Saeed Mahameed]

Hi Dave, Jakub

This series adds misc updates.
For more information please see tag log below.

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

Thanks,
Saeed.

---
The following changes since commit 8c8278a5b1a81e099ba883d8a0f9e3df9bdb1a74:

  Merge branch 'sfc-prerequisites-for-EF100-driver-part-3' (2020-07-02 14:47:40 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5-updates-2020-07-02

for you to fetch changes up to e62055642797a6de80f3576c18e212cbbf5b4361:

  net/mlx5e: Enhance TX timeout recovery (2020-07-02 21:05:19 -0700)

----------------------------------------------------------------
mlx5-updates-2020-07-02

Rx and Tx devlink health reporters enhancements.

1) Code cleanup
2) devlink output format improvements
3) Print more useful info on devlink health diagnose output
4) TX timeout recovery, on a single SQ recover failure, stop the loop
and reset all rings (re-open netdev).

----------------------------------------------------------------
Aya Levin (11):
      net/mlx5e: Add a flush timeout define
      net/mlx5e: Remove redundant RQ state query
      net/mlx5e: Align RX/TX reporters diagnose output format
      net/mlx5e: Move RQ helpers to txrx.h
      net/mlx5e: Add helper to get RQ WQE's head
      net/mlx5e: Add helper to get the RQ WQE counter
      net/mlx5e: Rename reporter's helpers
      net/mlx5e: Enhance CQ data on diagnose output
      net/mlx5e: Add EQ info to TX/RX reporter's diagnose
      net/mlx5e: Enhance ICOSQ data on RX reporter's diagnose
      net/mlx5e: Enhance TX timeout recovery

Eran Ben Elisha (1):
      net/mlx5e: Change reporters create functions to return void

 drivers/net/ethernet/mellanox/mlx5/core/en.h       |  20 ---
 .../net/ethernet/mellanox/mlx5/core/en/health.c    |  58 +++++++--
 .../net/ethernet/mellanox/mlx5/core/en/health.h    |  16 ++-
 .../ethernet/mellanox/mlx5/core/en/reporter_rx.c   | 145 ++++++++++++++-------
 .../ethernet/mellanox/mlx5/core/en/reporter_tx.c   |  80 ++++++++----
 drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h  |  46 ++++++-
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c  |  14 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c    |   4 +-
 drivers/net/ethernet/mellanox/mlx5/core/wq.h       |  15 +++
 9 files changed, 266 insertions(+), 132 deletions(-)

[net-next 01/12] net/mlx5e: Change reporters create functions to return void

[Saeed Mahameed]

From: Eran Ben Elisha <eranbe@mellanox.com>

Creation of devlink health reporters is not fatal for mlx5e instance load.
In case of error in reporter's creation, the return value is ignored.
Change all reporters creation functions to return void.

In addition, with this change, a failure in creating a reporter, will not
prevent the driver from trying to create the next reporter in the list.

Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Reviewed-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/health.c   | 15 +++------------
 .../net/ethernet/mellanox/mlx5/core/en/health.h   |  6 +++---
 .../ethernet/mellanox/mlx5/core/en/reporter_rx.c  |  5 ++---
 .../ethernet/mellanox/mlx5/core/en/reporter_tx.c  |  5 ++---
 4 files changed, 10 insertions(+), 21 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
index 7283443868f3..d0625ee923d6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -97,19 +97,10 @@ int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *
 	return 0;
 }
 
-int mlx5e_health_create_reporters(struct mlx5e_priv *priv)
+void mlx5e_health_create_reporters(struct mlx5e_priv *priv)
 {
-	int err;
-
-	err = mlx5e_reporter_tx_create(priv);
-	if (err)
-		return err;
-
-	err = mlx5e_reporter_rx_create(priv);
-	if (err)
-		return err;
-
-	return 0;
+	mlx5e_reporter_tx_create(priv);
+	mlx5e_reporter_rx_create(priv);
 }
 
 void mlx5e_health_destroy_reporters(struct mlx5e_priv *priv)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 38f97f79ef16..895d03d56c9d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -16,7 +16,7 @@ static inline bool cqe_syndrome_needs_recover(u8 syndrome)
 	       syndrome == MLX5_CQE_SYNDROME_WR_FLUSH_ERR;
 }
 
-int mlx5e_reporter_tx_create(struct mlx5e_priv *priv);
+void mlx5e_reporter_tx_create(struct mlx5e_priv *priv);
 void mlx5e_reporter_tx_destroy(struct mlx5e_priv *priv);
 void mlx5e_reporter_tx_err_cqe(struct mlx5e_txqsq *sq);
 int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq);
@@ -26,7 +26,7 @@ int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *
 int mlx5e_reporter_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name);
 int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg *fmsg);
 
-int mlx5e_reporter_rx_create(struct mlx5e_priv *priv);
+void mlx5e_reporter_rx_create(struct mlx5e_priv *priv);
 void mlx5e_reporter_rx_destroy(struct mlx5e_priv *priv);
 void mlx5e_reporter_icosq_cqe_err(struct mlx5e_icosq *icosq);
 void mlx5e_reporter_rq_cqe_err(struct mlx5e_rq *rq);
@@ -46,7 +46,7 @@ int mlx5e_health_recover_channels(struct mlx5e_priv *priv);
 int mlx5e_health_report(struct mlx5e_priv *priv,
 			struct devlink_health_reporter *reporter, char *err_str,
 			struct mlx5e_err_ctx *err_ctx);
-int mlx5e_health_create_reporters(struct mlx5e_priv *priv);
+void mlx5e_health_create_reporters(struct mlx5e_priv *priv);
 void mlx5e_health_destroy_reporters(struct mlx5e_priv *priv);
 void mlx5e_health_channels_update(struct mlx5e_priv *priv);
 int mlx5e_health_rsc_fmsg_dump(struct mlx5e_priv *priv, struct mlx5_rsc_key *key,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index c209579fc213..5161a1954577 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -563,7 +563,7 @@ static const struct devlink_health_reporter_ops mlx5_rx_reporter_ops = {
 
 #define MLX5E_REPORTER_RX_GRACEFUL_PERIOD 500
 
-int mlx5e_reporter_rx_create(struct mlx5e_priv *priv)
+void mlx5e_reporter_rx_create(struct mlx5e_priv *priv)
 {
 	struct devlink *devlink = priv_to_devlink(priv->mdev);
 	struct devlink_health_reporter *reporter;
@@ -575,10 +575,9 @@ int mlx5e_reporter_rx_create(struct mlx5e_priv *priv)
 	if (IS_ERR(reporter)) {
 		netdev_warn(priv->netdev, "Failed to create rx reporter, err = %ld\n",
 			    PTR_ERR(reporter));
-		return PTR_ERR(reporter);
+		return;
 	}
 	priv->rx_reporter = reporter;
-	return 0;
 }
 
 void mlx5e_reporter_rx_destroy(struct mlx5e_priv *priv)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
index 9805fc085512..b95dc15f23b9 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -406,7 +406,7 @@ static const struct devlink_health_reporter_ops mlx5_tx_reporter_ops = {
 
 #define MLX5_REPORTER_TX_GRACEFUL_PERIOD 500
 
-int mlx5e_reporter_tx_create(struct mlx5e_priv *priv)
+void mlx5e_reporter_tx_create(struct mlx5e_priv *priv)
 {
 	struct devlink_health_reporter *reporter;
 	struct mlx5_core_dev *mdev = priv->mdev;
@@ -421,10 +421,9 @@ int mlx5e_reporter_tx_create(struct mlx5e_priv *priv)
 		netdev_warn(priv->netdev,
 			    "Failed to create tx reporter, err = %ld\n",
 			    PTR_ERR(reporter));
-		return PTR_ERR(reporter);
+		return;
 	}
 	priv->tx_reporter = reporter;
-	return 0;
 }
 
 void mlx5e_reporter_tx_destroy(struct mlx5e_priv *priv)
-- 
2.26.2

[net-next 02/12] net/mlx5e: Add a flush timeout define

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

During queue's recovery, driver waits for flush. The flush timeout is
set to 2 seconds. Add a define for this value for the benefit of RX and
TX reporters.

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/health.h      | 1 +
 drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c | 3 ++-
 drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c | 3 ++-
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 895d03d56c9d..2938553a7606 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -33,6 +33,7 @@ void mlx5e_reporter_rq_cqe_err(struct mlx5e_rq *rq);
 void mlx5e_reporter_rx_timeout(struct mlx5e_rq *rq);
 
 #define MLX5E_REPORTER_PER_Q_MAX_LEN 256
+#define MLX5E_REPORTER_FLUSH_TIMEOUT_MSEC 2000
 
 struct mlx5e_err_ctx {
 	int (*recover)(void *ctx);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index 5161a1954577..495a3e6bf82b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -29,7 +29,8 @@ static int mlx5e_query_rq_state(struct mlx5_core_dev *dev, u32 rqn, u8 *state)
 
 static int mlx5e_wait_for_icosq_flush(struct mlx5e_icosq *icosq)
 {
-	unsigned long exp_time = jiffies + msecs_to_jiffies(2000);
+	unsigned long exp_time = jiffies +
+				 msecs_to_jiffies(MLX5E_REPORTER_FLUSH_TIMEOUT_MSEC);
 
 	while (time_before(jiffies, exp_time)) {
 		if (icosq->cc == icosq->pc)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
index b95dc15f23b9..6eb2971231d8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -5,7 +5,8 @@
 
 static int mlx5e_wait_for_sq_flush(struct mlx5e_txqsq *sq)
 {
-	unsigned long exp_time = jiffies + msecs_to_jiffies(2000);
+	unsigned long exp_time = jiffies +
+				 msecs_to_jiffies(MLX5E_REPORTER_FLUSH_TIMEOUT_MSEC);
 
 	while (time_before(jiffies, exp_time)) {
 		if (sq->cc == sq->pc)
-- 
2.26.2

[net-next 03/12] net/mlx5e: Remove redundant RQ state query

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

When received a CQE error, the driver inspect the syndrome given by the
firmware. RQ recovery is initiated only as a result of a fatal syndrome;
syndrome which set the RQ into an error state. Hence no need to query
the RQ state at the beginning of the recovery process. Add additional
debug prints before recovering.

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../mellanox/mlx5/core/en/reporter_rx.c        | 18 +-----------------
 .../net/ethernet/mellanox/mlx5/core/en/txrx.h  |  6 +++---
 .../net/ethernet/mellanox/mlx5/core/en_rx.c    |  4 +++-
 3 files changed, 7 insertions(+), 21 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index 495a3e6bf82b..b8b32aef1363 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -124,25 +124,9 @@ static int mlx5e_rq_to_ready(struct mlx5e_rq *rq, int curr_state)
 
 static int mlx5e_rx_reporter_err_rq_cqe_recover(void *ctx)
 {
-	struct mlx5_core_dev *mdev;
-	struct net_device *dev;
-	struct mlx5e_rq *rq;
-	u8 state;
+	struct mlx5e_rq *rq = ctx;
 	int err;
 
-	rq = ctx;
-	mdev = rq->mdev;
-	dev = rq->netdev;
-	err = mlx5e_query_rq_state(mdev, rq->rqn, &state);
-	if (err) {
-		netdev_err(dev, "Failed to query RQ 0x%x state. err = %d\n",
-			   rq->rqn, err);
-		goto out;
-	}
-
-	if (state != MLX5_RQC_STATE_ERR)
-		goto out;
-
 	mlx5e_deactivate_rq(rq);
 	mlx5e_free_rx_descs(rq);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
index e9d4a61b6bbb..be7692897fc1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
@@ -268,7 +268,7 @@ static inline void mlx5e_rqwq_reset(struct mlx5e_rq *rq)
 	}
 }
 
-static inline void mlx5e_dump_error_cqe(struct mlx5e_cq *cq, u32 sqn,
+static inline void mlx5e_dump_error_cqe(struct mlx5e_cq *cq, u32 qn,
 					struct mlx5_err_cqe *err_cqe)
 {
 	struct mlx5_cqwq *wq = &cq->wq;
@@ -277,8 +277,8 @@ static inline void mlx5e_dump_error_cqe(struct mlx5e_cq *cq, u32 sqn,
 	ci = mlx5_cqwq_ctr2ix(wq, wq->cc - 1);
 
 	netdev_err(cq->channel->netdev,
-		   "Error cqe on cqn 0x%x, ci 0x%x, sqn 0x%x, opcode 0x%x, syndrome 0x%x, vendor syndrome 0x%x\n",
-		   cq->mcq.cqn, ci, sqn,
+		   "Error cqe on cqn 0x%x, ci 0x%x, qn 0x%x, opcode 0x%x, syndrome 0x%x, vendor syndrome 0x%x\n",
+		   cq->mcq.cqn, ci, qn,
 		   get_cqe_opcode((struct mlx5_cqe64 *)err_cqe),
 		   err_cqe->syndrome, err_cqe->vendor_err_synd);
 	mlx5_dump_err_cqe(cq->mdev, err_cqe);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 8b42f729a4f7..350f9c54e508 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1187,8 +1187,10 @@ static void trigger_report(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
 	struct mlx5_err_cqe *err_cqe = (struct mlx5_err_cqe *)cqe;
 
 	if (cqe_syndrome_needs_recover(err_cqe->syndrome) &&
-	    !test_and_set_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state))
+	    !test_and_set_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state)) {
+		mlx5e_dump_error_cqe(&rq->cq, rq->rqn, err_cqe);
 		queue_work(rq->channel->priv->wq, &rq->recover_work);
+	}
 }
 
 void mlx5e_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
-- 
2.26.2

[net-next 04/12] net/mlx5e: Align RX/TX reporters diagnose output format

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Change the hierarchy of the RX reporter 'Common config' in the diagnose
output to match the 'Common config' of the TX reporter which reflects
that CQ is a helper to the traffic queues.

Before:
$ devlink health diagnose pci/0000:00:0b.0 reporter rx
Common config:
    RQ:
      type: 2 stride size: 2048 size: 8
    CQ:
      stride size: 64 size: 1024
    RQs:
    ...

After:
$ devlink health diagnose pci/0000:00:0b.0 reporter rx
Common config:
    RQ:
      type: 2 stride size: 2048 size: 8
      CQ:
        stride size: 64 size: 1024
    RQs:
    ...

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index b8b32aef1363..f1edde1ab8bc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -284,11 +284,11 @@ static int mlx5e_rx_reporter_diagnose(struct devlink_health_reporter *reporter,
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_reporter_cq_common_diagnose(&generic_rq->cq, fmsg);
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_cq_common_diagnose(&generic_rq->cq, fmsg);
+	err = mlx5e_reporter_named_obj_nest_end(fmsg);
 	if (err)
 		goto unlock;
 
-- 
2.26.2

[net-next 05/12] net/mlx5e: Move RQ helpers to txrx.h

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Use txrx.h to contain helper function regarding TX/RX. In the coming
patches, I will add more RQ helpers.

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  | 20 -------------------
 .../mellanox/mlx5/core/en/reporter_rx.c       |  1 +
 .../net/ethernet/mellanox/mlx5/core/en/txrx.h | 20 +++++++++++++++++++
 3 files changed, 21 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 2957edb7e0b7..c44669102626 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -852,26 +852,6 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq, int napi_budget);
 int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget);
 void mlx5e_free_txqsq_descs(struct mlx5e_txqsq *sq);
 
-static inline u32 mlx5e_rqwq_get_size(struct mlx5e_rq *rq)
-{
-	switch (rq->wq_type) {
-	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
-		return mlx5_wq_ll_get_size(&rq->mpwqe.wq);
-	default:
-		return mlx5_wq_cyc_get_size(&rq->wqe.wq);
-	}
-}
-
-static inline u32 mlx5e_rqwq_get_cur_sz(struct mlx5e_rq *rq)
-{
-	switch (rq->wq_type) {
-	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
-		return rq->mpwqe.wq.cur_sz;
-	default:
-		return rq->wqe.wq.cur_sz;
-	}
-}
-
 bool mlx5e_check_fragmented_striding_rq_cap(struct mlx5_core_dev *mdev);
 bool mlx5e_striding_rq_possible(struct mlx5_core_dev *mdev,
 				struct mlx5e_params *params);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index f1edde1ab8bc..bfdf9c185f02 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -3,6 +3,7 @@
 
 #include "health.h"
 #include "params.h"
+#include "txrx.h"
 
 static int mlx5e_query_rq_state(struct mlx5_core_dev *dev, u32 rqn, u8 *state)
 {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
index be7692897fc1..ed9e0b8a6f9e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
@@ -284,6 +284,26 @@ static inline void mlx5e_dump_error_cqe(struct mlx5e_cq *cq, u32 qn,
 	mlx5_dump_err_cqe(cq->mdev, err_cqe);
 }
 
+static inline u32 mlx5e_rqwq_get_size(struct mlx5e_rq *rq)
+{
+	switch (rq->wq_type) {
+	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
+		return mlx5_wq_ll_get_size(&rq->mpwqe.wq);
+	default:
+		return mlx5_wq_cyc_get_size(&rq->wqe.wq);
+	}
+}
+
+static inline u32 mlx5e_rqwq_get_cur_sz(struct mlx5e_rq *rq)
+{
+	switch (rq->wq_type) {
+	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
+		return rq->mpwqe.wq.cur_sz;
+	default:
+		return rq->wqe.wq.cur_sz;
+	}
+}
+
 /* SW parser related functions */
 
 struct mlx5e_swp_spec {
-- 
2.26.2

[net-next 06/12] net/mlx5e: Add helper to get RQ WQE's head

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Add helper which retrieves the RQ WQE's head. Use this helper in RX
reporter diagnose callback.

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/reporter_rx.c   |  5 +----
 drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h      | 10 ++++++++++
 drivers/net/ethernet/mellanox/mlx5/core/wq.h           |  4 ++++
 3 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index bfdf9c185f02..f0e639ef4ec5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -181,7 +181,6 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 						   struct devlink_fmsg *fmsg)
 {
 	struct mlx5e_priv *priv = rq->channel->priv;
-	struct mlx5e_params *params;
 	struct mlx5e_icosq *icosq;
 	u8 icosq_hw_state;
 	int wqes_sz;
@@ -189,7 +188,6 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 	u16 wq_head;
 	int err;
 
-	params = &priv->channels.params;
 	icosq = &rq->channel->icosq;
 	err = mlx5e_query_rq_state(priv->mdev, rq->rqn, &hw_state);
 	if (err)
@@ -200,8 +198,7 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 		return err;
 
 	wqes_sz = mlx5e_rqwq_get_cur_sz(rq);
-	wq_head = params->rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ ?
-		  rq->mpwqe.wq.head : mlx5_wq_cyc_get_head(&rq->wqe.wq);
+	wq_head = mlx5e_rqwq_get_head(rq);
 
 	err = devlink_fmsg_obj_nest_start(fmsg);
 	if (err)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
index ed9e0b8a6f9e..d787e8fc2e99 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
@@ -304,6 +304,16 @@ static inline u32 mlx5e_rqwq_get_cur_sz(struct mlx5e_rq *rq)
 	}
 }
 
+static inline u16 mlx5e_rqwq_get_head(struct mlx5e_rq *rq)
+{
+	switch (rq->wq_type) {
+	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
+		return mlx5_wq_ll_get_head(&rq->mpwqe.wq);
+	default:
+		return mlx5_wq_cyc_get_head(&rq->wqe.wq);
+	}
+}
+
 /* SW parser related functions */
 
 struct mlx5e_swp_spec {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/wq.h b/drivers/net/ethernet/mellanox/mlx5/core/wq.h
index 4cadc336593f..27dece35df7e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/wq.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/wq.h
@@ -290,4 +290,8 @@ static inline void mlx5_wq_ll_update_db_record(struct mlx5_wq_ll *wq)
 	*wq->db = cpu_to_be32(wq->wqe_ctr);
 }
 
+static inline u16 mlx5_wq_ll_get_head(struct mlx5_wq_ll *wq)
+{
+	return wq->head;
+}
 #endif /* __MLX5_WQ_H__ */
-- 
2.26.2

[net-next 07/12] net/mlx5e: Add helper to get the RQ WQE counter

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Add a helper which retrieves the RQ's WQE counter. Use this helper in
the RX reporter diagnose callback.

$ devlink health diagnose pci/0000:00:0b.0 reporter rx
Common config:
  RQ:
     type: 2 stride size: 2048 size: 8
     CQ:
      stride size: 64 size: 1024
RQs:
   channel ix: 0 rqn: 2113 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7 ICOSQ HW state: 1
   CQ:
    cqn: 1032 HW status: 0
   channel ix: 1 rqn: 2118 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7 ICOSQ HW state: 1
   CQ:
    cqn: 1036 HW status: 0

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/reporter_rx.c  |  6 ++++++
 drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h     | 10 ++++++++++
 drivers/net/ethernet/mellanox/mlx5/core/wq.h          | 11 +++++++++++
 3 files changed, 27 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index f0e639ef4ec5..bec804295c52 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -183,6 +183,7 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 	struct mlx5e_priv *priv = rq->channel->priv;
 	struct mlx5e_icosq *icosq;
 	u8 icosq_hw_state;
+	u16 wqe_counter;
 	int wqes_sz;
 	u8 hw_state;
 	u16 wq_head;
@@ -199,6 +200,7 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 
 	wqes_sz = mlx5e_rqwq_get_cur_sz(rq);
 	wq_head = mlx5e_rqwq_get_head(rq);
+	wqe_counter = mlx5e_rqwq_get_wqe_counter(rq);
 
 	err = devlink_fmsg_obj_nest_start(fmsg);
 	if (err)
@@ -220,6 +222,10 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 	if (err)
 		return err;
 
+	err = devlink_fmsg_u32_pair_put(fmsg, "WQE counter", wqe_counter);
+	if (err)
+		return err;
+
 	err = devlink_fmsg_u32_pair_put(fmsg, "posted WQEs", wqes_sz);
 	if (err)
 		return err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
index d787e8fc2e99..cf425a60cddc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
@@ -314,6 +314,16 @@ static inline u16 mlx5e_rqwq_get_head(struct mlx5e_rq *rq)
 	}
 }
 
+static inline u16 mlx5e_rqwq_get_wqe_counter(struct mlx5e_rq *rq)
+{
+	switch (rq->wq_type) {
+	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
+		return mlx5_wq_ll_get_counter(&rq->mpwqe.wq);
+	default:
+		return mlx5_wq_cyc_get_counter(&rq->wqe.wq);
+	}
+}
+
 /* SW parser related functions */
 
 struct mlx5e_swp_spec {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/wq.h b/drivers/net/ethernet/mellanox/mlx5/core/wq.h
index 27dece35df7e..e5c4dcd1425e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/wq.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/wq.h
@@ -172,6 +172,11 @@ static inline int mlx5_wq_cyc_cc_bigger(u16 cc1, u16 cc2)
 	return !equal && !smaller;
 }
 
+static inline u16 mlx5_wq_cyc_get_counter(struct mlx5_wq_cyc *wq)
+{
+	return wq->wqe_ctr;
+}
+
 static inline u32 mlx5_cqwq_get_size(struct mlx5_cqwq *wq)
 {
 	return wq->fbc.sz_m1 + 1;
@@ -294,4 +299,10 @@ static inline u16 mlx5_wq_ll_get_head(struct mlx5_wq_ll *wq)
 {
 	return wq->head;
 }
+
+static inline u16 mlx5_wq_ll_get_counter(struct mlx5_wq_ll *wq)
+{
+	return wq->wqe_ctr;
+}
+
 #endif /* __MLX5_WQ_H__ */
-- 
2.26.2

[net-next 08/12] net/mlx5e: Rename reporter's helpers

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Change prefix to match resident file:
%s/mlx5e_reporter_cq_diagnose/mlx5e_health_cq_diag_fmsg
%s/mlx5e_reporter_cq_common_diagnose/mlx5e_health_cq_common_diag_fmsg
%s/mlx5e_reporter_named_obj_nest_start/mlx5e_health_fmsg_named_obj_nest_start
%s/mlx5e_reporter_named_obj_nest_end/mlx5e_health_fmsg_named_obj_nest_end

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en/health.c   | 20 ++++----
 .../ethernet/mellanox/mlx5/core/en/health.h   |  8 ++--
 .../mellanox/mlx5/core/en/reporter_rx.c       | 48 +++++++++----------
 .../mellanox/mlx5/core/en/reporter_tx.c       | 32 ++++++-------
 4 files changed, 54 insertions(+), 54 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
index d0625ee923d6..1b735b54b3ab 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -5,7 +5,7 @@
 #include "lib/eq.h"
 #include "lib/mlx5.h"
 
-int mlx5e_reporter_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name)
+int mlx5e_health_fmsg_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name)
 {
 	int err;
 
@@ -20,7 +20,7 @@ int mlx5e_reporter_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name)
 	return 0;
 }
 
-int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg *fmsg)
+int mlx5e_health_fmsg_named_obj_nest_end(struct devlink_fmsg *fmsg)
 {
 	int err;
 
@@ -35,7 +35,7 @@ int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg *fmsg)
 	return 0;
 }
 
-int mlx5e_reporter_cq_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
+int mlx5e_health_cq_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
 {
 	struct mlx5e_priv *priv = cq->channel->priv;
 	u32 out[MLX5_ST_SZ_DW(query_cq_out)] = {};
@@ -50,7 +50,7 @@ int mlx5e_reporter_cq_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
 	cqc = MLX5_ADDR_OF(query_cq_out, out, cq_context);
 	hw_status = MLX5_GET(cqc, cqc, status);
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "CQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "CQ");
 	if (err)
 		return err;
 
@@ -62,14 +62,14 @@ int mlx5e_reporter_cq_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
 	return 0;
 }
 
-int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
+int mlx5e_health_cq_common_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
 {
 	u8 cq_log_stride;
 	u32 cq_sz;
@@ -78,7 +78,7 @@ int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *
 	cq_sz = mlx5_cqwq_get_size(&cq->wq);
 	cq_log_stride = mlx5_cqwq_get_log_stride_size(&cq->wq);
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "CQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "CQ");
 	if (err)
 		return err;
 
@@ -90,7 +90,7 @@ int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
@@ -282,7 +282,7 @@ int mlx5e_health_queue_dump(struct mlx5e_priv *priv, struct devlink_fmsg *fmsg,
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, lbl);
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, lbl);
 	if (err)
 		return err;
 
@@ -294,7 +294,7 @@ int mlx5e_health_queue_dump(struct mlx5e_priv *priv, struct devlink_fmsg *fmsg,
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 2938553a7606..6e48518d3d5b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -21,10 +21,10 @@ void mlx5e_reporter_tx_destroy(struct mlx5e_priv *priv);
 void mlx5e_reporter_tx_err_cqe(struct mlx5e_txqsq *sq);
 int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq);
 
-int mlx5e_reporter_cq_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg);
-int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg);
-int mlx5e_reporter_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name);
-int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg *fmsg);
+int mlx5e_health_cq_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg);
+int mlx5e_health_cq_common_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg);
+int mlx5e_health_fmsg_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name);
+int mlx5e_health_fmsg_named_obj_nest_end(struct devlink_fmsg *fmsg);
 
 void mlx5e_reporter_rx_create(struct mlx5e_priv *priv);
 void mlx5e_reporter_rx_destroy(struct mlx5e_priv *priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index bec804295c52..4e1a01d871b7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -238,7 +238,7 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_cq_diagnose(&rq->cq, fmsg);
+	err = mlx5e_health_cq_diag_fmsg(&rq->cq, fmsg);
 	if (err)
 		return err;
 
@@ -268,11 +268,11 @@ static int mlx5e_rx_reporter_diagnose(struct devlink_health_reporter *reporter,
 	rq_sz = mlx5e_rqwq_get_size(generic_rq);
 	rq_stride = BIT(mlx5e_mpwqe_get_log_stride_size(priv->mdev, params, NULL));
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "Common config");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "Common config");
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "RQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "RQ");
 	if (err)
 		goto unlock;
 
@@ -288,15 +288,15 @@ static int mlx5e_rx_reporter_diagnose(struct devlink_health_reporter *reporter,
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_cq_common_diagnose(&generic_rq->cq, fmsg);
+	err = mlx5e_health_cq_common_diag_fmsg(&generic_rq->cq, fmsg);
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		goto unlock;
 
@@ -329,7 +329,7 @@ static int mlx5e_rx_reporter_dump_icosq(struct mlx5e_priv *priv, struct devlink_
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
 		return 0;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "SX Slice");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "SX Slice");
 	if (err)
 		return err;
 
@@ -339,15 +339,15 @@ static int mlx5e_rx_reporter_dump_icosq(struct mlx5e_priv *priv, struct devlink_
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "ICOSQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "ICOSQ");
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "QPC");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "QPC");
 	if (err)
 		return err;
 
@@ -359,11 +359,11 @@ static int mlx5e_rx_reporter_dump_icosq(struct mlx5e_priv *priv, struct devlink_
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "send_buff");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "send_buff");
 	if (err)
 		return err;
 
@@ -374,11 +374,11 @@ static int mlx5e_rx_reporter_dump_icosq(struct mlx5e_priv *priv, struct devlink_
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	return mlx5e_reporter_named_obj_nest_end(fmsg);
+	return mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 }
 
 static int mlx5e_rx_reporter_dump_rq(struct mlx5e_priv *priv, struct devlink_fmsg *fmsg,
@@ -391,7 +391,7 @@ static int mlx5e_rx_reporter_dump_rq(struct mlx5e_priv *priv, struct devlink_fms
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
 		return 0;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "RX Slice");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "RX Slice");
 	if (err)
 		return err;
 
@@ -401,15 +401,15 @@ static int mlx5e_rx_reporter_dump_rq(struct mlx5e_priv *priv, struct devlink_fms
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "RQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "RQ");
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "QPC");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "QPC");
 	if (err)
 		return err;
 
@@ -421,11 +421,11 @@ static int mlx5e_rx_reporter_dump_rq(struct mlx5e_priv *priv, struct devlink_fms
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "receive_buff");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "receive_buff");
 	if (err)
 		return err;
 
@@ -435,11 +435,11 @@ static int mlx5e_rx_reporter_dump_rq(struct mlx5e_priv *priv, struct devlink_fms
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	return mlx5e_reporter_named_obj_nest_end(fmsg);
+	return mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 }
 
 static int mlx5e_rx_reporter_dump_all_rqs(struct mlx5e_priv *priv,
@@ -451,7 +451,7 @@ static int mlx5e_rx_reporter_dump_all_rqs(struct mlx5e_priv *priv,
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
 		return 0;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "RX Slice");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "RX Slice");
 	if (err)
 		return err;
 
@@ -461,7 +461,7 @@ static int mlx5e_rx_reporter_dump_all_rqs(struct mlx5e_priv *priv,
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
index 6eb2971231d8..9f712ff2faf9 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -166,7 +166,7 @@ mlx5e_tx_reporter_build_diagnose_output(struct devlink_fmsg *fmsg,
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_cq_diagnose(&sq->cq, fmsg);
+	err = mlx5e_health_cq_diag_fmsg(&sq->cq, fmsg);
 	if (err)
 		return err;
 
@@ -195,11 +195,11 @@ static int mlx5e_tx_reporter_diagnose(struct devlink_health_reporter *reporter,
 	sq_sz = mlx5_wq_cyc_get_size(&generic_sq->wq);
 	sq_stride = MLX5_SEND_WQE_BB;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "Common Config");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "Common Config");
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "SQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "SQ");
 	if (err)
 		goto unlock;
 
@@ -211,15 +211,15 @@ static int mlx5e_tx_reporter_diagnose(struct devlink_health_reporter *reporter,
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_cq_common_diagnose(&generic_sq->cq, fmsg);
+	err = mlx5e_health_cq_common_diag_fmsg(&generic_sq->cq, fmsg);
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		goto unlock;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		goto unlock;
 
@@ -257,7 +257,7 @@ static int mlx5e_tx_reporter_dump_sq(struct mlx5e_priv *priv, struct devlink_fms
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
 		return 0;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "SX Slice");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "SX Slice");
 	if (err)
 		return err;
 
@@ -267,15 +267,15 @@ static int mlx5e_tx_reporter_dump_sq(struct mlx5e_priv *priv, struct devlink_fms
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "SQ");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "SQ");
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "QPC");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "QPC");
 	if (err)
 		return err;
 
@@ -287,11 +287,11 @@ static int mlx5e_tx_reporter_dump_sq(struct mlx5e_priv *priv, struct devlink_fms
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "send_buff");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "send_buff");
 	if (err)
 		return err;
 
@@ -301,11 +301,11 @@ static int mlx5e_tx_reporter_dump_sq(struct mlx5e_priv *priv, struct devlink_fms
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-	return mlx5e_reporter_named_obj_nest_end(fmsg);
+	return mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 }
 
 static int mlx5e_tx_reporter_dump_all_sqs(struct mlx5e_priv *priv,
@@ -317,7 +317,7 @@ static int mlx5e_tx_reporter_dump_all_sqs(struct mlx5e_priv *priv,
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
 		return 0;
 
-	err = mlx5e_reporter_named_obj_nest_start(fmsg, "SX Slice");
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "SX Slice");
 	if (err)
 		return err;
 
@@ -327,7 +327,7 @@ static int mlx5e_tx_reporter_dump_all_sqs(struct mlx5e_priv *priv,
 	if (err)
 		return err;
 
-	err = mlx5e_reporter_named_obj_nest_end(fmsg);
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
 
-- 
2.26.2

[net-next 09/12] net/mlx5e: Enhance CQ data on diagnose output

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Add CQ's consumer index and size to the CQ's diagnose output retruved on
RX/TX reporter diadgnose.

$ devlink health diagnose pci/0000:00:0b.0 reporter rx
Common config:
    RQ:
      type: 2 stride size: 2048 size: 8
      CQ:
        stride size: 64 size: 1024
RQs:
    channel ix: 0 rqn: 2413 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7 ICOSQ HW state: 1
    CQ:
      cqn: 1032 HW status: 0 ci: 0 size: 1024
    channel ix: 1 rqn: 2418 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7 ICOSQ HW state: 1
    CQ:
      cqn: 1036 HW status: 0 ci: 0 size: 1024

$ devlink health diagnose pci/0000:00:0b.0 reporter tx
Common Config:
    SQ:
      stride size: 64 size: 1024
      CQ:
        stride size: 64 size: 1024
SQs:
    channel ix: 0 tc: 0 txq ix: 0 sqn: 2412 HW state: 1 stopped: false cc: 0 pc: 0
    CQ:
      cqn: 1030 HW status: 0 ci: 0 size: 1024
    channel ix: 1 tc: 0 txq ix: 1 sqn: 2417 HW state: 1 stopped: false cc: 5 pc: 5
    CQ:
      cqn: 1034 HW status: 0 ci: 5 size: 1024

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en/health.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
index 1b735b54b3ab..4bd46e109dbe 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -62,6 +62,14 @@ int mlx5e_health_cq_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg)
 	if (err)
 		return err;
 
+	err = devlink_fmsg_u32_pair_put(fmsg, "ci", mlx5_cqwq_get_ci(&cq->wq));
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "size", mlx5_cqwq_get_size(&cq->wq));
+	if (err)
+		return err;
+
 	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
 	if (err)
 		return err;
-- 
2.26.2

[net-next 10/12] net/mlx5e: Add EQ info to TX/RX reporter's diagnose

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Enhance TX/RX reporter's diagnose to include info about the
corresponding EQ.

$ devlink health diagnose pci/0000:00:0b.0 reporter rx
Common config:
    RQ:
      type: 2 stride size: 2048 size: 8
      CQ:
        stride size: 64 size: 1024
RQs:
    channel ix: 0 rqn: 1713 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7 ICOSQ HW state: 1
     CQ:
       cqn: 1032 HW status: 0 ci: 0 size: 1024
     EQ:
       eqn: 7 irqn: 42 vecidx: 1 ci: 93 size: 2048
     channel ix: 1 rqn: 1718 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7 ICOSQ HW state: 1
     CQ:
       cqn: 1036 HW status: 0 ci: 0 size: 1024
     EQ:
       eqn: 8 irqn: 43 vecidx: 2 ci: 2 size: 2048

$ devlink health diagnose pci/0000:00:0b.0 reporter tx
Common Config:
    SQ:
      stride size: 64 size: 1024
      CQ:
        stride size: 64 size: 1024
SQs:
   channel ix: 0 tc: 0 txq ix: 0 sqn: 1712 HW state: 1 stopped: false cc: 91 pc: 91
   CQ:
     cqn: 1030 HW status: 0 ci: 91 size: 1024
   EQ:
     eqn: 7 irqn: 42 vecidx: 1 ci: 93 size: 2048
   channel ix: 1 tc: 0 txq ix: 1 sqn: 1717 HW state: 1 stopped: false cc: 0 pc: 0
   CQ:
     cqn: 1034 HW status: 0 ci: 0 size: 1024
   EQ:
     eqn: 8 irqn: 43 vecidx: 2 ci: 2 size: 2048

Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en/health.c   | 31 +++++++++++++++++++
 .../ethernet/mellanox/mlx5/core/en/health.h   |  1 +
 .../mellanox/mlx5/core/en/reporter_rx.c       |  4 +++
 .../mellanox/mlx5/core/en/reporter_tx.c       |  4 +++
 4 files changed, 40 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
index 4bd46e109dbe..3dc200bcfabd 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -105,6 +105,37 @@ int mlx5e_health_cq_common_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *f
 	return 0;
 }
 
+int mlx5e_health_eq_diag_fmsg(struct mlx5_eq_comp *eq, struct devlink_fmsg *fmsg)
+{
+	int err;
+
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "EQ");
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u8_pair_put(fmsg, "eqn", eq->core.eqn);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "irqn", eq->core.irqn);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "vecidx", eq->core.vecidx);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "ci", eq->core.cons_index);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "size", eq->core.nent);
+	if (err)
+		return err;
+
+	return mlx5e_health_fmsg_named_obj_nest_end(fmsg);
+}
+
 void mlx5e_health_create_reporters(struct mlx5e_priv *priv)
 {
 	mlx5e_reporter_tx_create(priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 6e48518d3d5b..b9aadddfd000 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -23,6 +23,7 @@ int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq);
 
 int mlx5e_health_cq_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg);
 int mlx5e_health_cq_common_diag_fmsg(struct mlx5e_cq *cq, struct devlink_fmsg *fmsg);
+int mlx5e_health_eq_diag_fmsg(struct mlx5_eq_comp *eq, struct devlink_fmsg *fmsg);
 int mlx5e_health_fmsg_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name);
 int mlx5e_health_fmsg_named_obj_nest_end(struct devlink_fmsg *fmsg);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index 4e1a01d871b7..5f7fba74cfd4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -242,6 +242,10 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 	if (err)
 		return err;
 
+	err = mlx5e_health_eq_diag_fmsg(rq->cq.mcq.eq, fmsg);
+	if (err)
+		return err;
+
 	err = devlink_fmsg_obj_nest_end(fmsg);
 	if (err)
 		return err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
index 9f712ff2faf9..465c7cc8d909 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -170,6 +170,10 @@ mlx5e_tx_reporter_build_diagnose_output(struct devlink_fmsg *fmsg,
 	if (err)
 		return err;
 
+	err = mlx5e_health_eq_diag_fmsg(sq->cq.mcq.eq, fmsg);
+	if (err)
+		return err;
+
 	err = devlink_fmsg_obj_nest_end(fmsg);
 	if (err)
 		return err;
-- 
2.26.2

[net-next 11/12] net/mlx5e: Enhance ICOSQ data on RX reporter's diagnose

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

When the RQ is in striding RQ mode, it uses the ICOSQ as a helper queue.
In this mode, RX reporter dumps more info about the ICOSQ and its
related CQ.

$ devlink health diagnose pci/0000:00:0b.0 reporter rx
Common config:
    RQ:
      type: 2 stride size: 2048 size: 8
      CQ:
        stride size: 64 size: 1024
RQs:
    channel ix: 0 rqn: 2413 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7
    CQ:
      cqn: 1032 HW status: 0 ci: 0 size: 1024
    EQ:
      eqn: 7 irqn: 42 vecidx: 1 ci: 93 size: 2048
    ICOSQ:
      sqn: 2411 HW state: 1 cc: 74 pc: 74 WQE size: 128
      CQ:
        cqn: 1029 cc: 8 size: 128
    channel ix: 1 rqn: 2418 HW state: 1 SW state: 5 WQE counter: 7 posted WQEs: 7 cc: 7
    CQ:
      cqn: 1036 HW status: 0 ci: 0 size: 1024
    EQ:
      eqn: 8 irqn: 43 vecidx: 2 ci: 2 size: 2048
    ICOSQ:
      sqn: 2416 HW state: 1 cc: 74 pc: 74 WQE size: 128
      CQ:
        cqn: 1033 cc: 8 size: 128

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../mellanox/mlx5/core/en/reporter_rx.c       | 59 ++++++++++++++++++-
 1 file changed, 56 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
index 5f7fba74cfd4..32ed1067e6dc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -177,6 +177,59 @@ static int mlx5e_rx_reporter_recover(struct devlink_health_reporter *reporter,
 			 mlx5e_health_recover_channels(priv);
 }
 
+static int mlx5e_reporter_icosq_diagnose(struct mlx5e_icosq *icosq, u8 hw_state,
+					 struct devlink_fmsg *fmsg)
+{
+	int err;
+
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "ICOSQ");
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "sqn", icosq->sqn);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u8_pair_put(fmsg, "HW state", hw_state);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "cc", icosq->cc);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "pc", icosq->pc);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "WQE size",
+					mlx5_wq_cyc_get_size(&icosq->wq));
+	if (err)
+		return err;
+
+	err = mlx5e_health_fmsg_named_obj_nest_start(fmsg, "CQ");
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "cqn", icosq->cq.mcq.cqn);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "cc", icosq->cq.wq.cc);
+	if (err)
+		return err;
+
+	err = devlink_fmsg_u32_pair_put(fmsg, "size", mlx5_cqwq_get_size(&icosq->cq.wq));
+	if (err)
+		return err;
+
+	err = mlx5e_health_fmsg_named_obj_nest_end(fmsg);
+	if (err)
+		return err;
+
+	return mlx5e_health_fmsg_named_obj_nest_end(fmsg);
+}
+
 static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 						   struct devlink_fmsg *fmsg)
 {
@@ -234,15 +287,15 @@ static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
 	if (err)
 		return err;
 
-	err = devlink_fmsg_u8_pair_put(fmsg, "ICOSQ HW state", icosq_hw_state);
+	err = mlx5e_health_cq_diag_fmsg(&rq->cq, fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_health_cq_diag_fmsg(&rq->cq, fmsg);
+	err = mlx5e_health_eq_diag_fmsg(rq->cq.mcq.eq, fmsg);
 	if (err)
 		return err;
 
-	err = mlx5e_health_eq_diag_fmsg(rq->cq.mcq.eq, fmsg);
+	err = mlx5e_reporter_icosq_diagnose(icosq, icosq_hw_state, fmsg);
 	if (err)
 		return err;
 
-- 
2.26.2

[net-next 12/12] net/mlx5e: Enhance TX timeout recovery

[Saeed Mahameed]

From: Aya Levin <ayal@mellanox.com>

Upon a TX timeout handle, if the TX reporter was not able to recover
from the error, reopen the channels. If tried to reopen channels, do not
loop over TX queues for timeout.

With that, the reporters state and separation will better
expose the driver's state.

Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../mellanox/mlx5/core/en/reporter_tx.c       | 36 ++++++++++++++++---
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 14 ++------
 2 files changed, 33 insertions(+), 17 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
index 465c7cc8d909..826584380216 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -83,17 +83,40 @@ static int mlx5e_tx_reporter_err_cqe_recover(void *ctx)
 	return err;
 }
 
+struct mlx5e_tx_timeout_ctx {
+	struct mlx5e_txqsq *sq;
+	signed int status;
+};
+
 static int mlx5e_tx_reporter_timeout_recover(void *ctx)
 {
+	struct mlx5e_tx_timeout_ctx *to_ctx;
+	struct mlx5e_priv *priv;
 	struct mlx5_eq_comp *eq;
 	struct mlx5e_txqsq *sq;
 	int err;
 
-	sq = ctx;
+	to_ctx = ctx;
+	sq = to_ctx->sq;
 	eq = sq->cq.mcq.eq;
+	priv = sq->channel->priv;
 	err = mlx5e_health_channel_eq_recover(eq, sq->channel);
-	if (err)
-		clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state);
+	if (!err) {
+		to_ctx->status = 0; /* this sq recovered */
+		return err;
+	}
+
+	err = mlx5e_safe_reopen_channels(priv);
+	if (!err) {
+		to_ctx->status = 1; /* all channels recovered */
+		return err;
+	}
+
+	to_ctx->status = err;
+	clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state);
+	netdev_err(priv->netdev,
+		   "mlx5e_safe_reopen_channels failed recovering from a tx_timeout, err(%d).\n",
+		   err);
 
 	return err;
 }
@@ -389,9 +412,11 @@ int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq)
 {
 	struct mlx5e_priv *priv = sq->channel->priv;
 	char err_str[MLX5E_REPORTER_PER_Q_MAX_LEN];
+	struct mlx5e_tx_timeout_ctx to_ctx = {};
 	struct mlx5e_err_ctx err_ctx = {};
 
-	err_ctx.ctx = sq;
+	to_ctx.sq = sq;
+	err_ctx.ctx = &to_ctx;
 	err_ctx.recover = mlx5e_tx_reporter_timeout_recover;
 	err_ctx.dump = mlx5e_tx_reporter_dump_sq;
 	snprintf(err_str, sizeof(err_str),
@@ -399,7 +424,8 @@ int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq)
 		 sq->channel->ix, sq->sqn, sq->cq.mcq.cqn, sq->cc, sq->pc,
 		 jiffies_to_usecs(jiffies - sq->txq->trans_start));
 
-	return mlx5e_health_report(priv, priv->tx_reporter, err_str, &err_ctx);
+	mlx5e_health_report(priv, priv->tx_reporter, err_str, &err_ctx);
+	return to_ctx.status;
 }
 
 static const struct devlink_health_reporter_ops mlx5_tx_reporter_ops = {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 046cfb0ea180..b04c8572adea 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -4367,8 +4367,6 @@ static void mlx5e_tx_timeout_work(struct work_struct *work)
 {
 	struct mlx5e_priv *priv = container_of(work, struct mlx5e_priv,
 					       tx_timeout_work);
-	bool report_failed = false;
-	int err;
 	int i;
 
 	rtnl_lock();
@@ -4386,18 +4384,10 @@ static void mlx5e_tx_timeout_work(struct work_struct *work)
 			continue;
 
 		if (mlx5e_reporter_tx_timeout(sq))
-			report_failed = true;
+		/* break if tried to reopened channels */
+			break;
 	}
 
-	if (!report_failed)
-		goto unlock;
-
-	err = mlx5e_safe_reopen_channels(priv);
-	if (err)
-		netdev_err(priv->netdev,
-			   "mlx5e_safe_reopen_channels failed recovering from a tx_timeout, err(%d).\n",
-			   err);
-
 unlock:
 	mutex_unlock(&priv->state_lock);
 	rtnl_unlock();
-- 
2.26.2

Re: [pull request][net-next 00/12] mlx5 updates 2020-07-02

[David Miller]

From: Saeed Mahameed <saeedm@mellanox.com>
Date: Thu,  2 Jul 2020 21:08:20 -0700

> This series adds misc updates.
> For more information please see tag log below.
> 
> Please pull and let me know if there is any problem.

Pulled, thank you.