[pull request][net-next 0/8] Mellanox, mlx5 updates 2018-10-18

[pull request][net-next 0/8] Mellanox, mlx5 updates 2018-10-18

[Saeed Mahameed]

Hi Dave,

This pull request provides some misc updates to mlx5 core and netdevice
driver.

For more details please see tag log below.

Please pull and let me know if there's any problem.

Thanks,
Saeed.

---

The following changes since commit 3a3295bfa6f484bba91de0a804c1d0bf7d31dbd3:

  Merge branch 'sctp-fix-sk_wmem_queued-and-use-it-to-check-for-writable-space' (2018-10-18 11:23:47 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5-updates-2018-10-18

for you to fetch changes up to 4cb4e98e5bba6dce058582d78cc993fef6f0755a:

  net/mlx5e: Added 'raw_errors_laneX' fields to ethtool statistics (2018-10-18 13:32:57 -0700)

----------------------------------------------------------------
mlx5-updates-2018-10-18

This series provides misc updates to mlx5 core and netdevice driver.

1) From Tariq Toukan: Refactor fragmented buffer struct fields and init flow.

2) From Vlad Buslov, Flow counters cache improvements and fixes follow up.
as a follow up work for the previous series of the mlx5 flow counters,
Vlad provides two fixes:
  2.1) Take fs_counters dellist before addlist
Fixes: 6e5e22839136 ("net/mlx5: Add new list to store deleted flow counters")
  2.2) Remove counter from idr after removing it from list
Fixes: 12d6066c3b29 ("net/mlx5: Add flow counters idr")

>From Shay Agroskin,
3) Add FEC set/get FW commands and FEC ethtool callbacks support
4) Add new ethtool statistics to cover errors on rx, such as FEC errors.

----------------------------------------------------------------
Shay Agroskin (5):
      net/mlx5: Add FEC fields to Port Phy Link Mode (PPLM) reg
      net/mlx5e: Add port FEC get/set functions
      net/mlx5e: Ethtool driver callback for query/set FEC policy
      net/mlx5: Added "per_lane_error_counters" cap bit to PCAM
      net/mlx5e: Added 'raw_errors_laneX' fields to ethtool statistics

Tariq Toukan (1):
      net/mlx5: Refactor fragmented buffer struct fields and init flow

Vlad Buslov (2):
      net/mlx5: Take fs_counters dellist before addlist
      net/mlx5: Remove counter from idr after removing it from list

 drivers/infiniband/hw/mlx5/cq.c                    |  31 ++-
 drivers/infiniband/hw/mlx5/mlx5_ib.h               |   1 +
 drivers/net/ethernet/mellanox/mlx5/core/en/port.c  | 208 +++++++++++++++++++++
 drivers/net/ethernet/mellanox/mlx5/core/en/port.h  |  12 ++
 .../net/ethernet/mellanox/mlx5/core/en_ethtool.c   | 128 ++++++++++++-
 drivers/net/ethernet/mellanox/mlx5/core/en_stats.c |  46 ++++-
 .../net/ethernet/mellanox/mlx5/core/fs_counters.c  |  31 ++-
 drivers/net/ethernet/mellanox/mlx5/core/wq.c       | 120 +++++-------
 include/linux/mlx5/driver.h                        |  23 +--
 include/linux/mlx5/mlx5_ifc.h                      |  42 +++--
 10 files changed, 508 insertions(+), 134 deletions(-)

[net-next 1/8] net/mlx5: Refactor fragmented buffer struct fields and init flow

[Saeed Mahameed]

From: Tariq Toukan <tariqt@mellanox.com>

Take struct mlx5_frag_buf out of mlx5_frag_buf_ctrl, as it is not
needed to manage and control the datapath of the fragmented buffers API.

struct mlx5_frag_buf contains control info to manage the allocation
and de-allocation of the fragmented buffer.
Its fields are not relevant for datapath, so here I take them out of the
struct mlx5_frag_buf_ctrl, except for the fragments array itself.

In addition, modified mlx5_fill_fbc to initialise the frags pointers
as well. This implies that the buffer must be allocated before the
function is called.

A set of type-specific *_get_byte_size() functions are replaced by
a generic one.

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/infiniband/hw/mlx5/cq.c              |  31 ++---
 drivers/infiniband/hw/mlx5/mlx5_ib.h         |   1 +
 drivers/net/ethernet/mellanox/mlx5/core/wq.c | 120 ++++++++-----------
 include/linux/mlx5/driver.h                  |  22 ++--
 4 files changed, 69 insertions(+), 105 deletions(-)

diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 088205d7f1a1..cca1820802b8 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -393,7 +393,7 @@ static void handle_atomics(struct mlx5_ib_qp *qp, struct mlx5_cqe64 *cqe64,
 
 static void free_cq_buf(struct mlx5_ib_dev *dev, struct mlx5_ib_cq_buf *buf)
 {
-	mlx5_frag_buf_free(dev->mdev, &buf->fbc.frag_buf);
+	mlx5_frag_buf_free(dev->mdev, &buf->frag_buf);
 }
 
 static void get_sig_err_item(struct mlx5_sig_err_cqe *cqe,
@@ -728,16 +728,11 @@ static int alloc_cq_frag_buf(struct mlx5_ib_dev *dev,
 			     int nent,
 			     int cqe_size)
 {
-	struct mlx5_frag_buf_ctrl *c = &buf->fbc;
-	struct mlx5_frag_buf *frag_buf = &c->frag_buf;
-	u32 cqc_buff[MLX5_ST_SZ_DW(cqc)] = {0};
+	struct mlx5_frag_buf *frag_buf = &buf->frag_buf;
+	u8 log_wq_stride = 6 + (cqe_size == 128 ? 1 : 0);
+	u8 log_wq_sz     = ilog2(cqe_size);
 	int err;
 
-	MLX5_SET(cqc, cqc_buff, log_cq_size, ilog2(cqe_size));
-	MLX5_SET(cqc, cqc_buff, cqe_sz, (cqe_size == 128) ? 1 : 0);
-
-	mlx5_core_init_cq_frag_buf(&buf->fbc, cqc_buff);
-
 	err = mlx5_frag_buf_alloc_node(dev->mdev,
 				       nent * cqe_size,
 				       frag_buf,
@@ -745,6 +740,8 @@ static int alloc_cq_frag_buf(struct mlx5_ib_dev *dev,
 	if (err)
 		return err;
 
+	mlx5_init_fbc(frag_buf->frags, log_wq_stride, log_wq_sz, &buf->fbc);
+
 	buf->cqe_size = cqe_size;
 	buf->nent = nent;
 
@@ -934,7 +931,7 @@ static int create_cq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq,
 
 	*inlen = MLX5_ST_SZ_BYTES(create_cq_in) +
 		 MLX5_FLD_SZ_BYTES(create_cq_in, pas[0]) *
-		 cq->buf.fbc.frag_buf.npages;
+		 cq->buf.frag_buf.npages;
 	*cqb = kvzalloc(*inlen, GFP_KERNEL);
 	if (!*cqb) {
 		err = -ENOMEM;
@@ -942,11 +939,11 @@ static int create_cq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq,
 	}
 
 	pas = (__be64 *)MLX5_ADDR_OF(create_cq_in, *cqb, pas);
-	mlx5_fill_page_frag_array(&cq->buf.fbc.frag_buf, pas);
+	mlx5_fill_page_frag_array(&cq->buf.frag_buf, pas);
 
 	cqc = MLX5_ADDR_OF(create_cq_in, *cqb, cq_context);
 	MLX5_SET(cqc, cqc, log_page_size,
-		 cq->buf.fbc.frag_buf.page_shift -
+		 cq->buf.frag_buf.page_shift -
 		 MLX5_ADAPTER_PAGE_SHIFT);
 
 	*index = dev->mdev->priv.uar->index;
@@ -1365,11 +1362,10 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 		cqe_size = 64;
 		err = resize_kernel(dev, cq, entries, cqe_size);
 		if (!err) {
-			struct mlx5_frag_buf_ctrl *c;
+			struct mlx5_frag_buf *frag_buf = &cq->resize_buf->frag_buf;
 
-			c = &cq->resize_buf->fbc;
-			npas = c->frag_buf.npages;
-			page_shift = c->frag_buf.page_shift;
+			npas = frag_buf->npages;
+			page_shift = frag_buf->page_shift;
 		}
 	}
 
@@ -1390,8 +1386,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 		mlx5_ib_populate_pas(dev, cq->resize_umem, page_shift,
 				     pas, 0);
 	else
-		mlx5_fill_page_frag_array(&cq->resize_buf->fbc.frag_buf,
-					  pas);
+		mlx5_fill_page_frag_array(&cq->resize_buf->frag_buf, pas);
 
 	MLX5_SET(modify_cq_in, in,
 		 modify_field_select_resize_field_select.resize_field_select.resize_field_select,
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 320d4dfe8c2f..289c18db2611 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -435,6 +435,7 @@ struct mlx5_ib_qp {
 
 struct mlx5_ib_cq_buf {
 	struct mlx5_frag_buf_ctrl fbc;
+	struct mlx5_frag_buf    frag_buf;
 	struct ib_umem		*umem;
 	int			cqe_size;
 	int			nent;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/wq.c b/drivers/net/ethernet/mellanox/mlx5/core/wq.c
index 68e7f8df2a6d..9007e91ad53f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/wq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/wq.c
@@ -54,54 +54,37 @@ u32 mlx5_wq_ll_get_size(struct mlx5_wq_ll *wq)
 	return (u32)wq->fbc.sz_m1 + 1;
 }
 
-static u32 mlx5_wq_cyc_get_byte_size(struct mlx5_wq_cyc *wq)
+static u32 wq_get_byte_sz(u8 log_sz, u8 log_stride)
 {
-	return mlx5_wq_cyc_get_size(wq) << wq->fbc.log_stride;
-}
-
-static u32 mlx5_wq_qp_get_byte_size(struct mlx5_wq_qp *wq)
-{
-	return mlx5_wq_cyc_get_byte_size(&wq->rq) +
-	       mlx5_wq_cyc_get_byte_size(&wq->sq);
-}
-
-static u32 mlx5_cqwq_get_byte_size(struct mlx5_cqwq *wq)
-{
-	return mlx5_cqwq_get_size(wq) << wq->fbc.log_stride;
-}
-
-static u32 mlx5_wq_ll_get_byte_size(struct mlx5_wq_ll *wq)
-{
-	return mlx5_wq_ll_get_size(wq) << wq->fbc.log_stride;
+	return ((u32)1 << log_sz) << log_stride;
 }
 
 int mlx5_wq_cyc_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 		       void *wqc, struct mlx5_wq_cyc *wq,
 		       struct mlx5_wq_ctrl *wq_ctrl)
 {
+	u8 log_wq_stride = MLX5_GET(wq, wqc, log_wq_stride);
+	u8 log_wq_sz     = MLX5_GET(wq, wqc, log_wq_sz);
 	struct mlx5_frag_buf_ctrl *fbc = &wq->fbc;
 	int err;
 
-	mlx5_fill_fbc(MLX5_GET(wq, wqc, log_wq_stride),
-		      MLX5_GET(wq, wqc, log_wq_sz),
-		      fbc);
-	wq->sz    = wq->fbc.sz_m1 + 1;
-
 	err = mlx5_db_alloc_node(mdev, &wq_ctrl->db, param->db_numa_node);
 	if (err) {
 		mlx5_core_warn(mdev, "mlx5_db_alloc_node() failed, %d\n", err);
 		return err;
 	}
 
-	err = mlx5_frag_buf_alloc_node(mdev, mlx5_wq_cyc_get_byte_size(wq),
+	wq->db  = wq_ctrl->db.db;
+
+	err = mlx5_frag_buf_alloc_node(mdev, wq_get_byte_sz(log_wq_sz, log_wq_stride),
 				       &wq_ctrl->buf, param->buf_numa_node);
 	if (err) {
 		mlx5_core_warn(mdev, "mlx5_frag_buf_alloc_node() failed, %d\n", err);
 		goto err_db_free;
 	}
 
-	fbc->frag_buf = wq_ctrl->buf;
-	wq->db  = wq_ctrl->db.db;
+	mlx5_init_fbc(wq_ctrl->buf.frags, log_wq_stride, log_wq_sz, fbc);
+	wq->sz = mlx5_wq_cyc_get_size(wq);
 
 	wq_ctrl->mdev = mdev;
 
@@ -113,46 +96,19 @@ int mlx5_wq_cyc_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 	return err;
 }
 
-static void mlx5_qp_set_frag_buf(struct mlx5_frag_buf *buf,
-				 struct mlx5_wq_qp *qp)
-{
-	struct mlx5_frag_buf_ctrl *sq_fbc;
-	struct mlx5_frag_buf *rqb, *sqb;
-
-	rqb  = &qp->rq.fbc.frag_buf;
-	*rqb = *buf;
-	rqb->size   = mlx5_wq_cyc_get_byte_size(&qp->rq);
-	rqb->npages = DIV_ROUND_UP(rqb->size, PAGE_SIZE);
-
-	sq_fbc = &qp->sq.fbc;
-	sqb    = &sq_fbc->frag_buf;
-	*sqb   = *buf;
-	sqb->size   = mlx5_wq_cyc_get_byte_size(&qp->sq);
-	sqb->npages = DIV_ROUND_UP(sqb->size, PAGE_SIZE);
-	sqb->frags += rqb->npages; /* first part is for the rq */
-	if (sq_fbc->strides_offset)
-		sqb->frags--;
-}
-
 int mlx5_wq_qp_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 		      void *qpc, struct mlx5_wq_qp *wq,
 		      struct mlx5_wq_ctrl *wq_ctrl)
 {
-	u16 sq_strides_offset;
-	u32 rq_pg_remainder;
-	int err;
+	u8 log_rq_stride = MLX5_GET(qpc, qpc, log_rq_stride) + 4;
+	u8 log_rq_sz     = MLX5_GET(qpc, qpc, log_rq_size);
+	u8 log_sq_stride = ilog2(MLX5_SEND_WQE_BB);
+	u8 log_sq_sz     = MLX5_GET(qpc, qpc, log_sq_size);
 
-	mlx5_fill_fbc(MLX5_GET(qpc, qpc, log_rq_stride) + 4,
-		      MLX5_GET(qpc, qpc, log_rq_size),
-		      &wq->rq.fbc);
+	u32 rq_byte_size;
+	int err;
 
-	rq_pg_remainder   = mlx5_wq_cyc_get_byte_size(&wq->rq) % PAGE_SIZE;
-	sq_strides_offset = rq_pg_remainder / MLX5_SEND_WQE_BB;
 
-	mlx5_fill_fbc_offset(ilog2(MLX5_SEND_WQE_BB),
-			     MLX5_GET(qpc, qpc, log_sq_size),
-			     sq_strides_offset,
-			     &wq->sq.fbc);
 
 	err = mlx5_db_alloc_node(mdev, &wq_ctrl->db, param->db_numa_node);
 	if (err) {
@@ -160,14 +116,32 @@ int mlx5_wq_qp_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 		return err;
 	}
 
-	err = mlx5_frag_buf_alloc_node(mdev, mlx5_wq_qp_get_byte_size(wq),
+	err = mlx5_frag_buf_alloc_node(mdev,
+				       wq_get_byte_sz(log_rq_sz, log_rq_stride) +
+				       wq_get_byte_sz(log_sq_sz, log_sq_stride),
 				       &wq_ctrl->buf, param->buf_numa_node);
 	if (err) {
 		mlx5_core_warn(mdev, "mlx5_frag_buf_alloc_node() failed, %d\n", err);
 		goto err_db_free;
 	}
 
-	mlx5_qp_set_frag_buf(&wq_ctrl->buf, wq);
+	mlx5_init_fbc(wq_ctrl->buf.frags, log_rq_stride, log_rq_sz, &wq->rq.fbc);
+
+	rq_byte_size = wq_get_byte_sz(log_rq_sz, log_rq_stride);
+
+	if (rq_byte_size < PAGE_SIZE) {
+		/* SQ starts within the same page of the RQ */
+		u16 sq_strides_offset = rq_byte_size / MLX5_SEND_WQE_BB;
+
+		mlx5_init_fbc_offset(wq_ctrl->buf.frags,
+				     log_sq_stride, log_sq_sz, sq_strides_offset,
+				     &wq->sq.fbc);
+	} else {
+		u16 rq_npages = rq_byte_size >> PAGE_SHIFT;
+
+		mlx5_init_fbc(wq_ctrl->buf.frags + rq_npages,
+			      log_sq_stride, log_sq_sz, &wq->sq.fbc);
+	}
 
 	wq->rq.db  = &wq_ctrl->db.db[MLX5_RCV_DBR];
 	wq->sq.db  = &wq_ctrl->db.db[MLX5_SND_DBR];
@@ -186,17 +160,19 @@ int mlx5_cqwq_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 		     void *cqc, struct mlx5_cqwq *wq,
 		     struct mlx5_wq_ctrl *wq_ctrl)
 {
+	u8 log_wq_stride = MLX5_GET(cqc, cqc, cqe_sz) + 6;
+	u8 log_wq_sz     = MLX5_GET(cqc, cqc, log_cq_size);
 	int err;
 
-	mlx5_core_init_cq_frag_buf(&wq->fbc, cqc);
-
 	err = mlx5_db_alloc_node(mdev, &wq_ctrl->db, param->db_numa_node);
 	if (err) {
 		mlx5_core_warn(mdev, "mlx5_db_alloc_node() failed, %d\n", err);
 		return err;
 	}
 
-	err = mlx5_frag_buf_alloc_node(mdev, mlx5_cqwq_get_byte_size(wq),
+	wq->db  = wq_ctrl->db.db;
+
+	err = mlx5_frag_buf_alloc_node(mdev, wq_get_byte_sz(log_wq_sz, log_wq_stride),
 				       &wq_ctrl->buf,
 				       param->buf_numa_node);
 	if (err) {
@@ -205,8 +181,7 @@ int mlx5_cqwq_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 		goto err_db_free;
 	}
 
-	wq->fbc.frag_buf = wq_ctrl->buf;
-	wq->db  = wq_ctrl->db.db;
+	mlx5_init_fbc(wq_ctrl->buf.frags, log_wq_stride, log_wq_sz, &wq->fbc);
 
 	wq_ctrl->mdev = mdev;
 
@@ -222,30 +197,29 @@ int mlx5_wq_ll_create(struct mlx5_core_dev *mdev, struct mlx5_wq_param *param,
 		      void *wqc, struct mlx5_wq_ll *wq,
 		      struct mlx5_wq_ctrl *wq_ctrl)
 {
+	u8 log_wq_stride = MLX5_GET(wq, wqc, log_wq_stride);
+	u8 log_wq_sz     = MLX5_GET(wq, wqc, log_wq_sz);
 	struct mlx5_frag_buf_ctrl *fbc = &wq->fbc;
 	struct mlx5_wqe_srq_next_seg *next_seg;
 	int err;
 	int i;
 
-	mlx5_fill_fbc(MLX5_GET(wq, wqc, log_wq_stride),
-		      MLX5_GET(wq, wqc, log_wq_sz),
-		      fbc);
-
 	err = mlx5_db_alloc_node(mdev, &wq_ctrl->db, param->db_numa_node);
 	if (err) {
 		mlx5_core_warn(mdev, "mlx5_db_alloc_node() failed, %d\n", err);
 		return err;
 	}
 
-	err = mlx5_frag_buf_alloc_node(mdev, mlx5_wq_ll_get_byte_size(wq),
+	wq->db  = wq_ctrl->db.db;
+
+	err = mlx5_frag_buf_alloc_node(mdev, wq_get_byte_sz(log_wq_sz, log_wq_stride),
 				       &wq_ctrl->buf, param->buf_numa_node);
 	if (err) {
 		mlx5_core_warn(mdev, "mlx5_frag_buf_alloc_node() failed, %d\n", err);
 		goto err_db_free;
 	}
 
-	wq->fbc.frag_buf = wq_ctrl->buf;
-	wq->db  = wq_ctrl->db.db;
+	mlx5_init_fbc(wq_ctrl->buf.frags, log_wq_stride, log_wq_sz, fbc);
 
 	for (i = 0; i < fbc->sz_m1; i++) {
 		next_seg = mlx5_wq_ll_get_wqe(wq, i);
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 94ffd02af7cd..e10f61a1f77d 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -357,7 +357,7 @@ struct mlx5_frag_buf {
 };
 
 struct mlx5_frag_buf_ctrl {
-	struct mlx5_frag_buf	frag_buf;
+	struct mlx5_buf_list   *frags;
 	u32			sz_m1;
 	u16			frag_sz_m1;
 	u16			strides_offset;
@@ -994,10 +994,12 @@ static inline u32 mlx5_base_mkey(const u32 key)
 	return key & 0xffffff00u;
 }
 
-static inline void mlx5_fill_fbc_offset(u8 log_stride, u8 log_sz,
+static inline void mlx5_init_fbc_offset(struct mlx5_buf_list *frags,
+					u8 log_stride, u8 log_sz,
 					u16 strides_offset,
 					struct mlx5_frag_buf_ctrl *fbc)
 {
+	fbc->frags      = frags;
 	fbc->log_stride = log_stride;
 	fbc->log_sz     = log_sz;
 	fbc->sz_m1	= (1 << fbc->log_sz) - 1;
@@ -1006,18 +1008,11 @@ static inline void mlx5_fill_fbc_offset(u8 log_stride, u8 log_sz,
 	fbc->strides_offset = strides_offset;
 }
 
-static inline void mlx5_fill_fbc(u8 log_stride, u8 log_sz,
+static inline void mlx5_init_fbc(struct mlx5_buf_list *frags,
+				 u8 log_stride, u8 log_sz,
 				 struct mlx5_frag_buf_ctrl *fbc)
 {
-	mlx5_fill_fbc_offset(log_stride, log_sz, 0, fbc);
-}
-
-static inline void mlx5_core_init_cq_frag_buf(struct mlx5_frag_buf_ctrl *fbc,
-					      void *cqc)
-{
-	mlx5_fill_fbc(6 + MLX5_GET(cqc, cqc, cqe_sz),
-		      MLX5_GET(cqc, cqc, log_cq_size),
-		      fbc);
+	mlx5_init_fbc_offset(frags, log_stride, log_sz, 0, fbc);
 }
 
 static inline void *mlx5_frag_buf_get_wqe(struct mlx5_frag_buf_ctrl *fbc,
@@ -1028,8 +1023,7 @@ static inline void *mlx5_frag_buf_get_wqe(struct mlx5_frag_buf_ctrl *fbc,
 	ix  += fbc->strides_offset;
 	frag = ix >> fbc->log_frag_strides;
 
-	return fbc->frag_buf.frags[frag].buf +
-		((fbc->frag_sz_m1 & ix) << fbc->log_stride);
+	return fbc->frags[frag].buf + ((fbc->frag_sz_m1 & ix) << fbc->log_stride);
 }
 
 int mlx5_cmd_init(struct mlx5_core_dev *dev);
-- 
2.17.2

[net-next 2/8] net/mlx5: Take fs_counters dellist before addlist

[Saeed Mahameed]

From: Vlad Buslov <vladbu@mellanox.com>

In fs_counters elements from both addlist and dellist are removed by
mlx5_fc_stats_work() without any locking. This introduces race condition
when batch of new rules is created and then immediately deleted (for
example, when error occurred during flow creation). In such case some of
the rules might be in dellist, but not in addlist when mlx5_fc_stats_work()
is executed concurrently with tc, which will result rule deletion and
use-after-free on next iteration because deleted rules are still in
addlist.

Always take dellist first to guarantee that rules can only be deleted after
they were removed from addlist.

Fixes: 6e5e22839136 ("net/mlx5: Add new list to store deleted flow counters")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reported-by: Chris Mi <chrism@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/fs_counters.c   | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
index 1329bc5b7969..77bd4cd3a1db 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
@@ -179,19 +179,22 @@ static void mlx5_fc_stats_work(struct work_struct *work)
 	struct mlx5_core_dev *dev = container_of(work, struct mlx5_core_dev,
 						 priv.fc_stats.work.work);
 	struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats;
-	struct llist_node *tmplist = llist_del_all(&fc_stats->addlist);
+	/* Take dellist first to ensure that counters cannot be deleted before
+	 * they are inserted.
+	 */
+	struct llist_node *dellist = llist_del_all(&fc_stats->dellist);
+	struct llist_node *addlist = llist_del_all(&fc_stats->addlist);
 	struct mlx5_fc *counter = NULL, *last = NULL, *tmp;
 	unsigned long now = jiffies;
 
-	if (tmplist || !list_empty(&fc_stats->counters))
+	if (addlist || !list_empty(&fc_stats->counters))
 		queue_delayed_work(fc_stats->wq, &fc_stats->work,
 				   fc_stats->sampling_interval);
 
-	llist_for_each_entry(counter, tmplist, addlist)
+	llist_for_each_entry(counter, addlist, addlist)
 		mlx5_fc_stats_insert(dev, counter);
 
-	tmplist = llist_del_all(&fc_stats->dellist);
-	llist_for_each_entry_safe(counter, tmp, tmplist, dellist) {
+	llist_for_each_entry_safe(counter, tmp, dellist, dellist) {
 		list_del(&counter->list);
 
 		mlx5_free_fc(dev, counter);
-- 
2.17.2

[net-next 3/8] net/mlx5: Remove counter from idr after removing it from list

[Saeed Mahameed]

From: Vlad Buslov <vladbu@mellanox.com>

Fs_counters list can temporary become unsorted when new counters are
created/deleted concurrently. Idr is used to quickly lookup position to
insert new counter in logarithmic time. However, if new flows are
concurrently inserted during time window when flows with adjacent ids are
already removed from idr but are still present in counters list,
mlx5_fc_stats_work() observes counters list in inconsistent state, which
results following warning:

[ 1839.561955] mlx5_core 0000:81:00.0: mlx5_cmd_fc_bulk_get:587:(pid 729): Flow counter id (0x102d5) out of range (0x1c0a8..0x1c10b). Counter ignored.

Move idr_remove() call to be executed synchronously with counter deletion
from list. Extract this code to mlx5_fc_stats_remove() helper function that
is called by workqueue job handler mlx5_fc_stats_work().

Fixes: 12d6066c3b29 ("net/mlx5: Add flow counters idr")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/fs_counters.c  | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
index 77bd4cd3a1db..32accd6b041b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
@@ -99,6 +99,18 @@ static void mlx5_fc_stats_insert(struct mlx5_core_dev *dev,
 	list_add_tail(&counter->list, next);
 }
 
+static void mlx5_fc_stats_remove(struct mlx5_core_dev *dev,
+				 struct mlx5_fc *counter)
+{
+	struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats;
+
+	list_del(&counter->list);
+
+	spin_lock(&fc_stats->counters_idr_lock);
+	WARN_ON(!idr_remove(&fc_stats->counters_idr, counter->id));
+	spin_unlock(&fc_stats->counters_idr_lock);
+}
+
 /* The function returns the last counter that was queried so the caller
  * function can continue calling it till all counters are queried.
  */
@@ -195,7 +207,7 @@ static void mlx5_fc_stats_work(struct work_struct *work)
 		mlx5_fc_stats_insert(dev, counter);
 
 	llist_for_each_entry_safe(counter, tmp, dellist, dellist) {
-		list_del(&counter->list);
+		mlx5_fc_stats_remove(dev, counter);
 
 		mlx5_free_fc(dev, counter);
 	}
@@ -275,10 +287,6 @@ void mlx5_fc_destroy(struct mlx5_core_dev *dev, struct mlx5_fc *counter)
 		return;
 
 	if (counter->aging) {
-		spin_lock(&fc_stats->counters_idr_lock);
-		WARN_ON(!idr_remove(&fc_stats->counters_idr, counter->id));
-		spin_unlock(&fc_stats->counters_idr_lock);
-
 		llist_add(&counter->dellist, &fc_stats->dellist);
 		mod_delayed_work(fc_stats->wq, &fc_stats->work, 0);
 		return;
-- 
2.17.2

[net-next 4/8] net/mlx5: Add FEC fields to Port Phy Link Mode (PPLM) reg

[Saeed Mahameed]

From: Shay Agroskin <shayag@mellanox.com>

Added FEC related fields to PPLM layout.
These fields are needed to set and query FEC policy
for different link speeds.

Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 include/linux/mlx5/driver.h   |  1 +
 include/linux/mlx5/mlx5_ifc.h | 39 +++++++++++++++++++++++++----------
 2 files changed, 29 insertions(+), 11 deletions(-)

diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index e10f61a1f77d..696ed3f7f894 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -133,6 +133,7 @@ enum {
 	MLX5_REG_PVLC		 = 0x500f,
 	MLX5_REG_PCMR		 = 0x5041,
 	MLX5_REG_PMLP		 = 0x5002,
+	MLX5_REG_PPLM		 = 0x5023,
 	MLX5_REG_PCAM		 = 0x507f,
 	MLX5_REG_NODE_DESC	 = 0x6001,
 	MLX5_REG_HOST_ENDIANNESS = 0x7004,
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 963611820006..47b09a742ae5 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -7828,20 +7828,34 @@ struct mlx5_ifc_pplr_reg_bits {
 
 struct mlx5_ifc_pplm_reg_bits {
 	u8         reserved_at_0[0x8];
-	u8         local_port[0x8];
-	u8         reserved_at_10[0x10];
+	u8	   local_port[0x8];
+	u8	   reserved_at_10[0x10];
 
-	u8         reserved_at_20[0x20];
+	u8	   reserved_at_20[0x20];
 
-	u8         port_profile_mode[0x8];
-	u8         static_port_profile[0x8];
-	u8         active_port_profile[0x8];
-	u8         reserved_at_58[0x8];
+	u8	   port_profile_mode[0x8];
+	u8	   static_port_profile[0x8];
+	u8	   active_port_profile[0x8];
+	u8	   reserved_at_58[0x8];
 
-	u8         retransmission_active[0x8];
-	u8         fec_mode_active[0x18];
+	u8	   retransmission_active[0x8];
+	u8	   fec_mode_active[0x18];
 
-	u8         reserved_at_80[0x20];
+	u8	   rs_fec_correction_bypass_cap[0x4];
+	u8	   reserved_at_84[0x8];
+	u8	   fec_override_cap_56g[0x4];
+	u8	   fec_override_cap_100g[0x4];
+	u8	   fec_override_cap_50g[0x4];
+	u8	   fec_override_cap_25g[0x4];
+	u8	   fec_override_cap_10g_40g[0x4];
+
+	u8	   rs_fec_correction_bypass_admin[0x4];
+	u8	   reserved_at_a4[0x8];
+	u8	   fec_override_admin_56g[0x4];
+	u8	   fec_override_admin_100g[0x4];
+	u8	   fec_override_admin_50g[0x4];
+	u8	   fec_override_admin_25g[0x4];
+	u8	   fec_override_admin_10g_40g[0x4];
 };
 
 struct mlx5_ifc_ppcnt_reg_bits {
@@ -8137,7 +8151,10 @@ struct mlx5_ifc_pcam_enhanced_features_bits {
 struct mlx5_ifc_pcam_regs_5000_to_507f_bits {
 	u8         port_access_reg_cap_mask_127_to_96[0x20];
 	u8         port_access_reg_cap_mask_95_to_64[0x20];
-	u8         port_access_reg_cap_mask_63_to_32[0x20];
+
+	u8         port_access_reg_cap_mask_63_to_36[0x1c];
+	u8         pplm[0x1];
+	u8         port_access_reg_cap_mask_34_to_32[0x3];
 
 	u8         port_access_reg_cap_mask_31_to_13[0x13];
 	u8         pbmc[0x1];
-- 
2.17.2

[net-next 5/8] net/mlx5e: Add port FEC get/set functions

[Saeed Mahameed]

From: Shay Agroskin <shayag@mellanox.com>

Added functions to query and set link FEC policy.
To get/set FEC capabilities in PPLM reg we need to query
current link speed.
'mlx5_get_fec_speed_field' queries current link speed and returns
correct field offset.

FEC Query's return value is divided into 'active FEC policy', which is
the FEC policy used by the link, and 'configured FEC policy', which
is the FEC policy requested by the user.
The two values may differ if:
1) FEC policy was configured to 'auto',
   in which case the active FEC policy would be the default FEC policy
   for current link speed.

2) FEC policy was changed, but no link reset is performed. In which case,
   the active FEC policy would become the configured one after a link
   reset.

FEC set function sets FEC policy for all link speeds and perform link
reset.
1) If current link speed doesn't support requested FEC policy,
   the function fails.
2) If a different link speed doesn't support requested FEC policy,
   FEC capbilities for this speed are turned off and a warning message
   is printed.

Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en/port.c | 208 ++++++++++++++++++
 .../net/ethernet/mellanox/mlx5/core/en/port.h |  12 +
 2 files changed, 220 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/port.c b/drivers/net/ethernet/mellanox/mlx5/core/en/port.c
index 24e3b564964f..023dc4bccd28 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/port.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/port.c
@@ -235,3 +235,211 @@ int mlx5e_port_set_priority2buffer(struct mlx5_core_dev *mdev, u8 *buffer)
 	kfree(out);
 	return err;
 }
+
+static u32 fec_supported_speeds[] = {
+	10000,
+	40000,
+	25000,
+	50000,
+	56000,
+	100000
+};
+
+#define MLX5E_FEC_SUPPORTED_SPEEDS ARRAY_SIZE(fec_supported_speeds)
+
+/* get/set FEC admin field for a given speed */
+static int mlx5e_fec_admin_field(u32 *pplm,
+				 u8 *fec_policy,
+				 bool write,
+				 u32 speed)
+{
+	switch (speed) {
+	case 10000:
+	case 40000:
+		if (!write)
+			*fec_policy = MLX5_GET(pplm_reg, pplm,
+					       fec_override_cap_10g_40g);
+		else
+			MLX5_SET(pplm_reg, pplm,
+				 fec_override_admin_10g_40g, *fec_policy);
+		break;
+	case 25000:
+		if (!write)
+			*fec_policy = MLX5_GET(pplm_reg, pplm,
+					       fec_override_admin_25g);
+		else
+			MLX5_SET(pplm_reg, pplm,
+				 fec_override_admin_25g, *fec_policy);
+		break;
+	case 50000:
+		if (!write)
+			*fec_policy = MLX5_GET(pplm_reg, pplm,
+					       fec_override_admin_50g);
+		else
+			MLX5_SET(pplm_reg, pplm,
+				 fec_override_admin_50g, *fec_policy);
+		break;
+	case 56000:
+		if (!write)
+			*fec_policy = MLX5_GET(pplm_reg, pplm,
+					       fec_override_admin_56g);
+		else
+			MLX5_SET(pplm_reg, pplm,
+				 fec_override_admin_56g, *fec_policy);
+		break;
+	case 100000:
+		if (!write)
+			*fec_policy = MLX5_GET(pplm_reg, pplm,
+					       fec_override_admin_100g);
+		else
+			MLX5_SET(pplm_reg, pplm,
+				 fec_override_admin_100g, *fec_policy);
+		break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+/* returns FEC capabilities for a given speed */
+static int mlx5e_get_fec_cap_field(u32 *pplm,
+				   u8 *fec_cap,
+				   u32 speed)
+{
+	switch (speed) {
+	case 10000:
+	case 40000:
+		*fec_cap = MLX5_GET(pplm_reg, pplm,
+				    fec_override_admin_10g_40g);
+		break;
+	case 25000:
+		*fec_cap = MLX5_GET(pplm_reg, pplm,
+				    fec_override_cap_25g);
+		break;
+	case 50000:
+		*fec_cap = MLX5_GET(pplm_reg, pplm,
+				    fec_override_cap_50g);
+		break;
+	case 56000:
+		*fec_cap = MLX5_GET(pplm_reg, pplm,
+				    fec_override_cap_56g);
+		break;
+	case 100000:
+		*fec_cap = MLX5_GET(pplm_reg, pplm,
+				    fec_override_cap_100g);
+		break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+int mlx5e_get_fec_caps(struct mlx5_core_dev *dev, u8 *fec_caps)
+{
+	u32 out[MLX5_ST_SZ_DW(pplm_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(pplm_reg)] = {};
+	int sz = MLX5_ST_SZ_BYTES(pplm_reg);
+	u32 current_fec_speed;
+	int err;
+
+	if (!MLX5_CAP_GEN(dev, pcam_reg))
+		return -EOPNOTSUPP;
+
+	if (!MLX5_CAP_PCAM_REG(dev, pplm))
+		return -EOPNOTSUPP;
+
+	MLX5_SET(pplm_reg, in, local_port, 1);
+	err =  mlx5_core_access_reg(dev, in, sz, out, sz, MLX5_REG_PPLM, 0, 0);
+	if (err)
+		return err;
+
+	err = mlx5e_port_linkspeed(dev, &current_fec_speed);
+	if (err)
+		return err;
+
+	return mlx5e_get_fec_cap_field(out, fec_caps, current_fec_speed);
+}
+
+int mlx5e_get_fec_mode(struct mlx5_core_dev *dev, u32 *fec_mode_active,
+		       u8 *fec_configured_mode)
+{
+	u32 out[MLX5_ST_SZ_DW(pplm_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(pplm_reg)] = {};
+	int sz = MLX5_ST_SZ_BYTES(pplm_reg);
+	u32 link_speed;
+	int err;
+
+	if (!MLX5_CAP_GEN(dev, pcam_reg))
+		return -EOPNOTSUPP;
+
+	if (!MLX5_CAP_PCAM_REG(dev, pplm))
+		return -EOPNOTSUPP;
+
+	MLX5_SET(pplm_reg, in, local_port, 1);
+	err =  mlx5_core_access_reg(dev, in, sz, out, sz, MLX5_REG_PPLM, 0, 0);
+	if (err)
+		return err;
+
+	*fec_mode_active = MLX5_GET(pplm_reg, out, fec_mode_active);
+
+	if (!fec_configured_mode)
+		return 0;
+
+	err = mlx5e_port_linkspeed(dev, &link_speed);
+	if (err)
+		return err;
+
+	return mlx5e_fec_admin_field(out, fec_configured_mode, 0, link_speed);
+}
+
+int mlx5e_set_fec_mode(struct mlx5_core_dev *dev, u8 fec_policy)
+{
+	bool fec_mode_not_supp_in_speed = false;
+	u8 no_fec_policy = BIT(MLX5E_FEC_NOFEC);
+	u32 out[MLX5_ST_SZ_DW(pplm_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(pplm_reg)] = {};
+	int sz = MLX5_ST_SZ_BYTES(pplm_reg);
+	u32 current_fec_speed;
+	u8 fec_caps = 0;
+	int err;
+	int i;
+
+	if (!MLX5_CAP_GEN(dev, pcam_reg))
+		return -EOPNOTSUPP;
+
+	if (!MLX5_CAP_PCAM_REG(dev, pplm))
+		return -EOPNOTSUPP;
+
+	MLX5_SET(pplm_reg, in, local_port, 1);
+	err = mlx5_core_access_reg(dev, in, sz, out, sz, MLX5_REG_PPLM, 0, 0);
+	if (err)
+		return err;
+
+	err = mlx5e_port_linkspeed(dev, &current_fec_speed);
+	if (err)
+		return err;
+
+	memset(in, 0, sz);
+	MLX5_SET(pplm_reg, in, local_port, 1);
+	for (i = 0; i < MLX5E_FEC_SUPPORTED_SPEEDS && !!fec_policy; i++) {
+		mlx5e_get_fec_cap_field(out, &fec_caps, fec_supported_speeds[i]);
+		/* policy supported for link speed */
+		if (!!(fec_caps & fec_policy)) {
+			mlx5e_fec_admin_field(in, &fec_policy, 1,
+					      fec_supported_speeds[i]);
+		} else {
+			if (fec_supported_speeds[i] == current_fec_speed)
+				return -EOPNOTSUPP;
+			mlx5e_fec_admin_field(in, &no_fec_policy, 1,
+					      fec_supported_speeds[i]);
+			fec_mode_not_supp_in_speed = true;
+		}
+	}
+
+	if (fec_mode_not_supp_in_speed)
+		mlx5_core_dbg(dev,
+			      "FEC policy 0x%x is not supported for some speeds",
+			      fec_policy);
+
+	return mlx5_core_access_reg(dev, in, sz, out, sz, MLX5_REG_PPLM, 0, 1);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/port.h b/drivers/net/ethernet/mellanox/mlx5/core/en/port.h
index f8cbd8194179..cd2160b8c9bf 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/port.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/port.h
@@ -45,4 +45,16 @@ int mlx5e_port_query_pbmc(struct mlx5_core_dev *mdev, void *out);
 int mlx5e_port_set_pbmc(struct mlx5_core_dev *mdev, void *in);
 int mlx5e_port_query_priority2buffer(struct mlx5_core_dev *mdev, u8 *buffer);
 int mlx5e_port_set_priority2buffer(struct mlx5_core_dev *mdev, u8 *buffer);
+
+int mlx5e_get_fec_caps(struct mlx5_core_dev *dev, u8 *fec_caps);
+int mlx5e_get_fec_mode(struct mlx5_core_dev *dev, u32 *fec_mode_active,
+		       u8 *fec_configured_mode);
+int mlx5e_set_fec_mode(struct mlx5_core_dev *dev, u8 fec_policy);
+
+enum {
+	MLX5E_FEC_NOFEC,
+	MLX5E_FEC_FIRECODE,
+	MLX5E_FEC_RS_528_514,
+};
+
 #endif
-- 
2.17.2

[net-next 6/8] net/mlx5e: Ethtool driver callback for query/set FEC policy

[Saeed Mahameed]

From: Shay Agroskin <shayag@mellanox.com>

Driver callback function for 'ethtool --show-fec',
'ethtool --set-fec' commands.

The query function returns active and configured FEC policy
for current link speed.

The set function sets FEC policy for all supported link
speeds.
1) If current link speed doesn't support requested FEC policy,
   the function fails.
2) If a different link speed doesn't support requested FEC
   policy, FEC capbilities for this speed are turned off.

Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en_ethtool.c  | 128 +++++++++++++++++-
 1 file changed, 126 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index c61f6302cde2..3e770abfd802 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -547,6 +547,70 @@ static void ptys2ethtool_adver_link(unsigned long *advertising_modes,
 			  __ETHTOOL_LINK_MODE_MASK_NBITS);
 }
 
+static const u32 pplm_fec_2_ethtool[] = {
+	[MLX5E_FEC_NOFEC] = ETHTOOL_FEC_OFF,
+	[MLX5E_FEC_FIRECODE] = ETHTOOL_FEC_BASER,
+	[MLX5E_FEC_RS_528_514] = ETHTOOL_FEC_RS,
+};
+
+static u32 pplm2ethtool_fec(u_long fec_mode, unsigned long size)
+{
+	int mode = 0;
+
+	if (!fec_mode)
+		return ETHTOOL_FEC_AUTO;
+
+	mode = find_first_bit(&fec_mode, size);
+
+	if (mode < ARRAY_SIZE(pplm_fec_2_ethtool))
+		return pplm_fec_2_ethtool[mode];
+
+	return 0;
+}
+
+/* we use ETHTOOL_FEC_* offset and apply it to ETHTOOL_LINK_MODE_FEC_*_BIT */
+static u32 ethtool_fec2ethtool_caps(u_long ethtool_fec_code)
+{
+	u32 offset;
+
+	offset = find_first_bit(&ethtool_fec_code, sizeof(u32));
+	offset -= ETHTOOL_FEC_OFF_BIT;
+	offset += ETHTOOL_LINK_MODE_FEC_NONE_BIT;
+
+	return offset;
+}
+
+static int get_fec_supported_advertised(struct mlx5_core_dev *dev,
+					struct ethtool_link_ksettings *link_ksettings)
+{
+	u_long fec_caps = 0;
+	u32 active_fec = 0;
+	u32 offset;
+	u32 bitn;
+	int err;
+
+	err = mlx5e_get_fec_caps(dev, (u8 *)&fec_caps);
+	if (err)
+		return (err == -EOPNOTSUPP) ? 0 : err;
+
+	err = mlx5e_get_fec_mode(dev, &active_fec, NULL);
+	if (err)
+		return err;
+
+	for_each_set_bit(bitn, &fec_caps, ARRAY_SIZE(pplm_fec_2_ethtool)) {
+		u_long ethtool_bitmask = pplm_fec_2_ethtool[bitn];
+
+		offset = ethtool_fec2ethtool_caps(ethtool_bitmask);
+		__set_bit(offset, link_ksettings->link_modes.supported);
+	}
+
+	active_fec = pplm2ethtool_fec(active_fec, sizeof(u32) * BITS_PER_BYTE);
+	offset = ethtool_fec2ethtool_caps(active_fec);
+	__set_bit(offset, link_ksettings->link_modes.advertising);
+
+	return 0;
+}
+
 static void ptys2ethtool_supported_advertised_port(struct ethtool_link_ksettings *link_ksettings,
 						   u32 eth_proto_cap,
 						   u8 connector_type)
@@ -742,7 +806,7 @@ static int mlx5e_get_link_ksettings(struct net_device *netdev,
 	if (err) {
 		netdev_err(netdev, "%s: query port ptys failed: %d\n",
 			   __func__, err);
-		goto err_query_ptys;
+		goto err_query_regs;
 	}
 
 	eth_proto_cap    = MLX5_GET(ptys_reg, out, eth_proto_capability);
@@ -778,11 +842,17 @@ static int mlx5e_get_link_ksettings(struct net_device *netdev,
 							  AUTONEG_ENABLE;
 	ethtool_link_ksettings_add_link_mode(link_ksettings, supported,
 					     Autoneg);
+
+	err = get_fec_supported_advertised(mdev, link_ksettings);
+	if (err)
+		netdev_dbg(netdev, "%s: FEC caps query failed: %d\n",
+			   __func__, err);
+
 	if (!an_disable_admin)
 		ethtool_link_ksettings_add_link_mode(link_ksettings,
 						     advertising, Autoneg);
 
-err_query_ptys:
+err_query_regs:
 	return err;
 }
 
@@ -1277,6 +1347,58 @@ static int mlx5e_set_wol(struct net_device *netdev, struct ethtool_wolinfo *wol)
 	return mlx5_set_port_wol(mdev, mlx5_wol_mode);
 }
 
+static int mlx5e_get_fecparam(struct net_device *netdev,
+			      struct ethtool_fecparam *fecparam)
+{
+	struct mlx5e_priv *priv = netdev_priv(netdev);
+	struct mlx5_core_dev *mdev = priv->mdev;
+	u8 fec_configured = 0;
+	u32 fec_active = 0;
+	int err;
+
+	err = mlx5e_get_fec_mode(mdev, &fec_active, &fec_configured);
+
+	if (err)
+		return err;
+
+	fecparam->active_fec = pplm2ethtool_fec((u_long)fec_active,
+						sizeof(u32) * BITS_PER_BYTE);
+
+	if (!fecparam->active_fec)
+		return -EOPNOTSUPP;
+
+	fecparam->fec = pplm2ethtool_fec((u_long)fec_configured,
+					 sizeof(u8) * BITS_PER_BYTE);
+
+	return 0;
+}
+
+static int mlx5e_set_fecparam(struct net_device *netdev,
+			      struct ethtool_fecparam *fecparam)
+{
+	struct mlx5e_priv *priv = netdev_priv(netdev);
+	struct mlx5_core_dev *mdev = priv->mdev;
+	u8 fec_policy = 0;
+	int mode;
+	int err;
+
+	for (mode = 0; mode < ARRAY_SIZE(pplm_fec_2_ethtool); mode++) {
+		if (!(pplm_fec_2_ethtool[mode] & fecparam->fec))
+			continue;
+		fec_policy |= (1 << mode);
+		break;
+	}
+
+	err = mlx5e_set_fec_mode(mdev, fec_policy);
+
+	if (err)
+		return err;
+
+	mlx5_toggle_port_link(mdev);
+
+	return 0;
+}
+
 static u32 mlx5e_get_msglevel(struct net_device *dev)
 {
 	return ((struct mlx5e_priv *)netdev_priv(dev))->msglevel;
@@ -1699,4 +1821,6 @@ const struct ethtool_ops mlx5e_ethtool_ops = {
 	.self_test         = mlx5e_self_test,
 	.get_msglevel      = mlx5e_get_msglevel,
 	.set_msglevel      = mlx5e_set_msglevel,
+	.get_fecparam      = mlx5e_get_fecparam,
+	.set_fecparam      = mlx5e_set_fecparam,
 };
-- 
2.17.2

[net-next 7/8] net/mlx5: Added "per_lane_error_counters" cap bit to PCAM

[Saeed Mahameed]

From: Shay Agroskin <shayag@mellanox.com>

Added "Per lane raw errors" capability bit in
Ports Capabilities Mask (PCAM) enhanced features
layout.

This bit determines if the fields "phy_raw_errors_laneX"
in "Physical Layer statistical" counters group are supported.

Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 include/linux/mlx5/mlx5_ifc.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 47b09a742ae5..dbff9ff28f2c 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -8140,7 +8140,8 @@ struct mlx5_ifc_pcam_enhanced_features_bits {
 	u8         rx_icrc_encapsulated_counter[0x1];
 	u8	   reserved_at_6e[0x8];
 	u8         pfcc_mask[0x1];
-	u8         reserved_at_77[0x4];
+	u8         reserved_at_77[0x3];
+	u8         per_lane_error_counters[0x1];
 	u8         rx_buffer_fullness_counters[0x1];
 	u8         ptys_connector_type[0x1];
 	u8         reserved_at_7d[0x1];
-- 
2.17.2

[net-next 8/8] net/mlx5e: Added 'raw_errors_laneX' fields to ethtool statistics

[Saeed Mahameed]

From: Shay Agroskin <shayag@mellanox.com>

These are counters for errors received on rx side, such as
FEC errors.

Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/en_stats.c    | 46 +++++++++++++++++--
 1 file changed, 41 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
index 1c006869a642..1e55b9c27ffc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
@@ -614,46 +614,82 @@ static const struct counter_desc pport_phy_statistical_stats_desc[] = {
 	{ "rx_corrected_bits_phy", PPORT_PHY_STATISTICAL_OFF(phy_corrected_bits) },
 };
 
-#define NUM_PPORT_PHY_STATISTICAL_COUNTERS ARRAY_SIZE(pport_phy_statistical_stats_desc)
+static const struct counter_desc
+pport_phy_statistical_err_lanes_stats_desc[] = {
+	{ "rx_err_lane_0_phy", PPORT_PHY_STATISTICAL_OFF(phy_corrected_bits_lane0) },
+	{ "rx_err_lane_1_phy", PPORT_PHY_STATISTICAL_OFF(phy_corrected_bits_lane1) },
+	{ "rx_err_lane_2_phy", PPORT_PHY_STATISTICAL_OFF(phy_corrected_bits_lane2) },
+	{ "rx_err_lane_3_phy", PPORT_PHY_STATISTICAL_OFF(phy_corrected_bits_lane3) },
+};
+
+#define NUM_PPORT_PHY_STATISTICAL_COUNTERS \
+	ARRAY_SIZE(pport_phy_statistical_stats_desc)
+#define NUM_PPORT_PHY_STATISTICAL_PER_LANE_COUNTERS \
+	ARRAY_SIZE(pport_phy_statistical_err_lanes_stats_desc)
 
 static int mlx5e_grp_phy_get_num_stats(struct mlx5e_priv *priv)
 {
+	struct mlx5_core_dev *mdev = priv->mdev;
+	int num_stats;
+
 	/* "1" for link_down_events special counter */
-	return MLX5_CAP_PCAM_FEATURE((priv)->mdev, ppcnt_statistical_group) ?
-		NUM_PPORT_PHY_STATISTICAL_COUNTERS + 1 : 1;
+	num_stats = 1;
+
+	num_stats += MLX5_CAP_PCAM_FEATURE(mdev, ppcnt_statistical_group) ?
+		     NUM_PPORT_PHY_STATISTICAL_COUNTERS : 0;
+
+	num_stats += MLX5_CAP_PCAM_FEATURE(mdev, per_lane_error_counters) ?
+		     NUM_PPORT_PHY_STATISTICAL_PER_LANE_COUNTERS : 0;
+
+	return num_stats;
 }
 
 static int mlx5e_grp_phy_fill_strings(struct mlx5e_priv *priv, u8 *data,
 				      int idx)
 {
+	struct mlx5_core_dev *mdev = priv->mdev;
 	int i;
 
 	strcpy(data + (idx++) * ETH_GSTRING_LEN, "link_down_events_phy");
 
-	if (!MLX5_CAP_PCAM_FEATURE((priv)->mdev, ppcnt_statistical_group))
+	if (!MLX5_CAP_PCAM_FEATURE(mdev, ppcnt_statistical_group))
 		return idx;
 
 	for (i = 0; i < NUM_PPORT_PHY_STATISTICAL_COUNTERS; i++)
 		strcpy(data + (idx++) * ETH_GSTRING_LEN,
 		       pport_phy_statistical_stats_desc[i].format);
+
+	if (MLX5_CAP_PCAM_FEATURE(mdev, per_lane_error_counters))
+		for (i = 0; i < NUM_PPORT_PHY_STATISTICAL_PER_LANE_COUNTERS; i++)
+			strcpy(data + (idx++) * ETH_GSTRING_LEN,
+			       pport_phy_statistical_err_lanes_stats_desc[i].format);
+
 	return idx;
 }
 
 static int mlx5e_grp_phy_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx)
 {
+	struct mlx5_core_dev *mdev = priv->mdev;
 	int i;
 
 	/* link_down_events_phy has special handling since it is not stored in __be64 format */
 	data[idx++] = MLX5_GET(ppcnt_reg, priv->stats.pport.phy_counters,
 			       counter_set.phys_layer_cntrs.link_down_events);
 
-	if (!MLX5_CAP_PCAM_FEATURE((priv)->mdev, ppcnt_statistical_group))
+	if (!MLX5_CAP_PCAM_FEATURE(mdev, ppcnt_statistical_group))
 		return idx;
 
 	for (i = 0; i < NUM_PPORT_PHY_STATISTICAL_COUNTERS; i++)
 		data[idx++] =
 			MLX5E_READ_CTR64_BE(&priv->stats.pport.phy_statistical_counters,
 					    pport_phy_statistical_stats_desc, i);
+
+	if (MLX5_CAP_PCAM_FEATURE(mdev, per_lane_error_counters))
+		for (i = 0; i < NUM_PPORT_PHY_STATISTICAL_PER_LANE_COUNTERS; i++)
+			data[idx++] =
+				MLX5E_READ_CTR64_BE(&priv->stats.pport.phy_statistical_counters,
+						    pport_phy_statistical_err_lanes_stats_desc,
+						    i);
 	return idx;
 }
 
-- 
2.17.2

Re: [pull request][net-next 0/8] Mellanox, mlx5 updates 2018-10-18

[David Miller]

From: Saeed Mahameed <saeedm@mellanox.com>
Date: Thu, 18 Oct 2018 13:38:59 -0700

> This pull request provides some misc updates to mlx5 core and netdevice
> driver.
> 
> For more details please see tag log below.
> 
> Please pull and let me know if there's any problem.

Pulled, thanks Saeed.