From: Aya Levin <a...@mellanox.com>
Add support for report and recovery from error on completion on RQ by
setting the queue back to ready state. Handle only errors with a
syndrome indicating the RQ might enter error state and could be
recovered.
Signed-off-by: Aya Levin <a...@mellanox.com>
Reviewed-by: Tariq Toukan <tar...@mellanox.com>
Acked-by: Jiri Pirko <j...@mellanox.com>
Signed-off-by: Saeed Mahameed <sae...@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 3 +
.../ethernet/mellanox/mlx5/core/en/health.h | 9 +++
.../mellanox/mlx5/core/en/reporter_rx.c | 66 +++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/en_main.c | 9 +++
.../net/ethernet/mellanox/mlx5/core/en_rx.c | 11 ++++
5 files changed, 98 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h
b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 5f2a1d14de68..822f7b620640 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -295,6 +295,7 @@ struct mlx5e_dcbx_dp {
enum {
MLX5E_RQ_STATE_ENABLED,
+ MLX5E_RQ_STATE_RECOVERING,
MLX5E_RQ_STATE_AM,
MLX5E_RQ_STATE_NO_CSUM_COMPLETE,
MLX5E_RQ_STATE_CSUM_FULL, /* cqe_csum_full hw bit is set */
@@ -667,6 +668,8 @@ struct mlx5e_rq {
struct zero_copy_allocator zca;
struct xdp_umem *umem;
+ struct work_struct recover_work;
+
/* control */
struct mlx5_wq_ctrl wq_ctrl;
__be32 mkey_be;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 52e9ca37cf46..d3693fa547ac 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -8,6 +8,14 @@
#define MLX5E_RX_ERR_CQE(cqe) (get_cqe_opcode(cqe) != MLX5_CQE_RESP_SEND)
+static inline bool cqe_syndrome_needs_recover(u8 syndrome)
+{
+ return syndrome == MLX5_CQE_SYNDROME_LOCAL_LENGTH_ERR ||
+ syndrome == MLX5_CQE_SYNDROME_LOCAL_QP_OP_ERR ||
+ syndrome == MLX5_CQE_SYNDROME_LOCAL_PROT_ERR ||
+ syndrome == MLX5_CQE_SYNDROME_WR_FLUSH_ERR;
+}
+
int 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);
@@ -21,6 +29,7 @@ int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg
*fmsg);
int 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);
void mlx5e_reporter_rx_timeout(struct mlx5e_rq *rq);
#define MLX5E_REPORTER_PER_Q_MAX_LEN 256
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 4f5547ac4bee..b4f7e535dbc7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -111,6 +111,72 @@ void mlx5e_reporter_icosq_cqe_err(struct mlx5e_icosq
*icosq)
mlx5e_health_report(priv, priv->rx_reporter, err_str, &err_ctx);
}
+static int mlx5e_rq_to_ready(struct mlx5e_rq *rq, int curr_state)
+{
+ struct net_device *dev = rq->netdev;
+ int err;
+
+ err = mlx5e_modify_rq_state(rq, curr_state, MLX5_RQC_STATE_RST);
+ if (err) {
+ netdev_err(dev, "Failed to move rq 0x%x to reset\n", rq->rqn);
+ return err;
+ }
+ err = mlx5e_modify_rq_state(rq, MLX5_RQC_STATE_RST, MLX5_RQC_STATE_RDY);
+ if (err) {
+ netdev_err(dev, "Failed to move rq 0x%x to ready\n", rq->rqn);
+ return err;
+ }
+
+ return 0;
+}
+
+static int mlx5e_rx_reporter_err_rq_cqe_recover(void *ctx)
+{
+ struct mlx5e_rq *rq = ctx;
+ struct mlx5_core_dev *mdev = rq->mdev;
+ struct net_device *dev = rq->netdev;
+ u8 state;
+ int err;
+
+ 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);
+
+ err = mlx5e_rq_to_ready(rq, MLX5_RQC_STATE_ERR);
+ if (err)
+ goto out;
+
+ clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state);
+ mlx5e_activate_rq(rq);
+ rq->stats->recover++;
+ return 0;
+out:
+ clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state);
+ return err;
+}
+
+void mlx5e_reporter_rq_cqe_err(struct mlx5e_rq *rq)
+{
+ struct mlx5e_priv *priv = rq->channel->priv;
+ char err_str[MLX5E_REPORTER_PER_Q_MAX_LEN];
+ struct mlx5e_err_ctx err_ctx = {};
+
+ err_ctx.ctx = rq;
+ err_ctx.recover = mlx5e_rx_reporter_err_rq_cqe_recover;
+ sprintf(err_str, "ERR CQE on RQ: 0x%x", rq->rqn);
+
+ mlx5e_health_report(priv, priv->rx_reporter, err_str, &err_ctx);
+}
+
static int mlx5e_rx_reporter_timeout_recover(void *ctx)
{
struct mlx5e_rq *rq = ctx;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 76845bafd708..77f0c8fad9df 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -362,6 +362,13 @@ static void mlx5e_free_di_list(struct mlx5e_rq *rq)
kvfree(rq->wqe.di);
}
+static void mlx5e_rq_err_cqe_work(struct work_struct *recover_work)
+{
+ struct mlx5e_rq *rq = container_of(recover_work, struct mlx5e_rq,
recover_work);
+
+ mlx5e_reporter_rq_cqe_err(rq);
+}
+
static int mlx5e_alloc_rq(struct mlx5e_channel *c,
struct mlx5e_params *params,
struct mlx5e_xsk_param *xsk,
@@ -398,6 +405,7 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c,
rq->stats = &c->priv->channel_stats[c->ix].xskrq;
else
rq->stats = &c->priv->channel_stats[c->ix].rq;
+ INIT_WORK(&rq->recover_work, mlx5e_rq_err_cqe_work);
rq->xdp_prog = params->xdp_prog ? bpf_prog_inc(params->xdp_prog) : NULL;
if (IS_ERR(rq->xdp_prog)) {
@@ -907,6 +915,7 @@ void mlx5e_close_rq(struct mlx5e_rq *rq)
{
cancel_work_sync(&rq->dim.work);
cancel_work_sync(&rq->channel->icosq.recover_work);
+ cancel_work_sync(&rq->recover_work);
mlx5e_destroy_rq(rq);
mlx5e_free_rx_descs(rq);
mlx5e_free_rq(rq);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 43d790b7d4ec..2fd2760d0bb7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1130,6 +1130,15 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct
mlx5_cqe64 *cqe,
return skb;
}
+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))
+ queue_work(rq->channel->priv->wq, &rq->recover_work);
+}
+
void mlx5e_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
{
struct mlx5_wq_cyc *wq = &rq->wqe.wq;
@@ -1143,6 +1152,7 @@ void mlx5e_handle_rx_cqe(struct mlx5e_rq *rq, struct
mlx5_cqe64 *cqe)
cqe_bcnt = be32_to_cpu(cqe->byte_cnt);
if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
+ trigger_report(rq, cqe);
rq->stats->wqe_err++;
goto free_wqe;
}
@@ -1328,6 +1338,7 @@ void mlx5e_handle_rx_cqe_mpwrq(struct mlx5e_rq *rq,
struct mlx5_cqe64 *cqe)
wi->consumed_strides += cstrides;
if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
+ trigger_report(rq, cqe);
rq->stats->wqe_err++;
goto mpwrq_cqe_out;
}
--
2.21.0
