* [net-next v2 11/16] net/mlx5e: Report and recover from rx timeout
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Add support for report and recovery from rx timeout. On driver open we
post NOP work request on the rx channels to trigger napi in order to
fillup the rx rings. In case napi wasn't scheduled due to a lost
interrupt, perform EQ recovery.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../ethernet/mellanox/mlx5/core/en/health.h | 1 +
.../mellanox/mlx5/core/en/reporter_rx.c | 32 +++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/en_main.c | 1 +
3 files changed, 34 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 8acd9dc520cf..b4a2d9be17d6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -19,6 +19,7 @@ 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_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 ff49853b65d2..05450df87554 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -115,6 +115,38 @@ void mlx5e_reporter_icosq_cqe_err(struct mlx5e_icosq *icosq)
mlx5e_health_report(priv, priv->rx_reporter, err_str, &err_ctx);
}
+static int mlx5e_rx_reporter_timeout_recover(void *ctx)
+{
+ struct mlx5e_icosq *icosq;
+ struct mlx5_eq_comp *eq;
+ struct mlx5e_rq *rq;
+ int err;
+
+ rq = ctx;
+ icosq = &rq->channel->icosq;
+ eq = rq->cq.mcq.eq;
+ err = mlx5e_health_channel_eq_recover(eq, rq->channel);
+ if (err)
+ clear_bit(MLX5E_SQ_STATE_ENABLED, &icosq->state);
+
+ return err;
+}
+
+void mlx5e_reporter_rx_timeout(struct mlx5e_rq *rq)
+{
+ struct mlx5e_icosq *icosq = &rq->channel->icosq;
+ 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_timeout_recover;
+ sprintf(err_str, "RX timeout on channel: %d, ICOSQ: 0x%x RQ: 0x%x, CQ: 0x%x\n",
+ icosq->channel->ix, icosq->sqn, rq->rqn, rq->cq.mcq.cqn);
+
+ mlx5e_health_report(priv, priv->rx_reporter, err_str, &err_ctx);
+}
+
static int mlx5e_rx_reporter_recover_from_ctx(struct mlx5e_err_ctx *err_ctx)
{
return err_ctx->recover(err_ctx->ctx);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 0e43ea1c27b0..54f320391f63 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -808,6 +808,7 @@ int mlx5e_wait_for_min_rx_wqes(struct mlx5e_rq *rq, int wait_time)
netdev_warn(c->netdev, "Failed to get min RX wqes on Channel[%d] RQN[0x%x] wq cur_sz(%d) min_rx_wqes(%d)\n",
c->ix, rq->rqn, mlx5e_rqwq_get_cur_sz(rq), min_wqes);
+ mlx5e_reporter_rx_timeout(rq);
return -ETIMEDOUT;
}
--
2.21.0
^ permalink raw reply related
* [net-next v2 13/16] net/mlx5e: Report and recover from CQE with error on RQ
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@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 <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@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 | 69 +++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/en_main.c | 9 +++
.../net/ethernet/mellanox/mlx5/core/en_rx.c | 11 +++
5 files changed, 101 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 842adf3719ff..7316571a4df5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -300,6 +300,7 @@ 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 */
@@ -672,6 +673,8 @@ 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 05450df87554..b860569d4247 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -115,6 +115,75 @@ 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 mlx5_core_dev *mdev;
+ struct net_device *dev;
+ struct mlx5e_rq *rq;
+ u8 state;
+ 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);
+
+ 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_icosq *icosq;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 54f320391f63..7fdea6479ff6 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
^ permalink raw reply related
* [net-next v2 14/16] Documentation: net: mlx5: Devlink health documentation updates
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller; +Cc: netdev@vger.kernel.org, Aya Levin, Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Add documentation for devlink health rx reporter supported by mlx5.
Update tx reporter documentation.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../device_drivers/mellanox/mlx5.rst | 33 +++++++++++++++++--
1 file changed, 31 insertions(+), 2 deletions(-)
diff --git a/Documentation/networking/device_drivers/mellanox/mlx5.rst b/Documentation/networking/device_drivers/mellanox/mlx5.rst
index 214325897732..cfda464e52de 100644
--- a/Documentation/networking/device_drivers/mellanox/mlx5.rst
+++ b/Documentation/networking/device_drivers/mellanox/mlx5.rst
@@ -126,7 +126,7 @@ Devlink health reporters
tx reporter
-----------
-The tx reporter is responsible of two error scenarios:
+The tx reporter is responsible for reporting and recovering of the following two error scenarios:
- TX timeout
Report on kernel tx timeout detection.
@@ -135,7 +135,7 @@ The tx reporter is responsible of two error scenarios:
Report on error tx completion.
Recover by flushing the TX queue and reset it.
-TX reporter also support Diagnose callback, on which it provides
+TX reporter also support on demand diagnose callback, on which it provides
real time information of its send queues status.
User commands examples:
@@ -144,11 +144,40 @@ User commands examples:
$ devlink health diagnose pci/0000:82:00.0 reporter tx
+NOTE: This command has valid output only when interface is up, otherwise the command has empty output.
+
- Show number of tx errors indicated, number of recover flows ended successfully,
is autorecover enabled and graceful period from last recover::
$ devlink health show pci/0000:82:00.0 reporter tx
+rx reporter
+-----------
+The rx reporter is responsible for reporting and recovering of the following two error scenarios:
+
+- RX queues initialization (population) timeout
+ RX queues descriptors population on ring initialization is done in
+ napi context via triggering an irq, in case of a failure to get
+ the minimum amount of descriptors, a timeout would occur and it
+ could be recoverable by polling the EQ (Event Queue).
+- RX completions with errors (reported by HW on interrupt context)
+ Report on rx completion error.
+ Recover (if needed) by flushing the related queue and reset it.
+
+RX reporter also supports on demand diagnose callback, on which it
+provides real time information of its receive queues status.
+
+- Diagnose rx queues status, and corresponding completion queue::
+
+ $ devlink health diagnose pci/0000:82:00.0 reporter rx
+
+NOTE: This command has valid output only when interface is up, otherwise the command has empty output.
+
+- Show number of rx errors indicated, number of recover flows ended successfully,
+ is autorecover enabled and graceful period from last recover::
+
+ $ devlink health show pci/0000:82:00.0 reporter rx
+
fw reporter
-----------
The fw reporter implements diagnose and dump callbacks.
--
2.21.0
^ permalink raw reply related
* [net-next v2 12/16] net/mlx5e: RX, Handle CQE with error at the earliest stage
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Saeed Mahameed, Aya Levin, Tariq Toukan,
Jiri Pirko
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
Just to be aligned with the MPWQE handlers, handle RX WQE with error
for legacy RQs in the top RX handlers, just before calling skb_from_cqe().
CQE error handling will now be called at the same stage regardless of
the RQ type or netdev mode NIC, Representor, IPoIB, etc ..
This will be useful for down stream patch to improve error CQE
handling.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../ethernet/mellanox/mlx5/core/en/health.h | 2 +
.../net/ethernet/mellanox/mlx5/core/en_rx.c | 49 +++++++++++--------
2 files changed, 30 insertions(+), 21 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index b4a2d9be17d6..52e9ca37cf46 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -6,6 +6,8 @@
#include "en.h"
+#define MLX5E_RX_ERR_CQE(cqe) (get_cqe_opcode(cqe) != MLX5_CQE_RESP_SEND)
+
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);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 0f6033ea475d..43d790b7d4ec 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -48,6 +48,7 @@
#include "lib/clock.h"
#include "en/xdp.h"
#include "en/xsk/rx.h"
+#include "en/health.h"
static inline bool mlx5e_rx_hw_stamp(struct hwtstamp_config *config)
{
@@ -1069,11 +1070,6 @@ mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
prefetchw(va); /* xdp_frame data area */
prefetch(data);
- if (unlikely(get_cqe_opcode(cqe) != MLX5_CQE_RESP_SEND)) {
- rq->stats->wqe_err++;
- return NULL;
- }
-
rcu_read_lock();
consumed = mlx5e_xdp_handle(rq, di, va, &rx_headroom, &cqe_bcnt, false);
rcu_read_unlock();
@@ -1101,11 +1097,6 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
u16 byte_cnt = cqe_bcnt - headlen;
struct sk_buff *skb;
- if (unlikely(get_cqe_opcode(cqe) != MLX5_CQE_RESP_SEND)) {
- rq->stats->wqe_err++;
- return NULL;
- }
-
/* XDP is not supported in this configuration, as incoming packets
* might spread among multiple pages.
*/
@@ -1151,6 +1142,11 @@ void mlx5e_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
wi = get_frag(rq, ci);
cqe_bcnt = be32_to_cpu(cqe->byte_cnt);
+ if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
+ rq->stats->wqe_err++;
+ goto free_wqe;
+ }
+
skb = INDIRECT_CALL_2(rq->wqe.skb_from_cqe,
mlx5e_skb_from_cqe_linear,
mlx5e_skb_from_cqe_nonlinear,
@@ -1192,6 +1188,11 @@ void mlx5e_handle_rx_cqe_rep(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
wi = get_frag(rq, ci);
cqe_bcnt = be32_to_cpu(cqe->byte_cnt);
+ if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
+ rq->stats->wqe_err++;
+ goto free_wqe;
+ }
+
skb = rq->wqe.skb_from_cqe(rq, cqe, wi, cqe_bcnt);
if (!skb) {
/* probably for XDP */
@@ -1326,7 +1327,7 @@ void mlx5e_handle_rx_cqe_mpwrq(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
wi->consumed_strides += cstrides;
- if (unlikely(get_cqe_opcode(cqe) != MLX5_CQE_RESP_SEND)) {
+ if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
rq->stats->wqe_err++;
goto mpwrq_cqe_out;
}
@@ -1502,6 +1503,11 @@ void mlx5i_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
wi = get_frag(rq, ci);
cqe_bcnt = be32_to_cpu(cqe->byte_cnt);
+ if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
+ rq->stats->wqe_err++;
+ goto wq_free_wqe;
+ }
+
skb = INDIRECT_CALL_2(rq->wqe.skb_from_cqe,
mlx5e_skb_from_cqe_linear,
mlx5e_skb_from_cqe_nonlinear,
@@ -1537,26 +1543,27 @@ void mlx5e_ipsec_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe)
wi = get_frag(rq, ci);
cqe_bcnt = be32_to_cpu(cqe->byte_cnt);
+ if (unlikely(MLX5E_RX_ERR_CQE(cqe))) {
+ rq->stats->wqe_err++;
+ goto wq_free_wqe;
+ }
+
skb = INDIRECT_CALL_2(rq->wqe.skb_from_cqe,
mlx5e_skb_from_cqe_linear,
mlx5e_skb_from_cqe_nonlinear,
rq, cqe, wi, cqe_bcnt);
- if (unlikely(!skb)) {
- /* a DROP, save the page-reuse checks */
- mlx5e_free_rx_wqe(rq, wi, true);
- goto wq_cyc_pop;
- }
+ if (unlikely(!skb)) /* a DROP, save the page-reuse checks */
+ goto wq_free_wqe;
+
skb = mlx5e_ipsec_handle_rx_skb(rq->netdev, skb, &cqe_bcnt);
- if (unlikely(!skb)) {
- mlx5e_free_rx_wqe(rq, wi, true);
- goto wq_cyc_pop;
- }
+ if (unlikely(!skb))
+ goto wq_free_wqe;
mlx5e_complete_rx_cqe(rq, cqe, cqe_bcnt, skb);
napi_gro_receive(rq->cq.napi, skb);
+wq_free_wqe:
mlx5e_free_rx_wqe(rq, wi, true);
-wq_cyc_pop:
mlx5_wq_cyc_pop(wq);
}
--
2.21.0
^ permalink raw reply related
* [net-next v2 15/16] net/mlx5e: Fix deallocation of non-fully init encap entries
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller; +Cc: netdev@vger.kernel.org, Vlad Buslov, Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Vlad Buslov <vladbu@mellanox.com>
Recent rtnl lock dependency refactoring changed encap entry attach code to
insert encap entry to hash table before it was fully initialized in order
to allow concurrent tc users to wait on completion for encap entry to
finish initialization. That change required all the users of encap entry to
obtain reference to it first and for caller that creates encap to put
reference to it on error, instead of freeing the entry memory directly.
However, releasing reference to such encap entry that wasn't fully
initialized causes NULL pointer dereference in
mlx5e_rep_encap_entry_detach() which expects e->out_dev to be set and encap
to be attached to nhe:
[ 1092.454517] BUG: unable to handle page fault for address: 00000000000420e8
[ 1092.454571] #PF: supervisor read access in kernel mode
[ 1092.454602] #PF: error_code(0x0000) - not-present page
[ 1092.454632] PGD 800000083032c067 P4D 800000083032c067 PUD 84107d067 PMD 0
[ 1092.454673] Oops: 0000 [#1] SMP PTI
[ 1092.454697] CPU: 20 PID: 22393 Comm: tc Not tainted 5.3.0-rc3+ #589
[ 1092.454733] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017
[ 1092.454806] RIP: 0010:mlx5e_rep_encap_entry_detach+0x1c/0x630 [mlx5_core]
[ 1092.454845] Code: be f4 ff ff ff e9 11 ff ff ff 0f 1f 40 00 0f 1f 44 00 00 55 48 89 e5 41 57 41 56 41 55 41 54 49 89 fc 53 48 89 f3 48 83 ec 30 <48> 8b 87 28 16 04 00 48 89 f7 48 05 d0 03 00 00 48 89
45 c8 e8 cb
[ 1092.454942] RSP: 0018:ffffb6f08421f5a0 EFLAGS: 00010286
[ 1092.454974] RAX: 0000000000000000 RBX: ffff8ab668644e00 RCX: ffffb6f08421f56c
[ 1092.455013] RDX: ffff8ab668644e40 RSI: ffff8ab668644e00 RDI: 0000000000000ac0
[ 1092.455053] RBP: ffffb6f08421f5f8 R08: 0000000000000001 R09: 0000000000000000
[ 1092.455092] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000ac0
[ 1092.455131] R13: 00000000ffffff9b R14: ffff8ab63f200ac0 R15: ffff8ab668644e40
[ 1092.455171] FS: 00007fa195bdc480(0000) GS:ffff8ab66fa00000(0000) knlGS:0000000000000000
[ 1092.455216] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 1092.455249] CR2: 00000000000420e8 CR3: 0000000867522001 CR4: 00000000001606e0
[ 1092.455288] Call Trace:
[ 1092.455315] ? __mutex_unlock_slowpath+0x4d/0x2a0
[ 1092.455365] mlx5e_encap_dealloc.isra.0+0x31/0x60 [mlx5_core]
[ 1092.455424] mlx5e_tc_add_fdb_flow+0x596/0x750 [mlx5_core]
[ 1092.455484] __mlx5e_add_fdb_flow+0x152/0x210 [mlx5_core]
[ 1092.455534] mlx5e_configure_flower+0x4d5/0xe30 [mlx5_core]
[ 1092.455574] tc_setup_cb_call+0x67/0xb0
[ 1092.455601] fl_hw_replace_filter+0x142/0x300 [cls_flower]
[ 1092.455639] fl_change+0xd24/0x1bdb [cls_flower]
[ 1092.455675] tc_new_tfilter+0x3e0/0x970
[ 1092.455709] ? tc_del_tfilter+0x720/0x720
[ 1092.455735] rtnetlink_rcv_msg+0x389/0x4b0
[ 1092.455763] ? netlink_deliver_tap+0x95/0x400
[ 1092.455791] ? rtnl_dellink+0x2d0/0x2d0
[ 1092.455817] netlink_rcv_skb+0x49/0x110
[ 1092.455844] netlink_unicast+0x171/0x200
[ 1092.455872] netlink_sendmsg+0x224/0x3f0
[ 1092.455901] sock_sendmsg+0x5e/0x60
[ 1092.455924] ___sys_sendmsg+0x2ae/0x330
[ 1092.455950] ? task_work_add+0x43/0x50
[ 1092.455976] ? fput_many+0x45/0x80
[ 1092.456004] ? __lock_acquire+0x248/0x18e0
[ 1092.456033] ? find_held_lock+0x2b/0x80
[ 1092.456058] ? task_work_run+0x7b/0xd0
[ 1092.456085] __sys_sendmsg+0x59/0xa0
[ 1092.457013] do_syscall_64+0x5c/0xb0
[ 1092.457924] entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 1092.458842] RIP: 0033:0x7fa195da27b8
[ 1092.459918] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83
ec 28 89 54
[ 1092.462634] RSP: 002b:00007fff94409298 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
[ 1092.464011] RAX: ffffffffffffffda RBX: 000000005d515b0e RCX: 00007fa195da27b8
[ 1092.465391] RDX: 0000000000000000 RSI: 00007fff94409300 RDI: 0000000000000003
[ 1092.466761] RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000006
[ 1092.468121] R10: 0000000000404ec2 R11: 0000000000000246 R12: 0000000000000001
[ 1092.469456] R13: 0000000000480640 R14: 0000000000000016 R15: 0000000000000001
[ 1092.470766] Modules linked in: act_mirred act_tunnel_key cls_flower dummy vxlan ip6_udp_tunnel udp_tunnel sch_ingress nfsv3 nfs_acl nfs lockd grace fscache tun bridge stp llc sunrpc rdma_ucm rdma_cm
iw_cm ib_cm mlx5_ib ib_uverbs ib_core intel_rapl_msr intel_rapl_common sb_edac x86_pkg_temp_thermal intel_powerclamp coretemp mlx5_core kvm_intel kvm irqbypass crct10dif_pclmul mei_me crc32_pclmul crc32
c_intel igb iTCO_wdt ghash_clmulni_intel ses mlxfw intel_cstate iTCO_vendor_support ptp intel_uncore lpc_ich pps_core mei i2c_i801 joydev intel_rapl_perf ioatdma enclosure ipmi_ssif pcspkr dca wmi ipmi_
si ipmi_devintf ipmi_msghandler acpi_pad acpi_power_meter ast i2c_algo_bit drm_vram_helper ttm drm_kms_helper drm mpt3sas raid_class scsi_transport_sas
[ 1092.479618] CR2: 00000000000420e8
[ 1092.481214] ---[ end trace ce2e0f4d9a67f604 ]---
To fix the issue, set e->compl_result to positive value after encap was
initialized successfully. Check e->compl_result value in
mlx5e_encap_dealloc() and only detach and dealloc encap if the value is
positive.
Fixes: d589e785baf5 ("net/mlx5e: Allow concurrent creation of encap entries")
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index c57f7533a6d0..3917834b48ff 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1481,10 +1481,13 @@ void mlx5e_tc_update_neigh_used_value(struct mlx5e_neigh_hash_entry *nhe)
static void mlx5e_encap_dealloc(struct mlx5e_priv *priv, struct mlx5e_encap_entry *e)
{
WARN_ON(!list_empty(&e->flows));
- mlx5e_rep_encap_entry_detach(netdev_priv(e->out_dev), e);
- if (e->flags & MLX5_ENCAP_ENTRY_VALID)
- mlx5_packet_reformat_dealloc(priv->mdev, e->encap_id);
+ if (e->compl_result > 0) {
+ mlx5e_rep_encap_entry_detach(netdev_priv(e->out_dev), e);
+
+ if (e->flags & MLX5_ENCAP_ENTRY_VALID)
+ mlx5_packet_reformat_dealloc(priv->mdev, e->encap_id);
+ }
kfree(e->encap_header);
kfree(e);
@@ -2919,7 +2922,7 @@ static int mlx5e_attach_encap(struct mlx5e_priv *priv,
/* Protect against concurrent neigh update. */
mutex_lock(&esw->offloads.encap_tbl_lock);
- if (e->compl_result) {
+ if (e->compl_result < 0) {
err = -EREMOTEIO;
goto out_err;
}
@@ -2959,6 +2962,7 @@ static int mlx5e_attach_encap(struct mlx5e_priv *priv,
e->compl_result = err;
goto out_err;
}
+ e->compl_result = 1;
attach_flow:
flow->encaps[out_index].e = e;
--
2.21.0
^ permalink raw reply related
* [net-next v2 16/16] net/mlx5: Fix the order of fc_stats cleanup
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Gavi Teitz, Roi Dayan, Vlad Buslov,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Gavi Teitz <gavi@mellanox.com>
Previously, mlx5_cleanup_fc_stats() would cleanup the flow counter
pool beofre releasing all the counters to it, which would result in
flow counter bulks not getting freed. Resolve this by changing the
order in which elements of fc_stats are cleaned up, so that the flow
counter pool is cleaned up after all the counters are released.
Also move cleanup actions for freeing the bulk query memory and
destroying the idr to the end of mlx5_cleanup_fc_stats().
Signed-off-by: Gavi Teitz <gavi@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c | 9 ++++-----
1 file changed, 4 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 1804cf3c3814..ab69effb056d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c
@@ -402,21 +402,20 @@ void mlx5_cleanup_fc_stats(struct mlx5_core_dev *dev)
struct mlx5_fc *counter;
struct mlx5_fc *tmp;
- mlx5_fc_pool_cleanup(&fc_stats->fc_pool);
cancel_delayed_work_sync(&dev->priv.fc_stats.work);
destroy_workqueue(dev->priv.fc_stats.wq);
dev->priv.fc_stats.wq = NULL;
- kfree(fc_stats->bulk_query_out);
-
- idr_destroy(&fc_stats->counters_idr);
-
tmplist = llist_del_all(&fc_stats->addlist);
llist_for_each_entry_safe(counter, tmp, tmplist, addlist)
mlx5_fc_release(dev, counter);
list_for_each_entry_safe(counter, tmp, &fc_stats->counters, list)
mlx5_fc_release(dev, counter);
+
+ mlx5_fc_pool_cleanup(&fc_stats->fc_pool);
+ idr_destroy(&fc_stats->counters_idr);
+ kfree(fc_stats->bulk_query_out);
}
int mlx5_fc_query(struct mlx5_core_dev *dev, struct mlx5_fc *counter,
--
2.21.0
^ permalink raw reply related
* [net-next v2 10/16] net/mlx5e: Report and recover from CQE error on ICOSQ
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Add support for report and recovery from error on completion on ICOSQ.
Deactivate RQ and flush, then deactivate ICOSQ. Set the queue back to
ready state (firmware) and reset the ICOSQ and the RQ (software
resources). Finally, activate the ICOSQ and the RQ.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 8 ++
.../ethernet/mellanox/mlx5/core/en/health.h | 1 +
.../mellanox/mlx5/core/en/reporter_rx.c | 109 +++++++++++++++++-
.../net/ethernet/mellanox/mlx5/core/en_main.c | 22 +++-
.../net/ethernet/mellanox/mlx5/core/en_rx.c | 2 +
.../ethernet/mellanox/mlx5/core/en_stats.c | 3 +
.../ethernet/mellanox/mlx5/core/en_stats.h | 2 +
7 files changed, 140 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 7381d57b4b82..842adf3719ff 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -551,6 +551,8 @@ struct mlx5e_icosq {
/* control path */
struct mlx5_wq_ctrl wq_ctrl;
struct mlx5e_channel *channel;
+
+ struct work_struct recover_work;
} ____cacheline_aligned_in_smp;
struct mlx5e_wqe_frag_info {
@@ -1026,6 +1028,12 @@ void mlx5e_set_rx_cq_mode_params(struct mlx5e_params *params,
void mlx5e_set_rq_type(struct mlx5_core_dev *mdev, struct mlx5e_params *params);
void mlx5e_init_rq_type_params(struct mlx5_core_dev *mdev,
struct mlx5e_params *params);
+int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state);
+void mlx5e_activate_rq(struct mlx5e_rq *rq);
+void mlx5e_deactivate_rq(struct mlx5e_rq *rq);
+void mlx5e_free_rx_descs(struct mlx5e_rq *rq);
+void mlx5e_activate_icosq(struct mlx5e_icosq *icosq);
+void mlx5e_deactivate_icosq(struct mlx5e_icosq *icosq);
int mlx5e_modify_sq(struct mlx5_core_dev *mdev, u32 sqn,
struct mlx5e_modify_sq_param *p);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index a751c5316baf..8acd9dc520cf 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -18,6 +18,7 @@ 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);
#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 24b9c819b208..ff49853b65d2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c
@@ -27,6 +27,109 @@ static int mlx5e_query_rq_state(struct mlx5_core_dev *dev, u32 rqn, u8 *state)
return err;
}
+static int mlx5e_wait_for_icosq_flush(struct mlx5e_icosq *icosq)
+{
+ unsigned long exp_time = jiffies + msecs_to_jiffies(2000);
+
+ while (time_before(jiffies, exp_time)) {
+ if (icosq->cc == icosq->pc)
+ return 0;
+
+ msleep(20);
+ }
+
+ netdev_err(icosq->channel->netdev,
+ "Wait for ICOSQ 0x%x flush timeout (cc = 0x%x, pc = 0x%x)\n",
+ icosq->sqn, icosq->cc, icosq->pc);
+
+ return -ETIMEDOUT;
+}
+
+static void mlx5e_reset_icosq_cc_pc(struct mlx5e_icosq *icosq)
+{
+ WARN_ONCE(icosq->cc != icosq->pc, "ICOSQ 0x%x: cc (0x%x) != pc (0x%x)\n",
+ icosq->sqn, icosq->cc, icosq->pc);
+ icosq->cc = 0;
+ icosq->pc = 0;
+}
+
+static int mlx5e_rx_reporter_err_icosq_cqe_recover(void *ctx)
+{
+ struct mlx5_core_dev *mdev;
+ struct mlx5e_icosq *icosq;
+ struct net_device *dev;
+ struct mlx5e_rq *rq;
+ u8 state;
+ int err;
+
+ icosq = ctx;
+ rq = &icosq->channel->rq;
+ mdev = icosq->channel->mdev;
+ dev = icosq->channel->netdev;
+ err = mlx5_core_query_sq_state(mdev, icosq->sqn, &state);
+ if (err) {
+ netdev_err(dev, "Failed to query ICOSQ 0x%x state. err = %d\n",
+ icosq->sqn, err);
+ goto out;
+ }
+
+ if (state != MLX5_SQC_STATE_ERR)
+ goto out;
+
+ mlx5e_deactivate_rq(rq);
+ err = mlx5e_wait_for_icosq_flush(icosq);
+ if (err)
+ goto out;
+
+ mlx5e_deactivate_icosq(icosq);
+
+ /* At this point, both the rq and the icosq are disabled */
+
+ err = mlx5e_health_sq_to_ready(icosq->channel, icosq->sqn);
+ if (err)
+ goto out;
+
+ mlx5e_reset_icosq_cc_pc(icosq);
+ mlx5e_free_rx_descs(rq);
+ clear_bit(MLX5E_SQ_STATE_RECOVERING, &icosq->state);
+ mlx5e_activate_icosq(icosq);
+ mlx5e_activate_rq(rq);
+
+ rq->stats->recover++;
+ return 0;
+out:
+ clear_bit(MLX5E_SQ_STATE_RECOVERING, &icosq->state);
+ return err;
+}
+
+void mlx5e_reporter_icosq_cqe_err(struct mlx5e_icosq *icosq)
+{
+ struct mlx5e_priv *priv = icosq->channel->priv;
+ char err_str[MLX5E_REPORTER_PER_Q_MAX_LEN];
+ struct mlx5e_err_ctx err_ctx = {};
+
+ err_ctx.ctx = icosq;
+ err_ctx.recover = mlx5e_rx_reporter_err_icosq_cqe_recover;
+ sprintf(err_str, "ERR CQE on ICOSQ: 0x%x", icosq->sqn);
+
+ mlx5e_health_report(priv, priv->rx_reporter, err_str, &err_ctx);
+}
+
+static int mlx5e_rx_reporter_recover_from_ctx(struct mlx5e_err_ctx *err_ctx)
+{
+ return err_ctx->recover(err_ctx->ctx);
+}
+
+static int mlx5e_rx_reporter_recover(struct devlink_health_reporter *reporter,
+ void *context)
+{
+ struct mlx5e_priv *priv = devlink_health_reporter_priv(reporter);
+ struct mlx5e_err_ctx *err_ctx = context;
+
+ return err_ctx ? mlx5e_rx_reporter_recover_from_ctx(err_ctx) :
+ mlx5e_health_recover_channels(priv);
+}
+
static int mlx5e_rx_reporter_build_diagnose_output(struct mlx5e_rq *rq,
struct devlink_fmsg *fmsg)
{
@@ -167,9 +270,12 @@ static int mlx5e_rx_reporter_diagnose(struct devlink_health_reporter *reporter,
static const struct devlink_health_reporter_ops mlx5_rx_reporter_ops = {
.name = "rx",
+ .recover = mlx5e_rx_reporter_recover,
.diagnose = mlx5e_rx_reporter_diagnose,
};
+#define MLX5E_REPORTER_RX_GRACEFUL_PERIOD 500
+
int mlx5e_reporter_rx_create(struct mlx5e_priv *priv)
{
struct devlink *devlink = priv_to_devlink(priv->mdev);
@@ -177,7 +283,8 @@ int mlx5e_reporter_rx_create(struct mlx5e_priv *priv)
reporter = devlink_health_reporter_create(devlink,
&mlx5_rx_reporter_ops,
- 0, false, priv);
+ MLX5E_REPORTER_RX_GRACEFUL_PERIOD,
+ true, priv);
if (IS_ERR(reporter)) {
netdev_warn(priv->netdev, "Failed to create rx reporter, err = %ld\n",
PTR_ERR(reporter));
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 2c4e1cb54402..0e43ea1c27b0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -700,8 +700,7 @@ static int mlx5e_create_rq(struct mlx5e_rq *rq,
return err;
}
-static int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state,
- int next_state)
+int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state)
{
struct mlx5_core_dev *mdev = rq->mdev;
@@ -812,7 +811,7 @@ int mlx5e_wait_for_min_rx_wqes(struct mlx5e_rq *rq, int wait_time)
return -ETIMEDOUT;
}
-static void mlx5e_free_rx_descs(struct mlx5e_rq *rq)
+void mlx5e_free_rx_descs(struct mlx5e_rq *rq)
{
__be16 wqe_ix_be;
u16 wqe_ix;
@@ -891,7 +890,7 @@ int mlx5e_open_rq(struct mlx5e_channel *c, struct mlx5e_params *params,
return err;
}
-static void mlx5e_activate_rq(struct mlx5e_rq *rq)
+void mlx5e_activate_rq(struct mlx5e_rq *rq)
{
set_bit(MLX5E_RQ_STATE_ENABLED, &rq->state);
mlx5e_trigger_irq(&rq->channel->icosq);
@@ -906,6 +905,7 @@ void mlx5e_deactivate_rq(struct mlx5e_rq *rq)
void mlx5e_close_rq(struct mlx5e_rq *rq)
{
cancel_work_sync(&rq->dim.work);
+ cancel_work_sync(&rq->channel->icosq.recover_work);
mlx5e_destroy_rq(rq);
mlx5e_free_rx_descs(rq);
mlx5e_free_rq(rq);
@@ -1022,6 +1022,14 @@ static int mlx5e_alloc_icosq_db(struct mlx5e_icosq *sq, int numa)
return 0;
}
+static void mlx5e_icosq_err_cqe_work(struct work_struct *recover_work)
+{
+ struct mlx5e_icosq *sq = container_of(recover_work, struct mlx5e_icosq,
+ recover_work);
+
+ mlx5e_reporter_icosq_cqe_err(sq);
+}
+
static int mlx5e_alloc_icosq(struct mlx5e_channel *c,
struct mlx5e_sq_param *param,
struct mlx5e_icosq *sq)
@@ -1044,6 +1052,8 @@ static int mlx5e_alloc_icosq(struct mlx5e_channel *c,
if (err)
goto err_sq_wq_destroy;
+ INIT_WORK(&sq->recover_work, mlx5e_icosq_err_cqe_work);
+
return 0;
err_sq_wq_destroy:
@@ -1388,12 +1398,12 @@ int mlx5e_open_icosq(struct mlx5e_channel *c, struct mlx5e_params *params,
return err;
}
-static void mlx5e_activate_icosq(struct mlx5e_icosq *icosq)
+void mlx5e_activate_icosq(struct mlx5e_icosq *icosq)
{
set_bit(MLX5E_SQ_STATE_ENABLED, &icosq->state);
}
-static void mlx5e_deactivate_icosq(struct mlx5e_icosq *icosq)
+void mlx5e_deactivate_icosq(struct mlx5e_icosq *icosq)
{
struct mlx5e_channel *c = icosq->channel;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 60570b442fff..0f6033ea475d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -615,6 +615,8 @@ void mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
if (unlikely(get_cqe_opcode(cqe) != MLX5_CQE_REQ)) {
netdev_WARN_ONCE(cq->channel->netdev,
"Bad OP in ICOSQ CQE: 0x%x\n", get_cqe_opcode(cqe));
+ if (!test_and_set_bit(MLX5E_SQ_STATE_RECOVERING, &sq->state))
+ queue_work(cq->channel->priv->wq, &sq->recover_work);
break;
}
do {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
index 94a32c76c182..18e4c162256a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
@@ -109,6 +109,7 @@ static const struct counter_desc sw_stats_desc[] = {
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_cache_waive) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_congst_umr) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_arfs_err) },
+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_recover) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, ch_events) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, ch_poll) },
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, ch_arm) },
@@ -220,6 +221,7 @@ static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv)
s->rx_cache_waive += rq_stats->cache_waive;
s->rx_congst_umr += rq_stats->congst_umr;
s->rx_arfs_err += rq_stats->arfs_err;
+ s->rx_recover += rq_stats->recover;
s->ch_events += ch_stats->events;
s->ch_poll += ch_stats->poll;
s->ch_arm += ch_stats->arm;
@@ -1298,6 +1300,7 @@ static const struct counter_desc rq_stats_desc[] = {
{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, cache_waive) },
{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, congst_umr) },
{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, arfs_err) },
+ { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, recover) },
};
static const struct counter_desc sq_stats_desc[] = {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index bf645d42c833..c281e567711d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -116,6 +116,7 @@ struct mlx5e_sw_stats {
u64 rx_cache_waive;
u64 rx_congst_umr;
u64 rx_arfs_err;
+ u64 rx_recover;
u64 ch_events;
u64 ch_poll;
u64 ch_arm;
@@ -249,6 +250,7 @@ struct mlx5e_rq_stats {
u64 cache_waive;
u64 congst_umr;
u64 arfs_err;
+ u64 recover;
};
struct mlx5e_sq_stats {
--
2.21.0
^ permalink raw reply related
* Re: net: micrel: confusion about phyids used in driver
From: Uwe Kleine-König @ 2019-08-20 20:25 UTC (permalink / raw)
To: Nicolas Ferre
Cc: netdev, andrew, f.fainelli, kernel, hkallweit1, Ravi.Hegde,
Tristram.Ha, Yuiko.Oshino
In-Reply-To: <20190808083637.g77loqpgkzi63u55@pengutronix.de>
Hello Nicolas,
there are some open questions regarding details about some PHYs
supported in the drivers/net/phy/micrel.c driver.
On Thu, Aug 08, 2019 at 10:36:37AM +0200, Uwe Kleine-König wrote:
> On Tue, Jul 02, 2019 at 08:55:07PM +0000, Yuiko.Oshino@microchip.com wrote:
> > >On Fri, May 10, 2019 at 09:22:43AM +0200, Uwe Kleine-König wrote:
> > >> On Thu, May 09, 2019 at 11:07:45PM +0200, Andrew Lunn wrote:
> > >> > On Thu, May 09, 2019 at 10:55:29PM +0200, Heiner Kallweit wrote:
> > >> > > On 09.05.2019 22:29, Uwe Kleine-König wrote:
> > >> > > > I have a board here that has a KSZ8051MLL (datasheet:
> > >> > > > http://ww1.microchip.com/downloads/en/DeviceDoc/ksz8051mll.pdf, phyid:
> > >> > > > 0x0022155x) assembled. The actual phyid is 0x00221556.
The short version is that a phy with ID 0x00221556 matches two
phy_driver entries in the driver:
{ .phy_id = PHY_ID_KSZ8031, .phy_id_mask = 0x00ffffff, ... },
{ .phy_id = PHY_ID_KSZ8051, .phy_id_mask = MICREL_PHY_ID_MASK, ... }
The driver doesn't behave optimal for "my" KSZ8051MLL with both entries
... It seems to work, but not all features of the phy are used and the
bootlog claims this was a KSZ8031 because that's the first match in the
list.
So we're in need of someone who can get their hands on some more
detailed documentation than publicly available to allow to make the
driver handle the KSZ8051MLL correctly without breaking other stuff.
I assume you are in a different department of Microchip than the people
caring for PHYs, but maybe you can still help finding someone who cares?
> > >> > > I think the datasheets are the source of the confusion. If the
> > >> > > datasheets for different chips list 0x0022155x as PHYID each, and
> > >> > > authors of support for additional chips don't check the existing
> > >> > > code, then happens what happened.
> > >> > >
> > >> > > However it's not a rare exception and not Microchip-specific that
> > >> > > sometimes vendors use the same PHYID for different chips.
> > >>
> > >> From the vendor's POV it is even sensible to reuse the phy IDs iff the
> > >> chips are "compatible".
> > >>
> > >> Assuming that the last nibble of the phy ID actually helps to
> > >> distinguish the different (not completely) compatible chips, we need
> > >> some more detailed information than available in the data sheets I have.
> > >> There is one person in the recipents of this mail with an
> > >> @microchip.com address (hint, hint!).
> > >
> > >can you give some input here or forward to a person who can?
> >
> > I forward this to the team.
>
> This thread still sits in my inbox waiting for some feedback. Did
> something happen on your side?
This is still true, didn't hear back from Yuiko Oshino for some time
now.
Best regards
Uwe
--
Pengutronix e.K. | Uwe Kleine-König |
Industrial Linux Solutions | http://www.pengutronix.de/ |
^ permalink raw reply
* Re: regression in ath10k dma allocation
From: Tobias Klausmann @ 2019-08-20 20:24 UTC (permalink / raw)
To: Christoph Hellwig, Hillf Danton
Cc: Nicolin Chen, kvalo, davem, ath10k, linux-wireless, netdev,
linux-kernel, m.szyprowski, robin.murphy, iommu, tobias.klausmann
In-Reply-To: <20190820071250.GA28968@lst.de>
On 20.08.19 09:12, Christoph Hellwig wrote:
> On Tue, Aug 20, 2019 at 02:58:33PM +0800, Hillf Danton wrote:
>> On Tue, 20 Aug 2019 05:05:14 +0200 Christoph Hellwig wrote:
>>> Tobias, plase try this patch:
>>>
> New version below:
>
> ---
> From b8a805e93be5a5662323b8ac61fe686df839c4ac Mon Sep 17 00:00:00 2001
> From: Christoph Hellwig <hch@lst.de>
> Date: Tue, 20 Aug 2019 11:45:49 +0900
> Subject: dma-direct: fix zone selection after an unaddressable CMA allocation
>
> The new dma_alloc_contiguous hides if we allocate CMA or regular
> pages, and thus fails to retry a ZONE_NORMAL allocation if the CMA
> allocation succeeds but isn't addressable. That means we either fail
> outright or dip into a small zone that might not succeed either.
>
> Thanks to Hillf Danton for debugging this issue.
>
> Fixes: b1d2dc009dec ("dma-contiguous: add dma_{alloc,free}_contiguous() helpers")
> Reported-by: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> drivers/iommu/dma-iommu.c | 3 +++
> include/linux/dma-contiguous.h | 5 +----
> kernel/dma/contiguous.c | 9 +++------
> kernel/dma/direct.c | 10 +++++++++-
> 4 files changed, 16 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
> index d991d40f797f..f68a62c3c32b 100644
> --- a/drivers/iommu/dma-iommu.c
> +++ b/drivers/iommu/dma-iommu.c
> @@ -965,10 +965,13 @@ static void *iommu_dma_alloc_pages(struct device *dev, size_t size,
> {
> bool coherent = dev_is_dma_coherent(dev);
> size_t alloc_size = PAGE_ALIGN(size);
> + int node = dev_to_node(dev);
> struct page *page = NULL;
> void *cpu_addr;
>
> page = dma_alloc_contiguous(dev, alloc_size, gfp);
> + if (!page)
> + page = alloc_pages_node(node, gfp, get_order(alloc_size));
> if (!page)
> return NULL;
>
> diff --git a/include/linux/dma-contiguous.h b/include/linux/dma-contiguous.h
> index c05d4e661489..03f8e98e3bcc 100644
> --- a/include/linux/dma-contiguous.h
> +++ b/include/linux/dma-contiguous.h
> @@ -160,10 +160,7 @@ bool dma_release_from_contiguous(struct device *dev, struct page *pages,
> static inline struct page *dma_alloc_contiguous(struct device *dev, size_t size,
> gfp_t gfp)
> {
> - int node = dev ? dev_to_node(dev) : NUMA_NO_NODE;
> - size_t align = get_order(PAGE_ALIGN(size));
> -
> - return alloc_pages_node(node, gfp, align);
> + return NULL;
> }
>
> static inline void dma_free_contiguous(struct device *dev, struct page *page,
> diff --git a/kernel/dma/contiguous.c b/kernel/dma/contiguous.c
> index 2bd410f934b3..e6b450fdbeb6 100644
> --- a/kernel/dma/contiguous.c
> +++ b/kernel/dma/contiguous.c
> @@ -230,9 +230,7 @@ bool dma_release_from_contiguous(struct device *dev, struct page *pages,
> */
> struct page *dma_alloc_contiguous(struct device *dev, size_t size, gfp_t gfp)
> {
> - int node = dev ? dev_to_node(dev) : NUMA_NO_NODE;
> - size_t count = PAGE_ALIGN(size) >> PAGE_SHIFT;
> - size_t align = get_order(PAGE_ALIGN(size));
> + size_t count = size >> PAGE_SHIFT;
> struct page *page = NULL;
> struct cma *cma = NULL;
>
> @@ -243,14 +241,12 @@ struct page *dma_alloc_contiguous(struct device *dev, size_t size, gfp_t gfp)
>
> /* CMA can be used only in the context which permits sleeping */
> if (cma && gfpflags_allow_blocking(gfp)) {
> + size_t align = get_order(size);
> size_t cma_align = min_t(size_t, align, CONFIG_CMA_ALIGNMENT);
>
> page = cma_alloc(cma, count, cma_align, gfp & __GFP_NOWARN);
> }
>
> - /* Fallback allocation of normal pages */
> - if (!page)
> - page = alloc_pages_node(node, gfp, align);
> return page;
> }
>
> @@ -258,6 +254,7 @@ struct page *dma_alloc_contiguous(struct device *dev, size_t size, gfp_t gfp)
> * dma_free_contiguous() - release allocated pages
> * @dev: Pointer to device for which the pages were allocated.
> * @page: Pointer to the allocated pages.
> + int node = dev ? dev_to_node(dev) : NUMA_NO_NODE;
> * @size: Size of allocated pages.
> *
> * This function releases memory allocated by dma_alloc_contiguous(). As the
> diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c
> index 795c9b095d75..706113c6bebc 100644
> --- a/kernel/dma/direct.c
> +++ b/kernel/dma/direct.c
> @@ -85,6 +85,8 @@ static bool dma_coherent_ok(struct device *dev, phys_addr_t phys, size_t size)
> struct page *__dma_direct_alloc_pages(struct device *dev, size_t size,
> dma_addr_t *dma_handle, gfp_t gfp, unsigned long attrs)
> {
> + size_t alloc_size = PAGE_ALIGN(size);
> + int node = dev_to_node(dev);
> struct page *page = NULL;
> u64 phys_mask;
>
> @@ -95,8 +97,14 @@ struct page *__dma_direct_alloc_pages(struct device *dev, size_t size,
> gfp &= ~__GFP_ZERO;
> gfp |= __dma_direct_optimal_gfp_mask(dev, dev->coherent_dma_mask,
> &phys_mask);
> + page = dma_alloc_contiguous(dev, alloc_size, gfp);
> + if (page && !dma_coherent_ok(dev, page_to_phys(page), size)) {
> + dma_free_contiguous(dev, page, alloc_size);
> + page = NULL;
> + }
> again:
> - page = dma_alloc_contiguous(dev, size, gfp);
> + if (!page)
> + page = alloc_pages_node(node, gfp, get_order(alloc_size));
> if (page && !dma_coherent_ok(dev, page_to_phys(page), size)) {
> dma_free_contiguous(dev, page, size);
> page = NULL;
I can confirm this resolves the regression!
Tested-by: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de>
Thanks for the work of all involved,
Tobias
^ permalink raw reply
* [net-next v2 05/16] net/mlx5e: Extend tx reporter diagnostics output
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Enhance tx reporter's diagnostics output to include: information common
to all SQs: SQ size, SQ stride size.
In addition add channel ix, tc, txq ix, cc and pc.
$ devlink health diagnose pci/0000:00:0b.0 reporter tx
Common config:
SQ:
stride size: 64 size: 1024
SQs:
channel ix: 0 tc: 0 txq ix: 0 sqn: 4307 HW state: 1 stopped: false cc: 0 pc: 0
channel ix: 1 tc: 0 txq ix: 1 sqn: 4312 HW state: 1 stopped: false cc: 0 pc: 0
channel ix: 2 tc: 0 txq ix: 2 sqn: 4317 HW state: 1 stopped: false cc: 0 pc: 0
channel ix: 3 tc: 0 txq ix: 3 sqn: 4322 HW state: 1 stopped: false cc: 0 pc: 0
$ devlink health diagnose pci/0000:00:0b.0 reporter tx -jp
{
"Common config": {
"SQ": {
"stride size": 64,
"size": 1024
}
},
"SQs": [ {
"channel ix": 0,
"tc": 0,
"txq ix": 0,
"sqn": 4307,
"HW state": 1,
"stopped": false,
"cc": 0,
"pc": 0
},{
"channel ix": 1,
"tc": 0,
"txq ix": 1,
"sqn": 4312,
"HW state": 1,
"stopped": false,
"cc": 0,
"pc": 0
},{
"channel ix": 2,
"tc": 0,
"txq ix": 2,
"sqn": 4317,
"HW state": 1,
"stopped": false,
"cc": 0,
"pc": 0
},{
"channel ix": 3,
"tc": 0,
"txq ix": 3,
"sqn": 4322,
"HW state": 1,
"stopped": false,
"cc": 0,
"pc": 0
} ]
}
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../ethernet/mellanox/mlx5/core/en/health.c | 30 ++++++++
.../ethernet/mellanox/mlx5/core/en/health.h | 3 +
.../mellanox/mlx5/core/en/reporter_tx.c | 69 ++++++++++++++++---
3 files changed, 94 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
index fc3112921bd3..dab563f07157 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -4,6 +4,36 @@
#include "health.h"
#include "lib/eq.h"
+int mlx5e_reporter_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name)
+{
+ int err;
+
+ err = devlink_fmsg_pair_nest_start(fmsg, name);
+ if (err)
+ return err;
+
+ err = devlink_fmsg_obj_nest_start(fmsg);
+ if (err)
+ return err;
+
+ return 0;
+}
+
+int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg *fmsg)
+{
+ int err;
+
+ err = devlink_fmsg_obj_nest_end(fmsg);
+ if (err)
+ return err;
+
+ err = devlink_fmsg_pair_nest_end(fmsg);
+ if (err)
+ return err;
+
+ return 0;
+}
+
int mlx5e_health_sq_to_ready(struct mlx5e_channel *channel, u32 sqn)
{
struct mlx5_core_dev *mdev = channel->mdev;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 386bda6104aa..112771ad516c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -11,6 +11,9 @@ 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_named_obj_nest_start(struct devlink_fmsg *fmsg, char *name);
+int mlx5e_reporter_named_obj_nest_end(struct devlink_fmsg *fmsg);
+
#define MLX5E_REPORTER_PER_Q_MAX_LEN 256
struct mlx5e_err_ctx {
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 b9429ff8d9c4..a5d0fcbb85af 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -146,7 +146,7 @@ static int mlx5e_tx_reporter_recover(struct devlink_health_reporter *reporter,
static int
mlx5e_tx_reporter_build_diagnose_output(struct devlink_fmsg *fmsg,
- struct mlx5e_txqsq *sq)
+ struct mlx5e_txqsq *sq, int tc)
{
struct mlx5e_priv *priv = sq->channel->priv;
bool stopped = netif_xmit_stopped(sq->txq);
@@ -161,6 +161,18 @@ mlx5e_tx_reporter_build_diagnose_output(struct devlink_fmsg *fmsg,
if (err)
return err;
+ err = devlink_fmsg_u32_pair_put(fmsg, "channel ix", sq->ch_ix);
+ if (err)
+ return err;
+
+ err = devlink_fmsg_u32_pair_put(fmsg, "tc", tc);
+ if (err)
+ return err;
+
+ err = devlink_fmsg_u32_pair_put(fmsg, "txq ix", sq->txq_ix);
+ if (err)
+ return err;
+
err = devlink_fmsg_u32_pair_put(fmsg, "sqn", sq->sqn);
if (err)
return err;
@@ -173,6 +185,14 @@ mlx5e_tx_reporter_build_diagnose_output(struct devlink_fmsg *fmsg,
if (err)
return err;
+ err = devlink_fmsg_u32_pair_put(fmsg, "cc", sq->cc);
+ if (err)
+ return err;
+
+ err = devlink_fmsg_u32_pair_put(fmsg, "pc", sq->pc);
+ if (err)
+ return err;
+
err = devlink_fmsg_obj_nest_end(fmsg);
if (err)
return err;
@@ -184,24 +204,57 @@ static int mlx5e_tx_reporter_diagnose(struct devlink_health_reporter *reporter,
struct devlink_fmsg *fmsg)
{
struct mlx5e_priv *priv = devlink_health_reporter_priv(reporter);
- int i, err = 0;
+ struct mlx5e_txqsq *generic_sq = priv->txq2sq[0];
+ u32 sq_stride, sq_sz;
+
+ int i, tc, err = 0;
mutex_lock(&priv->state_lock);
if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
goto unlock;
+ 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");
+ if (err)
+ goto unlock;
+
+ err = mlx5e_reporter_named_obj_nest_start(fmsg, "SQ");
+ if (err)
+ goto unlock;
+
+ err = devlink_fmsg_u64_pair_put(fmsg, "stride size", sq_stride);
+ if (err)
+ goto unlock;
+
+ err = devlink_fmsg_u32_pair_put(fmsg, "size", sq_sz);
+ if (err)
+ goto unlock;
+
+ err = mlx5e_reporter_named_obj_nest_end(fmsg);
+ if (err)
+ goto unlock;
+
+ err = mlx5e_reporter_named_obj_nest_end(fmsg);
+ if (err)
+ goto unlock;
+
err = devlink_fmsg_arr_pair_nest_start(fmsg, "SQs");
if (err)
goto unlock;
- for (i = 0; i < priv->channels.num * priv->channels.params.num_tc;
- i++) {
- struct mlx5e_txqsq *sq = priv->txq2sq[i];
+ for (i = 0; i < priv->channels.num; i++) {
+ struct mlx5e_channel *c = priv->channels.c[i];
+
+ for (tc = 0; tc < priv->channels.params.num_tc; tc++) {
+ struct mlx5e_txqsq *sq = &c->sq[tc];
- err = mlx5e_tx_reporter_build_diagnose_output(fmsg, sq);
- if (err)
- goto unlock;
+ err = mlx5e_tx_reporter_build_diagnose_output(fmsg, sq, tc);
+ if (err)
+ goto unlock;
+ }
}
err = devlink_fmsg_arr_pair_nest_end(fmsg);
if (err)
--
2.21.0
^ permalink raw reply related
* [net-next v2 07/16] net/mlx5e: Add helper functions for reporter's basics
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Introduce helper functions for create and destroy reporters and update
channels. In the following patch, rx reporter is added and it will use
these helpers too.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en/health.c | 17 +++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/en/health.h | 4 ++++
.../net/ethernet/mellanox/mlx5/core/en_main.c | 9 +++------
3 files changed, 24 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
index ffd9a7a165a2..c11d0162eaf8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -96,6 +96,23 @@ int mlx5e_reporter_cq_common_diagnose(struct mlx5e_cq *cq, struct devlink_fmsg *
return 0;
}
+int mlx5e_health_create_reporters(struct mlx5e_priv *priv)
+{
+ return mlx5e_reporter_tx_create(priv);
+}
+
+void mlx5e_health_destroy_reporters(struct mlx5e_priv *priv)
+{
+ mlx5e_reporter_tx_destroy(priv);
+}
+
+void mlx5e_health_channels_update(struct mlx5e_priv *priv)
+{
+ if (priv->tx_reporter)
+ devlink_health_reporter_state_update(priv->tx_reporter,
+ DEVLINK_HEALTH_REPORTER_STATE_HEALTHY);
+}
+
int mlx5e_health_sq_to_ready(struct mlx5e_channel *channel, u32 sqn)
{
struct mlx5_core_dev *mdev = channel->mdev;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index 6725d417aaf5..b2c0ccc79b22 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -29,5 +29,9 @@ 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_destroy_reporters(struct mlx5e_priv *priv);
+void mlx5e_health_channels_update(struct mlx5e_priv *priv);
+
#endif
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 695c75a8d1ab..f3bcb9ddca58 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -2323,10 +2323,7 @@ int mlx5e_open_channels(struct mlx5e_priv *priv,
goto err_close_channels;
}
- if (priv->tx_reporter)
- devlink_health_reporter_state_update(priv->tx_reporter,
- DEVLINK_HEALTH_REPORTER_STATE_HEALTHY);
-
+ mlx5e_health_channels_update(priv);
kvfree(cparam);
return 0;
@@ -3201,7 +3198,6 @@ static void mlx5e_cleanup_nic_tx(struct mlx5e_priv *priv)
{
int tc;
- mlx5e_reporter_tx_destroy(priv);
for (tc = 0; tc < priv->profile->max_tc; tc++)
mlx5e_destroy_tis(priv->mdev, priv->tisn[tc]);
}
@@ -4969,12 +4965,14 @@ static int mlx5e_nic_init(struct mlx5_core_dev *mdev,
mlx5_core_err(mdev, "TLS initialization failed, %d\n", err);
mlx5e_build_nic_netdev(netdev);
mlx5e_build_tc2txq_maps(priv);
+ mlx5e_health_create_reporters(priv);
return 0;
}
static void mlx5e_nic_cleanup(struct mlx5e_priv *priv)
{
+ mlx5e_health_destroy_reporters(priv);
mlx5e_tls_cleanup(priv);
mlx5e_ipsec_cleanup(priv);
mlx5e_netdev_cleanup(priv->netdev, priv);
@@ -5077,7 +5075,6 @@ static int mlx5e_init_nic_tx(struct mlx5e_priv *priv)
#ifdef CONFIG_MLX5_CORE_EN_DCB
mlx5e_dcbnl_initialize(priv);
#endif
- mlx5e_reporter_tx_create(priv);
return 0;
}
--
2.21.0
^ permalink raw reply related
* [net-next v2 04/16] net/mlx5e: Extend tx diagnose function
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
The following patches in the set enhance the diagnostics info of tx
reporter. Therefore, it is better to pass a pointer to the SQ for
further data extraction.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../mellanox/mlx5/core/en/reporter_tx.c | 20 +++++++++----------
1 file changed, 10 insertions(+), 10 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 6f9f42ab3005..b9429ff8d9c4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -146,15 +146,22 @@ static int mlx5e_tx_reporter_recover(struct devlink_health_reporter *reporter,
static int
mlx5e_tx_reporter_build_diagnose_output(struct devlink_fmsg *fmsg,
- u32 sqn, u8 state, bool stopped)
+ struct mlx5e_txqsq *sq)
{
+ struct mlx5e_priv *priv = sq->channel->priv;
+ bool stopped = netif_xmit_stopped(sq->txq);
+ u8 state;
int err;
+ err = mlx5_core_query_sq_state(priv->mdev, sq->sqn, &state);
+ if (err)
+ return err;
+
err = devlink_fmsg_obj_nest_start(fmsg);
if (err)
return err;
- err = devlink_fmsg_u32_pair_put(fmsg, "sqn", sqn);
+ err = devlink_fmsg_u32_pair_put(fmsg, "sqn", sq->sqn);
if (err)
return err;
@@ -191,15 +198,8 @@ static int mlx5e_tx_reporter_diagnose(struct devlink_health_reporter *reporter,
for (i = 0; i < priv->channels.num * priv->channels.params.num_tc;
i++) {
struct mlx5e_txqsq *sq = priv->txq2sq[i];
- u8 state;
-
- err = mlx5_core_query_sq_state(priv->mdev, sq->sqn, &state);
- if (err)
- goto unlock;
- err = mlx5e_tx_reporter_build_diagnose_output(fmsg, sq->sqn,
- state,
- netif_xmit_stopped(sq->txq));
+ err = mlx5e_tx_reporter_build_diagnose_output(fmsg, sq);
if (err)
goto unlock;
}
--
2.21.0
^ permalink raw reply related
* [net-next v2 03/16] net/mlx5e: Generalize tx reporter's functionality
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Prepare for code sharing with rx reporter, which is added in the
following patches in the set. Introduce a generic error_ctx for
agnostic recovery despatch.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/Makefile | 5 +-
.../ethernet/mellanox/mlx5/core/en/health.c | 82 ++++++++++
.../ethernet/mellanox/mlx5/core/en/health.h | 14 ++
.../mellanox/mlx5/core/en/reporter_tx.c | 140 +++++-------------
4 files changed, 137 insertions(+), 104 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en/health.c
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
index 57d2cc666fe3..23d566a45a30 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
@@ -23,8 +23,9 @@ mlx5_core-y := main.o cmd.o debugfs.o fw.o eq.o uar.o pagealloc.o \
#
mlx5_core-$(CONFIG_MLX5_CORE_EN) += en_main.o en_common.o en_fs.o en_ethtool.o \
en_tx.o en_rx.o en_dim.o en_txrx.o en/xdp.o en_stats.o \
- en_selftest.o en/port.o en/monitor_stats.o en/reporter_tx.o \
- en/params.o en/xsk/umem.o en/xsk/setup.o en/xsk/rx.o en/xsk/tx.o
+ en_selftest.o en/port.o en/monitor_stats.o en/health.o \
+ en/reporter_tx.o en/params.o en/xsk/umem.o en/xsk/setup.o \
+ en/xsk/rx.o en/xsk/tx.o
#
# Netdev extra
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
new file mode 100644
index 000000000000..fc3112921bd3
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c
@@ -0,0 +1,82 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (c) 2019 Mellanox Technologies.
+
+#include "health.h"
+#include "lib/eq.h"
+
+int mlx5e_health_sq_to_ready(struct mlx5e_channel *channel, u32 sqn)
+{
+ struct mlx5_core_dev *mdev = channel->mdev;
+ struct net_device *dev = channel->netdev;
+ struct mlx5e_modify_sq_param msp = {};
+ int err;
+
+ msp.curr_state = MLX5_SQC_STATE_ERR;
+ msp.next_state = MLX5_SQC_STATE_RST;
+
+ err = mlx5e_modify_sq(mdev, sqn, &msp);
+ if (err) {
+ netdev_err(dev, "Failed to move sq 0x%x to reset\n", sqn);
+ return err;
+ }
+
+ memset(&msp, 0, sizeof(msp));
+ msp.curr_state = MLX5_SQC_STATE_RST;
+ msp.next_state = MLX5_SQC_STATE_RDY;
+
+ err = mlx5e_modify_sq(mdev, sqn, &msp);
+ if (err) {
+ netdev_err(dev, "Failed to move sq 0x%x to ready\n", sqn);
+ return err;
+ }
+
+ return 0;
+}
+
+int mlx5e_health_recover_channels(struct mlx5e_priv *priv)
+{
+ int err = 0;
+
+ rtnl_lock();
+ mutex_lock(&priv->state_lock);
+
+ if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
+ goto out;
+
+ err = mlx5e_safe_reopen_channels(priv);
+
+out:
+ mutex_unlock(&priv->state_lock);
+ rtnl_unlock();
+
+ return err;
+}
+
+int mlx5e_health_channel_eq_recover(struct mlx5_eq_comp *eq, struct mlx5e_channel *channel)
+{
+ u32 eqe_count;
+
+ netdev_err(channel->netdev, "EQ 0x%x: Cons = 0x%x, irqn = 0x%x\n",
+ eq->core.eqn, eq->core.cons_index, eq->core.irqn);
+
+ eqe_count = mlx5_eq_poll_irq_disabled(eq);
+ if (!eqe_count)
+ return -EIO;
+
+ netdev_err(channel->netdev, "Recovered %d eqes on EQ 0x%x\n",
+ eqe_count, eq->core.eqn);
+
+ channel->stats->eq_rearm++;
+ return 0;
+}
+
+int mlx5e_health_report(struct mlx5e_priv *priv,
+ struct devlink_health_reporter *reporter, char *err_str,
+ struct mlx5e_err_ctx *err_ctx)
+{
+ if (!reporter) {
+ netdev_err(priv->netdev, err_str);
+ return err_ctx->recover(&err_ctx->ctx);
+ }
+ return devlink_health_report(reporter, err_str, err_ctx);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index c7a5a149011e..386bda6104aa 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -11,4 +11,18 @@ 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);
+#define MLX5E_REPORTER_PER_Q_MAX_LEN 256
+
+struct mlx5e_err_ctx {
+ int (*recover)(void *ctx);
+ void *ctx;
+};
+
+int mlx5e_health_sq_to_ready(struct mlx5e_channel *channel, u32 sqn);
+int mlx5e_health_channel_eq_recover(struct mlx5_eq_comp *eq, struct mlx5e_channel *channel);
+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);
+
#endif
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 62b95f62e4dc..6f9f42ab3005 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -2,14 +2,6 @@
/* Copyright (c) 2019 Mellanox Technologies. */
#include "health.h"
-#include "lib/eq.h"
-
-#define MLX5E_TX_REPORTER_PER_SQ_MAX_LEN 256
-
-struct mlx5e_tx_err_ctx {
- int (*recover)(struct mlx5e_txqsq *sq);
- struct mlx5e_txqsq *sq;
-};
static int mlx5e_wait_for_sq_flush(struct mlx5e_txqsq *sq)
{
@@ -39,41 +31,20 @@ static void mlx5e_reset_txqsq_cc_pc(struct mlx5e_txqsq *sq)
sq->pc = 0;
}
-static int mlx5e_sq_to_ready(struct mlx5e_txqsq *sq, int curr_state)
+static int mlx5e_tx_reporter_err_cqe_recover(void *ctx)
{
- struct mlx5_core_dev *mdev = sq->channel->mdev;
- struct net_device *dev = sq->channel->netdev;
- struct mlx5e_modify_sq_param msp = {0};
+ struct mlx5_core_dev *mdev;
+ struct net_device *dev;
+ struct mlx5e_txqsq *sq;
+ u8 state;
int err;
- msp.curr_state = curr_state;
- msp.next_state = MLX5_SQC_STATE_RST;
-
- err = mlx5e_modify_sq(mdev, sq->sqn, &msp);
- if (err) {
- netdev_err(dev, "Failed to move sq 0x%x to reset\n", sq->sqn);
- return err;
- }
-
- memset(&msp, 0, sizeof(msp));
- msp.curr_state = MLX5_SQC_STATE_RST;
- msp.next_state = MLX5_SQC_STATE_RDY;
-
- err = mlx5e_modify_sq(mdev, sq->sqn, &msp);
- if (err) {
- netdev_err(dev, "Failed to move sq 0x%x to ready\n", sq->sqn);
- return err;
- }
-
- return 0;
-}
+ sq = ctx;
+ mdev = sq->channel->mdev;
+ dev = sq->channel->netdev;
-static int mlx5e_tx_reporter_err_cqe_recover(struct mlx5e_txqsq *sq)
-{
- struct mlx5_core_dev *mdev = sq->channel->mdev;
- struct net_device *dev = sq->channel->netdev;
- u8 state;
- int err;
+ if (!test_bit(MLX5E_SQ_STATE_RECOVERING, &sq->state))
+ return 0;
err = mlx5_core_query_sq_state(mdev, sq->sqn, &state);
if (err) {
@@ -96,7 +67,7 @@ static int mlx5e_tx_reporter_err_cqe_recover(struct mlx5e_txqsq *sq)
* pending WQEs. SQ can safely reset the SQ.
*/
- err = mlx5e_sq_to_ready(sq, state);
+ err = mlx5e_health_sq_to_ready(sq->channel, sq->sqn);
if (err)
goto out;
@@ -111,102 +82,66 @@ static int mlx5e_tx_reporter_err_cqe_recover(struct mlx5e_txqsq *sq)
return err;
}
-static int mlx5_tx_health_report(struct devlink_health_reporter *tx_reporter,
- char *err_str,
- struct mlx5e_tx_err_ctx *err_ctx)
-{
- if (!tx_reporter) {
- netdev_err(err_ctx->sq->channel->netdev, err_str);
- return err_ctx->recover(err_ctx->sq);
- }
-
- return devlink_health_report(tx_reporter, err_str, err_ctx);
-}
-
void mlx5e_reporter_tx_err_cqe(struct mlx5e_txqsq *sq)
{
- char err_str[MLX5E_TX_REPORTER_PER_SQ_MAX_LEN];
- struct mlx5e_tx_err_ctx err_ctx = {0};
+ struct mlx5e_priv *priv = sq->channel->priv;
+ char err_str[MLX5E_REPORTER_PER_Q_MAX_LEN];
+ struct mlx5e_err_ctx err_ctx = {0};
- err_ctx.sq = sq;
- err_ctx.recover = mlx5e_tx_reporter_err_cqe_recover;
+ err_ctx.ctx = sq;
+ err_ctx.recover = mlx5e_tx_reporter_err_cqe_recover;
sprintf(err_str, "ERR CQE on SQ: 0x%x", sq->sqn);
- mlx5_tx_health_report(sq->channel->priv->tx_reporter, err_str,
- &err_ctx);
+ mlx5e_health_report(priv, priv->tx_reporter, err_str, &err_ctx);
}
-static int mlx5e_tx_reporter_timeout_recover(struct mlx5e_txqsq *sq)
+static int mlx5e_tx_reporter_timeout_recover(void *ctx)
{
- struct mlx5_eq_comp *eq = sq->cq.mcq.eq;
- u32 eqe_count;
-
- netdev_err(sq->channel->netdev, "EQ 0x%x: Cons = 0x%x, irqn = 0x%x\n",
- eq->core.eqn, eq->core.cons_index, eq->core.irqn);
+ struct mlx5_eq_comp *eq;
+ struct mlx5e_txqsq *sq;
+ int err;
- eqe_count = mlx5_eq_poll_irq_disabled(eq);
- if (!eqe_count) {
+ sq = ctx;
+ eq = sq->cq.mcq.eq;
+ err = mlx5e_health_channel_eq_recover(eq, sq->channel);
+ if (err)
clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state);
- return -EIO;
- }
- netdev_err(sq->channel->netdev, "Recover %d eqes on EQ 0x%x\n",
- eqe_count, eq->core.eqn);
- sq->channel->stats->eq_rearm++;
- return 0;
+ return err;
}
int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq)
{
- char err_str[MLX5E_TX_REPORTER_PER_SQ_MAX_LEN];
- struct mlx5e_tx_err_ctx err_ctx;
+ struct mlx5e_priv *priv = sq->channel->priv;
+ char err_str[MLX5E_REPORTER_PER_Q_MAX_LEN];
+ struct mlx5e_err_ctx err_ctx;
- err_ctx.sq = sq;
- err_ctx.recover = mlx5e_tx_reporter_timeout_recover;
+ err_ctx.ctx = sq;
+ err_ctx.recover = mlx5e_tx_reporter_timeout_recover;
sprintf(err_str,
"TX timeout on queue: %d, SQ: 0x%x, CQ: 0x%x, SQ Cons: 0x%x SQ Prod: 0x%x, usecs since last trans: %u\n",
sq->channel->ix, sq->sqn, sq->cq.mcq.cqn, sq->cc, sq->pc,
jiffies_to_usecs(jiffies - sq->txq->trans_start));
- return mlx5_tx_health_report(sq->channel->priv->tx_reporter, err_str,
- &err_ctx);
+ return mlx5e_health_report(priv, priv->tx_reporter, err_str, &err_ctx);
}
/* state lock cannot be grabbed within this function.
* It can cause a dead lock or a read-after-free.
*/
-static int mlx5e_tx_reporter_recover_from_ctx(struct mlx5e_tx_err_ctx *err_ctx)
-{
- return err_ctx->recover(err_ctx->sq);
-}
-
-static int mlx5e_tx_reporter_recover_all(struct mlx5e_priv *priv)
+static int mlx5e_tx_reporter_recover_from_ctx(struct mlx5e_err_ctx *err_ctx)
{
- int err = 0;
-
- rtnl_lock();
- mutex_lock(&priv->state_lock);
-
- if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
- goto out;
-
- err = mlx5e_safe_reopen_channels(priv);
-
-out:
- mutex_unlock(&priv->state_lock);
- rtnl_unlock();
-
- return err;
+ return err_ctx->recover(err_ctx->ctx);
}
static int mlx5e_tx_reporter_recover(struct devlink_health_reporter *reporter,
void *context)
{
struct mlx5e_priv *priv = devlink_health_reporter_priv(reporter);
- struct mlx5e_tx_err_ctx *err_ctx = context;
+ struct mlx5e_err_ctx *err_ctx = context;
return err_ctx ? mlx5e_tx_reporter_recover_from_ctx(err_ctx) :
- mlx5e_tx_reporter_recover_all(priv);
+ mlx5e_health_recover_channels(priv);
}
static int
@@ -289,8 +224,9 @@ int mlx5e_reporter_tx_create(struct mlx5e_priv *priv)
{
struct devlink_health_reporter *reporter;
struct mlx5_core_dev *mdev = priv->mdev;
- struct devlink *devlink = priv_to_devlink(mdev);
+ struct devlink *devlink;
+ devlink = priv_to_devlink(mdev);
reporter =
devlink_health_reporter_create(devlink, &mlx5_tx_reporter_ops,
MLX5_REPORTER_TX_GRACEFUL_PERIOD,
--
2.21.0
^ permalink raw reply related
* [net-next v2 01/16] net/mlx5e: Rename reporter header file
From: Saeed Mahameed @ 2019-08-20 20:24 UTC (permalink / raw)
To: David S. Miller
Cc: netdev@vger.kernel.org, Aya Levin, Tariq Toukan, Jiri Pirko,
Saeed Mahameed
In-Reply-To: <20190820202352.2995-1-saeedm@mellanox.com>
From: Aya Levin <ayal@mellanox.com>
Rename reporter.h -> health.h so patches in the set can use it for
health related functionality.
Signed-off-by: Aya Levin <ayal@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../ethernet/mellanox/mlx5/core/en/{reporter.h => health.h} | 4 ++--
drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 2 +-
3 files changed, 4 insertions(+), 4 deletions(-)
rename drivers/net/ethernet/mellanox/mlx5/core/en/{reporter.h => health.h} (84%)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter.h b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
similarity index 84%
rename from drivers/net/ethernet/mellanox/mlx5/core/en/reporter.h
rename to drivers/net/ethernet/mellanox/mlx5/core/en/health.h
index ed7a3881d2c5..cee840e40a05 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.h
@@ -1,8 +1,8 @@
/* SPDX-License-Identifier: GPL-2.0 */
/* Copyright (c) 2019 Mellanox Technologies. */
-#ifndef __MLX5E_EN_REPORTER_H
-#define __MLX5E_EN_REPORTER_H
+#ifndef __MLX5E_EN_HEALTH_H
+#define __MLX5E_EN_HEALTH_H
#include "en.h"
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 817c6ea7e349..9ff19d69619f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c
@@ -1,7 +1,7 @@
/* SPDX-License-Identifier: GPL-2.0 */
/* Copyright (c) 2019 Mellanox Technologies. */
-#include "reporter.h"
+#include "health.h"
#include "lib/eq.h"
#define MLX5E_TX_REPORTER_PER_SQ_MAX_LEN 256
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 0c8e847a9eee..09a68b84cce0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -56,7 +56,7 @@
#include "en/xdp.h"
#include "lib/eq.h"
#include "en/monitor_stats.h"
-#include "en/reporter.h"
+#include "en/health.h"
#include "en/params.h"
#include "en/xsk/umem.h"
#include "en/xsk/setup.h"
--
2.21.0
^ permalink raw reply related
* Re: net: micrel: confusion about phyids used in driver
From: Heiner Kallweit @ 2019-08-20 20:30 UTC (permalink / raw)
To: Uwe Kleine-König, Nicolas Ferre
Cc: netdev, andrew, f.fainelli, kernel, Ravi.Hegde, Tristram.Ha,
Yuiko.Oshino
In-Reply-To: <20190820202503.xauhbrj24p3vcoxp@pengutronix.de>
On 20.08.2019 22:25, Uwe Kleine-König wrote:
> Hello Nicolas,
>
> there are some open questions regarding details about some PHYs
> supported in the drivers/net/phy/micrel.c driver.
>
> On Thu, Aug 08, 2019 at 10:36:37AM +0200, Uwe Kleine-König wrote:
>> On Tue, Jul 02, 2019 at 08:55:07PM +0000, Yuiko.Oshino@microchip.com wrote:
>>>> On Fri, May 10, 2019 at 09:22:43AM +0200, Uwe Kleine-König wrote:
>>>>> On Thu, May 09, 2019 at 11:07:45PM +0200, Andrew Lunn wrote:
>>>>>> On Thu, May 09, 2019 at 10:55:29PM +0200, Heiner Kallweit wrote:
>>>>>>> On 09.05.2019 22:29, Uwe Kleine-König wrote:
>>>>>>>> I have a board here that has a KSZ8051MLL (datasheet:
>>>>>>>> http://ww1.microchip.com/downloads/en/DeviceDoc/ksz8051mll.pdf, phyid:
>>>>>>>> 0x0022155x) assembled. The actual phyid is 0x00221556.
>
> The short version is that a phy with ID 0x00221556 matches two
> phy_driver entries in the driver:
>
> { .phy_id = PHY_ID_KSZ8031, .phy_id_mask = 0x00ffffff, ... },
> { .phy_id = PHY_ID_KSZ8051, .phy_id_mask = MICREL_PHY_ID_MASK, ... }
>
If two PHYs have same ID but need different drivers, then callback
match_phy_device may have to be implemented, provided that the PHYs
can be differentiated by some other register content.
See Realtek PHY driver for an example.
> The driver doesn't behave optimal for "my" KSZ8051MLL with both entries
> ... It seems to work, but not all features of the phy are used and the
> bootlog claims this was a KSZ8031 because that's the first match in the
> list.
>
> So we're in need of someone who can get their hands on some more
> detailed documentation than publicly available to allow to make the
> driver handle the KSZ8051MLL correctly without breaking other stuff.
>
> I assume you are in a different department of Microchip than the people
> caring for PHYs, but maybe you can still help finding someone who cares?
>
>>>>>>> I think the datasheets are the source of the confusion. If the
>>>>>>> datasheets for different chips list 0x0022155x as PHYID each, and
>>>>>>> authors of support for additional chips don't check the existing
>>>>>>> code, then happens what happened.
>>>>>>>
>>>>>>> However it's not a rare exception and not Microchip-specific that
>>>>>>> sometimes vendors use the same PHYID for different chips.
>>>>>
>>>>> From the vendor's POV it is even sensible to reuse the phy IDs iff the
>>>>> chips are "compatible".
>>>>>
>>>>> Assuming that the last nibble of the phy ID actually helps to
>>>>> distinguish the different (not completely) compatible chips, we need
>>>>> some more detailed information than available in the data sheets I have.
>>>>> There is one person in the recipents of this mail with an
>>>>> @microchip.com address (hint, hint!).
>>>>
>>>> can you give some input here or forward to a person who can?
>>>
>>> I forward this to the team.
>>
>> This thread still sits in my inbox waiting for some feedback. Did
>> something happen on your side?
>
> This is still true, didn't hear back from Yuiko Oshino for some time
> now.
>
> Best regards
> Uwe
>
^ permalink raw reply
* Re: [PATCH net-next 3/6] net: dsa: Delete the VID from the upstream port as well
From: Vladimir Oltean @ 2019-08-20 20:40 UTC (permalink / raw)
To: Florian Fainelli
Cc: Vivien Didelot, Andrew Lunn, Ido Schimmel, Roopa Prabhu, nikolay,
David S. Miller, netdev
In-Reply-To: <c359e0ca-c770-19da-7a3a-a3173d36a12d@gmail.com>
Hi Florian,
On Tue, 20 Aug 2019 at 22:40, Florian Fainelli <f.fainelli@gmail.com> wrote:
>
> On 8/20/19 10:52 AM, Vivien Didelot wrote:
> > Hi Vladimir,
> >
> > On Tue, 20 Aug 2019 12:54:44 +0300, Vladimir Oltean <olteanv@gmail.com> wrote:
> >> I can agree that this isn't one of my brightest moments. But at least
> >> we get to see Cunningham's law in action :)
> >> When dsa_8021q is cleaning up the switch's VLAN table for the bridge
> >> to use it, it is good to really clean it up, i.e. not leave any VLAN
> >> installed on the upstream ports.
> >> But I think this is just an academical concern at this point. In
> >> vlan_filtering mode, the CPU port will accept VLAN frames with the
> >> dsa_8021q ID's, but they will eventually get dropped due to no
> >> destination. The real breaker is the pvid change. If something like
> >> patch 4/6 gets accepted I will drop this one.
> >
> > I wish Ward had mentioned to submit such academical concern as RFC :)
> >
> > Please submit smaller series, targeting a single functional problem each,
> > with clear and detailed messages.
>
> Also, I don't think this change set is useful per-se, if we take care of
> removing VLANs on user facing ports, and VLAN filtering is turned on,
> then a frame ingressing an user port with a VLAN that is not part of the
> VLAN table/entries should simply be discarded on ingress, or on egress
> to the CPU port (depending on where the switch performs VID checking),
> so the CPU port cannot possibly receive such a frame, and so removing it
> from the CPU port is correct from a reference counting perspective, but
> useless in practice. Thoughts?
I don't need this patch. I'm not sure what my thought process was at
the time I added it to the patchset.
I'm still interested in getting rid of the vlan bitmap through other
means (picking up your old changeset). Could you take a look at my
questions in that thread? I'm not sure I understand what the user
interaction is supposed to look like for configuring CPU/DSA ports.
> --
> Florian
Thanks,
-Vladimir
^ permalink raw reply
* Re: [PATCH v2 net-next] netdevsim: Fix build error without CONFIG_INET
From: David Miller @ 2019-08-20 20:47 UTC (permalink / raw)
To: yuehaibing; +Cc: idosch, jiri, mcroce, jakub.kicinski, linux-kernel, netdev
In-Reply-To: <20190820141446.71604-1-yuehaibing@huawei.com>
From: YueHaibing <yuehaibing@huawei.com>
Date: Tue, 20 Aug 2019 22:14:46 +0800
> If CONFIG_INET is not set, building fails:
>
> drivers/net/netdevsim/dev.o: In function `nsim_dev_trap_report_work':
> dev.c:(.text+0x67b): undefined reference to `ip_send_check'
>
> Use ip_fast_csum instead of ip_send_check to avoid
> dependencies on CONFIG_INET.
>
> Reported-by: Hulk Robot <hulkci@huawei.com>
> Fixes: da58f90f11f5 ("netdevsim: Add devlink-trap support")
> Signed-off-by: YueHaibing <yuehaibing@huawei.com>
> ---
> v2: use ip_fast_csum instead of ip_send_check
Applied, thank you.
^ permalink raw reply
* Re: [PATCH net-next 0/9] s390/net: updates 2019-08-20
From: David Miller @ 2019-08-20 20:52 UTC (permalink / raw)
To: jwi; +Cc: netdev, linux-s390, heiko.carstens, raspl, ubraun
In-Reply-To: <20190820144643.64041-1-jwi@linux.ibm.com>
From: Julian Wiedmann <jwi@linux.ibm.com>
Date: Tue, 20 Aug 2019 16:46:34 +0200
> please apply the following patches to net-next. This series brings a mix
> of cleanups and small improvements for various parts of qeth's control
> path. Also, a minor cleanup for ctcm and lcs.
Looks good, series applied, thanks.
^ permalink raw reply
* Re: [PATCH mlx5-next 0/5] Mellanox, Updates for mlx5-next branch 2019-08-15
From: Saeed Mahameed @ 2019-08-20 20:54 UTC (permalink / raw)
To: Leon Romanovsky; +Cc: netdev@vger.kernel.org, linux-rdma@vger.kernel.org
In-Reply-To: <20190815194543.14369-1-saeedm@mellanox.com>
On Thu, 2019-08-15 at 19:46 +0000, Saeed Mahameed wrote:
> Hi All,
>
> This series includes misc updates for mlx5-next shared branch.
>
> mlx5 HW spec and bits updates:
> 1) Aya exposes IP-in-IP capability in mlx5_core.
> 2) Maxim exposes lag tx port affinity capabilities.
> 3) Moshe adds VNIC_ENV internal rq counter bits.
>
> Misc updates:
> 4) Saeed, two compiler warnings cleanups
>
>
Applied to mlx5-next.
^ permalink raw reply
* Re: [PATCH 0/6] Add ethernet support for Orange Pi 3
From: David Miller @ 2019-08-20 20:56 UTC (permalink / raw)
To: megous
Cc: robh+dt, mark.rutland, mripard, wens, peppe.cavallaro,
alexandre.torgue, joabreu, mcoquelin.stm32, netdev, devicetree,
linux-arm-kernel, linux-kernel, linux-stm32
In-Reply-To: <20190820145343.29108-1-megous@megous.com>
It looks like there will be some updates to this series either involving
adding -supply to the property names or adjusting some of the kernel log
messages.
Seriously, I would prefer less verbiage in the logs rather than more.
^ permalink raw reply
* Re: [PATCH] net: Fix detection for IPv4 duplicate address.
From: David Miller @ 2019-08-20 20:58 UTC (permalink / raw)
To: liudongxu3; +Cc: kuznet, yoshfuji, netdev, linux-kernel
In-Reply-To: <20190820151905.13148-1-liudongxu3@huawei.com>
From: Dongxu Liu <liudongxu3@huawei.com>
Date: Tue, 20 Aug 2019 23:19:05 +0800
> @@ -800,8 +800,11 @@ static int arp_process(struct net *net, struct sock *sk, struct sk_buff *skb)
> iptunnel_metadata_reply(skb_metadata_dst(skb),
> GFP_ATOMIC);
>
> - /* Special case: IPv4 duplicate address detection packet (RFC2131) */
> - if (sip == 0) {
> +/* Special case: IPv4 duplicate address detection packet (RFC2131).
> + * Linux usually sends zero to detect duplication, and windows may
> + * send a same ip (not zero, sip equal to tip) to do this detection.
> + */
> + if (sip == 0 || sip == tip) {
Regardless of whether this is a valid change or not, you've unindented the
comment which is completely inappropriate.
^ permalink raw reply
* Re: [PATCH net-next 3/6] net: dsa: Delete the VID from the upstream port as well
From: Vivien Didelot @ 2019-08-20 20:58 UTC (permalink / raw)
To: Vladimir Oltean
Cc: Florian Fainelli, Andrew Lunn, Ido Schimmel, Roopa Prabhu,
nikolay, David S. Miller, netdev
In-Reply-To: <CA+h21hqdXP1DnCxwuZOCs4H6MtwzjCnjkBf3ibt+JmnZMEFe=g@mail.gmail.com>
On Tue, 20 Aug 2019 23:40:34 +0300, Vladimir Oltean <olteanv@gmail.com> wrote:
> I don't need this patch. I'm not sure what my thought process was at
> the time I added it to the patchset.
> I'm still interested in getting rid of the vlan bitmap through other
> means (picking up your old changeset). Could you take a look at my
> questions in that thread? I'm not sure I understand what the user
> interaction is supposed to look like for configuring CPU/DSA ports.
What do you mean by getting rid of the vlan bitmap? What do you need exactly?
^ permalink raw reply
* Re: [PATCH] net/mlx5: Fix a memory leak bug
From: Saeed Mahameed @ 2019-08-20 21:00 UTC (permalink / raw)
To: wenwen@cs.uga.edu
Cc: linux-rdma@vger.kernel.org, davem@davemloft.net,
netdev@vger.kernel.org, leon@kernel.org,
linux-kernel@vger.kernel.org
In-Reply-To: <1565684495-2454-1-git-send-email-wenwen@cs.uga.edu>
On Tue, 2019-08-13 at 03:21 -0500, Wenwen Wang wrote:
> In mlx5_cmd_invoke(), 'ent' is allocated through kzalloc() in
> alloc_cmd().
> After the work is queued, wait_func() is invoked to wait the
> completion of
> the work. If wait_func() returns -ETIMEDOUT, the following execution
> will
> be terminated. However, the allocated 'ent' is not deallocated on
> this
> program path, leading to a memory leak bug.
>
> To fix the above issue, free 'ent' before returning the error.
Hi Wenewn, sorry i have to nack this.
As Moshe already pointed out, we intentionally don't free ent, since
even if the driver decided to timeout, FW might still send a
completion, until the FW sends the completion, this entry shouldn't be
freed and is not reusable by driver.
So this is not a memory leak, it just means that only FW completion is
allowed to free this entry or driver shutdown.. otherwise this command
entry is just dead until next fw completion.
Thanks,
Saeed.
>
> Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
> ---
> drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
> b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
> index 8cdd7e6..90cdb9a 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
> @@ -1036,7 +1036,7 @@ static int mlx5_cmd_invoke(struct mlx5_core_dev
> *dev, struct mlx5_cmd_msg *in,
>
> err = wait_func(dev, ent);
> if (err == -ETIMEDOUT)
> - goto out;
> + goto out_free;
>
> ds = ent->ts2 - ent->ts1;
> op = MLX5_GET(mbox_in, in->first.data, opcode);
^ permalink raw reply
* Re: [PATCH net-next 3/6] net: dsa: Delete the VID from the upstream port as well
From: Vladimir Oltean @ 2019-08-20 21:02 UTC (permalink / raw)
To: Vivien Didelot
Cc: Florian Fainelli, Andrew Lunn, Ido Schimmel, Roopa Prabhu,
nikolay, David S. Miller, netdev
In-Reply-To: <20190820165813.GB8523@t480s.localdomain>
On Tue, 20 Aug 2019 at 23:58, Vivien Didelot <vivien.didelot@gmail.com> wrote:
>
> On Tue, 20 Aug 2019 23:40:34 +0300, Vladimir Oltean <olteanv@gmail.com> wrote:
> > I don't need this patch. I'm not sure what my thought process was at
> > the time I added it to the patchset.
> > I'm still interested in getting rid of the vlan bitmap through other
> > means (picking up your old changeset). Could you take a look at my
> > questions in that thread? I'm not sure I understand what the user
> > interaction is supposed to look like for configuring CPU/DSA ports.
>
> What do you mean by getting rid of the vlan bitmap? What do you need exactly?
It would be nice to configure the VLAN attributes of the CPU port in
another way than the implicit way it is done currently. I don't have a
specific use case right now.
^ permalink raw reply
* Re: pull-request: can-next 2019-08-20,pull-request: can-next 2019-08-20
From: David Miller @ 2019-08-20 21:02 UTC (permalink / raw)
To: mkl; +Cc: netdev, kernel, linux-can
In-Reply-To: <827b7ab8-f678-d5ce-a0c6-2a4b5f42238d@pengutronix.de>
From: Marc Kleine-Budde <mkl@pengutronix.de>
Date: Tue, 20 Aug 2019 19:32:13 +0200
> this is a pull request for net-next/master consisting of 18 patches.
Pulled, thanks Marc.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox