You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
kmod-redhat-mlx5_core/SOURCES/0152-netdrv-net-mlx5e-TX-Er...

144 lines
4.9 KiB

From ca8424eebb93cf0c2b311f0f12ea911361b4833f Mon Sep 17 00:00:00 2001
From: Alaa Hleihel <ahleihel@redhat.com>
Date: Tue, 12 May 2020 10:54:15 -0400
Subject: [PATCH 152/312] [netdrv] net/mlx5e: TX, Error completion is for last
WQE in batch
Message-id: <20200512105530.4207-50-ahleihel@redhat.com>
Patchwork-id: 306921
Patchwork-instance: patchwork
O-Subject: [RHEL8.3 BZ 1789382 049/124] net/mlx5e: TX, Error completion is for last WQE in batch
Bugzilla: 1789382
RH-Acked-by: Tony Camuso <tcamuso@redhat.com>
RH-Acked-by: Kamal Heib <kheib@redhat.com>
RH-Acked-by: Jarod Wilson <jarod@redhat.com>
Bugzilla: http://bugzilla.redhat.com/1789382
Upstream: v5.6-rc1
commit b57e66ad42d051ed31319c28ed1b62b191299a29
Author: Tariq Toukan <tariqt@mellanox.com>
Date: Thu Jan 9 15:53:37 2020 +0200
net/mlx5e: TX, Error completion is for last WQE in batch
For a cyclic work queue, when not requesting a completion per WQE,
a single CQE might indicate the completion of several WQEs.
However, in case some WQE in the batch causes an error, then an error
completion is issued, breaking the batch, and pointing to the offending
WQE in the wqe_counter field.
Hence, WQE-specific error CQE handling (like printing, breaking, etc...)
should be performed only for the last WQE in batch.
Fixes: 130c7b46c93d ("net/mlx5e: TX, Dump WQs wqe descriptors on CQE with error events")
Fixes: fd9b4be8002c ("net/mlx5e: RX, Support multiple outstanding UMR posts")
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Alaa Hleihel <ahleihel@redhat.com>
Signed-off-by: Frantisek Hrbata <fhrbata@redhat.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 16 ++++++------
drivers/net/ethernet/mellanox/mlx5/core/en_tx.c | 33 +++++++++++--------------
2 files changed, 23 insertions(+), 26 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 64d6ecbece80..f0170fb2edbc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -614,13 +614,6 @@ void mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
wqe_counter = be16_to_cpu(cqe->wqe_counter);
- 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 {
struct mlx5e_sq_wqe_info *wi;
u16 ci;
@@ -630,6 +623,15 @@ void mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
ci = mlx5_wq_cyc_ctr2ix(&sq->wq, sqcc);
wi = &sq->db.ico_wqe[ci];
+ if (last_wqe && 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;
+ }
+
if (likely(wi->opcode == MLX5_OPCODE_UMR)) {
sqcc += MLX5E_UMR_WQEBBS;
wi->umr.rq->mpwqe.umr_completed++;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index 3ce27194ee7e..3bfeb7c06b25 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -452,34 +452,17 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq, int napi_budget)
i = 0;
do {
+ struct mlx5e_tx_wqe_info *wi;
u16 wqe_counter;
bool last_wqe;
+ u16 ci;
mlx5_cqwq_pop(&cq->wq);
wqe_counter = be16_to_cpu(cqe->wqe_counter);
- if (unlikely(get_cqe_opcode(cqe) == MLX5_CQE_REQ_ERR)) {
- if (!test_and_set_bit(MLX5E_SQ_STATE_RECOVERING,
- &sq->state)) {
- struct mlx5e_tx_wqe_info *wi;
- u16 ci;
-
- ci = mlx5_wq_cyc_ctr2ix(&sq->wq, sqcc);
- wi = &sq->db.wqe_info[ci];
- mlx5e_dump_error_cqe(sq,
- (struct mlx5_err_cqe *)cqe);
- mlx5_wq_cyc_wqe_dump(&sq->wq, ci, wi->num_wqebbs);
- queue_work(cq->channel->priv->wq,
- &sq->recover_work);
- }
- stats->cqe_err++;
- }
-
do {
- struct mlx5e_tx_wqe_info *wi;
struct sk_buff *skb;
- u16 ci;
int j;
last_wqe = (sqcc == wqe_counter);
@@ -517,6 +500,18 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq, int napi_budget)
napi_consume_skb(skb, napi_budget);
} while (!last_wqe);
+ if (unlikely(get_cqe_opcode(cqe) == MLX5_CQE_REQ_ERR)) {
+ if (!test_and_set_bit(MLX5E_SQ_STATE_RECOVERING,
+ &sq->state)) {
+ mlx5e_dump_error_cqe(sq,
+ (struct mlx5_err_cqe *)cqe);
+ mlx5_wq_cyc_wqe_dump(&sq->wq, ci, wi->num_wqebbs);
+ queue_work(cq->channel->priv->wq,
+ &sq->recover_work);
+ }
+ stats->cqe_err++;
+ }
+
} while ((++i < MLX5E_TX_CQ_POLL_BUDGET) && (cqe = mlx5_cqwq_get_cqe(&cq->wq)));
stats->cqes += i;
--
2.13.6