From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Maxim Mikityanskiy <maximmi@mellanox.com>,
Tariq Toukan <tariqt@mellanox.com>,
Saeed Mahameed <saeedm@mellanox.com>,
Sasha Levin <sashal@kernel.org>,
netdev@vger.kernel.org, linux-rdma@vger.kernel.org
Subject: [PATCH AUTOSEL 5.6 49/79] net/mlx5e: Don't trigger IRQ multiple times on XSK wakeup to avoid WQ overruns
Date: Thu, 30 Apr 2020 09:50:13 -0400 [thread overview]
Message-ID: <20200430135043.19851-49-sashal@kernel.org> (raw)
In-Reply-To: <20200430135043.19851-1-sashal@kernel.org>
From: Maxim Mikityanskiy <maximmi@mellanox.com>
[ Upstream commit e7e0004abdd6f83ae4be5613b29ed396beff576c ]
XSK wakeup function triggers NAPI by posting a NOP WQE to a special XSK
ICOSQ. When the application floods the driver with wakeup requests by
calling sendto() in a certain pattern that ends up in mlx5e_trigger_irq,
the XSK ICOSQ may overflow.
Multiple NOPs are not required and won't accelerate the process, so
avoid posting a second NOP if there is one already on the way. This way
we also avoid increasing the queue size (which might not help anyway).
Fixes: db05815b36cb ("net/mlx5e: Add XSK zero-copy support")
Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 3 ++-
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c | 3 +++
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 8 +++++---
drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c | 6 +++++-
4 files changed, 15 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index ddd2409fc8bef..5a5e6a21c6e1f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -367,6 +367,7 @@ enum {
MLX5E_SQ_STATE_AM,
MLX5E_SQ_STATE_TLS,
MLX5E_SQ_STATE_VLAN_NEED_L2_INLINE,
+ MLX5E_SQ_STATE_PENDING_XSK_TX,
};
struct mlx5e_sq_wqe_info {
@@ -950,7 +951,7 @@ void mlx5e_page_release_dynamic(struct mlx5e_rq *rq,
void mlx5e_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe);
void mlx5e_handle_rx_cqe_mpwrq(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe);
bool mlx5e_post_rx_wqes(struct mlx5e_rq *rq);
-void mlx5e_poll_ico_cq(struct mlx5e_cq *cq);
+int mlx5e_poll_ico_cq(struct mlx5e_cq *cq);
bool mlx5e_post_rx_mpwqes(struct mlx5e_rq *rq);
void mlx5e_dealloc_rx_wqe(struct mlx5e_rq *rq, u16 ix);
void mlx5e_dealloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
index fe2d596cb361f..3bcdb5b2fc203 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c
@@ -33,6 +33,9 @@ int mlx5e_xsk_wakeup(struct net_device *dev, u32 qid, u32 flags)
if (unlikely(!test_bit(MLX5E_SQ_STATE_ENABLED, &c->xskicosq.state)))
return 0;
+ if (test_and_set_bit(MLX5E_SQ_STATE_PENDING_XSK_TX, &c->xskicosq.state))
+ return 0;
+
spin_lock(&c->xskicosq_lock);
mlx5e_trigger_irq(&c->xskicosq);
spin_unlock(&c->xskicosq_lock);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 312d4692425b7..a9a96a630e4d6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -587,7 +587,7 @@ bool mlx5e_post_rx_wqes(struct mlx5e_rq *rq)
return !!err;
}
-void mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
+int mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
{
struct mlx5e_icosq *sq = container_of(cq, struct mlx5e_icosq, cq);
struct mlx5_cqe64 *cqe;
@@ -595,11 +595,11 @@ void mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
int i;
if (unlikely(!test_bit(MLX5E_SQ_STATE_ENABLED, &sq->state)))
- return;
+ return 0;
cqe = mlx5_cqwq_get_cqe(&cq->wq);
if (likely(!cqe))
- return;
+ return 0;
/* sq->cc must be updated only after mlx5_cqwq_update_db_record(),
* otherwise a cq overrun may occur
@@ -648,6 +648,8 @@ void mlx5e_poll_ico_cq(struct mlx5e_cq *cq)
sq->cc = sqcc;
mlx5_cqwq_update_db_record(&cq->wq);
+
+ return i;
}
bool mlx5e_post_rx_mpwqes(struct mlx5e_rq *rq)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
index 800d34ed8a96c..76efa9579215c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
@@ -145,7 +145,11 @@ int mlx5e_napi_poll(struct napi_struct *napi, int budget)
busy |= rq->post_wqes(rq);
if (xsk_open) {
- mlx5e_poll_ico_cq(&c->xskicosq.cq);
+ if (mlx5e_poll_ico_cq(&c->xskicosq.cq))
+ /* Don't clear the flag if nothing was polled to prevent
+ * queueing more WQEs and overflowing XSKICOSQ.
+ */
+ clear_bit(MLX5E_SQ_STATE_PENDING_XSK_TX, &c->xskicosq.state);
busy |= mlx5e_poll_xdpsq_cq(&xsksq->cq);
busy_xsk |= mlx5e_napi_xsk_post(xsksq, xskrq);
}
--
2.20.1
next prev parent reply other threads:[~2020-04-30 14:10 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20200430135043.19851-1-sashal@kernel.org>
2020-04-30 13:49 ` [PATCH AUTOSEL 5.6 14/79] SUNRPC/cache: Fix unsafe traverse caused double-free in cache_purge Sasha Levin
2020-04-30 13:49 ` [PATCH AUTOSEL 5.6 35/79] svcrdma: Fix trace point use-after-free race Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 40/79] wimax/i2400m: Fix potential urb refcnt leak Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 41/79] net: stmmac: fix enabling socfpga's ptp_ref_clock Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 42/79] net: stmmac: Fix sub-second increment Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 43/79] netfilter: nat: fix error handling upon registering inet hook Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 48/79] net/mlx5: Fix failing fw tracer allocation on s390 Sasha Levin
2020-04-30 13:50 ` Sasha Levin [this message]
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 50/79] net/mlx5e: Get the latest values from counters in switchdev mode Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 51/79] cpumap: Avoid warning when CONFIG_DEBUG_PER_CPU_MAPS is enabled Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 52/79] bpf: Forbid XADD on spilled pointers for unprivileged users Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 53/79] bpf: Fix handling of XADD on BTF memory Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 54/79] bpf, x86: Fix encoding for lower 8-bit registers in BPF_STX BPF_B Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 55/79] iwlwifi: actually check allocated conf_tlv pointer Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 67/79] net: phy: bcm84881: clear settings on link down Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 69/79] mac80211: sta_info: Add lockdep condition for RCU list usage Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 72/79] net: bcmgenet: suppress warnings on failed Rx SKB allocations Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 73/79] net: systemport: " Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 74/79] bpf, x86_32: Fix incorrect encoding in BPF_LDX zero-extension Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 75/79] bpf, x86_32: Fix clobbering of dst for BPF_JSET Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 76/79] bpf, x86_32: Fix logic error in BPF_LDX zero-extension Sasha Levin
2020-04-30 13:50 ` [PATCH AUTOSEL 5.6 77/79] bpf: Propagate expected_attach_type when verifying freplace programs Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200430135043.19851-49-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=maximmi@mellanox.com \
--cc=netdev@vger.kernel.org \
--cc=saeedm@mellanox.com \
--cc=stable@vger.kernel.org \
--cc=tariqt@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).