[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <577f7b9e99d1a86284a02792e038511293e56710.1480062521.git.jslaby@suse.cz>
Date: Fri, 25 Nov 2016 09:29:52 +0100
From: Jiri Slaby <jslaby@...e.cz>
To: stable@...r.kernel.org
Cc: linux-kernel@...r.kernel.org,
Daniel Jurgens <danielj@...lanox.com>,
Maor Gottlieb <maorg@...lanox.com>,
Leon Romanovsky <leon@...nel.org>,
Doug Ledford <dledford@...hat.com>, Jiri Slaby <jslaby@...e.cz>
Subject: [PATCH 3.12 079/127] IB/mlx5: Use cache line size to select CQE stride
From: Daniel Jurgens <danielj@...lanox.com>
3.12-stable review patch. If anyone has any objections, please let me know.
===============
commit 16b0e0695a73b68d8ca40288c8f9614ef208917b upstream.
When creating kernel CQs use 128B CQE stride if the
cache line size is 128B, 64B otherwise. This prevents
multiple CQEs from residing in a 128B cache line,
which can cause retries when there are concurrent
read and writes in one cache line.
Tested with IPoIB on PPC64, saw ~5% throughput
improvement.
Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Daniel Jurgens <danielj@...lanox.com>
Signed-off-by: Maor Gottlieb <maorg@...lanox.com>
Signed-off-by: Leon Romanovsky <leon@...nel.org>
Signed-off-by: Doug Ledford <dledford@...hat.com>
Signed-off-by: Jiri Slaby <jslaby@...e.cz>
---
drivers/infiniband/hw/mlx5/cq.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 706833ab7e7e..e5a6d839f1d1 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -684,8 +684,7 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
if (err)
goto err_create;
} else {
- /* for now choose 64 bytes till we have a proper interface */
- cqe_size = 64;
+ cqe_size = cache_line_size() == 128 ? 128 : 64;
err = create_cq_kernel(dev, cq, entries, cqe_size, &cqb,
&index, &inlen);
if (err)
--
2.10.2
Powered by blists - more mailing lists