patches for DPDK stable branches
 help / color / mirror / Atom feed
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
To: <luca.boccassi@gmail.com>
Cc: <stable@dpdk.org>, <xavier.huwei@huawei.com>
Subject: [dpdk-stable] [PATCH 19.11 v2 4/5] net/hns3: add free threshold in Rx
Date: Tue, 2 Jun 2020 09:28:06 +0800	[thread overview]
Message-ID: <1591061288-39518-5-git-send-email-xavier.huwei@huawei.com> (raw)
In-Reply-To: <1591061288-39518-1-git-send-email-xavier.huwei@huawei.com>

[ upstream commit ffd0ec015b9f51d3f39e320a02448fc5d43932be ]

This patch optimizes the Rx performance by adding the rx_free_thresh
related process in the '.rx_pkt_burst' ops implementation function named
hns3_recv_pkts. The related change as follows:
1. Adding the rx_free_thresh related process to reduce the number of
   writing the HNS3_RING_RX_HEAD_REG register.
2. Adjusting the internal macro named DEFAULT_RX_FREE_THRESH to 32 and
   adjusting HNS3_MIN_RING_DESC to 64 to make the effect of the thresh
   more obvious.

Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
---
 drivers/net/hns3/hns3_rxtx.c | 12 ++++++++++--
 drivers/net/hns3/hns3_rxtx.h |  2 +-
 2 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index 23bb115..34e7448 100644
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -30,7 +30,7 @@
 #include "hns3_logs.h"
 
 #define HNS3_CFG_DESC_NUM(num)	((num) / 8 - 1)
-#define DEFAULT_RX_FREE_THRESH	16
+#define DEFAULT_RX_FREE_THRESH	32
 
 static void
 hns3_rx_queue_release_mbufs(struct hns3_rx_queue *rxq)
@@ -610,6 +610,7 @@ hns3_dev_rx_queue_start(struct hns3_adapter *hns, uint16_t idx)
 
 	rxq->next_to_use = 0;
 	rxq->next_to_clean = 0;
+	rxq->nb_rx_hold = 0;
 	hns3_init_rx_queue_hw(rxq);
 
 	return 0;
@@ -624,6 +625,7 @@ hns3_fake_rx_queue_start(struct hns3_adapter *hns, uint16_t idx)
 	rxq = (struct hns3_rx_queue *)hw->fkq_data.rx_queues[idx];
 	rxq->next_to_use = 0;
 	rxq->next_to_clean = 0;
+	rxq->nb_rx_hold = 0;
 	hns3_init_rx_queue_hw(rxq);
 }
 
@@ -1577,7 +1579,13 @@ hns3_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts)
 	rxq->next_to_clean = rx_id;
 	rxq->pkt_first_seg = first_seg;
 	rxq->pkt_last_seg = last_seg;
-	hns3_clean_rx_buffers(rxq, nb_rx_bd);
+
+	nb_rx_bd = nb_rx_bd + rxq->nb_rx_hold;
+	if (nb_rx_bd > rxq->rx_free_thresh) {
+		hns3_clean_rx_buffers(rxq, nb_rx_bd);
+		nb_rx_bd = 0;
+	}
+	rxq->nb_rx_hold = nb_rx_bd;
 
 	return nb_rx;
 }
diff --git a/drivers/net/hns3/hns3_rxtx.h b/drivers/net/hns3/hns3_rxtx.h
index 00b92cf..771f3c9 100644
--- a/drivers/net/hns3/hns3_rxtx.h
+++ b/drivers/net/hns3/hns3_rxtx.h
@@ -5,7 +5,7 @@
 #ifndef _HNS3_RXTX_H_
 #define _HNS3_RXTX_H_
 
-#define	HNS3_MIN_RING_DESC	32
+#define	HNS3_MIN_RING_DESC	64
 #define	HNS3_MAX_RING_DESC	32768
 #define HNS3_DEFAULT_RING_DESC  1024
 #define	HNS3_ALIGN_RING_DESC	32
-- 
2.7.4


  parent reply	other threads:[~2020-06-02  1:33 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-02  1:28 [dpdk-stable] [PATCH 19.11 v2 0/5] improving I/O backport for hns3 PMD driver Wei Hu (Xavier)
2020-06-02  1:28 ` [dpdk-stable] [PATCH 19.11 v2 1/5] net/hns3: remove unnecessary assignments in Tx Wei Hu (Xavier)
2020-06-02  1:28 ` [dpdk-stable] [PATCH 19.11 v2 2/5] net/hns3: reduce judgements of free Tx ring space Wei Hu (Xavier)
2020-06-02  1:28 ` [dpdk-stable] [PATCH 19.11 v2 3/5] net/hns3: remove one IO barrier in Rx Wei Hu (Xavier)
2020-06-02  1:28 ` Wei Hu (Xavier) [this message]
2020-06-02  1:28 ` [dpdk-stable] [PATCH 19.11 v2 5/5] net/hns3: replace memory barrier with data dependency order Wei Hu (Xavier)
2020-06-02 13:27 ` [dpdk-stable] [PATCH 19.11 v2 0/5] improving I/O backport for hns3 PMD driver Luca Boccassi

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=1591061288-39518-5-git-send-email-xavier.huwei@huawei.com \
    --to=xavier.huwei@huawei.com \
    --cc=luca.boccassi@gmail.com \
    --cc=stable@dpdk.org \
    /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).