DPDK patches and discussions
 help / color / mirror / Atom feed
From: Gagandeep Singh <g.singh@nxp.com>
To: dev@dpdk.org, Hemant Agrawal <hemant.agrawal@nxp.com>,
	Sachin Saxena <sachin.saxena@nxp.com>
Subject: [PATCH 11/11] net/dpaa2: enable software taildrop for ordered queues
Date: Fri, 30 May 2025 12:43:44 +0530	[thread overview]
Message-ID: <20250530071344.2939434-12-g.singh@nxp.com> (raw)
In-Reply-To: <20250530071344.2939434-1-g.singh@nxp.com>

This patch adds support for software taildrop on ordered queues
in the DPAA2 driver.
It also enables congestion notification by default on
traffic management (TM) queues,  which is a prerequisite for
software taildrop functionality.

Signed-off-by: Gagandeep Singh <g.singh@nxp.com>
---
 drivers/net/dpaa2/dpaa2_rxtx.c | 24 +++++++++++++-
 drivers/net/dpaa2/dpaa2_tm.c   | 60 ++++++++++++++++------------------
 2 files changed, 52 insertions(+), 32 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_rxtx.c b/drivers/net/dpaa2/dpaa2_rxtx.c
index 67d065bb7c..dce1da80bb 100644
--- a/drivers/net/dpaa2/dpaa2_rxtx.c
+++ b/drivers/net/dpaa2/dpaa2_rxtx.c
@@ -1800,8 +1800,11 @@ dpaa2_dev_tx_ordered(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 		while (qbman_result_SCN_state(dpaa2_q->cscn)) {
 			retry_count++;
 			/* Retry for some time before giving up */
-			if (retry_count > CONG_RETRY_COUNT)
+			if (retry_count > CONG_RETRY_COUNT) {
+				if (dpaa2_q->tm_sw_td)
+					goto sw_td;
 				goto skip_tx;
+			}
 		}
 
 		frames_to_send = (nb_pkts > dpaa2_eqcr_size) ?
@@ -1961,6 +1964,25 @@ dpaa2_dev_tx_ordered(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 			rte_pktmbuf_free_seg(buf_to_free[loop].seg);
 	}
 
+	return num_tx;
+sw_td:
+	loop = 0;
+	while (loop < num_tx) {
+		if (unlikely(RTE_MBUF_HAS_EXTBUF(*bufs)))
+			rte_pktmbuf_free(*bufs);
+		bufs++;
+		loop++;
+	}
+
+	/* free the pending buffers */
+	while (nb_pkts) {
+		rte_pktmbuf_free(*bufs);
+		bufs++;
+		nb_pkts--;
+		num_tx++;
+	}
+	dpaa2_q->tx_pkts += num_tx;
+
 	return num_tx;
 }
 
diff --git a/drivers/net/dpaa2/dpaa2_tm.c b/drivers/net/dpaa2/dpaa2_tm.c
index dbf66c756e..36e815c356 100644
--- a/drivers/net/dpaa2/dpaa2_tm.c
+++ b/drivers/net/dpaa2/dpaa2_tm.c
@@ -611,41 +611,39 @@ dpaa2_tm_configure_queue(struct rte_eth_dev *dev, struct dpaa2_tm_node *node)
 	}
 	dpaa2_q->fqid = qid.fqid;
 
-	/* setting congestion notification */
-	if (!(priv->flags & DPAA2_TX_CGR_OFF)) {
-		struct dpni_congestion_notification_cfg cong_notif_cfg = {0};
-
-		cong_notif_cfg.units = DPNI_CONGESTION_UNIT_FRAMES;
-		cong_notif_cfg.threshold_entry = dpaa2_q->nb_desc;
-		/* Notify that the queue is not congested when the data in
-		 * the queue is below this thershold.(90% of value)
-		 */
-		cong_notif_cfg.threshold_exit = (dpaa2_q->nb_desc * 9) / 10;
-		cong_notif_cfg.message_ctx = 0;
-		iova = DPAA2_VADDR_TO_IOVA_AND_CHECK(dpaa2_q->cscn,
-				sizeof(struct qbman_result));
-		if (iova == RTE_BAD_IOVA) {
-			DPAA2_PMD_ERR("No IOMMU map for cscn(%p)", dpaa2_q->cscn);
-			return -ENOBUFS;
-		}
-		cong_notif_cfg.message_iova = iova;
-		cong_notif_cfg.dest_cfg.dest_type = DPNI_DEST_NONE;
-		cong_notif_cfg.notification_mode =
+	/* setting taildrop through congestion notification */
+	struct dpni_congestion_notification_cfg cong_notif_cfg = {0};
+
+	cong_notif_cfg.units = DPNI_CONGESTION_UNIT_FRAMES;
+	cong_notif_cfg.threshold_entry = dpaa2_q->nb_desc;
+	/* Notify that the queue is not congested when the data in
+	 * the queue is below this thershold.(90% of value)
+	 */
+	cong_notif_cfg.threshold_exit = (dpaa2_q->nb_desc * 9) / 10;
+	cong_notif_cfg.message_ctx = 0;
+
+	iova = DPAA2_VADDR_TO_IOVA_AND_CHECK(dpaa2_q->cscn,
+			sizeof(struct qbman_result));
+	if (iova == RTE_BAD_IOVA) {
+		DPAA2_PMD_ERR("No IOMMU map for cscn(%p)", dpaa2_q->cscn);
+		return -ENOBUFS;
+	}
+	cong_notif_cfg.message_iova = iova;
+	cong_notif_cfg.dest_cfg.dest_type = DPNI_DEST_NONE;
+	cong_notif_cfg.notification_mode =
 					DPNI_CONG_OPT_WRITE_MEM_ON_ENTER |
 					DPNI_CONG_OPT_WRITE_MEM_ON_EXIT |
 					DPNI_CONG_OPT_COHERENT_WRITE;
-		cong_notif_cfg.cg_point = DPNI_CP_QUEUE;
+	cong_notif_cfg.cg_point = DPNI_CP_QUEUE;
 
-		ret = dpni_set_congestion_notification(dpni, CMD_PRI_LOW,
-					priv->token,
-					DPNI_QUEUE_TX,
-					((node->parent->channel_id << 8) | tc_id),
-					&cong_notif_cfg);
-		if (ret) {
-			DPAA2_PMD_ERR("Error in setting tx congestion notification: "
-				"err=%d", ret);
-			return -ret;
-		}
+	ret = dpni_set_congestion_notification(dpni, CMD_PRI_LOW,
+			priv->token, DPNI_QUEUE_TX,
+			((node->parent->channel_id << 8) | tc_id),
+			&cong_notif_cfg);
+	if (ret) {
+		DPAA2_PMD_ERR("Error in setting tx congestion notification: "
+			"err=%d", ret);
+		return -ret;
 	}
 	dpaa2_q->tm_sw_td = true;
 
-- 
2.25.1


  parent reply	other threads:[~2025-05-30  7:15 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-05-30  7:13 [PATCH 00/11] NXP DPAA2 driver enhancements and fixes Gagandeep Singh
2025-05-30  7:13 ` [PATCH 01/11] net/dpaa2: fix issue of extract buffer preparation Gagandeep Singh
2025-05-30  7:13 ` [PATCH 02/11] net/dpaa2: fix shaper rate Gagandeep Singh
2025-05-30  7:13 ` [PATCH 03/11] bus/fslmc: add DPBP APIs for setting depletion thresholds Gagandeep Singh
2025-05-30  7:13 ` [PATCH 04/11] mempool/dpaa2: use unified VA to IOVA conversion Gagandeep Singh
2025-05-30  7:13 ` [PATCH 05/11] net/dpaa2: add dpmac MC header file Gagandeep Singh
2025-05-30  7:13 ` [PATCH 06/11] net/dpaa2: support dpmac counters in stats Gagandeep Singh
2025-05-30  7:13 ` [PATCH 07/11] net/dpaa2: support dpmac Tx stats Gagandeep Singh
2025-05-30  7:13 ` [PATCH 08/11] net/dpaa2: support dpmac Tx stats in xstats Gagandeep Singh
2025-05-30  7:13 ` [PATCH 09/11] net/dpaa2: retrieve DPNI API version at init time Gagandeep Singh
2025-05-30  7:13 ` [PATCH 10/11] net/dpaa2: setup the speed cap based on the actual MAC Gagandeep Singh
2025-05-30  7:13 ` Gagandeep Singh [this message]
2025-06-02 10:40 ` [PATCH v2 00/11] NXP DPAA2 driver enhancements and fixes Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 01/11] net/dpaa2: fix issue of extract buffer preparation Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 02/11] net/dpaa2: fix shaper rate Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 03/11] bus/fslmc: add DPBP APIs for setting depletion thresholds Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 04/11] mempool/dpaa2: use unified VA to IOVA conversion Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 05/11] net/dpaa2: add dpmac MC header file Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 06/11] net/dpaa2: support dpmac counters in stats Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 07/11] net/dpaa2: support dpmac Tx stats Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 08/11] net/dpaa2: support dpmac Tx stats in xstats Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 09/11] net/dpaa2: retrieve DPNI API version at init time Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 10/11] net/dpaa2: setup the speed cap based on the actual MAC Gagandeep Singh
2025-06-02 10:40   ` [PATCH v2 11/11] net/dpaa2: enable software taildrop for ordered queues Gagandeep Singh

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=20250530071344.2939434-12-g.singh@nxp.com \
    --to=g.singh@nxp.com \
    --cc=dev@dpdk.org \
    --cc=hemant.agrawal@nxp.com \
    --cc=sachin.saxena@nxp.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).