[dpdk-stable] patch 'net/qede: fix VF MTU update' has been queued to LTS release 17.11.4

[dpdk-stable] patch 'net/qede: fix VF MTU update' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 5ca5cdc9b0f682b7614156f182982d268cf0c5b1 Mon Sep 17 00:00:00 2001
From: Rasesh Mody <rasesh.mody@cavium.com>
Date: Tue, 22 May 2018 16:16:03 -0700
Subject: [PATCH] net/qede: fix VF MTU update

[ upstream commit d121a6b5f781525f65257920e97c1f033229f5e8 ]

This patch fixes VF MTU update to work without having to restart the
vport and there by not requiring port re-configuration. It adds a
VF MTU Update TLV to achieve the same. Firmware can handle VF MTU update
by just pausing the vport.

Fixes: dd28bc8c6ef4 ("net/qede: fix VF port creation sequence")

Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
---
 drivers/net/qede/base/ecore_sriov.c   |  44 +++++++++++++
 drivers/net/qede/base/ecore_vf.c      |  33 ++++++++++
 drivers/net/qede/base/ecore_vf.h      |   9 +++
 drivers/net/qede/base/ecore_vfpf_if.h |  16 +++++
 drivers/net/qede/qede_ethdev.c        | 120 ++++++++++++++++++++--------------
 drivers/net/qede/qede_ethdev.h        |   1 +
 6 files changed, 175 insertions(+), 48 deletions(-)

diff --git a/drivers/net/qede/base/ecore_sriov.c b/drivers/net/qede/base/ecore_sriov.c
index 68f40f8a0..1c885e1cc 100644
--- a/drivers/net/qede/base/ecore_sriov.c
+++ b/drivers/net/qede/base/ecore_sriov.c
@@ -61,6 +61,8 @@ const char *ecore_channel_tlvs_string[] = {
 	"CHANNEL_TLV_COALESCE_UPDATE",
 	"CHANNEL_TLV_QID",
 	"CHANNEL_TLV_COALESCE_READ",
+	"CHANNEL_TLV_BULLETIN_UPDATE_MAC",
+	"CHANNEL_TLV_UPDATE_MTU",
 	"CHANNEL_TLV_MAX"
 };
 
@@ -2854,6 +2856,45 @@ out:
 			       length, status);
 }
 
+static enum _ecore_status_t
+ecore_iov_vf_pf_update_mtu(struct ecore_hwfn *p_hwfn,
+				    struct ecore_ptt *p_ptt,
+				    struct ecore_vf_info *p_vf)
+{
+	struct ecore_iov_vf_mbx *mbx = &p_vf->vf_mbx;
+	struct ecore_sp_vport_update_params params;
+	enum _ecore_status_t rc = ECORE_SUCCESS;
+	struct vfpf_update_mtu_tlv *p_req;
+	u8 status = PFVF_STATUS_SUCCESS;
+
+	/* Valiate PF can send such a request */
+	if (!p_vf->vport_instance) {
+		DP_VERBOSE(p_hwfn, ECORE_MSG_IOV,
+			   "No VPORT instance available for VF[%d], failing MTU update\n",
+			   p_vf->abs_vf_id);
+		status = PFVF_STATUS_FAILURE;
+		goto send_status;
+	}
+
+	p_req = &mbx->req_virt->update_mtu;
+
+	OSAL_MEMSET(&params, 0, sizeof(params));
+	params.opaque_fid =  p_vf->opaque_fid;
+	params.vport_id = p_vf->vport_id;
+	params.mtu = p_req->mtu;
+	rc = ecore_sp_vport_update(p_hwfn, &params, ECORE_SPQ_MODE_EBLOCK,
+				   OSAL_NULL);
+
+	if (rc)
+		status = PFVF_STATUS_FAILURE;
+send_status:
+	ecore_iov_prepare_resp(p_hwfn, p_ptt, p_vf,
+			       CHANNEL_TLV_UPDATE_MTU,
+			       sizeof(struct pfvf_def_resp_tlv),
+			       status);
+	return rc;
+}
+
 void *ecore_iov_search_list_tlvs(struct ecore_hwfn *p_hwfn,
 				 void *p_tlvs_list, u16 req_type)
 {
@@ -4136,6 +4177,9 @@ void ecore_iov_process_mbx_req(struct ecore_hwfn *p_hwfn,
 		case CHANNEL_TLV_COALESCE_READ:
 			ecore_iov_vf_pf_get_coalesce(p_hwfn, p_ptt, p_vf);
 			break;
+		case CHANNEL_TLV_UPDATE_MTU:
+			ecore_iov_vf_pf_update_mtu(p_hwfn, p_ptt, p_vf);
+			break;
 		}
 	} else if (ecore_iov_tlv_supported(mbx->first_tlv.tl.type)) {
 		/* If we've received a message from a VF we consider malicious
diff --git a/drivers/net/qede/base/ecore_vf.c b/drivers/net/qede/base/ecore_vf.c
index 8a08911a3..334db6b97 100644
--- a/drivers/net/qede/base/ecore_vf.c
+++ b/drivers/net/qede/base/ecore_vf.c
@@ -1628,6 +1628,39 @@ exit:
 	return rc;
 }
 
+enum _ecore_status_t
+ecore_vf_pf_update_mtu(struct ecore_hwfn *p_hwfn, u16 mtu)
+{
+	struct ecore_vf_iov *p_iov = p_hwfn->vf_iov_info;
+	struct vfpf_update_mtu_tlv *p_req;
+	struct pfvf_def_resp_tlv *p_resp;
+	enum _ecore_status_t rc;
+
+	if (!mtu)
+		return ECORE_INVAL;
+
+	/* clear mailbox and prep header tlv */
+	p_req = ecore_vf_pf_prep(p_hwfn, CHANNEL_TLV_UPDATE_MTU,
+				 sizeof(*p_req));
+	p_req->mtu = mtu;
+	DP_VERBOSE(p_hwfn, ECORE_MSG_IOV,
+		   "Requesting MTU update to %d\n", mtu);
+
+	/* add list termination tlv */
+	ecore_add_tlv(&p_iov->offset,
+		      CHANNEL_TLV_LIST_END,
+		      sizeof(struct channel_list_end_tlv));
+
+	p_resp = &p_iov->pf2vf_reply->default_resp;
+	rc = ecore_send_msg2pf(p_hwfn, &p_resp->hdr.status, sizeof(*p_resp));
+	if (p_resp->hdr.status == PFVF_STATUS_NOT_SUPPORTED)
+		rc = ECORE_INVAL;
+
+	ecore_vf_pf_req_end(p_hwfn, rc);
+
+	return rc;
+}
+
 u16 ecore_vf_get_igu_sb_id(struct ecore_hwfn *p_hwfn,
 			   u16               sb_id)
 {
diff --git a/drivers/net/qede/base/ecore_vf.h b/drivers/net/qede/base/ecore_vf.h
index de2758cbd..e5555bb1e 100644
--- a/drivers/net/qede/base/ecore_vf.h
+++ b/drivers/net/qede/base/ecore_vf.h
@@ -319,5 +319,14 @@ void ecore_vf_set_vf_start_tunn_update_param(struct ecore_tunnel_info *p_tun);
 
 u32 ecore_vf_hw_bar_size(struct ecore_hwfn *p_hwfn,
 		     enum BAR_ID bar_id);
+
+/**
+ * @brief - ecore_vf_pf_update_mtu Update MTU for VF.
+ *
+ * @param p_hwfn
+ * @param - mtu
+ */
+enum _ecore_status_t
+ecore_vf_pf_update_mtu(struct ecore_hwfn *p_hwfn, u16 mtu);
 #endif
 #endif /* __ECORE_VF_H__ */
diff --git a/drivers/net/qede/base/ecore_vfpf_if.h b/drivers/net/qede/base/ecore_vfpf_if.h
index c6af9caf7..08b1f2484 100644
--- a/drivers/net/qede/base/ecore_vfpf_if.h
+++ b/drivers/net/qede/base/ecore_vfpf_if.h
@@ -531,6 +531,18 @@ struct pfvf_read_coal_resp_tlv {
 	u8 padding[6];
 };
 
+struct vfpf_bulletin_update_mac_tlv {
+	struct vfpf_first_tlv first_tlv;
+	u8 mac[ETH_ALEN];
+	u8 padding[2];
+};
+
+struct vfpf_update_mtu_tlv {
+	struct vfpf_first_tlv first_tlv;
+	u16 mtu;
+	u8 padding[6];
+};
+
 union vfpf_tlvs {
 	struct vfpf_first_tlv			first_tlv;
 	struct vfpf_acquire_tlv			acquire;
@@ -545,6 +557,8 @@ union vfpf_tlvs {
 	struct vfpf_update_tunn_param_tlv	tunn_param_update;
 	struct vfpf_update_coalesce		update_coalesce;
 	struct vfpf_read_coal_req_tlv		read_coal_req;
+	struct vfpf_bulletin_update_mac_tlv	bulletin_update_mac;
+	struct vfpf_update_mtu_tlv		update_mtu;
 	struct tlv_buffer_size			tlv_buf_size;
 };
 
@@ -675,6 +689,8 @@ enum {
 	CHANNEL_TLV_COALESCE_UPDATE,
 	CHANNEL_TLV_QID,
 	CHANNEL_TLV_COALESCE_READ,
+	CHANNEL_TLV_BULLETIN_UPDATE_MAC,
+	CHANNEL_TLV_UPDATE_MTU,
 	CHANNEL_TLV_MAX,
 
 	/* Required for iterating over vport-update tlvs.
diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index 7462f1ad7..c63b4d5c9 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -597,37 +597,6 @@ int qede_enable_tpa(struct rte_eth_dev *eth_dev, bool flg)
 	return 0;
 }
 
-/* Update MTU via vport-update without doing port restart.
- * The vport must be deactivated before calling this API.
- */
-int qede_update_mtu(struct rte_eth_dev *eth_dev, uint16_t mtu)
-{
-	struct qede_dev *qdev = QEDE_INIT_QDEV(eth_dev);
-	struct ecore_dev *edev = QEDE_INIT_EDEV(qdev);
-	struct ecore_sp_vport_update_params params;
-	struct ecore_hwfn *p_hwfn;
-	int rc;
-	int i;
-
-	memset(&params, 0, sizeof(struct ecore_sp_vport_update_params));
-	params.vport_id = 0;
-	params.mtu = mtu;
-	params.vport_id = 0;
-	for_each_hwfn(edev, i) {
-		p_hwfn = &edev->hwfns[i];
-		params.opaque_fid = p_hwfn->hw_info.opaque_fid;
-		rc = ecore_sp_vport_update(p_hwfn, &params,
-				ECORE_SPQ_MODE_EBLOCK, NULL);
-		if (rc != ECORE_SUCCESS) {
-			DP_ERR(edev, "Failed to update MTU\n");
-			return -1;
-		}
-	}
-	DP_INFO(edev, "MTU updated to %u\n", mtu);
-
-	return 0;
-}
-
 static void qede_set_ucast_cmn_params(struct ecore_filter_ucast *ucast)
 {
 	memset(ucast, 0, sizeof(struct ecore_filter_ucast));
@@ -1182,6 +1151,12 @@ static int qede_dev_start(struct rte_eth_dev *eth_dev)
 
 	PMD_INIT_FUNC_TRACE(edev);
 
+	/* Update MTU only if it has changed */
+	if (eth_dev->data->mtu != qdev->mtu) {
+		if (qede_update_mtu(eth_dev, qdev->mtu))
+			goto err;
+	}
+
 	/* Configure TPA parameters */
 	if (rxmode->enable_lro) {
 		if (qede_enable_tpa(eth_dev, true))
@@ -1932,6 +1907,72 @@ qede_set_mc_addr_list(struct rte_eth_dev *eth_dev, struct ether_addr *mc_addrs,
 	return qede_add_mcast_filters(eth_dev, mc_addrs, mc_addrs_num);
 }
 
+/* Update MTU via vport-update without doing port restart.
+ * The vport must be deactivated before calling this API.
+ */
+int qede_update_mtu(struct rte_eth_dev *eth_dev, uint16_t mtu)
+{
+	struct qede_dev *qdev = QEDE_INIT_QDEV(eth_dev);
+	struct ecore_dev *edev = QEDE_INIT_EDEV(qdev);
+	struct ecore_hwfn *p_hwfn;
+	int rc;
+	int i;
+
+	if (IS_PF(edev)) {
+		struct ecore_sp_vport_update_params params;
+
+		memset(&params, 0, sizeof(struct ecore_sp_vport_update_params));
+		params.vport_id = 0;
+		params.mtu = mtu;
+		params.vport_id = 0;
+		for_each_hwfn(edev, i) {
+			p_hwfn = &edev->hwfns[i];
+			params.opaque_fid = p_hwfn->hw_info.opaque_fid;
+			rc = ecore_sp_vport_update(p_hwfn, &params,
+					ECORE_SPQ_MODE_EBLOCK, NULL);
+			if (rc != ECORE_SUCCESS)
+				goto err;
+		}
+	} else {
+		for_each_hwfn(edev, i) {
+			p_hwfn = &edev->hwfns[i];
+			rc = ecore_vf_pf_update_mtu(p_hwfn, mtu);
+			if (rc == ECORE_INVAL) {
+				DP_INFO(edev, "VF MTU Update TLV not supported\n");
+				/* Recreate vport */
+				rc = qede_start_vport(qdev, mtu);
+				if (rc != ECORE_SUCCESS)
+					goto err;
+
+				/* Restore config lost due to vport stop */
+				qede_mac_addr_set(eth_dev, &qdev->primary_mac);
+
+				if (eth_dev->data->promiscuous)
+					qede_promiscuous_enable(eth_dev);
+				else
+					qede_promiscuous_disable(eth_dev);
+
+				if (eth_dev->data->all_multicast)
+					qede_allmulticast_enable(eth_dev);
+				else
+					qede_allmulticast_disable(eth_dev);
+
+				qede_vlan_offload_set(eth_dev,
+						      qdev->vlan_offload_mask);
+			} else if (rc != ECORE_SUCCESS) {
+				goto err;
+			}
+		}
+	}
+	DP_INFO(edev, "%s MTU updated to %u\n", IS_PF(edev) ? "PF" : "VF", mtu);
+
+	return 0;
+
+err:
+	DP_ERR(edev, "Failed to update MTU\n");
+	return -1;
+}
+
 static int qede_flow_ctrl_set(struct rte_eth_dev *eth_dev,
 			      struct rte_eth_fc_conf *fc_conf)
 {
@@ -2337,7 +2378,6 @@ static int qede_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
 			qede_mac_addr_remove(dev, 0);
 	}
 	rte_delay_ms(1000);
-	qede_start_vport(qdev, mtu); /* Recreate vport */
 	qdev->mtu = mtu;
 
 	/* Fix up RX buf size for all queues of the port */
@@ -2361,22 +2401,6 @@ static int qede_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
 	else
 		dev->data->dev_conf.rxmode.jumbo_frame = 0;
 
-	/* Restore config lost due to vport stop */
-	if (IS_PF(edev))
-		qede_mac_addr_set(dev, &qdev->primary_mac);
-
-	if (dev->data->promiscuous)
-		qede_promiscuous_enable(dev);
-	else
-		qede_promiscuous_disable(dev);
-
-	if (dev->data->all_multicast)
-		qede_allmulticast_enable(dev);
-	else
-		qede_allmulticast_disable(dev);
-
-	qede_vlan_offload_set(dev, qdev->vlan_offload_mask);
-
 	if (!dev->data->dev_started && restart) {
 		qede_dev_start(dev);
 		dev->data->dev_started = 1;
diff --git a/drivers/net/qede/qede_ethdev.h b/drivers/net/qede/qede_ethdev.h
index 2145aa6d7..cc1a409f6 100644
--- a/drivers/net/qede/qede_ethdev.h
+++ b/drivers/net/qede/qede_ethdev.h
@@ -34,6 +34,7 @@
 #include "base/nvm_cfg.h"
 #include "base/ecore_sp_commands.h"
 #include "base/ecore_l2.h"
+#include "base/ecore_vf.h"
 
 #include "qede_logs.h"
 #include "qede_if.h"
-- 
2.11.0

[dpdk-stable] patch 'net/mvpp2: check pointer before using it' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From e9f845f4a3f625aedbba64e6688328560a1f8d28 Mon Sep 17 00:00:00 2001
From: Tomasz Duszynski <tdu@semihalf.com>
Date: Mon, 4 Jun 2018 11:24:28 +0200
Subject: [PATCH] net/mvpp2: check pointer before using it

[ upstream commit 42ab8427d4c67532df0b91b8e55e695f8c4f3bb6 ]

Avoid NULL dereference by checking pointer before using it.

Note that the fixes tag is before the driver was renamed
to mvpp2, so the patch will not apply that far back.

Fixes: 3588aaa68eab ("net/mrvl: fix HIF objects allocation")

Signed-off-by: Tomasz Duszynski <tdu@semihalf.com>
---
 drivers/net/mrvl/mrvl_ethdev.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mrvl/mrvl_ethdev.c b/drivers/net/mrvl/mrvl_ethdev.c
index d7e9cefaf..2fc247855 100644
--- a/drivers/net/mrvl/mrvl_ethdev.c
+++ b/drivers/net/mrvl/mrvl_ethdev.c
@@ -1375,9 +1375,12 @@ mrvl_rx_queue_release(void *rxq)
 	if (core_id == LCORE_ID_ANY)
 		core_id = 0;
 
+	if (!q)
+		return;
+
 	hif = mrvl_get_hif(q->priv, core_id);
 
-	if (!q || !hif)
+	if (!hif)
 		return;
 
 	tc = q->priv->rxq_map[q->queue_id].tc;
-- 
2.11.0

[dpdk-stable] patch 'net/qede: fix link change event notification' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From e8ec136e04d8971a9211b7f70635c4b33070e1fb Mon Sep 17 00:00:00 2001
From: Shahed Shaikh <shahed.shaikh@cavium.com>
Date: Thu, 7 Jun 2018 09:30:52 -0700
Subject: [PATCH] net/qede: fix link change event notification

[ backported from upstream commit cfab13aa93897524a4c8d4f6e30d91e7dd9e247b ]

As per existing behavior, when firmware sends a link change
notification, PMD only updates the link structure but does
not notify applications about it.
This results in application sending packets even when link
status is down.

Fix this issue by issuing RTE_ETH_EVENT_INTR_LSC.

Fixes: ec94dbc57362 ("qede: add base driver")

Signed-off-by: Shahed Shaikh <shahed.shaikh@cavium.com>
---
 drivers/net/qede/qede_main.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/qede/qede_main.c b/drivers/net/qede/qede_main.c
index ae1873215..373139d7a 100644
--- a/drivers/net/qede/qede_main.c
+++ b/drivers/net/qede/qede_main.c
@@ -633,8 +633,11 @@ void qed_link_update(struct ecore_hwfn *hwfn)
 {
 	struct ecore_dev *edev = hwfn->p_dev;
 	struct qede_dev *qdev = (struct qede_dev *)edev;
+	struct rte_eth_dev *dev = (struct rte_eth_dev *)qdev->ethdev;
 
-	qede_link_update((struct rte_eth_dev *)qdev->ethdev, 0);
+	if (!qede_link_update(dev, 0))
+		_rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_LSC,
+					      NULL, NULL);
 }
 
 static int qed_drain(struct ecore_dev *edev)
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: add missing ids in xstats' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 20a03b31648fb2c67916452e1ff23a3b589d972a Mon Sep 17 00:00:00 2001
From: David Marchand <david.marchand@6wind.com>
Date: Thu, 7 Jun 2018 07:36:05 +0200
Subject: [PATCH] net/bnxt: add missing ids in xstats

[ upstream commit 6765998874ec7e1068b8646559e2f92ec0d3f308 ]

The xstats API expects that the driver fills both values and ids for
each filled entries.

Fixes: bfb9c2260be2 ("net/bnxt: support xstats get/reset")

Signed-off-by: David Marchand <david.marchand@6wind.com>
Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_stats.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/bnxt/bnxt_stats.c b/drivers/net/bnxt/bnxt_stats.c
index fe83d370a..f8bb4ed9e 100644
--- a/drivers/net/bnxt/bnxt_stats.c
+++ b/drivers/net/bnxt/bnxt_stats.c
@@ -296,6 +296,7 @@ int bnxt_dev_xstats_get_op(struct rte_eth_dev *eth_dev,
 	count = 0;
 	for (i = 0; i < RTE_DIM(bnxt_rx_stats_strings); i++) {
 		uint64_t *rx_stats = (uint64_t *)bp->hw_rx_port_stats;
+		xstats[count].id = count;
 		xstats[count].value = rte_le_to_cpu_64(
 				*(uint64_t *)((char *)rx_stats +
 				bnxt_rx_stats_strings[i].offset));
@@ -304,6 +305,7 @@ int bnxt_dev_xstats_get_op(struct rte_eth_dev *eth_dev,
 
 	for (i = 0; i < RTE_DIM(bnxt_tx_stats_strings); i++) {
 		uint64_t *tx_stats = (uint64_t *)bp->hw_tx_port_stats;
+		xstats[count].id = count;
 		xstats[count].value = rte_le_to_cpu_64(
 				 *(uint64_t *)((char *)tx_stats +
 				bnxt_tx_stats_strings[i].offset));
@@ -311,6 +313,7 @@ int bnxt_dev_xstats_get_op(struct rte_eth_dev *eth_dev,
 	}
 
 	/* The Tx drop pkts aka the Anti spoof coounter */
+	xstats[count].id = count;
 	xstats[count].value = rte_le_to_cpu_64(tx_drop_pkts);
 	count++;
 
-- 
2.11.0

[dpdk-stable] patch 'net/ena: fix GENMASK_ULL macro' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 3e0b04082d3afd012190ae8172a3c59017970bbb Mon Sep 17 00:00:00 2001
From: Rafal Kozik <rk@semihalf.com>
Date: Thu, 7 Jun 2018 11:43:20 +0200
Subject: [PATCH] net/ena: fix GENMASK_ULL macro

[ upstream commit 7544aee8d0b4ae0262b1ba7e1539cf8171664df7 ]

When use GENMASK_ULL(63,0) left shift by 64 bits is performed.
Shifting by number greater or equal then word length
is undefined operation and failed on some platforms.

Fixes: 9ba7981ec992 ("ena: add communication layer for DPDK")

Signed-off-by: Rafal Kozik <rk@semihalf.com>
Acked-by: Michal Krawczyk <mk@semihalf.com>
---
 drivers/net/ena/base/ena_plat_dpdk.h | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ena/base/ena_plat_dpdk.h b/drivers/net/ena/base/ena_plat_dpdk.h
index accecf518..da89bac94 100644
--- a/drivers/net/ena/base/ena_plat_dpdk.h
+++ b/drivers/net/ena/base/ena_plat_dpdk.h
@@ -116,11 +116,13 @@ typedef uint64_t dma_addr_t;
 #define ENA_MIN16(x, y) RTE_MIN((x), (y))
 #define ENA_MIN8(x, y) RTE_MIN((x), (y))
 
+#define BITS_PER_LONG_LONG (__SIZEOF_LONG_LONG__ * 8)
 #define U64_C(x) x ## ULL
 #define BIT(nr)         (1UL << (nr))
 #define BITS_PER_LONG	(__SIZEOF_LONG__ * 8)
 #define GENMASK(h, l)	(((~0UL) << (l)) & (~0UL >> (BITS_PER_LONG - 1 - (h))))
-#define GENMASK_ULL(h, l) (((U64_C(1) << ((h) - (l) + 1)) - 1) << (l))
+#define GENMASK_ULL(h, l) (((~0ULL) - (1ULL << (l)) + 1) & \
+			  (~0ULL >> (BITS_PER_LONG_LONG - 1 - (h))))
 
 #ifdef RTE_LIBRTE_ENA_COM_DEBUG
 #define ena_trc_dbg(format, arg...)					\
-- 
2.11.0

[dpdk-stable] patch 'net/nfp: fix field initialization in Tx descriptor' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From e04a4d32da10a7362e538de821cd7fe411f027f2 Mon Sep 17 00:00:00 2001
From: Alejandro Lucero <alejandro.lucero@netronome.com>
Date: Mon, 11 Jun 2018 10:50:37 +0100
Subject: [PATCH] net/nfp: fix field initialization in Tx descriptor

[ upstream commit 01f65293ceb94921a35554fa7b17bc0ec1764c96 ]

TX descriptor eop_offset field is not initialized and it could
contain garbage. This patch fixes the potential problem setting
EOP as the only subfield. The other subfield, data offset, is
not used by now.

Fixes: b812daadad0d ("nfp: add Rx and Tx")

Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>
---
 drivers/net/nfp/nfp_net.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/net/nfp/nfp_net.c b/drivers/net/nfp/nfp_net.c
index d9cd04732..96484d95e 100644
--- a/drivers/net/nfp/nfp_net.c
+++ b/drivers/net/nfp/nfp_net.c
@@ -2292,11 +2292,15 @@ nfp_net_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 				txq->wr_p = 0;
 
 			pkt_size -= dma_size;
-			if (!pkt_size)
-				/* End of packet */
-				txds->offset_eop |= PCIE_DESC_TX_EOP;
+
+			/*
+			 * Making the EOP, packets with just one segment
+			 * the priority
+			 */
+			if (likely(!pkt_size))
+				txds->offset_eop = PCIE_DESC_TX_EOP;
 			else
-				txds->offset_eop &= PCIE_DESC_TX_OFFSET_MASK;
+				txds->offset_eop = 0;
 
 			pkt = pkt->next;
 			/* Referencing next free TX descriptor */
-- 
2.11.0

[dpdk-stable] patch 'net/bonding: always update bonding link status' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From d4ee3056efafaca957b9d9e23c9fc8703bd1c76f Mon Sep 17 00:00:00 2001
From: Chas Williams <chas3@att.com>
Date: Thu, 17 May 2018 19:22:52 -0400
Subject: [PATCH] net/bonding: always update bonding link status

[ upstream commit 0e677a35349f316d50ba25485354363f3e65c12e ]

When the first ETH_LINK_UP slave is added to a bonding device, it is
immediately activated.  If these are polling slaves, there will be a
later link state event, when last_link_status doesn't match link_status.
But because this slave is already activated, the bonding device's link
status isn't updated.

To fix this, we always attempt to update the bonding device's link status
in the link state change callback.

Fixes: 46fb43683679 ("bond: add mode 4")

Signed-off-by: Chas Williams <chas3@att.com>
---
 drivers/net/bonding/rte_eth_bond_pmd.c | 15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)

diff --git a/drivers/net/bonding/rte_eth_bond_pmd.c b/drivers/net/bonding/rte_eth_bond_pmd.c
index e19a4a3e8..0cbe08ada 100644
--- a/drivers/net/bonding/rte_eth_bond_pmd.c
+++ b/drivers/net/bonding/rte_eth_bond_pmd.c
@@ -2535,10 +2535,8 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type,
 
 	rte_eth_link_get_nowait(port_id, &link);
 	if (link.link_status) {
-		if (active_pos < internals->active_slave_count) {
-			rte_spinlock_unlock(&internals->lsc_lock);
-			return rc;
-		}
+		if (active_pos < internals->active_slave_count)
+			goto link_update;
 
 		/* if no active slave ports then set this port to be primary port */
 		if (internals->active_slave_count < 1) {
@@ -2557,10 +2555,8 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type,
 				internals->primary_port == port_id)
 			bond_ethdev_primary_set(internals, port_id);
 	} else {
-		if (active_pos == internals->active_slave_count) {
-			rte_spinlock_unlock(&internals->lsc_lock);
-			return rc;
-		}
+		if (active_pos == internals->active_slave_count)
+			goto link_update;
 
 		/* Remove from active slave list */
 		deactivate_slave(bonded_eth_dev, port_id);
@@ -2579,6 +2575,7 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type,
 		}
 	}
 
+link_update:
 	/**
 	 * Update bonded device link properties after any change to active
 	 * slaves
@@ -2616,7 +2613,7 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type,
 
 	rte_spinlock_unlock(&internals->lsc_lock);
 
-	return 0;
+	return rc;
 }
 
 static int
-- 
2.11.0

[dpdk-stable] patch 'net/bonding: fix MAC address reset' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 349e12f146cce2400dd81220f622959b06cfe710 Mon Sep 17 00:00:00 2001
From: Kiran Kumar <kiran.kokkilagadda@caviumnetworks.com>
Date: Wed, 23 May 2018 14:41:21 +0530
Subject: [PATCH] net/bonding: fix MAC address reset

[ upstream commit fb6eb1e69c499feb5f36193b906c910553835cc6 ]

Currently when resetting bond mac address, we are getting the
persisted mac address from slave info considering primary port
as index. But we need to compare the port id from slave info
with the primary port to get the primary slave index and get
the persisted mac address. Without this fix, persisted mac addr
will be zero and rte_eth_dev_default_mac_addr_set will fail.

Fixes: a45b288ef21a ("bond: support link status polling")

Signed-off-by: Kiran Kumar <kiran.kokkilagadda@caviumnetworks.com>
---
 drivers/net/bonding/rte_eth_bond_api.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

diff --git a/drivers/net/bonding/rte_eth_bond_api.c b/drivers/net/bonding/rte_eth_bond_api.c
index 8fd90ae9f..e80338aa2 100644
--- a/drivers/net/bonding/rte_eth_bond_api.c
+++ b/drivers/net/bonding/rte_eth_bond_api.c
@@ -675,9 +675,21 @@ rte_eth_bond_mac_address_reset(uint16_t bonded_port_id)
 	internals->user_defined_mac = 0;
 
 	if (internals->slave_count > 0) {
+		int slave_port;
+		/* Get the primary slave location based on the primary port
+		 * number as, while slave_add(), we will keep the primary
+		 * slave based on slave_count,but not based on the primary port.
+		 */
+		for (slave_port = 0; slave_port < internals->slave_count;
+		     slave_port++) {
+			if (internals->slaves[slave_port].port_id ==
+			    internals->primary_port)
+				break;
+		}
+
 		/* Set MAC Address of Bonded Device */
 		if (mac_address_set(bonded_eth_dev,
-				&internals->slaves[internals->primary_port].persisted_mac_addr)
+			&internals->slaves[slave_port].persisted_mac_addr)
 				!= 0) {
 			RTE_BOND_LOG(ERR, "Failed to set MAC address on bonded device");
 			return -1;
-- 
2.11.0

[dpdk-stable] patch 'net/mlx5: fix crash in device probe' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From a1eef40349f4bf5156334b554baf0c7189d41183 Mon Sep 17 00:00:00 2001
From: Xueming Li <xuemingl@mellanox.com>
Date: Tue, 12 Jun 2018 19:38:11 +0800
Subject: [PATCH] net/mlx5: fix crash in device probe

[ upstream commit a9fc0b0ef0c276fcea082b20574e27cca1bac3e8 ]

This patch initializes counter descriptor struct before invoking Verbs
api to avoid segmentation fault.

Fixes: 9a761de8ea14 ("net/mlx5: flow counter support")

Signed-off-by: Xueming Li <xuemingl@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
---
 drivers/net/mlx5/mlx5.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5.c b/drivers/net/mlx5/mlx5.c
index 10ce33592..76f76c41b 100644
--- a/drivers/net/mlx5/mlx5.c
+++ b/drivers/net/mlx5/mlx5.c
@@ -680,7 +680,7 @@ mlx5_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
 	int i;
 	struct mlx5dv_context attrs_out;
 #ifdef HAVE_IBV_DEVICE_COUNTERS_SET_SUPPORT
-	struct ibv_counter_set_description cs_desc;
+	struct ibv_counter_set_description cs_desc = { .counter_type = 0 };
 #endif
 
 	assert(pci_drv == &mlx5_driver);
-- 
2.11.0

[dpdk-stable] patch 'eventdev: fix port in Rx adapter internal function' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 2725dbd296681c31ac6af3c87c9f5428e83b1977 Mon Sep 17 00:00:00 2001
From: Nikhil Rao <nikhil.rao@intel.com>
Date: Wed, 23 May 2018 02:36:42 +0530
Subject: [PATCH] eventdev: fix port in Rx adapter internal function

[ upstream commit d89fd4cd132720fea1d9497a375d50489e7dd0b7 ]

The dev_id parameter to fill_event_buffer() should be 16 bit,
also rename to to eth_dev_id to avoid confusion with event device
id elsewhere in the file.

Fixes: c2189c907dd1 ("eventdev: make ethdev port identifiers 16-bit")

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
---
 lib/librte_eventdev/rte_event_eth_rx_adapter.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.c b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
index 90106e6c9..f62972545 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
@@ -414,14 +414,14 @@ flush_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter)
 
 static inline void
 fill_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter,
-	uint8_t dev_id,
+	uint16_t eth_dev_id,
 	uint16_t rx_queue_id,
 	struct rte_mbuf **mbufs,
 	uint16_t num)
 {
 	uint32_t i;
 	struct eth_device_info *eth_device_info =
-					&rx_adapter->eth_devices[dev_id];
+					&rx_adapter->eth_devices[eth_dev_id];
 	struct eth_rx_queue_info *eth_rx_queue_info =
 					&eth_device_info->rx_queue[rx_queue_id];
 
-- 
2.11.0

[dpdk-stable] patch 'eventdev: fix missing update to Rx adaper WRR position' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From cff559723897834275579fce43e5ff8b420a8eff Mon Sep 17 00:00:00 2001
From: Nikhil Rao <nikhil.rao@intel.com>
Date: Sun, 3 Jun 2018 17:11:08 +0530
Subject: [PATCH] eventdev: fix missing update to Rx adaper WRR position

[ upstream commit 524dc20023b2e00cd9a021ae60cb4b91f51786e6 ]

After dequeuing Rx packets and enqueueing them to the
temporary buffer towards eventdev, the packet Rx loop exits
if the temporary buffer is full but the current WRR position
is not saved.

Save away the current value of the WRR position, so packets
are dequeued from the correct Rx queue in the next invocation.

Fixes: 9c38b704d280 ("eventdev: add eth Rx adapter implementation")

Suggested-by: Gage Eads <gage.eads@intel.com>
Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
---
 lib/librte_eventdev/rte_event_eth_rx_adapter.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.c b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
index f62972545..1cdbb848b 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
@@ -503,8 +503,10 @@ eth_rx_poll(struct rte_event_eth_rx_adapter *rx_adapter)
 		 */
 		if (buf->count >= BATCH_SIZE)
 			flush_event_buffer(rx_adapter);
-		if (BATCH_SIZE > (ETH_EVENT_BUFFER_SIZE - buf->count))
+		if (BATCH_SIZE > (ETH_EVENT_BUFFER_SIZE - buf->count)) {
+			rx_adapter->wrr_pos = wrr_pos;
 			break;
+		}
 
 		stats->rx_poll_count++;
 		n = rte_eth_rx_burst(d, qid, mbufs, BATCH_SIZE);
-- 
2.11.0

[dpdk-stable] patch 'eventdev: add event buffer flush in Rx adapter' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 9b4c8ab2cfbe9ed232c417cdc4a78f66e12374a5 Mon Sep 17 00:00:00 2001
From: Nikhil Rao <nikhil.rao@intel.com>
Date: Sun, 3 Jun 2018 18:12:25 +0530
Subject: [PATCH] eventdev: add event buffer flush in Rx adapter

[ upstream commit 6b83f59355437c0631a64e5ecb9f080c17a8ba24 ]

Add an event buffer flush when the current invocation
of the Rx adapter is completed.

This patch provides lower latency in case there is a
BATCH_SIZE of events in the event buffer.

Suggested-by: Narender Vangati <narender.vangati@intel.com>
Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
---
 lib/librte_eventdev/rte_event_eth_rx_adapter.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.c b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
index 1cdbb848b..377db42d9 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
@@ -476,7 +476,7 @@ fill_event_buffer(struct rte_event_eth_rx_adapter *rx_adapter,
  * the hypervisor's switching layer where adjustments can be made to deal with
  * it.
  */
-static inline uint32_t
+static inline void
 eth_rx_poll(struct rte_event_eth_rx_adapter *rx_adapter)
 {
 	uint32_t num_queue;
@@ -505,7 +505,7 @@ eth_rx_poll(struct rte_event_eth_rx_adapter *rx_adapter)
 			flush_event_buffer(rx_adapter);
 		if (BATCH_SIZE > (ETH_EVENT_BUFFER_SIZE - buf->count)) {
 			rx_adapter->wrr_pos = wrr_pos;
-			break;
+			return;
 		}
 
 		stats->rx_poll_count++;
@@ -521,7 +521,7 @@ eth_rx_poll(struct rte_event_eth_rx_adapter *rx_adapter)
 			if (nb_rx > max_nb_rx) {
 				rx_adapter->wrr_pos =
 				    (wrr_pos + 1) % rx_adapter->wrr_len;
-				return nb_rx;
+				break;
 			}
 		}
 
@@ -529,20 +529,18 @@ eth_rx_poll(struct rte_event_eth_rx_adapter *rx_adapter)
 			wrr_pos = 0;
 	}
 
-	return nb_rx;
+	if (buf->count >= BATCH_SIZE)
+		flush_event_buffer(rx_adapter);
 }
 
 static int
 event_eth_rx_adapter_service_func(void *args)
 {
 	struct rte_event_eth_rx_adapter *rx_adapter = args;
-	struct rte_eth_event_enqueue_buffer *buf;
 
-	buf = &rx_adapter->event_enqueue_buffer;
 	if (rte_spinlock_trylock(&rx_adapter->rx_lock) == 0)
 		return 0;
-	if (eth_rx_poll(rx_adapter) == 0 && buf->count)
-		flush_event_buffer(rx_adapter);
+	eth_rx_poll(rx_adapter);
 	rte_spinlock_unlock(&rx_adapter->rx_lock);
 	return 0;
 }
-- 
2.11.0

[dpdk-stable] patch 'eventdev: fix internal port logic in Rx adapter' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 3fd4954886683fc0b363bdae9a5993120e62786b Mon Sep 17 00:00:00 2001
From: Nikhil Rao <nikhil.rao@intel.com>
Date: Sun, 3 Jun 2018 19:40:46 +0530
Subject: [PATCH] eventdev: fix internal port logic in Rx adapter

[ upstream commit d411c4074fe6d90b695f9d9c4982b8c45f2a675d ]

Set the internal_event_port flag when the ethdev-eventdev
packet transfer is implemented in hardware and add a check
for the flag to ignore the connection when setting up the
WRR polling sequence.

Fixes: 9c38b704d280 ("eventdev: add eth Rx adapter implementation")

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
---
 lib/librte_eventdev/rte_event_eth_rx_adapter.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.c b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
index 377db42d9..e767591e8 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
@@ -220,6 +220,8 @@ eth_poll_wrr_calc(struct rte_event_eth_rx_adapter *rx_adapter)
 			nb_rx_queues = dev_info->dev->data->nb_rx_queues;
 			if (dev_info->rx_queue == NULL)
 				continue;
+			if (dev_info->internal_event_port)
+				continue;
 			for (q = 0; q < nb_rx_queues; q++) {
 				struct eth_rx_queue_info *queue_info =
 					&dev_info->rx_queue[q];
@@ -1032,6 +1034,7 @@ rte_event_eth_rx_adapter_queue_add(uint8_t id,
 				&rte_eth_devices[eth_dev_id],
 				rx_queue_id, queue_conf);
 		if (ret == 0) {
+			dev_info->internal_event_port = 1;
 			update_queue_info(rx_adapter,
 					&rx_adapter->eth_devices[eth_dev_id],
 					rx_queue_id,
@@ -1039,6 +1042,7 @@ rte_event_eth_rx_adapter_queue_add(uint8_t id,
 		}
 	} else {
 		rte_spinlock_lock(&rx_adapter->rx_lock);
+		dev_info->internal_event_port = 0;
 		ret = init_service(rx_adapter, id);
 		if (ret == 0)
 			ret = add_rx_queue(rx_adapter, eth_dev_id, rx_queue_id,
-- 
2.11.0

[dpdk-stable] patch 'eventdev: fix Rx SW adapter stop' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From ca24d9649e81e0be9305897cf744db239117843d Mon Sep 17 00:00:00 2001
From: Nikhil Rao <nikhil.rao@intel.com>
Date: Mon, 4 Jun 2018 18:25:17 +0530
Subject: [PATCH] eventdev: fix Rx SW adapter stop

[ upstream commit a66a83744667f31ee97e18eefa90105c462ed81b ]

The Rx adapter stop call does not guarantee that the
SW service function will not execute after the
rte_event_eth_rx_adapter_stop() call.

Add a "started" flag to prevent the adapter from executing
if stop has been called.

Fixes: 9c38b704d280 ("eventdev: add eth Rx adapter implementation")

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
---
 lib/librte_eventdev/rte_event_eth_rx_adapter.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/lib/librte_eventdev/rte_event_eth_rx_adapter.c b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
index e767591e8..d5c3fd56c 100644
--- a/lib/librte_eventdev/rte_event_eth_rx_adapter.c
+++ b/lib/librte_eventdev/rte_event_eth_rx_adapter.c
@@ -87,6 +87,8 @@ struct rte_event_eth_rx_adapter {
 	int socket_id;
 	/* Per adapter EAL service */
 	uint32_t service_id;
+	/* Adapter started flag */
+	uint8_t rxa_started;
 } __rte_cache_aligned;
 
 /* Per eth device */
@@ -542,6 +544,10 @@ event_eth_rx_adapter_service_func(void *args)
 
 	if (rte_spinlock_trylock(&rx_adapter->rx_lock) == 0)
 		return 0;
+	if (!rx_adapter->rxa_started) {
+		return 0;
+		rte_spinlock_unlock(&rx_adapter->rx_lock);
+	}
 	eth_rx_poll(rx_adapter);
 	rte_spinlock_unlock(&rx_adapter->rx_lock);
 	return 0;
@@ -831,8 +837,12 @@ rx_adapter_ctrl(uint8_t id, int start)
 						&rte_eth_devices[i]);
 	}
 
-	if (use_service)
+	if (use_service) {
+		rte_spinlock_lock(&rx_adapter->rx_lock);
+		rx_adapter->rxa_started = start;
 		rte_service_runstate_set(rx_adapter->service_id, start);
+		rte_spinlock_unlock(&rx_adapter->rx_lock);
+	}
 
 	return 0;
 }
-- 
2.11.0

[dpdk-stable] patch 'bus/dpaa: fix build' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 464972d7489cc6393c4a3769bc26d7753108f01d Mon Sep 17 00:00:00 2001
From: Thomas Monjalon <thomas@monjalon.net>
Date: Wed, 20 Jun 2018 12:38:06 +0200
Subject: [PATCH] bus/dpaa: fix build

[ upstream commit ce819b84220ef4abe87cccc71c78e7f4bc7a2f8b ]

The DPAA bus driver is defining some macros without prefix.
So it can conflict with other libraries like libbsd:

	drivers/bus/dpaa/include/compat.h:53:
		error: "__packed" redefined
	/usr/include/bsd/sys/cdefs.h:120:
		note: this is the location of the previous definition

Fixes: 39f373cf015a ("bus/dpaa: add compatibility and helper macros")

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
---
 drivers/bus/dpaa/include/compat.h | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/bus/dpaa/include/compat.h b/drivers/bus/dpaa/include/compat.h
index 42733aeb1..1f56a4b51 100644
--- a/drivers/bus/dpaa/include/compat.h
+++ b/drivers/bus/dpaa/include/compat.h
@@ -80,9 +80,15 @@
  */
 
 /* Required compiler attributes */
+#ifndef __maybe_unused
 #define __maybe_unused	__rte_unused
+#endif
+#ifndef __always_unused
 #define __always_unused	__rte_unused
+#endif
+#ifndef __packed
 #define __packed	__rte_packed
+#endif
 #define noinline	__attribute__((noinline))
 
 #define L1_CACHE_BYTES 64
-- 
2.11.0

[dpdk-stable] patch 'kni: fix build with gcc 8.1' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From e9106897f3d7ca13353df6ddea61f1c68a829ad3 Mon Sep 17 00:00:00 2001
From: Ferruh Yigit <ferruh.yigit@intel.com>
Date: Tue, 26 Jun 2018 12:38:07 +0100
Subject: [PATCH] kni: fix build with gcc 8.1
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

[ upstream commit a36b40dd8ecfb61648045e418551f75055ca77f9 ]

Error observed when CONFIG_RTE_KNI_KMOD_ETHTOOL config option is
enabled.

build error:
In function ‘strncpy’,
    inlined from ‘igb_get_drvinfo’ at
    kernel/linux/kni/igb_ethtool.c:814:2:
    include/linux/string.h:246:9: error: ‘__builtin_strncpy’ output
    may be truncated copying 31 bytes from a string of length 42
    [-Werror=stringop-truncation]
      return __builtin_strncpy(p, q, size);
               ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixed by using strlcpy instead of strncpy.

adapter->fw_version size kept same because of
c3698192940c ("kni: fix build with gcc 7.1")

Also next line strncpy usage replaced with strlcpy while arround.

Fixes: c3698192940c ("kni: fix build with gcc 7.1")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
---
 lib/librte_eal/linuxapp/kni/ethtool/igb/igb_ethtool.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/lib/librte_eal/linuxapp/kni/ethtool/igb/igb_ethtool.c b/lib/librte_eal/linuxapp/kni/ethtool/igb/igb_ethtool.c
index 95e262b78..aed14bcc8 100644
--- a/lib/librte_eal/linuxapp/kni/ethtool/igb/igb_ethtool.c
+++ b/lib/librte_eal/linuxapp/kni/ethtool/igb/igb_ethtool.c
@@ -826,9 +826,10 @@ static void igb_get_drvinfo(struct net_device *netdev,
 	strncpy(drvinfo->driver,  igb_driver_name, sizeof(drvinfo->driver) - 1);
 	strncpy(drvinfo->version, igb_driver_version, sizeof(drvinfo->version) - 1);
 
-	strncpy(drvinfo->fw_version, adapter->fw_version,
-		sizeof(drvinfo->fw_version) - 1);
-	strncpy(drvinfo->bus_info, pci_name(adapter->pdev), sizeof(drvinfo->bus_info) -1);
+	strlcpy(drvinfo->fw_version, adapter->fw_version,
+		sizeof(drvinfo->fw_version));
+	strlcpy(drvinfo->bus_info, pci_name(adapter->pdev),
+		sizeof(drvinfo->bus_info));
 	drvinfo->n_stats = IGB_STATS_LEN;
 	drvinfo->testinfo_len = IGB_TEST_LEN;
 	drvinfo->regdump_len = igb_get_regs_len(netdev);
-- 
2.11.0

[dpdk-stable] patch 'net/ixgbe: add support for VLAN in IP mode FDIR' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From a6db82e7ac8452c9cf70aedd167511374bb42494 Mon Sep 17 00:00:00 2001
From: Wei Zhao <wei.zhao1@intel.com>
Date: Wed, 13 Jun 2018 16:09:55 +0800
Subject: [PATCH] net/ixgbe: add support for VLAN in IP mode FDIR

[ upstream commit 2ee9fecfc3f449d6c2596cf34517d5728b541ee9 ]

In IP mode FDIR, X550 can support not only 4 tuple parameters
but also vlan tci in protocol, so add this feature to flow parser.

Fixes: 11777435c727 ("net/ixgbe: parse flow director filter")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
---
 drivers/net/ixgbe/ixgbe_flow.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ixgbe/ixgbe_flow.c b/drivers/net/ixgbe/ixgbe_flow.c
index 07abb3437..c6d952a0a 100644
--- a/drivers/net/ixgbe/ixgbe_flow.c
+++ b/drivers/net/ixgbe/ixgbe_flow.c
@@ -1665,7 +1665,8 @@ ixgbe_parse_fdir_filter_normal(struct rte_eth_dev *dev,
 				return -rte_errno;
 			}
 		} else {
-			if (item->type != RTE_FLOW_ITEM_TYPE_IPV4) {
+			if (item->type != RTE_FLOW_ITEM_TYPE_IPV4 &&
+					item->type != RTE_FLOW_ITEM_TYPE_VLAN) {
 				memset(rule, 0, sizeof(struct ixgbe_fdir_rule));
 				rte_flow_error_set(error, EINVAL,
 					RTE_FLOW_ERROR_TYPE_ITEM,
-- 
2.11.0

[dpdk-stable] patch 'net/ixgbe: fix tunnel id format error for FDIR' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 33f522cf34de6755eee382e532b43371401defad Mon Sep 17 00:00:00 2001
From: Wei Zhao <wei.zhao1@intel.com>
Date: Wed, 13 Jun 2018 16:11:22 +0800
Subject: [PATCH] net/ixgbe: fix tunnel id format error for FDIR

[ upstream commit e7d406851909b1c3cc24ee8c61044dccf9902559 ]

In cloud mode for FDIR, tunnel id should be set as protocol
request, the lower 8 bits should be set as reserved.

Fixes: 82fb702077f6 ("ixgbe: support new flow director modes for X550")
Fixes: 11777435c727 ("net/ixgbe: parse flow director filter")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
---
 drivers/net/ixgbe/ixgbe_fdir.c | 2 +-
 drivers/net/ixgbe/ixgbe_flow.c | 5 +----
 2 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_fdir.c b/drivers/net/ixgbe/ixgbe_fdir.c
index c1176472c..32720abc9 100644
--- a/drivers/net/ixgbe/ixgbe_fdir.c
+++ b/drivers/net/ixgbe/ixgbe_fdir.c
@@ -803,7 +803,7 @@ ixgbe_fdir_filter_to_atr_input(const struct rte_eth_fdir_filter *fdir_filter,
 		input->formatted.tunnel_type =
 			fdir_filter->input.flow.tunnel_flow.tunnel_type;
 		input->formatted.tni_vni =
-			fdir_filter->input.flow.tunnel_flow.tunnel_id;
+			fdir_filter->input.flow.tunnel_flow.tunnel_id >> 8;
 	}
 
 	return 0;
diff --git a/drivers/net/ixgbe/ixgbe_flow.c b/drivers/net/ixgbe/ixgbe_flow.c
index c6d952a0a..13cc121c1 100644
--- a/drivers/net/ixgbe/ixgbe_flow.c
+++ b/drivers/net/ixgbe/ixgbe_flow.c
@@ -2423,10 +2423,8 @@ ixgbe_parse_fdir_filter_tunnel(const struct rte_flow_attr *attr,
 			vxlan_spec = (const struct rte_flow_item_vxlan *)
 					item->spec;
 			rte_memcpy(((uint8_t *)
-				&rule->ixgbe_fdir.formatted.tni_vni + 1),
+				&rule->ixgbe_fdir.formatted.tni_vni),
 				vxlan_spec->vni, RTE_DIM(vxlan_spec->vni));
-			rule->ixgbe_fdir.formatted.tni_vni = rte_be_to_cpu_32(
-				rule->ixgbe_fdir.formatted.tni_vni);
 		}
 	}
 
@@ -2525,7 +2523,6 @@ ixgbe_parse_fdir_filter_tunnel(const struct rte_flow_attr *attr,
 			/* tni is a 24-bits bit field */
 			rte_memcpy(&rule->ixgbe_fdir.formatted.tni_vni,
 			nvgre_spec->tni, RTE_DIM(nvgre_spec->tni));
-			rule->ixgbe_fdir.formatted.tni_vni <<= 8;
 		}
 	}
 
-- 
2.11.0

[dpdk-stable] patch 'net/ixgbe: fix tunnel type set error for FDIR' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From d05de9eeed2675364772bcf8bcbd77c214578208 Mon Sep 17 00:00:00 2001
From: Wei Zhao <wei.zhao1@intel.com>
Date: Thu, 14 Jun 2018 16:17:28 +0800
Subject: [PATCH] net/ixgbe: fix tunnel type set error for FDIR

[ upstream commit 876b450ed99d2c96c5b71360cc84f7b0b8f506f4 ]

Tunnel type format should be translated to ixgbe required format
before register set in FDIR cloud mode, Ans also some register
not useful in cloud mode but only useful in IP mode should be set
to zero as datasheet request.

Fixes: 82fb702077f6 ("ixgbe: support new flow director modes for X550")
Fixes: 11777435c727 ("net/ixgbe: parse flow director filter")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
---
 drivers/net/ixgbe/ixgbe_ethdev.h |  5 +++++
 drivers/net/ixgbe/ixgbe_fdir.c   | 19 +++++++++++++++----
 drivers/net/ixgbe/ixgbe_flow.c   |  4 ++--
 3 files changed, 22 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_ethdev.h b/drivers/net/ixgbe/ixgbe_ethdev.h
index 51ddcfd4e..762f1ad34 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.h
+++ b/drivers/net/ixgbe/ixgbe_ethdev.h
@@ -123,6 +123,11 @@
 #define IXGBE_5TUPLE_MAX_PRI            7
 #define IXGBE_5TUPLE_MIN_PRI            1
 
+/* bit of VXLAN tunnel type | 7 bits of zeros  | 8 bits of zeros*/
+#define IXGBE_FDIR_VXLAN_TUNNEL_TYPE    0x8000
+/* bit of NVGRE tunnel type | 7 bits of zeros  | 8 bits of zeros*/
+#define IXGBE_FDIR_NVGRE_TUNNEL_TYPE    0x0
+
 #define IXGBE_RSS_OFFLOAD_ALL ( \
 	ETH_RSS_IPV4 | \
 	ETH_RSS_NONFRAG_IPV4_TCP | \
diff --git a/drivers/net/ixgbe/ixgbe_fdir.c b/drivers/net/ixgbe/ixgbe_fdir.c
index 32720abc9..cbe2d6ff8 100644
--- a/drivers/net/ixgbe/ixgbe_fdir.c
+++ b/drivers/net/ixgbe/ixgbe_fdir.c
@@ -800,8 +800,17 @@ ixgbe_fdir_filter_to_atr_input(const struct rte_eth_fdir_filter *fdir_filter,
 			input->formatted.inner_mac,
 			fdir_filter->input.flow.tunnel_flow.mac_addr.addr_bytes,
 			sizeof(input->formatted.inner_mac));
-		input->formatted.tunnel_type =
-			fdir_filter->input.flow.tunnel_flow.tunnel_type;
+		if (fdir_filter->input.flow.tunnel_flow.tunnel_type ==
+				RTE_FDIR_TUNNEL_TYPE_VXLAN)
+			input->formatted.tunnel_type =
+					IXGBE_FDIR_VXLAN_TUNNEL_TYPE;
+		else if (fdir_filter->input.flow.tunnel_flow.tunnel_type ==
+				RTE_FDIR_TUNNEL_TYPE_NVGRE)
+			input->formatted.tunnel_type =
+					IXGBE_FDIR_NVGRE_TUNNEL_TYPE;
+		else
+			PMD_DRV_LOG(ERR, " invalid tunnel type arguments.");
+
 		input->formatted.tni_vni =
 			fdir_filter->input.flow.tunnel_flow.tunnel_id >> 8;
 	}
@@ -1030,8 +1039,7 @@ fdir_write_perfect_filter_82599(struct ixgbe_hw *hw,
 			IXGBE_WRITE_REG(hw, IXGBE_FDIRSIPv6(2), 0);
 		} else {
 			/* tunnel mode */
-			if (input->formatted.tunnel_type !=
-				RTE_FDIR_TUNNEL_TYPE_NVGRE)
+			if (input->formatted.tunnel_type)
 				tunnel_type = 0x80000000;
 			tunnel_type |= addr_high;
 			IXGBE_WRITE_REG(hw, IXGBE_FDIRSIPv6(0), addr_low);
@@ -1039,6 +1047,9 @@ fdir_write_perfect_filter_82599(struct ixgbe_hw *hw,
 			IXGBE_WRITE_REG(hw, IXGBE_FDIRSIPv6(2),
 					input->formatted.tni_vni);
 		}
+		IXGBE_WRITE_REG(hw, IXGBE_FDIRIPSA, 0);
+		IXGBE_WRITE_REG(hw, IXGBE_FDIRIPDA, 0);
+		IXGBE_WRITE_REG(hw, IXGBE_FDIRPORT, 0);
 	}
 
 	/* record vlan (little-endian) and flex_bytes(big-endian) */
diff --git a/drivers/net/ixgbe/ixgbe_flow.c b/drivers/net/ixgbe/ixgbe_flow.c
index 13cc121c1..e60ecce77 100644
--- a/drivers/net/ixgbe/ixgbe_flow.c
+++ b/drivers/net/ixgbe/ixgbe_flow.c
@@ -2371,7 +2371,7 @@ ixgbe_parse_fdir_filter_tunnel(const struct rte_flow_attr *attr,
 	/* Get the VxLAN info */
 	if (item->type == RTE_FLOW_ITEM_TYPE_VXLAN) {
 		rule->ixgbe_fdir.formatted.tunnel_type =
-			RTE_FDIR_TUNNEL_TYPE_VXLAN;
+				IXGBE_FDIR_VXLAN_TUNNEL_TYPE;
 
 		/* Only care about VNI, others should be masked. */
 		if (!item->mask) {
@@ -2431,7 +2431,7 @@ ixgbe_parse_fdir_filter_tunnel(const struct rte_flow_attr *attr,
 	/* Get the NVGRE info */
 	if (item->type == RTE_FLOW_ITEM_TYPE_NVGRE) {
 		rule->ixgbe_fdir.formatted.tunnel_type =
-			RTE_FDIR_TUNNEL_TYPE_NVGRE;
+				IXGBE_FDIR_NVGRE_TUNNEL_TYPE;
 
 		/**
 		 * Only care about flags0, flags1, protocol and TNI,
-- 
2.11.0

[dpdk-stable] patch 'net/ixgbe: fix mask bits register set error for FDIR' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 16783d2af5f07018258c628f7f78bcc0b5a72581 Mon Sep 17 00:00:00 2001
From: Wei Zhao <wei.zhao1@intel.com>
Date: Fri, 15 Jun 2018 14:08:03 +0800
Subject: [PATCH] net/ixgbe: fix mask bits register set error for FDIR

[ upstream commit 5dadd757188e68260880924dea38c22858ac2ca9 ]

MAC address bits in mask registers should be set to zero
when the is mac mask is 0xFF, otherwise if it is 0x0
these bits should be to 0x3F.

Fixes: 82fb702077f6 ("ixgbe: support new flow director modes for X550")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
---
 drivers/net/ixgbe/ixgbe_fdir.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_fdir.c b/drivers/net/ixgbe/ixgbe_fdir.c
index cbe2d6ff8..02adfa437 100644
--- a/drivers/net/ixgbe/ixgbe_fdir.c
+++ b/drivers/net/ixgbe/ixgbe_fdir.c
@@ -423,9 +423,12 @@ fdir_set_input_mask_x550(struct rte_eth_dev *dev)
 				IXGBE_FDIRIP6M_TNI_VNI;
 
 	if (mode == RTE_FDIR_MODE_PERFECT_TUNNEL) {
-		mac_mask = info->mask.mac_addr_byte_mask;
-		fdiripv6m |= (mac_mask << IXGBE_FDIRIP6M_INNER_MAC_SHIFT)
-				& IXGBE_FDIRIP6M_INNER_MAC;
+		fdiripv6m |= IXGBE_FDIRIP6M_INNER_MAC;
+		mac_mask = info->mask.mac_addr_byte_mask &
+			(IXGBE_FDIRIP6M_INNER_MAC >>
+			IXGBE_FDIRIP6M_INNER_MAC_SHIFT);
+		fdiripv6m &= ~((mac_mask << IXGBE_FDIRIP6M_INNER_MAC_SHIFT) &
+				IXGBE_FDIRIP6M_INNER_MAC);
 
 		switch (info->mask.tunnel_type_mask) {
 		case 0:
-- 
2.11.0

[dpdk-stable] patch 'net/i40e: fix shifts of 32-bit value' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From cd21f29cdf3a08ecd689b9109732ff92726f69aa Mon Sep 17 00:00:00 2001
From: Beilei Xing <beilei.xing@intel.com>
Date: Wed, 23 May 2018 15:46:46 +0800
Subject: [PATCH] net/i40e: fix shifts of 32-bit value

[ upstream commit bc98bf3a2d0301a17707bc04e95fc32746a397c4 ]

Cppcheck reports following error,
(error) Shifting 32-bit value by 36 bits is undefined behaviour

According to datasheet, there's PHY type and PHY type extension
in setting PHY config command, should exclude PHY type extension
when setting PHY type.

Fixes: 1bb8f661168d ("net/i40e: fix link down and negotiation")

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
---
 drivers/net/i40e/i40e_ethdev.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
index 85baff9e9..660175b31 100644
--- a/drivers/net/i40e/i40e_ethdev.c
+++ b/drivers/net/i40e/i40e_ethdev.c
@@ -2015,8 +2015,8 @@ i40e_phy_conf_link(struct i40e_hw *hw,
 
 
 
-	/* To enable link, phy_type mask needs to include each type */
-	for (cnt = I40E_PHY_TYPE_SGMII; cnt < I40E_PHY_TYPE_MAX; cnt++)
+	/* PHY type mask needs to include each type except PHY type extension */
+	for (cnt = I40E_PHY_TYPE_SGMII; cnt < I40E_PHY_TYPE_25GBASE_KR; cnt++)
 		phy_type_mask |= 1 << cnt;
 
 	/* use get_phy_abilities_resp value for the rest */
-- 
2.11.0

[dpdk-stable] patch 'app/testpmd: fix VLAN TCI mask set error for FDIR' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From acbe3c72a6dd26a3ae6bfc018b35165ab2db1f0d Mon Sep 17 00:00:00 2001
From: Wei Zhao <wei.zhao1@intel.com>
Date: Tue, 5 Jun 2018 17:12:11 +0800
Subject: [PATCH] app/testpmd: fix VLAN TCI mask set error for FDIR

[ upstream commit 26f579aa274b46f42d77228dd6628c951cba05a1 ]

The vlan tci mask should be set to 0xEFFF, not 0x0,
the wrong mask will cause mask error for register set.

Fixes: d9d5e6f2f0ba ("app/testpmd: set default flow director mask")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
---
 app/test-pmd/testpmd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index a4b28e923..6aa2f2c7c 100644
--- a/app/test-pmd/testpmd.c
+++ b/app/test-pmd/testpmd.c
@@ -354,7 +354,7 @@ struct rte_fdir_conf fdir_conf = {
 	.pballoc = RTE_FDIR_PBALLOC_64K,
 	.status = RTE_FDIR_REPORT_STATUS,
 	.mask = {
-		.vlan_tci_mask = 0x0,
+		.vlan_tci_mask = 0xFFEF,
 		.ipv4_mask     = {
 			.src_ip = 0xFFFFFFFF,
 			.dst_ip = 0xFFFFFFFF,
-- 
2.11.0

[dpdk-stable] patch 'net/i40e: workaround performance degradation' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From f3d7c22b8be200277f09a457d8b0c5611a909298 Mon Sep 17 00:00:00 2001
From: Haiyue Wang <haiyue.wang@intel.com>
Date: Wed, 13 Jun 2018 13:52:41 +0800
Subject: [PATCH] net/i40e: workaround performance degradation

[ upstream commit 3320d4a240daab15559d3e671bd44573a89cf017 ]

The GL_SWR_PM_UP_THR value is not impacted from the link speed, its
value is set according to the total number of ports for a better
pipe-monitor configuration.

All bellowing relevant device IDs are considered (NICs, LOMs, Mezz
and Backplane):

Device-ID  Value        Comments
0x1572     0x03030303   10G SFI
0x1581     0x03030303   10G Backplane
0x1586     0x03030303   10G BaseT
0x1589     0x03030303   10G BaseT (FortPond)
0x1580     0x06060606   40G Backplane
0x1583     0x06060606   2x40G QSFP
0x1584     0x06060606   1x40G QSFP
0x1587     0x06060606   20G Backplane (HP)
0x1588     0x06060606   20G KR2 (HP)
0x158A     0x06060606   25G Backplane
0x158B     0x06060606   25G SFP28

Fixes: c9223a2bf53c ("i40e: workaround for XL710 performance")
Fixes: 75d133dd3296 ("net/i40e: enable 25G device")

Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
---
 drivers/net/i40e/i40e_ethdev.c | 71 +++++++++++++++++++++++++++++++++++++-----
 1 file changed, 64 insertions(+), 7 deletions(-)

diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
index 660175b31..afdeeab5a 100644
--- a/drivers/net/i40e/i40e_ethdev.c
+++ b/drivers/net/i40e/i40e_ethdev.c
@@ -9768,6 +9768,60 @@ i40e_pctype_to_flowtype(const struct i40e_adapter *adapter,
 #define I40E_GL_SWR_PM_UP_THR_SF_VALUE   0x06060606
 #define I40E_GL_SWR_PM_UP_THR            0x269FBC
 
+/*
+ * GL_SWR_PM_UP_THR:
+ * The value is not impacted from the link speed, its value is set according
+ * to the total number of ports for a better pipe-monitor configuration.
+ */
+static bool
+i40e_get_swr_pm_cfg(struct i40e_hw *hw, uint32_t *value)
+{
+#define I40E_GL_SWR_PM_EF_DEVICE(dev) \
+		.device_id = (dev),   \
+		.val = I40E_GL_SWR_PM_UP_THR_EF_VALUE
+
+#define I40E_GL_SWR_PM_SF_DEVICE(dev) \
+		.device_id = (dev),   \
+		.val = I40E_GL_SWR_PM_UP_THR_SF_VALUE
+
+	static const struct {
+		uint16_t device_id;
+		uint32_t val;
+	} swr_pm_table[] = {
+		{ I40E_GL_SWR_PM_EF_DEVICE(I40E_DEV_ID_SFP_XL710) },
+		{ I40E_GL_SWR_PM_EF_DEVICE(I40E_DEV_ID_KX_C) },
+		{ I40E_GL_SWR_PM_EF_DEVICE(I40E_DEV_ID_10G_BASE_T) },
+		{ I40E_GL_SWR_PM_EF_DEVICE(I40E_DEV_ID_10G_BASE_T4) },
+
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_KX_B) },
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_QSFP_A) },
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_QSFP_B) },
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_20G_KR2) },
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_20G_KR2_A) },
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_25G_B) },
+		{ I40E_GL_SWR_PM_SF_DEVICE(I40E_DEV_ID_25G_SFP28) },
+	};
+	uint32_t i;
+
+	if (value == NULL) {
+		PMD_DRV_LOG(ERR, "value is NULL");
+		return false;
+	}
+
+	for (i = 0; i < RTE_DIM(swr_pm_table); i++) {
+		if (hw->device_id == swr_pm_table[i].device_id) {
+			*value = swr_pm_table[i].val;
+
+			PMD_DRV_LOG(DEBUG, "Device 0x%x with GL_SWR_PM_UP_THR "
+				    "value - 0x%08x",
+				    hw->device_id, *value);
+			return true;
+		}
+	}
+
+	return false;
+}
+
 static int
 i40e_dev_sync_phy_type(struct i40e_hw *hw)
 {
@@ -9832,13 +9886,16 @@ i40e_configure_registers(struct i40e_hw *hw)
 		}
 
 		if (reg_table[i].addr == I40E_GL_SWR_PM_UP_THR) {
-			if (I40E_PHY_TYPE_SUPPORT_40G(hw->phy.phy_types) || /* For XL710 */
-			    I40E_PHY_TYPE_SUPPORT_25G(hw->phy.phy_types)) /* For XXV710 */
-				reg_table[i].val =
-					I40E_GL_SWR_PM_UP_THR_SF_VALUE;
-			else /* For X710 */
-				reg_table[i].val =
-					I40E_GL_SWR_PM_UP_THR_EF_VALUE;
+			uint32_t cfg_val;
+
+			if (!i40e_get_swr_pm_cfg(hw, &cfg_val)) {
+				PMD_DRV_LOG(DEBUG, "Device 0x%x skips "
+					    "GL_SWR_PM_UP_THR value fixup",
+					    hw->device_id);
+				continue;
+			}
+
+			reg_table[i].val = cfg_val;
 		}
 
 		ret = i40e_aq_debug_read_register(hw, reg_table[i].addr,
-- 
2.11.0

[dpdk-stable] patch 'net/pcap: fix multiple queues' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From abfc2bace9e58f72dd7ca63a7a3133876b0307f4 Mon Sep 17 00:00:00 2001
From: Ido Goshen <ido@cgstowernetworks.com>
Date: Tue, 19 Jun 2018 17:37:25 +0300
Subject: [PATCH] net/pcap: fix multiple queues

[ upstream commit 8ffda7673fa71e6a666963fc0ea4969d65dd09da ]

Change open_rx/tx_pcap/iface functions to open only a single pcap/dumper
and not loop num_of_queue times
The num_of_queue loop is already achieved by the
caller rte_kvargs_process

Fixing:
1. Opens N requested pcaps/dumpers instead of N^2
2. Leak of pcap/dumper's which are being overwritten by
   the sequential calls to open_rx/tx_pcap/iface functions
3. Use the filename/iface args per queue and not just the last one
   that overwrites the previous names

Fixes: 4c173302c307 ("pcap: add new driver")

Signed-off-by: Ido Goshen <ido@cgstowernetworks.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
---
 drivers/net/pcap/rte_eth_pcap.c | 86 +++++++++++++++++------------------------
 1 file changed, 36 insertions(+), 50 deletions(-)

diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 39947105b..f71ba001e 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/pcap/rte_eth_pcap.c
@@ -687,19 +687,19 @@ static const struct eth_dev_ops ops = {
 static int
 open_rx_pcap(const char *key, const char *value, void *extra_args)
 {
-	unsigned int i;
 	const char *pcap_filename = value;
 	struct pmd_devargs *rx = extra_args;
 	pcap_t *pcap = NULL;
 
-	for (i = 0; i < rx->num_of_queue; i++) {
-		if (open_single_rx_pcap(pcap_filename, &pcap) < 0)
-			return -1;
+	if (rx->num_of_queue >= RTE_PMD_PCAP_MAX_QUEUES)
+		return -1;
+	if (open_single_rx_pcap(pcap_filename, &pcap) < 0)
+		return -1;
 
-		rx->queue[i].pcap = pcap;
-		rx->queue[i].name = pcap_filename;
-		rx->queue[i].type = key;
-	}
+	rx->queue[rx->num_of_queue].pcap = pcap;
+	rx->queue[rx->num_of_queue].name = pcap_filename;
+	rx->queue[rx->num_of_queue].type = key;
+	rx->num_of_queue++;
 
 	return 0;
 }
@@ -711,19 +711,19 @@ open_rx_pcap(const char *key, const char *value, void *extra_args)
 static int
 open_tx_pcap(const char *key, const char *value, void *extra_args)
 {
-	unsigned int i;
 	const char *pcap_filename = value;
 	struct pmd_devargs *dumpers = extra_args;
 	pcap_dumper_t *dumper;
 
-	for (i = 0; i < dumpers->num_of_queue; i++) {
-		if (open_single_tx_pcap(pcap_filename, &dumper) < 0)
-			return -1;
+	if (dumpers->num_of_queue >= RTE_PMD_PCAP_MAX_QUEUES)
+		return -1;
+	if (open_single_tx_pcap(pcap_filename, &dumper) < 0)
+		return -1;
 
-		dumpers->queue[i].dumper = dumper;
-		dumpers->queue[i].name = pcap_filename;
-		dumpers->queue[i].type = key;
-	}
+	dumpers->queue[dumpers->num_of_queue].dumper = dumper;
+	dumpers->queue[dumpers->num_of_queue].name = pcap_filename;
+	dumpers->queue[dumpers->num_of_queue].type = key;
+	dumpers->num_of_queue++;
 
 	return 0;
 }
@@ -754,18 +754,18 @@ open_rx_tx_iface(const char *key, const char *value, void *extra_args)
 static inline int
 open_rx_iface(const char *key, const char *value, void *extra_args)
 {
-	unsigned int i;
 	const char *iface = value;
 	struct pmd_devargs *rx = extra_args;
 	pcap_t *pcap = NULL;
 
-	for (i = 0; i < rx->num_of_queue; i++) {
-		if (open_single_iface(iface, &pcap) < 0)
-			return -1;
-		rx->queue[i].pcap = pcap;
-		rx->queue[i].name = iface;
-		rx->queue[i].type = key;
-	}
+	if (rx->num_of_queue >= RTE_PMD_PCAP_MAX_QUEUES)
+		return -1;
+	if (open_single_iface(iface, &pcap) < 0)
+		return -1;
+	rx->queue[rx->num_of_queue].pcap = pcap;
+	rx->queue[rx->num_of_queue].name = iface;
+	rx->queue[rx->num_of_queue].type = key;
+	rx->num_of_queue++;
 
 	return 0;
 }
@@ -776,18 +776,18 @@ open_rx_iface(const char *key, const char *value, void *extra_args)
 static int
 open_tx_iface(const char *key, const char *value, void *extra_args)
 {
-	unsigned int i;
 	const char *iface = value;
 	struct pmd_devargs *tx = extra_args;
 	pcap_t *pcap;
 
-	for (i = 0; i < tx->num_of_queue; i++) {
-		if (open_single_iface(iface, &pcap) < 0)
-			return -1;
-		tx->queue[i].pcap = pcap;
-		tx->queue[i].name = iface;
-		tx->queue[i].type = key;
-	}
+	if (tx->num_of_queue >= RTE_PMD_PCAP_MAX_QUEUES)
+		return -1;
+	if (open_single_iface(iface, &pcap) < 0)
+		return -1;
+	tx->queue[tx->num_of_queue].pcap = pcap;
+	tx->queue[tx->num_of_queue].name = iface;
+	tx->queue[tx->num_of_queue].type = key;
+	tx->num_of_queue++;
 
 	return 0;
 }
@@ -977,15 +977,8 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
 	 * We check whether we want to open a RX stream from a real NIC or a
 	 * pcap file
 	 */
-	pcaps.num_of_queue = rte_kvargs_count(kvlist, ETH_PCAP_RX_PCAP_ARG);
-	if (pcaps.num_of_queue)
-		is_rx_pcap = 1;
-	else
-		pcaps.num_of_queue = rte_kvargs_count(kvlist,
-				ETH_PCAP_RX_IFACE_ARG);
-
-	if (pcaps.num_of_queue > RTE_PMD_PCAP_MAX_QUEUES)
-		pcaps.num_of_queue = RTE_PMD_PCAP_MAX_QUEUES;
+	is_rx_pcap = rte_kvargs_count(kvlist, ETH_PCAP_RX_PCAP_ARG) ? 1 : 0;
+	pcaps.num_of_queue = 0;
 
 	if (is_rx_pcap)
 		ret = rte_kvargs_process(kvlist, ETH_PCAP_RX_PCAP_ARG,
@@ -1001,15 +994,8 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
 	 * We check whether we want to open a TX stream to a real NIC or a
 	 * pcap file
 	 */
-	dumpers.num_of_queue = rte_kvargs_count(kvlist, ETH_PCAP_TX_PCAP_ARG);
-	if (dumpers.num_of_queue)
-		is_tx_pcap = 1;
-	else
-		dumpers.num_of_queue = rte_kvargs_count(kvlist,
-				ETH_PCAP_TX_IFACE_ARG);
-
-	if (dumpers.num_of_queue > RTE_PMD_PCAP_MAX_QUEUES)
-		dumpers.num_of_queue = RTE_PMD_PCAP_MAX_QUEUES;
+	is_tx_pcap = rte_kvargs_count(kvlist, ETH_PCAP_TX_PCAP_ARG) ? 1 : 0;
+	dumpers.num_of_queue = 0;
 
 	if (is_tx_pcap)
 		ret = rte_kvargs_process(kvlist, ETH_PCAP_TX_PCAP_ARG,
-- 
2.11.0

[dpdk-stable] patch 'net/thunderx: fix build with gcc optimization on' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 147d06d845676d519c46f8ca22d2d7625b19197f Mon Sep 17 00:00:00 2001
From: Ferruh Yigit <ferruh.yigit@intel.com>
Date: Thu, 21 Jun 2018 19:14:50 +0100
Subject: [PATCH] net/thunderx: fix build with gcc optimization on
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

[ upstream commit d90141961b9acead2a3cfacc7547ac1a256979eb ]

build error gcc version 6.3.1 20161221 (Red Hat 6.3.1-1),
with EXTRA_CFLAGS="-O3":

.../drivers/net/thunderx/nicvf_ethdev.c:907:9:
   error: ‘txq’ may be used uninitialized in this function
   [-Werror=maybe-uninitialized]
  if (txq->pool_free == nicvf_single_pool_free_xmited_buffers)
      ~~~^~~~~~~~~~~
.../drivers/net/thunderx/nicvf_ethdev.c:886:20:
   note: ‘txq’ was declared here
  struct nicvf_txq *txq;
                    ^~~

Same error on function 'nicvf_eth_dev_init' and 'nicvf_dev_start', it
seems 'nicvf_set_tx_function' inlined when optimization enabled.

Initialize the txq and add NULL check before using it to fix.

Fixes: 7413feee662d ("net/thunderx: add device start/stop and close")

Reported-by: Richard Walsh <richard.walsh@intel.com>
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
---
 drivers/net/thunderx/nicvf_ethdev.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/thunderx/nicvf_ethdev.c b/drivers/net/thunderx/nicvf_ethdev.c
index c62371cb1..3843084b9 100644
--- a/drivers/net/thunderx/nicvf_ethdev.c
+++ b/drivers/net/thunderx/nicvf_ethdev.c
@@ -897,7 +897,7 @@ nicvf_dev_tx_queue_release(void *sq)
 static void
 nicvf_set_tx_function(struct rte_eth_dev *dev)
 {
-	struct nicvf_txq *txq;
+	struct nicvf_txq *txq = NULL;
 	size_t i;
 	bool multiseg = false;
 
@@ -918,6 +918,9 @@ nicvf_set_tx_function(struct rte_eth_dev *dev)
 		dev->tx_pkt_burst = nicvf_xmit_pkts;
 	}
 
+	if (!txq)
+		return;
+
 	if (txq->pool_free == nicvf_single_pool_free_xmited_buffers)
 		PMD_DRV_LOG(DEBUG, "Using single-mempool tx free method");
 	else
-- 
2.11.0

[dpdk-stable] patch 'net/qede: fix legacy interrupt mode' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 13c7547c41d0504d6932421ce2f6e124b8754496 Mon Sep 17 00:00:00 2001
From: Shahed Shaikh <shahed.shaikh@cavium.com>
Date: Sat, 23 Jun 2018 14:20:32 -0700
Subject: [PATCH] net/qede: fix legacy interrupt mode

[ upstream commit 245aec2893380e442169c8a1914c36daa444efd3 ]

qede pmd does not have support for legacy interrupt mode.
This causes slow path completion failure with uio_pci_generic module,
since it uses legacy interrupt (INTx) mode.

Fix this issue by installing legacy interrupt handler.

Fixes: ec94dbc57362 ("qede: add base driver")

Signed-off-by: Shahed Shaikh <shahed.shaikh@cavium.com>
---
 drivers/net/qede/qede_ethdev.c | 40 +++++++++++++++++++++++++++++++++++++---
 drivers/net/qede/qede_main.c   |  2 +-
 2 files changed, 38 insertions(+), 4 deletions(-)

diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index c63b4d5c9..ee439ae6c 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -336,6 +336,24 @@ static void qede_interrupt_action(struct ecore_hwfn *p_hwfn)
 }
 
 static void
+qede_interrupt_handler_intx(void *param)
+{
+	struct rte_eth_dev *eth_dev = (struct rte_eth_dev *)param;
+	struct qede_dev *qdev = eth_dev->data->dev_private;
+	struct ecore_dev *edev = &qdev->edev;
+	u64 status;
+
+	/* Check if our device actually raised an interrupt */
+	status = ecore_int_igu_read_sisr_reg(ECORE_LEADING_HWFN(edev));
+	if (status & 0x1) {
+		qede_interrupt_action(ECORE_LEADING_HWFN(edev));
+
+		if (rte_intr_enable(eth_dev->intr_handle))
+			DP_ERR(edev, "rte_intr_enable failed\n");
+	}
+}
+
+static void
 qede_interrupt_handler(void *param)
 {
 	struct rte_eth_dev *eth_dev = (struct rte_eth_dev *)param;
@@ -2844,6 +2862,7 @@ static int qede_common_dev_init(struct rte_eth_dev *eth_dev, bool is_vf)
 	/* Fix up ecore debug level */
 	uint32_t dp_module = ~0 & ~ECORE_MSG_HW;
 	uint8_t dp_level = ECORE_LEVEL_VERBOSE;
+	uint32_t int_mode;
 	int rc;
 
 	/* Extract key data structures */
@@ -2888,8 +2907,22 @@ static int qede_common_dev_init(struct rte_eth_dev *eth_dev, bool is_vf)
 		return -ENODEV;
 	}
 	qede_update_pf_params(edev);
-	rte_intr_callback_register(&pci_dev->intr_handle,
-				   qede_interrupt_handler, (void *)eth_dev);
+
+	switch (pci_dev->intr_handle.type) {
+	case RTE_INTR_HANDLE_UIO_INTX:
+	case RTE_INTR_HANDLE_VFIO_LEGACY:
+		int_mode = ECORE_INT_MODE_INTA;
+		rte_intr_callback_register(&pci_dev->intr_handle,
+					   qede_interrupt_handler_intx,
+					   (void *)eth_dev);
+		break;
+	default:
+		int_mode = ECORE_INT_MODE_MSIX;
+		rte_intr_callback_register(&pci_dev->intr_handle,
+					   qede_interrupt_handler,
+					   (void *)eth_dev);
+	}
+
 	if (rte_intr_enable(&pci_dev->intr_handle)) {
 		DP_ERR(edev, "rte_intr_enable() failed\n");
 		return -ENODEV;
@@ -2897,7 +2930,8 @@ static int qede_common_dev_init(struct rte_eth_dev *eth_dev, bool is_vf)
 
 	/* Start the Slowpath-process */
 	memset(&params, 0, sizeof(struct qed_slowpath_params));
-	params.int_mode = ECORE_INT_MODE_MSIX;
+
+	params.int_mode = int_mode;
 	params.drv_major = QEDE_PMD_VERSION_MAJOR;
 	params.drv_minor = QEDE_PMD_VERSION_MINOR;
 	params.drv_rev = QEDE_PMD_VERSION_REVISION;
diff --git a/drivers/net/qede/qede_main.c b/drivers/net/qede/qede_main.c
index 373139d7a..95b4cd91b 100644
--- a/drivers/net/qede/qede_main.c
+++ b/drivers/net/qede/qede_main.c
@@ -279,7 +279,7 @@ static int qed_slowpath_start(struct ecore_dev *edev,
 	/* Start the slowpath */
 	memset(&hw_init_params, 0, sizeof(hw_init_params));
 	hw_init_params.b_hw_start = true;
-	hw_init_params.int_mode = ECORE_INT_MODE_MSIX;
+	hw_init_params.int_mode = params->int_mode;
 	hw_init_params.allow_npar_tx_switch = true;
 	hw_init_params.bin_fw_data = data;
 
-- 
2.11.0

[dpdk-stable] patch 'net/qede: remove primary MAC removal' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 9228fe31f31fc7194854707fd19e77986269ae23 Mon Sep 17 00:00:00 2001
From: Rasesh Mody <rasesh.mody@cavium.com>
Date: Sat, 23 Jun 2018 14:20:33 -0700
Subject: [PATCH] net/qede: remove primary MAC removal

[ upstream commit 475ff28951b69ce0c46514e8c6a6e8bfd91c5eb3 ]

This was added to dev_stop when set MTU requested vport restart.
We are not requiring vport restart any more with set MTU as it just
needs vport to be inactive and doesn't need the port reconfigured.

Fixes: d121a6b5f781 ("net/qede: fix VF MTU update")

Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
---
 drivers/net/qede/qede_ethdev.c | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index ee439ae6c..cd66a881c 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -1238,9 +1238,6 @@ static void qede_dev_stop(struct rte_eth_dev *eth_dev)
 	/* Disable traffic */
 	ecore_hw_stop_fastpath(edev); /* TBD - loop */
 
-	if (IS_PF(edev))
-		qede_mac_addr_remove(eth_dev, 0);
-
 	DP_INFO(edev, "Device is stopped\n");
 }
 
@@ -1963,8 +1960,6 @@ int qede_update_mtu(struct rte_eth_dev *eth_dev, uint16_t mtu)
 					goto err;
 
 				/* Restore config lost due to vport stop */
-				qede_mac_addr_set(eth_dev, &qdev->primary_mac);
-
 				if (eth_dev->data->promiscuous)
 					qede_promiscuous_enable(eth_dev);
 				else
@@ -2391,9 +2386,6 @@ static int qede_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
 		dev->data->dev_started = 0;
 		qede_dev_stop(dev);
 		restart = true;
-	} else {
-		if (IS_PF(edev))
-			qede_mac_addr_remove(dev, 0);
 	}
 	rte_delay_ms(1000);
 	qdev->mtu = mtu;
-- 
2.11.0

[dpdk-stable] patch 'net/ena: fix SIGFPE with 0 Rx queue' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 2387b625656cc0464f164e6ab0c45732b7da7115 Mon Sep 17 00:00:00 2001
From: Daria Kolistratova <daria.kolistratova@intel.com>
Date: Tue, 26 Jun 2018 18:38:56 +0100
Subject: [PATCH] net/ena: fix SIGFPE with 0 Rx queue

[ upstream commit 361913ad6f8c05fc541fe4bfdae3b0dc095ae3af ]

When the number of rx queues is 0 (what can be when application does
not receive) failed with SIGFPE.
It happens when the application is also requesting ETH_MQ_RX_RSS_FLAG
in the rte_dev->data->dev_conf.rxmode.mq_mode.
Fixed adding zero rx queues check.

Signed-off-by: Daria Kolistratova <daria.kolistratova@intel.com>
Acked-by: Michal Krawczyk <mk@semihalf.com>
---
 drivers/net/ena/ena_ethdev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ena/ena_ethdev.c b/drivers/net/ena/ena_ethdev.c
index aa24ef36a..63c36d490 100644
--- a/drivers/net/ena/ena_ethdev.c
+++ b/drivers/net/ena/ena_ethdev.c
@@ -907,7 +907,7 @@ static int ena_start(struct rte_eth_dev *dev)
 		return rc;
 
 	if (adapter->rte_dev->data->dev_conf.rxmode.mq_mode &
-	    ETH_MQ_RX_RSS_FLAG) {
+	    ETH_MQ_RX_RSS_FLAG && adapter->rte_dev->data->nb_rx_queues > 0) {
 		rc = ena_rss_init_default(adapter);
 		if (rc)
 			return rc;
-- 
2.11.0

[dpdk-stable] patch 'net/mlx5: fix Rx buffer replenishment threshold' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From b9e1edfe8ee644be2fe5c13e3239c8ee1f48afb4 Mon Sep 17 00:00:00 2001
From: Yongseok Koh <yskoh@mellanox.com>
Date: Tue, 26 Jun 2018 04:33:35 -0700
Subject: [PATCH] net/mlx5: fix Rx buffer replenishment threshold

[ upstream commit e10245a13b2e340f48ce80484f19bcbc13e9ebe6 ]

The threshold of buffer replenishment for vectorized Rx burst is a constant
value (64). If the size of Rx queue is comparatively small, device could
run out of buffers. For example, if the size of Rx queue is 128, buffers
are replenished only twice per a wraparound. This can cause jitter in
receiving packets and the jitter can cause unnecessary retransmission for
TCP connections.

Fixes: 6cb559d67b83 ("net/mlx5: add vectorized Rx/Tx burst for x86")
Fixes: 570acdb1da8a ("net/mlx5: add vectorized Rx/Tx burst for ARM")

Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Shahaf Shuler <shahafs@mellanox.com>
---
 drivers/net/mlx5/mlx5_defs.h          | 5 +++--
 drivers/net/mlx5/mlx5_rxtx_vec.h      | 4 ++--
 drivers/net/mlx5/mlx5_rxtx_vec_neon.h | 2 +-
 drivers/net/mlx5/mlx5_rxtx_vec_sse.h  | 2 +-
 4 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_defs.h b/drivers/net/mlx5/mlx5_defs.h
index d70635767..e9cda9a08 100644
--- a/drivers/net/mlx5/mlx5_defs.h
+++ b/drivers/net/mlx5/mlx5_defs.h
@@ -92,10 +92,11 @@
 #define MLX5_VPMD_MIN_TXQS 4
 
 /* Threshold of buffer replenishment for vectorized Rx. */
-#define MLX5_VPMD_RXQ_RPLNSH_THRESH   64U
+#define MLX5_VPMD_RXQ_RPLNSH_THRESH(n) \
+	(RTE_MIN(MLX5_VPMD_RX_MAX_BURST, (unsigned int)(n) >> 2))
 
 /* Maximum size of burst for vectorized Rx. */
-#define MLX5_VPMD_RX_MAX_BURST        MLX5_VPMD_RXQ_RPLNSH_THRESH
+#define MLX5_VPMD_RX_MAX_BURST 64U
 
 /*
  * Maximum size of burst for vectorized Tx. This is related to the maximum size
diff --git a/drivers/net/mlx5/mlx5_rxtx_vec.h b/drivers/net/mlx5/mlx5_rxtx_vec.h
index 1f08ed0b2..d504e2aee 100644
--- a/drivers/net/mlx5/mlx5_rxtx_vec.h
+++ b/drivers/net/mlx5/mlx5_rxtx_vec.h
@@ -106,9 +106,9 @@ mlx5_rx_replenish_bulk_mbuf(struct mlx5_rxq_data *rxq, uint16_t n)
 	volatile struct mlx5_wqe_data_seg *wq = &(*rxq->wqes)[elts_idx];
 	unsigned int i;
 
-	assert(n >= MLX5_VPMD_RXQ_RPLNSH_THRESH);
+	assert(n >= MLX5_VPMD_RXQ_RPLNSH_THRESH(q_n));
 	assert(n <= (uint16_t)(q_n - (rxq->rq_ci - rxq->rq_pi)));
-	assert(MLX5_VPMD_RXQ_RPLNSH_THRESH > MLX5_VPMD_DESCS_PER_LOOP);
+	assert(MLX5_VPMD_RXQ_RPLNSH_THRESH(q_n) > MLX5_VPMD_DESCS_PER_LOOP);
 	/* Not to cross queue end. */
 	n = RTE_MIN(n - MLX5_VPMD_DESCS_PER_LOOP, q_n - elts_idx);
 	if (rte_mempool_get_bulk(rxq->mp, (void *)elts, n) < 0) {
diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_neon.h b/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
index cf424778a..1604d0430 100644
--- a/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
+++ b/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
@@ -754,7 +754,7 @@ rxq_burst_v(struct mlx5_rxq_data *rxq, struct rte_mbuf **pkts, uint16_t pkts_n,
 	 *   N - (rq_ci - rq_pi) := # of buffers consumed (to be replenished).
 	 */
 	repl_n = q_n - (rxq->rq_ci - rxq->rq_pi);
-	if (repl_n >= MLX5_VPMD_RXQ_RPLNSH_THRESH)
+	if (repl_n >= MLX5_VPMD_RXQ_RPLNSH_THRESH(q_n))
 		mlx5_rx_replenish_bulk_mbuf(rxq, repl_n);
 	/* See if there're unreturned mbufs from compressed CQE. */
 	rcvd_pkt = rxq->cq_ci - rxq->rq_pi;
diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_sse.h b/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
index 793142922..01a93a4b4 100644
--- a/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
+++ b/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
@@ -735,7 +735,7 @@ rxq_burst_v(struct mlx5_rxq_data *rxq, struct rte_mbuf **pkts, uint16_t pkts_n,
 	 *   N - (rq_ci - rq_pi) := # of buffers consumed (to be replenished).
 	 */
 	repl_n = q_n - (rxq->rq_ci - rxq->rq_pi);
-	if (repl_n >= MLX5_VPMD_RXQ_RPLNSH_THRESH)
+	if (repl_n >= MLX5_VPMD_RXQ_RPLNSH_THRESH(q_n))
 		mlx5_rx_replenish_bulk_mbuf(rxq, repl_n);
 	/* See if there're unreturned mbufs from compressed CQE. */
 	rcvd_pkt = rxq->cq_ci - rxq->rq_pi;
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: fix HW Tx checksum offload check' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/28/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 7f55cf2748313b567f98bad0b73768623de57a2c Mon Sep 17 00:00:00 2001
From: Ajit Khaparde <ajit.khaparde@broadcom.com>
Date: Thu, 28 Jun 2018 13:15:33 -0700
Subject: [PATCH] net/bnxt: fix HW Tx checksum offload check

[ upstream commit f10258e39ec2a08d79585ecaf494198ba49bc6eb ]

Add more checks for checksum calculation offload.
Also check for tunnel frames and select the proper
buffer descriptor size.

Fixes: 6eb3cc2294fd ("net/bnxt: add initial Tx code")

Signed-off-by: Xiaoxin Peng <xiaoxin.peng@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Jason He <jason.he@broadcom.com>
Reviewed-by: Qingmin Liu <qingmin.liu@broadcom.com>
---
 drivers/net/bnxt/bnxt_txr.c | 51 ++++++++++++++++++++++++++++++++++++++++++---
 drivers/net/bnxt/bnxt_txr.h | 10 +++++++++
 2 files changed, 58 insertions(+), 3 deletions(-)

diff --git a/drivers/net/bnxt/bnxt_txr.c b/drivers/net/bnxt/bnxt_txr.c
index 03d2652c5..f5ed03f1b 100644
--- a/drivers/net/bnxt/bnxt_txr.c
+++ b/drivers/net/bnxt/bnxt_txr.c
@@ -161,7 +161,9 @@ static uint16_t bnxt_start_xmit(struct rte_mbuf *tx_pkt,
 
 	if (tx_pkt->ol_flags & (PKT_TX_TCP_SEG | PKT_TX_TCP_CKSUM |
 				PKT_TX_UDP_CKSUM | PKT_TX_IP_CKSUM |
-				PKT_TX_VLAN_PKT | PKT_TX_OUTER_IP_CKSUM))
+				PKT_TX_VLAN_PKT | PKT_TX_OUTER_IP_CKSUM |
+				PKT_TX_TUNNEL_GRE | PKT_TX_TUNNEL_VXLAN |
+				PKT_TX_TUNNEL_GENEVE))
 		long_bd = true;
 
 	tx_buf = &txr->tx_buf_ring[txr->tx_prod];
@@ -222,16 +224,46 @@ static uint16_t bnxt_start_xmit(struct rte_mbuf *tx_pkt,
 			/* Outer IP, Inner IP, Inner TCP/UDP CSO */
 			txbd1->lflags |= TX_BD_FLG_TIP_IP_TCP_UDP_CHKSUM;
 			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_OIP_IIP_TCP_CKSUM) ==
+			   PKT_TX_OIP_IIP_TCP_CKSUM) {
+			/* Outer IP, Inner IP, Inner TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_FLG_TIP_IP_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_OIP_IIP_UDP_CKSUM) ==
+			   PKT_TX_OIP_IIP_UDP_CKSUM) {
+			/* Outer IP, Inner IP, Inner TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_FLG_TIP_IP_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
 		} else if ((tx_pkt->ol_flags & PKT_TX_IIP_TCP_UDP_CKSUM) ==
 			   PKT_TX_IIP_TCP_UDP_CKSUM) {
 			/* (Inner) IP, (Inner) TCP/UDP CSO */
 			txbd1->lflags |= TX_BD_FLG_IP_TCP_UDP_CHKSUM;
 			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_IIP_UDP_CKSUM) ==
+			   PKT_TX_IIP_UDP_CKSUM) {
+			/* (Inner) IP, (Inner) TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_FLG_IP_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_IIP_TCP_CKSUM) ==
+			   PKT_TX_IIP_TCP_CKSUM) {
+			/* (Inner) IP, (Inner) TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_FLG_IP_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
 		} else if ((tx_pkt->ol_flags & PKT_TX_OIP_TCP_UDP_CKSUM) ==
 			   PKT_TX_OIP_TCP_UDP_CKSUM) {
 			/* Outer IP, (Inner) TCP/UDP CSO */
 			txbd1->lflags |= TX_BD_FLG_TIP_TCP_UDP_CHKSUM;
 			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_OIP_UDP_CKSUM) ==
+			   PKT_TX_OIP_UDP_CKSUM) {
+			/* Outer IP, (Inner) TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_FLG_TIP_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_OIP_TCP_CKSUM) ==
+			   PKT_TX_OIP_TCP_CKSUM) {
+			/* Outer IP, (Inner) TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_FLG_TIP_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
 		} else if ((tx_pkt->ol_flags & PKT_TX_OIP_IIP_CKSUM) ==
 			   PKT_TX_OIP_IIP_CKSUM) {
 			/* Outer IP, Inner IP CSO */
@@ -242,11 +274,23 @@ static uint16_t bnxt_start_xmit(struct rte_mbuf *tx_pkt,
 			/* TCP/UDP CSO */
 			txbd1->lflags |= TX_BD_LONG_LFLAGS_TCP_UDP_CHKSUM;
 			txbd1->mss = 0;
-		} else if (tx_pkt->ol_flags & PKT_TX_IP_CKSUM) {
+		} else if ((tx_pkt->ol_flags & PKT_TX_TCP_CKSUM) ==
+			   PKT_TX_TCP_CKSUM) {
+			/* TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_LONG_LFLAGS_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_UDP_CKSUM) ==
+			   PKT_TX_UDP_CKSUM) {
+			/* TCP/UDP CSO */
+			txbd1->lflags |= TX_BD_LONG_LFLAGS_TCP_UDP_CHKSUM;
+			txbd1->mss = 0;
+		} else if ((tx_pkt->ol_flags & PKT_TX_IP_CKSUM) ==
+			   PKT_TX_IP_CKSUM) {
 			/* IP CSO */
 			txbd1->lflags |= TX_BD_LONG_LFLAGS_IP_CHKSUM;
 			txbd1->mss = 0;
-		} else if (tx_pkt->ol_flags & PKT_TX_OUTER_IP_CKSUM) {
+		} else if ((tx_pkt->ol_flags & PKT_TX_OUTER_IP_CKSUM) ==
+			   PKT_TX_OUTER_IP_CKSUM) {
 			/* IP CSO */
 			txbd1->lflags |= TX_BD_LONG_LFLAGS_T_IP_CHKSUM;
 			txbd1->mss = 0;
@@ -270,6 +314,7 @@ static uint16_t bnxt_start_xmit(struct rte_mbuf *tx_pkt,
 	}
 
 	txbd->flags_type |= TX_BD_LONG_FLAGS_PACKET_END;
+	txbd1->lflags = rte_cpu_to_le_32(txbd1->lflags);
 
 	txr->tx_prod = RING_NEXT(txr->tx_ring_struct, txr->tx_prod);
 
diff --git a/drivers/net/bnxt/bnxt_txr.h b/drivers/net/bnxt/bnxt_txr.h
index 2feac51db..0bc217f09 100644
--- a/drivers/net/bnxt/bnxt_txr.h
+++ b/drivers/net/bnxt/bnxt_txr.h
@@ -71,10 +71,20 @@ uint16_t bnxt_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
 
 #define PKT_TX_OIP_IIP_TCP_UDP_CKSUM	(PKT_TX_TCP_CKSUM | PKT_TX_UDP_CKSUM | \
 					PKT_TX_IP_CKSUM | PKT_TX_OUTER_IP_CKSUM)
+#define PKT_TX_OIP_IIP_UDP_CKSUM	(PKT_TX_UDP_CKSUM | \
+					PKT_TX_IP_CKSUM | PKT_TX_OUTER_IP_CKSUM)
+#define PKT_TX_OIP_IIP_TCP_CKSUM	(PKT_TX_TCP_CKSUM | \
+					PKT_TX_IP_CKSUM | PKT_TX_OUTER_IP_CKSUM)
 #define PKT_TX_IIP_TCP_UDP_CKSUM	(PKT_TX_TCP_CKSUM | PKT_TX_UDP_CKSUM | \
 					PKT_TX_IP_CKSUM)
+#define PKT_TX_IIP_TCP_CKSUM		(PKT_TX_TCP_CKSUM | PKT_TX_IP_CKSUM)
+#define PKT_TX_IIP_UDP_CKSUM		(PKT_TX_UDP_CKSUM | PKT_TX_IP_CKSUM)
 #define PKT_TX_OIP_TCP_UDP_CKSUM	(PKT_TX_TCP_CKSUM | PKT_TX_UDP_CKSUM | \
 					PKT_TX_OUTER_IP_CKSUM)
+#define PKT_TX_OIP_UDP_CKSUM		(PKT_TX_UDP_CKSUM | \
+					PKT_TX_OUTER_IP_CKSUM)
+#define PKT_TX_OIP_TCP_CKSUM		(PKT_TX_TCP_CKSUM | \
+					PKT_TX_OUTER_IP_CKSUM)
 #define PKT_TX_OIP_IIP_CKSUM		(PKT_TX_IP_CKSUM |	\
 					 PKT_TX_OUTER_IP_CKSUM)
 #define PKT_TX_TCP_UDP_CKSUM		(PKT_TX_TCP_CKSUM | PKT_TX_UDP_CKSUM)
-- 
2.11.0