[dpdk-dev] [PATCH 1/5] net/qede: convert to new Rx/Tx offloads API
Rasesh Mody
rasesh.mody at cavium.com
Wed Mar 28 02:15:51 CEST 2018
From: Harish Patil <harish.patil at cavium.com>
Ethdev RX/TX offloads API has changed since:
commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API")
commit cba7f53b717d ("ethdev: introduce Tx queue offloads API")
This patch makes use of new offload flags as suggested and fix
the default RX/TX configuration. Also indent the code properly
around those changes.
Signed-off-by: Harish Patil <harish.patil at cavium.com>
Signed-off-by: Rasesh Mody <rasesh.mody at cavium.com>
---
drivers/net/qede/qede_ethdev.c | 103 ++++++++++++++++++++++------------------
drivers/net/qede/qede_rxtx.c | 2 +-
drivers/net/qede/qede_rxtx.h | 2 -
3 files changed, 59 insertions(+), 48 deletions(-)
diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index a91f436..6a51e3d 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -591,6 +591,8 @@ int qede_enable_tpa(struct rte_eth_dev *eth_dev, bool flg)
}
}
qdev->enable_lro = flg;
+ eth_dev->data->lro = flg;
+
DP_INFO(edev, "LRO is %s\n", flg ? "enabled" : "disabled");
return 0;
@@ -1166,10 +1168,10 @@ static int qede_vlan_offload_set(struct rte_eth_dev *eth_dev, int mask)
{
struct qede_dev *qdev = QEDE_INIT_QDEV(eth_dev);
struct ecore_dev *edev = QEDE_INIT_EDEV(qdev);
- struct rte_eth_rxmode *rxmode = ð_dev->data->dev_conf.rxmode;
+ uint64_t rx_offloads = eth_dev->data->dev_conf.rxmode.offloads;
if (mask & ETH_VLAN_STRIP_MASK) {
- if (rxmode->hw_vlan_strip)
+ if (rx_offloads & DEV_RX_OFFLOAD_VLAN_STRIP)
(void)qede_vlan_stripping(eth_dev, 1);
else
(void)qede_vlan_stripping(eth_dev, 0);
@@ -1177,7 +1179,7 @@ static int qede_vlan_offload_set(struct rte_eth_dev *eth_dev, int mask)
if (mask & ETH_VLAN_FILTER_MASK) {
/* VLAN filtering kicks in when a VLAN is added */
- if (rxmode->hw_vlan_filter) {
+ if (rx_offloads & DEV_RX_OFFLOAD_VLAN_FILTER) {
qede_vlan_filter_set(eth_dev, 0, 1);
} else {
if (qdev->configured_vlans > 1) { /* Excluding VLAN0 */
@@ -1187,7 +1189,8 @@ static int qede_vlan_offload_set(struct rte_eth_dev *eth_dev, int mask)
/* Signal app that VLAN filtering is still
* enabled
*/
- rxmode->hw_vlan_filter = true;
+ eth_dev->data->dev_conf.rxmode.offloads |=
+ DEV_RX_OFFLOAD_VLAN_FILTER;
} else {
qede_vlan_filter_set(eth_dev, 0, 0);
}
@@ -1195,13 +1198,11 @@ static int qede_vlan_offload_set(struct rte_eth_dev *eth_dev, int mask)
}
if (mask & ETH_VLAN_EXTEND_MASK)
- DP_INFO(edev, "No offloads are supported with VLAN Q-in-Q"
- " and classification is based on outer tag only\n");
+ DP_ERR(edev, "Extend VLAN not supported\n");
qdev->vlan_offload_mask = mask;
- DP_INFO(edev, "vlan offload mask %d vlan-strip %d vlan-filter %d\n",
- mask, rxmode->hw_vlan_strip, rxmode->hw_vlan_filter);
+ DP_INFO(edev, "VLAN offload mask %d\n", mask);
return 0;
}
@@ -1267,19 +1268,19 @@ static void qede_fastpath_start(struct ecore_dev *edev)
static int qede_dev_start(struct rte_eth_dev *eth_dev)
{
- struct rte_eth_rxmode *rxmode = ð_dev->data->dev_conf.rxmode;
struct qede_dev *qdev = QEDE_INIT_QDEV(eth_dev);
struct ecore_dev *edev = QEDE_INIT_EDEV(qdev);
+ struct rte_eth_rxmode *rxmode = ð_dev->data->dev_conf.rxmode;
PMD_INIT_FUNC_TRACE(edev);
/* Configure TPA parameters */
- if (rxmode->enable_lro) {
+ if (rxmode->offloads & DEV_RX_OFFLOAD_TCP_LRO) {
if (qede_enable_tpa(eth_dev, true))
return -EINVAL;
/* Enable scatter mode for LRO */
- if (!rxmode->enable_scatter)
- eth_dev->data->scattered_rx = 1;
+ if (!eth_dev->data->scattered_rx)
+ rxmode->offloads |= DEV_RX_OFFLOAD_SCATTER;
}
/* Start queues */
@@ -1294,7 +1295,7 @@ static int qede_dev_start(struct rte_eth_dev *eth_dev)
* Also, we would like to retain similar behavior in PF case, so we
* don't do PF/VF specific check here.
*/
- if (rxmode->mq_mode == ETH_MQ_RX_RSS)
+ if (eth_dev->data->dev_conf.rxmode.mq_mode == ETH_MQ_RX_RSS)
if (qede_config_rss(eth_dev))
goto err;
@@ -1411,15 +1412,15 @@ static int qede_dev_configure(struct rte_eth_dev *eth_dev)
/* Check requirements for 100G mode */
if (ECORE_IS_CMT(edev)) {
if (eth_dev->data->nb_rx_queues < 2 ||
- eth_dev->data->nb_tx_queues < 2) {
+ eth_dev->data->nb_tx_queues < 2) {
DP_ERR(edev, "100G mode needs min. 2 RX/TX queues\n");
return -EINVAL;
}
if ((eth_dev->data->nb_rx_queues % 2 != 0) ||
- (eth_dev->data->nb_tx_queues % 2 != 0)) {
+ (eth_dev->data->nb_tx_queues % 2 != 0)) {
DP_ERR(edev,
- "100G mode needs even no. of RX/TX queues\n");
+ "100G mode needs even no. of RX/TX queues\n");
return -EINVAL;
}
}
@@ -1439,20 +1440,8 @@ static int qede_dev_configure(struct rte_eth_dev *eth_dev)
if (qede_args(eth_dev))
return -ENOTSUP;
- /* Sanity checks and throw warnings */
- if (rxmode->enable_scatter)
- eth_dev->data->scattered_rx = 1;
-
- if (!rxmode->hw_strip_crc)
- DP_INFO(edev, "L2 CRC stripping is always enabled in hw\n");
-
- if (!rxmode->hw_ip_checksum)
- DP_INFO(edev, "IP/UDP/TCP checksum offload is always enabled "
- "in hw\n");
- if (rxmode->header_split)
- DP_INFO(edev, "Header split enable is not supported\n");
- if (!(rxmode->mq_mode == ETH_MQ_RX_NONE || rxmode->mq_mode ==
- ETH_MQ_RX_RSS)) {
+ if (!(rxmode->mq_mode == ETH_MQ_RX_NONE ||
+ rxmode->mq_mode == ETH_MQ_RX_RSS)) {
DP_ERR(edev, "Unsupported multi-queue mode\n");
return -ENOTSUP;
}
@@ -1467,19 +1456,23 @@ static int qede_dev_configure(struct rte_eth_dev *eth_dev)
return -ENOMEM;
/* If jumbo enabled adjust MTU */
- if (eth_dev->data->dev_conf.rxmode.jumbo_frame)
+ if (rxmode->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME)
eth_dev->data->mtu =
- eth_dev->data->dev_conf.rxmode.max_rx_pkt_len -
- ETHER_HDR_LEN - ETHER_CRC_LEN;
+ eth_dev->data->dev_conf.rxmode.max_rx_pkt_len -
+ ETHER_HDR_LEN - ETHER_CRC_LEN;
+
+ if (rxmode->offloads & DEV_RX_OFFLOAD_SCATTER)
+ eth_dev->data->scattered_rx = 1;
if (qede_start_vport(qdev, eth_dev->data->mtu))
return -1;
+
qdev->mtu = eth_dev->data->mtu;
/* Enable VLAN offloads by default */
ret = qede_vlan_offload_set(eth_dev, ETH_VLAN_STRIP_MASK |
- ETH_VLAN_FILTER_MASK |
- ETH_VLAN_EXTEND_MASK);
+ ETH_VLAN_FILTER_MASK |
+ ETH_VLAN_EXTEND_MASK);
if (ret)
return ret;
@@ -1534,26 +1527,46 @@ static int qede_dev_configure(struct rte_eth_dev *eth_dev)
dev_info->reta_size = ECORE_RSS_IND_TABLE_SIZE;
dev_info->hash_key_size = ECORE_RSS_KEY_SIZE * sizeof(uint32_t);
dev_info->flow_type_rss_offloads = (uint64_t)QEDE_RSS_OFFLOAD_ALL;
-
- dev_info->default_txconf = (struct rte_eth_txconf) {
- .txq_flags = QEDE_TXQ_FLAGS,
- };
-
- dev_info->rx_offload_capa = (DEV_RX_OFFLOAD_VLAN_STRIP |
- DEV_RX_OFFLOAD_IPV4_CKSUM |
+ dev_info->rx_offload_capa = (DEV_RX_OFFLOAD_IPV4_CKSUM |
DEV_RX_OFFLOAD_UDP_CKSUM |
DEV_RX_OFFLOAD_TCP_CKSUM |
DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM |
- DEV_RX_OFFLOAD_TCP_LRO);
-
+ DEV_RX_OFFLOAD_TCP_LRO |
+ DEV_RX_OFFLOAD_CRC_STRIP |
+ DEV_RX_OFFLOAD_SCATTER |
+ DEV_RX_OFFLOAD_JUMBO_FRAME |
+ DEV_RX_OFFLOAD_VLAN_FILTER |
+ DEV_RX_OFFLOAD_VLAN_STRIP);
+ dev_info->rx_queue_offload_capa = 0;
+
+ /* TX offloads are on a per-packet basis, so it is applicable
+ * to both at port and queue levels.
+ */
dev_info->tx_offload_capa = (DEV_TX_OFFLOAD_VLAN_INSERT |
DEV_TX_OFFLOAD_IPV4_CKSUM |
DEV_TX_OFFLOAD_UDP_CKSUM |
DEV_TX_OFFLOAD_TCP_CKSUM |
DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM |
- DEV_TX_OFFLOAD_TCP_TSO |
+ DEV_TX_OFFLOAD_QINQ_INSERT |
+ DEV_TX_OFFLOAD_MULTI_SEGS |
+ DEV_TX_OFFLOAD_TCP_TSO |
DEV_TX_OFFLOAD_VXLAN_TNL_TSO |
DEV_TX_OFFLOAD_GENEVE_TNL_TSO);
+ dev_info->tx_queue_offload_capa = dev_info->tx_offload_capa;
+
+ dev_info->default_txconf = (struct rte_eth_txconf) {
+ .txq_flags = DEV_TX_OFFLOAD_MULTI_SEGS,
+ };
+
+ dev_info->default_rxconf = (struct rte_eth_rxconf) {
+ /* Packets are always dropped if no descriptors are available */
+ .rx_drop_en = 1,
+ /* The below RX offloads are always enabled */
+ .offloads = (DEV_RX_OFFLOAD_CRC_STRIP |
+ DEV_RX_OFFLOAD_IPV4_CKSUM |
+ DEV_RX_OFFLOAD_TCP_CKSUM |
+ DEV_RX_OFFLOAD_UDP_CKSUM),
+ };
memset(&link, 0, sizeof(struct qed_link_output));
qdev->ops->common->get_link(edev, &link);
diff --git a/drivers/net/qede/qede_rxtx.c b/drivers/net/qede/qede_rxtx.c
index 0de7c6b..e9fe46c 100644
--- a/drivers/net/qede/qede_rxtx.c
+++ b/drivers/net/qede/qede_rxtx.c
@@ -87,7 +87,7 @@ static inline int qede_alloc_rx_buffer(struct qede_rx_queue *rxq)
/* Fix up RX buffer size */
bufsz = (uint16_t)rte_pktmbuf_data_room_size(mp) - RTE_PKTMBUF_HEADROOM;
- if ((rxmode->enable_scatter) ||
+ if ((rxmode->offloads & DEV_RX_OFFLOAD_SCATTER) ||
(max_rx_pkt_len + QEDE_ETH_OVERHEAD) > bufsz) {
if (!dev->data->scattered_rx) {
DP_INFO(edev, "Forcing scatter-gather mode\n");
diff --git a/drivers/net/qede/qede_rxtx.h b/drivers/net/qede/qede_rxtx.h
index f1d3666..a2dc9e7 100644
--- a/drivers/net/qede/qede_rxtx.h
+++ b/drivers/net/qede/qede_rxtx.h
@@ -76,8 +76,6 @@
ETH_RSS_VXLAN |\
ETH_RSS_GENEVE)
-#define QEDE_TXQ_FLAGS ((uint32_t)ETH_TXQ_FLAGS_NOMULTSEGS)
-
#define for_each_rss(i) for (i = 0; i < qdev->num_rx_queues; i++)
#define for_each_tss(i) for (i = 0; i < qdev->num_tx_queues; i++)
#define QEDE_RXTX_MAX(qdev) \
--
1.7.10.3
More information about the dev
mailing list