net/i40e: improve scalar Tx performance

[dpdk.git] / drivers / net / hns3 / hns3_rxtx.c

diff --git a/drivers/net/hns3/hns3_rxtx.c b/drivers/net/hns3/hns3_rxtx.c
index 9bb30fc..1d7a769 100644 (file)
--- a/drivers/net/hns3/hns3_rxtx.c
+++ b/drivers/net/hns3/hns3_rxtx.c
@@ -13,6 +13,7 @@
 #include <rte_malloc.h>
 #if defined(RTE_ARCH_ARM64)
 #include <rte_cpuflags.h>
+#include <rte_vect.h>
 #endif
 
 #include "hns3_ethdev.h"
@@ -2402,7 +2403,6 @@ hns3_recv_pkts_simple(void *rx_queue,
        struct rte_mbuf *nmb;           /* pointer of the new mbuf */
        struct rte_mbuf *rxm;
        uint32_t bd_base_info;
-       uint32_t cksum_err;
        uint32_t l234_info;
        uint32_t ol_info;
        uint64_t dma_addr;
@@ -2477,8 +2477,7 @@ hns3_recv_pkts_simple(void *rx_queue,
                /* Load remained descriptor data and extract necessary fields */
                l234_info = rte_le_to_cpu_32(rxd.rx.l234_info);
                ol_info = rte_le_to_cpu_32(rxd.rx.ol_info);
-               ret = hns3_handle_bdinfo(rxq, rxm, bd_base_info,
-                                        l234_info, &cksum_err);
+               ret = hns3_handle_bdinfo(rxq, rxm, bd_base_info, l234_info);
                if (unlikely(ret))
                        goto pkt_err;
 
@@ -2487,9 +2486,6 @@ hns3_recv_pkts_simple(void *rx_queue,
                if (rxm->packet_type == RTE_PTYPE_L2_ETHER_TIMESYNC)
                        rxm->ol_flags |= PKT_RX_IEEE1588_PTP;
 
-               if (likely(bd_base_info & BIT(HNS3_RXD_L3L4P_B)))
-                       hns3_rx_set_cksum_flag(rxm, rxm->packet_type,
-                                              cksum_err);
                hns3_rxd_to_vlan_tci(rxq, rxm, l234_info, &rxd);
 
                /* Increment bytes counter  */
@@ -2528,7 +2524,6 @@ hns3_recv_scattered_pkts(void *rx_queue,
        struct rte_mbuf *rxm;
        struct rte_eth_dev *dev;
        uint32_t bd_base_info;
-       uint32_t cksum_err;
        uint32_t l234_info;
        uint32_t gro_size;
        uint32_t ol_info;
@@ -2659,6 +2654,9 @@ hns3_recv_scattered_pkts(void *rx_queue,
                        continue;
                }
 
+               if (unlikely(bd_base_info & BIT(HNS3_RXD_TS_VLD_B)))
+                       hns3_rx_ptp_timestamp_handle(rxq, first_seg, rxdp);
+
                /*
                 * The last buffer of the received packet. packet len from
                 * buffer description may contains CRC len, packet len should
@@ -2702,17 +2700,16 @@ hns3_recv_scattered_pkts(void *rx_queue,
                l234_info = rte_le_to_cpu_32(rxd.rx.l234_info);
                ol_info = rte_le_to_cpu_32(rxd.rx.ol_info);
                ret = hns3_handle_bdinfo(rxq, first_seg, bd_base_info,
-                                        l234_info, &cksum_err);
+                                        l234_info);
                if (unlikely(ret))
                        goto pkt_err;
 
                first_seg->packet_type = hns3_rx_calc_ptype(rxq,
                                                l234_info, ol_info);
 
-               if (bd_base_info & BIT(HNS3_RXD_L3L4P_B))
-                       hns3_rx_set_cksum_flag(first_seg,
-                                              first_seg->packet_type,
-                                              cksum_err);
+               if (first_seg->packet_type == RTE_PTYPE_L2_ETHER_TIMESYNC)
+                       rxm->ol_flags |= PKT_RX_IEEE1588_PTP;
+
                hns3_rxd_to_vlan_tci(rxq, first_seg, l234_info, &rxd);
 
                /* Increment bytes counter */
@@ -2800,6 +2797,8 @@ static bool
 hns3_get_default_vec_support(void)
 {
 #if defined(RTE_ARCH_ARM64)
+       if (rte_vect_get_max_simd_bitwidth() < RTE_VECT_SIMD_128)
+               return false;
        if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_NEON))
                return true;
 #endif
@@ -2810,6 +2809,8 @@ static bool
 hns3_get_sve_support(void)
 {
 #if defined(RTE_ARCH_ARM64) && defined(__ARM_FEATURE_SVE)
+       if (rte_vect_get_max_simd_bitwidth() < RTE_VECT_SIMD_256)
+               return false;
        if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_SVE))
                return true;
 #endif
@@ -4208,17 +4209,46 @@ hns3_tx_check_simple_support(struct rte_eth_dev *dev)
        return (offloads == (offloads & DEV_TX_OFFLOAD_MBUF_FAST_FREE));
 }
 
+static bool
+hns3_get_tx_prep_needed(struct rte_eth_dev *dev)
+{
+#ifdef RTE_LIBRTE_ETHDEV_DEBUG
+       RTE_SET_USED(dev);
+       /* always perform tx_prepare when debug */
+       return true;
+#else
+#define HNS3_DEV_TX_CSKUM_TSO_OFFLOAD_MASK (\
+               DEV_TX_OFFLOAD_IPV4_CKSUM | \
+               DEV_TX_OFFLOAD_TCP_CKSUM | \
+               DEV_TX_OFFLOAD_UDP_CKSUM | \
+               DEV_TX_OFFLOAD_SCTP_CKSUM | \
+               DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM | \
+               DEV_TX_OFFLOAD_OUTER_UDP_CKSUM | \
+               DEV_TX_OFFLOAD_TCP_TSO | \
+               DEV_TX_OFFLOAD_VXLAN_TNL_TSO | \
+               DEV_TX_OFFLOAD_GRE_TNL_TSO | \
+               DEV_TX_OFFLOAD_GENEVE_TNL_TSO)
+
+       uint64_t tx_offload = dev->data->dev_conf.txmode.offloads;
+       if (tx_offload & HNS3_DEV_TX_CSKUM_TSO_OFFLOAD_MASK)
+               return true;
+
+       return false;
+#endif
+}
+
 static eth_tx_burst_t
 hns3_get_tx_function(struct rte_eth_dev *dev, eth_tx_prep_t *prep)
 {
        struct hns3_adapter *hns = dev->data->dev_private;
        bool vec_allowed, sve_allowed, simple_allowed;
-       bool vec_support;
+       bool vec_support, tx_prepare_needed;
 
        vec_support = hns3_tx_check_vec_support(dev) == 0;
        vec_allowed = vec_support && hns3_get_default_vec_support();
        sve_allowed = vec_support && hns3_get_sve_support();
        simple_allowed = hns3_tx_check_simple_support(dev);
+       tx_prepare_needed = hns3_get_tx_prep_needed(dev);
 
        *prep = NULL;
 
@@ -4229,7 +4259,8 @@ hns3_get_tx_function(struct rte_eth_dev *dev, eth_tx_prep_t *prep)
        if (hns->tx_func_hint == HNS3_IO_FUNC_HINT_SIMPLE && simple_allowed)
                return hns3_xmit_pkts_simple;
        if (hns->tx_func_hint == HNS3_IO_FUNC_HINT_COMMON) {
-               *prep = hns3_prep_pkts;
+               if (tx_prepare_needed)
+                       *prep = hns3_prep_pkts;
                return hns3_xmit_pkts;
        }
 
@@ -4238,7 +4269,8 @@ hns3_get_tx_function(struct rte_eth_dev *dev, eth_tx_prep_t *prep)
        if (simple_allowed)
                return hns3_xmit_pkts_simple;
 
-       *prep = hns3_prep_pkts;
+       if (tx_prepare_needed)
+               *prep = hns3_prep_pkts;
        return hns3_xmit_pkts;
 }