event/octeontx2: improve single flow performance

[dpdk.git] / drivers / net / octeontx2 / otx2_tx.c

diff --git a/drivers/net/octeontx2/otx2_tx.c b/drivers/net/octeontx2/otx2_tx.c
index 96be92a..1b75cd5 100644 (file)
--- a/drivers/net/octeontx2/otx2_tx.c
+++ b/drivers/net/octeontx2/otx2_tx.c
@@ -39,7 +39,7 @@ nix_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
        }
 
        /* Lets commit any changes in the packet */
-       rte_cio_wmb();
+       rte_io_wmb();
 
        for (i = 0; i < pkts; i++) {
                otx2_nix_xmit_prepare(tx_pkts[i], cmd, flags);
@@ -75,7 +75,7 @@ nix_xmit_pkts_mseg(void *tx_queue, struct rte_mbuf **tx_pkts,
        }
 
        /* Lets commit any changes in the packet */
-       rte_cio_wmb();
+       rte_io_wmb();
 
        for (i = 0; i < pkts; i++) {
                otx2_nix_xmit_prepare(tx_pkts[i], cmd, flags);
@@ -112,7 +112,6 @@ nix_xmit_pkts_vector(void *tx_queue, struct rte_mbuf **tx_pkts,
        uint64x2_t ltypes01, ltypes23;
        uint64x2_t xtmp128, ytmp128;
        uint64x2_t xmask01, xmask23;
-       uint64x2_t mbuf01, mbuf23;
        uint64x2_t cmd00, cmd01;
        uint64x2_t cmd10, cmd11;
        uint64x2_t cmd20, cmd21;
@@ -129,7 +128,7 @@ nix_xmit_pkts_vector(void *tx_queue, struct rte_mbuf **tx_pkts,
        txq->fc_cache_pkts -= pkts;
 
        /* Lets commit any changes in the packet */
-       rte_cio_wmb();
+       rte_io_wmb();
 
        senddesc01_w0 = vld1q_dup_u64(&txq->cmd[0]);
        senddesc23_w0 = senddesc01_w0;
@@ -139,9 +138,6 @@ nix_xmit_pkts_vector(void *tx_queue, struct rte_mbuf **tx_pkts,
        sgdesc23_w0 = sgdesc01_w0;
 
        for (i = 0; i < pkts; i += NIX_DESCS_PER_LOOP) {
-               mbuf01 = vld1q_u64((uint64_t *)tx_pkts);
-               mbuf23 = vld1q_u64((uint64_t *)(tx_pkts + 2));
-
                /* Clear lower 32bit of SEND_HDR_W0 and SEND_SG_W0 */
                senddesc01_w0 = vbicq_u64(senddesc01_w0,
                                          vdupq_n_u64(0xFFFFFFFF));
@@ -151,13 +147,11 @@ nix_xmit_pkts_vector(void *tx_queue, struct rte_mbuf **tx_pkts,
                senddesc23_w0 = senddesc01_w0;
                sgdesc23_w0 = sgdesc01_w0;
 
-               tx_pkts = tx_pkts + NIX_DESCS_PER_LOOP;
-
                /* Move mbufs to iova */
-               mbuf0 = (uint64_t *)vgetq_lane_u64(mbuf01, 0);
-               mbuf1 = (uint64_t *)vgetq_lane_u64(mbuf01, 1);
-               mbuf2 = (uint64_t *)vgetq_lane_u64(mbuf23, 0);
-               mbuf3 = (uint64_t *)vgetq_lane_u64(mbuf23, 1);
+               mbuf0 = (uint64_t *)tx_pkts[0];
+               mbuf1 = (uint64_t *)tx_pkts[1];
+               mbuf2 = (uint64_t *)tx_pkts[2];
+               mbuf3 = (uint64_t *)tx_pkts[3];
 
                mbuf0 = (uint64_t *)((uintptr_t)mbuf0 +
                                     offsetof(struct rte_mbuf, buf_iova));
@@ -929,6 +923,7 @@ nix_xmit_pkts_vector(void *tx_queue, struct rte_mbuf **tx_pkts,
                        lmt_status = otx2_lmt_submit(io_addr);
 
                } while (lmt_status == 0);
+               tx_pkts = tx_pkts + NIX_DESCS_PER_LOOP;
        }
 
        if (unlikely(pkts_left))
@@ -951,8 +946,8 @@ nix_xmit_pkts_vector(void *tx_queue, struct rte_mbuf **tx_pkts,
 }
 #endif
 
-#define T(name, f5, f4, f3, f2, f1, f0, sz, flags)                     \
-static uint16_t __rte_noinline __hot                                   \
+#define T(name, f6, f5, f4, f3, f2, f1, f0, sz, flags)                 \
+static uint16_t __rte_noinline __rte_hot                                       \
 otx2_nix_xmit_pkts_ ## name(void *tx_queue,                            \
                        struct rte_mbuf **tx_pkts, uint16_t pkts)       \
 {                                                                      \
@@ -968,8 +963,8 @@ otx2_nix_xmit_pkts_ ## name(void *tx_queue,                         \
 NIX_TX_FASTPATH_MODES
 #undef T
 
-#define T(name, f5, f4, f3, f2, f1, f0, sz, flags)                     \
-static uint16_t __rte_noinline __hot                                   \
+#define T(name, f6, f5, f4, f3, f2, f1, f0, sz, flags)                 \
+static uint16_t __rte_noinline __rte_hot                                       \
 otx2_nix_xmit_pkts_mseg_ ## name(void *tx_queue,                       \
                        struct rte_mbuf **tx_pkts, uint16_t pkts)       \
 {                                                                      \
@@ -986,8 +981,8 @@ otx2_nix_xmit_pkts_mseg_ ## name(void *tx_queue,                    \
 NIX_TX_FASTPATH_MODES
 #undef T
 
-#define T(name, f5, f4, f3, f2, f1, f0, sz, flags)                     \
-static uint16_t __rte_noinline __hot                                   \
+#define T(name, f6, f5, f4, f3, f2, f1, f0, sz, flags)                 \
+static uint16_t __rte_noinline __rte_hot                                       \
 otx2_nix_xmit_pkts_vec_ ## name(void *tx_queue,                                \
                        struct rte_mbuf **tx_pkts, uint16_t pkts)       \
 {                                                                      \
@@ -1006,12 +1001,13 @@ NIX_TX_FASTPATH_MODES
 
 static inline void
 pick_tx_func(struct rte_eth_dev *eth_dev,
-            const eth_tx_burst_t tx_burst[2][2][2][2][2][2])
+            const eth_tx_burst_t tx_burst[2][2][2][2][2][2][2])
 {
        struct otx2_eth_dev *dev = otx2_eth_pmd_priv(eth_dev);
 
-       /* [TSTMP] [NOFF] [VLAN] [OL3_OL4_CSUM] [IL3_IL4_CSUM] */
+       /* [SEC] [TSTMP] [NOFF] [VLAN] [OL3_OL4_CSUM] [IL3_IL4_CSUM] */
        eth_dev->tx_pkt_burst = tx_burst
+               [!!(dev->tx_offload_flags & NIX_TX_OFFLOAD_SECURITY_F)]
                [!!(dev->tx_offload_flags & NIX_TX_OFFLOAD_TSO_F)]
                [!!(dev->tx_offload_flags & NIX_TX_OFFLOAD_TSTAMP_F)]
                [!!(dev->tx_offload_flags & NIX_TX_OFFLOAD_MBUF_NOFF_F)]
@@ -1025,25 +1021,25 @@ otx2_eth_set_tx_function(struct rte_eth_dev *eth_dev)
 {
        struct otx2_eth_dev *dev = otx2_eth_pmd_priv(eth_dev);
 
-       const eth_tx_burst_t nix_eth_tx_burst[2][2][2][2][2][2] = {
-#define T(name, f5, f4, f3, f2, f1, f0, sz, flags)                     \
-       [f5][f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_ ## name,
+       const eth_tx_burst_t nix_eth_tx_burst[2][2][2][2][2][2][2] = {
+#define T(name, f6, f5, f4, f3, f2, f1, f0, sz, flags)                 \
+       [f6][f5][f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_ ## name,
 
 NIX_TX_FASTPATH_MODES
 #undef T
        };
 
-       const eth_tx_burst_t nix_eth_tx_burst_mseg[2][2][2][2][2][2] = {
-#define T(name, f5, f4, f3, f2, f1, f0, sz, flags)                     \
-       [f5][f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_mseg_ ## name,
+       const eth_tx_burst_t nix_eth_tx_burst_mseg[2][2][2][2][2][2][2] = {
+#define T(name, f6, f5, f4, f3, f2, f1, f0, sz, flags)                 \
+       [f6][f5][f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_mseg_ ## name,
 
 NIX_TX_FASTPATH_MODES
 #undef T
        };
 
-       const eth_tx_burst_t nix_eth_tx_vec_burst[2][2][2][2][2][2] = {
-#define T(name, f5, f4, f3, f2, f1, f0, sz, flags)                     \
-       [f5][f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_vec_ ## name,
+       const eth_tx_burst_t nix_eth_tx_vec_burst[2][2][2][2][2][2][2] = {
+#define T(name, f6, f5, f4, f3, f2, f1, f0, sz, flags)                 \
+       [f6][f5][f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_vec_ ## name,
 
 NIX_TX_FASTPATH_MODES
 #undef T