net/iavf: support flexible Rx descriptor in AVX path
authorLeyi Rong <leyi.rong@intel.com>
Mon, 20 Apr 2020 06:16:18 +0000 (14:16 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Tue, 21 Apr 2020 11:57:09 +0000 (13:57 +0200)
Support flexible Rx descriptor format in AVX
path of iAVF PMD.

Signed-off-by: Leyi Rong <leyi.rong@intel.com>
Reviewed-by: Qi Zhang <qi.z.zhang@intel.com>
drivers/net/iavf/iavf_rxtx.c
drivers/net/iavf/iavf_rxtx.h
drivers/net/iavf/iavf_rxtx_vec_avx2.c

index 8b2528d..e52e4e9 100644 (file)
@@ -2087,16 +2087,28 @@ iavf_set_rx_function(struct rte_eth_dev *dev)
                                    "Using %sVector Scattered Rx (port %d).",
                                    use_avx2 ? "avx2 " : "",
                                    dev->data->port_id);
-                       dev->rx_pkt_burst = use_avx2 ?
-                                           iavf_recv_scattered_pkts_vec_avx2 :
-                                           iavf_recv_scattered_pkts_vec;
+                       if (vf->vf_res->vf_cap_flags &
+                               VIRTCHNL_VF_OFFLOAD_RX_FLEX_DESC)
+                               dev->rx_pkt_burst = use_avx2 ?
+                                       iavf_recv_scattered_pkts_vec_avx2_flex_rxd :
+                                       iavf_recv_scattered_pkts_vec;
+                       else
+                               dev->rx_pkt_burst = use_avx2 ?
+                                       iavf_recv_scattered_pkts_vec_avx2 :
+                                       iavf_recv_scattered_pkts_vec;
                } else {
                        PMD_DRV_LOG(DEBUG, "Using %sVector Rx (port %d).",
                                    use_avx2 ? "avx2 " : "",
                                    dev->data->port_id);
-                       dev->rx_pkt_burst = use_avx2 ?
-                                           iavf_recv_pkts_vec_avx2 :
-                                           iavf_recv_pkts_vec;
+                       if (vf->vf_res->vf_cap_flags &
+                               VIRTCHNL_VF_OFFLOAD_RX_FLEX_DESC)
+                               dev->rx_pkt_burst = use_avx2 ?
+                                       iavf_recv_pkts_vec_avx2_flex_rxd :
+                                       iavf_recv_pkts_vec;
+                       else
+                               dev->rx_pkt_burst = use_avx2 ?
+                                       iavf_recv_pkts_vec_avx2 :
+                                       iavf_recv_pkts_vec;
                }
 
                return;
index 8246e79..84ec391 100644 (file)
@@ -414,9 +414,15 @@ uint16_t iavf_xmit_fixed_burst_vec(void *tx_queue, struct rte_mbuf **tx_pkts,
                                  uint16_t nb_pkts);
 uint16_t iavf_recv_pkts_vec_avx2(void *rx_queue, struct rte_mbuf **rx_pkts,
                                 uint16_t nb_pkts);
+uint16_t iavf_recv_pkts_vec_avx2_flex_rxd(void *rx_queue,
+                                         struct rte_mbuf **rx_pkts,
+                                         uint16_t nb_pkts);
 uint16_t iavf_recv_scattered_pkts_vec_avx2(void *rx_queue,
                                           struct rte_mbuf **rx_pkts,
                                           uint16_t nb_pkts);
+uint16_t iavf_recv_scattered_pkts_vec_avx2_flex_rxd(void *rx_queue,
+                                                   struct rte_mbuf **rx_pkts,
+                                                   uint16_t nb_pkts);
 uint16_t iavf_xmit_pkts_vec(void *tx_queue, struct rte_mbuf **tx_pkts,
                            uint16_t nb_pkts);
 uint16_t iavf_xmit_pkts_vec_avx2(void *tx_queue, struct rte_mbuf **tx_pkts,
index 2587083..04603e7 100644 (file)
@@ -614,6 +614,464 @@ _iavf_recv_raw_pkts_vec_avx2(struct iavf_rx_queue *rxq,
        return received;
 }
 
+static inline uint16_t
+_iavf_recv_raw_pkts_vec_avx2_flex_rxd(struct iavf_rx_queue *rxq,
+                                     struct rte_mbuf **rx_pkts,
+                                     uint16_t nb_pkts, uint8_t *split_packet)
+{
+#define IAVF_DESCS_PER_LOOP_AVX 8
+
+       const uint32_t *type_table = rxq->vsi->adapter->ptype_tbl;
+
+       const __m256i mbuf_init = _mm256_set_epi64x(0, 0,
+                       0, rxq->mbuf_initializer);
+       struct rte_mbuf **sw_ring = &rxq->sw_ring[rxq->rx_tail];
+       volatile union iavf_rx_flex_desc *rxdp =
+               (union iavf_rx_flex_desc *)rxq->rx_ring + rxq->rx_tail;
+
+       rte_prefetch0(rxdp);
+
+       /* nb_pkts has to be floor-aligned to IAVF_DESCS_PER_LOOP_AVX */
+       nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, IAVF_DESCS_PER_LOOP_AVX);
+
+       /* See if we need to rearm the RX queue - gives the prefetch a bit
+        * of time to act
+        */
+       if (rxq->rxrearm_nb > IAVF_RXQ_REARM_THRESH)
+               iavf_rxq_rearm(rxq);
+
+       /* Before we start moving massive data around, check to see if
+        * there is actually a packet available
+        */
+       if (!(rxdp->wb.status_error0 &
+                       rte_cpu_to_le_32(1 << IAVF_RX_FLEX_DESC_STATUS0_DD_S)))
+               return 0;
+
+       /* constants used in processing loop */
+       const __m256i crc_adjust =
+               _mm256_set_epi16
+                       (/* first descriptor */
+                        0, 0, 0,       /* ignore non-length fields */
+                        -rxq->crc_len, /* sub crc on data_len */
+                        0,             /* ignore high-16bits of pkt_len */
+                        -rxq->crc_len, /* sub crc on pkt_len */
+                        0, 0,          /* ignore pkt_type field */
+                        /* second descriptor */
+                        0, 0, 0,       /* ignore non-length fields */
+                        -rxq->crc_len, /* sub crc on data_len */
+                        0,             /* ignore high-16bits of pkt_len */
+                        -rxq->crc_len, /* sub crc on pkt_len */
+                        0, 0           /* ignore pkt_type field */
+                       );
+
+       /* 8 packets DD mask, LSB in each 32-bit value */
+       const __m256i dd_check = _mm256_set1_epi32(1);
+
+       /* 8 packets EOP mask, second-LSB in each 32-bit value */
+       const __m256i eop_check = _mm256_slli_epi32(dd_check,
+                       IAVF_RX_FLEX_DESC_STATUS0_EOF_S);
+
+       /* mask to shuffle from desc. to mbuf (2 descriptors)*/
+       const __m256i shuf_msk =
+               _mm256_set_epi8
+                       (/* first descriptor */
+                        15, 14,
+                        13, 12,        /* octet 12~15, 32 bits rss */
+                        11, 10,        /* octet 10~11, 16 bits vlan_macip */
+                        5, 4,          /* octet 4~5, 16 bits data_len */
+                        0xFF, 0xFF,    /* skip hi 16 bits pkt_len, zero out */
+                        5, 4,          /* octet 4~5, 16 bits pkt_len */
+                        0xFF, 0xFF,    /* pkt_type set as unknown */
+                        0xFF, 0xFF,    /*pkt_type set as unknown */
+                        /* second descriptor */
+                        15, 14,
+                        13, 12,        /* octet 12~15, 32 bits rss */
+                        11, 10,        /* octet 10~11, 16 bits vlan_macip */
+                        5, 4,          /* octet 4~5, 16 bits data_len */
+                        0xFF, 0xFF,    /* skip hi 16 bits pkt_len, zero out */
+                        5, 4,          /* octet 4~5, 16 bits pkt_len */
+                        0xFF, 0xFF,    /* pkt_type set as unknown */
+                        0xFF, 0xFF     /*pkt_type set as unknown */
+                       );
+       /**
+        * compile-time check the above crc and shuffle layout is correct.
+        * NOTE: the first field (lowest address) is given last in set_epi
+        * calls above.
+        */
+       RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, pkt_len) !=
+                       offsetof(struct rte_mbuf, rx_descriptor_fields1) + 4);
+       RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, data_len) !=
+                       offsetof(struct rte_mbuf, rx_descriptor_fields1) + 8);
+       RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, vlan_tci) !=
+                       offsetof(struct rte_mbuf, rx_descriptor_fields1) + 10);
+       RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, hash) !=
+                       offsetof(struct rte_mbuf, rx_descriptor_fields1) + 12);
+
+       /* Status/Error flag masks */
+       /**
+        * mask everything except Checksum Reports, RSS indication
+        * and VLAN indication.
+        * bit6:4 for IP/L4 checksum errors.
+        * bit12 is for RSS indication.
+        * bit13 is for VLAN indication.
+        */
+       const __m256i flags_mask =
+                _mm256_set1_epi32((7 << 4) | (1 << 12) | (1 << 13));
+       /**
+        * data to be shuffled by the result of the flags mask shifted by 4
+        * bits.  This gives use the l3_l4 flags.
+        */
+       const __m256i l3_l4_flags_shuf = _mm256_set_epi8(0, 0, 0, 0, 0, 0, 0, 0,
+                       /* shift right 1 bit to make sure it not exceed 255 */
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_BAD |
+                        PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_BAD |
+                        PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_GOOD |
+                        PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_GOOD |
+                        PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       (PKT_RX_L4_CKSUM_BAD | PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_L4_CKSUM_BAD | PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       (PKT_RX_L4_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_L4_CKSUM_GOOD | PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       /* second 128-bits */
+                       0, 0, 0, 0, 0, 0, 0, 0,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_BAD |
+                        PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_BAD |
+                        PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_GOOD |
+                        PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_GOOD |
+                        PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       (PKT_RX_L4_CKSUM_BAD | PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_L4_CKSUM_BAD | PKT_RX_IP_CKSUM_GOOD) >> 1,
+                       (PKT_RX_L4_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD) >> 1,
+                       (PKT_RX_L4_CKSUM_GOOD | PKT_RX_IP_CKSUM_GOOD) >> 1);
+       const __m256i cksum_mask =
+                _mm256_set1_epi32(PKT_RX_IP_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD |
+                                  PKT_RX_L4_CKSUM_GOOD | PKT_RX_L4_CKSUM_BAD |
+                                  PKT_RX_EIP_CKSUM_BAD);
+       /**
+        * data to be shuffled by result of flag mask, shifted down 12.
+        * If RSS(bit12)/VLAN(bit13) are set,
+        * shuffle moves appropriate flags in place.
+        */
+       const __m256i rss_vlan_flags_shuf = _mm256_set_epi8(0, 0, 0, 0,
+                       0, 0, 0, 0,
+                       0, 0, 0, 0,
+                       PKT_RX_RSS_HASH | PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED,
+                       PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED,
+                       PKT_RX_RSS_HASH, 0,
+                       /* end up 128-bits */
+                       0, 0, 0, 0,
+                       0, 0, 0, 0,
+                       0, 0, 0, 0,
+                       PKT_RX_RSS_HASH | PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED,
+                       PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED,
+                       PKT_RX_RSS_HASH, 0);
+
+       uint16_t i, received;
+
+       for (i = 0, received = 0; i < nb_pkts;
+            i += IAVF_DESCS_PER_LOOP_AVX,
+            rxdp += IAVF_DESCS_PER_LOOP_AVX) {
+               /* step 1, copy over 8 mbuf pointers to rx_pkts array */
+               _mm256_storeu_si256((void *)&rx_pkts[i],
+                                   _mm256_loadu_si256((void *)&sw_ring[i]));
+#ifdef RTE_ARCH_X86_64
+               _mm256_storeu_si256
+                       ((void *)&rx_pkts[i + 4],
+                        _mm256_loadu_si256((void *)&sw_ring[i + 4]));
+#endif
+
+               __m256i raw_desc0_1, raw_desc2_3, raw_desc4_5, raw_desc6_7;
+
+               const __m128i raw_desc7 =
+                       _mm_load_si128((void *)(rxdp + 7));
+               rte_compiler_barrier();
+               const __m128i raw_desc6 =
+                       _mm_load_si128((void *)(rxdp + 6));
+               rte_compiler_barrier();
+               const __m128i raw_desc5 =
+                       _mm_load_si128((void *)(rxdp + 5));
+               rte_compiler_barrier();
+               const __m128i raw_desc4 =
+                       _mm_load_si128((void *)(rxdp + 4));
+               rte_compiler_barrier();
+               const __m128i raw_desc3 =
+                       _mm_load_si128((void *)(rxdp + 3));
+               rte_compiler_barrier();
+               const __m128i raw_desc2 =
+                       _mm_load_si128((void *)(rxdp + 2));
+               rte_compiler_barrier();
+               const __m128i raw_desc1 =
+                       _mm_load_si128((void *)(rxdp + 1));
+               rte_compiler_barrier();
+               const __m128i raw_desc0 =
+                       _mm_load_si128((void *)(rxdp + 0));
+
+               raw_desc6_7 =
+                       _mm256_inserti128_si256
+                               (_mm256_castsi128_si256(raw_desc6),
+                                raw_desc7, 1);
+               raw_desc4_5 =
+                       _mm256_inserti128_si256
+                               (_mm256_castsi128_si256(raw_desc4),
+                                raw_desc5, 1);
+               raw_desc2_3 =
+                       _mm256_inserti128_si256
+                               (_mm256_castsi128_si256(raw_desc2),
+                                raw_desc3, 1);
+               raw_desc0_1 =
+                       _mm256_inserti128_si256
+                               (_mm256_castsi128_si256(raw_desc0),
+                                raw_desc1, 1);
+
+               if (split_packet) {
+                       int j;
+
+                       for (j = 0; j < IAVF_DESCS_PER_LOOP_AVX; j++)
+                               rte_mbuf_prefetch_part2(rx_pkts[i + j]);
+               }
+
+               /**
+                * convert descriptors 4-7 into mbufs, re-arrange fields.
+                * Then write into the mbuf.
+                */
+               __m256i mb6_7 = _mm256_shuffle_epi8(raw_desc6_7, shuf_msk);
+               __m256i mb4_5 = _mm256_shuffle_epi8(raw_desc4_5, shuf_msk);
+
+               mb6_7 = _mm256_add_epi16(mb6_7, crc_adjust);
+               mb4_5 = _mm256_add_epi16(mb4_5, crc_adjust);
+               /**
+                * to get packet types, ptype is located in bit16-25
+                * of each 128bits
+                */
+               const __m256i ptype_mask =
+                       _mm256_set1_epi16(IAVF_RX_FLEX_DESC_PTYPE_M);
+               const __m256i ptypes6_7 =
+                       _mm256_and_si256(raw_desc6_7, ptype_mask);
+               const __m256i ptypes4_5 =
+                       _mm256_and_si256(raw_desc4_5, ptype_mask);
+               const uint16_t ptype7 = _mm256_extract_epi16(ptypes6_7, 9);
+               const uint16_t ptype6 = _mm256_extract_epi16(ptypes6_7, 1);
+               const uint16_t ptype5 = _mm256_extract_epi16(ptypes4_5, 9);
+               const uint16_t ptype4 = _mm256_extract_epi16(ptypes4_5, 1);
+
+               mb6_7 = _mm256_insert_epi32(mb6_7, type_table[ptype7], 4);
+               mb6_7 = _mm256_insert_epi32(mb6_7, type_table[ptype6], 0);
+               mb4_5 = _mm256_insert_epi32(mb4_5, type_table[ptype5], 4);
+               mb4_5 = _mm256_insert_epi32(mb4_5, type_table[ptype4], 0);
+               /* merge the status bits into one register */
+               const __m256i status4_7 = _mm256_unpackhi_epi32(raw_desc6_7,
+                               raw_desc4_5);
+
+               /**
+                * convert descriptors 0-3 into mbufs, re-arrange fields.
+                * Then write into the mbuf.
+                */
+               __m256i mb2_3 = _mm256_shuffle_epi8(raw_desc2_3, shuf_msk);
+               __m256i mb0_1 = _mm256_shuffle_epi8(raw_desc0_1, shuf_msk);
+
+               mb2_3 = _mm256_add_epi16(mb2_3, crc_adjust);
+               mb0_1 = _mm256_add_epi16(mb0_1, crc_adjust);
+               /**
+                * to get packet types, ptype is located in bit16-25
+                * of each 128bits
+                */
+               const __m256i ptypes2_3 =
+                       _mm256_and_si256(raw_desc2_3, ptype_mask);
+               const __m256i ptypes0_1 =
+                       _mm256_and_si256(raw_desc0_1, ptype_mask);
+               const uint16_t ptype3 = _mm256_extract_epi16(ptypes2_3, 9);
+               const uint16_t ptype2 = _mm256_extract_epi16(ptypes2_3, 1);
+               const uint16_t ptype1 = _mm256_extract_epi16(ptypes0_1, 9);
+               const uint16_t ptype0 = _mm256_extract_epi16(ptypes0_1, 1);
+
+               mb2_3 = _mm256_insert_epi32(mb2_3, type_table[ptype3], 4);
+               mb2_3 = _mm256_insert_epi32(mb2_3, type_table[ptype2], 0);
+               mb0_1 = _mm256_insert_epi32(mb0_1, type_table[ptype1], 4);
+               mb0_1 = _mm256_insert_epi32(mb0_1, type_table[ptype0], 0);
+               /* merge the status bits into one register */
+               const __m256i status0_3 = _mm256_unpackhi_epi32(raw_desc2_3,
+                                                               raw_desc0_1);
+
+               /**
+                * take the two sets of status bits and merge to one
+                * After merge, the packets status flags are in the
+                * order (hi->lo): [1, 3, 5, 7, 0, 2, 4, 6]
+                */
+               __m256i status0_7 = _mm256_unpacklo_epi64(status4_7,
+                                                         status0_3);
+
+               /* now do flag manipulation */
+
+               /* get only flag/error bits we want */
+               const __m256i flag_bits =
+                       _mm256_and_si256(status0_7, flags_mask);
+               /**
+                * l3_l4_error flags, shuffle, then shift to correct adjustment
+                * of flags in flags_shuf, and finally mask out extra bits
+                */
+               __m256i l3_l4_flags = _mm256_shuffle_epi8(l3_l4_flags_shuf,
+                               _mm256_srli_epi32(flag_bits, 4));
+               l3_l4_flags = _mm256_slli_epi32(l3_l4_flags, 1);
+               l3_l4_flags = _mm256_and_si256(l3_l4_flags, cksum_mask);
+               /* set rss and vlan flags */
+               const __m256i rss_vlan_flag_bits =
+                       _mm256_srli_epi32(flag_bits, 12);
+               const __m256i rss_vlan_flags =
+                       _mm256_shuffle_epi8(rss_vlan_flags_shuf,
+                                           rss_vlan_flag_bits);
+
+               /* merge flags */
+               const __m256i mbuf_flags = _mm256_or_si256(l3_l4_flags,
+                               rss_vlan_flags);
+               /**
+                * At this point, we have the 8 sets of flags in the low 16-bits
+                * of each 32-bit value in vlan0.
+                * We want to extract these, and merge them with the mbuf init
+                * data so we can do a single write to the mbuf to set the flags
+                * and all the other initialization fields. Extracting the
+                * appropriate flags means that we have to do a shift and blend
+                * for each mbuf before we do the write. However, we can also
+                * add in the previously computed rx_descriptor fields to
+                * make a single 256-bit write per mbuf
+                */
+               /* check the structure matches expectations */
+               RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, ol_flags) !=
+                                offsetof(struct rte_mbuf, rearm_data) + 8);
+               RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, rearm_data) !=
+                                RTE_ALIGN(offsetof(struct rte_mbuf,
+                                                   rearm_data),
+                                          16));
+               /* build up data and do writes */
+               __m256i rearm0, rearm1, rearm2, rearm3, rearm4, rearm5,
+                       rearm6, rearm7;
+               rearm6 = _mm256_blend_epi32(mbuf_init,
+                                           _mm256_slli_si256(mbuf_flags, 8),
+                                           0x04);
+               rearm4 = _mm256_blend_epi32(mbuf_init,
+                                           _mm256_slli_si256(mbuf_flags, 4),
+                                           0x04);
+               rearm2 = _mm256_blend_epi32(mbuf_init, mbuf_flags, 0x04);
+               rearm0 = _mm256_blend_epi32(mbuf_init,
+                                           _mm256_srli_si256(mbuf_flags, 4),
+                                           0x04);
+               /* permute to add in the rx_descriptor e.g. rss fields */
+               rearm6 = _mm256_permute2f128_si256(rearm6, mb6_7, 0x20);
+               rearm4 = _mm256_permute2f128_si256(rearm4, mb4_5, 0x20);
+               rearm2 = _mm256_permute2f128_si256(rearm2, mb2_3, 0x20);
+               rearm0 = _mm256_permute2f128_si256(rearm0, mb0_1, 0x20);
+               /* write to mbuf */
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 6]->rearm_data,
+                                   rearm6);
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 4]->rearm_data,
+                                   rearm4);
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 2]->rearm_data,
+                                   rearm2);
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 0]->rearm_data,
+                                   rearm0);
+
+               /* repeat for the odd mbufs */
+               const __m256i odd_flags =
+                       _mm256_castsi128_si256
+                               (_mm256_extracti128_si256(mbuf_flags, 1));
+               rearm7 = _mm256_blend_epi32(mbuf_init,
+                                           _mm256_slli_si256(odd_flags, 8),
+                                           0x04);
+               rearm5 = _mm256_blend_epi32(mbuf_init,
+                                           _mm256_slli_si256(odd_flags, 4),
+                                           0x04);
+               rearm3 = _mm256_blend_epi32(mbuf_init, odd_flags, 0x04);
+               rearm1 = _mm256_blend_epi32(mbuf_init,
+                                           _mm256_srli_si256(odd_flags, 4),
+                                           0x04);
+               /* since odd mbufs are already in hi 128-bits use blend */
+               rearm7 = _mm256_blend_epi32(rearm7, mb6_7, 0xF0);
+               rearm5 = _mm256_blend_epi32(rearm5, mb4_5, 0xF0);
+               rearm3 = _mm256_blend_epi32(rearm3, mb2_3, 0xF0);
+               rearm1 = _mm256_blend_epi32(rearm1, mb0_1, 0xF0);
+               /* again write to mbufs */
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 7]->rearm_data,
+                                   rearm7);
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 5]->rearm_data,
+                                   rearm5);
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 3]->rearm_data,
+                                   rearm3);
+               _mm256_storeu_si256((__m256i *)&rx_pkts[i + 1]->rearm_data,
+                                   rearm1);
+
+               /* extract and record EOP bit */
+               if (split_packet) {
+                       const __m128i eop_mask =
+                               _mm_set1_epi16(1 <<
+                                              IAVF_RX_FLEX_DESC_STATUS0_EOF_S);
+                       const __m256i eop_bits256 = _mm256_and_si256(status0_7,
+                                                                    eop_check);
+                       /* pack status bits into a single 128-bit register */
+                       const __m128i eop_bits =
+                               _mm_packus_epi32
+                                       (_mm256_castsi256_si128(eop_bits256),
+                                        _mm256_extractf128_si256(eop_bits256,
+                                                                 1));
+                       /**
+                        * flip bits, and mask out the EOP bit, which is now
+                        * a split-packet bit i.e. !EOP, rather than EOP one.
+                        */
+                       __m128i split_bits = _mm_andnot_si128(eop_bits,
+                                       eop_mask);
+                       /**
+                        * eop bits are out of order, so we need to shuffle them
+                        * back into order again. In doing so, only use low 8
+                        * bits, which acts like another pack instruction
+                        * The original order is (hi->lo): 1,3,5,7,0,2,4,6
+                        * [Since we use epi8, the 16-bit positions are
+                        * multiplied by 2 in the eop_shuffle value.]
+                        */
+                       __m128i eop_shuffle =
+                               _mm_set_epi8(/* zero hi 64b */
+                                            0xFF, 0xFF, 0xFF, 0xFF,
+                                            0xFF, 0xFF, 0xFF, 0xFF,
+                                            /* move values to lo 64b */
+                                            8, 0, 10, 2,
+                                            12, 4, 14, 6);
+                       split_bits = _mm_shuffle_epi8(split_bits, eop_shuffle);
+                       *(uint64_t *)split_packet =
+                               _mm_cvtsi128_si64(split_bits);
+                       split_packet += IAVF_DESCS_PER_LOOP_AVX;
+               }
+
+               /* perform dd_check */
+               status0_7 = _mm256_and_si256(status0_7, dd_check);
+               status0_7 = _mm256_packs_epi32(status0_7,
+                                              _mm256_setzero_si256());
+
+               uint64_t burst = __builtin_popcountll
+                                       (_mm_cvtsi128_si64
+                                               (_mm256_extracti128_si256
+                                                       (status0_7, 1)));
+               burst += __builtin_popcountll
+                               (_mm_cvtsi128_si64
+                                       (_mm256_castsi256_si128(status0_7)));
+               received += burst;
+               if (burst != IAVF_DESCS_PER_LOOP_AVX)
+                       break;
+       }
+
+       /* update tail pointers */
+       rxq->rx_tail += received;
+       rxq->rx_tail &= (rxq->nb_rx_desc - 1);
+       if ((rxq->rx_tail & 1) == 1 && received > 1) { /* keep avx2 aligned */
+               rxq->rx_tail--;
+               received--;
+       }
+       rxq->rxrearm_nb += received;
+       return received;
+}
+
 /**
  * Notice:
  * - nb_pkts < IAVF_DESCS_PER_LOOP, just return no packet
@@ -625,6 +1083,18 @@ iavf_recv_pkts_vec_avx2(void *rx_queue, struct rte_mbuf **rx_pkts,
        return _iavf_recv_raw_pkts_vec_avx2(rx_queue, rx_pkts, nb_pkts, NULL);
 }
 
+/**
+ * Notice:
+ * - nb_pkts < IAVF_DESCS_PER_LOOP, just return no packet
+ */
+uint16_t
+iavf_recv_pkts_vec_avx2_flex_rxd(void *rx_queue, struct rte_mbuf **rx_pkts,
+                                uint16_t nb_pkts)
+{
+       return _iavf_recv_raw_pkts_vec_avx2_flex_rxd(rx_queue, rx_pkts,
+                                                    nb_pkts, NULL);
+}
+
 /**
  * vPMD receive routine that reassembles single burst of 32 scattered packets
  * Notice:
@@ -690,6 +1160,75 @@ iavf_recv_scattered_pkts_vec_avx2(void *rx_queue, struct rte_mbuf **rx_pkts,
                                rx_pkts + retval, nb_pkts);
 }
 
+/**
+ * vPMD receive routine that reassembles single burst of
+ * 32 scattered packets for flex RxD
+ * Notice:
+ * - nb_pkts < IAVF_DESCS_PER_LOOP, just return no packet
+ */
+static uint16_t
+iavf_recv_scattered_burst_vec_avx2_flex_rxd(void *rx_queue,
+                                           struct rte_mbuf **rx_pkts,
+                                           uint16_t nb_pkts)
+{
+       struct iavf_rx_queue *rxq = rx_queue;
+       uint8_t split_flags[IAVF_VPMD_RX_MAX_BURST] = {0};
+
+       /* get some new buffers */
+       uint16_t nb_bufs = _iavf_recv_raw_pkts_vec_avx2_flex_rxd(rxq,
+                                       rx_pkts, nb_pkts, split_flags);
+       if (nb_bufs == 0)
+               return 0;
+
+       /* happy day case, full burst + no packets to be joined */
+       const uint64_t *split_fl64 = (uint64_t *)split_flags;
+
+       if (!rxq->pkt_first_seg &&
+           split_fl64[0] == 0 && split_fl64[1] == 0 &&
+           split_fl64[2] == 0 && split_fl64[3] == 0)
+               return nb_bufs;
+
+       /* reassemble any packets that need reassembly*/
+       unsigned int i = 0;
+
+       if (!rxq->pkt_first_seg) {
+               /* find the first split flag, and only reassemble then*/
+               while (i < nb_bufs && !split_flags[i])
+                       i++;
+               if (i == nb_bufs)
+                       return nb_bufs;
+               rxq->pkt_first_seg = rx_pkts[i];
+       }
+       return i + reassemble_packets(rxq, &rx_pkts[i], nb_bufs - i,
+                                            &split_flags[i]);
+}
+
+/**
+ * vPMD receive routine that reassembles scattered packets for flex RxD.
+ * Main receive routine that can handle arbitrary burst sizes
+ * Notice:
+ * - nb_pkts < IAVF_DESCS_PER_LOOP, just return no packet
+ */
+uint16_t
+iavf_recv_scattered_pkts_vec_avx2_flex_rxd(void *rx_queue,
+                                          struct rte_mbuf **rx_pkts,
+                                          uint16_t nb_pkts)
+{
+       uint16_t retval = 0;
+
+       while (nb_pkts > IAVF_VPMD_RX_MAX_BURST) {
+               uint16_t burst =
+                       iavf_recv_scattered_burst_vec_avx2_flex_rxd
+                       (rx_queue, rx_pkts + retval, IAVF_VPMD_RX_MAX_BURST);
+               retval += burst;
+               nb_pkts -= burst;
+               if (burst < IAVF_VPMD_RX_MAX_BURST)
+                       return retval;
+       }
+       return retval + iavf_recv_scattered_burst_vec_avx2_flex_rxd(rx_queue,
+                               rx_pkts + retval, nb_pkts);
+}
+
 static inline void
 iavf_vtx1(volatile struct iavf_tx_desc *txdp,
          struct rte_mbuf *pkt, uint64_t flags)