* [PATCH 3/3] net/iavf: support vlan insertion offload for the avx-512 path
2025-06-19 13:36 [PATCH 1/3] net/iavf: support qinq insertion offload for scalar path Ciara Loftus
2025-06-19 13:36 ` [PATCH 2/3] net/iavf: fix tx vector path selection logic Ciara Loftus
@ 2025-06-19 13:36 ` Ciara Loftus
1 sibling, 0 replies; 3+ messages in thread
From: Ciara Loftus @ 2025-06-19 13:36 UTC (permalink / raw)
To: dev; +Cc: vladimir.medvedkin, Ciara Loftus
Re-enable support for VLAN insertion and implement support for QinQ
insertion on the iavf avx-512 transmit path.
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
---
drivers/net/intel/iavf/iavf_rxtx.h | 6 +-
drivers/net/intel/iavf/iavf_rxtx_vec_avx2.c | 24 +++----
drivers/net/intel/iavf/iavf_rxtx_vec_avx512.c | 62 ++++++++++++++-----
drivers/net/intel/iavf/iavf_rxtx_vec_common.h | 17 ++++-
4 files changed, 76 insertions(+), 33 deletions(-)
diff --git a/drivers/net/intel/iavf/iavf_rxtx.h b/drivers/net/intel/iavf/iavf_rxtx.h
index 0b5d67e718..d04722a5ed 100644
--- a/drivers/net/intel/iavf/iavf_rxtx.h
+++ b/drivers/net/intel/iavf/iavf_rxtx.h
@@ -30,8 +30,6 @@
#define IAVF_VPMD_TX_MAX_FREE_BUF 64
#define IAVF_TX_NO_VECTOR_FLAGS ( \
- RTE_ETH_TX_OFFLOAD_VLAN_INSERT | \
- RTE_ETH_TX_OFFLOAD_QINQ_INSERT | \
RTE_ETH_TX_OFFLOAD_MULTI_SEGS | \
RTE_ETH_TX_OFFLOAD_TCP_TSO | \
RTE_ETH_TX_OFFLOAD_VXLAN_TNL_TSO | \
@@ -48,7 +46,9 @@
#define IAVF_TX_VECTOR_OFFLOAD_CTX ( \
RTE_ETH_TX_OFFLOAD_OUTER_IPV4_CKSUM | \
- RTE_ETH_TX_OFFLOAD_OUTER_UDP_CKSUM)
+ RTE_ETH_TX_OFFLOAD_OUTER_UDP_CKSUM | \
+ RTE_ETH_TX_OFFLOAD_VLAN_INSERT | \
+ RTE_ETH_TX_OFFLOAD_QINQ_INSERT)
#define IAVF_RX_VECTOR_OFFLOAD ( \
RTE_ETH_RX_OFFLOAD_CHECKSUM | \
diff --git a/drivers/net/intel/iavf/iavf_rxtx_vec_avx2.c b/drivers/net/intel/iavf/iavf_rxtx_vec_avx2.c
index 88e35dc3e9..d08a3ac269 100644
--- a/drivers/net/intel/iavf/iavf_rxtx_vec_avx2.c
+++ b/drivers/net/intel/iavf/iavf_rxtx_vec_avx2.c
@@ -1649,14 +1649,14 @@ iavf_recv_scattered_pkts_vec_avx2_flex_rxd_offload(void *rx_queue,
static __rte_always_inline void
iavf_vtx1(volatile struct iavf_tx_desc *txdp,
- struct rte_mbuf *pkt, uint64_t flags, bool offload)
+ struct rte_mbuf *pkt, uint64_t flags, bool offload, uint8_t vlan_flag)
{
uint64_t high_qw =
(IAVF_TX_DESC_DTYPE_DATA |
((uint64_t)flags << IAVF_TXD_QW1_CMD_SHIFT) |
((uint64_t)pkt->data_len << IAVF_TXD_QW1_TX_BUF_SZ_SHIFT));
if (offload)
- iavf_txd_enable_offload(pkt, &high_qw);
+ iavf_txd_enable_offload(pkt, &high_qw, vlan_flag);
__m128i descriptor = _mm_set_epi64x(high_qw,
pkt->buf_iova + pkt->data_off);
@@ -1665,14 +1665,14 @@ iavf_vtx1(volatile struct iavf_tx_desc *txdp,
static __rte_always_inline void
iavf_vtx(volatile struct iavf_tx_desc *txdp,
- struct rte_mbuf **pkt, uint16_t nb_pkts, uint64_t flags, bool offload)
+ struct rte_mbuf **pkt, uint16_t nb_pkts, uint64_t flags, bool offload, uint8_t vlan_flag)
{
const uint64_t hi_qw_tmpl = (IAVF_TX_DESC_DTYPE_DATA |
((uint64_t)flags << IAVF_TXD_QW1_CMD_SHIFT));
/* if unaligned on 32-bit boundary, do one to align */
if (((uintptr_t)txdp & 0x1F) != 0 && nb_pkts != 0) {
- iavf_vtx1(txdp, *pkt, flags, offload);
+ iavf_vtx1(txdp, *pkt, flags, offload, vlan_flag);
nb_pkts--, txdp++, pkt++;
}
@@ -1683,25 +1683,25 @@ iavf_vtx(volatile struct iavf_tx_desc *txdp,
((uint64_t)pkt[3]->data_len <<
IAVF_TXD_QW1_TX_BUF_SZ_SHIFT);
if (offload)
- iavf_txd_enable_offload(pkt[3], &hi_qw3);
+ iavf_txd_enable_offload(pkt[3], &hi_qw3, vlan_flag);
uint64_t hi_qw2 =
hi_qw_tmpl |
((uint64_t)pkt[2]->data_len <<
IAVF_TXD_QW1_TX_BUF_SZ_SHIFT);
if (offload)
- iavf_txd_enable_offload(pkt[2], &hi_qw2);
+ iavf_txd_enable_offload(pkt[2], &hi_qw2, vlan_flag);
uint64_t hi_qw1 =
hi_qw_tmpl |
((uint64_t)pkt[1]->data_len <<
IAVF_TXD_QW1_TX_BUF_SZ_SHIFT);
if (offload)
- iavf_txd_enable_offload(pkt[1], &hi_qw1);
+ iavf_txd_enable_offload(pkt[1], &hi_qw1, vlan_flag);
uint64_t hi_qw0 =
hi_qw_tmpl |
((uint64_t)pkt[0]->data_len <<
IAVF_TXD_QW1_TX_BUF_SZ_SHIFT);
if (offload)
- iavf_txd_enable_offload(pkt[0], &hi_qw0);
+ iavf_txd_enable_offload(pkt[0], &hi_qw0, vlan_flag);
__m256i desc2_3 =
_mm256_set_epi64x
@@ -1721,7 +1721,7 @@ iavf_vtx(volatile struct iavf_tx_desc *txdp,
/* do any last ones */
while (nb_pkts) {
- iavf_vtx1(txdp, *pkt, flags, offload);
+ iavf_vtx1(txdp, *pkt, flags, offload, vlan_flag);
txdp++, pkt++, nb_pkts--;
}
}
@@ -1756,11 +1756,11 @@ iavf_xmit_fixed_burst_vec_avx2(void *tx_queue, struct rte_mbuf **tx_pkts,
if (nb_commit >= n) {
ci_tx_backlog_entry_vec(txep, tx_pkts, n);
- iavf_vtx(txdp, tx_pkts, n - 1, flags, offload);
+ iavf_vtx(txdp, tx_pkts, n - 1, flags, offload, txq->vlan_flag);
tx_pkts += (n - 1);
txdp += (n - 1);
- iavf_vtx1(txdp, *tx_pkts++, rs, offload);
+ iavf_vtx1(txdp, *tx_pkts++, rs, offload, txq->vlan_flag);
nb_commit = (uint16_t)(nb_commit - n);
@@ -1774,7 +1774,7 @@ iavf_xmit_fixed_burst_vec_avx2(void *tx_queue, struct rte_mbuf **tx_pkts,
ci_tx_backlog_entry_vec(txep, tx_pkts, nb_commit);
- iavf_vtx(txdp, tx_pkts, nb_commit, flags, offload);
+ iavf_vtx(txdp, tx_pkts, nb_commit, flags, offload, txq->vlan_flag);
tx_id = (uint16_t)(tx_id + nb_commit);
if (tx_id > txq->tx_next_rs) {
diff --git a/drivers/net/intel/iavf/iavf_rxtx_vec_avx512.c b/drivers/net/intel/iavf/iavf_rxtx_vec_avx512.c
index f2af028bef..ad21ada440 100644
--- a/drivers/net/intel/iavf/iavf_rxtx_vec_avx512.c
+++ b/drivers/net/intel/iavf/iavf_rxtx_vec_avx512.c
@@ -1853,14 +1853,14 @@ tx_backlog_entry_avx512(struct ci_tx_entry_vec *txep,
static __rte_always_inline void
iavf_vtx1(volatile struct iavf_tx_desc *txdp,
struct rte_mbuf *pkt, uint64_t flags,
- bool offload)
+ bool offload, uint8_t vlan_flag)
{
uint64_t high_qw =
(IAVF_TX_DESC_DTYPE_DATA |
((uint64_t)flags << IAVF_TXD_QW1_CMD_SHIFT) |
((uint64_t)pkt->data_len << IAVF_TXD_QW1_TX_BUF_SZ_SHIFT));
if (offload)
- iavf_txd_enable_offload(pkt, &high_qw);
+ iavf_txd_enable_offload(pkt, &high_qw, vlan_flag);
__m128i descriptor = _mm_set_epi64x(high_qw,
pkt->buf_iova + pkt->data_off);
@@ -1872,14 +1872,14 @@ iavf_vtx1(volatile struct iavf_tx_desc *txdp,
static __rte_always_inline void
iavf_vtx(volatile struct iavf_tx_desc *txdp,
struct rte_mbuf **pkt, uint16_t nb_pkts, uint64_t flags,
- bool offload)
+ bool offload, uint8_t vlan_flag)
{
const uint64_t hi_qw_tmpl = (IAVF_TX_DESC_DTYPE_DATA |
((uint64_t)flags << IAVF_TXD_QW1_CMD_SHIFT));
/* if unaligned on 32-bit boundary, do one to align */
if (((uintptr_t)txdp & 0x1F) != 0 && nb_pkts != 0) {
- iavf_vtx1(txdp, *pkt, flags, offload);
+ iavf_vtx1(txdp, *pkt, flags, offload, vlan_flag);
nb_pkts--, txdp++, pkt++;
}
@@ -1902,10 +1902,10 @@ iavf_vtx(volatile struct iavf_tx_desc *txdp,
((uint64_t)pkt[0]->data_len <<
IAVF_TXD_QW1_TX_BUF_SZ_SHIFT);
if (offload) {
- iavf_txd_enable_offload(pkt[3], &hi_qw3);
- iavf_txd_enable_offload(pkt[2], &hi_qw2);
- iavf_txd_enable_offload(pkt[1], &hi_qw1);
- iavf_txd_enable_offload(pkt[0], &hi_qw0);
+ iavf_txd_enable_offload(pkt[3], &hi_qw3, vlan_flag);
+ iavf_txd_enable_offload(pkt[2], &hi_qw2, vlan_flag);
+ iavf_txd_enable_offload(pkt[1], &hi_qw1, vlan_flag);
+ iavf_txd_enable_offload(pkt[0], &hi_qw0, vlan_flag);
}
__m512i desc0_3 =
@@ -1923,7 +1923,7 @@ iavf_vtx(volatile struct iavf_tx_desc *txdp,
/* do any last ones */
while (nb_pkts) {
- iavf_vtx1(txdp, *pkt, flags, offload);
+ iavf_vtx1(txdp, *pkt, flags, offload, vlan_flag);
txdp++, pkt++, nb_pkts--;
}
}
@@ -2101,7 +2101,7 @@ ctx_vtx1(volatile struct iavf_tx_desc *txdp, struct rte_mbuf *pkt,
((uint64_t)flags << IAVF_TXD_QW1_CMD_SHIFT) |
((uint64_t)pkt->data_len << IAVF_TXD_QW1_TX_BUF_SZ_SHIFT));
if (offload)
- iavf_txd_enable_offload(pkt, &high_data_qw);
+ iavf_txd_enable_offload(pkt, &high_data_qw, vlan_flag);
__m256i ctx_data_desc = _mm256_set_epi64x(high_data_qw, pkt->buf_iova + pkt->data_off,
high_ctx_qw, low_ctx_qw);
@@ -2149,6 +2149,22 @@ ctx_vtx(volatile struct iavf_tx_desc *txdp,
(uint64_t)pkt[1]->vlan_tci << IAVF_TXD_QW1_L2TAG1_SHIFT;
}
}
+ if (pkt[1]->ol_flags & RTE_MBUF_F_TX_QINQ) {
+ hi_ctx_qw1 |= IAVF_TX_CTX_DESC_IL2TAG2 << IAVF_TXD_CTX_QW1_CMD_SHIFT;
+ if (vlan_flag & IAVF_TX_FLAGS_VLAN_TAG_LOC_L2TAG2) {
+ /* Inner tag at L2TAG2, outer tag at L2TAG1. */
+ low_ctx_qw1 |= (uint64_t)pkt[1]->vlan_tci <<
+ IAVF_TXD_CTX_QW0_L2TAG2_PARAM;
+ hi_data_qw1 |= (uint64_t)pkt[1]->vlan_tci_outer <<
+ IAVF_TXD_QW1_L2TAG1_SHIFT;
+ } else {
+ /* Outer tag at L2TAG2, inner tag at L2TAG1. */
+ low_ctx_qw1 |= (uint64_t)pkt[1]->vlan_tci_outer <<
+ IAVF_TXD_CTX_QW0_L2TAG2_PARAM;
+ hi_data_qw1 |= (uint64_t)pkt[1]->vlan_tci <<
+ IAVF_TXD_QW1_L2TAG1_SHIFT;
+ }
+ }
if (IAVF_CHECK_TX_LLDP(pkt[1]))
hi_ctx_qw1 |= IAVF_TX_CTX_DESC_SWTCH_UPLINK
<< IAVF_TXD_CTX_QW1_CMD_SHIFT;
@@ -2164,13 +2180,29 @@ ctx_vtx(volatile struct iavf_tx_desc *txdp,
(uint64_t)pkt[0]->vlan_tci << IAVF_TXD_QW1_L2TAG1_SHIFT;
}
}
+ if (pkt[0]->ol_flags & RTE_MBUF_F_TX_QINQ) {
+ hi_ctx_qw0 |= IAVF_TX_CTX_DESC_IL2TAG2 << IAVF_TXD_CTX_QW1_CMD_SHIFT;
+ if (vlan_flag & IAVF_TX_FLAGS_VLAN_TAG_LOC_L2TAG2) {
+ /* Inner tag at L2TAG2, outer tag at L2TAG1. */
+ low_ctx_qw0 |= (uint64_t)pkt[0]->vlan_tci <<
+ IAVF_TXD_CTX_QW0_L2TAG2_PARAM;
+ hi_data_qw0 |= (uint64_t)pkt[0]->vlan_tci_outer <<
+ IAVF_TXD_QW1_L2TAG1_SHIFT;
+ } else {
+ /* Outer tag at L2TAG2, inner tag at L2TAG1. */
+ low_ctx_qw0 |= (uint64_t)pkt[0]->vlan_tci_outer <<
+ IAVF_TXD_CTX_QW0_L2TAG2_PARAM;
+ hi_data_qw0 |= (uint64_t)pkt[0]->vlan_tci <<
+ IAVF_TXD_QW1_L2TAG1_SHIFT;
+ }
+ }
if (IAVF_CHECK_TX_LLDP(pkt[0]))
hi_ctx_qw0 |= IAVF_TX_CTX_DESC_SWTCH_UPLINK
<< IAVF_TXD_CTX_QW1_CMD_SHIFT;
if (offload) {
- iavf_txd_enable_offload(pkt[1], &hi_data_qw1);
- iavf_txd_enable_offload(pkt[0], &hi_data_qw0);
+ iavf_txd_enable_offload(pkt[1], &hi_data_qw1, vlan_flag);
+ iavf_txd_enable_offload(pkt[0], &hi_data_qw0, vlan_flag);
iavf_fill_ctx_desc_tunnelling_field(&low_ctx_qw1, pkt[1]);
iavf_fill_ctx_desc_tunnelling_field(&low_ctx_qw0, pkt[0]);
}
@@ -2219,11 +2251,11 @@ iavf_xmit_fixed_burst_vec_avx512(void *tx_queue, struct rte_mbuf **tx_pkts,
if (nb_commit >= n) {
tx_backlog_entry_avx512(txep, tx_pkts, n);
- iavf_vtx(txdp, tx_pkts, n - 1, flags, offload);
+ iavf_vtx(txdp, tx_pkts, n - 1, flags, offload, txq->vlan_flag);
tx_pkts += (n - 1);
txdp += (n - 1);
- iavf_vtx1(txdp, *tx_pkts++, rs, offload);
+ iavf_vtx1(txdp, *tx_pkts++, rs, offload, txq->vlan_flag);
nb_commit = (uint16_t)(nb_commit - n);
@@ -2238,7 +2270,7 @@ iavf_xmit_fixed_burst_vec_avx512(void *tx_queue, struct rte_mbuf **tx_pkts,
tx_backlog_entry_avx512(txep, tx_pkts, nb_commit);
- iavf_vtx(txdp, tx_pkts, nb_commit, flags, offload);
+ iavf_vtx(txdp, tx_pkts, nb_commit, flags, offload, txq->vlan_flag);
tx_id = (uint16_t)(tx_id + nb_commit);
if (tx_id > txq->tx_next_rs) {
diff --git a/drivers/net/intel/iavf/iavf_rxtx_vec_common.h b/drivers/net/intel/iavf/iavf_rxtx_vec_common.h
index 38e9a206d9..68b694b39e 100644
--- a/drivers/net/intel/iavf/iavf_rxtx_vec_common.h
+++ b/drivers/net/intel/iavf/iavf_rxtx_vec_common.h
@@ -99,7 +99,8 @@ iavf_tx_vec_queue_default(struct ci_tx_queue *txq)
*/
if (txq->offloads & (IAVF_TX_VECTOR_OFFLOAD | IAVF_TX_VECTOR_OFFLOAD_CTX)) {
if (txq->offloads & IAVF_TX_VECTOR_OFFLOAD_CTX) {
- if (txq->vlan_flag == IAVF_TX_FLAGS_VLAN_TAG_LOC_L2TAG2) {
+ if (txq->vlan_flag == IAVF_TX_FLAGS_VLAN_TAG_LOC_L2TAG2 ||
+ txq->offloads & RTE_ETH_TX_OFFLOAD_QINQ_INSERT) {
txq->use_ctx = 1;
return IAVF_VECTOR_CTX_OFFLOAD_PATH;
} else {
@@ -167,7 +168,7 @@ iavf_tx_vec_dev_check_default(struct rte_eth_dev *dev)
static __rte_always_inline void
iavf_txd_enable_offload(__rte_unused struct rte_mbuf *tx_pkt,
- uint64_t *txd_hi)
+ uint64_t *txd_hi, uint8_t vlan_flag)
{
#if defined(IAVF_TX_CSUM_OFFLOAD) || defined(IAVF_TX_VLAN_QINQ_OFFLOAD)
uint64_t ol_flags = tx_pkt->ol_flags;
@@ -228,11 +229,21 @@ iavf_txd_enable_offload(__rte_unused struct rte_mbuf *tx_pkt,
#endif
#ifdef IAVF_TX_VLAN_QINQ_OFFLOAD
- if (ol_flags & (RTE_MBUF_F_TX_VLAN | RTE_MBUF_F_TX_QINQ)) {
+ if (ol_flags & RTE_MBUF_F_TX_VLAN && vlan_flag & IAVF_TX_FLAGS_VLAN_TAG_LOC_L2TAG1) {
td_cmd |= IAVF_TX_DESC_CMD_IL2TAG1;
*txd_hi |= ((uint64_t)tx_pkt->vlan_tci <<
IAVF_TXD_QW1_L2TAG1_SHIFT);
}
+
+ if (ol_flags & RTE_MBUF_F_TX_QINQ) {
+ td_cmd |= IAVF_TX_DESC_CMD_IL2TAG1;
+ if (vlan_flag & IAVF_TX_FLAGS_VLAN_TAG_LOC_L2TAG1)
+ *txd_hi |= ((uint64_t)tx_pkt->vlan_tci <<
+ IAVF_TXD_QW1_L2TAG1_SHIFT);
+ else
+ *txd_hi |= ((uint64_t)tx_pkt->vlan_tci_outer <<
+ IAVF_TXD_QW1_L2TAG1_SHIFT);
+ }
#endif
*txd_hi |= ((uint64_t)td_cmd) << IAVF_TXD_QW1_CMD_SHIFT;
--
2.34.1
^ permalink raw reply [flat|nested] 3+ messages in thread