[dpdk-dev,v8,5/5] ixgbe: support VxLAN & NVGRE TX checksum off-load
Commit Message
The patch add VxLAN & NVGRE TX checksum off-load. When the flag of
outer IP header checksum offload is set, we'll set the context
descriptor to enable this checksum off-load.
Also update release note for VxLAN & NVGRE checksum off-load support
and ABI change.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
---
doc/guides/rel_notes/release_16_04.rst | 14 +++++++++
drivers/net/ixgbe/ixgbe_ethdev.c | 4 +++
drivers/net/ixgbe/ixgbe_rxtx.c | 56 +++++++++++++++++++++++++++-------
drivers/net/ixgbe/ixgbe_rxtx.h | 6 +++-
4 files changed, 68 insertions(+), 12 deletions(-)
@@ -46,6 +46,15 @@ This section should contain new features added in this release. Sample format:
* **Added vhost-user live migration support.**
+* **Added support for VxLAN & NVGRE checksum off-load on X550.**
+
+ * Added support for VxLAN & NVGRE RX/TX checksum off-load on
+ X550. RX/TX checksum off-load is provided on both inner and
+ outer IP header and TCP header.
+ * Added functions to support VxLAN port configuration. The
+ default VxLAN port number is 4789 but this can be updated
+ programmatically.
+
Resolved Issues
---------------
@@ -113,6 +122,11 @@ ABI Changes
the previous releases and made in this release. Use fixed width quotes for
``rte_function_names`` or ``rte_struct_names``. Use the past tense.
+* New API ``rte_eth_dev_udp_tunnel_port_add`` has been introduced to replace
+ ``rte_eth_dev_udp_tunnel_add``.
+
+* New API ``rte_eth_dev_udp_tunnel_port_delete`` has been introduced to replace
+ ``rte_eth_dev_udp_tunnel_delete``.
Shared Library Versions
-----------------------
@@ -2811,6 +2811,10 @@ ixgbe_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
DEV_TX_OFFLOAD_SCTP_CKSUM |
DEV_TX_OFFLOAD_TCP_TSO;
+ if (hw->mac.type == ixgbe_mac_X550 ||
+ hw->mac.type == ixgbe_mac_X550EM_x)
+ dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM;
+
dev_info->default_rxconf = (struct rte_eth_rxconf) {
.rx_thresh = {
.pthresh = IXGBE_DEFAULT_RX_PTHRESH,
@@ -85,7 +85,8 @@
PKT_TX_VLAN_PKT | \
PKT_TX_IP_CKSUM | \
PKT_TX_L4_MASK | \
- PKT_TX_TCP_SEG)
+ PKT_TX_TCP_SEG | \
+ PKT_TX_OUTER_IP_CKSUM)
static inline struct rte_mbuf *
rte_rxmbuf_alloc(struct rte_mempool *mp)
@@ -364,9 +365,11 @@ ixgbe_set_xmit_ctx(struct ixgbe_tx_queue *txq,
uint32_t ctx_idx;
uint32_t vlan_macip_lens;
union ixgbe_tx_offload tx_offload_mask;
+ uint32_t seqnum_seed = 0;
ctx_idx = txq->ctx_curr;
- tx_offload_mask.data = 0;
+ tx_offload_mask.data[0] = 0;
+ tx_offload_mask.data[1] = 0;
type_tucmd_mlhl = 0;
/* Specify which HW CTX to upload. */
@@ -430,18 +433,35 @@ ixgbe_set_xmit_ctx(struct ixgbe_tx_queue *txq,
}
}
+ if (ol_flags & PKT_TX_OUTER_IP_CKSUM) {
+ tx_offload_mask.outer_l2_len |= ~0;
+ tx_offload_mask.outer_l3_len |= ~0;
+ tx_offload_mask.l2_len |= ~0;
+ seqnum_seed |= tx_offload.outer_l3_len
+ << IXGBE_ADVTXD_OUTER_IPLEN;
+ seqnum_seed |= tx_offload.l2_len
+ << IXGBE_ADVTXD_TUNNEL_LEN;
+ }
+
txq->ctx_cache[ctx_idx].flags = ol_flags;
- txq->ctx_cache[ctx_idx].tx_offload.data =
- tx_offload_mask.data & tx_offload.data;
+ txq->ctx_cache[ctx_idx].tx_offload.data[0] =
+ tx_offload_mask.data[0] & tx_offload.data[0];
+ txq->ctx_cache[ctx_idx].tx_offload.data[1] =
+ tx_offload_mask.data[1] & tx_offload.data[1];
txq->ctx_cache[ctx_idx].tx_offload_mask = tx_offload_mask;
ctx_txd->type_tucmd_mlhl = rte_cpu_to_le_32(type_tucmd_mlhl);
vlan_macip_lens = tx_offload.l3_len;
- vlan_macip_lens |= (tx_offload.l2_len << IXGBE_ADVTXD_MACLEN_SHIFT);
+ if (ol_flags & PKT_TX_OUTER_IP_CKSUM)
+ vlan_macip_lens |= (tx_offload.outer_l2_len <<
+ IXGBE_ADVTXD_MACLEN_SHIFT);
+ else
+ vlan_macip_lens |= (tx_offload.l2_len <<
+ IXGBE_ADVTXD_MACLEN_SHIFT);
vlan_macip_lens |= ((uint32_t)tx_offload.vlan_tci << IXGBE_ADVTXD_VLAN_SHIFT);
ctx_txd->vlan_macip_lens = rte_cpu_to_le_32(vlan_macip_lens);
ctx_txd->mss_l4len_idx = rte_cpu_to_le_32(mss_l4len_idx);
- ctx_txd->seqnum_seed = 0;
+ ctx_txd->seqnum_seed = seqnum_seed;
}
/*
@@ -454,16 +474,24 @@ what_advctx_update(struct ixgbe_tx_queue *txq, uint64_t flags,
{
/* If match with the current used context */
if (likely((txq->ctx_cache[txq->ctx_curr].flags == flags) &&
- (txq->ctx_cache[txq->ctx_curr].tx_offload.data ==
- (txq->ctx_cache[txq->ctx_curr].tx_offload_mask.data & tx_offload.data)))) {
+ (txq->ctx_cache[txq->ctx_curr].tx_offload.data[0] ==
+ (txq->ctx_cache[txq->ctx_curr].tx_offload_mask.data[0]
+ & tx_offload.data[0])) &&
+ (txq->ctx_cache[txq->ctx_curr].tx_offload.data[1] ==
+ (txq->ctx_cache[txq->ctx_curr].tx_offload_mask.data[1]
+ & tx_offload.data[1])))) {
return txq->ctx_curr;
}
/* What if match with the next context */
txq->ctx_curr ^= 1;
if (likely((txq->ctx_cache[txq->ctx_curr].flags == flags) &&
- (txq->ctx_cache[txq->ctx_curr].tx_offload.data ==
- (txq->ctx_cache[txq->ctx_curr].tx_offload_mask.data & tx_offload.data)))) {
+ (txq->ctx_cache[txq->ctx_curr].tx_offload.data[0] ==
+ (txq->ctx_cache[txq->ctx_curr].tx_offload_mask.data[0]
+ & tx_offload.data[0])) &&
+ (txq->ctx_cache[txq->ctx_curr].tx_offload.data[1] ==
+ (txq->ctx_cache[txq->ctx_curr].tx_offload_mask.data[1]
+ & tx_offload.data[1])))) {
return txq->ctx_curr;
}
@@ -492,6 +520,8 @@ tx_desc_ol_flags_to_cmdtype(uint64_t ol_flags)
cmdtype |= IXGBE_ADVTXD_DCMD_VLE;
if (ol_flags & PKT_TX_TCP_SEG)
cmdtype |= IXGBE_ADVTXD_DCMD_TSE;
+ if (ol_flags & PKT_TX_OUTER_IP_CKSUM)
+ cmdtype |= (1 << IXGBE_ADVTXD_OUTERIPCS_SHIFT);
return cmdtype;
}
@@ -588,8 +618,10 @@ ixgbe_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
uint64_t tx_ol_req;
uint32_t ctx = 0;
uint32_t new_ctx;
- union ixgbe_tx_offload tx_offload = {0};
+ union ixgbe_tx_offload tx_offload;
+ tx_offload.data[0] = 0;
+ tx_offload.data[1] = 0;
txq = tx_queue;
sw_ring = txq->sw_ring;
txr = txq->tx_ring;
@@ -623,6 +655,8 @@ ixgbe_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
tx_offload.l4_len = tx_pkt->l4_len;
tx_offload.vlan_tci = tx_pkt->vlan_tci;
tx_offload.tso_segsz = tx_pkt->tso_segsz;
+ tx_offload.outer_l2_len = tx_pkt->outer_l2_len;
+ tx_offload.outer_l3_len = tx_pkt->outer_l3_len;
/* If new context need be built or reuse the exist ctx. */
ctx = what_advctx_update(txq, tx_ol_req,
@@ -163,7 +163,7 @@ enum ixgbe_advctx_num {
/** Offload features */
union ixgbe_tx_offload {
- uint64_t data;
+ uint64_t data[2];
struct {
uint64_t l2_len:7; /**< L2 (MAC) Header Length. */
uint64_t l3_len:9; /**< L3 (IP) Header Length. */
@@ -171,6 +171,10 @@ union ixgbe_tx_offload {
uint64_t tso_segsz:16; /**< TCP TSO segment size */
uint64_t vlan_tci:16;
/**< VLAN Tag Control Identifier (CPU order). */
+
+ /* fields for TX offloading of tunnels */
+ uint64_t outer_l3_len:8; /**< Outer L3 (IP) Hdr Length. */
+ uint64_t outer_l2_len:8; /**< Outer L2 (MAC) Hdr Length. */
};
};