[dpdk-dev] [PATCH v2 0/9] bnxt patchset

[dpdk-dev] [PATCH v2 0/9] bnxt patchset

[Ajit Khaparde]

Apart form fixes support for Thor and vector PMD, this patchset
adds support for LRO on Thor based adapters and CoS classification.
Patchset against dpdk-next-net, has been compiled & tested on an
x86_64 system.

Please apply.


Lance Richardson (8):
  net/bnxt: increase tqm entry allocation
  net/bnxt: fix ring alignment for thor-based adapters
  net/bnxt: add support for LRO on thor adapters
  net/bnxt: use common receive transmit nq ring
  net/bnxt: fix stats context calculation
  net/bnxt: use correct default Rx queue for thor
  net/bnxt: advertise scatter receive offload capability
  net/bnxt: improve CPR handling in vector PMD

Venkat Duvvuru (1):
  net/bnxt: add support for CoS classification

 drivers/net/bnxt/bnxt.h                |  27 ++++-
 drivers/net/bnxt/bnxt_ethdev.c         |  45 +++++++-
 drivers/net/bnxt/bnxt_hwrm.c           | 131 ++++++++++++++++-------
 drivers/net/bnxt/bnxt_hwrm.h           |  14 ++-
 drivers/net/bnxt/bnxt_ring.c           | 142 +++++++++++++++++--------
 drivers/net/bnxt/bnxt_ring.h           |   3 +-
 drivers/net/bnxt/bnxt_rxq.c            |  11 +-
 drivers/net/bnxt/bnxt_rxq.h            |   1 -
 drivers/net/bnxt/bnxt_rxr.c            |  99 +++++++++--------
 drivers/net/bnxt/bnxt_rxr.h            |  41 +++++--
 drivers/net/bnxt/bnxt_rxtx_vec_sse.c   |  26 +----
 drivers/net/bnxt/bnxt_txq.c            |   4 +-
 drivers/net/bnxt/bnxt_txq.h            |   1 -
 drivers/net/bnxt/bnxt_txr.c            |  25 -----
 drivers/net/bnxt/bnxt_vnic.h           |   1 +
 drivers/net/bnxt/hsi_struct_def_dpdk.h |  28 ++++-
 16 files changed, 398 insertions(+), 201 deletions(-)

-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 1/9] net/bnxt: increase tqm entry allocation

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

The current TQM backing store size isn't sufficient to allow 512
transmit rings. Fix by correcting TQM SP queue size calculation.

Fixes: f8168ca0e690 ("net/bnxt: support thor controller")
Cc: stable@dpdk.org
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Kalesh Anakkur Purayil <kalesh-anakkur.purayil@broadcom.com>
---
 drivers/net/bnxt/bnxt_ethdev.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 02eacf7965..0e893cc956 100644
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -4292,7 +4292,9 @@ int bnxt_alloc_ctx_mem(struct bnxt *bp)
 	if (rc)
 		return rc;
 
-	entries = ctx->qp_max_l2_entries;
+	entries = ctx->qp_max_l2_entries +
+		  ctx->vnic_max_vnic_entries +
+		  ctx->tqm_min_entries_per_ring;
 	entries = bnxt_roundup(entries, ctx->tqm_entries_multiple);
 	entries = clamp_t(uint32_t, entries, ctx->tqm_min_entries_per_ring,
 			  ctx->tqm_max_entries_per_ring);
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 2/9] net/bnxt: fix ring alignment for thor-based adapters

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

When using transmit/receive queue sizes smaller than 256, alignment
requirements are not being met for Thor-based adapters. Fix by
forcing memory addresses used for transmit/receive/aggregation ring
allocations to be on 4K boundaries.

Fixes: f8168ca0e690 ("net/bnxt: support thor controller")
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Kumar Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_ring.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/bnxt/bnxt_ring.c b/drivers/net/bnxt/bnxt_ring.c
index a3e5a68714..4662e94132 100644
--- a/drivers/net/bnxt/bnxt_ring.c
+++ b/drivers/net/bnxt/bnxt_ring.c
@@ -162,18 +162,21 @@ int bnxt_alloc_rings(struct bnxt *bp, uint16_t qidx,
 	int nq_ring_len = BNXT_CHIP_THOR(bp) ? cp_ring_len : 0;
 
 	int tx_ring_start = nq_ring_start + nq_ring_len;
+	tx_ring_start = RTE_ALIGN(tx_ring_start, 4096);
 	int tx_ring_len = tx_ring_info ?
 	    RTE_CACHE_LINE_ROUNDUP(tx_ring_info->tx_ring_struct->ring_size *
 				   sizeof(struct tx_bd_long)) : 0;
 	tx_ring_len = RTE_ALIGN(tx_ring_len, 4096);
 
 	int rx_ring_start = tx_ring_start + tx_ring_len;
+	rx_ring_start = RTE_ALIGN(rx_ring_start, 4096);
 	int rx_ring_len =  rx_ring_info ?
 		RTE_CACHE_LINE_ROUNDUP(rx_ring_info->rx_ring_struct->ring_size *
 		sizeof(struct rx_prod_pkt_bd)) : 0;
 	rx_ring_len = RTE_ALIGN(rx_ring_len, 4096);
 
 	int ag_ring_start = rx_ring_start + rx_ring_len;
+	ag_ring_start = RTE_ALIGN(ag_ring_start, 4096);
 	int ag_ring_len = rx_ring_len * AGG_RING_SIZE_FACTOR;
 	ag_ring_len = RTE_ALIGN(ag_ring_len, 4096);
 
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 3/9] net/bnxt: add support for LRO on thor adapters

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

Add support for LRO for adapters based on Thor (BCM57508).

Reviewed-by: Kalesh Anakkur Purayil <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt.h        | 16 ++++++++
 drivers/net/bnxt/bnxt_ethdev.c |  4 ++
 drivers/net/bnxt/bnxt_hwrm.c   | 33 +++++++++++++---
 drivers/net/bnxt/bnxt_hwrm.h   |  1 +
 drivers/net/bnxt/bnxt_ring.c   | 14 ++++---
 drivers/net/bnxt/bnxt_ring.h   |  1 -
 drivers/net/bnxt/bnxt_rxq.c    |  4 +-
 drivers/net/bnxt/bnxt_rxr.c    | 72 +++++++++++++++++++++++++---------
 drivers/net/bnxt/bnxt_rxr.h    | 41 +++++++++++++++----
 9 files changed, 149 insertions(+), 37 deletions(-)

diff --git a/drivers/net/bnxt/bnxt.h b/drivers/net/bnxt/bnxt.h
index c34582c0ad..ad97e0e593 100644
--- a/drivers/net/bnxt/bnxt.h
+++ b/drivers/net/bnxt/bnxt.h
@@ -37,6 +37,21 @@
 #define BNXT_MAX_RX_RING_DESC	8192
 #define BNXT_DB_SIZE		0x80
 
+#define TPA_MAX_AGGS		64
+#define TPA_MAX_AGGS_TH		1024
+
+#define TPA_MAX_NUM_SEGS	32
+#define TPA_MAX_SEGS_TH		8 /* 32 segments in 4-segment units */
+#define TPA_MAX_SEGS		5 /* 32 segments in log2 units */
+
+#define BNXT_TPA_MAX_AGGS(bp) \
+	(BNXT_CHIP_THOR(bp) ? TPA_MAX_AGGS_TH : \
+			     TPA_MAX_AGGS)
+
+#define BNXT_TPA_MAX_SEGS(bp) \
+	(BNXT_CHIP_THOR(bp) ? TPA_MAX_SEGS_TH : \
+			      TPA_MAX_SEGS)
+
 #ifdef RTE_ARCH_ARM64
 #define BNXT_NUM_ASYNC_CPR(bp) (BNXT_STINGRAY(bp) ? 0 : 1)
 #else
@@ -525,6 +540,7 @@ struct bnxt {
 	uint16_t		max_rx_em_flows;
 	uint16_t		max_vnics;
 	uint16_t		max_stat_ctx;
+	uint16_t		max_tpa_v2;
 	uint16_t		first_vf_id;
 	uint16_t		vlan;
 #define BNXT_OUTER_TPID_MASK	0x0000ffff
diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 0e893cc956..4fc182b8cc 100644
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -4553,6 +4553,10 @@ static int bnxt_init_fw(struct bnxt *bp)
 	if (rc)
 		return rc;
 
+	rc = bnxt_hwrm_vnic_qcaps(bp);
+	if (rc)
+		return rc;
+
 	rc = bnxt_hwrm_func_qcfg(bp, &mtu);
 	if (rc)
 		return rc;
diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index 011cd05ae0..9f30214c99 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -700,6 +700,27 @@ int bnxt_hwrm_func_qcaps(struct bnxt *bp)
 	return rc;
 }
 
+int bnxt_hwrm_vnic_qcaps(struct bnxt *bp)
+{
+	int rc = 0;
+	struct hwrm_vnic_qcaps_input req = {.req_type = 0 };
+	struct hwrm_vnic_qcaps_output *resp = bp->hwrm_cmd_resp_addr;
+
+	HWRM_PREP(req, VNIC_QCAPS, BNXT_USE_CHIMP_MB);
+
+	req.target_id = rte_cpu_to_le_16(0xffff);
+
+	rc = bnxt_hwrm_send_message(bp, &req, sizeof(req), BNXT_USE_CHIMP_MB);
+
+	HWRM_CHECK_RESULT();
+
+	bp->max_tpa_v2 = rte_le_to_cpu_16(resp->max_aggs_supported);
+
+	HWRM_UNLOCK();
+
+	return rc;
+}
+
 int bnxt_hwrm_func_reset(struct bnxt *bp)
 {
 	int rc = 0;
@@ -1959,8 +1980,11 @@ int bnxt_hwrm_vnic_tpa_cfg(struct bnxt *bp,
 	struct hwrm_vnic_tpa_cfg_input req = {.req_type = 0 };
 	struct hwrm_vnic_tpa_cfg_output *resp = bp->hwrm_cmd_resp_addr;
 
-	if (BNXT_CHIP_THOR(bp))
-		return 0;
+	if (BNXT_CHIP_THOR(bp) && !bp->max_tpa_v2) {
+		if (enable)
+			PMD_DRV_LOG(ERR, "No HW support for LRO\n");
+		return -ENOTSUP;
+	}
 
 	if (vnic->fw_vnic_id == INVALID_HW_RING_ID) {
 		PMD_DRV_LOG(DEBUG, "Invalid vNIC ID\n");
@@ -1981,9 +2005,8 @@ int bnxt_hwrm_vnic_tpa_cfg(struct bnxt *bp,
 				HWRM_VNIC_TPA_CFG_INPUT_FLAGS_GRO |
 				HWRM_VNIC_TPA_CFG_INPUT_FLAGS_AGG_WITH_ECN |
 			HWRM_VNIC_TPA_CFG_INPUT_FLAGS_AGG_WITH_SAME_GRE_SEQ);
-		req.max_agg_segs = rte_cpu_to_le_16(5);
-		req.max_aggs =
-			rte_cpu_to_le_16(HWRM_VNIC_TPA_CFG_INPUT_MAX_AGGS_MAX);
+		req.max_agg_segs = rte_cpu_to_le_16(BNXT_TPA_MAX_AGGS(bp));
+		req.max_aggs = rte_cpu_to_le_16(BNXT_TPA_MAX_SEGS(bp));
 		req.min_agg_len = rte_cpu_to_le_32(512);
 	}
 	req.vnic_id = rte_cpu_to_le_16(vnic->fw_vnic_id);
diff --git a/drivers/net/bnxt/bnxt_hwrm.h b/drivers/net/bnxt/bnxt_hwrm.h
index 07181d4020..8912a4ed3e 100644
--- a/drivers/net/bnxt/bnxt_hwrm.h
+++ b/drivers/net/bnxt/bnxt_hwrm.h
@@ -110,6 +110,7 @@ int bnxt_hwrm_vnic_alloc(struct bnxt *bp, struct bnxt_vnic_info *vnic);
 int bnxt_hwrm_vnic_cfg(struct bnxt *bp, struct bnxt_vnic_info *vnic);
 int bnxt_hwrm_vnic_qcfg(struct bnxt *bp, struct bnxt_vnic_info *vnic,
 				int16_t fw_vf_id);
+int bnxt_hwrm_vnic_qcaps(struct bnxt *bp);
 int bnxt_hwrm_vnic_ctx_alloc(struct bnxt *bp, struct bnxt_vnic_info *vnic,
 			     uint16_t ctx_idx);
 int bnxt_hwrm_vnic_ctx_free(struct bnxt *bp, struct bnxt_vnic_info *vnic);
diff --git a/drivers/net/bnxt/bnxt_ring.c b/drivers/net/bnxt/bnxt_ring.c
index 4662e94132..14cfb8c155 100644
--- a/drivers/net/bnxt/bnxt_ring.c
+++ b/drivers/net/bnxt/bnxt_ring.c
@@ -187,13 +187,17 @@ int bnxt_alloc_rings(struct bnxt *bp, uint16_t qidx,
 			AGG_RING_SIZE_FACTOR)) : 0;
 
 	int tpa_info_start = ag_bitmap_start + ag_bitmap_len;
-	int tpa_info_len = rx_ring_info ?
-		RTE_CACHE_LINE_ROUNDUP(BNXT_TPA_MAX *
-				       sizeof(struct bnxt_tpa_info)) : 0;
+	int tpa_info_len = 0;
+
+	if (rx_ring_info && (rx_offloads & DEV_RX_OFFLOAD_TCP_LRO)) {
+		int tpa_max = BNXT_TPA_MAX_AGGS(bp);
+
+		tpa_info_len = tpa_max * sizeof(struct bnxt_tpa_info);
+		tpa_info_len = RTE_CACHE_LINE_ROUNDUP(tpa_info_len);
+	}
 
 	int total_alloc_len = tpa_info_start;
-	if (rx_offloads & DEV_RX_OFFLOAD_TCP_LRO)
-		total_alloc_len += tpa_info_len;
+	total_alloc_len += tpa_info_len;
 
 	snprintf(mz_name, RTE_MEMZONE_NAMESIZE,
 		 "bnxt_%04x:%02x:%02x:%02x-%04x_%s", pdev->addr.domain,
diff --git a/drivers/net/bnxt/bnxt_ring.h b/drivers/net/bnxt/bnxt_ring.h
index a31d59ea39..a5d5106986 100644
--- a/drivers/net/bnxt/bnxt_ring.h
+++ b/drivers/net/bnxt/bnxt_ring.h
@@ -27,7 +27,6 @@
 #define DEFAULT_RX_RING_SIZE	256
 #define DEFAULT_TX_RING_SIZE	256
 
-#define BNXT_TPA_MAX		64
 #define AGG_RING_SIZE_FACTOR	2
 #define AGG_RING_MULTIPLIER	2
 
diff --git a/drivers/net/bnxt/bnxt_rxq.c b/drivers/net/bnxt/bnxt_rxq.c
index 371534db6b..03b115dbaf 100644
--- a/drivers/net/bnxt/bnxt_rxq.c
+++ b/drivers/net/bnxt/bnxt_rxq.c
@@ -227,7 +227,9 @@ void bnxt_rx_queue_release_mbufs(struct bnxt_rx_queue *rxq)
 	/* Free up mbufs in TPA */
 	tpa_info = rxq->rx_ring->tpa_info;
 	if (tpa_info) {
-		for (i = 0; i < BNXT_TPA_MAX; i++) {
+		int max_aggs = BNXT_TPA_MAX_AGGS(rxq->bp);
+
+		for (i = 0; i < max_aggs; i++) {
 			if (tpa_info[i].mbuf) {
 				rte_pktmbuf_free_seg(tpa_info[i].mbuf);
 				tpa_info[i].mbuf = NULL;
diff --git a/drivers/net/bnxt/bnxt_rxr.c b/drivers/net/bnxt/bnxt_rxr.c
index b3cc0d8a04..1a6fb7944b 100644
--- a/drivers/net/bnxt/bnxt_rxr.c
+++ b/drivers/net/bnxt/bnxt_rxr.c
@@ -124,12 +124,13 @@ static void bnxt_tpa_start(struct bnxt_rx_queue *rxq,
 			   struct rx_tpa_start_cmpl_hi *tpa_start1)
 {
 	struct bnxt_rx_ring_info *rxr = rxq->rx_ring;
-	uint8_t agg_id = rte_le_to_cpu_32(tpa_start->agg_id &
-		RX_TPA_START_CMPL_AGG_ID_MASK) >> RX_TPA_START_CMPL_AGG_ID_SFT;
+	uint16_t agg_id;
 	uint16_t data_cons;
 	struct bnxt_tpa_info *tpa_info;
 	struct rte_mbuf *mbuf;
 
+	agg_id = bnxt_tpa_start_agg_id(rxq->bp, tpa_start);
+
 	data_cons = tpa_start->opaque;
 	tpa_info = &rxr->tpa_info[agg_id];
 
@@ -137,6 +138,7 @@ static void bnxt_tpa_start(struct bnxt_rx_queue *rxq,
 
 	bnxt_reuse_rx_mbuf(rxr, tpa_info->mbuf);
 
+	tpa_info->agg_count = 0;
 	tpa_info->mbuf = mbuf;
 	tpa_info->len = rte_le_to_cpu_32(tpa_start->len);
 
@@ -206,7 +208,7 @@ static int bnxt_prod_ag_mbuf(struct bnxt_rx_queue *rxq)
 
 static int bnxt_rx_pages(struct bnxt_rx_queue *rxq,
 			 struct rte_mbuf *mbuf, uint32_t *tmp_raw_cons,
-			 uint8_t agg_buf)
+			 uint8_t agg_buf, struct bnxt_tpa_info *tpa_info)
 {
 	struct bnxt_cp_ring_info *cpr = rxq->cp_ring;
 	struct bnxt_rx_ring_info *rxr = rxq->rx_ring;
@@ -214,14 +216,20 @@ static int bnxt_rx_pages(struct bnxt_rx_queue *rxq,
 	uint16_t cp_cons, ag_cons;
 	struct rx_pkt_cmpl *rxcmp;
 	struct rte_mbuf *last = mbuf;
+	bool is_thor_tpa = tpa_info && BNXT_CHIP_THOR(rxq->bp);
 
 	for (i = 0; i < agg_buf; i++) {
 		struct bnxt_sw_rx_bd *ag_buf;
 		struct rte_mbuf *ag_mbuf;
-		*tmp_raw_cons = NEXT_RAW_CMP(*tmp_raw_cons);
-		cp_cons = RING_CMP(cpr->cp_ring_struct, *tmp_raw_cons);
-		rxcmp = (struct rx_pkt_cmpl *)
+
+		if (is_thor_tpa) {
+			rxcmp = (void *)&tpa_info->agg_arr[i];
+		} else {
+			*tmp_raw_cons = NEXT_RAW_CMP(*tmp_raw_cons);
+			cp_cons = RING_CMP(cpr->cp_ring_struct, *tmp_raw_cons);
+			rxcmp = (struct rx_pkt_cmpl *)
 					&cpr->cp_desc_ring[cp_cons];
+		}
 
 #ifdef BNXT_DEBUG
 		bnxt_dump_cmpl(cp_cons, rxcmp);
@@ -258,29 +266,42 @@ static inline struct rte_mbuf *bnxt_tpa_end(
 		struct bnxt_rx_queue *rxq,
 		uint32_t *raw_cp_cons,
 		struct rx_tpa_end_cmpl *tpa_end,
-		struct rx_tpa_end_cmpl_hi *tpa_end1 __rte_unused)
+		struct rx_tpa_end_cmpl_hi *tpa_end1)
 {
 	struct bnxt_cp_ring_info *cpr = rxq->cp_ring;
 	struct bnxt_rx_ring_info *rxr = rxq->rx_ring;
-	uint8_t agg_id = (tpa_end->agg_id & RX_TPA_END_CMPL_AGG_ID_MASK)
-			>> RX_TPA_END_CMPL_AGG_ID_SFT;
+	uint16_t agg_id;
 	struct rte_mbuf *mbuf;
 	uint8_t agg_bufs;
+	uint8_t payload_offset;
 	struct bnxt_tpa_info *tpa_info;
 
+	if (BNXT_CHIP_THOR(rxq->bp)) {
+		struct rx_tpa_v2_end_cmpl *th_tpa_end;
+		struct rx_tpa_v2_end_cmpl_hi *th_tpa_end1;
+
+		th_tpa_end = (void *)tpa_end;
+		th_tpa_end1 = (void *)tpa_end1;
+		agg_id = BNXT_TPA_END_AGG_ID_TH(th_tpa_end);
+		agg_bufs = BNXT_TPA_END_AGG_BUFS_TH(th_tpa_end1);
+		payload_offset = th_tpa_end1->payload_offset;
+	} else {
+		agg_id = BNXT_TPA_END_AGG_ID(tpa_end);
+		agg_bufs = BNXT_TPA_END_AGG_BUFS(tpa_end);
+		if (!bnxt_agg_bufs_valid(cpr, agg_bufs, *raw_cp_cons))
+			return NULL;
+		payload_offset = tpa_end->payload_offset;
+	}
+
 	tpa_info = &rxr->tpa_info[agg_id];
 	mbuf = tpa_info->mbuf;
 	RTE_ASSERT(mbuf != NULL);
 
 	rte_prefetch0(mbuf);
-	agg_bufs = (rte_le_to_cpu_32(tpa_end->agg_bufs_v1) &
-		RX_TPA_END_CMPL_AGG_BUFS_MASK) >> RX_TPA_END_CMPL_AGG_BUFS_SFT;
 	if (agg_bufs) {
-		if (!bnxt_agg_bufs_valid(cpr, agg_bufs, *raw_cp_cons))
-			return NULL;
-		bnxt_rx_pages(rxq, mbuf, raw_cp_cons, agg_bufs);
+		bnxt_rx_pages(rxq, mbuf, raw_cp_cons, agg_bufs, tpa_info);
 	}
-	mbuf->l4_len = tpa_end->payload_offset;
+	mbuf->l4_len = payload_offset;
 
 	struct rte_mbuf *new_data = __bnxt_alloc_rx_data(rxq->mb_pool);
 	RTE_ASSERT(new_data != NULL);
@@ -395,6 +416,20 @@ static int bnxt_rx_pkt(struct rte_mbuf **rx_pkt,
 	rxcmp = (struct rx_pkt_cmpl *)
 	    &cpr->cp_desc_ring[cp_cons];
 
+	cmp_type = CMP_TYPE(rxcmp);
+
+	if (cmp_type == RX_TPA_V2_ABUF_CMPL_TYPE_RX_TPA_AGG) {
+		struct rx_tpa_v2_abuf_cmpl *rx_agg = (void *)rxcmp;
+		uint16_t agg_id = rte_cpu_to_le_16(rx_agg->agg_id);
+		struct bnxt_tpa_info *tpa_info;
+
+		tpa_info = &rxr->tpa_info[agg_id];
+		RTE_ASSERT(tpa_info->agg_count < 16);
+		tpa_info->agg_arr[tpa_info->agg_count++] = *rx_agg;
+		rc = -EINVAL; /* Continue w/o new mbuf */
+		goto next_rx;
+	}
+
 	tmp_raw_cons = NEXT_RAW_CMP(tmp_raw_cons);
 	cp_cons = RING_CMP(cpr->cp_ring_struct, tmp_raw_cons);
 	rxcmp1 = (struct rx_pkt_cmpl_hi *)&cpr->cp_desc_ring[cp_cons];
@@ -406,7 +441,6 @@ static int bnxt_rx_pkt(struct rte_mbuf **rx_pkt,
 				cpr->cp_ring_struct->ring_mask,
 				cpr->valid);
 
-	cmp_type = CMP_TYPE(rxcmp);
 	if (cmp_type == RX_TPA_START_CMPL_TYPE_RX_TPA_START) {
 		bnxt_tpa_start(rxq, (struct rx_tpa_start_cmpl *)rxcmp,
 			       (struct rx_tpa_start_cmpl_hi *)rxcmp1);
@@ -463,7 +497,7 @@ static int bnxt_rx_pkt(struct rte_mbuf **rx_pkt,
 	}
 #endif
 	if (agg_buf)
-		bnxt_rx_pages(rxq, mbuf, &tmp_raw_cons, agg_buf);
+		bnxt_rx_pages(rxq, mbuf, &tmp_raw_cons, agg_buf, NULL);
 
 	if (rxcmp1->flags2 & RX_PKT_CMPL_FLAGS2_META_FORMAT_VLAN) {
 		mbuf->vlan_tci = rxcmp1->metadata &
@@ -861,7 +895,9 @@ int bnxt_init_one_rx_ring(struct bnxt_rx_queue *rxq)
 	PMD_DRV_LOG(DEBUG, "AGG Done!\n");
 
 	if (rxr->tpa_info) {
-		for (i = 0; i < BNXT_TPA_MAX; i++) {
+		unsigned int max_aggs = BNXT_TPA_MAX_AGGS(rxq->bp);
+
+		for (i = 0; i < max_aggs; i++) {
 			rxr->tpa_info[i].mbuf =
 				__bnxt_alloc_rx_data(rxq->mb_pool);
 			if (!rxr->tpa_info[i].mbuf) {
diff --git a/drivers/net/bnxt/bnxt_rxr.h b/drivers/net/bnxt/bnxt_rxr.h
index 493b754066..76bf88d707 100644
--- a/drivers/net/bnxt/bnxt_rxr.h
+++ b/drivers/net/bnxt/bnxt_rxr.h
@@ -5,6 +5,7 @@
 
 #ifndef _BNXT_RXR_H_
 #define _BNXT_RXR_H_
+#include "hsi_struct_def_dpdk.h"
 
 #define B_RX_DB(db, prod)						\
 		(*(uint32_t *)db = (DB_KEY_RX | (prod)))
@@ -110,6 +111,36 @@
 		IS_L4_TUNNEL_PKT_ONLY_INNER_L4_CS(flags2_f)	\
 	)
 
+#define BNXT_TPA_START_AGG_ID_PRE_TH(cmp) \
+	((rte_le_to_cpu_16((cmp)->agg_id) & RX_TPA_START_CMPL_AGG_ID_MASK) >> \
+	 RX_TPA_START_CMPL_AGG_ID_SFT)
+
+#define BNXT_TPA_START_AGG_ID_TH(cmp) \
+	rte_le_to_cpu_16((cmp)->agg_id)
+
+static inline uint16_t bnxt_tpa_start_agg_id(struct bnxt *bp,
+					     struct rx_tpa_start_cmpl *cmp)
+{
+	if (BNXT_CHIP_THOR(bp))
+		return BNXT_TPA_START_AGG_ID_TH(cmp);
+	else
+		return BNXT_TPA_START_AGG_ID_PRE_TH(cmp);
+}
+
+#define BNXT_TPA_END_AGG_BUFS(cmp) \
+	(((cmp)->agg_bufs_v1 & RX_TPA_END_CMPL_AGG_BUFS_MASK) \
+	 >> RX_TPA_END_CMPL_AGG_BUFS_SFT)
+
+#define BNXT_TPA_END_AGG_BUFS_TH(cmp) \
+	((cmp)->tpa_agg_bufs)
+
+#define BNXT_TPA_END_AGG_ID(cmp) \
+	(((cmp)->agg_id & RX_TPA_END_CMPL_AGG_ID_MASK) >> \
+	 RX_TPA_END_CMPL_AGG_ID_SFT)
+
+#define BNXT_TPA_END_AGG_ID_TH(cmp) \
+	rte_le_to_cpu_16((cmp)->agg_id)
+
 #define RX_CMP_L4_CS_BITS	\
 	rte_cpu_to_le_32(RX_PKT_CMPL_FLAGS2_L4_CS_CALC)
 
@@ -144,14 +175,10 @@ enum pkt_hash_types {
 };
 
 struct bnxt_tpa_info {
-	struct rte_mbuf		*mbuf;
+	struct rte_mbuf			*mbuf;
 	uint16_t			len;
-	unsigned short		gso_type;
-	uint32_t			flags2;
-	uint32_t			metadata;
-	enum pkt_hash_types	hash_type;
-	uint32_t			rss_hash;
-	uint32_t			hdr_info;
+	uint32_t			agg_count;
+	struct rx_tpa_v2_abuf_cmpl	agg_arr[TPA_MAX_NUM_SEGS];
 };
 
 struct bnxt_sw_rx_bd {
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 4/9] net/bnxt: add support for CoS classification

[Ajit Khaparde]

From: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>

Class of Service (CoS) is a way to manage multiple types of
traffic over a network to offer different types of services
to applications. CoS classification (priority to cosqueue) is
determined by the user and configured through the PF driver.
DPDK driver queries this configuration and maps the cos queue
ids to different VNICs. This patch adds this support.

Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
Reviewed-by: Kalesh Anakkur Purayil <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt.h                | 10 ++-
 drivers/net/bnxt/bnxt_ethdev.c         | 30 +++++++--
 drivers/net/bnxt/bnxt_hwrm.c           | 84 +++++++++++++++++++-------
 drivers/net/bnxt/bnxt_hwrm.h           | 13 +++-
 drivers/net/bnxt/bnxt_ring.c           | 18 ++++--
 drivers/net/bnxt/bnxt_rxq.c            |  3 +-
 drivers/net/bnxt/bnxt_vnic.h           |  1 +
 drivers/net/bnxt/hsi_struct_def_dpdk.h | 28 ++++++++-
 8 files changed, 149 insertions(+), 38 deletions(-)

diff --git a/drivers/net/bnxt/bnxt.h b/drivers/net/bnxt/bnxt.h
index ad97e0e593..5cfe5ee2c7 100644
--- a/drivers/net/bnxt/bnxt.h
+++ b/drivers/net/bnxt/bnxt.h
@@ -470,8 +470,10 @@ struct bnxt {
 
 	uint32_t		flow_flags;
 #define BNXT_FLOW_FLAG_L2_HDR_SRC_FILTER_EN	BIT(0)
-
 	pthread_mutex_t         flow_lock;
+
+	uint32_t		vnic_cap_flags;
+#define BNXT_VNIC_CAP_COS_CLASSIFY	BIT(0)
 	unsigned int		rx_nr_rings;
 	unsigned int		rx_cp_nr_rings;
 	unsigned int		rx_num_qs_per_vnic;
@@ -523,8 +525,10 @@ struct bnxt {
 	uint16_t                        hwrm_max_ext_req_len;
 
 	struct bnxt_link_info	link_info;
-	struct bnxt_cos_queue_info	cos_queue[BNXT_COS_QUEUE_COUNT];
-	uint8_t			tx_cosq_id;
+	struct bnxt_cos_queue_info	rx_cos_queue[BNXT_COS_QUEUE_COUNT];
+	struct bnxt_cos_queue_info	tx_cos_queue[BNXT_COS_QUEUE_COUNT];
+	uint8_t			tx_cosq_id[BNXT_COS_QUEUE_COUNT];
+	uint8_t			rx_cosq_cnt;
 	uint8_t                 max_tc;
 	uint8_t                 max_lltc;
 	uint8_t                 max_q;
diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 4fc182b8cc..9adcd94ff8 100644
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -308,6 +308,25 @@ static int bnxt_init_chip(struct bnxt *bp)
 		goto err_out;
 	}
 
+	if (!(bp->vnic_cap_flags & BNXT_VNIC_CAP_COS_CLASSIFY))
+		goto skip_cosq_cfg;
+
+	for (j = 0, i = 0; i < BNXT_COS_QUEUE_COUNT; i++) {
+		if (bp->rx_cos_queue[i].id != 0xff) {
+			struct bnxt_vnic_info *vnic = &bp->vnic_info[j++];
+
+			if (!vnic) {
+				PMD_DRV_LOG(ERR,
+					    "Num pools more than FW profile\n");
+				rc = -EINVAL;
+				goto err_out;
+			}
+			vnic->cos_queue_id = bp->rx_cos_queue[i].id;
+			bp->rx_cosq_cnt++;
+		}
+	}
+
+skip_cosq_cfg:
 	rc = bnxt_mq_rx_configure(bp);
 	if (rc) {
 		PMD_DRV_LOG(ERR, "MQ mode configure failure rc: %x\n", rc);
@@ -4540,7 +4559,7 @@ static int bnxt_init_fw(struct bnxt *bp)
 	if (rc)
 		return -EIO;
 
-	rc = bnxt_hwrm_cfa_adv_flow_mgmt_qcaps(bp);
+	rc = bnxt_hwrm_vnic_qcaps(bp);
 	if (rc)
 		return rc;
 
@@ -4548,16 +4567,19 @@ static int bnxt_init_fw(struct bnxt *bp)
 	if (rc)
 		return rc;
 
-	/* Get the MAX capabilities for this function */
+	/* Get the MAX capabilities for this function.
+	 * This function also allocates context memory for TQM rings and
+	 * informs the firmware about this allocated backing store memory.
+	 */
 	rc = bnxt_hwrm_func_qcaps(bp);
 	if (rc)
 		return rc;
 
-	rc = bnxt_hwrm_vnic_qcaps(bp);
+	rc = bnxt_hwrm_func_qcfg(bp, &mtu);
 	if (rc)
 		return rc;
 
-	rc = bnxt_hwrm_func_qcfg(bp, &mtu);
+	rc = bnxt_hwrm_cfa_adv_flow_mgmt_qcaps(bp);
 	if (rc)
 		return rc;
 
diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index 9f30214c99..76ef004237 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -700,6 +700,7 @@ int bnxt_hwrm_func_qcaps(struct bnxt *bp)
 	return rc;
 }
 
+/* VNIC cap covers capability of all VNICs. So no need to pass vnic_id */
 int bnxt_hwrm_vnic_qcaps(struct bnxt *bp)
 {
 	int rc = 0;
@@ -714,6 +715,12 @@ int bnxt_hwrm_vnic_qcaps(struct bnxt *bp)
 
 	HWRM_CHECK_RESULT();
 
+	if (rte_le_to_cpu_32(resp->flags) &
+	    HWRM_VNIC_QCAPS_OUTPUT_FLAGS_COS_ASSIGNMENT_CAP) {
+		bp->vnic_cap_flags |= BNXT_VNIC_CAP_COS_CLASSIFY;
+		PMD_DRV_LOG(INFO, "CoS assignment capability enabled\n");
+	}
+
 	bp->max_tpa_v2 = rte_le_to_cpu_16(resp->max_aggs_supported);
 
 	HWRM_UNLOCK();
@@ -1199,11 +1206,13 @@ int bnxt_hwrm_queue_qportcfg(struct bnxt *bp)
 	int rc = 0;
 	struct hwrm_queue_qportcfg_input req = {.req_type = 0 };
 	struct hwrm_queue_qportcfg_output *resp = bp->hwrm_cmd_resp_addr;
+	uint32_t dir = HWRM_QUEUE_QPORTCFG_INPUT_FLAGS_PATH_TX;
 	int i;
 
+get_rx_info:
 	HWRM_PREP(req, QUEUE_QPORTCFG, BNXT_USE_CHIMP_MB);
 
-	req.flags = HWRM_QUEUE_QPORTCFG_INPUT_FLAGS_PATH_TX;
+	req.flags = rte_cpu_to_le_32(dir);
 	/* HWRM Version >= 1.9.1 */
 	if (bp->hwrm_spec_code >= HWRM_VERSION_1_9_1)
 		req.drv_qmap_cap =
@@ -1212,30 +1221,51 @@ int bnxt_hwrm_queue_qportcfg(struct bnxt *bp)
 
 	HWRM_CHECK_RESULT();
 
-#define GET_QUEUE_INFO(x) \
-	bp->cos_queue[x].id = resp->queue_id##x; \
-	bp->cos_queue[x].profile = resp->queue_id##x##_service_profile
-
-	GET_QUEUE_INFO(0);
-	GET_QUEUE_INFO(1);
-	GET_QUEUE_INFO(2);
-	GET_QUEUE_INFO(3);
-	GET_QUEUE_INFO(4);
-	GET_QUEUE_INFO(5);
-	GET_QUEUE_INFO(6);
-	GET_QUEUE_INFO(7);
+	if (dir == HWRM_QUEUE_QPORTCFG_INPUT_FLAGS_PATH_TX) {
+		GET_TX_QUEUE_INFO(0);
+		GET_TX_QUEUE_INFO(1);
+		GET_TX_QUEUE_INFO(2);
+		GET_TX_QUEUE_INFO(3);
+		GET_TX_QUEUE_INFO(4);
+		GET_TX_QUEUE_INFO(5);
+		GET_TX_QUEUE_INFO(6);
+		GET_TX_QUEUE_INFO(7);
+	} else  {
+		GET_RX_QUEUE_INFO(0);
+		GET_RX_QUEUE_INFO(1);
+		GET_RX_QUEUE_INFO(2);
+		GET_RX_QUEUE_INFO(3);
+		GET_RX_QUEUE_INFO(4);
+		GET_RX_QUEUE_INFO(5);
+		GET_RX_QUEUE_INFO(6);
+		GET_RX_QUEUE_INFO(7);
+	}
 
 	HWRM_UNLOCK();
 
+	if (dir == HWRM_QUEUE_QPORTCFG_INPUT_FLAGS_PATH_RX)
+		goto done;
+
 	if (bp->hwrm_spec_code < HWRM_VERSION_1_9_1) {
-		bp->tx_cosq_id = bp->cos_queue[0].id;
+		bp->tx_cosq_id[0] = bp->tx_cos_queue[0].id;
 	} else {
+		int j;
+
 		/* iterate and find the COSq profile to use for Tx */
-		for (i = 0; i < BNXT_COS_QUEUE_COUNT; i++) {
-			if (bp->cos_queue[i].profile ==
-				HWRM_QUEUE_SERVICE_PROFILE_LOSSY) {
-				bp->tx_cosq_id = bp->cos_queue[i].id;
-				break;
+		if (bp->vnic_cap_flags & BNXT_VNIC_CAP_COS_CLASSIFY) {
+			for (j = 0, i = 0; i < BNXT_COS_QUEUE_COUNT; i++) {
+				if (bp->tx_cos_queue[i].id != 0xff)
+					bp->tx_cosq_id[j++] =
+						bp->tx_cos_queue[i].id;
+			}
+		} else {
+			for (i = BNXT_COS_QUEUE_COUNT - 1; i >= 0; i--) {
+				if (bp->tx_cos_queue[i].profile ==
+					HWRM_QUEUE_SERVICE_PROFILE_LOSSY) {
+					bp->tx_cosq_id[0] =
+						bp->tx_cos_queue[i].id;
+					break;
+				}
 			}
 		}
 	}
@@ -1246,15 +1276,20 @@ int bnxt_hwrm_queue_qportcfg(struct bnxt *bp)
 		bp->max_tc = BNXT_MAX_QUEUE;
 	bp->max_q = bp->max_tc;
 
-	PMD_DRV_LOG(DEBUG, "Tx Cos Queue to use: %d\n", bp->tx_cosq_id);
+	if (dir == HWRM_QUEUE_QPORTCFG_INPUT_FLAGS_PATH_TX) {
+		dir = HWRM_QUEUE_QPORTCFG_INPUT_FLAGS_PATH_RX;
+		goto get_rx_info;
+	}
 
+done:
 	return rc;
 }
 
 int bnxt_hwrm_ring_alloc(struct bnxt *bp,
 			 struct bnxt_ring *ring,
 			 uint32_t ring_type, uint32_t map_index,
-			 uint32_t stats_ctx_id, uint32_t cmpl_ring_id)
+			 uint32_t stats_ctx_id, uint32_t cmpl_ring_id,
+			 uint16_t tx_cosq_id)
 {
 	int rc = 0;
 	uint32_t enables = 0;
@@ -1276,7 +1311,7 @@ int bnxt_hwrm_ring_alloc(struct bnxt *bp,
 		req.ring_type = ring_type;
 		req.cmpl_ring_id = rte_cpu_to_le_16(cmpl_ring_id);
 		req.stat_ctx_id = rte_cpu_to_le_32(stats_ctx_id);
-		req.queue_id = rte_cpu_to_le_16(bp->tx_cosq_id);
+		req.queue_id = rte_cpu_to_le_16(tx_cosq_id);
 		if (stats_ctx_id != INVALID_STATS_CTX_ID)
 			enables |=
 			HWRM_RING_ALLOC_INPUT_ENABLES_STAT_CTX_ID_VALID;
@@ -1682,6 +1717,11 @@ int bnxt_hwrm_vnic_cfg(struct bnxt *bp, struct bnxt_vnic_info *vnic)
 		ctx_enable_flag |= HWRM_VNIC_CFG_INPUT_ENABLES_MRU;
 		ctx_enable_flag |= HWRM_VNIC_CFG_INPUT_ENABLES_RSS_RULE;
 	}
+	if (bp->vnic_cap_flags & BNXT_VNIC_CAP_COS_CLASSIFY) {
+		ctx_enable_flag |= HWRM_VNIC_CFG_INPUT_ENABLES_QUEUE_ID;
+		req.queue_id = rte_cpu_to_le_16(vnic->cos_queue_id);
+	}
+
 	enables |= ctx_enable_flag;
 	req.dflt_ring_grp = rte_cpu_to_le_16(vnic->dflt_ring_grp);
 	req.rss_rule = rte_cpu_to_le_16(vnic->rss_rule);
diff --git a/drivers/net/bnxt/bnxt_hwrm.h b/drivers/net/bnxt/bnxt_hwrm.h
index 8912a4ed3e..fcbce60589 100644
--- a/drivers/net/bnxt/bnxt_hwrm.h
+++ b/drivers/net/bnxt/bnxt_hwrm.h
@@ -52,6 +52,16 @@ HWRM_CFA_ADV_FLOW_MGNT_QCAPS_OUTPUT_FLAGS_L2_HEADER_SOURCE_FIELDS_SUPPORTED
 	HWRM_FUNC_BACKING_STORE_CFG_INPUT_ENABLES_VNIC |       \
 	HWRM_FUNC_BACKING_STORE_CFG_INPUT_ENABLES_STAT)
 
+#define GET_TX_QUEUE_INFO(x) \
+	bp->tx_cos_queue[x].id = resp->queue_id##x; \
+	bp->tx_cos_queue[x].profile =	\
+		resp->queue_id##x##_service_profile
+
+#define GET_RX_QUEUE_INFO(x) \
+	bp->rx_cos_queue[x].id = resp->queue_id##x; \
+	bp->rx_cos_queue[x].profile =	\
+		resp->queue_id##x##_service_profile
+
 int bnxt_hwrm_cfa_l2_clear_rx_mask(struct bnxt *bp,
 				   struct bnxt_vnic_info *vnic);
 int bnxt_hwrm_cfa_l2_set_rx_mask(struct bnxt *bp, struct bnxt_vnic_info *vnic,
@@ -90,7 +100,8 @@ int bnxt_hwrm_set_async_event_cr(struct bnxt *bp);
 int bnxt_hwrm_ring_alloc(struct bnxt *bp,
 			 struct bnxt_ring *ring,
 			 uint32_t ring_type, uint32_t map_index,
-			 uint32_t stats_ctx_id, uint32_t cmpl_ring_id);
+			 uint32_t stats_ctx_id, uint32_t cmpl_ring_id,
+			 uint16_t tx_cosq_id);
 int bnxt_hwrm_ring_free(struct bnxt *bp,
 			struct bnxt_ring *ring, uint32_t ring_type);
 int bnxt_hwrm_ring_grp_alloc(struct bnxt *bp, unsigned int idx);
diff --git a/drivers/net/bnxt/bnxt_ring.c b/drivers/net/bnxt/bnxt_ring.c
index 14cfb8c155..cf0c24c9dc 100644
--- a/drivers/net/bnxt/bnxt_ring.c
+++ b/drivers/net/bnxt/bnxt_ring.c
@@ -424,7 +424,7 @@ static int bnxt_alloc_cmpl_ring(struct bnxt *bp, int queue_index,
 	}
 
 	rc = bnxt_hwrm_ring_alloc(bp, cp_ring, ring_type, cp_ring_index,
-				  HWRM_NA_SIGNATURE, nq_ring_id);
+				  HWRM_NA_SIGNATURE, nq_ring_id, 0);
 	if (rc)
 		return rc;
 
@@ -450,7 +450,7 @@ static int bnxt_alloc_nq_ring(struct bnxt *bp, int queue_index,
 	ring_type = HWRM_RING_ALLOC_INPUT_RING_TYPE_NQ;
 
 	rc = bnxt_hwrm_ring_alloc(bp, nq_ring, ring_type, nq_ring_index,
-				  HWRM_NA_SIGNATURE, HWRM_NA_SIGNATURE);
+				  HWRM_NA_SIGNATURE, HWRM_NA_SIGNATURE, 0);
 	if (rc)
 		return rc;
 
@@ -475,7 +475,7 @@ static int bnxt_alloc_rx_ring(struct bnxt *bp, int queue_index)
 
 	rc = bnxt_hwrm_ring_alloc(bp, ring, ring_type,
 				  queue_index, cpr->hw_stats_ctx_id,
-				  cp_ring->fw_ring_id);
+				  cp_ring->fw_ring_id, 0);
 	if (rc)
 		return rc;
 
@@ -510,7 +510,7 @@ static int bnxt_alloc_rx_agg_ring(struct bnxt *bp, int queue_index)
 	}
 
 	rc = bnxt_hwrm_ring_alloc(bp, ring, ring_type, map_idx,
-				  hw_stats_ctx_id, cp_ring->fw_ring_id);
+				  hw_stats_ctx_id, cp_ring->fw_ring_id, 0);
 
 	if (rc)
 		return rc;
@@ -701,6 +701,7 @@ int bnxt_alloc_hwrm_rings(struct bnxt *bp)
 		struct bnxt_tx_ring_info *txr = txq->tx_ring;
 		struct bnxt_ring *ring = txr->tx_ring_struct;
 		unsigned int idx = i + bp->rx_cp_nr_rings;
+		uint16_t tx_cosq_id = 0;
 
 		if (BNXT_HAS_NQ(bp)) {
 			if (bnxt_alloc_nq_ring(bp, idx, nqr))
@@ -710,12 +711,17 @@ int bnxt_alloc_hwrm_rings(struct bnxt *bp)
 		if (bnxt_alloc_cmpl_ring(bp, idx, cpr, nqr))
 			goto err_out;
 
+		if (bp->vnic_cap_flags & BNXT_VNIC_CAP_COS_CLASSIFY)
+			tx_cosq_id = bp->tx_cosq_id[i < bp->max_lltc ? i : 0];
+		else
+			tx_cosq_id = bp->tx_cosq_id[0];
 		/* Tx ring */
 		ring_type = HWRM_RING_ALLOC_INPUT_RING_TYPE_TX;
 		rc = bnxt_hwrm_ring_alloc(bp, ring,
 					  ring_type,
 					  i, cpr->hw_stats_ctx_id,
-					  cp_ring->fw_ring_id);
+					  cp_ring->fw_ring_id,
+					  tx_cosq_id);
 		if (rc)
 			goto err_out;
 
@@ -747,7 +753,7 @@ int bnxt_alloc_async_cp_ring(struct bnxt *bp)
 		ring_type = HWRM_RING_ALLOC_INPUT_RING_TYPE_L2_CMPL;
 
 	rc = bnxt_hwrm_ring_alloc(bp, cp_ring, ring_type, 0,
-				  HWRM_NA_SIGNATURE, HWRM_NA_SIGNATURE);
+				  HWRM_NA_SIGNATURE, HWRM_NA_SIGNATURE, 0);
 
 	if (rc)
 		return rc;
diff --git a/drivers/net/bnxt/bnxt_rxq.c b/drivers/net/bnxt/bnxt_rxq.c
index 03b115dbaf..5d291cbafd 100644
--- a/drivers/net/bnxt/bnxt_rxq.c
+++ b/drivers/net/bnxt/bnxt_rxq.c
@@ -76,6 +76,7 @@ int bnxt_mq_rx_configure(struct bnxt *bp)
 		switch (dev_conf->rxmode.mq_mode) {
 		case ETH_MQ_RX_VMDQ_RSS:
 		case ETH_MQ_RX_VMDQ_ONLY:
+		case ETH_MQ_RX_VMDQ_DCB_RSS:
 			/* FALLTHROUGH */
 			/* ETH_8/64_POOLs */
 			pools = conf->nb_queue_pools;
@@ -91,7 +92,7 @@ int bnxt_mq_rx_configure(struct bnxt *bp)
 				pools = max_pools;
 			break;
 		case ETH_MQ_RX_RSS:
-			pools = 1;
+			pools = bp->rx_cosq_cnt ? bp->rx_cosq_cnt : 1;
 			break;
 		default:
 			PMD_DRV_LOG(ERR, "Unsupported mq_mod %d\n",
diff --git a/drivers/net/bnxt/bnxt_vnic.h b/drivers/net/bnxt/bnxt_vnic.h
index de34b21eb8..4f760e0b08 100644
--- a/drivers/net/bnxt/bnxt_vnic.h
+++ b/drivers/net/bnxt/bnxt_vnic.h
@@ -45,6 +45,7 @@ struct bnxt_vnic_info {
 	uint16_t	cos_rule;
 	uint16_t	lb_rule;
 	uint16_t	rx_queue_cnt;
+	uint16_t	cos_queue_id;
 	bool		vlan_strip;
 	bool		func_default;
 	bool		bd_stall;
diff --git a/drivers/net/bnxt/hsi_struct_def_dpdk.h b/drivers/net/bnxt/hsi_struct_def_dpdk.h
index 26d12cf20a..c45d0883ac 100644
--- a/drivers/net/bnxt/hsi_struct_def_dpdk.h
+++ b/drivers/net/bnxt/hsi_struct_def_dpdk.h
@@ -21157,7 +21157,7 @@ struct hwrm_vnic_free_output {
  *****************/
 
 
-/* hwrm_vnic_cfg_input (size:320b/40B) */
+/* hwrm_vnic_cfg_input (size:384b/48B) */
 struct hwrm_vnic_cfg_input {
 	/* The HWRM command request type. */
 	uint16_t	req_type;
@@ -21300,6 +21300,9 @@ struct hwrm_vnic_cfg_input {
 	 */
 	#define HWRM_VNIC_CFG_INPUT_ENABLES_DEFAULT_CMPL_RING_ID \
 		UINT32_C(0x40)
+	/* This bit must be '1' for the queue_id field to be configured. */
+	#define HWRM_VNIC_CFG_INPUT_ENABLES_QUEUE_ID \
+		UINT32_C(0x80)
 	/* Logical vnic ID */
 	uint16_t	vnic_id;
 	/*
@@ -21345,6 +21348,19 @@ struct hwrm_vnic_cfg_input {
 	 * be chosen if packet does not match any RSS rules.
 	 */
 	uint16_t	default_cmpl_ring_id;
+	/*
+	 * When specified, only incoming packets classified to the specified CoS
+	 * queue ID will be arriving on this VNIC.  Packet priority to CoS mapping
+	 * rules can be specified using HWRM_QUEUE_PRI2COS_CFG.  In this mode,
+	 * ntuple filters with VNIC destination specified are invalid since they
+	 * conflict with the the CoS to VNIC steering rules in this mode.
+	 *
+	 * If this field is not specified, packet to VNIC steering will be
+	 * subject to the standard L2 filter rules and any additional ntuple
+	 * filter rules with destination VNIC specified.
+	 */
+	uint16_t	queue_id;
+	uint8_t	unused0[6];
 } __attribute__((packed));
 
 /* hwrm_vnic_cfg_output (size:128b/16B) */
@@ -21640,6 +21656,16 @@ struct hwrm_vnic_qcaps_output {
 	 */
 	#define HWRM_VNIC_QCAPS_OUTPUT_FLAGS_OUTERMOST_RSS_CAP \
 		UINT32_C(0x80)
+	/*
+	 * When this bit is '1', it indicates that firmware supports the
+	 * ability to steer incoming packets from one CoS queue to one
+	 * VNIC.  This optional feature can then be enabled
+	 * using HWRM_VNIC_CFG on any VNIC.  This feature is only
+	 * available when NVM option “enable_cos_classfication” is set
+	 * to 1.  If set to '0', firmware does not support this feature.
+	 */
+	#define HWRM_VNIC_QCAPS_OUTPUT_FLAGS_COS_ASSIGNMENT_CAP \
+		UINT32_C(0x100)
 	/*
 	 * This field advertises the maximum concurrent TPA aggregations
 	 * supported by the VNIC on new devices that support TPA v2.
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 5/9] net/bnxt: use common receive transmit nq ring

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

Thor queue scaling is currently limited by the number of NQs that
can be allocated. Fix by using a common NQ for all receive/transmit
rings instead of allocating a separate NQ for each ring.

Fixes: f8168ca0e690 ("net/bnxt: support thor controller")
Cc: stable@dpdk.org
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Kalesh Anakkur Purayil <kalesh-anakkur.purayil@broadcom.com>
---
 drivers/net/bnxt/bnxt.h        |   1 +
 drivers/net/bnxt/bnxt_ethdev.c |   5 ++
 drivers/net/bnxt/bnxt_hwrm.c   |   7 +--
 drivers/net/bnxt/bnxt_ring.c   | 107 ++++++++++++++++++++++-----------
 drivers/net/bnxt/bnxt_ring.h   |   2 +
 drivers/net/bnxt/bnxt_rxq.c    |   4 +-
 drivers/net/bnxt/bnxt_rxq.h    |   1 -
 drivers/net/bnxt/bnxt_rxr.c    |  27 ---------
 drivers/net/bnxt/bnxt_txq.c    |   4 +-
 drivers/net/bnxt/bnxt_txq.h    |   1 -
 drivers/net/bnxt/bnxt_txr.c    |  25 --------
 11 files changed, 84 insertions(+), 100 deletions(-)

diff --git a/drivers/net/bnxt/bnxt.h b/drivers/net/bnxt/bnxt.h
index 5cfe5ee2c7..ad0b18dddd 100644
--- a/drivers/net/bnxt/bnxt.h
+++ b/drivers/net/bnxt/bnxt.h
@@ -497,6 +497,7 @@ struct bnxt {
 
 	/* Default completion ring */
 	struct bnxt_cp_ring_info	*async_cp_ring;
+	struct bnxt_cp_ring_info	*rxtx_nq_ring;
 	uint32_t		max_ring_grps;
 	struct bnxt_ring_grp_info	*grp_info;
 
diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 9adcd94ff8..2845e9185a 100644
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -223,6 +223,7 @@ static void bnxt_free_mem(struct bnxt *bp, bool reconfig)
 		bnxt_free_rx_rings(bp);
 	}
 	bnxt_free_async_cp_ring(bp);
+	bnxt_free_rxtx_nq_ring(bp);
 }
 
 static int bnxt_alloc_mem(struct bnxt *bp, bool reconfig)
@@ -253,6 +254,10 @@ static int bnxt_alloc_mem(struct bnxt *bp, bool reconfig)
 	if (rc)
 		goto alloc_mem_err;
 
+	rc = bnxt_alloc_rxtx_nq_ring(bp);
+	if (rc)
+		goto alloc_mem_err;
+
 	return 0;
 
 alloc_mem_err:
diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index 76ef004237..b5211aea75 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -2325,11 +2325,8 @@ void bnxt_free_hwrm_rx_ring(struct bnxt *bp, int queue_index)
 			bp->grp_info[queue_index].ag_fw_ring_id =
 							INVALID_HW_RING_ID;
 	}
-	if (cpr->cp_ring_struct->fw_ring_id != INVALID_HW_RING_ID) {
+	if (cpr->cp_ring_struct->fw_ring_id != INVALID_HW_RING_ID)
 		bnxt_free_cp_ring(bp, cpr);
-		if (rxq->nq_ring)
-			bnxt_free_nq_ring(bp, rxq->nq_ring);
-	}
 
 	if (BNXT_HAS_RING_GRPS(bp))
 		bp->grp_info[queue_index].cp_fw_ring_id = INVALID_HW_RING_ID;
@@ -2361,8 +2358,6 @@ int bnxt_free_all_hwrm_rings(struct bnxt *bp)
 		if (cpr->cp_ring_struct->fw_ring_id != INVALID_HW_RING_ID) {
 			bnxt_free_cp_ring(bp, cpr);
 			cpr->cp_ring_struct->fw_ring_id = INVALID_HW_RING_ID;
-			if (txq->nq_ring)
-				bnxt_free_nq_ring(bp, txq->nq_ring);
 		}
 	}
 
diff --git a/drivers/net/bnxt/bnxt_ring.c b/drivers/net/bnxt/bnxt_ring.c
index cf0c24c9dc..19fc45395d 100644
--- a/drivers/net/bnxt/bnxt_ring.c
+++ b/drivers/net/bnxt/bnxt_ring.c
@@ -125,7 +125,7 @@ int bnxt_alloc_rings(struct bnxt *bp, uint16_t qidx,
 	int cp_vmem_len = RTE_CACHE_LINE_ROUNDUP(cp_ring->vmem_size);
 	cp_vmem_len = RTE_ALIGN(cp_vmem_len, 128);
 
-	int nq_vmem_len = BNXT_CHIP_THOR(bp) ?
+	int nq_vmem_len = nq_ring_info ?
 		RTE_CACHE_LINE_ROUNDUP(cp_ring->vmem_size) : 0;
 	nq_vmem_len = RTE_ALIGN(nq_vmem_len, 128);
 
@@ -159,7 +159,7 @@ int bnxt_alloc_rings(struct bnxt *bp, uint16_t qidx,
 	nq_ring_start = cp_ring_start + cp_ring_len;
 	nq_ring_start = RTE_ALIGN(nq_ring_start, 4096);
 
-	int nq_ring_len = BNXT_CHIP_THOR(bp) ? cp_ring_len : 0;
+	int nq_ring_len = nq_ring_info ? cp_ring_len : 0;
 
 	int tx_ring_start = nq_ring_start + nq_ring_len;
 	tx_ring_start = RTE_ALIGN(tx_ring_start, 4096);
@@ -403,12 +403,12 @@ static void bnxt_set_db(struct bnxt *bp,
 }
 
 static int bnxt_alloc_cmpl_ring(struct bnxt *bp, int queue_index,
-				struct bnxt_cp_ring_info *cpr,
-				struct bnxt_cp_ring_info *nqr)
+				struct bnxt_cp_ring_info *cpr)
 {
 	struct bnxt_ring *cp_ring = cpr->cp_ring_struct;
 	uint32_t nq_ring_id = HWRM_NA_SIGNATURE;
 	int cp_ring_index = queue_index + BNXT_NUM_ASYNC_CPR(bp);
+	struct bnxt_cp_ring_info *nqr = bp->rxtx_nq_ring;
 	uint8_t ring_type;
 	int rc = 0;
 
@@ -436,31 +436,85 @@ static int bnxt_alloc_cmpl_ring(struct bnxt *bp, int queue_index,
 	return 0;
 }
 
-static int bnxt_alloc_nq_ring(struct bnxt *bp, int queue_index,
-			      struct bnxt_cp_ring_info *nqr)
+int bnxt_alloc_rxtx_nq_ring(struct bnxt *bp)
 {
-	struct bnxt_ring *nq_ring = nqr->cp_ring_struct;
-	int nq_ring_index = queue_index + BNXT_NUM_ASYNC_CPR(bp);
+	struct bnxt_cp_ring_info *nqr;
+	struct bnxt_ring *ring;
+	int ring_index = BNXT_NUM_ASYNC_CPR(bp);
+	unsigned int socket_id;
 	uint8_t ring_type;
 	int rc = 0;
 
-	if (!BNXT_HAS_NQ(bp))
-		return -EINVAL;
+	if (!BNXT_HAS_NQ(bp) || bp->rxtx_nq_ring)
+		return 0;
+
+	socket_id = rte_lcore_to_socket_id(rte_get_master_lcore());
+
+	nqr = rte_zmalloc_socket("nqr",
+				 sizeof(struct bnxt_cp_ring_info),
+				 RTE_CACHE_LINE_SIZE, socket_id);
+	if (nqr == NULL)
+		return -ENOMEM;
+
+	ring = rte_zmalloc_socket("bnxt_cp_ring_struct",
+				  sizeof(struct bnxt_ring),
+				  RTE_CACHE_LINE_SIZE, socket_id);
+	if (ring == NULL) {
+		rte_free(nqr);
+		return -ENOMEM;
+	}
+
+	ring->bd = (void *)nqr->cp_desc_ring;
+	ring->bd_dma = nqr->cp_desc_mapping;
+	ring->ring_size = rte_align32pow2(DEFAULT_CP_RING_SIZE);
+	ring->ring_mask = ring->ring_size - 1;
+	ring->vmem_size = 0;
+	ring->vmem = NULL;
+
+	nqr->cp_ring_struct = ring;
+	rc = bnxt_alloc_rings(bp, 0, NULL, NULL, nqr, NULL, "l2_nqr");
+	if (rc) {
+		rte_free(ring);
+		rte_free(nqr);
+		return -ENOMEM;
+	}
 
 	ring_type = HWRM_RING_ALLOC_INPUT_RING_TYPE_NQ;
 
-	rc = bnxt_hwrm_ring_alloc(bp, nq_ring, ring_type, nq_ring_index,
+	rc = bnxt_hwrm_ring_alloc(bp, ring, ring_type, ring_index,
 				  HWRM_NA_SIGNATURE, HWRM_NA_SIGNATURE, 0);
-	if (rc)
+	if (rc) {
+		rte_free(ring);
+		rte_free(nqr);
 		return rc;
+	}
 
-	bnxt_set_db(bp, &nqr->cp_db, ring_type, nq_ring_index,
-		    nq_ring->fw_ring_id);
+	bnxt_set_db(bp, &nqr->cp_db, ring_type, ring_index,
+		    ring->fw_ring_id);
 	bnxt_db_nq(nqr);
 
+	bp->rxtx_nq_ring = nqr;
+
 	return 0;
 }
 
+/* Free RX/TX NQ ring.  */
+void bnxt_free_rxtx_nq_ring(struct bnxt *bp)
+{
+	struct bnxt_cp_ring_info *nqr = bp->rxtx_nq_ring;
+
+	if (!nqr)
+		return;
+
+	bnxt_free_nq_ring(bp, nqr);
+
+	bnxt_free_ring(nqr->cp_ring_struct);
+	rte_free(nqr->cp_ring_struct);
+	nqr->cp_ring_struct = NULL;
+	rte_free(nqr);
+	bp->rxtx_nq_ring = NULL;
+}
+
 static int bnxt_alloc_rx_ring(struct bnxt *bp, int queue_index)
 {
 	struct bnxt_rx_queue *rxq = bp->rx_queues[queue_index];
@@ -529,17 +583,10 @@ int bnxt_alloc_hwrm_rx_ring(struct bnxt *bp, int queue_index)
 	struct bnxt_rx_queue *rxq = bp->rx_queues[queue_index];
 	struct bnxt_cp_ring_info *cpr = rxq->cp_ring;
 	struct bnxt_ring *cp_ring = cpr->cp_ring_struct;
-	struct bnxt_cp_ring_info *nqr = rxq->nq_ring;
 	struct bnxt_rx_ring_info *rxr = rxq->rx_ring;
 	int rc;
 
-	if (BNXT_HAS_NQ(bp)) {
-		rc = bnxt_alloc_nq_ring(bp, queue_index, nqr);
-		if (rc)
-			goto err_out;
-	}
-
-	rc = bnxt_alloc_cmpl_ring(bp, queue_index, cpr, nqr);
+	rc = bnxt_alloc_cmpl_ring(bp, queue_index, cpr);
 	if (rc)
 		goto err_out;
 
@@ -644,16 +691,10 @@ int bnxt_alloc_hwrm_rings(struct bnxt *bp)
 	for (i = 0; i < bp->rx_cp_nr_rings; i++) {
 		struct bnxt_rx_queue *rxq = bp->rx_queues[i];
 		struct bnxt_cp_ring_info *cpr = rxq->cp_ring;
-		struct bnxt_cp_ring_info *nqr = rxq->nq_ring;
 		struct bnxt_ring *cp_ring = cpr->cp_ring_struct;
 		struct bnxt_rx_ring_info *rxr = rxq->rx_ring;
 
-		if (BNXT_HAS_NQ(bp)) {
-			if (bnxt_alloc_nq_ring(bp, i, nqr))
-				goto err_out;
-		}
-
-		if (bnxt_alloc_cmpl_ring(bp, i, cpr, nqr))
+		if (bnxt_alloc_cmpl_ring(bp, i, cpr))
 			goto err_out;
 
 		if (BNXT_HAS_RING_GRPS(bp)) {
@@ -697,18 +738,12 @@ int bnxt_alloc_hwrm_rings(struct bnxt *bp)
 		struct bnxt_tx_queue *txq = bp->tx_queues[i];
 		struct bnxt_cp_ring_info *cpr = txq->cp_ring;
 		struct bnxt_ring *cp_ring = cpr->cp_ring_struct;
-		struct bnxt_cp_ring_info *nqr = txq->nq_ring;
 		struct bnxt_tx_ring_info *txr = txq->tx_ring;
 		struct bnxt_ring *ring = txr->tx_ring_struct;
 		unsigned int idx = i + bp->rx_cp_nr_rings;
 		uint16_t tx_cosq_id = 0;
 
-		if (BNXT_HAS_NQ(bp)) {
-			if (bnxt_alloc_nq_ring(bp, idx, nqr))
-				goto err_out;
-		}
-
-		if (bnxt_alloc_cmpl_ring(bp, idx, cpr, nqr))
+		if (bnxt_alloc_cmpl_ring(bp, idx, cpr))
 			goto err_out;
 
 		if (bp->vnic_cap_flags & BNXT_VNIC_CAP_COS_CLASSIFY)
diff --git a/drivers/net/bnxt/bnxt_ring.h b/drivers/net/bnxt/bnxt_ring.h
index a5d5106986..833118391b 100644
--- a/drivers/net/bnxt/bnxt_ring.h
+++ b/drivers/net/bnxt/bnxt_ring.h
@@ -78,6 +78,8 @@ int bnxt_alloc_hwrm_rings(struct bnxt *bp);
 int bnxt_alloc_async_cp_ring(struct bnxt *bp);
 void bnxt_free_async_cp_ring(struct bnxt *bp);
 int bnxt_alloc_async_ring_struct(struct bnxt *bp);
+int bnxt_alloc_rxtx_nq_ring(struct bnxt *bp);
+void bnxt_free_rxtx_nq_ring(struct bnxt *bp);
 
 static inline void bnxt_db_write(struct bnxt_db_info *db, uint32_t idx)
 {
diff --git a/drivers/net/bnxt/bnxt_rxq.c b/drivers/net/bnxt/bnxt_rxq.c
index 5d291cbafd..9439fcd1fb 100644
--- a/drivers/net/bnxt/bnxt_rxq.c
+++ b/drivers/net/bnxt/bnxt_rxq.c
@@ -341,8 +341,8 @@ int bnxt_rx_queue_setup_op(struct rte_eth_dev *eth_dev,
 
 	eth_dev->data->rx_queues[queue_idx] = rxq;
 	/* Allocate RX ring hardware descriptors */
-	if (bnxt_alloc_rings(bp, queue_idx, NULL, rxq, rxq->cp_ring,
-			rxq->nq_ring, "rxr")) {
+	if (bnxt_alloc_rings(bp, queue_idx, NULL, rxq, rxq->cp_ring, NULL,
+			     "rxr")) {
 		PMD_DRV_LOG(ERR,
 			"ring_dma_zone_reserve for rx_ring failed!\n");
 		bnxt_rx_queue_release_op(rxq);
diff --git a/drivers/net/bnxt/bnxt_rxq.h b/drivers/net/bnxt/bnxt_rxq.h
index 3693d89a60..4f5182d9e9 100644
--- a/drivers/net/bnxt/bnxt_rxq.h
+++ b/drivers/net/bnxt/bnxt_rxq.h
@@ -39,7 +39,6 @@ struct bnxt_rx_queue {
 	uint32_t			rx_buf_size;
 	struct bnxt_rx_ring_info	*rx_ring;
 	struct bnxt_cp_ring_info	*cp_ring;
-	struct bnxt_cp_ring_info	*nq_ring;
 	rte_atomic64_t		rx_mbuf_alloc_fail;
 	const struct rte_memzone *mz;
 };
diff --git a/drivers/net/bnxt/bnxt_rxr.c b/drivers/net/bnxt/bnxt_rxr.c
index 1a6fb7944b..bda4f4c1b9 100644
--- a/drivers/net/bnxt/bnxt_rxr.c
+++ b/drivers/net/bnxt/bnxt_rxr.c
@@ -742,7 +742,6 @@ void bnxt_free_rx_rings(struct bnxt *bp)
 int bnxt_init_rx_ring_struct(struct bnxt_rx_queue *rxq, unsigned int socket_id)
 {
 	struct bnxt_cp_ring_info *cpr;
-	struct bnxt_cp_ring_info *nqr;
 	struct bnxt_rx_ring_info *rxr;
 	struct bnxt_ring *ring;
 
@@ -789,32 +788,6 @@ int bnxt_init_rx_ring_struct(struct bnxt_rx_queue *rxq, unsigned int socket_id)
 	ring->vmem_size = 0;
 	ring->vmem = NULL;
 
-	if (BNXT_HAS_NQ(rxq->bp)) {
-		nqr = rte_zmalloc_socket("bnxt_rx_ring_cq",
-					 sizeof(struct bnxt_cp_ring_info),
-					 RTE_CACHE_LINE_SIZE, socket_id);
-		if (nqr == NULL)
-			return -ENOMEM;
-
-		rxq->nq_ring = nqr;
-
-		ring = rte_zmalloc_socket("bnxt_rx_ring_struct",
-					  sizeof(struct bnxt_ring),
-					  RTE_CACHE_LINE_SIZE, socket_id);
-		if (ring == NULL)
-			return -ENOMEM;
-
-		nqr->cp_ring_struct = ring;
-		ring->ring_size =
-			rte_align32pow2(rxr->rx_ring_struct->ring_size *
-					(2 + AGG_RING_SIZE_FACTOR));
-		ring->ring_mask = ring->ring_size - 1;
-		ring->bd = (void *)nqr->cp_desc_ring;
-		ring->bd_dma = nqr->cp_desc_mapping;
-		ring->vmem_size = 0;
-		ring->vmem = NULL;
-	}
-
 	/* Allocate Aggregator rings */
 	ring = rte_zmalloc_socket("bnxt_rx_ring_struct",
 				   sizeof(struct bnxt_ring),
diff --git a/drivers/net/bnxt/bnxt_txq.c b/drivers/net/bnxt/bnxt_txq.c
index ea20d737fe..5ad4ee155e 100644
--- a/drivers/net/bnxt/bnxt_txq.c
+++ b/drivers/net/bnxt/bnxt_txq.c
@@ -141,8 +141,8 @@ int bnxt_tx_queue_setup_op(struct rte_eth_dev *eth_dev,
 	txq->port_id = eth_dev->data->port_id;
 
 	/* Allocate TX ring hardware descriptors */
-	if (bnxt_alloc_rings(bp, queue_idx, txq, NULL, txq->cp_ring,
-			txq->nq_ring, "txr")) {
+	if (bnxt_alloc_rings(bp, queue_idx, txq, NULL, txq->cp_ring, NULL,
+			     "txr")) {
 		PMD_DRV_LOG(ERR, "ring_dma_zone_reserve for tx_ring failed!");
 		bnxt_tx_queue_release_op(txq);
 		rc = -ENOMEM;
diff --git a/drivers/net/bnxt/bnxt_txq.h b/drivers/net/bnxt/bnxt_txq.h
index 7a442516d2..37a3f9539f 100644
--- a/drivers/net/bnxt/bnxt_txq.h
+++ b/drivers/net/bnxt/bnxt_txq.h
@@ -33,7 +33,6 @@ struct bnxt_tx_queue {
 
 	unsigned int		cp_nr_rings;
 	struct bnxt_cp_ring_info	*cp_ring;
-	struct bnxt_cp_ring_info        *nq_ring;
 	const struct rte_memzone *mz;
 	struct rte_mbuf **free;
 };
diff --git a/drivers/net/bnxt/bnxt_txr.c b/drivers/net/bnxt/bnxt_txr.c
index 0ed6581bed..6e2ee86c05 100644
--- a/drivers/net/bnxt/bnxt_txr.c
+++ b/drivers/net/bnxt/bnxt_txr.c
@@ -57,7 +57,6 @@ int bnxt_init_one_tx_ring(struct bnxt_tx_queue *txq)
 int bnxt_init_tx_ring_struct(struct bnxt_tx_queue *txq, unsigned int socket_id)
 {
 	struct bnxt_cp_ring_info *cpr;
-	struct bnxt_cp_ring_info *nqr;
 	struct bnxt_tx_ring_info *txr;
 	struct bnxt_ring *ring;
 
@@ -101,30 +100,6 @@ int bnxt_init_tx_ring_struct(struct bnxt_tx_queue *txq, unsigned int socket_id)
 	ring->vmem_size = 0;
 	ring->vmem = NULL;
 
-	if (BNXT_HAS_NQ(txq->bp)) {
-		nqr = rte_zmalloc_socket("bnxt_tx_ring_nq",
-					 sizeof(struct bnxt_cp_ring_info),
-					 RTE_CACHE_LINE_SIZE, socket_id);
-		if (nqr == NULL)
-			return -ENOMEM;
-
-		txq->nq_ring = nqr;
-
-		ring = rte_zmalloc_socket("bnxt_tx_ring_struct",
-					  sizeof(struct bnxt_ring),
-					  RTE_CACHE_LINE_SIZE, socket_id);
-		if (ring == NULL)
-			return -ENOMEM;
-
-		nqr->cp_ring_struct = ring;
-		ring->ring_size = txr->tx_ring_struct->ring_size;
-		ring->ring_mask = ring->ring_size - 1;
-		ring->bd = (void *)nqr->cp_desc_ring;
-		ring->bd_dma = nqr->cp_desc_mapping;
-		ring->vmem_size = 0;
-		ring->vmem = NULL;
-	}
-
 	return 0;
 }
 
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 6/9] net/bnxt: fix stats context calculation

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

The required number of statistics contexts is computed as the sum
of the number of receive and transmit rings plus one for the async
completion ring. A statistics context is not actually required for
the async completion ring, so remove it from the calculation.

Fixes: bd0a14c99f65 ("net/bnxt: use dedicated CPR for async events")
Cc: stable@dpdk.org

Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_hwrm.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index b5211aea75..1e65c3b80b 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -851,9 +851,7 @@ int bnxt_hwrm_func_reserve_vf_resc(struct bnxt *bp, bool test)
 	req.num_tx_rings = rte_cpu_to_le_16(bp->tx_nr_rings);
 	req.num_rx_rings = rte_cpu_to_le_16(bp->rx_nr_rings *
 					    AGG_RING_MULTIPLIER);
-	req.num_stat_ctxs = rte_cpu_to_le_16(bp->rx_nr_rings +
-					     bp->tx_nr_rings +
-					     BNXT_NUM_ASYNC_CPR(bp));
+	req.num_stat_ctxs = rte_cpu_to_le_16(bp->rx_nr_rings + bp->tx_nr_rings);
 	req.num_cmpl_rings = rte_cpu_to_le_16(bp->rx_nr_rings +
 					      bp->tx_nr_rings +
 					      BNXT_NUM_ASYNC_CPR(bp));
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 7/9] net/bnxt: use correct default Rx queue for thor

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

Use first receive queue assigned to VNIC as the default receive queue
when configuring Thor VNICs. This is necessary e.g. in order for flow
redirection to a specific receive queue to work correctly.

Fixes: f8168ca0e690 ("net/bnxt: support thor controller")
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Kumar Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_hwrm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index 1e65c3b80b..0d5362581e 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -1692,7 +1692,8 @@ int bnxt_hwrm_vnic_cfg(struct bnxt *bp, struct bnxt_vnic_info *vnic)
 	HWRM_PREP(req, VNIC_CFG, BNXT_USE_CHIMP_MB);
 
 	if (BNXT_CHIP_THOR(bp)) {
-		struct bnxt_rx_queue *rxq = bp->eth_dev->data->rx_queues[0];
+		struct bnxt_rx_queue *rxq =
+			bp->eth_dev->data->rx_queues[vnic->start_grp_id];
 		struct bnxt_rx_ring_info *rxr = rxq->rx_ring;
 		struct bnxt_cp_ring_info *cpr = rxq->cp_ring;
 
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 8/9] net/bnxt: advertise scatter receive offload capability

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

Scattered receive is supported but not included in receive offload
capabilities. Fix by adding it and including in scattered receive
calculation.

Fixes: 9c1507d96ab8 ("net/bnxt: switch to the new offload API")
Cc: stable@dpdk.org
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_ethdev.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 2845e9185a..5160ac002b 100644
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -163,7 +163,8 @@ static const struct rte_pci_id bnxt_pci_id_map[] = {
 				     DEV_RX_OFFLOAD_JUMBO_FRAME | \
 				     DEV_RX_OFFLOAD_KEEP_CRC | \
 				     DEV_RX_OFFLOAD_VLAN_EXTEND | \
-				     DEV_RX_OFFLOAD_TCP_LRO)
+				     DEV_RX_OFFLOAD_TCP_LRO | \
+				     DEV_RX_OFFLOAD_SCATTER)
 
 static int bnxt_vlan_offload_set_op(struct rte_eth_dev *dev, int mask);
 static void bnxt_print_link_info(struct rte_eth_dev *eth_dev);
@@ -749,6 +750,9 @@ static int bnxt_scattered_rx(struct rte_eth_dev *eth_dev)
 	uint16_t buf_size;
 	int i;
 
+	if (eth_dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_SCATTER)
+		return 1;
+
 	for (i = 0; i < eth_dev->data->nb_rx_queues; i++) {
 		struct bnxt_rx_queue *rxq = eth_dev->data->rx_queues[i];
 
-- 
2.20.1 (Apple Git-117)

[dpdk-dev] [PATCH v2 9/9] net/bnxt: improve CPR handling in vector PMD

[Ajit Khaparde]

From: Lance Richardson <lance.richardson@broadcom.com>

Reduce overhead of CPR descriptor validity checking in vector
receive and transmit functions.

Preserve raw cpr consumer index in vector transmit completion
function.

Remove an unneeded prefetch (per benchmarking) from vector
transmit completion function.

Fixes: bc4a000f2f53 ("net/bnxt: implement SSE vector mode")
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Kumar Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_rxtx_vec_sse.c | 26 ++++----------------------
 1 file changed, 4 insertions(+), 22 deletions(-)

diff --git a/drivers/net/bnxt/bnxt_rxtx_vec_sse.c b/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
index 029053e305..22d9f9e84a 100644
--- a/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
+++ b/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
@@ -245,10 +245,6 @@ bnxt_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
 		if (!CMP_VALID(rxcmp, raw_cons, cpr->cp_ring_struct))
 			break;
 
-		cpr->valid = FLIP_VALID(cons,
-					cpr->cp_ring_struct->ring_mask,
-					cpr->valid);
-
 		if (likely(CMP_TYPE(rxcmp) == RX_PKT_CMPL_TYPE_RX_L2)) {
 			struct rx_pkt_cmpl_hi *rxcmp1;
 			uint32_t tmp_raw_cons;
@@ -272,10 +268,6 @@ bnxt_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
 			rte_prefetch0(mbuf);
 			rxr->rx_buf_ring[cons].mbuf = NULL;
 
-			cpr->valid = FLIP_VALID(cp_cons,
-						cpr->cp_ring_struct->ring_mask,
-						cpr->valid);
-
 			/* Set constant fields from mbuf initializer. */
 			_mm_store_si128((__m128i *)&mbuf->rearm_data,
 					mbuf_init);
@@ -318,22 +310,13 @@ bnxt_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
 
 	rxq->rxrearm_nb += nb_rx_pkts;
 	cpr->cp_raw_cons = raw_cons;
+	cpr->valid = !!(cpr->cp_raw_cons & cpr->cp_ring_struct->ring_size);
 	if (nb_rx_pkts || evt)
 		bnxt_db_cq(cpr);
 
 	return nb_rx_pkts;
 }
 
-static inline void bnxt_next_cmpl(struct bnxt_cp_ring_info *cpr, uint32_t *idx,
-				  bool *v, uint32_t inc)
-{
-	*idx += inc;
-	if (unlikely(*idx == cpr->cp_ring_struct->ring_size)) {
-		*v = !*v;
-		*idx = 0;
-	}
-}
-
 static void
 bnxt_tx_cmp_vec(struct bnxt_tx_queue *txq, int nr_pkts)
 {
@@ -379,10 +362,8 @@ bnxt_handle_tx_cp_vec(struct bnxt_tx_queue *txq)
 		cons = RING_CMPL(ring_mask, raw_cons);
 		txcmp = (struct tx_cmpl *)&cp_desc_ring[cons];
 
-		if (!CMPL_VALID(txcmp, cpr->valid))
+		if (!CMP_VALID(txcmp, raw_cons, cp_ring_struct))
 			break;
-		bnxt_next_cmpl(cpr, &cons, &cpr->valid, 1);
-		rte_prefetch0(&cp_desc_ring[cons]);
 
 		if (likely(CMP_TYPE(txcmp) == TX_CMPL_TYPE_TX_L2))
 			nb_tx_pkts += txcmp->opaque;
@@ -390,9 +371,10 @@ bnxt_handle_tx_cp_vec(struct bnxt_tx_queue *txq)
 			RTE_LOG_DP(ERR, PMD,
 				   "Unhandled CMP type %02x\n",
 				   CMP_TYPE(txcmp));
-		raw_cons = cons;
+		raw_cons = NEXT_RAW_CMP(raw_cons);
 	} while (nb_tx_pkts < ring_mask);
 
+	cpr->valid = !!(raw_cons & cp_ring_struct->ring_size);
 	if (nb_tx_pkts) {
 		bnxt_tx_cmp_vec(txq, nb_tx_pkts);
 		cpr->cp_raw_cons = raw_cons;
-- 
2.20.1 (Apple Git-117)

Re: [dpdk-dev] [PATCH v2 0/9] bnxt patchset

[Ferruh Yigit]

On 10/4/2019 4:48 AM, Ajit Khaparde wrote:
> Apart form fixes support for Thor and vector PMD, this patchset
> adds support for LRO on Thor based adapters and CoS classification.
> Patchset against dpdk-next-net, has been compiled & tested on an
> x86_64 system.
> 
> Please apply.
> 
> 
> Lance Richardson (8):
>   net/bnxt: increase tqm entry allocation
>   net/bnxt: fix ring alignment for thor-based adapters
>   net/bnxt: add support for LRO on thor adapters
>   net/bnxt: use common receive transmit nq ring
>   net/bnxt: fix stats context calculation
>   net/bnxt: use correct default Rx queue for thor
>   net/bnxt: advertise scatter receive offload capability
>   net/bnxt: improve CPR handling in vector PMD
> 
> Venkat Duvvuru (1):
>   net/bnxt: add support for CoS classification
> 

Series applied to dpdk-next-net/master, thanks.