All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 0/2] net/mlx5: fix packet type support
@ 2017-07-26 19:29 Yongseok Koh
  2017-07-26 19:29 ` [PATCH 1/2] net/mlx5: fix missing packet type calculation Yongseok Koh
                   ` (2 more replies)
  0 siblings, 3 replies; 6+ messages in thread
From: Yongseok Koh @ 2017-07-26 19:29 UTC (permalink / raw)
  To: adrien.mazarguil, nelio.laranjeiro; +Cc: dev, Yongseok Koh

Even though HW supports more packet types, PMD doesn't parse it. Changed to have
more packet_type flags in mbuf.

Yongseok Koh (2):
  net/mlx5: fix missing packet type calculation
  net/mlx5: fix L4 packet type support

 drivers/net/mlx5/mlx5.c              |   2 +
 drivers/net/mlx5/mlx5_ethdev.c       |  10 +-
 drivers/net/mlx5/mlx5_rxtx.c         | 178 +++++++++++++++++++++++++----------
 drivers/net/mlx5/mlx5_rxtx.h         |   3 +-
 drivers/net/mlx5/mlx5_rxtx_vec_sse.c |  17 ++--
 5 files changed, 153 insertions(+), 57 deletions(-)

-- 
2.11.0

^ permalink raw reply	[flat|nested] 6+ messages in thread
* [PATCH 1/2] net/mlx5: fix missing packet type calculation
  2017-07-26 19:29 [PATCH 0/2] net/mlx5: fix packet type support Yongseok Koh
@ 2017-07-26 19:29 ` Yongseok Koh
  2017-07-30  9:08   ` Sagi Grimberg
  2017-07-26 19:29 ` [PATCH 2/2] net/mlx5: fix L4 packet type support Yongseok Koh
  2017-07-31 16:29 ` [PATCH 0/2] net/mlx5: fix " Ferruh Yigit
  2 siblings, 1 reply; 6+ messages in thread
From: Yongseok Koh @ 2017-07-26 19:29 UTC (permalink / raw)
  To: adrien.mazarguil, nelio.laranjeiro; +Cc: dev, Yongseok Koh

Calculation of packet type is currently enabled only when HW checksum is
enabled. This isn't related to HW checksum offload. Enable it regardless.

Fixes: 081f7eae242e ("mlx5: process offload flags only when requested")

Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
---
 drivers/net/mlx5/mlx5_rxtx.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_rxtx.c b/drivers/net/mlx5/mlx5_rxtx.c
index 29ce91b05..5a24e15ea 100644
--- a/drivers/net/mlx5/mlx5_rxtx.c
+++ b/drivers/net/mlx5/mlx5_rxtx.c
@@ -1755,7 +1755,7 @@ mlx5_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n)
 			pkt = seg;
 			assert(len >= (rxq->crc_present << 2));
 			/* Update packet information. */
-			pkt->packet_type = 0;
+			pkt->packet_type = rxq_cq_to_pkt_type(cqe);
 			pkt->ol_flags = 0;
 			if (rss_hash_res && rxq->rss_hash) {
 				pkt->hash.rss = rss_hash_res;
@@ -1773,10 +1773,8 @@ mlx5_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n)
 						mlx5_flow_mark_get(mark);
 				}
 			}
-			if (rxq->csum | rxq->csum_l2tun) {
-				pkt->packet_type = rxq_cq_to_pkt_type(cqe);
+			if (rxq->csum | rxq->csum_l2tun)
 				pkt->ol_flags |= rxq_cq_to_ol_flags(rxq, cqe);
-			}
 			if (rxq->vlan_strip &&
 			    (cqe->hdr_type_etc &
 			     htons(MLX5_CQE_VLAN_STRIPPED))) {
-- 
2.11.0

^ permalink raw reply related	[flat|nested] 6+ messages in thread
* [PATCH 2/2] net/mlx5: fix L4 packet type support
  2017-07-26 19:29 [PATCH 0/2] net/mlx5: fix packet type support Yongseok Koh
  2017-07-26 19:29 ` [PATCH 1/2] net/mlx5: fix missing packet type calculation Yongseok Koh
@ 2017-07-26 19:29 ` Yongseok Koh
  2017-07-30  9:08   ` Sagi Grimberg
  2017-07-31 16:29 ` [PATCH 0/2] net/mlx5: fix " Ferruh Yigit
  2 siblings, 1 reply; 6+ messages in thread
From: Yongseok Koh @ 2017-07-26 19:29 UTC (permalink / raw)
  To: adrien.mazarguil, nelio.laranjeiro; +Cc: dev, Yongseok Koh

TCP/UDP/NONFRAG/FRAG flags aren't counted for both outer and inner header
even though device supports it.

Fixes: 0603df73a077 ("net/mlx5: fix Rx packet validation and type")
Fixes: 6cb559d67b83 ("net/mlx5: add vectorized Rx/Tx burst for x86")

Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
---
 drivers/net/mlx5/mlx5.c              |   2 +
 drivers/net/mlx5/mlx5_ethdev.c       |  10 +-
 drivers/net/mlx5/mlx5_rxtx.c         | 172 ++++++++++++++++++++++++++---------
 drivers/net/mlx5/mlx5_rxtx.h         |   3 +-
 drivers/net/mlx5/mlx5_rxtx_vec_sse.c |  17 ++--
 5 files changed, 151 insertions(+), 53 deletions(-)

diff --git a/drivers/net/mlx5/mlx5.c b/drivers/net/mlx5/mlx5.c
index 49d4dba2a..1f438476a 100644
--- a/drivers/net/mlx5/mlx5.c
+++ b/drivers/net/mlx5/mlx5.c
@@ -887,6 +887,8 @@ RTE_INIT(rte_mlx5_pmd_init);
 static void
 rte_mlx5_pmd_init(void)
 {
+	/* Build the static table for ptype conversion. */
+	mlx5_set_ptype_table();
 	/*
 	 * RDMAV_HUGEPAGES_SAFE tells ibv_fork_init() we intend to use
 	 * huge pages. Calling ibv_fork_init() during init allows
diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
index b70b7b9a9..d1ad61c20 100644
--- a/drivers/net/mlx5/mlx5_ethdev.c
+++ b/drivers/net/mlx5/mlx5_ethdev.c
@@ -715,12 +715,20 @@ mlx5_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 {
 	static const uint32_t ptypes[] = {
 		/* refers to rxq_cq_to_pkt_type() */
+		RTE_PTYPE_L2_ETHER,
 		RTE_PTYPE_L3_IPV4_EXT_UNKNOWN,
 		RTE_PTYPE_L3_IPV6_EXT_UNKNOWN,
+		RTE_PTYPE_L4_NONFRAG,
+		RTE_PTYPE_L4_FRAG,
+		RTE_PTYPE_L4_TCP,
+		RTE_PTYPE_L4_UDP,
 		RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN,
 		RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN,
+		RTE_PTYPE_INNER_L4_NONFRAG,
+		RTE_PTYPE_INNER_L4_FRAG,
+		RTE_PTYPE_INNER_L4_TCP,
+		RTE_PTYPE_INNER_L4_UDP,
 		RTE_PTYPE_UNKNOWN
-
 	};
 
 	if (dev->rx_pkt_burst == mlx5_rx_burst ||
diff --git a/drivers/net/mlx5/mlx5_rxtx.c b/drivers/net/mlx5/mlx5_rxtx.c
index 5a24e15ea..2572a1652 100644
--- a/drivers/net/mlx5/mlx5_rxtx.c
+++ b/drivers/net/mlx5/mlx5_rxtx.c
@@ -79,31 +79,123 @@ mlx5_rx_poll_len(struct rxq *rxq, volatile struct mlx5_cqe *cqe,
 static __rte_always_inline uint32_t
 rxq_cq_to_ol_flags(struct rxq *rxq, volatile struct mlx5_cqe *cqe);
 
-/*
- * The index to the array should have:
- * bit[1:0] = l3_hdr_type, bit[2] = tunneled, bit[3] = outer_l3_type
- */
-const uint32_t mlx5_ptype_table[] = {
-	RTE_PTYPE_UNKNOWN,
-	RTE_PTYPE_L3_IPV6_EXT_UNKNOWN,               /* b0001 */
-	RTE_PTYPE_L3_IPV4_EXT_UNKNOWN,               /* b0010 */
-	RTE_PTYPE_UNKNOWN, RTE_PTYPE_UNKNOWN,
-	RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
-		RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN, /* b0101 */
-	RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
-		RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN, /* b0110 */
-	RTE_PTYPE_UNKNOWN, RTE_PTYPE_UNKNOWN,
-	RTE_PTYPE_L3_IPV6_EXT_UNKNOWN,               /* b1001 */
-	RTE_PTYPE_L3_IPV4_EXT_UNKNOWN,               /* b1010 */
-	RTE_PTYPE_UNKNOWN, RTE_PTYPE_UNKNOWN,
-	RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
-		RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN, /* b1101 */
-	RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
-		RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN, /* b1110 */
-	RTE_PTYPE_ALL_MASK			     /* b1111 */
+uint32_t mlx5_ptype_table[] __rte_cache_aligned = {
+	[0xff] = RTE_PTYPE_ALL_MASK, /* Last entry for errored packet. */
 };
 
 /**
+ * Build a table to translate Rx completion flags to packet type.
+ *
+ * @note: fix mlx5_dev_supported_ptypes_get() if any change here.
+ */
+void
+mlx5_set_ptype_table(void)
+{
+	unsigned int i;
+	uint32_t (*p)[RTE_DIM(mlx5_ptype_table)] = &mlx5_ptype_table;
+
+	for (i = 0; i < RTE_DIM(mlx5_ptype_table); ++i)
+		(*p)[i] = RTE_PTYPE_UNKNOWN;
+	/*
+	 * The index to the array should have:
+	 * bit[1:0] = l3_hdr_type
+	 * bit[4:2] = l4_hdr_type
+	 * bit[5] = ip_frag
+	 * bit[6] = tunneled
+	 * bit[7] = outer_l3_type
+	 */
+	/* L3 */
+	(*p)[0x01] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_NONFRAG;
+	(*p)[0x02] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_NONFRAG;
+	/* Fragmented */
+	(*p)[0x21] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_FRAG;
+	(*p)[0x22] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_FRAG;
+	/* TCP */
+	(*p)[0x05] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	(*p)[0x06] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	/* UDP */
+	(*p)[0x09] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	(*p)[0x0a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	/* Repeat with outer_l3_type being set. Just in case. */
+	(*p)[0x81] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_NONFRAG;
+	(*p)[0x82] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_NONFRAG;
+	(*p)[0xa1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_FRAG;
+	(*p)[0xa2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_FRAG;
+	(*p)[0x85] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	(*p)[0x86] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	(*p)[0x89] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	(*p)[0x8a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	/* Tunneled - L3 */
+	(*p)[0x41] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_NONFRAG;
+	(*p)[0x42] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_NONFRAG;
+	(*p)[0xc1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_NONFRAG;
+	(*p)[0xc2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_NONFRAG;
+	/* Tunneled - Fragmented */
+	(*p)[0x61] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_FRAG;
+	(*p)[0x62] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_FRAG;
+	(*p)[0xe1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_FRAG;
+	(*p)[0xe2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L4_FRAG;
+	/* Tunneled - TCP */
+	(*p)[0x45] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	(*p)[0x46] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	(*p)[0xc5] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	(*p)[0xc6] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_TCP;
+	/* Tunneled - UDP */
+	(*p)[0x49] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	(*p)[0x4a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	(*p)[0xc9] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+	(*p)[0xca] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN |
+		     RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN |
+		     RTE_PTYPE_L4_UDP;
+}
+
+/**
  * Return the size of tailroom of WQ.
  *
  * @param txq
@@ -1501,30 +1593,20 @@ mlx5_tx_burst_empw(void *dpdk_txq, struct rte_mbuf **pkts, uint16_t pkts_n)
 static inline uint32_t
 rxq_cq_to_pkt_type(volatile struct mlx5_cqe *cqe)
 {
-	uint32_t pkt_type;
-	uint16_t flags = ntohs(cqe->hdr_type_etc);
+	uint8_t idx;
+	uint8_t pinfo = cqe->pkt_info;
+	uint16_t ptype = cqe->hdr_type_etc;
 
-	if (cqe->pkt_info & MLX5_CQE_RX_TUNNEL_PACKET) {
-		pkt_type =
-			TRANSPOSE(flags,
-				  MLX5_CQE_RX_IPV4_PACKET,
-				  RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN) |
-			TRANSPOSE(flags,
-				  MLX5_CQE_RX_IPV6_PACKET,
-				  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN);
-		pkt_type |= ((cqe->pkt_info & MLX5_CQE_RX_OUTER_PACKET) ?
-			     RTE_PTYPE_L3_IPV6_EXT_UNKNOWN :
-			     RTE_PTYPE_L3_IPV4_EXT_UNKNOWN);
-	} else {
-		pkt_type =
-			TRANSPOSE(flags,
-				  MLX5_CQE_L3_HDR_TYPE_IPV6,
-				  RTE_PTYPE_L3_IPV6_EXT_UNKNOWN) |
-			TRANSPOSE(flags,
-				  MLX5_CQE_L3_HDR_TYPE_IPV4,
-				  RTE_PTYPE_L3_IPV4_EXT_UNKNOWN);
-	}
-	return pkt_type;
+	/*
+	 * The index to the array should have:
+	 * bit[1:0] = l3_hdr_type
+	 * bit[4:2] = l4_hdr_type
+	 * bit[5] = ip_frag
+	 * bit[6] = tunneled
+	 * bit[7] = outer_l3_type
+	 */
+	idx = ((pinfo & 0x3) << 6) | ((ptype & 0xfc00) >> 10);
+	return mlx5_ptype_table[idx];
 }
 
 /**
diff --git a/drivers/net/mlx5/mlx5_rxtx.h b/drivers/net/mlx5/mlx5_rxtx.h
index 534aaeb46..09882e173 100644
--- a/drivers/net/mlx5/mlx5_rxtx.h
+++ b/drivers/net/mlx5/mlx5_rxtx.h
@@ -334,8 +334,9 @@ uint16_t mlx5_tx_burst_secondary_setup(void *, struct rte_mbuf **, uint16_t);
 
 /* mlx5_rxtx.c */
 
-extern const uint32_t mlx5_ptype_table[];
+extern uint32_t mlx5_ptype_table[];
 
+void mlx5_set_ptype_table(void);
 uint16_t mlx5_tx_burst(void *, struct rte_mbuf **, uint16_t);
 uint16_t mlx5_tx_burst_mpw(void *, struct rte_mbuf **, uint16_t);
 uint16_t mlx5_tx_burst_mpw_inline(void *, struct rte_mbuf **, uint16_t);
diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_sse.c b/drivers/net/mlx5/mlx5_rxtx_vec_sse.c
index 95e96baa2..d43ab51d0 100644
--- a/drivers/net/mlx5/mlx5_rxtx_vec_sse.c
+++ b/drivers/net/mlx5/mlx5_rxtx_vec_sse.c
@@ -756,7 +756,7 @@ rxq_cq_to_ptype_oflags_v(struct rxq *rxq, __m128i cqes[4], __m128i op_err,
 	__m128i cv_flags;
 	const __m128i zero = _mm_setzero_si128();
 	const __m128i ptype_mask =
-		_mm_set_epi32(0xd06, 0xd06, 0xd06, 0xd06);
+		_mm_set_epi32(0xfd06, 0xfd06, 0xfd06, 0xfd06);
 	const __m128i ptype_ol_mask =
 		_mm_set_epi32(0x106, 0x106, 0x106, 0x106);
 	const __m128i pinfo_mask =
@@ -817,18 +817,23 @@ rxq_cq_to_ptype_oflags_v(struct rxq *rxq, __m128i cqes[4], __m128i op_err,
 	}
 	/*
 	 * Merge the two fields to generate the following:
-	 * bit[1]  = l3_ok,    bit[2]     = l4_ok
-	 * bit[8]  = cv,       bit[11:10] = l3_hdr_type
-	 * bit[12] = tunneled, bit[13]    = outer_l3_type
+	 * bit[1]     = l3_ok
+	 * bit[2]     = l4_ok
+	 * bit[8]     = cv
+	 * bit[11:10] = l3_hdr_type
+	 * bit[14:12] = l4_hdr_type
+	 * bit[15]    = ip_frag
+	 * bit[16]    = tunneled
+	 * bit[17]    = outer_l3_type
 	 */
 	ptype = _mm_and_si128(ptype, ptype_mask);
 	pinfo = _mm_and_si128(pinfo, pinfo_mask);
-	pinfo = _mm_slli_epi32(pinfo, 12);
+	pinfo = _mm_slli_epi32(pinfo, 16);
 	ptype = _mm_or_si128(ptype, pinfo);
 	ptype = _mm_srli_epi32(ptype, 10);
 	ptype = _mm_packs_epi32(ptype, zero);
 	/* Errored packets will have RTE_PTYPE_ALL_MASK. */
-	op_err = _mm_srli_epi16(op_err, 12);
+	op_err = _mm_srli_epi16(op_err, 8);
 	ptype = _mm_or_si128(ptype, op_err);
 	pkts[0]->packet_type = mlx5_ptype_table[_mm_extract_epi8(ptype, 0)];
 	pkts[1]->packet_type = mlx5_ptype_table[_mm_extract_epi8(ptype, 2)];
-- 
2.11.0

^ permalink raw reply related	[flat|nested] 6+ messages in thread
* Re: [PATCH 1/2] net/mlx5: fix missing packet type calculation
  2017-07-26 19:29 ` [PATCH 1/2] net/mlx5: fix missing packet type calculation Yongseok Koh
@ 2017-07-30  9:08   ` Sagi Grimberg
  0 siblings, 0 replies; 6+ messages in thread
From: Sagi Grimberg @ 2017-07-30  9:08 UTC (permalink / raw)
  To: Yongseok Koh, adrien.mazarguil, nelio.laranjeiro; +Cc: dev

Looks good,

Reviewed-by: Sagi Grimberg <sagi@grimberg.me>

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH 2/2] net/mlx5: fix L4 packet type support
  2017-07-26 19:29 ` [PATCH 2/2] net/mlx5: fix L4 packet type support Yongseok Koh
@ 2017-07-30  9:08   ` Sagi Grimberg
  0 siblings, 0 replies; 6+ messages in thread
From: Sagi Grimberg @ 2017-07-30  9:08 UTC (permalink / raw)
  To: Yongseok Koh, adrien.mazarguil, nelio.laranjeiro; +Cc: dev

Looks good,

Reviewed-by: Sagi Grimberg <sagi@grimberg.me>

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH 0/2] net/mlx5: fix packet type support
  2017-07-26 19:29 [PATCH 0/2] net/mlx5: fix packet type support Yongseok Koh
  2017-07-26 19:29 ` [PATCH 1/2] net/mlx5: fix missing packet type calculation Yongseok Koh
  2017-07-26 19:29 ` [PATCH 2/2] net/mlx5: fix L4 packet type support Yongseok Koh
@ 2017-07-31 16:29 ` Ferruh Yigit
  2 siblings, 0 replies; 6+ messages in thread
From: Ferruh Yigit @ 2017-07-31 16:29 UTC (permalink / raw)
  To: Yongseok Koh, adrien.mazarguil, nelio.laranjeiro; +Cc: dev

On 7/26/2017 8:29 PM, Yongseok Koh wrote:
> Even though HW supports more packet types, PMD doesn't parse it. Changed to have
> more packet_type flags in mbuf.
> 
> Yongseok Koh (2):
>   net/mlx5: fix missing packet type calculation
>   net/mlx5: fix L4 packet type support

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

^ permalink raw reply	[flat|nested] 6+ messages in thread
end of thread, other threads:[~2017-07-31 16:29 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-07-26 19:29 [PATCH 0/2] net/mlx5: fix packet type support Yongseok Koh
2017-07-26 19:29 ` [PATCH 1/2] net/mlx5: fix missing packet type calculation Yongseok Koh
2017-07-30  9:08   ` Sagi Grimberg
2017-07-26 19:29 ` [PATCH 2/2] net/mlx5: fix L4 packet type support Yongseok Koh
2017-07-30  9:08   ` Sagi Grimberg
2017-07-31 16:29 ` [PATCH 0/2] net/mlx5: fix " Ferruh Yigit

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.