From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by dpdk.org (Postfix) with ESMTP id 48C1D2E8A for ; Tue, 4 Apr 2017 03:04:44 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=intel.com; i=@intel.com; q=dns/txt; s=intel; t=1491267884; x=1522803884; h=from:to:cc:subject:date:message-id:in-reply-to: references; bh=NHCBKGJCBiGTREFuMjS7OSbbf/LByeHCdCyfSXANMEY=; b=DEVxr2NvQL/c6GZkdB2Zr8H3p72Os1VAp3a+G8k2MU618VSrfqe3k3SQ bv49U9aXLiQ8SpGiv6yRUORRD1cvKg==; Received: from orsmga004.jf.intel.com ([10.7.209.38]) by orsmga102.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 03 Apr 2017 18:04:43 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.36,273,1486454400"; d="scan'208";a="73237327" Received: from unknown (HELO localhost.localdomain.sh.intel.com) ([10.239.129.229]) by orsmga004.jf.intel.com with ESMTP; 03 Apr 2017 18:04:42 -0700 From: Qi Zhang To: ferruh.yigit@intel.com Cc: jingjing.wu@intel.com, helin.zhang@intel.com, dev@dpdk.org, Qi Zhang Date: Mon, 3 Apr 2017 21:55:14 -0400 Message-Id: <20170404015516.7784-2-qi.z.zhang@intel.com> X-Mailer: git-send-email 2.9.3 In-Reply-To: <20170404015516.7784-1-qi.z.zhang@intel.com> References: <20170404015516.7784-1-qi.z.zhang@intel.com> Subject: [dpdk-dev] [PATCH v7 1/3] net/i40e: enable per dev PTYPE mapping table X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 04 Apr 2017 01:04:44 -0000 The mapping from hardware defined packet type to software defined packet type is static for i40e device, the patch let each ethdev to to have their own copy of mapping table, this give the possibility that different ethdev can be set different PTYPE mapping rule which is the requirement to support following hardware's dynamic PTYPE feature. Signed-off-by: Qi Zhang --- v7: - Add Altivec vPMD support. v5: - Rebase to dpdk-next-net drivers/net/i40e/i40e_ethdev.c | 1 + drivers/net/i40e/i40e_ethdev.h | 5 +++++ drivers/net/i40e/i40e_ethdev_vf.c | 2 +- drivers/net/i40e/i40e_rxtx.c | 30 ++++++++++++++++++++++-------- drivers/net/i40e/i40e_rxtx.h | 3 ++- drivers/net/i40e/i40e_rxtx_vec_altivec.c | 18 ++++++++---------- drivers/net/i40e/i40e_rxtx_vec_neon.c | 8 +++++--- drivers/net/i40e/i40e_rxtx_vec_sse.c | 14 ++++++++------ 8 files changed, 52 insertions(+), 29 deletions(-) diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c index 00079bb..d7b9f8b 100644 --- a/drivers/net/i40e/i40e_ethdev.c +++ b/drivers/net/i40e/i40e_ethdev.c @@ -1058,6 +1058,7 @@ eth_i40e_dev_init(struct rte_eth_dev *dev) i40e_set_tx_function(dev); return 0; } + i40e_set_default_ptype_table(dev); pci_dev = I40E_DEV_TO_PCI(dev); intr_handle = &pci_dev->intr_handle; diff --git a/drivers/net/i40e/i40e_ethdev.h b/drivers/net/i40e/i40e_ethdev.h index a8ecea4..38fcd94 100644 --- a/drivers/net/i40e/i40e_ethdev.h +++ b/drivers/net/i40e/i40e_ethdev.h @@ -740,6 +740,8 @@ struct i40e_vf { uint64_t flags; }; +#define I40E_MAX_PKT_TYPE 256 + /* * Structure to store private data for each PF/VF instance. */ @@ -764,6 +766,9 @@ struct i40e_adapter { struct rte_timecounter systime_tc; struct rte_timecounter rx_tstamp_tc; struct rte_timecounter tx_tstamp_tc; + + /* ptype mapping table */ + uint32_t ptype_tbl[I40E_MAX_PKT_TYPE] __rte_cache_min_aligned; }; extern const struct rte_flow_ops i40e_flow_ops; diff --git a/drivers/net/i40e/i40e_ethdev_vf.c b/drivers/net/i40e/i40e_ethdev_vf.c index d3659c9..da74824 100644 --- a/drivers/net/i40e/i40e_ethdev_vf.c +++ b/drivers/net/i40e/i40e_ethdev_vf.c @@ -1478,7 +1478,7 @@ i40evf_dev_init(struct rte_eth_dev *eth_dev) i40e_set_tx_function(eth_dev); return 0; } - + i40e_set_default_ptype_table(eth_dev); rte_eth_copy_pci_info(eth_dev, pci_dev); eth_dev->data->dev_flags |= RTE_ETH_DEV_DETACHABLE; diff --git a/drivers/net/i40e/i40e_rxtx.c b/drivers/net/i40e/i40e_rxtx.c index ff70c06..e5471b1 100644 --- a/drivers/net/i40e/i40e_rxtx.c +++ b/drivers/net/i40e/i40e_rxtx.c @@ -61,7 +61,6 @@ #define DEFAULT_TX_RS_THRESH 32 #define DEFAULT_TX_FREE_THRESH 32 -#define I40E_MAX_PKT_TYPE 256 #define I40E_TX_MAX_BURST 32 @@ -458,6 +457,7 @@ i40e_rx_scan_hw_ring(struct i40e_rx_queue *rxq) int32_t s[I40E_LOOK_AHEAD], nb_dd; int32_t i, j, nb_rx = 0; uint64_t pkt_flags; + uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; rxdp = &rxq->rx_ring[rxq->rx_tail]; rxep = &rxq->sw_ring[rxq->rx_tail]; @@ -506,9 +506,9 @@ i40e_rx_scan_hw_ring(struct i40e_rx_queue *rxq) pkt_flags = i40e_rxd_status_to_pkt_flags(qword1); pkt_flags |= i40e_rxd_error_to_pkt_flags(qword1); mb->packet_type = - i40e_rxd_pkt_type_mapping((uint8_t)((qword1 & - I40E_RXD_QW1_PTYPE_MASK) >> - I40E_RXD_QW1_PTYPE_SHIFT)); + ptype_tbl[(uint8_t)((qword1 & + I40E_RXD_QW1_PTYPE_MASK) >> + I40E_RXD_QW1_PTYPE_SHIFT)]; if (pkt_flags & PKT_RX_RSS_HASH) mb->hash.rss = rte_le_to_cpu_32(\ rxdp[j].wb.qword0.hi_dword.rss); @@ -700,6 +700,7 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) uint16_t rx_id, nb_hold; uint64_t dma_addr; uint64_t pkt_flags; + uint32_t *ptype_tbl; nb_rx = 0; nb_hold = 0; @@ -707,6 +708,7 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) rx_id = rxq->rx_tail; rx_ring = rxq->rx_ring; sw_ring = rxq->sw_ring; + ptype_tbl = rxq->vsi->adapter->ptype_tbl; while (nb_rx < nb_pkts) { rxdp = &rx_ring[rx_id]; @@ -763,8 +765,8 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) pkt_flags = i40e_rxd_status_to_pkt_flags(qword1); pkt_flags |= i40e_rxd_error_to_pkt_flags(qword1); rxm->packet_type = - i40e_rxd_pkt_type_mapping((uint8_t)((qword1 & - I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)); + ptype_tbl[(uint8_t)((qword1 & + I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)]; if (pkt_flags & PKT_RX_RSS_HASH) rxm->hash.rss = rte_le_to_cpu_32(rxd.wb.qword0.hi_dword.rss); @@ -818,6 +820,7 @@ i40e_recv_scattered_pkts(void *rx_queue, uint64_t qword1; uint64_t dma_addr; uint64_t pkt_flags; + uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; while (nb_rx < nb_pkts) { rxdp = &rx_ring[rx_id]; @@ -925,8 +928,8 @@ i40e_recv_scattered_pkts(void *rx_queue, pkt_flags = i40e_rxd_status_to_pkt_flags(qword1); pkt_flags |= i40e_rxd_error_to_pkt_flags(qword1); first_seg->packet_type = - i40e_rxd_pkt_type_mapping((uint8_t)((qword1 & - I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)); + ptype_tbl[(uint8_t)((qword1 & + I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)]; if (pkt_flags & PKT_RX_RSS_HASH) first_seg->hash.rss = rte_le_to_cpu_32(rxd.wb.qword0.hi_dword.rss); @@ -2922,6 +2925,17 @@ i40e_set_tx_function(struct rte_eth_dev *dev) } } +void __attribute__((cold)) +i40e_set_default_ptype_table(struct rte_eth_dev *dev) +{ + struct i40e_adapter *ad = + I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); + int i; + + for (i = 0; i <= I40E_MAX_PKT_TYPE; i++) + ad->ptype_tbl[i] = i40e_get_default_pkt_type(i); +} + /* Stubs needed for linkage when CONFIG_RTE_I40E_INC_VECTOR is set to 'n' */ int __attribute__((weak)) i40e_rx_vec_dev_conf_condition_check(struct rte_eth_dev __rte_unused *dev) diff --git a/drivers/net/i40e/i40e_rxtx.h b/drivers/net/i40e/i40e_rxtx.h index 2fa7f37..20084d6 100644 --- a/drivers/net/i40e/i40e_rxtx.h +++ b/drivers/net/i40e/i40e_rxtx.h @@ -254,13 +254,14 @@ void i40e_set_rx_function(struct rte_eth_dev *dev); void i40e_set_tx_function_flag(struct rte_eth_dev *dev, struct i40e_tx_queue *txq); void i40e_set_tx_function(struct rte_eth_dev *dev); +void i40e_set_default_ptype_table(struct rte_eth_dev *dev); /* For each value it means, datasheet of hardware can tell more details * * @note: fix i40e_dev_supported_ptypes_get() if any change here. */ static inline uint32_t -i40e_rxd_pkt_type_mapping(uint8_t ptype) +i40e_get_default_pkt_type(uint8_t ptype) { static const uint32_t type_table[UINT8_MAX + 1] __rte_cache_aligned = { /* L2 types */ diff --git a/drivers/net/i40e/i40e_rxtx_vec_altivec.c b/drivers/net/i40e/i40e_rxtx_vec_altivec.c index 2f6f70a..5f525ff 100644 --- a/drivers/net/i40e/i40e_rxtx_vec_altivec.c +++ b/drivers/net/i40e/i40e_rxtx_vec_altivec.c @@ -209,7 +209,8 @@ desc_to_olflags_v(vector unsigned long descs[4], struct rte_mbuf **rx_pkts) #define PKTLEN_SHIFT 10 static inline void -desc_to_ptype_v(vector unsigned long descs[4], struct rte_mbuf **rx_pkts) +desc_to_ptype_v(vector unsigned long descs[4], struct rte_mbuf **rx_pkts, + uint32_t *ptype_tbl) { vector unsigned long ptype0 = vec_mergel(descs[0], descs[1]); vector unsigned long ptype1 = vec_mergel(descs[2], descs[3]); @@ -217,14 +218,10 @@ desc_to_ptype_v(vector unsigned long descs[4], struct rte_mbuf **rx_pkts) ptype0 = vec_sr(ptype0, (vector unsigned long){30, 30}); ptype1 = vec_sr(ptype1, (vector unsigned long){30, 30}); - rx_pkts[0]->packet_type = i40e_rxd_pkt_type_mapping( - (*(vector unsigned char *)&ptype0)[0]); - rx_pkts[1]->packet_type = i40e_rxd_pkt_type_mapping( - (*(vector unsigned char *)&ptype0)[8]); - rx_pkts[2]->packet_type = i40e_rxd_pkt_type_mapping( - (*(vector unsigned char *)&ptype1)[0]); - rx_pkts[3]->packet_type = i40e_rxd_pkt_type_mapping( - (*(vector unsigned char *)&ptype1)[8]); + rx_pkts[0]->packet_type = ptype_tbl[(*(vector unsigned char *)&ptype0)[0])]; + rx_pkts[1]->packet_type = ptype_tbl[(*(vector unsigned char *)&ptype0)[8])]; + rx_pkts[2]->packet_type = ptype_tbl[(*(vector unsigned char *)&ptype1)[0])]; + rx_pkts[3]->packet_type = ptype_tbl[(*(vector unsigned char *)&ptype1)[8])]; } /* Notice: @@ -242,6 +239,7 @@ _recv_raw_pkts_vec(struct i40e_rx_queue *rxq, struct rte_mbuf **rx_pkts, int pos; uint64_t var; vector unsigned char shuf_msk; + uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; vector unsigned short crc_adjust = (vector unsigned short){ 0, 0, /* ignore pkt_type field */ @@ -466,7 +464,7 @@ _recv_raw_pkts_vec(struct i40e_rx_queue *rxq, struct rte_mbuf **rx_pkts, vec_st(pkt_mb1, 0, (vector unsigned char *)&rx_pkts[pos]->rx_descriptor_fields1 ); - desc_to_ptype_v(descs, &rx_pkts[pos]); + desc_to_ptype_v(descs, &rx_pkts[pos], ptype_tbl); desc_to_olflags_v(descs, &rx_pkts[pos]); /* C.4 calc avaialbe number of desc */ diff --git a/drivers/net/i40e/i40e_rxtx_vec_neon.c b/drivers/net/i40e/i40e_rxtx_vec_neon.c index bd7239b..00be542 100644 --- a/drivers/net/i40e/i40e_rxtx_vec_neon.c +++ b/drivers/net/i40e/i40e_rxtx_vec_neon.c @@ -196,7 +196,8 @@ desc_to_olflags_v(uint64x2_t descs[4], struct rte_mbuf **rx_pkts) #define I40E_VPMD_DESC_DD_MASK 0x0001000100010001ULL static inline void -desc_to_ptype_v(uint64x2_t descs[4], struct rte_mbuf **rx_pkts) +desc_to_ptype_v(uint64x2_t descs[4], struct rte_mbuf **rx_pkts, + uint32_t *ptype_tbl) { int i; uint8_t ptype; @@ -205,7 +206,7 @@ desc_to_ptype_v(uint64x2_t descs[4], struct rte_mbuf **rx_pkts) for (i = 0; i < 4; i++) { tmp = vreinterpretq_u8_u64(vshrq_n_u64(descs[i], 30)); ptype = vgetq_lane_u8(tmp, 8); - rx_pkts[i]->packet_type = i40e_rxd_pkt_type_mapping(ptype); + rx_pkts[i]->packet_type = ptype_tbl[ptype]; } } @@ -225,6 +226,7 @@ _recv_raw_pkts_vec(struct i40e_rx_queue *rxq, struct rte_mbuf **rx_pkts, uint16_t nb_pkts_recd; int pos; uint64_t var; + uint32 *ptype_tbl = rxq->vsi->adapter->ptype_tbl; /* mask to shuffle from desc. to mbuf */ uint8x16_t shuf_msk = { @@ -429,7 +431,7 @@ _recv_raw_pkts_vec(struct i40e_rx_queue *rxq, struct rte_mbuf **rx_pkts, pkt_mb2); vst1q_u8((void *)&rx_pkts[pos]->rx_descriptor_fields1, pkt_mb1); - desc_to_ptype_v(descs, &rx_pkts[pos]); + desc_to_ptype_v(descs, &rx_pkts[pos], ptype_tbl); /* C.4 calc avaialbe number of desc */ var = __builtin_popcountll(stat & I40E_VPMD_DESC_DD_MASK); nb_pkts_recd += var; diff --git a/drivers/net/i40e/i40e_rxtx_vec_sse.c b/drivers/net/i40e/i40e_rxtx_vec_sse.c index deed7e2..c5d7083 100644 --- a/drivers/net/i40e/i40e_rxtx_vec_sse.c +++ b/drivers/net/i40e/i40e_rxtx_vec_sse.c @@ -218,7 +218,8 @@ desc_to_olflags_v(__m128i descs[4], struct rte_mbuf **rx_pkts) #define PKTLEN_SHIFT 10 static inline void -desc_to_ptype_v(__m128i descs[4], struct rte_mbuf **rx_pkts) +desc_to_ptype_v(__m128i descs[4], struct rte_mbuf **rx_pkts, + uint32_t *ptype_tbl) { __m128i ptype0 = _mm_unpackhi_epi64(descs[0], descs[1]); __m128i ptype1 = _mm_unpackhi_epi64(descs[2], descs[3]); @@ -226,10 +227,10 @@ desc_to_ptype_v(__m128i descs[4], struct rte_mbuf **rx_pkts) ptype0 = _mm_srli_epi64(ptype0, 30); ptype1 = _mm_srli_epi64(ptype1, 30); - rx_pkts[0]->packet_type = i40e_rxd_pkt_type_mapping(_mm_extract_epi8(ptype0, 0)); - rx_pkts[1]->packet_type = i40e_rxd_pkt_type_mapping(_mm_extract_epi8(ptype0, 8)); - rx_pkts[2]->packet_type = i40e_rxd_pkt_type_mapping(_mm_extract_epi8(ptype1, 0)); - rx_pkts[3]->packet_type = i40e_rxd_pkt_type_mapping(_mm_extract_epi8(ptype1, 8)); + rx_pkts[0]->packet_type = ptype_tbl[_mm_extract_epi8(ptype0, 0)]; + rx_pkts[1]->packet_type = ptype_tbl[_mm_extract_epi8(ptype0, 8)]; + rx_pkts[2]->packet_type = ptype_tbl[_mm_extract_epi8(ptype1, 0)]; + rx_pkts[3]->packet_type = ptype_tbl[_mm_extract_epi8(ptype1, 8)]; } /* @@ -248,6 +249,7 @@ _recv_raw_pkts_vec(struct i40e_rx_queue *rxq, struct rte_mbuf **rx_pkts, int pos; uint64_t var; __m128i shuf_msk; + uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; __m128i crc_adjust = _mm_set_epi16( 0, 0, 0, /* ignore non-length fields */ @@ -441,7 +443,7 @@ _recv_raw_pkts_vec(struct i40e_rx_queue *rxq, struct rte_mbuf **rx_pkts, pkt_mb2); _mm_storeu_si128((void *)&rx_pkts[pos]->rx_descriptor_fields1, pkt_mb1); - desc_to_ptype_v(descs, &rx_pkts[pos]); + desc_to_ptype_v(descs, &rx_pkts[pos], ptype_tbl); /* C.4 calc avaialbe number of desc */ var = __builtin_popcountll(_mm_cvtsi128_si64(staterr)); nb_pkts_recd += var; -- 2.9.3