From patchwork Thu May 11 09:25:07 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jianbo Liu X-Patchwork-Id: 99608 Delivered-To: patch@linaro.org Received: by 10.140.96.100 with SMTP id j91csp636605qge; Thu, 11 May 2017 02:26:33 -0700 (PDT) X-Received: by 10.28.144.15 with SMTP id s15mr34652wmd.137.1494494793649; Thu, 11 May 2017 02:26:33 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1494494793; cv=none; d=google.com; s=arc-20160816; b=rAPotzgSLo+sKArwkCXyTuC+yDiHyFAEoiEp1hBSITJ+BSzGeM5eNGBzdlmL89rELN CuWQfjzOOxYwar3FOjkuGasQyL06yRlJJ8lJq63WxE0cRSKgj/28lS0mldBEHYqE6YbO shndIOUhB70NtXqNkIbnUTIyRyDU3trmRrwZR9TqHt3XY4118aADzObhMJYS37TDT/wx ZRn6ETNQun92W/+pMWXNZclm24aGucbXph4pKCxH8niWuc0Kd9HNZdVSodWpv8I9huap 9C9RYH8uq30eiVIAEQwWoQhZm+ux2H8j9mJN9pJwGL2/Fij2guwPKtp62LiX6P+8QhBg kAAA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:list-subscribe:list-help:list-post:list-archive :list-unsubscribe:list-id:precedence:subject:references:in-reply-to :message-id:date:cc:to:from:arc-authentication-results; bh=EgWCcJUJKZO35IDyqkincffWpU424dyT5KAPIiAO7zY=; b=IhAx14uGtFjxSITWU+DVzBIFTUzRlCYs7Fl6j1xUYpgIjZi3ho+F9HUkLsTARoWRYw JAi9B+JaM9u4Wa4Z6wi9D0Visir71GG48pe5WMim99SChCYArAdHHmkEwuKMtdkCEiVz 1r+wso4AyAH+jsIfD1G758XtFS+QHEOd037zoIs3bMoWio4DZONwWvOSLzpK2UBwR64d tSgNrvwvIAQd3Ad3Yhp56KWb6brjGJcA2fCnUXaJGOQz+EsvVFPpbio7T/PKWkVPxT8a gEdvezzC3gqtUfMvyZsdhJdxM1qbK9uK9u9MJovMOzA0TjnoS89izjDQ6q/VqehjmnZO P8vQ== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: domain of dev-bounces@dpdk.org designates 2001:4b98:dc0:41:216:3eff:fe72:dd13 as permitted sender) smtp.mailfrom=dev-bounces@dpdk.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=linaro.org Return-Path: Received: from dpdk.org ([2001:4b98:dc0:41:216:3eff:fe72:dd13]) by mx.google.com with ESMTP id x82si35407wma.75.2017.05.11.02.26.33; Thu, 11 May 2017 02:26:33 -0700 (PDT) Received-SPF: pass (google.com: domain of dev-bounces@dpdk.org designates 2001:4b98:dc0:41:216:3eff:fe72:dd13 as permitted sender) client-ip=2001:4b98:dc0:41:216:3eff:fe72:dd13; Authentication-Results: mx.google.com; spf=pass (google.com: domain of dev-bounces@dpdk.org designates 2001:4b98:dc0:41:216:3eff:fe72:dd13 as permitted sender) smtp.mailfrom=dev-bounces@dpdk.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=linaro.org Received: from [92.243.14.124] (localhost [IPv6:::1]) by dpdk.org (Postfix) with ESMTP id B4DB67C98; Thu, 11 May 2017 11:25:40 +0200 (CEST) Received: from foss.arm.com (foss.arm.com [217.140.101.70]) by dpdk.org (Postfix) with ESMTP id 0E531567F for ; Thu, 11 May 2017 11:25:27 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.72.51.249]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 7DEFA1516; Thu, 11 May 2017 02:25:26 -0700 (PDT) Received: from localhost.localdomain.com (usa-sjc-imap-foss1.foss.arm.com [10.72.51.249]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id AA3533F220; Thu, 11 May 2017 02:25:25 -0700 (PDT) From: Jianbo Liu To: dev@dpdk.org, tomasz.kantecki@intel.com, jerin.jacob@caviumnetworks.com, ashwin.sekhar@caviumnetworks.com Cc: Jianbo Liu Date: Thu, 11 May 2017 17:25:07 +0800 Message-Id: <1494494708-20642-7-git-send-email-jianbo.liu@linaro.org> X-Mailer: git-send-email 1.8.3.1 In-Reply-To: <1494494708-20642-1-git-send-email-jianbo.liu@linaro.org> References: <1493709255-8887-1-git-send-email-jianbo.liu@linaro.org> <1494494708-20642-1-git-send-email-jianbo.liu@linaro.org> Subject: [dpdk-dev] [PATCH v3 6/7] examples/l3fwd: add the times of hash multi-lookup for different Archs 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: , Errors-To: dev-bounces@dpdk.org Sender: "dev" New macro to define how many times of hash lookup in one time, and this makes the code more concise. Signed-off-by: Jianbo Liu --- examples/l3fwd/l3fwd_em_hlm.h | 241 +++++++++++++----------------------------- 1 file changed, 71 insertions(+), 170 deletions(-) -- 1.8.3.1 diff --git a/examples/l3fwd/l3fwd_em_hlm.h b/examples/l3fwd/l3fwd_em_hlm.h index b9163e3..098b396 100644 --- a/examples/l3fwd/l3fwd_em_hlm.h +++ b/examples/l3fwd/l3fwd_em_hlm.h @@ -43,148 +43,65 @@ #include "l3fwd_em_hlm_neon.h" #endif +#ifdef RTE_ARCH_ARM64 +#define EM_HASH_LOOKUP_COUNT 16 +#else +#define EM_HASH_LOOKUP_COUNT 8 +#endif + + static inline __attribute__((always_inline)) void -em_get_dst_port_ipv4x8(struct lcore_conf *qconf, struct rte_mbuf *m[8], - uint8_t portid, uint16_t dst_port[8]) +em_get_dst_port_ipv4xN(struct lcore_conf *qconf, struct rte_mbuf *m[], + uint8_t portid, uint16_t dst_port[]) { - int32_t ret[8]; - union ipv4_5tuple_host key[8]; - - get_ipv4_5tuple(m[0], mask0.x, &key[0]); - get_ipv4_5tuple(m[1], mask0.x, &key[1]); - get_ipv4_5tuple(m[2], mask0.x, &key[2]); - get_ipv4_5tuple(m[3], mask0.x, &key[3]); - get_ipv4_5tuple(m[4], mask0.x, &key[4]); - get_ipv4_5tuple(m[5], mask0.x, &key[5]); - get_ipv4_5tuple(m[6], mask0.x, &key[6]); - get_ipv4_5tuple(m[7], mask0.x, &key[7]); - - const void *key_array[8] = {&key[0], &key[1], &key[2], &key[3], - &key[4], &key[5], &key[6], &key[7]}; - - rte_hash_lookup_bulk(qconf->ipv4_lookup_struct, &key_array[0], 8, ret); - - dst_port[0] = (uint8_t) ((ret[0] < 0) ? - portid : ipv4_l3fwd_out_if[ret[0]]); - dst_port[1] = (uint8_t) ((ret[1] < 0) ? - portid : ipv4_l3fwd_out_if[ret[1]]); - dst_port[2] = (uint8_t) ((ret[2] < 0) ? - portid : ipv4_l3fwd_out_if[ret[2]]); - dst_port[3] = (uint8_t) ((ret[3] < 0) ? - portid : ipv4_l3fwd_out_if[ret[3]]); - dst_port[4] = (uint8_t) ((ret[4] < 0) ? - portid : ipv4_l3fwd_out_if[ret[4]]); - dst_port[5] = (uint8_t) ((ret[5] < 0) ? - portid : ipv4_l3fwd_out_if[ret[5]]); - dst_port[6] = (uint8_t) ((ret[6] < 0) ? - portid : ipv4_l3fwd_out_if[ret[6]]); - dst_port[7] = (uint8_t) ((ret[7] < 0) ? - portid : ipv4_l3fwd_out_if[ret[7]]); - - if (dst_port[0] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[0]) == 0) - dst_port[0] = portid; - - if (dst_port[1] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[1]) == 0) - dst_port[1] = portid; - - if (dst_port[2] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[2]) == 0) - dst_port[2] = portid; - - if (dst_port[3] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[3]) == 0) - dst_port[3] = portid; - - if (dst_port[4] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[4]) == 0) - dst_port[4] = portid; - - if (dst_port[5] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[5]) == 0) - dst_port[5] = portid; - - if (dst_port[6] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[6]) == 0) - dst_port[6] = portid; - - if (dst_port[7] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[7]) == 0) - dst_port[7] = portid; + int i; + int32_t ret[EM_HASH_LOOKUP_COUNT]; + union ipv4_5tuple_host key[EM_HASH_LOOKUP_COUNT]; + const void *key_array[EM_HASH_LOOKUP_COUNT]; + + for (i = 0; i < EM_HASH_LOOKUP_COUNT; i++) { + get_ipv4_5tuple(m[i], mask0.x, &key[i]); + key_array[i] = &key[i]; + } + + rte_hash_lookup_bulk(qconf->ipv4_lookup_struct, &key_array[0], + EM_HASH_LOOKUP_COUNT, ret); + + for (i = 0; i < EM_HASH_LOOKUP_COUNT; i++) { + dst_port[i] = (uint8_t) ((ret[i] < 0) ? + portid : ipv4_l3fwd_out_if[ret[i]]); + if (dst_port[i] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[i]) == 0) + dst_port[i] = portid; + } } static inline __attribute__((always_inline)) void -em_get_dst_port_ipv6x8(struct lcore_conf *qconf, struct rte_mbuf *m[8], - uint8_t portid, uint16_t dst_port[8]) +em_get_dst_port_ipv6xN(struct lcore_conf *qconf, struct rte_mbuf *m[], + uint8_t portid, uint16_t dst_port[]) { - int32_t ret[8]; - union ipv6_5tuple_host key[8]; - - get_ipv6_5tuple(m[0], mask1.x, mask2.x, &key[0]); - get_ipv6_5tuple(m[1], mask1.x, mask2.x, &key[1]); - get_ipv6_5tuple(m[2], mask1.x, mask2.x, &key[2]); - get_ipv6_5tuple(m[3], mask1.x, mask2.x, &key[3]); - get_ipv6_5tuple(m[4], mask1.x, mask2.x, &key[4]); - get_ipv6_5tuple(m[5], mask1.x, mask2.x, &key[5]); - get_ipv6_5tuple(m[6], mask1.x, mask2.x, &key[6]); - get_ipv6_5tuple(m[7], mask1.x, mask2.x, &key[7]); - - const void *key_array[8] = {&key[0], &key[1], &key[2], &key[3], - &key[4], &key[5], &key[6], &key[7]}; - - rte_hash_lookup_bulk(qconf->ipv6_lookup_struct, &key_array[0], 8, ret); - - dst_port[0] = (uint8_t) ((ret[0] < 0) ? - portid : ipv6_l3fwd_out_if[ret[0]]); - dst_port[1] = (uint8_t) ((ret[1] < 0) ? - portid : ipv6_l3fwd_out_if[ret[1]]); - dst_port[2] = (uint8_t) ((ret[2] < 0) ? - portid : ipv6_l3fwd_out_if[ret[2]]); - dst_port[3] = (uint8_t) ((ret[3] < 0) ? - portid : ipv6_l3fwd_out_if[ret[3]]); - dst_port[4] = (uint8_t) ((ret[4] < 0) ? - portid : ipv6_l3fwd_out_if[ret[4]]); - dst_port[5] = (uint8_t) ((ret[5] < 0) ? - portid : ipv6_l3fwd_out_if[ret[5]]); - dst_port[6] = (uint8_t) ((ret[6] < 0) ? - portid : ipv6_l3fwd_out_if[ret[6]]); - dst_port[7] = (uint8_t) ((ret[7] < 0) ? - portid : ipv6_l3fwd_out_if[ret[7]]); - - if (dst_port[0] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[0]) == 0) - dst_port[0] = portid; - - if (dst_port[1] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[1]) == 0) - dst_port[1] = portid; - - if (dst_port[2] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[2]) == 0) - dst_port[2] = portid; - - if (dst_port[3] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[3]) == 0) - dst_port[3] = portid; - - if (dst_port[4] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[4]) == 0) - dst_port[4] = portid; - - if (dst_port[5] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[5]) == 0) - dst_port[5] = portid; - - if (dst_port[6] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[6]) == 0) - dst_port[6] = portid; - - if (dst_port[7] >= RTE_MAX_ETHPORTS || - (enabled_port_mask & 1 << dst_port[7]) == 0) - dst_port[7] = portid; + int i; + int32_t ret[EM_HASH_LOOKUP_COUNT]; + union ipv6_5tuple_host key[EM_HASH_LOOKUP_COUNT]; + const void *key_array[EM_HASH_LOOKUP_COUNT]; + + for (i = 0; i < EM_HASH_LOOKUP_COUNT; i++) { + get_ipv6_5tuple(m[i], mask1.x, mask2.x, &key[i]); + key_array[i] = &key[i]; + } + + rte_hash_lookup_bulk(qconf->ipv6_lookup_struct, &key_array[0], + EM_HASH_LOOKUP_COUNT, ret); + for (i = 0; i < EM_HASH_LOOKUP_COUNT; i++) { + dst_port[i] = (uint8_t) ((ret[i] < 0) ? + portid : ipv6_l3fwd_out_if[ret[i]]); + + if (dst_port[i] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[i]) == 0) + dst_port[i] = portid; + } } static inline __attribute__((always_inline)) uint16_t @@ -247,64 +164,48 @@ static inline __attribute__((always_inline)) uint16_t uint16_t dst_port[MAX_PKT_BURST]; /* - * Send nb_rx - nb_rx%8 packets - * in groups of 8. + * Send nb_rx - nb_rx % EM_HASH_LOOKUP_COUNT packets + * in groups of EM_HASH_LOOKUP_COUNT. */ - int32_t n = RTE_ALIGN_FLOOR(nb_rx, 8); + int32_t n = RTE_ALIGN_FLOOR(nb_rx, EM_HASH_LOOKUP_COUNT); - for (j = 0; j < 8 && j < nb_rx; j++) { + for (j = 0; j < EM_HASH_LOOKUP_COUNT && j < nb_rx; j++) { rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j], struct ether_hdr *) + 1); } - for (j = 0; j < n; j += 8) { + for (j = 0; j < n; j += EM_HASH_LOOKUP_COUNT) { + + uint32_t pkt_type = RTE_PTYPE_L3_MASK | + RTE_PTYPE_L4_TCP | RTE_PTYPE_L4_UDP; + uint32_t l3_type, tcp_or_udp; - uint32_t pkt_type = - pkts_burst[j]->packet_type & - pkts_burst[j+1]->packet_type & - pkts_burst[j+2]->packet_type & - pkts_burst[j+3]->packet_type & - pkts_burst[j+4]->packet_type & - pkts_burst[j+5]->packet_type & - pkts_burst[j+6]->packet_type & - pkts_burst[j+7]->packet_type; + for (i = 0; i < EM_HASH_LOOKUP_COUNT; i++) + pkt_type &= pkts_burst[j + i]->packet_type; - uint32_t l3_type = pkt_type & RTE_PTYPE_L3_MASK; - uint32_t tcp_or_udp = pkt_type & - (RTE_PTYPE_L4_TCP | RTE_PTYPE_L4_UDP); + l3_type = pkt_type & RTE_PTYPE_L3_MASK; + tcp_or_udp = pkt_type & (RTE_PTYPE_L4_TCP | RTE_PTYPE_L4_UDP); - for (i = 0, pos = j + 8; i < 8 && pos < nb_rx; i++, pos++) { + for (i = 0, pos = j + EM_HASH_LOOKUP_COUNT; + i < EM_HASH_LOOKUP_COUNT && pos < nb_rx; i++, pos++) { rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[pos], struct ether_hdr *) + 1); } if (tcp_or_udp && (l3_type == RTE_PTYPE_L3_IPV4)) { - em_get_dst_port_ipv4x8(qconf, &pkts_burst[j], portid, + em_get_dst_port_ipv4xN(qconf, &pkts_burst[j], portid, &dst_port[j]); } else if (tcp_or_udp && (l3_type == RTE_PTYPE_L3_IPV6)) { - em_get_dst_port_ipv6x8(qconf, &pkts_burst[j], portid, + em_get_dst_port_ipv6xN(qconf, &pkts_burst[j], portid, &dst_port[j]); } else { - dst_port[j] = em_get_dst_port(qconf, pkts_burst[j], - portid); - dst_port[j+1] = em_get_dst_port(qconf, pkts_burst[j+1], - portid); - dst_port[j+2] = em_get_dst_port(qconf, pkts_burst[j+2], - portid); - dst_port[j+3] = em_get_dst_port(qconf, pkts_burst[j+3], - portid); - dst_port[j+4] = em_get_dst_port(qconf, pkts_burst[j+4], - portid); - dst_port[j+5] = em_get_dst_port(qconf, pkts_burst[j+5], - portid); - dst_port[j+6] = em_get_dst_port(qconf, pkts_burst[j+6], - portid); - dst_port[j+7] = em_get_dst_port(qconf, pkts_burst[j+7], - portid); + for (i = 0; i < EM_HASH_LOOKUP_COUNT; i++) + dst_port[j + i] = em_get_dst_port(qconf, + pkts_burst[j + i], portid); } }