From patchwork Thu Jan 6 16:03:32 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Li, Xiaoyun" X-Patchwork-Id: 105663 X-Patchwork-Delegate: ferruh.yigit@amd.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id E6172A0351; Thu, 6 Jan 2022 17:13:48 +0100 (CET) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 476C54115A; Thu, 6 Jan 2022 17:13:44 +0100 (CET) Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by mails.dpdk.org (Postfix) with ESMTP id 5BB0841152 for ; Thu, 6 Jan 2022 17:13:42 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1641485622; x=1673021622; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=doBtOVZkZLMPhEOaWY1V3sLXPyRrvJO1NKMdX2LdUI8=; b=bvAz8pPou8NYx5P5ysbpNvyDqUoSMhATDqDb+nXm0HcxAGJjuZZOyncr Jq6yBUWJ+eqH2NZyGGLNyAP4iGLMKSaZSRSV5Rjncq7gVcAWcEhov4EF/ HIRurCh6Ysmp6XgTX+bn3s6FuRWk32CkYgLz0iiAK9cT+jALAi6OvjCr4 Y9k7AjJCK0eZ3XOT6EeMSoTEB/al6FzWd5VBGl5hkk7pmYdj3KgZK55vY L8HYM/aN1Q1DM50jCON1wYvYhklo0g0/JBf+xKZPM9pAVvjpuNPK4LMVl Jb7z5sGMITCp+hS4rfWP5shq8q7pFODfcn66z8+eSd1X2KkI13ItCZns9 g==; X-IronPort-AV: E=McAfee;i="6200,9189,10217"; a="241500759" X-IronPort-AV: E=Sophos;i="5.88,267,1635231600"; d="scan'208";a="241500759" Received: from orsmga007.jf.intel.com ([10.7.209.58]) by fmsmga104.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 06 Jan 2022 08:13:27 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.88,267,1635231600"; d="scan'208";a="513457505" Received: from dpdk-xiaoyun3.sh.intel.com ([10.67.119.214]) by orsmga007.jf.intel.com with ESMTP; 06 Jan 2022 08:13:24 -0800 From: Xiaoyun Li To: Aman.Deep.Singh@intel.com, ferruh.yigit@intel.com, olivier.matz@6wind.com, mb@smartsharesystems.com, konstantin.ananyev@intel.com, stephen@networkplumber.org, vladimir.medvedkin@intel.com Cc: dev@dpdk.org, Xiaoyun Li , Aman Singh , Sunil Pai G Subject: [PATCH v5 1/2] net: add functions to calculate UDP/TCP cksum in mbuf Date: Fri, 7 Jan 2022 00:03:32 +0800 Message-Id: <20220106160333.762686-2-xiaoyun.li@intel.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220106160333.762686-1-xiaoyun.li@intel.com> References: <20211015051306.320328-1-xiaoyun.li@intel.com> <20220106160333.762686-1-xiaoyun.li@intel.com> MIME-Version: 1.0 X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Add functions to call rte_raw_cksum_mbuf() to calculate IPv4/6 UDP/TCP checksum in mbuf which can be over multi-segments. Signed-off-by: Xiaoyun Li Acked-by: Aman Singh Tested-by: Sunil Pai G Acked-by: Ferruh Yigit --- doc/guides/rel_notes/release_22_03.rst | 11 ++ lib/net/rte_ip.h | 186 +++++++++++++++++++++++++ 2 files changed, 197 insertions(+) diff --git a/doc/guides/rel_notes/release_22_03.rst b/doc/guides/rel_notes/release_22_03.rst index 6d99d1eaa9..785fd22001 100644 --- a/doc/guides/rel_notes/release_22_03.rst +++ b/doc/guides/rel_notes/release_22_03.rst @@ -55,6 +55,14 @@ New Features Also, make sure to start the actual text at the margin. ======================================================= +* **Added functions to calculate UDP/TCP checksum in mbuf.** + + * Added the following functions to calculate UDP/TCP checksum of packets + which can be over multi-segments: + - ``rte_ipv4_udptcp_cksum_mbuf()`` + - ``rte_ipv4_udptcp_cksum_mbuf_verify()`` + - ``rte_ipv6_udptcp_cksum_mbuf()`` + - ``rte_ipv6_udptcp_cksum_mbuf_verify()`` Removed Items ------------- @@ -84,6 +92,9 @@ API Changes Also, make sure to start the actual text at the margin. ======================================================= +* net: added experimental functions ``rte_ipv4_udptcp_cksum_mbuf()``, + ``rte_ipv4_udptcp_cksum_mbuf_verify()``, ``rte_ipv6_udptcp_cksum_mbuf()``, + ``rte_ipv6_udptcp_cksum_mbuf_verify()`` ABI Changes ----------- diff --git a/lib/net/rte_ip.h b/lib/net/rte_ip.h index c575250852..534f401d26 100644 --- a/lib/net/rte_ip.h +++ b/lib/net/rte_ip.h @@ -400,6 +400,65 @@ rte_ipv4_udptcp_cksum(const struct rte_ipv4_hdr *ipv4_hdr, const void *l4_hdr) return cksum; } +/** + * @internal Calculate the non-complemented IPv4 L4 checksum of a packet + */ +static inline uint16_t +__rte_ipv4_udptcp_cksum_mbuf(const struct rte_mbuf *m, + const struct rte_ipv4_hdr *ipv4_hdr, + uint16_t l4_off) +{ + uint16_t raw_cksum; + uint32_t cksum; + + if (l4_off > m->pkt_len) + return 0; + + if (rte_raw_cksum_mbuf(m, l4_off, m->pkt_len - l4_off, &raw_cksum)) + return 0; + + cksum = raw_cksum + rte_ipv4_phdr_cksum(ipv4_hdr, 0); + + cksum = ((cksum & 0xffff0000) >> 16) + (cksum & 0xffff); + + return (uint16_t)cksum; +} + +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * Compute the IPv4 UDP/TCP checksum of a packet. + * + * @param m + * The pointer to the mbuf. + * @param ipv4_hdr + * The pointer to the contiguous IPv4 header. + * @param l4_off + * The offset in bytes to start L4 checksum. + * @return + * The complemented checksum to set in the L4 header. + */ +__rte_experimental +static inline uint16_t +rte_ipv4_udptcp_cksum_mbuf(const struct rte_mbuf *m, + const struct rte_ipv4_hdr *ipv4_hdr, uint16_t l4_off) +{ + uint16_t cksum = __rte_ipv4_udptcp_cksum_mbuf(m, ipv4_hdr, l4_off); + + cksum = ~cksum; + + /* + * Per RFC 768: If the computed checksum is zero for UDP, + * it is transmitted as all ones + * (the equivalent in one's complement arithmetic). + */ + if (cksum == 0 && ipv4_hdr->next_proto_id == IPPROTO_UDP) + cksum = 0xffff; + + return cksum; +} + /** * Validate the IPv4 UDP or TCP checksum. * @@ -426,6 +485,38 @@ rte_ipv4_udptcp_cksum_verify(const struct rte_ipv4_hdr *ipv4_hdr, return 0; } +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * Verify the IPv4 UDP/TCP checksum of a packet. + * + * In case of UDP, the caller must first check if udp_hdr->dgram_cksum is 0 + * (i.e. no checksum). + * + * @param m + * The pointer to the mbuf. + * @param ipv4_hdr + * The pointer to the contiguous IPv4 header. + * @param l4_off + * The offset in bytes to start L4 checksum. + * @return + * Return 0 if the checksum is correct, else -1. + */ +__rte_experimental +static inline uint16_t +rte_ipv4_udptcp_cksum_mbuf_verify(const struct rte_mbuf *m, + const struct rte_ipv4_hdr *ipv4_hdr, + uint16_t l4_off) +{ + uint16_t cksum = __rte_ipv4_udptcp_cksum_mbuf(m, ipv4_hdr, l4_off); + + if (cksum != 0xffff) + return -1; + + return 0; +} + /** * IPv6 Header */ @@ -538,6 +629,68 @@ rte_ipv6_udptcp_cksum(const struct rte_ipv6_hdr *ipv6_hdr, const void *l4_hdr) return cksum; } +/** + * @internal Calculate the non-complemented IPv6 L4 checksum of a packet + */ +static inline uint16_t +__rte_ipv6_udptcp_cksum_mbuf(const struct rte_mbuf *m, + const struct rte_ipv6_hdr *ipv6_hdr, + uint16_t l4_off) +{ + uint16_t raw_cksum; + uint32_t cksum; + + if (l4_off > m->pkt_len) + return 0; + + if (rte_raw_cksum_mbuf(m, l4_off, m->pkt_len - l4_off, &raw_cksum)) + return 0; + + cksum = raw_cksum + rte_ipv6_phdr_cksum(ipv6_hdr, 0); + + cksum = ((cksum & 0xffff0000) >> 16) + (cksum & 0xffff); + + return (uint16_t)cksum; +} + +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * Process the IPv6 UDP or TCP checksum of a packet. + * + * The IPv6 header must not be followed by extension headers. The layer 4 + * checksum must be set to 0 in the L4 header by the caller. + * + * @param m + * The pointer to the mbuf. + * @param ipv6_hdr + * The pointer to the contiguous IPv6 header. + * @param l4_off + * The offset in bytes to start L4 checksum. + * @return + * The complemented checksum to set in the L4 header. + */ +__rte_experimental +static inline uint16_t +rte_ipv6_udptcp_cksum_mbuf(const struct rte_mbuf *m, + const struct rte_ipv6_hdr *ipv6_hdr, uint16_t l4_off) +{ + uint16_t cksum = __rte_ipv6_udptcp_cksum_mbuf(m, ipv6_hdr, l4_off); + + cksum = ~cksum; + + /* + * Per RFC 768: If the computed checksum is zero for UDP, + * it is transmitted as all ones + * (the equivalent in one's complement arithmetic). + */ + if (cksum == 0 && ipv6_hdr->proto == IPPROTO_UDP) + cksum = 0xffff; + + return cksum; +} + /** * Validate the IPv6 UDP or TCP checksum. * @@ -565,6 +718,39 @@ rte_ipv6_udptcp_cksum_verify(const struct rte_ipv6_hdr *ipv6_hdr, return 0; } +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * Validate the IPv6 UDP or TCP checksum of a packet. + * + * In case of UDP, the caller must first check if udp_hdr->dgram_cksum is 0: + * this is either invalid or means no checksum in some situations. See 8.1 + * (Upper-Layer Checksums) in RFC 8200. + * + * @param m + * The pointer to the mbuf. + * @param ipv6_hdr + * The pointer to the contiguous IPv6 header. + * @param l4_off + * The offset in bytes to start L4 checksum. + * @return + * Return 0 if the checksum is correct, else -1. + */ +__rte_experimental +static inline int +rte_ipv6_udptcp_cksum_mbuf_verify(const struct rte_mbuf *m, + const struct rte_ipv6_hdr *ipv6_hdr, + uint16_t l4_off) +{ + uint16_t cksum = __rte_ipv6_udptcp_cksum_mbuf(m, ipv6_hdr, l4_off); + + if (cksum != 0xffff) + return -1; + + return 0; +} + /** IPv6 fragment extension header. */ #define RTE_IPV6_EHDR_MF_SHIFT 0 #define RTE_IPV6_EHDR_MF_MASK 1 From patchwork Thu Jan 6 16:03:33 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Li, Xiaoyun" X-Patchwork-Id: 105664 X-Patchwork-Delegate: ferruh.yigit@amd.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id A6FD3A0351; Thu, 6 Jan 2022 17:13:54 +0100 (CET) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 1680541160; Thu, 6 Jan 2022 17:13:45 +0100 (CET) Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by mails.dpdk.org (Postfix) with ESMTP id B748F41153 for ; Thu, 6 Jan 2022 17:13:42 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1641485622; x=1673021622; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=C6csC5pVZ1g9QtgHMHT0Ymwpm/adwh4cPmk8iOK5PHw=; b=KXIJVxgn2tpQMt54kymYvXb+gBIALvGpNLHQe/FfMxjQHenoj2neRDyo Oa3W5DXICpIZ14iht2zoVcwYWwxubCD/8vZc37of/llIVYxJbyQvUsc/D rk3iJBhUdYM1YFvsoimqUsLpBtQk8w/Ka8MVkCo4EubEcjToBQvtFydcs +vIfxPfWe29V3aDcusIEmQgEPs0rEqI6mb2cHU3vL+uHoFSCV84HvSUot Y/4BbosemPc8HGJTHPcj6c63upElfLD99QeOG5m/Y2vd4DW3pHsy/yt6O yod2pdCAKqlY5B3kAXRX16+aeF1Ak4MteTqP5OC9KeJTLunRCavxRswdS g==; X-IronPort-AV: E=McAfee;i="6200,9189,10217"; a="241500767" X-IronPort-AV: E=Sophos;i="5.88,267,1635231600"; d="scan'208";a="241500767" Received: from orsmga007.jf.intel.com ([10.7.209.58]) by fmsmga104.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 06 Jan 2022 08:13:32 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.88,267,1635231600"; d="scan'208";a="513457550" Received: from dpdk-xiaoyun3.sh.intel.com ([10.67.119.214]) by orsmga007.jf.intel.com with ESMTP; 06 Jan 2022 08:13:29 -0800 From: Xiaoyun Li To: Aman.Deep.Singh@intel.com, ferruh.yigit@intel.com, olivier.matz@6wind.com, mb@smartsharesystems.com, konstantin.ananyev@intel.com, stephen@networkplumber.org, vladimir.medvedkin@intel.com Cc: dev@dpdk.org, Xiaoyun Li , Sunil Pai G Subject: [PATCH v5 2/2] testpmd: fix l4 sw csum over multi segments Date: Fri, 7 Jan 2022 00:03:33 +0800 Message-Id: <20220106160333.762686-3-xiaoyun.li@intel.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220106160333.762686-1-xiaoyun.li@intel.com> References: <20211015051306.320328-1-xiaoyun.li@intel.com> <20220106160333.762686-1-xiaoyun.li@intel.com> MIME-Version: 1.0 X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org In csum forwarding mode, software UDP/TCP csum calculation only takes the first segment into account while using the whole packet length so the calculation will read invalid memory region with multi-segments packets and will get wrong value. This patch fixes this issue. Signed-off-by: Xiaoyun Li Tested-by: Sunil Pai G --- app/test-pmd/csumonly.c | 41 ++++++++++++++++++++++++++--------------- 1 file changed, 26 insertions(+), 15 deletions(-) diff --git a/app/test-pmd/csumonly.c b/app/test-pmd/csumonly.c index 2aeea243b6..0fbe1f1be7 100644 --- a/app/test-pmd/csumonly.c +++ b/app/test-pmd/csumonly.c @@ -96,12 +96,13 @@ struct simple_gre_hdr { } __rte_packed; static uint16_t -get_udptcp_checksum(void *l3_hdr, void *l4_hdr, uint16_t ethertype) +get_udptcp_checksum(struct rte_mbuf *m, void *l3_hdr, uint16_t l4_off, + uint16_t ethertype) { if (ethertype == _htons(RTE_ETHER_TYPE_IPV4)) - return rte_ipv4_udptcp_cksum(l3_hdr, l4_hdr); + return rte_ipv4_udptcp_cksum_mbuf(m, l3_hdr, l4_off); else /* assume ethertype == RTE_ETHER_TYPE_IPV6 */ - return rte_ipv6_udptcp_cksum(l3_hdr, l4_hdr); + return rte_ipv6_udptcp_cksum_mbuf(m, l3_hdr, l4_off); } /* Parse an IPv4 header to fill l3_len, l4_len, and l4_proto */ @@ -460,7 +461,7 @@ parse_encap_ip(void *encap_ip, struct testpmd_offload_info *info) * depending on the testpmd command line configuration */ static uint64_t process_inner_cksums(void *l3_hdr, const struct testpmd_offload_info *info, - uint64_t tx_offloads) + uint64_t tx_offloads, struct rte_mbuf *m) { struct rte_ipv4_hdr *ipv4_hdr = l3_hdr; struct rte_udp_hdr *udp_hdr; @@ -468,6 +469,7 @@ process_inner_cksums(void *l3_hdr, const struct testpmd_offload_info *info, struct rte_sctp_hdr *sctp_hdr; uint64_t ol_flags = 0; uint32_t max_pkt_len, tso_segsz = 0; + uint16_t l4_off; /* ensure packet is large enough to require tso */ if (!info->is_tunnel) { @@ -510,9 +512,15 @@ process_inner_cksums(void *l3_hdr, const struct testpmd_offload_info *info, if (tx_offloads & RTE_ETH_TX_OFFLOAD_UDP_CKSUM) { ol_flags |= RTE_MBUF_F_TX_UDP_CKSUM; } else { + if (info->is_tunnel) + l4_off = info->l2_len + + info->outer_l3_len + + info->l2_len + info->l3_len; + else + l4_off = info->l2_len + info->l3_len; udp_hdr->dgram_cksum = 0; udp_hdr->dgram_cksum = - get_udptcp_checksum(l3_hdr, udp_hdr, + get_udptcp_checksum(m, l3_hdr, l4_off, info->ethertype); } } @@ -527,9 +535,14 @@ process_inner_cksums(void *l3_hdr, const struct testpmd_offload_info *info, else if (tx_offloads & RTE_ETH_TX_OFFLOAD_TCP_CKSUM) { ol_flags |= RTE_MBUF_F_TX_TCP_CKSUM; } else { + if (info->is_tunnel) + l4_off = info->l2_len + info->outer_l3_len + + info->l2_len + info->l3_len; + else + l4_off = info->l2_len + info->l3_len; tcp_hdr->cksum = 0; tcp_hdr->cksum = - get_udptcp_checksum(l3_hdr, tcp_hdr, + get_udptcp_checksum(m, l3_hdr, l4_off, info->ethertype); } #ifdef RTE_LIB_GSO @@ -557,7 +570,7 @@ process_inner_cksums(void *l3_hdr, const struct testpmd_offload_info *info, /* Calculate the checksum of outer header */ static uint64_t process_outer_cksums(void *outer_l3_hdr, struct testpmd_offload_info *info, - uint64_t tx_offloads, int tso_enabled) + uint64_t tx_offloads, int tso_enabled, struct rte_mbuf *m) { struct rte_ipv4_hdr *ipv4_hdr = outer_l3_hdr; struct rte_ipv6_hdr *ipv6_hdr = outer_l3_hdr; @@ -611,12 +624,9 @@ process_outer_cksums(void *outer_l3_hdr, struct testpmd_offload_info *info, /* do not recalculate udp cksum if it was 0 */ if (udp_hdr->dgram_cksum != 0) { udp_hdr->dgram_cksum = 0; - if (info->outer_ethertype == _htons(RTE_ETHER_TYPE_IPV4)) - udp_hdr->dgram_cksum = - rte_ipv4_udptcp_cksum(ipv4_hdr, udp_hdr); - else - udp_hdr->dgram_cksum = - rte_ipv6_udptcp_cksum(ipv6_hdr, udp_hdr); + udp_hdr->dgram_cksum = get_udptcp_checksum(m, outer_l3_hdr, + info->l2_len + info->outer_l3_len, + info->outer_ethertype); } return ol_flags; @@ -957,7 +967,7 @@ pkt_burst_checksum_forward(struct fwd_stream *fs) /* process checksums of inner headers first */ tx_ol_flags |= process_inner_cksums(l3_hdr, &info, - tx_offloads); + tx_offloads, m); /* Then process outer headers if any. Note that the software * checksum will be wrong if one of the inner checksums is @@ -965,7 +975,8 @@ pkt_burst_checksum_forward(struct fwd_stream *fs) if (info.is_tunnel == 1) { tx_ol_flags |= process_outer_cksums(outer_l3_hdr, &info, tx_offloads, - !!(tx_ol_flags & RTE_MBUF_F_TX_TCP_SEG)); + !!(tx_ol_flags & RTE_MBUF_F_TX_TCP_SEG), + m); } /* step 3: fill the mbuf meta data (flags and header lengths) */