From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by passt.top (Postfix) with ESMTP id 73B805A0271 for ; Fri, 12 Apr 2024 20:00:33 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1712944832; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=CQ97mbcpV7re32BrARKKx4PmHZbLKLRaZgBBYLaXTPE=; b=bz/ilQdVN51ARPv3PIOgxN9pulpJZMgQeto3V6FY/zOnu6tVVoGP6iaCpRKzubCqw/LZXG 4fyyCJ/MQIdWW/ny2LTXd8ScDEV3nxyP+onX12tbZ2ExmrciQBZ+hGUsWtzKK+OqUhNDT7 kG7MTspU/qsK4Ypn4lqv4g6szrvAFzQ= Received: from mail-ej1-f72.google.com (mail-ej1-f72.google.com [209.85.218.72]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-557-grjpUPhWMe2e8sd9XfHoXA-1; Fri, 12 Apr 2024 14:00:30 -0400 X-MC-Unique: grjpUPhWMe2e8sd9XfHoXA-1 Received: by mail-ej1-f72.google.com with SMTP id a640c23a62f3a-a51c76700adso245053866b.0 for ; Fri, 12 Apr 2024 11:00:29 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1712944829; x=1713549629; h=content-transfer-encoding:mime-version:organization:references :in-reply-to:message-id:subject:cc:to:from:date:x-gm-message-state :from:to:cc:subject:date:message-id:reply-to; bh=CQ97mbcpV7re32BrARKKx4PmHZbLKLRaZgBBYLaXTPE=; b=boJMHRX5mJU6tfLvp3dIinEVtiRCh8Oad+/WaqjKBAPxwofatBBEVIZPIv5rU4WOON Geho6KCml6EoSCxp8wfpQwton+IKg8StZMNmpWAXYgmez8vsHgcs/a/I9JLvCsuXVhia yzrrdM7Z/cMrgWb6h/WTG2VdsiKayZV9BAKLa7ocdYlPseXxxyxUSR+w1aynNnTtx3DT EvW+4hZVh939iCNo0UGtFuniG8C2OR1fjnP3QOz4nST+yik0pPdqr0G0/q0Yenpi2nAW GoCxxIB58JxmlwCW+81xI3XkLxvL0BJMP9Piw+mrWGz5xZ8ewy3dOUmAGMvDcDCHa3Sw hVbg== X-Gm-Message-State: AOJu0Yz0+lmFJhJz62nNSYxEOZliDa+391dBHdoANU4cFtv46Y2ikas1 dZFkI4bD2vlkU2m7JJrZgJt2Xh+JkszEn++aiJ17GLIW8BwRkjXvVejmgi9ZuKGU8RyP97muw3d uJWz6fbdt9aDjp8vBWIZSpQyFZDlmutNuVBosU/84/Crh0J6ZLA== X-Received: by 2002:a17:906:1c1b:b0:a52:1635:9d16 with SMTP id k27-20020a1709061c1b00b00a5216359d16mr4144945ejg.38.1712944828592; Fri, 12 Apr 2024 11:00:28 -0700 (PDT) X-Google-Smtp-Source: AGHT+IFGQl0MhTjbM9qjMY7dbnSVg4hmcDZpCXBLZjOLUEBbgnmTVfaud7z/D2mst9xpIMMbQUE8yA== X-Received: by 2002:a17:906:1c1b:b0:a52:1635:9d16 with SMTP id k27-20020a1709061c1b00b00a5216359d16mr4144923ejg.38.1712944827707; Fri, 12 Apr 2024 11:00:27 -0700 (PDT) Received: from maya.cloud.tilaa.com (maya.cloud.tilaa.com. [164.138.29.33]) by smtp.gmail.com with ESMTPSA id w6-20020a1709064a0600b00a5209dc79c1sm2086907eju.146.2024.04.12.11.00.26 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Fri, 12 Apr 2024 11:00:26 -0700 (PDT) Date: Fri, 12 Apr 2024 19:59:52 +0200 From: Stefano Brivio To: Laurent Vivier Subject: Re: [RFC v4] tcp: Replace TCP buffer structure by an iovec array Message-ID: <20240412195952.679fc8e3@elisabeth> In-Reply-To: <20240321102655.2003763-1-lvivier@redhat.com> References: <20240321102655.2003763-1-lvivier@redhat.com> Organization: Red Hat X-Mailer: Claws Mail 4.2.0 (GTK 3.24.36; x86_64-pc-linux-gnu) MIME-Version: 1.0 X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Message-ID-Hash: ZIDNG2HSISDUD3OXKAY3ITJCCQ3NFYGF X-Message-ID-Hash: ZIDNG2HSISDUD3OXKAY3ITJCCQ3NFYGF X-MailFrom: sbrivio@redhat.com X-Mailman-Rule-Misses: dmarc-mitigation; no-senders; approved; emergency; loop; banned-address; member-moderation; nonmember-moderation; administrivia; implicit-dest; max-recipients; max-size; news-moderation; no-subject; digests; suspicious-header CC: passt-dev@passt.top, David Gibson X-Mailman-Version: 3.3.8 Precedence: list List-Id: Development discussion and patches for passt Archived-At: Archived-At: List-Archive: List-Archive: List-Help: List-Owner: List-Post: List-Subscribe: List-Unsubscribe: On Thu, 21 Mar 2024 11:26:55 +0100 Laurent Vivier wrote: > To be able to provide pointers to TCP headers and IP headers without > worrying about alignment in the structure, split the structure into > several arrays and point to each part of the frame using an iovec array. > > Using iovec also allows us to simply ignore the first entry when the > vnet length header is not needed. > > Signed-off-by: Laurent Vivier > Reviewed-by: David Gibson Sorry for the delay. A few comments: > --- > > Notes: > v4: > - fix static_assert() > > v3: > - use (again) a 2d array of iovecs > - fix pcap_multiple() call > - introduce IP_MAX_MTU to define MSS4 and MSS6 > - use VAR[0].data rather than ((TYPE *)0)->data in static_assert() > - cleanup tcp_iov_parts comments > - use sizeof(VAR) rather than sizeof(TYPE) > - fix tcp_l2_buf_fill_headers function comment > - don't mix network order and host order in vnet_len > - with dup_iov, update only TCP_IOV_PAYLOAD length > - remove _l2_ from the buffers names > > v2: > - rebased on top of David's series > "Some improvements to the tap send path" > - no performance improvement anymore > - remove the iovec functions of v1 introduced in tap.c to use new > functions from David > - don't use an array of array of iovec > - fix comments > - re-introduce MSS4 and MSS6 > - address comments from David and Stefano > > I have tested only passt, not pasta > > tcp.c | 524 +++++++++++++++++++++++++++------------------------------ > util.h | 3 + > 2 files changed, 253 insertions(+), 274 deletions(-) > > diff --git a/tcp.c b/tcp.c > index a1860d10b15f..cc705064f059 100644 > --- a/tcp.c > +++ b/tcp.c > @@ -318,39 +318,8 @@ > > /* MSS rounding: see SET_MSS() */ > #define MSS_DEFAULT 536 > - > -struct tcp4_l2_head { /* For MSS4 macro: keep in sync with tcp4_l2_buf_t */ > -#ifdef __AVX2__ > - uint8_t pad[26]; > -#else > - uint8_t pad[2]; > -#endif > - struct tap_hdr taph; > - struct iphdr iph; > - struct tcphdr th; > -#ifdef __AVX2__ > -} __attribute__ ((packed, aligned(32))); > -#else > -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))); > -#endif > - > -struct tcp6_l2_head { /* For MSS6 macro: keep in sync with tcp6_l2_buf_t */ > -#ifdef __AVX2__ > - uint8_t pad[14]; > -#else > - uint8_t pad[2]; > -#endif > - struct tap_hdr taph; > - struct ipv6hdr ip6h; > - struct tcphdr th; > -#ifdef __AVX2__ > -} __attribute__ ((packed, aligned(32))); > -#else > -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))); > -#endif > - > -#define MSS4 ROUND_DOWN(USHRT_MAX - sizeof(struct tcp4_l2_head), 4) > -#define MSS6 ROUND_DOWN(USHRT_MAX - sizeof(struct tcp6_l2_head), 4) > +#define MSS4 ROUND_DOWN(IP_MAX_MTU - sizeof(struct tcphdr) - sizeof(struct iphdr), sizeof(uint32_t)) > +#define MSS6 ROUND_DOWN(IP_MAX_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr), sizeof(uint32_t)) Would it be possible to keep those within 80 columns? That is, #define MSS4 ROUND_DOWN(IP_MAX_MTU - sizeof(struct tcphdr) - sizeof(struct iphdr), sizeof(uint32_t)) #define MSS6 ROUND_DOWN(IP_MAX_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr), sizeof(uint32_t)) doesn't look so bad to me. > > #define WINDOW_DEFAULT 14600 /* RFC 6928 */ > #ifdef HAS_SND_WND > @@ -445,133 +414,107 @@ struct tcp_buf_seq_update { > }; > > /* Static buffers */ > - > /** > - * tcp4_l2_buf_t - Pre-cooked IPv4 packet buffers for tap connections > - * @pad: Align TCP header to 32 bytes, for AVX2 checksum calculation only > - * @taph: Tap-level headers (partially pre-filled) > - * @iph: Pre-filled IP header (except for tot_len and saddr) > - * @uh: Headroom for TCP header > - * @data: Storage for TCP payload > + * struct tcp_payload_t - TCP header and data to send data While "data to send data" strictly speaking refers what this represents, it's a bit obscure, especially in combination with the tcp_payload_t name. What about: * struct tcp_payload_t - TCP header and data to send segments with payload ? > + * 32 bytes aligned to be able to use AVX2 checksum Not true if !__AVX2__. I would just drop this line and explain below (see below). > + * @th: TCP header > + * @data: TCP data > */ > -static struct tcp4_l2_buf_t { > -#ifdef __AVX2__ > - uint8_t pad[26]; /* 0, align th to 32 bytes */ > -#else > - uint8_t pad[2]; /* align iph to 4 bytes 0 */ > -#endif > - struct tap_hdr taph; /* 26 2 */ > - struct iphdr iph; /* 44 20 */ > - struct tcphdr th; /* 64 40 */ > - uint8_t data[MSS4]; /* 84 60 */ > - /* 65536 65532 */ > +struct tcp_payload_t { > + struct tcphdr th; > + uint8_t data[65536 - sizeof(struct tcphdr)]; We pick 65536 here because it's IP_MAX_MTU, now that you defined it, we might as well use it. > #ifdef __AVX2__ > -} __attribute__ ((packed, aligned(32))) > +} __attribute__ ((packed, aligned(32))); We could explain that here, e.g.: } __attribute__ ((packed, aligned(32))); /* For AVX2 checksum routines */ > #else > -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))) > +} __attribute__ ((packed, aligned(__alignof__(unsigned int)))); > #endif > -tcp4_l2_buf[TCP_FRAMES_MEM]; > - > -static struct tcp_buf_seq_update tcp4_l2_buf_seq_update[TCP_FRAMES_MEM]; > - > -static unsigned int tcp4_l2_buf_used; > > /** > - * tcp6_l2_buf_t - Pre-cooked IPv6 packet buffers for tap connections > - * @pad: Align IPv6 header for checksum calculation to 32B (AVX2) or 4B > - * @taph: Tap-level headers (partially pre-filled) > - * @ip6h: Pre-filled IP header (except for payload_len and addresses) > - * @th: Headroom for TCP header > - * @data: Storage for TCP payload > + * struct tcp_flags_t - TCP header and data to send option flags Actually, there's no such thing as "option flags", so, similarly to my remark above, what about: * struct tcp_flags_t - TCP header and data to send zero-length segments (flags) ? > + * @th: TCP header > + * @opts TCP option flags I'd say "TCP options". > */ > -struct tcp6_l2_buf_t { > -#ifdef __AVX2__ > - uint8_t pad[14]; /* 0 align ip6h to 32 bytes */ > -#else > - uint8_t pad[2]; /* align ip6h to 4 bytes 0 */ > -#endif > - struct tap_hdr taph; /* 14 2 */ > - struct ipv6hdr ip6h; /* 32 20 */ > - struct tcphdr th; /* 72 60 */ > - uint8_t data[MSS6]; /* 92 80 */ > - /* 65536 65532 */ > +struct tcp_flags_t { > + struct tcphdr th; > + char opts[OPT_MSS_LEN + OPT_WS_LEN + 1]; > #ifdef __AVX2__ > -} __attribute__ ((packed, aligned(32))) > +} __attribute__ ((packed, aligned(32))); > #else > -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))) > +} __attribute__ ((packed, aligned(__alignof__(unsigned int)))); > #endif > -tcp6_l2_buf[TCP_FRAMES_MEM]; > > -static struct tcp_buf_seq_update tcp6_l2_buf_seq_update[TCP_FRAMES_MEM]; > +/* Ethernet header for IPv4 frames */ > +static struct ethhdr tcp4_eth_src; > + > +static uint32_t tcp4_payload_vnet_len[TCP_FRAMES_MEM]; > +/* IPv4 headers */ > +static struct iphdr tcp4_payload_ip[TCP_FRAMES_MEM]; > +/* TCP headers and data for IPv4 frames */ Well, TCP headers and data make up segments, so perhaps: /* TCP segments with payload for IPv4 frames */ ...which is easier to understand as a distinction from tcp4_flags[] below: > +static struct tcp_payload_t tcp4_payload[TCP_FRAMES_MEM]; > + > +static_assert(MSS4 <= sizeof(tcp4_payload[0].data), "MSS4 is greater than 65516"); > + > +static struct tcp_buf_seq_update tcp4_seq_update[TCP_FRAMES_MEM]; > +static unsigned int tcp4_payload_used; > + > +static uint32_t tcp4_flags_vnet_len[TCP_FRAMES_MEM]; > +/* IPv4 headers for TCP option flags frames */ > +static struct iphdr tcp4_flags_ip[TCP_FRAMES_MEM]; > +/* TCP headers and option flags for IPv4 frames */ /* TCP segments without payload for IPv4 frames */ ? > +static struct tcp_flags_t tcp4_flags[TCP_FRAMES_MEM]; > + > +static unsigned int tcp4_flags_used; > > -static unsigned int tcp6_l2_buf_used; > +/* Ethernet header for IPv6 frames */ > +static struct ethhdr tcp6_eth_src; > + > +static uint32_t tcp6_payload_vnet_len[TCP_FRAMES_MEM]; > +/* IPv6 headers */ > +static struct ipv6hdr tcp6_payload_ip[TCP_FRAMES_MEM]; > +/* TCP headers and data for IPv6 frames */ > +static struct tcp_payload_t tcp6_payload[TCP_FRAMES_MEM]; > + > +static_assert(MSS6 <= sizeof(tcp6_payload[0].data), "MSS6 is greater than 65516"); > + > +static struct tcp_buf_seq_update tcp6_seq_update[TCP_FRAMES_MEM]; > +static unsigned int tcp6_payload_used; > + > +static uint32_t tcp6_flags_vnet_len[TCP_FRAMES_MEM]; > +/* IPv6 headers for TCP option flags frames */ > +static struct ipv6hdr tcp6_flags_ip[TCP_FRAMES_MEM]; > +/* TCP headers and option flags for IPv6 frames */ > +static struct tcp_flags_t tcp6_flags[TCP_FRAMES_MEM]; ...same here and above (three occurrences). > + > +static unsigned int tcp6_flags_used; > > /* recvmsg()/sendmsg() data for tap */ > static char tcp_buf_discard [MAX_WINDOW]; > static struct iovec iov_sock [TCP_FRAMES_MEM + 1]; > > -static struct iovec tcp4_l2_iov [TCP_FRAMES_MEM]; > -static struct iovec tcp6_l2_iov [TCP_FRAMES_MEM]; > -static struct iovec tcp4_l2_flags_iov [TCP_FRAMES_MEM]; > -static struct iovec tcp6_l2_flags_iov [TCP_FRAMES_MEM]; > +/* > + * enum tcp_iov_parts - I/O vector parts for one TCP frame > + * @TCP_IOV_TAP TAP header It's kind of funny that TCP_IOV_TAP is only used for passt, which, unlike pasta, does *not* use a tap device. I haven't really looked into how complicated this is at this point, but now that this is finally split out of the "Ethernet" part, could we call this IOV_VLEN or suchlike, and rename all related references? > + * @TCP_IOV_ETH ethernet header s/ethernet/Ethernet/ > + * @TCP_IOV_IP IP (v4/v6) header > + * @TCP_IOV_PAYLOAD IP payload (TCP header + data) > + * @TCP_NUM_IOVS the number of entries in the iovec array > + */ > +enum tcp_iov_parts { > + TCP_IOV_TAP = 0, > + TCP_IOV_ETH = 1, > + TCP_IOV_IP = 2, > + TCP_IOV_PAYLOAD = 3, > + TCP_NUM_IOVS > +}; > + > +static struct iovec tcp4_l2_iov [TCP_FRAMES_MEM][TCP_NUM_IOVS]; > +static struct iovec tcp6_l2_iov [TCP_FRAMES_MEM][TCP_NUM_IOVS]; > +static struct iovec tcp4_l2_flags_iov [TCP_FRAMES_MEM][TCP_NUM_IOVS]; > +static struct iovec tcp6_l2_flags_iov [TCP_FRAMES_MEM][TCP_NUM_IOVS]; > > /* sendmsg() to socket */ > static struct iovec tcp_iov [UIO_MAXIOV]; > > -/** > - * tcp4_l2_flags_buf_t - IPv4 packet buffers for segments without data (flags) > - * @pad: Align TCP header to 32 bytes, for AVX2 checksum calculation only > - * @taph: Tap-level headers (partially pre-filled) > - * @iph: Pre-filled IP header (except for tot_len and saddr) > - * @th: Headroom for TCP header > - * @opts: Headroom for TCP options > - */ > -static struct tcp4_l2_flags_buf_t { > -#ifdef __AVX2__ > - uint8_t pad[26]; /* 0, align th to 32 bytes */ > -#else > - uint8_t pad[2]; /* align iph to 4 bytes 0 */ > -#endif > - struct tap_hdr taph; /* 26 2 */ > - struct iphdr iph; /* 44 20 */ > - struct tcphdr th; /* 64 40 */ > - char opts[OPT_MSS_LEN + OPT_WS_LEN + 1]; > -#ifdef __AVX2__ > -} __attribute__ ((packed, aligned(32))) > -#else > -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))) > -#endif > -tcp4_l2_flags_buf[TCP_FRAMES_MEM]; > - > -static unsigned int tcp4_l2_flags_buf_used; > - > -/** > - * tcp6_l2_flags_buf_t - IPv6 packet buffers for segments without data (flags) > - * @pad: Align IPv6 header for checksum calculation to 32B (AVX2) or 4B > - * @taph: Tap-level headers (partially pre-filled) > - * @ip6h: Pre-filled IP header (except for payload_len and addresses) > - * @th: Headroom for TCP header > - * @opts: Headroom for TCP options > - */ > -static struct tcp6_l2_flags_buf_t { > -#ifdef __AVX2__ > - uint8_t pad[14]; /* 0 align ip6h to 32 bytes */ > -#else > - uint8_t pad[2]; /* align ip6h to 4 bytes 0 */ > -#endif > - struct tap_hdr taph; /* 14 2 */ > - struct ipv6hdr ip6h; /* 32 20 */ > - struct tcphdr th /* 72 */ __attribute__ ((aligned(4))); /* 60 */ > - char opts[OPT_MSS_LEN + OPT_WS_LEN + 1]; > -#ifdef __AVX2__ > -} __attribute__ ((packed, aligned(32))) > -#else > -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))) > -#endif > -tcp6_l2_flags_buf[TCP_FRAMES_MEM]; > - > -static unsigned int tcp6_l2_flags_buf_used; > - > #define CONN(idx) (&(FLOW(idx)->tcp)) > > /* Table for lookup from remote address, local port, remote port */ > @@ -967,25 +910,14 @@ static void tcp_update_check_tcp6(struct ipv6hdr *ip6h, struct tcphdr *th) > } > > /** > - * tcp_update_l2_buf() - Update L2 buffers with Ethernet and IPv4 addresses > + * tcp_update_l2_buf() - Update ethernet header buffers with addresses s/ethernet/Ethernet/ > * @eth_d: Ethernet destination address, NULL if unchanged > * @eth_s: Ethernet source address, NULL if unchanged > */ > void tcp_update_l2_buf(const unsigned char *eth_d, const unsigned char *eth_s) > { > - int i; > - > - for (i = 0; i < TCP_FRAMES_MEM; i++) { > - struct tcp4_l2_flags_buf_t *b4f = &tcp4_l2_flags_buf[i]; > - struct tcp6_l2_flags_buf_t *b6f = &tcp6_l2_flags_buf[i]; > - struct tcp4_l2_buf_t *b4 = &tcp4_l2_buf[i]; > - struct tcp6_l2_buf_t *b6 = &tcp6_l2_buf[i]; > - > - eth_update_mac(&b4->taph.eh, eth_d, eth_s); > - eth_update_mac(&b6->taph.eh, eth_d, eth_s); > - eth_update_mac(&b4f->taph.eh, eth_d, eth_s); > - eth_update_mac(&b6f->taph.eh, eth_d, eth_s); > - } > + eth_update_mac(&tcp4_eth_src, eth_d, eth_s); > + eth_update_mac(&tcp6_eth_src, eth_d, eth_s); > } > > /** > @@ -995,29 +927,42 @@ void tcp_update_l2_buf(const unsigned char *eth_d, const unsigned char *eth_s) > static void tcp_sock4_iov_init(const struct ctx *c) > { > struct iphdr iph = L2_BUF_IP4_INIT(IPPROTO_TCP); > - struct iovec *iov; > int i; > > - for (i = 0; i < ARRAY_SIZE(tcp4_l2_buf); i++) { > - tcp4_l2_buf[i] = (struct tcp4_l2_buf_t) { > - .taph = TAP_HDR_INIT(ETH_P_IP), > - .iph = iph, > - .th = { .doff = sizeof(struct tcphdr) / 4, .ack = 1 } > - }; > - } > + tcp4_eth_src.h_proto = htons_constant(ETH_P_IP); An extra newline here would be nice, as we're switching to Layer-3 headers now. > + for (i = 0; i < TCP_FRAMES_MEM; i++) { The existing version has two advantages over using i < TCP_FRAMES_MEM: the array sizes don't necessarily need to match (even though they do), but perhaps more importantly ARRAY_SIZE() clearly shows on what the loop boundary is based on. And two loops is probably a bit clearer than having one, because this gives the impression that the two arrays are somewhat linked (same amount, so it almost looks like there are header buffers first and payload buffers next). > + struct iovec *iov; > > - for (i = 0; i < ARRAY_SIZE(tcp4_l2_flags_buf); i++) { > - tcp4_l2_flags_buf[i] = (struct tcp4_l2_flags_buf_t) { > - .taph = TAP_HDR_INIT(ETH_P_IP), > - .iph = L2_BUF_IP4_INIT(IPPROTO_TCP) > - }; > - } > + /* headers */ > + tcp4_payload_ip[i] = iph; > + tcp4_payload[i].th.doff = sizeof(struct tcphdr) / 4; > + tcp4_payload[i].th.ack = 1; > + > + tcp4_flags_ip[i] = iph; > + tcp4_flags[i].th.doff = sizeof(struct tcphdr) / 4; > + tcp4_flags[i].th.ack = 1; > > - for (i = 0, iov = tcp4_l2_iov; i < TCP_FRAMES_MEM; i++, iov++) > - iov->iov_base = tap_frame_base(c, &tcp4_l2_buf[i].taph); > + /* iovecs */ > + iov = tcp4_l2_iov[i]; > + iov[TCP_IOV_TAP].iov_base = &tcp4_payload_vnet_len[i]; > + iov[TCP_IOV_TAP].iov_len = c->mode == MODE_PASST ? > + sizeof(tcp4_payload_vnet_len[i]) : 0; > + iov[TCP_IOV_ETH].iov_base = &tcp4_eth_src; > + iov[TCP_IOV_ETH].iov_len = sizeof(tcp4_eth_src); > + iov[TCP_IOV_IP].iov_base = &tcp4_payload_ip[i]; > + iov[TCP_IOV_IP].iov_len = sizeof(tcp4_payload_ip[i]); > + iov[TCP_IOV_PAYLOAD].iov_base = &tcp4_payload[i]; > > - for (i = 0, iov = tcp4_l2_flags_iov; i < TCP_FRAMES_MEM; i++, iov++) > - iov->iov_base = tap_frame_base(c, &tcp4_l2_flags_buf[i].taph); > + iov = tcp4_l2_flags_iov[i]; > + iov[TCP_IOV_TAP].iov_base = &tcp4_flags_vnet_len[i]; > + iov[TCP_IOV_TAP].iov_len = c->mode == MODE_PASST ? > + sizeof(tcp4_flags_vnet_len[i]) : 0; > + iov[TCP_IOV_ETH].iov_base = &tcp4_eth_src; > + iov[TCP_IOV_ETH].iov_len = sizeof(tcp4_eth_src); > + iov[TCP_IOV_IP].iov_base = &tcp4_flags_ip[i]; > + iov[TCP_IOV_IP].iov_len = sizeof(tcp4_flags_ip[i]); > + iov[TCP_IOV_PAYLOAD].iov_base = &tcp4_flags[i]; > + } > } > > /** > @@ -1026,29 +971,43 @@ static void tcp_sock4_iov_init(const struct ctx *c) > */ > static void tcp_sock6_iov_init(const struct ctx *c) > { > - struct iovec *iov; > + struct ipv6hdr ip6 = L2_BUF_IP6_INIT(IPPROTO_TCP); > int i; > > - for (i = 0; i < ARRAY_SIZE(tcp6_l2_buf); i++) { > - tcp6_l2_buf[i] = (struct tcp6_l2_buf_t) { > - .taph = TAP_HDR_INIT(ETH_P_IPV6), > - .ip6h = L2_BUF_IP6_INIT(IPPROTO_TCP), > - .th = { .doff = sizeof(struct tcphdr) / 4, .ack = 1 } > - }; > - } > + tcp6_eth_src.h_proto = htons_constant(ETH_P_IPV6); > + for (i = 0; i < TCP_FRAMES_MEM; i++) { Same here. > + struct iovec *iov; > > - for (i = 0; i < ARRAY_SIZE(tcp6_l2_flags_buf); i++) { > - tcp6_l2_flags_buf[i] = (struct tcp6_l2_flags_buf_t) { > - .taph = TAP_HDR_INIT(ETH_P_IPV6), > - .ip6h = L2_BUF_IP6_INIT(IPPROTO_TCP) > - }; > - } > + /* headers */ > + tcp6_payload_ip[i] = ip6; > + tcp6_payload[i].th.doff = sizeof(struct tcphdr) / 4; > + tcp6_payload[i].th.ack = 1; > + > + tcp6_flags_ip[i] = ip6; > + tcp6_flags[i].th.doff = sizeof(struct tcphdr) / 4; > + tcp6_flags[i].th .ack = 1; > > - for (i = 0, iov = tcp6_l2_iov; i < TCP_FRAMES_MEM; i++, iov++) > - iov->iov_base = tap_frame_base(c, &tcp6_l2_buf[i].taph); > + /* iovecs */ > + iov = tcp6_l2_iov[i]; > + iov[TCP_IOV_TAP].iov_base = &tcp6_payload_vnet_len[i]; > + iov[TCP_IOV_TAP].iov_len = c->mode == MODE_PASST ? > + sizeof(tcp6_payload_vnet_len[i]) : 0; > + iov[TCP_IOV_ETH].iov_base = &tcp6_eth_src; > + iov[TCP_IOV_ETH].iov_len = sizeof(tcp6_eth_src); > + iov[TCP_IOV_IP].iov_base = &tcp6_payload_ip[i]; > + iov[TCP_IOV_IP].iov_len = sizeof(tcp6_payload_ip[i]); > + iov[TCP_IOV_PAYLOAD].iov_base = &tcp6_payload[i]; > > - for (i = 0, iov = tcp6_l2_flags_iov; i < TCP_FRAMES_MEM; i++, iov++) > - iov->iov_base = tap_frame_base(c, &tcp6_l2_flags_buf[i].taph); > + iov = tcp6_l2_flags_iov[i]; > + iov[TCP_IOV_TAP].iov_base = &tcp6_flags_vnet_len[i]; > + iov[TCP_IOV_TAP].iov_len = c->mode == MODE_PASST ? > + sizeof(tcp6_flags_vnet_len[i]) : 0; > + iov[TCP_IOV_ETH].iov_base = &tcp6_eth_src; > + iov[TCP_IOV_ETH].iov_len = sizeof(tcp6_eth_src); > + iov[TCP_IOV_IP].iov_base = &tcp6_flags_ip[i]; > + iov[TCP_IOV_IP].iov_len = sizeof(tcp6_flags_ip[i]); > + iov[TCP_IOV_PAYLOAD].iov_base = &tcp6_flags[i]; > + } > } > > /** > @@ -1284,36 +1243,40 @@ static void tcp_rst_do(struct ctx *c, struct tcp_tap_conn *conn); > } while (0) > > /** > - * tcp_l2_flags_buf_flush() - Send out buffers for segments with no data (flags) > + * tcp_flags_flush() - Send out buffers for segments with no data (flags) > * @c: Execution context > */ > -static void tcp_l2_flags_buf_flush(const struct ctx *c) > +static void tcp_flags_flush(const struct ctx *c) > { > - tap_send_frames(c, tcp6_l2_flags_iov, 1, tcp6_l2_flags_buf_used); > - tcp6_l2_flags_buf_used = 0; > + tap_send_frames(c, &tcp6_l2_flags_iov[0][0], TCP_NUM_IOVS, > + tcp6_flags_used); > + tcp6_flags_used = 0; > > - tap_send_frames(c, tcp4_l2_flags_iov, 1, tcp4_l2_flags_buf_used); > - tcp4_l2_flags_buf_used = 0; > + tap_send_frames(c, &tcp4_l2_flags_iov[0][0], TCP_NUM_IOVS, > + tcp4_flags_used); > + tcp4_flags_used = 0; > } > > /** > - * tcp_l2_data_buf_flush() - Send out buffers for segments with data > + * tcp_payload_flush() - Send out buffers for segments with data > * @c: Execution context > */ > -static void tcp_l2_data_buf_flush(const struct ctx *c) > +static void tcp_payload_flush(const struct ctx *c) > { > unsigned i; > size_t m; > > - m = tap_send_frames(c, tcp6_l2_iov, 1, tcp6_l2_buf_used); > + m = tap_send_frames(c, &tcp6_l2_iov[0][0], TCP_NUM_IOVS, > + tcp6_payload_used); > for (i = 0; i < m; i++) > - *tcp6_l2_buf_seq_update[i].seq += tcp6_l2_buf_seq_update[i].len; > - tcp6_l2_buf_used = 0; > + *tcp6_seq_update[i].seq += tcp6_seq_update[i].len; > + tcp6_payload_used = 0; > > - m = tap_send_frames(c, tcp4_l2_iov, 1, tcp4_l2_buf_used); > + m = tap_send_frames(c, &tcp4_l2_iov[0][0], TCP_NUM_IOVS, > + tcp4_payload_used); > for (i = 0; i < m; i++) > - *tcp4_l2_buf_seq_update[i].seq += tcp4_l2_buf_seq_update[i].len; > - tcp4_l2_buf_used = 0; > + *tcp4_seq_update[i].seq += tcp4_seq_update[i].len; > + tcp4_payload_used = 0; > } > > /** > @@ -1323,8 +1286,8 @@ static void tcp_l2_data_buf_flush(const struct ctx *c) > /* cppcheck-suppress [constParameterPointer, unmatchedSuppression] */ > void tcp_defer_handler(struct ctx *c) > { > - tcp_l2_flags_buf_flush(c); > - tcp_l2_data_buf_flush(c); > + tcp_flags_flush(c); > + tcp_payload_flush(c); > } > > /** > @@ -1433,35 +1396,31 @@ static size_t tcp_fill_headers6(const struct ctx *c, > * tcp_l2_buf_fill_headers() - Fill 802.3, IP, TCP headers in pre-cooked buffers > * @c: Execution context > * @conn: Connection pointer > - * @p: Pointer to any type of TCP pre-cooked buffer > + * @iov: Pointer to an array of iovec of TCP pre-cooked buffer s/buffer/buffers/ now. > * @plen: Payload length (including TCP header options) > * @check: Checksum, if already known > * @seq: Sequence number for this segment > * > - * Return: frame length including L2 headers, host order > + * Return: IP payload length, host order > */ > static size_t tcp_l2_buf_fill_headers(const struct ctx *c, > const struct tcp_tap_conn *conn, > - void *p, size_t plen, > + struct iovec *iov, size_t plen, > const uint16_t *check, uint32_t seq) > { > const struct in_addr *a4 = inany_v4(&conn->faddr); > size_t ip_len, tlen; > > if (a4) { > - struct tcp4_l2_buf_t *b = (struct tcp4_l2_buf_t *)p; > - > - ip_len = tcp_fill_headers4(c, conn, &b->iph, &b->th, plen, > + ip_len = tcp_fill_headers4(c, conn, iov[TCP_IOV_IP].iov_base, > + iov[TCP_IOV_PAYLOAD].iov_base, plen, > check, seq); > - > - tlen = tap_frame_len(c, &b->taph, ip_len); > + tlen = ip_len - sizeof(struct iphdr); > } else { > - struct tcp6_l2_buf_t *b = (struct tcp6_l2_buf_t *)p; > - > - ip_len = tcp_fill_headers6(c, conn, &b->ip6h, &b->th, plen, > + ip_len = tcp_fill_headers6(c, conn, iov[TCP_IOV_IP].iov_base, > + iov[TCP_IOV_PAYLOAD].iov_base, plen, > seq); > - > - tlen = tap_frame_len(c, &b->taph, ip_len); > + tlen = ip_len - sizeof(struct ipv6hdr); > } > > return tlen; > @@ -1595,16 +1554,16 @@ static int tcp_send_flag(struct ctx *c, struct tcp_tap_conn *conn, int flags) > { > uint32_t prev_ack_to_tap = conn->seq_ack_to_tap; > uint32_t prev_wnd_to_tap = conn->wnd_to_tap; > - struct tcp4_l2_flags_buf_t *b4 = NULL; > - struct tcp6_l2_flags_buf_t *b6 = NULL; > + struct tcp_flags_t *payload; > struct tcp_info tinfo = { 0 }; > socklen_t sl = sizeof(tinfo); > int s = conn->sock; > + uint32_t vnet_len; > size_t optlen = 0; > - struct iovec *iov; > struct tcphdr *th; > + struct iovec *iov; > + size_t ip_len; > char *data; > - void *p; > > if (SEQ_GE(conn->seq_ack_to_tap, conn->seq_from_tap) && > !flags && conn->wnd_to_tap) > @@ -1627,19 +1586,17 @@ static int tcp_send_flag(struct ctx *c, struct tcp_tap_conn *conn, int flags) > return 0; > > if (CONN_V4(conn)) { > - iov = tcp4_l2_flags_iov + tcp4_l2_flags_buf_used; > - p = b4 = tcp4_l2_flags_buf + tcp4_l2_flags_buf_used++; > - th = &b4->th; > - > - /* gcc 11.2 would complain on data = (char *)(th + 1); */ > - data = b4->opts; > + iov = tcp4_l2_flags_iov[tcp4_flags_used++]; > + vnet_len = sizeof(struct ethhdr) + sizeof(struct iphdr); > } else { > - iov = tcp6_l2_flags_iov + tcp6_l2_flags_buf_used; > - p = b6 = tcp6_l2_flags_buf + tcp6_l2_flags_buf_used++; > - th = &b6->th; > - data = b6->opts; > + iov = tcp6_l2_flags_iov[tcp6_flags_used++]; > + vnet_len = sizeof(struct ethhdr) + sizeof(struct ipv6hdr); > } > > + payload = iov[TCP_IOV_PAYLOAD].iov_base; > + th = &payload->th; > + data = payload->opts; > + > if (flags & SYN) { > int mss; > > @@ -1688,8 +1645,11 @@ static int tcp_send_flag(struct ctx *c, struct tcp_tap_conn *conn, int flags) > th->syn = !!(flags & SYN); > th->fin = !!(flags & FIN); > > - iov->iov_len = tcp_l2_buf_fill_headers(c, conn, p, optlen, > - NULL, conn->seq_to_tap); > + ip_len = tcp_l2_buf_fill_headers(c, conn, iov, optlen, NULL, > + conn->seq_to_tap); > + iov[TCP_IOV_PAYLOAD].iov_len = ip_len; > + > + *(uint32_t *)iov[TCP_IOV_TAP].iov_base = htonl(vnet_len + ip_len); > > if (th->ack) { > if (SEQ_GE(conn->seq_ack_to_tap, conn->seq_from_tap)) > @@ -1705,24 +1665,27 @@ static int tcp_send_flag(struct ctx *c, struct tcp_tap_conn *conn, int flags) > if (th->fin || th->syn) > conn->seq_to_tap++; > > - if (CONN_V4(conn)) { > - if (flags & DUP_ACK) { > - memcpy(b4 + 1, b4, sizeof(*b4)); > - (iov + 1)->iov_len = iov->iov_len; > - tcp4_l2_flags_buf_used++; > - } > + if (flags & DUP_ACK) { > + struct iovec *dup_iov; > + int i; > > - if (tcp4_l2_flags_buf_used > ARRAY_SIZE(tcp4_l2_flags_buf) - 2) > - tcp_l2_flags_buf_flush(c); > - } else { > - if (flags & DUP_ACK) { > - memcpy(b6 + 1, b6, sizeof(*b6)); > - (iov + 1)->iov_len = iov->iov_len; > - tcp6_l2_flags_buf_used++; > - } > + if (CONN_V4(conn)) > + dup_iov = tcp4_l2_flags_iov[tcp4_flags_used++]; > + else > + dup_iov = tcp6_l2_flags_iov[tcp6_flags_used++]; > > - if (tcp6_l2_flags_buf_used > ARRAY_SIZE(tcp6_l2_flags_buf) - 2) > - tcp_l2_flags_buf_flush(c); > + for (i = 0; i < TCP_NUM_IOVS; i++) > + memcpy(dup_iov[i].iov_base, iov[i].iov_base, > + iov[i].iov_len); > + dup_iov[TCP_IOV_PAYLOAD].iov_len = iov[TCP_IOV_PAYLOAD].iov_len; > + } > + > + if (CONN_V4(conn)) { > + if (tcp4_flags_used > TCP_FRAMES_MEM - 2) > + tcp_flags_flush(c); > + } else { > + if (tcp6_flags_used > TCP_FRAMES_MEM - 2) > + tcp_flags_flush(c); > } > > return 0; > @@ -2169,30 +2132,43 @@ static void tcp_data_to_tap(const struct ctx *c, struct tcp_tap_conn *conn, > { > uint32_t *seq_update = &conn->seq_to_tap; > struct iovec *iov; > + size_t ip_len; > > if (CONN_V4(conn)) { > - struct tcp4_l2_buf_t *b = &tcp4_l2_buf[tcp4_l2_buf_used]; > - const uint16_t *check = no_csum ? &(b - 1)->iph.check : NULL; > + struct iovec *iov_prev = tcp4_l2_iov[tcp4_payload_used - 1]; > + const uint16_t *check = NULL; > > - tcp4_l2_buf_seq_update[tcp4_l2_buf_used].seq = seq_update; > - tcp4_l2_buf_seq_update[tcp4_l2_buf_used].len = plen; > + if (no_csum) { > + struct iphdr *iph = iov_prev[TCP_IOV_IP].iov_base; > + check = &iph->check; > + } > > - iov = tcp4_l2_iov + tcp4_l2_buf_used++; > - iov->iov_len = tcp_l2_buf_fill_headers(c, conn, b, plen, > - check, seq); > - if (tcp4_l2_buf_used > ARRAY_SIZE(tcp4_l2_buf) - 1) > - tcp_l2_data_buf_flush(c); > + tcp4_seq_update[tcp4_payload_used].seq = seq_update; > + tcp4_seq_update[tcp4_payload_used].len = plen; > + > + iov = tcp4_l2_iov[tcp4_payload_used++]; > + ip_len = tcp_l2_buf_fill_headers(c, conn, iov, plen, check, > + seq); > + iov[TCP_IOV_PAYLOAD].iov_len = ip_len; > + *(uint32_t *)iov[TCP_IOV_TAP].iov_base = > + htonl(sizeof(struct ethhdr) + > + sizeof(struct iphdr) + > + ip_len); A bit less cramped: uint32_t vnet_len; [...] vnet_len = sizeof(struct ethhdr) + sizeof(struct iphdr) + ip_len; *(uint32_t *)iov[TCP_IOV_TAP].iov_base = htonl(vnet_len); > + if (tcp4_payload_used > TCP_FRAMES_MEM - 1) > + tcp_payload_flush(c); > } else if (CONN_V6(conn)) { > - struct tcp6_l2_buf_t *b = &tcp6_l2_buf[tcp6_l2_buf_used]; > - > - tcp6_l2_buf_seq_update[tcp6_l2_buf_used].seq = seq_update; > - tcp6_l2_buf_seq_update[tcp6_l2_buf_used].len = plen; > + tcp6_seq_update[tcp6_payload_used].seq = seq_update; > + tcp6_seq_update[tcp6_payload_used].len = plen; > > - iov = tcp6_l2_iov + tcp6_l2_buf_used++; > - iov->iov_len = tcp_l2_buf_fill_headers(c, conn, b, plen, > - NULL, seq); > - if (tcp6_l2_buf_used > ARRAY_SIZE(tcp6_l2_buf) - 1) > - tcp_l2_data_buf_flush(c); > + iov = tcp6_l2_iov[tcp6_payload_used++]; > + ip_len = tcp_l2_buf_fill_headers(c, conn, iov, plen, NULL, seq); > + iov[TCP_IOV_PAYLOAD].iov_len = ip_len; > + *(uint32_t *)iov[TCP_IOV_TAP].iov_base = > + htonl(sizeof(struct ethhdr) + > + sizeof(struct ipv6hdr) + > + ip_len); > + if (tcp6_payload_used > TCP_FRAMES_MEM - 1) > + tcp_payload_flush(c); > } > } > > @@ -2247,19 +2223,19 @@ static int tcp_data_from_sock(struct ctx *c, struct tcp_tap_conn *conn) > iov_sock[0].iov_base = tcp_buf_discard; > iov_sock[0].iov_len = already_sent; > > - if (( v4 && tcp4_l2_buf_used + fill_bufs > ARRAY_SIZE(tcp4_l2_buf)) || > - (!v4 && tcp6_l2_buf_used + fill_bufs > ARRAY_SIZE(tcp6_l2_buf))) { > - tcp_l2_data_buf_flush(c); > + if (( v4 && tcp4_payload_used + fill_bufs > TCP_FRAMES_MEM) || > + (!v4 && tcp6_payload_used + fill_bufs > TCP_FRAMES_MEM)) { > + tcp_payload_flush(c); > > /* Silence Coverity CWE-125 false positive */ > - tcp4_l2_buf_used = tcp6_l2_buf_used = 0; > + tcp4_payload_used = tcp6_payload_used = 0; > } > > for (i = 0, iov = iov_sock + 1; i < fill_bufs; i++, iov++) { > if (v4) > - iov->iov_base = &tcp4_l2_buf[tcp4_l2_buf_used + i].data; > + iov->iov_base = &tcp4_payload[tcp4_payload_used + i].data; > else > - iov->iov_base = &tcp6_l2_buf[tcp6_l2_buf_used + i].data; > + iov->iov_base = &tcp6_payload[tcp6_payload_used + i].data; > iov->iov_len = mss; > } > if (iov_rem) > @@ -2304,7 +2280,7 @@ static int tcp_data_from_sock(struct ctx *c, struct tcp_tap_conn *conn) > plen = mss; > seq = conn->seq_to_tap; > for (i = 0; i < send_bufs; i++) { > - int no_csum = i && i != send_bufs - 1 && tcp4_l2_buf_used; > + int no_csum = i && i != send_bufs - 1 && tcp4_payload_used; > > if (i == send_bufs - 1) > plen = last_len; > diff --git a/util.h b/util.h > index 48f35604df92..0069df34a5d2 100644 > --- a/util.h > +++ b/util.h > @@ -31,6 +31,9 @@ > #ifndef ETH_MIN_MTU > #define ETH_MIN_MTU 68 > #endif > +#ifndef IP_MAX_MTU > +#define IP_MAX_MTU USHRT_MAX > +#endif > > #ifndef MIN > #define MIN(x, y) (((x) < (y)) ? (x) : (y)) ...everything else looks good to me! -- Stefano