/*
- * Copyright (c) 2008, 2009, 2010, 2011, 2012 Nicira, Inc.
+ * Copyright (c) 2008, 2009, 2010, 2011, 2012, 2013 Nicira, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
#include <config.h>
#include <sys/types.h>
#include "flow.h"
-#include <assert.h>
#include <errno.h>
#include <inttypes.h>
#include <limits.h>
#include "csum.h"
#include "dynamic-string.h"
#include "hash.h"
+#include "jhash.h"
#include "match.h"
#include "ofpbuf.h"
#include "openflow/openflow.h"
return ofpbuf_try_pull(packet, UDP_HEADER_LEN);
}
+static struct sctp_header *
+pull_sctp(struct ofpbuf *packet)
+{
+ return ofpbuf_try_pull(packet, SCTP_HEADER_LEN);
+}
+
static struct icmp_header *
pull_icmp(struct ofpbuf *packet)
{
return ofpbuf_try_pull(packet, sizeof(struct icmp6_hdr));
}
+static void
+parse_mpls(struct ofpbuf *b, struct flow *flow)
+{
+ struct mpls_hdr *mh;
+
+ while ((mh = ofpbuf_try_pull(b, sizeof *mh))) {
+ if (flow->mpls_depth++ == 0) {
+ flow->mpls_lse = mh->mpls_lse;
+ }
+ if (mh->mpls_lse & htonl(MPLS_BOS_MASK)) {
+ break;
+ }
+ }
+}
+
static void
parse_vlan(struct ofpbuf *b, struct flow *flow)
{
}
ofpbuf_pull(b, sizeof *llc);
- return llc->snap.snap_type;
+
+ if (ntohs(llc->snap.snap_type) >= ETH_TYPE_MIN) {
+ return llc->snap.snap_type;
+ }
+
+ return htons(FLOW_DL_TYPE_NONE);
}
static int
parse_ipv6(struct ofpbuf *packet, struct flow *flow)
{
- const struct ip6_hdr *nh;
+ const struct ovs_16aligned_ip6_hdr *nh;
ovs_be32 tc_flow;
int nexthdr;
nexthdr = nh->ip6_nxt;
- flow->ipv6_src = nh->ip6_src;
- flow->ipv6_dst = nh->ip6_dst;
+ memcpy(&flow->ipv6_src, &nh->ip6_src, sizeof flow->ipv6_src);
+ memcpy(&flow->ipv6_dst, &nh->ip6_dst, sizeof flow->ipv6_dst);
- tc_flow = get_unaligned_be32(&nh->ip6_flow);
+ tc_flow = get_16aligned_be32(&nh->ip6_flow);
flow->nw_tos = ntohl(tc_flow) >> 20;
flow->ipv6_label = tc_flow & htonl(IPV6_LABEL_MASK);
flow->nw_ttl = nh->ip6_hlim;
return EINVAL;
}
} else if (nexthdr == IPPROTO_FRAGMENT) {
- const struct ip6_frag *frag_hdr = packet->data;
+ const struct ovs_16aligned_ip6_frag *frag_hdr = packet->data;
nexthdr = frag_hdr->ip6f_nxt;
if (!ofpbuf_try_pull(packet, sizeof *frag_hdr)) {
/* We only process the first fragment. */
if (frag_hdr->ip6f_offlg != htons(0)) {
- if ((frag_hdr->ip6f_offlg & IP6F_OFF_MASK) == htons(0)) {
- flow->nw_frag = FLOW_NW_FRAG_ANY;
- } else {
+ flow->nw_frag = FLOW_NW_FRAG_ANY;
+ if ((frag_hdr->ip6f_offlg & IP6F_OFF_MASK) != htons(0)) {
flow->nw_frag |= FLOW_NW_FRAG_LATER;
nexthdr = IPPROTO_FRAGMENT;
break;
}
}
+static void
+parse_sctp(struct ofpbuf *packet, struct ofpbuf *b, struct flow *flow)
+{
+ const struct sctp_header *sctp = pull_sctp(b);
+ if (sctp) {
+ flow->tp_src = sctp->sctp_src;
+ flow->tp_dst = sctp->sctp_dst;
+ packet->l7 = b->data;
+ }
+}
+
static bool
parse_icmpv6(struct ofpbuf *b, struct flow *flow)
{
}
/* Initializes 'flow' members from 'packet', 'skb_priority', 'tnl', and
- * 'ofp_in_port'.
+ * 'in_port'.
*
* Initializes 'packet' header pointers as follows:
*
* - packet->l2 to the start of the Ethernet header.
*
+ * - packet->l2_5 to the start of the MPLS shim header.
+ *
* - packet->l3 to just past the Ethernet header, or just past the
* vlan_header if one is present, to the first byte of the payload of the
* Ethernet frame.
* - packet->l4 to just past the IPv4 header, if one is present and has a
* correct length, and otherwise NULL.
*
- * - packet->l7 to just past the TCP or UDP or ICMP header, if one is
+ * - packet->l7 to just past the TCP/UDP/SCTP/ICMP header, if one is
* present and has a correct length, and otherwise NULL.
*/
void
-flow_extract(struct ofpbuf *packet, uint32_t skb_priority,
- const struct flow_tnl *tnl, uint16_t ofp_in_port,
+flow_extract(struct ofpbuf *packet, uint32_t skb_priority, uint32_t pkt_mark,
+ const struct flow_tnl *tnl, const union flow_in_port *in_port,
struct flow *flow)
{
struct ofpbuf b = *packet;
memset(flow, 0, sizeof *flow);
if (tnl) {
- assert(tnl != &flow->tunnel);
+ ovs_assert(tnl != &flow->tunnel);
flow->tunnel = *tnl;
}
- flow->in_port = ofp_in_port;
+ if (in_port) {
+ flow->in_port = *in_port;
+ }
flow->skb_priority = skb_priority;
+ flow->pkt_mark = pkt_mark;
- packet->l2 = b.data;
- packet->l3 = NULL;
- packet->l4 = NULL;
- packet->l7 = NULL;
+ packet->l2 = b.data;
+ packet->l2_5 = NULL;
+ packet->l3 = NULL;
+ packet->l4 = NULL;
+ packet->l7 = NULL;
if (b.size < sizeof *eth) {
return;
}
flow->dl_type = parse_ethertype(&b);
+ /* Parse mpls, copy l3 ttl. */
+ if (eth_type_mpls(flow->dl_type)) {
+ packet->l2_5 = b.data;
+ parse_mpls(&b, flow);
+ }
+
/* Network layer. */
packet->l3 = b.data;
if (flow->dl_type == htons(ETH_TYPE_IP)) {
if (nh) {
packet->l4 = b.data;
- flow->nw_src = get_unaligned_be32(&nh->ip_src);
- flow->nw_dst = get_unaligned_be32(&nh->ip_dst);
+ flow->nw_src = get_16aligned_be32(&nh->ip_src);
+ flow->nw_dst = get_16aligned_be32(&nh->ip_dst);
flow->nw_proto = nh->ip_proto;
flow->nw_tos = nh->ip_tos;
parse_tcp(packet, &b, flow);
} else if (flow->nw_proto == IPPROTO_UDP) {
parse_udp(packet, &b, flow);
+ } else if (flow->nw_proto == IPPROTO_SCTP) {
+ parse_sctp(packet, &b, flow);
} else if (flow->nw_proto == IPPROTO_ICMP) {
const struct icmp_header *icmp = pull_icmp(&b);
if (icmp) {
parse_tcp(packet, &b, flow);
} else if (flow->nw_proto == IPPROTO_UDP) {
parse_udp(packet, &b, flow);
+ } else if (flow->nw_proto == IPPROTO_SCTP) {
+ parse_sctp(packet, &b, flow);
} else if (flow->nw_proto == IPPROTO_ICMPV6) {
if (parse_icmpv6(&b, flow)) {
packet->l7 = b.data;
}
}
- } else if (flow->dl_type == htons(ETH_TYPE_ARP)) {
+ } else if (flow->dl_type == htons(ETH_TYPE_ARP) ||
+ flow->dl_type == htons(ETH_TYPE_RARP)) {
const struct arp_eth_header *arp = pull_arp(&b);
if (arp && arp->ar_hrd == htons(1)
&& arp->ar_pro == htons(ETH_TYPE_IP)
flow->nw_proto = ntohs(arp->ar_op);
}
- flow->nw_src = arp->ar_spa;
- flow->nw_dst = arp->ar_tpa;
+ flow->nw_src = get_16aligned_be32(&arp->ar_spa);
+ flow->nw_dst = get_16aligned_be32(&arp->ar_tpa);
memcpy(flow->arp_sha, arp->ar_sha, ETH_ADDR_LEN);
memcpy(flow->arp_tha, arp->ar_tha, ETH_ADDR_LEN);
}
void
flow_get_metadata(const struct flow *flow, struct flow_metadata *fmd)
{
- BUILD_ASSERT_DECL(FLOW_WC_SEQ == 17);
+ BUILD_ASSERT_DECL(FLOW_WC_SEQ == 20);
fmd->tun_id = flow->tunnel.tun_id;
+ fmd->tun_src = flow->tunnel.ip_src;
+ fmd->tun_dst = flow->tunnel.ip_dst;
fmd->metadata = flow->metadata;
memcpy(fmd->regs, flow->regs, sizeof fmd->regs);
- fmd->in_port = flow->in_port;
+ fmd->pkt_mark = flow->pkt_mark;
+ fmd->in_port = flow->in_port.ofp_port;
}
char *
return ds_cstr(&ds);
}
+const char *
+flow_tun_flag_to_string(uint32_t flags)
+{
+ switch (flags) {
+ case FLOW_TNL_F_DONT_FRAGMENT:
+ return "df";
+ case FLOW_TNL_F_CSUM:
+ return "csum";
+ case FLOW_TNL_F_KEY:
+ return "key";
+ default:
+ return NULL;
+ }
+}
+
+void
+format_flags(struct ds *ds, const char *(*bit_to_string)(uint32_t),
+ uint32_t flags, char del)
+{
+ uint32_t bad = 0;
+
+ if (!flags) {
+ return;
+ }
+ while (flags) {
+ uint32_t bit = rightmost_1bit(flags);
+ const char *s;
+
+ s = bit_to_string(bit);
+ if (s) {
+ ds_put_format(ds, "%s%c", s, del);
+ } else {
+ bad |= bit;
+ }
+
+ flags &= ~bit;
+ }
+
+ if (bad) {
+ ds_put_format(ds, "0x%"PRIx32"%c", bad, del);
+ }
+ ds_chomp(ds, del);
+}
+
void
flow_format(struct ds *ds, const struct flow *flow)
{
struct match match;
match_wc_init(&match, flow);
- match_format(&match, ds, flow->skb_priority);
+ match_format(&match, ds, OFP_DEFAULT_PRIORITY);
}
void
return true;
}
-/* Initializes 'dst' as the combination of wildcards in 'src1' and 'src2'.
- * That is, a bit or a field is wildcarded in 'dst' if it is wildcarded in
- * 'src1' or 'src2' or both. */
+/* Sets 'dst' as the bitwise AND of wildcards in 'src1' and 'src2'.
+ * That is, a bit or a field is wildcarded in 'dst' if it is wildcarded
+ * in 'src1' or 'src2' or both. */
void
-flow_wildcards_combine(struct flow_wildcards *dst,
- const struct flow_wildcards *src1,
- const struct flow_wildcards *src2)
+flow_wildcards_and(struct flow_wildcards *dst,
+ const struct flow_wildcards *src1,
+ const struct flow_wildcards *src2)
{
uint32_t *dst_u32 = (uint32_t *) &dst->masks;
const uint32_t *src1_u32 = (const uint32_t *) &src1->masks;
}
}
+/* Sets 'dst' as the bitwise OR of wildcards in 'src1' and 'src2'. That
+ * is, a bit or a field is wildcarded in 'dst' if it is neither
+ * wildcarded in 'src1' nor 'src2'. */
+void
+flow_wildcards_or(struct flow_wildcards *dst,
+ const struct flow_wildcards *src1,
+ const struct flow_wildcards *src2)
+{
+ uint32_t *dst_u32 = (uint32_t *) &dst->masks;
+ const uint32_t *src1_u32 = (const uint32_t *) &src1->masks;
+ const uint32_t *src2_u32 = (const uint32_t *) &src2->masks;
+ size_t i;
+
+ for (i = 0; i < FLOW_U32S; i++) {
+ dst_u32[i] = src1_u32[i] | src2_u32[i];
+ }
+}
+
+/* Perform a bitwise OR of miniflow 'src' flow data with the equivalent
+ * fields in 'dst', storing the result in 'dst'. */
+static void
+flow_union_with_miniflow(struct flow *dst, const struct miniflow *src)
+{
+ uint32_t *dst_u32 = (uint32_t *) dst;
+ int ofs;
+ int i;
+
+ ofs = 0;
+ for (i = 0; i < MINI_N_MAPS; i++) {
+ uint32_t map;
+
+ for (map = src->map[i]; map; map = zero_rightmost_1bit(map)) {
+ dst_u32[raw_ctz(map) + i * 32] |= src->values[ofs++];
+ }
+ }
+}
+
+/* Fold minimask 'mask''s wildcard mask into 'wc's wildcard mask. */
+void
+flow_wildcards_fold_minimask(struct flow_wildcards *wc,
+ const struct minimask *mask)
+{
+ flow_union_with_miniflow(&wc->masks, &mask->masks);
+}
+
/* Returns a hash of the wildcards in 'wc'. */
uint32_t
flow_wildcards_hash(const struct flow_wildcards *wc, uint32_t basis)
{
- return flow_hash(&wc->masks, basis);;
+ return flow_hash(&wc->masks, basis);
}
/* Returns true if 'a' and 'b' represent the same wildcards, false if they are
if (fields.eth_type == htons(ETH_TYPE_IP)) {
fields.ipv4_addr = flow->nw_src ^ flow->nw_dst;
fields.ip_proto = flow->nw_proto;
- if (fields.ip_proto == IPPROTO_TCP) {
+ if (fields.ip_proto == IPPROTO_TCP || fields.ip_proto == IPPROTO_SCTP) {
fields.tp_port = flow->tp_src ^ flow->tp_dst;
}
} else if (fields.eth_type == htons(ETH_TYPE_IPV6)) {
ipv6_addr[i] = a[i] ^ b[i];
}
fields.ip_proto = flow->nw_proto;
- if (fields.ip_proto == IPPROTO_TCP) {
+ if (fields.ip_proto == IPPROTO_TCP || fields.ip_proto == IPPROTO_SCTP) {
fields.tp_port = flow->tp_src ^ flow->tp_dst;
}
}
- return hash_bytes(&fields, sizeof fields, basis);
+ return jhash_bytes(&fields, sizeof fields, basis);
+}
+
+/* Masks the fields in 'wc' that are used by the flow hash 'fields'. */
+void
+flow_mask_hash_fields(const struct flow *flow, struct flow_wildcards *wc,
+ enum nx_hash_fields fields)
+{
+ switch (fields) {
+ case NX_HASH_FIELDS_ETH_SRC:
+ memset(&wc->masks.dl_src, 0xff, sizeof wc->masks.dl_src);
+ break;
+
+ case NX_HASH_FIELDS_SYMMETRIC_L4:
+ memset(&wc->masks.dl_src, 0xff, sizeof wc->masks.dl_src);
+ memset(&wc->masks.dl_dst, 0xff, sizeof wc->masks.dl_dst);
+ if (flow->dl_type == htons(ETH_TYPE_IP)) {
+ memset(&wc->masks.nw_src, 0xff, sizeof wc->masks.nw_src);
+ memset(&wc->masks.nw_dst, 0xff, sizeof wc->masks.nw_dst);
+ } else if (flow->dl_type == htons(ETH_TYPE_IPV6)) {
+ memset(&wc->masks.ipv6_src, 0xff, sizeof wc->masks.ipv6_src);
+ memset(&wc->masks.ipv6_dst, 0xff, sizeof wc->masks.ipv6_dst);
+ }
+ if (is_ip_any(flow)) {
+ memset(&wc->masks.nw_proto, 0xff, sizeof wc->masks.nw_proto);
+ memset(&wc->masks.tp_src, 0xff, sizeof wc->masks.tp_src);
+ memset(&wc->masks.tp_dst, 0xff, sizeof wc->masks.tp_dst);
+ }
+ wc->masks.vlan_tci |= htons(VLAN_VID_MASK | VLAN_CFI);
+ break;
+
+ default:
+ NOT_REACHED();
+ }
}
/* Hashes the portions of 'flow' designated by 'fields'. */
switch (fields) {
case NX_HASH_FIELDS_ETH_SRC:
- return hash_bytes(flow->dl_src, sizeof flow->dl_src, basis);
+ return jhash_bytes(flow->dl_src, sizeof flow->dl_src, basis);
case NX_HASH_FIELDS_SYMMETRIC_L4:
return flow_hash_symmetric_l4(flow, basis);
|| fields == NX_HASH_FIELDS_SYMMETRIC_L4;
}
+/* Returns a hash value for the bits of 'flow' that are active based on
+ * 'wc', given 'basis'. */
+uint32_t
+flow_hash_in_wildcards(const struct flow *flow,
+ const struct flow_wildcards *wc, uint32_t basis)
+{
+ const uint32_t *wc_u32 = (const uint32_t *) &wc->masks;
+ const uint32_t *flow_u32 = (const uint32_t *) flow;
+ uint32_t hash;
+ size_t i;
+
+ hash = basis;
+ for (i = 0; i < FLOW_U32S; i++) {
+ hash = mhash_add(hash, flow_u32[i] & wc_u32[i]);
+ }
+ return mhash_finish(hash, 4 * FLOW_U32S);
+}
+
/* Sets the VLAN VID that 'flow' matches to 'vid', which is interpreted as an
* OpenFlow 1.0 "dl_vlan" value:
*
flow->vlan_tci |= htons((pcp << VLAN_PCP_SHIFT) | VLAN_CFI);
}
+/* Sets the MPLS Label that 'flow' matches to 'label', which is interpreted
+ * as an OpenFlow 1.1 "mpls_label" value. */
+void
+flow_set_mpls_label(struct flow *flow, ovs_be32 label)
+{
+ set_mpls_lse_label(&flow->mpls_lse, label);
+}
+
+/* Sets the MPLS TTL that 'flow' matches to 'ttl', which should be in the
+ * range 0...255. */
+void
+flow_set_mpls_ttl(struct flow *flow, uint8_t ttl)
+{
+ set_mpls_lse_ttl(&flow->mpls_lse, ttl);
+}
+
+/* Sets the MPLS TC that 'flow' matches to 'tc', which should be in the
+ * range 0...7. */
+void
+flow_set_mpls_tc(struct flow *flow, uint8_t tc)
+{
+ set_mpls_lse_tc(&flow->mpls_lse, tc);
+}
+
+/* Sets the MPLS BOS bit that 'flow' matches to which should be 0 or 1. */
+void
+flow_set_mpls_bos(struct flow *flow, uint8_t bos)
+{
+ set_mpls_lse_bos(&flow->mpls_lse, bos);
+}
+
/* Puts into 'b' a packet that flow_extract() would parse as having the given
* 'flow'.
*
ip->ip_tos = flow->nw_tos;
ip->ip_ttl = flow->nw_ttl;
ip->ip_proto = flow->nw_proto;
- ip->ip_src = flow->nw_src;
- ip->ip_dst = flow->nw_dst;
+ put_16aligned_be32(&ip->ip_src, flow->nw_src);
+ put_16aligned_be32(&ip->ip_dst, flow->nw_dst);
if (flow->nw_frag & FLOW_NW_FRAG_ANY) {
ip->ip_frag_off |= htons(IP_MORE_FRAGMENTS);
b->l4 = udp = ofpbuf_put_zeros(b, sizeof *udp);
udp->udp_src = flow->tp_src;
udp->udp_dst = flow->tp_dst;
+ } else if (flow->nw_proto == IPPROTO_SCTP) {
+ struct sctp_header *sctp;
+
+ b->l4 = sctp = ofpbuf_put_zeros(b, sizeof *sctp);
+ sctp->sctp_src = flow->tp_src;
+ sctp->sctp_dst = flow->tp_dst;
} else if (flow->nw_proto == IPPROTO_ICMP) {
struct icmp_header *icmp;
ip->ip_csum = csum(ip, sizeof *ip);
} else if (flow->dl_type == htons(ETH_TYPE_IPV6)) {
/* XXX */
- } else if (flow->dl_type == htons(ETH_TYPE_ARP)) {
+ } else if (flow->dl_type == htons(ETH_TYPE_ARP) ||
+ flow->dl_type == htons(ETH_TYPE_RARP)) {
struct arp_eth_header *arp;
b->l3 = arp = ofpbuf_put_zeros(b, sizeof *arp);
if (flow->nw_proto == ARP_OP_REQUEST ||
flow->nw_proto == ARP_OP_REPLY) {
- arp->ar_spa = flow->nw_src;
- arp->ar_tpa = flow->nw_dst;
+ put_16aligned_be32(&arp->ar_spa, flow->nw_src);
+ put_16aligned_be32(&arp->ar_tpa, flow->nw_dst);
memcpy(arp->ar_sha, flow->arp_sha, ETH_ADDR_LEN);
memcpy(arp->ar_tha, flow->arp_tha, ETH_ADDR_LEN);
}
}
+
+ if (eth_type_mpls(flow->dl_type)) {
+ b->l2_5 = b->l3;
+ push_mpls(b, flow->dl_type, flow->mpls_lse);
+ }
}
\f
/* Compressed flow. */
memcpy(dst->values, src->values, n * sizeof *dst->values);
}
+/* Initializes 'dst' with the data in 'src', destroying 'src'.
+ * The caller must eventually free 'dst' with miniflow_destroy(). */
+void
+miniflow_move(struct miniflow *dst, struct miniflow *src)
+{
+ int n = miniflow_n_values(src);
+ if (n <= MINI_N_INLINE) {
+ dst->values = dst->inline_values;
+ memcpy(dst->values, src->values, n * sizeof *dst->values);
+ } else {
+ dst->values = src->values;
+ }
+ memcpy(dst->map, src->map, sizeof dst->map);
+}
+
/* Frees any memory owned by 'flow'. Does not free the storage in which 'flow'
* itself resides; the caller is responsible for that. */
void
void
miniflow_expand(const struct miniflow *src, struct flow *dst)
{
- uint32_t *dst_u32 = (uint32_t *) dst;
- int ofs;
- int i;
-
- memset(dst_u32, 0, sizeof *dst);
-
- ofs = 0;
- for (i = 0; i < MINI_N_MAPS; i++) {
- uint32_t map;
-
- for (map = src->map[i]; map; map = zero_rightmost_1bit(map)) {
- dst_u32[raw_ctz(map) + i * 32] = src->values[ofs++];
- }
- }
+ memset(dst, 0, sizeof *dst);
+ flow_union_with_miniflow(dst, src);
}
static const uint32_t *
}
}
- return mhash_finish(hash, p - mask->masks.values);
+ return mhash_finish(hash, (p - mask->masks.values) * 4);
}
/* Returns a hash value for the bits of 'flow' where there are 1-bits in
}
}
- return mhash_finish(hash, p - mask->masks.values);
+ return mhash_finish(hash, (p - mask->masks.values) * 4);
}
\f
/* Initializes 'dst' as a copy of 'src'. The caller must eventually free 'dst'
miniflow_clone(&dst->masks, &src->masks);
}
+/* Initializes 'dst' with the data in 'src', destroying 'src'.
+ * The caller must eventually free 'dst' with minimask_destroy(). */
+void
+minimask_move(struct minimask *dst, struct minimask *src)
+{
+ miniflow_clone(&dst->masks, &src->masks);
+}
+
/* Initializes 'dst_' as the bit-wise "and" of 'a_' and 'b_'.
*
* The caller must provide room for FLOW_U32S "uint32_t"s in 'storage', for use